Calpain-6 Deficiency Promotes Skeletal Muscle Development and Regeneration

Calpains are Ca2+-dependent modulator Cys proteases that have a variety of functions in almost all eukaryotes. There are more than 10 well-conserved mammalian calpains, among which eutherian calpain-6 (CAPN6) is unique in that it has amino acid substitutions at the active-site Cys residue (to Lys in humans), strongly suggesting a loss of proteolytic activity. CAPN6 is expressed predominantly in embryonic muscles, placenta, and several cultured cell lines. We previously reported that CAPN6 is involved in regulating microtubule dynamics and actin reorganization in cultured cells. The physiological functions of CAPN6, however, are still unclear. Here, to elucidate CAPN6's in vivo roles, we generated Capn6-deficient mice, in which a lacZ expression cassette was integrated into the Capn6 gene. These Capn6-deficient mouse embryos expressed lacZ predominantly in skeletal muscles, as well as in cartilage and the heart. Histological and biochemical analyses showed that the CAPN6 deficiency promoted the development of embryonic skeletal muscle. In primary cultured skeletal muscle cells that were induced to differentiate into myotubes, Capn6 expression was detected in skeletal myocytes, and Capn6-deficient cultures showed increased differentiation. Furthermore, we found that CAPN6 was expressed in the regenerating skeletal muscles of adult mice after cardiotoxin-induced degeneration. In this experimental system, Capn6-deficient mice exhibited more advanced skeletal-muscle regeneration than heterozygotes or wild-type mice at the same time point. These results collectively showed that a loss of CAPN6 promotes skeletal muscle differentiation during both development and regeneration, suggesting a novel physiological function of CAPN6 as a suppressor of skeletal muscle differentiation.

The Capn5 and Capn6 genes were identified as orthologs of tra-3, a nematode gene for a sex-determination factor, TRA-3 [22,23]. The encoded two mammalian TRA-3 orthologs, CAPN5 (also called hTRA-3) and CAPN6, share more than 30% amino acid (aa) identity with TRA-3. These calpains have a C2 (not C2L) domain at the C-terminus, i.e., a ''CysPc-C2L-C2'' domain structure. CAPN5 has a Ca 2+ -dependent autolytic activity, is sensitive to several calpain inhibitors [24], and is expressed at varying levels in almost all tissues [25]. It is expressed by a subset of T cells, but the analysis of Capn5 2/2 mice showed that it is not required for development [26]. On the other hand, eutherian CAPN6 has a naturally occurring aa substitution at the active site Cys residue (Lys81 instead of Cys in human and mouse CAPN6), probably indicating a lack of proteolytic activity [25]. In terms of molecular evolution, metatherian (marsupial) and avian CAPN6 retain the active-site triad residues Cys-His-Asn, and frogs and fish have three TRA-3 homologs with conserved active-site residues.
Mammalian CAPN6 is predominantly expressed in embryonic muscles, placenta [27], and several cultured cell lines [28]. Using cultured cells, we previously found that CAPN6 regulates microtubule dynamics and actin reorganization by modulating the activity of a small G-protein [28,29]. The function of CAPN6 in vivo, however, is still unclear. In this study, we examined CAPN6's physiological role by analyzing Capn6-deficient mice. To our surprise, the Capn6-deficient mice displayed precocious development of embryonic skeletal muscle. We also found that CAPN6 was expressed in regenerating skeletal muscle in adulthood, and that Capn6 disruption promoted regeneration in cardiotoxin-injected skeletal muscle. Our results showed that a loss of CAPN6 promotes skeletal muscle differentiation during mouse development and regeneration, and suggest a novel physiological function for CAPN6 in suppressing differentiation in skeletal muscle.

Results
CAPN6 was successfully knocked-out in Capn6-lacZknock-in mice We disrupted the mouse Capn6 locus by replacing the proteincoding sequence in exon 2 with an nls-lacZ/PGKneo cassette ( Figure 1A), and obtained Capn6 lacZ/+ heterozygous mice. The heterozygotes, which were all female, as Capn6 is on the Xchromosome, appeared normal and were fertile. Capn6-deficient (Capn6 lacZ/Y male or Capn6 lacZ/lacZ female) mice were then obtained by intercrossing Capn6 lacZ/+ females with wild-type or Capn6 lacZ/Y male mice, respectively.
To confirm the absence of CAPN6 protein, whole embryos (at embryonic day 10.5 [E10.5], E12.5, and E14.5) were chosen for western blot analysis, because Capn6 mRNA is reported to be expressed during embryogenesis [27,28]. As shown in Figure 1C, in wild-type embryo lysate, an anti-CAPN6 antibody detected the expected 74-kDa band, and this band was absent in lysates of Capn6 lacZ/Y mice. Capn6 lacZ/lacZ mice gave essentially the same results (data not shown; see below and Figure 1E). These results indicated that the CAPN6 protein is present in wild-type mouse embryos, and that at least full-length CAPN6 was lost in the Capn6-deficient mice. The CAPN6 bands were most clearly detected in the lysates from the embryonic tongue and limb buds ( Figure 1D and E), which are rich in skeletal muscle. Notably, the signal intensities of the tongue CAPN6 were E14.5.E16.5.E18.5 ( Figure 1D).
The CAPN6 signal in Capn6 lacZ/+ (R) embryos was weaker than that in Capn6 +/Y (=) embryos ( Figure 1E), even though both possessed one Capn6 gene. This was probably because the Capn6 expression was partly suppressed by X-chromosome inactivation in the female mice, as with many other X-linked genes [30].

Capn6 is expressed in embryonic muscles and cartilages
In our constructs, the null allele, Capn6 lacZ , expressed ßgalactosidase (ß-Gal) in the cells where Capn6 is transcribed in wild-type mice. To verify that the expression of lacZ reflected that of authentic Capn6, the ß-Gal activity was examined in wholemount E10.5 Capn6 lacZ/+ embryos. As shown in Figures 2A and  2B, the mandibular arch, limb buds, somites, and heart were stained, consistent with the known Capn6 expression pattern [27,28]. At later stages (E11.5 and E14.5), ß-Gal was broadly expressed in the embryo ( Figure 2C and D). Cryosections of E16.5 Capn6 lacZ/+ embryos showed ß-Gal expression in all the skeletal muscles and bone cartilage throughout the embryonic body ( Figure 2E).
Capn6 lacZ/lacZ (Figures 2F, G, and S1A-C) and Capn6 lacZ/Y (data not shown) embryos, in which all the Capn6 locus-active cells expressed lacZ and no Capn6, gave essentially identical results: ß-Gal expression was observed in all muscles, cartilages, and bones at the epiphysis, throughout the craniofacial area and trunk. No obviously abnormal morphology was observed in these regions. These Capn6-deficient mice were born healthily and grew to adulthood without any apparent abnormality. In adult mice, only the placenta showed ß-Gal expression ( Figure 2H).

Capn6-deficient embryos exhibit advanced muscular development
Given the high levels of Capn6 expression in embryonic skeletal muscles, we focused on the physiological relevance of CAPN6 in skeletal muscle development. First, the morphology of the E16.5 tongue was compared between Capn6 lacZ/+ and Capn6 lacZ/lacZ mice. As shown in Figures 3A-D, more advanced tube formation and striated structures were observed in the tongue muscle of Capn6 lacZ/lacZ embryos than in their Capn6 lacZ/+ littermates. In addition, the diameters of the tongue muscle fibers in the Capn6 lacZ/lacZ mice were significantly larger than in their Capn6 lacZ/+ littermates ( Figures 3E and S2). To confirm these differences in skeletal muscle morphology, we examined the expression of desmin and a-sarcomeric actin, which are cytoskeletal components of skeletal muscle that play crucial roles in its development [31,32]. Western blot analysis of embryonic limb-

Author Summary
The calpains arose evolutionarily as a family of Ca 2+regulated intracellular proteases with divergent structures and functions. Unique among mammalian calpains, calpain-6 is ''non-proteolytic'', and its physiological role has remained unknown. In this study, using Capn6 knockout mice, we discovered that Calpain-6 is involved in skeletal muscle development and regeneration in vivo. Calpain-6 was predominantly expressed in embryonic muscles and bone cartilage. Capn6 knock-out promoted skeletal-muscle development as a result of increased progenitor cell differentiation. Furthermore, calpain-6 was detected in regenerating skeletal muscles, and suppressed this process. These results indicate that mammalian calpain-6 is a suppressive modulator for skeletal muscle differentiation and growth. Our findings indicate that calpain-6 might serve as a therapeutic target for muscular dystrophy/atrophy or as a useful tool in tissue engineering. For calpain studies, the physiological non-proteolytic role of calpain-6 may shed light on the elusive calpain functions that are independent of their well-studied proteolytic activities. In addition, as this structurally nonproteolytic feature is unique to the calpain-6 of placental mammals among vertebrates, our mutant mice might provide insight into the relationship between molecular and biological evolution.   Table S1 for primers). Neo r , neomycin-resistance gene; pA, poly A tail; TK, thymidine kinase; A, ApaI; B, BamHI; E, EcoRI. (B) Southern blot (left) and PCR (right) analyses of genomic DNA extracted from mouse tails. The bands of the Southern blot represent ApaI/BamHI-or EcoRI-digested genomic DNA from wild-type and Capn6 lacZ/Y mice, probed with the 59-or 39probe. (C-E) Western blot analyses confirming the absence of CAPN6 in Capn6-deficient mice. Total lysates were obtained from E10.5, E12.5, and E14.5 whole embryos (C), the tongue of E14.5, E16.5, and E18.5 embryos (D), and limb buds of E13.5 embryos (E). A band of around 70 kDa (indicated by arrows) was diminished in the Capn6-deficient embryos and was presumed to be CAPN6, and all the other bands were considered to be nonspecific. Wt, Capn6 +/Y (=); Hemi, Capn6 lacZ/Y (=); Het, Capn6 lacZ/+ (R); Ho, Capn6 lacZ/lacZ (R). doi:10.1371/journal.pgen.1003668.g001 bud lysates showed that the expressions of desmin and asarcomeric actin increased with development, and that they were stronger in Capn6 lacZ/Y embryos than in their Capn6 +/Y littermates at E16.5 and E18.5 ( Figure 3F). These results indicated that the deletion of CAPN6 advances the progression of skeletal muscle development during embryogenesis. Lack of CAPN6 promotes the differentiation of skeletal muscle progenitor cells To examine how the loss of CAPN6 promotes skeletal muscle development, we observed the differentiation of primary cultured skeletal muscle (skm-primary) cells from 7-week (wk) old Capn6 +/Y and Capn6 lacZ/Y littermates. The Capn6 +/Y skm-primary cells were induced to differentiate into myotubes with horse-serum-containing medium. In these cultures, Capn6 mRNA was detected in the differentiating myocytes as well as in undifferentiated skm-primary cells ( Figure 4A). The CAPN6 protein was detected in the   Figure 4B). Its expression was upregulated after the induction of differentiation, and then gradually decreased.
Next, the differentiation of skm-primary cells was compared between Capn6 +/Y and Capn6 lacZ/Y littermates. DAPI staining demonstrated that the average number of nuclei in each cell derived from the Capn6 lacZ/Y mice was greater than in their Capn6 +/Y littermates 5-9 days after induction ( Figure 4C and D). These results indicate that, in myotube differentiation, the skm-primary cells from Capn6 lacZ/Y mice underwent fusion more rapidly than did those from Capn6 +/Y mice. In other words, the disruption of CAPN6 promotes the differentiation of myocytes to myotubes, suggesting that CAPN6 acts as a suppressor of skeletal muscle differentiation.

CAPN6 is involved in muscle regeneration
The above experiments using skm-primary cells suggested that CAPN6 might be involved in skeletal muscle regeneration as well as its development. To examine skeletal muscle regeneration in adult mice, cardiotoxin was injected into the mouse tibialis anterior (TA) muscle. Cardiotoxin is a snake venom that selectively causes myofiber degeneration, but leaves nerves, blood vessels, and satellite cells morphologically intact [33]. Two days after the cardiotoxin injection, normal myofibers distinguished by marginal nuclei (shown as in the controls in Figure 5C) were diminished (data not shown). Then, 4d after the injection, the expression of skeletal muscle embryonic and skeletal muscle perinatal myosin heavy chain genes (Myh3 and Myh8, respectively, in Figure 5A) and the cells' central nucleation ( Figure 5C) showed that the degenerated TA skeletal muscle had begun to regenerate.
During this process, we examined Capn6 expression by RT-PCR, using PBS-injected TA muscle as a control. Capn6 mRNA was detected after 4d, similar to Myh3 and Myh8, when muscle regeneration was taking place ( Figure 5A). The CAPN6 protein was also detected in the regenerating muscle by western blot analysis ( Figure 5B). Notably, histological analysis showed more nuclei in each myofiber in the regenerating skeletal muscles of the Capn6 lacZ/lacZ mice than in the Capn6 lacZ/+ mice ( Figure 5C and D). Furthermore, the cross-sectional area (CSA) of the regenerating skeletal muscles of the Capn6 lacZ/lacZ mice was significantly larger than that of the Capn6 lacZ/+ mice ( Figure 5C and E). Consistent with these observations, the expression of MYH8 was also higher at 7 days after cardiotoxin injection in the regenerating Capn6 lacZ/Y muscles than the Capn6 +/Y muscles ( Figure 5F). To our surprise, the expression of CAPN1 was reciprocally suppressed with significance (P = 0.019; n = 3). These results indicated that the loss of CAPN6 promotes both the regeneration of skeletal muscle and its development.

Discussion
In this study, we have shown a physiological function for CAPN6 for the first time: CAPN6 is involved in the suppressive modulation of skeletal muscle development and regeneration. This is the second reported calpain gene (after CAPN3/Capn3) to be genetically shown to be responsible for a skeletal muscle phenotype [12,13,34].
What is the physiological relevance of CAPN6's function as a skeletal muscle growth suppressor like myostatin and muscle RING-finger proteins (MuRFs) [35][36][37]? One possible function of CAPN6 is to restrict the embryo size of eutherians, because an embryo that is too large could harm the mother's body (known as ''maternal-fetal conflict'' [38]). This idea is compatible with the uniquely eutherian non-proteolytic character of CAPN6. During embryogenesis, CAPN6 is expressed not only in skeletal muscles but also in cartilage, the mandibular arch, somites, and the myocardium from early (E9.5) to late (E18.5) stages. Therefore, it is likely that CAPN6 also functions as a growth suppressor in tissues other than skeletal muscles. The role of CAPN6, however, must be accessory and/or stress-responsive, because the CAPN6 deficiency caused milder phenotype in the skeletal muscle systems under normal conditions than those of myostatin-or MuRFsdeficient mice [36,39].
What is the molecular mechanism underlying CAPN6's action? We previously showed that CAPN6 modulates lamellipodial formation and cell motility by stabilizing microtubules and regulating Rac1 activity [28,29]. Microtubule organization and the activity of small G-proteins are also crucial regulators of skeletal myogenesis [40,41]. Thus, our collective results suggest that CAPN6 suppresses skeletal muscle differentiation and growth by modulating basal cellular functions such as cell division, growth, and cell migration through cytoskeletal reorganization. Molecules and a mechanism involved in these phenomena, however, may be different between skeletal muscles and cultured fibroblasts. It has been discussed that CAPN6 may antagonize other calpains to protect their substrates from cleavage [23,27], as in the case of cFLIP, which is a protease-deficient caspase homolog that functions as an apoptosis inhibitor [42]. These points should be clarified in detail in future studies.
The changes in the level of another muscle-differentiationrelated calpain, CAPN1, during TA muscle degeneration and regeneration may provide another hint ( Figure 5F). It was reported that overexpressed CAPN1 suppresses muscle cell differentiation during myogenesis [43]. During the cardiotoxin-induced muscle degeneration, CAPN1's expression decreased, and it recovered during the regeneration process (see Figure 5F), but the absence of CAPN6 delayed the recovery of CAPN1 expression in the regenerating skeletal muscle (compare CAPN1 in Figure 5F lanes 6 and 8). Capn1 mRNA expression was not changed during this process ( Figure 5A), suggesting the involvement of a proteolytic mechanism in this alteration of CAPN1 expression. This CAPN1 change was not observed in primary cultured cell differentiation (Figure 4, data not shown), suggesting an unknown mechanism specific to in vivo. At present, however, it cannot be eliminated that the observed CAPN1 change is a result of different muscle conditions caused by Capn6 disruption. A relationship between CAPN1 and CAPN6 is another interesting subject to be addressed.
In addition, Capn6 is reported to be overexpressed in the skeletal muscle of patients with limb-girdle muscular dystrophy type 2A (LGMD2A), which is caused by pathological mutations of CAPN3 [44]. How the physiological role of CAPN6 in skeletal muscle development is related to the pathological condition of muscular dystrophy, where skeletal muscles are in a continuous cycle of cells were established from the skeletal muscles of 7-wk-old Capn6 +/Y and Capn6 lacZ/Y mice, and incubated in horse serum-containing medium for the indicated number of days to induce myotube differentiation. The nuclei and actin filaments of differentiated cells were visualized with DAPI (blue) and rhodamine-labeled phalloidin (red), respectively.  degeneration and regeneration, is an intriguing question. In this respect, CAPN6 could be a therapeutic target for muscular dystrophies, since its downregulation is expected to enhance the de novo formation of functional skeletal muscles. In terms of tissue engineering, repressing CAPN6 could be helpful for generating skeletal muscle from embryonic or induced pluripotent stem cells in vitro.
Despite the predominant expression of CAPN6 during embryogenesis, however, Capn6-deficient mice were born and grew healthily, and showed slightly but significantly higher skeletal muscle and body weight (see Figure S3). It is possible that Capn6 deficiency did not cause severe phenotypes such as lethality or dysgenesis because of a redundancy of CAPN6's non-proteolytic function among calpains. Indeed, recent genetic studies have suggested that calpain family members have non-proteolytic roles in addition to their proteolytic functions [14,34,45]. These nonproteolytic functions are ascribed to the CysPc domain, the other domains, or both, and may include the structural regulation of other molecules and/or a chaperone function. Some calpains of schistosomes, insects, and nematodes, and all of the calpains of Trypanosoma are non-proteolytic members with substitutions in one or more of the well-conserved active-site triad residues, as seen in eutherian CAPN6 [46][47][48][49].
A recent bioinformatics analysis revealed that non-classical calpains are considered as the majority among all eukaryotes, and that classical calpains are relatively new in the evolution of calpains, being confined to a small subset of eukaryotes [49]. As mentioned, non-classical calpains contain various non-proteolytic calpains. These findings highlight importance of the calpains' nonproteolytic functions, which are considered to be rather recent functional modifications evolved during calpains' diversification. Although calpains' proteolytic activities are important for understanding many of their physiological functions, recent reports on the non-proteolytic functions of calpains, including this one, shed light on their more elusive physiological roles, and hence may lead to important breakthroughs in biological research and medicine.

Experimental animals
All procedures using experimental animals were approved by the Experimental Animal Care and Use Committee of Tokyo Metropolitan Institute of Medical Science, and the animals were treated according to the committee's guidelines. C57BL/6 and ICR (a strain established in The Institute of Cancer Research in London, United Kingdom) mice were purchased from Nihon CLEA Inc. For embryos, mice were mated in the evening and if a vaginal plug was observed the next morning, fertilization was assumed to have taken place at midnight. Pregnant mice were sacrificed by cervical dislocation, and the embryos or fetuses were dissected from the uterine decidua.

Generation of Capn6-targeted embryonic stem cells and mutant mice
Genomic clones for mouse Capn6 were obtained by screening the C57BL6/J-derived BAC libraries from the BACPAC Resource Center. The nls-lacZ cassette was made using the lacZ gene with a nuclear localization signal (nls-lacZ) and flanked with lox71 at the 59-end and lox2272 at the 39-end, to allow recombinase-mediated cassette exchange. A pKO Scrambler NTKV-1904 plasmid (Stratagene) was used as the backbone vector. For the targeting construct, a PCR-amplified 1.1-kb fragment and a 5.8-kb NotI-SalI fragment from intron 1 were placed on each side of the nls-lacZ cassette. The targeting vector was linearized using NotI and electroporated into the F1 hybrid ES cell line G4 [50]. Clones surviving positive-negative selection with neomycin and FIAU were screened for homologous recombination with diagnostic PCR primers (see Table S1). Of 292 ES cell clones screened, five correctly targeted clones were identified by PCR and confirmed by Southern blot analysis. Three independent ES cell clones hemizygous for the disrupted allele were injected into ICR blastocysts to generate germline chimeras. Female mice homozygous for the Capn6 lacZ allele were obtained by intercrossing F1 hemizygous males (Capn6 lacZ/Y ) and heterozygous females (Capn6 lacZ/+ ). The genotypes of offspring were determined by PCR or Southern blot analysis of tail-tip DNA ( Figure 1B). After backcrossing progeny with C57BL6/J mice more than 10 times, the mice were maintained on the B6 genetic background.

Western blotting
To prepare protein lysates, tissues, and cultured cells were homogenized with TED buffer (10 mM Tris/Cl [pH 7.5], 1 mM EDTA, 1 mM DTT) containing 2% SDS and protease inhibitors, as described above. The protein concentration was determined using a DCA protein assay kit (BioRad), and samples containing equal amounts of protein were separated by 7.5 or 10% SDSpolyacrylamide gel electrophoresis (PAGE) and then electrotrans-the primers that detect full-length Capn6 mRNA. Myh3 and Myh8 mRNA was amplified to confirm that skeletal muscles successfully regenerated after their cardiotoxin-induced degeneration; Capn1 mRNA was detected as an internal control. (B) Western blot analysis using antibodies against CAPN6, of skeletal muscle lysates of 7-wk-old Capn6 lacZ/+ and Capn6 lacZ/lacZ mice, 5 days after cardiotoxin or PBS (control) injection. The specific CAPN6 band, indicated by an arrow, was detected around 70 kDa, and was not detected in Capn6 lacZ/lacZ mice. (C-E) Histological analysis of the TA muscles of 7wk-old Capn6 lacZ/+ and Capn6 lacZ/lacZ mice after a cardiotoxin or PBS (control) injection. Regenerating skeletal muscles with central nuclei were observed 4 d after the cardiotoxin injection. Remarkable degeneration was not observed in TA muscles at 4d after a PBS injection (shown as control) (C). Scale bar: 50 mm. Seven days after the injection, the average number of nuclei per cell was significantly greater in Capn6 lacZ/lacZ (1.43 [mean]60.06 [s.e.m.]) than in Capn6 lacZ/+ (1.2160.03) (D), and the mean CSA of the regenerating muscle myofibers was also significantly greater in Capn6 lacZ/lacZ (1608.0 [mean]678.5 [s.e.m.] mm) than in Capn6 lacZ/+ (938.0667.2 mm) (E). *, P,0.05; **, P,0.01 by Student's t-test. Each value is the mean from five experiments, and 50 cells were counted in each experiment. (F) Western blot analysis with skeletal muscle lysates of 7-wk-old Capn6 lacZ/Y and Capn6 +/Y mice 2d and 7 d after cardiotoxin or PBS (control) injection using antibodies against MYH3, MYH8 and CAPN1. CBB staining was used as an internal control. Wt, Capn6 +/Y (=); Hemi, Capn6 lacZ/Y (=); Het, Capn6 lacZ/+ (R); Ho, Capn6 lacZ/lacZ (R). doi:10.1371/journal.pgen.1003668.g005 ferred to a polyvinylidene difluoride (PVDF) membrane. After being blocked with 5% skim milk in 0.1% Tween 20 in Trisbuffered saline, pH 7.6, the blot was probed with primary antibody. The membrane was then washed with 0.1% Tween 20 in Tris-buffered saline, pH 7.6, and incubated with peroxidaseconjugated anti-rabbit or anti-mouse immunoglobulin G (Dako). The signals were detected using the ECL chemiluminescence detection system (Amersham Bioscience).

b-Galactosidase staining
LacZ expression in Capn6 mutant mice was detected by staining with X-gal (5-bromo-4-chloro-3-indoyl-b-D-galactoside). Whole embryos were isolated in ice-cold KPP buffer (0.1 M potassium phosphate), then fixed with 4% PFA/0.1% glutaraldehyde in KPP buffer, containing 5 mM EGTA and 2 mM MgCl 2 . The embryos were then rinsed 3 times with wash buffer (0.01% Na deoxycholate and 0.02% Nonidet P-40 in KPP buffer containing 5 mM EGTA and 2 mM MgCl2). Samples for histological analysis were embedded in OTC compound and cryosectioned. The embryos or sections were incubated overnight at 37uC in lacZ-staining buffer (10 mM potassium ferrocyanide, 10 mM potassium ferricyanide, and 2 mg/ml X-gal in wash buffer).

Skeletal muscle primary cell cultures
Mouse skeletal muscle primary cells were prepared as previously described [51], and maintained in F-10 medium containing 10% FBS (fetal bovine serum) and 5 mg/ml FGF-8. To induce muscle differentiation, the medium was switched to differentiation medium (5% horse serum [Invitrogen] and 0.2 M ascorbic acid in Dulbecco's modified Eagle's medium) for further culture. All media were supplemented with 100 units/ml penicillin, 0.1 mg/ml streptomycin, and 2 mM L-glutamine (Invitrogen). To quantify the average number of nuclei per cell, nuclei and cells were labeled by DAPI and rhodaminephalloidin, respectively, and 100 cells were randomly selected in the fixed area.
Cardiotoxin injection and tissue preparation 0.1 ml of 10 M cardiotoxin (Wako Pure Chemical Industries) in 0.9% saline was injected directly into the left TA muscle with a 27gauge needle under ether anesthesia. Mice were killed by cervical dislocation, and the cardiotoxin-injected TA muscles (left) and PBS-injected contralateral TA muscles (right) were removed for analysis. Several of the muscles were frozen in isopentane cooled by liquid nitrogen for histological analysis, and the other muscles were frozen directly in liquid nitrogen for RNA and Protein extraction, and stored at 280uC.

RT-PCR
Total RNA was extracted using TRIzol reagent (Invitrogen). The reverse transcription reaction was carried out using the First Strand cDNA Synthesis kit (Amersham Bioscience) according to the manufacturer's instructions. The resultant cDNAs were amplified with ExTaq polymerase (Takara) in a thermocycler. The amplified mouse Capn1 and Gapdh were used as internal controls.

H-E staining and CSA analysis of skeletal muscle
For H-E staining, sections were stained with Mayer's hematoxylin, followed by eosin staining. H-E-stained sections were mounted with Mount-Quick (Daido Sangyo), and viewed with a BX60 microscope (Olympus). The CSA of H-E-stained myofiber sections was determined with NIS-Elements Ar 3.0 software (Nikon).

Statistical analysis
Student's t-test was used to test whether or not the means of two normally distributed populations were equal. P values less than 0.05 were considered statistically significant.