Pioglitazone Increases Insulin Sensitivity by Activating Insulin Receptor Kinase MASASHI KOBAYASHI, MASANORI IWANISHI, KATSUYA EGAWA, AND YUKIO SHIGETA
A new oral agent, 5-[4-(2-(5- ethyl 12-pyridyl)ethoxy]benzoyl]-2,4-thiazolidinedione (pioglitazone), has been developed for treatment of non-insulin-dependent diabetes mellitus (NIDDM). This agent increases insulin sensitivity in vivo in genetically obese Wistar fatty rats. Administration of the agent (3 mg/kg/day) for 10 days to the rats ameliorated hyperglycemia and hyperinsulinemia, indicating that it decreased insulin resistance. To clarify the mechanism of the drug to increase insulin sensitivity, we examined insulin binding and kinase activity of insulin receptors from muscles of both untreated and treated rats. Pioglitazone treatment did not change insulin binding in Wistar fatty rats but increased insulin-stimulated autophosphorylation of insulin receptors to 78% over the level in the control but not the basal state. Kinase activity toward exogenous substrate, poly Glu 4 Tyr\ was also increased to 87% over the level of untreated control obese rats. In contrast, in lean rats, pioglitazone treatment did not increase autophosphorylation and kinase activity toward exogenous substrates. To further elucidate the mechanism, we incubated insulin receptors with the agent and measured kinase activity. Incubation of solubilized receptors with the agent did not increase kinase activity. However, the receptors from IM-9 cells, which were incubated with 10~ 8 M pioglitazone for 7 days, showed a 46% increase over the control in insulin-stimulated autophosphorylation and kinase activity. These results suggested that pioglitazone increased insulin sensitivity in part by activating kinase of the receptors through indirect effect on insulin receptors and that the drug may have useful benefits in insulin resistance of NIDDM. Diabetes 41:476-83, 1992
From the Third Department of Medicine, Shiga University of Medical Science, Ohtsu, Japan. Address correspondence and reprint requests to Masashi Kobayashi, MD, First Department of Medicine, Toyama Medical and Pharmaceutical University, Sugitani, Toyama 930-01, Japan. Received for publication 7 May 1991 and accepted in revised form 18 December 1991.
476
T
he insulin receptor consists of two a-subunits that bind insulin and two (3-subunits that have tyrosine-specific kinase activity and respond to insulin by autophosphorylation and the phosphorylation of exogenous substrates (1-3). The importance of insulin receptor function for glucose metabolism is highlighted by the insulin receptor disease characterized by markedly decreased insulin binding or kinase defect due to genetic abnormalities (4-6). Thus, the insulin receptor function should be evaluated in various insulin resistance states. Non-insulin-dependent diabetes mellitus (NIDDM) is characterized by insulin resistance along with decreased insulin secretion and is often associated with obesity. Previous studies on insulin receptor function of NIDDM revealed decreased kinase activity of p-subunit, which may be in part responsible for decreased insulin action in NIDDM (7-9). Therefore, the best way to treat patients with NIDDM is to increase insulin sensitivity and insulin secretion. For this reason, sulfonylurea has been used, and the mechanism of increased insulin sensitivity by sulfonylurea has been investigated. The initial step of insulin action, i.e., insulin binding and kinase of insulin receptors, is not directly influenced by the agent, indicating that its action appears to be at the postkinase step (10,11). Pioglitazone, a thiazolidine derivative that has potent activity to increase insulin sensitivity, has been developed for treatment of NIDDM. Previous studies showed that diabetic obese rats could be treated successfully with this oral agent and that insulin resistance of these rats was ameliorated (12,13). To clarify the mechanism of the agent, we examined the function of insulin receptors by measuring the insulin binding and receptor kinase activity of insulin receptors from the pioglitazone-treated rats. We report that pioglitazone is a potent agent that increases insulin sensitivity by activating kinase of insulin receptors, which sulfonylurea does not exhibit, and that it
DIABETES, VOL. 41, APRIL 1992
KOBAYASHI AND ASSOCIATES
C2H5
•o
CH 2 CH 2 O FIG. 1. Structure of pioglitazone.
could be a potentially effective oral agent with a new mode of action for NIDDM. RESEARCH DESIGN AND METHODS Wistar fatty rats {fa/fa) were bred by mating heterozygous lean rats (fa/+) as reported previously (14). Eleven-weekold male rats with elevated plasma glucose, triglyceride, and insulin levels were used to examine the effects of pioglitazone on insulin receptor binding and insulin receptor kinase activity in muscles (Fig. 1). Lean rats {fa/+) were also used to study the effects of the agent on insulin receptor function for comparison. Pioglitazone was administered by a stomach tube after the compound was suspended in 5% gum arabic solution. Three and 10 mg/kg of the agent was given to fatty and lean rats for 10 days, respectively. Previous studies revealed increased activity of the agent toward obese rats and relative ineffectiveness toward lean rats (12,13). In lean rats, 3- to 10-mg/kg dosage did not significantly affect insulin sensitivity or produce toxic effects. Therefore, considering the lean mass of these rats and the more clear-cut results of the effect of the agent, decreased amount of pioglitazone (i.e., 3 mg/kg), was given to the fatty rats, which was enough to ameliorate insulin resistance. The rats were killed by decapitation and blood samples were collected for measurement of blood glucose, insulin, and triglyceride levels. Purified pork insulin was a gift from Shimizu (Shizuoka, Japan). [ 125 I] Na and 1000-3000 Ci/mmol [ 7 - 32 P]ATP were purchased from Du Pont-NEN (Boston, Massachussetts). Fatty acid-free bovine serum albumin and 8 0 0 0 - H polyethylene glycol were from Sigma (St. Louis, MO). Disuccinimidyl suberate was from Pierce (Rockford, IL). Protein A (pansorbin) was from Calbiochem (La Jolla,
CA). Wheat-germ agglutinin (WGA) agarose was from Pharmacia (Uppsala, Sweden). Anti-insulin antibody was obtained from a type B insulin-resistant patient with Sjogren's syndrome. Anti-phosphotyrosine antibody was kindly supplied by Dr. H. Fujio (Research Institute for Microbial Diseases, Osaka University, Osaka, Japan). Preparation of partially purified insulin receptor Hindlimb muscles were trimmed of fat and immediately frozen in liquid N 2 . The frozen muscle was finely ground in a motor precooled to - 2 0 ° C and an ice-cold buffer consisting of 25 mM HEPES, 1 % Triton X-100, 4 mM EDTA, 1 typsin inhibitory unit (TIU)/ml aprotinin, and 2 mM phenylmethylsulfonyl fluoride (pH 7.4) as described by Burant et al. (15). The frozen slurry was homogenized as it thawed in a Potter-Elvejhem homogenizer and then centrifuged at 10,000 x g for 10 min at 4°C. The resulting supernate was slowly stirred at 21 °C for 60 min and then centrifuged at 150,000 x g for 90 min at 4°C. The supernate was applied to WGA agarose column. The column was washed with 25 mM HEPES and 0 . 1 % Triton X-100, pH 7.4. Receptors were eluted with 0.3 M A/-acetyl glucosamine in the buffer for washing the column. Insulin binding to solubilized receptors. Lectin- purified insulin receptors (30 |xl) were incubated with 1 2 5 I labeled insulin at 4°C for 16 h in the presence of various concentrations of unlabeled insulin in 200 uJ of 25 mM HEPES (pH 7.6) containing 0.05% Triton X-100, 15 mM NaCI, 0.1 mg/ml bovine serum albumin, and 150 mM N-acetyl-D-glucosamine. With human 7-globulin as carrier protein, receptor-bound insulin was precipitated with polyethylene glycol as previously described (16). Nonspecific binding was defined as the radioactivity precipitated in the presence of 3.1 \xM unlabeled insulin.
TABLE 1 Characteristics of rats Concentration in plasma
Lean Control Treated Fatty Control Treated
Body weight (g)
Glucose (mM)
Triglyceride (mM)
Nonesterified fatty acid (M-Eq/L)
Insulin (M.U/ml)
Liver weight (g)
Muscle weight (g)
235 ± 10 245 ± 10*
7.6 ± 0.3 7.3 ±0.1
5.4 ± 0.8 4.6 ± 0.4+
84 ± 2 1 59 ± 36*
92 ± 6 66 ± 15*
11.6 ±0.8 12.1 ±0.8*
11.2 ±0.9 11.2 ±0.7*
284 ± 9§ 290 ± 9*
9.6 ± 1.1§ 6.6 ± 0.2||
17.6±4.1§ 7.1 ± 0.9||
136±21§ 93 ± 19*
368 ± 45§ 220 ± 23||
15.0±0.7§ 14.3 ±0.7*
8.3 ± 1.5* 9.3 ±0.9*
Values are means ± SD determined 20 h after administration of last dosage of pioglitazone for 6 rats/group. Pioglitazone was administered orally to 6-wk-old male Wistar lean and fatty rats for 10 days. Ten and 3 mg • kg" 1 • day" 1 pioglitazone was administered to lean and fatty rats, respectively. *Not significant, lean vs. fatty and treated vs. control rats. t P < 0.05, $P < 0.01, vs. control rats. §P < 0.001, lean vs. fatty rats. ||P < 0.001 vs. control rats.
DIABETES, VOL 41, APRIL 1992
477
NEW DRUG FOR ACTIVATING INSULIN RECEPTOR KINASE
FIG. 2. Insulin binding to solubilized insulin receptors from hindlimb skeletal muscles of fatty (9,0) and lean (A,A) rats with (O,A) and without (•.A) pioglitazone treatment (3 or 10 mg day" 1 kg" 1 ) for 10 days. Values are means ± SE (n = 6). The Scatchard plot (B) was calculated from the binding data (A). No significant difference between treated and nontreated groups but there was statistically significant difference between fatty and lean rats in receptor number (P < 0.05).
antiphosphotyrosine serum. Immune complexes were precipitated by addition of a 10% suspension of pansorbin. After incubation for 2 h at 4°C, the immune complexes bound to pansorbin were sedimented by centrifugation at 10,000 x g for 10 min. After washing, the pellet was dissolved in the Laemmli buffer containing 2% sodium dodecyl sulfate (SDS), 10% glycerol, 0.01% bromophenol blue, and 10 mM sodium phosphate with or without 100 mM dithiothreitol, and boiled for 5 min and used for SDS-polyacrylamide gel electrophoresis (SDSPAGE) as previously described (17). Electrophoretic separation of the labeled components in 7.5% or 4-12 linear gradient PAGE was performed as previously described. The specific bands were excised from the gel, and the radioactivity was quantitated by liquid-scintillation counting. Phosphorylation of exogenous substrates by purified insulin receptors. Solubilized insulin receptors (20 |xl) were preincubated with various concentrations of insulin at 24°C for 90 min in a total vol of 60 |jJ containing the same components as the preincubation mixture for the autophosphorylation assay. Glutamine-tyrosine (Glu:Tyr, 4:1) polymer and the divalent cation Mg 2+ were added to the solution. After an additional 10-min incubation, phosphorylation was initiated by adding 50 |iM [7-32P]ATP. The final concentrations of Glu-Tyr, MgCI2, and [7-32P] ATP in the incubation mixture were 2 mg/ml, 20 mM, and 12.5 |xM, respectively. After incubation for 10 min at 24°C, the reaction was terminated by adding 70 mM ATP, and the sample was applied to filter papers (Whatman 3 MM). Papers were washed with 10% trichloroacetic acid three times and dried. Radioactivities were counted with a liquid-scintillation counter (9). In vitro kinase activation study by pioglitazone. To study the direct effect of pioglitazone, we incubated IM-9 cells with the agent and examined autophosphorylation of insulin receptors from the IM-9 cells. The cells were incubated with 10~8 M pioglitazone for 7 days, and insulin receptors were purified by the same method used for purification of insulin receptors (17). The direct incubation of solubilized insulin receptors from IM-9 cells with pioglitazone was also carried out, and autophosphorylation of the insulin receptor was determined. Statistical analysis. Data are means ± SD or SE as indicated. P values were determined by Student's t test, and P
A new oral agent, 5-[4-(2-(5- ethyl 12-pyridyl)ethoxy]benzoyl]-2,4-thiazolidinedione (pioglitazone), has been developed for treatment of non-insulin-dependent diabetes mellitus (NIDDM). This agent increases insulin sensitivity in vivo in genetically obese Wistar fatty rats. Administration of the agent (3 mg/kg/day) for 10 days to the rats ameliorated hyperglycemia and hyperinsulinemia, indicating that it decreased insulin resistance. To clarify the mechanism of the drug to increase insulin sensitivity, we examined insulin binding and kinase activity of insulin receptors from muscles of both untreated and treated rats. Pioglitazone treatment did not change insulin binding in Wistar fatty rats but increased insulin-stimulated autophosphorylation of insulin receptors to 78% over the level in the control but not the basal state. Kinase activity toward exogenous substrate, poly Glu 4 Tyr\ was also increased to 87% over the level of untreated control obese rats. In contrast, in lean rats, pioglitazone treatment did not increase autophosphorylation and kinase activity toward exogenous substrates. To further elucidate the mechanism, we incubated insulin receptors with the agent and measured kinase activity. Incubation of solubilized receptors with the agent did not increase kinase activity. However, the receptors from IM-9 cells, which were incubated with 10~ 8 M pioglitazone for 7 days, showed a 46% increase over the control in insulin-stimulated autophosphorylation and kinase activity. These results suggested that pioglitazone increased insulin sensitivity in part by activating kinase of the receptors through indirect effect on insulin receptors and that the drug may have useful benefits in insulin resistance of NIDDM. Diabetes 41:476-83, 1992
From the Third Department of Medicine, Shiga University of Medical Science, Ohtsu, Japan. Address correspondence and reprint requests to Masashi Kobayashi, MD, First Department of Medicine, Toyama Medical and Pharmaceutical University, Sugitani, Toyama 930-01, Japan. Received for publication 7 May 1991 and accepted in revised form 18 December 1991.
476
T
he insulin receptor consists of two a-subunits that bind insulin and two (3-subunits that have tyrosine-specific kinase activity and respond to insulin by autophosphorylation and the phosphorylation of exogenous substrates (1-3). The importance of insulin receptor function for glucose metabolism is highlighted by the insulin receptor disease characterized by markedly decreased insulin binding or kinase defect due to genetic abnormalities (4-6). Thus, the insulin receptor function should be evaluated in various insulin resistance states. Non-insulin-dependent diabetes mellitus (NIDDM) is characterized by insulin resistance along with decreased insulin secretion and is often associated with obesity. Previous studies on insulin receptor function of NIDDM revealed decreased kinase activity of p-subunit, which may be in part responsible for decreased insulin action in NIDDM (7-9). Therefore, the best way to treat patients with NIDDM is to increase insulin sensitivity and insulin secretion. For this reason, sulfonylurea has been used, and the mechanism of increased insulin sensitivity by sulfonylurea has been investigated. The initial step of insulin action, i.e., insulin binding and kinase of insulin receptors, is not directly influenced by the agent, indicating that its action appears to be at the postkinase step (10,11). Pioglitazone, a thiazolidine derivative that has potent activity to increase insulin sensitivity, has been developed for treatment of NIDDM. Previous studies showed that diabetic obese rats could be treated successfully with this oral agent and that insulin resistance of these rats was ameliorated (12,13). To clarify the mechanism of the agent, we examined the function of insulin receptors by measuring the insulin binding and receptor kinase activity of insulin receptors from the pioglitazone-treated rats. We report that pioglitazone is a potent agent that increases insulin sensitivity by activating kinase of insulin receptors, which sulfonylurea does not exhibit, and that it
DIABETES, VOL. 41, APRIL 1992
KOBAYASHI AND ASSOCIATES
C2H5
•o
CH 2 CH 2 O FIG. 1. Structure of pioglitazone.
could be a potentially effective oral agent with a new mode of action for NIDDM. RESEARCH DESIGN AND METHODS Wistar fatty rats {fa/fa) were bred by mating heterozygous lean rats (fa/+) as reported previously (14). Eleven-weekold male rats with elevated plasma glucose, triglyceride, and insulin levels were used to examine the effects of pioglitazone on insulin receptor binding and insulin receptor kinase activity in muscles (Fig. 1). Lean rats {fa/+) were also used to study the effects of the agent on insulin receptor function for comparison. Pioglitazone was administered by a stomach tube after the compound was suspended in 5% gum arabic solution. Three and 10 mg/kg of the agent was given to fatty and lean rats for 10 days, respectively. Previous studies revealed increased activity of the agent toward obese rats and relative ineffectiveness toward lean rats (12,13). In lean rats, 3- to 10-mg/kg dosage did not significantly affect insulin sensitivity or produce toxic effects. Therefore, considering the lean mass of these rats and the more clear-cut results of the effect of the agent, decreased amount of pioglitazone (i.e., 3 mg/kg), was given to the fatty rats, which was enough to ameliorate insulin resistance. The rats were killed by decapitation and blood samples were collected for measurement of blood glucose, insulin, and triglyceride levels. Purified pork insulin was a gift from Shimizu (Shizuoka, Japan). [ 125 I] Na and 1000-3000 Ci/mmol [ 7 - 32 P]ATP were purchased from Du Pont-NEN (Boston, Massachussetts). Fatty acid-free bovine serum albumin and 8 0 0 0 - H polyethylene glycol were from Sigma (St. Louis, MO). Disuccinimidyl suberate was from Pierce (Rockford, IL). Protein A (pansorbin) was from Calbiochem (La Jolla,
CA). Wheat-germ agglutinin (WGA) agarose was from Pharmacia (Uppsala, Sweden). Anti-insulin antibody was obtained from a type B insulin-resistant patient with Sjogren's syndrome. Anti-phosphotyrosine antibody was kindly supplied by Dr. H. Fujio (Research Institute for Microbial Diseases, Osaka University, Osaka, Japan). Preparation of partially purified insulin receptor Hindlimb muscles were trimmed of fat and immediately frozen in liquid N 2 . The frozen muscle was finely ground in a motor precooled to - 2 0 ° C and an ice-cold buffer consisting of 25 mM HEPES, 1 % Triton X-100, 4 mM EDTA, 1 typsin inhibitory unit (TIU)/ml aprotinin, and 2 mM phenylmethylsulfonyl fluoride (pH 7.4) as described by Burant et al. (15). The frozen slurry was homogenized as it thawed in a Potter-Elvejhem homogenizer and then centrifuged at 10,000 x g for 10 min at 4°C. The resulting supernate was slowly stirred at 21 °C for 60 min and then centrifuged at 150,000 x g for 90 min at 4°C. The supernate was applied to WGA agarose column. The column was washed with 25 mM HEPES and 0 . 1 % Triton X-100, pH 7.4. Receptors were eluted with 0.3 M A/-acetyl glucosamine in the buffer for washing the column. Insulin binding to solubilized receptors. Lectin- purified insulin receptors (30 |xl) were incubated with 1 2 5 I labeled insulin at 4°C for 16 h in the presence of various concentrations of unlabeled insulin in 200 uJ of 25 mM HEPES (pH 7.6) containing 0.05% Triton X-100, 15 mM NaCI, 0.1 mg/ml bovine serum albumin, and 150 mM N-acetyl-D-glucosamine. With human 7-globulin as carrier protein, receptor-bound insulin was precipitated with polyethylene glycol as previously described (16). Nonspecific binding was defined as the radioactivity precipitated in the presence of 3.1 \xM unlabeled insulin.
TABLE 1 Characteristics of rats Concentration in plasma
Lean Control Treated Fatty Control Treated
Body weight (g)
Glucose (mM)
Triglyceride (mM)
Nonesterified fatty acid (M-Eq/L)
Insulin (M.U/ml)
Liver weight (g)
Muscle weight (g)
235 ± 10 245 ± 10*
7.6 ± 0.3 7.3 ±0.1
5.4 ± 0.8 4.6 ± 0.4+
84 ± 2 1 59 ± 36*
92 ± 6 66 ± 15*
11.6 ±0.8 12.1 ±0.8*
11.2 ±0.9 11.2 ±0.7*
284 ± 9§ 290 ± 9*
9.6 ± 1.1§ 6.6 ± 0.2||
17.6±4.1§ 7.1 ± 0.9||
136±21§ 93 ± 19*
368 ± 45§ 220 ± 23||
15.0±0.7§ 14.3 ±0.7*
8.3 ± 1.5* 9.3 ±0.9*
Values are means ± SD determined 20 h after administration of last dosage of pioglitazone for 6 rats/group. Pioglitazone was administered orally to 6-wk-old male Wistar lean and fatty rats for 10 days. Ten and 3 mg • kg" 1 • day" 1 pioglitazone was administered to lean and fatty rats, respectively. *Not significant, lean vs. fatty and treated vs. control rats. t P < 0.05, $P < 0.01, vs. control rats. §P < 0.001, lean vs. fatty rats. ||P < 0.001 vs. control rats.
DIABETES, VOL 41, APRIL 1992
477
NEW DRUG FOR ACTIVATING INSULIN RECEPTOR KINASE
FIG. 2. Insulin binding to solubilized insulin receptors from hindlimb skeletal muscles of fatty (9,0) and lean (A,A) rats with (O,A) and without (•.A) pioglitazone treatment (3 or 10 mg day" 1 kg" 1 ) for 10 days. Values are means ± SE (n = 6). The Scatchard plot (B) was calculated from the binding data (A). No significant difference between treated and nontreated groups but there was statistically significant difference between fatty and lean rats in receptor number (P < 0.05).
antiphosphotyrosine serum. Immune complexes were precipitated by addition of a 10% suspension of pansorbin. After incubation for 2 h at 4°C, the immune complexes bound to pansorbin were sedimented by centrifugation at 10,000 x g for 10 min. After washing, the pellet was dissolved in the Laemmli buffer containing 2% sodium dodecyl sulfate (SDS), 10% glycerol, 0.01% bromophenol blue, and 10 mM sodium phosphate with or without 100 mM dithiothreitol, and boiled for 5 min and used for SDS-polyacrylamide gel electrophoresis (SDSPAGE) as previously described (17). Electrophoretic separation of the labeled components in 7.5% or 4-12 linear gradient PAGE was performed as previously described. The specific bands were excised from the gel, and the radioactivity was quantitated by liquid-scintillation counting. Phosphorylation of exogenous substrates by purified insulin receptors. Solubilized insulin receptors (20 |xl) were preincubated with various concentrations of insulin at 24°C for 90 min in a total vol of 60 |jJ containing the same components as the preincubation mixture for the autophosphorylation assay. Glutamine-tyrosine (Glu:Tyr, 4:1) polymer and the divalent cation Mg 2+ were added to the solution. After an additional 10-min incubation, phosphorylation was initiated by adding 50 |iM [7-32P]ATP. The final concentrations of Glu-Tyr, MgCI2, and [7-32P] ATP in the incubation mixture were 2 mg/ml, 20 mM, and 12.5 |xM, respectively. After incubation for 10 min at 24°C, the reaction was terminated by adding 70 mM ATP, and the sample was applied to filter papers (Whatman 3 MM). Papers were washed with 10% trichloroacetic acid three times and dried. Radioactivities were counted with a liquid-scintillation counter (9). In vitro kinase activation study by pioglitazone. To study the direct effect of pioglitazone, we incubated IM-9 cells with the agent and examined autophosphorylation of insulin receptors from the IM-9 cells. The cells were incubated with 10~8 M pioglitazone for 7 days, and insulin receptors were purified by the same method used for purification of insulin receptors (17). The direct incubation of solubilized insulin receptors from IM-9 cells with pioglitazone was also carried out, and autophosphorylation of the insulin receptor was determined. Statistical analysis. Data are means ± SD or SE as indicated. P values were determined by Student's t test, and P
