|Home | About | Journals | Submit | Contact Us | Français|
The function of insulin receptor substrate-1 (IRS-1) is regulated by both tyrosine and serine/threonine phosphorylation. Phosphorylation of some serine/threonine residues in IRS-1 dampens insulin signaling, whereas phosphorylation of other serine/threonine residues enhances insulin signaling. Phosphorylation of human IRS-1 at Ser629 was increased by insulin in Chinese hamster ovary cells expressing the insulin receptor (1.26 ± 0.09-fold; P < 0.05) and L6 cells (1.35 ± 0.29-fold; P < 0.05) expressing human IRS-1. Sequence analysis surrounding Ser629 revealed conformity to the consensus phosphorylation sequence recognized by Akt. Phosphorylation of IRS-1 at Ser629 in cells was decreased upon treatment with either an Akt inhibitor or by coexpression with kinase dead Akt, whereas Ser629 phosphorylation was increased by coexpression with constitutively active Akt. In addition, Ser629 of IRS-1 is directly phosphorylated by Akt in vitro. In cells, preventing phosphorylation of Ser629 by a Ser629Ala mutation resulted in increased phosphorylation of Ser636, a known negative regulator of IRS-1, without affecting phosphorylation of Tyr632 or Ser616. Cells expressing the Ser629Ala mutation, along with increased Ser636 phosphorylation, had decreased insulin-stimulated association of the p85 regulatory subunit of phosphatidylinositol 3′-kinase with IRS-1 and decreased phosphorylation of Akt at Ser473. Finally, in vitro phosphorylation of a Ser629-containing IRS-1 fragment with Akt reduces the subsequent ability of ERK to phosphorylate Ser636/639. These results suggest that a feed-forward mechanism may exist whereby insulin activation of Akt leads to phosphorylation of IRS-1 at Ser629, resulting in decreased phosphorylation of IRS-1 at Ser636 and enhanced downstream signaling. Understanding the complex phosphorylation patterns of IRS-1 is crucial to elucidating the factors contributing to insulin resistance and, ultimately, the pathogenesis of type 2 diabetes.
Many proteins are regulated by tyrosine phosphorylation, and others are regulated by phosphorylation on serine and threonine residues. Insulin receptor substrate-1 (IRS-1), a central component of both insulin and IGF-I signaling, is tyrosine phosphorylated by the insulin or IGF-I receptor in response to hormone activation of the receptor. The phosphotyrosine residues that reside in pYXXM motifs serve as recognition sites for proteins that contain Src homology 2 (SH2) domains, such as the p85 regulatory subunit of phosphatidylinositol (PI) 3′-kinase (1). Besides being tyrosine phosphorylated, IRS-1 also is phosphorylated heavily on serine and threonine residues. Serine and threonine residues account for 242 of the 1242 total amino acids in the IRS-1 sequence, about 20% total. A number of specific serine and threonine phosphorylation sites have been identified (2–16), and some have the ability either to decrease (2, 12) or enhance (10) insulin signaling. Some sites may function either as positive or negative modulators of insulin signaling, depending upon the particular experimental conditions (17).
Ser312 (Ser307 in the rat and mouse IRS-1 sequence) has been the focus of considerable research as a negative regulator of IRS-1 function (2, 4, 9, 18). This site originally was identified as a potential c-jun N-terminal kinase phosphorylation site that decreases association of IRS-1 with the insulin receptor (IR) (2), likely due to the proximity of Ser312 to the phosphotyrosine binding domain of IRS-1 (amino acid residues 160–263). More recently, Ser312 has been implicated in inflammation and lipid-induced insulin resistance (19). On the other hand, Ser1223, which was identified using in vitro kinase assays and mass spectrometry analysis, appears to have a positive role in IRS-1 function. A mutation of Ser1223 to alanine increases association of the tyrosine phosphatase SH2-containing phosphatase-2 (SHP-2) with IRS-1 and decreases insulin-stimulated IRS-1 tyrosine phosphorylation (10). Proximity of phosphorylated Ser1223 to the SH2 recognition motifs responsible for association of SHP-2 with IRS-1 may interfere with this binding event.
Using mass spectrometry analysis of recombinant human IRS-1 (hIRS-1) proteins phosphorylated in vitro with various serine kinases, we recently reported a number of novel phosphorylation sites in this protein (16). Among the sites phosphorylated by protein kinase A (PKA) was Ser629. Ser629, which is present in human but not rodent IRS-1, lies in a critical region of IRS-1, with the sequence 627KGSGDYMPMSPK638. This peptide contains a phosphorylated pYMXM motif (Tyr632) important for association of PI 3′-kinase with IRS-1, as well as the phosphorylation site Ser636, which resides in a PXpSP proline-directed kinase motif. Phosphorylation of Ser636 has had negative effects on downstream insulin signaling (12, 20); it may be that steric factors are involved in the negative effects of Ser636. In light of the proximity of Ser629 to this critical region of IRS-1 containing a recognition site for p85 (Tyr632) and another regulatory site, Ser636, we hypothesized that phosphorylation of Ser629 might also be involved in regulating the function of this region of IRS-1. In this study we report that Ser629 is phosphorylated both in vivo in human muscle and in cells, and that Ser629 is phosphorylated by Akt in vitro. Phosphorylation of IRS-1 at Ser629 in cells is decreased upon treatment with either an Akt inhibitor or by coexpression with kinase dead (KD) Akt, whereas Ser629 phosphorylation is increased by coexpression with constitutively active (CA) Akt. In cell culture, phosphorylation of Ser629 is increased by insulin, and expression of a Ser629Ala IRS-1 mutant results in increased phosphorylation of Ser636, a known negative regulator of IRS-1, and decreased insulin-stimulated association of the p85 regulatory subunit of PI 3′-kinase with IRS-1 and decreased phosphorylation of Akt at Ser473.
Finally, phosphorylation at Ser629 decreases phosphorylation of Ser636 in vitro, suggesting that insulin activation of Akt leads to phosphorylation of Ser629 of IRS-1, resulting in decreased phosphorylation of IRS-1 at Ser636 and enhanced downstream signaling.
A lean, healthy subject with normal glucose tolerance (75 g oral glucose tolerance test) underwent a percutaneous biopsy of the vastus lateralis muscle under local anesthesia, under basal conditions, as previously described (21). The protocol was approved by the Arizona State University Institutional Review Board, and informed written consent was obtained from the volunteer.
The Chinese hamster ovary (CHO) cell line overexpressing the human IR was a gift from Dr. Feng Liu (Pharmacology, The University of Texas Health Science Center at San Antonio, San Antonio, TX). The L6 myoblast cell line was a gift from Dr. Amira Klip (Programme in Cell Biology, The Hospital for Sick Children, Toronto, Ontario, Canada). The cDNA encoding full-length wild-type hIRS-1 was a gift from Dr. C. Ronald Kahn (Department of Medicine, Joslin Diabetes Center, Boston, MA). The cDNA encoding wild-type, CA, and KD Akt were gifts from Dr. Feng Liu.
Anti-HA.11 was purchased from Covance (Berkeley, CA). Anti-Akt, anti-phospho-Akt (Ser473) anti-phospho-IRS-1 (Ser636/639), anti phospho-IRS-1 (Ser612) (Ser616 in human sequence), and anti-phospho-Akt substrate antibodies were purchased from Cell Signaling Technology (Beverly, MA). Anti-phospho-IRS-1 (Tyr632) and anti-myc antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), whereas anti-IRS-1 and anti-p85 antibodies were purchased from Upstate (Charlottesville, VA). Akt Inhibitor III was purchased from CALBIOCHEM (EMD Biosciences, Inc, San Diego, CA). Active recombinant Akt-1 and ERK-2 were purchased from Upstate.
The cDNA encoding full-length wild-type hIRS-1 (pBEX-hIRS-1) was described previously (10). Ser629 and Ser636 in hIRS-1 were mutated to Ala (pBEX-hIRS-1-S629A and pBEX-hIRS-1-S636A) with the Quick-Change Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA) according to the manufacturer’s instructions. For pBEX-hIRS-1-S629A, the sense primer was 5′-CCCAGTGGCCGAAAGGGCGCTGGAGACTATATGCCC-3′, and antisense primer was 5′-GGGCATATAGTCTCCAGCGCCCTTTCGGCCACTGGG-3′. For pBEX-hIRS-1-S636A, the sense primer was 5′-GACTATATGCCCATGGCCCCCAAGAGCGTATCTG-3′, and antisense primer was 5′-CAGATACGCTCTTGGGGGCCATGGGCATATAGTC-3′. The plasmid and mutant sites were verified by restriction digestion and sequencing. Adenoviruses encoding green fluorescence protein (as an adenovirus control) and hIRS-1 (Ad-hIRS-1) were produced as previously described (10). The cDNA encoding the hIRS-1539–706 fragment was generated by PCR using hIRS-1 cDNA (courtesy of Dr. C. Ronald Kahn) as a template. Sac I and Xba I directed sequences were introduced to allow the in-frame insertion of the cDNA into the His-tagged and myc-tagged prokaryotic expression plasmid pTrcHis2 (Invitrogen, Carlsbad, CA). All recombinant plasmid constructs were verified by restriction mapping and DNA sequencing.
CHO/IR cells were grown in Ham’s F-12 medium (Life Technologies, Inc., Invitrogen) and L6 cells in αMEM (Life Technologies, Inc., Invitrogen), supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. Transfections of cells were performed in 60-mm plates with 5 µg of each recombinant plasmid, using Lipofectamine (Invitrogen) according to the manufacturer’s protocol. In some experiments, cells were transduced with an adenovirus encoding full-length, HA-tagged, wild-type hIRS-1 (10). Twenty-four hours after transfection/after transduction, cells were serum starved for 2–4 h and then treated with or without inhibitors before being exposed to 100 nm insulin for the assigned time. Cells were then washed three times with ice-cold PBS and lysed in 300–400 µl lysis buffer [50 mm HEPES (pH 7.6), 150 mm Na CI, 1% Triton X-100, 10 mm NaF, 20 mm sodium pyrophosphate, 20 mm β-glycerol phosphate, 1 mm sodium orthovanadate, 10 µg/ml leupeptin, 10 µg/ml aprotinin, 1 µm microcystin-LR, and 1 mm phenylmethylsulfonyl fluoride]. Cell lysates were centrifuged (10,000 × g, 4 C, 10 min), and the clarified supernatants were used for immunoprecipitation or Western blot experiments. Immunoprecipitation and Western blot analyses were performed as described previously (10).
GST-hIRS-1524–698 fusion proteins were expressed as described previously (16). In brief, XL1-blue cells containing plasmids encoding for recombinant GST-hIRS-1524–698 fusion protein or DH5α cells containing plasmids encoding for hIRS-1539–706 fusion proteins were grown in Luria Broth (LB) medium at 37 C overnight. This culture was diluted 1:10 and grown at 30 C for 3 h. Expression of the fusion proteins was induced by the addition of isopropyl β-d-thiogalactoside to a final concentration of 0.2 mm. After 4-h induction, the cells were harvested by centrifugation at 5000 × g for 10 min, washed with 10 mmTris-HCl (pH 8.0), suspended in bacterial lysis buffer containing 50 mm Tris-HCl (pH 7.5), 50 mm KCl, 1 mm dithiothreitol, 5 mm EDTA, 1 mm phenylmethane sulfonyl fluoride, 0.1% (vol/vol) Triton X-100, and 1 mg/ml lysozyme, and lysed by sonication. Cell lysates were clarified by centrifugation at 12,000 × g for 15 min. GST-IRS-1524–698 fusion proteins were purified by affinity chromatography with glutathione-sepharose beads. The His-myc-hIRS-1539–706 fusion proteins were purified using Ni-NTA Agarose beads (Invitrogen) following the protocol provided by Invitrogen except that the elution step was omitted so that the His-myc-tagged hIRS-1 would stay bound to the beads.
hIRS-1539–706 fragment proteins adsorbed on the Ni-NTA agarose beads were phosphorylated by in vitro treatment with kinases, including Akt or ERK2, at 30 C for 30 min, using a final volume of 40 µl kinase buffer [20 mm HEPES (pH 7.5), 100 mm NaCl, 10 mm MgCl2, and 15 µm cold ATP]. After extensive washing with PBS, proteins were eluted by boiling at 95 C for 4 min before being subjected to SDS-PAGE. GST-IRS-1524–698 fusion proteins were phosphorylated by in vitro treatment with the Akt-1 (Upstate) at 30 C for 30 min, using assay dilution buffer, magnesium/ATP cocktail (Upstate), and [γ-32P]ATP (PerkinElmer, Boston, MA). The reactions were terminated by addition of SDS-PAGE sample loading buffer and heated at 95 C for 4 min. The proteins were separated on 10% SDS-PAGE gels, stained with Coomassie blue, and phosphorylation was visualized by autoradiography. For mass spectrometry analysis, the in vitro kinase assay was performed with unlabeled ATP instead of [γ-32P]ATP.
Approximately 100 mg (wet weight) human vastus lateralis muscle was homogenized in detergent lysis buffer containing protease and phosphatase inhibitors (21). IRS-1 was immunoprecipitated from homogenate containing 1 mg Lowry protein, the immunoprecipitated proteins were then resolved on 10% SDS-PAGE gels, and the gels were stained with Coomassie blue. The region of the gel containing IRS-1 was excised; no protein band was visible for IRS-1, but the position of IRS-1 was established in preliminary experiments using immunoblot analysis and confirmed by mass spectrometry. Mass spectrometry analysis was performed as described under Mass spectrometry analysis.
Mass spectrometry analysis was performed as described previously (10, 16). Briefly, IRS-1 was immunoprecipitated from either human skeletal muscle homogenates, CHO/IR cell lysates, or L6 cell lysates, and the proteins were resolved by 1-DSDS-PAGE. The IRS-1 bands were excised, digested in situ with trypsin, and the resulting peptides were analyzed by HPLC-electrospray ionization tandem mass spectrometry (HPLC-ESI/MS/MS). HPLC-ESI/MS/MS was performed on a Thermo Finnigan LTQ-FT/ICR (San Jose, CA), which has been adapted for nanospray ionization. On-line HPLC separation of the IRS-1 digests was accomplished with a Michrom BioResources MAGIC 2002 micro HPLC (Michrom BioResources, Inc., Auburn, CA), using a PicoFrit column (75 µm inner diameter; New Objective, Inc., Woburn, MA) packed to 10 cm with C18 adsorbent (218MS5 µm, 300 Å; Vydac, Hesperia, CA); mobile phase, linear gradient of 2–65% acetonitrile in 0.5% acetic acid/0.005% trifluoroacetic acid followed by a hold at 65% acetonitrile and then a step to 80% acetonitrile; flow rate, 0.4 µl/min. Programs used for data analysis include SEQUEST (Thermo Finnigan), Mascot (Matarix Science, London, UK), ProteinProspector (University of California, San Francisco MS facility web site), and GPMAW (Lighthouse Data, Odense, Denmark).
Lysates of CHO/IR and L6 myoblasts ectopically expressing HA-tagged hIRS-1 were immunoprecipitated with either anti-HA or anti-IRS-1 antibodies. Immunoprecipitated proteins were resolved by SDS-PAGE, and the area of the gel corresponding to IRS-1 was analyzed by mass spectrometry. Using a hypothesis driven approach (16, 22), a peptide monophosphorylated at Ser629 was identified in both cell types; a typical collision-induced dissociation (CID) spectrum from hIRS-1 expressed in L6 cells is shown in Fig. 1A. Ser629 is present in human, primate, and pig IRS-1, but the corresponding residue in rat, mouse, and chicken IRS-1 is the hydrophilic amino acid asparagine (Fig. 1B).
To determine whether Ser629 is an authentic phosphorylation site in human muscle in vivo, a muscle biopsy taken under basal conditions (no insulin stimulation) was homogenized, the IRS-1 immunoprecipitated, and analyzed by mass spectrometry from tryptic digests. IRS-1 was easily observed under these conditions, with 40–60% sequence coverage. Using a targeted analysis designed to detect Ser629 based on the chromatographic behavior of the tryptic peptide containing this site and its ionization characteristics, we identified the Ser629 phosphopeptide. MS/MS analysis localized the site of phosphorylation to Ser629 (Fig. 1C).
To determine whether insulin regulates phosphorylation of Ser629, CHO/IR and L6 cells overexpressing hIRS-1 were incubated with or without 100 nm insulin for 15 min. Ser629 phosphorylation in insulin-treated cells was compared with that in untreated cells. Quantification by mass spectrometry analysis showed that insulin modestly, but significantly, increased phosphorylation of Ser629 in CHO/IR cells (1.26 ± 0.09-fold; n = 5; P < 0.05, paired t test) and L6 cells (1.35 ± 0.29-fold; n = 5,;P < 0.05, paired t test).
Ser629 of hIRS-1 was initially identified as a PKA phosphorylation site in vitro (16). However, the observation that Ser629 phosphorylation is increased by insulin argues against PKA being the only kinase for Ser629 phosphorylation because PKA activity is not likely to be increased by insulin. Upon inspection, it was noted that the amino acid sequence surrounding Ser629 (626RKGpSGDYMPMSPK638) bore a modest resemblance to a minimal Akt phosphorylation consensus motif, so we hypothesized that Akt could be an insulin-stimulated Ser629 kinase. To test this hypothesis, first we assessed whether Akt was capable of phosphorylating Ser629 in vitro. GST-hIRS-1524–698 was expressed in Escherichia coli, and purified GST-hIRS-1524–698 was phosphorylated in vitro by recombinant Akt-1 in the presence of γ32P-ATP. The phosphorylated IRS-1 was resolved by SDS-PAGE, stained with Coomassie blue, and the gel was subjected to autoradiography. Figure 2A shows an autoradiograph confirming that Akt efficiently phosphorylates IRS-1524–698, the sequence that encompasses Ser629. Meanwhile, to establish whether Ser629 itself is phosphorylated by Akt in vitro, phosphorylated GST-hIRS-1524–698 was resolved by SDS-PAGE, excised, and digested by trypsin in situ. The phosphopeptide (R)KGpSGDYMPMSPK was analyzed by HPLC-ESI-MS/MS. Mass spectrometry analysis of the tryptic peptides derived from the recombinant, phosphorylated IRS-1 protein confirmed that Ser629 is phosphorylated by Akt in vitro (spectrum not shown).
To determine whether Akt can phosphorylate Ser629 in response to insulin in cells, CHO/IR cells were transfected with wild-type pBEX-hIRS-1 for 24 h. The transfected cells were fasted for 2 h and incubated with or without 15 µm Akt inhibitor III for 2 h before being exposed to 100 nm insulin for 15 min. IRS-1 was immunoprecipitated from cell lysates, and the immunoprecipitated IRS-1 was immunoblotted with an anti-phospho-Akt substrate antibody, which detects phosphorylated amino acid sequence motifs for Akt. Figure 2B shows that phosphorylation of Akt substrate sites in hIRS-1 is stimulated by insulin, and this insulin stimulation decreases upon incubation of the cells with Akt inhibitor. Although this experiment shows the effectiveness of the Akt inhibitor in reducing insulin-stimulated phosphorylation of these sites, it is not specific. To determine whether the Akt inhibitor specifically inhibited insulin stimulation of Ser629 phosphorylation, immunoprecipitated IRS-1 from the same experiments used to obtain the results in Fig. 2B was resolved by SDS-PAGE, the band containing IRS-1 was excised from the gel, and the proteins digested in situ with trypsin. The resulting peptides were analyzed and quantified by HPLC-ESI/MS/MS. Figure 2C shows that insulin-stimulated phosphorylation of Ser629 specifically was inhibited by preincubation with Akt inhibitor III, suggesting that Akt is a Ser629 kinase in cells. To confirm further a role for Akt as a kinase capable of phosphorylating IRS-1 at Ser629 in cells, CHO/IR cells were transfected with pcDNA-Akt, pcDNA-CA Akt, or pcDNA-KD Akt for 24 h, followed by infection with Ad-hIRS-1 for another 24 h. Cells were lysed, and cell lysates were immunoprecipitated with anti-HA antibody for IRS-1. Immunoprecipitates of IRS-1 were resolved by SDS-PAGE and subjected to mass spectrometry quantification for Ser629 phosphorylation. Figure 2D shows that CA-Akt, compared with wild-type Akt, increases Ser629 phosphorylation, whereas KD-Akt reverses the phosphorylation of Ser629 observed in the presence of CA-Akt, further confirming that Akt acts as a Ser629 kinase in cells.
Inspection of the sequence surrounding Ser629 suggests that phosphorylation of Ser629, because of its proximity to two other important phosphorylation sites (Tyr632 and Ser636), might influence the phosphorylation or function of these two sites. To address this, site-directed mutagenesis was used to generate HA-tagged S629A and S636A mutant IRS-1 proteins. HA-tagged pBEX-hIRS-1, pBEX-hIRS-1 S629A, and pBEX-hIRS-1 S636A were transfected into CHO/IR cells for 24 h. Cells were serum starved for 4 h before being exposed to 100 nm insulin for 15 min. The immunoprecipitated proteins were analyzed for phosphorylation at Ser636, Tyr632, and Ser616 by immunoblotting with site-specific antibodies, and the results are shown in Fig. 3. In wild-type IRS-1, insulin stimulated the phosphorylation of Ser636 and Tyr632. As expected, the S636A mutant showed little or no immunoreactivity with the anti-phospho-Ser636/639 antibody (Fig. 3, A and B). When Ser629 was mutated to alanine to prevent phosphorylation at this site, both basal and insulin insulin-stimulated phosphorylation at Ser636 were increased. Quantification of these results is shown in Fig. 3B. In contrast, phosphorylation at neither Tyr632 nor Ser616 was affected by the S629A mutation (Fig. 3, C and D, respectively), indicating the specificity of this phenomenon.
To gain more insight into the effect of Ser629 phosphorylation on Ser636 phosphorylation, time course experiments were performed in the presence of insulin using wild-type and S629A mutant IRS-1 proteins expressed in CHO/IR cells. Cells were serum starved for 2 h before being exposed to 100 nm insulin for 0, 1, 5, 15, 30, 60, and 120 min. Cell lysates were immunoprecipitated using anti-HA antibody, and immunoprecipitates were subjected to immunoblots for Ser636 phosphorylation. The results are shown in Fig. 4A. Phosphorylation of Ser636 in wild-type IRS-1 increased by 15 min reached a maximum at 30 min and returned toward basal values thereafter. In IRS-1 in which phosphorylation at Ser629 was prevented by the S629A mutation, Ser636 phosphorylation increased more rapidly, peaking by 15 min, before returning to baseline. Quantification of the results of three such independent experiments is shown in Fig. 4B, demonstrating the leftward shift in Ser636 phosphorylation caused by the S629A mutation.
In light of data suggesting Ser636 phosphorylation negatively regulates insulin signaling (12), we hypothesized that because the S629A mutant possesses increased Ser636 phosphorylation, the S629A mutation in IRS-1 would decrease downstream insulin signaling. This hypothesis, if confirmed, would provide evidence that phosphorylation of Ser629 positively modulates downstream insulin signaling. To test this, we assessed association of the p85 regulatory subunit of PI 3′-kinase with wild-type and S629A mutant IRS-1. CHO/IR cells were transfected with wild-type pBEX-hIRS-1 or mutant pBEX-hIRS-1S629A. Twenty-four hours after transfection, cells were serum starved for 4 h before being exposed to 100 nm insulin for 15 min. Cells were lysed, and cell lysates were immunoprecipitated with an anti-p85 antibody. The immunoprecipitates were analyzed for coprecipitated hIRS-1 by immunoblotting. Figure 5A shows the results of a typical experiment, and Fig. 5B is the quantification data for five independent experiments. The results demonstrate that prevention of Ser629 phosphorylation decreases insulin-stimulated association of p85 with IRS-1, under the same conditions that increase Ser636 phosphorylation. To analyze further downstream insulin signaling, we tested the ability of insulin to increase Akt phosphorylation at Ser473 in CHO/IR cells ectopically expressing either wild-type or S629A mutant IRS-1. As can be seen in Fig. 6, A and B, as was the case for inhibited p85 association with IRS-1, cells expressing the S629A mutant showed significantly decreased Akt phosphorylation, again supporting the notion that phosphorylation at Ser629 positively regulates insulin signaling.
Together, the aforementioned data suggest that by phosphorylating Ser629, Akt can enhance downstream insulin signaling by enhancing association of the p85 regulatory subunit of PI 3′-kinase with IRS-1, and this effect may be mediated by regulating the level of Ser636 phosphorylation. To test whether phosphorylation of IRS-1 in this region by Akt directly interferes with the subsequent phosphorylation of Ser636, a His-tagged fragment of IRS-1 containing Ser629 and Ser636 (hIRS-1539–706) was expressed in E. coli and purified using nickel beads. hIRS-1539–706, still coupled to the beads, was subjected to an in vitro kinase assay with Akt or no added kinase as a control. After washing to remove Akt, a second kinase reaction was performed using ERK as the kinase. The hIRS-1539–706 was resolved by SDS-PAGE, transferred to nitrocellulose membranes, and probed with an antibody directed against phospho-Ser636/639. The results of three independent experiments showed that phosphorylation of this region of IRS-1 with Akt, under conditions known to phosphorylate Ser629 (see Fig. 2A), reduced subsequent phosphorylation of Ser636/639 by ERK (Fig. 7, A and B). This finding supports the hypothesis that Ser629 phosphorylation is involved in regulating the level of Ser636 phosphorylation.
Although it has long been known that tyrosine phosphorylation of IRS-1 is required for the downstream transmission of the insulin signal, serine and threonine phosphorylation of IRS-1 is emerging as an important regulatory feature of insulin action. A large number of phosphorylation sites have been identified in IRS-1, and many of these sites are reported to affect the function of this protein (2–20). Some sites appear likely to be negative regulators of IRS-1 signaling. For example, Ser312 (Ser307 in the rodent IRS-1 sequence) is phosphorylated in response to inflammatory stimuli that activate kinases such as c-jun N-terminal kinase (JNK)and inhibitor κB kinase (IKK) (2, 9, 14, 19, 23). Ser312 lies in close proximity to the phosphotyrosine binding domain of IRS-1, and evidence from yeast trihybrid experiments indicates that phosphorylation at this site interferes with the interaction between IRS-1 and the IR (2). Furthermore, some evidence supports the notion that excessive phosphorylation at Ser312 is important for lipid-induced insulin resistance (19). Other sites appear to be positive regulators of IRS-1 function. In this regard, Ser1223, which we identified in a mass spectrometry screening, is involved in regulation of SHP-2-mediated dephosphorylation of tyrosine residues of IRS-1 (10). Ser1223 lies in proximity to an SH2 recognition motif used by SHP-2 to associate with IRS-1 (24). When Ser1223 is phosphorylated, interaction between SHP-2 and IRS-1 is decreased, and IRS-1 tyrosine phosphorylation increases (10).
Ser636 is a well-recognized phosphorylation site in IRS-1, and the preponderance of evidence suggests that Ser636 phosphorylation can negatively influence insulin signaling (12, 20). Ser636 lies in the sequence 627KGSGDYMPMSPK638. Tyr632, when phosphorylated, serves as an SH2 recognition sequence (pYMXM) recognized by an SH2 domain of the p85 regulatory subunit of PI 3′-kinase. Ser636 lies in a classic proline-directed kinase consensus motif (PXpSP), and evidence suggests that when it is phosphorylated, it interferes with the association of p85 with IRS-1. In mouse IRS-1, this sequence is 623KGNGDYMPMSPK634, with an asparagine residue in the homologous position to Ser629 in the hIRS-1 sequence. Initially, we used mass spectrometry analysis to identify Ser629 of hIRS-1 as a site phosphorylated by PKA in vitro (16). The proximity of this serine residue to Tyr632 and Ser636 suggested that Ser629 also could be involved in regulating the function of this important p85 binding region. In addition, the fact that it is present only in human (and other primate) IRS-1 suggests that regulation of IRS-1 function by this site represents a site-specific regulation in this region that would be critical to analyze in humans. Moreover, the fact that we observed Ser629 to be phosphorylated in vivo in human muscle lends additional importance to these findings, and enhances the importance of studying the regulation of phosphorylation of this site under physiological and pathophysiological conditions in humans.
Having ascertained that Ser629 of IRS-1 is phosphorylated in vivo in human muscle, the question of what kinase or kinases phosphorylate this site gains importance. Although PKA is capable of phosphorylating Ser629 in vitro (16), it is unlikely to be the kinase responsible for the insulin-stimulated increase in phosphorylation at this site because PKA activity is not thought to be stimulated by insulin. Moreover, PKA action is more generally considered to be in opposition to insulin action. Because the amino acid sequence surrounding Ser629 bore some resemblance to an Akt consensus phosphorylation motif, and because Akt activity is increased by insulin, we considered Akt as a candidate kinase for this site. Several approaches were used to determine if this site is a bona fide Akt site. First, mass spectrometry analysis of in vitro kinase assays using Akt and an IRS-1 substrate confirmed that Akt is capable of directly phosphorylating IRS-1 at Ser629. Second, phosphorylation of IRS-1 at Ser629 in cells was decreased upon treatment with either an Akt inhibitor or by coexpression with KD Akt, whereas Ser629 phosphorylation was increased by coexpression with CA Akt.
The findings of the present study support a novel role for the phosphorylation of Ser629 in regulating the function of hIRS-1. Prevention of phosphorylation at this site using a S629A mutation markedly increased phosphorylation at Ser636/639 in cells. Time course studies suggest that Ser629 phosphorylation delays Ser636 phosphorylation. This indicates that phosphorylation events at Ser629 and Ser636 play sequential roles, possibly by first enhancing downstream insulin signaling (Ser629) and then desensitizing insulin signaling (Ser636). Therefore, it is feasible that decreasing phosphorylation of IRS-1 at Ser629 leads to earlier and increased phosphorylation at Ser636, which in turn contributes to insulin resistance. In fact, phosphorylation at Ser636 is increased in primary muscle cell cultures from insulin-resistant subjects, leading to decreased activation of PI 3′-kinase (20). Because phosphorylation of Ser636 has a negative impact on insulin signaling, we reasoned that these results suggest that phosphorylation of Ser629, by dampening phosphorylation at Ser636, should have a positive effect on insulin signaling. We found that this prediction was upheld because IRS-1 with the S629A mutation not only had increased Ser636/639 phosphorylation but also decreased association of the p85 regulatory subunit of PI 3′-kinase with IRS-1. Moreover, insulin-stimulated Akt phosphorylation also was decreased under these conditions.
Finally, phosphorylating hIRS-1539–706 with Akt inhibited the subsequent ability of ERK to phosphorylate Ser636/639, as assessed by immunoblot analysis. The latter observation indicates that phosphorylation of IRS-1 by Akt directly interferes with subsequent phosphorylation of Ser636 by ERK or, by implication, other kinases that may serve as inhibitors of IRS-1 function by phosphorylating Ser636 (25), which is known to inhibit downstream insulin signaling (12, 20).
The results of this study show that Akt phosphorylates Ser629 in vitro, that most, if not all, of the insulin-stimulated increase in Ser629 phosphorylation can be attributed to Akt in cells, and that phosphorylation of this region of IRS-1 by Akt interferes with phosphorylation of Ser636. Together, these data suggest the model for interactions in this region of IRS-1 shown in Fig. 8. According to this model, after insulin-induced phosphorylation of Tyr632 and subsequent activation of PI 3′-kinase, Akt is phosphorylated and activated rapidly. Active Akt then phosphorylates IRS-1 at Ser629. This delays the activation of a negative feedback loop involving IRS-1 phosphorylation at Ser636 by some unknown mechanism, perhaps steric hindrance, thus lessening the dampening effect of Ser636 phosphorylation on Tyr632 function. Therefore, phosphorylation of IRS-1 at Ser629 enhances and sustains insulin signaling.
The results of these studies lead us to propose that insulin-induced phosphorylation at Ser629 represents a feed-forward mechanism to enhance insulin action in humans. By extension, factors interfering with phosphorylation at Ser629 could be candidates for factors that cause insulin resistance.
This work was supported by National Institutes of Health Grants R01DK47936 (to L.J.M.) and R01DK66483 (to L.J.M.), and the American Diabetes Association.
Disclosure Statement: The authors have nothing to declare.