Patients with very long-chain acyl-CoA dehydrogenase (VLCAD) deficiency frequently present cardiomyopathy and heartbeat disorders. However, the underlying factors, which may be of cardiac or extra cardiac origins, remain to be elucidated. In this study, we tested for metabolic and functional alterations in the heart from 3- and 7-mo-old VLCAD null mice and their littermate counterparts, using validated experimental paradigms, namely, 1) ex vivo perfusion in working mode, with concomitant evaluation of myocardial contractility and metabolic fluxes using 13C-labeled substrates under various conditions; as well as 2) in vivo targeted lipidomics, gene expression analysis as well as electrocardiogram monitoring by telemetry in mice fed various diets. Unexpectedly, when perfused ex vivo, working VLCAD null mouse hearts maintained values similar to those of the controls for functional parameters and for the contribution of exogenous palmitate to β-oxidation (energy production), even at high palmitate concentration (1 mM) and increased energy demand (with 1 μM epinephrine) or after fasting. However, in vivo, these hearts displayed a prolonged rate-corrected QT (QTc) interval under all conditions examined, as well as the following lipid alterations: 1) age- and condition-dependent accumulation of triglycerides, and 2) 20% lower docosahexaenoic acid (an omega-3 polyunsaturated fatty acid) in membrane phospholipids. The latter was independent of liver but affected by feeding a diet enriched in saturated fat (exacerbated) or fish oil (attenuated). Our finding of a longer QTc interval in VLCAD null mice appears to be most relevant given that such condition increases the risk of sudden cardiac death.
- isolated working mouse heart perfusion
- metabolic fluxes
- docosahexaenoic acid
very long-chain acyl-CoA dehydrogenase (VLCAD) catalyzes the first reaction of mitochondrial β-oxidation. This enzyme utilizes fatty acyl-CoA of chain lengths varying between 14 and 24 carbons but shows its highest activity with palmitoyl-CoA (25). In humans, VLCAD deficiency, first reported in 1993 (2, 4, 62), is the most common inherited long-chain fatty acid (LCFA) oxidation disorder, with an incidence currently estimated to be between 1/42,500 and 1/120,000 (34, 52). Because VLCAD is highly expressed in the liver, heart, and skeletal muscles, VLCAD deficiency evokes multiple clinical symptoms, including cardiomyopathy and heartbeat disorders (6). The cardiac symptoms are currently believed to arise from an energy deficit due to impaired β-oxidation of LCFAs (which normally supply ∼70% of energy to the normal beating heart) and/or accumulation of toxic metabolites, such as long-chain (LC)-acylcarnitines or LC-CoAs (6). In the most severe cases, without appropriate treatments, VLCAD-deficient infants will succumb from sudden cardiac death, mainly caused by severe arrhythmias between 2 and 5 mo of age, often after fasting or an infectious illness (30, 47, 57). However, ∼30% of these patients never develop cardiac symptoms and display a milder clinical phenotype characterized by hypoketotic hypoglycemia, indicating metabolic derangements in the liver (57). This diversity in clinical phenotype severity as well as the large variety of mutations known to cause VLCAD deficiency make the diagnosis and treatment of patients particularly difficult (21, 37).
VLCAD null mice were created in 2003 (15) to improve our understanding of the clinical phenotype of VLCAD-deficient patients and, hopefully, suggest more efficient treatments. VLCAD null mice recapitulate the human phenotype to some extent. Specifically, they display cold intolerance, arrhythmia, and LC-acylcarnitine accumulation in the liver, serum, and skeletal muscles when exposed to exercise or cold under fasting conditions (10, 14, 15, 53, 54, 56). Furthermore, newborn VLCAD null mice display alterations in metabolic protein and gene expression in the heart, liver, and brown adipose tissues (14, 15, 20) as well as in calcium homeostasis in the heart (60). However, in general, VLCAD null mice present a milder phenotype than humans. In this regard, Chegary et al. (9) recently reported that fibroblasts from VLCAD-deficient mice and humans differ in their capacity to oxidize LCFA, and their finding was explained by the functional overlap of VLCAD with LC-acyl-CoA dehydrogenase (LCAD), an enzyme that is active in mice but apparently not in humans. However, much remains to be elucidated about the causes underlying the wide variety of symptoms manifested by patients with VLCAD deficiency. To the best of our knowledge, the impact of VLCAD deficiency on lipid metabolism, particularly in the heart, has not yet been specifically examined nor has its metabolic or functional capacity to withstand a fat challenge been tested.
Hence, we investigated whether VLCAD null mice hearts display alterations in their metabolic, contractile, and electrical functions when subjected to conditions that mimic those believed to induce decompensation in human patients, namely, fasting, adrenergic stress, and high-fat loading. These parameters were assessed in 3- and/or 7-mo-old VLCAD null mice and their littermates using validated experimental paradigms, namely, 1) ex vivo perfusion in working mode with concomitant evaluation of myocardial contractility and metabolic fluxes, employing 13C-labeled LCFA (28); and 2) in vivo telemetry. While ex vivo working VLCAD null mouse hearts unexpectedly maintained normal contractile function and utilization of exogenous LCFAs for β-oxidation under all conditions examined, other in vivo analyses revealed that these mice displayed other cardiac-specific lipid and functional alterations beyond β-oxidation, namely age- and condition-dependent accumulation of LCFAs in triglycerides (TGs), lower docosahexaenoic acid (DHA) level in phospholipids (PLs), as well as a prolonged rate-corrected QT (QTc) interval, a condition that increases the risk of sudden cardiac death.
MATERIALS AND METHODS
Materials and Animals
The sources of chemicals, biological products, and 13C-labeled substrates as well as the BSA dialysis procedure (BSA fraction V; Intergen, Purchase, NY) have been identified previously (58). Antibodies against phosphorylated and total extracellular-regulated kinase (ERK) forms 1 and 2 (ERK1/2), Akt, and beclin-1, Atg7, and Atg12 were purchased from Cell Signaling Technologies (Danvers, MA). Animal experiments were all approved by the local animal care committee in agreement with the guidelines of the Canadian Council on Animal Care. Acadvl−/− (VLCAD−/−) mice (129svJ X C57BL/6J genetic background) were kindly provided by Dr. Arnold Strauss, University of Cincinnati, College of Medicine (15). Male VLCAD−/− mice were backcrossed with C57Bl/6J females to produce heterozygous VLCAD+/− mice, which were then bred to generate VLCAD−/− and control littermate (VLCAD+/+) offspring. The latter were genotyped, as described below. All animals, housed in a specific pathogen-free facility with a 12:12-h light-dark cycle starting at 7:00 AM, had free access to water and standard chow (except if specified otherwise, as described below).
Quantitative PCR Genotyping Assay
All animals were genotyped at weaning via tail biopsies (20 mg/ml), which were incubated at 55°C for 4 h in buffer (50 mM Tris, 30 mM EDTA, and 0.25% SDS) containing (1,000 μg/ml) proteinase K (Sigma, St. Louis, MO). After digestion, the samples were diluted 1:250 in DNAse/RNAse free water along with 15 mg chelex-resin (Bio-Rad Laboratories, Hercules, CA) and heated at 95°C for 5 min. They were then analyzed in duplicate by quantitative (q)PCR with 2X Platinum SYBR Green qPCR Supermix-UDG, according to the manufacturer's specifications (Invitrogen Life Technologies, Carlsbad, CA). Cycling was achieved in a MX3005p cycler (Stratagene, Mississauga, Ontario, Canada) with the following conditions: 95°C for 10 min and 40 cycles of 95°C for 30 s, 47°C for 45 s, and 72°C for 45 s. The primers for qPCR were designed to target the NEOr gene cassette (reverse: TGGCTACCCGTGATATTGC; forward: GGCGATAGAAGGCGATGC), and the corresponding PCR signal was expressed relative to cyclophilin A (Ppia) genomic DNA (reverse: GCCGCCAGTGCCATTATG; forward: CCGATGACGAGCCCTTGG).
Effect of a 24-h Fast and Various Diets
Male VLCAD+/+ and VLCAD−/− mice were fed either a standard diet (3- and 7-mo-old; Harlan Teklab no. 2018; 3.1 kcal/g), a high-fat diet (HFD: Harlan Teklab no. 03584; 5.4 kcal/g), or a fish oil diet (FOD: Harlan Teklab no. 10789; 3.4 kcal/g) ad libitum and were kept for 2 (HFD) or 5 (FOD) wk in the same room. Percent calories from carbohydrates, lipids, and proteins for the various diets were, respectively, as follows: 1) standard: 58:18:24; 2) HFD: 26.6:58.4:15; and 3) FOD: 54.2:25.4:20.4 (for the FA profile of these diets, see Supplemental Table 1; Supplemental Material for this article is available online at the Am J Physiol Heart Circ Physiol website). To document the impact of fasting, selected animals from all groups were housed individually in cages, with fasting initiated at 9:00 AM and lasting 24 h. All animals were weighed before death under anesthesia induced by injection of a solution (1 μl/g ip) of ketamine (100 mg/ml) and xylazine (20 mg/ml). Body weights were not different between groups of mice fed different diets, under all conditions and at all ages (Supplemental Table 2). In one group of 3-mo-old mice from both genotypes, hearts were perfused ex vivo in working mode, as described below. In other groups, hearts, livers, and skeletal muscles (gastrocnemius) were excised rapidly, freeze clamped with metal tongs chilled in liquid nitrogen, and stored in liquid nitrogen. At the same time, blood samples were collected (18) and stored at −80°C for further analyses.
Heart Perfusion in Semirecirculating Mode
Male VLCAD−/− and VLCAD+/+ mice, which had free access to water and food unless specified otherwise, were anesthetized (1 μl/g ip) with a mixture of ketamine (100 mg/ml) and xylazine (20 mg/ml) and heparinized (5,000 U/kg sc) 15 min before surgery. Previous publications describe the procedure: 1) for mouse heart isolation and its ex vivo perfusion in working mode as well as for continuous monitoring of functional and physiological parameters throughout perfusion as well as of concentrations in influent (arterial) and effluent (coronary) of Po2, Pco2, pH, Ca2+, and other ions (18, 28); and 2) for calculating myocardial oxygen consumption (MV̇o2), intracellular pH, rate pressure product, cardiac power, and cardiac efficiency from previously reported equations (28). At the end of the perfusion period, hearts were freeze clamped with metal tongs chilled in liquid nitrogen, weighed, and stored at −80°C for further analyses.
Working mouse hearts were perfused for 30 min with semirecirculating Krebs-Heinselet buffer containing various substrates, cofactors, and hormones according to four different protocols. For all perfusions, the buffer contained 1.5 mM lactate, 0.2 mM pyruvate, 0.8 nM insulin, 50 μM l-carnitine, and 0.5 nM epinephrine, while glucose and palmitate (complexed to 3% albumin) concentrations were varied to mimic fed or fasting conditions. For any given perfusion, we included [U13-C16]palmitate (initial molar percent enrichment: 25%) as the labeled substrate and assessed its metabolism through β-oxidation and TG formation. For protocol 1, hearts from 3- and 7-mo-old VLCAD+/+ (n = 6) and VLCAD−/− (n = 5) mice were perfused at preload and afterload pressures of 15 and 50 mmHg, respectively, in the presence of 11 mM glucose and 0.4 mM palmitate to mimic the fed state. For protocol 2, hearts from 3-mo-old VLCAD+/+ (n = 6) and VLCAD−/− (n = 6) mice were perfused, as in protocol 1 except for the presence of 7 mM glucose and 1 mM palmitate to mimic the fasted state. For protocol 3, hearts from 3-mo-old VLCAD+/+ (n = 6) and VLCAD−/− mice (n = 5) were perfused with 7 mM glucose and 1 mM palmitate as in protocol 2 for the first 5 min. Thereafter, afterload pressure was increased from 50 to 60 mmHg, and 1 μM of epinephrine was added to mimic adrenergic stress. For protocol 4, hearts isolated from 24-h fasted, 3-mo-old VLCAD+/+ (n = 5) and VLCAD−/− (n = 6) mice were perfused as described in protocol 2.
Electrocardiogram Analysis In Vivo
Electrocardiogram (ECG) parameters were monitored continuously in live mice via in vivo telemetry, as described previously (8, 12) in 7-mo-old control and VLCAD null mice (n = 6 in each group). In brief, 12 mice were simultaneously instrumented under isoflurane anesthesia, with OpenHeart radiofrequency transmitters (Data Sciences International, Arden Hills, MN). After a 1-wk recovery period, during which the mice were fed the standard diet, ECG parameters were recorded for 24 h to track the complete circadian rhythm. Then, the standard diet was replaced by a HFD for 2 wk or FOD for 5 wk, and ECG parameters were recorded for another 24 h. Twelve signals corresponding to 12 mice were acquired simultaneously with IOX 220.127.116.11 (EMKA Technologies, Falls Church, VA) at a sample rate of 1 kHz. Recordings were analyzed with ECG Auto 18.104.22.168 (EMKA Technologies) for heart rate as well as for RR, QRS, and QT intervals. The data reported represent the means of values recorded during the first 10 min of each hour over the 24-h recording period.
Gas chromatography-mass spectrometry (GC-MS) assay and flux parameters.
Our previously published studies (28, 58) provide the following: 1) detailed descriptions of assays of 13C enrichment of Krebs cycle intermediates (citrate, succinate, fumarate, and malate) by GC-MS (Agilent 6890N gas chromatograph equipped with a HP-5 column coupled to a 5973N mass selective detector); 2) definitions of 13C terminology and calculations of flux ratios relevant to substrate selection for citrate synthesis from 13C enrichment of the acetyl (carbons 4 and 5) and oxaloacetate (carbons 1, 2, 3, and 6) moiety of citrate; and 3) determinations of glucose and lactate concentrations by enzymatic assays.
FA profiling in TGs and PLs.
FAs from heart, liver, and skeletal muscle tissue homogenates (TGs and PLs) were quantified using a previously described modified method (29), which included tissue lipid extraction (16) and separation into classes with an aminoisopropyl column (Varian, Harbor City, CA; Ref. 46). Briefly, pulverized tissues (25 mg) were incubated overnight at 4°C in a solution of chloroform/methanol (2:1) containing 0.004% butylated hydroxytoluene, filtered through gauze, and dried under nitrogen gas. The samples were resuspended in 2 ml of hexane/chloroform/methanol (95:3:2). TGs and PLs were separated on the aminoisopropyl column (which had been activated with hexane) by the addition of 1) 4 ml chloroform (for TGs), followed by 2) 2.5 ml methanol/chloroform (6:1) and 2.5 ml 0.05 M sodium acetate in methanol/chloroform (6:1) (for PLs). All samples were dried under nitrogen gas after the addition of 50 μl of 2 mM nonadecanoic acid as external standard. FAs were transmethylated according to a modified method described by Lepage and Roy (33). Briefly, FAs were resuspended in 2 ml of hexane/methanol (1:4) with 0.004% butylated hydroxytoluene and 200 μl of acetyl chloride added; the samples were then heated at 100°C for 1 h. The remaining acetyl chloride was neutralized by incorporating 5 ml of 6% potassium carbonate; thereafter, the upper hexane phase containing FA methyl esters was collected and subjected to GC-FID or GC-MS. GC-FID analyses were achieved in an Agilent DB-WAX polar capillary column (60 m; 0.25 mm ID; 0.25 μm thickness) with high-purity hydrogen as carrier gas at a constant flow of 1.9 ml/min under the following conditions: 40°C for 7 min, increased by 10°C/min until 120°C, then by 1°C/min until 180°C, and, finally, by 0.5°C/min until 218°C. GC-MS was operated in chemical ionization mode with methane as reagent gas. LCFAs were analyzed as their [M+H]+ ion by selective ion monitoring, after separation in a Varian CP7420 FA methyl ester polar capillary column (100 m; 0.25 mm ID; 0.23 μm thickness) with high-purity helium as carrier gas at a constant flow rate of 0.5 ml/min under the following conditions: 190°C for 25 min and increased by 2°C/min until 245°C. GC-FID and GC-MS have been internally validated, and similar results were obtained for samples analyzed by both techniques. FAs were identified according to their retention time, and concentrations were calculated using standard curves and the external standard.
13C enrichment of palmitate in TGs.
Palmitate from perfused heart tissue TGs was extracted and processed, as described above, except that it was analyzed by GC-MS (29) with selective ion monitoring at m/z of 270 (for unlabeled palmitate) and 286 (for [U-13C16]palmitate).
RT-qPCR gene expression analysis.
Supplemental Table 3 enumerates the qPCR primer pairs for the selected genes, crafted with the Beacon Designer program (version 5.0), using mouse sequences available in GenBank and not reported previously (18). Only qPCR primer pairs giving >90% efficiency were retained. qPCR analyses were conducted as described previously (18). We documented the gene expression of enzymes/proteins involved in β-oxidation, namely, VLCAD (Acadvl), LCAD (Acadl), medium-chain acyl-CoA dehydrogenase (MCAD: Acadm), the recently discovered ACAD9 (Acad9), as well as hormone-sensitive lipase (HSL: Lipe) and adipose triglyceride lipase (ATGL: Pnpla2), which are implicated in TG hydrolysis, and, finally, two enzymes involved in polyunsaturated fatty acids (PUFAs) metabolism, delta-5 (Fads1) and delta-6 desaturase (Fads2). Gene transcript levels in each sample were averaged and normalized to the housekeeping gene, namely, hypoxanthine guanine phosphoribosyl transferase 1 (Hprt1), after ΔΔCt correction. This gene was selected because among all tested housekeeping genes (Gapdh, Hprt1, and Ppia), it displayed the lowest variations between different experimental conditions.
Protein expression and phosphorylation status.
The abundance and phosphorylation of several proteins were evaluated by standard Western blotting in hearts and livers clamped in vivo or in perfused hearts. We used primary antibodies against phospho-ERK1/2, total ERK1/2, phospho-Akt (ser 473), total Akt, beclin-1, Atg7, Atg12 (mouse specific), or GAPDH as a loading control. Thereafter, the blots were washed and incubated with secondary antibody conjugated to horseradish peroxidase (1:10,000). Note that, for all measurements, we obtained similar results twice with different animals.
The data are expressed as means ± SE of n = 4–20 hearts, livers, or plasma samples. Statistical significance was reached at P < 0.05 by unpaired t-test or two-way ANOVA, followed by the Bonferroni posttest. The two-way ANOVA was used to compare the effects of conditions (fasting or diet) within the control and VLCAD null groups and to assess the interaction between these groups.
Ex Vivo Working VLCAD Null Mouse Hearts Maintain Physiological Parameters and Exogenous LCFA β-Oxidation Similar to Control Values, Even When LCFA Concentrations and Energy Demand Are Increased
To test the hypothesis that VLCAD-deficient hearts manifest alterations in contractile function and LCFA β-oxidation, particularly in the presence of high LCFA concentrations, we perfused ex vivo working hearts with a mixture of substrates (28), including 13C-labeled LCFA. This provides indicators of the dynamics of cardiac energy substrate metabolism, information that is not accessible from static measurements of mRNA, protein expression, or enzyme activity. Surprisingly, when perfused ex vivo, working hearts from 3- and 7-mo-old VLCAD null mice maintained physiological parameters similar to their age-matched littermate counterparts during the entire 30-min perfusion under all conditions examined. These included the presence of 1 mM palmitate, acute adrenergic stimulation, or hearts isolated from mice after a 24-h fast, a condition that has been linked to cardiac decompensation in patients with LCFA β-oxidation defects. Figure 1A reports rate pressure product values under all perfusion conditions. Other functional data on working hearts isolated from 3-mo-old fed or 24-h fasted mice, which are representative of data obtained under all conditions, are listed in Supplemental Table 4.
At the metabolic level, flux through β-oxidation, which is reflected by the contribution of exogenous palmitate to acetyl-CoA formation for citrate synthesis, was also similar between control and VLCAD null hearts under all conditions (Fig. 1B). In addition, 13C enrichment of palmitate in TGs was low, variable (between 1 and 3%, data not included) and did not significantly differ between the two groups, indicating similar partitioning of this LCFA between β-oxidation (energy production) and esterification to TGs (storage). It is noteworthy that when exogenous palmitate concentration was augmented to 1 mM, there was a three- to fourfold increase in the contribution of LCFA to acetyl-CoA formation in both control and VLCAD null hearts. This result clearly illustrates the capacity of VLCAD null hearts to respond to a LCFA challenge at least ex vivo. Finally, we also discerned no decrease in the contribution of LCFA to β-oxidation flux when 0.4 mM palmitate, a saturated LCFA, was replaced by 0.4 of unsaturated LCFA, either oleate or linoleate (data not included).
Further to our findings of normal β-oxidation fluxes and contractile function ex vivo, we expanded our lipidomic analyses to test for the presence of alterations 1) in LCFA partitioning between oxidation and storage in vivo, as evidenced by LCFA accumulation in TGs; and 2) the LCFA composition of PLs, specifically DHA, which is known to be antiarrhythmogenic (35). These analyses were conducted under various nutritional conditions in freeze-clamped hearts, but also in other tissues, to test whether the observed changes were cardiac specific.
VLCAD Null Mice Present Age- and Condition-Dependent Changes in LCFA Accumulation in TGs in Heart and Liver
Analyses of LCFA accumulation in TGs were conducted in 3- and 7-mo-old VLCAD mouse hearts freeze clamped after ex vivo perfusion under the aforementioned conditions or in situ in the fed state or after a 24-h fast or 2 wk of feeding of the HFD consisting of lard (58.4% calories from fat, predominantly palmitate, to mimic our ex vivo perfusion conditions). Compared with the controls, 7-mo-old VLCAD mouse hearts freeze clamped in situ manifested significantly higher levels (∼80%) of the most abundant LCFAs (oleate: C18:1n9 and linoleate: C18:2n6) in TGs in the fed state. Figure 2 reports data acquired with the standard diet, which are also representative of those obtained with the HFD (Fig. 2, A and B). At 3 mo of age, however, there was no difference in LCFAs between the two groups unless the hearts were freeze clamped after ex vivo perfusion with 0.4 mM palmitate (data not included).
In contrast to heart, we found an approximate twofold accumulation of LCFAs (C18:1n9; C18:2n6; and also palmitate: C16:0) in TGs from VLCAD null mouse livers compared with the controls, which occurred following fasting already at 3 mo of age (Fig. 2, C and D).
Factors underlying the various lipid alterations in VLCAD null mice.
We investigated a number of potential explanations for the observed condition-dependent changes in TGs, respectively, in the heart as well as differences in LCFA accumulation in TGs between the heart (at 7 mo in the fed state) and liver (at 3 mo in the fasted state) of VLCAD null mice. Firstly, in the heart, these values cannot be ascribed to transcriptional changes (i.e., decrease) of genes encoding for enzymes that may potentially compensate for VLCAD activity, namely, LCAD or ACAD9. In fact, although we noted a lower transcript level (25%) of Acadl in VLCAD null compared with littermate control mice at 7 (Fig. 3, A–D) and not at 3 (Supplemental Fig. 1) mo of age, this was observed for the fasting condition both on the standard (Fig. 3, A–D) and HFD (data not included). In contrast, already at 3 mo of age, VLCAD mouse livers displayed differences in Acadl transcript levels in the fasting condition (∼25% decreased; P < 0.001), albeit not for Acad9 (Fig. 3, E–H).
Secondly, we tested for alterations in signaling and metabolic pathways regulating TG accumulation. We observed no difference in the expression levels of several proteins (beclin-1, Atg12-Atg5, and Atg7) involved in autophagy, a process suggested to affect lipid metabolism (31, 48, 49) in livers from 3-mo-old fasted VLCAD−/− and control mice clamped in situ (Supplemental Fig. 2). Furthermore, we uncovered no evidence of ERK1/2 activation, two kinases that may regulate cardiac TG lipolysis through HSL activation by Ser600 phosphorylation (22, 29) in freeze-clamped livers from fed mice and hearts from fed and fasted mice (data not presented). However, the ratio of phosphorylated-to-total ERK1/2 was measurable in hearts perfused with 0.4 mM palmitate and found to be significantly reduced from 150 and 200%, respectively, in 7 mo-old, albeit not in 3-mo-old, VLCAD null hearts compared with the controls (P < 0.05; Fig. 4, A and B). It is noteworthy, however, that in these perfused hearts the phosphorylated Akt (Ser473) level (Supplemental Fig. 3) was similar in all groups. Supporting the notion of a potential decrease in TG hydrolysis in VLCAD null mouse hearts, we found that among gene encoding for enzymes involved in TG hydrolysis (Lipe and Pnpla2), the mRNA level for HSL (Lipe) was significantly decreased in 7-mo-old VLCAD null mouse hearts, albeit only in the fasted condition (Fig. 4, C and D).
VLCAD Mice Display a Cardiac Specific Decrease of DHA in PLs
One consistent finding of our targeted lipidomic analysis is a ∼20–31% lower proportion of DHA in PLs of VLCAD-deficient hearts in vivo after being fed the standard diet as well as after perfusion ex vivo and at all ages. Representative data on 3-mo-old mice are reported in Fig. 5A. Interestingly, HFD of 7-mo-old mice decreased DHA in PLs from 20 to 14% in VLCAD null mice (Fig. 5B). The proportion of other LCFAs in PLs displayed marginal age- and condition-dependent differences between groups of mice, despite marked changes in the proportion of some LCFAs, specifically stearate and arachidonate, following HFD in both groups (Supplemental Table 5). It is noteworthy that transcript levels of enzymes involved DHA biosynthesis from linolenic acid (Fads1 and Fads2) were similar between groups of mice fed the HF diet, except for a ∼45% increase for Fads2 in the fed state for VLCAD group (Fig. 5, C and D).
Our finding of a lower DHA level in heart PLs appears to be of great potential pathophysiological significance given the reported antiarrhythmogenic effects of ω-3 PUFAs and, more specifically, DHA (35). In this regard, because the heart seems to rely mainly on liver synthesis and export to maintain its DHA level, since it would be devoid of 2-elongase (3, 23), and a number of studies (17, 24, 39, 61) have emphasized the role of the liver in cardiac functional and metabolic abnormalities, we tested the possibility that VLCAD null mice exhibit lipid alterations also in this tissue. In contrast to heart, the hepatic LCFA profile in PLs and, consequently, the DHA percent were similar in both groups of mice under all conditions examined: 3- or 7-mo-old mice, fasted or fed the standard diet or the HFD (Table 1). A similar conclusion was reached for the level of DHA in total PLs of skeletal muscles and for total DHA level in plasma (Table 1), supporting the notion that the liver export of this LCFA was also not affected. It is noteworthy that PLs from mouse hearts were found to contain almost twofold more DHA than livers, concurring with data from Watkins et al. (59).
Further to our findings of a cardiac specific decrease in DHA in PLs, we tested the impact of a 5-wk FOD. This diet increased the cardiac level of DHA in PLs in both control (59%) and VLCAD (82%) groups of mice (Fig. 6). While there was still a ∼11% lower proportion of DHA in PLs of VLCAD mouse hearts, this difference did not reach significance (P = 0.06; n = 6–7/group).
The proportion of almost all other LCFAs in cardiac PLs was also markedly affected by the FOD compared with the standard diet in both groups of mice (Supplemental Table 5), and for some of them (C16:0; C18:0; C18:1n7; and C18:2n6), the proportion was significantly different between the VLCAD and control groups.
VLCAD Null Mouse Hearts Display Prolonged QTc Interval In Vivo Under All Conditions Examined
Finally, further to our finding of a cardiac specific decrease in DHA in PLs in VLCAD null mice and considering 1) the proposed role of DHA in modulating the electrical activity of the heart and that 2) arrhythmias are part of the cardiac phenotype of patients with LCFA oxidation defect (6) and VLCAD-deficient mice (14, 60), we conducted a telemetry study, which delivers precise ECG characterization in conscious animals without the impact of acute anesthesia (38). This was achieved in 7-mo-old VLCAD null mice and their littermate counterparts, which were consecutively fed a standard diet followed by either 1) a HFD for 2 wk or 2) FOD for 5 wk.
Under all conditions examined, both groups of mice maintained similar heart rate and R-R and QRS values (Table 2). The QTc interval, calculated with a modified Bazzett correction formula, QTc = QT/(RR/100)1/2 (38), varied between 45 and 55 ms, concurring with published values (38). VLCAD-deficient mice displayed, however, a significantly longer average QTc interval when fed a standard diet or HFD (∼11 and ∼17%, respectively; P < 0.05), which persisted with the FOD (∼24%; P < 0.05), suggesting that factors beyond DHA contribute to the QTc prolongation at least under our conditions.
In this study, we adopted both ex vivo and in vivo approaches to test the metabolic and functional capacity of VLCAD null mouse hearts under various conditions that mimic those believed to induce decompensation in human patients, namely, fasting, adrenergic stress, or high-fat loading. Firstly, our results demonstrate the ability of 3- and 7-mo-old VLCAD null mouse hearts perfused ex vivo to maintain normal values for 1) the various physiological parameters measured as well as 2) the contribution of exogenous palmitate, oleate, or linoleate (which are the most abundant LCFAs occurring in plasma) to β-oxidation under all conditions examined. These included 1) increasing the concentration of palmitate to 1 mM, a condition under which palmitate β-oxidation contributed 60% to energy production; 2) simulating a rise in energy demand combined with adrenergic stimulation; and 3) fasting conditions. While the results contrast with our previous findings in another model of LCFA oxidation defects, namely, peroxisome proliferator-activated receptor-α null mouse hearts (18), they are consistent with the notion that compensatory mechanisms of LCFA β-oxidation are present in VLCAD null mice, as suggested by studies in fibroblasts with radioactive oleate (9) and measurement of acyl-CoA dehydrogenase activity with palmitoyl-CoA in the hearts of these mice (10). However, the latter studies did not evaluate whether these compensatory mechanisms are able to maintain normal substrate flux in the intact, beating heart through β-oxidation as well as contractile activity, even under conditions of high-energy demand and LCFA supply. Here, we show that this is the case, at least ex vivo.
Additional data on LCFA accumulation in TGs in freeze-clamped hearts support the notion that such is also the case in the heart, as well as in the liver in vivo, although they also indicate that the efficiency of compensatory mechanisms for VLCAD deficiency may decrease under some conditions and differ between organs. Regarding the latter, it should be emphasized that both LCAD and ACAD9 have been shown to present activity that overlaps with VLCAD, although the specificity of all these enzymes towards the various LCFAs differs according to chain length or saturation vs. unsaturation level (9, 13). Interestingly, VLCAD and ACAD9, but not LCAD, were reported to interact with the mitochondrial membrane (13, 25, 50). However, based on a recent study (43) suggesting that ACAD9 is involved in the biogenesis of electron transport chain complex 1 rather than LCFA β-oxidation, LCAD is more likely to be the compensatory mechanism for VLCAD deficiency.
From our other data, we would like to emphasize the following notions. First, the observed age-, condition-, and tissue-dependent changes in LCFA accumulation in TGs (which reflect altered LCFA partitioning between β-oxidation and storage) as well as in Acadl transcript levels suggest that VLCAD deficiency is more efficiently compensated in the heart than in the liver. Second, our qPCR and Western blotting data suggest other potential mechanisms beyond reduced β-oxidation flux due to decreased efficiency of compensatory mechanisms, which may contribute to the condition- and age-dependent accumulation of LCFAs in TGs in VLCAD null mouse hearts. These include potential alterations in the activity of 1) enzymes or 2) signaling pathways involved in TG hydrolysis. This is suggested, respectively, by lower levels in VLCAD null mouse hearts of 1) transcripts for Lipe, which encodes for HSL (for review, see Ref. 7); and 2) phosphorylated ERK1/2, known to enhance HSL activity by Ser600 phosphorylation (22, 29) in working hearts perfused ex vivo, although further investigations are needed to explore these issues. Third, although VLCAD null mouse hearts and livers displayed age- and condition-dependent LCFA accumulation in TGs that was greater than in the controls, the two groups did not differ in their response to a HFD. This contrasts, to some extent, with a recent study (63) reporting that VLCAD null mice were resistant to high-fat-induced obesity, as evidenced by an approximately two- to fourfold decrease in TG levels in the liver and skeletal muscles in the fed state. Among possible explanations of this discrepancy are differences in the type of diet, which consisted of hydrogenated vegetable oils, compared with lard (i.e., from animal fats, predominantly palmitate, stearate, and oleate) in our study, which may differentially impact on the levels of compensatory mechanisms.
Beyond considerations about TG accumulation and compensatory mechanisms for VLCAD deficiency, one unexpected and consistent finding of our targeted lipidomic study was the lower levels of DHA in cardiac PLs from VLCAD-deficient mice. This was found both in hearts perfused ex vivo and freeze clamped in situ, under all conditions examined and at all ages, and, interestingly, this was the only parameter to be exacerbated by high-fat feeding in VLCAD null mice. Little is known about the mechanisms regulating DHA level in PLs in the heart. DHA is synthesized from its precursor linolenic acid, which is an essential PUFA, through a series of enzymatic steps that include delta-5 and delta-6 desaturases. However, the heart appears to rely mainly on liver synthesis and export via lipoproteins to maintain its DHA level since it would be devoid of 2-elongase (3, 23), although this view has been challenged recently (44).
In this regard, our findings of similar DHA levels in both groups of mice in hepatic PLs, as well as in skeletal muscle and in plasma, suggest that DHA exportation and trafficking from liver to the heart are unaffected in VLCAD null mice despite the fact that we did document some abnormalities in lipid accumulation in the liver of these mice, as discussed above. This notion if further supported by the fact that DHA level in PLs was increased by 82% when VLCAD null mice were fed the FOD. Our gene expression data for delta-5 and delta-6 desaturase along with the absence of accumulation of eicosapentanoic acid, a DHA precursor, do not support the notion of restricted cardiac DHA synthesis. Hence, collectively, these results point out to a cardiac-specific decrease in DHA level in PLs from VLCAD null mouse hearts. While the underlying mechanism remains to be clarified, potential explanations include increased DHA metabolism (through oxidation or conversion to bioactive metabolites) (11) and/or restricted incorporation of DHA into specific classes of phospholipids (27, 59).
Nevertheless, our finding of a cardiac specific reduction of DHA in PLs of VLCAD null mice appeared particularly important, given that DHA, which is a PUFA of the n-3 series, has numerous biological roles (55) as well as beneficial effects on the heart. Interestingly, DHA supplementation is recommended in patients with long-chain 3-hydroxyacyl-CoA dehydrogense deficiency (19, 36) to improve visual acuity (51). However, to the best of our knowledge, there is no specific recommendation for VLCAD-deficient patients. However, since plasma or blood DHA level is most commonly used as an indicator of its tissue level, a cardiac-specific decrease in DHA level cannot be excluded. Among its many biological roles, DHA is proposed to be antiarrhythmogenic and to reduce sudden cardiac death (32, 35), which is most relevant to cardiac symptoms reported in both VLCAD null mice (namely, arrhythmias; Refs. 14, 60) and patients with LCFA oxidation defects (i.e., arrhythmias and conduction defects, leading to sudden death). In this regard, we found that VLCAD null mice display prolonged QTc intervals (>10%) compared with their littermates under all conditions examined, as revealed by in vivo telemetry. This observation alone appeared to bear a strong pathophysiological significance given that prolonged QTc is considered to be an independent predictor of sudden cardiac death in humans (1). While little is known about the impact of QTc prolongation in mice, genetic murine models with long QT syndrome have been shown to be prone to inducible polymorphic ventricular tachycardia (26).
Interestingly, a potential cause-effect relationship between DHA and QTc interval is suggested by literature data. In fact, n-3 PUFA administration or a fish diet enriched in n-3 PUFAs decreases the likelihood of prolonged QT interval in dogs following ischemia (5) and in humans (40), respectively, while DHA supplementation (but not its precursor eicosapentaenoic acid) shortens QT interval in spontaneously hypertensive rats (45). However, in this study, feeding VLCAD null mice a FOD increased the cardiac level of DHA in PLs above that observed for controls under the standard diet, yet the prolonged QTc interval persisted. Potential explanations for the apparent discrepancy between our results and previous reports in dogs or rats include the absence under our basal condition of sympathetic stimulation (in contrast to what occurs during ischemia in dogs or with hypertension in rats), which is one mechanism by which DHA and n-3 PUFAs are proposed to exert their effect. This action appears to involve factors such as intracellular calcium homeostasis, inositol phosphate, protein kinase C signaling, and diacyglycerol (27, 41, 42). Interestingly, the latter are the specific substrate for HSL, for which we report decreased level of 1) mRNA and 2) stimulatory signaling stimulus (ERK1/2) in VLCAD mouse hearts, while abnormal calcium homeostasis is a hallmark of these mice (60). Hence, it appears warranted in future studies to 1) test for the impact of FOD on QTc under conditions that are known to precipitate arrhythmia to further clarify the exact mechanism(s) underlying prolonged QTc interval in VLCAD null mouse hearts, as well as 2) to assess DHA level in the various classes of PLs, particularly those potentially involved in cell signaling (27).
In conclusion, the results from this investigation demonstrate that despite normal exogenous cardiac LCFA β-oxidation and contractile function ex vivo, VLCAD null mice display other lipid alterations as well as a prolonged QTc interval in vivo. The reported age- and condition-dependent changes in LCFA accumulation in TGs and Acadl mRNA levels predict that VLCAD deficiency would be more efficiently compensated in the heart than in the liver. However, more importantly, our study highlighted lower DHA in PLs in VLCAD-deficient hearts at all ages and conditions examined, which appears to be independent of liver DHA supply and export via lipoproteins. Future experiments are warranted to clarify the exact cause and potential consequence of lower DHA in cardiac PLs as well as the exact mechanism underlying the QTc prolongation in VLCAD null mouse hearts.
This study was supported by the Canadian Institutes of Health Research (CIHR Grant #9575 to C. Des Rosiers), National Institute of Diabetes and Digestive and Kidney Diseases Grant RO1-DK-069752, and Studentships from the CIHR and the Department of Nutrition, Université de Montréal (to to R. Gélinas).
No conflicts of interest, financial or otherwise, are declared by the author(s).
Part of this work was presented at the International Society for the Study of Fatty Acids and Lipids Meeting in Maastricht in May 2010, at the Society for Heart and Vascular Metabolism Meeting in Boston in June 2008 and in Padova in August 2009, at the Mitochondrial Biology in Cardiovascular Health and Diseases Conference in Bethesda in October 2008, at the Experimental Biology Meeting in April 2009, and at the International Society for Heart Research American Section in Baltimore in May 2009. We thank Drs. C. Fiset, M.-C. Guertin, and V.J. Exil for helpful comments; N. Duquette and A. Sanscartier for assistance with animal care; and Ovid Da Silva and Luce Begin for editorial and secretarial assistance, respectively.
- Copyright © 2011 the American Physiological Society