Suppressors of cytokine signaling-1 and -6 associate with and inhibit the insulin receptor. A potential mechanism for cytokine-mediated insulin resistance.

Insulin resistance contributes to a number of metabolic disorders, including type II diabetes, hypertension, and atherosclerosis. Cytokines, such as tumor necrosis factor-alpha, interleukin-1 beta, and interleukin-6, and hormones, such as growth hormone, are known to cause insulin resistance, but the mechanisms by which they inhibit the cellular response to insulin have not been elucidated. One mechanism by which these agents could cause insulin resistance is by inducing the expression of cellular proteins that inhibit insulin receptor (IR) signaling. Suppressors of cytokine signaling (SOCS) proteins are negative regulators of cytokine signaling pathways, the expression of which is regulated by certain cytokines. SOCS proteins are therefore attractive candidates as mediators of cytokine-induced insulin resistance. We have found that SOCS-1 and SOCS-6 interact with the IR when expressed in human hepatoma cells (HepG2) or in rat hepatoma cells overexpressing the human IR. In SOCS-1-expressing cells, insulin treatment increases the extent of interaction with the IR, whereas in SOCS-6-expressing cells the association with the IR appears to require insulin treatment. SOCS-1 and SOCS-6 do not inhibit insulin-dependent IR autophosphorylation, but both proteins inhibit insulin-dependent activation of ERK1/2 and protein kinase B in vivo and IR-directed phosphorylation of IRS-1 in vitro. These results suggest that SOCS proteins may be inhibitors of IR signaling and could mediate cytokine-induced insulin resistance and contribute to the pathogenesis of type II diabetes.

Insulin resistance contributes to a number of important metabolic disorders, including type II diabetes, hypertension, and atherosclerosis (reviewed in Refs. 1 and 2). At the cellular level, insulin resistance has been shown to be associated with decreased insulin receptor (IR) 1 tyrosine kinase activity and de-creased phosphorylation of target proteins, such as IRS-1. This results in a decrease in downstream signaling and an attenuation in biological responses. Several cytokines, including interleukin-1␤ (IL-1␤) (3), interleukin-6 (IL-6) (4), and tumor necrosis factor-␣ (TNF␣) (5,6), and hormones, such as growth hormone, have been shown to cause insulin resistance, but the mechanism(s) by which they inhibit signaling has not been elucidated. One mechanism by which these agents could cause insulin resistance is by inducing the expression of cellular proteins that inhibit IR signaling; however, thus far no candidate protein has been definitively identified.
It has recently been reported that certain suppressors of cytokine signaling (SOCS) proteins interact with the IGF receptor in vivo, in vitro, and in the yeast two-hybrid system (7,8). The SOCS proteins are attractive candidates for the putative mediator(s) of cytokine-induced insulin resistance. The SOCS family of proteins was first identified as inhibitors of cytokine signaling (9 -12). Eight SOCS family members have been identified by various techniques, including DNA data base searches (13)(14)(15). At least four SOCS proteins (CIS and SOCS-1, 2, and 3) (9 -12) have been shown to be expressed in mammalian cells. The SOCS family is defined by a characteristic structure composed of a highly variable amino-terminal region, a central SH2 domain, and a highly conserved 40 -50 amino acid motif (called the SOCS box) at the C terminus (13)(14)(15). A wide range of cytokine family members, including IL-1 through -7 (9 -12), granulocyte/macrophage colony-stimulating factor (9), granulocyte colony-stimulating factor (9,11), erythropoietin (12), leukemia inhibitory factor (11), TNF␣ (9,17,18), interferon-␣ and -␥ (9,19), growth hormone (GH) (20,21), and leptin (22), induce the expression of one or more SOCS proteins in different tissues. CIS, SOCS-1, and SOCS-3 have been shown to inhibit cytokine signaling, and it has been postulated that they play a role in the negative feedback control of cytokine pathways (9 -12, 16, 17). Interestingly, SOCS expression is increased by agents known to induce insulin resistance, including IL-1␤, IL-6, TNF␣, and GH. These observations suggest that SOCS proteins could be the putative mediators of cytokine-induced insulin resistance. These studies were undertaken to test this hypothesis by determining whether SOCS proteins interact with the insulin receptor and alter its signaling activity. ria). The pBIG-2i expression plasmid was a gift of Dr. Craig Strathdee (The John P. Robarts Research Institute, London, Ontario) (23). The antibody to the C terminus of the insulin receptor (␣-IRct) was a gift of Dr. J. N. Livingston (Bayer Corp., West Haven, CT), and ␣-IR-1 (24) was a gift of Dr. S. Jacobs (Wellcome Research Laboratories, Research Triangle Park, NC). The antibodies to phospho-ERK1/2 and phospho-Akt (Ser-473) were purchased from Cell Signaling Technology (Beverly, MA) and New England Biolabs, Inc. (Beverly, MA), respectively. The anti-FLAG antibodies M2 and M5 were from Sigma. The anti-phosphotyrosine antibody (4G10) and recombinant rat IRS-1 were from Upstate Biotechnology Inc., Lake Placid, NY. The transfection reagents Gene-PORTER and Fugene 6 were purchased from Gene Therapy Systems, Inc. (San Diego, CA) and Roche Molecular Biochemicals, respectively. Other reagents were obtained from commercial sources as indicated in the text and figure legends.
Plasmid Construction-The SOCS-6 expression vector was constructed by amplifying the entire coding sequence of the murine SOCS-6 cDNA by polymerase chain reaction using upstream and downstream primers containing EcoRI and XhoI sites, respectively. The polymerase chain reaction product was first inserted into the matching sites in the yeast expression plasmid pJG4 -5 (25). Construction of the pJG-SOCS-1 plasmid has been described previously (7). The sequences of all polymerase chain reaction-generated fragments were confirmed by automatic ABI Prism DNA sequencing. The pBIG-2i-FLAG-SOCS-1 and -6 plasmids were constructed by ligating the EcoRI-XhoI fragments of the corresponding pJG derivatives into the pBIG-2i expression plasmid engineered to express an amino-terminal-linked FLAG epitope tag.
Coimmunoprecipitation of SOCS-1 and -6 with the Insulin Receptor-McA-RH7777 cells were transfected with the human insulin receptor construct pHIRc (26) using the technique described above. A population of cells stably expressing the human insulin receptor was isolated with two cycles of cell sorting using a monoclonal anti-human insulin receptor antibody (␣-IR-1). This cell population was designated McA-HIR. McA-HIR cells or HepG2 cells were transiently transfected with the SOCS-1 or SOCS-6 constructs as described above. After 6 h, the cells were placed in serum-free medium for 18 h. Cells were then treated with insulin at 10 Ϫ7 M for 1 min. After two washes with cold phosphate-buffered saline, cells were lysed in Lysis Buffer A containing 50 mM Tris, pH 7.4, 140 mM NaCl, 1% Triton X-100, 50 mM NaF, 10 mM tetrasodium pyrophosphate, 25 mM benzamidine, protease inhibitor mixture (Calbiochem), 2.5 mM pervanadate, 2 mM phenylmethylsulfonyl fluoride, and 10% glycerol. Lysates were passed 10 times through an 18-gauge needle, centrifuged at 10,000 ϫ g for 10 min, and then adjusted to contain equal amounts of protein as determined by the Bradford method (27). Insulin receptors were immunoprecipitated using an anti-human insulin receptor antibody (␣-IR-1) bound to protein G-Sepharose. Immune complexes were washed 4 times with wash buffer (1% Triton X-100, 100 mM Tris-HCl, pH 7.4, and 150 mM NaCl) and separated by SDS-PAGE (28), and associated FLAG-tagged SOCS-1 and -6 were detected by Western blotting using anti-FLAG antibody.
In Vivo Insulin Receptor Autophosphorylation and Activation of Akt and ERK1/2-McA-RH7777 cells transiently expressing SOCS-1 or -6 were harvested after treatment with insulin for 5 min as described above. To analyze the phosphorylation of the insulin receptor, tyrosinephosphorylated proteins were immunoprecipitated with 4G10, and insulin receptors were detected by Western blotting using an antibody directed against the C terminus of the insulin receptor (␣-IRct). Alternatively, receptors were immunoprecipitated with ␣-IRct, and phosphorylated receptors were detected by Western blotting using 4G10. To examine the effects of SOCS proteins on the insulin-dependent activation of Akt and ERK1/2, cell lysates obtained from insulin-treated (5 min) McA-RH7777 cells transiently expressing SOCS-1 or -6 were probed with antibodies selective for the phosphorylated, active forms of the kinases. Total ERK1/2 and Akt content was assayed by stripping the anti-phospho ERK1/2 transfer membrane and reprobing with anti-ERK1/2 or with anti-Akt antibodies.
In Vitro Interactions between the Insulin Receptor and SOCS-1 and -6 -FLAG-tagged SOCS-1 and SOCS-6 were harvested from COS cells transiently expressing the proteins (see above) by immunoprecipitation with anti-FLAG antibody. COS cells were used for the isolation rather than McA-RH7777 cells because the former express few insulin receptors, thus avoiding the possible in vivo complexing of SOCS proteins with endogenous insulin receptors. Insulin receptors were isolated from McA-HIRc cells using wheat germ agglutinin-agarose affinity chromatography. Briefly, ϳ5 ϫ 10 6 cells were lysed in a buffer containing 50 mM Tris, pH 7.4, 100 mM NaCl, 1% Triton X-100, 1 mg/ml bacitracin, 25 mM benzamidine, and 1 mM phenylmethylsulfonyl fluoride. After homogenization in a Dounce homogenizer and centrifugation, the supernatant was applied to a column containing 1 ml of wheat germ agglutinin-conjugated agarose beads. After 30 min at 4°C, the column was drained dry and washed three times with wash buffer (50 mM Tris, pH 7.4, and 0.1% Triton X-100), and the IR was eluted with buffer containing 300 mM N-acetylglucosamine, 50 mM Tris, pH 7.4, and 0.1% Triton X-100.
Insulin (10 Ϫ7 M) and ATP (10 M) were added to aliquots of the wheat germ agglutinin-purified receptors and incubated at room temperature for 30 min. The receptor preparations were then added to the SOCS protein immunoprecipitates and incubated for an additional 30 min. Finally, 10 M [␥-32 P]ATP and recombinant IRS-1 (0.2 g/50 l reaction) were added, and the reaction mixtures were incubated for an additional 30 min at room temperature. The reactions were stopped with Laemmli sample buffer, and the constituents were separated by SDS-PAGE. The SDS-polyacrylamide gel was treated with 1 N NaOH for 60 min at 55°C, fixed, and dried. Analysis was by autoradiography.

RESULTS
Previous studies have demonstrated that SOCS-2 and SOCS-3 interacted with the IGF-1 and/or insulin receptors in vitro, in vivo, and in the yeast two-hybrid system (7,8). We have found that SOCS-1 and SOCS-6 also interact with the IR in the yeast two-hybrid system (data not shown). The present studies were undertaken to determine whether SOCS-1 and SOCS-6 also interact with the IR in mammalian cells and whether they alter IR signal transduction.
To determine whether SOCS-1 or SOCS-6 interacts with the IR in vivo, we transiently expressed the FLAG-tagged proteins in McA-HIR cells that stably overexpress the human IR. Transfected cells were harvested prior to or after insulin stimulation. The extracts were immunoprecipitated with insulin receptor antibody, and the coimmunoprecipitated proteins were resolved by SDS-PAGE and immunoblotted with FLAG antibody (Fig. 1, upper panels). SOCS expression in the transfected cells was confirmed by immunoblotting the total lysates with the FLAG antibody (Fig. 1, lower panels). A small amount of SOCS-1 protein was detected in IR immunoprecipitates from unstimulated cells. SOCS-6 was not detected in association with the IR in the absence of insulin. However, the detection of SOCS-1 and SOCS-6 was markedly increased in the immuno- precipitates following insulin stimulation (Fig. 1, upper panels). The corresponding bands were not observed in cells transfected with the empty vector.
To confirm that the association of SOCS-1 and SOCS-6 with the IR was not limited to cells that overexpress the receptor, SOCS-1 and -6 were also expressed in HepG2 cells. The association of SOCS-1 and SOCS-6 with the endogenous IR was assessed by analyzing the FLAG-tagged SOCS proteins in the IR immunoprecipitates (Fig. 2). In agreement with the results observed with the McA-HIR cells, SOCS-1 association with the IR was detectable in the absence of insulin, but the hormone markedly increased the interaction, whereas the association of SOCS-6 with the IR required insulin stimulation (Fig. 2, upper  panels). These results demonstrate that both SOCS-1 and SOCS-6 associate with the IR in vivo and that receptor activation increases the association.
To determine whether SOCS-1 or SOCS-6 affects IR phosphorylation, the FLAG-tagged proteins or the null vector was transiently expressed in the McA-RH7777 cells. The McA-RH7777 rat hepatoma cell line was chosen because it has a well characterized insulin receptor signaling pathway (29). Extracts prepared from insulin-stimulated or unstimulated cells were immunoprecipitated with a C terminus-specific IR antibody (␣-IRct), and the precipitates were examined by Western blotting for phosphotyrosine (4G10) (Fig. 3, upper panels). Insulin stimulated the receptor autophosphorylation comparably in cells transfected with either the empty vector or the SOCS constructs when examined by this method. However, when the cell lysates were first immunoprecipitated with the phosphotyrosine antibody (4G10) and the precipitates were then examined by Western blotting for the IR, insulin-dependent receptor autophosphorylation was not observed in cells expressing either SOCS protein (Fig. 3, middle panels). The expression of SOCS proteins produced only a small decrease in the total cellular levels of IR (Fig. 3, lower panels). These results demonstrate that the expression of either SOCS-1 or SOCS-6 in rat hepatoma cells in vivo does not inhibit insulin-stimulated IR phosphorylation. However, association of SOCS proteins with the IR blocks recognition of phosphotyrosine residues on the IR by the phosphotyrosine antibody, supporting our observation that these SOCS species form a stable complex with the IR in vivo.
The next series of experiments was designed to determine whether SOCS-1 or SOCS-6 alters IR signal transduction. To this end, the effects of these SOCS proteins on IR-dependent activation of Akt and ERK1/2 were examined. McA-RH7777 cells expressing FLAG-tagged SOCS-1 or SOCS-6 or cells containing the empty vector were treated with insulin for 5 min before harvesting. The activation of Akt and ERK1/2 was assessed by Western blotting using antibodies selective for the active (phosphorylated) forms of these proteins. The expression of either SOCS-1 or SOCS-6 markedly inhibited insulin-dependent activation of both ERK1/2 (Fig. 4, A and B) and Akt (Fig. 4, C and D). It is noteworthy that there was no evidence that SOCS expression altered the low levels of phosphorylated ERK1/2 and Akt present in the unstimulated cells. SOCS expression did not alter the total cellular levels of either ERK1/2 or Akt (Fig. 4, A and C). These results suggest that SOCS-1 and SOCS-6 inhibit insulin-dependent signal transduction.
To determine whether SOCS proteins directly inhibit IR tyrosine kinase activity, immuno-isolated SOCS-1 or SOCS-6 was added to wheat germ agglutinin-purified IR prior to an in vitro kinase assay in which recombinant IRS-1 was used as the substrate. In this assay both SOCS proteins significantly inhibited insulin-stimulated IRS-1 phosphorylation (Fig. 5) with SOCS-6 consistently being more potent than SOCS-1 in multiple experiments. These results suggest that SOCS-1 and SOCS-6 directly inhibit IR substrate phosphorylation, at least in vitro. DISCUSSION The studies reported here were undertaken to determine whether specific SOCS proteins interact with the IR and inhibit its tyrosine kinase activity. They were based on the observation that certain SOCS species, including SOCS-1, -2, -3, and -6, interact with the type I IGF receptor and/or the IR in the yeast two-hybrid system (7,8). 2 The current studies focused on the interactions of SOCS-1 and -6 with the IR. We found that SOCS-1 and SOCS-6 interact with the IR in human hepatoma cells (HepG2) and in rat hepatoma cells overexpressing the human IR (McA-HIR). In both systems, an association of SOCS-1 with the IR was detectable in the absence of insulin treatment but was markedly increased by the hormone; in contrast, the association of SOCS-6 with the receptor required insulin treatment. In McA-RH7777 cells, neither SOCS-1 nor SOCS-6 inhibited insulin-stimulated IR autophosphorylation, but both SOCS proteins inhibited insulin-dependent activation of ERK1/2 and Akt kinase. relatively tightly bound to the inactive receptor because the interaction was maintained after wheat germ agglutinin chromatography of the receptor and prior to its activation with insulin. In total, these findings suggest that SOCS-1 and SOCS-6 could play a role in cytokine-mediated insulin resistance.
The results reported herein suggest that the IR signaling pathway can be suppressed through association with SOCS-1 and SOCS-6. Although this is the first report to demonstrate the inhibition of the IR tyrosine kinase by SOCS proteins, the association of this family of proteins with the IR family of receptors has been reported previously. We earlier reported that SOCS-2 (7) and SOCS-3 (8) associate with the IGF-1 receptor both in vivo and in vitro. Emanuelli et al. (30) implicated SOCS-3 as an inhibitor of insulin-dependent activation of STAT-5B but did not observe an inhibition of receptor tyrosine kinase activity. Nonetheless, their data suggested that binding to tyrosine 960 of the IR was essential for SOCS-3-dependent inhibition. We have not yet determined whether SOCS-2 or SOCS-3 inhibits IR kinase activity or signaling in our system nor have we mapped the sites of interaction of SOCS-1 or SOCS-6 with the IR.
Several studies have examined the mechanisms by which CIS, SOCS-1, and SOCS-3 inhibit cytokine signaling, and it appears that these proteins employ several different mechanisms. CIS, the first SOCS family member to be identified, inhibits erythropoietin signaling by binding to a phosphotyrosine residue on the receptor (Y401), which is also a STAT-5 binding site, thereby competing with STAT-5 for binding and preventing its phosphorylation (12). In contrast, SOCS-1 and SOCS-3 inhibit cytokine signaling by inhibiting the JAK kinases, although by slightly different mechanisms. SOCS-1 inhibits IL-1, -4, and -6 and interferon-␥ signaling by binding directly to and inhibiting the associated Janus kinases (10,11,31,32). In contrast, SOCS-3 inhibits GH signaling by binding to phosphotyrosine residues on the GH receptor itself; this association allows SOCS-3 to then interact with and inhibit the receptor-bound JAK2 kinase (33,34). The JAK kinase inhibitory activities of both SOCS-1 and SOCS-3 have been mapped to a common motif in the distal amino-terminal regions of these proteins, which appears to function as a pseudosubstrate that interacts with the active site and blocks access of legitimate substrates (31,32,35). In support of this, it has been shown that the amino-terminal domains of SOCS-1 and -3 can bind to JAK2 when overexpressed in cells (31,32,33). It is noteworthy that we found that (overexpressed) SOCS-1 also binds to the IR independent of receptor activation. It has also been reported that the binding of SOCS-3 to the IGF-1 receptor does not require receptor activation (8). In both instances, however, the binding of the native proteins (i.e. containing the SH2 domain) Wheat germ agglutinin-purified insulin receptors, previously activated at room temperature for 30 min with 10 Ϫ7 M insulin and 10 M ATP, were added to SOCS-1 or SOCS-6 immunoprecipitates and incubated for an additional 30 min at room temperature. Recombinant rat IRS-1 and [␥-32 P]ATP were then added, and the reaction was allowed to proceed for 30 min at room temperature. Assay components were separated by SDS-PAGE, and the gel was treated with 1 N NaOH for 1 h at 55°C and analyzed by autoradiography. is enhanced by receptor activation. These observations suggest a model in which receptor autophosphorylation provides binding sites for the SH2 domain of SOCS-1, which then enhances the interaction of the amino-terminal inhibitory domain with the active site of the receptor. Our observation that the interaction of SOCS-6 with the IR requires receptor activation suggests that SOCS-6 may inhibit substrate phosphorylation by a different mechanism, but additional experiments will be necessary to confirm this speculation.
The SOCS proteins have been shown to inhibit signaling by two additional mechanisms. SOCS-1 has been shown to associate with Grb-2, the p85 regulatory subunit of PI-3 kinase, and the Rho-family guanine nucleotide exchange factor Vav through determinants in the SOCS-1 amino-terminal domain (36). These interactions are believed to play a role in the suppression of Kit receptor signaling by SOCS-1. These or similar interactions could also play a role in the inhibition by SOCS-1 of IR signaling in vivo, but they could not account for the inhibition of IR kinase activity that we observed in the in vitro kinase assay. Such a mechanism would not apply to SOCS-6 because there is no evidence that SOCS-6 binds to other signaling proteins, and analysis of its amino acid sequence revealed no recognizable binding motifs for these proteins. Finally, it has been shown that the SOCS proteins can inhibit signaling by targeting their partners for proteosome degradation (37,38,39). This activity appears to be mediated by the SOCS motif, which contains a functional BC box similar to that in Elongin A and the von Hippel-Lindau protein (37). Enhanced proteosomal degradation of the IR could contribute to the inhibition of IR signaling in vivo but would not be expected to contribute to the rapid inhibition observed in the experiments reported here (i.e. within 1-5 min of insulin addition) or in the in vitro experiments.
It has been known for some time that inflammatory processes, sepsis, cachexia, GH excess, and obesity are associated with reversible insulin resistance. Importantly, these clinical conditions are also associated with major increases in circulating cytokine levels (1,2). Most relevant to the current report, several of the cytokines implicated in insulin resistance, such as Il-1␤, Il-6, GH, and TNF␣, have been shown to induce SOCS-1 expression in target tissues. In mice, IL-6 and GH rapidly induce hepatic expression of SOCS-1 in vivo (9,18,21,22), and IL-6 and TNF␣ induce its expression in 3T3-L1 adipocytes (17). There are no data on the expression of SOCS-1 in other insulin-responsive tissues in vivo, on the long-term effects of cytokines on its expression, or on its expression in insulin-resistant or diabetic animals. Very little is known about the expression, regulation, or function of SOCS-6, which was identified based on its shared C-terminal SOCS box consensus sequence (15). SOCS-6 has not previously been implicated in any cytokine signaling pathway, and it does not interact with JAK2, lck, c-Kit, or the erythropoietin, IL-2, or fibroblast growth factor receptors (15). In fact, the IR represents the only SOCS-6 binding partner identified to date. The tissue distribution and regulation of expression of SOCS-6 have not been established.
In summary, we propose a model to explain the mechanism of insulin resistance mediated by cytokines. SOCS protein expression is induced in target tissues in response to the cytokines. SOCS-1 and SOCS-6 can associate with the IR and suppress receptor signaling at the level of the receptor tyrosine kinase. In light of evidence demonstrating increases in cytokine levels in inflammatory states, acromegaly, and obesity, the potential contribution of this inhibitory pathway to insulin resistance and type II diabetes should be explored.