OBJECTIVE— Adiponectin increases insulin sensitivity and contributes to insulin's indirect effects on hepatic glucose production.

RESEARCH DESIGN AND METHODS— To examine adiponectin's contribution to insulin action, we analyzed adiponectin levels and activation of AMP-activated protein kinase (AMPK) in insulin receptor transgenic/knockout mice (L1), a genetic model of resistance to insulin's indirect effects on hepatic glucose production.

RESULTS— In euglycemic, insulin-resistant L1 mice, we detected hyperadiponectinemia with normal levels of adiponectin receptor-1 and -2. Moreover, adiponectin administration is unable to lower glucose levels or induce activation of AMPK, consistent with a state of adiponectin resistance. In a subset of hyperglycemic L1 mice, we observed decreased mRNA expression of AdipoR2 in liver and muscle, as well as decreased peroxisome proliferator–activated receptor (PPAR)α target gene expression in liver, raising the possibility that deterioration of adiponectin/AdipoR2 signaling via PPARα activation contributes to the progression from compensated insulin resistance to diabetes. In contrast, we failed to detect changes in other markers of the systemic or local inflammatory response.

CONCLUSIONS— These data provide evidence for a mechanism of adiponectin resistance and corroborate the notion that adiponectin potentiates hepatic insulin sensitivity.

Type 2 diabetes is characterized by insulin resistance and impaired pancreatic β-cell function (1). Fasting hyperglycemia in diabetes results primarily from the inability of insulin to inhibit endogenous glucose production (2). Insulin regulates hepatic glucose production (HGP) through both direct (hepatic) and indirect (extrahepatic) mechanisms (3). Chronic effects of insulin are primarily mediated by direct mechanisms via hepatic insulin receptor (Insr)/phosphatidylinositol 3-kinase/forkhead box O1 (FoxO1) signaling to suppress the expression of gluconeogenic enzymes (4). Acute effects of insulin on HGP are mediated by both direct and indirect mechanisms (59). Multiple mechanisms have been proposed to account for insulin's indirect effects on HGP, including glucagon (10,11), gluconeogenic substrates released from muscle (12) and fat (13), and hypothalamic signals (9,14). In addition, adipocytokines have been shown to either increase (e.g., resistin) or decrease (e.g., adiponectin) glucose production (15,16).

In previous studies, we have carried out genetic reconstitution experiments with an allelic series of tissue-specific transgenes to disentangle the complex interactions underlying the integrated physiology of insulin action (1720). These studies have emphasized the central role of the liver in insulin action (19,20) and the role of indirect mechanisms in insulin control of HGP (18). Thus, in mice in which insulin signaling is restricted to liver, selected regions of the brain and pancreatic β-cells (referred to as L1) are, surprisingly, resistant to insulin's direct effect on HGP (18). In this study, we asked whether the impairment of insulin's indirect actions in the liver of L1 mice could be explained by altered adipocytokine action. We report that L1 mice display hyperadiponectinemia, associated with the inability to lower plasma glucose levels and blunted hepatic AMP-activated protein kinase (AMPK) response to adiponectin. These findings delineate a condition of “adiponectin resistance,” previously recognized in IGF-1 receptor dominant-negative transgenic mice (21), that may contribute to the impairment of insulin's direct control of HGP, as previously reported (18).

Transgenic mice expressing human INSR cDNA from the transthyretin (Ttr) promoter were intercrossed with Insr+/− mice. The resulting progeny were further intercrossed to generate Insr−/− and Ttr-INSR (L1) mice and Insr+/+ littermates. Animals were maintained on a mixed background derived from 129/Sv, C57BL/6, and FVB. Genotyping was performed as previously described (19) with the following modification. Wild-type and null Insr alleles were detected using primers 5′-AGCTGTGCACTTCCCTGCTCAC-3′, 5′-TTAAGGGCCAGCTCATTCCTCC-3′, and 5′-TCTTTGCCTGTGCTCCACTCTCA-3′. The product of the wild-type allele is 232 bp in length and that of the null allele is 355 bp. All animal procedures have been approved by the Columbia University Institutional Animal Care and Utilization Committee.

Metabolic analyses.

Plasma glucose was measured by the OneTouch Ultra meter (LifeScan, Milpitas, CA). Free fatty acids, triglycerides, and β-hydroxybutyrate were measured by nonesterified fatty acid C test kit (Wako Chemicals, Richmond, VA), free glycerol reagent (Sigma, St. Louis, MO), and β-hydroxybutyrate reagent set (Pointe Scientific, Canton, MI), respectively. Insulin was measured by enzyme-linked immunosorbent assay, and adiponectin and resistin were measured by radioimmunoassay (Linco Research, St. Charles, MO). Cytokines were measured by the mouse cytokine 10-plex Luminex kit (Biosource, Camarillo, CA). Body composition was determined using GE Lunar PIXImus scan.

Adiponectin treatment.

Production and purification of recombinant full-length murine adiponectin were performed as described (22). In the morning, food was withdrawn 2 h before the experiment. Mice were then injected with either saline or adiponectin via the tail vein (2 μg/g body wt). Blood glucose was measured at 0, 1, 2, 3, 4, 6, 8, and 48 h after injection from tail sampling. Mice were refed 8 h after injection. For Western analysis, mice were fasted for 6 h, injected with saline or adiponectin via the tail vein (2 μg/g body wt), and killed 20 min after injection.

RNA isolation and RT-PCR analyses.

We extracted total RNA using RNeasy Mini kit, RNeasy Mini Fibrous Tissue kit, and RNase-Free DNase set (Qiagen, Valencia, CA). RNA was reverse transcribed using oligo-dT and SuperScript II First-Strand Synthesis System (Invitrogen, Carlsbad, CA). Quantitative PCRs were performed in triplicate using a DNA Engine Opticon 2 System (MJ Research, Bio-Rad, Hercules, CA) and DyNAmo HS SYBR green Q-PCR kit (New England Biolabs, Ipswich, MA). The following sequences of the primers were used: 5′-CCCTGAACCCTAAGGCCAACCGTGAAAA-3′ and 5′-TCTCCGGAGTCCATCACAATGCCTGTG-3′ for β-actin, 5′-AAGGGCAAGCGGGCAGCCAGCA-3′ and 5′-ATCTTCTCCATGGCATGGTGGGCTTG-3′ for AdipoR1, 5′-CAACTACCAAGGAGATTTGGAGCCCAGC-3′ and 5′-GCGGGGACATGCCCATAAACCCTTCA-3′ for AdipoR2, 5′-TTGGCCAAGCTATTGCGACA-3′ and 5′-GCAAAGGCATTGGCTGGAAG-3′ for CD36, 5′-GTGCAGCTCAGAGTCTGTCCAA-3′ and 5′-TACTGCTGCGTCTGAAAATCCA-3′ for Acox1, 5′-CCATCATGGGTCCTTCTGGAG-3′ and 5′-GAACAGTGAGGTGAGGCAGCA-3′ for ABCg1, 5′-TGCACTACGGAGTCCTGCAA-3′ and 5′-GGACAACCTCCATGGCTCAG-3′ for Cpt1α, 5′-GGGTACCACTACGGAGTTCACG-3′ and 5′-CAGACAGGCACTTGTGAAAACG-3′ for peroxisome proliferator–activated receptor (PPAR)α, and 5′-AGTCCTTCCCGCTGACCAAAG-3′ and 5′-TCGAAACTGGCACCCTTGAAA-3′ for PPARγ. Relative mRNA levels were calculated using standard curves, with the PCR product for each primer set normalized to β-actin RNA.

Western blotting.

We prepared detergent extract from liver in buffer containing 20 mmol/l Tris (pH 7.6), 150 mmol/l NaCl, 1 mmol/l dithiothreitol, 10 mmol/l EGTA, 1% NP40, 2.5 mmol/l Na4P2O7, 1 mmol/l NaVO3, 1 mmol/l β-glycerophosphate, and protease inhibitor cocktail (Roche, Indianapolis, IN). Protein concentration of extracts was determined by BCA assay (Pierce, Rockford, IL). We resolved equal amounts of protein (100 μg) on SDS-PAGE, transferred them onto nitrocellulose membranes (Schleicher & Schuell, Dassel, Germany), and probed the membranes with antibodies against phospho-AMPKα (Thr172), total AMPKα, and actin (Cell Signaling, Danvers, MA).

Immunohistochemistry.

Adipose and liver samples were fixed for 12–16 h at room temperature in Z-Fix (Anatech, Battle Creek, MI) and embedded in paraffin. Sections (5-μm thick) were mounted on slides and stained to visualize F4/80 as described (23), using a rat monoclonal antibody against mouse F4/80 and rat IgG2a isotype as control (Caltag, Burlingame, CA). For each liver sample, five different high-power fields (×40 objective) were analyzed, and F4/80-positive cells were counted. For each adipose sample, 10 different high-power fields were analyzed.

General characteristics of the experimental animals.

L1 mice lack endogenous Insr but carry an INSR transgene driven by the transthyretin promoter that reactivates INSR expression in liver, pancreatic β-cells, and various brain regions (19). At 3 months of age, ∼65% of the male L1 mice showed normal glucose levels, while 35% were frankly diabetic. All L1 mice showed increased circulating insulin levels compared with wild-type mice under both fasting and fed conditions (19) (Fig. 1A), indicative of systemic insulin resistance. In addition, we have shown that despite restored insulin signaling in liver by transgenic expression of INSR, L1 mice remain resistant to insulin suppression of HGP during hyperinsulinemic-euglycemic clamps (18).

Hyperadiponectinemia and normoresistinemia in L1 mice.

We hypothesized that hepatic insulin resistance in L1 mice is due to impaired adipocytokine signaling. We measured plasma adiponectin levels in 3-month-old L1 mice and wild-type littermates under fed conditions and detected a 140% increase in L1 mice (Table 1 and Fig. 1C). Notably, every L1 mouse examined had higher adiponectin values than the mean wild-type value, and only two wild-type mice had adiponectin values in the low range of the L1 mice. Furthermore, we observed an inverse correlation between adiponectin and glucose values in L1 mice (Fig. 1D), indicating that falling adiponectinemia correlates with hyperglycemia. Resistin levels in 3-month-old L1 and wild-type mice were similar, but by 5 months of age, L1 mice had a 42% increase of resistin levels compared with wild-type littermates (Fig. 1B). This difference appears to result from diverging age-dependent trends, such that resistinemia decreased with age in wild-type (3.31 ng/ml at 3 months vs. 2.47 ng/ml at 5 months, P = 0.06) but not in L1 (3.76 ng/ml at 3 months vs. 3.51 ng/ml at 5 months, P = 0.53) mice. Since L1 mice exhibited hepatic insulin resistance at 3 months of age, when their resistinemia was similar to wild-type mice, it is unlikely that altered resistin levels contribute to insulin resistance in L1 mice. In addition, we previously showed that leptin levels are not significantly different between L1 and wild-type mice (18). Finally, there were no differences in mean body weight, fat mass, and fat-free mass between 3-month-old wild-type and normoglycemic L1 mice. Diabetic L1 mice displayed increased body weight that was largely accounted for by increased fat-free mass (Table 1). Histological analyses of skeletal muscle of diabetic L1 mice did not reveal significant changes in intramuscular lipid accumulation or in slow-twitch/fast-twitch fiber composition (data not shown). Postprandial free fatty acid and triglyceride levels were increased in both groups of L1 mice (Table 1), while fasting levels were similar to wild-type mice (18). Fasting β-hydroxybutyrate levels were increased by 75% in normoglycemic L1 mice compared with wild-type mice (Table 1).

L1 mice are resistant to the rapid glucose-lowering action of adiponectin.

The correlation between hyperadiponectinemia and hyperinsulinemia in L1 mice led us to hypothesize that hyperadiponectinemia is a marker of adiponectin resistance and that impaired adiponectin action contributes to hepatic insulin resistance in L1 mice. To test this hypothesis, we administered a single dose of purified recombinant full-length murine adiponectin by intravenous injections in 3-month-old normoglycemic L1 mice and wild-type littermates. The recombinant protein was produced in mammalian cells as previously described (22) and had normal distribution of trimer, hexamer, and high–molecular weight complexes (21 and data not shown). Consistent with previous reports (21,22), administration of an adiponectin bolus in wild-type mice resulted in a 38% decrease in glucose levels 2 h after injection (Fig. 2A), followed by a return to normal values by 6 h postinjection (Fig. 2A). In contrast, we detected no difference in glucose levels following adiponectin injection in L1 mice compared with saline-injected controls over an 8-h period (Fig. 2B). Adiponectin-treated L1 mice displayed a trend toward hyperglycemia at 2 days after injection (134% of basal glycemia, P = 0.08), while control (saline-injected) animals did not. These data indicate that L1 mice are resistant to the acute action of adiponectin to lower plasma glucose levels.

Reduced AdipoR2 expression correlates with diabetes susceptibility in L1 mice.

The metabolic effects of adiponectin are mediated by two putative receptors, AdipoR1 and AdipoR2. The former is primarily expressed in muscle, while the latter is in liver (24). In addition, T-cadherin has been proposed to serve as an adiponectin receptor (25), although the metabolic effects of this interaction have not been studied. We investigated whether the decrease in adiponectin's ability to lower plasma glucose levels in L1 mice was due to decreased expression of AdipoR1 and/or -R2. We measured levels of mRNAs encoding AdipoR1 and -R2 in liver and skeletal muscle samples from 3-month-old normoglycemic and diabetic L1 mice, as well as wild-type littermates, by quantitative RT-PCR. Hepatic AdipoR1 mRNA levels were similar in L1 and wild-type mice. Hepatic AdipoR2 mRNA levels in normoglycemic L1 mice were also similar to those in wild-type mice, whereas diabetic L1 mice showed a 33 ± 9% reduction in hepatic AdipoR2 levels compared with wild-type mice (Fig. 3A). We observed a similar expression profile in skeletal muscle. AdipoR1 mRNA expression in muscle exhibited no significant difference between wild-type and L1 mice. AdipoR2 expression in muscle was unaffected in normoglycemic L1 mice but was reduced by 33 ± 11% in diabetic L1 mice compared with wild-type mice (Fig. 3B). These data suggest that the systemic “adiponectin resistance” observed in normoglycemic L1 mice (Fig. 2) is independent of mRNA expression levels of AdipoR1 and -R2 in liver or muscle. However, reduction in AdipoR2 mRNA in both liver and muscle is associated with hyperglycemia in L1 mice.

Adiponectin-induced AMPK phosphorylation and PPARα activation in L1 mice.

Adiponectin has been shown to inhibit glucose production via activation of AMPK (2628) and fatty acid oxidation via PPARα (24,29). We used Western blotting with a phospho-Thr172–specific antibody to examine AMPK activation in response to intravenous adiponectin injection in normoglycemic L1 and wild-type littermates. As expected, adiponectin treatment in wild-type mice significantly increased phospho-AMPK levels in liver (Fig. 4), without affecting AMPK expression. Interestingly, basal phospho-AMPK levels were elevated in liver of saline-injected L1 mice compared with wild-type mice, possibly resulting from hyperadiponectinemia. However, acute adiponectin treatment in L1 mice failed to enhance AMPK phosphorylation (Fig. 4). These data indicate that adiponectin resistance in L1 mice correlates with impaired hepatic AMPK activation by adiponectin.

The insulin-sensitizing effects of adiponectin can be partly explained by activation of PPARα (26). We examined PPARα mRNA expression in liver and skeletal muscle of L1 mice. Hepatic PPARα mRNA levels in both normoglycemic and diabetic L1 mice were similar to those in wild-type mice (Fig. 5A), as were those of PPARα target genes acyl-coA oxidase (Acox1) and carnitine palmitoyl-transferase 1α (Cpt1α). In contrast, mRNA expression of both Acox1 and Cpt1α in liver of diabetic L1 mice was reduced compared with wild-type mice (Fig. 5A). Expression of CD36, another PPARα target gene, was not significantly altered in L1 mice (Fig. 5A). This suggests that PPARα expression/activity is not affected in normoglycemic L1 mice, while diabetic L1 mice may exhibit reduced PPARα activity; hence, there is reduced fatty acid oxidation, independent of PPARα mRNA levels. This is consistent with the trend toward increased hepatic triglycerides in diabetic L1 mice (15.4 ± 7.6 mg/g liver) compared with wild-type mice (4.6 ± 0.5 mg/g liver, P = 0.19) but not in normoglycemic L1 mice (4.4 ± 2.0 mg/g liver, n = 4 for all groups). Additionally, mRNA levels of PPARα, Acox1, Cpt1α, and CD36 in muscle of L1 mice and wild-type littermates showed no significant differences (Fig. 5B). Therefore, changes in PPARα expression and activity cannot account for adiponectin resistance in L1 mice.

Systemic inflammatory markers and reduced resident hepatic macrophages.

A chronic, low-grade inflammatory stage, characterized by macrophage accumulation in adipose tissue and changes in circulating proinflammatory cytokines, has been demonstrated in insulin resistance, obesity, and diabetes (15). L1 mice macrophages lack Insr (30) and may therefore contribute to systemic inflammation and predispose to hepatic insulin resistance. To address this possibility, we measured circulating levels of interleukin (IL)-1β, IL-2, IL-6, IL-12, granulocyte macrophage-colony stimulating factor, and tumor necrosis factor-α in plasma and found no significant difference between L1 mice and wild-type littermates (Table 2). In addition, histological surveys of macrophages in perigonadal adipose tissue by immunostaining for F4/80, a marker specific for mature macrophages (23,31), showed no morphometric differences between wild-type, normoglycemic L1, and diabetic L1 mice (data not shown). Quantitative analyses revealed that the percentage of F4/80-positive cells was similar between wild-type and normoglycemic L1 mice (6.8 ± 0.5% vs. 6.1 ± 0.6%, P = 0.35) but was moderately increased in the diabetic L1 group (8.1 ± 0.8%, P = 0.16 and 0.05 for wild-type and normoglycemic L1, respectively). Since no difference was detected between wild-type and normoglycemic L1 mice, these data indicate that systemic inflammation does not contribute to insulin resistance in L1 mice.

We next investigated whether there are alterations of Kupffer cell numbers or distribution in L1 mice, potentially leading to intrahepatic inflammation and cytokine production, thus causing hepatic insulin resistance. F4/80 immunostaining failed to reveal differences in Kupffer cell morphology between wild-type, normoglycemic L1, and diabetic L1 mice (Fig. 6A–C). Indeed, quantitative analyses revealed that both normoglycemic and diabetic L1 mice had fewer Kupffer cells than wild-type mice (80 and 61% of wild type, P < 0.05 and P < 0.01, respectively) (Fig. 6D). We also examined expression of macrophage-specific transcripts in liver samples of L1 and wild-type mice. Hepatic CD36 expression failed to show significant differences between wild-type and normoglycemic or diabetic L1 mice (Fig. 5A). In addition, mRNA expression of ATP-binding cassette transporter G1, a macrophage-specific cholesterol efflux transporter, in liver of normoglycemic L1 mice was not significantly different from that of wild-type mice but was reduced by 60% in diabetic L1 mice (Fig. 6E). The transcriptional activity of PPARγ is required for the activation of both CD36 and ABCg1 in macrophages (32,33). We did not detect differences in PPARγ mRNA levels in liver lysates (Fig. 6F), although it is possible that changes in macrophage PPARγ expression were masked by PPARγ transcripts in hepatocytes. In summary, our data indicate that liver of L1 mice has a reduced Kupffer cell population, which correlates with decreases in macrophage-specific gene expression. Therefore, we found no evidence for local activation of inflammation, which may account for hepatic insulin resistance in L1 mice.

Tissue-specific insulin resistance is associated with hyperadiponectinemia and adiponectin resistance.

The antidiabetic properties of adiponectin have attracted considerable attention. A number of studies in humans, nonhuman primates, and rodents have shown that hypoadiponectinemia is a common feature of obesity and insulin resistance (34). In contrast, L1 mice have elevated adiponectin levels. The association between hyperadiponectinemia and insulin resistance was previously reported in another model of selective Insr ablation, the adipocyte-specific Insr knockout (FIRKO) (35), suggesting that hyperadiponectinemia in L1 mice may be explained by the lack of Insr signaling in adipose cells. This hypothesis is also consistent with a recent study demonstrating that FoxO1, a transcription factor inhibited by insulin signaling, activates adiponectin gene transcription in adipocytes (36). Moreover, Kim et al. (21) have recently shown that mice with muscle-specific insulin resistance due to trans-dominant inhibition of Insr and IGF-1 receptors are also hyperadiponectinemic and adiponectin resistant. When considered with the present data, this burgeoning literature begins to outline a mechanism by which adiponectin production is controlled in response to changes in systemic insulin sensitivity. This mechanism is likely to be conserved in humans, as a recent study of patients with severe insulin resistance found that subjects with INSR mutations have elevated plasma adiponectin levels, whereas moderate insulin resistance correlates in hypoadiponectinemia (37). Since three major forms of adiponectin are present in circulation, potentially with different signaling activities (34), it will be interesting to determine whether all three are elevated in Insr-deficient models.

In this study, we show that acute adiponectin treatment failed to lower glycemia in L1 mice (Fig. 2), consistent with systemic adiponectin resistance. The mechanism for this failure appears to correlate with a blunted response of AMPK activation but not with changes in AdipoR1 and -R2 mRNA expression. The reduced AMPK response is associated with increased basal AMPK phosphorylation. Therefore, although our data are consistent with adiponectin resistance in L1 mice, we cannot rule out the possibility that adiponectin/AMPK signaling is maximally activated in L1 mice under basal conditions and is no longer sensitive to acute increases in circulating adiponectin. Interestingly, mice with muscle-specific expression of dominant-negative IGF-1 receptor (muscle IGF-1 receptor K-[lysine]-R[arginine] mutant [MKR]) also display systemic adiponectin resistance, although adiponectin-induced AMPK activation was normal (21). Therefore, adiponectin resistance may occur via multiple mechanisms, and MKR mice may represent a state of “AMPK resistance.”

Progressive reduction in adiponectin signaling and diabetes.

The metabolic phenotype of L1 mice is inherently heterogeneous. While all L1 mice are markedly hyperinsulinemic and insulin resistant, ∼35% of male L1 mice are frankly diabetic, possibly as a result of differences in their genetic make up (38). In this subset of mice, we identified two differences in the adiponectin pathway: a trend toward lower adiponectin levels compared with normoglycemic L1 mice and a reduction of AdipoR2 mRNA levels in liver and muscle compared with wild-type mice. These data are consistent with the possibility that hyperglycemia acts as an independent variable in the control of adiponectin sensitivity. In this regard, we have previously shown that AdipoR1 and -R2 are negatively regulated by insulin via FoxO1 (39). Since FoxO1 is also regulated by glucose via oxidative stress (40), it is possible that the changes in AdipoR2 levels reflect decreased FoxO1-dependent transcription. In contrast, AdipoR1 mRNA expression was not changed in liver and tended to be reduced in skeletal muscle of both normoglycemic and diabetic L1 mice. These data may reflect differential regulation of AdipoR1 and -R2 under insulin-resistant/diabetic conditions. In addition, the tendency for reduced AdipoR1 expression in muscle of L1 mice may also contribute to adiponectin resistance and insulin resistance.

Inflammation in a nonobese model of insulin resistance.

A mechanistic link between innate immunity and metabolism has emerged over the past few years (15). Obesity is associated with chronic low-grade inflammation, macrophage recruitment to white adipose tissue, as well as increased proinflammatory cytokines (41). However, whether inflammation also plays a role in insulin sensitivity in the lean state is unknown. Moreover, the relative contribution of insulin resistance in adipocytes versus macrophages to activation of the inflammatory state is unclear. L1 mice maintain normal body weight and composition, despite lack of Insr in macrophages and adipocytes. Thus, they represent an excellent nonobese model of insulin resistance/diabetes to address these questions. Our findings indicate that the serum cytokine profile of L1 and wild-type mice was similar, as was the number and morphology of adipose tissue macrophages. Although we observed a 33% increase in macrophage number in the diabetic L1 group, it was far less pronounced than the four- to sixfold increase seen in diet-induced obesity or Lepob/ob mice (23). Therefore, these data indicate that, in the absence of obesity, insulin resistance in macrophages and adipocytes is insufficient to trigger macrophage accumulation and stimulate cytokine production.

Previous studies have also suggested that activation of liver-resident macrophages, or Kupffer cells, plays a role in obesity/insulin resistance (42,43). Since L1 mice retained hepatic insulin resistance despite restored Insr signaling in hepatocytes, we investigated whether the neighboring Insr-deficient Kupffer cells contribute to local activation of inflammation and act in a paracrine fashion to cause insulin resistance. Interestingly, L1 mice showed a moderate decrease in Kupffer cell population compared with wild-type mice and a corresponding decrease in macrophage-specific gene expression in liver (Figs. 5 and 6). It is unclear whether this represents defects in Kupffer cell development and/or infiltration in the absence of Insr. In summary, our data suggest that insulin resistance in Kupffer cells does not play a major role in exacerbating hepatic insulin resistance in L1 mice.

Adiponectin signaling pathways and hepatic insulin sensitivity.

Both intracellular effectors of adiponectin (AMPK and PPARα) have been shown to regulate glucose homeostasis and lipid metabolism (44,45). In liver, AMPK activation suppresses gluconeogenesis (46). The blunting of adiponectin-dependent AMPK activation in liver of L1 mice can thus explain our prior observations of increased HGP during hyperinsulinemic clamps in this model (18). In contrast, PPARα expression/activity was not affected in normoglycemic L1 mice, while diabetic L1 mice exhibited decreases in PPARα target gene expression in liver (Fig. 5). These data raise the possibility that decreases in PPARα activation by adiponectin/AdipoR2 signaling, and, consequently, reduction of PPARα-dependent fatty acid oxidation, exacerbates hepatic insulin resistance and results in diabetes. This finding is consistent with our recent demonstration that gain-of-function in the transcription factor FoxO1, as seen in insulin resistance, results in decreased levels of PPARα targets (47).

The interactions between insulin and adiponectin are increasingly complex. In this study, we used L1 mice (18) to probe the contribution of adiponectin to insulin sensitivity in a model of tissue-specific insulin resistance. Our data indicate that insulin resistance can result in hyperadiponectinemia and adiponectin resistance. Moreover, stepwise decreases in adiponectin and AdipoR1 and -R2 levels are associated with the progression from insulin resistance to overt diabetes, raising the question of whether “adiponectin failure” is an integral component of the natural history of type 2 diabetes. Further work will be required to test this hypothesis in a mechanistic fashion.

FIG. 1.

Metabolic profile. A: Scatter plot of insulin versus glucose values in 3-month-old L1 and wild-type male littermates. B: Resistin values in 3- and 5-month-old mice. C: Adiponectin values in 3-month-old mice. D: Scatter plot of adiponectin versus glucose values in 3-month-old mice. **P < 0.001 by two-tailed Student's t test. Blue circles, wild type; red diamonds, L1; light blue line, mean of each group.

FIG. 1.

Metabolic profile. A: Scatter plot of insulin versus glucose values in 3-month-old L1 and wild-type male littermates. B: Resistin values in 3- and 5-month-old mice. C: Adiponectin values in 3-month-old mice. D: Scatter plot of adiponectin versus glucose values in 3-month-old mice. **P < 0.001 by two-tailed Student's t test. Blue circles, wild type; red diamonds, L1; light blue line, mean of each group.

FIG. 2.

Effects of intravenous adiponectin administration (2 μg/g body wt) on blood glucose levels in wild-type (A) and L1 (B) mice. Mice were fasted for 2 h before injection and refed 8 h after injection. n = 4 for wild-type adiponectin injection; n = 6–10 for all other groups. Data are means ± SE. *P < 0.05 vs. saline group by two-tailed Student's t test; **P < 0.01 vs. saline group; †P < 0.05 vs. wild-type adiponectin group. ○, saline injection; •, adiponectin injection.

FIG. 2.

Effects of intravenous adiponectin administration (2 μg/g body wt) on blood glucose levels in wild-type (A) and L1 (B) mice. Mice were fasted for 2 h before injection and refed 8 h after injection. n = 4 for wild-type adiponectin injection; n = 6–10 for all other groups. Data are means ± SE. *P < 0.05 vs. saline group by two-tailed Student's t test; **P < 0.01 vs. saline group; †P < 0.05 vs. wild-type adiponectin group. ○, saline injection; •, adiponectin injection.

FIG. 3.

mRNA expression of AdipoR1 and AdipoR2 in liver (A) and hind limb skeletal muscle (B) of wild-type (□), normoglycemic (▒), and diabetic L1 (▪) mice. Mice were killed, and liver and hind limb skeletal muscles were dissected and snap frozen in liquid nitrogen. Total RNA was isolated. AdipoR1 and AdipoR2 mRNA expression was measured by quantitative RT-PCR and normalized against β-actin mRNA. Data are means ± SE. n = 8–14. *P < 0.05 vs. wild type.

FIG. 3.

mRNA expression of AdipoR1 and AdipoR2 in liver (A) and hind limb skeletal muscle (B) of wild-type (□), normoglycemic (▒), and diabetic L1 (▪) mice. Mice were killed, and liver and hind limb skeletal muscles were dissected and snap frozen in liquid nitrogen. Total RNA was isolated. AdipoR1 and AdipoR2 mRNA expression was measured by quantitative RT-PCR and normalized against β-actin mRNA. Data are means ± SE. n = 8–14. *P < 0.05 vs. wild type.

FIG. 4.

Adiponectin-stimulated AMPKα phosphorylation in liver of 8-week-old wild-type (WT) and normoglycemic L1 mice. Mice were fasted for 6 h, intravenously injected with saline or adiponectin (2 μg/g body wt), and killed 20 min after injection. Representative phospho-Thr12 and total AMPKα immunoblots of liver lysates are shown. Ratio of phosphorylated versus total AMPKα was calculated and is shown as means ± SE. n = 4–6. *P < 0.05 vs. wild-type saline group; **P < 0.01 vs. wild-type saline group. AU, arbitrary units.

FIG. 4.

Adiponectin-stimulated AMPKα phosphorylation in liver of 8-week-old wild-type (WT) and normoglycemic L1 mice. Mice were fasted for 6 h, intravenously injected with saline or adiponectin (2 μg/g body wt), and killed 20 min after injection. Representative phospho-Thr12 and total AMPKα immunoblots of liver lysates are shown. Ratio of phosphorylated versus total AMPKα was calculated and is shown as means ± SE. n = 4–6. *P < 0.05 vs. wild-type saline group; **P < 0.01 vs. wild-type saline group. AU, arbitrary units.

FIG. 5.

mRNA expression of Pparα, Acox1, Cpt1α, and CD36 in liver (A) and hind limb skeletal muscle (B) of wild-type (□), normoglycemic (▒), and diabetic L1 (▪) mice. mRNA expression was measured by quantitative RT-PCR and normalized against β-actin mRNA. Data are means ± SE. n = 6–8. *P < 0.05 vs. wild type. AU, arbitrary units.

FIG. 5.

mRNA expression of Pparα, Acox1, Cpt1α, and CD36 in liver (A) and hind limb skeletal muscle (B) of wild-type (□), normoglycemic (▒), and diabetic L1 (▪) mice. mRNA expression was measured by quantitative RT-PCR and normalized against β-actin mRNA. Data are means ± SE. n = 6–8. *P < 0.05 vs. wild type. AU, arbitrary units.

FIG. 6.

Immunohistochemical and gene expression analyses of liver macrophages of wild-type and L1 mice. Kupffer cells were immunostained with F4/80 using DAB (brown). Representative images of wild-type (A), normoglycemic L1 (B), and diabetic L1 (C) mice are shown. D: Kupffer cells were quantified as number of F4/80-positive cells per five opitical fields (×40 objective) in each animal (n = 8 for each group). mRNA expression of ABCg1 (E) and Pparγ (F) was determined by quantitative RT-PCR in liver of wild-type, normoglycemic, and diabetic L1 mice (n = 6–8). Data were normalized against β-actin mRNA. Data are means ± SE. *P < 0.05 vs. wild type; **P < 0.01 vs. wild type. AU, arbitrary units; □, wild type;▒, normoglycemic L1; ▪, diabetic L1.

FIG. 6.

Immunohistochemical and gene expression analyses of liver macrophages of wild-type and L1 mice. Kupffer cells were immunostained with F4/80 using DAB (brown). Representative images of wild-type (A), normoglycemic L1 (B), and diabetic L1 (C) mice are shown. D: Kupffer cells were quantified as number of F4/80-positive cells per five opitical fields (×40 objective) in each animal (n = 8 for each group). mRNA expression of ABCg1 (E) and Pparγ (F) was determined by quantitative RT-PCR in liver of wild-type, normoglycemic, and diabetic L1 mice (n = 6–8). Data were normalized against β-actin mRNA. Data are means ± SE. *P < 0.05 vs. wild type; **P < 0.01 vs. wild type. AU, arbitrary units; □, wild type;▒, normoglycemic L1; ▪, diabetic L1.

TABLE 1

Metabolic characteristics of the animals

Wild typeNormo-glycemic L1Diabetic L1
Body weight (g) 25.6 ± 0.9 24.9 ± 0.9 29.3 ± 1.2* 
Fat mass (g) 4.8 ± 0.5 5.1 ± 0.4 5.6 ± 0.6 
Fat-free mass (g) 20.7 ± 1.1 19.8 ± 0.7 23.7 ± 1.2 
Glucose (mg/dl) 160 ± 9 154 ± 8 240 ± 23* 
FFA (mEq/l) 0.20 ± 0.04 0.37 ± 0.06* 0.33 ± 0.06 
Triglycerides (mg/ml) 0.10 ± 0.04 0.24 ± 0.04* 0.24 ± 0.06* 
β-OH butyrate (mmol/l) 1.31 ± 0.15 2.29 ± 0.38* ND 
Insulin (ng/ml) 1.0 ± 0.3 12.3 ± 2.1§ 19.3 ± 3.5§ 
Adiponectin (μg/ml) 8.5 ± 1.6 21.2 ± 3.1§ 15.4 ± 2.1* 
Resistin (ng/ml) 3.3 ± 0.4 4.1 ± 0.5 3.1 ± 0.4 
Wild typeNormo-glycemic L1Diabetic L1
Body weight (g) 25.6 ± 0.9 24.9 ± 0.9 29.3 ± 1.2* 
Fat mass (g) 4.8 ± 0.5 5.1 ± 0.4 5.6 ± 0.6 
Fat-free mass (g) 20.7 ± 1.1 19.8 ± 0.7 23.7 ± 1.2 
Glucose (mg/dl) 160 ± 9 154 ± 8 240 ± 23* 
FFA (mEq/l) 0.20 ± 0.04 0.37 ± 0.06* 0.33 ± 0.06 
Triglycerides (mg/ml) 0.10 ± 0.04 0.24 ± 0.04* 0.24 ± 0.06* 
β-OH butyrate (mmol/l) 1.31 ± 0.15 2.29 ± 0.38* ND 
Insulin (ng/ml) 1.0 ± 0.3 12.3 ± 2.1§ 19.3 ± 3.5§ 
Adiponectin (μg/ml) 8.5 ± 1.6 21.2 ± 3.1§ 15.4 ± 2.1* 
Resistin (ng/ml) 3.3 ± 0.4 4.1 ± 0.5 3.1 ± 0.4 

Data are means ± SE. β-OH butyrate was measured after a 16-h fast. All other measurements were made in 3-month-old fed male mice.

*

P < 0.05 vs. wild type;

§

P < 0.01 vs. wild type;

P < 0.05 vs. normoglycemic L1;

P < 0.01 vs. normoglycemic L1. FFA, free fatty acid; ND, not determined.

TABLE 2

Plasma cytokine levels

Cytokines (pg/ml)Wild typeNormo-glycemic L1Diabetic L1
IL-1β 37.6 ± 4.4 46.9 ± 9.7 46.2 ± 5.2 
IL-2 33.6 ± 1.4 34.0 ± 1.5 ND 
IL-6 24.5 ± 2.1 28.3 ± 4.7 27.8 ± 2.2 
IL-12 119.9 ± 27.4 121.4 ± 13.6 ND 
GM-CSF 24.4 ± 3.8 31.7 ± 7.1 31.2 ± 4.2 
TNFα 36.4 ± 0.9 37.7 ± 2.2 38.6 ± 1.6 
Cytokines (pg/ml)Wild typeNormo-glycemic L1Diabetic L1
IL-1β 37.6 ± 4.4 46.9 ± 9.7 46.2 ± 5.2 
IL-2 33.6 ± 1.4 34.0 ± 1.5 ND 
IL-6 24.5 ± 2.1 28.3 ± 4.7 27.8 ± 2.2 
IL-12 119.9 ± 27.4 121.4 ± 13.6 ND 
GM-CSF 24.4 ± 3.8 31.7 ± 7.1 31.2 ± 4.2 
TNFα 36.4 ± 0.9 37.7 ± 2.2 38.6 ± 1.6 

Data are means ± SE. Measurements were made in 3-month-old male mice under fed conditions. GM-CSF, granulocyte-macrophage colony-stimulating factor; ND, not determined; TNF, tumor necrosis factor.

Published ahead of print at http://diabetes.diabetesjournals.org on 2 May 2007. DOI: 10.2337/db07-0127.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

This study was supported by National Institutes of Health Grants DK58282 (to D.A.), DK071030 (to P.E.S.), and P30DK63608 (to the Columbia Diabetes and Endocrinology Research Center).

We thank members of the Accili, Scherer, Rossetti, and Shapiro laboratories for helpful discussions and critical review of the data.

1.
Accili D: Lilly Lecture 2003: The struggle for mastery in insulin action: from triumvirate to republic.
Diabetes
53
:
1633
–1642,
2004
2.
Cherrington AD: Banting Lecture 1997: Control of glucose uptake and release by the liver in vivo.
Diabetes
48
:
1198
–1214,
1999
3.
Cherrington AD: The role of hepatic insulin receptors in the regulation of glucose production.
J Clin Invest
115
:
1136
–1139,
2005
4.
Accili D, Arden KC: FoxOs at the crossroads of cellular metabolism, differentiation, and transformation.
Cell
117
:
421
–426,
2004
5.
Boden G, Chen X, Ruiz J, White JV, Rossetti L: Mechanisms of fatty acid-induced inhibition of glucose uptake.
J Clin Invest
93
:
2438
–2446,
1994
6.
Rebrin K, Steil GM, Mittelman SD, Bergman RN: Causal linkage between insulin suppression of lipolysis and suppression of liver glucose output in dogs.
J Clin Invest
98
:
741
–749,
1996
7.
Lewis GF, Vranic M, Harley P, Giacca A: Fatty acids mediate the acute extrahepatic effects of insulin on hepatic glucose production in humans.
Diabetes
46
:
1111
–1119,
1997
8.
Sindelar DK, Chu CA, Venson P, Donahue EP, Neal DW, Cherrington AD: Basal hepatic glucose production is regulated by the portal vein insulin concentration.
Diabetes
47
:
523
–529,
1998
9.
Obici S, Zhang BB, Karkanias G, Rossetti L: Hypothalamic insulin signaling is required for inhibition of glucose production.
Nat Med
8
:
1376
–1382,
2002
10.
Ishihara H, Maechler P, Gjinovci A, Herrera PL, Wollheim CB: Islet beta-cell secretion determines glucagon release from neighbouring alpha-cells.
Nat Cell Biol
5
:
330
–335,
2003
11.
Diao J, Asghar Z, Chan CB, Wheeler MB: Glucose-regulated glucagon secretion requires insulin receptor expression in pancreatic alpha-cells.
J Biol Chem
280
:
33487
–33496,
2005
12.
Sindelar DK, Balcom JH, Chu CA, Neal DW, Cherrington AD: A comparison of the effects of selective increases in peripheral or portal insulin on hepatic glucose production in the conscious dog.
Diabetes
45
:
1594
–1604,
1996
13.
Sindelar DK, Chu CA, Rohlie M, Neal DW, Swift LL, Cherrington AD: The role of fatty acids in mediating the effects of peripheral insulin on hepatic glucose production in the conscious dog.
Diabetes
46
:
187
–196,
1997
14.
Obici S, Feng Z, Karkanias G, Baskin DG, Rossetti L: Decreasing hypothalamic insulin receptors causes hyperphagia and insulin resistance in rats.
Nat Neurosci
5
:
566
–572,
2002
15.
Shoelson SE, Lee J, Goldfine AB: Inflammation and insulin resistance.
J Clin Invest
116
:
1793
–1801,
2006
16.
Yamauchi T, Kaburagi Y, Ueki K, Tsuji Y, Stark GR, Kerr IM, Tsushima T, Akanuma Y, Komuro I, Tobe K, Yazaki Y, Kadowaki T: Growth hormone and prolactin stimulate tyrosine phosphorylation of insulin receptor substrate-1, -2, and -3, their association with p85 phosphatidylinositol 3-kinase (PI3-kinase), and concomitantly PI3-kinase activation via JAK2 kinase.
J Biol Chem
273
:
15719
–15726,
1998
17.
Okamoto H, Hribal ML, Lin HV, Bennett WR, Ward A, Accili D: Role of the forkhead protein FoxO1 in beta cell compensation to insulin resistance.
J Clin Invest
116
:
775
–782,
2006
18.
Okamoto H, Obici S, Accili D, Rossetti L: Restoration of liver insulin signaling in Insr knockout mice fails to normalize hepatic insulin action.
J Clin Invest
115
:
1314
–1322,
2005
19.
Okamoto H, Nakae J, Kitamura T, Park BC, Dragatsis I, Accili D: Transgenic rescue of insulin receptor-deficient mice.
J Clin Invest
114
:
214
–223,
2004
20.
Lauro D, Kido Y, Castle AL, Zarnowski MJ, Hayashi H, Ebina Y, Accili D: Impaired glucose tolerance in mice with a targeted impairment of insulin action in muscle and adipose tissue.
Nat Genet
20
:
294
–298,
1998
21.
Kim CH, Pennisi P, Zhao H, Yakar S, Kaufman JB, Iganaki K, Shiloach J, Scherer PE, Quon MJ, LeRoith D: MKR mice are resistant to the metabolic actions of both insulin and adiponectin: discordance between insulin resistance and adiponectin responsiveness.
Am J Physiol Endocrinol Metab
291
:
E298
–E305,
2006
22.
Berg AH, Combs TP, Du X, Brownlee M, Scherer PE: The adipocyte-secreted protein Acrp30 enhances hepatic insulin action.
Nat Med
7
:
947
–953,
2001
23.
Weisberg SP, McCann D, Desai M, Rosenbaum M, Leibel RL, Ferrante AW Jr: Obesity is associated with macrophage accumulation in adipose tissue.
J Clin Invest
112
:
1796
–1808,
2003
24.
Yamauchi T, Kamon J, Ito Y, Tsuchida A, Yokomizo T, Kita S, Sugiyama T, Miyagishi M, Hara K, Tsunoda M, Murakami K, Ohteki T, Uchida S, Takekawa S, Waki H, Tsuno NH, Shibata Y, Terauchi Y, Froguel P, Tobe K, Koyasu S, Taira K, Kitamura T, Shimizu T, Nagai R, Kadowaki T: Cloning of adiponectin receptors that mediate antidiabetic metabolic effects.
Nature
423
:
762
–769,
2003
25.
Hug C, Wang J, Ahmad NS, Bogan JS, Tsao TS, Lodish HF: T-cadherin is a receptor for hexameric and high-molecular-weight forms of Acrp30/adiponectin.
Proc Natl Acad Sci U S A
101
:
10308
–10313,
2004
26.
Yamauchi T, Kamon J, Minokoshi Y, Ito Y, Waki H, Uchida S, Yamashita S, Noda M, Kita S, Ueki K, Eto K, Akanuma Y, Froguel P, Foufelle F, Ferre P, Carling D, Kimura S, Nagai R, Kahn BB, Kadowaki T: Adiponectin stimulates glucose utilization and fatty-acid oxidation by activating AMP-activated protein kinase.
Nat Med
8
:
1288
–1295,
2002
27.
Tomas E, Tsao TS, Saha AK, Murrey HE, Zhang Cc C, Itani SI, Lodish HF, Ruderman NB: Enhanced muscle fat oxidation and glucose transport by ACRP30 globular domain: acetyl-CoA carboxylase inhibition and AMP-activated protein kinase activation.
Proc Natl Acad Sci U S A
99
:
16309
–16313,
2002
28.
Combs TP, Berg AH, Obici S, Scherer PE, Rossetti L: Endogenous glucose production is inhibited by the adipose-derived protein Acrp30.
J Clin Invest
108
:
1875
–1881,
2001
29.
Yamauchi T, Kamon J, Waki H, Terauchi Y, Kubota N, Hara K, Mori Y, Ide T, Murakami K, Tsuboyama-Kasaoka N, Ezaki O, Akanuma Y, Gavrilova O, Vinson C, Reitman ML, Kagechika H, Shudo K, Yoda M, Nakano Y, Tobe K, Nagai R, Kimura S, Tomita M, Froguel P, Kadowaki T: The fat-derived hormone adiponectin reverses insulin resistance associated with both lipoatrophy and obesity.
Nat Med
7
:
941
–946,
2001
30.
Liang CP, Han S, Okamoto H, Carnemolla R, Tabas I, Accili D, Tall AR: Increased CD36 protein as a response to defective insulin signaling in macrophages.
J Clin Invest
113
:
764
–773,
2004
31.
Cecchini MG, Dominguez MG, Mocci S, Wetterwald A, Felix R, Fleisch H, Chisholm O, Hofstetter W, Pollard JW, Stanley ER: Role of colony stimulating factor-1 in the establishment and regulation of tissue macrophages during postnatal development of the mouse.
Development
120
:
1357
–1372,
1994
32.
Akiyama TE, Sakai S, Lambert G, Nicol CJ, Matsusue K, Pimprale S, Lee YH, Ricote M, Glass CK, Brewer HB Jr, Gonzalez FJ: Conditional disruption of the peroxisome proliferator-activated receptor gamma gene in mice results in lowered expression of ABCA1, ABCG1, and apoE in macrophages and reduced cholesterol efflux.
Mol Cell Biol
22
:
2607
–2619,
2002
33.
Li AC, Binder CJ, Gutierrez A, Brown KK, Plotkin CR, Pattison JW, Valledor AF, Davis RA, Willson TM, Witztum JL, Palinski W, Glass CK: Differential inhibition of macrophage foam-cell formation and atherosclerosis in mice by PPARalpha, beta/delta, and gamma.
J Clin Invest
114
:
1564
–1576,
2004
34.
Kadowaki T, Yamauchi T, Kubota N, Hara K, Ueki K, Tobe K: Adiponectin and adiponectin receptors in insulin resistance, diabetes, and the metabolic syndrome.
J Clin Invest
116
:
1784
–1792,
2006
35.
Bluher M, Michael MD, Peroni OD, Ueki K, Carter N, Kahn BB, Kahn CR: Adipose tissue selective insulin receptor knockout protects against obesity and obesity-related glucose intolerance.
Dev Cell
3
:
25
–38,
2002
36.
Qiao L, Schaack J, Shao J: Suppression of adiponectin gene expression by histone deacetylase inhibitor valproic acid.
Endocrinology
147
:
865
–874,
2006
37.
Semple RK, Soos MA, Luan J, Mitchell CS, Wilson JC, Gurnell M, Cochran EK, Gorden P, Chatterjee VK, Wareham NJ, O'Rahilly S: Elevated plasma adiponectin in humans with genetically defective insulin receptors.
J Clin Endocrinol Metab
91
:
3219
–3223,
2006
38.
Kido Y, Philippe N, Schaffer AA, Accili D: Genetic modifiers of the insulin resistance phenotype in mice.
Diabetes
49
:
589
–596,
2000
39.
Tsuchida A, Yamauchi T, Ito Y, Hada Y, Maki T, Takekawa S, Kamon J, Kobayashi M, Suzuki R, Hara K, Kubota N, Terauchi Y, Froguel P, Nakae J, Kasuga M, Accili D, Tobe K, Ueki K, Nagai R, Kadowaki T: Insulin/Foxo1 pathway regulates expression levels of adiponectin receptors and adiponectin sensitivity.
J Biol Chem
279
:
30817
–30822,
2004
40.
Kitamura YI, Kitamura T, Kruse JP, Raum JC, Stein R, Gu W, Accili D: FoxO1 protects against pancreatic beta cell failure through NeuroD and MafA induction.
Cell Metab
2
:
153
–163,
2005
41.
Wellen KE, Hotamisligil GS: Inflammation, stress, and diabetes.
J Clin Invest
115
:
1111
–1119,
2005
42.
Cai D, Yuan M, Frantz DF, Melendez PA, Hansen L, Lee J, Shoelson SE: Local and systemic insulin resistance resulting from hepatic activation of IKK-beta and NF-kappaB.
Nat Med
11
:
183
–190,
2005
43.
Arkan MC, Hevener AL, Greten FR, Maeda S, Li ZW, Long JM, Wynshaw-Boris A, Poli G, Olefsky J, Karin M: IKK-beta links inflammation to obesity-induced insulin resistance.
Nat Med
11
:
191
–198,
2005
44.
Kahn BB, Alquier T, Carling D, Hardie DG: AMP-activated protein kinase: ancient energy gauge provides clues to modern understanding of metabolism.
Cell Metab
1
:
15
–25,
2005
45.
Haluzik MM, Haluzik M: PPAR-alpha and insulin sensitivity.
Physiol Res
55
:
115
–122,
2006
46.
Zhou G, Myers R, Li Y, Chen Y, Shen X, Fenyk-Melody J, Wu M, Ventre J, Doebber T, Fujii N, Musi N, Hirshman MF, Goodyear LJ, Moller DE: Role of AMP-activated protein kinase in mechanism of metformin action.
J Clin Invest
108
:
1167
–1174,
2001
47.
Matsumoto M, Han S, Kitamura T, Accili D: Dual role of transcription factor FoxO1 in controlling hepatic insulin sensitivity and lipid metabolism.
J Clin Invest
116
:
2464
–2472,
2006