Advertisement

Platelet-derived Growth Factor Inhibits Insulin Stimulation of Insulin Receptor Substrate-1-associated Phosphatidylinositol 3-Kinase in 3T3-L1 Adipocytes without Affecting Glucose Transport*

      Phosphatidylinositol 3-kinase (PI3K) activation is necessary for insulin-responsive glucose transporter (GLUT4) translocation and glucose transport. Insulin and platelet-derived growth factor (PDGF) stimulate PI3K activity in 3T3-L1 adipocytes, but only insulin is capable of stimulating GLUT4 translocation and glucose transport. We found that PDGF causes serine/threonine phosphorylation of insulin receptor substrate 1 (IRS-1) in 3T3-L1 cells, measured by altered mobility on SDS-polyacrylamide gel, and this leads to a decrease in insulin-stimulated tyrosine phosphorylation of IRS-1. The PI3K inhibitors wortmannin and LY294002 inhibit the PDGF-induced phosphorylation of IRS-1, whereas the MEK inhibitor PD98059 was without a major effect. PDGF pretreatment for 60–90 min led to a marked 80–90% reduction in insulin stimulatable phosphotyrosine and IRS-1-associated PI3K activity. We examined the functional consequences of this decrease in IRS-1-associated PI3K activity. Interestingly, insulin stimulation of GLUT4 translocation and glucose transport was unaffected by 60–90 min of PDGF preincubation. Furthermore, insulin activation of Akt and p70s6kinase, kinases downstream of PI3K, was unaffected by PDGF pretreatment. Wortmannin was capable of blocking these insulin actions following PDGF pretreatment, suggesting that PI3K was still necessary for these effects. In conclusion, 1) PDGF causes serine/threonine phosphorylation of IRS-1, and PI3K, or a kinase downstream of PI3K, mediates this phosphorylation. 2) This PDGF-induced phosphorylation of IRS-1 leads to a significant decrease in insulin-stimulated PI3K activity. 3) PDGF has no effect on insulin stimulation of Akt, p70s6kinase, GLUT4 translocation, or glucose transport. 4) This suggests the existence of an IRS-1-independent pathway leading to the activation of PI3K, Akt, and p70s6kinase; GLUT4 translocation; and glucose transport.
      EGF
      epidermal growth factor
      IRS
      insulin receptor substrate
      PI3K
      phosphatidylinositol 3-kinase
      PDGF
      platelet-derived growth factor
      MAPK
      microtubule-associated protein kinase
      GLUT4
      insulin-responsive glucose transporter
      TNFα
      tumor necrosis factor α
      pY
      phosphotyrosine
      PAGE
      polyacrylamide gel electrophoresis
      PIP2
      phosphatidylinositol 3,4-bisphosphate
      PIP3
      phosphatidylinositol 3,4,5-trisphosphate
      ATP
      adenosine triphosphate.
      Tyrosine kinase receptors such as the insulin receptor, the epidermal growth factor (EGF)1 receptor, and the platelet-derived growth factor (PDGF) receptor activate many of the same signaling cascades. The two most prominent shared pathways are the phosphatidylinositol 3-kinase (PI3K) and the microtubule-associated protein kinase (MAPK) pathways. Activation of MAPK leads to mitogenic progression and may be involved in differentiation (for review, see Ref.
      • Seger R.
      • Krebs E.G.
      ). Numerous studies using inhibitors of PI3K activity, expression of constitutively active PI3K, and microinjection of dominant negative inhibitors of PI3K have all demonstrated that PI3K is necessary for the mitogenic effects of many growth factors and is both necessary and sufficient for the metabolic effects of insulin (
      • Conricode K.M.
      ,
      • Margalet-Sanchez V.
      • Goldfine I.D.
      • Vlahos C.J.
      • Sung C.K.
      ,
      • Okada T.
      • Kawano Y.
      • Sakakibara T.
      • Hazeki O.
      • Ui M.
      ,
      • Cheatham B.
      • Vlahos C.J.
      • Cheatham L.
      • Wang L.
      • Blenis J.
      • Kahn C.R.
      ,
      • Berger J.
      • Hayes N.
      • Szalkowksi D.M.
      • Zhang B.
      ,
      • Clarke J.F.
      • Young P.W.
      • Yonezawa K.
      • Kasuga M.
      • Holman G.D.
      ).
      PI3K is composed of two subunits, an 85-kDa regulatory subunit (p85) and a 110-kDa catalytic subunit (p110). The p85 subunit is composed of an N-terminal Src-homology 3 (SH3) domain and 2 Src-homology 2 (SH2) domains. The SH2 domains flank the region where the p110 associates with p85. The SH2 domains interact with phosphotyrosine residues, leading to subsequent activation of the p110 catalytic subunit (for review, see Ref.
      • Kapeller R.
      • Cantley L.C.
      ). Insulin stimulation leads to the tyrosine phosphorylation of insulin receptor substrates 1 and 2 (IRS-1/2) and the association of PI3K with the SH2 binding sites contained within these proteins. PDGF and EGF activate PI3K through its association with phosphotyrosines located in the C terminus of their receptors.
      PI3K exhibits both lipid and protein kinase activities. PI3K phosphorylates the D3 position of the inositol ring of phosphatidylinositols, causing the production of phosphatidylinositol 3,4-bisphosphate (PIP2), and phosphatidylinositol 3,4,5-trisphosphate (PIP3) (
      • Kapeller R.
      • Cantley L.C.
      ). The role of PIP2 and PIP3, the primary products of PI3K, has been elusive. Recently, a role for PIP2 and PIP3 in the activation of an Akt kinase has been discovered (
      • Stokoe D.
      • Stephens L.R.
      • Copeland T.
      • Gaffney P.R.J.
      • Reese C.B.
      • Painter G.F.
      • Holmes A.B.
      • McCormick F.
      • Hawkins P.T.
      ). PI3K also phosphorylates proteins on serine/threonine residues (
      • Dhand R.
      • Hiles I.
      • Panayotou G.
      • Roche S.
      • Fry M.J.
      • Gout I.
      • Totty N.F.
      • Truong O.
      • Vicendo P.
      • Yonezawa K.
      • Kasuga M.
      • Courtneidge S.A.
      • Waterfield M.D.
      ,
      • Lam K.
      • Carpenter C.L.
      • Ruderman N.B.
      • Friel J.C.
      • Kelly K.L.
      ,
      • Tanti J.-F.
      • Gremeaux T.
      • Van Obberghen E.
      • Le Marchand-Brustel Y.
      ). The insulin receptor substrate-1 (IRS-1) is an in vitro substrate for this activity (
      • Lam K.
      • Carpenter C.L.
      • Ruderman N.B.
      • Friel J.C.
      • Kelly K.L.
      ,
      • Tanti J.-F.
      • Gremeaux T.
      • Van Obberghen E.
      • Le Marchand-Brustel Y.
      ).
      3T3-L1 adipocytes provide a means to study the metabolic effects of insulin in a cell culture system. These cells express insulin, PDGF, and EGF receptors. Each of these ligands is capable of activating MAPK in these cells, but only insulin and PDGF activate PI3K (
      • Wiese R.J.
      • Mastik C.C.
      • Lazar D.F.
      • Saltiel A.R.
      ). Despite the disparity in insulin and PDGF receptor number in these cells, PDGF and insulin activate PI3K to a similar extent. Because the activation of PI3K has been shown to be necessary for the metabolic effects of insulin, such as the translocation of the insulin-stimulated glucose transporter, GLUT4, to the cell membrane, glucose uptake, and glycogen synthesis, it has been puzzling as to how insulin, but not PDGF, causes translocation of GLUT4 and glucose transport when both of these ligands cause stimulation of PI3K (
      • Conricode K.M.
      ,
      • Margalet-Sanchez V.
      • Goldfine I.D.
      • Vlahos C.J.
      • Sung C.K.
      ,
      • Okada T.
      • Kawano Y.
      • Sakakibara T.
      • Hazeki O.
      • Ui M.
      ,
      • Cheatham B.
      • Vlahos C.J.
      • Cheatham L.
      • Wang L.
      • Blenis J.
      • Kahn C.R.
      ,
      • Berger J.
      • Hayes N.
      • Szalkowksi D.M.
      • Zhang B.
      ,
      • Clarke J.F.
      • Young P.W.
      • Yonezawa K.
      • Kasuga M.
      • Holman G.D.
      ).
      In this study, we investigated the interplay between insulin and PDGF signaling cascades in 3T3-L1 adipocytes. Insulin and PDGF lead to different time courses of activation of PI3K, with insulin causing a sustained activation of PI3K and PDGF generating transient effects. PDGF pretreatment stimulates the serine/threonine phosphorylation of IRS-1, interfering with the ability of insulin to activate PI3K. Serine phosphorylation of serine 612, one of the major serine phosphorylation sites in IRS-1, may inhibit tyrosine phosphorylation of tyrosine 608, a site of IRS-1/PI3K interaction (
      • De Fea K.
      • Roth R.A.
      ). Importantly, this inhibition of IRS-1-associated PI3K activity had no effect on insulin induction of glucose transport, GLUT4 translocation, or Akt and p70s6kinase phosphorylation. These biologic effects of insulin remain PI3K-dependent, however, because they are blocked by PI3K inhibitors. These results indicate that PI3K can be activated under the direction of insulin by an IRS-1/IRS-2-independent pathway. It appears that whereas PI3K activity is necessary for the metabolic effects of insulin, IRS-1 phosphorylation is not essential.

      EXPERIMENTAL PROCEDURES

       Materials

      Dulbecco's modified Eagle's medium, high glucose, Glutamax, and PDGF-BB were obtained from Life Technologies, Inc. Penicillin-streptomycin and fetal calf serum were obtained from Omega Scientific (Tarzana, CA). Insulin was a gift from Eli Lilly (Indianapolis, IN). Nitrocellulose was obtained from Schleicher & Schuell. Silica-coated thin-layer chromatography plates and all chemicals, unless otherwise noted, were obtained from Sigma. LY294002 and PD98059 were obtained from Calbiochem. [γ-32P]ATP was obtained from ICN. 2-[3H]Deoxyglucose andl-[3H]glucose were obtained from NEN Life Science Products. Akt antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Phospho-Akt and phospho-p70s6kinase were obtained from New England Biolabs (Beverly, MA). IRS-1 and p85α-N-SH3 antibodies and recombinant protein A-agarose were from Upstate Biotechnology (Lake Placid, NY). pY20 antibody and RC-20 were from Transduction Laboratories (Lexington, KY). The GLUT4 antibody, 1F8, was obtained from East Acres Biologicals (Southbridge, MA). Horseradish peroxidase-conjugated secondary antibodies were purchased from Amersham Pharmacia Biotech. Enhanced chemiluminescence reagent was obtained from Pierce.

       Cell Culture

      3T3-L1 fibroblasts were maintained in Dulbecco's modified Eagle's medium, high glucose, containing 10% calf serum. Postconfluency fibroblasts were differentiated into adipocytes by changing the medium with Dulbecco's modified Eagle's medium, high glucose, containing 10% fetal calf serum, 1 μg/ml insulin, 0.1 μg/ml dexamethasone, and 112 μg/ml isobutylmethylxanthine. The medium was removed after 2 days and replaced with Dulbecco's modified Eagle's medium, high glucose, containing 10% fetal calf serum, Glutamax, and 1% penicillin-streptomycin. Seven days after the addition of the differentiation mix, the cells were plated in 6- or 12-well dishes at densities of 8 × 105 and 4 × 105, respectively. The medium was changed every 2–3 days until use, 12–15 days postdifferentiation. Approximately 80–90% of the cells exhibited large lipid droplets indicative of adipocytes.

       Measurement of PI3K Activity

      3T3-L1 adipocytes were stimulated with 100 ng/ml insulin or 50 ng/ml PDGF or pretreated with PDGF for various times and then stimulated with insulin for 5 min. Cells were lysed at 4 °C in a lysis buffer containing 50 mm, 150 mm NaCl, 1% Triton X-100, 4 mm sodium orthovanadate, 20 mm sodium pyrophosphate, 200 mm sodium fluoride, 10 mmEDTA, 2 mm phenylmethylsulfoxide, and 10% glycerol, pH 7.4. Lysates were centrifuged at 14,000 × g for 10 min at 4 °C. Supernatants were incubated with pY20 antibody and anti-mouse IgG agarose or IRS-1 antibody and recombinant protein A-agarose overnight at 4 °C. Beads were pelleted by a 10-min spin at 14,000 × g. Pellets were washed three times with Buffer A (Tris-buffered saline, pH 7.5, 1% Nonidet P-40, and 100 μm Na3VO4), three times with Buffer B (100 mm Tris, pH 7.5, 500 mmLiCl2, and 100 μmNa3VO4), and twice with Buffer C (10 mm Tris, pH 7.5, 100 mm NaCl, 1 mmEDTA, and 100 μm Na3VO4). Pellets were resuspended in Buffer C without the Na3VO4. As described previously (
      • Ruderman N.B.
      • Kapeller R.
      • White M.F.
      • Cantley L.C.
      ), PI3K activity was assessed by the phosphorylation of phosphatidylinositol in the presence of 20 μCi of [γ-32P]ATP. Following the separation of lipids by thin-layer chromatography using the borate-buffered system (
      • Walsh J.P.
      • Caldwell K.K.
      • Majerus P.W.
      ), phosphatidylinositol phosphate was visualized by autoradiography. National Institutes of Health Image software was used to analyze the autoradiographs.

       Assessment of IRS-1 Tyrosine and Serine/Threonine Phosphorylation and Association with PI3K

      Adipocytes plated in 6-well dishes were stimulated at 37 °C with insulin, PDGF, or PDGF followed by insulin for the times indicated. Some wells were treated with wortmannin (100 nm), LY294002 (20 μm), or PD98059 (30 μm) for 20 min prior to the addition of PDGF. Cells were lysed in the lysis buffer mentioned above and spun for 10 min. Antibodies to IRS-1, pY20, or p85 were added to supernatants overnight at 4 °C. Recombinant protein A-agarose or anti-mouse IgG agarose was added for 2–4 h at 4 °C to precipitate IRS-1 complex or p85 and pY20 protein complexes, respectively. Pellets were washed three times in lysis buffer. Laemmli's buffer was added to the pellets and boiled for 5 min. Samples were separated by SDS-PAGE on 5 or 7.5% polyacrylamide gels. Proteins were transferred to nitrocellulose and blotted with IRS-1, pY20, or RC-20 antibodies according to the manufacturer's instructions. Following incubation with horseradish peroxidase-conjugated secondary antibodies (except in the case of RC-20), proteins were visualized by enhanced chemiluminescence. Serine/threonine phosphorylation of IRS-1 was determined by a reduction in the electrophoretic mobility of IRS-1.

       Assessment of GLUT4 Translocation and Glucose Uptake

      3T3-L1 adipocytes, 12 days postdifferentiation, were preincubated with or without PDGF for 90 min followed by 10 min of insulin stimulation. The translocation assay was performed as described previously (
      • Haruta T.
      • Morris A.J.
      • Rose D.W.
      • Nelson J.G.
      • Mueckler M.
      • Olefsky J.M.
      ). Cells were washed with ice-cold phosphate-buffered saline, fixed in 3.7% formaldehyde, and incubated with anti-GLUT4 (1F8) antibody overnight at 4 °C. After rinsing in phosphate-buffered saline, a second incubation was performed with fluorescein isothiocyanate-conjugated anti-mouse IgG antibody for 30 min. Cells positive for GLUT4 translocation show an increase in plasma membrane-associated fluorescent staining. The percentage of cells positive for GLUT4 translocation was determined by analyzing 150–200 cells. For glucose transport, 3T3-L1 adipocytes, 12 or 15 days postdifferentiation, were washed four times with KRP-Hepes containing 1% bovine serum albumin and 2 mm pyruvic acid, pH 7.4. Cells were then stimulated at 37 °C with 1, 10, or 100 ng/ml insulin for 10 min or with 10 or 50 ng/ml PDGF for 60 or 90 min or pretreated with PDGF for various times followed by 10 min of insulin stimulation. Following stimulation, cells were removed from the incubator, and the buffer was replaced with room temperature buffer. Glucose uptake was assessed by the addition of 0.1 mm 2-deoxyglucose containing 0.2 μCi of 2-[3H]deoxyglucose as described previously (
      • Klip A.
      • Li G.
      • Logan W.J.
      ). Nonspecific uptake was assessed by the addition of 0.1 mml-glucose containing 0.2 μCi ofl-[3H]glucose. The reaction was stopped by aspiration and extraneous glucose was removed by four washes with ice-cold phosphate-buffered saline. Cells were lysed in 1 nNaOH, and uptake was assessed by scintillation counting. Samples were normalized for protein content.

       Assessment of Akt and p70s6kinaseActivation

      Adipocytes were stimulated at 37 °C with 100 ng/ml insulin for 10 min or with 50 ng/ml PDGF for 10, 60, or 90 min, or they were pretreated with 50 ng/ml PDGF for 60 or 90 min followed by 100 ng/ml insulin for 10 min. Cells were or were not pretreated with wortmannin (100 nm) and LY294002 (20 μm) for 20 min prior to the addition of PDGF. Cells were lysed in the buffer mentioned above and spun for 10 min at 14,000 × g at 4 °C. Equal fractions of the total lysates were mixed 1:1 with 2× Laemmli's buffer, boiled for 5 min, and loaded on a 7.5% polyacrylamide gel. Proteins were separated by SDS-PAGE and transferred to nitrocellulose membranes. Membranes were blocked in 5% milk overnight at 4 °C. Following blocking, membranes were incubated with antibodies to Akt, phospho-Akt, or phospho-p70s6kinase for 2–4 h at room temperature, and a blotting procedure was followed according to the specifications of the manufacturer. Membranes were incubated with horseradish peroxidase-conjugated secondary antibodies and proteins were visualized by enhanced chemiluminescence.

      RESULTS

       Insulin and PDGF Stimulation of PI3K in 3T3-L1 Adipocytes

      Activation of PI3K by the tyrosine kinase receptors for insulin, PDGF, and EGF has been well documented. However, in 3T3-L1 adipocytes, only insulin and PDGF can activate PI3K (
      • Wiese R.J.
      • Mastik C.C.
      • Lazar D.F.
      • Saltiel A.R.
      ). Both activated PI3K rapidly, within 1–5 min following stimulation (Fig. 1). However, the time course of activation differed for these two growth factors. The insulin-induced PI3K activity precipitable by either anti-phosphotyrosine or anti-IRS-1 antibodies was maintained for up to 3 h (Fig. 1, A and B). As assessed in anti-phosphotyrosine precipitates, PDGF-induced PI3K activity peaked at approximately 1–2 min, sharply decreases by 30 min and was undetectable 60 min after stimulation (Fig. 1 C). PDGF-stimulated PI3K activity in IRS-1 antibody precipitates was negligible (data not shown). The ability of insulin to maintain a sustained activation of PI3K may contribute to insulin-specific biological effects.
      Figure thumbnail gr1
      Figure 1Time course of insulin- and PDGF-stimulated PI3K activation. 3T3-L1 adipocytes (12 days postdifferentiation) were stimulated with 100 ng/ml insulin (A and B) or 50 ng/ml PDGF (C) for the times indicated. PI3K activity was precipitated with anti-phosphotyrosine antibody (A and C) or anti-IRS-1 antibody (B). PI3K activity in the immunoprecipitates was assessed by incorporation of 32P into phosphatidylinositol and separation of phospholipids by thin-layer chromatography as described under “Experimental Procedures.” The position of the PI3K product is indicated (PIP). Representative autoradiographs are shown. Quantitative results, obtained using the NIH Image program, for B and Care shown to the right of the respective autoradiograph and represents fold increase in PI3K activity over basal activity.

       The Effect of PDGF on Insulin-stimulated PI3K Activity and Association with IRS-1

      We wanted to determine whether the down-regulation of PI3K activity seen following 60 min of PDGF treatment could inhibit the ability of insulin to stimulate PI3K activity. We treated 3T3-L1 adipocytes with 50 ng/ml PDGF for various times followed by 5 min of 100 ng/ml insulin stimulation. PI3K activity was immunoprecipitated with phosphotyrosine or IRS-1 antibodies. Insulin-stimulated, phosphotyrosine-precipitated, PI3K activity was dramatically inhibited by preincubation with PDGF and was reduced to basal levels by 60 min of PDGF pretreatment (Fig. 2 A). Insulin-stimulated, IRS-1-precipitated PI3K activity was partially inhibited by 60 min of PDGF pretreatment, but 90 min of PDGF reduced insulin-stimulated PI3K to basal levels (Fig. 2 B).
      Figure thumbnail gr2
      Figure 2Effect of PDGF pretreatment on insulin-stimulated PI3K activity. 3T3-L1 adipocytes (12 days postdifferentiation) were stimulated for the times indicated with 50 ng/ml PDGF, for 5 min with 100 ng/ml insulin, for 5 min with 100 ng/ml EGF, or pretreated with 50 ng/ml PDGF for the times indicated followed by 5 min of insulin stimulation. Cell lysates were immunoprecipitated with anti-phosphotyrosine antibody (A), anti-IRS-1 antibody (B), or anti-p85 antibodies (C).PDGFR, PDGF receptor. A and B, PI3K activity in the immunoprecipitates was assessed by incorporation of32P into phosphatidylinositol and separation of phospholipids by thin-layer chromatography as described under “Experimental Procedures.” The position of phosphatidylinositol 3-phosphate (PIP) is shown. Representative autoradiographs are shown. Quantitative analysis(NIH Image program) is shown to the right of the respective autoradiographs. Quantitation of three experiments is expressed as fold stimulation over basal, unstimulated activity. Columns,means; bars, S.E. C, immunoprecipitated proteins were separated by SDS-PAGE and blotted with anti-phosphotyrosine antibodies.
      To confirm these results, we assessed the ability of PI3K to associate with IRS-1 by co-precipitation and Western blotting. 3T3-L1 cells were stimulated with insulin for 5 min, PDGF for 5 or 90 min, or PDGF for 90 min followed by insulin for 5 min. PI3K was immunoprecipitated from cell lysates with anti-p85 antibodies. Following separation by SDS-PAGE, the immunoprecipitated proteins were blotted with anti-phosphotyrosine antibodies (Fig. 2 C). As shown in Fig. 2 C, insulin stimulation led to the association of PI3K with IRS-1 (second lane from left). PDGF stimulation led to a transient association of PI3K with the PDGF receptor that was no longer apparent following 90 min of PDGF stimulation (third and fourth lanes). Preincubation of the cells for 90 min with PDGF followed by 5 min of insulin decreased the insulin-induced association of PI3K with IRS-1 (right lane).

       The Effect of PDGF on IRS-1 Tyrosine and Serine/Threonine Phosphorylation

      The ability of PDGF to inhibit insulin-stimulated PI3K may be due to inhibition of the insulin signal transduction pathway that leads to the activation of PI3K. As indicated in Fig. 2 C, IRS-1, the major route by which insulin activates PI3K, is a likely target for the PDGF effect. To assess this idea, we stimulated 3T3-L1 adipocytes with PDGF for various times followed by insulin stimulation for 5 min. IRS-1 was immunoprecipitated from lysates with antibodies to the C terminus of IRS-1 (Fig. 3 A, top panel) or to phosphotyrosine residues (Fig. 3 A, bottom panel). Proteins were blotted with a phosphotyrosine antibody. As seen in Fig. 3 A, insulin alone caused tyrosine phosphorylation of IRS-1 and the insulin receptor β subunit (Fig. 3 A, lane 2), whereas PDGF alone led to tyrosine phosphorylation of the PDGF receptor and a 120–130-kDa protein(s) (Fig. 3 A, lane 3). However, with increasing times of PDGF pretreatment, insulin-induced tyrosine phosphorylation of IRS-1 was inhibited, and after 90 min of PDGF pretreatment, IRS-1 phosphorylation was decreased by approximately 60% (Fig. 3 A, lanes 4–9). Interestingly, PDGF-induced autophosphorylation of the PDGF receptor was transient, returning to basal values by 30–45 min. (Fig. 3 A, lanes 6–9).
      Figure thumbnail gr3
      Figure 3Effect of PDGF on IRS-1 tyrosine and serine/threonine phosphorylation and association with PI3K. 3T3-L1 adipocytes (12 days postdifferentiation) were stimulated with or without (–) 100 ng/ml insulin for 5 min, with or without (–) 50 ng/ml PDGF for the times indicated followed with or without (–) 100 ng/ml insulin for 5 min. A, lysates were immunoprecipitated with anti-IRS-1 antibodies (top panel) or anti-phosphotyrosine antibodies (bottom panel). Immunoprecipitated proteins were separated by SDS-PAGE and blotted with anti-phosphotyrosine antibodies.B, prior to PDGF stimulation, cells were incubated with or without (–) 20 μm LY294002, 100 nmwortmannin, or 30 μm PD98059 for 20 min. Following SDS-PAGE, lysates were immunoprecipitated and blotted with anti-IRS-1 antibodies. Representative autoradiographs are shown.
      3T3-L1 adipocytes were stimulated with insulin for 5 min following PDGF stimulation for various times. IRS-1 was precipitated from cell lysates following Western blotting with an IRS-1 antibody. The same antibody was used for blotting. As seen in Fig. 3 B, an IRS-1 gel shift occurred following PDGF stimulation and was evident within 45 min of PDGF treatment (Fig. 3 B, lane 3). Thus, the inhibition of IRS-1 tyrosine phosphorylation was paralleled by a decreased electrophoretic mobility of IRS-1, indicating serine or threonine phosphorylation (
      • De Fea K.
      • Roth R.A.
      ). It is important to note that whereas PDGF inhibited IRS-1 tyrosine phosphorylation (Fig. 3 A), IRS-1 expression was not affected by PDGF pretreatment (Fig. 3 B). PDGF efficiently activated both the PI3K and MAPK pathways in 3T3-L1 adipocytes (
      • Wiese R.J.
      • Mastik C.C.
      • Lazar D.F.
      • Saltiel A.R.
      ). Therefore, we investigated the possibility that PI3K, MAPK, or proteins downstream of these pathways were responsible for the serine/threonine phosphorylation of IRS-1. We used the PI3K inhibitors wortmannin and LY294002 and the MEK inhibitor PD98059 to determine whether either of these proteins is involved. As shown in Fig. 3 B, pretreatment with LY294002 or wortmannin reversed the PDGF-induced IRS-1 gel shift. Pretreatment with PD98059 caused a modest reversal of the IRS-1 gel shift (
      • De Fea K.
      • Roth R.A.
      ), but the effect was not as great as that observed with the PI3K inhibitors (Fig. 3 B, lanes 5–7). These results indicate that PI3K or a protein downstream of PI3K is involved in the serine/threonine phosphorylation of IRS-1 but that MAPK does not play a major role. This is consistent with previous reports that PI3K phosphorylates IRS-1 on serine residues in vitro (
      • Lam K.
      • Carpenter C.L.
      • Ruderman N.B.
      • Friel J.C.
      • Kelly K.L.
      ,
      • Tanti J.-F.
      • Gremeaux T.
      • Van Obberghen E.
      • Le Marchand-Brustel Y.
      ). The serine/threonine phosphorylation and corresponding decrease in tyrosine phosphorylation of IRS-1 may act in concert to prevent the association of PI3K with IRS-1.

       The Effect of PDGF Pretreatment on Insulin-induced Glucose Transport and GLUT4 Translocation

      It has previously been shown that activation of PI3K is a necessary step for insulin-stimulated GLUT4 translocation (
      • Gould G.W.
      • Holman G.D.
      ), and it is known that activated PI3K is sufficient for this biologic effect (
      • Cheatham B.
      • Vlahos C.J.
      • Cheatham L.
      • Wang L.
      • Blenis J.
      • Kahn C.R.
      ,
      • Clarke J.F.
      • Young P.W.
      • Yonezawa K.
      • Kasuga M.
      • Holman G.D.
      ). Therefore, because PDGF pretreatment has a major inhibitory effect on insulin-stimulated PI3K activity, we evaluated the impact of PDGF treatment on insulin-stimulated glucose transport. 3T3-L1 adipocytes were stimulated with insulin or PDGF or were pretreated with PDGF for various times followed by insulin stimulation. As seen in Fig. 4 A, insulin alone caused a severalfold increase in glucose transport, whereas PDGF alone had only a modest effect. Importantly, preincubation of adipocytes with PDGF for up to 60 min did not inhibit insulin-stimulated glucose transport. This is in contrast to the marked effect of PDGF pretreatment to almost completely inhibit insulin-stimulated IRS-1-precipitated or antiphosphotyrosine antibody-precipitated PI3K activity.
      Figure thumbnail gr4
      Figure 4Effect of PDGF on insulin-stimulated 2-deoxyglucose transport. 3T3-L1 adipocytes (12–15 days postdifferentiation) were stimulated with 100 ng/ml insulin for 10 min or 10 ng/ml PDGF for 60 min or were preincubated with 10 ng/ml PDGF for various times followed by 100 ng/ml insulin for 10 min (A) or pretreated for 20 min with or without 100 nm wortmannin and stimulated with or without 10 ng/ml PDGF for 90 min followed by stimulation with 1, 10, or 100 ng/ml insulin for 10 min (B). 2-Deoxyglucose uptake was assessed as described under “Experimental Procedures.” The results from four (A) or five (B) experiments are expressed as fold increase in 2-deoxyglucose uptake over basal, unstimulated cells.Columns, means; bars, S.E.
      In the above experiments, insulin was used at a maximally effective concentration, and it seemed possible that an inhibitory effect of PDGF pretreatment might be elicited at submaximal insulin levels. Thus, we conducted insulin dose-response studies in 3T3-L1 adipocytes, with and without PDGF pretreatment. As demonstrated in Fig. 4 B, a 90-min PDGF pretreatment had no effect on insulin-induced glucose transport throughout the insulin concentration range. The inability of PDGF to inhibit insulin-stimulated glucose transport, despite its marked inhibitory effect on insulin-stimulated PI3K activity, does not mean that PI3K activation is unnecessary for glucose transport stimulation. In fact, the PI3K inhibitor wortmannin was fully effective at inhibiting insulin-stimulated glucose transport with and without PDGF treatment (Fig. 4 B).
      In parallel experiments, we directly measured GLUT4 translocation. Following insulin stimulation, cells were fixed and permeabilized and GLUT4 translocation was visualized with immunofluorescence microscopy using anti-GLUT4 antibodies and fluorescein isothiocyanate-conjugated secondary antibodies (
      • Haruta T.
      • Morris A.J.
      • Rose D.W.
      • Nelson J.G.
      • Mueckler M.
      • Olefsky J.M.
      ,
      • Morris A.J.
      • Martin S.S.
      • Haruta T.
      • Nelson J.G.
      • Vollenweider P.
      • Gustafson T.A.
      • Mueckler M.
      • Rose D.W.
      • Olefsky J.M.
      ). The formation of a fluorescent ring around the cell surface indicates that a cell is positive for GLUT4 translocation, whereas diffuse fluorescent staining is scored as negative (
      • Haruta T.
      • Morris A.J.
      • Rose D.W.
      • Nelson J.G.
      • Mueckler M.
      • Olefsky J.M.
      ,
      • Morris A.J.
      • Martin S.S.
      • Haruta T.
      • Nelson J.G.
      • Vollenweider P.
      • Gustafson T.A.
      • Mueckler M.
      • Rose D.W.
      • Olefsky J.M.
      ). The results of these experiments are summarized in Fig. 5, which shows that PDGF pretreatment did not inhibit the subsequent effect of insulin to stimulate GLUT4 translocation. Furthermore, because PDGF alone does not stimulate GLUT 4 translocation (
      • Ricort J.-M.
      • Tanti J.-F.
      • Van Obberghen E.
      • Le Marchand-Brustel Y.
      ), it is likely that the modest effect of PDGF to increase glucose transport involves stimulation of GLUT1 translocation.
      Figure thumbnail gr5
      Figure 5Effect of PDGF on insulin-stimulated GLUT4 translocation. 3T3-L1 cells were incubated with or without PDGF (10 ng/ml) for 90 min followed by stimulation with 10 or 100 ng/ml insulin for 10 min. GLUT4 translocation was visualized by the presence of a fluorescent ring around the cell, indicating translocation from the cytosol to the plasma membrane. Basal represents unstimulated cells. Results are presented as percentage of cells positive for translocation from two separate experiments.Columns, means; bars, S.E.

       The Effect of PDGF on Insulin-stimulated Akt and p70s6kinase Activation

      It has been suggested that the activation of Akt (protein kinase B), a serine/threonine kinase downstream of PI3K, is necessary for several metabolic effects of insulin, including GLUT4 translocation, glucose uptake, and glycogen synthase activation. Therefore, we sought to determine the effect of PDGF pretreatment on insulin-stimulated Akt activity in the adipocytes. 3T3-L1 adipocytes were stimulated with insulin or PDGF or were pretreated with PDGF for various times followed by insulin stimulation. Cell lysates were then analyzed by SDS-PAGE followed by Western blotting with either an anti-Akt antibody or an antibody directed against phosphorylated Akt. With the anti-Akt antibody, retarded gel mobility indicates serine/threonine phosphorylation and activation of Akt. The phospho-Akt antibody is directed against phospho-serine at position 473, and again reflects activation of Akt kinase. Ten min of insulin stimulation led to a significant increase in Akt activation (Fig. 6 A, lane 2), as assessed with the phospho-Akt antibody (Fig. 6 A, top panel) or by gel shift with the Akt antibody (Fig. 6 A, bottom panel). PDGF alone also activated Akt, but not to the same extent as insulin (Fig. 6 A, lanes 3–5). Preincubation of the cells with PDGF for 60 or 90 min had no effect on the ability of a subsequent addition of insulin to fully stimulate Akt activation (Fig. 6 A, lanes 6and 7). The activation of Akt, with or without PDGF preincubation, was inhibitable by treatment with wortmannin (Fig. 6 A, lanes 8 and 9), demonstrating that PI3K is necessary for Akt activation.
      Figure thumbnail gr6
      Figure 6Effect of PDGF on insulin-stimulated Akt and p70s6kinase phosphorylation. 3T3-L1 adipocytes were pretreated with or without (–) 100 nm wortmannin for 20 min and preincubated with or without (–) 50 ng/ml PDGF for 10, 60, or 90 min, followed by stimulation with or without (–) 100 ng/ml insulin for 10 min. Proteins from total cell lysates were separated by SDS-PAGE and blotted with antibodies to phospho-Akt (A, top panel), Akt (A, bottom panel), or phospho-p70s6kinase (B). Results are representative of at least three experiments.
      It has been shown that p70s6kinase is downstream of PI3K and Akt and that activation of PI3K and/or Akt is necessary for p70s6kinase stimulation (
      • Burgering B.M.T.
      • Coffer P.J.
      ). To assess the effect of PDGF pretreatment on this pathway, adipocytes were treated with either insulin alone or PDGF alone or were preincubated with PDGF followed by insulin stimulation. The cells were then lysed, followed by Western blotting with a phospho-p70s6kinase antibody. The results of these studies are shown in Fig. 6 B and are exactly parallel to those seen for Akt activation. Thus, insulin led to an increase in p70s6kinase phosphorylation, and the effect of insulin was greater than that of PDGF, which had only a modest and transitory effect. Importantly, PDGF pretreatment had no impact on subsequent insulin stimulation of p70s6kinasephosphorylation. Therefore, despite the effect of PDGF to compromise the ability of insulin to stimulate PI3K activity, neither Akt nor p70s6kinase activation was affected. The maintenance of the signaling pathway leading from PI3K to Akt and p70s6kinase may explain why insulin-stimulated GLUT4 translocation and glucose uptake were unaffected by PDGF pretreatment.

      DISCUSSION

      We and others have previously shown that PI3K activity, as well as proper targeting of this enzyme complex, are necessary for insulin-stimulated glucose transport (
      • Conricode K.M.
      ,
      • Margalet-Sanchez V.
      • Goldfine I.D.
      • Vlahos C.J.
      • Sung C.K.
      ,
      • Okada T.
      • Kawano Y.
      • Sakakibara T.
      • Hazeki O.
      • Ui M.
      ,
      • Cheatham B.
      • Vlahos C.J.
      • Cheatham L.
      • Wang L.
      • Blenis J.
      • Kahn C.R.
      ,
      • Berger J.
      • Hayes N.
      • Szalkowksi D.M.
      • Zhang B.
      ,
      • Clarke J.F.
      • Young P.W.
      • Yonezawa K.
      • Kasuga M.
      • Holman G.D.
      ,
      • Nave B.T.
      • Haigh R.J.
      • Hayward A.C.
      • Siddle K.
      • Shepherd P.R.
      ,
      • Sharma P.M.
      • Egawa K.
      • Gustafson T.A.
      • Martin J.L.
      • Olefsky J.M.
      ). PDGF can stimulate PI3K activity in 3T3-L1 adipocytes but does not lead to stimulation of glucose transport (
      • Conricode K.M.
      ,
      • Margalet-Sanchez V.
      • Goldfine I.D.
      • Vlahos C.J.
      • Sung C.K.
      ,
      • Okada T.
      • Kawano Y.
      • Sakakibara T.
      • Hazeki O.
      • Ui M.
      ). In addition, PDGF treatment can lead to serine/threonine phosphorylation of IRS-1 (
      • Ricort J.-M.
      • Tanti J.-F.
      • Van Obberghen E.
      • Le Marchand-Brustel Y.
      ). In the current study, we have taken advantage of these findings to study the role of PDGF on the mechanisms controlling insulin-stimulated glucose transport. The major findings of these experiments are that PDGF can cause serine/threonine phosphorylation of IRS-1, as measured by altered mobility on a SDS-polyacrylamide gel. This effect of PDGF is inhibited by preincubation with wortmannin, indicating that the p110 subunit of PI3K is responsible for IRS-1 phosphorylation. We find that this serine/threonine phosphorylation of IRS-1 leads to a marked decrease in the ability of IRS-1 to associate with PI3K following insulin stimulation. Surprisingly, this marked decrease in insulin-stimulated IRS-1-associated PI3K activity did not lead to any measurable impairment of several downstream signaling effects of insulin, such as stimulation of GLUT4 translocation, glucose transport, and Akt and p70s6kinase phosphorylation. These results lead to several conclusions and interpretations.
      Insulin leads to extensive tyrosine phosphorylation of IRS-1, which promotes association with the SH2 domains of the p85 subunit of PI3K, leading to stimulation of PI3K activity (
      • Backer J.M.
      • Myers M.G.
      • Shoelson S.E.
      • Chin D.J.
      • Sun X.J.
      • Mirapaipeix M.
      • Hu P.
      • Margolis B.
      • Skolnik E.Y.
      • Schlessinger J.
      • White M.F.
      ). IRS-1 can be serine/threonine-phosphorylated in vitro by the p110 subunit (
      • Lam K.
      • Carpenter C.L.
      • Ruderman N.B.
      • Friel J.C.
      • Kelly K.L.
      ,
      • Tanti J.-F.
      • Gremeaux T.
      • Van Obberghen E.
      • Le Marchand-Brustel Y.
      ), as well as MAPK (
      • De Fea K.
      • Roth R.A.
      ), and in vivo, tumor necrosis factor α (
      • Hotamisligil G.S.
      • Peraldi P.
      • Budavari A.
      • Ellis R.
      • White M.F.
      • Spiegelman B.M.
      ) and PDGF (
      • Ricort J.-M.
      • Tanti J.-F.
      • Van Obberghen E.
      • Le Marchand-Brustel Y.
      ) can cause serine/threonine phosphorylation of IRS-1. Serine phosphorylation of IRS-1 has been shown to prevent its tyrosine phosphorylation by the insulin receptor (
      • Hotamisligil G.S.
      • Peraldi P.
      • Budavari A.
      • Ellis R.
      • White M.F.
      • Spiegelman B.M.
      ) and, consequently, IRS-1 association with PI3K and subsequent PI3K activity (
      • Mothe I.
      • Van Obberghen E.
      ). We explored the functional consequences of these effects. We found that PDGF treatment of 3T3-L1 adipocytes led to serine/threonine phosphorylation of IRS-1, as monitored by mobility shift on SDS-polyacrylamide gel. This effect of PDGF can be inhibited by treatment with the PI3K inhibitors wortmannin and LY294002, but it is not greatly affected by treatment with the MEK inhibitor PD98059. These results indicate that although serine kinases, such as MAPK, can phosphorylate IRS-1 in vitro, (
      • De Fea K.
      • Roth R.A.
      ) in vivophosphorylation following stimulation with PDGF is mediated by the p110 subunit of PI3K or a serine/threonine kinase downstream of PI3K activity.
      More importantly, we studied the functional consequences of PDGF-induced IRS-1 serine/threonine phosphorylation. We found that prior treatment with PDGF caused a decrease in insulin-stimulated IRS-1 tyrosine phosphorylation and a marked impairment in the ability of IRS-1 to associate with PI3K following insulin stimulation. Cells that were pretreated with PDGF showed an 80–90% decrease in insulin stimulatable PI3K activity, precipitated with either anti-phosphotyrosine or IRS-1 antibodies. This inhibition presumably results from a PDGF-induced modification of IRS-1 preventing PI3K from binding to IRS-1. De Fea et al. (
      • De Fea K.
      • Roth R.A.
      ) recently demonstrated that the phosphorylation of IRS-1 on serine 612 causes a conformational change in IRS-1 preventing tyrosine phosphorylation of tyrosine 608, a site of IRS-1/PI3K interaction. The PDGF-induced inhibition of IRS-1-associated PI3K activity was greater than the reduction in insulin-stimulated IRS-1 tyrosine phosphorylation, indicating either that the phosphotyrosines involved in PI3K association were inhibited to a greater extent than overall IRS-1 phosphorylation or that some other biochemical event, such as a conformational change, in addition to decreased tyrosine phosphorylation, is responsible for the block in IRS-1/PI3K association. Surprisingly, this striking inhibition of insulin-stimulated IRS-1-associated PI3K activity had no measurable influence on the downstream insulin signaling events that we examined. For example, it is known that insulin stimulation of GLUT4 translocation and glucose transport is PI3K-dependent (
      • Conricode K.M.
      ,
      • Margalet-Sanchez V.
      • Goldfine I.D.
      • Vlahos C.J.
      • Sung C.K.
      ,
      • Okada T.
      • Kawano Y.
      • Sakakibara T.
      • Hazeki O.
      • Ui M.
      ,
      • Cheatham B.
      • Vlahos C.J.
      • Cheatham L.
      • Wang L.
      • Blenis J.
      • Kahn C.R.
      ,
      • Berger J.
      • Hayes N.
      • Szalkowksi D.M.
      • Zhang B.
      ,
      • Clarke J.F.
      • Young P.W.
      • Yonezawa K.
      • Kasuga M.
      • Holman G.D.
      ), but PDGF pretreatment had no effect on the subsequent ability of insulin to stimulate translocation of GLUT4 proteins to the cell surface or enhance glucose transport activity. On the other hand, even with PDGF pretreatment, insulin stimulation of GLUT4 translocation and glucose transport activity was still completely inhibited by wortmannin and LY294002. This indicates that these effects of insulin remain completely dependent on PI3K activity and eliminates the possibility that a PI3K-independent mechanism, such as that induced by exercise or osmotic shock (
      • Sakaue H.
      • Ogawa W.
      • Takata M.
      • Kuroda S.
      • Kotani K.
      • Matsumoto M.
      • Sakaue M.
      • Nishio S.
      • Ueno H.
      • Kasuga M.
      ,
      • Chen D.
      • Elmendorf J.S.
      • Olson A.L.
      • Li X.
      • Earp H.S.
      • Pessin J.E.
      ), is responsible for the preserved insulin effect under the conditions of PDGF pretreatment.
      Our results show that PDGF treatment inhibits insulin-stimulated IRS-1-associated PI3K activity but does not affect the ability of insulin to stimulate glucose transport, even though glucose transport remains PI3K-dependent. In essence, these results dissociate the function of IRS-1 from stimulation of glucose transport. There are several possible interpretations for these results. The simplest interpretation is that insulin stimulation of IRS-1 tyrosine phosphorylation, with its subsequent binding to and activation of PI3K, is not important for the action of insulin to stimulate glucose transport. Alternatively, it is possible that insulin stimulates glucose transport by at least two parallel pathways, which may be interacting, or redundant. If one of these pathways involves IRS-1, then the blockade of this input is counterbalanced by the alternate pathway. In this event, the alternate pathway would also involve PI3K stimulation, because our results show that under all circumstances, insulin stimulation of glucose transport is strictly PI3K-dependent. A third possibility is that only a small fraction of IRS-1-associated PI3K is necessary for full stimulation of glucose transport. Thus, although only a small amount of IRS-1-associated PI3K activity remains after prolonged PDGF pretreatment, it is possible that this amount of activity, particularly if it is sublocalized to a critical intracellular compartment, can sustain full transport activation. However, if this were the case, one would expect to observe an inhibition of transport at submaximal insulin concentrations. In other words, the dose-response curve for insulin-stimulated glucose transport should be shifted to the right. This is the case, because in the absence of PDGF pretreatment, a low concentration of insulin will lead to a greater amount of IRS-1-associated PI3K activity than will a maximal amount of insulin under conditions of PDGF treatment. As shown in Fig. 4 B, PDGF had no effect on the insulin/glucose transport dose-response curve. On the other hand, it remains possible that a small amount of properly compartmentalized PI3K activity allows glucose transport to be stimulated in a permissive way and that some other insulin-regulatable input into the stimulatory process provides the gating function leading to a dose response.
      The literature bearing on the role of IRS-1 in insulin-stimulated glucose transport is somewhat inconsistent. Some studies argue for a significant role, whereas other papers are more consistent with the view that IRS-1 plays only a minor role, if any, in this particular biologic effect of insulin. For example, IRS-1 knockout mice show marked growth retardation but are only mildly insulin resistant and are not diabetic (

      Bernal, D., Brooks, J., Bruening, J., Towery, H. H., White, M. F., and Yenush, L. (1997) Diabetes 46, Suppl. 1, 5A

      ,
      • Araki E.
      • Lipes M.A.
      • Patti M.E.
      • Bruning J.C.
      • Haag B.
      • Johnson R.S.
      • Kahn C.R.
      ,
      • Yamauchi T.
      • Tobe K.
      • Tamemoto H.
      • Ueki K.
      • Kaburagi Y.
      • Yamamoto-Honda R.
      • Takahashi Y.
      • Yoshizawa F.
      • Aizawa S.
      • Akanuma Y.
      • Sonenberg N.
      • Yazaki Y.
      • Kadowaki T.
      ,
      • Kaburagi Y.
      • Satoh S.
      • Tamemoto H.
      • Yamamoto-Honda R.
      • Tobe K.
      • Veki K.
      • Yamauchi T.
      • Kono-Sugita E.
      • Sekihara H.
      • Aizawa S.
      • Cushman S.W.
      • Akanuma Y.
      • Yazaki Y.
      • Kadowaki T.
      ). These findings are consistent with the formulation that, with respect to metabolic signaling, alternate or redundant pathways exist. Furthermore, fat cells derived from IRS-1 knockout animals show either normal glucose transport (

      Bernal, D., Brooks, J., Bruening, J., Towery, H. H., White, M. F., and Yenush, L. (1997) Diabetes 46, Suppl. 1, 5A

      ) or a 40–50% decrease in glucose transport (
      • Araki E.
      • Lipes M.A.
      • Patti M.E.
      • Bruning J.C.
      • Haag B.
      • Johnson R.S.
      • Kahn C.R.
      ,
      • Yamauchi T.
      • Tobe K.
      • Tamemoto H.
      • Ueki K.
      • Kaburagi Y.
      • Yamamoto-Honda R.
      • Takahashi Y.
      • Yoshizawa F.
      • Aizawa S.
      • Akanuma Y.
      • Sonenberg N.
      • Yazaki Y.
      • Kadowaki T.
      ,
      • Kaburagi Y.
      • Satoh S.
      • Tamemoto H.
      • Yamamoto-Honda R.
      • Tobe K.
      • Veki K.
      • Yamauchi T.
      • Kono-Sugita E.
      • Sekihara H.
      • Aizawa S.
      • Cushman S.W.
      • Akanuma Y.
      • Yazaki Y.
      • Kadowaki T.
      ,
      • Lavan B.E.
      • Lienhard G.E.
      ), despite a complete absence of IRS-1 function. In some studies using IRS-1 knockout mice or cells derived from these animals (

      Bernal, D., Brooks, J., Bruening, J., Towery, H. H., White, M. F., and Yenush, L. (1997) Diabetes 46, Suppl. 1, 5A

      ,
      • Araki E.
      • Lipes M.A.
      • Patti M.E.
      • Bruning J.C.
      • Haag B.
      • Johnson R.S.
      • Kahn C.R.
      ), it has been suggested that IRS-2 might compensate for the loss of IRS-1 activity, although this is not the case in all studies (
      • Kaburagi Y.
      • Satoh S.
      • Tamemoto H.
      • Yamamoto-Honda R.
      • Tobe K.
      • Veki K.
      • Yamauchi T.
      • Kono-Sugita E.
      • Sekihara H.
      • Aizawa S.
      • Cushman S.W.
      • Akanuma Y.
      • Yazaki Y.
      • Kadowaki T.
      ). This is unlikely to provide an explanation for our results, because IRS-2-associated PI3K activity would have been detected in the anti-phosphotyrosine immunoprecipitates. It is interesting to note that in fat cells derived from IRS-1 knockout mice, in which IRS-1 is absent and IRS-2 is negligible, only a 50% reduction in maximal insulin-stimulated glucose transport is observed, with no change in insulin sensitivity, as detected by the insulin dose-response curve. From these studies, Kaburagi et. al. (
      • Kaburagi Y.
      • Satoh S.
      • Tamemoto H.
      • Yamamoto-Honda R.
      • Tobe K.
      • Veki K.
      • Yamauchi T.
      • Kono-Sugita E.
      • Sekihara H.
      • Aizawa S.
      • Cushman S.W.
      • Akanuma Y.
      • Yazaki Y.
      • Kadowaki T.
      ) conclude that an IRS1/2-independent but PI3K-dependent pathway must exist. They suggest that this alternate pathway might involve IRS-3 (
      • Kaburagi Y.
      • Satoh S.
      • Tamemoto H.
      • Yamamoto-Honda R.
      • Tobe K.
      • Veki K.
      • Yamauchi T.
      • Kono-Sugita E.
      • Sekihara H.
      • Aizawa S.
      • Cushman S.W.
      • Akanuma Y.
      • Yazaki Y.
      • Kadowaki T.
      ). However, signaling through IRS-3 is unlikely to explain our results, because tyrosine-phosphorylated IRS-3 (pp60) cannot be detected in 3T3-L1 cells (Ref.
      • Kaburagi Y.
      • Satoh S.
      • Tamemoto H.
      • Yamamoto-Honda R.
      • Tobe K.
      • Veki K.
      • Yamauchi T.
      • Kono-Sugita E.
      • Sekihara H.
      • Aizawa S.
      • Cushman S.W.
      • Akanuma Y.
      • Yazaki Y.
      • Kadowaki T.
      and data not shown). Interestingly, when endogenous IRS-1 was decreased with an antisense ribozyme (
      • Quon M.J.
      • Butte A.J.
      • Zarnowski M.J.
      • Sesti G.
      • Cushman S.W.
      • Taylor S.I.
      ), a shift in the insulin/glucose dose-response curve was observed, whereas no change in maximal responsiveness occurred. In contrast, in fat cells derived from IRS-1 knockout mice, a decrease in maximal responsiveness was noted with no change in the dose-response curve (
      • Araki E.
      • Lipes M.A.
      • Patti M.E.
      • Bruning J.C.
      • Haag B.
      • Johnson R.S.
      • Kahn C.R.
      ). Other studies, in which the interleukin-4 receptor and GLUT4 were overexpressed in L6 myoblasts, show that stimulation with interleukin-4 had no effect on glucose transport, despite the fact that interleukin-4 strongly stimulated tyrosine phosphorylation of IRS-1 and its association with PI3K (
      • Isakoff S.J.
      • Taha C.
      • Rose E.
      • Marcusohn J.
      • Klip A.
      • Skolnik E.Y.
      ). Similarly, Krook et al. (
      • Krook A.
      • Moller D.E.
      • Dib K.
      • O'Rahilly S.
      ) have shown that the expression in CHO cells of two insulin receptor mutants (Arg1174 → Gln,Pro1178 → Leu) can mediate IRS-1 phosphorylation but fail to stimulate glycogen synthesis. Thus, the studies by Isakoff et. al. (
      • Isakoff S.J.
      • Taha C.
      • Rose E.
      • Marcusohn J.
      • Klip A.
      • Skolnik E.Y.
      ) and Krook et. al. (
      • Krook A.
      • Moller D.E.
      • Dib K.
      • O'Rahilly S.
      ) are quite consistent; both show that IRS-1 phosphorylation, with PI3K activation, is not sufficient to initiate metabolic signaling, indicating that an additional insulin-derived metabolic input is necessary.
      In other studies, Morris et al. (
      • Morris A.J.
      • Martin S.S.
      • Haruta T.
      • Nelson J.G.
      • Vollenweider P.
      • Gustafson T.A.
      • Mueckler M.
      • Rose D.W.
      • Olefsky J.M.
      ) microinjected IRS-1 inhibitory antibodies and peptides into 3T3-L1 cells and found no inhibitory effect on insulin-stimulated GLUT4 translocation, despite the fact that other insulin actions, such as cytoskeletal rearrangement and DNA synthesis, were inhibited by these reagents. Similarly, Sharmaet al. (
      • Sharma P.M.
      • Egawa K.
      • Gustafson T.A.
      • Martin J.L.
      • Olefsky J.M.
      ) used an adenovirus system to overexpress the IRS-1 PTB and SAIN domains in 3T3-L1 cells. These domains behaved as competitive inhibitors of insulin receptor/IRS-1 interactions, leading to a markedly decreased IRS-1 phosphorylation and IRS-1-associated PI3K activity. Despite this inhibition of IRS-1/PI3K, insulin-stimulated Akt activation, GLUT4 translocation, and glucose transport were normal in PTB or SAIN domain-expressing 3T3-L1 adipocytes (
      • Sharma P.M.
      • Egawa K.
      • Gustafson T.A.
      • Martin J.L.
      • Olefsky J.M.
      ).
      To further explore the mechanisms of these effects, we examined additional targets of insulin action that are thought to be downstream of PI3K. Akt, a serine/threonine kinase downstream of PI3K, is activated following insulin or PDGF stimulation (
      • Burgering B.M.T.
      • Coffer P.J.
      ). Akt is activated by a dual mechanism involving the binding of PIP3 to the Akt PH domain (
      • Stokoe D.
      • Stephens L.R.
      • Copeland T.
      • Gaffney P.R.J.
      • Reese C.B.
      • Painter G.F.
      • Holmes A.B.
      • McCormick F.
      • Hawkins P.T.
      ), as well as serine/threonine phosphorylation by one or more Akt kinases, which may themselves be stimulated by the lipid products of PI3K (
      • Stokoe D.
      • Stephens L.R.
      • Copeland T.
      • Gaffney P.R.J.
      • Reese C.B.
      • Painter G.F.
      • Holmes A.B.
      • McCormick F.
      • Hawkins P.T.
      ). Furthermore, recent data indicate that activation of Akt may be necessary for GLUT4 translocation (
      • Tanti J.-F.
      • Grillo S.
      • Gremeaux T.
      • Coffer P.J.
      • Van Obberghen E.
      • Le Marchand-Brustel Y.
      ,
      • Kohn A.D.
      • Summers S.A.
      • Birnbaum M.J.
      • Roth R.A.
      ). The expression of a constitutively active Akt construct in rat adipocytes (
      • Tanti J.-F.
      • Grillo S.
      • Gremeaux T.
      • Coffer P.J.
      • Van Obberghen E.
      • Le Marchand-Brustel Y.
      ) or 3T3-L1 adipocytes (
      • Kohn A.D.
      • Summers S.A.
      • Birnbaum M.J.
      • Roth R.A.
      ) increases GLUT4 translocation and glucose uptake. Consistent with these findings, we show that PDGF pretreatment of cells does not impair the ability of insulin to activate Akt. Despite the fact that insulin-stimulated IRS-1-associated PI3K activity is markedly impaired by PDGF pretreatment, the effect of insulin to stimulate Akt in the presence of PDGF is still entirely inhibited by wortmannin and LY294002, indicating that Akt activation is dependent on PI3K. These results support a role for Akt as a PI3K-dependent upstream activator of GLUT4 translocation. Similarly, insulin-induced p70s6kinase phosphorylation was unaffected by PDGF pretreatment, demonstrating that the pathway from PI3K to p70s6kinase was still intact. Taken together, our data suggest that an insulin-regulated pathway for PI3K stimulation, which is relatively independent of IRS-1/IRS-2, must exist and that this pathway is fully sufficient to activate Akt and p70s6kinase, as well as stimulate glucose transport.
      It has been suggested that the ability of insulin, but not PDGF, to stimulate GLUT4 translocation is due to unique subcompartmentalization of PI3K following insulin treatment. Insulin stimulation leads to the localization of PI3K activity to the low-density microsomal, cytosolic, and plasma membrane fractions, whereas PDGF stimulation results only in PI3K activity in the plasma membrane fraction (
      • Ricort J.-M.
      • Tanti J.-F.
      • Van Obberghen E.
      • Le Marchand-Brustel Y.
      ,
      • Nave B.T.
      • Haigh R.J.
      • Hayward A.C.
      • Siddle K.
      • Shepherd P.R.
      ). Our results are fully consistent with this idea, but they further indicate that a pathway independent of, or in addition to, IRS-1, is fully capable of affecting this unique subcompartmentalization of PI3K, such that activated PI3K in this compartment can fully stimulate Akt phosphorylation, p70s6kinase phosphorylation, GLUT4 translocation, and glucose transport.
      We propose the existence of another insulin-stimulated protein that interacts with and activates PI3K independent of IRS-1. IRS-2 has been shown to substitute for IRS-1 in adipocytes when expression of IRS-1 is compromised (

      Bernal, D., Brooks, J., Bruening, J., Towery, H. H., White, M. F., and Yenush, L. (1997) Diabetes 46, Suppl. 1, 5A

      ,
      • Araki E.
      • Lipes M.A.
      • Patti M.E.
      • Bruning J.C.
      • Haag B.
      • Johnson R.S.
      • Kahn C.R.
      ). However, because phosphotyrosine antibodies would precipitate IRS-2/PI3K activity, this is an unlikely explanation for our results. Thus, PI3K may be capable of interacting with another insulin receptor substrate, such as IRS-4 or some other protein. Alternatively, cross-talk between insulin receptor signal transduction and G-protein coupled receptors may be an explanation for our observations. The βγ-subunits of G-proteins activate a PI3K that is wortmannin-sensitive (
      • Tang X.
      • Downes C.P.
      ) and leads to Akt activation (
      • Tilton B.
      • Andjelkovic M.
      • Didichenko S.A.
      • Hemmings B.A.
      • Thelen M.
      ). Future experiments are necessary to determine the route of insulin-stimulated PI3K activity following PDGF treatment.

      REFERENCES

        • Seger R.
        • Krebs E.G.
        FASEB J. 1995; 9: 726-735
        • Conricode K.M.
        Biochem. Mol. Biol. Int. 1995; 36: 835-843
        • Margalet-Sanchez V.
        • Goldfine I.D.
        • Vlahos C.J.
        • Sung C.K.
        Biochem. Biophys. Res. Commun. 1994; 204: 446-452
        • Okada T.
        • Kawano Y.
        • Sakakibara T.
        • Hazeki O.
        • Ui M.
        J. Biol. Chem. 1994; 269: 3568-3573
        • Cheatham B.
        • Vlahos C.J.
        • Cheatham L.
        • Wang L.
        • Blenis J.
        • Kahn C.R.
        Mol. Cell. Biol. 1994; 14: 4902-4911
        • Berger J.
        • Hayes N.
        • Szalkowksi D.M.
        • Zhang B.
        Biochem. Biophys. Res. Commun. 1994; 205: 570-576
        • Clarke J.F.
        • Young P.W.
        • Yonezawa K.
        • Kasuga M.
        • Holman G.D.
        Biochem. J. 1994; 300: 631-635
        • Kapeller R.
        • Cantley L.C.
        BioEssays. 1994; 16: 565-576
        • Stokoe D.
        • Stephens L.R.
        • Copeland T.
        • Gaffney P.R.J.
        • Reese C.B.
        • Painter G.F.
        • Holmes A.B.
        • McCormick F.
        • Hawkins P.T.
        Science. 1997; 277: 567-570
        • Dhand R.
        • Hiles I.
        • Panayotou G.
        • Roche S.
        • Fry M.J.
        • Gout I.
        • Totty N.F.
        • Truong O.
        • Vicendo P.
        • Yonezawa K.
        • Kasuga M.
        • Courtneidge S.A.
        • Waterfield M.D.
        EMBO J. 1994; 13: 522-533
        • Lam K.
        • Carpenter C.L.
        • Ruderman N.B.
        • Friel J.C.
        • Kelly K.L.
        J. Biol. Chem. 1994; 269: 20648-20652
        • Tanti J.-F.
        • Gremeaux T.
        • Van Obberghen E.
        • Le Marchand-Brustel Y.
        Biochem. J. 1994; 304: 17-21
        • Wiese R.J.
        • Mastik C.C.
        • Lazar D.F.
        • Saltiel A.R.
        J. Biol. Chem. 1995; 270: 3442-3446
        • De Fea K.
        • Roth R.A.
        J. Biol. Chem. 1997; 272: 31400-31406
        • Ruderman N.B.
        • Kapeller R.
        • White M.F.
        • Cantley L.C.
        Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 1411-1415
        • Walsh J.P.
        • Caldwell K.K.
        • Majerus P.W.
        Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 9184-9187
        • Haruta T.
        • Morris A.J.
        • Rose D.W.
        • Nelson J.G.
        • Mueckler M.
        • Olefsky J.M.
        J. Biol. Chem. 1995; 270: 27991-27994
        • Klip A.
        • Li G.
        • Logan W.J.
        Am. J. Physiol. 1984; 247: E291-E296
        • Gould G.W.
        • Holman G.D.
        Biochem. J. 1993; 295: 329-341
        • Morris A.J.
        • Martin S.S.
        • Haruta T.
        • Nelson J.G.
        • Vollenweider P.
        • Gustafson T.A.
        • Mueckler M.
        • Rose D.W.
        • Olefsky J.M.
        Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 8401-8406
        • Ricort J.-M.
        • Tanti J.-F.
        • Van Obberghen E.
        • Le Marchand-Brustel Y.
        Eur. J. Biochem. 1996; 239: 17-22
        • Burgering B.M.T.
        • Coffer P.J.
        Nature. 1995; 376: 599-602
        • Nave B.T.
        • Haigh R.J.
        • Hayward A.C.
        • Siddle K.
        • Shepherd P.R.
        Biochem. J. 1996; 318: 55-60
        • Sharma P.M.
        • Egawa K.
        • Gustafson T.A.
        • Martin J.L.
        • Olefsky J.M.
        Mol. Cell. Biol. 1997; 17: 7386-7397
        • Ricort J.-M.
        • Tanti J.-F.
        • Van Obberghen E.
        • Le Marchand-Brustel Y.
        J. Biol. Chem. 1997; 272: 19814-19818
        • Backer J.M.
        • Myers M.G.
        • Shoelson S.E.
        • Chin D.J.
        • Sun X.J.
        • Mirapaipeix M.
        • Hu P.
        • Margolis B.
        • Skolnik E.Y.
        • Schlessinger J.
        • White M.F.
        EMBO J. 1992; 11: 3469-3479
        • Hotamisligil G.S.
        • Peraldi P.
        • Budavari A.
        • Ellis R.
        • White M.F.
        • Spiegelman B.M.
        Science. 1996; 271: 665-668
        • Mothe I.
        • Van Obberghen E.
        J. Biol. Chem. 1996; 271: 11222-11227
        • Sakaue H.
        • Ogawa W.
        • Takata M.
        • Kuroda S.
        • Kotani K.
        • Matsumoto M.
        • Sakaue M.
        • Nishio S.
        • Ueno H.
        • Kasuga M.
        Mol. Endocrinol. 1997; 11: 1552-1562
        • Chen D.
        • Elmendorf J.S.
        • Olson A.L.
        • Li X.
        • Earp H.S.
        • Pessin J.E.
        J. Biol. Chem. 1997; 272: 27401-27410
      1. Bernal, D., Brooks, J., Bruening, J., Towery, H. H., White, M. F., and Yenush, L. (1997) Diabetes 46, Suppl. 1, 5A

        • Araki E.
        • Lipes M.A.
        • Patti M.E.
        • Bruning J.C.
        • Haag B.
        • Johnson R.S.
        • Kahn C.R.
        Nature. 1994; 372: 186-190
        • Yamauchi T.
        • Tobe K.
        • Tamemoto H.
        • Ueki K.
        • Kaburagi Y.
        • Yamamoto-Honda R.
        • Takahashi Y.
        • Yoshizawa F.
        • Aizawa S.
        • Akanuma Y.
        • Sonenberg N.
        • Yazaki Y.
        • Kadowaki T.
        Mol. Cell. Biol. 1996; 16: 3074-3084
        • Kaburagi Y.
        • Satoh S.
        • Tamemoto H.
        • Yamamoto-Honda R.
        • Tobe K.
        • Veki K.
        • Yamauchi T.
        • Kono-Sugita E.
        • Sekihara H.
        • Aizawa S.
        • Cushman S.W.
        • Akanuma Y.
        • Yazaki Y.
        • Kadowaki T.
        J. Biol. Chem. 1997; 272: 25839-25844
        • Lavan B.E.
        • Lienhard G.E.
        J. Biol. Chem. 1993; 268: 5921-5928
        • Quon M.J.
        • Butte A.J.
        • Zarnowski M.J.
        • Sesti G.
        • Cushman S.W.
        • Taylor S.I.
        J. Biol. Chem. 1994; 269: 27920-27924
        • Isakoff S.J.
        • Taha C.
        • Rose E.
        • Marcusohn J.
        • Klip A.
        • Skolnik E.Y.
        Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 10247-10251
        • Krook A.
        • Moller D.E.
        • Dib K.
        • O'Rahilly S.
        J. Biol. Chem. 1996; 271: 7134-7140
        • Tanti J.-F.
        • Grillo S.
        • Gremeaux T.
        • Coffer P.J.
        • Van Obberghen E.
        • Le Marchand-Brustel Y.
        Endocrinology. 1997; 138: 2005-2010
        • Kohn A.D.
        • Summers S.A.
        • Birnbaum M.J.
        • Roth R.A.
        J. Biol. Chem. 1996; 271: 31372-31378
        • Tang X.
        • Downes C.P.
        J. Biol. Chem. 1997; 272: 14193-14199
        • Tilton B.
        • Andjelkovic M.
        • Didichenko S.A.
        • Hemmings B.A.
        • Thelen M.
        J. Biol. Chem. 1997; 272: 28096-28101