SEARCH

SEARCH BY CITATION

Keywords:

  • development;
  • myogenic regulatory factors;
  • Pax3;
  • Pax7;
  • regeneration;
  • skeletal muscle;
  • stem cells

Abstract.

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Establishing skeletal muscle in the embryo from founder stem cells
  5. Genetic strategies to identify prenatal skeletal muscle stem and progenitor cells
  6. Genetic strategies to identify postnatal skeletal muscle stem and progenitor cells
  7. Distinguishing juvenile from adult satellite cells and embryonic founder stem cells
  8. Building muscle in the embryo versus adult muscle regeneration
  9. The skeletal muscle stem cell niche and signalling pathways
  10. Conclusions and perspectives
  11. Conflict of interest statement
  12. Acknowledgements
  13. References

Tissue and organ regeneration proceed in a coordinated manner to restore proper function after trauma. Vertebrate skeletal muscle has a remarkable ability to regenerate after repeated and complete destruction of the tissue, yet limited information is available on how muscle stem and progenitor cells, and other nonmuscle cells, reestablish homeostasis after the regenerative process. The genetic pathways that regulate the establishment of skeletal muscle in the embryo have been studied extensively, and many of the genes that govern muscle stem cell maintenance and commitment are redeployed during adult homeostasis and regeneration. Therefore, correlates can be made between embryonic muscle development and postnatal regeneration. However, there are some important distinctions between prenatal development and regeneration – in the context of the cells, niche, anatomy and the regulatory genes employed. The similarities and distinctions between these two scenarios are the focus of this review.


Introduction

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Establishing skeletal muscle in the embryo from founder stem cells
  5. Genetic strategies to identify prenatal skeletal muscle stem and progenitor cells
  6. Genetic strategies to identify postnatal skeletal muscle stem and progenitor cells
  7. Distinguishing juvenile from adult satellite cells and embryonic founder stem cells
  8. Building muscle in the embryo versus adult muscle regeneration
  9. The skeletal muscle stem cell niche and signalling pathways
  10. Conclusions and perspectives
  11. Conflict of interest statement
  12. Acknowledgements
  13. References

Regenerative medicine has gained important insights through the study of developmental and regenerative biology. Tissue development and regeneration share common features as modules of regulatory pathways and transcription factors which are crucial for prenatal development are redeployed for tissue reconstruction after trauma [1–4]. However, there are also some notable differences between establishing a tissue during development, and regenerating it after damage (Figs 1 and 2). For skeletal muscle, numerous experimental models have been used to investigate the regeneration process, which involves an interplay between diverse cell types as the regenerate forms [2]. The relative outcome and the contribution of stem cells with respect to the supporting tissue during tissue regeneration can be different depending on the paradigm used [1, 2]. In contrast, during normal embryonic development, skeletal muscle is established in a multistep fashion concomitant with a more complex and disjointed programme of gene activations and the generation of different precursor cell types [3, 5–8]. Skeletal muscle in the mouse is established from embryonic day 8.5/9 (E8.5/9) to E18.5 (birth ∼ 19 days) with a further maturation during the postnatal period for about 2–3 weeks (Fig. 1). After destruction of skeletal muscle in the adult, regeneration is usually completed and homeostasis is established within 3–4 weeks after injury.

image

Figure 1.  Scheme depicting skeletal muscle development and homeostasis in vertebrates. Multiple phases of growth mobilize stem and progenitor cells to establish skeletal muscles, while maintaining a population that resists differentiation throughout life. Those that establish muscles prenatally are referred to here as founder stem cells (FSC1, FSC2, FSC3) to distinguish them from the juvenile and adult satellite cell population, which will contain the adult stem cells. It is proposed here that FSC1 will be largely exhausted in the embryo, whereas FSC2 (short range displacement from somite in trunk) and FSC3 (long range migration from somites to limbs, diaphragm, tongue), will give rise to the majority of adult skeletal muscle stem cells. In all regions, an anlage is established prior to Pax7 expression. Subsequently Pax7 expression marks all of the founding stem and progenitor cells from mid-embryonic stages onwards. Emerging juvenile satellite cells are ensheathed underneath a basement membrane from about 2 days before birth in mice. They continue to proliferate until about 12 days after birth. From that time, and for about 2 weeks, quiescent ‘adult’ satellite cells emerge. Note that satellite cells are referred to as adult muscle stem cells, yet this population is heterogeneous. The true stem cell entity within this heterogeneous population remains to be identified.

Download figure to PowerPoint

image

Figure 2.  Distinctions between myogenesis in the head and that in the body and developmental versus regenerative myogenesis. EOMs, extraocular muscles; PA pharyngeal arch.

Download figure to PowerPoint

As most skeletal muscles are not vital for the survival of the organism, this tissue provides an excellent paradigm to investigate the degeneration and regeneration process. Its accessibility and remarkable ability to regenerate even after multiple rounds of injury make this tissue amenable for investigating the role of stem cells in this process, and over extended periods. An understanding of how stem and associated cells construct a tissue can provide insights into the development of clinical strategies to combat the normal decline in skeletal muscle performance during ageing, or its reconstitution after trauma and during disease. This review addresses these issues, focusing on skeletal muscle regeneration and how different factors including genes, cells and the environment impinge on this process. Similarities and differences between developmental and regenerative myogenesis are highlighted. Definitions of stem, progenitor and precursor cells and their application to skeletal muscle and other tissues, are discussed elsewhere [9]. As a convention, a lineage progression is referred to here as stem (fate not acquired)[RIGHTWARDS ARROW]progenitor (fate being acquired; determination genes activated)[RIGHTWARDS ARROW]precursor (myoblast; fate acquired)[RIGHTWARDS ARROW]differentiated (myofibres; functional entity). In addition, other aspects of muscle development or regeneration have been reviewed extensively previously [2–8, 10–13].

Establishing skeletal muscle in the embryo from founder stem cells

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Establishing skeletal muscle in the embryo from founder stem cells
  5. Genetic strategies to identify prenatal skeletal muscle stem and progenitor cells
  6. Genetic strategies to identify postnatal skeletal muscle stem and progenitor cells
  7. Distinguishing juvenile from adult satellite cells and embryonic founder stem cells
  8. Building muscle in the embryo versus adult muscle regeneration
  9. The skeletal muscle stem cell niche and signalling pathways
  10. Conclusions and perspectives
  11. Conflict of interest statement
  12. Acknowledgements
  13. References

Skeletal muscle development in vertebrates is initiated throughout the organism [3, 8]. A number of regulatory molecules are crucial for myogenic fate to be acquired from stem cells – the discussions here will be confined to a subset of these genes, particularly in the mouse. Pax3 and Pax7 are paired/homeodomain transcription factors that play crucial roles in multiple tissues and organs including skeletal muscle [3, 14, 15]. Although Pax3 exerts its influence primarily during embryonic development (but not in the head) and Pax7 during perinatal growth, the double mutant results in the loss of body muscles from mid-embryonic stages (E12.5) [16] due to a loss of the founding stem/progenitor cell populations [16–19] [called here founder stem cell-2 (FSC-2); Fig. 1)]. During perinatal development, Pax3 does not appear to play a critical role in muscle progenitor cells, but Pax7 null mutants are severely deficient in the numbers of juvenile satellite cells [20–23] (Fig. 1, see below).

The myogenic regulatory factors (Mrfs) Myf5, Mrf4 and Myod determine skeletal muscle cell identity, and are therefore considered to be the gatekeepers for entry into the myogenic lineage [24–26]. Mice triple mutant for Myf5, Mrf4 and Myod do not generate myoblasts and muscle fibres, albeit stem/progenitor cells largely persist in the organism in the absence of myogenic commitment [19, 25] (Fig. 3). Accordingly, even the culture of the triple mutant embryonic cells fails to bypass the requirement for these genes to generate myoblasts [25]. The fourth myogenic regulatory factor Myogenin plays a critical role in muscle differentiation as mice carrying a germline mutation in this gene lack differentiated muscles, and they die at birth [27–30]. However, embryonic Myogenin null myoblasts do differentiate when cultured in vitro, probably via the function of Myod and Mrf4 [30]. This interesting observation, which bypasses the in vivo requirement for Myogenin, might provide insights into the report that Myogenin is dispensable for muscle differentiation in adult mice [28, 31] (see below).

image

Figure 3.  Head and body myogenic programmes are distinct. (a) The majority of head muscles originate from paraxial (major contributor) and prechordal (minor contributor) cranial mesoderm. Muscles in the body (trunk, limbs) and some muscles in the head (e.g. tongue) arise from founder stem and progenitor cells in somites. Pax3 (body, but not in the head) and Pax7 act as specification genes for establishing the founding cell population once the anlagen are established. (b) The myogenic regulatory determination genes Myf5, Mrf4 and Myod are essential gatekeepers for the acquisition of muscle cell fate. In the triple mutant, no muscle forms. However, for several days, muscle progenitor cells can be traced by nlacZ expression from the Myf5nlacZ allele. PAs, pharyngeal arches; H, heart; FL, forelimb.

Download figure to PowerPoint

A recurrent theme in skeletal myogenesis is the disjointed nature in which diverse transcriptional regulators act in different stem cell populations. Skeletal musculature in vertebrates can be compartmentalized into body (trunk, limbs) and head muscles [3, 32–34]. Myogenesis has been investigated more extensively in the body than in the head in different model organisms [3, 8, 10, 32–34]. All skeletal muscles in the body proper arise from muscle founder stem cells located in somites – transitory structures that form in pairs, on either side of the neural tube, and as epithelial spheres from the presomitic mesoderm [3, 8]. Cells in the epithelial somite are naïve, and cell fates are acquired by signalling molecules emanating from the surrounding environment [8, 10, 32, 34, 35]. Multiple cell types including cartilage (sclerotome), skeletal muscle, dermis, endothelial, and connective tissue are derived from this structure. As somites mature, a transitory epithelial dermomyotome forms, and this contains multipotent cells that generate diverse cell types including all of the skeletal muscle stem cells in the body proper.

Although the core determination genes play a crucial role in executing muscle cell fates, they are not used in the same combinatorial manner in all regions where muscles are established. Furthermore, Pax3 and Pax7 play important roles prenatally and perinatally in regulating muscles in the body, but head myogenesis has evolved with a distinct transcriptional code. Indeed, recent studies underscored the notion that the site of origin of the founder stem cells is important in determining the regulatory genes that govern their behaviour [36–39]. These findings are also relevant in the context of myopathies, some of which affect only a subset of muscles, whilst other muscle groups escape the disease [40]. Therefore, the developmental ontology of skeletal muscle might provide insights into the susceptibility of certain muscles to succumb to the disease. It was reported previously that Pax3:Myf5(Mrf4) mutants (Mrf4 compromised in cis) uncouple myogenesis in the head from that in the body, where muscles do not form in the latter [41]. These studies showed that head and body myogenesis are regulated differently, and that Pax3 and Myf5 act in parallel genetic pathways (Fig. 4). Other studies are in keeping with this notion (see [8, 32, 34]). These findings raise the question: what regulators cooperate with the Mrfs to establish MPC fate in the head?

image

Figure 4.  Core and complementary genetic networks regulating skeletal muscle stem cell fates. Myogenesis in the somites (trunk, limbs) relies on Pax3 to complement the core myogenic regulatory network – Myf5, Mrf4 and Myod [25, 41]. EOMs have dispensed with this complementary regulatory pathway, hence, in absence of Myf5 and Mrf4, Myod expression is compromised and EOMs do not form. Like myogenesis in the somites, PA muscles rely on a complementary genetic pathway: Tbx1 co-operates with Myf5, and in their combined absence, Myod expression is compromised and PA muscles are essentially all missing. Mrf4 determines embryonic but not foetal MPC fate in the body and EOMs [25, 39]. Pitx2 null as well as Myf5:Mrf4 double mutants lack EOMs (see [3]). With low penetrance, some EOMs are observed in Myf5:Mrf4 double mutants and some PA muscle are observed in Tbx1:Myf5 double mutants. In these cases, it is possible that Pitx2 rescues Myod expression [39]. Arrows do not necessarily imply direct interactions.

Download figure to PowerPoint

Head muscles have diverse functions including facial expression, feeding and eye movements, and they are classified broadly into three groups: 1) tongue and posterior neck muscles, 2) pharyngeal muscles including those in the jaw, anterior neck and face and 3) extraocular muscles (EOMs). The majority of these muscles are derived from cranial paraxial mesoderm (CPM) comprising paraxial head (PHM), and the more anterior prechordal (PCM) mesoderm (Fig. 3a) [22–34]. CPM, which is contiguous with somitic mesoderm, gives rise to EOMs and those in the pharyngeal arches (PAs). In addition to the genetic diversity among the muscle groups that originate from these regions, a cellular complexity characterises muscle progenitors in the first PA. Interestingly, a subset of progenitors in the first PA migrates ventrally to contribute to cardiac development in the anterior heart field [32–34]. A second resident population contributes to first PA skeletal myogenesis [37, 42–44]. Furthermore, a number of markers of skeletal myogenesis, such as the transcription factors Tbx1 and Pitx2, are shared with cardiogenesis [32, 34]. Thus, divergent myogenic programmes (skeletal and cardiac) arise from PA muscle progenitors. In contrast, extraocular MPCs give rise to a specialized subset of seven muscle groups in the mouse (six in human). Curiously, in Duchenne muscular dystrophy where essentially all skeletal muscle groups are affected, the EOMs are spared [40, 45].

Extraocular muscles have an obligate requirement for Myf5 or Mrf4 for initiating myogenic fate, and in the absence of these Mrfs, muscle progenitors are lost rapidly by apoptosis - this is not the case for somite or PA-derived MPCs [39] (Fig. 4). Therefore EOM progenitors employ a distinct combination of genetic tools from those used elsewhere. Myogenesis in pharyngeal muscles operates with a genetic code which is distinct from that in the body and the EOMs. Here myogenesis is severely impaired in Tbx1:Myf5 double mutants [39]. Tbx1 is a T-box containing gene and mice mutant for this transcription factor have a random complement of hypomorphic first arch-derived muscles [37]. Interestingly, reduced levels of TBX1 in humans appear to contribute to pharyngeal muscle hypotonia [37].

Therefore, skeletal myogenesis is defined by complementary genetic pathways, governed by Pax3 in body, and Tbx1, in pharyngeal muscles, but this complementary pathway is absent in EOMs (Fig. 4). Why different transcriptional codes are employed in skeletal muscles in distinct locations is not clear. In any case, this observed patchwork in stem cell behaviour raises the possibility that their adult muscle derivatives might retain some of these ontological features.

At this stage, it is appropriate to discuss the nature of cells in which these critical genes act first to specify, then determine cell fates. In other words, stem cells in skeletal muscle. In more general terms, what is the stem cell entity in skeletal muscle, be it embryonic or adult? Is this an invariant self-renewing cellular entity? Can committed myogenic cells return to the stem cell state to replenish this population in extreme cases of injury, or after the loss of stem cells? Many of these key questions remain unanswered for virtually all tissues and organs, including skeletal muscle [9, 46]. Consequently, most ‘stem cell’ populations that are currently being investigated are heterogeneous in nature. They might constitute truly distinct cell entities (stem-1 + stem-2, etc. or stem + committed cells). Alternatively, a single stem cell state defines this tissue in any given location. Whichever the case, all cells, including stem cells, exercise some range of variations in gene expression. This may explain, in part, some of the observed heterogeneities. One major challenge is characterizing this range in cell behaviour and determining which cells are more susceptible to revert (from committed to stem cell), remain as stem cells, or progress to the differentiated state during normal growth or regeneration. Skeletal muscle has a well-defined repertoire of transcription factors which, when mutated, can affect the behaviour of myogenic cells dramatically. Myf5 (Mrf4):Myod mutant embryos, and in some cases Myf5nlacZ/nlacZ(Mrf4) mutants, do not display a loss of the ancestral Pax+ cell population during early stages of development (Pax3+, Pax7+, Tbx1+, or Pitx2+; depending on location and timing), but myogenic lineage progression is blocked at the progenitor cell state [19, 24, 47] (Fig. 2). Therefore, it is appropriate to distinguish the “founding” stem/progenitor cell populations, which are better characterized in the body, into classes: 1) FSC1 establish the primary muscle mass (myotome) from E9 in the mouse. These cells express Pax3 (and later Myf5/Mrf4/(Myod)), and they arise primarily from the dorsal and ventral lips of the dermomyotome, but they do not all require this gene for myogenesis, since a myotome forms in Pax3 null embryos [7, 16, 41]; 2) the FSC2 population is released from the central dermomyotome into the underlying myotome from about E10.5 [16–19], and it expresses Pax3/Pax7 (and later Myod/(Myf5)). These cells are lost in the body in Pax3:Pax7 double null germline mutants from mid-late embryonic stages [16] (Figs. 1, 2); 3) the FSC3 population leaves the ventral dermomyotome of some somites to establish skeletal muscles in the limbs, diaphragm and tongue from about E9.5–E11.5. This population expresses Pax3, Met, Lbx1 and Meox1 (and later Myf5/Myod/(Mrf4)) (see [3, 7, 8, 25]). Current knowledge suggests that FSC1 is largely exhausted early, whilst FSC2 and FSC3 persist to contribute to the majority of adult skeletal muscle stem cells (see [3]). This notion needs to be explored further. Therefore combinations of Myf5, Mrf4, Myod, Pax3 and Pax7 mutants affect distinct classes of FSCs in the embryo. Subpopulations of founder stem cells in the head are beginning to be explored. Recent lineage studies in the head have provided some insights into these populations in the chick and mouse [36]. Notably, Islet1, which is a marker of cardiac progenitor cells, also identifies a subset of jaw muscle-derived satellite cells, thereby underscoring the notion that postnatal satellite cells in the head have unique genetic requirements and origins, and that they arise from distinct cell populations [36].

In summary, given the distinctions between different muscle groups and the maintenance of a certain developmental ‘memory’ linking prenatal and postnatal stem cells, it is noteworthy that the majority of experimental manipulations with adult skeletal muscles is performed on skeletal muscles below the knee (e.g. tibialis anterior, gastrocnemius, extensor digitorum longus). Therefore, the interpretation of those studies needs to be considered in the context of the discussions above.

Genetic strategies to identify prenatal skeletal muscle stem and progenitor cells

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Establishing skeletal muscle in the embryo from founder stem cells
  5. Genetic strategies to identify prenatal skeletal muscle stem and progenitor cells
  6. Genetic strategies to identify postnatal skeletal muscle stem and progenitor cells
  7. Distinguishing juvenile from adult satellite cells and embryonic founder stem cells
  8. Building muscle in the embryo versus adult muscle regeneration
  9. The skeletal muscle stem cell niche and signalling pathways
  10. Conclusions and perspectives
  11. Conflict of interest statement
  12. Acknowledgements
  13. References

As noted earlier, genetic strategies in mice have unraveled the roles of key regulators of muscle function [3, 5, 6, 10]. Recently, two studies reported that Myf5 and Myod expression can distinguish subpopulations of embryonic muscle progenitor cells. Using a genetic strategy where Myf5Cre mice were crossed to RosaStop-DTA reporter mice (diphtheria toxin A; potent inhibitor of translation), cells that expressed the Cre recombinase were ablated [48, 49]. Interestingly, in both studies, and using different Myf5Cre alleles, early embryonic cells were eliminated after DTA expression. However, mice were born and muscles were present in these genetic mutants. These findings indicate that a subpopulation of myogenic cells does not express sufficient amounts of Myf5 but, more Myod instead. Alternatively, ablation of the Myf5-expressing subpopulation results in a compensatory expansion of those cells expressing more Myod [48, 49]. A similar strategy using Mrf4Cre crossed to RosaStop-DTA resulted in the loss of differentiated muscles [49], in keeping with the observation that Mrf4 is expressed predominantly after muscle differentiation [8]. However, Mrf4 is also expressed in embryonic progenitors in the trunk and head [25, 39], therefore, in some cases, the cells were ablated prior to muscle differentiation in this genetic cross.

Although mice triple mutant for Myf5 (Mrf4): Myod lack muscles and their progenitor cells do not acquire a myogenic fate [25], subpopulations of cells within the muscle lineage express these transcription factors at different levels in normal development. Whether these subpopulations constitute bona fide separate lineages is a matter of debate. They all originate from somites, for muscles in the body. Genetic ablation strategies presume that sufficient levels of Cre are expressed to recombine loxP sites (and intervening sequences) and allow DTA to be expressed. However, low levels of Myf5 (Cre) expression in some cells may be inadequate to trigger recombination and provoke cell ablation. Direct lineage analysis should be performed to assess the developmental potential and ancestral history of these subpopulations that express the Mrfs at different levels.

A genetic cell-ablation strategy was employed also with the more upstream markers Pax3Cre and Pax7iCre [50]. As with Myf5Cre, the onset of expression of these genes is important for the interpretation of the outcome. Pax3 is expressed prior to somite formation, therefore Cre recombinase in Pax3Cre embryos is expected to be active in all of the cell lineages that are derived from somites (e.g. sclerotome, ribs and vertebral column, endothelial cells, dorsal dermis, smooth muscle) [3, 5, 10]. In the case of Pax7, the onset of its expression occurs later, from E9.5 in the dermomyotome of the mouse somite, and particularly in its central domain that will give rise to the future skeletal muscle stem/progenitor cells (FSC2) [16, 19, 51]. Crossing of Pax3Cre or Pax7iCre mice with R26RStop-lacZ reporter mice (conditional expression of lacZ in Rosa locus containing a ubiquitously expressed promoter) revealed interesting cell lineage relationships, whereas crossing the same Cre-driver mice with the RosaStop-DTA line ablated specific cell populations [50]. These experiments showed that Pax3- and Pax7-expressing stem/progenitor cells contribute in a differential manner to embryonic (Pax3; downregulated in foetus) and foetal (Pax7) myogenesis. In addition, Pax3 is more widely expressed and it marks endothelial as well as myogenic cells, whereas Pax7 expression is more restricted to myogenic cells derived from somites [50].

On this latter point, it is noteworthy that both endothelial and myogenic cells originate from the dermomyotome, yet they have different requirements for β-catenin [50]. Pax3-expressing dermomyotomal myogenic cells migrate to the limbs and they establish limb muscles. Mutations in Pax3, or other genes such as Met, Lbx1, Scatter factor/Hepatocyte growth factor (Sf/Hgf; ligand of Met receptor), and Meox1, eliminate all or some muscles in this location [7, 8, 52]. Some endothelial cells in the limb also originate from somites opposing the limbs. Cell lineage studies in the chick showed that somite-derived endothelial and muscle progenitors in the limb have a common ancestor in the ventral dermomyotome, and that their respective fates is determined by extrinsic signals in the limb environment [53].

This is an interesting paradigm for divergent cell fates being determined at a late stage – both cell types migrate from the somite to the limb, and each do so with gene specific expression patterns [54]. Notably, when Pax3Cre or Pax7iCre mice were crossed with β-catenin conditional mutant mice, β-catenin was shown to be dispensible for somite-derived (Pax3+) limb endothelial cells. In contrast, the loss of β-catenin at early (but not later) stages results in a loss of muscle progenitors in the limb [50]. Whether this phenotype is related to a role for this gene in cell adhesion or Wnt signalling is not clear [50]. Therefore these cell lineages have differential requirements for β-catenin function.

Although loss of β-catenin at later embryonic stages does not affect significantly embryonic myogenesis, during foetal myogenesis in the limb, progenitor and myofibre numbers are reduced in the absence β-catenin, pointing to a role for β-catenin in regulating myogenic lineage progression at this stage [50]. These observations are in keeping with the reported role for β-catenin and canonical Wnt signalling in regulating the self-renewal of adult satellite cells after muscle injury ([50, 55] and references therein). Finally, β-catenin regulates slow myosin expression and therefore the balance between slow and fast muscle fibres during foetal myogenesis [50]. It would be interesting if this will be the case also during skeletal muscle regeneration in the adult.

In summary, cell lineage and cell ablation analyses have revealed that different subpopulations of prenatal myogenic cells arise during development, and that these have different requirements for Myf5, Mrf4, Myod, Pax3, Pax7 and β-catenin. A point of interest in relation to adult myogenesis is cell plasticity. Given the numerous reports of transdifferentiation from distinct cell types to skeletal muscle [56, 57], it is tempting to speculate that those cells that crossover to another fate might be primed if they share common developmental origins, or transcriptional regulatory modules, and this could facilitate their transition to the other cell fate.

Genetic strategies to identify postnatal skeletal muscle stem and progenitor cells

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Establishing skeletal muscle in the embryo from founder stem cells
  5. Genetic strategies to identify prenatal skeletal muscle stem and progenitor cells
  6. Genetic strategies to identify postnatal skeletal muscle stem and progenitor cells
  7. Distinguishing juvenile from adult satellite cells and embryonic founder stem cells
  8. Building muscle in the embryo versus adult muscle regeneration
  9. The skeletal muscle stem cell niche and signalling pathways
  10. Conclusions and perspectives
  11. Conflict of interest statement
  12. Acknowledgements
  13. References

Heterogeneity is also a defining feature of postnatal skeletal muscle stem/progenitor cells [3, 5, 58–60]. Pax7 expression marks essentially all satellite cells [22, 61–63]. The expression of the Mrfs, and their relation to satellite cell heterogeneity is less clear. It is generally believed that Myf5 and Myod protein are not detectable in quiescent adult mouse satellite cells, albeit in the knockin Myf5nlacZ mouse, nlacZ expression marks the majority of quiescent satellite cells [3, 64]. After muscle injury, essentially all activated satellite cells express Myod protein, and this is a hallmark of the activated satellite cell. Curiously, in spite of the presence of Myod protein, differentiation does not occur for several days in vivo and in culture [65]. Subsequently, some cells will lose Myod protein expression, maintain Pax7 expression, leave the cell cycle and self-renew [63]. This suggests that other factors, such as the repressor Id (helix-loop-helix protein lacking a DNA binding domain) which can interact with Myod [66], may be instigating a ‘pause’ to allow myoblast proliferation and cell expansion to occur prior to differentiation. Whether subpopulations of satellite cells differentially express Myf5 and Myod as in the embryo is not clear. Notably, some clonogenic satellite cell-derived myoblasts do not express Myf5nlacZ [67]. This is an important consideration in the context of the reports that the asymmetric distribution or expression of these transcription factors occurs in daughters of satellite cells after cell division ([59, 63, 68]; B. Gayraud-Morel and ST, unpublished observations; see below).

Further evidence for heterogeneity in the satellite cell population comes from cell lineage studies carried out during development, and followed through in postnatal mice. Here, genetic strategies remove an intervening (stop) sequence to activate the expression of a reporter gene driven by the Rosa promoter [68]. Using Myf5Cre crossed with a RosaStop-YFP reporter mouse, a subset of postnatal Pax7-expressing satellite cells were found to be YFP-negative (about 10%), in other words, they did not historically express Myf5 [68]. The interpretation of these observations is that the YFP-negative population represents a more ancestral stem cell entity with a lesser extent of commitment to myogenesis [68].

In other experiments, it was shown that cell divisions that are perpendicular to the myofibre generate a Pax7+/Myf5+ daughter that is myofibre-proximal and a Pax7+/Myf5− daughter adjacent to the basement membrane [68]. The prospective isolation and subsequent transplantation of these two cell populations [YFP+/(Myf5+) and YFP−/(Myf5−) isolated by α7+/β1+ integrins and Lineage−/Sca1−/CD31−/CD45−] into skeletal muscles of perinatal Pax7 null animals (depleted in satellite cells), showed that YFP− cells engrafted and repopulated the niche three to four times more efficiently. By contrast, the YFP+ subpopulation had a greater propensity to differentiate. These findings suggest that a small fraction of satellite cells are more ‘stem-like’ whereas the majority are more committed [59, 68, 69].

This genetic strategy using Myf5Cre mice has been highly informative and it represents a powerful approach for performing cell lineage studies, or prospectively fractionating functionally distinct subpopulations of cells in a particular lineage. Like any method, however, there are also some pitfalls that must be considered when interpreting these results: (i) the strength of the promoter driving Cre expression is a key concern, as defining the subpopulations depends on producing enough Cre protein to promote a recombination event. A similar argument was made for the cell ablation experiments discussed above. Here, there are several potential concerns. First, the Cre transcript (containing a heterologous poly-adenylation signal) is not regulated like the endogenous transcript, thereby raising the possibility that it may not reflect faithful expression of the endogenous gene. Secondly, for any strategy using Cre, some loxP sites in the genome will be refractory to removal by this recombinase [70], and this cannot always be predicted. Therefore detailed characterization of the genetic tools is paramount. Thirdly, for a variety of reasons, some genes exhibit leaky/inappropriate or weak transcription. In the former case, this could potentially result in the precocious activation of the Cre gene, and recombination of the floxed sites. In the case of weak transcription, the effect is the opposite – potentially too little Cre is produced to perform the recombination and the cell will be negative for the expression of the reporter gene in spite of the expression of the gene of interest.

So one question that arises with any genetic strategy of this sort is: what constitutes ‘functional’ expression? The answer varies depending on the context. Generally, functional expression equates with a measurable phenotypic output. For transcription factors, for example, this could be the expression of a target gene. It is not always possible to test this event in all instances. In the specific context of the experiments mentioned above, i.e. Myf5Cre crossed with a RosaStop-YFP reporter mouse – subpopulations of satellite cells with apparently distinct properties were identified [68], so the genetic tools were exploited for that purpose. The biological relevance of this result needs to be explored further.

A recent report used a similar genetic strategy, but this time MyodCre mice were crossed with a RosaStop-YFP reporter mice [71]. The outcome of this result was strikingly different from that with the Myf5Cre mice. Here, virtually all (about 99%) of satellite cells in different muscle groups were YFP+, during perinatal and adult stages, indicating that historically (probably before birth) all future satellite cells expressed MyodCre [71]. Given the caveats concerning the genetic strategies discussed above, how can these contrasting results be reconciled? One possibility is that the functional consequences of the expression of Myf5 or Myod are not interpreted equivalently by the myogenic cell. Alternatively, the level of Myf5 expression allowed the fortuitous identification of subpopulations of satellite cells due to the relatively weaker output of Cre from this locus compared to that of Myod. These reports underscore the necessity of carrying out a detailed characterization of genetic tools. Although the interpretation of the outcome remains precarious, the tools may still be informative if used appropriately.

Distinguishing juvenile from adult satellite cells and embryonic founder stem cells

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Establishing skeletal muscle in the embryo from founder stem cells
  5. Genetic strategies to identify prenatal skeletal muscle stem and progenitor cells
  6. Genetic strategies to identify postnatal skeletal muscle stem and progenitor cells
  7. Distinguishing juvenile from adult satellite cells and embryonic founder stem cells
  8. Building muscle in the embryo versus adult muscle regeneration
  9. The skeletal muscle stem cell niche and signalling pathways
  10. Conclusions and perspectives
  11. Conflict of interest statement
  12. Acknowledgements
  13. References

Founder stem cells which establish the tissue prenatally give rise to ‘emerging’ or juvenile satellite cells during perinatal life. These cells assure muscle growth and regeneration [3, 4]. A clear distinction between foetal myoblasts and emerging satellite cells is required to fully appreciate and functionally distinguish prenatal from postnatal myogenesis. In other words, it is not clear when foetal progenitors and myoblasts become exhausted, and how ‘satellite’ cells can be distinguished from foetal MPCs. Furthermore, the population of mononuclear cells that are associated with perinatal myofibres is heterogeneous (see Fig. 1), consisting of: (i) future satellite cells; it is currently unclear how these Pax7+/(Myf5/Myod+) cells will be selected; (ii) myoblasts that may upregulate or downregulate Pax7 expression, but express Myf5 and Myod; these cells can potentially give rise to future satellite cells or commit to differentiation; (iii) differentiated Pax7-negative/Myogenin-positive cells (Fig. 1). For this reason, it is misleading to refer to this mixed population of cells as satellite cells. They were referred to as ‘emerging’ or ‘juvenile’ satellite cells previously [3, 19] to maintain a distinction between this prequiescent population and those that are in G0 and emerge from about 2–3 weeks after birth [72].

Skeletal muscle contains the functional myofibre unit, which has permanently exited the cell cycle, and the regenerative satellite cell population, which is quiescent and in G0 during homeostasis in the adult [3, 4, 11, 64, 73, 74] (Fig. 5). Adult satellite cells express both NCAM/CD56 in human [75] and markers such as M-cadherin, Pax7 and Myf5, or Myf5nlacZ in the mouse [23, 47, 69, 76, 77]. Other cell types that make up the tissue include vessels (endothelial cells, pericytes and smooth muscles), fibroblasts and mesenchymal stem cells. Many of these cell types are relatively poorly defined because of the lack of cell-specific markers [2, 56].

image

Figure 5.  Scheme of skeletal muscle and associated structures. Satellite cells are located between the basement membrane and the plasmalemma of the myofibre. Image in upper right is a section of the tibialis anterior muscle of a mouse that was perfused with India ink to label the vasculature and stained with X-gal to reveal satellite cells in a Myf5nlacZ adult mouse (courtesy of W. McCord). Note the close proximity of the vessel and satellite cell. The vessel is composed of endothelial cells which form the lumen, and they are separated from pericytes (located intermittently along vessels) by a basement membrane. The fluorescent image (upper right) shows a satellite cell marked with cytoplasmic GFP expression from a Myf5GPF-P adult mouse. Here the entire satelite cell in aligned along the vessel, and the two are separated by a basement membrane (courtesy of F. Chrétien and V. Shinin). Note in both insets the close proximity of satellite cells with blood vessels (nucleus as well as cytoplasm). Three connective tissue layers can be distinguished in skeletal muscle and these form the lattice network and associated basement membranes in which myofibres regenerate after injury. The epimysium is the deep facia component that ensheaths the entire muscle and it is contiguous with the tendon (muscle to bone) and endosteum (facia surrounding bone). The perimysium ensheaths individual muscle fibres into fascicules (bundles). The endomysium is located between fibres and it ensheaths individual muscle fibres. Within the muscle cell (myofibre) the major intracellular source of calcium needed for muscle contraction is the sarcoplasmic reticulum, which connects to the transverse (T) tubules, and these surround the sarcomeres.

Download figure to PowerPoint

Satellite cells enter the cell cycle after muscle injury and they generate myoblasts which will differentiate, then fuse homotypically to form nascent multinucleate myofibres or fuse with regenerating myofibre end-fragments [4, 12, 78]. Signalling molecules that participate in muscle regeneration include fibroblast growth factor, members of the transforming growth factor-β family, insulin-like growth factor-I and II, hepatocyte growth factor (Hgf) and interleukins (see Refs [4, 79–81]). Some of these molecules are key regulators of embryonic myogenesis. For example, Wnts can stimulate myogenesis in somites, but repress the programme in the head [8, 34, 35]. Also, mutations in RBPJk, which mediates Notch signalling, and the Notch ligand Deltalike1, result in the loss of MPCs and muscles during development [51, 82]. Therefore, common signalling molecules act on embryonic and adult MPCs to regulate their fate. How and when this signalling intervenes during lineage progression from stem cells is unclear.

Gene disruptions have revealed a greater complexity than previously anticipated, and they have exposed fundamental differences between developmental and regenerative myogenesis. For example, in contrast to prenatal development, the conditional inactivation of Myogenin in the adult does not result in a dramatic skeletal muscle phenotype. Therefore, Myogenin is not essential for postnatal life [28, 31]. Like their embryonic counterparts, adult Myogenin null myoblasts do differentiate in vitro. However, Myogenin has different target genes in the adult compared to the embryo [28], thereby highlighting its distinct roles in embryonic versus regenerative myogenesis. Thus, the adult stem/progenitor cells and their committed progeny have developed other regulatory pathways to promote lineage progression and differentiation, and importantly, these are distinct from those exerted during development.

Another significant finding that was reported recently could alter our views on how Pax7 (and Pax3) regulate myogenesis, whilst underscoring important distinctions that can be made between development and adult muscle gene regulation (Figs. 1, 2). Although Pax3/Pax7 play crucial roles early during development, conditional inactivation of these genes in the adult results in normal myogenesis and regeneration after injury of the tibialis anterior muscle [83]. Strikingly, conditional inactivation of these genes prior to satellite cell quiescence during perinatal stages up to postnatal day 21 resulted in profound deficits in skeletal muscle regeneration. This phenotype became progressively milder as the mouse pups approached 21 days in age [83]. These findings raise a number of important issues. First, as Pax7 (Pax3) appears not to be a key regulator of adult muscle regeneration and satellite cell self-renewal, which genes play this role? Amongst those that have been reported, Foxk1 [84] and Sox15 [85] affect skeletal muscle regeneration as well as satellite cell numbers, albeit not all satellite cells are lost in these mutants, suggesting that other regulators also intervene. A second point concerns the distinction between prequiescent and postquiescent muscle progenitors that appear after birth. As noted above, perinatal muscle progenitors were called ‘emerging’ juvenile satellite cells due to uncertainties regarding their true nature [3]. What proportion of this cell population corresponds to foetal myoblasts as opposed to future satellite cells? The original description of the satellite cell [86] and subsequent studies, lead to the notion that satellite cells are sublaminar and quiescent. The conditional inactivation of Pax7 (Pax3) in adult mice [83] now provides functional evidence that supports this view and highlights the differences between perinatal progenitors that continue to divide, and those that are quiescent in the adult.

These findings also permit a reinterpretation of the germline Pax7 null mutant phenotype that was described previously, and the role that this gene plays in regulating juvenile myogenic cells perinatally. Prior studies provided convincing evidence that the numbers of juvenile myogenic cells declined significantly after birth, due in part to apoptosis in this cell population [20, 22, 23, 87]. However, it was unclear from those studies, due to the lack of specific markers, which of the presumed subpopulations indicated above (Fig. 1) were lost, and whether future satellite cells, which emerge from a subset of the juvenile myogenic cell pool, resist the loss of Pax7. Secondly, postnatal Pax7 null progenitors were suggested to undergo accelerated differentiation [83]. Perhaps in this case less myoblasts are generated due to a more rapid transition to the differentiated state. Alternatively, Pax7 regulates cell cycle kinetics and it may be necessary for the amplification of progenitors and/or myoblasts. These points need to be clarified in future studies.

Building muscle in the embryo versus adult muscle regeneration

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Establishing skeletal muscle in the embryo from founder stem cells
  5. Genetic strategies to identify prenatal skeletal muscle stem and progenitor cells
  6. Genetic strategies to identify postnatal skeletal muscle stem and progenitor cells
  7. Distinguishing juvenile from adult satellite cells and embryonic founder stem cells
  8. Building muscle in the embryo versus adult muscle regeneration
  9. The skeletal muscle stem cell niche and signalling pathways
  10. Conclusions and perspectives
  11. Conflict of interest statement
  12. Acknowledgements
  13. References

During embryogenesis, an anlage (the founding element for a tissue or organ) forms first in essentially all sites where skeletal muscle is established. Subsequently, Pax7-expressing stem/progenitor cells emerge in these locations, and this marker identifies these cells throughout the organism from this period onwards (Fig. 1). Prenatal muscle development occurs in multiple waves, where embryonic (primary) fibres act as a scaffold for the formation of foetal (secondary) fibres, the latter comprising the bulk of future adult muscles after cell fusion and hypertrophy [4, 64, 88] (Fig. 1). There is an important distinction to be made here between this strategy, in building the tissue during development, and the task of doing so during regeneration – the latter does not pass through an intermediate phase of primary fibre production. Given these diverse strategies, how then are regenerating fibres built in the absence of a primary fibre scaffold, to ultimately generate exquisitely aligned myofibres ready for the task ahead? One possibility is that a scaffold is not required and somehow myofibres self-organize. If that is the case, what determines this organization, the number of myofibres generated and the fusion index (number of myofibre nuclei)? Is there an upper and lower limit to the number of myoblasts that fuse with a fibre? A model which eliminates primary muscle fibres but allows foetal myogenesis to proceed would be interesting to investigate these questions.

Given that myofibre organisation can be achieved during muscle regeneration, it is possible that the basement membrane, which surrounds each myofibre, plays a determinant role in this process. This lattic structure, which is composed of collagen, laminins and extracellular matrix molecules, does not get eliminated after muscle injury ([2]; F. Chrétien, B. Gayraud-Morel and ST, unpublished observations). This lattice network of extracellular matrix also delimits the zone in which future myofibres will be generated, and perhaps it also regulates myofibre number to some extent.

After trauma or during disease, skeletal muscle degenerates and myofibre necrosis ensues [2, 4, 12, 89]. Concurrently, three distinct phases characterize the regeneration process: inflammation, tissue reconstruction and tissue remodelling. The regenerate is almost perfectly functional, with some scarring or fibrosis depending on the extent of the injury [89]. Interestingly, during muscle regeneration embryonic and neonatal myosin heavy chains are re-expressed briefly, but in ‘secondary’ fibres [2, 4, 11, 64]. Whether the same regulators that govern their expression during embryonic development, also do so during regeneration is not known.

Tissue reconstitution after injury involves repairing or replacing the supporting structures including vasculature, connective tissue and interstitial cells which are poorly characterized. Less is known about how this network of associated cells coordinates myofibre growth and homeostasis. As in the embryo, neovascularization occurs very early and it is a critical feature of the regenerate, providing essential blood cells, macrophages and molecules. A major objective is maintaining cell survival in a hostile necrotic environment for satellite cells and other cell types to proliferate, differentiate and self-renew. Macrophages play a key role in this process by eliminating debris in the first few days after severe muscle injury; then they play a permissive role for myoblast proliferation and survival [90–92]. Timing is also crucial, as most of the proliferation, at least of satellite cells, occurs within the first 5 days after acute injury (V. Shinin, P. Rocheteau and S Tajbakhsh, unpublished observations).

Heterotopic transplantation assays in the adult provide a useful paradigm to address how satellite cells of diverse ontological origin behave in a foreign muscle environment. Recently, an analysis of satellite cells isolated from adult mice demonstrated that the molecular signature of adult progenitors reflects their developmental history [36, 39]. First, this finding is in accordance with the notions discussed above, indicating that skeletal muscle stem cells are not all alike, but that they retain distinct molecular signatures from their time of birth, and in the adult. The extent to which this initial molecular signature affects the differentiated phenotype compared to the phenotypic changes that occur as muscles are remodelled postnatally (contractile activity via nerve, hormonal control, etc) is not clear. The heterotopic transplantation of satellite cells from one location to another one which is distinct, addresses this issue. In spite of the initial differences in properties of muscle stem/progenitor cells, when EOM [39] or masseter [36] satellite cells are transplanted into a preinjured limb muscle environment a robust contribution of donor cells to the regenerate takes place. Notably, however, the heterotopic transplantations of EOM satellite cells to limb muscle failed to generate eye muscle specific markers [39]. This observation indicates that satellite cells alone cannot assure the original muscle phenotype, suggesting that they are reprogrammed in the ectopic location. Another issue in this context concerns cell plasticity. Remarkably, PA- but not limb-derived adult satellite cells express cardiogenic markers after exposure to bone morphogenetic protein (BMP). Interestingly, this outcome indicates that a developmental plasticity persists in these adult progenitor cells, which is reminiscent of their ontogeny, and this, in spite of their residence in a skeletal muscle environment [36].

The skeletal muscle stem cell niche and signalling pathways

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Establishing skeletal muscle in the embryo from founder stem cells
  5. Genetic strategies to identify prenatal skeletal muscle stem and progenitor cells
  6. Genetic strategies to identify postnatal skeletal muscle stem and progenitor cells
  7. Distinguishing juvenile from adult satellite cells and embryonic founder stem cells
  8. Building muscle in the embryo versus adult muscle regeneration
  9. The skeletal muscle stem cell niche and signalling pathways
  10. Conclusions and perspectives
  11. Conflict of interest statement
  12. Acknowledgements
  13. References

The niche is generally defined as the microenvironment in which tissue-specific stem cells reside [93–95]. Here, it is appropriate to ask whether one exists in the embryo or if this is a unique feature of the postnatal animal. The appearance of the basement membrane 2–3 days before birth in the mouse (Fig. 1) may be a defining step in formulating the stem cell niche for skeletal muscle. However, the emerging satellite cell population continues to proliferate actively and it will not enter quiescence until about 2–3 weeks after birth. When this occurs, it is not clear whether satellite cells stop to migrate along the myofibre, or if they retain some motility as quiescent cells. If the latter is true, the location of the niche will be in constant flux.

The satellite cell niche is asymmetric in nature, where one side of the satellite cell is in contact with the plasmalemma, and the other side, with the basement membrane. Different molecules that can potentially exert differential cell behaviour [59, 68, 69] reside in these structures and they may be important in defining the orientations of cell division of the satellite cell: perpendicular or planar (see below).

Endothelial cells or pericytes of the vasculature might play a key role in the satellite cell niche, as over 95% of satellite cells are subjacent to a vessel, and separated from it by the basement membrane [96] (Fig. 5, and insets). The neural and haematopoietic niches also have a vascular component, although in the latter, its significance is currently being debated [93, 95]. Cell–cell contacts via extracellular molecules are key features of well-defined niches [95]. M-Cadherin [97], and integrins (α1, α7, β1), and Syndecans 3 and 4 (co-receptors for tyrosine kinases) [98] are examples of molecules that link the satellite cell with its micro-environment. CD34, which is implicated in cell adhesion and signalling, is used routinely to isolate blood stem cells, and it is also expressed on satellite cells [76, 99].

Extrinsic signals in the form of growth factors and receptor–ligand interactions mediated by cell–cell contacts will eventually lead to the intrinsic control of cell behaviour. To what extent intrinsic signals operate autonomously in satellite cells that are separated from their niche is not known. Some properties, such as asymmetric cytoplasmic marker segregation, as well as template DNA strand segregation can occur independently of the niche, at least to a limited extent [72, 100]. This suggests that continued extrinsic cues are not essential to mediate these phenotypes, at least in the short term. Factors that activate satellite cells are generally better understood than those that maintain their quiescence. For example, Sf/Hgf can potentially regulate myoblast proliferation and differentiation, probably via the Met receptor (see [2, 101]). As indicated above, Sf/Hgf and Met govern myogenic cell migration from the somites to other locations [8, 52]. Fgfr2 and Fgfr4 (fibroblast growth factor receptors) are expressed in skeletal muscles of the embryonic somite as well as in satellite cells (see [2]). The Fgf ligands are expressed by their associated differentiated muscle fibres [102]. These factors, in addition to those released by endothelial cells, promote the survival and proliferation of myogenic cells [96]. Given these observations, it is unexpected that satellite cells remain quiescent and unresponsive to these growth factors during muscle homeostasis. One possibility is that potent repressors prevent the relay of intracellular signalling. Alternatively, the receptors are not occupied by the ligand, the latter being sequestered by the basal lamina in the niche, until muscle injury releases sequestered ligand and allows signalling to proceed.

Notch is another regulator of satellite cells where a role in the expansion of the myoblast pool was suggested. This brief requirement for Notch activity appears to be followed by canonical Wnt signalling to promote differentiation [79]. Therefore, in adults, as in the embryo where muscle masses were reduced and Pax3/Pax7 stem/progenitor cells were lost in the absence of Notch function, Notch plays a critical role. In addition, when Notch activity is impaired in aged muscle due to a defect in signalling by its ligand Delta muscle regeneration is compromised [103]. Interestingly, satellite cell numbers were reported to decline in aged mouse [104, 105] and human [106] skeletal muscle. Nevertheless, those that remain tend to maintain their intrinsic capacity to regenerate muscle as efficiently as satellite cells from younger mice [104]. These experiments underscore the importance of extrinsic and intrinsic signalling and also raise the possibility that the nature of the niche may be changing during ageing.

Some questions remain unanswered concerning Notch function in the embryo and adult: is Notch signalling required in the stem cells for regulating this cell state, or immediately after, in progenitor cells during cell commitment? Which Notch receptors, ligands and downstream effectors intervene in these decisions, and what is the mechanism of action? Notably, Notch3 was expressed in higher levels in the more ‘stem-like’ Pax7+/(Myf5−) satellite cells whereas Notch1 was preferentially expressed in the more committed Pax7+/(Myf5+) cells, suggesting that a subtle interplay between Notch mediators might be regulating this cell population [68].

Transcription factors tend to be degraded during mitosis, therefore it has not been possible to monitor their segregation to distinct daughter cells during mitosis. Nevertheless, the asymmetric distribution of Myod [63] and Pax7 to distinct daughter cells after mitosis has also been reported [68] (B. Gayraud-Morel and ST, unpublished data), suggesting that those cells that are Myod-negative correspond to self-renewing satellite cells. As mentioned above, both asymmetric and symmetric types of divisions involving these transcription factors, and a genetic readout of Myf5 expression (YFP+ or YFP− or Myf5nlacZ), suggests that self-renewal and exponential growth of satellite cells occur after muscle injury. It is not clear how this balance is regulated in the organism, but Wnt signalling appears to play a role as Wnt7a null mice have fewer satellite cells and symmetric type divisions on isolated myofibres. In contrast, overexpression of Wnt7a results in more satellite cells being produced [107]. Interestingly, Wnt7a is part of the noncanonical signalling pathway, in this case acting through the Fzd7 receptor and Vangl2, which mediates the downstream effects in this planar cell polarity signalling pathway [107]. In the embryonic myotome, planar cell polarity also plays a role in myotome formation, but in this case Wnt11 is the effector molecule [108]. The basal lamina of the basement membrane and the orientation of the mitotic spindle [109] probably play critical roles in mediating the distribution of cell fate regulators selectively to daughter cells in the adult, and perhaps also in the embryo. Whether myoblast numbers and self-renewal of stem/progenitor cells in the embryo and adult are regulated similarly remains to be determined.

Conclusions and perspectives

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Establishing skeletal muscle in the embryo from founder stem cells
  5. Genetic strategies to identify prenatal skeletal muscle stem and progenitor cells
  6. Genetic strategies to identify postnatal skeletal muscle stem and progenitor cells
  7. Distinguishing juvenile from adult satellite cells and embryonic founder stem cells
  8. Building muscle in the embryo versus adult muscle regeneration
  9. The skeletal muscle stem cell niche and signalling pathways
  10. Conclusions and perspectives
  11. Conflict of interest statement
  12. Acknowledgements
  13. References

A number of cellular and molecular strategies that regulate the establishment of skeletal muscles in the embryo, and the regeneration of the muscle tissue after trauma, are shared in these two contexts. However, some notable differences between these two programmes highlight the unique requirements for each scenario. From the architectural viewpoint, constructing the tissue in the embryo occurs after a scaffold is established – initially an anlage forms, then 1° fibres that contain a few myonuclei, followed by 2° fibres which constitute the bulk of the tissue. During this period, connective tissue and vasculature integrate into the muscle masses to sculpt and sustain the tissue. Adult regeneration occurs somewhat differently where the basement membrane which forms a lattice structure might provide a natural network in which nascent myofibres can be established. Although this process apparently short-circuits the formation of an anlage and 1° fibres, interestingly, embryonic muscle markers are re-expressed for several days following adult skeletal muscle regeneration.

Intriguingly, many of the genes that play critical roles in determining stem cell identity, myogenic fate and cell differentiation are redeployed during adult regenerative myogenesis, yet in some cases, null mutants produce less severe, or no phenotypes. Some examples include Pax3, Pax7, Myf5 and Myogenin mutants. Therefore, an undefined gene regulatory programme in the adult compensates for the loss of these genes that play critical roles in myogenesis prenatally. The unravelling of these networks will be a major area of interest in the future.

Finally, recent genetic studies described in this review underscore the importance of distinguishing embryonic, foetal, perinatal and adult stem/progenitor cells (Fig. 1). Notably, during perinatal growth, a juvenile satellite cell population exhibits cellular and gene regulatory heterogeneity that is distinct from adult quiescent satellite cells. Future investigations in this area should lead to the identification and characterization of skeletal muscle stem cells, and their modes of division and self-renewal.

Acknowledgements

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Establishing skeletal muscle in the embryo from founder stem cells
  5. Genetic strategies to identify prenatal skeletal muscle stem and progenitor cells
  6. Genetic strategies to identify postnatal skeletal muscle stem and progenitor cells
  7. Distinguishing juvenile from adult satellite cells and embryonic founder stem cells
  8. Building muscle in the embryo versus adult muscle regeneration
  9. The skeletal muscle stem cell niche and signalling pathways
  10. Conclusions and perspectives
  11. Conflict of interest statement
  12. Acknowledgements
  13. References

I thank members of the laboratory including R. Sambasivam and I. Le Roux for critical comments, and F. Chrétien, W. McCord and V. Shinin for the images, S. Oliver for the artwork, and F. Chrétien for advice for Figure 5. I acknowledge also the support from the Institut Pasteur, Association Française contre le Myopathies, Agence Nationale de la Recherche (ANR-06-BLAN-0039), MyoRes (EU FP 6) EuroSysStem (EU FP 7) and the Fondation pour la Recherche Medicale.

References

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Establishing skeletal muscle in the embryo from founder stem cells
  5. Genetic strategies to identify prenatal skeletal muscle stem and progenitor cells
  6. Genetic strategies to identify postnatal skeletal muscle stem and progenitor cells
  7. Distinguishing juvenile from adult satellite cells and embryonic founder stem cells
  8. Building muscle in the embryo versus adult muscle regeneration
  9. The skeletal muscle stem cell niche and signalling pathways
  10. Conclusions and perspectives
  11. Conflict of interest statement
  12. Acknowledgements
  13. References