Calcium-dependent Epidermal Growth Factor Receptor Transactivation Mediates the Angiotensin II-induced Mitogen-activated Protein Kinase Activation in Vascular Smooth Muscle Cells*

We have recently reported that angiotensin II (Ang II)-induced mitogen-activated protein kinase (MAPK) activation is mainly mediated by Ca2+-dependent activation of a protein tyrosine kinase through Gq-coupled Ang II type 1 receptor in cultured rat vascular smooth muscle cells (VSMC). In the present study, we found Ang II rapidly induced the tyrosine phosphorylation of the epidermal growth factor (EGF) receptor and its association with Shc and Grb2. These reactions were inhibited by the EGF receptor kinase inhibitor, AG1478. The Ang II-induced phosphorylation of the EGF receptor was mimicked by a Ca2+ionophore and completely inhibited by an intracellular Ca2+chelator. Thus, AG1478 abolished the MAPK activation induced by Ang II, a Ca2+ ionophore as well as EGF but not by a phorbol ester or platelet-derived growth factor-BB in the VSMC. Moreover, Ang II induced association of EGF receptor with catalytically active c-Src. This reaction was not affected by AG1478. These data indicate that Ang II induces Ca2+-dependent transactivation of the EGF receptor which serves as a scaffold for pre-activated c-Src and for downstream adaptors, leading to MAPK activation in VSMC.

Angiotensin II (Ang II), a major effector peptide of the reninangiotensin system, is now believed to play a critical role in the pathogenesis of cardiovascular remodeling associated with hypertension, heart failure, and atherosclerosis (7). We and others (8,9) have previously cloned the Ang II type 1 receptor (AT 1 ) which not only mediates diverse hemodynamic effects of Ang II (10) but also promotes hypertrophy and/or hyperplasia of vascular smooth muscle cells (VSMC) (11)(12)(13), cardiomyocytes (14), and cardiac fibroblasts (15). AT 1 belongs to the superfamily of heterotrimeric G protein-coupled receptors (GPCR) (8,9). In cultured VSMC, AT 1 activates phospholipase C (PLC), which initiates the generation of inositol trisphosphate and diacylglycerol, causing intracellular calcium mobilization and protein kinase C activation, respectively (16,17). In addition, Ang II induces several signaling events commonly evoked by growth factor receptors, such as the activation of MAPK (18,19) and the ribosomal S6 kinase (20), and the expression of the nuclear proto-oncogenes, c-fos, c-jun, and c-myc (21)(22)(23) in VSMC.
We have recently reported that Ang II-induced Ras and MAPK activation is mainly mediated by a calcium-dependent protein tyrosine kinase through G q -mediated PLC activation via AT 1 in cultured rat quiescent VSMC (36). However, the identity of the tyrosine kinase and its pathophysiological significance in the growth promoting signal of Ang II have remained unclear. In the present study, we found that Ang II induces Ca 2ϩ -dependent tyrosine phosphorylation of the EGF receptor to recruit Shc and Grb2, thereby activating MAPK in VSMC. The transactivation of the EGF receptor seems to be an essential point of convergence in this growth promoting cascade because it provides docking sites for the upstream tyro-sine kinase c-Src and downstream adaptors at the plasma membrane, and because its activity is required for the MAPK activation induced by Ang II.

EXPERIMENTAL PROCEDURES
Chemicals and Reagents-Chemicals and reagents were obtained from the following sources: Dulbecco's modified Eagle's medium (DMEM), fetal calf serum, penicillin, and streptomycin from Life Technologies, Inc.; Ang II from Peninsula Laboratories; recombinant human EGF and PDGF-BB from Upstate Biotechnology Inc.; AG1478, AG1295, A23187, and BAPTA-AM from Calbiochem; phorbol 12-myristate 13acetate and EGTA from Sigma; an agarose-conjugated glutathione Stransferase (GST)-Grb2-(1-217) fusion protein and protein A/G-agarose from Santa Cruz Biotechnology. CV11974 was a generous gift of Takeda Pharmaceutical Co.
Antibodies and Their Specificities-The rabbit polyclonal phosphospecific MAPK antibody (9101) raised against a synthetic phosphotyrosine peptide corresponding to amino acids 196 -209 (DHTGFLTEY-(P)VATRWC, where P indicates phosphate) of human p44 MAPK (ERK1) was obtained from New England Biolabs that detects only the catalytically active form of p42/44 MAPKs which are phosphorylated at Tyr 204 . We have previously shown (36) that this antibody specifically recognizes Tyr 204 -phosphorylated p42/44 MAPK in cultured rat VSMC. Anti-EGF receptor polyclonal antibody (pAb)(1005) raised against a synthetic peptide corresponding to amino acid residues 1005-1016 of human EGF receptor (identical to the corresponding mouse sequence) was obtained from Santa Cruz Biotechnology that also specifically recognizes rat EGF receptor in both immunoblotting and immunoprecipitation. Anti-Shc pAb (06-203) and monoclonal antibody (mAb) (S52420: clone 8) raised against a GST-tagged fusion protein corresponding to the SH2 domain (amino acid residues 366 -473) of the human p46/p52 Shc was obtained from Upstate Biotechnology and Transduction Laboratories, respectively. Anti-Shc pAb specifically reacts with p46/52/66 Shc of rat origin by immunoblotting and immunoprecipitation (34,37), and anti-Shc mAb also reacts with p46/52/66 Shc of rat origin by immunoblotting (34,38). Anti-Grb2 pAb (C-23) raised against a peptide corresponding to amino acid residues 195-217 of human Grb2 was obtained from Santa Cruz Biotechnology that is also specific for Grb2 of the rat origin by immunoprecipitation (34,38). Anti-Grb2 mAb (G16720: clone 24) raised against the entire 24-kDa Grb2 protein from rat brain was obtained from Transduction Laboratories that specifically reacts with rat Grb2 by immunoblotting (30). Anti-Sos pAbs (S15530) raised against a protein fragment of mouse Sos1 corresponding to amino acid residues 1-109 and a protein fragment of mouse Sos2 corresponding to amino acid residues 1095-1297, respectively, were obtained from Transduction Laboratories that also specifically react with Sos1 and Sos2 of rat origin by immunoblotting (34). Anti-PDGF ␤ receptor pAb (06-498) raised against a synthetic peptide corresponding to amino acid residues 1013-1025 of human PDGF ␤ receptor was obtained from Upstate Biotechnology. It specifically reacts with PDGF ␤ receptor by immunoblotting and immunoprecipitation. We have previously confirmed that this antibody also specifically reacts with rat PDGF ␤ receptor in cultured rat VSMC (39). The mAb directed to Tyr 530 -dephosphorylated c-Src (clone 28) was prepared as described previously which selectively recognizes the active form of rat c-Src (40). A horseradish peroxidase-conjugated recombinant antibody fragment specific for phosphotyrosine (RC20) and anti-phosphotyrosine mAb (4G10) were from Transduction Laboratories and Upstate Biotechnology Inc., respectively. Horseradish peroxidase-conjugated second antibodies were from Amersham Pharmacia Biotech.
The specificities of the antibodies used in the present study have been described in publications as referred to above as well as elsewhere. In addition, the rabbit pAbs used in the present immunoprecipitation studies (1005, 06-203, C-23), except control normal rabbit IgG or the pAb preincubated with its immunogen, specifically immunoprecipitated target proteins that, upon SDS-polyacrylamide gel electrophoresis, migrated to positions calculated from their molecular weights when visualized by the respective pAbs (1005) or mAbs (S52420, G16720) used in the present immunoblotting studies ( Fig. 4C and data not shown), thus confirming that the precipitated proteins are specific to the precipitating antibodies and that the bands detected by the blotting antibodies are specifically recognized by the antibodies.
Cell Culture-VSMC were prepared from the thoracic aorta of 12week-old Sprague-Dawley rats (Charles River Breeding Laboratories) by the explant method and cultured in DMEM containing 10% fetal calf serum, penicillin, and streptomycin as described previously (41). Subcultured VSMC from passages 3-15, used in the experiments, showed Ͼ99% positive immunostaining of smooth muscle ␣-actin antibody and were negative for mycoplasma infection. The expression of AT 1 but not AT 2 receptors was confirmed on the basis of binding studies with specific receptor antagonists (36). Cells at ϳ80% confluence in culture wells were made quiescent by incubation with serum-free DMEM for 3 days.
MAPK Activity-VSMC grown on a 24-well plate were stimulated with agonists at 37°C in serum-free DMEM for specified durations. The reaction was terminated by the replacement of medium with the icecold lysis buffer (10 mM Tris-HCl, pH 7.4, 20 mM NaCl, 2 mM EGTA, 2 mM dithiothreitol, 1 mM Na 3 VO 4 , 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, and 10 g/ml aprotinin). After brief sonication (10 s), the samples were centrifuged for 5 min at 14,000 ϫ g, and the supernatant was assayed for MAPK activity with an assay kit (Amersham Pharmacia Biotech) that measures the incorporation of [ 33 P]phosphate from [␥-33 P]ATP into a synthetic peptide (KRELVEPLTPAGEAP-NQALLR) as a specific MAPK substrate (36). The reaction was carried out with the cell lysate (ϳ1 g of protein) in 75 mM HEPES buffer, pH 7.4, containing 1.2 mM MgCl 2 , 2 mM substrate peptide, and 1.2 mM ATP, 1 Ci of [␥-33 P]ATP for 30 min at 30°C. The resultant solution was applied to a phosphocellulose membrane and extensively washed in 1% acetic acid and then in deionized water. The radioactivity trapped on the membrane was measured by liquid scintillation counting.
Immunoprecipitation and Immunoblotting-Cells were lysed by adding ice-cold lysis buffer, pH 7.5, containing 50 mM HEPES, 50 mM NaCl, 1% Triton X-100, 10% glycerol, 1.5 mM MgCl 2 , 1 mM EDTA, 10 mM sodium pyrophosphate, 1 mM Na 3 VO 4 , 100 mM NaF, 30 mM 2-(p-nitrophenyl) phosphate, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, and 10 g/ml aprotinin and centrifuged for 5 min at 14,000 ϫ g. Supernatant was mixed with the immunoprecipitation antibody and rocked at 4°C for 2-16 h, and then protein A/G-Sepharose was added for an additional 2 h to overnight. Immunoprecipitates were washed 3 times in the lysis buffer, solubilized in Laemmli sample buffer with 2-mercaptoethanol, resolved by SDS-polyacrylamide gel electrophoresis, and transferred to nitrocellulose membranes (Amersham Pharmacia Biotech). After blocking with 5% milk, the membrane was treated with a primary antibody followed by a secondary antibody conjugated with horseradish peroxidase. Immunoreactive proteins were detected by enhanced chemiluminescence (Amersham Pharmacia Biotech). For repeated immunoblotting, membranes were stripped in 62.5 mM Tris-HCl, pH 6.7, 2% SDS, and 0.1 M 2-mercaptoethanol for 30 -45 min at 50°C. For immunoblot analysis of MAPK phosphorylation, VSMC grown on a 6-well plate were directly lysed by Laemmli sample buffer with 2-mercaptoethanol, resolved by SDS-polyacrylamide gel electrophoresis, and subjected to immunoblotting. For immunoblot analysis of Grb2-associated proteins, agarose-conjugated GST-Grb2 fusion protein was rocked with Triton X-100 lysate of VSMC for 2 h to overnight at 4°C and washed 3 times with lysis buffer. Bound proteins were solubilized, resolved by SDS-polyacrylamide gel electrophoresis, and subjected to immunoblotting, as described above.
Reproducibility of the Results-Unless stated otherwise, results are representative of at least three experiments giving similar results.

Inhibition of Ang II-induced MAPK Activation by AG1478, the EGF Receptor Kinase
Inhibitor-In many MAPK activation systems, a tyrosine-phosphorylated scaffold is needed to assemble adaptor proteins (1-3). Highly likely candidates in VSMC for such a scaffold mediating MAPK activation by Ang II are the PDGF or EGF receptors. In VSMC, Ang II has been shown to cause PDGF ␤ receptor phosphorylation which leads to recruitment of the Shc⅐Grb2 complex to the receptor (25). This pathway may account for Ras and MAPK activation. A recent study revealed that several GPCRs use EGF receptor transactivation for MAPK activation, c-fos induction, and DNA synthesis in Rat-1 fibroblasts (31). To clarify the role of these receptor tyrosine kinases in Ang II-induced signal transduction in VSMC, we first tested the effect of selective receptor tyrosine kinase inhibitors (42) on the MAPK activity in VSMC. As shown in Fig. 1A, the EGF receptor kinase inhibitor tyrphostin AG1478 dose-dependently and completely blocked EGF-induced MAPK activation, whereas it had no effect on the PDGFinduced activation confirming its stringent selectivity. We then observed marked inhibition of Ang II-induced MAPK activation by AG1478 (Fig. 1B) but not by the PDGF receptor-selective tyrosine kinase inhibitor, AG1295 (Fig. 1C). AG1478 also inhibited the tyrosine phosphorylation of MAPK by Ang II and EGF without affecting that by PDGF-BB (Fig. 1D). Moreover, AG1478 inhibited Ca 2ϩ ionophore (A23187)-induced MAPK activation, whereas it had no effect on the phorbol ester-induced activation (Fig. 1C). This is in good agreement with our previous observation that the Ang II-induced Ras and MAPK activation requires a Ca 2ϩ -sensitive tyrosine kinase but not protein kinase C in VSMC (36).
Ang II Induces Calcium-dependent Tyrosine Phosphorylation of the EGF Receptor and Its Association with Shc, Grb2, and Sos-The activated EGF receptor can recruit the Grb2⅐Sos complex directly and indirectly via tyrosine phosphorylation of Shc, thereby activating the Ras/MAPK signaling pathway (1)(2)(3). To examine the possibility that Ang II signaling utilizes the EGF receptor to provide a docking site for Grb2, we examined proteins that interact with a GST-Grb2 fusion protein in the lysate of VSMC upon stimulation by Ang II. As shown in Fig.  2A, Ang II transiently increased association of several tyrosinephosphorylated proteins with the GST-Grb2 fusion protein as detected by anti-phosphotyrosine antibody. The association of tyrosine-phosphorylated proteins is specific to Grb2 because no band was seen when GST-agarose alone was used (Fig. 2D). A similar pattern of Grb2-associating tyrosine-phosphorylated proteins was observed following treatment of VSMC with the Ca 2ϩ ionophore, A23187 (Fig. 2B). The major phosphoprotein (ϳ170 kDa) associated with the GST-Grb2 fusion protein upon treatment with Ang II was identified as the EGF receptor because it was recognized by the anti-EGF receptor antibody (Fig. 2C) and was diminished by pretreatment with AG1478 (data not shown). We confirmed that the phosphorylated EGF receptor was coprecipitated with endogenous Grb2 upon Ang II treatment and was also diminished by AG1478 (Fig. 3). It should be noted that the tyrosine-phosphorylated bands of ϳ50 kDa were practically wiped out by AG1478, whereas the ϳ120-kDa band was not visibly affected (Fig. 3). In cultured rat VSMC, Ang II was shown to induce tyrosine phosphorylation of 46, 52, and 66 kDa Shc isoforms which subsequently form a complex with Grb2 (25). The ϳ50-kDa phosphoprotein associated with Grb2 upon treatment with Ang II shown in Fig. 3 should be p52 Shc. Indeed, the Ang II treatment resulted in association of 46-, 52-, and 66-kDa Shc isoforms to the GST-Grb2 fusion protein in which p52 Shc was the dominant form in VSMC ( Fig. 2A). A23187 also increased p52 Shc association to the GST-Grb2 fusion protein in VSMC (Fig. 2B).
Since Shc and Grb2 specifically recognize tyrosine-phosphorylated proteins (1-3), the Ang II-induced association of the EGF receptor and Shc with the GST-Grb2 fusion protein suggests that Ang II causes tyrosine phosphorylation of the EGF receptor. To confirm the involvement of the EGF receptor and to examine the effect of Ang II on the phosphotyrosine content of the EGF receptor, it was immunoprecipitated and analyzed by immunoblotting with an anti-phosphotyrosine antibody. As shown in Fig. 4A, Ang II induced tyrosine phosphorylation of the EGF receptor within 1 min, peaked at 2 min, and declined in 5 min. We further confirmed that the band recognized by anti-phosphotyrosine antibody specifically represents the tyrosine-phosphorylated EGF receptor because neither phosphorylated band nor immunoprecipitated EGF receptor could be observed when immunoprecipitation antibody (anti-EGF receptor antibody) was preabsorbed with its immunogen (Fig.  4C). Moreover, Ang II increased the amount of Shc, Grb2, and Sos which was coprecipitated with the EGF receptor (Fig. 4A). The phosphorylation of the EGF receptor was also observed upon treatment of VSMC with the Ca 2ϩ ionophore A23187 (Fig.  4B). In good agreement with our previous observation that Ang II-induced Ras and MAPK activation is mainly mediated through intracellular Ca 2ϩ mobilization (36), the Ang II-induced phosphorylation of the EGF receptor was completely inhibited by an intracellular Ca 2ϩ chelator, BAPTA-AM, but only partially affected by an extracellular Ca 2ϩ chelator, EGTA (Fig. 4C).
Ang II has been shown to stimulate the secretion of several growth factors in cultured VSMC, some of which have not been identified yet (13). It is possible that the Ang II-induced EGF receptor activation occurs secondarily to autocrine secretion of EGF receptor ligands such as EGF or transforming growth factor-␣. Since there are no reliable antibodies to rapidly neutralize multiple endogenous rat EGF receptor ligands (note that the transactivation takes place within 1 min) or to completely neutralize rat EGF receptor, we have excluded this possibility by examining the effect of the Ang II-treated conditioned medium on EGF receptor phosphorylation. As shown in Fig. 4D, the ability of the conditioned medium to phosphorylate the EGF receptor was lost when VSMC was pretreated with the AT 1 receptor antagonist, CV11974. Although there exists a small degree of basal phosphorylation of the PDGF ␤ receptor, we could not detect further phosphorylation of the PDGF ␤ receptor by Ang II during the time course in which we observed its EGF receptor phosphorylation (1ϳ5 min) or its MAPK activation (2ϳ10 min) in our VSMC (Fig. 5). These data indicate that the Ang II-induced EGF receptor transactivation, presumably through the intracellular Ca 2ϩ elevation coupled to the AT 1 receptor, may account for the induced association of the phosphorylated receptor with Shc, Grb2, and Sos, and resultant Ras and MAPK activation.
Interaction of c-Src with Shc and EGF Receptor upon Ang II Stimulation-The Src family tyrosine kinases have been implicated in the GPCR-induced MAPK activation (32-35). c-Src mediates Shc phosphorylation, Shc⅐Grb2 complex formation, and ensuing MAPK activation elicited by G i -coupled receptors in COS-7 cells (35). In VSMC, Ang II was shown to activate c-Src (43) which may be required for the Ras activation by Ang II (44). Ca 2ϩ -dependent c-Src activation was also reported in epidermal keratinocytes (45) and neuronal cells (46). To examine whether c-Src is involved in the MAPK cascade initiated by Ang II, the proteins inducibly associated with the GST-Grb2 fusion protein were immunoblotted by a monoclonal antibody, clone 28, which selectively recognizes the active (Tyr 530 -dephosphorylated) form of c-Src (40). Ang II increased transient association of active c-Src with the GST-Grb2 fusion protein (Fig. 6A). Since Shc is known to be tyrosine-phosphorylated by Src kinases (47) presumably through the interaction with the SH3 domain of the kinases (48), we examined whether active c-Src forms a coprecipitable complex with Shc in response to Ang II. As shown in Fig. 6B, Ang II induced complex formation of active c-Src with Shc that was correlated with p52 Shc tyrosine phosphorylation in VSMC.
Recent studies indicate that c-Src is required for mitogenic effects of EGF (49). In human breast carcinoma cell lines, c-Src has been shown to be associated with the activated EGF receptor through its SH2 domain (50). It has been suggested that catalytically active c-Src phosphorylates the non-autophosphorylation site of the EGF receptor, Tyr 891 , which serves as a binding site for c-Src (51). Therefore, we further tested the possibility that c-Src interacts with the EGF receptor upon stimulation by Ang II in VSMC. As shown in Fig. 6C, Ang II enhanced the association of active c-Src with the EGF receptor within 30 s. A23187 also increased the association (data not shown). The Ang II-induced association of active c-Src with the EGF receptor was still observable even in the presence of AG1478 (Fig. 6C). These data further suggest that Ang II could utilize c-Src to phosphorylate and associate with the EGF receptor, leading to recruitment of the downstream adaptors, Shc and Grb2, at the plasma membrane. DISCUSSION VSMC in culture has proven to be a useful model to examine the molecular mechanisms whereby vasoactive substances such as Ang II contribute to abnormal vascular hypertrophy. Recently, we have reported (36) that in VSMC, Ras and MAPK activation through AT 1 was mediated by a tyrosine kinase which may respond to G q -coupled intracellular Ca 2ϩ mobilization but not to protein kinase C activation. In the present study, we have demonstrated that an EGF receptor kinase inhibitor, AG1478, selectively inhibited MAPK activation induced by Ang II and the Ca 2ϩ ionophore A23187, whereas it had no effect on the activation induced by a phorbol ester. Furthermore, Ang II and A23187 induced tyrosine phosphorylation of the EGF receptor, which was sufficient to recruit the adaptor proteins that are involved in Ras activation. Thus, it is reasonable to speculate that the Ca 2ϩ -dependent tyrosine phosphorylation of the EGF receptor may be a common mechanism to activate MAPK shared by several GPCRs coupled to G q , such as AT 1 in VSMC. This notion is supported by the recent findings that several GPCR agonists (31) as well as KCl-induced depolarization (38) elicited tyrosine phosphorylation of the EGF receptor and subsequent recruitment of the adaptor proteins to the receptor. In general, phosphorylation of the EGF receptor by GPCR agonists and Ca 2ϩ agonists is relatively weaker than EGF itself as observed by us and others (31,38) indicating that there may exist a threshold phosphorylation level of the receptor that is sufficient for the recruitment of the adaptors and subsequent MAPK activation.
AT 1 has been shown to mobilize intracellular Ca 2ϩ by activating PLC-␤ through G q in cultured VSMC (16,17). However, recent reports by Marrero et al. (28,52) indicate that Ang II can mobilize intracellular Ca 2ϩ by PLC-␥ activation mediated by Src family tyrosine kinases in cultured rat VSMC. Since the EGF receptor can recruit and activate PLC-␥ through an autophosphorylation site at Tyr 992 (2), Ang II could elevate intracellular Ca 2ϩ through PLC-␥ activated by the EGF receptor thereby activating the MAPK cascade in VSMC. However, AG1478 (250 nM) did not influence Ang II-induced intracellular Ca 2ϩ mobilization in VSMC. 2 We also showed that intracellular but not extracellular Ca 2ϩ chelation was sufficient to inhibit EGF receptor phosphorylation induced by Ang II in the present study. Thus, the EGF receptor should be functionally downstream of the intracellular Ca 2ϩ mobilization in the Ras-MAPK signal cascade originating at AT 1 .
Although AG1478 is highly selective for the EGF receptor over other receptor tyrosine kinases (42), it is still possible that it affects other non-receptor kinases or signaling intermediates nonspecifically. This possibility is suspected by the fact that a higher dose of AG1478 was required for the inhibition of the MAPK activation by EGF than by Ang II. A similar phenomenon was also observed in Rat-1 fibroblasts which was attributed to the relatively weaker receptor phosphorylation by GPCR agonists than by EGF itself (31). In addition, pretreatment of VSMC with AG1478 tended to increase the amount of precipitated EGF receptor (see Fig. 6C). This may be due to the inhibition of basal level ubiquitination and subsequent proteolytic degradation of the receptor which requires the receptor tyrosine kinase activity (53). On the contrary, we found AG1478 had no effect on Ang II-induced association of c-Src to the EGF receptor nor on phorbol ester-or PDGF-BB-induced MAPK activation in the present study. Daub et al. (31) have reported that AG1478 inhibited the MAPK activation, c-fos mRNA expression, and DNA synthesis induced by endothelin-1 and thrombin, but it did not affect the tyrosine phosphorylation of focal adhesion kinase and paxillin by these agonists in Rat-1 fibroblasts. They have further demonstrated that these GPCR agonists failed to stimulate MAPK when Rat-1 cells were transfected with the dominant negative EGF receptor mutant. We 2 S. Eguchi, unpublished data.

FIG. 4. Ang II stimulates tyrosine phosphorylation of EGF receptor through AT 1 receptor-mediated intracellular Ca 2؉ mobilization but not through autocrine release of the EGF receptor ligand.
A, VSMC were stimulated with Ang II (100 nM) for indicated durations. After cell lysis, immunoprecipitation (IP) was performed with anti-EGF receptor pAb. Precipitates were analyzed by immunoblotting with anti-phosphotyrosine (pTyr) mAb, anti-Shc mAb, anti-Grb2 mAb, anti-Sos pAb, and anti-EGF receptor pAb by repeated reprobing. Arrowheads indicate the positions of 170-kDa tyrosinephosphorylated EGF receptor, p46 and p52 Shc, Grb2, Sos1 and Sos2, PYK2, and EGF receptor (EGFR), respectively. B, VSMC were stimulated with A23187 (10 M) or EGF (10 ng/ml) for indicated durations. After cell lysis, immunoprecipitation was performed with anti-EGF receptor pAb. Precipitates were analyzed by immunoblotting with antiphosphotyrosine mAb and anti-EGF receptor pAb as indicated. Arrowheads indicate the position of EGF receptor. C, VSMC were pretreated with or without 10 M BAPTA-AM or 5 mM EGTA and then stimulated with Ang II (100 nM) for 2 min. After cell lysis, immunoprecipitation was performed with anti-EGF receptor pAb or the anti-EGF receptor pAb preabsorbed with its immunogen (10 ϫ) for 2 h at 4°C as a control. Precipitates were analyzed by immunoblotting with anti-phosphotyrosine mAb and anti-EGF receptor pAb. Arrowheads indicate the position of EGF receptor. D, VSMC were pretreated with or without the AT 1 receptor antagonist CV11974 (1 M) for 30 min and then stimulated with conditioned medium from VSMC treated with Ang II (100 nM) for 2 min. After cell lysis, immunoprecipitation was performed with anti-EGF receptor pAb or the anti-EGF receptor pAb. Precipitates were analyzed by immunoblotting with anti-phosphotyrosine mAb and anti-EGF receptor pAb. Arrowheads indicate the position of EGF receptor. have also found that AG1478 (250 nM) inhibits Ang II-induced c-Fos expression and protein synthesis but not its intracellular Ca 2ϩ mobilization or c-Jun induction in cultured rat VSMC. 2 These data further confirm the specificity of AG1478 and strongly support the general observations that AG1478 acts at the point of the EGF receptor transactivation induced by GPCRs, leading to specific inhibition of GPCR-coupled MAPKdependent growth promoting signals, but does not interfere with functional coupling of GPCRs to other downstream kinases or signaling intermediates.
In addition, incomplete inhibition of Ang II-induced MAPK activation by AG1478 indicates that the activation is not exclusively mediated by the AG1478-sensitive pathway. The alternative activation signal(s) of MAPK by Ang II may involve other upstream transducers such as a novel Ca 2ϩ -sensitive tyrosine kinase, PYK2 (30)/CAK␤ (54)/RAFTK (55)/CADTK (56), Src family kinases (32)(33)(34)(35) as discussed below, ErbB2 (31,38), or protein kinase C (57). Further studies are required to determine relative contribution and possible cross-talks of these mechanisms leading to global growth promoting signaling.
Linseman et al. (25) showed that Ang II induced PDGF ␤ receptor phosphorylation and subsequent complex formation with Shc, Grb2, as well as c-Src in cultured rat VSMC. However, the contribution of the PDGF ␤ receptor to the MAPK activation by Ang II in VSMC is not likely. This view is supported by the observations that Ang II-induced MAPK activation was minimally affected by the selective PDGF receptor kinase inhibitor, AG1295, which almost completely abolished the PDGF-BB-induced MAPK activation (Fig. 1C) and that we could not detect the enhanced phosphotyrosine content of PDGF ␤ receptor by Ang II in our VSMC (Fig. 5) during the time course in which the Ang II-induced maximum Ras (3ϳ4 min) and MAPK activation (5 min) took place (36). However, Ang II-induced phosphorylation of the EGF receptor and its complex formation with the adaptors are detectable within 1 min. Given that the reported PDGF ␤ receptor phosphorylation (25) was detected in 5 min, plateaued in 10 min, and sustained up to 120 min, it may signal to different downstream transducers rather than MAPK.
In cultured rat VSMC, Ang II has been shown to induce tyrosine phosphorylation of all three Shc isoforms of p46, p52, and p66 which subsequently form a complex with Grb2 (25). In the present study, tyrosine phosphorylation of Shc by Ang II was observed in the immunoprecipitates of the three Shc isoforms in which p52 phosphorylation was dominant (Fig. 6B). Furthermore, Ang II treatment resulted in association of these Shc isoforms to the GST-Grb2 fusion protein ( Fig. 2A) and to the EGF receptor (Fig. 4A) in VSMC. In addition to Shc, Grb2 and Sos were also coprecipitated with the EGF receptor upon Ang II stimulation. Taken together with the recent findings that the expression of mutant Shc proteins defective in Grb2 binding displays a dominant negative effect on the pertussis toxin-insensitive MAPK activation induced by thrombin in fibroblasts (58), it is possible to speculate that the Ras and MAPK activation by Ang II may be at least partly mediated through Shc by linking the EGF receptor to a Grb2⅐Sos complex in VSMC. Some differences are noted in relative changes in the intensity of Shc bands immunoblotted by anti-phosphotyrosine mAb between Fig. 2A and Fig. 3 and between Fig. 2A blotted with anti-Shc pAb and Fig. 4A with anti-Shc mAb (the latter being needed to eliminate a thick rabbit IgG band). These may be due to a difference in efficiency of exogenous GST-Grb2 and endogenous Grb2 in binding Shc and also likely due to a difference in selectivity of anti-Shc pAb and mAb to each Shc isoform. It may also be possible that the relative difference in Shc band intensities is due to different affinity of each isoforms to Grb2 and the EGF receptor, respectively.
Ang II was reported to activate c-Src in cultured VSMC (43) which was proposed as an essential step for Ras activation by Ang II (44). Recently, Sadoshima and Izumo (34) reported that a Src family tyrosine kinase, Fyn, is activated by Ang II which recruits and phosphorylates Shc, leading to Ras activation in cardiac myocytes. We also found that Ang II increased transient association of the active c-Src with Shc which is contingent on Shc phosphorylation, suggesting a similar mechanism involving c-Src may operate the recruitment of Shc by Ang II in VSMC. Furthermore, the present study showed that Ang II and a Ca 2ϩ ionophore enhanced the association of c-Src with the EGF receptor. Although the exact hierarchical order of activation of the kinases and adapters has yet to be clarified, given that c-Src has been shown to phosphorylate the EGF receptor FIG. 6. Interaction of c-Src with Shc, Grb2, and EGF receptor upon Ang II stimulation. A, VSMC were stimulated with Ang II (100 nM) for indicated durations. After cell lysis, GST-Grb2 fusion protein immobilized on glutathione-agarose beads was added. Proteins associated with the fusion protein were separated by SDS-PAGE and immunoblotted with anti-c-Src mAb (clone 28) which selectively recognizes the catalytically active form of c-Src. B, VSMC were stimulated with Ang II (100 nM) for indicated durations. After cell lysis, immunoprecipitation (IP) was performed with anti-Shc pAb. Precipitates were analyzed by immunoblotting with anti-c-Src mAb (clone 28), anti-phosphotyrosine mAb, and anti-Shc mAb by repeated reprobing. Arrowheads indicate the positions of c-Src, tyrosine-phosphorylated p52 Shc, and p46/p52/p66 Shc, respectively. C, VSMC were pretreated with or without AG1478 (250 nM) for 30 min and stimulated with Ang II (100 nM) for indicated durations. After cell lysis, immunoprecipitation was performed with anti-EGF receptor pAb. Precipitates were analyzed by immunoblotting with anti-c-Src mAb (clone 28) and anti-EGF receptor pAb as indicated. Arrowheads indicate the position of c-Src and EGF receptor (EGFR), respectively. DMSO, dimethyl sulfoxide. (51) and that the enhanced association of c-Src with the EGF receptor by Ang II was not affected by AG1478 (Fig. 6C), we submit the scenario in which the active c-Src phosphorylates the EGF receptor.
In the case of pertussis toxin-sensitive GPCR, the ␤␥ subunits of G protein play a crucial role in the Ras and MAPK activation which also involve Src kinases (35). Recently, it has been proposed that Src kinase is downstream of the wortmannin-sensitive phosphoinositide 3-kinase ␥ in this cascade (59). However, as we reported, Ang II induces pertussis toxin-insensitive Ras and MAPK activation in VSMC (36), and wortmannin has no effect on the Ang II-induced MAPK activation. 3 In agreement with our concept that c-Src phosphorylates the EGF receptor, a recent report by Luttrel et al. (60) showed that G i -coupled receptor utilizes c-Src to phosphorylate the EGF receptor and for subsequent recruitment of the adaptors in MAPK activation. Thus, G i -and G q -coupled receptor-mediated MAPK cascades could converge on the Src kinase-operated EGF receptor transactivation.
In conclusion, we have demonstrated several lines of evidence that Ang II induces Ca 2ϩ -dependent tyrosine phosphorylation of the EGF receptor which serves as docking sites for presumably pre-activated c-Src and downstream adaptors at the plasma membrane, leading to MAPK activation in cultured rat VSMC. The identification and characterization of the putative transducer(s) which directly sense intracellular Ca 2ϩ mobilization to activate the kinases are under investigation.