Endocrinology and Metabolism

Stearoyl-CoA desaturase-1 deficiency reduces ceramide synthesis by downregulating serine palmitoyltransferase and increasing β-oxidation in skeletal muscle

Agnieszka Dobrzyn, Pawel Dobrzyn, Seong-Ho Lee, Makoto Miyazaki, Paul Cohen, Esra Asilmaz, D. Grahame Hardie, Jeffrey M. Friedman, James M. Ntambi

Abstract

Stearoyl-CoA desaturase (SCD) has recently been shown to be a critical control point of lipid partitioning and body weight regulation. Lack of SCD1 function significantly increases insulin sensitivity in skeletal muscles and corrects the hypometabolic phenotype of leptin-deficient ob/ob mice, indicating the direct antilipotoxic action of SCD1 deficiency. The mechanism underlying the metabolic effects of SCD1 mutation is currently unknown. Here we show that SCD1 deficiency reduced the total ceramide content in oxidative skeletal muscles (soleus and red gastrocnemius) by ∼40%. The mRNA levels and activity of serine palmitoyltransferase (SPT), a key enzyme in ceramide synthesis, as well as the incorporation of [14C]palmitate into ceramide were decreased by ∼50% in red muscles of SCD1−/− mice. The content of fatty acyl-CoAs, which contribute to de novo ceramide synthesis, was also reduced. The activity and mRNA levels of carnitine palmitoyltransferase I (CPT I) and the rate of β-oxidation were increased in oxidative muscles of SCD1−/− mice. Furthermore, SCD1 deficiency increased phosphorylation of AMP-activated protein kinase (AMPK), suggesting that AMPK activation may be partially responsible for the increased fatty acid oxidation and decreased ceramide synthesis in red muscles of SCD1−/− mice. SCD1 deficiency also reduced SPT activity and ceramide content and increased AMPK phosphorylation and CPT I activity in muscles of ob/ob mice. Taken together, these results indicate that SCD1 deficiency reduces ceramide synthesis by decreasing SPT expression and increasing the rate of β-oxidation in oxidative muscles.

  • adenosine 5′-monophosphate-activated protein kinase
  • ob/ob mouse
  • leptin
  • carnitine palmitoyltransferase

mammalian stearoyl-CoA desaturase (SCD) belongs to a family of desaturases that has been highly conserved throughout evolution in animals, plants, and yeast. SCD catalyzes the desaturation of saturated fatty acyl-CoAs, the preferred substrates being palmitoyl- and stearoyl-CoA, which are converted to palmitoleoyl- and oleoyl-CoA, respectively (16). The saturated-to-unsaturated fatty acid ratio has been implicated in the regulation of cell growth and differentiation through effects on membrane fluidity and signal transduction (31). Recent studies of asebia (abJ and ab2J) mouse strains that have a natural mutation in the SCD1 gene (46) and a mouse model with a targeted disruption of the SCD1 gene (30) have provided evidence that SCD is a critical control point in lipid partitioning and body weight regulation. SCD1 deficiency results in increased energy expenditure (33) and decreased body adiposity due to the upregulation of genes of fatty acid oxidation (15, 33) and the downregulation of genes of lipid synthesis in liver (reviewed in Ref. 32). SCD1 knockout mice are considerably leaner than their wild-type littermates and have decreased hepatic triglyceride synthesis, increased rate of β-oxidation in liver (15) and brown adipose tissue, and higher basal thermogenesis (27). SCD1-deficient animals are also resistant to diet-induced obesity (33). Recent studies show that SCD1 is required for the fully developed obese phenotype of leptin-deficient ob/ob mice (9) and suggest that a significant proportion of leptin’s metabolic effect results from inhibition of this enzyme (9, 24). In addition, SCD1 deficiency significantly increases insulin sensitivity in skeletal muscle (34) and corrects the hypometabolic phenotype of leptin deficiency (9), suggesting that inhibition of SCD1 could be of benefit for the treatment of obesity and the metabolic syndrome.

An increasing body of evidence indicates that several manifestations of the metabolic syndrome and type 2 diabetes mellitus, including insulin resistance and dysfunction of the pancreatic β-cells, occur secondarily to overaccumulation of lipids in nonadipose tissues, such as skeletal muscle and pancreatic β-cells (44). Although the lipotoxicity correlates well with increased intracellular triglyceride content, lipid-induced damage to tissues is probably not caused by triglyceride (28, 41, 43) but rather by derivatives of unoxidized palmitoyl-CoA, with ceramide being the most likely suspect (43, 44). In support of this notion, ceramide has been shown to accumulate in various rodent models of insulin resistance, including obese fa/fa rats (42) and mice overexpressing lipoprotein lipase (25), as well as in the muscle of obese insulin-resistant humans (1). Ceramide-induced inhibition of Akt activity has been implicated in the impaired translocation of GLUT4 in muscle (40) and might also play an important role in lipoapoptosis. Ceramide is also involved in the lipotoxicity of pancreatic β-cells (36) and the heart (47) that occurs in the unleptinied tissues of Zucker diabetic fatty rats and in transgenic mice with myocardial overexpression of acyl-CoA synthase (8). Moreover, recent studies have shown that endogenously produced ceramide is not only sufficient but also necessary for the development of palmitate-induced insulin resistance (7) and apoptosis (3, 36).

The signaling pool of ceramide is generated by sphingomyelin hydrolysis and/or by de novo synthesis (21). The first committed step in de novo ceramide synthesis, the esterification of palmitoyl-CoA and serine, is catalyzed by serine palmitoyltransferase (SPT). Transcriptional regulation of SPT has been observed in response to several types of inflammatory and stress stimuli (19), leptin receptor mutation (36), and activation of AMP-activated protein kinase (AMPK) (3) and has been suggested to be in response to changes in intracellular fatty acid content. A recent study on Chinese hamster ovary cells has shown that oleic acid, the major product of SCD, may play an important role in the regulation of intracellular ceramide synthesis (28). Thus it is possible that SCD1 deficiency protects against lipotoxicity in muscle by regulation of ceramide synthesis.

The present work was therefore undertaken to study the effect of SCD1 deficiency on ceramide synthesis in different skeletal muscle types. We found that mRNA levels and activity of SPT as well as the ceramide content were decreased in the soleus and red gastrocnemius of SCD1−/− mice. The content of fatty acyl-CoAs was also reduced. The effect of SCD1 deficiency on ceramide synthesis was less pronounced in white gastrocnemius. Further experiments revealed that SCD1 deficiency increased the activity of carnitine palmitoyltransferase I (CPT I) and the rate of β-oxidation in oxidative muscles. Phosphorylation of AMPK was increased in soleus and red gastrocnemius muscles of SCD1−/− mice, suggesting that AMPK activation is partially responsible for the increased fatty acid oxidation and decreased ceramide synthesis observed in SCD1 deficiency. The SCD1 mutation also reduced SPT activity and ceramide content and increased AMPK phosphorylation in the skeletal muscle of leptin-deficient ob/ob mice. The data indicate that SCD1 deficiency reduces ceramide synthesis by decreasing SPT expression and increasing the rate of β-oxidation in oxidative muscles.

EXPERIMENTAL PROCEDURES

Animals.

The generation of SCD1−/− and abJ/abJ;ob/ob mice has been previously described (30, 9). Twelve-week-old purebred homozygous SCD1−/− and wild-type male mice were used. The abJ/ab+;ob/ob and abJ/abJ;ob/ob mice were 16 wk old. Mice were housed in a pathogen-free barrier facility operating a 12:12-h light-dark cycle and fed a normal nonpurified diet (5008 test diet; PMI Nutrition International, Richmond, IN). The breeding of these animals was in accordance with the protocols approved by the Animal Care Research Committee of the University of Wisconsin-Madison and The Rockefeller University Laboratory Animal Research Center. Mice were euthanized, and the soleus and the red and white sections of gastrocnemius muscles were extracted, frozen in liquid nitrogen, and stored at −80°C.

Materials.

[14C]palmitic acid, l-[3-14C]serine, l-[3H]carnitine, and [α-32P]dCTP were purchased from American Radiolabeled Chemicals (St. Louis, MO) and Perkin-Elmer Life Sciences (Boston, MA). Anti-AMPKα1 and -α2 and anti-phospho-AMPK (Thr172) antibodies were obtained as described (39, 45). All other chemicals were purchased either from Sigma (St. Louis, MO) or Fisher Scientific (Pittsburgh, PA).

Measurement of lipids and fatty acyl-CoAs.

Fatty acyl-CoAs were extracted from muscle tissues and analyzed as described (23). Sphingomyelin, ceramide, and free fatty acids (FFA) were extracted by the method of Bligh and Dyer (4) and measured as described (12, 18). Briefly, the lipids were fractionated into different classes by thin-layer chromatography on silica gel-60 plates (Merck). The bands corresponding to sphingomyelin, ceramide, and FFA were scraped off the plate and transferred to screw-cap glass tubes containing methylpentadecanoic acid as an internal standard. Fatty acids were then transmethylated in the presence of 14% boron trifluoride in methanol. The resulting methyl esters were extracted with hexane and analyzed by gas liquid chromatography (12). Total contents of sphingomyelin, ceramide, and FFA were calculated from individual fatty acid content in each fraction.

[14C]palmitic acid incorporation into ceramide.

The mice were anesthetized, and 0.8 μCi of [14C]palmitic acid (55 mCi/mmol) per 20 g body wt, conjugated to albumin, was administered into the tail vein of SCD1−/− and wild-type mice. Ten minutes after administration of the label, muscle samples were taken. The muscle lipids were extracted, and ceramide was isolated as described above. The lipid bands corresponding to ceramide were scraped off the plates, and the radioactivity was counted in a liquid scintillation counter.

Enzyme activity.

Activity of SPT in isolated microsomes was measured with l-[3-14C]serine as substrate, as described (11). The reaction was terminated by addition of 1.5 ml of chloroform-methanol (2:1, vol/vol). The radiolabeled product, 3-ketosphinganine, was separated from radiolabeled serine by phase partitioning. Sphinganine (25 μg) was added as a carrier, followed by the addition of 1 ml of chloroform and 2 ml of 0.5 N NH4OH. The chloroform layer was counted in a scintillation counter.

CPT I activity was measured in isolated mitochondria with the use of l-[3H]carnitine as substrate, as described (5). The reaction was stopped by 0.5 ml of 4 M ice-cold perchloric acid (PCA), and samples were centrifuged at 13,000 g for 10 min. The resulting pellet was washed with 500 μl of 2 mM PCA and then resuspended in 800 μl of H2O and extracted with 600 μl of butanol. Three hundred microliters of the butanol phase were counted by liquid scintillation.

Measurement of mitochondrial fatty acid oxidation.

Fatty acid oxidation was determined as previously described (15, 26) using [14C]palmitic acid and presented as the sum of labeled CO2 released during oxidation, ketone body and acid soluble-labeled oxidation products.

Isolation and analysis of RNA.

Total RNA was isolated from muscles using TRIzol reagent. CPT I gene expression was analyzed by Northern blotting. Twenty micrograms of total RNA were fractionated on 1% agarose-2.2 M formaldehyde gels and transferred to Hybond N+ nylon membranes. After UV cross-linking, the membrane was hybridized with cDNA probes labeled with [32P]dCTP by a random primer labeling kit (Promega, Madison, WI). After a washing, the membranes were exposed to X-ray film at −80°C, and signals were quantified by densitometry. The probes for CPT I were obtained from Dr. Jeffrey M. Peters (The Pennsylvania State University, University Park, PA). The pAL15 mRNA was used as an internal control. Semiquantitative RT-PCR was performed to analyze LCB1 and LCB2 gene expression. Primers used to amplify LCB1 were 5′-ACCTGGAGCGACTGCTAAAA-3′ (sense) and 5′-CAGTGACCACAACCCTGATG-3′ (antisense). Primers used to amplify LCB2 were 5′-CCTGTCAGCAGCTCATACCA-3′ (sense) and 5′-GGAAAGTCAGCAGAAGGCAC-3′ (antisense). The amplified products were resolved on 1.5% agarose gel, and the levels of mRNA were expressed as the ratio of signal intensity relative to that for GAPDH.

Western blot analysis.

Muscle samples were homogenized in ice-cold 50 mM HEPES buffer (pH 7.4) containing 150 mM NaCl, proteases inhibitors, and 10% glycerol and centrifuged at 10,000 g for 10 min. Proteins were separated on a 9% SDS-PAGE gel and then transferred to and immobilized on nitrocellulose membrane. The membranes were immunoblotted with antibody against phosphopeptides, based on the amino acid sequence surrounding Thr172 of the α-subunit of AMPK (39). Protein levels of α1- and α2-subunits of AMPK were determined with specific antibodies. The proteins were visualized with enhanced chemiluminescence (Amersham Biosciences) and quantified by densitometry.

Protein content.

The protein concentration was determined with a Bio-Rad protein assay (Hercules, CA), using BSA as a standard.

Statistical analysis.

Results were analyzed by Student’s t-test. A difference of P < 0.05 was considered significant. Values are presented as means ± SD (n = 6 mice/group in the case of SCD1−/− and wild-type mice, and n = 4 in the case of abJ/abJ;ob/ob and abJ/ab+;ob/ob mice).

RESULTS

SCD1 deficiency decreases the total contents of ceramide, sphingomyelin, free fatty acids, and fatty acyl-CoAs in soleus and red gastrocnemius muscles.

The total content of ceramide in the soleus and red gastrocnemius (the oxidative muscles) was reduced by 42 and 48%, respectively, in SCD1−/− mice compared with wild-type controls (Fig. 1A). In the white gastrocnemius, which is a glycolytic muscle, SCD1 deficiency did not significantly change the total ceramide content (Fig. 1A).

Fig. 1.

Total content of ceramide (A), sphingomyelin (B), free fatty acids (FFA; C), and fatty acyl-CoAs (FA-CoAs; D) in the skeletal muscles of SCD1−/− and wild-type mice. Gastroc (R), red gastrocnemius; Gastroc (W), white gastrocnemius; SCD, stearoyl-CoA desaturase. *P < 0.05 vs. wild-type mice.

A signaling pool of ceramide is formed through sphingomyelin hydrolysis by neutral and/or acid sphingomyelinases and de novo synthesis (21). Therefore, we analyzed the total content of sphingomyelin and the activities and gene expression of acid and neutral sphingomyelinases in each muscle type. The content of sphingomyelin was significantly reduced in soleus and red gastrocnemius muscles of SCD1−/− mice by 24 and 29%, respectively, compared with wild-type mice. In the white gastrocnemius muscle, the sphingomyelin content was similar in both SCD1−/− and wild-type mice (Fig. 1B). SCD1 deficiency did not affect gene expression or activities of sphingomyelinases in the three muscle types (data not shown).

Fatty acyl-CoAs (FA-CoAs), the biochemically active fatty acids, are the substrates for de novo ceramide synthesis. The total contents of both FFA and FA-CoAs were decreased by 34 and 55%, respectively, in soleus and by 37 and 50%, respectively, in red gastrocnemius of SCD1−/− mice compared with wild-type controls (Fig. 1, C and D). In white gastrocnemius muscle, SCD1 deficiency did not significantly affect the total contents of FFA and FA-CoAs (Fig. 1, C and D).

SCD1 deficiency alters the fatty acid composition of FA-CoA and FFA fractions.

The contents of major FA-CoAs, palmitoyl-CoA (16:0), stearoyl-CoA (18:0), oleoyl-CoA (18:1), palmitoleoyl-CoA (16:1), and linoleoyl-CoA (18:2), were decreased by 40, 27, 86, 90, and 28%, respectively, in soleus (Fig. 2A) and by 55, 26, 74, 85, and 20%, respectively, in red gastrocnemius (Fig. 2B) of SCD1−/− mice compared with wild-type controls. In white gastrocnemius, the total content of FA-CoA was not changed (Fig. 1D); however, the contents of palmitoleoyl-CoA and oleoyl-CoA, the major products of SCD, were decreased by 66 and 46%, whereas the contents of palmitoyl-CoA and stearoyl-CoA were increased by 30 and 40%, respectively, in SCD1−/− mice relative to wild-type mice (Fig. 2C). In oxidative muscles, SCD1 deficiency also caused a significant decrease in the content of myristic (14:0), palmitic (16:0), stearic (18:0), palmitoleic (16:1), oleic (18:1), and linoleic (18:2) acids (Table 1). The content of arachidonic acid (20:4) was increased by 54% in soleus and by 40% in red gastrocnemius of SCD1−/− mice (Table 1). In the white gastrocnemius of SCD1−/− mice, the contents of palmitoleic (16:1) and oleic (18:1) acids were significantly decreased, whereas the contents of corresponding saturated fatty acids, 16:0 and 18:0, respectively, were slightly increased. The levels of 14:0, 18:2, and 20:4 fatty acids were unchanged in the white gastrocnemius (Table 1).

Fig. 2.

Fatty acid composition of intramuscular FA-CoA in soleus (A), red gastrocnemius (B), and white gastrocnemius (C) of SCD1−/− and wild-type mice. *P < 0.05 vs. wild-type mice.

View this table:
Table 1.

FFA content in skeletal muscles of SCD1−/− and WT mice

SCD1 deficiency reduces LCB gene expression and decreases SPT activity.

SPT is the rate-limiting enzyme in de novo synthesis of ceramide (19). The mRNA levels of both SPT subunits (LCB1 and LCB2) were reduced by 80 and 70% in soleus and by 60 and 86% in red gastrocnemius, respectively, of SCD1−/− mice relative to wild-type controls (Fig. 3A). In white gastrocnemius, SCD1 deficiency caused only a 15% decrease in expression of the LCB2 gene but did not affect the LCB1 mRNA level (Fig. 3A). The levels of GAPDH mRNA used as the loading control were not significantly altered. The activities of SPT in the soleus and red gastrocnemius muscles were reduced by 45 and 48%, respectively, in SCD1−/− mice relative to wild-type controls (Fig. 3B). In the white gastrocnemius muscle, SPT activity was similar in both groups of animals (Fig. 3B). To confirm that SCD1 deficiency affects de novo ceramide synthesis, we analyzed the incorporation of [14C]palmitic acid into the ceramide fraction in each muscle type. The incorporation of labeled palmitate into ceramide was decreased in both soleus and red gastrocnemius muscles by 53 and 56%, respectively, but did not significantly change in white gastrocnemius of SCD1−/− mice relative to wild-type controls (Fig. 3C).

Fig. 3.

mRNA levels of the two serine palmitoyltransferase (SPT) subunits (LCB1 and LCB2; A), SPT activity (B), and [14C]palmitic acid incorporation into ceramide fraction (C) in the skeletal muscles of SCD1−/− and wild-type mice. RT-PCR was performed to measure LCB mRNA levels, normalized to GAPDH mRNA control. *P < 0.05 vs. wild-type mice.

The activity and gene expression of CPT I and mitochondrial β-oxidation are increased in the red skeletal muscles of SCD1−/− mice.

The decreased level of FA-CoAs and FFA in the red muscles of SCD1−/− mice could result from increased rates of fatty acid β-oxidation. Therefore, we measured gene expression and activity of CPT I, the rate-limiting enzyme for mitochondrial β-oxidation. Northern blot analysis showed that the levels of CPT I mRNA were increased 2.2- and 3-fold in soleus and red gastrocnemius, respectively, in SCD1−/− mice compared with wild-type controls (Fig. 4A). The activities of CPT I in soleus and red gastrocnemius muscle were increased by 1.8- and 1.9-fold, respectively, in SCD1−/− mice compared with wild-type controls (Fig. 4B). Palmitate β-oxidation in soleus and red gastrocnemius was increased by 1.7- and 2-fold in SCD1−/− mice relative to wild-type controls (Fig. 4C). The effect of SCD1 deficiency on CPT I activity and gene expression as well as on the rate of β-oxidation was less pronounced in white gastrocnemius (Fig. 4).

Fig. 4.

Carnitine palmitoyltransferase I (CPT I) mRNA level (A), CPT I activity (B), and palmitate oxidation (C) in the skeletal muscles of SCD1−/− and wild-type mice. Northern blot analysis was performed to measure CPT I mRNA level, normalized to pAL15 mRNA control. *P < 0.05 vs. wild-type mice.

SCD1 deficiency increases phosphorylation of AMPK in oxidative skeletal muscles.

AMPK, the metabolic sensor that monitors cellular AMP and ATP levels, is involved in regulation of CPT I (22) and has recently been shown to be an essential regulator of fatty acid oxidation in skeletal muscles (29, 35, 41). SCD1 deficiency increased phosphorylation of Thr172 in the α-subunits of AMPK in soleus and red gastrocnemius by 85 and 92%, respectively (Fig. 5). Phosphorylation of this residue is essential for activation of AMPK (37). The level of AMPK phosphorylation was not significantly changed in white gastrocnemius of SCD1−/− mice relative to wild-type mice (Fig. 5). Loss of SCD1 function did not affect the protein levels of α1- and α2-subunits of AMPK in the three muscle types (Fig. 5).

Fig. 5.

Phospho-AMP-activated protein kinase (P-AMPK) and AMPKα1 and AMPKα2 protein subunits in skeletal muscles of SCD1−/− and wild-type mice. Phosphorylation of AMPK as well as AMPKα1 and AMPKα2 protein contents was quantified with Western blotting.

SCD1 deficiency increases AMPK phosphorylation and CPT I activity and decreases ceramide synthesis in leptin-deficient ob/ob mice.

Leptin is a known activator of AMPK and β-oxidation in skeletal muscle (29). To determine whether the increased AMPK phosphorylation observed in SCD1−/− mice is leptin dependent, we measured the phosphorylation of AMPK and CPT I activity in ob/ob leptin-deficient mice crossed with abJ/abJ mice, which have a natural mutation in the SCD1 gene (9). In the soleus and red gastrocnemius of double-mutant abJ/abJ;ob/ob mice, we found a 46 and 58% increase in phosphorylation of AMPK, respectively, relative to the abJ/ab+;ob/ob control (Fig. 6A). This corresponded to 34 and 38% higher activities of CPT I in soleus and red gastrocnemius of double-mutant mice (Fig. 6B). SCD1 deficiency did not significantly affect AMPK phosphorylation or the activity of CPT I in white gastrocnemius muscle of abJ/ab+;ob/ob mice (Fig. 6, A and B). The activity of SPT was decreased by 29% in soleus and 35% in red gastrocnemius of double-mutant abJ/abJ;ob/ob mice relative to abJ/ab+;ob/ob mice (Fig. 6C). SCD1 deficiency also decreased the total content of ceramide by 25% in soleus and by 28% in red gastrocnemius of abJ/ab+;ob/ob mice (Fig. 6D). SPT activity and ceramide content were not significantly altered in the white gastrocnemius of double-mutant abJ/abJ;ob/ob mice relative to abJ/ab+;ob/ob mice (Fig. 6, C and D).

Fig. 6.

SCD1 deficiency increases AMPK phosphorylation (A) and CPT I activity (B) and decreases SPT activity (C) and ceramide content (D) in skeletal muscles of ob/ob mice. Phospho-AMPK and AMPKα1 and AMPKα2 proteins in skeletal muscles of abJ/ab+;ob/ob mice and double-mutant abJ/abJ;ob/ob mice were assayed by Western blotting. *P < 0.05 vs. wild-type mice.

DISCUSSION

SCD, the rate-limiting enzyme in the de novo synthesis of unsaturated fatty acids from saturated acyl-CoAs, has recently been shown to be a regulator of hepatic lipogenesis (14, 32) and a critical enzyme for synthesis of triglycerides, alkyl-2,3-diacylglycerol, and cholesterol esters (30, 32). The present study shows that loss of SCD1 function decreases de novo ceramide synthesis by downregulating SPT activity and the expression of both SPT subunits (LCB1 and LCB2) in the oxidative skeletal muscles. The contents of FA-CoAs were also reduced. Furthermore, the activity and gene expression of CPT I as well as the rate of fatty acid β-oxidation were increased in oxidative muscles of SCD1−/− mice. Because the effect of SCD1 deficiency on ceramide content and the rate of β-oxidation was less pronounced in white gastrocnemius, reduction of ceramide synthesis in oxidative muscles of SCD1−/− mice appears to be largely the result of increased rates of β-oxidation. SCD1 deficiency increases phosphorylation of AMPK in soleus and red gastrocnemius muscles, suggesting that increased fatty acid oxidation and a reduced rate of ceramide synthesis in the skeletal muscles of SCD1−/− mice are at least partially caused by AMPK activation. In addition to the downregulation of protein tyrosine phosphatase 1B expression (34), a decrease in ceramide synthesis might be a possible mechanism for the increased insulin sensitivity observed in the skeletal muscles of SCD1−/− mice.

The accumulation of ceramide in skeletal muscles has usually been considered to occur via receptor-mediated activation of neutral and/or acid sphingomyelinases (12, 38). However, SCD1 deficiency reduces the total content of ceramide in both soleus and red gastrocnemius by >40% (Fig. 1A) without changes in the activities of sphingomyelinases (data not shown), despite the lower amount of sphingomyelin (Fig. 1B). The decrease in the content of ceramide associated with SCD1 deficiency appears to be due to decreased de novo synthesis as evidenced by decreased SPT activity (Fig. 3B) and reduced incorporation of [14C]palmitate into ceramide (Fig. 3C). Because ceramide is one of the substrates for sphingomyelin biosynthesis (21), a decrease in de novo ceramide synthesis would lead to reduction in sphingomyelin content. The changes in SPT activity in red muscles of SCD1−/− mice parallel the decrease in both LCB1 and LCB2 mRNA levels (Fig. 3A). The mechanism underlying the regulation of LCB mRNA levels in skeletal muscles is largely unknown. The data obtained in the present study support the possibility that changes in intracellular FFA content may affect LCB mRNA levels, as suggested recently (19).

AMPK has been shown to downregulate various ATP-consuming anabolic pathways, including fatty acid and cholesterol synthesis (22), and might also participate in the regulation of ceramide synthesis. Indeed, 5-aminoimidazole-4-carboxamide (AICA) riboside, which enters cells and is converted to AICA ribotide, an ATP analog (10), has been shown to inhibit palmitate-induced SPT activity, LCB2 gene expression, and de novo ceramide synthesis in rat astrocytes (3) and bovine retinal pericytes (6). Previously, we (15) showed that SCD1 deficiency activates the AMPK pathway in liver. In the present study, we found increased AMPK phosphorylation in soleus and red gastrocnemius of SCD1−/− mice relative to wild-type controls (Fig. 5). Consequently, in both oxidative muscles of SCD1−/− mice, changes in AMPK phosphorylation parallel the downregulation of SPT and decrease in total ceramide content.

How does AMPK regulate ceramide synthesis? Activation of AMPK leads to inhibition of acetyl-CoA carboxylase activity, resulting in a drop in malonyl-CoA levels. This disinhibits CPT I, causing an increase in fatty acid oxidation accompanied by a decrease in intracellular FFA (22). SCD1 deficiency upregulates both gene expression and the activity of CPT I (Fig. 4, A and B) and increases rates of β-oxidation (Fig. 4C) in the soleus and red gastrocnemius. Consequently, the total contents of FFA and FA-CoA were decreased in oxidative muscles of SCD1−/− mice (Fig. 1, C and D). The fatty acid composition in muscle FFA and FA-CoA fractions was also changed by SCD1 deficiency (Table 1 and Fig. 2). The reduction in palmitoleoyl-CoA and oleoyl-CoA in SCD1-deficient mice was expected to be accompanied by an increase in their saturated fatty acyl-CoA precursors, palmitoyl-CoA and stearoyl-CoA. However, in soleus and red gastrocnemius of SCD1−/− mice, the contents of palmitoyl-CoA and stearoyl-CoA as well as palmitic and stearic acids were significantly reduced (Table 1 and Fig. 2), possibly because of increased β-oxidation. The activity of SPT is strongly affected by the availability of its substrates serine and palmitoyl-CoA and shows a bell-shaped dependence on the content of palmitic acid (but not other fatty acids) in vitro (17, 20). Thus a decrease in the cellular levels of palmitoyl-CoA (Fig. 2) in soleus and red gastrocnemius of SCD1−/− mice would result in decreased SPT activity and a decrease in the rate of ceramide synthesis. This hypothesis is also supported by data obtained in white gastrocnemius, where palmitoyl-CoA level (Fig. 2) and the rate of β-oxidation (Fig. 5) were unaffected by SCD1 deficiency, and consequently, the rate of ceramide synthesis was unchanged. In addition, the effect of SCD1 deficiency on ceramide content in red skeletal muscles is similar to that observed in muscles under conditions that enhance fatty acid oxidation, such as troglitazone treatment (47) and exercise (12, 13). Taken together, these data indicate that activation of AMPK and an increase in β-oxidation, via decreasing intramuscular palmitoyl-CoA content, are the main factors causing downregulation of SPT activity and reduction in ceramide synthesis in skeletal muscle of SCD1−/− mice.

The mechanism by which SCD1 deficiency upregulates AMPK in skeletal muscle is currently unknown. Because AMPK is activated in muscle by contraction (2) and because SCD1−/− mice have a higher energy expenditure (33), one would expect that increased physical activity might be responsible for activation of AMPK in muscles of SCD1−/− mice. However, although exercise stimulates AMPK in both oxidative and glycolytic muscles (2), SCD1−/− mice show higher AMPK phosphorylation only in oxidative muscles and not in white gastrocnemius (Fig. 5). The effect of SCD1 deficiency on the AMPK pathway is thus similar to that observed after leptin treatment, because leptin was shown to activate AMPK in oxidative muscles (soleus and red gastrocnemius) but not in glycolytic muscles (white gastrocnemius or extensor digitorum longus) (29). However, the activation of AMPK in SCD1−/− mice seems to be leptin independent, because increased AMPK phosphorylation, higher rates of CPT I activity, and decreases in the rate of ceramide synthesis were still observed in the oxidative muscles of abJ/abJ;ob/ob double-mutant mice (Fig. 6). In the present report, we show for the first time that SCD1 deficiency mimics the effect of leptin on the AMPK pathway and β-oxidation in the skeletal muscle of ob/ob mice (Fig. 6). This finding raises the possibility that SCD1 may be a downstream component of the leptin-signaling pathway in muscle, as it was previously shown to be in liver (9, 24). Further studies are necessary to determine the role of SCD1 in leptin signaling in skeletal muscle.

In summary, the results presented here show that SCD1 deficiency reduces ceramide synthesis in oxidative muscles through downregulation of SPT and increased β-oxidation. As shown schematically in Fig. 7, disruption of the SCD1 gene leads to activation of AMPK, which increases CPT I activity and enhances FA-CoA transport into the mitochondria for β-oxidation. Increased β-oxidation causes a drop in the content of intracellular FA-CoAs, including palmitoyl-CoA, which is required for the first step of de novo ceramide synthesis and regulates SPT activity (17, 20). Thus low SPT activity, together with the decreased level of intracellular FA-CoAs, accounts for the decreased ceramide formation in the oxidative muscle of SCD1−/− mice. In addition, SCD1 deficiency leads to downregulation of gene expression of both SPT subunits (LCB1 and LCB2). The mechanisms by which loss of SCD1 function activates AMPK and downregulates SPT gene expression are currently unknown. The data suggest that the antilipotoxic action of SCD1 deficiency on skeletal muscle might be at least partially mediated by reduction in ceramide synthesis. It remains to be determined whether SCD1 deficiency protects against ceramide-induced lipotoxicity in other tissues such as the heart and pancreatic islets.

Fig. 7.

Proposed model of the effect of SCD1 deficiency on ceramide synthesis in oxidative skeletal muscles. Activation of AMPK in the muscles of SCD1−/− mice increases CPT I activity and enhances mitochondrial β-oxidation. A drop in palmitic acid content reduces SPT activity, which, together with the decreased level of intracellular FA-CoAs, accounts for the decreased ceramide formation. SCD1 deficiency also leads to a downregulation of SPT gene expression. The mechanism of AMPK activation by SCD1 deficiency is currently unknown.

GRANTS

This work was supported by National Institutes of Health Grants RO1-DK-62388 (J. M. Ntambi), RO1-DK-41096 (J. M. Friedman), and MSTP-GM-07739 (P. Cohen); American Heart Association Postdoctoral Fellowship 0420051Z (A. Dobrzyn); Wellcome Trust Grant 065565; and European Commission Grant QLG1-CT-2001-01488 (D. G. Hardie).

Acknowledgments

We thank Dr. Jeffrey M. Peters (The Pennsylvania State University) for providing the probes for CPT I.

Footnotes

  • 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.

REFERENCES

View Abstract