Roles of Protein-tyrosine Phosphatases in Stat1α-mediated Cell Signaling

Different Stat proteins are activated through phosphorylation of unique tyrosine residues in response to different cytokines and growth factors. Interferon-γ activates Stat1 molecules that form homodimers and bind cognate DNA elements. Here we show that treatment of permeabilized cells with 200-500 μM peroxo-derivatives of vanadium, molybdenum, and tungsten results in the accumulation of constitutively phosphorylated Stat1α molecules. In contrast, treatment of permeabilized cells with orthovanadate, vanadyl sulfate, molybdate, and tungstate at the same range of concentrations does not result in the accumulation of activated Stat1α molecules in the absence of ligand. However, these compounds inhibit the inactivation of interferon-γ-induced DNA-binding activity of Stat1α. A 4-6-h exposure of the permeabilized cells to orthovanadate, molybdate, and tungstate, but not vanadyl sulfate, results in a ligand-independent activation of Stat1α, which is blocked by the inhibition or depletion of NADPH oxidase activity in the cells, indicating that NADPH oxidase-catalyzed superoxide formation is required for the bioconversion of these metal oxides to the corresponding peroxo-compounds. Interestingly, ligand-independent Stat1α activation by peroxo-derivatives of these transition metals does not require Jak1, Jak2, or Tyk2 kinase activity, suggesting that other kinases can phosphorylate Stat1α on tyrosine 701.

Soluble extracellular signaling polypeptides like cytokines and growth factors transmit their signals from cell surface to nucleus through a number of protein messengers (1)(2)(3). Propagation of signals down the pathways can be envisioned as a cascade of physical and chemical changes in these signaling molecules. In the IFN 1 -␥-activated pathway, the signal is initiated by a physical interaction of the cytokine with its receptor on the plasma membrane that leads to structural change(s) in the engaged receptor complex. This in turn triggers tyrosyl phosphorylation of Jak1 and Jak2 kinases resulting in the induction of their catalytic activities, and phosphorylation of a unique tyrosyl residue on the receptor tail that serves as a docking site for the latent Stat1 molecules, Stat1␣ (p91) and Stat1␤ (p84) (1,3,4). The immobilized Stat1 molecules are phosphorylated at tyrosine 701 by Jak activities (1,4,5). The activated Stat1 molecules form homodimers through an intermolecular SH2-domain and tyrosyl-phosphate interactions (1,6). The Stat1 dimers are translocated to the nucleus and bind cognate DNA elements. However, only the Stat1␣ homodimer (GAF) can activate the transcription of the cognate genes (1,7).
Tyrosine phosphorylation is a pivotal reaction that activates IFN-␥ and many other cytokine signals (8). While the downregulation of the signals is less well understood, removal of the phosphate group from the tyrosine residues of the activated signaling proteins could be a potential mechanism for the termination of cytokine signals. A SH2-domain-containing protein-tyrosine phosphatase (PTP), SH-PTP1 (also known as PTP1C/HCP), which is preferentially expressed in hematopoietic cells, acts as a negative regulator of cell signaling by erythropoietin (9, 10), interleukin-3 (11), c-Kit (12), and antigen receptor (13). In contrast, SH-PTP2 (also known as PTP1D/ Syp), another SH2-domain-containing PTP, acts as positive regulator of mitogenic signal transduction pathways (14 -17). Larner and co-workers (18,19) have shown that pre-but not simultaneous treatment of reconstituted cell-free signaling systems with PTP inhibitors blocks the formation of ISGF3 and GAF by IFN-␣ and IFN-␥, respectively, indicating a positive role of PTP in IFN signaling pathways. Based on the observation that pervanadate prevents the inactivation of IFN-induced signals, the involvement of a nuclear PTP in the down-regulation of IFN-induced gene activation has been implicated (20). However, the levels of Stat activation resulting from the individual actions of IFN and pervanadate were not distinguished (20), although it was subsequently demonstrated that incubation of cells with pervanadate causes IFN-independent Stat activation (21,22).
Termination of cytokine or growth factor signals has been recognized as an important cellular event, particularly when a signal acts as an activator of genes that are constitutively silent or expressed at a very low basal levels (23). The precise mechanism of termination of IFN-induced signals and the consequent deactivation of IFN-stimulated genes remained to be resolved. The down-regulation of the IFN-stimulated gene transcription may be a consequence of either the inactivation of transcription factors or the induction/activation of repressor molecules or a combination of both. For example, induction of repressors, like IRF-2 and positive regulatory domain 1 binding factor 1, has been demonstrated in the down-regulation of IFN-␤ gene transcription (24 -26). Some of the IFN-stimulated genes have IRF-1/IRF-2-binding sites in their promoters, which is similar to or same as an IFN-stimulated response element (1,27); however, a role for IRF-2 or any other proteins in the deactivation of IFN-stimulated genes has not been described so far.
To address the roles of PTP activities in regulating ligand (IFN-␥)-dependent and constitutive activation of Stat1 pro-teins, we investigated the kinetics of Stat1␣ activation and deactivation in cells treated with regulators of PTPs. We present evidence that more than one PTP activity is involved in the regulation of both the constitutive and IFN-␥-dependent signal transduction mediated through the tyrosyl phosphorylation of Stat1␣. By using cell lines lacking Jak1, Jak2, and Tyk2, we show that none of these kinases individually is required for the ligand-independent activation of Stat1␣ molecules.

MATERIALS AND METHODS
Oligonucleotides-The wild-type (5Ј-TCGAGCCTGATTTCCCCGA-AATGACGGC-3Ј) and mutant (5Ј-TCGAGCCTGATTTCCCCGACAT-GACGGC-3Ј) GAS, also known as the inverted repeat element of the human IRF-1 gene (28), and the corresponding complementary strands were synthesized with a SalI linker (TCGA) at the 5Ј-ends. The duplex oligonucleotides were used for electrophoretic mobility shift assay as described previously (29).
Anti-Stat1␣ Antibody-Stat1␣-specific antibody was raised in rabbits using a GST-fusion protein encoding the C-terminal 37 amino acids of human Stat1␣ protein (7,30). The GST-fusion protein construct was prepared by ligating a BamHI-linked 118-bp polymerase chain reaction product (encoding C-terminal 37 amino acids of Stat1␣) to the BamHI site of pGEX-3X vector. The fusion protein was expressed in Escherichia coli (DH-5␣) and purified by glutathione-agarose affinity chromatography.
Luciferase Reporter Plasmid Construction-A 1.3-kb SstI fragment of the human IRF-1 gene promoter was excised from pBluescript vector (28) and subcloned into the SstI site of luciferase expression vector, pGL2-Basic (Promega).
Cell Culture-HeLa S3 cells were obtained from ATCC. HeLa and Cos-1 cells were grown in Dulbecco's modified Eagle's medium containing 10% heat-inactivated fetal calf serum and 50 units of penicillin and 50 units of streptomycin/ml of medium. 2fTGH and its mutant-derivative cell lines were grown as described previously (29). Namalwa and HL-60 cells were grown in RPMI 1640 medium containing 10% heatinactivated fetal calf serum and 50 units of penicillin and 50 units of streptomycin/ml of medium. Daudi cells were grown in RPMI 1640 medium containing 10% heat-inactivated fetal calf serum, 1 mM sodium pyruvate, 100 M Dulbecco's modified Eagle's medium nonessential amino acids, 5 mM Hepes buffer, pH 7.5, and 50 units of penicillin and 50 units of streptomycin/ml of medium.
Preparation of Whole Cell Extract (WCE) and Electrophoretic Mobility Shift Assay (EMSA)-Cells were grown to an appropriate density and treated with the indicated agents. WCE was prepared as described previously (29). For EMSA, WCE containing 5-10 g of protein was incubated at room temperature in a 18-l final reaction volume containing 20 mM Na ϩ -Hepes, pH 7.9, 10% glycerol, 80 mM NaCl, 1 mM dithiothreitol, 0.6 mM EDTA, pH 8.0, 4 mM Tris⅐Cl, pH 7.9, 5 mM MgCl 2 , 3 g of poly(dI⅐dC), and 0.2 ng of end-labeled GAS/inverted repeat oligonucleotide duplex probe for 20 min. The protein-DNA complex was resolved on a 5.5% polyacrylamide (acrylamide/bis, 75/2) gel using 0.4 ϫ TBE (1 ϫ TBE ϭ 89 mM Tris borate, 89 mM boric acid, 2 mM EDTA, pH 8.35) at room temperature for 1.5-2.0 h at 0.65 V/cm 2 , and the gel was dried and exposed to film at Ϫ70°C.
Permeabilization of Cells-Cells were washed with phosphate-buffered saline and incubated in permeabilization buffer (10 ml of buffer/ 10-cm plate containing 10 -15 ϫ 10 6 cells) for 30 min at 37°C. The cells were washed 2 times with fresh medium and incubated in fresh medium for indicated length of time. The composition of the permeabilization buffer was essentially the same as described by Bernier et al. (31) except the digitonin was used at a concentration of 10 g/ml.
Preparation of Pervanadate Solution-20 mM sodium orthovanadate was prepared in water and mixed with an equal volume of 20 mM hydrogen peroxide for 15 min at room temperature (32). Sodium molybdate and sodium tungstate were treated similarly with hydrogen peroxide to prepare the peroxo-derivatives of these metals.
Western Blot Analysis-Whole cell extracts containing 10 g of protein were analyzed by 10% SDS-polyacrylamide gel electrophoresis and proteins transferred to Immobilon-P membranes and blocked with 5% carnation milk in TBST (10 mM Tris⅐Cl, pH 8.0, 150 mM NaCl, 0.1% Tween-20) at 4°C overnight. The blot was briefly rinsed 2 times with TBST and incubated with 3% bovine serum albumin for 1 h at room temperature followed by an anti-Stat1␣ antibody in TBST containing 3% bovine serum albumin at room temperature for 3 h. The membrane was washed for 40 min with TBST at room temperature and incubated with a goat anti-rabbit IgG conjugated with horseradish peroxidase in 5% Carnation milk in TBST for 45 min, washed with TBST for 1 h, and developed with ECL reagent (Amersham Corp.).
Transfection of Cos-1 Cells and Determination of Luciferase Activity-Cos-1 cells were transfected with appropriate plasmids using Lipofectin (Life Technologies, Inc.). Briefly, the cells were incubated with OPTI-MEM I (Life Technologies, Inc.) containing DNA-lipofectin complex for 8 h, washed, and incubated with Dulbecco's modified Eagle's medium containing 10% heat-inactivated fetal calf serum and penicillin-streptomycin (50 units/ml) for 16 h before treatment with 50 M pervanadate for 5 h. Luciferase activity was measured by using Promega kit according to the manufacturer's protocol. The reference plasmid expressed the ␤-galactosidase gene driven by Rous sarcoma virus long terminal repeat, and the ␤-galactosidase activity was measured as described elsewhere (33).

RESULTS
Time Course of GAF Inactivation-The GAS-binding activity of Stat1␣ (GAF) in most cell lines was maximal after 30 min of IFN-␥ treatment. When HeLa S3 cells were exposed to IFN-␥ (50 units/ml) for 30 min and the treatment was withdrawn, 40 -50% of the GAF activity remained by 4 -6 h falling to 10 -20% at 8 -10 h, compared with that observed after a 30-min exposure of the cells to IFN (Fig. 1). Under similar conditions of treatment, Namalwa cells exhibited a different time course of GAF inactivation. In these cells, only 40 -50% of GAF activity (compared with 30 min of treatment with IFN-␥) remained at 2 h (Fig. 1). Continuous treatment of HeLa or Namalwa cells with IFN-␥ prolonged GAF activity (data not shown). Like HeLa cells, HT1080 and Jurkat showed longer GAF inactivation time compared with human fibroblast cells WI38 and GM2667 (data not shown). Since the gene encoding the Stat1 proteins is transcriptionally induced by both type I and type II IFNs (34), changes in GAF activity may reflect different levels of Stat1␣ protein. However, in HeLa S3 cells, the level of Stat1␣ increased only by 1.5-2.0-fold after 4 h of IFN-␥ treatment, and in Namalwa cells there was no change in Stat1␣ protein levels (Fig. 2). Thus, the persistence of GAF activity in HeLa cells may be partly attributed to the increase in Stat1␣ protein level, but clearly the inactivation of the DNA-binding activity of Stat1␣ is not due to degradation of the Stat1␣ protein.
Inhibition of protein synthesis by cycloheximide prolongs the induction period of IFN-induced genes (35)(36)(37). To determine whether protein synthesis inhibition altered the inactivation rate of the DNA-binding activity of Stat1␣, GAF activity was followed in the presence and absence of cycloheximide for several hours after HeLa cells were treated with IFN-␥ for 30 min. The rate of inactivation of Stat1␣ was not markedly affected by protein synthesis inhibition (Fig. 3). A similar observation was made with Namalwa cells (data not shown).
Effect of PTP Inhibitors on Stat1␣ Activation/Inactivation-We investigated the effects of commonly used PTP inhibitors, such as vanadate, molybdate, and tungstate on Stat1 phosphorylation. However, since these agents do not efficiently enter intact cells (38 -41), membrane-permeable derivatives of these metal oxides must be used or cells must be permeabilized. The peroxo-derivative of vanadium (perhydroxy-vanadate or pervanadate) prepared by treating sodium orthovanadate with hydrogen peroxide is a very potent inhibitor of PTP activity and readily penetrates cell membranes (38 -44). Pervanadate activates Stat proteins in a ligand-independent manner (Fig. 4). Treatment of cells with pervanadate has previously been shown to result in the accumulation of tyrosyl phosphate in many cellular proteins (33, 38 -44) including Stat1␣ (21,22,45). Treatment of vanadyl sulfate, molybdate, and tungstate with an equimolar concentration of hydrogen peroxide and exposure of intact cells to the resulting mixture fails to induce any detectable DNA-binding of Stat1␣ (Fig. 4). However, when the peroxo-derivatives of these metals were added to permeabilized cells, the molybdate and tungstate, but not the vanadyl derivative, induced the DNA-binding activity of Stat1␣ (Fig. 5). We confirmed that the PTP inhibitor-activated complex contained Stat1␣ and hydrogen peroxide did not induce Stat1␣ activation (data not shown). To determine if pervanadate-activated Stat1␣ is able to activate the transcription of the IFN-␥induced genes, we carried out transient transfection assays in Cos-1 cells by using a reporter construct containing the luciferase coding sequence driven by a 1.3-kb IFN-responsive promoter fragment of the human IRF-1 gene (28). Pervanadate increased the luciferase activity by 2-3-fold when treated with a nontoxic dose of pervanadate for 5 h (Fig. 6). This clearly indicates that the pervanadate-mediated activation of Stat1␣ is biologically functional. Thus, the ligand-independent activation of Stat proteins by pervanadate restricts its use in demonstrating the role of PTP in the deactivation of ligand-induced Stat activity.
A short term (1-2 h) treatment of permeabilized HeLa cells with sodium orthovanadate did not induce DNA-binding activity of Stat1␣ but prevented the inactivation of IFN-␥ induced DNA-binding activity of Stat1␣ (Fig. 7). Like orthovanadate, vanadyl sulfate, molybdate, and tungstate also prevented the inactivation of IFN-␥-activated DNA-binding activity of Stat1␣ molecules (data not shown). Moreover, longer (4 -6 h) treatment of permeabilized cells with molybdate and tungstate but not vanadyl sulfate resulted in the ligand-independent Stat1 activation (data not shown). These results clearly demonstrate that tyrosine dephosphorylation is a mechanism for the inactivation of ligand-activated Stat1␣.
Ligand-independent Stat1␣ Activation Does Not Require

FIG. 2. Stat1␣ level in IFN-␥-treated cells.
Western blot analysis was performed with WCE containing 10 g of protein, derived from HeLa and Namalwa cells as described in Fig. 1. Equivalent amount of WCE prepared from 2fTGH and U3A cells were used as positive and negative controls, respectively, for the detection of Stat1␣ band. The immunoblot was probed with a Stat1␣-specific polyclonal antibody.

FIG. 3. Inactivation time course of IFN-␥-induced GAF in HeLa cells in the presence of cycloheximide (CHX).
EMSA was performed using an end-labeled GAS probe and WCE derived from cells that were treated with IFN-␥ (50 units/ml) for 30 min and then washed as described in Fig. 1. The cells were incubated in fresh medium with or without cycloheximide (50 g/ml) for the indicated length of time minus 30 min. The leftmost lane represents a control of 6 h of cycloheximide treatment.

FIG. 4. GAF activation by PTP inhibitors in intact HeLa cells.
EMSA was performed using an end-labeled GAS probe and WCE derived from cells that were treated with 500 M PTP inhibitor (OV, sodium orthovanadate; VS, sodium vanadyl sulfate; Mo, sodium molybdate; W, sodium tungstate; F, sodium fluoride) for 30 min. Either an aqueous solution of the compound or its peroxo-derivative (prepared by treatment with H 2 O 2 ) was used for the treatment.
Jak1 or Jak2 Activity-Pervanadate prevents dephosphorylation of tyrosine residues in many cellular proteins (22, 32, 38 -45). A role for superoxide, a product of a NADPH oxidasecatalyzed reaction, in the bioconversion of orthovanadate to pervanadate has been suggested (32). To elaborate this possibility, we treated permeabilized HeLa cells with sodium orthovanadate in the presence of diphenylene iodonium (DPI), an inhibitor of NADPH oxidase activity (46). DPI effectively blocked the orthovanadate-mediated Stat1␣ activation (Fig. 8).
Interestingly, the conversion of orthovanadate to pervanadate (as measured by the Stat1␣ activation) was not found in permeabilized HL-60 cells. However, when these cells were induced to differentiate along the granulocyte lineage by Me 2 SO, they supported the bioconversion of orthovanadate to pervanadate (data not shown). Unlike their Me 2 SO-differentiated counterparts, HL-60 cells harbor very little NADPH oxidase activity (47). Like orthovanadate, molybdate and tungstate are also converted to permolybdate and pertungstate when exposed to permeabilized HeLa cells for 6 h (data not shown). These data confirmed that peroxo-derivatives of vanadium, molybdenum, and tungstate are the active species that prevent tyrosine dephosphorylation in Stat1␣ and other Stat proteins. 2 In order to further address the mechanism of the ligandindependent phosphorylation of Stat proteins, we determined whether Jak family kinases are necessary. Both Jak1 and Jak2 activities have been shown to be required for Stat1␣ activation by IFN-␥ (1,48,49), and both are tyrosine-phosphorylated in pervanadate-treated cells (22,45). The mutant cell line, U4A, which does not contain Jak1 (49), supported the pervanadateactivated GAF formation (Fig. 9). Similarly, Daudi cells, which do not respond to IFN-␥ due to the lack of Jak2 activity 3 also induced GAF formation in response to pervanadate. In the U1A cell line, which lacks Tyk2 activity (50), pervanadate-mediated Stat1␣ activation was observed albeit at much lower level. In U2A cells, the absence of p48, another component of IFN sig-

FIG. 5. GAF activation by PTP inhibitors in permeabilized
HeLa cells. EMSA was performed using an end-labeled GAS probe and WCE prepared from the permeabilized HeLa cells. The cells were permeabilized in a hypotonic buffer containing 10 g/ml digitonin for 10 min and then added PTP inhibitor (described in Fig. 5) in the same buffer followed by an incubation for 20 min at 37°C. naling pathways (51), did not affect the action of pervanadate on Stat1␣ activation (Fig. 9). U3A cells, which do not contain Stat1 proteins (34), served as a negative control for pervanadate-mediated Stat1␣ activation. Thus, the peroxo-derivatives of vanadium, molybdenum, and tungsten, which are known inhibitors of PTPs, activate the phosphorylation of Stat1␣ or prevent the dephosphorylation of constitutively activated Stat1␣ in a ligand-independent mechanism that does not absolutely require Jak1 or Jak2 activity. DISCUSSION We have demonstrated that PTP inhibitors orthovanadate and vanadyl sulfate prevent the inactivation of IFN-␥-induced DNA-binding activity of Stat1␣ molecules in permeabilized cells. Molybdate and tungstate having structural similarities with vanadate inhibit PTP activity in cell-free systems (40,52). We have shown that like vanadium salts, molybdate and tungstate also block the inactivation of IFN-␥-activated Stat1␣ molecules in permeabilized cells. In the absence of any stimulation by extracellular signaling proteins, Jak kinases exhibit low levels of constitutive kinase activity in normal cells 4 and can phosphorylate Stat molecules. Recently Heim et al. (53) reported that overexpression of Jak and Stat in Cos cells leads to the ligand-independent phosphorylation of Stat molecules. Orthovanadate, vanadyl sulfate, molybdate, and tungstate (group I PTP inhibitors) do not readily penetrate the cell membrane (38 -44), but when added to permeabilized cells they inhibit the PTP activity that dephosphorylates the ligand-activated Stat molecules but not the constitutively activated Stat molecules. In contrast, the second group of PTP inhibitors namely, peroxoderivatives of vanadium, molybdenum, and tungsten, which seem to be more potent inhibitors of PTPs than the first group, inhibits the PTP activity that dephosphorylates the constitutively activated Jaks and Stats. Accordingly, an accumulation of constitutively activated Stat molecules in the absence of any ligand is detected in the cells that are exposed to the second group of PTP inhibitors.
Based on the observed differential sensitivity of the intracellular PTP activities involved in the Jak-Stat pathways, to the group I PTP inhibitors, we propose a simple model that explains how reversible tyrosine phosphorylation regulates cell signaling through Jak-Stat pathways (Fig. 10). According to this model, the dephosphorylation of the activated Jak molecules is catalyzed by PTP-x, which is insensitive to the micromolar concentrations of group I PTP inhibitors (orthovanadate, vanadyl sulfate, molybdate, tungstate, etc.). On the other hand, the dephosphorylation of activated Stat molecules is catalyzed by a downstream PTP activity, termed PTP-y. Unlike PTP-x, PTP-y is partially sensitive to the micromolar concentrations of group I PTP inhibitors. However, both PTP-x and PTP-y are sensitive to the micromolar concentrations of group II PTP inhibitors (peroxo-derivatives of vanadium, molybdenum, and tungsten).
PTP-x and PTP-y may represent two different families of PTPs rather than two individual enzymes. Recently a SH2domain-containing PTP, SH-PTP1 has been identified as phosphatase that inactivates Jak2 in erythropoietin signaling pathway (9,10). The binding site of SH-PTP1 has been identified in the cytoplasmic domain of the erythropoietin receptor (9). The binding is mediated through an interaction of the SH2 domain of SH-PTP1 and a unique phosphotyrosyl residue (Tyr-429) of erythropoietin receptor (9). Association of SH-PTP1 with other cytokine receptors that activate Jak2, like interleukin-3, has been reported (11). To prevent spurious and nonspecific dephosphorylation of proteins and regulate the availability of specific substrates, subcellular localization of the intracellular PTPs through the interaction of adhesive protein-domains would be a potential mechanism to downregulate the receptorassociated or receptor-intrinsic protein-tyrosine kinase activities (54). Accordingly, we assume that SH-PTPs represent the PTP-x family in our model. It is interesting to note that recombinant SH-PTP1 is not sensitive to micromolar concentrations of group I PTP inhibitors 5 suggesting this or a similar PTP as a candidate for PTP-x.
We have shown that group I PTP inhibitors are converted to group II inhibitors inside the cells with the exception of vanadyl sulfate. It has previously been suggested by Grinstein and colleagues that orthovanadate is converted to pervanadate inside the cells through a superoxide anion-mediated reaction (32). We confirmed this observation and showed that molyb-4 S. J. Haque and B. R. G. Williams, unpublished observation. 5 T. Yi, personal communication.
FIG. 9. Pervanadate activates Stat1␣ in the cells lacking protein-tyrosine kinases involved in IFN signaling pathways. EMSA was performed using an end-labeled GAS probe and WCE prepared from cells that were treated with 500 M sodium orthovanadate (OV) or 500 M pervanadate (PV) for 30 min or left untreated as control. date and tungstate are also converted to their peroxo-derivatives in vivo through superoxide-mediated reaction. Like many other transition metals, vanadium, molybdenum, and tungsten form peroxo-compounds in their highest oxidation states (55), which may explain why vanadyl sulfate (VOSO 4 ) in which vanadium is in oxidation state of ϩ4 does not form pervanadate either when treated with hydrogen peroxide in vitro or through superoxide anion-mediated reaction in permeabilized cells. Notably, in the group I PTP inhibitors that are converted to peroxo-derivatives, the transition metals are in their highest oxidation states.
The peroxo-derivatives of these transition metals, classified as group II PTP inhibitors, have a broad spectrum of action. Treatment of cells with pervanadate results in the accumulation of tyrosyl phosphate in many cellular proteins depending on the cell types. This results from the inhibition of PTP activities, which not only dephosphorylate nonenzyme proteins, e.g. Stats, but also activate protein-tyrosine kinases, e.g. Jaks (45). Thus, the group II PTP inhibitors indirectly activate proteintyrosine kinase activities by preventing the dephosphorylation of unique tyrosyl residues that positively regulate the proteintyrosine kinase activity.
We have shown that cell lines lacking individual Jak kinases, namely Jak1, Jak2, and Tyk2 can support the spontaneous phosphorylation of the unique tyrosyl residue 701 of Stat1␣ molecules in a ligand-independent fashion. These data indicate that Stat1␣ phosphorylation at tyrosine 701 is not a Jak-specific catalysis. This has also recently been demonstrated by others in different experimental contexts (53,56). The pervanadate-mediated activation of Stat1␣ molecules in Jak-minus cell lines rules out the possibility that the constitutive activation of Stat1␣ is due to the action of IFNs secreted by the cells or present in the serum-containing medium. In general the PTPs having 10 -1000-fold higher specific activities than the protein-tyrosine kinases in vitro are expected to prevent the accumulation of phosphate on tyrosyl residues of many cellular proteins resulting from the constitutive activities protein-tyrosine kinases (57). Thus, a strong inhibitory action of group II PTP inhibitors leads to the accumulation of phosphate on the tyrosyl residue of cellular proteins including Stats.
PTP activities play crucial roles in the regulation of constitutive and ligand-dependent signal transduction through Jak-Stat and other pathways. Although we have exploited the differences in relative potencies of different PTP inhibitors to address the roles of multiple PTP activities in Jak-Stat pathways, molecular identification of these PTPs (PTP-x and PTP-y) will be necessary to comprehend the complex regulation of cytokine signal transduction pathways.