Endothelial Nitric Oxide Synthase (eNOS) Knockout Mice Have Defective Mitochondrial β-Oxidation

Experiments were carried out under protocols approved by the Institutional Animal Care and Use Committee. eNOS^−/− and C57BL6 mice, as previously described, were used (1). Male knockout (eNOS^−/−) and control mice (eNOS^+/+) were generated by mating heterozygous animals from our colony. Mice of generations 9–12 were used for our studies. Mice were fed a normal diet (UAR, Epinay sur Orge, France; energy content: 12% fat, 28% protein, and 60% carbohydrate, low nitrates). Throughout the study period, the mice were housed with lights on from 7:00 a.m. to 7:00 p.m., with food and water ad libitum.

A tail fragment was digested during 12 h at 55°C, by 30 μl proteinase K (Sigma-Aldrich, Buch, Switzerland), in 720 μl digestion buffer (50 mmol/l Tris HCL, pH 8, 100 mmol/l EDTA, 1% SDS). After the addition of 5 mol/l NaCl (250 μl), the solution was centrifuged 10 min at 14,000g. The supernatant was incubated 30 min at −20°C with 600 μl isopropanol and centrifuged for 10 min at 14,000g. The DNA was washed with 70% ethanol and suspended in 100 μl water. The amplification by PCR was made with three oligonucleotides (oligo94: TGGCTACCCGTGATATTGCT, oligo1823: ATTTCCTGTCCCCTGCCTTC, oligo1824: GGCCAGTCTCAGAGCCATAC) designed by The Jackson Laboratory (Bar Harbor, ME).

For body composition (percent fat), mice were killed by cervical dislocation, and the whole carcasses were incised, dried to a constant weight at 70°C, and subsequently homogenized. Total body fat content was determined by the Soxhlet extraction method using petroleum benzine as previously described (11). Results are presented as absolute weight (grams) and as percentage of total body weight.

Animals undergoing indirect calorimetry were acclimatized to the respiratory chambers for 12 h before the gas exchange measurements. Mice were individually housed in the calorimeter cages, and data on gas exchanges, movements, physical activity, and food intake were collected for 36 h. Indirect calorimetry was performed with a computer-controlled Oxymax open-circuit calorimetry system (Columbus Instruments, Columbus, OH). Each chamber was equipped with a water bottle, a food tray connected to a balance, and an activity monitor. Oxygen consumption (Vo₂) and carbon dioxide production (Vco₂) were measured for each mouse at 6-min intervals. The energy expenditure was calculated as follows: energy expenditure = (3.815 + 1.232 × Vco₂/Vo₂) × Vo₂, where Vco₂ is the expired Co₂ volume (ml · kg⁻¹ · h⁻¹), and Vo₂ is the expired O₂ volume (ml · kg⁻¹ · h⁻¹). Mice were weighted before each recording, and energy expenditure was corrected for body weight. Ambulatory activity of individually housed mice was assessed simultaneously using photo beams. Consecutive photo-beam breaks occurring in adjacent photo beams were scored as an ambulatory movement. Cumulative ambulatory activity counts were recorded every 6 min throughout the light and dark cycles.

During the 36 h period of recording, mice display phases with almost no movement alternated with phases of increased ambulatory activity. To be able to better compare energy expenditure between eNOS^−/− and control mice during these different activity phases, we report energy expenditure according to increasing levels of physical activity divided into four quartiles (0–50, 50–200, 200–400, and 400–800 movements/h).

For measurement of mitochondrial DNA content (mtDNA), a marker of total mitochondrial number, total DNA was isolated from skeletal muscle by proteinase K digestion followed by a phenol extraction as previously described (12). The mtDNA and the genomic DNA were measured by quantitative RT-PCR (see below) with primers designed to target the subunit I of cytochrome-c oxidase (mtDNA) and the DNA telomerase (genomic DNA). Results are given as the ratio between mtDNA and genomic DNA.

Mitochondria were prepared from muscle as previously described by Jimenez et al. (13). In brief, muscle (gastrocnemius, 100–150 mg) was minced with scissors and homogenized using a Teflon pestle in 8 ml ice-cold homogenization buffer containing 100 mmol/l sucrose, 180 mmol/l KCl, 10 mmol/l EDTA, 5 mmol/l MgCl₂, 50 mmol/l Tris/HCl, pH 7.4, and 0.06% protease inhibitor cocktail (Sigma-Aldrich). The homogenate was centrifuged at 1,600g for 10 min at 4°C. The supernatant was centrifuged at 9,200g for 15 min at 4°C, and the resulting subsarcolemmal mitochondrial pellet was suspended in 0.25 mol/l sucrose buffer. Mitochondrial protein concentration was determined according to Bradford (14).

Mitochondrial β-oxidation was determined using [1-¹⁴C]palmitoyl-CoA (Perkin Elmer, Schwerzenbach, Switzerland) as a substrate (15). The assay medium (220 μl) contained 9.1 mmol/l HEPES (pH 7.2), 100 mmol/l KCl, 0.1 mmol/l EDTA, 0.9 mmol/l Na₂PO₄, 0.9 mmol/l dithiothreitol, 0.9 mmol/l ADP, 0.9 mmol/l NAD⁺, 4.5 mmol/l malonate, 0.1 mmol/l l-carnitine, 0.18 mmol/l coenzyme A, 1% BSA, and 36 μmol/l palmitoyl-CoA (all products were supplied by Sigma-Aldrich). Palmitoyl-CoA oxidation was measured with 0.01 μCi [1-¹⁴C]palmitoyl-CoA. The part of β-oxidation induced by peroxisomal contamination in subsarcolemmal mitochondrial extract was assessed for each assay by the addition of 0.5 mmol/l potassium cyanide. The peroxisomal β-oxidation (cyanide insensitive) was then subtracted from the total β-oxidation (peroxisomal and mitochondrial) measured in the same manner for each sample. All samples were preincubated for 3 min at 37°C. The oxidation reaction was initiated by the addition of 30 μg subsarcolemmal mitochondrial proteins. After 10 min, the oxidation was stopped by addition of 200 μl HClO₄ 600 mmol/l. The total solution was then centrifuged at 3,000 rpm for 10 min. Aliquots of 200 μl were transferred to a scintillation vial containing 10 ml liquid scintillation cocktail (OptiPhase “HiSafe”3; Perkin Elmer, Schwerzenbach, Switzerland) and assayed for radioactivity in a liquid scintillation counter (Wallac 1409; Perkin Elmer, Huenenberg, Switzerland). All assays were normalized per cytochrome-c oxidase activity.

To assess whether NO could have a direct effect on mitochondrial activity, we measured β-oxidation in isolated mitochondria, obtained from control mice, after incubation with substrates or inhibitors of NOS. Specifically, we incubated isolated skeletal muscle mitochondria either with 500 μmol/l l-arginine (the substrate of eNOS) or 500 μmol/l of d-arginine as a control for the specificity of the l-arginine effect, 500 μmol/l of N^G-nitro-l-arginine methyl ester (l-NAME) (a NOS inhibitor), and 500 μmol/l of l-NAME concomitantly with 500 μmol/l of l-arginine. Mitochondria were incubated for 10 min at room temperature in each of these conditions, before starting the β-oxidation assay.

Cytochrome-c oxidase activity was measured by following the oxidation of reduced cytochrome-c, as previously described by Trounce et al. (16). Mitochondrial protein (10 μg) and reduced cytochrome-c (10 mmol/l; Sigma-Aldrich) were added to KH₂PO₄ (10 mmol/l, pH 7.4) and sucrose (0.25 mol/l). The cytochrome-c oxidase activity was calculated from the initial linear rate of cytochrome-c oxidation at 550 nm (WPA S2000; Biolabo Scientific, Chatel St. Denis, Switzerland).

Total NOS activity was quantified by measuring the conversion of [³H]l-arginine to [³H]l-citrulline in gastrocnemius muscle as described previously, and neuronal NOS (nNOS) expression was quantified by quantitative RT-PCR (see below) (17).

Lipolysis in epididymal adipose tissue was assessed by measuring lipoprotein lipase (LPL) activity and hormone-sensitive lipase expression. LPL activity was determined as previously described by Olivecrona et al. (18). Results are given as activity per quantity of LPL (units/mg). Hormone-sensitive lipase expression was quantified by real-time quantitative RT-PCR (see below).

Mitochondrial breakdown in skeletal muscle was assessed by measuring the expression levels of proapoptotic (BAD, BAX) and antiapoptotic (BCL2, BCLX) genes using quantitative RT-PCR (see below). In addition, we also measured plasma levels of IL-6, IL-10, IFN-γ, and tumor necrosis factor using BD Cytometric Bead Array Flex Set System combined with a BD FACSArray Bioanalyzer System (Becton Dickinson, Basel, Switzerland) to assess whether high levels of these cytokines may induce muscle wasting and decrease mitochondrial content in skeletal muscle in eNOS^−/− mice.

Total RNA was isolated from skeletal muscle (hindlimb immediately frozen in liquid nitrogen and stored at −80°C after dissection) with TriPure isolation reagent (Roche Diagnostics, Rotkreuz, Switzerland) and reverse transcribed, after RNase free DNAse I (Roche Diagnostics) treatment, with M-MLV reverse transcriptase (Promega, Catalys, Wallisellen, Switzerland) to cDNA. Real-time PCR was performed with the Light Cycler detection system (Roche Diagnostics) and the QuantiTect SYBR Green PCR kit (Qiagen, Hombrechtikon, Switzerland). The housekeeping gene encoding β-actin was chosen as an internal control.

For triglyceride estimation, the frozen muscle sample (100–150 mg) was powdered under liquid nitrogen with a pestle and mortar. A total of 6 ml chloroform-methanol (2:1) solution and 1 ml NaCl 0.9% were added, and the organic and aqueous phases were separated by centrifugation at 5,000g for 10 min at 4°C (19). The aqueous phases were removed, and the organic phase was dried under N₂ at 40°C. Lipid extracts were dissolved in chloroform (1 ml) and applied to an aminopropyl silica column (Bond Elut NH₂; Varian International, Zug, Switzerland) to remove phospholipids. Lipid extracts containing triglyceride were eluted by chloroform (four times, 1 ml), dried under N₂ at 40°C, and dissolved in chloroform (1 ml).

The triglyceride content of lipid extracts was determined by a colorimetric method involving the quantification of glycerol after triglyceride saponification (20). In brief, a 200 μl lipid sample was dried under N₂ at 40°C and resuspended in 2.6 ml isopropanol/water (9:1) and 5% KOH. After 30 min saponification at 70°C, 1 ml NaIO₄ 0.03 M/isopropanol/acetate 1 N (1:1.7:5.7) and 0.5 ml acetylacetone/isopropanol/ammonium acetate (1:3.3:15.4) were added. The mixture was incubated for 30 min at 50°C, and the absorbance was read at 405 nm. To assess the intramyocellular triglyceride content (micromoles triglyceride per milligram tissue), a tripalmitate solution was used to determined the standard curve.

Statistical analysis was performed with the two-tailed t test for single comparisons. A P value <0.05 was considered to indicate statistical significance. Data are given as means ± SE.

Body weight (28.5 ± 0.5 vs. 28.8 ± 0.6 g), percent of body fat (10.58 ± 1.34 vs. 9.53 ± 1.01%), and total food intake (27.6 ± 2.3 vs. 25.6 ± 0.9 mg · h⁻¹ · g⁻¹) were comparable in control (eNOS^+/+) and eNOS^−/− mice throughout the study period (P > 0.1 for all comparisons, n = 6) (Table 1). Under basal conditions (food and water ad libitum), the mean energy expenditure (0.013 ± 0.001 vs. 0.014 ± 0.001 kcal · h⁻¹ · g⁻¹, P > 0.1, n = 6) measured during a 36 h period was comparable in the two groups of mice, whereas the ambulatory activity (110 ± 5 vs. 168 ± 6 movements/h, P < 0.05, n = 6) was significantly higher in eNOS^−/− (Table 1 and Fig. 1A and B). However, when analyzed according to similar levels of physical activity, energy expenditure was consistently lower in eNOS^−/−: 0–50 movements/h: 0.011 ± 0.001 vs. 0.013 ± 0.001 kcal · h⁻¹ · g⁻¹, 50–200 movements/h: 0.014 ± 0.001 vs. 0.015 ± 0.001 kcal · h⁻¹ · g⁻¹, 200–400 movements/h: 0.016 ± 0.001 vs. 0.017 ± 0.001 kcal · h⁻¹ · g⁻¹ (P < 0.05, for all comparisons) (Fig. 1C).

Since in the fasting state, whole-body energy production depends on mitochondrial activity, which is reflected by oxygen utilization, we measured oxygen consumption (VO₂) in eNOS^−/− and control mice during food deprivation. The mean oxygen consumption was significantly lower in eNOS^−/− mice than in control mice (2,530 ± 137 vs. 2,919 ± 88 ml · kg⁻¹ · h⁻¹, P < 0.05, n = 6) (Fig. 1D).

To determine whether in eNOS^−/− mice the lower oxygen consumption in the fasting state was related to defective mitochondrial biogenesis or impaired fatty acid–dependent energy production in mitochondria, we assessed mitochondrial content in muscle tissue and measured β-oxidation in isolated mitochondria. We found that in eNOS^−/− mice, mitochondrial protein content in gastrocnemius muscle was roughly 30% lower than in control mice (2.0 ± 0.2 vs. 2.8 ± 0.2 μg/mg, P < 0.05, n = 8) and tended to be lower in heart muscle (5.2 ± 0.2 vs. 6.0 ± 0.4 μg/mg, P = 0.06, n = 8) (Fig. 2A). In gastrocnemius muscle, the decreased mitochondrial protein content was paralleled by a significant decrease of the mtDNA-to-DNA ratio (2,436 ± 116 vs. 3,180 ± 200, P < 0.05, n = 8). We next assessed β-oxidation in isolated mitochondria and found that it was roughly 30% lower in mitochondria harvested from eNOS^−/− than from control mice (gastrocnemius: 0.037 ± 0.003 vs. 0.025 ± 0.002 mmol · mg⁻¹ · min⁻¹, P < 0.05, n = 6; heart: 0.031 ± 0.003 vs. 0.021 ± 0.002 mmol · mg⁻¹ · min⁻¹, P < 0.05, n = 6) (Fig. 2B).

To provide additional evidence that in eNOS^−/− mice the decreased mitochondrial β-oxidation was related to the lack of eNOS activity, we measured in vitro β-oxidation in mitochondria isolated from skeletal muscle of control mice, while modulating NO availability. Figure 3 shows that an NO donor and the NOS substrate l-arginine increased β-oxidation (control: 0.042 ± 0.002 mmol · mg⁻¹ · min⁻¹, l-arginine: 0.080 ± 0.003 mmol · mg⁻¹ · min⁻¹, DETA-NONOate: 0.097 ± 0.001 mmol · mg⁻¹ · min⁻¹, P < 0.0001 vs. control for all, n = 3), whereas d-arginine had no detectable effect (0.044 ± 0.002 mmol · mg⁻¹ · min⁻¹, P > 0.1 vs. control, n = 3), indicating the specificity of the effect observed with l-arginine. Conversely, l-NAME, an inhibitor of NOS, inhibited mitochondrial oxidation by roughly 50% (l-NAME: 0.022 ± 0.001 mmol · mg⁻¹ · min⁻¹, P < 0.0001 vs. control, n = 3), and this effect was reversed by the addition of l-arginine (l-NAME + l-arginine: 0.045 ± 0.002 mmol · mg⁻¹ · min⁻¹, P > 0.1 vs. control, n = 3). These data suggest that the observed alterations in mitochondrial β-oxidation in eNOS^−/− mice are at least in part directly related to decreased nitric oxide availability.

To ascertain that eNOS^−/− mice had indeed decreased NOS activity in skeletal muscle and that there was no compensation by overexpression of other NOS isoforms, we measured total NOS activity and nNOS expression in skeletal muscle (gastrocnemius) of control and eNOS^−/− mice. We found that NOS activity was significantly lower in muscle from eNOS^−/− mice (1.12 ± 0.04 vs. 1.31 ± 0.07 pmol · mg⁻¹ · min⁻¹, P < 0.05, n = 5) (Fig. 4), whereas nNOS expression was comparable in the two groups of mice (118 ± 6 vs. 100 ± 11%, n = 12, P > 0.1).

LPL activity and hormone-sensitive lipase expression levels in adipose tissue were similar in the two groups (LPL activity: 40.1 ± 4.1 vs. 35.0 ± 3.6 units/mg, P > 0.1, n = 8; hormone-sensitive lipase expression: 100 ± 5 vs. 90 ± 4%, P > 0.1, n = 9). These data suggest that increased FFA levels in eNOS^−/− mice were not related to increased lipolysis in adipose tissue.

The expression levels of proapoptotic and anti-apoptotic genes and circulating cytokine levels were comparable in eNOS^−/− mice and control mice (proapoptotic genes: BAD, 100 ± 1 vs. 99 ± 2%, BAX, 100 ± 3 vs. 90 ± 5%, P > 0.1 for all comparisons, n = 12; anti-apoptotic genes: BCL2, 100 ± 3 vs. 98 ± 2%, BCLX, 100 ± 3 vs. 88 ± 4%, P > 0.1 for all comparisons, n = 12; IL-6, 15.6 ± 1.4 vs. 13.7 ± 1.7 pg/μl, IL-10, 2.6 ± 0.3 vs. 2.3 ± 0.3 pg/μl, IFN-γ, 40.4 ± 5.1 vs. 38.1 ± 5.3 pg/μl, tumor necrosis factor, 52.5 ± 0.2 vs. 52.2 ± 5.8 pg/μl, P > 0.1 for all comparisons, n = 9). Thus, decreased mitochondrial content did not appear to be related to increased mitochondrial breakdown or muscle wasting.

NO is known to modulate the expression of genes at mRNA levels. We therefore quantified (by quantitative RT-PCR) a cluster of genes involved in intracellular FFA metabolism in eNOS^−/− and control mice. To simplify the presentation of the results, genes of interest were separated in three groups, according to their metabolic function: fatty acid transport (FATP, CD36, FABP3), neolipogenesis (ACCβ, FAS, MCD), and β-oxidation (CPT-1, VLCAD, FACoA) (Fig. 5).

FATP and FABP3 expression in skeletal muscle was roughly 40% lower in eNOS^−/− than in control mice (P < 0.05), whereas CD36 expression was similar in the two groups (Fig. 5A).

ACCβ and FAS were increased by 60% (P < 0.05) and 220% (P < 0.05), respectively, whereas MCD was decreased by 50% (P < 0.05) in eNOS^−/− mice (Fig. 5B).

CPT-1 and VLCAD were decreased by 30% (P < 0.05) and 40% (P < 0.05), respectively, in eNOS^−/− mice, whereas FACoA was comparable in the two groups (Fig. 5C).

To test for the functional consequences of defective fatty acid oxidation, we assessed intramyocellular triglyceride content in skeletal muscle of eNOS^−/− and control mice. We found that the intramyocellular triglyceride content was roughly 33% higher in eNOS^−/− mice than in control mice (12.05 ± 0.60 vs. 8.99 ± 0.60 μmol/g, P < 0.05, n = 12) (Fig. 6).

Using mice deficient for the eNOS gene, we have assessed the role of eNOS-derived NO on energy metabolism, oxygen consumption, and β-oxidation. We found that eNOS^−/− mice had markedly lower energy expenditure and oxygen consumption than control mice. This impairment of energy expenditure and oxygen consumption in eNOS^−/− mice was associated with a roughly 30% decrease of the mitochondrial content in cardiac and skeletal muscle, with a decreased total NOS activity in skeletal muscle, and, most importantly, with mitochondrial dysfunction, as evidenced by a >30% lower β-oxidative activity in isolated mitochondria. This defect of mitochondrial biogenesis and β-oxidative function was associated with a roughly 30% greater intramyocellular lipid content in eNOS^−/− mice. These data show, for the first time, that eNOS-derived NO is not only an important determinant of mitochondrial biogenesis, but also of energy expenditure and mitochondrial FFA β-oxidation.

Increased FFA levels in eNOS^−/− mice prompted us to study the role of eNOS-derived NO in the regulation of lipid oxidation and metabolism, since we had no evidence of increased lipolysis in adipose tissue of these animals (1). In the fasting state, fatty acids are the main metabolic substrate for energy production in skeletal muscle. In a first step, we therefore assessed fatty acid oxidation by indirect calorimetry. We found that oxygen consumption in the fasting state, an index of mitochondrial activity, was significantly lower in eNOS^−/− mice than in control mice. This decrease in energy consumption was seen despite the fact that eNOS^−/− mice had significantly increased ambulatory activity. The underlying mechanism for the increased locomotion in eNOS^−/− mice needs further study but could represent an adaptive response to decreased energy consumption or be related to a central neural effect of eNOS deletion (21). In line with this latter concept, chronic NO inhibition increases locomotion activity in rats (22). Energy expenditure, normalized for ambulatory activity, was >30% lower in eNOS^−/− mice than in control mice, suggesting decreased FFA oxidation in skeletal muscle of eNOS^−/− mice.

In a next step, we therefore turned our attention to mitochondria, the primary site of FFA oxidation. Decreased β-oxidation in eNOS^−/− mice could be related either to a decrease of the number of mitochondria and/or defective β-oxidative capacity of the mitochondria. We found that eNOS-derived NO synthesis is an important determinant of mitochondrial quantity, as evidenced by a markedly lower mitochondrial protein content and mtDNA-to-DNA ratio in skeletal muscle of eNOS^−/− mice. This decreased mitochondrial content was not related to mitochondrial breakdown or muscle wasting. This is in line with previous findings by Nisoli et al. (9,10) who have shown that NO produced by eNOS is important for mitochondrial biogenesis through a PGC1α-mediated mechanism. Moreover, this impairment of mitochondrial biogenesis is associated with a smaller size of mitochondria. Since mitochondrial size is a determinant of mitochondrial function, the impaired FFA oxidation in eNOS^−/− mice could be related, at least in part, to smaller mitochondria. Alternatively, there is evidence that NO deficiency may alter mitochondrial function independently of mitochondrial biogenesis and size (9,10). To test for this possibility, we directly assessed β-oxidation in mitochondria isolated from heart and skeletal muscle of eNOS^−/− and control mice. We found that the β-oxidative activity in mitochondria isolated from heart and skeletal muscle was >30% lower in eNOS^−/− mice than in control mice. It is important to note that these direct measurements in isolated mitochondria allowed us to study β-oxidation in the absence of confounding factors, such as FFA transport across the cell membrane. Moreover, β-oxidation was normalized for cytochrome-c activity and mitochondrial protein content, allowing us to assess β-oxidative function independently of mitochondrial size. Thus, eNOS deficiency in mice causes defective mitochondrial β-oxidation that is independent of the size of mitochondria. Taken together, these data suggest that in eNOS^−/− mice, defective FFA use is related to both decreased mitochondrial size and an intrinsic defect of mitochondrial β-oxidative activity.

The underlying mechanism by which NO deficiency causes defective β-oxidation is not known yet, but there is evidence that NO regulates CPT-1 activity, a mitochondrial enzyme and rate-limiting step of fatty acid β-oxidation (23,24). The present data provide additional evidence that NO may directly regulate mitochondrial activity by showing that the NOS substrate l-arginine and the NO donor DETA-NONOate increased β-oxidation in isolated mitochondria from control mice, whereas the NOS inhibitor l-NAME significantly inhibited this activity. This rapid and direct effect of NO may be related to modifications of the activity of enzymes involved in mitochondrial β-oxidation. In addition, the quantitative PCR studies provide evidence for a dysregulation of genes involved in fatty acid metabolism in skeletal muscle of eNOS^−/− mice. This included a downregulation of genes involved in the regulation of fatty acid transport and β-oxidation, such as FATP, FABP3, CPT-1, VLCAD, FACoA, and an upregulation of mitochondrial genes (ACCβ and FAS) implicated in fatty acid uptake inhibition. These data suggest that alterations in the expression of genes involved in fatty metabolism may contribute to hyperlipidemia in eNOS^−/− mice, even though we have not directly measured the enzymatic activity of these genes. In line with this hypothesis, NO has been shown to modulate the expression of an increasing number of genes by regulating the activity of transcription factors (25). Whereas no such information is available for genes studied in the present experiments, it appears plausible that this may represent a mechanism underlying alteration in the expression levels of genes involved in lipid metabolism in eNOS^−/− mice. Taken together, these data suggest that the lack of eNOS-derived NO in eNOS^−/− mice contributes directly to the observed alterations in lipid metabolism. Finally, we cannot rule out the alternative possibility that an adaptive phenomenon consecutive to the reduction in β-oxidation and mitochondrial content in skeletal muscle may also play a role. This is suggested by studies showing that the expression of genes involved in FFA β-oxidation is correlated with the β-oxidative activity of the tissue (26,27). Further studies are needed to determine the exact underlying mechanisms by which eNOS-derived NO mediates the regulation of these genes.

In conclusion, the present studies indicate that the lack of eNOS-derived NO in insulin-resistant skeletal muscle of eNOS^−/− mice is associated with a reduction of mitochondrial content and defective fatty acid β-oxidation. These defects were coupled with the dysregulation of several genes involved in cellular fatty acid metabolism. As a consequence of these alterations of mitochondrial function, we found a marked accumulation of intramyocellular triglycerides in eNOS^−/− mice.

This work was supported by the Placide Nicod Foundation.

We would like to thank Dr. Josianne Seydoux for performing the studies for body composition measurement and Dr. Gunilla Olivecrona for LPL activity assessment in eNOS^−/− mice.