Vinculin Phosphorylation in Response to Calcium and Phorbol Esters in Intact CeIls*

Vinculin phosphorylation in both chick embryo fi- broblasts and Swiss 3T3 cells was increased by either calcium or biologically active phorbol esters. Increased phosphorylation of vinculin was noted as early as 10 min following phorbol 12-myristate 13-acetate treatment and was maximal at about 1 h. Maximal increases in phosphorylation were noted at approximately 100 nM phorbol 12-myristate 13-acetate. Phorbol 12,13-dibutyrate (80 nM), a less potent phorbol ester, resulted in smaller increases in vinculin phosphorylation than phorbol 12-myristate 13-acetate at equimolar concen- trations. Phorbol, dibutyryl CAMP, and dibutyryl cGMP had no significant effect on phosphorylation. No correlation was found between vinculin phosphorylation and the morphological changes induced by phorbol esters. Tryptic peptide analysis of vinculin revealed multisite phosphorylation. Phosphorylation of only three of the peptides was significantly increased following phorbol 12-myristate 13-acetate treatment. Phosphoamino acid analysis revealed increases at both serine and threonine residues. The low level of phosphotyrosine present in control cells was not signifi- cantly increased by phorbol 12-myristate 13-acetate treatment. These findings combined with studies of vinculin phosphorylation by purified protein kinase C (Werth, D. K.,

Vinculin phosphorylation in both chick embryo fibroblasts and Swiss 3T3 cells was increased by either calcium or biologically active phorbol esters. Increased phosphorylation of vinculin was noted as early as 10 min following phorbol 12-myristate 13-acetate treatment and was maximal at about 1 h. Maximal increases in phosphorylation were noted at approximately 100 nM phorbol 12-myristate 13-acetate. Phorbol 12,13dibutyrate (80 nM), a less potent phorbol ester, resulted in smaller increases in vinculin phosphorylation than phorbol 12-myristate 13-acetate at equimolar concentrations. Phorbol, dibutyryl CAMP, and dibutyryl cGMP had no significant effect on phosphorylation. No correlation was found between vinculin phosphorylation and the morphological changes induced by phorbol esters. Tryptic peptide analysis of vinculin revealed multisite phosphorylation. Phosphorylation of only three of the peptides was significantly increased following phorbol 12-myristate 13-acetate treatment. Phosphoamino acid analysis revealed increases at both serine and threonine residues. The low level of phosphotyrosine present in control cells was not significantly increased by phorbol 12-myristate 13-acetate treatment. These findings combined with studies of vinculin phosphorylation by purified protein kinase C (Werth, D. K., Niedel, J. E., and Pastan I. (1983) J . Biol. Chem. 258, 11423-11426) suggest the hypothesis that protein kinase C may be involved in regulation of phosphorylation of vinculin, a cytoskeletal protein.
Phorbol esters have been shown to produce many biological effects (I), but the mechanisms responsible for their diverse effects are only beginning to be elucidated. Recently, it was shown that phorbol esters activate protein kinase C, the calcium, phospholipid-activated protein kinase (2, 3). Activation of protein kinase C with subsequent phosphorylation of specific substrate proteins could provide a mechanism by which phorbol esters produce multiple biochemical effects.
One of the biological effects of phorbol esters in cells is to produce morphological changes which mimic those seen in transformed cells (4, 5 ) . One protein that has been postulated to have a role in the morphological changes seen in Rous sarcoma virus transformation is vinculin. Vinculin, a cytoskeletal protein, has been found to be a substrate for pp6O3",' * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisernent" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
-. _ _ _ _ -______ ______ the transforming protein of the Rous sarcoma virus (6, 7). Using purified proteins, we have recently found that vinculin is a better substrate for protein kinase C than several other serine-and threonine-specific protein kinases (8). In addition, biologically active phorbol esters were found to stimulate vinculin phosphorylation by protein kinase C. These findings could establish a link between the morphological changes induced by phorbol esters and those seen following cellular transformation by viruses. Studies involving phosphorylation with purified components allow detailed investigation of enzymatic properties, but the results of such studies may not accurately reflect physiological responses. Another approach to the study of protein phosphorylation is to metabolically label intact cells with 'j2P1 and examine phosphorylation in response to various agents. In the present study, we have examined vinculin phosphorylation in intact cells in response to phorbol esters, cyclic nucleotides, and calcium. We have also compared the effects of phorbol esters on cellular morphology with their effects on vinculin phosphorylation.

EXPERIMENTAL PROCEDURES
Cells and Radioactive Labeling-Chick embryo fibroblasts were prepared and grown at 39 "C as previously described (9). Third passage chick embryo fibroblasts were plated a t a density of 8 X lo6 cells/lOO-mm dish or 4.8 X 1O6cells/60-mm dish and grown overnight. Cell monolayers were washed three times with Tris-buffered saline (0.15 M NaCI, 20 mM Tris-HC1, pH 7.2). Phosphoproteins were labeled by incubation with 0.5-1.0 mCi of ["P]orthophosphate/ml of media in phosphate-free Dulbecco's modified Eagle's medium supplemented with 5% heated calf serum, L-glutamine (2 mM), penicillin (50 units/ml), and streptomycin (50 rg/ml). Cells were labeled for 8 h at 37 "C. [35S]Methionine-labeled cells were prepared by incubation for 8 h a t 37 "C with 250 WCi/ml of [35S]methionine in methioninefree Eagle's media supplemented as for the 3' P labeling.
Treatment with phorbol esters was performed for the time periods indicated during the last portion of the labeling period. Phorbol esters, dibutyryl CAMP, and dibutyryl cGMP were dissolved in dimethyl sulfoxide. The final concentration of dimethyl sulfoxide was 0.2%. Control dishes were treated with the same concentration of dimethyl sulfoxide. Calcium concentration in the supplemented media (1.6 mM) was determined using a Perkin-Elmer model 5000 atomic absorption spectrophotometer. To obtain the desired calcium concentration, appropriate microliter amounts of stock solutions of 0.1 M EGTA and 0.05 M CaCL were added. Following the labeling and treatment periods, the cell monolayers were washed three times with ice-cold Tris-buffered saline containing 1% aprotinin and 1 mM EDTA. Cell lysates were prepared by adding modified RIPA buffer (Tris-buffered saline containing 1% Nonidet P-40, 0.5% deoxycholate, 1 mM EDTA, and 1% aprotinin) a t a concentration of 8 X lo6 cells/ml. The lysates were clarified by centrifugation at 100,000 X g for 30 min prior to immunoprecipitation.
Cells were plated a t a density of IO6 cells/lOO-mm dish. Labeling, treatment, and preparation of cell lysates were as descrihed above, except that the lysate concentration was 10' cells/ml. Antisrra and Immunopwcipitation-Antisera to purified chicken gizzard vinculin were prepared by methods similar to those previously descrihed for actin (10). The antisera were purified by affinity chromatography using a column ofvinculin (1.5 mg) coupled to cyanogen bromide-act ivated Sepharose 4R (Pharmacia) (1 1 ).
For immunoprecipitation, 200 pl of cell lysate were precipitated with excess antibody preadsorhed to 50 p1 of Staphylococcus aurrus ( 1 0 5 suspension) in 1 ml of modified RIPA as described (12). After incubation for 30 min on ice, the immunoprecipitates were collected hy centrifugation at 3000 X g for 10 min and washed successively with 1 ml o f modified RIPA, 1 ml of modified RIPA with 2.5 M KCI, and 1 ml of modified RIPA. The complexes were eluted by boiling in SIX gel sample buffer for 1 min and analyzed hy SDS-polyacrylamide gel electrophoresis in 10% or 5-105; gradient gels according to the procedure of Laemmli (13). Gels containing [:%]methionine-labeled proteins were fluorographed by the procedure of Ronner and Laskey (14).
Coomassie hlue-stained gels were dried on cellophane memhrane hacking (Rio-Rad). The amount of vinculin was quantitated by densitometric scanning of the stained vinculin hand. The radioactivity was quantified hy excising the vinculin hand and determined by scintillation counting in 10 ml of Aquasol (New England Nuclear).
Tryptic Prptidc. Ana/.vsk-Vinculin hands from SDS-polyacrylamide gels of the immunoprecipitates were excised, and tryptic digestion was performed as previously described (15). Separation of phosphorylated peptides was hy electrophoresis in 30% formic acid in the first dimension on cellulose thin layer plates (Schleicher & Schull) followed by chromatography in the second dimension in hutanokacetic acid:pyridine:water (60:12:4048 v/v).
Cellulose chromatography plates were exposed to Kodak X-Omat film with a Chronex intensifying screen at -70 "C. The individual peptides were marked hy comparison with the autoradiograms and aspirated from the plate, and radioactivity was determined by scintillation counting.
I'hosphoamino Acid Ana/,vsi.s-For phosphoamino acid analysis, 2 ml of cell lysates for each condition was immunoprecipitated with excess antibody. The '"P-labeled vinculin bands were excised from the gels and electroeluted for 12 h at 1 W in 0.025 M Tris, 0.192 M glycine, and 0.1 ?E SDS with an ISCO model 1750 sample concentrator. The concentrated sample (1 ml) of eluted vinculin was dialyzed overnight against 1 liter of 10 mM ammonium bicarbonate and lyophilized. The lyophilized material was washed with 200 pl of ethanokether (1:l v/v) at 0 "C and centrifuged at 3000 X g for 10 min.
The pellet was dissolved in 0.5 ml of 6 N HCI by heating at 100 "C for 1 min and hydrolyzed under N, for 2 h at 110 "C. The hydrolyzed material was lyophilized and redissolved in 0.5 ml of distilled water followed again by lyophilization to remove the acid. Amino acid separation was performed at pH 3.5 as previously descrihed (16). The individual amino acids were identified by comparison with internal standards detected by ninhydrin staining. Autoradiograms were ohtained as described above, and radioactivity was determined by scintillation counting.
Materials-Vinculin was purified from chicken gizzards as previously descrihed (17). Formalin-fixed S. aureus was prepared by the method of Kessler (18) and boiled in Tris-buffered saline containing 30% SDS and 10% 8-mercaptoethanol prior to use. Phorhol esters, dibutyryl CAMP, and dibutyryl cGMP were obtained from Sigma.

RESULTS
The present study utilized biosynthetic labeling of proteins in intact cells with :' ' Pi to ascertain whether factors reported to alter protein kinase C activity affect vinculin phosphorylation. Because of the many phosphoproteins present in the cell lysate, no clear increases in discrete proteins could be identified by one-dimensional separation of crude cell lysates by SDS-polyacrylamide gel electrophoresis. Therefore, immunoprecipitation of the cell lysates with antivinculin antibody was employed to examine vinculin phosphorylation. The antibody employed specifically precipitated vinculin (Fig. 1) as seen with both the Coomassie blue-stained gel (lanes 1-4) M, While immunoprecipitation was performed with lysates containing equal numbers of cells, the readily visible Coomassie blue-stained vinculin band enabled us to perform densitometric analysis to exclude small differences in protein due to processing and extraction.
Vinculin Phosphorylation in Response to Phorbol Esters-Phorbol esters, particularly PMA, have been reported to activate protein kinase C ( 2 , 3) and stimulate vinculin phosphorylation by purified protein kinase C (8). Therefore, we examined whether treatment of intact chick embryo fibroblasts with PMA would activate protein kinase C and result in increased phosphorylation of vinculin. Fig. 2 demonstrates that exposure of the chick embryo fibroblasts to 50 ng/ml(80 nM) of PMA results in significantly increased phosphorylation of vinculin. B shows the autoradiogram of immunoprecipitates of ["'S]methionine-labeled cells treated in the same manner. No differences in the amount of vinculin precipitated by the antibody were found during the labeling and treatment period. Thus, the increased phosphorylation seen in response to phorbol esters represents increased protein phosphorylation.
Vinculin phosphorylation in response to PMA was rapid (Fig. 3). Significant increases in vinculin phosphorylation were noted in minutes and began to plateau around 1 h of PMA treatment. Again no changes were noted in the amount of vinculin labeled by [:"S]methionine during the short time periods of phorbol treatment (data not shown). Fig. 4  Time course of the effect of PMA on vinculin phosphorylation. Chick emhryo fibroblasts were '"P-labeled in 0.6 mM calcium and exposed to 80 nM PMA for the time periods specified. All methods were as described under "Experimental Procedures." Hesults are expressed as per cent increase over control. very low concentrations (8 nM) resulted in small, but reproducible, increases in vinculin phosphorylation.
To test the specificity of the effect of phorbol esters on vinculin phosphorylation, several other agents were examined (Table I). Dimethyl sulfoxide, the solvent used for these compounds, had no effect on vinculin phosphorylation a t twice the final concentration used in the studies. Phorbol 12,13-dibutyrate, a less potent compound for protein kinase C activation, produced a small increase in vinculin phospho- The vinculin band in the Coomassie blue-stained gel for Swiss 3T3 cells was at the lower limit of detection by densitometry and precluded accurate protein determination. Therefore, results are expressed as per cent increase over control.

TABLE I
Effect of phorbol esters, dibutyryl cAMP and dibutyryl cGMP on vinculin phosphorylation in chick embryo fibroblasts Cells were labeled with [R2P]orthophosphate as described under "Experimental Procedures." Stock solutions (7.5 pl) of the compounds dissolved in dimethyl sulfoxide or dimethyl sulfoxide (20 pl) alone were added to achieve a final concentration as indicated in individual dishes. Control dishes were treated with distilled water. After 60 min incubation a t 37 "C, medium was removed, cell lysates were prepared, and immunoprecipitation was performed as described under "Experimental Procedures." rylation, but not of the same magnitude as PMA at equimolar concentrations. Phorbol, an inactive compound, had no effect on vinculin phosphorylation. Two cyclic nucleotides, dibutyryl cAMP and dibutyryl cGMP, were also tested. Both had negligible effects on the phosphorylation of vinculin in the intact cells.
Vinculin Phosphorylation in Response to Calcium-Protein kinase C has an absolute requirement for Ca'+ and is reversibly activated by increasing Ca'+ concentrations. Fig. 5 shows the effect on vinculin phosphorylation of lowering the calcium concentration from 1.6 mM (lane 3 ) to 0.6 mM (lane 2) and to 0.2 mM (lane I). Lowering the calcium concentrations produced decreases in vinculin phosphorylation. At both 0.2 mM Ca2+ and 0.6 mM Ca2+, PMA treatment resulted in increased vinculin phosphorylation (Fig. 2). However at 1.6 mM Ca", the addition of up to 160 nM PMA produced no further increases in vinculin phosphorylation (data not shown). Thus, the effect of PMA on vinculin phosphorylation was only seen a t low calcium concentrations. Tryptic Peptide Analysis-Tryptic peptide analysis was done to determine whether phorbol ester treatment resulted in new sites of phorphorylation. Control cells revealed a complex pattern of phosphorylation as shown in Fig. 6A. Multiple phosphate-containing peptides with different degrees of phosphorylation were found. Treatment with PMA did not affect phosphorylation of all peptides. PMA resulted in increased phosphorylation of the two peptides (arrows) in the upper middle portion of the electropherogram (Fig. 6 B ) . In addition, the appearance of a new phosphopeptide was detected in vinculin from PMA-treated cells (large arrow). Raising the calcium content of the medium resulted in changes in phosphorylation of the same three peptides (data not shown).
Phosphoamino Acid Analysis of Vinculin-Protein kinase C FIG. 6. Tryptic peptide analysis of 32P-labeled vinculin from control and PMA-treated chick embryo fibroblasts. All methods were as described under "Experimental Procedures." Cells were labeled with in 0.6 mM calcium and 80 nM PMA was added during the last 60 min of the labeling period. Autoradiograms were exposed for 24 h at -70 "C. A, '*P-labeled peptides from control cells: R , :%'-labeled peptides from PMA-treated cells. The lowest peptide in A was more diffuse and did not photograph. is a serine-and threonine-specific protein kinase (8,19). However, vinculin has also been found to contain phosphotyrosine in both normal and transformed cells. The amount of phosphotyrosine in vinculin was increased 8-fold in chick embryo fibroblasts transformed by Rous sarcoma virus (6). Since phorbol esters belong to a group of compounds known as tumor promoters, we examined whether increases in phosphotyrosine might occur in vinculin from cells treated with PMA. Fig. 7 is an autoradiogram of the phosphoamino acid analysis of vinculin from control and PMA-treated chick embryo fibroblasts. Vinculin from control cells was found to contain all three phosphoamino acids. However, the amount of phosphotyrosine was very low. PMA-treated cells had increased amounts of both phosphoserine and phosphothreonine. The level of phosphotyrosine remained barely above background, and no significant increases were detected.
Morphology-Phorbol esters produce morphological changes in many cells which mimic those seen with transformation. Consequently, the effect of phorbol esters on the morphology of both these fibroblasts was compared. In Swiss 3T3 cells small concentrations of PMA produced increased phosphorylation of vinculin (Fig. 4) and very dramatic morphological changes in the same time period in which changes in vinculin phosphorylation were noted (Fig. 8, A and B ) .
However, as can be seen in Fig. 8, C and D, even the highest concentration of PMA tested produced no obvious morphological changes in chick embryo fibroblasts, even though comparable increases in vinculin phosphorylation were seen in these cells. Therefore, no simple correlation existed between increased phosphorylation of vinculin and the morphological change produced by PMA in these cells.

DISCUSSION
The present study demonstrates that increased phosphorylation of vinculin occurs in intact cells in response to both calcium and several biologically active phorbol esters . Neither cyclic nucleotide (CAMP and cGMP) nor phorbol, a biologically inactive compound, significantly increased vinculin phosphorylation (Table I). These findings parallel results of experiments using purified components (8) in which both purified CAMP-dependent protein kinase catalytic subunit and cGMP-dependent protein kinase phosphorylated vinculin at a 100-fold lower rate than protein kinase C.
PMA has been shown to activate protein kinase C directly by decreasing the calcium concentration necessary for enzyme activation (2, 3). Therefore, at high calcium concentration M ) in the presence of phospholipid and/or diolein, no further effects of PMA on protein kinase C activity were found (2, 3) suggestion that PMA may also increase Ca2+ influx (2, 20), although in mouse thymocytes 10 nM PMA has been shown to decrease intracellular calcium (21). In both cell types, PMA resulted in increased vinculin phosphorylation in low calcium media but had no further effect on vinculin phosphorylation at high calcium concentrations. The present results do not. allow us to distinguish the exact mechanism of the phorbolinduced increases in vinculin phosphorylation. Either mechanism, increased Ca2+ influx resulting in protein kinase C activation or direct interaction with protein kinase C to lower the K,, for Ca" activation, could have produced the increases in vinculin phosphorylation.
T h e finding that vinculin serves as a substrate for purified src kinase ( 2 2 ) and contains significantly increased levels of phosphotyrosine in Rous sarcoma virus-transformed cells (6) has led to the hypothesis that vinculin may be involved in the altered cytoskeletal structure of transformed cells. However, the role of vinculin in these morphological changes remains to be clarified. In view of previous reports that phorbol esters induce morphological changes which mimic those seen in transformed cells (4,5), it was of obvious interest to compare the effects of phorbol esters on morphology with effects on vinculin phosphorylation. No morphological changes were observed in chick embryo fibroblasts at the time when significant increases in vinculin phosphorylation were observed. A previous study of the effect of PMA on actin-containing structures in chick embryo fibroblasts did show a loss of ordered actin structures resembling those seen in virally transformed cells (6). However, these changes were observed a t 16 h of PMA exposure and required both RNA and protein synthesis. These discrepancies suggest that the effects of phorbol esters on morphology are complex, and vinculin phosphorylation alone was not sufficient to induce morphologic changes.
Two recent reports demonstrate that in chick embryo fibroblasts phorbol 12-myristate 13-acetate increases tyrosine phosphorylation on a protein or proteins ( M , = 40,000-43,000) (23,24). The increased phosphorylation of vinculin seen in the present study occurred on serine and threonine residues (Fig. 7). Protein kinase C has been shown to phosphorylate both of these residues, but not tyrosine in vinculin and other substrates. In agreement with a previous report (6), vinculin from normal cells did contain low levels of phosphotyrosine. No distinct increases in phosphotyrosine were found in vinculin from phorbol-treated cells. Because of the low level of phosphotyrosine, smail changes might have gone undetected. However, comparable techniques were used in a study of Rous sarcoma virus-transformed chick embryo fibroblasts and revealed significant increases in phosphotyrosine in vinculin.
In addition, small increases in phosphotyrosine were detected in several other proteins (6).
The tryptic peptide analysis reveals a complex pattern of phosphorylation (Fig. 6). T h e findings indicate that multiple protein kinases are involved in the phosphorylation of vinculin. Two major and several minor phosphorylated peptides were found in purified vinculin phosphorylated by protein kinase C (8). One of the major peptides and one of the minor peptides phosphorylated in purified vinculin correspond with the peptides whose phosphorylation was increased by PMA and calcium in intact cells. However, the other major phosphate-containing peptide from purified vinculin did not correspond to any of the phosphorylated peptides seen in the present study. These findings are not surprising in view of the mult.iplicity of sites which can be phosphorylated. Discrepancies between phosphorylation sites in intact tissues or cells when compared to purified proteins have been reported for other proteins ( 2 5 ) and vinculin (6, 8). We have recently shown that vinculin may undergo a conformational change by binding to anionic phospholipids (26). Thus, data obtained with purified vinculin may not correspond exactly to results obtained in intact cells. These differences and the reports of increased tyrosine phosphorylation following PMA treatment (23, 24) may indicate that phorbol esters affect other protein kinases or phosphatases. Therefore, it is not possible to establish the exact identity of the kinase responsible for the increased phosphorylation of vinculin seen following treatment with calcium or phorbol esters.
T h e significance of multisite phosphorylation is not very well understood. Myosin light chain kinase has been shown to be a substrate for CAMP-dependent protein kinase with phosphorylation occurring on two separate peptides. Phosphorylation of one site decreased the affinity of the enzyme for its activator, calmodulin. Phosphorylation of the other site did not affect binding of calmodulin (27). Differential phosphorylation and its probable regulation are necessary effects on the activity of glycogen synthase following site-prerequisites to the understanding of the role of vinculin in specific phosphorylation has also been reported (28). T h e transmembrane interactions. The present findings that indifferential phosphorylation ofvinculin on specific sites (Fig. creases in vinculin phosphorylation occur in intact cells in 6) suggests that phosphorylation of different sites regulates response to factors which activate protein kinase C, coupled different functions.
with data obtained with purified proteins (8), suggest that vincu1in has been shown to be able to bind F actin (29,30) protein kinase c, the calcium, phospholipid-activated protein and is proposed to link actin microfilament bundles to the kinase may directly Or indirectly regulate vinculin Phosphomembrane. An actin-independent interaction of vinculin with rylation. the cell membrane has also been demonstrated (31). Thus, VinCUlin may be involved in Several types Of molecular inter-Achnon,led~menls-Wewould like to thank Elizabeth Lovelace and actions. Identification ofthe enzymes responsible for vinculin Annie Harris for maintaining the cell cultures. Dr. Mark C. Wil-