Dependence of activated Galpha12-induced G1 to S phase cell cycle progression on both Ras/mitogen-activated protein kinase and Ras/Rac1/Jun N-terminal kinase cascades in NIH3T3 fibroblasts.

We evaluated the roles of mitogen-activated protein kinase (MAPK) and Jun N-terminal kinase (JNK) signaling cascades in Gα12-induced G1 to S phase cell cycle progression in NIH3T3(M17) fibroblasts. Transient expression of a constitutively active mutant of Gα12, Gα12(R203C), resulted in a 2-fold increase in the number of bromodeoxyuridine-positive S phase cells over vector control level under serum-deprived conditions. Consistent with the ability of Gα12(R203C) to induce G1/S transition, its expression led to a 2-fold increase in cyclin A promoter activity, which showed a marked synergism with a low concentration of serum, resulting in up to a 15-fold elevation over the basal level. In addition, Gα12(R203C) caused a 2-fold stimulation in E2F-mediated transactivation. Wild type Gα12 showed similar stimulatory effects on cyclin A promoter activity and E2F-mediated transactivation, although of lesser magnitude. We observed a modest but constitutive activation of MAPK in cells transfected with Gα12(R203C), which was abolished by a dominant negative form of Ras. Gα12(R203C) also induced a 3-fold increase in JNK activity, which was abolished by dominant negative forms of either Rac1 or Ras. The expression of dominant negative forms of Ras, MAPK, Rac1, or JNK inhibited Gα12(R203C)-induced increases in bromodeoxyuridine-positive cells. Also, the dominant negative forms of Ras, MAPK, and JNK strongly inhibited Gα12(R203C)-induced stimulation of cyclin A promoter activity. These results demonstrate that both the Ras/MAPK and Ras/Rac1/JNK pathways convey necessary, if not sufficient, mitogenic signals induced by Gα12 activation.

Heterotrimeric G proteins are critical components in transmembrane signaling via heptahelical receptors and regulate the activities of a variety of effector enzymes and ion channels. Heterotrimeric G proteins are composed of three polypeptides, an ␣ subunit, a ␤ subunit, and a ␥ subunit, the latter two of which form a dimer (1). The G␣ subunits are a family of over 20 different proteins that share 45-95% amino acid identity. They have been divided into four classes: G s ␣, G i ␣, G q ␣, and G␣12 (1). The G␣12 class includes G␣12 and G␣13 (2,3) and is ubiquitously expressed in mammalian tissues and cells (4). In contrast to G s , G i , and G q class members for which functional roles have been well established (1), the biological activity of the G␣12 class is not fully clarified. However, accumulating evidence suggests that G␣12 is involved in cell growth and transformation (5). First, mutationally activated G␣12, G␣12(Q229L), stimulated cell proliferation and induced neoplastic transformation in NIH3T3 cells (3,6,7) and Rat-1 cells (8). Second, expression cloning of a transforming gene from a human sarcoma cDNA library resulted in isolation of wild type human G␣12, suggesting that overexpression of wild type G␣12 was sufficient to cause neoplastic transformation in the presence of serum growth factors (3). Third, DNA synthesis stimulated by thrombin and serum, but not basic fibroblast growth factor, was abrogated by microinjection of anti-G␣12 antibody in 1321N1 astrocytoma cells (9).
The molecular mechanisms by which activated G␣12 promotes cell growth and causes transformation have not yet been clearly understood. The expression of G␣12(Q229L) was shown to increase the activity of Na ϩ /H ϩ exchanger in COS-1 cells (10). It was demonstrated for Rat-1 cells stably expressing G␣12(Q229L) that the basal activity of MAPK 1 (extracellular signal-regulated kinase) under serum-starved conditions was not elevated, but MAPK activation induced by epidermal growth factor or serum was potentiated, as compared with cells transfected with an empty vector (8). Very recently, constitutive activation of JNK (stress-activated protein kinase), a new member of the MAPK family, has been demonstrated in G␣12(Q229L)-transformed NIH3T3 cells (11). However, it remains unknown whether activation of either of these signaling pathways contributes to G␣12-elicited cell growth or transformation.
Recent progress in the understanding of mammalian cell cycle regulation has elucidated the principal molecular mechanism of G 1 to S phase progression (12,13). A number of studies demonstrate that the E2F family transcription factors play pivotal roles for entry into the S phase (14). It was reported that microinjection of E2F1 expression plasmid into quiescent REF- 52 fibroblasts is sufficient to induce DNA synthesis (15). In G 0 and early G 1 phases, E2F is complexed with pRb, the product of the retinoblastoma tumor suppressor gene, and the pRb-related proteins, p107 and p130. The E2F complex found during G 0 and early G 1 phases is inactive as a transcrip-tion activator for E2F target genes (12)(13)(14) or may have transcriptional repressor function (16). In middle to late G 1 phase, activation of cyclin-dependent kinases occurs in a temporally ordered manner, and pRb and the pRb-related proteins become progressively phosphorylated by activated cyclin-dependent kinases, resulting in inactivation of their growth-suppressive function (12,13). Consequently, E2F is liberated and transactivates a number of the late G 1 /S phase genes, including cyclin E, cyclin A, B-myb, cdc2, ribonucleotide reductase, thymidylate synthase, DNA polymerase ␣, and E2F1 itself, through consensus E2F binding sites in their promoter regions (14,17). Cyclin A thus induced at the G 1 /S border forms complexes with its catalytic partner Cdk2. Accumulated evidence indicates that the cyclin A-Cdk2 complex mediates initiation and progression of the S phase (12,13). It has also been demonstrated that MAPK is necessary for growth factor-stimulated G 1 /S progression (18). However, it is poorly understood how MAPK activation leads to activation of the critical events in late G 1 , i.e. E2F activation and induction of the E2F target genes that are necessary for G 1 to S progression. Information about how JNK is involved in G 1 to S progression is even more scanty.
In the present study, we observed that the expression of the G␣12 mutant, G␣12(R203C), caused activation of both MAPK and JNK and stimulation of S phase entry in NIH3T3 fibroblasts. The mutation corresponding to G␣12(R203C) in G s ␣, G i ␣2, and G q ␣ renders them constitutively active and oncogenic (19 -22). We evaluated the role of the MAPK and the JNK signaling cascades in G␣12(R203C)-induced G 1 to S phase progression. We further explored the involvement of the Ras and Rho subfamilies of low molecular weight G proteins in G␣12(R203C)-induced activation of MAPK and JNK and tried to elucidate whether E2F-mediated transactivation and cyclin A gene expression, two major events occurring at the G 1 /S boundary, are downstream targets of these signaling molecules. The present results demonstrate that the expression of G␣12(R203C) leads to constitutive activation of MAPK and JNK in a small G protein-dependent manner and that both Ras/MAPK and Ras/Rac1/JNK pathways are indispensable for G␣12(R203C)-evoked activation of G 1 /S gene expression and S phase entry.

EXPERIMENTAL PROCEDURES
Molecular Cloning of Rat G␣12 cDNA-A 219-bp fragment (679 -891 when the A residue of the initiation codon ATG is numbered as 1) of rat G␣12 was obtained from the cDNAs reverse-transcribed from rat liver poly(A) ϩ RNA by PCR amplification using two sets of degenerate oligonucleotide primers corresponding to the two six-amino acid sequences ( 182 DVGGQR and 290 FLNKQD in G s ␣) conserved between the G s ␣ and G i ␣ class members that were designed by Strathmann et al. (24). The sense primer was GTCTAGAGA(C/T)GTC(A/C/G/T)GG(A/C/G/T)GG(A/ G)(A/C)G, and the antisense primer was GGAATTC(A/G)TC(C/T)TT(C/ T)TT(A/G)TT(A/C/G/T)AG(A/G)AA. The conditions for the PCR were 1 min at 94°C, 1.5 min at 37°C, and 2 min at 72°C. A full-length G␣12 cDNA was isolated form a rat brain ZAPII cDNA library (Stratagene) by hybridization screening using the 219-bp PCR fragment as a probe under high stringency condition (at 42°C in the presence of 50% formamide and 0.9 M NaCl). The nucleotide sequence of the cloned cDNA insert was sequenced by the dideoxy chain termination method with Sequenase (U.S. Biochemical Corp.). The nucleotide sequence of rat G␣12 has been deposited in the GenBank TM /EMBL Data Bank with accession number D85760. A constitutively active mutant of G␣12, G␣12(R203C), was created with a site-directed mutagenesis kit (Muta-Gene M13, Bio-Rad) according to the manufacturer's instructions. The mutagenic primer was CATCCTGTTGGCATGCAAGGACACCAAG. The created mutation was confirmed by DNA sequencing. It was previously demonstrated for G s ␣, G i ␣2, and G q ␣ that substitution of the corresponding arginine with cysteine makes the G␣ proteins constitutively active (19 -23 (25). The cells were maintained in DMEM supplemented with 5% iron-enriched calf serum (25) and 200 g/ml Geneticin at 37°C in subconfluent states. pSV-␤gal, the expression plasmid for ␤-galactosidase, was purchased from Promega. pactEF-MAPK and pactEF-MAPK-DN, the expression vectors of Xenopus MAPK and its dominant negative form with Asp 170 to Ala substitution, respectively, were kindly provided by Dr. Okazaki (26) (Kurume University Institute of Life Science, Kurume, Japan). The expression vector for a Myc epitope (EQKLISEEDL)-tagged MAPK (pME18S-Myc MAPK) was created by a PCR-based method (27). The cDNAs of human JNK1 with a Myc epitope tag at its N terminus and human Rac1 were obtained from a human WI-38 fibroblast cDNA library by PCR amplification. The cDNA of dominant negative JNK1 (JNK-DN) with Thr 183 to Ala and Tyr 185 to Phe substitutions (28) was generated by a PCR-based method (27). The cDNAs of Myc epitopetagged MAPK and dominant negative Rac1 (Rac1(N17)) were also prepared by a PCR-based method (27). The nucleotide sequences of the cDNAs obtained by the PCR method were confirmed by sequencing with an ALFred DNA sequencer (Pharmacia Biotech Inc.). The cDNAs were subcloned into the EcoRI site or the BstX1 site of a mammalian expression vector, pME18S (generously provided by Dr. Maruyama at the Institute of Medical Science, University of Tokyo, Tokyo, Japan). A luciferase reporter vector, E2F-Luc, was created as described previously by Slansky et al. (29). The oligonucleotide 5Ј-CTAGCAGCTGCTGC-GATTTCGCGCCAAACTTGACG-3Ј, which contains a Ϫ20 to ϩ9 from the dihydrofolate reductase promoter plus a PvuII site for screening and XhoI and NheI sites at the 5Ј-and 3Ј-ends, was inserted into the vector pGL3basic (Promega). A 2.5-kilobase pair BglII fragment of the cyclin A genomic DNA was isolated from a human leukocyte genomic library (Clontech). A luciferase reporter vector, CycA-Luc, was constructed by ligating the blunted 1.3-kilobase pair HindIII-SmaI fragment of cyclin A genomic DNA (30) to pGL2 basic vector (Promega) at the SmaI site. The bacterial expression plasmid of truncated c-Jun (amino acids 5-89) fused to glutathione S-transferase, pGEX-2T-c-Jun-(5-89), was kindly provided by Dr. A. Kraft (University of Alabama School of Medicine, Birmingham, AL). The plasmids were purified by two cycles of CsCl 2 density gradient centrifugation and introduced into cells by the calcium phosphate precipitation procedure. The day after transfection, the cells were serum-deprived by incubating in DMEM containing 0.2% bovine serum albumin for 24 h. To induce Ras(N17), dexamethasone (5 ϫ 10 Ϫ7 M) was added into media at least 8 h prior to each experiment (25).
Western Blot Analysis-Two or three days after transfection, as indicated in the figure legends, the cells were washed twice with Ca 2ϩand Mg 2ϩ -free Dulbecco's phosphate-buffered saline and lysed in 2 ϫ SDS sample buffer. Gel-loaded sample volumes were adjusted on the basis of protein concentration (5) determined on parallel dishes so that an equal amount of total cellular protein was loaded onto gels per well. Proteins were separated on SDS-10% polyacrylamide gel electrophoresis and transferred onto Immobilon P membranes (Millipore Corp.). Western blot analysis was performed by probing with anti-G␣12 antibody (Santa Cruz) or a mouse monoclonal anti-Myc epitope antibody (9E10), which recognized the N-terminal amino acid sequence (EQKLI-SEEDL) of c-Myc, and respective alkaline phosphatase-conjugated secondary antibody (Zymed).
BrdUrd Incorporation-NIH3T3 cells, seeded at 1 ϫ 10 5 cells/35-mm dish, were co-transfected with pSV-␤gal and various expression plasmids at a total of 3 g of DNA/dish as indicated in the figure legends. The cells were serum-deprived in the presence or absence of dexamethasone. Then 10 mM of BrdUrd (Boehringer Mannheim) was added and further incubated for 17 h in the presence or absence of dexamethasone. The cells were washed with Ca 2ϩ -and Mg 2ϩ -free Dulbecco's phosphate-buffered saline, fixed in 3.7% formaldehyde, and permeabilized in 0.25% Triton X-100. Cells were first incubated with a rabbit polyclonal anti-␤-galactosidase antibody (Cappel) and then with a rhodamine-conjugated goat anti-rabbit IgG antibody (Cappel). After fixation in 3.7% formalin and treatment in 1.5 N HCl (31), BrdUrd was probed sequentially with a mouse monoclonal anti-BrdUrd antibody (Sigma) and a fluorescein isothiocyanate-conjugated rabbit anti-mouse IgG antibody (Zymed). Each incubation was performed according to the manufacturers' recommendations. More than 200 ␤-galactosidase-positive (transfected) cells were examined, and BrdUrd-positive cells were counted under a fluorescent microscope (Olympus, Tokyo, Japan).
Luciferase Assay-NIH3T3(M17) cells (seeded at 5 ϫ 10 4 /well in 12-well plates) were co-transfected with expression plasmids and either CycA-Luc or E2F-Luc (a total of 2 g of DNA/well), serum-deprived for 24 h, and then incubated in DMEM containing 0.2% bovine serum albumin with or without calf serum (0.5%) and dexamethasone for 17 h. Cell lysates were prepared, and luciferase activity was measured with a Lumat LB95001 luminometer (Berthold) using the luciferase assay system (Promega). Protein concentrations of the same samples were determined using Bio-Rad protein assay reagent, and luciferase activity was normalized for protein content.
Measurement of MAPK Activity-NIH3T3(M17) cells in 35-mm dishes were co-transfected with pME18S-Myc-MAPK and either pMT2-G␣12(R203C) or pMT2 empty vector (a total of 3 g of DNA/dish). After incubation for 24 h in DMEM containing 0.2% bovine serum albumin in the presence or absence of dexamethasone (5 ϫ 10 Ϫ7 M), the cells were lysed in a lysis buffer containing 50 mM Tris (pH 8.0), 1 mM EDTA, 150 mM NaCl, 1 mM Na 3 VO 4 , 1 mM phenylmethylsulfonyl fluoride, 0.5% sodium deoxycholate, 0.1% SDS, and 1% NP-40. Myc-tagged MAPK was immunoprecipitated by using 9E10 anti-Myc epitope antibody. MAPK activity associated with the immune complex was assayed in vitro at 30°C for 10 min using myelin basic protein (Sigma) as a substrate as described (32). The reaction was terminated by adding 4 ϫ SDS sample buffer, and samples were analyzed on SDS-15% polyacrylamide gel electrophoresis followed by autoradiography. The radioactivity in the band corresponding to myelin basic protein was determined with a Fuji BAS 2000 Bio-Image Analyzer (Fuji Film Co., Tokyo, Japan).
Measurement of JNK Activity-NIH3T3(M17) cells in 35-mm dishes were co-transfected with pME18S-Myc-JNK and indicated expression plasmids (a total of 3 g/dish). The cells were deprived of serum for 24 h and then an immune complex JNK assay was performed as described by Derijard et al. (33) with a slight modification. Briefly, cells were lysed in an ice-cold lysis buffer (25 mM HEPES (pH 7.5), 1% Triton X-100, 20 mM Tris (pH 7.6), 0.5% Nonidet P-40, 500 mM NaCl, 50 mM NaF, 5 mM EDTA, 3 mM EGTA, 1 mM Na 3 VO 4 , 10 g/ml each of leupeptin and aprotinin, and 1 mM phenylmethylsulfonyl fluoride). A soluble fraction was obtained by centrifugation and precleared with protein A-conjugated Sepharose 4B beads (Pharmacia). The supernatant was incubated with 9E10 antibody and subsequently with rabbit anti-mouse IgG (Zymed)-bound protein A-Sepharose beads. The immunoprecipitates were washed three times with the lysis buffer and twice with a kinase assay buffer (20 mM HEPES (pH 7.6), 20 mM MgCl 2 , 20 mM ␤-glycerophosphate, 20 mM p-nitrophenyl phosphate, 2 mM dithiothreitol, 0.1 mM Na 3 VO 4 ). The pelleted beads were incubated with 30 l of the kinase assay buffer containing 3 g of GST-c-Jun-(5-89), 20 M ATP, and 5 Ci of [␥-32 P]ATP at 30°C for 20 min. The reaction was terminated by adding 10 l of 4 ϫ Laemmli's SDS-sample buffer and boiled. The samples were analyzed by SDS-12% polyacrylamide gel electrophoresis, and the radioactivity in the band corresponding to GST-c-Jun-(5-89) was measured.

Molecular Cloning and Expression of Rat G␣12 cDNA-A
2.1-kilobase pair cDNA clone containing the entire coding region of rat G␣12 was isolated. Sequence analysis of this clone revealed that the predicted amino acid sequence was identical to that of mouse G␣12 except for one residue (341), where serine in mouse G␣12 is changed to glycine in rat G␣12. NIH3T3(M17) cells were transiently transfected with expression plasmids carrying either wild type G␣12 or the constitutively active mutant G␣12(R203C). The expression of either form was confirmed by Western blot analysis using anti-G␣12 antibody (Fig. 1), which detected expression of 45-kDa proteins corresponding to G␣12 and G␣12(R203C) proteins. A trace level of endogenous G␣12 was detected in empty vector-transfected control cells.

G␣12(R203C) Elicits G 1 /S Cell Cycle Progression, E2F-mediated Transactivation, and Cyclin A Promoter Activation-
Previous studies demonstrated that expression of G␣12-(Q229L), another constitutively active mutant with deficient GTPase activity, resulted in an increased thymidine incorporation into DNA (8). To evaluate whether G␣12(R203C) has a similar stimulatory effect on cell cycle progression, cells were transfected with pMT2-G␣12(R203C), pMT2-G␣12, or pMT2 empty vector together with a ␤-galactosidase expression plasmid, and BrdUrd incorporation into nuclei in a ␤-galactosidasepositive cell population was determined (Fig. 2). Under serumdeprived conditions, the expression of G␣12(R203C) resulted in an approximately 2-fold increase in BrdUrd-positive cells over the vector control level. The stimulatory effect of G␣12(R203C) on S phase progression was also evident in the presence of a low concentration of serum (0.5%). The expression of wild type G␣12 tended to increase BrdUrd-positive cells in the presence and absence of serum but without statistical significance.
Recent investigations have elucidated that G 1 to S phase progression requires activation of E2F family transcription factors, a process dependent on cyclin-dependent kinase-mediated phosphorylation and inactivation of pRb family proteins (12)(13)(14). It is also known that S phase entry is associated with transcriptional activation of the cyclin A gene (12,13). To investigate whether or not activation of G␣12 elicits these events, we performed a series of co-transfection experiments using luciferase-reporter plasmids. As shown in Fig. 3 (left), under serum-deprived conditions, transfection of G␣12(R203C) resulted in a 2-fold increase in transactivation of a luciferase gene that is under the control of a consensus E2F binding site in the 5Ј-upstream region. It was of note that transfection of wild type G␣12 also resulted in stimulation of E2F activity, although to a smaller extent. In the presence of a low concentration of serum, the basal luciferase activity was slightly stimulated as compared with serum-deprived cells. Under this condition, wild type G␣12 and G␣12(R203C) also caused stimulation of luciferase activity over the vector control level.  Fig. 3 (right) is G␣12-induced stimulation of cyclin A promoter activity. Transfection of G␣12(R203C) resulted in a weak but significant increase in luciferase activity in the absence of serum stimulation, which was potentiated by a low concentration of serum in a synergistic manner. Again, the wild type G␣12 showed stimulatory effects, although to a lesser extent. These results demonstrate that G␣12(R203)-induced G 1 /S transition is associated with stimulation of E2F activity and up-regulation of cyclin A gene transcription.

G␣12(R203C) Induces Ras-dependent Activation of MAPK-
We evaluated whether G␣12(R203C) activates MAPK in a transient co-transfection assay with a Myc epitope-tagged MAPK (Fig. 4). In cells transfected with G␣12(R203C), MAPK activity showed a 1.5-fold increase over the vector control in the absence of external growth factors. This increase was statistically significant (p Ͻ 0.01) and reproducible. By taking advantage of inducible expression of the dominant negative mutant of Ras(N17) by dexamethasone treatment, we examined whether G␣12(R203C)-induced activation of MAPK is dependent on Ras. When Ras(N17) was induced, MAPK activity in G␣12(R203C)-transfected cells was not significantly different from the vector control level. Western blot analysis using anti-Myc epitope antibody revealed that the expression level of a Myc epitope-tagged MAPK was not changed by co-expression of Ras(N17) and/or G␣12(R203C). The results demonstrate that G␣12(R203C) induces stimulation of MAPK in a Ras-dependent manner.
G␣12(R203C) Induces JNK Activation in a Ras-and Rac1dependent Manner-Very recently Prasad et al. (11) have reported that G␣12(Q229L) activates JNK. We examined whether G␣12(R203C) activated JNK and, if so, whether G␣12(R203C)-induced JNK activation was dependent on the small G proteins, Ras and Rac. As shown in Fig. 5A, co-transfection of G␣12(R203C) and a Myc epitope-tagged JNK resulted in an approximately 2.5-fold increase in JNK activity over the vector control level. G␣12(R203C)-induced activation of JNK was completely abolished when Ras(N17) was expressed. In addition, G␣12(R203C)-induced JNK activation was entirely dependent on Rac1, as demonstrated by the inhibition of JNK activation by the dominant negative Rac1(N17). Western blot analysis using anti-Myc epitope antibody revealed that the expression level of a Myc epitope-tagged JNK was not changed by co-expression of either Rac1(N17) or Ras(N17) (Fig. 5B).

G␣12(R203C)-elicited Activation of Cyclin A Gene Transcription Is Completely Abolished by the Dominant Negative Forms of Ras, MAPK, and JNK-
We investigated whether Ras, MAPK, and JNK are involved in G 1 /S cell cycle progression evoked by G␣12. To this end, we studied the effects of expression of dominant negative forms of these signaling molecules on G␣12(R203C)-induced activation of cyclin A promoter and BrdUrd incorporation into DNA. As shown in Fig. 6A, stimulation of cyclin A promoter activity by activated G␣12 was completely abolished by induced expression of Ras(N17), both in the presence and absence of a low concentration of serum. The dominant negative MAPK also strongly inhibited the stimulatory effect of G␣12(R203C) down to the serum-starved vector control level (Fig. 6B). Similarly, co-expression of the dominant negative JNK potently inhibited G␣12(R203C)-induced cyclin A promoter activation (Fig. 6C).
G␣12(R203C)-mediated S Phase Entry Is Inhibited by the Dominant Negative Forms of Ras, Rac1, MAPK, and JNK-As described in Fig. 2, BrdUrd-positive cells among the ␤-galactosidase-positive population were monitored (Table I). The expression of either dominant negative MAPK or Ras(N17) strongly inhibited basal and G␣12(R203C)-induced BrdUrd incorporation into nuclei. Both dominant negative JNK and Rac1(N17) also inhibited basal and G␣12(R203C)-induced BrdUrd incorporation; however, they were less potent as compared with dominant negative MAPK and Ras(N17). We confirmed the expression of the dominant negative proteins by Western blot analysis of samples from parallel cultures using specific antibodies (data not shown). These results clearly indicate that activation of Ras/MAPK and Ras/Rac1/JNK pathways are both necessary for G␣12(R203C)-induced G 1 to S phase progression. DISCUSSION In the present work we studied the mitogenic signaling in G␣12(R203C)-induced G 1 to S phase cell cycle progression. The mutation corresponding to G␣12(R203C) in G s ␣ renders it constitutively active and oncogenic (gsp) (20). It is also the site of ADP-ribosylation by cholera toxin, which activates it (19). The corresponding mutation in G i ␣2 also activates it (21) and creates the gip2 oncogene (22). Finally, the corresponding muta-  /well), i.e. pGL3-E2F (left) or pGL2-cycA (right). After serum deprivation, the cells were incubated for 17 h in the presence or absence of 0.5% calf serum, and the luciferase activity in cell lysate was measured as described under "Experimental Procedures." The data shown represent means Ϯ S.E. from three separate experiments each performed in triplicate and are expressed as -fold increase over the mean value for vector control in the absence of serum. *, **, and NS, statistically significant differences (**, p Ͻ 0.01; *, p Ͻ 0.05) and statistically not significant difference, respectively, as compared with vector controls. tion in G q ␣ renders it constitutively active (23). As a general mechanism, the mutation causes activation by decreasing the GTPase activity (19,20). In view of these findings, we think it is reasonable to infer that the R203C mutation in G␣12 would also render it constitutively active, particularly since the data are consistent with this conclusion. The present study demonstrated that G␣12(R203C) stimulated G 1 to S phase cell cycle progression in NIH3T3 cells in a manner dependent upon both MAPK and JNK cascades. Consistent with the ability of G␣12(R203C) to promote G 1 to S progression, we observed that the expression of G␣12(R203C) led to stimulations of E2Fmediated transactivation and the cyclin A promoter activity, the latter of which was also dependent upon both MAPK and JNK. In addition, we found that the small G protein Ras was required for G␣12(R203C)-induced activation of MAPK, while both Ras and Rac were required for JNK activation and S phase entry. The results clearly indicate the critically important roles of both Ras/MAPK and Ras/Rac/JNK cascades in activated G␣12-induced G 1 /S cell cycle progression.
Besides well studied growth factor ligands for receptor protein-tyrosine kinases (34), certain agonists for G protein-coupled receptors, including thrombin, lysophosphatidic acid, and gastrin-releasing peptide/bombesin, are capable of stimulating mitogenesis (35)(36)(37). Several previous studies demonstrated that pertussis toxin caused substantial inhibition of thrombinand lysophosphatidic acid-induced mitogenesis in fibroblasts, indicating a role for a heterotrimeric G protein of the G i or G o classes (36 -38). Subsequent studies demonstrated that activation of pertussis toxin-sensitive G proteins by these ligands led to activation of Ras, which was shown to mediate stimulation of DNA synthesis (39 -41). On the other hand, there are also many reports describing that agonists for G protein-coupled receptors stimulate DNA synthesis through pertussis toxininsensitive G proteins (1,42). In addition, more recent studies demonstrated that thrombin and angiotensin II can activate Ras through a pertussis toxin-insensitive G protein in astroglial cells (9,43) and cardiac myocytes (44). A very recent study by Aragay et al. (9) using microinjection of anti-G␣12 antibody indicated that G␣12 functions as a pertussis toxin-insensitive G protein that mediates thrombin-induced, Ras-dependent mitogenesis in human astroglial cells. Consistent with this report, the present results directly document that G␣12 causes stimulation of DNA synthesis in a Ras-dependent manner (Table I).
In view of the ubiquitous expression of G␣12 in various tissues and cell types (4), it is a likely possibility that G␣12 is involved in the pertussis toxin-insensitive, Ras-dependent mitogenesis evoked by G protein-coupled receptor agonists in a variety of cell types.
There is now much evidence for the notion that Ras activates multiple signaling pathways for cell proliferation and differentiation. The Raf/MEK/MAPK cascade is the best characterized example of Ras-dependent signaling. The activation of the MAPK cascade is critically important for growth factor-induced G 1 /S progression, because expression of dominant negative forms of MEK (18) and a synthetic inhibitor of MEK (45) were shown to inhibit growth factor-induced DNA synthesis, and, conversely, expression of a constitutively active MEK induced cell proliferation and transformation (18,46). It is known that activation of MAPK can be brought about by both Ras-dependent and -independent mechanisms in response to various external stimuli (18, 39 -41, 43, 44, 47). The present study demonstrated that the expression of G␣12(R203C) led to activation of MAPK via a Ras-dependent mechanisms (Fig. 4). G␣12(R203C)-induced activation of MAPK is not as potent as that observed with acute growth factor stimulation but is very likely sustained, i.e. still evident 2 days after transfection. A previous study demonstrated that in fibroblasts stably expressing activated G␣12, MAPK activity was not increased under serum-deprived conditions as compared with cells transfected with an empty vector, although epidermal growth factor-stimulated MAPK activation was enhanced in G␣12-transformed cells (8). The reason for the discrepancy between the previous report and the present result is not clear. However, one reason might be that we adopted a transient transfection method with the G␣12(R203C) expression vector and the Myc-tagged MAPK expression vector. This may have allowed for the detection of a small increase in MAPK in the present study. The activation of MAPK in G␣12(R203C)-transfected cells appears to be critically important for G 1 /S cell cycle progression, because the dominant negative MAPK totally abolished G␣12(R203C)-induced DNA synthesis in the present study (Table I). It is poorly understood how the MAPK functions in mitogenic signaling, although the MAPK cascade was shown to be associated with activation of several different transcription factors including the Ets family proteins (48,49). In the present study we demonstrated for the first time that MAPK is involved in activation of cyclin A promoter. It is now established that E2F functions as a crucial cell cycle-specific transcription factor to activate a number of genes (including cyclin A) that control cell cycle progression (14,17). Therefore, G␣12(R203C)-induced activation of E2F is likely an important mechanism underlying G␣12(R203C)-evoked G 1 /S progression. Accumulated evidence indicates that E2F activation is brought about by phosphorylation of pRb and its related proteins by cyclin-dependent kinases (12)(13)(14). In middle to late G 1 , cyclin D-Cdk4 or cyclin D-Cdk6 complex is first activated to initiate phosphorylation of pRb family proteins. With regard to the activation mechanisms of Cdk4 or Cdk6, a very recent study demonstrated that a dominant negative MAPK inhibited epidermal growth factorinduced activation of cyclin D1 promoter (50). Since the induction of cyclin D1 in middle to late G 1 is a prerequisite for activation of Cdk4 and/or Cdk6 (12)(13)(14), which in turn activates E2F, MAPK-mediated cyclin D1 induction may be an upstream mechanism for MAPK-mediated cyclin A induction at the G 1 /S boundary.
Recent evidence suggests that Ras may activate a second signaling pathway for cell proliferation and transformation, which involves Rac, a member of the Rho subfamily of small G proteins. Several studies demonstrated that expression of a dominant negative form of Rac1 blocked transformation by oncogenic Ras, suggesting that Rac functions downstream of Ras (51,52). It was also shown that an activated form of Rac1 markedly synergized with activated Raf to enhance transformation (52). Further, it was demonstrated that microinjection of activated forms of Rac into quiescent fibroblasts stimulated G 1 /S cell cycle progression, whereas expression of a dominant  ). B and C, the cells were transfected with 1 g/well of pGL2-cycA, 0.5 g/well of pMT2-G␣12(R203C), and 0.5 g/well of pactEF-MAPK-DN (B) or pME18S-JNK-DN (C). For control, the respective empty vector was included instead of the expression plasmid. The cyclin A promoter activity was measured, and data were expressed as described in the legend to Fig. 3. , and Rac1(N17) NIH3T3(M17) cells were transfected with a combination of pSV-␤gal, pMT2-G␣12(R203C) and one of the expression plasmids carrying dominant negative forms of MAPK (MAPK-DN) and JNK (JNK-DN), Rac1(N17), or an empty vector. Expression of Ras(N17) was induced by dexamethasone treatment as described under "Experimental Procedures." The cells were incubated with BrdUrd in the presence of 0.5% calf serum for 17 h. Percentage of BrdUrd-positive cells of the ␤-galactosidase-positive cell population was scored as described for Fig. 2 negative form of Rac1 blocked serum-induced DNA synthesis (53). These observations suggest that in addition to the Raf/ MEK/MAPK cascade, the Rac1 signaling cascade is required for Ras-mediated cell proliferation and transformation. In the present study we have shown that G␣12(R203C)-induced G 1 to S progression is also dependent upon Rac (Table I). Recent studies identified JNK as a downstream effector of Rac (54,55). JNK phosphorylates and activates the transcription factors c-Jun and ATF-2, which results in induction of c-Jun itself via two AP-1 sites in its promoter region (28,54,55). We demonstrated by employing a dominant negative form of JNK (Figs. 4 and 5, Table I) that JNK is actually involved in the Rac-dependent mitogenic signaling. Previous studies demonstrated that c-Jun, in conjunction with several related AP-1 proteins, promotes G 1 phase progression and S phase entry (56,57). The important AP-1 target genes implicated in G 1 to S progression include cyclin D1 (50,56). Therefore, G␣12(R203C)-induced, JNK-mediated cyclin A promoter activation (Fig. 4) may be mediated partly by cyclin D1 induction via the AP-1 site. Thus, both MAPK and JNK appear to be involved in sequential induction of cyclins and resultant activation of cyclin-dependent kinases at G 1 and S. Recent data demonstrate that ␤␥ subunits of heterotrimeric G proteins can mediate activation of both MAPK and JNK in a Ras-dependent manner (58,59). However, the present study demonstrated that the ␣ subunit of G␣12 alone is able to activate MAPK and JNK in Ras-and Ras/Rac-dependent manners, respectively. The Ras dependence of G␣12or G␣13induced JNK activation was also recently shown by another group (11). These observations provide compelling evidence that there is a Ras-dependent pathway activated by the ␣ subunits of G12 and G13. The direct effector of G␣12 in terms of Ras activation is presently unknown. Recent studies have demonstrated that G q -coupled receptor agonists activate Ras via activation of nonreceptor tyrosine kinases through a mechanism involving tyrosyl phosphorylation of Shc, a SH2 domaincontaining adaptor protein (44,60). Shc thus phosphorylated at specific tyrosyl residues mediates recruitment of the Ras GDP/ GTP exchange factor mSOS to the plasma membrane (61). Rac1 activation that appears to be required for G␣12-induced JNK activation is probably caused by recruitment and activation of a GDP/GTP exchange factor for Rac1. The activation of Ras might be necessary for the activation of the Rac1 GDP/GTP exchange factor. Further work will be required to understand the G␣12 effector pathway more thoroughly.