Portal de Periódicos da CAPES

Two SH2 Domains of p120 Ras GTPase-activating Protein Bind Synergistically to Tyrosine Phosphorylated p190 Rho GTPase-activating Protein

1995; Elsevier BV; Volume: 270; Issue: 30 Linguagem: Inglês

10.1074/jbc.270.30.17947

1083-351X

Sophia S. Bryant, Scott Briggs, Thomas E. Smithgall, George A. Martin, Frank McCormick, Jinhong Chang, Sarah J. Parsons, Richard Jove,

Hippo pathway signaling and YAP/TAZ

p120 GTPase-activating protein (GAP) is a negative regulator of Ras that functions at a key relay point in signal transduction pathways that control cell proliferation. Among other proteins, p120 GAP associates with p190, a GAP for the Ras-related protein, Rho. To characterize the p120•p190 interaction further, we used bacterially expressed glutathione S-transferase fusion polypeptides to map the regions of p120 necessary for its interactions with p190. Our results show that both the N-terminal and the C-terminal SH2 domains of p120 are individually capable of binding p190 expressed in a baculovirus/insect cell system. Moreover, the two SH2 domains together on one polypeptide bind synergistically to p190, and this interaction is dependent on tyrosine phosphorylation of p190. In addition, mutation of the highly conserved Arg residues in the critical FLVR sequences of both SH2 domains of full-length p120 reduces binding to tyrosine-phosphorylated p190. The dependence on p190 phosphorylation for complex formation with p120 SH2 domains observed in vitro is consistent with analysis of the native p120•p190 complexes formed in vivo. These findings suggest that SH2-phosphotyrosine interaction is one mechanism by which the cell regulates p120•p190 association and thus may be a means for coordinating the Ras- and Rho-mediated signaling pathways. p120 GTPase-activating protein (GAP) is a negative regulator of Ras that functions at a key relay point in signal transduction pathways that control cell proliferation. Among other proteins, p120 GAP associates with p190, a GAP for the Ras-related protein, Rho. To characterize the p120•p190 interaction further, we used bacterially expressed glutathione S-transferase fusion polypeptides to map the regions of p120 necessary for its interactions with p190. Our results show that both the N-terminal and the C-terminal SH2 domains of p120 are individually capable of binding p190 expressed in a baculovirus/insect cell system. Moreover, the two SH2 domains together on one polypeptide bind synergistically to p190, and this interaction is dependent on tyrosine phosphorylation of p190. In addition, mutation of the highly conserved Arg residues in the critical FLVR sequences of both SH2 domains of full-length p120 reduces binding to tyrosine-phosphorylated p190. The dependence on p190 phosphorylation for complex formation with p120 SH2 domains observed in vitro is consistent with analysis of the native p120•p190 complexes formed in vivo. These findings suggest that SH2-phosphotyrosine interaction is one mechanism by which the cell regulates p120•p190 association and thus may be a means for coordinating the Ras- and Rho-mediated signaling pathways. In mammalian cells, Ras acts as a molecular relay in a mitogenic signal transduction network that ultimately regulates initiation of DNA replication(1Kung H.F. Smith M.R. Bekesi E. Manne V. Stacey D.W. Exp. Cell Res. 1986; 162: 363-371Google Scholar, 2Stacey D.W. DeGudicibus S.R. Smith M.R. Exp. Cell Res. 1987; 171: 232-242Google Scholar). Upon growth factor stimulation of receptor tyrosine kinases or activation of nonreceptor tyrosine kinases, Ras is activated by binding to GTP(3Lowy D. Willumsen B. Annu. Rev. Biochem. 1993; 62: 851-891Google Scholar). Activated Ras subsequently binds downstream effectors, including the Raf-1 serine/threonine kinase, which in turn activates a cytoplasmic kinase cascade, thus conveying mitogenic signals from the cell surface to their eventual nuclear destination (4Heidecker G. Kolch W. Morrison D.K. Rapp U.R. Adv. Cancer Res. 1992; 58: 53-73Google Scholar, 5Roberts T.M. Nature. 1992; 360: 534-535Google Scholar, 6Crews C.M. Erikson R.L. Cell. 1993; 74: 215-217Google Scholar, 7Moodie S.A. Willumsen B.M. Weber M.J. Wolfman A. Science. 1993; 260: 1658-1661Google Scholar, 8Van Aelst L. Barr M. Marcus S. Polverino A. Wigler M. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 6213-6217Google Scholar, 9Vojtek A.B. Hollenberg S.M. Cooper J.A. Cell. 1993; 74: 205-214Google Scholar). In normal cells, inactivation of Ras is mediated by the 120 kDa Ras GAP,1 1The abbreviations used are: GAPGTPase-activating proteinGSTglutathione S-transferasePAGEpolyacrylamide gel electrophoresis. p120, which was originally identified by its ability to stimulate the intrinsic GTPase activity of Ras over 100-fold(10Trahey M. McCormick F. Science. 1987; 238: 542-545Google Scholar, 11Trahey M. Wong G. Halenbeck R. Rubinfeld B. Martin G.A. Ladner M. Long C.M. Crosier W.J. Watt K. Koths K. McCormick F. Science. 1988; 242: 1697-1700Google Scholar, 12Vogel U.S. Dixon R.A. Schaber M.D. Diehl R.E. Marshall M.S. Scolnick E.M. Sigal I.S. Gibbs J.B. Nature. 1988; 335: 90-93Google Scholar, 13Boguski M.S. McCormick F. Nature. 1993; 366: 643-654Google Scholar). GTPase-activating protein glutathione S-transferase polyacrylamide gel electrophoresis. The p120 GAP can be subdivided into several domains, including a species-specific N-terminal hydrophobic region, an SH3 domain flanked by two SH2 domains, pleckstrin homology, and putative calcium-dependent binding domains, as well as the C-terminal catalytic domain(14Marshall M. Hill W. Ng A. Vogel U. Schaber M. Scolnick E. Dixon R. Sigal I. Gibbs J. EMBO J. 1989; 8: 1105-1110Google Scholar). In many signaling proteins, SH2 domains have been shown to mediate protein-protein interactions through their ability to bind phosphorylated tyrosine residues(15Pawson T. Gish G.D. Cell. 1992; 71: 359-362Google Scholar, 16Mayer B.J. Hamaguchi M. Hanafusa H. Nature. 1988; 332: 272-275Google Scholar, 17Musacchio A. Noble M. Pauptit R. Wierenga R. Saraste M. Nature. 1992; 359: 851-855Google Scholar, 18Waksman G. Kominos D. Robertson S.C. Pant N. Baltimore D. Birge R.B. Cowburn D. Hanafusa H. Mayer B.J. Overduin M. Resh M. Rios C. Silverman L. Kuriyan J. Nature. 1992; 358: 646-653Google Scholar, 19Birge R.B. Hanafusa H. Science. 1993; 262: 1522-1524Google Scholar, 20Ren R. Mayer B.J. Cicchetti P. Baltimore D. Science. 1993; 259: 1157-1161Google Scholar, 21Marengere L.E. Songyang Z. Gish G.D. Schaller M.D. Parsons J.T. Stern M.J. Cantley L.C. Pawson T. Nature. 1994; 369: 502-505Google Scholar). This role is also performed by the p120 SH2 domains, which are necessary for interactions between p120 and kinases, including the platelet-derived growth factor receptor and the Src tyrosine kinase, as well as cytoplasmic proteins such as the p120 GAP-associated protein, p62 (22Kaplan D.R. Morrison D.K. Wong G. McCormick F. Williams L.T. Cell. 1990; 61: 125-133Google Scholar, 23Kazlauskas A. Ellis C. Pawson T. Cooper J. Science. 1990; 247: 1578-1581Google Scholar, 24Brott B. Decker S. Shafer J. Gibbs J. Jove R. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 755-759Google Scholar, 25Park S. Liu X. Pawson T. Jove R. J. Biol. Chem. 1992; 267: 17194-17200Google Scholar). While the importance of the N-terminal SH2 domain in p120 for protein-protein interactions has been illustrated(26Anderson D. Koch C. Grey L. Ellis C. Moran M. Pawson T. Science. 1990; 250: 979-982Google Scholar, 27Moran M. Koch A. Anderson D. Ellis C. England L. Martin G. Pawson T. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 8622-8626Google Scholar, 28Marengere L. Pawson T. J. Biol. Chem. 1992; 267: 22779-22786Google Scholar), the necessary presence of the C-terminal SH2 domain has not previously been demonstrated, raising the question of the significance of two SH2 domains in p120. p120 GAP associates with several proteins that are potentially important for the regulation or function of p120 and Ras, including the 190-kDa phosphoprotein, p190. Despite its large size and multiple domains, only two functions have been ascribed thus far to p190, that of being a GAP for the Rho/Rac family of GTPases (29Settleman J. Albright C.F. Foster L.C. Weinberg R.A. Nature. 1992; 359: 153-154Google Scholar) and binding directly to GTP(30Foster R. Hu K.-Q. Shaywitz D. Settleman J. Mol. Cell. Biol. 1994; 14: 7173-7181Google Scholar). 2 2R. Roof, J.-H. Chang, and S. J. Parsons, manuscript submitted. This subfamily of Ras-related proteins is involved in membrane ruffling and formation of actin stress fibers in response to growth factor stimulation(32Ridley A.J. Paterson H.F. Johnston C.L. Diekmann D. Hall A. Cell. 1992; 70: 401-410Google Scholar, 33Ridley A.J. Hall A. Cell. 1992; 70: 389-399Google Scholar). In v-Src-transformed or epidermal growth factor-stimulated cells, the majority of cytoplasmic p120 is complexed with p190(34Ellis C. Moran M. McCormick F. Pawson T. Nature. 1990; 343: 377-381Google Scholar, 35Moran M. Polakis P. McCormick F. Pawson T. Ellis C. Mol. Cell. Biol. 1991; 11: 1804-1812Google Scholar). The biological significance of this interaction between the p120 and p190 GAPs is not yet clear. One attractive possibility is that through this complex, cells may couple the Ras-mediated signaling pathway to that of other GTPases, such as Rho or Rac, thereby coordinating DNA replication with changes in cell morphology. To characterize in more detail the p120•p190 interaction, we have used a series of bacterially expressed GST fusion proteins containing isolated regions of p120 to precisely map the domains that interact with p190 expressed in a baculovirus/insect cell system. Our results show that the two SH2 domains of p120 bind in a synergistic manner to tyrosine phosphorylated p190. Upon mutating critical SH2 residues, which are involved in mediating interactions with phosphotyrosine, binding of full-length p120 to tyrosine phosphorylated p190 is reduced. Consistent with these in vitro binding studies, p120 association with p190 in rat fibroblasts transformed by v-Src correlates with tyrosine phosphorylation of p190. These results suggest that SH2-phosphotyrosine interactions contribute to regulation of p120•p190 complex formation in the cell and in this way may coordinate the Ras- and Rho-mediated signaling pathways during cell proliferation. Spodoptera frugiperda (Sf9) insect cells (American Type Culture Collection) were cultured as described previously(36O'Reilly D. Miller L. Luckow V. Baculovirus Expression Vectors: A Laboratory Manual. W. H. Freeman and Company, New York, NY1992Google Scholar). For protein production, Sf9 cells were singly infected or coinfected with recombinant baculovirus stocks using a multiplicity of infection of 10 for each virus(25Park S. Liu X. Pawson T. Jove R. J. Biol. Chem. 1992; 267: 17194-17200Google Scholar). Untransformed (3Y1) and v-Src-transformed (SR3Y1) rat fibroblasts (37Kawai S. J. Virol. 1980; 34: 772-776Google Scholar) were maintained in Dulbecco's modified Eagle's medium as described previously(38Park S. Jove R. J. Biol. Chem. 1993; 268: 25728-25734Google Scholar). Construction of the following recombinant baculoviral vectors has been described elsewhere(39Park S. Marshall M. Gibbs J. Jove R. J. Biol. Chem. 1992; 267: 11612-11618Google Scholar) : bSrc coding for full-length chicken c-Src, bGAP encoding full-length bovine p120 GAP, GAPdSH encoding p120 GAP with a deletion of amino acids 166-518, and GAPdCAT encoding p120 GAP with a deletion of amino acids 751-983. The baculovirus RSH2E mutant encodes full-length bovine p120 GAP containing two Arg ⟶ Glu point mutations at amino acids 203 and 373.3 3S. Bryant and R. Jove, unpublished data. The complete human p190 cDNA was also expressed from a baculoviral recombinant.4 4G. Martin and F. McCormick, unpublished data. Regions of human p120 GAP containing the variable N-terminal region, various combinations of the SH2 and SH3 domains, or the C-terminal domains were subcloned into pGEX-2T and expressed in Escherichia coli(40Hjermstad S. Briggs S. Smithgall T. Biochemistry. 1993; 32: 10519-10525Google Scholar). Cells were pelleted and resuspended in buffer containing phosphate-buffered saline, 1% Triton X-100, 1 mM EDTA, 0.1%β-mercaptoethanol, 0.2 mM phenylmethanesulfonyl fluoride, and 5 mM benzamidine. Lysozyme was added to a final concentration of 0.5 mg/ml, and then cells were lysed by sonication and clarified. The soluble fraction was incubated with glutathione-coated Sepharose beads (Pharmacia Biotech Inc.) for 30 min at 4°C and washed 3 times in washing buffer. Beads were stored in 50 mM HEPES, pH 8.0, 150 mM NaCl, 10% glycerol, 0.1 mM dithiothreitol, and 5 mM benzamidine at 4°C. Immobilized fusion proteins were used in subsequent binding assays. Lysates of Sf9 insect cells infected with various baculovirus recombinants were incubated with different purified GST-p120 domain fusion proteins immobilized on glutathione-Sepharose beads for 1 h at 4°C. The resulting complexes were collected by centrifugation, washed 3 times with RIPA buffer (150 mM NaCl, 50 mM Tris-HCl, pH 7.5, 1% Nonidet P-40, 0.25% sodium deoxycholate, 1 mM sodium orthovanadate, 1 mM phenylmethanesulfonyl fluoride, 1 μM leupeptin, 1 μM antipain, 0.1 μM aprotinin, 10 μg/ml α2-macroglobulin, 2 mM EGTA) containing 1 mM sodium orthovanadate and lacking protease inhibitors, boiled in SDS-gel sample buffer, and resolved by SDS-PAGE. Levels of baculoviral-expressed recombinant p190 bound to the GST-p120 fusion proteins were assayed by Western blot analysis using polyclonal anti-p190 antibodies followed by horseradish peroxidase-conjugated goat anti-rabbit IgG antibodies for detection by enhanced chemiluminescence (ECL) (Amersham Corp.). Normalization of GST-p120 fusion proteins used in each incubation with p190 was confirmed by Coomassie stain of the lower portion of the same gel used for Western blot analysis of p190. Levels of p190 bound to GST fusion proteins were quantified from films exposed in the linear range using the AMBIS optical imaging system. 48 h postinfection, Sf9 cells were lysed in RIPA buffer for 15 min at 4°C. Lysates were clarified by centrifugation for 15 min and used in subsequent binding assays. Confluent 3Y1 or SR3Y1 fibroblasts (10-cm plates) were lysed in 1 ml of RIPA lysis buffer as described above. Proteins in cell lysates were immunoprecipitated with antibodies for 1 h at 4°C, incubated with protein A-Sepharose beads (Pharmacia) for 30 min at 4°C and then collected by centrifugation. Immunoprecipitates were washed 3 times with RIPA buffer, and resolved by SDS-PAGE. Levels of p120 GAP, p190, and their tyrosine phosphorylation states were determined by Western blot analysis using monoclonal anti-phosphotyrosine antibodies, and monoclonal or polyclonal anti-p120 or anti-p190 antibodies followed by horseradish peroxidase-conjugated anti-mouse or anti-rabbit antibodies for detection by ECL. Anti-p120 rabbit polyclonal sera was raised against the N terminus of human p120 GAP (amino acid residues 1-181). Preparation of anti-p120 and anti-p190 monoclonal antibodies as well as anti-p190 polyclonal sera have been described(41Chang J.-H. Sutherland W.M. Parsons S.J. Methods Enzymol. 1995; 254: 430-445Google Scholar). Anti-phosphotyrosine monoclonal antibody 4G10 was obtained from UBI (Lake Placid, NY). Previous studies indicated that the majority of p120 GAP associates with a cellular phosphoprotein, p190, in v-Src-transformed cells(35Moran M. Polakis P. McCormick F. Pawson T. Ellis C. Mol. Cell. Biol. 1991; 11: 1804-1812Google Scholar). To determine which domains of p120 GAP mediate this interaction, we used a panel of GST fusion proteins spanning the entire length of human p120 (Fig. 1). Within this panel, p120 SH2 and SH3 domains are present singly or in combination with each other, allowing us to define individual as well as any possible synergistic contributions of these domains to formation of the p120•p190 complex. After expressing these constructs in E. coli, cells were lysed, and fusion proteins were purified using glutathione-Sepharose beads. These immobilized GST fusion proteins were then incubated with Sf9 insect cell lysates containing baculoviral-expressed p190, and the levels of p190 that co-precipitated with the p120 fusion proteins were analyzed by SDS-PAGE and immunoblotting. Results show that none of the individual domains of p120 GAP bound p190 to an appreciable extent (Fig. 2A). Similarly, each SH2 domain in conjunction with the SH3 domain bound very little p190. In contrast, the N- and C-terminal SH2 domains together with the SH3 domain contained in the (N+C)SH fusion protein bound p190 to a greater extent. This enhanced binding is not due simply to an effect of the SH3 domain upon the individual SH2 domains, as neither NSH2+SH3 nor CSH2+SH3 fusion proteins bound p190 at levels significantly above those seen with NSH2 or CSH2 fusion proteins alone. Only when both SH2 domains were present did the levels of binding increase, suggesting that the binding of p120 to p190 is mediated by both SH2 domains of p120. Comparison with the levels of p190 in whole cell lysate indicates that approximately 2-3% of total p190 bound to (N+C)SH fusion protein in this experiment. Levels of GST fusion proteins used in each precipitation were determined by Coomassie staining of the same gel (Fig. 2B). Because SH2 domains interact with phosphorylated tyrosine residues, we examined the effect of increasing p190 tyrosine phosphorylation upon its interaction with the GST-p120 fusion proteins. To explore the effects of phosphorylation on p120•p190 interactions, p190 was phosphorylated on tyrosine using a baculoviral c-Src recombinant that exhibits an elevated kinase activity compared with endogenous rodent or chicken c-Src(25Park S. Liu X. Pawson T. Jove R. J. Biol. Chem. 1992; 267: 17194-17200Google Scholar). Previous studies suggested that p190 is a substrate of c-Src tyrosine kinase in intact rodent fibroblasts(42Chang J.-H. Wilson L. Moyers J. Zhang K. Parsons S.J. Oncogene. 1993; 8: 959-967Google Scholar). Sf9 insect cells were singly infected with wild-type baculovirus or recombinant baculovirus encoding either activated c-Src or p190 or were coinfected with both c-Src and p190 viruses. Cell lysates were loaded in triplicate for separation by SDS-PAGE and assayed by Western blot analysis using anti-Src, anti-p190, or anti-phosphotyrosine antibodies (Fig. 3, A-C). Results show very little tyrosine phosphorylation of p190 in infected cells by endogenous insect cell kinases; however, the level of tyrosine phosphorylation of p190 is markedly increased upon coinfection with the activated c-Src baculovirus, despite the lower overall levels of p190 in the p190/Src virus-coinfected cells. We assayed for p190 binding after incubating Sf9 cell lysates of p190 virus singly-infected cells or p190/Src virus-coinfected cells with purified GST-p120 fusion proteins. Levels of p190 bound to the p120 fusion proteins were determined by Western blot using anti-p190 antibodies (Fig. 4A), and levels of GST-p120 fusion proteins used in each incubation are shown by Coomassie staining of the same gel (Fig. 4B). We quantified the relative binding of p190 to GST-p120 fusion proteins after normalization with respect to levels of p190, as well as to levels of GST fusion proteins used in each incubation (Fig. 4C). Significantly, the SH3 domain of p120 bound neither the phosphorylated nor unphosphorylated form of p190, leading us to conclude that it does not directly mediate the interaction between p120 and p190. Similarly, GST-p120 fusion proteins containing C-terminal p120 GAP regions did not bind phosphorylated p190 (data not shown), implying that these domains do not directly mediate the interaction between p120 GAP and phosphorylated p190. On the other hand, tyrosine phosphorylation of p190 did enhance its binding to the isolated N- and C-terminal SH2 domains to similar low extents. p190 phosphorylation also enhanced its binding to the NSH2+SH3 and CSH2+SH3 constructs to approximately the same extent as the individual SH2 domains, demonstrating that the SH3 domain does not substantially enhance binding of p190 to the individual SH2 domains. Importantly, when both SH2 domains were present on the same polypeptide, binding to phosphorylated p190 was increased in a synergistic manner compared with that of the individual SH2 domains. In this experiment, approximately 30-40% of total phosphorylated p190 bound to (N+C)SH fusion protein, as determined by comparison with the levels of p190 in whole cell lysates. Tyrosine phosphorylation of p190 increased its binding to (N+C)SH by at least 10-fold compared with unphosphorylated p190. These findings show the dependence on tyrosine phosphorylation of p190 and the necessary presence of both N- and C-terminal p120 SH2 domains for maximal complex formation. To examine the contribution of p120 SH2 domains to complex formation in the context of the full-length protein, we constructed a p120 GAP baculoviral recombinant (RSH2E) in which the highly conserved Arg residue of the FLVR sequence in both SH2 domains was mutated to Glu using polymerase chain reaction mutagenesis. This Arg residue is one critical component of the SH2 phosphotyrosine binding pocket and forms an ion pair with the phosphate group(18Waksman G. Kominos D. Robertson S.C. Pant N. Baltimore D. Birge R.B. Cowburn D. Hanafusa H. Mayer B.J. Overduin M. Resh M. Rios C. Silverman L. Kuriyan J. Nature. 1992; 358: 646-653Google Scholar). Others have shown using GST-p120 fusion proteins that when this residue is mutated, binding of the individual SH2 domains to p120 GAP-associated protein p62 and the EGF receptor is abolished(28Marengere L. Pawson T. J. Biol. Chem. 1992; 267: 22779-22786Google Scholar). We expressed the p120 double point mutant as well as wild-type p120 GAP in Sf9 cells, and mixed these p120-containing lysates with either p190 or p190/Src virus-infected cell lysates. The p120•p190 GAP complexes were immunoprecipitated with an anti-p120 antibody and detected by Western blot analysis using anti-p120, anti-p190, and anti-phosphotyrosine antibodies. Equivalent amounts of wild-type and mutant p120 GAP were immunoprecipitated by the anti-p120 antibody (Fig. 5A). While both p120 GAPs bound little unphosphorylated p190, the p120 double point mutant bound significantly less phosphorylated p190 compared with wild-type p120 GAP (Fig. 5, B and C). These results are consistent with our GST-p120 GAP findings suggesting that the p120•p190 GAP interaction is mediated by p120 SH2 domains and depends on p190 tyrosine phosphorylation for maximal complex formation. Because p120 protein preferentially binds phosphorylated p190 in vitro, we compared the levels of native p120•p190 complex formation in normal 3Y1 rat fibroblasts and v-Src-transformed 3Y1 cells (SR3Y1). To study the interaction between endogenous p120 GAP and p190 in rat fibroblasts, we characterized several anti-p120 monoclonal antibodies(41Chang J.-H. Sutherland W.M. Parsons S.J. Methods Enzymol. 1995; 254: 430-445Google Scholar). We found that those antibodies that specifically immunoprecipitated p120 and maintained the p120•p190 complex all recognized the SH3 domain of p120 GAP (data not shown). 3Y1 or SR3Y1 cells were lysed and incubated with either anti-p120 or anti-p190 monoclonal antibodies. Resulting precipitates were resolved by SDS-PAGE, and immunoblots were probed with anti-p120, anti-p190 or anti-phosphotyrosine antibodies (Fig. 6, A-C). The p190 in the v-Src transformed SR3Y1 whole cell lysates is phosphorylated on tyrosine to a higher extent than that in normal 3Y1 cells (panel C). Consistent with our findings using Sf9 cell lyates, the level of p120 GAP bound to p190 in anti-p190 immunoprecipitates was increased in SR3Y1 cells (panel A). Essentially the same result was obtained when monoclonal anti-p120 antibodies were used to immunoprecipitate the p120•p190 complex (panel B). To ensure that the observed phosphorylation-dependent increase in the level of p190 binding to p120 was not an artifact of the particular antibodies being used, the same immunoprecipitation experiment shown in Fig. 6 was performed using three additional monoclonal anti-p120 antibodies. The polyclonal anti-p190 antibody did not immunoprecipitate the p120•p190 complex well, but it did show the difference in the phosphorylation state of p190 in v-Src-transformed versus normal 3Y1 cells (Fig. 7B). These additional monoclonal anti-p120 antibodies confirmed our earlier results; more p190 bound to the full-length p120 GAP when p190 was phosphorylated in SR3Y1 cells compared with when it was not, despite higher levels of p190 in 3Y1 lysates than in SR3Y1 lysates (Fig. 7, A-C). Together, these results suggest that the p120•p190 interaction in intact cells is enhanced by p190 tyrosine phosphorylation.Figure 7:Enhanced binding of tyrosine phosphorylated p190 to full-length p120 GAP detected with different monoclonal anti-p120 antibodies. Confluent 10 cm plates of either 3Y1 or SR-3Y1 cells were lysed in RIPA buffer and clarified. p120 GAP and p190 were precipitated from lysates with anti-p120 monoclonal antibody 3F3E6 (3F3) or anti-p190 polyclonal antibody (p190) as indicated at the top of each lane. Resulting complexes were analyzed by Western blot with anti-p190 polyclonal antibodies (A) or anti-phosphotyrosine monoclonal antibodies (B). C, to extend results shown in panel A, 3Y1 and SR-3Y1 lysates were also precipitated with 2A5B4 (2A5) or 10E6F3 (10E) anti-p120 monoclonal antibodies. Immunopreciptates, along with 2% of the whole cell lysates (WCL) used in each precipitation were analyzed by immunoblotting with polyclonal anti-p190 antibodies for levels of p190 co-precipitating with p120 GAP.View Large Image Figure ViewerDownload (PPT) While the catalytic activity of p120 GAP toward Ras has been studied extensively(10Trahey M. McCormick F. Science. 1987; 238: 542-545Google Scholar, 11Trahey M. Wong G. Halenbeck R. Rubinfeld B. Martin G.A. Ladner M. Long C.M. Crosier W.J. Watt K. Koths K. McCormick F. Science. 1988; 242: 1697-1700Google Scholar, 12Vogel U.S. Dixon R.A. Schaber M.D. Diehl R.E. Marshall M.S. Scolnick E.M. Sigal I.S. Gibbs J.B. Nature. 1988; 335: 90-93Google Scholar, 14Marshall M. Hill W. Ng A. Vogel U. Schaber M. Scolnick E. Dixon R. Sigal I. Gibbs J. EMBO J. 1989; 8: 1105-1110Google Scholar, 43DeClue J. Zhang K. Redford P. Vass W. Lowy D. Mol. Cell. Biol. 1991; 11: 2819-2825Google Scholar, 44Gideon P. John J. Frech M. Lautwein A. Clark R. Scheffler J. Wittinghofer A. Mol. Cell. Biol. 1992; 12: 2050-2056Google Scholar), the characterization of its interactions with associated proteins, such as p62 and p190, has been less well studied. In v-Src-transformed cells, only a small fraction of total cellular p120 GAP is found in the particulate membrane fraction along with Ras and p62. The majority of p120 GAP is complexed with p190 in the cytosolic fraction of the cell(34Ellis C. Moran M. McCormick F. Pawson T. Nature. 1990; 343: 377-381Google Scholar, 35Moran M. Polakis P. McCormick F. Pawson T. Ellis C. Mol. Cell. Biol. 1991; 11: 1804-1812Google Scholar, 38Park S. Jove R. J. Biol. Chem. 1993; 268: 25728-25734Google Scholar). Interest in p190 was heightened after the discovery that it too had GAP-like activity, but it was toward members of the Rho subfamily of small GTP-binding proteins(29Settleman J. Albright C.F. Foster L.C. Weinberg R.A. Nature. 1992; 359: 153-154Google Scholar, 45Settleman J. Narasimhan V. Foster L. Weinberg R. Cell. 1992; 69: 539-549Google Scholar). The Rho GTPases (RhoA, RhoB, RhoC, Rac1, Rac2, CDC42Hs, and TC10) are Ras-related proteins that contribute to changes in cell morphology, including membrane ruffling and actin stress fiber formation(13Boguski M.S. McCormick F. Nature. 1993; 366: 643-654Google Scholar). Because p190 exhibits Rho GAP activity, an intriguing model has emerged in which the cell, through p120•p190 association, may coordinate changes in cell morphology with stimulation of DNA replication. To characterize further the p120•p190 association, we have used a panel of GST-p120 fusion proteins and the baculovirus/Sf9 insect cell system to precisely determine what regions in p120 GAP are necessary for complex formation with p190 and to define some of the parameters that affect this interaction. Our results show that, while none of the individual p120 domains bound significant levels of p190, expression of the entire SH2/SH3/SH2 region in the (N+C)SH construct mediated p190 binding. That the p120•p190 complex is mediated by the p120 SH2/SH3/SH2 region is consistent with other studies, which have localized this interaction to the N-terminal half of p120 GAP protein (27Moran M. Koch A. Anderson D. Ellis C. England L. Martin G. Pawson T. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 8622-8626Google Scholar, 46Koch C. Moran M. Anderson D. Liu X. Mbmalu G. Pawson T. Mol. Cell. Biol. 1992; 12: 1366-1374Google Scholar, 47McGlade J. Brunkhorst B. Anderson D. Mbamalu G. Settleman J. Dedhar S. Rozakis-Adcock M. Chen L. Pawson T. EMBO J. 1993; 12: 3073-3081Google Scholar). Because SH2 domains bind to phosphorylated tyrosine residues(18Waksman G. Kominos D. Robertson S.C. Pant N. Baltimore D. Birge R.B. Cowburn D. Hanafusa H. Mayer B.J. Overduin M. Resh M. Rios C. Silverman L. Kuriyan J. Nature. 1992; 358: 646-653Google Scholar), we explored the effect of p190 tyrosine phosphorylation upon its association with p120 GAP domains. We show that tyrosine phosphorylation of p190 enhances its binding to the individual p120 SH2 domains to an equivalent limited extent and that the two SH2 domains together synergistically bind approximately 10-fold more p190. Although others have shown examples of p120 SH2/SH3 domains synergistically mediating interactions with the platelet-derived growth factor receptor in fibroblasts, only a small increase in binding of the entire SH2/SH3/SH2 region to the receptor relative to the individual N-terminal SH2 domains was observed(26Anderson D. Koch C. Grey L. Ellis C. Moran M. Pawson T. Science. 1990; 250: 979-982Google Scholar). In addition, only a low level of the isolated C-terminal SH2 domain bound to the receptor in comparison with the isolated N-terminal SH2 domain, raising a question as to the significance of the presence of the C-terminal SH2 domain(26Anderson D. Koch C. Grey L. Ellis C. Moran M. Pawson T. Science. 1990; 250: 979-982Google Scholar). In the cases of p120 GAP-associated proteins p62 and the epidermal growth factor receptor, it has been reported that these interactions are mediated primarily by the N-terminal SH2 domain and only weakly by the C-terminal SH2 domain, although potential synergistic effects of both SH2 domains were not determined(28Marengere L. Pawson T. J. Biol. Chem. 1992; 267: 22779-22786Google Scholar). In our experiments, we show that the presence of the C-terminal SH2 domain is necessary for maximal p120•p190 complex formation, providing evidence for the importance of p120 having two SH2 domains. Even with both SH2 domains present, however, tyrosine phosphorylation of p190 is needed for maximum complex formation. Together, these results suggest that the two SH2 domains in p120 bind synergistically to phosphorylated tyrosine residues in p190. Two additional lines of evidence support the important contribution of SH2-phosphotyrosine interactions to p120•p190 complex formation. First, point mutations of the critical Arg residues in the FLVR sequences of both SH2 domains in full-length p120 reduced binding to p190 (Fig. 5). Second, earlier studies showed that a tyrosine phosphorylated peptide corresponding to the region of PDGF receptor that binds p120 SH2 domains competes with p190 for binding to p120 (31Fantl W.J. Escobedo J.A. Martin G.A. Turck C.W. del Rosario M. McCormick F. Williams L.T. Cell. 1992; 69: 413-423Google Scholar). While the necessity of both p120 SH2 domains for maximal binding is evident from our results, the role that the SH3 domain plays in this interaction is less clear. From these experiments, we can conclude that the SH3 domain does not bind directly to p190 nor does it contribute to the ability of individual SH2 domains to bind p190. We cannot, however, rule out the possibility that the SH3 domain might have a structural role, such as providing a spatial requirement between both SH2 domains, that is only possible when all three domains are on the same polypeptide. Consistent with the conclusion that the SH3 domain does not directly bind p190 is our finding that when full-length p120 GAP was immunoprecipitated from rat fibroblasts with monoclonal anti-p120 antibodies directed against the SH3 domain, these antibodies did not disrupt the p120•p190 complex. To extend our findings from the baculovirus/Sf9 cell system, we examined p120•p190 native complex formation in vivo using rodent fibroblasts. Consistent with our insect cell data, more p190 co-precipitated with endogenous full-length p120 GAP from lysates of v-Src-transformed rat fibroblasts than from their normal counterparts in anti-p120 immunoprecipitates. Similarly, more p120 GAP complexed with p190 in anti-p190 immunoprecipitates prepared from v-Src-transformed SR3Y1 cells as compared with normal 3Y1 cells. This increase in co-precipitation of p120•p190 complexes paralleled the increase in tyrosine phosphorylation of p190, consistent with the importance of tyrosine phosphorylation for complex formation. Previous studies of p120•p190 complex formation yielded apparently conflicting results. Early experiments indicated that the majority of p120 GAP associates with p190 after growth factor stimulation and in v-Src-transformed Rat-1 fibroblasts(35Moran M. Polakis P. McCormick F. Pawson T. Ellis C. Mol. Cell. Biol. 1991; 11: 1804-1812Google Scholar). Other studies have suggested that similar levels of p190 associate with p120 GAP regardless of the level of p190 tyrosine phosphorylation in C3H10T1/2 fibroblasts(42Chang J.-H. Wilson L. Moyers J. Zhang K. Parsons S.J. Oncogene. 1993; 8: 959-967Google Scholar). Still others have found that stable expression in Rat-2 fibroblasts of an N-terminal polypeptide derived from p120 GAP leads to its association with p190 in both growth factor-stimulated and serum-deprived cells(47McGlade J. Brunkhorst B. Anderson D. Mbamalu G. Settleman J. Dedhar S. Rozakis-Adcock M. Chen L. Pawson T. EMBO J. 1993; 12: 3073-3081Google Scholar). One possible explanation for these contrasting findings is that the association between p120 and p190 could be regulated at multiple levels in addition to the SH2-phosphotyrosine interactions described here. Indeed, we detected a low level of p190 binding to (N+C)SH fusion protein in the absence of p190 tyrosine phosphorylation by c-Src (Fig. 2). Additional regulatory events might be dependent upon the mammalian cell type, and may not be evident using isolated domains of p120 expressed in bacterial cells. For example, we cannot exclude that there may be other structural features of p120 GAP that contribute to p190 association only in the context of the full-length, native p120 protein. Consistent with this possibility, we observed that point mutations in the FLVR sequences of both SH2 domains did not completely abolish binding of full-length p120 GAP to p190 (Fig. 5). A model consistent with our results is one in which a kinase like Src phosphorylates p190 on two tyrosine residues, thereby stabilizing its association with p120 GAP via the SH2 domains. There is some evidence suggesting that p120 in a complex with p190 has reduced catalytic activity toward Ras(35Moran M. Polakis P. McCormick F. Pawson T. Ellis C. Mol. Cell. Biol. 1991; 11: 1804-1812Google Scholar), raising the possibility that complex formation contributes to Ras activation. Similarly, p190 in the complex may have reduced GAP activity toward Rho family proteins. Alternatively, because the p120•p190 complex is predominantly cytosolic, complex formation may prevent access of both GAP proteins to their membrane-associated substrates. Thus elevated p120•p190 complex formation, promoted by tyrosine phosphorylation and possibly by other regulatory events, might simultaneously activate both Ras- and Rho-mediated signaling pathways. We thank members of the lab for stimulating discussions and K. Pumiglia for comments on the manuscript.