The COOH-terminal Tyrosine Phosphorylation Sites on IRS-1 Bind SHP-2 and Negatively Regulate Insulin Signaling*

Activation of tyrosine kinases by numerous growth factor and cytokine receptors leads to tyrosine phosphorylation of the insulin receptor substrate (IRS)-proteins. Tyrosine-phosphorylated motifs on the IRS proteins bind to the SH2 domains in proteins that mediate downstream signals, including phosphatidylinositol 3′-kinase, GRB-2, and SHP-2. We investigated the function of the two SHP-2 binding COOH-terminal tyrosines of IRS-1 by replacing them with phenylalanine (IRS-1FCT). IRS-1FCT failed to bind SHP-2 or mediate its tyrosine phosphorylation during insulin stimulation. Although several reports suggest a critical role for SHP-2 in insulin stimulated mitogen-activated protein kinase activation and cell proliferation, IRS-1FCT mediated these effects normally in 32D cells. Indeed, IRS-1FCT exhibited increased tyrosine phosphorylation, phosphatidylinositol 3′-kinase binding and activation of protein synthesis in response to insulin. These results suggest that SHP-2 attentuates the phosphorylation and downstream signal transmission of IRS-1 and that the interaction of IRS-1 and SHP-2 is an important regulatory event which attenuates insulin metabolic responses.

Insulin transmits cellular signals by binding the cell surface insulin receptor and activating the insulin receptor tyrosine kinase (1). The activated tyrosine kinase transmits downstream signals largely through the action of a class of intermediary docking proteins, including GAB-1, Shc, and a set of signaling molecules known as the insulin receptor substrate (IRS) 1 proteins, IRS-1, IRS-2, IRS-3, and IRS-4 (2-6). The NH 2 -terminal region of the IRS proteins contains extended domains that cooperate in receptor recognition (1,2,7), while the COOH-terminal region possesses numerous tyrosine-containing motifs that become phosphorylated and bind specific Src homology 2 (SH2) domains in signaling proteins during insulin stimulation. While each IRS protein contains unique tyrosine phosphorylation sites, three classes of sites are well conserved across the IRS proteins, multiple YMXM motifs responsible for binding regulatory subunits of phosphatidylin-sositol (PI) 3Ј-kinase, a YXNX motif that binds GRB-2, and a pair of COOH-terminal motifs (YIDL and YASI) responsible for binding the SHP-2 tyrosine phosphatase. The role of the PI 3Ј-kinase and GRB-2 binding motifs on IRS-1 have been extensively studied, but the role of the SHP-2 binding sites in IRS-1 signaling is unclear (1).
SHP-2 is a cytoplasmic protein tyrosine phosphatase that contains two SH2 domains (8). It is ubiquitously expressed in mammalian cells and appears to be the homolog of the Drosophila corkscrew gene responsible for activation of p21 ras downstream of a variety of tyrosine kinases (9,10). Coordinate binding of the two SH2 domains of SHP-2 to double tyrosine motifs in target proteins such as IRS-1 increases the affinity of the association and activates the phosphatase by moving the NH 2 -terminal SH2 domain away from the phosphatase active site (11)(12)(13)(14).
Homozygous disruption of SHP-2 in mice is early embryonic lethal, suggesting that SHP-2 plays a critical role in cell growth or differentiation. Consistent with this result, microinjection or overexpression of a catalytically inactive SHP-2, the SHP-2 SH2 domains, or ␣SHP-2 antibodies inhibits insulin-stimulated DNA synthesis and MAP kinase activation in a number of cell types (15)(16)(17)(18). These approaches, however, interfere with all SHP-2-mediated signaling, rather than the IRS-1/SHP-2 signaling pathway specifically.
In this report, we examine signaling by a mutant IRS-1 molecule that lacks the COOH-terminal tyrosine phosphorylation sites and does not bind SHP-2. During insulin stimulation, this mutant is more highly phosphorylated and more strongly activates the PI 3Ј-kinase pathway, including protein synthesis, than does wild-type IRS-1. In contrast, MAP kinase activation and DNA synthesis occur normally. Thus, although SHP-2 may serve multiple roles, its association with IRS-1 appears to attenuate certain insulin signals important for metabolic responses.

MATERIALS AND METHODS
Cell Lines-Chinese hamster ovary cells overexpressing the human insulin receptor (CHO IR ) or the human insulin receptor and rat IRS-1 (CHO IR /IRS-1) have been described (19). CHO IR cells were maintained in Ham's F-12 medium (Life Technologies, Inc.) containing 10% fetal bovine serum (Sigma) and were made quiescent for experiments by incubation in F-12 medium containing 0.5% bovine serum albumin for 12-18 h. The mutant cDNA for IRS-1 FCT was co-transfected into CHO IR cells with pHyg (containing the hygromycin B phosphotransferase gene) by calcium phosphate precipitation (20). The transfected cells were selected in medium containing 300 g/ml hygromycin (ICN, Aurora, Ohio).
32D cell lines expressing the human insulin receptor (32D IR ) and/or rat IRS-1 have been described (21,22); these cells were maintained in RPMI 1640 medium (Life Technologies, Inc.) containing 10% fetal bovine serum and 5% WEHI-3 conditioned medium (a source of IL-3). 32D cell lines were made quiescent by incubation in unsupplemented Dul-becco's modified Eagle's medium for 3-4 h before experiments. 32D cell lines were transfected with the cDNAs for isoforms of IRS-1 by electroporation (21,22) and were selected and maintained in the presence of 10 mM histidinol (Bachem).
Antibodies and Growth Factors-Bovine insulin was from Calbiochem. Antibodies directed against SHP-2 were rabbit polyclonal antisera raised against human SHP-2 expressed as a glutathione S-transferase fusion protein (8). Antiphosphotyrosine antibodies (␣PY) were rabbit polyclonal antibodies affinity-purified on phosphotyrosine (23) or monoclonal antibody 4G10 (Upstate Biotechnology Inc., Lake Placid, NY). ␣IRS-1 antibodies were rabbit polyclonal antisera raised against the amino-terminal 135 amino acids of rat IRS-1 expressed as a glutathione S-transferase fusion protein (24) or against synthetic peptides comprising amino acids 1-15 of rat IRS-1 coupled to keyhole limpet hemocyanin (25). ␣PhosphoMAP kinase antibody was from New England Biolabs.
Generation of IRS-1 FCT cDNA-The cDNA for rat IRS-1 in pBluescript (3) was used as a template for polymerase chain reaction-mediated oligonucleotide-directed mutagenesis (26). The mutagenic primer 5Ј-AAG/AGT/CTT/AAC/TTT/ATA/GAC/TTG/GAT-3Ј and its complement were used with outside primers flanking the EcoRI and AatII sites in the rat IRS-1 cDNA to mutate Tyr 1172 . This product was subjected to another round of mutagenesis using the mutagenic primers 5Ј-GAT/ TTA/AGC/ACC/TTT/GCC/AGC/ATC/ACC-3Ј and its complement and the same outside primers. The rat IRS-1 cDNA in pBluescript and mutant polymerase chain reaction products were digested with EcoRI and AatII, and the mutant polymerase chain reaction fragments were inserted in place of the wild-type sequence. The presence of the desired mutations and the absence of adventitious mutations was confirmed by DNA sequencing of the mutant cDNA constructs. All mutant cDNAs were sucloned into the pCMVhis expression vector for expression in mammalian cells using the restriction enzymes SacI and HindIII.
Immunoprecipitation-Quiescent confluent CHO IR cells growing on 15-cm dishes were incubated in the absence or presence of insulin as indicated, the medium was aspirated, and cells were immediately washed with ice-cold PBS. Quiescent 32D IR cells were stimulated as indicated and collected by centrifugation after the addition of ice-coldphosphate-buffered saline. Cells were lysed in 137 mM NaCl, 20 mM Tris, 1 mM NaCl 2 , 1 mM CaCl 2 , pH 7.5, supplemented with 10% glycerol, 1% Nonidet P-40, 2 mM phenylmethylsulfonyl fluoride, 0.01 mg/ml aprotinin, 0.01 mg/ml leupeptin, and 0.1 mM sodium orthovanadate. Nuclei and insoluble materials were removed by centrifugation at 10,000 ϫ g for 15 min. Supernatants were incubated at 4°C with antibodies for several hours to overnight before being collected with protein A-Sepharose 6 MB (Amersham Pharmacia Biotech). Finally, immunoprecipitates were washed three times with lysis buffer, denatured by boiling in Laemmli sample buffer for 5 min, and separated by SDS-PAGE.
Phosphatidylinositol 3Ј-Kinase Activity-In vitro phosphorylation of PI was carried out in immune complexes as described previously (27). Cells were made quiescent, stimulated, lysed, and immunoprecipitated as above. Immune complexes were precipitated from the supernatant with protein A-Sepharose (Amersham Pharmacia Biotech) and washed (27). Immune complexes were incubated with phosphatidylinositol (Avanti) and [␥-32 P]ATP for 10 min at 22°C. The reactions were stopped with 20 l of 8 N HCl and 160 l of CHCl 3 :methanol (1:1) and centrifuged, and the lower organic phase was removed and applied to a silica gel TLC plate (Merck) that had been coated with 1% potassium oxalate (27). TLC plates were developed in CHCl 3 :CH 3 OH:H 2 O:NH 4 OH (60:47: 11.3:2), dried, and visualized and quantitated on a Molecular Dynamics PhosphorImager.
Immunoblotting-Proteins in cell lysates prepared as for PI 3Ј-kinase assays or from immunoprecipitates similarly prepared were denatured by boiling in Laemmli sample buffer containing 100 mM dithiothreitol and resolved by SDS-PAGE. Gels were transferred to nitrocellulose membranes (Schleicher & Schuell) in Towbin buffer containing 0.02% SDS and 20% methanol (28). Membranes were blocked, probed, and developed as described previously (29,30). Blots were exposed to Kodak XAR film or imaged on a Molecular Dynamics PhosphorImager.

Incorporation of [ 3 H]Thymidine into DNA in 32D
Cells-Insulinstimulated thymidine incorporation was assayed as described previously (22,29). Briefly, cells in log phase growth were washed and seeded into RPMI with 10% fetal bovine serum alone, 5% IL-3-containing WEHI-3-conditioned medium, or with various concentrations of insulin. Cells were grown for 48 h at 37°C, [ 3 H]thymidine (ICN) was added to 0.5 mCi/ml, and incubation was continued for 3 h. Cells were collected onto glass microfiber filters and lysed, and unincorporated nucleotide was removed by repeated washing with water. Filters were dried and counted in scintillation fluid for 1 min.

S]Methionine Incorporation into Protein in 32D
Cells-32D cell lines were maintained at concentrations of 5 ϫ 10 5 cells/ml and starved in RPMI 1640 medium supplemented with 10% fetal bovine serum for 6 h. Following starvation, 1-ml aliquots of cells were taken and incubated in the presence of IL-3 or various concentrations of insulin. Cells were incubated with [ 35 S]methionine (10 Ci/aliquot) in methioninefree RPMI 1640 for 60 min, washed once in phosphate-buffered saline containing 10 mM methionine, and lysed in 0.5 M NaOH for 30 min at 37°C. Protein was precipitated with 12% ice-cold trichloroacetic acid containing 10 mM methionine, collected on glass filters, and washed with 5% trichloroacetic acid and ethanol. Filters were air-dried, and incorporated radionuclide was quantitated by liquid scintigraphy in an aqueous fluor.

RESULTS
The COOH-terminal Tyrosine Phosphorylation Sites in IRS-1 Mediate Interactions with SHP-2 and the Tyrosine Phosphorylation of SHP-2-In order to investigate the role of IRS-1/SHP-2 interactions in insulin signaling, we generated IRS-1 FCT , in which the two predicted SHP-2-binding residues (Tyr 1172 and Tyr 1222 ) on IRS-1 were replaced by phenylalanine, and we expressed IRS-1 FCT in CHO IR and 32D IR cells at similar levels as wild-type IRS-1 (Fig. 1). We assayed insulin-stimulated tyrosine phosphorylation of IRS-1 isoforms in the CHO IR cells by ␣PY immunoblotting immunoprecipitated IRS-1. While tyrosine phsophorylation of endogenous hamster IRS-1 was weakly detected in the CHO IR cells, robust insulin-stimulated tyrosine phosphorylation of overexpressed IRS-1 and IRS-1 FCT was readily detected (Fig. 2A).
Insulin-stimulated Tyrosine Phosphorylation of SHP-2-␣PY immunoblots of ␣SHP-2 immunoprecipitates from CHO IR cells revealed that tyrosine phosphorylated IRS-1 associated with SHP-2 during insulin stimulation, consistent with previous results (32,33) (Fig. 2B). In addition, overexpression of wild-type IRS-1 greatly increased the amount of SHP-2-associated tyrosine-phosphorylated IRS-1. Furthermore, SHP-2 became tyrosine-phosphorylated during insulin stimulation of CHO IR cells, and the overexpression of wild-type IRS-1 similarly increased the tyrosine phosphorylation of SHP-2 during insulin stimulation. The identity of the approximately 60-kDa tyrosyl phosphoprotein in ␣SHP-2 immunoprecipitates was confirmed to be SHP-2 by reblotting with ␣SHP-2 antibodies and demonstrating that the bands overlay (data not shown). In contrast, IRS-1 FCT did not associate with SHP-2; nor did IRS-1 FCT increase the tyrosine phosphorylation of SHP-2. Thus, the COOH-terminal tyrosine phosphorylation sites in IRS-1 mediated the binding of SHP-2 and IRS-1 in these cells, and this interaction mediated the tyrosine phosphorylation of SHP-2 during insulin stimulation. A 95-100-kDa phosphoprotein (p100) also associated with ␣SHP-2 immunoprecipitates during insulin stimulation; its association similarly depended upon the presence of the COOH-terminal tyrosines on IRS-1 Insulin-stimulated Tyrosine Phosphorylation and PI 3Ј-Kinase Binding of IRS-1 Isoforms-In order to examine the function of IRS-1 FCT during insulin signaling, we used the 32D IR cells, which contain no endogenous IRS-proteins (21,22). We first examined tyrosine phosphorylation of IRS-1 and IRS-1 FCT during insulin stimulation in the 32D IR cell lines (Fig. 3). At all doses of insulin the tyrosine phosphorylation of IRS-1 FCT was greater than that of wild-type IRS-1 (Fig. 3); after normalizing for variations in expression of IRS-1 and IRS-1 FCT , the phosphotyrosine content of IRS-1 FCT was approximately 2-fold greater than for wild-type IRS-1 during stimulation with a range of insulin concentrations (Fig. 3, graph).
We investigated whether the increased insulin-stimulated tyrosine phosphorylation of IRS-1 FCT enhanced IRS-1 downstream signals by assaying PI 3Ј-kinase activity in ␣IRS-1 immunoprecipitates: Indeed, PI 3Ј-kinase activity associated with IRS-1 FCT was increased compared with wild-type IRS-1, and to a similar extent as tyrosine phosphorylation of IRS-1 FCT was increased (Fig. 4A).
We furthermore assessed downstream signaling by IRS-1 FCT by assaying generalized insulin-stimulated protein synthesis, which requires IRS-1-mediated PI 3Ј-kinase activity in the 32D cells (34,35) (Fig. 4B). As previously shown, expression of the insulin receptor alone in 32D IR cells was insufficient for the stimulation of protein synthesis by insulin, whereas co-expression of IRS-1 in these cells mediated insulin-stimulated protein synthesis. As with tyrosine phosphorylation and PI 3Ј-kinase binding, IRS-1 FCT enhanced insulin-stimulated protein synthesis compared with wild-type IRS-1. Therefore, the inhibitory effect of Tyr 1172 and Tyr 1222 on tyrosine phosphorylation and PI 3Ј-kinase activity carries over to the stimulation of protein synthesis.
MAP Kinase and Proliferative Signaling by IRS-1 FCT -The insulin receptor partially activates MAP kinase via the Shc/ GRB-2 pathway in 32D IR cells and expression of IRS-1 enhances this signal by recruiting GRB-2 to Tyr 895 in IRS-1 (29). Since reagents which interfere with SHP-2 block insulin-stim-

FIG. 1. Construction and expression of COOH-terminal tyrosine mutants of IRS-1.
A, Schematic diagram of IRS-1. Shown is a linear diagram of IRS-1, including the pleckstrin homology (PH) and phosphotyrosine binding (PTB) domains required for upstream kinase interactions and 18 potential tyrosine phosphorylation sites (* ϭ known tyrosine phosphorylation site). IRS-1 F*CT is shown below IRS-1, with the two mutated COOH-terminal tyrosine residues in boxes. Expression of IRS-1 and IRS-1 mutants in B, CHO IR and C, 32D IR cells. Lysates of control (-) cells and cells expressing IRS-1 (IRS-1) or IRS-1 FCT (FCT) were resolved by SDS-PAGE and immunoblotted with ␣IRS-1 before being exposed to autoradiography. There are two nonspecific bands in panel C (90 and 170 kDa) due to the cross-reactivity of the antisera when used in 32D cells.

FIG. 2. Insulin-stimulated tyrosine phosphorylation and SHP-2 association of IRS-1 isoforms in CHO IR cells.
Quiescent CHO IR cell lines expressing the indicated IRS-1 isoforms were incubated for 5 min in the absence (Ϫ) or presence (ϩ) of 100 nM insulin before being lysed. Lysates were clarified and immunoprecipitated with either A, ␣IRS-1 or B, ␣SHP-2 antisera. Immunoprecipitates were collected, washed, and resolved by SDS-PAGE before being analyzed by immunoblotting with the indicated antisera and being exposed to autoradiography. Migration of molecular weight standards is shown on the left and the position of SHP-2, IRS-1, and p100 are indicated on the right. ulated MAP kinase signaling, we assayed the insulin-stimulated activation of MAP kinase in 32D IR cells expressing the SHP-2 binding-defective IRS-1 FCT (Fig. 5A). Insulin stimulated the activation of MAP kinase in 32D IR cells, and IRS-1 expression increased the sensitivity and intensity of the response. Furthermore, insulin stimulated MAP kinase activity similarly in 32D IR /IRS-1 FCT and 32D IR /IRS-1 cells, demonstrating that IRS-1 and IRS-1 FCT signal similarly to MAP kinase in 32D IR cells. Therefore, the SHP-2-binding COOH-terminal tyrosine phosphorylation sites in IRS-1 are not involved in the insulinstimulated MAP kinase signaling pathway in the 32D IR cells.
Although insulin receptor expression alone mediates some activation of MAP kinase in the 32D cells, insulin-stimulated DNA synthesis, like protein synthesis, requires IRS-1 expres-sion (Fig. 5B). In contrast to the effects observed in signaling to protein synthesis, however, IRS-1 FCT did not increase the proliferative response to insulin compared with wild-type IRS-1. Thus, the COOH-terminal tyrosines on IRS-1 mediate neither positive nor negative proliferative signals in 32D IR cells. DISCUSSION We have examined the role of the COOH-terminal tyrosines (Tyr 1172 and Tyr 1222 ) of IRS-1 for interactions with SHP-2 during insulin signaling. When phosphorylated by the insulin receptor, the motif containing Tyr 1172 binds to the NH 2 -terminal SH2 domain and the motif containing Tyr 1222 binds to the COOH-terminal SH2 domain of SHP-2 (11). The binding of these motifs to the SH2 domains of SHP-2 activates the tyro- sine phosphatase activity of the SHP-2 molecule by altering the conformation of the NH 2 -terminal SH2 domain such that the active site of SHP-2 becomes available (12)(13)(14)(31)(32)(33). Thus, SHP-2 likely regulates insulin signaling by dephosphorylating tyrosine residues on IRS-1 or associated molecules.
Previous work on IRS-1 and SHP-2 has focused largely on interactions between purified proteins. Here we confirm in intact cells that the association of IRS-1 and SHP-2 depends upon Tyr 1172 and Tyr 1222 . Using the yeast two-hybrid system or purified proteins in vitro, others have suggested that SHP-2 and the insulin receptor can interact directly, without IRS-1 (36,37). Our data in CHO IR cells suggest that IRS-1 is necessary to link SHP-2 to the insulin receptor; insulin stimulates the tyrosine phosphorylation of SHP-2 when it is associated with IRS-1. The simplest interpretation of this data is that SHP-2 is phosphorylated when recruited to the insulin receptor-IRS-1 complex by the phosphorylated Tyr 1172 and Tyr 1222 of IRS-1. Furthermore, although previous reports suggest that SHP-2 enhances the coupling between IRS-1 and the insulin receptor, IRS-1 FCT , which does not bind SHP-2, mediates signaling as well or better than wild-type IRS-1. Thus SHP-2 is not required for insulin receptor-IRS-1 coupling in this system.
Numerous studies have suggested that SHP-2 activity is required for the insulin-stimulated activation of MAP kinases and proliferation. We examined the function of the SHP-2binding defective IRS-1 FCT in insulin signal transmission in 32D IR cells, as these cells contain no endogenous IRS proteins and facilitate the analysis of IRS-1 mutants (21,22); IRS-1 FCT mediates MAP kinase activation and proliferation similarly to wild-type IRS-1 in the 32D IR cells. Thus, at least in this system, IRS-1/SHP-2 interactions are not required for these two insulin-mediated events. In contrast, mutation of the SHP-2 binding site in IRS-1 FCT results in enhanced insulin-stimulated tyrosine phosphorylation of IRS-1; this increases the association of PI 3Ј-kinase with IRS-1 FCT , which in turn enhances insulin-stimulated protein synthesis. These data suggest a negative regulatory role for the COOH-terminal tyrosine phosphorylation sites on IRS-1 and, by extension, for the IRS-1/SHP-2 interaction.
A reasonable model for this effect is that tyrosine phosphorylation of the COOH-terminal sites on IRS-1 recruits and activates the SHP-2 phosphatase in complex with IRS-1. In this complex, SHP-2 dephosphorylates IRS-1 and/or the insulin receptor, reducing tyrosine phosphorylation of PI 3Ј-kinase binding YMXM motifs in IRS-1 (Fig. 6); indeed, others have shown that dephosphorylation of IRS-1 by SHP-2 in vitro is mediated by the IRS-1/SHP-2 interaction (38). Hence, our model of SHP-2/IRS-1 interactions is consistent with the emerging model of SHP-2/gp130 interaction; mutation of the SHP-2 binding sites in gp130 (the shared subunit of many IL-6-type cytokine receptors) prolongs activation of the associated JAK tyrosine kinase and the downstream STAT signaling proteins during oligomerization of gp130 or during CNTF signaling (39 -41).
FIG. 6. Model of SHP-2 function in IRS-1 signaling. Activation of the insulin receptor causes tyrosine phosphorylation of multiple sites on IRS-1, including PI 3Ј-kinase and SHP-2 binding sites. PI 3Ј-kinase binding causes activation of downstream Ser/Thr kinases such as protein kinase B (PKB), p70s6k, and protein kinase C (PKC). Activation of protein kinase C controls insulin-stimulated general protein synthesis. SHP-2 binding to IRS-1 causes dephosphorylation of some PI 3Јkinase-binding YMXM motifs, reducing the strength of the protein synthesis signal. Unlike protein synthesis, mitogenesis is not controlled only by PI 3Ј-kinase, but likely requires only a certain threshold level of PI 3Ј-kinase activation in conjunction with other, undefined signals mediated by IRS-1. SHP-2 also binds and dephosphorylates other classes of tyrosyl phosphoproteins, including the SIRP/SHPS proteins and an approximately 95-100-kDa GAB-1-like tyrosyl-phosphorylated docking protein (p95-100) (42)(43)(44)(45)(46)(47)(48)(49)(50)(51). Although, like IRS-1, these SHP-2 binding proteins are dephosphorylated by SHP-2, dephosphorylation in this case correlates with increased MAP kinase activation and proliferation. Thus, while SHP-2 dephosphorylates a variety of bound tyrosyl phosphoproteins, this dephosphorylation may either inhibit downstream signals (as in the case of IRS-1 and gp130) or mediate positive signals as in the case of SIRP/SHPS and p95-100.
While our data show no link between IRS-1/SHP-2 interactions and MAP kinase and proliferative signaling, a large body of data suggests that SHP-2 is required for insulin-stimulated activation of the p21 ras 3 MAP kinase 3 c-Fos pathway and proliferation (15)(16)(17)(18)52). These studies, however, have investigated SHP-2 signaling by overexpressing or microinjecting reagents (such as catalytically inactive SHP-2 mutants, SHP-2 SH2 domains, or ␣SHP-2 antibodies) which interfere with SHP-2 signaling. This approach blocks not only the SHP-2/ IRS-1 interaction, but the interaction of SHP-2 with other proteins, such as SIRP/SHPS and p95-100. In contrast, we have examined the SHP-2/IRS-1 interaction specifically. Thus, we are able to draw conclusions specifically about the role of the SHP-2/IRS-1 interaction without disrupting signaling by other important SHP-2 interacting molecules which may mediate positive signals.
The function of IRS-1 FCT also highlights differences between the signaling pathways connecting IRS-1 to protein synthesis and proliferation (Fig. 6). Although both pathways require PI 3Ј-kinase, the increased PI 3Ј-kinase activity generated by the IRS-1 FCT enhances insulin-stimulated protein synthesis without altering insulin-stimulated proliferation. Similarly, an IRS-1 mutant whose only tyrosines are three YMXM motifs (and thus lacks the other 15 tyrosines in wild-type IRS-1, including six other YMXM motifs and both COOH-terminal tyrosine) mediates enhanced activation of PI 3Ј-kinase and protein synthesis, but impaired proliferation compared with wild-type IRS-1 (34,53). Furthermore, insulin receptor variants which bind and activate PI 3Ј-kinase directly mediate protein synthesis but not proliferation in response to insulin (54). 2 Thus, activation of PI 3Ј-kinase and its downstream pathways suffices to stimulate general protein synthesis, while signals in addition to PI 3Ј-kinase (and which are not modified by the COOH-terminal tyrosine phosphorylation sites) are required for proliferation. A certain level of PI 3Ј-kinase may be required for proliferation, but PI 3Ј-kinase independent pathways mediated by IRS-1 must be limiting for this effect.
PI 3Ј-kinase and its downstream mediators control insulin stimulated protein synthesis directly, and may suffice to control other aspects of energy storage, such as glucose transport, as well (1,35,55,56). The COOH-terminal tyrosine phosphorylation sites on IRS-1 may therefore offer potential therapeutic targets for the treatment of insulin resistance, since they appear to regulate PI 3Ј-kinase-mediated metabolic signals without affecting proliferative signaling.