Epigenetic control of skeletal muscle regeneration: integrating genetic determinants and environmental changes

Abstract

During embryonic development, pluripotent cells are genetically committed to specific lineages by the expression of cell type-specific transcriptional activators that direct the formation of specialized tissues and organs in response to developmental cues. Chromatin-modifying proteins are emerging as essential components of the epigenetic machinery, which establishes the nuclear landscape that ultimately determines the final identity and functional specialization of adult cells. Recent evidence has revealed that discrete populations of adult cells can retain the ability to adopt alternative cell fates in response to environmental cues. These cells include conventional adult stem cells and a still poorly defined collection of cell types endowed with facultative phenotype and functional plasticity. In physiological conditions or adaptive states, these cells cooperate to support tissue and organ homeostasis, and to promote growth or compensatory regeneration. However, during chronic diseases and aging these cells can adopt a pathological phenotype and mediate maladaptive responses, such as formation of fibrotic scars and fat deposition that progressively replaces structural and functional units of tissues and organs. The molecular determinants of these phenotypic transitions are only emerging from recent studies that reveal how dynamic chromatin states can generate flexible epigenetic landscapes, which confer on cells the ability to retain partial pluripotency and adapt to environmental changes. This review summarize our current knowledge on the role of the epigenetic machinery as a “filter” between genetic commitment and environmental signals in cell types that can alternatively promote skeletal muscle regeneration or fibro-adipogenic degeneration.

Introduction

During embryonic development pluripotent cells progressively restrict their ability to adopt multiple lineages to ultimately give rise to specialized cell types. This process is spatially and temporally regulated, and is orchestrated by diffusible signals, transcription factors and chromatin modifiers, thereby revealing a complex interplay between environmental, genetic and epigenetic determinants of lineage commitment and formation of differentiated tissues and organs (Palacios and Puri 2006; Ho and Crabtree 2010). The identification in adult organisms of cell types that retain the ability to adopt alternative fates in response to environmental signals supports the concept that terminally differentiated tissues and organs are more dynamic than previously predicted, and that processes highly reminiscent of development can be resumed in certain conditions, such as tissue regeneration. For instance, the regeneration potential of skeletal muscles appears to rely on multiple cell types that, upon exposure to specific cues, cooperate to generate new myofibers (Brack and Rando 2012; Wang and Rudnicki 2012). This potential implicates that different cell types exposed to common regeneration cues coordinately respond with the activation of cell type-specific programs. In particular conditions, such as chronic diseases and aging, the ability of these cells to support the regenerative response declines, leading to maladaptive responses – e.g. formation of fibrotic scars and fatty infiltration. While the knowledge on the cellular and molecular determinants of such a switch is still incomplete, it appears that reciprocal interactions between genetic and environmental factors might influence the epigenetic landscape that ultimately determines the functional phenotype of cell types that have retained a partial plasticity.

This review summarizes the current knowledge on the epigenetic mechanism that controls cell fate determination of muscle resident cells. In particular, skeletal muscle will be adopted as paradigm to discuss the importance of the chromatin landscape in determining cell fate decisions and commitment to alternative lineages during post-natal life in physiological and pathological conditions. We will focus on the epigenetic control of lineage identity and functional phenotype of adult muscle stem cells (MuSCs - otherwise indicated as satellite cells)(Mauro 1961) and of heterogenic populations of muscle resident/interstitial cells, which can adopt alternative phenotypes in response to environmental signals and influence directly and indirectly the repair ability skeletal muscles (Joe et al. 2010; Uezumi et al. 2010; Gussoni et al. 1999; Qu et al. 1998; L De Angelis et al. 1999; a Dellavalle et al. 2011; Mitchell et al. 2010). Overall, understanding the mechanism that regulates lineage commitment in the adult tissues is a key issue that could open the way to the therapeutic modulation of specific pluripotent cell populations to treat chronic degenerative muscle disorders.

Transcriptional control in muscle stem cells during development and adult life

The genetic circuitry that regulates muscle formation has been extensively studied. Developmental signals and transcription factors constitute the core network that directs the differentiation of pluripotent progenitors into cellular components of specialized tissues and organs during embryonic development and in postnatal-life. For instance, during embryogenesis muscle progenitors are specified by the sequential expression of a network of transcription factors comprising Six1, Six4, Pax3 and Pax7, and the basic helix-loop-helix (bHLH) myogenic activators MyoD, Myf5, MRF4 and myogenin, also indicated as muscle regulatory factors (MRFs) (reviewed in (Guasconi & P L Puri 2009; Thomas Braun & Gautel 2011). Pax3 and Pax7 are both expressed in the dermomyotome, with Pax3 playing a predominant role during embryonic myogenesis (Bober et al. 1994; Daston et al. 1996; Tremblay et al. 1998) and Pax7 being mostly required for postnatal myogenesis (P Seale et al. 2000; Montarras et al. 2005; Oustanina et al. 2004). Once MRFs are expressed, they activate and sustain the skeletal myogenic program in cooperation with MEF2 family members (reviewed in Puri and Sartorelli 2000). MRFs show the unique ability to initiate myogenesis and “convert” non muscle cells to the myogenic lineage (Davis et al. 1987). Knockout studies have placed MyoD and Myf5 hierarchically upstream of MRF4 and Myogenin, with the first two MRFs specifying the lineage of two distinct populations of skeletal muscle progenitors (Haldar et al. 2008; Gensch et al. 2008), while the others execute the terminal differentiation program (reviewed in (H. H. Arnold & T Braun 1996; Giulio Cossu & Borello 1999). Functional studies have revealed that MyoD and Myf5 are endowed with the unique ability to remodel the chromatin at previously silent muscle loci (Gerber et al. 1997), whereas Myogenin supports high rate of muscle gene transcription (Ohkawa et al. 2006). This model is consistent with a multistep process where progression towards myogenic lineage restriction can be achieved using parallel and distinct nodes. The plasticity of the circuitry ensures that loss of a single connection is not sufficient to compromise the process. This regulatory axis is essential to establish the myogenic identity and supports the differentiation ability of muscle progenitors during development and in adult life.

Epigenetic regulation of cell fate determination and differentiation of adult skeletal muscle progenitors

The epigenetic circuitry that determines the identity of pluripotent-derived cells, including adult muscle progenitors, is currently being intensely investigated. Specific chromatin states and epigenetic events are required to establish and maintain the myogenic identity in quiescent MuSC, and to enable proper response to external cues once MuSCs are activated and exposed to the regenerative environment (reviewed in (V Sartorelli & A. Juan 2011). MuSCs pose technical challenges for global genome analysis due to the limited number of available cells as well as because of their difficult synchronization in an unambiguous functional state – e.g. uniform populations of quiescent, proliferating and differentiated cells. Thus, most of the information has been derived from studies of next generation sequencing (RNA-seq and ChIP-seq) in skeletal muscle cell lines, such as C2C12 cells.

In general the epigenetic regulation of gene expression is operated by highly interconnected events, such as post-translational modifications of histones, chromatin remodeling and nucleosome patterning, DNA methylation and a network of non-coding RNA (reviewed in Kouzarides 2007; Goldberg et al. 2007; Jones 2012; Iyer 2012; Castel and Martienssen 2013). There are several classes of post-translational histone modifications (e.g. phosphorylation, acetylation, methylation and ubiquitination) that cooperatively affect chromatin accessibility and therefore regulate the recruitment of DNA-interacting proteins (Kouzarides 2007; Jenuwein & C D Allis 2001; Sadeh & C David Allis 2011). While a detailed description of histone modifications is beyond the scope of this review, for the purpose of the topics discussed here it is opportune to mention that methylation of lysines or arginines, in combination with lysine acetylation, establishes specific chromatin signatures predictive of transcriptional activation or repression at enhancers and promoter regions. For instance, lysine methylation at specific residues is catalyzed by distinct families of methyltransferases and has been linked to formation of euchromatin and transcriptional activation (H3K4 H3K36 and H3K79 methylation) or formation of heterochromatin and transcriptional repression (H3K9, H3K27 and H4K20 methylation) (Schuettengruber et al. 2011; Morey & Helin 2010).

Arginine methylation is catalyzed by protein arginine methyltransferases (PRMTs) and adds a further level of complexity, by regulating docking of key co-transcriptional effectors (Lorenzo and Bedford 2011). Histone acetylation has invariably been linked to transcriptional activation and is dynamically regulated by the opposing activities of histone acetyltransferases (HAT) and histone deacetylases (HDAC)(Haberland et al. 2009) Seminal works have revealed the first evidence of an epigenetic regulation of muscle gene expression by HATs p300/CBP and PCAF (Eckner et al. 1996; Yuan et al. 1996; Puri et al. 1997; Puri, et al. 1997; Sartorelli et al. 1999; Polesskaya et al. 2000). Further work has shown that three distinct classes of HDACs are involved in the repression of muscle gene transcription by countering the activities of HATs during myoblast proliferation (Puri et al. 2001; Mal et al. 2001; Lu et al. 2000; McKinsey et al. 2000) and in response to metabolic perturbations (Fulco et al. 2003). During myoblast proliferation class I HDACs can associate and inhibit MyoD (P L Puri et al. 2001; Mal et al. 2001), while class II HDACs (specifically HDAC4 and HDAC5) are dedicated repressors of MEF2 activity (J Lu et al. 2000; Jianrong Lu et al. 2000; T a McKinsey et al. 2000), Moreover, a distinct class of histone deacetylases – such as the NAD(+)-dependent histone deacetylase Sir2 - form a repressor complex with PCAF and MyoD (Fulco et al. 2003). Class I and II HDAC contribute respectively to the hypoacetylation of MyoD and downregulation of MEF2 target genes during proliferation; Sir2 works as a redox sensor adjusting muscle gene expression in response to metabolic variations (Fulco et al. 2003). Recent genome-wide studies based on ChIP-seq technology have revealed a link between MyoD chromatin binding in myoblasts, recruitment of HATs and generation of local hyperacetylation at MyoD binding sites (Ebox) in myoblasts that is likely to preset regions of chromatin accessibility prior to the activation of the differentiation program (Cao et al. 2010). A similar pattern was observed for the neurogenic bHLH transcriptional activator NeuroD (Fong et al. 2012). These studies support a model of cellular differentiation that is genetically determined by Ebox sequences, with a superimposed “epigenetic presetting” of the nuclear landscape determined by the local acetylation that marks the preferential availability of Ebox sequences for transcriptional initiation. Further studies have detected MyoD-bound distal enhancers where additional transcription factors are assembled around Ebox sequences, leading to regional enrichment of H3K4 mono-methylation (H3K4me1) and H3K27 acetylation (H3K27ac) – two typical markers of active enhancers (Blum et al. 2012). Altogether these data are consistent with the general concept that cell type-specific transcriptional activators expressed in tissue progenitors before differentiation (e.g. MyoD and Myf5 in myoblasts) pre-determine chromatin accessibility at specific loci for activation of tissue-specific gene expression (Biddie et al. 2011; John et al. 2011). Thus, an emerging hypothesis is that in MuSC the generation of an optimal epigenetic landscape determines the lineage identity and prepares the genome for signal dependent activation of skeletal myogenesis. Several studies have illustrated key epigenetic networks that coordinate transition from quiescence to differentiation of MuSC sequentially exposed to different extrinsic signals. In quiescent satellite cells, which are typically insulated from the regeneration environment, Polycomb Repressive Complex 2 (PRC2) mediated repression of muscle gene transcription (Caretti et al. 2004) and Pax7-activated gene expression (Soleimani et al. 2012) cooperate to maintain the latent myogenic lineage. Upon activation by muscle injury, satellite cells are exposed to distinct waves of regeneration signals that sequentially promote expansion and asymmetric division, giving rise to self-renewing or differentiating progenies (Brack & Rando 2012; Y. X. Wang et al. 2013; Yennek & Shahragim Tajbakhsh 2013). Studies from McKinnel and colleagues have shown that in differentiation committed MuSCs Pax7 activates Myf5 expression by recruiting Wdr5-Ash2L-MLL2 histone methyltransferase to Myf5 gene (McKinnell et al. 2008). Pax7-mediated activation of Myf5 is tightly regulated, via Carm1-dependent methylation of Pax7 itself (Kawabe et al. 2012). The differentiation-committed population of MuSCs is exposed to regeneration signals, which elicit pro-differentiative cascades, such as the p38 signaling (Troy et al. 2012), that direct chromatin binding and activity of chromatin-modifying complexes to coordinate the expression of distinct subsets of genes. In particular, p38 pathway on one side directs the recruitment of the chromatin-remodeling SWI/SNF complex on muscle loci (Simone et al. 2004; Serra et al. 2007; Forcales et al. 2012; Albini et al. 2013) and the deposition of H3K4me3 mark on the chromatin of muscle genes, via an Ash2L-containing methyltransferase complexes (Rampalli et al. 2007). On the other side p38 kinases alpha promotes the chromatin binding of the Polycomb Repressive Complex 2 to the regulatory elements of Pax7 gene and cell cycle genes to extinguish their expression in differentiating MuSCs (Palacios et al. 2010) – (Figure 1 illustrates these sequence of events).

Figure 1.

Illustration of the different functional states of MuSCs and related regulatory networks that coordinate expression of different subsets of genes at each stage. In quiescent cells the muscle stem cells lineage is maintained through the cooperation of Pax7 and PRC2 that promote the expression of MuSC lineage genes and repressed muscle differentiation genes, respectively, and through miR31 that blocks Myf5 translation. Upon activation by muscle injury, the exposure to regeneration cues promotes asymmetric division, giving rise to differentiation-committed MuSCs in which Pax7 activates the expression of cell cycle genes and, with the collaboration of the Wdr5-Ash2L-MLL2 histone methyltransferase complex (HMT) promotes Myf5 transcription. At the same time HDAC class I and II contribute to repress the activity of MEF2 and MyoD and hold the cells at the stage of proliferating myoblasts. Subsequent exposure to differentiation signals triggers pro-myogenic cascades (e.g. p38) and causes the recruitment of SWI/SNF complex on muscle loci, the deposition of H3K4 marks on Myog and CKM by the HMT complex and the downregulation of the PRC2 enzymatic subunit EzH2. MyoD and SWI/SNF complex also promote the expression of Myomirs. In addition different miR contribute to the repression of Pax7 YY1 and Ezh2. The overall decrease in Ezh2 levels lead to de-repression of muscle genes and formation of limiting amounts of EzH2-based PRC2 complexes that extinguish the expression of Pax7 and cell cycle genes.

Other histone modifications (Asp et al. 2011) and networks involving non-coding RNAs (reviewed (A. Williams et al. 2009) contribute to generate an epigenetic landscape that is permissive for maintenance of the myogenic lineage in quiescent MuSCs and that allows to activate muscle gene expression and form differentiated myotubes in response to regeneration signals. In particular, a cluster of miRNAs (miRs 1, 133 and 206) that are specifically expressed in skeletal muscles and hence are defined as “myomiRs” control muscle differentiation by targeting both transcription factors and epigenetic remodelers (J.-F. Chen et al. 2006; Rao et al. 2006; Kim et al. 2006; Rosenberg et al. 2006; Cacchiarelli et al. 2010; J.-F. Chen et al. 2010; Dey et al. 2011). Interestingly, these myomiRs are activated by MRF and require the activity of SWI/SNF chromatin remodeling complex (Mallappa et al. 2010) – see figure 1. Other miRNAs with a ubiquitous expression pattern have been reported to control a peculiar biological feature of MuSCs, such as cellular quiescence. In quiescent MuSCs, miR-31 ensures the functional inactivation of Myf5, by retaining its transcripts inside cytoplasmic mRNP granules, thereby preventing its translation (Crist et al. 2012). MuSC quiescence is also maintained by miR489-mediated targeting of DEK, an oncogene that localizes to differentiated daughter cells during asymmetric division of satellite cells and promotes transient proliferation of myogenic progenitors (Cheung et al. 2012). As muscle progenitors differentiate the combined activity of miR214 (A. H. Juan et al. 2009), miR26a (Wong & Tellam 2008) and miR29 (H. Wang et al. 2008) contribute to downregulate the enzymatic subunit of PRC2 - EzH2 – and its partner YY1, thereby relieving PRC2-mediated repression of muscle genes (Caretti et al. 2004). Finally, competing endogenous RNAs (ceRNAs) have recently been described to play a role in the activation of the muscle differentiation program, with linc-MD1 competing with miR-133 and miR-135 for the control of expression of MAML1 and MEF2C, which are required for muscle-specific gene transcription (Cesana et al. 2011). A current gap of information remains on the epigenetic regulation of MyoD expression in differentiation-competent satellite cells and more in general the control of the segregation of self-renewing MuSCs that do not express MyoD.

Adult skeletal muscle resident cells

Despite being considered the major source of progenitors in adult skeletal muscles, MuSCs are not the only cellular effectors of muscle regeneration, as during the past two decades several other muscle-derived populations have been identified that participate either directly or indirectly to the regeneration process. While the origin and the roles of these populations are still far from being firmly established, it appears that a heterogeneous population of cells might contribute to muscle regeneration in physiological conditions, but might turn into pathological phenotypes that contribute to fibrosis and fatty infiltration in chronic diseases and during aging. It is likely that dynamic transitions from one population to another might bias the composition of these cell types and ultimately dictate their activity (Malecova and Puri 2012). These transitions reflect specific epigenetic states that avail or repress the expression of distinct sub-sets of genes – e.g. muscle or fibro-adipogenic genes.

Due to the different methods used to isolate these cells, it is still unclear the phenotypic and functional relationship and the overlap between the multitudes of cell populations described so far to reside in the adult muscle. Nevertheless, these populations can be empirically divided on the basis of their ability to adopt the muscle lineage (Mitchell et al. 2010; Gussoni et al. 1999; Tamaki et al. 2002; Zheng et al. 2007; A. Dellavalle et al. 2007) or to adopt alternative mesodermal lineages (Uezumi et al. 2010; Joe et al. 2010; Dulauroy et al. 2012). However, it remains possible that most of these cell types are endowed with a facultative phenotype that is highly dependent on the environment and therefore their lineage plasticity might not have been entirely appreciated (figure 2).

Figure 2.

Schematic representation of the possible lineages that can be adopted by the different muscle resident cells. List of the reference used in the figure: 1 (Gussoni et al. 1999); 2 (Uezumi et al. 2006); 3 (Qu-Petersen et al. 2002); 4 (Cao et al. 2003); 5 (A. Dellavalle et al. 2007); 6 (Alliot-Licht et al. 2005); 7 (Farrington-Rock et al. 2004); 8 (Schor et al. 1990); 9 (Doherty et al. 1998); 10 (Minasi et al. 2002); 11 (Mitchell et al. 2010); 12 (Joe et al. 2010);13 (Uezumi et al. 2010)

Myogenic Muscle Resident Cells

In 1999 Gussoni and colleagues showed that a population refractory to Hoechst 33342 staining (and thereby indicated as side population – SP) was present in bone marrow and in adult skeletal muscles, and was endowed by an inducible myogenic potential (Gussoni et al. 1999). SP cells could repopulate the muscle (and generate satellite cells) and the hematopoietic system in irradiated animals (Gussoni et al. 1999; Jackson et al. 1999). Later on, it was shown that the differential expression of c-kit and CD45 discriminate two distinct SP populations - with the hematopoietic potential being restricted to the CD45+ sub-population (Asakura et al. 2002). Unexpectedly the myogenic potential of SP cells was shown to be independent on Pax7 expression (Asakura et al. 2002), supporting an origin of these cells distinct from that of satellite cells. Further studies demonstrated that the majority of SP cells in healthy muscle belong to the CD31+ CD45− fraction (Uezumi et al. 2006). Interestingly, when exposed to specific signals in vitro SP cells could adopt alternative mesodermal-derived phenotypes, such as osteogenic and adipogenic lineages (Uezumi et al. 2006), revealing the dependency of fate commitment upon exposure to environmental cues. More recently, Tanaka et al reported that a subset of SP cells actually express the MuSC markers Syndecan-4 and Pax7 (Tanaka et al. 2009). This rare (ABCG2+ SCA1+ Syndecan-4+ Pax7+) subpopulation can differentiate into myotubes in vitro and when transplanted in a regenerating muscle can engraft at high efficiency (Tanaka et al. 2009).

Another approach to identify stem cells in the muscle was based on pre-plating. Several populations have been identified based on temporal difference in adhering to collagen-coated plates. Increasing the frequency of pre-plating enriches for the presence of cells that express the muscle marker desmin, and this correlates with an enhanced survival after engraftment (Qu et al. 1998). These cells are typically positive for Sca1 and CD34 expression and display stem-cell like characteristics, being able to replenish the myogenic compartment and self renew in vivo; hence, they were termed muscle derived stem cells (MDSC) (Qu-Petersen et al. 2002). As further proof of the plasticity of MDSC Cao and colleagues demonstrated that these cells can repopulate the bone marrow in a primary irradiated animal and if re-isolated and re-injected can reconstitute the bone marrow in a secondary lethally irradiated recipient or adopt the myogenic lineage (Cao et al. 2003).

An additional muscle resident population that retains the ability to differentiate along mesenchymal lineage and adopt the muscle phenotype are the pericytes (A. Dellavalle et al. 2007; Crisan et al. 2008). These cells derive the from blood vessels and can be isolated as CD146^high, CD34^neg, CD45^neg, CD56^neg cells (Crisan et al. 2008); Kutcher and Herman 2009). Surprisingly these cells never express myogenic markers, such as Pax7 Myf5 or MyoD in the proliferation stage; however, low levels of expression of MyoD were detected when their differentiation was induced (A. Dellavalle et al. 2007). From a regulatory perspective pericytes do not seem to follow the canonical pathway of skeletal muscle differentiation and for this reason it would be of great interest to characterize the steps that lead to myogenesis in these cells. Recent studies have addressed the potential of this cell type to differentiate into myoblasts, adipocytes, osteoblasts and odontocytes (Schor et al. 1990; Doherty et al. 1998; Farrington-Rock et al. 2004; Alliot-Licht et al. 2005). Since pericytes are present in several tissues they might provide a general source of mesenchymal stem cells, rather than a have a specific role in muscle regeneration. A pericytes-like population has been isolated from vessels of post-natal tissues. These cells are referred as to mesoangioblasts (De Angelis et al. 1999; Minasi et al. 2002; Sampaolesi et al. 2003; Sampaolesi et al. 2006) and were first isolated from the dorsal aorta during embryonic development (De Angelis et al. 1999). Mesoangioblasts express several endothelial markers, such as CD34, Sca-1, Flk1 and CD144 (Minasi et al. 2002). Embryonic mesoangioblast can be isolated more efficiently and their proliferation rate is higher compared to adult mesoangioblasts; nevertheless both cell types are similar in terms of expression profile and differentiation capabilities (Sampaolesi et al. 2003) and can differentiate into several lineages, including skeletal muscle, smooth muscle, chondrocyte, osteocytes, blood cells and hematopoietic tissues (Minasi et al. 2002). Of note, the myogenic potential of these cells can be exploited for therapeutic purposes in muscular diseases, as these cells are able to restore dystrophin expression when transplanted in mouse and canine models of muscular dystrophies (Sampaolesi et al. 2003; Sampaolesi et al. 2006). Due to their ability to migrate inside the muscle after systemic injection and to the fact that can be expanded in vitro, mesoangioblasts provide one of the most promising tools to treat muscular dystrophies. From a molecular perspective is important to note that skeletal muscle lineage commitment in mesoangioblasts relies on Pax3 and not on Pax7. This might reveal their embryonic origin. However, their actual role in adult myogenesis during physiological and pathological conditions has not been elucidated yet.

More recently, another myogenic non-satellite muscle interstitial population has been identified by Sassoon lab, based on the expression of the cell stress mediator PW1 and their anatomical (interstitial) position (Mitchell et al. 2010). PW1 positive interstitial cells (PICs) are isolated by FACS as CD45^neg, Ter119^neg, Sca1^high, CD34^pos and can generate both smooth and skeletal muscle populations in vitro, upon exposure to stimuli that induce expression of Pax7 and MyoD (Mitchell et al. 2010). Consistently, PICs derived from Pax7^−/− mice are unable to differentiate (Mitchell et al. 2010). Upon transplantation, PICs can self renew, generate other PICs and satellite cells, and participate to myofiber formation.

Non Myogenic Muscle Resident Cells

Not all muscle interstitial cells posses the ability to adopt the myogenic lineage. Indeed, a number of cell types have been reported to provide an environment supportive (functional niche) the activity of MuSCs. Among these cells, a heterogeneous population of bipotent, non myogenic muscle interstitial cells collectively indicated as fibro adipogenic progenitors (FAPs) have recently emerged as key cellular determinant of muscle ability to regenerate or undergo fibro-adipogenic degeneration (Joe et al. 2010; Uezumi et al. 2010). These cells show a basal fibroblast-like phenotype, with an inducible potential to differentiate in vitro and in vivo into adipocytes when exposed to pro-adipogenic cues (Joe et al. 2010; Uezumi et al. 2010). Both groups isolated these cells as CD31 and CD45 negative cells; however, while Joe and colleagues used Sca1 and CD34 as positive markers, Uezumi and colleagues isolated these cells by virtue of their expression of PDGF receptor alpha (PDGFRa). Furthermore, a substantial overlap between the populations was observed, leading to the conclusion that FAPs represent the majority of PDGFRa^pos cells in the muscle interstitium (Joe et al. 2010). These cells are developmentally distinct from MuSCs, because they never express MRFs nor differentiate into myotubes; however, they could provide a transient “niche” to support MuSCs during differentiation (Joe et al. 2010). These cells quickly proliferate in response to injury and thereby provide a pro-regenerative transient environment. Intriguingly, FAPs show a completely different behavior in a fatty degeneration model. When transplanted in a fibrotic environment they mediate deposition of collagen (Uezumi et al. 2011); and if transplanted in glycerol treated muscles (a fat infiltration model) they differentiate into adipocytes (Joe et al. 2010). These data strongly suggest that the environment regulates the functional phenotype of FAPs. Supporting this idea, one of the molecular switches that control FAPs fate has been identified in the IL-4/IL13 signaling that is triggered by the inflammatory infiltrate typical of regenerating muscles (Heredia et al. 2013). IL4 produced by the eosinophils recruited upon muscle injury promotes the proliferation in FAPs and prevents their adipogenic differentiation. In response to IL4 FAPs also participate to the phagocytosis of necrotic debris. This study provides a direct evidence of the regulatory activity of signals released by dynamic changes of the environment to direct the fate and activity of FAPs. Interestingly, FAP ability to support MuSC mediated regeneration of dystrophic muscles is highly influenced by changes in the environment imposed by the disease progression in mdx mice - the mouse models of Duchenne Muscular Dystrophy (DMD) (Mozzetta et al. 2013), suggesting that FAPs are potential contributors to DMD pathogenesis (Mozzetta et al. 2013; Uezumi et al. 2011). Overall, FAP bi-potency could play a key role in conditions of chronic regeneration, such as DMD. In conditions conducive for regeneration FAPs might support the activity of MuSCs; however, when the regeneration potential has been exhausted or the signals from the environment become de-regulated (e.g. chronic inflammation), as in the case of advanced stages of muscular dystrophies or aging, FAPs might turn into the effectors of a maladaptive repair by fibrotic scars and fatty infiltration. Thus, FAPs appear to be key sensors of skeletal muscle perturbations and are able to transmit environmental changes to the direct cellular effectors (MuSCs) of muscle regeneration.

A recent report from Dularoy and colleagues has identified a specific subset of PDGFRa positive perivascular muscle cells that during acute injury transiently upregulates ADAM12 and differentiates into profibrotic cells (Dulauroy et al. 2012). ADAM12^pos cells disappear in fully recovered muscles (40 days after injury), supporting the idea that ADAM12 de novo expression marks a subpopulation of cells that indirectly contribute to muscle regeneration and eliminated upon muscle healing (Dulauroy et al. 2012). Although these cells share with FAPs the expression of PDGFRa, they are likely distinct populations that might cooperate to direct the regeneration activity of injured muscles, but might turn into cell types that mediate fibrotic and fatty infiltration when their clearance is compromised, such as in the presence of pathological signals or subverted environment typical of degenerating dystrophic muscles. In support of this concept, freshly isolated FAPs are negative for adipogenic markers and begin to differentiate into fibro-adipocytes only when exposed to adipogenic cues.

Epigenetic regulation of lineage determination in muscle-resident cell types

Although the epigenetic basis of the lineage plasticity of muscle resident cells is currently unknown, the knowledge gained on the epigenetic network that regulates gene expression in pluripotent cells, such as embryonic stem cells and adult muscle stem cells, can help to formulate predictions and testable hypothesis.

The induced expression of markers of distinct lineages in the same cell type, upon exposure to different environmental cues in vivo and culture conditions in vitro, clearly indicates that lineage determination is ultimately dictated by the environment, but also that a specific chromatin landscape must have been developmentally preset to permit responsiveness to the environmental cues. For instance, it is likely that retention of chromatin signatures of pluripotency at specific loci are needed to support the phenotypic plasticity in muscle resident cells and allow key events such as the reprogramming to alternative mesodermal–derived fates, including the myogenic, adipogenic and fibrogenic phenotypes, upon exposure to regeneration cues.

In this regard, an impervious or permissive ground for lineage transitions can be determined by special combinations of histone modifications. Studies in pluripotent embryonic stem cells have revealed the importance of the co-existence of activatory and repressive histone marks on the regulatory elements of developmental genes (also referred to as chromatin bivalency) for rapid activation of gene expression, upon resolution of the bivalence (Bernstein et al. 2006; Mikkelsen et al. 2007). Thus, it is likely that retention of a chromatin bivalence at particular loci (e.g. developmentally marked enhancers) confers on adult muscle-resident cells the ability to adopt one of the facultative fates upon resolution of the bivalence by environmental signals. Moreover, tridimensional structures generated by long-distance interactions might also contribute to determine a nuclear landscape permissive or repressive for lineage transitions. While the chromatin dynamics that underlie embryonic stem cells (ESCs) differentiation and induced pluripotency of somatic cells is currently being investigated by multiple studies the paucity and heterogeneity of adult stem cells has limited the ability to perform genome-wide epigenetic analysis of the chromatin dynamics directing their phenotypic transition in response to environmental cues.

For muscle resident cells, that can adopt a muscle phenotype, an essential event is clearly the activation of MRFs (MyoD and Myf5) that typically promote skeletal myogenesis. Thus, future studies should investigate the profiles of DNA methylation and chromatin modifications that control the activation of MyoD (and other MRFs) and downstream target genes. Likewise, recent studies indicate the importance of selection of specific structural variants of the SWI/SNF chromatin-remodeling complex to promote chromatin accessibility at lineage-specific loci and determine specific lineages (Lessard et al. 2007; A. Yoo et al. 2009; J. I. Wu et al. 2007) reviewed in (Lessard & G. R. Crabtree 2010). For example, formation of BAF60C-based SWI/SNF complex is clearly required to support MRF-mediated myogenic conversion of embryonic and somatic cells (Forcales et al. 2012; Albini et al. 2013) reviewed in (Puri & Mercola 2012)

Targeting epigenetic network that control the lineage plasticity of muscle resident cells can be exploited to pharmacologically reprogram in situ pluripotent cells types for therapeutic purposes. Indeed, epigenetic drugs such as HDAC inhibitors (HDACi) have been successfully used to promote therapeutic regeneration and prevent fibrosis and fatty infiltration in dystrophic muscles (Minetti et al. 2006; Mozzetta et al. 2013; Consalvi et al. 2013). Indeed, HDACi can unmask a latent myogenic potential of muscle resident cells, while inhibiting their fibro-adipogenic differentiation (Mozzetta et al. 2013 and our unpublished data), by correcting a pathological epigenetic landscape generated by primary genetic mutations.

Conclusions

The results illustrated and discussed in this review support the existence of a complex interplay between genetic, epigenetic players and environmental signals in the control lineage determination and differentiation ability of MuSCs and muscle resident cells. Due to the heterogeneity of these populations and their functional crosstalk, it is clear that a successful therapeutic approach should not only target MuSCs, but also other muscle resident cells. The control of lineage plasticity in mesodermal-derived adult muscle resident cells by epigenetic drugs is indeed a powerful tool that is currently exploited to promote compensatory regeneration at the expense of fibro-adipogenic degeneration in dystrophic muscle. The optimal application of these drugs appears to depend on the availability of external cues from a permissive environment (Chiara Mozzetta et al. 2013)) and this might inspire future protocols for effective treatments of muscular dystrophies and other neuromuscular disorders.

Abbreviation list

bHLH

basic helix-loop-helix

MRFs

muscle regulatory factors

MuSCs

muscle stem cells

PRMTs

protein arginine methyltransferases

HATs

histone acetyltransferases

HDACs

histone deacetylases

PRC2

Polycomb repressive complex 2

SP

side population

MDSCs

muscle derived stem cells

PICs

PW1 positive cells

FAPs

fibro-adipogenic progenitors

DMD

Duchenne muscular dystrophy

ESCs

embryonic stem cells

HDACi

histone deacetylases inhibitors