DGAT1 deficiency decreases PPAR expression and does not lead to lipotoxicity in cardiac and skeletal muscle.

Diacylglycerol (DAG) acyl transferase 1 (Dgat1) knockout (−/−) mice are resistant to high-fat-induced obesity and insulin resistance, but the reasons are unclear. Dgat1−/− mice had reduced mRNA levels of all three Ppar genes and genes involved in fatty acid oxidation in the myocardium of Dgat1−/− mice. Although DGAT1 converts DAG to triglyceride (TG), tissue levels of DAG were not increased in Dgat1−/− mice. Hearts of chow-diet Dgat1−/− mice were larger than those of wild-type (WT) mice, but cardiac function was normal. Skeletal muscles from Dgat1−/− mice were also larger. Muscle hypertrophy factors phospho-AKT and phospho-mTOR were increased in Dgat1−/− cardiac and skeletal muscle. In contrast to muscle, liver from Dgat1−/− mice had no reduction in mRNA levels of genes mediating fatty acid oxidation. Glucose uptake was increased in cardiac and skeletal muscle in Dgat1−/− mice. Treatment with an inhibitor specific for DGAT1 led to similarly striking reductions in mRNA levels of genes mediating fatty acid oxidation in cardiac and skeletal muscle. These changes were reproduced in cultured myocytes with the DGAT1 inhibitor, which also blocked the increase in mRNA levels of Ppar genes and their targets induced by palmitic acid. Thus, loss of DGAT1 activity in muscles decreases mRNA levels of genes involved in lipid uptake and oxidation.

centrations in lipid extracts were determined enzymatically with colorimetric kits (Sigma-Aldrich). DAG and ceramide levels were measured using a DAG kinase-based method. Lipids extracted from muscle specimens were dried under a stream of N 2 , redissolved in 7.5% octyl-␤ -D-glucoside containing 5 mM cardiolipin and 1 mM diethylenetriamine pentaacetate, in which DAG and ceramide are quantitatively phosphorylated to form 32 P-labeled phosphatidic acid and ceramide-1-phosphate, respectively, which were then quantifi ed. The reaction was carried out at room temperature for 30 min in 100 mM imidazole HCl, 100 mM NaCl, 25 mM MgCl 2 , 2 mM EGTA (pH 6.6), 2 mM DTT, 10 g/100 l DAG kinase (Sigma-Aldrich), 1 mM ATP, and 1 Ci/100 l [ ␥ -32 P]ATP. The reaction was stopped by addition of chloroform/methanol (v/v, 2:1) and 1% HClO 4 , and lipids were extracted and washed twice with 1% HClO 4 . Lipids were resolved by TLC (Partisil K6 adsorption TLC plates, Whatman catalog no. LK6D); mobile phase contained chloroform/methanol/acetic acid (v/v/v, 65: 15:5). The bands corresponding to phosphatidic acid and ceramide-1-phosphate were identifi ed with known standards, and silicon was scraped into a scintillation vial for radioactivity measurement. [ 3 H] triolein bands from the same TLC plates were identifi ed and quantifi ed in the same way and were used as controls for lipid recovery.

DGAT activity assay
DGAT activity was measured in membrane fractions isolated from muscle specimens using [ 14 C]labeled palmitoyl-CoA as previously described ( 5 ). The amounts of both DAG and fatty acyl-CoA in the reaction mixture were in excess, and the reaction rate was of the fi rst order with respect to the amount of input DGAT activity to be assayed.

Real-time PCR
Total RNA was extracted using a Trizol kit from Invitrogen. An amount of 1 g of RNA was initially treated with DNase I (Invitrogen). The RNA samples were then reverse-transcribed using the SuperScript III First-Strand Synthesis System for RT-PCR (Invitrogen). Real-time amplifi cation was performed using iQ SYBR Green Supermix (Bio-Rad). Primers used for PCR amplifi cation are listed in supplementary Table I. Analysis was performed using iCycler iQ Real-Time Detection System software (Bio-Rad).

Dual energy X-ray absorptiometry
The dual energy X-ray absorptiometry (DEXA) machine was calibrated before use. WT, Dgat1 +/ Ϫ , and Dgat1 Ϫ / Ϫ mice were anesthetized with intraperitoneal injection of ketamine/xylazine (86.98 mg/kg and 13.4 mg/kg, respectively), and whole body measurements, excluding the head, were made (DEXA, PIXImus Lunar-GE). Measurements included body weight, lean mass, fat mass, and percentage of fat/body weight. One person performed all scans, and mice were always in the prostrate position on the imaging-positioning tray.

Exercise protocol
Mice were forced to swim in tanks. The swimming protocol was a modifi cation of the procedure used by Ryder et al. ( 13 ). Water temperature was maintained at 34-35°C. Mice swam for 10 min sessions twice a day separated by a 10 min break. Sessions were increased by 10 min each day until the fourth day; mice were then allowed to swim for six 30 min intervals separated by 10-15 min rest periods. After the last swim interval, mice were dried and put back in their cages for 18-20 h.

Western blotting
An amount of 10-30 mg of tissue was homogenized in lysis buffer containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% In fact, this alteration in lipid metabolism appears to mimic that found with athletic training, in which TG stores increase, fatty acid oxidation increases, and insulin sensitivity improves ( 7,8 ). Overexpression of DGAT1 in cardiomyocytes increased TG and decreased levels of toxic lipids, such as DAG, ceramide, and free fatty acids ( 6 ). When a cardiomyocyte-expressing human DGAT1 transgene was crossed onto a heart failure model overexpressing acyl CoA synthetase 1, heart function and survival rate were signifi cantly improved ( 6 ). Similarly, overexpression of DGAT1 in macrophages alleviated fatty acid-induced infl ammation and high-fat feeding-induced insulin resistance ( 9 ).
Despite these seemingly benefi cial actions of DGAT1 overexpression, its deletion does not lead to an opposite phenotype. Dgat1 knockout ( Dgat1 Ϫ / Ϫ ) mice have reduced diet-induced obesity and less insulin resistance ( 10 ). Greater insulin sensitivity was also found in skeletal muscle in these mice ( 22 ); however, the reasons remain unclear. Other experiments showed that the transplantation of Dgat1 Ϫ / Ϫ white adipose tissue decreases adiposity and enhances glucose disposal in wild-type (WT) mice; adiponectin secreted by adipose tissue was postulated to contribute to the increased energy expenditure and insulin sensitivity in Dgat1 Ϫ / Ϫ mice ( 11,12 ). This observation has spurred pharmaceutical development of DGAT1 inhibitors. However, the effect of loss of DGAT1 activity in other important organs, such as the heart, has not been reported. We hypothesized that Dgat1 Ϫ / Ϫ mice would have reduced conversion of DAG to triglyceride. This might lead to accumulation of DAG, ceramide, and fatty acids in cardiac muscle, which could lead to cardiomyopathy, especially in the setting of high-fat diets.
In this study, we describe the effects of acute and chronic loss of DGAT1 activity in muscles. We show that loss of this enzyme modulates counter-regulatory pathways that reduce muscle lipid and increase glucose uptake. This fi nding suggests that DGAT inhibitors may be of benefi t in the treatment of reperfusion injury.

Mice and diet
Animal protocols were in compliance with accepted standards of animal care and were approved by the Columbia University Institutional Animal Care and Use Committee. Male mice were used in experiments unless otherwise indicated. WT, C57BL/6J, and Dgat1 Ϫ / Ϫ C57BL/6J mice ( 10 )

Palmitic acid and DGAT1i treatment of myocytes
AC16 human cardiomyocytes and C2C12 myoblasts were grown in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin at 37°C. C2C12 myoblasts were maintained below 60% confl uence to prevent confl uence-induced differentiation. For experiments, cells were allowed to become confl uent, and the media were supplemented with 2% horse serum instead of FBS for myotube formation. After 4-5 days of differentiation, myotubes were formed and used for experiments. AC16 human cardiomyocytes were changed to 2% FBS 24 h before experiments. Each experiment was performed on three separate occasions. Palmitic acid was administered to cells as a conjugate with 1.25% (w/v) fatty acid-free BSA at the fi nal concentration of 0.5 mM for 12 h followed by 4 h of incubation in the presence or absence of DGAT1i (5 mM). Both cells were washed two times with ice-cold PBS prior to RNA isolation and quantitative realtime PCR.
To determine palmitic acid uptake, AC16 cardiomyocytes were pretreated with 5 mM DGAT1i for 4 h and then incubated for 15 min with 0.5 mM palmitic acid with 1 Ci/ml of 9,10-[ 3 H]palmitic acid and 1.25% BSA. Cells were washed three times with icecold PBS prior to 0.1N Na 2 OH addition. An aliquot of homogenate was taken for radioactivity measurement.

Statistics
All data are presented as mean ± SD, unless indicated otherwise. Comparisons between two groups were performed using unpaired two-tailed Student's t -test with Statistica version 6.0 (StatSoft). A P value less than 0.05 was used to determine statistical signifi cance.

Dgat1 ؊ / ؊ mice have decreased heart and skeletal muscle DGAT activity
To determine if loss of Dgat1 in heart and skeletal muscle reduces or eliminates DGAT function, we measured DGAT activity in the heart and skeletal muscle tissue of Dgat1 +/ Ϫ and Dgat1 Ϫ / Ϫ mice. DGAT activity was decreased ‫ف‬ 20% and ‫ف‬ 65% in hearts of Dgat1 +/ Ϫ and Dgat1 Ϫ / Ϫ mice, respectively ( Fig. 1A ). Dgat2 mRNA did not change in these tissues ( Fig. 1B ). In skeletal muscle, DGAT activity was also reduced ( Fig. 1C ), but Dgat2 mRNA did not change signifi cantly ( Fig. 1D ). Dgat1 Ϫ / Ϫ mouse liver had a ‫ف‬ 57% decrease in total DGAT activity, but DGAT activity was 4-fold greater than skeletal muscle both under WT and Dgat1 Ϫ / Ϫ conditions. Therefore, Dgat1 Ϫ / Ϫ liver had an activity level that was greater than that found in WT soleus muscle ( Fig. 1E ). These animals also had no compensatory increase in Dgat2 mRNA ( Fig. 1F ). Plasma levels of lipids and glucose in these mice are shown in supplementary Table II. Nonidet P40, 0.5% sodium deoxycholate, and 1 protease inhibitor cocktail tablet (Roche)/25-50 ml. Cell lysates (50 g per sample) obtained after centrifugation at 15,000 g for 10 min at 4°C were resolved by SDS-PAGE and then transferred onto nitrocellulose membranes. Immunoblotting was carried out using the following primary antibodies: phospho-AKT (p-AKT), phospho-mTOR (p-mTOR), total AKT, total mTOR, glyceraldehyde-3phosphate dehydrogenase (GAPDH) (Cell Signaling), and PPAR ␣ , ␤ , and ␥ (Santa Cruz).

Echocardiography
Two-dimensional echocardiography was performed using a high-resolution imaging system with a 30-MHz imaging transducer (Vevo 770; VisualSonics) in unconscious 3-4-month-old mice. The mice were anesthetized with 1.5-2% isofl urane and thereafter maintained on 1-1.5% isofl urane throughout the procedure. Two-dimensional echocardiographic images were obtained and recorded in a digital format. Images were analyzed offl ine by a researcher blinded to the murine genotype. Left ventricular end-diastolic dimension (LVDd) and left ventricular endsystolic dimension (LVDs) were measured. Percentage fractional shortening, which quantifi es contraction of the ventricular wall and is an indication of muscle function, was calculated as % FS = ([LVDd Ϫ LVDs] / LVDd) × 100.

In vitro fatty acid oxidation and in vivo glucose and [ 14 C] triolein-VLDL uptake
Cardiac muscle fatty acid oxidation was measured with heart slices as described previously ( 6 ). Hearts were removed and immediately sliced across the ventricles forming rings ‫ف‬ 0.1-0.2 mm thick and usually 20-30 mg wet weight. The cardiac muscles were fi rst washed for 1 h at 37°C in Krebs-Henseleit bicarbonate buffer [10 mM HEPES (pH 7.35), 24 mM NaHCO 3 , 114 mM NaCl, 5 mM KCl, 1 mM MgCl 2 , and 2.2 mM CaCl 2 ] containing 8 mM glucose, 32 mM mannitol, and 0.1% BSA. The reaction was initiated by adding to the mix preconjugated BSA/fatty acid solution (final concentrations: 0.2 mM palmitate with 1 Ci/ml of 9,10-[ 3 H] palmitate and 1.25% BSA). The fatty acid oxidation product 3 H 2 O was determined in muscles incubated for 2 h at 37°C. An aliquot of the completed reaction mixture was transferred into an open microtube, which was placed inside a scintillation vial containing 0.5 ml of distilled water. The scintillation vial was then capped and incubated at 50°C for 18-24 h to reach vapor-phase equilibrium. After the transfer of 3 H 2 O from the reaction mixture to the scintillation vial (by water equilibration), the 3 H 2 O in the scintillation vial was scintillation-counted. The transfer and counting effi ciency was calibrated using a standard solution containing 0.1 Ci of 3 H 2 O for re-equilibration. Glucose uptake was measured as described ( 12 ). Basal glucose uptake was measured following an intravenous administration of 3 Ci of 2-deoxy-D-[1-3 H]glucose (PerkinElmer Life Sciences) in 4 h fasted mice. Blood was collected 0.5, 15, 30, and 60 min following injection. At 60 min, tissues were excised after the body cavity was perfused with 10 ml of PBS by cardiac puncture.
VLDL was labeled using cholesteryl ester transfer protein to incorporate [ 14 C]triolein into VLDL as described previously ( 14,15 ). After intravenous injection, plasma counts at 0.5, 5, 10, and 15 min were measured, and the mice were perfused with PBS as above. Radioactivity in the tissues was measured in a scintillation counter, and tissue uptake was normalized to recovered counts in the liver.

Acute treatment with DGAT1 inhibitor (DGAT1i)
Male 8-week-old mice fed a chow diet or male 14-week-old mice after 6 weeks of high-fat (Harlan TD 88137) feeding were gavaged with 200 l of vehicle (0.5% hydroxypropylmethyl contents of DAG and ceramide were reduced ‫ف‬ 16% and ‫ف‬ 22%, respectively, in hearts of high-fat-diet Dgat1 Ϫ / Ϫ mice. Heart and muscle weights were increased in Dgat1 ؊ / ؊ mice Heart weight was increased 29% ( P < 0.01) in chow diet and 7% ( P < 0.05) in high-fat diet Dgat1 Ϫ / Ϫ mice. Soleus and extensor digitorum longus (EDL) muscle weight increased 28% and 21%, respectively, in chow diet Dgat1 Ϫ / Ϫ DAG and ceramide levels were decreased in heart from Dgat1 ؊ / ؊ mice Surprisingly the content of TG was not signifi cantly decreased in cardiac muscle in Dgat1 Ϫ / Ϫ mice ( Table 1 ). Also unlike what we expected, the contents of TG precursors DAG and ceramide were not increased but, in fact, were decreased ‫ف‬ 20% in Dgat1 Ϫ / Ϫ hearts ( Table 1 ). We then fed the mice a high-fat diet. High-fat-diet control mice had an increase in TG levels ( ‫ف‬ 15%), DAG levels (26%), and FFA levels (37%) in the heart ( Table 1 ). As with chow, the Ϫ / Ϫ mice. A, C, and E: Total DGAT activity levels in membrane fractions of heart, soleus muscle, and liver from WT, Dgat1 Ϫ / Ϫ mice (n = 5, * P < 0.05 vs. WT, ** P < 0.01 vs. WT, # P < 0.05 vs. Dgat1 ). B, D, and E: Realtime-PCR quantifi cation of Dgat2 mRNA levels in heart, soleus muscle, and liver of WT, Dgat1

Dgat1
Ϫ / Ϫ hearts, mRNA levels were the same as those found with chow ( Table 2 ). Glucose transporter ( Glut ) 1 and 4 mRNA levels were upregulated 115% and 379% ( P < 0.05) in Dgat1 Ϫ / Ϫ hearts; this implies an energy supply shift from fatty acid consumption to glucose use. High-fat diets increased mRNA levels of a number of genes, but loss of Dgat1 signifi cantly reduced these genes ( Table 2 ). Skeletal muscle showed similar results to heart (supplementary Table III).

Dgat1 ؊ / ؊ mice have normal heart function Cardiac function of Dgat1
Ϫ / Ϫ mice evaluated by echocardiography was normal. We also exercised the Dgat1 Ϫ / Ϫ mice for two weeks to see if there was any change in heart function after exercise challenge. LVDd, LVDs, and fractional shortening were not signifi cantly altered in Dgat1 Ϫ / Ϫ mouse hearts before and after exercise ( Fig. 3A , B ). Representative echocardiograms are shown in Fig. 3C .

Exercise alters mRNA levels of genes involved in fatty acid oxidation in Dgat1 ؊ / ؊ cardiac muscle
Heart Ppara mRNA expression was doubled by exercise ( Fig. 4A ). This fi nding was associated with greater mRNA levels of Atgl , Cd36 , and Pdk4 ( Fig. 4B ); these changes are consistent with greater lipid uptake and oxidation as reported by others (20)(21)(22)(23). Dgat1 Ϫ / Ϫ mice also had exerciseinduced increases in mRNA levels of Ppara and Ppard and their downstream target genes ( Fig. 4A, B ). Thus, although the exercised Dgat1 Ϫ / Ϫ hearts still had lower mRNA levels of many of these genes compared with exercised WT mice, these genes were increased in the exercised Dgat1 Ϫ / Ϫ hearts, leading to similar mRNA levels as were found in the unexercised WT mouse hearts. These results indicated that exercise was able to induce metabolic changes independent of DGAT1 actions.

DGAT1 defi ciency reduces cardiac fatty acid oxidation and alters VLDL and glucose uptake
Previous studies showed that Dgat1 Ϫ / Ϫ skeletal muscle is more insulin-sensitive when exposed to hyperinsulinemic conditions ( 24 ), but the reasons were not revealed. Because our study demonstrated that mRNA levels of genes involved in lipid metabolism were reduced in Dgat1 Ϫ / Ϫ mice, we hypothesized that under basal conditions (without addition of insulin) Dgat1 Ϫ / Ϫ mice switch from lipid to more glucose usage. First, we tested fatty acid oxidation in isolated heart slices. Fatty acid oxidation rates were reduced in Dgat1 Ϫ / Ϫ mice ( Fig. 5A ). mice ( P < 0.05) and 10% and 9%, respectively, in high-fat diet Dgat1 Ϫ / Ϫ mice ( P < 0.05) ( Fig. 2A , B). Soleus and EDL muscle weights also were increased 13% ( P < 0.05) and 20% ( P < 0.01) in chow diet Dgat1 +/ Ϫ mice ( Fig. 2A). Thus, Dgat1 defi ciency increased muscle weight, especially on a chow diet. DEXA showed that lean tissue weight was increased 8% in chow diet Dgat1 Ϫ / Ϫ mice (Fig. 2C). No signifi cant change was observed for lean weight in high-fat diet mice with DEXA. However, as has been reported ( 10 ), fat tissue was reduced 56% in high-fat diet Dgat1 Ϫ / Ϫ mice (Fig. 2D). Photographs of DEXA are shown in supplementary Fig. I.
There are several signaling pathways that can mediate cardiac hypertrophy ( 18,19 ). We assessed markers for both pathological and physiologic hypertrophy. mRNA levels of heart failure markers brain-type natriuretic peptide (BNP) and atrial natriuretic factor (ANF) were not increased in the heart (supplementary Fig. IC, D). However, protein levels of p-AKT, a mediator of physiologic hypertrophy ( 18,19 ), and its downstream target p-mTOR were increased in heart (Fig. 2E); p-AKT increased 183%, and p-mTOR increased 126%. These increases were also found in high-fat diet mice (supplementary  Table 2 ). The protein levels of PPARs were also reduced (supplementary Fig. III). Cd36 mRNA level was decreased ‫ف‬ 60-80%, and adipose TG lipase ( Atgl ) and lipoprotein lipase ( Lpl ) mRNA levels were decreased ‫ف‬ 65%. mRNA levels of genes coding for pyruvate dehydrogenase kinase ( Pdk ) 4, a regulator of glucose oxidation, acylCoA oxidase ( Aox ), and carnitine palmitoyl transferase ( Cpt ) 1b decreased ‫ف‬ 60-80% ( Table 2 ). Proliferator-activated receptor ␥ coactivator ( Pgc)1a mRNA level, a coactivator of PPARs involved in fatty acid oxidation and mitochondria genesis, was not changed in chow fed Dgat1 Ϫ / Ϫ cardiac muscle. It was decreased in cardiac muscle from high-fat-diet mice, but in

uptake was increased in Dgat1
Ϫ / Ϫ heart (34.4%, P < 0.05) and skeletal muscle (soleus: 39%, P < 0.05). Glucose uptake by the liver did not change. VLDL-TG uptake into muscle was compared with that of liver to assess distribution of To assess the metabolic effects of DGAT1 defi ciency in vivo, plasma turnover and tissue distribution of tracer 2-deoxy-glucose and VLDL-triglyceride were determined. Despite normal plasma decay ( Fig. 5B ), 2-deoxy-glucose  of WT ( Table 4 )]. mRNA levels of genes coding for heart failure markers ( Anf and Bnp ) were not changed (data not shown). We next tested whether the DGAT1i had similar effects in mice fed a high-fat diet. As reported by others ( 25,26 ), the levels of mRNA expression of genes involved in fatty acid oxidation were increased in hearts of mice on a high-fat diet. However, many of these increases were totally blocked by the acute administration of DGAT1i; increases in mRNA levels corresponding to Atgl and Cpt1b were totally blocked by the DGAT1i ( Table 4 ). The mRNA levels for genes encoding Cd36 and Pdk4 were increased by high fat, but they remained much lower than was seen in untreated control hearts from high-fat diet mice.

Expression of genes involved in fatty acid oxidation in liver was increased in Dgat1 ؊ / ؊ mice and in DGAT1i-treated WT mice
Although our results demonstrated reduced mRNA levels of genes involved in fatty acid oxidation in heart and skeletal muscle in Dgat1 Ϫ / Ϫ mice, other studies reported that Dgat1 Ϫ / Ϫ mice on high-fat diets had increased fatty acid oxidation in whole body ( 10 ) and liver ( 27 ). However, hepatic mRNA levels of many metabolic genes were not reported. Unlike in muscle, some mRNA levels of genes involved in hepatic fatty acid oxidation were increased in Dgat1 Ϫ / Ϫ mice ( Table 5 ). mRNA levels of Ppard, Cpt1a, and Pdk4 were increased to 159 ± 49, 137 ± 18, and 151 ± 29% of WT ( P < 0.05). mRNA levels of other Ppar genes that mediate fatty acid uptake and metabolism and hepatocyte nuclear factor ( Hnf ) 4a did not change in Dgat1 Ϫ / Ϫ livers. Similar results were found in liver from DGAT1i-treated mice; mRNA levels of genes coding for Cpt1a and Pdk4 increased to 189 ± 58 and 146 ± 16% of control, respectively ( P < 0.05) ( Table 5 ).

Inhibition of DGAT1 reduces mRNA levels of genes involved in lipid metabolism in AC16 cardiomyocytes and C2C12 myocytes
To determine if the effects of DGAT1i treatment were direct or secondary to some other in vivo metabolic effect of this drug, AC16 and C2C12 cells were treated with DGAT1i in the presence or absence of palmitic acid. DGAT1i markedly reduced expression of Ppar and genes involved in fatty acid oxidation in both cell types ( Table  6 ). In addition, the increased mRNA levels due to 0.5 mM palmitic acid ( Ppara , Ppard, Pparg , Cd36 , Atgl , and Pdk4 increased from 1-to 4-fold) were totally prevented by DGAT1i treatment. DGAT1i attenuated, but did not completely block, the palmitic acid-mediated increase of mRNA levels of the gene coding for Cpt1b . Treating AC16 cardiomyocytes with DGAT1i prior to addition of labeled palmitic acid reduced palmitate uptake (supplementary Fig. IV). These results are similar to what we observed in vivo with the effects of high-fat diets and indicate that DGAT1 deficiency in muscle cells leads to marked reduction in mRNA levels of genes involved in lipid metabolism, an effect that is only partially compensated by addition of exogenous fatty acids. this tracer among organs. Plasma decay of VLDL is shown in Fig. 5C . VLDL-TG uptake was 47% greater in Dgat1 Ϫ / Ϫ liver. Although the amount of recovered tracer in heart and soleus was unchanged, when compared with liver uptake, relatively less VLDL-TG was distributed to muscles; 31% less TG was recovered in hearts, and 26% less in soleus ( Fig. 5B ). Thus, Dgat1 defi ciency led to greater uptake of lipid into the liver, while muscles switched to more glucose use.

Effect of acute DGAT1 inhibition on lipid contents and gene expression
The metabolic changes found in Dgat1 Ϫ / Ϫ mice could refl ect developmental compensation. For this reason, we tested whether pharmacological inhibition of DGAT1 also leads to reduced mRNA levels of genes mediating lipid uptake and oxidation. First, we measured the lipid content of DGAT1i-treated muscles. DGAT1i reduced the content of TG ‫ف‬ 16% ( P < 0.05) and ceramides ‫ف‬ 18% ( P < 0.05) in chow diet WT mouse hearts; DAG levels were not changed ( Table 3 ). DGAT1i-treated mice had reduced mRNA levels of all three forms of Ppar , ranging from 47 to 63% of control ( Table 4 ). mRNA levels of genes coding for proteins involved in lipid uptake were reduced [ Cd36 to ‫ف‬ 50%, Atgl to ‫ف‬ 50%, Pdk4 to ‫ف‬ 50%, and Cpt1b to ‫ف‬ 20%  , and Dgat1 Ϫ / Ϫ mice. C: Plasma VLDL-TG clearance, liver VLDL-TG uptake, and VLDL-TG uptake distribution in heart and soleus muscle from WT, Dgat1
Much of the interest in DGAT1 resulted from studies in mice with a genetic deletion of this enzyme. Dgat1 Ϫ / Ϫ mice have less diet-induced insulin resistance and obesity ( 10 ). Although older Dgat1 Ϫ / Ϫ mice have alopecia ( 36 ), these abnormalities were not observed in the young mice in our studies. In addition, this skin phenotype is not seen with partially reduced Dgat1 expression. Thus, inhibition of DGAT1 appeared to be a reasonable target for treatment of metabolic disorders.
In some situations, defective lipid storage in adipose leads to greater muscle lipid content and insulin resistance ( 37,38 ). The reason this does not occur in Dgat1 Ϫ / Ϫ mice or with acute pharmacologic DGAT1 inhibition became clear when we studied muscles from these mice. Chronic Dgat1 defi ciency did not grossly alter heart content of TG, although acute use of DGAT1i led to reduced heart TG content. In both situations, we had expected that reduced conversion of DAG to TG would elevate tissue levels of DAG and perhaps ceramide (i.e., we should fi nd the opposite of the effects of genetic overexpression of DGAT1 in muscles ( 5 )). This did not occur. We previously incubated isolated Dgat1 Ϫ / Ϫ skeletal muscle in high concentrations of fatty acids (a combination of oleic and palmitic acids) to test whether DGAT1 actions mediated detoxifi cation of fatty acids by their conversion to TG, resulting in less DAG and ceramide ( 5 ). This acute intervention and its effect on muscle insulin resistance did not use any conditions that were comparable to those found in vivo or in our in vitro studies of gene expression in cultured cells. In contrast, other studies have shown that Dgat1 Ϫ / Ϫ mice have improved insulin sensitivity ( 10 ), and our data suggest that this is due, at least in part, to changes in skeletal muscle metabolism.
The reasons that Dgat1 defi ciency did not lead to lipotoxicity were shown. Levels of mRNA of genes involved in lipid uptake were markedly reduced. Although not specifically studied, a reduction in cluster of differentiation 36 (CD36) is likely to have led to reduced uptake of free fatty acids ( 39 ), while a decrease in LpL would decrease muscle uptake of lipoprotein-derived fatty acids ( 14 ). These changes are also likely to have caused the redistribution of VLDL-TG with relatively more lipid directed to the liver and less to muscle. Although both overexpression and knockout of DGAT1 in cardiac muscle resulted in the DISCUSSION Lipid uptake and oxidation regulate cellular lipid stores. When this regulation is dysfunctional due to greater lipid accumulation than usage, it leads to a group of diseases classifi ed as lipotoxicities, including nonalcoholic fatty liver disease, type 2 diabetes, and metabolic cardiomyopathy. Data from our lab ( 5,6 ) and others ( 28 ) have shown that TG accumulation itself is not toxic. In contrast, it is believed that several lipid intermediates, especially ceramides and DAGs, are harmful (29)(30)(31)(32)(33)(34). Ceramides are thought to directly cause cellular apoptosis ( 35 ), and DAGs via activation of PKCs lead to insulin resistance ( 33,34 ). For this reason, we initially hypothesized that loss of DGAT1 activity would prevent conversion of DAG to TG and lead to cellular toxicity, if not on chow then on a highfat diet. We were wrong! But, in the process of our studies on Dgat1 defi ciency, we obtained information that explains the increased insulin sensitivity in Dgat1 Ϫ / Ϫ muscles and the reason that Dgat1 Ϫ / Ϫ mice have greater fatty acid oxidation (lower respiratory quotient). We show that buildup of toxic lipids does not occur because expression of genes involved in lipid uptake ( Cd36 and Lpl ) is reduced. This fi nding is a direct effect of loss of DGAT1 activity in myocytes, not a secondary phenomenon. Finally, we show that defi ciency of DGAT1 has a different effect on the liver.  control and exercise training. In addition, the increase in the p-AKT/p-mTOR pathway without increased expression of genes coding for the natiuretic proteins ANF and BNP is consistent with physiologic hypertrophy. We performed two experiments to determine if mRNA levels of proteins mediating fatty acid uptake and oxidation would increase with physiologic interventions. Although both exercise and high-fat diets increased gene expression, Dgat1 Ϫ / Ϫ mouse hearts were not normalized. Nonetheless, it was apparent that some activation of fatty acid oxidation pathways was via processes exclusive of those blocked by Dgat1 defi ciency.
As for DGAT1, studies of overexpression and deletion of PPAR in the heart have not led to the expected opposite phenotypes. Three PPAR genes have been upregulated in the heart. PPARa and PPARg lead to lipotoxic cardiomyopathy ( 44,45 ). PPARd, which does not cause increased lipid accumulation, does not cause cardiomyopathy ( 46 ). Reduced lipid use would be expected to occur with deletion of these transcription factors. Presumably that is why PPARa, PPARd, and PPARg cardiomyocyte deletion all lead to toxicity (47)(48)(49). We should note that unlike the situation in Dgat1 Ϫ / Ϫ mice, these deletions of Ppar are total and do not cause partial loss of the transcription factor.
Why does loss of DGAT1 activity lead to such marked changes in Ppar mRNA levels and expression of their downstream target genes? By performing studies in cultured cells, we showed that these changes are not secondary to other metabolic changes occurring in vivo. We postulate that Dgat1 defi ciency affects a pathway needed for uptake of lipids into the cells and movement of endogenous PPAR agonists to the nucleus. This might occur because Dgat1 defi ciency reduces input into the lipid droplet and creation of PPAR ligands via the activity of ATGL. LpL also generates PPAR ligands and is downregulated in Dgat1 Ϫ / Ϫ mice. How downregulation of DAG and ceramide, the mechanisms are different. Overexpression of DGAT1 decreases DAG as more DAG is used for TG synthesis and fatty acid oxidation in the muscle. Fatty acid levels in cardiac muscle of DGAT1 overexpression were decreased, whereas CD36 was increased as compensation for the outcome of this result ( 6 ). In contrast, the reduction in ceramide and DAG with DGAT1 defi ciency is, we assume, due to a decrease in lipid uptake and the availability of precursors.
Neither heart nor skeletal muscle appeared to have pathological changes due to reduced lipid uptake. In part, these tissues compensated by increasing glucose uptake. In many isolated perfused heart models of ischemia/ reperfusion, reduced fatty acid and increased glucose oxidation leads to less cardiac damage ( 40 ). Dgat1 Ϫ / Ϫ hearts appeared to retain their metabolic fl exibility with changes in metabolic genes, allowing greater fatty acid oxidation with exercise and high-fat diets. The greater heart and skeletal glucose uptake in the presence of nonelevated insulin is due to greater insulin sensitivity, as shown previously in clamp studies ( 41 ). In our study, greater glucose uptake was associated with a marked reduction in mRNA levels of the Ppar target genes regulating fatty acid oxidation. A reduction in PDK4 would be expected to lead to greater glucose oxidation.
Chronic Dgat1 defi ciency was also associated with increased p-AKT and p-mTOR, well-established causes of muscle hypertrophy ( 18,19 ). Ceramide content may reduce AKT phosphorylation ( 30,42 ). In our study, ceramide levels were signifi cantly decreased in hearts of Dgat1 Ϫ / Ϫ and DGAT1i-treated mice. This change might have accounted for the increase of p-AKT and p-mTOR. In the heart, muscle hypertrophy occurs with both physiologic compensation and in response to pathologic stimuli ( 43 ). Hearts of Dgat1 Ϫ / Ϫ mice functioned normally under other ligands, such as the phospholipid ligand described by the Semenkovich Lab ( 50 ), get into tissues is unclear. It is also not clear whether the reduction in lipid uptake or Ppar gene downregulation is primary; this is the direction of future investigation. Note that overexpression of DGAT1 in heart did not lead to as marked an alteration in PPAR downstream gene expression or regulation of PPAR mRNA levels as was found with knockout. Of all tissues, we suspect that the heart is bathed in PPAR agonists and any DGAT1-mediated increase in ligands leads to a relatively less dramatic effect than is seen with the knockout. Dgat1 Ϫ / Ϫ mice on high-fat diets had increased fatty acid oxidation in whole body ( 10 ) and liver ( 27 ). Some mRNA levels of genes Ppard, Cpt1a, and Pdk4 involved in hepatic fatty acid oxidation were increased in Dgat1 Ϫ / Ϫ and DGAT1i-treated mice. It is likely that the high level of expression of DGAT2 in the liver is suffi cient to obviate effects of DGAT1 loss in TG production ( 2,4 ).
In summary, studies of muscle effects of Dgat1 defi ciency uncovered a new PPAR regulatory pathway. Downregulation of genes involved in lipid uptake and oxidation is likely responsible for improved insulin sensitivity and lack of toxic lipid accumulation with Dgat1 defi ciency. Muscles from chow diet Dgat1 Ϫ / Ϫ mice are hypertrophied but do not have reduced function. Thus, as a therapeutic target, DGAT1 inhibition appears to be relatively free of cardiac and skeletal muscle toxicity.