Muscle Fiber Splitting is a Physiological Response to Extreme Loading in Animals

Abstract

Skeletal muscle fiber branching and splitting is typically associated with damage and regeneration and is considered pathological when observed during loading-induced hypertrophy. We hypothesize that fiber splitting is a non-pathological component of extreme loading and hypertrophy, which is primarily supported by evidence in animals, and propose that the mechanisms and consequences of fiber splitting deserve further exploration.

Summary for TOC:

Skeletal muscle fiber splitting is proposed to be a physiological response to extreme loading and hypertrophy across species and is distinct from regeneration.

INTRODUCTION

It is generally accepted that skeletal muscle fiber hypertrophy in response to loading is a consequence of individual muscle fiber growth, and not through the creation of de novo muscle fibers. Under circumstances of extreme loading, however, the appearance of newly formed muscle fibers has been reported under a variety of conditions (1). The source of these new muscle fibers, usually quantified on histological cross-sections, may be attributed to new muscle fiber formation from satellite cells or may be an artifact of changes in muscle architecture. The longitudinal “branching”, “fragmenting”, or “splitting” of existing muscle fibers during robust overload may also contribute to the appearance of new muscle fibers, or perhaps even result in bona fide hyperplasia. Greater muscle fiber number during hypertrophy usually occurs concomitant with signs of regeneration, and is therefore considered pathological. Alternatively, we hypothesize that muscle fiber branching (which we define as the appearance of a smaller muscle fiber stemming from a larger one) or splitting (which we define as a roughly symmetrical division that may or may not run the entire length of the fiber) and regeneration may be distinct processes that occur during “extreme” loading. The precise triggers for fiber splitting are unknown, but for the purpose of this review, we classify extreme loading as a stimulus that is extraordinarily high volume (i.e. constant loading characteristic of surgical overload models in animals) or high intensity (i.e. uncommonly heavy and frequent loading associated with powerlifting or perhaps bodybuilding) that generally results in muscle hypertrophy of abnormal magnitude. Our position is supported by observations of increased muscle fiber number following robust hypertrophy in animals, as well as with synergist ablation overload in the absence of muscle stem cells (satellite cells).

EXTREME MODELS OF LOADING IN ANIMALS

Although not a mammalian model, the most dramatic example of extreme mechanical-mediated loading and subsequent skeletal muscle hypertrophy is wing-weighting to induce stretch-overload of accessory muscles in birds. Building on earlier work (2), Antonio et al. showed a 111% increase in latissimus dorsi muscle fiber area by 12 days and a 318% increase in muscle mass by 28 days using this model; muscle fiber number increased 82% on histological cross-sections by 28 days as well (3). Some studies that employed a weighted stretch approach reported increased muscle fiber number on cross-sections concomitant with apparent degeneration-regeneration and nascent fiber formation from satellite cells (4–6). These studies also provided evidence for the provocative notion that satellite cells can migrate from under the basal lamina into the extrafascicular space and initiate myogenesis in developed muscle. Alternatively, other studies reported longitudinal “branching”, “fragmentation”, or “splitting” of isolated muscle fibers following stretch-overload as an explanation for increased muscle fiber number (5, 7, 8). Splitting during stretch overload seemingly can occur at the level of the myofibril (9), which is consistent with what has been observed during developmental growth in mammals (10). Whether fiber splitting during hypertrophy can result in true hyperplasia, or the formation of new muscle fibers complete with innervation, is still unclear, but branching and splitting can certainly contribute to increased fiber number on cross-section. Wing-weighting stretch overload has provided valuable insight into hypertrophic processes, but the fiber branching and splitting associated with it is regarded as pathological since the model is generally not translatable to other species.

SYNERGIST ABLATION AS A MODEL OF EXTREME LOADING FOR STUDYING ROBUST SKELETAL MUSCLE HYPERTROPHY IN RODENTS

Due to its relative simplicity and effectiveness for inducing hypertrophy, the synergist ablation surgical overload technique has been employed for studying skeletal muscle adaptations in rodents for ~50 years (11, 12). This approach involves the removal of a primary-mover muscle so that a less prominent synergistic muscle must compensate during everyday ambulatory activity, thereby causing mechanically-mediated hypertrophy of the synergist. In our laboratory, we mechanically overload the plantaris muscle of the mouse hind limb by removing a portion of the gastrocnemius/soleus complex (13). We typically observe a doubling in muscle mass within 1–2 weeks and muscle fiber hypertrophy in excess of 35% after a few months (14–16); this rate and magnitude of hypertrophy exceeds that which occurs in humans following similar durations of resistance exercise training. Given the invasive and potentially traumatic nature of the surgery as well as the chronic and considerable load placed on the plantaris muscle following synergist ablation, a damage-mediated regenerative response is usually observed within two weeks. This regenerative response is evidenced by developmental myosin expression, centrally-located myonuclei, and the appearance of small-caliber muscle fibers (16, 17). Concomitant with new muscle fiber formation, there is usually an increase in muscle fiber number as early as 1–2 weeks when quantified on histological cross-sections (15–17) which persists until at least 8 weeks (unpublished data from Fry et al.) (14).

The plantaris muscle in rodents is a pennated muscle, and changes in internal muscle geometry (including pennation angle) are characteristic of the hypertrophic process (18). It has been postulated that altered pennation angle may largely explain increased fiber number on histological cross-sections following synergist ablation (19, 20). Roy et al. reported that plantaris pennation angle in the mouse increased modestly but significantly after 8 weeks of synergist ablation (14° to 16°, P<0.05) (21). This magnitude of pennation angle change during hypertrophy is consistent with what is observed after prolonged resistance training in the vastus lateralis of humans (22), but the degree of hypertrophy is much greater with synergist ablation. Assuming that muscle fiber growth is in lockstep with pennation angle during hypertrophy of the mouse plantaris, fiber counts should remain stable via histology when sectioning is performed perpendicular to the long axis of the muscle (as is almost always the case). However, if muscle fiber hypertrophy and architectural changes are out of sync during synergist ablation, perhaps due to edema or because the plantaris has adapted aberrantly to its new role as a primary plantarflexor, more muscle fibers could appear on whole muscle cross-section as an artifact of increased pennation angle (see Figure 1). Alternatively, the longitudinal growth of fibers that terminate intrafasicularly could also mediate the appearance of more muscle fibers on cross-section after overload, most noticeably in pennated muscles (19, 20, 23). Another explanation is that the location of mid-muscle belly may change with overload, which could be problematic when fiber number on cross-sections can differ at different locations along the length of the muscle (24). It is our estimation that the aforementioned possibilities contribute to the increase in muscle fiber number on cross-sections observed after overload, but are unlikely to account for the entire increase. Hyperplasia via a regenerative response from satellite cells is therefore the “classic” explanation for increased muscle fiber number in response to synergist ablation (20).

Figure 1. Skeletal muscle architectural adaptations in the mouse plantaris following synergist ablation overload, an extreme model for rodent hypertrophy.

The synergist ablation technique induces robust hypertrophy of the plantaris muscle, which can be evaluated at the muscle fiber level via histology. During eight weeks of overload, pennation angle in the mouse plantaris increases from ~14° to ~16° (21). If muscle fiber hypertrophy (lines) keeps pace with increased pennation angle, then more muscle fibers will not appear on cross section when the muscle is mounted upright and cut perpendicular to the long axis of the muscle (circles). However, if muscle fiber hypertrophy does not sync with increased pennation angle, more muscle fibers could theoretically appear on cross-section as a consequence of increased pennation angle. The plantaris is a bi-articular muscle (61) and may not adapt architecturally as expected during hypertrophy. It is conceivable that changes in muscle geometry during hypertrophy influences the number of fibers appearing on cross section, but we do not think this could entirely account for increased muscle fiber number during hypertrophy.

THE APPERANCE OF MUSCLE FIBER SPLITTING IN DISEASE

The notion that muscle fiber branching and splitting during hypertrophy is purely a consequence of damage and regeneration via satellite cells is supported by evidence in disease states. For example, muscular dystrophies and other myopathic/neurogenic disorders are characterized by chronic degeneration and regeneration and branched or split muscle fibers are a hallmark characteristic (25–27). In murine dystrophy models specifically, compensatory muscle fiber hypertrophy is primarily mediated via apparent muscle fiber splitting (28), and branched fibers may compromise muscle function via mechanisms independent from a lack of dystrophin (29–32). Alternatively, in healthy amphibian muscle, branched skeletal muscle fibers are common and not associated with excitability or contractile defects (33), suggesting that fiber branching or splitting is not necessarily pathological.

EVIDENCE FOR FIBER SPLITTING UNCOUPLED FROM DEGENERATION-REGENERATION IN MAMMALS

There is considerable evidence for muscle fiber splitting with models of extreme overload in birds (see above) as well as rodents (12, 34–38). It is worth noting that two reports do not show increased fiber number in the rat plantaris (19) or mouse soleus (24) via individual counting of isolated fibers after 4–8 weeks of synergist ablation, nor significant evidence of fiber branching or splitting, but the latter stands in contrast to other reports. Evidence from non-surgical models of mammalian muscle hypertrophy in which a regenerative response is minimized suggests that muscle fiber splitting does occur. The appearance of fiber splitting is a prominent feature with weight-lifting paradigms in rats (39, 40) and cats (41, 42), and also occurs during weighted wheel running in mice (43). While these models are comparatively more translatable to resistance exercise training in humans, they are not totally devoid of a regenerative response.

In order to study the hallmarks of muscle fiber splitting in greater detail, our laboratory modified the synergist ablation surgical approach to be less invasive and provide a more gentle hypertrophic stimulus (44). After 14 days of synergist tenotomy, where the tendons of the gastrocnemius were excised but the soleus was left intact in order to support the plantaris during overload, we observed minimal developmental myosin expression in the plantaris (<1% versus ~30% with a traditional synergist ablation) (16), but a preponderance of large muscle fibers with centrally-located myonuclei. We also observed the appearance of muscle fiber splitting on histological cross sections and isolated single muscle fibers (Figure 2). On serial cross-sections, the appearance of a central nucleus preceded the splitting of a muscle fiber concomitant with continuous dystrophin (sarcolemma) and laminin (basal lamina) in each new branch. The appearance of a central myonucleus prior to a branch is consistent with previous reports that observed fiber splitting during overload in rats (35, 39). These data collectively suggest that fiber splitting is characteristic of robust hypertrophy in animals, and that mis-positioned myonuclei signify this process. However, given the presence of satellite cells in this and all aforementioned models, the possibility that fiber splitting during extreme hypertrophy is actually “defect regeneration” or “incomplete regeneration” cannot be ruled out (31, 45, 46) since satellite cells are the engines of muscle regeneration (16). The process of defective regeneration is portrayed in Figure 3A-G, where a muscle fiber subjected to a robust hypertrophic stimulus undergoes degeneration due to damage, and subsequent regeneration results in the incomplete fusion of satellite cells to create the appearance of a branched or split muscle fiber. In Figure 3A’-G’, the concept of myocyte grafting is depicted, where focal damage to the muscle fiber sarcolemma during hypertrophy ultimately produces a splitting-like phenotype mediated by satellite cells.

Figure 2. The appearance of split muscle fibers following 14 days of modified synergist ablation overload of the mouse plantaris.

A-D illustrate fiber splitting over ~50 µm on serial cross sections of a frozen plantaris muscle. Representative images show laminin and dystrophin to identify muscle fiber borders, and myonuclei. Dark arrows point to central myonuclei that appear prior to the appearance of each new branch in the muscle fiber (light arrows). E shows a phase-contrast image of a trifurcated single muscle fiber from this same mouse, along with myonuclei. F illustrates the extent of fiber splitting morphology in this mouse, with minimal eMyHC expression (a marker of regeneration). Note that many fibers contain >1 mis-positioned myonucleus, which appears to signify the splitting process. Immunohistochemistry images were captured at 20x and 40x magnification, and single fiber image was captured at 40x; Scale bars = 50 µm. (Reprinted from (44). Copyright © 2017 BioMed Central. Used with permission.)

Figure 3. Potential explanations for increased fiber number on cross-section during robust hypertrophy.

In scenario A-G, a muscle fiber response to an extreme hypertrophic stimulus but is damaged in the process. This leads to a degeneration-regeneration response in which the sarcolemma is damaged but the basal lamina initially remains intact. The muscle fiber is ultimately degraded while satellite cells are activated to initiate regeneration. During regeneration, aberrant fusion of satellite cells leads to the appearance of a branched (assymetrical, not depicted) or split (symmetrical) muscle fiber. In scenario A’-G’, there is focal damage to the sarcolemma in response to a robust hypertrophic stimulus. Satellite cells are activated and form a myocyte that “grafts” to the existing muscle fiber, ultimately resulting in a branching or splitting phenotype. In scenario A”-G”, an existing muscle fiber grows and reaches a critical point, upon which it splits. Whether a split fiber can become two separate muscle fibers, and whether satellite cells are required to fuse in to maintain a new muscle fiber is unknown.

In contrast to the notion that muscle fiber splitting during hypertrophy is primarily defective regeneration or myocyte grafting, we point to findings from our Pax7-DTA mouse model. Tamoxifen treatment of adult Pax7-DTA mice (>4 months old) activates diphtheria toxin A specifically in satellite cells and results in depletion of satellite cells by >90%, thereby severely reducing the ability of a muscle to regenerate (16). We then overload the mice via synergist ablation for various durations in the absence of satellite cells. Robust muscle fiber hypertrophy ensues after 10–14 days of overload without satellite cells, concomitant with a significant increase in muscle fiber number on histological cross-sections (16, 17, 44). Furthermore, a 26% increase in muscle fiber number persists for up to 8 weeks of overload without satellite cells (unpublished data from Fry et al.) (14). Since central myonuclei and split muscle fibers are generally not abundant in un-stressed muscle (39), we speculate that muscle fiber splitting could represent a non-pathological component of robust muscle fiber hypertrophy in animal models that may occur independent from satellite cells and regeneration (see Figure 3A”-F”).

EVIDENCE FOR FIBER SPLITTING IN A NON-SURGICAL MODEL OF HYPERTROPHY IN MICE

Our laboratory recently developed a progressive weighted wheel running approach (PoWeR) that employs a gradual progression of loading combined with an unbalanced wheel that provides a novel resistive component and encourages continued running activity at heavier loads. We did not observe small-caliber muscle fibers, developmental myosin expression, or increased muscle fiber number on histological cross-sections in the soleus after 4 or 8 weeks of PoWeR, and muscle wet weight and muscle fiber size was significantly larger than controls by 8 weeks. Although muscle wet weight remained higher than controls after 12 weeks of PoWeR, muscle fiber size returned toward untrained levels and muscle fiber number was elevated (Figure 4). The shape of the fiber size distribution curve expectedly shifted to the right at 8 weeks indicating muscle fiber hypertrophy but shifted back toward medium-sized fibers with a loss of very large fibers at 12 weeks (Figure 4). The central peak at 12 weeks suggests that fiber number increased due to the splitting of large muscle fibers. Consistent with our findings, the appearance of muscle fiber splitting on serial cross-sections was apparent in soleus muscles of mice that completed 10 weeks of balanced weighted wheel running (43). It is important to note that the soleus muscle is ideal for studying muscle fiber hypertrophy and splitting on cross-sections in mice due to its shallow pennation angle relative to the plantaris (5° versus 14°) (21, 47) and because most or all fibers can be captured on cross-sections if cut at mid-belly (20, 24, 34). If spontaneous regeneration between 8 and 12 weeks of training were the explanation for greater fiber number at 12 weeks, then a population of small-caliber muscle fibers (<500 µm²) would perhaps be expected. In contrast to early hypotheses (48), these data support a model where muscle fibers reach a critical size before splitting (3).

Figure 4. Muscle fiber size and number in the soleus muscle with progressive weighted wheel running (PoWeR) in adult mice.

PoWeR is a non-invasive minimally-injurious method for inducing muscle fiber hypertrophy via exercise in mice. A Relative to untrained age-matched control mice, muscle fiber cross sectional area (CSA) measured on entire histological cross-sections using dystrophin immunohistochemistry and automated software (62) is greater after 8 weeks and muscle fiber number is not different. After 12 weeks of PoWeR, muscle fiber size is lower relative to 8 week-trained mice, and muscle fiber number is greater. B The leftward shift in the muscle fiber size distribution curve suggests that the largest muscle fibers may have split sometime between 8 and 12 weeks of training. Data were normally distributed, significance was determined via one-way ANOVA with a Tukey’s post-hoc test and significance set at P<0.05, shared letters = NOT significantly different, ^#p=0.12 vs untrained, n = 5 female >4 month old C57BL6 mice per group.

EVIDENCE FOR FIBER SPLITTING WITH RESISTANCE TRAINING IN HUMANS

In general, the evidence for muscle fiber splitting during hypertrophy in humans is limited, indirect, and circumstantial; however, some human studies with small samples sizes and unique populations suggest that fiber splitting in humans during extreme hypertrophy is possible. Using indirect measures, MacDougall et al. (49), Alway et al. (50), and D’Antona et al. (51) reported that muscle fiber hypertrophy alone could not explain the extreme whole muscle hypertrophy observed in competitive bodybuilders. Their observations suggest that: 1) elite bodybuilders are born with more muscle fibers, or 2) more muscle fibers can manifest during extreme muscle loading or hypertrophy. Using histochemical and immunohistochemical techniques, several investigations report the appearance of muscle fiber splitting in muscle biopsy samples from individuals undergoing extreme loading (45, 52, 53). The studies by Eriksson et al. show the morphology of fiber splitting on serial cross-sections in a small sampling of power lifters that frequently subject themselves to abnormally heavy muscle loading (45, 52). However, the incidence of split fibers may be influenced by anabolic steroid usage in these individuals. Exogenous androgens can activate satellite cells (54, 55) which could theoretically increase the likelihood of an event such as myocyte grafting; however, steroid usage also allows for more intense and frequent loading which may increase the likelihood of muscle fiber splitting.

Eriksson et al.’s anecdotal observations are likely mediated in part by “defective” or “incomplete” regeneration since satellite cells are surely playing a role in the muscle remodeling process in elite strength athletes. Nevertheless, it is possible that fiber branching or splitting of existing muscle fibers also contributes to hypertrophy in some circumstances. The extent to which fiber branching, splitting, or true hyperplasia occurs during muscle hypertrophy may never be determined in humans and may be negligible with normal exercise training. A genetic predisposition to having more muscle fibers, as suggested by MacDougall et al. in intermediate caliber bodybuilders (56), cannot be ruled out. It is also conceivable that fiber splitting is the consequence of frequent exposure to eccentric muscle actions, which also occurs throughout the course of competitive/elite high-load weight-lifting, and could result in pronounced hypertrophy with splitting or defective regeneration/myocyte grafting (46). The creation of new muscle fibers in adult human muscle may be possible with extreme and very prolonged loading and hypertrophy, especially in combination with anabolic steroids, but further work is needed to draw firm conclusions.

POTENTIAL REASONS FOR MUSCLE FIBER SPLITTING

If muscle fiber branching, splitting, and or hyperplasia from existing muscle fibers occurs during extreme muscle loading and subsequent hypertrophy, it is attractive to speculate on the underlying causes. One possible explanation is that possessing multiple smaller fibers versus one large fiber confers a biomechanical advantage during loading; dividing one large fiber into two or more smaller fibers would distribute force over a larger surface area. This explanation implies that load, and not hypertrophy itself drives splitting. An alternative explanation is that when muscle fibers grow too large, an upper limit in oxygen diffusion capacity is reached that then may trigger splitting in order to maintain functionality (57). The appearance of muscle fiber splitting during aging in rats occurs concomitant with mitochondrial defects (58), which provides some rational for the “oxygen diffusion” hypothesis when applied to hypertrophy. Further evidence is found in crab swimming muscles, where large oxidative muscle fibers become “sub-divided” during hypertrophy in order to maintain oxygen diffusion capacity (59). A third possibility is that an upper limit for the area that a myonucleus can transcriptionally govern, or “myonuclear domain” is exceeded, but this limit has yet to be identified definitively with experimental evidence in animals or humans (60). Further research into the cause(s) of muscle fiber splitting could reveal novel mechanism(s) by which skeletal muscle mass is regulated; this knowledge could be leveraged as a therapeutic in circumstances of muscle fiber loss (e.g. aging).

FUTURE DIRECTIONS AND CONCLUSIONS

It is contentious that adult skeletal muscle can grow via mechanisms other than muscle fiber hypertrophy. In this review, we outline the evidence for muscle fiber branching and splitting during extreme loading and hypertrophy, and hypothesize that these processes may represent a non-pathological adaptation. It is presently unknown to what extent split muscle fibers contribute to muscle fiber hypertrophy, and how much of the appearance of split fibers is attributable to defective or incomplete regeneration mediated by satellite cells. More detailed investigations using the Pax7-DTA satellite cell depletion mouse model, as well as translatable methods of murine exercise training such as PoWeR, may help provide clarity. Also unknown is what specifically triggers fiber splitting (e.g. heavy, frequent, or high-volume loading, extreme hypertrophy, rapid hypertrophy, eccentric muscle actions, genetic predisposition, anabolic steroid usage, or a combination of some or all factors), the magnitude to which it may occur, and the time course in which it can happen. A final unknown is whether a branched or split muscle fiber results in bona fide hyperplasia, complete with innervation and a fully functioning contractile apparatus. Collectively, the literature indicates that skeletal muscle plasticity extends beyond the ability of individual muscle fibers to hypertrophy, and future investigations will provide insight into the mechanism(s) of muscle fiber splitting.

Key Points.

• The branching or splitting of skeletal muscle fibers occurs during extreme muscle loading and subsequent hypertrophy.

• Fiber splitting has been documented in a variety of animal hypertrophy models that are not associated with a marked degeneration-regeneration response, which suggests the process may not be pathological.

• Increased muscle fiber number during hypertrophy may occur in the absence of muscle stem cells (satellite cells).

• Some evidence suggests that muscle fiber splitting in response to robust muscle tension or hypertrophy is can occur in humans but may be limited to those using exogenous steroids during training.

• The extent to which fiber splitting or bona fide hyperplasia contributes to resistance training-induced muscle fiber hypertrophy in humans is not well-understood.