|Home | About | Journals | Submit | Contact Us | Français|
Recent studies have shown that bone morphogenetic proteins (BMPs) are important regulators in the pituitary–gonadal endocrine axis. We here investigated the effects of BMPs on GNRH production controlled by estrogen using murine GT1-7 hypothalamic neuron cells. GT1-7 cells expressed estrogen receptor α (ERα; ESR1 as listed in MGI Database), ERβ (ESR2 as listed in MGI Database), BMP receptors, SMADs, and a binding protein follistatin. Treatment with BMP2 and BMP4 had no effect on Gnrh mRNA expression; however, BMP6 and BMP7 significantly increased Gnrh mRNA expression as well as GnRH production by GT1-7 cells. Notably, the reduction of Gnrh expression caused by estradiol (E2) was restored by cotreatment with BMP2 and BMP4, whereas it was not affected by BMP6 or BMP7. E2 activated extracellular signal-regulated kinase (ERK) 1/2 and stress-activated protein kinase/c-Jun NH2-terminal kinase (SAPK/JNK) signaling but did not activate p38-mitogen-activated protein kinase (MAPK) signaling in GT1-7 cells. Inhibition of ERK1/ERK2 reversed the inhibitory effect of estrogen on Gnrh expression, whereas SAPK/JNK inhibition did not affect the E2 actions. Expression levels of Erα and Erβ were reduced by BMP2 and BMP4, but were increased by BMP6 and BMP7. Treatment with an ER antagonist inhibited the E2 effects on Gnrh suppression including reduction of E2-induced ERK phosphorylation, suggesting the involvement of genomic ER actions in Gnrh suppression. BMP2 and BMP4 also suppressed estrogen-induced phosphorylation of ERK1/ERK2 and SAPK/JNK signaling, suggesting that BMP2 and BMP4 downregulate estrogen effects by attenuating ER–MAPK signaling. Considering that BMP6 and BMP7 increased the expression of α1E-subunit of R-type calcium channel (Cacna1e), which is critical for GNRH secretion, it is possible that BMP6 and BMP7 directly stimulate GNRH release by GT1-7 cells. Collectively, a newly uncovered interaction of BMPs and ER may be involved in controlling hypothalamic GNRH production and secretion via an autocrine/paracrine mechanism.
Reproduction in mammals is controlled by the coordination between episodic release of GNRH and GNRH-dependent secretion of pituitary gonadotropins. GNRH is a highly conserved decapeptide, which is secreted by a scattered group of cells located in the hypothalamus. The pattern of GNRH secretion is regulated by the intrinsic oscillatory activity of GNRH neurons and the integration of presynaptic inputs of various neurotransmitters. GNRH synthesis and release are tightly regulated by gonadal steroids, which maintain control through a negative feedback system (Petersen et al. 2003). In females, estrogens act directly or indirectly on GNRH neuronal network to modulate the final output of GNRH into the median eminence (Herbison 1998, Herbison & Pape 2001). Estrogen receptors (ER) belong to the superfamily of nuclear receptors and are classified as class I ligand-activated transcription factors. There are two identified subtypes of ERα and ERβ (ESR1 and ESR2 as listed in MGI Database), which are encoded by separate genes and differ in structure, function, and anatomical distribution.
The characterization of specific estrogen effects and identification of ER-dependent signaling pathways have been difficult in GNRH neurons. In this respect, immortalized GNRH-producing GT1-7 cells have proven to be a valuable tool to study the biology of GNRH neurons (Liposits et al. 1991, Wetsel et al. 1991). Although earlier studies indicated that GNRH neurons generally lack steroid hormone receptors, recent studies have provided evidence that GT1-7 cells express functional ER, suggesting that estrogen regulates GNRH production and secretion directly in GNRH neuron in the reproductive system composed of hypothalamo-pituitary–ovarian (HPO) axis.
Bone morphogenetic proteins (BMPs), which belong to the transforming growth factor-β (TGF-β or TGFB1 as listed in MGI Database) superfamily, were originally identified as the active components in bone extracts capable of inducing bone formation at ectopic sites. Recent studies have shown that BMPs are crucial molecules in normal folliculogenesis by regulating gonadotropin-induced steroidogenesis and mitosis in ovarian granulosa cells (Otsuka et al. 2000, Otsuka & Shimasaki 2002a, Shimasaki et al. 2004). The major regulatory process by BMPs in ovarian steroidogenesis is the control of FSH receptor signaling (Otsuka et al. 2001a,b), leading to normal follicular development in the ovary with prevention of immature ovulation. Furthermore, the pituitary BMP system acts as a regulator not only for pituitary differentiation but also for the transformation of differentiated pituitary cells (Otsuka & Shimasaki 2002b, Takeda et al. 2003, 2007, Miyoshi et al. 2008). It is of note that BMPs activate FSH production by the pituitary gonadotrope (Huang et al. 2001, Otsuka & Shimasaki 2002b, Nicol et al. 2008). Functional communication between the ovary and central nervous system ensures that the neural signal for ovulation occurs when ovarian follicles are fully matured. Hence, BMPs play crucial roles in female reproduction not only by regulating ovarian steroidogenesis and mitogenesis but also by activating pituitary gonadotropin secretion in an autocrine/paracrine manner.
However, the effects of BMPs on hypothalamic GNRH production and secretion have yet to be elucidated. In the present study, we here investigated the effects of BMPs on GNRH production controlled by estrogen using murine GT1-7 hypothalamic neuron cells. A novel interaction between BMPs and ER, which is involved in controlling hypothalamic GNRH production and secretion, was here uncovered.
DMEM, dimethyl sulfoxide, penicillin–streptomycin solution, 17β-estradiol (E2), BSA-conjugated E2 (E2-BSA), ICI 182 780 (also called fulvestrant), and recombinant human activin A were purchased from Sigma–Aldrich Co. Ltd. Recombinant human BMP-2, -4, -6, and -7 were purchased from R&D Systems Inc. (Minneapolis, MN, USA), recombinant human platelet-derived growth factor (PDGF)-BB was purchased from PeproTech (London, UK), and U0126 and SB203580 were purchased from Promega Corp. SP600125 was purchased from Biomol Lab. Inc (Plymouth Meeting, PA, USA). Mouse ovary total RNA was purchased from Ambion (Austin, TX, USA). For the experiments using E2-BSA, the preparation of E2-BSA free of estradiol was performed as previously reported (Taguchi et al. 2004, Takahashi et al. 2008): 4 ml E2-BSA (1·25 mM in estrogen dissolved in 50 mM Tris–HCl (pH 8·5)) was added to a Amicon Ultra centrifugal filter unit with a MW cut off of 5000 (Millipore, Bedford, MA, USA) and centrifuged at 4000 g until 50 ml retentate remained. The retentate was washed with Tris buffer, and final volume was adjusted to 5 ml (1 mM).
GT1-7 cells were kindly provided by Dr Pamela L Mellon, University of California, San Diego, CA, USA. GT1-7 cells were maintained in DMEM supplemented with 10% FCS, penicillin, and streptomycin (Sigma–Aldrich Corp.) at 37 °C in 5% CO2 humidified atmosphere. The culture medium was changed twice a week, and cultures were passaged at ~80% confluence. Changes in cell morphology and growing conditions were carefully monitored under an inverted microscope.
GT1-7 cells (2×105 viable cells/ml) were precultured in serum-free DMEM, and cells were treated with indicated concentrations of BMPs in combination with E2 and various chemical inhibitors including U0126, SB203580, SP600125, and ICI 182 780. After 24-h culture, the medium was removed, and total cellular RNAs were extracted using TRIzol (Invitrogen Corp.), subsequently quantified by measuring absorbance at 260 nm and stored at –80 °C until assay. The expression of BMP receptors (Bmprs), Smads, Ers, and Gnrh was detected by RT-PCR analysis. The extracted RNA (1 μg) was subjected to a RT reaction using First-Strand cDNA Synthesis System (Invitrogen Corp.) with random hexamer (2 ng/μl), reverse transcriptase (200 U), and dNTP (0·5 mM) at 42 °C for 50 min, 70 °C for 10 min. Subsequently, hot-start PCR was performed using MgCl2 (1·5 mM), dNTP (0·2 mM), and Taq DNA polymerase (2·5 U) (Invitrogen Corp). Oligonucleotides used for PCR were custom ordered from Invitrogen Corp. PCR primer pairs were selected from different exons of the corresponding genes to discriminate PCR products that might arise from possible chromosome DNA contaminants. The primer pairs for mouse BMP receptors, Smads, and a housekeeping gene, ribosomal protein-L19 (Rpl19), were selected as reported (Kano et al. 2005, Takeda et al. 2007, Miyoshi et al. 2008). For Ers, Gnrh, and calcium channel α1E-subunit (Cacna1e) genes, the following sequences were used: Erα, 1523–1543 and 1737–1757 from GenBank accession number NM_007956; Erβ, 701–721 and 1004–1024 from NM_010157; Gnrh, 90–110 and 316–336 from BC116897; and calcium channel α1E, 3181–3201 and 3403–3423 from NM_009782. The aliquots of PCR products were electrophoresed on 1·5% agarose gels and visualized after ethidium bromide staining. For the quantification of Gnrh, Erα, Erβ, Cacna1e, and Rpl19 mRNA levels, real-time PCR was performed using LightCycler FastStart DNA master SYBR Green I system (Roche Diagnostic Co.) under each optimized condition of annealing at 59–63 °C with 4 mM MgCl2, following the manufacturer's protocol. Accumulated levels of fluorescence for each product were analyzed by the second derivative method after the melting curve analysis (Roche Diagnostic Co.), and then, following the assay validation by calculating each amplification efficiency (Rpl19, 98·5%; Gnrh, 89%; Erα, 88%; Erβ, 92%; and calcium channel α1E, 105%), the expression levels of target genes were quantified based on standard curve analysis for each product and normalized by Rpl19 level in each target.
To assess the effects of BMPs on GNRH protein synthesis, GT1-7 cells (2×105 viable cells/ml) were cultured in 96-well plates with DMEM containing 10% FCS for 24 h. The medium was then changed to serum-free DMEM and subsequently treated with the indicated concentrations of BMPs and E2. After 24-h culture, the supernatant of the culture media was collected and stored at –80 °C until assay. GNRH concentration (pg/ml) in the conditioned medium was determined by RIA as previously reported (Chappell et al. 2003). Briefly, 0·1 ml aliquots of each sample were incubated for 48 h with GNRH antibody EL-14 at 4 °C, after which time 10 000 c.p.m. sample of radio-iodinated GNRH (Amersham Pharmacia) was added. Following ethanol precipitation of the bound fractions after 48-h incubation, radioactivity was detected by a gamma scintillation counter (Micromedic, Huntsville, AL, USA), and the standard curve analysis was performed at each assay. The inter- and intra-assay variability are 8·7 and 7·7% respectively in the current RIA.
GT1-7 cells (1×105 viable cells/ml) were precultured in 12-well plates with DMEM containing 10% FCS for 24 h. After preculture medium was replaced with fresh serum-free medium and indicated concentrations of E2, BMPs and PDGF-BB were added to the culture medium. After 24-h culture, 0·5 μCi/ml [methyl-3H] thymidine (Amersham Pharmacia) was added and incubated for 3 h at 37 °C. The incorporated thymidine was detected as previously reported (Takahashi et al. 2008). Cells were then washed with PBS, incubated with 10% ice-cold trichloroacetic acid for 60 min at 4 °C, and solubilized in 0·5 M NaOH, and radioactivity was determined with a liquid scintillation counter (TRI-CARB 2300TR, Packard Co., Meriden, CT, USA).
GT1-7 cells (1×105 viable cells/ml) were precultured in 12-well plates in serum-free DMEM. After 24-h preculture, BMPs and activin A (100 ng/ml) were added to the culture medium either alone or in combination with E2 (100 nM). After acute (10–60 min) or chronic (24 h) stimulations with indicated concentrations of BMPs in combination with E2 and various chemical inhibitors including U0126, SB203580, SP600125, and ICI 182 780, cells were solubilized in 100 μl RIPA lysis buffer (Upstate Biotechnology, Lake Placid, NY, USA) containing 1 mM Na3VO4, 1 mM NaF, 2% SDS, and 4% β-mercaptoethanol. The cell lysates were then subjected to SDS-PAGE/immunoblotting analysis as previously reported (Inagaki et al. 2006, Otani et al. 2007), using anti-phospho- and anti-total-extracellular signal-regulated kinase (ERK) 1/2 mitogen-activated protein kinase (MAPK) antibody (Cell Signaling Technology, Inc., Beverly, MA, USA), anti-phospho- and anti-total-p38 MAPK antibody (Cell Signaling Technology, Inc.), anti-phospho- and anti-total-stress-activated protein kinase/c-Jun NH2-terminal kinase (SAPK/JNK) MAPK antibody (Cell Signaling Technology, Inc.), anti-phospho-Smad1, 5, 8 (pSmad1, 5, 8) antibody (Cell Signaling Technology, Inc.), anti-Smad5 antibody (Cell Signaling Technology, Inc.), anti-actin antibody (Sigma–Aldrich Co. Ltd), and anti-ERα (MC-20), and anti-ERβ (H-150) antibodies (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA).
For the immunofluorescence study, GT1-7 cells were precultured in serum-free DMEM using chamber slides (Nalge Nunc Int, Naperville, IL, USA). Cells at ~50% confluency were treated with BMP-2 and BMP-4 (100 ng/ml) for 1 h. Cells were then fixed with 4% paraformaldehyde in PBS, permeabilized with 0·5% Triton X-100 in PBS at room temperature, and washed thrice with PBS. The cells were then incubated with anti-phospho-SMAD1, 5, 8 antibody (Cell signaling Technology, Inc.) for 1 h and washed thrice with PBS, then with Alexa Fluor 488 anti-rabbit IgG (Invitrogen Corp.) in humidified chamber for 1 h and washed with PBS, followed by application of the counter medium containing 4′,6′-diaminodino-2-phenylindole (Invitorogen Corp.), and then stained cells were visualized under a fluorescent microscope.
GT1-7 cells (1×105 viable cells/ml) were precultured in 12-well plates in DMEM with 10% FCS for 24 h. The cells were transiently transfected with 500 ng Gnrh-luc reporter plasmid, which contains 173-bp rat Gnrh promoter region (Nelson et al. 2000), and 50 ng cytomegalovirus-β-galactosidase plasmid (pCMV-β-gal) using FuGENE6 (Roche Molecular Biochemicals) for 24 h. The cells were then treated with indicated concentrations of BMPs and E2 for 24 h in serum-free fresh medium. After 24-h culture, the cells were washed with PBS and lysed with Cell Culture Lysis Reagent (Toyobo, Osaka, Japan). Luciferase activity and β-gal activity of the cell lysate were measured by luminescencer-PSN (ATTO, Tokyo, Japan) as previously reported (Miyoshi et al. 2006). The data were shown as the ratio of luciferase to β-gal activity.
All results are shown as mean±s.e.m. of data from at least three separate experiments, each performed with triplicate samples. Differences between groups were analyzed for statistical significance using ANOVA with Fisher's protected least significant difference (PLSD) test (StatView 5.0 software, Abacus Concepts, Inc., Berkeley, CA, USA). P values <0·05 were accepted as statistically significant.
We first examined mRNA expression of the BMP type I and type II receptors in GT1-7 cells by RT-PCR. As shown in Fig. 1A, the BMP type I receptors including activin receptor-like kinase (Alk2) (also called Actr1a), Alk3 (Bmpr1a), and Alk4 (Actr1b), BMP type II receptors including Bmpr2, Actr2, and Actr2b, and a binding protein Fst were clearly expressed in GT1-7 cells. Compared with mouse ovary RNA positive control, the expression of Alk6 (Bmpr1b) was not detected in GT1-7 cells. In addition, mRNA expression of Smad signaling molecules including Smad1, 2, 3, 4, 5, and 8, and inhibitory Smad6 and 7 was also detected (Fig. 1A). The expression of two subtypes of ERs including Erα and Erβ, and Gnrh mRNA was also confirmed in GT1-7 cells. As shown in Fig. 1B, the immunofluorescence study demonstrated the nuclear localization of phosphorylated SMAD1, 5, 8 molecules in GT1-7 cells stimulated by BMP-2 and -4 (100 ng/ml), indicating that the BMP receptors and SMADs are functionally active in GT1-7 cells.
To investigate the BMP effects on Gnrh transcription, Gnrh-promoter activity was examined in GT1-7 cells transiently transfected with a Gnrh-luc reporter plasmid containing 173-bp rat Gnrh promoter region (Nelson et al. 2000). In GT1-7 hypothalamic neurons, a neuron-specific enhancer responsible for directing expression of the rat Gnrh gene was identified (Whyte et al. 1995), in which ~300-bp upstream region confers enhancer activity to a 173-bp Gnrh promoter specifically in GT1-7 cells. In the present study, in order to simply examine Gnrh promoter activity induced by BMPs, only the key promoter region containing 173-bp rat Gnrh promoter was utilized for this promoter assay. As a result, BMP6 and BMP7 stimulated transcriptional activity of Gnrh gene, while BMP2 and BMP4 had no significant effects on Gnrh-promoter activity (Fig. 2A). To further examine the BMP effects on steady-state Gnrh mRNA levels in GT1-7 cells, cells were treated with BMP ligands for 24 h in serum-free conditions. Treatment with BMP2 and BMP4 (10 and 100 ng/ml) had no effect on Gnrh mRNA expression by GT1-7 cells (Fig. 2B). In accordance with the results of Gnrh promoter assay, BMP6 and BMP7 (10 and 100 ng/ml) significantly increased Gnrh mRNA expression (Fig. 2B).
In addition, treatment with E2 (1–100 nM) significantly decreased Gnrh-luc activity (Fig. 3A) and Gnrh mRNA expression levels (Fig. 3B). Notably, E2 (100 nM)-induced reduction of Gnrh mRNA was reversed by treatment with BMP2 and BMP4. In contrast, E2-induced Gnrh suppression was not apparently affected by treatment with BMP6 and BMP7 (100 ng/ml; Fig. 3B).
The effects of BMPs on GNRH production were evaluated by measuring GNRH concentration in the conditioned medium by RIA (Fig. 4A). In accordance with the results of Gnrh transcription, BMP6 and BMP7 increased GNRH concentration (pg/ml) by GT1-7 cells, while E2, BMP2, and BMP4 had no effects on the GNRH levels (Fig. 4A). To investigate the involvement of BMPs in GNRH release by GT1-7 cells, the expression changes in a key calcium channel, α1E-subunit were examined (Watanabe et al. 2004). Various stimuli including potassium and acetylcholine have known to induce GNRH secretion through increased influx of intracellular calcium by GT1-7 cells. The α1E-subunit of R-type calcium channel was recently shown to be critical for GNRH secretion in GT1-7 cells (Watanabe et al. 2004). BMP6 and BMP7 increased mRNA expression of α1E-subunit of R-type calcium channel, although E2, BMP2, and BMP4 had no effects on α1E-subunit mRNA (Fig. 4B). In addition, the levels of thymidine incorporation in GT1-7 cells were not increased in response to treatment with E2, BMP-2, -4, -6, and -7, although GT1-7 cells were responsive to exogenous PDGF-BB (Fig. 5). It is thus possible that BMP6 and BMP7 may directly enhance GNRH release through upregulating a critical calcium channel expression in GT1-7 cells.
Western blots also showed SMAD1, 5, 8 signaling activation by BMP-2, -4, -6, and -7 in GT1-7 cells (Fig. 6A). Treatments with BMP-2, -4, -6, and -7, in particular BMP2 and 4, activated Smad1, 5, 8 phosphorylation in GT1-7 cells (Fig. 6B). The BMP-induced SMAD1, 5, 8 activation was not affected in the presence of E2 (100 nM). It was of note that treatment with E2 (100 nM) significantly activated MAPK phosphorylation including ERK1/ERK2 and SAPK/JNK but not p38-MAPK signaling (Fig. 6A). BMP-2 and BMP-4 significantly suppressed E2 (100 nM)-induced phosphorylation of ERK1/ERK2 and SAPK/JNK signaling, whereas BMP6 and BMP7 did not affect E2-induced activation of ERK1/ERK2 and SAPK/JNK (Fig. 6B). Activin A (100 ng/ml) induced p38 phosphorylation in the presence or absence of E2, while activin A did not affect E2-induced ERK1/ERK2 activation.
Since treatment with E2 activated MAPK pathways including ERK1/ERK2 and SAPK/JNK signaling, the functional roles of these pathways in Gnrh transcription were examined. Treatment with 3 μM U0126 specifically inhibited basal and E2-induced ERK1/ERK2 phosphorylation (Fig. 7A). SP600125 (3 and 10 μM) inhibited basal and E2-induced SAPK/JNK phosphorylation in GT1-7 cells, although SB203580 (3 mM) did not affect either ERK1/ERK2 or SAPK/JNK phosphorylation (Fig. 7A). As shown in Fig. 7B, U0126 (3 μM) restored the E2-induced suppression of Gnrh expression, although either U0126 or SP600125 alone had no significant effects on Gnrh expression in GT1-7 cells. In contrast, SP600125 (10 μM) failed to reverse the inhibitory effect of estrogen on Gnrh expression. Thus, the effects of estrogen on Gnrh suppression occur through ERK1/ERK2 activation in GT1-7 cells.
We next investigated whether the E2 actions on MAPK activation and Gnrh suppression are mediated via ER. As shown in Fig. 8A, ERK1/ERK2 phosphorylation induced by E2 (100 nM) was inhibited in the presence of a competitive ER antagonist ICI 182 780 (30–300 nM), although ICI 182 780 alone had no effects on ERK phosphorylation. A membrane-impermeable E2-BSA (100 nM) formulation failed to activate ERK1/ERK2 pathway (Fig. 8A). Notably, treatment with ICI 182 780 reversed the E2-induced suppression of Gnrh mRNA, whereas ICI 182 780 (30 nM) or E2-BSA (100 nM) alone did not change the Gnrh expression in GT1-7 cells (Fig. 8B). These suggest that estrogen inhibits Gnrh expression in GT1-7 cells at least in part by activating ERK1/ERK2 pathway via classic ER. Given the finding that BMP2 and BMP4 suppressed estrogen-induced phosphorylation of MAPK signaling, BMP2 and BMP4 may regulate the ER activity leading to GNRH reduction in GT1-7 cells.
To further investigate the involvement of ER in BMP-induced Gnrh regulation, the expression levels of Erα and Erβ mRNA were examined in GT1-7 cells treated with BMPs and E2. BMP2 and BMP4 (10–100 ng/ml) significantly decreased mRNA levels of Erα and Erβ, while BMP6 and BMP7 increased the expression of Erα and Erβ (Fig. 9A and B). E2 (100 nM) itself also decreased Erα mRNA levels in GT1-7 cells. In addition, the changes of ERα and ERβ protein levels were examined by western immunoblotting in GT1-7 cell lysates treated with BMP-2, -4, -6, and -7 for 24 h. As shown in Fig. 9C, the expression levels of ERα (66 kDa) and ERβ (56 kDa) proteins were decreased by BMP2 and BMP4, while BMP6 and BMP7 in turn increased ERα and ERβ protein expression, although the levels of an internal control actin (42 kDa) were unchanged in GT1-7 cells. Thus, BMP2 and BMP4 are likely to downregulate estrogen effects by attenuating estrogen-induced ERK1/ERK2 activation and by suppressing the expression of ER in GT1-7 cells.
The episodic secretion of pituitary gonadotropin is essential for the maintenance of the mammalian reproductive cycle and normal gonadal function. In the present study, we showed that hyopothalamic GT1-7 neuron cells express functional machinery of the BMP system, and that the BMP system acts to modulate estrogen effects on Gnrh regulation. E2 suppressed GNRH production by activating the MAPK pathway through ER action. In this feedback loop between E2 and GNRH, BMP2 and BMP4 suppress estrogen effects by attenuating estrogen ERK signaling with reduction of ERα and ERβ expression. In contrast, BMP6 and BMP7 directly stimulated Gnrh transcription and secretion by GT1-7 cells. This novel interaction of BMPs and ER action is likely involved in controlling hypothalamic GNRH release in an autocrine/paracrine mechanism (Fig. 10).
Estrogen has rapid regulatory effects on membrane-associated and intracellular responses (often referred to as nongenomic actions), in addition to the well-defined genomic actions in the nucleus. In the present study, E2 readily activated MAPK phosphorylation including ERK1/ERK2 and SAPK/JNK but not p38-MAPK signaling, implying the involvement of nongenomic effects of estrogen in GT1-7 cells. BMP2 and BMP4 suppressed the estrogen-induced phosphorylation of ERK1/ERK2 and SAPK/JNK signaling, whereas BMP6 and BMP7 did not affect the E2-induced activation of ERK1/ERK2 and SAPK/JNK. Since the inhibition of ERK1/ERK2 restored the E2-induced suppression of Gnrh expression by GT1-7 cells, the effects of estrogen on Gnrh suppression is likely to occur through ERK1/ERK2 pathway in GT1-7 cells.
These estrogen effects may be distinct from the genomic actions of estrogen and are attributable to its interactions with membrane-bound receptors and other cellular components (Pietras & Szego 1999, Pietras et al. 2005). Membrane ER has been observed in the hypothalamus and other brain regions, including the cortex, hippocampus, and brain stem. The extent to which GNRH neurons are directly responsive to steroid hormones has been controversial. However, it is now being accepted that hypothalamic cells express both membrane-associated and nuclear receptors, and that they are capable of directly responding to changes in systemic estrogen (Malyala et al. 2005).
In our data, ERK1/ERK2 phosphorylation induced by E2 was inhibited in the presence of ICI 182 780, an ER antagonist that inhibits both ERα and ERβ activities. In addition, the effects of membrane-impermeable E2-BSA were much less effective than E2 for activating ERK1/ERK2 signaling in GT1-7 cells. ICI 182 780 reversed the E2-induced suppression of Gnrh mRNA and E2–BSA failed to suppress Gnrh expression in GT1-7 cells. These findings suggest that estrogen inhibits Gnrh expression, at least in part, by activating ERK1/ERK2 pathway via classic ER signaling. Given the finding that BMP2 and BMP4 suppressed estrogen-induced phosphorylation of MAPK signaling, including ERK1/ERK2, BMP2, and BMP4, may directly regulate the ER activity leading to GNRH reduction in GT1-7 cells.
In agreement with our present results, Roy et al. (1999) demonstrated that E2 represses Gnrh gene expression in GT1-7 cells. This estrogen effect was blocked by ICI 182 780 and also mimicked by 2,2-bis-(p-hydroxyphenyl)-1,1,1-trichloroethane, a putative ERα agonist and ERβ antagonist (Gaido et al. 1999, Roy et al. 1999), suggesting that ERα may be responsible for the repressive effects of E2 on Gnrh expression. However, a phytoestrogen, coumestrol, also suppressed Gnrh expression, which was in turn blocked by an ERβ antagonist, R,R-tetrahydrochrysene (Bowe et al. 2003). These results also imply that estrogen negatively regulates Gnrh promoter activity via not only ERα but also ERβ in hypothalamic cells. In the present study, the expression levels of both Erα and Erβ were decreased in GT1-7 cells treated with BMP2 and BMP4, which could be linked to the suppression of estrogen-induced MAPK activation. To the contrary, the expression levels of Erα and Erβ were increased by BMP6 and BMP7. This effect may be associated with the upregulation of estrogen–GNRH feedback function following the increase in GNRH production caused by BMP6 and BMP7.
The mechanism whereby estrogen regulates GNRH secretion may be partly due to an indirect effect on the neighboring neurons, wherein estrogen acts in part to stimulate the GNRH and LH surge by turning off inhibitory neurons and turning on excitatory neurons. The principal determinants of GNRH neuronal activity include gonadal steroids, various neuropeptides, and neurotransmitters. In this regard, work from several laboratories has shown that hypothalamic astrocytes release the factors that can stimulate GNRH release (Gallo & Russell 1995, Melcangi et al. 1995a, Avola et al. 2000, Buchanan et al. 2000, Ojeda et al. 2000, Dhandapani et al. 2003). Conditioned media collected from purified hypothalamic astrocytes markedly stimulated GNRH release from GT1-7 neurons, an immortalized GNRH neuronal cell line (Buchanan et al. 2000). A similar GNRH stimulatory effect has been reported with conditioned media from cerebral cortical astrocytes (Melcangi et al. 1995b). In our current study, E2 suppressed Gnrh transcription, whereas GNRH protein concentration was not significantly changed in the culture medium. The suppression of Gnrh mRNA was observed in GT1-7 cells in our data and by others (Roy et al. 1999). Regarding GNRH secretion, Navarro et al. (2003) reported that GNRH release was not affected by picomolar of estrogen in static culture of GT1-7 cells and hypothalamic cells, which was rather up-regulated in the presence of cyclic treatment with progesterone. Given that E2 did not affect the expression of a key calcium channel for GNRH secretion in GT1-7 cells, E2 may not directly regulate the GNRH release at least for a certain culture period regardless of the reduction of Gnrh transcription. The discrepancy between transcriptional regulation and secretory process of GNRH needs to be elucidated in a future study.
It has been reported that TGF-β1 as a neuroregulatory factor secreted by astrocytes can modulate GNRH secretion (Melcangi et al. 1995b, Galbiati et al. 1996, Buchanan et al. 2000, Zwain et al. 2002). Additionally, Ojeda et al. have provided evidence in support of a neuroregulatory role for astrocyte-derived TGF-α (Ma et al. 1992, 1997, Ojeda et al. 2000). Furthermore, neuroactive steroid metabolites such as 3α-hydroxy-5α-pregnan-20-one (3α,5α-THP) are also known to be produced by astrocytes, which are capable of stimulating the release of GNRH both in vitro and in vivo (el-Etr et al. 1995, Sim et al. 2001). Given the findings that several BMPs share a neurotropic capacity for dopaminergic neurons with other members of the TGF-β superfamily through glial cells (Jordan et al. 1997, Reiriz et al. 1999, Samanta et al. 2007), it is presumable that hypothalamic GNRH production and/or secretion may also be indirectly regulated by BMP system in the neural network.
In addition, GT1-7 cells generate spontaneous action potentials, exhibit transient oscillations of the intracellular calcium concentration, and secrete GnRH in a pulsatile manner (Charles & Hales 1995, Sun et al. 1998, Morales et al. 2005). The expression of the R-, L-, N-, and T-type calcium channels was demonstrated in GT1-7 cells, in which the R-type subunit plays a key role as a major current component for regulating calcium-dependent GNRH release compared with L-, N-, and T-type channels (Watanabe et al. 2004). Since BMP6 and BMP7 increased mRNA expression of α1E-subunit of R-type calcium channel, which is a critical subunit for hypothalamic GNRH secretion, it seems likely that BMP6 and BMP7 directly stimulate GNRH release by GT1-7 cells. Further investigation would be necessary to clarify whether BMP6 and BMP7 physiologically stimulate calcium channel activity in GNRH neural cells in the future study.
Recent studies have shown that BMPs regulate the pituitary–gonadal endocrine axis. However, roles of BMPs in hypothalamic GNRH production have yet to be clarified. Understanding the neural mechanisms through which E2 regulates GNRH neurons is key for elucidating reproductive control of the HPO axis. We uncovered here the effects of BMPs on GNRH regulation controlled by estrogen in hypothalamic neuron cells. In particular, BMP2 and BMP4 downregulate estrogen effects on suppressing GNRH production by attenuating estrogen ERK signaling and by suppressing ER expression. On the contrary, BMP6 and BMP7 contribute to increase in GNRH secretion. Hence, a new interaction of BMPs and ER may be involved in controlling hypothalamic GNRH production and secretion in an autocrine/paracrine mechanism (Fig. 10). The BMP system may play a key role in regulating not only in the ovarian folliculogenesis by regulating gonadotropin sensitivity but also in modulating hypothalamic GNRH secretions leading to fine-tuning of gonadotropin secretion from the pituitary gonadotrope.
We thank Dr Pamela L Mellon for providing GT1-7 cells. We also thank Dr R Kelly Moore for helpful discussion and critical reading of the manuscript.
This work was supported in part by Grants-in-Aid for Scientific Research, The Naito Foundation, Kanae Foundation for The Promotion of Medical Science, and Kanzawa Medical Research Foundation. M T was supported by the National Institute of Child Health and Human Development/National Institutes of Health through Grant R01 HD 037568 and U54 HD12303 as part of the Specialized Co-operative Centers Program in Reproduction and Infertility Research to M A L.
Declaration of interest
All authors have nothing to disclose.