Mouse Cardiac Acyl Coenzyme A Synthetase 1 Deficiency Impairs Fatty Acid Oxidation and Induces Cardiac Hypertrophy

Abstract

Long-chain acyl coenzyme A (acyl-CoA) synthetase isoform 1 (ACSL1) catalyzes the synthesis of acyl-CoA from long-chain fatty acids and contributes the majority of cardiac long-chain acyl-CoA synthetase activity. To understand its functional role in the heart, we studied mice lacking ACSL1 globally (Acsl1^T−/−) and mice lacking ACSL1 in heart ventricles (Acsl1^H−/−) at different times. Compared to littermate controls, heart ventricular ACSL activity in Acsl1^T−/− mice was reduced more than 90%, acyl-CoA content was 65% lower, and long-chain acyl-carnitine content was 80 to 90% lower. The rate of [¹⁴C]palmitate oxidation in both heart homogenate and mitochondria was 90% lower than in the controls, and the maximal rates of [¹⁴C]pyruvate and [¹⁴C]glucose oxidation were each 20% higher. The mitochondrial area was 54% greater than in the controls with twice as much mitochondrial DNA, and the mRNA abundance of Pgc1α and Errα increased by 100% and 41%, respectively. Compared to the controls, Acsl1^T−/− and Acsl1^H−/− hearts were hypertrophied, and the phosphorylation of S6 kinase, a target of mammalian target of rapamycin (mTOR) kinase, increased 5-fold. Our data suggest that ACSL1 is required to synthesize the acyl-CoAs that are oxidized by the heart, and that without ACSL1, diminished fatty acid (FA) oxidation and compensatory catabolism of glucose and amino acids lead to mTOR activation and cardiac hypertrophy without lipid accumulation or immediate cardiac dysfunction.

The mitochondrial oxidation of long-chain fatty acids (FAs) provides 60 to 90% of heart ATP (9, 43, 49). Reduced cardiac FA oxidation and increased glucose utilization are a proposed consequence of pathological left ventricular hypertrophy (LVH) (22, 33). However, when genes that encode enzymes of FA oxidation are knocked out in mice, LVH develops (11, 20). Thus, it remains unclear whether the shift in substrate use is a cause or consequence of cardiac hypertrophy and whether the increased use of glucose interferes with cardiac function.

Long-chain acyl coenzyme A (acyl-CoA) synthetase (ACSL) isoenzymes convert FAs to acyl coenzyme A (acyl-CoA) in an ATP-dependent manner, simultaneously activating and trapping FAs within cells (4). Activation to acyl-CoA is required before FAs can be either oxidized to provide ATP or esterified to synthesize triacylglycerol (TAG) or membrane phospholipids (PL). The activation of FA is catalyzed by one of a family of five long-chain acyl-CoA synthetases (ACSLs), long-chain acyl-CoA synthetase isoform 1 (ACSL1), ACSL3, ACSL4, ACSL5, and ACSL6, which differ in substrate preference, enzyme kinetics, subcellular location, and tissue-specific expression (10). Because amphipathic acyl-CoAs can move freely within a membrane monolayer or be transported to distant membranes, all acyl-CoAs should, theoretically, be metabolically equivalent, no matter which ACSL isoenzyme catalyzes their formation and no matter which subcellular organelle is the site of their synthesis. Yet, both loss-of-function and gain-of-function studies suggest that each ACSL isoenzyme has a distinct function in directing acyl-CoAs to one or more specific downstream pathways (5, 29, 32). We have reported that mice lacking ACSL1 specifically in adipose tissue have defects in adipose FA oxidation (15); however, mice lacking ACSL1 in the liver have minor defects in both hepatic FA oxidation and triacylglycerol synthesis (28), suggesting that the ACSL isoforms may have roles that differ in different tissues. The role that ACSL1 plays in cardiac FA metabolism has remained unclear.

Although FAs provide the major substrate for oxidation in the heart, it is unknown whether ACSL activity affects FA oxidation rates or whether a particular ACSL isoenzyme activates FAs destined for oxidation. The total ACSL activity of the mouse embryonic heart (embryonic day 16.5 [E16.5]) increases 14-fold during the week after birth and 90-fold by adulthood (13). This dramatic increase in ACSL activity parallels the transition of the developing heart's substrate preference from glucose prenatally to FAs after birth. In mice, this transition is accompanied by a 6-fold increase in Acsl1 mRNA abundance and a >90% decrease in the mRNA abundance of Acsl3, the predominant Acsl1 in fetal heart (13). In rats, the increases in ACSL activity and Acsl1 mRNA expression parallel the postnatal heart's increased workload, rate of ATP generation, oxidative preference for FAs, and expression of FA oxidative genes (19). Together, these data suggested that ACSL1 might be the major activator of the FAs that are oxidized in postnatal cardiac tissue.

In order to understand the relationship between cardiac substrate use and hypertrophy, we created a mouse model that lacks acyl-CoA synthetase 1 (ACSL1) in multiple tissues, including the heart. We reasoned that the absence of ACSL1 would enable us to learn whether a block in FA activation prevents potential lipotoxicity and abnormal heart function and to determine the mechanism by which a shift in substrate use from FAs to glucose causes cardiac hypertrophy. Herein we report that mice lacking ACSL1 in a multitissue-specific manner and in a heart-specific manner have markedly reduced FA oxidation in the heart and develop cardiac hypertrophy.

MATERIALS AND METHODS

Animal treatment.

All protocols were approved by the University of North Carolina Institutional Animal Care and Use Committee. Mice were housed in a pathogen-free barrier facility (12-h light/12-h dark cycle) with free access to water and food (Prolab RMH 3000 SP76 chow). Mice with Loxp sequences inserted on either side of exon 2 in the Acsl1 gene (9) were backcrossed to the C57BL/J6 strain six times and then interbred with mice in which Cre expression is driven either by a ubiquitous promoter enhancer or by an α-myosin heavy-chain promoter, both of which are induced by tamoxifen [B6.Cg-Tg(cre/Esr1)5Amc/J or B6.Cg-Tg(Myh6-cre/Esr1)1Jmk/J; Jackson Labs] to generate tamoxifen-inducible multitissue-specific (Acsl1^T−/−) or heart-specific (Acsl1^H−/−) Acsl1 knockout mice. When the mice were 6 to 8 weeks old, tamoxifen (Sigma, St. Louis, MO), dissolved in corn oil (20 mg/ml), was injected intraperitoneally (i.p.) for 4 consecutive days (3 mg/40 g of body weight) into Acsl1^T−/−, Acsl1^H−/−, and littermate Acsl1^flox/flox control mice. Tissues were removed and snap-frozen in liquid nitrogen. To isolate the mitochondria, the hearts were removed, rinsed in phosphate-buffered saline (PBS), minced, and homogenized with 10 up-and-down strokes, using a motor-driven Teflon pestle and glass mortar in ice-cold buffer (0.2 mM EDTA, 0.25 M sucrose, 10 mM Tris-HCl [pH 7.8], protease inhibitor [Roche, Florence, SC]). Homogenates were centrifuged at 1,000 × g for 10 min at 4°C, the supernatant was spun at 12,000 × g for 15 min at 4°C, and the resulting pellet was washed once and resuspended in oxidation buffer. Protein content was determined by using the bicinchoninic acid (BCA) assay kit (Pierce, Rockford, IL) with bovine serum albumin (BSA) as the standard. Plasma was collected from mice in 5% 0.5 M EDTA. Plasma triacylglycerol (TAG), β-hydroxybutyrate (Stanbio, Boerne, TX), cholesterol, free FAs (FFAs), glucose (Wako, Richmond, VA), and free and total glycerol (Sigma) were measured with colorimetric assays. Insulin tolerance tests were performed by i.p. injection with insulin (0.5 U/kg of body weight), and the tail blood glucose level was measured at baseline, 15, 30, 60, and 120 min using a One Touch Ultra glucometer (Lifescan, Inc., Milpitas, CA). Blood pressure was measured by CODA high-throughput noninvasive tail blood pressure system (Kent Scientific, Torrington, CT). Mouse echocardiograms were performed on unanesthetized mice with the VisualSonics (Toronto, Ontario, Canada) Vevo 770 ultrasound biomicroscopy system. M-mode images of the left ventricle were analyzed using VisualSonics software. Transverse arch banding was performed in mice using a slipknot technique as previously described (44).

ACSL assay.

ACSL initial rates were measured with 50 μM [1-¹⁴C]palmitic acid (Perkin Elmer, Waltham, MA), 10 mM ATP, and 0.25 mM coenzyme A (CoA) in total membrane fractions (0.5 to 4.0 μg) or in ventricular mitochondria (38).

Reverse transcription-PCR (RT-PCR).

Total RNA was isolated from ventricles (RNeasy fibrous tissue kit; Qiagen, Alameda, CA), and cDNA was synthesized (Applied Biosystems high-capacity cDNA reverse transcription kit). Total DNA was isolated using the QIAmp DNA microkit (Qiagen). DNA or cDNA was amplified by real-time PCR using SYBR green (Applied Biosystems, Foster City, CA) detection with primers specific to the gene of interest. Results were normalized to the housekeeping gene Gapdh for mRNA or H19 for DNA and expressed as arbitrary units of 2^−ΔΔCT relative to the control group.

Immunoblots.

Total protein lysates were isolated in lysis buffer (20 mM Tris base, 1% Triton X-100, 50 mM NaCl, 250 mM sucrose, 50 mM NaF, 5 mM Na₂P₂O₇, plus protease inhibitor [catalog no. 11836153; Roche, Florence, SC]). Total membrane fractions were isolated in medium I (10 mM Tris [pH 7.4], 1 mM EDTA, 0.25 M sucrose, 1 mM dithiothreitol). Equal amounts of protein (40 to 60 μg) were loaded and resolved on 10% SDS-polyacrylamide gels and transferred to nitrocellulose membranes. Blots were probed with antibodies against ACSL1 (catalog no. 4047), phosphorylated AMP-activated protein kinase (P-AMPK) (catalog no. 2535), and phosphorylated p70 (P-p70) S6 kinase (S6K) (catalog no. 9234) and were then stripped and reprobed with either total AMPKα (catalog no. 2532) or p70 S6K (catalog no. 9202) antibody (all antibodies from Cell Signaling, Danvers, MA). The purity of mitochondrial fractions was verified by immunoblotting with antibodies against the mitochondrial protein voltage-dependent anion channel protein (VDAC) (ab16816) and the endoplasmic reticulum (ER) protein calnexin (ab13504) (both antibodies from Abcam, Cambridge, MA).

Histology.

The hearts were gravity perfused and fixed for 24 h in PBS containing 4% paraformaldehyde and transferred to 70% ethanol. The fixed tissue samples were embedded in paraffin, serial sectioned, and stained with hematoxylin and eosin or Masson's trichrome. For lectin staining, paraformaldehyde-fixed cardiac tissue samples were deparaffinized, hydrated, and incubated with Triticum vulgaris lectin tetramethyl rhodamine isothiocyanate (TRITC) conjugate (Sigma). The sections were subsequently examined by fluorescence microscopy. For electron micrograph analysis, the animals were euthanized, and their hearts were perfused with a freshly made solution containing 2% paraformaldehyde and 2.5% glutaraldehyde in 0.15 M sodium phosphate buffer, pH 7.4. Cardiac tissue was imaged using a LEO EM910 transmission electron microscope at 80 kV (LEO Electron Microscopy, Thornwood, NY) and photographed using a Gatan BioScan digital camera (Gatan, Inc., Pleasanton, CA). Mitochondrial area and myocyte size were quantified using NIH ImageJ software.

Fatty acid oxidation.

Freshly isolated heart ventricles and livers were minced and homogenized with 10 up-and-down strokes using a motor-driven Teflon pestle and glass mortar in ice-cold buffer (100 mM KCl, 40 mM Tris-HCl, 10 mM Tris base, 5 mM MgCl₂·6H₂O, 1 mM EDTA, and 1 mM ATP [pH 7.4]) at a 20-fold dilution (wt/vol), or 40 μg of isolated mitochondria was used for oxidation (37). Oxidation was measured in a 200-μl reaction mixture containing 100 mM sucrose, 10 mM Tris-HCl, 10 mM KPO₄, 100 mM KCl, 1 mM MgCl₂·6H₂O, 1 mM l-carnitine, 0.1 mM malate, 2 mM ATP, 0.05 mM coenzyme A, and 1 mM dithiothreitol (pH 7.4) with either 8 μM [1-¹⁴C]palmitate (0.1 μCi/reaction) and 100 μM sodium palmitate complexed to BSA, 80 μM [1-¹⁴C]pyruvate (0.1 μCi/reaction mixture) and 5 mM pyruvate, 4 μM [1-¹⁴C]palmitoyl-CoA (0.04 μCi/reaction mixture) and 50 μM palmitoyl-CoA, or 8 μM [U-¹⁴C]glucose (0.1 μCi/reaction mixture) and 200 μM glucose. Oxidation studies measured the production of ¹⁴C-labeled carbon dioxide (CO₂) and acid-soluble metabolites (ASM) for 30 min without substrate competition in a two-well oxidation system: one well contained the reaction mixture with the tissue homogenate, and the adjoining well contained 1 NaOH. The reaction was terminated by adding 70% perchloric acid to the assay well, and then the plate was incubated for 1 h to drive the CO₂ into the NaOH. Radioactivity of ASM in the supernatant of the reaction mixture and CO₂ was determined by liquid scintillation. Fatty acid oxidation was quantified using the following formula: [(dpm − BL)/SA]/[gram of tissue [wet weight] × time (in hours) of reaction mixture incubation]), where dpm is the disintegrations per minute, BL is the dpm of blank wells, and SA is the FA-specific radioactivity.

Tissue lipid, nucleotide, and glucose-6-phosphate content.

Acyl-carnitines were quantified by liquid chromatography and tandem mass spectrometry (1). Complex lipid content was analyzed by Lipomics Technologies, Inc. (West Sacramento, CA) (50). Acyl-CoA and diacylglycerol composition and content were assessed from whole hearts extracted with internal standards (36). After separation, purification, and elution, lipid metabolite extracts were separated by high-performance liquid chromatography (HPLC), and individual and total lipid species were analyzed by liquid chromatography and tandem mass spectrometry (36). Frozen hearts were homogenized in 0.4 M perchlorate and neutralized in 4 M K₂CO₃ as described previously (31). Nucleotides were separated by HPLC (7) using a Varian Prostar solvent delivery system (PS-210; Varian, Palo Alta, CA) and a Luna 5μm C₁₈ 100A column (Phenomenex, Torrance, CA). Peaks were detected using a Gilson 118 UV detector (Middleton, WI). Glucose-6-phosphate was measured by a spectrophotometric enzymatic analysis using glucose-6-phosphate dehydrogenase coupled to NADPH production with extinction at 340 nm (26).

PPARα activity.

Nuclear fractions were isolated from fresh hearts (nuclear extraction kit; Cayman Chemicals, Ann Arbor, MI) and used with the peroxisome proliferator-activated receptor α (PPARα) complete transcription factor assay kit (Cayman Chemicals, Ann Arbor, MI) according to the manufacturer's protocol.

2-Deoxyglucose uptake.

Anesthetized mice were injected retro-orbitally with 10 μCi of 2-[1-¹⁴C]deoxyglucose in saline (Moravek Biochemicals, Brea, CA). Tissues were harvested and flash frozen in liquid N₂ 30 min after injection. Radioactivity was measured in tissue homogenates and normalized to the counts present in 10 μl of serum obtained 5 min after injection [(dpm/mg of tissue)/dpm in 10 μl serum].

RESULTS

Cardiac ACSL activity and acyl-CoA content are reduced in mice that lack ACSL1.

In mice lacking ACSL1 globally (Acsl1^T−/−) mice, the largest and most persistent reduction in total ACSL enzyme activity and ACSL1 protein was observed in hearts, compared to livers, kidneys, and adipose tissue (Fig. 1A to D). Ten weeks after tamoxifen was injected, total ACSL activity in heart ventricles was 97% lower than in the littermate controls (Fig. 1A). The ACSL activity that remained in tissues was probably due to other ACSL isoenzymes and to residual ACSL1 activity in cells that had not been exposed to tamoxifen. The appearance, histology, and weights of the liver and adipose tissue did not differ for the different genotypes (Table 1; histology not shown). Two weeks after the tamoxifen injection, heart Acsl1 mRNA was nearly absent, but 40% of ACSL1 protein remained (Fig. 1B and E), suggesting that the half-life of ACSL1 may be as long as 2 weeks. By 10 weeks after the tamoxifen injection, virtually no ACSL1 protein was present. Other differences in the measurements at the two time points included 37% lower plasma triacylglycerol (TAG) and 2.5-fold-higher plasma fatty acid (FA) concentrations in the Acsl1^T−/− mice compared to the controls at 2 weeks, but not at 10 weeks after the tamoxifen injection (Table 1). In hearts from Acsl1^T−/− mice, loss of ACSL1 reduced the total pool of long-chain acyl-CoAs by 65% (Fig. 1F), with 67 to 75% reductions in 16:0-, 16:1-, 18:1-, 18:2-, and 18:3-CoA species and a 26% reduction in 18:0-CoA (Fig. 1G). Expressed as a percentage of total acyl-CoA content, 18:0-CoA was twice as high in Acsl1^T−/− hearts as in the control hearts, and all other species except 16:0-CoA were reduced ∼20%. Thus, aside from the relative increase in 18:0-CoA, lack of ACSL1 did not markedly change the FA composition of the acyl-CoA pool. These data indicate that ACSL1 is the major activator of long-chain FA in the heart and that the knockdown of Acsl1 was virtually complete and persistent in cardiac tissue for as long as 10 weeks. Despite the 65% reduction in long-chain acyl-CoA content in Acsl1^T−/− hearts, total cardiac TAG content did not change, suggesting that ACSL1 does not limit the acyl-CoA pool used for the synthesis of TAG (Fig. 1H).

FIG. 1.

ACSL activity and acyl-CoA content are reduced in mice that lack ACSL1. (A to D) Total ACSL activity (A and C) and ACSL1 protein (B and D) in control (Con) and Acsl1^T−/− tissues 2 weeks (2wk) and 10 weeks (10wk) after the tamoxifen injection (n = 5 to 7). Kid, kidney; WAT, white adipose tissue; BAT, brown adipose tissue. (E) Acsl isoenzyme mRNA abundance in ventricles 2 and 10 weeks after the tamoxifen injection in control and Acsl1^T−/− mice (n = 6). (F and G) Heart total (F) and individual (G) long-chain acyl-CoA content in control and Acsl1^T−/− mice 10 weeks after the tamoxifen injection (n = 7 or 8). (H) Heart triacylglycerol (TAG) content in control and Acsl1^T−/− mice 10 weeks after the tamoxifen injection (n = 7 or 8). The values are means plus standard errors of the means (SEMs) (error bars). Values that are significantly different (P ≤ 0.05) from the values for the control are indicated by an asterisk. Values that are significantly different 2 weeks versus 10 weeks after the tamoxifen injection within mice of the same genotype are indicated by a # symbol.

TABLE 1.

Body weight, adipose weight, plasma metabolites, and blood pressure in male mice^a

Characteristicb	Value (mean ± SEM) at the following time after the tamoxifen injection in micec:
	2 wk		10 wk
	Control	Acsl1T−/−	Control	Acsl1T−/−
Body wt (g)	25.5 ± 1.5	23.3 ± 1.7	31.4 ± 0.8†	30.1 ± 0.4†
G-WAT wt (% of BW)	0.87 ± 0.1	0.92 ± 0.1	1.7 ± 0.2†	1.8 ± 0.2†
Plasma TAG concn (mg/dl)	47.2 ± 3	29.7 ± 2.0*	36.7 ± 5†	33.8 ± 2
Plasma FA concn (mmol/liter)	0.08 ± 0.03	0.28 ± 0.04*	0.08 ± 0.01	0.09 ± 0.01†
Glucose concn (mg/dl)	136 ± 8	131 ± 9	164 ± 3.6	136 ± 5.3*
Insulin concn (ng/ml)			2.73 ± 0.8	1.67 ± 0.2
AUC (mg/min/dl)			4,981 ± 298	4,828 ± 357
Blood pressure
Systolic (mm Hg)			155 ± 4.0	158 ± 3.4
Diastolic (mm Hg)			120 ± 4.3	125 ± 3.6

^aMale control and Acsl1^T−/− mice (n = 8 to 15) were studied 2 and 10 weeks after the tamoxifen injection. The body weight and adipose tissue weight, plasma metabolites, area under the curve in response to insulin tolerance tests, and blood pressure were determined.

^bG-WAT, gonadal white adipose tissue; TAG, triacylglycerol; FA, fatty acid; area under the curve (AUC) in response to the insulin tolerance test (ITT).

^cValues that are significantly different (P ≤ 0.05) from the value for the control mice are indicated by an asterisk. Values that are significantly different, (P ≤ 0.05) at 2 weeks versus 10 weeks after the tamoxifen injection within mice of the same genotype are indicated by a † symbol.

ACSL1 is required for heart FA oxidation.

To determine whether ACSL1 provides acyl-CoAs for FA oxidation, we assessed the rates of [1-¹⁴C]palmitate incorporation into CO₂ and acid-soluble metabolites (ASM), which are measures of complete and incomplete oxidation, respectively. Compared to the controls, the rate of maximal FA oxidation in Acsl1^T−/− heart homogenates was 95% lower (Fig. 2A). Thus, the absence of ACSL1 nearly abolished FA oxidative capacity. In contrast, in Acsl1^T−/− livers, in which ACSL1 protein and activity remained similar to those of controls, the rate of [1-¹⁴C]palmitate oxidation was unchanged (Fig. 2A). The oxidation rates for the ACSL1 product [1-¹⁴C]palmitoyl-CoA were similar for the two genotypes (Fig. 2B), indicating that carnitine palmitoyltransferase 1 (CPT1)-mediated transport of acyl-CoA into the mitochondria and its subsequent oxidation to CO₂ were unimpaired and that the severe block in FA oxidation in Acsl1^T−/− hearts was due to the lack of FA activation.

FIG. 2.

Acsl1^T−/− hearts have impaired fatty acid (FA) oxidation. (A and B) [1-¹⁴C]palmitate oxidation to CO₂ and acid-soluble metabolites (ASM) in heart and liver homogenates (A) and heart [1-¹⁴C]palmitoyl-CoA oxidation into CO₂ from control (Con) and Acsl1^T−/− (T^−/−) mice (n = 5 or 6) (B). (C) Representative immunoblots against ACSL1, VDAC, and calnexin in control and Acsl1^T−/− mitochondrial fractions 2 and 10 weeks after the tamoxifen injection. (D) [1-¹⁴C]palmitate oxidation to CO₂ and ASM from control (Con) and Acsl1^T−/− (T^−/−) ventricular mitochondria 2 and 10 weeks after the tamoxifen injection (n = 5 or 6). (E to H) Free acyl-carnitines (E), acetyl-carnitine (F), medium-chain acyl-carnitines (G), and long-chain acyl-carnitines (H) in control and Acsl1^T−/− hearts (n = 5 to 7). The values are means plus SEMs (error bars). Values that are significantly different (P ≤ 0.05) from the values for the control are indicated by an asterisk.

The 95% lower rates of FA oxidative capacity in Acsl1^T−/− heart homogenates strongly suggested that ACSL1 activates the pool of FAs that is directed toward mitochondrial oxidation. This interpretation is further supported by the presence of residual ACSL1 in purified heart mitochondria 2 weeks after tamoxifen was injected (Fig. 2C). At that time, when ∼40% of ACSL1 protein still remained, [1-¹⁴C]palmitate oxidation in Acsl1^T−/− heart mitochondria was 40% lower than in the controls; however, 10 weeks after the tamoxifen injection, when ACSL1 protein was absent, [1-¹⁴C]palmitate oxidation was 90% lower than in the controls (Fig. 2D). These data showing that the rate of FA oxidative capacity and the amount of ACSL1 protein decreased in parallel support the idea that ACSL1 is required to activate cardiac FAs before they can be oxidized.

Cellular acyl-carnitines are markers of metabolite flux through degradative and oxidative pathways (30). In Acsl1^T−/− cardiac tissue, total free carnitine was 77% higher than in control hearts, indicating that carnitine was not limiting (Fig. 2E), whereas acetyl-carnitine was 51% lower, consistent with reduced mitochondrial acetyl-CoA levels (Fig. 2F). Confirming the impaired FA oxidation, nearly all medium-chain (8- to 12-carbon) acyl-carnitine species in the Acsl1^T−/− hearts were 60 to 80% lower than in the control hearts (Fig. 2G), and long-chain acyl-carnitines were 80 to 90% lower than in the control hearts (Fig. 2H). Total cholesteryl ester and diacylglycerol were reduced ∼20% in Acsl1^T−/− hearts, but total phospholipid (PL) and TAG were unaffected (see Fig. S1 in the supplemental material). Because the levels of PL and TAG were normal in Acsl1^T−/− hearts but the abundance of nearly all acyl-carnitine species was reduced ∼90%, it appears that ACSL1 activity does not affect the synthesis of acyl-CoAs used in the pathways of PL and TAG synthesis but, instead, specifically provides the acyl-CoAs required for FA oxidation.

Altered FA composition of Acsl1^T−/− cardiac lipids.

If all acyl-CoAs were metabolically equivalent, the composition of complex lipids would reflect changes in the acyl-CoA pool. However, despite the fact that the Acsl1^T−/− hearts contained twice as much 18:0-CoA as the control hearts (as a percentage of total long-chain acyl-CoAs), the amount of 18:0-CoA did not change in the TAG, phosphatidylethanolamine (PE), and phosphatidylserine (PS) fractions of Acsl1^T−/− hearts and was significantly lower in the cardiolipin (CL) and phosphatidylcholine (PC) fractions (see Fig. S1 in the supplemental material). Acsl1^T−/− 18:1-CoA content was ∼60% lower than in the controls and, as a percentage of total long-chain acyl-CoAs, was 10 to 20% lower; however, the total levels of 18:1-CoA were not altered in the TAG, PE, and PS fractions and were increased in the CL and PC fractions. As a percentage of total esterified FA, 18:1-CoA was increased in TAG, PE, CL, and PC fractions, but not in the PS fraction (see Fig. S1 in the supplemental material). These data indicate that the FA composition of TAG and PL in Acsl1^T−/− hearts did not reflect the changes in the acyl-CoA pool and suggest that other ACSL isoenzymes contributed to the acyl-CoAs that were esterified in the pathways of glycerolipid synthesis.

Pathological cardiac hypertrophy and mitochondrial biogenesis in the Acsl1^T−/− mice.

The weights of Acsl1^T−/− female and male hearts as a percentage of body weight were 23% and 17% greater than the control hearts, respectively (Fig. 3A). Cardiomyocyte area was 27% larger than for the controls (Fig. 3B), and ventricular walls appeared thickened (Fig. 3C). Echocardiography confirmed that Acsl1^T−/− ventricular walls were 10 to 41% thicker than for the controls (Table 2) with a 34% heavier left ventricular mass (Fig. 3D and E). Echocardiography analysis also showed that the Acsl1^T−/− left ventricular inner diameter was ∼30% smaller than for the controls, suggesting concentric ventricular hypertrophy (Table 2). Despite left ventricular hypertrophy (LVH), the percent fractional shortening and ejection fraction were similar to the values for the controls, indicating that cardiac function remained normal for at least 2 months after the ablation of Acsl1 (Fig. 3F and Table 2).

FIG. 3.

Acsl1^T−/− mice develop cardiac hypertrophy. (A) Weight (Wt) (wet weight) of control and Acsl1^T−/− female and male hearts expressed as a percentage of body weight (n = 10 to 20). (B) Quantification of cardiomyocyte area from lectin-stained control and Acsl1^T−/− male hearts (n = 3). (C) Representative hematoxylin-and-eosin-stained control and Acsl1^T−/− hearts. (D) Echocardiography calculation of left ventricular (LV) mass to the body mass of male control and Acsl1^T−/− mice (n = 8 to 10). (E and F) Representative echocardiogram M-mode images (E) and percent fractional shortening (%FS) (F) in control and Acsl1^T−/− male mice (n = 8 to 10). (G) mRNA abundance of fetal gene markers, α-skeletal actin (αSkAc) and brain natriuretic peptide (Bnp) in control and Acsl1^T−/− hearts (n = 6). (H to J) Heart weight (H), percent fractional shortening (I), and ACSL1 protein content (in arbitrary units [Au]) (J) in control and Acsl1^T−/− mice 4 weeks after transverse arch banding (TAB) (n = 4 to 6). The values are means plus SEMs (error bars). Values that are significantly different (P ≤ 0.05) from the values for the control are indicated by an asterisk.

TABLE 2.

Echocardiogram diameter and functional characteristics^a

Characteristicb	Value (mean ± SEM) for micec:
	Control	Acsl1T−/−
Diastole
IVST (mm)	0.98 ± 0.02	1.29 ± 0.02*
LVID (mm)	2.88 ± 0.06	2.47 ± 0.06*
LVPWT (mm)	0.86 ± 0.01	1.23 ± 0.02*
Systole
IVST (mm)	1.67 ± 0.03	1.89 ± 0.04*
LVID (mm)	1.31 ± 0.04	1.08 ± 0.03*
LVPWT (mm)	1.48 ± 0.03	1.80 ± 0.04*
% EF	86.6 ± 0.50	88.3 ± 0.51
LV mass (mg)	86.5 ± 3.57	115.4 ± 4.24*

^aTransthoracic echocardiogram imaging was performed on conscious control and Acsl1^T−/− male mice (n = 6 to 8) 10 weeks after the tamoxifen injection. M-mode images were analyzed.

^bIVST, interventricular septum thickness; LVID, left ventricule inner diameter; LVPWT, left ventricule posterior wall thickness; % EF, percent ejection fraction; LV, left ventricule.

^cValues that are significantly different (P ≤ 0.05) from the values for the control are indicated by an asterisk.

Upregulation of fetal genes occurs with pathological hypertrophy (18); thus, the 4- to 5-fold-higher mRNA expression of the fetal genes for α-skeletal actin and brain natriuretic peptide (Fig. 3G), together with the 6-fold upregulation of Acsl3, the major Acsl mRNA expressed in fetal heart (13) (Fig. 1E), is compatible with pathological hypertrophy. Fibrosis was not present (data not shown). Thus, the absence of ACSL1 for 10 weeks severely impaired FA oxidation and induced ventricular hypertrophy with upregulation of fetal genes but without cardiac dysfunction or failure.

To evaluate the responses of hearts from Acsl1^T−/− mice under stress conditions, we induced pressure overload by transverse arch banding in control and Acsl1^T−/− mice. The banding procedure resulted in a doubling of heart weight (Fig. 3H) and a decrease in fractional shortening to ∼40% (Fig. 3I) in both control and Acsl1^T−/− hearts. With transverse arch banding-induced hypertrophy, the ACSL1 protein decreased (Fig. 3). These data show that ACSL1 is repressed and not required for pressure overload-induced cardiac hypertrophy.

Evaluation of cardiomyocyte organelles by electron microscopy showed that the mitochondrial area in Acsl1^T−/− ventricles was 54% greater than in the controls (Fig. 4A and B). Confirming the enhanced mitochondrial biogenesis in Acsl1^T−/− ventricles, the levels of mitochondrial DNA for NADH dehydrogenase subunit 1, cytochrome b, and cytochrome c oxidase 1 were at least 2-fold higher than for the controls (Fig. 4C). Mitochondrial biogenesis is driven by several transcription factors, including peroxisome proliferator-activated receptor (PPAR) coactivator 1α (PGC1α) and estrogen-related receptor α (ERRα) (35). In Acsl1^T−/− ventricles, the mRNA levels for the Pgc1α and Errα genes were 100% and 41% greater, respectively, than in controls 2 weeks after tamoxifen was injected (Fig. 4D). PPARα transcription factor activity in nuclear extracts from control and Acsl1^T−/− hearts did not differ for the two genotypes (Fig. 4E), but several PPARα/δ target genes were markedly upregulated in Acsl1^T−/− hearts, including muscle carnitine palmitoyltransferase 1, cytosolic thioesterase 1, and Acsl3 (6) (Fig. 4F and G and 1E). The increased mRNA abundance of mitochondrial β-oxidation genes could reflect either increased mitochondrial content in Acsl1^T−/− hearts or the activation of PPARδ by mammalian target of rapamycin (mTOR) (46). Because FA oxidation was severely impaired, we suspected that an energy deficit might exist that would activate the energy-sensing enzyme, AMP-activated kinase (AMPK), which can upregulate Pgc1α. However, compared to the controls, AMPK phosphorylation (Thr¹⁷²) was ∼40% lower in Acsl1^T−/− ventricles (Fig. 4H). Further, myocardial ATP and AMP content did not differ for the two genotypes (Fig. 4I), suggesting that glucose oxidation had increased sufficiently to maintain the cellular ATP/AMP ratio.

FIG. 4.

Mitochondrial excess in Acsl1^T−/− hearts. (A and B) Representative electron microscopy images (bars = 2 μm) (A) and quantification of mitochondrial area from the ventricles of control and Acsl1^T−/− male mice (B) 10 weeks after the tamoxifen injection (n = 3). (C) Quantification of the mitochondrial DNA genes for cytochrome c oxidase 1 (Co1), cytochrome b (Cytb), and NADH dehydrogenase subunit 1 (Nd1) relative to nuclear DNA in the ventricles from male and female control and Acsl1^T−/− mice 10 weeks after the tamoxifen injection (n = 6). (D) mRNA abundance of Pgc1α and Errα genes in control and Acsl1^T−/− ventricles 2 weeks after the tamoxifen injection (n = 6). (E) PPARα transcription factor activity (TFA) (in arbitrary units [Au] per microgram of protein) in nuclear extracts from control and Acsl1^T−/− hearts (n = 6 to 8). (F and G) mRNA abundance of Cte1, mCpt1, Mcad, Cd36, Fas, and Pparα genes in control and Acsl1^T−/− ventricles (n = 6). (H) Quantification of AMPK phosphorylation (AMPK-P) at Thr¹⁷² over total AMPK in control and Acsl1^T−/− ventricles 10 weeks after the tamoxifen injection (n = 5 to 7). (I) Myocardial ATP and AMP content determined by HPLC in control and Acsl1^T−/− hearts. The values are means ± SEMs (error bars). Values that are significantly different (P ≤ 0.05) from the values for the control are indicated by an asterisk.

Increased glucose use in Acsl1^T−/− hearts.

Because the capacity of Acsl1^T−/− hearts to oxidize FAs was markedly impaired yet cellular ATP content was preserved, we questioned whether the oxidative capacity for non-FA substrates had increased to compensate. In Acsl1^T−/− heart homogenates, the maximal rate of [U-¹⁴C]glucose incorporation into [¹⁴C]CO₂ was 20% higher than in the controls (Fig. 5A), and in Acsl1^T−/− ventricular mitochondria, the oxidation of [1-¹⁴C]pyruvate to [¹⁴C]CO₂ was 39% higher, suggesting increased pyruvate dehydrogenase activity (Fig. 5B). Consistent with increased glucose use, plasma glucose in Acsl1^T−/− mice was 17% lower than in the control mice (Table 1). Confirming enhanced glucose metabolism, the uptake of [1-¹⁴C]2-deoxyglucose into Acsl1^T−/− hearts compared to the control hearts was 9-fold greater (Fig. 5C). 2-Deoxyglucose uptake into Acsl1^T−/− liver and gonadal adipose tissue was similar to that in control mice, but the uptake of 2-deoxyglucose into brown adipose tissue (BAT) was 2-fold greater (Fig. 5D). The amount of glucose-6-phosphate was 3-fold higher in the Acsl1^T−/− hearts than in the control hearts, suggesting an enhanced rate of glycolysis with accumulation of glycolytic intermediates (Fig. 5E). In addition, propionyl- and methyl-malonyl/succinyl-carnitines were 71% and 174% higher in Acsl1^T−/− hearts than in the control hearts (Fig. 5F). These carnitine species are metabolites of glucose and of branched-chain and ketogenic amino acids, suggesting that oxidation of these alternate substrates had increased in Acsl1^T−/− hearts. Consistent with the idea that protein catabolism provided an alternative fuel source, the total amino acid content in Acsl1^T−/− hearts was 32% higher than in the control hearts (Fig. 5G). This increase in amino acid content was reflected by increases in nearly every amino acid species (Fig. 5H and I). Together, these data strongly support the conclusion that reduced FA oxidation in Acsl1^T−/− hearts leads to compensatory increases in glycolysis and protein catabolism.

FIG. 5.

Glucose oxidation, amino acid catabolism, and S6 kinase activation increased in Acsl1^T−/− hearts. (A) [U-¹⁴C]glucose oxidation to CO₂ in heart homogenates from control and Acsl1^T−/− mice (n = 5 or 6). (B) [1-¹⁴C]pyruvate oxidation to CO₂ in mitochondria from the hearts of control and Acsl1^T−/− mice (n = 3 or 4). (C and D) [1-¹⁴C]2-deoxyglucose ([1-¹⁴C]2DG) uptake into the heart (C) and liver, gastrocnemius muscle, gonadal white adipose tissue (WAT), and brown adipose tissue (BAT) (D) (see Materials and Methods) (n = 3 or 4). (E) Glucose-6-phosphate content in control and Acsl1^T−/− ventricles (n = 5 to 7). (F) Short-chain acyl-carnitine content in control and Acsl1^T−/− hearts (n = 5 to 7). Short-chain acyl-carnitine abbreviations: 3, propionyl-carnitine; 5OH/3DC, 3-hydroxy-isovalerly- or malonyl-carnitine; 4DC/i4DC, methylmalonyl- or succinyl-carnitine. (G to I) Total amino acids (AA) (G) and individual amino acids (H and I) in control and Acsl1^T−/− hearts (n = 5 to 7). (J) Representative immunoblot (n = 4 or 5) and quantification (n = 10) of p70-S6K phosphorylation (S6K-P) at Thr³⁸⁹ and total p70-S6K in control and Acsl1^T−/− ventricles. The values are means plus SEMs (error bars). Values that are significantly different (P ≤ 0.05) from the values for the control are indicated by an asterisk.

mTOR stimulates hypertrophy in Acsl1^T−/− hearts.

The mTOR kinase activates pathways that increase cell growth (3, 12). mTOR kinase activation is required for both thyroid hormone-induced and spontaneous hypertensive rat cardiac hypertrophy (14, 25). To determine whether mTOR was linked to the hypertrophy observed in Acsl1^T−/− hearts, we quantified the phosphorylation of the mTOR target, p70-S6 kinase (S6K). Compared to the controls, S6K phosphorylation at Thr³⁸⁹ was 5-fold greater in Acsl1^T−/− hearts (Fig. 5J). These data suggest that the cardiac hypertrophy in Acsl1^T−/− mice was mediated by activation of the mTOR pathway. Possible causes of enhanced mTOR signaling include elevations in amino acid availability (2), enhanced glycolytic flux (41), elevated plasma insulin (21), and hypertension (42). mTOR activation in Acsl1^T−/− hearts probably resulted from a combination of several of these factors, including the 32% higher amino acid content and the increase in glucose uptake and glycolysis (Fig. 5). In addition, diminished AMPK activity, as shown by the ∼40% reduction in Acsl1^T−/− heart AMPK phosphorylation (Fig. 4H), would relieve the inhibition of mTOR by AMPK (23). Other activators of mTOR signaling were not present; neither plasma insulin nor the area under the curve during an insulin tolerance test differed for mice with the two genotypes, and hypertension was not present (Table 1). Because the downstream substrate of mTOR, S6 kinase, is not hyperphosphorylated in brown adipose or gonadal adipose tissue from Acsl1^T−/− mice (data not shown), it appears that mTOR is activated by the deficiency of ACSL1 specifically in cardiac tissue. Glucose-6-phosphate is also an activator of mTOR (41), and compared to the control hearts, the Acsl1^T−/− hearts contained 3-fold-more glucose-6-phosphate (Fig. 5E). These data implicate enhanced glycolytic flux, the accumulation of glycolytic intermediates, and decreased AMPK activation as the major signals that mediate mTOR activation and hypertrophy in Acsl1^T−/− hearts.

Heart-specific ACSL1 deficiency causes hypertrophy.

Because the Acsl1^T−/− mice were deficient in ACSL1 in multiple tissues (Fig. 1A), it might be argued that the observed cardiac hypertrophy had resulted from some aspect of the whole-body ACSL1 deficiency. Thus, we examined mice that were deficient in ACSL1 only in ventricular cardiomyocytes (Acsl1^H−/−). In these Acsl1^H−/− mice, real-time PCR showed nearly absent expression of Acsl1 mRNA in the ventricles (Fig. 6A) and a 50% lower expression of Acsl5 mRNA (data not shown). ACSL1 protein was absent in Acsl1^H−/− ventricles but was present in the liver and skeletal muscle and at levels similar to those of the controls (Fig. 6B). The activity of ACSL in Acsl1^H−/− atria was similar to that of controls, whereas ventricle ACSL activity was 90% lower (Fig. 6C). Confirming that ACSL1 is required for cardiomyocyte mitochondrial fatty acid oxidation, the rate of [1-¹⁴C]palmitate incorporation into CO₂ and ASM in isolated mitochondria from Acsl1^H−/− ventricles was reduced ∼90% compared to control ventricles (Fig. 6D). Supporting the conclusion that cardiac hypertrophy was a direct result of ACSL1 cardiomyocyte deficiency, the weights of the hearts from female and male Acsl1^H−/− mice were 25% and 20% greater than the hearts from the control mice, respectively (Fig. 6E). The left ventricle masses, calculated from echocardiograph M-mode images, were 34% heavier (Fig. 6F) with a 30% larger left ventricular wall diameter in the Acsl1^H−/− hearts than in the control hearts. The percent fractional shortening and ejection fraction were similar in mice with both genotypes, indicating that cardiac function remained normal despite ventricular hypertrophy (data not shown). To determine whether mTOR was linked to the hypertrophy observed in Acsl1^H−/− hearts, we quantified the phosphorylation of S6K. S6K phosphorylation at Thr³⁸⁹ was 5-fold greater in the Acsl1^H−/− hearts than in the control hearts (Fig. 6G). These data strongly suggest that the cardiac hypertrophy in both Acsl1^T−/− and Acsl1^H−/− mice was mediated by activation of the mTOR pathway and that hypertrophy is a direct consequence of ACSL1 deficiency in cardiomyocytes.

FIG. 6.

Impaired oxidation, hypertrophy, and S6 kinase activation in the cardiomyocyte-specific ACSL1 knockout mice. (A) Acsl1 mRNA abundance in the heart 10 weeks after the tamoxifen injection in control and Acsl1^H−/− mice (n = 6). (B) Representative immunoblot against ACSL1 protein in heart, liver, and gastrocnemious (gastroc) muscle 10 weeks after the tamoxifen injection. (C) Total ACSL activity in control and Acsl1^H−/− atria and ventricles 10 weeks after tamoxifen (n = 5 to 7). (D) [1-¹⁴C]palmitate oxidation into carbon dioxide (CO₂) and acid-soluble metabolites (ASM) in heart homogenates from control and Acsl1^H−/− mice (n = 5 or 6). (E) Weight (wet weight) of control and Acsl1^H−/− male and female hearts expressed as a percentage of body weight (n = 10 to 20). (F) Echocardiography calculation of left ventricular (LV) mass in control and Acsl1^H−/− male mice (n = 8 to 10). (G) Representative immunoblot of p70-S6K phosphorylation at Thr³⁸⁹ and total p70-S6K in control and Acsl1^H−/− ventricles 20 weeks after tamoxifen (n = 4 or 5). The values are means plus SEMs (error bars). Values that are significantly different (P ≤ 0.05) from the values for the control are indicated by an asterisk.

DISCUSSION

The primary finding of this study is that ACSL1 provides the FAs used for cardiac oxidation, and that in the absence of ACSL1, the heart compensates by increasing the oxidation of glucose and amino acids. Normally, FAs provide 60 to 90% of the heart's energy requirements, and glucose and lactate oxidation provide the remaining ATP (43, 49). The use of these three energy sources is controlled by substrate availability, physiological conditions, and transcriptional and hormonal regulation. Our data suggest that the use of FAs as an oxidative substrate is additionally controlled by ACSL1.

Whether a shift from lipolytic to glycolytic oxidation is a cause or a consequence of cardiac hypertrophy has remained unclear. Data from human deficiencies in FA oxidation and rodent models with genetically or chemically impaired FA oxidative capacity support the interpretation that reduced FA oxidation causes cardiac hypertrophy. For example, in humans, defects in the Na⁺-carnitine cotransporter (Na⁺-driven organic cation transporter 2 [OCTN2]) cause a cardiomyopathy characterized by cardiac lipid accumulation and hypertrophy (34). Similarly, mice with juvenile visceral steatosis (JVS) mouse mimics human systemic carnitine deficiency because mice with JVS have a spontaneous deficiency of OCTN2 that results in cardiac lipid accumulation and hypertrophy, as well as a 2-fold increase in cardiac mitochondrial area (45). Inhibition of CPT1 and acyl-carnitine synthesis in rats leads to cardiac hypertrophy (39), and mice deficient in long-chain acyl-CoA dehydrogenase (LCAD) (11) or very-long-chain acyl-CoA dehydrogenase (VLCAD) (17) also develop cardiac hypertrophy, increased mitochondrial biogenesis, and TAG accumulation. Our data also strongly support the idea that the shift to glycolysis promotes cardiac hypertrophy.

When ACSL1 is overexpressed 12-fold in the mouse heart, left ventricular mass doubles by the time the mouse is 3 weeks old, and heart failure, characterized by a 50% reduction in percent fractional shortening, occurs by the time the mouse is 4 weeks old (8). The rates of FA oxidation were not assessed in these hearts, but the levels of TAG and phosphatidylcholine in the hearts were 12-fold and 1.5-fold higher, respectively, than in the control hearts. These data suggest that when ACSL1 is markedly overexpressed, the acyl-CoAs produced are used to synthesize TAG and phospholipid (PL). In this model, ACSL1 overexpression probably activated large amounts of acyl-CoA within cardiomyocytes, without a concomitant increase in oxidative demand. Thus, similarly to VLCAD and LCAD null mice, in which mitochondrial FA oxidation is partially blocked (11, 17), excess and potentially toxic acyl-CoAs are esterified to form triacylglycerol (TAG). In contrast to these models, lack of ACSL1 prevents FA activation, thereby both blocking FA metabolism and preventing FAs from being trapped as acyl-CoA within cells. The reduction in ACSL activity allows nonactivated FAs to leave cells, thereby decreasing the apparent rate of FA uptake and the amount of FAs retained (48). Thus, Acsl1^T−/− and Acsl1^H−/− mice have severely impaired cardiac FA oxidation that results in hypertrophy without TAG accumulation.

The fact that the levels of PL and TAG were unchanged in the hearts of Acsl1^T−/− mice despite a 90 to 97% reduction in ACSL activity suggests that ACSL1 does not substantially contribute to the synthesis of acyl-CoAs that are incorporated into glycerolipids. If the metabolic fate of a cellular acyl-CoA were nonspecific, then a 65% decrease in long-chain acyl-CoA content should diminish both oxidation and TAG synthesis equally. The normal content of TAG, the 80 to 90% reduction in long-chain acyl-carnitines, and the 90% decrease in palmitate oxidation, all support the conclusions that ACSL1 channels FAs specifically toward oxidation and that acyl-CoAs synthesized by other ACSL isoenzymes are used for complex lipid synthesis, but not for oxidation (Fig. 7).

FIG. 7.

Overview of metabolic disturbance and pathway activation in the Acsl1^T−/− heart. The loss of ACSL1 prevents uptake and activation of fatty acids (FAs) for oxidation. Other ACSL isoforms (ACSLx) activate FAs that are used for triacylglycerol (TAG) and phospholipid (PL) synthesis. The inability of Acsl1^T−/− heart to oxidize FA is compensated for by increased glucose and amino acid catabolism. The shift in oxidative metabolism leads to reduced AMPK phosphorylation and the activation of the mTOR pathway causes cardiac hypertrophy in Acsl1^T−/− and Acsl1^H−/− mice. S6 Kinase-P, S6 kinase phosphorylation; Glucose 6-P, glucose-6-phosphate.

A major regulator of cell size is the mTOR kinase, which activates S6 kinase (S6K) to initiate transcriptional activity that induces cell growth (27). Activated mTOR is present in several models of cardiac hypertrophy and is likely to be an important signal in the pathway that mediates hypertrophy. For example, S6K phosphorylation is increased in the hypertrophied hearts of spontaneous hypertensive rats (SHR), and treatment with the mTOR inhibitor rapamycin attenuates the hypertrophy without altering the hypertension (42). Similarly, rapamycin inhibits thyroid hormone-induced S6K phosphorylation and cardiac hypertrophy (25). Exercise-induced cardiac hypertrophy in mice also increases S6K phosphorylation (24), whereas mouse hearts lacking acetyl-CoA carboxylase 2 (ACC2) have reduced S6K phosphorylation and smaller hearts (16). These data strongly suggest that mTOR activation mediates cell growth in several models of cardiac hypertrophy and that when mTOR is inhibited, the hypertrophy reverses or diminishes.

Several features present in Acsl1^T−/− hearts could enhance mTOR activation, including elevated amino acid content, enhanced glycolysis, and diminished AMP-activated kinase (AMPK) activity. AMPK activity decreases in isolated skeletal muscle after exposure to glucose or to branched-chain amino acids and results in increased phosphorylation of mTOR and S6K and increased protein synthesis (40). This induction of the mTOR pathway in muscle depends on the reduced AMPK phosphorylation and activation (40), probably because AMPK phosphorylates the tumor suppressor complex that coverts GTP-Rheb to GDP-Rheb and inactivates mTOR (23). Thus, the decrease in AMPK activity relieves the GDP-Rheb inhibition of mTOR. Although it remains unclear how increased glucose and amino acid metabolism activates the mTOR pathway in cardiac muscle, AMPK suppression is likely to play a key role in this process. It has been proposed that long-chain acyl-CoAs inhibit AMPK by allosterically interacting with LKB1/AMPKK; thus, the 65% reduction in long-chain acyl-CoAs in Acsl1^T−/− hearts could lead to reduced AMPK activation (47). The resulting diminished content of activated AMPK would relieve LKB1/AMPKK inhibition and allow greater mTOR activity. In the isolated working heart of a mouse, the initial phosphorylation of glucose appears to be critical for glucose-induced S6K phosphorylation; S6K phosphorylation occurred after exposing cardiomyocytes to 2-deoxyglucose, which can be phosphorylated, but not after exposure to 3-O-methylglucose, which cannot be phosphorylated (41). The 3-fold increase in glucose-6-phosphate in the Acsl1^T−/− hearts suggests that glycolytic intermediates may activate the mTOR pathway. Our data suggest that the shift from FA to glucose oxidation, the consequent increase in glycolytic flux, the increase in amino acid availability, and the reduction in AMPK phosphorylation all contributed to mTOR activation in Acsl1^T−/− hearts (Fig. 7).

In summary, the loss of ACSL1 in the mouse heart results in severely impaired FA oxidation, a compensatory increase in glucose oxidation, and cardiac hypertrophy without systolic dysfunction or lipid accumulation. The 5-fold increase in S6K phosphorylation suggests that activation of the mTOR pathway initiates the observed cardiac hypertrophy in Acsl1^T−/− and Acsl1^H−/− hearts. In support of the role of ACSL1 in activating FAs for oxidation, mice with an adipose-tissue-specific knockout of ACSL1 had reduced FA oxidation rates in white and brown adipocytes and were severely cold intolerant (15). Like ACSL1 in adipose tissue, it appears that ACSL1 functions in cardiac tissue to activate FAs destined for β-oxidation. In this model, the shift in substrate use from FAs to glucose causes the ensuing cardiac hypertrophy, although cardiac function was not impaired during the time period studied.