Insulin-like Growth Factor-I-stimulated Insulin Receptor Substrate-1 Negatively Regulates Src Homology 2 Domain-containing Protein-tyrosine Phosphatase Substrate-1 Function in Vascular Smooth Muscle Cells*

Vascular smooth muscle cells maintained in normal (5.6 mm) glucose respond to insulin-like growth factor-I (IGF-I) with increased protein synthesis but do not proliferate. In contrast, hyperglycemia alters responsiveness to IGF-I, resulting in increased SHPS-1 phosphorylation and assembly of a signaling complex that enhances MAPK and phosphatidylinositol 3-kinase pathways. Hyperglycemia also reduces the basal IRS-1 concentration and IGF-I-stimulated IRS-1-linked signaling. To determine if failure to down-regulate IRS-1 alters vascular smooth muscle cell (VSMC) responses to IGF-I, we overexpressed IRS-1 in VSMCs maintained in high glucose. These cultures showed reduced SHPS-1 phosphorylation, transfer of SHP-2 to SHPS-1, and impaired Shc and MAPK phosphorylation and cell proliferation in response to IGF-I. In vitro studies demonstrated that SHPS-1 was a substrate for type I IGF receptor (IGF-IR) and that IRS-1 competitively inhibited SHPS-1 phosphorylation. Exposure of VSMC cultures to a peptide that inhibited IRS-1/IGF-IR interaction showed that IRS-1 binding to IGF-IR impairs SHPS-1 phosphorylation in vivo. IRS-1 also sequestered SHP-2. Expression of an IRS-1 mutant (Y1179F/Y1229F) reduced IRS-1/SHP-2 association, and exposure of cells expressing the mutant to the inhibitory peptide enhanced SHPS-1 phosphorylation and SHP-2 transfer. This result was confirmed by expressing an IRS-1 mutant that had both impaired binding to IGF-IR and to SHP-2 IGF-I increased SHPS-1 phosphorylation, SHP-2 association with SHPS-1, Shc MAPK phosphorylation, and proliferation in cells expressing the mutant. We conclude that IRS-1 is an important factor for maintaining VSMCs in the non-proliferative state and that its down-regulation is a component of the VSMC response to hyperglycemic stress that results in an enhanced response to IGF-I.

Insulin-like growth factor-I (IGF-I) 2 binds selectively to the type I IGF receptor, which results in stimulation of several bio-logical functions. IGF-I is a known stimulant of vascular smooth muscle cell (VSMC) proliferation and migration, and these effects are mediated through the phosphatidylinositol 3-kinase and MAPK pathways (1)(2)(3). Upon ligand occupancy, the receptor undergoes a conformational change that activates its tyrosine kinase, which phosphorylates downstream signaling molecules. IRS-1 (insulin receptor substrate-1) is one of the important substrates for both the insulin and IGF-I receptor tyrosine kinases (4,5). IRS-1 is a docking protein that contains multiple domains that mediate protein-protein interactions, including a pleckstrin homology domain, a phosphotyrosinebinding (PTB) domain, and the carboxyl terminus (5,6). The PTB domain binds directly to the phosphorylated NPXY motif located within the activated insulin, IGF-I, or interleukin-4 receptors (7)(8)(9)(10). In addition, the pleckstrin homology domain also couples IRS-1 to activated receptors (11). The carboxyl terminus of IRS-1 contains several phosphotyrosines that interact with Src homology 2 domains of signaling molecules. Following their phosphorylation, several signaling intermediates, such as Grb-2, p85, and SHP-2 (Src homology 2 domain-containing protein-tyrosine phosphatase 2), bind to IRS-1, leading to activation of downstream signaling events (12,13). IRS-1 has been linked to many of the insulinor IGF-I-mediated biological responses, including changes in cell size, mitogenesis, and cell migration (14 -16). IGF-Istimulated IGF-IR has been shown to bind to IRS-1 and phosphorylate it in several cell types, including vascular smooth muscle cells (17,18).
Previous studies have shown that the trans-membrane scaffolding protein SHPS-1 (SHP substrate-1) is a substrate for activated tyrosine kinases (19). Our studies have shown that SHPS-1 phosphorylation and the subsequent assembly of a signaling complex that includes SHP-2, Src, Shc (Src homology 2 domain-containing protein), and Grb2 (growth factor receptor-bound protein 2) is essential for these cells to respond to hyperglycemic stress and that it enhances the ability of IGF-I to activate both MAPK and phosphatidylinositol 3-kinase pathways, leading to increased proliferation and migration (2,3,20). Following their tyrosine phosphorylation, IRS-1 and Shc bind independently to Grb2, which increases activation of MAPK/ extracellular signal-regulated kinase (ERK) (2,21,22). However, the relative contribution of Shc and IRS-1 in regulating this cascade has not been clearly delineated for many different cell types (12,23). Additionally, IRS-1 concentrations are regulated by several factors, including glucose, insulin, IGF-I, and cell/tissue type (24 -30). Several investigators have demonstrated that IRS-1 concentrations are down-regulated in response to hyperglycemia in insulin-sensitive tissues as skeletal muscle and fat (13,31,32). Previously, we reported that VSMCs maintained in high glucose concentrations (Ͼ12 mM) had a significant decrease in IRS-1. We also showed that the IRS-1 tyrosine phosphorylation response and subsequent Grb2 binding are significantly reduced in response to IGF-I (29). These studies were undertaken to determine the significance of the reduction in IRS-1 and whether failure to reduce the IRS-1 concentration alters assembly of the SHPS-1 complex, thus leading to altered cell proliferation and migration in response to IGF-I.

MATERIALS AND METHODS
Human IGF-I was a gift from Genentech (South San Francisco, CA). Immobilon-P membranes were purchased from Millipore Corp. (Bedford, MA). DMEM containing 4500 mg/liter glucose (DMEM-HG), DMEM containing 900 mg/liter glucose (DMEM-NG), and streptomycin and penicillin were purchased from Invitrogen. Antibodies against phospho-MAPK or total MAPK and ␤-actin were purchased from Cell Signaling (Beverly, MA). Polyclonal antibodies for IRS-1, SHPS-1, and Shc were obtained from Upstate Biotechnology (Lake Placid, NY). Anti-phosphotyrosine (Tyr(P) 99 ), the hemagglutinin epitope (HA), anti-IRS-1, anti-IGF-IR, anti-SHP-2, and anti-Grb2 antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). SHPS-1 and IGF-IR polyclonal antiserum were prepared in our laboratory. IGF-IR tyrosine kinase inhibitor, PQ401, was purchased from Tocris Bioscience, (Ellisville, MO). Recombinant rat IRS-1 expressed in Sf9 cells with His tag was purchased from Millipore Corp. Synthetic peptides containing the cell permeability sequence of protein transduction domain 4 and the sequence from IGF-IR YARA-AARQARA 947 NPEYFSAADV 956 (underlined and hereafter referred to as peptide 299), the tyrosine-phosphorylated form of p299 peptide (corresponding to Tyr 950 of IGF-IR), and the sequence from the IRS-1 PTB domain YARAAA-RQARA 206 LQLMNIRRCGHS 217 (underlined and hereafter referred to as peptide 301, where C is italicized to denote reduction) were synthesized by the Protein Chemistry Core Facility at the University of North Carolina at Chapel Hill. Purity and sequence confirmation were determined by mass spectrometry. We used a heterologous system in this study. It is important to note that the sequences in porcine IGF-IR that bind to IRS-1 are conserved in human IGF-IR. Likewise, the IGF-I receptor tyrosine kinase sequence is completely conserved when human and porcine IGF-IR are compared. Similarly, the PTB domain of porcine IRS-1 has 100% homology to human IRS-1. Finally, porcine IGF-I and human IGF-I have 100% sequence identity. Therefore, we believe that this degree of sequence conservation enables us to study these interactions interchangeably.
Construction, Expression, and Purification of the Cytoplasmic Domain of SHPS-1 (SHPS-1/CD) and IGF-IR-The SHPS-1 cytoplasmic domain containing amino acids 394 -503 was PCR-amplified with the SP6 promotor sequence, following which reverse transcription was performed to obtain mRNA (33). Using this template, SHPS-1/CD was generated, employing in vitro translation using wheat germ extract as per the manufacturer's recommendation (Promega).
Cell Culture-Porcine SMCs were isolated from the porcine aortic explants according to the protocol described by Ross (74) and were maintained as described previously (2,29). Cells were maintained in DMEM-high glucose growth medium (HG) or DMEM-normal glucose growth medium (NG) with 10% fetal bovine serum (Hyclone, Logan, UT), streptomycin (100 ng/ml), and penicillin (100 units/ml). All of the mutant cells were maintained in HG. The SMCs that were used in these experiments were used between passages 4 and 12. Vectors pLenti-HA-IGF-IR  Wild Type, pLenti-HA-IRS-1 wild type (IRS-1-WT), pLenti-HA-IRS-1-FF mutant (IRS-FF), and pLenti-HA-IRS-1-R3Q-FF  Double Mutant (IRS-R3Q-FF)-Wild type IGF-IR in pBluescript II KS was a kind gift from Dr. Derek LeRoith. Using this as a template, PCR amplification was performed with forward primer 5Ј-caccatgaagtctggctccggaggagggt-3Ј, including a Kozak sequence (in boldface type) placed 5Ј to the ATG (underlined) start site. The reverse primer 5Ј-ttaagcgtaatctggaacatcgtatgggtagcaggtcgaagactggggcagcgg-3Ј contained the sequence encoding HA epitope (underlined), followed by the stop codon (boldface type). The amplified product was cloned into the pENTR/D-TOPO Gateway entry vector according to the manufacturer's instructions (Invitrogen). The resultant product was confirmed by sequencing (University of North Carolina Genome Analysis Facility, Chapel Hill, NC) and transferred from the entry vector into the pLentiCMV DEST vector using the LR clonase reaction following the manufacturer's instructions (Invitrogen). Full-length human IRS-1 cDNA was purchased from Open Biosystems in pCMV-sport6 vector and after PCR amplification was cloned into the pENTR/D-TOPO Gateway entry vector according to the manufacturer's instructions (Invitrogen). The forward and reverse primers used to generate the PCR product were as follows: forward primer, 5Ј-caccatggcgagccctccggagagcgat-3Ј; reverse primer, 5Ј-ttaagcgtaatctggaacatcgtatgggtactgacggtcctctggctgcttctgg-3Ј. The forward primer includes a Kozak sequence (boldface type) placed 5Ј to the ATG (underlined) start site. The reverse primer contained the sequence encoding HA epitope (underlined) followed by the stop codon (boldface type). The resultant product was sequenced (UNC Genome Analysis Facility, Chapel Hill, NC) to confirm the correct full-length human IRS-1 using M13 forward and reverse primers and five other primers: IRS-1Seq1A, 5Ј-gggtcagacaaagaacctgatt-3Ј; IRS-1Seq2A, 5Ј-cgcccagcctcggtgga-3Ј; IRS-1Seq3A, 5Ј-ctaccattacccaccagaag-3Ј; IRS-1Seq4A, 5Ј-tgactacatgaacatgtcacca-3Ј; IRS-1Seq5C, 5Ј-tgaccatgcagatgagttgtcc-3Ј.

Generation of pLenti Expression
Using the pENTR IRS-1-WT as template, an IRS mutant was generated where the tyrosines at positions 1179 and 1229 were mutated to phenylalanine (IRS-FF). PCR amplifications were carried out using complementary primer 5Ј-ggagaatggtcttaact-Tcatagacctggatttgg-3Ј and reverse primer 5Ј-ccaaatccaggtctat-gAagttaagaccattctcc-3Ј to generate Y1179F and another complementary primer 5Ј-gaggatttaagcgcctTtgccagcatcag-3Ј and reverse primer 5Ј-ctgatgctggcaAaggcgcttaaatcctc-3Ј to generate Y1229F (where capitalized bases indicate the substitutions). After selection of correct clone based on DNA sequencing, using pENTR IRS-FF as template, a double mutant was constructed where three arginines at positions 212, 213, and 227 were mutated to glutamine (IRS-R3Q-FF) using doublestranded mutagenesis. Initially, PCR amplification was carried out using forward primer 5Ј-gctgatgaacatcCAgcAAtgtggccactcgg-3Ј and reverse primer 5Ј-ccgagtggccacaTTgcTGgatgttcatcagc-3Ј (where capitalized bases indicate the substitutions), by which two of the residues, Arg 212 and Arg 213 , were mutated to glutamine. The resultant DNA was used as a template, and the arginine at position 227 was mutated to glutamine using complementary primers 5Ј-catcgaggtgggccAAtctgccgtgac-gggg-3Ј and reverse primer 5Ј-ccccgtcacggcagaTTggcccacctcgatg-3Ј (where capitalized bases indicate the substitution). The products were sequenced (University of North Carolina Genome Analysis Facility) to confirm the correct incorporation of the changes, following which the cDNAs encoding the wild type, single mutant, and double mutant forms of the protein were transferred from the entry vector into pLentiCMV DEST vector using the LR clonase reaction following the manufacturer's instructions (Invitrogen).
Construction of a Plasmid Containing Short Hairpin RNA Template for IRS-1 Silencing-Based on Invitrogen web site design tools, a sequence containing 21 oligonucleotides (GCGAGGATTTAAGCGCCTATG) located within the porcine IRS-1 (accession number EU681268.1; nucleotides 3515-3536) was used to construct the short hairpin RNA template plasmid. The selected sequences were not found in any other mammalian protein searched using the NCBI Blast search program. The oligonucleotides were synthesized by the Nucleic Acids Core Facility at the University of North Carolina, annealed, and ligated into BLOCK-iT U6 RNAi entry vector (catalogue no. K4945-00, Invitrogen), following the manufacturer's instructions. The complete sequence was verified by DNA sequencing. The expression vector was generated using the Gateway LR recombination reaction between the entry vector and BLOCK-iT Lentiviral RNAi Gateway Vector (catalogue no. K4943-00, Invitrogen). The empty vector was used as a control (EVC). After confirmation of the sequence, plasmid DNA was prepared by a Plasmid minikit (Promega, Madison, WI).
Generation of Virus Stocks-293FT cells (Invitrogen) were prepared for generation of virus stocks of each individual pLenti construct. All of the transfections and viral stock generations were carried out as described previously (2,3).
Establishment of Porcine VSMCs Expressing pLenti Constructs-VSMCs (passages 4 and 5) were seeded at 3 ϫ 10 5 / well in each of two wells of a 6-well plate (catalog no. 353046, Falcon) the day before transduction. The viral stocks were thawed, and the viral complexes were precipitated as described previously (2,3). The expression of the HA-tagged IRS-1-WT and IRS-1 mutants were detected by immunoblotting with a 1:1000 dilution of anti-HA antibody using 20 g of protein.
Immunoprecipitation and Immunoblotting-Cells were seeded at 5 ϫ 10 5 cells/10-cm plate (BD Biosciences) and grown for 5-7 days to reach confluence. The cultures were incubated in serum-free DMEM-HG for 14 -18 h before the addition of IGF-I (as indicated). In some experiments, non-transduced SMCs were preincubated with or without the synthetic peptides for 1-2 h before IGF-I was added. Following cell lysis, all immunoprecipitations and immunoblotting were carried out as described previously (2,3). The proteins were detected using enhanced chemiluminescence (Pierce). The images obtained were also scanned using DuoScan scanning densitometer (Agfa, Morstel, Belgium). Densitometric analyses of the images were determined using National Institutes of Health ImageJ, version 1.37v. To determine differences in MAPK phosphorylation, arbitrary scanning units obtained for phospho-MAPK band intensities were divided by arbitrary scanning units obtained for total ␤-actin protein band intensities for each time point analyzed. The -fold values or average ratio values obtained from at least three independent experiments for each lane were pooled together.
Cell Proliferation Assay and Wounding and Migration Assay-Assessment of SMC proliferation was performed as described previously (35). IGF-I is known to stimulate smooth muscle cell chemokinesis, and cellular migration was analyzed as previously described (36). Briefly, VSMCs were plated in 6-well dishes and grown to confluence. The monolayer of cells was wounded with a razor blade. After wounding, cells were incubated with SFM plus 0.2% FBS in the presence or absence of IGF-I (50 ng/ml) at 37°C for 48 h. The wounded monolayers were then fixed and stained (Diff Quick; Dade Behring, Inc., Newark, DE), and the number of cells migrating into the wound area was counted. At least five of the previously selected 1-mm areas at the edge of the wound were counted for each data point.
Statistical Analysis-Student's t test was used to compare the differences between the treatments. The results that are shown in all experiments are representative of at least three separate experiments, and p Յ 0.05 was considered statistically significant.
In order to determine the effect of maintaining a higher level of IRS-1 than would normally be present in hyperglycemic con-FIGURE 1. Hyperglycemia decreases total IRS-1 protein levels and IRS-1-mediated signaling. A, confluent VSMC cultures were maintained in medium containing 5 mM (NG) or 25 mM (HG) glucose prior to these analyses. They were changed to serum-free medium containing the same glucose concentration for 12, 24, or 48 h without (top) or with (second panel) IGF-I (50 ng/ml). The anti-␤-actin antibody (loading control) is shown in the lower panel. After cell lysis, aliquots containing equal amounts of protein were separated directly by SDS-PAGE and immunoblotted (IB) with anti-IRS-1 antibody. The bar graph is representative of three independent experiments without IGF-I. Error bars, mean Ϯ S.E. **, p Ͻ 0.01. B, the extent of IRS-1 tyrosine phosphorylation was determined by immunoprecipitating (IP) cell lysates with an anti-IRS-1 antibody and then immunoblotting with an anti-Tyr(P) 99 (PY99) antibody (first panel), and Grb-2 association with IRS-1 was determined by immunoprecipitating with an anti-Grb2 antibody and then immunoblotting with anti-IRS-1 antibody (third panel). The membranes were stripped and reprobed with either anti-IRS-1 or anti-Grb2 antibody to detect the total levels of IRS-1 (second panel) or Grb2 (fourth panel), respectively. The amount of immunoprecipitate analyzed was adjusted to correct for differences in total IRS-1 between the NG and HG cultures. The bar graph represents three independent experiments. Error bars, mean Ϯ S.E. ***, p Ͻ 0.001 when the amount of tyrosine phosphorylation of IRS-1 or the amount of Grb2 associated with IRS-1 at 5 min in response to IGF-I is compared between NG and HG cultures. C, VSMCs expressing either LacZ or IRS-WT were serum-starved in DMEM-HG and analyzed for recombinant protein expression. Cell lysates were immunoblotted using an anti-HA antibody (top), anti-IRS-1 antibody (middle), and an anti-␤-actin antibody loading control (bottom). D, confluent VSMCs expressing LacZ or IRS-WT cultured in HG were serum-deprived and stimulated without or with IGF-I (50 ng/ml) for 12 or 24 h. Equal amounts of protein were separated by SDS-PAGE and immunoblotted with anti-IRS-1 antibody (without IGF-I (top) or with IGF-I (middle)) or with anti-␤-actin antibody (bottom). The bar graph is representative of three independent experiments. Error bars, mean Ϯ S.E. **, p Ͻ 0.01. ditions on IGF-I actions, VSMCs overexpressing HA-tagged IRS-1 (IRS-WT) were constructed. The basal total IRS-1 protein level was higher (5.2 Ϯ 0.1-fold, n ϭ 5, p Ͻ 0.001) in IRS-WT cells maintained in DMEM-HG compared with LacZ control cells (Fig. 1C). Following exposure to IGF-I, IRS-1 was reduced to a significantly greater extent in the LacZ cells (2.6 Ϯ 0.2-fold decrease) compared with a 1.9 Ϯ 0.1-fold decrease in IRS-1 WT cells (n ϭ 3, p Ͻ 0.01; Fig. 1D).
Effect of IRS-1 Overexpression on IGF-I-mediated SHPS-1 Phosphorylation, SHPS-1 Complex Assembly, Downstream Signaling, and Cellular Proliferation-Cellular proliferation in IRS-WT-overexpressing cells maintained in DMEM-HG was significantly impaired (1.1 Ϯ 0.3-fold increase) compared with LacZ control cells (1.98 Ϯ 0.1-fold increase) in response to IGF-I (n ϭ 5, p Ͻ 0.05; Fig. 2A). Similarly, VSMC migration in response to IGF-I was significantly reduced (1.97 Ϯ 0.1-fold reduction, n ϭ 4, p Ͻ 0.05) in IRS-WT cells maintained in DMEM-HG compared with LacZ control cells (Fig. 2B). Because proliferation of VSMCs in high glucose has been shown to be dependent on Shc phosphorylation and subsequent MAPK activation, we determined the Shc and MAPK phosphorylation responses in SMCs overexpressing IRS-1 that were maintained in DMEM-HG. The Shc phosphorylation response to IGF-I was reduced 2.1 Ϯ 0.2-fold and 2.6 Ϯ 0.1-fold at 5 and 10 min, respectively (n ϭ 4, p Ͻ 0.01) in IRS-WT cells compared with LacZ cells, and Shc-Grb2 binding was also reduced 2.7 Ϯ 0.1-fold both at 5 and 10 min after IGF-I stimulation (n ϭ 4, p Ͻ 0.001) (Fig. 2C, top panels). Additionally, phosphorylation of MAPK was found to be significantly attenuated (4.1 Ϯ 0.2-fold, n ϭ 5, p Ͻ 0.01) in cells overexpressing IRS-WT compared with LacZ cells in response to IGF-I after 10 min (Fig. 2C, bottom panels).
Phosphorylation of SHPS-1 and subsequent assembly of a signaling complex is required for IGF-I stimulation of p52 shc phosphorylation and MAPK activation (2). SHP-2 is a key component of that complex that binds directly to phosphorylated SHPS-1 (33); hence, SHPS-1 phosphorylation and SHP-2 recruitment were evaluated. Phosphorylation of SHPS-1 was found to be significantly impaired (5.1 Ϯ 0.2-fold reduction, n ϭ 3, p Ͻ 0.001) in IRS-WT cells compared with LacZ cells in response to IGF-I (Fig. 2D). Similarly, the association of SHP-2 with SHPS-1 was found to be reduced (4.2 Ϯ 0.1-fold, n ϭ 3, p Ͻ 0.001) in IRS-WT cells compared with LacZ cells in response to IGF-I (Fig. 2D). In contrast, the IGF-IR phosphorylation levels remained the same in both IRS-WT cells and LacZ cells in response to IGF-I (Fig. 2D). Taken together, the above data demonstrate that maintenance of an IRS-1 level in SMCs maintained in high glucose that was similar to the IRS-1 level in VSMCs maintained in normal glucose was associated with a reduction in several of the cellular responses to IGF-I, including SHPS-1 phosphorylation, SHP-2 recruitment to SHPS-1, p52 shc phosphorylation, MAPK activation, and cell proliferation.
Because IRS-1 is a known substrate for both insulin receptor and IGF-IR tyrosine kinases, we wished to determine whether IRS-1 could compete with SHPS-1 as an IGF-IR substrate. Recombinant IRS-1 was added to the in vitro phosphorylation assays. Increasing concentrations of IRS-1 decreased the phosphorylation of SHPS-1/CD (3.9 Ϯ 0.3-fold decrease, n ϭ 3, p Ͻ 0.01) (Fig. 3F), but there was no effect on IGF-IR autophosphorylation. These observations, taken together with our results showing that IRS-1-overexpressing cells exhibit impaired SHPS-1 phosphorylation (Fig. 2D), suggest that IRS-1 is competing with SHPS-1 for binding to IGF-IR, thereby altering SHPS-1 phosphorylation.
To control for potential differences in tyrosine phosphorylation of peptide 299, the experiment was repeated using a tyrosinephosphorylated form. Following exposure to this peptide, the amount of IRS-1 associated with IGF-IR and the tyrosine phosphorylation of IRS-1 were significantly impaired (2.3 Ϯ 0.2-and 2.1 Ϯ 0.1-fold, respectively) in response to IGF-I after 5 min in cells exposed to peptide compared with non-exposed cells (n ϭ 3, p Ͻ 0.01). Further, the level of association of SHPS-1 with IGF-IR and SHPS-1 phosphorylation was significantly increased in the cells exposed to the peptide in response to IGF-I at 5 min compared with non-exposed cells (3.1 Ϯ 0.1-and 3.3 Ϯ 0.1-fold, respectively) (n ϭ 3, p Ͻ 0.001; supplemental Fig. 1A).
Formation of the SHPS-1 signaling complex that leads to MAPK pathway activation is dependent upon SHP-2 recruitment to phosphorylated SHPS-1 (2). When we determined the effect of peptide 301 on SHP-2 transfer, it failed to restore the SHP-2 transfer to SHPS-1 despite its ability to increase SHPS-1 phosphorylation. Furthermore, the amount of IRS-1 associated with SHP-2 in the presence of peptide 301 in response to IGF-I was not reduced significantly (1.0 Ϯ 0.1-fold, n ϭ 3, p ϭ not significant; Fig. 4B).
In order to confirm the findings obtained in cells overexpressing IRS-1, we used non-transfected VSMCs cultured in DMEM-NG (Fig. 1). When those cultures were exposed to peptide 301, they showed a significant reduction in IGF-IR/IRS-1 association compared with cells not exposed to the peptide in response to IGF-I (3.6 Ϯ 0.1-fold, n ϭ 3, p Ͻ 0.001; Fig. 4C). This reduction in IGF-IR/IRS-1 association led to impairment of tyrosine phosphorylation of IRS-1 in response to IGF-I (2.0 Ϯ 0.2-fold) compared with control cells. Peptide exposure increased the SHPS-1 phosphorylation (2.2 Ϯ 0.3-fold) and SHP-2 transfer to SHPS-1 (1.9 Ϯ 0.1-fold) in response to IGF-I (n ϭ 3, p Ͻ 0.05; Fig. 4C). Further-more, a significant increase was observed in IGF-IR/SHPS-1 association (2.2 Ϯ 0.2-fold, n ϭ 3, p Ͻ 0.01), Shc tyrosine phosphorylation (1.9 Ϯ 0.1-fold, n ϭ 3, p Ͻ 0.05), and phosphorylation of MAPK (1.9 Ϯ 0.1-fold, n ϭ 3, p Ͻ 0.05; supplemental Fig. 1C) in cells exposed to peptide 301 in response to IGF-I compared with non-exposed cells. Concomitant incubation of VSMCs maintained in DMEM-NG with peptide 301 increased cellular proliferation in response to IGF-I (1.8 Ϯ 0.1-fold, n ϭ 3, p Ͻ 0.01; supplemental Fig. 1D). Taken together, these results suggest that IRS-1 and SHPS-1 compete for binding to IGF-IR kinase and that the level of IRS-1 that is present in VSMCs maintained in normal glucose is sufficient to suppress SHPS-1 phosphorylation and the subsequent assembly of the SHPS-1-associated signaling complex.

Effect of SHP-2 Sequestration by IRS-1 on SHPS-1 Complex
Assembly-IRS-1 contains several YXX(L/I/V) motifs that are known to bind to SHP-2 following stimulation of IRS-1 tyrosine phosphorylation. Therefore, we proposed that IRS-1 overexpression resulted in sequestration of SHP-2 and prevented its recruitment to phosphorylated SHPS-1 (Fig. 4B). To test this hypothesis, we prepared a HA-tagged IRS-1 mutant (IRS-FF), wherein tyrosine residues 1179 and 1229 were mutated to phenylalanine (Fig. 5A). The association of IRS-1 and SHP-2 was reduced 2.7 Ϯ 0.1-fold in the IRS-FF mutant (n ϭ 5, p Ͻ 0.01) compared with IRS-WT cells in response to IGF-I after 5 min (Fig. 5B). Recent studies have shown that IRS-1 binding to SHP-2 could potentially be mediated by five other sites, Tyr 759 , Tyr 760 , Tyr 891 , Tyr 935 , and Tyr 1006 (38). This could explain the residual IRS-1 binding to SHP-2 that is detected using the IRS-FF mutant. Despite the reduction in SHP-2 binding to IRS-1, the amount of SHP-2 transferred to SHPS-1 was still impaired in IRS-FF mutant cells in response to IGF-I compared with LacZ cells. This is probably due to the fact that SHPS-1 phosphorylation was still significantly impaired in the IRS-FF cells compared with LacZ cells (Fig. 5C). Furthermore, the levels of SHPS-1 phosphorylation and SHP-2 transfer failed to increase significantly in the IRS-FF cells compared with IRS-WT cells in response to IGF-I.
To determine why inhibiting SHP-2 association with IRS-1 did not alter SHP-2 binding to SHPS-1, we assessed IRS-1/ IGF-IR association using the cells that were overexpressing the IRS-FF mutant. When those cells were compared with the IRS-WT cells, there was no difference in IRS-1/IGF-IR association in response to IGF-I after 5 min (1.1 Ϯ 0.1-fold, n ϭ 3, p ϭ not significant; Fig. 5C). The difference in IGF-IR/IRS-1 association after 10 min of IGF-I stimulation could be explained by the ability of SHP-2 to dephosphorylate IRS-1 in WT cells and not in FF mutant cells (data not shown). To determine if inhib-iting IGF-IR/IRS-1 interaction by incubating the blocking peptide with the IRS-FF mutant cells could rescue the SHPS-1 phosphorylation and SHP-2 recruitment to SHPS-1, cells expressing the IRS-1-FF mutant were exposed to peptide 301. The addition of peptide 301 resulted in increased IGF-IR association with SHPS-1 in response to IGF-I compared with cells not exposed to the peptide (2.2 Ϯ 0.1-fold, n ϭ 5, p Ͻ 0.05; Fig.  5D). Similarly, there was an increase in SHPS-1 phosphorylation (2.6 Ϯ 0.1-fold, n ϭ 3, p Ͻ 0.001) and the subsequent SHPS-1/SHP-2 association (2.3 Ϯ 0.3-fold, n ϭ 3, p Ͻ 0.05; Fig.  5D). More importantly, tyrosine phosphorylation of p52 shc was significantly increased in the IRS-FF cells exposed to peptide 301 (2.9 Ϯ 0.2-fold, n ϭ 3, p Ͻ 0.001) compared with cells not exposed to the peptide in response to IGF-I (Fig. 5D). Taken together, these results suggest that disruption of association IGF-IR/IRS-1 association and inhibition of SHP-2 binding to IRS-1 appears to be required for optimal SHP-2 transfer to SHPS-1 and Shc activation.

Expression of an IRS-1 Double Mutant That Impairs IGF-IR/ IRS-1 Association and IRS-1/SHP-2 Association Restores SHPS-1 Phosphorylation and SHP-2 Association with SHPS-1-To defin-
itively prove that IRS-1 binding to both IGF-IR and SHP-2 accounted for inhibition of SHPS-1 phosphorylation and downstream signaling, we used the IRS-FF mutant as a template and mutated the arginine residues 212, 213, and 217 in the IRS-1 PTB domain to glutamine because these residues have been shown to mediate IRS-1 association with IR/IGF-IR (7,10). The combined mutant is termed IRS-R3Q-FF. The expressions of HA-tagged IRS-WT and IRS-R3Q-FF were found to be similar when the cell lysates were compared (Fig. 6A). As shown in Fig. 6B, this IRS-1 mutant had significantly impaired binding to IGF-IR in response to IGF-I compared with cells expressing IRS-WT. Further, the tyrosine phosphorylation of IRS-1 and the subsequent IRS-1 binding to Grb2 was found to be significantly reduced in IRS-R3Q-FF compared with IRS-WT cells in response to IGF-I (3.3 Ϯ 0.1-fold, n ϭ 3, p Ͻ 0.001). The amount of SHP-2 bound to IRS-1 was found to be reduced (1.9 Ϯ 0.2-fold, n ϭ 3, p Ͻ 0.05) but not completely abolished in  (fourth panel). E, confluent IGF-IR-overexpressing cells were serum-starved for 16 h in DMEM-HG and then exposed to IGF-I for the times indicated. The cell lysates were immunoprecipitated with anti-IGF-IR antiserum. Some aliquots of reconstituted immunoprecipitate were incubated with SHPS-1/CD in in vitro phosphorylation assays (lanes 1-3) or with buffer alone (Con). After centrifugation, the resultant supernatants were immunoblotted using a phosphotyrosine antibody IGF-IR (top) or phospho-SHPS-1/CD (middle). Similar lysates were immunoprecipitated and immunoblotted with anti-IGF-IR antibody as a loading control (bottom). To ascertain the specificity of the bands, an antibody only control was also used. The bar graph represents the relative increase in SHPS-1/CD phosphorylation in response to IGF-I at 10 min compared with non-stimulated immunoprecipitates; ***, p Ͻ 0.001 based on at least three independent experiments. Error bars, mean Ϯ S.E.

. Inhibition of IRS-1 binding to IGF-IR increases association of SHPS-1 with IGF-IR and increases SHPS-1 phosphorylation.
A and B, confluent cultures expressing WT-IRS-1 were serum-starved overnight and then incubated with or without cell-permeable peptides, IGF-IR (peptide 299) (A) or IRS-1 (peptide 301) (B) at 20 g/ml for 2 h. IGF-I was added for either 5 or 10 min. Cell lysates were immunoprecipitated (IP) with anti-IGF-IR antibody and then immunoblotted for IRS-1 (top panel) or for SHPS-1 (second panel). The blots were stripped and reprobed with anti-IGF-IR antibody (third panel). Cell lysates from the same experiments were immunoprecipitated with anti-SHPS-1 antibody and immunoblotted with Tyr(P) 99 antibody (fourth panel). The blots were stripped and reprobed with anti-SHPS-1 antibody (fifth panel). B, in addition, cell lysates were immunoprecipitated with anti-SHPS-1 antibody and immunoblotted for SHP-2 (sixth panel). Similar lysates were immunoprecipitated with anti-SHP-2 antibody and immunoblotted for IRS-1 (seventh panel) or SHP-2 (eighth panel). The top bar graph shows the relative increase in IGF-IR/SHPS-1 association for at least three independent experiments. Error bars, mean Ϯ S.E. *, p Ͻ 0.05 when increase in IGF-IR/SHPS-1 association in response to IGF-I at 10 min is compared between IRS-WT cells with and without peptide 299. The middle bar graph shows the relative increase in SHPS-1 phosphorylation for at least three independent experiments. Error bars, mean Ϯ S.E. ***, p Ͻ 0.001 when increase in SHPS-1 phosphorylation in response to IGF-I after 10 min is compared in IRS-WT-expressing cells with or without peptide 301. C, VSMCs cultured and maintained in DMEM-NG were serum-starved and then incubated with or without cell-permeable IGF-IR/IRS-1 peptide 301 at a concentration of 5 g/ml for 2 h. IGF-I was added for 5 min with and without the peptide. Cell lysates from the above experiments were immunoprecipitated with anti-IRS-1 antibody and immunoblotted for either IGF-IR (upper panel) or Tyr(P) 99 (second panel). The blots were stripped and reprobed with anti-IRS-1 antibody (third panel). Cell lysates from the same experiments were immunoprecipitated with anti-SHPS-1 antibody and immunoblotted with Tyr(P) 99 antibody (fourth panel) or with anti-SHP-2 antibody (fifth panel). The blots were stripped and reprobed with anti-SHPS-1 antibody (sixth panel). The bar graph shows the relative increase in SHPS-1 phosphorylation for at least three independent experiments. Error bars, mean Ϯ S.E. **, p Ͻ 0.01 when the increase in SHPS-1 phosphorylation in response to IGF-I is compared with or without peptide 301.
the double mutant compared with IRS-WT cells (Fig. 6B). Furthermore, following IGF-I stimulation, there was a significant increase in SHPS-1 binding to IGF-IR (2.9 Ϯ 0.4-fold, n ϭ 3, p Ͻ 0.01) and an increase in SHPS-1 phosphorylation (3.3 Ϯ 0.1fold, n ϭ 3, p Ͻ 0.001) in the cells expressing the IRS-R3Q-FF mutant cells compared with IRS-WT cells (Fig. 6C). The subsequent transfer of SHP-2 to SHPS-1 was also increased in IRS-R3Q-FF mutant cells compared with IRS-WT cells in response to IGF-I (2.1 Ϯ 0.2-fold, n ϭ 3, p Ͻ 0.01; Fig. 6C). When the IRS-R3Q-FF cells were analyzed, p52 shc phosphorylation was found to be increased in response to IGF-I compared with IRS-WT cells (3.0 Ϯ 0.2-fold, n ϭ 3, p Ͻ 0.001). Cells expressing IRS-R3Q-FF cells showed increased phosphorylation of MAPK (3.3 Ϯ 0.2-fold, n ϭ 3, p Ͻ 0.01) compared with cells overexpressing IRS-WT in response to IGF-I (Fig. 6C). Using a lower degree of cellular confluence and increased IGF-I concentration, we observed a decrease in basal Shc phosphorylation response in the mutant cells. Further, in response to IGF-I, the phosphorylation of MAPK increased significantly (3.9 Ϯ 0.1, n ϭ 3, p Ͻ 0.001) in the IRS-R3Q-FF cells (supplemental Fig. 2). The proliferation response of cells expressing IRS-R3Q-FF to IGF-I was found to be significantly greater (2.1 Ϯ 0.3-fold, n ϭ 4, p Ͻ 0.01) compared with cells expressing IRS-WT (Fig. 6D). These results demonstrate that inhibiting IRS-1 binding to IGF-IR and to SHP-2 attenuated IRS-1 tyrosine phosphorylation and Grb2 binding. Furthermore, these changes resulted in increased SHPS-1 association with IGF-IR kinase, SHPS-1 phosphorylation, SHP-2 transfer, phosphorylation of Shc and MAPK, and cellular proliferation in response to IGF-I.

DISCUSSION
IRS-1 is a major substrate for IGF-IR (4,9) and has been shown to mediate several signaling events, including activation of both the phosphatidylinositol 3-kinase and MAPK signaling pathways (4,9,39,40). Previously, we showed that IGF-I-stim-  ulated IRS-1 phosphorylation and subsequent Grb2 binding were markedly down-regulated in VSMCs that had been maintained in high glucose (e.g. 25 mM) (29). The results from the current study confirm the previous observation that VSMCs in high glucose had a significant reduction in total IRS-1 protein levels at 12, 24, and 48 h compared with cells exposed to normal glucose. Exposure to high glucose has been shown to decrease IRS-1 levels in a variety of cell types (32,41). IRS-1 degradation has been extensively studied in insulin-sensitive cell types, such as adipocytes and skeletal muscle cells and also in these tissues in animal models (13,31,(42)(43)(44). Hyperglycemia-induced IRS-1 down-regulation has been shown in vascular smooth muscle following PKC activation (45), ␤-adrenergic stimulation (46), and angiotensin II exposure (47,48). Exposure to growth factors, including IGF-I, also enhances IRS-1 degradation in a variety of cells (25,26,49,50). In the present study, exposure to high glucose and IGF-I for 24 or 48 h reduced the IRS-1 levels significantly compared with cells maintained in high glucose alone. This observation is supported by the previous studies that showed that the combination of high glucose and insulin reduced IRS-1 levels in several model systems and in humans with type 2 diabetes and obesity (30,31,51,52). Furthermore, in the current study, IRS-1 tyrosine phosphorylation and Grb2 binding were impaired in response to IGF-I, suggesting that exposure to high glucose with subsequent IRS-1 down-regulation attenuates IRS-1-mediated signaling.
In an attempt to determine the consequences of the failure to down-regulate IRS-1, we transduced VSMCs with an IRS-1 cDNA and maintained them in 25 mM glucose. Maintenance of this increased IRS-1 level led to impaired IGF-I stimulated SHPS-1 phosphorylation, but it had no effect on IGF-I-stimulated IGF-I receptor phosphorylation. That this effect is direct is supported by our observation that IRS-1 inhibited IGF-IRmediated SHPS-1 phosphorylation in vitro in a concentrationdependent manner. In VSMCs exposed to high glucose, IGF-I induces SHPS-1 phosphorylation (53). SHPS-1 has been demonstrated to be a substrate of various growth factor receptor tyrosine kinases, including the insulin and IGF-I receptor (19, 37, 54 -58). The role of IRS-1 in inhibiting SHPS-1 phosphorylation in intact SMCs was confirmed by using three independent methods. First, in SMCs maintained in normal glucose, SHPS-1 phosphorylation was suppressed, and the addition of a synthetic peptide that inhibited IGF-IR/IRS-1 association enhanced SHPS-1 phosphorylation. Second, the addition of two synthetic peptides that disrupted IRS-1/IGF-IR association to cells maintained in hyperglycemia that were overexpressing IRS-1 led to enhanced SHPS-1 phosphorylation. Third, expression of a mutant form of IRS-1 that could not bind to IGF-IR at levels that were comparable with WT IRS-1 expression also led to enhanced IGF-I-stimulated SHPS-1 phosphorylation. These findings support the conclusion that IRS-1 functions by binding directly to the IGF-I receptor and suppressing IGF-I receptor-mediated SHPS-1 phosphorylation.
Crystal structure and mutational analysis of the IRS-1 PTB domain showed that arginines 212, 213, and 217 are responsible for binding to the NPXY motif in the receptor, and substitution with glutamine resulted in a complete loss of binding (7,10). Our findings confirm the importance of these residues because the mutation of the three arginine residues in the PTB domain of the IRS-1 significantly impaired IRS-1/IGF-IR association and IRS-1 tyrosine phosphorylation following IGF-I stimulation. That IRS-1/IGF-IR association was not completely inhibited can be explained by the fact that IRS-1 also binds to the major autophosphorylation sites (Tyr 1131 , Tyr 1135 , and Tyr 1136 ) of the IGF-IR (9). In addition, inhibition of IGF-IR/IRS-1 by both phosphorylated and non-phosphorylated forms of peptide 299 suggests that both peptides compete with intact IGF-IR for binding to IRS-1 and supports the conclusion that IRS-1/IGF-IR association is occurring through this interaction.
Following phosphorylation of YXX(L/I/V) motifs within the SHPS-1 cytoplasmic domain, SHP-2 is recruited to SHPS-1, and this is essential for optimal phosphorylation of Shc and MAPK activation in response to IGF-I (2). Our study shows that in addition to inhibiting IGF-IR-mediated SHPS-1 phosphorylation, IRS-1 also sequesters SHP-2, limiting its translocation to SHPS-1 following SHPS-1 phosphorylation. IRS-1 also contains seven YXX(L/I/V) motifs that could bind to the SHP-2 Src homology 2 domains, and substitution of IRS-1 Tyr 1172 and Tyr 1222 to phenylalanine has been shown to reduce SHP-2/IRS-1 association (59,60). In this study, we used several approaches to show that IRS-1 functions in part by sequestering SHP-2 and preventing its recruit- A, in VSMCs under normal glucose conditions, in response to IGF-I, competitive binding of IRS-1 to IGF-IR impairs the access of SHPS-1 to IGF-I receptor kinase, thereby decreasing SHPS-1 phosphorylation. Further, IRS-1 sequesters SHP-2, leading to impaired assembly of the signaling complex on SHPS-1. B, in high glucose conditions, reduction in IRS-1 levels allows SHPS-1 improved access to the IGF-I receptor kinase, leading to enhanced phosphorylation of SHPS-1 and subsequent SHP-2 transfer. This facilitates formation of the SHP-2⅐Src⅐Shc⅐Grb-2 signaling complex, which augments phosphatidylinositol 3-kinase and MAPK pathway activation. C, overexpression of IRS-1 decreases SHPS-1 access to the IGF-IR kinase, SHPS-1 phosphorylation, and subsequent complex assembly. The IRS-1 double mutant with impaired binding to IGF-IR and SHP-2 allows increased SHPS-1 access to IGF-IR kinase, thus increasing SHPS-1 phosphorylation and SHP-2 transfer, thereby leading to increased phosphorylation of MAPK and cellular proliferation in response to IGF-I. ment to SHPS-1. First, the cells expressing the IRS-1 Y1179F/ Y1229F substitution mutant had decreased IRS-1/SHP-2 binding in response to IGF-I. Additionally, exposure of IRS1-FF mutant cells to a peptide that inhibited IRS-1 association with IGF-IR increased SHPS-1 phosphorylation and restored the subsequent SHP-2 transfer. Second, expression of a combined mutant (IRS-R3Q-FF) in which both IRS-1/IGF-IR and IRS-1/ SHP-2 binding were decreased led to increases in SHPS-1 phosphorylation and subsequent SHP-2 transfer to SHPS-1. Previous studies have shown that IRS-1/SHP-2 association is differentially regulated in different tissues (61), in response to changes in glucose (62) and state of SMC differentiation (63). These findings suggest that this could be an important mechanism for regulating cellular responsiveness to IGF-I.
In VSMCs, Shc phosphorylation in response to IGF-I depends on SHPS-1 phosphorylation and subsequent SHPS-1 complex assembly (2,29). Although some studies have shown that IRS-1 and Shc are competitive substrates for receptor tyrosine kinase (8,9,64), our results show that competition occurs between SHPS-1 and IRS-1, and this leads to impaired SHPS-1 phosphorylation. Further, IGF-IR overexpression increased SHPS-1 phosphorylation in response to IGF-I. Therefore, attenuation of Shc phosphorylation is not due to direct Shc/ IRS-1 competition for access to IGF-IR. Shc has been shown to co-immunoprecipitate with several activated tyrosine kinase receptors, such as epidermal growth factor receptors, erbB-2/ neu, and Trk (21,(65)(66)(67); however, it does not co-immunoprecipitate with activated IR (68,69). Therefore, the activation of Shc in VSMCs in response to IGF-I is regulated by SHP-2⅐Src transfer to SHPS-1 and Src activation (2,20).
These observations raise the question of the role of IRS-1 in VSMCs maintained in normoglycemic conditions. Our prior studies have shown that when IRS-1 levels are substantially increased (e.g. VSMCs maintained in 5 mM glucose), SHPS-1 phosphorylation, SHP-2 transfer, and Shc/MAPK activation are suppressed (29,70). Consequently, although the protein synthesis response to IGF-I is maintained, the cell migration and proliferation responses are significantly decreased. Our finding that inhibition of IGF-IR/IRS-1 association in VSMCs maintained in 5 mM glucose resulted in significant enhancement of SHPS-1 phosphorylation suggests that this mechanism is operative in VSMCs exposed to normal glucose. This observation could have major physiologic significance for maintenance of normal vascular function. In normal mice, VSMCs are maintained in the quiescent state (71,72) and have a minimal proliferative response to IGF-I administration. In contrast, the presence of diabetes cell proliferation is increased substantially (70). Because differentiated VSMCs in blood vessels are maintained in a non-proliferative state, expression of IRS-1 may be one of the important determinants that allow normal protein synthesis and hypertrophic responses to IGF-I and insulin while functioning simultaneously to inhibit cell replication. In contrast, transition to the dedifferentiated phenotype could possibly require a major decrease in IRS-1 in order to permit MAPK pathway activation (17,73). This conclusion is supported by the results of this study, wherein IRS-1 silencing increased basal IGF-IR phosphorylation, which in turn, in the absence of IRS-1 binding, had increased access to SHPS-1, thereby augmenting basal SHPS-1 phosphorylation. Future studies should investigate the role of IRS-1 in animal models and determine if IRS-1 is regulating the mitogenic response to IGF-I in vascular tissue in vivo.
In summary, exposure of VSMCs to high glucose and IGF-I results in down-regulation of IRS-1, thus allowing IGF-I-stimulated SHPS-1 phosphorylation and SHPS-1 complex assembly. The mechanism by which IRS-1 functions is due to its ability to bind to IGF-IR and inhibit IGF-IR mediated SHPS-1 tyrosine phosphorylation as well as sequestration of SHP-2, which inhibits the assembly of the signaling complex on SHPS-1. These changes result in inhibition of IGF-I-stimulated increases in MAPK activation and cell proliferation (Fig. 7). Taken together with the findings from our previous studies, these results have significant implications for understanding the role of IGF-I in stimulating the proliferation of neointimal VSMCs that have been exposed to hyperglycemic stress.