SHP-2 Is a Dual-specificity Phosphatase Involved in Stat1 Dephosphorylation at Both Tyrosine and Serine Residues in Nuclei

Abstract

Signal transducer and activator of transcription (STAT) proteins are both tyrosine- and serine-phosphorylated, mediating signal transduction and gene regulation. Following gene regulation, STAT activity in the nucleus is then terminated by a nuclear protein phosphatase(s), which remains unidentified. Using novel antibody arrays to screen the Stat1-specific protein phosphatase(s), we identified a SHP-2-Stat1 interaction in the A431 cell nucleus. SHP-2 and Stat1 nuclear localization and their association in response to either epidermal growth factor or interferon-γ (IFNγ) were confirmed by immunofluorescent staining and affinity precipitation assays. The SHP-2 C-terminal region containing protein-tyrosine phosphatase activity interacted with the C-terminal SH2 transcriptional activation domain of Stat1. In SHP-2−/− mouse fibroblast cells, Stat1 phosphorylation at both the tyrosine residue Tyr701 and the serine residue Ser727 by IFNγ was enhanced and prolonged. Consistently, purified GST-SHP-2 dephosphorylated Stat1 at both tyrosine and serine residues when immunoprecipitated phospho-Stat1 or a peptide corresponding to the sequence surrounding Tyr(P)701 or Ser(P)727 of Stat1 was used as the substrate. Overexpression of SHP-2 in 293T cells inhibited IFNγ-dependent Stat1 phosphorylation and suppressed Stat1-dependent induction of luciferase activity. Our findings demonstrate that SHP-2 is a dual-specificity protein phosphatase involved in Stat1 dephosphorylation at both tyrosine and serine residues and plays an important role in modulating STAT function in gene regulation. Signal transducer and activator of transcription (STAT) proteins are both tyrosine- and serine-phosphorylated, mediating signal transduction and gene regulation. Following gene regulation, STAT activity in the nucleus is then terminated by a nuclear protein phosphatase(s), which remains unidentified. Using novel antibody arrays to screen the Stat1-specific protein phosphatase(s), we identified a SHP-2-Stat1 interaction in the A431 cell nucleus. SHP-2 and Stat1 nuclear localization and their association in response to either epidermal growth factor or interferon-γ (IFNγ) were confirmed by immunofluorescent staining and affinity precipitation assays. The SHP-2 C-terminal region containing protein-tyrosine phosphatase activity interacted with the C-terminal SH2 transcriptional activation domain of Stat1. In SHP-2−/− mouse fibroblast cells, Stat1 phosphorylation at both the tyrosine residue Tyr701 and the serine residue Ser727 by IFNγ was enhanced and prolonged. Consistently, purified GST-SHP-2 dephosphorylated Stat1 at both tyrosine and serine residues when immunoprecipitated phospho-Stat1 or a peptide corresponding to the sequence surrounding Tyr(P)701 or Ser(P)727 of Stat1 was used as the substrate. Overexpression of SHP-2 in 293T cells inhibited IFNγ-dependent Stat1 phosphorylation and suppressed Stat1-dependent induction of luciferase activity. Our findings demonstrate that SHP-2 is a dual-specificity protein phosphatase involved in Stat1 dephosphorylation at both tyrosine and serine residues and plays an important role in modulating STAT function in gene regulation. STAT 1The abbreviations used are: STAT, signal transducer and activator of transcription; SIE, c-sis inducible element; PTP, protein-tyrosine phosphatase; SH, Src homology domain; SHP-2, SH2 domain-containing PTP-2; SHP-1, SH2 domain-containing PTP-1; DSP, dual-specificity phosphatase; GST, glutathioneS-transferase; MyBP, myosin-binding protein; EGFR, epidermal growth factor receptor; IFN, interferon; Tyr(P), phosphotyrosine; Ser(P), phosphoserine; PVDF, polyvinylidene difluoride; PBS, phosphate-buffered saline; NTA, nitrilotriacetic acid; FITC, fluorescein isothiocyanate; MEF, mouse embryonic fibroblast; MAP, mitogen-activated protein proteins are latent cytoplasmic transcription factors that can be activated by a variety of polypeptide hormones and cytokines. Activation of STAT proteins requires post-translational modification, i.e.phosphorylation at both tyrosine and serine residues. Upon translocation into the nucleus, these proteins activate specific target genes by binding to regulatory sequences termed c-sis inducible elements (SIE). A variety of protein-tyrosine kinases, transmembranal or non-transmembranal, such as JAK, EGFR, and Src induce STAT tyrosine phosphorylation. Phosphorylation of the single tyrosine residue in the transcriptional activation domain of STAT is essential for STAT dimer formation. STAT serine phosphorylation, on the other hand, requires both ERK-dependent and -independent activation pathways (1Chung J. Uchida E. Grammer T.C. Blenis J. Mol. Cell. Biol. 1997; 17: 6508-6516Google Scholar, 2David M. Petricoin III, E. Benjamin C. Pine R. Weber M.J. Larner A.C. Science. 1995; 269: 1721-1723Google Scholar). Serine phosphorylation is critical for STAT to achieve full transcriptional activity even though it is not essential for STAT nuclear translocation and DNA binding. Thus, phosphorylation at both the positional-conserved tyrosine and serine residues corresponding to Tyr701 and Ser727 of human Stat1 is required for its maximal transcription activity (3Darnell Jr., J.E. Science. 1997; 277: 1630-1635Google Scholar). IFNγ activation of STAT proceeds to phosphorylate both tyrosine and serine residues, and these two events appear to be independent of each other (4Wen Z. Zhong Z. Darnell Jr., J.E. Cell. 1995; 82: 241-250Google Scholar). Jak1 and Jak2 functions are both required for IFNγ to induce Stat1 tyrosine phosphorylation. However, Jak2 has also been reported to be involved in Stat1 serine phosphorylation in response to IFNγ (5Zhu X. Wen Z. Xu L.Z. Darnell Jr., J.E. Mol. Cell. Biol. 1997; 17: 6618-6623Google Scholar). EGFR- activated STATs are also tyrosine- and serine-phosphorylated. Although associated with the EGFR C-terminal, JAKs are not involved in EGFR-mediated STAT activation (6David M. Wong L. Flavell R. Thompson S.A. Wells A. Larner A.C. Johnson G.R. J. Biol. Chem. 1996; 271: 9185-9188Google Scholar, 7Olayioye M. Beuvink A.I. Horsch K. Daly J.M. Hynes N.E. J. Biol. Chem. 1999; 274: 17209-17218Google Scholar). STAT post-translational modification such as tyrosine and serine phosphorylation has been elegantly studied. The precise mechanism of the inactivation, however, has only recently started to be investigated. JAK-specific inhibitors such as SOCS and inhibitors of activated STAT such as PIAS have been discovered (8Chung C.D. Liao J. Liu B. Rao X. Jay P. Berta P. Shuai K. Science. 1997; 278: 1803-1805Google Scholar, 9Endo T.A. Masuhara M. Yokouchi M. Suzuki R. Sakamoto H. Mitsui K. Matsumoto A. Tanimura S. Ohtsubo M. Misawa H. Miyazaki T. Leonor N. Taniguchi T. Fujita T. Kanakura Y. Komiya S. Yoshimura A. Nature. 1997; 387: 921-924Google Scholar, 10Naka T. Narazaki M. Hirata M. Matsumoto T. Minamoto S. Aono A. Nishimoto N. Kajita T. Taga T. Yoshizaki K. Akira S. Kishimoto T. Nature. 1997; 387: 924-929Google Scholar). STAT inactivation includes both maintaining STAT latency and terminating activated STAT by dephosphorylation. Dephosphorylation, however, may represent one of the most important mechanisms by which cells negatively control STAT activity. Treatment of cells with protein-tyrosine phosphatase (PTP) inhibitors such as vanadate leads to STAT activation, indicating that PTP(s) is required for maintaining STAT in its inactive form. Upon cytokine or growth factor stimulation, STATs are rapidly phosphorylated at both tyrosine and serine residues and bind DNA, which is followed by STAT dephosphorylation and dissociation from DNA. Phosphatases therefore control STAT activity at two levels: (i) maintaining STAT in its latent form by blocking phosphorylation and (ii) terminating STAT activity by removing available phosphates. This suggests that the balance between protein kinases and protein phosphatases play an important role in modulating the activity of STATs in vivo (11Ram P.A. Waxman D.J. J. Biol. Chem. 1997; 272: 17694-17702Google Scholar, 12Shuai K. Stark G.R. Kerr I.M. Darnell Jr J.E. Science. 1993; 261: 1744-1746Google Scholar). The protein phosphatases responsible for STAT dephosphorylation in nucleus has yet to be confirmed (13Haspel R.L. Darnell Jr., J.E. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 10188-10193Google Scholar). The present study was initiated to identify the protein phosphatase(s) responsible for STAT dephosphorylation in nuclei. By employing our novel antibody array screening for protein-protein interaction, we identified SHP-2 and Stat1 association in A431 cells. Treatment of A431 cells with either EGF or IFNγ led to Stat1 and SHP-2 accumulation in nucleus where these two proteins interact with each other within their C-terminal regions. We provide evidence here that SHP-2 is a dual-specificity protein phosphatase involved in Stat1 dephosphorylation at both Tyr(P)701 and Ser(P)727 residues. The cDNA of full-length SHP-2 (amino acid residues 1–593), N-SH2 domain (residues 1–107), C-SH2 domain (residues 108–220), N-SH2-C-SH2 (residues 1–220), C-SH2PTP (residues 108–593), or PTP domain (residues 221–593) was inserted at the 3′-end and in-frame to the GST open reading frame in the expression vector pGEX-3T (Amersham Biosciences). Bacterially expressed GST fusion proteins were purified using glutathione-agarose beads (Sigma) and dialyzed overnight. Human Stat1 cDNA of full-length (residues 1–750), C-terminal (residues 489–750), N-terminal (residues 1–488), and SH2-TA domain (residues 577–750) were tagged with His6by inserting them into the pET28A vector. Expression of recombinant His6-Stat1 proteins was confirmed by immunoblotting using both anti-His6 and and control mouse M. Mol. Cell. Biol. 1999; were in with A431 cells, 293T cells, and both SHP-2−/− and control mouse were in with cell were by using precipitation PBS, with inhibitors each of and and as M. Y. Science. 1996; 272: Scholar, T. Mol. Cell. Biol. Scholar). The cytoplasmic was by cells in with and the inhibitors a the were as cytoplasmic The nuclear was by the in for and the were at for M. K. Flavell Mol. Cell. Biol. 1997; 17: Scholar). were at of antibody arrays was as T. Mol. Cell. Biol. Scholar). and protein phosphatases were from and were with for at which binding of other proteins in the antibody arrays were with cell for an at the antibody arrays were with antibody followed by (Amersham was from SHP-2−/− by using the antibody immunoprecipitated phospho-Stat1 was then with GST-SHP-2 in the and for at followed by in for The were then in an were used as the a was to as H. U. M.J. J. Biol. Chem. 1997; 272: Scholar, J. Biol. Chem. Scholar). in with or was with of or GST control in for at were terminated by of of H. U. M.J. J. Biol. Chem. 1997; 272: Scholar). for at was using a at for of were by with a of The of is as of of The serine dephosphorylation was to the Proc. Natl. Acad. Sci. U. S. A. was by phosphorylation on serine residues with the in the of to the The was terminated by acid to the and the was purified from by with acid and serine phosphatase activity was then by of MyBP, of and and of purified GST fusion protein as for at The was by of acid and on for at for The containing was then in peptide was at the serine residue with with the from a serine peptide was with of GST fusion proteins at for followed by precipitation with The were then for and immunoblotting were as K. Stark G.R. Kerr I.M. Darnell Jr J.E. Science. 1993; 261: 1744-1746Google Scholar, T. Mol. Cell. Biol. Scholar, M. K. Flavell Mol. Cell. Biol. 1997; 17: Scholar). cells on the were with EGF or IFNγ for were for in and with blocking in for to the on the were then with or antibody for cells were with antibody for an and with a binding cell were with of beads for at Following precipitation by were with proteins were by and to for immunoblotting with from A431 cells were also with of for and on an as M. K. Flavell Mol. Cell. Biol. 1997; 17: Scholar). the Stat1-dependent 293T cells were in with the The the cells were with and were using to the the expression and the which two of the Stat1 T. Mol. Cell. Biol. Scholar). the cells IFNγ an the cells were in and in luciferase The were for luciferase activity using an the of STAT protein we A431 cells with EGF and IFNγ for and activated Stat1 proteins were immunoprecipitated using and in both EGF and IFNγ Stat1 phosphorylation on both Tyr701 and Ser727 residues. Stat1 was in the cells with in the cells with the serine dephosphorylation at Ser(P)727 was with tyrosine dephosphorylation at Tyr(P)701 residue tyrosine phosphorylation is for we STAT DNA binding activity in response to EGF and in A431 cells both EGF and IFNγ in by using as the This is with the Stat1 tyrosine phosphorylation in where an was even In to was that EGF and IFNγ to the Stat1 identify the protein phosphatase(s) responsible for STAT we an antibody array by and protein phosphatases and inhibitors of the on a T. Mol. Cell. Biol. Scholar). or A431 cell were with these antibody arrays for followed by immunoblotting in were in with protein phosphatases and SHP-2 a interaction between Stat1 and these protein phosphatases in A431 cells. in the arrays and However, it that SHP-2 and Stat1 association at of was the interaction with antibody array and immunoblotting were Although a interaction between Stat1 and SHP-2 was in the of A431 cells, the interaction between these two proteins was enhanced in the from the cells both EGF and IFNγ and was most with EGF Thus, SHP-2 may Stat1 proteins by interaction with them in cells. the where SHP-2 with we the localization in cells. A431 cells were with and in and Stat1 nuclear were SHP-2 in both and to antibody the in both and cells, cytoplasmic staining also be We then the of these two proteins in nuclear EGF both SHP-2 and Stat1 nuclear within a of to SHP-2 nuclear translocation has been at and a of SHP-2 in was EGF for nuclear phospho-Stat1 was for at the SHP-2 nuclear was to to the of In the cell SHP-2 nuclear localization was that was not on not Consistently, protein were in the nuclear of that of A431 cells EGFR and Jak1 were from SHP-2 in the Jak2 was not to with SHP-2 not Although SHP-2 can STAT activation activity in as reported M. Mol. Cell. Biol. 1999; the that SHP-2 may Stat1 activity in nuclei. the of SHP-2 and Stat1 involved in their interaction, we a of GST recombinant fusion proteins of SHP-2 either full-length or with as in A. of GST-SHP-2 fusion proteins to beads were with A431 cell and proteins with the GST fusion proteins by immunoblotting in and GST-SHP-2 of containing the C-terminal region (residues Stat1. The N-terminal region of SHP-2 (residues SH2 is important for both protein-protein interaction and C-terminal PTP regulatory M.J. S. T. Nature. 1996; Scholar, P. S. S. M.J. Cell. Scholar, H. K. M. T. Mol. Cell. Biol. 1996; Scholar). However, our that these SH2 these two SH2 residues was to Stat1 A signal was indicating that a interaction between Stat1 and domain of SHP-2 that the SHP-2 C-terminal domain (PTP) its interaction with Stat1. We to the of Stat1 important for its interaction with This was using in full-length or as in B. recombinant Stat1 proteins were with A431 cell followed by SHP-2 antibody SHP-2 were by immunoblotting with anti-His6 in that full-length the C-terminal region of residues and the SH2-TA domain of residues 577–750) of the N-terminal region of Stat1 (residues was present in the SHP-2 SHP-2 was to interact with and other T. R. J. Biol. Chem. 1997; 272: Scholar). the interaction between the of these two proteins in A and a interaction or was by proteins such as we an in protein-protein interaction Bacterially expressed and were with GST, and full-length GST-SHP-2 followed by with in the of form with both full-length GST-SHP-2 and not with with the in A the interaction between and or full-length GST-SHP-2 was not The of of Stat1 in interaction with SHP-2 was then in 293T Stat1 with or interact with SHP-2 Stat1 with a from to was with which the domain (residues the critical of the DNA binding domain (residues and the essential of the SH2 domain from were can be in form of Stat1 interacted with in vivo findings were with the in that Stat1 and SHP-2 interaction within their C-terminal to their SH2 A interaction between the SHP-2 domain and Stat1 containing both and residues required for transcription activity suggests that SHP-2 can Stat1. SHP-2 Stat1 was in SHP-2−/− and control A a of Stat1 tyrosine phosphorylation in response to IFNγ in these cells. in SHP-2−/− cells, Stat1 dephosphorylation was or tyrosine the cells were with the tyrosine for an Stat1 tyrosine phosphorylation in was enhanced and in SHP-2−/− In to IFNγ only a Stat1 tyrosine phosphorylation in the cells. with a M. Mol. Cell. Biol. 1999; Scholar). Stat1 serine phosphorylation at Ser727 by IFNγ was also enhanced and in SHP-2−/− Stat1 tyrosine and serine were in SHP-2−/− cells a role of SHP-2 on Stat1 SHP-2−/− and control cells were with IFNγ for and then activity was by the was for and was only in SHP-2−/− cells. In control cells, phospho-Stat1 was the for the However, an or phospho-Stat1 by at and the signal within of We then Stat1 DNA binding activity in SHP-2−/− with control from these were with Stat1 precipitation by was then confirmed in in SHP-2−/− cells, Stat1 DNA binding activity was enhanced and in control cells, Stat1 DNA binding activity was by that in these cells with SHP-2, STAT activation by was enhanced or M. Mol. Cell. Biol. 1999; Scholar). that SHP-2 Stat1 protein phosphorylation in cells at both tyrosine and serine residues. Jak1 tyrosine phosphorylation in SHP-2 cell has been reported M. Mol. Cell. Biol. 1999; Scholar). The and Stat1 activation here be the of enhanced Jak1 activity. We the and nuclear from the SHP-2−/− embryonic with IFNγ for in phosphorylated Jak1 was in both and nuclear Jak1 phosphorylation in response to IFNγ for was enhanced in the Jak1 tyrosine phosphorylation was and in the was also and Although both were in nucleus, Jak1 and phospho-Stat1 not form Thus, Jak1 activation in in SHP-2−/− cells was responsible for Stat1 activation in these cells. We then the of STAT in SHP-2 cell was in both SHP-2−/− and control EGF led to that a PTP other SHP-2 was responsible for was not in both SHP-2−/− and control cells in response to EGF and were phosphorylated in SHP-2−/− and control of the of these cells with the J. Biol. Chem. Scholar). the activity of SHP-2 on we protein phosphatase cell of were with of GST-SHP-2 or proteins were in A Stat1 dephosphorylation at Tyr701 in GST-SHP-2 in a We then used immunoprecipitated phospho-Stat1 as the to and were SHP-2 serine dephosphorylation on as in immunoblotting with antibody In immunoprecipitated from A431 cells was by GST-SHP-2 the and were used as the Stat1 not was dephosphorylated by GST-SHP-2 a PTP GST-SHP-2 activity in Stat1 at both Tyr701 and Ser727 residues the that protein-tyrosine phosphatase activity was involved in the in The of Stat1 as a of SHP-2 was also by using the Tyr701 of Stat1 as in a were with and was to the was at as H. U. M.J. J. Biol. Chem. 1997; 272: Scholar, J. Biol. Chem. Scholar). not GST-SHP-2 with the or GST control from as by the in at both SHP-2 and the sequence with most dual-specificity phosphatases T. Proc. Natl. Acad. Sci. U. S. A. 1997; Scholar, J.E. J. Biol. Chem. Scholar). PTP and an essential the was in region between these two of the the Stat1 was used as the GST-SHP-2 was of with is for most dual-specificity protein phosphatases that their serine dephosphorylation are with their tyrosine dephosphorylation Proc. Natl. Acad. Sci. U. S. A. Scholar). that SHP-2 a phosphatase was was at serine residues with by protein and was used as a substrate. Upon of with GST-SHP-2 not GST control a of was in and was both and GST control protein GST-SHP-2 with the on from which the of a serine phosphatase activity in GST fusion with a J. Biol. Chem. 1993; a of serine dephosphorylation was in response to GST-SHP-2 when was used as the peptide containing the Stat1 Ser727 residue was and with by with GST-SHP-2 for were and in not the or GST control associated with the beads findings that SHP-2 is a that can both and residues. Stat1 to DNA and gene provide evidence for SHP-2 on of Stat1 the of SHP-2 on luciferase activity was luciferase was with the either SHP-2 or SHP-2 with the serine a inactive form J.E. J. Biol. Chem. Scholar). IFNγ luciferase activity in control cells in which an vector was with the luciferase vector However, luciferase activity was when luciferase was with vector not with Although both Stat1 and are of the of SHP-2 on tyrosine phosphorylation in response to IFNγ in SHP-2−/− mouse fibroblast M. Mol. Cell. Biol. 1999; Scholar). Consistently, tyrosine phosphorylation and its Stat1 activity in response to IFNγ were not in 293T cells with SHP-2 However, nuclear phospho-Stat1 in 293T cells with SHP-2, not was Jak2 was reported as the of SHP-1, an of SHP-2 (11Ram P.A. Waxman D.J. J. Biol. Chem. 1997; 272: 17694-17702Google Scholar, H. K. M. T. Mol. Cell. Biol. 1996; Scholar). In Jak2 activity was by SHP-2 M. Mol. Cell. Biol. 1999; Scholar, H. J. Scholar). In 293T cell however, the was these that SHP-2 can nuclear Stat1 and Stat1 activity for gene regulation. on the we that SHP-2 is a that Stat1 at both the Tyr(P)701 and Ser(P)727 residues. that SHP-2 is involved in the of STAT M. Mol. Cell. Biol. 1999; Scholar, T. R. J. Biol. Chem. 1997; 272: Scholar, M. Pine R. J.E. Larner A.C. J. Biol. Chem. 1996; 271: Scholar). Our that EGF or IFNγ stimulation, both Stat1 and SHP-2 proteins into where may with each SOCS protein phosphatases can STAT in an inactive form by protein-tyrosine activity M. Mol. Cell. Biol. 1999; Scholar). In SHP-2−/− Jak1 tyrosine phosphorylation was in response to which led to a between Jak1 and Stat1. However, tyrosine phosphorylation of by was in SHP-2−/− M. Mol. Cell. Biol. 1999; in 293T cells immunoprecipitated Stat1 can be dephosphorylated by GST-SHP-2 in these SHP-2 in phospho-Stat1 Thus, SHP-2 Stat1 activity at at two blocking tyrosine activity and removing the from full activation of STAT to be both tyrosine- and serine-phosphorylated. Serine in nucleus a of STAT activity in SHP-2 was of Stat1 at both and residues in with the that in SHP-2−/− Stat1 tyrosine and serine were both prolonged. and Jak2 were to STAT serine phosphorylation, (1Chung J. Uchida E. Grammer T.C. Blenis J. Mol. Cell. Biol. 1997; 17: 6508-6516Google Scholar, 2David M. Petricoin III, E. Benjamin C. Pine R. Weber M.J. Larner A.C. Science. 1995; 269: 1721-1723Google Scholar, X. Wen Z. Xu L.Z. Darnell Jr., J.E. Mol. Cell. Biol. 1997; 17: 6618-6623Google Scholar). activity was in SHP-2−/− cells, a of SHP-2 on J. Biol. Chem. Scholar, M. M. Mol. Cell. Biol. Scholar). A of SHP-2 on Jak2 activity was also reported H. J. Scholar, J. Biol. Chem. Scholar). the of SHP-2 on Stat1 was by these proteins in the nucleus. most the activity of SHP-2 in serine dephosphorylation was with its activity in tyrosine dephosphorylation. This is to the of SHP-2, and which is to be to a residue and it may Stat1 serine phosphorylation was in vivo and that enhanced tyrosine phosphorylation and DNA binding activity were with either or STATs K. Liao J. Mol. Cell. Biol. 1996; Scholar, R. R. S. C. Liu X. L. A. B. Mol. Cell. Biol. 1996; Scholar). However, our in protein-protein interaction that SHP-2 associated with Stat1 only within its C-terminal region containing the SH2-TA domain and both Tyr701 and Ser727 residues. interaction between SHP-2 and the N-terminal region of Stat1 was In of these the STAT N-terminal domain may be important in STAT tyrosine phosphorylation by as a phosphatase The N-terminal SH2 domain of SHP-2 was reported to with growth factor M.J. S. T. Nature. 1996; the SHP-2 PTP domain with kinases Jak1 and U. U. Cell. 1995; Scholar). the C-terminal region of SHP-2 is critical for its interaction with both and STAT The PTP domain enhanced activity in in with the that the domain negatively the activity of SHP-2 P. S. S. M.J. Cell. Scholar). The interaction between the domain of SHP-2 and the C-terminal region of Stat1 was not on Stat1 phosphorylation on Tyr701 or Ser727 both the and Stat1 interact with SHP-2 tyrosine phosphorylation on tyrosine residues and in its C-terminal PTP domain cytokine T. R. J. Biol. Chem. 1997; 272: Scholar). Although the Stat1 with the C-terminal region of SHP-2, domain is not for such an Stat1-dependent gene expression the of M. Y. Science. 1996; 272: Scholar, Wen Z. Darnell Jr., J.E. Proc. Natl. Acad. Sci. U. S. A. 1996; Scholar, Z. C. J. Darnell Jr., J.E. Scholar). SHP-2 as as SHP-1, both and on cell growth have been reported Proc. Natl. Acad. Sci. U. S. A. Scholar, U. U. Cell. 1995; Scholar, L. Z. P. J. Biol. Chem. 1996; 271: Scholar, P. M. T. J. Biol. Chem. Scholar). However, the growth of SHP-2−/− cells was that of control cells In 293T cells with SHP-2 the cells with an vector which was with a M. M. Mol. Cell. Biol. Scholar). Thus, SHP-2 cell growth by factors such as Stat1. array screening a of protein-protein interaction, both and in cells The of the screening from antibody array on the of the on the The from the antibody array screening that Stat1 may have the to interact with phosphatases in cells. both and are and were to interact with in the array to be confirmed phosphatases as in the antibody array are involved in STAT dephosphorylation. a nuclear of the PTP has been identified in cells to both Stat1 and J. M. Y. M. M. Shuai K. Mol. Cell. Biol. Scholar). of evidence that is also involved in the of the activity. and SHP-2 have been to interact with and may their M. Mol. Cell. Biol. 1999; Scholar, H. U. M.J. J. Biol. Chem. 1997; 272: Scholar, M. S. Larner A.C. Mol. Cell. Biol. 1995; Scholar). In either cells or cells SHP-2, and and STAT phosphorylation by IFNγ was reported M. Mol. Cell. Biol. 1999; Scholar, H. K. M. T. Mol. Cell. Biol. 1996; Scholar, M. S. Larner A.C. Mol. Cell. Biol. 1995; Scholar). A dephosphorylation of SHP-2 on was not in SHP-2 was to binding to and its activation by T. K. T. K. Y. T. S. M. H. T. M. H. M. T. Scholar). it was reported that be dephosphorylated by SHP-2 in J. Biol. Chem. Scholar). However, SHP-2−/− cells in the tyrosine phosphorylation EGF a dephosphorylation in these SHP-2−/− mouse a activity other SHP-2 was responsible for dephosphorylation in these cells. STAT proteins be dephosphorylated can be for activation by protein The activity SHP-2 to from both tyrosine and serine residues of STAT This an important mechanism by which cells STAT activity. We for reading and critical on The was from SHP-2 and SHP-2 were by A. M. and B. and were from The was from J. E.