Molecular regulation of myocyte fusion

Introduction

The vertebrate skeletal muscle system is vital for fundamental functions of life, from movement to breathing and metabolism. Skeletal muscle is formed during embryonic development, where muscle precursor cells undergo myogenesis, a stepwise process resulting in the formation of myotome segments of tightly packed myofibers. During myogenesis, precursor cells expand in number and differentiate into specialized cells known as myocytes, which possess the remarkable ability to fuse their cell membranes with one another and form new myofibers containing multiple nuclei within a shared cytoplasm. Hundreds to thousands of nuclei are required in mature myofibers to maintain fully functional skeletal musculature; thus, during postnatal development, myocytes continuously fuse into the new myofibers to achieve the necessary level of multinucleation [1, 2].

Myogenesis is not only confined to developmental stages. Mature adult myofibers can adapt to various cues and demands, such as muscle injury and exercise, through the addition of new nuclei. A population of quiescent, self-renewing stem cells called satellite cells, are activated by signals from muscle damage or exercise-induced stimuli and undergo proliferation, differentiation, and fusion. Activated satellite cells (myocytes) can not only fuse with each other and form new myofibers, but they can also fuse with existing myofibers to assist in muscle repair or growth [3, 4]. Thus, aside from myocyte fusion being necessary for skeletal muscle development, it also supports the repair and maintenance of muscle throughout life. Without an efficient cell fusion system, muscle is not able to form properly and adapt to physiological demands.

In this review, we will explore the latest advances in the biology of myocyte fusion, examining the process from membrane, cellular, and tissue morphological perspectives. We will focus on recent advances in both mammalian and zebrafish model systems. Extensive research in with these organisms has focused on unraveling the molecular regulators of the fusion process at the subcellular scale [5]. In concert with these subcellular processes, there is coordination at both the cell and tissue scale that is important for the timing and positioning of fusion events. Here, we take a multiscale approach, covering both molecular processes (e.g., action of the skeletal muscle fusion proteins) and cell and tissue processes that ensure robust myocyte fusion and organization of a functional myotome.

The term fusogen is used to describe a protein that directly drives the fusion of two distinct plasma membranes, in contrast to proteins that regulate steps of the process preceding membrane merger. Fusogens are essential for the development of multicellular organisms, supporting processes like gamete fusion during fertilization, cell-cell fusion during organ development, and tissue repair [6, 7]. Since the discovery of the skeletal muscle fusogens that are conserved across vertebrates, Myomaker (Mymk) [8] and Myomerger/Myomixer/Minion (Mymx) [9-11], there is an improved understanding of the molecular requirements for myocyte membrane fusion, which is highlighted here. Of note, in this review we use both Mymg and Mymx to refer to Myomerger/Myomixer/Minion. Mutations in each fusogen have been linked to congenital myopathies in humans, which opens new systems to learn about the biology of myocyte fusion and its implication in disease [12, 13]. Additionally, we discuss how the myocyte fusion reaction can be repurposed to pave the way for innovative therapeutic strategies to address skeletal muscle diseases. Finally, we highlight future research directions that are anticipated to deepen our understanding of myocyte fusion.

Overview of Vertebrate Myocyte Fusion

The current model to explain the subcellular mechanisms leading to myocyte fusion involves a series of coordinated steps that ensures the successful merger of individual cells into functional myofibers (Figure 1). This has been extensively discussed in previous reviews [14-17]; thus, we only provide a brief overview. Initially, myocytes recognize and adhere to other myocytes or to existing myofibers, which is thought to be necessary to ensure that fusion occurs between compatible cells of the same lineage. Several factors have been proposed to regulate cell recognition during myocyte fusion in zebrafish, but there a bona fide recognition molecule that is essential for vertebrate myocyte fusion has not been identified. This has led to the idea that the extracellular matrix (ECM) could provide the forces to bring cells in close enough proximity for fusogen engagement [18]. Following cell recognition, signaling pathways induce actin remodeling and cause F-actin polymerization, which drives the formation of finger-like membrane protrusions that extend from the surface of one cell towards another and assist in the formation of the fusion synapse between the membranes of two fusing cells [14, 15]. These protrusions are suggested to assist in the arrangement of the muscle fusogens (Mymk/Mymx), which likely remodel membrane lipids to lower the energy barrier for membrane fusion [19]. The interplay of fusogens and lipids is proposed to result in the mixing of outer leaflet membrane lipids (hemifusion) and the formation of a fusion pore - a specialized channel that connects the cytoplasm of the two fusing cells [20, 21]. Upon successful pore formation, the contents of each cell, such as organelles, proteins, and genetic material, get intermingled. In summary, the multifaceted process of myocyte fusion facilitates the merger of two independent membranes into a singular entity.

Figure 1.

A) General process myocyte fusion highlighting critical molecules. B) Schematic showing interaction between cytoskeletal proteins in myocytes and ECM molecules thought to contribute to myoblast fusion potentially by enhancing fusogen engagement.

Regulation of Myocyte Fusion by Skeletal Muscle Fusogens

Despite the proposed importance of fusogens for cell-cell fusion events, they remain a poorly understood group of proteins in vertebrates, but have been more extensively studied for their role in viral pathogenesis [22, 23]. In this review, we focus on the role of fusogens exclusively utilized for skeletal muscle formation.

The Skeletal Muscle Fusogens

Mymk and Mymx possess multiple characteristics that are consistent with them being the skeletal muscle fusogens. While multiple factors influence the cellular processes involved in myocyte fusion, the majority of these factors are present and active in non-muscle cells that do not fuse. By contrast, Mymk and Mymx are expressed specifically in myocytes during times of myogenesis, highlighting their centrality for the fusion process [8-11, 24, 25]. Genetic deletion of either fusogen results in a robust lack of skeletal muscle fusion in mice [8-11, 24, 25] and zebrafish [26-30]. In mice, disruption of either factor results in embryonic lethality, but in contrast, zebrafish that lack Mymk have no myocyte fusion in fast myofibers, but live to adulthood with mononucleated myofibers, reduced muscle mass, craniofacial abnormalities, and a failure to thrive during later stages of life. The discrepancy in lifespan between mice and zebrafish after fusion blockade could be explained by the magnitude of multinucleation required for muscle function in the two organisms. Complimentary to an absolute requirement of Mymk and Mymx for myocyte fusion, forced expression of these factors in cells that normally do not fuse is sufficient to induce membrane fusion [9-11], which is a barometer historically used to classify proteins as a fusogen. In addition, Mymk and Mymg have the ability to function within simpler systems that do not involve two cells. Enveloped viruses pseudotyped with Mymk and Mymg can fuse with muscle cells, opening up new approaches for skeletal muscle gene therapy [31]. Together, Myomaker and Mymerger/Myomixer emerge as central regulators of the myocyte fusion reaction.

Two models have been proposed to explain how Mymk and Mymx collaborate to induce membrane fusion. One of those models proposes that Mymk and Mymx act in a bipartite mechanism, where both fusogens regulate independent lipid remodeling steps of the fusion process [32]. In this situation, Mymk functions at or before hemifusion of outer leaflet lipids, which is supported by the necessity of Mymk for hemifusion in myoblasts. Moreover, the loss of Mymk does not impact the formation of actin protrusions, indicating that the protein functions downstream of actin remodeling [19]. After Mymk-mediated hemifusion, Mymx functions to catalyze the hemifusion to full pore formation transition completing the cell fusion reaction. An alternative model proposes that Mymx is capable of activating the fusogenic potential of Mymk through a direct interaction [9, 33, 34]. However, this second model is complicated by data that Mymk and Mymx can each function in systems where both proteins are not present [31-33]. If one combines these two models, it is possible that Mymk and Mymx have independent membrane remodeling functions, the effects of which synergize to indirectly activate the other fusogen.

Myomaker

As noted above, the expression of Mymk is restricted to specific myoblast cell lineages that are undergoing fusion. This specificity is transcriptionally controlled in part by MyoD and Myogenin [33, 35]. Recent work has shown Mymk expression in the myotome can also be influenced by Notch signaling in chick embryos [36]. In contrast to the transcriptional regulation of its myoblast lineage-specific expression, we still do not know the biochemical function of Mymk that enables it to drive hemifusion. The difficulty in identifying the precise function of Mymk lies in that it does not have conserved domains that can obviously drive fusion, such as those that are present in fusogens which work in virus-cell and intracellular membrane fusion events. Vesicle fusogens, like v-SNARES, contain long extracellular domains that bind t-SNARES on the target membrane and undergo a ‘folding’ conformational change that brings the fusing membranes in close proximity to induce membrane merger [37, 38]. In contrast, Mymk is highly embedded in the lipid bilayer, palmitoylated on its C-terminus (which is necessary for function) and has no domains that could easily reach an opposing membrane (Figure 2). It remains possible that Mymk could interact between opposing cells through short extracellular domains but this would likely require prior engagement of adhesion factors between those cells [39]. While Mymk does not have homology to fusogens that function in other fusion systems, it does have distant similarity to ceramidases - enzymes that regulate levels of ceramide, a bioactive sphingolipid [34, 40-42]. Together, the unique structure of Mymk implies that the regulation of myocyte membrane fusion is distinct from other membrane fusion systems.

Figure 2.

General topology of membrane fusogens.

Current efforts to reveal the function of Mymk are dedicated to determining protein domains that could drive the biochemical activity of the protein. Structural prediction from AlphaFold and previous work suggest Mymk adopts a seven-pass transmembrane domain topology, similar to other seven-pass transmembrane ceramidases like Acer3 and Adiponectin receptors [34, 43-46]. In these ceramidases with known structures, the active site is oriented internally towards the center of the protein and are marked by three histidine residues that coordinate a zinc ion as well as two hydrogen bond donors that stabilize a water molecule. All of these residues are proposed to be needed for the general acid-base hydrolysis of a ceramide molecule [43, 44, 47]. Interestingly, Mymk has the conserved histidine residues, but it lacks a critical hydrogen bond donor that is responsible for hydrolysis. Recently, these structural predictions were mostly confirmed experimentally using cryo-electron microscopy (cryo-EM) [39]. The structures of mouse and Ciona robusta Mymk were resolved in both nanodiscs and detergents, and this work demonstrated that the protein forms a homodimer and binds zinc through interactions with the histidine residues. Mymk homodimers were proposed to recognize homodimers on another fusing cell to facilitate cell-to-cell contact and membrane fusion and disrupting homodimerization destroyed the fusogenic capacity of Mymk in heterologous fusion assays. The current structures available for Mymk do not yet resolve the question as to whether Mymk uses its active site, which has similarity to active sites in ceramidases, to remodel lipids that drive hemifusion.

Myomixer/Myomerger

Mymx is another fusogen protein that plays a critical role in skeletal muscle development, and is proposed to act to drive fusion pore formation and final step of cell fusion. In the mouse, its expression is primarily restricted to differentiating myoblasts, aligning with its specialized role in the skeletal muscle lineage [9-11, 24]. Unlike Mymk, which is largely a bilateral fusogen, Mymx is required on only one of the two fusing membranes [9-11, 33, 34]. Mechanistically, Mymx resides in the cell membrane with a single transmembrane domain and two extracellular α-helical domains thought to destabilize cell membranes to induce pore formation. Specifically, helix one of Mymx is amphipathic and relatively unstructured in nature. However, when a negatively charged lipid, such as phosphatidylserine, is present in the membrane, helix one can insert itself into the head groups of the lipids (Figure 2) causing deeper insertion of helix two. These insertions generates curvature in the membrane, which then catalyzes the hemifusion stalk-to-pore formation transition and fusion completion [48, 49].

Other Myocyte Fusion Enabling Proteins

Skeletal muscle cell fusion is a tightly regulated process involving several key proteins beyond Mymk and Mymx. One such protein is dysferlin, a transmembrane component of the dystrophin-glycoprotein complex (DGC), which acts as a link between the primary components of the DGC and the actin cytoskeleton [50, 51]. Dysferlin has long been found to contribute to myocyte fusion during muscle repair in cultured C2C12 myoblasts and in studies involving mice [52-55]. Dysferlin deficiency can lead to impaired muscle regeneration and weakness [53], mutations and deficiencies of which have been attributed to myopathy and muscular dystrophies [56-58]. While considered an ancillary factor for myocyte fusion, the true extent of the contribution of Dysferlin to muscle development and repair remains unclear. Additionally, related Ferlin proteins, such as Myoferlin and Dysferlin orthologues, harbor similar membrane repair and fusogenic properties [54], the exact extent of which has yet to be explored in an in vivo context.

The RAC/DOCK1-associated proteins, Elmo1 and Elmo2, have been discovered to play an essential role in myogenesis. The loss of both Elmo1 and Elmo2 results in myoblast fusion defects, while Elmo2 expression in dysferlin null mice can rescue animals with a dystrophic phenotype [52]. While dissection of the pathways and proteins involved remains an open question, these results provide an exciting new avenue for manipulating muscle regeneration in therapeutic applications.

Another example of a pair of molecules essential to myocyte fusion in zebrafish are the cell surface heterophilic receptor pair, Jamb and Jamc. Both of these cell adhesion molecules (CAMs), containing immunoglobulin (Ig) repeats in their extracellular domains, are co-expressed in fast myocytes of the zebrafish, and mutation in either of the proteins results in mononucleated fast-twitch myofibers [59, 60]. Transplantation experiments in zebrafish jamb and jamc mutants have shown that the molecules must interact between neighboring cells for fusion to occur [59]. jamc, but not jamb, is ectopically expressed in the slow myocytes in prdm1a mutants (that have defective specification of the slow-twitch fate and adopt the fast-twitch differentiation program) and can cause inappropriate fusion of the mutant slow-twitch myocytes. This suggests that Jamc is a key molecule for driving cell fusion and overall patterning during myogenesis. Despite the unusual mononuclear phenotype in embryonic fish, jamb and jamc single and double mutant fish are indistinguishable from wild-type fish after reaching adulthood, suggesting they are not necessarily required beyond early myogenesis [59, 61]. jamb; mymk double mutant zebrafish displayed a mononucleated myofiber phenotype into adulthood in addition to significant intracellular adipocyte infiltration in the adult mutant fish [59]. Taken together, this suggests that mymk is essential for myoblast fusion in embryonic muscle development and during adulthood, whilst jamb and jamc function is dispensable during post-embryonic muscle development.

In addition to the above CAMs, other Ig domain-containing proteins have been discovered to play a role in zebrafish myocyte fusion. Like the jam genes, kirrel3l is specifically expressed in fast muscle precursors and disruption of its activity caused failure of fast myocyte fusion, resulting in a population of mononucleated fast-twitch myofibers within the somites [62]. The paralogous protein, Nephrin, the function of which is primarily associated with the maintenance of the renal filtration barrier [63], has also been implicated in fast myocyte fusion during zebrafish embryonic myogenesis [64]. However, unlike kirrel3l, the nephrin gene is not obviously expressed in the skeletal muscle lineage, making its requirement in the fusion process even more tenuous. It should be noted, however, that the roles of both Kirrel3l and Nephrin in fast myocyte fusion were established using transient knock-down with morpholinos: investigating their roles with stable mutants will be required to firmly establish their necessity for fusion and the degree of redundancy with each other as well as with other CAMs like the Jams. Further, it is uncertain whether these proteins act analogously in the mouse and humans.

In summary, we have described the two key fusogen proteins along with a range of accessory factors, such as adhesion molecules, that play crucial roles in myocyte fusion. Yet, understanding the interplay between these proteins requires knowledge of their dynamics. Such information will be crucial to unravel the intricacies of muscle development, regeneration, and potential therapeutic interventions in muscle-related disorders.

Cell and Tissue Morphogenesis

The fusion of myocytes is a highly regulated event, and it is tightly coordinated with changes in cell shape and tissue organization. During early muscle development, myocytes undergo a series of morphogenetic changes, such as elongation and alignment, to facilitate their recognition and adhesion to neighboring myocytes [30, 65, 66]. Furthermore, the fusion process is closely intertwined with the establishment of cell-cell contacts, driven by key molecules including cadherins and integrins, CAMs and cytoskeletal proteins. The interplay between cell and tissue morphogenesis and myocyte fusion underpins the formation of functional skeletal muscles with implications for both developmental biology and regenerative medicine.

Tissue-scale Morphogenesis

Cell size, shape, organization and remodeling during development represent pivotal facets of tissue formation [67] (Figure 3A). In the context of skeletal muscle, dynamic reorganization of tissues not only accommodates the spatial expansion of developing muscle but also shapes the mechanical microenvironment in which myocytes operate. The mechanical cues, direction and magnitude of forces generated by the extracellular matrix and neighboring cells are known to impact myoblast and myocyte behavior and cytoskeletal dynamics. These processes potentially have knock-on effects on the timing and efficiency of cell-cell contact prior to cell-cell fusion [66, 67].

Figure 3.

A) Fusion events are coordinated on a tissue scale. Cross-sectional view through the developing zebrafish myotome. Left: early somite stage, just after segmentation from the presomitic mesoderm. Slow fibers are specified. Middle: Subsequently, fusion events (magenta) begin amongst the fast fiber population, with a bias towards the medial myotome surface. Right: The slow fibers (cyan) migrate to the lateral surface, which appear to coordinate the fast myocyte fusion events. B) Corresponding images to A from a 24 - 32 hour zebrafish embryo, showing the formation of muscle fibers (from a dorsal-ventral view). Images courtesy of Mario Mendieta-Serrano. C) Pictorial graph demonstrating fusion events at a local scale occurring in an uncoordinated manner.

The zebrafish myotome has a distinctive chevron shape [68], which has been hypothesized to be important for efficient swimming [69]. Recent work has shown that the emergence of this chevron shape is likely in part due to active forces generated as the constituent myofibers elongate and fuse [70]. Disruption to fusion and muscle fate alters the chevron angle and size of the myotome. Though fusion plays a role in determining tissue shape, it remains an open question as to how the underlying physical forces drive tissue morphogenesis. Studies involving C2C12 cultured myoblasts have demonstrated that the mechanical properties of the surrounding developing muscle tissue, such as tissue tension, cell adhesion, cell density, stiffness, and even the direction and magnitude of the morphological forces involved all play a crucial role in regulating cell fusion during myogenesis [71-77]. Further understanding of these processes is essential not only for developmental biology but also for regenerative medicine and therapeutic interventions targeting muscle-related disorders, though extensive research delving into the role of tissue-scale forces and dynamics in the 4D environment constituting myogenesis remain few and far-between.

Cell-scale Morphogenesis

Changes in cell shape constitute a fundamental aspect of myogenesis, playing an important role in facilitating the intricate process of myocyte fusion during skeletal muscle development. Myoblasts undergo dynamic alterations in their morphology to enable effective fusion. Early in myogenesis, myoblasts typically exhibit a heterotypic rounded shape; then as myocytes, they begin to flatten and elongate prior to fusion [30, 78] (Figure 3B). While this process was thought to be a factor determining how and when cells might fuse, it has recently been shown in zebrafish that fast myocyte elongation and the positions of myocytes relative to their fusion partners, is highly heterogeneous [66] (Figure 3C). Individual cell-cell fusion events have been demonstrated to be largely stochastic. Yet, at the tissue scale, distinct waves of fast myocyte fusion occur along the medial to lateral regions of the developing myotome segments. This wave of fusion correlates with the migration of non-fusogenic slow-twitch muscle cells medio-laterally through the tissue. Thus, it appears that the slow muscle cells may provide an instructional cue to trigger fast myocyte fusion, although what exactly this cue is remains to be determined. One potential mechanism is that the slow fibers restrict the space for the fast myocytes, forcing them closer together, thereby increasing the probability of successful fusion. However, other mechanisms, including direct cell signaling cannot be currently discounted.

Cell elongation is accompanied by cytoskeletal rearrangements, including the reorganization of actin filaments, which promote cell motility and the formation of specialized structures such as actin foci and podosomes at cell-cell contact points [19]. These localized changes in cell shape and cytoskeletal architecture may enable myocytes to properly adhere to one another and bring their cell membranes close enough to initiate the fusion process [30, 65]. Furthermore, the elongated morphology of the myocytes facilitates their alignment and eventual extension to the myotome boundaries following membrane fusion, ultimately culminating in the formation of multinucleated myotubes anchored at the myotendinous junctions [79, 80]. Thus, the dynamic modulation of cell shape is a pivotal morphogenetic event that underpins the effective fusion of myocytes during myogenesis.

One intriguing feature that has emerged from these analyses of fusion behavior of zebrafish fast myocytes is that despite the key role of Mymk in the fusion process, it appears to play a permissive rather than an instructive role. For example, mymk expression levels only weakly correlate with the spatio-temporal occurrence of fusion events [57]. Cells with high levels of mymk did not immediately fuse nor did they fuse with their immediate neighbor, suggesting that the levels of its expression are not deterministic as to when and where fusion will occur. Rather, the domain of mymk expression within the developing myotome appear to circumscribe a rather broad region of fusion competence among the fast-twitch myocytes. While cellular levels of mymk mRNA are not correlated with fusion, it is possible that protein levels or localization on the plasma membrane, could be instructive.

Extracellular Matrix

The ECM and accompanying adhesion molecules play integral roles during myogenesis, helping govern structural and functional aspects of skeletal muscle development. The ECM, composed of a complex array of proteins including collagens, laminins, and proteoglycans, provides a supportive microenvironment for myocyte fusion [81-83]. It also functions as a reservoir of signaling molecules and growth factors that can regulate myogenic differentiation [84]. Moreover, the ECM serves as a physical scaffold, guiding myocyte alignment and migration.

Adhesion proteins, such as integrins and cadherins, mediate the interaction between myoblasts and myocytes with the ECM, as well as between the cells themselves. This cellular adhesion, as previously discussed, is a critical step for myocyte fusion. Integrin-mediated adhesion and signaling cascades are required for cell fusion and myogenesis [85, 86], facilitating the recruitment of signaling molecules controlling downstream gene expression whilst also playing its own role in binding myofibers to the basal lamina alongside the dystrophin-glycoprotein complexes [78, 87, 88].

Cadherins are, similarly to integrins, transmembrane glycoproteins which are expressed in skeletal muscle among other tissues. Cadherins mediate Ca2⁺-dependent cell-cell interactions and bind indirectly to the actin cytoskeleton via association with the catenin family of adhesion molecules [89]. M-cadherin is the most abundant cadherin type to be expressed during myogenesis. Knock-down studies in cultured rat L6 myocytes concluded that a reduction in m-cadherin expression resulted in a reduction in myoblast adhesion and differentiation within developing muscle despite an apparently normal adult muscle phenotype [90]. M-cadherin also plays a role in mediating β-integrin/Wnt signaling [91], with a recent study suggesting m-cadherin as a Pax3 target gene, implicating it heavily in the process of cell specification in the myotome [92].

Cytoskeletal Proteins

The cytoskeletal protein actin assumes a paramount role during myogenesis. It has been long understood that actin is essential in maintaining cell shape and driving cell migration. At the onset of myogenesis, myoblasts exhibit dynamic changes in their actin cytoskeleton, transitioning from a monomeric globular form (G-actin) to a highly organized, polarized network of F-actin filaments [88]. These F-actin structures are pivotal for several events in myogenesis, including myocyte fusion via formation of filopodia-like structures protruding from the membrane of fusion-competent myocytes to trigger pore formation and sarcomere assembly [19]. During fusion, F-actin-rich structures, such as actin foci and podosomes, facilitate cell migration and cell-cell adhesion [93], ensuring the alignment and fusion of myocytes into multinucleated myotubes. Furthermore, the sarcomeric units within developing myofibers rely extensively on the actin filaments to form the contractile apparatus, allowing for the generation of force in mature myofibers [94]. In essence, the multifaceted involvement of actin in myogenesis underscores an indispensable role in sculpting functional skeletal muscle tissue.

Like actin, microtubules are also intrinsically linked within skeletal muscle development [95]. As a major cytoskeletal protein, microtubules play a significant role during myogenesis. It has long been established that disruption to microtubular architecture and the resultant mispositioning of nuclei underlies the severe phenotype in dystrophic myofibers [96-98]. During early myogenesis, microtubules facilitate intracellular transport of components such as mRNAs, proteins and organelles, facilitating their delivery to specific cellular locations [99, 100]. This transport is essential for myocyte alignment, fusion and formation of multinucleated myotubes [101]. Moreover, microtubules are involved in the establishment and maintenance of cell polarity, overall cell structure and remodeling and asymmetric divisions of myoblasts; events critical for myofiber formation. Additionally, microtubules serve as tracks for the directional movement of motor proteins, like dynein and kinesin, which govern myoblast migration and intracellular trafficking. This is important for the translation of extracellular signals to changes in gene expression within the nucleus [72, 76, 102].

Recent studies have implicated microtubules in the ability of cells to sense changes in the local environment, such as alterations in substrate/tissue rigidity and mechanical forces exerted upon the cells [72]. Microtubules can alter their physical properties and stability (via chemical modification such as acetylation and tyrosination) [75], which can impact the behavior and physical properties of cells. This reactivity is particularly important in tissues experiencing significant changes in tissue tension or mechanical pressures, such as for myocytes within muscle tissue. Given this consideration, microtubules are likely key regulators of myogenesis, ensuring the spatiotemporal coordination of various cellular processes necessary for the development of functional skeletal muscle. However, the specific mechanisms by which they act – especially in vivo – remain poorly characterized. This is in part due to the challenges of imaging their action (which are on subcellular spatial scales and act on second time scales) at the appropriate tissue spatial (50-100mm) and temporal (minutes to hours) scales. Recent developments in live-imaging, such as light-sheet microscopy [103], and controlled perturbations using optogenetics [104] provide new opportunities to study the processes of cytoskeletal proteins in both an in vitro and in vivo settings.

Exploring Myocyte Fusion in Disease: Unique Myopathies and Emerging Therapeutic Prospects

Beyond the role of myocyte fusion in muscle development and regeneration, recent work has highlighted that fusion can, either directly or indirectly, impact muscle diseases. There is an emerging concept that hypomorphic mutations in MYMK and MYMX reduce levels of myocyte fusion and cause a myopathy termed Carey-Fineman-Ziter Syndrome (CFZS) [12, 13, 105-107]. In addition, in Duchenne Muscular Dystrophy (DMD), which is the most common genetic muscle disease, muscle pathology in a mouse model of DMD is exacerbated due to the residual expression of Mymk in myofibers post-fusion [108]. Outside of myocyte fusion being a determinant of muscle pathology in CFZS and DMD, a recent study demonstrated that the myocyte fusion reaction can be leveraged to regulate fusion of viral membranes with muscle [31]. These viruses can be used as vehicles to deliver therapeutic genes specifically to skeletal muscle. Together, the sections below highlight that dysregulation of myocyte fusion impacts muscle pathologies, and that understanding fusion biology has practical consequences on the treatment of muscle diseases.

Fusogen Mutations Cause a Unique Myopathy

CFZS patients present a diverse range of clinical manifestations, from facial weakness and developmental delays to generalized muscle abnormalities[12, 109] (Figure 4). The cause of the myopathy was difficult to identify because, at birth, the facial abnormalities, weaknesses, and feeding issues resembled features seen in congenital craniofacial disorders with neurological implications. However, the presence of muscle hypotonia and generalized hypoplasia suggested that the disorder was influenced directly by the muscle. Even more perplexing though were the muscle biopsies from these patients, because both biopsies (14-month-old male, 19-year-old female) showed some myofiber size variances, but there were no detectable changes in the ratio of myofiber types or activation of the myogenic program, which are commonly seen in myopathies [109]. Recent studies have now identified that MYMK and MYMX mutations are causative of the phenotypes observed in CFZS, suggesting the syndrome is a primary muscle disorder caused by defects in the myocyte fusion machinery [12, 13]. Further explorations have aimed to determine the effect of these mutations on myocyte fusion in humans and their association with phenotypic variation.

Figure 4.

Overview of phenotypes associated with Carey-Fineman-Ziter syndromes.

Impact of CFZS Mutations in Humans

Upon discovering mutated fusogens in CFZS patients, questions emerged regarding the mechanistic implications of these mutations on myocyte fusion and their role in causing the observed phenotypes. Genetic analyses revealed each CFZS patient has a combination of hypomorphic and null alleles, or the pairing of two hypomorphic alleles (compound heterozygous). This genetic configuration points to a critical threshold that each fusogen needs to be maintained at for normal myocyte fusion. If the critical threshold of fusogen expression is not reached during times of myocyte fusion, it is expected to decrease the magnitude of myocyte fusion events, leading to reduced myofiber numbers and myonuclear accretion. In response to a theoretical reduction in fusion events, myofibers may attempt to increase muscle mass by undergoing hypertrophy, a myopathic feature that is often observed in patient biopsies [12, 106]. This hypertrophy does not correlate with enhanced muscle function though, as muscle function is diminished throughout the lives of CFZS patients. If CFZS patients have a fully functional fusion system that supports muscle regeneration and maintenance, then muscle health may gradually improve through regenerative fusion. However, adult CFZS patients have a diminished ability to exercise and thrive at later stages in life compared to their peers, which suggests that the extent of fusion and myonuclear accretion in adults is reduced by mutations in the fusogens.

The simplest explanation for CFZS-associated mutations causing pathology in humans is that functional levels of fusogens are reduced below a critical threshold resulting in fewer myonuclei within myofibers. There is, however, minimal evidence regarding the magnitude of fusion events in CFZS patients and whether a threshold of fusogen expression is truly needed for normal development in humans. One study has tested the efficiency of myocyte fusion in primary myoblasts isolated from one CFZS patient [12]. These data suggest there is a reduction in the extent of myocyte fusion events, but in vitro systems may not accurately reflect the complex environment during myocyte fusion in vivo.

In Vivo Models to Study CFZS

Given the limited number of individuals diagnosed with CFZS and the scarcity of available biopsies, the need to develop novel models for investigating the effects of mutations in myocyte fusion machinery became apparent. Efforts to address this challenge have involved the utilization of zebrafish and mouse models to probe the impact of CFZS and its associated mutants in vivo. In the context of zebrafish, the seminal study that initially identified mutations in MYMK, introduced CFZS variants by injecting RNA encoding the mutants into mymk knock-out (KO) zebrafish [12]. Unlike Mymk KO mice which do not survive, mymk KO zebrafish at developmental and juvenile stages exhibit viability, highlighting the residual function of slow-twitch mononucleated fibers which evade fusion and/or the ability of the non-fusing fast-twitch myonuclei to produce contractile fibers. Nevertheless, adult mymk KO zebrafish display reduced body size, craniofacial deformities, and a variable degree of muscle fatty infiltration, reminiscent of features often observed in CFZS patients [26, 30]. Injection of RNA encoding hypomorphic MYMK mutants, but not the null mutants, was able to partially correct the myocyte fusion defects and supported the findings that the hypomorphic mutations retain residual MYMK function. Although this investigation contributed to elucidating the functional nature of MYMK mutations as hypomorphic or null alleles, it did not directly assess the impact of the mutations on Mymk-mediated myocyte fusion in the context of mammalian muscle.

To directly assess how CFZS-associated mutations impact myocyte fusion in mammalian muscle, attention has shifted towards exploring the valuable insights that can be gained from a mouse model that contains a MYMX 136C>T mutation [13]. This mutation yields a truncated Mymx protein, characterized by a premature stop codon at Arginine 46 (Mymx R46*). Homozygosity of this mutation in mice causes mortality at postnatal day 14, severe hypotonia, heterogenous myofiber sizes, and a lack of myonuclei within myofibers. While deficiencies in embryonic myogenesis are evident likely because of compromised myocyte fusion, the regenerative potential of the adult muscular system and disease progression throughout life could not be evaluated because of the early lethality of the mice.

The reasons for the contrasting survival durations between CFZS-afflicted humans and the mouse model is unclear; yet, one can speculate that compensatory factors replace Mymx function in humans, which are not active in mice. Alternatively, disparities between the skeletal muscle systems of mice and humans could be present, such that myogenic progenitor cells in mice and humans may have different properties and functions, but determining the differences has remained difficult due to the lack of studies characterizing and identifying human myogenic cell populations [4, 110]. Furthermore, it is crucial to acknowledge the absence of mouse models that faithfully mimic pathogenic MYMK mutations in humans, which underpins the need for further exploration in this direction.

Unveiling Fusogenic Mechanisms Through Disease-Associated Mutations

Fusogen mutations found in CFZS patients have the potential to illustrate the molecular and biochemical events essential for proper myocyte membrane fusion. Each mutation has an independent impact on the protein, where some missense mutations result in a hypomorphic protein product, and some lead to a total loss of expression or activity. The MYMK mutants that are undetectable when overexpressed in cells (Met1 mutation of initiating methionine, Gly100Ser, Cys185Arg, Trp79Arg, etc.) are not as informative for biochemical investigations as mutations that do not disrupt expression. The lack of detectable protein suggests that these mutations affect protein expression, folding, or the stability of the protein in a membrane. By contrast, the proposed hypomorphic Mymk mutant proteins, Pro91Thr and Ile154Thr, have some ability to drive fusion in a heterologous fibroblast fusion assay [12]. Since the biochemical activity of Mymk is not known, the precise effect of these mutations on protein function cannot currently be determined. Further explorations could reveal how hypomorphic and null mutations impact the structure and dimerization status of Mymk.

In contrast to the Mymk mutant proteins that lack expression, the only reported mutation in Mymx is R46*, which impacted protein activity, not necessarily expression or localization [13]. Mymx R46* may show a similar expression pattern compared to the wild-type protien, but it failed to increase fusion in a heterologous fibroblast fusion assay. This suggests the Mymx R46* protein, which loses the second extracellular helix, has a reduced ability to induce pore formation in fusing cells. This is consistent with in vitro studies that demonstrate the first helix of Mymx interacts with negatively charged lipids and destabilizes lipid bilayers independent of helix two. However, both helices are necessary for optimal function [48, 49]. An intriguing hypothesis to explain the synergism of helix one and two is that helix two, with its leucine-rich zipper, may control the clustering of helix one, which then amplifies its function. These data, combined with findings that Mymx R46* mice have some, but minimal, multinucleated myofibers, indicate that Mymx R46* is a hypomorph, capable of inducing low levels of pore formation. In summary, investigating CFZS-associated mutations in the fusogens has shed light on how changes to their cellular processing, structure, and function impact the molecular events occurring during myocyte fusion.

Indirect Implications of Myocyte Fusion in Disease

Myocyte fusion, aside from its direct role in myopathies such as CFZS, also influences muscle diseases through indirect mechanisms, notably in more prevalent conditions like DMD. This X-linked recessive condition results from mutations affecting Dystrophin, a crucial protein that maintains muscle sarcolemma stability and shields it from damage caused by muscle contractions [111, 112]. In the absence of functional Dystrophin, muscle degeneration occurs, prompting satellite cells to undergo regenerative myogenesis and fuse with compromised myofibers. While fusion is tightly regulated during regeneration in healthy animals, the absence of Dystrophin leads to continuous, damage-triggered fusion. These fusion events, however, have unintended consequences that may adversely impact fiber health.

It remains unclear precisely how myocyte fusion, especially in a dystrophic setting, exacerbates pathology. One hypothesis posits that the fusion process, marked by changes in membrane structure and characteristics, may promote membrane instability [20, 113, 114]. Additionally, myocyte membranes that fuse with myofibers could introduce proteins and lipids that compromise myofiber membrane integrity. In healthy myofibers, the effects of these myocyte-derived components may be regulated through dispersion into areas of lower concentration or through endo/exocytosis events [115, 116]. In dystrophic fibers, excessive fusion may strain the mechanisms controlling these processes, potentially resulting in localized membrane destabilization.

For instance, cell fusion contributes to elevated levels of Mymk in mature mouse myofibers which can exacerbate pathology [108]. This might be due to the fact that Mymk is a membrane-active protein, potentially causing further damage to already unstable dystrophic myofibers. Intriguingly, forced expression of both Mymk and Mymx in adult myofibers adversely affects muscle health and membrane stability [117]. Acute Mymk expression can lead to substantial myofiber membrane damage, while prolonged exposure to a lower dose correlates with elevated central nucleation and muscle weight loss. Mymx, on its own, exhibits similar membrane-damaging features, though potentially at a lower toxicity compared to Mymk. Notably, atomic force microscopy revealed both fusogens reduce myofiber membrane stiffness, and the loss of Mymk in a dystrophic context can restore membrane stiffness to levels seen in healthy myofibers. This impact on membrane stiffness by Mymk suggests its residual capacity to influence membrane fluidity, a feature that frequently accompanies membrane fusion. Given the harmful role of Mymk and Mymx in adult myofibers, it is plausible to hypothesize that myocytes fusing with dystrophic myofibers contribute to the expression of Mymk and Mymx on the myofiber membrane. These potential contributions of myocyte fusion to membrane instability complicate our understanding of disease progression in conditions like DMD.

Manipulating the activity of fusogens in DMD to reduce the amount of myocyte fusion in dystrophic mice could be a double edge sword. Reducing the amount of fusion in a dystrophic setting may improve muscle health and stability by increasing myofiber membrane stability. However, the long-term consequences could include limitations in muscle regeneration, leading to muscle wasting over time. This conundrum underscores the importance of comprehending both the short- and long-term impacts of Mymk and Mymx within a dystrophic setting. Therefore, unraveling how myocyte fusion can be manipulated during disease progression becomes a crucial research pursuit, and suggests the need to develop inhibitors of the fusogens that can restrain their activities without completely eradicating their functions.

Harnessing the Myoblast Fusion Reaction for Gene Therapy

In recent years, significant progress has been made in the field of gene therapy that offers a promising avenue for correcting the genetic underpinnings of DMD. These therapies involve the delivery of truncated versions of Dystrophin (μ-dystrophin) for gene replacement or CRISPR/Cas9 components to repair the endogenous mutations. Adeno-associated virus (AAV) serotypes have been extensively utilized as carriers for genetic payloads, where AAV9-μ-dystrophin has recently received FDA approval, and novel serotypes like MyoAAV and AAVMyo exhibit more specific tropism for skeletal muscle [118-124]. While the AAV delivery approaches are promising, AAVs can provoke immune responses and the factors that influence the immune response are unpredictable and vary between patients [120, 125]. Recent research has tapped into the membrane fusion reaction mediated by the fusogens, Mymk and Mymg, offering an intriguing strategy to enhance the precision, efficacy, and safety of gene delivery to muscle tissues.

Pseudotyping was used to engineer lentiviral particles with Mymk and Mymg proteins on the viral envelope (Mymk+Mymg−LVs) [31]. When Mymk+Mymg−LVs were administered systemically in mice, they targeted and fused exclusively with muscle cells. A stimulus that activates satellite cells and induces the expression of Mymk, and consequently cell fusion, will permit the muscle cells to be transduced with the Mymk+Mymg−LVs. These specialized vehicles serve as a precise means to administer therapeutic payloads, including μ-dystrophin, to dystrophic myofibers and muscle progenitors, facilitating membrane stabilization and promoting muscle function. These findings illustrate that the muscle-specific fusogens, beyond their crucial roles in myogenesis, exhibit a remarkable ability to catalyze membrane fusion reactions across diverse membrane systems, conferring a novel dimension to the enhancement of viral-mediated gene transfer.

The viral-cell fusion system described above has implications beyond therapeutic gene therapy; it may uncover mechanisms underlying myocyte-to-myocyte fusion. It is worth noting that Mymk, but not Mymg, is needed on both the viral and recipient cell membrane for successful transduction. The absence of Mymk in dystrophic myofibers results in a lack of transduction by Mymk+Mymg−LVs, aligning with the pivotal role of the protein on both fusing membranes. In the pursuit of optimizing viral vector-mediated gene therapy, it is crucial to explore both viral delivery and production processes. Understanding how Mymk+Mymg−LVs target and interact with recipient cells, their mechanisms for cellular entry and exit, and their interactions with the immune system is of paramount importance.

Future Directions in Myocyte Fusion

Despite recent findings that contribute to the understanding of myocyte fusion, there remain many areas that require clarification to fully understand the process. Most of the knowledge behind membrane fusion intermediates in other systems originates from in vitro work using vesicle fusion systems, likely because it is easier to visualize and characterize fusion intermediates. It is more difficult to study cell-to-cell fusion systems because it requires the concerted functions of many factors that influence the fusion process through distinct mechanisms. It is also more difficult to visualize cell-to-cell fusion intermediates due to imaging limitations and the dynamic nature of cells. The development of live markers that do not perturb function will be a major advance. With these reagents, the spatiotemporal localization of fusogens can be quantitatively assessed and compared with when and where fusion occurs. The relationship between cell shape and fusion capability remains unclear, especially in an in vivo context. Boundaries and spatial constraints appear to play a role in guiding where and when fusion occurs, but better in vivo analyses are required to dissect the specific mechanisms. The lipid remodeling step of myocyte fusion represents an understudied part of the process, because there are challenges associated with detecting lipid remodeling events that occur specifically on the plasma membrane during cell-cell fusion. In terms of the skeletal muscle fusogens Mymk and Mymx, there remain many open questions regarding their mechanisms of action, membrane trafficking, localization, and interactions between themselves and other proteins. It also remains plausible that additional components of the fusion machinery remain to be discovered, and their identification will allow us to build a more complete picture of the mechanisms of myocyte fusion. Relatedly, as noted above, the ECM and adhesion molecules play important roles in fusion, yet the mechanisms by which these function remain largely unknown. Given the recent molecular discoveries and the advances in imaging technology, we expect further mechanistic insights in the myocyte fusion field in the upcoming years.