|Home | About | Journals | Submit | Contact Us | Français|
We examined the role of butyric acid, a short-chain fatty acid formed by fermentation in the large intestine, in the regulation of insulin sensitivity in mice fed a high-fat diet.
In dietary-obese C57BL/6J mice, sodium butyrate was administrated through diet supplementation at 5% wt/wt in the high-fat diet. Insulin sensitivity was examined with insulin tolerance testing and homeostasis model assessment for insulin resistance. Energy metabolism was monitored in a metabolic chamber. Mitochondrial function was investigated in brown adipocytes and skeletal muscle in the mice.
On the high-fat diet, supplementation of butyrate prevented development of insulin resistance and obesity in C57BL/6 mice. Fasting blood glucose, fasting insulin, and insulin tolerance were all preserved in the treated mice. Body fat content was maintained at 10% without a reduction in food intake. Adaptive thermogenesis and fatty acid oxidation were enhanced. An increase in mitochondrial function and biogenesis was observed in skeletal muscle and brown fat. The type I fiber was enriched in skeletal muscle. Peroxisome proliferator–activated receptor-γ coactivator-1α expression was elevated at mRNA and protein levels. AMP kinase and p38 activities were elevated. In the obese mice, supplementation of butyrate led to an increase in insulin sensitivity and a reduction in adiposity.
Dietary supplementation of butyrate can prevent and treat diet-induced insulin resistance in mouse. The mechanism of butyrate action is related to promotion of energy expenditure and induction of mitochondria function.
Recent studies suggest that natural compounds represent a rich source for small thermogenic molecules, which hold potential in the prevention and treatment of obesity and insulin resistance. Several natural products, such as resveratrol (1,2), bile acid (3), and genipin (4), have been reported to increase thermogenic activities in animal or cellular models. In the current study, we provide evidence for the thermogenic activity and therapeutic value of a short-chain fatty acid, butyric acid, in a mouse model of metabolic syndrome. Butyric acid has four carbons in the molecule (CH3CH2CH2-COOH) and becomes sodium butyrate after receiving sodium. Sodium butyrate is a dietary component found in foods such as cheese and butter. It is also produced in large amounts from dietary fiber after fermentation in the large intestine, where butyric acid is generated together with other short-chain fatty acids from nondigestible carbohydrates, such as nonstarch polysaccharides, resistant starch, and miscellaneous low-digestible saccharides (5,6). The bioactivities of sodium butyrate are related to inhibition of class I and class II histone deacetylases (7). Histone deacetylases regulate gene transcription through modification of chromatin structure by deacetylation of proteins, including histone proteins and transcription factors. To our knowledge, there is no report about butyrate in the regulation of insulin sensitivity or energy metabolism.
Dietary intervention is a potential strategy in the prevention and treatment of metabolic syndrome. Peroxisome proliferator–activated receptor (PPAR)-γ coactivator (PGC)-1α, a transcription coactivator, is a promising molecular target in dietary intervention (1,2). PGC-1α controls energy metabolism by interaction with several transcription factors, e.g., estrogen-related receptor-α, nuclear respiratory factor-1 and -2, PPAR-α and -δ, and thyroid hormone receptor, that direct gene transcription for mitochondrial biogenesis and respiration (8). In the muscle, PGC-1α increases oxidative (type I) fiber differentiation and enhances fatty acid metabolism (9). In brown fat, PGC-1α stimulates adaptive thermogenesis through upregulation of uncoupling protein (UCP)-1 expression (10). A reduction in PGC-1α function is associated with mitochondrial dysfunction, reduction in fatty acid oxidation, and risk for insulin resistance or type 2 diabetes (11–14). Dietary intervention of PGC-1α activity holds promise in the prevention and treatment of metabolic syndrome. However, our knowledge is limited in the dietary components or derivatives that are able to regulate the PGC-1 activity. In the current report, we provide evidence that sodium butyrate induced PGC-1 activity in skeletal muscle and brown fat in mice.
Male C57BL/6J (4 weeks old) mice were purchased from the Jackson Laboratory (Bar Harbor, ME). After 1 week quarantine, the C57BL/6J mice were fed a high-fat diet (D12331; Research Diets, New Brunswick, NJ), in which 58% of calories is from fat. All of the mice were housed in the animal facility with a 12-h light/dark cycle and constant temperature (22–24°C). The mice had free access to water and diet. All procedures were performed in accordance with National Institutes of Health guidelines for the care and use of animals and were approved by the institutional animal care and use committee at the Pennington Biomedical Research Center.
Sodium butyrate (303410; Sigma) was incorporated into the high-fat diet at 5% wt/wt. Sodium butyrate was blended into the diet using a food processor at 400 rpm. The sodium butyrate–containing diet was pelleted and stored in a –20°C freezer until usage. On the supplemented diet, mice could receive sodium butyrate at 5 g · kg−1 · day−1 at the normal daily rate of calorie intake.
Intraperitoneal insulin tolerance testing was conducted by intraperitoneal injection of insulin (I9278; Sigma) at 0.75 units/kg body wt in mice after a 4-h fast, as described previously (15).
Body composition was measured for the fat content using quantitative nuclear magnetic resonance as previously described (15).
Total RNA was extracted from frozen tissues (kept at −80°C) using Tri-Reagent (T9424; Sigma), as described previously (16). TaqMan RT-PCR primer and probe were used to determine mRNA for PGC-1α (Mm00447183_m1), UCP-1 (Mm00494069_m1), PPAR-δ (Mm01305434_m1), and carnitine palmitoyltransferase-1b (CPT1b; Mm00487200_m1). The reagents were purchased from Applied Biosystems (Foster City, CA). Mouse ribosomal 18S rRNA_s1 (without intron-exon junction) was used as an internal control. The reaction was conducted with a 7900 HT Fast real-time PCR system (Applied Biosystems).
Energy expenditure, respiratory exchange ratio, spontaneous physical movement, and food intake were measured simultaneously for each mouse with a comprehensive laboratory animal monitoring system (Columbus Instruments, Columbus, OH) as described previously (15).
Body temperature was measured in the cold room with ambient temperature at 4°C. Animals were sedated and restrained for <30 s during the measurement. A Thermalert model TH-8 temperature monitor (Physitemp, Clifton, NJ) was used with the probe placed in the rectum at 2.5 cm in depth.
Fresh fat and muscles were collected and frozen immediately in liquid nitrogen. The whole-cell lysate was prepared in a lysis buffer with sonication as described previously (16). Antibodies and their sources are myoglobin (sc-25607; Santa Cruz Biotechnology), phosphorylated Akt (threonine 308 [Thr308], no. 9275; Cell Signaling), tubulin (ab7291; Abcam), phosphorylated insulin receptor substrate-1 (tyrosine 632 [Tyr632], no. sc-17196; Santa Cruz Biotechnology), myosin (M8421; Sigma), phosphorylated AMP kinase (AMPK; Thr172, no. 2531; Cell Signaling), and pP38 (sc-7975; Santa Cruz Biotechnology). The antibodies to PGC-1α and UCP-1 were from Dr. Thomas Gettys at our institute (Pennington Biomedical Research Center).
The fiber types in skeletal muscle were examined using two methods: succinate dehydrogenase staining for ATPase and immunostaining of type I myosin heavy chain. In the succinate dehydrogenase staining, midbelly cross-sections of muscle were cut at 8 μm in a cryostat (–20°C). After drying for 5 min at room temperature, the sections were incubated at 37°C for 60 min in incubation solution containing 6.5 mmol/l sodium phosphate monobasic, 43.5 mmol/l sodium phosphate biphasic, 0.6 mmol/l nitroblue tetrazolium (74032; Sigma), and 50 mmol/l sodium succinate (14160; Sigma). The sections were rinsed three times (30 s each time) in 0.9% saline, 5 min in 15% ethanol, and then mounted with aqueous mounting medium (DakoCytomation).
Fresh skeletal muscle was collected, embedded in gum tragacanth mixed with OCT freezing matrix, and quickly frozen in liquid nitrogen. The tissue slides were obtained through serial cross-section cutting at 8 μm thickness. The slide was blotted with type I myosin heavy-chain antibody (M8421; Sigma) at 1:200 dilution. After being washed, the slide was incubated with a biotinylated secondary antibody (BA-2000), and the color reaction was performed using ABC elite reagent (PK-6101). A 3,3′-diaminobenzidine substrate kit (SK-4100) was used to obtain signal for myosin I.
Fresh tissues (brown adipose tissue [BAT]) were collected at 16 weeks of age after 12 weeks on the sodium butyrate–containing diet. Tissue was fixed in 10% formalin solution (HT50–1-2; Sigma). Tissue slides were obtained through serial cross-section cutting 8 μm in thickness and processed with standard procedure.
Histone deacetylase assay was conducted using a histone deacetylase assay kit (17-320; Upstate). Briefly, the muscle nuclear extract (10 μg) was incubated with [3H]acetyl CoA (TRK688; Amersham) radio-labeled histone H4 peptide (25,000 cpm, as a substrate) at 37°C for 12 h by shaking. Released [3H]acetate was measured using a scintillation counter. The nuclear extract was prepared according to a protocol described previously (17).
The serum fatty acids, including butyrate, were examined using a protocol described elsewhere (18). The detail is presented in Supplement 1, available in an online appendix at http://diabetes.diabetesjournals.org/cgi/content/full/db08-1637/DC1. Fatty acids in feces were determined using the protocol from a study by Schwarz et al. (19). Triglyceride and cholesterol were measured in whole blood with a Cardiochek portable test system.
Data are the means ± SE from multiple samples. All of the in vitro experiments were conducted at least three times. Student's t test or two-way ANOVA was used in the statistical analysis with significance set at P ≤ 0.05.
We first tested butyrate in prevention of dietary obesity. In the diet-induced obesity model, butyrate supplementation started at the beginning of high-fat diet feeding. Plain high-fat diet was used in the control group. Calorie intake was monitored four times in the first 10 weeks. After normalization with body weight, calorie intake was reduced with the increase in age. In the butyrate group, it was significantly higher at all of the time points (Fig. 1A). Energy expenditure, oxygen consumption, and substrate utilization were monitored using a metabolic chamber. In the butyrate group, energy expenditure and oxygen consumption were elevated at night (Fig. 1B and C). The respiratory exchange ratio was reduced during the day and night (Fig. 1D), suggesting an increase in fatty acid oxidation in response to butyrate. These data suggest that sodium butyrate may increase energy expenditure in the diet-induced obesity model.
Body weight and fat content were monitored in this study. In the control mice, body weight increased from 23 to 40 g after 16 weeks on high-fat diet (Fig. 1E), and fat content (adiposity) increased from 10 to 35% of body weight (Fig. 1F). Accordingly, lean mass was reduced from 80 to 65% (Fig. 1G). In the butyrate group, these parameters were not significantly changed during the 16 weeks on high-fat diet (Fig. 1E, F, and G), suggesting that butyrate prevented diet-induced obesity. Mouse growth was not influenced by butyrate because body length was identical between the two groups. Dietary fat digestion and absorption in the gastrointestinal tract was examined by measuring fatty acid content in feces. Fat content was identical in the feces of the two groups (Fig. 1H), suggesting that butyrate does not influence fat absorption by the gastrointestinal tract. Spontaneous physical activity was monitored during the day and night in the mice. The data suggest that physical activity was not reduced by butyrate (Fig. 1I). Increased activity was observed in the butyrate group at night. These data suggest that dietary supplementation with butyrate protected the mice from diet-induced obesity. This effect is associated with an increase in energy expenditure and fatty acid oxidation. The food intake and physical activity suggests that no toxicity was observed for butyrate in the mice.
The increase in energy metabolism suggests that butyrate may protect mice from high-fat diet–induced insulin resistance. To test this possibility, systemic insulin sensitivity was analyzed by fasting glucose, fasting insulin, and insulin tolerance. In the control group, fasting glucose was increased significantly after 10 weeks on high-fat diet (Fig. 2A). In the butyrate group, this increase was not observed (Fig. 2A). Fasting insulin was 50% lower in the butyrate group at 16 weeks on high-fat diet (Fig. 2B). In the intraperitoneal insulin tolerance test, the butyrate group exhibited much better response to insulin at all time points (30, 60, 120, and 180 min) (Fig. 2C). Homeostasis model assessment for insulin resistance (HOMA-IR) was 60% lower in the butyrate group (Fig. 2D). These data suggest that insulin resistance was prevented in the butyrate group. Insulin signaling was examined in the skeletal muscle lysate with Tyr632 (Y632) phosphorylation of insulin receptor substrate-1 protein and Thr308 phosphorylation of Akt (Fig. 2E). Both signals were increased in the butyrate-treated mice (Fig. 2E and F), suggesting a molecular mechanism of insulin sensitization.
The association of increased food intake with elevated energy expenditure led us to study BAT, which is responsible for adaptive thermogenesis in response to diet or cold (20–22). Diet-induced thermogenesis reduces obesity in both humans and animals (23). In the butyrate group, the increase in energy expenditure was observed at night when mice actively took food (Fig. 1A and B). This result suggests an increase in thermogenesis. To determine the thermogenic function, we conducted a cold-response experiment. Mice were exposed to a cold environment with an ambient temperature of 4°C for 90 min. Core body temperature was monitored three times by measuring the rectal temperature. In control mice, body temperature decreased with time and was 34.5°C after 90 min in the cold (Fig. 3A). In butyrate-treated mice, body temperature dropped to 35°C transiently at 30 min and then returned to 36°C for the remainder of the time. These data suggest that thermogenic function is enhanced in the butyrate group.
Brown fat is critical for adaptive thermogenesis in mice. Morphology and gene expression were examined in brown fat. Compared with the control mice, the size of brown adipocytes was much smaller in the butyrate group (Fig. 3B), suggesting higher thermogenic activity that leads to the reduction in fat accumulation. Mitochondrial function is regulated by gene expression (24). To understand the molecular basis of the increased thermogenesis, we examined the expression of two thermogenesis-related genes, PGC-1α and UCP-1, in BAT. mRNA of both genes was increased in the butyrate-treated mice (Fig. 3C). The increase was observed in their proteins in the brown fat (Fig. 3D). The increased gene expression provides a molecular basis for enhanced thermogenesis by butyrate treatment.
To understand the cellular basis of enhanced fatty acid utilization in the butyrate group, we assessed muscle fiber types. PGC-1α was reported to induce transformation of skeletal muscle fiber from glycolytic type (type II) into oxidative type (type I) in transgenic mice (9). Type I fibers are distinct from type II fibers in several properties (25). Type I fibers (oxidative and slow-twitch fibers) are rich in mitochondria, red in color, and active in fat oxidation for ATP biosynthesis. Type II fibers (glycolytic and fast-twitch fibers) are relatively poor in mitochondrial activity, lighter in color, and dependent on glycolysis in ATP production. The butyrate effect on PGC-1α in BAT suggests that skeletal muscle fibers may be changed by butyrate.
Compared with the control group, the butyrate group exhibited a deep red color (Fig. 4A). Fiber type analysis was conducted in the vastus lateralis, gastrocnemius (rich in glycolytic fibers), and soleus (rich in oxidative fiber). Type I fiber was determined with type I myosin heavy-chain immunohistostaining. The ratio of type I fibers were increased in all of the skeletal muscles of butyrate-treated mice (Fig. 4B). The increase was confirmed with a metachromatic dye–ATPase assay, in which the type I fibers were blue in color (Fig. 4C). To support the change in muscle morphology, the proteins of type 1 myosin heavy chain and PGC-1α were quantified in muscle lysate in an immunoblot. A significant increase was observed in both proteins in butyrate-treated mice (Fig. 4D). Myoglobin (another marker of oxidative type I fiber) was also increased by butyrate (Fig. 4D). A mean value of each protein is presented in Fig. 4D. These data suggest that the ratio of type I fiber was increased by butyrate in skeletal muscle.
AMPK and p38 activities were examined by their phosphorylation status. Their activities may contribute to elevation of PGC-1α protein through enhanced protein stability (26–28). It was not clear whether this mechanism was activated by butyrate. To test this possibility, we examined activity of AMPK and p38 in skeletal muscle. An increase in their phosphorylation was observed in muscle lysate of butyrate-treated mice (Fig. 4D), suggesting increased activation of the two kinases by butyrate. In the L6 cell line, AMPK and p38 phosphorylation was increased by butyrate in the cell culture (Fig. 5A), suggesting butyrate is able to activate AMPK directly. In the same culture, PGC-1α protein was increased (Fig. 5A). In the liver of butyrate-treated mice, a similar pattern of changes was observed in AMPK, p38, and PGC-1α (Fig. 5B). These data consistently suggest that AMPK and p38 were activated by butyrate and that their activation may contribute to the increase in PGC-1α activity.
Mitochondrial function was examined in skeletal muscle tissue and L6 muscle cells under butyrate treatment. Fatty acid oxidation was monitored in gastrocnemius muscle with 14C-labeled palmitic acid. A 200% increase in 14C-labeled CO2 was observed in butyrate-treated mice (Fig. 6A). Fatty acid oxidation was associated with expression of PGC-1α target genes, such as CPT1b and COX-I (cytochrome c oxidase I) (9). Expression of these two genes was increased in skeletal muscle of butyrate-treated mice (Fig. 6B and C). The nuclear receptor PPAR-δ promotes fatty acid oxidation in skeletal muscle (29). PPAR-δ expression was also increased in butyrate-treated mice (Fig. 6B). In cultured L6 cells, a similar increase was observed in fatty acid oxidation and gene expression after butyrate treatment (Fig. 6D and E). These data consistently support the role of butyrate activity in the promotion of mitochondrial function.
The butyrate concentration was analyzed in serum collected from the butyrate and control groups. In the fasted condition (overnight fast), the butyrate concentration was 7.23 ± 0.93 μg/ml in the butyrate group and 5.71 ± 0.38 μg/ml in the control animals. In the fed condition, the butyrate concentration was 9.40 ± 1.36 μg/ml in the butyrate group versus 5.48 ± 0.60 μg/ml in the control mice (P < 0.05, n = 5) (Fig. 6F). The data suggest that dietary supplementation increased butyrate levels in the blood. It is likely that the metabolic activity of butyrate is related to inhibition of histone deacetylase. Sodium butyrate inhibits the class I and class II histone deacetylases. To test this possibility, histone deacetylase activity was examined in skeletal muscle of mice at 16 weeks on high-fat diet (Fig. 6F). The assay was conducted with nuclear extracts of muscle samples. Histone deacetylase activity was reduced by 50% in the butyrate group (Fig. 6G). Trichostatin A (TSA), a typical histone deacetylase inhibitor, was used as a positive control in the parallel treatment. Histone deacetylase activity was decreased in the skeletal muscle of TSA-treated mice (Supplement 2). These data suggested that dietary supplementation of butyrate leads to suppression of histone deacetylase activity in the body. Total triglyceride and cholesterol were examined in the blood. These lipids were reduced in the butyrate group (Fig. 6H and I).
In the prevention studies, butyrate was administrated together with high-fat diet during the induction of obesity. To test butyrate in the treatment of obesity and insulin resistance, we administrated butyrate to obese mice that had been on a high-fat diet for 16 weeks. After a 5-week treatment with butyrate, the obese mice lost 10.2% of their original body weight, which dropped from 37.6 to 34.4 g (Fig. 7A). In the control group, body weight increased by 15.8% (from 35.9 to 41.6 g) during the same time period. Consistent with the change in body weight, fat content was reduced by 10% in the butyrate group (Fig. 7B). Furthermore, fasting glucose was reduced by 30% from 131 to 98.6 mg/dl (P < 0.016), HOMA-IR was reduced by 50%, and intraperitoneal insulin tolerance was improved significantly in the butyrate group (Fig. 7C and D). These data suggest that butyrate is effective in the treatment of obesity and insulin resistance in the dietary obese model.
Metabolic activities of butyric acid were examined in this study in diet-induced obese mice. The most important observation is that butyrate supplementation at 5% wt/wt in high-fat diet prevented development of dietary obesity and insulin resistance. It also reduced obesity and insulin resistance in obese mice. In butyrate-treated mice, the plasma butyrate concentration was increased, and blood lipids (triglycerides, cholesterol, and total fatty acids) were decreased (Fig. 6H–I and Supplement 1). The change in insulin sensitivity may be a consequence of a reduction in adiposity in our model. The increase in energy expenditure and fatty acid oxidation may be responsible for the antiobesity effect of butyrate. Butyrate supplementation did not reduce food intake, fat absorption, or locomotor activity, suggesting that there was no toxicity from butyrate. Butyrate was tested at 5 and 2.5% wt/wt in the high-fat diet in this study. At the lower (2.5% wt/wt) dosage, similar metabolic activity was observed (Supplement 3). At 5% in the high-fat diet, butyrate increased the calorie content from 58 to 64.4% in the fat. The increase in fat calories may not contribute to our observation of the antiobesity activity for butyrate. A recent study of weight-loss diets suggests that total calorie intake, not diet composition, is responsible for weight reduction in humans (30). At the cellular level, butyrate increased mitochondrial respiration, as indicated by the increase in oxygen consumption and CO2 production. At the molecular level, increased expression of PGC-1α, PPAR-δ, and CPT1b may be involved in the stimulation of mitochondrial function by butyrate.
The current study indicates that in vivo butyrate is a novel activator of PGC-1α. PGC-1α activity may be regulated by butyrate at three levels. PGC-1α expression was increased in both mRNA and protein. The protein elevation was observed in brown fat, skeletal muscle, and liver in butyrate-treated mice. It may be a result of increased mRNA expression or extended half-life of the PGC-1α protein. The change in protein stability is supported by the activities of AMPK and p38 in tissues and cells after butyrate treatment. These kinases phosphorylate the PGC-1α protein and inhibit its degradation (27,28,31–34). As a transcriptional coactivator, PGC-1α transcription activity may be induced by phosphorylation, which leads to removal of a suppressor protein (p160 myb) that is associated with PGC-1α in the basal condition (35). P38 acts downstream of AMPK in the phosphorylation of PGC-1α (36). Therefore, AMPK may increase PGC-1α phosphorylation through direct and indirect (p38) mechanisms. It is not clear how AMPK is activated by butyrate. Butyrate may act through induction of AMP levels in cells from increased consumption of ATP. It was reported that butyrate increases ATP consumption (37). Induction of PGC-1α activity may be a molecular mechanism by which butyrate stimulates mitochondrial function.
Inhibition of histone deacetylase may contribute to increased mRNA expression of PGC-1α, PPAR-δ, and CPT1b. Histone deacetylase inhibition promotes gene expression through transcriptional activation, which is determined by gene promoter activity. Promoter activation requires histone acetylation, which opens chromatin DNA to the general transcription factors for transcription initiation and mRNA elongation. Histone deacetylase inhibits gene promoter activity through deacetylation of histone proteins. In the presence of butyrate, promoter inhibition is prevented by butyrate suppression of histone deacetylase. Histone deacetylase suppression will enhance histone acetylation. This chromatin modification may occur in the gene promoters for PGC-1α, PPAR-δ, and CPT1b for the upregulation of gene transcription.
Butyrate induces type I fiber differentiation in skeletal muscle. In skeletal muscle cells, inhibition of histone deacetylase enhances myotube differentiation in vitro (28–30) and protects muscle from dystrophy in vivo (29–31). In transgenic mice, knockout of class II histone deacetylases was shown to promote differentiation of type I (oxidative) fibers in skeletal muscle (32). This is consistent with our data that type I fiber was increased by butyrate, which inhibits histone deacetylase activities in the skeletal muscle of butyrate-treated mice. TSA, a typical histone deacetylase inhibitor, was tested in parallel treatment with butyrate. TSA exhibited activity similar to that of butyrate in mice (Supplement 2). TSA prevented dietary obesity, insulin resistance, and increased the type I fiber in the skeletal muscle. The activity was associated with elevation of PGC-1α protein. The current study suggests that the metabolic activities of butyrate may be dependent on the inhibition of histone deacetylase.
In summary, dietary supplementation of butyrate can prevent and treat diet-induced obesity and insulin resistance in mouse models of obesity. These activities of butyrate are similar to those of resveratrol (1,2). The mechanism of butyrate action is related to promotion of energy expenditure and induction of mitochondrial function. Stimulation of PGC-1α activity may be a molecular mechanism of butyrate activity. Activation of AMPK and inhibition of histone deacetylases may contribute to the PGC-1α regulation. Butyrate and its derivatives may have potential application in the prevention and treatment of metabolic syndrome in humans.
This study was supported by National Institutes of Health grants DK68036 and P50AT02776-020002 (to J.Y.), American Diabetes Association (ADA) Research Award 7-07-RA-189 (to J.Y.), a pilot grant from the Functional Food Division at the Pennington Biomedical Research Center, ADA Junior Faculty Award 1-09-JF-17 (to Z.G.), and CNRU Grant 1P30 DK072476 sponsored by the NIDDK.
No potential conflicts of interest relevant to this article were reported.
We thank Zhong Wang, Can Pang, Jong-Seop Rim, Qing He, Ms. Xin Ye, and Wei Tseng for their excellent technical support. We thank Thomas Gettys for the antibodies to PGC-1α and UCP-1.
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.