Glutaredoxin Modulates Platelet-derived Growth Factor-dependent Cell Signaling by Regulating the Redox Status of Low Molecular Weight Protein-tyrosine Phosphatase*

Glutaredoxin (GRX) is a glutathione-disulfide oxidoreductase involved in various cellular functions, including the redox-dependent regulation of certain integral proteins. Here we demonstrated that overexpression of GRX suppressed the proliferation of myocardiac H9c2 cells treated with platelet-derived growth factor (PDGF)-BB. After stimulation with PDGF-BB, the phosphorylation of PDGF receptor (PDGFR) β was suppressed in GRX gene-transfected cells, compared with controls. Conversely, the phosphorylation was enhanced by depletion of GRX by RNA interference. In this study we focused on the role of low molecular weight protein-tyrosine phosphatase (LMW-PTP) in the dephosphorylation of PDGFRβ via a redox-dependent mechanism. We found that depletion of LMW-PTP using RNA interference enhanced the PDGF-BB-induced phosphorylation of PDGFRβ, indicating that LMW-PTP works for PDGFRβ. The enhancement of the phosphorylation of PDGFRβ was well correlated with inactivation of LMW-PTP by cellular peroxide generated in the cells stimulated with PDGF-BB. In vitro, with hydrogen peroxide treatment, LMW-PTP showed decreased activity with the concomitant formation of dithiothreitol-reducible oligomers. GRX protected LMW-PTP from hydrogen peroxide-induced oxidation and inactivation in concert with glutathione, NADPH, and glutathione disulfide reductase. This strongly suggests that retention of activity of LMW-PTP by enhanced GRX expression suppresses the proliferation of cells treated with PDGF-BB via enhanced dephosphorylation of PDGFRβ. Thus, GRX plays an important role in PDGF-BB-dependent cell proliferation by regulating the redox state of LMW-PTP.

The redox status of sulfhydryl groups in proteins plays an important role in the regulation of cellular functions such as the synthesis and folding of proteins and regulation of the structure and activity of enzymes, receptors, and transcription factors. To maintain the cellular thiol-disulfide redox status under reducing conditions, living cells possess two major systems, the thioredoxin (TRX) 3 /thioredoxin reductase system and the glutathione (GSH)/glutaredoxin (GRX) system (1).
GRX, also known as thiol transferase, was first discovered as a GSH-dependent hydrogen donor for ribonucleotide reductase in Escherichia coli mutants lacking TRX (2). GRX catalyzes the reduction of protein disulfide via a disulfide exchange reaction by utilizing the active site Cys-Pro-Tyr-Cys with a dithiol mechanism involving both active-site thiols (3). In addition, GRX has a unique ability to reduce protein-S-S-glutathione mixed disulfide (deglutathionylation) or to participate in its formation (glutathionylation) through a monothiol mechanism, which requires only the more N-terminal active site Cys (3,4). Oxidized GRX is selectively recycled to the reduced form by GSH with the formation of glutathione disulfide (GSSG) and regeneration of GSH through coupling with NADPH and GSSG reductase, a system termed the GSH-regenerating system (5,6). These characteristic interactions with GSH distinguish GRX from TRX, which favors intramolecular disulfide substrates and is turned over by NADPH and thioredoxin reductase independent of GSH. Functional overlap or cross-talk between the two systems, however, has been indicated (7,8).
Mammalian GRX is known to have two isoforms, GRX1 and GRX2 (4). GRX1 is a cytosolic form of GRX. GRX1 and S-glutathionylation are thought to be involved in a variety of cellular events such as signal transduction, stress response, and metabolic regulation by regulating the redox status of various cellular proteins including human immunodeficiency virus-1 protease (9), glyceraldehyde-3-phosphate dehydrogenase (10), nuclear factor I (11), ASK1 (12), EphA2 kinase (13), actin (14), Ras (15), tubulin (16), tau and microtubule-associated protein-2 (17), annexin A2 (18), and protein-tyrosine phosphatase 1B (4,19). We also have reported that GRX plays an important role in protecting cells from apoptosis by regulating the redox status of Akt/ protein kinase B (20). The other form of GRX, GRX2, is distributed in the mitochondria and nucleus and also functions as part of a cellular antioxidant defense system by regulating the redox status of mitochondrial proteins such as complex I protein (21)(22)(23).
Platelet-derived growth factor (PDGF) is a major mitogen for connective tissue cells and certain other cell types. PDGF is a family of homo-and heterodimers of disulfide-bonded polypeptide chains A, B, C, and D (i.e. PDGF-AA, BB, CC, DD, and AB) (24). PDGFs bind to two cell-surface receptor-tyrosine kinases, PDGF receptor (PDGFR) ␣ and ␤. The PDGF molecules are bivalent, and PDGF-dependent activation of receptors causes a mitogenic signal transduction through the phosphorylation of specific tyrosines in the receptors. The in vivo function of PDGF signaling has been studied with gene targeting experiments (25,26). PDGF-BB or PDGFR␤ knock-out mice die during late gestation from cardiovascular complications (27,28), suggesting the importance of PDGF-BB/PDGFR␤ signaling in the myocardiac development of mice. Although the physiological functions of PDGF-BB/PDGFR␤ signaling in myocardiac cells are not fully understood, PDGFR␤ is expressed in adult rat heart (29) and rat cardiomyocyte-derived H9c2 cells (30), and PDGF-BB/PDGFR␤ signaling also controls the proliferation of neonatal rat cardiomyocytes (31). These reports suggest that the PDGF-BB/PDGFR␤ signaling pathway plays an important role in the cellular physiology of myocardiac cells, although the precise regulatory mechanism for the signaling pathway and its relation to cellular redox-dependent regulation are not yet clarified.
In the present study we explored the effect of overexpression of GRX on PDGF-BB-dependent signaling and cell proliferation in myocardiac H9c2 cells. We found that overexpression of GRX suppressed the tyrosine phosphorylation of PDGFR␤ after stimulation with PDGF-BB, resulting in a suppression of the PDGF-BB-dependent cell proliferation. Furthermore, we describe a novel regulatory mechanism for PDGF-BB signaling involving the redox-dependent regulation of low molecular weight protein-tyrosine phosphatase (LMW-PTP) by GRX in a GSH-dependent manner.
Cell Line and Culture-H9c2 cells, a clonal line derived from embryonic rat heart, were obtained from American Type Culture Collection (CRL-1446). H9c2 cells which had been transfected with the expression vector for mouse GRX1 cDNA have been described previously (20). Cell lines (GRX22 and -30) expressing high levels of FLAG-tagged GRX protein were used in the study. H9c2 cells and gene-transfected cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum (FCS) containing 75 g/ml G418 and 75 g/ml hygromycin B in a humidified atmosphere of 95% air and 5% CO 2 at 37°C.
Cell Proliferation-The proliferation of cultured cells was evaluated by measuring attached live cells photometrically after staining with crystal violet. The cells (5000) were placed in 100 l of medium/well in 96-well plates and cultured in medium containing 0.5% FCS with or without 0.5 nM PDGF-BB for specific periods. Then the cells were fixed with 4% paraformaldehyde in phosphate-buffered saline (pH 7.5), washed, and stained with 0.01% crystal violet at room temperature for 20 min. Each well was extensively washed with water and dried. The stained cells were lysed by adding 100 l of lysis buffer A (10% SDS and 0.1 N HCl), and the cell number was then estimated photometrically by measuring the absorbance at 570 nm using a microplate reader.
Immunoblot Analysis-Cultured cells were harvested and lysed for 20 min at 4°C in lysis buffer B (20 mM Tris (pH 7.2), 150 mM NaCl, and 1% Nonidet P-40, including protease inhibitors (20 M phenylmethylsulfonyl fluoride, 50 M pepstatin, and 50 M leupeptin)). The protein concentration was determined using a BCA assay kit (Pierce). Protein samples were electrophoresed on SDS-polyacrylamide gels (7.5-15%) under reducing conditions as described previously (32). The proteins in the gels were transferred onto a nitrocellulose membrane. The membranes were blocked in Tris-buffered saline (TBS, 10 mM Tris-HCl (pH 7.5) and 150 mM NaCl) containing 0.1% (v/v) Tween 20 (TBST) and 5% (w/v) nonfat dry milk and then reacted with primary antibodies in TBST containing 5% (w/v) bovine serum albumin or 5% (w/v) nonfat dry milk overnight with constant agitation at 4°C. After several washes with TBST, the membranes were incubated with peroxidase-conjugated secondary antibodies. Proteins in the membranes were then visualized using the enhanced chemiluminescence (ECL) detection kit (Amersham Biosciences) according to the manufacturer's instructions.
Immunoprecipitation-Cultured cells were harvested and lysed in lysis buffer B. Cell lysates normalized for protein levels were immunoprecipitated using anti-c-Myc monoclonal antibody-conjugated agarose beads for 2 h at 4°C. The beads were washed three times with cold lysis buffer B. For the immunoblot analysis, the beads were resuspended in 100 l of SDS sample buffer containing DTT and boiled for 3 min. Then the supernatant was directly subjected to SDS-PAGE as described above.
Construction of LMW-PTP Gene Expression Vector-Fulllength human LMW-PTP cDNA was cloned from total RNA of human colon cancer HCT8 cells by reverse transcription-PCR using Super Script II RNase H reverse transcriptase (Invitrogen) and Advantage HF2 Taq polymerase (BD Biosciences Clontech) with the following primer pair, which was designed on the basis of reported nucleotide sequences for human LMW-PTP (34) (GenBank TM accession number M83653): forward, 5Ј-CGT GGA TCC GAG AAG ATG GCG-3Ј, and reverse, 5Ј-CCC GAA TTC TCA GTG GGC CTT-3Ј. The amplified DNA fragments were subcloned into a TA cloning vector (pCRII, Invitrogen), and the nucleotide sequences were confirmed by sequencing with an ALFexpress II system (Amersham Biosciences). The LMW-PTP cDNA was cloned into a C-terminal Myc-tagged expression vector, pcDNA3.1/Myc-His (Invitrogen) under the control of the CMV promoter for expression in mammalian cells and also was cloned into a bacterial expression vector, pGEX6p-1 (Amersham Biosciences). The gene transfection into mammalian cells was performed using Lipofectamine Plus reagent (Invitrogen) according to the manufacturer's directions.
LMW-PTP Activity Assay-The activity of LMW-PTP was measured according to the methods of Chiarugi et al. (35) with a slight modification. Briefly, the LMW-PTP expression vector was transfected into control and GRX gene-transfected H9c2 cells. After 24 h of transfection, the cells were lysed, and the cell lysates normalized for protein levels were subjected to immunoprecipitation using anti-c-Myc monoclonal antibody-conjugated agarose beads at 4°C for 2 h. In each experimental condition, duplicate immunoprecipitates were prepared and subjected to the PTP activity assay or to immunoblot analysis for estimation of the level of LMW-PTP obtained in each immunoprecipitation. After washing, the beads were incubated at 37°C for 30 min in 0.1 M sodium acetate (pH 5.5), 1 mM EDTA, and 4 mM p-nitrophenyl phosphate, then the reaction was stopped by the addition of 6 N NaOH. Aliquots of the mixture were used to determine the phosphatase activity spectrophotometrically by measuring the absorbance at 410 nm. Each activity was standardized by comparing with the level of LMW-PTP in the immunoprecipitates densitometrically estimated in immunoblot analysis and expressed as the relative value to the activity for the untreated control (% of untreated control). Furthermore, LMW-PTP activity in vitro was also measured for recombinant LMW-PTP after incubation with or without H 2 O 2 . The aliquots were diluted 20-fold by adding 0.1 M sodium acetate (pH 5.5), 1 mM EDTA and subjected to the PTP activity assay as described above. Activity was quantified by comparison with a standard curve (0.01-0.2 mM p-nitrophenol).
Purification of Recombinant LMW-PTP and Generation of Antibody against LMW-PTP-LMW-PTP was purified with the glutathione S-transferase (GST) gene fusion system (Amersham Biosciences) according to the manufacturer's instructions. In brief, E. coli strain BL21 cells were transformed with pGEX6p-LMW-PTP, and protein expression was induced by isopropyl ␤-D-1-thiogalactopyranoside. GST-fused LMW-PTP (GST-LMW-PTP) was affinity-purified from cell lysates using glutathione-Sepharose 4B (Amersham Biosciences) and then digested with PreScission protease. The cleaved GST was removed from glutathione-Sepharose 4B, and LMW-PTP was purified. The LMW-PTP was used to immunize rabbits to generate anti-LMW-PTP antibodies and also used for in vitro experiments concerning redox regulation of LMW-PTP as described below.
Determination of the Redox State of LMW-PTP in Vitro-The redox state of LMW-PTP was assessed by modifying free thiol with AMS (20). Briefly, recombinant LMW-PTP (2 g) was reduced by incubation with 50 mM DTT for 30 min on ice. DTT was then removed by gel filtration on NAP-10 columns (Amersham Biosciences). Reduced LMW-PTP was incubated at room temperature for the periods indicated with H 2 O 2 in buffer with or without the GRX (4 -40 g) and/or GSH-regenerating system as described (20). The GSH-regenerating system was composed of GSH/GSSG (1 mM GSH and 0.05 mM GSSG), NADPH (1 mM), and GSSG reductase (GR) (1.2 units). After incubation, LMW-PTP protein was incubated with buffer containing a final concentration of 50 mM Tris-HCl (pH 7.4), 1% SDS, and 15 mM AMS at 4°C for 1 h. The reaction was stopped by the addition of 5ϫ SDS sample buffer. Proteins were separated by 15% SDS-PAGE with or without DTT and blotted to a nitrocellulose membrane. Proteins in the membrane were then visualized by immunoblot analysis as described above.
Peroxide Quantification-Peroxide was quantified using the PeroXOquant quantitative peroxide assay kit (Pierce) according to the manufacturer's instructions. In brief, cells were incu-bated with PDGF-BB for given periods. Cells were then harvested and lysed on ice in 100 l of distilled water by sonication using an ultrasonic applicator (Bioruptor UCD-200T, COSMO BIO, Tokyo, Japan). Working reagent (100 l) was mixed with 10 l of lysate and incubated at room temperature for 15-20 min. The purple product composed of Fe 3ϩ -xylenol orange complex was detected spectrophotometrically at 570 nm in the reaction mixture. Activity was quantified by comparison with a standard curve (1-50 M H 2 O 2 ). The protein concentration was determined using a BCA assay kit (Pierce).
Statistical Analysis-Statistical analysis was performed using Student's t test or analysis of variance (StatView software). Significance was set at p Ͻ 0.05.

Overexpression of the GRX Gene Inhibits PDGF-induced Proliferation of H9c2
Cells-Rat myocardiac H9c2 cells were transfected with mouse GRX1 gene expression vectors to obtain two cell lines overexpressing GRX. Fig. 1A shows that expression of GRX increased in the overexpressors (GRX22 and GRX30) and was immunologically undetectable in mock-transfected (control) H9c2 cells. Recently it has been reported that an antioxidant inhibited cell growth by blocking the phosphorylation of tyrosines in PDGFR␤ elicited by PDGF-BB (36). In H9c2 cells PDGF-BB induces cell proliferation through PDGFR␤ (30). We also examined the PDGF-BB-dependent proliferation of H9c2 cells in culture medium containing 0.5% FCS with different concentrations of PDGF-BB (0 -0.5 nM) and found that the cells proliferated with PDGF-BB in a dose-dependent manner (Fig. 1B, inset). The result showed that 0.5 nM PDGF-BB could sufficiently stimulate the proliferation of H9c2 cells and was consistent with a previous report by Brostrom et al. (30). To investigate the functional role of overexpressed GRX in myocardiac cell proliferation, control and GRX gene-transfected cells were cultured in medium supplemented with 0.5% FCS with or without 0.5 nM PDGF-BB, then cell proliferation was analyzed as described under "Materials and Methods." In Fig.  1B, after 3 days of culture with 0.5 nM PDGF-BB, the cell count had doubled in control cells, but the increase was suppressed among the transfectants. On the other hand, without PDGF-BB, proliferation was apparently suppressed in all cells. To investigate whether PDGF-BB signaling is influenced by overexpression of GRX, the phosphorylation status of PDGFR␤ was examined immunologically in the cells treated with 0.5 nM PDGF-BB (Fig. 1C). In control cells phosphorylation of PDGFR␤ was markedly increased by PDGF-BB, but the increase was apparently suppressed in GRX gene-transfected cells treated with PDGF-BB. Together, these results indicate that PDGF-dependent cell proliferation was suppressed by overexpression of GRX in H9c2 cells through the suppressed phosphorylation of a specific receptor, PDGFR␤. Using 125 Ilabeled PDGF-BB, the binding of PDGF-BB both to control and GRX gene-transfected cells was investigated as described in the supplemental material. The results showed that overexpression of GRX did not influence the PDGF-BB binding to the cells (supplemental Fig. S1, A and B). Furthermore, we examined cell surface expression of PDGFR␤ in control and GRX gene-transfected cells after biotinylation of cell surface proteins and found that overex-pression of GRX did not change the expression level of cell surface PDGFR␤ (supplemental Fig. S1C). Together, these results indicate that overexpression of GRX does not influence the ligand-receptor binding mechanism of PDGF-BB in H9c2 cells.
LMW-PTP Regulates Phosphorylation of PDGFR␤-For the dephosphorylation of PDGFR, there are various PTPs reported, which include Src homology 2-containing protein-tyrosine phosphatase (SHP)-1 and SHP-2, density-enhanced proteintyrosine phosphatase-1 (DEP-1), PTP-1B, PTP-PEST, T-cell protein-tyrosine phosphatase (TC-PTP), and LMW-PTP (37,38). Among them, we focused on LMW-PTP because its redoxdependent regulation was extensively studied in the case of PDGF-BB signaling (39). We examined the expression of LMW-PTP in H9c2 cells by reverse transcription-PCR and immunoprecipitation followed by immunoblotting using anti-LMW-PTP antibodies as described in the supplemental material. The results showed that the levels for LMW-PTP were similar in control and GRX gene-transfected cells, indicating that LMW-PTP is expressed in H9c2 cells, and the expression level is not influenced by overexpression of the GRX gene (supplemental Fig. S2). To investigate whether LMW-PTP is responsible for the dephosphorylation of PDGFR␤ in H9c2 cells, LMW-PTP expression was down-regulated by siRNA transfection, and the effect of suppressed LMW-PTP expression on PDGF signaling was examined. Cells were transfected with siRNAs for LMW-PTP for 48 h, then the transcriptional level of LMW-PTP was examined by reverse transcription-PCR as described under "Materials and Methods." As shown in Fig. 2A, the expression of LMW-PTP was apparently suppressed in cells transfected with siRNAs compared with cells transfected with or without control siRNA bearing scramble sequences for LMW-PTP. After transfection with siRNA, cells underexpressing LMW-PTP were treated with PDGF-BB, and the level of tyrosine phosphorylation of PDGFR␤ was examined in immunoblot analysis (Fig. 2C). The result showed that the level of phosphorylation was apparently increased in cells transfected with siRNAs for LMW-PTP compared with controls. Together, these results demonstrate that LMW-PTP is an endogenous PTP responsible for the dephosphorylation of PDGFR␤ in H9c2 cells.
Next, to investigate whether GRX affects the dephosphorylation of PDGFR␤ in H9c2 cells, the expression of GRX was down-regulated by siRNA transfection, and the effect of suppressed GRX expression on PDGF signaling was examined. Cells were transfected with siRNAs targeting GRX for 48 h, then the transcriptional level of GRX was examined by reverse transcription-PCR as described under "Materials and Methods." As shown in Fig. 2B, the expression of GRX was apparently suppressed in cells transfected with siRNAs compared with cells transfected with or without control siRNA. After transfection with siRNA, cells underexpressing GRX were treated with PDGF-BB, and the level of tyrosine phosphorylation of PDGFR␤ was examined in immunoblot analysis (Fig.  2C). The result showed that the level of phosphorylation was apparently increased in cells transfected with siRNAs for GRX compared with controls. Together, these results demonstrate that the expression of GRX negatively regulates PDGF signaling to suppress the phosphorylation of PDGFR␤ in H9c2 cells treated with PDGF-BB.
To further examine the involvement of LMW-PTP in the mechanism regulating PFGF-BB signaling, we prepared the cells transiently transfected with the expression vector for LMW-PTP, because the endogenous level of LMW-PTP protein in H9c2 cells was too low to investigate its biochemical characteristics. To clarify whether LMW-PTP regulates the phosphorylation of PDGFR␤ in H9c2 cells, the level of tyrosine phosphorylation of PDGFR␤ was examined in an immunoblot analysis using H9c2 cells transfected with two amounts of LMW-PTP gene expression vector, pcDNA-LMW-PTP. As shown in Fig. 3A, the level of phosphotyrosine in PDGFR␤ was decreased by transfection with pcDNA-LMW-PTP. To investigate whether transient expression of LMW-PTP gene influences the expression level of cell surface PDGFR␤, we immunoprecipitated PDGFR␤ from cell lysates of the cells transfected with or without pcDNA-LMW-PTP after biotinylation of cell surface proteins, and immunoprecipitates were examined by immunoblot analysis to detect PDGFR␤ or biotinylated proteins. The results showed that the expression of exogenous LMW-PTP does not change the expression level of cell surface PDGFR␤ (data not shown). To confirm that the decrease in the tyrosine phosphorylation of PDGFR␤ was due to an increase in LMW-PTP activity caused by the gene transfection, the effect of PTP inhibitor III (PTPI-III) on the phosphorylation status of PDGFR␤ was examined in the cells transfected with pcDNA-LMW-PTP. PTPI-III is a cell-permeable broad-range inhibitor of PTPs (40). In Fig. 3B, upper, the cells transfected with pcDNA-LMW-PTP were pretreated with various concentrations of PTPI-III for 30 min. Expressed Myctagged LMW-PTP was immunoprecipitated, then the activity for LMW-PTP was assayed in the immunoprecipitates as described under "Materials and Methods." The LMW-PTP activity was decreased by PTPI-III in a dose-dependent manner. In Fig. 3B, lower, the status of PDGFR␤ was also examined by immunoblot analysis in the LMW-PTP gene-transfected cells treated with PTPI-III. Although the phosphorylation of PDGFR␤ was apparently suppressed by transfection with pcDNA-LMW-PTP, the level gradually increased on treatment with PTPI-III in a dose-dependent manner. These results indicate that overexpression of the LMW-PTP gene causes a decrease in the phosphorylation of PDGFR␤ simply through increased LMW-PTP activity.
Next, to investigate how endogenous LMW-PTP activity influenced the PDGF-BB-stimulated phosphorylation of PDGFR␤, the status of PDGFR␤ was examined in the control and GRX gene-transfected cells treated with or without PDGF-BB and/or various concentrations of PTPI-III. Control and GRX gene-transfected cells were pretreated with or without PTPI-III for 30 min and then stimulated with PDGF-BB. In Fig. 3C, the level of phosphorylated PDGFR␤ after stimulation with PDGF-BB was highly increased in control cells compared with GRX gene-transfected cells. This result was consistent with the result shown in Fig. 1C. However, when endogenous LMW-PTP activity was suppressed with PTPI-III, the phosphorylation of PDGFR␤ was apparently increased in GRX gene-transfected cells treated with PDGF-BB. On the other hand, in control cells treated with PDGF-BB, the phosphorylation level showed no further increase even in the presence of PTPI-III. Thus, phosphorylation levels were similar between control and GRX gene-transfected cells treated with PTPI-III and PDGF-BB. These findings demonstrate that the activity of endogenous PTP such as LMW-PTP plays an important role in the regulation of the phosphorylation status of PDGFR␤ in GRX gene-transfected cells treated with PDGF-BB.
The Activity of LMW-PTP Is Up-regulated by Overexpression of GRX-The PTP activity of LMW-PTP is dependent on the level of cellular reduced glutathione (GSH) (35). To know the effect of overexpression of GRX on LMW-PTP, the activity of LMW-PTP was examined in control and GRX-overexpressing cells after transfection with the LMW-PTP gene expression vector. Control and GRX-overexpressing cells were transiently transfected with pcDNA-LMW-PTP. After 24 h of transfection, the cells were treated with PDGF-BB for given periods, and the activity of LMW-PTP was estimated as described under "Materials and Methods." As shown in Fig. 4A, the activity was always stronger in GRX-overexpressing cells than controls during the treatment with PDGF-BB, indicating that the expression level of GRX affected the activity of LMW-PTP. To investigate whether the up-regulation of LMW-PTP activity was due to the effect of overexpressed GRX, the effect of cadmium (CdCl 2 ), an inhibitor for GRX (41), on the LMW-PTP activity was examined in control and GRX-overexpressing cells transiently transfected with pcDNA-LMW-PTP. After 24 h of transfection, the cells were treated with or without PDGF-BB and/or CdCl 2 . Then the activity of LMW-PTP was estimated as described under "Materials and Methods." Treatment with 200 M CdCl 2 reduced the thiol transferase activity of GRX by more than 60% in GRX-overexpressing H9c2 cells (20). As shown in Fig. 4B, the LMW-PTP activity was significantly suppressed by CdCl 2 in control and GRX-overexpressing cells with or without PDGF-BB, indicating that GRX activity contributed to the activation of LMW-PTP. In Fig. 4C, the effect of CdCl 2 on the phosphorylation of PDGFR␤ was examined in control and GRX-overexpressing cells treated with PDGF-BB. The results showed that the phosphorylation was apparently increased by CdCl 2 in control and GRX-overexpressing cells treated with PDGF-BB. The phosphorylation levels were similarly increased in control and GRX-overexpressing cells treated with PDGF-BB. Collectively, these results indicate that the activity for LMW-PTP is up-regulated by overexpression of GRX through the increased thiol transferase activity of GRX. This also suggests that GRX plays an important role in regulating the redox state of LMW-PTP under PDGF-BB treatment. Expressed Myc-tagged LMW-PTP was immunoprecipitated using anti-c-Myc monoclonal antibody-conjugated agarose beads, and then the activity for LMW-PTP was assayed in the immunoprecipitates as described under "Materials and Methods" (upper). Each value represents the mean Ϯ S.D. of three independent experiments. The phosphorylation status of PDGFR␤ was examined by immunoblot analysis as described above (lower). The data represent three independent experiments. C, control and GRX genetransfected H9c2 cells were serum-starved for 24 h, treated with the indicated concentration of PTPI-III for 30 min, and then stimulated with or without 0.5 nM PDGF-BB. The phosphorylation status of PDGFR␤ was examined by immunoblot analysis as described above. The data represent three independent experiments.

Peroxide Is Involved in the Down-regulation of LMW-PTP
Activity during PDGF-BB Signaling-The intracellular generation of reactive oxygen species (ROS) is a pivotal event in growth factor-receptor signaling (42)(43)(44). There are several reports showing that the phosphorylation status of growth factor receptors is regulated through redox regulation of specific phosphatases (45). LMW-PTP has eight Cys residues in the structure and is oxidized and inactivated by oxidative stress such as nitric oxide and H 2 O 2 with the formation of a disulfide bond between Cys-12 and Cys-17 (46,47). At first, to investigate whether the treatment with PDGF-BB could generate ROS in the cell, control and GRX-overexpressing cells were treated with PFGF-BB for given periods, and the amount of intracellular peroxide was estimated as described under "Materials and Methods." As shown in Fig. 5A, peroxide assays revealed that the level of cellular H 2 O 2 was increased by PDGF-BB with a peak at 5-10 min then decreased to the initial level at 30 min. In this respect, the results suggest that the PDGF-dependent generation of peroxide may lead to the inactivation of LMW-PTP. Next we investigated the effect of chemical antioxidants such as Trolox, an analog of vitamin E, or N-acetyl cysteine (NAC), a supplier of cellular thiols, on the PDGF signaling and LMW-PTP activity. When control and GRX-overexpressing cells were treated with Trolox (500 M) or NAC (1 mM) for 30 min, production of peroxide almost ceased (data not shown). In Figs. 5, B and C, control cells were transiently transfected with the  LMW-PTP gene expression vector and then treated with or without 0.5 nM PDGF-BB in the presence or absence of Trolox or NAC. The phosphorylation of PDGFR␤ was examined by immunoblot analysis (Fig. 5B), and LMW-PTP activity was also estimated as described under "Materials and Methods" (Fig.  5C). In the presence of Trolox or NAC, the PDGF-dependent phosphorylation of PDGFR␤ was apparently suppressed compared with untreated controls. On the other hand, without Trolox or NAC, LMW-PTP activity was significantly decreased by PDGF-BB. However, with Trolox or NAC, LMW-PTP activity was up-regulated compared with untreated cells. As a result, the PDGF-dependent decrease in LMW-PTP activity was reversed by the reduction of peroxide with Trolox or NAC.
Peroxide Causes the Formation of DTT-reducible High Molecular Weight Oligomers of LMW-PTP in Vitro with a Concomitant Decrease of PTP Activity-Next, to investigate the redox-dependent effect of H 2 O 2 on LMW-PTP, the redox-dependent structural change of LMW-PTP was examined in vitro in the presence of H 2 O 2 . LMW-PTP was prepared using a bacterial GST fusion protein expression system as described under "Materials and Methods." For reduction, LMW-PTP was initially incubated with 50 mM DTT for 30 min, then excess DTT was removed by gel filtration as described under "Materials and Methods." LMW-PTP was then treated for 10 min with various concentrations of H 2 O 2 , and the samples were incubated with AMS to see the status of thiols by blocking free sulfhydryls. Then the samples were subjected to SDS-PAGE under nonreducing or reducing conditions, and LMW-PTP was detected by immunoblot analysis using specific antibodies (Fig. 6A). With H 2 O 2 treatment, LMW-PTP showed a slower migration and formed oligomers under non-reducing conditions (ϪDTT). Under reducing conditions (ϩDTT), high molecular weight protein bands (oligomers) disappeared, and a single low molecular weight band (ϳ25 kDa) was observed with or without H 2 O 2 . In Fig. 6B, LMW-PTP was treated with H 2 O 2 as in Fig.  6A, then the LMW-PTP activity in vitro was assayed. The results showed that LMW-PTP was inactivated by H 2 O 2 in a dose-dependent manner. These results indicate that H 2 O 2 induces inactivation and oxidation-induced oligomer formation of LMW-PTP. Collectively, these results support that the generation of ROS such as peroxide plays an important role in regulating the redox state of LMW-PTP to down-regulate the activity leading to an increase in the level of phosphorylated PDGFR␤ in cells stimulated with PDGF-BB.
GRX Requires the GSH-regenerating System for the Redox-dependent Regulation of LMW-PTP in Vitro-To further clarify the molecular mechanism by which GRX regulates the redox state of LMW-PTP, we examined whether GRX could protect LMW-PTP from oxidation-induced structural change in the presence of H 2 O 2 in vitro. GRX was purified using the GST fusion protein expression system as described previously (20). GST-fused GRX was affinity-purified from E. coli cell lysates using glutathione-Sepharose 4B and then digested with Pre-Scission protease. The cleaved GST was removed from glutathione-Sepharose 4B, and GRX was purified for the experiments. The LMW-PTP was incubated with or without 200 M H 2 O 2 under various conditions. The treatment with H 2 O 2 decreased the PTP activity of LMW-PTP with or without GRX (Fig. 7A, left). However, in the presence of the GSH-regenerating system (GSH/GSSG and NADPH/GSSG reductase), GRX significantly suppressed the oxidation-induced inactivation of LMW-PTP, although the regenerating system itself also slightly suppressed the inactivation (Fig. 7A, right). Furthermore, in the presence of the GSH-regenerating system, GRX protected LMW-PTP from oxidation-induced inactivation in a dose-dependent manner (Fig. 7B). Then we examined the effect of GRX and/or the GSH-regenerating system on the redox state of LMW-PTP after treatment with H 2 O 2 in vitro. In Fig. 7C, H 2 O 2 caused oxidation-induced oligomer formation in LMW-PTP. The change was not suppressed solely by GRX. However, in the presence of the GSH-regenerating system, GRX could reduce the oxidation-induced change of LMW-PTP in a dose-dependent manner. On the other hand, when LMW-PTP was initially oxidized by H 2 O 2 , it was completely inactivated and not reactivated even in the presence of DTT or GRX with a GSH-regenerating system (data not shown). Taken together, these results indicate that GRX regulates the redox state of LMW-PTP in concert with the GSH-regenerating system to protect the activity against oxidative stress.

DISCUSSION
We have shown that when overexpressed, GRX suppressed the tyrosine phosphorylation of PDGFR␤ through an up-regu- lation of the activity of LMW-PTP, resulting in a suppression of proliferation in H9c2 cells stimulated with PDGF-BB and that the GSH/GRX system protected LMW-PTP from H 2 O 2 -induced inactivation and structural changes.
GRX shows a variety of cellular functions by modulating the redox status of a number of cellular proteins (4). In terms of cell proliferation, there are several signaling pathways controlled by redox-dependent regulation in the cell (45). However, it was not clear how GRX could be involved in the regulation of cell proliferation. Recent findings indicate that treatment with growth factors or ROS promotes the tyrosine phosphorylation of growth factor receptors (48). This effect can be achieved by the activation of protein tyrosine kinases and/or inactivation of PTPs. It has been shown that ROS such as superoxide and H 2 O 2 are transiently generated intracellularly when cells are stimulated with cytokines or growth factors (49). In addition, exogenous oxidants could be produced under various physiological conditions such as during oxidative burst by neutrophils, monocytes, and macrophages, or pathological conditions such as during reperfusion after ischemia. PTPs have been shown to be regulated by a redox mechanism, although there is no convincing evidence that protein tyrosine kinases are activated by ROS (37,45). Oxidation of PTPs always takes place at the catalytic site cysteine where the sulfhydryl residue is transformed to sulfenic acid. Oxidized PTPs are catalytically inactive because they cannot form the cysteinyl-phosphate intermediate during the first step of the catalysis. These observations suggest that PTPs might undergo ROS-dependent inactivation in cells, resulting in a shift in the equilibrium of the protein tyrosine kinases toward phosphorylation.
The PDGF pathway is a well characterized growth factor signaling pathway, and its importance is also recognized in the development of certain cell types. Gene knock-out experiments in mice emphasized the importance of PDGF-BB/PDGFR␤ signaling in myocardiac development. Furthermore, PDGFR␤ is expressed in adult rat heart (29) and rat myocardiac H9c2 cells (30), and PDGF-BB/PDGFR␤ signaling also controls the proliferation of neonatal rat cardiomyocytes (31). These findings suggest that the PDGF-BB/PDGFR␤ signaling pathway plays an important role in the cellular physiology of myocardiac cells, although the precise regulatory mechanism for the pathway has yet to be clarified.
Among PTPs that are involved in PDGF signaling we focused on LMW-PTP because the redox-dependent regulation of the molecule and its role in PDGF-BB/PDGFR␤ signaling have been studied extensively (39). LMW-PTP was expressed in H9c2 cells, and the expression level was not affected by transfecting the GRX gene expression vector (supplemental Fig. S2). When the expression of LMW-PTP was suppressed by siRNA transfection, PDGF-BB-induced tyrosine phosphorylation was apparently enhanced compared with untransfected controls (Fig. 2C). This strongly suggests that LMW-PTP is responsible for the dephosphorylation of PDGFR␤ in H9c2 cells, although other PTPs may also be involved in the mechanism. On the other hand, the PDGF-induced phosphorylation of PDGFR␤ was apparently enhanced in GRX-overexpressing cells treated with tyrosine phosphatase inhibitor PTPI-III but was not enhanced in control cells treated with PTPI-III (Fig. 3C). In GRX-overexpressing cells, the enhancing effect of PTPI-III on the PDGF-induced phosphorylation of PDGFR␤ was consistent with the data using siRNA for LMW-PTP. However, at present it is not clear why suppression of endogenous LMW-PTP by PTPI-III does not much influence on the PDGF-BBinduced phosphorylation of PDGFR␤ in control cells. It is reported that ␣-bromoacetophenone derivatives such as PTPI-III act as PTP inhibitors by covalently alkylating the conserved catalytic Cys residue in the PTP active site (58). In this respect it is possible that the susceptibility of endogenous LMW-PTP to PTPI-III could be changed when the redox state of endogenous LMW-PTP is different between control and GRX-overexpressing cells. This suggests that the effect of PTPI-III on endogenous LMW-PTP may be highly enhanced in GRX-overexpressing cells compared with controls.
LMW-PTP is an 18-kDa enzyme that is widely expressed (39,59). Previous studies on the molecular biology of LMW-PTP in NIH3T3 cells demonstrated a well defined role for this enzyme in PDGF-BB-induced mitogenesis showing that activated PDGFR␤ is a substrate for LMW-PTP (56). We showed that cellular LMW-PTP was inactivated by peroxide, which could be generated with PDGF-BB, and the inactivation was suppressed by antioxidants such as Trolox or NAC (Figs. 5, B and C). These results support previous findings that LMW-PTP is oxidized by ROS generated in cells stimulated with PDGF-BB, resulting in up-regulation of the phosphorylation of PDGFR␤. We also showed that GRX could up-regulate the activity of LMW-PTP even in the cells treated with PDGF-BB, resulting in an enhanced dephosphorylation of PDGFR␤. Furthermore, in vitro we found that treatment with H 2 O 2 causes the formation of high molecular weight oligomers of LMW-PTP in SDS-PAGE under non-reducing conditions. The high molecular weight form disappeared and shifted to a single molecule of 25 kDa in SDS-PAGE under reducing conditions. This suggested that LMW-PTP forms DTT-reducible oligomers in vitro through oxidation-induced disulfide-bond formation between the molecules. In the presence of GRX-and GSH-regenerating systems, the H 2 O 2 -induced oligomer formation of LMW-PTP was apparently suppressed, and the activity of LMW-PTP was also protected from H 2 O 2 -induced inactivation. Collectively, these results indicate that GSH/GRX systems effectively work to protect against oxidation-dependent structural change and inactivation of LMW-PTP.
During PDGF signaling, LMW-PTP is regulated by a redoxdependent mechanism involving the two Cys residues of the catalytic site, which change reversibly from the reduced to oxidized state after PDGF treatment. The reversibility of in vivo LMW-PTP oxidation is glutathione-dependent (35). The additional catalytic Cys-17 retains an interesting role in the formation of the S-S intramolecular bond, which protects the catalytic Cys-12 from further and irreversible oxidation. The presence of an additional Cys near the catalytic one confers upon LMW-PTP the ability to rapidly recover its activity and finely regulate PDGF receptor activation. To investigate whether specific Cys residues of the catalytic site of LMW-PTP are involved in the oxidation-induced oligomer formation of LMW-PTP in vitro, the LMW-PTP mutant, in which Cys-17 was changed to Ser by site-directed mutagenesis, was produced and characterized by non-reducing SDS-PAGE after treatment with H 2 O 2 . The results showed that high molecular weight oligomers were formed by H 2 O 2 even with the mutant LMW-PTP, indicating that Cys-17 had no particular contribution to oxidation-induced oligomer formation (data not shown). This suggests that there is no direct correlation between the inactivation status and formation of high molecular weight oligomers of LMW-PTP under conditions with H 2 O 2 in vitro, although oxidation-induced oligomer formation was effectively suppressed by GRX-and GSH-regeneration systems. It is also noteworthy that the oxidation-induced oligomer formation of LMW-PTP was not detected in non-reducing SDS-PAGE of cell lysate samples from LMW-PTP gene-transfected cells treated with H 2 O 2 (data not shown). Thus, further investigation is required to clarify precise mechanical correlation between the activation status and structure of LMW-PTP in cells treated with PDGF.
PTP-1B has been extensively studied in terms of its redoxdependent regulation related to the signaling receptors for epidermal growth factor (44,60) and insulin (61,62). The oxidation state of the active-Cys in PTP1B was also clearly characterized by structural analyses (63,64). PTP1B has been reported to be involved in the redox-dependent dephosphorylation of PDGFR. Lee et al. (19) reported that oxidized PTP-1B is reduced by TRX rather than GSH/GRX systems. On the other hand, Barrett et al. (65) reported that GSH could recycle oxidized PTP1B, showing the involvement of GSH-related systems in the mechanism. Thus, the precise mechanism of the redox regulation of PTP-1B is still controversial. SHP-2 is also an example of a redox-regulated PTP involved in the PDGFR signaling (52). Recently, the tumor suppressor protein, PTEN (phosphate and tensin homolog deleted on chromosome ten), has also been reported to dephosphorylate PDGFR (66). Although the activity of PTEN is regulated in a redox-dependent manner, oxidized PTEN is reduced by TRX rather than GRX in vitro (67). In the present study the expression of TRX was slightly suppressed in the GRX-overexpressing cells compared with controls (data not shown), but its relation to the redox-dependent regulation of LMW-PTP was not elucidated. On the other hand, in the case of mouse embryonic fibroblasts in which the PTP-1B gene was knocked out, the phosphorylation status of PDGFR␤ was not much influenced by stimulation with PDGF-BB, suggesting that other PTPs could compensate for the loss of PTP-1B function (68). Taken together, these findings suggest that several different PTPs are involved in the redox-dependent regulation of the dephosphorylation of phosphorylated PDGFR. Thus, to clarify the precise mechanism behind the regulation of PDGF signaling, further investigations are required. In conclusion, the present study shows that GRX and GSH-regenerating systems are involved in the regulation of PDGF-BB signaling through the redox-dependent regulation of LMW-PTP.