Cardiac-Specific Overexpression of Peroxisome Proliferator–Activated Receptor-α Causes Insulin Resistance in Heart and Liver

Male MHC-PPARα mice and wild-type littermates (10) weighing ∼27 g (12–14 weeks of age) were housed under controlled temperature (23°C) and lighting (12 h of light, 0700–1900; 12 h of dark, 1900–0700), with free access to water and standard mouse diet. At least 4 days before in vivo experiments, whole-body fat and lean mass were measured in awake mice using ¹H-magnetic resonance spectroscopy (Bruker Mini-spec Analyzer; Echo Medical Systems, Houston, TX). Following the body composition measurement, mice were anesthetized with an intraperitoneal injection of ketamine (100 mg/kg body wt) and xylazine (10 mg/kg body wt), and an indwelling catheter was inserted in the right internal jugular vein. On the day of experiment (4–5 days postsurgery), a 3-way connector was attached to the jugular vein catheter to intravenously deliver solutions (e.g., glucose, insulin). Also, mice were placed in a rat-size restrainer (to minimize stress during experiments in awake state) and tail restrained using a tape to obtain blood samples from the tail vessels. All procedures were approved by Yale University Animal Care and Use Committee.

Basal rates of glucose and palmitate uptake in heart were assessed in awake MHC-PPARα mice and wild-type littermates (n = 6 for both groups). Basal cardiac glucose uptake was assessed in overnight-fasted mice by intravenously bolus injecting 2-deoxy-d-[1-¹⁴C]glucose (2-[¹⁴C]DG; 10 μCi; PerkinElmer Life and Analytical Sciences, Boston, MA) and obtaining heart samples 30 min later for the measurement of 2-[¹⁴C]DG-6-phosphate (2-[¹⁴C]DG-6-P). Plasma samples were taken at 5- to 10-min intervals following injection to measure the concentrations of 2-[¹⁴C]DG. Basal cardiac palmitate uptake was similarly assessed in mice by intravenously injecting [¹⁴C]palmitate (10 μCi; PerkinElmer Life and Analytical Sciences) and obtaining heart samples 5 min later for the measurement of [¹⁴C] incorporation into triglyceride. Plasma samples were taken at 30-s to 1-min intervals following injection to measure the concentrations of [¹⁴C]palmitate.

After overnight fast, a 2-h hyperinsulinemic-euglycemic clamp was conducted in awake MHC-PPARα mice and wild-type littermates (n = 6–13). The experiment began with a primed-continuous infusion of human regular insulin (Humulin; Eli Lilly, Indianapolis, IN) at a rate of 15 pmol · kg⁻¹ · min⁻¹ to raise plasma insulin within a physiological range (∼200 pmol/l) (16). Blood samples (20 μl) were collected at 10- to 20-min intervals for the immediate measurement of plasma glucose concentration, and 20% glucose was infused at variable rates to maintain glucose at basal concentrations (∼6 mmol/l). Basal and insulin-stimulated whole-body glucose uptake was estimated with a continuous infusion of [3-³H]glucose (PerkinElmer Life and Analytical Sciences) for 2 h before the clamps (0.05 μCi/min) and throughout the clamps (0.1 μCi/min), respectively. All infusions were performed using the microdialysis pumps (CMA/Microdialysis, North Chelmsford, MA). To estimate insulin-stimulated glucose uptake in individual tissues, 2-[¹⁴C]DG was administered as a bolus (10 μCi) at 75 min after the start of clamps. Blood samples were taken before, during, and at the end of clamps for measurement of plasma [³H]glucose, ³H₂O, 2-[¹⁴C]DG concentrations, and/or insulin concentrations. At the end of the clamps, mice were anesthetized with sodium pentobarbital injection, and tissue samples (gastrocnemius, tibialis anterior, and quadriceps from both hindlimbs, epididymal white adipose tissue, interscapular brown adipose tissue, liver, and heart) were rapidly taken for biochemical and molecular analysis.

Glucose concentration during clamps was analyzed using 10 μl plasma by a glucose oxidase method on a Beckman Glucose Analyzer 2 (Beckman, Fullerton, CA). Plasma insulin concentration was measured by radioimmunoassay using kits from Linco Research (St. Charles, MO). Circulating levels of interleukin (IL)-6 were measured using an enzyme-linked immunosorbent assay kit from R&D Systems (Quantikine, Minneapolis, MN). Plasma concentrations of [3-³H]glucose, 2-[¹⁴C]DG, and ³H₂O were determined following deproteinization of plasma samples as previously described (16). The radioactivity of ³H in tissue glycogen was determined by digesting tissue samples in KOH and precipitating glycogen with ethanol. For the determination of tissue 2-[¹⁴C]DG-6-P content, tissue samples were homogenized, and the supernatants were subjected to an ion-exchange column to separate 2-[¹⁴C]DG-6-P from 2-[¹⁴C]DG.

Rates of basal hepatic glucose production (HGP) and insulin-stimulated whole-body glucose uptake were determined as the ratio of the [³H]glucose infusion rate (disintegrations per minute; dpm/min) to the specific activity of plasma glucose (dpm/μmol) at the end of basal period and during the final 30 min of clamps, respectively. Insulin-stimulated rates of HGP during clamps were determined by subtracting the glucose infusion rate from the whole-body glucose uptake. Whole-body glycolysis was calculated from the rate of increase in plasma ³H₂O concentration, determined by linear regression of the measurements at 80, 90, 100, 110, and 120 min of clamps. Whole-body glycogen plus lipid synthesis from glucose was estimated by subtracting whole-body glycolysis from whole-body glucose uptake (16). Since 2-deoxyglucose is a glucose analog that is phosphorylated but not further metabolized, insulin-stimulated glucose uptake in individual tissues can be estimated by determining the tissue (e.g., skeletal muscle, heart) content of 2-[¹⁴C]DG-6-P. Based on this, glucose uptake in individual tissues was calculated from plasma 2-[¹⁴C]DG profile, which was fitted with a double exponential or linear curve using MLAB (Civilized Software, Bethesda, MD) and tissue 2-[¹⁴C]DG-6-P content. Tissue glycogen synthesis was assessed as previously described (16).

Heart and liver samples were obtained at the end of insulin clamps to measure in vivo activities of insulin receptor substrate (IRS)-1 and IRS-2–associated phosphatidylinositol (PI) 3-kinase, respectively. These activities were assessed by immunoprecipitating tissue lysates with polyclonal IRS-1/IRS-2 antibodies (Upstate Biotechnology, Lake Placid, NY) coupled to protein A-Sepharose beads (Sigma, St. Louis, MO) and assessing the incorporation of ³²P into PI to yield PI 3-monophosphate. The immune complex was washed, and the activity was determined as previously described (16). In additional groups of mice (n = 6 for MHC-PPARα mice and wild-type littermates), basal and insulin-stimulated insulin signaling were assessed following intraperitoneal injection of saline and insulin, respectively, and the heart samples were taken 15 min following the injection. Insulin-stimulated Akt activity was determined in heart samples obtained at the end of clamps as previously described (17). Basal Akt activities were measured in overnight-fasted MHC-PPARα mice and wild-type mice (n = 3 for each group).

For the measurement of signal transducer and activator of transcription 3 (STAT3) protein expression and tyrosine phosphorylation of STAT3, heart and liver samples were obtained at the end of clamps. Sample lysates were resolved by SDS-PAGE (8% gel) and transferred to nitrocellulose membranes. The nitrocellulose membranes were blocked and incubated with phospho-specific STAT3 (Tyr705) polyclonal antibody (Cell Signaling, Beverly, MA). The membranes were washed and incubated with horseradish peroxidase secondary antibody (Amersham Pharmaceuticals, Arlington Heights, IL), and the bands were visualized using the enhanced chemiluminescence system (Amersham Pharmaceuticals). For the PPARα expression in individual organs, whole-cell protein extracts were prepared from heart and liver samples, and Western blotting studies were performed using rabbit-derived antibodies directed against the Flag epitope tag (Sigma) as previously described (18). For the p38 mitogen-activated protein kinase (MAPK) activity, heart samples were homogenized in 20 vol of homogenizing buffer (20 mmol/l Hepes, pH 7.4; 150 mmol/l NaCl; 2 mmol/l EDTA; 1% TritonX-100; 0.1% SDS; 10% glycerol; 1 mmol/l dithiothreitol; 0.5% deoxycholate, supplemented with 5 μg/ml leupeptin; 5 μg/ml aprotinin; 1 μg/ml pepstatin A; 1 mmol/l phenylmethylsulfonyl fluoride; 1 mmol/l benzamidine; 2 mmol/l Na₃VO₄; and 5 mmol/l NaF) and lysed for 30 min before clarification at 20,800g for 30 min. Protein concentration was determined by Coomassie protein reagent (Pierce, Rockford, IL). p38 MAPK was immunoprecipitated using anti-p38 MAPK antibody (Santa Cruz) and collected on protein-G Sepharose beads. The immune complexes were subjected to an in vitro kinase assay as previously described (19). Glycogen synthase kinase (GSK)3 phosphorylation was assessed by Western blot analysis using antibodies directed against phosphorylated GSK3-α/-β (SER21/9) (Cell Signaling Technologies) and total GSK3-β. A total of 20 μg protein were run on SDS-PAGE, transferred to nitrocellulose, and reacted with primary antibody and horseradish peroxidase–conjugated secondary antibody. Signal was detected using enhanced chemiluminescence plus (Amersham Biosciences) and the ratio of phosphorylated to total GSK3 quantified on using a STORM 860 (Molecular Dynamics).

For the measurement of tissue-specific triglyceride concentration, heart and liver samples were digested in chloroform-methanol (20). Lipid layer was separated using H₂SO₄, and concentrations were determined using triglyceride assay kit (Sigma Diagnostics) and spectrophotometry. To determine the intramyocardial and intrahepatic concentrations of fatty acid metabolites (long-chain acyl CoAs, diacylglycerol) using LC/MS/MS (liquid-chromatography tandem mass spectrometry), heart and liver samples were homogenized and extracted as previously described (21,22). Briefly, frozen tissue samples (∼100 mg) were grounded under liquid nitrogen and homogenized in 1 ml of 100 mmol/l KH₂PO₄ (pH 4.9) and 1 ml of 2-propanol. Heptadecanoyl CoA was added as internal standard. A total of 125 μl saturated (NH₄)₂SO₄ and 2 ml acetonitrile were added to the suspension then vortexed for 2 min. The emulsion was centrifuged for 10 min at 4,000 rpm, and then the supernatant was diluted with 5 ml of 100 mmol/l KH₂PO₄ (pH 4.9) for the solid-phase extraction. Before the loading, oligonucleotide purification cartridge columns were conditioned with 5 ml acetonitrile and 2 ml of 25 mmol/l KH₂PO₄ (pH 4.9). After loading the samples, the cartridges were washed with at least 10 ml distilled H₂O, and then long-chain acyl CoAs were eluted slowly with 0.5 ml of 60% acetonitrile. The eluent was dried in Speedvac and finally reconstituted in 100 μl methanol/H₂O for ESI/MS/MS (electrospray ionization tandem mass spectrometry) analysis.

PerkinElmer sciex API 3000 tandem mass spectrometer interfaced with TurboIonSpray ionization source was used for the analysis. The intracellular concentrations of long-chain fatty acyl CoAs (C16:0, C16:1, C18:0, C18:1, C18:2, and C18:3) were detected in negative electrospray mode. The doubly charged ions of these compounds were transmitted, and singly charged product ions were quantified in multiple reaction mode. Long-chain acyl CoA standards were purchased from Sigma Chemical (St. Louis, MO). The calibration of long-chain acyl CoAs showed consistent linearity from 0.2 to 20 ng/μl, and coefficient of variance was 2.1–5.5% for all long-chain acyl CoA species. The intracellular concentrations of diacylglycerol were measured as previously described (20).

M-mode echocardiography was performed in additional groups of age-matched MHC-PPARα mice and wild-type littermates (n = 5 for both groups) using the Phillips Sonos 5500 System with a 15-MHz probe. Mice were lightly anesthetized with inhaled isoflurane, and images were collected in the short and long axis. Data represents the averaged values of 3–5 cardiac cycles.

To determine the mechanism of hepatic insulin resistance, hyperinsulinemic-euglycemic clamps were performed in additional groups of male MHC-PPARα mice and wild-type littermates at 8 weeks of age (n = 6–7). Basal Akt activities in heart were measured in overnight-fasted MHC-PPARα mice and wild-type littermates at ∼8 weeks of age (n = 4 for each group). Echocardiography was performed to assess ventricular functions in additional group of 8-week-old male MHC-PPARα and wild-type mice (n = 5 for each group).

Data are expressed as means ± SE. The significance of the difference in mean values between wild-type mice versus MHC-PPARα mice was evaluated using the Student’s t test. The statistical significance was at the P < 0.05 level, unless stated otherwise.

Whole-body lean and fat mass as well as overnight fasted plasma glucose and insulin levels did not differ between the male MHC-PPARα mice and wild-type littermates (12–14 weeks of age) (Table 1). To determine the effects of heart-selective overexpression of PPARα on basal cardiac energy metabolism in vivo, [¹⁴C]palmitate was intravenously administered in awake mice. Basal palmitate uptake was increased by more than twofold in the MHC-PPARα mice (Fig. 1A), consistent with previous micro-PET measurement (10). Additionally, basal cardiac glucose uptake, as assessed using 2-[¹⁴C]deoxyglucose, was reduced by 70% in the MHC-PPARα mice (Fig. 1B), also consistent with previous measurements using micro-PET and in isolated working hearts (10).

Cardiac insulin signaling and insulin-stimulated glucose metabolism were assessed during a 2-h hyperinsulinemic-euglycemic clamp in awake mice. During the clamp experiments, plasma insulin concentration was raised to ∼200 pmol/l, while the plasma glucose concentration was maintained at ∼6 mmol/l by a variable infusion of glucose in both groups (Table 1). Insulin clamp increased cardiac glucose uptake by 10-fold in the wild-type mice compared with the basal state (Fig. 1C). In contrast, insulin-stimulated cardiac glucose uptake was significantly blunted in the MHC-PPARα mice (Fig. 1C), which was consistent with profoundly reduced GLUT4 content in the heart of MHC-PPARα mice (10). Defects in cardiac insulin action were further associated with an 80% decrease in insulin-stimulated glycogen synthesis in the heart of MHC-PPARα mice (Fig. 1D). Since glucose transport is a rate-controlling step in glucose utilization (23), decreases in insulin-stimulated glycogen synthesis may be partly accounted for by cardiac insulin resistance in the MHC-PPARα mice.

Previous studies (20,22) have shown that insulin resistance mediated by increased flux of fatty acids into skeletal muscle was secondary to alteration in insulin signaling activity. To determine whether cardiac insulin resistance in the MHC-PPARα mice was associated with defects in cardiac insulin signaling in vivo, we assessed insulin-stimulated activities of IRS-1–associated PI 3-kinase and Akt, which are key intracellular mediators of insulin signaling (24,25), in tissues obtained at the end of 2-h insulin clamp experiments. Defects in insulin-mediated cardiac glucose metabolism were associated with 30–50% decreases in insulin-stimulated IRS-1–associated PI 3-kinase and Akt activity in the MHC-PPARα mice compared with the wild-type mice (Figs. 2A and B). Additionally, basal and insulin-stimulated IRS-1–associated PI 3-kinase activities in heart were assessed following intraperitoneal injection of insulin in another group of mice and showed no significant differences in the MHC-PPARα mice (Fig. 2C). However, insulin increased IRS-1–associated PI 3-kinase activity by approximately sevenfold over basal in the heart of wild-type mice, whereas insulin-mediated increase in IRS-1–associated PI 3-kinase activity over basal was significantly blunted in the heart of MHC-PPARα mice (approximately fourfold) (Fig. 2D). Basal Akt activity also showed a tendency to be reduced in the MHC-PPARα mice, although the difference did not reach statistical significance (data not shown). Altogether, these results suggest that cardiac insulin resistance in the MHC-PPARα mice may be attributed to both reduced GLUT4 contents and blunted insulin signaling activity.

Since our recent studies demonstrated an important causative role of intracellular fatty acid–derived metabolites (i.e., fatty acyl CoA, diacylglycerol) on skeletal muscle insulin resistance (20,22), we assessed intramyocardial levels of triglyceride, fatty acyl CoAs, and diacylglycerol. Myocardial triglyceride levels were increased by twofold in the MHC-PPARα mice, which is consistent with increased lipid uptake in these mice (Fig. 3A). In contrast, individual and total (sum of C16:0, C16:1, C18:0, C18:1, C18:2, and C18:3) levels of intramyocardial fatty acyl CoAs were significantly (C18:2) or showed a tendency to be reduced in the MHC-PPARα mice (Fig. 3B). Consistent with the previous observation, intramyocardial levels of total diacylglycerol were not significantly altered in the MHC-PPARα mice (Fig. 3C).

Inoue et al. (26) have recently shown that liver-specific STAT3 knockout mice developed hepatic insulin resistance and suggested the role of STAT3 in hepatic glucose metabolism. Thus, we examined the expression of total STAT3 and insulin-stimulated STAT3 activity in the heart of MHC-PPARα mice. Interestingly, insulin-stimulated tyrosine phosphorylation of STAT3 was significantly reduced by 40% in the MHC-PPARα mice compared with the wild-type littermates, whereas total STAT3 expression levels were not altered in the MHC-PPARα mice (Fig. 4A and C). In contrast, myocardial activities of p38 MAPK, GSK3-α/-β, or expression of GSK3-β were not altered in the MHC-PPARα mice (Fig. 4A and D).

We found that the steady-state rates of glucose infusion during clamp experiments were reduced by ∼30% in the MHC-PPARα mice, reflecting whole-body insulin resistance in the MHC-PPARα mice (Fig. 5A). Basal HGP did not differ between the groups, but HGP during the insulin-stimulated state (clamps) was significantly increased in the MHC-PPARα mice (Fig. 5B). As a result, hepatic insulin action, reflected as the insulin-mediated suppression of basal HGP, was markedly reduced in the MHC-PPARα mice compared with the wild-type littermates (Fig. 5C). In contrast, insulin-stimulated skeletal muscle glucose uptake was not altered in the MHC-PPARα mice (Fig. 5D). Additionally, insulin-stimulated whole-body glucose uptake and metabolic fluxes (i.e., glycolysis, glycogen plus lipid synthesis) were unaltered in the MHC-PPARα mice (data not shown). These findings indicate that the MHC-PPARα mice developed insulin resistance selectively in heart and liver, which accounted for the reduced glucose infusion rates during clamps in the MHC-PPARα mice.

Hepatic insulin resistance in the MHC-PPARα mice was associated with a 30% decrease in insulin-stimulated IRS-2–associated PI 3-kinase activity in the liver of MHC-PPARα mice (Fig. 6A). Similar to the findings in heart, hepatic triglyceride content was significantly elevated in the MHC-PPARα mice, whereas individual and total levels of intrahepatic fatty acyl CoAs were either significantly reduced or showed a tendency to be reduced in the MHC-PPARα mice (Fig. 6B and C). Additionally, intrahepatic total diacylglycerol levels were not altered in the MHC-PPARα mice (data not shown). In contrast to the findings in heart, insulin-mediated tyrosine phosphorylation of STAT3 was not altered in the liver of MHC-PPARα mice (data not shown). To determine whether altered hepatic lipid metabolism may be due to increased gene expression, we measured hepatic expression of PPARα and found no detectable levels of FLag PPARα in the liver of MHC-PPARα mice (Fig. 4B). Also, changes in the expression of endogenous PPARα were not detected in the MHC-PPARα mice (data not shown). Interestingly, circulating levels of IL-6 were increased by more than fourfold in the MHC-PPARα mice compared with the wild-type mice (∼20 weeks of age) or both groups of mice at 8 weeks of age (Fig. 6D).

To determine the mechanism of hepatic insulin resistance, hyperinsulinemic-euglycemic clamps were conducted in the MHC-PPARα mice and wild-type littermates at 8 weeks of age. The metabolic parameters (i.e., glucose, insulin) were comparable in the 8-week and the 12- to 14-week-old mice. Insulin-stimulated cardiac glucose uptake was reduced by ∼60% in the 8-week-old MHC-PPARα mice compared with the age-matched wild-type mice (Fig. 7A), consistent with the notion that reduced glucose metabolism is secondary to PPARα-mediated alteration in lipid metabolism. Also, cardiac Akt activity was significantly reduced in the 8-week-old MHC-PPARα mice (1,541 ± 95 vs. 2,100 ± 167 arbitrary units in the age-matched wild-type mice, P < 0.05). In contrast, ventricular fractional shortening was not altered in the 8-week-old MHC-PPARα mice (Fig. 7B), suggesting that PPARα-induced metabolic changes in cardiomyocytes precede and are causally associated with altered cardiac function in the 12- to 14-week-old MHC-PPARα mice. Additionally, hepatic insulin action was normal in the 8-week-old MHC-PPARα mice (Fig. 7C), and this was associated with normal intrahepatic triglyceride content in these mice (Fig. 7D). These results are in contrast to the 12- to 14-week-old MHC-PPARα mice, which developed profound hepatic insulin resistance associated with elevated intrahepatic triglyceride content. Overall, these findings indicate that hepatic insulin resistance in the MHC-PPARα mice is age dependent and may be secondary to altered cardiac function in the MHC-PPARα mice.

Recent findings (3–5) indicate that the perturbation in cardiac energy metabolism and insulin resistance are among the earliest diabetes-induced events in the myocardium, preceding both functional and pathological changes. Studies (11,12) using isolated perfused heart preparations or cultured cardiomyocytes uniformly showed increased lipid oxidation in diabetic heart. Others (13,14) found markedly reduced rates of myocardial glucose uptake in type 2 diabetic subjects using PET combined with ¹⁸F-fluoro-deoxy-glucose. The importance of cardiac insulin action and glucose metabolism was further demonstrated in mouse models of cardiac-specific ablation of GLUT4 or deletion of insulin receptor that developed cardiac hypertrophy and other metabolic phenotypes resembling diabetic heart (27–30). Additionally, mice with cardiac-specific overexpression of Akt were shown to develop ventricular hypertrophy associated with altered insulin signaling and glucose metabolism in the cardiomyocytes (31,32). Taken together, these findings indicate the important role of cardiac insulin resistance in the pathogenesis of diabetic heart failure.

Although increased lipid oxidation and reduced glucose metabolism are characteristic features of diabetic heart (3–5), it is unclear whether these changes are causally related. Our findings that basal cardiac glucose metabolism is reduced in the MHC-PPARα mice suggest that increased lipid oxidation, at least in part due to increased activity of PPARα, is causally associated with altered glucose metabolism in diabetic heart. More importantly, our results indicate that increased lipid oxidation may also be causally associated with reduced insulin action in the diabetic heart, and the mechanism may involve defects in cardiac insulin signaling activity and reduced GLUT4 contents (10). Previous studies indicated that IRS-1 and -2 are important intracellular mediators of insulin signaling and insulin-mediated glucose metabolism in skeletal muscle and liver, respectively (i.e., glucose transport and glycogen synthesis) (33). Defects in insulin-stimulated tyrosine phosphorylation of IRS-1 and -2 as well as IRS-associated PI 3-kinase were shown to mediate insulin resistance in skeletal muscle and liver, respectively (33). Additionally, mice lacking Akt2 (protein kinase B-β) were shown to develop whole-body insulin resistance (25). Our findings that insulin-stimulated activities of IRS-1–associated PI 3-kinase and Akt were markedly reduced in the MHC-PPARα mice implicate that defects in cardiac insulin signaling, in addition to reduced GLUT4 content (10), were responsible for reduced cardiac glucose metabolism in the MHC-PPARα mice.

The mechanism by which increased cardiac lipid metabolism causes defects in insulin signaling may involve intramyocardial accumulation of fatty acid–derived metabolites (i.e., fatty acyl CoA, diacylglycerol, ceramide) (34,35). The underlying mechanism may involve activation of serine kinase cascade, of which protein kinase C-θ and/or IκB kinase-β may play a role, by fatty acid metabolites leading to the serine phosphorylation of IRS-1 (35–37). In this regard, serine phosphorylation of IRS-1 has been shown to reduce tyrosine phosphorylation of IRS-1 and interfere with its ability to recruit and activate PI 3-kinase, as occurs upon treatment with tumor necrosis factor-α (38). However, despite increases in lipid oxidation, uptake, and myocardial triglyceride content, intracellular fatty acyl CoAs and diacylglycerol levels were not elevated in the MHC-PPARα mice. Instead, intramyocardial levels of fatty acid metabolites showed a tendency to be reduced in the MHC-PPARα mice, and this is consistent with previous findings (39) of unaltered ceramide content in the heart of MHC-PPARα mice. These results suggest that while such mechanism plays a role in fat-induced insulin resistance in skeletal muscle, it may not play a major role in the cardiac insulin resistance of MHC-PPARα mice. Additionally, paradoxical changes in myocardial triglyceride content and intracellular fatty acid metabolites suggest differential effects of PPARα on lipid uptake/synthesis and mitochondrial lipid oxidation. In other words, overexpression of PPARα may have caused a greater increase in mitochondrial lipid oxidation than lipid uptake, resulting in elevated triglyceride levels but unaltered fatty acyl CoAs. Although our findings contrast to the previously observed increases in triacylglycerol-containing long-chain fatty acid species in the heart of MHC-PPARα mice (39), the discrepancy may be due to the measurement taken in mice at a different age and/or metabolic state (i.e., fed versus fasted state).

A recent study (26) showed that liver-specific STAT3 knockout mice developed hepatic insulin resistance, suggesting a potential role of STAT3 in hepatic glucose metabolism. In this regard, the role of STAT3 on glucose metabolism has been recently identified in different mouse models of STAT3 deletion that have commonly shown alteration in glucose homeostasis (40,41). Since insulin-mediated tyrosine phosphorylation of STAT3 was significantly reduced in the MHC-PPARα mice, our results suggest the potential role of STAT3 in cardiac glucose metabolism and further indicate that blunted STAT3 activity may be responsible for cardiac insulin resistance in the MHC-PPARα mice (Fig. 8). Consistent with this notion, mice with cardiomyocyte-restricted knockout of STAT3 were shown to be more susceptible to cardiomyopathy following doxorubicin treatment (42), which may be due to altered glucose metabolism in these mice. In this regard, Hrelia et al. (43) showed that doxorubicin-induced cardiomyopathy was associated with protective increases in glucose metabolism in cultured cardiomyocytes. Furthermore, recent studies (44–46) reported the antagonistic effects of and bidirectional regulation between STATs and PPARs and suggested a potential cross-talk between Janus kinase-STAT and PPAR pathways. Additionally, Gervois et al. (47) demonstrated a downregulation of STAT3 in liver following chronic treatment of the PPARα agonist, fenofibrate. Taken together, these findings indicate the important and novel role of STAT3 in cardiac glucose metabolism and that blunted STAT3 activity may partly mediate cardiac insulin resistance in the MHC-PPARα mice. Importantly, our findings also suggest that the reciprocal regulation between PPARα and STAT3 may be the mechanism by which increased cardiac lipid oxidation leads to insulin resistance in the MHC-PPARα mice (Fig. 8). Alternatively, cardiac insulin resistance may be due to prolonged state of high lipid oxidation in the heart of MHC-PPARα mice that leads to accumulation of other lipid intermediates, mitochondrial or peroxisomal generation of reactive oxygen species, or excessive oxygen consumption (48–50).

Surprisingly, MHC-PPARα mice developed defects in hepatic insulin action associated with blunted IRS-2–associated PI 3-kinase activity in liver. Hepatic insulin resistance may be due to increased intrahepatic triglyceride content in the MHC-PPARα mice (20). Since heart is a major organ for fatty acid clearance (51), increased cardiac lipid oxidation in the MHC-PPARα mice may alter hepatic lipid metabolism in order to meet the increasing demand of the transgenic heart (Fig. 8). This may be reflected by a tendency for reduced intrahepatic total fatty acyl CoAs and significantly reduced intrahepatic levels of C18:1 and C18:2 in the MHC-PPARα mice. To further determine the mechanism of hepatic insulin resistance in the MHC-PPARα mice, we assessed cardiac insulin action, function, and hepatic glucose/lipid metabolism in the 8-week-old mice. Insulin-stimulated cardiac glucose uptake was similarly reduced in the 8-week-old MHC-PPARα mice compared with the 12- to 14-week-old MHC-PPARα mice. Despite early changes in cardiac glucose metabolism, ventricular fractional shortening was not altered in the 8-week-old MHC-PPARα mice. These results indicate that cardiac insulin resistance and blunted glucose metabolism developed before the onset of cardiac dysfunction in the MHC-PPARα mice. Our findings further support the notion that altered cardiac energy metabolism may precede and cause changes in function of diabetic heart. Moreover, hepatic insulin action was normal in the 8-week-old MHC-PPARα mice, which was associated with normal hepatic triglyceride content. The observation of age dependence in the development of hepatic insulin resistance in the MHC-PPARα mice as well as overlapping pattern of blunted cardiac function and hepatic glucose metabolism further supports the causal relationship between heart and liver. In this regard, the scenario involving potential cross-talk between heart and liver may be mediated by circulating inflammatory cytokines, such as IL-6 (Fig. 8). Our findings that circulating IL-6 levels were not altered in the MHC-PPARα mice at 8 weeks of age but significantly elevated at ∼20 weeks of age, when cardiac dysfunction was evident, suggest the physiological role of IL-6 in the MHC-PPARα mice. In this regard, IL-6 and tumor necrosis factor-α have been shown to be secreted by cardiomyocytes and in failing heart (52,53). Altogether, more studies are clearly needed to examine the cross-talk relationship between the heart and liver and to determine the underlying mechanism.

In summary, heart-specific overexpression of PPARα caused early defects in cardiac glucose metabolism and insulin resistance that led to ventricular hypertrophy and blunted cardiac functions. Cardiac insulin resistance was associated with reduced STAT3, IRS-1–associated PI 3-kinase, and Akt activities, suggesting the possible role of STAT3 in modulating cardiac glucose metabolism. Furthermore, age-associated development of hepatic insulin resistance in the MHC-PPARα mice may be due to altered cardiac metabolism, function, and/or IL-6 level. Overall, our results demonstrate the important role of cardiac energy metabolism in the development of diabetic heart failure and provide novel insights into the potential cross-talk between heart and liver.

This study is supported by grants from the U.S. Public Health Service (U24 DK-59635 to J.K.K. and HL-04429 to K.S.R.), American Diabetes Association (7-01-JF-05 and 1-04-RA-47 to J.K.K. and 7-02-JF-26 to Y.-B.K.), The Robert Leet and Clara Guthrie Patterson Trust Award to J.K.K., and American Heart Association to K.S.R.

We are grateful to Dongyan Zhang, Anthony Romanneli, Aida Groszmann, and Jonas Lai for technical assistance.