Potentiation of insulin-related signal transduction by a novel protein-tyrosine phosphatase inhibitor, Et-3,4-dephostatin, on cultured 3T3-L1 adipocytes.

We previously isolated dephostatin from Streptomyces as a novel inhibitor of CD45-associated protein-tyrosine phosphatase. We prepared Et-3,4-dephostatin as a stable analogue and found it to inhibit PTP-1B and SHPTP-1 protein-tyrosine phosphatases selectively but not to inhibit CD45 and leukocyte common antigen-related phosphatase ones effectively. Et-3,4-dephostatin increased the tyrosine phosphorylation of the insulin receptor and insulin receptor substrate-1 with or without insulin in differentiated 3T3-L1 mouse adipocytes. The increase of tyrosine phosphorylation by Et-3,4-dephostatin was more prominent in 6-h than in 30-min incubation. It also increased phosphorylation and activation of Akt with or without insulin. Et-3,4-dephostatin also enhanced translocation of glucose transporter 4 from the cytoplasm to the membrane and 2-deoxy-glucose transport. Et-3,4-dephostatin-induced glucose uptake was inhibited by SB203580, a p38 inhibitor, but not by PD98059, a MEK inhibitor, or by cycloheximide as insulin-induced uptake. Interestingly, although LY294002, a phosphatidylinositol 3-kinase inhibitor, inhibited the insulin-induced glucose uptake completely, it only partially inhibited the Et-3,4-dephostatin-induced uptake. It also blocked insulin-induced glucose transporter 4 translocation but not the Et-3,4-dephostatin-induced one. The increase in c-Cbl tyrosine phosphorylation caused by Et-3,4-dephostatin was stronger than that in insulin receptor phosphorylation. These observations indicate that a phosphatidylinositol 3-kinase-independent pathway involving c-Cbl is more important in Et-3,4-dephostatin-induced glucose uptake than in insulin-induced uptake. Et-3,4-dephostatin showed an in vivo antidiabetic effect in terms of reducing the high blood glucose level in KK-A(y) mice after oral administration. Thus, Et-3,4-dephostatin potentiated insulin-related signal transductions in cultured mouse adipocytes and showed an antidiabetic effect in mice.

Type 2 diabetes mellitus is mainly characterized by impaired signal transduction downstream of the insulin receptor in pe-ripheral tissues, such as skeletal muscles and adipocytes (1)(2)(3)(4). Insulin secretion from the pancreatic ␤-cells is also affected in many cases of type 2 diabetes mellitus (5,6), since the secondary effect of impaired signal transduction may affect insulin secretion (it was reported that long-term exposure to a high glucose concentration induced apoptosis in pancreatic ␤-cells (7)). A functional insulin receptor was also shown to be essential for the secretion of insulin by cultured mouse ␤-cells (8).
Thiazolidine derivatives such as troglitazone are being or have been widely used for the treatment of type 2 diabetes mellitus (9). Troglitazone is a ligand of peroxisome proliferatoractivated receptor ␥ that transcriptionally regulates a number of adipose tissue-specific genes by binding to peroxisome proliferator-activated receptor response elements of these genes as a heterodimer with retinoid X receptor. These heterodimers potentiate adipose tissue differentiation in vivo (10), and their side effects include obesity. Several drugs such as sulfonylureas enhance insulin secretion. Sulfonylureas bind to receptors on pancreatic ␤-cells to induce Ca 2ϩ influx, thereby increasing insulin secretion (11). Compared with the insulin therapy for type I diabetes mellitus, chemotherapy for type 2 has been poorly developed, and less toxic chemotherapy based on the insulin receptor-associated signal transduction should be developed.
The insulin receptor is a heterotetramer consisting of two ␣-subunits that are entirely extracellular and two ␤-subunits that span the plasma membrane and contain intrinsic tyrosine kinase activity (12,13). Binding of insulin to its receptor results in activation of the receptor tyrosine kinase and receptor autophosphorylation, followed by tyrosine phosphorylation of IRS-1, -2, and -3, which then bind to the p85 regulatory subunit of phosphatidylinositol (PI) 1 3-kinase. Then the p110 catalytic subunit of PI 3-kinase produces phosphatidylinositol 3,4bisphosphate and phosphatidylinositol 3,4,3-trisphosphate, which can bind to the PH domain of PKB␣(␤)/Akt1 (2), which then up-regulate glucose transporter 4 (GLUT4) translocation. These Akts are phosphorylated to become activated by 3-phosphoinositide-dependent protein kinase 1 or 2 (14). The insulinstimulated phosphorylations of Akt on threonine 308 by 3-phosphoinositide-dependent protein kinase 1 and on serine 473 by 3-phosphoinositide-dependent protein kinase 2 are required for maximal Akt activity (15)(16)(17). Kupriyanova and Kandror (18) reported that the phosphorylated Akt2 directly binds to GLUT4-containing vesicles to phosphorylate GLUT4 to trigger translocation of GLUT4 from the cytoplasm to the membrane (19), resulting in an increase in cellular glucose uptake (20). On the other hand, very recently, Pessin and co-workers (21) showed the existence of a PI 3-kinase-independent c-Cbl pathway in insulin-induced signal transduction. The c-Cbl protein is a proto-oncogene product that can be tyrosine-phosphorylated. The Cbl-associated protein forms complex with c-Cbl. This complex mediates GLUT4 translocation through the PI 3-kinase-independent pathway in 3T3-L1 adipocytes (21).
Several protein-tyrosine phosphatases (PTPases) have been identified in major insulin-sensitive tissues such as skeletal muscle, liver, and adipose tissue. These include transmembrane PTPases such as CD45 and leukocyte common antigenrelated phosphatase (LAR)/RPTP-␣ and nontransmembrane PTPases such as SHPTP-1, SHPTP-2, PTP-1B, and PTP-1C (22). Various PTPases are considered to be involved in the etiology of diabetes mellitus. Especially, PTP-1B, a cytoplasmic PTPase, is known to be a negative regulator of insulin receptorassociated signal transduction (23). PTP1B-deficient mice showed increased insulin sensitivity and resistance to obesity (24). Overexpression of PTP-1B in Rat1 fibroblasts is known to reduce ligand-stimulated autophosphorylation of the insulin receptor (23). Therefore, inhibitors of PTP-1B may enhance the insulin sensitivity and glucose uptake .
We previously isolated dephostatin from Streptomyces as a novel inhibitor of T-cell receptor-associated protein-tyrosine phosphatase CD45 (26). Since dephostatin was unstable in cell culture media, later we synthesized Me-3,4-dephostatin as a stable analogue (27). Me-3,4-dephostatin was shown to enhance nerve growth factor-or epidermal growth factor-induced morphological differentiation in rat pheochromocytoma PC12 h cells (28). This PTPase inhibitor prolonged the tyrosine-phosphorylated and activated state of MAP kinase. More recently, we synthesized Et-3,4-dephostatin as another stable analogue (29).
In this report, we studied the effects of Et-3,4-dephostatin on tyrosine phosphorylation of cellular proteins, activation of Akt, GLUT4 translocation, and hexose transport. We found that Et-3,4-dephostatin mimicked or potentiated the various activities of insulin; especially, it activated the PI 3-kinase-independent signaling pathway. It also showed antidiabetic activity in mice.

EXPERIMENTAL PROCEDURES
Materials-Dephostatin analogues were synthesized as described previously (29). Insulin and p-nitrophenyl phosphate were obtained from Sigma. Human recombinant PTP-1B, SHPTP-1, and anti-phosphotyrosine antibody (4G10) were obtained from Upstate Biotechnology, Inc. (Lake Placid, NY). CD45 and LAR were purchased from Biomol (Plymouth Meeting, PA). Anti-insulin receptor ␤ subunit antibody, anti-Akt antibody, anti-GLUT1 antibody, anti-GLUT4 antibody, and horseradish peroxidase-conjugated anti-goat IgG were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-phosphorylated Akt antibody was obtained from New England Biolabs (Mississauga, Canada). Horseradish peroxidase-conjugated anti-mouse IgG and anti-rabbit IgG were obtained from Amersham Pharmacia Biotech. Western Blot Chemiluminescence Reagent and Tyramide Signal Amplification (TSA TM -Direct) were obtained from PerkinElmer Life Sciences. Cell Culture and Differentiation-3T3-L1 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% calf serum at 5% CO 2 . For differentiation, cells (1 ϫ 10 6 ) were cultured in a 60-mm dish for 3 days to postconfluence in medium containing 10% fetal bovine serum, 500 M isobutylmethylxanthine, 1 M dexamethasone, and 1 g/ml insulin. Then the cells were grown in medium containing only 10% fetal bovine serum and insulin for another 2 days. After that, the medium was changed, and the cells were cultured in 10% fetal bovine serum medium for another 5-7 days for full differentiation (30).
Enzyme Assays-PTP-1B, SHPTP-1, CD45, and LAR were assayed in a reaction mixture (final volume 200 l) containing 25 mM HEPES (pH 7.2), 50 mM NaCl, 5 mM dithiothreitol, 2.5 mM EDTA, 100 g/ml bovine serum albumin, 5 mM p-nitrophenyl phosphate, 0.5 unit of each enzyme, and the test chemical. After incubation for 10 min at 37°C, the reaction was terminated by the addition of 100 l of 2 M Na 2 CO 3 , and the absorbance at 405 nm was measured.
Immunoprecipitation of Signaling Proteins-Differentiated 3T3-L1 adipocytes were grown in 60-mm dishes. After having been treated with test chemicals, the cells were washed with 1.5 ml of ice-cold phosphatebuffered saline (PBS) containing 1 mM Na 3 VO 4 , scraped off the dishes, and centrifuged at 1000 ϫ g for 5 min. After solubilization with 300 l of lysis buffer, the soluble fractions were incubated with a corresponding antibody at 2 g/300 l of lysate overnight at 4°C followed by incubation for 6 h with protein G-agarose. The immunocomplexes were pelleted by centrifugation, washed four times with IP buffer (50 mM HEPES, 150 mM NaCl, 2.5 mM EGTA, 1 mM EDTA, 1 mM dithiothreitol, 0.1% Tween 20), and then boiled in a loading buffer.
Western Blot Analysis-3T3-L1 adipocytes (1 ϫ 10 6 ) were grown in a 60-mm dish for 10 days. After having been treated with test chemicals, the cells were washed with 1.5 ml of ice-cold PBS containing 1 mM Na 3 VO 4 and centrifuged at 1000 ϫ g for 5 min. After solubilization of the cells with 200 l of lysis buffer consisting of 50 mM HEPES (pH 7.4), 150 mM NaCl, 10 mM EDTA, 1% Triton X-100, 200 mM sodium fluoride, 10% glycerol, 20 mM sodium pyrophosphate, 2 mM phenylmethanesulfonyl fluoride, 4 mM Na 3 VO 4 , 0.1 mg/ml leupeptin, and 15 mM benzamidine, the cell lysate was centrifuged at 14,000 ϫ g for 10 min. The supernatant was assayed for protein content by using a Bio-Rad protein assay kit. The proteins were resolved by SDS-polyacrylamide gel electrophoresis and electrotransferred to polyvinylidene difluoride membranes. After having been blocked with 5% (w/v) nonfat dry milk in Tris-buffered saline-Tween buffer containing 20 mM Tris (pH 7.6), 0.14 M NaCl, and 0.1% (w/v) Tween 20, the membranes were treated with antibodies, and then the proteins were visualized by use of an ECL Western blotting detection system.
Plasma Membrane Sheet Assay-Plasma membrane sheets were prepared from differentiated 3T3-L1 adipocytes as described by Robinson et al. (31). Briefly, the cells cultured on a microcover glass were washed once with ice-cold PBS and incubated with 500 l of 0.5 mg/ml poly-Llysine for 30 s. The cells were then swollen in 500 l of hypotonic buffer (23 mM KCl, 10 mM HEPES, pH 7.5, 2 mM MgCl 2 , 1 mM EGTA) with three successive rinses. The swollen cells were sonicated for 5 s in a sonication buffer (70 mM KCl, 30 mM HEPES, pH 7.5, 5 mM MgCl 2 , 3 mM EGTA, 1 mM dithiothreitol, 0.1 mM phenylmethanesulfonyl fluoride). The bound plasma membrane sheets were washed three times with the sonication buffer and used for immunofluorescence studies.
2-Deoxyglucose Uptake in 3T3-L1 Adipocytes-Differentiated 3T3-L1 adipocytes were grown in 12-well plates. After preincubation with the desired test chemical for 1 h, 10 nM insulin was added to the medium. After the incubation, the cells were incubated in assay buffer consisting of 140 mM NaCl, 2.7 mM KCl, 6.5 mM Na 2 HPO 4 , 1.5 mM KH 2 PO 4 , 0.9 mM CaCl 2 , 0.5 mM MgCl 2 , and 1 mg/ml bovine serum albumin, pH 7.0, at 37°C for 10 min, and then the cells were incubated with 0.01 Ci of 2-[ 3 H]deoxy-D-glucose for 10 min. The uptake was terminated by adding 500 l of ice-cold PBS containing 0.1 mM phloretin. The cells were washed twice with ice-cold PBS and then solubilized in 0.5 N NaOH. Thereafter, the radioactivity was measured by a scintillation counter. Nonspecific glucose uptake was measured by treating cells with 20 g/ml cytochalasin B, and this value was subtracted from all of the data.
Immunostaining of GLUT4 -Immunostaining of GLUT4 was carried out as previously described (32). Differentiated 3T3-L1 adipocytes were cultured on microcover glasses in 12-well plates. After having been treated with 10 g/ml Et-3,4-dephostatin for 6 h, the adipocytes were fixed in 500 l of 2% formaldehyde in PBS for 5 min on ice and 5 min at room temperature. Then they were washed with 500 l of 100 mM glycine in PBS for 15 min and blocked with 500 l of 5% calf serum in PBS for 60 min. The cells were next incubated with 2 g/ml polyclonal anti-GLUT4 antibody in PBS overnight at 4°C. After that, they were incubated with horseradish peroxidase-conjugated anti-goat IgG (diluted 1:100 with PBS) at room temperature for 60 min and then washed with TNT buffer (0.1 M Tris-HCl, pH 7.5, 0.15 M NaCl, 0.05% Tween 20) three times. The cells were then incubated with fluorophore tyramide (diluted 1:50 with the amplification diluent; PerkinElmer Life Sciences) for 7 min at room temperature in the dark, washed three times, 10 min each time, in TNT buffer at room temperature, and visualized under a fluorescence microscope.
In Vivo Blood Glucose Level-KK-A y mice (10-week-old males; average weight, 40 g) were purchased from Japan Clea Co. The test chemical was added to CE-2 diet containing 12% sucrose, and the mice were given the drug-containing diet for 24 h. Then the blood was taken from the orbital capillary bed, and 0.01-ml aliquots of the serum were taken after centrifugation. The glucose content was measured with a glucose test kit (Wako, Tokyo).
Effect of Et-3,4-Dephostatin on Intracellular Tyrosine Phosphorylation in 3T3-L1 Adipocytes-Mouse 3T3-L1 fibroblasts were differentiated into adipocytes for 8 -12 days. Et-3,4-dephostatin was added to the differentiated 3T3-L1 adipocytes (day 10) with or without suboptimal 10 nM insulin; the cells were incubated for 30 min, 2 h, or 6 h; and the tyrosine phosphorylation was then examined by immunoprecipitation with anti-insulin receptor and anti-IRS-1 antibodies. Insulin at 10 nM induced tyrosine phosphorylation of the 95-kDa insulin receptor ␤-subunit most prominently in 30 min, and Et-3,4dephostatin alone at 10 g/ml increased the tyrosine phosphorylation more slowly, as shown in Fig. 2. When added together, insulin and Et-3,4-dephostatin enhanced the phosphorylation synergistically in 6 h. Insulin and Et-3,4-dephostatin did not change the cellular amount of insulin receptor and IRS-1. In a 30-min incubation, Et-3,4-dephostatin alone increased tyrosine phosphorylation of insulin receptor. But unexpectedly, it even lowered the insulin-induced tyrosine phosphorylation of insu-lin receptor and IRS-1 in a 30-min incubation. In a 6-h incubation, insulin and Et-3,4-dephostatin also increased the phosphorylation of other proteins at 135, 115, 85, and 60 kDa, which were identified to be phospholipase C␥, c-Cbl, PI 3-kinase regulatory subunit, and IRS-3, respectively, by the mobility of each authentic protein in a Western blotting analysis (data not shown).
Effect of Et-3,4-dephostatin on Akt Activation-PKB/Akt is a downstream signal transducer of the IR-PI3K pathway that up-regulates insulin signaling. So we examined whether Et-3,4-dephostatin could induce Akt activation, which can be monitored by phosphorylation. We employed a 6-h incubation, since the effect of Et-3,4-dephostatin on tyrosine phosphorylation was more prominent in 6-h incubation. As a result, Et-3,4dephostatin induced Akt phosphorylation and enhanced insulin-induced activation synergistically, as shown in Fig. 3, whereas it did not alter the amount of Akt. As expected, either insulin-induced or Et-3,4-dephostatin-induced Akt activation was inhibited by LY294002, a PI 3-kinase inhibitor.
Effect of Et-3,4-dephostatin on GLUT4 Translocation-Insulin induces GLUT4 translocation from intracellular vesicles to the plasma membrane (20), so we studied the effect of Et-3,4-   -3,4-dephostatin PTP-1B, SHPTP-1, CD45, and LAR were assayed in a reaction mixture containing 5 mM p-nitrophenyl phosphate, 0.5 unit of each enzyme, and the chemical. After incubation for 10 min at 37°C, the reaction was terminated by the addition of 100 l of 2 M Na 2 CO 3 , and then absorbance at 405 nm was measured. dephostatin on GLUT4 translocation by employing the plasma membrane sheet assay. In this assay, the upper membrane is removed after treatment of the cells, and translocation to the membrane can be detected by the antibody to the intracellular portion of GLUT4 (31). As shown in Fig. 4A, Et-3,4-dephostatin or insulin induced translocation of GLUT4 to the membrane in 3T3-L1 adipocytes in 6 h. GLUT1 was located at the membrane without stimulation, and the location was not altered by Et-3,4-dephostatin or insulin (Fig. 4B). As shown in Fig. 4C, the cellular amount of GLUT4 or GLUT1 was not changed by insulin or Et-3,4-dephostatin. Thus, Et-3,4-dephostatin specifically induced translocation of GLUT4, like insulin. insulin in the target tissues. As shown in Fig. 5A, insulin at 10 nM induced glucose uptake most prominently at 30 min. On the other hand, Et-3,4-dephostatin at 10 g/ml alone induced glucose uptake time-dependently up to 6 h. When added together, Et-3,4-dephostatin increased the uptake additively and timedependently up to 6 h. Fig. 5B shows the dose effect of Et-3,4dephostatin on glucose uptake at 6 h. Et-3,4-dephostatin increased the uptake at 1-10 g/ml with or without insulin, and the effect of the combination was greater than that of 100 nM (30.4 g/ml) sodium vanadate. On the other hand, 4-O-Me-Et-3,4-dephostatin only slightly enhanced the uptake. The slight increase would be caused by the 4-O-Me compound still weakly inhibiting PTPases. Thus, Et-3,4-dephostatin stimulated glucose uptake in differentiated 3T3-L1 adipocytes with or without a suboptimal concentration of insulin.

Inhibition of Et-3,4-dephostatin-induced 2-[ 3 H]deoxy-D-glucose Uptake by Signal Transduction
Inhibitors-Next we studied which signaling pathways are involved in Et-3,4-dephostatin-induced glucose uptake in 3T3-L1 adipocytes by using enzyme inhibitors of small molecular weight. 3T3-L1 adipocytes were pretreated with various inhibitors for 1 h and then stimulated with 10 nM insulin and/or 10 g/ml Et-3,4-dephostatin for 6 h. As shown in Fig. 6A, insulin-or Et-3,4-dephostatin-stimulated glucose uptake was not affected by 50 M PD98059, a MEK inhibitor, or 30 g/ml cycloheximide. SB203580, a p38 inhibitor, at 10 nM inhibited both insulin and Et-3,4-dephostatin-stimulated glucose uptake significantly. On the other hand, although 100 M LY294002, a PI 3-kinase inhibitor, completely inhibited the insulin-stimulated glucose uptake, it only partially inhibited the Et-3,4-dephostatin-stimulated glucose uptake. Fig. 6B shows the effect of LY294002 on insulin-and Et-3,4-dephostatin-induced glucose uptake in 30 min. Unexpectedly, LY294002 inhibited both insulin-induced and the inhibitor-induced glucose uptake completely. Therefore, in long term incubation, Et-3,4-dephostatin may employ different signals from those of insulin for activation of glucose transport, which are independent of PI 3-kinase.
Effect of Et-3,4-dephostatin on c-Cbl Phosphorylation-Cellular Cbl was reported to be an important mediator of the PI 3-kinase-independent signaling pathway induced by insulin (21). Therefore, we studied the effect of Et-3,4-dephostatin on c-Cbl phosphorylation in 3T3-L1 adipocytes by immunoprecipitation. As shown in Fig. 8, insulin induced tyrosine phosphorylation of c-Cbl at 10 min, but the phosphorylation disappeared at 6 h. When the cells were treated with Et-3,4-dephostatin with or without insulin, the phosphorylation was clearly observed after 6 h. This phosphorylation of c-Cbl was not inhibited by LY294002. These results were parallel to those of 2-deoxyglucose uptake in Fig. 6. Thus, Et-3,4-dephostatin is likely to inhibit tyrosine dephosphorylation of c-Cbl, activating PI 3-kinaseindependent signaling for GLUT4 translocation.
Effect of Et-3,4-dephostatin on Blood Glucose Level in Vivo-Due to the malfunction of their melanocortin receptor, KK-A y mice have yellow hair and show obesity and a high blood glucose level of 400 -500 mg/dl (33). When Et-3,4-dephostatin was given to KK-A y mice at 500 mg/kg orally with the diet, the glucose level significantly decreased by about 50%, as shown in Fig. 9. In this model, troglitazone at 500 mg/kg also lowered the glucose level by about 50% (data not shown). Thus, Et-3,4dephostatin decreased the blood glucose level in vivo.

DISCUSSION
Few PTPase inhibitors have been used as biochemical tools except sodium vanadate. We previously developed Me-3,4-dephostatin as a stable analogue of dephostatin (27) and showed it to enhance the effect of nerve growth factor and epidermal growth factor on differentiation of rat pheochromocytoma PC12 h cells (28). In this paper, we demonstrate that an additional stable analogue, Et-3,4-dephostatin, potentiated or mimicked the effect of insulin in differentiated 3T3-L1 adipocytes and showed an antidiabetic effect in vivo.
First, we found that Et-3,4-dephostatin selectively inhibits PTP-1B and SHPTP-1, both of which are known to be involved FIG. 5. Enhancement of glucose uptake in insulin-treated or untreated 3T3-L1 adipocytes. A, 3T3-L1 adipocytes were untreated (E) or treated with 10 nM insulin (q), 10 g/ml Et-3,4-dephostatin (‚), or insulin and Et-3,4-dephostatin (OE) for the indicated periods. B, dose effect of Et-3,4-dephostatin on glucose uptake. After incubation, the cells were incubated with 2-[ 3 H]deoxy-D-glucose for 10 min, and the reaction was then terminated. After having been washed twice, the cells were solubilized and neutralized, and the radioactivity was measured by a scintillation counter. Values are means Ϯ S.D. of triplicate determinations.
in the regulation of the insulin receptor (22,23), as shown in Table I. It only weakly inhibited CD45. The IC 50 value for CD45 is larger than that in our previous report (27). This difference occurred because we changed the enzyme source to a commercially available preparation from the crude preparation of Jurkat cell membranes. Et-3,4-dephostatin may inhibit other untested PTPases; therefore, involvement of PTPases other than PTP-1B and SHPTP-1 cannot be excluded in its cellular effects.
Et-3,4-dephostatin increased the phosphorylation of Akt, GLUT4 translocation, and hexose transport in 3T3-L1 adipocytes. 4-O-Me-Et-3,4-dephostatin did not induce these phenotypic changes, indicating that these effects are due to the inhibition of PTPases. Since protein synthesis was blocked by the addition of cycloheximide, the increase of hexose transport shown in Fig. 5 should not be due to the increase of glucose transporters. Vanadate was also reported to prolong insulin action by increasing the tyrosine phosphorylation of the insulin receptor (34). Vanadate activated PI 3-kinase-dependent glucose transport (35). But Et-3,4-dephostatin appears more effective than vanadate, as is shown in Fig. 5B.
Unexpectedly, Et-3,4-dephostatin alone mimicked insulin action to induce several phenotypic changes. Compared with the quick action of insulin in 30 min, Et-3,4-dephostatin alone increased the tyrosine phosphorylation of the insulin receptor and IRS-1 rather slowly in 6 h, as shown in Fig. 2. This is because accumulation of phosphorylation by inhibition of PTPases takes time. Et-3,4-dephostatin strongly induced GLUT4 translocation and glucose transport. Therefore, Et-3,4dephostatin may inhibit not only insulin receptor dephosphorylation but also dephosphorylation of other components that regulate glucose uptake. In fact, Et-3,4-dephostatin induced tyrosine phosphorylation in several other proteins, including c-Cbl, PLC␥, PI 3-kinase, and IRS-3. In Fig. 2, Et-3,4-dephostatin unexpectedly lowered the insulin-induced tyrosine phosphorylation of insulin receptor and IRS-1 in a 30-min incubation. Insulin is known to activate PTPases such as PTP-1B (36) that can dephosphorylate both insulin and IRS-1. Et-3,4-dephostatin might enhance induction of the inhibitor-insensitive PTPases by insulin in a 30-min incubation. On the other hand, Et-3,4-dephostatin did not affect insulin-induced activation of hexose transport even in a 30-min incubation (Fig. 5A). This is because Et-3,4-dephostatin might also inhibit tyrosine dephosphorylation of downstream components.
Insulin-induced glucose transport is mediated largely by PI 3-kinase (37) and partly by p38 (38). On the other hand, the Ras/mitogen-activated protein kinase pathway and protein synthesis are not necessary in the effect of insulin. In Fig. 6A, insulin-induced glucose uptake was inhibited by LY294002 and SB203580 but not by a MEK inhibitor or by cycloheximide. Interestingly, LY294002 inhibited insulin-induced glucose transport and GLUT4 translocation completely in our assay system; but when Et-3,4-dephostatin was used, the inhibitor only partially suppressed glucose transport and GLUT4 translocation in a 6-h incubation (Fig. 6A and 7). On the other hand, in a 30-min incubation, LY294002 completely blocked the increase of glucose transport by Et-3,4-dephostatin, as shown in Fig. 6B. Therefore, it is likely that Et-3,4-dephostatin activates a PI 3-kinase-independent pathway in addition to the PI 3-kinase/Akt pathway only in long term incubation. Very recently, Pessin and co-workers (21) showed the existence of a PI 3-kinase-independent c-Cbl pathway in insulin-induced signal transduction. The proto-oncogene product c-Cbl is known to be a component of protein-tyrosine kinase-mediated signaling cascades downstream of the activated cell surface receptors. These receptors include the T cell, B cell, and cytokine receptors (39). The c-Cbl protein contains the RING finger domain, an extensive proline-rich region that provides binding sites for the SH3 domain, and the C-terminal leucine zipper domain that shows significant homology to ubiquitin-associated proteins (40). In hematopoietic cells, c-Cbl is known to be a negative regulator of Syk tyrosine kinase through the RING finger domain (41). In 3T3-L1 cells, c-Cbl is prominently tyrosine-phosphorylated in response to insulin in 3T3-L1 adipocytes and not in 3T3-L1 fibroblasts (42). The tyrosine-phosphorylated c-Cbl protein associates with Cbl-associated protein at the site of caveola, and this protein complex up-regulates GLUT4 translocation through the PI 3-kinase independent pathway (21). Therefore, it is possible that Et-3,4-dephostatin inhibits PTPases to protect phosphorylated c-Cbl from dephosphorylation. Actually, Et-3,4-dephostatin markedly enhanced tyrosine phosphorylation of c-Cbl, as shown in Fig. 8. Insulin-induced c-Cbl tyrosine phosphorylation greatly decreased at 6 h, but when Et-3,4dephostatin was added, the phosphorylation strongly remained. Et-3,4-dephostatin alone also induced phosphorylation of c-Cbl. Thus, it is likely that Et-3,4-dephostatin potentiates both PI 3-kinase/Akt-dependent and PI 3-kinase/Akt-independent pathways including activation of c-Cbl.
Kayali et al. (25) reported that phospholipase C-␥ is associated with the insulin receptor and can be phosphorylated by the receptor. Therefore, if a PI 3-kinase is not involved downstream of phospholipase C-␥, activation of phospholipase C-␥ by inhibition of dephosphorylation may also be a possible mechanism.
When Et-3,4-dephostatin was orally given to mice with high blood glucose, it lowered their blood glucose significantly, as shown in Fig. 9. Therefore, Et-3,4-dephostatin is likely to be absorbed from the intestine and to be stable in the body. During the in vivo experiment, no toxicity, including the loss of body weight, was observed. In 24 h, there was no marked difference in food uptake between the test and control group. There was also no abnormality in coat of fur, behavior, and FIG. 8. Effect of LY294002 on insulin or Et-3,4-dephostatininduced intracellular c-Cbl tyrosine phosphorylation. After 100 M, LY294002 was pretreated for 30 min, 3T3-L1 adipocytes were treated with 10 nM insulin and/or 10 g/ml Et-3,4-dephostatin for 6 h, and then the cells were lysed with the lysis buffer and immunoprecipitated with monoclonal anti-c-Cbl antibody. Following electrophoretic transfer to the membranes, the proteins were immunoblotted with anti-phosphotyrosine antibody.
FIG. 9. Decrease in blood glucose level in Et-3,4-dephostatintreated KK-A y mice. KK-A y mice were treated with a CE-2 diet containing 12% sucrose and Et-3,4-dephostatin for 24 h. A blood sample was then taken from the orbital capillary bed, and its glucose level was measured. The results are representative of two independent experiments.
feces. The methoxime-type derivative of dephostatin (29) belongs to the PTPase inhibitors of the second generation. Long term experiments with this compound also decreased the blood glucose level in mice without toxicity. 2 Thiazolidine compounds are being widely used clinically for the treatment of type 2 diabetes mellitus. Its target is known to be peroxisome proliferator-activated receptor ␥, which is a unique transcription factor for adipocyte differentiation. In our assay system shown under "Experimental Procedures," troglitazone significantly enhanced differentiation of 3T3-L1 cells into adipocytes. However, Et-3,4-dephostatin did not increase the differentiation significantly (data not shown). Therefore, the mechanism of the antidiabetic effect by Et-3,4-dephostatin in vivo should be different from that by troglitazone, and its side effects may not include obesity, unlike the case for troglitazone.
Thus, Et-3,4-dephostatin potentiated the insulin-related signal transduction in cultured 3T3-L1 adipocytes and showed antidiabetic effects in mice. Especially, it was orally active in vivo. Since the nitrosamine group in Et-3,4-dephostatin may be mutagenic and carcinogenic, a nitrosamine-free analogue of Et-3,4-dephostatin (29) may be a prototype of new antidiabetic agents.