“Muscling” Throughout Life: Integrating Studies of Muscle Development, Homeostasis, and Disease in Zebrafish

Abstract

The proper development and function of skeletal muscle is vital for health throughout the lifespan. Skeletal muscle function enables posture, breathing, and locomotion; and also impacts systemic processes—such as metabolism, thermoregulation, and immunity. Diseases of skeletal muscle (myopathies, muscular dystrophies) and even some neurological, age-related, and metabolic diseases compromise muscle function and negatively affect health span and quality of life. There have been numerous, recent examples of studies on skeletal muscle development with exciting, therapeutic implications for muscle diseases. The zebrafish (Danio rerio) is a vertebrate model organism well accepted for developmental biology and biomedical research and thus an ideal system in which to elucidate the translational implications of mechanisms regulating skeletal muscle development and homeostasis. Muscle fiber types (slow- vs fast-twitch) are spatially segregated in zebrafish allowing for the opportunity to identify distinct mechanisms regulating fiber type specification during development as well as observe fiber type-specific effects in zebrafish models of muscle diseases. Accessible genetics coupled with transparent zebrafish embryos has enabled in vivo cell biology experiments allowing for the visualization and understanding of never-before-seen cellular processes occurring in muscle development, regeneration, and disease. In addition, high-throughput drug screening provides a platform for efficient drug discovery. The purpose of this chapter is to review the studies in zebrafish that significantly contributed to our understanding of cellular and molecular mechanisms regulating skeletal muscle development, homeostasis, or disease in vertebrates, with a particular emphasis on the basic developmental biology studies with promising therapeutic implications.

1. INTRODUCTION

This review focuses on areas where understanding mechanisms underlying skeletal muscle development has led to insights into muscle diseases (Table 2). Muscle diseases, such as muscular dystrophies and myopathies, are stereotypically characterized by muscle wasting, weakness, and impaired locomotion. One of the major challenges regarding therapy development is the phenotypic variability within and between dystrophies and myopathies. For example, within a subset of muscular dystrophies called dystroglycanopathies, phenotypes span the clinical spectrum from the most severe (congenital muscular dystrophy (CMD)–dystroglycanopathy with eye and brain abnormalities) to the mildest (adult onset limb-girdle muscular dystrophy–dystroglycanopathy). We predict that use of zebrafish models for these diseases will not only provide new insight into the connection between muscle development and disease, but also should identify the cellular and molecular etiology of phenotypic variability, which then could help in therapy design.

Table 2.

Zebrafish Models of Muscle Diseases not in Gibbs, Horstick, and Dowling (2013) Review

Myopathy	Gene Product	Method of Gene Manipulation	References
Centronuclear myopathy	Dynamin2	Translation and splice blocking dnm2a MOs; overexpression of human DNM2 R522H or S619L mutants	Gibbs, Davidson, Telfer, Feldman, and Dowling (2014) and Bragato et al. (2016)
	Bin1	Translation and splice blocking bin1 MOs; overexpression of human BIN1 K35N or K575* mutants	Smith, Gupta, and Beggs (2014)
X-linked myotubular myopathy (subtype of centronuclear myopathy)	Mtm1, Mtmr12	Translation (mtm1, mtmr12) and splice (mtmr12) blocking MOs; overexpression of human MTM1, MTMR12	Gupta, Hnia, et al. (2013)
Limb-girdle muscular dystrophy (LGMD)	Popdc1	Splice blocking popcd1 MOs; overexpression of human POPDC1 S201F; TALEN-mediated mutagenesis popdc1 s191f	Schindler et al. (2016)
	Hnrpdl	Translation blocking hnrpdl MOs	Vieira et al. (2014)
LGMD, movement myopathy, congenital muscular dystrophy (CMD) with fatty liver and infant onset cataract	Trappc11	Mutant line. Liver and eye phenotypes previously characterized in zebrafish. Skeletal muscle phenotype not characterized in zebrafish	Liang et al. (2015)
Mitochondrial myopathy	Slc25a42	Translation and splice blocking slc25a42 MOs; overexpression of human SLC25A42 N291D mutant	Shamseldin, Smith, et al. (2016)
Distal myopathy	Adssl1	Splice blocking adssl1 MOs; overexpression of human ADSSL1 D304N and I350fs mutants	Park et al. (2016)
Ullrich congenital muscular dystrophy (UCMD), Bethlem myopathy	Col6a1	TALEN-mediated mutagenesis of splice site resulting in exon skipping	Radev et al. (2015)
	Col6a1	Splice blocking col6a1 MOs	Zulian et al. (2014)
Myofibrillar myopathy	FilaminC	flnc MOs and mutant line; overexpression of human FLNC W2710X mutant	Ruparelia, Oorschot, Ramm, and Bryson-Richardson (2016) and Ruparelia, Zhao, Currie, and Bryson-Richardson (2012)
	Abcrystallina, Abcrystallinb, Desmina, Desminb, Bag3, Dnajb6a, Dnajb6b, Fhl1a, Fhl1b, Filaminca, Filamincb, Myotilin, Plectina, Plectinb, Vcp, Zasp	Translation or splice blocking MOs	Bührdel et al. (2015)
	Desmina	Mutant line from ENU mutagenesis screen; translation and splice blocking desma MOs	Ramspacher et al. (2015)
	Bag3	Translation and splice blocking MOs; overexpression of human BAG3 P209L mutant	Ruparelia, Oorschot, Vaz, Ramm, and Bryson-Richardson (2014)
Nemaline myopathy	Acta1, Neb	Translation (actc1b) and splice (actc1b, neb) blocking MOs; overexpression of human ACTA1 D286G mutant	Sztal et al. (2015)
	Lmod3	Splice blocking lmod3 MOs	Yuen et al. (2014)
	Tmod4	Mutant line from ENU mutagenesis screen; translation blocking tmod4 MOs	Berger et al. (2014)
	Klhl40, Klhl41	Translation and splice blocking MOs for klhl41a and klhl41b; Translation blocking MOs for klhl40a and klhl40b	Gupta, Ravenscroft, et al. (2013), and Ravenscroft et al. (2013)
Myotonia congenita	Clcn1	Overexpression of human CLCN1 I553F/H555N and CLCN1 L844F mutants	Cheng, Tian, Burgunder, Hunziker, and Eng (2014)
Duchenne muscular dystrophy (DMD)	DMD	Mutant lines (sapje, sapje-like, dmdpc2)	Waugh et al. (2014), Li, Andersson-Lendahl, Sejersen, and Arner (2014), Kawahara et al. (2014), Kawahara and Kunkel (2013), Giacomotto et al. (2013), Johnson, Farr, and Maves (2013), Winder, Lipscomb, Angela Parkin, and Juusola (2011), and Bassett et al. (2003)
Stormorken syndrome	Stim1, Orai1	Overexpression of human STIM1 R304W, STIM1 D76A, and ORAI1 P245L mutants. Thrombocytopenia, but not tubular myopathy phenotype investigated in zebrafish	Nesin et al. (2014)
GNE myopathy	Gne	Translation and splice blocking gne MOs	Daya et al. (2014)
Myotonic dystrophy	Dmpk 3′UTR	Overexpression of RNA containing 91 uninterrupted CUG repeats	Todd et al. (2014)
Secondary dystroglycanopathy	Gmppb	Splice blocking gmppb MOs	Carss et al. (2013)
	B3galnt2	Translation and splice blocking b3galnt2 MOs	Stevens et al. (2013)
	Pomk	Translation and splice blocking pomk MOs; overexpression of human POMK Q109* mutant mRNA	Di Costanzo et al. (2014)
Native American myopathy	Stac3	Mutant line; overexpression of zebrafish stac3 mutant mRNA corresponding to human STAC3 W284S mutation	Horstick et al. (2013)
Arthrogryposis	Zc4h2	Translation and splice blocking zc4h2 MOs; overexpression of mouse Zc4h2 mutant mRNAs corresponding to human ZC4H2 mutants (Val63Leu, Arg198Gln, Pro201Ser, Arg213Trp)	Hirata et al. (2013)
Arthrogryposis, lethal congenital contracture syndrome (LCCS)	Mybpc1	Translation and splice blocking mybpc1 MOs, overexpression of zebrafish Mybpc1 W220R, Y845H mutants (correspond to pathogenic human mutations)	Ha et al. (2013)
LCCS	Zbtb42	Translation and splice blocking zbtb42 MOs; overexpression of human ZBTB42 Arg397His mutant	Patel et al. (2014)
Nutritional myopathy		Vitamins E and C deficiency	Lebold et al. (2013)
Ceroid lipofuscinosis, neuronal, 10	Cathepsin D	Translation blocking ctsd MOs	Follo, Ozzano, Montalenti, Santoro, and Isidoro (2013)
Unnamed neuromuscular disease	Golga2	Translation and splice blocking golga2 MOs	Shamseldin, Bennett, Alfadhel, Gupta, and Alkuraya (2016)
No known human disease	Apobec2	Translation blocking apobec2 MOs	Etard, Roostalu, and Strähle (2010)
	Atrogin1	Translation and splice blocking atrogin1 MOs	Bühler et al. (2016)
	Col22a1	Translation blocking col22a1 MOs, overexpression of human wild-type COLXXII	Charvet et al. (2013)
	Mybpc2	MOs	Li, Andersson-Lendahl, Sejersen, and Arner (2016)

^*Stop codon.

2. FROM MESODERM TO MUSCULATURE: OVERVIEW OF SKELETAL MUSCLE DEVELOPMENT

The first step in vertebrate muscle development is segmentation of the paraxial mesoderm into somites. Somites give rise to multiple tissues: muscle (fast-twitch, slow-twitch, and muscle progenitor cells that contribute to muscle growth), elements of the skeleton and tendons, and hematopoietic stem cells (Murayama et al., 2015; Nguyen et al., 2014; Qiu et al., 2016). After somites form, the anterior border cells (ABCs) of each somite give rise to the zebrafish equivalent of the amniote dermomyotome (Hollway et al., 2007; Stellabotte, Dobbs-McAuliffe, Fernández, Feng, & Devoto, 2007) (Fig. 1, red cells). The amniote dermomyotome contains multipotent progenitor cells that proliferate and express Pax3 and Pax7 (Buckingham et al., 2003). In zebrafish, ABCs migrate laterally after somitogenesis and form an extremely thin layer of cells, the external cell layer (ECL). The ECL expresses Pax3 and Pax7 and contributes to muscle growth (Hollway et al., 2007; Stellabotte et al., 2007); thus, the zebrafish ECL is considered the functional equivalent of the amniote dermomyotome.

Fig. 1.

Zebrafish skeletal muscle structure and development. (A) Medial focal plane, side oriented, anterior left, dorsal top, 22 somite stage zebrafish embryo showing stages of skeletal muscle development along the anterior–posterior axis. The anterior most segments contain long muscle cells and are called myotomes (blue cells, slow-twitch muscle pioneers; dark green cells, fast-twitch fibers (multinucleate); black chevrons, myotendinous junctions (MTJs)). Muscle segments toward the posterior of the embryo contain cells in different phases of fast-twitch muscle morphogenesis (called the transition zone). From left to right within the transition zone, dark-medium green cells are elongated but irregularly shaped myotubes, medium green cells are elongating via protrusion extension and filling, and light green cells are short and extending protrusions in all directions. All muscle cells within the transition zone are mononucleate. At this developmental stage, muscle segments in the very posterior of the embryo are not yet undergoing fast muscle morphogenesis and are called somites. (B) Rotated, transverse, 3D views of one myotome, one muscle segment in the transition zone, and one somite (from left to right, respectively). Brown cylinders represent the neural tube on top and the notochord below. Myotomes contain slow-twitch fibers that have completed their migration (blue) and fast-twitch fibers that have completed their morphogenesis (dark green). Also shown are tendon progenitors (gray cells), sclerotome (purple cells), and satellite cells in the external cell layer (red). Muscle segments in the transition zone contain slow-twitch fibers part way through their migration (blue), fast-twitch fibers medial to the migrating slow fibers that have undergone morphogenesis (dark green), presumptive fast cells undergoing morphogenesis (lighter shades of green), and satellite cells of the external cell layer (red). Somites consist of epithelial cells surrounding an inner mesenchymal group of cells (both shown in green), medially located slow muscle cells that have not yet migrated (blue), and anterior border cells (red) that will become satellite cells of the external cell layer. M, medial; A, anterior; V, ventral.

Next, muscle cells elongate and attach to myotendinous junctions (MTJs) (Fig. 1). After elongation, the terminal ends of muscle fibers attach to somite boundaries, which become myotome boundaries and give rise to MTJs (Long, Adcock, & Root, 2002) (Fig. 1, black chevrons). Then, fast-twitch muscle cells fuse into multinucleate myofibers and myofibrillogenesis generates sarcomeres, the contractile units of muscle.

Slow-twitch muscle progenitors are initially located medially but, shortly after somite formation, they migrate laterally and become the most superficial layer of muscle (Devoto, Melançon, Eisen, & Westerfield, 1996) (Fig. 1B, blue cells). Slow-twitch fiber migration is necessary and sufficient to trigger elongation of the fast-twitch fibers (Henry & Amacher, 2004) (Fig. 1, dark green cells).

2.1. The Tortoise or the Hare? Fiber Type Specification During Development

Muscle fibers are frequently classified as slow-twitch/type I or fast-twitch/type II, but intermediate fiber types provide for an almost continuous spectrum of physiological possibilities (Schiaffino & Reggiani, 2011). Fiber types are defined by the particular myosin heavy chain isoforms that they express. Many other factors also contribute to a fiber’s phenotype—such as contraction rate, response to neural stimulation, mode of metabolism, and ATP usage. Understanding development and regeneration of different fiber types is important because some genetic/acquired muscle diseases have more deleterious effects on particular muscle fiber types (Ciciliot, Rossi, Dyar, Blaauw, & Schiaffino, 2013; Schiaffino & Reggiani, 2011). Thus, it is possible that some muscle and metabolic diseases could be treated by shifting fiber type proportions or by supporting the development of a specific fiber type (Ljubicic, Burt, & Jasmin, 2014). Toward this end, we need to understand the developmental processes that generate muscle fiber types in embryos.

The molecular mechanisms of zebrafish muscle fiber type specification have been reviewed recently (Jackson & Ingham, 2013; Talbot & Maves, 2016). Four genetically, morphologically, and/or physiologically distinct fiber types have been identified in zebrafish embryos: superficial slow-twitch fibers, medial slow-twitch muscle pioneers, medial fast fibers, and fast-twitch fibers (Wolff, Roy, & Ingham, 2003) (Table 1). High levels of hedgehog signaling early in zebrafish myogenesis are needed to specify slow fiber types as well as medial fast fibers (Currie & Ingham, 1996; Wolff et al., 2003); and hedgehog is important for later fast muscle differentiation (Feng, Adiarte, & Devoto, 2006; Hammond et al., 2007). Hedgehog signaling also promotes slow muscle fiber identity in mouse and chick embryonic limb muscles (Cann, Lee, & Stockdale, 1999; Hu, McGlinn, Harfe, Kardon, & Tabin, 2012; Li et al., 2004). Recent studies in zebrafish have identified roles for new factors in slow- and/or fast-twitch fiber specification, including transcriptional regulators (Devakanmalai, Zumrut, & Ozbudak, 2013; Yao, Farr, Tapscott, & Maves, 2013), components of signaling pathways (Tu, Tsao, Lee, & Yang, 2014), and microRNAs (Ketley et al., 2013; Lin et al., 2013; O’Brien, Hernandez-Lagunas, Artinger, & Ford, 2014). Whether these new factors play conserved roles in mammalian fiber type specification remains to be determined.

Table 1.

Markers for Cell Types in Zebrafish Skeletal Muscle

Cell Type	Transgenic Marker	Gene Marker	Antibody
Satellite cells	Tg(pax7a:eGFP), Tg(pax7a: GFP), Tg(pax7b:gal4;UAS: GFP), Tg(pax3a:GFP)	pax7a	Pax3, Pax7
Superficial slow-twitch fibers	Tg(smyhc1:GFP), Tg(prdm1a: GFP), Tg(tnnc1b:eGFP)	tnnt1, mybpc1, smyhc1, prdm1a, tnnc1b	F59, S58, Smyhc, Prox1
Muscle pioneers	Tg(eng2a:eGFP)	tnnt1, mybpc1, smyhc1, eng1a, eng1b, eng2a, eng2b	Engrailed, F59, S58, Smyhc, Prox1
Medial fast-twitch fibers	Tg(eng2a:eGFP), Tg(mylpfa: H2B-GFP), Tg(mylpfa:GFP)	tpma, tnnt3a/b, tnnc2, myhz2, mylpfa, eng1a, eng1b, eng2a, eng2b	Engrailed, F310
Fast-twitch fibers	Tg(mylpfa:GFP), Tg(mylpfa: H2B-GFP)	tpma, tnnt3a/b, tnnc2, myhz2, mylpfa	F310

Genes common to slow- and fast-twitch fibers: acta1, myhz1, myod, myf5, myog.

Antibodies or stains common to slow- and fast-twitch fibers: MF20, phalloidin.

2.2. Early Muscle Fiber Morphogenesis

2.2.1. Slow-Twitch Muscle Elongation

Presumptive slow-twitch muscle cells are initially stacked like short bricks: they are cuboidal cells arranged in a sheet of approximately 4×5 cells. Next, this sheet of epithelial cells transforms into a stack, about 20 deep, of elongated slow-twitch muscle cells. Live imaging elucidated the cellular behaviors that underlie this transformation (Daggett, Domingo, Currie, & Amacher, 2007; Yin & Solnica-Krezel, 2007). Slow-twitch cells first elongate in the dorsal–ventral direction (perpendicular to their final anterior–posterior alignment). Next, slow-twitch cells rotate such that they are aligned anterior-posteriorly (Yin & Solnica-Krezel, 2007). The actin regulatory protein Cap1 is specifically expressed in these cells and is required for elongation (Daggett et al., 2007). In the future, it will be interesting to determine the molecular mechanisms that mediate slow-twitch fiber elongation.

2.2.2. Slow-Twitch Muscle Migration

After they have elongated, slow-twitch fibers migrate laterally through the presumptive fast-twitch muscle domain. There are two notable aspects of this cellular behavior. One is that slow-twitch fibers are both necessary and sufficient to trigger fast-twitch muscle cell elongation (Henry & Amacher, 2004). It will be important to identify the molecular cues involved, because that knowledge could be applied to promote the growth and/or regeneration of fast-twitch fibers. Another interesting aspect of slow-twitch fiber migration is that these cells migrate after they have differentiated. These aspects have made slow-twitch muscle migration an enticing phenomenon.

One mechanism that regulates slow-twitch muscle migration is differential cell–cell adhesion (Cortes et al., 2003). N-cadherin is initially expressed in both presumptive slow and presumptive fast muscle cells (Cortes et al., 2003). M-cadherin is initially expressed only in slow-twitch muscle fibers but is upregulated in the fast-twitch muscle fibers medial to slow-twitch muscle migration (Cortes et al., 2003). This results in a “wave” of cadherin expression as slow-twitch muscle fibers migrate (Fig. 2). Ectopic expression of N-cadherin traps individual slow muscle cells in the fast muscle domain (Cortes et al., 2003). Taken together, these data suggest that differential cell adhesion is one mechanism that contributes to proper slow-twitch muscle cell migration.

Fig. 2.

Mechanisms regulating slow-twitch muscle migration in zebrafish. A medial focal plane of the dorsal half of one muscle segment in the transition zone of a side oriented, anterior left, dorsal top zebrafish embryo is cartooned. Cell–cell adhesion mediated by N- and M-cadherin plays a role in efficient and coordinated slow muscle migration. Dynamic zones of N- and/or M-cadherin expression instruct migrating slow-twitch muscle cells through the presumptive fast-twitch muscle domain toward the lateral surface of the embryo. Adjacent to the notochord, a zone of N-cadherin expression (red) occurs where the nonmigratory slow-twitch muscle pioneer fibers (light blue) reside. Fast-twitch muscle cells undergo morphogenesis (dark gray) medial to the migrating front of slow-twitch fibers in a zone of M-cadherin expression (yellow). A zone of M- and N-cadherin expression (orange) marks the migrating front of slow-twitch fibers (blue). The transcription factor Prdm1a is required cell autonomously to polarize filopodia to the leading edge of migrating slow-twitch fibers and the scaffolding protein Akap12 is also required for slow-twitch muscle migration in zebrafish. Presumptive fast-twitch muscle cells (light gray) lateral to the migrating front of slow-twitch fibers (blue) have not yet undergone morphogenesis and are in a zone of N-cadherin expression (red).

Evidence from experiments with zebrafish mutants and in vivo imaging studies recently showed that slow-twitch muscle cell migration is an active, cell autonomous migration mediated by the transcription factor Prdm1a (Ono, Yu, Jackson, Parkin, & Ingham, 2015). Slow-twitch muscle fibers in prdm1a mutant embryos migrate more slowly and not as far as slow fibers in wild-type embryos (Ono et al., 2015). Elegant experiments using genetic mosaic analysis and time-lapse confocal microscopy indicate that slowtwitch muscle cells actively migrate by extending filopodia along their leading edge (Fig. 2 and Ono et al., 2015). This process is disrupted in prdm1a mutant embryos, where slow-twitch muscle cells extend filopodia along both their leading and trailing edges (Ono et al., 2015). Taken together, these data suggest the model that Prdm1a, which is induced by Hedgehog signaling (Baxendale et al., 2004), mediates polarization of slow-twitch muscle fibers and that this polarization enables directed cell migration (Ono et al., 2015). One candidate for this is Akap12, a slow muscle-specific cytoskeletal remodeling protein required for slow muscle migration (Kim, Kim, Jeong, Han, & Kim, 2014) (Fig. 2). In light of these recent data, what role does differential cadherin expression play in this process? Given the compelling evidence that both M- and N-cadherin are required for slow muscle migration, we suggest that these proteins contribute to efficient migration, that their differential expression functions as an additional guidance cue, and that homophilic cell–cell adhesion enables the collective migration of the slow muscle cell cohort.

2.2.3. Fast-Twitch Muscle Development

Live imaging studies identified morphogenetic cell behaviors mediating fast muscle fiber morphogenesis (Snow, Goody, et al., 2008). Similar behaviors have been identified during muscle morphogenesis in the chick embryo (Gros, Scaal, & Marcelle, 2004; Gros, Serralbo, & Marcelle, 2009), suggesting potential conservation of muscle morphogenesis in vertebrates.

First, short fast muscle precursor cells exhibit dynamic protrusive activity in all directions (Snow, Goody, et al., 2008) (Fig. 3). Second, protrusions are only formed in the direction of elongation (Snow, Goody, et al., 2008). During this phase, intercalation/elongation of muscle fibers results in elongation of fast muscle cells. Next is boundary capture, where muscle cells interact with the anterior and posterior myotome boundaries and cease elongating. After boundary capture, cellular shape changes generate more regularly shaped myotubes (myotube formation) (Snow, Goody, et al., 2008). Fusion occurs concurrently with myotube formation. However, fusion is not necessary for myotube formation because some zebrafish mutant strains with defective muscle cell fusion have mononucleate, regularly shaped myotubes (Powell & Wright, 2011). The next phase includes myofibril formation, the generation of strings of sarcomeres, which are the basic units that generate muscle contraction. The earlier studies have provided a rudimentary understanding of cellular mechanisms and have allowed for the identification of some of the molecular mechanisms that drive muscle morphogenesis in zebrafish.

Fig. 3.

The cellular morphologies of fast-twitch muscle cells during zebrafish muscle development. Developmental time is represented vertically with earlier developmental time on the top and later developmental time on the bottom. Presumptive fast-twitch muscle fibers are short, extend protrusions in all directions, and contain one nucleus (light green cell on top, gray circle represents nucleus). Next, cells extend protrusions and elongate. Elongating fast-twitch muscle cells then interact with the extracellular matrix at segment boundaries (i.e., boundary capture) and cease elongating. Fusion generates multinucleated cells (green cell, fourth from top). Proteins are organized into repeating contractile units called sarcomeres during myofibrillogenesis (white outlines in dark green cells represent sarcomeres, fifth from top). Sarcomeres and interactions between functional muscle fibers and the extracellular matrix at segment boundaries must be actively maintained during muscle use and further growth of zebrafish embryos. Fusion also continually occurs (dark green cell, sixth from top).

2.2.4. Fast-Twitch Muscle Cell Elongation

The basement membrane (BM) protein laminin 111, and its receptor, Integrin alpha6beta1 are necessary for efficient and oriented fast muscle cell elongation in zebrafish (Goody et al., 2012; Snow, Goody, et al., 2008). Muscle cell adhesion to the extracellular matrix (ECM) is also critical for fast muscle cell elongation, as laminin beta1 and gamma1 are necessary for timely fast muscle cell elongation (Snow, Goody, et al., 2008). Hedgehog signaling is necessary for the delayed fast-twitch muscle cell elongation that occurs in laminin beta1 or gamma1 mutant zebrafish (Peterson & Henry, 2010). Understanding the latent secondary, or “back-up,” mechanisms for developmental processes will provide deeper insight into the mechanisms of development.

2.2.5. Boundary Capture and Myotube Formation

Axial skeletal muscle fibers stop elongating when they contact the anterior and posterior segment boundaries. This process is called boundary capture and controls the length of muscle fibers. In zebrafish, boundary capture is mediated by laminin 111 signaling (Snow, Goody, et al., 2008). The ECM glycoprotein Fibronectin (Fn) also plays a role in regulating length of axial skeletal muscle fibers in zebrafish (Snow, Peterson, Khalil, & Henry, 2008). After fast-twitch muscle cells have elongated, they are initially irregularly shaped and mononucleate. The next phase of muscle morphogenesis involves transition from an irregularly shaped cell to a rod-shaped myotube (Snow, Goody, et al., 2008). Increased organization of the actin cytoskeleton correlates with myotube formation. Actin organization occurs prior to the formation of sarcomeres (Mei, Li, & Gui, 2009; Naganawa & Hirata, 2011), but the mechanisms that regulate this distinct cellular process in vertebrates in vivo are unclear. What little is known came from work in zebrafish, which identified the actin regulatory proteins flightless homolog 1 (Naganawa & Hirata, 2011) and myotubularin-related 8 (Mei et al., 2009). A major question is what is the function of this increased actin organization and is this change in actin a necessary step for sarcomere formation? Altogether, the transparency of zebrafish embryos along with the spatial segregation of slow- and fast-twitch muscle fibers has enabled visualization of cellular behaviors during morphogenesis.

3. TRADEMARK TRAITS OF SKELETAL MUSCLE: MUSCLE CELL FUSION AND SARCOMERES

3.1. Muscle Cell Fusion

A defining trait of skeletal muscle cells is multiple nuclei per cell. This multinucleate state is achieved via cell fusion events. Skeletal muscle cells fuse during primary muscle morphogenesis, during hypertrophy (when skeletal muscle cells become bigger in response to exercise), and during repair (in response to damage, injury, or disease). It is important to understand the regulation of vertebrate skeletal muscle cell fusion at the cellular and molecular levels because this information could potentially hold clues to understanding muscle regeneration and to identifying targets for muscle disease therapies. The zebrafish model has been particularly informative in terms of identifying genes and their roles in vertebrate muscle cell fusion.

3.1.1. Vertebrate-Specific Innovations

A landmark study identified two vertebrate-specific cell surface receptors required for fast-twitch muscle fusion in zebrafish: Jamb and Jamc (Powell & Wright, 2011). These proteins are deuterostome-specific Ig superfamily cell surface proteins with two extracellular Ig superfamily domains, a transmembrane domain, and a short cytoplasmic domain. Although fast-twitch fibers are specified and appear morphologically normal in jamb or jamc mutants, fusion does not occur (Powell & Wright, 2011). Biochemical and genetic mosaic analyses demonstrated that Jamb and Jamc interact and this interaction is required for fusion (Powell & Wright, 2011).

Another landmark study identified the transmembrane protein Tmem8c, also known as Myomaker, as necessary for myoblast fusion and muscle formation in mouse embryos (Millay et al., 2013). Myomaker is the only vertebrate muscle-specific factor identified to date that is both necessary and sufficient to direct myoblast fusion. Although the intracellular C-terminal region of Myomaker is necessary for myoblast fusion and Myomaker exerts its function through the cytoskeleton (Millay et al., 2016, 2013), it is not yet clear how Myomaker interacts with other factors known to play roles in myoblast fusion. Myomaker knockdown via morpholino injection in zebrafish embryos results in short mononucleate presumptive fast muscle cells when analyzed at a developmental stage where fast muscle fibers should be long (Landemaine, Rescan, & Gabillard, 2014). One interpretation of these findings is that Myomaker is required for muscle cell fusion in zebrafish. An alternate interpretation is that Myomaker is required for timely fast-twitch muscle cell elongation. Given that the initial myotome contains mononucleate muscle cells in zebrafish, any putative fusion mutant would be expected to have long muscle cells that are mononucleate, as is the case for jamb or jamc mutants (Powell & Wright, 2011).

The distinction between fast muscle cell elongation vs fusion phenotypes impacts the interpretation of the function of Ckip-1 in muscle cell fusion in zebrafish. Ckip-1 was shown to play a role in mammalian muscle fusion in C2C12 myoblasts (Baas et al., 2012). Like Myomaker, knockdown of Ckip-1 in zebrafish resulted in short mononucleate cells that could result from either fusion or elongation defects (Baas et al., 2012). Future studies should examine whether fast muscle cell elongation recovers in Myomaker- or Ckip-1-deficient zebrafish and, if so, whether long fast-twitch fibers contain fewer nuclei.

3.1.2. Do Proteins From Insects Mediate Cell Fusion in Vertebrates?

Zebrafish orthologs of Drosophila Kirre (Kirrel3l), Drosophila CDM and Crk family members (Dock1, Dock5, Crk, and Crkl), and of Drosophila Sns (Nephrin) have been identified (Moore, Parkin, Bidet, & Ingham, 2007; Sohn et al., 2009; Srinivas, Woo, Leong, & Roy, 2007). Loss of function for many of these factors results in short mononucleate muscle cells in zebrafish that could result from either elongation or fusion defects (Moore et al., 2007; Sohn et al., 2009; Srinivas et al., 2007). By immunohistochemically analyzing phenotypes later in development, it was shown that fast-twitch muscle elongation recovered and there were fewer nuclei in long fast-twitch muscle fibers in Kirrel3l-deficient zebrafish (Srinivas et al., 2007). Therefore, Kirrel3l does appear to play a conserved role in muscle cell fusion. Short fast-twitch muscle cells were also observed upon knockdown of Nephrin in zebrafish (Sohn et al., 2009). Whether muscle cell elongation recovers and whether there were actual fusion defects in these embryos remains to be determined. Taken together, these data suggest the hypothesis that the heterophilic adhesion mediated by Kirrel3l and Nephrin may play dual roles in zebrafish muscle development: initially mediating fast-twitch muscle cell elongation and then promoting fusion. Improved visualization of muscle cell fusion using fluorescent markers will greatly assist with this issue. For example, the Tg(mylpfa:H2B-GFP) zebrafish line, in which fast muscle cell nuclei fluoresce, should be very useful toward this end (Zhang & Roy, 2016).

3.2. Myofibrillogenesis

A second trademark feature of skeletal muscle cells is the repeating, contractile units called sarcomeres. Movement is generated by the coordinated contraction and relaxation of sarcomeres within muscle cells. Sarcomeres consist of densely packed and highly ordered myofilaments (i.e., actin and myosin) and other proteins that bundle them together, anchor them to the sarcolemma, or are otherwise involved. Myofibrillogenesis is the process by which long strings of sarcomeres are generated in muscle cells. Myofibrillogenesis occurs during primary muscle development as well as during muscle growth and repair. Disruptions of myofibrillogenesis and/or the inability to maintain sarcomeres leads to a subset of myopathies/cardiomyopathies called the myofibrillar myopathies. Thus, understanding how sarcomeres are assembled and how sarcomeres are incorporated into myofibrils are critical issues in muscle development and disease (see Section 6.4).

3.2.1. Sarcomeres as Dynamic Structures

The ability to image myofibrillogenesis in zebrafish has dramatically contributed to our understanding of sarcomere formation and homeostasis. Fluorescence recovery after photobleaching (FRAP) analysis demonstrated that sarcomeric proteins—Actin, Alpha-actinin, FATZ, Myotilin, and Telethonin—are dynamically exchanged between sarcomeres and a cytoplasmic pool (Sanger, Wang, Holloway, Du, & Sanger, 2009). Live imaging also provided insight into the biomechanics of muscle contractions. Muscle contraction is powered by skeletal myosin transducing ATP into mechanical work. Understanding how myosin lever-arm orientation changes during aging and disease may provide insight into the etiology of how myopathies linked to myosin deleteriously affect muscle function. The application of superresolution microscopy to measuring the orientation of single myosin molecules in zebrafish skeletal muscle revealed differences in the orientations of myosin lever-arms in skeletal vs cardiac muscle (Sun et al., 2014). Applying this technology to different types of muscle diseases will be an important approach to learn more about how myosin mediates muscle function. Additionally, developing methods to utilize superresolution microscopy to simultaneously analyze dynamics of multiple sarcomeric proteins and their conformations during different types of exercise will greatly enhance our understanding of how myofibrils regulate muscle performance and health.

3.2.2. It’s All About Connections: Impacts of the Membrane, ECM, and Neuromuscular Junctions on Myofibrillogenesis

Recent insight indicates that myofibrillogenesis is integrated with, and regulated by, adhesion to the sarcolemma and to the ECM (Raeker, Shavit, Dowling, Michele, & Russell, 2014; Weitkunat, Kaya-Çopur, Grill, & Schnorrer, 2014). In Drosophila, attachment to the ECM and organization of myosin into myofibrils results in an increase in tension, which induces myofibrillogenesis simultaneously throughout the length of whole muscle cells (Weitkunat et al., 2014). In support of the idea that adhesion to the sarcolemma and the ECM could regulate myofibrillogenesis in vertebrates, it has been shown in zebrafish that the sarcolemma protein Caveolin-3 (Nixon et al., 2005) and the ECM proteins Periostin (Kudo, Amizuka, Araki, Inohaya, & Kudo, 2004), Collagen XV (Pagnon-Minot et al., 2008), and transforming growth factor beta induced (TGFBi) (Kim & Ingham, 2009) are necessary for myofibrillogenesis. Additionally, disrupting cholesterol in cell membranes via statins disrupted myofibrillogenesis in zebrafish (Campos et al., 2015; Huang et al., 2011). It was hypothesized that cholesterol might be important for maintenance of membrane micro-domains and potentially for secretion of ECM proteins to the extracellular space (Campos et al., 2015). Thus, studies in the zebrafish support the recent hypothesis, generated from experiments in Drosophila, that tension mediated by adhesion to the ECM may mediate myofibrillogenesis. Furthermore, myofibrillogenesis may also depend on the neuromuscular junction (NMJ) because genes encoding multiple acetylcholine receptor (AchR) subunits have been shown to have roles in myofibrillogenesis in zebrafish (Behra et al., 2002; Brennan, Mangoli, Dyer, & Ashworth, 2005; van der Meulen, Schipper, van Leeuwen, & Kranenbarg, 2005). Integrating studies of the sarcolemma, muscle cell–ECM adhesion, and NMJ with myofibrillogenesis certainly adds layers of complexity, but also the opportunity for an integrated physiological understanding of myofibrillogenesis and muscle development and homeostasis. The zebrafish system, with its accessible genetics and live imaging capabilities, is poised to have a tremendous impact on our understanding of myofibrillogenesis and sarcomere maintenance.

3.2.3. Interactions Between Myofibrillogenesis and Muscle Fusion?

Myofibrillogenesis and muscle cell fusion may be more closely linked than has been appreciated. In particular, a recent study links metabolism with both muscle cell fusion and muscle growth (Tixier et al., 2013). In this study, a screen for genes expressed in both zebrafish and Drosophila muscle identified the glycolysis gene pgam2. Injection of morpholinos against zebrafish pgam2 resulted in thinner fast muscle fibers with significantly fewer nuclei (Tixier et al., 2013). The observation of thinner fast fibers in zebrafish embryos is highly suggestive of defects in myofibrillogenesis (Kim and Ingham, 2009; Kotani et al., 2015). While the observation of thinner slowtwitch fibers in zebrafish would suggest that myofibrillogenesis could regulate fiber thickness independently of fusion events, the occurrence of thinner fast fibers suggests the possibility that there is crosstalk between the processes of fusion and myofibrillogenesis in regulating fast fiber thickness. Thus, when thinner fast-twitch muscle fibers are observed, we believe investigating whether fusion and/or myofibrillogenesis defects are contributing factors may be a fruitful avenue of investigation. The functional/biochemical consequences and potential translational implications of this thinner fiber phenotype also need to be addressed in future studies.

4. MTJ DEVELOPMENT

4.1. Developmental Defects in the MTJ May Lead to Muscle Disease

Muscles and tendons function as an integrated unit to transduce force to the skeletal system and stabilize joints. Muscle and tendons anchor to each other through receptor complexes that indirectly link the actin cytoskeleton inside muscle cells to the tendon ECM. Cell–ECM adhesions mechanically link muscles to tendons and are required for muscle physiology and function. The main receptor complexes linking muscle cells to their surrounding ECM are the dystrophin–glycoprotein complex and Integrins (Marshall & Crosbie-Watson, 2013).

Many muscle diseases, such as Duchenne, Becker, Merosin-deficient muscular dystrophies, and CMD with integrin deficiency, result from mutations that disrupt adhesion of muscle fibers to their surrounding BM. This weakened link between muscle fibers and their surrounding BM results in increased susceptibility to fiber damage and death during repeated cycles of contraction and relaxation. Data from mouse and zebrafish models of muscular dystrophies implicate MTJ damage and/or fiber detachment from the MTJ as the primary etiology behind muscle dysfunction in some types of muscular dystrophy (Hall et al., 2007; Jacoby et al., 2009; Mayer et al., 1997). It is not known if detachment of muscle fibers from the MTJ BM contributes to human muscular dystrophies because biopsies are excised away from the MTJ to avoid injury to the tendon. However, MRI studies in humans do suggest that muscle damage is more severe closer to the MTJ (Hasegawa et al., 1992; Nagao et al., 1991). Thus, studies of MTJ formation during zebrafish development could provide important clinical insights into how defects in this structure contribute to the onset or progression of muscular dystrophy.

4.2. MTJ Composition During Early Development

The composition of the MTJ ECM changes during MTJ development (Fig. 4). The ECM protein Fn is abundant at initial somite boundaries (Julich, Geisler, & Holley, 2005; Koshida et al., 2005; Trinh & Stainier, 2004), but is downregulated after fast-twitch muscle fibers elongate and attach to the MTJ (Snow & Henry, 2009). Increased organization of the BM protein laminin 111 is observed concomitant with Fn downregulation. These changes occur by approximately 1 day postfertilization (dpf) in zebrafish embryos. At this stage, there are sparse collagen fibrils that are not well organized or anchored to the BM (Charvet, Malbouyres, Pagnon-Minot, Ruggiero, & Le Guellec, 2011). Similar to muscle development in mouse embryos, the predominant laminin isoform in the BM shifts from laminin 111 to laminin 211; this likely occurs between 2 and 3 dpf in zebrafish (Hall et al., 2007). By 3 dpf, the myoseptum has enlarged to 500 nm and contains thick collagen fibrils that are aligned either transversely or longitudinally (Charvet et al., 2011).

Fig. 4.

Molecular changes at the developing zebrafish MTJ. Simplified cartoon of some of the major proteins involved in muscle cell–ECM adhesion at the zebrafish MTJ over developmental time. Developmental time progresses horizontally from left to right. Concurrent with the cellular changes that accompany fast-twitch muscle morphogenesis (cells in various shades of gray), the molecular composition of the MTJ ECM changes from Fn-rich somite boundaries, to laminin 111-rich nascent MTJs, to laminin 211-rich mature MTJs (chevron-shaped lines represent segment boundaries). Blue muscle fibers represent slow-twitch muscle pioneers. Magnifications of the boxed section of each segment boundary cartoon the muscle cell plasma membrane (sarcolemma), transmembrane receptors, and major components of the MTJ ECM (see key). Multiple transmembrane receptor complexes indirectly link the intracellular actin cytoskeleton (not shown) to the ECM. At the developmental stage furthest to the right, the lamininrich basement membrane (BM) attaches to the collagen-rich interstitial matrix. Mutations in the genes encoding these adhesion components cause many muscle diseases.

Many other ECM proteins expressed at the MTJ such as Periostin, Thrombospondin, and multiple collagens (Kudo et al., 2004; Pagnon-Minot et al., 2008; Subramanian & Schilling, 2014; Telfer, Busta, Bonnemann, Feldman, & Dowling, 2010) play a critical role in maintaining the MTJ during repeated cycles of contraction and relaxation, injury, and repair. Zebrafish mutant or morpholino-treated embryos deficient for laminin alpha2, Thrombospondin 4b, or Collagen 22a1 proteins have muscle defects characterized by muscle fibers retracting from the MTJ and the defects are enhanced when muscle contraction is stimulated (either mechanically or electrically) (Charvet et al., 2013; Hall et al., 2007; Subramanian & Schilling, 2014). In these examples, the muscle fibers do not rupture but retract from the MTJ intact. Therefore, these proteins play a role in muscle health by maintaining integrity of muscle cell adhesion to the ECM at the MTJ.

4.3. Mechanisms of Laminin 111 Organization and Regulation

Given the critical role that specific ECM proteins play in MTJ development and homeostasis, it is necessary to determine the mechanisms underlying the dynamic changes in ECM composition during MTJ development. Understanding these mechanisms may provide insight into therapeutic strategies in muscle disease.

Nicotinamide riboside kinase (Nrk)-mediated NAD+ synthesis is conserved from yeast to humans, and members of the Nrk family function to generate nicotinamide adenine dinucleotide (NAD+) (Bieganowski & Brenner, 2004; Tempel et al., 2007). MTJ morphogenesis in Nrk2b-deficient zebrafish embryos can be rescued by providing exogenous NAD+, and it was shown that this metabolite is critical for laminin 111 organization at the MTJ. This is important because some models of CMDs present with disorganized laminin 111 at the MTJ (Goody et al., 2012). NAD+ acts as a small molecule agonist of laminin 111 organization and muscle fiber–MTJ adhesion; exogenous NAD+ supplementation reduced muscle fiber degeneration and improved mobility in zebrafish models of some CMDs (Goody et al., 2012). These data highlight the important contribution that asking basic questions about developmental biology can have with regards to potential therapies. It will be interesting to identify additional components in the Nrk2b pathway. In this regard, laminin 111 organization also appears to be reduced in a zebrafish model of autosomal recessive adolescent onset distal myopathy caused by mutations in adssl1 (Park et al., 2016). Adssl1 is an adenylosuccinate synthase that is expressed mainly in muscle. Adssl1 may interact with the Nrk2b pathway. Nrk2b is a nicotin-amide riboside kinase and thus requires nicotinamide riboside. Human purine nucleotide phosphorylase is required for nicotinamide riboside utilization (Belenky, Christensen, Gazzaniga, Pletnev, & Brenner, 2009) and Adssl1 plays a role in purine biosynthesis (Park et al., 2016). Thus, in the future, it will be interesting to determine the mechanisms of action of Adssl1 and whether Adssl1 interacts with the Nrk2b pathway to promote laminin 111 organization at the MTJ.

4.4. Mechanisms of Fn Regulation

Fn is critically important in multiple developmental contexts (somite segmentation, cell migration, cellular branching) and also for regeneration (reviewed in Bentzinger et al., 2013; Goody & Henry, 2010). Fn adjacent to fast-twitch muscle fibers becomes downregulated during MTJ development (Snow & Henry, 2009) but, until recently, it was unknown how Fn was degraded in vivo. Recent work showed that laminin 111 organization acts as a “checkpoint” for Fn downregulation. Laminin 111 organization potentiates the localization of a matrix metalloproteinase (Mmp11), which is necessary and sufficient for Fn downregulation at the zebrafish fast-twitch muscle MTJ, providing one mechanism by which laminin 111 acts as a signal that results in Fn downregulation (Jenkins, Alrowaished, Goody, Crawford, & Henry, 2016). Thus, understanding how crosstalk between ECM proteins influences cellular outputs during development resulted in the identification of a MMP that may have therapeutic utility in the many conditions where fibrosis contributes to disease progression.

5. FROM DEVELOPMENT TO HOMEOSTASIS

An interesting chapter in a recent Current Topics in Developmental Biology volume asks (1) if we should define developmental biology and, if so, (2) what would that definition be (Pradeu et al., 2016)? These questions are especially important given that the impact of development on health, aging, and congenital diseases is becoming more recognized. Here, we discuss evidence for the concept that muscle development does not “end” per se, but, rather, active homeostasis maintains muscle structure even after initial muscle development occurs.

5.1. Homeostasis Part 1: The MTJ and Regulation of Muscle Fiber Length

The axis of time is critical when interpreting phenotypes because sometimes a given phenotype could result from disruptions at multiple stages of development. For example, defects in muscle fiber length can be due to either early boundary defects or later homeostasis defects. As mentioned earlier, elongating muscle fibers are “captured” by the ECM at the MTJ, which is derived from somite boundaries. If somite boundaries do not form properly, then elongating fibers are not captured and abnormally long muscle fibers are observed (Snow, Peterson, et al., 2008). Such cases are examples of early developmental defects. However, in zebrafish muscle disease models, we and others provided examples where somite boundaries form normally, but later in development, muscle fibers become abnormally long (Etard et al., 2010; Goody, Kelly, Lessard, Khalil, & Henry, 2010; Goody et al., 2012; Snow, Goody, et al., 2008) (Fig. 5A). This phenotype shows that maintenance of fiber length and MTJ integrity is an active process, which we refer to as active homeostasis.

Fig. 5.

Schematic of muscle disease phenotypes in zebrafish embryos. Two muscle segments of a side oriented, anterior left, dorsal top zebrafish embryo are diagrammed at the different developmental stages indicated in the figure. Muscle fibers are gray cylinders and segment boundaries are black chevron-shaped lines. Muscle fibers with disease phenotypes are dark gray or red. (A) In zebrafish deficient in laminin beta1, laminin gamma1, Integrin alpha6, Nrk2b, Apo2a, or Periostin gene products, MTJs fail and some muscle fibers are abnormally long (dark gray fibers). (B) In 72 hpf dmd mutant zebrafish, fiber deta’chment is observed. Detached (short red fiber) and even some attached fibers (long red fiber) uptake Evans blue dye, showing that dmd mutations result in loss of sarcolemma integrity. Fiber detachment occurs independently of loss of sarcolemma integrity in laminin alpha2 or dystroglycan mutant zebrafish (short dark gray fiber), suggesting that loss of adhesion to the ECM can also cause fiber detachment. Fibers detach yet remain viable in laminin beta2 mutant zebrafish by forming an ectopic MTJ (dark gray fiber with white line in the middle representing an ectopic MTJ).

Evidence from zebrafish studies elucidated a critical role for laminin 111 in fiber length and MTJ integrity. The laminin chains laminin beta1 and laminin gamma1 as well as an integrin receptor for laminin 111, Integrin alpha6, are required to maintain fiber length/MTJ integrity (Goody et al., 2012; Snow, Goody, et al., 2008). Data also indicate that just the presence of laminin 111 is not sufficient: laminin 111 is present but disrupted in zebrafish deficient for Nrk2b or Apo2a (Etard et al., 2010). In the latter case, Hsp45b binds to Apo2a and its silencing causes a similar MTJ failure phenotype, suggesting that both are required for MTJ integrity (Etard et al., 2010). Laminin 111 is not the only ECM protein required for MTJ maintenance, as zebrafish lacking Periostin also show MTJ failure and abnormally long muscle fibers (Kudo et al., 2004). In contrast, although collagens are critical components of the MTJ, disruption of many of the collagen chains studied thus far (Col15a1, Col6a2, Col6a4a, Col4a4b, Col22a1) does not result in MTJ failure or abnormally long fibers in zebrafish (Charvet et al., 2013; Pagnon-Minot et al., 2008; Ramanoudjame et al., 2015). These data highlight the utility of the zebrafish model in identifying discrete roles for different MTJ ECM proteins during development and homeostasis.

5.2. Homeostasis Part 2: When the Resilience of Muscle Fibers Fails in Dystrophies

Muscle homeostasis involves maintenance of sarcolemma integrity, fiber adhesion to the ECM, ECM integrity, and the satellite cell pool. Muscle use can result in small tears in the sarcolemma (plasma membrane) that need to be repaired. Muscle fibers can also sustain enough damage to the sarcolemma that they die, at which point the muscle resident stem cells (satellite cells) mediate muscle repair. There is a great deal of interest in understanding muscle repair as it pertains to muscle aging, traumatic muscle injury, and muscle diseases, and zebrafish provides a powerful model for such studies.

Dysferlin is a membrane protein important for membrane resealing. Mutations in human DYSF can result in multiple types of myopathies (Bashir et al., 1998; Liu et al., 1998). In mammalian cell culture models, Dysferlin has been associated with sites of damaged muscle cell membranes, suggesting that it might play a role in membrane repair (Lennon et al., 2003). Important questions being addressed using zebrafish are (1) how does Dysferlin potentiate membrane healing in vivo, (2) how does the membrane reseal and where does the membrane patch come from, and (3) can any membrane domain contribute to repair? These questions were addressed by using live imaging to monitor membrane repair in zebrafish embryos (Roostalu & Strähle, 2012). The fact that Dysferlin, but not many other membrane/vesicle markers tested, rapidly relocates to damage suggests the hypothesis that there is a specialized membrane domain, rich in Dysferlin, that participates in membrane repair.

By combining live imaging and genetic approaches in zebrafish, the steps for the highly ordered process of membrane repair involving Annexins and Dysferlin have been uncovered (Roostalu & Strähle, 2012). First, the lipidbinding Annexin A6 accumulates rapidly and might “clog” the membrane. Simultaneously, Dysferlin from a specialized membrane subdomain accumulates at the damage site. Next, Annexin A2a is added to the patch. Finally, Annexin A1a arrives. Thus, these data support the model where a multilayered scaffold forms in a sequential fashion to promote membrane repair (Roostalu & Strähle, 2012). The hope is that this information could be used to develop methodologies for sarcolemma repair to either enhance muscle recovery after injury or potentially reduce degeneration of diseased muscle.

6. NEW INSIGHTS INTO MUSCLE DISEASE FROM ZEBRAFISH

One premise underlying the effort dedicated to understanding muscle development, growth, and repair is that knowledge gained can be leveraged to develop therapies to combat muscle disease. The zebrafish model has made many important contributions to our understanding of muscle disease, particularly with regards to muscular dystrophies and myopathies. Here, we will discuss some recent studies that used the zebrafish model to provide insight into muscle disease.

6.1. Testing Candidate Disease Causing Mutations

Morpholinos can be used to knockdown gene expression to help validate potential human disease causing genes identified by whole-exome sequencing: if injection of morpholinos against the gene in question results in muscle disease then this result bolsters the conclusion that the disease causing gene has been identified. Some recent examples include a distal myopathy caused by mutation in ADSSL1 (Park et al., 2016) and Native American myopathy in which Stac3 is mutated (Horstick et al., 2013). Similar approaches identified the causative genes for a nemaline myopathy (Yuen et al., 2014) and a mitochondrial myopathy (Shamseldin, Smith, et al., 2016). This approach was also used to identify a novel lethal congenital contracture syndrome caused by mutation in ZBTB42 (Patel et al., 2014). Thus, one use of the zebrafish system has been to fairly readily narrow down potential candidates to identify and/or validate mutations responsible for disease phenotypes.

6.2. Identification of Candidate Genes for Muscle Diseases of Unknown Genetic Cause

A recent study investigated how fast skeletal myosin-binding protein C (MyBPC-2) mediates sarcomere length and muscle contraction (Li et al., 2016). The finding that injection of morpholinos against MyBPC-2 results in myopathy may aid clinicians endeavoring to identify the genetic basis for unknown muscular dystrophies.

Genetic screens in zebrafish have also led to identification of genes disrupting muscle integrity. One such mutant, trage, had disruption of muscle integrity as assessed by birefringence and impaired swimming behavior (Berger et al., 2014). Positional cloning identified a nonsense mutation in tropomodulin4 (Berger et al., 2014). Phenotypic characterization showed that cytoplasmic rods, similar to the nemaline rods found in patients with nemaline myopathy, were prevalent in tmod4/trg mutants (Berger et al., 2014). Thus, tmod4/trg is a new candidate gene for nemaline myopathies of unknown etiology.

6.3. Novel Disease Mechanisms

The cellular pathology underlying disease progression is not well understood in muscular dystrophies and myopathies. This lack of understanding is a significant barrier for therapy development. Thus, identification of the cellular pathology in zebrafish disease models is a high priority because this knowledge could lead to new therapeutic avenues.

One example is the divergent cellular pathology between a Duchenne muscular dystrophy model and a dystroglycanopathy model. Muscle fibers in the zebrafish model of Duchenne muscular dystrophy uptake Evans blue dye, which only infiltrates cells that have damaged membranes. This dye was used to label and track the behavior of muscle fibers with damaged membranes. Evans blue dye-labeled fibers were observed to detach from the BM at the MTJ, and then undergo necrosis/apoptosis (Hall et al., 2007) (Fig. 5B). In contrast, muscle fibers in some zebrafish muscular dystrophy models actually detach from the BM at the MTJ prior to being permeable to Evans blue dye (Hall et al., 2007; Jacoby et al., 2009) (Fig. 5B). These data indicate that the cellular pathology of muscular dystrophies varies depending upon the genetic basis for disease. Remarkably, muscle fibers in one zebrafish mutant with muscular dystrophy survive after they detach from the MTJ (Jacoby et al., 2009). Muscle fibers in softy/lamb2 mutant zebrafish generate and attach to ectopic BMs and thus remain viable (Fig. 5B). Softy mutants are unique in that they are homozygous viable and it is hypothesized that these ectopic fiber terminations function to stabilize the damaged myotome (Jacoby et al., 2009). This stabilization could allow more robust regeneration or slow degeneration. Certainly, understanding the mechanisms underlying the recovery from early and severe muscle degeneration is critical because it may provide novel insight into therapeutic approaches.

Another example of zebrafish contributing new knowledge regarding disease mechanisms is autosomal dominant centronuclear myopathy, which is caused by mutations in dynamin-2 (DNM2) (Hanisch et al., 2011; Jeub et al., 2008). Generation of a zebrafish model revealed defects in excitation-contraction coupling that were hypothesized to be caused by defective membrane tubulation (Gibbs et al., 2014). This hypothesis was tested in an in vitro tubulation assay, where the dominant effect of the mutant DNM2 mRNA on normal tubule formation was confirmed. Thus, this study provides new insight into the cellular mechanisms of disease, and this knowledge can be used as a platform for therapy development.

6.4. Myofibrillar Myopathies

Myofibrillar myopathies are heterogeneous in the muscle groups affected and the timing of onset, but they share progressive muscle weakness and protein aggregation. Zebrafish models have been developed for most known myofibrillar myopathy-causing genes (Table 2). Interestingly, overexpression of mutated human BAG3 or FLNC, despite causing protein aggregation, rescues the fiber phenotype caused by knockdown of the corresponding zebrafish protein, suggesting that these disease causing mutations still result in functional proteins (Ruparelia et al., 2016, 2014). Zebrafish studies have linked impaired autophagy to disrupted myofibrillogenesis and shown that pharmacological activation of autophagy may be a potential therapeutic intervention for myofibrillar myopathies (Ruparelia et al., 2016; Skobo et al., 2014). The importance of autophagy in the dynamic maintenance of sarcomeres and the contractile function of muscle tissue suggests that investigating autophagy in muscle development, regeneration, aging, and other myopathies is an important goal for future studies.

7. CONCLUSIONS AND FUTURE PROSPECTS

One major premise underlying efforts to elucidate how muscle develops during embryogenesis is that understanding normal developmental processes and homeostasis can inform regenerative medicine (traumatic muscle injuries), aging (sarcopenia), and disease (cachexia, muscular dystrophies, myopathies). Studies in zebrafish have made seminal contributions to our understanding of vertebrate muscle specification and morphogenesis. The zebrafish model has also been invaluable for identification of disease causing genes and novel disease mechanisms. What are the major challenges that remain? With regards to muscular dystrophies, there are at least two pressing issues. One is the dramatic phenotypic spectrum of disease progression and the other is therapeutic crossover. One of the challenges in understanding muscular dystrophies and myopathies is that patients present with varying phenotypes, both within a particular type of dystrophy/myopathy, and between different dystrophies/myopathies. The extent to which potential aberrations in muscle development are linked to such variations in the age of onset and speed of progression is not well understood. Use of the zebrafish model has already made significant insights into the genetic basis of multiple muscle diseases, and revealed new cellular pathologies in these diseases. We predict that leveraging the advantages of the zebrafish model will have an enormous impact on understanding the molecular genetics underlying the dramatic phenotypic spectrum of muscle diseases. It is difficult to imagine another vertebrate model where it is feasible to quantify multiple phenotypic traits in large populations through time, followed by genomic analyses to identify biomarkers and loci that interact with disease progression. Thus, the zebrafish model, with its facile genetics and in vivo cell biology, stands poised to contribute tremendously to our understanding of muscle diseases. The question of whether there is any potential therapeutic crossover between treatments for Duchenne/Becker muscular dystrophy and other types of muscular dystrophies has yet to be answered. This is an important question because although therapeutic strides are being made for the most common type of muscular dystrophy, Duchenne/Becker muscular dystrophy, treatment for most other types of muscle diseases lags behind. Therefore, in the future, it is critical to determine whether differences in cellular pathology affect treatment options. This is an area where zebrafish models of muscle disease have high potential to contribute to pathologic mechanisms and questions regarding therapeutic crossover.