The Cycling of Acetyl-Coenzyme A Through Acetylcarnitine Buffers Cardiac Substrate SupplyClinical Perspective
A Hyperpolarized 13C Magnetic Resonance Study
Background—Carnitine acetyltransferase catalyzes the reversible conversion of acetyl-coenzyme A (CoA) into acetylcarnitine. The aim of this study was to use the metabolic tracer hyperpolarized [2-13C]pyruvate with magnetic resonance spectroscopy to determine whether carnitine acetyltransferase facilitates carbohydrate oxidation in the heart.
Methods and Results—Ex vivo, following hyperpolarized [2-13C]pyruvate infusion, the [1-13C]acetylcarnitine resonance was saturated with a radiofrequency pulse, and the effect of this saturation on [1-13C]citrate and [5-13C]glutamate was observed. In vivo, [2-13C]pyruvate was infused into 3 groups of fed male Wistar rats: (1) controls, (2) rats in which dichloroacetate enhanced pyruvate dehydrogenase flux, and (3) rats in which dobutamine elevated cardiac workload. In the perfused heart, [1-13C]acetylcarnitine saturation reduced the [1-13C]citrate and [5-13C]glutamate resonances by 63% and 51%, respectively, indicating a rapid exchange between pyruvate-derived acetyl-CoA and the acetylcarnitine pool. In vivo, dichloroacetate increased the rate of [1-13C]acetylcarnitine production by 35% and increased the overall acetylcarnitine pool size by 33%. Dobutamine decreased the rate of [1-13C]acetylcarnitine production by 37% and decreased the acetylcarnitine pool size by 40%.
Conclusions—Hyperpolarized 13C magnetic resonance spectroscopy has revealed that acetylcarnitine provides a route of disposal for excess acetyl-CoA and a means to replenish acetyl-CoA when cardiac workload is increased. Cycling of acetyl-CoA through acetylcarnitine appears key to matching instantaneous acetyl-CoA supply with metabolic demand, thereby helping to balance myocardial substrate supply and contractile function.
The enzyme complex pyruvate dehydrogenase (PDH) is the key enzyme involved in balancing cardiac glucose and fatty acid oxidation, catalyzing the irreversible oxidation of pyruvate to form acetyl-coenzyme A (CoA), NADH, and CO2. Several PDH-mediated mechanisms exist to promote fatty acid oxidation over glucose oxidation, including phosphorylation inhibition of PDH1 and end product inhibition of PDH by high acetyl-CoA/CoA and NADH/NAD+ ratios.2 After eating and in response to energetic stressors, such as aerobic exercise or ischemia, glucose and other carbohydrates become important cardiac fuels.3,4 Additionally, mobilization of the heart's limited buffer stores (phosphocreatine, triacylglyceride, and glycogen) help to maintain ATP production for short periods of increased energetic demand or reduced substrate supply.5–9
Editorial see p 171
Clinical Perspective on p 209
Recently, we used hyperpolarized [2-13C]pyruvate as a metabolic tracer in the isolated perfused rat heart and observed the relationship between PDH flux and Krebs cycle metabolism in real time using 13C magnetic resonance spectroscopy (MRS).10–12 We observed conversion of [2-13C]pyruvate into the metabolites citrate and glutamate, with a substantial fraction of pyruvate oxidized by PDH being converted to acetylcarnitine.10 Acetylcarnitine is produced by carnitine acetyltransferase (CAT), an enzyme found in cardiac mitochondria13 that is stimulated by acetyl-CoA and carnitine.14 Myocardial acetylcarnitine levels vary widely with species and metabolic state. In rat, acetylcarnitine levels were >20-fold higher than that of acetyl-CoA, ranging from ≈180 nmol/g of frozen heart tissue in fat-fed rats to ≈380 nmol/g of frozen heart tissue in normal and carbohydrate-fed rats.15 In pigs, acetylcarnitine levels were between 4- and 10-fold greater than acetyl-CoA levels, ranging from 320 nmol/g dry weight on inhibition of fatty acid oxidation to ≈800 nmol/g dry weight in controls.16 The equilibrium for the acetyl transfer reaction is maintained despite different steady-state concentrations of each reactant, as shown in Equation 1 as follows15,17:(1)
Studies performed in isolated mitochondria and in rat heart tissue have suggested 3 physiological roles for CAT (Figure 1). First, formation of acetylcarnitine from acetyl-CoA has reduced the mitochondrial acetyl-CoA/CoA ratio,15,18 preventing end product inhibition of PDH2 and, importantly, liberating acetylated CoA for other mitochondrial functions.18 Second, acetylcarnitine formation enables transport of acetyl-CoA produced in the mitochondria into the cytosol17,19 for potential use as a source of malonyl-CoA for fatty acid synthesis17 or inhibition of fatty acid oxidation through carnitine palmitoyl transferase.20 Third, acetylcarnitine itself may serve as an acetyl-CoA buffer within cardiomyocytes.15,16
In patients with heart failure21,22 and in experimental pressure overload cardiac hypertrophy,23,24 carnitine deficiency has been implicated as a cause of disease, whereas carnitine supplementation has improved cardiac function. Our observation of significant levels of [1-13C]acetylcarnitine production from [2-13C]pyruvate in the intact heart was consistent with the in vitro observation that acetylcarnitine is produced readily from pyruvate-derived acetyl-CoA.25 When taken together, ex vivo hyperpolarized 13C MRS and in vitro results suggest that CAT and the acetylcarnitine pool may play an essential role in regulating cardiac energy metabolism. However, neither the dynamic interaction of the acetylcarnitine pool with acetyl-CoA nor the real-time response of the acetylcarnitine pool to in vivo changes in cardiac substrate supply and energy demand have been investigated.
The aim of this study was to use hyperpolarized [2-13C]pyruvate with high temporal resolution MRS to investigate the hypothesis that acetyl-CoA produced through PDH must cycle through CAT into the acetylcarnitine pool before incorporation into the Krebs cycle. To achieve this aim, we used a novel approach of saturation transfer hyperpolarized MRS to crush the [1-13C]acetylcarnitine resonance, thus determining how much [1-13C]acetylcarnitine dynamically generated from [2-13C]pyruvate is immediately incorporated back into the acetyl-CoA pool and, subsequently, into the Krebs cycle. The second aim was to characterize [2-13C]pyruvate metabolism in vivo with hyperpolarized MRS and to use this technique to monitor the acetylcarnitine pool in response to increases in pyruvate oxidation and cardiac workload. The results indicate that acetylcarnitine acts to fine-tune acetyl-CoA availability in the heart and that use of hyperpolarized [2-13C]pyruvate with MRS may enable noninvasive clinical diagnosis of cardiomyopathies involving carnitine deficiency.26
[2-13C]pyruvic acid 40 mg (Sigma), mixed with 15 mmol/L trityl radical (OXO63; GE-Healthcare) and a trace amount of Dotarem (Guerbet), was polarized and dissolved as previously described.10,11 Further details are presented in the online-only Data Supplement.
Saturation Transfer Experiments
We chose to perform the saturation transfer MRS studies in the isolated perfused heart instead of in vivo to ensure maximum signal-to-noise ratio from all metabolites derived from [2-13C]pyruvate and to minimize magnetic field inhomogeneity for improved spectral saturation. Five male Wistar rat hearts (heart weight, ≈1 g) were perfused in the Langendorff mode, as in a published protocol.10 The Krebs-Henseleit bicarbonate perfusion buffer used contained 11 mmol/L glucose and 2.5 mmol/L pyruvate and was aerated with a mixture of 95% O2 and 5% CO2 to give a final pH of 7.4 at 37°C. Three experiments were then performed in each perfused heart: (1) control, (2) acetylcarnitine saturation, and (3) control saturation (Figure 2). First, hyperpolarized [2-13C]pyruvate was diluted to 2.5 mmol/L and perfused into the heart, and 13C metabolites were detected, with 1-s temporal resolution over the course of 2 minutes (repetition time, 1 s; excitation flip angle, 30°; acquisitions, 120; sweep width, 180 parts/million [ppm]; acquired points, 4096; frequency centered at 125 ppm). The precise frequency of the [1-13C]acetylcarnitine resonance was noted.
Second, after a second dose of [2-13C]pyruvate was polarized (45 minutes), 2.5 mmol/L [2-13C]pyruvate was again delivered to the heart and 13C spectra acquired as before while radiofrequency saturation was continually applied throughout the experiment (described later) exactly to the [1-13C]acetylcarnitine resonance, as determined in the previous experiment (≈175 ppm).
Third, a third dose of 2.5 mmol/L [2-13C]pyruvate was delivered to the heart, and a control saturation experiment was performed. 13C spectra were acquired while saturation was continuously applied off resonance at ≈187 ppm. The location of this pulse replicated the frequency separation between the [1-13C]citrate and the [1-13C]acetylcarnitine peaks and ensured that the radiofrequency saturation pulse bandwidth was narrow enough to prevent direct saturation of the [1-13C]citrate resonance.
To achieve the continuous saturation at the appropriate resonance frequency, a cascade of 8 SNEEZE pulses27 were repeatedly applied. Each SNEEZE pulse had a time-bandwidth product of 5.82 and 100 ms duration, with a 100-μs interval between pulses. No saturation was applied during signal acquisition (180 μs). The effective full-width half-maximum bandwidth was ≈160 Hz when tested on a sample of acetone. An experiment in which the acetyl-CoA resonance was also saturated was not performed because the small [1-13C]acetyl-CoA resonance at ≈202 ppm (Figure 328) could not be saturated without also affecting the input [2-13C]pyruvate signal (207.8 ppm10).
Animal Handling and In Vivo MRS Experiments
Investigations conformed to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication no. 85–23, revised 1996); the Home Office Guidance on the Operation of the Animals (Scientific Procedures) Act, 1986; and institutional guidelines. Animals (weight, ≈300 g) were prepared as previously described.29 dichloroacetate (DCA) (n=6) 2 mg/mL per minute was dissolved into 1.5 mL of PBS, and the solution was infused intravenously starting 15 minutes before [2-13C]pyruvate infusion. Dobutamine (n=7) 0.1 mg/kg per minute was infused into rats intravenously starting 5 minutes before [2-13C]pyruvate infusion. In vivo cardiac hyperpolarized 13C MRS experiments were performed in a 7-T horizontal bore MR scanner as previously described,30 with further details presented in the online-only Data Supplement.
A 1H/13C butterfly surface coil was placed over the rat chest to localize signals to the heart. It is possible that this approach may have allowed for MR signal contamination from neighboring organs; however, because of the high rates of PDH and Krebs cycle flux in the heart compared with the liver, resting skeletal muscle, and blood, it is extremely likely that the MR signals measured from [1-13C]acetylcarnitine, [1-13C]citrate, and [5-13C]glutamate were predominantly from cardiac metabolism.29
MR Data Analysis
All cardiac 13C spectra were analyzed using the AMARES algorithm in the jMRUI software package.31 For acetylcarnitine saturation experiments, the first 50 spectra acquired following pyruvate arrival at the perfused heart were summed (50 s worth) to yield a single spectrum with peak heights for [1-13C]citrate, [5-13C]glutamate, [1-13C]acetylcarnitine, and [2-13C]pyruvate that were well above the level of noise. These single spectra were then quantified in jMRUI, and the resulting ratios of [1-13C]citrate/[2-13C]pyruvate, [5-13C]glutamate/[2-13C]pyruvate, and [1-13C]acetylcarnitine/[2-13C]pyruvate were used for subsequent analyses. Spectra were summed before analysis because acetylcarnitine saturation experiments had a low signal-to-noise ratio to the degree that accurate peak quantification of individual spectra was not feasible.
The peak areas of in vivo [2-13C]pyruvate, [1-13C]acetylcarnitine, [1-13C]citrate, and [5-13C]glutamate at each time point were quantified and used as input data for a kinetic model developed for the analysis of hyperpolarized [2-13C]pyruvate MRS data32,33 based on a model initially developed by Zierhut et al.34 First, the change in [2-13C]pyruvate signal over the 60-s acquisition time was fit to the integrated [2-13C]pyruvate peak area data using Equation 2, as follows:(2)
In this equation, Mpyr(t) was the [2-13C]pyruvate peak area as a function of time. This equation fit the parameters kpyr, the rate constant for pyruvate signal decay (s−1); rateinj, the pyruvate arrival rate (arbitrary units s−1); tarrival, the pyruvate arrival time (s); and tend, the time correlating with the end of the injections. These parameters were used in Equation 3, along with the dynamic [1-13C]acetylcarnitine, [1-13C]citrate, and [5-13C]glutamate data, to calculate kpyr→X, the rate constant for exchange of pyruvate to each of its metabolites (s−1) and kX, the rate constant for signal decay of each metabolite (s−1), which was assumed to consist of metabolite T1 decay and signal loss from the 7.5° radiofrequency flip angle pulses.(3)
In Equation 3, t′=t−tdelay, where tdelay was the delay between pyruvate arrival and metabolite appearance.
This model was implemented to facilitate comparisons among experimental conditions within each metabolite pool. The calculated rate constants kpyr→X reflect the rate of 13C accumulation in each metabolite pool, not necessarily the overall amount of production of each metabolite.
In Vitro Measurement of Acetylcarnitine Pool Size
A spectrophotometric assay was performed on heart tissue extracts to quantify acetylcarnitine levels in 4 separate groups of rats by the methods of Marquis and Fritz35 and Grizard et al.36 In one of these groups, acetylcarnitine levels were measured in normal heart tissue without pyruvate infusion (n=4). The other 3 groups were prepared identically to the groups of rats studied in vivo with hyperpolarized [2-13C]pyruvate: one group received 0.25 mmol/kg of aqueous pyruvate alone (n=7), a second group received DCA at 2 mg/mL per minute for 15 minutes before 0.25 mmol/kg pyruvate infusion (n=4), and a third group received dobutamine at 0.1 mg/kg per minute for 5 minutes before 0.25 mmol/kg pyruvate infusion (n=4). After each intervention, the heart from each rat was rapidly removed from the body cavity at a time point chosen to correspond with the time of maximum [1-13C]acetylcarnitine MR signal during in vivo data acquisition. Details of the protocols used for tissue dissection and for the acetylcarnitine assay are described in the online-only Data Supplement.
Values are presented as mean±SEM, and all analyses were performed with GraphPad Prism software. Statistical significance was considered at P<0.05.
For ex vivo MRS experiments in which 3 experiments were performed sequentially in the same heart, 1-way repeated-measures ANOVA with a Tukey multiple comparison test was used to identify statistical differences among groups. For in vivo MRS experiments and in vitro measurements of acetylcarnitine levels in which each experimental group comprised independent subjects, 1-way ANOVA was used to identify statistical differences. For the groups in which a significant difference was detected, t tests were performed with a Holm-Sidak correction for multiple comparisons.
Acetylcarnitine Saturation Experiments
In spectra acquired from the isolated perfused rat heart, the acetylcarnitine resonance was selectively saturated to crush the MR signal in all nuclei incorporated into the acetylcarnitine pool. The effect of acetylcarnitine saturation on the citrate resonance area qualitatively distinguished the fractions of [2-13C]pyruvate-derived acetyl-CoA that were (1) converted immediately into citrate, (2) converted into acetylcarnitine and stored, and (3) converted into acetylcarnitine before rapid reconversion to acetyl-CoA and incorporation into the Krebs cycle (Figure 2).
The acetylcarnitine saturation pulse reduced the acetylcarnitine peak area by 99.5±0.2% compared with control experiments in which no saturation was applied. Further, the control saturation experiments confirmed that the saturation pulse had a sufficiently low bandwidth to avoid radiofrequency effects on the citrate peak. The control saturation pulse did, however, affect the glutamate peak, which was 300 Hz closer to the control saturation pulse than was citrate (Figure 2).
Selective saturation of the acetylcarnitine resonance reduced the magnitude of the citrate resonance by 63% (Figure 2) compared with the control experiments in which no saturation was applied. When compared with the control experiments in which saturation was applied off resonance, acetylcarnitine saturation reduced the magnitude of the citrate resonance by 51%. This indicated that of all the [2-13C]pyruvate-derived acetyl-CoA incorporated into the Krebs cycle over 50 s, approximately one half was initially converted into acetylcarnitine before release from the acetylcarnitine pool as acetyl-CoA and incorporation into the Krebs cycle. The remaining [1-13C]citrate, unaffected by saturation, was produced from [2-13C]pyruvate-derived acetyl-CoA that immediately condensed with citrate synthase. This effect was also observed as a 51% reduction in the magnitude of the glutamate resonance compared with control experiments in which no saturation was applied.
In Vivo Metabolism of Hyperpolarized [2-13C]Pyruvate
Peaks relating to the infused [2-13C]pyruvate as well as to the metabolic products [1-13C]acetylcarnitine (175 ppm), [1-13C]citrate (181 ppm), [5-13C]glutamate (183 ppm), [2-13C]lactate (71 ppm), and [2-13C]alanine (53 ppm) were clearly observed in vivo. A broad, low signal-to-noise resonance corresponding with [1-13C]acetyl-CoA was detectable ≈202 ppm, as previously described.28 A representative spectrum is shown in Figure 3, with an example of the time course of the primary metabolic products.
The effects of increased PDH activity, caused by DCA infusion, or increased cardiac workload, caused by dobutamine infusion, on [1-13C]citrate, [5-13C]glutamate, and [1-13C]acetylcarnitine production are shown in Figure 4. In control rat hearts, the rate constant for the conversion of [2-13C]pyruvate to [1-13C]citrate (kpyr→cit) obtained from the in vivo data was 0.0012±0.0002 s−1. This value did not change with either metabolic intervention. The mean kpyr→glu in control rats was 0.0025±0.0003 s−1, and this did not change in response to DCA or dobutamine infusions.
In control rats, the rate constant for the production of [1-13C]acetylcarnitine (kpyr→AC) was 0.0075±0.0007 s−1, ≈3-fold higher than that of [5-13C]glutamate production and 6-fold higher than the rate constant of [1-13C]citrate production. DCA infusion caused a significant (35%) increase in the rate of 13C-label incorporation into the acetylcarnitine pool (kpyr→AC=0.010±0.001 s−1, P=0.043). Dobutamine infusion caused a 37% decrease in the rate of 13C-label incorporation into the acetylcarnitine pool (kpyr→AC=0.0047±0.0007 s−1, P=0.014).
Spectra acquired with 0.5 s temporal resolution revealed that the 13C label appeared (defined as the time tdelay) in the [5-13C]glutamate pool 3.1±0.4 s after the arrival of [2-13C]pyruvate and 1.6 s after the 13C label appeared in the [1-13C]citrate pool. There were no differences between the time of appearance of [1-13C]citrate at 1.5±0.5 s and [1-13C]acetylcarnitine at 2.0±0.3 s.
Acetylcarnitine in Cardiac Extracts
Infusion of 0.25 mmol/kg pyruvate, using the same protocol as was used for in vivo hyperpolarized 13C MRS experiments, increased myocardial acetylcarnitine levels by 76% compared with tissue that received no pyruvate (from 230±60 to 400±40 nmol/g wet weight tissue) (Figure 5A). In rats that had received 0.25 mmol/kg pyruvate and were thus identical to the rats studied in vivo with hyperpolarized 13C MRS (Figure 5B), DCA infusion caused a significant (33%) increase in the size of the myocardial acetylcarnitine pool (540±30 nmol acetylcarnitine/g wet weight). Dobutamine infusion caused a 40% decrease in the size of the myocardial acetylcarnitine pool (240±20 nmol acetylcarnitine/g wet weight).
The maintenance of near equilibrium in the CAT reaction such that myocardial acetylcarnitine content remains approximately one order of magnitude greater than myocardial acetyl-CoA content suggests both high enzyme activity and a role for acetylcarnitine as an acetyl-CoA buffer. Therefore, in this study, we hypothesized that CAT was dynamically involved in pyruvate oxidation by means of rapid cycling of acetyl-CoA generated from pyruvate through the acetylcarnitine pool. Use of saturation transfer 13C MRS with hyperpolarized [2-13C]pyruvate uniquely provided the temporal resolution necessary to test our hypothesis in the ex vivo heart.10
Saturating the MR signal of the [1-13C]acetylcarnitine resonance reduced the magnitudes of the Krebs cycle-derived [1-13C]citrate and [5-13C]glutamate resonances by 63% and 51%, respectively, over a 50-s period in the isolated perfused heart. Therefore, in the healthy heart, approximately one half of the PDH-derived acetyl-CoA that contributes to ATP production cycles through CAT into the acetylcarnitine pool before reconversion to acetyl-CoA and incorporation into the Krebs cycle. Constant acetylcarnitine/acetyl-CoA turnover may allow the cardiomyocyte to respond to multiple stimuli that may affect mitochondrial acetyl-CoA content, analogous to rapid turnover of other energetic buffer pools. For example, cardiac triacylglyceride turnover maintains homeostasis of free fatty acids in the cytosol,6,37 and the rapid conversion of phosphocreatine to ATP prevents altered cardiac ATP content.9
The second aim of this study was to determine whether hyperpolarized [2-13C]pyruvate metabolism could be followed in vivo with MRS and, if so, to use the technique to monitor the acetylcarnitine pool in response to altered substrate supply and energy demands. Using a protocol similar to that applied previously in living rats to follow hyperpolarized [1-13C]pyruvate metabolism,29 we detected the conversions of [2-13C]pyruvate into [1-13C]citrate, [5-13C]glutamate, and [1-13C]acetylcarnitine with 0.5-s temporal resolution. A delay between the appearance of 13C in the citrate and glutamate pools of ≈1.5 s was measured, marking the first, to our knowledge, in vivo assessment of instantaneous rates of flux though the first span of the Krebs cycle and the oxoglutarate-malate carrier. The result indicates the potential utility of our hyperpolarized 13C MRS protocol to contribute to real-time metabolic models.
A limitation of the applied protocol was that the infused pyruvate dose raised plasma pyruvate levels to ≈250 μmol/L (25% higher than physiological levels) 30 s after infusion.33 This increased PDH flux, manifested as a 76% increase in cardiac acetylcarnitine content to 400 nmol/g wet weight (a value still within the physiological range). However, the elevation in plasma pyruvate was minor compared with the physiological disturbances typical of ex vivo or in vitro preparations. Further, physiological acetylcarnitine levels vary widely under normal physiological conditions,15,16 and the in vivo DCA and dobutamine experiments confirmed that the acetylcarnitine pool remained sensitive to metabolic perturbation.
Infusion of DCA, an acetate analog that stimulates PDH activity without affecting cardiac workload,38 caused a 35% increase in the production of [1-13C]acetylcarnitine from [2-13C]pyruvate alongside a 33% increase in the myocardial acetylcarnitine pool size measured in frozen tissue. However, despite the increase in PDH flux and resultant increase in 13C incorporated into the acetyl-CoA pool,33 the rate of 13C-label incorporation into [1-13C]citrate and [5-13C]glutamate was unchanged. Taken together, these observations suggest that in vivo, if acetyl-CoA is formed from PDH in excess of the energetic demand, then mitochondrial CAT continually recovers free CoA while allowing for the disposal of the excess acetyl moieties to maintain a low acetyl-CoA/CoA ratio (Figure 6A). To our knowledge, this is the first study to demonstrate CAT-mediated acetyl-CoA disposal in real time and in vivo.
Dobutamine, a selective β-agonist known to increase flux through PDH and the Krebs cycle,39,40 was used to acutely increase myocardial workload in rats in vivo. Five minutes after in vivo dobutamine infusion, we observed a 37% decrease in the production of [1-13C]acetylcarnitine from [2-13C]pyruvate alongside a 40% decrease in the size of the myocardial acetylcarnitine pool itself. The decreased rate of [1-13C]acetylcarnitine production and acetylcarnitine pool size, particularly in the context of increased PDH flux,39 strongly indicated that acetylcarnitine itself was consumed in vivo to buffer myocardial acetyl-CoA levels in response to increased ATP demand (Figure 6B).
A second limitation to this study was that acetyl-CoA and free CoA levels were not measured alongside acetylcarnitine levels because of the technical difficulties in accurately assessing CoA ester content in in vivo myocardial tissue. In the future, our conclusions regarding the role of acetylcarnitine to reduce acetyl-CoA/CoA could be confirmed by measuring these quantities in response to pathophysiological perturbations of cardiac substrate supply and energy demands.
Also of interest was the unaltered production rate of [1-13C]citrate and [5-13C]glutamate following dobutamine infusion, despite the documented and observed effects of the β-agonist on PDH, CAT, and Krebs cycle fluxes. A previous steady-state, ex vivo 13C MRS study revealed that following dobutamine stimulation, 13C-glutamate production was reduced because of elevated flux through the Krebs cycle enzyme α-ketoglutarate dehydrogenase compared with flux through the oxoglutarate-malate carrier.40 A similar mechanism may have influenced our instantaneous hyperpolarized 13C MRS measurements, preventing alteration to the observed [5-13C]glutamate production despite the increased Krebs cycle flux. In addition, balanced increases to [1-13C]citrate production by citrate synthase and [1-13C]citrate consumption by the next Krebs cycle enzyme aconitase may have masked any differences to alterations in the rate of [1-13C]citrate production. Further mathematical modeling is warranted to fully characterize how various pathophysiological conditions affecting Krebs cycle and oxoglutarate-malate carrier fluxes also influence the [1-13C]citrate and [5-13C]glutamate resonances.
To summarize, the results of the present study suggest that acetyl-CoA produced from pyruvate dynamically cycles through the acetylcarnitine pool, fine-tuning the instantaneous acetyl-CoA supply from carbohydrates to match Krebs cycle demand (Figure 6C). This buffering may enable acetyl moiety storage to encourage constant rates of oxidative ATP production. Further, acetyl-CoA:acetylcarnitine cycling may prevent static acetyl-CoA accumulation and its associated inhibitory effects.
Carnitine deficiency has been reported in numerous pathophysiological conditions in which cardiac energetics are also depleted, including old age,41 pressure overload hypertrophy,23 and heart failure.21,42 Generally, it has been assumed that carnitine deficiency compromises myocardial energy metabolism by limiting fatty acid translocation across the mitochondrial membrane, thus decreasing fatty acid oxidation.24 However, there is preliminary evidence that carnitine deficiency may also have a negative impact on glucose oxidation. For example, in the aged heart, acetyl-l-carnitine supplementation increases rates of overall ATP production, pyruvate uptake, and mitochondrial pyruvate oxidation,41 whereas in the hypertrophic heart, proprionyl-l-carnitine supplementation improves contractility by means of a 2-fold increase in the contribution of glucose oxidation to overall ATP production.24 The present work, alongside studies of carnitine depletion and repletion in disease, suggests that acetyl-CoA:acetylcarnitine cycling fine-tunes mitochondrial acetyl-CoA availability in the heart, which appears key for the maintenance of optimal cardiac function and energetics.
We have demonstrated that hyperpolarized [2-13C]pyruvate metabolism can be measured in vivo with subsecond temporal resolution. Further studies with hyperpolarized [2-13C]pyruvate and MRS may help to clarify the mechanism by which carnitine deficiency contributes to heart failure in patients21 and, potentially, to enable noninvasive clinical diagnosis of cardiomyopathies involving carnitine deficiency.26 The impact of understanding and diagnosing carnitine deficiency in humans could be high, given the ease by which treatment, by normalizing carnitine stores through oral supplementation, can already be achieved.22
Sources of Funding
This work was supported by the National Institutes of Health (grant 1-F31-EB006692-01A1), the Wellcome Trust (a Sir Henry Wellcome Postdoctoral Fellowship to Dr Schroeder), the Medical Research Council (grant G0601490), the British Heart Foundation (grants FS/10/002/28078 and PG/07/070/23365), and GE-Healthcare.
Dr Tyler has received research support from GE-Healthcare.
We thank Drs Richard Veech, Rhys Evans, and Jan Henrik Ardenkjær-Larsen for their helpful suggestions.
The online-only Data Supplement is available with this article at http://circimaging.ahajournals.org/lookup/suppl/doi:10.1161/CIRCIMAGING.111.969451/-/DC1.
- Received September 8, 2011.
- Accepted December 22, 2011.
- © 2012 American Heart Association, Inc.
- Linn TC,
- Pettit FH,
- Reed LJ
- Cross HR,
- Opie LH,
- Radda GK,
- Clarke K
- Saddik M,
- Lopaschuk GD
- Kreisberg RA
- Kreisberg RA
- Schroeder MA,
- Atherton HJ,
- Ball DR,
- Cole MA,
- Heather LC,
- Griffin JL,
- Clarke K,
- Radda GK,
- Tyler DJ
- Ardenkjaer-Larsen JH,
- Fridlund B,
- Gram A,
- Hansson G,
- Hansson L,
- Lerche MH,
- Servin R,
- Thaning M,
- Golman K
- Golman K,
- in 't Zandt R,
- Thaning M
- Pearson DJ,
- Tubbs PK
- Fritz IB,
- Schultz SK,
- Srere PA
- Lysiak W,
- Lilly K,
- DiLisa F,
- Toth PP,
- Bieber LL
- Fritz IB,
- Yue KT
- Saddik M,
- Gamble J,
- Witters LA,
- Lopaschuk GD
- Schonekess BO,
- Allard MF,
- Lopaschuk GD
- Lysiak W,
- Toth PP,
- Suelter CH,
- Bieber LL
- Schroeder MA,
- Clarke K,
- Neubauer S,
- Tyler DJ
- Jensen PR,
- Peitersen T,
- Karlsson M,
- In 't Zandt R,
- Gisselsson A,
- Hansson G,
- Meier S,
- Lerche MH
- Schroeder M,
- Cochlin L,
- Heather L,
- Clarke K,
- Radda G,
- Tyler D
- Atherton HJ,
- Dodd MS,
- Heather LC,
- Schroeder MA,
- Griffin JL,
- Radda GK,
- Clarke K,
- Tyler DJ
- Marquis NR,
- Fritz IB
- Grizard G,
- Lombard-Vignon N,
- Boucher D
- O'Donnell JM,
- Zampino M,
- Alpert NM,
- Fasano MJ,
- Geenen DL,
- Lewandowski ED
- O'Donnell JM,
- Kudej RK,
- LaNoue KF,
- Vatner SF,
- Lewandowski ED
- Shigenaga MK,
- Hagen TM,
- Ames BN
Increasingly, changes in metabolic substrate use and depleted myocardial energetics are considered to be causes, rather than consequences, of heart failure. Recent experimental studies have demonstrated that monitoring hyperpolarized 13C-labeled tracers with magnetic resonance spectroscopy offers a new, noninvasive method to investigate cardiac metabolism, and that the technology may be useful to determine diagnosis and prognosis and to optimize management of patients with heart disease. In the present study, we used the metabolic tracer hyperpolarized [2-13C]pyruvate with magnetic resonance spectroscopy to show that in healthy hearts, acetyl-coenzyme A formed from pyruvate rapidly cycles through the mitochondrial acetylcarnitine pool. Further, we demonstrated that the acetylcarnitine pool functions in vivo to continually fine-tune mitochondrial acetyl-coenzyme A levels, providing an overflow pool to prevent potentially deleterious acetyl-coenzyme A accumulation and a source of energetic substrate in response to rapidly increased cardiac workload. Many physiological and pathological conditions, including aging, hypertrophy, and heart failure, are characterized by depleted myocardial carnitine stores and energy depletion. Noninvasive assessment of the acetylcarnitine pool and other metabolic processes using hyperpolarized 13C magnetic resonance spectroscopy may advance our understanding of the importance of carnitine in heart failure, related abnormalities in heart failure, and targeted metabolic treatments.