Gα12 and Gα13 Stimulate Rho-dependent Tyrosine Phosphorylation of Focal Adhesion Kinase, Paxillin, and p130 Crk-associated Substrate*

We examined whether constitutively active mutants of the Gα proteins Gα12 and Gα13, which together comprise the G12 subfamily of Gα proteins, induce Rho-dependent tyrosine phosphorylation of the focal adhesion proteins p125 focal adhesion kinase, paxillin, and p130 Crk-associated substrate. We report that transient expression of the constitutively active mutants of Gα12 or of Gα13 in human embryonic kidney 293 cells stimulates tyrosine phosphorylation of a set of proteins ofM r of 110,000–130,000, 97,000, and 60,000–70,000. We identified p125 focal adhesion kinase, paxillin, and p130 Crk-associated substrate as prominent tyrosine-phosphorylated proteins in human embryonic kidney 293 cells expressing constitutively active Gα12 and Gα13. In common with the increased tyrosine phosphorylation of these proteins mediated by mitogens acting through heptahelical receptors, the Gα12- and Gα13-mediated increase in tyrosine phosphorylation is blocked by cytochalasin D, which specifically disrupts the actin cytoskeleton, and by the Clostridium botulinum C3 exoenzyme, which ADP-ribosylates and inactivates Rho. Our results support the hypothesis that Gα12 and Gα13activate Rho and suggest that Gα12 and Gα13may mediate the tyrosine phosphorylation of p125 focal adhesion kinase, paxillin, and p130 Crk-associated substrate.

Many mitogens, including bioactive lipids such as lysophosphatidic acid (LPA) 1 and neuropeptides such as bombesin, signal through distinct cell surface proteins that are members of the superfamily of G-protein-coupled receptors (GPCRs) with seven putative transmembrane domains. In Swiss 3T3 cells, a useful model system to analyze the mechanisms of growth factor signaling, an early event in the signal transduction pathways activated by these mitogens is the tyrosine phosphorylation of a common set of focal adhesion proteins (1). Treatment of Swiss 3T3 cells with GPCR agonists including bioactive lipids (2)(3)(4) and mitogenic neuropeptides such as bombesin (5)(6)(7) stimulates tyrosine phosphorylation of proteins with apparent M r of 110,000 -130,000, 97,000, and 60,000 -70,000 on SDS-polyacrylamide gels. The nonreceptor tyrosine kinase p125 fak (8,9) and the adaptor proteins paxillin (10,11) and p130 cas (12), which localize to focal adhesion plaques, have been identified as prominent tyrosine-phosphorylated proteins in agonist-stimulated Swiss 3T3 cells (3,4,(13)(14)(15)(16)(17)(18)(19)(20). The increases in tyrosine phosphorylation of p125 fak , paxillin, and p130 cas are accompanied by profound alterations in the organization of the actin cytoskeleton in Swiss 3T3 cells, leading to the formation of actin stress fibers and the assembly of focal adhesions (3,4,21,22). RhoA, a member of the Ras superfamily of low molecular weight GTPases, has been shown to direct stress fiber formation and focal adhesion assembly in Swiss 3T3 cells (23,24). We proposed that the tyrosine phosphorylation of p125 fak , paxillin, and p130 cas mediated by heterotrimeric G-proteins is downstream of RhoA activation and focal adhesion assembly (1,(25)(26)(27)(28). Signaling through p125 fak , paxillin, and p130 cas has been implicated in the regulation of cell migration, proliferation, and transformation (20, 29 -32).
The immediate mechanism(s) coupling mitogen-induced activation of GPCRs to tyrosine phosphorylation of focal adhesion proteins is not understood. Although many GPCRs couple to the pertussis toxin-sensitive G i , G i does not appear to mediate increases in protein tyrosine phosphorylation (1,33). G␣ 12 and G␣ 13, which together comprise the ubiquitously expressed G 12 subfamily of G␣ proteins, are distantly related to other Gprotein ␣ subunits and are pertussis toxin-insensitive (34,35). There is increasing evidence indicating that G␣ 12 and G␣ 13 are involved in cell migration, proliferation, and transformation. Expression of mutationally activated G␣ 12 stimulates cellular entry into DNA synthesis, proliferation, and malignant transformation in NIH3T3 and Rat-1 cell lines (36 -41) and promotes stress fiber formation and focal adhesion assembly in Swiss 3T3 cells (42). Furthermore, in astrocytoma cells, thrombin-induced stimulation of DNA synthesis was prevented by the microinjection of anti-G␣ 12 antibody (43). Gene disruption experiments have implicated G␣ 13 in the regulation of cell migration (44). The downstream targets through which G␣ 12 and G␣ 13 induce these effects have not been identified, although Ras-, Rac-, Rho-, and Cdc42-dependent pathways leading to cytoskeletal reorganization and to the activation of mitogen-activated protein kinase, Jun N-terminal kinase, and the Na ϩ -H ϩ exchanger have been implicated (40,(45)(46)(47)(48). However, it is not known whether activation of G␣ 12 and/or G␣ 13 can also promote the tyrosine phosphorylation of the nonreceptor protein tyrosine kinase p125 fak and the adaptor proteins paxillin and p130 cas .
To examine whether the G 12 subfamily of heterotrimeric G-proteins induce protein tyrosine phosphorylation, we have transiently transfected human embryonic kidney HEK 293 cells with expression vectors encoding constitutively active G␣ 12 and G␣ 13 proteins and determined the effect of the expression of these activated G␣ subunits on the tyrosine phos-* 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 1 The abbreviations used are: LPA, lysophosphatidic acid; anti-Tyr(P), anti-phosphotyrosine; C3 toxin, C. botulinum C3 exoenzyme; DMEM, Dulbecco's modified Eagle's medium; G␣ 12 QL, G␣ 13 QL, G␣ i-2QL, and G␣ q QL, constitutively active mutant G␣ 12 -Q229L, G␣ 13 -Q226L, G␣ i-2 -Q205L, and G␣ q -Q209L subunits, respectively; G-protein, guanine nucleotide regulatory protein; GPCR, G-protein-coupled receptors; mAb, monoclonal antibody; p125 fak , p125 focal adhesion kinase; p130 cas , p130 Crk-associated substrate; PKC, protein kinase C; PIPES, piperazine-N,NЈ-bis(2-ethanesulfonic acid) ; EF, elongation factor. phorylation of p125 fak , paxillin, and p130 cas . Our results demonstrate that expression of active G␣ 12 and G␣ 13 in HEK 293 cells induces tyrosine phosphorylation of these focal adhesion proteins through a pathway that requires the integrity of the actin cytoskeleton and functional Rho.
Cell Culture and Transfection-Cultures of transformed HEK 293 cells were maintained in DMEM supplemented with 10% fetal bovine serum in a humidified atmosphere containing 10% CO 2 at 37°C. HEK 293 cells were transfected with the various plasmids by calcium phosphate precipitation (52) or with LipofectAMINE TM (Life Technologies, Inc.). The conditions for transient transfection of HEK 293 cells using calcium phosphate precipitation and LipofectAMINE were each optimized to yield peak efficiency of transfection using an expression plasmid encoding ␤-galactosidase under the control of the human polypeptide chain elongation factor 1␣ (EF-1␣) promoter (pEFlacZ, a gift from R. Treisman). Both transfection methods yielded greater than a 50% transfection efficiency optimally, and we were unable to match this transfection efficiency with Lipofectin or DEAE-dextran-mediated transfection methods. Briefly, HEK 293 cells were plated at a density of 1.2 ϫ 10 6 cells/60-mm Nunc dish and transfected the following day. For calcium phosphate transfections, 12.5 g of plasmid DNA (precipitated in 0.25 ml of 140 mM NaCl, 0.75 mM Na 2 HPO 4 , 125 mM CaCl 2 , 20 mM PIPES, pH 6.95) was added per dish containing 2.25 ml of Optimem-1 medium, and the dishes were incubated for 4 h at 5% CO 2 . For Lipo-fectAMINE transfections, HEK 293 cells were transfected with 3 g of plasmid DNA and 25 l of LipofectAMINE in 3 ml of Optimem-1/dish for 18 h at 5% CO 2 according to the manufacturer's specifications. In the cotransfection experiments, total amounts of DNA were kept constant, and equal amounts of G␣ and of C3 expression constructs (or vector without cDNA insert) were mixed. After transfection with calcium phosphate or LipofectAMINE, the cells were washed and maintained in DMEM supplemented with 10% fetal bovine serum. All cultures were used for experimental purposes 4 days after plating.
Immunoprecipitation-Cultures (one 60-mm dish/condition) were washed twice with DMEM and equilibrated in 5 ml of the same medium at 37°C for 1-2 h. Some dishes were treated with inhibitors during this equilibration period or with growth factors for 10 min at the end of this period as indicated. Cultures were lysed in 1.4 ml of ice-cold lysis buffer A (10 mM Tris-HCl, pH 7.6, 5 mM EDTA, 50 mM NaCl, 30 mM sodium pyrophosphate, 50 mM NaF, 1% Triton X-100) supplemented with 2 mM Na 3 VO 4 , 1 mM 4-(aminoethyl)-benzenesulfonyl fluoride and 0.1% SDS. Lysates were clarified by centrifugation at 20,800 ϫ g for 10 min at 4°C. In some experiments, the cultures were washed and lysed as described above, with the exception that DMEM without phenol red was used, and the lysates were adjusted to 1 mg protein/ml after protein determination using the bicinchoninic acid protein assay (Pierce). 1 ml of the lysate was immunoprecipitated at 4°C for 3-4 h with anti-mouse IgG-agarose-linked mAb directed against phosphotyrosine (4G10 or PY72), p125 fak (2A7), paxillin, or p130 cas or with Protein A-agaroselinked rabbit antisera directed against p125 fak .
Western Blot Analysis-Immunoprecipitates were washed three times by centrifugation with lysis buffer A supplemented with 1 mM Na 3 VO 4 and extracted in 1ϫ Laemmli sample buffer containing 1 mM EDTA for 10 min at 100°C. The solubilized proteins were fractionated by discontinuous SDS-polyacrylamide gel electrophoresis (53) under reducing conditions on an 8% acrylamide resolving gel. Proteins were transferred onto polyvinylidene fluoride membranes (Immobilon TM -P) at 4°C for 2 h at 35 V and 2 h at 70 V in 48 mM Tris, 386 mM glycine, 0.1% SDS, and 20% methanol. The membranes were then blocked with 1% polyvinylpyrrolidone, 0.05% Tween 20, 0.02% sodium azide in phosphate-buffered saline, pH 7.4 (blocking buffer) (54) and incubated for 2 h with anti-phosphotyrosine (anti-Tyr(P)) mAb (4G10, 1 g/ml) in blocking buffer. For some experiments, cultures were immunoprecipi-tated with anti-paxillin and with anti-p130 cas mAb in duplicate, and parallel Western blots were probed with anti-Tyr(P) antibody and with the immunoprecipitating antibody to verify that equivalent amounts of protein were immunoprecipitated. Bound antibodies were visualized by the binding of 125 I-labeled anti-mouse IgG (0.1 Ci/ml) followed by autoradiography. After autoradiographic detection of bound anti-Tyr(P), membranes from anti-p125 fak immunoprecipitates were reprobed with rabbit antiserum directed against p125 fak (1:500) and horseradish peroxidase-conjugated donkey antibodies to rabbit IgG (1:5,000). Immunoreactive p125 fak was detected by enhanced chemiluminescence Western blotting ECL TM reagents (Amersham Pharmacia Biotech).
For analysis of G␣ subunit expression, identical samples of 50 l (50 g of protein) of the excess clarified lysate were diluted with 50 l of 2ϫ Laemmli sample buffer containing 2 mM EDTA and heated for 10 min at 100°C. The proteins were fractionated by SDS-polyacrylamide gel electrophoresis on 10% acrylamide resolving gels. Proteins were transferred onto Immobilon TM -P polyvinylidene fluoride membranes at 4°C for 3 h at 400 mA in 25 mM Tris, 192 mM glycine, and 20% methanol. After blocking the membranes as above, immunoreactive G␣ subunits were visualized with rabbit antisera directed against G␣ 12 , G␣ 13, G␣ i -1/G␣ i-2, or G␣ q at 1:500 -1:1,000 dilution, 125 I-labeled Protein A (0.1 Ci/ml), and autoradiography. The G␣ 13 antiserum was raised against the synthetic peptide CLHDNLKQLMLQ (which corresponds to the carboxyl-terminal peptide 367-377 of murine G␣ 13 with an N-terminal cysteine added for coupling) cross-linked to keyhole limpet hemocyanin with the heterobifunctional reagent sulfosuccinimidyl 4-(p-maleimidophenyl) butyrate as described by Offermanns et al. (55). The G␣ q antiserum was raised against a synthetic peptide (corresponding to amino acid residues 119 -133 of murine G␣ q ) cross-linked to keyhole limpet hemocyanin with glutaraldehyde. Reproducibility-Although both the calcium phosphate and the Li-pofectAMINE transfection methods appeared to yield little variation in levels of expression of G␣ proteins between dishes in a single transfection, both methods yielded variation in levels of expression of G␣ proteins between experiments. Although this variation in the levels of expression of the G␣ proteins was reflected in the variation in the levels of increased protein tyrosine phosphorylation, the reported experiments were repeated at least two times with results equivalent to those illustrated.
Materials-LipofectAMINE TM and Optimem-1 were obtained from Life Technologies, Inc., and the bicinchoninic acid BCA* TM protein assay reagent was obtained from Pierce. The anti-Tyr(P) mAb 4G10 and the anti-p125 fak mAb 2A7 were obtained from Upstate Biotechnology, Inc.; mAbs directed against paxillin and p130 cas were obtained from Transduction Laboratories; the rabbit anti-p125 fak antiserum C-20 and rabbit anti-G␣ 12 were obtained from Santa Cruz Biotechnology, Inc.; and the rabbit anti-G␣ i-1 /G␣ i-2 was obtained from Calbiochem. 125 I-Labeled sheep anti-mouse IgG (100 Ci/ml), 125 I-labeled protein A (100 Ci/ml), horseradish peroxidase-conjugated donkey anti-rabbit Ig, and the ECL TM Western blotting detection reagents were from Amersham International. Polyvinylidene fluoride Immobilon TM -P Western blotting membranes were obtained from Millipore. Anti-mouse IgG-linked agarose, Na 3 VO 4 , thrombin, 1-oleoyl-LPA, phorbol 12,13-dibutyrate, epidermal growth factor, cytochalasin D, pertussis toxin, polyvinylpyrrolidone 40, and thapsigargin were obtained from Sigma. GF109203X and U73122 were obtained from LC Laboratories. The synthetic G␣ 13 peptide CLHDNLKQLMLQ was obtained from Research Genetics. The anti-Tyr(P) mAb PY72 was obtained from the Hybridoma Development Unit, Imperial Cancer Research Fund. All other reagents were of the highest available grade from standard commercial sources.

Thrombin-mediated Tyrosine Phosphorylation of Focal Adhesion Proteins in HEK 293
Cells-Elements of the signal transduction pathways mediating the tyrosine phosphorylation of focal adhesion proteins downstream of GPCRs have been well characterized in Swiss 3T3 cells. We were interested in using a transient transfection system to analyze the ability of the G 12 subfamily of heterotrimeric G-proteins to induce tyrosine phosphorylation of focal adhesion proteins. We elected to use HEK 293 cells for these studies because these cells can be efficiently transfected. However, signal transduction pathways exhibit a degree of heterogeneity in different cells types. Consequently, we initially examined tyrosine phosphorylation induced by ligands of GPCRs in HEK 293 cells.
As thrombin causes GPCR-mediated activation of G␣ 12 and G␣ 13 (43,44,55), we examined the effects of thrombin on tyrosine phosphorylation in HEK 293 cells. Lysates from these cells treated with or without thrombin were immunoprecipitated with anti-Tyr(P) mAb PY72 or mAbs directed against p125 fak , paxillin, and p130 cas . The immunoprecipitates were then subjected to Western blot analysis with the anti-Tyr(P) mAb 4G10. As illustrated by Fig. 1 (left panel), treatment of HEK 293 cells with thrombin resulted in increased tyrosine phosphorylation of proteins with apparent M r of 110,000 -130,000, 97,000, and 60,000 -70,000. This pattern is identical to that elicited by GPCR agonists in Swiss 3T3 cells (3-6, 13, 14). Fig. 1 (right panels) shows that thrombin stimulated tyrosine phosphorylation of p125 fak , paxillin, and p130 cas in HEK 293 cells. p125 fak and paxillin exhibit a greater increase in tyrosine phosphorylation in response to thrombin than does p130 cas . Western blot analysis confirmed that similar amounts of p125 fak , paxillin, and p130 cas protein were immunoprecipitated (data not shown). Stimulation of HEK 293 cells with LPA resulted in an identical pattern of protein tyrosine phosphorylation and in increased tyrosine phosphorylation of p125 fak , paxillin, and p130 cas (data not shown).
Thrombin-induced Tyrosine Phosphorylation of p125 fak , Paxillin, and p130 cas Requires an Intact Cytoskeleton but Is Independent of PKC, Ca 2ϩ , and Pertussis Toxin-sensitive G-proteins-Agonist-mediated tyrosine phosphorylation of focal adhesion proteins in Swiss 3T3 cells occurs through a PKC-, Ca 2ϩ -, and Gi-independent pathway that is critically dependent on the integrity of the actin cytoskeleton and of functional Rho (1,3,4,14,15,25,26). Direct PKC activation by phorbol esters can also lead to tyrosine phosphorylation of these proteins in Swiss 3T3 cells (14,15,20). As illustrated by Fig. 2  (upper left panel), the phorbol ester phorbol 12,13-dibutyrate also induced tyrosine phosphorylation of p125 fak , paxillin, and p130 cas in HEK 293 cells. The PKC inhibitor GF109203X (56,57) prevented the tyrosine phosphorylation induced by phorbol 12,13-dibutyrate but did not inhibit the thrombin-induced tyrosine phosphorylation of p125 fak , paxillin, or p130 cas in HEK 293 cells, suggesting that PKC activation does not mediate agonist-induced tyrosine phosphorylation of focal adhesion proteins in HEK 293 cells. Similarly, treatment with pertussis toxin to inactivate Gi, thapsigargin to deplete the endoplasmic reticulum Ca 2ϩ pool (58), or U73122 to inhibit phospholipase C-mediated phosphoinositide breakdown (59) did not affect thrombin-induced tyrosine phosphorylation of focal adhesion proteins in HEK 293 cells (Fig. 2).
To determine whether the thrombin-induced increase in the tyrosine phosphorylation of p125 fak , paxillin, and p130 cas is dependent upon an intact actin cytoskeleton, we tested the effect of cytochalasin D on tyrosine phosphorylation in HEK 293 cells stimulated with thrombin. As shown in Fig. 2 (upper middle panels), cytochalasin D pretreatment inhibited the thrombin-induced tyrosine phosphorylation of p125 fak , paxillin, and p130 cas in HEK 293 cells, indicating that an intact actin cytoskeleton is required. These results indicate that the signal transduction pathway leading to the tyrosine phophorylation of focal adhesion proteins in HEK 293 cells is very similar to that extensively studied in Swiss 3T3 cells. Thus, HEK 293 cells appear to be an appropriate cellular system to test the effects of mutationally active G␣ 12 and G␣ 13 .
G␣ 12 QL and G␣ 13 QL Induce Tyrosine Phosphorylation of p125 fak , Paxillin, and p130 cas in HEK 293 Cells-Mutations in the catalytic domain of G␣ subunits that inhibit their intrinsic GTPase activity are known to convert these proteins into con- stitutively active ␣ -subunits (60). To examine the effects of G␣ subunits on tyrosine phosphorylation, HEK 293 cells were transiently transfected with expression plasmids encoding the constitutively active G␣ mutants G␣ 12 QL, G␣ 13 QL, G␣ i-2 QL, and G␣ q QL, which are deficient in GTPase activity (37,49,50). Given that pertussis toxin does not interfere with agoniststimulated tyrosine phosphorylation of focal adhesion proteins, G␣ i was tested as a negative control. Conversely, because PKC activation leads to tyrosine phosphorylation of p125 fak , paxillin, and p130 cas , G␣ q , which stimulates phospholipase C-mediated phosphoinositide hydrolysis and thereby PKC (61-63), was included as a positive control. The extracts of transfected cells were immunoprecipitated with the anti-Tyr(P) mAb PY72, and the immunoprecipitates were analyzed by SDS-polyacrylamide gel electrophoresis and Western blotting with the anti-Tyr(P) mAb 4G10. As illustrated by Fig. 3 (upper panel), HEK 293 cells overexpressing constitutively active mutant G␣ 12 QL, G␣ 13 QL, or G␣ q QL proteins exhibited increased tyrosine phosphorylation of a set of proteins with apparent M r of 110,000 -130,000, 97,000, and 60,000 -70,000. In contrast, transient transfection of HEK 293 cells with activated G␣ i-2 QL expression plasmids did not increase tyrosine phosphorylation. Overexpression of wild-type G␣ 12 and G␣ 13 in HEK 293 cells also did not induce tyrosine phosphorylation (data not shown), suggesting that the effects of G␣ 12 QL, G␣ 13 QL, and G␣ q QL were specific for the activated state of these G␣ subunits rather than due to the modulation of ␤␥ subunit availability. No stimulation of protein tyrosine phosphorylation was observed after treatment of HEK 293 cells with medium conditioned by cells transfected transiently with activated G␣12QL or G␣13QL expression plasmids (data not shown), suggesting that the increased tyrosine phosphorylation was not mediated by secreted factors.
The pattern of increased total tyrosine phosphorylation in response to the constitutively active mutants G␣ 12 QL, G␣ 13 QL, or G␣ q QL was identical to that induced by treatment of HEK 293 cells with thrombin, LPA, or phorbol 12,13-dibutyrate. As p125 fak and paxillin are tyrosine-phosphorylated in response to these stimuli in HEK 293 cells, we investigated whether these same proteins are tyrosine phosphorylated in response to activated G␣ 12 , G␣ 13, and G␣ q in HEK 293 cells. As illustrated by Fig. 3 (middle panel), HEK 293 cells transfected with G␣ 12 QL, G␣ 13 QL, or G␣ q QL subunits exhibited increased tyrosine phosphorylation of p125 fak and paxillin. Western blot analysis confirmed that similar amounts of p125 fak were immunoprecipitated under all conditions. As shown in Fig. 3 (lower panel), we confirmed that the cells transfected with the G␣ 12 QL, G␣ 13 QL, G␣ i-2 QL, or G␣ q QL expression plasmids were overexpressing these G␣ subunits.
Tyrosine Phosphorylation in Response to G␣ 12 QL and G␣ 13 QL Requires an Intact Actin Cytoskeleton-Agonist-stimulated tyrosine phosphorylation of p125 fak , paxillin, and p130 cas is inhibited by treatment of Swiss 3T3 cells or HEK 293 cells with cytochalasin D, which selectively disrupts the actin cytoskeleton (3,14,15,64). To determine whether the increase in the tyrosine phosphorylation of p125 fak and paxillin induced by activated G␣ 12 or G␣ 13 is dependent upon an intact actin cytoskeleton, we tested the effect of cytochalasin D on tyrosine phosphorylation in HEK 293 cells transfected transiently with the G␣ 12 QL or G␣ 13 QL expression vectors. As shown in Fig. 4 (upper panels), cytochalasin D inhibited the G␣ 12 QL-and G␣ 13 QL-induced tyrosine phosphorylation of p125 fak and paxillin. Western blotting of total cell lysates with antisera to the G␣ subunits demonstrated that cytochalasin D treatment did not alter expression levels of the transfected G␣ subunits (Fig.  4, lower panels). Cytochalasin D also inhibited G␣ q QL-induced tyrosine phosphorylation (data not shown).
␣ 12 QL and G␣ 13 QL Induce Tyrosine Phosphorylation of p125 fak , Paxillin, and p130 cas in a Rho-dependent Manner-The C. botulinum C3 toxin, which specifically ADP-ribosylates RhoA, RhoB, and RhoC, resulting in the functional inhibition of these GTP-binding proteins, can be used to determine whether a cellular response requires functional Rho (65). This approach has been used to demonstrate that in Swiss-3T3 cells, bombesin-, LPA-and toxin-stimulated tyrosine phosphorylation of p125 fak , paxillin, and p130 cas requires functional Rho (25)(26)(27).
To determine whether the G␣ 12 QL-or G␣ 13 QL-induced increases in the tyrosine phosphorylation of p125 fak and paxillin are dependent upon Rho, we cotransfected HEK 293 cells with either the G␣ 12 QL or the G␣ 13 QL expression vector and with a C3 toxin expression vector. We then immunoprecipitated lysates from the transfected cells with the anti-Tyr(P) mAb PY 72, and the immunoprecipitates were analyzed by Western   FIG. 3. G␣ 12 QL, G␣ 13 QL, and G␣ q QL induce tyrosine phosphorylation of p125 fak and paxillin in HEK 293 cells. HEK 293 cells were transfected using lipofectAMINE with the following vectors (G␣): pcDNA1 without G␣ insert (Ϫ) or pcDNA1 encoding constitutively active mutant G␣ 12 QL (12*), constitutively active mutant G␣ 13 QL (13*), constitutively active mutant G␣ i-2 QL (i*), or constitutively active mutant G␣ q QL (q*). Three days after transfection, the cells were washed and incubated for 1 h in DMEM before treatment for 10 min with 1 NIH unit thrombin/ml (ϩ) or solvent (Ϫ), and the cells were lysed. Upper Panel, total tyrosine phosphorylation was analyzed by immunoprecipitation (IP) of the cell lysates with anti-Tyr(P) mAb PY72 (PY) and Western blotting (Blot) the immunoprecipitates with anti-Tyr(P) mAb 4G10 (PY). The positions of the migration of the molecular weight markers (in kDa) are indicated to the left. Middle panels, tyrosine phosphorylation of individual proteins was also analyzed by immunoprecipitation (IP) with rabbit antiserum directed against p125 fak (FAK) or mAb directed against paxillin (PAX) and Western blotting (Blot) the immunoprecipitates with anti-Tyr(P) mAb 4G10 (PY). The Western blot of the mAb anti-p125 fak immunoprecipitates (FAK) was reblotted with rabbit antiserum directed against p125 fak (FAK). The positions of immunoreactive p125 fak at apparent M r 125,000 and paxillin at apparent M r 60,000 -70,000 are indicated by the arrows on the left. Lower panels, levels of expression of G␣ subunits were analyzed by Western blotting (Blot) aliquots of total cell lysates with antisera directed against G␣ 12 (␣12), G␣ 13 (␣13), G␣ i-1 /␣ i-2 (␣i), or G␣ q (␣q). The positions of immunoreactive G␣ 12 (␣12*), G␣ 13 (␣13*), G␣ i-2 (␣ *), and G␣ q (␣q*) at apparent M r 43,000 are indicated by the arrows to the left. blotting with the anti-Tyr(P) mAb 4G10. As illustrated by Fig.  5 (upper panel), the increase in the tyrosine phosphorylation of proteins with apparent M r of 110,000 -130,000, 97,000, and 60,000 -70,000 induced by either G␣ 12 QL or G␣ 13 QL was inhibited by cotransfection with the C3 toxin expression vector. The G␣ q QL-induced tyrosine phosphorylation of this set of proteins was also inhibited by cotransfection with the C3 toxin expression vector (data not shown). In addition, we immunoprecipitated lysates of cotransfected cells with mAbs directed against p125 fak , paxillin, and p130 cas and subjected the immunoprecipitates to Western blot analysis with the anti-Tyr(P) mAb 4G10. As shown in Fig. 5 (middle panels), the G␣ 12 QLand G␣ 13 QL-induced tyrosine phosphorylation of p125 fak , paxillin, and p130 cas was inhibited by cotransfection with the C3 toxin expression vector. Western blotting with antisera to the G␣ subunits demonstrated that cotransfection with C3 toxin plasmid did not alter the expression levels of the transfected G␣ subunits (Fig. 5, lower panels).
We verified that thrombin-induced tyrosine phosphorylation of the focal adhesion proteins p125 fak , paxillin, and p130 cas in HEK 293 cells was also inhibited by transfection with the C3 toxin expression vector. In contrast, epidermal growth factorinduced tyrosine phosphorylation of a protein of apparent M r 170,000 (presumably the epidermal growth factor receptor) was not inhibited by transfection of HEK 293 cells with the expression plasmid encoding C3 toxin, indicating that this toxin inhibited tyrosine phosphorylation of focal adhesion proteins in a selective manner (results not shown).

DISCUSSION
Activation of GPCRs for mitogenic neuropeptides and bioactive lipids leads to a pertussis toxin-insensitive tyrosine phosphorylation of multiple protein substrates including the nonreceptor tyrosine kinase p125 fak and the adaptor proteins paxillin and p130 cas . The mechanism(s) linking activation of GPCRs to increases in protein tyrosine phosphorylation remains poorly understood. In the present study, we have used transient transfection of HEK 293 cells with expression plasmids encoding constitutively active G␣ 12 and G␣ 13 to investigate the involvement of the G 12 subfamily of G␣ proteins in tyrosine phosphorylation of the nonreceptor tyrosine kinase p125 fak and of the adaptor proteins paxillin and p130 cas . We report for the first time that HEK 293 cells expressing constitutively active G␣ 12 and G␣ 13 mutants exhibit increased tyrosine phosphorylation of p125 fak , paxillin, and p130 cas . Identical increases in tyrosine phosphorylation of these focal adhesion proteins were observed in HEK 293 cells treated with thrombin. This is noteworthy with reference to the previous reports that thrombin receptors couple to G␣ 12 and G␣ 13 in platelets and to G␣ 12 in astrocytoma cells (43,55) and, in particular, that fibroblasts lacking G␣ 13 exhibit a greatly diminished migratory  13 QL-induced tyrosine phosphorylation of p125 fak , paxillin, and p130 cas is blocked by the Rho-inhibitor C3 toxin. HEK 293 cells were cotransfected with G␣ and C3 expression vectors using lipofectAMINE as indicated. The cells were transfected with empty pcDNA1 vectors (Ϫ) or pcDNA1 encoding constitutively active mutants G␣ 12 QL (12*) or G␣ 13 QL (13*) in combination with the C3 vector (C3 ϩ). Three days after transfection, the cells were washed and incubated for 2 h in DMEM before being lysed. The protein concentration in each of the lysates was adjusted to 1 mg protein/ml. Upper panel, total tyrosine phosphorylation was analyzed by immunoprecipitation (IP) of the cell lysates with anti-Tyr(P) mAb PY72 (PY) and Western blotting (Blot) the immunoprecipitates with anti-Tyr(P) mAb 4G10 (PY). The positions of the migration of the molecular mass markers (in kDa) are indicated to the left. Middle panels, tyrosine phosphorylation of individual proteins was also analyzed by immunoprecipitation (IP) with mAb 2A7 directed against p125 fak (FAK), mAb directed against paxillin (PAX), or mAb directed against p130 cas (CAS) and Western blotting (Blot) the immunoprecipitates with anti-Tyr(P) mAb 4G10 (PY). The Western blot of the mAb anti-p125 fak immunoprecipitates (FAK) was reblotted with rabbit antiserum directed against p125 fak (FAK). The positions of immunoreactive p125 fak at apparent M r 125,000, paxillin at apparent M r 60,000 -70,000, and p130 cas at apparent M r 130,000 are indicated by the arrows on the left. Lower panels, levels of expression of G␣ subunits were analyzed by Western blotting (Blot) aliquots of total cell lysates with antisera directed against G␣ 12 (␣12) or G␣ 13 (␣13). The positions of immunoreactive G␣ 12 (␣12*) and G␣ 13 (␣13*) at apparent M r 43,000 are indicated by the arrows to the left. response to thrombin (44). Our observations reveal a novel connection between activated G␣ 12 and G␣ 13 and increases in tyrosine phosphorylation of p125 fak , paxillin, and p130 cas .
The signal transduction mechanism that mediates G␣ 12 -and G␣ 13 -induced increases in tyrosine phosphorylation of p125 fak , paxillin, and p130 cas in HEK 293 cells has several features in common with the mechanism extensively studied in Swiss 3T3 cells mediating tyrosine phosphorylation of these proteins in response to LPA and bombesin, which signal by activation of GPCRs. The common features in the signal transduction pathway leading to tyrosine phosphorylation of p125 fak , paxillin, and p130 cas include a requirement for an intact actin cytoskeleton, as indicated by the block of the tyrosine phosphorylation response by treatment with cytochalasin D, which specifically disrupts the actin cytoskeleton (3,4,14,15,20,64). Another common feature is the requirement for functional Rho, as indicated by inhibition with the C. botulinum C3 toxin, which ADP-ribosylates and functionally inactivates Rho (25,26,66). These observations suggest that G␣ 12 and G␣ 13 may couple GPCRs to increases in tyrosine phosphorylation of p125 fak , paxillin, and p130 cas .
It has been proposed that the mitogen-induced increases in tyrosine phosphorylation of p125 fak , paxillin, and p130 cas are downstream of Rho activation, stress fiber formation, and focal adhesion assembly (4, 25-28, 67, 68). Activated G␣ 12 and G␣ 13 have been shown to induce Rho activation (69) and Rho-dependent biological responses, including stress fiber formation and focal adhesion assembly (42,70,71). These results suggest that these G-proteins may couple GPCRs to the activation of Rho. Our observation that activated G␣ 12 and G␣ 13 mediate a Rho-dependent increase in tyrosine phosphorylation of p125 fak , paxillin, and p130 cas further supports the hypothesis that G␣ 12 and G␣ 13 , which together comprise the G 12 subfamily of G␣ proteins, couple the activation of GPCRs to the activation of Rho. Taken together, these findings suggest the existence of a signal transduction pathway whereby ligand occupation of GPCRs activates G␣ 12 and/or G␣ 13 , which subsequently activate Rho, presumably by a mechanism involving recruitment and activation of a GDP/GTP exchange factor. Activated Rho would induce the formation of actin stress fibers and promote the assembly of focal adhesions, resulting in recruitment of p125 fak to focal adhesions and tyrosine phosphorylation of p125 fak , paxillin, and p130 cas . This hypothesis warrants further experimental work.
Direct activation of PKC also leads to tyrosine phosphorylation of focal adhesion proteins in either Swiss 3T3 cells (14,15,20) or in HEK 293 cells (Fig. 2). Accordingly, we found that HEK 293 cells expressing constitutive active G␣ q , which couples to phospholipase C and PKC (61-63), exhibit increased tyrosine phosphorylation of focal adhesion proteins, which was dependent upon an intact cytoskeleton and functional Rho. In light of the report that, in contrast to G␣ 12 and G␣ 13 , G␣ q does not induce stress fiber formation and focal adhesions in Swiss 3T3 cells (42), G␣ q has not been thought to couple GPCRs to the activation of Rho. However, a recent report has suggested that GPCRs can induce the phosphorylation of Tiam1, a Rac-1specific exchange factor, in a manner that is dependent upon PKC and independent of pertussis toxin-sensitive G-proteins (72). In Swiss 3T3 cells, Cdc42, Rac, and Rho have been placed in a hierarchical cascade wherein Cdc42 activates Rac, which in turn activates Rho (73). Therefore, G␣ q may induce the Rho-dependent tyrosine phosphorylation of focal adhesion proteins via an activation of Rac, which may subsequently activate Rho. Because agonist-mediated tyrosine phosphorylation of focal adhesion proteins occurs independently of PKC, the precise physiological significance of the PKC-dependent pathway is unclear.
There is increasing evidence indicating that G␣ 12 and G␣ 13 are involved in cell migration, proliferation, and transformation (36, 38 -40, 43, 44, 48). Downstream targets through which G␣ 12 and G␣ 13 act may include Ras-, Rac-, Rho-, and Cdc42-dependent pathways leading to cytoskeletal reorganization and to the activation of mitogen-activated protein kinase, Jun N-terminal kinase, the Na ϩ -H ϩ exchanger, and the c-fos serum response element (40 -42, 46 -48). In this context, our results revealing that activated G␣ 12 and G␣ 13 induce increases in the level of tyrosine phosphorylation of p125 fak , paxillin, and p130 cas suggest novel mechanisms of action of these G␣ subunits. These findings assume an added importance in view of increasing evidence implicating p125 fak and p130 cas in cell migration, proliferation, and transformation. Gene disruption experiments have demonstrated a critical role of p125 fak in embryonic development, cell migration, and turnover of focal adhesions (29,74), and microinjection of dominant negative fragments of p125 fak prevented serum stimulation of DNA synthesis (75). The adaptor protein p130 cas has also been implicated in agonist-stimulated mitogenesis and in cell transformation (12,20,76) and has recently been identified as a mediator of p125 fak -mediated cell migration (31). Interestingly, embryonic fibroblasts lacking G␣ 13 also display a greatly impaired migratory response to thrombin receptor activation (44). Taken together with the results presented here, these findings raise the attractive possibility that p125 fak and p130 cas are downstream targets of G␣ 13 in a signal transduction pathway that regulates cell migration in response to GPCR agonists. Future studies should assess the contribution of increased tyrosine phosphorylation of p125 fak , paxillin, and p130 cas to the growth promoting and transforming activities of G␣ 12 and G␣ 13 .