Angiotensin II Induces Transactivation of Two Different Populations of the Platelet-derived Growth Factor β Receptor

Several signal transduction events induced by angiotensin II (AngII) binding to the angiotensin II type 1 receptor resemble those evoked by platelet-derived growth factor (PDGF) binding to the PDGF-β receptor (PDGFβ-R). We report here, in agreement with previous data, that AngII and PDGF-B-chain homodimer (PDGF-BB) stimulate tyrosine phosphorylation of the PDGFβ-R. Both AngII and PDGF-BB stimulated the phosphorylation of PDGFβ-R via the binding of tyrosine-phosphorylated Shc to PDGFβ-R. Both PDGF-BB- and AngII-induced phosphorylation of the Shc·PDGFβ-R complex was inhibited by antioxidants such as N-acetylcysteine and Tiron, but not by calcium chelation. However, transactivation of PDGFβ-R by AngII (measured by PDGFβ-R tyrosine phosphorylation) differed significantly from PDGF-BB. Evidence to support different mechanisms of PDGFβ-R phosphorylation includes differences in the time course of PDGFβ-R phosphorylation, differing effects of inhibitors of the endogenous PDGFβ-R tyrosine kinase and Src family tyrosine kinases, differing results when the PDGFβ-R was directly immunoprecipitated (PDGFβ-R-antibody) versuscoimmunoprecipitated (Shc-antibody), and cell fractionation studies that suggested that the Shc·PDGFβ-R complexes phosphorylated by AngII and PDGF-BB were located in separate subcellular compartments. These studies are the first to suggest that transactivation of tyrosine kinase receptors by G protein-coupled receptors involves a unique pathway that regulates a population of tyrosine kinase receptors different from the endogenous tyrosine kinase ligand.

* This work was supported in part by Grants R01 HL49192 and R01 HL59975 from the NHLBI, National Institutes of Health (to B. C. B.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ‡ The first two authors contributed equally to this work.  1 The abbreviations used are: AngII, angiotensin II; AT1R, angiotensin type 1 receptor; GPCR, G-protein coupled receptor; VSMC, vascular smooth muscle cells; ERK(1/2), extracellular signal-regulated kinases; JAK2, janus kinase; EGF, epidermal growth factor; EGF-R, epidermal growth factor receptor; PDGF-BB, platelet-derived growth factor Bchain homodimer; PDGF␤-R, platelet-derived growth factor ␤ receptor; TNF, tumor necrosis factor; BAPTA-AM, 1,2-bis(2-aminophenoxy)ethane-N,N,NЈ,NЈ-tetraacetic acid-acetoxymethyl; ROS, reactive oxygen species. and PDGF␤-R phosphorylation induced by PDGF-BB and An-gII were completely abolished by the antioxidants Tiron and N-acetylcysteine. However, phosphorylation of the PDGF␤-R by AngII and PDGF-BB occurred via different pathways as shown by different subcellular location and sensitivity to kinase inhibitors.

EXPERIMENTAL PROCEDURES
Reagents-Cell culture media and protein G-agarose were from Life Technologies, Inc. Polyclonal antibody against Shc, monoclonal antiphosphotyrosine antibody (4G10), and polyclonal antibody against PDGF␤-R were from Upstate Biotechnology (Lake Placid, NY). Polyclonal anti-phospho-specific ERK1/2 antibody was from New England Biolabs. Recombinant (human) TNF and recombinant (human) PDGF-BB were from Sigma. BAPTA-AM, AG1296 (PDGF-R tyrosine kinase inhibitor), and PP-1 were from Calbiochem.
Cell Culture-Rat aortic VSMC were isolated from the thoracic aorta of 200 -250-g male Harlan Sprague-Dawley rats and maintained in Dulbecco's modified Eagle's medium supplemented with 10% serum as described (14). Passages 8 -14 at 60 -70% confluence were growth arrested by incubation in Dulbecco's modified Eagle's medium with 0.1% serum for 48 h before use.
Immunoprecipitation and Immunoblot Analysis-The methods for immunoprecipitation and immunoblot analysis were described previously (15). In brief, growth-arrested VSMC were stimulated with AngII or PDGF-BB as indicated in each experiment. Cells were lysed in Triton/Nonidet P-40 lysis buffer (0.5% Triton, 0.5% Nonidet P-40, 10 mM Tris (pH 7.5), 2.5 mM KCl, 150 mM NaCl, 20 mM ␤-glycerol phosphate, 50 mM NaF, 1 mM orthovanadate, 10 g/ml leupeptin, 1 mM dithiothreitol, 10 g/ml soybean trypsin inhibitor, and 200 mM benzamidine), scraped off the dish, and centrifuged. Lysates containing equal amounts of protein were precleared and incubated with antibodies overnight at 4°C. Antibody complexes were collected by incubation with protein G-agarose for 3 h at 4°C. Precipitates were washed four times with the Triton/Nonidet P-40 lysis buffer and then resuspended in SDS sample buffer. Samples were separated by SDS-polyacrylamide gel electrophoresis (8 -12%) and transferred to nitrocellulose membranes. After incubation in blocking solution (1% bovine serum albumin, 10 mM Tris (pH 7.5), 100 mM NaCl, 0.1% Tween 20), membranes were incubated with primary antibodies. After washing (Tris-buffered saline, 0.03% Tween 20), the blots were incubated with the appropriate secondary antibodies. The membranes were washed and proteins were detected by the ECL system (Amersham Pharmacia Biotech).

RESULTS
AngII Stimulates Tyrosine Phosphorylation of the PDGF␤-R-To compare phosphorylation of the PDG␤-R by AngII and PDGF-BB, the PDGF␤-R was immunoprecipitated from cell lysates treated with AngII (100 nM) or PDGF-BB (30 ng/ml) for 2 min. Immunoblotting with an antibody against tyrosinephosphorylated proteins showed phosphorylation of a ϳ180-kDa band (Fig. 1A, upper panel) which was identified as the PDGF␤-R by stripping and reprobing with PDGF␤-R antibody (Fig. 1A, lower panel). Phosphorylation of the PDGF␤-R by AngII at 2 min was ϳ50% of phosphorylation by PDGF-BB itself.
The adaptor protein Shc exists as three isoforms, 66 kDa, 52 kDa, and 46 kDa, and is a likely candidate to mediate crosstalk between the PDGF␤-R and the AT1R by assembling a signal transduction complex at the PDGF␤-R (7, 10). Therefore, AngII-induced PDGF␤-R phosphorylation may be studied by coimmunoprecipitation of Shc and the PDGF␤-R. AngII and PDGF-BB both stimulated tyrosine phosphorylation of only the 66-kDa isoform of Shc (Fig. 1B, upper panel), whereas the two smaller Shc isoforms were not involved in PDGF␤-R transactivation in our cell system. Furthermore, tyrosine-phosphorylated Shc was part of a signal transduction complex that included the PDGF␤-R as shown by coimmunoprecipitation of the PDGF␤-R with Shc (Fig. 1B, lower panel).
PDGF␤-R Tyrosine Phosphorylation Stimulated by AngII Is More Rapid Than by PDGF-BB-To determine whether the signaling events induced by AngII and PDGF-BB which cause PDGF␤-R phosphorylation were similar, the time course for tyrosine phosphorylation of the PDGF␤-R was studied. AngII caused a more rapid and transient phosphorylation of the PDGF␤-R than did PDGF-BB (Fig. 2). AngII-induced phosphorylation of the PDGF␤-R peaked at 1 min (23.4 Ϯ 3.8-fold) and returned to base line after 10 min, when immunoprecipitated with antibody against the PDGF␤-R. A similar increase (24.6 Ϯ 3.4-fold) in PDGF␤-R phosphorylation was observed when coimmunoprecipitated with antibody against Shc (Fig. 2). Thus, it appeared that 50% of the PDGF␤-R phosphorylated in re- Cell lysates underwent immunoprecipitation (IP) with anti-Shc or anti-PDGF␤-R and were immunoblotted with anti-phosphotyrosine. The extent of PDGF␤-R phosphorylation was quantified by scanning densitometry using the NIH image program and normalized to total PDGF␤-R content after reprobing the membranes with anti-PDGF␤-R antibody. Data are the mean Ϯ S.E. of four independent experiments. sponse to AngII was bound to Shc, and 50% was not bound. In contrast, PDGF-BB maximally phosphorylated the PDGF␤-R at 2 min, and phosphorylation was sustained up to 30 min (Fig.  2, and data not shown). Phosphorylation of the 66-kDa isoform of Shc by AngII was also more rapid than phosphorylation of the PDGF␤-R (data not shown).
AngII-and PDGF-BB-induced Shc⅐PDGF␤-R Complex Phosphorylation Is Calcium-independent-Because AngII increases intracellular calcium (16,17) and the transactivation of the EGF-R by AngII was reported to be calcium-dependent (12), we determined the role of calcium in PDGF␤-R phosphorylation. Treatment with BAPTA-AM, a chelator of intracellular calcium, did not alter phosphorylation of the PDGF␤-R by either AngII or PDGF-BB when coimmunoprecipitated with an anti-body against Shc (Fig. 3, B and C) or when immunoprecipitated with an antibody against PDGF␤-R (Fig. 3A, upper panel, and 3B). Furthermore, tyrosine phosphorylation of the 66-kDa Shc isoform was also not significantly inhibited (Fig. 3C). Activation of ERK1/2 by both AngII and PDGF-BB was also not calcium-dependent in these VSMC (Fig. 3A, lower panel). BAPTA-AM treatment stimulated ERK1/2 activation to a small extent in control cells. Because chelation of intracellular calcium blocks activation of the MAP kinase phosphatase-1 (18), we presume that activation of ERK1/2 in unstimulated control cells is a consequence of changes in MAP kinase phosphatase-1 activity. As a positive control for calcium chelation, VSMC were also treated with TNF. BAPTA-AM completely inhibited ERK1/2 activation by TNF, confirming that pretreatment with BAPTA-AM inhibits calcium-regulated signaling events (Fig.  3A, lower panel). (19). Furthermore, previous studies have shown that ROS are required for PDGF-BB-mediated signal transduction (20). To determine the role of ROS in AngII-induced Shc⅐PDGF␤-R phosphorylation, VSMC were preincubated with the antioxidants, N-acetylcysteine and Tiron. Both antioxidants completely abolished AngII-stimulated tyrosine phosphorylation of the Shc⅐PDGF␤-R complex (Fig. 4).

AngII-and PDGF-BB-induced Shc⅐PDGF␤-R Complex Phosphorylation Is Completely Abolished by Antioxidants-It is well known that AngII stimulates reactive oxygen species (ROS) production in vitro and in vivo
Role of PDGF␤-R Kinase in the PDGF␤-R Tyrosine Phosphorylation Induced by AngII and PDGF-BB-To evaluate the role of the endogenous PDGF␤-R tyrosine kinase in phosphorylation of the Shc⅐PDGF␤-R complex by AngII, we used tyrphostin AG1296, a potent and specific inhibitor of this kinase (21). First, we studied the effect of AG1296 on phosphorylation of PDGF␤-R that was associated with Shc as measured by immunoprecipitation with anti-Shc antibody. At concentrations reported to inhibit PDGF-mediated events (21), AG1296 (10 M) completely inhibited tyrosine phosphorylation of PDGF␤-R and Shc induced by PDGF-BB (Fig. 5A). In contrast, AG1296 did not affect AngII-induced phosphorylation of the Shc⅐PDGF␤-R complex (Fig. 5A). Second, we studied the effect of AG1296 on total PDGF␤-R phosphorylation by immunoprecipitation with anti-PDGF␤-R antibody (Fig. 5B). Under these experimental conditions, AG1296 caused complete inhibition of the AngIIinduced tyrosine phosphorylation of the PDGF␤-R. These results suggest that two different populations of the PDGF␤-R exist, differentiated by whether the PDGF␤-R is bound to the 66-kDa Shc isoform or is not bound. The endogenous PDGF␤-R tyrosine kinase is therefore responsible for phosphorylation of the uncomplexed PDGF␤-R but not for phosphorylation of the Shc⅐PDGF␤-R complex (Fig. 5B).
Role of the Src Kinase Family in PDGF␤-R Phosphorylation Induced by AngII and PDGF-BB-Previous studies from our laboratory and others showed that the Src tyrosine kinase family is required for AngII-and PDGF-BB-induced signaling events in VSMC, such as cytoskeletal reorganization at focal adhesions (22), migration (23), and proliferation (23). To determine the role of Src in AngII-and PDGF-BB-mediated phosphorylation of the Shc⅐PDGF␤-R complex, we used PP-1, a pyrazolopyrimidine that interacts specifically with Src family kinases and is a competitive inhibitor of ATP (24). PP-1 did not inhibit AngII-induced phosphorylation of the PDGF␤-R (Fig. 6, upper panel), nor AngII phosphorylation of Shc (Fig. 6A, bottom  panel). In contrast, PP-1 completely inhibited PDGF-BB-mediated PDGF␤-R phosphorylation and Shc phosphorylation (Fig.  6A), in agreement with findings of Waltenberger et al. (23). PP-1 also completely inhibited the tyrosine phosphorylation of PDGF␤-R not bound to Shc that was stimulated by AngII (Fig.  6B). These results further support the concept that PDGF-BB and AngII activate the PDGF␤-R via two different pathways in VSMC, differentiated by their interaction with Shc.
Role of the Tyrosine Kinase JAK2 in PDGF␤-R Phosphorylation Induced by AngII and PDGF-BB-To define further the kinases involved in AngII-mediated Shc⅐PDGF␤-R complex phosphorylation, we investigated JAK2, which is rapidly activated by AngII (26). AG490 (60 M), a specific JAK2 inhibitor (25,26), did not inhibit phosphorylation of the PDGF␤-R induced by PDGF-BB (Fig. 7). AG490 also had no effect on AngIImediated PDGF␤-R phosphorylation measured by Shc coimmunoprecipitation (Fig. 7). These data demonstrate that JAK2 activity is not responsible for PDGF␤-R phosphorylation by AngII.

PDGF␤-R Phosphorylation by AngII and PDGF Occurs in Different Compartments
: Differential Centrifugation-A possible explanation for different pathways of Shc⅐PDGF␤-R complex phosphorylation by AngII and PDGF-BB could be that the Shc⅐PDGF␤-R complex is localized to a different subcellular compartment when stimulated with AngII. This hypothesis is supported by the findings of Lotti et al. (27), who showed that Shc proteins were localized to endoplasmic reticulum membranes and redistributed after activation of tyrosine kinase receptors. To determine whether separate populations of Shc⅐PDGF␤-R complexes could be purified, we performed cell fractionation studies. Tyrosine-phosphorylated PDGF␤-R was present in the supernatant after centrifugation of lysates at 3,500 ϫ g for 15 min from cells stimulated with both AngII and PDGF-BB (Fig. 8A, left panel). In contrast, tyrosine-phosphorylated PDGF␤-R was no longer present in the supernatant of lysates prepared from AngII-stimulated cells after centrifugation at 22,500 ϫ g for 10 min (Fig. 8A, right panel). However, when cells were stimulated with PDGF-BB, tyrosine-phosphorylated PDGF␤-R was still present in the 22,500 ϫ g supernatant (Fig. 8A, right panel). To determine the location of Shc after AngII and PDGF-BB stimulation, we dissolved the 22,500 ϫ g pellet in 2% SDS-Triton/Nonidet P-40 lysis buffer and immunoprecipitated Shc with an anti-Shc antibody. As shown in Fig. 8B, the Shc⅐PDGF␤-R complex was only present in the pellet of cells stimulated with AngII. In PDGF-BBstimulated lysates only a small amount of Shc was present in the 22,500 ϫ g pellet. These results suggest that the PDGF␤-R phosphorylated in response to AngII (present as a Shc⅐PDGF␤-R complex) was located in a compartment physically different from the PDGF␤-R phosphorylated in response to PDGF-BB. We measured the relative amounts of caveolin-1 present in the Shc⅐PDGF␤-R complexes purified from AngIIand PDGF-BB-stimulated cells. Because there was no difference in caveolin-1 coprecipitated (data not shown) we do not believe that translocation to caveolae explains the difference between AngII and PDGF-BB. DISCUSSION The major findings of the present study are that AngII and PDGF-BB stimulate tyrosine phosphorylation of the PDGF␤-R via different pathways (Fig. 9). Although transactivation of the PDGF␤-R and EGF-R by AngII has been reported previously (10,12), the novel aspect of the present study is the demonstration of two different pathways that are mechanistically separate as shown by the presence of two populations of the PDGF␤-R. One population of PDGF␤-R exists in which Shc is bound to the receptor in the basal state, whereas in the other population the PDGF␤-R is not complexed to Shc. Evidence to support different mechanisms for PDGF␤-R phosphorylation related to these separate populations includes differences in the time course of PDGF␤-R phosphorylation (Fig. 2), differing effects of inhibitors of the endogenous PDGF␤-R tyrosine kinase and Src family tyrosine kinases (Figs. 4 and 5), differing results when the PDGF␤-R was directly immunoprecipitated (PDGF␤-R-antibody) versus coimmunoprecipitated (Shc-antibody), and cell fractionation studies that suggested that the PDGF␤-R phosphorylated by AngII and PDGF-BB is located in separate subcellular compartments (Fig. 8). These studies are the first to suggest that transactivation of tyrosine kinase receptors by GPCRs involves a unique pathway that regulates a population of tyrosine kinase receptors different from the endogenous tyrosine kinase ligand (Fig. 9).
In the discussion below, the different mechanisms for AngIIand PDGF-mediated PDGF␤-R tyrosine phosphorylation are examined based on the model shown in Fig. 9. It has become clear that ROS mediate many of the rapid responses of VSMC to growth-promoting agents (19, 28 -30). ROS are essential for both PDGF-mediated stimulation of VSMC DNA synthesis (20) and for AngII-mediated increases in VSMC hypertrophy (32). In addition, ROS are necessary for increases in signal events related to cell survival and protein synthesis such as activation of the Akt/protein kinase B pathway (33). Finally, inhibiting ROS-dependent signal transduction events by overexpression of catalase (34) or by treatment with antioxidants such as N-acetylcysteine (35) inhibits VSMC DNA synthesis and cell survival. In the present study we observed that AngII-dependent phosphorylation of both the PDGF␤-R and Shc required ROS generation (Fig. 4). Similar results have been shown by others for PDGF and its receptor (20).
What is the nature of the ROS-stimulated tyrosine kinase that phosphorylates the PDGF␤-R and Shc in response to AngII? In the present study we show that the kinase is not PYK2 (no calcium requirement, Fig. 3), not Src (no effect of the Src kinase inhibitor PP1, Fig. 6), and not JAK2 (no effect of the JAK2 inhibitor AG490, Fig. 7). The PDGF␤-R tyrosine kinase plays a role in AngII-mediated tyrosine phosphorylation of the PDGF␤-R but not of Shc (Fig. 5). Specifically, we found that the population of PDGF␤-R which is not bound to Shc is completely dependent on the endogenous PDGF␤-R tyrosine kinase as shown by inhibition with AG1296 (Fig. 5B). In contrast, tyrosine phosphorylation of the PDGF␤-R that is complexed to Shc was not altered by AG1296 (Fig. 5B). As expected, PDGF-BBmediated tyrosine phosphorylation of both populations of the PDGF␤-R was inhibited by AG1296 (Fig. 5). A candidate tyrosine kinase that could transactivate the PDGF␤-R is Syk. Syk is a nonreceptor 72-kDa tyrosine kinase, first isolated from porcine spleen and capable of phosphorylating Shc (36). There are reports showing that Syk is involved in ERK1/2 activation following activation of GPCRs (37). Syk is expressed mainly in cells of the lymphoid system, although high level expression has been observed in mouse heart (38). Future studies will be required to determine whether Syk is expressed in VSMC and is activated by AngII.
The adaptor protein Shc appears to be central to the differences in AngII-and PDGF-mediated phosphorylation of the PDGF␤-R. Shc proteins are adaptors, involved in the signaling events of GPCRs, as well as receptor tyrosine kinases (39,40). In the present study, we found that association of Shc with the PDGF␤-R in VSMC determines the nature of PDGF␤-R phosphorylation and subcellular location. Stimulation of PDGF␤-R that is not complexed to Shc by either AngII or PDGF-BB causes tyrosine phosphorylation of the PDGF␤-R, and the receptor remains in a low molecular weight complex. We characterize this population of PDGF␤-R as residing in compartment I, possibly cytosol (Fig. 9). PDGF-BB binding to the PDGF␤-R causes Shc tyrosine phosphorylation, and the Shc-phosphorylated PDGF␤-R complex (when stimulated by PDGF-BB) also resides in compartment I as shown by our inability to sediment the Shc⅐PDGF␤-R complex (Fig. 8, A and B). In contrast, AngII has a very different effect as shown by the finding that high speed centrifugation after AngII stimulation sedimented the Shc-phosphorylated PDGF␤-R complex (Fig. 8B, compartment II). After solubilizing the pellet resulting from the high speed centrifugation at 22,500 ϫ g we could recover all of the Shcphosphorylated PDGF␤-R in that pellet, which leads to two possible conclusions. One is that the yet unidentified "Shckinase" phosphorylates the PDGF␤-R bound to Shc on different tyrosine residues than the intrinsic PDGF␤-R kinase, and this results in the translocation to a different compartment. A second explanation is that the Shc-kinase itself forms a multiprotein complex with the Shc-phosphorylated PDGF␤-R complex, which then leads to translocation to a different compartment. Further studies in our laboratory will be done to address these possibilities.
Candidate locations for compartments I and II are the cytosol and endoplasmic reticulum membranes, respectively, based on work of Lotti et al. (27). These investigators showed that in quiescent NIH/3T3 cells, Shc is localized to the rough endoplasmic reticulum (27). Upon receptor tyrosine kinase stimulation, Shc redistributed toward the cell periphery (plasma membrane and endocytic structures) (27). Shc redistribution was lost in cells transfected with a phosphorylation-defective mutant of the EGF-R, suggesting that the interaction of Shc with specific phosphotyrosine residues in the cytoplasmic tail of the activated receptor is required to redistribute Shc to the cell periphery. The mechanism for Shc redistribution by GPCRs, however, remains to be elucidated. There is evidence that the different Shc proteins (66-, 52-, and 46-kDa isoforms) are localized in different compartments of the cell (41). Clark et al. (42) showed that in 3T3-L1 adipocytes, the 66-kDa isoform was localized mainly to the endoplasmic reticulum, the 46-kDa isoform to the plasma membrane, and the 52-kDa isoform was distributed evenly throughout the cytosol (42). It is possible that GPCRs and tyrosine kinase receptors couple preferentially to a specific Shc isoform. However, in the present study, both AngII and PDGF-BB stimulated phosphorylation predominantly of 66-kDa Shc with minimal phosphorylation of the 52and 46-kDa isoforms. In addition, the different nature of receptor phosphorylation by either a cytosolic tyrosine kinase (in the case of AngII) or an intrinsic receptor tyrosine kinase (PDGF-BB) may influence the subcellular location of Shc.
Transactivation of the EGF-R by AngII in cardiac fibroblasts has been shown to be important in the regulation of fibronectin and transforming growth factor-␤ synthesis (13). The importance of PDGF␤-R phosphorylation by AngII remains unclear, but the mechanism differs clearly from EGF-R phosphorylation. One model for EGF-R-mediated transactivation by AngII involves an increase in intracellular calcium that stimulates tyrosine phosphorylation of the tyrosine kinase PYK2. PYK2 then forms a complex with c-Src, and phosphorylation of EGF-R occurs via interactions with c-Src and/or PYK2 (12,43). EGF-R transactivation by AngII as reported by Eguchi et al. (31) involves an increase in intracellular calcium and leads to the EGFR-dependent Ras activation and subsequent p70 s6K kinase activation via two parallel pathways. However, the activation of p70 s6K kinase via these two pathways was similar whether AngII or EGF was the agonist. In contrast, in the present study, we show that PDGF␤-R phosphorylation by AngII did not require calcium or c-Src. Furthermore, we show that PDGF␤-R phosphorylation by AngII involves a mechanism distinct from that activated by PDGF-BB. The phosphorylation, localization, and binding of Shc to the PDGF␤-R seem to be important for the signal transduction pathway identified here. Thus, Shc appears to be an important mediator to study the functional significance of receptor tyrosine kinase transactivation by AngII.