Mutation of Tyrosine 960 within the Insulin Receptor Juxtamembrane Domain Impairs Glucose Transport but Does Not Inhibit Ligand-mediated Phosphorylation of Insulin Receptor Substrate-2 in 3T3-L1 Adipocytes*

CSF-1 is equipotent to insulin in its ability to stimulate 2-[3H]deoxyglucose uptake in 3T3-L1 adipocytes expressing the colony stimulating factor-1 receptor/insulin receptor chimera (CSF1R/IR). However, CSF-1-stimulated glucose uptake and glycogen synthesis is reduced by 50% in comparison to insulin in 3T3-L1 cells expressing a CSF1R/IR mutated at Tyr960(CSF1R/IRA960). CSF-1-treated adipocytes expressing the CSF1R/IRA960 were impaired in their ability to phosphorylate insulin receptor substrate 1 (IRS-1) but not in their ability to phosphorylate IRS-2. Immunoprecipitation of IRS proteins followed by Western blotting revealed that the intact CSF1R/IR co-precipitates with IRS-2 from CSF-1-treated cells. In contrast, the CSF1R/IRA960 co-precipitates poorly with IRS-2. These observations suggest that Tyr960is important for interaction of the insulin receptor cytoplasmic domain with IRS-2, but it is not essential to the ability of the insulin receptor tyrosine kinase to use IRS-2 as a substrate. These observations also suggest that in 3T3-L1 adipocytes, tyrosine phosphorylation of IRS-2 by the insulin receptor tyrosine kinase is not sufficient for maximal stimulation of receptor-regulated glucose transport or glycogen synthesis.

A complete understanding of insulin action requires the identification of the intracellular pathways that regulate insulin-stimulated growth, development, and metabolism. Molecular analysis of structures within the insulin receptor cytoplasmic domain that are mediators of insulin-stimulated signal transduction pathways has provided one strategy for the study of insulin-sensitive intracellular signaling. The usual approach has been to create mutations within specific receptor sequences that result in the selective disruption of some insulin-regulated pathways, while leaving others intact. Deletion mutagenesis of the insulin receptor cytoplasmic domain has generated insulin receptors with altered biological properties (1)(2)(3)(4) and led to the suggestion that different regions of the insulin receptor cytoplasmic domain play distinct roles in modulating the biological effects of insulin. Most mutations of the insulin receptor cytoplasmic domain that altered autophosphorylation sites within the cytoplasmic domain (5)(6)(7)(8) furthered the notion that tyrosine phosphorylation and the tyrosine kinase encoded within the receptor ␤ subunit are essential components of normal insulin action. Previously, structure/function analysis required the introduction of mutated insulin receptors into cells expressing low levels of endogenous insulin receptors, e.g. Chinese hamster ovary or Rat-1 fibroblast cell lines (for a review, see Ref. 9) to minimize the background signal generated by insulin activation of endogenous receptors. However, the relative insensitivity of fibroblast cell lines to insulin treatment has made them less than ideal models for studying the action of insulin on intermediary metabolism.
Conversely, insulin-responsive cell types are excellent systems in which to study insulin action, but the fact that they express large numbers of endogenous insulin receptors is an impediment to the expression and analysis of mutated recombinant insulin receptors. To circumvent this concern, the extracellular ligand binding domain of the human CSF-1 1 receptor (10) was spliced to the transmembrane and cytoplasmic domains of the human insulin receptor (11)(12)(13) to construct a chimeric receptor (CSF1R/IR). Recently, we showed that the CSF1R/IR has CSF-1-dependent enzymatic and biological properties expected of the insulin receptor tyrosine kinase in insulin-responsive cells (14). Deletion of 12 amino acids within the juxtamembrane domain of the CSF1R/IR inhibits CSF-1stimulated phosphorylation of IRS-1 and Shc (14) and blocks CSF1R/IR-mediated glucose uptake (15) but does not block the ability of the chimeric receptor to stimulate the differentiation of 3T3-L1 cells into adipocytes (14) or promote cell survival in the presence of an apoptotic stimulus (15). In the experiments presented here we examined the signaling properties of a CSF1R/IR construct bearing a single mutation of the tyrosine at the position corresponding to Tyr 960 of the normal human insulin receptor to alanine. Our data demonstrate that the Ala 960 mutation partially inhibits the ability of the CSF1R/IR to stimulate glucose uptake and blocks the ability of the chimera to stimulate the phosphorylation of IRS-1 but not IRS-2. Despite the ability of the CSF1R/IRA960 to phosphorylate IRS-2, co-precipitation studies indicate that the Ala 960 mutation decreases the ability of the insulin receptor cytoplasmic domain to interact with IRS-2.
EXPERIMENTAL PROCEDURES 3T3-L1 cells expressing the CSF1R/IR were described previously (14). Antibodies Ca-1 and CT-1 directed against the cytoplasmic domain of the human insulin receptor (17) were a gift from Ken Siddle, Cambridge, UK. Antibodies to IRS-1 were a gift from L. Sweet, West Haven, CT. Antibodies to IRS-2 were described previously (18). Anti-phosphotyrosine (RC20H) antibodies were from Transduction Laboratories. Dulbecco's modified Eagle medium (DMEM), penicillin, streptomycin and neomycin antibiotics, G418, restriction endonucleases and Lipo-fectAMINE were purchased from Life Technologies, Inc. Fetal bovine serum was purchased from Atlanta Biologicals. Porcine insulin was from Calbiochem. Dexamethasone and 3-isobutyl-1-methylxanthine were from Sigma. Human recombinant CSF-1 was a gift from Genetics Institute.
Construction and Expression of the CSF1R/IR and the CSF1R/ IRA960 -Construction of the CSF1R/IR was described previously (14). The CSF1R/IR chimera containing an alanine substitution for Tyr 960 in the insulin receptor cytoplasmic domain (CSF1R/IRA960) was constructed by replacing the SpeI/BglII fragment from the CSF1R/IR with the corresponding fragment from pSGHIRcA960 (19). Expression of the CSF1R/IR and the CSF1R/IRA960 cDNAs was accomplished by subcloning the chimera cDNAs into the vector pEF-1 (14). Transfection of 3T3-L1 preadipocytes was performed with LipofectAMINE according to the manufacturer's instructions. Cells expressing the CSF1R/IR were selected with 800 g/ml G418 and isolated by flow cytometry with a monoclonal antibody against the extracellular domain of the human CSF-1 receptor (20) (Oncogene Science) and fluorescent secondary antibody. Flow cytometry was performed by the University of Nebraska Medical Center Flow Cytometry Core Laboratory.

2-[ 3 H]Deoxy-D-glucose Uptake in 3T3-L1
Cells-Glucose uptake in 3T3-L1 adipocytes was measured using a modification of the method described previously (21). 3T3-L1 adipocytes were used 14 days after induction of differentiation, at which time Ͼ90% of the cells expressed an adipocyte phenotype. Cell monolayers were washed twice with buffer A (130 mM NaCl, 2.7 mM KCl, 1 mM KH 2 PO 4, 8.1 mM Na 2 HPO 4 , 0.68 mM CaCl 2 ,0.49 mM MgCl 2 , pH 7.4) and then incubated for 2 h at 37°C in 1 ml of serum-free DMEM. To determine nonspecific glucose uptake, the cells were treated for 10 min in serum-free DMEM containing 20 M cytochalasin B and 200 M phloretin. Cell monolayers were washed once with 1 ml of Krebs-Ringer's phosphate buffer (130 mM NaCl, 5 mM KCl, 1.3 mM CaCl 2 , 1.3 mM MgSO 4, 10 mM Na 2 HPO 4 , pH 7.4) with or without CSF-1 or insulin as indicated and incubated for 20 min at 37°C in 1 ml of the same buffer. The glucose uptake assay was initiated by replacement of Krebs-Ringer's phosphate buffer in each well with Krebs-Ringer's phosphate buffer containing 0.1 mM 2-deoxy-D[2,6-3 H]glucose (1 Ci) for 10 min at 37°C. Termination of glucose uptake was performed by the rapid removal of assay buffer followed by four rapid washes of each well with 1 ml of ice-cold buffer A containing 500 M phloretin. Cells were removed from each well with 0.4 ml of 0.1% SDS and counted for radioactivity after the addition of 4 ml of Optifluor (Packard Instrument Co).
Glycogen Synthesis-The incorporation of glucose into glycogen was measured as described previously (22). Fully differentiated adipocytes in 2-cm wells were serum-starved for 3 h in DMEM with 5.5 mM glucose and 2 mM glutamine. Cells were washed in glycogen assay buffer (140 mM NaCl, 1.7 mM KCl, 0.9 mM CaCl 2 , 1.47 mM K 2 HPO 4 , 0.9 mM MgSO 4 , 25 mM Tris-HCl, pH 7.5, and 2 mg/ml bovine serum albumin), stimulated with the indicated concentrations of hormones for 20 min, and incubated with D-[U-14 C] glucose (1 Ci/ml, final specific activity 0.26 Ci/mol) for 30 min. Incubations were stopped by 3 rapid washes on ice with phosphate-buffered saline, and the cells were solubilized in 1 ml of 0.1 M NaOH. Carrier glycogen (1 mg) was added, and the samples were boiled for 30 min. Glycogen was precipitated with 70% ethanol overnight at Ϫ20°C. Each sample was centrifuged at 2100 ϫ g, washed once with 70% ethanol, and resuspended in water. The amount of radioactivity incorporated into glycogen was determined by scintillation counting.
PI 3Ј Kinase Assay-Differentiated 3T3-L1 adipocytes in 35-mm culture dishes were serum-starved for 5 h in DMEM before hormone stimulation. Cells were then stimulated with 10 nM CSF-1 or 100 nM insulin for 2 min. The media was aspirated, and the cell monolayers were washed once with ice-cold phosphate-buffered saline containing 0.1 mM Na 3 VO 4 . The proteins were solubilized for 10 min in lysis buffer (40 mM Tris-HCl, pH 7.4, 2 mM Na 3 VO 4 , 20 mM NaF, 10 mM sodium inorganic pyrophosphate, 1 mM EDTA, 1 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 20 g/ml aprotinin) containing 1% Triton X-100. Cell lysates containing equal amounts of protein were incubated with the indicated antisera or antibodies for 2 h at 4°C. Protein G-Sepharose (25 l) was added to each immune complex, and incubation was continued for 2 h at 4°C. The immunoprecipitates were washed twice with each of the three following buffers: (a) phosphate-buffered saline, pH 7.4, containing 1% Triton X-100, (b) 100 mM Tris, pH 7. . The samples were briefly separated by centrifugation, and the lower organic phase (20 l) was spotted on silica gel 6OA plates (Merck) that had been pretreated with 1% potassium oxalate at 100°C. The plates were developed in chloroform/acetone/methanol/acetic acid/water (35:15:13:12:8), dried, and detected by autoradiography. Phosphorylated PI was quantified with a Molecular Dynamics densitometer.

RESULTS
Tyrosine Phosphorylation of the CSF1R/IR and CSF1R/ IRA960 Chimeras-Quiescent 3T3-L1 preadipocytes expressing the CSF1R/IR and the CSF1R/IRA960 were incubated in the presence or absence of CSF-1, and aliquots of cell lysates containing equal amounts of protein were immunoprecipitated with Ca-1 antibody directed against the cytoplasmic domain of the insulin receptor. Treatment of 3T3-L1 preadipocytes expressing the CSF1R/IR chimera with CSF-1 for 10 min resulted in the tyrosine phosphorylation of the 170-kDa chimeric receptor protein (Fig. 1, 2nd lane). The same treatment of 3T3-L1 cells expressing the CSF1R/IRA960 chimera resulted in a slightly reduced extent of the tyrosine phosphorylation of the mutant receptor (Fig. 1, 4th lane) compared with the intact chimeric receptor, which is consistent with prior studies (8). The intact and mutated chimera were expressed equally, as determined by Western blot with Ca-1 antibody (Fig. 1). The same effect of CSF-1 on phosphorylation of the CSF1R/IR and CSF1R/IRA960 was observed in fully differentiated 3T3-L1 adipocytes (data not shown).
Effect of the CSF1R/IRA960 on Glucose Transport and Glycogen Synthesis-Tyr 960 in the insulin receptor juxtamembrane domain is essential for the ability of the receptor kinase to phosphorylate IRS-1 and couple to downstream effectors (1,25). The tyrosine in the CSF1R/IR corresponding to Tyr 960 in the insulin receptor was changed to alanine, and the resulting CSF1R/IRA960 chimera was tested for its ability to stimulate glucose uptake upon stimulation with different doses of CSF-1. CSF-1 was equipotent to insulin in its ability to stimulate maximal glucose uptake in 3T3-L1 adipocytes expressing the CSF1R/IR (Fig. 3A). At submaximal concentrations, CSF-1 appears more potent than insulin in its ability to stimulate glucose uptake and glycogen synthesis (Fig. 3). This may result from the fact that the expression of CSF1R/IR is slightly higher than endogenous insulin receptors in the transfected 3T3-L1 adipocytes (data not shown). In adipocytes expressing the CSF1R/IRA960, glucose uptake in response to 100 nM CSF-1 was no greater than that observed when the same cells were treated with 1 nM insulin (Fig. 3A). However, glucose uptake stimulated by 1 nM and 10 nM CSF-1 was unaffected by the Ala 960 mutation (Fig. 3A).
To determine whether the mutation of Tyr 960 inhibited only the ability of the chimera to stimulate maximal glucose uptake, we examined glycogen synthesis in 3T3-L1 adipocytes expressing the CSF1R/IR and CSF1R/IRA960. In cells expressing the CSF1R/IR, glycogen synthesis stimulated by CSF-1 resembled the dose-dependent increases in synthesis observed with insulin (Fig. 3B). As seen with glucose uptake, however, CSF-1 was unable to stimulate maximal glycogen synthesis in cells expressing the CSF1R/IRA960 above the level of synthesis observed with 1 nM insulin (Fig. 3B).
CSF-1-stimulated Tyrosine Phosphorylation of IRS-1 and IRS-2 in 3T3-L1 Adipocytes Expressing the CSF1R/IR and CSF1R/IRA960 -The ability of the Ala 960 mutation to blunt but not abolish CSF-1-stimulated glucose uptake and glycogen synthesis in adipocytes expressing the CSF1R/IRA960 indicated that the juxtamembrane mutation did not completely uncouple the CSF1R/IR from downstream effectors that regulate these biological activities. IRS-1 is readily detectable in adipocytes (26) permitting examination of the ability of the CSF1R/IR and CSF1R/IRA960 to tyrosine-phosphorylate this endogenous substrate of the insulin receptor kinase. Cells were left untreated or treated with CSF-1 and lysed, and IRS-1 was immunoprecipitated and analyzed by Western blotting with anti-phosphotyrosine antibodies. CSF-1 stimulated tyrosine phosphorylation of IRS-1 in adipocytes expressing the intact CSF1R/IR (Fig. 4). Consistent with previous studies (27), substitution of Tyr 960 with alanine-blocked CSF-1 stimulated tyrosine phosphorylation of IRS-1 in adipocytes expressing the CSF1R/IRA960 (Fig. 4). Stripping and reprobing of the blot verified that equivalent amounts of IRS-1 were present in each immunoprecipitate from each cell line (Fig. 4, middle panel). In contrast, CSF-1-treated 3T3-L1 adipocytes expressing the CSF1R/IRA960 were comparable with 3T3-L1 adipocytes expressing the CSF1R/IR in their ability to phosphorylate IRS-2 (Fig. 4). Stripping and reprobing of the blot also revealed comparable levels of expression of IRS-2 in each immunoprecipitate from each cell line (Fig. 4).
Interaction of IRS-2 with the CSF1R/IR and the CSF1R/ IRA960 -A second tyrosine-phosphorylated protein co-precipitated with IRS-2 from CSF-1-treated cells (Fig. 4). To determine the identity of this protein, the CSF1R/IR or the CSF1R/ IRA960 were left untreated or treated with 10 nM CSF-1, and lysates were immunoprecipitated with antibodies to IRS-2 or to the carboxyl terminus of the insulin receptor cytoplasmic domain (CT-1). IRS-2 was phosphorylated in CSF-1-treated cells expressing the CSF1R/IR or the CSF1R/IRA960. However, the second tyrosine-phosphorylated protein co-precipitated with IRS-2 in cells expressing the CSF1R/IR but did not co-precipitate with IRS-2 in cells expressing the CSF1R/IRA960 (Fig. 5). Reprobing of the blots with CT-1 antibody revealed that the CSF1R/IR, but not the CSF1R/IRA960, coprecipitated with phosphorylated IRS-2 protein.
CSF-1-stimulated Activation of PI 3Ј Kinase in 3T3-L1 Adipocytes Expressing the CSF1R/IR or the CSF1R/IRA960 -PI 3Ј kinase is activated by the insulin receptor via association with tyrosine-phosphorylated intermediate proteins including IRS-1 (28). To further investigate the role of the juxtamembrane Tyr 960 in downstream signaling by the insulin receptor, we examined the ability of the CSF1R/IRA960 to activate PI 3Ј kinase associated with IRS-1 and IRS-2 or with tyrosine-phosphorylated proteins. 3T3-L1 adipocytes expressing the CSF1R/IR or the CSF1R/IRA960 were serum-deprived for 5 h and then treated with 10 nM CSF-1 or with 100 nM insulin for 2 min. The cells were lysed and immunoprecipitated with antibodies to IRS-1, IRS-2, or phosphotyrosine. In adipocytes expressing the CSF1R/IR, insulin and CSF-1 stimulated PI 3Ј kinase activity in anti-phosphotyrosine immunoprecipitates as well as in anti-IRS-1 or anti-IRS-2 immunoprecipitates (Fig. 6). In contrast, CSF-1 was unable to stimulate PI 3Ј kinase in anti-IRS-1 immunoprecipitates from adipocytes expressing the CSF1R/IRA960 (Fig. 6). CSF-1 stimulated PI 3Ј kinase activity in anti-IRS-2 and anti-phosphotyrosine immunoprecipitates from adipocytes expressing the CSF1R/IRA960 (Fig.  6), consistent with its ability to stimulate the phosphorylation of IRS-2. DISCUSSION These data demonstrate that the juxtamembrane phosphorylation site Tyr 960 within the insulin receptor cytoplasmic domain is an essential determinant for the tyrosine phosphorylation of IRS-1 but not for the phosphorylation of IRS-2 by the insulin receptor kinase (Fig. 4). Phosphorylation of IRS-2 by the CSF1R/IRA960 is functional as it stimulates the activation of PI 3Ј kinase bound to IRS-2 (Fig. 6). The Ala 960 mutation blocks the ability of the chimeric CSF1R/IR to maximally stimulate glucose uptake and glycogen synthesis but does not impair stimulation of glucose transport by submaximal amounts of CSF-1 (Fig. 3). If IRS-2 is capable of mediating insulinstimulated GLUT4 translocation (29), then these data suggest that interaction of IRS-2 with the insulin receptor juxtamembrane domain may not be a required event to induce glucose uptake. The ability of the CSF1R/IRA960 to phosphorylate IRS-2 and to stimulate glucose transport may explain why ectopic expression of IRS-1 domains that interact with the insulin receptor juxtamembrane domain were able to block insulin-stimulated mitogenic effects mediated by IRS-1 but did not inhibit insulin-stimulated GLUT 4 translocation or glucose transport (30,31). The data presented here demonstrate that competitive inhibition of interaction with the juxtamembrane domain is not required for tyrosine phosphorylation of IRS-2.
These data also suggest an important but not exclusive role for Tyr 960 and IRS-1 in signaling by the insulin receptor kinase (14,15,30,31). The conclusion that other mechanisms, in addition to the phosphorylation of IRS-1, specifically contribute to insulin-stimulated glucose uptake is consistent with several forms of evidence. Microinjection of antibodies and peptides or the exogenous expression of IRS-1 domains disrupt protein interaction with the insulin receptor juxtamembrane region (30,31). IRS-1 Ϫ/Ϫ mice show mild resistance to insulin and impaired glucose tolerance (32,33). Adipocytes isolated from IRS-1 Ϫ/Ϫ mice are impaired but not completely resistant to insulin-stimulated glucose uptake (33). IRS-2 Ϫ/Ϫ mice also demonstrate peripheral insulin resistance, but the severity of the phenotype is compounded by a coinciding lack of pancreatic ␤-cell compensation (34).
The Ala 960 mutation does appear to affect the interaction of the insulin receptor cytoplasmic domain with IRS-2. Western blotting revealed that the CSF1R/IR co-precipitated with IRS-2, but the CSF1R/IRA960 was not detected in anti-IRS-2 immunoprecipitates. These data indicate that the juxtamem-brane domain is a point of interaction between the insulin receptor cytoplasmic domain and IRS-2 as demonstrated previously (35,36). The observation that IRS-2 is still tyrosine-phosphorylated upon activation of the CSF1R/IRA960 may emphasize the importance of additional domains of interaction between the insulin receptor cytoplasmic domain and IRS-2. Recent observations that IRS-1 and IRS-2 are distributed differently in 3T3-L1 cells and that they translocate from intracellular membrane compartments to cytosol after insulin stimulation (37) also raises the possibility of differential interactions between the insulin receptor cytoplasmic domain and these intracellular substrates.
Two-hybrid analyses (35,36) have demonstrated that IRS-2 interacts with phosphotyrosines in the insulin receptor kinase activation loop through a region termed the kinase regulatory loop binding domain (38). Furthermore, it has been suggested that phosphorylation of Tyr 624 and Tyr 628 within the kinase regulatory loop binding domain of IRS-2 may inhibit interaction of this domain with the insulin receptor kinase (38). If these data are correct, then the phosphorylation-induced decrease in affinity between the insulin receptor and IRS-2 might explain the inability to co-precipitate IRS-2 and the CSF1R/ IRA960 (Fig. 5).
The insulin receptor has multiple intracellular substrates (39 -45) that may serve as distinct signaling intermediates for the pleiotropic effects of insulin. The ability of the insulin FIG. 5. Interaction of IRS-2 with the CSF1R/IR and the CSF1R/IRA960. 3T3-L1 adipocytes expressing the CSF1R/IR or the CSF1R/IRA960 were treated with or without 10 nM CSF-1 and lysed. Identical samples were immunoprecipitated with antibodies against IRS-2 or against the carboxyl terminus of the insulin receptor cytoplasmic domain (CT-1). Immunoprecipitated proteins were resolved by SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and detected on blots with antibodies against phosphotyrosine, followed by stripping and reprobing the blots with antibody CT-1 or with anti-IRS-2 antibody. M.W., molecular mass.
FIG. 6. Activation of PI 3 kinase associated with IRS proteins or tyrosine-phosphorylated proteins in 3T3-L1 adipocytes. 3T3-L1 adipocytes expressing the CSF1R/IR or the CSF1R/IRA960 were left untreated or treated with 10 nM CSF-1 or 100 nM insulin and lysed. Cell lysates were immunoprecipitated with antibodies to IRS-1, IRS-2, or phosphotyrosine. Immunoprecipitates were analyzed for the presence of PI 3Ј kinase activity as described under "Experimental Procedures." Quantitative analysis of PI 3Ј kinase activity associated with immunoprecipitates of IRS-1, IRS-2, and tyrosine-phosphorylated proteins was performed using storage phosphor technology. The data are the mean ϮS.D. of four independent experiments. receptor kinase to mediate diverse signals through distinct intracellular substrates may be determined not only by the phosphorylation state of the substrate but by the nature of the signaling complex formed upon interaction between the receptor and each substrate. The data presented here demonstrate the utility of the chimeric CSF1R/IR for probing the molecular details of the interactions between the insulin receptor kinase and its various intracellular effectors.