Syp associates with gp130 and Janus kinase 2 in response to interleukin-11 in 3T3-L1 mouse preadipocytes.

Protein tyrosine phosphorylation and thus dephosphorylation are part of the interleukin (IL)-11 response in mouse 3T3-L1 cells. We report here for the first time the involvement and interactions of the SH2-containing protein tyrosine phosphatase Syp in the IL-11 signal transduction pathway. Addition of IL-11 to 3T3-L1 cells resulted in an increase in the tyrosine phosphorylation of Syp. When cell lysates were precipitated with glutathione S-transferase fusion products of Syp, the C-terminal SH2 domain of Syp was shown to precipitate several proteins of 70, 130, 150, and 200 kDa that were tyrosine phosphorylated in response to IL-11. Reciprocal immunoprecipitation experiments showed that Syp was inducibly associated with both gp130 and Janus kinase 2 (JAK2). A phosphopeptide containing the sequence for a potential Syp binding site (YXXV) was used to compete with the associations of Syp with gp130 and JAK2. The phosphopeptide reduced the Syp association with both gp130 and JAK2. To summarize, Syp has multiple interactions in IL-11 signal transduction. In addition to the IL-11-induced tyrosine phosphorylation of Syp, Syp coprecipitated with gp130, JAK2, and other tyrosine-phosphorylated proteins in response to IL-11. These findings may have extensive significance to IL-11 and related cytokine signal transduction, suggesting new pathways and mechanisms.

IL-11 1 is a stromal fibroblast-derived cytokine involved in lymphopoiesis and hematopoiesis and is also important in proliferation and differentiation of a variety of cell types (1). IL-11 binds to a specific receptor, which requires the signal transducer gp130 for transmission of IL-11 signals into the cell (2). gp130 is the common signal transducer for IL-11 (3), ciliary neurotrophic factor (4), leukemia inhibitory factor (4), IL-6 (5), and oncostatin M (6). It has been shown that gp130 associates with Janus kinase 2 (JAK2) constitutively and that JAK2 is activated upon IL-11, IL-6, ciliary neurotrophic factor, oncostatin M, or leukemia inhibitory factor stimulation of cells (7,8). JAK2 is a 130-kDa tyrosine-phosphorylated kinase whose response to IL-6 is blocked by point mutations or deletions of the membrane proximal portion of gp130 (9).
Tyrosine phosphorylation and dephosphorylation is a common means for cytokine-mediated intracellular signal transduction. Induction with IL-2 (10), IL-6 (11), IL-7 (12), or IL-11 (13) results in the activation of kinases leading to protein tyrosine phosphorylation. IL-6 also stimulates tyrosine dephosphorylation by unspecified phosphatases leading to the reduction of tyrosine phosphorylation (14). Recently, it was reported that ciliary neurotrophic factor can induce the tyrosine phosphorylation of the protein-tyrosine phosphatase Syp (also named SH-PTP2, PTP1D, and PTP2C) (15). Syp is a recently discovered protein-tyrosine phosphatase containing two SH2 domains (16,17). Syp is ubiquitously expressed and found to be involved in platelet-derived growth factor (PDGF) and epidermal growth factor signal transduction by direct association with the activated receptors (18 -20). Syp is also involved in insulin receptor substrate 1 signal transduction (21)(22)(23). In addition, Syp functions as an adaptor protein for the signal transduction from PDGF through GRB2-Sos, leading to the activation of the Ras signaling pathway (24).
In this study, we identify and characterize a new member of the IL-11 signal transduction mechanism, Syp, and determine its interactions with established components of this pathway, gp130, and JAK2. Both gp130 and JAK2 have potential binding sites for Syp (25,26). We suggest that the interactions of Syp and these components are integral parts of the IL-11 signaling pathway.
Coimmunoprecipitations-3T3-L1 mouse preadipocytes were grown to confluence on 150 ϫ 25-mm dishes in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, 100 units/ml penicillin, 100 g/ml streptomycin, and 2 mM L-glutamine (pH 7.2) (complete Dulbecco's modified Eagle's medium). No IL-11 or 500 ng/ml IL-11 was added as indicated under "Results" at 37°C in the presence of 10% CO 2 . * This work was supported by U. S. Public Health Service Grants R01HL48819 and R01DK43105 (to Y.-C. Y.). 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.  1 The abbreviations used are: IL, interleukin; PDGF, platelet-derived growth factor; GST, glutathione S-transferase; JAK2, Janus kinase 2.
Reactions were stopped by placing the dish of attached cells in ice water and adding ice-cold phosphate-buffered saline containing 2 mM sodium orthovanadate to the cells. Cells were scrapped off the plate, washed in ice-cold phosphate-buffered saline, and then centrifuged at 500 ϫ g for 5 min at 4°C. Cells were resuspended in 0.5 ml of digitonin lysis buffer, vortexed, and incubated on ice for at least 10 min. Insoluble material was removed by centrifugation. The remaining lysate was used for immunoprecipitations with the appropriate antibodies for 12-14-h incubations. Immunoprecipitations of Syp to determine if IL-11 induced the tyrosine phosphorylation of Syp were performed as described above. Densitometry for quantitation of immunoblots was completed with a Bio-Rad GS-670 imaging densitometer coupled to a power Macintosh 8100/80.
Peptide Competition-The peptide YSTV and phosphopeptide VEY*STVVHS were prepared by a stepwise Fmoc (N-(9-fluorenyl)methoxycarbonyl) synthetic strategy as described (27), and their purity was verified by high performance liquid chromatography. The peptides were added immediately after addition of digitonin lysis buffer or to cells 5 min before IL-11 addition, in case any adaptation to the peptide was necessary, and freshly added after washes and with lysis buffer so that the peptide was continuously present at 1 M until washed out at the end of the immunoprecipitation.

IL-11-induced Tyrosine Phosphorylation of Syp-Syp has not
been previously shown to be involved in the signaling pathway of IL-11. An assay for the IL-11-induced tyrosine phosphorylation of Syp would suggest its involvement in the IL-11 signal transduction pathway. Therefore, we tested whether IL-11 induces the tyrosine phosphorylation of Syp. In this analysis, 3T3-L1 cells were stimulated with IL-11 for various amounts of time so that peak phosphorylation could be seen. Increased tyrosine phosphorylation of Syp with IL-11 addition was seen in as little as 30 s (data not shown) and peaked at 3 min (Fig.  1A). This suggests a relatively early involvement for Syp in the IL-11 signal transduction pathway. Dose response experiments showed tyrosine phosphorylation of Syp increasing in a dosedependent manner peaking at 100 ng/ml IL-11 (Fig. 1B).

C-terminal SH2 Domain of GST-Syp Fusion Proteins Associated with Tyrosine-phosphorylated Proteins in IL-11-stimulated Cells-Since
Syp apparently responds to IL-11, GST-Syp fusion constructs were prepared to study potential Syp-associated proteins. To determine if IL-11 induced association of specific SH2 domains of Syp with tyrosine-phosphorylated proteins in 3T3-L1 cells, the following procedures were performed. GST-Syp fusion proteins containing the N/C-terminal SH2 domains and the separate N-or C-terminal SH2 domains were purified from Escherichia coli as described (16). Fig. 2 shows the purified GST-Syp proteins of the expected molecular weights and a GST control. Upon precipitation of 3T3-L1 lysates with GST-Syp fusion proteins, several tyrosine-phosphorylated protein bands were seen when the Western blot was probed with an antiphosphotyrosine antibody (Fig. 3). With IL-11 stimulation, tyrosine-phosphorylated protein bands of apparent molecular masses of 200 and 150 kDa associated with the N/C-and C-terminal SH2 constructs of GST-Syp (Fig. 3,  lanes 4 and 6). These protein bands were not seen or were seen only weakly with unstimulated cells, a GST control, the Nterminal SH2-GST-Syp fusion protein (Fig. 2), and unrelated SH2 domains (data not shown). Other protein bands induced by IL-11 and associated with the C-terminal SH2 domain construct of GST-Syp have molecular masses of approximately 70 and 130 kDa (Fig. 3, lane 4). Similar results were seen with equimolar GST-Syp constructs or when 1% Triton-X 100 was used instead of 1% digitonin (data not shown).
Immunoprecipitations with Anti-gp130 and Anti-JAK2 Showed Coprecipitation of Syp-GST-Syp precipitated proteins of molecular weights potentially corresponding to the IL-11 signal transduction proteins of gp130 and JAK2. To determine if these proteins interact with Syp, immunoprecipitations with anti-gp130 and anti-JAK2 were performed. Immunoprecipitations with anti-gp130 showed IL-11-induced tyrosine phosphorylation of gp130 (Fig. 4A). When this same immunoblot resulting from a gp130 precipitation was tested for the presence of Syp, Syp was clearly seen with IL-11 induction (Fig. 4B). As a control for the gp130 concentration, the same blot was assayed with anti-gp130, which showed that equal amounts of immunoprecipitated gp130 were in each lane (Fig. 4C). The lack of tyrosine-phosphorylated Syp seen with gp130 in Fig. 4A is due to lower levels of Syp, which coprecipitated with gp130 as compared to a direct immunoprecipitation of Syp (Fig. 1).
To investigate JAK2 association with Syp, JAK2 immunoprecipitations were carried out. JAK2 immunoprecipitations coprecipitated Syp transiently in response to IL-11 with the highest level of association at 3 min (Fig. 5A), which corresponds to the peak level of Syp tyrosine phosphorylation (Fig.  1). JAK2 was shown to be present in equal amounts in each lane in the same experiment (Fig. 5B). The association between Syp and JAK2 appears to be constitutive in unstimulated cells but can be further induced by addition of IL-11.
The YXXV-containing Phosphopeptide Disrupted the IL-11induced Binding of Syp to gp130 and JAK2-To further investigate protein interactions with Syp in response to IL-11, immunoprecipitations with anti-Syp were performed and analyzed by antiphosphotyrosine, anti-JAK2, anti-gp130, and anti-Syp immunoblots (Fig. 6). An antiphosphotyrosine immunoblot showed the presence of tyrosine-phosphorylated proteins in response to IL-11 that may correspond to Syp, JAK2, and gp130 (Fig. 6A). The protein band that may represent  gp130 (150 kDa) is less intense than the other protein bands. Proteins of approximately 90 and 200 kDa were also seen, but their identities were not determined. Protein bands of approximately 200, 150, and 130 kDa were also seen associated with GST-Syp (Fig. 3).
The Syp immunoprecipitation was assayed for the coprecipitation of JAK2 and showed that IL-11 induced the increased association of JAK2 to Syp (Fig. 6B). Further analysis of the Syp immunoprecipitation involved testing for the coprecipitation of gp130. gp130 was seen coprecipitating with Syp when the cells were induced with IL-11 (Fig. 6C). In addition, peptides corresponding to the putative Syp binding site (YSTV) on gp130 were used in a peptide competition experiment to study this association. Immunoprecipitation of Syp in the presence of the unphosphorylated YSTV-containing peptide reduced the association of Syp to gp130 (Fig. 6C, lane 3) but did not decrease the induced association of JAK2 to Syp significantly (Fig. 6B, lane 3). The phosphorylated YSTV-containing peptide blocked significant association of JAK2 and gp130 to Syp. A control assay with anti-Syp was done on the same membrane to test for the amount of Syp with each variable and demonstrated equal quantities of Syp immunoprecipitated (Fig. 6D). Peptide addition before or after lysis coupled with the continued peptide presence throughout the immunoprecipitation gave similar results (data not shown).

DISCUSSION
Our current understanding of IL-11 signal transduction is that after IL-11 joins with its cognate receptor, it then complexes with the signal transducer gp130. gp130 increases its association with JAK2, and JAK2 shows an increase in activity and tyrosine phosphorylation (7) along with an increase in the tyrosine phosphorylation of gp130 (3). IL-6, which has a similar mechanism to IL-11, causes dimerization of gp130 (5), and it has been shown with IL-6 that gp130 tyrosine phosphorylation occurs after JAK2 activation (28). JAK2 is a relatively large (130 kDa) tyrosine-phosphorylated cytoplasmic signaling protein involved in many signaling pathways (29). JAK2 is known to activate signal transducers and activators of transcription proteins (30) and associate with receptors (31,32). In the present study, we have identified a component of the IL-11 pathway that may be equally pervasive. We have shown that the ubiquitously expressed phosphotyrosine phosphatase Syp associates with gp130 and JAK2 and that IL-11 stimulation increases these associations. The association of Syp with gp130 was reduced with the peptide YSTV and blocked with the YSTV-containing phosphopeptide. The YSTV peptide sequence was derived from gp130 and is similar to sequences known to bind Syp (11,33). This finding suggests that Syp is involved in the interaction with the YSTV site on gp130. The unphosphorylated YSTV peptide, however, does not compete away JAK2 association with Syp as much, suggesting that the Syp-JAK2 association is unique and may not require the gp130-Syp association. The induced association of JAK2 and Syp has not been previously presented, although JAK2 has several potential Syp binding sites (26,33,34). The association of Syp in all pathways that activate JAK2 such as erythropoietin, growth hormone, prolactin, granulocyte colony-stimulating factor, IL-3, IL-5, and granulocyte-macrophage colony-stimulating factor will require further investigation (35).
Experiments with GST-Syp fusion proteins indicate specific IL-11-induced associations between the C-terminal SH2 of Syp and tyrosine-phosphorylated proteins. There are reports of the C-terminal SH2 domain having a higher affinity to the PDGF receptor (36) and the N-terminal SH2 domain binding preferably to the erythropoietin receptor (37). The differences seen in the tyrosine-phosphorylated protein bands precipitated between the N/C-and C-terminal GST-SH2-Syp precipitations may indicate a role for the N-terminal GST-SH2 domain of Syp of increasing the specificity and affinity of binding of Syp toward associated proteins. This synergistic role for the two Syp SH2 domains has been reported recently in response to PDGF (36). Tyrosine-phosphorylated protein bands of the size of JAK2 (130 kDa) and gp130 (150 kDa) were seen precipitated with GST-Syp in addition to 90-and 200-kDa proteins of undetermined origin. The large protein may be the insulin receptor substrate-1 docking protein, which binds with Syp (21).
The time course of tyrosine phosphorylation for Syp differs from the results found for A31 mouse fibroblasts induced by PDGF (16). Tyrosine phosphorylation of Syp may correspond directly to activation (19) or protein targeting (38), and either may be the case here. Since activated Syp has not been found to be phosphorylated during insulin receptor substrate-1 activation (21), tyrosine phosphorylation may not be required for activation. Syp-gp130 or Syp-JAK2 interactions may lead to activation of Syp as does a portion of the PDGF receptor and Syp (39).
The IL-11-inducible binding of Syp to both gp130 and JAK2 was shown by reciprocal immunoprecipitations. It was shown previously that the 61 amino acids proximal to the transmembrane region on gp130 are required for the mitogenic signal through gp130 (9). The suggested Syp binding site on gp130 (15) is not included in this region (25) and therefore not required for the mitogenic signal. A possible function of Syp-gp130 binding is the localization of Syp to the membrane to allow easier association with JAK2 since IL-11 induces association of gp130 and JAK2 (38). The necessary site for gp130 and JAK2 association with each other is within the membrane proximal region of gp130 that is required for the mitogenic response (9). Further evidence of Syp being involved in the mitogenic response is a series of microinjection experiments by Xiao and others (41), which showed that the functional inhibition of Syp also inhibited the mitogenic signal. Another potential mitogenic pathway for Syp is as an adaptor protein for the Grb2-Sos-Ras pathway in response to PDGF (24). The induced tyrosine phosphorylation of Shc and other members of the Ras pathway has been shown previously in the gp130 signal transduction pathway (15).
Incubation of cells and lysates with the unphosphorylated YSTV peptide appeared to significantly reduce the IL-11-in-duced association of Syp with gp130 but not reduce the JAK2 association with Syp. Phosphotyrosine-independent peptide binding has been reported previously at a lower affinity than the same tyrosine-phosphorylated peptide at sites of SH2 domain interactions (42,43). A similar interaction may be occurring here. The affinities of the activated gp130 and YSTV peptide may be similar enough for them to compete and reduce gp130-Syp association. The activated JAK2 affinity for Syp may be stronger and thus less sensitive to competition with the unphosphorylated YSTV peptide. An alternative explanation is that the YSTV peptide bound to gp130 and altered the gp130 configuration or blocked gp130 dimerization and thus reduced its association with Syp.
Binding to Syp is strongly enhanced in phosphopeptides containing the protein motif Y*XXV (33), of which JAK2 has two possible sites (Y 201 NSV and Y 1007 YKV) (26) and gp130 has one (YSTV). The presence of these sites and the subsequent inhibition of Syp binding to gp130 and JAK2 by the phosphopeptide implies competitive inhibition with the peptide and suggests possible sites for binding of Syp to gp130 and JAK2. In vivo peptide addition produced no significant effects, implying that in vitro peptide interactions alone, such as competition, suppress association of Syp and gp130 or JAK2.
In summary, the ubiquitous tyrosine phosphatase Syp associates with the signal transducer gp130 and cytoplasmic kinase JAK2. This association increases with IL-11 stimulation and is decreased by a tyrosine-phosphorylated peptide containing the sequence YSTV. A non-phosphorylated peptide, YSTV, affects gp130-Syp but does not affect JAK2-Syp-induced association. These results support a model in which Syp interacts with gp130 and JAK2 independently. Further experiments must be conducted to determine the nature of these interactions.
Addendum-While this work was in review, a paper supporting the importance of the gp130 YSTV sequence for Syp tyrosine phosphorylation was published by Stahl et al. (40). In this paper, a point mutation (Y to F, of YSTV) of an over-expressing gp130-neurotrophin-3 fusion protein blocked the induced antiphosphotyrosine immunoprecipitation of Syp.