Ezrin Prompted Myoblast Differentiation and Muscle Fiber Specialization and Gastrocnemius Muscle Repair in Peroneal Nerve Injury Model through PKA-NFATs Signaling Pathway


 Background: Muscular dystrophy is a destructive neuromuscular disease characterized by progressive muscle weakness and muscle atrophy. The role of Ezrin in myoblast differentiation/fusion and muscle atrophy is still unknown.Method: Gastrocnemius muscle atrophy model were prepared by mechanical clamp of peroneal nerve. Differentiating C2C12 cells treated with Ad-Ezrin or Ad-shEzrin were detected by gene chip, Q-PCR, immunofluorescence staining and Western blot.Results: Ezrin was expressed in MyHC I/II myofibers in vivo, and time-dependently increased during myoblast differentiation/fusion characterized by MyoG+/MEF2c nuclei and MyHC+ myotubes in vitro. Overexpression of Ezrin promoted myoblast differentiation/fusion in time-dependent manner, inducing the increased MyHC-I+ and MyHC-II+ muscle fiber specialization, the specific effects could be abolished by addition of Ad-Periaxin. Ad-Ezrin did not alter PKA and PKAreg II α levels, but PKAreg I α/β. The PKA inhibitor, H-89, remarkably abolished the over-expression effects by Ezrin on an increased myoblast differentiation/fusion. By contrast, Knockdown of Ezrin by shRNA significantly delayed myoblast differentiation/fusion accompanied by the decreased PKA reg I/II ratio, the inhibitory effects could be eliminated by PKAreg I activator N6-Bz-cAMP. Meanwhile, Ad-Ezrin enhanced type I muscle fiber specialization, accompanied by the increased levels of NFATc1/c2.Furthermore, Ad-NFATc2 or Ad-NFATc4 reversed the inhibitory effects of Ad-shEzrin on myoblast differentiation/fusion. Importantly, in vivo transfection of Ad-Ezrin into gastrocnemius muscles in peroneal nerve injury model increased the numbers of MyHC-I+ and MyHC-II+ myofibers, reducing muscle atrophy and fibrosis.Conclusions: Ezrin activated PKA-NFAT-MyoD/MyoG/MEF2C signaling pathway, triggering myoblast differentiation/fusion and muscle fiber specialization in periaxin-depentdent manner, contributing to gastrocnemius muscles repair.


Introduction
Muscular dystrophy (MD) is a destructive neuromuscular disease characterized by progressive muscle weakness, muscle atrophy, and cardiac dysfunction 1 . In addition to dysfunction of neurotrophic effect, another possible contributor to the generation of pathology in MD is the primary muscular disorders 2, 3 .
Indeed, muscle satellite cells function in the physiological self-renewal and repair of pathological injury 4 .
However, the underlying mechanism of muscle satellite cells remains unclear during the development of skeletal myopathy.
Published data showed that Periaxin could be a candidate gene for CMT4F by destroying the myelin sheath formed by Schwann cells 5 . Of interest, Ezrin, a member of the Ezrin/radixin/moesin (ERM) protein family, is encoded byvillin2 located on chromosome 6q25.2-q26, which could inhibit the self-association of L-periaxin to participate in myelin sheath maintenance 6 . Evidence that the N-terminal of Ezrin connecting with CD44 on the cell membrane 7 , and the C-terminal linking into actin laments were involved in tumor invasion and metastasis indicated that Ezrin could play an important role in MD because CD44 and F-actin performed crucial function in muscle satellite cells differentiation and fusion.
Accumulated data showed that Ezrin could anchor cAMP dependent protein kinases, resulting in the activation of PKA, which phosphorylates enzyme proteins or channel proteins, such as Na+/H+ exchanger (NHE3), and changes their gating properties 8 . Recent reports have shown that the changes of pH were closely related to the regeneration and repair of skeletal muscle mediated by satellite cells 9,10 . Furthermore, the activated protein kinase A (PKA) has been linked to the unique phenomenon of myoblast differentiation/fusion and myotube formation, ascribing to the alteration in PKA regulatory subunit I (PKA RI) under normal differentiation condition 11 . Our previous study suggested that it abated the ratio of PKA RI/RII in myoblast cells, resulting in the postponement of myoblast differentiation and fusion 12 . Further evidence shows that ERM proteins act as PKA-anchoring proteins and sequester PKA close to its target proteins for their effective phosphorylation and functional regulation 13 . NFATs (Nuclear factor of activated T cells) activation mediated by PKA plays a crucial role in myoblast differentiation and fusion, myotube size, and altered muscle ber specialization 4,13 . In this study, overexpression of Ezrin increased the number of MyHC-I/II myo bers, resulting in the recovery of gastrocnemius muscle atrophy in peroneal nerve injury model, which were attributed to the higher potential of myoblast cells differentiation/fusion and myo ber specialization through PKA-NFAT-MEF2C signaling pathway.

Method
Animals, In vivo transfection and Peroneal nerve injury model preparement According to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health and China, animal studies were performed accordingly. Experimental Animal Centre of Hubei Medical University provided C57BL/6 mice (male, 3-5 months) that meet the criteria. The Institutional Animal Care and Use Committee of Hubei Medical University approved animal protocols.
For the transduction ofadult muscles, C57BL/6male micewere anesthetized by using an iso urane vaporizer maintained at 2% iso urane, 1 L/m oxygen. Gastrocnemius and soleus (SL) muscles were exposed and injected with Ad-Ezrin (1×10 10 pfu, two point, 50μm/each) 15 . Muscles were removed 7 days after transfection, frozen inisopentane cooled in liquid nitrogen, and stored at -80°C. The e ciency of transfection was measured through detecting His-tag.
C2C12 myoblast culture and differentiation induction C2C12 myoblast (Purchased from cell resource center of Shanghai Academy of life sciences, Chinese Academy of Sciences) were inoculated in 75-cm 2 culture dishesand cultured with proliferation medium (PM) containing high glucose DMEM (Gibco, USA, HG-DMEM) supplemented with 10% FBS (C0225, AusGenex Fetal Bovine Serum Excellent ) at 37℃and 5% CO 2 . When the con uence of the cells reached 75%, the PM was replaced with differentiation medium (MD) containing HG-DMEM supplemented with 2% horse serum (HS, BI 04-124-1A, Sigma, USA) to induce C2C12 myoblast cell differentiation. Traits of myotube formation from myoblast differentiation were observed daily under a microscope 14 .

Adenoviral vector preparement and in vitro transfection
Construction of Ezrin, L-Periaxin, NFATc1/c2 overexpression adenoviral vector were prepared as previously described 15 . Constructions of Ezrin, L-Periaxin and NFATc3/c4 short hairpin RNA (shRNA) adenoviral vector were prepared as previously described 15 . These overexpression adenoviral vectors containing Ad-NFATc1, Ad-NFATc2, Ad-shNFATc3 and Ad-shNFATc4 were obtained from Vigenebio. To con rm the role of L-Periaxin in myoblast, the addition of Ad-Null, Ad-Periaxin, or Ad-shPeriaxin (1×10 9 pfu) into corresponding culture dish one day before Ad-Ezrin or Ad-shEzrin was added. To con rm the role of NFATc3 or NFAtc4 on myoblast, the addition of Ad-Null, Ad-shNFATc3, or Ad-shNFATc4 (1×10 9 pfu) into corresponding culture dish one day before Ad-Ezrin or Ad-shEzrin was added.
And then proliferation medium was replaced with differentiation medium for further observation.

Myoblast differentiation
After myoblast were treated under DM for the indicated time, the differentiated myoblast was stained for MyoG or MEF2C using the primary polyclonal antibody MyoG (sc-12732, 1:150, Santa Cruze) or MEF2C (5030S, 1:400, CST) and appropriative TRITC-labeled secondary antibody (Jackson Lab, 1:500, USA). The nuclei were stained with DAPI. C2C12 myoblast with only 1-2 nuclei within a cellular structure were evaluated with MyoG or MEF2C staining. The MyoG+ or MEF2C+ cells were de ned as the differentiated cells that did not fuse to form myotubes. Myoblast with 3 or more nucleiin the structure of a cell were de ned as myotubes. The number of double-positive nuclei under high power eld (HPF, 50 μm) were analyzed after double staining of MyoG/DAPI or MEF2C/DAPI. Two individuals who did not know the results evaluated the images using Image J (Java) software (National Institutes of Health, USA).
To analyze myotube morphology, we divided the cells into 2 groups, including short myotube with 3~5 myoblast fusion and long myotube with more than 5 myoblast fusions 15,17 . Two individuals who did not know the results evaluated the images using Image J (Java) software (National Institutes of Health, USA).

Microarray analysis and Quantitative RT-PCR
The C2C12 myoblast treated with Ad-Ezrin or Ad-shEzrin for 6 days were prepared according to the manufacturer's protocol, and detected and analyzed (Shanghai OE Biotech CO, LTD). Total RNA from C2C12 myoblast was obtained using TRIzol (Invitrogen, Life Technologies) and transcribed into cDNA using the SuperScript II cDNA kit (Invitrogen, Life Technologies). Quantitative PCR was carried out using SYBR green PCR master mix (Thermo Fisher Scienti c, Applied Biosystems, CN) in Real-Time PCR System (RotorGene 6000, Qiagen, Germany). The transcript levels of the gene of interest in each group were normalized to GAPDH levels 18 . The primers used are listed in Table 1.

Table1
The sequences of primers of qPCR.

Gene
Forward Reverse qPCRs were performed to identi ed satellite cell differentiation and muscle bers traits by using the speci c primers of satellite cell differentiation markers including MyoD and MyoG, type I muscle ber makers like MyHC1, and type II muscle ber makers such as MyHC2a, MyHC2b, and MyHC2X.

Statistical analysis
Data of quantitative and semi-quantitative analysis presented are mean ± SD. Paired or unpaired Student's t-test determined statistical signi cance between the two groups. One-way ANOVA was used to compare the results for more than two experimental groups to specify the differences between groups. P<0.05 is considered meaningful.

Results
Ezrin was speci cally expressed in myoblasts and their skeletal muscle bers To con rm the role of Ezrin in skeletal muscle, we rstly detected if Ezrin was expressed in gastrocnemius muscle, as shown in gure1A, myo bers partially were expressing Ezrin. To distinguish the traits of Ezrin expression in different myo bers including MyHC-I and MyHC-II, showing that more MyHC-II myo bers were positive for Ezrin through double immuno uorescence staining, in addition to MyHC-I myo bers ( Figure.1B-1D). Subsequently, to further con rm if Ezrin was expressed in myoblast cells and differentiating myoblast cells, the expression of Ezrin during the process of C2C12 cells differentiation/fusion was analyzed by and immuno uorescence staining and western blotting. We found that Ezrin expression gradually increased during myoblast differentiation, reaching peak levels on day 4 of differentiation ( Figure.1E-1G). Furthermore, in line with gure1B, differentiated C2C12 myoblast cells and formed myotubes showed the positive traits for Ezrin ( Figure.1H). These results indicated that speci c expression of Ezrin in myoblasts and their skeletal muscle bers could involve in myoblast differentiation/fusion and muscle ber specialization.

Ezrin involved in myoblast differentiation/fusion
To determine the effect of Ezrinon myoblast differentiation and fusion, we transfected C2C12 cells with an adenovirus and shRNA to overexpress or knockdown the expression ofEzrin, respectively. Assessment of the transfection e ciency revealed that following the application of 100 optimal multiplication of infection (MOI) with the adenovirus, C2C12 myoblast almost reached a con uence of 95%(s g.1A-1C).
His-tag was used to con rm the successful expression of the exogenous genes and their respective functions (s g.1C-1E).
Our results further showed that the number of MyHC+ myotubes with either 3-5 or 5 + nuclei increased upon treatment with Ad-Ezrin in a time-dependent manner ( Figure.2A-2D). Conversely, knockdown of Ezrin by shRNA did not only obviously reduce MyHC+ cell number, but it also dramatically decreased myotube numbers with either 3-5 or 5 + nuclei ( Figure. We then performed MyoG and MEF2C immunostaining assay to explore the relationship between Ezrin and the myogenic markers. Our results showed that following treatment with Ad-Ezrin, the MyoG and MEF2C expression increased, with the number of MyoG + nuclei signi cantly higher than that of MEF2C+ nuclei ( Figure.2F-2I). Meanwhile, the knockdown effects of Ezrin with shRNA resulted in markedly declined number of MyoG+ and MEF2C+ nuclei cells in differentiated myoblast cells ( Figure.2F-2I). In a word, these results indicated that Ezrin affects myoblast differentiation via the modulation of myogenic markers, MyoG and MEF2C, but more especially, MyoG.

Ezrin involved in myo ber specialization
Because skeletal muscle function differences were associated with muscle bers types 14 ,we used gene chip to analyze the traits of myo bers types in differentiated myoblast cells treated with Ad-Ezrin or Ad-shEzrin. As shown in Figure Interaction between Ezrin and L-Periaxin involved in myoblast differentiation and myo ber specialization Since mutation and deletion of L-Periaxin was associated with Charcot-Marie-Tooth (CMT) characterized by progressive muscle weakness and atrophy of distal extremities with sensory impairment through destroying the myelin sheath formed by Schwann cells 4,5 . Interestingly, Ezrin could inhibit the selfassociation of L-periaxin to participate in myelin sheath maintenance. These results push us to explore if Ezrin-mediated myoblast differentiation involved in L-Periaxin, we rstly found that L-Periaxin expression were gradually increased and reached the peak at 6 day of myoblast differentiation (sFigure3A-3C), which was different from the peak expression of Ezrin at 4 day of myoblast differentiation (Figure1F), indicating that Ezrin could affect myoblast differentiation through regulating L-periaxin. Unfortunately, as shown in Figure 4A-4E and sFigure4, overexpression of alone L-Periaxin (Ad-Periaxin) did not alter Ezrin-induced myoblast differentiation, but overexpression of L-Periaxin obviously abolished the bene cial effect of overexpressing-Ezrin on myoblast differentiation while knockdown of L-Periaxin by shRNA slightly enhanced overexpressing-Ezrin' effects. These results indicated that interaction between Ezrin and L-Periaxin involved in myoblast differentiation, at least in part.
To further con rm if Ezrin and L-Periaxin involved in myoblast differentiation and myo ber specialization, we established peroneal nerve injury (PNI) model to partially mimic CMT-associated muscle atrophy, results showed the traits of gastrocnemius muscle atrophy (sFigure.5). Meanwhile, we found that overexpression of Ezrin markedly increased MyHC-I and MyHC-II positive myo bers numbers (Figure5A-5D) while it reduced muscle brosis (sFigure.5). Unfortunately, overexpression of L-Periaxin (Ad-Periaxin) did not recover the PNI-induced gastrocnemius muscle atrophy but deteriorated its brosis (sFigure.5). Furthermore, knockdown of L-Periaxin by shRNA could not enhance the bene cial effects of Ad-Ezrin on repairing gastrocnemius muscle injury mediated by PNI. Of interest, Ad-Periaxin reduced the numbers of MyHC-II weakly positive myo bers compared with sham and PNI groups. And knockdown of L-Periaxin by shRNA did not abolish the enhanced effects of Ezrin on MyHC-II weakly positive myo bers ( Figure.5A-5D). These results suggested that myoblast differentiation and myo ber specialization did not directly involve in L-Periaxin, but Ezrin.

Ezrin regulated myoblast differentiation and fusion through PKA signaling pathway
Previous studies have shown that the PKA and PKAreg I/II ratio play crucial roles in controlling myoblast differentiation and fusion 11,12 . To con rm if Ezrin's role in myoblast differentiation and fusion was involved in PKA signaling pathway, we used gene chip and western blot to analyze the traits of PKA signaling pathwayin differentiated myoblast cells treated with Ad-Ezrin or Ad-shEzrin. Our results revealed that, the overexpression of Ezrin did not alter PKAreg II levels, but it signi cantly increased the levels of PKAα, PKAreg Iα, and PKAreg Iβ, resulting in an increased PKAreg I/II ratio. By contrast, knockdown of Ezrin by shRNA signi cantly reduced PKAreg Iα and PKAreg Iβ levels, but it did not alter PKAreg II levels, resulting in a decreased PKAreg I/II ratio ( Figure.6A-6H).
Since the activity of PKA has been reported to have effects on myoblast differentiation and fusion 11,12 .
Combining with the above observations that the overexpression or knockdown of Ezrin affected myoblast differentiation and fusion by altering PKA activity, we treated C2C12 cells with a PKA inhibitor (H-89). We found that the inhibition of PKA abolished the bene cial effects of Ezrin onmyoblast differentiation and fusion. By contrast, the PKA activator reversed the inhibitory effects of Ezrin knockdown on C2C12 myoblast differentiation and fusion ( Figure.6I-6K). These results indicated that Ezrin mediation of myoblast differentiation and fusion is associated with the PKA signaling pathway.
Ezrin regulated myoblast differentiation and fusion through PKA-MyoG/MEF2C signaling pathway Indeed, myotube is formed by the fusion of differentiated myoblasts, which is characterized by three (3 + ) or morenuclei in the cell structure 15,17 . As shown in Figure. 6I and sFigure.6-7, we found that knockdown of Ezrin by shRNA markedly decreased MyoG + and MEF2C + less than 3 nuclei, and these effects could be abolished by PKA activator. By contrast, overexpression of Ezrin substantially increased the number of MyoG+ or MEF2C+ nuclei in less than 3 and 3 + myotubes. Furthermore, western blot showed that overexpression of Ezrin increased the nuclear levels of MyoG and MEF2C while knockdown of Ezrin by shRNA reduced it. More important, these speci c changes could be cancelled by the PKA inhibitor and PKA activator, respectively (sFigure.6-7). These results indicated that Ezrin participated in C2C12 myoblast differentiation and fusion through PKA-MyoG/MEF2C signaling pathway.

NFATs signaling involved in the regulation of myoblast differentiation/fusionmediated by Ezrin
Since NFATs play a crucial important role in myoblast differentiation/fusion, especially myotube speci cation 26-32 .We used gene chip and western blot to analyze the traits of NFATs signaling pathway in differentiated myoblast cells treated with Ad-Ezrin or Ad-shEzrin. Results showed that the overexpression of Ezrin promoted nuclear translocation of NFATc1/c2 that led to increased levels of NFATc1/c2 nuclei, and decreased NFATc3/c4 nuclei level. Conversely, knocking down Ezrin induced nuclear translocation of NFATc3/c4 resulting in increased NFATc3/c4 nuclei level, and decreased the NFATc1/c2 nuclei levels ( Figure.7A-7G, sFigure8). Secondly, the overexpression of NFATc2 or knockdown of NFATc4 almost completely reversed the inhibitory effects of knocking-down Ezrin during myoblast differentiation and fusion, resulting in the recovery of slim-long myotubes (nuclei numbers more than 5) ( Figure.7F-7G). Eventually, NFATc1/c3 knockdown did not signi cantly recuperate the inhibitory effects of knocking-down Ezrin by shRNA in myoblast differentiation/fusion ( Figure.7F-7G).
Since MyoG and MEF2C are involved in the initiation and later stage of myoblast differentiation, respectively 26 . As shown in Figure.7F-7GandsFigure. 9, we found that knockdown of Ezrin by shRNA markedly decreased the number and percentage of MyoG+ and MEF2C+ nuclei in less than 3-nuclei cells and 3-nuclei + myotubes, and treatment with Ad-NFATc2 or Ad-shNFATc4 could abolish these effects. In addition, Ad-NFATc1 or Ad-shNFATc3 reversed the number of MEF2C+ nuclei in 3-nuclei + myotubes. These results indicated that Ezrin participated in C2C12 myoblast differentiation and fusion with coordination of MyoG and MEF2C, which were associated with NFATc2/c4, at least in part.

NFATs signaling involved in the regulation of myo ber specialization mediated by Ezrin
Previous studies have reported on two types of muscle bers, including slow (slim-long) and fast (thickshort) myotubes, MyHC-I forms the former, and MyHC-2a, MyHC-2b or MyHC-2X the latter 24,25 .As shown in Figure. 8A-8D, we found that overexpression of Ezrin increased the numbers of MyHC-I and MyHC-II positive myotubes while knockdown of Ezrin decreased it, the speci c effects could be abolished by Ad-NFATc1/c2 or Ad-shNFATc3/c4. More importantly, the speci c increase in MyHC-2a and MyHC-2b mediated by knockdown of Ezrin could be obviously inversed by Ad-NFATc1/c2 or Ad-shNFATc4 (sFigure.10), respectively. Thus, Ezrin mainly regulated myo ber speci cation through integration role of NFATs signaling pathway.

Discussion
In this study, we made three novel observations. Firstly, we found that the Ezrin was expressed in MyHC-I + and MyHC-II + myo bers and showing a time-dynamic expression traits during myoblast differentiation and fusion. Secondly, Ezrin signi cantly controlled myoblast differentiation and fusion, which were related with PKA RI-NFAT-MyoD/MEF2C signaling pathway. And lastly, Ezrin contributed to the regeneration and repair of damaged gastrocnemius muscle induced by peroneal nerve injury.
Clinically, exploring the potential target of gene therapy for muscle atrophy is a hot spot. As one of ERM proteins, Ezrin activated the intracellular signal pathways through transferring extracellular signal molecules into actin cytoskeleton, affecting several key cellular processes, including membrane dynamics, cell adhesion, cell survival, motility, and determination of cell shape 6-10 . Indeed, these cellular processes were associated with myoblast differentiation and fusion 21 . Herein, for the rst time, we found that MyHC-I + and MyHC-II + myo bers showed the positive expression of Ezrin, and its levels gradually increased during myoblast differentiation/fusion, and rapidly decreased when myotubes approached and/or reached maturity. Although Periaxin could be a candidate gene for CMT4F by destroying the myelin sheath formed by Schwann cells 5 . Our novel nding further showed that overexpression of Lperiaxin (Ad-periaxin) obviously abolished the effect of Ezrin on both myoblast differentiation/fusion in vitro and muscle injury repair in vivo. Unfortunately, knockdown of L-periaxin by shRNA did not markedly enhance the therapeutic effect of Ezrin on muscle injury. Considering the recent report that Ezrin could inhibit the self-association of L-periaxin to participate in myelin sheath maintenance 6 , we conjectured that not L-periaxin, but Ezrin was bene cial way of gene therapy for the injured muscles including CMT4F. the involvement of PKA RI and RII during myoblast differentiation and myotube formation 11,12 . Of interest, we found that overexpressingEzrin markedly increased PKA RI levels, leading to an increased PKA RI/RII ratio, accompanied by an acceleration of myoblast differentiation and fusion. Furthermore, the knockdown of Ezrin, resulting in a lower PKA RI/RII ratio, inhibitory effects observed during myoblast differentiation/fusion could be reversed by the PKA RI activator. More importantly, PKA RI activator almost completely recovered the number of MyoG or MEF2C positive myotubes.