Starring or Supporting Role? Satellite Cells and Skeletal Muscle Fiber Size Regulation

Abstract

Recent loss-of-function studies show that satellite cell depletion does not promote sarcopenia or unloading-induced atrophy, and does not prevent regrowth. Although overload-induced muscle fiber hypertrophy is normally associated with satellite cell-mediated myonuclear accretion, hypertrophic adaptation proceeds in the absence of satellite cells in fully grown adult mice, but not in young growing mice. Emerging evidence also indicates that satellite cells play an important role in remodeling the extracellular matrix during hypertrophy.

Introduction

In 1865, structures that were termed “muskelkörperchen,” or “muscle corpuscles,” were first identified during the study of muscle regeneration (193). Although the precise origin of these cells was unknown, they were thought to give rise to new muscle fibers (164). Almost 100 years later, Alexander Mauro (106) and Nobel laureate Sir Bernard Katz (82) independently but simultaneously observed these same cells while studying the peripheral regions of extrafusal and intrafusal fibers, respectively. They referred to these cells as “satellite cells,” given their satellite-like position along the outside boundary of the sarcolemma. Although Mauro is almost exclusively credited with coining the term “satellite cell,” J. David Robertson used the term “satellite cell” in investigations of intrafusal muscle fibers that predated both Mauro and Katz; however, Robertson believed that peripherally located satellite cells were related to Schwann cells (154). Nevertheless, in the opening paragraph of his seminal report, Mauro speculated that satellite cells could be the engines of muscle regeneration. Shortly thereafter, numerous investigations linked satellite cells to the progression of the regenerative process after injury (4, 5, 31, 169). Although satellite cells were identified, their origin and precise function remained elusive. Erroneous early reports indicated that satellite cells were not found in uninjured skeletal muscles, leading some researchers to conjecture that satellite cells were mononuclear cells that “broke off” the muscle fiber during injury (Refs. 45, 152, 153; reviewed in Ref. 23). In the closing paragraph of his report, Mauro stated, “the correct explanation of the . . . role of the satellite cell must await the outcome of further studies.”

At the current time, a PubMed search for “satellite cells” returns >11,000 articles that collectively address the contribution of satellite cells to skeletal muscle maturation, regeneration, health, disease, aging, and exercise adaptation across numerous species. It is now known that satellite cells comprise an autonomous cell population located underneath the basal lamina that is essential for proper postnatal muscle development (168) and, as Mauro initially postulated, are indispensable for muscle regeneration following injury (94, 109, 125, 162). Since myonuclei contained within syncytial muscle fibers are considered post-mitotic (23, 119, 120, 143, 165, 178), it is accepted that satellite cell-fusion into muscle fibers is required for myonuclear replacement or addition (46, 118, 166). The “myonuclear domain” theory posits that the cytoplasmic area that a myonucleus can transcriptionally govern is relatively fixed in adult skeletal muscle (29, 66, 137). It has therefore been assumed that satellite cell-dependent myonuclear accretion is unconditionally required for adult skeletal muscle fiber hypertrophy (128, 144, 182). Although muscle fiber hypertrophy is normally associated with myonuclear addition (6, 139, 140, 147, 160, 166), hypertrophy in the presence of satellite cells but absence of myonuclear accretion has also been reported (70, 78, 139, 140, 175, 187, 190), suggesting the myonuclear domain is flexible (186, 187). A major advance in the field was the development of conditional satellite cell knockout mice in 2011 (94, 109, 125, 162), which has enabled researchers to directly test the necessity of satellite cells for postnatal skeletal muscle adaptation. The purpose of this review is to provide background and perspective on the varied roles of satellite cells in muscle fiber size regulation, highlighting results from recent satellite cell loss-of-function investigations.

Satellite Cells Are Necessary for Postnatal Skeletal Muscle Growth

It is generally accepted that postnatal skeletal muscle development in mammals is primarily driven by muscle fiber hypertrophy and not hyperplasia (76, 134, 197). As such, the principal role of satellite cells during maturational skeletal muscle growth is myonuclear accretion to support the transcriptional demands of postnatal development. The work of White et al. indicates that mouse extensor digitorum longus (EDL) muscle fiber size increases approximately eightfold and length increases approximately fourfold by postnatal day 56, whereas myonuclear accretion stabilizes by postnatal day 21 (197). The maturational growth that occurs in adolescence (between days 21 and 56) therefore involves a rapid and significant expansion of the myonuclear domain that is preceded by myonuclear accretion. The steep increase in myonuclear domain size during early muscle maturation has been confirmed in humans (36). Robust myonuclear accretion in early postnatal development appears necessary to increase ribosomal RNA transcription, facilitate ribosome biogenesis, and augment translational capacity to support the profound rate of postnatal muscle growth (34). Satellite cell and myonuclear accumulation at the ends of growing fibers may also enable muscle fiber lengthening during postnatal growth (8, 158).

Although myonuclear accretion into muscle fibers is reportedly complete within the first few weeks of postnatal development (197), other work suggests that satellite cell fusion extends into young adulthood. The Gundersen laboratory reported significant myonuclear accretion into EDL and soleus muscle fibers between 2 and 14 mo of age (20). A few investigations using fluorescent reporter mice suggest that satellite cell contribution to muscle fibers does not stabilize until ~3 mo of age (138) and persists at basal levels throughout the lifespan across various skeletal muscles (85, 138). Since skeletal muscle fibers are multi-nucleated, tracking the fate of satellite cell-derived myonuclei presents a unique challenge. Unless satellite cell-derived myonuclei are permanently labeled and tracked, satellite cell-mediated myonuclear turnover rates cannot be accurately quantified. It is conceivable that satellite cell fusion into muscle fibers for homeostatic purposes does not occur appreciably beyond a certain age, but non-fusion-mediated communication between satellite cells and muscle fibers should not be ruled out. Regardless, correlational findings indicate that satellite cell-mediated myonuclear accretion is an integral component of postnatal muscle growth, and the requirement for myonuclear accretion to support radial and perhaps longitudinal fiber growth may persist at least into late adolescence.

Paired box 7 (Pax7) is a necessary transcription factor for maintaining satellite cell quiescence and is the most well-accepted marker for identifying satellite cells in skeletal muscle (90, 161, 168, 192). The Pax7 knockout mouse provided the first experimental insight into the effects of satellite cell depletion in early postnatal maturation. Pax7-null mice are generally born without an overt phenotype (104) but experience a precipitous decline in satellite cells (~80%) by 10–11 days after birth (168). Most Pax7 knockout mice die by 2 wk of age (135, 168). Small muscle fibers with few nuclei (90, 150, 168) and overall muscle weakness is reported in mice that survive (90). These loss-of-function studies support the observational findings that satellite cells are necessary for proper postnatal skeletal muscle growth.

Indirect Evidence of a Requirement for Satellite Cell-Mediated Myonuclear Accretion During Hypertrophy

Given the importance of satellite cell-mediated myonuclear accretion during postnatal muscle fiber growth, it has been postulated that satellite cells are also necessary for hypertrophic adaptation in mature skeletal muscle. In 1968, Reger and Craig (147) were perhaps the first to suggest that satellite cells were the source of increased myonuclear number observed previously during muscle growth in animals (47, 100). Evaluating a muscle biopsy sample from a young woman presenting with clinical dysplasia and “bizarre muscle hypertrophy,” these authors insightfully speculated that “satellite cells may assume myoblast-like properties and serve as a focal point for muscle fiber enlargement during muscle hypertrophy” and “the increase in number of nuclei [within muscle fibers] evidenced elsewhere may result from cellular fusion involving satellite cells.”

The work of Moss and Leblond in 1970 using thymidine labeling in rats provided evidence that satellite cells are the source of new myonuclei (119). Schiaffino et al. subsequently reported that satellite cells seemed to account for increased myonuclear density following compensatory hypertrophy (166), thereby supporting Reger and Craig’s hypothesis. Around this time, Cheek et al. proposed that “each nucleus has jurisdiction over a finite volume of cytoplasm” in muscle since the protein-to-DNA ratio generally scaled during growth (29). This “DNA unit” concept ultimately gave rise to the notion that the “myonuclear domain” must remain stable, requiring satellite cells to contribute new myonuclei to meet transcriptional demands during muscle fiber hypertrophy (6, 66, 160). Implicit in this theory is that each myonucleus in the fiber is working at its transcriptional maximum.

Numerous human investigations have since reported satellite cell expansion with acute and chronic resistance exercise (12, 14, 32, 33, 39, 50, 73, 102, 103, 110, 113, 123, 130, 139, 149, 159, 174, 190), as well as increased myonuclear density when muscle fibers hypertrophy (14, 50, 78, 79, 132, 139, 140, 173). Additionally, the efficacy of satellite cell-mediated myonuclear accretion has been linked to the extent of muscle fiber hypertrophy from resistance training in humans (14, 139, 140, 173), and a hypertrophic “threshold” beyond which myonuclear accretion becomes necessary (~26%) has also been proposed (78, 139). Work from our laboratory suggests that oxidative, slow myosin heavy chain (MyHC) type 1 muscle fibers experience preferential satellite cell expansion (55, 123) and myonuclear accretion (55) compared with glycolytic, fast-twitch MyHC type 2 fibers during hypertrophy induced by bicycle ergometer training in humans, suggesting potential fiber type-specific differences in satellite cell requirement. A more stringent reliance on myonuclear accretion during hypertrophy of oxidative muscle fibers may be due to their greater relative metabolic activity and protein synthesis rate (FIGURE 1) (13, 38, 160, 184, 188). Fiber type-specific gene expression analysis in humans further supports the notion that transcriptional regulation of hypertrophy differs in MyHC type 1 vs. 2a muscle fibers (122, 146).

FIGURE 1.

Theoretical model of fiber type-specific satellite cell dependence during hypertrophy

In the presence of satellite cells (SC), oxidative myosin heavy chain (MyHC) type 1 and glycolytic MyHC type 2 muscle fibers respond to a hypertrophic stimulus via satellite cell proliferation and myonuclear accretion. In the absence of satellite cells, resident myonuclei of aerobic fibers may lack the transcriptional capacity to sustain hypertrophy due to the biosynthetic demands of oxidative metabolism combined with growth processes, whereas glycolytic fibers may possess a more flexible myonuclear domain that still permits muscle fiber growth.

Contrary to a proposed “rigid” myonuclear domain in adult skeletal muscle fibers, there is evidence for “flexibility” within the myonuclear domain in humans (78, 139, 186, 191). For instance, resistance training-induced muscle fiber hypertrophy of >26% in the absence of myonuclear accretion has been reported (70, 190), as well as myonuclear accretion in the complete absence of hypertrophy (102). Nevertheless, the positive relationship between exercise-induced hypertrophy and satellite cell-mediated myonuclear accretion has perpetuated the idea that adult skeletal muscle fibers are constrained by a “myonuclear domain ceiling” (139, 140, 186, 187). The question of whether satellite cells are required for hypertrophy has inspired debates and commentaries (108, 128, 129, 148). In a 2007 Point:Counterpoint editorial, McCarthy and Esser pointed out that various lines of evidence suggest robust pharmacologically driven muscle fiber hypertrophy occurs without satellite cell-mediated myonuclear accretion in adult skeletal muscle of rodents (108, 128). Unfortunately, direct evidence regarding the necessity of satellite cells for adult muscle fiber hypertrophy was not available at that time.

Since the 2007 editorial, additional indirect evidence further supports the conclusion that non-mechanically induced hypertrophy can occur in the absence of myonuclear accretion. Inhibiting myostatin (a powerful negative regulator of skeletal muscle mass) results in enlarged muscle fibers (10–30% larger than adult wild-type mice) that are not accompanied by increased satellite cell or myonuclear density (10, 92, 133, 194, 196). Furthermore, overexpression of AKT (a key signaling node in hypertrophic pathways) (15) and JunB (a transcription factor regulating atrophy signaling) (145) results in significant hypertrophy of adult skeletal muscle without myonuclear accretion. Hypertrophy and force production are often interrelated, but some models of non-mechanically induced satellite cell-independent hypertrophy do not result in a commensurate increase in muscle force production (9, 19). Hypertrophic adaptation by mechanical means may fundamentally differ from non-mechanically mediated hypertrophy. Contraction, and not hypertrophy itself, may drive satellite cell-mediated myonuclear accretion, which is evidenced by myonuclear accretion via exercise in the absence of muscle fiber growth (102, 111).

The first attempts at determining whether satellite cell-mediated myonuclear accretion was required for mechanically-induced adult muscle fiber hypertrophy employed γ-irradiation. DNA damage induced by low-dose radiation exposure prevents satellite cell proliferation and, consequently, myonuclear accretion. Utilizing rodent models, some investigations reported that irradiation prevented satellite cell proliferation and muscle hypertrophy in response to mechanical overload (2, 141, 155–157). Other studies in various species inferred that γ-irradiation had no effect during mechanically induced muscle hypertrophy (51, 99). Conclusions from this model are contentious, and the model itself has limitations that hinder interpretation. Classic studies suggest that 50–60% of nuclei in whole skeletal muscle are post-mitotic myonuclei (47, 167). Ameliorating the proliferative potential of other mononucleated cell types within skeletal muscle, or perhaps the transcriptional effectiveness of resident myonuclei, could influence the hypertrophic process. It is also conceivable that irradiation interrupted mTOR signaling and protein synthesis, thereby inhibiting hypertrophy (2, 17). The ability to specifically delete satellite cells in adult mice now provides the tool necessary to assess the requirement of satellite cell-dependent myonuclear accretion for mechanically induced hypertrophy.

A Mouse Model of Satellite Cell Depletion

Four separate research teams simultaneously generated conditional satellite cell-specific knockout mice in 2011 (94, 109, 125, 162). In our laboratory’s strategy (109), the driver mouse (Pax7-Cre^ER) contains a chimeric transgene encoding the viral Cre recombinase fused to a mutated ligand binding domain of the modified estrogen receptor, such that it no longer binds estrogen but does bind tamoxifen (125, 127). This construct was knocked into the Pax7 locus so that expression of Cre^ER in muscle is restricted to satellite cells. The modified estrogen receptor of the Cre^ER protein keeps it sequestered in the cytoplasm, bound to HSP90, until tamoxifen binding allows Cre^ER to translocate to the nucleus and induce Cre-mediated recombinase activity. A second mouse strain contains a modified diphtheria toxin A chain gene [DTA; active in the absence of binding to the DTA receptor (198)] knocked into the Rosa26 locus (Rosa-DTA). Rosa26 is a constitutively active promoter; however, a stop codon, flanked by loxP sites and recognized by the Cre recombinase, was inserted between the promoter and the DTA transgene, effectively silencing the transgene. Crossing these two mouse strains generates the Pax7-DTA strain. Allowing Pax7-DTA offspring to grow to maturity (4 mo of age) followed by intraperitoneal injection of tamoxifen results in Cre-mediated activation of the DTA gene, which subsequently kills satellite cells, effectively depleting them by >90%. For a more detailed overview of Cre-mediated satellite cell depletion, refer to Relaix and Zammit (151). By effectively deleting satellite cells, all four laboratories confirmed the initial postulation of Mauro, demonstrating that satellite cells are indispensable for muscle regeneration (94, 109, 125, 162). For a comprehensive overview of the role of satellite cells in muscle regeneration, refer to recent reviews (40, 151, 195).

Satellite Cells Are Not Required for Adult Muscle Fiber Size Maintenance

Utilizing the Pax7-DTA mouse model, our laboratory and others have shown that satellite cells are not required for homeostatic maintenance of muscle fiber size. Short-term deletion of satellite cells (2–8 wk) in adult mice does not result in muscle fiber atrophy (43, 52, 53, 109, 124). Even with increased activity via wheel running for 8 wk, muscle fiber size is not compromised in satellite cell-depleted hind limb (74) or diaphragm muscles (121). Furthermore, when satellite cells are depleted in adulthood (>4 mo of age), sarcopenia is generally not exacerbated 6–12 mo or 14–20 mo post-depletion in cage-dwelling mice (54, 85, 91). One recent investigation reported that satellite cell depletion accelerated sarcopenia by 12 mo of age (97), but this should be interpreted with caution; a limited number of isolated fibers from one dorsiflexor muscle (EDL) were analyzed for size, and mice were depleted of satellite cells before full maturity (3 mo of age). Instead of eliciting premature sarcopenia, satellite cell depletion may have inhibited maturational growth in this scenario. Both Fry et al. and Keefe et al. reported that myonuclear number does not decline throughout adulthood in the absence of satellite cells in the plantaris muscle (54, 85). In the EDL, myonuclear number decreased to the same extent by late life in the presence and absence of satellite cells (85), but satellite cell-depleted muscle fibers were 15% smaller, suggesting the potential for muscle-specific satellite cell contribution to muscle mass maintenance. Challenging the conclusion drawn from satellite cell reporter mice with a non-nuclear signal (85, 138), data from the Pax7-DTA model generally suggest that satellite cell-mediated myonuclear turnover is low under non-stressed conditions.

The identification of non-satellite stem cells in skeletal muscle with myogenic potential (37, 96, 115) raises the possibility that alternative cell populations could substitute for satellite cells to support homeostatic myonuclear replacement; however, these investigations did not specifically track stem cell nuclear fate. Use of lineage-tracing reporter mice with a nuclear-restricted signal will provide clarity on the contribution of satellite cells or alternative stem cells to myonuclear accretion during homeostasis and remodeling (25, 26). A ubiquitous finding in studies utilizing the Pax7-DTA model is that satellite cell depletion does not affect fiber-type distribution, regardless of depletion length or activity status (53, 54, 74, 85, 97, 109, 121, 124). Furthermore, satellite cells are not required for serial sarcomere addition in adult muscle that has been chronically stretched (86). Collectively, it can be concluded that satellite cells are dispensable for maintaining adult muscle fiber size and fiber-type distribution in non-stressed conditions, but more work is needed to determine the frequency and source of homeostatic myonuclear replacement.

Satellite Cell-Independent Hypertrophy During Mechanical Overload

Our laboratory was the first to report that satellite cells were not necessary for short-term (2 wk) mechanically induced hypertrophy in fully grown (>4 mo old) Pax7-DTA mice (109). To accomplish this, we utilized a synergist ablation overload approach where portions of the gastrocnemius and soleus muscles were excised, causing compensatory hypertrophy of the plantaris muscles (87). Using this model, we went on to show that resident myonuclei in mature skeletal muscle can compensate by sustaining transcriptional output in the absence of satellite cell-mediated myonuclear accretion (88). Robust transcriptional reserve capacity in resident myonuclei directly challenges the myonuclear domain theory within growing muscle fibers. Interestingly, resident myonuclei may be functionally distinct from satellite cell-derived myonuclei during hypertrophy, although this notion requires further investigation. Regenerated muscle fibers containing predominantly satellite cell-derived myonuclei lose the ability to grow during 2 wk of overload, suggesting resident myonuclei primarily facilitate growth processes (84). Muscle fiber hypertrophy without significant myonuclear accretion persists when plantaris overload is extended to 8 wk (52, 53). However, the large magnitude of muscle fiber hypertrophy during prolonged overload without satellite cells (26%) appears attenuated compared to overload to satellite cells (36%). Satellite cells play an integral role in regulating the extracellular matrix (ECM) (52, 54, 98, 125, 177), and depletion of satellite cells results in excessive fibrosis that appears to restrict long-term hypertrophy (see discussion in A Novel Role for Satellite Cells in Adult Skeletal Muscle Hypertrophy below).

In contrast to mature adult mice, hypertrophy in still-growing mice seems to require satellite cell-dependent myonuclear accretion (124). Our laboratory recently found that young growing mice (<3 mo old) depleted of satellite cells by >90% did not hypertrophy in response to mechanical overload (124). In adult satellite cell-depleted mice (>4 mo old), we also reported greater average muscle fiber hypertrophy than in our previous report (~20% vs. ~10%) (109) due to a modified surgical approach (124). By removing less gastrocnemius/soleus muscle and shortening the overload duration, we effectively overloaded the plantaris and reduced muscle fiber splitting (107, 109, 124). Muscle fiber splitting is consistently shown to occur in mice (124, 176, 189), rats (42, 67, 68, 72, 180), cats (61, 62), and humans (48, 49, 77, 101, 199) in response to extreme loading and/or hypertrophy, and does not appear to depend on satellite cell activity or regeneration (109, 124).

The lack of hypertrophy in young satellite cell-depleted mice (<4 mo of age) may explain recent claims that satellite cell fusion is absolutely required for overload-induced muscle growth. Guerci et al. (63) depleted serum response factor (SRF) in muscle of 2-mo-old mice, which impaired satellite cell proliferation and fusion. SRF-depleted mice did not grow in response to overload, whereas restoring satellite cell fusion in this model rescued the hypertrophic response. The authors concluded that myonuclear accretion is necessary for overload-induced hypertrophy, which is consistent with our findings in young growing mice (124). Goh and Millay (60) prevented myonuclear accretion during overload by deleting MyoMaker, a protein essential for myogenic cell fusion (114), specifically in satellite cells. Although satellite cell-specific MyoMaker deletion prevented overload-induced hypertrophy, the mice in this study were only ~3 mo old. Utilizing the Pax7-DTA model, Egner et al. (43) reported that mice do not hypertrophy following satellite cell depletion. However, in context with data from our laboratory and others using young growing mice, ambiguity regarding the age of satellite cell depletion (3–4 mo of age) in Egner et al.’s study makes it challenging to interpret these findings. Variation in mouse strains (60, 63), overload approaches (43, 60, 63), and tamoxifen treatment strategies (60, 63) could account for why the aforementioned studies arrived at different conclusions than our original investigation (107, 109). Alternatively, we speculate that myonuclei in growing muscle have a low transcriptional reserve capacity, necessitating satellite cell-dependent myonuclear accretion to provide the requisite transcriptional output to adapt to a hypertrophic stimulus (124). There appears to be a transition between 2 and 4 mo of age from satellite cell-dependence for overload-induced hypertrophy to satellite cell-independence, where hypertrophy can be accomplished by myonuclear domain expansion in adult muscle lacking satellite cells (FIGURE 2).

FIGURE 2.

Dynamic continuum of satellite cell contribution to hypertrophy across the lifespan

Young growing mice (<4 mo of age) require SCs for overload-induced hypertrophy, whereas adult mice (>4 mo) do not. However, aged anabolic-resistant mice (24 mo old) do not grow, regardless of the presence of satellite cells.

A study in aged anabolic-resistant mice indicated that satellite cell fusion does not drive overload-induced hypertrophy (91). Neither satellite cell-replete or -depleted mice grow with overload at 24 mo of age, even though satellite cell-mediated myonuclear accretion occurs in the former. The myonuclear domain consequently shrinks during overload in aged satellite cell-replete mice, which provides strong evidence that myonuclear accretion cannot facilitate hypertrophy in adulthood, at least in aged mice. A fiber-type transition with overload persists in old mice independent from satellite cells, confirming that satellite cells are not necessary for fiber-type plasticity (91). Collectively, Pax7-DTA overload experiments indicate that the satellite cell requirements during hypertrophy change along a dynamic continuum throughout the lifespan (FIGURE 2) (124).

A caveat to the interpretation of plantaris overload studies is that the mouse plantaris is composed primarily of fast-twitch MyHC type 2a, 2x, and 2b muscle fibers; humans do not generally express the MyHC type 2b protein but do express MyHC type 2x (69, 172). Although mouse MyHC type 2a fibers appear similar in oxidative capacity to human slow-twitch MyHC type 1 fibers (16), it is conceivable that fast-twitch muscle fibers of the mouse plantaris have a less-stringent requirement for myonuclear accretion during growth, similar to MyHC type 2 fibers in humans (see FIGURE 1 and discussion above). Indirect evidence in rodents suggests that augmenting the oxidative profile of mouse muscle fibers could alter satellite cell requirements during hypertrophy. For instance, relative to myostatin knockout-mice, elevated myonuclear density was recently reported in a myostatin knockout/estrogen-related receptor γ overexpressing mouse model characterized by very large yet oxidative muscle fibers (133). More research is needed to determine the effect of satellite cell depletion on the hypertrophic response in oxidative vs. glycolytic muscle fibers, and in muscles of varying function.

The ability to conditionally delete satellite cells has provided valuable insights into the necessity of satellite cells for mechanically induced muscle hypertrophy. One major criticism of the Pax7-DTA murine model relates to the discrepancy in muscle fiber size between mice and humans, which may limit translatability. Human muscle fibers are approximately three to five times larger than mouse muscle fibers (16). Mouse muscle fibers may never achieve a large enough absolute size during overload to exceed a hypothetical “myonuclear domain” ceiling. Scalability between species can be inferred because mouse muscle fibers have comparable myonuclear domains relative to humans (95), and the synergist ablation technique causes extreme muscle fiber hypertrophy (~30% or more by 8 wk) (53, 87). Regardless, an even greater growth stimulus in mouse models may be required to identify whether a true myonuclear domain ceiling exists. Gaps in knowledge notwithstanding, research within the last decade shows that adult mouse muscle fiber hypertrophy occurs in the absence of satellite cells, whereas hypertrophy in young mice (<4 mo of age) requires satellite cell-mediated myonuclear accretion (124).

Satellite Cells, Unloading-Induced Atrophy, and Regrowth

In addition to mechanical overload, the role of satellite cells in unloading-induced atrophy and regrowth during recovery has also been explored. One early investigation depleted satellite cells in mouse hindlimbs via γ-irradiation before 2 wk of unloading by hindlimb suspension (116). Satellite cell depletion did not exacerbate atrophy of the soleus, but myonuclear number decreased as fibers atrophied with and without irradiation. Furthermore, in the absence of satellite cells, the soleus did not fully regrow upon return to ambulation. These initial observations suggested that satellite cell depletion did not promote atrophy, but a lack of satellite cell-mediated myonuclear accretion during reloading inhibited muscle regrowth (116). Worth noting is that the mice in this investigation were 9–11 wk old and were likely still growing. Utilizing the Pax7-DTA model, our laboratory subsequently repeated these experiments in adult mice (>4 mo old). We confirmed that satellite cell depletion did not cause or exacerbate muscle fiber atrophy during 2 wk of unloading; however, myonuclei were not lost in the soleus, and regrowth occurred independent of satellite cells (75).

Several investigations support the conclusion that myonuclei in mouse skeletal muscle are not lost during 14–56 days of unloading (21, 22, 75, 105). Conversely, after long-term microgravity exposure (~90 days), myonuclear number appears to decline in mice (163). The data regarding myonuclear loss with unloading are conflicting in rats. A few investigations report that myonuclear number does not decline with varying lengths of unloading (21, 80, 81), but numerous studies from several laboratories find that short-term unloading (2 wk) results in myonuclear loss concomitant with atrophy (7, 41, 56, 71, 83, 93, 170, 171). Evidence for changes in myonuclear density during unloading in humans is limited. After 11 days of microgravity exposure (35) or 14 days of bed rest (11), myonuclear number declined in muscle fibers of the mixed fiber-type vastus lateralis. Following 28 days of bed rest, vastus lateralis myonuclear number was reduced by 11% (18). In soleus muscle fibers (primarily slow-twitch), myonuclear number declined during 2 or 4 mo of bed rest, concomitant with significant atrophy (131). Collectively, changes in myonuclear density with unloading across species, muscle groups, and unloading periods are inconsistent and sometimes contradictory. These discrepancies make it difficult to definitively determine the role of satellite cells during regrowth upon return to ambulation. It is also difficult to attribute rate of regrowth following atrophy or detraining to myonuclear density, since the frequency of myonuclear loss during atrophy and/or turnover may differ across species and conditions (65). Resident myonuclei in adult mice have a robust transcriptional reserve capacity that is apparent in overloaded satellite cell-depleted muscle (88). We speculate that, even if myonuclei are lost during unloading, resident myonuclei in adult mice could support considerable regrowth in the absence of satellite cells. More research in this area is warranted to draw definitive conclusions.

A Novel Role for Satellite Cells During Adult Skeletal Muscle Hypertrophy

Among the initial investigations reporting the effects of satellite cell depletion during regeneration using genetically modified mouse models (94, 109, 125, 162), one study revealed an important interaction between satellite cells and fibroblasts (the primary cell type responsible for ECM production) (125). Deletion of satellite cells before severe chemically mediated muscle injury caused misregulation of fibroblasts and exaggerated ECM accumulation. This work expanded on a body of literature that pointed to a role for satellite cells in regulating the muscle microenvironment, specifically the ECM (3, 44, 64, 183). The ECM provides the structural scaffolding for muscle remodeling, and facilitates mechanical tension sensing and the resulting intracellular signaling response (89). ECM composition also has a profound effect on myogenic cell behavior (58, 59, 117, 185). Uncovering mechanisms regulating ECM remodeling is an area of increasing interest, specifically as it relates to the natural process of loading-induced hypertrophy (57, 112, 179).

As outlined in Satellite Cell-Independent Hypertrophy During Mechanical Overload, muscle fiber hypertrophy is apparent during prolonged overload without satellite cells in Pax7-DTA mice, but the magnitude is reduced (36% vs. 26% with and without satellite cells, respectively) concomitant with increased ECM accumulation in the latter (53). Utilizing a coculture system, activated satellite cells downregulated expression of ECM genes in fibroblasts through a paracrine mechanism (53). A subsequent investigation revealed an exosome-mediated mechanism by which satellite cells remodel the ECM during hypertrophic growth. Once activated, satellite cells release exosomes containing micro-RNA-206. Satellite cell-derived exosomes dock with fibroblasts, delivering miR-206 that targets Rrbp1 mRNA (a master regulator of collagen production from fibroblasts) and represses ECM deposition (52) (FIGURE 3). Furthermore, the presence of satellite cells during the first week of overload is sufficient to ensure proper ECM remodeling; depletion of satellite cells after the first week of overload results in uncompromised muscle fiber hypertrophy with minimal myonuclear accretion by 8 wk (52). Since myonuclei have a robust transcriptional reserve capacity (88), we hypothesize that excessive fibrosis, and not a myonuclear domain ceiling, limits long-term hypertrophy in the absence of satellite cells in our previous work (53).

FIGURE 3.

Satellite cells regulate fibrosis during hypertrophy

SC proliferation in response to a hypertrophic stimulus mediates proper extracellular matrix (ECM) remodeling by preventing fibrosis, which appears to allow for continued muscle fiber hypertrophy. Absence of satellite cells results in blunted hypertrophy in response to long-term mechanical overload.

Recent investigations lend further support to the important role for satellite cells in regulating ECM deposition during muscle remodeling. The TEAD1 mouse is characterized by elevated baseline satellite cell density. When this mouse is crossed with the mdx mouse, which displays excessive fibrosis and satellite cell turnover, the dystrophic phenotype is partially rescued (177). Similarly, loss of syndecan-3 destabilizes satellite cells in their niche and increases proliferative activity. Concomitant with increased satellite cell density, aged as well as dystrophic syndecan-3-null mice experience less fibrosis than their syndecan-3+ counterparts (142). However, increased baseline satellite cell density in adult mice does not result in larger muscle fibers, indicating that satellite cells themselves do not drive adult muscle fiber growth (177). As discussed above, deletion of MyoMaker in young mice prevented satellite cell fusion and overload-induced hypertrophy after 2 wk, and also caused excessive fibrosis (60). It is possible that MyoMaker deletion in satellite cells created a pro-fibrotic environment within the muscle through an unknown mechanism, which may have contributed directly to attenuated hypertrophy with overload.

Numerous cell populations besides fibroblasts exist within muscle (27, 181), and it is conceivable that satellite cells communicate with many of them (including muscle fibers) in the absence of fusion. Satellite cells and endothelial cells are consistently shown to reside in close proximity to one another in vivo (30, 126), which could facilitate endothelial cell-satellite cell communication. Similarly, satellite cells were recently shown to interact directly with macrophages during regeneration (24), and a lack of satellite cells during regeneration alters macrophage dynamics (136). Satellite cells are also able to attract non-muscle cells (specifically monocytes) in vitro (28). Satellite cell depletion impairs neuromuscular junction regeneration and accelerates age-related neuromuscular deficits, suggesting communication between satellite cells and neurons (97, 98). Finally, satellite cells communicate with bone tissue through the release of growth factors to facilitate skeletal remodeling after acute trauma (1). These findings collectively indicate that satellite cells may act as a central node for intercellular communication across a spectrum of cell types within and outside of muscle to coordinate adaptation.

Summary

Understanding of the role of satellite cells during adaptation has been refined and redefined in recent years. The Pax7-DTA mouse has allowed for rigorous investigation into the necessity of satellite cells in muscle fiber size regulation. The loss of satellite cells does not promote sarcopenia, unloading-induced atrophy, or regrowth. Although hypertrophy is normally associated with satellite cell-mediated myonuclear accretion, evidence is accumulating that muscle fiber hypertrophy can proceed in the absence of satellite cells in fully grown adult mice, with no alternative stem cell population substituting, since growth occurs without appreciable myonuclear accretion. Furthermore, the role of satellite cells during hypertrophy changes throughout the lifespan. Muscle fiber hypertrophy in young, growing mice has a more stringent requirement for satellite cells, whereas satellite cell-dependent myonuclear accretion is not permissive for hypertrophy in aged anabolic-resistant mice. Recent evidence also indicates that satellite cells play an important role in remodeling the ECM during hypertrophy. Satellite cells signal to fibroblasts via miRNA-carrying exosomes to regulate ECM deposition and promote proper muscle remodeling during overload. Thus the role of satellite cells in fiber size regulation extends beyond providing a myonucleus to growing muscle fibers. A number of important questions still remain to be answered:

• • What is the prevalence of satellite cell-mediated myonuclear turnover throughout the lifespan and under stressed conditions, and can alternative stem cell populations within muscle substitute for satellite cells?
• • What regulates the transition from satellite cell-dependent to -independent hypertrophy during maturation?
• • How do adult myonuclei communicate to coordinate transcriptional activity in the absence of satellite cell-dependent myonuclear accretion during mechanically induced hypertrophy?
• • Is there a fiber type- and/or muscle-specific requirement for satellite cells during hypertrophy?
• • Do satellite cells communicate with other cells within skeletal muscle (e.g., endothelial cells or muscle fibers) in the absence of fusion, and, if so, by what mechanisms?

The results from recent loss-of-function studies highlight the importance of continued investigation into satellite cell dynamics to acquire an in-depth understanding of their contribution to muscle size regulation.