Integrin Engagement, the Actin Cytoskeleton, and c-Src Are Required for the Calcitonin-induced Tyrosine Phosphorylation of Paxillin and HEF1, but Not for Calcitonin-induced Erk1/2 Phosphorylation*

We have previously shown that in a HEK-293 cell line that overexpresses the C1a isoform of the calcitonin receptor (C1a-HEK), calcitonin induces the tyrosine phosphorylation of the focal adhesion-associated proteins HEF1 (a p130Cas-like docking protein), paxillin, and focal adhesion kinase and that it also stimulates the phosphorylation and activation of Erk1 and Erk2. We report here that cell attachment to the extracellular matrix, an intact actin cytoskeleton, and c-Src are absolutely required for the calcitonin-induced phosphorylation of focal adhesion-associated proteins. In contrast to the phosphorylation of paxillin and HEF1 in cells attached to fibronectin-coated dishes, calcitonin failed to stimulate the phosphorylation of paxillin and HEF1 in suspended cells, in cells attached to poly-d-lysine-coated dishes, and in attached cells pretreated with the RGD-containing peptide GRGDS. Overexpression of wild-type c-Src increased calcitonin-induced paxillin and HEF1 phosphorylation, whereas overexpression of kinase-dead Src or Src lacking a functional SH2 domain inhibited the calcitonin-stimulated tyrosine phosphorylation of these proteins. Overexpression of Src lacking the SH3 domain did not affect the calcitonin-induced phosphorylation of paxillin and HEF1. In contrast to the regulation of paxillin and HEF1 phosphorylation, the calcitonin-induced phosphorylation of Erk1 and Erk2 did not appear to involve c-Src and was only partially dependent on cell adhesion to the extracellular matrix and an intact actin cytoskeleton. Furthermore, inhibition of Erk1 and Erk2 phosphorylation had no effect on the calcitonin-induced phosphorylation of paxillin and HEF1. Thus, in C1a-HEK cells, the calcitonin receptor is coupled to the tyrosine phosphorylation of focal adhesion-associated proteins and to Erk1/2 phosphorylation by mechanisms that are in large part independent.

The activation of G protein-coupled receptors (GPCRs) 1 can lead to a large number of downstream cellular responses. Among these is the tyrosine phosphorylation of multiple proteins that localize at focal adhesion plaques (1)(2)(3)(4)(5)(6)(7)(8)(9)(10), which suggests that there is cross-talk between GPCR and integrin signaling pathways (11). Recent reports showed that stimulation of some GPCRs activates integrins by a protein kinase C-dependent mechanism, a process called "inside-out signaling." The activated integrins then transmit signals into the interior of the cells, leading to the tyrosine phosphorylation of focal adhesion-associated proteins (8,12,13). The focal adhesionassociated adaptor proteins paxillin and p130 Cas are among the proteins that are prominently tyrosine-phosphorylated following the activation of GPCRs and integrins. The GPCR-stimulated tyrosine phosphorylation of paxillin and p130 Cas is accompanied by profound cytoskeletal reorganization (1-3, 8, 10, 14) and is inhibited by cytochalasins (1, 4 -7), which disrupt the actin cytoskeleton.
The mechanism by which integrin ligation induces protein tyrosine phosphorylation has been well studied. Upon cell adhesion to extracellular matrix (ECM) proteins, focal adhesion kinase (FAK) is autophosphorylated, creating a binding site for the SH2 domain of Src (15). The subsequent binding of Src to FAK activates Src kinase, resulting in the Src-catalyzed phosphorylation of several other FAK tyrosine residues. The Src⅐FAK complex then binds to and phosphorylates other proteins such as Grb2, paxillin, and p130 Cas (16 -19). Although FAK directly binds to the SH3 domain of p130 Cas , it has been shown that Src rather than FAK catalyzes p130 Cas phosphorylation (17)(18)(19)(20). Although the mechanisms by which GPCRs stimulate tyrosine phosphorylation are not as clear, recent reports showed that the thrombin-and angiotensin II-induced cytoskeletal reorganization and the tyrosine phosphorylation of p130 Cas and paxillin require c-Src (10,21).
Calcitonin (CT) is a 32-amino acid peptide that increases renal calcium excretion and inhibits bone resorption by osteoclasts. The CT receptor (CTR) is a member of the calcitonin/ parathyroid hormone/secretin family of GPCRs. CT inhibits motility and induces marked cellular retraction of isolated osteoclasts, two effects that are related to cell attachment and cytoskeleton organization and that are thought to be essential to the CT-induced inhibition of bone resorption (22). Using a HEK-293 cell line that overexpresses the C1a isoform of the CTR (C1a-HEK), we have previously shown that CT stimulates the tyrosine phosphorylation and association of HEF1 (human enhancer of filamentation 1), paxillin, and FAK (23). These responses are inhibited by cytochalasin D and thus appear to require the integrity of the actin cytoskeleton (23). HEF1 is a multiple-domain docking protein that is 64% homologous to p130 Cas . Like p130 Cas , HEF1 is detected at focal adhesions and is activated following integrin ligation in lymphocytes (24,25) by a mechanism that is disrupted by cytochalasin D (24 -26). In T cell receptor signaling, HEF1 is a substrate of the Src family tyrosine kinases Fyn and Lck (27).
We have also shown that CT stimulates the phosphorylation of the serine/threonine kinases Erk1 and Erk2 in C1a-HEK cells (28). Erk1 and Erk2 are MAPKs whose phosphorylation is induced by the activation of both GPCRs (6, 29 -36) and integrins (16,19,(37)(38)(39). In some cells, the Erk1/2 phosphorylation induced by the activation of some GPCRs, including adrenergic receptors, the lysophosphatidic acid receptor, and the thrombin receptor, and by integrin ligation involves c-Src (32-34, 36, 40). However, although the Erk1/2 phosphorylation that results from integrin-mediated cell adhesion is inhibited by cytochalasin D (38), indicating the need for an intact actin cytoskeleton, this requirement may not apply to the GPCR-induced phosphorylation of Erk1/2 (6,9,41). Thus, the mechanisms that couple GPCRs to Erk1/2 phosphorylation, on the one hand, and to the tyrosine phosphorylation of focal adhesion-associated cytoskeletal proteins, on the other, appear to differ in some respects.
In this study, we show that the CT-induced phosphorylation of paxillin and HEF1 is completely dependent on integrinmediated cell adhesion and an intact actin cytoskeleton, but that CT-stimulated Erk1/2 phosphorylation is partly independent of these requirements. The tyrosine phosphorylation of paxillin and HEF1 is dependent on c-Src; but, in contrast to what has been reported for some other GPCRs, c-Src is not required for CT-induced Erk1/2 phosphorylation. Furthermore, inhibition of Erk1/2 phosphorylation does not have an effect on the CT-induced tyrosine phosphorylation of paxillin and HEF1. Therefore, the downstream mechanisms that couple the CTR to the tyrosine phosphorylation of the focal adhesion-associated proteins paxillin and HEF1, on the one hand, and to the phosphorylation and activation of Erk1/2, on the other, are apparently independent.
Cell Culture-Cells were maintained as described (23) in Dulbecco's modified Eagle's medium with 10% fetal bovine serum, 100 units/ml penicillin, 100 g/ml streptomycin, and 1 mg/ml G418. Cells were cultured in medium with 0.5% fetal bovine serum for 24 h prior to experimental treatment. Cell lysis, immunoprecipitation, and immunoblotting were performed as described previously (23).
Src Constructs and Transient Transfection-The cDNA for wild-type avian c-Src (WT-Src) was a gift from Dr. Joan Brugge (Harvard University). cDNAs for Src-⌬SH3, Src-⌬SH2, and Src-K295M were gifts from Drs. Harold Varmus and Pamela Schwartzberg (National Institutes of Health, Rockville, MD). Src-⌬SH3 has a deletion of homologue box A (positions 88 -137) of the Src SH3 domain; Src-⌬SH2 has a deletion of homologue box B (positions 148 -187) of the Src SH2 domain; and Src-K295M is a kinase-dead Src with a single amino acid mutation at residue 295. The various Src constructs were transiently transfected into C1a-HEK cells by LipofectAMINE (Life Technologies, Inc.). 48 h after transfection, serum was reduced to 0.5%, and the cells were cultured overnight before stimulation.

Calcitonin-induced Erk1/2 Phosphorylation and Calcitonininduced Paxillin and HEF1 Phosphorylation Differ in Their
Requirements for Cell Adhesion-The tyrosine phosphorylation of paxillin and HEF1 can be induced by both integrin-mediated cell adhesion to ECM proteins (18,24,25,(42)(43)(44) and the stimulation of GPCRs (1-4, 7, 8, 10, 23). The report that integrin engagement to the ECM is required for the tyrosine phosphorylation of paxillin downstream of muscarinic m3 receptors (8) suggests that there may be cross-talk between the integrin and GPCR signaling pathways. The activation and phosphorylation of the serine/threonine kinases Erk1 and Erk2 can also be activated by both integrin-and GPCR-induced signaling. We have previously shown that CT induces both paxillin/HEF1 and Erk1/2 phosphorylation in C1a-HEK cells via protein kinase C-and calcium-dependent mechanisms (23,28). We therefore examined the effect of cell adhesion on the CT-induced phosphorylation of paxillin, HEF1, and Erk1/2 in C1a-HEK cells to determine the role of integrin engagement in these responses.
As previously reported (23), 1 nM CT stimulated the tyrosine phosphorylation of paxillin and HEF1 in serum-starved C1a-HEK cells attached to culture dishes. However, when the cells were detached and kept suspended in serum-free medium, CT failed to stimulate paxillin and HEF1 phosphorylation ( Fig. 1, left panels). In contrast, CT stimulated Erk1/2 phosphorylation in both the suspended and the attached cells, although the degree of stimulation was less in the suspended cells ( Fig. 1, left panels).
We also examined CT-induced paxillin, HEF1, and Erk1/2 phosphorylation in cells that were replated on dishes coated with fibronectin or poly-D-lysine. After replating, the cells were cultured for 2.5 h, an interval shown by others to be sufficient for the initial attachment-induced tyrosine phosphorylation to return to basal levels (8), and then treated with 1 nM CT for an FIG. 1. Effect of cell adhesion on calcitonin-induced phosphorylation of paxillin, HEF1, and Erk1/2. Left panels, C1a-HEK cells were serum-starved for 24 h and then detached with 2 mM EDTA and kept in suspension in serum-free medium for 2 h. Quiescent cells in culture (attached (att)) or cells in suspension (susp) were treated with 1 nM CT or with vehicle (control (ctrl)) for 3 min and lysed, and paxillin (pax; upper panels) and HEF1 (center panels) were immunoprecipitated (IP) from 500 g of protein. The immune complexes were processed for immunoblotting with anti-phosphotyrosine antibody (PY). The membranes were stripped and reprobed with an antibody against the target antigen in the immunoprecipitation. Total cell lysates (TCL) were immunoblotted with anti-phospho-Erk1/2 antibody (P-Erk1/2), and the upper portion of the membrane was immunoblotted with anti-C/H antibody (C/H) to confirm equal protein loading (lower panels). Middle panels, suspended cells were replated on poly-D-lysine (PDL)-or fibronectin (FN)-coated dishes and allowed to attach for 2.5 h before treatment with 1 nM CT for 3 min. Samples were analyzed for phosphorylated HEF1, paxillin, and Erk1/2 as described above. Right panels, quiescent cells in culture were pretreated with 200 g/ml GRGDS or GRGES for 20 min before treatment with 1 nM CT for 3 min. Samples were analyzed for phosphorylated paxillin, HEF1, and Erk1/2 as described above. additional 3 min. As shown in Fig. 1 (middle panels), CT stimulated the tyrosine phosphorylation of paxillin and HEF1 in cells plated on fibronectin, but not in cells plated on poly-Dlysine. Again, in contrast to the lack of phosphorylation of paxillin and HEF1, CT induced an increase in Erk1/2 phosphorylation in cells replated on poly-D-lysine-coated dishes, although the level of phosphorylation was somewhat less than that seen in cells replated on fibronectin-coated dishes.
Finally, we examined the effect of the GRGDS peptide on CT-induced paxillin, HEF1, and Erk1/2 phosphorylation. This peptide mimics the integrin recognition site on ECM molecules such as fibronectin and vitronectin (45) and has been shown to inhibit integrin-mediated cell adhesion (46). The GRGDS peptide blocked CT-induced paxillin and HEF1 phosphorylation, but only weakly inhibited CT-induced Erk1/2 phosphorylation. An inactive analogue, GRGES, had no effect on any of the responses (Fig. 1, right panels). Together, these results suggest that the CTR, like the muscarinic m3 receptor, absolutely requires integrin engagement to induce the phosphorylation of focal adhesion-associated proteins. In contrast, CT induces the phosphorylation of Erk1/2 by both attachment-independent and attachment-dependent mechanisms.
Disrupting the Actin Cytoskeleton Does Not Abolish Calcitonin-stimulated Erk1/2 Phosphorylation in C1a-HEK Cells-We showed previously that the CT-induced tyrosine phosphorylation of paxillin and HEF1 is inhibited by cytochalasin D, which prevents actin polymerization and disrupts focal adhesions (23). Since we found that CT-induced Erk1/2 phosphorylation is partly independent of integrin engagement, unlike paxillin and HEF1 phosphorylation, we considered the possibility that CT-induced Erk1/2 phosphorylation might also be insensitive to agents that disrupt actin filaments. To test this, C1a-HEK cells were pretreated with either cytochalasin D or latrunculin A, both of which prevent the formation of actin filaments by different mechanisms (47). After stimulating the cells with CT, the phosphorylation states of paxillin and Erk1/2 were characterized. Both cytochalasin D and latrunculin A blocked the CT-induced phosphorylation of paxillin, whereas Erk1/2 phosphorylation was only slightly inhibited (Fig. 2). Thus, in contrast to the case of paxillin and HEF1 phosphorylation, the actin cytoskeleton apparently does not play a major role in the CT-stimulated phosphorylation of Erk1/2. c-Src Is Required for the Calcitonin-induced Tyrosine Phos-phorylation of Paxillin and HEF1, but Not for Erk1/2 Phosphorylation-It has been reported that, in at least some instances, the tyrosine kinase c-Src plays a role in the tyrosine phosphorylation of paxillin and p130 Cas and in Erk1/2 phosphorylation downstream of both GPCRs and integrins (10, 16, 17, 19, 21, 24, 25, 27, 32-34, 36, 48, 49). Src family tyrosine kinases have also been shown to play a role in regulating the tyrosine phosphorylation of HEF1 in B cells and T cells (24,27). We therefore investigated the role of c-Src in the coupling of the CTR to the phosphorylation of paxillin/HEF1 and Erk1/2. C1a-HEK cells were transiently transfected with WT-Src or various Src mutants and then stimulated with CT. As shown in Fig. 3A (left panels), overexpression of WT-Src increased the CT-stimulated (but not the basal) tyrosine phosphorylation of paxillin and HEF1, whereas the phosphorylation of these proteins was inhibited by overexpression of kinase-dead Src-K295M. Pretreatment with cytochalasin D abolished the CT-induced increases in paxillin and HEF1 phosphorylation in cells transfected with WT-Src (Fig. 3B), indicating that the overexpressed Src was in fact potentiating the cytoskeleton-dependent signaling mechanism. In contrast, CT-induced Erk1/2 phosphorylation was not influenced by overexpression of either WT-Src or Src-K295M (Fig. 3A, left panels). We then examined the involvement of the different domains of c-Src in these CT-induced signaling events (Fig. 3A, right panels). Overexpression of Src lacking the SH2 domain (Src-⌬SH2) had no effect on basal paxillin and HEF1 phosphorylation, but inhibited the CT-induced increases in the phosphorylation of these proteins (Fig.  3A, upper and middle right panels). In contrast, Src lacking the SH3 domain (Src-⌬SH3) actually increased the basal phosphorylation of both paxillin and HEF1, but the CT-induced in-  3. c-Src is required for calcitonin-stimulated paxillin and HEF1 phosphorylation, but not for Erk1/2 phosphorylation. A, C1a-HEK cells were transiently transfected with empty vector or with cDNA encoding WT-Src, kinase-inactivated Src (K295M), Src lacking the SH3 domain (⌬SH3), or Src lacking the SH2 domain (⌬SH2). Transfected cells were serum-starved as described under "Experimental Procedures" and then stimulated with 1 nM CT for 3 min and lysed. Paxillin (pax) and HEF1 were immunoprecipitated (IP) from 500 g of protein.
The immune complexes were processed for immunoblotting with antiphosphotyrosine antibody (PY). The blots were then stripped and reprobed with an antibody against the target antigen in the immunoprecipitation. Total cell lysates (TCL) were immunoblotted with anti-phospho-Erk1/2 antibody (P-Erk1/2) and with anti-C/H antibody (C/H) to determine the protein load. B, cells were transiently transfected with WT-Src and then pretreated with cytochalasin D (cyto.D) as described in the legend to Fig. 2. The calcitonin-induced phosphorylation of paxillin and HEF1 was analyzed as described for A. ctrl, control. creases in phosphorylation still occurred. Neither Src-⌬SH2 nor Src-⌬SH3 had an effect on CT-induced Erk1/2 phosphorylation (Fig. 3A, lower right panels). Thus, the CT-induced tyrosine phosphorylation of paxillin and HEF1 requires catalytically active c-Src with an intact SH2 domain, but c-Src is not required for the coupling of the CTR to Erk1/2 phosphorylation.

The Phosphorylation of Erk1/2 Is Not Upstream of Paxillin and HEF1 Phosphorylation in Calcitonin-induced Signaling-
Our findings that cell attachment, an intact cytoskeleton, and catalytically active Src are required for paxillin and HEF1 phosphorylation, but not for Erk1/2 phosphorylation, show that the tyrosine phosphorylation of paxillin and HEF1 plays little or no role in coupling the CTR to Erk1/2 activation. However, both the paxillin/HEF1 and Erk1/2 responses are dependent on the activation of protein kinase C and an increase of intracellular calcium (23,28), which raises the possibility that Erk1 and Erk2 contribute to the coupling of the CTR to the phosphorylation of paxillin and HEF1. We therefore examined the effect of PD98059, which inhibits Erk1/2 phosphorylation by MEK (50), on the CT-stimulated phosphorylation of paxillin and HEF1. Neither paxillin phosphorylation nor HEF1 phosphorylation was inhibited by PD98059, whereas, as expected, Erk1/2 phosphorylation was markedly inhibited (Fig. 4). Thus, the tyrosine phosphorylation of paxillin and HEF1, on the one hand, and Erk1/2 phosphorylation, on the other, appear to lie on independent signaling pathways downstream of the G qcoupled activation of protein kinase C and elevation of cytosolic calcium.

DISCUSSION
In studies of distal signaling events downstream of the CTR, we have found that CT induces the phosphorylation and activation of the MAPKs Erk1 and Erk2 (28) and the tyrosine phosphorylation and association of HEF1, paxillin, and FAK (23). Both of these distal signaling events can also be induced by integrin ligation (16 -19, 24, 25, 37-39, 42-44); and in at least some cases, the signaling pathways that couple GPCRs and integrins to these responses involve Src or other Src family members (10, 16 -19, 21, 24, 25, 27, 32-34, 36, 49). We have therefore examined the roles of integrin engagement and Src kinase in the mechanisms that couple the CTR to paxillin and HEF1 phosphorylation, on the one hand, and to Erk1/2 phosphorylation, on the other.
The phosphorylation of paxillin and HEF1 required both the binding of integrins to the extracellular matrix and an intact actin cytoskeleton, indicating that there is indeed cross-talk between the CTR and integrins. Thus, CT failed to induce paxillin or HEF1 phosphorylation in suspended cells, in cells plated on poly-D-lysine, or in the presence of the GRGDS peptide. Preincubation of the attached cells with cytochalasin D or latrunculin A, both of which interfere with actin filament formation by different mechanisms (47), also blocked CT-induced paxillin phosphorylation. These requirements are consistent with the reported mechanisms that couple other GPCRs to the phosphorylation of focal adhesion proteins (1, 4 -9, 41, 49). In contrast to their potent inhibition of paxillin and HEF1 phosphorylation, these treatments had little or no effect on CTinduced Erk1/2 phosphorylation. Conversely, inhibiting the phosphorylation and activation of Erk1/2 with the MEK inhibitor PD98059 failed to block the CT-induced tyrosine phosphorylation of paxillin and HEF1, indicating that the two responses (Erk activation and the phosphorylation of focal adhesion-associated proteins) are induced by independent coupling mechanisms. Similar independent regulation of focal adhesion proteins and the Erk proteins has been reported by others (6,9,41,49).
Both paxillin and p130 Cas are substrates of c-Src and are phosphorylated following integrin-mediated cell adhesion to the ECM in a Src-dependent manner (16 -19). HEF1 is also reported to be a substrate for the Src family tyrosine kinases Fyn and Lck during T cell receptor signal transduction (27). The mechanisms by which the tyrosine phosphorylation of paxillin and p130 Cas is induced by GPCRs are less well understood. Recent reports suggest that the GPCR-induced phosphorylation of these proteins requires c-Src (10,21) and that GPCR agonists induce the association of FAK with Src in fibroblasts (51). Our data suggest that the CTR-mediated tyrosine phosphorylation of paxillin and HEF1 also requires c-Src since overexpression of WT-Src enhanced CT-induced paxillin and HEF1 phosphorylation, whereas overexpression of dominant-negative kinase-dead Src-K295M or Src lacking the SH2 domain decreased CT-induced paxillin and HEF1 phosphorylation.
Paxillin and p130 Cas can potentially associate with c-Src in the focal adhesion complexes by binding directly to the SH3 domain of c-Src or by binding to FAK (52). However, the interaction with the Src SH3 domain is apparently not required for CT-induced paxillin and HEF1 phosphorylation since the deletion of the SH3 domain does not prevent the increase in phosphorylation that occurs in response to CT. In fact, overexpressing a Src protein lacking the SH3 domain resulted in an increase in the basal phosphorylation of paxillin and HEF1. This may be related to the known stabilizing effect of the SH3 domain on the inactive conformation of Src (53,54). Deletion of the SH3 domain would therefore increase the number of Src molecules in the active conformation and also promote the availability of the SH2 domain to bind to tyrosine-phosphorylated proteins in the adhesion complexes.
In contrast to the phosphorylation of paxillin and HEF1, the CT-induced phosphorylation of Erk1/2 was not significantly enhanced by overexpression of WT-Src or inhibited by kinase- FIG. 4. Calcitonin-induced paxillin and HEF1 phosphorylation is independent of Erk1/2 activation. C1a-HEK cells were pretreated with a 50 M concentration of the MEK inhibitor PD98059 or with vehicle for 30 min before stimulation with 1 nM CT for 3 min. Cells were then lysed, and paxillin (pax) and HEF1 were immunoprecipitated (IP) from 500 g of protein. The immune complexes were processed for immunoblotting with anti-phosphotyrosine antibody (PY). The membranes were stripped and reprobed with an antibody against the target antigen in the immunoprecipitation. To analyze Erk1/2 phosphorylation, total cell lysates (TCL) were immunoblotted with anti-phospho-Erk1/2 antibody (P-Erk1/2), and the upper portion of the membrane was immunoblotted with anti-C/H antibody (C/H) to confirm equal protein loading. ctrl, control.
dead Src-K295M, Src-⌬SH3, or Src-⌬SH2. These results differ from those of Lefkowitz and co-workers (33,49), who have reported that the coupling of the G i -coupled ␣ 2A -adrenergic and G q/11 -coupled ␣ 1B -adrenergic receptors to Erk1/2 phosphorylation in HEK-293 cells was attenuated by overexpression of kinase-dead Src or of Csk, which phosphorylates and inactivates c-Src. This difference between our results and those of Lefkowitz and co-workers may indicate that the CTR activates Erk1/2 in HEK-293 cells via a Src-independent mechanism other than or in addition to the Src-dependent mechanism that has been elucidated by these authors (30,31,33,49) and others (29). There is other evidence that supports such a possibility. We have found that overexpressing constitutively active Src-E378G in C1a-HEK cells induced the phosphorylation of Erk1/2 to a level somewhat lower than that induced by CT (data not shown), indicating that Src-dependent phosphorylation of Erk1/2 can occur in these cells. However, treatment of the Src-E378G-expressing cells with CT further stimulated Erk1/2 phosphorylation, consistent with the presence of both Src-dependent and Src-independent coupling mechanisms. In addition, our earlier study of the activation of Erk1/2 downstream of the CTR (28) found that the level of CT-induced Erk1/2 phosphorylation was similar to that induced by epidermal growth factor. In contrast, the GPCR-dependent Erk1/2 phosphorylation reported by Lefkowitz and co-workers was only 25-50% of the epidermal growth factor-induced phosphorylation. Interestingly, although we found that CT and epidermal growth factor induced similar levels of Erk1/2 phosphorylation, the CT-induced tyrosine phosphorylation of Shc, the adaptor protein that lies downstream of Src in the model of Lefkowitz and co-workers (33), was much less than the epidermal growth factor-induced phosphorylation of Shc. Kranenburg et al. (35) have identified a Src-and Shc-independent mechanism that couples the lysophosphatidic acid receptor to Erk1/2 phosphorylation in fibroblasts and COS cells, and a similar pathway may couple the CTR to Erk1/2 phosphorylation in C1a-HEK cells.
In conclusion, we have shown that the mechanisms by which the CTR couples to the MAPKs Erk1 and Erk2 and to the focal adhesion-associated proteins paxillin and HEF1 are largely independent. An intact actin cytoskeleton and the engagement of integrins with extracellular matrix proteins are required for the CT-stimulated tyrosine phosphorylation of paxillin and HEF1, but CT-activated Erk1/2 phosphorylation is largely independent of these factors. The CT-induced phosphorylation of paxillin and HEF1, but not of Erk1/2, is dependent on c-Src. These results offer new insight into the mechanisms by which CT may modulate cell function and, more generally, into GPCR-induced signaling events in HEK-293 cells.