Association of SH2-Containing Inositol Phosphatase 2 With the Insulin Resistance of Diabetic db/db Mice

Human crystal insulin was provided by Novo Nordisk Pharmaceutical (Copenhagen, Denmark). [γ-³²P]ATP (111 TBq/mmol) was purchased from NEN Life Science Products (Boston, MA). Two polyclonal anti-SHIP-2 antibodies were described previously (12). A polyclonal anti-Ser (21/9) phosphospecific GSK-3 antibody, a polyclonal anti-Thr (410/403) phosphospecific PKC-ζ/λ antibody, and the Akt kinase assay kit were obtained from New England Biolabs (Beverly, MA). A polyclonal anti-Akt antibody and a polyclonal anti-PKC-ζ/λ antibody were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). A monoclonal anti-phosphotyrosine antibody (PY20) and a monoclonal anti-p85 antibody of PI 3-kinase were obtained from Transduction Laboratories (Lexington, KY). Enhanced chemiluminescence reagents were obtained from Amersham Pharmacia Biotech (Uppsala, Sweden). All other routine reagents were analytical grade and were purchased from Sigma Chemicals (St. Louis, MO) or Wako Pure Chemical Industries (Osaka, Japan).

Male C57BL/KsJ-db/db mice and their lean heterozygote littermates (db/+m) and C57BL/6J were purchased from Clea Japan (Tokyo, Japan) at age 6 weeks and maintained under standard light (12 h light/dark) and temperature conditions. These mice were caged in groups of four and were provided with food and water ad libitum, and were used for the experiments at approximately ages 8–12 weeks. The mice group with a high-fat diet were fed for 2 weeks with a diet containing 32% safflower oil, 33.1% casein, 0.5% dl-methionine, 17.6% sucrose, 1.4% vitamin, 9.8% mineral mixture, and 5.6% cellulose powder (28). The treatment group of db/db and db/+m mice received rosiglitazone, provided from SmithKline Beecham, in their food at a level of 0.005%, which resulted in the ingestion of 10 mg · kg⁻¹ · day⁻¹ for 2 weeks.

Food was withdrawn 12 h before the experiments, and mice were intraperitoneally injected with saline or insulin (5 mU/g body wt) in saline. After 10 min, the mice were killed, and their quadriceps muscle, epididymal fat tissue, and liver were rapidly excised. Then 50 mg of these tissues were homogenized using a polytron at half maximum speed (15,000 rpm) for 1 min on ice in 500 μl of a homogenization buffer containing 20 mmol/l Tris, 5 mmol/l EDTA, 10 mmol/l Na₄P₂O₇, 100 mmol/l NaF, 2 mmol/l Na₃VO₄, 1% NP-40, 1 mmol/l phenylmethylsulfonyl fluoride (PMSF), 10 μg/ml aprotinin, and 10 μg/ml leupeptin (pH 7.5). The tissue lysates were solubilized by continuous stirring for 1 h at 4°C, and centrifuged for 10 min at 14,000g. The supernatants were used for the immunoblot analysis (23,29). For total membrane preparation, the tissue samples were homogenized with the polytron homogenizer in HEPES EDTA sucrose (HES) buffer containing 20 mmol/l HEPES, 5 mmol/l EDTA, 250 mmol/l sucrose, 50 nmol/l okadaic acid, 1 mmol/l Na₃VO₄, 2 μg/ml pepstatin, 1 mmol/l PMSF, 10 μg/ml aprotinin, and 2 μg/ml leupeptin (pH 7.5) at 4°C. The relative volumes of the tissues to HES buffer were 1:3 for the adipose tissue and 1:10 for skeletal muscle and liver. The homogenate was centrifuged at 1,200g at −4°C for 15 min, and the fat cake and pellet were discarded. The infranatant was then centrifuged at 220,000g for 90 min at 4°C. The pellet containing the total membrane fraction was resuspended in the HES buffer and subjected to immunoblot analysis (23,29). Blood samples were also collected at the time the mice were killed. Serum glucose was measured by an automated glucose analyzer (A&T, Tokyo, Japan) and serum insulin was detected by an enzyme-linked immunosorbent assay kit (MIA Institute of Biological Science, Yokohama, Japan). All procedures were approved by the Committee of Animal Experiment at Toyama Medical & Pharmaceutical University.

The equal protein amount of the samples was separated by 7.5% SDS-PAGE and transferred onto polyvinylidene difluoride membranes (PVDMs) using a Bio-Rad Transblot apparatus. The membranes were blocked in a buffer containing 50 mmol/l Tris, 150 mmol/l NaCl, 0.1% Tween 20, and 2.5% BSA or 5% non-fat milk (pH 7.5) for 2 h at 20°C. The membranes were then probed with the specified antibodies for 2 h at 20°C or for 16 h at 4°C. After washing the membranes in a buffer containing 50 mmol/l Tris, 150 mmol/l NaCl, and 0.1% Tween 20 (pH 7.5), blots were incubated with horseradish peroxidase-linked second antibody and then examined by enhanced chemiluminescence (ECL) detection using an ECL reagent according to the manufacturer’s instructions (Amersham) (13,14).

The tissue lysates (0.5 mg protein) were immunoprecipitated with anti-phosphotyrosine antibody (PY20) coupled to protein G-sepharose for 16 h at 4°C. The immune complex was washed twice with PBS containing 1% NP-40, 100 μmol/l Na₃VO₄, and 1 mmol/l dithiothreitol (DTT); twice with 100 mmol/l Tris-HCl containing 500 mmol/l LiCl₂, 100 μmol/l Na₃VO₄, and 1 mmol/l DTT (pH 7.5); and twice with 10 mmol/l Tris containing 0.1 mol/l NaCl, 1 mmol/l EDTA, and 1 mmol/l DTT (pH 7.5). The pellet was resuspended in 10 μl Tris-NaCl buffer. The PI 3-kinase reaction was started by adding 20 μl of PI solution containing 0.5 mg PI/ml, 50 mmol/l HEPES, 1 mmol/l NaH₂PO₄, and 1 mmol/l EGTA (pH 7.6) at 20°C, followed by the addition of 10 μl of a reaction mixture containing 250 μmol/l [γ-³²P]ATP (0.37 MBq/tube), 100 mmol/l HEPES, and 50 mmol/l MgCl₂ (pH 7.6) for 5 min. The reaction was stopped by the addition of 15 μl of 8N HCl. The products were extracted by adding 130 μl chloroform-methanol (1:1) followed by centrifugation. The organic phase was removed and spotted on a silica gel thin-layer chromatography plate (Merck), and the plates were developed and dried. The phosphorylated inositol was visualized by autoradiography and quantitated with the BAS 2000 image analyzer (Fuji Film, Tokyo, Japan) (13,14).

The Akt kinase assay was performed using an Akt kinase assay kit (New England Biolabs, Beverly, MA). In brief, the tissue lysates were immunoprecipitated with immobilized anti-Akt antibody for 3 h at 4°C. The precipitates were washed twice with the cell lysis buffer and twice with a kinase buffer containing 25 mmol/l Tris, 5 mmol/l β-glycerophosphate, 2 mmol/l DTT, 0.1 mmol/l Na₃VO₄, and 10 mmol/l MgCl₂ (pH 7.5). Then 40 μl of the pellets were suspended with 200 μmol/l ATP and 1 μg GSK-3 fusion protein and incubated for 30 min at 30°C. The reaction was terminated by adding SDS sample buffer containing 187.5 mmol/l Tris, 6% wt/vol SDS, 30% glycerol, 150 mmol/l DTT, and 0.03% wt/vol bromophenol blue (pH 6.8). The samples were then separated by 12% SDS-PAGE and transferred onto PVDMs. The membranes were blocked and probed with anti-Ser^21/9-phosphospecific GSK-3 antibody for 16 h at 4°C. After the membranes were washed, the blots were incubated with horseradish peroxidase-linked second antibody then examined by ECL using ECL reagent according to the manufacturer’s instructions (Amersham) (13,14).

Data are presented as means ± SE. P values were determined by a paired t test or Bonferroni test with ANOVA; P < 0.05 was considered statistically significant.

The db/db mouse is known to be an insulin-resistant diabetic animal model (27). We examined the expression level of SHIP-2 in skeletal muscle, epididymal fat tissue, and liver of db/db and db/+m mice. SHIP-2 was expressed in the quadriceps muscle, epididymal fat tissue, and liver of db/+m mice age 8 weeks. Similar amounts of SHIP-2 were also expressed in these tissues in C57BL/6J mice age 8 weeks (data not shown). Interestingly, the amount of SHIP-2 protein was increased in the skeletal muscle and fat tissue, but not in the liver of db/db mice compared with db/+m mice (Fig. 1A). Thus the amount of expressed SHIP-2 protein in the skeletal muscle and fat tissue of db/db mice was significantly increased by 45.4 ± 7.9 and 105.1 ± 13.5%, respectively, compared with db/+m mice (Fig. 1B). The increase in SHIP-2 expression was already seen in these tissues of db/db mice at age 4 weeks (data not shown). The amount of expressed SHIP-2 protein in db/+m mice fed the high-fat diet was examined to investigate whether the increase seen was limited to db/db mice. The amount of expressed SHIP-2 protein was also increased in the skeletal muscle and fat tissue of db/+m mice with insulin resistance induced by a high-fat diet (Fig. 1C). The expression of SHIP-2 was increased by 52.3 ± 6.5 and 81.4 ± 8.9%, respectively, compared with control db/+m mice (Fig. 1D).

After ligand stimulation, SHIP-2 translocates to the plasma membrane where SHIP-2 can access its substrate PI(3,4,5)P3 for hydrolysis to PI(3,4)P2 (30,31). We next compared the amount of SHIP-2 in the total membrane preparations in the skeletal muscle, fat tissue, and liver of db/db mice with that of db/+m mice. Insulin stimulated translocation of SHIP-2 to the membrane preparation in all three tissues of db/+m mice. The amount of membrane-associated SHIP-2 was increased in the skeletal muscle and fat tissue, but not in the liver of db/db mice compared with db/+m mice (Fig. 2A). The amount of SHIP-2 located at the membrane was increased by 40.6 ± 7.9% in the skeletal muscle and 82.3 ± 8.4% in the fat tissue of db/db mice, respectively, compared with db/+m mice. Thus, the increase of SHIP-2 at the total membrane preparation is well correlated with the elevated amount of total SHIP-2 (Fig. 2B).

Insulin-induced PI 3-kinase activation is the key for the biological action of insulin (1–3). Insulin stimulation increased PI 3-kinase activity by 2.6 ± 0.1-fold in the skeletal muscle, 2.3 ± 0.3-fold in the fat tissue, and 2.8 ± 0.7-fold in the liver of db/+m mice. Insulin-induced PI 3-kinase activity was decreased by 38.1 ± 6.4, 48.2 ± 4.2, and 54.3 ± 9.7%, respectively, in db/db mice. Therefore, the moderate decrease in insulin-induced PI 3-kinase activation was seen in all three tissues of db/db mice compared with db/+m mice (Figs. 3A and 4). Akt and atypical PKC are downstream target molecules of PI 3-kinase, which are important for further transmission of the metabolic signaling of insulin (8,32–34). Because the expression of SHIP-2 was enhanced in the skeletal muscle and fat tissue of db/db mice, we examined how insulin-stimulated activity of Akt and PKC-ζ/λ was affected in db/db mice compared with db/+m mice. Insulin induced Akt activation by 2.8 ± 0.3-fold in the skeletal muscle, 2.6 ± 0.4-fold in the fat tissue, and 2.8 ± 0.6-fold in the liver of db/+m mice. Insulin-induced Akt activation was decreased by 59.0 ± 3.2, 65.0 ± 4.4, and 58.3 ± 5.2%, respectively, in db/db mice. Thus, insulin-induced Akt activation in the skeletal muscle and fat tissue, but not in the liver, was more severely affected than PI 3-kinase in db/db mice (Figs. 3B and 4). On the other hand, insulin increased atypical PKC phosphorylation by 2.2 ± 0.4-fold in the skeletal muscle, 2.3 ± 0.3-fold in the fat tissue, and 2.2 ± 0.5-fold in the liver of db/+m mice. PKC-ζ/λ phosphorylation was decreased by 88.5 ± 3.6, 86.3 ± 4.0, and 51.0 ± 5.5%, respectively, in db/db mice. Apparently, insulin-induced phosphorylation of PKC-ζ/λ in the skeletal muscle and fat tissue, but not in the liver, was decreased more markedly than PI 3-kinase in db/db mice (Figs. 3C and 4). The amount of the p85 regulatory subunit of PI 3-kinase, PKB, and PKC-ζ/λ in the skeletal muscle, fat tissue, and liver was similar between db/db mice and db/+m mice (data not shown). These results indicate that the activity of molecules downstream of PI 3-kinase in the skeletal muscle and fat tissue of db/db mice is decreased, at least in part, by increased expression of SHIP-2.

The db/db mice in this study were obese, hyperglycemic, and hyperinsulinemic compared with the control db/+m mice. The 2-week treatment with rosiglitazone resulted in a reduction in serum glucose and insulin levels in db/db mice, but did not affect these levels in db/+m mice. Rosiglitazone treatment did not affect body weight in either db/db mice or db/+m mice (Table 1). We next examined whether rosiglitazone affects the amount of SHIP-2 protein in these tissues of db/db mice and db/+m mice. Rosiglitazone treatment had a tendency (which was not statistically significant) to decrease the amount of SHIP-2 in the skeletal muscle and fat tissue of db/+m mice. Importantly, treatment with rosiglitazone significantly reduced the amount of SHIP-2 protein in the skeletal muscle and fat tissue, but not in the liver, of db/db mice (Fig. 5A). The amount of SHIP-2 protein was decreased by 46.1 ± 4.3% in the skeletal muscle and 50.2 ± 5.1% in the fat tissue of db/db mice (Fig. 5B).

Because rosiglitazone treatment decreased the amount of elevated SHIP-2 in the skeletal muscle and fat tissue of db/db mice, we further examined the effect of rosiglitazone on insulin-induced activation of Akt and PKC-ζ/λ compared with PI 3-kinase. Rosiglitazone treatment did not significantly affect insulin-induced PI 3-kinase activity, Akt activity, or PKC-ζ/λ phosphorylation in all three tissues of db/+m mice (data not shown). On the other hand, treatment with rosiglitazone partially restored insulin-induced PI 3-kinase activation in the skeletal muscle and fat tissue, but not in the liver, of db/db mice. It is interesting that, in accordance with the reduction in the amount of SHIP-2 by rosiglitazone, insulin-induced Akt activation and PKC-ζ/λ phosphorylation were completely restored to the control level in the skeletal muscle and fat tissue, but not in the liver, of db/db mice (Figs. 3 and 4). The amount of the p85 regulatory subunit of PI 3-kinase, Akt, and PKC-ζ/λ in the skeletal muscle, fat tissue, and liver was not affected by treatment with rosiglitazone in either db/db mice or db/+m mice (data not shown). These results further indicate that SHIP-2 is an important factor associated with the insulin resistance in db/db mice.

It is reported that targeted disruption of the SHIP-2 gene in mice resulted in increased insulin sensitivity without affecting biological systems other than insulin signaling (15). As the molecular mechanism by which SHIP-2 is involved in the regulation of insulin signaling, we have reported that overexpression of SHIP-2 inhibits insulin-induced activation of Akt and atypical PKC via the 5′-phosphatase activity of SHIP-2, and that PI(3,4,5)P3 rather than PI(3,4)P2 is important for further transmission of insulin signaling via PI 3-kinase in 3T3-L1 adipocytes and L6 myotubes (13,14). These previous findings indicated that SHIP-2 is a physiologically important negative regulator of insulin signaling. In the present study, we demonstrate that the amount of SHIP-2 protein is increased in the skeletal muscle and fat tissue of obese, insulin-resistant, diabetic db/db mice. In association with the increased amount of SHIP-2 protein, insulin signaling appears to be impaired, at least in part, at the level distal to PI 3-kinase in the activation of Akt and atypical PKC. Taken together, our findings appear to indicate that SHIP-2 is not only involved in the physiological regulation of insulin signaling, but is also associated with the state of insulin resistance. However, we cannot rule out the possibility that a primary defect exists at the level of phosphoinositide dependent kinase 1 or the activation of Ser/Thr phosphatase, such as PP2A, in the regulation of Akt and atypical PKC in db/db mice. The increased SHIP-2 expression seen in the skeletal muscle and fat tissue was not limited to obese, hyperglycemic, insulin-resistant db/db mice. The increased amount of SHIP-2 was also observed in mice with insulin resistance induced by a high-fat diet. Therefore, the increased expression of SHIP-2 might be a relatively common phenomenon seen in insulin resistance. Further investigations, including possible alteration of SHIP-2 expression in human type 2 diabetes, are required to clarify this issue.

It is known that the total cellular 5′-phosphatase activity of SHIP-2 is not altered by ligand stimulation (30,31). Instead, SHIP-2 translocates from cytosol to plasma membrane compartment upon ligand stimulation, whereby SHIP-2 is able to access to its substrate PI(3,4,5)P3 for hydrolysis to PI(3,4)P2 (30,31). Based on these previous reports, it is possible that the function of SHIP-2 is also affected by the altered localization of SHIP-2 in db/db mice. However, this does not appear to be the case, as the amount of SHIP-2 in the total membrane preparation was elevated in parallel with the increased amount of total SHIP-2 in db/db mice compared with control db/+m mice. Because the amount of SHIP-2 in the membrane is normal or supernormal after insulin stimulation, the insulin-induced process of the translocation of SHIP-2 does not appear to be impaired in the skeletal muscle and fat tissue of db/db mice. Taken together, the increased amount of total SHIP-2 appears to be a predominant alteration seen in the skeletal muscle and fat tissue of db/db mice. Although more studies will be required to examine the possible functional alteration of SHIP-2 in the state of insulin resistance, this notion is further supported by the fact that rosiglitazone treatment decreased the elevated expression of SHIP-2 together with improved insulin signaling. Importantly, rosiglitazone treatment efficiently improved insulin-induced activation of Akt and PKC-ζ/λ, whereas the decreased PI 3-kinase activity in the skeletal muscle and fat tissue of db/db mice was only moderately restored by the treatment. These findings further indicated that SHIP-2 is associated with impaired insulin signaling at the level of Akt and PKC-ζ/λ activation, an association that is at least in part independent of the PI 3-kinase in the state of insulin resistance. In addition, they demonstrated that SHIP-2 is a novel target molecule of thiazolidinediones in the treatment of insulin resistance. Rosiglitazone also appeared to decrease the amount of SHIP-2 in the skeletal muscle and fat tissue of control db/+m mice, although this decrease was not statistically significant. In contrast to the results with db/db mice, the possible decrease in the amount of SHIP-2 did not significantly affect the metabolic parameters in db/+m mice. Although the reason is uncertain, the effect of the reduction of SHIP-2 amounts by treatment with rosiglitazone is likely to have great implication for improving the metabolic parameters in mice with insulin resistance.

It is uncertain which factor(s) implicated in insulin resistance regulates SHIP-2 expression in the skeletal muscle and fat tissue of db/db mice. Enhanced expression of SHIP-2 may be a compensatory mechanism to limit insulin signaling against chronic hyperinsulinemia seen in db/db mice, as SHIP-2 is the physiological negative regulator of insulin signaling (13–15). Alternatively, it is also possible to consider the more direct implication of SHIP-2 in insulin resistance. In type 2 diabetes, hyperglycemia contributes to the development of insulin resistance (1,2,19,20). It has been reported that insulin-induced PI 3-kinase activity is relatively reduced in soleus muscle, but is not impaired in EDL muscle of GK rats, whereas Akt activity and glucose transport are severely impaired in both types of muscles (18). It should be noted that Akt activity was fully restored to control levels concomitantly with improved glycemic control (18). In addition, insulin-induced Akt activity has been reported to be severely suppressed, whereas PI 3-kinase activity is increased in ZDF rat liver (19). Interestingly, the restoration to normoglycemia by sodium-dependent glucose transporter inhibitor normalized impaired insulin-induced Akt activation (19). These reports indicate that steps in insulin signaling downstream of PI 3-kinase appear to be sensitive to changes in the glycemic milieu, although the mechanism by which this occurs is unknown. Along this line, SHIP-2 appears to be a candidate factor associated with the insulin resistance with hyperglycemia. To clarify this issue, we incubated the cells in medium containing high glucose concentrations for 24 h and found an increased expression of SHIP-2 in 3T3-L1 adipocytes. Treatment with rosiglitazone partially reduced the elevated amount of SHIP-2 (K. Fukui, T. Sasaoka, and M. Kobayashi, unpublished observations). Therefore, hyperglycemia appears to partially affect the expression of SHIP-2 in skeletal muscle and fat tissue of db/db mice.

On the other hand, it has been reported that decreased effects of insulin-induced glucose transport in the skeletal muscle and adipocytes of GK-diabetic rats are attributable to markedly diminished phosphorylation and activation of PKC-ζ/λ, at least in part, independent of PI 3-kinase activity (23,24). Thiazolidinediones enhance insulin-induced glucose transport in the skeletal muscle and adipocyte of diabetic GK rats through increased phosphorylation and activation of PKC-ζ/λ (23,24). Improvement in insulin-induced PKC-ζ/λ activation occurs after rosiglitazone treatment, despite only modest decreases in serum glucose levels (23,24). Thus, even though serum glucose levels in diabetic db/db mice were markedly improved by treatment with rosiglitazone in the current study, we may need to consider additional factors other than hyperglycemia when determining the effect of rosiglitazone on the restoration of PKC-ζ/λ activation via the change of SHIP-2 expression. In view of this, recent research has focused on the role of adipocytokines secreted from adipocytes (1). Leptin, tumor necrosis factor-α, plasminogen-activator inhibitor type 1, adipsin, resistin, and adiponectin are adipocytokines known to be implicated in the insulin resistance (1). Clearly these are interesting candidates that are possibly implicated in the regulation of SHIP-2 expression. Future studies are needed to clarify the mechanism by which the expression of SHIP-2 is regulated in the state of insulin resistance.

In summary, the amount of SHIP-2 protein was increased in the skeletal muscle and fat tissue of db/db mice. Insulin-induced activity of molecules downstream of PI 3-kinase appeared to be decreased, at least in part, by the elevated expression of SHIP-2. Rosiglitazone ameliorated the elevated expression of SHIP-2 with marked improvement of insulin-induced Akt and PKC-ζ/λ activation. Because SHIP-2 appears to be associated with the insulin resistance, in addition to its physiological role in the negative regulation of insulin signaling, SHIP-2 would be a potential therapeutic target molecule for the treatment of insulin resistance in type 2 diabetes.

This work was supported in part by a Grant-in-Aid for Scientific Research (C) from the Japan Society for the Promotion of Science (T.S.).

We thank Drs. Takayoshi Sugihara, Hiroshi Tsuneki, Kosei Kikuchi, Mika Kishida, and Ikuko Kimura for their technical assistance and comments in the tissue preparation of animals.