Attenuation of Insulin-stimulated Insulin Receptor Substrate-1 Serine 307 Phosphorylation in Insulin Resistance of Type 2 Diabetes *

insulin-stimulated phosphorylation serine about 0.2 n M with a t 1 ⁄ 2 of about 2 min. The amount of IRS1 was similar in cells from non-diabetic and diabetic subjects. These findings identify a molecular mechanism for insulin resistance in non-selected patients with type 2 diabetes. Theincidenceoftype2diabetesisrapidlyincreasinginallpartsoftheworld that allows a sedentary and affluent lifestyle, as the disease is closely associated This is a first demonstration of a critical function of serine phosphorylation of IRS1 in primary human cells that also are target cells for insulin. important, in two separate unselected groups of patients with type 2 diabetes these findings identify a mechanism with the potential to explain the insulin resistance of the disease: a curtailed phosphorylation of serine 307 reduces the steady-state tyrosine phosphorylation of IRS1 at physiological concentrations of insulin. Type 2 diabetes has a strong genetic component, and it therefore serine sensitize IRS1 tyrosine phosphorylation. Site-directed mutagenesis of serine 302 to alanine in murine 32D cells, insulin-stimulated phosphorylation of serine downstream protein kinase activation. Serine consensus sequence(R XX S/T)forphosphorylationbythedownstreamproteinkinase rapamycin phosphorylation mammalian-target-of-rapamycin phosphorylation.

Insulin resistance is a primary characteristic of type 2 diabetes and likely causally related to the pathogenesis of the disease. It is a result of defects in signal transduction from the cell surface receptor of insulin to target effects. We found that insulin-stimulated phosphorylation of serine 307 (corresponding to serine 302 in the murine sequence) in the immediate downstream mediator protein of the insulin receptor, insulin receptor substrate-1 (IRS1), is required for efficient insulin signaling and that this phosphorylation is attenuated in adipocytes from patients with type 2 diabetes. Inhibition of serine 307 phosphorylation by rapamycin mimicked type 2 diabetes and reduced the sensitivity of IRS1 tyrosine phosphorylation in response to insulin, while stimulation of the phosphorylation by okadaic acid, in cells from patients with type 2 diabetes, rescued cells from insulin resistance. EC 50 for insulin-stimulated phosphorylation of serine 307 was about 0.2 nM with a t1 ⁄ 2 of about 2 min. The amount of IRS1 was similar in cells from non-diabetic and diabetic subjects. These findings identify a molecular mechanism for insulin resistance in non-selected patients with type 2 diabetes.
The incidence of type 2 diabetes is rapidly increasing in all parts of the world that allows a sedentary and affluent lifestyle, as the disease is closely associated with obesity. It has been estimated that over 200 million people will be afflicted with the disease by the end of this decade. The pathogenesis of diabetes is not understood in any great detail, but it is generally believed that insulin resistance in skeletal muscle and especially in adipose tissue is an early, if not primary, event. The peripheral insulin resistance is compensated for by increased concentrations of circulating insulin, masking development to diabetes. Eventually the pancreatic ␤-cells fail to compensate for the insulin resistance, and type 2 diabetes can be diagnosed. The insulin resistance of muscle and fat tissues is the result of impaired signal transduction from the cell surface receptor of insulin to metabolic effects, such as to increase glucose transport and inhibit lipolysis in adipocytes (1,2). Epidemiological studies have failed to provide a strong link between the disease in general and genetic polymorphisms in the genes coding for the signaling proteins. The occupied insulin receptor autophosphorylates and then phosphorylates the insulin receptor substrate-1 (IRS1) 2 protein on tyrosine residues to provide docking sites for downstream activation of signal-transducing proteins (1).
Recently, the tyrosine phosphorylation of IRS1 was shown to be the first signaling step to exhibit reduced sensitivity to insulin in adipocytes from patients with type 2 diabetes (3). The finding depended on the realization that adipocytes obtained from human beings become insulin-resistant from the surgical cell isolation procedures and that the resistance is reversed by overnight incubation of the cells (3). This insulin resistance is manifest downstream of IRS1, and if cells from diabetic patients are not allowed to recover overnight the mix of two different types of insulin resistance will be examined. IRS1 has also been identified as a key step in insulin resistance of human skeletal muscle (4 -6). Naturally occurring mutations in the IRS1 protein are present in some subjects with type 2 diabetes (7)(8)(9)(10)(11), and obese individuals and relatives of diabetic patients have been reported to present with lower adipocyte expression of IRS1 (12).
Along a different line of research, in cell-free systems and cell cultures, phosphorylation of IRS1 on serine residues has been found to affect, usually inhibit, insulin receptor-catalyzed phosphorylation/activation of IRS1 on tyrosine residues (13)(14)(15)(16)(17)(18)(19)(20)(21)(22). Specifically phosphorylation of murine serine 307 (corresponding to serine 312 in human IRS1) has been reported to block IRS1 interaction with the insulin receptor (18) or to enhance its proteolytic degradation in the cell (23). This site has also been reported to be phosphorylated in response to tumor necrosis factor-␣ (24) and in animal models of type 2 diabetes (19,25). Recently, phosphorylation of IRS1 on serine 302 in murine 32D cells (corresponding to serine 307 in human IRS1) was reported to be stimulated by insulin and abolishing of this phosphorylation by introduction of an alanine residue in place of serine 302 impaired insulin-stimulated tyrosine-specific phosphorylation of IRS1 (26). However, negative effects by serine 302 phosphorylation on insulin stimulation of IRS1 tyrosine phosphorylation (27) and disruption of insulin receptor-IRS1 interaction using a yeast two-hybrid method (19) have also been reported.
Herein we have examined the effect of IRS1 serine 307 phosphorylation on IRS1 tyrosine phosphorylation in response to insulin and the ability of insulin to induce phosphorylation of serine 307 in primary adipocytes from patients with type 2 diabetes.

EXPERIMENTAL PROCEDURES
Subjects-Subject characteristics are described in the respective figure legends. Informed consent from patients and approval by the Ethics Committee were obtained.
Isolation and Incubation of Adipocytes-Subcutaneous adipose tissue was obtained as described in the figure legends to Figs. 1 and 4. Adipocytes were isolated by collagenase (type 1, Worthington) digestion as described (28). Tissue pieces from biopsies performed with local anesthesia using xylocain were extensively washed in large volumes of 0.15 M NaCl before digestion with collagenase. At a final concentration of 100 l packed cell volume per ml, cells were incubated in Krebs-Ringer solution (0.12 M NaCl, 4.7 mM KCl, 2.5 mM CaCl 2 , 1.2 mM MgSO 4 , 1.2 mM KH 2 PO 4 ) containing 20 mM Hepes, pH 7.40, 1% (w/v) fatty acid-free bovine serum albumin, 100 nM phenylisopropyladenosine, 0.5 unit⅐ml Ϫ1 adenosine deaminase with 2 mM glucose, at 37°C on a shaking water bath. For analysis after 20 -24-h incubation cells were incubated at 37°C, 10% CO 2 in the same solution mixed with an equal volume of Dulbecco's modified Eagle's medium containing 7% (w/v) albumin, 200 nM phenylisopropyladenosine, 20 mM Hepes, 50 UI/ml penicillin, 50 g/ml streptomycin, pH 7.40 (3). Before analysis cells were washed and transferred to the Krebs-Ringer solution.
SDS-PAGE and Immunoblotting-Cell incubations were terminated by separating cells from medium by centrifugation through dinonylphtalate. To minimize postincubation signaling modifications in the cells and protein modifications during immunoprecipitation, the cells were immediately dissolved in SDS and ␤-mercaptoethanol with protease and protein phosphatase inhibitors, frozen within 10 s, and thawed in boiling water (28). Equal amounts of cells as determined by lipocrit, which is total cell volume, was subjected to SDS-PAGE and immunoblotting. After SDS-PAGE and electrotransfer membranes were incubated with the indicated antibodies that were detected using ECLϩ (Amersham Biosciences, Amersham, UK) with horseradish peroxidaseconjugated anti-IgG as secondary antibody and evaluated by chemiluminescence imaging (Las 1000, Image-Gauge, Fuji, Tokyo, Japan).
* This work was supported by the Ö stergö tland County Council, the Swedish Diabetes Association, and the Swedish Research Council. 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.

RESULTS
Insulin-stimulated Phosphorylation of IRS1 Serine 307-In primary subcutaneous adipocytes from a group of non-diabetic subjects insulin stimulated the phosphorylation of IRS1 on serine 307 (Fig. 1A), thus establishing the physiological relevance of this phosphorylation event in primary human cells. However, in adipocytes from an unselected group of obese female (Fig. 1, B and C) or male (Fig.  1D) patients with type 2 diabetes this insulin-induced serine 307 phosphorylation of IRS1 was curtailed, suggesting its involvement in the disease. In these same diabetic patients the insulin-stimulated steady-state phosphorylation of IRS1 on tyrosine was inhibited at submaximal insulin concentrations (cf. Ref. 3; Fig. 2).
Amount of IRS1 in Cells-It has been suggested that the amount of IRS1 protein is reduced in adipocytes from patients with diabetes (12). We found the amount of IRS1 protein to be identical in the adipocytes from the group of patients with type 2 diabetes and the non-diabetic control group (Fig. 1E). However, it can be noted that inter-individual variations were large, thus minor differences in the mean value may not have been revealed.
Dose Response and Time Course for Insulin-stimulated Phosphorylation of Serine 307-Serine 307 was phosphorylated in response to insulin in a dosedependent manner with EC 50 ϭ 0.2 nM in cells from non-diabetic control subjects (Fig. 3A). This is considerably more insulin-sensitive than the insulin receptor autophosphorylation (EC 50 ϭ 1-2 nM) (3) indicating a downstream effect involving enzymatic signal amplification by e.g. additional protein kinases; for comparison see Ref. 3. Time course studies in cells from nondiabetic control subjects demonstrated a half-maximal phosphorylation of serine 307 after about 2 min (Fig. 3B), which was slower than the insulin-induced tyrosine phosphorylation of IRS1, exhibiting half-maximal effect already within 1 min (Fig. 3C).
Phosphorylation of Serine 307 in Cells from Diabetics Makes Cells Insulinsensitive-The phosphoserine/threonine protein phosphatase inhibitor okadaic acid completely inhibits protein phosphatase-2A at 1 nM and protein phosphatase-1 at 1 M concentrations in intact cells, including adipocytes (29). Okadaic acid enhanced the phosphorylation of serine 307 in adipocytes from control subjects and also caused a general increase in the extent of serine/threonine phosphorylation of IRS1 as indicated by a shift in its gel mobility (data not shown). In adipocytes from patients with type 2 diabetes okadaic acid enhanced the phosphorylation of serine 307 on average 1.7 Ϯ 0.4-fold (mean Ϯ S.E., n ϭ 4) and at the same time enhanced the sensitivity to insulin for tyrosine-specific phosphorylation of IRS1 (Fig. 4). It is interesting that okadaic acid has been reported to enhance FIGURE 1. Insulin-induced phosphorylation of IRS1 serine 307 in primary adipocytes from non-diabetic subjects and from patients with type 2 diabetes. Abdominal subcutaneous adipose tissue was obtained during elective surgery in seven consecutively recruited non-diabetic Caucasian female control subjects (62 Ϯ 3 years; BMI: 27 Ϯ 6 kg/m 2 , mean Ϯ S.E.) and seven consecutively recruited Caucasian female type 2 diabetic patients (60 Ϯ 4 years; BMI: 35 Ϯ 3 kg/m 2 , HbA1c: 6.6 Ϯ 0.5%) or during local anesthesia with xylocain and needle biopsy through the skin from five Caucasian males with type 2 diabetes (56 Ϯ 5 years; BMI: 36 Ϯ 5 kg/m 2 , HbA1c: 6.2 Ϯ 0.6%). Isolated adipocytes were incubated with or without 100 nM insulin for 10 min, then cells were separated from medium, lysed in SDS and protein kinase, protein phosphatase, and proteinase inhibitors, and frozen within 10 s (3). IRS1 serine 307 phosphorylation was determined by immunoblotting with site-specific antibodies and chemiluminescence imaging and normalized for the amount of IRS1 protein in each sample (3). A, effect of insulin on cells from non-diabetic control subjects. B, Effect of insulin on cells from female patients with type 2 diabetes. C, comparison of the level of insulin-induced phosphorylation of serine 307 in cells from the female non-diabetic and diabetic subjects. D, effect of insulin on cells from male patients with type 2 diabetes. E, comparison of the amount of IRS1 protein in cells from the female non-diabetic and diabetic subjects. The amount of IRS1 protein was determined by immunoblotting and normalized for the amount of actin. In the comparisons all samples were analyzed on the same SDS-PAGE gel; A, paired and C, p Ͻ 0.05; B, paired; D, paired; and E, p Ͼ 0.05. Mean Ϯ S.E., n ϭ 7 (A, B, C, E) or 5 (D) subjects. Student's (paired) t test was used. F, blots for IRS1-serine 307 phosphorylation and IRS1 protein, corresponding to indicated panel; lanes without and with insulin are in the same order.

Inhibition of Serine 307 Phosphorylation in Cells from Control Subjects Makes Cells
Insulin-resistant-Rapamycin has been reported to inhibit insulininduced phosphorylation of the corresponding residue (serine 302) in mouse 32D myeloid cells and mouse embryo fibroblasts (26,27). In the human adipocytes from non-diabetic subjects in which insulin increased the phosphorylation of serine 307, this increase was inhibited by rapamycin (Fig. 5A). Importantly, rapamycin inhibition of phosphorylation at serine 307 impeded insulin-induced phosphorylation of IRS1 on tyrosine residues, resulting in decreased insulin sensitivity (Fig.  5B) without affecting the maximal response (data not shown). Thus rapamycin inhibited serine 307 phosphorylation of IRS1 and concomitantly transformed the cells into a diabetes-like phenotype (3). Rapamycin has been found to enhance insulin stimulation of glucose uptake in adipocytes (32,33), demonstrating other effects of rapamycin downstream of IRS1 (34), which makes it impossible to extend the analysis of rapamycin effects on serine 307 phosphorylation in human adipocytes to effects on glucose transport (Fig. 5C).  Fig. 1B (the biopsies from two of the patients did not produce enough cells for the dose-response analysis) and five female non-diabetic subjects (dotted lines) were incubated with the indicated concentration of insulin for 10 min, and IRS1 tyrosine phosphorylation was determined. The phosphorylation was normalized to maximal effect. Dose-response curves were fitted to experimental data using the sigmoidal dose-response algorithm in GraphPad Prism 4 software. and on tyrosine residues in response to insulin. A, adipocytes from non-diabetic subjects were incubated with the indicated concentration of insulin for 10 min, and serine 307 phosphorylation was determined (mean Ϯ S.E., n ϭ 5 subjects). The phosphorylation was normalized to maximal effect. Dose-response curves were fitted to experimental data using the sigmoidal dose-response algorithm in GraphPad Prism 4 software. B and C, cells from nondiabetic subjects were incubated with 100 nM insulin for the indicated time, and IRS1 serine 307 phosphorylation was determined (B) or on the same samples, IRS1 tyrosine phosphorylation was determined (C) (mean Ϯ S.E., n ϭ 3 subjects).

DISCUSSION
This is a first demonstration of a critical function of serine phosphorylation of IRS1 in primary human cells that also are target cells for insulin. More important, in two separate unselected groups of patients with type 2 diabetes these findings identify a mechanism with the potential to explain the insulin resistance of the disease: a curtailed phosphorylation of serine 307 reduces the steady-state tyrosine phosphorylation of IRS1 at physiological concentrations of insulin. Type 2 diabetes has a strong genetic component, and it therefore appears that different genetic traits, providing susceptibility to the disease, may converge at curtailed phosphorylation of IRS1 on serine 307.
We used two different approaches to manipulate serine 307 phosphorylation that yielded predicted effects on the sensitivity to tyrosine phosphorylation by insulin. (i) Phosphorylation of serine 307 by protein phosphatase inhibition in cells from patients with type 2 diabetes made the cells more insulin-sensitive, mimicking adipocytes from non-diabetic subjects. (ii) Inhibition of serine 307 phosphorylation by rapamycin in adipocytes from non-diabetic subjects made the cells insulin-resistant, mimicking cells from type 2 diabetic patients (3). Rapamycin and okadaic acid affect the phosphorylation state of many phosphorylation sites in the cells. They do not, however, directly affect tyrosine phosphorylation of the insulin receptor or IRS1, and they have entirely different mechanisms of action. In two different and opposite approaches, in normal and diabetic cells, the sensitivity of IRS1 to be tyrosine-phosphorylated in response to insulin was affected in a predicted manner. It is therefore reasonable to interpret these findings as supporting a cause-effect relationship between curtailed serine 307 phosphorylation and reduced insulin sensitivity in cells from patients with type 2 diabetes. To completely establish this it will be necessary to knock out wild-type IRS1 and replace it with serine 307 3 alanine-mutated IRS1, to block phosphorylation, in normal adipocytes and with serine 307 3 glutamate, to mimic phosphorylation, in adipocytes from patients with type 2 diabetes.
It cannot be ruled out that additional sites are phosphorylated and cooperate with serine 307 to sensitize IRS1 to tyrosine phosphorylation. Site-directed mutagenesis of serine 302 to alanine in murine 32D cells, however, indicates that blocking the phosphorylation of this serine alone is enough to inhibit IRS1 tyrosine phosphorylation in response to insulin (26).
The EC 50 for insulin-stimulated phosphorylation of serine 307 indicates downstream protein kinase activation. Serine 307 is located in a consensus sequence (RXRXXS/T) for phosphorylation by the downstream protein kinase B (PKB/Akt) (35). The ability of rapamycin to inhibit the phosphorylation suggests the involvement of protein kinase mammalian-target-of-rapamycin (mTOR) in relaying the phosphorylation. The rictor-mTOR complex has indeed very recently been demonstrated to activate PKB by phosphorylation of serine 473, thus complementing PDK1-catalyzed threonine 308 phosphorylation (34). Rictor-mTOR does not seem to be acutely inhibited by rapamycin, but prolonged presence of the inhibitor may inhibit the kinase (34). As a second potential candidate for serine 307 phosphorylation the insulin receptor and mTOR downstream p70 S6 kinase has been shown to effectively phosphorylate IRS1 on serine 307 in vitro (27). On the other hand, mTOR-mediated phosphorylation of IRS1-serine 307 in murine cells (serine 312 in human cells) represents a negative feedback inhibition of prolonged insulin stimulation (23,25). This mechanism is clearly different from the acute insulin-induced phosphorylation of serine 307 in human adipocytes, which sensitizes IRS1 to activation by the insulin receptor. It should be stressed that earlier experiments describing negative (19,27) or positive (26) effects of murine serine 302 phosphorylation (corresponding to human serine 307) on insulin-induced IRS1 tyrosine phosphorylation have been performed in cell lines, whereas the findings herein are from primary human cells.
The physiological feed-forward mechanism from insulin via a putative protein kinase mediating the serine 307 phosphorylation would be an example of sensitivity enhancement through positive feed-forward. The mechanism may be critical for maintaining insulin sensitivity in adipocytes of human beings. The delayed IRS1 serine 307 compared with tyrosine phosphorylation indicates that serine 307 phosphorylation is not required for the initial tyrosine phosphorylation but for the subsequently established steady-state level of tyrosine phosphorylation at physiological insulin concentrations, herein determined after 10 min. This could be achieved by inhibition of phosphotyrosineprotein phosphatase-catalyzed dephosphorylation of IRS1, as has been described for phosphorylation of the cluster of serines 265, 302, 325, and 358 (murine sequence) in IRS1 of CHO-T cells (36).
It will be important to establish whether this mechanism is operating in other target tissues of insulin such as muscle, liver, brain, and in islet ␤-cells. The protein kinase(s) and phosphatase(s) involved need to be defined and how they are con-trolled by insulin and the nature of the interference in type 2 diabetes. The inability of insulin to enhance the phosphorylation of serine 307 may be a distinguishing feature of insulin resistance and predict type 2 diabetes. The protein kinase(s) and the corresponding protein phosphatase(s) involved in controlling serine 307 phosphorylation are potential targets for novel pharmaceutical intervention.