The Roles of miRNAs in Adult Skeletal Muscle Satellite Cells

Abstract

Satellite cells are bona fide muscle stem cells that are indispensable for successful post-natal muscle growth and regeneration after severe injury. These cells also participate in adult muscle adaptation in several capacities. microRNA (miRNA) are post-transcriptional regulators of mRNA that are implicated in several aspects of stem cell function. There is evidence to suggest that miRNAs affect satellite cell behavior in vivo and myogenic progenitor behavior in vitro, but the role of miRNAs in adult skeletal muscle satellite cells is less studied. In this review, we provide evidence for how miRNAs control satellite cell behavior with emphasis on satellite cells of adult muscle in vivo. We first outline how miRNAs are indispensable for satellite cell viability and control the phases of myogenesis. Next, we discuss the interplay between miRNAs and myogenic cell redox status, senescence, and communication to other muscle-resident cells during muscle adaptation. Results from recent satellite cell miRNA profiling studies are also summarized. In vitro experiments in primary myogenic cells and cell lines have been invaluable for exploring the influence of miRNAs, but we identify a need for novel genetic tools to further interrogate how miRNAs control satellite cell behavior in adult skeletal muscle in vivo.

Graphical Abstract

Brief History of microRNAs (miRNAs) and Satellite Cells in Skeletal Muscle

miRNAs are small non-coding RNAs that canonically bind to the 3’ UTR of target mRNA to prevent translation via transcript destabilization and/or impeding translation initiation. The first miRNA (lin-4) was identified in Caenorhabditis elegans, or C. elegans [1, 2]. Subsequent studies revealed that miRNA were phylogenetically conserved [3] and may have an important role in the regulation of protein translation in animals [4–6], including mice [7, 8] and humans [8]. Roughly half of miRNA reside in a protein-coding gene and are transcribed simultaneous with that gene [9, 10]. Some evidence suggests that up to 35% of intronic miRNAs are expressed as an independent transcription unit under regulation of its own promoter [11]. Regardless of genomic context, following pri-miRNA processing, the 60–70 bp precursor pre-miRNA is transported from the nucleus by Exportin 5 to the cytoplasm [12, 13]. In the cytoplasm, a second RNase III endonuclease, Dicer, cleaves the pre-miRNA to produce a ~22 nucleotide double-stranded RNA molecule in which one strand, termed the guide strand, is transferred to the RISC (RNA-induced silencing complex) that contains Argonaute 2 (Ago2) and the RNA binding protein TARBP2 (TAR RNA binding protein 2); the non-guide strand is subject to degradation [14]. Guide strand switching is a more recently discussed biological phenomenon which can modulate gene expression because each strand has a unique seed sequence and thus target genes [15]. The mature miRNA guides the RISC to target the 3’ UTR of target mRNA; however, only the 5’ region of the miRNA (~2–8 bp) needs to be entirely complementary to the target mRNA, permitting miRNAs to have multiple targets [16–18]. Emerging evidence suggests miRNAs can also regulate mRNA translation via non-canonical binding to other transcript sites [19], and that miRNAs can vary in length which affects their targeting (i.e. isomiRs) [20, 21]. Additionally, miRNA can bind with non-coding transcripts (e.g. long non-coding RNA “sponges”) [22]. Due to their flexibility in targeting mRNA, miRNAs are estimated to target >60% of human protein coding genes, giving miRNAs strong mechanistic implications in the regulation of skeletal muscle mass [23]. Mature miRNAs generally reside in the cytoplasm of a cell to target mature mRNAs, but they can also be transported back into the nucleus where they may have a role in miRNA biogenesis, stability of nuclear transcripts, splicing, and/or DNA interactions with promoters via AGO proteins [24–28].

Following the discovery that miRNAs are conserved in different species, evidence emerged showing that the expression of some miRNAs are heavily enriched in certain tissues. Among the first examples of a tissue-specific miRNA was miR-1. This miRNA was initially found to be expressed primarily in the human heart [4, 6]. Lagos-Quintana et al. confirmed the finding that some miRNAs are expressed in a tissue-specific fashion and showed that miR-1, −122a, and −124a expression were enriched in striated muscle, liver and brain, respectively [4]. Sempere et al. subsequently identified 30 miRNAs that were enriched or specifically expressed within a particular tissue [29]. These authors provided the first description of striated muscle-specific miR-1, −133a and −206, later designated as myomiRs [29, 30].

The myomiR family has expanded since its original description to include miR-208a, miR-208b, miR-499 and, more recently, miR-486 [31–33]. Northern blot analyses confirmed that these new members of the myomiR family are striated muscle-specific (miR-208a, miR-208b and miR-499), being derived from the intron of different muscle-specific myosin heavy chain genes, or are highly enriched in muscle (miR-486) [31, 33]. Most myomiR family members are expressed in both the heart and skeletal muscle except for miR-208a, which is cardiac-specific, and miR-206, which is skeletal muscle-specific and enriched in oxidative muscles [34] as well as muscle stem cells (satellite cells) [35–37]. It is worth mentioning that myomiRs are sometimes dysregulated in tumor cells of various origin, and may also have roles in the development of non-muscle tissues [38].

Skeletal muscle development in vertebrates is mediated by myogenic progenitor cells that ultimately become bona fide resident muscle stem cells, called satellite cells [39, 40]. These myogenic progenitors fuse together to form multinuclear myotubes during embryonic development (i.e. primary myogenesis). Myocytes (differentiated with one to a few nuclei) and immature myotubes (larger with numerous myonuclei) are the scaffold for muscle fiber hypertrophy during late embryonic then postnatal muscle growth (i.e. secondary myogenesis). Myogenic progenitors ultimately assume a “satellite” position outside the muscle fiber sarcolemma and are characterized by Pax7 expression [41]. Satellite cells continue to fuse to maturing muscle fibers after birth and throughout early postnatal development to drive the establishment of force-producing adult skeletal muscle [42–45]. This process has been reviewed in detail by several others to which we refer the reader [42, 45–47]. In adult skeletal muscle, satellite cells are the engines of muscle fiber regeneration [48–50] and help facilitate reinnervation following injury [51]. Satellite cells also contribute to exercise adaptation in a variety of ways including myonuclear contribution to muscle fibers, influencing coordination via muscle spindles, and regulating ECM deposition [52–56].

MyomiRs have historically been the focus of miRNA research related to the behavior of satellite cells (or myogenic progenitor cells during early development and in culture). This focus on myomiRs is understandable given their specificity to muscle and overall abundance. However, numerous non-myomiRs have emerged as potentially important regulators of myogenic cell function. Furthermore, satellite cells in vivo use miRNAs to communicate to other cell types and coordinate adaptation. These insights have been enabled and expanded on by the growing availability and usage of small RNA-sequencing and other technologies. The purpose of this review is to provide information on how miRNAs affect satellite cell behavior in skeletal muscles (specifically limb and diaphragm) with a specific emphasis on in vivo evidence in adult muscles.

miRNAs Are Essential for Satellite Cell Viability

The most compelling evidence for the essentiality of miRNAs in satellite cell biology comes from experiments involving knockout of the miRNA processing enzyme Dicer. Removing Dicer should deplete cells of miRNA [57–59]. When Dicer is eliminated from the myogenic compartment during in vivo murine embryogenesis, reduced muscle mass and myofiber malformation ensues concomitant with perinatal death [60]. A similar result occurs when Dicer is deleted in Pax3-expressing cells [61] (Figure 1A). PAX3+ cells ultimately become satellite cells in developed muscle [62–65]. Some satellite cells can retain PAX3 expression, specifically in intrafusal muscle fibers (i.e. muscle spindles) [66] and the diaphragm [67, 68], and these cells may compensate for a loss of PAX7+ cells in the diaphragm [69]. The differential expression of PAX3+ satellite cells between limb and diaphragm muscles in adulthood is controlled by miR-206 and alternative polyadenylation of the Pax3 transcript [70–72]. Organism-wide knockout of Dicer in young growing mice (8 weeks old) impairs limb muscle regeneration after severe injury [73]; successful muscle regeneration is entirely dependent on satellite cells [74–77]. Conditional satellite cell-specific knockout of Dicer in young adult mice (8 weeks old) causes satellite cells to activate then undergo apoptosis [78]. Deletion of Dicer in satellite cells also impairs regeneration after injury [78] and prevents these cells from contributing to loading-induced muscle hypertrophy in adult mice (>4 months of age) [36] (Figure 1B).

Figure 1. Consequences of Dicer KO In Vivo.

A) Pax3/7 cell specific Dicer KO results in perinatal death and reduced muscle mass [58], B) Satellite cell specific Dicer KO results in spontaneous activation of satellite cells followed by apoptosis [78], C) Organism-wide Dicer KO results in impaired regeneration following cardiotoxin induced muscle injury of the TA muscle in developing mice [73], D) Adult mice with skeletal-muscle specific Dicer KO display similar muscle alterations as WT mice in models of aging [79], muscle hypertrophy [79], atrophy [79, 80], and physical activity [81]: synergist ablation MOV of the plantaris, denervation, hindlimb unloading, and voluntary wheel running. CTX = Cardiotoxin. TA = Tibialis Anterior. KO = Knockout. MOV = Mechanical Overload.

It is worth mentioning that the result of Dicer deletion differs dramatically between satellite cells and mature muscle fibers. Conditional knockout of Dicer in adult myofibers does not produce a remarkable phenotype or affect muscle adaptation to activity, unloading, or aging, and miRNA/myomiR levels are only reduced ~50% regardless of depletion duration [79–81]. Some evidence suggests that muscle-specific Dicer deletion can affect innervation during development and after injury, but these experiments were limited in scope [82]. Recent evidence posits that certain miRNAs in muscle fibers have shorter half-lives than previously assumed (although not as short as mRNA), so skeletal muscle may possess a unique miRNA biogenesis mechanism that is revealed by Dicer depletion [83]. Collectively, these fundamental studies established how miRNAs can influence satellite cell behavior through different stages of muscle development and during stress (Figure 1C).

Evidence for miRNA Control of Myogenic Cell Function

Early insights on how miRNAs regulate satellite cells came from in vitro studies of myogenic cells and focused primarily on myomiRs. Complimentary to the in vivo studies outlined above, Dicer knockout impairs myoblast survival and differentiation in vitro which coincides with reduced myomiR levels [84, 85]. Similar results were found in primary myoblasts deficient in DGCR8, which is also involved in miRNA processing [86]. miRNA knockout experiments in myogenic cells revealed that miR-1 and miR-206 target Pax3 to regulate MyoD-controlled apoptosis [84]. Goljanek-Whysall et al. further defined and corroborated miR-1/miR-206 targeting of Pax3 to control myogenesis in vivo [87]. This group and others went on to show that miR-1, miR-206, and miR-133 controls C2C12 myoblast behavior via a variety of target gene mechanisms [88–93]. The myomiRs miR-1 and miR-133 are co-transcribed (expanded on below) but may have distinct roles in myogenesis; miR-1 targets HDAC4 to promote myogenesis by reducing Myocyte enhancer factor 2 (MEF2), while miR-133 may enhance myoblast proliferation by targeting serum response factor (SRF) [92]. There is also evidence to suggest that myomiRs are controlled by myogenic transcription factors [94–96], and that miR-1 is regulated by the master cell growth regulator mTOR [97]. A feedback circuit between myogenic/growth genes and myomiRs may therefore direct satellite cell behavior in vivo. Furthermore, miR-1 and miR-206 are induced during myogenic cell differentiation and target the satellite cell identity gene Pax7 to facilitate differentiation [85], suggesting an important role in satellite cells (Figure 2).

Figure 2. miRNA Control Network of Satellite Cell Behavior.

Canonical myomiRs and non-myomiRs dictate myogenic cell fate by targeting numerous genes implicated in myogenesis. miRNAs expression is controlled by MRFs, feedback loops, and other upstream transcriptional regulators. Satellite cells are responsive to their environment in conditions such as hypoxia, oxidative stress, and aging skeletal muscle. Certain miRNAs are sensitive to these conditions and are induced to modulate homeostasis by controlling antioxidant-or senescence-related genes. ARE = Antioxidant Response Element. HRE = Hypoxia Response Element. MRFs = Muscle Regulatory Factors. ROS = Reactive Oxygen Species.

Satellite cells undergo metabolic reprogramming during myogenesis which may, at least in part, be controlled by miRNAs. miR-206 targets glucose-6-phosphate dehydrogenase (G6pdh), the rate limiting step in the pentose phosphate pathway (PPP), in turn suppressing proliferation and promoting differentiation [98]. The PPP is similarly utilized by cancer cells [99] where metabolic substrates are shunted towards aerobic glycolysis to produce electron transport carriers such as NADPH and nucleotide synthesis (i.e. the Warburg effect). The end result is biomass accumulation, clearance of reactive oxygen species, and protein synthesis to support growth [100]. Incidentally, miRNAs may control PPP intermediate production in rapidly hypertrophying skeletal muscle [37, 101]. Myogenic cell quiescence is reliant on low-level glycolysis and the TCA cycle, activation and proliferation is associated with “aerobic glycolysis” and mitochondrial expansion [102–104], while differentiation relies on oxidative phosphorylation [105, 106]. The myomiRs miR-1 and miR-133a are implicated in controlling mitochondrial biogenesis [107]. In contrast to other work mentioned above, genetic knockout of both these myomiRs in myogenic progenitor cells did not affect proliferation or differentiation in vitro despite differentiated cells having severe mitochondrial defects [107]. miR-133 is implicated in satellite cell determination between a myogenic or brown fat adipogenic fate and influences uncoupled respiration during muscle regeneration [108]. Further research is warranted on how miRNAs, and specifically myomiRs, affect satellite cell metabolic preferences and control myogenic cell commitment.

To corroborate in vitro observations, several studies have explored the role of myomiRs in satellite cell behavior in vivo (Figure 3A). Early work from the Olson laboratory showed that organism-wide miR-206 deficiency from birth results in grossly normal muscle [109], but delays regeneration and aggravates muscular dystrophy and amyotrophic lateral sclerosis phenotypes [109, 110]. On balance, miR-206/133b knockout mice appear to have a normal regenerative response during dystrophy disease progression [111]; this is interesting since miR-133b can suppress Pax7 [112]. PAX7 is thought to be necessary for satellite cell function in mature muscle [113–115]. Recent in vivo evidence suggests that knockout of miR-1 (consisting of miR-1a1 and miR-1a2) along with miR-206 does not affect myogenic cell differentiation in vitro despite impaired mitochondrial function, but leads to smaller muscles with reduced function and considerably greater PAX7+ satellite cells during development in vivo [116]. Transfection of C2C12 myoblasts with miR-1 promotes mitochondrial activity by directly enhancing translation of mitochondrial encoded transcripts, namely, Nd1 and Cox1 [117]. Recent work by the Braun laboratory indicates deletion of the miR-1/133/206 family in myogenic progenitors dysregulates the DOK7-CRK-RAC1 cascade, resulting in improper neuromuscular junction development and respiratory function [118]. Deletion of the miR-1/133/206 family is also lethal during embryogenesis despite being dispensable for skeletal muscle development [118]. Other myomiRs, while less well-characterized, may also affect satellite cells in vivo. For instance, germline miR-486 knockout results in muscle malformation and dysfunction [119]. Following glycerol-induced muscle injury, mice with global miR-378 deletion have a greater abundance of satellite cells in the early stages of regeneration, but ultimately have smaller muscles despite an otherwise normal regenerative process [120]. Numerous excellent reviews focus on the role of miRNAs during myogenesis, to which we refer the reader [121–128]. Most studies to date, however, have employed in vitro models, young developing mice, and/or non-cell type specific manipulation of miRNAs in myogenic cells (primary cells or immortalized cell lines). The focus has generally not been on evidence from satellite cells in adult muscle in vivo [122].

Figure 3. In Vivo Evidence of miRNA Control of Satellite Cells During Muscle Regeneration And Intercellular Communication of Satellite Cells.

A) In vivo models of muscle injury implicate satellite cells during regeneration. Organism-wide deletion of miR-206 results in an augmented dystrophic and ALS phenotype [109, 110]. Organism-wide of miR-378 causes enhanced proliferation of satellite cells early in regeneration and smaller myofibers later in regeneration [120]. B) Satellite cells and senescent cells (which can be satellite cells) engage in intercellular communication by secretion of EVs with miRNA cargo, subsequently influencing the transcriptome, proteome, and behavior of recipient cells such as FAPs and muscle fibers. ALS = Amyotrophic Lateral Sclerosis. EVs = Extracellular Vesicles. FAPs = Fibro-Adipogenic Progenitor Cells. KO = Knockout. Mdx = Duchenne Muscular Dystrophy. TA = Tibialis Anterior

miRNA Regulation of Redox Status in Satellite Cells – “redoximiRs” and “hypoxamiRs”

Emerging evidence suggests that certain miRNAs can modulate cellular redox status by targeting antioxidant responsive elements (AREs) and antioxidant genes. These specific miRs are called “redoximiRs” [129–131] (Figure 2). Quiescent satellite cells express high levels of antioxidant genes, including thioredoxin reductase 1 (TXNRD11) and glutathione peroxidase 3 (GPX3). Impaired antioxidant defense or elevated oxidative stress can alter the function of satellite cells [132, 133]. Several studies have thus explored the connection between miRNAs and the oxidative stress response in myoblasts, particularly focusing on the role of nuclear factor (erythroid-derived 2)-like 2 (Nfe2l2, or Nrf2). NRF2, one of the principal regulators of the ARE induces transcriptional activation of several ARE-binding antioxidants, including heme oxygenase-1 (HMOX1) and GPX [134]. HMOX1 overexpression in C2C12 myoblasts reduces the expression of the myomiRs miR-133 and miR-206 and inhibits differentiation [135]. Induction of miR-133 and miR-206 along with HMOX1 overexpression partially restores myoblast differentiation, indicating that the effect of HMOX1 on cell maturation is mediated in part by the inhibition of miR-133 and miR-206 [135]. Similarly, a recent study showed Nrf2 overexpression in C2C12 myoblasts reduced miR-1, miR-133, and miR-206 expression and decreased myoblast differentiation [136]. Satellite cells isolated from elderly human skeletal muscle were also shown to have increased superoxide levels concomitant with elevated miR-1 and miR-133 expression compared to satellite cells from young muscle [137]. These findings suggest a relationship between reactive oxygen species (ROS) production, the NRF2 antioxidant pathway, and myomiR expression in satellite cells. Furthermore, evidence from other cell types indicate that miRNA biogenesis can be modulated by ROS [138]. In endothelial and epithelial cells, ROS can activate the transcription factors p53 and β-catenin to promote miRNA expression via induction of miRNA processing factors such as Drosha [139, 140]. High ROS levels can also increase miRNA levels through inhibition of histone deacetylases (HDACs) that are direct targets of miRNAs [141]. Whether such a ROS-miRNA regulatory network is operative in satellite cells requires further investigation, especially in vivo.

In addition to the potential role for NRF2 in miRNA biogenesis, reciprocal regulation of Nrf2 by miRNAs is also being investigated. In C2C12 myoblasts, miR-128 may negatively regulate the small MAF transcriptional regulator v-Maf avian musculoaponeurotic fibrosarcoma oncogene homolog G (MAFG) [142]. Downregulation of MAFG by miR-128 overexpression reduces basal mRNA levels of NRF2 target antioxidants and attenuates the induction of ARE-dependent NRF2 targets following glutathione depletion [142]. miR-340-5p may be another regulator of Nrf2 in myoblasts. In C2C12 cells treated with miR-340-5p mimics, Nrf2 levels are reduced. Conversely, treatment with miR-340-5p inhibitors increased Nrf2 [143]. Importantly, recent work showed that satellite cells from old mice can maintain function by compensatory upregulation of Nrf2 [144]. Experimental downregulation of miRNAs to restore Nrf2 activation can protect against cerebral and cardiac injury [145, 146]. Therefore, miRNAs that function as redoximiRs to interfere with the oxidative stress-induced induction of the NRF2/ARE pathway may also be potential targets to modulate satellite cell dysfunction during aging.

One of the most well-characterized redox-modulating miRNAs is the hypoxia-responsive “hypoxamiR” miR-210 [147] (Figure 2). Hypoxia inducible factor (HIF) 1α induces miR-210 expression by binding directly to a hypoxia responsive element (HRE) on the miR-210 proximal promoter [147]. Consequences of miR-210 induction include repression of mitochondrial metabolism, which in turn can reduce ROS production and oxidative stress [148]. miR-210 expression is enhanced during myogenic differentiation of C2C12 myoblasts as well as during in vivo regeneration following cardiotoxin injection of mouse tibialis anterior muscles [149]. Moreover, miR-210 inhibition increases superoxide production in C2C12 myoblasts [149]. In both C2C12 myoblasts with impaired mitochondrial function as well as in response to hydrogen peroxide treatment, blocking miR-210 decreases myotube formation and myotube survival [149]. Thus, miR-210 may be potentially involved in cytoprotection during satellite cell differentiation.

Satellite cells also release extracellular vesicles (EVs) containing miRNAs that can be delivered to fibrogenic cells and myofibers (expanded on below) [36, 150]. Oxidative-stress associated increases in redoximiRs in satellite cells has the potential to influence other cell types via EV-mediated communication. To this end, Fulzele et al. treated C2C12 myoblasts with hydrogen peroxide and found that miR-34a levels were greater in EVs isolated from the conditioned medium of hydrogen peroxide-treated myoblasts than untreated myoblasts [151]. Mouse bone marrow mesenchymal cells (BMSCs) were then treated with EVs isolated from the C2C12 medium and found that EVs from hydrogen peroxide-exposed myoblasts led to decreased cell viability in BMSCs compared to EVs from untreated myoblasts [151]. EVs from C2C12 cells overexpressing miR-34a were administered to mice in vivo, which resulted in reduced Sirtuin 1 (SIRT1) expression during ex vivo culture of primary bone marrow cells from these mice [151]. These data provide evidence that satellite cell-derived EVs may transport oxidative stress-associated miRNAs to neighboring cells in skeletal muscle, which may regulate their molecular and biochemical makeup.

Perspectives on Unambiguous Evidence for How miRNAs Control Adult Satellite Cells In Vivo

To understand how myomiRs affect satellite cell behavior in adult skeletal muscle, it is essential to manipulate them in a cell-type specific fashion after developmental growth has ceased. In this way, the contribution of specific myomiRs in satellite cells can be studied in the context of homeostasis, injury, exercise, mechanical overload, aging, and disease without the potential confounding influence of developmental myogenesis. Furthermore, such an approach limits the influence of myomiR expression from the muscle fiber and/or other cell types in the muscle milieu and captures the behavior of satellite cells in their native niche. The niche is known to strongly affect satellite cell behavior [152, 153]. Unfortunately, the complex genomic positioning of myomiRs can make it difficult to design genetic mouse models to test in vivo satellite cell-specific questions in a targeted fashion. For instance, miR-1 is co-transcribed with miR-133 in a bicistronic cluster and has two distinct loci in the mouse, one of which is found in the Mind Bomb 1 (Mib1) gene [154]. Simultaneous knockout of miR-1/miR-133a specifically in satellite cells during development, without affecting Mib1 expression, results in reduced muscle mass and mitochondrial defects in muscle fibers [107]. The effects of knockout of each myomiR in isolation on satellite cell behavior in vivo is unknown. It is generally assumed that myomiRs are expressed exclusively in muscle fibers and/or satellite cells, so broadly manipulating them across cell types is still specific to myogenic cells. This assumption is fair but not always safe. For example, recent evidence suggests miR-206 can be expressed endogenously in muscle fibro/adipogenic progenitor cells under certain conditions [155, 156]. Deletion of myomiRs with similar seed sequences might also exert compensatory responses. Drastic upregulation of miR-206 (similar sequence to miR-1/133a) occurs following germline deletion of miR-1/133a which prevents postnatal lethality [118]. This same group deleted the miR-1/133/206 cluster in myogenic progenitors and found it does not alter the MEF2A/Dlk1/Dio3 axis during embryogenesis. This contrasts with previous findings that germline deletion of miR-1/133a results in upregulation of the MEF2A/Dlk1/Dio3 axis in adult mice [107], suggesting that the function of miR-1/133 may depend on developmental stage [118]. The delivery of antagomirs in skeletal muscle is effective and can be targeted and insightful [157], but more specific genetic approaches are still warranted. To our knowledge, there is no evidence of individual myomiRs being inducibly deleted or overexpressed specifically in satellite cells of adult skeletal muscle in vivo.

Although not a myomiR, Crist et al. genetically overexpressed miR-27 in PAX3+ embryonic cells in vivo which resulted in premature differentiation. These authors also provided initial evidence for a role of miR-27 in satellite cells during in vivo muscle regeneration using an injected antagomir [158]. Furthermore, miR-31 antagomir-mediated knockdown in satellite cells enhanced muscle regeneration in vivo [159]. Using an inducible satellite cell-specific genetic approach, miR-29 knockout impaired satellite cell proliferation and myotube formation after chemical or exercise-induced injury, potentially via targeting Fbn1, Lamc1, Col4a1, Hspg2, and Sparc [160]. This work is important since it unambiguously implicates a single miRNA in satellite cell function in intact muscle under stress. Follow-up in vitro work involving miRNA screening and sequencing confirmed miR-29 as a “highly-active” miRNA in myogenic cells [86]. However, in vivo knockdown of miR-29 alongside let-7, miR-125, miR-199, and miR-221 using antagomirs three days after injury enhanced muscle regeneration via regulation of the focal adhesion complex and extracellular matrix factors [86]. This result may be difficult to reconcile with genetic miR-29 knockout experiments but raises important considerations when studying how miRNAs affect satellite cell behavior: 1) when a miRNA is manipulated (e.g. during quiescence, activation, proliferation, or fusion) can reveal unique functions for that miRNA, 2) complicated regulation of miRNA via a cooperative and integrated network involving numerous miRNAs and targets is likely, and 3) miRNAs beyond canonical myomiRs may have an important role in influencing satellite cell behavior in vivo.

Transcriptome Profiling of Isolated Satellite Cells Reveals Novel miRNAs That May Affect Cell Behavior

Array technology and RNA-sequencing approaches allow for global profiling of miRNAs in myogenic progenitors as they progress through different stages of commitment. The Rando laboratory was among the first to profile miRNAs in purified satellite cells from young adult mice using microarrays. Their analysis revealed that 351 miRNAs were differentially regulated in the transition from satellite cell quiescence to activation [78]. These experiments in combination with satellite cell-specific Dicer knockout experiments identified evolutionarily conserved miR-489 as a crucial regulator of satellite cell quiescence [78]. The repression of miR-489 in satellite cells caused spontaneous activation and proliferation which was in part the result of targeting Dek [78]. Using a similar experimental approach in 12-week-old mice, miR-195/497 was identified as a regulator of satellite cell quiescence in diaphragm muscle by targeting the cell cycle regulators Cdc25 and Ccnd [161]. In human myoblasts, array profiling revealed that miRNA abundance was lower in quiescence [162]. The Tajbakhsh laboratory more recently performed small RNA-sequencing on purified quiescent, activated, and differentiating myogenic progenitor cells from mice. This comprehensive approach identified 209 miRNAs differentially regulated from quiescence to activation, 126 from quiescent to differentiating, and 110 between activated to differentiating [163]. Another laboratory isolated satellite cells from bovine skeletal muscle and identified 72 miRNAs that were differentially expressed between early and late-stage differentiation [164]. miRNAs are often expressed in clusters of ≥2, and sequencing analysis revealed that miR-127/miR-136 and miR-379/miR-410 in the Dlk1/Dio3 locus may be implicated in maintaining satellite cell quiescence [163]. Follow-up work from this same laboratory provided evidence that miR-708, which is enriched in satellite cells, opposes satellite cell migration and maintains quiescence in vivo [165].

The aforementioned profiling studies showed that ubiquitously-expressed miR-16 drops precipitously during myogenic cell differentiation [78, 163]. Furthermore, miRNA profiling of isolated satellite cells during a time course of in vivo muscle regeneration revealed miR-16 as one of the most highly expressed and regulated miRNAs [166]. Low miR-16 levels during development is associated with muscle hypertrophy [167, 168]. Overexpression of miR-16 prevents myogenic cell differentiation and myotube formation in vitro [169], and knockdown enhances myotube formation [169, 170]. Manipulating miR-16 on myofiber-associated satellite cells in vitro via hairpin inhibitors affects their fate [171]. Thus, we recently focused on the role of miR-16 in myogenic cells [172]. Proteomic profiling of C2C12 myoblasts after miR-16 knockdown suggested that miR-16 may influence myogenic cell differentiation through upregulation of EEF1A2, OPA1, and PTEN [172]; all of these factors have defined roles in controlling satellite cell fate [173–180]. Recent advances in single cell and spatial sequencing technology will enable the characterization of small RNAs in individual satellite cells at different phases during adult muscle adaptation [181]. These approaches will surely lead to the discovery of more novel miRNAs that regulate satellite cell fate progression and heterogeneity in vivo.

Another important aspect of miRNA biology is identifying biologically-relevant miRNA targets. Traditionally, defining miRNA interactomes has primarily relied on computational prediction of miRNA binding sites mostly in 3’ UTRs to identify targets [182]. However, these bioinformatic approaches tend to overpredict sites, inaccurately model target site accessibility due to RNA structure/competitive binding, and fail to predict non-canonical binding of miRNAs [182]. Additionally, approaches to validate miRNA targets such as luciferase reporter assay constructs fail to factor the dependence of miRNA functions on stoichiometric capacity to bind to available AGO protein and potential targets, and how changes in miRNA levels may have dramatic effects on the AGO ‘pool’ [183]. To investigate functional relationships between miRNAs and mRNA transcripts, early research focused on mRNA profiling following miRNA manipulation [184]. However, a limitation of this methodology is that it fails to account for indirect consequences of miRNA overexpression and knockdown on the abundance of other miRNAs [185, 186]. To address these limitations, emerging research has focused on using AGO crosslinking immunoprecipitation coupled with high-throughput sequencing (HITS-CLIP) to allow for biochemical determination of AGO:miRNA binding sites on a transcriptome-wide scale [187–189]. Moreover, recent advances using enhanced CLIP (eCLIP) allows increased efficiency and improved specificity in the discovery of miRNA binding sites [190]. CLIP analysis of the RNA-binding protein AGO has led to identification of miRNA targets in C. elegans [191] and subsequently in mouse and human tissues and cells [192–197]. AGO CLIP sequencing was specifically applied to C2C12 myoblasts and myotubes, which led to the identification of miRNA targeting to mitochondrial transcripts [117]. Furthermore, AGO CLIP sequencing during C2C12 differentiation identified myomiR enrichment and targeting of JNK/MAPK signaling [198]. Finally, in mouse skeletal muscle tissue, targeted chimeric AGO-miR-486 eCLIP sequencing was applied to identify miR-486 targets associated with dystrophic muscle [199]. Future application of HITS-CLIP methodologies to decipher transcriptome-wide maps of miRNA binding sites in skeletal muscle tissue will be critical to define regulatory targets of important skeletal muscle and satellite cell-related miRNAs.

The Contributions of miRNA to Aging and Senescence in Satellite Cells

Cellular senescence is now a recognised mechanism contributing to deterioration of tissue function during aging and disease, including in skeletal muscle [200–202]. Although a consensus on senescence in postmitotic tissues, such as muscle, is yet to be established [203], there is evidence that stem cells, including satellite cells, undergo senescence [204–206]. Single cell and single nuclei sequencing studies demonstrate that senescence occurs in multiple cell types of skeletal muscle during aging, including satellite cells [207–209]. Senescent satellite cells could contribute to muscle loss through defective regeneration and/or the senescence-associated secretory phenotype, and dysregulation of signalling [210, 211]. It is noteworthy that satellite cell senescence has been shown to delay muscle regeneration in mice during aging [211], but other literature suggests that satellite cells from old mice and humans retain their intrinsic regenerative capacity [212–214]. Post-developmental lifelong depletion of satellite cells does not exacerbate sarcopenia [215, 216], but elicits excess extracellular matrix accumulation [216] and impaired exercise adaptation concomitant with muscle spindle cell dysregulation [217]. It is possible that senescence of other muscle resident cells, e.g. fibroadipogenic progenitors or immune cells, has a higher contribution than satellite cells to muscle wasting during aging [208, 218, 219].

The importance of miRNAs in regulating senescence was first demonstrated by Dicer knock-out induction of senescence [220]. Moreover, multiple signalling pathways associated with cellular senescence are regulated by microRNAs [221]. Soriano et al. demonstrated an important role of miR-143 in regulating myogenic progenitor cell senescence in vitro [222]. Downregulation of miR-143 in myogenic cells from old mice was associated with increased cell viability. miR-24 was also shown by the same group to regulate redox homeostasis through Prdx6 gene expression regulation in myogenic progenitors from aged mouse and human muscle [223]. miR-29 is increased in the skeletal muscle during aging and is characterized as a senescence-related miRNA [201]. miR-29 enhances cellular senescence by targeting insulin-like growth factor-1 (Igf1), p85, and B-myb. The widely-available C2C12 myogenic cell line is sometimes used to study senescence [224–229]. It is important to note that studying senescence in the C2C12 line can be complicated because this line does not contain the Cdkn2a/Ink4a locus [230]; this locus codes for p16, a key component of senescence regulation [231].

In addition to the intracellular role of microRNAs in regulating signalling pathways, miRNAs have also been shown to be released from cells, including via EVs such as exosomes, and can play a role in cell-to-cell communication (discussed below) [232]. These exosome miRNAs released by either senescence satellite cells or other muscle resident cells could participate in propagation of cellular senescence to neighbouring cells, similar to the SASP, or be a part of mechanisms maintaining function of skeletal muscle fibers [233, 234] (Figure 3B). This hypothesis is supported by evidence from exosome miRNA contributing to muscle wasting in the context of cancer cachexia [235]. Furthermore, prematurely senescent human myoblasts have enhanced EV release, and these EVs induce senescence markers and reduce proliferation of endothelial cells in vitro [236]. As the importance of senescence in skeletal muscle wasting continues to be established, the role of miRNAs in regulating senescence, either intracellularly or though intercellular communication, will begin to be unravelled.

Satellite Cell Communication via miRNAs During Adult Muscle Adaptation

Most studies on miRNAs in myogenic cells focus on how miRNAs regulate cell commitment and fate via mRNA 3’ UTR targeting. Beyond these intracellular functions, however, it is generally accepted that miRNAs can be transported intercellularly via various mechanisms. The most well-studied mechanism of miRNA transport between cells is via EVs [237, 238]. Initial work showed that myogenic cells in culture release EVs containing miRNAs that may have a role in intercellular communication during myogenesis [239, 240] (Figure 3B).

In 2017, evidence emerged for satellite cell communication to other cell types in adult muscle via miRNAs in EVs [36]. Myogenic progenitors are highly enriched with miR-206 as are their EVs [35, 53]. During a hypertrophic stimulus in skeletal muscle of adult mice (>4 months old), activated satellite cells release EVs containing miR-206 that is delivered to fibrogenic cells. These miR-206 enriched EVs regulate extracellular matrix deposition from fibrogenic cells via a central regulator of collagen deposition, Rrbp1 [36]. Additional work has since expanded the scope of satellite cell-mediated EV communication via miRNAs from fibrogenic cells to other cell types during adult muscle adaptation to loading [241]. Furthermore, satellite cells may deliver miRNAs to muscle fibers prior to fusion to control Mmp9 levels in the muscle fiber during hypertrophy [35]; these experiments were conducted in adult mice (>4 months of age), and supporting evidence in this study comes from an in vivo satellite cell depletion model (Pax7-DTA) [35, 74]. Mmp9 is robustly induced in myonuclei during mechanical overload [242] and is a pleiotropic factor that can affect ECM turnover [243–245] as well as muscle fiber growth [244, 246]. Satellite cells influencing myofiber Mmp9 levels via delivered miRNAs is noteworthy since muscle fiber nuclei (myonuclei) contribute to myofiber growth and ECM remodeling during adaptation [55, 242, 247]. Certain microRNAs are also secreted in EVs during disease states such as mdx, termed “dystromiRs”, and can reflect the state of muscle regeneration, which is largely mediated by satellite cells [248]. These dystromiRs may also have distinct release/degradation kinetics and changes in EV-packaging over time which may further impact their ability to convey information between cells [249]. MicroRNAs are thought to contribute to nearly every aspect of myogenic cell behavior including maintenance of quiescence [71, 78, 159, 162], activation [70, 171], proliferation [85, 250], self-renewal [251], migration [165], differentiation [61, 85, 252], and possibly fusion [253, 254]. Satellite cells can now be considered cellular coordinators during adult muscle adaptation via miRNA delivery to other cell types, and conceivably vice versa, to influence recipient cell behavior [53].

Conclusion

Satellite cell and myogenic progenitor behavior is tightly controlled by a coordinated network of miRNAs. Deletion of miRNAs via Dicer in adult skeletal muscle satellite cells results in apoptosis and is fatal in the perinatal period. To the best of our knowledge, there have not been any published studies involving the genetic modulation of individual miRNAs in adult muscle satellite cells in vivo. The bulk of our understanding of miRNA control of satellite cells has largely come from in vitro experiments or usage of non-cell type-specific antago-miRs in vivo. The widespread usage of these approach is largely a technical barrier arising from complex genomic positions of miRNAs (i.e., clusters or intragenic). Apparent redundancy in miRNA targets and cooperative miRNA networks provides an additional challenge to studying their effects on satellite cells. To elucidate how miRNAs control satellite cells with precision in vivo will require the development of new genetic models capable of conditional and inducible deletion or overexpression of individual and/or clusters of miRNAs exclusively in satellite cells, combined with transcriptome and proteome profiling. Closing this fundamental gap in our understanding of resident muscle stem cells could ultimately lead to targeted miRNA-based therapies to improve human muscle quality and healthspan.

• MicroRNAs are essential for the viability and function of myogenic cells

• Strong in vitro evidence for how specific microRNAs control myogenic cell behavior

• Satellite cells communicate in vivo with other cells via exosome-bound microRNAs

• Need for novel tools to study microRNA control of adult muscle satellite cells