Background
Exercise has been recognized as among the most effective interventions including nutrition at promoting wellbeing and alleviating chronic disease [
1‐
4]. As reported in a review of meta-analyses, exercise, in particular, is as effective as drug treatment for the prevention of heart disease and diabetes, the treatment of heart failure, and the rehabilitation from stroke [
5]. Resistance exercise, in particular, is being considered and/or encouraged to address muscle atrophy associated with such conditions such as cachexia, aging, and type 2 diabetes [
1,
2]. Specifically, resistance exercise promotes adaptation in terms of greater muscular strength/power and muscle hypertrophy [
2]. Such hypertrophy can come in the form of increased muscle fiber size and, in some cases, increased muscle fiber number [
2,
6‐
11].
Daily activity and exercise (including resistance exercise) typically consist of movements requiring a sequence of different types of contractions, isometric, lengthening, and shortening contractions, i.e. SSCs [
12]. To investigate the response to contractions with high control and precision, dynamometer based animal studies have been utilized [
13]. The majority of these studies have investigated isometric, lengthening, or shortening contractions in isolation in order to characterize the impact of each contraction mode alone [
13,
14]. An exception to this has been the study of SSCs in their entirety in a rat model [
13,
15‐
17]. However, this animal model has been limited in regard to diversity of genetically defined strains available.
Mice are the most commonly used animal model for biomedical research. This is largely because they have biological similarities to humans, have been characterized as experimental models for nearly a century, and are available in thousands of genetically defined strains to aid in the investigation of molecular mechanisms (e.g. more than 7,000 genetically defined strains are maintained at Jackson Laboratories) [
18]. Given these advantages, establishment of a SSC training protocol for mice would provide a valuable research tool for determining the molecular mechanisms of muscle adaptation.
The aim of the present study was to develop a dynamometer-based SSC training protocol to induce adaptation of muscle mass and performance gains for plantarflexor muscles of mice. We hypothesized that the plantarflexor muscles of mice would adapt to one month of training with sessions of 80 SSCs, a protocol based on that which has been demonstrated previously to be effective for hindlimb muscles of rats [
15,
16,
19]. Consistent with the hypothesis, the 80 SSC protocol induced an increase in plantarflexion performance and PLT muscle mass. Additionally, unanticipated outcomes emerged which raised the value of the training model further. Following SSC training, the performance increase and agonist muscle mass gain was accompanied by an increase in muscle fiber number in transverse muscle sections while the antagonist muscle, the TA muscle, atrophied. These outcomes were not observed in previous research regarding the rat model [
13,
15‐
17]. Furthermore, we characterized fiber type transitions and utilized the Myogenesis & Myopathy RT
2 Profiler™ PCR Array to establish gene expression profiles for these plantarflexion SSC training induced alterations. The outcomes of this research establish this SSC training mouse model as an exceptional research model and provides insights into muscle fiber type transitions, fiber number modulation, and muscle imbalance.
Discussion
In the present study, we developed a novel adaptive SSC protocol for plantarflexor PLT muscles of mice, thereby effectively addressing the aim and hypothesis of the study in establishing in mice a practical and influential model for the study of SSC training. In developing this model, two major findings were observed. First, a shift in fiber type distribution accompanied by increased total fiber number (rather than enlarged fiber size) in cross-section as responsible for the increase in PLT muscle mass. This result highlights the present study as providing a unique model of SSC-induced adaptation given that our previous research regarding SSC training in rats demonstrated the more typical response of muscle fiber hypertrophy concomitant with increased muscle mass [
19]. The second major finding of the study was the 30% decrease in the antagonist TA muscle mass with selective atrophy of type IIb and IIx fibers. Such an induced muscle imbalance indicates the profound detrimental effects of limiting resistance-type training to specific muscles without regard to surrounding muscles. Overall, the results have impact on understanding SSC-induced adaptation in regards to implementation in the mouse model, fiber type transitions, fiber number modulation, and the induction of muscle imbalance.
In developing this SSC training for the induction of increased plantarflexion torque and muscle mass, the serendipitous consequence of a shift to a more oxidative metabolism was evident by analysis of fatigue recovery, gene expression, and fiber type data. Increased maximum isometric force recovery has been observed previously for muscles of rats exposed to contraction-induced training [
16,
25,
26]. In this study, we demonstrate that not only did fatigue recovery improve in terms of maximum isometric force but also capacity to maintain force generation during the entirety of a contraction shortly following a SSC session improved with training. Interestingly, resistance to fatigue during the first set of SSCs was largely unaltered with training. A comparable result of a shift away from type IIb fibers and differential adaptation for enhanced fatigue recovery rather than fatigue resistance has been observed for SSC-trained rats and endurance trained human subjects [
16,
27]. This supports the concept that recovery from fatigue, rather than the susceptibility to fatigue, presents as more sensitive to certain training-induced alterations in fiber type [
16,
27]. The training-induced alteration in metabolism for PLT muscles was apparent at the transcriptional level. A shift away from glycolysis was evident by the downregulation of genes for Slc2a4 (Glut4; glucose transporter), Hk2 (prepares glucose for glycolysis), and AMPK (a regulator of
Glut4 and
Hk2 expression) [
28‐
30]. The upregulation of
Myog may underlie this metabolic alteration.
Myog overexpression in mice induces a shift from glycolytic to oxidative metabolism [
31,
32]. At the fiber type level, the metabolic alteration of the present study is further supported with the shift from type IIb to IIx fibers and the upregulation of
Myh1, coding for MHC IIx.
The observation of increased fiber number per PLT muscle cross-section was a novel finding in an animal model of SSC training. For the SSC training model in rats, increased muscle fiber size can account for the induced muscle hypertrophy [
19]. With training, mean fiber size for rats of that study increased by 14% (2867 ± 107 vs. 2521 ± 84 μm, trained vs non-trained,
P = 0.01) while fiber number did not (16,227 ± 529 vs. 15,966 ± 646 fibers, trained vs. non-trained,
P = 0.65) [
19]. This differential outcome may be due to differences in the mechanical strain of the SSC training protocols in terms of the extent muscles stretched and shortened each contraction. For the protocol utilized previously, muscles were stretched and shortened through an ankle rotation of 50° [
15‐
17,
19]. Because of limitations in the muscle lever system of the present study, the ankle rotated through a decreased excursion (20° of ankle rotation). Consequently, muscles were exposed to SSCs which were more moderate when compared with those of previous studies. The finding that only PLT muscles increased in muscle mass while the other plantarflexor muscle masses were unaltered is consistent with the moderate nature of the training. Based on the glycolytic fiber type characteristics of the PLT muscle relative to the other plantarflexor muscles (soleus and gastrocnemius), the PLT muscle is recruited less often during daily activity and, consequently, activation during SSC training would be especially novel and sensitive to adaptation [
33].
Evidence for contraction-induced increases in muscle fiber number has been observed in several studies regarding voluntary resistance training in animal models and human subjects [
6‐
11]. The present study demonstrates that such a response can occur after high-intensity SSC training consisting of maximal activation of muscles. This increased fiber number potentially could have been mediated by the downregulation of
Mstn. Postnatal muscle-specific inactivation of
Mstn results in increased cross-sectional fiber number accounting for ~30% of the increase in whole muscle section area [
34]. Furthermore, a missense mutation in
Mstn causes increased muscle fiber number without fiber hypertrophy [
35]. Consistent with the downregulation of
Mstn in the present study,
Igf2 was upregulated. Myostatin is a negative regulator of
Igf2 expression [
36]. Despite the common belief that
Igf2 has only a marginal role during postnatal muscle growth, recent studies have indicated otherwise [
36,
37]. Pigs with a SNP within intron 3 of the paternal
Igf2 allele have increased postnatal
Igf2 expression and greater muscle mass [
37,
38]. High postnatal levels of
Igf2 also accompany the increased muscle mass of
Mstn null mice [
36]. In addition to gene expression consistent with increased fiber number, the gene expression analysis was indicative of enhanced muscle fiber survival as evident by the downregulation of genes involved in protein degradation. Several genes associated with Fbxo32 (Atrogin-1) induced protein degradation (via the ubiquitin-proteasome pathway) were downregulated -
Fbxo32,
Foxo1,
Myod1, and
Hdac5 [
39,
40].
Capn3, the gene encoding the calpain 3 protease involved in an alternative proteolytic pathway, was also downregulated [
40].
Whether the fiber number increase of the present study represents a completely different form of adaptation than that observed in the SSC trained rat model or characterizes an earlier stage of adaptation remains unknown. For instance, SSC training may induce first a fiber number increase, as observed in the present study, and then fusion of fibers (e.g. fusion of myofibers or branched fibers) with a return to control value fiber numbers, as observed in previous research regarding the rat SSC model [
15‐
17,
19,
41]. This order of adaptation has been proposed as a possibility by others during regeneration following muscle injury [
41,
42]. The present study is distinctive in suggesting that such a sequence of adaptation may also occur for muscles exposed to non-injurious training.
The concept of muscle imbalance has been characterized as hypertrophy or high recruitment of one muscle or muscle group as linked to an atrophy response in opposing muscles [
43,
44]. Muscle imbalance has been implicated in underlying such conditions as lumbopelvic instability, lower back pain, susceptibility to knee injury, and shoulder pathology [
43,
45,
46]. Based on observations of patients, phasic muscles (rather than tonic postural muscles) have been considered susceptible to possible atrophy and weakness from muscle imbalance [
44]. Our findings regarding the phasic non-weight bearing TA muscles provide direct evidence for this concept. Furthermore, the present study establishes a gene expression profile and fiber type phenotype for muscle imbalance. The severity of the TA muscle atrophy was apparent at the gene expression level as evident by the downregulation for genes encoding proteins associated with titin and dystroglycan, proteins essential for force transmission during contractions [
47]. Such downregulation suggested that the muscle was compromised in terms of quality (decreased expression of genes for cytoskeletal/sarcomeric proteins per unit of muscle mass) as well as quantity (i.e. overall muscle mass decrement). The depreciated state of the TA muscle was also evident by the upregulation of
Cryab, an opposite change of regulation to that of the PLT muscle.
Cryab encodes αB chain crystallin (i.e. heat shock protein B5), a molecular chaperone that binds misfolded proteins to prevent the accumulation of protein aggregates and is upregulated in stress/pathological conditions [
48]. Gene expression analysis indicates that such a compromised condition in TA muscles was accompanied by the upregulation of gene expression promoting muscle atrophy (e.g.
Casp3) and downregulation of genes which repress such atrophy (e.g.
Pax3 and
Rps6kb1) [
49,
50]. The downregulation of
Rps6kb1 was especially noteworthy given that
Rps6kb1 encodes p70 Ribosomal protein S6 kinase 1, a protein intimately linked to protein synthesis via regulation of major components of translation - Ribosomal S6 protein, eukaryotic translation initiation factor 4B, and eukaryotic elongation factor 2 [
50,
51]. This indicated that the gene expression profile was consistent with decreased protein synthesis in addition to enhanced atrophy.
A neuromuscular imbalance may underlie the atrophy response for the TA muscle [
44]. One instance of an onset of such an imbalance is when antagonist muscle activity is reduced secondary to training-induced tension of the agonist [
52,
53]. This response contributes to the increased force production in trained agonist muscles. However, if antagonistic muscle inhibition is pervasive enough to be prevalent (even during daily activity between training sessions) such a response could be detrimental to antagonistic muscles. The notion of decreased antagonist activity is supported by the gene expression of
Myog and
Musk. Muscle-specific receptor tyrosine kinase (Musk) signals clustering of acetylcholine receptors and muscle activity suppresses expression outside of the neuromuscular junction [
54]. When activity is abolished by denervation, myogenin upregulates
Musk in an embryonic pre-innervation pattern of gene expression [
54]. In the present study,
Myog and
Musk were highly upregulated (9- and 2-fold increases, respectively), a finding consistent with diminished TA muscle activity. Preliminary research regarding dorsiflexion SSC training also supports the notion that decreased muscle activity may underlie the antagonistic atrophy of the present study.
. Following dorsiflexion SSC training of young (3 month old) Sprague Dawley rats (
N = 4), TA muscles, which are agonists in this case, increase normalized muscle mass by the expected 20% (trained vs non-trained; 17.3 ± 0.6 vs 14.4 ± 0.7 mg/mm;
P = 0.03). Interestingly, for the same animals, such training had no effect on antagonist muscle masses in this situation, the tonic weight-bearing plantarflexor muscles (trained vs non-trained; PLT – 8.3 ± 0.2 vs 8.3 ± 0.2 mg/mm; gastrocnemius - 39.3 ± 1.2 vs 39.6 ± 1.3 mg/mm; and soleus - 3.3 ± 0.2 vs 3.4 ± 0.1). When these results are considered along with the present study, phasic non-weight bearing phasic TA muscle may be largely inhibited without significant impact on daily activity as in the present study while plantarflexor muscles may be protected from severe inhibition because of their weight-bearing activity in the preliminary study.
The shift away from MHC IIb observed for TA muscles in the present study is consistent with several reports regarding the MHC distribution alteration observed for fast-twitch muscles (e.g. TA and laryngeal muscles) following immobilization, nerve transection, and toxin-induced muscle paralysis [
55‐
58]. Although the distribution alteration away from MHC IIb is in common with that of the PLT muscles, the downregulation of
Ppargc1a suggests a diminished metabolic capacity for TA muscles after plantarflexion SSC training.
Ppargc1a encodes peroxisome proliferator-activated receptor gamma coactivator-1α (PGC-1α), a master regulator of mitochondrial biogenesis and oxidative enzymes [
59]. The downregulation of
Ppargc1a coupled with a shift from MHC IIb fibers to MHC IIx and MHC IIa fibers suggests metabolic changes without the typical MHC alteration. A disconnect between MHC composition and metabolic capacity has already been shown to be possible previously for transgenic mice which overexpress myogenin [
31]. The transition away from type IIb fibers in the present study was accompanied by atrophy of IIb and IIx fibers while type IIa fiber size was preserved. This tendency for a greater propensity for type IIb/x to atrophy has also been observed following immobilization of TA muscles, a finding consistent with specific catabolic conditions [
60,
61]. Therefore, the atrophy of TA muscles in the present study was a function of both a shift in fiber type distribution (from large IIb fibers to smaller IIx and IIa fibers) and atrophy within fiber type, specifically for type IIb and IIx fibers.
The plantarflexion SSC training induced 30% atrophy of the TA muscle is striking given that this occurred at a young age, an age with high adaptive capacity and resistance to maladaptation as evident in resistance training for human subjects [
62,
63]. Likewise in a previous report, when muscles of young rats were directly exposed to a severe SSC protocol – a protocol which induced performance deficits at maturity and old age – the response was muscle hypertrophy and enhanced performance [
16]. However, these past studies were limited in that only agonist muscles were evaluated. The present study demonstrates that antagonist muscles are more susceptible to such maladaptation following a distinct SSC protocol; and, therefore, supports the concept of investigating antagonist and surrounding muscles in addition to the agonist muscle when evaluating training regimes. Such investigation is especially warranted because muscle imbalance can be largely subclinical until severe joint mobility, maladaptation, or possibly injury manifests. General activity/training alone is not sufficient to prevent this maladaptation as evident by studies demonstrating that elite athletes incur muscle imbalance [
43,
46].
Overall, the outcomes of the present investigation effectively address the rational and aim of the study – to develop a dynamometer-based SSC training protocol to induce muscle mass and performance gains for the plantarflexor PLT muscle of mice. Regarding the major finding of increased muscle fiber number per PLT cross-section, a limitation was not being able to determine the precise mechanism (i.e. muscle fiber splitting vs
de novo fiber formation) with the available data. Despite this, the finding establishes that increased fiber number is a possible response to high-intensity SSC training and warrants further investigation. The other major finding of the study - plantarflexion SSC training induced muscle imbalance in the antagonist TA muscle of mice - provides the scientific community with a novel model to investigate muscle imbalance. In addition, the finding is consistent with the proposal that antagonist phasic muscles are susceptible to imbalance when agonist muscles are exposed to contractions in isolation [
43,
44]. Fortunately, such training-induced muscle imbalance is not inevitable. Research regarding volitional weight-lifting for rats demonstrates that squat-like training induces hypertrophy in the agonist plantarflexor muscles without atrophy of the antagonist TA muscle, a muscle activated for stabilization during squats [
64]. Resistance-type training requiring antagonist stabilization should be considered as a potential strategy to prevent the occurrence of muscle imbalance. The further development of concepts for exercise prescription are justified to ensure the prevention of such deleterious outcomes as those observed for the phasic antagonist muscle of the present study.