GnRH Pulse Frequency Control of Fshb Gene Expression Is Mediated via ERK1/2 Regulation of ICER

Abstract

The pulsatile release of GnRH regulates the synthesis and secretion of pituitary FSH and LH. Two transcription factors, cAMP-response element-binding protein (CREB) and inducible cAMP early repressor (ICER), have been implicated in the regulation of rat Fshb gene expression. We previously showed that the protein kinase A pathway mediates GnRH-stimulated CREB activation. We hypothesized that CREB and ICER are activated by distinct signaling pathways in response to pulsatile GnRH to modulate Fshb gene expression, which is preferentially stimulated at low vs high pulse frequencies. In the LβT2 gonadotrope-derived cell line, GnRH stimulation increased ICER mRNA and protein. Blockade of ERK activation with mitogen-activated protein kinase kinase I/II (MEKI/II) inhibitors significantly attenuated GnRH induction of ICER mRNA and protein, whereas protein kinase C, calcium/calmodulin-dependent protein kinase II, and protein kinase A inhibitors had minimal effects. GnRH also stimulated ICER in primary mouse pituitary cultures, attenuated similarly by a MEKI/II inhibitor. In a perifusion paradigm, MEKI/II inhibition in LβT2 cells stimulated with pulsatile GnRH abrogated ICER induction at high GnRH pulse frequencies, with minimal effect at low frequencies. MEKI/II inhibition reduced GnRH stimulation of Fshb at high and low pulse frequencies, suggesting that the ERK pathway has additional effects on GnRH regulation of Fshb, beyond those mediated by ICER. Indeed, induction of the activating protein 1 proteins, cFos and cJun, positive modulators of Fshb transcription, by pulsatile GnRH was also abrogated by inhibition of the MEK/ERK signaling pathway. Collectively, these studies indicate that the signaling pathways mediating GnRH activation of CREB and ICER are distinct, contributing to the decoding of the pulsatile GnRH to regulate FSHβ expression.

The hypothalamic decapeptide, GnRH, binds to its native high-affinity 7-transmembrane spanning receptor, gonadotropin-releasing hormone receptor, on the cell surface of anterior pituitary gonadotropes. Upon binding to the gonadotropin-releasing hormone receptor, GnRH stimulates the synthesis and secretion of FSH and LH (1). Subsequently, FSH and LH mediate steroidogenesis and gametogenesis, critical for reproductive function (2). GnRH is released intermittently to produce pulses, and changes in GnRH pulse frequencies and amplitudes have differential effects on FSH and LH synthesis and release (3, 4). In humans, GnRH pulse frequency varies throughout the menstrual cycle, with pulses on average every 6 hours in mid-to-late luteal phases and every 90 minutes during follicular and early luteal phases (5). In rodents, GnRH pulse intervals are shorter (ranging from 8 minutes to 240 min) (6). High GnRH pulse frequencies preferentially stimulate LH synthesis and secretion, whereas FSH synthesis and secretion are preferentially stimulated at lower frequencies. Sharing a common α-glycoprotein subunit, FSH and LH are differentiated by distinct β-subunits (FSHβ and LHβ, respectively) (7). Cga, Fshb, and Lhb gene transcription are also differentially regulated by varying GnRH pulse frequencies (8, 9), although α-glycoprotein subunit is produced in excess, regardless of GnRH pulse frequency, and thought to be less important in the ultimate control of FSH and LH synthesis. Studies using rodent models have demonstrated that Fshb gene expression is optimally stimulated by GnRH pulses every 120 minutes, whereas Lhb gene expression is higher at shorter pulse intervals, every 30 minutes (8–10). The pulsatile manner by which GnRH is released serves as a primary mechanism by which the synthesis and secretion of FSH and LH from the gonadotropes are controlled (as reviewed in Ref. 11). Changes in the pulse frequency and amplitude of GnRH release result in differential FSH and LH synthesis and secretion, contributing to the regulation of the female menstrual cycle (3, 4).

Disruption of the hormonal control and regulation of the menstrual cycle can result in reproductive disorders, often resulting in subfertility (12). For example, accelerated GnRH frequency release in humans, corresponding with subsequent excessive LH over FSH secretion, is seen in polycystic ovarian syndrome, which affects 5%–15% of the female population within reproductive age. Polycystic ovarian syndrome often results in infertility (13) and has been associated with obesity, insulin resistance, and metabolic and cardiovascular abnormalities (14–16).

It has been shown that a cAMP-response element (CRE)/activating protein 1 (AP1) site in the Fshb gene promoter, fully conserved in humans, is GnRH responsive and is important for the GnRH pulse frequency-dependent control of Fshb gene expression (17–20). GnRH stimulates the phosphorylation of the transcription factor, CRE-binding protein (CREB) (17, 21), leading to the recruitment of CREB-binding protein and subsequent Fshb transcription. There is also evidence of a role for transcriptional repressors in this system. The inducible cAMP early repressor (ICER) functionally antagonizes CREB in other systems (22, 23). It has been shown that ICER expression is induced preferentially at high GnRH pulse frequencies and binds to the CRE/AP1 site in the LβT2 cell line, resulting in reduced CREB binding and a significantly lower level of induction of Fshb transcription compared with lower frequencies of pulsatile GnRH (20). We hypothesized that the pathways by which pulsatile GnRH induces CREB (a stimulator of Fshb transcription) at slow GnRH pulse frequencies and ICER (a repressor of Fshb transcription) at fast pulse frequencies would be distinct (as reviewed in Ref. 24), to result in GnRH pulse frequency-dependent differential control of Fshb transcription. In the current study, we demonstrate that ICER is preferentially stimulated at high GnRH pulse frequencies through an ERKI/II dependent mechanism. Inhibition of the mitogen-activated protein kinase kinase/extracellular signal-regulated kinase (MEK/ERK) signaling pathway prevented the induction of ICER by pulsatile GnRH and resulted in loss of the frequency dependent pattern of ICER induction. However, inhibition of the ERK pathway also completely abrogated induction of Fshb expression at both high and low GnRH pulse frequencies, and similarly prevented induction of Lhb by pulsatile GnRH. These findings suggest that the MEK/ERK signaling pathway is involved in the stimulation of Fshb and Lhb gene expression by pulsatile GnRH directly, in addition to its role in the induction of the repressor, ICER, at fast GnRH pulse frequencies.

Materials and Methods

Materials

Materials were purchased either from Fischer Scientific or Sigma Chemical Co. The protein kinase A (PKA) (H-89) inhibitor was purchased from Sigma. The protein kinase C (PKC) (GF109203X) and calcium/calmodulin-dependent protein kinase (CamK)II (KN-93) inhibitors were obtained from Tocris. The MEKI/II (PD0325901) inhibitor was purchased from LC Laboratories. cFos and cJun antibodies (sc-52 and sc-1694 respectively) were purchased from Santa Cruz Biotechnology, Inc. ICER antibody was kindly provided by Carlos Molina (New Jersey Medical School, Newark, NJ). The LβT2 cell line was a kind gift from Pamela L. Mellon (University of California, San Diego, CA).

Cell culture

LβT2 cells were maintained in high-glucose DMEM (Mediatech, Inc), supplemented with 10% (vol/vol) fetal bovine serum (FBS) (Omega), 100 U of penicillin/mL, and 100 μg of streptomycin sulfate/mL (Thermo Fisher Scientific, Inc) in 5% CO₂ humidified 37°C air. Unless otherwise stated, cells were serum starved for 24 hours with FBS-free DMEM (static cultures only), before treatment with 10nM GnRH (Sigma) with or without pharmacologic pathway inhibitors. This concentration of GnRH was selected for consistency with the perifusion studies presented in this manuscript, as well as other studies from our group (20, 21, 25–27) and others (19, 28–30).

For primary pituitary cell cultures, 12 CD-1 IGS male mice (22–24-g body weight, 7 wk old) were purchased (Charles River Laboratories), killed by CO₂ asphyxiation, decapitated, and pituitaries collected in 10% FBS supplemented DMEM, followed by centrifugation at 2500 rpm for 5 minutes. The pituitaries were washed with Hanks' buffered saline solution (HBSS) (Thermo Fisher Scientific, Inc), followed by centrifugation at 2500 rpm for 5 minutes, aspiration of HBSS, and digestion with 2 mL of collagenase solution (HBSS supplemented with 30-mg/mL BSA [Sigma] and 1.5-mg/mL collagenase [Sigma]) for 30 minutes at 37°C with agitation. Finally, cells were washed once with, and subsequently resuspended in, 10% FBS supplemented DMEM and seeded at a density of 1.0 × 10⁶ cells/well (12-well plate; Corning Costar). After incubation for 48 hours in 5% CO₂ humidified 37°C air, the dispersed pituitary cells were treated with 100nM GnRH for up to 12 hours in serum-free DMEM, with or without the MEKI/II (PD0325901) pharmacologic pathway inhibitor, followed by RNA or protein extraction as described below.

Perifusion studies

As noted, preferential induction of LHβ vs FSHβ subunit synthesis in response to varying GnRH pulse frequencies has been observed in both primary pituitary (26, 31) and pituitary cell line derived cultures (20, 21, 27, 32), with Fshb gene expression optimally stimulated by GnRH pulses every 120 minutes, whereas Lhb gene expression is higher at shorter pulse intervals, every 30 minutes, in rodent models (10). Consistent with these studies, LβT2 cells in the perifusion experiments presented in this manuscript were treated with pulsatile GnRH at intervals of 30 or 120 minutes. The amplitude and duration of GnRH pulses (10nM, 5 min) were selected based on previous optimization studies demonstrating their ability to result in sustained gonadotropin responses without desensitization from dispersed rodent primary pituitary cell cultures (33). After the final pulse, total RNA or protein was isolated at the indicated times with either TRI Reagent (Sigma) or RIPA buffer (Santa Cruz Biotechnology, Inc), respectively.

mRNA and protein quantification

As previously described (21), LβT2 cells harvested for total RNA with TRI Reagent were processed using the chloroform/isopropanol method. One microgram of total RNA was reverse transcribed using the Superscript III cDNA synthesis kit (Thermo Fisher Scientific, Inc), followed by quantitative real-time PCR analysis performed on an ABI PRISM 7000 sequence detection system (Applied Biosystems) using iQ SYBR Green (Bio-Rad Laboratories) according to manufacturer's instructions. Results were analyzed using ABI PRISM 7000 SDS software (Applied Biosystems), and products were subsequently electrophoresed on an agarose gel to verify that a single product was amplified. Levels of mRNA were normalized to ribosomal protein L19 (Rpl19) as an internal control.

Total protein isolated from LβT2 cells using RIPA buffer was quantified, and 60 μg were separated by SDS-PAGE and transferred onto a nitrocellulose membrane (Whatman) in preparation for Western blot analysis. The membranes were blocked in 3% nonfat milk in Tris-buffered saline plus Tween 20 (TBST) for 1 hour at room temperature with shaking, followed by incubation for 16 hours at 4°C with ICER, cFos, cJun, or β-actin primary antibodies (diluted 1:3000, 1:2000, 1:2000, and 1:5000, respectively) with gentle shaking. After washes with TBST, the membranes were exposed to either donkey antirabbit (ICER, cFos, and cJun), or donkey antimouse (β-actin) IgG-horseradish peroxidase (1:5000 dilution, sc-2313 or sc-2314, respectively; Santa Cruz Biotechnology, Inc) secondary antibody for 1 hour at room temperature with shaking. After further washes in TBST, complexes were detected using HyGLO Quick Spray Chemiluminescent horseradish peroxidase antibody detection reagent (Denville Scientific, Inc). ICER, cFos, and cJun signal intensities were measured by ImageJ software (National Institutes of Health), normalized to β-actin detected on separate SDS-PAGE gels run concurrently, and expressed as fold change over control samples.

Statistical analysis

All graphs were generated and statistical analysis performed using Prism software (GraphPad Software, Inc). Unless otherwise stated, one-way ANOVA followed by post hoc Newman-Keuls multiple comparison tests were carried out to determine statistical significance. P < .05 was considered to be statistically significantly different. Numerical data presented in this manuscript represent the mean ± SEM from at least 3 independent experiments, each performed in triplicate.

Results

MEKI/II inhibition attenuates GnRH induction of ICER mRNA expression

In order to identify the signaling pathways that are involved in the decoding of the pulsatile GnRH signal, we used the murine gonadotrope-derived LβT2 cell line model (34). Early studies investigating the response of this cell line to pulsatile GnRH showed that varying pulse frequency differentially regulates Fshb and Lhb expression (27), mirroring results from studies conducted in vivo and ex vivo which demonstrated differential Fshb and Lhb expression in response to pulsatile GnRH in primary gonadotropes (9, 26, 31). These data support the use of LβT2 cells to examine the signaling pathways stimulated by varying pulses of GnRH, leading to subsequent Fshb or Lhb transcription. Recently, we demonstrated that the PKA pathway was preferentially stimulated in LβT2 cells in response to pulsatile GnRH at low (every 120 min) pulse frequencies (21), complementing a similar study (35). Additional studies have also shown that GnRH stimulates the PKA pathway in the gonadotrope, as well as the CamK, PKC, and MAPK pathways, as reviewed recently (24).

To investigate whether these pathways are important for GnRH-stimulated ICER induction, LβT2 cells were treated in static culture with 10nM GnRH for 0, 12, and 24 hours, in the presence of increasing concentrations of inhibitors for PKA, CamKII, PKC, or MEKI/II as described (Figure 1), inclusive of the predicted IC₅₀ as reported by the manufacturer. As previously demonstrated (20), GnRH stimulation of LβT2 cells induced ICER mRNA expression by a mean of 53.0 ± 7.3-fold (P < .0001) after 12 hours and 20.0 ± 3.8-fold (P < .05) after 24 hours (Figure 1, A–D, red bars). The MEKI/II inhibitor, PD0325901, significantly reduced GnRH-stimulated ICER induction (Figure 1D). In the presence of PD0325901, at both concentrations, GnRH did not significantly increase ICER mRNA levels compared with basal controls. Similar results were obtained using a second MEKI/II inhibitor, U0126, validating this finding (data not shown). Analysis of the effects of the PKA and CamKII inhibitors, H89 and KN-93, respectively, demonstrated that in the presence of increasing concentrations, 12 hours of GnRH treatment still significantly stimulated ICER mRNA levels compared with controls (Figure 1, A and B). The activity of both inhibitors was confirmed in control experiments, in which H89 inhibited GnRH-stimulated CREB phosphorylation (21), and KN-93 inhibited CamKII phosphorylation (36; data not shown). A modest decrease of GnRH-stimulated ICER mRNA induction was observed in the presence of the PKC inhibitor, GF109203X (Figure 1C). This decrease was in proportion to the inhibition of ERK phosphorylation by GF109203X (data not shown), suggesting that ERK phosphorylation in response to GnRH occurred only in part through PKC mediated pathways. Regardless of GF109203X concentration, ICER mRNA levels were still significantly increased after 12 hours of stimulation with GnRH, compared with controls (Figure 1C). Taken together, these data suggest that the MEK/ERK pathway is primarily responsible for the induction of ICER mRNA expression in LβT2 cells stimulated with GnRH in static culture.

Figure 1.

GnRH stimulation of ICER mRNA levels is attenuated by a MEKI/II inhibitor. LβT2 cells in static culture were pretreated for 30 minutes with the indicated concentrations of inhibitors of (A) PKA (H89), (B) CamKII (KN-93), (C) PKC (GF109203X), and (D) MEKI/II (PD0325901), followed by treatment with 10nM GnRH for 0 (red bars), 12 (light blue), or 24 (dark blue) hours. Bar graphs show relative ICER mRNA levels (mean ± SEM from 3 independent experiments, each performed in triplicate, normalized to Rpl19 mRNA levels). Significant differences (P < .05), measured by one-way ANOVA with a post hoc Newman-Keuls multiple comparison test, are indicated by different letters.

GnRH stimulates ICER mRNA synthesis in cultured primary mouse pituitary cells

In both primary rat pituitary (25, 26, 31, 37–40) as well as mouse pituitary (41–47) cell cultures, the effects of both GnRH and activin on gonadotropin subunit transcription and expression have been investigated. In order to confirm the effect of GnRH treatment on ICER induction in LβT2 cells in a more physiologically relevant model, dispersed primary pituitary cells from CD-1 adult male mice were studied. Over a 0-, 4-, and 12-hour time course of static GnRH, we observed that 12 hours was optimal to stimulate ICER mRNA synthesis in these primary pituitary cells (data not shown). Primary murine pituitary cell cultures were then treated with 100nM GnRH in the presence or absence of 10nM of the MEKI/II inhibitor, PD0325901 (Figure 2). GnRH stimulation of the primary pituitary cells induced ICER mRNA expression by 1.5 ± 0.1-fold (P < .01) after 12 hours. MEKI/II inhibition completely eliminated GnRH-stimulated ICER induction (Figure 2).

Figure 2.

GnRH stimulation of ICER mRNA levels in dispersed primary mouse pituitary cells is attenuated by a MEKI/II inhibitor. Dispersed mouse pituitary cells in static culture were pretreated for 30 minutes in the presence (blue bars) or absence (red bars) of 10nM PD0325901, followed by treatment with 100nM GnRH for 0 or 12 hours. Bar graphs show relative ICER mRNA levels (mean ± SEM from 3 independent experiments, each performed in triplicate, normalized to Rpl19 mRNA levels). Significant differences (P < .05), measured by one-way ANOVA with a post hoc Newman-Keuls multiple comparison test, are indicated by different letters.

MEKI/II inhibition attenuates GnRH-stimulated ICER protein induction

Because MEKI/II blockade significantly reduced GnRH-stimulated ICER mRNA expression (Figures 1D and 2), we also explored the role of the ERK pathway in GnRH induction of cellular ICER protein levels. GnRH has been shown previously to stimulate ICER protein levels (20); therefore, we wanted to confirm that MEKI/II inhibition would attenuate the ability of GnRH to stimulate ICER protein levels in addition to its effects on ICER mRNA expression.

Using the identical experimental paradigm described for Figure 1D, LβT2 cells were treated in static culture with 10nM GnRH for 0, 12, and 24 hours, in the presence of increasing concentrations of the MEKI/II inhibitor, PD0325901. Lysates were extracted from LβT2 cells after 12 or 24 hours treatment with GnRH. Figure 3A illustrates a Western blotting from a representative experiment. A significant increase in ICER protein levels after GnRH stimulation was noted (2.0 ± 0.1-fold, P < .001 at 12 h and 2.3 ± 0.2-fold, P < .0001 at 24 h) (Figure 3B, red bars), compared with untreated controls. Similarly to the effects observed on ICER mRNA induction, MEKI/II inhibition at either concentration completely abrogated GnRH stimulation of cellular ICER protein levels, compared with control samples (Figure 3B, shaded blue bars). The induction of ICER protein after 24 hours is equivalent to the induction at 12 hours (Figure 3B, red bars). This differs from ICER mRNA induction (Figure 1), suggesting 2 distinct time courses, induction of mRNA followed by protein production. In addition, 100nM GnRH induced ICER protein levels in dispersed primary mouse pituitary cell cultures, and this effect was also inhibited by MEKI/II blockade using 10nM PD0325901 (data not shown). Thus, similarly to the ICER mRNA data presented in Figures 1 and 2, these data demonstrate the importance of the MEK/ERK pathway in the induction of ICER protein by GnRH.

Figure 3.

GnRH stimulation of ICER protein levels is attenuated by a MEKI/II inhibitor. LβT2 cells in static culture were pretreated for 30 minutes with the indicated concentrations of PD0325901, followed by treatment with 10nM GnRH for 0, 12, or 24 hours. A, Western immunoblotting shows a representative experiment. B, Bar graph shows relative ICER levels (mean ± SEM from 3 independent experiments, each performed in triplicate, normalized to β-actin). Significant differences (P < .05), measured by one-way ANOVA with a post hoc Newman-Keuls multiple comparison test, are indicated by different letters.

Stimulation of ICER by pulsatile GnRH is attenuated by MEKI/II inhibition at high GnRH pulse frequency

Our initial studies have firmly established a role of the MEK/ERK pathway in mediating the induction of ICER mRNA and protein by GnRH in static culture. However, the goal of our study is to identify the pathways that regulate ICER synthesis after stimulation with pulsatile GnRH. This paradigm, in which LβT2 cells are perifused with pulsatile GnRH at high (every 30 min) or low (every 120 min) pulse frequencies for 20 hours, is more physiologically relevant and has been shown to result in preferential stimulation of Fshb gene expression at low GnRH pulse frequencies and of Lhb at high GnRH pulse frequencies (20, 21, 27). The time points selected were optimized in previous studies (8, 9, 20, 21, 27, 37, 40, 48).

Pulsatile GnRH stimulates ICER mRNA expression (20); the aim of this experiment was to determine whether this effect is also mediated via MEK/ERK pathways. LβT2 cells were perifused and stimulated with pulsatile GnRH at either high (every 30 min) or low (every 120 min) pulse frequencies or with medium alone for 20 hours, as previously described (20, 21, 27), but in the presence or absence of the MEKI/II inhibitor, PD0325901. ICER mRNA levels were quantified by real-time RT-qPCR. As shown previously (20), ICER mRNA levels were stimulated by pulsatile GnRH, with the greatest response at the high GnRH pulse frequency (high frequency 10.1 ± 0.5-fold, P < .0001; low frequency 2.4 ± 0.2-fold, P < .05) (Figure 4, red bars). MEKI/II blockade completely abrogated GnRH stimulation of ICER mRNA levels at the high pulse frequency (P < .0001) (Figure 4, blue bar, column 2), to a level not significantly different from control-treated (media alone) cells (1.4 ± 0.01-fold, P > .05). At the low GnRH pulse frequency, ICER mRNA levels were increased only modestly in the presence of the MEKI/II inhibitor (1.8 ± 0.2-fold, P > .05) compared with control-treated cells, and a modest but not statistically significant decrease in ICER mRNA levels was observed compared with cells without the MEKI/II inhibitor was observed (P > .05) (Figure 4, blue bar, column 3). In sum, these data suggest that the MEK/ERK pathway is important for mediating induction of ICER mRNA expression by pulsatile GnRH at high GnRH pulse frequencies. In addition, these data indicate that MEKI/II inhibition results in the loss of GnRH pulse frequency dependent induction of ICER.

Figure 4.

GnRH induction of ICER mRNA is attenuated by MEKI/II inhibition at high GnRH pulse frequency. LβT2 cells were perifused with high (⋀⋀⋀⋀; every 30 min) or low (⋀___⋀; every 120 min) frequencies of pulsatile GnRH or medium alone (____) in the presence (blue bars), or absence (red bars) of 10nM PD0325901 for 20 hours, after which ICER mRNA levels were quantified by real-time RT-PCR. Bar graph shows relative change compared with control samples (mean ± SEM from 3 independent experiments, each performed in triplicate, normalized to Rpl19 mRNA levels). Significant differences (P < .05), measured by one-way ANOVA with a post hoc Newman-Keuls multiple comparison test, are indicated by different letters.

ICER protein levels are stimulated by pulsatile GnRH in a frequency-dependent manner that is attenuated by MEKI/II inhibition at high GnRH pulse frequency

To extend our findings for the effects of pulsatile GnRH on ICER mRNA expression to cellular ICER protein levels, we again treated LβT2 cells with pulsatile GnRH at either high (every 30 min) or low (every 120 min) pulse frequencies or with medium alone for 20 hours, followed by protein extraction and Western immunoblot detection of ICER. Figure 5A illustrates a Western blotting from a representative experiment. ICER protein levels were increased at both high and low GnRH pulse frequencies, with the greatest induction at the high pulse frequency, as observed previously (high frequency 3.5 ± 0.6-fold, P < .001; low frequency 2.3 ± 0.2-fold, P < .05) (Figure 5B, red bars) (20). As observed for ICER mRNA expression (Figure 4), MEKI/II blockade completely abrogated GnRH stimulation of ICER protein levels at the high pulse frequency (P < .001) (Figure 5B, blue bars). Specifically, GnRH stimulation of ICER protein at high pulse frequency in the presence of the MEKI/II inhibitor was not significantly different from control samples (high frequency 1.5 ± 0.2-fold, P > .05) (Figure 5B, blue bar). At the low GnRH pulse frequency, the modest induction of ICER protein levels was prevented by MEKI/II inhibition, but levels were not significantly different from the GnRH-stimulated levels at this pulse frequency (low frequency P > .05) (Figure 5B, blue bar). Compared with control samples, ICER protein induction was not significant at low GnRH pulse frequency in the presence of the MEKI/II inhibitor (1.3 ± 0.2-fold, P > .05).

Figure 5.

ICER protein levels are induced by pulsatile GnRH in a frequency-dependent manner that is attenuated by MEKI/II inhibition. LβT2 cells were perifused with high (⋀⋀⋀⋀; every 30 min) or low (⋀___⋀; every 120 min) frequencies of pulsatile GnRH or medium alone (____) for 20 hours, in the presence (blue bars), or absence (red bars), of 10nM PD0325901, followed by ICER protein quantification by Western immunoblot analysis. A, A representative Western immunoblotting is shown. B, Bar graphs show relative ICER protein levels (mean ± SEM from at least 3 independent experiments, each performed in triplicate, normalized to β-actin). Significant differences (P < .05), measured by one-way ANOVA with a post hoc Newman-Keuls multiple comparison test, are indicated by different letters.

The regulation of ICER mRNA and protein by pulsatile GnRH in the presence or absence of a MEKI/II inhibitor, as presented in Figures 4 and 5, further supports a critical role for the MEK/ERK signaling cascade in the GnRH pulse frequency dependent induction of ICER, an inhibitor of GnRH induction of FSHβ expression.

Stimulation of Fshb and Lhb by pulsatile GnRH is attenuated by MEKI/II inhibition at both high and low GnRH pulse frequencies

The overarching goal of these studies is to elucidate the mechanisms and pathways by which pulsatile GnRH modulates Fshb gene expression and FSH synthesis. ICER competes with CREB for binding to the CRE/AP1 site within the rat Fshb gene promoter after stimulation with pulsatile GnRH at high pulse frequencies, resulting in reduced stimulation of Fshb gene transcription compared with pulsatile GnRH at low frequencies (20). Therefore, based on our data (Figure 5), we hypothesized that MEKI/II inhibition, by reducing the induction of ICER at high GnRH pulse frequencies, would result in a greater induction of Fshb mRNA levels. Because the GnRH pulse frequency-dependent regulation of ICER was lost in the presence of the MEK inhibitor, we hypothesized that the frequency dependent response of Fshb gene expression would also be lost.

LβT2 cells were again perifused and stimulated by pulsatile GnRH at high and low pulse frequencies for 20 hours, and Fshb mRNA levels were measured by real-time RT-qPCR. In this study, Fshb mRNA levels were stimulated by pulsatile GnRH, with the greatest response at the low GnRH pulse frequency, as expected (high frequency 14.3 ± 2.1-fold, P < .0001; low frequency 40.0 ± 2.1-fold, P < .0001) (Figure 6A, red bars) (20, 21). Interestingly, MEKI/II inhibition reduced Fshb mRNA induction at both high and low GnRH pulse frequencies (high frequency stimulation 5.3 ± 0.4-fold, not significantly different from controls P > .05; low frequency stimulation 8.5 ± 0.8-fold, significant compared with controls P < .01) (Figure 6A, blue bars), and removed the GnRH pulse frequency dependent differential regulation of Fshb gene expression between high and low GnRH pulse frequencies (P > .05).

Figure 6.

GnRH induction of Fshb and Lhb mRNA at both high and low GnRH pulse frequencies is attenuated by MEKI/II inhibition. LβT2 cells were perifused with high (⋀⋀⋀⋀; every 30 min) or low (⋀___⋀; every 120 min) frequencies of pulsatile GnRH or with medium alone (____), in the presence (blue bars), or absence (red bars), of 10nM PD0325901 for 20 hours. Fshb (A) and Lhb (B) mRNA levels were quantified by real-time RT-qPCR. Bar graphs show fold changes in mRNA levels compared with control samples (mean ± SEM from 3 independent experiments, each performed in triplicate, normalized to Rpl19 mRNA levels). Significant differences (P < .05), measured by one-way ANOVA with a post hoc Newman-Keuls multiple comparison test, are indicated by different letters.

Lhb gene expression is well known to be dependent on ERK activation, mediated through early growth response protein-1 (Egr1) (49–55). As expected (20, 21), Lhb mRNA levels were stimulated by pulsatile GnRH, with the highest response at the high GnRH pulse frequency (high frequency 6.4 ± 0.5-fold, P < .0001; low frequency 2.0 ± 0.1-fold, P < .01) (Figure 6B, red bars). Furthermore, MEKI/II inhibition completely abrogated the induction of Lhb mRNA levels at both high and low GnRH pulse frequencies (high frequency stimulation 0.7 ± 0.05-fold, not significant compared with controls P > .05; low frequency stimulation 1.0 ± 0.04-fold, not significant compared with controls P > .05) (Figure 6B, blue bars). There is no statistically significant difference between the high and low frequency pulsatile GnRH stimulation of Lhb mRNA levels in the presence of the MEKI/II inhibitor (P > .05). These data are consistent with previous studies that showed that stimulation of LHβ promoter activity, as reflected by LHβLUC reporters, by static or pulsatile GnRH was reduced after inhibition of the MEKI/II signaling pathway (32, 56–59).

Taken together, these data suggest a role for the MEK/ERK pathway in the inhibition of both Fshb and Lhb gene expression by pulsatile GnRH at both high and low GnRH pulse frequencies, despite the inhibition of ICER induction at high frequencies of pulsatile GnRH, which would be expected to result in an augmented response of FSHβ.

Stimulation of AP1 family members by pulsatile GnRH is inhibited by MEKI/II blockade

We have found that MEKI/II inhibition reduces GnRH-stimulated ICER induction to a greater extent at high rather than low GnRH pulse frequencies. However, the expected increase in Fshb mRNA levels in response to pulsatile GnRH at the high GnRH pulse frequency after MEKI/II blockade was not observed. In contrast, Fshb mRNA levels were actually decreased, at both high and low GnRH pulse frequencies, after MEKI/II inhibition. We hypothesized that MEKI/II inhibition might be interfering with the activity of other positive regulators of murine Fshb gene expression. A candidate for such positive FSHβ regulators is the AP1 protein family. AP1 homo- and heterodimers, comprised of a combination of Jun and Fos intermediate-early gene family members, are well known to be induced by GnRH through MAPK/ERK pathways, resulting in increases in Fshb gene expression (19, 60–64).

Indeed, in agreement with previous studies demonstrating the importance of AP1 proteins, acting via the AP1-binding site in the Fshb promoter, in regulating murine Fshb transcription (19, 60–64), cFos protein levels were stimulated by pulsatile GnRH (Figure 7A illustrates a Western blotting from a representative experiment), with the greatest induction at the high GnRH pulse frequency (high frequency 5.8 ± 0.9-fold, P < .0001; low frequency 2.8 ± 0.4-fold, P < .05; high vs low frequency difference P < .001) (Figure 7B, red bars). A similar frequency-dependent response (Figure 7A illustrates a Western blotting from a representative experiment) was observed for cJun protein levels (high frequency 6.8 ± 0.5-fold, P < .0001; low frequency 5.1 ± 0.6-fold, P < .0001; high vs low frequency difference P < .01) (Figure 7C, red bars), as also reported previously (60). MEKI/II blockade completely abrogated GnRH induction of cFos and cJun protein levels at both high and low GnRH pulse frequencies (high frequency stimulation of 0.9 ± 0.2-fold, P > .05; low frequency stimulation of 1.1 ± 0.2-fold, P > .05 [cFos, Figure 7B, blue bars]; high frequency stimulation of 1.4 ± 0.04-fold, P > .05; low frequency stimulation of 2.0 ± 0.2-fold, P > .05 [cJun, Figure 7C, blue bars]). MEKI/II inhibition eliminated GnRH frequency-dependent differences in cFos and cJun protein levels (P > .05).

Figure 7.

AP1 protein levels are induced by pulsatile GnRH in a frequency-dependent manner that is attenuated by MEKI/II inhibition. LβT2 cells were perifused and stimulated with pulsatile GnRH at high (⋀⋀⋀⋀; every 30 min) or low (⋀___⋀; every 120 min) frequencies, or with medium alone (____), in the presence (blue bars), or absence (red bars), of 10nM PD0325901. Representative Western immunoblottings (A) are shown. Bar graphs show relative cFos (B) and cJun (C) levels (mean ± SEM from 3 independent experiments, each performed in triplicate, normalized to β-actin). Significant differences (P < .05), measured by one-way ANOVA with a post hoc Newman-Keuls multiple comparison test, are indicated by different letters.

In summary, our data demonstrate that inhibition of the ERK pathway with a MEKI/II inhibitor disrupts the GnRH pulse frequency dependent synthesis of both positive (cFos and cJun) and negative (ICER) regulators of Fshb transcription, resulting in the loss of differential Fshb expression in response to pulsatile GnRH.

Discussion

Early studies have demonstrated that different frequencies of pulsatile GnRH, released from the hypothalamus, result in altered secretion patterns of FSH and LH (3, 4). Increasing frequencies result in preferential secretion of LH, whereas conversely decreasing frequencies lead to greater FSH secretion. Because FSH and LH synthesis and secretion are both modulated by GnRH (1, 3, 4), through the same G protein-coupled GnRH receptor, one possible mechanism is that distinct signaling pathways are activated, dependent on GnRH pulse frequency, to lead to the differential regulation of FSH and LH production (65–67). It remains unknown how gonadotropes decode the pulsatile GnRH signal to preferentially synthesize FSH or LH.

Comprising 5%–15% of the anterior pituitary (68, 69), primary gonadotropes are difficult to isolate in the numbers required to conduct investigations of the signaling pathways involved in the responses to pulsatile GnRH. Two murine gonadotrope-derived cell lines are available to circumvent these challenges, αT3–1 (70) and LβT2 cells (34). The more mature LβT2 gonadotrope-derived cell line has been shown to express FSHβ in response to activin A and GnRH, as well as LHβ, and to synthesize and secrete both FSH and LH (71–75). LβT2 (51, 60) and primary pituitary (25, 26, 31, 37, 38) cells have been used most commonly in studies investigating the effects of pulsatile GnRH. Studies from our group have demonstrated that LβT2 cells, when exposed to high (every 30 min) vs low (every 120 min) pulses of GnRH, can respond with differential expression of FSHβ and LHβ (20, 21, 27, 32). Therefore, this makes LβT2 cells the best available model for our studies to identify the mechanisms by which gonadotropes decode the pulsatile GnRH signal to control Fshb and Lhb gene expression.

As previously described (17–19), the rat Fshb promoter contains a partial CRE, predominantly bound by the CREB, which is also conserved in humans. CREB phosphorylation in response to low frequency pulsatile GnRH signaling and subsequent PKA activation increases Fshb gene expression (21). It has been shown previously that mutation of the GnRH-responsive CRE results in the loss of preferential Fshb transcription at low GnRH pulse frequencies (20), leading to the hypothesis that a transcriptional repressor is induced at high GnRH pulse frequencies, attenuating the actions of CREB. This transcriptional repressor of Fshb transcription was identified as ICER (20).

In this study, using a method similar to that used to identify PKA as a mediator of CREB phosphorylation and hence Fshb gene expression (21), we tested possible signaling pathways leading to the induction of ICER by GnRH using a series of pharmacological inhibitors. Studies have shown that calcium and CamKI/II signaling, MAPK pathways, and the PKA pathway, are involved in the response to GnRH, as reviewed in (24). We found that a MEKI/II inhibitor, PD0325901, had the greatest relative impact on GnRH-induced ICER expression, with a lesser effect by PKC inhibition and none by CamKII or PKA inhibitors (Figure 1). This finding was validated using a second MEKI/II inhibitor, U0126. In addition, the activity of H89 and KN-93 was confirmed in control experiments, in which H89 inhibited GnRH-stimulated CREB phosphorylation as expected (21), and KN-93 inhibited CAMKII phosphorylation, as reported previously (36). GnRH stimulates the PKC-Raf-MEK-ERK pathway in LβT2 cells (76, 77). Blockade of MEKI/II results in a 95% reduction in GnRH stimulated ERK phosphorylation, compared with 40% as a result of PKC inhibition (76), suggesting that pathways in addition to PKC contribute to ERK activation by GnRH. Indeed, we found that GF109203X only partially inhibited GnRH stimulated ERK phosphorylation in LβT2 cells, whereas ERK phosphorylation by PMA was completely inhibited by GF109203X, a positive control confirming the activity of the inhibitor. These findings support the previous reports which suggest that pathways in addition to PKC contribute to ERK activation by GnRH. Consistent with the effects of MEK inhibition on ICER mRNA levels, GnRH-induced ICER protein expression was also completely blocked by MEKI/II inhibition (Figure 3). Paracrine effects on gonadotropes from other pituitary cell types, as well as from folliculostellate cells, may play a role in studies that use dispersed primary pituitary cells (41, 78–80). The significant increase in ICER induction in mouse primary pituitary cells treated with GnRH, and subsequent blockade of this effect by MEKI/II inhibition (Figure 2), suggests that similar pathways are involved in both the LβT2 cell line and in primary gonadotropes.

A better understanding of the cellular decoding of the pulsatile GnRH signal is essential to elucidate the mechanisms that lead to GnRH pulse frequency dependent differential stimulation of synthesis and secretion of FSH and LH. The data presented from static cultures in this study suggest a role for the MEK/ERK pathway in the induction of ICER synthesis by GnRH in the gonadotrope, but our goal was to determine the contributions of this signaling pathway to the induction of ICER and the subsequent reduction in FSHβ activation at high GnRH pulse frequencies (20). We therefore stimulated perifused LβT2 cells with pulsatile GnRH for 20 hours and, as expected, observed preferential induction of ICER at high (every 30 min) vs low (every 120 min) GnRH pulse frequencies (Figure 4), as detected by real-time RT-qPCR and in agreement with previous end-point PCR data (20). Furthermore, GnRH also stimulated ICER protein levels preferentially at high pulse frequencies (Figure 5). These data, together with our previous studies, suggest that GnRH pulse frequency-dependent signaling results in preferential Fshb transcription by the activation of CREB through phosphorylation at Ser¹³³ at low pulse frequencies (21), and the induction of ICER and functional antagonism of CREB at high pulse frequencies (20, 21).

MEKI/II inhibition had similar effects on both ICER mRNA and protein levels in response to pulsatile GnRH (Figures 4 and 5). At high GnRH pulse frequencies, MEKI/II inhibition completely abrogated ICER induction, whereas at low pulse frequencies, a modest but not significant reduction was observed. A previous study from our group (32), using a similar experimental paradigm, showed that ERKI/II activation occurred after both high and low frequencies of pulsatile GnRH. In this previous study, it was shown that low GnRH pulse frequencies resulted in a more rapid and sustained activation of ERK, compared with high GnRH pulse frequencies, posing another potential mechanism by which GnRH signaling can differentially stimulate FSHβ or LHβ expression in the gonadotrope.

By inhibiting ERK activation and subsequently reducing ICER mRNA and protein levels after exposure to high GnRH pulse frequencies, the mediator of the functional antagonism of CREB has been removed. We therefore predicted that Fshb mRNA levels would increase in response to high frequency GnRH pulses in the presence of the MEKI/II inhibitor. In the absence of any pathway inhibitors, pulsatile GnRH increased Fshb mRNA levels preferentially at the low pulse frequency (Figure 6A), as demonstrated previously (20, 21). Conversely, as expected (21), Lhb mRNA levels were increased preferentially at the high GnRH pulse frequency (Figure 6B). However, somewhat surprisingly, MEKI/II inhibition significantly lowered GnRH-induced Fshb mRNA levels at both pulse frequencies, although the GnRH pulse frequency dependent differential response was indeed lost (Figure 6A). GnRH stimulation of Lhb mRNA levels (Figure 6B) was completely abrogated by MEKI/II inhibition at both high and low GnRH pulse frequencies. Similar effects on Fshb and Lhb transcription have been reported previously (32), GnRH stimulation of −2000/+698 FSHβLUC and −797/+5 LHβLUC reporters were completely abrogated by MEKI/II inhibition, using U0126 as the inhibitor. The effect of inhibition of the ERK signaling pathway by MEKI/II blockade on GnRH-stimulated Lhb mRNA levels was expected. Egr1 is critical for Lhb transcription, with 2 Egr1-binding elements present in the proximal Lhb promoter (81, 82). As recently reviewed in more detail (24), the importance of ERK signaling in Egr1 induction has been shown by several studies. Potential downstream targets of ERK activation and translocation are activators of Fshb transcription such as AP1 proteins and repressors of Fshb transcription such as ski-oncogene-like protein, TG interacting factor 1, and ICER, each of which appear to be stimulated dependent on GnRH pulse frequency. If the inactivation of ERK signaling reduces the activation of stimulators as well as repressors of Fshb transcription, this provides a potential explanation for our observation that in the presence of the MEK inhibitor, Fshb mRNA levels no longer respond differentially to pulsatile GnRH, yet that the stimulation of Fshb mRNA by GnRH is also reduced (Figure 6A).

Our initial prediction that MEKI/II inhibition and the subsequent decrease in ICER induction would result in an augmentation of the induction of Fshb mRNA by high frequency pulsatile GnRH was not borne out by our experimental results. Several studies that address the role of the MAPK pathway in gonadotropin transcription offer potential explanations for the decrease in GnRH-stimulated Fshb transcription observed after MEKI/II inhibition (Figure 6A). Stimulation of the MEK/ERK pathway results in the activation of several transcription factors, including CREB, cFos, cJun, and activating transcription factor 1 (17, 76, 83). Stimulation of FSHβ expression by GnRH is in part driven by activation of cFos and cJun (19), which bind to the corresponding CRE/AP1 element (19) in the murine Fshb gene promoter, and inhibition of the ERK pathway reduces the ability of GnRH to induce Fshb transcription (19). Therefore, we hypothesized that MEKI/II inhibition was likely blocking cFos and cJun induction by pulsatile GnRH, resulting in the observed decrease in Fshb gene expression in response to pulsatile GnRH. Similarly to a recent study (60), we demonstrate that expression of Fos and Jun family members (cFos and cJun) are induced by pulsatile GnRH to a greater extent at the higher GnRH pulse frequency (Figure 7, A and B). MEKI/II inhibition in our perifusion paradigm completely prevented the induction of cFos and cJun expression in response to pulsatile GnRH (Figure 7, A and B). Therefore, despite the loss of ICER induction after MEKI/II inhibition, we conclude that additional inhibition of AP1 protein induction prevents any increase in Fshb transcription at high pulse frequency, compared with cells treated with GnRH alone (ie, without the MEKI/II inhibitor).

Although LβT2 cells express both activin and follistatin (84), 2 autocrine effectors of Fshb transcription, studies conducted with these cells lack the contributions of endocrine factors, such as (but not limited to) inhibin and sex steroids secreted from the gonads. Further validation of our findings will require investigation in a more physiological context, such as genetic animal models, although it was reassuring to observe ICER induction by GnRH in a MEK-dependent manner in dispersed primary mouse pituitary cell culture. Despite the technical challenges involved in isolating sufficient primary gonadotrope cell numbers for validation of our perifusion experiments, the gonadotrope-specific ERKI/II murine mouse model (46) would provide an interesting model for investigation of the effects of silencing the MEK/ERK signaling pathway on ICER induction and Fshb and Lhb expression in response to varying frequencies of pulsatile GnRH. Interestingly, pituitary-specific ERK deletion in these mice resulted in female, but not male, infertility (46). This effect was demonstrated to be the result of reduced Egr1, as well as LHβ, whereas FSHβ was unaffected (46). Nonetheless, the female mice were anovulatory and failed to display normal estrous cyclicity, suggesting impaired responses to the changes in GnRH pulsatility. Furthermore, the expected increase in Fshb expression after ovariectomy was blunted compared with control wild type mice. Although we might predict that loss of ERK in the pituitary would reduce ICER levels and increase Fshb expression, effects on other AP1 proteins, such as cFos and cJun, could again explain the findings. A direct assessment of the pituitary gonadotrope responses to varying GnRH pulse frequencies, in terms of Fshb and Lhb expression, would provide additional insights into the physiologic role of MEK/ERK signaling in the gonadotropin responses to pulsatile GnRH. In addition, studies of the ICER knockout mouse model (85) to better characterize their pituitary gonadotrope phenotypes would further develop our understanding of the role of ICER in female and male reproduction. ICER is transcribed from within the cAMP-responsive modulator (CREM) (a CREB family member with transcriptional repressive activity) gene locus, and mice lacking CREM expression have been reported to have defects in spermatogenesis (86, 87), but these mice do not address ICER-specific functions, because they lack CREM as well as ICER. Therefore, the effects of ICER-specific knock-down on reproductive function in mice remain to be elucidated.

In conclusion, taken together, the data presented in this study demonstrate the critical role of the MEK/ERK pathway as an important mediator of GnRH-stimulated ICER, cFos, cJun, as well as subsequent Fshb and Lhb expression in LβT2 cells. Inhibition of ERK activation results in a loss of GnRH pulse frequency dependent expression of these positive and negative regulators of FSHβ synthesis, suggesting that the MEK/ERK signaling pathway may play a key role in decoding pulsatile GnRH inputs.

Additional material

Supplementary data supplied by authors.