Abstract
The fusion of myoblasts into multinucleated muscle fibers is vital to skeletal muscle development, maintenance and regeneration. Genetic mutations in the Myomaker (mymk) gene cause Carey–Fineman–Ziter syndrome (CFZS) in human populations. To study the regulation of mymk gene expression and function, we generated three mymk mutant alleles in zebrafish using Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) technology and analyzed the effects of mymk knockout on muscle development and growth. Our studies demonstrated that knockout of mymk resulted in defective myoblast fusion in zebrafish embryos and increased mortality at larval stage around 35–45 days post-fertilization. The viable homozygous mutants were smaller in size and weighed approximately one-third the weight of the wild type (WT) sibling at 3 months old. The homozygous mutants showed craniofacial deformities, resembling the facial defect observed in human populations with CFZS. Histological analysis revealed that skeletal muscles of mymk mutants contained mainly small-size fibers and substantial intramuscular adipocyte infiltration. Single fiber analysis revealed that myofibers in mymk mutant were predominantly single-nucleated fibers. However, myofibers with multiple myonuclei were observed, although the number of nuclei per fiber was much less compared with that in WT fibers. Overexpression of sonic Hedgehog inhibited mymk expression in zebrafish embryos and blocked myoblast fusion. Collectively, these studies demonstrated that mymk is essential for myoblast fusion during muscle development and growth.
Introduction
Skeletal muscle arises from the fusion of myoblasts into multinucleated myofibers. Myoblast fusion is essential for skeletal muscle development, maintenance and regeneration. Genetic analyses in Drosophila and mouse have provided important insights into the molecular and cellular mechanisms of myoblast fusion (1–3). It appears that the molecular regulation of myoblast fusion is conserved during evolution. Several evolutionarily conserved regulators have been identified, ranging from cell adhesion molecules to actin polymerization regulators and mechanical sensors (4–7). Recent genetic studies in mouse and zebrafish have identified several membrane proteins that play important roles in myoblast fusion in vivo. These include two cell adhesion molecules, Jamb and Jamc (8), a transmembrane protein, Myomaker (mymk; 9–11) and its interacting protein Myomixer, also called Myomerger and Minion (12–15).
mymk is a seven-pass transmembrane protein of the TMEM8A and TMEM8B family (9). Expression analysis revealed that mymk is expressed in myoblasts of early stage mouse embryos. mymk expression is downregulated in adult mouse skeletal muscle but reactivated during muscle regeneration. Genetic analysis demonstrated that mymk is required for myoblast fusion during myogenesis. Knockout of mymk in mice resulted in defective myoblast fusion and postnatal lethality between days 1–7 after birth (9). Conditional knockout of mymk in muscle stem cells impaired skeletal muscle regeneration in adult mice (10).
mymk has been identified in other vertebrates from fish to chick and human (16–18). A recent multinational study demonstrated that autosomal recessive mutations in human mymk gene cause Carey–Fineman–Ziter syndrome (CFZS; 18). CFZS is associated with developmental and growth delay, feeding difficulty, facial weakness and microcephalus (19,20). Genetic studies in zebrafish showed that mymk is required for myoblast fusion in zebrafish embryos (11,16). However, its function in myoblast fusion at later stages during muscle growth and in adult fish remains elusive. Moreover, little is known about the regulation of mymk expression in skeletal muscles. It has been reported that mymk is expressed specifically in fast muscle precursors that are able to fuse in zebrafish embryos (16). However, a recent study revealed that mymk transcripts could also be detected in slow muscle precursors that normally do not fuse in zebrafish embryos (14).
Zebrafish is an excellent vertebrate model to study myoblast fusion in vivo due to its external development and the transparent nature of the embryos. Two types of myoblast precursors have been identified in early stage embryos that give rise to topologically separated embryonic slow and fast muscle fibers, respectively (21). The slow muscle precursors, also known as adaxial cells adjacent to the notochord, are induced by Hedgehog (Hh) signaling to differentiate into slow muscle fibers (22,23). In contrast, lateral presomitic cells remaining deep in the myotome differentiate into fast muscle fibers. Strikingly, embryonic slow and fast myoblasts show different characteristic in cell fusion. Embryonic slow muscles are single-nucleated myofibers that are formed without myoblast fusion (21), whereas fast myofibers are multinucleated muscle cells that are generated by myoblast fusion (21).
In this study, we characterized mymk gene expression and function in zebrafish skeletal muscles. Our studies demonstrated that overexpression of Sonic Hedgehog (Shh) in zebrafish embryos inhibited mymk expression and blocked myoblast fusion. To investigate the role of mymk in myoblast fusion during muscle growth, we generated three mymk mutant alleles in zebrafish using the CRISPR technology. All three mutant alleles contained a reading frame shift mutation in the mymk gene. The homozygous mymk mutants showed defective myoblast fusion and partial lethality at larval and juvenile stages. The surviving homozygous mutants were smaller in size and showed craniofacial deformities. Histological analysis revealed that mymk mutant contained mainly smaller single-nucleated myofibers and increased intramuscular adipocyte infiltration in skeletal muscles. Collectively, these studies demonstrated that mymk is essential for myoblast fusion during muscle development and growth.
Results
Characterization of mymk gene in zebrafish
The zebrafish mymk gene is approximately 8 kb long containing 6 exons. It encodes a predicted mymk protein of 220 amino acids. Bioinformatics analysis revealed the presence of mymk orthologues in vertebrates from fish to human. mymk protein sequences are highly conserved during evolution. Zebrafish mymk protein shares 75 and 73% sequence identities with the human and mouse mymk protein, respectively (Fig. 1A). mymk is a member of the TMEM8 family designated as TEMEM8C, which contains seven transmembrane domains (Fig. 1B). The majority of mymk protein is embedded in the plasma membrane with small extra cellular domains or N-terminal extracellular domain (Fig. 1B).
It is known that the zebrafish genome underwent an extra round of genome duplication during evolution, resulting in multiple copies for many zebrafish genes. To clarify the number of mymk gene in zebrafish, we performed a Basic Local Alignment Search Tool (BLAST) analysis in zebrafish genome. Only one mymk gene was identified, which was located on chromosome 5. Two other members of the TEMEM8 family, TEME8A and TEME8B, which share partial homology with the mymk protein sequence, were also identified in the zebrafish genome. The mymk gene in zebrafish is closely linked with the adamts12 and the slc2a6/8 genes at the 5′ and 3′ regions, respectively. Comparative genomic analysis revealed a conserved syntony among these genetic loci in zebrafish, mice and human genomes.
Protein sequence alignment of mymk protein and transmembrane domain prediction.
(A) Zebrafish mymk protein shares 75 and 73% sequence identities with mymk proteins in human and mice, respectively. Sequences of identical and similar amino acids are indicated by (*) and (:), respectively. (B) Predictions of transmembrane domains in zebrafish mymk protein using the TMHMM Server v. 2.0 (http://www.cbs.dtu.dk/services/TMHMM-2.0/). The probability of each amino acid to be in the transmembrane region, inside or outside of the cell membrane, is indicated on the Y axis.
sgRNA design and indel mutations generated by sgRNA/Cas9 in the zebrafish mymk gene.
(A) Two sgRNAs (sgRNA-1 and sgRNA-2) were designed to target the exon 3 sequence in mymk gene of zebrafish. sgRNA-1 or sgNA-2 was co-injected with Cas9 mRNA into WT zebrafish embryos at 1–2 cell stages. Several indel mutations were found at the sgRNA-1 or sgRNA-2 target sites. The bolded italic sequences represent the sgRNA-1 and sgRNA-2 target sites. (B) Three mutant alleles (mymkmb13/mb13, mymkmb14/mb14 and mymkmb15/mb15) were recovered containing a 7 or 8 bp deletion, or a 4 bp insertion, respectively. All three mutant alleles contained a reading frame shift mutation which resulted in the loss of mymk function. The bolded italic sequences represent the transmembrane domains. The underlined sequences represent the C-terminal sequence from reading frame shift mutations.
mymk mutation disrupts myoblast fusion in fast muscles.
Myoblast fusion was analyzed in WT and three mymk mutant alleles by DAPI and phalloidin staining at 48 hpf. Compared with WT control that contained mostly multinucleated fast myofibers (A), fast myofibers in homozygous mutant embryos contained one centrally located nucleus per fiber in the mymkmb13/mb13, mymkmb14/mb14 or mymkmb15/mb15 mutant alleles, respectively (B, C, D). Scale bar (30 μm).
Partial rescue of myoblast fusion by mymk mRNA injection.
Myoblast fusion was analyzed in WT or mymkmb14/mb14 mutant embryos injected with the pmCherry-NLS-C1 DNA plasmid alone (A, D) or together with mymk mRNA (B, C, E, F). Myofibers in mymkmb14/mb14 mutant embryo injected with pmCherry-NLS-C1 DNA alone contained only one centrally located nucleus per fiber at 48 hpf (A and D). In contrast, multinucleated myofibers (arrow) were detected in mymkmb14/mb14 mutant embryos co-injected with mymk mRNA and pmCherry-NLS-C1 DNA plasmid (B, E). Multiple nucleated myofibers observed in WT embryos co-injected with mymk mRNA and pmCherry-NLS-C1 DNA plasmid (C, F). Scale bar (30 μm).
mymk overexpression does not induce myoblast fusion in slow muscle fibers.
(A, B, D, E). Uninjected control (A, B) and mymk mRNA-injected (D, E) WT zebrafish embryos at 24 hpf. Myonuclei in slow muscles were labeled with anti-Prox-1antibody at 24 hpf (A, B and D, E). The slow fibers were counter stained with F59 monoclonal antibody that specifically labels slow myosin heavy chain (A and D). (C and F) The pmCherry-NLS-C1 plasmid was injected alone (C) or with mymk mRNA (F) into WT zebrafish embryos. The slow fibers were stained with F59 monoclonal antibody. pmCherry-NLS-C1 positive slow fibers contained only one nucleus per fiber in pmCherry-NLS-C1 injected (C) or mymk mRNA and pmCherry-NLS-C1 co-injected embryos (F) at 48 hpf.
Scale bar (30 μm).
Generation of mymk zebrafish mutants
To carry out a genetic analysis of mymk function in zebrafish, we generated three mutant alleles (mymkmb13, mymkmb14 and mymkmb15) in the mymk gene using the CRISPR technology. To ensure the success of mutagenesis by CRISPR, two guide RNAs (sgRNAs) were synthesized targeting to distinct sequences in exon 3 (Fig. 2A). The efficacy of these two sgRNAs in guiding Cas9-mediated mymk mutation was tested in the F0-injected embryos. Sequence analysis showed that both sgRNA-1 and sgRNA-2 were effective in guiding Cas9-mediated mutagenesis (Fig. 2A). One mutant allele (mymkmb13) was selected at the F1 generation that contained a 7 base pairs (bp) deletion within the sgRNA-1 target site. The 7 bp deletion occurred at the position 314–320 with respect to the ATG start codon. Two mutant alleles (mymkmb14 and mymkmb15) were identified at the sgRNA-2 target site. The mymkmb14 allele contained an 8 bp deletion (355–362), whereas the mymkmb15 had a 4 bp insertion (359–362), respectively (Fig. 2). The indel mutations of the 8 bp deletion or 4 bp insertion destroyed a HaeIII site (GGCC) at the sgRNA-2 target site (Fig. 2A), thus facilitating easy genotyping and mutant screening. All these three mutations led to reading frame shift and resulted in premature translational termination, likely producing nonfunctional truncated proteins (Fig. 2B).
mymk mutant embryos show defective myoblast fusion
Myoblast fusion was analyzed in the F2 and F3 homozygous mutant zebrafish embryos of the mymkmb13, mymkmb14 or mymkmb15 alleles by DAPI and phalloidin staining. In contrast to myofibers from WT embryos that contain multiple myonuclei per fiber (Fig. 3A), myofibers from all three mymkmb13, mymkmb14 and mymkmb15 mutant alleles had only one nucleus per fiber at the center of the segmented myotomes (Fig. 3B–D).
To determine whether the myoblast fusion defect could be rescued by mymk mRNA injection, DNA constructs (pmCherry-NLS-C1) expressing a nuclear mCherry were injected either alone or together with mymk mRNA into mymkmb14/mb14 mutant zebrafish embryos at 1–2 cell stages. As shown in Figure 4, mymkmb14/mb14 mutant embryos injected with pmCherry-NLS-C1 DNA construct alone only contained single-nucleated myofibers (Fig. 4A and D). DAPI staining showed a central nuclear localization in the myotome (Fig. 4D). In contrast, co-injection with mymk mRNA resulted in pmCherry-NLS-C1 positive fibers with both single and multiple nuclei in the mymkmb14/mb14 mutant embryos (Fig. 4B and E). Moreover, mymk mRNA-injected embryos displayed myonuclei localization in a broader pattern within the myotome (Fig. 4E). However, myoblast fusion did not appear to be fully rescued in the mymkmb14/mb14 mutant by mymk mRNA injection, compared with the WT control embryos co-injected with mymk mRNA and pmCherry-NLS-C1 construct (Fig. 4C and F). Together, these data indicate that microinjection of mymk mRNA might partially rescue the fusion defect in early stage mymkmb14/mb14 mutant fish embryos.
To determine whether mymk mRNA injection was able to induce myoblast fusion in slow myofibers that normally do not fuse in zebrafish embryos, the mymk mRNA-injected embryos were stained with anti-prox-1 antibody, which labeled myonuclei specifically in slow myofibers. As shown in Figure 5, no multinucleated slow myofibers were detected in the mymk mRNA-injected embryos (Fig. 5D and E), similar to the un-injected control (Fig. 5A and B). In addition, we performed pmCherry-NLS-C1 DNA injection into zebrafish embryos either alone or with mymk mRNA. Slow fibers expressing nuclear mCherry were identified by F59 antibody staining, which labeled myosin heavy chain specifically expressed in slow fibers. As shown in Figure 5F, only single-nucleated pmCherry-NLS-C1 positive slow fibers were found in the mymk mRNA-injected embryos. This pattern was similar to that observed in the control embryos (Fig. 5C). Together, these data indicate that although mymk mRNA injection was able to partially rescue myoblast fusion in fast myofibers, it was not sufficient to induce myoblast fusion in slow fibers of zebrafish embryos.
mymk mutation does not affect slow and fast muscle specification and differentiation
To determine whether blocking myoblast fusion in fast muscles affects the fiber-type specification and differentiation, we analyzed the expression of myosin heavy chain genes (slow myosin heavy chain or smhc and fast myosin heavy chain or fmhc) which were specifically expressed in respective slow or fast muscles by whole mount in situ hybridization. The data showed that mymk mutation had no effect on smhc and fmhc gene expression, suggesting that blocking myoblast fusion did not affect fiber-type specification during early myogenesis (Fig. 6A–H). To further determine whether blocking myoblast fusion in fast muscles affected their differentiation, we performed immunostaining with antibodies against myosin light chain and Myomesin expressed specifically in fast fibers. The results revealed no notable difference between mymkmb14/mb14 mutant and WT control at 72 hpf (Fig. 6I–L). Collectively, these data demonstrate that blocking myoblast fusion in fast muscles does not affect their specification and differentiation.
mymk mutation does not affect slow and fast muscle specification and differentiation.
(A–H) Whole mount in situ hybridization shows the expression of smhc 1 mRNA in WT (A, B) and mymkmb14/mb14 mutant (C, D) embryos at 24 hpf. fmhc mRNA expression in WT (E, F) and mymkmb14/mb14 mutant (G, H) embryos at 24 hpf. Scale bar (120 μm). (I–L) Immunostaining with anti-myosin light chain antibody (F310) or anti-myomesin antibody (mMaC myomesin B4) that specifically label fast muscle fibers in mymkmb14/mb14 mutant (I, J) or WT (K, L) zebrafish embryos at 72 hpf. Scale bar (30 μm).
mymk mutant shows reduced growth and increased adipocyte infiltration in skeletal muscle
It has been reported that knockout of mymk in mice resulted in postnatal lethality at day 1–7 after birth (9). To determine whether knockout of mymk in zebrafish affected zebrafish survival, we followed the fish survival rate generated from heterozygous mutant in cross for 3 months. The data showed that unlike mice, the zebrafish homozygous mutants were able to survive normally within 1 month after fertilization. A typical Mendelian inheritance pattern was observed at 7, 14, 21 and 28 dpf. However, morphological differences became apparent 1 month after fertilization. The homozygous mutants appeared skinnier and darker compared with wild type (WT) and heterozygous mutant, starting around 30–40 dpf, and some of the homozygous mutant larvae started to die. By 90 dpf, only 9 homozygous mutants were identified from 87 fish generated from in-cross of mymkmb14/+ heterozygous mutants. This is below the expected 25% from the pattern of Mendelian inheritance. The number of heterozygous mutant (53) and WT (25) fish were significantly increased. However, their ration (2.12:1) is close to the expected ration of (2:1) from the Mendelian inheritance. The reduced survival rate was found in all three mymkmb13, mymkmb14 and mymkmb15 mutant alleles (data not shown). Collectively, these data indicate that the homozygous mutants were subject to partial lethality during larval growth.
The survived mymk mutant adults showed reduced growth (Fig. 7A). Compared to WT and heterozygous mutants, homozygous mutants were smaller in size and weighed about one-third the weight of WT sibling when analyzed at 3 months old (Fig. 7B and C). In addition, a characteristic craniofacial deformity was observed in the homozygous mutant with an open lower jaw (Fig. 7B). Histological analysis showed that the skeletal muscles were severely malformed in the homozygous mutants at 3 months old. In contrast to WT control, in which skeletal myofibers were highly organized and packed tightly within the muscle mass (Fig. 7D and E), myofibers were smaller and not well organized in the mymk mutants (Fig. 7F and G). Moreover, mutant fish contained a large number of adipocytes imbedded in the skeletal muscles (Fig. F and G). Collectively, these data indicate that mymk is essential for normal skeletal muscle growth in fish, and malformation of skeletal muscle results in increased adipocyte infiltration in skeletal muscles of mymk mutant.
mymk mutant showed reduced growth, craniofacial deformity, disorganized skeletal muscles and increased adipocyte infiltration.
(A–C) Photographs of WT and mymkmb14/mb14 mutants at 90 dpf (A). The mymk mutants exhibited craniofacial deformities indicated by an arrow (B). Weight comparison among WT, heterozygous and homozygous mutant sibling at 90 dpf (C). (D–G) HE staining showing cross (D, F) and sagittal (E, G) sections of trunk skeletal muscles in WT control (D, E) and mymkmb14/mb14 mutant fish (F and G) at 90 dpf. Noticeable adipocyte infiltration was observed in mutant skeletal muscles (F, G). Scale bar (100 μm).
Myofibers in adult mymk mutant contained fewer nuclei
To determine whether or not mymk mutation had a continuing effect on myoblast fusion, we characterized myofibers in mymkmb14/mb14 adult mutant. Individual myofibers were dissected from skeletal muscles of WT and homozygous mutants at 7 months old. The myofibers were stained with DAPI and counter stained with Phalloidin-TRITC. The results showed that in contrast to myofibers from WT fish that contained a large number of nuclei per myofiber (Fig. 8A), myofibers from the mymkmb14/mb14 mutant were smaller in fiber diameter and length (Fig. 8B and C). In addition, the mymk mutant fibers had significantly fewer number of myonuclei compared with fibers from the WT control (Fig. 8D). However, a similar range of nucleo-cytoplasmic ratios were found for WT and mutant myofibers, suggesting that the difference in the number of myonuclei in WT and mutant might contribute to the smaller diameter of mutant myofibers. Together, these data indicate that mymk mutation had a perpetual effect on blocking myoblast fusion and fiber growth from embryos to adult.
Single fiber analysis of fiber size and nuclear numbers in skeletal muscles from WT and mymk mutant.
Single fibers were isolated from skeletal muscles of WT adult fish (A) and mymkmb14/mb14 mutant (B, C) at 7 months old. The isolated fibers were stained with DAPI and phalloidin-TRIC, and photographed by confocal microcopy. Representative of myofibers from mymkmb14/mb14 mutant contained a single nucleus (B) or multiple nuclei (C) per fiber. (D). Statistical analysis of nuclear numbers in myofibers from WT and mymk mutant fish. Scale bar (50 μm).
The effects of Shh mRNA injection on mymk expression and myoblast fusion.
(A–F)In vitro-transcribed Shh mRNA was injected into fertilized zebrafish eggs at 1–2 cell stage. Whole mount in situ hybridization shows the expression of mymk, fast myosin heavy chain (fmyhc) or slow myosin heavy chain (smyhc) in WT control (A, C, E), or Shh mRNA-injected (B, D, F) embryos at 24 hpf. Scale bar (120 μm). (G–H)In vitro-transcribed Shh mRNA was injected alone into fertilized zebrafish eggs at 1–2 cell stage. Myoblast fusion was analyzed by DAPI and phalloidin staining in WT control (G), or Shh (H) injected embryos at 48 hpf. Scale bar (30 μm). (I–J) pmCherry-NLS-C1 DNA was injected alone (I) or together with Shh mRNA (J) into fertilized WT zebrafish embryos. Nuclear mCherry expression was observed in myofibers from control (I), or Shh mRNA (J) injected embryos at 48 hpf. Scale bar (30 μm).
Overexpression of Hedgehog (Hh) inhibited mymk expression and myoblast fusion
Previous studies demonstrated that mymk is expressed specifically in fast muscle precursors that are able to fuse to generate multinucleated fast myofibers (16). It is well known that overexpression of Hh in zebrafish embryos results in the conversion of fast muscle myoblasts into slow muscles (22,23). To determine whether ectopic expression of Shh inhibits mymk expression in zebrafish embryos, Shh mRNA was microinjected into zebrafish embryos at 1–2 cell stages. mymk expression was analyzed by in situ hybridization at 22–24 hpf. The data showed that Shh mRNA injection significantly reduced mymk mRNA expression in developing smites and skeletal muscles of zebrafish embryos (Fig. 9B). To confirm that ectopic expression of Shh indeed converted fast muscles into slow muscles, we analyzed the expression of fiber-type specific myosin heavy chain isoforms in slow or fast muscles. The data showed ectopic expression of Shh blocked the expression of fast muscle specific myosin (Fig. 9D) and expanded the expression domain of slow muscle heavy chain into the whole myotome region (Fig. 9F).
To determine whether overexpression of Shh in zebrafish embryos could inhibit myoblast fusion, we analyzed the myonuclei in Shh-injected embryos at 48 hpf by DAPI and phalloidin staining. Unlike the un-injected control (Fig. 9G), Shh mRNA-injected embryos contained myofibers with a single central nucleus (Fig. 9H). To further validate this finding, DNA construct (pmCherry-NLS-C1) expressing a nuclear mCherry was injected either alone or together with Shh mRNA into WT zebrafish embryos at 1–2 cells stages. The nuclear mCherry could be quantified in myofibers of the injected embryos. The data showed that 90% (n = 50) of the mCherry positive myofibers contained 2–6 myonuclei in the control embryos (Fig. 9I). In contrast, more than 80% (n = 30) of the mCherry positive myofibers in the Shh mRNA-injected embryos contained only one nucleus per fiber (Fig. 9J). Collectively, these studies indicate that converting fast muscles into slow muscles by Shh inhibits mymk expression and blocks myoblast fusion.
Discussion
In this study, we analyzed the molecular regulation of myoblast fusion in zebrafish skeletal muscles. Our studies demonstrated that mymk is essential for myoblast fusion and fish skeletal muscle growth. Knockout of mymk gene in zebrafish resulted in defective myoblast fusion, poor muscle growth, partial lethality and craniofacial deformity. The viable homozygous mutants were smaller in size and contained poorly organized small myofibers with extensive intramuscular adipocyte infiltration in skeletal muscles. Single fiber analysis showed that most myofibers in adult mymk mutant were single-nucleated fibers. However, multiple-nucleated myofibers were detected in mutant fish. Ectopic expression of Shh inhibited mymk expression in developing somites and resulted in significant reduction of myoblast fusion. Our data are mostly in agreement with the previous findings by Zhang and Roy (11) and Di Gioia et al. (18), namely the defective myoblast fusion in fast muscles and the poor muscle growth and increased adipocyte infiltration in skeletal muscles. Furthermore, our studies also extended the work of Zhang and Roy (11) and Di Gioia et al. (18), specifically, the characterization of myonuclear numbers in adult fibers and the effect of Hh on mymk expression.
mymk expression and myoblast fusion
In this study, we generated three novel mymk mutant alleles in zebrafish. These three new mutant alleles contained reading frame shift mutations, leading to premature translation termination. By characterizing these mutants, we demonstrated that mymk is essential for myoblast fusion in fast muscles of zebrafish embryos. Our data are consistent with previous findings in other zebrafish mymk mutant alleles (11,18).
It is well established that zebrafish embryonic slow fibers are single-nucleated fibers, whereas the fast fibers are multinucleated formed by myoblast fusion. The molecular mechanism behind the lack of myoblast fusion in slow muscles is not clear. It has been suggested that distinct patterns of gene expression involved in myoblast fusion may dictate the different fusion characteristics of slow and fast muscles. Previous studies showed that two key regulators in myoblast fusion, Kirrel and Jam B (4,8), are exclusively expressed in fusion competent fast muscle precursors in zebrafish embryos.
It is uncertain whether mymk has a different pattern of expression in fast and slow muscles, underlying the different fusion capability of slow and fast muscles in zebrafish. Landemaine and colleagues (16) reported that mymk was specifically expressed in fast muscles. However, Shi and colleagues (14) showed that mymk transcripts could be deleted in adaxial cells, the slow muscle precursors. Our studies demonstrated that ectopic expression of Shh, which converted fast to slow muscles, downregulated mymk expression in myotome. This was accompanied by decreased myoblast fusion. Collectively, these studies suggested that mymk may express differently in slow and fast muscles during myogenesis in zebrafish.
Low or lack of mymk expression may preclude myoblast fusion in slow muscles of zebrafish embryos (16). It has been demonstrated that overexpression of mymk by DNA injection was sufficient to drive the myoblast fusion in slow myocytes of zebrafish embryos (11). However, the fusion efficiency was very low. Only 5.8% of the mymk positive slow fibers underwent fusion and differentiated as binucleated syncytia (11). The majority of the mymk expressing slow fibers failed to fuse. Consistent with the idea that it might be more demanding to induce non-fusogenic slow muscle precursors to fuse, we found that microinjection of mymk mRNA did not induce myoblast fusion in slow muscles of WT or mutant embryos although myoblast fusion was detected in fast muscles of the injected mymk mutant embryos.
It is possible that other factors expressed in fusogenic fast muscles may work in concert with mymk to promote myoblast fusion. Recent studies revealed that myomixer, also called minion and myomerger, plays an important role in myoblast fusion in mouse and zebrafish embryos (12–15). Myomixer is expressed specifically in fast muscle precursors during myoblast differentiation in zebrafish embryos (14). Myomixer has been suggested to serve as a membrane anchor of mymk protein (12), consistent with finding of myoblast fusion requires mymk activity at the plasma membrane (24). It is tempting to suggest that mymk may require Myomixer to function efficiently in fast muscles. Without Myomixer, mymk function is compromised. However, with high levels of mymk ectopic expression in slow muscle, mymk may induce myoblast fusion in slow muscles at a low efficiency.
Defective skeletal muscle growth and organization in mymk mutants
In mice, genetic disruption of mymk resulted in perinatal death (9). No live mymk mutant mice could be detected at postnatal day 7 although mymk mutant mice were observed at a normal Mendelian ratio at E15 and E17.5 (9). It has been suggested that the perinatal lethality was due to muscle dysfunction, likely caused by the absence of multinucleated muscle fibers in breathing muscles (9). Mice mutants with conditional mymk knockout in adult muscle stem cells (satellite cells) could survive, but showed compromised muscle regeneration after injury (10). We showed here that zebrafish mymk mutants were partially lethal at the juvenile stage around 35–45 days old. Approximately 40% of the expected mymk zebrafish mutant survived to adults when analyzed at 3 months old. The mutant fish were smaller in size and weighed approximately one-third the weight of their WT siblings.
We showed in this study that mymk mutants contained poorly organized skeletal muscles of smaller size and a fewer number of myofibers. Given that fish muscle growth involves several phases of myogenesis (25), it is possible that mymk is also required in post embryonic muscle growth. consistent with data from conditional knockout studies in mice demonstrating that mymk is reactivated in adult mice after muscle injury and mymk function is required for muscle regeneration and growth (10). This is not surprising given that muscle regeneration and growth also involve myoblast fusion. Supporting this idea, our studies of single fiber analysis showed that myofibers in adult mymk mutant were predominantly single-nucleated fibers. The fibers in mymk mutant fish appeared to be 4 to 10 times smaller in diameter than the fibers in WT animal. It is well established that a change in muscle fiber size was accompanied by a similar change in nuclear number in aging skeletal muscle (26–28). We compared the number of nuclei per volume cytoplasm in WT and mutant fibers. A similar range of nucleo-cytoplasmic ratio was found for WT and mutant myofibers. Collectively, these data indicate that the difference in the number of nuclei between WT and mutant might explain the smaller diameter of mutant myofibers.
Our histological analysis revealed that mymk zebrafish mutants contained an extensive number of adipocytes in their skeletal muscles. This is in agreement with a recent study reporting that the mymk mutant fish have reduced body width and increased fat infiltration (18). This characteristic muscle defect referred as intermuscular adipose tissue (IMAT) accumulation has been observed in variety of myopathies, including Duchenne muscular dystrophy, mitochondrial myopathy and neurogenic myotrophy (29). Interestingly, CFZS patients with mymk mutations showed myopathic changes in their skeletal muscles. They contained disproportionately high numbers of atrophic/hypotrophic Type I fibers and hypertrophic Type II fibers (18). In addition, magnetic resonance imaging (MRI) analysis revealed fatty infiltration in their adductor magnus and Sartorius muscles (18). The fatty infiltration was more pronounced in the biopsy from the older CFZS patient compared to the younger patients (18). Consistent with this finding, we found that the adipocyte infiltration was less dramatic in younger fish mutant (data not shown). Together, these data indicate that the increase in fatty infiltration likely correlates with the slow progression of muscle defects.
The mechanisms underlying the increased adipocyte infiltration in mymk mutant muscles is not clear. Fibrous-adipose replacement is an unspecific change observed in skeletal muscle biopsy whenever the muscular component is reduced due to pathological processes. It has been shown that adipocytes accumulate and replace a large proportion of muscle fibers in patients with muscular dystrophies. The IMAT accumulation has also been detected in Type II diabetes, aged or denervated muscles (30,31). Increased IMAT was also found in chronic disuse-induced muscle atrophy and even healthy young adults after only 1 month of immobilization (32).
Primers for cDNA cloning, sgRNA synthesis and PCR analyses
| Primers | Sequences |
|---|---|
| zfMyomaker-P1 | ATGGGAGCGTTTATCGCCAAGATG |
| zfMyomaker-P2 | TGCATCATACACAGCAGCAGAGGG |
| myomaker-sgRNA-1-F1 | 5′-GATCACTAATACGACTCACTATAGGGTGTTGACCGCAGCTGTGGTTTTAGAGCTAGAAAT-3′ |
| myomaker-sgRNA-2-F1 | 5′-GATCACTAATACGACTCACTATAGGCATTTACTCCGGCCCCATGTTTTAGAGCTAGAAAT-3′ |
| P4 | 5′-AAAAGCACCGACTCGGTGCC-3′ |
| mymkE3-F2 | 5′- CACGAGCATGTTGACAGGTGTTTAC-3′ |
| mymkE3-R2 | 5′-CCTCCAGGACAACTCTACCTTAAC -3′ |
| Primers | Sequences |
|---|---|
| zfMyomaker-P1 | ATGGGAGCGTTTATCGCCAAGATG |
| zfMyomaker-P2 | TGCATCATACACAGCAGCAGAGGG |
| myomaker-sgRNA-1-F1 | 5′-GATCACTAATACGACTCACTATAGGGTGTTGACCGCAGCTGTGGTTTTAGAGCTAGAAAT-3′ |
| myomaker-sgRNA-2-F1 | 5′-GATCACTAATACGACTCACTATAGGCATTTACTCCGGCCCCATGTTTTAGAGCTAGAAAT-3′ |
| P4 | 5′-AAAAGCACCGACTCGGTGCC-3′ |
| mymkE3-F2 | 5′- CACGAGCATGTTGACAGGTGTTTAC-3′ |
| mymkE3-R2 | 5′-CCTCCAGGACAACTCTACCTTAAC -3′ |
The underlined sequences in myomaker-sgRNA-1-F1 and myomaker-sgRNA-2-F1 primers represent specific target sequences in the zebrafish mymk gene.
Primers for cDNA cloning, sgRNA synthesis and PCR analyses
| Primers | Sequences |
|---|---|
| zfMyomaker-P1 | ATGGGAGCGTTTATCGCCAAGATG |
| zfMyomaker-P2 | TGCATCATACACAGCAGCAGAGGG |
| myomaker-sgRNA-1-F1 | 5′-GATCACTAATACGACTCACTATAGGGTGTTGACCGCAGCTGTGGTTTTAGAGCTAGAAAT-3′ |
| myomaker-sgRNA-2-F1 | 5′-GATCACTAATACGACTCACTATAGGCATTTACTCCGGCCCCATGTTTTAGAGCTAGAAAT-3′ |
| P4 | 5′-AAAAGCACCGACTCGGTGCC-3′ |
| mymkE3-F2 | 5′- CACGAGCATGTTGACAGGTGTTTAC-3′ |
| mymkE3-R2 | 5′-CCTCCAGGACAACTCTACCTTAAC -3′ |
| Primers | Sequences |
|---|---|
| zfMyomaker-P1 | ATGGGAGCGTTTATCGCCAAGATG |
| zfMyomaker-P2 | TGCATCATACACAGCAGCAGAGGG |
| myomaker-sgRNA-1-F1 | 5′-GATCACTAATACGACTCACTATAGGGTGTTGACCGCAGCTGTGGTTTTAGAGCTAGAAAT-3′ |
| myomaker-sgRNA-2-F1 | 5′-GATCACTAATACGACTCACTATAGGCATTTACTCCGGCCCCATGTTTTAGAGCTAGAAAT-3′ |
| P4 | 5′-AAAAGCACCGACTCGGTGCC-3′ |
| mymkE3-F2 | 5′- CACGAGCATGTTGACAGGTGTTTAC-3′ |
| mymkE3-R2 | 5′-CCTCCAGGACAACTCTACCTTAAC -3′ |
The underlined sequences in myomaker-sgRNA-1-F1 and myomaker-sgRNA-2-F1 primers represent specific target sequences in the zebrafish mymk gene.
mymk mutant showed craniofacial defects
We demonstrated that mymk mutant fish exhibited craniofacial deformities. This craniofacial phenotype resembles that observed in human populations with the CFZS carrying mymk genetic mutations. The CFZS is characterized by facial weakness, upturned/broad nasal tip, micro/retrognathia and small/retracted lower jaw (18). The molecular and cellular mechanisms underlying the craniofacial deformities are not clear. It is known that muscle can influence bone development. Congenital myopathies with facial weakness often appeared with a secondary disorder of retrognathia, palatal defects and facial dysmorphisms (20,33). Genetic studies also demonstrate the critical role of early muscle contraction and static loading by striated muscle in skeletogenesis in mice. Blocking jaw and tongue muscle development by knockout of Myf5 and MyoD also resulted in secondary retrognathia and cleft palate in mice (34). It will be interesting to determine if mymk is expressed in craniofacial muscles, and whether the craniofacial defects could result from a secondary effect from defective jaw muscle development.
Materials and Methods
Zebrafish lines and maintenance
Zebrafish were maintained at the zebrafish facility of the Aquaculture Research Center, Institute of Marine and Environmental Technology, University of Maryland at Baltimore. The fish were kept in 1 and 8-gallon aquaria at 28°C with a photoperiod of 14 h light and 10 h dark. All animal studies were carried out according to the guideline for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Institutional Animal Care and Use Committee of the University of Maryland (Permit Number: 0516005).
Isolation of zebrafish mymk cDNA and generation of mymk expression construct
The full length mymk cDNA was cloned from zebrafish by reverse transcription polymerase chain reaction (RT-PCR). Total RNA was isolated from zebrafish embryos at 24 hpf using the Trizol reagents (Invitrogen). The PCR was carried out using zfMyomaker-P1 and zfMyomaker-P2 primers (Table 1) and the high-fidelity Phusion DNA polymerase (F530S, ThermoFisher Scientific). The DNA product was treated with Taq DNA polymerase to add an extra nucleotide A at the 3′ end and followed by cloning into the pGEM-T easy vector (Promega) to generate the pGEM-T-zfmyomaker plasmid construct.
To generate an expression construct for making mymk mRNA for microinjection, the full-length coding sequence of mymk was released from the pGEM-T-zfmyomaker plasmid by EcoRI digestion and re-subcloned into the EcoRI site of CS2+ expression vector. The CS2+-zfmyomaker construct was confirmed by DNA sequencing and used as template for in vitro mRNA transcription.
sgRNAs design and synthesis
Two sgRNAs were synthesized targeting to two distinct sites in the exon 3 of zebrafish mymk gene. The first target site was named mymk exon-3-1 (5′-GGGTGTTGACCGCAGCTGTGAGG-3′). The second target site was designated as mymk exon-3-2 (5′-GGCATTTACTCCGGCCCCATCGG-3′). The first 2 bp GG sequences are required for correct T7 transcription. The PAM sequences (NGG) are underlined. The second target site contains a unique HaeIII restriction enzyme cutting site (GGCC).
For sgRNA synthesis, a DNA template was generated by PCR using the DR274 plasmid as template and mymk specific forward primers (myomaker-sgRNA-1-F1 or myomaker-sgRNA-2-F1) together with the reverse primer P4 (Table 1). The gene-specific forward primers were 60 nucleotide long, including a common protective base (5′-GATCAC-3′), the T7 promoter sequence (5′-TAATACGACTCACTATA-3′), the 20 gene-specific target sequence without the PAM and the pDR274 5′ sequence (5′-GTTTTAGAGCTAGAAAT-3′). The common reverse primer (P4, 5′-AAAAGCACCGACTCGGTGCC-3′) was based the complementary strand at the 3′ end of the DR274 vector (Table 1).
PCR was carried out using DR274 plasmid as a template purchased from Addgene and Phusion DNA Polymerase (F530S, ThermoFisher Scientific) for 35 cycles. The PCR product was purified using the gel extraction kit (Qiagen). sgRNA synthesis was carried out using the MAXIscript T7 kit (AM1312, ThermoFisher Scientific) for 4 h containing 0.3 μg purified DNA as template. The sgRNA products were purified using the MEGAclear Transcription Clean-Up Kit (AM1908, ThermoFisher Scientific).
Cas9, mymk and Shh mRNA synthesis by in vitro transcription
Cas9 mRNA was synthesized in vitro using XbaI linearized pT3TS-nCas9n plasmid as the template. The pT3TS-nCas9n plasmid was purchased from Addgene (35). The RNA synthesis was carried out using the mMESSAGE mMACHINE® T3 Transcription Kit (AM1348, ThermoFisher Scientific). Cas9 mRNA was purified using the MEGAclear™ Transcription Clean-Up Kit (AM1908, ThermoFisher Scientific).
Zebrafish mymk mRNA was synthesized in vitro using NotI linearized CS2+-zfmyomaker construct as DNA template. The plasmid was linearized with NotI and followed by purification with QIAquick Gel Extraction Kit (#28706, QIAGEN). Capped mymk mRNA was synthesized by in vitro transcription using the mMESSAGE mMACHINE Sp6 kit (AM1340, ThermoFisher Scientific). The mRNA was purified using MEGAclear Transcription Clean-Up kit (AM1908, ThermoFisher Scientific).
The Shh mRNA was synthesized in vitro using NotI linearized T7TS shh plasmid as described (36). Capped mRNAs were transcribed from linearized DNA template using the mMESSAGE mMACHINE T7 in vitro transcription kit (AM1344, ThermoFisher Scientific) according to the manufacturer’s instructions. The mRNA was purified using MEGAclear Transcription Clean-Up kit (AM1908, ThermoFisher Scientific).
mRNA and DNA microinjection
Cas9 mRNA (∼400 ng/μl) was mixed with equal volume of mymk sgRNA-1 (∼50 ng/μl) or sgRNA-2 (∼50 ng/μl), respectively. Approximately, 2 nl of the mixed solution containing 0.1% phenol red was microinjected into each fertilized egg. 200–300 embryos were injected for each sgRNA. For mymk rescue assay and overexpression experiment, approximately 1 ng of mymk mRNA (2.5 nl of 400 ng/μl) was microinjected into each mutant or WT zebrafish embryo at 1–2 cell stages. For overexpression of Shh mRNA, approximately 0.25 ng of Shh mRNA (2.5 nl of 100 ng/μl) was microinjected into each zebrafish embryos at 1–2 cell stages. For pmCherry-NLS-C1 DNA injection, the DNA plasmid was dissolved in water at 50 ng/μl, and approximately 2.5 nl of the DNA solution was injected into each embryo. For co-injection with capped mymk mRNA or Shh mRNA, the pmCherry-NLS-C1 DNA was mixed with mymk or Shh mRNA. The final concentration of pmCherry-NLS-C1plasmid DNA was 50 ng/μl, and the final concentration of mymk or Shh mRNA was 350 ng/μl and 100 ng/μl, respectively. Approximately 2.5 nl of the mRNA and DNA mixture solution was microinjected into each embryo.
DNA isolation from fish embryos and caudal fin clips for genotyping
For genotyping of fish embryos or adult fish, genomic DNA was isolated from individual zebrafish embryos at 48 hpf or caudal fin clip of adult fish using the alkaline method. Briefly, each embryo was lysed with 20 μλ of 50 mM NaOH by incubation at 95°C for 10 min. During the 10 min incubation, the samples were mixed gently by tapping the tube every 2–3 min. By the end of 10 min, the samples were cooled on ice for 3 min, and 2 μl of 1 M Tris–HCl (pH. 8) was added to each tube. The sample was mixed gently and then centrifuged at 10 000 RPM for 5 min. 1 μl of the supernatant solution was used for each PCR reaction. For genotyping with fin clip DNA, each small fin clip was lysed in 50 μl of 50 mM NaOH. The lysate was gently vortexed and neutralized with 5 μl of 1 M Tris–HCL (pH 8.0). The crude lysate was centrifuged for 4 min at 12 K and 1 μl of the DNA lysate was used for each PCR reaction.
For genotyping of fish embryos after immunostaining or in situ hybridization, the head region was dissected from each embryos after the experiment. DNA was isolated from the head region using the alkaline method described above. For genotyping with a pool of 20 embryos, the embryos were lysed in 200 μl of 50 mM NaOH for 10 min at 95°C, and then neutralized with 20 μl of 1 M Tris–HCl (pH 8). The crude lysate was centrifuged for 4 min at 12 K and 1 μl of the DNA lysate was used for each PCR reaction.
The PCR reaction was carried out in a 25 μl reaction using GoTaq® DNA polymerase (Progema). The myomaker-E3-F2 and myomaker-E3-R2 primer set was used (Table 1). The PCR product was treated with ExoSAP-IT (USB, #78200) and used for DNA sequencing. For genotyping by HaeIII digestion, 10 μl of the PCR product was digested with HaeIII in a 20 μl reaction, and followed by electrophoresis on a 2% agarose gel.
Whole-mount immunostaining and nuclear staining by Hoechst 32258
Immunostaining was carried out using whole-mount zebrafish embryos as previously described (37). Embryos were anesthetized in a 0.02% MS222 and fixed in 4% paraformaldehyde for 1 h at room temperature. The fixed embryos were washed with PBS containing 0.1% Tween-20 (PBST) three times (10 min each). To increase the antibody penetration, the embryos were digested in 1 mg/ml collagenase for 7, 45 and 75 min, for 28, 48 and 72 hpf embryos, respectively. Immunostaining was performed with anti-MyHC (F59), anti-MyLC (F310) or anti-myomesin (mMaC myomesin B4) primary monoclonal antibodies (DSHB, USA) or anti-Prox1 antibody (AB5475, Millipore, USA) overnight at 4°C. After washing with PBST, the embryos were incubated with respective Alexa Fluor® 488- or Alexa Fluor 555-conjugated anti-IgG secondary antibodies (A11001 and A31630, ThermoFisher Scientific) for 1 h at room temperature in the dark. For nuclear staining, the embryos were incubated with 1 μg/ml Hoechst 32258 (Sigma B2883) for 1 h at room temperature in the dark. For double staining with Hoechst and phalloidin, the fixed embryos were incubated with 1 μg/ml Hoechst 32258 (Sigma B2883) and 50 ng/ml Phalloidin-TRITC (Sigma P1951) for 1 h at room temperature in the dark. After washed with PBST three times (30 min each), the trunk region of fish embryos was dissected and mounted in Vectashied (Vector lab, H-1000). The embryos were photographed using a Leica SP8 confocal microscope.
Whole mount in situ hybridization
Whole mount in situ hybridization was carried out using digoxigenin-labeled antisense probes as previously described (38). The pGEM-T-zfmyomaker plasmid was linearized with ApaI, whereas pGEM-T-JamB and pGEM-T-JamC were linearized with NcoI. The linearized DNA was purified using the gel extraction kit (Qiagen). The purified DNAs were used as template for in vitro transcription with Sp6 RNA polymerase to synthesize digoxigenin-labeled antisense RNA probes. The antisense digoxigenin-labeled probes were purified using the Megaclear kit (Invitrogen).
HE staining of the adult zebrafish muscle
Three-month-old adult zebrafish were euthanized and fixed in 4% paraformaldehyde for 48 h at 4°C. The fixed samples were washed with pure water for 4 h at room temperature. After dehydration in a series of graded ethanol and cleared by xylene, samples were embedded with paraffin. Paraffin sectioning and hematoxylin & eosin (HE) staining were conducted based on standard procedures and photographed under a light microscope.
Isolation of single muscle fibers and immunostaining from adult zebrafish
Seven-month-old adult zebrafish were euthanized in MS-222 (0.03%) and then fixed in 4% paraformaldehyde for 20 min at room temperature. The skin was removed in 4% paraformaldehyde and the deskinned fish were fixed for another 40 min. Muscle bundles were dissected after the fixation and placed in 0.04% saponin. Muscle bundles were dissected after the fixation and placed in 0.04% saponin. Single fibers were then gently pulled apart from muscle bundles with tweezers under a dissecting microscope. Once isolated, single muscle fibers were put on a slide and immunofluorescence staining was conducted with 1 μg/ml Hoechst 32258 (Sigma B2883) for nuclei and 50 ng/ml Phalloidin-TRITC (Sigma P1951) for myofibril a-actin filaments. After 20 min staining, single fibers were washed three times with PBST and then mounted in Vectashied (Vector lab, H-1000). Ten individual fibers were analyzed from the WT or mutant fish by confocal microscope to determine the number of myonuclei per myofiber.
Statistical analysis
Statistical analysis in nuclear number comparison between WT and mutant myofibers was performed with one-way ANOVA followed by Duncan’s comparison tests to determine whether there is a significant difference between WT and mymk mutant using SPSS 19.0. Statistical analysis in weight comparison was done by one-way ANOVA followed by LSD test to determine significant differences between different genotypes using SPSS 19.0. Statistical significance was determined at P-value <0.01. Values represent mean ± SEM. ∗∗∗P-value < 0.01.
Acknowledgements
We thank Nick Du for proofreading the manuscript.
Conflict of Interest statement. None declared.
Funding
Seed funding from University of Maryland (to S.D.); National Natural Science Foundation of China (No. 31230076); Chinese Scholarship Council (to M.C. and Y.S.).
References
Author notes
The first three authors contributed equally to this project.
