Extrinsic regulation of satellite cell specification

Cellular commitment during vertebrate embryogenesis is controlled by an interplay of intrinsic regulators and morphogenetic signals. These mechanisms recruit a subset of cells in the developing organism to become the ancestors of skeletal muscle. Signals that control progression through the myogenic lineage converge on a battery of hierarchically organized transcription factors which modulate the cells to either remain in a primitive state or allow their commitment and differentiation into skeletal muscle fibers. A small population of cells will retain a largely unspecified state throughout development. Such stem cells, in conjunction with more committed myogenic progenitors, form a heterogeneous population that colonizes adult skeletal muscle as satellite cells. The satellite cell pool is responsible for the remarkable regenerative capacity of skeletal muscle. Similar to their counterparts during embryonic development, satellite cells are capable of self-renewal and can give rise to myogenic progeny. Impaired satellite cell homeostasis has been associated with numerous muscular disorders. Due to intense research efforts in the past two decades, the complex biology of muscle stem cells has now revealed some of its secrets and new avenues for the development of therapeutic molecules have emerged. In the present review we focus on the extrinsic mechanisms that control self-renewal, specification and differentiation of satellite cells and their significance for the development of biologic drugs.


Introduction
Myogenic specifi cation initially takes place in the somites of the developing vertebrate embryo and is thenceforth reiterated throughout the life of the organism [1]. Th is process will establish and maintain one of the major constituents of the body: skeletal muscle. Th e presence of tissue-specifi c stem cells, the satellite cells, gives adult muscle the capacity for extensive regeneration in response to trauma and disease [2]. Despite a fundamentally diff erent hormonal and anatomical environment, muscle regeneration in the adult organism recapitulates many aspects of embryonic myogenesis [3]. However, the capacity of adult muscle for regeneration seems to be limited and repeated degeneration is accompanied by increasingly ineffi cient tissue reconstitution [4]. Since the discovery of the satellite cell 50 years ago, research has provided valuable insights into the molecular mechanisms that regulate the satellite cell pool and ultimately the potential for regenerative myogenesis [5,6]. Particularly, a recently discovered subpopulation of satellite cells with an extensive capacity for self-renewal and the characterized signaling molecules that control these cells hold great potential for therapeutic manipulation [7,8].

Developmental myogenesis
Skeletal muscle in all vertebrates originates from cells found in the mesoderm, one of the three primary germ layers [9,10]. Parts of the mesoderm give rise to segmented clusters called somites, which are aligned along the anterior-posterior axis of the embryo. Th e somites, the paraxial head mesoderm and the prechordal mesoderm are the source of primitive myogenic cells, most of which are marked by the expression of two paired-box (Pax) transcription factors, Pax3 and Pax7. Later during development, a subpopulation of these cells will diff er en tiate into terminally committed myocytes. Th e embryonic body axes then orient the fusion of these cells, generating the fi rst multinucleated myofi bers. In several subsequent waves, more embryonic myocytes align and fuse into precisely arranged postmitotic muscle fi bers that will give rise to the organism's skeletal muscle. Limb, trunk and some head muscles arise from cells of somitic origin, whereas the remainder of the head muscles derive from cells of the paraxial head mesoderm and the prechordal mesoderm [1,[11][12][13][14][15].
Myogenic specifi cation during development is regulated by signaling factors released from the surrounding tissue. Among such factors are sonic hedgehog (Shh),

Abstract
Cellular commitment during vertebrate embryogenesis is controlled by an interplay of intrinsic regulators and morphogenetic signals. These mechanisms recruit a subset of cells in the developing organism to become the ancestors of skeletal muscle. Signals that control progression through the myogenic lineage converge on a battery of hierarchically organized transcription factors which modulate the cells to either remain in a primitive state or allow their commitment and diff erentiation into skeletal muscle fi bers. A small population of cells will retain a largely unspecifi ed state throughout development. Such stem cells, in conjunction with more committed myogenic progenitors, form a heterogeneous population that colonizes adult skeletal muscle as satellite cells. The satellite cell pool is responsible for the remarkable regenerative capacity of skeletal muscle. Similar to their counterparts during embryonic development, satellite cells are capable of self-renewal and can give rise to myogenic progeny. Impaired satellite cell homeostasis has been associated with numerous muscular disorders. Due to intense research eff orts in the past two decades, the complex biology of muscle stem cells has now revealed some of its secrets and new avenues for the development of therapeutic molecules have emerged. In the present review we focus on the extrinsic mechanisms that control self-renewal, specifi cation and diff erentiation of satellite cells and their signifi cance for the development of biologic drugs.
which is released from the neural tube, lateral mesoderm derived bone morphogenetic proteins (BMPs), and members of the wingless-type MMTV integration site (Wnt) family of proteins, which emanate from both the ectoderm and the neural tube [1]. On the genetic level, myogenic determination is modulated by Pax3/Pax7 and a family of transcription factors known as myogenic regulatory factors (MRFs) [6].
In the perinatal period, the niche in between the basal lamina and the muscle fi ber membrane is populated by juvenile satellite cells that proliferate extensively. A subset of theses cells will remain as quiescent satellite cells in the adult organism [16].

Satellite stem cells
With the exception of some head muscles, satellite cells in the adult are generally considered to be the progeny of Pax3-and Pax7-expressing cells of somitic origin [11][12][13][14]17]. Pax transcription factors are genetic master switches that can imprint stem cells towards a myogenic fate but repress genes involved in diff erentiation. All adult satellite cells are marked by the expression of Pax7 whereas Pax3 is postnatally down regulated in most muscles [18]. Other molecular markers of satellite cells include vascular cell adhesion molecule 1 (VCAM-1), c-Met (receptor for hepatocyte growth factor), chemokine C-X-C motif receptor 4 (CXCR4), M-cadherin, neural cell adhesion molecule 1 (NCAM1), forkhead box protein K1 (Foxk1), integrin α7β1, CD34, and syndecans 3 and 4 [19].
In adult skeletal muscle only a small subpopulation of Pax7-expressing satellite cells derives from a lineage that has never expressed myogenic factor 5 (Myf5), a transcription factor belonging to the MRFs. It has been demonstrated that these cells are capable of extensive self-renewal and can very effi ciently repopulate their niche in transplantation experiments into satellite-celldepleted muscle. To date, this cell type makes up the most primitive and stem-like population that has been identifi ed in adult muscle tissue and these cells are therefore referred to as 'satellite stem cells' . Conversely, a cell that expresses Myf5 or is descended from an ancestor that expressed this factor once is more prone to diff erentiation and is therefore termed a 'satellite myogenic cell' [7].

Regenerative myogenesis
Injury of adult muscle causes an infl ux of infl ammatory cells that remove necrotic debris from the tissue. Subsequently, Pax7-expressing satellite cells enter mitosis to generate progeny that will go through repeated rounds of proliferation and then migrate to the site of damage. A high percentage of this progeny will undergo myogenic diff erentiation in order to restore the destroyed muscle fi bers, whereas others will self-renew and, upon complete regeneration, repopulate the muscle as satellite stem cells [2].
Once activated by muscle injury, proliferating satellite cells become myoblasts through upregulation of a MRF called myoblast determination protein (MyoD) [18]. MyoD drives proliferation by controlling reentry into the cell cycle and activates the transcription of musclespecifi c genes [20]. As a last step, downregulation of Pax7 and upregulation of myogenin primes the myoblasts to become myocytes. Th ese cells are terminally committed, exit the cell cycle and fuse with other myoblasts or existing fi bers. Th is process will fi nally repair the damaged muscle tissue [6]. Satellite cell self-renewal and the transcription factors controlling lineage progression during regeneration are thought to be regulated by a variety of extrinsic cues [19]. For the remaining part of this review, we will focus on these factors.

The satellite cell niche
A stem cell niche is defi ned as a specifi c anatomical location that participates in tissue generation, maintenance and repair. Stem cells reside in their niche for an indefi nite period and it protects its host from depletion or uncontrolled proliferation [21].
Th e satellite cell niche is exceptionally complex and the sources of environmental infl uences are diverse. For instance, satellite cells are often localized in close proximity to capillaries, which might be a means to effi ciently supply them with signaling factors [22]. Furthermore, several cell types, such as fi broblasts and immune cells, can colonize muscle tissue. Th ese cells secrete cytokines that may infl uence satellite cells [19]. Neural input leads to depolarization of the muscle fi ber, which can aff ect satellite cells through paracrine factors and adhesion molecules [23,24].
Th e basement membrane and the muscle fi ber sarcolemma, in-between which the satellite cell is wedged, are the main anatomical hallmarks of its niche. Th e function of the extracellular matrix (ECM) for satellite cells during myogenesis is a matter of ongoing investigation. It has been documented that certain ECM proteins like laminin and collagen are reciprocally regulated by fi bronectin, hyaluronic acid and tenascin during muscle regeneration [25]. Fibronectin inhibits the diff erentiation of cultured myoblasts whereas laminin promotes it [26,27]. Mice suff ering from muscular dystrophy, which is caused by a null mutation in the laminin alpha 2 chain, display dramatic defects in muscle regeneration [28]. Furthermore, knockout of the matrix modifying enzyme membrane type 1 metalloprotease (MT1-MMP) in mice results in impaired skeletal muscle recovery after injury [29]. Th is evidence demonstrates that an intact ECM is essential for muscle repair and suggests that transitional changes in its composition provide instructive cues to satellite cells. Next to its structural composition, the niche controls satellite cells through several signaling molecules that emanate from a wide range of sources ( Figure 1).

Signaling factors regulating satellite cells
Th e amount of satellite cells decreases with age, although their myogenic potential does not diminish throughout the life of the organism [30]. Th is notion has been supported by experiments demonstrating a signifi cant myogenic capacity of satellite cells transplanted from old tissue into young muscle. On the other hand, when young cells were transplanted into old muscle, they did not perform well [31,32]. Furthermore, exposure of old mice to serum from young mice dramatically improved their regenerative capacity [33]. Taken together, this evidence indicates that satellite cells are highly regulated by extrinsic cues. Th e identifi cation and understanding of these factors will open new avenues for the development of therapeutic strategies. Drugs with the ability to increase or restore the regenerative potential of skeletal muscle by boosting satellite cell function or number could help to preserve muscle mass in degenerative muscular disorders.

Development of biologics for therapeutic manipulation of satellite cells
Satellite stem cell transplantation could theoretically be a promising approach to restore or enhance the regenerative potential of diseased muscle. In reality, such cellbased approaches face serious limitations, including the need to cultivate satellite cells, their incompatibility with systemic delivery, and their poor survival following intramuscular injection [34]. For these reasons, alter na tive approaches need to be taken into consideration. Th e existence of numerous physiological factors that are involved in the regulation of satellite cells provides an opportunity for the development of drugs that mimic or interfere with these molecules [18].
Biologic drugs are pharmaceuticals inherently biological in nature and manufactured using biotechnology [35]. Th e advantages of biologics over classic small molecular compounds are that they harness the same principles employed by endogenous proteins: high functional specifi city resulting in a well defi ned biological eff ect with little or no off -target activity [36]. In the following sections we discuss some approaches for the rational design of biologics mimicking or inhibiting signaling molecules that control satellite cell function.

Transforming growth factor-β
An impaired regenerative capacity and chronic infl amma tion with hyperplasia of the interstitial connective tissue are pathological hallmarks of muscular dystrophy. Infl ammation of dystrophic muscle is dominated by macrophages and T lymphocytes that secrete pro-fi brotic cytokines. Th is causes a gradual development of fi brosis, which hinders muscle regeneration and ultimately leads to incomplete functional recovery [37]. Consequently, anti-infl ammatory and immunosuppressive drugs such as corticosteroids and cyclosporine are benefi cial for these conditions. However, weight gain and infections are common side eff ects of these pharmaceutical agents [38]. Th is provides a rationale for the development of therapeutics that target the molecular pathways involved in the fi brotic muscle pathology more specifi cally.
Transforming growth factor-β (TGF-β), a cytokine that is released by infl ammatory cells, has become a focus of attention in research on such signaling mechanisms [39]. Its expression is dramatically increased in muscular dystrophy [40]. Stimulation of muscle resident fi broblasts with this factor increases their collagen and fi bronectin production, which results in increased interstitial fi brosis [41][42][43]. Furthermore, TGF-β stimulates the diff erentiation of cultured myoblasts into a fi brogenic cell type [44]. Th is suggests that TGF-β also drives satellite cells into this alternative lineage, which would ultimately deplete them from muscle and prevent effi cient regeneration. More over, an increase in fi brogenic cells could further exacerbate muscle pathology by contributing to tissue fi brosis.
Inhibition of TGF-β or its downstream eff ectors by large molecule ligand traps and antisense oligo nucleotides are novel therapeutic approaches being explored today [45]. It is promising that some of these therapeutics have already been demonstrated to ameliorate muscular dystrophy in certain mouse models [46]. Interestingly, a recent study has demonstrated a dose-dependent requirement of cultured myoblasts for TGF-β. Th is suggests that specifi c concentrations of TGF-β have permissive infl uences on satellite cells and it raises some concerns about the therapeutic window for inhibitors of this factor [47].

Myostatin
Myostatin has been demonstrated to be a powerful antagonist of muscle growth. Inhibition or genetic ablation of myostatin triggers dramatic increases in skeletal muscle mass across many diff erent species [48]. Th e total number of muscle fi bers per muscle is increased in myostatin knockout animals [49,50]. Th is is indicative of increased myogenic activity. Furthermore, the application of myostatin to cultured myoblasts prevents diff erentiation by suppressing MRFs. Conversely, silencing of myostatin in myoblast culture increases diff erentiation [51].
Muscle fi broblasts do express the myostatin receptor at high levels and its activation induces their proliferation and the secretion of fi brotic ECM proteins [52]. Myostatin inhibition could therefore be a means to improve regeneration of dystrophic muscle by reducing fi brosis while, at the same time, promoting the activation of satellite cells. Indeed, it has been demonstrated that myostatin defi ciency or systemic application of inhibitory antibodies slows down the degeneration in dystrophic muscle of mdx mice [53].
Th e direct relevance of myostatin for satellite cells remains a matter of ongoing investigation. It has been reported that satellite cells of myostatin-null mice proliferate more in the resting state while the total number of satellite cells per muscle fi ber increases. Further more, the regenerative potential of myostatin knockout muscle seems to be superior to the wild-type counterpart [51]. Intriguingly, a recent report rebutted that a myostatin defi ciency is benefi cial for mdx mice and also excluded a direct eff ect of myostatin on satellite cells [54]. Th is study challenges several previous reports and it will be interesting to follow the future debate on this topic.
A number of circulating factors that control myostatin activity have been discovered. Among these is follistatin, which can function as a potent myostatin antagonist. Overexpression of follistatin in mice causes muscle gains beyond myostatin inhibition, which suggests that it inhibits additional negative regulators of muscle growth and/or satellite cells [51]. Similar to lowering myostatin levels, delivery of follistatin to dystrophic mice reversed muscle pathology and improved strength [55].
Despite the controversy regarding which cell type mediates the eff ects of myostatin on muscle, antagonists that target this factor seem to be promising candidates for the treatment of several muscular disorders. Future studies will have to clarify the potential of myostatin inhibitors for direct therapeutic manipulation of satellite cells.

Insulin-like growth factor-1
Insulin-like growth factor 1 (IGF-1) is either secreted by the liver as an endocrine hormone or produced locally by other tissues where it can act in a paracrine/autocrine fashion [56]. Exogenous IGF-1 or genetic overexpression results in increased muscle mass and enhanced regeneration in mice. Furthermore, IGF-1 levels are upregulated in regenerating muscle and hypertrophy induced by this factor not only involves an increase in protein synthesis but also augments the DNA content of muscle [51]. Th is evidence suggests that IGF-1 supports the mobilization of satellite cells for muscle regeneration.
After stimulation, muscle tissue expresses high levels of a splice variant of IGF-1, which is termed mechanogrowth factor (MGF) [57]. MGF promotes the proliferation and inhibits the diff erentiation of cultured myoblasts [58]. Th e main circulating isoform of IGF-1 has the same eff ect on proliferation but also facilitates diff erentiation [59][60][61]. Th is indicates that the paracrine/autocrine eff ects of MGF rather expand the satellite cell pool, while other IGF-1 variants are generally activating and allow diff erentiation. In agreement with this idea, addition of a synthetic MGF peptide increased the number of desminpositive myogenic cells isolated from healthy and diseased muscles [62]. However, because desmin is expressed by quiescent and activated satellite cells, MGF's potential for a sustained expansion of the satellite cell pool remains to be demonstrated [63]. Interestingly, mitotically active young muscle has been found to be most responsive to MGF, which argues against a direct eff ect on quiescent satellite cells [64].
In summary, there is evidence that both IGF-1 and MGF could be used to stimulate satellite cell function and proliferation under pathologic conditions. Th e fi rst recombinant IGF-1 drugs have recently received US Food and Drug Administration (FDA) approval for the treatment of IGF-1 defi ciency, and intense eff orts for the expansion of its use in diseases such as sarcopenia, muscular dystrophy and amyotrophic lateral sclerosis are ongoing [65][66][67].

Hepatocyte growth factor
Th e expression of hepatocyte growth factor (HGF) spikes during the early phase of muscle regeneration and decreases subsequently during regeneration. HGF is sequestered in the ECM of skeletal muscle and released upon injury. Th is factor can stimulate the proliferation of cultured myoblasts while inhibiting diff erentiation [2]. In agreement with these fi ndings, the application of recombinant HGF to injured muscle slows regeneration while increasing the number of activated satellite cells [68]. Th is suggests that HGF serves to activate quiescent satellite cells in the immediate phase after injury while blocking diff erentiation. Once the pool of proliferating satellite myogenic cells is suffi ciently expanded in the later stages of muscle regeneration, HGF levels decline and diff erentiation is initiated. It remains to be determined whether treatment with HGF could be used to mobilize satellite cells in situations of inactivity, such as atrophy or cachexia.
It has been demonstrated that nitric oxide (NO) signaling may act upstream of HGF to modulate the activation state of satellite cells. NO seems to augment active HGF by triggering its release from the ECM through metalloproteinases [69]. Furthermore, increased levels of NO promotes regeneration of normal and dystrophic muscle [70,71]. Th e delivery of pharmaceutical compounds to increase NO signaling in diseased skeletal muscle is feasible and candidate drugs are currently being investigated [72].

Fibroblast growth factor
Several fi broblast growth factors (FGFs), particularly FGF-6 and FGF-2, have been demonstrated to induce the proliferation of cultured myoblasts while inhibiting their diff erentiation [2,73]. Th is indicates a role for these factors in the expansion of the satellite cell compartment. For instance, FGF-6 expression is muscle specifi c and is increased during regeneration [74]. However, results from studies of injury-challenged muscle in FGF-6defi cient mice are controversial. Some reports demonstrated impaired regeneration whereas others could not confi rm such eff ects [74,75].
FGF2 is sequestered in the basement membrane surrounding developing and adult myotubes, and neutrali zing antibodies against FGF2 seem to delay or prevent muscle regeneration in the immediate period after injury [76,77]. Furthermore, FGF2 appears to facilitate satellite cell divisions and muscle regeneration in dystrophic mice [78]. Moreover, the combined loss of the FGF2 and FGF6 genes increases the dystrophic pathology in the musculature of mdx mice, whereas the transgenic delivery of both factors to damaged muscle enhances regeneration [79,80].
Th erapeutic recombinant FGFs or biologically active derivatives could potentially be used to enhance the regenerative potential of muscle by increasing satellite cell number. However, further studies will have to clarify whether FGF treatment leads to the therapeutically desirable expansion of the satellite stem cell compartment as opposed to the presumably rather transient eff ects caused by an increase in satellite myogenic cells.

Wnt
Wnt proteins are secreted lipid-modifi ed glycoproteins that act through Frizzled (Fzd) receptors. Mammals harbor 19 diff erent wnt and 10 fzd genes. Historically, Wnt signaling has been divided into 'canonical' β-catenin-dependent and 'noncanonical' β-catenin-independent signaling path ways. More recently, this classifi cation has been undermined and crosstalk between the canonical and noncanonical pathways has been described [81].
Given the variety of Wnts, it is not surprising that different species can excerpt both inhibitory and per missive eff ects on myogenic diff erentiation. For instance, some Wnt molecules facilitate myogenesis during regenera tion while others have been demonstrated to drive the diff erentiation of satellite cells into a fi brotic lineage [82,83].
It has been reported that a temporal switch from activation of the Notch pathway to increased Wnt3a signaling is required for myogenic lineage progression and consequently for eff ective muscle regeneration. Th e eff ects of Wnt3a and Notch signaling are mediated by the modulation of the common intracellular eff ector glycogen synthase kinase 3β (GSK3β) [84].
A recent study demonstrated that components of the noncanonical planar cell polarity (PCP) pathway regulate the expansion of satellite stem cells during muscle regeneration in concert with Wnt7a. Wnt7a is expressed by muscle fi bers in the immediate period after myocyte fusion. Th is suggests that Wnt7a release is a physiological means to expand the satellite stem cell pool after the initial phase of myogenesis. In agreement with this theory, muscle-injury-challenged Wnt7a-defi cient mice display a reduced number of satellite stem cells. Furthermore, Wnt7a application enhances the regenera tive capacity of skeletal muscle dramatically [8]. Th is demonstrates conclusively that an expansion of the satellite stem cell population is benefi cial for skeletal muscle regeneration and suggests that manipulation of Wnt7a/ PCP signaling could be therapeutically relevant. Future studies will have to address the feasibility of recombinant Wnt7a, or mimetics of this factor, for the treatment of diseases that are accompanied by a disequilibrium in the satellite stem cell pool.

Conclusions
Knowledge about the factors that regulate satellite cell activity is not only crucial for their direct manipulation but will also foster the success of other approaches, such as stem cell therapy. Our understanding of the molecular mechanisms that control the specifi cation and diff erentiation of satellite cells into a postmitotic muscle fi ber has grown tremendously and a complexity far beyond expectations has emerged. In parallel, however, promising new starting points for the development of therapeutics have been discovered and it is only a matter of time until this translates into eff ective treatment options for degenerative muscular diseases.