5 -AMP-activated Protein Kinase Phosphorylates IRS-1 on Ser-789 in Mouse C2C12 Myotubes in Response to 5-Aminoimidazole-4- carboxamide Riboside*

Transcription

1 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 276, No. 50, Issue of December 14, pp , by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. 5 -AMP-activated Protein Kinase Phosphorylates IRS-1 on Ser-789 in Mouse C2C12 Myotubes in Response to 5-Aminoimidazole-4- carboxamide Riboside* Received for publication, August 22, 2001, and in revised form, September 28, 2001 Published, JBC Papers in Press, October 11, 2001, DOI /jbc.C Søren N. Jakobsen, D. Grahame Hardie, Nick Morrice, and Hans E. Tornqvist From Diabetes Biology, Novo Nordisk A/S, Novo Alle, 2880 Bagsvaerd, Denmark and the Division of Molecular Physiology and Medical Research Council Protein Phosphorylation Unit, School of Life Sciences, University of Dundee, Wellcome Trust Biocentre, Dow Street, Dundee DD1 5EH, Scotland, United Kingdom Exercise is known to increase insulin sensitivity and is an effective form of treatment for the hyperglycemia observed in type 2 diabetes. Activation of 5 -AMP-activated protein kinase (AMPK) by 5-aminoimidazole-4- carboxamide riboside (AICAR), exercise, or electrically stimulated contraction leads to increased glucose transport in skeletal muscle. Here we report the first evidence of a direct interaction between AMPK and the most upstream component of the insulin-signaling cascade, insulin receptor substrate-1 (IRS-1). We find that AMPK rapidly phosphorylates IRS-1 on Ser-789 in cellfree assays as well as in mouse C2C12 myotubes incubated with AICAR. In the C2C12 myotubes activation of AMPK by AICAR matched the phosphorylation of IRS-1 on Ser-789. This phosphorylation correlates with a 65% increase in insulin-stimulated IRS-1-associated phosphatidylinositol 3-kinase activity in C2C12 myotubes preincubated with AICAR. The binding of phosphatidylinositol 3-kinase to IRS-1 was not affected by AICAR. These results demonstrate the existence of an interaction between AMPK and early insulin signaling that could be of importance to our understanding of the potentiating effects of exercise on insulin signaling. Both exercise and insulin stimulate glucose transport in skeletal muscle (1, 2). Activation of the IRS-1 1 /PI 3-kinase axis is an absolute requirement for the effects mediated by insulin (3). In the absence of insulin, in vitro contraction of isolated skeletal muscle can increase glucose transport, suggesting that this stimulus uses a mechanism distinct from that of insulin (1, 4). However, in animals (5) and in humans (6), there appears to be a requirement for the presence of insulin to obtain fully stimulated glucose uptake in muscle following exercise. The question as to whether exercise mediates muscle glucose uptake completely independently of insulin or whether exercise and insulin have a common interaction point in vivo is therefore still unresolved. * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. To whom correspondence should be addressed. Tel.: ; Fax: ; snyj@novonordisk.com. 1 The abbreviations used are: IRS-1, insulin receptor substrate-1; PI, phosphatidylinositol; AMPK, 5 -AMP-activated protein kinase; AICAR, 5-aminoimidazole-4-carboxamide riboside; DMEM, Dulbecco s modified Eagle s medium; PAGE, polyacrylamide gel electrophoresis; TLE, thin-layer electrophoresis; TLC, thin-layer chromatography; HPLC, high pressure liquid chromatography; PI3P, phosphatidylinositol 3-phosphate Studies from several laboratories have suggested that AMPK is a major mediator of contraction-induced glucose uptake in skeletal muscle based on the observation that AMPK activity is rapidly elevated by exercise or electrically stimulated contraction (7, 8). In addition, AICAR, a cell-permeable activator of AMPK (9), stimulates glucose uptake in perfused and isolated muscle (10, 11). Initially AMPK was found to phosphorylate and inactivate 3-hydroxy-3-methylglutaryl-CoA reductase. Later AMPK was demonstrated to modulate a number of key enzymes in energy-regulating metabolic processes including fatty acid and cholesterol metabolism, another aspect of the cellular response to energy demand that is highly regulated by insulin (12). However, direct cross-talk between AMPK and the most central axis of insulin signaling has not yet been demonstrated. The insulin receptor substrate IRS-1 contains multiple potential tyrosine phosphorylation sites in consensus motifs recognized by the insulin receptor tyrosine kinase (13). A number of proteins including the PI 3-kinase can bind to these tyrosine phosphorylation sites via their respective Src homology 2 domains (14). In addition, IRS-1 contains a large number of potential serine and threonine phosphorylation sites (15). Relatively little is known about the protein-serine/threonine kinases that act on IRS-1, and the mechanisms by which insulin signaling is influenced by serine phosphorylation of IRS-1 are not yet well understood (16 21). Here we provide the first evidence for a direct interaction between AMPK and the most immediate insulin-signaling events by identifying Ser-789 on IRS-1 as a specific phosphorylation site for AMPK. Our results from C2C12 myotubes stimulated with AICAR suggest the existence of a close interaction between AMPK and early insulin signaling that could be of importance to the potentiating effects of exercise on the metabolic effect of insulin. EXPERIMENTAL PROCEDURES Materials Human insulin was from Novo Nordisk (Bagsvaerd, Denmark). [ - 32 P]ATP was obtained from Amersham Pharmacia Biotech. Monoclonal mouse anti-phosphotyrosine antibody PY100 was from Cell Signal Technology. Rat recombinant IRS-1 protein, anti-irs-1, antip85, and anti-phosphoserine 473-Akt antibodies were from Upstate Biotechnology. Production of anti-phosphothreonine 172-AMPK, anti- AMPK 1, and anti-ampk 2 antibodies and purified AMPK and SAMS peptide has been described elsewhere (22 24). A sheep was immunized with a keyhole limpet hemocyanin-conjugated phosphopeptide, LRLSSSpSGRLR where Sp phosphoserine, corresponding to amino acids on rat IRS-1. The Ser(P)-789 phospho-specific antibodies were collected by affinity purification of the antiserum by sequential chromatographies on LRLSSSpSGRLR-CH-Sepharose and LRLSSSS- GRLR-CH-Sepharose (CH-Sepharose from Amersham Pharmacia Biotech). Sequencing grade trypsin, endoproteinase Glu-C, and Complete Protease Inhibitor Mixture were from Roche Molecular Biochemicals. This paper is available on line at

2 Serine Phosphorylation of IRS-1 by AMPK Other chemicals were obtained from Sigma Chemical Co. The peptide LRLSSSpSGRLR corresponding to amino acids in rat IRS-1 was synthesized by Dr. G. Bloomberg, University of Bristol. Treatment of Cells Mouse C2C12 myoblasts (ATCC) were maintained in Dulbecco s modified Eagle s medium (DMEM), 4500 mg/liter glucose supplemented with 10% fetal calf serum, 50 units/ml penicillin, and 50 g/ml streptomycin. Confluent cells were differentiated in DMEM containing 2% horse serum and fused into myotubes after 4 days. Before experiments cells were starved for 3 h in serum-free DMEM and then incubated with 100 nm wortmannin (35 min), 0.5 mm AICAR (30 min), and 10 nm insulin (10 min) alone or in combinations. Wortmannin was included 5 min prior to AICAR and 25 min prior to insulin. AICAR was included 20 min prior to insulin stimulation. Cells were lysed on ice and either subjected to subcellular fractionation as described below or lysed as total cell lysate including 1% Triton X-100 in the lysis buffer (20 mm Tris acetate, ph 7.0, 270 mm sucrose, 1 mm EDTA, 1 mm EGTA, 1 mm Na 3 VO 4,50 M NaF, 5 mm sodium pyrophosphate, 10 mm sodium -glycerophosphate, 1 mm benzamidine, 4 g/ml leupeptin, 4 g/ml aprotinin, 0.2 mm 4-(2-aminoethyl)benzenesulfonyl fluoride). Subcellular Fractionation of Cell Lysates C2C12 myotubes were fractionated following a procedure simplified from that described by Clark et al. (25). Cells were lysed in cold lysis buffer, without Triton X-100, by passing 12 times through a 25-gauge needle. Lysates were centrifuged for 20 min at 38,000 g to remove plasma membranes and high-density microsomes. The resultant supernatant was centrifuged for 75 min at 170,000 g to give a pellet (dissolved in lysis buffer) designated high-speed pellet and a supernatant designated cytosol. Immunoblot Analysis Total cell lysate and high-speed pellet protein was subjected to SDS-PAGE and immunoblotted with the indicated antibodies (1:1000 dilution) followed by detection using enhanced chemiluminescence (ECL, Amersham Pharmacia Biotech). Membranes were routinely stripped and reblotted with a second primary antibody. In Vitro AMPK Assay Rat recombinant IRS-1 protein was phosphorylated in vitro by AMPK, purified from rat liver as described previously (26). The reactions were carried out on 0.4 g of IRS-1 in AMPK assay buffer (50 mm Hepes, ph 7.2, 1 mm dithiothreitol, 0.2 mm AMP, 0.02% Brij-35, Complete Protease Inhibitor Mixture (1:50 dilution)), 10 mm magnesium acetate, 100 M [ - 32 P]ATP (specific activity of 10 6 cpm/nmol) using 1 unit/ml AMPK (1 unit of AMPK phosphorylates 1 nmol of SAMS peptide/min at 30 C) in a final volume of 30 l at30 C for various times. Following SDS-PAGE the IRS-1 bands were cut out, and radioactivity was measured by liquid scintillation counting. The K m and V max values for AMPK activity toward the IRS-1 peptide (LRLSSSSGRLR), its corresponding phosphopeptide (LRLSSSpS- GRLR), and the SAMS peptide (HMRSAMSGLHLVKRR) were determined with increasing concentrations of substrate peptide in a total volume of 20 l under the above-described conditions After 20 min the reaction was stopped and spotted onto phosphocellulose Whatman P81 paper that was washed with 75 mm phosphoric acid and counted by Cerenkov counting. Two-dimensional Phosphopeptide Mapping Phosphorylated IRS-1 was separated from AMPK by SDS-PAGE, cut out from the dried gel, and washed sequentially in water; 50% acetonitrile, H 2 O; 100 mm NH 4 HCO 3, and 50 mm NH 4 HCO 3, 50% acetonitrile followed by in-gel digestion with 0.2 g/ml of sequencing grade trypsin in 50 mm NH 4 HCO 3 overnight at 37 C. 32 P-Radiolabeled tryptic peptides were recovered by extracting with 50% acetonitrile, 50 mm NH 4 HCO 3 and subjected to two-dimensional TLE/TLC (27). Radioactive phosphopeptides were detected by PhosphorImager technology. Determination of Phosphorylated Site on IRS-1 Peptides IRS-1-derived tryptic 32 P-peptides were lyophilized, reconstituted in 50% acetonitrile, 0.1% trifluoroacetic acid, loaded onto a C18 reverse phase peptide/protein column (Vydac C mm, catalog no. 218TP54), and separated against an acetonitrile gradient in 0.1% trifluoroacetic acid at a flow rate of 300 l/min. 32 P-Labeled phosphopeptides were detected by liquid scintillation counting of the 0.4-ml fractions collected (28). 32 P-Labeled IRS-1 phosphopeptides isolated by HPLC fractionation were subjected to phosphoamino acid analysis (27). The site of phosphorylation was determined by solid phase Edman degradation of the 32 P-labeled tryptic IRS-1 phosphopeptides (29). IRS-1 digested with 0.2 g/ml endopeptidase Glu-C in 50 mm ammonium bicarbonate overnight at 37 C was analyzed by the same procedure as for trypsindigested IRS-1. Determination of IRS-1-associated PI 3-Kinase Activity IRS-1 and p85-associated IRS-1 was immunoprecipitated from cell lysates (250 g of protein) using 1.5 g of anti-irs-1 or anti-p85 antibodies, respectively, and collected with protein A-Sepharose beads. IRS-1-associated PI 3-kinase activity in the immunocomplex was measured and detected as described previously (30). RESULTS AMPK Phosphorylates IRS-1 on Ser-789 in Vitro From the screening of a number of purified kinases for their ability to phosphorylate IRS-1, purified AMPK from rat liver was found to rapidly phosphorylate recombinant rat IRS-1 in vitro with a stoichiometry of phosphorylation reaching 0.4 mol of phosphate/mol of IRS-1 (Fig. 1A). Two-dimensional tryptic phosphopeptide mapping of IRS-1 established that the phosphate was primarily incorporated into one single tryptic peptide (Fig. 1B, inset). For the identification of the site phosphorylated on IRS-1 by AMPK we resolved the tryptic digest of IRS-1 by reverse phase HPLC and detected a single peak of radioactivity (Fig. 1B). This HPLC fraction was first subjected to solid phase Edman degradation with release of radioactivity during cycle 4 (Fig. 1C). Phosphoamino acid analysis showed phosphorylation exclusively on serine (Fig. 1D). Assuming no missed cleavages, trypsin would be expected to generate eight peptides from IRS-1 containing a serine residue 4 amino acids distal to the cleavage site. The consensus recognition motif for AMPK is Hyd-(X,Basic)XX(S/T)XXX-Hyd (where Hyd M, L, I, F, or V and Basic R K H (31)). A search of the rat IRS-1 sequence using this motif with the program Findpatterns (from the GCG suite (32)) revealed a single fit to this motif, i.e. Ser-789. This sequence, which is conserved in the human, and the mouse IRS-1 sequence also contain a histidine 6 residues N-terminal to the phosphorylated serine. A basic residue at this position is an additional positive determinant for AMPK. 2 This predicted phosphorylation site at Ser-789 was confirmed by digesting IRS-1 with endopeptidase Glu-C. The HPLC separation of AMPK-phosphorylated IRS-1 digested with endopeptidase Glu-C gave rise to two peaks containing phosphorylated serine that eluted at similar times. Both peaks gave release of phosphate on Edman degradation cycle 12 (data not shown). Ser-789 is the only serine in the rat IRS-1 sequence that is 4 amino acids distal to a trypsin protease cleavage site (residues ) and 12 amino acids distal to an endopeptidase Glu-C cleavage site (residues /801). In addition AMPK readily phosphorylated the synthetic dephosphopeptide but not the phosphopeptide corresponding to the amino acid sequence of rat IRS-1 (Fig. 1E). The dephosphopeptide was a good substrate for AMPK with a K m of 33 5 M compared with a K m of 49 4 M for the SAMS peptide, the standard substrate for AMPK. AMPK had a similar V max toward both peptides. On the basis of these data we conclude that AMPK readily phosphorylates IRS-1 on Ser-789 in vitro. AMPK Phosphorylates IRS-1 on Ser-789 in Intact Cells To determine whether AMPK phosphorylates IRS-1 on Ser-789 in intact cells, we raised phospho-specific antibodies in sheep against the phosphopeptide LRLSSSpSGRLR. This antibody recognized rat recombinant IRS-1 that had been incubated with MgATP in the presence but not the absence of AMPK (not shown). Incubation of the Ser(P)-789 antibody with the LRLSSSpSGRLR peptide used to raise the antibody abolished its recognition of IRS-1 phosphorylated by AMPK (not shown). Whether IRS-1 is phosphorylated on Ser-789 in intact cells was investigated in mouse C2C12 myotubes stimulated with AICAR. IRS-1 in the high-speed pellet fraction of untreated C2C12 myotubes did not react with the Ser(P)-789 antibody by immunoblot analysis. However, after incubation of the cells with 0.5 mm AICAR for min, IRS-1 from C2C12 myo- 2 D. G. Hardie and J. W. Scott, unpublished.

3 46914 Serine Phosphorylation of IRS-1 by AMPK FIG. 1. AMPK phosphorylates IRS-1 on Ser-789 in vitro. A, recombinant IRS-1 (3.5 pmol) was incubated with purified AMPK for various times in the presence of 100 M Mg[ - 32 P]ATP and separated by SDS-PAGE. Radioactivity incorporated into IRS-1 was determined by liquid scintillation counting, and the stoichiometry in phosphorylation was assessed knowing the specific activity of the [ - 32 P]ATP used. The bottom panel shows the autoradiogram from one representative experiment, whereas the graph at the top shows pooled data from four experiments. P, phosphate. B, recombinant IRS-1, maximally phosphorylated by AMPK (45 min), was separated from AMPK by SDS-PAGE and digested with trypsin. The recovered 32 P-radiolabeled tryptic peptide mixture was either separated by reverse phase HPLC and fractions were counted by liquid scintillation or separated by TLE/TLC and detected by PhosphorImager technology (inset). C, the single reverse phase HPLC fraction of tryptic IRS-1 peptides containing radioactivity was subjected to solid phase Edman sequencing, and radioactivity released at each cycle was counted by liquid scintillation. D, phosphoamino acid analysis of the isolated 32 P-radiolabeled tryptic peptide. ps, phosphoserine; pt, phosphothreonine; py, phosphotyrosine. E, kinetics in AMPK activity toward the IRS-1 peptide (LRLSSSSGRLR) (f), its corresponding phosphopeptide (LRLSSSpS- GRLR) ( ) and the SAMS peptide (HMRSAMSGLHLVKRR) ( ) was determined. Results are presented as means of one representative triplicate assay S.D. K m and V max values were determined by nonlinear regression using GraphPad Prism 3 software. tubes was clearly detected using the Ser(P)-789-specific antibody for immunoblotting (Fig. 2A). Phosphorylation on Ser-789 of IRS-1 was also found in primary rat hepatocytes and in primary rat adipocytes after stimulation with AICAR (data not shown). To estimate the extent of Ser-789 phosphorylation on IRS-1 in C2C12 myotubes we analyzed, by Western blot analysis using the anti-ser(p)-789 antibody, samples from AICARtreated C2C12 myotubes together with recombinant IRS-1 phosphorylated by purified AMPK in vitro. Phosphorylation of IRS-1 on Ser-789 in C2C12 myotubes appeared comparable to or greater than the stoichiometry of phosphorylation achieved in vitro (Fig. 2B). Insulin did not induce phosphorylation on Ser-789, and phosphorylation induced by AICAR could not be inhibited by wortmannin, demonstrating that the phosphorylation on Ser-789 does not occur through phosphorylation by Akt or other insulin-activated kinases (Fig. 3). AMPK 1 activation in C2C12 myotubes, judged by assessment of Thr-172 phosphorylation using an anti-thr(p)-172 antibody, was induced by AICAR treatment and not by insulin stimulation (Fig. 3). AMPK 2 activation by AICAR was also detected but with a much weaker signal than AMPK 1 (data not shown). No modulation of insulin-induced phosphorylation of IRS-1 on tyrosine or activation of Akt as judged by phosphorylation of Akt on Ser-473 was seen in response to AICAR (Fig. 3). Total levels of IRS-1, p85 subunit of PI 3-kinase, or Akt in the high-speed pellet were also unchanged in response to AICAR and/or insulin (Fig. 3). IRS-1-associated PI 3-Kinase Activity To further investigate whether phosphorylation of Ser-789 on IRS-1 induces changes in the formation or activity of the IRS-1-Tyr(P) p85-pi 3-kinase complex we measured the binding and activation of PI 3-kinase in IRS-1 immunoprecipitates of whole C2C12 cell extracts treated with insulin and/or AICAR. In the absence of insulin no change in p85 or PI 3-kinase activity associated with IRS-1 was seen. In the presence of insulin AICAR induced a significant 65% additional increase in IRS-1-associated PI 3-kinase activity, whereas no change in IRS-1 tyrosine phosphorylation or binding of p85 to IRS-1 (measured as IRS-1 in p85 immunoprecipitates) was detected (Fig. 4). This synergistic increase in IRS-1-associated PI 3-kinase activity was sustained

4 Serine Phosphorylation of IRS-1 by AMPK FIG. 2. Western blot analysis of AMPK phosphorylation of IRS-1 on Ser-789. A, C2C12 myotubes were incubated with 0.5 mm AICAR for the times indicated, and high-speed pellets (15 g of protein) of cell lysates were analyzed for Ser-789 phosphorylation on IRS-1. B, Ser-789 phosphorylation in the same samples from time 0 and 30 min were compared with Ser-789 phosphorylation of 5, 10, and 50 ng of recombinant IRS-1 that had been incubated with MgATP in the presence ( ) or absence ( ) of purified AMPK. Samples were subjected to SDS-PAGE and probed for phosphorylation of Ser-789 in IRS-1 by immunoblotting using anti-phospho-ser-789 (ps789) antibody and reprobed using a nonphospho-specific anti-irs-1 antibody to allow comparison of the amounts of IRS-1 loaded onto the gel. FIG. 3. Effects of Ser-789 phosphorylation on tyrosine phosphorylation of IRS-1 in C2C12 myotubes. C2C12 myotubes were stimulated with insulin (10 nm) and/or AICAR (0.5 mm) in the presence or absence of wortmannin as indicated, and cell extracts were analyzed by immunoblot analysis. High-speed pellet fractions (15 g of protein) or total cell lysate (25 g of protein, AMPK only) were subjected to SDS-PAGE and immunoblotted using selective Ser(P)-789-IRS-1, phosphoserine 473-Akt (ps473akt), or phosphothreonine 172-AMPK (pt172-ampk) antibodies as indicated. Blots were then stripped and reblotted as appropriate with anti-tyr(p), anti-irs-1, anti-p85, anti- Akt, and anti-ampk 1 antibodies as indicated. py, phosphotyrosine. for up to 60 min (data not shown). AICAR on its own did not increase PI 3-kinase activity. FIG. 4.IRS-1-associated PI 3-kinase activity following insulin and AICAR stimulation. C2C12 myotubes were stimulated with insulin (10 nm) and/or AICAR (0.5 mm), and IRS-1-associated PI 3-kinase activity in anti-irs-1 immunoprecipitates from total cell lysates was measured using phosphatidylinositol as substrate. For comparison tyrosine phosphorylation of IRS-1 in the IRS-1 immunoprecipitate and IRS-1 associated with the p85 subunit of PI 3-kinase in an anti-p85 immunoprecipitate of the total cell extracts was analyzed by immunoblotting using anti-phosphotyrosine (py) and anti-irs-1 antibodies, respectively (upper panels). For estimating PI 3-kinase activity the 32 P radioactivity incorporated into PI3P was determined by thin-layer chromatography and PhosphorImager analysis of the TLC plates. One representative experiment of TLC-resolved PI3 32 P spots is shown. The combined results of four independent experiments are presented as -fold radioactivity in PI3 32 P relative to that in the insulin-stimulated sample S.E. DISCUSSION By taking an in vitro based approach to screen for purified kinases that demonstrate activity toward IRS-1 we identified AMPK as a potential modifier of insulin signaling through phosphorylation of IRS-1 on Ser-789. The in vitro activity of AMPK toward recombinant IRS-1 was directed toward a single peptide suggesting that AMPK had a high specificity toward this sequence that is conserved in human, rat, and mouse IRS-1. This was supported through determination of the K m for AMPK toward the IRS-1 peptide, which showed that this primary sequence is a good substrate compared with the SAMS peptide, a well characterized AMPK substrate. By producing an antibody that is specific for the detection of IRS-1 phosphorylated on Ser-789 we were able to investigate the occurrence of this phosphorylation in the intact cell by using the high-speed pellet fraction where IRS-1 is highly enriched (25). The degree of phosphorylation of IRS-1 on Ser-789 was rapid to a substantial stoichiometry with a time course consistent with previous studies of AMPK activation by AICAR in intact cells (9). Given that AMPK phosphorylates recombinant IRS-1 exclusively on Ser-789 in vitro, that this site appears to be an excellent substrate for AMPK, and that the same site is phosphorylated in intact cells in response to AICAR, it is very likely that AMPK does indeed phosphorylate IRS-1 in the intact cell. We have no evidence that other protein kinases phosphorylate the same site. If a kinase other than AMPK is responsible, it does not appear to lie downstream of the insulin receptor. Previously serine phosphorylation of IRS-1 has been demonstrated to correlate with a reduced capacity of insulin to activate PI 3-kinase (17, 18, 20, 21). In this study we did not observe impaired insulin signaling in C2C12 myotubes despite the observation that IRS-1 appears to be phosphorylated on Ser-789 to a high stoichiometry following AICAR treatment. Importantly and in contrast, we observed an increased activity of IRS-1-associated PI 3-kinase as seen in anti-irs-1 immunoprecipitates. Since neither insulin-induced association between the p85 subunit of PI 3-kinase and IRS-1 nor cellular distribution of p85, IRS-1, or IRS-1-Tyr(P) is changed by AICAR stimulation it appears that the PI 3-kinase associated with Ser-789- phosphorylated IRS-1 is in a more active form than the PI 3-kinase associated with IRS-1 not phosphorylated on Ser-789. The reason why we do not see altered insulin-stimulated Akt phosphorylation in the presence of AICAR is not known. One explanation could be given by previous findings suggesting that only a small amount of insulin is necessary to fully activate Akt (33, 34). Recently it has been reported that expression of a dominant inhibitory mutant of AMPK in mouse muscle completely blocks the ability of AICAR to activate hexose uptake, whereas con-

5 46916 Serine Phosphorylation of IRS-1 by AMPK traction-induced hexose uptake is only partially reduced indicating that there are at least partly divergent pathways responsible for the effects of AICAR versus contraction on glucose uptake (35). Partly divergent pathways for AICAR- versus contraction-induced effects on glucose uptake could possibly explain the divergence between our findings and the reported decrease in IRS-1-associated PI 3-kinase activity in insulinstimulated muscle after contraction (36, 37). Our findings that IRS-1 phosphorylation on Ser-789 also occurs in primary rat adipocytes and in hepatocytes exposed to AICAR imply that there has to be a stimulus for phosphorylation of Ser-789 other than contraction. It points to the existence of a more general mechanism for insulin-responsive cells to react to changes in ATP levels. Possibly this includes direction of signaling via IRS-1 toward intracellular energy-restoring processes, e.g. by increasing glucose uptake and reducing fatty acid and cholesterol synthesis. Several reports from studies on incubated muscles have shown that the PI 3-kinase inhibitor wortmannin inhibits insulin-stimulated glucose uptake but not contraction-stimulated glucose uptake (11, 38). However, another report demonstrates that wortmannin does inhibit some of the contraction-stimulated glucose transport in rat skeletal muscle suggesting that the signaling pathways activated by insulin and exercise might also converge on a common effector (39). This would explain the circumstances where the combination of insulin and exercise show a synergistic effect on glucose transport (5, 6). In conclusion we suggest that phosphorylation of IRS-1 on Ser-789 by AMPK might represent a novel mechanism for insulin-responsive cells to react to changes in cellular energy status. This could be by directing signaling through IRS-1 toward energy-restoring processes or via an overall potentiating effect on insulin signaling mediated by an increased activity of the PI 3-kinase associated with IRS-1. Exercise could be the effect that initiates these mechanisms in muscle. We believe this is the first report that identifies a specific serine phosphorylation on IRS-1 that may positively affect insulin signaling. Acknowledgments We thank Dr. Erica Nishimura for constructive suggestions and comments on this manuscript. We also thank Dr. Jon Whitehead and Professor David E. James for introducing Søren Nyboe Jakobsen to the subcellular fractionation protocol. REFERENCES 1. Nesher, R., Karl, I. E., and Kipnis, D. M. (1985) Am. J. Physiol. 249, C226 C Ploug, T., Galbo, H., Vinten, J., Jorgensen, M., and Richter, E. A. (1987) Am. J. Physiol. 253, E12 E20 3. Shepherd, P. R., Withers, D. J., and Siddle, K. (1998) Biochem. J. 333, Ploug, T., Galbo, H., and Richter, E. A. (1984) Am. J. Physiol. 247, E726 E Vranic, M., Kawamori, R., Pek, S., Kovacevic, N., and Wrenshall, G. A. (1976) J. Clin. Invest. 57, DeFronzo, R. A., Ferrannini, E., Sato, Y., Felig, P., and Wahren, J. (1981) J. Clin. Invest. 68, Winder, W. W., and Hardie, D. G. (1996) Am. J. Physiol. 270, E299 E Vavvas, D., Apazidis, A., Saha, A. K., Gamble, J., Patel, A., Kemp, B. E., Witters, L. A., and Ruderman, N. B. (1997) J. Biol. Chem. 272, Corton, J. M., Gillespie, J. G., Hawley, S. A., and Hardie, D. G. (1995) Eur. J. Biochem. 229, Merrill, G. F., Kurth, E. J., Hardie, D. G., and Winder, W. W. (1997) Am. J. Physiol. 273, E1107 E Hayashi, T., Hirshman, M. F., Kurth, E. J., Winder, W. W., and Goodyear, L. J. (1998) Diabetes 47, Hardie, D. G., and Carling, D. (1997) Eur. J. Biochem. 246, Shoelson, S. E., Chatterjee, S., Chaudhuri, M., and White, M. F. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, Yenush, L., and White, M. F. (1997) Bioessays 19, Sun, X. J., Rothenberg, P., Kahn, C. R., Backer, J. M., Araki, E., Wilden, P. A., Cahill, D. A., Goldstein, B. J., and White, M. F. (1991) Nature 352, Tanasijevic, M. J., Myers, M. G., Jr., Thoma, R. S., Crimmins, D. L., White, M. F., and Sacks, D. B. (1993) J. Biol. Chem. 268, De Fea, K., and Roth, R. A. (1997) J. Biol. Chem. 272, Eldar-Finkelman, H., and Krebs, E. G. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, Li, J., DeFea, K., and Roth, R. A. (1999) J. Biol. Chem. 274, Ravichandran, L. V., Esposito, D. L., Chen, J., and Quon, M. J. (2001) J. Biol. Chem. 276, Aguirre, V., Uchida, T., Yenush, L., Davis, R., and White, M. F. (2000) J. Biol. Chem. 275, Sugden, C., Crawford, R. M., Halford, N. G., and Hardie, D. G. (1999) Plant J. 19, Woods, A., Salt, I., Scott, J., Hardie, D. G., and Carling, D. (1996) FEBS Lett. 397, Hawley, S. A., Davison, M., Woods, A., Davies, S. P., Beri, R. K., Carling, D., and Hardie, D. G. (1996) J. Biol. Chem. 271, Clark, S. F., Martin, S., Carozzi, A. J., Hill, M. M., and James, D. E. (1998) J. Cell Biol. 140, Carling, D., Clarke, P. R., Zammit, V. A., and Hardie, D. G. (1989) Eur. J. Biochem. 186, Boyle, W. J., van der, G. P., and Hunter, T. (1991) Methods Enzymol. 201, Scheper, G. C., Morrice, N. A., Kleijn, M., and Proud, C. G. (2001) Mol. Cell. Biol. 21, Stokoe, D., Campbell, D. G., Nakielny, S., Hidaka, H., Leevers, S. J., Marshall, C., and Cohen, P. (1992) EMBO J. 11, Ridderstrale, M., Degerman, E., and Tornqvist, H. (1995) J. Biol. Chem. 270, Dale, S., Wilson, W. A., Edelman, A. M., and Hardie, D. G. (1995) FEBS Lett. 361, Devereux, J., Haeberli, P., and Smithies, O. (1984) Nucleic Acids Res. 12, Kim, Y. B., Nikoulina, S. E., Ciaraldi, T. P., Henry, R. R., and Kahn, B. B. (1999) J. Clin. Invest. 104, Ceresa, B. P., Kao, A. W., Santeler, S. R., and Pessin, J. E. (1998) Mol. Cell. Biol. 18, Mu, J., Brozinick, J. T., Valladares, O., Bucan, M., and Birnbaum, M. J. (2001) Mol. Cell 7, Whitehead, J. P., Soos, M. A., Aslesen, R., O Rahilly, S., and Jensen, J. (2000) Biochem. J. 349, Wojtaszewski, J. F., Hansen, B. F., Kiens, B., and Richter, E. A. (1997) Diabetes 46, Lund, S., Holman, G. D., Schmitz, O., and Pedersen, O. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, Wojtaszewski, J. F., Hansen, B. F., Urso, B., and Richter, E. A. (1996) J. Appl. Physiol. 81,