Role of Zebrafish fhl1A in Satellite Cell and Skeletal Muscle Development.

BACKGROUND: Four-and-a-half LIM domains protein 1 (FHL1) mutations are associated with human myopathies. However, the function of this protein in skeletal development remains unclear.
METHODS: Whole-mount in situ hybridization and embryo immunostaining were performed.
RESULTS: Zebrafish Fhl1A is the homologue of human FHL1. We showed that fhl1A knockdown causes defective skeletal muscle development, while injection with fhl1A mRNA largely recovered the muscle development in these fhl1A morphants. We also demonstrated that fhl1A knockdown decreases the number of satellite cells. This decrease in satellite cells and the emergence of skeletal muscle abnormalities were associated with alterations in the gene expression of myoD, pax7, mef2ca and skMLCK. We also demonstrated that fhl1A expression and retinoic acid (RA) signalling caused similar skeletal muscle development phenotypes. Moreover, when treated with exogenous RA, endogenous fhl1A expression in skeletal muscles was robust. When treated with DEAB, an RA signalling inhibitor which inhibits the activity of retinaldehyde dehydrogenase, fhl1A was downregulated.
CONCLUSION: fhl1A functions as an activator in regulating the number of satellite cells and in skeletal muscle development. The role of fhl1A in skeletal myogenesis is regulated by RA signaling. Copyright© Bentham Science Publishers; For any queries, please email at epub@benthamscience.org.

Introduction

Four-and-a-half LIM domain protein 1 (FHL1) is highly expressed in skeletal muscles, suggesting that the FHL1 gene is associated with the development of human muscle diseases. Indeed, FHL1 mutations have been identified in a number of human myopathies [1-3]. As all clinical subtypes with mutations in exons of FHL1 are associated with myofibril changes and diminished bodies, these muscular diseases may be investigated 
as resulting from these mutations [1, 2]. FHL1-related muscle diseases can be divided into two broad categories: 1) missense mutations in the LIM2 domain affecting the zinc-coordinating histidine and cysteine, and 2) missense mutations in the C-terminal domain-encoding region of FHL1 isoforms [4-10]. In mice, the LIM-only protein FHL1 is primarily expressed in skeletal and cardiac muscles [11, 12]. Overexpression of this protein enhances myoblast fusion, resulting in hypertrophic myotubes in C2C12 cells, as well as modulating muscle mass and enhancing strength via the regulation of NFATc1 signalling [13]. The transgenic expression of FHL1 protein promotes hypertrophy in mouse skeletal muscle and performs various functions in muscle cell lines [14, 15]. It is also reported that the loss of FHL1 function causes age-dependent myopathy [16]. Although the role of FHL1 in the development of human myopathies is becoming better understood, the pathological and molecular mechanisms of FHL1 in skeletal muscle development and disease have yet to be clarified.

Recently, zebrafish have been identified as a useful model of certain human diseases, including myopathies [17]. In zebrafish, axial skeletal muscles are made up of four types of cells: slow-twitch muscle cells, muscle pioneer cells, fast-twitch muscle cells and medial fast fibre cells [18]. These cells, marked by the expression of transcription factors such as myoD, initially form from the somites that are patterned into myotomes and sclerotomes [19]. The progenitors of slow-twitch muscle cells, including muscle pioneers, are located on the lateral surfaces of the somites, whereas the progenitors of fast-twitch muscle cells remain deeper within the somites [20]. In addition, the satellite cells, marked by their expression of transcription factors like Pax7, are also progenitor cells and can generate the myoblasts required for muscle growth, repair and regeneration [21]. By 28 hpf, slow-twitch and fast-twitch muscle differentiation are well underway. Muscle development requires co-ordinated expression of a family of basic helix-loop-helix (bHLH) myogenic regulatory factors (MRFs): myoD, myogenin, myf5 and MRF4 [22-24]. MyoD directly activates the expression of the other transcription factors, such as myogenin, and plays a central role in the induction of myogenic differentiation [25]. Myocyte enhancer factor 2C (MEF2C) is expressed at all stages of satellite cell myogenesis and is required for the development of skeletal muscle [26]. A skeletal muscle myosin light chain kinase (skMLCK) is important for the activation of satellite cells and their subsequent differentiation [27], as well as for skeletal muscle development. Seale et al. showed that deletion of Pax7 in mice compromises muscle regeneration due to increased satellite cell apoptosis [28]. Moreover, a number of signalling pathways are involved in the process of muscle development: Wnt signalling modulates both the number of terminally differentiated myogenic cells and the intricate slow/fast patterning of the limb musculature [29], while TGF-beta inhibits myogenic differentiation in myoblasts [30]. At present, the molecular mechanism that modulates skeletal muscle development is not fully understood.

Retinoic Acid (RA) has been demonstrated to regulate heart chamber development and promote myogenic differentiation [31-34]. Our previous study revealed that RA signalling restricts heart chamber formation via fhl1A, and that fhl1A is expressed in the early stages of skeletal muscle and cardiac myocyte development. We, therefore, hypothesised that RA signalling may also affect skeletal muscle development through fhl1A. In the present study, we observed the role of zebrafish fhl1A, the homologue of human FHL1, on skeletal muscle development. We demonstrated that fhl1A functions as an activator in regulating the number of satellite cells and in skeletal muscle development.

Materials and methods

Zebrafish Maintenance

Wild-type AB strain zebrafish was raised under standard laboratory conditions at 28.5°C in non-buffered E3.

Whole-Mount In Situ Hybridization

In situ RNA hybridization was performed using digoxigenin-labelled antisense RNA probes. Total RNA was extracted as previously described, then reverse transcribed to cDNA [34]. To construct probes, 400–1000 bp exon sequences of antisense probes for fhl1A, myoD, skMLCK and mef2ca were amplified from cDNA using PCR and cloned into pGEM-T or pGEM-T easy vectors (Promega, Madison, WI, USA) [35-37]. They were then transcribed using RNA transcription reagents (Promega). The primer sequence used was as follows: skMLCK S: 5′-GCTACAACCCTCCTAAACT-3′; R: 5′-CACTCATACGACCACTTCTT-3′.

DEAB and RA Treatments

Zebrafish embryos were treated with all-trans RA (Sigma, St. Louis, MO, USA) and DEAB (Sigma) beginning at 40% epiboly as previously described [34]. The embryos were washed in PBST (1X PBS, 0.1% Tween 20) and fixed overnight at the indicated stages at 4°C in 4% paraformaldehyde.

Whole-Mount Antibody Staining

Embryos at the desired developmental stages were gathered and processed as previously described. Whole-mount embryo immunostaining was performed following standard protocols [38]. The primary antibodies used were as follows: MF20 (1:200 dilutions, DSHB) and Pax7 (1:200 dilutions, DSHB). The fluorescent secondary anti-mouse antibody used for detection was Alex-594-conjugated anti-mouse IgG (Invitrogen, Carlsbad, CA, USA). Stained embryos were photographed using fluorescence optics from Nikon (Nikon, Japan).

Antisense Morpholinos, Rescue and Overexpression Experiments

Two non-overlapping antisense morpholino oligonucleotides of fhl1A-MO1 and fhl1A-MO2 targeting the translational start site of zebrafish fhl1A (fhl1A MO) were synthesised at Gene-Tools (OR, USA), along with standard control morpholinos. The fhl1A MO was used as previously described [37]. The fhl1A-capped mRNA transcripts of fhl1A were synthesised and used as previously described [37]. After injection, embryos were incubated at 28.5°C in embryo medium.

Statistical Analysis

All data were expressed as mean ± standard deviation using SigmaPlot 11.0. Statistical differences

between the control group and MO-treated groups were determined using the rank sum test. For phenotype classification in the later developmental stages (after 24 hpf), we observed the overall phenotype; fhl1A MO-injected embryos with short trunks, abnormal dorsal curved body shape and embryos with pigment loss were defined as defective embryos. Embryos without these phenotypic traits were classified as normal. For embryos in the early developmental stages (before 24 hpf), we carefully checked the somite shape and the body axis morphology, and those with shorter, round somites were defined as defective embryos. Those with longer, narrow somites were defined as normal. Results were considered statistically significant if p < 0.05.

Western Blotting

Wild-type AB strain embryos were treated as described above with DEAB and RA. Zebrafish embryos were manually dechorionated and deyolked before homogenisation. Deyolked samples were dissolved in 2 μL of 2X sodium dodecyl sulphate (SDS) sample buffer per embryo and incubated for 5 min at 95°C. After full-speed centrifugation for 1 min in a microcentrifuge to remove insoluble particles, samples were separated on a 10% SDS polyacrylamide gel. Western blot was performed according to the previous report using the human anti-FHL1 antibody (1:1500 dilution, Proteintech) and anti-β-Actin antibody (1:4000 dilution, Proteintech) [34].

RT-PCR

PCR amplification was performed in a 25 µL reaction mixture containing 1 µM pax7a primers (F: 5′-CCAGAACTACCCACGAAC-3′ and R: 5′-CTTCCACCAATAGCACCC-3′), pax7b primers (F: 5′-AAGACGACGAGGATGATTGT-3′ and R: 5′-TCTGGTGTAGATGTCAGGGTAG-3′), skMLCK primers (F: 5′-GCTGTCGGAGACCGTAAA-3′ and R: 5′-AATCAGCAGCCAGTCAAA-3′) or β-actin primers (F: 5′-TATTGTGATGGACTCTGGTGATG-3′ and R: 5′-TCGGCTGTGGTGGTGAAG-3′), 1.5 mM MgCl, 0.2 mM dNTP, and 1 U Taq (Takara, Japan). The samples were incubated in a thermal cycler (Hybaid MultiBlock System) at 95°C for 5 min; then subjected to 30 cycles of 95°C for 30 s, 55°C for 30 s and 72°C for 30 s; followed by final extension at 72°C for 8 min. All PCR products were separated using electrophoresis on a 1.2% agarose gel.

Results

Expression of fhl1A and its Role in Skeletal Muscle Development

We cloned the ORF of fhl1A and synthesised a DIG-labelled fhl1A antisense RNA probe as previously described [37]. The sense RNA probe was used as a negative control. The expression pattern of zebrafish fhl1A during embryogenesis was analysed by whole-mount in situ hybridization. As shown in Fig. , fhl1A expression was not detected in somites at the 3-somite stage but was detected in somites at the 10- and 20-somite stages, implying that it may not be involved in progenitor skeletal muscle initiation but may play a role in skeletal muscle development. To assess the functions of fhl1A, we performed loss-of-function experiments in zebrafish using fhl1A MO, which blocks the expression of Fhl1A protein in zebrafish embryos [37]. As shown in Fig. and , the morphant embryos injected with 2 ng fhl1A-MO exhibited a defective phenotype, with shorter somites at the 12-somite stage, compared with control embryos injected with 2 ng Control-MO (Fig. and ). When fhl1A MO and fhl1A-capped mRNA were co-injected, 89% of the embryos were restored to normal phenotypes (Fig. , and ). Stronger defects were observed at 24 hpf; instead of the healthy chevron shape observed in control somites (Fig. , ), 67% of the embryos showed slash-shaped somites (Fig. , ). These somites became shorter and rounder, with broadened somitic furrows throughout the trunk (Fig. ). When embryos were co-injected with fhl1A MO and fhl1A-capped mRNA, 76% of these defects were restored to normal (Fig. , and ). These results demonstrate that alterations in the gene expression of fhl1A are associated with morphologic deformities in the somites and skeletal muscles.

Effects of fhl1A on the Expression of Skeletal Myogenesis Markers

Previous studies have shown that contractile muscles are derived from several populations of skeletal muscle precursors in somites [39]. The specific tempo-spatial expression pattern of fhl1A raises the question of whether it plays any role in skeletal myogenesis. Expression of myoD takes place in the somites and adaxial cells and plays a central role in the induction of myogenic differentiation [36]. We used whole-mount in situ hybridization analysis to characterise the effects of fhl1A knockdown and overexpression on myoD expression during skeletal myogenesis.

The results demonstrated that, compared with control embryos (Fig. ), fhl1A knockdown (Fig. ) and overexpression (Fig. ) had no effect on MyoD expression at the 3-somite stage, excepting a faint curvature of the axis (Fig. ). At the 12-somite stage, compared with control embryos (Fig. ), fhl1A knockdown led to a reduction of myoD expression in 82% of embryos (Fig. , ), but fhl1A overexpression resulted in a broader and stronger myoD expression in 78% of embryos (Fig. , ). When fhl1A MO and fhl1A-capped mRNA were co-injected, 89% of the embryos showed near normal expression of myoD (Fig. , ). The question arose whether the protein level in skeletal muscle development was affected by fhl1A. At the 12-somite stage, 89% of fhl1A knockdown embryos showed a narrow expression of muscle fibres (Fig. , ) relative to controls (Fig. ), and 89% of fhl1A overexpression embryos exhibited broader expression domains of muscle fibres (Fig. , , arrowhead indicated), which is consistent with the phenotypes detected by whole-mount in situ hybridization analysis. A similar phenotype was observed at the 25-somite stage, during which anti-MF20 antibody staining indicated that the knockdown of fhl1A was associated with defective expression of muscle fibres (Fig. ). At 24 hpf, fhl1A depletion abolished the expression of myoD (71%, Fig. , ). This was further confirmed by the analysis at 28 hpf (81%, Fig. , ). In embryos co-injected with fhl1A mRNA, myoD expression was robust (65%, Fig. , ). This was further confirmed by the analysis at 28 hpf when delayed disappearance of myoD relative to control embryos was observed in the tails of the injected embryos (79%, Fig. , ). Recovery of myoD expression was observed in 93% of embryos (Fig. , ). These data support the conclusion that fhl1A affects skeletal muscle development; more specifically, fhl1A may delay the degradation of skeletal muscle differentiation marker, myoD.

Effects of fhl1A on Satellite Cell Formation

Previous reports have shown that satellite cells are progenitors of skeletal muscles and that the satellite cell marker pax7 is expressed in the anterior part of somites during the 14-somite stage [40]. We examined whether fhl1A expression regulates the formation of satellite cells by first evaluating its effects on pax7 mRNA levels using RT-PCR analysis. The results demonstrated that, in fhl1A knockdown embryos, the mRNA expression of pax7a and pax7b was significantly decreased at 24 hpf (Fig. , , p < 0.01). These results were further confirmed by immunostaining with anti-Pax7 antibody, which demonstrated that the number of Pax7-positive satellite cells at 24 hpf was significantly lower (p < 0.001) in fhl1A knockdown embryos (with an average of 183 cells counted from somitic 10 to 17, n = 8) than in controls (with an average of 279 cells, n = 8) as shown in Fig. , suggesting that fhl1A affected the number of satellite cells.

Previous studies have shown that mef2c is highly expressed in quiescent satellite cells and during satellite cell myogenesis [41]. They have established that skeletal muscle commitment takes place through the amplification of and synergism with myoD [42, 43]. We examined whether skeletal muscle development affected by fhl1A could bypass mef2ca, a zebrafish homologue of human mef2c, at the 12-somite stage. In fhl1A knockdown embryos, mef2ca expression was reduced in anterior somites (Fig. , arrowhead indicated), with 77% of the embryos showing defective mef2ca expression (Fig. ). However, fhl1A overexpression caused upregulation of mef2ca expression in somites, with high ectopic expression in a bilateral pattern along the somitic paraxial mesoderm. This took place in 65% of embryos with defective phenotypes (Fig. , ). These mef2ca expression phenotypes were recovered in 82% of the fhl1A-MO embryos co-injected with fhl1A mRNA (Fig. , ). Recently, skMLCK was found to regulate myoD and myf5 expression in ES and satellite cells, and the specificity of skMLCK regulating activities may be mediated by the ability of mef2c to recruit p300 to skeletal muscle promoters during muscle commitment [27]. We explored whether expression of fhl1A affects the expression of skMLCK in skeletal myogenesis. As shown in Fig. and , the mRNA expression level of skMLCK was significantly decreased in fhl1A knockdown embryos (p < 0.001). These results were further confirmed by whole-mount in situ hybridization, which showed that skMLCK expression was significantly reduced in 90% of fhl1A knockdown embryos relative to controls (Fig. ). These data support the observation that fhl1A plays a role in regulating satellite cells for skeletal muscle development.

Effects of RA Signalling on fhl1A Expression and Skeletal Muscle Development

As the expression of fhl1A is regulated by RA signalling, we proposed that the role of fhl1A in skeletal muscle development could be modulated by RA signalling [37]. We analysed fhl1A expression during myogenesis at the 10-somite stage. In the presence of exogenous RA, higher levels of fhl1A expression were induced in somites than in controls (90%, n = 76, Fig. , ). In contrast, fhl1A expression was lower in somites in the presence of exogenous DEAB (89%, n = 65 Fig. , ) than in controls (n = 70, Fig. ). Next, we monitored the protein levels of Fhl1A after treatment with RA and DEAB, and under these conditions, the endogenous Fhl1A was upregulated and downregulated, respectively (Fig. ). These data suggest that endogenous fhl1A expression in skeletal muscles is influenced by RA signalling.

We next determined whether the myoD expression in somites could be modulated by RA signalling. As expected, the myoD expression in the somites of RA-treated embryos at the 10- and 16-somite stages (98%, Fig. , , and ) was higher than in controls (100%, Fig. , ), but the expression of myoD in adaxial cells appeared to be unchanged. In DEAB-treated embryos, myoD expression was reduced in somites (93% Fig. , ), similar to results observed in fhl1A morphant embryos (Fig. , ). Phenotypes in the somites regulated by RA signalling resembled those of fhl1A knockdown and overexpression embryos (Fig. , ), suggesting that fhl1A expression has effects comparable to those of RA signalling in myogenesis. These data collectively point to a role of fhl1A in specifying embryonic skeletal muscle downstream of RA signalling in zebrafish.

Discussion

In the present study, we demonstrated that fhl1A is specifically expressed in skeletal muscles during the 8-, 12- and 20-somite stages. We provide evidence that the expression of fhl1A plays a pivotal role in the development of skeletal muscle.

To examine the effects of fhl1A loss of function on skeletal muscle development, we used fhl1A MO to knock down Fhl1A protein expression. The effectiveness of fhl1A MO used in this analysis was demonstrated in our previous paper, in which we examined the fhl1A-GFP expression in zebrafish embryos by detecting Fhl1A-GFP fluorescence and analysed the Fhl1A protein expression by western blot. These results suggest that fhl1A MO specifically blocks the expression of the fhl1A gene in vivo [37]. In the present study, we demonstrated that the tempo-spatial specific expression of fhl1A had associated effects on skeletal muscle development. During the 12- through 18- somite stages, fhl1A expression was only detectable in the skeletal muscle cells, suggesting that fhl1A impacts the development of skeletal muscle. Knockdown of fhl1A caused defective skeletal muscle development. These defects were found to be largely recovered in fhl1A morphants co-injected with fhl1A mRNA. These results suggest that fhl1A is required for skeletal muscle development.

Fhl1A directly activates the expression of myogenic regulatory factors myoD, myf5, mrf4 and myogenin, thus activating a skeletal muscle differentiation gene expression programme [44, 45]. We demonstrated that the expression of myoD is significantly reduced in fhl1A knockdown embryos (Fig. ), and that fhl1A overexpression led to an increased expression of myoD (Fig. ). We further analysed myoD expression at 24 hpf and found it to be minimal. No myoD expression was detected in fhl1A knockdown embryos (Fig. 2Q), while expression was delayed in embryos overexpressing fhl1A. Abnormal expression of myoD could be corrected at 24 phf (Fig. ), with similar results observed at 28 hpf (Fig. ).

Satellite cells are progenitor cells that generate the myoblasts required for muscle growth, repair and regeneration [21]. Mef2c is expressed at all stages of satellite cell myogenesis and is required for the development of skeletal muscle [26]. Expression of skMLCK is important for the activation of satellite cells and their subsequent differentiation [27]. Our results demonstrated that, at 24 hpf, knockdown of fhl1A produced significant downregulation of the expression of satellite cell markers pax7a and pax7b, and the regulation genes mef2ca and skMLCK, suggesting that fhl1A plays a pivotal role in the formation of satellite cells. Our results further demonstrated that fhl1A expression in somites is controlled by RA signalling. Previous studies have shown that RA signalling activates the expression of fast muscle myosin by inducing premature myoD expression in the presomitic mesoderm, but that DEAB represses it by reducing expression of myoD in somites [46]. Our results demonstrated that both fhl1A expression and RA signalling produced a similar phenotype, suggesting that the role of fhl1A in skeletal muscle development is likely regulated by RA signalling. Collectively, these results demonstrate that the fhl1A gene functions as an activator of skeletal muscle development in living organisms.

Mutations in the FHL1 gene have been identified in a number of human myopathies. In the present study, we demonstrated that skeletal muscle defects occur after alterations in fhl1A expression, indicating that the zebrafish fhl1A mutation model may be suitable for studying the effects of genetic variations identified in patients with myopathies.

CONCLUSION

In summary, our results demonstrate that fhl1A plays pivotal roles in regulating the number of satellite cells and skeletal muscle development. However, a complete understanding of the specific molecular mechanism by which fhl1A promotes skeletal myogenesis requires further investigation.

ETHICS APPROVAL AND CONSENT TO PARTICIPATE

Full details of the study were approved by Ethics Committee of Hunan Normal University.

HUMAN AND ANIMAL RIGHTS

No humans were used for studies that are base of this research. All embryos were handled according to relevant national and international guidelines “Act on Welfare and Management of Animals”.

CONSENT FOR PUBLICATION

Not applicable.