Cross-talk between Serine/Threonine Protein Phosphatase 2A and Protein Tyrosine Phosphatase 1B Regulates Src Activation and Adhesion of Integrin αIIbβ3 to Fibrinogen*

Integrin αIIbβ3 signaling mediated by kinases and phosphatases participate in hemostasis and thrombosis, in part, by supporting stable platelet adhesion. Our previous studies indicate that the genetic manipulation of PP2Acα (α isoform of the catalytic subunit of protein phosphatase 2A) negatively regulate the adhesion of human embryonal kidney 293 cells expressing αIIbβ3 to fibrinogen. Here, we demonstrated that small interference RNA (siRNA) mediated knockdown of PP2Acα in 293 αIIbβ3 cells led to the dephosphorylation of Src Tyr-529, phosphorylation of Src Tyr-418 and an increased Src kinase activity. Conversely, overexpression of PP2Acα decreased the basal Src activity. Pharmacological inhibition of PP2Ac in human platelets or PP2Acα knockdown in primary murine megakaryocytes resulted in Src activation. PP2Acα-depleted 293 αIIbβ3 cells did not alter the serine (Ser) phosphorylation of Src but enhanced the Ser-50 phosphorylation of protein tyrosine phosphatase 1B (PTP-1B) with a concomitant increase in the PTP-1B activity. Src activation in the PP2Acα-depleted 293 αIIbβ3 cells was abolished by siRNA mediated knockdown of PTP-1B. Pharmacological inhibition of Src or knockdown of Src, PTP-1B blocked the enhanced activation of extracellular signal-regulated kinase (ERK1/2) and the increased adhesiveness of PP2Acα-depleted 293 αIIbβ3 cells to fibrinogen, respectively. Thus, inactivation of PP2Acα promotes hyperphosphorylation of PTP-1B Ser-50, elevates PTP-1B activity, which dephosphorylates Src Tyr-529 to activate Src and its downstream ERK1/2 signaling pathways that regulate αIIbβ3 adhesion. Moreover, these studies extend the notion that a cross-talk between Ser/Thr and Tyr phosphatases can fine-tune αIIbβ3 outside-in signaling.

Stable platelet-platelet and platelet-extracellular matrix interactions play a critical role in hemostasis and thrombosis. These interactions can be mediated by integrin ␣ IIb ␤ 3 signaling at the sites of vascular injury. The binding of ␣ IIb ␤ 3 to an extracellular ligand-like fibrinogen triggers outside-in signals in platelets via modulation of the activities of several kinases and phosphatases. These signals in turn regulate the cytoskeletal reorganization, which contributes to stable platelet adhesion and spreading events (1). Signaling molecules, including the tyrosine kinase c-Src (2), protein tyrosine phosphatase 1 B (PTP-1B) 3 (3), and the catalytic subunit of protein phosphatase 2A (PP2Ac) (4) either associate constitutively or are recruited to the ␣ IIb ␤ 3 complex in response to integrin engagement. Kinases and phosphatases that assemble within the protein complexes organized by the cytoplasmic tails of ␣ IIb and ␤ 3 could facilitate reversible phosphorylation of multiple effector proteins and mediate outside-in signaling process.
Protein phosphatase 2A (PP2A) is a serine/threonine (Ser/ Thr) phosphatase that regulate cell growth, development, and apoptosis. PP2A holoenzyme contains the catalytic subunit C (PP2Ac) and the structural or scaffolding subunit A (PP2Aa). The A subunit of PP2A interacts with multiple regulatory B subunits and regulates the subcellular localization, substrate specificity, and the catalytic activity of PP2A (5). Although, PP2Ac exhibit ␣ and ␤ isoforms, a 10-fold abundant expression of PP2Ac␣ in most cell types (6) had led us to consider PP2Ac␣ in our previous integrin studies (4). These studies revealed that the depletion of PP2Ac␣ in 293 ␣ IIb ␤ 3 cells resulted in an enhanced adhesion to immobilized fibrinogen (4). However, an underlying mechanism is incompletely understood.
Alternative model systems like 293 ␣ IIb ␤ 3 cells or Chinese Hamster Ovary (CHO) ␣ IIb ␤ 3 cells have proved useful in elucidating the signaling properties of integrin ␣ IIb ␤ 3 (7)(8)(9). In particular, such models are appealing if the role of the signaling molecules under study will be deciphered using a genetic approach. In the context of this study, embryonic lethality of PP2Ac␣-null mice (10), lack of a specific PP2Ac␣ pharmacological inhibitor, and the non feasibility of gene expression in anucleate platelets have limited the extensive use of platelets. The aim of this study in 293 ␣ IIb ␤ 3 cells was to elucidate the molecular events that underpin the PP2Ac␣ mediated negative regulation of ␣ IIb ␤ 3 cell adhesion. In particular; we wished to examine the regulation of c-Src activity (referred to as Src in this manuscript) by PP2Ac␣. c-Src signaling constitutes one of the early signaling events in the outside-in ␣ IIb ␤ 3 signaling process (11).
In this study, we identified that loss of PP2Ac␣ activated Src via PTP-1B. Furthermore, Src and PTP-1B activity was required for the enhanced adhesive phenotype displayed by the PP2Ac␣-depleted 293 ␣ IIb ␤ 3 cells.
siRNA Construct, Transfection, and Adhesion-Stable human embryonal kidney 293 cells overexpressing ␣ IIb ␤ 3 were generated by flow cytometric sorting using monoclonal antibodies specific to ␣ IIb ␤ 3 and cultured in DMEM with 10% FBS as previously described (7). A preformed mix of four independent (SMART pool) siRNAs targeting human PP2Ac␣, PTP-1B, c-Src, PP1c␣, PP1c␤, PP1c␥ 1 , mouse PP2Ac␣, and a nonspecific control siRNA pool were purchased from Dharmacon (Thermo Fisher Scientific, Lafayette, CO). 293 ␣ IIb ␤ 3 cells were transfected with 100 nM siRNA oligonucleotides using siImporter according to the manufacturer's instruction. In some experiments, these cells were transiently transfected using Lipofectamine with cDNA for HA-tagged PP2Ac␣ or the control vector (gift from Dr. A. Verin, University of Chicago, Chicago, IL) After 48 -72 h, the cells were used for either Western blotting, immunoprecipitation or cell adhesion experiments. For ERK1/2 immunoblotting studies, cell lysate was prepared following the treatment with 10 M Src inactive control (PP3) or 5-10 M Src inhibitor (PP2) for 30 min. For adhesion experiments, 1 ϫ 10 5 siRNAtreated cells in Tyrode's buffer, were incubated for 15 min with 5% BSA or 12.5 g/ml fibrinogen-coated 96-well plate as we have described previously (4). In certain experiments, cells were pretreated with either control DMSO, 5-10 M PP2 (Src inhibitor), or 10 M PP3 (inactive analog of PP2). Unbound cells were washed, and the adhered cell was quantified by assaying for acid phosphatase activity at 405 nm. In certain experiments, megakaryocytes were generated from the bone marrow cultures of the BALB/c mice and transfected with control or mouse PP2Ac␣ siRNAs using Mirus transfecting reagent as we have described previously (4).
Src Kinase Activity Assay-Src activity was assayed using HTScan Src kinase assay kit with the modification protocol suggested by the manufacturer. Lysates were generated from PP2Ac␣-depleted 293 ␣ IIb ␤ 3 cells, PP2Ac␣ overexpressing 293 ␣ IIb ␤ 3 cells or platelets treated with DMSO or 50 nM okadaic acid. Src was immunoprecipitated from the lysate by incubating overnight at 4°C with rabbit anti-Src antibody and protein A beads. After rinsing with kinase buffer, the immunoprecipitate was resuspended with 50 l of IX kinase buffer containing 1.5 M biotinylated Src substrate peptide, 20 M ATP and 1.25 M DTT for 30 min and mixed with equal volume of 50 nM EDTA to stop the reaction. 25 l of the reaction solution was transferred to a 96-well streptavidin plates containing 75 l of water and incubated at room temperature for 60 min. The wells were rinsed three times with 0.1% Triton X-100 in TBS and incubated with anti-phosphotyrosine antibody pY-100 (1:1000 dilution) for 120 min at 37°C. Following washing, the wells were incubated with 100 l of HRP-linked goat anti-mouse IgG antibody (1:4000 dilution) and incubated for 30 min. After rinsing, 100 l of 1 mg/ml TMB (3,3Ј,5,5Ј tetramethybenzidine dihydrochloride) was added to the each well and incubated for 10 min at 37°C. The reaction was stopped by adding 100 l of stop solution, and the resulting absorbance at 450 nm was noted.
PTP-1B Activity Assay-Lysate obtained from the control and PP2Ac␣ siRNA-treated 293 ␣ IIb ␤ 3 cells were immunoprecipitated with anti-PTP-1B antibody or mouse IgG. These immunoprecipitates were then assessed for the PTP-1B activity by evaluating the dephosphorylation of a phosphopeptide (RRLIEDAEpYAARG) using a Malachite green assay within the PTP-1B assay kit according to the manufacturer's instructions.
Immunoprecipitation, GST Pull-down and Immunoblotting-Blood was drawn in an acid/citrate/dextrose anticoagulant from normal donors. These donors signed informed consent approved by the Institutional Review Board of Baylor College of Medicine, Houston, TX. Washed platelets were prepared as we previously described (12). Washed platelets were incubated with 50 nM OA or DMSO for 30 min. Platelets, 293 ␣ IIb ␤ 3 (PP2Ac␣ depleted and PP2Ac␣ overexpressing) were lysed with cell lysis buffer containing 1% Triton X-100. Lysate were immunoprecipitated with antiphosphoserine, anti-Src, anti-PP2Ac, or rabbit and mouse IgG antibodies using protein A/G-Sepharose beads. Beads were washed three times, and the proteins were separated by SDS-PAGE and immunoblotted with anti-Src, anti-phosphoserine, and anti-PP2Ac antibodies. The cDNA for PP2Ac␣ in pcMV vector was amplified by PCR and subcloned into a glutathione S-transferase (GST) vector pGEX 4T-1 using BamH1 and Sal1 restriction enzymes. GST or GST-tagged PP2Ac␣ proteins were expressed in Escherichia coli following induction with isopropyl-␤-D-thiogalactopyranoside and purified using glutathione beads. Purified GST or GST-PP2Ac␣ were pre-coupled with glutathione beads and mixed with 100 g of lysate from 293 ␣ IIb ␤ 3 for 3 h at 4°C. Beads were washed three times, and the PP2Ac␣-interacting proteins separated by SDS-PAGE and immunoblotted with Src antibody. 40 g of protein lysate from siRNA-treated cells were separated by a 10% reducing SDS-PAGE, transferred to nitrocellulose, and immunoblotted with anti-phospho Src Tyr-529, anti-phospho Src Tyr-418 and total Src antibodies using ECL. The signals on the films were scanned and the densitometric quantification performed using Image J software from NIH.
Statistics-Statistical significance of the data was analyzed by using a paired Student's t test. Data are expressed as mean Ϯ S.E.

RESULTS
Src Interacts with PP2Ac in 293␣ IIb ␤ 3 Cells and Platelets-To gain mechanistic insights by which PP2Ac␣ might negatively regulate ␣ IIb ␤ 3 signaling, we sought to identify PP2Ac␣-interacting proteins in 293 ␣ IIb ␤ 3 cells, using GST pull-down assays. PP2Ac␣ was expressed as a GST fusion protein in E. coli (Fig. 1A). GST-PP2Ac␣ protein in the pull-down assays detected the scaffolding A subunit of PP2A (PP2Aa), a known binding partner of PP2Ac (5) (Fig. 1B). This indicates that the GST-PP2Ac␣ fusion protein is func- tional in protein interaction studies. More importantly, Src was identified to interact with GST-PP2Ac␣ but not GST alone (Fig.  1C). To verify these findings, co-immunoprecipitation studies were also performed. Immunoblots of PP2Ac but not control IgG immunoprecipitates from 293 ␣ IIb ␤ 3 cells and resting human platelets detected Src (Fig. 1D). These studies indicate that Src is a part of the protein complex that interacts with PP2Ac in 293 ␣ IIb ␤ 3 cells and platelets.
PP2Ac␣ Negatively Regulates Src Activation-The close proximity of PP2Ac and Src prompted us to test whether PP2Ac␣ plays a role in Src activation. Therefore, we analyzed Src activation in 293 ␣ IIb ␤ 3 cells transfected with control siRNAs and four independent siRNAs targeting PP2Ac␣. Fig.  2A revealed specific knockdown of PP2Ac because the levels of a PP2Ac-related phosphatase, PP1c (catalytic subunit of all protein phosphatase 1 isoforms), and actin were comparable between the control and PP2Ac␣ siRNA-treated cells. Phosphorylation of Src Tyr-529 promotes the intramolecular interaction of the C-terminal domain of Src with the SH2 domain and inhibits Src kinase activity (13). Src Tyr-529 phosphorylation was markedly (p ϭ 0.01) decreased in response to the depletion of PP2Ac␣ (Fig. 2B). Conversely, phosphorylation of Src Tyr-418 within the kinase activation domain, which promotes Src activity (13), was increased (p ϭ 0.03) in PP2Ac␣depleted cells (Fig. 2C). Total Src levels in the control and PP2Ac␣ siRNA-treated cells were comparable (Fig. 2, B and C, lower panels). Next, we directly tested whether the depletion of PP2Ac␣ potentiated Src kinase activity. Src was immunoprecipitated from the control and PP2Ac␣ siRNA-treated 293 ␣ IIb ␤ 3 cells (Fig. 2D) and Src kinase activity analyzed as previously described (14). Consistent with the immunoblotting data obtained with the Src phosphospecific antibodies, the kinase activity of Src isolated from the PP2Ac␣-depleted cells were increased (p ϭ 0.03) compared with the control cells (Fig. 2E).
To ascertain whether the Src activation was specific to PP2Ac␣ inhibition, we examined Src Tyr-529 and Src Tyr-418 phosphorylation in 293 ␣ IIb ␤ 3 cells that were depleted for the catalytic subunits of ␣, ␤, and ␥ isoforms of PP1 (PP1c). Fig. 2F reveals that siRNA treatment reduced the expression of PP1c. This knockdown was specific to PP1c because the levels of PP2Ac and actin were comparable between the control and PP1c siRNA-treated cells (Fig. 2F, lower two panels). However, Src Tyr-529 and Src Tyr-418 phosphorylation levels were comparable between the control and PP1c-depleted cells (Fig. 2G). These observations indicate that the Src activation was specific to PP2Ac␣ inhibition.
To further confirm the findings from the siRNA studies, we took a complementary approach and overexpressed a HAtagged PP2Ac␣ in 293 ␣ IIb ␤ 3 cells. Immunoblotting with anti-HA and anti-PP2Ac antibodies confirmed the overexpression of PP2Ac␣ (Fig. 3A). Src was immunoprecipitated from these cells (Fig. 3B) and evaluated for Src kinase activity. Compared with the vector-treated cells (control), PP2Ac␣-overexpressing cells revealed a modest, but significantly lower (by ϳ22%; p ϭ 0.005) basal Src activity (Fig. 3C). Collectively, these studies suggest that Src activity in 293 ␣ IIb ␤ 3 cells can be regulated by PP2Ac␣.
To further establish that the depletion of PP2Ac␣ activates Src in primary cells, we studied Src activation in a bone marrow-derived murine megakaryocytes, an additional model system that has direct relevance to platelet biology. siRNA treatment resulted in the reduction in PP2Ac protein in megakaryocytes (Fig. 4A). Depletion of PP2Ac␣ in megakaryocytes resulted in the dephosphorylation of Src Tyr-529 along with a concomitant increase in the Src Tyr-418 phosphorylation (Fig. 4B). Finally, to examine the role of PP2Ac in a more physiological context, we examined Src activation in human platelets treated with PP2Ac selective inhibitor-OA. OA at concentrations of 10 -100 nM is required to inhibit PP2Ac in intact cells, while 10 M can inhibit cellular PP1c activity (15). Compared with the DMSO-treated platelets, phosphorylation of Src Tyr-529 was distinctly reduced (p ϭ 0.009), whereas the level of Src Tyr-418 phosphorylation was enhanced (p ϭ 0.05) in response to 50 nM OA treatment (Fig. 4, C and D). Similar results were obtained with 100 nM OA (not shown). Total Src level in these studies was not altered (Fig. 4, C and D, lower panel). Consistent with the immunoblotting data obtained with the Src phosphospecific antibodies, the kinase activity of Src isolated from OA-treated platelets were increased (p, 0.05) (Fig.  4E). Taken together, these studies indicate that PP2Ac␣ negatively regulates Src activity.
Depletion of PP2Ac␣ Enhances PTP-1B but Not Src Serine Phosphorylation-Besides phosphorylation on Tyr residues, Src is also phosphorylated by serine/threonine (Ser/Thr) kinases at the N terminus, including the phosphorylation of Src Ser-12 that associate with an increased Src kinase activity (16,17). Because Src is one of the proteins that complex with PP2Ac␣, we considered whether the loss of PP2Ac␣ modulated the overall Ser phosphorylation of Src. Lysates from the control and PP2Ac␣ siRNA-treated cells were immunoprecipitated with control rabbit IgG and phosphoserine antibodies, and the resulting immunoprecipitates were immunoblotted with anti-Src antibody. Src was detected in phosphoserine but not in control IgG immunoprecipitates (Fig. 5A, upper panel). In a reciprocal study, Src immunoprecipitates from the control and PP2Ac␣ siRNA-treated cells revealed comparable immunostaining with anti-phosphoserine antibody (Fig. 5A, lower  panel). Densitometric studies revealed comparable levels of the ratio of phospho Src/total Src in arbitrary units (1.9 Ϯ 0.5 versus 2.01 Ϯ 0.4) in control and PP2Ac␣-depleted cells. Although these studies cannot identify phosphorylation of any specific Ser residues on Src, they imply that the loss of PP2Ac␣ does not significantly alter the overall Ser phosphorylation of Src.
Because Src is also activated by protein tyrosine phosphatases that dephosphorylate Src Tyr-529, we focused our attention on protein tyrosine phosphatases. Although there are several candidate tyrosine phosphatases that can activate Src, we considered protein tyrosine phosphatase 1B (PTP-1B) for the following reasons. 1) Phosphorylation of PTP-1B at Ser-50 by Ser/ Thr kinases enhances PTP-1B activity in vivo (18). 2) PTP-1B has a prominent role in regulating outside-in integrin ␣ IIb ␤ 3 signaling (3). We considered whether the loss of PP2Ac␣ enhanced PTP-1B Ser-50 phosphorylation. PTP-1B immunoprecipitate from the lysates of PP2Ac␣ siRNA-treated ␣ IIb ␤ 3 cells demonstrated moderately enhanced Ser-50 phosphorylation compared with the control cells (Fig. 5B). In a complementary approach, direct immunoblotting studies with anti-PTP-1B Ser-50 antibody revealed ϳ2-fold increased (p ϭ 0.02) PTP-1B Ser-50 phosphorylation in PP2Ac␣ siRNA-treated 293 ␣ IIb ␤ 3 cells (Fig.  5C). We ascertained whether the hyperphosphorylation of PTP-1B Ser-50 in PP2Ac␣-depleted cells affected its enzymatic activity. Compared with the control siRNAtreated cells, loss of PP2Ac␣ resulted in a 1.6-fold (p ϭ 0.009) enhanced PTP-1B activity (Fig. 5D). PTP-1B recovery in the immunoprecipitates from the control, and PP2Ac␣-depleted cells were comparable (Fig. 5E, lower panel). PTP-1B enzymatic activity was specific because the mouse IgG immunoprecipitates only detected the baseline phosphate levels. Collectively, these studies indicated that the loss of PP2Ac␣ in 293 ␣ IIb ␤ 3 cells resulted in an increased PTP-1B Ser-50 phosphorylation that was accompanied by enhanced PTP-1B activity. Because the depletion of PP2Ac enhanced PTP-1B activity and also activated Src, we examined if there exist protein complexes between Src and PTP-1B. Src was detected in the PTP-1B immunoprecipitate from the PP2Ac␣depleted cells but not in the control siRNA-treated cells (Fig. Lysates were immunoblotted with anti-phospho (Src Tyr-529 and Src Tyr-418) antibodies. The blot was reprobed with anti-Src antibody. Densitometric analysis of phospho Src/total Src from four experiments is expressed in arbitrary units. The decreased Src Tyr-529 in PP2Ac␣-depleted cells were significant at *, p ϭ 0.009, while the increased Src Tyr-418 was significant at †, p ϭ 0.05. E, Src kinase activity from DMSO-and OA-treated platelets. The increased Src kinase activity in OA-treated platelets were significant at *, p ϭ 0.05. n ϭ 3. 5E). Although the in vivo association of Src with the protein complex containing PTP-1B appears modest, the influence of such an association on Src signaling is relatively profound (see below in Fig. 6, B and C).
Src Activation in PP2Ac␣-depleted 293 ␣ IIb ␤ 3 Cells Has Functional Consequences for Integrin ␣ IIb ␤ 3 -We have previously shown that the knockdown of PP2Ac␣ in 293 ␣ IIb ␤ 3 cells enhanced their ability to adhere to fibrinogen and increased the activation of extracellular signal-regulated kinase (ERK1/ 2) (4). Importantly, ERK1/2 signaling was linked to ␣ IIb ␤ 3 adhesion in PP2Ac␣-depleted cells because blockade of ERK1/2 activation abolished their increased adhesiveness (4). Although the regulation of Src activity by PP2Ac␣ is integrin ␣ IIb ␤ 3 independent, we evaluated whether the increased Src activity in PP2Ac␣depleted 293 ␣ IIb ␤ 3 cells can regulate signaling and adhesive functions of ␣ IIb ␤ 3 . Therefore, ERK1/2 activation was assessed in PP2Ac␣-depleted cells treated with a widely used Src inhibitor (PP2) or an inactive analog (PP3) (2,14). Treatment with PP2 but not PP3 blocked the enhanced ERK1/2 activation observed in PP2Ac␣-depleted cells (Fig. 7, A and B). These studies suggest that ERK1/2 activation in PP2Ac␣-depleted cells is dependent, in part, on Src. To determine if the increased Src activity in PP2Ac␣-depleted cells contributes to the adhesive phenotype, adhesion assays were performed with Src inhibitors. Consistent with our previous findings, knockdown of PP2Ac␣ significantly (p Ͻ 0.0001) enhanced the adhesion of 293 ␣ IIb ␤ 3 cells to immobilized fibrinogen (Fig. 7C). The inability of parental (plain) 293 cells to adhere to fibrinogen and the inhibition of 293 ␣ IIb ␤ 3 cell adhesion to fibrinogen in the presence of 10E5 (␣ IIb ␤ 3 specific blocking antibody) demonstrates that the adhesion is ␣ IIb ␤ 3 dependent. Importantly, the enhanced adhesion associated with the PP2Ac␣-depleted 293 ␣ IIb ␤ 3 cells to fibrinogen was abolished (p ϭ 0.07) with 10 M PP2. In contrast, 10 M of PP3 did not alter (p Ͻ 0001) the increased adhesiveness of PP2Ac␣-depleted cells (Fig. 7C).
To validate the findings from the Src pharmacological inhibitor studies, adhesion assays were also performed with cells treated with dual PP2A-and Src-directed siRNAs. Compared with the cells treated with only control siRNA, dual treatment of PP2Ac␣ and Src siRNAs decreased the expression of PP2Ac and Src proteins (Fig. 8A). The enhanced adhesive phenotype displayed by the PP2Ac␣-depleted 293 ␣ IIb ␤ 3 cells was abrogated by siRNA-mediated knockdown of Src (Fig. 8B). Because Src activation in the PP2Ac␣-depleted cells is dependent on PTP-1B, we examined the role of PTP-1B in the adhesion of PP2Ac␣-depleted cells. Fig. 8A shows decreased protein expression of PTP-1B and PP2Ac␣ in PTP-1B and PP2Ac␣ siRNA-treated cells. Dual knockdown of PTP-1B and PP2Ac␣ abolished the increased adhesion exhibited by the PP2Ac␣depleted cells (Fig. 8B). Knockdown of only Src or PTP-1B in 293 ␣ IIb ␤ 3 cells did not alter the adhesion. Collectively, these studies indicate that the enhanced Src and PTP-1B activation in PP2Ac␣-depleted 293 ␣ IIb ␤ 3 cells contribute to the increased ␣ IIb ␤ 3 -mediated adhesion to fibrinogen.

DISCUSSION
It is widely accepted that integrin ␣ IIb ␤ 3 function is regulated by cellular signaling that is generated by the opposing actions of protein kinases and protein phosphatases. However, the contribution of Ser/Thr protein kinases but not Ser/ Thr phosphatases in integrin ␣ IIb ␤ 3 function has been extensively investigated. Here, using 293 ␣ IIb ␤ 3 cells, primary murine megakaryocytes and human platelets, we established that PP2Ac␣ negatively regulates Src activity. Mechanistically, loss of PP2Ac␣ leads to the hyperphosphorylation of PTP-1B Ser-50 along with an increased PTP-1B activity. PTP-1B dephosphorylates Src Tyr-529 and activates Src. Moreover, Src activation in PP2Ac␣-depleted cells contributes to the activation of downstream ERK1/2 signaling and the increased adhesiveness to fibrinogen. Thus, this study uncovers a previously unappreciated mechanism, wherein, inhibition of PP2Ac␣ promotes ␣ IIb ␤ 3 adhesiveness by a crosstalk between the family members of Ser/Thr and Tyr phosphatases.
Studies with immortalized cell lines in oncogene-induced signaling have implicated PP2A in regulating Src activity (16). However, these studies were performed with micro molar concentrations of OA that could potentially inhibit PP1, PP2A, and PP4 (19). Thus, the specific role for PP2A in regulating Src activity is still ambiguous. In an in vitro study with purified protein components, PP2A containing the catalytic (C) and the scaffolding (A) subunits interacted with V-Src and inhibited v-Src activity (20). However, it is unclear as to which subunit of the PP2A participated in the regulation of v-Src activity. Using a genetic approach in 293 ␣ IIb ␤ 3 cells, our data indicate that the inhibition of catalytic subunit of PP2A (PP2Ac␣) but not catalytic subunit of PP1 (PP1c) can result in Src activation (Fig. 2). Similar observations were noted in PP2Ac␣-depleted murine megakaryocytes and in human platelets treated with nanomolar concentrations of OA (Fig. 4). Despite these similarities, differences in the level of Src activation were noted between the model system and platelets. Densitometric quantification studies revealed that the depletion of PP2Ac␣ in 293 ␣ IIb ␤ 3 cells decreased Src Tyr-529 phosphorylation by ϳ75%, increased Src Tyr-418 phosphorylation by ϳ45% and produced a ϳ30% increase in Src kinase activity (Fig. 2, B-E). In platelets, OA decreased Src Tyr-529 phosphorylation by ϳ95%, increased the Src Tyr-418 phosphorylation by ϳ300% and produced a ϳ35% increased Src activity (Fig. 4, C-E). The extent of increase in Src activity in OA-treated platelets was not as substantial as the data from the phosphospecific antibodies would have pre-dicted. These observations suggest that potential differences in the mechanisms of Src activation between the PP2Ac␣-depleted 293 ␣ IIb ␤ 3 cells and the OA-treated platelets may exist. Thus, findings from the model system should be cautiously interpreted in the context of platelets. Additional studies using platelets from a megakaryocytic lineage specific PP2Ac␣-null mouse, as and when available in the future, should be considered to validate these findings.
Activated Src family kinases like c-Src and lck phosphorylates PP2Ac at Tyr-307 and inhibits the activity of PP2A (21). Such an interelationship between Src and PP2A implies that a pool of PP2A and Src could be in close proximity to each other. Indeed, our findings in resting platelets and 293 ␣ IIb ␤ 3 cells reveal a complex of PP2Ac and Src (Fig. 1). It is likely that protein complex containing Src and PP2A might be a part of the multiprotein complex anchored by integrin ␣ IIb ␤ 3 in platelet and/or 293 ␣ IIb ␤ 3 cells. Association of c-Src with the ␤ 3 subunit (2) and the interaction of PP2Ac with the ␣ IIb subunit (4) support this notion. Indeed, previous studies have identified a multiprotein complex containing the SV40 small t-antigen and polyoma virus middle T-antigen to contain PP2A and Src (22,23).
The data in this manuscript raise two interesting questions. First, how can inhibition of a Ser/Thr phosphatase (PP2Ac␣) activate a tyrosine kinase (Src)? While the occurrence of such a cross-talk has been widely accepted (24), the underpinning mechanisms are not clear. Inhibition of PP2Ac is likely to alter the Ser/Thr phosphorylation of multiple cellular proteins, including some that have the potential to regulate Src activity. Indeed, Ser phosphorylation of the N terminus of Src by okadaic acid is reported to stimulate c-Src kinase activity (16,17). Ser phosphorylation of PTP-1B by okadaic acid (25) and the specific phosphorylation of PTP-1B Ser-50 by Ser/ Thr kinases are reported to stimulate ϳ3 fold basal PTP-1B activity (18) or mildly (ϳ25%) impair the dephosphorylation of insulin receptor substrate following insulin stimulation (26). Our studies show that the depletion of PP2Ac␣ resulted in the hyperphosphorylation of PTP-1B Ser-50 with a concomitant increase in PTP-1B activity (Fig. 5). Importantly, Src activation in PP2Ac␣-depleted cells was largely mediated by PTP-1B because the dephosphorylation of Src Tyr-529 and the phosphorylation of Src Tyr-418 were not evident in cells with dual FIGURE 7. Pharmacological inhibition of Src by PP2 but not by an inactive control (PP3) blocks the enhanced ERK1/2 signaling and abrogates the increased adhesiveness of PP2Ac ␣ -depleted cells. A, lysates was prepared from either control or PP2A c␣ -depleted cells followed by treatment with PP3 (inactive Src inhibitor) or PP2 (Src inhibitor). They were separated by 10% SDS-PAGE and ERK activation was assessed by immunoblotting using antibodies specific for the active p44/42 ERK (pERK1/20 or total ERK. B, densitometric quantification of pERK/total ERK from three experiments in arbitrary units. C, effect of Src inhibitor (PP2) or inactive control for PP2 (PP3) on the increased adhesion of PP2Ac ␣ -depleted cells to fibrinogen. n ϭ 4 -5, *, p Ͻ 0.0001. Error bars are too narrow to be seen with 10E5 inhibition and plain 293 cell adhesion studies.
knockdown of PP2Ac and PTP-1B (Fig. 6). Second, how can loss of PP2Ac␣ lead to an increased PTP-1B Ser-50 phosphorylation? It is currently unclear whether PTP-1B Ser-50 is a direct substrate of PP2Ac. However, our preliminary studies with GST-PP2Ac␣ pull-down assays and PP2Ac co-immunoprecipitation studies failed to detect a complex between PP2Ac and PTP-1B (not shown). Whether loss of PP2Ac␣ leads to the activation of Ser/Thr kinases that phosphorylates PTP-1B Ser-50 is currently being investigated.
PTP-1B is recruited to the multiprotein complex anchored by the integrin ␣ IIb ␤ 3 , including c-Src protein. On the other hand, C terminus Src kinase (CSK) that phosphorylates Src Tyr-529 dissociates from the ␣ IIb ␤ 3 protein complex. Such dynamic interactions constitutes an early mechanism that leads to integrin-dependent c-Src activation following fibrinogen binding to ␣ IIb ␤ 3 (2, 3). Interestingly, Src activation by the inhibition of PP2Ac␣ in our studies is independent of integrin activation. Nevertheless, the regulation of Src by PPAc␣ could also occur in the milieu of integrin-associated protein complexes. For example, in resting platelets, the presence of PP2Ac in the ␣ IIb protein complex could contribute to maintain ␤ 3 -bound c-Src in an inactive conformation. Fibrinogen binding to the activated ␣ IIb ␤ 3 suppresses integrin ␣ IIb -bound PP2Ac activity, which in turn could facilitate ␤ 3 -associated c-Src activation via PTP-1B activity within the integrin complex. Finally, activated Src could phosphorylate PP2Ac Tyr-307 to further inhibit PP2Ac activity and thereby maintain Src activation. In line with this argument, we have previously shown that PP2Ac associates with integrin ␣ IIb ␤ 3 complex and fibrinogen binding suppresses ␣ IIb ␤ 3 -associated PP2Ac activity (4).
Studies in human platelets (2,27) and in mouse models (28,29) have revealed an essential role for Src and/or Src-␣ IIb ␤ 3 interaction in outside-in ␣ IIb ␤ 3 integrin signaling dependent functions like adhesion, and spreading on fibrinogen. We observed that PP2Ac␣-depleted cells exhibited Src activation and increased adhesion to fibrinogen. siRNA-mediated knockdown of Src or pharmacological inhibition of Src with PP2 but not an inactive analog PP3 selectively abolished the increased adhesion of PP2Ac␣-depleted 293 ␣ IIb ␤ 3 cells. Consistent with the role of PTP-1B in Src activation, knockdown of PTP-1B abrogated the increased adhesion of PP2Ac␣-depleted 293 ␣ IIb ␤ 3 cells (Fig. 8). It is likely that ERK1/2 in PP2Ac␣-depleted cells could be one of the downstream effectors of Src in regulating ␣ IIb ␤ 3 adhesion. Such a notion is supported by the following observations: (a) ERK1/2 activation was noted in PP2Ac␣-depleted cells (4), (b) Src inhibitor partially ablates the increased ERK activation in PP2Ac␣-depleted cells (Fig. 7A), and (c) the ability of ERK and Src inhibitors to block the increased adhesiveness of PP2Ac␣ cells to fibrinogen (4). Other investigators have reported that inhibition of Src activity blocked the increased migratory property of the endothelial and carcinoma cells treated with generic PP2A inhibitor (30,31).
In summary, our studies indicate that the inhibition of PP2Ac␣ leads to the hyperphosphorylation of PTP-1B Ser-50 along with increased PTP-1B activity, which dephosphorylates Src Tyr-529. Increased Src activity due to the depletion of PP2Ac␣ contributes to the enhanced ERK1/2 signaling and the increased ␣ IIb ␤ 3 adhesiveness to fibrinogen. In a broader context, regulation of Src activity by a cross-talk between the Ser/ Thr and Tyr phosphatases may provide additional mode of regulatory mechanisms, during processes like cell differentiation, proliferation and the pathological conditions associated with tumor growth and progression.