Absence of IQGAP1 Protein Leads to Insulin Resistance

Abstract

Insulin binds to the insulin receptor (IR) and induces tyrosine phosphorylation of the receptor and insulin receptor substrate-1 (IRS-1), leading to activation of the PKB/Akt and MAPK/ERK pathways. IQGAP1 is a scaffold protein that interacts with multiple binding partners and integrates diverse signaling cascades. Here we show that IQGAP1 associates with both IR and IRS-1 and influences insulin action. In vitro analysis with pure proteins revealed that the IQ region of IQGAP1 binds directly to the intracellular domain of IR. Similarly, the phosphotyrosine-binding domain of IRS-1 mediates a direct interaction with the C-terminal tail of IQGAP1. Consistent with these observations, both IR and IRS-1 co-immunoprecipitated with IQGAP1 from cells. Investigation of the functional effects of the interactions revealed that in the absence of IQGAP1, insulin-stimulated phosphorylation of Akt and ERK, as well as the association of phosphatidylinositol 3-kinase with IRS-1, were significantly decreased. Importantly, loss of IQGAP1 results in impaired insulin signaling and glucose homeostasis in vivo. Collectively, these data reveal that IQGAP1 is a scaffold for IR and IRS-1 and implicate IQGAP1 as a participant in insulin signaling.

Introduction

Insulin resistance, in which cells are unresponsive to normal insulin concentrations, may lead to type 2 diabetes. Although the underlying mechanism remains to be fully elucidated, dysregulated signaling by the insulin receptor substrate proteins IRS-1⁵ and IRS-2 is implicated as a common mechanism for insulin resistance in humans (1). While most receptor tyrosine kinases bind directly to cytoplasmic effectors, insulin receptor (IR) and insulin-like growth factor receptor associate with effectors via IRS proteins (2). IRS phosphotyrosine binding (PTB) and pleckstrin homology (PH) domains are responsible for its interaction with IR (3). Upon binding insulin, the IR undergoes autophosphorylation at several tyrosine residues, leading to its activation (reviewed in Ref. 2). Active IR phosphorylates the IRS proteins on tyrosine residues, mediating recruitment of phosphatidylinositol 3-kinase (PI3K) and Grb2 that stimulate the Akt and mitogen-activated protein kinase (MAPK) pathways, respectively. PI3K phosphorylates phosphatidylinositol 4,5-bisphosphate and increases intracellular phosphatidylinositol 3,4,5-trisphosphate (PIP₃) at the plasma membrane. PIP₃ binds to Akt, resulting in its translocation to the membrane, where it is phosphorylated by 3-phosphoinositide-dependent protein kinase (PDK-1), leading to its activation (4, 5). IR/IRS also activate the MAPK pathway to control cell growth and differentiation (2). Together, PI3K and MAPK signaling regulate glucose homeostasis and proliferation, respectively, in response to insulin.

Protein scaffolds exert a crucial role in regulating signal transduction. Scaffolds organize components of signaling pathways and increase the efficiency of protein interactions (6, 7). In mammals, scaffolds have been most extensively investigated in the MAPK signaling cascade. More recently, a few scaffold proteins that regulate insulin signaling have been identified. For example, interaction of the scaffold CNK1 (connector enhancer of kinase suppressor of Ras) with cytohesin is critical for activation of the PI3K/AKT cascade downstream of IR and insulin-like growth factor receptor-1 (8). Another scaffold protein, sequestosome 1/p62, has been identified as one of the binding partners of IRS-1 (9). Its overexpression led to increased phosphorylation of Akt, GLUT4 translocation, and glucose uptake. β-Arrestins are adaptor proteins that link G-protein-coupled receptors to downstream signaling pathways (10). In mice, insulin induces formation of a complex in which β-arrestin-2 scaffolds Akt and Src to IR (11). Consistent with these findings, β-arrestin-2 knock-out mice exhibit insulin resistance.

The Ras GTPase-activating-like protein, IQGAP1, is a ubiquitously expressed scaffold protein with multiple interaction motifs (12–14). These motifs include a calponin homology domain near the N terminus, followed by a WW domain, four tandem IQ motifs, a Ras GAP-related domain, and a Ras GAP_C terminus (RGCT). Through its interactions with numerous proteins, IQGAP1 modulates many cellular functions, such as cell-cell adhesion, transcription, and cytoskeletal architecture (15, 16).

IQGAP1 also participates in signaling cascades that are mediated by a heterogeneous group of receptors (12, 16). It binds to and regulates the function of selected receptors such as estrogen receptor (17), platelet-derived growth factor receptor (PDGFR) (18), epidermal growth factor receptor (EGFR) (19), human epidermal growth factor receptor-2 (HER2) (20), and vascular endothelial growth factor receptor-2 (VEGFR2) (21). We previously reported that IQGAP1 is a scaffold in the MAPK signaling pathway. IQGAP1 binds directly to both mitogen-activated protein kinases (MEK1/2) and extracellular signal-regulated kinase (ERK1/2) and modulates MAPK activity in response to EGF (22, 23). IQGAP1 also interacts with MAPK components proximal to MEK, namely B-Raf (24). Furthermore, the loss of IQGAP1 reduced VEGF-stimulated Akt signaling (21). Collectively, these observations indicate that IQGAP1 has the capacity to integrate multiple signaling cascades downstream of cell surface receptors.

An important unresolved question in insulin signaling is the precise mechanism by which IRS-1 is coupled to IR. Because IQGAP1 binds to several growth factor receptors and modulates their signaling cascades, we hypothesized that IQGAP1 may serve as an adaptor protein for IR. In this study, we investigated this possibility and show that IQGAP1 binds directly to both IR and IRS-1 and appears to be necessary for normal insulin signaling and glucose homeostasis.

Results

IQGAP1 Interacts with IR in Cells

To evaluate whether IQGAP1 interacts with IR, several complementary strategies were adopted. In the first approach, we used CHO/IR cells that stably overexpress IR. CHO/IR cells were lysed and incubated with GST-IQGAP1, and complexes were isolated with glutathione-Sepharose beads. Endogenous IR in cell lysates bound to GST-IQGAP1 (Fig. 1A, top). No IR was present in the samples that were incubated with GST alone, validating the specificity of the interaction.

FIGURE 1.

IR and IQGAP1 interact in cells and in vitro. A, GST pull-down of IR from CHO/IR cell lysates using GST-IQGAP1 or GST alone analyzed by immunoblotting (top panels) and Coomassie Blue staining (bottom panels). B, immunoprecipitation (IP) with anti-IQGAP1 antibodies or NIRS from CHO/IR/IRS-1 cells transfected with Myc-IQGAP1. C, IP with anti-IQGAP1 antibodies or NIRS of endogenous IQGAP1 from C2C12 cells. D, schematic representation of IR-β and IR_cyt. TM/JM, transmembrane and juxtamembrane domains; TK, tyrosine kinase; C-term, C-terminal tail. E, GST pull-down of purified IQGAP1 using GST-IR_cyt or GST. Samples were analyzed by immunoblotting (top) and Coomassie Blue staining (bottom). Shown is GST pull-down of His₆-IR-β (F) or His₆-IR_cyt (G) using GST-IQGAP1 or GST alone. The middle portion of the gel was processed by Western blotting (top), whereas the top and bottom portions of the gel were stained with Coomassie Blue (middle and bottom). H, binding of biotin-labeled IQGAP1 proteins expressed using T_NT to GST-IR_cyt and GST alone. All data are representative of at least three independent experiments.

Next, we examined the binding between IQGAP1 and IR using immunoprecipitation. CHO/IR/IRS-1 cells were transiently transfected with Myc-tagged IQGAP1 and exposed to the cross-linker dithiobis(succinimidyl propionate) (DSP) before lysis. DSP reacts with primary amine groups and cross-links bound proteins. Immunoprecipitation with anti-IQGAP1 polyclonal antibodies revealed that IR co-immunoprecipitated with IQGAP1 from CHO/IR/IRS-1 cells (Fig. 1B). Minimal IR was detected in samples precipitated with non-immune rabbit serum (NIRS), confirming the specificity of the interaction.

Binding of endogenous proteins was evaluated in C2C12 myoblasts. IR specifically co-immunoprecipitated with IQGAP1 but not with NIRS (Fig. 1C). Collectively, these data reveal that IR associates with IQGAP1 in cells.

IQGAP1 and IR Bind Directly in Vitro

To determine whether IQGAP1 interacts directly with IR, in vitro binding analysis with pure proteins was performed with IR-β (the β-subunit of IR, encompassing amino acids 763–1382) or IR_cyt (the cytosolic domain of the β-subunit of IR, encompassing amino acids 980–1382) proteins (Fig. 1D).

Purified GST-IR_cyt was incubated with purified IQGAP1, and complexes were isolated with glutathione-Sepharose. Western blotting reveals that IQGAP1 bound to IR_cyt (Fig. 1E, top). Minimal IQGAP1 was present in samples incubated with GST alone, revealing the specificity of the interaction. A Coomassie-stained gel of purified IR_cyt protein is shown in Fig. 1E (bottom). The lower molecular weight bands are probably degradation products that are generated during the assay.

Additionally, binding was examined using purified His₆-tagged constructs of IR and GST-IQGAP1. His₆-IR-β or His₆-IR_cyt was incubated with GST-IQGAP1, and complexes were isolated with glutathione-Sepharose. Western blotting reveals that both IR-β and IR_cyt bound to IQGAP1 (Fig. 1, F and G, top panels). No IR was present in samples incubated with GST alone, revealing the specificity of the interactions. Coomassie-stained gels of purified GST-IQGAP1 (Fig. 1, F and G, middle panels) and GST proteins (Fig. 1, F and G, bottom panels) are shown. Collectively, these data reveal a direct interaction between IQGAP1 and IR and indicate that the cytosolic domain of the β-subunit of IR is sufficient for binding to IQGAP1.

Identification of IR Binding Region on IQGAP1

We used the TNT system to identify the region of IQGAP1 to which IR binds. Biotinylated IQGAP1 constructs were expressed in rabbit reticulocyte lysate using the TNT system. As observed with pure IQGAP1 (Fig. 1, E–G), full-length IQGAP1 expressed in reticulocyte lysate binds to IR (Fig. 1H, middle). Whereas IR_cyt bound the N-terminal half of IQGAP1, no binding to the C-terminal half was detected. The isolated IQ domain (residues 717–916) bound IR_cyt (Fig. 1H)_. The IQGAP1 constructs did not bind to GST alone (Fig. 1H, right), verifying the specificity of the interaction. The amount of IQGAP1 fragments used (input) was similar. These data reveal that the IQ domain of IQGAP1 is sufficient for IR binding.

IRS-1 Binds to IQGAP1

IQGAP1 binds to EGFR (19) and several EGFR downstream effector proteins (23, 24). Therefore, we hypothesized that IQGAP1 may interact with components downstream of IR, specifically IRS-1. We first evaluated the interaction of IRS-1 and IQGAP1 in a normal cell milieu. 3T3-L1 cells were lysed and incubated with GST-IQGAP1. Analysis by Western blotting shows that endogenous IRS-1 binds to GST-IQGAP1 (Fig. 2A). The specificity of the binding is validated by the absence of IRS-1 from samples that were incubated with GST alone.

FIGURE 2.

IRS-1 binds to IQGAP1 in cell lysates. A, GST pull-down of endogenous IRS-1 from 3T3-L1 preadipocyte lysates using GST-IQGAP1 or GST alone. B, IP with anti-IQGAP1 polyclonal antibody from CHO/IR/IRS-1 cells transfected with Myc-IQGAP1. NIRS was the negative control. C, IP with anti-IQGAP1 antibodies or NIRS of endogenous IQGAP1 from C2C12 cells. D, schematic representation of IRS-1 constructs. Binding to IQGAP1 is depicted as follows: +, binding detected; −, no binding. E, GST pull-down of endogenous IQGAP1 from HEK-293H cells using GST-tagged IRS-N (N), IRS-C (C) or GST alone. Samples were resolved by SDS-PAGE. The top panel was probed for IQGAP1, and the bottom panel was stained with Coomassie Blue. F, GST pull-down of [³⁵S]methionine-labeled IRS-1 proteins using full-length GST-IQGAP1 (IQF) and GST alone. G, quantification of binding. Data are means ± S.E. (error bars) with binding of IRS-1 peptide 162–267 to IQGAP1 set as 1. The data are representative of three independent experiments.

This interaction was further examined by immunoprecipitation. CHO/IR/IRS-1 cells were transiently transfected with Myc-tagged IQGAP1. Immunoprecipitation of cell lysates with anti-IQGAP1 polyclonal antibodies revealed that IRS-1 co-precipitates with IQGAP1 (Fig. 2B, top). Minimal IRS-1 is present in precipitates from cells incubated with NIRS.

Finally, endogenous IRS-1 co-immunoprecipitates with endogenous IQGAP1 from C2C12 myoblasts (Fig. 2C). These data indicate that IQGAP1 and IRS-1 interact in cells.

IQGAP1 Binds to the PTB Domain of IRS-1

To determine the IQGAP1 binding region on IRS-1, binding studies with multiple fragments of IRS-1 were performed. The IQGAP1 binding sites on IRS-1 are summarized in Fig. 2D. Lysates of HEK-293H cells were incubated with the fusion proteins, namely GST-tagged IRS-N (N-terminal IRS-1; amino acids 2–564), IRS-C (C-terminal IRS-1; amino acids 511–1242), or GST alone. Complexes were isolated with glutathione-Sepharose, resolved by SDS-PAGE, and processed by Western blotting. Examination of the blots reveals that IQGAP1 binds only to the N-terminal portion of IRS-1 (Fig. 2E, top). IQGAP1 does not bind to the C-terminal region of IRS-1 or to GST alone. The Coomassie-stained gel shows that the amounts of IRS-N and IRS-C used were equivalent in the samples (Fig. 2E, bottom).

To narrow the IQGAP1 binding region on IRS-1, N-terminal fragments of IRS-1 were expressed, including amino acids 2–300, which contains both the PH (amino acid residues 12–115) and the PTB (amino acid residues 160–264) domains. In addition, the isolated PH and PTB were expressed alone. These IRS-1 fragments were labeled with [³⁵S]methionine, and their binding to purified full-length IQGAP1 was examined. IRS-1(2–300) binds IQGAP1 (Fig. 2F, left). In addition, IRS-1(162–267), which comprises the PTB domain, binds GST-IQGAP1, but IRS-1(2–115) does not bind. The input gel reveals that equivalent amounts of IRS-1 peptides were present in the samples (Fig. 2F, right). Quantification of three independent experiments shows minimal binding to the PH domain (amino acids 2–115) but clear binding to amino acids 2–300 and 162–267, which contain the PTB (Fig. 2G). These data indicate that the PTB domain, but not the PH domain, of IRS-1 is necessary for binding IQGAP1.

Identification of IRS-1 Binding Region on IQGAP1

The region on IQGAP1 to which IRS-1 binds was investigated using GST pull-down assays of selected IQGAP1 fragments (Fig. 3A). HEK-293H cells were transfected with GFP-tagged IQGAP1 constructs, namely full-length IQGAP1 (Fig. 3, FL), the N-terminal half of IQGAP1 (N), and the C-terminal half of IQGAP1 (C). pEGFP-C1 (GFP) alone was used as the negative control. Cells were lysed and incubated with GST-tagged IRS-1(2–300), which contains both the PH and PTB, or GST alone. Complexes were isolated with glutathione-Sepharose beads. As observed with the interaction of GST-IQGAP1 with endogenous IRS-1 (Fig. 2), full-length IQGAP1 binds to the N-terminal region of GST-IRS-1(2–300) (Fig. 3B). Examination of portions of IQGAP1 reveals that the C-terminal half (amino acids 864–1657), but not the N-terminal half (amino acids 2–863), of IQGAP1 binds to IRS-1.

FIGURE 3.

IRS-1 binds to the C-terminal region of IQGAP1. A, schematic representation of IQGAP1 constructs. Binding to IRS-1 is depicted as follows: +, binding detected; −, no binding. B, GST pull-down from HEK-293H cells transiently transfected with GFP-tagged full-length IQGAP1 (FL), IQGAP1-N (N), or IQGAP1-C (C) using GST-IRS-1(2–300) or GST alone. Blots were probed with anti-GFP antibody. C and D, GST pull-down (PD) of [³⁵S]methionine-labeled IQGAP1 fragments using GST-IRS-1(2–300) or GST alone. Input is shown in the bottom panel. Gels were dried and processed by autoradiography. The data are representative of at least two independent experiments.

To narrow the IRS-1 binding region, selected fragments spanning the C-terminal half of IQGAP1 were generated. Amino acids 914–1657 of IQGAP1 interact with IRS-1(2–300) (Fig. 3C, top). The amount of radiolabeled IQGAP1 fragments used (input) was similar in all of the samples (Fig. 3C, bottom). Deletion of amino acids from the N terminus of peptide 914–1657 revealed that amino acids 1358–1657 of IQGAP1 bind IRS-1 (Fig. 3D, top). Truncation of this fragment by 50 amino acids from the N terminus to yield amino acids 1408–1657 abrogated binding (Fig. 3D, top). Similarly, removal of 100 or 150 residues from the C-terminal end (to yield fragments 1308–1557 and 1308–1507, respectively) eliminated binding. Binding of IRS-1 to the amino acids 1358–1657 of IQGAP1 was confirmed by dot blotting (data not shown). The amount of radiolabeled IQGAP1 fragments used (input) was similar in all of the samples (Fig. 3D, bottom). These data reveal that amino acids 1358–1657 of IQGAP1 bind to IRS-1. Thus, the Ras GAP_C terminus (RGCT) and distal C terminus of IQGAP1 are required for IRS-1 binding.

IRS-1 Associates with IR and IQGAP1

A GST fusion protein of full-length IRS-1 was incubated with purified His₆-IR_cyt alone or with His₆-IR_cyt and IQGAP1 together. Complexes were isolated with glutathione-Sepharose. Examination of the blot revealed that IR_cyt binds directly to IRS-1 (Fig. 4A). When both IR_cyt and IQGAP1 were incubated with GST-IRS-1, all three proteins were detected in the pull-down. Note that the addition of IQGAP1 did not significantly alter the amount of IR that bound to IRS-1, whereas there was a slight, but significant, increase in the amount of IQGAP1 bound to IR-IRS-1 compared with IQGAP1 bound to IRS-1 alone (Fig. 4, A and B). Equal amounts of IRS-1 were present in the samples. Specificity of the interactions was revealed by the minimal amounts of IQGAP1 and IR detected in the samples incubated with GST alone (Fig. 4A). Collectively, these data reveal that IR, IRS-1, and IQGAP1 associate in vitro.

FIGURE 4.

IRS-1, IR, and IQGAP1 associate. A, GST-IRS-1 or GST was preincubated with purified His₆-IR_cyt or IQGAP1. After removing unbound proteins, IQGAP1 or IR_cyt was added where indicated. IQGAP1 and IR_cyt binding were assessed by Western blotting. B, quantification of binding. Data are means ± S.E. (error bars) with binding to IQGAP1 or IR_cyt incubated with GST-IRS-1 alone set as 1 (n = 4). C, serum-starved CHO/IR/IRS1 cells, transfected with GFP-IQGAP1, were stimulated with (+) or without (−) 100 nm insulin for 10 min. Samples were immunoprecipitated with anti-IQGAP1 or rabbit IgG antibodies. Western blots were probed for IRS1, phospho-IR (pIR) Tyr-972, and IR, and then blots were stripped and reprobed with anti-IQGAP1 antibody, which detects both GFP-IQGAP1 and endogenous IQGAP1. D, quantification of binding corrected for the amount of IQGAP1 immunoprecipitated. Data are expressed as means ± S.E. with binding minus insulin set as 1 (n = 4).

IQGAP1 Association with IR and IRS-1 Is Insulin-independent

To assess whether the interaction of IR or IRS-1 with IQGAP1 is regulated by insulin, CHO/IR/IRS1 cells were transfected with GFP-IQGAP1. After starving them of serum overnight, insulin or vehicle was added to cells, and then IQGAP1 was immunoprecipitated. Immunoblots (Fig. 4C) demonstrate that IR and IRS-1 associate constitutively with IQGAP1. Insulin enhanced the tyrosine phosphorylation of IR (Fig. 4C, right). Analysis of four independent experiments revealed that insulin slightly, but not significantly, reduced binding of IR and IRS-1 to IQGAP1 (Fig. 4D). This result indicates that IR and IRS-1 associate constitutively with IQGAP1 in cells, and insulin has a minimal effect on this interaction.

IQGAP1 Modulates Insulin Signaling in Cells

MEF cells derived from control and IQGAP1-null mice were used to determine whether IQGAP1 has a role in insulin signaling. Cells were starved for 16 h and then stimulated with 100 nm insulin to activate the insulin signaling cascade. To examine its phosphorylation, IR was immunoprecipitated and resolved by Western blotting. Probing with anti-phosphotyrosine antibody demonstrates that there is no significant difference in tyrosine phosphorylation of IR-β between control and IQGAP1-null MEFs (Fig. 5, A and B). These data indicate that IQGAP1 does not influence IR-β activation by insulin.

FIGURE 5.

Effect of IQGAP1 on IR and IRS-1. Serum-starved control (WT) and IQGAP1-null (−/−) MEFs were incubated with 100 nm insulin for the indicated times. A, lysates were immunoprecipitated with anti-IR-β or IgG (control) antibodies, and blots were probed with anti-phosphotyrosine (pTyr) and anti-IR-β antibodies. B, quantification with Image Studio (LI-COR Biosciences). The amount of phosphotyrosine was corrected for IR-β immunoprecipitated from the same sample. Data are expressed as the means ± S.E. (error bars) with control cells at 2 min set as 1 (n = 3). C, IP of endogenous IRS-1. Blots were probed with anti-IRS-1 and anti-phosphotyrosine antibodies. D, the amount of phosphotyrosine signal was quantified as described for B and corrected for the amount of IRS-1 immunoprecipitated from the same sample. Data are means ± S.E. with control cells at 15 min set as 1 (n = 8). E, CHO/IR/IRS-1 cells were transfected with siRNA against IQGAP1 (siIQ) or nonspecific control siRNA sequences (Ctrl). Cells were treated with (+) or without (−) 100 nm insulin for 1 min, and IRS-1 was immunoprecipitated. Western blotting of lysates and immunoprecipitates was performed, and blots were probed for the proteins indicated. F, the amount of co-immunoprecipitated IR was quantified and corrected for IRS-1 immunoprecipitated from the same sample. Data are expressed as means ± S.E. with cells transfected with control siRNA and treated with vehicle set as 1 (n = 5). G, MEFs were incubated with (+) or without (−) 100 nm insulin for 5 min, and IRS-1 was immunoprecipitated. Western blotting of lysates and immunoprecipitates was performed, and blots were probed for the proteins indicated. H, the amount of co-immunoprecipitated p85α was quantified and corrected for the amount of IRS-1 immunoprecipitated from the same sample. Data are expressed as means ± S.E. with wild type MEFs incubated with insulin set as 1 (n = 5). *, p < 0.05; **, p < 0.01.

To evaluate insulin signaling downstream of IR, activation of IRS-1 was examined. IRS-1 was immunoprecipitated from MEFs stimulated with insulin for 0, 5, or 15 min. There is a modest (17.7 and 23.9%), yet significant, reduction in tyrosine phosphorylation of IRS-1 at 5 and 15 min, respectively, after insulin stimulation in IQGAP1-null MEFs (Fig. 5, C and D).

To ascertain whether IQGAP1 influences the association between IR and IRS-1, we compared the binding between IR and IRS-1 in control cells and cells with IQGAP1 knockdown. Because we were unable to reliably co-immunoprecipitate IR with IRS-1 from MEFs, we used CHO/IR/IRS-1 cells. siRNA knockdown reduced the amount of IQGAP1 in CHO/IR/IRS-1 cells by 66% (Fig. 5E). The interaction of IR with IRS-1 was examined by co-immunoprecipitation. In cells transfected with control siRNA, insulin stimulated the association of IR with IRS-1 by 2.1-fold (Fig. 5, E and F). By contrast, cells with knockdown of IQGAP1 showed greater constitutive association of IR with IRS-1, and insulin was unable to enhance the binding. These data suggest that loss of IQGAP1 may alter, but not prevent, the interaction of IR with IRS-1.

To determine whether the loss of IQGAP1 affects signaling downstream of IRS-1, the association of PI3K with IRS-1 was investigated. Insulin enhanced by 3.7-fold the amount of the p85α subunit of PI3K that co-immunoprecipitated with IRS-1 from control MEFs (Fig. 5, G and H). The ability of insulin to augment the association of p85α with IRS-1 was significantly reduced (by 35%) in cells lacking IQGAP1. These data reveal that IQGAP1 is necessary for optimal coupling of IRS-1 to PI3K in response to insulin.

In addition, we evaluated activation of the PI3K/Akt pathway using antibodies that specifically recognize the phosphorylation of Akt at Ser-473. Phosphorylation at Ser-473 of Akt is required for maximal activation of Akt and activation of other downstream signaling proteins (25). Akt phosphorylation in control MEFs increased significantly (p < 0.05) after 5 min of insulin stimulation and remained elevated for up to 30 min (Fig. 6, A and B). Although insulin slightly promoted Akt phosphorylation in IQGAP1-null MEFs, the increase was not statistically significant and was considerably less than that in control MEFs. Insulin-stimulated Akt phosphorylation in control MEFs was significantly higher (p < 0.05) than that in IQGAP1-null MEFs at all of the time points examined (Fig. 6, A and B).

FIGURE 6.

Effect of IQGAP1 on Akt and ERK. A, MEFs were serum-starved and incubated with insulin for 0, 5, 15, or 30 min. After lysis, samples were processed by Western blotting. A, blots were probed for phospho-Akt Ser-473 (pAkt), Akt, and β-tubulin (loading control). B, after quantification, phospho-Akt was corrected for total Akt present in the same sample. Data are means ± S.E. (error bars) with control cells at 0 min set as 1 (n = 6). C, blots were probed with anti-pERK1/2 Thr-202/Tyr-204, ERK1/2, and β-tubulin antibodies. D, after quantification, phospho-ERK (pERK) was corrected for total ERK in the same sample. Data are means ± S.E. with control cells at 0 min set as 1 (n = 5). E, control and IQGAP1^−/− MEFs were transfected with empty vector (V) or Myc-tagged IQGAP1 full-length (FL) or N-half (N) plasmids. Cells were serum-starved and incubated with (+) or without (−) insulin for 10 min and processed by Western blotting with the antibodies indicated. F, quantification of phospho-Akt, corrected for Akt. Data are means ± S.E. with insulin-treated IQGAP1^−/− cells transfected with empty vector set as 1 (n = 6). G, control and IQGAP1^−/− MEFs were transfected with empty vector (V) or GFP-tagged C-half of IQGAP1 (C) plasmids. Cells were serum-starved and incubated with (+) or without (−) insulin for 10 min and processed by Western blotting with the antibodies indicated. H, quantification of phospho-Akt, corrected for Akt. Data are means ± S.E. with insulin-treated IQGAP1^−/− cells transfected with empty vector set as 1 (n = 4). *, p < 0.05; **, p < 0.01; n.s., not significant.

The MAPK pathway is also stimulated by insulin downstream of IRS-1 (2). Activation of the MAPK pathway was evaluated using an antibody specific for ERK Thr-202/Tyr-204 phosphorylation. Following insulin stimulation, robust ERK phosphorylation was observed in control MEFs (Fig. 6, C and D). ERK phosphorylation in IQGAP1-null cells was significantly lower than in control MEFs, both 5 (p < 0.05) and 30 (p < 0.01) min after the addition of insulin. Collectively, these observations demonstrate that IQGAP1 is required for maximal insulin-stimulated phosphorylation of Akt and ERK.

To verify that loss of IQGAP1 from MEFs is responsible for the reduced Akt response to insulin, IQGAP1-null MEFs were transfected with IQGAP1. Re-expression of full-length IQGAP1 significantly increased insulin-stimulated phosphorylation of Akt over that in cells transfected with empty vector (Fig. 6, E and F). The incomplete rescue of Akt signaling by full-length IQGAP1 is probably due to the relatively low expression of the IQGAP1 protein during transient transfection. By contrast, IQGAP1-null cells transfected with the N-half of IQGAP1, which is unable to bind IRS-1 but is capable of interacting with IR and Akt, had Akt activation comparable with cells transfected with empty vector (Fig. 6, E and F). Similarly, expression of the C-half of IQGAP1, which associates with IRS-1, but not with IR, Akt, or MAPK components, failed to alter the reduced Akt phosphorylation in IQGAP1-null cells (Fig. 6, G and H).

Insulin Signaling Is Impaired in IQGAP1-null Mice

To evaluate the role of IQGAP1 in insulin signaling in vivo, control and IQGAP1-null mice fed a normal diet were fasted for 4 h, followed by intraperitoneal injection of saline or insulin. Insulin-sensitive tissues (liver, epididymal fat pad, and skeletal muscle (quadriceps)) were harvested and homogenized, and insulin signaling was assessed by IR and Akt phosphorylation. Consistent with the lack of an effect of IQGAP1 on the ability of insulin to activate IR in cultured cells (Fig. 5, A and B), the extent of tyrosine phosphorylation of IR in liver samples from insulin-treated control and IQGAP1-null mice did not differ significantly (Fig. 7, A and B). Investigation of signaling downstream of IR revealed that insulin stimulated a robust increase in phosphorylation of Akt in all tissues examined from control mice (Fig. 7, C–H). Strikingly, IQGAP1-null mice displayed impaired activation of Akt in response to insulin, with statistically significant differences detected in liver (36% less than control mice), quadriceps muscle (64% less), and epididymal fat (67% less). Collectively, these data reveal that IQGAP1 is required for insulin to maximally stimulate Akt in classic insulin-sensitive tissues.

FIGURE 7.

Insulin-stimulated Akt phosphorylation is impaired in IQGAP1-null mice. Control (WT) and IQGAP1-null (−/−) mice fed a normal diet were fasted for 4 h. 15 min after intraperitoneal saline or insulin injection, mice were euthanized, and tissues were frozen and homogenized. A, IR was immunoprecipitated from lysates, and Western blots were probed with antibodies to phosphotyrosine and IR-β. B, quantification of tyrosine phosphorylation in A corrected for the amount of IR immunoprecipitated from the same sample. Data are means ± S.E. (error bars) with WT mice treated with insulin set to 1 (n = 3). To compare Akt activation, Western blotting for pAkt (Ser-473) and total Akt was performed using liver (n = 6) (C), quadriceps muscle (n = 6) (E), and epididymal fat (n = 4) (G) isolated from the mice. D, F, and H, pAkt levels were quantified and corrected for total Akt in the same sample. Data, normalized to insulin-treated control mice set as 1, are means ± S.E. *, p < 0.05; **, p < 0.01; n.s., not significant.

IQGAP1-null Mice Have Impaired Glucose Tolerance

To determine whether the absence of IQGAP1 protein influences glucose tolerance, control and IQGAP1-null mice were subjected to an intraperitoneal glucose tolerance test. After an overnight fast, mice received an intraperitoneal injection of glucose (0.01 ml of 20% glucose/g of body weight), and blood glucose concentrations were monitored over a 2-h period. In the control mice, blood glucose concentrations reached a peak 15 min after glucose injection and approached baseline by 120 min (Fig. 8A). The glucose tolerance curve was altered in the absence of IQGAP1. In IQGAP1-null mice, peak glucose concentration was reached at 30 min and declined more slowly than in control mice (Fig. 8A). Moreover, glucose concentrations at 120 min in IQGAP1-null mice were significantly higher (p < 0.01) than those in control mice. Fasting glucose concentrations in control and IQGAP1-null mice were the same.

FIGURE 8.

Glucose regulation is impaired in IQGAP1-null mice. Glucose tolerance tests were performed using control (filled circles) and IQGAP1-null (empty circles) mice fed with either a regular diet (n = 20) (A) or a high fat diet (n = 15) (B). Data are means ± S.E. (error bars). C, insulin tolerance tests were performed in control (filled circles) and IQGAP1-null (empty circles) mice fed a regular diet. Data are means ± S.E. (n = 6). D, serum insulin was measured in mice fed or fasted for 16 h. Data are means ± S.E. (fed, 7 control and 4 IQGAP1^−/−; fasted, 5 control and 7 IQGAP1^−/−). p values are for control versus IQGAP1-null mice. *, p < 0.05; **, p < 0.01.

We also evaluated glucose tolerance in mice that had been fed a high fat diet (60% fat calories) for 8 weeks. This analysis was performed to exacerbate any potential aberrations in glucose tolerance (26). As expected, fasting glucose concentrations in both control and IQGAP1-null mice fed a high fat diet were higher than those in mice fed a regular diet (Fig. 8, A and B). In control mice, peak glucose concentrations were reached at 15 min (as observed with the control mice on the regular diet), and baseline was nearly reached by 120 min (Fig. 8B). In IQGAP1-null mice on a high fat diet, glucose concentrations peaked at 30 min, as observed in the mice on regular diet. However, the difference in glucose clearance in control and IQGAP1-null mice was significant at all the time points examined (except for the fasting glucose concentrations) (Fig. 8B). At 120 min, glucose concentration in IQGAP1-null mice remained 1.7-fold higher than the fasting concentrations and was highly significant (p < 0.01) (Fig. 8B). These data reveal that the IQGAP1-null mice are not able to clear glucose as well as the control mice, indicating probable involvement of IQGAP1 in glucose homeostasis.

To expand on these results, an intraperitoneal insulin tolerance test was performed on mice fed a regular diet. After a short fast (4 h), control and IQGAP1-null mice were injected with insulin intraperitoneally. Blood glucose concentrations were measured at baseline (t = 0, 10 min before injection) and 30, 60, and 90 min postinjection. The ability of insulin to lower blood glucose concentrations was reduced in IQGAP1-null mice. In addition, blood glucose concentrations returned to basal levels 90 min after insulin in IQGAP1-null mice, whereas glucose values in control mice remained significantly lowered (Fig. 8C). Additionally, insulin concentrations in fed and fasted mice were measured to determine whether there was a difference between control and IQGAP1-null mice in global insulin secretion. In fed mice, serum insulin concentrations in control (1.35 ± 0.12 ng/ml) and IQGAP1-null (1.25 ± 0.11 ng/ml) mice were essentially the same (Fig. 8D). Moreover, there were no significant differences in insulin concentrations between these mice when fasted. The data suggest that the defects in glucose homeostasis in the mice were probably due to impaired insulin signaling in the absence of IQGAP1.

Discussion

Scaffold proteins exert critical functions in signal transduction (6, 7). IQGAP1 is a scaffold protein that participates in signaling initiated by a variety of receptors, ranging from EGFR and HER2 to PDGFR and estrogen receptor (12, 13, 16). By binding to diverse proteins, IQGAP1 assembles multiprotein complexes that facilitate and modulate signaling. In this study, we describe a previously unidentified role for IQGAP1 as a component of IR signaling. We observed that IQGAP1 binds directly in a phosphotyrosine-independent manner to both IR and IRS-1. Using IQGAP1-null mice, we demonstrate that IQGAP1 is required for maximum insulin action and normal glucose homeostasis.

IQGAP1 associates directly with the cytoplasmic region of the β-subunit of IR. Interestingly, loss of IQGAP1 does not significantly alter insulin-stimulated phosphorylation of IR (Figs. 5 (A and B) and 7 (A and B)). By contrast, prior studies documented that decreased expression of IQGAP1 reduced EGF-stimulated EGFR activation (19), reduced tyrosine phosphorylation of HER2 (which is ligand-independent) (20), and decreased PDGF-stimulated phosphorylation of PDGFRβ (18). Thus, IQGAP1 binding regulates phosphorylation and activation of several growth factor receptors. Although IQGAP1 does not influence IR phosphorylation by insulin, it modulates signaling downstream of insulin.

The best characterized substrates of IR are the IRS proteins, which are phosphorylated at multiple tyrosine residues by IR (2). Here we document direct binding between IQGAP1 and IRS-1. The vast majority of proteins that bind to IRS-1 are recruited via their SH2 domains to phosphotyrosine residues on IRS-1. There are no SH2 domains in IQGAP1, and tyrosine phosphorylation of IRS-1 is not required for the association. Several other proteins have been identified that do not contain SH2 domains and interact with non-phosphorylated IRS-1. These include calmodulin (27), the σ3A subunit of the adaptor protein complex-3 (28), and the Ca²⁺ ATPase SERCA (29). We now identify IQGAP1 as an additional phosphotyrosine-independent binding partner of IRS-1.

Using a panel of deletion constructs and fusion peptides, we identified the regions of IRS-1 and IQGAP1 that mediate association. Importantly, the PTB domain of IRS-1 is sufficient for binding IQGAP1. PTB domains are present in several signaling molecules, including Shc, IRS-1, and DOK1 (30). The PTB domains usually bind to phosphotyrosine residues, many of which are located in NPXY motifs. For example, the PTB of IRS-1 binds to NPXY in the IR (31). A recent report described binding of IQGAP1 to the PTB of Shc (32). Analogous to our observations with IRS-1, tyrosine phosphorylation of IQGAP1 was not required for the Shc interaction. NMR analysis revealed that residues 401–533 of IQGAP1 mediated Shc binding (32). By contrast, we established that amino acid residues 1358–1657 of IQGAP1 are required for binding IRS-1. Moreover, we observed no binding of IRS-1 to amino acids 401–533 of IQGAP1 (data not shown). These data reveal that the interaction of IQGAP1 with the IRS-1 PTB is distinct from its interaction with the Shc PTB.

Although the PTB of IRS-1 improves its coupling to IR, the IRS-1 PH domain is the principle element that links it to IR (33). We observed that IRS-1(162–267), which contains all of the residues that are involved in binding the juxtamembrane region of IR (31), interacts with IQGAP1. Notwithstanding the overlap, we did not detect competition between IR and IQGAP1 for binding IRS-1. Because the PTB can associate with IR and IQGAP1, there may be distinct binding sites for each protein within the PTB. More intricate interactions among IR, IRS-1, and IQGAP1 may mediate complex formation. For example, both IR (reviewed in Ref. 34) and IQGAP1 (35) dimerize. Our data with pure proteins and immunoprecipitation suggest that IR and IRS-1 form a complex with IQGAP1. This concept is supported by our finding that IR and IRS-1 bind to distinct regions on IQGAP1, namely the IQ and C-terminal regions, respectively. Furthermore, there was a significant, albeit small, increase in the amount of IQGAP1 bound to IR-IRS-1 compared with IRS-1 alone (Fig. 4B).

Consistent with the hypothesis that IQGAP1 functions downstream of insulin, we observed that signaling distal to IR is impaired by loss of IQGAP1. The ability of insulin to induce tyrosine phosphorylation of IRS-1 was slightly attenuated in IQGAP1-null MEFs (Fig. 5, C and D). Although relatively small changes were observed in IRS-1 phosphorylation, insulin signaling downstream of IRS-1 was impaired more substantially in IQGAP1-null MEFs. Insulin-stimulated association of the p85α subunit of PI3K with IRS-1 was reduced by 35% when IQGAP1 was not present (Fig. 5, G and H). This impaired recruitment of p85α may reduce synthesis of PIP₃ and contribute to decreased activation of Akt. Congruent with this concept, the ability of insulin to activate Akt in MEFs was significantly reduced in the absence of IQGAP1. Re-expression of full-length IQGAP1 in IQGAP1-null MEFs increased Akt activation, whereas re-expression of the N-terminal half (that binds IR and Akt) or the C-terminal half (that binds IRS-1) had no effect on Akt. These rescue experiments indicate that both the N- and C-halves of IQGAP1 are required for insulin-stimulated Akt signaling. Potentially, IQGAP1 may function by scaffolding several insulin signaling components, such as IR to IRS-1 and/or IRS-1 to p85α and Akt.

In MEFs, insulin activates both IR and IGF-1R to promote phosphorylation of Akt and ERK (36). The contributions of IQGAP1 to insulin-stimulated IGF-1R signaling have yet to be investigated, but they may contribute to the observations in MEFs. More specific signaling from the IR was observed in IQGAP1-knock-out mice, which also displayed impaired insulin-stimulated Akt activation in classic insulin-responsive tissues (liver, skeletal muscle, and adipose tissue). These results in tissues demonstrate that IQGAP1 regulates insulin signaling in cells and tissues.

The impaired insulin-stimulated Akt activation in IQGAP1-null mice is relevant, because defects in Akt activation have been correlated with impaired glucose transport in several mouse models (37, 38). Therefore, it is possible that IQGAP1 is required for optimal glucose transport. This concept is supported by our data in mice. In an in vivo model of glucose homeostasis, IQGAP1-null mice had an exaggerated response to a glucose load, and their ability to restore euglycemia was impaired when compared with wild type littermate controls. Consistent with these findings, insulin had an impaired ability to lower blood glucose concentrations in mice devoid of IQGAP1. These data are also congruent with the impaired insulin signaling observed in tissues, where altered Akt activation may affect glucose uptake. Collectively, these results reveal that mice lacking IQGAP1 are insulin-resistant.

Humans and mice have three IQGAP family members, namely IQGAP1, IQGAP2, and IQGAP3 (16). Of these, IQGAP1 is by far the best characterized. Although less is known about IQGAP2, studies indicate some opposite functions for IQGAP1 and IQGAP2. Published evidence suggests that IQGAP1 is an oncogene, whereas IQGAP2 is a tumor suppressor (for a review, see Ref. 39). For example, loss of IQGAP1 reduces breast cancer cell proliferation and tumorigenicity in mice (40), whereas IQGAP2-null mice develop hepatocellular carcinoma (41). In keeping with these opposite functions, mice lacking IQGAP2 were shown to have increased insulin sensitivity, with enhanced glucose clearance in response to parenteral glucose administration (42). Insulin stimulation of Akt in livers of IQGAP2-null mice was greater than that in control mice, leading the authors to conclude that lack of IQGAP2 leads to increased hepatic insulin sensitivity. We observed that IQGAP1-null mice had impaired insulin-stimulated Akt activation in liver. These observations suggest that IQGAP1 and IQGAP2 have opposing functions in the regulation of insulin response in the liver. While IQGAP2 is expressed predominantly in the liver, IQGAP1 is ubiquitously expressed (41). In our study, IQGAP1 also affected insulin signaling in adipose tissue and skeletal muscle, suggesting roles for IQGAP1 regulation of signaling in multiple tissues. In addition to insulin, several other hormones (e.g. glucagon and cortisol) regulate glucose homeostasis. Therefore, studies are needed to investigate whether the impaired glucose homeostasis in IQGAP1-null mice is exclusively due to impaired insulin signaling or whether disruption of other regulatory mechanisms contributes.

IQGAP1 has been shown to be a component of signaling induced by growth factors other than insulin. For example, activation of the MAPK cascade by EGF and IGF-1 is decreased in cells with reduced IQGAP1 (22, 23). Similarly, cells deficient in IQGAP1 displayed both impaired Akt and ERK phosphorylation downstream of HER2 (20) and decreased activation of Akt by VEGF (21). Moreover, Akt and ERK phosphorylation in response to cardiac pressure overload is decreased in IQGAP1-null mice (43).

The multidomain composition of IQGAP1 enables it to assemble a variety of protein complexes (14). Here, IQGAP1 is shown to interact with both IR and IRS-1. Our data (Fig. 4, A and D) and prior publications (31, 44, 45) reveal that IR and IRS-1 associate directly in vitro. The precise physiological role of IQGAP1 in the coupling of these molecules in insulin signaling is not clear. Cells lacking IQGAP1 exhibit no significant change in IR phosphorylation and only a moderate decrease in IRS-1 phosphorylation in response to insulin. When IQGAP1 was reduced in CHO/IR/IRS-1 cells, the constitutive association of IR with IRS-1 was increased, and insulin failed to promote the interaction (Fig. 5, E and F). The molecular mechanism underlying these observations remains to be elucidated. Potentially, IQGAP1 may regulate protein-protein interactions or signaling pathways that affect the insulin response.

Although the role of IQGAP1 in the association of IR with IRS-1 is fairly subtle, loss of IQGAP1 leads to clear impairment in IRS-1 association with p85α and considerably reduced activation of Akt and ERK. Modest changes in IRS-1 phosphorylation may contribute to the significantly impaired recruitment of PI3K (Fig. 5). Reduced PI3K association with IRS-1 would reduce PIP₃ synthesis, which would contribute to decreased Akt activation. Thus, small changes in upstream signaling may lead to more dramatic reductions in downstream pathways due to impaired signal amplification.

Additionally, IQGAP1 may modulate insulin signaling by scaffolding components of the Akt (46) and MAPK (22–24) pathways for efficient signaling. Thus, the association of IQGAP1 with IR and IRS-1 may primarily serve as a scaffold to bring IQGAP1 in close proximity to the receptor complex to facilitate signaling of downstream components. Moreover, IQGAP1 may mediate communication between these signaling components for multifaceted regulation of downstream signaling. IQGAP1 does not appear to be a direct target of IR because no tyrosine phosphorylation of IQGAP1 was detected after insulin stimulation (data not shown). Note that loss of IQGAP1 does not abrogate insulin-stimulated phosphorylation of IRS-1, Akt, or ERK or the binding of p85 to IRS-1. Rather, the findings in this study indicate that IQGAP1 is required for optimal activation of at least some insulin-regulated signaling pathways. This hypothesis is strongly supported by the impaired insulin signaling and glucose homeostasis in IQGAP1-null mice. Because impaired glucose tolerance is a hallmark of type 2 diabetes, further elucidation of the molecular mechanisms by which IQGAP1 participates in insulin action may lead to the development of novel therapeutic agents to treat patients with type 2 diabetes.

Experimental Procedures

Materials

All cell culture reagents were obtained from Invitrogen. Glutathione-Sepharose and Protein A-Sepharose beads were purchased from GE Healthcare. Nickel-nitrilotriacetic acid-agarose beads were purchased from Qiagen. PVDF membranes were purchased from Millipore Corp. (Bedford, MA). Anti-IRS-1, anti-IR, anti-phospho-Akt (Ser-473), anti-Akt, anti-phospho-ERK (Thr-202/Tyr-204), anti-ERK, anti-PI3K p85α, and anti-phosphotyrosine antibodies were obtained from Cell Signaling Technology (Danvers, MA). Anti-GFP antibody was purchased from Life Technologies, Inc. Anti-β-tubulin antibody was obtained from Sigma. Anti-IRS-1 antibody 06-248 from Millipore was used for immunoprecipitation. Anti-IR-β and normal rabbit IgG for immunoprecipitation were from Santa Cruz Biotechnology, Inc. The anti-IQGAP1 polyclonal antibody has been characterized previously (47). Blocking buffer and infrared dye-conjugated (IRDye) secondary antibodies were purchased from LI-COR Biosciences (Lincoln, NE). The EGFP (enhanced green fluorescent protein) (pEGFP-C1) and pGEX-2T plasmids were obtained from Clontech and GE Healthcare, respectively. Insulin was obtained from Sigma. Halt protease inhibitor mixture was obtained from Thermo Scientific. DC protein assay reagents were purchased from Bio-Rad. Unless otherwise stated, all other reagents used were of standard analytical grade.

Plasmid Constructs

GST-tagged constructs comprising full-length IQGAP1 (amino acids 2–1657), IQGAP1-N (amino acids 2–863), and IQGAP1-C (amino acids 864–1657) cloned in the pGEX-2T vector have been described previously (47). Myc-tagged IQGAP1 in pcDNA3 (23) and pEGFPC1-IQGAP1 (48) have been described previously. GST-IRS-1 was cloned in two steps using the vector pCMV-IRS-1 (Addgene Plasmid 11238) (49) as the template. First, the IRS-C-terminal portion (amino acids 511–1242) was cut from pCMV-IRS-1 with BglII, and blunt ends were created using T4 DNA polymerase. The plasmid was then inserted into the pGEX2T-TEV vector (35) at an SmaI site, resulting in the construct pGEX2T-TEV-IRS-C. Then, the N-terminal portion, IRS-N (amino acids 2–564), was synthesized by PCR using pCMV-IRS-1 as the template with 5′-CGGGATCCGCGAGCCCTCCGGATACC-3′ as the forward primer and 5′-GTGCTTGCCTTGGGATCC-3′ as the reverse primer. It was then cut with BamHI and cloned into pGEX2T-TEV-IRS-C at a BamHI site.

IRS fragments 2–115, 2–300, and 162–267 were amplified using pCMV-IRS-1 as the template with forward primers starting at region IRS-1 (amino acid 2) 5′-CGGGATCCGCGAGCCCTCCGGATACCGATGGC-3′, and (amino acid 162) 5′-CGGGATCCATCCTGAAACCCAAGGGCTTAGG-3′, and the following reverse primers were used: (amino acid 115) 5′-CCGGAATTCTCTAGACTAATTATGCAGCTGCAGAAGAGCC-3′, (amino acid 267) 5′-CCGGAATTCTCTAGACTAAGATTGGCTTTTCGTGCGCGGG-3′, and (amino acid 300) 5′-CCGGAATTCTCAAGTTCGAGATCTCCGAGTCAG-3′. Then these fragments were cloned into BamHI and XbaI sites of pcDNA3-myc.

GST-IR-β (β-subunit of IR; amino acids 763–1382) and GST-IR_cyt (cytoplasmic domain of IR; amino acids 980–1382) were amplified by PCR using GFP-tagged human insulin receptor (Addgene plasmid 22286) (50) as the template. The forward primers used for IR-β and IR_cyt were 5′-CGGGATCCTCCCTTGGCGATGTTGGGAATGTG-3′ and 5′-CGGGATCCAGAAAGAGGCAGCCAGATGGGCCG-3′, respectively, and5′-CCGGAATTCTCTAGATTAGGAAGGATTGGACCGAGGC-3′ was used as the reverse primer for both constructs. They were then cloned into the BamHI/EcoRI sites of pGEX2T-TEV vector. The sequences of all constructs were confirmed by DNA sequencing. The plasmids were purified with the QIAprep Spin Mini Qiagen Prep Kit (Qiagen).

Preparation of Fusion Proteins

GST-IQGAP1 was expressed in Escherichia coli and isolated using glutathione-Sepharose essentially as described previously (47). Where indicated, the GST tag was cleaved from GST-IQGAP1 using tobacco etch virus protease as described previously (35). GST-IRS-1 was expressed and isolated as described for GST-IQGAP1. Briefly, the recombinant construct pGEX2T-GST-IRS-1 was expressed in the BL21 strain of E. coli, and expression was induced with 0.5 μm isopropyl-1-β-d-thiogalactopyranoside (IPTG) at 25 °C for 8 h. Bacteria were then harvested by centrifugation at 1000 × g for 20 min, and the cell pellet was resuspended in buffer (1× PBS, pH 7.4, 2 mm EDTA, 10 mm DTT, 1% Triton X-100, and 1 mm PMSF). The cell suspension was then sonicated, and the lysate was centrifuged at 16,000 × g for 30 min at 4 °C. The resulting supernatant was loaded onto glutathione-Sepharose beads, and the recombinant protein was isolated.

The recombinant proteins, GST-IR-β and GST-IR_cyt, were expressed in the BL21 strain of E. coli and induced with 100 μm IPTG for 16 h at 16 °C. Bacteria were harvested by centrifugation at 1000 × g for 20 min, and the cell pellet was resuspended in buffer (50 mm Tris-HCl, pH 7.5, 150 mm NaCl, 0.1 mm EGTA, 270 mm sucrose, 0.1% 2-mercaptoethanol, and 1 mm PMSF). The cell suspension was then sonicated, and the lysate was centrifuged at 20,000 × g for 30 min at 4 °C. The resulting supernatant was loaded onto glutathione-Sepharose beads, and the recombinant proteins were isolated. The purity of all fusion proteins as evaluated by Coomassie staining of SDS-PAGE gels was ≥90%.

His₆-IR-β and His₆-IR_cyt were expressed in BL21 DE3 strain of E. coli and induced with 1 mm IPTG. Bacteria were then harvested by centrifugation at 1000 × g for 30 min, and the cell pellet was resuspended in buffer (50 mm NaH₂PO₄, pH 8.0, 300 mm NaCl, and 10 mm imidazole). Lysozyme was added to the cell suspension to a final concentration of 1 mg/ml and incubated for 30 min at 4 °C. The cell suspension was sonicated, followed by centrifugation at 20,000 × g for 30 min at 4 °C. The resulting supernatant was loaded onto nickel-nitrilotriacetic acid-agarose beads, and the recombinant proteins were isolated.

Cell Culture and Transfection

C2C12, 3T3-L1, and MCF-7 cells were obtained from the American Type Culture Collection (Manassas, VA). HEK-293H cells were obtained from Invitrogen. CHO/IR cells (which stably overexpress IR) and CHO/IR/IRS-1 cells (which stably overexpress both IR and IRS-1) were kindly provided by Xiao Sun (University of Maryland). Mouse embryonic fibroblast (MEF) cells derived from IQGAP1-null and normal littermate controls have been described previously (24). 3T3-L1, HEK-293H, MCF-7, and MEF cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (v/v) fetal bovine serum and 1% penicillin/streptomycin solution (Life Technologies). CHO/IR and CHO/IR/IRS-1 cells were cultured in Ham's/F-12 medium (Life Technologies) supplemented with 10% (v/v) fetal bovine serum and 1% penicillin/streptomycin solution.

The pEGFP-C1 constructs, full-length GFP-IQGAP1, N-terminal GFP-IQGAP1, C-terminal GFP-IQGAP1, and p-EGFP-C1 (GFP) were transfected into HEK-293H cells using Lipofectamine 2000 (Invitrogen) transfection reagent according to the manufacturer's instructions. The pcDNA3-Myc-IQGAP1 or GFP-IQGAP1 construct was transfected into CHO/IR/IRS-1 cells using Lipofectamine LTX transfection reagent according to the manufacturer's instructions. Wild type or IQGAP1^−/− MEFs were transfected with Myc-tagged full-length IQGAP1, Myc-N-terminal IQGAP1, pcDNA-Myc empty vector, pEGFP-C1 empty vector, or GFP-C-terminal IQGAP1 using TransIT-2020 (Mirus) as described (51).

In Vitro Binding Assays

Purified untagged IQGAP1 was precleared with glutathione beads for 1 h at 4 °C, and then IQGAP1 was incubated with GST constructs of IR-β (amino acids 763–1382) or IR_cyt (amino acids 980–1382) in 500 μl of buffer A (50 mm Tris-HCl, pH 7.4, 150 mm NaCl, and 1% (v/v) Triton X-100) with 1 mm PMSF and protease inhibitor mixture (Buffer B) for 3 h at 4 °C. GST alone was used as the negative control. After washing the beads five times with buffer A, samples were resolved by SDS-PAGE. The gel was cut at the 150-kDa region. The top part was transferred to PVDF, blocked with blocking buffer (LI-COR Biosciences) for 1 h at 22 °C, and then probed with anti-IQGAP1 polyclonal antibodies overnight at 4 °C. The membrane was incubated with IRDye-conjugated anti-rabbit antibody for 1 h, and antigen-antibody complexes were detected using the Odyssey Imaging System (LI-COR). The lower portion of the gel containing GST-IR-β, GST-IR_cyt, and GST was stained with Coomassie Blue.

GST-IRS-1 or GST alone was incubated with purified His₆-IR_cyt alone or with both His₆-IR_cyt and purified IQGAP1 in 1 ml of buffer A for 3 h at 4 °C. Complexes were isolated using glutathione-Sepharose, and the samples were washed and resolved by SDS-PAGE. After electrophoresis, the gel was cut at 150 kDa and transferred to PVDF membrane. The top portion was probed with anti-IQGAP1 antibody and incubated with IRDye-conjugated anti-rabbit antibody. The blot was subsequently stripped by incubating with stripping buffer (62.5 mm Tris-HCl (pH 6.8), 2% (w/v) SDS, and 0.7% (v/v) β-mercaptoethanol) for 30 min at 50 °C and reprobed with anti-IRS-1 antibody. The bottom portion of the PVDF membrane was probed with anti-IR-β and anti-GST antibodies.

For the protein binding assays depicted in Fig. 4A, GST or GST-IRS-1 (40 nm) bound to glutathione-Sepharose was preincubated with a molar excess of IR_cyt or IQGAP1 for 1 h at 4 °C. Beads were washed, and IR_cyt or IQGAP1 was added to the samples as indicated in the legend to Fig. 4A. After 1 h, samples were washed and analyzed by SDS-PAGE and immunoblotting to evaluate binding of IQGAP1 and IR_cyt to IRS-1.

GST Pull-down Assays

Glutathione-Sepharose pull-down assays were performed essentially as described previously (52). Briefly, cells were grown to 90% confluence, washed with ice-cold PBS, and lysed in 500 μl of buffer B. The lysates were sonicated, and the insoluble pellet was centrifuged at 16,000 × g for 10 min at 4 °C. The resulting supernatants were precleared by incubating with glutathione-Sepharose beads for 1 h at 4 °C. Clarified cell lysates were equalized for protein concentration using the DC protein assay, and equal amounts of protein were incubated with GST, GST-IQGAP1, or GST-IRS-1 for 3 h at 4 °C. After centrifugation, samples were washed five times with buffer A and resolved by SDS-PAGE. The gel was cut into two portions. One portion was stained with Coomassie Blue, whereas the other portion was transferred to PVDF and processed by Western blotting. For GST-IQGAP1 pull-down, the top portion containing GST-IQGAP1 was stained with Coomassie Blue, and the bottom portion was transferred to PVDF and processed by Western blotting. The membranes were blocked with blocking buffer for 1 h at 22 °C and then probed with anti-IRS-1 or anti-IR antibodies overnight at 4 °C. Then the membranes were incubated with IRDye-conjugated anti-rabbit antibody for 1 h, and antigen-antibody complexes were detected using the Odyssey Imaging System (LI-COR). In Figs. 2A and 3E, the blot was incubated with the antibodies indicated and developed by enhanced chemiluminescence (ECL).

Immunoprecipitation

Immunoprecipitation was performed essentially as described previously (22). Briefly, cells were plated in 10-cm dishes to obtain 80% confluence. The following day, they were transfected with Myc-tagged IQGAP1 and allowed to grow for 48 h. The cells were then washed with ice-cold PBS and treated with 1 mm DSP (Pierce) for 30 min at 22 °C. The reaction was stopped by the addition of Tris, pH 7.4, to a final concentration of 50 mm and incubated for 10 min. The cells were washed again with PBS and lysed in buffer A. For immunoprecipitation of endogenous IR and IRS-1 with IQGAP1, cells were lysed and washed with a buffer containing 100 mm Tris, pH 7.4, 5 mm EDTA, 1% Triton X-100, and protease and phosphatase inhibitors. Lysates were subjected to sonication for 5 s, and the insoluble fraction was pelleted by centrifugation at 16,000 × g at 4 °C. Supernatants were precleared with glutathione-Sepharose beads for 1 h at 4 °C. Clarified cell lysates were equalized for protein concentration using the DC protein assay, and equal amounts of protein were incubated with anti-IQGAP1 polyclonal antibodies for 3 h at 4 °C. Samples were processed in parallel with NIRS as the negative control. Immune complexes were isolated using protein A-Sepharose beads, washed five times in buffer A, resolved by SDS-PAGE, and processed by Western blotting.

For immunoprecipitation of IR from MEFs, control and IQGAP1-null cells were grown to confluence and serum-starved overnight. Following stimulation with 100 nm insulin for 0, 2, and 15 min, cells were washed with PBS, lysed in 500 μl of buffer B, sonicated, and centrifuged at 16,000 × g at 4 °C. Lysates were precleared with glutathione-Sepharose for 1 h. Equal concentrations of protein lysate were incubated with 0.5 μg of anti-IR-β or normal rabbit IgG for 1 h, protein A-Sepharose was added for 4 h, and beads were washed five times with buffer A. Immunoprecipitates were eluted in sample buffer. Lysate samples were taken before the addition of antibody. Samples were examined by SDS-PAGE, followed by probing Western blots with anti-phosphotyrosine and anti-IR-β antibodies.

TNT Product Production and Binding Analysis

[³⁵S]Methionine-labeled and biotinylated lysine-labeled TNT (transcription and translation) products were synthesized with the TNT Quick Coupled Transcription/Translation system (Promega) according to the manufacturer's instructions. Briefly, 1 μg of pcDNA3-IQGAP1, IQGAP1-N, IQGAP1-C, IQGAP1 (785–1307), IQGAP1(864–1357), IQGAP1(914–1657), IQGAP1 (950–1407), IQGAP1(950–1457), or IQGAP1(1010–1457) was incubated with 40 μl of TNT Quick Master Mix (Promega) and 20 μCi of [³⁵S]methionine (PerkinElmer Life Sciences) at 30 °C for 90 min. Alternatively, 1 μg of pcDNA3-IQGAP1-F, IQGAP1-N, IQGAP1-C, IQGAP1(2–764), or IQGAP1(717–916) was incubated with 40 μl of TNT Quick Master Mix (Promega) and 2 μl of biotinylated lysine (PerkinElmer Life Sciences) at 30 °C for 90 min. TNT products were diluted in buffer A and used in pull-down assays. Products were confirmed by SDS-PAGE and autoradiography (for proteins labeled with [³⁵S]methionine) or chemiluminescence with HRP-conjugated streptavidin (for proteins labeled with biotinylated lysine).

Measuring Association of IR with IRS-1

CHO/IR/IRS-1 were reverse transfected using Lipofectamine RNAiMax with siRNA targeting IQGAP1 (sc-35701) or control siRNA (sc-37007) from Santa Cruz Biotechnology following the manufacturer's instructions. Cells were serum-starved overnight and 48 h after transfection they were incubated with 100 nm insulin for 1 min. Cells were lysed in a buffer containing 100 mm Tris, pH 7.4, 5 mm EDTA, 1% Triton X-100, and protease and phosphatase inhibitors. IRS-1 was immunoprecipitated with anti-IRS-1 antibody 06-248 from Millipore. Binding of IR to IRS-1 was assessed by Western blotting. Knockdown of IQGAP1 was confirmed with an IQGAP1-specific antibody.

Measurement of IR, IRS-1, Akt, and ERK Phosphorylation

Control and IQGAP1-null MEFs were plated and allowed to attach overnight. The following day, cells were serum-starved for 16 h and then incubated with 100 nm insulin at 37 °C for the times indicated. To examine phosphorylation of IR and IRS-1, MEFs were lysed in 500 μl of buffer B, and IR or IRS-1 was immunoprecipitated as described above. Samples were examined by Western blots probed with anti-phosphotyrosine and anti-IR-β or anti-IRS-1 antibodies. Additionally, MEFs were stimulated with or without insulin for 5 min, and IRS-1 was immunoprecipitated. Blots were probed for the association of PI3K p85α subunit. For Akt and ERK phosphorylation, cells were washed once with ice-cold PBS and lysed in 500 μl of buffer B. The lysates were cleared by centrifugation at 16,000 × g for 10 min. Clarified cell lysates were equalized for protein concentration using the DC protein assay, resolved by SDS-PAGE, and processed by Western blotting. Phosphorylation of Akt was measured by probing blots with anti-phospho-Akt antibody. In all experiments, blots were stripped as described above and then reprobed with an antibody against total Akt. ERK and phospho-ERK blots were probed with specific antibodies to measure ERK activation. All blots were also probed with anti-β-tubulin antibody to assess protein loading. For rescue studies, 48 h after transfection, MEFs were starved of serum for 6 h and incubated with or without 100 nm insulin for 10 min, and cells were processed as described above.

Animals

IQGAP1^−/− (53) and wild type (littermate control) mice were bred in the animal facility at the National Institutes of Health (NIH) and maintained according to NIH guidelines. The studies were performed with approval of the NIH Animal Care and Use Committee. The high fat diet, which contained 60% fat calories, was obtained from Bio-Serv.

Insulin Signaling in Tissues

Wild type or IQGAP1^−/− male mice were fasted for 4 h, followed by an intraperitoneal injection of saline or insulin (Humulin-R, 0.5 units/ml, 0.01 ml/g body mass). 15 min postinjection, mice were euthanized, and the liver, quadriceps muscle, and epididymal fat pads were harvested and frozen immediately. Tissues were homogenized with CKMIX beads (Bertin) using a Precellys24 homogenizer. Debris was removed by centrifugation, and equal amounts of protein were analyzed by SDS-PAGE and immunoblotting.

Glucose Tolerance Tests

Wild type and IQGAP1^−/− mice were matched for age and weight and divided into two groups. Each group comprised 10 mice, 5 control and 5 IQGAP1^−/− mice. One group was kept on a regular chow diet. The other group was placed on a high fat diet for 8 weeks. Intraperitoneal glucose tolerance tests (20% d-glucose; 0.01 ml/g of body weight) were conducted after animals had fasted for 16 h. Blood was obtained from the tail vein 0, 15, 30, 60, 90, and 120 min after glucose injection, and the blood glucose concentrations were measured using a hand-held glucose meter (Sure Step Flexx, LifeScan).

Insulin Tolerance Tests

Wild type and IQGAP1^−/− mice were matched for age and weight. Intraperitoneal insulin tolerance tests (0.1 units/ml Humulin-R; 0.005 ml/g of body weight) were conducted after animals had fasted for 4 h. Blood was obtained from the tail vein at baseline (t = 0), and 30, 60, and 90 min after insulin injection, and blood glucose concentrations were measured using a hand-held glucose meter (Stat Strip, Nova).

Measurement of Serum Insulin

Wild type and IQGAP1^−/− mice were fed or fasted for 16 h. Blood was collected, and serum insulin concentrations were measured using the Ultra Sensitive Mouse Insulin ELISA kit (Crystal Chem) following the manufacturer's instructions.

Statistical Analysis

Statistical analysis for insulin signaling was performed by two-tailed Student's t test with Excel (Microsoft) with statistical significance set at p < 0.05. Western blotting images were quantified with Image Studio (LI-COR) according to the manufacturer's instructions.

Author Contributions

B. C., A. C. H., S. S., Z. L., and D. B. S. designed experiments. B. C., A. C. H., S. S., H. H. E., and Z. L. performed experiments. B. C., A. C. H., S. S., H. H. E., Z. L., and D. B. S. analyzed data. B. C., A. C. H., and D. B. S. wrote the manuscript.