Isometric resistance training increases strength and alters histopathology of dystrophin-deficient mouse skeletal muscle

Abstract

Mutation to the dystrophin gene causes skeletal muscle weakness in patients with Duchenne muscular dystrophy (DMD) or Becker muscular dystrophy (BMD). Deliberation continues regarding implications of prescribing exercise for these patients. The purpose of this study was to determine whether isometric resistance exercise (~10 tetanic contractions/session) improves skeletal muscle strength and histopathology in the mdx mouse model of DMD. Three isometric training sessions increased in vivo isometric torque (22%) and contractility rates (54%) of anterior crural muscles of mdx mice. Mice expressing a BMD-causing missense mutated dystrophin on the mdx background showed comparable increases in torque (22%), while wild-type mice showed less change (11%). Increases in muscle function occurred within 1 h and peaked 3 days posttraining; however, the adaptation was lost after 7 days unless retrained. Six isometric training sessions over 4 wk caused increased isometric torque (28%) and contractility rates (22–28%), reduced fibrosis, as well as greater uniformity of fiber cross-sectional areas, fewer embryonic myosin heavy-chain-positive fibers, and more satellite cells in tibialis anterior muscle compared with the contralateral untrained muscle. Ex vivo functional analysis of isolated extensor digitorum longus (EDL) muscle from the trained hindlimb revealed greater absolute isometric force, lower passive stiffness, and a lower susceptibility to eccentric contraction-induced force loss compared with untrained EDL muscle. Overall, these data support the concept that exercise training in the form of isometric tetanic contractions can improve contractile function of dystrophin-deficient muscle, indicating a potential role for enhancing muscle strength in patients with DMD and BMD.

NEW & NOTEWORTHY We focused on adaptive responses of dystrophin-deficient mouse skeletal muscle to isometric contraction training and report that in the absence of dystrophin (or in the presence of a mutated dystrophin), strength and muscle histopathology are improved. Results suggest that the strength gains are associated with fiber hypertrophy, reduced fibrosis, increased number of satellite cells, and blunted eccentric contraction-induced force loss in vitro. Importantly, there was no indication that the isometric exercise training was deleterious to dystrophin-deficient muscle.

INTRODUCTION

Loss of dystrophin protein causes Duchenne muscular dystrophy (DMD) (31). Expression of a truncated and partially functional dystrophin protein causes the milder Becker muscular dystrophy (BMD). Because dystrophin localizes to the sarcolemma of striated muscle fibers and anchors the dystrophin-glycoprotein complex (DGC) to the intracellular cytoskeleton (19, 20), muscle from DMD and BMD patients lack stability and are susceptible to sarcolemmal damage (35, 49). DMD is characterized by progressive skeletal muscle wasting, overt cardiomyopathies, and loss of ambulation (18), while BMD patients suffer from a less severe form of myopathy, the extent to which depends on the mutation of the dystrophin gene and protein expression level (8). Interventions aiming to ameliorate striated muscle pathology in DMD are in development (14, 17, 30, 41, 44), but there is still a necessity to explore complementary treatments that either slow disease progression or provide an improvement in quality of life.

Prescription of exercise for patients with DMD is not standard practice because of a lack of definitive guidelines (45), as well as some cautionary results in studies with the mdx mouse model of DMD, which show swimming or running may exacerbate respiratory and cardiac histopathology (61). However, other studies with the mdx mouse show that endurance types of exercise like swimming or running can result in adaptations that improve the dystrophin-deficient skeletal muscle phenotype (7, 12, 27, 28, 32, 43, 60). The potential for skeletal muscle adaptations to be induced by resistance exercise training is even less clear, and a cautionary point of this type of exercise for patients with DMD is eccentric muscle contractions (3). This caution emanates from the use of maximal eccentric contractions in mdx mouse muscle showing high susceptibility of dystrophin-deficient skeletal muscle to force loss (48, 56). Despite the widespread demonstration of eccentric contractions causing loss of strength, it has also been shown that strength loss following eccentric contractions in vivo can be fully recovered in mdx mice and that repeated bouts of eccentric contraction in vivo can improve muscle strength in mdx mice (11). Therefore, while dystrophin-deficient muscle is exceedingly susceptible to sarcolemmal damage and acute loss of strength following eccentric contractions, it is a highly adaptable tissue that may benefit from exercise.

Isometric contractions generate force or torque without a change in muscle length or joint angle. In healthy adult muscle, resistance training using isometric contractions can induce muscle hypertrophy (1), and increase fiber cross-sectional area (22), torque (36), as well as rate of torque development (36). There is also evidence that isometric resistance exercise may improve muscle strength in patients with amyotrophic lateral sclerosis (15) and spinal muscular atrophy (39). Therefore, the purpose of this study was to determine whether isometric resistance training improved skeletal muscle strength and histopathology in dystrophin-deficient skeletal muscle of mice.

MATERIALS AND METHODS

Ethical Approval and Experimental Mice

Male mdx (C57BL/10ScSn-DMD^mdx/J) and C57BL/10 (C57BL/10ScSn/J) mice were purchased from Jackson Laboratory (Bar Harbor, ME). Generation of Dys^{∆R4–23/∆CT}-mdx (40) and mL172H-mdx (46) were previously reported. All mice used in this study were 6 mo old and chosen to represent a period of stable skeletal muscle degeneration and regeneration (55). All mice were housed in groups of 3 or 4 per cage on a 14:10-h light-dark cycle with food and water provided ad libitum. All animal protocols were approved by the University of Minnesota Institutional Animal Care and Use Committee (1707–34941A) and conform to the European Convention for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes.

Study 1: Determine Whether Dystrophin-Deficient Skeletal Muscle Adapts to Short-Term Isometric Strength Training

Study design.

C57BL/10 (n = 10) and mdx (n = 15) mice completed three isometric training sessions over 1 wk. Sessions were completed on days 1 (doubled as baseline strength assessment), 3, and 7, which were followed by a set of eccentric contractions immediately following the last session. Submaximal and maximal isometric tetanic torque and maximal tetanic rates of contraction and relaxation were measured during each session. Recovery of isometric torque following eccentric injury was measured on days 8, 10, 14, and 17 posteccentric contractions. Following the last contraction for each session (except day 17), mice recovered in their home cage.

In vivo isometric training.

Mice were anesthetized with isoflurane, and peak isometric torque generation of the anterior crural muscles (tibialis anterior, extensor digitorum longus, and extensor halluces longus) was measured as previously described (12, 13). To do this, the left hindlimb was depilated and aseptically prepared, and the foot was placed in a foot plate attached to a servomotor (model no. 300B‐LR; Aurora Scientific, Aurora, ON, Canada) and Pt-Ir electrode wires (model no. E2‐12; Grass Technologies, West Warwick, RI) were inserted percutaneously on either side of the peroneal nerve. Contractions were induced via stimulation of the peroneal nerve by a stimulator and stimulus isolation unit (model no. S48 and SIU5, respectively; Grass Technologies). Peak isometric tetanic torque was measured by manipulating voltage (4.3 ± 1.1 for C57BL/10 and 4.8 ± 0.9 for mdx) at 250 Hz every minute until a plateau was attained (3–5 contractions). One minute later, a torque-frequency relationship was established at varying stimulation frequencies (20, 40, 60, 80, 100, 125, 150, 200, 250, 300, 350, and 400 Hz). Because the anterior crural muscles of a mouse hindlimb reaches maximal torque at ~125–150 Hz (13), the total number of isometric tetanic contractions completed by each mouse/session was ~10 (i.e., the number of contractions required to identify peak torque + contractions of the torque-frequency assessment). Training sessions used tetanic contractions to mimic the maximal voluntary contractions used in human isometric resistance training (16a). Frequency 50 (the frequency required to reach 50% maximal torque), maximal rates of contraction, and relaxation were calculated.

Eccentric contraction protocol.

After the torque frequency on the 7th day, the anterior crural muscles were injured by performing a series of eccentric contractions (13). Briefly, the foot was passively rotated from 0° to 19° dorsiflexion followed by 38° of plantarflexion at 2000°/s. The peroneal nerve was stimulated for 100 ms with no movement about the ankle while at 19° dorsiflexion, immediately followed by a 20-ms stimulation at optimal frequency (125–150 Hz, as determined by the torque frequency) during the plantarflexion to elicit an eccentric contraction of the anterior crural muscles. Each eccentric contraction was separated by 10 s. Because dystrophin-deficient mdx muscle is highly susceptible to eccentric contraction-induced force loss, we completed eccentric contractions until 70% force loss was attained. Because C57BL/10 mice are more resilient to eccentric contractions than mdx mice, eccentric contractions were performed until 50% force loss was attained. These values were chosen because C57BL/10 mice would likely need to complete hundreds of contractions to reach 70% torque loss in vivo (65), and it is difficult to limit torque loss of mdx skeletal muscle to 50% because of the large incremental drops in torque from 0 through 60% loss (53). Five minutes following completion, another torque frequency was performed. Recovery of isometric torque following injury was determined by remeasuring isometric torque using the torque-frequency analysis at 1, 3, 7, and 10 days later.

Study 2: Determine the Rate of Dystrophin-Deficient Skeletal Muscle Adaptation to Isometric Strength Training and the Dependence on Dystrophin

Study design.

Because of the adaptations in mdx mice measured in Study 1, we wanted to first determine the rate of change in isometric torque. Using five separate cohorts of mdx mice (n = 5–13/time point), a single session of isometric training was completed followed by a second session of either 1 h, 6 h, 1 day, 3 days, or 7 days later. Each isometric training session was identical to that described in Study 1. Changes in isometric torque, frequency 50, and rates of contraction and relaxation were measured. Mice were euthanized with 200 mg/kg pentobarbital sodium, and the tibialis anterior (TA) was carefully excised and snap frozen in liquid nitrogen.

We also wanted to determine whether the adaptations associated with isometric training were dependent on the protein dystrophin. We utilized two mdx mouse models that transgenically express a skeletal muscle-specific microdystrophin (Dys^{∆R4–23/∆CT}-mdx) (40) or a missense mutated dystrophin replicating a Becker muscular dystrophy (BMD) model (mL172H) (46). Dys^{∆R4–23/∆CT}-mdx mice that are rescued for most skeletal muscle phenotypes (53) and, thus, represent a model similar to C57BL/10. mL172H mice present with a dystrophic phenotype that is partially rescued above the skeletal muscle phenotype of mdx mice and, thus, represents an intermediate model for testing the question of whether isometric training adaptation is dystrophin-dependent. Dys^{∆R4–23/∆CT}-mdx (n = 7) and mL172H mice (n = 8) each completed the short-term isometric training, as described in Study 1.

Study 3: Determine Whether Dystrophin-Deficient Skeletal Muscle Adapts to Long-Term Isometric Strength Training

Study design.

Mdx mice completed six isometric training sessions over 4 wk (n = 15) to assess whether the short-term adaptations observed in Study 1 over 1 wk were maintained. Sessions were completed on days 1, 3, 7, 14, 21, and 28, and each isometric training session was identical to that described in Study 1. A 4-wk training protocol was chosen because it limited the number of times the mice were anesthetized and the number of times the skin was pierced with stimulating electrodes. Immediately before the training session on days 1 and 28, mice completed an in vivo passive torque assessment of the ankle because mdx mice have elevated passive mechanical properties of the anterior hindlimb muscles (9). As resistance exercise can affect skeletal muscle blood flow (29), 3 days following the sixth training session, eight randomly selected mice were assessed for in vivo blood flow to measure any effect of training. Then, 1 wk following the sixth session, the group was split into two separate cohorts (n = 7 and 8) for completion of ex vivo muscle function analysis of the extensor digitorum longus (EDL) and histological analyses of the TA for myosin/actin content, fibrosis, central nucleation, fiber cross-sectional area (CSA), embryonic myosin heavy chain (eMHC) expression, and satellite cell number. We chose these parameters because each is perturbed in the mdx mouse or is affected by resistance exercise.

Ankle joint mechanics.

Although mice were anesthetized with isoflurane (n = 15), passive torque about the left ankle joint was assessed by stabilizing the knee and placing the foot in a plate attached to a servomotor, as previously described (24). Briefly, the servomotor randomly rotated the ankle to four angles of dorsiflexion (5°, 10°, 15°, and 20°) and four angles of plantarflexion (5°, 10°, 15°, and 20°), and torque was measured at each angle.

Doppler imaging.

Red blood cell flux was measured in mice (n = 15) using the moorLab laser-Doppler flow meter with the MP7a probe (Moor Instruments, Millwey, Axminster, UK), as previously described (63). We used these experiments because isoflurane causes a reduction in systemic vascular resistance through vasodilation (16). The fur from both legs was removed using a avertin (2,2,2 tribromoethanol at 250 mg/kg body mass; Sigma Aldrich; cat. no. T48402) for chemical depilatory to allow for direct contact with the tissue. The TA muscle was gently placed on top of the probe to avoid compressing the tissue. Single readings were taken using the probe from at least 10 different positions on the TA muscles. An arbitrary unit (AU) was determined as the average AU value during a plateau phase of each measurement. Muscle perfusion was measured in the trained muscle and compared with the contralateral untrained control. The investigator was blinded to the training status of each mouse.

Ex vivo muscle preparation and physiology.

Mice were anesthetized with pentobarbital sodium (75 mg/kg body mass for mdx). Baseline contractile functions of EDL muscles (n = 5/group) were assessed according to the methods of Moran et al. (51). Maximal isometric tetanic force (P_o) was measured every 2 min by stimulating the muscle to contract for 200 ms at 175 Hz until force plateaued (within 5 mN from one contraction to the next). Two minutes later, a force frequency was performed (10, 20, 25, 30, 40, 50, 60, 80, 100, 120, and 160 Hz for 200 ms with 3 min between contractions). Two minutes later, a single concentric contraction was performed by passively lengthening to 102.5% L_o and then stimulated at 175 Hz for 200 ms, while the muscle was simultaneously shortened to 97.5% L_o at 0.5 L_o/s. Two minutes later, five eccentric contractions were performed. For each eccentric contraction, the muscle was passively shortened to 97.5% L_o and then stimulated at 175 Hz for 200 ms, while the muscle was simultaneously lengthened to 102.5% L_o at 0.5 L_o/s. Each eccentric contraction was separated by 3 min of rest to prevent fatigue (42). The force measured at each eccentric contraction was expressed as a percentage of the force produced during the first (“initial”) contraction.

Myosin/actin Western blot analysis.

Frozen TA muscles (n = 5/group) were analyzed for myosin and actin content via Western blot analysis, as previously described (52).

Muscle histological analysis.

TA muscle (n = 5/group) was frozen in liquid nitrogen-cooled isopentane and stored at −80°C until analysis. Central nucleation and CSA were measured using hematoxylin-and-eosin (H&E) staining of 10-µm-thick sections cut midbelly of the muscle. All fibers within the TA were counted for central nucleation, while CSA was measured for 800 fibers/muscle using 4 × 200 fiber regions located in four distinct areas of the muscle (top, bottom, left and right). Fibrosis was quantified by Sirius red/fast-green stain for collagen using 10-µm-thick sections cut midbelly of the muscle. Visualization and quantification of images were conducted using bright-field (Leica DM5500 B) at ×10 magnification, and ImageJ was used to adjust the color threshold of Sirius red/fast-green stain and quantify the fibrotic area (red) as a percentage of the total muscle area.

Embryonic myosin heavy-chain immunofluorescence.

Ten-micrometer cross sections (n = 5/group) of the TA muscle were fixed in 4% PFA for 10 min. Slides were then blocked with 3% BSA in PBS for 1 h at room temperature, which was followed by incubation with 1:20 dilution eMHC antibody (F1.652; Developmental Studies Hybridoma Bank, Iowa City, IA) and anti-laminin (rabbit polyclonal 1:250, ab11575; Abcam, Cambridge, MA) in 3% BSA in 1 × PBS for a further 1 h at room temperature. Slides were then washed 3 × 5 min with 1 × PBS and incubated with a 1:750 dilution of Alexa Fluor 555 goat anti-mouse IgG and Alexa Fluor 480 anti-rabbit IgG 1:750 in 3% BSA in 1 × PBS at room temperature for 45 min. Slides were washed again in 1 × PBS 3 × 5 min and DAPI immunomount anti-fade ProLong Gold reagent was added to each slide. Visualization and quantification of images were conducted using fluorescence (Leica DM5500 B) at ×20 magnification, and ImageJ was used to count the total number of positive eMHC fibers as a percentage of total fibers in the cross section.

Satellite cell isolation and quantification.

Isolation of satellite cells from TA muscle (n = 8/group) was performed, as described previously (4). Muscles were carefully dissected and chopped in parallel with muscle fibers using razor blade and forceps to separate the fibers. Muscles were incubated shaking for 75 min in 0.2% collagenase type II (17101-015; Gibco, Grand Island, NY) in high glucose DMEM without phenol red containing 4.00 mM l-glutamine, 4,500 mg/l glucose, and sodium pyruvate (SH30284.01; Hyclone, Logan, UT) supplemented with 1% Pen/Strep (15140122; Gibco) at 37°C. Samples were washed with rinsing solution (F-10+), Ham’s/F-10 medium (SH30025.01; HyClone) supplemented with 10% horse serum, 1% HEPES buffer solution (15630080; Gibco), and 1% Pen/Strep (Gibco) and centrifuged at 1,500 rpm × 5 min at 4°C. Samples were washed and centrifuged a second time. Samples were pulled into a sheared Pasteur pipette, centrifuged, and washed again. Following aspiration, samples were resuspended in F-10+ with collagenase type II and dispase (17105–041; Gibco), vortexed, and incubated shaking at 37°C for 30 min. Samples were vortexed again, drawn, and released into a 3-ml syringe with 16-gauge needle four times and then with a 18-gauge needle four times and passed through a 40-µm cell strainer (Falcon, Hanover Park, IL). Three ml of F-10+ was added to each sample and centrifuged at 1,500 rpm × 5 min 4°C. Following aspiration, samples were resuspended in FACS staining medium (2% FBS in PBS).

Muscle samples were stained using an antibody mixture of PE-Cy7 rat anti-mouse CD31 (clone 390), PE-Cy7 rat anti-mouse CD45 (clone 30-F11), Biotin rat anti-mouse CD106 [clone 429(MVCAM.A)], PE streptavidin from BD Biosciences (San Diego, CA) and Itga7 647 (clone R2F2) from AbLab (Vancouver, BC, Canada). An antibody cocktail was added to samples and incubated on ice for 30 min. Samples were washed and resuspended with FACS staining media containing propidium iodide for FACS analysis on the FACSAriaII SORP (BD Biosciences, San Diego, CA). Total number of satellite cells (lineage negative; VCAM, alpha7 double-positive cells) was quantified in the entire TA muscle sample.

Statistics.

Differences between groups in study 1 were analyzed by two-way repeated-measures ANOVA. When a difference was measured between groups, Dunnett’s post hoc analysis was completed to identify individual differences. Differences between groups for the rate of adaptation measurements in Study 2 were analyzed by one-way ANOVA using Tukey’s post hoc analysis if a significant effect was calculated. A one-way, repeated-measures ANOVA was used to analyze differences between training sessions for Study 3 using Tukey’s post hoc analysis if a significant effect was calculated. For all experiments in Study 4, groups were analyzed using either a paired t-test or repeated-measures, one-way ANOVA with Tukey’s post hoc analysis if a significant effect were calculated. All data are presented as means ± SE, with significance set at P < 0.05.

RESULTS

Study 1

Peak isometric torque of healthy anterior crural muscles from C57BL/10 mice did not change in response to the short-term training (Fig. 1A; P = 0.165), nor did specific isometric torque (Fig. 1B; P = 0.099). In contrast, both peak and specific isometric torque of mdx anterior crural hindlimb muscles increased by more than 20% from training session 1 to 3 (Fig. 1A and B; P < 0.001). Mdx muscle also had increased maximal tetanic rates of contraction and relaxation with training (>42%; Fig. 1, C and D; P < 0.001), while only the rate of relaxation at training session 2 was increased in C57BL/10 muscle (24%; Fig. 1D; P = 0.013).

Fig. 1.

Three sessions of isometric strength training increased anterior hindlimb muscle torque and contractility of mdx mice in vivo. Peak isometric torque (A), peak isometric torque normalized by body mass (B), maximal tetanic rates of contraction (C), and maximal rates of tetanic relaxation (D) in C57BL/10 (n = 10) and mdx (n = 15) mice. *Significantly different from training session 1 within genotype, A–C: P < 0.001; D: P < 0.013.

As part of the isometric strength training, a torque frequency analysis was completed. In both C57BL/10 and mdx muscle, there was a rightward shift in the curve as a function of peak torque (Fig. 2, A and B). Absolute torque of C57BL/10 mice was lower at 40 Hz following training (> 28%; Fig. 2C; P < 0.010), while torque as a percentage of maximum was lower at frequencies 20, 40, 60, and 80 Hz following session 2 and 3 (Fig. 2A; P < 0.05). For mdx mice, an increase in absolute torque was measured at 150 and 200 Hz (Fig. 2D; P < 0.05), while torque as a percentage of maximum was lower at 20, 40, 60, 80, and 100 Hz for training sessions 2 and 3 (Fig. 2B; P < 0.05) and further at 125 Hz for session 3 only (Fig. 2D; P = 0.017). Similarly, the frequency required to generate 50% of maximum torque increased from 30 ± 1.2 to 51 ± 1.5 Hz in C57BL/10 mice and from 34 ± 1.4 to 53 ± 1.3 Hz for mdx mice with training (P < 0.001).

Fig. 2.

Torque-frequency relationships for C57BL/10 and mdx mice following three sessions of isometric strength training. Isometric torque as a percentage of maximum for C57BL/10 (n = 10; A) and mdx (n = 15; B) mice. Absolute torque for C57BL/10 (C) and mdx (D) mice. *Significantly different from training session 1, A, B, and D: P < 0.05; C: P < 0.010.

Immediately following session 3, an eccentric contraction protocol was completed to identify whether the increase in isometric torque of mdx muscle affected muscles’ sensitivity to force loss. For C57BL/10 mice, peak eccentric torque did not differ between trained and untrained mice (7.7 ± 0.5 vs. 6.5 ± 0.3 mN·m; P = 0.058). The trained mice also took fewer contractions to reach 50% torque loss (53 ± 10 vs. 139 ± 13 contractions; Fig. 3A, inset; P < 0.001). For mdx mice, isometric training resulted in a trend toward greater peak eccentric torque (6.1 ± 0.5 vs. 7.6 ± 0.6 mN·m; P = 0.057). Similar to C57BL/10, the trained mdx mice reached 70% torque loss with fewer contractions (22 ± 3 vs. 68 ± 10 contractions; Fig. 3B, inset; P = 0.001). Because dystrophin-deficient mouse muscle can completely recover force following eccentric contractions in vivo (11), we measured the effect of short-term isometric training on torque recovery. C57BL/10 and mdx mice recovered 100% of isometric torque by 10 days in both trained and untrained groups (Fig. 3, A and B; P > 0.05).

Fig. 3.

Recovery of isometric torque 1 day to 10 days following a series of eccentric contractions. Recovery of isometric torque (A) and number of eccentric contractions (A, inset) to reach 50% torque loss in C57BL/10 mice (n = 10). Recovery of isometric torque (B) and number of eccentric contractions (B, inset) to reach 70% torque loss and in mdx mice (n = 15). *Significantly different from untrained, A: P < 0.001; B: P = 0.001.

Study 2

To determine whether the increased isometric torque by trained mdx muscle in Study 1 was transient, mice completed a single session of isometric training followed by a second session at various time points from 1 h to 7 days postexperiment. A single session caused an increase in isometric torque (Fig. 4A), maximal tetanic rate of contraction (Fig. 4B), and relaxation (Fig. 4C) that was not affected by whether the second session occurred between 1 h and 3 days post (P > 0.05). However, when the second session was completed 7 days post, isometric torque was lower than 1 and 3 days (P < 0.05), and maximal rates of contraction and relaxation were lower than all other time points (P < 0.05).

Fig. 4.

A single isometric strength training session elicits an immediate increase in torque. Isometric torque (A), maximal tetanic rate of contraction (B), and maximal rate of relaxation (C) following a single isometric training session. Each time point represents a separate group of mice (n ≥ 8/group) trained once and retested at the specified time post-training. *Significantly different from 1 day, P < 0.05. #Significantly different from 3 days, P < 0.05. $Significantly different from all other time points, P < 0.05.

Because we did not measure a change in isometric torque in C57BL/10 mice in response to training in study 1, as we did in mdx mice, we wanted to determine whether the strength adaptation was dystrophin-dependent. In mL172H mice, which express a missense mutant dystrophin associated with BMD, we measured a 19% increase in specific isometric torque in response to two and three sessions of training (Fig. 5A; P = 0.010). We also assessed the Dys^{∆R4–23/∆CT}-mdx mouse, which transgenically overexpresses a microdystrophin (40) and is rescued for most skeletal muscle phenotypes, but we did not measure any training-induced change in specific isometric torque (Fig. 5B; P = 0.952). We also measured increases in maximal tetanic rates of contraction in mL172H (Fig. 5C; P < 0.001) and Dys^{∆R4–23/∆CT}-mdx mice (Fig. 5D; P = 0.043) and an increase in maximal tetanic rates of relaxation in mL172H (Fig. 5E; P < 0.001) with no change in Dys^{∆R4–23/∆CT}-mdx mice (Fig. 5F; P = 0.558). The data suggest that isometric strength training adaptations are possible in the absence of dystrophin or in the presence of a mutated dystrophin but not a microdystrophin.

Fig. 5.

Isometric strength training increases in vivo isometric torque and muscle contractility in mL172H [expresses a Becker muscular dystrophy (BMD) mutated skeletal muscle-specific dystrophin; n = 8] but not Dys^{∆R4–23/∆CT}-mdx mice (expresses a skeletal muscle-specific microdystrophin; n = 7). Isometric torque (A and B), maximal rates of contraction (C and D), and maximal rates of relaxation (E and F) following three isometric training sessions. *Significantly different from training session 1, A: P = 0.010; C and E: P < 0.001; D: P = 0.043.

Study 3

Here, we completed six training sessions over 4 wk to determine whether the increase in isometric torque in mdx mice could be maintained with continued training. Mdx mice increased isometric torque throughout the six training sessions (Fig. 6A; P < 0.001), which also resulted in increased specific isometric torque (Fig. 6B; P = 0.005). The increase in strength was associated with increased maximal tetanic rates of contraction (Fig. 6C; P < 0.001) and relaxation (Fig. 6D; P < 0.001). We also measured a change in torque across the training period at multiple stimulation frequencies (sessions 2–6 differed from session 1 at 40, 60, 125, 150 and 200 Hz; Fig. 6E; P < 0.001), which was associated with a rightward shift in the torque frequency curve as a function peak torque (Fig. 6F; P < 0.001) and greater frequency required to produce 50% of maximal torque (Fig. 6G; P < 0.001).

Fig. 6.

Six isometric training sessions of the anterior hindlimb muscles over 4-wk increases and maintains in vivo isometric torque and muscle contractility of mdx mice (n = 15). Isometric torque (A), specific isometric torque (B), maximal tetanic rate of contraction (C), maximal rate of tetanic relaxation (D), absolute torque frequency (E), isometric torque as a percentage of maximum (F), and frequency required to generate 50% max torque (frequency 50) (G) following six sessions of isometric strength training. *Significantly different from training session 1, A C, D, and G: P < 0.001; B: P = 0.005. Statistical significance is not presented on graphs E and F for simplicity.

Resistance exercise is known to affect skeletal muscle blood flow (29), and dystrophin-deficient muscle is ischemic (37). Therefore, we analyzed in vivo blood flow of both the untrained and trained TA muscle. We measured no difference in red blood cell concentration (187 ± 6 vs. 194 ± 11 AU; P = 0.619), blood flow speed (22.6 ± 2.2 vs. 23.7 ± 6.5 AU; P = 0.883) or flux (82.3 ± 9.9 vs. 82.7 ± 19.6 AU; P = 0.987) between the untrained and trained hindlimbs, indicating that a change in blood flow did not contribute to improved strength with training.

Because dystrophin-deficient muscle has elevated fibrosis and passive stiffness (26), which can impact strength, we determined whether isometric training affected these characteristics. Compared with session 1, passive torque about the ankle was reduced when the ankle was at 10°, 15°, and 20° of dorsiflexion by session 6 (Fig. 7; P < 0.001). We subsequently stained for and quantified lower fibrosis in trained compared with untrained TA muscles (5.1 ± 0.5 vs. 2.9 ± 0.4% of total muscle area; Fig. 7, B and C; P = 0.021).

Fig. 7.

Six isometric training sessions over 4 wk reduces passive torque and lowers fibrosis of the anterior hindlimb muscles about the ankle of mdx mice. A: passive torque of the anterior hindlimb muscles about the ankle (n = 15). *Significantly different from training session 1, P < 0.001. Representative images of Sirius red/fast green stain for collagen from untrained and trained tibialis anterior (TA) muscles (B) and quantification of collagen staining (n = 5; C). *Significantly different from untrained, C: P = 0.021.

Because the TA contributes ~90% of torque produced during anterior crural muscle isometric contractions (66), we measured muscle wet mass, total protein content, and actin and myosin content of the untrained and trained TA to determine whether the increase in torque was associated with a change in contractile protein content. The TA wet mass of the trained leg was lighter (79.0 ± 1.8 vs. 75.3 ± 1.4 mg; P = 0.048), while there were no differences in total protein (3.9 ± 0.5 vs. 3.5 ± 0.2 mg; P = 0.187), or actin and myosin protein content between untrained and trained TA muscles (P ≥ 0.825). We did measure a rightward shift in CSA of the trained TA muscle fibers (Fig. 8, A and B). Noticeably, there were fewer fibers <1000 µm² and a greater number of fibers >1000 µm² in the trained compared with untrained TA (P ≤ 0.037).

Fig. 8.

Six isometric training sessions over 4-wk shifts the distribution of fiber size in tibialis anterior (TA) muscle fibers of mdx mice (n = 5). Representative images of hematoxylin-and-eosin (H&E)-stained fibers from untrained and trained TA muscles (A), cross-sectional area (CSA) of muscle fibers from untrained and trained TA muscle (B), and centrally nucleated muscle fibers (C) from untrained and trained TA muscle.

Central nucleation is also a robust phenotype of mdx mice (10). Therefore, we measured central nucleation of the untrained and trained TA muscle by H&E. Central nucleation was not affected by training status (Fig. 8C; P = 0.837). To determine whether the greater population of small fibers in the untrained TA were regenerative, we stained for eMHC (Fig. 9A) and measured a lower number of positive eMHC fibers in the trained TA compared with the untrained (3.6 ± 0.8 vs. 0.9 ± 0.2%; Fig. 9B; P = 0.031).

Fig. 9.

Six isometric training sessions over 4 wk results in fewer embryonic myosin heavy chain (eMHC) positive fibers in the tibialis anterior (TA) muscle of mdx mice (n = 5). Representative images of positively stained muscle fibers for DAPI, laminin and eMHC from untrained and trained TA muscles (A) and quantification of fibers positive for eMHC as a percentage of total fibers (B). *Significantly different from untrained, P = 0.031.

Because we measured a shift toward larger fibers (Fig. 8B) and lower eMHC-positive fibers (Fig. 9) following isometric training, we wanted to determine whether the regenerative capacity of the trained TA had been affected. Thus, we analyzed the total satellite cell pool of the untrained and trained TA muscles (Fig. 10A) and measured a 30% increase in the number of satellite cells in the trained TA muscles (Fig. 10B; P = 0.041).

Fig. 10.

Six isometric training sessions over 4 wk results in a greater number of satellite cells in the tibialis anterior (TA) muscle of mdx mice (n = 8). Representative flow cytometric analysis of satellite cells double positive for VCAM and alpha7 integrin from untrained and trained TA muscles normalized to TA muscle mass (A) and satellite cell number normalized to TA mass (B). *Significantly different from untrained, P = 0.041.

Neural adaptations play important roles in skeletal muscle strength gains (21). To get insight into whether the torque increases were associated with a neural adaptation, we measured ex vivo physiological function of isolated EDL muscles where contraction is initiated by opening of sodium channels on the plasmalemma. We measured higher isometric, eccentric, and concentric forces of trained compared with untrained EDL muscles (Table 1). However, when isometric force was normalized to muscle CSA, the difference became insignificant. Interestingly, we also found a significant partial protection against eccentric contraction-induced force loss (Fig. 11A; P = 0.02), which was corroborated by the postisometric tetanic force following the fifth eccentric contraction (Fig. 11B; P = 0.030). Although these data do not rule out some type of neural adaptation with training, it is not sufficient because strength gains occurred even when muscle was stimulated to contract downstream of the neuromuscular junction.

Table 1.

Physiological characteristics of isolated EDL muscles from untrained and trained mdx mice

Parameter	Untrained	Trained	P Value
EDL mass, mg	16.7 ± 0.8	17.6 ± 0.5	0.056
Lo, mm	13.1 ± 0.2	13.1 ± 0.1	0.418
CSA, mm2	2.7 ± 0.1	2.9 ± 0.1	0.080
Passive stiffness, N/m	15.4 ± 1.3	13.8 ± 1.0	0.044*
Active stiffness, N/m	683 ± 61	782 ± 55	0.063
Peak twitch, mN	95 ± 14	114 ± 10	0.094
Twitch force development time, ms	17.7 ± 0.3	17.8 ± 0.4	0.406
Twitch ½ relaxation time, ms	17.2 ± 0.9	17.4 ± 1.3	0.420
Po, mN	410 ± 49	461 ± 40	0.042*
Specific Po, N/cm2	15.0 ± 1.8	16.0 ± 1.4	0.155
Tetanic maximal rate of contraction, N/s	11.3 ± 1.3	12.9 ± 1.7	0.070
Tetanic maximal rate of relaxation, N/s	21.2 ± 3.2	23.5 ± 2.7	0.121
Peak eccentric force, mN	499 ± 57	608 ± 61	<0.001*
Peak concentric force, mN	264 ± 38	327 ± 31	0.014*
Frequency 50, Hz	35.5 ± 0.9	33.9 ± 1.1	0.200

Values are expressed as means ± SE. L_o, optimal muscle length; CSA, physiological cross-sectional area; P_o, maximal isometric tetanic force; frequency 50, Hz required to produce 50% of peak force. n = 5/group. Statistics were performed using paired t-test analyses with

^*P < 0.05 considered significant.

Fig. 11.

Six isometric training sessions over 4 wk results in a partial protection of isolated extensor digitorum longus (EDL) muscle from eccentric contraction-induced force loss. Eccentric force as a percentage of the first contraction of untrained and trained EDL muscle during five eccentric contractions (A) and isometric tetanic force of EDL muscles immediately following the 5th eccentric contraction relative to isometric force before the eccentric contractions (B). *Significantly different from untrained, P = 0.030. n = 5/group.

DISCUSSION

Exercise prescription for patients with DMD remains questionable (23). Although endurance exercise (7, 12, 27, 28, 32, 43, 60) and eccentric contractions (11) have been shown to improve skeletal muscle function in the dystrophin-deficient mdx mouse, exercise is not a standard therapy for those with DMD, likely due to fear of exacerbating the disease. The results of our study clearly demonstrate that repeated sessions of isometric tetanic contractions improve contractile function (Figs. 1, 4, and 6) and ameliorate several distinct histopathological features of dystrophin-deficient skeletal muscle of mice (Figs. 7–9). An advantage of isometric training is that it theoretically would circumvent the concern of potentially injurious lengthening (eccentric) contractions.

There is a dearth of clinical trials on boys with DMD that have examined the efficacy, reliability, and tolerability of exercise therapy. Between the 1960s and 1980s, clinical trials on patients with neuromuscular diseases showed that resistance-type exercise maintained or improved skeletal muscle strength, while also finding no evidence for exacerbation of disease (38, 59, 64). A recent study demonstrated that ergometry exercise improved endurance and muscular strength in DMD patients (2). Two active clinical trials are also examining exercise as a therapy for DMD patients. The effects of aerobic exercise on physical strength are being studied in ambulatory and nonambulatory boys with DMD (NCT03319030), and isometric resistance exercise on muscle strength and safety is being investigated in 7–10-yr-old boys with DMD (NCT02421523). The improved skeletal muscle strength and histopathology of the current study on mdx mice provides additional evidence to support the rationale for the latter clinical trial and also corroborates recent evidence on the use of exercise in the mdx mouse (61). However, consideration of the isometric training protocol is required because the tetanic contractions elicited in our study on 6-mo-old mdx mice (i.e., 100% of motor units were activated) resulted in significant benefits, but realistically, submaximal exercise prescription may be more appropriate for boys with DMD. Therefore, the submaximal voluntary isometric contractions proposed in the latter clinical trial on boys with DMD over a 12-wk program will need to be monitored and may be worth exploring in follow-up mouse studies.

Absence of the dystrophin protein causes a multitude of functional defects, including skeletal muscle weakness and susceptibility to injury in mdx mice. Our study provides evidence that repeated isometric tetanic contractions by the anterior hindlimb muscles of mdx mice can increase torque generation by those skeletal muscles. Importantly, the change in strength can be consistently maintained and even increased for up to 4 wk through a weekly isometric resistance training program (Fig. 6). The change in strength was also associated with an increase in the frequency required to generate 50% of peak torque, which is indicative of a shift in fiber type or a change in calcium kinetics following contraction. Because adaptation to the training protocol happened within an hour, it is more likely that training improved calcium kinetics of the anterior crural muscles rather than fiber-type shifting. The results also provide further support to studies that clearly define dystrophin-deficient muscle as a highly adaptive tissue (7, 11). Although an increase in strength was not associated with an improvement in resisting strength loss from eccentric contractions in vivo (Fig. 3), potentially through an improved circuit breaker adaptation that prevents injury from eccentric contraction (54), isometric torque still recovered to 100% within 10 days and highlights the adaptive and regenerative properties of dystrophin-deficient skeletal muscle.

Like DMD, BMD patients suffer from skeletal muscle wasting and weakness (47). While slowing disease progression and improving quality of life for DMD patients is critical, managing symptoms of BMD is equally justified. Increased isometric strength in an mdx mouse model expressing a mutated dystrophin from a BMD patient (L172H; Fig. 5) indicates that isometric resistance training could similarly become useful for this patient population. However, the isometric resistance training protocol did not affect muscle contractility in C57BL/10 mice or Dys^{∆R4–23/∆CT}-mdx mice that express a therapeutically relevant microdystrophin (Fig. 5). These results indicate that the isometric training protocol used in the current study is not adequate to elicit strength gains when fully functional dystrophin protein is present. Moreover, although Dys^{∆R4–23/∆CT}-mdx mice did not increase their strength using this study design, mice that transgenically express Dys^{∆R4–23/∆CT} present with skeletal muscle strength, morphology, and pathology similar to C57BL/10 mice, which still makes it a potential suitable therapy for DMD.

The practical implications of the current protocol are highlighted by the immediate and sustainable increase in isometric strength. The increase in strength after only 1 h (Fig. 4) suggests it may be associated with a neurological adaptation, as has been observed following isometric training in humans (58), in addition to possible calcium kinetic adaptations. A neurological adaptation could be associated with an improvement in neuromuscular junction (NMJ) function or excitation-contraction coupling. Because dystrophin deficiency affects NMJ morphology (57) and plasmalemmal excitability (13), isometric training may have improved these aberrations resulting in an increase in strength. Results also suggest the changes in strength and histopathology are associated with a hypertrophic response (Fig. 8). An increase in muscle fiber CSA (Fig. 8) may be associated with an improved regenerative capacity, as indicated by lower eMHC positive fibers (Fig. 9) and increased total satellite cell number (Fig. 10). Because isometric resistance training can induce a hypertrophic response in healthy adult skeletal muscle (1), and DMD patients suffer from skeletal muscle wasting, exercise in the form of isometric tetanic contractions may provide a stimulus that increases skeletal muscle fiber cross-sectional area and attenuates atrophy.

Dystrophin-deficient skeletal muscle is highlighted by consistent cycles of skeletal muscle degeneration and regeneration that eventually lead to deposition of fibrotic tissue and muscle weakness. Several preclinical therapies primarily target the molecular mechanisms associated with muscle degeneration and regeneration with the intention of improving or sustaining a homeostatic environment (25, 50). Here, we show that isometric resistance training in the dystrophin-deficient mouse can significantly reduce skeletal muscle fibrosis, reduce passive stiffness of isolated muscle and the hindlimb compartment, and increase the total satellite cell pool. However, isometric training did not affect central nucleation, which indicates not all histopathology can be improved. While resistance exercise has been shown to increase satellite cell numbers in healthy human muscle (34), our data provide the first evidence that isometric training can increase total satellite cell number in a dystrophin-deficient environment. Interestingly, the 19% and 31% increase in satellite cells at 30 and 90 days into a resistance training program in healthy muscle (34) is similar to the 30% increase that we measured in dystrophin-deficient mouse TA muscle. Combined with a lower number of eMHC-positive fibers (Fig. 9), these data suggest an improvement in the quality of skeletal muscle, which may contribute to the strength improvements. However, while isometric training resulted in greater total satellite cells, lowered fibrosis, and lower eMHC positive fibers, the changes need to be considered when evaluating the physiological relevance. Because dystrophin deficiency causes a multitude of severe histopathological changes that contribute to muscle wasting, stiffness, and weakness, even small changes induced by therapeutic or pharmaceutical interventions could be considered physiologically relevant and improve quality of life.

The current study investigated the effects of isometric training on a hindlimb muscle that predominantly comprises Type IIb MHC (5). Although the adaptations are promising in the TA muscle of mice, isometric resistance training in Type Ia MHC-dominant skeletal muscle of human vastus lateralis muscle (62) resulted in a similar adaptive response (33). Therefore, isometric training could be applied to a collection of skeletal muscles in DMD patients with potential positive adaptations. Moreover, isometric training circumvents the potential injurious activity of eccentric contractions in dystrophin-deficient skeletal muscle, while still providing a stimulus that enables strength gains and hypertrophy (1).

In summary, results of our studies indicate that isometric strength training of dystrophin-deficient skeletal muscle immediately improves strength that is sustainable for up to 4 wk. The data also show that 4 wk of isometric training elicits changes in dystrophin-deficient skeletal muscle, including fiber hypertrophy, reduced fibrosis, and increased number of satellite cells. Collectively, the evidence suggests that noninjurious isometric strength training improves the dystrophin-deficient phenotype and supports translation to patients with BMD and DMD.

GRANTS

This work was supported by NIH Grant RO1-AR042423 (to J. M. Ervasti), RO1 AR049899 (to J. M. Ervasti), T32-AG029796 (to A. A. Lindsay), and T32-GM008244 (M. Verma).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

A.L., J.M.E., and D.A.L conceived and designed research; A.L., A.A.L., and M.V. performed experiments and analyzed data; A.L., A.A.L., M.V., J.M.E., and D.A.L interpreted results of experiments; A.L., A.A.L., and M.V. prepared figures; A.L., J.M.E., and D.A.L. drafted manuscript; A.L., A.A.L., M.V., J.M.E., and D.A.L edited, revised and approved final version of manuscript.