STAT3 Is a Serine Kinase Target in T Lymphocytes

Interleukin 2 (IL-2) induces tyrosine phosphorylation of STATs 3 and 5 (signal transducer and activator of transcription). We now show that IL-2 regulation of STAT3 proteins in T cells is a complex response involving activation of two forms of STAT3: 90-kDa STAT3α and an 83-kDa carboxyl-terminal truncated STAT3β. The phosphorylation of STAT proteins on serine residues is also required for competent STAT transcription. A critical serine phosphorylation site in STAT3α is at position 727. In this study we have produced an antisera specific for STAT3α proteins phosphorylated on serine 727 and used this to monitor the phosphorylation of this residue during T lymphocyte activation. Our results show that phosphorylation of STAT3α on serine 727 is not constitutive in quiescent T cells but can be induced by the cytokine IL-2. Interestingly, triggering of the T cell antigen receptor complex or activation of protein kinase C with phorbol esters also induces phosphorylation of serine 727 but without simultaneously inducing STAT3 tyrosine phosphorylation or DNA binding. Hence, the present results show that STAT3 serine phosphorylation can be regulated independently of the tyrosine phosphorylation of this molecule. IL-2 and T cell antigen receptor complex induction of STAT3α serine 727 phosphorylation is dependent on the activity of the MEK/ERK pathway. Previous studies have identified H-7-sensitive kinase pathways that regulate STAT3 DNA binding. We show that H-7-sensitive pathways regulate STAT3 DNA binding in T cells. Nevertheless, we show that H-7-sensitive kinases do not regulate STAT3 tyrosine phosphorylation or phosphorylation of serine 727. These results thus show that STAT3 proteins are targets for multiple kinase pathways in T cells and can integrate signals from both cytokine receptors and antigen receptors.

IL-2-activated transcription factors include members of the STAT (signal transducer and activator of transcription) family: STAT3 and STAT5 (15)(16)(17)(18)(19). STAT activation involves tyrosine phosphorylation which allows for Src homology (SH) 2 domainmediated homodimerization or heterodimerization. Activated and tyrosine-phosphorylated STATs then translocate to the nucleus to form sequence-specific DNA binding complexes and enable cytokine-mediated gene transcription (reviewed in Ref. 20). Serine kinases are critical for STAT activation and can regulate STAT DNA binding or transcriptional activity (21)(22)(23)(24)(25)(26). In reconstitution experiments, it has been recently shown that for STAT1 to elicit IFN-␥-mediated antigrowth activity, the serine phosphorylation site (Ser-727) is required (27). We have also recently shown that IL-2 regulation of STAT5 is mediated by both tyrosine and serine/threonine kinase pathways (28). STAT5 serine phosphorylation in T cells is regulated by an as yet uncharacterized kinase whose activity is required for STAT5 transcriptional activity.
IL-2 induces the tyrosine phosphorylation and DNA binding of STAT3. The phosphorylation of serine 727 in the STAT3 carboxyl terminus is important for STAT3 transcriptional function, but neither has there been any characterization of STAT3 serine phosphorylation nor any study of the identity of STAT3 serine kinases in T cells. Recent debates have focused particularly on the role of MEK/ERK kinase pathways in STAT activation (29). However, the identity of the STAT3 serine 727 kinase has not been fully resolved in any cellular system. This site forms an in vitro substrate for the MAP kinase ERKs which raised the possibility that STAT3 integrates signals from the MAP kinases. This hypothesis has not been fully explored in vivo, in particular, for certain cell systems phosphorylation of serine 727 is constitutive and in others, subject to regulation by extracellular stimuli (25). Herein we present evidence that IL-2 induces tyrosine phosphorylation and DNA binding of two different forms of STAT3: 90-kDa STAT3␣ and 83-kDa STAT3␤. IL-2 also induces serine phosphorylation of serine 727 in STAT3␣ via a MEK/ERK pathway. The present data also show that STAT3 tyrosine and serine phosphorylation can be independently regulated: activation of MAP kinases by phorbol esters or T cell antigen receptor cross-linking induces phosphorylation of STAT3␣ on serine 727 without concurrently inducing STAT3 DNA binding or tyrosine phosphorylation. The phosphorylation of serine 727 is not obligatory for DNA binding or tyrosine phosphorylation of STAT3. Nevertheless, we find a H-7-sensitive kinase pathway that does not regulate STAT3 tyrosine phosphorylation or phosphorylation of serine 727 in STAT3␣ is able to modulate STAT3 DNA binding. These data collectively indicate the role of multiple T cell signaling pathways in mediating differential control over STAT activation.

MATERIALS AND METHODS
Cell Culture-Kit225 cells (30) were maintained in RPMI 1640 with 10% fetal calf serum and 1 nM recombinant IL-2 (Eurocetus). IL-2 * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. receptor expressing human peripheral blood-derived T lymphoblasts were generated and maintained as described previously (31). Cells were quiesced by washing three times in RPMI 1640, and replacing in RPMI 1640 with 10% serum in the absence of IL-2 for 48-72 h. The inhibitors H-7, 1-(5-isoquinolinylsulfonyl)-2-methylpiperazine (Sigma), and PD098059 (a gift of L. O'Neill, Trinity College, Dublin) were preincubated with cells in RPMI 1640 for 30 min prior to cytokine stimulation. Recombinant IL-2 (1 nM), interferon ␣ (1000 units/ml), or phorbol 12,13dibutyrate (50 ng/ml) was incubated at indicated times. Interferon ␣ (IFN␣) was a gift from I. Kerr. Phorbol 12,13-dibutyrate (PDBu) was from Calbiochem.
Affinity Purification of DNA and Proteins-Whole cell extracts were prepared by lysis of approximately 10 -20 ϫ 10 6 cells/ml in a lysis buffer comprised of 50 mM Tris-Cl, pH 8.0, 1% Nonidet P-40, 150 mM NaCl, 0.1 mM EDTA, 10 mM sodium fluoride, 10% glycerol, 1 mM Na 3 VO 3 , 1 mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol, 1 g/ml aprotinin, 1 g/ml leupeptin, 1 g/ml chymostatin. The oligonucleotide sequence used was derived from the high-affinity SIE of the c-fos gene (SIEM67) GTCGACATTTCCCGTAAATC. DNA-binding proteins were isolated from extracts of approximately 15 ϫ 10 6 cells in the above buffer by incubation at 4°C for 1 h with 1 g of double-stranded, 5Ј-biotinylated oligonucleotide coupled to 30 l of a 50% suspension of streptavidinagarose (Sigma). For affinity purification of proteins binding to the phosphopeptide IFN␥R Y440P, the phosphotyrosylpeptide TSFGYD-KPH was coupled to Affi-Gel 10 (Bio-Rad) at 5 mg/ml, and 30 l were used to purify binding proteins from 1 ml of cell lysate. Binding was at 4°C for 1 h, and complexes were washed twice in lysis buffer prior to elution by boiling in reducing sample buffer. Proteins were separated by SDS-PAGE on 7.5% acrylamide, 0.2% bis gels, and transferred to polyvinylidene difluoride membranes (Millipore). Western blot analysis was performed with antibodies to phosphotyrosine (4G10 and PY20; 1 g/ml each; Upstate Biotechnology Inc. and Transduction Laboratories, respectively) or antibodies against STAT proteins; STAT5, STAT3, or STAT1(anti-ISGF3) (Transduction Laboratories unless indicated otherwise). Immunoreactive bands were visualized with the epichemiluminescence Western blotting system (Amersham).
For the analysis of STAT3␣/␤ immunoreactivity on Western blot analyses, the following antibodies were used: amino-terminal (residues 1-175) raised anti-STAT3 N-T fusion protein, carboxyl-terminal (residues 750 -769) raised anti-STAT3 C-T peptide, and carboxyl-terminal (residues 688 -727) raised anti-STAT3 C-T fusion protein (Transduction Laboratories, Santa Cruz, CA, and a gift from I. Kerr, respectively). For the analysis of STAT3 tyrosine phosphorylation by Western blot analysis, antibodies raised to the consensus tyrosine phosphorylation motif were used (residues 701-709); either using a tyrosine phosphopeptide (␣-Py STAT3) or the equivalent non-phosphorylated peptide (␣-STAT3 control). Both antibodies were from New England Biolabs.
Generation of Phosphoserine 727 STAT3 Antisera-The peptide sequence corresponding to human STAT3 at site 717-736, TC-SNTIDLPMSPRALDSLMQ, were synthesized with serine 727 as a phosphopeptide or a control non-phosphorylated peptide. These peptides were then coupled to keyhole limpet hemocyanin via the glutaraldehyde coupling method. Rabbit antisera to these serine 727 phosphopeptide were generated by standard immunization protocols. Rabbit sera was screened for antigen reactivity by standard ELISA and Western blot analysis. On final screening, the phosphoserine 727 antisera was used at a dilution of 1:5000 for Western analysis and at a dilution of 1:100 for immunoprecipitation studies.
Western Blot Analysis of Total Cell Lysates-Cells were lysed in the buffer used for affinity purification experiments, and proteins were concentrated by precipitation with 2.0 volumes of acetone. Proteins isolated from 1 to 3 ϫ 10 6 cells were separated by SDS-PAGE using the following gel conditions: for STAT5, STAT3, Sos, and Raf-1 analysis, 7.5% acrylamide, 0.2% bisacrylamide gels and 15% acrylamide, 0.075% bisacrylamide for MAP kinase analysis. Proteins were transferred to polyvinylidene difluoride membranes, and detected by Western blot with the indicated antisera: for STAT5 and STAT3, pan-Erk antisera (Transduction Labs.), Sos (Upstate Biotechnology Inc.), Raf-1 (Santa Cruz Inc.), or phosphotyrosine STAT3 (New England Biolabs), unless indicated otherwise.

IL-2 Induces DNA Binding of STAT3␣ and STAT3␤ in T
Cells-Biotinylated oligonucleotides comprising the high-affinity mutant sis-inducible element (SIEM) was used to affinity purify STAT proteins from cell extracts of quiescent or IL-2stimulated human peripheral blood-derived T lymphoblasts.
The data in Fig. 1A show that IL-2 induces DNA binding of STAT3 and STAT5 in peripheral blood-derived T cells. IL-2 induced DNA binding of STAT3 and STAT5 is clearly discernible within 1 min of exposure to the cytokine and maximal by 10 min (Fig. 1A). Two forms, a 90-kDa and a 83-kDa form of STAT3 proteins, were seen in both Kit225 cells and peripheral blood-derived T lymphoblasts (Fig. 1, A and B). Previous studies have not noted STAT3 heterogeneity in T cells. However, two different STAT3 proteins of 90 and 83 kDa were detected in the IL-2 STAT⅐DNA complexes in both peripheral blood-derived T cells and Kit225 cells (Fig. 1, A and B). Two different forms of STAT3 have been described: STAT3␣ and STAT3␤; STAT3␤ being a carboxyl-terminal truncated spliced variant of a single STAT3 gene (32,33). To further characterize the 90and 83-kDa IL-2 induced STAT3 proteins (termed hereafter as STAT3␣ and STAT3␤, respectively), antisera raised against different domains of STAT3 were used in Western blot analyses of the STAT3 proteins in total T cell lysates and in DNA binding complexes (Fig. 1B). Additionally, a tyrosine phosphopeptide comprising the tyrosine-phosphorylation site (Tyr-440) of the ␥-interferon receptor ␣-chain was used to generate an affinity matrix for the purification of STAT3 from cell extracts. The data in Fig. 1B show that antisera raised against an NH 2 -terminal STAT3 fusion protein and antisera generated against a fusion protein comprising residues 688 -727 of STAT3 recognize two STAT3 proteins in T cell lysates and in the Tyr-440 peptide complexes. They also recognize both IL-2induced STAT3 proteins in SIEM DNA binding complexes. In contrast, antisera raised against a COOH-terminal STAT3 peptide comprising residues 750 -769 recognize only the 90- Whole cell extracts were then prepared. DNA-binding proteins were affinity purified from whole cell extracts with an agarose-coupled SIEM oligonucleotide. Bound proteins were separated by SDS-PAGE and detected by Western blot with STAT5 or STAT3 antibodies. B, Kit225 cells were left untreated or treated for 10 min with IL-2 (lanes 4, 8, and 12). Cell lysates were then subjected to acetone precipitation or affinity purification with tyrosine-phosphorylated peptides or SIEM oligonucleotides, as indicated. Proteins were separated by SDS-PAGE and subjected to Western blot analysis using antisera raised to an NH 2terminal fusion protein (lanes 1-4), distal COOH-terminal peptide (residues 750 -769) (lanes 5-8), or antisera generated against a proximal COOH-terminal fusion protein (residues 688 -727), of STAT3 (lanes 9 -12). The positions of molecular weight markers are indicated on the left in kDa. kDa STAT3 and not the 83-kDa protein (Fig. 1B). The 83-kDa STAT3␤ protein in T cells thus corresponds to the recently described carboxyl-terminal truncated spliced variant of STAT3 (32,33). Activated and tyrosine phosphorylated STAT proteins were readily detected in STAT3 immunoprecipitates prepared from IL-2 activated T cells using anti-phosphotyrosine antibodies raised to the tyrosine phosphorylation site of STAT3 (Tyr-705) (Fig. 2).
IL-2 Induces Phosphorylation of Serine 727 in STAT3␣-There was a discernible reduction in the electrophoretic mobility of STAT3␣ but not STAT3␤ isolated from IL-2-activated Kit225 cells and in IL-2 activated peripheral blood T lymphoblasts ( Fig. 3A and data not shown). Both forms of STAT3␣ (termed s, slow, and f, fast; for the different electrophoretic mobility shift properties) can equally bind DNA upon IL-2 activation and are both tyrosine-phosphorylated. The ability of STAT3␣ to undergo an electrophoretic mobility shift in response to IL-2 could reflect IL-2-regulated serine phosphorylation of STAT3. There is a serine phosphorylation site at residue 727 in STAT3␣ that is required for transcriptional activation (25). Generation of phosphorylation state-specific antibodies has been described previously (34) and had been shown to be successful in the analysis of site-specific phosphorylations of a variety of proteins (35)(36)(37), including transcription factors (38). Thus to circumvent radiolabeling techniques and to explore whether IL-2 is inducing the phosphorylation of STAT3␣ on serine 727, we generated rabbit antisera selectively reactive to a phosphopeptide corresponding to the serine 727 site in STAT3. The selectivity of this antisera for the phosphorylated serine 727 STAT3 peptide compared with the nonphosphorylated peptide is shown by ELISA (Fig. 3B).
The phosphoserine 727 STAT3 antisera could weakly immunoprecipitate STAT3␣ proteins from quiescent peripheral blood-derived T cells but levels of STAT3␣ immunoprecipitable with the phosphoserine 727 STAT3 antisera were increased markedly in cells lysates isolated from IL-2 stimulated T cells (Fig. 4A). The carboxyl-terminal truncated STAT3␤ protein lacking the serine 727 site could not be immunoprecipitated with the phosphoserine 727 STAT3 antisera from either quiescent or IL-2-activated cells. On Western blot analysis, IL-2 induced STAT3␣ DNA bound proteins were also reactive with the phosphoserine 727 STAT3 antisera (Fig. 4B). The reactivity of the phosphoserine 727 STAT3 antisera with STAT3␣ proteins isolated from IL-2-activated T cells could be blocked by competition with the phosphorylated serine 727 STAT3 peptide but not the nonphosphorylated peptide (data not shown). STAT3 proteins, affinity purified using the IFN␥R Tyr-440 tyrosine phosphopeptide matrix were similarily analyzed for immunoreactivity with the phosphoserine 727 STAT3 antisera. There was a low level reactivity of phosphoserine 727 STAT3 antisera with STAT3 proteins isolated from quiescent cells but this was markedly increased upon IL-2 stimulation (Fig. 4C). No reactivity of phosphoserine 727 STAT3 antisera with STAT3␤ could be detected. These data collectively show that IL-2 induces serine phosphorylation of STAT3␣ on serine 727 in peripheral blood-derived T lymphoblasts.
Serine 727 in STAT3␣ Is Regulated by the MEK/ERK Pathway in T Cells-Xenopus MAP kinase can phosphorylate STAT1 in vitro at the carboxyl terminus consensus MAPK phosphorylation site (25). The equivalent in vitro ERK site in STAT3 is serine 727. To explore directly the role of the MEK/ ERK pathway in STAT3 regulation by IL-2 we used an inhibitor of the ERK stimulatory kinase, MAP kinase kinase (MEK), PD098059. The specificity of PD098059 as a MEK inhibitor has been previously described (39,40). To monitor the effectiveness  1, 4, and 7), IL-2 (lanes 2, 5, and 8), or interferon ␣ (lanes 3, 6, and 9)-treated Kit225 cells and subjected to immunoprecipitation using an antibody raised to residues (750 -769) of STAT3. Proteins were separated by SDS-PAGE and analyzed by Western blot with phosphotyrosine (Tyr(P)-705) STAT3 (lanes 1-3), STAT3 (lanes 4 -6), and STAT5 (lanes 7-9) antibodies. , and subsequently blocked with gelatin. Antisera was then incubated at the indicated dilutions, starting at 1:100, followed with serial 2-fold dilutions. Horseradish peroxidaselinked donkey anti-rabbit antisera was then incubated at 1:1000. Plates were subsequently washed in phosphate-buffered saline-Tween 0.05% and immunoabsorbance was detected using ABTS substrate (Boehringer Mannheim). Plates were read at 405 nm after 30 min incubation. Data shown represents two such experiments expressed as an average. of the MEK inhibitor in T cells, we examined the phosphorylation of ERK induced by IL-2 in T cells pretreated with PD098059. As an additional means to monitor ERK activation in IL-2-treated cells, hyperphosphorylation of ERK substrates were also analyzed. Cellular substrates for ERK in T cells, as in many cells, include the p21 ras effector molecule Raf-1, and the guanine nucleotide exchange protein Sos (41,42). These proteins are both upstream regulators of the MAP kinases, but are phosphorylated by the MAP kinases through feedback signaling mechanisms. Hyperphosphorylated ERK, Raf-1, and Sos have reduced electrophoretic mobility when analyzed by SDS-PAGE. The data in Fig. 5A show that IL-2 regulated hyperphosphorylation of ERK, Sos, and Raf-1 are inhibited by the MEK inhibitor PD098059, demonstrating that this inhibitor is effective at blocking the MEK/ERK pathway in T cells. Analysis of STAT3␣ and STAT3␤ in DNA bound complexes (Fig. 5B) shows that IL-2 induced changes in STAT3␣ electrophoretic mobility were inhibited by the MEK inhibitor PD098059, however, this inhibitor did not prevent IL-2 induced DNA binding of STAT3.
To examine directly the effects of PD098059 on IL-2 induced phosphorylation of serine 727 of STAT3␣ we used the phosphoserine 727 STAT3 antisera. The data in Fig. 5C shows a marked increase in reactivity of the phosphoserine 727 STAT3 antisera with STAT3␣, isolated from IL-2-activated T cells. This increased reactivity was inhibited in STAT3␣ proteins isolated from IL-2-treated T cells that had been preincubated with the MEK inhibitor PD098059. Hence, inhibition of the . Whole cell extracts were prepared and subjected to immunoprecipitation (I.P.) using phosphoserine 727 STAT3 antisera. Proteins were separated by SDS-PAGE, followed by Western blot analysis with STAT3 antibodies. B, whole cell extracts were prepared as above and DNA bound proteins were affinity purified with SIEM oligonucleotides. Proteins were separated on SDS-PAGE and STAT proteins were detected using phosphoserine 727 STAT3 (lanes 1 and 2), STAT3 (lanes 3 and 4), or phosphotyrosine (Tyr(P)-705) STAT3 antibodies (lanes 5 and 6). In Western blot analysis, phosphotyrosine STAT3 antibodies were cross-reactive with tyrosine-phosphorylated STAT5, hence indicated as ␣-phosphotyrosine (␣-pY) STAT. C, whole cell extracts were prepared as above and STAT proteins were affinity purified using agarose-conjugated tyrosine phosphopeptides, as previously. Proteins were separated on SDS-PAGE and followed by Western blot analysis using phosphoserine 727 STAT3 (lanes 1 and 2) or STAT3 antibodies (lanes 3 and 4).  1 and 2) of PD098059 (50 M). Whole cell extracts were prepared and SIEM-binding proteins were affinity purified and subjected to SDS-PAGE and Western blot using STAT3 antibodies. C, Kit225 cells were treated as above and STAT proteins were affinity purified using agarose-conjugated tyrosine phosphopeptide. Proteins were separated on SDS-PAGE followed by Western analysis using phosphoserine 727 STAT3. MEK/ERK pathway prevents the IL-2-induced phosphorylation of serine 727 in STAT3␣.
The Induction of STAT3␣ Phosphorylation on Serine 727 Does Not Require Tyrosine Phosphorylation of STAT3␣-The phorbol ester PDBu stimulates protein kinase C and activates the ERK/MAP kinase pathway independently of IL-2 receptor activation. If STAT3␣ is regulated by MEK/ERK then one possibility is that PDBu, which activates ERK via a protein kinase C/Raf-1 pathway independent of IL-2, could also induce the phosphorylation of serine 727 in STAT3␣. This is assuming that serine phosphorylation of STAT proteins can proceed independently of their tyrosine phosphorylation. The data in Fig.  6A show that PDBu did not induce changes in the electrophoretic mobility of STAT5 but similar to IL-2, PDBu induced changes in the electrophoretic mobility of STAT3␣ but not STAT3␤. Moreover, immunoprecipitation and Western blot analyses of STAT3 proteins showed that PDBu induced STAT3␣ reactivity with the phosphoserine 727 STAT3 antisera (Fig. 6B, panels a and b). PDBu did not induce DNA binding of STAT3 probably because of the failure of phorbol esters to induce STAT3 tyrosine phosphorylation ( Fig. 6C and data not shown). The PDBu data thus show that phosphorylation of STAT3␣ on serine 727 can occur independently of DNA binding or tyrosine phosphorylation.
Phorbol esters are pharmacological activators of the ERK pathway and to examine whether a more physiological signal for ERK activation can induce phosphorylation of STAT3 proteins in the absence of STAT3 tyrosine phosphorylation, we examined the effects of triggering the T cell antigen receptor complex on STAT3. Previous studies have shown that triggering of the T cell antigen receptor (TCR) complex does not induce tyrosine phosphorylation of STAT3 (31). However, the data in Fig. 6D show that STAT3␣ proteins isolated from TCR-activated T cells are reactive with the phosphoserine 727 STAT3 antisera compared with the weak reactivity seen in STAT3␣ proteins isolated from quiescent T cells. These results show that STAT3␣ can be phosphorylated on serine 727 in response to triggering the TCR complex. This effect is prevented by pretreating T cells with the MEK inhibitor PD098059; consistent with a model in which TCR regulation of the phosphorylation of serine 727 being mediated by the MEK/ERK pathway (Fig. 6D).
IL-2 Induced STAT3 but Not STAT5 DNA Binding Is Regulated by a H-7 Serine Kinase Pathway-The data above show that IL-2 regulates a MEK-sensitive serine kinase that is acting on serine 727 in the COOH terminus of STAT3␣. The activity of this MEK-dependent pathway is not necessary or Whole cell extracts were prepared and acetone-precipitated proteins were subjected to SDS-PAGE and Western blot using STAT5 and STAT3 antibodies. B, panel a, Kit225 cells were prepared as above followed by Western analysis using phosphoserine 727 STAT3 (lanes 1-3) or STAT3 antibodies (lanes  4 -6). B, panel b, Kit225 cells were treated as above and whole cell extracts were prepared, followed by immunoprecipitation using phosphoserine 727 STAT3 antibodies. Proteins were separated by SDS-PAGE and detected by Western analysis using STAT3 antibodies. C, Kit225 were treated as above and DNA-binding proteins capable of binding to the SIEM-oligonucleotide were affinity purified and subjected to SDS-PAGE and Western blot using STAT5 and STAT3 antibodies (as indicated). D, quiescent T lymphoblasts were untreated or stimulated with IL-2 (1 nM) (lane 2) or UCHT-1 (1 g/ml) (lanes 3 and 4) for 10 min following a 30-min preincubation in the absence (lanes Whole cell extracts were prepared and STAT proteins were affinity purified using agarose-conjugated tyrosine phosphopeptides. Proteins were separated by SDS-PAGE followed by Western blot using phosphoserine 727 STAT3 antibodies. Positions of molecular weight markers on the left are indicated in kDa. sufficient for induction of STAT3␣ DNA binding ( Fig. 5B and  6C, respectively). Although the MEK inhibitor PD098059 inhibits STAT3␣ serine 727 phosphorylation (Fig. 5C), the data in Fig. 7A show that pretreatment of T cells with the MEK inhibitor does not abrogate IL-2 induced STAT3␣/␤ tyrosine phosphorylation (Fig. 7A, lane 2 versus lane 3). Thus, although the MEK pathway can control STAT3␣ phosphorylation on serine 727, its activity is not required for IL-2 induced DNA binding of STAT3␣ or STAT3␤ or tyrosine phosphorylation. In certain cells, cytokine induction of STAT3 DNA binding can be regulated by a serine kinase pathway sensitive to the serine/ threonine kinase inhibitor H-7 (24). H-7 does not block the activation of the MEK/ERK pathway in T cells (28) but the role of H-7-sensitive kinases in IL-2/STAT3 regulation in T cells had not been examined. We considered the possibility that STAT3 could be regulated by two different serine kinases. We therefore examined the effect of H-7 pretreatment of T cells on IL-2 regulation of STAT3. The data in Fig. 7B shows that IL-2 induced STAT3␣/␤ DNA binding was inhibited in T cells pretreated with H-7. H-7 did not inhibit IL-2-induced tyrosine phosphorylation of STAT3␣/␤ (Fig. 7A, lane 2 versus 7). The effect of H-7 on STAT3 DNA binding showed selectivity since this inhibitor did not prevent IL-2 induced tyrosine phosphorylation of STAT3 or STAT5 (Fig. 7A). Nor was DNA binding of STAT5 proteins affected by H-7 (Fig. 7B) although as described previously, H-7 inhibits the generation of the serine-phosphorylated STAT5 p2 form as shown by the effect of H-7 on IL-2induced electrophoretic mobility shift of STAT5 proteins. (28) (Fig. 7B).
The inhibitor H-7 also did not prevent the formation of hyperphosphorylated STAT3␣ in IL-2-activated cells in contrast to the effects of the MEK inhibitor PD098059 (Fig. 7A lane 7  versus lane 3, respectively). Concurrently, STAT3␣ proteins from IL-2 activated, H-7-pretreated T cells were also reactive with phosphoserine 727 STAT3 antisera (Fig. 7C), demonstrating that H-7 in T cells does not inhibit IL-2-induced phosphorylation of serine 727 in STAT3␣. Thus, H-7 inhibits IL-2 induction of STAT3 DNA bound complexes without preventing tyrosine phosphorylation or phosphorylation of serine 727 in STAT3␣.

DISCUSSION
The present study reveals that IL-2 activates DNA binding of two forms of STAT3 in T cells: 90-kDa STAT3␣ and an 83-kDa carboxyl-terminal truncated STAT3␤. We show moreover that STAT3 proteins are both tyrosine-and serine-phosphorylated in IL-2-activated T cells. Notably, we developed a specific antibody reactive with phosphorylated serine 727 in STAT3␣ and show that STAT3␣ is regulated by a MEK regulated pathway that is not required for STAT3 DNA binding but which specifically targets the site serine 727 required for STAT3␣ maximal transcriptional activity (25). The present results also reveal that phosphorylation of serine 727 in STAT3␣ is not only regulated by IL-2 but can be regulated by the TCR or phorbol esters (Fig. 8). IL-2 activation of STAT3 concurrently induces phosphorylation on serine 727 and STAT3 tyrosine phosphorylation. However, STAT3 phosphorylation on serine 727 in response to TCR ligation or phorbol esters occurs without concomitant induction of tyrosine phosphorylation or DNA binding of STAT3. Accordingly, the serine and tyrosine phosphorylation pathways that target the STATs can be controlled independently by extracellular stimuli. A continual challenge in T cell biology is to understand the mechanisms that regulate the immune specificity of T cell responses. Phosphorylation of serine 727 in STAT3␣ is required for maximal STAT3 transcriptional activity (25). The ability of T cells to respond to TCR triggering by inducing phosphorylation of STAT3 on a residue that is key for STAT3 transcriptional activity reveals that in T cells there is "crosstalk" regulation of the STATs by members of the antigen receptor family. This allows the transcriptional  -7) with IL-2 for 10 min following preincubation with the indicated micromolar concentrations of PD098059 (lanes 3-6) or 200 M H-7. Whole cell extracts were prepared and subjected to acetone precipitation, followed by SDS-PAGE and Western blot analysis with STAT3 and phosphotyrosine (Tyr(P)-705) STAT3 antibodies (both from New England Biolabs). A low exposure is also shown to highlight the electrophoretic mobility shift of STAT3␣, no difference in mobility shift of STAT3␤ was detected. On total lysates, the phosphotyrosine (Tyr(P) 705) STAT3 antibody is also cross-reactive with STAT5 (indicated). B, Kit225 cells were untreated (lanes 1 and 2) or stimulated for the indicated times with IL-2, after a 30-min preincubation in the absence (odd lanes) or presence (even lanes) of H-7 (200 M). SIEM-binding proteins were affinity purified, and detected by Western blot analysis with 4G10/PY20 (phosphotyrosine), STAT5 and STAT3 antibodies. The inhibitor H-7 has been one tool with which to explore STAT serine kinases in a variety of cell systems, particularly, the regulation of STAT3 by H-7-sensitive kinases has been described. Because of these previous experiments with H-7 we were prompted to see whether H-7 could regulate the phosphorylation of serine 727 in STAT3␣. The present study shows that IL-2-induced STAT3 DNA binding can be regulated by a H-7sensitive pathway. Nevertheless, the H-7-sensitive pathway does not regulate STAT3␣ serine 727 phosphorylation. The H-7 target is unknown but is intriguing because of the effects of H-7 on STAT3 DNA binding. The simplest interpretation of the H-7 data is that STAT3␣ integrates signals from two serine kinase pathways in T cells: a MEK-regulated pathway and a H-7sensitive pathway. These pathways are independent as H-7 does not prevent IL-2 activation of the MEK/ERK2 pathway (28) or the phosphorylation of serine 727 in STAT3␣. H-7, however, does inhibit IL-2-activated STAT3 DNA binding in T cells. Conversely, the MEK/ERK2 pathway regulates phosphorylation of serine 727 in STAT3␣ but is not necessary for IL-2 induced STAT3 DNA binding. The regulation of both STAT3␣ and STAT3␤ DNA binding indicates that the H-7-sensitive pathway must regulate a site outside the variant carboxyl termini of the two forms of STAT3. It is possible that through a H-7-sensitive kinase, phosphorylation at, or close to, the DNA-binding domain increases DNA binding affinity of STAT3 proteins. Nevertheless, it is also possible that the H-7 effect on STAT3 proteins is indirect and mediated through association with other transcription factors or co-factors; H-7-inhibited phosphorylation of these associated factors could contribute to a loss in STAT3 DNA binding affinity.
We have shown previously that IL-2 activation of STAT5 involves the convergent action of both tyrosine and serine/ threonine kinases (28). The STAT5 serine kinase is H-7 sensitive and not regulated by MEK and cannot be stimulated by activators of the MAP kinase cascade (28). Herein we show that the STAT3 727 serine kinase is sensitive to a MEK inhibitor and can be stimulated by receptors that activate the MEK cascade independently of STAT tyrosine phosphorylation. The involvement of the MEK pathway clearly distinguishes IL-2 regulation of STAT3␣ and STAT5 and shows how a single cytokine working in the same cell can use distinct serine kinase pathways to regulate different STATs. There has been considerable debate as to the identity of STAT serine kinases. The present results show that regarding the identity of STAT serine kinases may reflect that cytokines can target STATs involving multiple and functionally different serine kinase pathways.
The MEK/ERK pathway is known to have an essential role in the activation responses of mature T cells and in thymocyte differentiation: the transition of thymocytes from double negative to double positive cells requires MEK/ERK2 activity as does positive selection of thymocytes (43,44). Despite the importance of this pathway, few MEK regulated pathways have been identified in T lymphocytes. We now show that serine 727, a residue critical for STAT3 function, is a substrate for the action of the MEK in T cells. Interestingly, serine 727 in STAT3␣ can be phosphorylated by the MEK/ERK pathway not only in response to the cytokine IL-2 but also in response to triggering of the antigen receptor complex. The present results demonstrate serine phosphorylation of residue 727 can occur independently of STAT3␣ tyrosine phosphorylation, indicating that the regulation of STAT3 DNA binding and maximal transcriptional activity are separate processes. These data collectively reveal the complexity of STAT regulation and show that STATs are not only targets for members of the cytokine receptor family in T cells but can also integrate diverse MAP kinase signals notably, signals from antigen receptors. FIG. 8. Schematic representation of the pathways that converge to regulate STAT3 in T cells. During T cell activation, the T cell antigen receptor activates the MEK/ERK pathway. This phosphorylates STAT3␣ on serine 727. Subsequent engagement of the IL-2 receptor results in phosphorylation of tyrosine 705 of STAT3␣ and its SH2-mediated dimerization and induction of DNA binding. IL-2 can also induce the MEK/ERK-mediated phosphorylation of serine 727. A second, unidentified, H-7-sensitive pathway can also regulate STAT3 DNA binding in IL-2-activated T cells.