Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Magn Reson Med. Author manuscript; available in PMC 2014 January 1.
Published in final edited form as:
PMCID: PMC3767464

Dynamic Monitoring of Carnitine and Acetylcarnitine in the Trimethylamine Signal after Exercise in Human Skeletal Muscle by 7T 1H MRS


A trimethylamine moiety is present in carnitine and acetylcarnitine, and both molecules play critical roles in muscle metabolism. At 7 T the chemical shift dispersion was sufficient to routinely resolve the trimethylamine signals from carnitine at 3.20 and acetylcarnitine at 3.17 ppm in the 1H MRS of human soleus muscle with a temporal resolution of about 2 minutes. In healthy, sedentary adults, the concentration of acetylcarnitine increased nearly 10-fold, to 4.1 ± 1.0 mmol/kg, in soleus muscle after 5 minutes of calf-raise exercise and recovered to a baseline concentration of 0.5 ± 0.3 mmol/kg. While the half-time for decay of acetylcarnitine was the same whether measured from the trimethylamine signal (18.8 ± 5.6 min) or measured from the methyl signal (19.4 ± 6.1 min), the detection of acetylcarnitine by its trimethylamine signal in soleus has the advantage of higher sensitivity and without overlapping from lipid signals. Although the activity of carnitine acetyltransferase is sufficient to allow equilibrium between carnitine and coenzyme-A pools, the exchange in trimethylamine signal between carnitine and acetylcarnitine is slow in soleus following exercise on 7T 1H NMR time scale. The trimethylamine signal provides a simple and direct measure of the relative amounts of carnitine and acetylcarnitine.

Keywords: carnitine, carnitine acetyltransferase, skeletal muscle, exercise, acetylcarnitine, MRS


Carnitine plays a crucial role in skeletal muscle metabolism. Its participation in translocation of long-chain fatty acids from the cytosol into the mitochondrial matrix followed by β oxidation and metabolism of acetyl-CoA in the citric acid cycle has been exhaustively studied (1). More recently the role of carnitine as a buffer of acetyl groups has received attention. Carnitine acetyltransferase (EC2.3.1.7) (CAT), a monomeric enzyme bound to the inner mitochondrial membrane, reversibly catalyzes the transfer of acetyl groups between carnitine and coenzyme A (13). This process plays an essential role in muscle energetics by buffering acetyl groups generated at varying rates from various sources (Figure 1) (4,5). During strenuous exercise the rate of generation of acetyl groups may considerably exceed the oxidative capacity of muscle, thereby depleting free CoASH (2). However, free CoASH serves as a substrate for two critical processes in exercising muscle, the continuous oxidation of fats and carbohydrates to acetyl-CoA, and oxidative decarboxylation of α-ketoglutarate to succinyl-CoA. Given the low concentration of CoASH, ~10 µmol/kg ww, even a small mismatch between the rate of generation of acetyl groups and the rate of consumption in the citric acid cycle would substantially alter AcCoA/CoASH ratio. A very low concentration of free CoASH presumably would arrest citric acid cycle flux (69).

Figure 1
Role of Carnitine Acetyl Transferase in Skeletal Muscle Energetics

In addition to preserving a pool of free CoASH, other important effects of CAT in fatty acid and carbohydrate oxidation have been described. For example, mediated by CAT, acetylation of carnitine reduces availability of free carnitine, potentially inhibiting the transport of long-chain fatty acid into mitochondria for β-oxidation (10). Since flux through multiple enzymes in glycolysis is inhibited by a high AcCoA/CoASH ratio, transfer of acetyl groups from AcCoA to carnitine may increase glucose oxidation (11). Finally, the availability of acetyl groups in acetylcarnitine provides a pool for rapid regeneration of acetyl-CoA and its oxidation in the citric acid cycle (5,1214).

The ability to continuously monitor carnitine and acetylcarnitine would be important not only because of the significance of carnitine in muscle metabolism but also because the ratio acetylcarnitine/carnitine may provide an index of AcCoA/CoASH. After exercise, Kreis et al (15) observed a 1H NMR signal assigned to the methyl protons of acetylcarnitine at 2.13 ppm. Interestingly, in the same study, the trimethylamine (TMA) signal of carnitine shifted slightly upfield after exercise consistent with an assignment to acetylcarnitine. Based on the known turnover number of CAT (29,000 moles of substrate transformed per minute per mole of enzyme, at pH 7.8 and 30°) and the content of CAT in muscle (0.3 – 3.2 µmol/kg (16,17)), the rate constant (k) for carnitine ↔ acetylcarnitine exchange would be in the range of 0.015 – 0.16 s−1 for a high substrate concentration of 10 mmol/kg, or 0.3 – 3.2 s−1 for a low substrate concentration of 0.5 mmol/kg, which should always satisfy the slow exchange regime for 1H NMR observations at fields used for in vivo studies. This conclusion has two implications. First, the absence of a resolved TMA signal assignable to acetylcarnitine separate from carnitine is due to the relatively broad lines relative to the chemical shift difference at 1.5 T. Second, resolution of two TMA signals is consistent with rapid exchange in the acetylcarnitine-carnitine pool.

The purpose of this study was to determine whether both acetylcarnitine and carnitine could be resolved in the TMA signal at 7T, and to compare the rate constant for recovery of acetylcarnitine measured directly from the methyl signal at 2.13 ppm to the same measurement from the TMA signal. In this study, the soleus of 20 healthy sedentary subjects was studied at 7T after relatively simple but intense exercise. The TMA signal was resolved in two components assigned to acetylcarnitine and carnitine. The rate constant for disappearance of acetylcarnitine during recovery, ~ 19 min, was the same using analysis of either the acetyl signal or the TMA signal. This internal consistency supports earlier assignments (15). However, in comparison to acetyl signal, the detection of acetylcarnitine by its TMA signal has the advantage of high sensitivity, and virtually without chemical shift displacement artifact and overlapping from lipid signals in human soleus muscle. The trimethylamine signal provides a simple and direct measure of the relative amounts of carnitine and acetylcarnitine.

Materials and Methods


Carnitine, acetylcarnitine and other biochemicals were purchased from Sigma-Aldrich, St. Louis, MO.

Human Subjects

The protocol was approved by the Institutional Review Board of the University of Texas Southwestern Medical Center. Prior to the MRS study, informed written consent was obtained from all 20 participants. All subjects,, young (20 – 35 yr) and middle aged (36 – 56 yr), are healthy, sedentary adults with no history of peripheral vascular, systemic or myopathic diseases. Other details are summarized in Table 1. To avoid possible exercise-associated physiological variations among subjects, all subjects were asked to refrain from any physical exercise other than walking for two days prior to the study, and subjects sat at rest for 20 minutes prior to the study. Subjects were instructed to move slowly in the scan room. The entire scanning session including the pre-exercise data acquisition, exercise and post-exercise data acquisition was 90 min or less and was tolerated by all subjects. Blood oxygen saturation level and heart rate were monitored during the scan. All subjects were interviewed after the exam and again at 24 hours after the exam. There were no reports of headaches, vertigo or related neurological complaints at 24 hours, but all subjects reported muscle soreness in the exercised leg lasting up to 2 days after exercise.

Table 1
Anthropometric measurements of the subjects studied (n = 20)

Exercise Protocol

Subjects were asked to perform brief warm-up stretching lasting less than two minutes and without significant subjective exertion. After a pre-exercise MRS scan, subjects performed standing single-leg calf-raise exercise. The exercise routine consisted of 75 repetitions of calf lift-and-return for a total exercise time of 5 min. The subject stood with the ball of the left foot on a ~1” block of wood. The right foot was placed behind the left ankle and the subject placed both hands on a wall for balance. The subject lifted himself or herself by plantar flexion and then returned to the rest position with the heel on the ground in a 4 second cycle. Immediately after the exercise, subjects were placed back in scanner at the pre-marked original position for a series of post-exercise dynamic 1H MRS scans.


All subjects were positioned supine in a 7T system (Achieva, Philips Medical Systems, Cleveland, OH). The left leg of each subject was positioned parallel to the magnetic field with the foot constrained in a soft stabilizer, and the location of the coil center line relative to the leg muscle was marked in the pre-exercise scanning session for repositioning reference in the post-exercise session. 1H MRS spectra were acquired with a partial volume quadrature transmit-receive coil (Philips Medical Systems, Cleveland, OH). The coil uses a butterfly configuration with two tilted, partially overlapping loops (~10 cm each, with 10% cross overlapping), customized to fit the shape of a human calf. Axial, coronal, and sagittal T2-weighted turbo spin echo images were initially acquired of the left calf muscle. Typical parameters were: field-of-view (FOV) 180 × 180mm, time of repetition (TR) 2 s, echo time (TE) 75 ms, turbo factor 16, in-plane spatial resolution 0.6 × 0.7 mm2, slice thickness 5 mm, gap 3 mm, bandwidth 178 Hz, number of acquisitions (NA) one, and acquisition time 2 min. The T2w images provided clear visualization of fasciae separating different muscle groups which allows the placement of the voxel exclusively within the soleus. Skeletal muscle size was measured from the selected area-of-interest on the T2w image that was acquired cross the calf girth (largest overall cross sectional area). The measurement was done on soleus, gastrocnemius lateral and medial (Table 1).

Single-voxel 1H MR spectra were collected from soleus with a typical volume of 10.7 ± 3.5 mL (anterior-posterior, 9 – 20 mm; right-to-left, 15 – 25 mm; foot to head, 27 – 35 mm), using a STEAM sequence. The placement of the voxel was guided by T2w images so that the geometry and orientation of the voxel were adjusted to fit the shape of subject’s soleus muscle and to avoid fat tissues and the boundaries of muscle. The 2nd order pencil-beam or iterative VOI shimming and two dummy scans were applied prior to spectral data collection. New shimming was carried out for the first post-exercise dynamic scan (~ 2 min). The final shimming yielded a water signal linewidth LW1/2 of 16.6 ± 3.1 Hz (in the range of 12 – 22 Hz, n = 20) for pre-exercise, and 13.5 ± 2.1 Hz (in the range of 10 – 20 Hz, n = 20) for post-exercise, respectively.

Parameters for dynamic scans were: TR, 2 s; TE, 140 ms; spectral bandwidth, 4 kHz; 16-step phase cycling; number of points, 4096; number of dummy acquisition, 4; number of data acquisitions (NA), 64; and net data acquisition time, 2.1 min. Spectra were zero-filled to 8192 prior to Fourier transformation with apodization filtering parameters adjusted to optimize resolution in the TMA region. The total number of scans run per subject was 10 – 31, depending on individual’s acetyl-carnitine decay process. The time from the end of the exercise to the start of data acquisition ranged from 2.4 to 3.5 min among subjects (average 3.0 ± 0.2), and was standardized to 3 min for data comparison by extrapolation from the dynamic data fit. The RF pulse used for acquisition had a flat excitation profile with a uniformly excited window in the chemical shift region of interest (from 0.5 to 6.0 ppm). The spectra were acquired without water suppression.

The T1 and T2 values of carnitine, acetylcarnitine and creatine 1H signals were measured using inversion recovery and varying TE methods, respectively. To correct for the effect of dynamic concentration change for post-exercise study, calibration spectra (four data points) were acquired prior to, and after the T1 and T2 scans for evaluation of rate of decay/increase (see below). The scans were done with TR 4 s, TE 140 ms, and NA 16. For T1 measurement, the inversion bandwidth was 3 kHz, and inversion delay time TI was varied in the range 100 – 3000 ms (seven data points). For T2 measurement, the TE was varied from 60 to 300 ms (five to eight data points). Signal intensities were used to evaluate T1 and T2 relaxation times, with correction for the effect of signal partial saturation, as well as the decrease and increase of acetylcarnitine and carnitine, respectively, occurring during post-exercise recovery. The T1 and T2 measurements were performed on four subjects before and after a single bout of exercise (n = 4), two of them also participated in the main group study with the uniformed acquisition protocol (TR = 2 s and TE = 140 ms, NA 64).

To evaluate the 1H MRS detection limit for acetylcarnitine, a mixture of carnitine and acetylcarnitine was prepared in a 500 mL sphere-shaped flask containing 10 mM carnitine and varied acetyl-carnitine at concentration of 0.1, 0.25, 0,5, 0.75, 1.0 and 2.0 mM and at pH ~7.4. The samples were scanned on a voxel size of 5.1 ml (19.1 × 18.7 × 14.3 mm3), using the same coil and sequence parameters (TR 2 s, TE 140 ms, NA 64) as in the in vivo study but with water-suppression applied to overcome the artifacts related to large unsuppressed water signal.

The 1H MRS phantom sample with 10 mM carnitine and 5 mM acetylcarnitine in 500 ml aqueous solution was also used to measure relaxation times with the same coil and the methods as in human calf study. The scan parameters for T1 relaxation time measurement by inversion-recovery method was TR 9 s, TE 140 ms and TI varied between 100 and 6500 ms (eight data points). The T2 relaxation times were measured using varying TE method (eight data points from 60 to 1000 ms) with TR 9 s. High-resolution 1H NMR spectra of mixtures of carnitine and acetylcarnitine (50 mM, 1:1) at pH 6.4 and 7.4 were also collected on 400 MHz Varian Inova spectrometer at 25°C and 37°C to examine the possible effect of temperature and pH on TMA chemical shift difference between carnitine and acetylcarnitine.

Analysis of Spectra

The 1H chemical shifts of all metabolite resonances in the muscle were referenced to creatine -CH3 protons set to 3.02 ppm. The area of each metabolite 1H resonance signal was determined by fitting the spectrum using ACD software (Advanced Chemistry Development, Inc., Toronto, Canada), as reported previously (18). The ratio acetylcarnitine/total carnitine or facetylcarnitine was determined from the TMA signals of free carnitine at 3.20 ppm relative to acetylcarnitine at 3.17 ppm, after correction for differences in relaxation times.

The time-course data for decay of the acetylcarnitine TMA signal were fit to a mono-exponential equation

equation M1

in which c(t), represents the concentration of acetylcarnitine at any time point t, co is acetylcarnitine at the assumptive time point t = 0 (the time when the exercise was stopped), and cr represents the residual acetylcarnitine at t → ∞, τ is the decay constant. The concentration of carnitine and acetylcarnitine was referenced to the signal of the creatine methyl resonance at 3.02 ppm, which has been widely used as an internal standard (as 30 mmol/kg wet weight) (19), with correction for the differences in T1 and T2 values between the creatine methyl Cr3 (3 protons) and acetylcarnitine TMA (9 protons).

The initial forward acetylcarnitine depletion rate Rf(0) (toward conversion of acetylcarnitine to free carnitine) was calculated by the following equation

equation M2

Dividing by total carnitine concentration (carnitine + acetylcarnitine), Equation [1] can be rewritten to give the acetyl fraction, facetylcarnitine = acetylcarnitine/(carnitine + acetylcarnitine):

equation M3

in which f0 and fr denote the faction of acetylcarnitine in total carnitine pool at the beginning (t = 0) of post-exercise recovery and at infinite time (t = ∞), respectively.

Statistical analysis

All data regression and correlation analysis was done with Microsoft Excel. The correlation of anthropometric data with post-exercise acetylcarnitine concentration were done using TTEST software package (Microsoft Excel) to evaluate the possible effect of gender, age and BMI, based on the p-values (using type-two variance and one-tail distribution).


Assignment of Acetylcarnitine

Two resonances, not present or not obviously visible at rest, were significantly elevated in soleus muscle after calf-raise exercise (Figures 2 and and3).3). The signal at 2.12 ppm matches the resonance of the acetyl group of acetylcarnitine reported by Kreis et al (15); the other signal at 3.17 ppm agrees with the chemical shift of TMA resonance of acetylcarnitine in aqueous solution at 400 MHz (Figure 3d, also see Figure 2 in ref. 41). This latter signal is clearly distinguishable from the TMA resonance of free carnitine at 3.20 ppm (Figure 2a) and was not previously resolved from the TMA signal of free carnitine in human skeletal muscle (15). The appearance of the acetylcarnitine TMA peak at 3.17 ppm was accompanied by a corresponding decrease in the intensity of free carnitine TMA peak at 3.20 ppm (Figures 2 and and3),3), as anticipated from conversion of free carnitine to acetylcarnitine (Figure 1).

Figure 2
Single-voxel 7T 1H MR Collected from Soleus Muscle
Figure 3
Single-voxel 7T 1H MR from a Subject and a Phantom

The high-resolution NMR spectrum of a mixture of carnitine and acetylcarnitine is shown in Figure 3a. A change in either pH (from pH 7.4 to 6.4) or temperature (from 37°C to 25°C) did not measurably alter the chemical shift difference between carnitine and acetylcarnitine TMA (Δδ = 0.03 ppm) or between acetylcarnitine TMA and acetyl resonances (Δδ = 1.08 ppm).

Measurements of T1 and T2

The relaxation times of the TMA group of carnitine and acetylcarnitine were similar (Table 2). The TMA T2 of carnitine was about 4-fold shorter in resting soleus muscle than in aqueous solution, while T1 was increased by about 30% (Table 2). After exercise, carnitine TMA T2 increased by about 10%, but the T1 was essentially the same (Table 2). T1 and T2 measurements on acetylcarnitine resonances in resting muscle were difficult due to the low concentration (Figures 4 and and5);5); measurements from activated muscle yielded TMA T1 and T2 values close to those of free carnitine (Table 2). Although the T2 of TMA in carnitine and acetylcarnitine were similar to that of the methyl group in creatine, their TMA T1 values were significantly shorter, as shown by the inversion recovery spectra (Figure 4). To convert TMA intensity into concentration under these acquisition conditions using the creatine methyl signal as an internal reference, a T1 correction factor (0.91 for carnitine and 0.88 for acetylcarnitine) was used. After correction for partial saturation, the post-exercise soleus TMA signals in Figures 2 and and33 corresponded to a free carnitine concentration of 5.4 and 4.8 mM, and acetylcarnitine concentration of 3.4 and 4.1 mM, respectively.

Figure 4
Measuring T1 of Carnitine and Acetylcarnitine in Pre- and Post-exercise Soleus Muscle
Figure 5
Measuring T2 of Carnitine and Acetylcarnitine in Pre- and Post-exercise Soleus Muscle
Table 2
Relaxation times of carnitine, acetylcarnitine and creatine in phantom and soleus muscle (n = 4)

Since the correction for relaxation differences between carnitine and acetylcarnitine TMA signal intensities is small, as a good approximation one may convert their intensity ratio directly into concentration ratio. Provided the TMA signal at 3.2 ppm was exclusively contributed by free carnitine, then the estimated acetylcarnitine : carnitine was 0.40 : 0.60 (Figure 2) and 0.48 : 0.52 (Figure 3) by lineshape fitting of the post-exercise spectra.

Threshold for Detecting Acetylcarnitine

It was found that, for the soleus acetylcarnitine acetyl signal at 2.12 ppm and its TMA at 3.17 ppm, the 1H MR intensity ratio was about 1 : 2 at TE = 140 ms (Figures 2 and and3),3), indicating a possible advantage of TMA over the acetyl signal for sensitive detection of acetylcarnitine. The detection limit of this pulse sequence was evaluated using a spherical phantom containing aqueous 10 mM carnitine in 500 mL. The concentration of acetylcarnitine was varied. With well-resolved TMA signals (LW1/2 = 3.8 Hz), a linear relation was detected between the acetylcarnitine TMA signal intensity and its concentration in the phantom in the range from 0.25 to 2 mM (Figure 6). At lower acetylcarnitine concentration, about ~0.25 mM, the acetylcarnitine TMA signal was difficult to detect (Figure 6). Given the linewidth of acetylcarnitine in vivo, the threshold for detecting acetylcarnitine in soleus was estimated to be ~ 0.7 mmol/kg ww, taking into consideration the unfavorable linewidth in vivo (3.8 Hz vs 8.0 ± 1.9 Hz, n = 20) due to the short T2 relaxation time.

Figure 6
Threshold for Detecting Acetylcarnitine

Detection and Slow Decay of Acetylcarnitine

As summarized in Table 3, acetylcarnitine increased dramatically after exercise. The concentration of acetylcarnitine decayed slowly during the recovery period (Figure 7a) to the pre-exercise level at about 50 minutes. Three methods were used to describe the rate of disappearance of acetylcarnitine. In the first, the acetyl signal was fit to a mono-exponential model yielding a decay constant τ of 18.2 (2.3) min (Note: the value in bracket represents the estimate of fitting error). A second approach was to measure the fraction of acetylcarnitine in the total carnitine pool by comparing the area of the acetylcarnitine TMA resonance to the total TMA signal (Figure 7b). This approach yielded a nearly equal decay constant, 17.8 (1.7) min, for the disappearance of acetylcarnitine. Finally, using the post-exercise acetylcarnitine TMA decay curve, the half-time constant τ was 17.1 (1.9) min (Figure 7c). The rate of disappearance of the acetylcarnitine acetyl resonance occurred synchronously with that of the acetylcarnitine TMA resonance (Figure 7). For the group of the subjects studied (n = 20), the average τ value was 18.8 ± 5.6 min for acetylcarnitine TMA and 19.4 ± 6.1 min for the acetyl signal. Using the methyl signal of creatine as internal reference (Equation [1]), the estimated initial post-exercise acetylcarnitine was 4.9 ± 0.2 mmol/kg ww and the residual acetylcarnitine concentration was 0.4 ± 0.1 mmol/kg ww.

Figure 7
Single-voxel 7T 1H MR Spectra after Exercise
Table 3
Pre- and post-exercise carnitine and acetylcarnitine concentrations, and the fitting results of post-exercise acetylcarnitine decays in human soleus muscle (n = 20)

An interesting feature of the post-exercise muscle 1H MRS spectra was the reduced linewidth in the carnitine signal illustrated in Figure 7a. The average LW1/2 was decreased by 22% for creatine methyl Cr3 (pre–exercise 10.9 ± 2.7 Hz vs post-exercise 8.5 ± 2.0 Hz, n = 20) and by 26% for carnitine TMA (pre-exercise 11.6 ± 3.0 Hz vs post-exercise 8.6 ± 2.1 Hz) (Figure 8a and and7b).7b). However, as muscles continued to recover from exercise, the linewidth was gradually increased toward its pre-exercise level (Figure 7a). After correction for the difference in relaxation times between carnitine and acetylcarnitine, the post-exercise total carnitine concentration was not different from the pre-exercise concentration (9.5 ± 1.6 mmol/kg ww vs 8.9 ± 1.8 mmol/kg ww).

Figure 8
Average 1H MR Spectra, Recovery Half Time for Acetylcarnitine, and the Effect of Exercise on the Concentration of Carnitine and Acetylcarnitine

For this group of twenty subjects, there was no gender difference in facetylcarnitine and τ values, although male subjects (n = 12) showed higher free carnitine concentration prior to exercise (9.0 ± 1.7 mmol/kg ww vs 7.9 ± 1.4 mmol/kg ww). However, with similar free carnitine concentration, young subjects (male and female, n = 13) had significantly lower post-exercise carnitine (5.8 ± 1.2 vs 6.9 ± 1.6 mmol/kg ww, p = 0.05) and significantly larger facetylcarnitine (38 ± 9% vs 31 ± 4%, p < 0.05) than the middle-aged (n = 7). The difference in facetylcarnitine were even more prominent between the young males (n = 7) and middle-aged males (n = 5) (41 ± 8% vs 29 ± 4%, p < 0.05). Neither BMI nor calf muscle size was found to be correlated with facetylcarnitine or τ value.


The assignment of the signal at 3.17 ppm in post-exercise skeletal muscle to the TMA resonance of acetylcarnitine is reasonable in view of the chemical shift, the apparent singlet structure, the appearance and decay coincident with the acetyl signal at 2.12 ppm, and the measured concentration after correction for small relaxation time differences. In comparison to acetyl signal, the detection of acetylcarnitine by its TMA signal has the advantage of high sensitivity, and virtually without chemical shift displacement artifact and lipid signal overlapping in human soleus muscle (Figure 8a). The capacity to resolve two closely-spaced TMA signals is interesting because they are exchanging substrates of a very active enzyme, carnitine acetyltransferase (CAT). This reaction is freely reversible, the enzyme is present in high activity, and the reaction has no need for energy input (20). Consequently, the relative concentration of acetylcarnitine and carnitine reflect the ratio acetyl-CoA/CoASH (21), at least in systems without compartmentation. In spite of the high activity of this enzyme, the observation of well-resolved TMA signals in the post-exercise soleus muscle at 7T (Figures 2, ,3,3, ,4,4, ,55 and and7)7) is in agreement with the analysis of slow exchange kinetics (k << Δδ = 9 Hz). For a clinical 3T system, the TMA exchange between carnitine and acetylcarnitine may still be in slow kinetic regime at high acetylcarnitine level; therefore a similar observation of TMA separation may be possible provided that the TMA linewidth would be proportionally reduced to about 4 Hz.

The current protocol - a short pre-scan, a 5-min calf-raise exercise conducted out-of-magnet, followed by subsequent post-exercise scans - was completed by all participants. Though standing calf-raises is a common resistance exercise to strengthen calf muscles, it has not been well-explored in metabolic studies. The major considerations for adopting such an exercise protocol were its efficiency in acetylcarnitine production and compatibility with 7T working environment, in addition to its simplicity. No extra equipment was needed to carry out the exercise, and no extra visits or pre-tests were required for the participants to familiarize the procedure in advance. Furthermore, since each subject was asked to just lift up his or her own body weight, similar to the daily activity of walking and stair-climbing, a common but often difficult standardization procedure was avoided for determination of individual’s maximum force. The main limitation of this exercise protocol is calf tenderness and discomfort for 1 – 2 days.

The measured acetylcarnitine is consistent with earlier reports from biopsy studies, in three respects. First, the initial concentration of acetylcarnitine after exercise was about 4.1 ± 1.0 mmol/kg wet weight, a result very close to biopsy measurements. The concentration of acetylcarnitine after exercise in the vastus lateralis is approximately 14–16 mmol/kg dry weight or 3.6–3.8 mmole/kg wet weight assuming known water content in muscle (22,23). Second, in the majority of subjects (15/20), the concentration of acetylcarnitine at rest in our study was below 7T 1H MRS detection limit, which is estimated to be ~ 0.7 mmol/kg wet weight. The remaining 5 subjects with detectable acetylcarnitine TMA signal had an average acetylcarnitine concentration of 1.0 ± 0.6 mmol/kg wet weight. The residual acetylcarnitine from fitting the dynamic decay curves was 0.5 ± 0.3 mmol/kg wet weight (Table 2). This is in agreement with the reported results 2.0–2.2 mmol/kg dry weight, or ~ 0.5 mmol/kg wet weight for human muscle at rest (10,23,24). Third, although repetitive biopsy sampling is not feasible in humans to determine a detailed evaluation of acetylcarnitine kinetics with recovery, the observed acetylcarnitine decay constant, about 19 minutes, is consistent with full recovery of free carnitine by 3 hours after exhaustive exercise reported by invasive studies (25).

The ratio of free CoASH to acetyl-CoA serves as an essential regulatory role in skeletal muscle fuel metabolism. Because of their low concentrations, noninvasive measurements are not feasible and for this reason it is attractive to consider that acetyl-CoA/CoASH can be measured from the equilibrium constant of the enzyme and the MR-observed acetylcarnitine/carnitine. Indeed, MR methods now provide two approaches, each with advantages and limitations, to simultaneously quantify carnitine and acetylcarnitine in vivo during recovery from exercise. The proton resonances of the trimethylamine groups of free carnitine and acetyl-carnitine, arising from 9 equivalent protons, have higher detection sensitivity, and the relative signal amplitudes were insensitive to pulsing conditions because of similar relaxation times. Quantitation of acetylcarnitine from the methyl signal may be less attractive in soleus muscle because of fewer protons for detection, the need for correcting for relaxation time differences, and chemical shift displacement effects are more significant in single-voxel spectroscopy. The disadvantage of quantifying acetylcarnitine from the TMA signal is that high spectral resolution is essential.

With caveats, it is worth exploring an estimate of acetyl-CoA/CoASH in muscle. Given the known equilibrium constant Keq for the acetyl transfer reaction (Figure 1), the MRS-determined ratio of acetylcarnitine to free carnitine can be used as follows: acetyl-CoA/[CoASH = acetylcarnitine/(carnitine · Keq· Keq is approximately unity (1) and numerous investigators have reported that the acetyl transfer reaction is in equilibrium or near equilibrium in muscle (3,7,1922). For example, Friolet et al (26) demonstrated that the ratio acetylcarnitine/carnitine linearly correlated with the ratio AcCoA/CoASH in human skeletal muscle after exhaustive exercise with a linear coefficient value of near unity and an intercept near zero (y = 0.96 × − 0.02). Constantin-Teodosiu et al (24) found a close match between the ratio acetylcarnitine/carnitine and acetyl-CoA/CoASH at rest and after 3 min of exercise. Carlin et al (27) observed a linear relation between acetylcarnitine and AcCoA incremental content in skeletal muscle of horses after short duration speed exercises, and attributed the observation to a well-maintained equilibrium in the acetyl transfer reaction between CoA and carnitine pools. However, any calculation of acetyl-CoA/CoASH assumes that all MR-detected carnitine has access to CAT. The assumption is open to question since this enzyme, bound to the inner mitochondrial membrane, presumably is accessible only to mitochondrial carnitine, yet MR detects the entire carnitine pool of muscle.

The advantage of the large dynamic range of acetylcarnitine concentration (~10-fold) in the post-exercise recovery period (~1 hr) enables a reliable determination of the decay constant τ, with very high temporal resolution, at least compared to biopsy studies. Kimber et al reported that acetylcarnitine, measured by biopsy at exhaustion and 3, 6, and 18 hours after recovery, decreased significantly during recovery (25). The high temporal resolution enables an estimate of CAT’s enzymatic activity kf in converting acetylcarnitine to carnitine. Since the resting acetylcarnitine concentration is negligibly low compared to the peak value immediately after the exercise, the kf-value is simply determined as the reciprocal of the acetylcarnitine decay rate (= 1/τ). The faster the acetylcarnitine signal decays, the more rapidly CAT catalyzes the conversion of acetylcarnitine to free carnitine. The kinetic rate constant kf in soleus muscle was found to be 0.05 ± 0.02 min−1, which is a fraction of biochemical assay results of 1.74 – 18.6 min−1 (estimated by CAT turnover number 29,000 moles of substrate transformed per mole of CAT enzyme per minute, at pH7.8 and 30° (16), 5 mmol/kg ww acetylcarnitine substrate, and the enzyme content in muscle 0.3 – 3.2 µmol/kg ww (16,17)). When expressed in terms of quantity of substrate conversion, upper limit of post-exercise CAT activity Rf was 0.21 ± 0.11 mmol/min/kg ww (= 0.05 min−1 × 4.1 mmol/kg ww) for the depletion of acetylcarnitine to regenerate AcCoA, which is comparable to the down-stream TCA cycle turnover rate 0.10 mmol/kg ww/min obtained from human skeletal muscle by 13C MRS with [2-13C]acetate infusion (28). The fact that the CAT activity in vivo is only a fraction of that found with the isolated enzyme implies that the acetylcarnitine access to the transferase may be spatially restricted.

In addition to acetylcarnitine-to-carnitine conversion, acetylcarnitine depletion could also be caused by acetylcarnitine leaking into the circulation and by inter-tissue acetylcarnitine-carnitine exchange. The accumulated acetylcarnitine, like lactate, may escape from muscle into the blood pool, and lead to a decreased 1H MR signal intensity in acetylcarnitine in muscle. Though elevated acetylcarnitine was indeed found in plasma post-exercise (29), its concentration (less than 0.01 mM (29)) was too low (100-fold lower than in muscle) to be a significant contributor for the observed acetylcarnitine decay phenomenon. Furthermore, if acetylcarnitine was leaked rather than converted to free carnitine, one would expect a significantly decreased carnitine TMA signal at 3.2 ppm at the end of full muscle recovery from exercise, as compared to the pre-exercise TMA signal. Such a hypothesis contradicts 1H MRS findings (Figure 7) and also is not consistent with biopsy results for a conserved total carnitine concentration, independent of exercise (23). Alternatively, the accumulated acetylcarnitine in muscle may be exported into circulation in exchange for the import of free carnitine from blood into muscle. However, it seems less likely that this inter-tissue acetylcarnitine-carnitine exchange mechanism would have a significant impact in a time-frame of 10–20 min (Figure 7), given the fact of very low free carnitine level in blood and extremely slow uptake of free carnitine by muscle (10).

It should be noted that the synthesis of acetylcarnitine during exercise, also catalyzed by this enzyme, is much more rapid than its decay following exercise, which would suggest that the decay of the acetylcarnitine following exercise yields a measure of the flux of acetylcarnitine incorporation into the TCA cycle, or the resting metabolic rate of skeletal muscle. Although the activity of CAT must be sufficient high to support the equilibrium between carnitine and CoA pools following exercise, its upper limit in human soleus in vivo is apparently defined by the rate of kinetic exchange shown by TMA signals of carnitine and acetylcarnitine, which falls in slow exchange regime on NMR time scale at 7T (9 s−1).

In summary, detection of resolved carnitine and acetylcarnitine TMA signals by 7T 1H MRS offers a convenient tool for the study of CAT-mediated acetyl transfer process in mitochondria of human skeletal muscle in vivo. The measurement of acetylcarnitine production and its decay constant may serve as a new surrogate for non-invasive testing of mitochondrial function. One may expect that the method could find potential clinical application in neuromuscular disorders such as mitochondrial myopathies. Although the advantages of MR such as the ability to monitor metabolism repeatedly and noninvasively are well-known, perhaps equally important is its ability to integrate measures of carnitine and acetylcarnitine with other metabolites important in fatty acid oxidation and mitochondrial function such as the concentration of intramyocellular lipids that may be linked to impaired mitochondrial oxidation of fatty acids and insulin resistance (19,3036).


The authors are grateful for the helpful discussion with Dr Ivan Dimitrov (Philips Medical Systems), Jing Yang and Deborah Douglas for operational assistance. Jeannie Davis and Sonya Rios recruited and managed the human subjects. This study was supported by the National Institutes of Health (RR02584, DK081186, and RO1AR050597) and the Department of Defense (Contract number W81XWH-06-2-0046).


1. Ramsay RR, Zammit VA. Carnitine acyltransferases and their influence on CoA pools in health and disease. Molecular Aspects of Medicine. 2004;25(5–6):475–493. [PubMed]
2. Zammit VA. Carnitine acyltransferases: functional significance of subcellular distribution and membrane topology. Progress in Lipid Research. 1999;38(3):199–224. [PubMed]
3. Colucci WJ, Gandour RD. Carnitine acetyltransferase: a review of its biology, enzymology, and bioorganic chemistry. Biochem J. 1988;16:307–334.
4. Ventura-Clapier Re, Mettauer B, Bigard X. Beneficial effects of endurance training on cardiac and skeletal muscle energy metabolism in heart failure. Cardiovascular Research. 2007;73(1):10–18. [PubMed]
5. Stephens FB, Constantin-Teodosiu D, Greenhaff PL. New insights concerning the role of carnitine in the regulation of fuel metabolism in skeletal muscle. The Journal of Physiology. 2007;581(2):431–444. [PubMed]
6. Stephens FB, Constantin-Teodosiu D, Laithwaite D, Simpson EJ, Greenhaff PL. Insulin stimulates L-carnitine accumulation in human skeletal muscle. FASEB J. 2006;20(2):377–379. [PubMed]
7. Constantin-Teodosiu D, Carlin JI, Cederblad G, Harris RC, Hultman E. Acetyl group accumulation and pyruvate dehydrogenase activity in human muscle during incremental exercise. Acta Physiol Scand. 1991;143(4):367–372. [PubMed]
8. Greenhaff PL, Campbell-O'Sullivan SP, Constantin-Teodosiu D, Poucher SM, Roberts PA, Timmons JA. An acetyl group deficit limits mitochondrial ATP production at the onset of exercise. Biochem Soc Trans. 2002;30(2):275–280. [PubMed]
9. Roberts PA, Loxham SJG, Poucher SM, Constantin-Teodosiu D, Greenhaff PL. Acetyl-CoA provision and the acetyl group deficit at the onset of contraction in ischemic canine skeletal muscle. Am J Physiol Endocrinol Metab. 2005;288(2):E327–E334. [PubMed]
10. Stephens FB, Constantin-Teodosiu D, Laithwaite D, Simpson EJ, Greenhaff PL. An acute increase in skeletal muscle carnitine content alters fuel metabolism in resting human skeletal muscle. J Clin Endocrinol Metab. 2006;91(12):5013–5018. [PubMed]
11. Randle PJ. Pyruvate-dehydrogenase complex: Meticulous regulator of glucose disposal in animals. Trends Biochem Sci. 1978;3:217–225.
12. Fernández-Pastor VJ, Ruiz M, Diego-Acosta AM, Avila C, García JC, Pérez F, Guirado F, Noguer N. Metabolic and hormonal changes during aerobic exercise in distance runners. J Physiol Biochem. 1999;55(1):7–16. [PubMed]
13. Bet L, Bresolin N, Moggio M, Meola G, Prelle A, Schapira AH, Binzoni T, Chomyn A, Fortunato F, Cerretelli P. A case of mitochondrial myopathy, lactic acidosis and complex I deficiency. J Neurol. 1990;237(7):399–404. [PubMed]
14. Peluchetti D, Antozzi C, Roi GS, Rimoldi M, Cornelio F. Muscle weakness and intolerance to exercise in mitochondrial myopathies. Acta Neurol (Napoli) 1989;11(5):335–345. [PubMed]
15. Kreis R, Jung B, Rotman S, Slotboom J, Boesch C. Non-invasive observation of acetyl-group buffering by 1H-MR spectroscopy in exercising human muscle. NMR in Biomedicine. 1999;12(7):471–476. [PubMed]
16. Chase JFA. Carnitine acetyltransferase from pigeon breast muscle. Methods Enzmol. 1969;13:387–393.
17. Fritz IB, Schultz SK, Srere PA. Properties of Partially Purified Carnitine Acetyltransferase. Journal of Biological Chemistry. 1963;238(7):2509–2517. [PubMed]
18. Khuu A, Ren J, Dimitrov I, Woessner D, Murdoch J, Sherry AD, Malloy CR. Orientation of lipid strands in the extracellular compartment of muscle: Effect on quantitation of intramyocellular lipids. Magnetic Resonance in Medicine. 2009;61(1):16–21. [PMC free article] [PubMed]
19. Boesch C, Machann J, Vermathen P, Schick F. Role of proton MR for the study of muscle lipid metabolism. NMR Biomed. 2006;19(7):968–988. [PubMed]
20. Jogl G, Hsiao Y, Tong L. Structure and function of carnitine acyltransferase. Ann N Y Acad of Sci. 2004;1033:17–29. [PubMed]
21. Jogl G, Tong L. Crystal Structure of Carnitine Acetyltransferase and Implications for the Catalytic Mechanism and Fatty Acid Transport. Cell. 2003;112(1):113–122. [PubMed]
22. Bergstrom J, Furst P, Noree LO, Vinnars E. Intracellular free amino acid concentration in human muscle tissue. J Appl Physiol. 1974;36(6):693–697. [PubMed]
23. Constantin-Teodosiu D, Howell S, Greenhaff PL. Carnitine metabolism in human muscle fiber types during submaximal dynamic exercise. J Appl Physiol. 1996;80(3):1061–1064. [PubMed]
24. Constantin-Teodosiu D, Cederblad G, Hultman E. PDC activity and acetyl group accumulation in skeletal muscle during prolonged exercise. J Appl Physiol. 1992;73(6):2403–2407. [PubMed]
25. Kimber NE, Heigenhauser GJF, Spriet LL, Dyck DJ. Skeletal muscle fat and carbohydrate metabolism during recovery from glycogen-depleting exercise in humans. The Journal of Physiology. 2003;548(3):919–927. [PubMed]
26. Friolet R, Hoppeler H, Krahenbuhl S. Relationship between the coenzyme A and the carnitine pools in human skeletal muscle at rest and after exhaustive exercise under normoxic and acutely hypoxic conditions. J Clin Invest. 1994;94(4):1490–1495. [PMC free article] [PubMed]
27. Carlin JI, Harris RC, Cederblad G, Constantin-Teodosiu D, Snow DH, Hultman E. Association between muscle acetyl-CoA and acetylcarnitine levels in the exercising horse. J Appl Physiol. 1990;69(1):42–45. [PubMed]
28. Befroy D, Petersen K, Dufour S, Mason G, de Graaf R, Rothman D, Shulman G. Impaired mitochondrial substrate oxidation in muscle of insulin-resistant offspring of type 2 diabetic patients. Diabetes. 2007;56(5):1376–1381. [PMC free article] [PubMed]
29. Gatti R, De Palo CB, Spinella P, De Palo EF. Free carnitine and acetyl carnitine plasma levels and their relationship with body muscular mass in athletes. Amino Acids. 1998;14(4):361–369. [PubMed]
30. Liska D, Dufour S, Zern TL, Taksali S, Calí AM, Dziura J, Shulman GI, Pierpont BM, Caprio S. Interethnic differences in muscle, liver and abdominal fat partitioning in obese adolescents. PLoS One. 2007;2(6):e569. [PMC free article] [PubMed]
31. Machann J, Stefan N, Schick F. 1H MR spectroscopy of skeletal muscle, liver and bone marrow. Eur J Radiol. 2008;67(2):275–284. [PubMed]
32. Malloy C, Ren J, White C. Intramyocyte Lipids May Impair Insulin Signaling. Am J Psychiatry. 2007;164(10):1475. [PubMed]
33. Petersen KF, Dufour S, Shulman GI. Decreased insulin-stimulated ATP synthesis and phosphate transport in muscle of insulin-resistant offspring of type 2 diabetic parents. PLoS Med. 2005;2(9):e233. [PMC free article] [PubMed]
34. Machann J, Steidle G, Thamer C, Mader I, Schick F, G AW. Annual Reports on NMR Spectroscopy. Volume Volume 50. Academic Press; 2003. In Vivo Proton NMR Studies in Skeletal Musculature; pp. 1–74.
35. Schrauwen P, Schrauwen-Hinderling V, Hoeks J, Hesselink MKC. Mitochondrial dysfunction and lipotoxicity. Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids. 2010;1801(3):266–271. [PubMed]
36. Ren J, Sherry AD, Malloy CR. 1H MRS of intramyocellular lipids in soleus muscle at 7 T: Spectral simplification by using long echo times without water suppression. Magnetic Resonance in Medicine. 2010;64(3):662–671. [PMC free article] [PubMed]