Galpha12 stimulates c-Jun NH2-terminal kinase through the small G proteins Ras and Rac.

The pertussis toxin (PTX) insensitive heterotrimeric G protein G12 has been implicated in mitogenesis and transformation, but its direct effectors remain unknown. To define potential signaling pathways utilized by G12, we expressed an activated mutant of its α subunit, Gα12(Q229L), in HEK293 cells and examined its effects on Ras and mitogen-activated protein kinases (MAPKs). Transient expression of activated Gα12 increased the percentage of Ras in the active, GTP-bound state, stimulated c-Jun NH2-terminal kinase (JNK) activity, and enhanced the transcriptional activity of c-Jun. Dominant negative Ras (N17Ras) inhibited Gα12-mediated JNK activation in NIH3T3 cells but failed to do so in HEK293 cells. In contrast, dominant negative Rac (N17Rac1) inhibited JNK activation by Gα12 in HEK293 cells as well as three other cell lines. In 1321N1 cells, where thrombin stimulates G12-dependent mitogenesis, coexpression of N17Rac1 or a dominant negative mutant of MEKK1 (MEKKΔ(K432M)) inhibits c-Jun/AP-1 sensitive reporter gene expression stimulated by thrombin or Gα12. These data demonstrate that the α subunit of the heterotrimeric G protein G12, like tyrosine kinase growth factor receptors, activates Ras and recruits a signal transduction pathway involving the small GTP-binding protein Rac that leads to JNK activation.

The pertussis toxin (PTX) insensitive heterotrimeric G protein G 12 has been implicated in mitogenesis and transformation, but its direct effectors remain unknown. To define potential signaling pathways utilized by G 12 , we expressed an activated mutant of its ␣ subunit, G␣ 12 12 . These data demonstrate that the ␣ subunit of the heterotrimeric G protein G 12 , like tyrosine kinase growth factor receptors, activates Ras and recruits a signal transduction pathway involving the small GTPbinding protein Rac that leads to JNK activation.

(Q229L), in HEK293 cells and examined its effects on Ras and mitogen-activated protein kinases (MAPKs). Transient expression of activated G␣ 12 increased the percentage of Ras in the active, GTP-bound state, stimulated c-Jun NH 2 -terminal kinase (JNK) activity, and enhanced the transcriptional activity of c-Jun. Dominant negative Ras (N17Ras) inhibited G␣ 12mediated JNK activation in NIH3T3 cells but failed to do so in HEK293 cells. In contrast, dominant negative Rac (N17Rac1) inhibited JNK activation by G␣ 12 in HEK293 cells as well as three other cell lines. In 1321N1 cells, where thrombin stimulates G 12 -dependent mitogenesis, coexpression of N17Rac1 or a dominant negative mutant of MEKK1 (MEKK⌬(K432M)) inhibits c-Jun/AP-1 sensitive reporter gene expression stimulated by thrombin or G␣
The c-Jun NH 2 -terminal kinases (JNKs) 1 , which belong to the mitogen activated family of protein kinases (MAPKs), stimulate the transactivation potential of the c-Jun component of the transcription factor AP-1 by phosphorylating it at residues Ser-63 and Ser-73 (1,2). The JNKs are strongly activated by exposure to UV irradiation and osmotic stress (2)(3)(4). The JNKs are also activated by mitogens including epidermal growth factor and the oncogenic v-Src tyrosine kinase and v-Ras, and by proinflammatory cytokines such as tumor necrosis factor ␣ (2,5,6). The activation of JNK by growth factors and cytoplasmic oncogenes suggests a role for JNK in cellular proliferation or differentiation (7). Consistent with this hypothesis, blockade of the JNK pathway inhibits transformation of NIH3T3 fibroblasts by v-Src (5).
JNK regulation by cytoplasmic oncogenes and growth factors has recently been shown to require activation of the low molecular weight G protein Ras (8) and two Rho subfamily low molecular weight G proteins, Cdc42 and Rac (5,9). While Ras is also an efficient activator of the extracellular signal regulated kinases (ERKs), Rac and Cdc42 are involved only in activation of JNK and the related p38 MAPK (5, 10 -12). It has also been demonstrated that G protein-coupled muscarinic receptors expressed in NIH3T3 and Rat 1a cells can activate JNKs (13,14). A recent report demonstrates that M 1 and M 2 muscarinic acetylcholine receptors (mAChRs) activate JNKs through the G ␤␥ subunit, suggesting this as the mechanism of JNK regulation for other G protein-coupled receptors (15).
Although its effectors are unknown, the pertussis toxin (PTX) insensitive heterotrimeric G protein G 12 has been implicated in cellular transformation and proliferation. Expression of an activated form of G␣ 12 , G␣ 12 (Q229L) (referred to here as activated G␣ 12 ) induces transformation in NIH3T3 and Rat1 cells (16,17). We recently demonstrated by microinjection of antibodies to G␣ 12 that the PTX insensitive DNA synthesis induced by thrombin in 1321N1 astrocytoma cells requires G 12 (18). In addition, transient expression of activated G␣ 12 can dramatically stimulate AP-1-dependent gene expression in a Ras-dependent manner (18). These results suggest that G␣ 12 stimulates a Ras-dependent signaling pathway leading to mitogenesis and AP-1 activation. The studies reported here examine the possibility that the ␣ subunit of G 12 activates Rasand Rac-dependent pathways and thereby stimulates JNK and AP-1 activity.

Transfection
HEK293 cells were plated at 2 ϫ 10 5 cells per 60-mm dish 3 days before transfection and grown at 5% CO 2 in ␣-minimum essential medium supplemented with 10% fetal calf serum and penicillin/streptomycin. Cells were transfected by calcium phosphate coprecipitation (19) using a total of 12 g of DNA. After 48 h, cells were harvested either for immunoblotting, Ras-GTP, kinase, or reporter gene assays.
NIH3T3, HeLa, and COS1 cells were grown on 35-mm dishes in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 1 mM glutamate, and penicillin/streptomycin. Cells were transfected by the LipofectAMINE method (5) using a total of 2 g of DNA.
After 48 h, cells were harvested for immunoblotting or kinase assays.
1321N1 cells were plated at 6 ϫ 10 5 cells/60-mm dish and grown at 10% CO 2 in Dulbecco's modified Eagle's medium supplemented with 5% fetal calf serum and penicillin/streptomycin. Cells were transfected by calcium phosphate coprecipitation (19) using the AP-1-dependent reporter gene 2 ϫ TRE-luciferase and other constructs as described. After 48 h, cells were harvested and results were obtained by analytical luminescence.

Immunoblotting
Verification of G␣ Subunit Expression-48 h after transfection, membranes were isolated from HEK293 cells as described previously (20). Membrane proteins were denatured by boiling in Laemmli buffer and * This work was supported in part by National Institutes of Health Grants GM36927 (to J. H. B.) and HL35018 (to M. K.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ Supported by National Institutes of Health Predoctoral Fellowship GM17277. Work was in partial fulfillment of the Ph.D. degree in the Biomedical Sciences Graduate Program.
§ Supported by National Institutes of Health Postdoctoral Fellowship F32CA65105.
Determination of JNK and ERK Expression-48 h after transfection, cells were harvested in JNK lysis buffer described below. HA-JNK or HA-ERK was immunoprecipitated from 200 g of HEK293 lysate using an anti-HA antibody (Boehringer Mannheim). Immunoprecipitates were washed twice in lysis buffer and twice in kinase buffer (see below). After the final wash, beads were boiled in 1 ϫ Laemmli buffer and resolved on 12% SDS-PAGE and transferred to Immobilon as above. Blots were blocked and then probed with the appropriate specific antibodies (either anti-JNK1 or anti-ERK2, Santa Cruz) and visualized by chemiluminescence.

Determination of Ras-GTP Loading
Cells were transfected by calcium phosphate coprecipitation with wild type Ha-Ras and either vector or activated G␣ subunits. As a positive control for the assay, one plate of cells was transfected with activated Ha-Ras (V12) and processed simultaneously with the experimental conditions. The transfection of wild type Ha-Ras was necessary because of the low levels of endogenous Ha-Ras in HEK293 cells (21). Forty eight hours after transfection, cells were labeled for 3 h with 1 mCi/ml of [ 32 P]orthophosphate in phosphate-free culture media, and the content of Ras-associated nucleotides was assayed as described previously (22). Briefly, Ras was immunoprecipitated using Y13-259 antibody (Santa Cruz), immunoprecipitates were washed extensively, and guanine nucleotides bound to Ras were eluted and separated by thin layer chromatography (Bakerflex PEI cellulose). After autoradiography, GDP and GTP spots were isolated from the plates and quantitated by scintillation counting. The percentage of Ras that was bound to GTP was determined by the following formula: Ras-GTP ϭ (GTP cpm/ (GTP cpm ϩ GDP cpm)).
Protein Kinase Assays-HA-JNK1 and HA-ERK1 were assayed by collecting transiently transfected cells in lysis buffer containing 20 mM Tris, pH 7.6, 250 mM NaCl, 3 mM EDTA, 3 mM EGTA, 20 mM ␤-glycerophosphate, 0.5% (v/v) Nonidet P-40. The following inhibitors were added to the lysis buffer immediately before use: 100 M Na 3 VO 4 , 1 mM p-nitrophenylphosphate, 2 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, and 10 g/ml aprotinin. Kinases were immunoprecipitated in lysis buffer using anti HA antibodies (Boehringer Mannheim). Immunoprecipitates were washed twice in lysis buffer and twice in a kinase buffer containing 20 mM Hepes, pH 7.6, 20 mM ␤-glycerophosphate, 10 mM MgCl 2 . 100 M Na 3 VO 4 , 2 mM dithiothreitol, and 1 mM p-nitrophenylphosphate, were added to kinase buffer before use. HA-JNK1 activity was determined in a 25-l reaction mixture of kinase buffer with 1 M ATP, 5 Ci of [␥-32 P] ATP, and 2 g GST-c-Jun-(1-79) as the substrate for 20 min at 30°C. HA-ERK1 activity was determined similarly, but 2 g of myelin basic protein per sample were used as the substrate, and the reactions were performed for 15 min at 30°C. Phosphorylated substrate was resolved by SDS-PAGE, gels were then dried, and results were quantitated either by AMBIS or PhosphorImager.
Since 1321N1 astrocytoma cells have a very low transfection efficiency, we utilized HEK293 cells to determine directly whether G␣ 12 could activate Ras. In these experiments, we expressed constitutively activated G␣ 12 and examined its ef-fects upon Ras-GTP loading. For comparison, the effects of a similarly activated mutant of G␣ q (R183C), another PTX insensitive G␣ subunit, were also examined. First we verified that expression levels of both G␣ 12 and G␣ q proteins were increased in HEK293 cells following transfection of the activated mutants of these proteins. Both G␣ 12 and G␣ q are efficiently expressed in membranes isolated from transiently transfected HEK293 cells as demonstrated by the increase in ␣ subunit relative to the vector control (Fig. 1A). Next, cells were labeled for 3 h with [ 32 P]orthophosphate, Ras was isolated by immunoprecipitation, and the fraction of GTP in Ras-bound guanine nucleotides was determined. As shown in Fig. 1B, expression of activated G␣ 12 increases the fraction of Ras in the GTP-bound, active state, while expression of activated G␣ q does not. In four experiments, G␣ 12 produced an average 1.7-fold increase in the GTP content of Ras (Fig. 1B).
Possible downstream effectors of Ras include ERK and JNK (1,10,26). We therefore asked whether G␣ 12 can activate these MAPKs. The activated mutants of G␣ 12 and G␣ q were expressed in HEK293 cells along with hemagglutinin-tagged ERK2 (HA-ERK2). We next determined kinase activity and verified by immunoblotting that HA-ERK2 was expressed to similar extents in G␣ q -and G␣ 12 -transfected cells ( Fig. 2A). While transfected v-Raf stimulates HA-ERK2 activity 7-fold (data not shown) demonstrating that the HA-tagged kinase is functional, neither G␣ q nor G␣ 12 significantly stimulate ERK2 activity. In contrast, expression of G␣ 12 increases HA-JNK1 activity by 10-fold as assessed by phosphorylation of a specific substrate GST-c-Jun-(1-79) (Fig. 2B). Activated G␣ q does not significantly increase JNK activity. To demonstrate that effects of G␣ 12 reflect an increase in HA-JNK activity rather than simply increased expression of the epitope-tagged construct, we performed Western blot analysis on kinase assay samples. In some experiments, HA-JNK immunoreactivity was higher in cells cotransfected with G␣ 12 ; however, G␣ 12 consistently increased HA-JNK activity whether or not expression levels were increased. An immunoblot from an experiment in which G␣ 12 strongly stimulated HA-JNK1 activity is presented in Fig. 2B and demonstrates no appreciable change in HA-JNK1 expression.
To confirm that JNK is activated by G␣ 12 and can transduce its signal to a downstream target, we measured JNK-dependent transcription. A reporter system used to assess Jun phosphorylation-dependent transcription has been developed (1). This system utilizes a chimeric transcription factor in which the DNA binding domain of Gal4 is fused to the transactivation domain of c-Jun and a reporter gene containing five Gal4 binding sites upstream of the luciferase gene. When the transactivation domain of c-Jun is phosphorylated, luciferase production increases (1). Reporter activity is increased 15-fold by G␣ 12 demonstrating that G␣ 12 can regulate transcription through c-Jun. More modest stimulation was observed with G␣ q (Fig.  2C). The activation of Jun-dependent transcription by G␣ 12 is consistent with our earlier finding that this G protein induces AP-1-dependent gene expression (18).
Since G␣ 12 stimulates both Ras and JNK activity, and oncogenic Ras is able to stimulate JNK activity in HEK293 cells (data not shown), we hypothesized that G␣ 12 might stimulate JNK through a Ras-dependent pathway. Surprisingly, G␣ 12induced JNK activity in HEK293 cells is not inhibited by cotransfection of either of two dominant negative Ha-Ras mutants, N17Ras or A15Ras (Fig. 3A). Expression of dominant negative Ras in G␣ 12 -transfected HEK293 cells was confirmed by Western blot (data not shown). Moreover, coexpression of N17Ras inhibited G␣ 12 -induced JNK activation in NIH3T3 cells, demonstrating that the dominant negative Ras is functional (Fig. 3B).
Recent work has shown that the JNKs can be activated by two members of the Rho family, Cdc42Hs and Rac, but not by a third member RhoA (5,9). Consistent with this observation, a dominant negative form of RhoA (N19RhoA) failed to inhibit JNK activation by G␣ 12 in HEK293 cells (data not shown). However, coexpression of a dominant interfering Rac mutant, N17Rac1, reduced JNK activation by G␣ 12 by approximately 90% compared to vector control in HEK293 cells (Fig. 4A). Inhibition of G␣ 12 -mediated JNK activation by N17Rac1 was also observed in three other cell lines. As shown in Fig. 4, B-D, N17Rac1 reduced G␣ 12 -mediated JNK activity by 60% in NIH3T3s, 75% in COS1 cells, and 93% in HeLa cells. These data demonstrate that activated G␣ 12 signals to Rac1 and through this signal communicate to the JNK cascade. Expression of a dominant negative mutant of Cdc42, N17Cdc42, was somewhat less effective in HEK293 and COS1 cells (35% and 24% inhibition, respectively), but nevertheless reduced G␣ 12 stimulation of HA-JNK to similar extents as N17Rac1 in 3T3 and HeLa cells (71% and 94%, respectively, data not shown).
In 1321N1 astrocytoma cells, thrombin and G␣ 12 stimulate AP-1-mediated reporter gene (2 ϫ TRE luciferase) expression in a Ras-dependent manner. In addition, thrombin-stimulated mitogenesis in these cells requires G 12 . Since G␣ 12 stimulates JNK activity, we hypothesized that thrombin might stimulate AP-1 gene expression and mitogenesis through JNK. We first determined whether thrombin stimulates JNK activity in 1321N1 cells. In three experiments, thrombin stimulation of 1321N1 cells for 20 min produced a 4.4 Ϯ 0.5 fold stimulation of JNK activity (data not shown). AP-1-dependent gene expression is regulated by JNK (6,27). As shown in Fig. 5A, AP-1 activity induced by thrombin was markedly attenuated by expression of a dominant inhibitory form of Rac, N17Rac1. Expression of a dominant negative form of MEKK1 (MEKK1⌬(K432M)) almost completely inhibits both thrombin and G␣ 12 -stimulated AP-1 gene expression (Fig. 5, A and B). Since both MEKK1 and Rac1 have been reported to preferentially mediate activation of the JNK cascade (5,8,9), these data suggest that the thrombin receptor stimulates JNK through G 12 , Rac, and MEKK1. DISCUSSION The purpose of this work was to examine the signal transduction pathways by which G 12 effects AP-1-dependent gene expression. Since our previous work demonstrated that Ras mediates thrombin and G␣ 12 -induced AP-1 activation, we examined the possibility that G␣ 12 increases the level of Ras in the active state (18,22). We demonstrate here that expression of constitutively activated G␣ 12 stimulates Ras-GTP loading in HEK293 cells, while G␣ q , another PTX-insensitive G␣ subunit that can interact with the thrombin receptor (22,28), does not. To our knowledge, this is the first demonstration that expression of a specific G␣ subunit can increase formation of Ras-GTP.
The mechanism of Ras activation by G␣ 12 is currently unknown. It is possible that expression of activated G␣ 12 activates Ras by stimulating Ras exchange factors or inhibiting Ras-GAPs; alternatively, prolonged expression of G␣ 12 may activate Ras through an indirect mechanism such as inducing the expression of other proteins that regulate Ras. Regardless of the mechanism, these data demonstrate the existence of a pathway by which a heterotrimeric G protein ␣ subunit can communicate with small G proteins of the Ras family. The communication between Ras and G 12 is consistent with previous observations that G␣ 12 -mediated responses are Ras-dependent (18,29) and may explain the ability of G␣ 12 to transform fibroblasts (16,17) and its requirement for thrombin-stimulated mitogenesis (18).
A recent report demonstrated that expression of G ␤1␥2 in-FIG. 5. Thrombin and G␣ 12 (Q229L) require Rac1 and MEKK1 to induce AP-1-regulated gene expression. A, 1321N1 astrocytoma cells were transiently transfected with 4 g of the 2XTRE-luciferase reporter gene and either 4 g of control vector, 4 g of N17Rac1, or 4 g of MEKK1⌬(K432M). Following transfection, cells were stimulated with either thrombin or control vehicle (0.1% fatty acid-free bovine serum albumin) for 48 h. Stimulation was determined relative to control vehicle. B, 1321N1 cells were transfected with 4 g of 2XTRE luciferase and either 4 g of N17Rac1 or 4 g of MEKK1⌬(K432M). Either control vector or 4 g of G␣ 12 (Q229L) was coexpressed for 48 h as the stimulus. Stimulation was determined relative to control vector. The data in each panel represent the mean Ϯ S.E. from two experiments performed in triplicate. In both A and B, total DNA was adjusted to 12 g with control vector as needed. creases JNK activity (15), presumably due to the ability of free G ␤␥ subunits to activate Ras (21). While this appears to be the mechanism of JNK activation by M1 and M2 muscarinic acetylcholine receptors (M1 and M2 mAChRs) (15), our data and that of Prasad et al. (29) demonstrate that expression of a G␣ subunit, G␣ 12 , stimulates JNK providing another pathway by which G protein-coupled receptors may activate JNK. The thrombin receptor, which induces proliferation in several cell types including fibroblasts, vascular smooth muscle cells, and astrocytes (22, 30 -32), can couple to G 12 (20,28), while the M 3 mAChR appears to do so only poorly relative to the thrombin receptor. A combined signal input from both G␣ 12 and G ␤␥ subunits could explain our observation that thrombin activates Ras and induces mitogenesis in 1321N1 cells while the M 3 mAChR, which does not appear to couple to G 12 , cannot (20).
The finding that Ras is required for JNK activation by G␣ 12 in 3T3 cells is in agreement with the observations reported by Prasad et al. (29) in COS1 cells. However, since dominant negative Ras does not affect G␣ 12 -induced JNK activation in HEK293 cells, the requirement for Ras to activate JNK is cell type-specific. In contrast, the low molecular weight G protein Rac is required for JNK activation by G␣ 12 in all the cell lines examined. Furthermore, the observation that both thrombinand G␣ 12 -induced AP-1-dependent gene expression in 1321N1 cells require Rac and MEKK1 activity suggests that the thrombin receptor signals through G␣ 12 , Rac, and MEKK to stimulate JNK and AP-1. It has recently been reported that G␣ 12 affects stress fiber and focal adhesion formation through the low molecular weight G protein Rho (33). Since Rho may be activated subsequent to Ras and Rac activation (24,34), our data are consistent with a pathway in which G␣ 12 activates Rac and Rho.
In addition to regulating the activity of the JNKs and the cytoskeleton, the low molecular weight G proteins may be involved in proliferation. Olson et al. (35) have demonstrated that Rac and Cdc42 are required for Swiss 3T3 cells to progress through the G1 phase of the cell cycle, and that activated forms of Rac and Cdc42 promote DNA synthesis; the effectors responsible for this phenomenon are unknown. The ERK family of MAP kinases are required for differentiation and proliferation in response to several growth factors (36,37), but Rac and Cdc42 stimulate DNA synthesis without activating the ERKs (35). Since both Rac and Cdc42 stimulate JNK activity, it was suggested that the JNKs and not the ERKs mediate the effects of Rac and Cdc42 on cell proliferation (5,35). In support of this hypothesis, JNKs are required for v-Src-induced transformation of NIH3T3 cells (5). Our observations that G␣ 12 stimulates JNK in a Rac-dependent manner without appreciably stimulating ERK also support a model in which G 12 promotes its proliferative and transforming effects through Rac and JNK. We are currently testing such a model.