Pioglitazone Ameliorates Insulin Resistance and Diabetes by Both Adiponectin-dependent and -independent Pathways*

*Thiazolidinediones have been shown to up-regulate adiponectin expression in white adipose tissue and plasma adiponectin levels, and these up-regulations have been proposed to be a major mechanism of the thiazolidinedione-induced amelioration of insulin resistance linked to obesity. To test this hypothesis, we generated adiponectin knock-out (adipo-/-) ob/ob mice with a C57B/6 background. After 14 days of 10 mg/kg pioglitazone, the insulin resistance and diabetes of ob/ob mice were significantly improved in association with significant up-regulation of serum adiponectin levels. Amelioration of insulin resistance in ob/ob mice was attributed to decreased glucose production and increased AMP-activated protein kinase in the liver but not to increased glucose uptake in skeletal muscle. In contrast, insulin resistance and diabetes were not improved in adipo-/-ob/ob mice. After 14 days of 30 mg/kg pioglitazone, insulin resistance and diabetes of ob/ob mice were again significantly ameliorated, which was attributed not only to decreased glucose production in the liver but also to increased glucose uptake in skeletal muscle. Interestingly, adipo-/-ob/ob mice also displayed significant amelioration of insulin resistance and diabetes, which was attributed to increased glucose uptake in skeletal muscle but not to decreased glucose production in the liver. The serum-free fatty acid and triglyceride levels as well as adipocyte sizes in ob/ob and adipo-/-ob/ob mice were unchanged after 10 mg/kg pioglitazone but were significantly reduced to a similar degree after 30 mg/kg pioglitazone. Moreover, the expressions of TNFα and resistin in adipose tissues of ob/ob and adipo-/-ob/ob mice were unchanged after 10 mg/kg pioglitazone but were decreased after 30 mg/kg pioglitazone. Thus, pioglitazone-induced amelioration of insulin resistance and diabetes may occur adiponectin dependently in the liver and adiponectin independently in skeletal muscle.

Thiazolidinediones (TZDs) 2 have been shown to act as insulin sensitizers in animal models of obesity-linked insulin resistance and diabetes, and they have been widely used as therapeutic agents for the treatment of type 2 diabetes (1)(2)(3)(4)(5). TZDs have been proposed to ameliorate insulin resistance by binding to and activating peroxisome proliferator-activated receptor ␥ (PPAR␥) in adipose tissue, thereby promoting adipose differentiation and increasing the number of small adipocytes that are more sensitive to insulin (6 -9). Generation of small insulin-sensitive adipocytes by TZDs appears to be associated with amelioration of insulin resistance (10,11). TZDs also lower circulating serum triglyceride and free fatty acid levels and down-regulate the production and secretion of TNF␣ and resistin (6, 10 -15).
Adiponectin is an adipose tissue-derived secreted protein that circulates in plasma (16 -19). We previously reported (20) finding that replenishment of adiponectin ameliorated insulin resistance in obese mice with decreased plasma adiponectin levels and that a combination of physiological doses of adiponectin and leptin reversed insulin resistance in lipoatrophic mice. Independently, administration of adiponectin has been reported to decrease plasma glucose levels by suppressing hepatic glucose production (21,22), and administration of globular adiponectin reportedly lowers elevated fatty acid concentrations by oxidizing fatty acids in muscle (23). We and others (24,25) have also demonstrated that adiponectin-deficient (adipo Ϫ/Ϫ ) mice are insulin-resistant and glucose-intolerant. Previous studies have shown that adiponectin stimulates fatty acid oxidation in skeletal muscle and inhibits glucose production in the liver by activating AMP-activated protein kinase (AMPK) (26) through its specific receptors, AdipoR1 and AdipoR2 (27). As a result, adiponectin has come to be recognized as a major insulinsensitizing hormone (28,29).
TZDs increase plasma adiponectin levels in animal models of obesity and diabetes, nondiabetic subjects, and patients with type 2 diabetes, and the improvement in insulin sensitivity in response to TZD administration is associated with an increase in circulating adiponectin (20, 30 -35). Thus, it is reasonable to speculate that the action whereby TZDs increase insulin sensitivity is mediated, at least in part, by increased adiponectin. However, whether the TZD-induced increase in plasma adiponectin is causally involved in TZD-mediated insulin-sensitizing effects has not been addressed experimentally.
To address this issue, in the present study, we used adipo Ϫ/Ϫ ob/ob mice with a C57Bl/6 background to investigate whether the PPAR␥ agonist pioglitazone is capable of ameliorating insulin resistance in the absence of adiponectin. The absence of adiponectin had no effect on either the obesity or the diabetic phenotype of these mice. We found that the insulin resistance and diabetes of ob/ob mice was significantly improved in association with significant up-regulation of serum adiponectin levels after 14 days of 10 mg/kg pioglitazone treatment. Amelioration of insulin resistance in ob/ob mice was attributed to decreased glucose production and increased AMPK in the liver but not to increased glucose uptake in skeletal muscle. In contrast, insulin resistance and diabetes were not improved in adipo Ϫ/Ϫ ob/ob mice. After 14 days of 30 mg/kg pioglitazone treatment, insulin resistance and diabetes of ob/ob mice were again significantly ameliorated, which was attributed not only to decreased glucose production in the liver but also to increased glucose uptake in skeletal muscle. Interestingly, adipo Ϫ/Ϫ ob/ob mice also displayed significant amelioration of insulin resistance and diabetes, which was attributed to increased glucose uptake in skeletal muscle but not to decreased glucose production in the liver. Thus, pioglitazone-induced amelioration of insulin resistance and diabetes is mediated via both adiponectin-dependent pathway in the liver and adiponectin-independent pathway in skeletal muscle.

Animals and Genotyping
Mice were housed on a 12-h light-dark cycle and fed standard chow CE-2 (CLEA Japan Inc., Tokyo, Japan) with the following composition: 25.6% (w/w) protein, 3.8% fiber, 6.9% ash, 50.5% carbohydrates, 4% fat, and 9.2% water. To rule out the potential impact of the expression cassettes for the selection of targeted ES cells in the targeted allele on the expression of genes surrounding the adiponectin locus, selection cassettes were deleted by the Cre-Pac method as described previously (36), with some modification. We then backcrossed the original adipo Ϫ/Ϫ mice (C57Bl/6 and 129/sv background) (24) with C57Bl/6 mice more than seven times. ob/ob and adipo Ϫ/Ϫ ob/ob mice were prepared by adipo ϩ/Ϫ ob/ϩ mouse intercrosses. All experiments in this study were conducted on male littermates. The animal care and procedures of the experiments were approved by the Animal Care Committee of the University of Tokyo.

Pioglitazone Treatment Study
10 mg/kg pioglitazone (AD-4833-HCl) or vehicle (0.25% carboxymethylcellulose) was adnimistered to ob/ob and adipo Ϫ/Ϫ ob/ob mice by oral gavage once daily for 14 consecutive days. 30 mg/kg pioglitazone or vehicle was also adnimistered to ob/ob and adipo Ϫ/Ϫ ob/ob mice by oral gavage once daily for 14 consecutive days. Pioglitazone was kindly provided by Takeda Chemical Industries Co., Ltd. (Osaka, Japan).

Hyperinsulinemic-Euglycemic Clamp Study
Clamp studies were carried out as described previously (37) with slight modifications. In brief, 2-3 days before the study, an infusion catheter was inserted into the right jugular vein under general anesthesia with sodium pentobarbital. Studies were performed on mice under conscious and unstressed conditions after a 6-h fast. A primed continuous infusion of insulin (Humulin R, Lilly) was given (5.0 milliunits/kg/ min), and the blood glucose concentration, monitored every 5 min, was maintained at ϳ120 mg/dl by administration of glucose (5 g of glucose per 10 ml enriched to ϳ20% with [6,6-2 H 2 ]glucose (Sigma)) for 120 min. Blood was sampled via tail tip bleeds at 90, 105, and 120 min for determination of the rate of glucose disappearance (R d ). R d was calculated according to nonsteady-state equations (37), and endogenous glucose production (EGP) was calculated as the difference between R d and exogenous glucose infusion rates (GIR) (37).

Serum Adiponectin and Lipid Measurements
Mice were fasted for more than 16 h before the measurements. Serum adiponectin levels were determined with a mouse adiponectin enzymelinked immunosorbent assay kit (Otsuka Pharmaceutical Co., Ltd., Tokyo, Japan). Serum triglyceride and free fatty acids (Wako Pure Chemical Industries Ltd., Osaka, Japan) were assayed by enzymatic methods.

In Vivo Glucose Homeostasis
Glucose Tolerance Test-Mice were fasted for more than 16 h before the study and then orally loaded with glucose, 1.5 mg/g body weight. Blood samples were collected from the orbital sinus at different times, and glucose was measured with an automatic glucometer (Glutest Ace, Sanwa Chemical Co., Nagoya, Japan). Whole blood was collected and centrifuged in heparinized tubes, and the plasma was stored at Ϫ20°C. Insulin levels were determined with an insulin radioimmunoassay kit (Biotrak, Amersham Biosciences, Buckinghamshire, UK) with rat insulin as the standard (38).
Insulin Tolerance Test-Mice were given free access to food and then fasted during the study. They were intraperitoneally challenged with human insulin, 0.75 milliunits/g of body weight (Humulin R), and venous blood samples were drawn at different times (38). The changes were plotted as a percentage of basal glucose versus time.

Measurement of Adipocyte Size
Epididymal white adipose tissue was routinely processed for paraffin embedding, and 2-m sections were cut and mounted on silanized slides. The adipose tissue was stained with hematoxylin and eosin, and total adipocyte area was manually traced and analyzed with Win ROOF software (Mitani Co. Ltd., Chiba, Japan). White adipocyte area was measured in 200 or more cells per mouse in each group according to methods described previously (39), with slight modifications.

RNA Preparation and Taqman PCR
Total RNA was prepared from adipose tissue with an RNeasy Mini Kit (Qiagen Co., Düsseldorf, Germany) according to the manufacturer's instructions. mRNA levels in white adipose tissue were quantitatively analyzed by fluorescence-based reverse transcriptase-PCR. The reverse transcription mixture was amplified with specific primers, using an ABI Prism 7000 sequence detector equipped with a thermocycler. The primers used for ␤-actin were as described previously (40). The primers used for phosphoenolpyruvate carboxykinase (PEPCK), TNF␣, and resistin were purchased from Applied Biosystems (Foster City, CA). Relative expression levels were compared after normalization to ␤-actin.

Absence of Adiponectin Had No Effects on Obesity, Fasting Hyperglycemia, or Fasting Hyperinsulinemia-
The ob/ob mice and adipo Ϫ/Ϫ ob/ob mice on a standard diet gained total body weight at comparable rates (Fig. 1A). Moreover, the ob/ob and adipo Ϫ/Ϫ ob/ob mice showed comparable fasting hyperglycemia (Fig. 1B) and comparable fasting hyperinsulinemia (Fig. 1C). These data indicate that the absence of adiponectin had no effect on either the obesity or the diabetic phenotype of these mice.
10 mg/kg Pioglitazone for 14 Days Improved Diabetes in ob/ob Mice but Not in adipo Ϫ/Ϫ ob/ob Mice-Ob/ob mice showed diabetic glucose tolerance ( Fig. 2A). 10 mg/kg pioglitazone for 14 days significantly increased serum adiponectin levels in the ob/ob mice ( Fig. 2A, inset). After 14 days of 10 mg/kg pioglitazone treatment, an oral glucose tolerance test (OGTT) showed that the blood glucose level of pioglitazonetreated ob/ob mice 15 min after glucose loading was significantly lower than that of untreated ob/ob mice ( Fig. 2A). Adipo Ϫ/Ϫ ob/ob mice showed comparable diabetic glucose tolerance to ob/ob mice (Fig. 2B). Serum adiponectin levels were not detectable in adipo Ϫ/Ϫ ob/ob mice before and after 14 days of 10 mg/kg pioglitazone treatment (Fig. 2B, inset). Unlike ob/ob mice, the blood glucose levels before and after glucose loading were indistinguishable between untreated and treated adipo Ϫ/Ϫ ob/ob mice (Fig. 2B). We calculated the area under the curve (AUC) during the OGTT to quantitate glucose intolerance. Before pioglitazone treatment, the AUCs were indistinguishable between ob/ob and adipo Ϫ/Ϫ ob/ob mice (Fig. 2C). The AUCs became significantly  smaller after pioglitazone treatment in the ob/ob/mice but not in the adipo Ϫ/Ϫ ob/ob mice (Fig. 2C). Both ob/ob and adipo Ϫ/Ϫ ob/ob mice showed hyperinsulinemia before and after glucose loading; however, the plasma insulin levels 15 min after glucose loading tended to be reduced in the adipo Ϫ/Ϫ ob/ob mice as compared with ob/ob mice (p ϭ 0.11), as we reported previously in the adipo Ϫ/Ϫ mice as compared with wildtype mice (24) (Fig. 2D). The plasma insulin levels of pioglitazonetreated ob/ob mice before and after glucose loading were significantly lower than those of untreated ob/ob mice (Fig. 2D). In contrast, the plasma insulin levels before and after glucose loading were indistinguishable between untreated and treated adipo Ϫ/Ϫ ob/ob mice (Fig. 2D). These findings indicate that pioglitazone ameliorates diabetes in mice with an ob/ob background in an adiponectin-dependent manner.
10 mg/kg Pioglitazone for 14 Days Improved Insulin Resistance in the Liver of ob/ob Mice but Not of adipo Ϫ/Ϫ ob/ob Mice-We next carried out hyperinsulinemic-euglycemic clamp studies in ob/ob and adipo Ϫ/Ϫ ob/ob mice to investigate the effect of pioglitazone on amelioration of insulin resistance in the liver and skeletal muscle. Before pioglitazone treatment, GIR were comparable in ob/ob and adipo Ϫ/Ϫ ob/ob mice (Fig.  3A). After 14 days of 10 mg/kg pioglitazone treatment, the GIR of pioglitazone-treated ob/ob mice was significantly higher than that of untreated ob/ob mice, indicating insulin resistance in ob/ob mice to be improved (Fig. 3A). In contrast, the GIR were indistinguishable between untreated and pioglitazone-treated adipo Ϫ/Ϫ ob/ob mice (Fig. 3A). The amelioration of insulin resistance in ob/ob mice was, at least in part, due to decreased EGP (Fig. 3B). Rates of R d were indistinguishable between ob/ob and adipo Ϫ/Ϫ ob/ob mice, and 10 mg/kg pioglitazone for 14 days had no effect on these levels in either genotype (Fig. 3C). PEPCK expression levels in the liver were comparable in ob/ob and adipo Ϫ/Ϫ ob/ob mice before pioglitazone treatment (Fig. 3D). 10 mg/kg pioglitazone for 14 days significantly decreased PEPCK expression in ob/ob, but not adipo Ϫ/Ϫ ob/ob, mice (Fig. 3D). AMPK expression levels in the liver did not differ significantly between ob/ob and adipo Ϫ/Ϫ ob/ob mice before or after pioglitazone treatment (Fig. 3E). AMPK activities before pioglitazone treatment were comparable in ob/ob and adipo Ϫ/Ϫ ob/ob mice (Fig. 3E). AMPK phosphorylation in ob/ob mice was significantly increased after 10 mg/kg pioglitazone for 14 days but was unchanged in adipo Ϫ/Ϫ ob/ob mice (Fig. 3E). These findings indicate that pioglitazone ameliorates hepatic, but not muscle, insulin resistance in mice with an ob/ob background in an adiponectin-dependent manner via, at least in part, decreased gluconeogenesis and increased AMPK activation.
30 mg/kg Pioglitazone for 14 Days Improved Diabetes to a Similar Degree in ob/ob and adipo Ϫ/Ϫ ob/ob Mice-We next administered 30 mg/kg pioglitazone to ob/ob and adipo Ϫ/Ϫ ob/ob mice for 14 days.
30 mg/kg pioglitazone also significantly increased serum adiponectin levels in the ob/ob mice (Fig. 4A, inset), but the serum adiponectin levels after 30 mg/ml pioglitazone were not significantly different from those after 10 mg/kg pioglitazone ( Fig. 2A, inset, and Fig. 4A, inset). The blood glucose levels of pioglitazone-treated ob/ob mice before and after glucose loading were significantly lower than those of untreated ob/ob mice (Fig. 4A). Interestingly, the blood glucose levels of pioglitazonetreated adipo Ϫ/Ϫ ob/ob mice before and after glucose loading became significantly lower than those of untreated adipo Ϫ/Ϫ ob/ob mice, being similar to the levels seen in ob/ob/mice (Fig. 4B). The AUCs during the OGTT of both groups became smaller after pioglitazone treatment, but they were indistinguishable between the ob/ob and adipo Ϫ/Ϫ ob/ob mice (Fig. 4C). The plasma insulin levels of pioglitazone-treated ob/ob mice before and after glucose loading were significantly lower than those of untreated ob/ob mice (Fig. 4D). Similarly, the plasma insulin levels before glucose loading of pioglitazone-treated adipo Ϫ/Ϫ ob/ob mice also became significantly lower than those of untreated adipo Ϫ/Ϫ ob/ob mice (Fig. 4E).

mg/kg Pioglitazone for 14 Days Improved Insulin Resistance in the Liver and Skeletal Muscle of ob/ob Mice but Only in Skeletal
Muscle of adipo Ϫ/Ϫ ob/ob Mice-We next carried out hyperinsulinemic-euglycemic clamp studies in ob/ob and adipo Ϫ/Ϫ ob/ob mice to investigate the effect of 30 mg/kg pioglitazone treatment on amelioration of insulin resistance in the liver and skeletal muscle. After 14 days of 30 mg/kg pioglitazone treatment, the GIR of pioglitazone-treated ob/ob mice was significantly higher than that of untreated ob/ob mice (Fig. 5A). Interestingly, the GIR of pioglitazone-treated adipo Ϫ/Ϫ ob/ob mice was also significantly higher than that of untreated adipo Ϫ/Ϫ ob/ob mice, indicating insulin resistance in adipo Ϫ/Ϫ ob/ob mice to be improved (Fig.  5A). The EGP was significantly decreased in ob/ob mice after 30 mg/kg pioglitazone treatment but not in adipo Ϫ/Ϫ ob/ob mice (Fig. 5B). In contrast, the Rd was significantly increased in ob/ob and adipo Ϫ/Ϫ ob/ob mice to a similar degree after 30 mg/kg pioglitazone treatment (Fig. 5C). 30 mg/kg pioglitazone significantly decreased PEPCK expression in ob/ob mice but not in adipo Ϫ/Ϫ ob/ob mice (Fig. 5D). AMPK phosphorylation in ob/ob mice was significantly increased after 30 mg/kg pioglitazone for 14 days but was unchanged in adipo Ϫ/Ϫ ob/ob mice (Fig. 5E). These findings suggest that the amelioration of insulin resistance in adipo Ϫ/Ϫ ob/ob mice was, at least in part, due to increased glucose uptake in skeletal muscle.  adipocytes and decreased the number of large adipocytes, thereby ameliorating insulin resistance (6). To determine whether the presence of adiponectin is required for the reduction of average adipocyte size induced by TZDs to occur, we histologically analyzed epididymal fat pads after fixation and quantitation of adipocyte size. The adipocyte sizes of ob/ob and adipo Ϫ/Ϫ ob/ob mice were indistinguishable and were not changed by 10 mg/kg pioglitazone for 14 days (Fig. 6, A and B). 30 mg/kg pioglitazone for 14 days, however, significantly reduced adipocyte sizes of ob/ob and adipo Ϫ/Ϫ ob/ob mice to a similar degree (Fig. 6, C and D). These results suggest that pioglitazone can induce a reduction in adipocyte size in the absence of adiponectin or leptin or the absence of both.

mg/kg Pioglitazone, but Not 10 mg/kg, Significantly Decreased Serum Triglyceride and Free Fatty Acid Levels in ob/ob and adipo Ϫ/Ϫ ob/ob
Mice-In addition to improving insulin resistance, TZDs reportedly reduce serum triglyceride (TG) and free fatty acid (FFA) levels (6, 10 -12). However, since the possible involvement of adiponectin in this action of TZDs remains unclear, we investigated the effects of pioglitazone treatment on serum lipid levels. The serum TG levels of the ob/ob and adipo Ϫ/Ϫ ob/ob mice were essentially the same (Fig. 7A), and 10 mg/kg pioglitazone for 14 days did not change the serum TG levels in either genotype (Fig. 7A). Serum FFA levels were also indistinguishable between ob/ob and adipo Ϫ/Ϫ ob/ob mice (Fig. 7B), and 10 mg/kg pioglitazone for 14 days again had no effect on these levels in either group of mice (Fig. 7B). However, 30 mg/kg pioglitazone for 14 days significantly decreased serum TG levels, to a similar degree, in ob/ob and adipo Ϫ/Ϫ ob/ob mice (Fig.  7C). 30 mg/kg pioglitazone for 14 days also lowered FFA levels in both genotypes, and the serum FFA levels of ob/ob and adipo Ϫ/Ϫ ob/ob mice became similar after pioglitazone treatment (Fig. 7D).
30 mg/kg Pioglitazone, but Not 10 mg/kg, Reduced TNF␣ and Resistin Expressions in ob/ob and adipo Ϫ/Ϫ ob/ob Mice-TNF␣ and resistin have been shown to be important mediators of insulin resistance linked to obesity (13)(14)(15). TZDs reportedly reduce the expressions of TNF␣ and resistin (13)(14)(15), but whether adiponectin was involved in this action remains unclear. TNF␣ expression tended to be higher in the adipo Ϫ/Ϫ ob/ob mice than in the ob/ob mice (Fig. 8A). After 14 days of 10 mg/kg pioglitazone treatment, TNF␣ expression was not significantly changed in either ob/ob or adipo Ϫ/Ϫ ob/ob mice (Fig. 8A). Resistin expressions were indistinguishable between ob/ob and adipo Ϫ/Ϫ ob/ob mice before and after 14 days of 10 mg/kg pioglitazone (Fig.  8B). After 14 days of 30 mg/kg pioglitazone, however, TNF␣ expressions were significantly decreased in both ob/ob and adipo Ϫ/Ϫ ob/ob mice (Fig. 8C), and resistin expressions tended to be lower in both pioglitazone-treated groups than in the untreated groups (Fig. 8D).

DISCUSSION
TZDs have been reported to alleviate insulin resistance in adipose tissue, skeletal muscle, and the liver (5,(7)(8)(9)(10)(11). However, since PPAR␥, which is bound and activated by TZDs, is predominantly expressed in adipose tissue, it is reasonable to speculate that the effect of TZDs on insulin resistance in skeletal muscle and the liver is mediated largely via the effects of TZDs on adipose tissue including alterations of adipokine expression and secretion by adipocytes (5, 7-11). Adiponectin has been proposed to be a major insulin-sensitizing adipokine (20 -25) and is a plausible candidate for one of the adipokines that may mediate the TZD-induced amelioration of insulin resistance. Therefore, in this study, we used obesity models, ob/ob and adipo Ϫ/Ϫ ob/ob mice, to investigate whether the TZD-induced increase in plasma adiponectin is causally involved in TZD-mediated insulin-sensitizing effects.
Insulin resistance and diabetes improved significantly in ob/ob mice in association with significant up-regulation of serum adiponectin levels with 14 days of 10 mg/kg pioglitazone treatment. Amelioration of insulin resistance in ob/ob mice was attributed to improvement of hepatic, but not muscle, insulin resistance. These improvements by pioglitazone were significantly obliterated in adipo Ϫ/Ϫ ob/ob mice, indicating that adiponectin is causally involved in the 10 mg/kg pioglitazone-mediated amelioration of hepatic insulin resistance and diabetes. In fact, while PEPCK expression levels were significantly decreased and AMPK activity was significantly increased in the livers of ob/ob mice, these changes were not seen in mice without adiponectin. Interestingly, 10 mg/kg pioglitazone for 14 days failed to improve adipocyte hypertrophy, or the elevations of TG and FFA in serum and TNF␣ and resistin in white    adipose tissue, despite elevated serum adiponectin concentrations in ob/ob mice. This suggests that adiponectin-dependent amelioration of hepatic insulin resistance and diabetes occurs independently of adipocyte size, serum TG, and FFA levels or TNF␣ and resistin levels in adipose tissue.
On the other hand, 30 mg/kg pioglitazone for 14 days unexpectedly ameliorated insulin resistance and diabetes both in ob/ob and adipo Ϫ/Ϫ ob/ob mice. Although the hepatic insulin resistance was not improved in adipo Ϫ/Ϫ ob/ob mice as seen after 10 mg/kg pioglitazone treatment, muscle insulin resistance was alleviated in adipo Ϫ/Ϫ ob/ob mice to a similar degree in ob/ob mice after 30 mg/kg pioglitazone for 14 days. This suggests that 30 mg/kg pioglitazone for 14 days can ameliorate muscle insulin resistance and diabetes via mechanisms which do not require the presence of adiponectin.
We previously reported that TZD treatment resulted in smaller adipocytes and a decreased number of large adipocytes, both of which were accompanied by decreases in TNF␣, resistin, and FFA and an increase in adiponectin (6,20). In fact, 30 mg/kg pioglitazone for 14 days, but not 10 mg/kg pioglitazone for 14 days, significantly reduced adipocyte size in ob/ob and adipo Ϫ/Ϫ ob/ob mice to a similar degree. Moreover, insulin resistance-causing adipokines, such as TNF␣, resistin, and FFA, were similarly reduced by 30 mg/kg pioglitazone for 14 days in adipo Ϫ/Ϫ ob/ob mice as well as ob/ob mice. Thus, adiponectin was not absolutely required for 30 mg/kg pioglitazone-induced reductions in TNF␣, resistin, or FFA. The smaller adipocytes, as well as the decreases in TNF␣, resistin, and FFA, may have played a role in the amelioration of muscle insulin resistance and diabetes produced by 30 mg/kg pioglitazone for 14 days.
Recently, Wellen et al. (41) investigated whether the ability of TZDs to block TNF␣ action may be relevant to its ability to improve insulin action and the metabolism of lipids such as serum TG and FFA, using obese ob/ob mice lacking TNF␣ function. TZDs significantly ameliorated blood glucose and lipid levels in mice with and without TNF␣ function. This suggests that TZDs can improve blood glucose and lipid levels in a TNF␣-independent manner. Therefore, resistin, FFA, and/or other cytokine(s), but perhaps not TNF␣, may play roles in TZD-induced amelioration of insulin resistance and diabetes.
Although both low (10 mg/kg) and high (30 mg/kg) doses of pioglitazone ameliorated insulin resistance and diabetes, the underlying mechanisms may be different. The low dose of pioglitazone may be largely dependent on the adiponectin pathway, while the high dose of pioglitazone also improves adiponectin-independent pathways. High dose pioglitazone treatment in this study showed no significant difference in the ability to ameliorate insulin resistance and diabetes between ob/ob and adipo Ϫ/Ϫ ob/ob mice, but the presence of adiponectin might have affected insulin resistance and diabetes in these mouse models, if the duration of 30 mg/kg pioglitazone treatment had been longer. The degree to which the adiponectin-dependent pathway is involved in TZD-induced amelioration of insulin resistance and diabetes merits further study in murine models and eventually in humans.
In this study, we addressed the important question of whether TZDinduced up-regulation of plasma adiponectin levels is causally involved in the insulin sensitizing actions of TZDs and demonstrated that TZDinduced amelioration of insulin resistance and diabetes may occur adiponectin-dependently in the liver and adiponectin-independently in skeletal muscle.