Abstract
Duchenne Muscular Dystrophy (DMD) is caused by mutations in the gene coding for dystrophin and leads to muscle degeneration, wheelchair dependence and death by cardiac or respiratory failure. Physical exercise has been proposed as a palliative therapy for DMD to maintain muscle strength and prevent contractures for as long as possible. However, its practice remains controversial because the benefits of training may be counteracted by muscle overuse and damage.
The effects of physical exercise have been investigated in muscles of dystrophin-deficient mdx mice and in patients with DMD. However, a lack of uniformity among protocols limits comparability between studies and translatability of results from animals to humans. In the present review, we summarize and discuss published protocols used to investigate the effects of physical exercise on mdx mice and DMD patients, with the objectives of improving comparability between studies and identifying future research directions.
ABBREVIATIONS
Acetyl-CoA carboxylase
β-Hydroxy acyl-CoA dehydrogenase
Diaphragm
Duchenne muscular dystrophy
Extensor digitorum longus
Extracellular signal-regulated kinase 1/2
Gastrocnemius muscle
Hypoxia-inducible factor-1
c-Jun N-terminal kinase
Mitogen-activated protein kinase
Myosin heavy chain 2a
oxidase Nicotinamide adenine dinucleotide phosphate-oxidase
Peroxisome proliferator-activated receptor gamma coactivator 1α
Peroxisome proliferator-activated receptor gamma
Quadriceps muscle
Reactive oxygen species
L-Sorbose 1-dehydrogenase
Sirtuin 1
INTRODUCTION
Duchenne muscular dystrophy (DMD) is an X-linked muscular disease caused by mutations in the DMD gene, which codes for dystrophin, a cytoskeletal scaffolding protein important in signalling and muscle stability. An absence of dystrophin results in muscle degeneration and death by cardiac or respiratory failure. Symptoms usually appear in boys at 2–5 years of age, manifesting as difficulty standing unaided. As muscle wasting progresses, patients experience increasing difficulty in performing daily activities, become wheelchair-dependent between 11 and 13 years of age [1], and die before age 30 [2].
Therapeutic approaches for DMD include reducing inflammatory symptoms with glucocorticoids [3], correcting scoliosis by surgical intervention [4] and aiding respiratory function using mechanical ventilation [5]. Recently, restoration of dystrophin expression has been achieved by ribosomal readthrough of premature stop codons [6] and exon-skipping therapy [7].
Regular physical exercise stimulates muscle protein synthesis and mitochondrial biogenesis [8]. Exercise has therefore been proposed as treatment for DMD, to maintain muscle strength and prevent contractures [9, 10]; however, this recommendation has not been unanimously accepted because exercise might damage dystrophic muscles [11]. The five mechanisms rendering dystrophin-deficient muscles vulnerable to exercise (reviewed elsewhere [12]) are the weakening of the sarcolemma, increased calcium influx and oxidative stress, recurrent muscle ischemia and aberrant signalling to surrounding tissues such as nerves or cells of the immune system. A mechanistic basis for exercise intolerance [13] and recommendations for the management of DMD [9, 14] have also been reviewed. The lack of uniformity between protocols for exercise of dystrophin-deficient muscles, however, has been pointed out [15], but not reviewed.
Here, we summarize and discuss studies addressing physical training in the context of DMD. We focus on articles describing the effects of exercise in dystrophin-deficient mdx mice and in patients with DMD (Fig. 1A).
EXPERIMENTAL PROCEDURES FOR THE STUDY OF EXERCISE IN MDX MICE
A murine model for DMD: The mdx mouse
The dystrophin-deficient mdx mouse is the most common animal model for DMD. This mutant bears a spontaneous nonsense mutation in exon 23 of the dystrophin gene [16]. Its phenotype is mild compared to the symptoms of DMD in patients. This difference in disease severity between mice and humans arises from differences in size, mechanical loading and lifespan [17, 18]. First, the size difference between mice and humans is 2000–3000 folds. According to the square/cube rule, the mechanical stress experienced by an organism increases with the cube of the linear size; therefore, the difference in mechanical stress experienced by mice and humans is not linear but exponential [18]. Second, humans have a bipedal posture, meaning that the body weight is distributed between the lower limbs and the backbone [17], rather than across four limbs. Third, the difference in lifespan means that humans endure more degeneration–regeneration cycles than mice, resulting in extended muscle deterioration [17].
Compared to DMD patients, mdx mice recover from the progressive muscle wasting and show much less accumulation of connective and adipose tissue. The necrotic process persists throughout their life, but the regenerative capacity does not decline until an advanced age (>65 weeks) [19, 20]. These differences must be considered in investigations of physical exercise on dystrophic muscle across species.
Exercise studies in mdx mice
A literature search in PubMed, using the keywords “mdx mice” and “exercise” was performed on 25 May 2015. A total of 175 articles were examined, of which 57 investigated the effects of physical exercise, and were selected to form the basis of this part of the review. Of these studies, 37 reported only the negative effects of physical exercise, 15 only beneficial effects, and 5 both negative and positive effects. Studies were classified according to experimental protocol and are listed in Table 1.
The purpose of exercise studies in mdx mice
Physical training of mdx mice served three purposes: assessing the physical capacities of the mice; investigating the effects of training on dystrophic muscles; worsening the phenotype before assessing the effects of a drug (Fig. 1B). Depending on the study goals, researchers used acute exercise protocols to reveal immediate effects (Table 1A), or chronic protocols to study long term effects (Table 1B). The mildest methods used included swimming, voluntary wheel running, rotarod or low-speed treadmill (<9 m/min). The hardest ones employed high-speed treadmill (>12 m/min) or downhill running. Further variables included the age of the mice and training duration. Voluntary and brief (<30 min per session) exercise of young mice (<8 weeks old) was defined as low intensity training [30, 59], whereas exercise of older mice, or under intensive or prolonged conditions (>30 min), was defined as high intensity training [63].
Assessment of the physical capacities of mdx mice
Measuring the running capabilities of mdx mice using voluntary wheel or downhill running is a simple way to assess their physical abilities. All studies using voluntary wheel running followed the same protocol, namely measuring the total running distance. High inter-individual variability was reported: 4-week-old mdx mice ran 0.5 [38] to 9 km [36] per day; 6-week-old mice ran 2 km/day [36], while 10-week-old mice ran 0.03±0.005 to 4.48±0.96 km/day [49]. Performance of mdx mice peaked at 8 weeks of age (5.8 [39] to 9 km [36] per day) and decreased to 2.6 km/day at 10 weeks [39] or to 5 km/day at 14 weeks of age [36]. In consequence, a large number of animals should be used when performing experiments with mdx mice. The downhill exercise study [76] adapted the 6 minute walking test, used for patients with DMD, to allow comparison of performance between 10-week-old wild type and mdx mice. Results show that wild type mice run an average of 500 m in 6 minutes, but mdx mice run only 300 m.
Measuring ex vivo the properties of specific muscles is another way to assess physical capacity in mice. However, muscle type and choice of protocol varied too much between studies to allow comparison (Table 1). The parameters assessed after voluntary wheel running included tetanic stress and stiffness of the extensor digitorum longus [36], grip strength and specific tetanic force of the soleus [38], maximal isometric torque and fatigue resistance of the plantar flexor [39] or specific and absolute maximal force of the tibialis anterior muscle [41]. Overall experiments reported that phenotype of hind limb and diaphragm of mdx mice improved when training began before 7 weeks old [45], but worsened if exercise began later period [47]. However, worsening of heart phenotype was observed when training started at 4 weeks old mdx mice [41] In spite of these general considerations, important differences in outcome can be observed between the studies, corroborating the need for a common protocol for measurements in individual muscles after exercise.
Different muscles in mdx mice, such as the hindlimb muscles or the diaphragm, are not equally affected by an absence of dystrophin; for example, hindlimb muscles show more necrotic events than the diaphragm, but less fibrosis, following regeneration [78].Consequently, studies should investigate different muscles simultaneously. However, most studies focused on the effects of exercise on hindlimbs; others investigated the diaphragm [43, 77] or the heart [38, 75]. The hindlimbs and the diaphragm of 4-week-old mdx mice tolerate the effects of voluntary running well (Table 1B), whereas necrosis and fibrosis occur after 10 weeks of age. Conversely, cardiac complications appear after voluntary running in 4-week-old mdx mice, with increased cardiac mass [39] and impaired function [41, 43]. Effects of swimming on cardiac function have only been investigated preliminarily. Our own results have shown that 30 minutes daily swimming from 8 to 16 weeks of age had no influence on cardiac weight (Hyzewicz, unpublished data), but further investigations are necessary.
The studies cited above were performed using male mdx mice. Studies using female mdx mice at 24–28 weeks of age did not reveal signs of cardiomyopathy after voluntary running [52]. Interestingly, female mdx mice were more susceptible than males to develop cardiac problems [79]. Further studies must be conducted to determine whether voluntary running can protect female hearts from complications.
Investigations of physical exercise on mdx mouse muscle
Acute exercise (Table 1A) leads to membrane leakiness, even under mild conditions such as swimming using 4-week-old mice [21]. Voluntary running in 10-week-old mdx mice also causes apoptotic events in the tibialis anterior muscle [23]. Necrosis, thiol oxidation and increased expression of interleukin (IL)-6 mRNA have been reported in quadriceps muscle of 12-week-old mice after 30 minutes of treadmill running at 12 m/min [24]. These results show that even single bouts of exercise can cause muscle damage in mdx mice.
In wild type mice, adaptation to chronic exercise leads to large changes in signal transduction mechanisms [8], including subfamilies of the mitogen-activated protein kinase (MAPK) signalling pathways, namely: extracellular signal-regulated kinase (ERK) 1/2, c-Jun N-terminal kinase (JNK), and p38 MAPK [80]. These signalling pathways are activated by reactive oxygen species and lead to activation of genes involved in mitochondrial adaptation, such as PGC1-α, and muscle differentiation [81]. In mdx skeletal muscle, production of such species is abnormally elevated, owing to either mitochondrial Ca2 + overload [82] or over-activation of membrane-bound NADPH oxidase 2 [83]. Chronic exercise studies using 6–8-week-old mdx mice that performed 10 weeks of downhill running on a 7° slope at 23 m/min for 1 hour showed that proteins downstream of MAPK were over-phosphorylated [74, 75]. This protocol also resulted in infiltration of immune cells, fibrosis, and deposits of adipose tissues in skeletal muscle and heart. Chronic treadmill running at 12 m/min for 30 min caused downregulation of Sirt1, PGC1-α, PPARγ and myogenin in 4-week-old animals [62].
In contrast, low intensity training is beneficial. Expression of mitochondrial [39, 45] and muscle differentiation [41] genes was increased after voluntary wheel running in 4–8-week-old animals (Table 1B). Low intensity swimming and running in young mdx mice also stimulated a switch from fast glycolytic muscle (type IIb) to oxidative (type IIa) and slow (type I) muscle [34, 37]. Despite this switch, protein expression in slow and fast skeletal muscle increased after low intensity swimming in 4-week-old mice [30]. Comparison of wild type and mdx muscle following low intensity swimming also pointed to higher protein expression levels in mdx fibres [30]. This observation suggests that dystrophic muscles could benefit from smaller quantities of training than wild type muscles. The fact that hypoxia is more severe in muscles of mdx mice than in wild types could explain this phenomenon, since stimulation of HIF-1 initiates adaptation to training via the MAPK signalling pathway [34].
Exercise as a means to worsen the dystrophic phenotype
The mild disease phenotype in mdx mice causes a bias when assessing effectiveness of potential drugs for DMD therapy. To worsen the mdx phenotype, researchers increase the mechanical stress using voluntary wheel running [50], treadmill [24, 66] or downhill running [70, 77] (Table 1). We compared these protocols to determine which types of exercise are likely to make mdx muscles become more like those in patients with DMD.
Voluntary running in 10–12-week-old animals caused a suitable worsening of the mdx phenotype, showing fibrosis, kyphosis [50] and necrosis of quadriceps muscle [48]. Similar results were obtained with treadmill running in 4-week-old mice for 4 weeks at 12 m/min, leading to fibrosis [61], gastrocnemius degeneration [59], decreased forelimb strength [24, 55] and elevated levels of reactive oxygen species in plasma [56]. Twelve weeks of training also caused downregulation of Sirt1, PGC1-α and PPARγ mRNA expression [62]. However, the acute damaging effects of exercise tended to disappear 96 hours after training, as shown by decreased levels of mRNA coding for pro-inflammatory IL-1β and IL-6 [63].
Downhill running was reported to result in increased muscle damage and decreased grip strength [70, 71], but no information about fibrosis or inflammation was available in these reports. The most complete studies on downhill running showed evidence of fibrosis, adipose tissue and infiltration of immune cells in the hearts of 18-week-old mice after 10 weeks of running on a 7° slope at 23 m/min [75], or decreased muscle strength, increased myoglobinuria and inflammation in 12-week-old mice after 2 weeks of training on a 15° slope at 15 m/min [76].
Based on these observations, we conclude that to worsen the phenotype of mdx mice, a minimum of 4 weeks of voluntary exercise from 10 weeks of age, or at least 4 weeks of treadmill running at 12 m/min from 4 weeks of age, is required. Furthermore, in order to avoid the bias due to acute exercise effects, the ability of drugs to prevent exercise-induced damage should be ideally measured with a proper lag time (around 2-3 days) after last exercise bout [54].
TREAT-NMD protocols for exercise in mdx mice
TREAT-NMD is a network for research on neuromuscular diseases that proposes standard operating procedures (SOPs) for experiments with the aim of improving comparability between studies [84]. Two exercise protocols were proposed for mdx mice: one to worsen the phenotype [85] and the other to assess the progression of the dystrophic state [86]. Both are based on previous publications on wheel [21, 36] or treadmill [53, 56] running.
The first protocol advised that 3–4-week-old mice perform voluntary wheel running 1–7 days/week, or treadmill exercise at 12 m/min for 30 minutes twice per week. Based on our review (section 2.3.3), treadmill exercise in mice aged 3–4 weeks is suitable for worsening the mdx phenotype, but voluntary wheel running requires the mice to be at least 10 weeks old; in younger mice, the benefits of exercise counteract the aggravation of the dystrophic phenotype [36–38].
The second protocol also suggested voluntary wheel running 1–7 days/week, or treadmill exercise at 9 or 12 m/min for 30 minutes twice a week, in young mice. Authors recommended avoiding downhill running since mdx mice barely tolerate this exercise. They also pointed out that all mice should perform the same amount of exercise, especially during voluntary wheel running. We agree with these recommendations.
SOPs are an important tool for harmonizing experiments between laboratories. We suggest adding protocols for swimming training based on previous publications [21, 30–35].
EXPERIMENTAL PROCEDURES FOR EXERCISE IN PATIENTS WITH DMD
Literature search for exercise studies in patients with DMD
We performed a literature search in PubMed, using the keywords “DMD” and “exercise”, which was completed on 25 May 2015. A total of 167 articles were examined. Twenty-five of them reported effects of exercise in patients with DMD and described the protocol or results; these formed the basis of this part of the review. They were classified according to the type of muscle performing the exercise and are listed in Table 2.
The purpose of exercise studies in DMD patients
Physical training was mainly used to assess therapeutic methods for improving dystrophic muscle capacity in wheelchair-dependent patients with DMD, measuring the effects of exercise on respiratory or masticatory muscles. Several studies investigated the effects of acute exercise (Table 2A). Studies assessing therapeutic exercises at early stages of the disease (before 10 years of age) were rare, and mainly used chronic training of upper and lower limbs (Table 2B).
Therapeutic training in patients with DMD before wheelchair dependence
Only three studies focused on improving ambulation in young patients by arm and leg training [95, 98] (Table 1B). Interestingly, documentation for parents of DMD patients recommends physical exercise during the early stages of the disease [10, 111] based on observations in mdx mice [111]. Moderate exercise is recommended, without pushing the child, and stopping before the threshold of exhaustion, switching to cycling or swimming when difficulties become apparent.
Bicycle [98] or ergometer [95, 97] training in young DMD patients confirmed the benefit of long-term physical exercise from early stages of the disease. However, two studies showed that running or step exercise damaged the tibialis anterior muscle and caused myoglobinuria immediately after training in patients aged 6–10 years [93, 94]. Theoretically, damage induced by short-term exercise does not prevent long-term improvement in muscle status. But the benefits of training have been demonstrated on bicycle and ergometer, whereas studies reporting short-term negative effects have involved running or step exercises without equipment. The long-term benefits of non-assisted leg training remain to be demonstrated.
Swimming is often recommended for DMD patients [98, 112], but only one study has investigated its effects, and found that myoglobin and creatine kinase levels were elevated after training [88].
Therapeutic training in patients with DMD after wheelchair dependence
The first demonstration that muscle training could improve the daily life of late-stage DMD patients was in 1952 [98], but used no control, and was therefore hard to interpret [113]. Subsequent studies demonstrated that appropriate training could improve the capacity of masticatory and respiratory muscles in wheelchair-dependent patients with DMD.
For masticatory muscles, two studies showed that jaw and tongue training for 24 weeks, accompanied by massage of the masseter muscle, could improve jaw performance and ease of eating [99, 100] (Table 2B).
For respiratory muscles, two approaches were followed: non-assisted or assisted training. Non-assisted training involved resistive inspiratory muscle training [103, 104] or yoga [109]. Assisted training involved the use of special apparatus [101, 108], video games [102], breathing through a valve [105, 110] or resistance to a load [107] (Table 2B).
All studies except one [101] reported an improvement in patients’ respiratory capacity. Non-assisted training improved maximal resistance, duration of ventilation [103], inspiratory airway pressure [104], forced expiratory volume in 1 s [109], and vital capacity [104, 109]. Assisted training improved maximal voluntary respiration, maximal achieved respiration [102], maximal sniff assessed oesophageal and transdiaphragmatic pressure [105], static inspiratory/expiratory pressures [107] and inspiratory mouth pressure, as well as 12 s maximal voluntary ventilation [108, 110], duration of progressive isocapnic hyperventilation manoeuvre [102] and respiratory muscle endurance [105, 106].
Non-assisted and assisted respiratory training data are not comparable because different parameters were measured. Only one study compared the effects of non-assisted and load-assisted training and concluded that non-assisted training had no effect [107].
In conclusion, training of respiratory muscles successfully delays the need for mechanical ventilation [104], but there is a lack of studies comparing non-assisted and assisted respiratory training to determine whether training equipment is useful for therapeutic purposes or could be replaced by non-assisted inspiratory training.
Investigating acute exercise in patients with DMD
An early ergometer study showed that DMD patients have limited adaptation to exercise owing to reduced cardiorespiratory capacity, weaker leg strength and limited use of peripheral oxygen [87] (Table 2A). Arm muscle training studies revealed that the intracellular pH at the end of the exercise was higher in DMD muscle fibres than in healthy patients and that inorganic phosphate and pH recovery rates were lower [89, 90]. These differences were explained in part by the fact that the vasoconstrictor response of dystrophic muscles is not blunted in response to exercise [92]. In contrast, DMD patients feel less fatigue and have less muscle injury at the end of most types of exercise [89, 91].
DIFFERENCES IN RESEARCH APPROACHES BETWEEN MICE AND HUMANS
Research in animal models is commonly the first step before clinical trials in patients. However, because the mdx mouse was discovered only in 1984 [16], much after the first report in humans in 1868 [114], experiments investigating the effect of physical exercise on dystrophic muscles began in DMD patients 40 years before the first studies in mdx mice. Here, we have reviewed 80 articles and found none reporting results from both the murine model and patients. The consequence is a large number of differences between results of studies in mice and humans.
Research focusing on respiratory function in patients with DMD
The majority of investigations in patients with DMD aimed to improve respiratory function alone, and recruited mainly wheelchair-dependent patients [101–110]. In comparison, only three studies document results of investigations of respiratory function in exercised mdx mice [43, 77]. However, limb and diaphragm muscles were investigated together in mdx mice, but separately in patients with DMD.
Effect of exercise on cardiac function in patients with DMD
Since the development of mechanical ventilation, cardiac failure has become the primary cause of death of DMD patients. Experiments in exercised mdx mice demonstrated a vulnerability of cardiac muscle with running, even with low intensity training [39, 43]. However, no study in patients with DMD has ever investigated the impact of exercise on the heart. The limits of exercise for DMD patients should be adjusted by taking into account the limits of the cardiac muscle, especially because experiments in mice have shown that voluntary training can also damage the heart. This aspect is important because it suggests that children with DMD might exercise over their limit, without considering the damage occurring in their heart.
Swimming is recommended for patients, but without evidences
Swimming exercise appears intuitively to be beneficial for DMD patients, because water supports a large part of the patient’s weight and thus reduces mechanical stress. However, benefits of water training have never been assessed in patients and scarcely investigated in mice. Vulnerability of the mdx heart has been observed with running [39, 43], but results from preliminary studies of swimming (Hyzewicz, unpublished data) suggest that this exercise might spare the cardiac muscle. If further studies confirm the harmlessness of swimming, then it should become a research priority in patients with DMD.
CONCLUSIONS
We have reviewed here the present state of research into the effects of physical exercise on dystrophic muscles, and suggested further investigations to establish evidence-based recommendations regarding optimal training modes. The main conclusions we have drawn are summarized in Fig. 2 and outlined below.
Studies in mdx mice have demonstrated that voluntary running exercise in 4-week-old animals improves hindlimb and diaphragm capacity but is harmful for the heart. The effects of swimming on cardiac tissue have not yet been studied. Forced treadmill running of at least 4-week-old mice at 12 m/min for 4 weeks renders the mdx phenotype closer to that of DMD patients.
Studies in human have shown that respiratory and masticatory muscle training successfully improves functional capacity in patients aged 12 years or older. However, further comparisons between machine-assisted respiratory training and non-assisted training are necessary. Bicycle training can also delay the impairment of motor function in young patients, although the effects of exercise on cardiac function have not yet been investigated. There is also a lack of studies investigating the effect of running and swimming in DMD.
In order to fill these knowledge gaps, we suggest the following approaches for future research regarding the effects of physical exercise on mdx mice and DMD patients: As DMD causes degeneration of respiratory, cardiac and limb muscles, future studies should assess the effects of exercise on these three types of muscle simultaneously. Running has a negative effect on the heart in mdx mice. It is crucial to determine whether a similar effect occurs in the hearts of patients with DMD after running. Effects of exercise on young DMD patients (4–12 years old) should be investigated. Even though swimming is recommended, its cardiac consequences have not been studied in mdx mice or patients. This should be performed using appropriate technology, such as MRI and biomarker measurement. Studies should compare non-assisted respiratory training with machine-assisted training.
CONFLICT OF INTEREST
The authors have no conflict of interest to report.
Footnotes
ACKNOWLEDGMENTS
We thank the Department of Molecular Therapy, National Center of Neurology and Psychiatry, for support and useful discussions. We are also grateful to the Library of the National Center of Neurology and Psychiatry for help in gathering bibliographic material. This work was supported by an Intramural Research Grant (25-5) for Neurological and Psychiatric Disorders of the National Center of Neurology and Psychiatry.
