SKELETAL MUSCLE OXIDATIVE CAPACITY IN AMYOTROPHIC LATERAL SCLEROSIS TERENCE E. RYAN, PhD,1 MELISSA L. ERICKSON, MS,1 AJAY VERMA, MD, PhD,2 JUAN CHAVEZ, MD,2 MICHAEL H. RIVNER, MD,3 and KEVIN K. MCCULLY, PhD1 1

Department of Kinesiology, University of Georgia, Athens, Georgia, USA Experimental Medicine, Biogen Idec, Cambridge, Massachusetts, USA 3 Department of Neurology, Georgia Regents University, Augusta, Georgia, USA Accepted 21 February 2014 2

ABSTRACT: Introduction: Mitochondrial dysfunction in the motor neuron has been suspected in amyotrophic lateral sclerosis (ALS). If mitochondrial abnormalities are also found in skeletal muscle, assessing skeletal muscle could serve as an important biomarker of disease progression. Methods: Using 31 P magnetic resonance (31P-MRS) and near infrared (NIRS) spectroscopy, we compared the absolute values and reproducibility of skeletal muscle oxidative capacity in people with ALS (n 5 6) and healthy adults (young, n 5 7 and age-matched, n 5 4). Results: ALS patients had slower time constants for phosphocreatine (PCr) and muscle oxygen consumption (mVO2) compared with young, but not age-matched controls. The coefficient of variation for the time constant was 10% (SD 5 2.8%) and 17% (SD 5 6.2%) for PCr and mVO2, respectively. Conclusions: People with ALS had, on average, a small but not statistically significant, impairment in skeletal muscle mitochondrial function measured by both 31P-MRS and NIRS. Both methods demonstrated good reproducibility. Muscle Nerve 50: 767–774, 2014

Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease that affects motor neurons.1 The majority of cases are sporadic, and the etiology is still largely unknown. Because of the difficulties in early diagnosis and evaluation of disease progression, there is a growing need to identify clinically useful biomarkers that can also aid in the assessment of drug response in clinical trials. Abbreviations: ALSFRS, ALS functional rating scale; ATP, adenosine triphosphate; CV, coefficient of variation; FOV, field of view; FWHM, fullwidth half maximum; HHb, deoxygenated hemoglobin/myoglobin; mVO2, muscle oxygen consumption; MRI, magnetic resonance imaging; MRS, magnetic resonance spectroscopy; NIRS, near infrared spectroscopy; O2Hb, oxygenated hemoglobin/myoglobin; PCr, phosphocreatine; [PCr], concentration of phosphocreatine; PDE, phosphodiesters; Pi, inorganic phosphate; 31P-MRS, phosphorus magnetic resonance spectroscopy; SD, standard deviation; SNR, signal-to-noise ratio Key words: mitochondrial bioenergetics; mitochondrial function; motor neuron disease; MRS; NIRS T.E.R, M.L.E, J.C., A.V. and K.K.M. did the conception and design of the research; T.E.R. and M.L.E., performed the experiments; M.H.R. provided clinical support and recruited and screened age-matched controls and ALS patients. T.E.R wrote the analysis routines and analyzed the data; T.E.R., M.L.E., and K.K.M. interpreted the results of the experiments; T.E.R. prepared the figures; T.E.R. drafted the manuscript; T.E.R., M.L.E., J.C., A.V., M.H.R., and K.K.M. edited and revised the manuscript; T.E.R., M.L.E., J.C., A.V., M.H.R., and K.K.M. approved the final version of the manuscript. The authors report no conflicts of interest. This study was funded by Biogen IDEC. Correspondence to: T. E. Ryan, Department of Physiology, East Carolina Diabetes and Obesity Institute, Brody School of Medicine, East Carolina University, Greenville North Carolina, USA; e-mail: [email protected] C 2014 Wiley Periodicals, Inc. V

Published online 24 February 2014 in Wiley Online Library (wileyonlinelibrary. com). DOI 10.1002/mus.24223

Muscle Metabolism in ALS

Mitochondrial dysfunction has been implicated in the pathophysiology of ALS and other neuromuscular disorders. Mitochondrial-related oxidative stress,2 mitochondrial DNA mutations/deletions,3 and abnormalities in mitochondrial respiratory chain complexes have been reported in ALS patients and various experimental models (cell and animal) of the disease.2,4–8 If ALS-related mitochondrial abnormalities are widespread beyond motor neurons, skeletal muscles could be used as surrogate tissues to test human mitochondrial function with noninvasive measurements. Noninvasive measurements of mitochondrial function are ideal for human testing. Currently, phosphorus magnetic resonance spectroscopy (31PMRS) is considered the gold standard for assessing skeletal muscle metabolism in vivo. In particular, the kinetics of phosphocreatine (PCr) resynthesis after exercise have been used as a direct assessment of mitochondrial capacity.9 Despite its widespread use in research, 31P-MRS is seldom used clinically due to its cost, requirement of specialized skills by investigators, and limited availability of multinuclear MRI systems and equipment. Two previous studies have used 31P-MRS to investigate skeletal muscle metabolism in people with ALS,10,11 and both reported no difference from control subjects. A potential limitation of these studies is the use of voluntary exercise protocols to induce changes in skeletal muscle metabolism, which results in activation of healthy, intact motor units that are less likely to exhibit metabolic alterations. In contrast, near infrared spectroscopy (NIRS) provides a noninvasive measure of muscle oxygenation.12 Commercially available NIRS devices provide information about the relative changes in oxygenated hemoglobin/myoglobin (O2Hb), deoxygenated hemoglobin/myoglobin (HHb), and total hemoglobin or blood volume. In comparison to MRS, NIRS devices are much smaller, less expensive, portable, and easier to use. Recent studies have used transient arterial occlusion and NIRS to measure the rate of recovery of muscle oxygen consumption (mVO2) after exercise.13–15 The recovery of mVO2 follows a monoexponential MUSCLE & NERVE

November 2014

767

function similar to that of PCr resynthesis, where the time constant is inversely proportional to mitochondrial oxidative capacity.16 In this study, 31P-MRS and NIRS were used to examine skeletal muscle mitochondrial function in 3 cohorts: young healthy subjects, ALS patients, and their age-matched controls. This study also examined the reproducibility, consistency of measurements over time, and signal-to-noise ratios (SNRs) of the 31P-MRS and NIRS tests. MATERIALS AND METHODS Participants. This study

recruited women and men aged 18–75 years. Patients with a diagnosis of at least clinically probable ALS in accordance with the revised El Escorial criteria17 were enrolled at the Georgia Regents University ALS clinic (Augusta, Georgia). Eligible participants had onset of first ALS symptoms within 24 months of screening and an upright slow vital capacity of at least 65% percent of the predicted value for age, gender, and height. In addition, participants had calf muscle strength of grade 3 or 4 based on the MRC scale.18 Subjects were excluded based on the following criteria: other neurodegenerative disease, use of statins, musculoskeletal injury, or any condition that would affect lower extremity strength (other than ALS). Age-matched controls were excluded if they had a history of regular exercise (>30 min) 2 or more days per week. The study was conducted with the approval of the Institutional Review Boards at the University of Georgia (Athens, Georgia) and Georgia Regents University (Augusta, Georgia), and all participants gave written, informed consent before testing. 31

P-MRS and NIRS measurements were made a total of 8 times across 4 testing sessions. For ALS patients, 31P-MRS and NIRS measurements were made within a month of the screening visit. Each testing session consisted of 2 31 P-MRS and 2 NIRS measurements. Both tests were performed in the dominant leg as selfreported by each subject. On day 1, participants arrived at the UGA BioImaging Research Center (Athens, GA) and were tested twice, using both 31 P-MRS and NIRS, separated by approximately 90 min. Participants returned to the MRI facility within 5–14 days and repeated the same protocol as day 1. For each participant, testing was performed at the same time of day for each visit. Study Design.

Clinical Screening and Determination of Lower Leg Function. We evaluated each patient clinically within a month before the first testing session. Three patients had clinically definite ALS, and 3 had clinically probable ALS. All patients had lower motor neuron findings in 3 regions, and the 768

Muscle Metabolism in ALS

clinical definite ALS patients also had upper motor neuron findings in at least 3 regions. We tested muscle strength and evaluated them using the ALS Functional Rating Scale (ALSFRS).19 The ALSFRS score was determined by a clinical interview with patient/caregiver-reported answers. In addition, an investigator who was a board-certified neurologist, with more than 25 years’ experience in ALS, rankordered the ALS patients in terms of clinical involvement of the triceps surae (gastrocnemius/ soleus) muscle group. The ranking took into account several factors including manual muscle testing results, patient-reported ALSFRS scores, and the potentially subjective clinical assessment of disease severity by the study neurologist (M.H.R.). This evaluation was performed without knowledge of the results of any of the tests of muscle metabolism. The rank ordering was performed because of the variable rate of involvement of the triceps surae muscle group in the progression of ALS symptoms, and because we believed the amount of clinical involvement might help predict or understand the metabolic measurements. NIRS Experimental Protocol. NIRS testing was performed as described previously.13,16 Two aluminum foil electrodes attached to a Theratouch 4.7 stimulator (Rich-mar, Inola, Oklahoma) were placed on the skin over the triceps surae, 1 proximal and 1 distal to the NIRS optode. A blood pressure cuff (Hokanson SC12D or SC12L; Bellevue, WA) was placed above the knee and connected to a rapidinflation system (Hokanson E20, Bellevue, Washington) and 2 horsepower, 10-gallon capacity commercial air compressor. Adipose tissue thickness was measured at the site of the NIRS optode using B-mode ultrasound (LOGIQ e; GE HealthCare, USA). The test protocol consisted of measurement of resting muscle oxygen consumption, followed by 2 exercise/recovery measurements and an ischemic calibration to normalize NIRS signals. To increase mVO2, 15 s of continuous electrical stimulation at 4–6 HZ was applied to the muscle. The current intensity was adjusted for each individual to produce twitch contractions at the maximum tolerable level. Immediately following the electrical stimulation, a series of brief (5–10 s) arterial occlusions was applied to measure the rate of recovery of mVO2 back to resting levels. The repeated transient arterial occlusion protocol was as follows: 5 s on / 5 s off for cuffs 1–5, 7 s on / 7 s off for cuffs 6–10, 10 s on / 10 s off for cuffs 11–15, and 10 s on / 20 s off for the remaining cuffs. NIRS Measurements. NIRS signals were obtained using a battery-powered, portable, continuous-wave NIRS device (PORTAMON, Artinis Medical MUSCLE & NERVE

November 2014

Systems, The Netherlands). The probe has 3 source-detector separation distances (30, 35, and 40 mm). The NIRS data were collected at 10 HZ. Muscle oxygen consumption (mVO2) was calculated as the slope of change in O2Hb and HHb during the arterial occlusion using simple linear regression. mVO2 was expressed as a percentage of the ischemic calibration per unit time. The postexercise repeated measurements of mVO2 were fit to a mono-exponential curve according to the formula: y5End2Delta
e 2kt For this equation, y represents relative mVO2 during the arterial occlusion, End is the mVO2 immediately after the cessation of exercise, Delta is the change in mVO2 from rest to end exercise, t is time, and k is the fitting rate constant (time constant 5 1/k). NIRS data were analyzed using custom-written routines for Matlab v. 7.13.0.564 (The Mathworks, Natick, Massachusetts). NIRS signals were corrected for changes in blood volume as described previously.13 The SNR of NIRS data was also evaluated in each group. The SNR was calculated for each arterial occlusion on all data. The noise was calculated as the standard deviation of 600 data points (60 s), after linear detrending to remove respiratory and blood pressure waves from the NIRS signal. The signal was calculated as the change in the NIRS signal during an arterial occlusion, thus the size of the signal is a function of the duration of the occlusion (chosen to be 3 s in this study) and the rate of oxygen consumption.

Calculation of NIRS SNR.

31

P-MRS Measurements. Participants were tested in a 3 Tesla whole body magnet (GE Healthcare, Waukesha, Wisconsin). An 1H and 31P radiofrequency (RF) dual surface coil (Clinical MR Solutions, Brookfield, Wisconsin) was placed over the triceps surae muscle of the participant’s leg. The size of the 31P coil was 13 cm 3 13 cm, placed orthogonal to the 1H coil (2 loops, side by side, 20 cm 3 20 cm in size). Manual shimming on 1H was applied to get a better SNR and less spectrum distortion after an auto-shimming by a prescan sequence [all participants 1H full-width half maximum (FWHM) mean 6 SD; 0.50 6 0.12 ppm]. A nonlocalized free induction decay chemical shift imaging pulse sequence was applied to acquire the 31 P spectrum with the following scan parameters: repetition time (TR) 5 3 s, field of view (FOV) 5 8 cm, slice thickness 5 8 cm, number of excitations 5 1, rfpulse 5 hard. Two aluminum foil electrodes attached to a Theratouch 4.7 stimulator (Rich-Mar, Inola, Muscle Metabolism in ALS

Oklahoma) were placed on the triceps surae, 1 proximal and 1 distal to the surface coil. The stimulation protocol consisted of 1 min of rest, 1 min of electrical stimulation with 4–6 HZ continuous stimulation, and 5.5 min of recovery. The same current intensity was used for NIRS and 31P-MRS. No ergometer was used, and force levels were not recorded. The leg was positioned straight horizontally (zero degrees of knee flexion). Resting spectra were acquired every 3 s until 150 scans were taken. The resulting spectra were summed and analyzed in a custom analysis program (Winspa, Ronald Meyer, Michigan State University). The summed spectrum was apodized using 2 HZ exponential line broadening and the area under the curve for each peak (Pi, PDE, PCr, a ATP, b ATP, and c ATP) was determined using integration. Absolute concentrations were assumed using a value of 8.2 mM for the gamma ATP peak.20 pH was calculated using the equation21:

 pH 56:771log Pishift 23:27 = 5:682Pishift where Pishift is the chemical shift of Pi relative to PCr in parts per million (ppm). The recovery kinetics of PCr was determined using a custom-written routine in Matlab v. 7.13.0.564 (The Mathworks, Natick, Massachusetts) that uses peak heights instead of the area under the PCr peak.22 PCr peaks were calculated on individual spectra after being apodized using 2 HZ exponential line broadening, followed by zero filling to 8,192 points. The FWHM of each PCr peak was also calculated to ensure no changes in homogeneity occurred during the recovery, which would influence the assumption that the peak height is representative of the concentration. PCr peak heights during recovery after exercise were fit to an exponential curve: PCr 5PCrend 2DPCr
e 2kt where PCrend is the percent PCr immediately after cessation of exercise, DPCr is the change in PCr from rest to end exercise, t is time, and k is the fitting time constant (time constant 5 1/k). PCr peaks were corrected for saturation effects using T1 of 6.7 s.23 31

P-MRS SNR. The SNR was also calculated for each individual spectrum on all data. The signal was calculated as the peak height of PCr for each spectrum (TR 5 3 s) after apodization (2–5 HZ exponential filtering), zero-filling (from 2,048 to 8,192 points), and Fourier transformation. The noise was calculated as the standard deviation of the first 3,000 data points of the spectra after Calculation of

MUSCLE & NERVE

November 2014

769

Table 1. Physical characteristics of participants.*

Young (n 5 7) Age-matched (n 5 4) ALS (n 5 6)

Age (yr)

Height (cm)

Weight (kg)

ATT (mm)

23.2 6 1.4 45.8 6 13.2 47.6 6 9.0

173 6 13 175.9 6 1.3 176.5 6 11.5

78 6 19 87.3 6 15.1 76.0 6 14.1

3.5 6 1.8 6.1 6 2.0 7.6 6 1.7

*All data are presented as mean 6 SD. ATT, adipose tissue thickness.

Between-group comparisons were made using the average of all tests (i.e., each individual had 1 value, calculated by averaging the results of all tests). Post hoc analysis was performed using a Tukey Honestly Significant Difference test. The test–retest reliability was analyzed using coefficient of variation (CV), expressed as a percentage. Correlational analyses were performed when deemed appropriated using Pearson correlation coefficients. Spearman rank correlations were performed to examine the relationship between lower leg function and skeletal muscle oxidative capacity. Statistical analyses were performed using SPSS 19.0 R , Armonk, New York). Significance was (IBMV accepted when P < 0.05.

linear detrending to remove any residual baseline artifact. T1-weighted anatomical images of both legs (from the knee to ankle joints) were acquired with the whole-body transmit/receive coil using fast gradient recalled echo sequence with the following parameters: TR/echo time 5 700/9.4 ms, FA 5 90, echo-train length 5 3, number of excitations 5 3, FOV 5 30 3 30 cm, slice thickness 5 1 cm, gap thickness 5 1.5 cm, number of slices 5 20, acquired matrix 5 320 3 224 (reconstructed 512 3 512). Images were exported and analyzed offline using Image J software (National Institutes of Health, http://rsbweb.nih.gov/ij/index.html). Regions of interest were drawn manually around the triceps surae muscle group and the tibia bone (both cortical bone and marrow). A histogram of all pixels and signal intensities within the region of interest was produced. To differentiate between skeletal muscle and fat in the histogram, a midpoint between skeletal muscle and fat was chosen, corresponding to the average signal intensity of pure skeletal muscle pixels. Percent fat was calculated using values of skeletal muscle and fat pixels from the histogram. Cortical bone and marrow area of the tibia was used as an internal reference to ensure accurate comparisons of images between tests.

Anatomical MRI.

RESULTS

Physical characteristics of all participants are shown in Table 1. Clinical characteristics of ALS patients are shown in Table 2. One ALS patient had bulbar onset; all others had limb onset. All participants completed testing without any adverse events. One participant (young healthy group) was only tested on day 1 due to inability to schedule the second day of testing. NIRS Measurements. Resting mVO2 was 0.45 6 0.29 %/s. There was no difference in resting mVO2 between the 3 groups (main effect P 5 0.527). The average time constants for the recovery of mVO2 were 28.6 6 7.6, 36.7 6 10.0, and 50.8 6 20.4 s for young, age-matched, and ALS participants, respectively (Fig. 1a). The time constant was significantly slower for ALS compared with

Statistical Analysis. Data are presented as mean 6 SD. Comparisons between groups were made using a one-way analysis of variance.

Table 2. Clinical Characteristics of ALS Patients.* ALSFRS-R

Patient

Age yr

Height cm

Weight kg

Gender M/W

Time with symptoms Months

Walking

Stairs

Vital capacity (% Predicted)

1† 2 3† 4† 5 6†

47 50 38 58 54 36

180.3 182.8 157.5 182.8 175.3 180.3

75.0 93.8 63.6 81.8 64.5 77.3

M M W M M M

9.4 38.9 34.2 32.5 10.8 10.5

3 2 3 3 3 3

1 0 1 2 1 1

85 82 106 45 82 89

*Patients are ranked by lower leg function (see the Methods section), with 1 being the worst and 6 being the best. †

Denotes patients currently taking Riluzole.

ALSFRS-R; the revised Amyotrophic Lateral Sclerosis Functional Rating Scale.

770

Muscle Metabolism in ALS

MUSCLE & NERVE

November 2014

younger controls (P 5 0.029) but not significantly different between young and age-matched (P 5 0.675) or age-matched and ALS (P 5 0.247). In patients with ALS, the time constant correlated well with lower leg function (Spearman r 5 20.71) (Fig. 1d). The CVs for the time constants were 13.7 6 2.1, 24.3 6 5.0, and 18.0 6 6.5% for young, age-matched, and ALS participants, respectively. The CV was significantly different between young and age-matched controls (P 5 0.008), but it was not different between young and ALS (P 5 0.269) or age-matched and ALS (P 5 0.132). There was no evidence of changes in the CVs from within a test session, between testing sessions (same day), or between days. The SNR of resting NIRS data was 13.0 6 6.6, 10.0 6 4.0, and 16.5 6 4.3 for young, age-matched, and ALS participants, respectively. No significant differences in the resting SNR were detected between groups (all comparisons P > 0.14). There was a small relationship (linear regression equation: Y 5 20.45X 1 24; R2 5 0.17) between the SNR and CV in the time constant. 31

P-MRS Measurements. There were no differences in resting [PCr] or pH between groups (main

effect P 5 0.614). The time constant for the recovery of PCr was 36.1 6 7.2, 51.1 6 13.3, and 59.3 6 11.1 s for young, age-matched, and ALS participants, respectively (Fig. 1a). The time constant was significantly different between young and ALS (P 5 0.004), but not young and age-matched controls (P 5 0.08) or age-matched controls and ALS (P 5 0.445). In patients with ALS, the time constant correlated well with lower leg function (Spearman r 5 20.70) (Fig. 1c). 31P-MRS data from 1 ALS participant was excluded due to an inability to decrease [PCr] to a level that would allow confidence in curve fitting the recovery (less than 10 percent decrease in PCr). The CVs for the time constants were 10.1 6 2.7, 10.6 6 3.7, and 10.6 6 2.7% for young, age-matched, and ALS participants, respectively. The CV was not different significantly for any between-group comparisons (P 5 0.951 for all comparisons). There was no evidence of changes in the CVs from within a test session, between testing session (same day), or between days. The SNR of resting phosphorus data was 143.6 6 19.7, 126.0 6 31.4, and 100.5 6 25.8 for young, age-matched, and ALS participants,

FIGURE 1. (A) Postexercise recovery time constants for PCr and mVO2, measured with 31P-MRS and NIRS, respectively. (B) MRI image analysis for muscle volume and percent fat of the gastrocnemius and soleus muscle groups (both legs summed). (C) Spearman rank correlation between PCr time constants and ranked lower leg function. (D) Spearman rank correlation between mVO2 time constants and ranked lower leg function. *denotes P < 0.05 versus young controls. Muscle Metabolism in ALS

MUSCLE & NERVE

November 2014

771

respectively. The SNR was significantly different between young controls and ALS (P 5 0.02), but not different between young and age-matched controls (P 5 0.512) or age-matched controls and ALS (P 5 0.285). There was a small relationship (linear regression equation: Y 5 20.03X 1 14; R2 5 0.09) between the SNR and variation (CV) in the time constant. We also found a small negative correlation between the SNR of phosphorus data and the fat percentage from MRI images (Pearson r 5 0.33). 31

P-MRS. There was a significant correlation (Pearson r 5 0.66; P 5 0.005) between NIRS and 31P-MRS time constants. For all groups, NIRS time constants were faster than PCr time constants (37.1 6 13.7 vs. 47.3 6 13.7 for NIRS and 31P-MRS, respectively). Comparison of NIRS and

MR Imaging of Muscle Volume and Composition.

Muscle volume of the triceps surae muscles was calculated as the sum of cross-sectional areas from each image. Muscle volume was 1134 6 299, 667 6 200, and 567 6 149 cm3 for young controls, age-matched controls, and ALS, respectively (Fig. 1b). The muscle volume was significantly different between young and age-matched controls (P 5 0.019) and between young controls and ALS (P 5 0.002), but not different between age-matched controls and ALS (P 5 0.779). The mean fat percentage was 7.1 6 1.3, 15.0 6 5.5, and 13.5 6 3.4 % for young controls, age-matched controls, and ALS, respectively. The fat percent was significantly different between young and age-matched controls (P 5 0.009) and between young controls and ALS (P 5 0.018), but it was not different between agematched controls and ALS (P 5 0.782). DISCUSSION

We did not find a statistically significant impairment in mitochondrial function in the triceps surae muscles of people with ALS compared with age-matched controls. These results are consistent with a previous study that did not find differences between ALS and controls using 31P-MRS to measure mitochondrial function in vivo.10,11 Of interest, there were trends for people with greater clinical involvement of the calf muscles to have slower recovery rates for both 31P-MRS and NIRS. Because of the rapid progression of symptoms of ALS, we found it difficult to recruit and test participants who had significant muscle involvement and yet had enough muscle function so that we could activate their muscles in the tests. Mitochondrial function measured in vitro from muscle biopsies has identified abnormalities in people with ALS. These skeletal muscle mitochondrial abnormalities include morphological changes 772

Muscle Metabolism in ALS

with electron microscopy,24 mitochondrial DNA mutations,3 mitochondrial uncoupling,25 and altered mitochondrial enzyme concentrations.7 Assessments of the mitochondrial respiratory chain (i.e., ATP producing capacity) in ALS have produced mixed results. Echaniz-Laguna et al.7 reported an elevated maximal respiratory capacity (State 3 respiration) in muscle tissue of newly diagnosed patients with ALS. In contrast, Vielhaber and colleagues3 reported decreased maximal mitochondrial respiratory capacity and specific defects in Complexes I and IV of the electron transport system in patients with ALS. Krasnianski et al.26 reported no difference in mitochondrial respiratory chain activities between patients with ALS and healthy controls. The key difference between the in vitro and in vivo measurements may be the presence and contribution of denervated skeletal muscle fibers. In in vitro measurements all muscle fibers in the biopsy are evaluated, and denervated muscle contributes to the mitochondrial assessments. In contrast, the in vivo approaches such as NIRS and 31P-MRS provide information only from muscle fibers with intact nerves, because any change in muscle oxygen consumption or [PCr] are due to muscle fibers activated by the exercise bout. Of interest, in 1 participant with ALS, the electrical stimulation exercise did not decrease the concentration of PCr to a degree that would allow recovery measurements (less than 10% depletion). However, the NIRS testing was successful in this participant, with the electrical stimulation increasing mVO2 fivefold resting. Future studies of skeletal muscle mitochondrial function in patients who have a combination of muscle inactivity, muscle paralysis, and muscle denervation may need to consider the relative values of in vivo and in vitro measurements in their design. We found generally good agreement between the 2 in vivo measurements of mitochondrial function. The test–retest variability was generally very good and similar to previous studies,27–29 particularly for 31P-MRS measurements (10% for CV). We found slightly higher CV in our NIRS studies than have been reported in previous studies (up to 25% in the older control group).13 To develop a better understanding of the variability in our study, we evaluated SNRs for both 31P-MRS and NIRS measurements. Clear presentations of SNR are rarely reported for these measurements, even though they may be critical in interpreting the results. Interestingly, we found evidence that low SNR contributed to the increased CV in the NIRS measurements, but not in the 31P-MRS measurements. The higher SNR we found in the PCr values relative to the SNR for the NIRS measures may explain this. MUSCLE & NERVE

November 2014

The 31P-MRS and NIRS measurements provide similar assessments of skeletal muscle mitochondrial function.30 The advantage of the 31P-MRS measurements is that they could be combined with MRI measurements of muscle size and intramuscular fat. We found the older groups (ALS and controls) to have smaller muscles and higher levels of intramuscular fat, although there was no difference between ALS and older controls. The NIRS measured recovery rates of mVO2 were faster compared with 31P-MRS measured rates of PCr resynthesis. This contrasts with previous studies by Nagasawa et al.30 and Ryan et al.22 There are possible explanations for this difference. First, the difference in electrical stimulation exercise duration between 31P-MRS and NIRS (60 seconds vs. 15 s, respectively) could have resulted in acidosis of the exercised muscles, which has been shown to slow the recovery of PCr.31,32 This study used nonlocalized 31P-MRS. Given the architecture of the phosphorus coil used (13 cm), the sensitive volume of the coil is estimated to be approximately 6.5 cm. Therefore, it is possible that a significant portion of nonexercised muscle contributed to the phosphorus spectra collected. This could result in greater depletion of PCr (and decreased pH) in the exercised muscle than is measured in the nonlocalized spectra. Previous studies have shown slower nonlocalized PCr recovery kinetics (and lower initial rates of PCr resynthesis) in comparison to localized measurements.33 The resolution of phosphorus spectra in this study was not great enough to resolve a split inorganic phosphate peak, which may result from having muscle fiber pools with different pH. We examined the residuals between the PCr data and the monoexponential curve fitting. A positive slope (of the residuals) was found in all data during the last 150 s of recovery. This suggests a slower component of PCr resynthesis, which is consistent with the dependence of PCrTc on pH reported in previous studies.32,34 Differences in muscle size or volume between participants could exacerbate the effects of nonlocalization in some participants. Furthermore, the depth of activation using electrical stimulation is unknown in this study. Submaximal electrical stimulation has been shown to result in incomplete and nonuniform activation patterns.35 Maximal stimulation currents are difficult to obtain using surface electrical stimulation in humans due to limitations in pain thresholds. In contrast, the NIRS device has a much smaller sensitive volume. The sampling depth of NIR light has been estimated to be approximately half the source-detector distance or less. Therefore, the NIRS device used in this study had a maximal sampling depth of 2 cm, and the NIRS data are less Muscle Metabolism in ALS

likely to be contaminated by nonactive musculature than 31P-MRS data. Limitations. The sample size of this study was small. There were some unique challenges of the recruitment process that contributed to this. First, due to the rapid progression of ALS, it was necessary to enroll participants early in the disease process. Second, we attempted to recruit patients with ALS who had some lower leg impairment, without a large degree of denervation. Third, these patients were required to have good respiratory function to lie safely inside the bore of the MRI for the duration of testing (30 min). Due to these requirements, ALS patients with only upper limb or bulbar impairment were excluded, thereby decreasing the number of potential patients. The correlation between muscle oxidative capacity and our clinical assessment of calf function was exploratory. The clinical assessment in this study was a rank based on expert clinical opinion. The potential relationship between mitochondrial function and calf function suggests that future studies should include more quantitative measures of clinical muscle involvement. This will be important to determine if muscle metabolism measurements can be useful to help evaluate disease status and progression in ALS. CONCLUSION

In summary, in vivo skeletal muscle mitochondrial function, measured with NIRS and 31P-MRS, was similar in ALS patients compared with agematched controls, but it was lower compared with a group of younger controls. We found a relationship between lower leg function and skeletal muscle mitochondrial function in patients with ALS, supporting the notion that variations in mitochondrial dysfunction may be related to the degree of disease progression and functional ability of patients. Both NIRS and 31P-MRS measurements could be performed successfully in patients with ALS, although NIRS may be a more sensitive technique for patients with greater amounts of denervated muscle. We thank Sarah Stoddard, Jared Brizendine, Hui-Ju (Zoe) Young, Michael Smith, W. Michael Southern, and the other students in the Exercise Vascular Biology Laboratory for their help with data collection. We also thank Brandy Quarles who played a major role in recruiting patients for this study. REFERENCES 1. Rowland LP, Shneider NA. Amyotrophic lateral sclerosis. N Engl J Med 2001;344:1688–1700. 2. Bowling AC, Schulz JB, Brown RH Jr, Beal MF. Superoxide dismutase activity, oxidative damage, and mitochondrial energy metabolism in familial and sporadic amyotrophic lateral sclerosis. J Neurochem 1993;61:2322–2325. 3. Vielhaber S, Kunz D, Winkler K, Wiedemann FR, Kirches E, Feistner H, et al. Mitochondrial DNA abnormalities in skeletal muscle of patients with sporadic amyotrophic lateral sclerosis. Brain 2000; 123(Pt 7):1339–1348. 4. Menzies FM, Cookson MR, Taylor RW, Turnbull DM, ChrzanowskaLightowlers ZM, Dong L, et al. Mitochondrial dysfunction in a cell

MUSCLE & NERVE

November 2014

773

5. 6. 7.

8. 9. 10.

11. 12. 13.

14.

15.

16.

17.

18. 19.

20.

culture model of familial amyotrophic lateral sclerosis. Brain 2002; 125(Pt 7):1522–1533. Menzies FM, Ince PG, Shaw PJ. Mitochondrial involvement in amyotrophic lateral sclerosis. Neurochem Int 2002;40:543–551. Dupuis L, Gonzalez de Aguilar JL, Echaniz-Laguna A, Loeffler JP. Mitochondrial dysfunction in amyotrophic lateral sclerosis also affects skeletal muscle. Muscle Nerve 2006;34:253–254. Echaniz-Laguna A, Zoll J, Ponsot E, N’Guessan B, Tranchant C, Loeffler JP, et al. Muscular mitochondrial function in amyotrophic lateral sclerosis is progressively altered as the disease develops: a temporal study in man. Exp Neurol 2006;198:25–30. Johri A, Beal MF. Mitochondrial dysfunction in neurodegenerative diseases. J Pharmacol Exp Ther 2012;342:619–630. Meyer RA. A linear-model of muscle respiration explains monoexponential phosphocreatine changes. Am J Physiol 1988;254:C548–C553. Grehl T, Fischer S, Muller K, Malin JP, Zange J. A prospective study to evaluate the impact of 31P-MRS to determinate mitochondrial dysfunction in skeletal muscle of ALS patients. Amyotroph Lateral Scler 2007;8:4–8. Sharma KR, Kent-Braun JA, Majumdar S, Huang Y, Mynhier M, Weiner MW, et al. Physiology of fatigue in amyotrophic lateral sclerosis. Neurology 1995;45:733–740. McCully KK, Hamaoka T. Near-infrared spectroscopy: what can it tell us about oxygen saturation in skeletal muscle? Exerc Sport Sci Rev 2000;28:123–127. Ryan TE, Erickson ML, Brizendine JT, Young HJ, McCully KK. Noninvasive evaluation of skeletal muscle mitochondrial capacity with near-infrared spectroscopy: correcting for blood volume changes. J Appl Physiol 2012;113:175–183. Motobe M, Murase N, Osada T, Homma T, Ueda C, Nagasawa T, et al. Noninvasive monitoring of deterioration in skeletal muscle function with forearm cast immobilization and the prevention of deterioration. Dyn Med 2004;3:2. Hamaoka T, McCully KK, Niwayama M, Chance B. The use of muscle near-infrared spectroscopy in sport, health and medical sciences: recent developments. Philos Trans A Math Phys Eng Sci 2011;369: 4591–4604. Ryan TE, Brizendine JT, McCully KK. A comparison of exercise type and intensity on the noninvasive assessment of skeletal muscle mitochondrial function using near-infrared spectroscopy. J Appl Physiol 2013;114:230–237. Brooks BR, Miller RG, Swash M, Munsat T. World federation of neurology research group on motor neuron diseases. El Escorial revisited: revised criteria for the diagnosis of amyotrophic lateral sclerosis. Amyotroph Lateral Scler Other Motor Neuron Disord 2000;1:293– 299. Medical Research Council. Aids to the examination of the peripheral nervous system, Memorandum No. 45. London: Her Majesty’s Stationery Office; 1976. The Amyotrophic Lateral Sclerosis Functional Rating Scale. Assessment of activities of daily living in patients with amyotrophic lateral sclerosis. The ALS CNTF treatment study (ACTS) phase I-II Study Group. Arch Neurol 1996;53:141–147. Harris RC, Hultman E, Nordesjo LO. Glycogen, glycolytic intermediates and high-energy phosphates determined in biopsy samples of

774

Muscle Metabolism in ALS

21. 22.

23.

24. 25.

26.

27.

28.

29. 30.

31.

32. 33.

34. 35.

musculus quadriceps femoris of man at rest. Methods and variance of values. Scand J Clin Lab Invest 1974;33:109–120. Moon RB, Richards JH. Determination of intracellular Ph by P-31 magnetic-resonance. J Biol Chem 1973;248:7276–7278. Ryan TE, Southern WM, Reynolds MA, McCully KK. A crossvalidation of near infrared spectroscopy measurements of skeletal muscle oxidative capacity with phosphorus magnetic resonance spectroscopy. J Appl Physiol 2013;115:1757–1766. Bogner W, Chmelik M, Schmid AI, Moser E, Trattnig S, Gruber S. Assessment of (31)P relaxation times in the human calf muscle: a comparison between 3 T and 7 T in vivo. Magn Reson Med 2009;62: 574–582. Afifi AK, Aleu FP, Goodgold J, MacKay B. Ultrastructure of atrophic muscle in amyotrophic lateral sclerosis. Neurology 1966;16:475–481. Dupuis L, di Scala F, Rene F, de Tapia M, Oudart H, Pradat PF, Meininger V, et al. Up-regulation of mitochondrial uncoupling protein 3 reveals an early muscular metabolic defect in amyotrophic lateral sclerosis. FASEB J 2003;17:2091–2093. Krasnianski A, Deschauer M, Neudecker S, Gellerich FN, Muller T, Schoser BG, et al. Mitochondrial changes in skeletal muscle in amyotrophic lateral sclerosis and other neurogenic atrophies. Brain 2005; 128(Pt 8):1870–1876. Lanza IR, Bhagra S, Nair KS, Port JD. Measurement of human skeletal muscle oxidative capacity by 31P-MR spectroscopy: a crossvalidation with in vitro measurements. J Magn Reson Imaging 2011; 34:1143–1150. Layec G, Bringard A, Le Fur Y, Vilmen C, Micallef JP, Perrey S, et al. Reproducibility assessment of metabolic variables characterizing muscle energetics in vivo: a P-31-MRS study. Magn Reson Med 2009;62: 840–854. McCully KK, Turner TN, Langley J, Zhao Q. The reproducibility of measurements of intramuscular magnesium concentrations and muscle oxidative capacity using 31P MRS. Dyn Med 2009;8:5. Nagasawa T, Hamaoka T, Sako T, Murakami M, Kime R, Homma T, et al. A practical indicator of muscle oxidative capacity determined by recovery of muscle O 2 consumption using NIR spectroscopy. Eur J Sport Sci 2003;3:1–10. Iotti S, Lodi R, Frassineti C, Zaniol P, Barbiroli B. In-vivo assessment of mitochondrial functionality in human gastrocnemius-muscle by P-31 MRS - the role of Ph in the evaluation of phosphocreatine and inorganic-phosphate recoveries from exercise. NMR Biomed 1993;6: 248–253. Walter G, Vandenborne K, McCully KK, Leigh JS. Noninvasive measurement of phosphocreatine recovery kinetics in single human muscles. Am J Physiol 1997;41:C525–C534. Meyerspeer M, Robinson S, Nabuurs CI, Scheenen T, Schoisengeier A, Unger E, et al. Comparing localized and nonlocalized dynamic 31P magnetic resonance spectroscopy in exercising muscle at 7 T. Magn Reson Med 2012;68:1713–1723. McCann DJ, Mole PA, Caton JR. Phosphocreatine kinetics in humans during exercise and recovery. Med Sci Sports Exerc 1995; 27:378–389. Adams GR, Harris RT, Woodard D, Dudley GA. Mapping of electrical muscle stimulation using MRI. J Appl Phsyiol 1993;74:532– 537.

MUSCLE & NERVE

November 2014