Maintenance of the MuSC pool is essential for regeneration of the skeletal muscle

MuSCs are a rare population of cells in the skeletal muscle, making up less than 5% of all nuclei in the skeletal muscle. The importance of the integrity of the MuSC pool was elegantly illustrated by genetic approaches, whereby the Pax7-expressing MuSC pool was conditionally ablated by the expression of diphtheria toxin (DTA) after tamoxifen administration to Pax7CreERT2/+; R26DTA mice, with consequent collapse of the skeletal muscle regeneration after acute injury [22,23,24]. These studies demonstrate that the skeletal muscle does not regenerate without MuSCs and also confirm that MuSCs are the only cell population with myogenic potential that regenerates muscle in vivo [25]. With only a single and rare cellular source fueling lifelong regeneration of the skeletal muscle, it is critical that this population is maintained.

A decline in skeletal muscle regeneration occurs when the integrity of the MuSC pool is compromised

The importance of maintaining the MuSC pool is also illustrated in muscle disease, the progression of which often coincides with a loss of MuSC numbers and function. DMD is a devastating X-linked skeletal muscle degenerative disease affecting approximately 1 in 5000 boys [26] with 100% mortality by early adulthood. The disease is caused by mutations in the DMD gene, which lead to impaired synthesis of full-length dystrophin protein, the absence of which causes myofibre fragility. In DMD patients, cycles of muscle degeneration and regeneration lead to the exhaustion of the MuSC pool, in part because chronically activated MuSCs exhibit severe proliferation defects and undergo premature senescence [19]. Dystrophin protein is now understood to be expressed in activated MuSCs, where it regulates MuSC polarity and asymmetric cell divisions that are required to maintain the MuSC pool [27, 28]. In the Dmdmdx mouse model of DMD, dystrophin deficiency also leads to chronic degeneration of skeletal muscle. However, the phenotype is mild and Dmdmdx mice have a normal lifespan, in part due to greater proliferation capacity of mouse MuSCs that fuels regeneration of the muscle [29] and likely in part due to their shorter (27 months) lifespan. In addition to DMD, mutations in PAX7 have also been linked to the pathology of a new myopathy with variable severity in humans. The lack of PAX7-expressing MuSCs in the human muscle may lead in part to muscle atrophy, hypotonia, scoliosis, and mild dysmorphic facial features that are present in individuals with these mutations [30].

The progression of the muscle wasting associated with sarcopenia is also accompanied by a decrease in the fitness and numbers of MuSCs [31], with consequent loss of skeletal muscle regeneration [32, 33]. Compared to MuSCs isolated from young adult muscle, MuSCs isolated from old mice are prone to apoptosis and senescence when placed in culture [32, 34]. Aged MuSCs appear to activate a number of stress response pathways associated with p38MAPK [34,35,36], which is activated in response to a variety of cell stress and inflammation [37], Jak-Stat3 [38, 39], a pro-survival pathway activated in response to stress [40] and p16INK4a [32], which negatively regulates the cell cycle in response to cell stress. In young adult MuSCs, p16INK4A expression is epigenetically silenced. In geriatric mice, ubiquitination of H2A leads to permissive chromatin marks that enable p16INK4A expression. MuSCs with elevated p16INK4A expression do not activate and transit into the G1 phase, but instead irreversibly become senescent [32], with consequent depletion of the functional MuSC pool and impaired muscle regeneration.

Altogether, genetic ablation strategies and myopathies that are characterised by a loss of MuSCs number and function illustrate the importance of maintaining the MuSC pool to fuel lifelong regeneration of the muscle. The notion that the tissue microenvironment, or niche, protects MuSCs from cellular and environmental stress is challenged by evidence that quiescent MuSCs actively initiate multiple stress response pathways (Fig. 1). Next, we review the cellular responses to stress utilised by MuSCs and further discuss the fate of MuSCs when these stress response pathways are compromised.

MuSC adaptations to cellular stressMuSC responses to stress by reprogramming mRNA translation

Regulation of mRNA translation contributes to many aspects of cell physiology, including cell growth, proliferation, differentiation, and cell survival when exposed to stress. The coordinated regulation of transcription and translation provides optimal levels of required proteins that is balanced against the energy expenditure of protein synthesis [41, 42]. Under conditions of stress, the survival of cells depends on the rapid reprogramming of translation to selectively translate mRNAs required to initiate a stress response, while globally repressing mRNA translation to reduce the energy requirements of protein synthesis [43, 44].

The arrest of translation initiation, the rate limiting step of protein synthesis, is a major hallmark of stress-induced translational control. Two translation initiation factors play central roles in the regulation of mRNA translation in response to stress. These are eukaryotic initiation factor 2 (eIF2), which is central to the ISR [45] (Fig. 2) and eukaryotic initiation factor 4E (eIF4E), which is a key component of a stress response regulated by the mechanistic target of rapamycin complex 1 (mTORC1) signalling pathway (Fig. 3) via eIF4E binding proteins (4E-BPs) [46].

Fig. 3

The mTORC1 pathway in stress. The presence of growth factors (grey, green, purple circles), abundant amino acids (brown diamonds), and cellular energy (light brown mitochondria) positively regulated mTORC1 to regulate cell growth pathways through the phosphorylation of S6K1 kinase (green arrow), and cell proliferation pathways (red arrow) through the phosphorylation of eIF4E binding protein (4E-BP). Phosphorylated 4E-BP no longer competes for eIF4E binding, permitted eIF4E to initiate cap-dependent translation. eIF4E is the cap binding protein that functions within the eIF4F tertiary complex (purple) along with eIF4G and eIF4A (not shown). Cellular and environmental stress activates the TSC1/TSC2 complex to inhibit mTORC1 signalling, leading to the repression of these pathways, with consequent decrease in cell growth and proliferation

The integrated stress response

In response to a broad range of cellular stress, eukaryotes activate the ISR [8, 45] (Fig. 2). The central event in this pathway is the phosphorylation of eukaryotic initiation factor 2α (P-eIF2α) by one of four members of the eIF2α kinase family. General control nonderepressible 2 (GCN2) responds to amino acid starvation [8], protein kinase R (PKR) responds to the presence of viral double-stranded RNA [47], heme-regulated inhibitor (HRI) responds to the absence of heme in erythroid cells [48, 49], and PKR-like endoplasmic reticulum kinase (PERK) is activated in response to endoplasmic reticulum stress [50]. Additional environmental stresses that induce eIF2α phosphorylation for which the specific kinase remains unknown are exposure to arsenite, osmotic stress, heat shock, and nutrient starvation (Fig. 2).

The eIF2 complex (eIF2α, eIF2β, and eIF2γ) is a trimeric protein complex that is essential for protein synthesis and responsible for recycling the methionine loaded tRNA (Met-tRNA) initiation complex to the 40S ribosomal subunit to form the 43S preinitiation complex. P-eIF2α turns eIF2 into a competitive inhibitor of the guanine nucleotide exchange factor eIF2B, to prevent recycling of the eIF2-GTP-initiatior methionyl tRNA ternary complex needed to initiate translation [51]. The resultant block in translation initiation has two important consequences to initiate a stress response (Fig. 2). First, translation reprogramming occurs in the cell whereby a global arrest in translation of mRNA is countered by selective translation of specific mRNAs required for the initiation of a stress response. Selective mRNA translation is mediated in part by inhibitory upstream ORFs (uORFs) in the 5’UTRs of transcripts, exemplified by transcripts for activating transcription factor 4 (Atf4) [52] (Fig. 2). P-eIF2α-dependent readthrough of inhibitory uORFs in the 5’UTR of Atf4 enables the initiation of translation at the main ORF encoding for ATF4, and ATF4 in turn activates the expression of genes required for cell recovery in response to stress [53]. Although the ISR is a pro-survival pathway, exposure to severe stress or prolonged stress leads to the induction of cell death pathways [54,55,56]. Cells that have genetic modifications to remove the phosphorylated serine residue at position 51 of eIF2α (S51A) are unable to cope with acute stress. Moreover, the importance of eIF2α phosphorylation in mammals is illustrated by perinatal lethality in eIF2αS51A/S51A mice [57].

Second, P-eIF2α leads to a pool of mRNAs paused at the initiation step of translation, which through liquid-liquid phase separation seed the assembly of stress granules, membrane-less organelles of ribonucleoprotein complexes composed of RNA binding proteins and stalled mRNAs [48, 49] (Fig. 2). When the eIF2-GTP-initiatior methionyl tRNA ternary complexes are reduced, RNA binding proteins TIA1 and TIAR promote the assembly of non-canonical preinitiation complexes that lack the methionine loaded tRNA. TIA1 and TIAR dynamically triage translationally incompetent mRNAs into stress granules [58]. Despite that stress granule composition, assembly and disassembly have been studied for many years, their true function in the cell remains unclear. They presumably serve as sites of mRNA triage, help the cell cope with stress, and possibly facilitate the recovery and rapid reinitiation of translation after stress removal and stress granule disassembly [43, 59].

The phosphorylation of eIF2α is a translational control mechanism regulating MuSC quiescence and self-renewal

Quiescent MuSCs maintain low levels of protein synthesis, by PERK phosphorylation of eIF2α. The activity of PERK and P-eIF2α are both essential for MuSC quiescence and self-renewal [7]. Upon MuSC activation, eIF2α is rapidly dephosphorylated, coincident with translation and rapid accumulation of myogenic regulatory factors MYF5 and MYOD. When cultured ex vivo, rare MuSCs expressing only PAX7 maintain P-eIF2α, while the bulk of proliferating MuSCs that activate the myogenic programme dephosphorylate eIF2α. Like all cells, MuSCs require P-eIF2α to initiate a pro-survival stress response when challenged with an acute stress, for example brief exposure to ER stress inducer thapsigargin. However, MuSCs do not require P-eIF2α for cell survival under physiological conditions, nor is P-eIF2α required for MuSC survival during a regenerative response after acute injury. Instead, MuSCs that are unable to phosphorylate eIF2α are prone to spurious activation, proliferation, and contribution to new or existing myofibres in vivo [7].

P-eIF2α is also required for the assembly of RNA granules within the cytoplasm of quiescent MuSCs [7] (Fig. 1). These RNA granules are similar to size and RNA binding protein composition to stress granules, marked by RNA binding proteins DDX6, TIAR, FMRP, and GW182 [2, 3, 6]. They do not contain mRNA decapping enzyme DCP1, which is a marker of P bodies that are considered sites of mRNA decay. Instead, DCP1-positive P-bodies predominate in activated MuSCs [6]. Quiescent MuSC RNA granules are thought to be sites of storage for transcripts required for activation of the myogenic programme and proliferation. For example, Myf5 transcripts colocalize to RNA granules and can be immunoprecipitated with antibodies against DDX6 [2]. Upon MuSC activation, the dissolution of RNA granules and rapid accumulation of MYF5 protein are amongst the earliest markers of MuSC activation, which coincides with reengagement of Myf5 mRNA with translating ribosomes and rapid accumulation of MYF5 protein. Therefore, RNA granules possibly ‘prime’ quiescent MuSCs for rapid activation by their disassembly and rapid initiation of Myf5 mRNA translation [2, 7].

MuSCs appear not only to activate the P-eIF2α stress response pathway to maintain quiescence and self-renewal, but also thrive under ex vivo conditions that promote eIF2α phosphorylation [7, 60]. Under normal culture conditions, a subset of PAX7-expressing MuSCs maintain P-eIF2α, while activated MuSCs that express MYOD dephosphorylate eIF2α. Fresh isolated MuSCs that are cultured in the presence of the eIF2α phosphatase inhibitor sal003 expand ex vivo as a population of PAX7(+), MYOD(-) cells. These cells retain their stem cell properties to regenerate muscle and self-renew, illustrated by their engraftment into the Dmdmdx preclinical mouse model of Duchenne muscular dystrophy [7].

How MuSCs expand under culture conditions that promote the eIF2α phosphorylation stress response and lower global rates of protein synthesis is an interesting paradox that is potentially resolved by translational reprogramming. Culture of MuSCs in the presence of sal003 revealed hundreds of genes that are upregulated at the level of protein, without a corresponding increase in mRNA levels [61], suggesting post-transcriptional regulation. The most significantly represented class of genes were for those involved in spindle assembly, suggesting that Pax7-expressing MuSCs use eIF2α phosphorylation to maintain the fidelity of cell division. For example, P-eIF2α enables the translation of an mRNA for the mitotic spindle assembly gene transforming acidic coiled coil protein 3 (Tacc3) by virtue of inhibitory uORFs present in the 5’ untranslated region (5’UTR) of Tacc3 mRNA. In the absence of Tacc3, MuSCs expand poorly, with consequent depletion of the integrity of the MuSC pool and compromised regeneration of the skeletal muscle after acute injury [61].

There are a number of remaining questions related to the eIF2α pathway in MuSCs. What are the identity and fate of mRNAs that localise to P-eIF2α dependent RNA granules? Conversely, which mRNAs are translated in a P-eIF2α-dependent manner in quiescent MuSCs? These questions are potentially addressed with next generation RNA-seq technologies compatible with low amounts of mRNA isolated from ribosomes [60] or new strategies to isolate and determine the RNA component of RNA granules [61, 62]. Another important question is to what extent is P-eIF2α dependent changes in mRNA translation modified in muscle disease. While in normal healthy muscle, MuSC quiescence is maintained by PERK phosphorylation of eIF2α, the extent to which eIF2α phosphorylation is modified by kinases responding to other forms of cellular stress, for example PKR or GCN2, within the context of aging or chronically degenerating muscle, remains unclear.

MuSC quiescence is mediated in part by the stress response pathway regulated by mTORC1

Mechanistic target of rapamycin (mTOR) is a serine threonine kinase belonging to the family of phosphatidylinositol 3-kinase (PI3K)–related kinase (PIKKs) and is a main activator of the cellular biosynthesis machinery needed for increase cell growth and proliferation [62]. Mechanistically, mTOR functions in multiprotein complexes mTORC1 (Fig. 3) and mTORC2 and is activated by growth factors, nutrients, and energy [63]. The two most extensively studied downstream effectors of mTORC1 signalling are p70 S6 kinase (p70S6K; RPS6K1/2) and the eIF4E binding protein 1/2/3 family (4E-BP) (Fig. 3). p70S6K regulates cell growth by phosphorylation of ribosome protein S6 to increase rates of ribosome biogenesis and protein synthesis [63, 64] (Fig. 3). Phosphorylation of 4E-BPs regulates cell proliferation by disrupting their inhibition of eIF4E to enable 7-methylguanosine 5-triphosphate (m7GTP) cap-dependent translation of mRNAs encoding cell cycle regulators [46] (Fig. 3). Translational reprogramming of mRNA is also a feature of the mTORC1 pathway, since 4E-BPs regulate the translation of specific mRNAs that have established 5’ terminal oligopyrmidine (TOP) motifs [65].

The activity of mTOR is sensitive to complex signalling networks, including those that are activated in response to cell stress. The bulk of mTORC1 inhibition is channelled through the tuberous sclerosis (TSC) proteins TSC1 and TSC2 [62] (Fig. 3), which together serve to promote inactivating GTP hydrolysis of components of mTORC1. Cellular stresses that activate TSC1/TSC2 include growth factor deficiencies, low cellular energy, hypoxia, ROS, and DNA damage [62]. The resultant decrease in p70S6K activity leads to decreased ribosome biogenesis and reduced cell growth. The decrease in phosphorylated 4E-BP enables 4E-BP binding to eIF4E, leading to inhibition of cap dependent mRNA translation, and reduced cell proliferation (Fig. 3).

Quiescent G0 MuSCs activate mTORC1 signalling to transition to Galert

An important role for mTORC1 signalling has been elucidated in the early activation of MuSCs, termed Galert [16]. Tissue injury at distal sites leads to the accumulation and circulation of growth factors like hepatocyte growth factor (HGF) that activate mTORC1 to increase rates of protein synthesis [66]. The Galert phase is characterised by increased mitochondria, more ATP and an increase in cell size, but not by an increase in cell proliferation. Mechanistically, the Galert phenotype, characterised by an increase in MuSC size, is associated with an increase in S6 kinase phosphorylation. Moreover, genetic inactivation of Tsc1 specifically in Pax7-expressing MuSCs leads to acquisition of the Galert phenotype independent of distal injury, suggesting that MuSC quiescence is also regulated by cellular or environmental stresses that together inhibit mTORC1 signalling via TSC1 (Fig. 1). Altogether, the inhibition of mTORC1 signalling by TSC1 maintains MuSC quiescence, while the activation of mTORC1 by circulating growth factors like HGF is an early stage of MuSC activation [16, 66].

How the cell growth arm of the mTOR pathway, regulated by S6 phosphorylation, is specifically activated in Galert, while the cell proliferation arm of the mTOR pathway, potentially regulated by 4E-BP, remains resistant, is unknown. Positive mTORC1 regulation of cell proliferation potentially becomes the dominant response in activated MuSCs, since inactivation of Raptor, a specific component of the mTORC1 signalling pathway, limits MuSC proliferation, with consequent perturbation in muscle differentiation and regeneration [67]. Lastly, mTORC2 is a second mTOR complex that responds to growth factors to regulate cell proliferation, but the study of mTORC2 has lagged behind mTORC1 and has also not yet been investigated within the context of MuSC quiescence and activation.

The DNA damage response

When challenged with irradiation induced genotoxic stress, MuSCs resist apoptosis compared to non-myogenic cells and differentiated muscle present in the skeletal muscle. Mechanistically, quiescent MuSCs more accurately and efficiently repair DNA double-stranded breaks (DSBs) than activated MuSCs and committed progeny. Resistance to DNA damage is mediated in part due to the activity of DNA-PKcs [68] (Fig. 1), which is a central effector of the DNA damage response (DDR), a stress response pathway that senses DNA damage and replication stress to activate a protective response. DNA-PKcs is another member of the PIKK kinase family that function to phosphorylate a large number of substrates that are required for efficient and accurate DNA repair and also coordinate DNA repair with stalls on transcription, replication, and cell proliferation. While all quiescent MuSCs exhibit increased DNA damage repair compared to their activated and differentiated progeny, the subset of Pax3-expressing MuSCs is particularly resistant to genotoxic stress (Fig. 1). These cells have reduced levels of ROS, exhibit low levels gamma histone family member X (γH2Ax) foci, and reduced DNA damage in response to irradiation than MuSCs that only express Pax7. These cells are rare, exhibit limited contribution to normal regeneration and repair, but exhibit stress tolerance and are capable of clonal expansion and contribution to repair under stress [69].

Environmental stress

Quiescent MuSCs have developed resistance to xenobiotics, genotoxics, and oxidative stress. Toxic substances may be pumped out of the quiescent MuSC by virtue of high-level expression of genes for efflux channels Abcb1a, Abca5, and Abcc9. Moreover, quiescent MuSCs may have developed strategies to solubilise toxic substances. The aryl hydrocarbon receptor (Ahr) is also expressed at high levels in MuSCs, where it plays a role to sense toxic molecules like dioxin derivatives or polycyclic aromatic hydrocarbons [