Okadaic acid exerts a full insulin-like effect on glucose transport and glucose transporter 4 translocation in human adipocytes. Evidence for a phosphatidylinositol 3-kinase-independent pathway.

The effects of the serine/threonine phosphatase inhibitor, okadaic acid, and insulin on glucose transport activity, glucose transporter 4 translocation to the plasma membrane, and the signaling pathway of insulin were examined in human adipocytes. Okadaic acid consistently produced a greater increase than insulin in the rate of glucose transport, and both agents together had a partial additive effect. Both insulin alone and okadaic acid alone stimulated the translocation of glucose transporter 4 to the plasma membrane. Insulin, but not okadaic acid, stimulated phosphatidylinositol 3-kinase (PI 3-kinase) activity, and wortmannin completely inhibited the effect of insulin on glucose transport. When the cells were incubated with both agents, okadaic acid inhibited insulin-stimulated PI 3-kinase activity but did not block the association of the p85 or p110 subunits of PI 3-kinase with insulin receptor substrate 1. Insulin-stimulated tyrosine phosphorylation of insulin receptor substrate 1 was only slightly reduced (15-30%) by okadaic acid. These data demonstrate that okadaic acid exerts a full insulin-like effect independent of the activation of PI 3-kinase. Thus, PI 3-kinase lipid kinase is not essential for glucose transporter 4 translocation in human adipocytes, and different pathways exist that lead to glucose transporter 4 translocation and increased glucose transport.

The effects of the serine/threonine phosphatase inhibitor, okadaic acid, and insulin on glucose transport activity, glucose transporter 4 translocation to the plasma membrane, and the signaling pathway of insulin were examined in human adipocytes. Okadaic acid consistently produced a greater increase than insulin in the rate of glucose transport, and both agents together had a partial additive effect. Both insulin alone and okadaic acid alone stimulated the translocation of glucose transporter 4 to the plasma membrane. Insulin, but not okadaic acid, stimulated phosphatidylinositol 3-kinase (PI 3-kinase) activity, and wortmannin completely inhibited the effect of insulin on glucose transport. When the cells were incubated with both agents, okadaic acid inhibited insulin-stimulated PI 3-kinase activity but did not block the association of the p85 or p110 subunits of PI 3-kinase with insulin receptor substrate 1. Insulin-stimulated tyrosine phosphorylation of insulin receptor substrate 1 was only slightly reduced (15-30%) by okadaic acid. These data demonstrate that okadaic acid exerts a full insulin-like effect independent of the activation of PI 3-kinase. Thus, PI 3-kinase lipid kinase is not essential for glucose transporter 4 translocation in human adipocytes, and different pathways exist that lead to glucose transporter 4 translocation and increased glucose transport.
Insulin plays a key role for the regulation of metabolism in many mammalian cells, principally liver, muscle, and adipose cells (1,2). The ability of insulin to increase glucose transport into muscle and fat cells is mediated by the translocation of a specific glucose transporter, GLUT4, 1 from intracellular vesicles to the cell surface (3)(4)(5). The pathways mediating this translocation are poorly understood. The initial mechanism of insulin action involves its binding to specific cell surface receptors, leading to the autophosphorylation and activation of an intrinsic tyrosine kinase associated with the ␤ receptor subunit. A major target for the insulin receptor kinase is insulin receptor substrate 1 (IRS-1) (6 -8). In its tyrosine-phosphorylated form, IRS-1 acts as a docking protein that forms a signal-ing complex with other proteins with Src homology 2 domains, thus initiating divergent signaling cascades (2,7,9,10). One such target protein is the phosphatidylinositol (PI) 3-kinase, a heterodimeric enzyme consisting of an 85-kDa regulatory subunit with Src homology 2 domains capable of binding to phosphorylated IRS-1 and a 110-kDa catalytic subunit (11) that phosphorylates the inositol ring of phosphatidylinositol and its phosphorylated derivatives (12). On the basis of experiments with the PI 3-kinase inhibitors, wortmannin, and LY 294002 (13)(14)(15)(16) and with a dominant negative mutant regulatory subunit of this enzyme (17), PI 3-kinase has been implicated as one of the key signal transducers in insulin-stimulated glucose uptake and GLUT4 translocation.
Okadaic acid is a tumor promoter, originally isolated from the sea sponge, that potently inhibits the activity of protein phosphatases 1 and 2A (18). It rapidly stimulates protein phosphorylation in intact cells (19) and also partially stimulates glucose uptake in rat and 3T3-L1 adipocytes (19 -21). However, okadaic acid also blocks insulin-stimulated glucose transport, translocation of glucose transporters, and PI 3-kinase activation in rat adipocytes (20 -22).
In this study, we examined the effects of insulin and okadaic acid on glucose transport, protein tyrosine phosphorylation, and PI 3-kinase activation in human adipocytes. We demonstrate that okadaic acid exerts a full insulin-like effect in terms of increasing glucose transport activity through the translocation of GLUT4 to the plasma membrane. Furthermore, this full insulin-like effect of okadaic acid is independent of PI 3-kinase activation. The effects of insulin and okadaic acid are also in part additive, demonstrating the presence of multiple pathways for glucose transport activation and GLUT4 protein translocation.

EXPERIMENTAL PROCEDURES
Materials-Insulin was from Novo Nordisk (Copenhagen, Denmark). Okadaic acid was purchased from LC Laboratories (Woburn, MA). Bovine serum albumin (BSA) (fraction V) and wortmannin were from Sigma. Radiochemicals were from Amersham Corp. or ICN (U.K.). Anti-phosphotyrosine monoclonal antibodies, anti-IRS-1 polyclonal antibodies, and anti-PI 3-kinase (p85) whole antiserum were purchased from Upstate Biotechnology Inc. (Lake Placid, NY). GLUT1 and GLUT4 polyclonal antibodies were kindly provided by Dr. Sam Cushman (National Institutes of Health) and supplied by Hoffmann-La Roche. Rabbit antisera against the GLUT1 and GLUT4 glucose transporter isoforms were prepared using synthetic COOH-terminal peptides (EELFHPL-GADSQV and STELEYLGPDEND, respectively) as described (23). Protein A-Sepharose was from Pharmacia Biotech Inc.
Patients and Source of Adipose Tissue-Specimens of human subcutaneous adipose tissue were obtained from the abdominal region of nondiabetic subjects undergoing elective surgery for nonmalignant disease. The biopsies were obtained immediately after induction of anesthesia and placed in Medium 199 containing 4% BSA with 5.5 mM glucose at 37°C. The tissue was immediately transported to the labo-ratory for further processing. The study was approved by the Ethical Committee of the Goteborg University.
Preparation of Adipose Cells and Plasma Membranes-Adipose cells were prepared according to methods previously described (24). The tissue was cut into small fragments visibly free of connective tissue and clotted blood. About 0.6 g of tissue was incubated at 37°C in Medium 199 containing 4% BSA, 5.5 mM glucose, and 0.8 mg/ml collagenase in a shaking water bath. After ϳ50 min, liberated cells were filtered through a nylon mesh with a pore size of 400 m, washed four times in fresh BSA-containing medium, and finally resuspended at 2% cytocrit. Cell size and number were measured as described previously (24). Cells were then incubated with the various additions as indicated. Subcellular fractions enriched in plasma membranes were isolated following homogenization and differential centrifugation as described previously (25). The plasma membrane fraction was characterized by measuring 5Ј-nucleotidase. There was no evidence that insulin or okadaic acid altered the recovery of plasma membranes following the fractionation procedures (data no shown). The membranes were resuspended in TES buffer (20 mM Tris-HCl, 1 mM EDTA, 0.25 M sucrose) and assayed for proteins by the bicinchoninic acid method (26) with a commercial kit (Pierce). Aliquots (30 g of protein) were mixed with same volume of sample buffer containing 8 M urea, loaded on polyacrylamide 4 -20% gradient gels (Bio-Rad), and electrophoretically transferred onto nitrocellulose paper; immunoblotting was performed with GLUT4 or GLUT1 antibodies as described below.
Glucose Transport in Human Adipose Cells-Cellular uptake of [U-14 C]glucose was measured during a 1-h incubation of cells in glucosefree medium at lipocrit 3-5%. Under the conditions used, glucose transport was shown to be the rate-limiting step (27). Following preincubation in the absence or presence of okadaic acid for 5 min, 6.9 nM insulin was added, and after 15 min, 0.86 M [U-14 C]glucose was added. After 1 h, the cells were separated from the incubation medium by centrifugation through silicone oil, and the radioactivity associated with the cells was measured by scintillation counting.
Immunoprecipitations-All incubations were carried out at 37°C in Medium 199 buffered with 25 mM Hepes, 0.1 M N 6 -(2-phenylisopropyl)adenosine, and 1 unit/ml adenosine deaminase, with the addition of 4% BSA. Isolated human adipocytes were distributed into plastic vials (12-15% cell suspension) in a final incubation volume of 400 l. Cells were preincubated with the indicated concentrations of okadaic acid and/or 5 M wortmannin for 5 min. The incubations were then continued in the presence or absence of 6.9 nM insulin for 10 min. Cells were immediately separated by centrifugation through silicone oil and lysed in 0.4 ml of lysis buffer containing 25 mM Tris-HCl, pH 7.4, 0.5 mM EGTA, 25 mM NaCl, 1% Nonidet P-40, 1 mM Na 3 VO 4 , 10 mM NaF, 0.2 mM leupeptin, 1 mM benzamidine, and 0.1 mM 4-(2-aminoethyl)benzenesulfonyl fluoride hydrochloride and rocked for 40 min at 4°C. Detergent-insoluble material was sedimented by centrifugation at 12,000 ϫ g for 10 min at 4°C, and the supernatants were collected. Immunoprecipitations of tyrosine-phosphorylated proteins were performed with monoclonal anti-phosphotyrosine antibodies at a concentration of 5 g/ml for 2 h. Subsequently, the immune complexes were bound to protein A-Sepharose CL-4B (40 l of hydrated beads per sample) by incubation for 1 h at 4°C. The immunoprecipitates were washed twice with phosphate-buffered saline containing 1% Nonidet P- 40  Phosphatidylinositol 3-Kinase Activity-PI 3-kinase assay was performed directly on the immunoprecipitates as described previously (28). After the washings described above, 6 l of a mixture of phosphatidylinositol (10 g/sample) and phosphatidylserine (2.5 g/sample) were added to the beads, and the reaction was started by the addition of 30 l of a reaction mixture consisting of 40 mM Hepes pH 7.5, 20 mM MgCl 2 , and 50 M [␥-32 P]ATP (0.2 Ci/l). After 15 min at 30°C, the reaction was stopped by the addition of 40 l of HCl (4 N) and 160 l of CHCl 3 /methanol (1:1). The organic phase was extracted and applied to a silica gel thin layer chromatography plate precoated with 1% potassium oxalate (Analtech). Thin layer chromatography plates were developed in CHCl 3 /CH 3 OH/H 2 O/NH 4 OH (60:47:11.3:2), dried, and visualized by autoradiography. The radioactivity was quantified with a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).
Immunoblotting-Whole cell lysates or immunoprecipitations were boiled in Laemmli buffer containing 55 mM dithiothreitol for 5 min. The samples were analyzed by SDS-polyacrylamide gel electrophoresis. Proteins were transferred from the gel to the nitrocellulose sheets and blocked in 5% milk. The blots were probed with the different primary antibodies according to the recommendations of the manufacturer, and the proteins were detected by enhanced chemiluminescence using horseradish peroxidase-labeled second antibodies (Amersham), and the intensity of the bands was quantified with a laser densitometer (Molecular Dynamics, Sunnyvale, CA) and expressed in arbitrary units. Fig. 1 shows that incubation of human adipocytes with a maximal insulin concentration resulted in a 3-fold increase in glucose uptake, similar to the effect of okadaic acid at a submaximal concentration (0.25 M). A maximal concentration of okadaic acid (1 M) consistently produced a greater increase in glucose transport than that of a maximally effective insulin concentration. When the cells were incubated with maximal concentrations of both okadaic acid and insulin, a partial additive effect of the two agents was seen (Fig. 1).

Effect of Okadaic Acid and Insulin on Glucose Transport and GLUT4 Translocation-
To examine the role of PI 3-kinase in human fat cells for the effects of insulin and okadaic acid on glucose transport, incubations were performed with these agents in the absence or presence of the PI 3-kinase inhibitor wortmannin. Fig. 2 shows that when the cells were preincubated for 5 min with wortmannin, insulin-stimulated glucose transport was completely inhibited. In contrast, the effect of okadaic acid on glucose transport was essentially unchanged. Addition of wortmannin to the combination of insulin and okadaic acid produced a partial inhibition, probably caused by the inhibitory effect on insulin action while leaving that of okadaic acid unchanged.
To determine whether both insulin and okadaic acid stimulated glucose transport in human adipocytes by increasing the number of glucose transporters in the plasma membrane, the relative amounts of GLUT4 were determined in immunoblots of proteins from purified plasma membranes separated by SDS-PAGE in 4 -20% gradient minigels. In these gradient gels, both GLUT1 and GLUT4 proteins are seen as sharp bands. A typical blot and the quantification are shown in Fig. 3. In the absence of insulin or okadaic acid, a low GLUT4 content was found in the plasma membranes. However, following insulin exposure, the GLUT4 content in the plasma membranes increased 2-fold. Okadaic acid was as effective as insulin in increasing the amount of GLUT4 associated with the plasma membranes. When the cells were incubated with the combina- tion of okadaic acid and insulin, an additive effect of the two agents on GLUT4 content was observed (Fig. 3B). In contrast, the GLUT1 content in the plasma membranes was not altered by okadaic acid and/or insulin (Fig. 3A).
Effect of Insulin and Okadaic Acid on PI 3-Kinase Activity-PI 3-kinase activity was measured in anti-phosphotyrosine immunoprecipitates of cellular proteins obtained from human adipocytes treated with okadaic acid and/or insulin. As shown in Fig. 4, okadaic acid per se had no effect on PI 3-kinase activity, while insulin induced a 20-fold increase in activity. Okadaic acid markedly reduced the stimulatory effect of insulin on PI 3-kinase activity (Fig. 4), while at the same time there was an additive effect in GLUT4 translocation (Fig. 3).
To determine whether the inhibitory effect of okadaic acid on insulin-stimulated PI 3-kinase activity results from a decrease in the amount of PI 3-kinase recovered in the anti-phosphotyrosine immunoprecipitates or associated with IRS-1, the presence of PI 3-kinase was assessed by immunodetection in antiphosphotyrosine and anti-IRS-1 immunoprecipitates. As shown in Fig. 5, the p85 subunit of PI 3-kinase was present in both the anti-phosphotyrosine (Fig. 5A) and the anti-IRS-1 (Fig. 5B) immunoprecipitates from cells that had been incubated with insulin, while okadaic acid alone had no effect. When the cells were pretreated with okadaic acid before exposure to insulin, the p85 protein band was unchanged, showing that okadaic acid did not interfere with the insulin-stimulated association of p85 with IRS-1. Similar results were obtained when the membranes were reprobed with antibodies against the catalytic subunit (p110) of PI 3-kinase (data not shown).
Effect of Okadaic Acid and Insulin on Protein Tyrosine Phosphorylation-Tyrosine phosphorylation of cellular proteins from human adipocytes treated with okadaic acid and/or insulin was studied by immunoblotting proteins from whole cell lysates with anti-phosphotyrosine antibodies. When the cells were stimulated with insulin, two major tyrosine-phosphorylated bands appeared corresponding to the insulin receptor ␤ subunit and IRS-1 (Fig. 6A). Okadaic acid per se did not increase the tyrosine phosphorylation of these proteins. When the cells were pretreated for 5 min with okadaic acid and then incubated with insulin, the tyrosine phosphorylation of the insulin receptor was unaltered but the band corresponding to the tyrosine phosphorylation of IRS-1 was inhibited, and a new band migrating at higher molecular weight appeared. To determine whether this new band corresponded to a hyperphosphorylated form of IRS-1 or to a new protein, the same blot was reprobed with anti-IRS-1 polyclonal antibodies, which recognized this band as IRS-1 (data not shown). In addition, cells treated with okadaic acid and/or insulin were immunoprecipitated with anti-IRS-1 antibodies, and the extent of tyrosine phosphorylation of IRS-1 was determined. Fig. 6B shows that the insulin-stimulated tyrosine phosphorylation of IRS-1 was partially inhibited by okadaic acid (15-30%). However, this reduction in tyrosine phosphorylation of IRS-1 did not inhibit the association of p85 (or p110, not shown) with IRS-1 (Fig. 5). DISCUSSION The salient finding of our study with human adipocytes is that okadaic acid exerts a full insulin-like effect on both glucose transport activity and translocation of GLUT4 proteins to the plasma membrane. However, the insulin-like effects occur in the absence of any measurable increase in PI 3-kinase activity. In fact, the maximal effect of okadaic acid alone on glucose transport was consistently greater than that of insulin alone. The combination of insulin and okadaic acid produced an additive effect on both glucose transport activity and GLUT4 content in the plasma membranes, suggesting the presence of different pathways to activate glucose transport and GLUT4 FIG. 2. Effect of wortmannin on glucose uptake stimulated by okadaic acid and/or insulin. Isolated adipocytes were incubated as described in Fig. 1 in the absence or presence of 5 M wortmannin (Wort) added 5 min before the addition of 1 M okadaic acid (OA) and/or 6.9 nM insulin (INS). Glucose uptake was measured as described. Results are means Ϯ S.E. of six separate experiments performed in triplicates.

FIG. 3. Effect of okadaic acid and insulin on the GLUT1 and GLUT4 content in purified plasma membranes.
Human adipocytes were incubated with insulin (INS) and/or okadaic acid (OA) for 20 min. Cells were homogenized, and subcellular fractionation was performed as described under "Experimental Procedures." Plasma membranes (30 g of protein) were separated by SDS-PAGE using 4 -20% gradient minigels, transferred to membranes, and immunoblotted with antibodies against GLUT1 and GLUT4 as described. A, immunoblot of a representative experiment; B, quantification of the bands by laser densitometry. The experiment was repeated twice with similar results.

translocation.
These findings differ from those recently reported with rat and mouse adipocytes (20,22,29) as well as our own unpublished experiments. In these cells, okadaic acid alone elicits only a small increase in glucose transport, and when combined with insulin, okadaic acid markedly inhibits the effect of insulin (29). Thus, there seem to be species differences in the responsiveness to the okadaic acid-related pathway to increase glucose transport. This concept is further supported by the recent finding that okadaic acid also seems to exert a full insulin-like effect on glucose transport in human skeletal muscle, although no additive effect with insulin was demonstrated (30).
In agreement with previous studies in rat adipocytes (21), okadaic acid alone did not increase tyrosine phosphorylation of the insulin receptor or IRS-1 in human adipocytes and altered the mobility of IRS-1, probably due to a hyperphosphorylation involving serine/threonine sites as also demonstrated in animal cells (31). Okadaic acid also produced a partial inhibition (15-30%) of the insulin-stimulated tyrosine phosphorylation of IRS-1. However, this reduction in tyrosine phosphorylation was not associated with an inhibition of the binding of the p85 or p110 subunits of PI 3-kinase to IRS-1. Taken together, the effects of insulin on the upstream signaling pathway, including tyrosine phosphorylation of the insulin receptor as well as the binding of the PI 3-kinase subunits to IRS-1, were not impaired by okadaic acid. This is consistent with the finding that the effect of insulin on glucose transport activity and GLUT4 translocation was also unimpaired. However, a surprising finding was that the activation of PI 3-kinase lipid kinase by insulin was markedly reduced by okadaic acid, in spite of the normal insulin signaling, including the binding of p85 and p110 to IRS-1. These results lead to the conclusion that either PI 3-kinase lipid kinase is not necessary for glucose transport activation or that cellular PI 3-kinase activity is redundant and that only a small activation is necessary to elicit the full insulin effect or that the association of PI 3-kinase with IRS-1 is by itself an important factor. This latter possibility is supported by recent findings showing that the association of PI 3-kinase with the signaling complex appears to be the most important factor rather than the absolute level of phosphorylated phosphoinositides (9,32).
The finding that okadaic acid alone exerts a full insulin-like effect on glucose transport activity and GLUT4 translocation in the absence of any measurable increase in PI 3-kinase activity clearly supports the presence of an alternative and PI 3-kinaseindependent pathway to stimulate glucose transport. The molecular mechanisms of this putative pathway are currently unclear. PI 3-kinase is a dual specificity kinase with both lipid and serine kinase activity (33,34). However, it is unlikely that the PI 3-kinase serine kinase plays a critical role since wortmannin, which inhibits the effect of insulin but not that of okadaic acid, also inhibits the PI 3-kinase serine kinase activity (35). This finding does not exclude a critical role of other serine kinases in mediating the insulin-like effect of okadaic acid alone. Interestingly, serine phosphorylation of the p85 and p110 subunits of PI 3-kinase markedly impairs the PI 3-kinase lipid kinase activity (33,34). Such an effect of okadaic acid may thus explain the inhibition of the PI 3-kinase activity in the presence of insulin.
There is much evidence to support a key role of PI 3-kinase activity in eliciting the stimulating effect of insulin on glucose uptake. Various studies have shown that the effect of insulin on this enzyme is reduced in states of insulin resistance (36 -38), that the inhibition of PI 3-kinase by wortmannin and LY 294002 leads to the blockade of insulin-stimulated glucose transport (Fig. 2) (14,16,39), and that the p110 subunit of the PI 3-kinase presents homology with the Vsp34 gene product that is involved in vesicular trafficking in yeast (40). In addition, it has been shown that the ability of insulin to stimulate the recruitment of GLUT4-containing vesicles to the plasma membrane is associated with the recruitment of p85/p110 PI 3-kinase to the glucose transporter-carrying vesicles and potential participation of the lipid products of this enzyme in the trafficking of these vesicles (16,41,42). However, a recent study demonstrated that PI 3-kinase activation is necessary but not by itself sufficient to stimulate glucose transport and GLUT4 translocation in 3T3-L1 adipocytes in response to insulin (43).
Our results show that wortmannin blocks the insulin-but not the okadaic acid-stimulated glucose transport in human adipocytes. In a similar way, it was recently shown that wortmannin does not interfere with the stimulatory effects of contraction and hypoxia on glucose transport in muscle while the effect of insulin is inhibited (44). Taken together, these results suggest that the effect of okadaic acid on glucose transport is independent of PI 3-kinase activation and that the ability of wortmannin to block insulin-stimulated glucose transport is specific for the insulin-signaling system and not for the mechanisms responsible for the movement of transporters to the cell surface.
The GLUT4 isoform is the principal glucose transporter isoform in human adipose cells (45). The GLUT1 content is only 10% of that of GLUT4. Since GLUT1 also has low affinity for the glucose, this isoform makes only a minor contribution to the total glucose transport activity in human fat cells (45). Neither insulin nor okadaic acid altered the subcellular distribution of GLUT1.
Our results also clearly indicate that different pathways exist in human adipocytes for GLUT4 translocation and activation of glucose transport. This may explain the additive effect of insulin and okadaic acid on GLUT4 content in the plasma membranes. It is clear that insulin alone does not promote the complete redistribution of GLUT4 proteins from the intracellular pool to the plasma membranes even in highly insulinresponsive rat adipocytes (46). Although it may be due to the continued recycling of GLUT4 even in the presence of insulin (46), recent studies with skeletal muscle have also suggested the presence of different intracellular pools of GLUT4 susceptible to different stimuli (44). Measurements of the redistribution of the intracellular pool of GLUT4 transporting proteins following the exposure to insulin and/or okadaic acid are necessary to define the mechanisms for the additive effect of these agents. Unfortunately, we were unable to obtain enough human tissue to quantify the number of intracellular GLUT4 proteins. The possibility that okadaic acid acts through the inhibition of GLUT4 endocytosis is highly unlikely since kinetic experiments with 3-O-methylglucose uptake have shown that the half-time for the stimulating effect of okadaic acid on glucose transport is similar to that of insulin (ϳ3 min, data not shown). Irrespective of mechanisms, the finding that the okadaic acid-stimulated pathway is PI 3-kinase-independent, but additive to that of insulin, makes it a potentially important therapeutic target in diabetes and other insulin-resistant states.