Divergent signaling pathways requiring discrete calcium signals mediate concurrent activation of two mitogen-activated protein kinases by gonadotropin-releasing hormone.

Receptors coupled to heterotrimeric G proteins are linked to activation of mitogen-activated protein kinases (MAPKs) via receptor- and cell-specific mechanisms. We have demonstrated recently that gonadotropin-releasing hormone (GnRH) receptor occupancy results in activation of extracellular signal-regulated kinase (ERK) through a mechanism requiring calcium influx through L-type calcium channels in alphaT3-1 cells and primary rat gonadotropes. Further studies were undertaken to explore the signaling mechanisms by which the GnRH receptor is coupled to activation of another member of the MAPK family, c-Jun N-terminal kinase (JNK). GnRH induces activation of the JNK cascade in a dose-, time-, and receptor-dependent manner in clonal alphaT3-1 cells and primary rat pituitary gonadotrophs. Coexpression of dominant negative Cdc42 and kinase-defective p21-activated kinase 1 and MAPK kinase 7 with JNK and ERK indicated that specific activation of JNK by GnRH appears to involve these signaling molecules. Unlike ERK activation, GnRH-stimulated JNK activity does not require activation of protein kinase C and is not blocked after chelation of extracellular calcium with EGTA. GnRH-induced JNK activity was reduced after treatment with the intracellular calcium chelator BAPTA-AM (1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid acetoxymethyl ester), whereas activation of ERK was not affected. Chelation of intracellular calcium also reduced GnRH-induced activation of JNK in rat pituitary cells in primary culture. GnRH-induced induction and activation of the JNK target c-Jun was inhibited after chelation of intracellular calcium, whereas induction of c-Fos, a known target of ERK, was unaffected. Therefore, although activation of ERK by GnRH requires a specific influx of calcium through L-type calcium channels, JNK activation is independent of extracellular calcium but sensitive to chelation of intracellular calcium. Our results provide novel evidence that GnRH activates two MAPK superfamily members via strikingly divergent signaling pathways with differential sensitivity to activation of protein kinase C and mobilization of discrete pools of calcium.

by various stimuli and are known to play critical roles in the control of multiple cell functions. Three major MAPK family members are known to exist: ERKs (p42 and p44 MAPKs), JNK, and p38 MAPK (1). ERKs are often activated by growth factors and have been shown to regulate growth and differentiation in many cells (2). JNK and p38 are often activated by stress stimuli such as ultraviolet irradiation or osmotic shock and, in many cases, inhibit cell growth or cause apoptosis (3). However, it is becoming increasingly more obvious that these three MAPKs have diverse functions in differentiated cells and that the balance of their concurrent activation may be regulated in a complex manner and instrumental in the control of differentiated cell function. Definition of MAPK signaling pathway architecture and regulation is necessary for examining the consequences of concurrent activation of multiple MAPK superfamily members.
G protein-coupled receptors activate MAPK signaling pathways through mechanisms that vary with specific ligand-receptor interactions, heterotrimeric G protein subtypes, and cellular phenotype (1). The hypothalamic decapeptide GnRH is coupled to concurrent activation of all three family members of the MAPK superfamily. Activation of the Gq/11-coupled receptor for GnRH in the clonal gonadotrope ␣T3-1 cell line or in pituitary cells in primary culture results in stimulation of a complex signaling cascade that includes activation of phospholipase C, production of IP3 and diacylglycerol, and subsequent activation of PKC (4,5). In addition, GnRH receptor occupancy is linked with an increase in intracellular calcium through mobilization of two distinct pools in both ␣T3-1 cells and rat pituitary gonadotrophs (4 -6). Extracellular calcium enters the cell through VGCCs in the plasma membrane while IP3 releases calcium from intracellular stores. The IP3-released calcium has been shown to be the critical signal required for secretion of the gonadotropic hormones, luteinizing hormone, and follicle-stimulating hormone (7,8). Influx of calcium through L-and T-type VGCCs is initiated independently from the IP3-mediated signal, but calcium influx through the plasma membrane is ultimately required for replenishment of intracellular stores (9).
We have demonstrated recently that calcium influx through VGCCs is absolutely required for activation of ERK by the GnRH receptor agonist buserelin in both ␣T3-1 cells and pituitary cells maintained in primary culture (10). Further, stimulation of VGCCs was a sufficient signal for activation of ERK in the absence of hormone. IP3-released calcium did not appear to be involved in the signaling cascade linking the GnRH receptor to ERK activation. Activation of ERK by GnRH in ␣T3-1 cells required PKC (11,12), and our results led to the hypothesis that the GnRH-induced VGCC signal required for activation of ERK may be located downstream of PKC activation (10). The goal of the experiments described here was to investigate possible differences in mechanisms integrating activation of the ERK and JNK pathways induced by GnRH receptor occupancy. We report that the signaling cascade linking the GnRH receptor to JNK may involve Cdc42-, PAK1-and MKK7-like molecules. In contrast to the signaling pathway for activation of ERK, GnRH-induced activation of JNK occurs independently of both PKC activation and extracellular calcium. Activation of JNK, but not ERK, by GnRH was inhibited by chelation of intracellular calcium. Inhibition of GnRH-induced JNK activity by chelation of intracellular calcium was also observed in studies using rat pituitary cells in primary culture. Our results support the conclusion that there is divergence in the signaling pathways coupling the GnRH receptor to multiple MAPKs with selective requirements for classical downstream signaling molecules as well as PKC and pharmacologically distinct calcium signals. Defining specific elements that are critical for activating MAPKs is fundamental to the understanding of GnRHmediated regulation of immediate early and late response genes that control gonadotrope function.

MATERIALS AND METHODS
Cells and Tissue Culture-␣T3-1 cells, an immortalized mouse pituitary cell line of the gonadotrope lineage, were cultured in monolayer in Dulbecco's modified Eagle's medium supplemented with 5% fetal bovine serum and 5% horse serum (Life Technologies, Inc.). Cells were grown to approximately 70% confluence before use in studies. For kinase assays and immunoblot studies, cells were serum starved for 2 h before receiving hormone. Some experiments were carried out using EGTA to chelate extracellular calcium. The ionic concentration of calcium chloride in Dulbecco's modified Eagle's medium is 1.8 mM. The GnRH agonist buserelin ((D-SER(tBU) 6 ,Pro 9 -ethylamide)GnRH) was applied to the cells at 10 nM for various lengths of time. Drugs (nifedipine, PMA, BAPTA-AM, H-89, staurosporine, BIM 1, and BIM 5) were prepared as stock solutions in Me 2 SO, acetone, or ethanol and applied to the cells in Dulbecco's modified Eagle's medium. Cells were never exposed to Ͼ 0.1% Me 2 SO, acetone, or ethanol, and these concentrations of vehicle had no effect on responses of ␣T3-1 cells.
Pituitary cells for primary cultures were taken from the anterior pituitary of adult female Harlan Sprague-Dawley rats. Pituitaries were placed in filter-sterilized dissociation medium consisting of 137 mM NaCl, 25 mM Hepes, 1 mM KCl, 2 mM glucose, pH 7.3. The pituitaries were sliced into small fragments and digested with collagenase type II (1 mg/ml) and hyaluronidase type V (1 mg/ml) for 30 min at 37°C. After digestion, cells were triturated, collected, and placed in Dulbecco's modified Eagle's medium containing 5% fetal bovine serum and 5% horse serum. This procedure was repeated three times, and cells were collected after each trituration. Cells were plated on poly-L-lysinecoated dishes and maintained in culture for 48 h before treatment with hormone.
Antibodies, Immunoprecipitation, Immunoblotting, and Kinase Assays-For immunoprecipitations, cells were treated with drugs for specified time periods and then washed with ice-cold buffer containing 0.15 M NaCl and 10 mM Hepes (pH 7.5). The cells were lysed in a lysis buffer containing 20 mM Tris (pH 8.0), 137 mM NaCl, 10% glycerol, 1% Nonidet P-40, 0.1% SDS, 0.5% deoxycholate, 2 mM EDTA, 5 mM sodium vanadate, 5 mM benzamidine, and 1 mM phenylmethylsulfonyl fluoride on ice for 10 min. The cell lysates were collected, and debris was cleared by centrifugation. For Western blotting, proteins were resolved using SDSpolyacrylamide gel electrophoresis and transferred to polyvinyldiene membrane by electroblotting. Polyclonal (Santa Cruz Biotechnology) and phospho-specific (Promega) antibodies to ERKs were used according to the manufacturers' instructions. Immunostained proteins were visualized using enhanced chemiluminescence reagents (NEN Life Science Products). C-Fos and c-Jun antisera were obtained from Santa Cruz Biotechnology. The c-Fos antibody is specific to p62 c-Fos and, according to the manufacturer, is not cross-reactive with other c-Fos family members. Polyvinylidene membranes were stripped by soaking for 30 min at 55°C in a solution containing 62.5 mM Tris (pH 6.8), 2% SDS, and 100 mM 2-mercaptoethanol. For immunoprecipitation, JNK-1 was immunoprecipitated by adding JNK-1 antibody (0.5 g; Santa Cruz Biotechnology) and 25 l of protein G-and A-agarose beads. Samples were rotated gently for 2 h at 4°C. The beads were washed once in 1 ml of lysis buffer, twice in 1 ml of ice-cold wash buffer containing 20 mM Tris (pH 8.0), 137 mM NaCl, 10% glycerol, 1% Nonidet P-40, 2 mM EDTA, 5 mM sodium vanadate, 5 mM benzamidine, and 1 mM phenylmethylsulfonyl fluoride and once in 0.5 ml of a kinase buffer containing 20 mM Hepes (pH 7.5), 20 mM MgCl 2 , 25 mM ␤-glycerol phosphate, 100 M sodium vanadate, 20 M ATP, and 2 mM dithiothreitol. The reaction mixture (50 l) contained the agarose beads suspended in kinase buffer, [␥-32 P]ATP and substrate GST-ATF2 for JNK assay. Samples were subjected to kinase assay for 30 min at 30°C with frequent mixing. After kinase assay the reaction was stopped with the addition of SDS loading buffer, then samples were boiled for 2 min, resolved by SDSpolyacrylamide gel electrophoresis, and visualized by autoradiography. All of the experiments presented were conducted at least three times with equivalent results.
Plasmids and Transfection Experiments-The expression vector for Gal4 DNA binding domain c-Jun and the luciferase reporter containing 5 Gal4 DNA binding sites upstream of the E1B TATAA box and luciferase coding sequences have been described previously (13). MKK7 was a gift from Lynn Heasley (University of Colorado, Health Sciences Center), PAK1 was a gift from Melanie Cobb (University of Texas, Southwestern), and Cdc42N17 was a gift from Dr. Rick Cerione (Cornell University). All plasmids used in transfection studies were prepared by centrifugation through cesium chloride using standard methods. Prior to all studies, cells were split to fresh medium and cultured to approximately 60 -70% confluence. Transient transfections were accomplished using the calcium phosphate method (Life Technologies, Inc.) for Gal4c-Jun studies, as there was a requirement for adherent cells immediately after transfection. Overexpression of dominant negative molecules was accomplished by electroporation as described previously (13). Briefly, for transient transfection studies with dominant negative and kinase-defective molecules, cells were transfected by electroporation using a single electrical pulse at 220 V and 950 microfarads. Some transfected cells received the calcium chelator BAPTA-AM (20 M) approximately 8 h after electroporation. Cells were collected by scraping 6 h after the final administration of inhibitors, lysed by three freeze-thaw cycles, and luciferase activity was determined as described (13).
Electrophysiology-All whole cell recordings were made at room temperature with the whole cell perforated patch technique (14). Whole cell recordings were performed 1-2 days after cells were plated. Patchclamp recordings were done on cells that were not in contact with other cells and had no cell processes to avoid possible cell-cell coupling artifacts and to maintain good space clamp. Patch clamp electrodes were made with soft capillary glass and adjusted to obtain a tip resistance of approximately 2-4 megohms. Amphotericin B (Sigma) was added from a stock solution to obtain a final concentration of 200 G/ml. Applying a 1-mV square pulse of 10 ms at 10 Hz monitored the electrode tip resistance. Once a high resistance seal (Ͼ5 gigohms) was formed between the recording pipette and the cell, the access resistance was monitored until it reached values less than 30 megohms. Capacitance was monitored before and after each experiment. All voltage clamp protocols were generated and currents recorded using an Axopatch 1-C amplifier and pClamp 5.51 acquisition system (Axon Instruments). The extracellular solution was exchanged continuously during recordings to add/wash out drugs. Bath perfusion was be performed by exchanging the content of the 35-mm culture dish with recording solutions at a rate of ϳ2 ml/min using a gravity flow system. Drugs were added in the extracellular solution. For recording currents through VGCCs, barium was used as the charge carrier and the solutions were, in mM, external: 30 BaCl 2 , 109 N-methyl-D-glucamine, 2.5 KCl, 1 MgCl 2 , 10 Hepes, and 8 mM glucose (pH 7.2 with NaOH) and internal: 120 CsCl, 3 MgCl 2 , and 25 Hepes (pH 7.2 with CsOH). Currents were recorded from cells stepped for 90 ms from a holding potential of Ϫ60 mV to test potentials between Ϫ50 and ϩ40 mV (10-mV increments).
Fluorometry-␣T3-1 cells were treated with trypsin and resuspended in the standard physiological saline solution (see above) at a density of 10 6 cells/ml. Cells were loaded with 1 M indo-1 acetoxymethyl ester (indo-1 AM, purchased from Molecular Probes, Junction City, OR) for 30 min at 37°C in the presence of 0.1% bovine serum albumin. After loading, cells were washed and used within 2 h. 3-ml aliquots of cell suspension were placed in acrylic cuvettes and maintained at 37°C with constant stirring. Indo-1 fluorescence at 405 nm was monitored with a Perkin-Elmer LS-5 fluorescence spectrophotometer. Indo-1 was excited at 355 nm for measurement of free ionized calcium. All of the experiments presented were conducted at least three times with equivalent results.

GnRH-induced Activation of JNK Is Dose-, Time-, and
Receptor-dependent-␣T3-1 cells were treated with multiple doses of buserelin, a specific GnRH receptor agonist, to determine the optimal dose of buserelin for activation of JNK (Fig. 1A). Cell lysates were collected and assayed for JNK activity via an immune complex kinase assay using GST-ATF2 as a substrate. For all subsequent experiments, buserelin was administered at 10 nM. The time course of JNK activation was examined by stimulating ␣T3-1 cells with buserelin for multiple time points ranging from 15 min to 2 h. Buserelin stimulated an increase in JNK activity by 15 min, which persisted for up to 1 h and was decreased by 2 h (Fig. 1B). To provide evidence that GnRH receptor occupancy was required for JNK activation, the cells were pretreated with the specific GnRH receptor antagonist, antide. Antide was applied to ␣T3-1 cells 30 min before and during exposure to buserelin for 5, 15, or 30 min. Buserelininduced JNK activity was blocked by antide ( Fig. 1C), providing direct evidence that the observed buserelin-induced increase in JNK activity was GnRH receptor-mediated. Consistent with results observed in studies using ␣T3-1 cells, stimulation of rat pituitary cells in primary culture with buserelin for 0, 15, or 30 min induced a time-dependent increase in JNK activity (Fig. 1D). This observation suggests that the ␣T3-1 cell model accurately reflects fully differentiated pituitary cells in primary culture.
Cdc42-, PAK1-, and MKK7-like Molecules Are Involved in the Signaling Pathway Linking the GnRH Receptor to JNK-One reported signaling cascade leading to activation of JNK consists of Cdc42, PAK1, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase kinase, and JNK kinase (MKK7) (3,(15)(16)(17). To determine whether the buserelin signal to JNK involves these signaling molecules, ␣T3-1 cells were cotransfected with 1 g of wild type JNK, 1 g of wild type ERK2, and 5 g of either dominant negative Cdc42 (Cdc42N17), truncated PAK1(1-231) molecule, or kinase-defective MKK7. Transfected cells were stimulated with buserelin for 0, 15, or 30 min, lysed, and divided for examination by kinase assay for JNK activity and by Western blot for ERK activation. Overexpression of either Cdc42 or PAK1 reduced buserelin-stimulated JNK activity but had no affect on activation of ERK ( Fig. 2A). Overexpression of dominant negative MKK7 also reduced buserelin-induced JNK activity, whereas ERK activation was not influenced (Fig. 2B). Similar results were observed at higher doses of dominant negative or kinasedefective expression vectors (data not shown), suggesting that the doses used provided maximal responses.
GnRH-stimulated Activation of ERK and JNK Exhibits Differential Sensitivity to PKC Inhibition-␣T3-1 cells express multiple PKC isozymes, including PKC-␣, -⑀, and - (11,18). Previous studies have suggested that activation of both GnRHinduced ERK and JNK in ␣T3-1 cells requires PKC (12,13,19). Initial studies comparing the PKC requirement for buserelininduced ERK and JNK activation were performed using the PKC inhibitor BIM 1 (GF 109203X) and its structurally related but functionally inactive negative control, BIM 5. Cells were pretreated for 30 min with control vehicle (Me 2 SO), 2 M BIM 1, or 2 M BIM 5 before buserelin application for 0, 15, or 30 min. After hormone treatment cells were lysed and analyzed for JNK activity by kinase assay or ERK activation by Western blot. As expected, buserelin-induced ERK activation was reduced in the presence of BIM 1 and essentially unaffected by treatment with BIM 5 (Fig. 3A). Surprisingly, both BIM 1 and the negative control BIM 5 reduced buserelin-induced JNK activity (Fig. 3A). These results suggest that the reduction of JNK activity by the BIM drugs is likely not caused by PKC inhibition but by a nonspecific effect of the drugs. Therefore, contribution of PKC isozymes to JNK activation by GnRH was examined further by treating ␣T3-1 cells with 100 nM PMA for 16 -20 h to deplete diacylglycerol-dependent PKC isozymes. Chronic treatment with PMA results in nearly a complete loss of detectable PKC-␣ and PKC-⑀ isozymes (18). Depletion of PKC isozymes in the current studies resulted in no obvious reduction in the magnitude of JNK activation at 15 and 30 min, whereas buserelin-induced activation of ERKs was blocked completely by PKC depletion at all time points (Fig. 3B). Similar results were obtained after treatment with the PKC inhibitor staurosporine (Fig. 3C) and H-89 at doses previously shown to inhibit PKC activity (20; data not shown). These results indicate that upstream effectors of ERK and JNK diverge at the level of PKC activity. PKC is absolutely required FIG. 1. Buserelin stimulates receptor-mediated JNK activity in ␣T3-1 cells. Panel A, ␣T3-1 cells were treated with doses of buserelin, a GnRH receptor agonist, ranging from 0.1 to 1,000 nM for 15 min before collecting whole cell lysates for immunoprecipitation with JNK1 antibody and analysis by kinase assay. The phosphorylated substrate is GST-ATF2. Panel B, ␣T3-1 cells were treated with 10 nM buserelin for 15, 30, 60, or 120 min before analysis by kinase assay with immunoprecipitated JNK. Panel C, ␣T3-1 cells were exposed to 10 nM buserelin for 5, 15, or 30 min in the presence or absence of 100 nM antide, a specific GnRH receptor antagonist. Antide was applied 30 min before and during buserelin application. Whole cell lysates were collected and examined for JNK activity by immunoprecipitation followed by kinase assay. Panel D, primary cultures of rat anterior pituitary cells were stimulated for 0, 15, or 30 min with 10 nM buserelin. Cells were lysed and immunoprecipitated with antibody to JNK. Imunoprecipitates were analyzed by kinase assay using GST-ATF2 as a substrate. Western blotting with an antibody to JNK was used to demonstrate equivalent protein amounts present in immunoprecipitates (data not shown).
for ERK activation, whereas activation of JNK is not affected remarkably by pharmacological inhibition of PKC inhibition or depletion of isozymes after chronic administration of PMA.
It has been shown that PMA increases the current flow through ␣T3-1 cell VGCCs at a magnitude similar to that of the current increase observed after GnRH treatment (21). One likely scenario is that buserelin-stimulated PKC may mediate an increase in current flow through VGCCs (10) in ␣T3-1 cells, as has been described in other cells (22,23). We demonstrated in a recent report that treatment with the VGCC channel blocker nifedipine inhibits buserelin-induced activation of ERK but not JNK (10), and in the current paper we present evidence indicating that buserelin-induced ERK but not JNK activation requires PKC activity. To gain additional insight into the role of PKC in buserelin-stimulated cells, we performed a direct examination of the buserelin-induced increase in current influx through VGCCs in the absence or presence of the PKC inhibitor staurosporine using the perforated patch method of whole cell recording (Fig. 3D). Barium was used as the charge carrier, and the peak current was observed at Ϫ10 mV. The peak voltageactivated barium current in control cells was Ϫ62.9 Ϯ 7.3 pA, whereas in the presence of 100 nM buserelin the peak current was Ϫ86.1 Ϯ 6.8 pA (n ϭ 10 cells). Our data are similar to results reported in a previous study examining GnRH-stimulated barium current through VGCCs using the standard method of whole cell recording (21). We report that buserelin induced an approximately 20 -25% increase in the barium current through VGCCs. This increase was abolished completely in buserelin-stimulated ␣T3-1 cells pretreated for 30 min with 500 nm staurosporine (n ϭ 4, Fig. 3D). Data shown are from a representative cell. In addition, we have shown that ERK can be activated by the VGCC activator Bay K 8644 in control or PKC-depleted cells, suggesting that PKC action is upstream of calcium influx through VGCCs (10). This evidence supports the hypothesis that PKC functions to mediate the buserelin-induced increase in calcium influx through VGCCs and provides additional insight into the mechanism by which the GnRH   FIG. 2. Overexpression of dominant negative Cdc42 or kinasedefective PAK or MKK7 reduces buserelin-stimulated JNK activity. Panel A, ␣T3-1 cells were cotransfected by electroporation with 1 g of FLAG-JNK, 1 g of ERK2, and either 10 g of control plasmid (pcDNA3), dominant negative Cdc42, or kinase-defective PAK. 18 h later buserelin was administered for 0, 15, or 30 min, and cell lysates were collected and analyzed for JNK activity by immunoprecipitation followed by kinase assay or for ERK activity by Western blotting using an antibody for phospho-ERK. For all of the experiments the blots were stripped and reprobed with ERK antibody demonstrating equal protein amounts in each lane. Panel B, ␣T3-1 cells were cotransfected by electroporation with 1 g of FLAG-JNK, 1 g of ERK2, and either 10 g of control plasmid (pcDNA3) or kinase-defective MKK7. 18 h later buserelin was administered for 0, 15, or 30 min, and cell lysates were collected and analyzed for JNK activity by immunoprecipitation followed by kinase assay or for ERK activity by Western blotting using an antibody for phospho-ERK. For all of the experiments the blots were stripped and reprobed with ERK antibody demonstrating equal protein amounts in each lane.

FIG. 3. Effects of PKC down-regulation on buserelin-stimulated JNK activation.
Panel A, ␣T3-1 cells were treated with the PKC inhibitor BIM 1 (GF 109203X, 2 M) or the structurally related but functionally inactive BIM 5 (2 M) for 30 min before stimulation with buserelin for 0, 15, or 30 min. After treatment, cell lysates were collected and analyzed for JNK activity by immunoprecipitation followed by kinase assay or for ERK activity by Western blotting using an antibody for phospho-ERK. Panel B, ␣T3-1 cells were exposed to chronic (approximately 16 h) treatment with 100 nM PMA or control vehicle (Me 2 SO) before stimulation with 10 nM buserelin for 0, 15, or 30 min. After treatment, cell lysates were collected and analyzed for JNK activity by immunoprecipitation followed by kinase assay or for ERK activity by Western blotting using an antibody for phospho-ERK. Panel C, ␣T3-1 cells were treated with the PKC inhibitor staurosporine (500 nM) for 30 min before stimulation with buserelin for 0, 15, or 30 min. After treatment, cell lysates were collected and analyzed for JNK activity by immunoprecipitation followed by kinase assay or for ERK activity by Western blotting using an antibody for phospho-ERK. For all of the experiments the blots were stripped and reprobed with ERK antibody demonstrating equal protein amounts in each lane (not shown). Panel D, current-voltage relations were constructed for barium currents before and after the addition of 100 nM buserelin to the bath. Cells were stepped from a holding potential of Ϫ50 to ϩ50 mV at 10-mV increments. The current-voltage relation was measured as the average of 30 ms corresponding to the middle portion of the pulse. Cells were either maintained in control solution or pretreated with 500 nM staurosporine for 30 min before recording. Current-voltage relationships are shown for two representative cells.
receptor is coupled to ERK, but not JNK, activation.
GnRH-induced Activation of JNK Is Not Inhibited by Acute Chelation of Extracellular Calcium-Activation of the GnRH receptor results in IP3-mediated increases in intracellular calcium which have been shown to be the primary signal for exocytosis in ␣T3-1 cells and primary gonadotropes (7-9). Calcium influx through the plasma membrane is thought to play a role in maintaining intracellular calcium stores. Previous studies 2 (9) have indicated that ␣T3-1 cells maintained in calciumfree medium for longer than 20 min undergo a time-dependent depletion or rundown of internal IP3-sensitive calcium stores. Therefore, accurate examination of a requirement for extracellular calcium was accomplished by rapidly chelating extracellular calcium and subsequently measuring JNK activity after brief exposures to buserelin. Treatment of cells with 5 mM (Fig.  4) or 15 mM EGTA (data not shown) for 2 min before stimulation with buserelin for 2, 5, or 10 min did not block GnRHinduced JNK activity but inhibited buserelin-induced ERK activation completely. These data demonstrate unequivocally that there is divergence in the signaling pathways coupling the GnRH receptor to activation of ERK and JNK (Fig. 4).
BAPTA-AM Reduces GnRH-stimulated JNK Activity-To investigate further a requirement for intracellular calcium mobilization in buserelin-stimulated JNK activity, ␣T3-1 cells were pretreated with control vehicle (Me 2 SO) or the calcium chelator BAPTA-AM for 30 min before exposure to buserelin for 0, 15, or 30 min. After hormone stimulation, cells were lysed, and cell lysates were examined for JNK activity (kinase assay) or ERK activation (Western blot). Pretreatment with either 10 or 50 M BAPTA-AM greatly reduced buserelin-induced JNK activity but had only a slight effect on ERK activation (Fig. 5A).
Although BAPTA-AM is a known and commonly used chelator of intracellular calcium, it can have varying degrees of effectiveness depending on cell type and experimental conditions. An examination of indo-1 fluorescence observed upon buserelin stimulation after BAPTA-AM treatment suggested that the spike phase of the fluorescence signal, corresponding to IP3-released calcium, was reduced (Fig. 5B). In contrast, the plateau phase, corresponding to calcium entry through VGCCs, was still quite prominent in the BAPTA-loaded cells (Fig. 5B). The buserelin-stimulated indo-1 fluorescence in cells pretreated with BAPTA-AM is greatly reduced in cells that have also been treated with nifedipine (Fig. 5C) effectively chelates calcium released from internal stores by IP3 (spike phase) but is not as effective at chelating calcium entering the cell through VGCCs (plateau phase).
We report above that buserelin induces JNK activity in rat pituitary cells maintained in primary culture. To determine whether a requirement for intracellular calcium mobilization exists for buserelin-induced JNK activation in cultured rat pituitary cells, cells were pretreated with control vehicle (Me 2 SO) or the calcium chelator BAPTA-AM for 30 min before exposure to buserelin for 0, 15, or 30 min. After hormone stimulation, cells were lysed and cell lysates examined for JNK activity (kinase assay). A portion of the immunoprecipitate was reserved for examination of the amount of JNK protein (Western blot). Pretreatment with BAPTA-AM inhibited buserelininduced JNK activity completely (Fig. 5D).
BAPTA-AM Pretreatment Blocks Buserelin-induced Activation of JNK Targets, but ERK Targets Are Unaffected-It has been shown that stimulation of ␣T3-1 cells with GnRH results in increased mRNA and protein for the immediate early genes c-fos and c-jun (10,24). To examine for effects of BAPTA-AM on downstream targets of GnRH-induced MAPKs, ␣T3-1 lysates were examined for the amounts of c-Fos and c-Jun protein after 1 or 2 h of hormone stimulation after a 30-min pretreatment with control vehicle or BAPTA-AM. Buserelin-induced increases in c-Fos protein amounts were unaffected in the BAPTA-AM-treated cells, whereas c-Jun protein amounts and phosphorylation, as measured by retarded electrophoretic mobility shift, were reduced in BAPTA-AM-treated cells compared with control (Fig. 6A). An alternative strategy for examining the sensitivity of c-Jun activation to intracellular calcium was to cotransfect ␣T3-1 cells with a Gal4-c-Jun fusion protein and a Gal4-dependent luciferase reporter and examine buserelinstimulated luciferase activity in the absence and presence of BAPTA-AM. These studies revealed that buserelin-stimulated c-Jun transcriptional activity was dramatically inhibited by pretreatment with BAPTA-AM (Fig. 6B).

DISCUSSION
It is becoming increasingly evident that the specific nature of the way extracellular signals are interpreted within a cell to regulate gene transcription can vary greatly. Outcomes can vary depending upon cell type, the type of stimulus and the duration and intensity of the stimulus, as well as influences from other signaling pathways that may be concurrently activated. The mechanisms by which an extracellular signal uses ubiquitous intracellular signaling molecules to elicit highly specific cellular responses are not well understood. The goal of the present studies was to enhance our knowledge and understanding of the mechanisms linking the GnRH receptor to multiple signaling pathways that influence gene transcription. Such information is critical for understanding how GnRH regulates gonadotrope function and exerts control over key reproductive processes.
GnRH receptor occupancy results in simultaneous activation of multiple MAPK superfamily members (11-13, 18, 19, 25). These observations are consistent with ligand activation of the endothelin B, ␣ 1a -adrenergic, and angiotensin II G proteincoupled receptors (26 -29). Although it is known that buserelin activates JNK in ␣T3-1 cells (25), we believe that data presented here provide the first demonstration of JNK activation coupled to GnRH receptor stimulation in rat pituitary cells in primary culture. Translational studies are essential in that they serve to reinforce the fidelity and appropriateness of the ␣T3-1 cell model for the study of GnRH receptor-linked signal transduction.
Results from the present studies support the conclusion that signaling pathways linking the GnRH receptor to activation of ERK and JNK in ␣T3-1 cells diverge at the level of PKC. Although diacylglycerol-dependent PKC isozymes are absolutely required for ERK and p38 MAPK activation (12,13,18), GnRH-induced JNK activation in ␣T3-1 cells occurs via a PKCindependent pathway. This finding is similar to results reported for the G␣q/11-coupled angiotensin II receptor, which is coupled to JNK in a PKC-independent manner in rat liver epithelial cells as well as hypothalamic and brainstem neurons (29,30). However, angiotensin II receptors are coupled to JNK in a PKC-dependent mechanism in cardiac myocytes (31), demonstrating the variability of signaling pathways in heterologous cells. We cannot exclude the possibility that a phorbol ester-insensitive form of PKC, such as PKC-(32), may play a role in GnRH-induced JNK activation. Examination of this possibility awaits development of potent inhibitors that have specificity for individual PKC isozymes.
We have reported recently that activation of ERK (but not JNK) by GnRH in clonal and primary gonadotropes requires calcium entry though VGCCs (10). It has been demonstrated previously in ␣T3-1 cells that PMA stimulates an increase in current flow though VGCCs which is similar to the current observed after treatment with GnRH (21). In this study we provide direct electrophysiological evidence that the GnRHinduced VGCC signal is blocked when cells have been treated with the PKC inhibitor staurosporine. If PKC is functioning to facilitate the buserelin-induced increase in current flux through VGCCs, and JNK does not require PKC activation, it is not surprising that buserelin-stimulated JNK activity was not reduced by treatment with the VGCC antagonist nifedipine. These data lend support to our conclusion that buserelin-stimulated JNK activity in ␣T3-1 cells requires neither PKC nor calcium entry through VGCCs.
Pharmacological blockade of L-type VGCCs had no influence on buserelin-induced JNK activity (10), and JNK catalytic activity was still observed after chelation of extracellular calcium with EGTA. BAPTA-AM-treated cells exhibited reduced JNK activity but normal ERK activity. An examination of the fluorescence profile of BAPTA-AM-treated cells indicated that the rapid spike phase of the response, corresponding to release of calcium from IP3-gated intracellular stores, was reduced, whereas the VGCC plateau phase was still quite robust. These results, along with those obtained from previous studies examining the calcium requirement for ERK activation, indicate that the GnRH receptor makes use of two different and discrete calcium signals for activation of ERK and JNK. We suggest that because of the discrete nature of these calcium signals, it is likely that they are confined to distinct subcellular compartments.
It is presently not clear how the IP3-released calcium signal influences the signaling pathway linking the GnRH receptor to JNK. Dominant negative PAK1, which we have suggested may be important in linking the GnRH receptor to JNK, markedly reduced JNK activation by angiotensin II in Chinese hamster ovary and COS cells expressing the angiotensin II type 1 receptor (33). Further, angiotensin II-mediated activation of PAK1 and JNK was inhibited by chelation of intracellular calcium with BAPTA-AM (33), suggesting that PAK1 itself or another upstream signaling molecule was calcium-sensitive. Angiotensin II has been shown to stimulate activation of the proline-rich tyrosine kinase in cultured vascular smooth muscle cells via a mechanism that requires release of calcium from internal stores (34). The authors also reported that angiotensin II was associated with proline-rich tyrosine kinase-Src complex formation (34). A recent report by Levi et al. (19) suggested that Src is required for GnRH-induced JNK activity in ␣T3-1 cells. Future experiments will aim at addressing the potential involvement of additional JNK-related signaling molecules as well as investigating the specific sites for calcium regulation for GnRH-stimulated MAPK activity.
Data presented here and in a previous study (10) indicate that the requirement for specific GnRH-stimulated calcium signals affects the downstream targets c-Fos and c-Jun via a MAPK-specific mechanism. Concurrent activation of multiple MAPKs by GnRH could have varying effects on generation of transcriptionally active AP-1 heterodimers formed by the dimerization of c-Fos and c-Jun, possibly enhancing the ability of AP-1 to bind to its target genes and regulate numerous cellular processes. The role and importance of MAPK signaling and AP-1 in the regulation of gonadotrope cell function and GnRH-dependent gene expression have recently been the focus of a detailed study of the GnRH receptor gene (35). Expression of the GnRH receptor gene requires a putatively tissue-specific tripartite enhancer consisting of a steroidogenic factor 1 binding site, an element required for tissue-specific expression, and a consensus AP-1 site. Disruption of the AP-1 site by mutagenesis resulted in a blockade of GnRH inducibility of a GnRH receptor promoter reporter gene (35). Little is known about the ability of GnRH-induced JNK to influence transcription of the GnRH receptor gene. It is possible that through the use of discrete calcium signals and multiple MAPKs the GnRH receptor possesses an exquisitely sensitive mechanism by which the immediate early genes c-fos and c-jun can be regulated as well as multiple late response genes that have an AP-1 site for regulation of transcription, such as the GnRH receptor gene.
Data from the present studies, combined with our recent findings regarding buserelin-induced activation of ERK (10), support the following mechanistic model for activation of ERK and JNK by the GnRH receptor (Fig. 7). The relevance of the model is supported by critical parallel studies done with rat pituitary cells in primary culture. After ligand binding of the GnRH receptor, activation of ERK is dependent upon diacylglycerol-sensitive PKC isozymes and requires calcium influx though plasma membrane VGCCs. We have demonstrated that activation of the ERK signaling pathway involves activation of Raf kinase and does not appear to require release of calcium from intracellular stores. We also suggest that PKC functions to facilitate GnRH-induced calcium flux through VGCCs. In contrast, activation of JNK by the GnRH receptor does not require PKC or calcium entry through VGCCs but appears to require release of calcium from internal stores. JNK activation occurs via a pathway that involves Cdc42-, PAK-, and MKK7like signaling molecules. These data are in agreement with a previously reported role for Cdc42 (19) in the GnRH to JNK cascade and provide new evidence for the involvement of two additional signaling molecules, PAK1 and MKK7, in the GnRH pathway.
The sensitivity of JNK activation to intracellular calcium fluctuations suggests that release of calcium from IP3-gated intracellular stores may have functions in addition to the widely accepted role as being the primary signal for hormone secretion. We suggest that in addition to serving as a trigger for hormone release, GnRH-induced mobilization of intracellular calcium may exert some control over JNK-related nuclear events. Further studies are required to determine how the two different MAPK pathways induced by ligand binding to the GnRH receptor have such selective requirements for distinct calcium signals. In addition, defining the roles of calciumsensitive ERK and JNK activity upon transcription of immediate early and late response genes in gonadotropes will be critical for understanding how GnRH manages reproductive processes. FIG. 7. Proposed model for the mechanism of JNK and ERK activation by the GnRH receptor. GnRH binds to its receptor and initiates an increase in calcium influx through VGCCs, possibly mediated by PKC and an IP3-mediated release of calcium from internal stores. JNK activation requires Cdc42, PAK, and MKK7 and requires calcium release from internal stores, whereas activation of ERK is dependent upon PKC, calcium entry through VGCCs, and Raf activation.