Platelet-derived Growth Factor-stimulated Migration of Murine Fibroblasts Is Associated with Epidermal Growth Factor Receptor Expression and Tyrosine Phosphorylation*

Previous studies have shown that epidermal growth factor (EGF) synergizes with various extracellular matrix components in promoting the migration of B82L fibroblasts expressing wild-type EGF receptors and that functional EGF receptors are critical for the conversion of B82L fibroblasts to a migratory cell type (1). In the present study, we examined the effects of platelet-derived growth factor (PDGF) on the motility of B82L fibroblasts using a microchemotaxis chamber. We found that PDGF can enhance fibronectin-induced migration of B82L fibroblasts expressing wild-type EGF receptors (B82L-clone B3). However, B82L cells that lack the EGF receptor (B82L-parental) or that express an EGF receptor that is kinase-inactive (B82L-K721M) or C-terminally truncated (B82L-c′973) exhibit little PDGF-stimulated migration. In addition, none of these three cell lines exhibit the capacity to migrate to fibronectin alone. These observations indicate that, similar to cell migration toward fibronectin, PDGF-induced cell migration of B82L fibroblasts is augmented by the expression of an intact EGF receptor kinase. The loss of PDGF-stimulated motility in B82L cells that do not express an intact EGF receptor does not appear to result from a gross dysfunction of PDGF receptors, because ligand-stimulated tyrosine phosphorylation of the PDGF-β receptor and the activation of mitogen-activated protein kinases are readily detectable in these cells. Moreover, an interaction between EGF and PDGF receptor systems is supported by the observation that the EGF receptor exhibits an increase in phosphotyrosine content in a time-dependent fashion upon the addition of PDGF. Altogether, these studies demonstrate that the expression of EGF receptor is critical for PDGF-stimulated migration of murine B82L fibroblasts and suggest a role for the EGF receptor downstream of PDGF receptor activation in the signaling events that lead to PDGF-stimulated cell motility.

PDGF 1 plays an important role in the control of cell proliferation, differentiation, and migration-associated processes such as wound healing and tumor cell metastasis (2)(3)(4). Binding of PDGF induces receptor dimerization and activation of its intrinsic tyrosine kinase activity (2,3). There are two homologous, yet distinct, isoforms of the PDGF receptor. The PDGF-␤ receptor binds PDGF-BB with high affinity and PDGF-AB with low affinity, whereas the PDGF-␣ receptor binds all three PDGF forms with high affinity (5)(6)(7). Both receptor isoforms can mediate mitogenic signals, yet they exhibit a differential capacity to mediate chemotactic signals (8,9). Activation of the receptor tyrosine kinase leads to receptor autophosphorylation and the regulation of various cellular substrates containing Src homology 2 or phosphotyrosine binding domains including phospholipase-C␥, phosphatidylinositol 3-kinase, GTPase-activating protein for Ras, Src family kinases, and several adaptor proteins such as Grb2, Shc, and Nck (2,7). These events initiate diverse signal transduction cascades that culminate in growth factor-stimulated end points such as cell proliferation and migration. Emerging evidence supports the concept that PDGF-induced cell motility is associated with the activation of phospholipase-C␥ and phosphatidylinositol 3-kinase and the suppression of Ras GTPase-activating protein (10 -12). Despite the similarities between EGF and PDGF actions (13), certain cell types respond differently to these two growth factors (14), and the unique signaling mechanisms that impart PDGF and EGF specificity at post-receptor levels have yet to be fully elucidated.
PDGF has been reported to be a potent chemotactic agent for a number of cell types, including fibroblasts (15,16), neutrophils, monocytes (17), and smooth muscle cells (18). Considerable progress has been made recently toward an understanding of how cell migration occurs and the linkages that are needed between PDGF receptor signaling and the modulation of the cell migration machinery (19 -21). Cell motility is a process requiring the temporal and spatial coordination of cellular events such as actin assembly/disassembly and adhesion formation/detachment. PDGF appears to stimulate actin filament reorganization through activation of the small GTP-binding proteins Cdc42, Rho, and Rac (19 -21). Similarly, several proteins localized at focal adhesions such as focal adhesion kinase, paxillin (22,23), and talin (24) are tyrosine-phosphorylated following PDGF treatment, whereas another cytoskeletal protein, vinculin, has been found to redistribute from focal adhesions to the perinuclear regions upon PDGF administration. In addition, activated PDGF-␤ receptors have been found to associate with the ␣v␤3 integrin, and this association potentiates both mitogenic and chemotactic responses induced by PDGF (25). Conversely, the activation of integrins can also lead to the modulation of PDGF receptors. For example, extracellular matrix (ECM) components, either alone (26) or together with PDGF (27), can promote the tyrosine phosphorylation of PDGF-␤ receptors. Moreover, PDGF-induced phosphatidylinositol 4,5-bisphosphate hydrolysis and Ca 2ϩ mobilization have been found to induce actin filament disassembly (28), consistent with the ability of PDGF to relax stress fibers and to induce edge membrane ruffles (23,29).
Another member of the receptor tyrosine kinase family, the EGF receptor, has been shown to share many structural and functional similarities with the PDGF receptor. In a previous study from our laboratory, we reported that EGF can stimulate integrin-mediated migration toward various ECM components in mouse B82L fibroblasts transfected with the human EGF receptor (1). Co-presentation (co-positioning) of EGF with laminin or fibronectin is essential for EGF-stimulated migration (1). The expression of functional EGF receptors, including an intact kinase domain, C terminus, and the intracellular juxtamembrane region, is necessary not only for the EGF-stimulated enhancement of cell migration but also for cell migration toward fibronectin alone (1). 2 The importance of the EGF receptor for conferring B82L cell motility is further revealed by the inhibition of fibronectin-and laminin-induced cell migration by the addition of neutralizing anti-EGF receptor antibodies, suggesting that the EGF receptor may act downstream of integrin activation and may be engaged in the signal transduction events leading to cell motility (1,30).
Because PDGF initiates many of the same signal transduction events as EGF (2,7) and has been shown to be chemotactic for a number of cell types (15)(16)(17)(18), together with the observation that the EGF receptor appears essential for integrin-mediated cell migration in certain fibroblasts (1), we evaluated the effects of PDGF on fibronectin-induced migration using mouse B82L fibroblasts transfected with various EGF receptor constructs. We observed that PDGF can enhance fibronectin-induced cell migration to the same extent as EGF, but that this enhancement of cell migration occurs only in cells expressing wild-type EGF receptors. PDGF did not stimulate the migration of B82L cells that either do not express EGF receptors or that express EGF receptors that are kinase-inactive or C-terminally truncated. However, PDGF can induce the tyrosine phosphorylation of the PDGF-␤ receptor and can result in MAP kinase (ERK1 and ERK2) activation in all of these cells, indicating that they are responsive to PDGF in the absence of an intact EGF receptor. Furthermore, the critical role of the EGF receptor in PDGF stimulation of cell motility is supported by the observation that the EGF receptor is tyrosine-phosphorylated upon PDGF treatment in B82L cells expressing wild-type EGF receptors.
Cells and Culture Conditions-Mouse B82L-parental fibroblasts, which contain no detectable endogenous EGF receptor, were transfected with various EGF receptor constructs (see Fig. 1). B82L-parental fibroblasts were maintained in Dulbecco's modified Eagle's medium containing 10% cosmic calf serum (Hyclone, Logan, UT). Cells transfected with wild-type or mutated human EGF receptors were cultured in the same medium, except that it contained 10 M methotrexate because a mutant dihydrofolate reductase gene was used as a selectable marker (32). The B82L-clone B3 cells, which were derived from B82L cells transfected with wild-type EGF receptors based on their ability to bind laminin (30), were cultured in the same medium as the B82L transfectants.
Cell Motility Assays-The experiments assessing cell motility were performed using a 48-well chemotaxis chamber as described previously (1,30). In these studies, cells were grown for 3-4 days as described under "Cells and Culture Conditions." After being detached from the plastic dishes using a 0.1% trypsin solution, the cells were stabilized for 1 h at 37°C in Dulbecco's modified Eagle's medium containing 0.1% bovine serum albumin. The cells were then counted and resuspended in Dulbecco's modified Eagle's medium plus 0.1% bovine serum albumin at a final concentration of 1 ϫ 10 5 cells/50 l. The lower compartment of the migration chamber was filled with the indicated proteins dissolved in Dulbecco's modified Eagle's medium ϩ 0.1% bovine serum albumin (29 l/well), and the cells were added to the upper compartment of the migration chamber (50 l/well). To ascertain the effects of PDGF or EGF, the growth factor was added either to the upper or to the lower compartments at indicated concentrations. The two compartments of the migration chamber were separated by a polycarbonate filter (5 m). The cells were allowed to migrate for 4 h at 37°C in a humidified atmosphere containing 5% CO 2 . The cells that did not migrate through the membrane were located on the upper surface of the filter and were removed mechanically by scraping; the migrant cells on the lower surface were fixed in methanol/acetone (1:1) for 2 min and then stained with 1% crystal violet. The filters were densitometrically analyzed using the OFOTO program (Light Source Computer Images, Inc.) and quantified by Scanner Analysis (Biosoft, Ferguson, MO).
Analysis of MAP Kinase Activation-Cells expressing various EGF receptor constructs were serum-starved for 4 h before they were treated with PDGF (10 or 100 ng/ml), EGF (1 nM), or the control buffer for 5 min at 37°C. Cells were then lysed with 2ϫ sample buffer (2% SDS, 20 mM dithiothreitol, 1 mM Na 3 VO 4 , 2 mM EDTA, 10% glycerol, 20 mM Tris, pH 8.0) and boiled for 5 min. The protein content in each sample was determined by microBCA assays. The samples were separated by 7.5% SDS-polyacrylamide gel electrophoresis using equal amounts of protein loading (60 g of protein/sample). The proteins were then transferred to a polyvinylidene difluoride membrane and blocked with 5% nonfat milk dissolved in TBST overnight at 4°C. The membrane was incubated 1 h at 37°C with antibodies directed against active MAP kinases ERK1 and ERK2. After being washed 3 ϫ 10 min with TBST, the blots were incubated 1 h at 37°C with horseradish peroxidase-conjugated antimouse secondary antibodies and washed at room temperature 6 times during a period of 2 h. Bound antibodies were visualized by chemiluminescent detection method.
Immunoprecipitation and Immunoblotting-Cells expressing various EGF receptor constructs were serum-starved for 4 h before they were treated with porcine PDGF-BB or recombinant rat PDGF-BB at 37°C, using the concentrations and incubation time indicated in the figure legends. The cells were then lysed in radioimmunoprecipitation assay buffer (1% Nonidet P-40, 0.25% deoxycholate, 1 mM Na 3 VO 4 , 1 mM phenylmethysulfonyl fluoride, 1 g/ml leupeptin, 1 g/ml aprotinin, 1 mM EDTA, 1 mM EGTA, 150 mM NaCl, 10 mM Tris, pH 8.0), and the cell lysates were centrifuged at 10,000 ϫ g for 10 min. The supernatants were subjected to microBCA protein analysis, and the same amount of protein was taken from each sample for the subsequent immunoprecipitation. The cell lysates were precleared with protein A-conjugated agarose beads and then immunoprecipitated with polyclonal antibodies against the PDGF-␤ receptor cytoplasmic tail or with polyclonal antibodies against a fragment of the EGF receptor cytoplasmic domain, followed by incubation with protein A-conjugated agarose beads. The immunoprecipitates were subjected to SDS-polyacrylamide gel electrophoresis, and the proteins were transferred to a polyvinylidene difluoride membrane. The membrane was blocked in 0.25% gelatin in TBST overnight at 4°C and then incubated 1 h at 37°C with antibodies against phosphotyrosines (a mixture of 0.2 g of Upstate Biotechnology antibody and 0.2 g Santa Cruz antibody in 5 ml 0.25% gelatin, TBST). After being washed three times for 10 min with TBST, the blots were incubated for 1 h at 37°C with horseradish peroxidase-conjugated anti-mouse secondary antibodies followed by washing 6 times at room temperature over a period of 2 h. The bound antibodies were visualized using chemiluminescence detection.
After probing with anti-phosphotyrosine antibodies, the same membrane was stripped with 2% SDS and 100 mM dithiothreitol in 62.5 mM Tris, pH 6.8, for 30 min at 70°C, washed extensively with TBST, and then blocked in 5% nonfat milk. After checking the stripping efficiency by subjecting the blot to incubation with horseradish peroxidase-conjugated secondary antibodies and chemiluminescent detection, the blot was reprobed with anti-PDGF-␤ receptor or anti-EGF receptor antibodies. After washing three times for 10 min with TBST, the blots were incubated for 1 h at 37°C with horseradish peroxidase-conjugated anti-mouse secondary antibodies, followed by washing 6 times at room temperature over a period of 2 h. The bound antibodies were visualized using chemiluminescence detection.
For the quantification of PDGF-induced EGF receptor tyrosine phosphorylation, the film was densitometrically analyzed using the OFOTO program and quantified by Scanner Analysis. For the statistical analysis of the data, the ratio of phosphotyrosine to EGF receptor mass after subtracting background readings was calculated for each sample. Variance comparison of the ratio values before and after PDGF treatment indicates that PDGF does not change the distribution of the variance (p value Ͼ 0.2). These ratio values were then compared using a paired-t test.

Influence of PDGF on the Fibronectin-induced Migration of Cells Expressing Wild-type EGF Receptors-
We have shown previously that EGF can synergize with various ECM components in promoting the integrin-mediated migration of murine B82L fibroblasts transfected with wild-type EGF receptors (B82L-wt) and of a clone isolated from B82L-wt that possesses laminin binding activity (B82L-clone B3) (see Fig. 1 for receptor constructs). In these earlier studies, the synergistic action of EGF was found to require the co-positioning of EGF and the chemoattractant, such as laminin and fibronectin, in the lower wells of the 48-well chemotaxis chamber (1). Another extensively studied growth factor, PDGF, initiates many of the same signal transduction events as EGF (2, 7) and is a potent chemoattractant for human neutrophils, monocytes (17), fibroblasts (15,16), and smooth muscle cells (18). To investigate whether PDGF can regulate integrin-mediated migration in a manner similar to EGF, we examined the effects of PDGF on fibronectin-induced migration of B82L-clone B3 cells using a 48-well chemotaxis chamber. PDGF-BB was used in all these studies due to its ability to bind to and activate all PDGF receptor isoforms, and thus, induces a strong chemotactic response (2,3,5,7). PDGF at different doses was mixed with fibronectin and placed in the lower wells of the chemotaxis chamber. B82L-clone B3 cells were then added to the upper wells in the presence or absence of PDGF. As shown in Fig. 2, PDGF alone at all concentrations tested had little detectable effect on cell migration. However, PDGF synergized with fibronectin in promoting the migration of B82L-clone B3 cells; this effect of PDGF was dose-dependent, and maximal migration occurred in the presence of approximately 30 ng/ml PDGF. The highest extent of cell migration stimulated by PDGF was similar to that stimulated by EGF, which was about a 2-fold increase compared with cell migration toward fibronectin alone. Interestingly, the synergistic action of PDGF on fibronectin-induced chemotaxis was only achieved when PDGF was co-present with fibronectin in the lower wells of the migration chamber. When PDGF was present in the upper wells together with cells, it had little effect on the migration as compared with fibronectin alone. These effects of PDGF on fibronectin-induced migration in B82L-clone B3 cells were comparable with those of EGF (1,30) in terms of the maximal stimulation elicited by the growth factors and the requirement for co-presentation of the growth factor and fibronectin.

Effect of PDGF on the Fibronectin-induced Migration of B82L Fibroblasts Expressing Various EGF Receptor Constructs-
B82L-clone B3 cells were derived from B82L-parental cells that have no measurable endogenous EGF receptors (30) (Fig. 1). Our previous studies have indicated that the B82L-parental cells exhibit no response to EGF treatment and do not migrate toward fibronectin alone (1), although they express functional fibronectin receptors and can adhere to a fibronectin-coated surface (30). These studies also indicated that the lack of ECMstimulated motility in B82L-parental cells was attributable to the absence of intact EGF receptor expression (1). To assess whether the inability of B82L-parental cells to migrate to fibronectin alone could be overcome by PDGF administration, we examined the capacity of B82L-parental cells to migrate toward fibronectin in the presence of PDGF. As shown in Fig. 3, PDGF at all the concentrations tested (0ϳ100 ng/ml), whether present in the upper or lower wells, could not stimulate fibronectin-induced migration of B82L-parental cells. These results indicate that PDGF treatment of B82L-parental cells is not sufficient to induce their migration, and based on the data shown in Figs. 2 and 3, the expression of a functional EGF receptor is critical for the PDGF-stimulated increase in the migration of B82L-clone B3 cells.
Ligand-induced EGF receptor autophosphorylation initiates FIG. 1. Schematic representation of the human EGF receptor constructs that were expressed in mouse B82L-parental fibroblasts (which contain no detectable endogenous EGF receptor). A, wild-type (full-length) EGF receptor (B82L-wt); B82L-clone B3 cells also express the wild-type EGF receptor and are a clone isolated from B82L-wt cells based on their capacity to bind laminin (30). B, kinaseinactive EGF receptor that contains a lysine to methionine substitution at residue 721 (B82L-K721M). C, kinase-active EGF receptor that has been truncated at residue 973 (B82L-cЈ973). This mutant lacks all the major EGF receptor tyrosine autophosphorylation sites, which are designated by a Y .   FIG. 2. The influence of PDGF and fibronectin on the migration of B82L-clone B3 fibroblasts. Cell migration toward fibronectin was determined using a 48-well microchemotaxis chamber as discussed under "Materials and Methods." In the experiments using fibronectin as the chemoattractant, 100 g/ml fibronectin was placed in the lower wells. PDGF (0 -100 ng/ml) was mixed either with the fibronectin (FN) in the lower wells (diamonds) or with clone B3 fibroblasts in the upper wells (circles). Cell motility in response to PDGF alone (placed in lower wells) was also measured (triangles). EGF (1 nM), which has been shown to stimulate fibronectin-induced migration in clone B3 cells, was used as a positive control (bars) (1). In all cases, the data points represent the mean Ϯ S.D. of triplicate determinations. Similar results were obtained in three different experiments. a cascade of signaling events leading to cellular responses such as cell proliferation and cell motility (33). Both the receptor kinase activity and the receptor C terminus, which provides docking sites for Src homology 2 or phosphotyrosine binding domain containing cellular proteins, are critical for EGF-induced signal transduction (34). To map the EGF receptor elements that are critical for the synergistic stimulation of PDGF on integrin-mediated motility, B82L-parental cells were transfected EGF receptors that are kinase-inactive (B82L-K721M) or C-terminally truncated (B82L-cЈ973) (Fig. 1). Lysine residue at 721 within the kinase domain is believed to be involved in ATP binding, and replacement of this lysine with methionine abrogates receptor tyrosine kinase activity. As shown in Fig.  4A, EGF receptor kinase-inactive cells (B82L-K721M) did not migrate toward fibronectin in the presence or absence of PDGF. Similar results were obtained with cells expressing EGF receptors that lack all five C-terminal autophosphorylation sites (B82L-cЈ973) (Fig. 4B), which exhibit increased intracellular tyrosine kinase activity due to the removal of a C-terminal autoinhibitory restraint (35,36). The observation that the migratory behavior of B82L-K721M or B82L-cЈ973 cells is indistinguishable from B82L-parental cells suggests that both the EGF receptor tyrosine kinase activity and its C-terminal region are important for the PDGF-induced enhancement of cell motility in B82L fibroblasts. Taken together, these studies demonstrate that the PDGF-stimulated increase of fibronectin-induced migration in B82L fibroblasts depends on the expression of functional EGF receptors. Cells that do not carry fully functional EGF receptors, such as B82L-parental, K721M, and cЈ973 cells, do not show PDGF-stimulated cell migration. Nor do they migrate toward fibronectin alone.
Expression and Ligand-dependent Activation of the PDGF-␤ Receptor in B82L Cell Lines-To test whether the differences in the PDGF-induced migratory response of B82L fibroblasts expressing various EGF receptor constructs are due to altered PDGF receptor expression or function, we examined the tyrosine phosphorylation of PDGF-␤ receptors immunoprecipitated from the cell lysates of B82L-parental, clone B3, K721M, and cЈ973 fibroblasts treated with or without PDGF (50 ng/ml). As shown in Fig. 5A, anti-phosphotyrosine blotting revealed that the levels of basal and ligand-stimulated PDGF-␤ receptor In the experiments using fibronectin (FN) as the chemoattractant, 100 g/ml fibronectin was placed in the lower wells. PDGF (0 -100 ng/ml) was mixed either with the fibronectin in the lower wells (diamonds) or with B82L-parental fibroblasts in the upper wells (circles). Cell motility induced by PDGF alone (placed in lower wells) was also measured (triangles). B82L-clone B3 fibroblasts were used as a positive control for cell motility (bars) (1). Each data points represent the mean Ϯ S.D. of triplicate determinations. Similar results were obtained in three different experiments.

FIG. 4. The influence of the EGF receptor tyrosine kinase activity and C terminus on the PDGF-stimulated migration of B82L fibroblasts toward fibronectin. B82L-parental cells express-
ing no endogenous EGF receptors were transfected with an EGF receptor that is kinase-inactive (B82L-K721M) (A) or a kinase-active EGF receptor that has been C-terminally truncated at residue 973 (B82L-cЈ973) (B). The migration of B82L cells toward medium alone, PDGF alone (10 ng/ml), fibronectin (FN, 100 g/ml) alone, and fibronectin (100 g/ml) and PDGF (10 ng/ml) co-positioned in the lower wells or fibronectin in the lower well and PDGF in the upper well (co-positioned with the cells) was measured. Each bar represents the mean Ϯ S. D. of triplicate determinations. Similar results were obtained in three different experiments. tyrosine phosphorylation were similar in all four cell lines. Re-probing of the same membrane with anti-PDGF-␤ receptor antibody showed that all four cell lines exhibit similar levels of PDGF-␤ receptor expression (Fig. 5B). These results indicate that B82L-parental, clone B3, K721M, and cЈ973 fibroblasts express comparable levels of functional PDGF-␤ receptors, and the differences in their migration are not likely the result of variations in PDGF-␤ receptor level or tyrosine phosphorylation.
PDGF-stimulated MAP Kinase Activation in B82L Cell Lines-To determine if the PDGF receptors in various B82L cells can transduce intracellular signals other than receptor phosphorylation later, we examined PDGF-stimulation of MAP kinase activation. As shown in Fig. 6, the MAP kinases (ERK1/ ERK2) were activated by PDGF (10 ng/ml and 100 ng/ml) in all four cell lines, suggesting that the PDGF receptors are functional in these cell lines, and transfection of the wild-type or mutated EGF receptor does not alter the PDGF receptor responsiveness in B82L cells. It is noteworthy that B82L-parental cells appear to be less responsive at 10 ng/ml PDGF than the three EGF receptor transfectants. This could be linked to a potential role of the EGF receptor in PDGF-induced intracellular signaling, a hypothesis supported by the evidence presented below that demonstrates that the EGF receptor can be tyrosine-phosphorylated by PDGF administration. In addition, the ability of kinase-inactive (B82L-K721M) or C-terminally truncated (B82L-cЈ973) EGF receptor to mediate EGF-elicited ERK1/ERK2 activation, as shown in Fig. 6, is consistent with previous reports (37,38). For example, Wright et al. (38) observe that upon EGF stimulation of B82L cells expressing EGF receptor mutant K721M, ERK2 and ERK1 MAP kinases are activated. Heterodimerization with and activation of the endogenous ErbB2 is a possible mechanism by which kinasedefective receptors stimulate the MAP kinase pathway, because kinase-defective receptors can induce increased ErbB2 enzymatic activity and binding to Shc (38). In sum, our present observations indicate that certain PDGF actions in B82L fibroblasts, such as MAP kinase activation, are not largely affected by EGF receptor transfection, and that the lack of PDGFstimulated cell motility in B82L-parental, K721M, and cЈ973 cells does not appear to result from a substantial alteration of PDGF receptor responsiveness.
EGF Receptor Tyrosine Phosphorylation in Response to PDGF Treatment in B82L-clone B3 Fibroblasts-Because functional EGF receptors are linked to cell migration toward fibronectin alone as well as to PDGF-stimulated cell motility, PDGF may be acting, at least in part, through the EGF receptor to induce motility of B82L cells. To test the hypothesis that the EGF receptor can act as a downstream mediator in PDGFstimulated signal transduction pathways, we examined the tyrosine phosphorylation state of the EGF receptor in PDGFstimulated B82L clone B3 cells. As shown in Fig. 7A, which contains data from three separate experiments, we observed that the stimulation of B82L-clone B3 cells with PDGF (50 ng/ml) resulted in a readily detectable increase in the phosphotyrosine content of the EGF receptors. Densitometric quantification of these data indicates that the PDGF-stimulated tyrosine phosphorylation of the EGF receptor is significantly increased over control levels (p value Ͻ 0.002) (Fig. 7B). This alteration in the EGF receptor phosphorylation state is not due to a change in the receptor levels, given that PDGF does not appear to have effects on EGF receptor levels (Fig. 7A, lower  panel). The specificity of the anti-EGF receptor immunoprecipitations performed in these experiments was confirmed by using B82L-parental cells that lack EGF receptors, i.e. no EGF receptors were immunoprecipitated from B82L-parental cells (data not shown). These observations suggest that EGF receptor may be involved in mediating certain signal transduction events initiated by PDGF. However, we were unable to detect the co-precipitation of the EGF receptor with the PDGF receptor using B82L-clone B3 cell extracts (data not shown), suggesting that the EGF receptor may not be a direct substrate of PDGF-␤ receptor kinase. Furthermore, this observation supports that the PDGF-induced tyrosine phosphorylation of the EGF receptor (approximately 170 kDa) we observed in Fig. 7A is not likely to result from the possible co-precipitation of the PDGF receptor (approximately 180 and 160 kDa) in the similar size range.
The PDGF used in the aforementioned experiments was isolated from porcine platelets to greater than 97% purity. However, to eliminate the possibility that the PDGF-induced tyrosine phosphorylation of the EGF receptor is due to EGF contamination of the PDGF preparation, B82L-clone B3 cells were also treated with recombinant rat PDGF-BB. We observed that recombinant rat PDGF could also stimulate the migration of these cells by 2.89 Ϯ 0.01 fold (data not shown). Furthermore, as illustrated in the inset of Fig. 7B, recombinant rat PDGF (30 ng/ml) could also induce the tyrosine phosphorylation of the EGF receptor in B82L-clone B3 cells. The induction of EGF receptor phosphorylation by recombinant rat PDGF was transient, detectable by 1 min after PDGF addition, and was maximal at 10 min. The amount of phosphorylation declined to slightly above the basal level by 2 h after recombinant rat PDGF addition (Fig. 8). These observations indicate that PDGF can stimulate the tyrosine phosphorylation of the EGF receptor in mouse fibroblasts in a time-dependent manner.

DISCUSSION
In the present study, we have observed that the stimulatory effect of PDGF on the fibronectin-induced migration of B82L fibroblasts is dependent on the expression of an intact EGF receptor and that the EGF receptor is tyrosine-phosphorylated FIG. 5. PDGF-induced tyrosine phosphorylation of the PDGF-␤ receptor in B82L fibroblasts expressing various EGF receptor constructs. All four B82L cell lines (B82L-parental, clone B3, K721M, and cЈ973) were treated with PDGF (50 ng/ml) for 5 min at 37°C before they were lysed in radioimmunoprecipitation assay buffer. Cell lysates containing equal amounts of protein (0.5 mg) were subjected to immunoprecipitation using an anti-PDGF-␤ receptor antibody. A, the immunoprecipitates were probed with antiphosphotyrosine (PY) antibodies as described under "Materials and Methods." B, the blot shown in panel A was stripped and reprobed with an anti-PDGF-␤ receptor (PBGFR-␤) antibody that recognizes both the mature (180 kDa) and immature (160 kDa) form of PDGF receptor ␤ subunits.
in response to PDGF treatment. It has been well documented that PDGF is a potent chemoattractant for several cell types (15)(16)(17)(18), and our finding that cell migration in response to PDGF treatment requires an intact EGF receptor provides further evidence for cross-regulation between these two growth factor receptors.
Previous investigations have revealed a functional interaction between EGF and PDGF receptor action with regard to the mitogenesis of several cell types. In BALB/c-3T3 cells, different phases of the cell cycle can be regulated by various growth factors (39 -41). For example, a brief exposure of cells to PDGF induces BALB/c-3T3 cells to leave G 0 and enter the cell cycle, whereas EGF facilitates progression through G 0 /G 1 and entry into S phase (42). In addition, PDGF stimulates porcine theca cell proliferation in a dose-dependent manner, but although EGF alone is not mitogenic for these cells, it can markedly enhance the proliferative capacity of PDGF (43). Conversely, in granulosa cells, EGF-or transforming growth factor-␣-stimu-FIG. 6. PDGF-induced MAP kinase activation in B82L fibroblasts expressing various EGF receptor constructs. All four B82L cell lines (B82L-parental, clone B3, K721M, and cЈ973) were treated with PDGF (10 or 100 ng/ml) or EGF (1 nM) for 5 min at 37°C before they were lysed as described under "Material and Methods." Equal amounts of protein (30 g) were separated by SDS-polyacrylamide gel electrophoresis and immunoblotted (IB) with a polyclonal antibody directed against active ERK1/ERK2. Similar results were obtained in three separate experiments. 1 and 2, or 3 and 4, or 5 and 6) were examined. Serum-starved B82L-clone B3 cells were treated with control buffer (lanes 1, 3, and 5) or PDGF (50 ng/ml) (lanes 2, 4, 6, and 7) for 5 min at 37°C before they were lysed. Cell lysates were normalized within each treatment pair for their protein content before they were immunoprecipitated using an anti-EGF receptor antibody. Additionally, lane 7 contains nonspecific rabbit IgG as a negative control for the anti-EGF receptor immunoprecipitation step. The immunoprecipitates were probed with antiphosphotyrosine (PY) antibodies as described under "Materials and Methods." The same blot was stripped and reprobed with the anti-EGF receptor antibodies. B, the amount of phosphotyrosine and EGF receptor mass in the samples shown in panel A were densitometrically quantified, and the ratio of phosphotyrosine content to total EGF receptor mass before and after PDGF treatment was calculated. Each bar represents the mean ratio (ϮS.D.) of three separate measurements. Statistical analysis revealed a significant difference (p Ͻ 0.002) between control and PDGF treatment. Inset, serum-starved B82L-clone B3 were treated, analyzed, and densitometrically quantified as described for panels A and B, except that recombinant rat PDGF (rr-PDGF); 30 ng/ml) was used. Similar observations were made in three separate experiments.  1-7), The immunoprecipitates (IP) were probed with antiphosphotyrosine (PY) antibodies as described under "Materials and Methods." The same blot (IB) was stripped and reprobed with the anti-EGF receptor antibodies. In lane 8, nonspecific rabbit IgG was used as a negative control for the anti-EGF receptor immunoprecipitation step. Similar results were obtained in three separate experiments. lated proliferation is enhanced by PDGF (42)(43)(44). Interestingly, a synergistic action between EGF and PDGF is not limited to mitogenesis, i.e. PDGF sensitizes BALB/c-3T3 fibroblasts to EGF action with regard to the redistribution of vinculin from focal contacts to the perinuclear region of the cells, whereas EGF alone cannot induce vinculin loss from focal adhesions. In comparison, exposure of these cells to PDGF at concentrations that do not induce vinculin redistribution allows for EGF to induce vinculin redistribution (45).

FIG. 7. EGF receptor tyrosine phosphorylation following PDGF treatment in B82L fibroblasts expressing wild-type EGF receptors. A, three independent treatment pairs (lanes
PDGF-stimulated tyrosine phosphorylation of the EGF receptor, as shown in the present studies via the use of both purified porcine PDGF-BB and recombinant rat PDGF-BB, may provide one potential mechanism by which PDGF and EGF receptors cooperate to promote cell motility in B82L fibroblasts. Because PDGF does not appear to bind to the EGF receptor and because ligand-dependent tyrosine phosphorylation of the PDGF-␤ receptors is observed in B82L-clone B3 cells, it appears that PDGF-induced tyrosine phosphorylation of the EGF receptor can be initiated by the activated PDGF-␤ receptor. This conclusion is consistent with previous reports demonstrating that both PDGF and EGF receptors are colocalized in the highly specialized plasma membrane fraction, caveolae (46,47). In addition, Habib et al. (48) show that the PDGF receptor and the EGF receptor form a constitutive physical complex in several cell types (48), although we were unable to detect complex formation between these two receptors in our systems using immunoprecipitation approach. In these earlier studies (48), it was also demonstrated that the addition of EGF results in the rapid tyrosine phosphorylation of PDGF-␤ receptors and the recruitment of phosphatidylinositol 3-kinase to the PDGF-␤ receptor. However, the authors did not observe any increase in EGF receptor tyrosine phosphorylation in response to PDGF (48), perhaps due to the different cell types that were used or because of issues related to assay sensitivity. Nonetheless, the present investigations suggest that the EGF receptor may either be a direct substrate of the PDGF-␤ receptor in B82L cells or that other signaling molecules, such as Srcrelated kinases or phosphotyrosine phosphatases such as phosphotyrosine phosphatase 1B, may contribute to the PDGFstimulated increase in EGF receptor phosphotyrosine content.
Several models may explain why EGF receptor expression can convert a nonmigratory cell type to a migratory one (whereas the PDGF receptor cannot) as well as why PDGF relies on the EGF receptor to induce cell migration. One feature of the EGF receptor that may be linked to this process is the capacity of the receptor to associate with intracellular actin microfilaments. The EGF receptor co-purifies with F-actin from A431 cells (49) and binds to both G-and F-actin directly in in vitro studies. 3 Moreover, EGF treatment shifts a population of EGF receptors with high affinity for EGF from the Triton X-100soluble membrane fraction to the detergent-insoluble actincontaining cytoskeleton fraction (50,51). In addition, the EGF receptor contains an actin binding motif (DDVVDADEYLIPQ) at its C-terminal region (52), whereas neither PDGF receptor isoform appears to have such a sequence (53,54). Given that cell migration is a process involving dynamic and coordinated actin polymerization and depolymerization, the direct association between F-actin and the EGF receptor may in part account for the dependence of B82L cell migration on the EGF receptor but not the PDGF receptor. Additionally, it has been shown that the EGF receptor can be activated by stimulating the receptors for ECM components, namely the integrins (55). Several of the integrin-elicited intracellular signaling events are regulated by the EGF receptor through mechanisms that are dependent on PKC, Src, and PTP1B (56,57), suggesting a role for the EGF receptor in the modulation of cell motility.
Another possibility for the differential capacity of EGF and PDGF receptors to confer cell motility may come from the existence of two similar yet distinct PDGF receptors, i.e. the ␣ and ␤ receptor isoforms. It has been shown that only cells expressing the ␤ receptor, but not the ␣ receptor, are able to migrate chemotactically toward a concentration gradient of PDGF and form circular membrane ruffles, although both receptors are able to induce a loss of stress fibers and appearance of edge ruffles (9,29). Other studies indicate that the ␣ receptor is capable of mediating chemotaxis as well as the ␤ receptor but that the chemotactic response induced by ligand binding to PDGF-␤ receptor may be different from that induced by the ligand binding to the ␣ receptor form (8). Because PDGF-BB, which binds and activates both the ␣ and ␤ receptor (5,7), was used in our study, we cannot determine whether both receptor isoforms are expressed in B82L fibroblasts and involved in the signaling events that lead to cell motility. It is possible that a particular isoform of the PDGF receptor may require the EGF receptor whereas the other does not. Lastly, it is conceivable that differences in structure or substrate specificity between the EGF receptor and the PDGF receptor can account for their distinct ability in conferring cell mobility.
Our finding that B82L fibroblasts require the expression of an intact EGF receptor to respond to PDGF with respect to fibronectin-induced cell motility is likely to be important in physiological processes such as wound healing (4). Upon injury, platelets and subsequently monocytes/lymphocytes are deposited at the site of injury. PDGF secreted by platelets, monocytes/macrophages, and possibly injured endothelial cells, initiate the repair process by stimulating chemotaxis and proliferation of fibroblasts at the fibronectin-fibrin wound interface where granulation tissue forms (4). In addition, the locally increased PDGF may also stimulate the synthesis of a new collagen-containing matrix, and expression of other growth factors such as transforming growth factor-␤ (4, 31). The requirement for the presence of the EGF receptor for fibroblast migration in response to PDGF may thus provide an additional control on the mechanisms associated with the wound-healing process by preferentially promoting the attraction of fibroblasts to the sites of injury. In addition to wound healing, interaction between the EGF receptor and the PDGF receptor in the control of cell motility may have a broad impact on other physiological and pathological events, where regulation of cell migration is a key element. Such events may include organ development, tumor metastasis, arteriosclerosis, and inflammatory responses.