Protein phosphatase 2A interacts with the Src kinase ... - Nature

Oncogene (2001) 20, 6057 ± 6065 ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc

Protein phosphatase 2A interacts with the Src kinase substrate p130CAS Noriko Yokoyama1 and W Todd Miller*,1 1

Department of Physiology and Biophysics, School of Medicine, State University of New York at Stony Brook, Stony Brook, New York, NY 11794-8661, USA

In this study, we report that the Src substrate Cas (p130 Crk-associated substrate) associates with protein phosphatase 2A (PP2A), a serine/threonine phosphatase. We investigated this interaction in cells expressing a temperature-sensitive mutant form of v-Src. v-Src activation (by shifting cells from the nonpermissive to the permissive temperature) led to an increase in the tyrosine phosphorylation of v-Src and Cas, as well as in the association between v-Src and Cas. v-Src has previously been shown to bind to PP2A and to phosphorylate the catalytic subunit of PP2A, resulting in inhibition of phosphatase activity. We found that the association between v-Src and PP2A decreased as cells were shifted to the permissive temperature. In contrast, the levels of PP2A that co-immunoprecipitated with Cas increased when v-Src was activated. We obtained similar results in pull-down experiments with immobilized Microcystin, a PP2A inhibitor. Serine/threonine phosphorylation of Cas has previously been shown to occur in a cell cycle regulated matter. Treatment of NIH3T3 cells with okadaic acid, a PP2A inhibitor, augments the serine/threonine phosphorylation of Cas that occurs at mitosis. Furthermore, PP2A dephosphorylates serine residues on Cas in vitro. Taken together, our results suggest that PP2A may be involved in the cell cyclespeci®c dephosphorylation of Cas. Oncogene (2001) 20, 6057 ± 6065. Keywords: Src; PP2A; Cas; serine dephosphorylation Introduction Src family nonreceptor tyrosine kinases play essential roles in a variety of cellular functions, including proliferation, survival, dierentiation and apoptosis (Brown and Cooper, 1996). Src kinases are composed of an N-terminal membrane binding region, a unique domain, an SH3 domain that binds proline-rich sequences, a phosphotyrosine binding SH2 domain, a tyrosine kinase catalytic domain, and a C-terminal regulatory tail.

*Correspondence: WT Miller; E-mail: [email protected] Received 22 February 2001; revised 5 June 2001; accepted 14 June 2001

Cas (p130 Crk-associated substrate) was originally identi®ed as a major tyrosine-phosphorylated protein of 130 kDa in cells transformed by v-Src or v-Crk (Reynolds et al., 1989; Sakai et al., 1994a). Cas becomes tyrosine phosphorylated following various physiological stimuli, such as cell adhesion, cytokine receptor engagement, and growth factor stimulation (O'Neill et al., 2000; Schlaepfer and Hunter, 1998). Evidence that Src is directly responsible for phosphorylating Cas includes: (i) tyrosine phosphorylation of Cas correlates well with transformation of NIH3T3 cells by Src (Sakai et al., 1994a); (ii) Cas is not phosphorylated in ®broblasts derived from Src7 `knockout' mice (Schlaepfer et al., 1997; Vuori et al., 1996); and (iii) Cas is an in vitro substrate of Src (Sakai et al., 1994b). Cas contains an SH3 domain as well as multiple potential Src phosphorylation sites and a Cterminal domain that contains Src SH3- and SH2binding sequences (Sakai et al., 1994a; O'Neill et al., 2000). This C-terminal region is required for extensive tyrosine phosphorylation of Cas in cell expressing activated c-Src. After the substrate region of Cas is hyperphosphorylated, Cas forms a signaling complex with the adaptor proteins Crk and Nck. The formation of the Cas-Crk complex has been shown to be important in the induction of cell migration mediated by Cas (Klemke et al., 1998), and similar multiprotein complexes are likely to be involved in transducing other downstream signals. There is evidence that serine phosphorylation of Cas also plays a role in modulating its biological function. Cas possesses a serine-rich domain adjacent to the region containing potential Src phosphorylation sites. 14-3-3 proteins have been shown to interact with Cas via phosphoserine-dependent interactions (Garcia-Guzman et al., 1999). Furthermore, Cas, focal adhesion kinase (FAK), and paxillin are prominently serine/ threonine phosphorylated (and tyrosine dephosphorylated) at mitosis (Yamakita et al., 1999). Serine/ threonine phosphorylation of Cas occurs concomitantly with a dissociation of the FAK/Cas/Src complex. The mitosis-speci®c serine/threonine phosphorylation of Cas continues past cytokinesis and is reversed after mitosis is complete (Yamakita et al., 1999). These observations suggest that the serine/ threonine phosphorylation of Cas, paxillin, and FAK during mitosis may contribute to the arrest of integrinmediated signaling in mitosis. The kinase(s) and phosphatase(s) responsible for serine/threonine phos-

Cas-PP2A interaction N Yokoyama and WT Miller

6058

phorylation of Cas and the other focal adhesion components have not been identi®ed. Protein phosphatase 2A (PP2A), a serine/threonine phosphatase, is widely distributed in the cytoplasm of mammalian cells. The PP2A holoenzyme is a heterotrimer composed of catalytic (C), structural (A) and regulatory (B) subunits (Shenolikar and Nairn, 1991; Goldberg, 1999; Millward et al., 1999). In cells, the ABC heterotrimer is considered the predominant form of PP2A, and multiple regulatory subunits provide dierent substrate speci®cities and intracellular localization. PP2A has been shown to be required at various stages of the eucaryotic cell cycle. In the yeast S. cerevisiae, PP2A is required for the progression into mitosis (Lin and Arndt, 1995; Yanagida et al., 1992). A recent mutational analysis of the PP2A C subunit indicated that the proper regulation of PP2A is required for proper exit from mitosis (Evans and Hemmings, 2000). Okadaic acid (OA), an inhibitor of PP1 and PP2A, causes mitotic cell arrest in human leukemia K562 cells (Zheng et al., 1991) and HeLa cells (Chaudhuri et al., 1997). In this paper, we examine the possibility that PP2A may participate in signaling complexes containing Src and Cas. We have used a cell line expressing a temperature-sensitive mutant form of v-Src, in which Src activity is regulated by a shift from the nonpermissive temperature (398C) to the permissive temperature (348C). This has allowed us to manipulate the activity of v-Src in order to investigate the eects on complex formation with PP2A and Cas. We report here that PP2A forms speci®c complexes with v-Src and Cas in these cells, and that modulating the activity of Src has dierential eects on the individual protein ± protein complexes. In NIH3T3 cells that are not transformed by v-Src, we show that interfering with PP2A leads to increased serine/threonine phosphorylation of Cas at mitosis. Our studies suggest that Cas may be a cell cycle speci®c substrate of PP2A.

Results Activation of v-Src in Ts-72 cells Ts-72 cells were maintained at the Src-nonpermissive temperature of 398C. To activate v-Src, cells at 70% con¯uency were transferred to the permissive temperature (348C) for 2 days. Tyrosine phosphorylation of proteins in cell lysates was then examined at both temperatures. As shown in Figure 1, overall tyrosine phosphorylation levels were dramatically increased at 348C. We next investigated the tyrosine phosphorylation of two proteins in the lysates: (i) v-Src itself and (ii) Cas, a known v-Src substrate. Cell extracts were subjected to immunoprecipitation with Src and Cas antibodies, and proteins in the immunoprecipitates were separated by SDS ± PAGE and analysed by Western blot analysis using anti-phosphotyrosine antibody. As expected, these experiments clearly showed that the tyrosine phosphorylation of v-Src and Cas was Oncogene

Figure 1 Phosphorylation of Src and Cas in Ts-72 cells. Ts-72 cells were maintained at the Src-nonpermissive temperature of 398C or transferred to the permissive temperature (348C) for 48 h before harvest. Whole cell lysates (WCL; 50 mg total protein) were analysed by SDS ± PAGE and Western blotting using antiphosphotyrosine antibody 4G10. Cell lysates (1 mg proteins) were immunoprecipitated by either Src or Cas antibody and probed with phosphotyrosine antibody. Results from cells grown at the nonpermissive and permissive temperature are denoted 39 and 34, respectively. The results are representative of four independent experiments

increased after the temperature was shifted from 39 to 348C (Figure 1). Additional tyrosine-phosphorylated proteins were present in the immunocomplexes with vSrc and Cas, including a prominent band at &180 ± 200 kD that we have not identi®ed (Figure 1). These data con®rmed that Src is activated at the permissive temperature in Ts-72 cells, and that activation of Src leads to tyrosine phosphorylation of Cas in these cells. Association of PP2A with v-Src and Cas In the next series of experiments, we investigated the association of PP2A with v-Src and Cas in Ts-72 cells at the nonpermissive and permissive temperatures. Cell extracts were immunoprecipitated with PP2A antibodies and then probed with either Src or PP2A antibodies (Figure 2a). A complex between v-Src and PP2A was detected at the nonpermissive temperature, consistent with previous reports of Src-PP2A binding (Pallas et al., 1990; Glenn and Eckhart, 1993; Campbell et al., 1995; Glover et al., 1999; Ogris et al., 1999) and v-Src phosphorylation of PP2A (Chen et al., 1992). Activation of v-Src by shifting the temperature to 348C resulted in a decrease in the amount of v-Src that coimmunoprecipitated with PP2A (Figure 2a). The same result was obtained in a reciprocal experiment in which cell lysates were immunoprecipitated with Src antibodies and probed by Western blotting with anti-PP2A (Figure 2b). We also detected the PP2A regulatory subunit Aa in the anti-PP2A immunoprecipitation reaction (Figure 2a). Because the association between v-Src and Cas increases when Ts-72 cells are shifted to 348C (Figures 1 and 2b), we tested for the formation of a complex between PP2A and Cas. A small amount of Cas was present in PP2A immunoprecipitates prepared from Ts72 cells at the nonpermissive temperature, and this


6059

Figure 2 PP2A interacts with Src and Cas in Ts-72 cells. (a) Ts-72 cell lysates were prepared from cells grown at 34 or 398C. Lysates (1 mg protein) were immunoprecipitated with monoclonal antibody against the C subunit of PP2A antibody. Control precipitation reactions (IgG) were carried out with mouse IgG (control for PP2A and Cas Ab) or rabbit IgG (control for Src Ab). Immunoprecipitated proteins were subjected to SDS ± PAGE and analysed by Western blotting against Src, PP2A, Cas, and the PP2A Aa regulatory subunit. HC=antibody heavy chain, LC=antibody light chain. (b) Lysates were immunoprecipitated using polyclonal anti-Src antibody and analysed by Western blotting with Src, PP2A, and Cas antibodies. (c) Lysates were immunoprecipitated with Cas antibody and Western blotting was performed with PP2A or Cas antibody. The results are representative of six independent experiments

amount increased dramatically when v-Src was activated by shifting to 348C (Figure 2a). To con®rm this ®nding, reciprocal immunoprecipitations were carried out with Cas antibody and the reactions were analysed using PP2A antibody (Figure 2c). In parallel experiments, we were unable to detect any interaction between Cas and PP2B, another serine/threonine phosphatase (data not shown). Thus, these experiments demonstrated a speci®c interaction between PP2A and Cas that was dependent on the activation state of vSrc. We next tested whether the catalytic activity of vSrc was important in these associations. We treated Ts72 cells with PP2, a Src speci®c inhibitor, and then tested for the formation of PP2A-Cas and PP2A-Src complexes (Figure 3). The amount of Cas present in PP2A immunoprecipitates decreased with treatment of PP2, suggesting that v-Src kinase activity is involved in the interaction between Cas and PP2A (Figure 3a). The association between v-Src and PP2A was increased by

treatment with PP2 (Figure 3b), consistent with our previous observation of a decrease in the Src-PP2A interaction upon activation of v-Src (Figure 2a,b). We con®rmed the results of the co-immunoprecipitation experiments using Microcystin anity resin. Microcystin is a potent inhibitor of PP2A and PP1 that binds to the phosphatase catalytic subunits (Goldberg et al., 1995). Microcystin-Sepharose was used as an anity resin to isolate complexes containing PP2A from Ts-72 cells grown at 34 and 398C. As shown in Figure 4a, when cells were grown at 398C, Src and Cas were present in Microcystin-Sepharose complexes containing PP2A. Activation of v-Src by shifting Ts-72 cells to the permissive temperature resulted in decreased association between PP2A and Src, and increased association between PP2A and Cas. These results coincide with those obtained in the coimmunoprecipitation experiments (Figure 2). Activation of v-Src results in pronounced changes in complex Oncogene


6060

Figure 3 Eects of the Src inhibitor PP2 on Src-PP2A and CasPP2A association. Ts-72 cells grown at the permissive temperature (348C) were treated with 100 nM PP2 for 1 h before harvesting. Cell lysates (1 mg protein) were subjected to immunoprecipitation with PP2A antibody and analysed by Western blotting using Cas antibody (a) or Src antibody (b). In each case, blots were reprobed with PP2A antibody. Control precipitation reactions were carried out with mouse IgG. Results from cells grown at the nonpermissive and permissive temperatures are denoted 39 and 34, respectively. The results are representative of ®ve independent experiments

Figure 4 Microcystin-Sepharose pull-down experiments. Ts-72 cell lysates were prepared from cells grown at 34 or 398C. Lysates (1 mg protein) were incubated with Microcystin-Sepharose (15 ml) for 4 h. Bound proteins were eluted by SDS-sample buer and analysed by Western blotting with Src, Cas and PP2A antibodies (a) or phosphotyrosine and PP2A antibodies (b). Results from cells grown at the nonpermissive and permissive temperatures are denoted 39 and 34, respectively. The results are representative of six independent experiments Oncogene

formation with PP2A and Cas. Using MicrocystinSepharose pulldown experiments, we also analysed the tyrosine phosphorylation state of PP2A in Ts-72 cells grown at 34 and 398C. Activation of v-Src led to an increase in the tyrosine phosphorylation of PP2A (Figure 4b), consistent with previous studies (Chen et al., 1992). To determine the importance of PP2A activity on the formation of PP2A-Src and PP2A-Cas complexes, Ts72 cells were treated with 40 nM okadaic acid (OA), a PP1 and PP2A inhibitor (Sheppeck et al., 1997; Favre et al., 1997). PP2A was isolated by immunoprecipitation, and complex formation was detected by Western blotting with anti-Src antibody (Figure 5a). Treatment with OA decreased the association of v-Src with PP2A (Figure 5a), implying that PP2A activity plays a role in complex formation. Treatment with OA also decreased the interaction of Cas and v-Src (Figure 5b). Because OA could potentially have eects in Ts-72 cells that are unrelated to PP2A inhibition, we tested the importance of PP2A activity on PP2A/Src/Cas complex formation using a complementary strategy. We transfected Cos-7 cells with HA epitope-tagged PP2A or a catalytically inactive H118N PP2A mutant. Histidine 118 of PP2A is a critical residue for catalysis, and it is conserved in a variety of enzymes involved in the hydrolysis of phosphate esters (Ogris et al., 1999).

Figure 5 Inhibition of the Src-PP2A interaction by okadaic acid. (a) Ts-72 cell lysates (1 mg protein) from cells grown at the nonpermissive and permissive temperatures were immunoprecipitated with PP2A antibody and probed with Src and PP2A antibodies. (b) Cell lysates (1 mg protein) were immunoprecipitated with Cas antibody or Src antibody and probed with Src antibody or Cas antibody, respectively. For experiments with OA, Ts-72 cells at the permissive temperature (348C) were treated with 40 nM OA for 1 h before harvesting


We then isolated PP2A by anti-HA immunoprecipitation reactions, and mixed the immobilized PP2A with Ts-72 cell lysates (from cells grown at 348C) to measure binding of v-Src and Cas to PP2A. As shown in Figure 6, the pull-down reactions contained approximately equal amounts of PP2A (wild-type or mutant). v-Src associated with the wild-type form of PP2A and showed less interaction with the inactive PP2A mutant (Figure 6). In contrast, Cas bound better to the inactive form of PP2A (Figure 6). When similar experiments were carried out using lysates from cells grown at 398C, there were no dierences in binding of wild-type or mutant PP2A to Cas or Src (results not shown). The results presented in Figures 5 and 6 suggest that PP2A activity is important for maximum binding between v-Src and PP2A at the Src-permissive temperature. This partial requirement for PP2A activity may explain why a stronger PP2A-Src interaction was observed at 348C by PP2A immunoprecipitation (Figure 2) than by Microcystin pulldown (Figure 4). We also investigated whether PP2A directly in¯uences v-Src activity in vitro. We incubated puri®ed PP2A with puri®ed v-Src (either full-length v-Src or a truncated version consisting of the SH3, SH2 and catalytic domains). v-Src kinase activity (as measured toward a synthetic peptide substrate) was inhibited in a concentration- and time-dependent manner by PP2A. After PP2A treatment, the activity of the truncated form of v-Src was 35.2+5.2% of a control sample (data not shown). Inhibition of PP2A was reversed in the presence of OA. Roles of PP2A in Cas signaling We next investigated whether PP2A plays a role in Cas signaling in cells that are not overexpressing Src. It has previously been demonstrated that Cas is serine/ threonine phosphorylated at mitosis (Yamakita et al., 1999). This phenomenon was observed in REF-2A and NIH3T3 cells which were treated with nocodazole, an

Figure 6 Pull-down experiments with wild-type and inactive PP2A C subunits. Immobilized PP2A C subunits (wild-type or H118N) were incubated with cell lysates (1 mg protein) from Ts72 cells grown at the nonpermissive and permissive temperatures. Bound proteins were analysed by immunoblotting with PP2A, Src or Cas antibodies. WT and Mut indicate wild-type and H118N PP2A C subunits, respectively. The results are representative of six independent experiments

anti-microtubule reagent. Mitosis-speci®c serine/threonine phosphorylation of Cas continues past cytokinesis and is reversed after mitosis is complete (Yamakita et al., 1999). We carried out similar experiments using NIH3T3 cells treated with nocodazole or OA. Consistent with the results reported by Yamakita et al. (1999) serine phosphorylation of Cas was increased after nocodazole treatment. This was evident by enhanced reactivity toward anti-phosphoserine antibody (Figure 7a) or by a shift to a higher apparent molecular weight on SDS ± PAGE (B Craddock, N Yokoyama, and WT Miller, unpublished observations). The levels of phosphoserine were also elevated upon treatment of the cells with 80 nM OA (Figure 7a). The results with OA are made more striking by the observation that OA or nocodazole treatment resulted in lower recovery of Cas in the anti-Cas immunoprecipitation reactions (Figure 7a). Serine phosphorylation of Cas was dramatically increased when cells were treated with a combination of nocodazole and okadaic acid (Figure 7a), although this is partially explained by a reproducible increase in Cas recovery under these conditions. The enhanced serine phosphorylation of Cas observed in the presence of OA suggests that PP2A may be involved in serine-dephosphorylation of Cas. We investigated the possibility that Cas is a substrate of PP2A. Cas immunoprecipitates from cells treated with nocodazole alone, or nocodazole plus OA, were treated with puri®ed PP2A and then analysed by antiphosphoserine Western blotting. As shown in Figure 7a, PP2A treatment markedly reduced the levels of phosphoserine in Cas. Tyrosine phosphorylation of Cas was detected in cells that were treated with a combination of nocodazole and OA. However, tyrosine phosphorylation of Cas did not change after treatment with puri®ed PP2A, although serine phosphorylation of Cas diminished dramatically. Thus, under these in vitro conditions, PP2A is able to dephosphorylate Cas. We also detected an increase in the association between PP2A and Cas when cells were treated with nocodazole (Figure 7b). 14-3-3z has been shown to associate with Cas in a phosphoserine-dependent manner (Garcia-Guzman et al., 1999). We tested for an interaction between Cas and 14-3-3z in mitotic NIH3T3 cells. 14-3-3z immunoprecipitates from cells treated with nocodazole alone (Figure 7c) or nocodazole plus OA (results not shown) contained elevated amounts of Cas in comparison with untreated NIH3T3 cells. These observations support the ®nding that mitotic cells show enhanced serine phosphorylation of Cas.

6061

Discussion To investigate the importance of Src activity on associations with Cas and PP2A, we took advantage of NIH3T3 cells that express a temperature-sensitive mutant (Ts-72) of v-Src (Mayer et al., 1986). Temperature-sensitive mutants of Rous sarcoma virus Oncogene


6062

Figure 7 Serine phosphorylation of Cas in NIH3T3 cells. (a). NIH3T3 cells were untreated (7) or treated with: 0.2 mg/ml nocodazole (Noc), 80 nM OA, or 0.2 mg/ml nocodazole plus 80 nM OA for 3 h. Cells were lysed and proteins (1 mg) were immunoprecipitated with polyclonal Cas antibody or rabbit IgG (C) as control. In some reactions, immunoprecipitated Cas was incubated with PP2A (0.125 mg) at 308C for 1 h. In those reactions, after incubation with PP2A, immunoprecipitated Cas was washed with 0.5% NP-40 in phosphate buered saline and eluted by SDS-sample buer. Proteins were analysed by SDS ± PAGE and analysed by Western blotting with phosphoserine, phosphotyrosine, and PP2A antibodies (ECL method) and Cas (colorimetric method). (b) Lysates (4 mg protein) from untreated or nocodazole-treated NIH3T3 cells were subjected to Cas immunoprecipitation. Bound proteins were analysed by immunoblotting with Cas and PP2A antibodies. (c) Lysates (1 mg protein) from untreated or nocodazole-treated NIH3T3 cells were immunoprecipitated using anti-14-3-3 antibody. Bound proteins were analysed by immunoblotting with Cas antibody. The results are representative of four independent experiments

(RSV) were used in early studies to demonstrate that the tyrosine kinase activity of v-Src is important for cell transformation (Sefton et al., 1980; Jove and Hanafusa, 1987). Lesions in v-Src which confer temperature-sensitivity have been mapped to the kinase catalytic domain (Mayer et al., 1986). The tsNY72-4 mutant form of v-Src expressed in these cell lines results in a tumorigenic phenotype only when cells are cultured at 348C, the permissive temperature (Mayer et al., 1986). At 398C, the restrictive temperature, the mutant form of Src is inactive. The two growth conditions provide an `internal control'; the cells are identical except for the presence of active vs inactive Src (Sefton et al., 1980; Jove et al., 1986). As expected, shifting Ts-72 cells to the permissive temperature resulted in heightened v-Src kinase activity, increased phosphorylation of Cas, and increased association between v-Src and Cas (Figures 1 and 2). Using this system, we were able to show for the ®rst time that Cas associates with the serine/ threonine phosphatase PP2A (Figure 2). Moreover, this association was strongly dependent on the activity of vSrc; we observed a large increase in the association Oncogene

between Cas and PP2A when Ts-72 cells were shifted to conditions where v-Src is active (Figure 2). Furthermore, treatment of Ts-72 cells with PP2, a Src-speci®c inhibitor, blocked the association between PP2A and Cas (Figure 3a). In contrast to results for the Cas-PP2A interaction, we showed that Src-PP2A association decreased when v-Src was activated by shift of temperature (Figure 2). Inhibition of v-Src by PP2 led to an increase in the association of v-Src and PP2A (Figure 3b). Previously, v-Src, insulin receptor, EGF receptor and Lck have been reported to phosphorylate the C-subunit of PP2A and inhibit the enzyme (Chen et al., 1992, 1994). This suggests a model in which v-Src activation in Ts-72 cells is accompanied by tyrosine phosphorylation of PP2A, followed by release of PP2A from the complex. In support of this model, we observed an increase in tyrosine-phosphorylated PP2A in lysates from cells grown at the permissive temperature, as compared with the non-permissive temperature (Figure 4b). The principal sites for serine/threonine phosphorylation of Src lie in the N-terminal unique region (Brown and Cooper, 1996). The presence of these phosphoryla-


tion sites suggests that the unique region might be the point of contact with PP2A. However, a v-Src construct lacking the unique region interacts with PP2A as strongly as full-length v-Src in vitro (N Yokoyama and WT Miller, unpublished observations), suggesting that the N-terminal membrane binding and unique domains of Src are dispensable for binding to PP2A. Because Cas can also bind to the SH3-SH2-catalytic domains of Src (O'Neill et al., 2000), it is possible that Cas and PP2A compete with each other for Src binding, and release of PP2A might partially explain the increase in Src-Cas association observed at the permissive temperature. Another possible explanation for the reciprocal eects of Src activity on the PP2A-Src and PP2A-Cas complexes is that phosphorylated Cas might have a high anity for PP2A. The available evidence suggests that Src is one of the principal tyrosine kinases involved in phosphorylation of Cas. Cas becomes processively tyrosine phosphorylated following various stimuli, such as cell adhesion, cytokine receptor engagement, and growth factor stimulation (O'Neill et al., 2000; Schlaepfer and Hunter, 1998). We investigated the importance of PP2A activity on complex formation with Src and with Cas. Depending on the experimental approach, previous studies have found varying degrees of importance of PP2A activity in complex formation with Src and polyoma virus middle T antigen (Glover et al., 1999; Ogris et al., 1999). We initially used okadaic acid, a PP2A inhibitor, to assess the importance of PP2A activity. At the concentrations used in our studies, PP2A has been reported to be selective for PP2A (Sheppeck et al., 1997). Treatment with OA resulted in a decrease in the formation of PP2A-Src complex (Figure 5), implying that phosphatase activity is important in association. Because OA could have eects on unidenti®ed cellular targets that could complicate the interpretation, we also carried out experiments using catalytically-inactive PP2A. In these experiments, wild-type PP2A showed a somewhat greater amount of complex formation with Src than did the inactive PP2A mutant (Figure 6). The ability of PP2A to associate with Cas was not ablated by the H118N mutation (Figure 6). The observation that Src and Cas retain some ability to associate with mutant PP2A suggests the possibility that the structural or regulatory subunits of PP2A might play a role in complex formation. The more dramatic decrease in Src and Cas binding seen in the presence of okadaic acid (Figure 5) could be due to a direct interaction between these molecules and the OA binding site (residues 265 ± 269) on the C-subunit of PP2A. However, nonspeci®c eects of OA cannot be ruled out in our experiments. Cas possesses a serine-rich domain adjacent to its tyrosine phosphorylation sites. Serine phosphorylation of Cas has not been well characterized with respect to sites of modi®cation, or the relevant kinase(s) and phosphatase(s). However, the importance of serine phosphorylation of Cas was highlighted in a recent report showing mitosis-speci®c serine phosphorylation of Cas. The increase in serine phosphorylation of Cas was accompanied by a decrease in tyrosine phosphor-

ylation and dissociation of the FAK/Cas/c-Src signaling complex. Because PP2A activity has previously been shown to be important at various stages in the eucaryotic cell cycle, and because of our observation of complexes between PP2A and Cas (Figure 2), we investigated the possibility that PP2A might be a Cas phosphatase. We carried out these studies in synchronized NIH3T3 cells using methods previously established to examine Cas serine phosphorylation (Yamakita et al., 1999). As reported by Yamakita et al. (1999) mitotic NIH3T3 cells (prepared by treatment with nocodazole) showed higher levels of Cas serine phosphorylation than interphase cells (Figure 7a). Treatment of interphase NIH3T3 cells with OA led to a similar increase in Cas serine phosphorylation (Figure 7a). Mitotic cells are rounded with disassembled focal adhesions and stress ®bers, and treatment of NIH3T3 cells with okadaic acid led to a similar morphology (data not shown). Treatment of cells with both OA and nocodazole led to an even greater level of Cas serine phosphorylation (Figure 7a). Using serine phosphorylated Cas obtained from mitotic NIH3T3 cells, we demonstrate that PP2A has the capacity to dephosphorylate Cas in vitro. PP2A has been reported to possess tyrosine phosphatase activity in the presence of a speci®c activator (Cayla et al., 1990). However, we do not observe any evidence for tyrosine dephosphorylation of Cas in our system (Figure 7a). The amount of PP2A that co-immunoprecipitated with Cas was higher in mitotic NIH3T3 cells than in interphase cells (Figure 7b). Taken together, our results suggest that PP2A may be involved in the serine dephosphorylation of Cas that occurs as cells exit mitosis. Additionally or alternatively, it is possible that the Cas-PP2A complex mediates other protein ± protein interactions in Cas signaling. Recently it has been shown that one of the signaling molecules downstream of Cas is the 14-3-3z protein (Garcia-Guzman et al., 1999). The interaction between Cas and 14-3-3z is stimulated by integrin-mediated cell adhesion and is dependent on serine phosphorylation of Cas. We observed an increased association between 14-3-3z and Cas in NIH3T3 cells treated with nocodazole (Figure 7c), consistent with previous ®ndings (Garcia-Guzman et al., 1999, Yamakita et al., 1999), and suggesting that PP2A may play a mitosis-speci®c role in regulating the interaction of these two signaling molecules. PP2A activity is regulated by association of dierent regulatory subunits with the AC core dimer and by dierential subcellular localization. In addition, PP2A activity is modulated by phosphorylation and carboxymethylation. These post-translational modi®cations may aect the subcellular localization and substrate speci®city of PP2A. Recent data suggest that the cell cycle-speci®c activity of PP2A may be due to dierential methylation of the C-subunit and dierential binding of the PR55/B subunit (Turowski et al., 1995; Evans and Hemmings, 2000). Our future studies will address the potential role of carboxymethylation in the PP2A-Cas interaction.

6063

Oncogene


6064

Here we demonstrated that PP2A forms a complex with Cas in cells expressing active Src. Inhibition of PP2A leads to an increase in serine phosphorylation of Cas, similar to the increase observed in mitotic cells. We also show that Cas is a direct in vitro substrate of PP2A. The data suggest a potential role for PP2A in cell cycle regulation through the focal adhesion protein Cas.

Materials and methods Materials PP2A protein (heterodimer of C subunit and A subunit), Microcystin-Sepharose, and antibodies against the PP2A C subunit, phosphotyrosine (4G10), and Src (clone GD 11) were obtained from Upstate Biotechnology, Inc. (Lake Placid, NY, USA). Antibodies against the PP2A regulatory subunit Aa, Src (N-16)-G, Cas(N-17), 14-3-3z and HA were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Monoclonal Cas antibody was from Transduction Laboratories (Lexington, KY, USA). Polyclonal phosphoserine antibody was from Zymed Laboratories Inc. (South San Francisco, CA, USA). Okadaic acid, mouse IgG, rabbit IgG and protein A-Sepharose were from Sigma. Ni-NTA resin was from Qiagen. 5-bromo-4-chloro-3-indolyl phosphate (BCIP) and nitroblue tetrazolium chloride (NBT) were from Roche Molecular Biochemicals (Indianapolis, IN, USA). PP2 was from Calbiochem (San Diego, CA, USA). Trans IT.LT1 was purchased from Mirus (Madison, WI, USA). The wild-type (HA)3-PP2A C subunit and inactive (HA)3-H118N-PP2A C subunit expression plasmids were a gift from Dr David Brautigan (University of Virginia, Charlottesville, Virginia USA). Cells and cell culture The Ts-72 cell line used in these studies (NIH3T3 ®broblasts transformed by a temperature-sensitive mutant of Src) was obtained from Dr Steven M Anderson (University of Colorado Health Science Center, Denver, Colorado, USA). The tsNY72-4 mutant form of v-Src expressed in these cell lines results in a tumorigenic phenotype only when cells are cultured at the permissive temperature (Mayer et al., 1986). Cells were cultured in Dulbecco's modi®ed Eagle's medium supplemented with 10% of fetal bovine serum and antibiotics at 398C. For activation of Src, cells were cultured at 348C for 2 days. For okadaic acid or PP2 treatment, cells were treated with 40 nM OA or 100 nM PP2 for 1 h before harvesting. Immunoprecipitation and Western blotting Cells were harvested, washed twice with ice-cold phosphatebuered saline, and lysed in lysis buer I [10 mM Tris-HCl (pH 7.5), 5 mM EDTA, 50 mM NaF, 2 mM Na3VO4, 10 mM DTT, 1% Nonidet P-40 (NP-40), 0.05% SDS, 0.25% sodium deoxycholate, 1 mM PMSF, 10 mg/ml leupeptin, 10 mg/ml aprotinin] at 48C for 30 min. Cell lysates were centrifuged at 15 000 g for 10 min. The resulting supernatants were collected and protein concentrations were determined using the Bradford method (Bio-Rad). The supernatants were precleared with protein A-Sepharose for 1 h at 48C and centrifuged at 15 000 g for 10 min. The resulting superOncogene

natants (1 mg cellular proteins) were used for immunoprecipitation analysis with polyclonal or monoclonal antibodies. After addition of 10 ml of protein A-Sepharose, incubations were continued 5 h or overnight at 48C. The resin was collected and washed four times with phosphate-buered saline plus 0.5% NP-40. The precipitated proteins were analysed on a 7.5% SDS-polyacrylamide gel. For Western blotting experiments, proteins in the immunoprecipitates were separated by SDS ± PAGE and transferred to Immobilon membrane (Millipore, Bedford, MA, USA) in the presence of 0.1% SDS. The membranes were blocked using 5% milk in Tris-buered saline plus 0.1% Tween 20, then probed with the appropriate antibodies. Blots were visualized either by the alkaline phosphatase conjugated second antibody with BCIP/ NBT as substrate, or the horseradish peroxidase conjugated second antibody with ECL (Enhanced Chemiluminescence, Amersham). Microcystin-Sepharose affinity isolations Cell extracts (1 mg) were incubated with 10 ml of Microcystin-Sepharose for 4 h at 48C. The beads were washed four times with 0.5% NP-40 in phosphate-buered saline, and bound proteins were eluted with Laemmli sample buer and subjected to immunoblot analysis. Pull-down experiments using HA-tagged PP2A Cos-7 cells were cultured to 70% con¯uence in 100 mm dishes. Transient transfections were performed in OPTI medium with Trans IT polyamine transfection reagent according to the manufacturer's instructions. Plasmids encoding wild-type (HA)3-tagged PP2A C subunit or inactive (HA)3-tagged H118N-PP2A C subunit (5 mg) were used in the transfections (Chung and Brautigan, 1999). Cells were incubated for 7 h and replaced with complete Dulbecco's modi®ed Eagle's medium. After 24 h of incubation, the cells were washed with phosphate-buered saline, harvested, and lysed. Cell lysates (20 mg) were used to con®rm protein expression by Western blotting. Anti-HA monoclonal antibody was added to the remaining cell lysates. After addition of protein A-Sepharose, incubation was continued for 5 h or overnight. The resin was then washed extensively with 0.5% NP-40 in phosphate buered saline and suspended in phosphate-buered saline (50% slurry). These immobilized wild-type and catalytically inactive PP2A C subunits were used for pull-down experiments, as described in the legend to Figure 6. Effect of PP2A on v-Src in vitro Full-length v-Src was produced in Spodoptera frugiperda (Sf9) cells by infection with a recombinant baculovirus harboring the v-Src gene (Garcia et al., 1993). Immunoanity puri®cation of v-Src was performed as described previously (Garcia et al., 1993). His-tagged Src (SH3-SH2-catalytic domains) was expressed in Sf9 cells and puri®ed on a NiNTA column followed by hydroxyapatite chromatography, as described (Yokoyama and Miller, 1999). Src activity was measured using the phosphocellulose binding assay, as described previously (Yokoyama and Miller, 1999). The assays were carried out at 308C in 20 ml volumes with or without PP2A (Src:PP2A molar ratio=1 : 1.2). The reaction mixtures contained 20 mM Tris-HCl (pH 7.4), 10 mM MgCl2, 0.1 mM Na3VO4, 5 mg/ml BSA, 0.25 mM ATP, 0.5 mM peptide substrate (Glu-Glu-Glu-Glu-Ile-Tyr-Met-Met-MetMet) and [g-32P]ATP (200 ± 300 c.p.m./pmol). Reactions were


terminated by the addition of 50% acetic acid and spotted on p81 paper. After washing with 0.5% phosphoric acid, incorporation of 32P into peptides was measured by liquid scintillation counting. Serine/threonine phosphorylation of Cas in NIH3T3 cells NIH3T3 cells were left untreated or treated for 3 h with (i) 0.2 mg/ml nocodazole; (ii) 80 nM OA and 0.2 mg/ml nocodazole; or (iii) 80 nM OA. OA- or nocodazole-treated cells were washed twice with phosphate-buered saline and collected by the shake-o method. Untreated cells were harvested by trypsinization. Cells were then washed with 25 mM Tris-HCl (pH 7.5) and lysed in lysis buer II [10 mM Tris-HCl (pH 7.5), 5 mM EDTA, 0.2 mM Na3VO4, 10 mM DTT, 1% Nonidet P-40, 1 mM PMSF, 10 mg/ml leupeptin, 10 mg/ml aprotinin] at 48C for 30 min. Cells were centrifuged and resulting supernatants (1 mg cellular proteins) were used for immunoprecipitations with Cas polyclonal antibody or rabbit

IgG as a control. Immunoprecipitates were washed four times with 25 mM Tris-HCl (pH 7.5) plus 0.5% NP-40, and incubated with or without PP2A in the reaction mixture for 1 h at 308C. The reaction mixtures contained 50 mM TrisHCl (pH 7.5), 0.1 mM EDTA, 0.1 mM DTT, 1 mM caeine and 100 mg BSA (Yokoyama, 1995). Immunoprecipitates were then washed with 0.5% NP-40 plus phosphate-buered saline four times and proteins were analysed by SDS ± PAGE and Western blotting analysis using phosphoserine, Cas and PP2A antibodies.

6065

Acknowledgments We thank Dr Steven Anderson for the Ts-72 cells and Dr David Brautigan for the wild-type and inactive (HA)3tagged H118N-PP2A C subunit plasmids. We thank Barbara Craddock for technical assistance. This work was supported by National Institutes of Health Grants CA28146 and CA58530 (to WT Miller).

References Brown MT and Cooper JA. (1996). Biochim. Biophys. Acta., 1287, 121 ± 149. Campbell KS, Auger KT, Hemmings BA, Roberts TM and Pallas DC. (1995). J. Virol., 69, 3721 ± 3728. Cayla X, Goris J, Hermann J, Hendrix P, Ozon R and Merlevede W. (1990). Biochemistry, 29, 658 ± 667. Chaudhuri SK, Ghosh S, Paweletz N and Schroeter D. (1997). Indian J. Exp. Biol., 35, 1044 ± 1054. Chen J, Martin BL and Brautigan DL. (1992). Science, 257, 1261 ± 1264. Chen J, Parsons S and Brautigan DL. (1994). J. Biol. Chem., 269, 7957 ± 7962. Chung H and Brautigan DL. (1999). Cell Signal., 11, 575 ± 580. Evans DR and Hemmings BA. (2000). Mol. Gen. Genet., 264, 425 ± 432. Favre B, Turowski P and Hemmings BA. (1997). J. Biol. Chem., 272, 13856 ± 13863. Garcia P, Shoelson SE, George ST, Hinds DA, Goldberg AR and Miller WT. (1993). J. Biol. Chem., 268, 25146 ± 25151. Garcia-Guzman M, Dol® F, Russello M and Vuori K. (1999). J. Biol. Chem., 274, 5762 ± 5768. Glenn GM and Eckhart W. (1993). J. Virol., 67, 1945 ± 1952. Glover HR, Brewster CE and Dilworth SM. (1999). Oncogene, 18, 4364 ± 4370. Goldberg Y. (1999). Biochem. Pharmacol., 57, 321 ± 328. Goldberg J, Huang HB, Kwon YG, Greengard P, Nairn AC and Kuriyan J. (1995). Nature, 376, 745 ± 753. Jove R and Hanafusa H. (1987). Annu. Rev. Cell Biol., 3, 31 ± 56. Jove R, Mayer BJ, Iba H, Laugier D, Poirier F, Calothy G, Hanafusa T and Hanafusa H. (1986). J. Virol., 60, 840 ± 848. Klemke RL, Leng J, Molander R, Brooks PC, Vuori K and Cheresh DA. (1998). J. Cell Biol., 140, 961 ± 972. Lin FC and Arndt KT. (1995). EMBO J., 14, 2745 ± 2759. Mayer BJ, Jove R, Krane JF, Poirier F, Calothy G and Hanafusa H. (1986). J. Virol., 60, 858 ± 867. Millward TA, Zolnierowicz S and Hemmings BA. (1999). Trends Biochem. Sci., 24, 186 ± 191.

O'Neill GM, Fashena SJ and Golemis EA. (2000). Trends Cell Biol., 10, 111 ± 119. Ogris E, Mudrak I, Mak E, Gibson D and Pallas DC. (1999). J. Virol., 73, 7390 ± 7398. Pallas DC, Shahrik LK, Martin BL, Jaspers S, Miller TB, Brautigan DL and Roberts TM. (1990). Cell, 60, 167 ± 176. Reynolds AB, Roesel DJ, Kanner SB and Parsons JT. (1989). Mol. Cell Biol., 9, 629 ± 638. Sakai R, Iwamatsu A, Hirano N, Ogawa S, Tanaka T, Mano H, Yazaki Y and Hirai H. (1994a). EMBO J., 13, 3748 ± 3756. Sakai R, Iwamatsu A, Hirano N, Ogawa S, Tanaka T, Nishida J, Yazaki Y and Hirai H. (1994b). J. Biol. Chem., 269, 32740 ± 32746. Schlaepfer DD, Broome MA and Hunter T. (1997). Mol. Cell Biol., 17, 1702 ± 1713. Schlaepfer DD and Hunter T. (1998). Trends Cell Biol., 8, 151 ± 157. Sefton BM, Hunter T and Beemon K. (1980). J. Virol., 33, 220 ± 229. Shenolikar S and Nairn AC. (1991). Adv. Second Messenger Phosphoprotein Res., 23, 1 ± 121. Sheppeck JE, Gauss CM and Chamberlin AR. (1997). Bioorg. Med. Chem., 5, 1739 ± 1750. Turowski P, Fernandez A, Favre B, Lamb NJ and Hemmings BA. (1995). J. Cell Biol., 129, 397 ± 410. Vuori K, Hirai H, Aizawa S and Ruoslahti E. (1996). Mol. Cell Biol., 16, 2606 ± 2613. Yamakita Y, Totsukawa G, Yamashiro S, Fry D, Zhang X, Hanks SK and Matsumura F. (1999). J. Cell Biol., 144, 315 ± 324. Yanagida M, Kinoshita N, Stone EM and Yamano H. (1992). Ciba. Found Symp., 170, 130 ± 146. Yokoyama N. (1995). Mol. Cell Biochem., 148, 123 ± 132. Yokoyama N and Miller WT. (1999). FEBS Lett., 456, 403 ± 408. Zheng B, Woo CF and Kuo JF. (1991). J. Biol. Chem., 266, 10031 ± 10034.

Oncogene