J Biol Chem, Vol. 273, Issue 48, 31890-31900, November 27, 1998

Protein-tyrosine Phosphatase  Regulates Src Family Kinases and Alters Cell-Substratum Adhesion^*

Kenneth W. Harder^§, Niels P. H. Moller^¶, James W. Peacock, and Frank R. Jirik

From the  Centre for Molecular Medicine and Therapeutics and the Department of Medicine, University of British Columbia, Vancouver, British Columbia V5Z 4H4, Canada and ^¶ Novo Nordisk, Novo Alle, DK-2880 Bagsvaerd, Denmark

	ABSTRACT

Top Abstract Introduction Procedures Results Discussion References

The roles of protein-tyrosine phosphatases (PTPs) in processes such as cell growth and adhesion are poorly understood. To explore the ability of specific PTPs to regulate cell signaling pathways initiated by stimulation of growth factor receptors, we expressed the receptor-like PTP, PTP, in A431 epidermoid carcinoma cells. These cells express high levels of the epidermal growth factor (EGF) receptor and proliferate in response to the autocrine production of transforming growth factor-. Conversely, EGF stimulation of A431 cells in vitro leads to growth inhibition and triggers the rapid detachment of these cells from the substratum. Although PTP expression did not alter the growth characteristics of either unstimulated or EGF-stimulated cells, this phosphatase was associated with increased cell-substratum adhesion. Furthermore, PTP-expressing A431 cells were strikingly resistant to EGF-induced cell rounding. Overexpression of PTP in A431 cells was associated with the dephosphorylation/activation of specific Src family kinases, suggesting a potential mechanism for the observed alteration in A431 cell-substratum adhesion. Src kinase activation was dependent on the D1 catalytic subunit of PTP, and there was evidence of association between PTP and Src kinase(s). PTP expression also led to increased association of Src kinase with the integrin-associated focal adhesion kinase, pp125^FAK. In addition, paxillin, a Src and/or pp125^FAK substrate, displayed increased levels of tyrosine phosphorylation in PTP-expressing cells and was associated with elevated amounts of Csk. In view of these alterations in focal adhesion-associated molecules in PTP-expressing A431 cells, as well as the changes in adhesion demonstrated by these cells, we propose that PTP may have a role in regulating cell-substratum adhesion.

	INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

Reversible protein phosphorylation is a widely employed mechanism for regulating enzyme activity, assembly, and localization of protein complexes and gene transcription within eukaryotic cells (1, 2). Protein phosphorylation also controls cytoskeleton organization during cell adhesion to extracellular matrix (ECM)¹ or to other cells during processes such as morphogenesis, cell migration, differentiation, and metastases (3). Indeed, similar to growth factor receptor-mediated signal transduction, engagement of cell adhesion molecules is followed by the rapid activation of specific protein tyrosine kinases and the ensuing assembly of multimeric protein complexes at sites of cell adhesion (4). Moreover, many of the signal transduction molecules activated or phosphorylated in response to ligand binding to growth factor receptors are also regulated by cell adhesion (5, 6).

Due to their ability to regulate protein phosphotyrosine levels, PTPs are undoubtedly essential to such processes as proliferation, differentiation, and cell adhesion (7-9). However, the roles of specific PTPs in the regulation of growth factor receptor initiated signal transduction events and cell adhesion remain poorly understood. In this study, we have evaluated the ability of the receptor-like phosphatase PTP to regulate EGF-receptor-dependent cell signaling processes in the human epidermoid carcinoma cell line A431. Structurally, the widely expressed PTP is composed of a heavily N- and O-glycosylated 123- or 132-residue alternatively spliced extracellular domain, a transmembrane region, and two tandem cytoplasmic phosphatase domains, as is characteristic of the majority of receptor-like PTPs (10-13).

A431 cells have been well characterized with respect to EGF-dependent signal transduction. These cells express high levels of the EGF receptor on their surface (0.5-3.0 × 10⁶/cell) (14), and proliferate in response to autocrine production of transforming growth factor  (15). Although EGF enhances the growth of A431 cell-derived tumors in nude mice, when grown as a monolayer in culture, these cells, similar to squamous cell carcinoma cell lines overexpressing EGF receptors, are inhibited by high concentrations of EGF (14, 16-21). In addition, EGF stimulation of A431 cells causes dramatic change in cell morphology, including extensive membrane ruffling, filipodia extension, and changes in cytoskeletal organization and cell adhesion. These processes culminate in the rounding-up and retraction of these cells from the substratum (22-24). Thus, A431 cells have proven invaluable to studies of cell growth, differentiation, and cell adhesion as regulated by receptor tyrosine kinase activity.

We found that PTP expression in A431 cells led to a PTP D1-dependent increase in cell-substratum adhesion and inhibited EGF-induced cell rounding and lift-off. This PTP-dependent phenotype was not restricted to EGF stimulated A431 cells, as PTP expression in BHK-IR cells also led to resistance of these cells to insulin-induced cell rounding and detachment from the substratum. These results suggest that PTP might be capable of regulating the activities of molecules involved in cell-substratum adhesion. In keeping with this hypothesis, we found that in A431 cells PTP could be co-immunoprecipitated with Src kinase(s). PTP expression was also associated with the dephosphorylation and/or activation of specific Src kinases. Moreover, Src kinases immunoprecipitated from PTP-expressing A431 cells were associated with elevated levels of focal adhesion kinase (FAK), a molecule activated by stimuli such as integrin-dependent cell adhesion to the substratum, v-src transformation, and growth factor or neuropeptide stimulation (reviewed in Ref. 25). PTP expression was also associated with an increase in the tyrosine phosphorylation of paxillin, a protein localized to focal adhesions and a putative substrate of FAK and/or Src kinases (26-28). Paxillin obtained from PTP-expressing cells was complexed with increased levels of Csk, suggesting a possible feedback loop in the regulation of Src activity in these cells and supporting previous observations linking the activation of Src kinases with changes in the intracellular localization of Csk (29-33). The dephosphorylation of Src kinases, together with the observed PTP D1-dependent changes in cell adhesion and the tyrosine phosphorylation and association of FAK and paxillin with Src and Csk kinases, support the hypothesis that PTP may be involved in the regulation of cell-substratum adhesion.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Cells and Plasmids-- A431 cells were obtained from the American Type Culture Collection and were grown in Dulbecco's modified Eagle's medium containing 10% fetal calf serum, antibiotics, and 50 µM -mercaptoethanol. The PTP cDNA was obtained from a human HepG2 cell line cDNA library (Stratagene) as described previously (10). The wild type and catalytically inactive forms of PTP (containing a cysteine-to-alanine mutation at residue 433 within the first catalytic domain, D1 C433A) were subcloned into the eukaryotic expression vector pBCMGNeo. This vector contains the cytomegalovirus immediate-early gene promoter and 79% of the bovine papilloma virus genome, allowing episomal replication of transfected plasmids (34). Plasmids were introduced into A431 cells by electroporation, and G418 resistant clones were selected. The BHK cell line overexpressing the human insulin receptor (BHK-IR) (35) was maintained at 37 °C under 5% CO₂ in Dulbecco's modified Eagle's medium containing 4.5 g/liter glucose, 10% fetal calf serum, 2 mM L-glutamine, 1 µM methotrexate, and penicillin/streptomycin. This cell line was used to establish stable BHK-IR cell lines overexpressing PTP in a functionally dependent way as described previously (36). BHK-IR/PTP cells were maintained in complete medium in the presence of 100 nM insulin.

Antibodies-- PTP-specific antibodies were produced by immunization of New Zealand White rabbits with recombinant PTP cytoplasmic domain containing residues 167-793 (PTP-2) or with N-terminal cysteine-linked keyhole limpet hemocyanin conjugated synthetic peptides (37), corresponding to amino acids 20-60 within the extracellular domain of PTP (PTP-ext), or residues 512-558, corresponding to the region separating the two catalytic domains (PTP-1). Antibodies were affinity purified on thiol-Sepharose peptide or CNBr-Sepharose recombinant PTP specific affinity columns. The anti-Src mAb 327 (provided by J. Brugge, ARIAD Pharmaceuticals), anti-Fyn mAb (provided by R. Perlmutter, University of Washington, Seattle, WA), anti-Yes antiserum (from J. Bolen, Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ), and anti-Csk antiserum (from J. Cooper and B. W. Howell, Fred Hutchinson Cancer Research Center) were used to immunoprecipitate each kinase. The antibody SRC2 (Santa Cruz Biotechnology), specific for the C-terminal peptide sequence 509-533 of Src and the conserved sequences of Fyn and Yes, was also used to immunoprecipitate and immunoblot Src, Fyn, and Yes. Antibodies against phosphotyrosine (4G10), paxillin, FAK, and Src (GD11) were obtained from Upstate Biotechnology Inc. and Transduction Laboratories. Anti-Lyn kinase antibodies were obtained from Santa Cruz Biotechnology Inc.

A431 Cell Lysis-- A431 cells were lysed on ice for 30 min in buffer containing 1% Nonidet P-40, 10% glycerol, 50 mM NaCl, 50 mM Tris, pH 7.5, 2 mM EDTA, 1 mM sodium orthovanadate, 10 mM sodium fluoride, 1 mM phenylmethylsulfonyl fluoride, 1 mM soybean trypsin inhibitor, and 100 µM leupeptin. Insoluble cellular debris was removed by ultracentrifugation at 30,000 × g for 30 min. Protein concentrations were estimated with the bicinchoninic acid assay (Pierce).

Kinase Assays-- Src kinase activity from transfected A431 clones was assessed by autokinase and enolase assays. Src kinase was immunoprecipitated from 500 µg of cell lysate with 1 µg of mAb 327. Immune complexes were collected with 75 µl of a 30% slurry of rabbit anti-mouse IgG preabsorbed protein A-Sepharose. Beads were washed in cell lysis buffer, radioimmune precipitation buffer (lysis buffer containing 1% Nonidet P-40, 0.5% deoxycholic acid, and 0.1% SDS), and then kinase buffer (100 mM PIPES, pH 7.0, 5 mM MnCl₂, and 10 µM vanadate) before resuspension in kinase buffer containing 25 µM ATP and 10 µCi [-³²P]ATP (3000 Ci/mmole) with or without 10 µg of acid-denatured enolase. Kinase assays were performed at 25 °C for 15 min before termination by addition of 2× Laemmli buffer. 50% of each immunoprecipitate was immunoblotted with anti-Src antiserum (GD11) to ensure that equal quantities of Src were used in each assay.

Bacterial Expression-- Regions corresponding to the entire cytoplasmic domain of PTP (PTP-D1+D2, amino acids 167-793), the first catalytic domain (PTP-D1, amino acids 167-555), and the C-terminal phosphatase domain (PTP-D2, amino acids 510-793), were polymerase chain reaction-amplified with Vent DNA polymerase (New England Biolabs). The cDNA sequence of each PTP fragment was confirmed by DNA sequencing before subcloning into pGex 2T-tag, a modified pGex 2T plasmid, containing an expanded polylinker and sequence encoding the 10-residue hemagglutinin epitope tag derived from influenza virus (38). Thrombin cleavage of glutathione S-transferase (GST) fusion proteins generates proteins containing the hemagglutinin epitope at the N terminus. Luria broth cultures (500 ml) of UT5600 bacteria (New England Biolabs) containing the various pGex plasmids were grown to absorbance 0.6-0.9 at 37 °C. Cells were then shifted to 26 °C and induced overnight with 100 µM isopropyl-1-thio--D-galactopyranoside. Bacteria were sedimented and lysed by sonication in buffer composed of 50 mM Tris, pH 7.5, 150 mM NaCl, 5 mM -mercaptoethanol and 1 mM phenylmethylsulfonyl fluoride. Triton X-100 was then added to 1% final concentration, and cellular debris were removed by ultracentrifugation at 30,000 × g. The supernatant was removed and incubated with 1 ml of a 50% slurry of glutathione-Sepharose (Amersham Pharmacia Biotech) for 1 h. The beads were thoroughly washed before cleavage in 1 ml of buffer containing 50 mM Tris, pH 8.0, 2.5 mM CaCl₂, 150 mM NaCl, 10 mM -mercaptoethanol, and 50 µl of thrombin (400 µg/ml). Glycerol was added to a final concentration of 15% before storage at 80 °C.

Malachite Green Phosphatase Assay-- Phosphatase activity was determined using the malachite green microtiter plate (MGMP) phosphatase assay to detect the release of phosphate from phosphotyrosine-containing synthetic peptides as described previously (38). Briefly, recombinant PTP or PTP immunoprecipitated directly from A431 cell lysates was incubated with phosphopeptides in buffer containing 25 mM MES, pH 6.0, and 0.1 mM -mercaptoethanol. Enzyme reactions were carried out in half-volume microtiter plates (Costar) in a final volume of 25 µl for the indicated times. Phosphatase reactions were terminated and free phosphate detected by addition of 100 µl of malachite green solution to each well. Changes in absorbance at 620 nm of each well were measured in an enzyme-linked immunosorbent assay plate reader and phosphate release was determined by comparison to a standard curve (38). Phosphopeptides Src-⁵²⁷YTSTEPQpYQPGENL and CSF-1 receptor ⁷⁰⁸YIHLEKKpYVRRDSG were synthesized as described previously (38, 39). The activity of PTP-D2 toward para-nitrophenylphosphate (40) was detected using the MGMP assay as above.

A431/EGF Growth Inhibition Assay-- Passage number 10 or less control vector alone-transfected and PTP-overexpressing A431 cell lines were plated at 10⁴ cells/well in quadruplicate in 96-well plates. EGF was added, and cell proliferation was determined 4 days later by WST-1 (4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate) assay (Boehringer Mannheim).

BHK-IR and BHK-IR/PTP Growth Curves-- The effect of PTP expression on the growth characteristics of BHK-IR cells was analyzed by comparing growth curves obtained with and without insulin as described previously (36). In short, five 6-well plates were plated at 2 × 10⁴ cells/well. After 24 h, the number of adherent cells per well was determined in one plate (day 0). Three wells in each of the remaining four plates then received insulin at a final concentration of 100 nM (+ insulin) with the other three wells serving as controls ( insulin). The number of adherent cells was determined in one plate after a further 24 h of incubation (day 1), whereas the remaining plates were washed thoroughly three times to remove nonadherent cells. Fresh medium with and without insulin was added. This procedure was repeated every 24 h for the next three days. Similar results were obtained in two independent experiments.

Cell-Substratum Adhesion Assay-- During routine passaging of PTP-overexpressing A431 clones, we observed that these cells were resistant to removal from the substratum by low concentrations of trypsin (0.05%). To quantitate this characteristic, an adhesion assay was developed to assess cell-substratum adhesion. Cells containing vector alone, PTP, or PTP (D1 C433A) were plated in 96-well flat-bottom plates in quadruplicate at 10⁴ cells/well. Cells were allowed to adhere and spread in serum-containing medium to 75-85% confluency (1-2 days). Medium was then carefully removed so as not to disturb the cell monolayer, and the cells were gently washed 3-5 times with 100 µl of Ca²⁺- and Mg²⁺-free PBS/well/wash over a period of 20-30 min. During this treatment, control cells rounded and lifted off the substratum in a manner that was dependent on the number and duration of the PBS washes. To be able to discern the lower adherence of control and PTP (D1 C433A)-transfected A431 cells, the number and duration of PBS washes was reduced. Wash solutions were then discarded, and 100 µl of culture medium was added to each well. One hour later, 25 µl of 3-(4,5-dimethyl thiazol-2-yl)-2,5-diphenyl tetrazolium bromide (5 mg/ml) was added, and the cells remaining in each well were stained for 2.5 h. Medium was then carefully removed, 100 µl of Me₂SO was added to each well to dissolve the precipitated 3-(4,5-dimethyl thiazol-2-yl)-2,5-diphenyl tetrazolium bromide crystals, and the absorbance at 550 nm was determined using an enzyme-linked immunosorbent assay plate reader. The 3-(4,5-dimethyl thiazol-2-yl)-2,5-diphenyl tetrazolium bromide assay allowed linear detection of between 1000 and 70,000 cells/well. The absorbance of each well was compared with untreated wells to determine the percentage of cells removed by each washing protocol. For each cell line, the percentage difference in absorbance was found to be equivalent to the percentage of cells removed. To analyze adherence changes in response to agents known to alter cell-substratum adhesion, EGF (100 ng/ml), pervanadate (100 µM vanadate, 2 mM H₂O₂), and EDTA (10 mM) were added to the PBS washes, and the cells were treated as above.

Peptide Binding Assay-- Lysates from A431 cells containing vector alone or from PTP-expressing or PTP (D1 C433A)-expressing cells were incubated for 1 h with Src Tyr⁵²⁷ phosphopeptide immobilized on CNBr-activated Sepharose. Beads were thoroughly washed in cell lysis buffer before resuspension in Laemmli buffer. Src family kinases precipitated by the beads were separated by 10% SDS-PAGE, transferred to Duralose membrane (Stratagene) and immunoblotted with SRC2 antisera or the Src-specific antibody (GD11). Immobilized nonphosphorylated Src Tyr⁵²⁷ peptide was used as a control. Blots were developed using horseradish peroxidase-linked goat anti-rabbit or goat anti-mouse antiserum and the ECL system (Amersham).

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Expression of PTP in A431 Cells-- To investigate the effects of PTP expression, A431 cells were transfected with cDNAs encoding the 123-residue extracellular domain-containing isoform of PTP, as well as a catalytically inactive mutant form of PTP (PTP D1 C433A), using the expression vector pBCMGNeo (34). G418-resistant clones were obtained with similar frequency in all three instances. Three vector alone transfected control A431 lines and 10 clones of PTP-transfected and PTP (D1 C433A)-transfected cells were examined for PTP expression. Three clones of each, in the case of PTP and PTP (D1 C433A) transfectants, were then selected based on their having similar levels of PTP expression. Results presented are representative examples of each group of clones.

A431 cells normally express relatively low levels of endogenous PTP (Fig. 1A). PTP expression was determined by immunoblot analysis of A431 total cell lysates with polyclonal anti-peptide antibodies (anti-PTP-1) (Fig. 1A) or with anti-recombinant PTP-specific antibodies (anti-PTP-2) (data not shown). Both of these antibodies recognized proteins of approximately 130-150, 100, 85, and 68 kDa. N- and O-glycosylation of the predicted 85-kDa PTP polypeptide chain results in a mature protein of between 130 and 150 kDa (11). A 100-kDa form of PTP, observed in immunoblots of lysates derived from PTP-expressing cells (Fig. 1A), was also immunoprecipitated from these lysates with antibodies specific for the extracellular region of PTP (Fig. 1B) and likely corresponds to an N-glycosylated precursor of the larger form. This is in agreement with Daum et al. (11), who reported that antibodies against baculovirus-expressed PTP recognized a glycosylation-dependent epitope in the extracellular domain of PTP. The anti-PTP-extracellular domain antibodies used in our study were generated against residues 20-60 of the extracellular domain, a region containing multiple potential N- and O-glycosylated residues. Thus, the PTP-extracellular domain antiserum likely recognized an incompletely glycosylated PTP species. This antibody, however, bound to the surface of PTP-overexpressing cells with little or no binding observed to vector alone transfected control cells, demonstrating PTP cell-surface expression in the transfected clones (data not shown).

View larger version (43K): [in this window] [in a new window]   Fig. 1.   Expression of PTP in A431 cells. A, cell lysates from vector alone control (Ctr), PTP, and PTP (D1-C433A) transfected cells were separated by SDS-PAGE, transferred to membrane, and immunoblotted with anti-PTP-1 antibodies. B, alternatively, PTP was immunoprecipitated from the above lysates with antibodies against the extracellular domain of PTP (anti-PTP-ext) and immunoblotted with anti-PTP-1 antibodies.

A431 and BHK-1R Cells Expressing PTP Are Resistant to the Cell-rounding and Adhesion-disrupting Effects of Growth Factors-- To assess whether PTP expression would alter the EGF-induced growth inhibition response of A431 cells, we treated vector alone control cells and PTP-transfected clones with various concentrations of EGF. Cell numbers were then determined by WST-1 assay (as described under "Experimental Procedures"). Maximal inhibition of vector control transfected cell growth was observed at EGF concentrations between 5-10 ng/ml (Fig. 2A). The response of two PTP-expressing A431 clones (PTP-1 and PTP-2) to EGF are shown in Fig. 2A. The expression of PTP in A431 cells was unable to rescue these cells from the growth inhibitory effects of EGF. However, a dramatic change in cell morphology was evident following exposure of A431 cells to EGF in serum-free conditions. Whereas EGF caused control A431 cells to round-up and lift-off the substratum within 5-10 min (Fig. 2, B and C, rounded phase-bright cells in A431-Ctr 1 and A431-Ctr 2 EGF-treated panels) clones expressing PTP remained adherent and spread following exposure to EGF (Fig. 2, B and C, clones PTP-1 and PTP-3). This phenotype was observed in all of the PTP-expressing clones and was dependent on the catalytic activity of PTP D1 (data not shown).

View larger version (60K): [in this window] [in a new window]   Fig. 2.   The effects of EGF on cell growth and cell morphology. A, inhibition of A431 cell growth by EGF. The growth of vector alone (Ctr) and two A431 clones expressing PTP (PTP-1 and PTP-2) was assessed in the presence of indicated concentrations of EGF. Curves represent the percentage of growth inhibition of each A431 cell-line after 4 days of growth in the specified concentration of EGF. Cell numbers were determined by WST-1 staining of cells as outlined under "Experimental Procedures." Similar EGF-dependent growth inhibition was observed in three independent A431 clones expressing PTP. B, PTP-expressing A431 cells are resistant to the cell rounding effects of EGF. Control vector alone (Ctr-1) and PTP-transfected A431 cells (clone PTP-1) were washed in PBS and either stimulated with 100 ng/ml EGF for 10 min (+) or left untreated () (magnification, × 10). (C) A431 clones Ctr-2 and PTP-3 were treated as in B except that cells were stimulated with EGF for 60 min (magnification, × 32).

A previously established BHK-IR cell line (35) responds to insulin stimulation with growth inhibition, cell rounding, and detachment from the substratum of cell culture dishes (36). These cells round up quickly after addition of insulin, with detachment being detectable as early as 1 h after the addition of insulin. The optimal insulin concentration required to bring about this effect was approximately 100 nM, with the half-maximal effect being observed at a concentration of approximately 1 nM. In contrast, stable BHK-IR cell lines overexpressing PTP failed to round up and remained attached to the substratum (data not shown). Thus, in these two different cell model systems, PTP expression was capable of inhibiting adhesion-disrupting effects induced by the activation of two distinct receptor tyrosine kinases.

PTP Expression Prevents Insulin-induced Growth Inhibition of BHK-IR Cells-- We have previously shown that insulin completely abolishes the growth of BHK-IR cells while having no effect on the parental BHK cell line (36). Based on these phenotypic changes, it was possible to introduce a novel selection procedure that allowed establishment of stable BHK-IR cell lines expressing PTPs in a functionally dependent manner. To further assess the effect of PTP expression on insulin-induced growth inhibition, we compared the growth of BHK-IR/PTP cells with the BHK-IR cell line with and without insulin. As shown in Fig. 3 and in contrast to the effect of PTP expression in A431 cells, this PTP rendered the BHK-IR cell line partially resistant to the growth inhibitory effect of insulin.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   BHK-IR cells overexpressing PTP are resistant to insulin-dependent growth inhibition and cell detachment. BHK-IR and BHK-IR/PTP cell proliferation, in the absence (solid lines) or presence (dashed lines) of 100 nM insulin, was determined by counting adherent cells. Insulin caused the rounding up and detachment of BHK-IR cells from the substratum with cessation of cell growth (left panel). BHK-IR cells expressing PTP remained partially resistant to the adhesion disrupting, growth inhibiting effects of insulin (right panel) (± insulin).

Increased Cell-Substratum Adhesion of PTP-expressing A431 Cells-- A431 cells overexpressing PTP, grown in the absence of EGF, were also more adherent to the substratum than control vector alone transfected A431 cells. This characteristic was quantified using an adhesion assay based on measuring the resistance of PTP-expressing cells to rounding-up and release from the substratum following incubation in PBS (described under "Experimental Procedures"). Although greater than 80% of control A431 cells were removed from the substratum after four gentle washes in PBS (e.g. Ctr (untreated) A₅₅₀ ~ 2.3, Ctr (PBS-washed) A₅₅₀ ~ 0.3), only 30% of cells expressing PTP were removed (e.g. PTP-1 (untreated) A₅₅₀ ~ 2.2, PTP-1 (PBS-washed) A₅₅₀ ~ 1.4) (Fig. 4A). Results of the adhesion assay shown in Fig. 4A were obtained with A431 clone PTP-1, whereas those results shown in Fig. 2B were obtained with clones PTP-1 and PTP-3. To determine whether this characteristic was dependent on the catalytic activity of PTP, or possibly due to a PTP extracellular domain-ECM interaction, we also evaluated the adhesion phenotype of PTP (D1 C433A)-expressing cells. To be able to assess the differences in adhesion of control vector and PTP (D1 C433A)-transfected cells, the number and duration of the PBS washes was reduced as compared with Fig. 4A. Although PTP (D1 C433A) protein levels were similar to those of cells transfected with wild type PTP (Fig. 1, A and B), these cells were removed from the substratum more readily than control vector-alone transfected cells (Fig. 4B) (e.g. Ctr (untreated) A₅₅₀ ~ 2.3, Ctr (PBS-washed) A₅₅₀ ~ 2.3, PTP-D1 C433A (untreated) A₅₅₀ ~ 2.4, PTP-D1 C433A (PBS-washed) A₅₅₀ ~ 1.8). The results indicated that the catalytic activity of PTP was essential to the increased cell-substratum adhesion phenotype of the PTP-expressing A431 clones. Consistent with this conclusion, addition of the PTP inhibitor pervanadate to the PBS washes eliminated the adhesion differences between PTP-expressing and untreated control A431 cells (Fig. 4A). Pervanadate treatment also reduced the adhesion of control and PTP (D1 C433A)-expressing cells, hinting at a general role for PTPs in the maintenance of A431 cell-substratum adhesion (Fig. 4, A and B). Indeed, extended incubation of all A431 lines with vanadate resulted in the complete removal of cells from the substratum. The observed changes in A431 cell adhesion did not reflect clonal variability, as all PTP-expressing lines, obtained either as isolated clones or as bulk selected populations, exhibited increased cell-substratum adhesion as compared with vector alone control A431 cell lines.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   A431 cells overexpressing PTP exhibit increased cell-substratum adhesion. A, the adhesion of vector alone control (Ctr) and PTP-expressing cells (clone PTP-1) was assessed by measurement of cell-substratum adhesion as outlined under "Experimental Procedures." The percentage of cells remaining bound to the substratum after 3-5 washes (10 min each) in PBS, PBS + EGF, PBS + pervanadate, or PBS + EDTA was compared with untreated wells and plotted as adherent cells as percentage of control. B, the adhesion of control vector alone-transfected A431 cells was compared with cells expressing catalytically inactive PTP (D1 C433A). Cells were treated as in A except that the cells were washed three times (5 min each) in order to discriminate between the lower adhesive properties of vector alone control and the PTP (D1 C433A)-expressing A431 cells. Cells were plated in 96-well plates in quadruplicate, columns represent the mean, and bars represent the S.E. for each set of measurements. Similar results were obtained in three independent experiments.

The effect of EGF stimulation on A431 cell-substratum adhesion was also evaluated using this assay. As anticipated and reflected by the cell rounding and detachment depicted in Fig. 2B, C, EGF stimulation profoundly reduced cell-substratum adhesion of control A431 cells in these adhesion assays. However, whereas almost 100% of vector control A431 cells were released from the substratum by exposure to EGF in serum-free conditions (e.g. Ctr (untreated) A₅₅₀ ~ 2.3, Ctr (EGF-treated) A₅₅₀ ~ 0.004), approximately 40% of PTP-transfected cells remained adherent and spread (e.g. PTP-1 (untreated) A₅₅₀ ~ 2.2, PTP-1 (EGF-treated) A₅₅₀ ~ 0.8) (Figs. 2, B and C, and 4A). In contrast, PTP (D1 C433A)-expressing cells appeared more sensitive to the adhesion-disrupting actions of EGF than were vector control-transfected cells (e.g. Ctr (untreated) A₅₅₀ ~ 2.5, Ctr (EGF-treated) A₅₅₀ ~ 1.1, PTP-D1 C433A (untreated) A₅₅₀ ~ 2.4, PTP-D1 C433A (EGF-treated) A₅₅₀ ~ 0.4) (Fig. 4B).

The increase in adhesion may have been due to altered expression of ECM proteins by PTP-overexpressing cells. However, the transfer of control A431 cells to plates on which PTP-expressing cells had been previously grown did not alter the adherence characteristics of the transferred cells (data not shown). This suggested that the augmented adhesion of PTP-expressing cells was an intrinsic property and not simply a result of alterations in ECM composition.

Potential Substrates of PTP-- To investigate the basis for both the resistance of PTP-expressing cells to the EGF induced cell rounding and lift-off, and the altered cell adhesion of unstimulated cells, total cell lysates were evaluated for changes in anti-phosphotyrosine antibody (4G10) immunoreactivity. Anti-phosphotyrosine immunoblots of total cell lysates from unstimulated control vector alone-, PTP-, or PTP (D1 C433A)-transfected cells revealed reduced phosphotyrosine levels in proteins with molecular masses of 50-65 kDa only in the PTP-expressing cells (Fig. 5A). In contrast, other proteins with molecular masses of approximately 70 and 120-130 kDa exhibited enhanced anti-phosphotyrosine immunoreactivity in these cells. As previously discussed and shown in Fig. 4A, differences in the adhesion of control and PTP-expressing cells were readily apparent in unstimulated cells. Moreover, because EGF stimulation led to dramatic increases in whole scale tyrosine phosphorylation that were indistinguishable between cell lines (data not shown), we focused on the identification of potential substrates of PTP in unstimulated cells. In addition, we did not observe any alteration in the basal level of EGF receptor tyrosine phosphorylation in cells expressing PTP (data not shown).

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Expression of PTP results in the dephosphorylation of specific Src family kinases. A, total cell lysate from vector alone control-transfected (Ctr), PTP-expressing, and PTP (D1 C433A)-expressing A431 cells was separated by SDS-PAGE, transferred to membrane, and immunoblotted with anti-phosphotyrosine antibodies. Arrows indicate proteins containing diminished anti-phosphotyrosine immunoreactivity. Src family kinases were immunoprecipitated with antibodies specific for Src, Yes, Fyn, or Lyn from Ctr (lane 1), PTP (lane 2), and PTP (D1 C433A) (lane 3) A431 cell lysates, and immunoblotted with anti-phosphotyrosine (B) or kinase-specific antiserum (C). Shown are representative results from single clones corresponding to Ctr, PTP-expressing, and PTP (D1 C433A)-expressing A431 cell lines. All PTP-expressing clones examined exhibited a similar PTP D1-dependent reduction in phosphotyrosine levels of proteins in the 50-65-kDa range.

It was previously reported that expression of PTP in Fischer rat embryo fibroblasts (41) and P19 cells (42) was associated with pp60^c-Src kinase activation. In rat embryo fibroblasts, PTP expression was associated with cell transformation, whereas PTP expression in P19 cells shifted the differentiation of these cells toward a neuronal cell phenotype (41, 42). These studies did not specify whether PTP expression had resulted in any general changes in cellular protein phosphotyrosine content or if only Src-specific alterations were observed. To determine whether the dephosphorylated 50-65-kDa proteins in A431 cell lysates corresponded to Src kinase(s), we immunoprecipitated Src, Yes, Fyn, and Lyn and assessed their phosphotyrosine content. Reduced phosphotyrosine levels were observed for Src, Yes, and Fyn, whereas Lyn phosphotyrosine levels were unchanged (Fig. 5B). Interestingly, dephosphorylation of Src and Yes was observed only in cells expressing wild type PTP, whereas Fyn phosphotyrosine content appeared to be reduced in PTP (D1 C433A)-expressing cells. However, in contrast to the other Src family kinases, Fyn protein levels were strikingly reduced in cell lines overexpressing either PTP or PTP (D1 C433A) (Fig. 5C). Thus, the reduction in Fyn anti-phosphotyrosine immunoreactivity is likely a result of reduced Fyn protein levels. The reduction in Fyn protein did not appear to be due to changes in the solubility of this protein, and was typical of all A431 clones expressing PTP (data not shown).

PTP Expression Results in Src Kinase Activation-- Src family kinases are both positively and negatively regulated by tyrosine phosphorylation (43, 44). C-terminal phosphorylation leads to the inhibition of enzyme activity through a proposed intramolecular association of the N-terminal SH2 and SH3 domains with the C terminus (45-51). In contrast, tyrosine phosphorylation within the catalytic domain increases enzyme activity (52). To determine the effect of PTP expression on Src kinase activity, we immunoprecipitated Src from two other A431 clones expressing PTP. These two clones contained equivalent levels of PTP protein (Fig. 6A) and possessed similar levels of PTP specific activity against a phosphopeptide based on Src Tyr⁵²⁷ protein sequence (Fig. 6B). In addition, Src kinase immunoprecipitated from these cells contained lower phosphotyrosine levels than Src immunoprecipitated from control A431 cells (Fig. 7, A and B). The relative activity of Src was determined by autokinase and enolase assays (Fig. 7, C and D). Src kinase isolated from these PTP-expressing cells was approximately 3-fold more active than control A431 derived Src kinase (as determined by scintillation counting of enolase excised from bands shown in Fig. 7D), suggesting that PTP induced the activation of Src kinase by reducing the phosphorylation of Src Tyr⁵²⁷. The diminution of overall Src kinase phosphotyrosine levels, however, suggested either that PTP dephosphorylation of Src was not accompanied by Tyr⁴¹⁶ phosphorylation or that PTP expression also resulted in Tyr⁴¹⁶ dephosphorylation. The latter interpretation is supported by the results of Zheng et al. (41) and den Hertog et al. (42), who observed PTP-dependent Src Tyr⁵²⁷ and Tyr⁴¹⁶ dephosphorylation.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Analysis of PTP-overexpressing cells with regard to PTP expression and enzyme activity. A, immunoblot analysis of lysates derived from vector alone control (Ctr) and two A431 clones expressing similar levels of PTP (PTP 1 and PTP 2). B, PTP was immunoprecipitated from A431 cell lysates obtained from these cells using anti-peptide antibodies against residues 512-558 of the intercatalytic region of PTP and PTP activity was assessed by measuring the dephosphorylation of Src Tyr527 phosphopeptide using the MGMP assay. Pi release was determined by absorbance measurement at 620 nm at the indicated times.

View larger version (33K): [in this window] [in a new window]   Fig. 7.   PTP expression results in Src kinase dephosphorylation and activation. Src kinase was immunoprecipitated (mAb-327) from lysates derived from vector alone control (Ctr) and two A431 clones expressing similar levels of PTP (PTP 1 and PTP 2) and immunoblotted with anti-phosphotyrosine (A) or anti-Src specific antisera (GD11) (B). The relative activity of Src kinase obtained from either Ctr or PTP-expressing cell lysates was determined by autokinase (C) or enolase (D) assays.

Recombinant PTP Phosphatase Activity-- The ability of the various domains of PTP to dephosphorylate Src Tyr⁵²⁷ was also investigated in vitro using recombinant PTP and synthetic phosphopeptide substrates. The entire cytoplasmic region of PTP (PTP-D1+D2), the membrane-proximal catalytic domain (PTP-D1), and the second catalytic domain alone (PTP-D2) were expressed in bacteria as GST fusion proteins. Recombinant proteins were incubated with phosphopeptides corresponding to the C-terminal sequence of Src (⁵²⁷YTSTEPQpYQPGENL) or peptides corresponding to sequences surrounding an autophosphorylation site in the CSF-1 receptor (⁷⁰⁸YIHLEKKpYVRRDSG). Phosphate release was then determined using the MGMP assay (38). PTP-D1+D2 and PTP-D1 catalyzed the dephosphorylation of the two phosphopeptide substrates, with both enzymes showing a preference for the Src Tyr⁵²⁷ peptide (Fig. 8, A and B). Relative K_m values for the individual recombinant PTP proteins were PTP D1 + D2, Src Tyr⁵²⁷, 70 µM; CSF1r Tyr⁷⁰⁸, 197 µM; PTP D1, Src Tyr⁵²⁷, 46 µM; CSF1r Tyr⁷⁰⁸, 184 µM. The second catalytic domain of PTP was also active, as demonstrated by its ability to dephosphorylate p-nitrophenylphosphate and the Src Tyr⁵²⁷ phosphopeptide (Fig. 8C). Thrombin-cleaved preparations from GST-alone lysates exhibited no detectable PTP activity in any of the PTP assays (data not shown). PTP-D2, like PTP-D1+D2 and PTP-D1, was more active toward the Src Tyr⁵²⁷ phosphopeptide than the CSF-1 receptor Tyr⁷⁰⁸ peptide. However, detection of PTP-D2 activity required approximately 100-fold more enzyme (approx. 1 µg versus 10 ng) than those assays with fusion proteins containing domain-1. In keeping with the low activity of recombinant PTP-D2 toward Src Tyr⁵²⁷, immunoprecipitates of PTP from A431 cell lysates derived from A431 cells expressing the PTP C433A D1 mutant exhibited dramatically reduced activity as compared with immunoprecipitates of wt. PTP (Fig. 8D). Thus, whereas recombinant PTP-D2 displayed low but detectable activity in vitro, no domain-2-specific activity was detected in the context of an inactive domain-1. However, this may have been a consequence of insufficient protein being obtained in the immunoprecipitation assay as compared with the recombinant PTP assay. The lack of detectable activity against the Src Tyr⁵²⁷ phosphopeptide in PTP (D1 C433A) immunoprecipitates and the PTP D1-dependent reduction in both Src and Yes phosphotyrosine levels in transfected A431 cells (Fig. 5B) suggested that the first catalytic domain of PTP was sufficient for Src and Yes kinase dephosphorylation. Similarly, den Hertog et al. (42) demonstrated that bacterially expressed PTP (D1 C433A) lost the ability to dephosphorylate both in vivo ³²P-labeled Src (tyrosine-phosphorylated predominantly at Tyr⁵²⁷), and Src phosphorylated in vitro at Tyr⁴¹⁶.

View larger version (29K): [in this window] [in a new window]   Fig. 8.   PTP bacterial expression and enzymatic activity. The entire cytoplasmic region of PTP (PTP-D1+D2), the membrane-proximal catalytic domain (PTP-D1), and the second catalytic domain alone (PTP-D2) were expressed as GST fusion proteins, purified with glutathione-Sepharose and thrombin cleaved for in vitro analysis of enzyme activity. Enzymes corresponding to PTP-D1 + D2 (A), PTP-D1 (B), and PTP-D2 (C) were assayed for activity against p-nitrophenylphosphate () (for PTP-D2) or Src Tyr527 () and CSF-1 Tyr708 () phosphopeptides. Substrate dephosphorylation was assessed by measurement of liberated phosphate using the malachite green microtiter plate assay. Pi was detected by spectrophotometer measurement at 620 nm and compared with a Pi/malachite green standard curve. Control lysates expressing GST alone displayed no detectable PTP activity (data not shown). In D, PTP was immunoprecipitated from cells transfected with either wt. PTP or PTP (D1-C433A) using anti-peptide antibodies against residues 512-558 of the intercatalytic region (anti-PTP-1). Dephosphorylation of Src Tyr527 phosphopeptide by immunoprecipitated PTP was then assessed as above.

Dephosphorylation and activation of Src correlates with changes in cellular localization and association with other proteins by freeing both the SH2 and SH3 domains to bind specific targets. Thus, Src activated by dephosphorylation of Tyr⁵²⁷ would be predicted to exhibit enhanced SH2 domain-mediated binding to phosphoproteins or phosphopeptides (50, 53). Consistent with this model, we detected a marked increase in the amount of Src kinase precipitated from lysates derived from PTP-expressing cells by beads containing immobilized Src Tyr⁵²⁷ phosphopeptide (Fig. 9, middle panel). Beads containing immobilized non-phosphorylated Src Tyr⁵²⁷ peptide did not retain Src kinase from any of the cell lysates (Fig. 9, right panel). Interestingly, SRC2 immunoblots of phosphopeptide immunoprecipitates revealed a band of approximately 62 kDa retained by Src Tyr⁵²⁷ phosphopeptide-Sepharose (Fig. 9, left panel). Although an antibody able to specifically immunoblot Yes was not available, the SRC2 antibody immunoblot, and the reduction in phosphotyrosine content of Yes in anti-Yes immunoprecipitates (Fig. 5B) suggested that PTP expression was accompanied by dephosphorylation of the C-terminal regulatory site of this kinase. The enhanced binding of Src kinases to Src Tyr⁵²⁷ peptide beads, like the reduction in Src and Yes phosphotyrosine levels observed in Fig. 5, A and B, required PTP D1, as Src Tyr⁵²⁷ peptide beads incubated with lysates derived from PTP (D1 C433A)-expressing cells retained similar levels of Src and Yes protein as control A431 cell derived lysates (Fig. 9, left and middle panels, lane 3).

View larger version (27K): [in this window] [in a new window]   Fig. 9.   Src Tyr527 phosphopeptide binding assay. Lysates from control vector alone (lane 1), PTP-expressing (lane 2), and PTP (D1 C433A)-expressing (lane 3) A431 cells were incubated with Src Tyr527 phosphopeptide conjugated CNBr-Sepharose (panels SRC2 and anti-src). Bead-precipitated enzyme was then detected with anti-Src antiserum (panel anti-src) or SRC2 antiserum, which recognize Src, Yes, and Fyn (panel SRC2). Beads conjugated to nonphosphorylated Src Tyr527 peptide were used as a control in this assay (panel SRC2). Arrows indicate the position of Src (in panels SRC2 and anti-src) and a band which likely corresponds to Yes kinase that was detected with the SRC2 antisera (panel SRC2). The sequence of the 13-residue Src Tyr527 peptide used for precipitations is shown at bottom.

PTP Co-immunoprecipitates with Src Kinase(s)-- To further investigate the relationship between PTP and Src kinases, we assessed whether these enzymes were associated within cells using co-precipitation studies. SRC2 antibody immunoprecipitates of Src, Yes, and Fyn from A431 cells expressing either PTP, or PTP (D1 C433A), were able to co-immunoprecipitate PTP. Although Fig. 10 (right lanes) suggests reduced levels of PTP co-immunoprecipitated with Src kinases in cells expressing PTP (D1 C433A), this was not consistently observed. We also investigated whether PTP co-immunoprecipitated with Csk, as PTP Y579 and Y789 were phosphorylated by this kinase in vitro.² However, we were unable to co-immunoprecipitate PTP and Csk from A431 cells (Fig. 10, right panel). Antibodies directed against FAK, paxillin, or the -1 integrin also failed to co-immunoprecipitate PTP in these cells (data not shown). The stoichiometry of association between Src kinase(s) and PTP may be low as detection of co-precipitating PTP required immunoblotting with anti-recombinant PTP antibodies (anti-PTP-2) and was not detected with anti-PTP peptide antiserum (anti-PTP-1) (data not shown). Moreover, we have been unable to detect PTP in SRC2 immunoprecipitates from transiently transfected HEK 293 cells overexpressing PTP. However, specific antibodies against Src (GD11) and Yes were also able to co-immunoprecipitate PTP from A431 cells (data not shown). These antibodies, however, were less efficient at co-immunoprecipitating PTP than the SRC2 antiserum.

View larger version (40K): [in this window] [in a new window]   Fig. 10.   PTP co-immunoprecipitates with Src family kinase(s) from A431 cells. Approximately 2 mg of total cell lysate from control vector alone (Ctr), PTP-expressing, and PTP (D1 C433A)-expressing A431 cells was subjected to immunoprecipitation with either rabbit polyclonal antibodies against Src, Yes, and Fyn (SRC2) (left panel) or rabbit polyclonal antibodies directed against Csk (right panel). Approximately 2 µg of SRC2 antisera and 5 µg of anti-Csk antiserum were used in each immunoprecipitation. Immunoprecipitations were then immunoblotted with antisera raised against the entire cytoplasmic region of PTP (residues 167-793, anti-PTP-2). The positions of PTP and Ig heavy chain are indicated by arrows. The center panel shows the mobility of PTP in total cell lysates derived from PTP-overexpressing A431 cells. The left panel was exposed to film for approximately 1 min, and the right panel was exposed for 5 min.

To look for sites in PTP that might mediate an association with Src, we investigated a membrane-proximal proline-rich motif (STNRKYPPLPVDKLE) in PTP for its capacity to associate with various SH3 domains. Although this region bound to the SH3 domains of Src, Yes, Fyn, Lyn, and other SH3 domains in vitro, mutation of the two Pro residues in this sequence to Ala failed to prevent PTP from co-precipitating with Src kinases.³ Thus, the association between PTP and Src kinase(s) does not appear to be mediated exclusively via Src family SH3 domains. The structural basis for the observed PTP-Src association, which could be either direct or mediated via a linker molecule, requires further investigation.

Protein Phosphotyrosine Levels in PTP-expressing Cells-- An interesting consequence of PTP expression in A431 cells was the selective reduction of Src family kinase phosphotyrosine levels. Indeed, arguing for the specificity of PTP for Src and Yes kinases, proteins of 70 and 110-130 kDa exhibited an increase in anti-phosphotyrosine immunoreactivity in PTP-expressing cells (Fig. 11A). Previous studies have demonstrated that v-Src, Tyr⁵²⁷F mutants of c-Src, and Src kinases activated within Csk/ mouse embryo fibroblasts induce the tyrosine phosphorylation of specific proteins (32, 43, 44). Many of these Src family kinase substrates are localized and/or involved in the organization of the cytoskeleton. The changes in cell adhesion of PTP-expressing A431 cells therefore suggested this phosphatase might be involved in regulating Src kinase activity at focal adhesions.

View larger version (47K): [in this window] [in a new window]   Fig. 11.   Expression of PTP in A431 cells leads to increased FAK tyrosine phosphorylation and results in enhanced Src/FAK association. Lysates from vector alone control (lane 1), PTP-expressing (lane 2), and PTP (D1 C433A)-expressing (lane 3) A431 cells were separated by SDS-PAGE and immunoblotted with anti-phosphotyrosine antibodies (A), subjected to immunoprecipitation with SRC2 antisera and immunoblotted with anti-phosphotyrosine antibodies (B) or anti-FAK antisera (C), or used to immunoprecipitate FAK (D and E). Anti-FAK immunoprecipitates were then immunoblotted with either anti-phosphotyrosine antibodies (D) or anti-FAK antisera (E). Similar PTP D1-dependent increases in FAK tyrosine phosphorylation and Src kinase association were observed in all three PTP-expressing A431 clones analyzed.

Increased Src Kinase Association with FAK in PTP-expressing Cells-- To establish whether the proteins displaying increased anti-phosphotyrosine immunoreactivity in PTP-expressing A431 cells were associated with Src kinase(s), Src, Yes, and Fyn were immunoprecipitated with SRC2 antiserum. Immunoblotting of these immunoprecipitates with anti-phosphotyrosine antibodies revealed proteins of 110-130 and approximately 220 kDa co-immunoprecipitating with Src kinase(s) (Fig. 11B). The activation of Src enhances the phosphorylation and activation of FAK, a 125-kDa kinase localized at focal adhesions and activated in response to integrin-dependent cell adhesion (32, 54-57). Others have reported the SH2 domain-dependent association of Src family kinases with FAK (58, 59). To determine whether the 125-kDa phosphoprotein co-immunoprecipitating with Src kinase(s) was FAK, SRC2 immunoprecipitates were immunoblotted with anti-FAK antibodies. Increased quantities of FAK were observed in SRC2 immunoprecipitates from PTP-overexpressing cells (Fig. 11C). In keeping with an enhanced association of Src kinase(s) with FAK, we observed a small but consistent increase (1.5-2-fold) in the level of FAK tyrosine phosphorylation in all A431 clones expressing PTP (Fig. 11D). In contrast, in cells expressing PTP (D1 C433A), FAK phosphotyrosine levels were approximately 50% lower than control A431 cell-derived FAK as judged by densitometry (Fig. 11D and data not shown). FAK protein levels were not altered by expression of either PTP or PTP (D1 C433A) (Fig. 11E). Immunoblots of SRC2 immunoprecipitates also detected elevated levels of both p130 Cas and p120 Cbl co-immunoprecipitating with Src kinases in cells expressing PTP (data not shown). Cas has been previously identified as being associated with Src kinases and is a target of Src phosphorylation (60-62). Moreover, p130 Cas is tyrosine phosphorylated in response to cell-substratum adhesion and associates with FAK (63, 64). We have not determined the identity of the ~ 200-kDa phosphoprotein co-immunoprecipitating with Src kinase(s) (Fig. 11B); however, p220 tensin is a likely candidate (65).

Enhanced Paxillin Tyrosine Phosphorylation in PTP-expressing Cells-- The protein(s) exhibiting the most dramatic increase in phosphotyrosine content in PTP-expressing cells was ~ 70 kDa, and did not co-immunoprecipitate with Src kinase(s) (Fig. 11B and data not shown). We hypothesized that this phosphoprotein(s) might include paxillin, a 68-70-kDa protein previously shown to be a substrate of Src and/or FAK (25, 26). We immunoprecipitated paxillin from control, PTP-expressing, and PTP (D1 C433A)-expressing cells and immunoblotted these immunoprecipitates with anti-phosphotyrosine antibodies. As predicted, increased levels of anti-phosphotyrosine immunoreactivity were detected within paxillin obtained from PTP-overexpressing cells (Fig. 12A). Paxillin protein levels were similar among the three cell lines shown (Fig. 12B). Densitometric analysis of paxillin immunoprecipitates revealed approximately 5-fold greater anti-phosphotyrosine immunoreactivity within paxillin obtained from PTP-expressing cells (data not shown). The increase in paxillin phosphorylation was a feature of all A431 cell lines expressing wild type PTP.

View larger version (41K): [in this window] [in a new window]   Fig. 12.   A431 PTP expression enhances paxillin tyrosine phosphorylation and Csk-paxillin association. Paxillin was immunoprecipitated from control (lane 1), PTP-expressing (lane 2), and PTP (D1 C433A)-expressing (lane 3) A431 cell lysates, separated by SDS-PAGE, and immunoblotted with anti-phosphotyrosine (A) and anti-paxillin antibodies (B). Csk was immunoprecipitated as above and immunoblotted with anti-phosphotyrosine (C) or anti-paxillin antibodies (D). All clones analyzed displayed a similar PTP D1-dependent increase in paxillin tyrosine phosphorylation.

Increased Csk/Paxillin Association within PTP-expressing Cells-- Src kinase-dependent tyrosine phosphorylation of paxillin is thought to result in the recruitment of Csk to phosphorylated paxillin via an interaction between the Csk SH2 domain and phosphotyrosine residues in paxillin (29, 31). Sabe et al. (29, 33) demonstrated that v-Crk express