Human Insulin Receptor Substrate-1 (IRS-1) Polymorphism G972R Causes IRS-1 to Associate with the Insulin Receptor and Inhibit Receptor Autophosphorylation*

The most commonly detected polymorphism in human insulin receptor substrate-1 (IRS-1), a glycine to arginine change at codon 972 (G972R), is associated with an increased risk of Type 2 diabetes and insulin resistance. To determine the molecular mechanism by which this polymorphism may be linked to insulin resistance, we produced recombinant peptides comprising amino acid residues 925–1008 from IRS-1 that contain either a glycine or arginine at codon 972 and the two nearby tyrosine phosphorylation consensus sites (EY941MLM and DY989MTM), which are known binding sites for the p85α regulatory subunit of phosphatidylinositol 3-kinase. The wild type peptide could be phosphorylated at these sites in vitro by purified insulin receptor. Introduction of the G972R polymorphism into the peptide reduced the amount of tyrosine phosphorylation by >60%. Pull-down experiments indicated that there was an association between the IRS-1-(925–1008) peptide and the insulin receptor that was markedly enhanced by the presence of the G972R polymorphism. The use of additional overlapping fragments localized this interaction to domains between residues 950–986 of IRS-1 and residues 966–1271 of the insulin receptor, containing the tyrosine kinase domain of the receptor. In addition, the IRS-1-(925–1008) G972R peptide acted as a competitive inhibitor of insulin receptor and insulin-like growth factor-1 receptor autophosphorylation. Taken together, these data indicate that the G972R naturally occurring polymorphism of IRS-1 not only reduces phosphorylation of the substrate but allows IRS-1 to act as an inhibitor of the insulin receptor kinase, producing global insulin resistance.

The most commonly detected polymorphism in human insulin receptor substrate-1 (IRS-1), a glycine to arginine change at codon 972 (G972R), is associated with an increased risk of Type 2 diabetes and insulin resistance. To determine the molecular mechanism by which this polymorphism may be linked to insulin resistance, we produced recombinant peptides comprising amino acid residues 925-1008 from IRS-1 that contain either a glycine or arginine at codon 972 and the two nearby tyrosine phosphorylation consensus sites (EY 941 MLM and DY 989 MTM), which are known binding sites for the p85␣ regulatory subunit of phosphatidylinositol 3-kinase. The wild type peptide could be phosphorylated at these sites in vitro by purified insulin receptor. Introduction of the G972R polymorphism into the peptide reduced the amount of tyrosine phosphorylation by >60%. Pulldown experiments indicated that there was an association between the IRS-1-(925-1008) peptide and the insulin receptor that was markedly enhanced by the presence of the G972R polymorphism. The use of additional overlapping fragments localized this interaction to domains between residues 950 -986 of IRS-1 and residues 966 -1271 of the insulin receptor, containing the tyrosine kinase domain of the receptor. In addition, the IRS-1-(925-1008) G972R peptide acted as a competitive inhibitor of insulin receptor and insulin-like growth factor-1 receptor autophosphorylation. Taken together, these data indicate that the G972R naturally occurring polymorphism of IRS-1 not only reduces phosphorylation of the substrate but allows IRS-1 to act as an inhibitor of the insulin receptor kinase, producing global insulin resistance.
Type 2 diabetes is a complex disease involving both impaired insulin secretion and peripheral insulin resistance impacted by genetics and the environment (1). Insulin receptor substrate-1 (IRS-1) 1 functions as one of the key downstream signaling molecules in both the insulin receptor and the insulin-like growth factor-1 receptor signaling pathways (IGF-1R). Thus genetic changes in IRS-1 may potentially contribute toward the development of insulin resistance. Screening of IRS-1 has identified a number of genetic variants, the most common of these being a glycine to arginine change at codon 972 (G972R) (Fig. 1) (2). This polymorphism is found in ϳ5% of the normal population and 10% of Type 2 diabetics (3) and carriers of the G972R variant have a 25% increased risk for developing Type 2 diabetes (4).
The G972R polymorphism appears to be associated with obesity because it is found more commonly among obese populations compared with lean populations (5), and decreased insulin sensitivity has been observed in obese carriers of the polymorphism (6). Normal glucose-tolerant subjects carrying the G972R change have also been shown to have decreased insulin secretion (7). In addition, an association with an increased risk of developing coronary artery disease has been reported in people with this polymorphism (8 -10), and two recent studies have also demonstrated an association between this polymorphism and Type 1 diabetes (11,12).
After autophosphorylation of the insulin receptor, the receptor kinase is activated and phosphorylates IRS-1 and other intracellular substrates. This signaling molecule then acts as a docking protein for multiple Src homology-2 domain (SH2)containing proteins, including PI 3-kinase, Grb-2, and SHP2 (13). In vitro studies in 32D(IR) cells have demonstrated that when the G972R polymorphism was expressed, it caused impaired insulin-stimulated signaling along the PI 3-kinase pathway (14). Other studies in different tissues have shown altered glucose metabolism in muscle over-expressing IRS-1 G972R (15), defects in insulin secretion and increased susceptibility to apoptosis in islets from G972R carriers (16 -18), and impaired regulation of Akt/endothelial nitric-oxide synthase activation in endothelial cells from G972R carriers (19), demonstrating that this polymorphism has detrimental effects in multiple insulin-sensitive tissues.
In this study we have investigated the detailed molecular mechanism by which this polymorphism may be linked to insulin resistance. The G972R polymorphism is found near the C terminus of IRS-1 flanked by two tyrosine phosphorylation consensus sites (EY 941 MLM and DY 989 MTM), which are known binding sites for the p85␣ regulatory subunit of PI 3-kinase (Fig. 1). A series of recombinant fragments from this region of human IRS-1 were produced with either a glycine or an arginine at codon 972. These included fragments containing amino acid residues 925-1008, both of the flanking tyrosine phosphorylation consensus sites and overlapping fragments containing residues 950 -1027 or 910 -986, which contain either the downstream (DY 989 MTM) or the upstream (EY 941 MLM) tyrosine phosphorylation consensus sites, respectively (Fig. 2). These fragments were used in in vitro experiments with purified insulin receptor to investigate the mechanism by which the G972R polymorphism contributes to altered insulin signaling.

EXPERIMENTAL PROCEDURES
Cloning of the IRS-1 Fragments-DNA fragments (nucleotides 3793-4044, 3868 -4101, and 3748 -3975 of wild type human IRS-1 cDNA) corresponding to amino acid residues 925-1008, 950 -1027, and 910 -986 respectively were amplified using standard PCR techniques with Pfu polymerase (Stratagene, La Jolla, CA). Each of the primers contained a restriction site (BamHI in the forward primer, HindIII in the reverse primer) at the 5Ј end for cloning purposes. After digestion, the amplified DNA was cloned into BamHI/HindIII-cleaved pQE30 vector (Qiagen, Valencia, CA). Plasmid pQE30 contains a sequence encoding a His 6 tag located upstream of the cloned fragments. The G972R polymorphism (codon change GGG 3 CGG) was introduced into each of the wild type plasmids using PCR-based site-directed mutagenesis, and the clones were sequenced to confirm introduction of the polymorphism. The DNA was used to transform Escherichia coli strain M15 containing a lac repressor plasmid (pRep4, Qiagen).
Production of Recombinant IRS-1 Fragments-Recombinant clones were grown at 37°C in LB containing 25 g/ml kanamycin and 100 g/ml ampicillin. Expression of the recombinant protein was induced by the addition of isopropyl 1-thio-␤-D-galactopyranoside to a concentration of 1 mM at the end of the log phase. Growth was continued for an additional 4 h. Cells were lysed in loading buffer (100 mM NaH 2 PO 4 , 10 mM Tris, 300 mM NaCl, and 10 mM imidazole, pH 8) on ice, sonicated to reduce viscosity, and clarified by centrifugation at 10,000 rpm for 30 min. The cleared lysate was incubated with Talon cobalt metal-affinity resin (BD Biosciences). The resin bound to the His-tagged protein was then loaded on to a column, washed exhaustively with loading buffer, and eluted with loading buffer supplemented with 150 mM imidazole, 1 mM dithiothreitol, and 25 mM EDTA. Eluted protein was dialyzed against 50 mM Tris, 33 mM NaCl, 0.66 mM dithiothreitol, pH 7.4.
Purification of the Insulin-like Growth Factor-1 Receptor-Insulin receptor knock-out brown preadipocytes stably over-expressing human IGF-1R were lysed in 10 mM Tris HCl, 1% Triton X-100, 0.5% Nonidet P-40, 150 mM NaCl, 10 mM sodium orthophosphate, 10 mM sodium pyrophosphate, 10 mM sodium orthovanadate, 100 nM sodium fluoride, 1 mM EDTA, 1 mM EGTA, and complete protease inhibitor tablets (one per 50 ml, pH 7.6, Roche Applied Science). After sonication, the cell lysate was centrifuged at 3,000 rpm for 5 min, and the cleared lysate was loaded onto a wheat germ agglutinin-agarose column (Amersham Biosciences) and passed through three times. The column was washed exhaustively with 50 mM HEPES, 150 mM NaCl, and 0.1% Triton X-100, pH 7.6, and the IGF-1R was eluted from the column using wash buffer supplemented with 0.6 M N-acetylglucosamine (Sigma).
In Vitro Phosphorylation Assays-Purified full-length insulin receptor (kind gift of Jongsoon Lee, Joslin Diabetes Center) was incubated with 100 nM insulin for 60 min at room temperature and was then supplemented with 50 M ATP, 10 mM MgCl 2 , and 8 mM MnCl 2 (final concentrations). Recombinant IRS-1 fragments were added, and a time course of phosphorylation was performed. To investigate the inhibition of insulin receptor autophosphorylation, we incubated recombinant IRS-1 fragments with the insulin-activated insulin receptor prior to the addition of 50 M ATP, 10 mM MgCl 2 , and 8 mM MnCl 2 and performed a time course. At each time point, the reaction was stopped by the addition of Laemmli's sample buffer containing 100 mM dithiothreitol, and the samples were analyzed by SDS-PAGE. A similar protocol was used for the phosphorylation assays performed using commercial insulin receptor kinase fragments, IR-(966 -1271) (InsR active, Upstate Biotechnology, Lake Placid, NY) or IR-(947-1343) (␤IRK, BioMol, Plymouth Meeting, PA), except no preincubation with insulin was required to activate the kinase.
To investigate the inhibition of IGF-1R autophosphorylation, we incubated full-length purified IGF-1R with 100 nM IGF-1 for 60 min at room temperature to activate the receptor. Recombinant IRS-1-(925-1008) fragments were added prior to the addition of 50 M ATP, 10 mM MgCl 2 , and 8 mM MnCl 2 , and a time course of phosphorylation was performed.
Pull-down Experiments-Recombinant IRS-1-(925-1008) fragments were added to full-length insulin receptor activated with insulin, supplemented with 50 M ATP, 10 mM MgCl 2 , and 8 mM MnCl 2 (final concentrations), and were incubated for 30 min at room temperature. After incubation with Talon metal-affinity resin for 90 min, the affinity resin-protein complexes were washed three times with loading buffer (see above), solubilized in Laemmli's sample buffer, and analyzed by SDS-PAGE.
Western Blots-Proteins were transferred to polyvinylidene difluoride membranes and identified by immunoblotting with primary antibodies including anti-His (Qiagen), anti-phosphotyrosine (Upstate Biotechnology), and anti-insulin receptor ␤-subunit (Upstate Biotechnology). Blots were incubated with the appropriate peroxidase-conjugated secondary antibodies (Amersham Biosciences) and developed with enhanced chemiluminescence according to the manufacturer's instructions (Roche Applied Science).

In Vitro Phosphorylation of the IRS-1-(925-1008) Fragment
Is Reduced by the G972R Polymorphism-In vivo association of IRS-1 via the PTB domain ( Fig. 1) with Tyr 960 of the insulin receptor is normally required for optimal phosphorylation of IRS-1 by the insulin receptor (20,21). The IRS-1-(925-1008) wild type and G972R His-tagged fusion proteins, which contain both of the flanking tyrosine phosphorylation consensus sites (Fig. 2), were incubated with purified full-length human insulin receptor, and a time course of tyrosine phosphorylation was performed. Under the conditions used, the wild type (WT) recombinant fragment was able to undergo tyrosine phosphorylation in vitro (Fig. 3) despite the absence of the PTB domain in this fragment. Introduction of the G972R polymorphism into the IRS-1-(925-1008) fragment reduced the level of phosphorylation seen by Ͼ60% (Fig. 3). Mass spectrometry of the WT and G972R phosphopeptides confirmed that phosphorylation was significantly reduced at Tyr 941 in the EY 941 MLM phosphorylation consensus site upstream of the polymorphism compared with the WT (data not shown). The phosphopeptide containing the DY 989 MTM consensus site could not be detected under the conditions used for mass spectrometry; therefore, the extent of reduction at this site could not be determined.
The IRS-1-(925-1008) G972R Fragment Associates with the Insulin Receptor-Pull-down experiments were performed to investigate any association between the insulin receptor and IRS-1-(925-1008). The WT and G972R His-tagged fragments were incubated with insulin-activated insulin receptor followed by pull-down experiment with Talon metal-affinity resin specific for the His tag. Immunoblotting of the metalaffinity resin-protein complexes for the insulin receptor indicates that there was an association between the insulin receptor and the IRS-1-(925-1008) fragments (Fig. 4) and that this was markedly enhanced when the fragment contained the G972R polymorphism.
The IRS-1-(925-1008) G972R Fragment Inhibits Autophosphorylation of the Insulin Receptor and the IGF-1R-To deter- FIG. 1. Schematic diagram of human IRS-1. Shown underneath the molecule are all known polymorphisms discovered to date (38,39). Selected tyrosine phosphorylation sites are indicated, and the Src homology-2 domain-containing proteins that are known to interact with IRS-1 at these sites are shown (PI 3-K, GRB-2 and SHP2). The G972R polymorphism is found near the C terminus between two flanking tyrosine phosphorylation sites (EY 941 MLM (941) and DY 989 MTM (989)) that are known binding sites for the p85␣ regulatory subunit of PI 3-kinase. mine the mechanism by which the phosphorylation of the IRS-1-(925-1008) G972R fragment was reduced, we preincubated the insulin receptor with the IRS-1-(925-1008) fragments prior to the addition of ATP, and we performed a time course of insulin receptor autophosphorylation after the addition of ATP. Fig. 5A illustrates that, in the presence of the IRS-1-(925-1008), WT fragment autophosphorylation of the insulin receptor was robust and increased over time. In the presence of the IRS-1-(925-1008) G972R fragment, the autophosphorylation of the insulin receptor was markedly reduced (Fig. 5A), demonstrating that the IRS-1-(925-1008) G972R fragment acts as a competitive inhibitor of insulin receptor autophosphorylation.
This effect occurred in a dose-dependent manner with decreasing insulin receptor autophosphorylation as increasing molar concentrations of IRS-1-(925-1008) G972R were added (Fig. 6).  4. IRS-1-(925-1008) binds to the insulin receptor. The Histagged fusion proteins IRS-1-(925-1008) WT and G972R were incubated with full-length insulin receptor and 50 M ATP prior to pulldown experiment using Talon metal-affinity resin specific for the His tag. The control sample contained insulin receptor and Talon resin only. Samples were analyzed by 6% reducing SDS-PAGE and immunoblotted (IB) with antibody to the insulin receptor ␤-subunit. The IRS-1-(925-1008) G972R fragment was also able to significantly reduce autophosphorylation of the IGF-1R compared with the IRS-1-(925-1008) WT fragment (Fig. 5B), demonstrating that the IRS-1 G972R may also act as a competitive inhibitor of IGF-1R activity.

DISCUSSION
The aim of these experiments was to provide an understanding at a molecular level of how the most common human polymorphism in IRS-1, the G972R polymorphism, may be contributing to insulin resistance and diabetes. This study has demonstrated that the G972R variant alters the ability of the recombinant IRS-1-(925-1008) fragment to be phosphorylated at specific tyrosine residues (EY 941 MLM and DY 989 MTM) flanking the polymorphism, which normally form high affinity binding sites for the p85␣ regulatory subunit of PI 3-kinase. Reduced phosphorylation of IRS-1 at the flanking tyrosine phosphorylation consensus sites would lead to impaired association of the Src homology-2 domains (SH2) of p85␣ with IRS-1 at these sites and lowered IRS-1-associated PI 3-kinase activ- FIG. 7. IRS-1-(925-1008)  ity. This is consistent with the decrease in IRS-1-associated PI 3-kinase activity that Almind et al. (14) observed in 32D(IR) cells expressing the full-length IRS-1 G972R molecule. However, no significant decrease in insulin-stimulated tyrosine phosphorylation of the full-length IRS-1 G972R molecule was observed compared with the wild type molecule, suggesting that phosphorylation at other sites was likely intact and demonstrating the importance of studies like this one using small recombinant fragments from IRS-1 to determine the specific effect of the G972R polymorphism.
Our results are in agreement with several other studies in muscle and ␤-cells that have shown decreased binding of p85␣ to IRS-1 G972R, reduced IRS-1-associated PI 3-kinase activity, and decreased glycogen synthase kinase-3 phosphorylation in these tissues. Taken together with the present study, these studies suggest that the IRS-1 G972R polymorphism might be contributing to insulin resistance by impairing the ability of insulin to activate the IRS-1/PI 3-kinase/Akt/glycogen synthase kinase-3 signaling pathway in these two vitally important insulin-sensitive tissues. This would lead to multiple defects in processes including glucose transport, glucose transporter translocation and glycogen synthesis in muscle (15), and insulin secretion and cell survival in the ␤-cell (16 -18).
Type 1 diabetes is characterized by rapid decline and loss of ␤-cell function. Studies from the Sesti laboratory (16,17) showing that the G972R variant impairs human pancreatic function and survival in vitro may provide a possible link between Type 1 and Type 2 diabetes. Because Type 2 diabetes involves other tissues including liver, fat, and brain, it would also be interesting to investigate the effect of this polymorphism in these other tissues.
In addition to decreased phosphorylation of the fragment, we found that the IRS-1-(925-1008) G972R fragment associates with the insulin receptor directly and inhibits autophosphorylation of the receptor. We have shown that this association occurs between residues 966 and 1271 of the insulin receptor, the region of the receptor ␤-subunit that includes the tyrosine kinase autophosphorylation domain, and between residues 950 and 986 of IRS-1. Normally after ligand binding and autophosphorylation of the insulin receptor, IRS-1 associates with the receptor via the PH and PTB domains in the N terminus of IRS-1 (Fig. 1). The PH domain targets the molecule to the plasma membrane, and the PTB domain associates with phosphorylated Tyr 960 in the juxtamembrane region of the receptor (22). This interaction is transient and occurs with low affinity; therefore it is unexpected that the IRS-1-(925-1008) fragment containing the G972R polymorphism and lacking both the PH and PTB domains associates with the insulin receptor kinase domain, a region not normally involved in insulin receptor/ IRS-1 interactions, with such high affinity.
We have also demonstrated that the IRS-1-(925-1008) G972R fragment is able to inhibit autophosphorylation of the closely related IGF-1R, which also signals through IRS-1. It is reasonable to speculate that the mechanism of inhibition would be similar to that for the insulin receptor because the tyrosine kinase domains of the two receptors are highly homologous.
The question remains as to whether the reduced phosphorylation seen is associated with a direct effect of the G972R polymorphism because of conformational changes in IRS-1 or a change in charge state related to the substitution of glycine 972 for arginine or whether it is associated with an indirect effect of the polymorphism because of inhibition of insulin receptor tyrosine kinase autophosphorylation by IRS-1 G972R. The G972R polymorphism may interfere with the ATP-binding site at Lys 1018 in the insulin receptor tyrosine kinase domain. The substituted Arg 972 in IRS-1 has charge characteristics similar to those of Lys 1018 , and it has been shown previously that if ATP binding is abolished at this site, tyrosine kinase activity is lost (23)(24)(25). It seems likely that all of these mechanisms may contribute. Furthermore, this indicates how, even in a het- erozygous state, the G972R polymorphism may alter insulin sensitivity. Co-crystallization of the IRS-1-(925-1008) G972R fragment with the insulin receptor tyrosine kinase domain would give detailed information about the altered interaction between the two proteins and would provide insight into how the G972R polymorphism enhances the association. Because of the ability to bind with high affinity to the insulin receptor and to inhibit insulin receptor and IGF-1R autophosphorylation, the IRS-1-(925-1008) G972R peptide could also be a potentially useful tool to inhibit receptor function in vivo.
Type 2 diabetes is a complex polygenic disorder, and the true extent of the genes involved is unclear (26). Polymorphisms that cause only a modest change in gene or protein function may become clinically significant when expressed along with other genetic or acquired defects. Genetic alterations have been reported in numerous insulin-signaling proteins in addition to IRS-1 including the insulin receptor (27), IRS-2 (28 -30), IRS-4 (31), and p85␣ (32,33), as well as other modulators of insulin signaling including peroxisome proliferator-activated receptor-␥ (34,35) and peroxisome proliferator-activated receptor coactivator-1 (PGC-1) (36,37). A combination of changes in multiple genes/proteins involved in insulin signaling may lead to the end-disease phenotype. Therefore the IRS-1 G972R polymorphism may be just one contributing factor of many that predispose an individual to Type 2 diabetes.
Taken together, these data indicate that the primary defect of the G972R polymorphism is the increased association of IRS-1 G972R with the insulin receptor and the inhibition of insulin receptor autophosphorylation leading to reduced downstream signaling along the PI 3-kinase pathway. These effects in combination with changes in other genes encoding insulinsignaling proteins may contribute to global insulin resistance.