Growth hormone regulation of SIRP and SHP-2 tyrosyl phosphorylation and association.

SIRPs (signal-regulatory proteins) are a family of transmembrane glycoproteins that were identified by their association with the Src homology 2 domain-containing protein-tyrosine phosphatase SHP-2 in response to insulin. Here we examine whether SIRPalpha and SHP-2 are signaling molecules for the receptors for growth hormone (GH), leukemia inhibitory factor (LIF), or interferon-gamma (IFNgamma), cytokine receptor superfamily members that bind to and activate Janus kinase 2 (JAK2). In 3T3-F442A fibroblasts, GH rapidly stimulates tyrosyl phosphorylation of both SIRPalpha and SHP-2 and enhances association of SHP-2 with SIRPalpha. Consistent with JAK2 binding and phosphorylating SIRPalpha in response to GH, co-expression of SIRPalpha and JAK2 in COS cells results in tyrosyl phosphorylation of SIRPalpha and JAK2 association with SIRPalpha. LIF does not stimulate tyrosyl phosphorylation of SIRPalpha but stimulates greater tyrosyl phosphorylation of SHP-2 than GH. Additionally, LIF enhances association of SHP-2 with the gp130 subunit of the LIF receptor signaling complex. IFNgamma, which stimulates JAK2 to a greater extent than LIF, is ineffective at stimulating tyrosyl phosphorylation of SIRPalpha or SHP-2. These results suggest that SIRPalpha is a signaling molecule for GH but not for LIF or IFNgamma. Differential phosphorylation of SIRPalpha and SHP-2 may contribute to the distinct physiological effects of these ligands.

The protein tyrosine phosphatase SHP-2 (Src homology 2 domain-containing protein tyrosine phosphatase 2) is a member of a protein-tyrosine phosphatase family characterized by two tandem SH2 1 domains in the N terminus and a catalytic domain in the C terminus. This family also includes the hematopoietic cell phosphatase SHP-1 and the Drosophila phosphatase Csw (1)(2)(3)(4). Unlike SHP-1, SHP-2 is ubiquitously expressed in vertebrate cells and tissues (5,6). SHP-2 plays a critical role in development because disruption of the shp2 gene in mice results in embryonic lethality of mice homozygous for the mutant shp2 (7,8). At the cellular level, SHP-2 has been implicated in signaling by receptor tyrosine kinases and cytokine receptors. In response to ligand, the SH2 domains of SHP-2 associate with IRS-1, insulin receptor, EGF receptor, PDGF receptor, or erythropoietin receptor (5, 9 -12). Binding of the SH2 domains of SHP-2 to tyrosyl phosphorylated signaling molecules is thought to regulate SHP-2 phosphatase activity because phosphopeptides corresponding to SHP-2 binding sites in PDGF receptor and IRS-1 stimulate SHP-2 phosphatase activity (13,14). In addition to binding tyrosyl phosphorylated signaling molecules in response to ligand, SHP-2 is tyrosyl phosphorylated in response to epidermal growth factor and PDGF as well as ligands for cytokine receptors that activate the Janus tyrosine kinases JAK1 and JAK2, such as erythropoietin, IL-11, IL-3, granulocyte-macrophage colony stimulating factor, IFN␣/␤, prolactin, and ciliary neurotrophic factor (5,11,13,(15)(16)(17)(18)(19)(20).
The tyrosines in SHP-2 that are phosphorylated by PDGF receptor, JAK1, or JAK2 are within consensus binding sites for the adapter protein Grb2 and have been shown to mediate Grb2-SHP-2 association in response to PDGF, IL-3, or granulocyte-macrophage colony-stimulating factor (17,(21)(22)(23). Thus, SHP-2 has been hypothesized to play a positive role in signal transduction by serving as an adapter protein between the receptor and Grb2, linking SHP-2 phosphorylation to the Grb2-SOS-Ras-MAPK pathway (22,24). However, the role of SHP-2 in growth factor and cytokine receptor signaling pathways may be more complex than that of an adapter molecule. The catalytic activity of SHP-2 is critical for its ability to modulate certain cellular responses to ligands for some receptor tyrosine kinases and cytokine receptors. For example, SHP-2 phosphatase activity is required for fibroblast growth factor-induced Xenopus development, EGF-induced cell cycle progression, and PDGF-induced mitogenesis (25)(26)(27). Additionally, overexpression of catalytically inactive SHP-2 but not of wild-type SHP-2 inhibits ligand-stimulated reporter gene activity or MAPK activity in response to insulin, EGF, prolactin, or IFN␣/␤ (19, 20, 26, 28 -30).
How phosphatase-inactive SHP-2 negatively regulates signaling is poorly understood. One possible mechanism is that overexpression of inactive SHP-2 prevents dephosphorylation by endogenous SHP-2 of a negative regulator of signaling. In support of this model, overexpression of catalytically inactive but not of active SHP-2 results in hyper tyrosyl phosphorylation of an SHP-2-associated 115-kDa protein in response to insulin in NIH3T3 and CHO cells (28,31,32). The rat and human SHP-2-associated 115-kDa proteins have been cloned and designated SHPS-1 (33) and SIRP (34), respectively, and are both substrates for SHP-2. SIRPs are a family of transmembrane glycoproteins that are divided into two subgroups, based on the presence (SIRP␣) or the absence (SIRP␤) of a cytoplasmic domain. SIRP␣ proteins contain four potential tyrosine phosphorylation sites and a proline-rich region in their cytoplasmic region. The tyrosine phosphorylated cytoplasmic domain is required for growth factor-induced association of SHP-2 with SIRP (34).
Four SIRPs, SIRP␣1, SIRP␣2, SIRP␣3, and SIRP␤1, have been cloned. The best characterized is SIRP␣1. SIRP␣1 is believed to be a negative regulator of growth factor signaling because overexpression of wild-type but not mutant SIRP␣1 lacking the SHP-2 binding region inhibits growth factor-induced MAPK activation and cell proliferation (34). Whether SIRP␣ family members are regulated by cytokine receptor signaling is unknown. We therefore examined the role of SIRP␣ in signaling initiated by GH, LIF, or IFN␥, ligands that activate the Janus tyrosine kinases (35)(36)(37) . Endoglycosidase-F/N-Glycosidase F, Triton X-100, leupeptin, and aprotinin were from Boehringer Mannheim. Recombinant protein A-agarose was from Repligen. The ECL detection system and anti-mouse and anti-rabbit IgG conjugated to horseradish peroxidase were from Amersham. Prestained protein molecular weight standards were from Life Technologies, Inc.
Antisera-Anti-SHP-2 antibody (␣SHP-2) raised against a peptide corresponding to amino acids 576 -593 of human SHP-2 (Santa Cruz) was used for immunoprecipitations at a dilution of 1:250. Monoclonal ␣SHP-2 antibody raised against a peptide corresponding to amino acids 1-177 of human SHP-2 (Transduction Laboratories) was used for immunoblotting at a dilution of 1:2500. Anti-phosphotyrosine antibody 4G10 (␣PY) (Upstate Biotechnology, Inc.) was used at 1:7500 for Western blotting. Antibody to gp130 raised against amino acids 895-914 of murine gp130 (Santa Cruz) was used at 1:2500 for Western blotting. Antibody to GHR (␣GHR) raised against recombinant rat GH-binding protein, was kindly provided by W. Baumbach (American Cyanamid, Princeton, NJ) and was used at a dilution of 1:1000 for immunoprecipitation. Antibody to JAK2 raised against a peptide corresponding to amino acids 758 -776 of murine JAK2 was used at a dilution of 1:1000 for immunoprecipitation and 1:15,000 for Western blotting (38). Antibody to SIRP (␣SIRP) raised against a glutathione S-transferase fusion protein containing amino acids 336 -503 of human SIRP␣1 was used at dilutions of 1:1000 and 1:4000 for immunoprecipitation and Western blotting, respectively (34).
Transfection and Cell Culture-The stock of 3T3-F442A cells was provided by H. Green (Harvard University). 3T3-F442A cells were cultured as described previously (39). COS-7 cells were transfected with prk5 expression vectors encoding the indicated cDNAs by calcium phosphate precipitation (40). cDNAs for SIRP␣1 (10 g), murine JAK2 (2.5 g), or kinase-inactive K882E JAK2 (2.5 g) were used for transient transfection. Empty prk5 expression vector was added to ensure equivalent amounts of DNA in each transfection. After 24 h, cells were washed twice with Dulbecco's modified Eagle's medium and incubated for an additional 24 h with Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, 1 mM L-glutamine and antibiotics. Cell lysates were prepared 48 h post-transfection.
Immunoprecipitation, Western Blotting, and Protein Deglycosylation-Confluent 3T3-F442A fibroblasts were incubated in serum-free medium (41). GH, LIF, or IFN␥ were then added at 37°C for the indicated times at the indicated concentrations. COS or 3T3-F442A cells were washed twice with ice-cold 10 mM sodium phosphate, pH 7.4, 150 mM NaCl, 1 mM Na 3 VO 4 and solubilized in lysis buffer (50 mM Tris, pH 7.5, 0.1% Triton X-100, 150 mM NaCl, 2 mM EGTA, 1 mM Na 3 VO 4 , 1 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin, 10 g/ml leupeptin). Cell lysates were centrifuged at 12,000 ϫ g for 10 min, and the supernatants were incubated on ice for 2 h with the indicated antibody. Immune complexes were collected with protein A-agarose for 1 h at 8°C, washed three times with 50 mM Tris, pH 7.5, 0.1% Triton X-100, 137 mM NaCl, 2 mM EGTA, and boiled for 5 min in a mixture of lysis buffer and 5ϫ SDS-polyacrylamide gel electrophoresis sample buffer (250 mM Tris, pH 6.8, 10% SDS, 10% ␤-mercaptoethanol, and 40% glycerol). Samples were resolved by SDS-polyacrylamide gel electro-phoresis followed by Western blot analysis with the indicated antibodies using the ECL detection system (41). Blots were either directly reprobed with antibody as indicated or stripped in stripping buffer (100 mM ␤-mercaptoethanol, 2% SDS, 62.5 mM Tris-HCl) at 55°C for 30 min and then reprobed with the indicated antibody. For deglycosylation experiments, immunoprecipitated proteins were deglycosylated as described previously (34). Samples were analyzed by Western blotting as described above.

GH Stimulates Tyrosyl Phosphorylation of SHP-2 and an
SHP-2-associated 120-kDa Protein-We first examined whether GH treatment induces tyrosyl phosphorylation of SHP-2. 3T3-F442A cells were incubated with GH for various times, and SHP-2 was immunoprecipitated using ␣SHP-2. GH induces a rapid and transient tyrosyl phosphorylation of SHP-2, with tyrosyl phosphorylation apparent at 2.5 min after GH treatment (the earliest time tested) and decreasing to near basal levels 60 min after GH stimulation ( Fig. 1, lanes 1-6). Reprobing the blot with ␣SHP-2 indicated that the amount of SHP-2 did not change ( Fig. 1, lower panel). A prominent tyrosyl phosphorylated protein with an M r of ϳ120,000 (p120) coprecipitated with SHP-2. p120 displayed a rapid and transient time course of tyrosyl phosphorylation in response to GH, similar to that of SHP-2. To determine if GH enhances the association of p120 with SHP-2 or stimulates tyrosyl phosphorylation of p120 constitutively associated with SHP-2, we first had to determine the identity of p120.
GH Stimulates Tyrosyl Phosphorylation of SIRP␣ and Association of SIRP␣ with SHP-2-To confirm that the SHP-2-associated p120 protein is SIRP␣ and to determine if GH stimulates SHP-2 association with SIRP␣, ␣SHP-2 immunoprecipitates from untreated or GH-treated cells were immunoblotted with ␣SIRP. Because the specificity of this antibody for SIRP␣1, SIRP␣2, and SIRP␣3 has not been determined, we shall refer to the SIRP that this antibody Western blots or immunoprecipitates as SIRP␣. As shown in Fig. 3A, lanes 1 and 2, GH stimulates the association of SIRP␣ with SHP-2. No GHR was detectable by immunoblotting ␣SHP-2 immunoprecipitates with ␣GHR (data not shown), as predicted from the small amount of tyrosyl phosphorylated GHR that seems to co-precipitate with ␣SHP-2 (Fig. 2). Stripping and reprobing the blot in Fig. 3A (lanes 1 and 2) with ␣PY (Fig. 3A, lanes 3 and  4) shows that SIRP␣ co-migrates with the tyrosyl phosphorylated p120 in ␣SHP-2 immunoprecipitates, providing additional evidence that the tyrosyl phosphorylated 120-kDa protein that associates with SHP-2 is SIRP␣. An additional tyrosyl phosphorylated protein of 130 kDa, a mass appropriate for JAK2, also co-precipitated with ␣SHP-2 (Fig. 3A, lane 4). To determine whether GH enhances tyrosyl phosphorylation of SIRP␣, SIRP␣ was immunoprecipitated from cells that were treated with or without GH and immunoblotted with ␣PY. As shown in Fig. 3B, GH enhances tyrosyl phosphorylation of SIRP␣. The results of Fig. 3 therefore indicate that GH promotes tyrosyl phosphorylation of SIRP␣ and its association with SHP-2.
JAK2 Tyrosyl Phosphorylates and Associates with SIRP␣1-The GH-activated tyrosine kinase JAK2 was thought to be the most likely candidate for the SIRP␣ tyrosine kinase. To determine if JAK2 tyrosyl phosphorylates SIRP␣ proteins, we transiently expressed in COS-7 cells JAK2 or kinase-inactive JAK2 in which the critical lysine 882 in the ATP binding region is replaced with glutamate (JAK2 K/E) or SIRP␣1. As shown in Fig. 4, co-expression of SIRP␣1 with JAK2 (lane 3), but not kinase-inactive JAK2 (lane 5), results in tyrosyl phosphorylation of SIRP␣1. JAK2 associates with SIRP␣1, as shown by co-precipitation by ␣SIRP of a tyrosyl phosphorylated protein with a size appropriate for JAK2 in cells overexpressing wildtype JAK2 (Fig. 4, lane 3). The identity of this phosphoprotein as JAK2 was confirmed by immunoblotting with antibody to JAK2 (data not shown). Upon reprobing the nitrocellulose membrane with ␣SIRP, multiple bands corresponding to SIRP␣1 were apparent at 60 -80 kDa in cells transfected with SIRP␣1 (Fig. 4, lanes 2, 3, and 5, lower panels). Because deglycosylation of ␣SIRP immunoprecipitates reduces the multiple bands to a single 60-kDa band (data not shown), the multiple bands most likely represent glycosylation intermediates of overexpressed SIRP␣1 in COS cells.
LIF Stimulates Tyrosyl Phosphorylation of SHP-2 and SHP-2-associated LIFR␤ and gp130 but IFN␥ Stimulates Tyrosyl Phosphorylation of Neither SHP-2 nor of SHP-2-associated Proteins-Because JAK2 is activated in response to LIF and IFN␥ as well as GH (35)(36)(37), we compared the abilities of GH, LIF, and IFN␥ to induce tyrosyl phosphorylation of SHP-2 and SHP-2-associated proteins such as SIRP␣. 3T3-F442A cells were incubated with 25 ng/ml LIF or 10 ng/ml IFN␥, concentrations shown previously to induce maximal JAK2 tyrosyl phosphorylation. 2 Although LIF stimulates tyrosyl phosphorylation of JAK2 poorly in comparison with GH (Fig. 5, lanes 1 and 4), LIF induces a more robust tyrosyl phosphorylation of SHP-2 than GH (Fig. 5, lanes 2, 3, 5 and 6). Our results support previous findings that IL-6, oncostatin M, or LIF, ligands that utilize gp130 in their receptor signaling complexes, stimulate tyrosyl phosphorylation of a protein that migrates with a size appropriate for SHP-2 (18). LIF did not increase the amount of tyrosyl phosphorylated SIRP␣ associated with SHP-2 (Fig. 5,  lanes 5 and 6). Instead, LIF stimulates tyrosyl phosphorylation of two other SHP-2-associated proteins that migrate at sizes appropriate for gp130 and LIFR␤. In contrast to GH and LIF, IFN␥ does not stimulate tyrosyl phosphorylation of SHP-2 or SHP-2-associated proteins such as SIRP␣, even though IFN␥ promotes a more robust tyrosyl phosphorylation of JAK2 than LIF does (Fig. 5, lanes 7-9).
LIF Stimulates SHP-2 Association with gp130 -To determine whether LIF stimulates association of SHP-2 with gp130 or tyrosyl phosphorylation of gp130 constitutively associated with SHP-2, ␣SHP-2 immunoprecipitates from untreated or LIF-treated cells were immunoblotted with antibody to gp130. As shown in Fig. 6 (lanes 1 and 2), LIF stimulates association of SHP-2 with gp130. Reprobing this blot with ␣PY demonstrates that gp130 and the SHP-2-associated 130-kDa phosphoprotein co-migrate (Fig. 6, lanes 3 and 4). In this experiment, LIF induces tyrosyl phosphorylation of SHP-2-associated SIRP␣ to a small extent. However, LIF-induced tyrosyl phosphorylation of SHP-2-associated SIRP␣ was not reproducible, occurring in only three out of ten experiments.
GH, but Not LIF nor IFN␥, Potently Stimulates Tyrosyl Phosphorylation of SIRP␣-To determine whether SIRP␣ is tyrosyl phosphorylated in response to LIF or IFN␥, 3T3-F442A cells were treated with GH, LIF, or IFN␥, and ␣SIRP immunoprecipitates were immunoblotted with ␣PY. Consistent with the previous data (Fig. 5, lanes 2 and 3), GH potently stimulates tyrosyl phosphorylation of SIRP␣ (Fig. 7, lanes 1 and 2). However, neither LIF nor IFN␥ stimulate tyrosyl phosphorylation of SIRP␣ (Fig. 7, lanes 3 and 4). These results suggest that GH, but not LIF or IFN␥, regulates SIRP␣. DISCUSSION SIRP␣ Is a Signaling Molecule for GH-Our findings that GH induces tyrosyl phosphorylation of SIRP␣ and association of SHP-2 with SIRP␣ demonstrate that SIRP␣ is involved in GHR signaling. To our knowledge, this is the first evidence that SIRPs are involved in cytokine receptor signaling. JAK2 is the most likely candidate for the tyrosine kinase that phosphoryl-2 L. S. Argetsinger and C. Carter-Su, unpublished data. ates SIRP␣ in response to GH. JAK2 is potently activated by GH (35), and co-expression of wild-type but not kinase-inactive JAK2 with SIRP␣1 in COS cells results in tyrosyl phosphorylation of SIRP␣1 (Fig. 4). The association of SIRP␣1 with JAK2 in COS cells may not be dependent on phosphorylated tyrosines in either JAK2 or SIRP␣1, because neither JAK2 nor SIRP␣1 contains a known phosphotyrosine binding domain (e.g., SH2 or PTB). However, SIRP␣1 contains a proline-rich region (PKQPAPKP) that conforms to the consensus sequence (hydrophobic-XXX-aliphatic-PXP) of the proline-rich box1 region of cytokine receptors that mediates JAK2-cytokine receptor association (43,44) Therefore, it is tempting to speculate that the association of JAK2 with SIRP␣ involves binding of JAK2 to the proline-rich region of SIRP␣. Whether or not GH-induced tyrosyl phosphorylation of SIRP␣ in 3T3-F442A cells is a consequence of a direct interaction of JAK2 with SIRP␣, as suggested by the COS cell transfection experiments, or involves an additional protein(s) remains to be determined. Interestingly, not all ligands that activate JAK2 induce detectable levels of tyrosyl phosphorylated SIRP␣ in 3T3-F442A cells. The ability of GH, but not LIF or IFN␥, to potently stimulate tyrosyl phosphorylation of SIRP␣ may be a consequence of the much greater ability of GH to stimulate JAK2 tyrosyl phosphorylation ( Fig. 5 and Refs. 45 and 46). Alternatively, tyrosyl phosphorylation of SIRP␣ may require a factor(s) in addition to JAK2 that is recruited by GHR but not the receptors for LIF or IFN␥.
JAK2 is thought to be mediated primarily via GHR because JAK2 binds to a glutathione S-transferase fusion protein containing the SH2 domains of SHP-2 only when JAK2 is coexpressed with GHR. 3 Similarly, the GH-dependent increase in the amount of SHP-2 co-precipitated with ␣IRS-2 suggests that a small amount of SHP-2 also binds to IRS proteins (48). The amount of SHP-2 associated with IRS-2 is probably a small subset of total SHP-2 because we did not observe IRS-2 in ␣SHP-2 immunoprecipitates.
As expected from the inability of LIF or IFN␥ to stimulate tyrosyl phosphorylation of SIRP␣, SHP-2 does not seem to associate with tyrosyl phosphorylated SIRP␣ in response to these ligands. Instead, LIF stimulates association of SHP-2 with tyrosyl phosphorylated gp130-LIFR␤ complexes, presumably via the SH2 domains of SHP-2. In support of this, IL-11 has been shown to stimulate association of SHP-2 with gp130 via a YSTV motif in gp130 and association of the SH2 domains of SHP-2 with a phosphoprotein of the appropriate size for gp130 (16). LIFR␤ also contains an SHP-2 binding motif, but whether LIFR␤ also binds SHP-2 remains to be determined.
In contrast to LIF and GH, IFN␥ did not stimulate association of SHP-2 with tyrosyl phosphorylated signaling proteins in 3T3-F442A cells even though IFN␥ receptor ␣ contains a YSLV motif that could be a potential binding site for the SH2 domain of SHP-2. This suggests that SHP-2 may not play a major role in IFN␥ signaling. However, our data do not exclude the alternative possibilities that SHP-2 does associate with IFN␥ receptor signaling molecules but that the interaction is so rapid and transient that we do not detect it in our assays or that SHP-2 binds to non-tyrosyl phosphorylated signaling molecules in response to IFN␥.
SHP-2 Is Tyrosyl Phosphorylated in Response to GH and LIF, but Not IFN␥-GH and LIF, but not IFN␥, stimulate the tyrosyl phosphorylation of SHP-2, with LIF being more effective than GH. Thus, the levels of tyrosyl phosphorylation of SHP-2 induced by GH, LIF, or IFN␥ do not correspond to the relative levels of JAK2 tyrosyl phosphorylation induced by these ligands. One explanation of this apparent discrepancy is that the levels of SHP-2 tyrosyl phosphorylation may reflect the differential abilities of the ligands to recruit SHP-2 to cytokine receptor-JAK signaling complexes, where SHP-2 can be tyrosyl phosphorylated by JAKs. For example, in response to LIF, SHP-2 appears to be primarily recruited into the gp130-LIFR␤-JAK signaling complex where it is likely to be in sufficiently close proximity to LIFR␤-gp130-associated JAKs to be highly phosphorylated. Consistent with this model, mutation of the SHP-2 binding site in gp130 abolishes neurotrophin-3-induced tyrosyl phosphorylation of SHP-2 by JAKs in cells overexpressing TrkC-gp130 chimeric proteins (49). In the case of GH, the majority of SHP-2 appears to be recruited to tyrosyl phosphorylated SIRP␣. This subset of SHP-2 may not be in close enough proximity to JAK2 to be phosphorylated. The SHP-2 that is phosphorylated may reflect the smaller amount of SHP-2 that is recruited to GHR-JAK2 complexes. The greater binding of SHP-2 to SIRP␣ compared with GHR-JAK2 may reflect a higher affinity of SHP-2 for SIRP␣ than for GHR or a greater degree of phosphorylation of SHP-2 binding motifs within SIRP␣ compared with GHR. Consistent with the lack of association of SHP-2 with tyrosyl phosphorylated IFN␥ receptor-JAK signaling complexes, IFN␥ does not induce SHP-2 tyrosyl phosphorylation in 3T3-F442A cells.
Role of SIRP␣ and SHP-2 in GH, LIF, and IFN␥ Signaling-Our findings that GH stimulates SIRP␣ tyrosyl phosphorylation and the association of SHP-2 with SIRP␣ suggests that SIRP␣ plays a role in signaling by GH and potentially other members of the cytokine receptor superfamily. Overexpression of SIRP␣1, but not mutant SIRP␣1, which cannot bind SHP-2, inhibits insulin and EGF-induced DNA synthesis and MAPK activation (34), suggesting that SIRP␣ may be a negative regulator of these cellular functions. These effects are mimicked by overexpression of catalytically inactive SHP-2. The observation that overexpression of inactive SHP-2 results in a dramatic accumulation of SHP-2 at the cell membrane (32), coincident with enhanced SIRP␣ tyrosyl phosphorylation, raises the possibility that a major function of SIRP␣ is to recruit SHP-2 to the cell membrane, where it dephosphorylates SIRP␣ and other membrane-associated proteins. Binding of the SH2 domains of SHP-2 to phosphorylated tyrosine-containing motifs has been shown to stimulate SHP-2 phosphatase activity (13,14). In support of SHP-2 being activated by binding to SIRP␣, SHP-2 associated with SIRP␣ via SHP-2 SH2 domains dephosphorylates SIRP␣ (32,34).
The differential abilities of GH, LIF, and IFN␥ to induce tyrosyl phosphorylation of SHP-2 and SIRP␣ may contribute to the distinct biological effects induced by these ligands. example, because SHP-2 is so highly tyrosyl phosphorylated in response to LIF as compared with GH or IFN␥, a major function of SHP-2 in LIF signaling may be to bind SH2 domaincontaining proteins such as Grb2. In a COS cell overexpression system, JAK2 tyrosyl phosphorylates SHP-2 on tyrosine 304, a consensus binding site for Grb2 (21,23). Therefore, SHP-2 tyrosyl phosphorylation may link LIFR and to a lesser extent GHR to the Grb2-Ras-MAPK pathway. In contrast to LIF, GH stimulates robust tyrosyl phosphorylation of SIRP␣ and weaker tyrosyl phosphorylation of SHP-2. Because SIRP␣ appears to be the major tyrosyl phosphorylated protein that binds SHP-2 in response to GH, a major function of SHP-2 in GH signaling may be to bind and dephosphorylate SIRP␣. Because SIRP␣ proteins contain an extracellular domain with three immunoglobulin-like repeats, it is possible that SIRP␣ is recruited into GHR signaling pathways in response to binding an as yet unidentified ligand or that GH highly regulates the cellular response to that ligand. The effects of SIRPs on GH signal transduction have yet to be determined but should offer new insights into signaling pathways utilized by GHR and potentially other cytokine receptors that activate Janus tyrosine kinases.