The Src and signal transducers and activators of transcription pathways as specific targets for low molecular weight phosphotyrosine-protein phosphatase in platelet-derived growth factor signaling.

The low molecular weight phosphotyrosine-protein phosphatase (LMW-PTP) is a cytosolic phosphotyrosine-protein phosphatase specifically interacting with the activated platelet-derived growth factor (PDGF) receptor through its active site. Overexpression of the LMW-PTP results in modulation of PDGF-dependent mitogenesis. In this study we investigated the effects of this tyrosine phosphatase on the signaling pathways relevant for PDGF-dependent DNA synthesis. NIH 3T3 cells were stably transfected with active or dominant negative LMW-PTP. The effects of LMW-PTP were essentially restricted to the G1 phase of the cell cycle. Upon stimulation with PDGF, cells transfected with the dominant negative LMW-PTP showed an increased activation of Src, whereas the active LMW-PTP induced a reduced activation of this proto-oncogene. We observe that c-Src binding to PDGF receptor upon stimulation is prevented by overexpression of LMW-PTP. These effects were associated with parallel changes in myc expression. Moreover, wild-type and dominant negative LMW-PTP differentially regulated STAT1 and STAT3 activation and tyrosine phosphorylation, whereas they did not modify extracellular signal-regulated kinase activity. However, these modifications were associated with changes in fos expression despite the lack of any effect on extracellular signal-regulated kinase activation. Other independent pathways involved in PDGF-induced mitogenesis, such as phosphatidylinositol 3-kinase and phospholipase C-gamma1, were not affected by LMW-PTP. These data indicate that this phosphatase selectively interferes with the Src and the STATs pathways in PDGF downstream signaling. The resulting changes in myc and fos proto-oncogene expression are likely to mediate the modifications observed in the G1 phase of the cell cycle.

Protein-tyrosine phosphorylation plays a key role in the regulation of many cellular processes in eukaryotes such as cellular metabolism, proliferation, differentiation, and oncogenic trans-formation (1). Accumulating evidence indicates that the contribution of phosphotyrosine-protein phosphatases (PTPs) 1 to the control of phosphorylation state is as relevant as that of phosphotyrosine-protein kinases. PTPs activity is carefully regulated and appears to be, in most cases, highly specific. The PTPs consist of a family of over 40 enzymes often classified into three groups: 1) receptor-like PTPs; 2) intracellular PTPs; and 3) dual specificity PTPs (2). All PTPs share the signature active site motif CXXXXXR, in which the catalytic cysteine residue is involved in formation of a phosphoenzyme reaction intermediate (3). The low molecular weight phosphotyrosine-protein phosphatase (LMW-PTP) is a cytosolic enzyme without extensive sequence homology with other PTPs family members, but it contains the CXXXXXR motif and shares the same catalytic mechanism of classical PTPs (4,5). Furthermore, the LMW-PTP crystal structure (6) revealed a tridimensionally folded phosphate binding loop that is structurally identical to that contained in the human placenta PTP1B and Yersinia PTP (7,8).
We have previously demonstrated that the mutation of the cysteine residue in the signature motif to serine (C12S), causes the complete loss of catalytic activity (9). Nevertheless, this dominant negative mutant (dnLMW-PTP) is still able to bind to specific substrates (10). Overexpression of the active phosphatase (wtLMW-PTP) causes a reduction of cell proliferation (11), whereas dnLMW-PTP induces a remarkable increase in DNA synthesis, indicating that this latter molecule behaves as a dominant negative in vivo (12). In addition we have demonstrated that this phenotype is associated with a specific and direct interaction between the active site of dnLMW-PTP and the activated platelet-derived growth factor receptor (PDGF-R) (13). These data indicate that the PDGF-R is a specific substrate of the LMW-PTP and that this phosphatase may be involved in the control of one or more signaling pathways triggered by PDGF-R activation.
Binding of PDGF induces receptor dimerization and autophosphorylation. Tyrosine phosphorylation sites in the PDGF-R function as high affinity binding sites for several molecules involved in downstream signal transduction (14). These proteins bind the phosphorylated receptor through their Src homology domain 2 or phosphotyrosine binding domains, * This work was supported by the Italian Association for Cancer Research (AIRC) and in part by the Ministero della Università e Ricerca Scientifica e Tecnologica (MURST) and Consiglio Nazionale delle Ricerche (CNR). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ Supported by grants from MURST (Liver cirrhosis and viral hepatitis project) and from the Italian Liver Foundation.
ʈ To whom correspondence should be addressed. Tel.: 39-55-413765; Fax: 39-55-4222725; E-mail: raugei@cesit1.unifi.it. and in many cases are themselves substrates of the receptor kinase activity. These proteins include enzymes such as phospholipase C-␥1 (PLC␥1), phosphatidylinositol 3-kinase (PI3K), Src, or molecular adapters such as Shc, Grb2, and Nck, leading to different routes that mediate the biological effects of PDGF. One of the main routes leading to cell proliferation is the so-called Ras/extracellular signal-regulated kinase (ERK) pathway. Upon stimulation with growth factors, Ras activation can be achieved through a variety of signal transducers like Shc, Grb2, and Sos (15). Alternatively, Syp phosphatase may function as an adapter molecule for Grb2 and therefore activate Ras (16). Activated Ras triggers a kinase cascade with sequential activation of Raf-1, MAP/ERK, and ERK. Upon activation, ERK translocates to the nucleus where it induces fos transcriptional activation (17,18). An alternative route leading to fos transcription independent of Ras/ERK, is the activation of the signal transducers and activators of transcription (STAT) pathway, leading to formation of three transcription factors that bind to the fos promoter. These transcription factors, known as sis-inducible factor (SIF) A, B, and C, are STAT-3 homodimers, STAT1/3 heterodimers, and STAT1 homodimers, respectively. In addition, another Ras-independent pathway relevant for mitogenesis is the activation of the cytosolic tyrosine kinase Src. Recently, Barone and Courtneidge (19) have demonstrated that the kinases of the Src family regulate myc activation, and that myc is necessary for PDGF-dependent DNA synthesis in NIH 3T3 cells. In fact, myc but not fos rescue the PDGFsignaling block caused by dominant negative Src. Both the Ras/ERK and the Src pathways are necessary for DNA synthesis in response to PDGF, epidermal growth factor, and colonystimulating factor-1 (19). In addition, the Src and STAT pathways seems to be linked (20,21), although it remains to be established if this connection is directly mediated by Src kinase. Finally, other pathways relevant for PDGF mitogenic signaling are PLC␥1 and PI3K. PLC␥1 catalyzes the breakdown of phosphatidylinositol bisphosphate, triggering a cytosolic calcium increase and protein kinase C activation. PI3K is a protein and lipid kinase that activates the product of the proto-oncogene akt, protein kinase C, and other targets, which have been only partially elucidated.
In this study, we examine the role of the LMW-PTP in signaling pathways originated by PDGF-R activation. We report that LMW-PTP is involved in the regulation of Myc expression through Src activation, and of Fos through an ERK independent pathway mediated by the STAT proteins.

EXPERIMENTAL PROCEDURES
Materials-Unless specified all reagents were obtained from Sigma. NIH 3T3 cells were purchased from ATCC; human recombinant platelet-derived growth factor BB (PDGF-BB) was from Peprotech; RC20 anti-phosphotyrosine antibodies were from Affiniti; enhanced chemiluminescence kit was from Amersham; c-fos murine cDNA was from ATCC; human c-myc cDNA was a generous gift from G. I. Evans; and Src kinase assay kit, anti-Src antibodies, and PDGF receptor antibodies were from Upstate Biotechnology, Inc. Agarose-conjugated phosphotyrosine antibodies were purchased from Oncogene Science. Anti-ERK1, anti-phosphotyrosine, anti-PDGF receptor, and anti-PLC␥1 antibodies were from Santa Cruz Biotechnology, and anti-STAT1 and anti-STAT3 antibodies were from Transduction Laboratories.
Cell Culture and Transfections-NIH 3T3 cells were routinely cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum in 5% CO 2 humidified atmosphere. 10 g of pSVT7PTPC12S (12) or pSVT7PTP (22) and 0.5 g of pSV2neo, conferring neomycin resistance, were cotransfected in NIH 3T3 cells using the calcium phosphate method. Stable transfected clonal cell lines were isolated by selection with G418 (400 g/ml). Control cell lines were obtained by transfecting 2 g of pSV2neo alone. The clonal lines were screened for expression of the transfected genes by a) Northern blot analysis and b) enzyme-linked immunosorbent assay using polyclonal anti-LMW-PTP rabbit antibodies, which do not crossreact with murine endogenous LMW-PTP.
Immunoprecipitations and Western Blot Analysis-1 ϫ 10 6 cells were seeded in 10-cm plates in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. Cells were serum-starved for 24 h before receiving PDGF-BB. Cells were then lysed for 20 min on ice in 500 l of modified RIPA lysis buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Nonidet P-40, 0.25% sodium deoxycholate, 2 mM EGTA, 1 mM sodium orthovanadate, 1 mM phenylmethanesulfonyl fluoride, 10 g/ml aprotinin, 10 g/ml leupeptin). Lysates were clarified by centrifugation and immunoprecipitated for 4 h at 4°C with 5 g of the specific antibodies. Immune complexes were collected on protein A-Sepharose (Pharmacia Biotech Inc.), separated by SDS-polyacrylamide gel electrophoresis, and transferred onto nitrocellulose (Sartorius). Immunoblots were incubated in 3% bovine serum albumin, 10 mM Tris-HCl, pH 7.5, 1 mM EDTA, and 0.1% Tween-20, for 1 h at room temperature and then probed first with specific antibodies and then with secondary antibodies conjugated with horseradish peroxidase, washed and developed with the enhanced chemiluminescence kit (Amersham).
Cell Cycle Analysis-Growth data were obtained plating 2 ϫ 10 5 cells for 48 h. Cell number was estimated by counting in a Burker chamber. Cell doubling times (cdt) were calculated as: cdt ϭ 48/(b Ϫ a), where a is log 2 of the cell number at time of plating and b is log 2 of cell number at 48 h of standard exponential growing conditions.
Cytofluorimetric analysis was performed according to Liu (23). Briefly, 2 ϫ 10 5 cells were seeded into 6-cm dishes the day before analysis to obtain exponentially growing cultures. Cells were rinsed twice with cold phosphate-buffered saline and harvested after trypsinization in 1 ml of 50 mg/liter of propidium iodide. Analysis was performed in a Becton Dickinson FACScan using the Lysis II and Cell Fit cell analysis software according to the manufacturer's procedure.
mRNA Analysis-1 ϫ 10 7 cells were serum-starved for 24 h and than stimulated with 30 ng/ml of PDGF. Total RNA was purified according to a method previously described (24). 10 g of total RNA were separated onto a denaturing 15% formaldehyde-agarose gel and blotted onto nitrocellulose filters using standard methods. Hybridization was performed overnight in 4ϫ standard saline citrate, 0.1% SDS, 5ϫ Denhart's, 1 mM EDTA, 20 mM sodium phosphate, pH 7.2, at 65°C using a random priming labeled cDNA with specific activity of about 10 9 cpm/ g. Washings were performed twice in 0.5ϫ standard saline citrate, 0.1% SDS at 55°C; the filter was then autoradiographed and the signals quantitated with a densitometer. Normalization was done on the basis of a human actin cDNA probe hybridization.
ERK Activity Assay-ERK activity was assayed by phosphorylation of microtubule-associated protein 2 (MAP-2) as described by Mihasaka (25). Briefly, 3 ϫ 10 5 cells were seeded in 6-cm dishes and serumstarved for 24 h. Cells stimulated with 5-50 ng/ml PDGF were harvested by scraping in 0.2 ml of lysis buffer (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 2 mM EGTA, 1 mM orthovanadate, 0.1 mM molybdate, 1% Triton X-100, 1 mM phenylmethanesulfonyl fluoride, 10 g/ml aprotinin, 10 g/ml leupeptin). 10 l of cell lysate (1.5 mg/ml of total protein) were incubated with 2 g of MAP-2 for 10 min at 30°C in a final volume of 25 l, containing 50 mM Tris-HCl, pH 7.4, 2 mM EGTA, 10 mM MgCl 2 , 40 M [␥-32 P]ATP (1 Ci, 3000 Ci/mmol). The reaction was stopped by the addition of 4ϫ Laemmli SDS-polyacrylamide gel electrophoresis sample buffer. Phosphorylated MAP-2 was resolved by SDS-polyacrylamide gel electrophoresis (7.5% acrylamide gel). Coomassie Bluestained bands containing MAP-2 protein were excised from the gel, and incorporated radioactivity was measured by scintillation counting. ERK activity was also measured by an in vitro kinase assay after immunoprecipitation. 70 g of total proteins in modified RIPA lysis buffer were immunoprecipitated with anti-ERK1 polyclonal antibodies and with the use of protein A-Sepharose. After five washings with 1 ml of modified RIPA buffer, the immunobeads were incubated in 25 l of a buffer containing 15 mM Tris-HCl, pH 7.4, 15 mM MgCl 2 , 0.5 mM EGTA, 40 mM ATP, 1 Ci [␥-32 P]ATP (3000 Ci/mmol), and 10 g of a peptide derived from myelin basic protein (Santa Cruz Biotechnology). After 5 min of incubation at 30°C the reaction was stopped by adding 20 l of 40% trichloroacetic acid. 25 l of the reaction mixture were spotted onto 2 ϫ 2-cm phosphocellulose disks, washed four times in 0.75% phosphoric acid and once in acetone. Radioactivity was evaluated by scintillation counting. The amount of immunoprecipitated ERK1 from each sample was evaluated by anti-ERK1 immunoblot. Densitometric analysis of the immunoblot was used for normalization of the immunokinase assay.
Cell Motility Assay-Migration of NIH 3T3 cells expressing wtLMW-PTP or dnLMW-PTP was assayed with the Boyden chamber system for chemotaxis (Nucleopore Corp., Pleasanton, CA) equipped with 8-m pore polyvinylpyrrolidone-free polycarbonate filters (13-mm diameter) (26). Polycarbonate filters were precoated with human type I collagen (20 g/ml in phosphate-buffered saline, pH 7.4) for 30 min at 37°C and placed between the chemoattractant (lower chamber) and the upper chamber. The lower chamber was filled with medium supplemented with different concentrations of PDGF-BB. Cells cultured in serum-free Dulbecco's modified Eagle's medium were suspended by trypsinization, and 3 ϫ 10 4 cells in 200 l was added to the top wells and incubated at 37°C in 5% CO 2 for 6 h. After incubation the cells that had attached to the upper side but had not migrated through the filter were mechanically removed using cotton swabs. The filters were fixed in 96% methanol and stained with Harris hematoxylin solution. Chemotaxis was quantitated by counting the cells that had migrated to the lower surface of the polycarbonate filters. For each filter the number of cells in six randomly chosen fields was determined, and the counts were averaged (mean Ϯ S.D.).
Phosphatidylinositol 3-Kinase Activity-PI3K assay was performed as described elsewhere (27,28). Briefly, serum-starved cells were incubated with 30 ng/ml of PDGF for 10 min and then lysed in RIPA buffer. Equal amounts of protein were immunoprecipitated using an agaroseconjugated anti-phosphotyrosine antibody. After washing, the immunobeads were resuspended in 50 l of 20 mM Tris-HCl, pH 7.5, 100 mM NaCl, and 0.5 mM EGTA. 0.5 l of 20 mg/ml phosphatidylinositol was added, mixed, and incubated at 25°C for 10 min. 1 l of 1 M MgCl 2 and 10 Ci of [␥-32 P]ATP (3000 Ci/mmol) were then added simultaneously and incubated at 25°C for an additional 10 min. The reaction was stopped by the addition of 150 l of chloroform, methanol, 37% HCl, 10:20:0.2. The samples were extracted with chloroform and dried. Radioactive lipids were separated by thin-layer chromatography using chloroform, methanol, 30% ammonium hydroxide, water 46:41:5:8. After drying, the plates were autoradiographed. Identity of the 3-OH phosphorylated lipids after separation by thin-layer chromatography has been previously confirmed using high-pressure liquid chromatography (28). The radioactive spots corresponding to phosphatidylinositol phosphate were scraped and counted in a liquid scintillation counter.
The samples were then separated on a native 5% polyacrylamide gel (50 mM Tris, 380 mM glycine, 10% glycerol). After electrophoresis, the gel was dried and autoradiographed. For "supershift" experiments, the cell extracts were incubated with a monoclonal anti-STAT1 or a polyclonal anti-STAT3 antibody for 30 min on ice before the binding reaction.
Src Kinase Assay-1.5 ϫ 10 6 cells were seeded in 10-cm dishes, serum-starved for 24 h and then stimulated with 50 ng/ml of PDGF for 5 min at 37°C. Cell were lysed in modified RIPA buffer (50 mM Tris-HCl, pH 7.5, 2 mM EGTA, 1 mM orthovanadate, 1% Nonidet P-40, 0.25% sodium deoxycholate, 1 mM phenylmethanesulfonyl fluoride, 10 g/ml aprotinin, 10 g/ml leupeptin). Lysates were clarified by centrifugation and assayed for protein content. 5 g of anti-Src antibodies were added to 500 g of total proteins. The reaction mixtures were incubated overnight at 4°C. Immune complexes were collected using 50 l of protein A-Sepharose resin that had been preincubated with 5 g of rabbit anti-mouse IgG. The beads were incubated in the reaction mixture for 1 h at 4°C. The beads were washed three times with 500 l of modified RIPA buffer and once with 1 ml of cold phosphate-buffered saline. The beads were then incubated for 15 min at 30°C in 40 l kinase buffer (25 mM Tris-HCl, pH 7.2, 30 mM MgCl 2 , 6 mM MnCl 2 , 0.5 mM EGTA, 0.1 mM orthovanadate, 0.5 mM dithiothreitol, 125 M [␥-32 P]ATP, corresponding to 1 Ci) containing 10 g of a Src specific peptide. The reaction was stopped by adding 20 l of 40% trichloroacetic acid. After centrifugation 25 l of supernatant were spotted onto 2 ϫ 2-cm phosphocellulose paper, that was subsequently washed in a 0.75% phosphoric acid solution and then in pure acetone. Radioactivity was quantitated in a scintillation counter. The amount of immunoprecipitated c-Src from each sample was evaluated by anti-c-Src immunoblot. Densitometric analysis of the immunoblot was used for normalization of the immunokinase assay.

RESULTS
LMW-PTP Expression-NIH 3T3 cell line was chosen to assess the role of LMW-PTP in PDGF signaling. Expression plasmids containing the coding sequence for either wtLMW-PTP and dnLMW-PTP, in which the critical active site cysteine residue had been mutated to serine (C12S), were transfected into the NIH 3T3 cell line. Clonal selection of neomycin-resistant cells resulted in the establishment of cell lines overexpressing active LMW-PTP or the dominant negative form of LMW-PTP, respectively. To verify the overexpression of LMW-PTP and to choose clones with comparable LMW-PTP expression for subsequent studies, we used polyclonal anti-LMW-PTP antibodies in an enzyme-linked immunosorbent assay analysis of whole cell lysates from each of the derivative cell lines in comparison to neomycin cells. Enzyme-linked immunosorbent assay quantification data of LMW-PTP overexpression are presented in Fig. 1A. For further analysis, we chose four NIH 3T3-transfected clones with comparable LMW-PTP expression levels (wtLMW-PTP cl.2 and cl.12 and dnLMW-PTP cl.6 and cl.14, respectively). We previously demonstrated that overexpression of wtLMW-PTP and dnLMW-PTP in NIH 3T3 cells resulted in an association between the phosphatase and the PDGF receptor, leading, in the case of wtLMW-PTP, to receptor dephosphorylation (12,13). We analyzed the PDGF receptor tyrosine phosphorylation of both cl.2 and cl.12 overexpressing wtLMW-PTP and cl.6 and cl.14 overexpressing dnLMW-PTP, respectively. PDGF receptor was immunoprecipitated from cells stimulated with 30 ng/ml PDGF-BB for 5Ј and then its tyrosine phosphorylation was analyzed by anti-phosphotyrosine immunoblotting. Fig. 1B presents the results obtained. Scanning densitometry of these data revealed that both dnLMW-PTP-overexpressing clones showed an increased tyrosine phosphorylation of PDGF receptor (lanes 2 and 3) with respect to control (lane 1). In contrast, clones overexpressing the active phosphatase (lanes 4 and 5) showed the opposite phenotype, displaying a decreased level of receptor phosphorylation. As already published, overexpression of the dnLMW-PTP in NIH 3T3 cells causes an increased mitogenic response to PDGF, whereas overexpression of the wtLMW-PTP has an opposite effect (11,12). Moreover, we observed that cell morphology, viability, and protein content were not affected by LMW-PTP overexpression. In any case, no difference among clones overexpressing either wtLMW-PTP or dnLMW-PTP was ever observed. For these reasons, clones 2 and 14 were randomly chosen for further analysis.
LMW-PTP Acts Specifically in G 1 Phase-To investigate the role of LMW-PTP in the downstream signaling events originating from PDGF-R, we analyzed the possible variations in cell cycle phases distribution in response to LMW-PTP overexpression. The cellular doubling time during exponential growth in complete medium was 20.7 Ϯ 0.4 h (mean Ϯ S.D.) for neomycin cells (expressing neomycin resistance alone), 17.6 Ϯ 0.4 and 29.9 Ϯ 0.7 h for NIH 3T3 cells overexpressing dnLMW-PTP or wtLMW-PTP, respectively. These data are in agreement with our previous results clearly showing an opposite phenotypic effect of dnLMW-PTP with respect to wtLMW-PTP. The effects on cell cycle distribution was analyzed by flow cytometry, and results are presented in Table I. dnLMW-PTP-overexpressing cells showed a shorter G 1 phase in comparison to neomycin cells, whereas the duration of the other phases remained almost identical. In contrast, cells transfected with wtLMW-PTP showed a marked increase in the duration of G 1 phase with respect to neomycin cells but also presented modifications in S and G 2 /M phases, suggesting the existence of possible additional effects of wtLMW-PTP overexpression.
To confirm the influence of LMW-PTP on G 1 phase, cell synchronization was performed by aphidicolin treatment in dnLMW-PTP-overexpressing cells and neomycin cells. Results of this experiment confirmed a shortening of G 1 phase of about 3 h in response to dominant negative overexpression (data not shown).
LMW-PTP Modulates Src Activity and Myc Expression-The cytosolic tyrosine kinase Src is recruited and activated by phosphorylated PDGF-R. To investigate the possible interaction of LMW-PTP with this pathway, we evaluated the effects of phosphatase overexpression in NIH 3T3 cells on Src activity in response to PDGF stimulation. Overexpression of the dnLMW-PTP ( Fig. 2A) greatly increased Src activity upon stimulation with PDGF in comparison to neomycin cells (about 60% increase), whereas overexpression of wtLMW-PTP resulted in a dramatic decrease of Src kinase activity (less than 30% with respect to neomycin cells). No significant change was observed in the Src activity of serum-starved unstimulated cells.
Growth factor-mediated activation of the Src pathway has recently been shown to be responsible for the activation of myc proto-oncogene (19). Therefore, we evaluated whether the observed changes induced by LMW-PTP on Src activity were paralleled by comparable changes in Myc expression. A Northern blot analysis using a human c-myc cDNA probe is shown in Fig. 2B. Using total RNA from neomycin cells (lanes 1-3), the expected increase in the specific transcript level (about 4.5-fold) was observed 1 and 3 h after stimulation with PDGF. Overexpression of the dnLMW-PTP (lanes 7-9) resulted in an increased response to PDGF both 1 and 3 h after stimulation. Conversely, overexpression of the wtLMW-PTP nearly abolished myc mRNA expression (lanes 4 -6). These data indicate that LMW-PTP controls Myc expression through the regulation of Src activity.
Furthermore we have explored the possibility that LMW-PTP, while interacting directly with PDGF receptor, could prevent the recruitment of Src kinase by the receptor upon activation. For this purpose, PDGF receptor was immunoprecipitated from cell lysates after PDGF stimulation. Anti-Src immunoblot revealed that LMW-PTP really precludes Src association with the receptor (Fig. 3). In cells overexpressing dnLMW-PTP (lanes 1 and 2) we observed a dramatic increase in Src recruitment with respect to neomycin cells (lanes 3 and  4), whereas in cells overexpressing wtLMW-PTP the association of Src with the receptor was almost completely prevented. Taken together these data indicate that LMW-PTP regulation of Src pathway is at the very beginning in the signal transduction route leading to Myc activation. It is likely that LMW-PTP interferes with the binding of Src to PDGF receptor.
LMW-PTP Affects PDGF-stimulated Chemotaxis-Recent evidences indicates that the Src tyrosine kinase directly phosphorylates the PDGF-R in Tyr-934 (31). This leads to a negative modulation of the signal transduction pathway leading to motility response and shifts the response to increased mitogenicity. Since LMW-PTP appears to modulate the Src kinase activity, dnLMW-PTP or wtLMW-PTP-overexpressing cells were analyzed for their ability to migrate against a gradient of PDGF-BB, employing the leading front assay in a modified Boyden chamber. Measurements were done also in comparison with NIH 3T3 overexpressing c-Src. Cells at a density of 1.5 ϫ 10 5 /ml were resuspended after trypsinization and centrifugation in Dulbecco's modified Eagle's medium and seeded in the upper part of the modified Boyden chamber, while medium containing PDGF-BB was added below the 150-m thick micropore filter. Compared with neomycin cells (Fig. 4), dnLMW-PTP-overexpressing cells displayed decreased PDGF-stimulated chemotaxis, whereas the wtLMW-PTP showed a clear extension in chemotactic response. As expected, c-Src-overexpressing cells showed a clear decrease in cell mobility as previously reported by Hansen (31). These data suggest that LMW-PTP influences the PDGF receptor-regulated chemotaxis. LMW-PTP probably modulates a signal transduction   1 and 2), neomycin cells (lanes 3 and 4), and wtLMW-PTP (lanes 5 and 6) were serumstarved for 24 h and then incubated with 30 ng/ml of PDGF for 2 min where indicated. Cell lysates were immunoprecipitated with polyclonal anti-PDGF-R antibodies and subjected to anti-Src immunoblot as de- dnLMW-PTP or wtLMW-PTP was also evaluated (Fig. 5B). Maximal PDGF stimulation of ERK activity was always achieved with about 5 ng/ml of the growth factor. It is also evident that the potency of PDGF on ERK activation was not influenced by overexpression of LMW-PTP either in its dominant negative or active form.
Furthermore, we also measured ERK activity in immunoprecipitates obtained with the use of anti-ERK antibodies. An immune complex kinase assay using a peptide derived from myelin basic protein as a substrate was performed (Fig. 5C). No significant difference is present in the different samples.
ERK is activated by phosphorylation on tyrosine and threo-nine residues by the dual-specificity MAPK/ERK kinase (32). We evaluated the electrophoretic mobility of ERK as an index of its phosphorylation state since phosphorylated ERK shows a reduced mobility. Also in this case, the shift induced by treatment with PDGF was identical, irrespective of LMW-PTP overexpression (data not shown). Taken together, these data indicate that the phenotypic effect of the LMW-PTP is not mediated by its interference on the Ras/ERK pathway. LMW-PTP Modulates Fos Expression via STAT1 and STAT3 Activation-Several soluble mediators activate the so called STATs pathway with differing specificity with respect to the activated kinases and the STAT protein complexes formed (33). PDGF and epidermal growth factor induce activation of STAT1 and STAT3 through tyrosine phosphorylation, which binds to the promoter of fos as homo-and heterodimers known as SIF-A, SIF-B, and SIF-C (34). We analyzed STAT1 and STAT3 tyrosine phosphorylation after PDGF stimulation for 15 min. (Fig.  6). In dnLMW-PTP-overexpressing cells both STAT1 and STAT3 showed an increased tyrosine phosphorylation with respect to mock transfected cells, whereas in wtLMW-PTPoverexpressing cells a decrease of STAT1 and STAT3 tyrosine phosphorylation was observed. In addition, we analyzed STAT1 and STAT3 activation in a gel mobility shift assay using a high affinity mutated oligonucleotide from the fos promoter (m67) according to Wagner (30). Formation of the three DNA binding complexes in response to PDGF stimulation (Fig. 7A) was markedly increased in cells expressing the dnLMW-PTP (lane 6), whereas it was dramatically reduced in those transfected with wtLMW-PTP (lane 5). As expected, no complex was formed in the absence of PDGF stimulation (lanes 1 to 3). Competition with an excess of unlabeled m67 oligonucleotide resulted in the disappearance of the three complexes (lanes 10 -12 and 16), whereas addition of an unrelated oligonucleotide (specific for NFB, lane 17) did not affect the DNA binding. The identity of the proteins involved in DNA binding was confirmed by preincubating the cell lysates with anti-STAT1 (lanes 7-9) or STAT3 antibodies (lane 14). Anti-STAT1 antibodies resulted in a further reduction ("supershift") of the electrophoretic mobility of SIF-B and SIF-C, the two complexes which contain STAT1. Anti-STAT3 antibodies caused the disappearance of the two slower migrating complexes corresponding to SIF-A and SIF-B, which contain STAT3. As a control, non-immune mouse IgG did not affect DNA binding of any of the SIF complexes.
Signals generated by PDGF at the cell membrane and transduced through the Ras/ERK or the STATs pathways converge at the level of the fos promoter, leading to the transcriptional activation of this oncogene. ERK and STATs pathways act on different regulatory elements of the fos promoter. Our results indicate that whereas LMW-PTP does not modify the Ras/ERK activation cascade, it modulates the STATs pathway and that this effect should be responsible for Fos induction. We evaluated the fos mRNA level in response to PDGF stimulation by Northern blot analysis (Fig. 7B). As expected, an increase of fos mRNA was observed in serum-starved neomycin cells as early as 5 min after the addition of the growth factor (lanes 1-3). This effect was clearly more pronounced in cells transfected with the dnLMW-PTP (lanes 7-9). On the other hand, in wtLMW-PTPoverexpressing cells this effect was observed as having a marked reduction of fos transcript accumulation (lanes 4 -6) 15 and 30 min after the mitogenic stimulation with respect to neomycin cells. According to these data, it is very likely that modulatory activity of LMW-PTP on PDGF-regulated STAT activation is sufficient to induce changes in Fos expression.
PI3K or PLC␥1 Are not Targets of the LMW-PTP-PI3K and PLC␥1 have been shown to contribute in the transduction of the mitogenic signal originating from PDGF-R activation (14). To investigate if LMW-PTP overexpression caused perturbations in these pathways, we analyzed PI3K and PLC␥1 activity. PDGF induced a dramatic up-regulation of PI3K activity (Fig. 8A), but this variation was not significantly influenced by the overexpression of either dnLMW-PTP or wtLMW-PTP (lanes 5 and 6) in comparison to neomycin cells (lane 4). Furthermore, we evaluated the amount of receptor-recruited PLC␥1. Immunoprecipitates obtained with anti-PDGF receptor antibodies were analyzed on Western blot using anti-PLC␥1 antibodies (Fig. 8B). No marked difference was present in the three samples.
According to these data, it is very likely that the effects of the LMW-PTP on mitogenesis are not mediated by PI3K and PLC␥1 activation. DISCUSSION Recent data from our laboratory have provided evidence for a role of LMW-PTP in the regulation of PDGF-dependent mitogenesis. In addition, we have reported that the PDGF receptor is a specific in vivo substrate of the LMW-PTP in NIH 3T3 cells and that this phosphatase physically associates with the activated receptor (12). However, no data were available regarding the molecular mechanisms by which the LMW-PTP can shut off the mitogenic signals initiated by PDGF. In this report, we first established that the overexpression of dnLMW-PTP induces a significant shortening of G 1 cell cycle phase whereas S and G 2 /M remain almost unchanged, indicating that the LMW-PTP acts specifically in G 1 phase. This is in agreement with the hypothesis that LMW-PTP acts on the activated PDGF receptor (11,12), which is a specific effector of G 1 progression. We have also observed that overexpression of the wtLMW-PTP markedly prolongs the doubling time although the changes are not uniquely restricted to the G 1 phase. It is possible that overexpression of the active enzyme can artifactually interact with other cellular districts and/or functions, whereas the dnLMW-PTP would cause fewer and more specific perturbations of the physiology of the cells. However, actions on phases of the cell cycle other than G 1 cannot be completely ruled out by the present data.
The main finding of this study is that in cells transfected with the LMW-PTP not all the pathways originating from the activated PDGF receptor are affected. Rather, this phosphatase appears to be highly specific in modulating the Src and the STATs pathways without interacting with other important routes such as Ras/ERK, PI3K, and PLC␥1. The effects on ERK activation are particularly striking since many extracellular and intracellular effectors may converge at this level. Along these lines, the lack of changes in ERK activity further confirms that neither PI3K nor PLC␥1 are involved in mediating the effects of the LMW-PTP. In fact, PI3K has been shown to signal through routes that are at least in part dependent on Ras (35), and activation of PLC␥1 results in the generation of signals that ultimately lead to ERK activation via Ras-dependent and -independent pathways (36,37). Overexpression of the active and inactive forms of the LMW-PTP determine a reduced or increased activation of both Src and STATs pathways. Furthermore, LMW-PTP prevents the direct association of Src with the tyrosine-phosphorylated PDGF receptor. These modifications are associated with similar changes in the nuclear targets of these signaling pathways, namely myc and fos. Therefore, not only LMW-PTP modulate post-receptor signaling, but the resulting modifications are sufficient to affect activation of specific transcription factors.
Src kinase transduces PDGF signaling in a ERK-independent manner (19). Recent results indicate that Src is the starting point of a route that leads to Myc expression, whereas it is not necessary for Ras activation. This so called "Src pathway" is in turn independent of Ras and appears to be required for PDGFinduced DNA synthesis. In fact, microinjection of a dominant negative Myc (Myc In 373) completely inhibits the mitogenic response to PDGF (19). Furthermore, microinjection of antibodies directed against all three Src kinase family proteins blocks both PDGF-induced Myc expression and entry of NIH 3T3 cells into S phase (38). Hence, these data strongly support the hypothesis that myc transcription is under the control of Src family kinases and is required for mitogenesis. Data presented herein indicate that the LMW-PTP is involved in the modulation of the Src pathway activated by the PDGF-R since Src recruitment by the phosphorylated PDGF-R is modified by LMW-PTP overexpression. Src kinase receptor association directly up-regulates its tyrosine kinase activity. It is likely that the modulation in the Src kinase activity observed in LMW-PTP-overexpressing cells is mediated by the LMW-PTP preven-FIG. 7. LMW-PTP modulates STAT activation through their tyrosine phosphorylation and c-fos expression. A, NIH 3T3 cells transfected with dnLMW-PTP or wtLMW-PTP or neomycin resistance alone (neo) were serum-starved for 24 h and then incubated with 100 ng/ml of PDGF for 15 min where indicated. 10 g of total proteins from cell lysate were used for a gel mobility shift assay using a radiolabeled oligonucleotide (m67) as described under "Experimental Procedures." In the indicated lanes, the lysates were preincubated with a 100-fold molar excess of unlabeled m67-SIE or an unrelated oligonucleotide (NFB). In addition, we incubated the lysates with monoclonal anti-STAT1 or anti-STAT3 antibodies or with an unrelated anti-mouse IgG (Ab). tion of binding of Src to the receptor. Recent data reported by our group indicate that LMW-PTP is a target of v-Src in vitro (39). However, we failed to observe a physical interaction between c-Src and either dnLMW-PTP or wtLMW-PTP in vivo. A recent paper reports that the recruitment of the Src tyrosine kinase by PDGF-R leads to decreased PDGF-stimulated chemotaxis directly mediated by PDGF-R Tyr-934 phosphorylation by Src kinase (31). Our data on the modulation of the PDGF chemotactic response by LMW-PTP are in agreement with a specific action of LMW-PTP on the Src-mediated pathways leading to increased mitogenicity through Myc activation and to decreased PDGF-stimulated chemotaxis.
Fos activation appears to be strongly implicated in PDGF mitogenic signaling (17,18). Activation of fos transcription by PDGF is mediated by the serum response element (SRE) and the sis-inducible element (SIE) of the promoter. The SRE interacts with the ternary complex factor, whose components Elk-1 and SAP-1 are direct targets of ERK (18). Conversely, SIE interacts with proteins of the STAT family, STAT1 and STAT3, which form three transcription factors, namely SIF-A, -B, or -C (30). We find that the LMW-PTP modulates Fos expression, which is greater in cells transfected with the dnLMW-PTP and reduced in those expressing wtLMW-PTP. These changes are associated with a modulation of both tyrosine phosphorylation and DNA binding activity of STAT1 and STAT3. On the other hand, ERK activity, measured by three independent methods, is unaffected by the overexpression of the LMW-PTP. The SRE and the SIE are both able to drive fos transcription in studies using deletion mutants of the promoter and a reporter gene (30). However, the relative contribution of the SIE to fos transcription is controversial (40). Data from the present study provide evidence that Fos expression can be modulated by the STATs pathway in vivo in the absence of any changes in the ERK/SRE pathway, even though the possible contribution of other factors cannot be completely ruled out. Therefore, the effects of the LMW-PTP suggest that both the SRE and the SIE are necessary to achieve optimal fos transcription.
The mechanisms involved in the reduced activation of STAT proteins by the LMW-PTP remain to be clarified. Three different Jak kinases, Jak-1, Jak-2, and Tyk-2 have been shown to interact with the PDGF receptor (36). However, none of these enzymes is necessary per se to obtain STATs phosphorylation. Recent data indicate that v-Src transformed cells show greater activation of STAT3 (20,21), indicating a possible cross-talk between the Src and the STATs pathways. In addition, we have observed that in PDGF-stimulated NIH 3T3 cells c-Src overexpression leads to an up-regulation of both STAT1 and STAT3 (41).
In this study we present evidences showing that the activation of both STAT1 and STAT3 are clearly affected by the LMW-PTP, a finding that could be explained by the concomitant action of LMW-PTP on Src activity, thus suggesting that the STATs and Src pathways could be related. However, the direct LMW-PTP action on a Src-independent STATs pathway cannot be excluded.
In summary, the results of this study indicate that LMW-PTP decreases PDGF mitogenic signaling by selectively interacting with the Src and the STATs pathway, resulting in modulated expression of myc and fos, two proto-oncogenes crucial for G 1 progression. Whether the LMW-PTP affects PDGF receptor signal transduction pathways through the dephosphorylation of different receptor tyrosines remains to be established. On the other hand, this phosphatase may directly and specifically interact with only one phosphorylated tyrosine, leading to pleiotropic effects. The available data indicate that LMW-PTP may act as a physiological modulator of the effects of PDGF on cell growth.