Sterol regulatory element binding protein-1c (SREBP-1c), a transcription factor that is important for mediating insulin effects on metabolic gene expression in liver during the fasted-to-fed transition, is also expressed in skeletal muscle. However, the regulation and role of SREBP-1c in skeletal muscle are poorly understood. The present study compared the effects of nutritional status, physiological hyperinsulinemic clamps, and adenovirus-mediated hyperleptinemia (HLEP) in rats on expression of SREBP-1c and other metabolic genes in skeletal muscle. Three- and 6-h refeeding of 18-h-fasted animals increased levels of SREBP-1c mRNA and the SREBP-1 protein (full length and mature) in gastrocnemius muscle (P < 0.05). Fatty acid synthase (FAS) and hexokinase II (HKII) mRNA levels were also increased by refeeding, and uncoupling protein 3 (UCP3) mRNA level was decreased (all P < 0.05). Surprisingly, 3-h hyperinsulinemic clamps did not increase gastrocnemius muscle SREBP-1c and FAS mRNA levels or SREBP-1 protein levels but did increase HKII mRNA levels and decrease UCP3 mRNA levels (P < 0.05). HLEP reduced refeeding-induced increases of SREBP-1c and FAS mRNA levels but did not reduce the level of SREBP-1 protein. We conclude that 1) skeletal muscle SREBP-1c gene expression is regulated by nutritional status in a fashion similar to that observed in liver and adipose tissue, 2) physiological hyperinsulinemia is not sufficient to imitate the effects of refeeding on SREBP-1c gene expression, and 3) leptin suppresses refeeding effects on SREBP-1c mRNA levels.
- nutrient partitioning
- gene expression
recent work has identified sterol regulatory element binding protein-1c (SREBP-1c) as an insulin-activated transcription factor that regulates the expression of a number of metabolic genes involved in the partitioning of nutrients into lipid storage (9, 10, 17, 19, 43, 46). SREBP-1c is strongly expressed in tissues of high lipogenic capacity, such as liver and adipose tissue. Indeed, most of what is known regarding the regulation and function of SREBP-1c has come from studies in these tissues. However, recent studies (8, 11, 42) have demonstrated that SREBP-1c is also expressed in skeletal muscle. Unlike adipose tissue and liver, skeletal muscle is not regarded as being highly lipogenic. Nonetheless, muscle triglyceride stores are substantial and vary under a number of metabolic conditions (2, 26, 27, 34, 47, 48). Skeletal muscle is also a major site for fatty acid oxidation (39), and dysregulated muscle lipid metabolism and elevated muscle lipid storage (lipotoxicity) have been implicated in the pathogenesis of insulin resistance in diabetes and obesity. Thus it is clear that an understanding of the regulation and role of SREBP-1c in skeletal muscle lipid metabolism is required.
Guillat-Deniau and colleagues (11) demonstrated that the levels of SREBP-1c mRNA in skeletal muscle are fiber-type dependent, and levels in gluteus muscle (fast-twitch glycolytic) are comparable to those observed in liver. The same group demonstrated in rat primary myotubes that adenovirus-mediated overexpression of constitutively active SREBP-1c increases the levels of fatty acid synthase (FAS) mRNA and citrate lyase mRNA, two genes that are SREBP-1c responsive in liver (11). Also similar to liver and adipose tissue, insulin increases SREBP-1c mRNA levels in rat primary myotubes (11) and in human skeletal muscle (8, 42). Finally, exercise training increases SREBP-1c mRNA levels in mouse skeletal muscle, implicating a role for SREBP-1c in exercise training-induced increases in triglyceride synthesis (15). Taken together, these data suggest an important role for SREBP-1c in the regulation of lipid metabolism in skeletal muscle. However, despite these studies, the role of SREBP-1c in skeletal muscle in the in vivo setting remains largely unknown. Likewise, the in vivo regulation of SREBP-1c expression in skeletal muscle has been minimally explored.
The primary goal of the present study was to further investigate the in vivo regulation of SREBP-1c gene expression by comparing the effects of nutritional status, physiological hyperinsulinemia and hyperglycemia, and hyperleptinemia on SREBP-1c mRNA and protein levels in skeletal muscle. We demonstrate that similar to liver, refeeding increases SREBP-1c gene expression in skeletal muscle. Furthermore, we demonstrate that physiological increases in insulin and glucose concentrations are not sufficient to imitate the effects of refeeding on SREBP-1c gene expression. Finally, we demonstrate that hyperleptinemia suppresses refeeding-induced increases in SREBP-1c mRNA levels.
Male Wistar rats (Charles Rivers, Madison, WI) were treated according to Institutional Animal Care and Use Committee guidelines at the University of Pittsburgh Medical Center. Rats were exposed to a 12:12-h light-dark cycle and were allowed ad libitum access to food and water. For the fasting-refeeding studies and the hyperinsulinemic clamp studies, food was removed 18–24 h before the study. For hyperleptinemia studies, feeding regimens are described below.
Rats were either 18-h fasted or were 18-h fasted and then allowed to refeed for either 3 or 6 h. Food intake was recorded for the refeeding period. At the end of the refeeding period, nembutal sodium (60 mg/kg ip) was administered. Once corneal and pedal responses were absent, tissues were harvested in the following order: blood, liver, gastrocnemius muscle, and epididymal fat. Blood (cardiac puncture) was dispensed into tubes containing EDTA (1.75 mM final concentration) and was centrifuged for 1 min at room temperature. Resulting plasma was placed on ice for ∼30 min before storage at −80°C. Tissues were placed immediately into liquid nitrogen on removal and were stored at −80°C.
Catheters were placed into the carotid artery (advanced to the aortic arch) and jugular vein as described previously (37). Rats were given 4–7 days to recover and were studied only if body weight exceeded 90% of presurgery body weight. The day before study, food was removed so that rats entered clamps 18 to 24 h fasted. Approximately 45 min before beginning clamps, a baseline blood sample was taken. Clamps commenced once glucose values were below 120 mg/dl. A venous infusion of insulin (Humulin, Eli Lilly, Indianapolis, IN) at 15 mU·kg−1·min−1 was then begun in conjunction with a variable glucose (30%) infusion to maintain glucose concentrations (Beckman Glucose II analyzer, Fuller, CA) at either euglycemia (∼110 mg/dl, n = 7, Eug) or hyperglycemia (∼210 mg/dl, n = 6, Hyp). This rate of insulin infusion was used to achieve a circulating insulin concentration that was within the high physiological range but not supraphysiological (see Tables 2 and 3) to allow comparison with refeeding studies. To achieve supraphysiological plasma insulin concentrations, insulin was infused at 30 mU·kg−1·min−1 (n = 5; exogenous glucose, 50%), and glucose concentrations were maintained at ∼110 mg/dl. Arterial blood samples (∼30 μl) were taken every 8–10 min throughout the clamp, with larger samples (∼100 μl) taken at 60, 120, and 180 min. At the end of the clamp period, nembutal sodium (60 mg/kg) was administered through the arterial line, and removal and storage of tissues were handled in a manner similar to that described for the fasting-refeeding studies. Total blood volume taken throughout the clamp period did not exceed 10% of total blood volume (approximated from body weight). Control rats received a venous saline infusion at a rate that provided over 3 h a total volume approximately equal to that administered with the euglycemic or hyperglycemic clamps or supraphysiological insulin clamps.
Four days before study, rats received a recombinant adenovirus containing the leptin cDNA (HLEP, n = 5) or a control recombinant adenovirus containing the Esherichia coli β-galactosidase cDNA (βGal, n = 4) as described previously (6, 13). βGal rats were calorically matched to HLEP rats for the next 72 h because hyperleptinemia decreases food intake in rodents (6). The day before study, all rats were fasted 18 h and were then allowed access to food for 6 h. Harvesting of tissues at the end of the 6-h refeeding period was done as described for the fasting-refeeding studies. In addition to liver and gastrocnemius muscle (mixed fiber type), superficial vastus lateralis (predominantly type II fibers) and soleus (predominantly type I fibers) muscles were harvested.
Hormones and metabolites.
Insulin and leptin concentrations were determined by radioimmunoassay (Linco Research, St. Louis, MO), and glucose (fed/fasted studies) and free fatty acid concentrations were determined spectrophotometrically (Sigma Kit 510-A, St. Louis, MO and Waco NEFA-C kit, Richmond, VA, respectively). For clamp studies, glucose concentrations were determined by the glucose oxidase method using a Beckman Glucose Analyzer II.
RNA isolation and Northern blots.
Total RNA from 50–200 mg tissue was isolated using Trizol Reagent (Invitrogen/BRL-GIBCO, Carlsbad, CA) as described by the manufacturer. RNA (15 μg) was denatured in formaldehyde/formamide (1:3 vol/vol), was size-fractionated in formaldehyde-containing agarose gels (1.2%) containing ethidium bromide (1.25 ng/ml), and was then transferred to a Hybond-N nylon membrane (Amersham, Piscataway, NJ). After crosslinking and hybridization with an [α-32P]dCTP-labeled cDNA probe at 42°C overnight, membranes were exposed to autoradiographic film (Kodak X-OMAT AR film) at −80°C overnight. The cDNA probe containing the sequence corresponding to the rat SREBP-1 mRNA sequence (+20 to +303) was prepared by PCR cloning and the amplified cDNA was labeled using the Random Primed DNA Labeling Kit (Roche Diagnostics, Indianapolis, IN) with [α-32P]dCTP (6,000 Ci/mmol). Because SREBP-1c expression is 8- to 30-fold greater than SREBP-1a expression in skeletal muscle (11), we assumed that changes in SREBP-1 mRNA levels were reflective of changes in SREBP-1c mRNA levels. To confirm this, data obtained using Northern blots were checked by RT-PCR using SREBP-1c- and SREBP-1a-specific oligonucleotides. Northern blots were scanned and quantified using an image processor program (NIH Image J, v1.62). Ethidium bromide-stained 28S RNA was used as a loading control.
Total RNA (1 μg) was annealed to oligo(dT)18 primer (1 μM) at 70°C for 2 min. After quenching on ice for 5 min, the product was incubated in reaction buffer (40 mM Tris·HCl, 75 mM KCl, and 3 mM MgCl2) containing murine leukemia virus reverse transcriptase (10 U), recombinant RNase inhibitor (1 U), and 0.5 mM each of dCTP, dGTP, dATP, and dTTP for 1 h at 42°C. Diethyl pyrocarbonate-treated water was used to bring the final volume to 100 μl after heating at 94°C for 5 min. The PCR was carried out using 8 μl of the RT reaction in 4 mM KCl, 0.8 mM Tris·HCl, and 0.008% Triton X-100, pH 9.0 (Promega, Madison, WI), with 1 mM MgCl2, 0.4 mM dNTP, 1.25 U Taq polymerase (Promega, Madison, WI), to which ∼100 nmol (2 mM final concentration) of each of the primer pairs (IDT Technologies, Coralville, IA, Table 1) was added in a total volume of 50 μl. Primers for each gene were selected using either the PRIMER3 program or from work cited by others (Table 1). PCR products were designed to be approximately 200–600 bp with primers having an annealing temperature of ∼60°C. The PCR reaction was as follows: 5 min at 94°C, 22–30 cycles of 30 s at 94°C, 60 s at 60°C, and 120 s at 72°C. PCR products were separated on a 1.5% agarose gel containing ethidium bromide (1.25 ng/ml) in TAE buffer (40 mM Trizma, 20 mM acetic acid, 16.4 mM EDTA, pH = 8.5). The abundance of the PCR products was quantified using optical densiotometry (NIH Image J, v1.62). All samples were run in triplicate. Before use in experiments the number of cycles used for each primer pair was determined to be in the linear range for product formation.
Tissues (50–200 mg) were homogenized on ice in lysis buffer containing 20 mM Tris, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM pyrophosphate, 1 mM β-glycerophosphate, 1 mM Na3VO4, 1 μg/ml leupeptin, 3 μM aprotinin, 25 μg/ml calpain inhibitor I, and 0.1 mM phenylmethylsulfonyl fluoride, pH 7.5. Protein concentration was determined using the bicinchoninic acid method (Pierce, Rockford, IL) after centrifugation at 14,000 rpm for 10 min at 4°C. Proteins were denatured in 6% SDS (0.19 M Tris, 30% glycerol, 150 mM DTT, 0.03% bromophenol blue, pH 6.8) by boiling for 5 min. Equal amounts (10 μg for liver and 15 μg for gastrocnemius) were resolved by SDS-PAGE (10% Tris·HCl, Bio-Rad, Hercules, CA) at 180 V. Proteins were then electrotransferred (260 mA, 2 h) to Immobilon-P transfer membranes (PVDF, Millipore, Billerica, MA). After blocking (2 h at room temperature in 5% nonfat dry milk in lysis buffer), membranes were incubated overnight at 4°C with an antibody against SREBP-1 (Santa Cruz Biotechnology, Santa Cruz, CA) at 0.2 μg IgG1/ml (1:1,000) in 1% nonfat dry milk in lysis buffer. After incubation with horseradish peroxidase-linked goat anti-mouse IgG antibody (200 μg/ml; 1:2,000; Santa Cruz Biotechnology), membrane-bound antibodies were visualized by luminol chemiluminescence (LumiGLO, Cell Signaling Technology, Beverly, MA). Band intensity was determined by optical density (NIH Image J, v 1.62).
All results are expressed as means ± SE. For RT-PCR data, one-way ANOVA was used to determine differences among fasted vs. 6-h refed vs. 3-h refed and between saline-infused controls vs. euglycemic vs. hyperglycemic clamps. Student-Newman-Keuls multiple-comparison tests were run when appropriate. Unpaired t-tests assuming equal variance were used to determine significance for Northern blot analyses and for Western blots. Significance was set at P ≤ 0.05.
Skeletal muscle SREBP-1c mRNA and protein levels are increased by refeeding.
Substrate and hormonal responses to refeeding are presented in Table 2. In gastrocnemius muscle, refeeding for 3 or 6 h increased SREBP-1c mRNA levels (Fig. 1A, 110 and 190%, respectively; P ≤ 0.05 vs. fasted controls). As expected, refeeding also increased SREBP-1c mRNA levels in liver (3 h, 150%, P ≤ 0.001; 6 h, 300%, P ≤ 0.001, Fig. 1B). Furthermore, SREBP-1c mRNA levels in gastrocnemius muscle and liver as determined by RT-PCR using SREBP-1c-specific primers were consistent with those observed by Northern blot analysis (data not shown). Of note, SREBP-1a mRNA was not detectable at the number of cycles used to assess the refeeding effects (22 cycles), requiring 40 cycles for detection. These data support previous observations that SREBP-1c mRNA is by far the predominant SREBP transcript in skeletal muscle (11). In addition to SREBP-1c mRNA, levels of the mature 68-kDa (6 h only) and full-length 125-kDa forms (3 and 6 h) of the SREBP-1 protein were increased by refeeding in gastrocnemius muscle (mature at 3 h, 20%, P = 0.32 and at 6 h, 46%, P = 0.04; full length at 3 h, 35%, P = 0.04 and at 6 h, 55%, P = 0.008, Fig. 1C). Again as expected, liver SREBP-1 protein level was also increased by refeeding (mature at 3 h, 220%, P = 0.0001 and at 6 h, 38%, P = 0.04; full length at 3 h, 148%, P = 0.0004 and at 6 h, 141%, P = 0.0001; Fig. 1D).
We next determined the effects of refeeding on the expression of a number of SREBP-1c responsive [fatty acid synthase (FAS), acetyl CoA synthase (ACC)] and nutritionally responsive [uncoupling protein 3 (UCP3), hexokinase II (HKII)] genes, as a first step in understanding the role of SREBP-1c in the regulation of metabolic gene expression in skeletal muscle (Fig. 2). In gastrocnemius muscle, UCP3 mRNA levels decreased with 3 h (−49%; P = 0.0007) and 6 h (−75%; P ≤ 0.0001) of refeeding, and HKII mRNA levels increased with 6 h (126%, P = 0.0029) of refeeding (Fig. 2A). The level of FAS mRNA increased with both 3 h (102%, P = 0.0520) and 6 h (69%, P = 0.0026) of refeeding, while ACC2 mRNA levels remained unchanged with refeeding. In liver, changes in expression of the selected metabolic gene transcripts were as expected (Fig. 2B): refeeding suppressed levels of phosphoenolpyruvate carboxykinase (PEPCK) mRNA (3 h, −31%, P = 0.1820; 6 h, −84%, P = 0.0003) and G6Pase mRNA (3 h, −29%, P = 0.0225; 6 h, −46%, P = 0.0075), whereas FAS mRNA levels increased with refeeding at both 3 and 6 h (57%, P = 0.0034 and 49%, P = 0.0332, respectively). ACC1 mRNA levels increased with 6 h of refeeding only (146%, P = 0.0002).
Nutritional effects on SREBP-1c expression are not mimicked by physiological hyperinsulinemia in skeletal muscle or liver.
Both direct and indirect evidence suggest that insulin and glucose regulate SREBP-1c expression in liver (9, 10, 12, 14, 44), adipose tissue (19, 45), and skeletal muscle (8, 11, 42). However, it is unclear if physiological elevations of insulin and glucose levels alone are sufficient to explain the effects of refeeding on SREBP-1c gene expression. Thus we next assessed the effects of physiological hyperinsulinemia and/or hyperglycemia using the glucose-clamp technique on SREBP-1c expression in skeletal muscle and liver (Fig. 3). Substrate and hormonal responses in euglycemic- and hyperglycemic-hyperinsulinemic clamp studies are presented in Table 3. Unlike refeeding, neither 3 h of euglycemia-hyperinsulinemia nor hyperglycemia-hyperinsulinemia increased SREBP-1c mRNA levels in gastrocnemius muscle (Fig. 3A). Similar to muscle, liver SREBP-1c mRNA levels were not increased above those in saline-infused rats by euglycemia-hyperinsulinemia (Fig. 3B). However, hyperglycemia-hyperinsulinemia did increase SREBP-1c mRNA levels in liver (P = 0.0146), but the increase was blunted relative to 3 h of refeeding. RT-PCR analysis of SREBP-1c mRNA levels confirmed these observations (data not shown). Similar to SREBP-1c mRNA levels, 3 h of hyperinsulinemia, whether at euglycemia or hyperglycemia, did not increase the level of 68- or 125-kDa SREBP-1 protein in gastrocnemius muscle or in liver (Fig. 3, C and D), presenting further evidence for differential responses of SREBP-1c expression to refeeding and physiological hyperinsulinemia. Similar results were obtained with 4.5 h euglycemic-hyperinsulinemic clamps, where insulin levels were comparable to those observed at 6 h of refeeding (data not shown). Additionally, we assessed the effects of a supraphysiological concentration of circulating insulin using clamps in which insulin was administered at 30 mU·kg−1·min−1 for 3 h. In these studies (n = 5), glucose concentrations were maintained at 100 ± 5 mg/dl. Basal insulin was 0.9 ± 0.1 ng/ml and rose to 38.7 ng/ml by the end of the clamp procedure. Free fatty acids fell from 0.666 ± 0.17 mM to 0.13 ± 0.02 mM by the end of the clamp, and the glucose infusion rate was 33.9 ± 3.7 mg·kg−1·min−1. Levels of SREBP-1c mRNA were not increased significantly in gastrocnemius muscle or liver compared with control (n = 5) saline-infused animals (data not shown).
To ensure that metabolic gene expression was responsive to the insulin concentrations obtained in our study, we next assessed the effects of a 15 mU·kg−1·min−1 insulin clamp on the expression of the same genes measured in refeeding studies (Fig. 4). In gastrocnemius muscle (Fig. 4A), UCP3 mRNA levels decreased with both 3-h hyperinsulinemia-euglycemia and 3-h hyperinsulinemia-hyperglycemia (−36%, P = 0.0032 and −36%, P = 0.0282, respectively), similar to the effects of 3-h refeeding. HKII mRNA levels did not increase with 3-h hyperinsulinemia-euglycemia, again similar to 3-h refeeding, but hyperinsulinemia-hyperglycemia produced a 44% increase (P = 0.0051). In liver (Fig. 4B), PEPCK (−51% and −43%, P = 0.0005 and 0.0090, for hyperinsulinemia-euglycemia and hyperinsulinemia-hyperglycemia, respectively) and G6Pase (−95% for hyperinsulinemia-euglycemia, −78% for hyperinsulinemia-hyperglycemia, P ≤ 0.0001 and 0.0083, respectively) mRNA levels decreased markedly with hyperinsulinemia. Indeed, the effects of 3-h hyperinsulinemia on G6Pase and PEPCK mRNA levels were even more pronounced than the effects of 3 h of refeeding. In skeletal muscle and liver, mRNA levels of FAS, an SREBP-1c-responsive gene, were unchanged by hyperinsulinemia-euglycemia or hyperinsulinemia-hyperglycemia, unlike the effects of 3 h of refeeding. Similar to refeeding, neither ACC1 nor ACC2 mRNA changed with 3 h of hyperinsulinemia/hyperglycemia. Taken together, these data demonstrate that clamp studies elicited changes in expression of a number of insulin and/or glucose-responsive metabolic genes in a fashion similar to refeeding studies.
Hyperleptinemia decreases feeding-induced increases in SREBP-1c mRNA in skeletal muscle.
Previous studies have suggested that liver SREBP-1c gene expression is regulated by leptin (17). Furthermore, leptin decreases skeletal muscle lipid levels (5) and reduces fatty acid esterification (40), suggesting that alterations in SREBP-1c gene expression may play a role in the lipopenic effects of leptin in skeletal muscle. Thus we next investigated the effects of hyperleptinemia on the response of SREBP-1c expression to refeeding (Fig. 5). Substrate and hormone responses to 72 h of hyperleptinemia (HLEP) and refeeding are presented in Table 4. In gastrocnemius muscle, refeeding significantly increased SREBP-1c mRNA levels in both βGal (130%, P ≤ 0.0001) and HLEP (100%, P ≤ 0.0001) rats relative to fasted controls (data not shown). However, the refeeding-induced increase in SREBP-1c mRNA levels was blunted in HLEP rats relative to βGal rats (P = 0.0018). Thus, when examined across muscle groups of varying proportions of type I and type II fibers, refed SREBP-1c mRNA levels were suppressed 40% in gastrocnemius (P = 0.0257) and 41% in vastus lateralis (P = 0.0011), but not in soleus (P = 0.07), in HLEP relative to βGal. Unlike SREBP-1c mRNA levels, GAS full-length or mature SREBP-1 protein levels were not decreased by hyperleptinemia. SREBP-1 protein levels were not estimated in vastus lateralis or soleus muscle due to lack of tissue. Hyperleptinemia also suppressed refed levels of FAS mRNA in all three muscle groups studied (Fig. 6, gastrocnemius, −17%, P = 0.0060; vastus lateralis, −46%, P = 0.0411, soleus, −57%, P = 0.0050). There were no significant effects of hyperleptinemia on refed levels of ACC2 mRNA in any skeletal muscle group studied.
SREBP-1c is a critical regulator of lipid metabolism in adipose tissue and liver. Increases in the expression and processing of this transcription factor result in elevated lipogenesis, resulting from increases in the expression of a number of lipogenic metabolic genes (9, 10, 46). In this study we have investigated the in vivo regulation of SREBP-1c expression in rat skeletal muscle. These studies were undertaken as a first step toward understanding the potential role of SREBP-1c in regulating lipid metabolism in a tissue that is not generally regarded as having a high lipogenic capacity. Three major novel conclusions arise from the current study. First, we demonstrate that the expression of skeletal muscle SREBP-1c mRNA and protein are regulated by nutritional status in a fashion similar to that observed in liver and adipose tissue. Second, we demonstrate that elevation of insulin and glucose to concentrations associated with the high physiological range does not imitate the effects of refeeding on SREBP-1c expression. Third, we demonstrate that hyperleptinemia inhibits the effects of refeeding on SREBP-1c mRNA levels but does not suppress expression of SREBP-1 protein.
Nutritional regulation of SREBP-1c expression is well described in liver (12, 14, 18, 55–57) and adipose tissue (19). The current study extends these observations to demonstrate that there is a robust stimulation of skeletal muscle SREBP-1c gene expression with as little as a 3-h refeeding period. Both SREBP-1c mRNA and SREBP-1 protein levels were increased by refeeding, and the extent of the stimulation of SREBP-1c mRNA level was similar to that observed in liver. These data are in good agreement with the recent study of Bizeau et al. (3). Taken together, these data offer further indirect evidence for an important functional role for SREBP-1c in skeletal muscle. In addition to the effects on SREBP-1c expression, refeeding increased the level of HKII mRNA and FAS mRNA, while UCP3 mRNA level was decreased. FAS is a target gene for SREBP-1c in liver (9, 10, 43) and adipose tissue (19), suggesting that FAS expression may also be regulated by SREBP-1c in skeletal muscle in vivo. In support of this hypothesis, Guillet-Deniau et al. (11) demonstrated in vitro that overexpression of a dominant positive SREBP in cultured myotubes increases FAS mRNA levels. Previous studies have demonstrated that HKII gene expression is regulated by insulin (28, 38, 53) and glucose (38) and that UCP3 gene expression is increased in the fasted state (4, 51, 52, 54). Additionally, Guillet-Deniau et al. (11) demonstrated that overexpression of a dominant positive SREBP-1c decreased UCP3 expression in cultured myotubes. Taken together with the observations in the current study, these data support the hypothesis UCP3 is a target gene of SREBP-1c in skeletal muscle in vivo.
A primary goal of the current study was to determine if the effects of refeeding on SREBP-1c gene expression are reproduced by physiological increases in insulin and glucose. An unanticipated finding of our study was that SREBP-1c expression in skeletal muscle and liver did not increase in response to a physiological hyperinsulinemic clamp. Liver SREBP-1c mRNA was increased by a hyperglycemic-hyperinsulinemic clamp, but even here the increase was blunted compared with that observed with refeeding. These results were unexpected because a number of in vitro studies in cell culture have demonstrated that insulin and glucose metabolism are positive regulators of SREBP-1c expression (9, 19, 20, 31). Furthermore, administration of insulin to streptozotocin diabetic animals increases hepatic SREBP-1c expression (1). Also, two studies (8, 42) have demonstrated an increase in skeletal muscle SREBP-1c mRNA, as measured by RT-PCR, in a 3-h hyperinsulinemic clamp in normal humans. However, a number of important issues should be considered when comparing the data from our clamp studies with data from previous in vivo and in vitro studies. Previous insulin clamp studies (8, 42) were performed in humans; the control tissues in these studies were preclamp biopsies, rather than biopsies taken after a saline infusion, and the effects of insulin alone were not compared with the effects of refeeding. In many in vitro studies addressing the effects of insulin on SREBP-1c expression, including a study in primary myotubes (11), insulin concentrations are extremely high (∼10× to 200× physiological). Also, SREBP-1c mRNA levels in the presence of insulin are compared with SREBP-1c mRNA levels in the absence of insulin, a comparison that would maximize the likelihood of observing an insulin effect. This is also the case in streptozotocin-induced diabetes (1), where SREBP-1c mRNA levels and insulin levels are substantially suppressed before the administration of a large insulin bolus that results in unknown plasma insulin concentrations. In our study, basal insulin concentrations are present in the saline-infused animals, and the insulin concentration during the clamp was maximally (in supraphysiological insulin clamps) approximately five times greater than the insulin concentration in the fed state. However, it is clear from our data that metabolic gene expression was responsive to the insulin and glucose levels achieved during the clamp. Thus levels of gene transcripts (PEPCK, G6Pase, UCP3, and HKII) that are responsive to insulin and/or glucose (16, 25, 29, 30, 38, 41, 51, 53) changed in directions and magnitudes in the clamp study that matched the effects of refeeding. Finally, no study has compared the effects of refeeding vs. the effects of insulin alone on SREBP-1c gene expression. Thus, based on the data presented in the current study, it is plausible to suggest that refeeding has a greater effect on SREBP-1c gene expression than an insulin clamp alone. This may be explained by changes in the levels of other nutritionally responsive positive and/or negative regulators of SREBP-1c expression that are not altered under clamp conditions alone. Negative regulators of SREBP-1c expression in liver include free fatty acids (18, 55–57) and glucagon/cAMP (10) and could be expected to alter SREBP-1c expression in the current study. The circulating concentration of free fatty acids is decreased ∼60% in rats refed for 3 h after a 48-h fast (7), a reduction that is similar to that observed in our clamp studies (a 75% decrease in clamped rats vs. saline-infused rats). Thus differences in levels of free fatty acids between the refeeding studies and clamp studies do not explain the inability of the clamp conditions to increase SREBP-1c gene expression. Refeeding (3 h) of 18-h fasted rats does not reduce circulating glucagon concentration below fasted concentrations (22). Conversely, glucagon concentrations are maximally suppressed at insulin infusion rates used in the present study (21). Thus neither elevated free fatty acids nor glucagon concentrations in the clamp studies compared with the refeeding studies can be invoked to explain the lack of an effect of physiological hyperinsulinemia on SREBP-1c gene expression. Clearly, further studies are required to determine if other nutritional factors are involved in the regulation of SREBP-1c gene expression during refeeding.
The effects of leptin on lipid metabolism are well described. Thus leptin increases fatty acid oxidation and suppresses fatty acid synthesis in liver and adipose tissue (6, 17, 24). In skeletal muscle, leptin increases fatty acid oxidation and decreases fatty acid esterification (23, 32, 33, 49, 50). Previous studies have suggested that SREBP-1c expression is regulated by leptin. Thus, in adenovirus-mediated hyperleptinemic rats, hepatic SREBP-1c mRNA levels are reduced (17). The current study extends these observations by addressing the effects of hyperleptinemia on skeletal muscle SREBP-1c expression in the physiological condition most relevant for SREBP-1c function, i.e., the absorptive phase. The data demonstrate that refeeding-induced increases in SREBP-1c mRNA levels are blunted by hyperleptinemia but that hyperleptinemia does not decrease SREBP-1 protein levels. This latter observation is somewhat surprising, although small but functionally important changes in SREBP-1 protein levels may have been missed because of the lack of robustness of Western blots for quantitative analysis. Of interest is the observation that FAS gene expression, a target gene of SREBP-1c, is also reduced in hyperleptinemic rats compared with controls. However, whether this decrease is a result of decreased SREBP-1c activity is unknown.
In conclusion, the present study offers important insight into the in vivo regulation of expression of the transcription factor SREBP-1c in skeletal muscle. Expression of the SREBP-1c gene is nutritionally responsive in skeletal muscle, together with a number of other metabolic genes important for carbohydrate and lipid metabolism. Furthermore, the nutritional regulation of SREBP-1c is not replicated by physiological elevations of insulin, implicating other factors in the rapid increases in SREBP-1c induced by refeeding. Finally, leptin suppresses refeeding effects on SREBP-1c and FAS gene expression, an observation that may partially explain the lipopenic effects of leptin. While the role of SREBP-1c in skeletal muscle nutrient metabolism requires further study, the similarities in regulation of this transcription factor in liver and skeletal muscle suggest that it is an important determinant of nutrient partitioning in both tissues.
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Grant RO1-DK-58855–01 (to R. M. O'Doherty). S. R. Commerford is supported by NIDDK Grant T32-DK07052 (Research Training in Diabetes and Endocrinology).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society