R3hdml regulates satellite cell proliferation and differentiation

Abstract

In this study, we identified a previously uncharacterized skeletal satellite cell‐secreted protein, R3h domain containing‐like (R3hdml). Expression of R3hdml increases during skeletal muscle development and differentiation in mice. Body weight and skeletal muscle mass of R3hdml knockout (KO) mice are lower compared to control mice. Expression levels of cell cycle‐related markers, phosphorylation of Akt, and expression of insulin‐like growth factor within the skeletal muscle are reduced in R3hdml KO mice compared to control mice. Expression of R3hdml increases during muscle regeneration in response to cardiotoxin (CTX)‐induced muscle injury. Recovery of handgrip strength after CTX injection was significantly impaired in R3hdml KO mice, which is rescued by R3hdml. Our results indicate that R3hdml is required for skeletal muscle development, regeneration, and, in particular, satellite cell proliferation and differentiation.

Introduction

Other than adipose tissue, skeletal muscle is the most abundant tissue in our bodies, constituting approximately 50% of adipose tissue‐free weight in humans 1. Skeletal muscles also function as the largest metabolic organ in the body, consuming the majority of blood glucose 2.

Skeletal muscles are mainly composed of myofibers. During embryonic and fetal myogenesis, myogenic progenitors (myoblasts) expressing paired box (Pax) 3/7 and MyoD proliferate and fuse together to form myofibers. A subpopulation of myoblasts loses MyoD and paired Pax3 expression and become Pax7⁺ quiescent satellite cells, which are the resident muscle stem cells 3.

Satellite cells become important for the rapid growth of skeletal muscle by the perinatal stage. Skeletal muscle mass increases several fold in the first few weeks after birth. During this time, satellite cells supply new myonuclei, leading to myofiber hypertrophy 4.

Adult skeletal muscle is stable with a slow turnover rate under physiological conditions but with high regenerative capacity. Satellite cells remain mostly quiescent, which is characterized by high levels of Pax‐7 expression 5. After muscle injury, satellite cells are activated, enter the cell cycle, and rapidly proliferate. Most of these cells differentiate and fuse to regenerate myofibers, whereas a minority self‐renew and return to quiescence to replenish the satellite cell pool 6. The deletion of inducible Pax7, a gene expressed in satellite cells, has been reported to lead to disturbance of skeletal muscle regeneration 7. Therefore, maintenance of satellite cell number is important for muscle development and regeneration.

In this paper, we present the discovery of a novel gene expressed in myogenic satellite cells termed R3h domain containing‐like (R3hdml). R3hdml was originally identified as a glomerular podocyte‐expressed gene through microarray‐based genomic screening 8. R3hdml expression was found increased in skeletal muscle not only during development but also following muscle injury. In our current findings, R3hdml protein was secreted upon satellite cell differentiation. Inactivation of R3hdml led to a reduced amount of skeletal muscle and decreased proliferation of myogenic satellite cells. Overall, our results indicate that R3hdml is important for skeletal muscle development, regeneration, and, in particular, satellite cell proliferation and differentiation.

Results

R3hdml is expressed during skeletal muscle development and satellite cell differentiation

First, we examined the expression of R3hdml during skeletal muscle development. Lower leg skeletal muscles were dissected from wild‐type mice on post‐natal days 0, 7, and 21. As shown in Fig 1A and B, R3hdml mRNA levels were highest on post‐natal day 0, after which the levels decreased by days 7 and 21 after birth. This suggests that R3hdml may be important for skeletal muscle development. During development, satellite cells proliferate and differentiate actively; however, satellite cells enter a quiescent state by post‐natal day 21 in vivo. Therefore, R3hdml may be expressed transiently in activated satellite cells and downregulated after their return to the quiescent state; meanwhile, R3hdml expression is turned off in fully differentiated muscle cells (myofibers). To further investigate this, we isolated satellite cells from adult mice and differentiated them to myotubes in vitro (Fig 1C). As shown in Fig 1D and E, R3hdml expression was observed in satellite cells, indicating that R3hdml is a novel satellite cell‐expressed protein and that R3hdml levels increased in conjunction with satellite cell differentiation. The murine muscle cell line C2C12 was also used to determine that the presence of R3hdml, both at the mRNA and protein level, was increased during C2C12 cell differentiation (Appendix Fig S1A–D). Because it is not possible to induce cultured satellite cells and C2C12 cells into fully differentiated skeletal muscles in vitro, it may be possible that R3hdml continued to be expressed during the differentiation stages of cell culture. Nevertheless, our results indicate that R3hdml expression increased during satellite cell differentiation and proliferation.

Figure 1. Expression of R3hdml during skeletal muscle development and satellite cell differentiation.

• A
  Skeletal muscles in the lower legs were dissected, and RNA was extracted on post‐natal days 0, 7, and 21. The expression of R3hdml was examined by RT–PCR. GAPDH was used as an internal control. MM, 100 bp DNA marker.
• B
  Quantification of RT–PCR results. Data are expressed as the means ± standard error of the mean (SEM). The experiments were repeated at least three times. ****P < 0.0001 versus day 0 group; one‐way ANOVA followed by Bonferroni post hoc test.
• C–E
  Satellite cells were isolated from skeletal muscles of adult mice by enzymatic digestion and placed in culture medium. Then, differentiation was induced in satellite cells (C). The proteins were extracted at the indicated time points. R3hdml expression was examined by Western blotting using an anti‐R3hdml antibody (D). β‐actin was used as internal control. Quantification of Western blotting results (E). Data are expressed as the means ± standard error of the mean (SEM). The experiments were repeated at least three times. Representative data are shown. *P < 0.05; Student's t‐test. Scale bar = 500 μm.

R3hdml may be a novel protein secreted from satellite cells

R3hdml protein is predicted to have a signal peptide sequence at the N‐terminus and to possess a putative furin cleavage site, indicating that R3hdml is a secreted protein (Fig EV1A; highlighted in blue). We examined R3hdml protein expression in cell lysates and conditioned media during satellite cell differentiation. The predicted R3hdml protein consists of 253 amino acids, corresponding to a protein of around 28 kDa. As shown in Fig EV1B, we detected R3hdml protein in differentiated myotube lysates, while a protein of around 21 kDa was observed in conditioned medium. To analyze the amino acid sequence of R3hdml protein secreted into the medium, R3hdml was overexpressed in Ad293 cells. Then, the conditioned medium of R3hdml‐overexpressed cells was collected, and mass spectrometry analysis of the 21 kDa protein revealed that the secreted protein lacks the putative furin cleavage sequences, indicating that R3hdml was cleaved at the furin cleavage site prior to secretion into the medium (Fig EV1A; highlighted in red).

Figure EV1. R3hdml protein is secreted from satellite cells.

1. The predicted amino acid sequence of R3hdml. RHRR motifs, which are highlighted in blue, are putative furin cleavage sequences. Conditioned medium from Ad293 cells with or without R3hdml overexpression was separated by agarose gel electrophoresis followed by Coomassie Brilliant Blue staining. A protein signal corresponding to 21 kDa was present under R3hdml overexpression but not without; this protein was cut out from the gel and subjected to mass spectrometric analysis. The amino acid sequence of the secreted protein is highlighted in red.
2. Satellite cells were isolated from wild‐type control mice and differentiated in the presence of differentiation induction medium. Cell lysates and conditioned medium were obtained before and 5 days after differentiation and subjected to SDS–PAGE followed by immunoblotting using anti‐R3hdml antibody.

MyoD, a crucial transcriptional factor for skeletal muscle development, regulates R3hdml expression

The expression of other myogenic factors such as Pax7, MyoD, and myogenin during C2C12 cell differentiation was examined. Pax7 and MyoD, markers associated with early skeletal muscle differentiation, were expressed prior to R3hdml (Appendix Fig S1C and D). Therefore, we examined the 5ʹ flanking sequence upstream of the translation site of the R3hdml gene in silico to identify promoter sequences and identified five putative MyoD binding sites within 781 bp upstream (Appendix Fig S2A and B). This 781‐bp 5ʹ flanking sequence from the murine genome was subcloned into a luciferase reporter vector and transferred into C2C12 cells, and then, the cells were differentiated into myotubes. As shown in Fig 2A, R3hdml promoter activity increased during C2C12 cell differentiation. We also constructed luciferase constructs with or without a putative MyoD binding site (Fig 2B). When the MyoD gene was overexpressed in C2C12 cells (Fig 2C), R3hdml promoter activity increased, even when not in the presence of differentiation medium (Fig 2D). On the other hand, luciferase activity was significantly attenuated when the MyoD high‐affinity binding site was lacking (Construct A), indicating that R3hdml expression during development is regulated in the presence of MyoD but independently of the degree of differentiation.

Figure 2. Expression of R3hdml is controlled by the expression of MyoD.

1. Cultured C2C12 cells were transfected with the firefly luciferase construct containing the 781‐bp fragment upstream of the translational start codon of the R3hdml coding sequence (R3hdml‐luc), together with Renilla luciferase vectors as controls; the cells were then differentiated in the presence of 2% horse serum. Cells were harvested at the indicated time points, after which luciferase activity of the cell lysate was measured.
2. The table represents the schematic of each construct. All inserts in the constructs contain the same 3ʹ end. The 781‐bp fragment of the 5ʹ flanking sequence from the murine genome, which contains several putative MyoD binding sites (black box), was subcloned into a luciferase reporter vector (Construct C). We also subcloned the genomic sequence containing one high‐affinity MyoD binding site CACCTG (Construct B) and one without a high‐affinity MyoD binding site (Construct A) into a luciferase reporter vector.
3. Plasmids containing MyoD cDNA or empty vector were transfected into C2C12 cells. One day and 3 days after transfection, cells were lysed and subjected to SDS–PAGE and Western blotting using anti‐MyoD antibodies.
4. Plasmids containing MyoD cDNA or empty vector together with R3hdml‐luc and Renilla luciferase vectors were transfected into C2C12 cells. One day and 3 days after transfection, the luciferase activity of the cell lysate with or without MyoD overexpression was measured. Relative luciferase activity is expressed as the fold activity above the background conferred by the empty vector together with R3hdml‐luc.

Data information: Data are expressed as the means ± standard error of the mean (SEM). The experiments were repeated at least six times. **P < 0.01; ****P < 0.0001 (versus day 0 group in A); one‐way ANOVA followed by Bonferroni post hoc test.

R3hdml functions in myotubule formation and skeletal muscle development

As the expression of R3hdml increased in developing skeletal muscle, we speculated that R3hdml has roles in skeletal muscle development. To investigate this, we first silenced the R3hdml gene in C2C12 cells using siRNA (Fig 3A), which was followed by the induction of myotube formation. We stained the myotubes with anti‐myosin heavy chain antibodies and counted the number of nuclei in each myotube. Fewer multinuclear myotube formation events were observed in the absence of R3hdml, and these changes were reversed by adding purified R3hdml protein (50 ng/μl; Fig 3B–D). These findings indicate that R3hdml is essential for myoblast differentiation.

Figure 3. Silencing the R3hdml gene ameliorates C2C12 cell differentiation.

1. R3hdml siRNA or non‐targeting siRNA was transfected into C2C12 cells. Then, C2C12 cells were differentiated into myotubes. RNA was extracted from the cells after induction of differentiation at the indicated time points. Closed bars, control; open bars, silenced R3hdml gene; **P < 0.01; *P < 0.05; Student's t‐test.
2. Immunofluorescence after C2C12 differentiation. MyHC is shown in green and the nuclei in blue (counterstained with Hoechst). Representative images of MyHC immunofluorescence of C2C12 myocytes transfected with the siRNA or after addition of R3hdml protein. Scale bar = 100 μm.
3. The number of nuclei per MyHC⁺ cell was calculated on day 5 after differentiation. **P < 0.01; one‐way ANOVA followed by Bonferroni post hoc test.
4. The distribution of nuclei per myotube was calculated on day 5 after differentiation. **P < 0.01; *P < 0.05; one‐way ANOVA followed by Bonferroni post hoc test.

Data information: Data are expressed as the means ± standard error of the mean (SEM). The experiments were repeated at least three times, and at least three fields of view were evaluated.

R3hdml knockout (KO) mice exhibit disorganized skeletal muscle development

R3hdml KO mice were created by homologous recombination. R3hdml KO mice developed normally, were fertile, and did not exhibit gross developmental defects; however, body weight was significantly lower (91.3%, P < 0.001) than that of littermate controls at 20 weeks of age (Fig EV2A). While fat mass measurements using dual‐energy X‐ray absorptiometry and femoral length indicated that the amount of adipose tissue and body length was unchanged in R3hdml KO mice (Appendix Fig S3A–C), leg muscle weight was significantly lower (70.7%) than that of the controls at 20 weeks of age (Fig EV2B). The diameter of the rectus femoris, dissected from mice on post‐natal day 0, was examined histologically and found to be significantly shorter (68.7%) in KO mice, while the femur diameter was not significantly different from that of wild‐type controls (Appendix Fig S3D–F).

Figure EV2. In R3hdml KO mice, body weight and skeletal muscles of the lower legs are lighter than those of the wild‐type control mice.

• A
  Weight of R3hdml KO (n = 10) and wild‐type control (n = 10) mice was evaluated starting from 6 weeks after birth until 20 weeks at 2‐week intervals. Data are expressed as the means ± standard error of the mean (SEM). **P < 0.01; ****P < 0.0001; two‐way ANOVA followed by a Bonferroni post hoc comparison of the individual time points.
• B
  Skeletal muscles from the lower legs were dissected from R3hdml KO mice (n = 10) and wild‐type controls (n = 10) and weighed. The weight of skeletal muscle was compared with the whole‐body weight. Closed columns, wild‐type controls; open columns, R3hdml KO mice; quad, quadriceps; TA, tibialis anterior; GA, gastrocnemius. Data are expressed as the means ± standard error of the mean (SEM). *P < 0.05; **P < 0.01; Student's t‐test.
• C, D
  Skeletal muscles were dissected from post‐natal day 0 (C) and 3‐month‐old R3hdml KO (n = 4) and wild‐type control (n = 4) mice (D). Transmission electron microscopy analysis was performed. Arrowheads indicate Z bands. Scale bar in low magnification images = 2 μm; in high magnification images = 500 nm.

Electron microscopy revealed that R3hdml KO mice exhibited diffuse Z lines and shorter and denser sarcomeres than those of wild‐type control mice on post‐natal day 0. It was also difficult to separate A bands from I bands, and the alignment of myosin and actin filaments was disorganized in R3hdml KO mice compared with wild‐type control mice on post‐natal day 0 (Fig EV2C). R3hdml KO in 3‐month‐old mice also exhibited diffuse Z lines and disorganized alignment of myosin and actin filaments compared with those of wild‐type control mice (Fig EV2D). The frequency of occurrence of muscle fiber disorganization was about one per 12‐mm² muscle area in KO mice. On the other hand, the shorter and denser sarcomeres that were seen in post‐natal day 0 R3hdml KO mice were not observed in 3‐month‐old mice, indicating that there is some compensatory mechanism (Fig EV2D).

Expression of skeletal muscle cell development and proliferation‐related markers was disturbed in R3hdml KO mice

Because the lack of R3hdml appeared to affect skeletal muscle development, we examined known factors required for this process and found that levels of Pax7, Myod1, and Myogenin (Myog) were decreased in R3hdml KO mice compared to those in littermate controls (Fig EV3A). We also examined the global transcriptional landscape using mRNA extracted from both R3hdml KO and wild‐type mouse skeletal muscle dissected from post‐natal day 0 mice. RNA sequencing indicated that, in total, 659 genes were differentially expressed. Among those, 142 genes were significantly downregulated in R3hdml KO compared with wild‐type mice (Table 1). Notably, the expression of genes involved in cell cycle regulation was lower in skeletal muscles from R3hdml KO mice than in skeletal muscles from wild‐type controls (Fig EV3B). We then evaluated the expression of Mki67 (proliferation marker), Cdkn1a (p21; proliferation inhibition marker), and E2f1 and Ccnd1 (CyclinD1; important G1 to S transition markers) and found that Mki67, E2f1, and CyclinD1 levels were significantly decreased (72.5, 34.6, and 39.0%, respectively), whereas p21 expression was significantly increased (124.0%, Fig EV3C).

Figure EV3. Gene expression profiles in skeletal muscles.

1. Skeletal muscles of the lower legs were dissected from R3hdml KO (n = 5) and wild‐type control (n = 9) mice. Known myogenic factors were evaluated by real‐time PCR. Closed columns, wild‐type controls; open columns, R3hdml KO mice. Data are expressed as the means ± standard error of the mean (SEM). *P < 0.05; **P < 0.01; Student's t‐test.
2. Total RNA was extracted from neonatal skeletal muscles for RNA sequencing, followed by gene set enrichment analysis for the KEGG‐derived cell cycle pathway in R3hdml KO and wild‐type control mice. FC W/KO, fold change WT/R3hdml KO. Normal P‐value 0.00, FDR q‐value 0.00.
3. RNA sequencing results evaluated by real‐time PCR. Mki67, a marker for proliferation; Cdkn1a, a marker for inhibition of proliferation; E2f1 and Ccnd1, markers important in the G1 to S transition. Closed columns, wild‐type controls; open columns, R3hdml KO mice. Data are expressed as the means ± standard error of the mean (SEM). *P < 0.05; **P < 0.01; ***P < 0.001; Student's t‐test.

Table 1.

Global transcriptional screening using mRNA extracted from R3hdml KO and wild‐type skeletal muscle by RNA sequencing

Gene sets	NES	FDR q‐value
Gene set enriched in WT
REACTOME_CELL_CYCLE_MITOTIC	−2.37	0
REACTOME_DNA_REPLICATION	−2.35	0
REACTOME_MITOTIC_M_M_G1_PHASES	−2.31	0
REACTOME_DNA_STRAND_ELONGATION	−2.3	0
KEGG_DNA_REPLICATION	−2.25	0
REACTOME_ACTIVATION_OF_THE_PRE_REPLICATIVE_COMPLEX	−2.23	0
REACTOME_E2F_MEDIATED_REGULATION_OF_DNA_REPLICATION	−2.21	0
REACTOME_G2_M_CHECKPOINTS	−2.21	0
REACTOME_MITOTIC_G1_G1_S_PHASES	−2.19	0
REACTOME_ACTIVATION_OF_ATR_IN_RESPONSE_TO_REPLICATION_STRESS	−2.17	0
Gene set enriched in R3hdml
REACTOME_PEPTIDE_CHAIN_ELONGATION	2.73	0
KEGG_RIBOSOME	2.65	0
REACTOME_3_UTR_MEDIATED_TRANSLATIONAL_REGULATION	2.63	0
REACTOME_INFLUENZA_VIRAL_RNA_TRANSCRIPTION_AND_REPLICATION	2.47	0
REACTOME_FORMATION_OF_THE_TERNARY_COMPLEX_	2.44	0
REACTOME_NONSENSE_MEDIATED_DECAY_ENHANCED_BY_THE_EXON_JUNCTION_COMPLEX	2.41	0
REACTOME_ACTIVATION_OF_THE_MRNA_UPON_BINDING_OF_THE_CAP_BINDING_COMPLEX_	2.31	0
REACTOME_SRP_DEPENDENT_COTRANSLATIONAL_PROTEIN	2.18	0
REACTOME_RESPIRATORY_ELECTRON_TRANSPORT_ATP_SYNTHESIS	2.16	0
REACTOME_TRANSLATION	2.12	0

Subsequently, satellite cells were isolated both from wild‐type control and R3hdml KO mice and cultured in dishes in vitro to examine their proliferative activities. Ki‐67 staining revealed that satellite cells from R3hdml KO mice exhibited decreased (by 17.1%) proliferative activity (P < 0.05, Fig 4A and B). However, the number of satellite cells per myofiber was not significantly different between wild‐type controls and R3hdml KO mice, which may be due to compensatory mechanisms.

Figure 4. R3hdml KO mice‐derived satellite cells show reduced proliferation and differentiation compared with those from wild‐type controls.

• A
  Satellite cells were isolated by enzymatic digestion from R3hdml KO (n = 6) and wild‐type control (n = 6) mice and placed in culture medium. Then, cell proliferation was evaluated by Ki67 staining. Cell nuclei were stained with DAPI.
• B
  The ratio of Ki67‐positive cells to total cells (DAPI‐positive) was evaluated. The number of cells was counted under approximately 60 fields of view for each well. Closed columns, wild‐type controls; open columns, R3hdml KO mice. *P < 0.05; Student's t‐test.
• C–F
  Thirty freshly isolated myofibers from 8‐ to 12‐week‐old R3hdml KO (n = 7) and wild‐type mice (n = 7) were cultured for 72 h. Then, the isolated myofibers were fixed with 4% PFA followed by immunostaining using anti‐Pax‐7, anti‐MyoD, and anti‐Myog antibodies. (C, D) Co‐immunostaining parallel cultures for Pax7 and MyoD showed that there were fewer cells with the self‐renewing Pax7⁺/MyoD⁻ phenotype in muscles from R3hdml KO mice compared with control: Pax7⁺/MyoD⁻ cells (arrowheads), Pax7⁻/MyoD⁺ cells (yellow arrows), and Pax7⁺/MyoD⁺ cells (white arrows). *P < 0.05; Student's t‐test. (E, F) Co‐immunostaining parallel cultures for MyoD and Myog showed that there were fewer cells with the differentiating MyoD⁻/Myog⁺ phenotype in muscles from R3hdml KO mice compared to control: MyoD⁻/Myog⁺ cells (arrowheads), MyoD⁻/Myog⁺ cells (yellow arrows), and MyoD⁺/Myog⁺ cells (white arrows). *P < 0.05; Student's t‐test.

Data information: Data are expressed as the means ± standard error of the mean (SEM).

To examine differentiation efficiency, isolated myofibers from R3hdml KO mice and wild‐type controls were cultured for 72 h and then immunostained with Pax 7/MyoD or MyoD/Myog. In this study, Pax7⁺/MyoD⁻ represents quiescent cells, Pax7⁺/MyoD⁺ represents cells in transit toward differentiation, and Pax7⁻/MyoD⁺ represents activated satellite cells. Moreover, MyoD⁺/Myog⁻ represents activated satellite cells, MyoD⁺/Myog⁺ transient cells, and MyoD⁻/Myog⁺ cells committed to myogenic differentiation. As shown in Fig 4C–F, satellite cells from R3hdml KO mice exhibited 18.8% less Pax7 positivity (Fig 4C and D) and 50.3% fewer Myog‐positive cells (Fig 4E and F), indicating that satellite cells from R3hdml KO mice exhibited reduced self‐renewal and differentiation abilities compared with that of controls.

R3hdml regulates Akt signaling in skeletal muscles by controlling the expression of insulin‐like growth factor 1

The insulin‐like growth factor 1 (IGF‐1) signaling pathway plays an important role in skeletal muscle development. We thus examined the expression of IGF‐1 and the IGF‐1 receptor (IGF‐1R) as well as Akt phosphorylation, which is a downstream target of the IGF‐1/IGF‐1R signaling pathway, as the histological changes in skeletal muscles from R3hdml KO mice observed by electron microscopy demonstrated histological features that resemble a lack of local IGF‐1 production 9. As shown in Fig EV4A and B, IGF‐1 expression decreased by 27.2% in R3hdml KO mice skeletal muscle compared with wild‐type controls; IGF‐1R expression was not significantly different. Furthermore, Akt signaling was decreased in R3hdml KO mice, as evidenced by 49.2% reduced Akt phosphorylation in R3hdml KO skeletal muscle compared with wild‐type controls (Fig EV4C and D). As IGF‐1 was decreased in R3hdml KO mice, we examined the association of R3hdml and IGF‐1 expression. As shown in Fig EV4E–H, silencing R3hdml in C2C12 cells led to reduced IGF‐1 mRNA levels, whereas overexpression of R3hdml in C2C12 cells increased IGF‐1 mRNA expression. These findings indicate that R3hdml regulates Akt signaling in skeletal muscles by directly or indirectly controlling the expression of IGF‐1.

Figure EV4. Akt signaling is attenuated in skeletal muscles of R3hdml KO mice compared with wild‐type control mice.

• A, B
  Skeletal muscles of the lower legs were dissected from R3hdml KO (closed bar, n = 4) and wild‐type control (open bar, n = 4) mice. The expression of insulin‐like growth factor 1 (Igf1) was evaluated by real‐time PCR. Data are expressed as the means ± standard error of the mean (SEM). ***P < 0.001. n.s. not significant; Student's t‐test.
• C, D
  Skeletal muscles of the lower legs were dissected from R3hdml KO (n = 4) and wild‐type control (n = 4) mice, lysed, and subjected to SDS–PAGE, followed by immunoblotting using anti‐Akt and phosphor‐Akt antibodies. Data are expressed as the means ± standard error of the mean (SEM). *P < 0.05; Student's t‐test.
• E–H
  The R3hdml gene was either silenced by siRNA or overexpressed by gene transfer in C2C12 cells, after which the expression of R3hdml (E, G) and Igf1 (F, H) was examined by real‐time PCR. Closed bars, control (n = 3, biological replicates); open bars, silenced R3hdml gene (n = 3, biological replicates); hatched bars, overexpressed R3hdml gene (n = 3, biological replicates). Data are expressed as the means ± standard error of the mean (SEM). *P < 0.05, **P < 0.01; Student's t‐test.

Source data are available online for this figure.

R3hdml plays a role in skeletal muscle regeneration following injury

Due to its apparent role in skeletal muscle development, we further examined the function of R3hdml in skeletal muscle regeneration. We thus injected CTX into skeletal muscle, which destroys skeletal muscles. This is followed by muscle regeneration, which is completed after approximately 1 month. The CTX model is used to estimate skeletal muscle regeneration; therefore, we are able to examine the role of R3hdml in skeletal muscle regeneration. After CTX injection, we first examined R3hdml expression. As shown in Appendix Fig S4A and B, R3hdml mRNA levels increased 154.8‐fold 3 days after CTX injection (P < 0.05). Next, we examined which cell types expressed R3hdml during skeletal muscle injury and found R3hdml mRNA in MyoD‐positive cells, indicating that R3hdml is expressed in activating satellite cells undergoing proliferation and development (Fig 5A).

Figure 5. Expression of R3hdml increases during skeletal muscle regeneration, and R3hdml is important both for skeletal muscle regeneration and function.

1. Seven days after CTX injection, skeletal muscles were dissected, fixed, and embedded in paraffin. Two consecutive sections were utilized, and the first section was subjected to in situ hybridization using an R3hdml sequence‐specific RNA probe to detect Rh3dml, while the second section was used for IHC using the anti‐MyoD antibody. MyoD protein was visualized with DAB stain (brown). R3hdml expression is highly detected in MyoD‐positive cells. Scale bar in low magnification images = 100 μm; in high magnification images = 25 μm.
2. Fifty microliters of 10 μM CTX was intramuscularly injected into the forearm muscle of anesthetized R3hdml KO ([b], n = 10) and wild‐type control ([a], n = 10) mice. Subsequently, handgrip was evaluated 0, 3, 6, and 14 days after CTX injection. R3hdml genes were overexpressed in the forearm muscle of anesthetized R3hdml KO mice ([c], n = 10), and handgrip was also evaluated 0, 3, 6, and 14 days after CTX injection. Data are expressed as the means ± standard error of the mean (SEM). **P < 0.01; ****P < 0.0001 versus wild‐type control; ^# P < 0.05; ^## P < 0.01 versus R3hdml KO mice; two‐way ANOVA followed by a Bonferroni post hoc comparison of the individual time points.
3. Myofibers were freshly dissected from 8‐ to 12‐week‐old R3hdml KO (n = 4) and control mice (n = 4) 35 days after CTX injection, and the number of satellite cells was evaluated by Pax‐7 staining. Arrowheads indicate Pax‐7‐positive cells. The number of Pax‐7‐positive cells per myofiber was counted. At least 50 myofibers in each animal were evaluated. Closed columns, wild‐type controls; open columns, R3hdml KO mice. Data are expressed as the means ± standard error of the mean (SEM). *P < 0.05; Student's t‐test.

CTX was then injected into both R3hdml KO mice and wild‐type controls, after which handgrip strength was examined. As shown in Fig 5B, muscle strength recovery was significantly slower in R3hdml KO mice than in wild‐type controls. The expression of Pax‐7 and the number of satellite cells per myofiber were then evaluated after CTX injection in the forearm; Pax‐7 levels were found significantly increased in wild‐type controls but not in R3hdml KO mice (Appendix Fig S4C). The mean number of satellite cells per myofiber was 20.2 ± 3.8 for wild‐type controls and 9.9 ± 2.27 for R3hdml KO mice (Fig 5C), indicating a reduced number of satellite cells in the skeletal muscle of R3hdml KO mice compared with wild‐type controls following CTX injection. This finding may help explain the skeletal muscle regeneration defects seen in R3hdml KO mice.

In addition, overexpression of R3hdml in the forearm of R3hdml KO mice rescued the recovery of handgrip strength after CTX injection to levels comparable to that of wild‐type controls (Fig 5B and Appendix Fig S4D and E). These results indicate that R3hdml functions in skeletal muscle regeneration.

Recently, it has been reported that fibronectin1 (FN1), which is a component of the stem cell niche, affects the regenerative capacity of skeletal muscle in mice 10. Therefore, we examined FN1 and found that FN1 expression in skeletal muscle was significantly lower (56.5%) in R3hdml KO mice compared with controls at post‐natal day 0 (Fig EV5A). Silencing R3hdml in C2C12 cells led to a 31.8% decrease in FN1 mRNA expression on day 2 after differentiation. On the other hand, R3hdml overexpression significantly increased FN1 mRNA levels starting from day 2 after differentiation (Fig EV5B and C). These findings indicate that R3hdml regulates satellite cell regeneration capacity through direct or indirect regulation of FN1 expression.

Figure EV5. Fibronectin1 (FN1) expression.

1. Skeletal muscles in the lower legs were dissected, and RNA was extracted on post‐natal days 0, 7, and 21 in both R3hdml KO (n = 5) and wild‐type control (n = 4) mice. FN1 expression was examined by RT–PCR. GAPDH was used as an internal control. Data are expressed as the means ± standard error of the mean (SEM). *P < 0.05; Student's t‐test.
2. R3hdml gene silencing in C2C12 cells. C2C12 cells with (controls, closed bars) or without (R3hdml silenced, open bars) were differentiated after which FN gene expression was evaluated at the indicated time points (n = 3, biological replicates). Data are expressed as the means ± standard error of the mean (SEM). *P < 0.05; Student's t‐test.
3. R3hdml gene overexpression in C2C12 cells. FN gene expression was evaluated in mice in which R3hdml was overexpressed compared with controls at the indicated time points (n = 3, biological replicates). Data are expressed as the means ± standard error of the mean (SEM). *P < 0.05; **P < 0.01; Student's t‐test.

Discussion

In this study, we report a novel gene, R3hdml, which is expressed during skeletal muscle development. Although R3hdml was not expressed in adult skeletal muscle, R3hdml expression was detected in the skeletal muscle of post‐natal mice and transiently increased during muscle regeneration following pharmacological muscle injury. Using in vitro cell culture experiments, it was shown that R3hdml expression is induced in response to muscle differentiation and is controlled in the presence of MyoD. Silencing of the R3hdml gene in C2C12 cells attenuated myotube formation, and those phenotypes were rescued in the presence of R3hdml protein. A lack of R3hdml expression in mice led to decreased skeletal muscle weight and reduced expression of cell cycle‐related genes and Akt phosphorylation compared with that of wild‐type controls. R3hdml was also expressed during skeletal muscle regeneration. Handgrip strength recovery was delayed after CTX injection in R3hdml KO mice compared with littermate controls. Furthermore, overexpression of R3hdml at forearm rescued the wild‐type phenotype.

Skeletal muscle is important not only for supporting the body and enabling movement, but also for its role as the largest metabolic organ. Loss of muscle mass or atrophy causes a number of diseases, such as congenital myodystrophy, aging sarcopenia, and diabetes. Therefore, elucidating the mechanism of skeletal muscle development and regeneration is highly relevant. A number of myogenic factors have been reported as having important roles, including early myogenic factors such as pax7 11, MyoD 12, and myf5 13, followed by myog and Myh5 in later myogenesis 14. Coordinated expression of these myogenic factors is critical for skeletal muscle development and regeneration. As described here, R3hdml, which is regulated by MyoD, has been identified as a novel myogenic factor. The expression of R3hdml was observed in MyoD‐positive satellite cells but not in skeletal muscle cells, indicating that R3hdml promotes proper satellite cell function. Indeed, it appears that R3hdml is critical for satellite cell proliferation and differentiation.

Several reports have suggested that satellite cells are important for skeletal muscle development 15, 16. The deletion of inducible Pax7, a gene expressed in satellite cells, has been reported to lead to disturbances in skeletal muscle regeneration after barium chloride injection 7. In addition, a decreased capacity of satellite cell regeneration is reportedly associated with the development of sarcopenia 15 and loss of muscle mass during aging, which is due to the low capacity of satellite cells to regenerate upon skeletal muscle injury. Many signaling pathways influencing satellite cells are also reportedly involved in the development of sarcopenia. For instance, the NOTCH‐delta signaling factor is important for self‐replication of Pax7⁺/Myf5⁻ satellite cells, and decreased activity of this pathway has been associated with aging 17, 18, 19. Activation of the Frizzled 7‐WNT7 pathway was observed in aging skeletal muscle, leading to the differentiation of satellite cells into adipose cells instead of skeletal muscle cells 20. Increased expression of fibroblast growth factor‐2 (FGF‐2) due to the decreased expression of Sprouty1 in the aging state reportedly compromised satellite cell function 21. It has also been reported that increased JAK‐STAT signaling impairs muscle regeneration during aging in mice 22. Therefore, using RNA sequencing, we examined whether the signaling pathways mentioned above were disrupted when R3hdml was lacking in mice. However, none of these signaling pathways demonstrated differential expression between R3hdml KO mice and wild‐type controls; only cell cycle‐related genes were found differentially expressed.

It has also been reported that levels of the cyclin‐dependent kinase (Cdk) inhibitors p21^WAF1/CIP1 and p27 ^Kip1 were increased along with nuclear accumulation of p53 and Foxo1 in isolated satellite cells from old rats compared to young rats 23. These changes may be associated with decreased skeletal muscle regeneration. Therefore, it is important to increase our knowledge regarding factors functioning as cell cycle regulators of satellite cells to better understand the mechanisms of skeletal muscle development and regeneration.

In this study, we discovered that R3hdml is expressed in satellite cells and is involved in regulating their proliferation. R3hdml belongs to the cysteine‐rich secretory proteins, antigen 5, and pathogenesis‐related 1 proteins (CAP) superfamily based on sequence similarities 24. In humans and mice, there are 31 and 33 individual family members, respectively, although many members are poorly characterized. Moreover, most members display a notable expression bias for the reproductive tract and immune tissues and are dysregulated in cancers 24. CAP superfamily proteins are involved in a range of processes, including regulation of extracellular matrix production and branching morphogenesis (potentially as either proteases or protease inhibitors), ion channel regulation in fertility, tumor suppressor or pro‐oncogenic functions in tissues including the prostate, and in cell–cell adhesion during fertilization 24. However, there have been no reports concerning R3hdml.

The mechanisms by which R3hdml promotes satellite cell proliferation require further investigation. Our data indicate that IGF‐1/Akt signaling, which is known to be important for satellite cell proliferation, was decreased in R3hdml KO skeletal muscles. However, we were unable to show direct binding of R3hdml to Akt. It has been reported that elements of the extracellular matrix (ECM) of satellite cells such as fibronectin1 and type 1 collagen alpha are also important for satellite cell proliferation and maturation 10. Indeed, we found decreased expression of FN1 in R3hdml KO compared with wild‐type controls, although we do not yet know the molecular mechanism by which R3hdml regulates FN1 expression. The precise mechanism by which R3hdml inhibits IGF‐1/Akt also requires further investigation.

There were several limitations to this study. We only analyzed the expression and function of R3hdml in satellite cells isolated from limbs, which may or may not be applicable to satellite cells from skeletal muscle in other parts of the body. Moreover, we do not know how R3hdml affects the function of satellite cells. Since R3hdml is a secreted protein, we speculate that it may work through the activation of certain receptors or by regulation of the ECM surrounding satellite cells via the predicted protease inhibitor functions based on R3hdml protein structure. Indeed, adding the R3hdml protein to the culture media rescued the phenotypes seen in R3hdml‐silenced C2C12 cells, indicating that secreted R3hdml affects the cells from the outside. We intend to address these questions in future research.

In conclusion, we have identified R3hdml, a novel satellite cell‐secreted protein. Knockout of this gene inhibited satellite cell proliferation, skeletal muscle development, and regeneration. Because the lower proliferative capacity of satellite cells in aged muscle is reportedly associated with the development of sarcopenia, our results suggest that R3hdml is a potential therapeutic target in skeletal muscle‐related diseases. Further research on R3hdml is needed to confirm our findings and hypotheses.

Materials and Methods

Antibodies and reagents

CTX was purchased from Latoxan (Valence, France). Hoechst 33342 was obtained from Dojindo Laboratories (Kumamoto, Japan). Antibodies included rat monoclonal anti‐Ki67 (Dako Ltd., Kyoto, Japan), mouse polyclonal anti‐myogenin, rabbit polyclonal anti‐MyoD (Santa Cruz Biotechnology, Santa Cruz, CA), anti‐myosin heavy chain antibody (Millipore, Burlington, MA), anti‐Pax7, and anti‐myogenin (F5D; Developmental Studies Hybridoma Bank, Iowa City, IA).

Antibody production for R3hdml

Rabbit anti‐mouse R3hdml antibody was raised against the C‐terminal sequence of mouse R3hdml (CFSGLKSNRLPWV). The antibody was produced by Innovagen (Lund, Sweden).

R3hdml stable expression and protein purification

The full‐coding sequence of R3hdml was cloned into pQCXIH, a retrovirus vector, according to manufacturer's instructions (Clontech Laboratories, Inc.). Then, PT‐67 cells were transiently transfected with pQCXIH‐mouse R3hdml construct. Cell‐free viral supernatants were harvested at 24 and 48 h, mixed with 1 volume of complete media in the presence of 8 μg/ml polybrene (Sigma‐Aldrich, St. Louis, MO), and then used for transfection into Ad293 cells. Transfected cells were selected by 200 μg/ml hygromycin to establish the R3hdml‐stable‐expression cell line. For purification of R3hdml protein, the conditioned medium was collected, purified with anti‐R3dhml antibody‐conjugated column, and eluted with 0.2 M glycine at pH 2.5. Finally, the elution buffer was replaced with Dulbecco's phosphate‐buffered saline.

Cell culture

The mouse skeletal muscle cell line C2C12 myoblasts were maintained in DMEM supplemented with 10% fetal bovine serum (FBS; Sigma‐Aldrich), 100 U/ml penicillin, and 100 U/ml streptomycin (Sigma‐Aldrich) at 37°C and 5% CO₂. For biochemical experiments, the cells were grown on 6‐well plates at a density of 2.5 × 10⁵ cells/well in 2 ml of growth medium. After 24‐h incubation, differentiation was induced by switching the growth medium to DMEM supplemented with 2% horse serum, 100 U/ml penicillin, and 100 U/ml streptomycin (differentiation medium). The differentiation medium was changed every 48 h.

Isolation of satellite cells

Satellite cells were prepared as described previously 25. The extensor digitorum longus (EDL) muscles were isolated from R3hdml KO mice or C57BL6 wild‐type mice and digested in type I collagenase. Satellite cells were obtained from isolated myofibers by trypsinization in a 0.125% trypsin‐EDTA solution for 10 min at 37°C. Satellite cells were incubated in growth medium consisting of GlutaMAX DMEM (Thermo Fisher Scientific, Waltham, MA) supplemented with 30% FBS, 1% chicken embryo extract, 10 ng/ml bFGF, and 1% penicillin‐streptomycin at 37°C with 5% CO₂. Differentiation was induced in differentiation medium (GlutaMAX DMEM supplemented with 5% horse serum and 1% penicillin–streptomycin) at 37°C with 5% CO₂.

Reverse transcription polymerase chain reaction (RT–PCR)

The presence of R3hdml transcript was analyzed by RT–PCR. PCR was performed using a standard PCR kit with 1‐μl aliquots of cDNA and Go Taq Hot Start Polymerase (Promega, Madison, WI) with specific primer pairs. Primer pair sequences are listed in Appendix Table S1.

Real‐time PCR

Real‐time PCR was carried out as previously described 26. The primers used are shown in Appendix Table S2. Relative gene expression was calculated by the comparative ΔΔCt method using the 7500 Fast System SDS software (Applied Biosystems, Waltham, MA).

Mass spectrometric analysis

Conditioned medium from Ad293 cells with or without R3hdml overexpression was separated by agarose gel electrophoresis followed by Coomassie Brilliant Blue staining. A protein signal corresponding to 21 kDa was present with R3hdml overexpression but not without. This protein was cut out from the gel and subjected to mass spectrometric analysis.

For protein identification by peptide mass fingerprinting, protein spots were excised, digested with trypsin (Promega), mixed with α‐cyano‐4‐hydroxycinnamic acid in 50% acetonitrile/0.1% TFA, and subjected to MALDI‐TOF analysis (Microflex LRF 20; Bruker Daltonics, Billerica, MA) as described previously 27. Spectra were collected from 300 shots per spectrum over an m/z range of 600–3,000 and calibrated by two‐point internal calibration using trypsin auto‐digestion peaks (m/z 842.5099, 2211.1046). The peak list was generated using Flex Analysis 3.0. The threshold used for peak‐picking was as follows: 500 for minimum resolution of monoisotopic mass and 5 for S/N. The search program MASCOT, developed by Matrixscience (http://www.matrixscience.com/), was used for protein identification by peptide mass fingerprinting. The following parameters were used for the database search: trypsin as the cleaving enzyme, a maximum of one missed cleavage, iodoacetamide as a complete modification, oxidation as a partial modification, monoisotopic masses, and a mass tolerance of ± 0.1 Da. The PMF acceptance criteria were probability scoring. Mass spectrometric analyses were performed by Cosmo Bio Company Ltd. (Tokyo, Japan).

Silencing of the R3hdml gene by siRNA

R3hdml siRNA (stealth RNAi) was prepared by Invitrogen (Carlsbad, CA). The sequences used are shown in Appendix Table S3. Transfection was performed using Lipofectamine RNAiMAX (Invitrogen) according to manufacturer's instructions. Cells were seeded into a 15‐cm dish and after 24 h, and the cells were transfected with 4 pmol of iRNA using 4 μl of Lipofectamine RNAiMAX. The cells were incubated for 24 h at 37°C, and then, cells were seeded into 6‐well plates at 2.5 × 10⁵ cells/well and incubated using DMEM with 10% FBS. After 24‐h incubation, differentiation was induced by switching the growth medium to DMEM supplemented with 2% horse serum, 100 U/ml penicillin, and 100 U/ml streptomycin (differentiation medium) with or without purified R3hdml protein (50 ng/μl). The differentiation medium and R3hdml protein were changed every 48 h.

Western blotting

Cells were lysed in boiling Laemmli's sample buffer and then separated via 10% SDS‐polyacrylamide gel electrophoresis (PAGE). Immunoblotting was performed as described previously 26. Briefly, the samples were transferred to a PVDF membrane (Immobilon‐P, Merck Millipore, Burlington, MA), after which the blots were blocked with 5% BSA and probed with the primary antibodies overnight. The antibodies used were as follows: anti‐phospho‐Akt (Ser 473; 1:1,000; Cell Signaling Technology, Danvers, MA), anti‐Akt (1:1,000; Cell Signaling Technology), anti‐GAPDH (1:1,000; Cell Signaling Technology), and anti‐β‐actin (1:1,000 dilution; Cell Signaling Technology). Then, the blots were washed and incubated with peroxidase‐conjugated anti‐rabbit (1:2,500; GE Healthcare, Chicago, IL) IgG. After washing in PBS, immunoreactivity was visualized using an ECL Western blotting detection system (GE Healthcare).

Plasmids

We isolated the 781‐bp fragment upstream of the translational start codon of the R3hdml coding sequence using PCR. This fragment was then fused to the luciferase gene into the pGL4.14 vector (Construct C). We also subcloned the genomic sequence containing one high‐affinity MyoD binding site CACCTG (Construct B) and one without a high‐affinity MyoD binding site (Construct A) into a luciferase reporter vector. The CMV‐MyoD expression vector (plasmid #8398; Addgene, Watertown, MA) was a gift from Andrew Lassar (Harvard University, MA).

Luciferase assay

C2C12 cells were plated in 24‐well dishes at 50% confluency and transfected with 0.8 μg R3hdml‐luc (Construct A, B, C) per well and/or 0.8 μg of the CMV‐MyoD construct per well using Lipofectamine LTX and PLUS Reagents (Invitrogen). At 24 h after transfection, the medium was changed to DMEM supplemented with 10% FBS or supplemented with 2% horse serum to induce myogenic differentiation. Luciferase activity in the cell lysate was measured using a 1420 ARVO SX multilabel counter (Wallac, Inc., Gaithersburg, MD) using the Dual Luciferase Reporter Assay System (Promega).

Targeting strategy and generation of R3hdml mutant mice

VelociGene technology 28 was used to generate a specific deletion of the genomic sequence encoding the entire R3hdml gene, extending from exon 1 to the TAG termination codon. This sequence, which corresponds to a 10‐kb region encompassing all R3hdml coding exons and intervening introns, was replaced in frame by the coding sequence of GFP followed by a loxP‐flanked neomycin selection cassette. Briefly, a bacterial artificial chromosome containing the 10‐kb R3hdml coding region and flanking sequences was modified to generate a bacterial artificial chromosome‐based targeting vector, which was then linearized and used to replace the R3hdml gene sequence in F1H4 (C57BL/6‐129 hybrid) mouse embryonic stem (ES) cells. Correctly targeted ES cells were identified using the loss‐of‐native‐allele assay as previously described 28. Two independent correctly targeted ES lines were used to generate chimeric male mice which were then bred to C57BL/6 and/or ICR females to generate F1 mice. Heterozygous F1 mice (backcrossed to C57BL/6) were bred to homozygosity, and correct targeting was reconfirmed by RT and real‐time PCR. The animals were backcrossed further against C57BL/6 mice. Animals backcrossed more than seven generations were used for experiments, and littermate wild‐type animals served as controls.

All experiments in this study were performed in accordance with the Guidelines of the Animal Care and Use Committee of Chiba University, Japan, which follows the Guide for the Care and Use of Laboratory Animals (NIH publication no. 85‐23, revised 1985). The ethics committee for animal research at Chiba University approved all animal experiments. These mice were kept in a temperature‐controlled room with a 12:12 h light–dark cycle. Food and drinking water were provided ad libitum. Semi‐blinded cohorts were used meaning it was known that, e.g., three experimental and three control mice were within the group, but the genotype of each individual mouse was not known for the experimenter. Post‐mortem mice were re‐genotyped via PCR to ensure the correct genotype.

Genotyping of mutant mice

PCR amplification of genomic DNA extracted from tail biopsies was used to genotype R3hdml ^GFP/GFP mice. Briefly, tail samples were incubated in 100 μl of 50 mM NaOH for 20 min at 95°C. Next, samples were added to 100 μl of 0.5 M Tris–HCl (pH 8.0) and centrifuged at max speed for 1 min. PCR was performed using the following three specific oligonucleotide primers amplifying the wild‐type and knock‐in alleles, forward 1, 5‐GCACTTCTCATCACCTAACC‐3 in 5ʹ‐UTR; reverse 1, 5‐CTGCTGCCAACATGGAGTAT‐3 in exon 1; and reverse 2, 5‐TGAACTTCAGGGTCAGCTTG‐3 in the GFP cassette. The forward 1 and reverse 1 primers produced a 475‐bp signal corresponding to the wild‐type allele, while forward 1 and reverse 2 amplified a 366‐bp mutant allele.

PCR was performed using GoTaq Hot Start Polymerase (Promega). The reactions contained 5 μl 5× Green GoTaq Flezi buffer (Promega), 0.5 μl MgCl₂ (1.5 mM), 2.5 μl deoxynucleoside triphosphate (2 mM), 0.25 μl primers (0.3 mM), 0.125 μl GoTaq HotStart Polymerase (5 U/μl), 1 μl DNA template, and distilled H₂O in a final volume of 25 μl. Amplification was performed using a PerkinElmer 9700 PCR machine with the following program, 94°C for 15 min followed by 30 cycles of 94°C for 30 s, 56°C for 30 s, and 72°C for 30 s, and finally 72°C for 2 min.

Transmission electron microscopy

Tissue samples were fixed in 2.5% glutaraldehyde/2% paraformaldehyde (PFA) in 30 mM HEPES buffer and then postfixed in 1% OsO₄ and dehydrated in graded alcohol as described previously 29. Briefly, after immersion in propylene oxide, the specimens were embedded in Epon 812 (Sigma‐Aldrich). Ultrathin sections were cut, doubly stained with uranyl acetate and lead citrate, and examined with a transmission electron microscope (TEM‐1010; JEOL, Tokyo, Japan).

RNA sequencing

Total RNA was extracted from neonatal skeletal muscle using the RNeasy Plus Micro Kit (Qiagen, Hilden, Germany), after which cDNA libraries were generated using the NEBNext Ultra RNA Library Prep Kit (New England BioLabs, Beverly, MA). Sequencing was performed using a HiSeq1500 (Illumina, San Diego, CA) with a single‐read sequencing length of 60 bp. TopHat (version 2.0.13; with default parameters) was used to map sequences to the reference genome (UCSC/mm10 or UCSC/hg19) with annotation data from iGenomes (Illumina). Finally, gene expression levels were quantified using Cuffdiff (Cufflinks version 2.2.1; with default parameters).

Immunocytochemistry

Immunocytochemistry was performed as previously described 30. After fixation with 4% PFA, cells were cultured with primary antibodies at 4°C overnight. The primary antibodies were visualized by appropriate species‐specific 488 and 594 fluorochrome‐conjugated secondary antibodies (Invitrogen). Nuclei were stained with Vectashield fluorescent mounting medium containing 4,6‐diamidino‐2‐phenylindole (DAPI; Vector Laboratories, Burlingame, CA) or Hoechst 33342. Nuclear number was determined by calculating the percentage of nuclei incorporated into the MyHC‐positive myotubes (> 2 nuclei) with the indicated number of nuclei.

Muscle regeneration

Analysis of muscle regeneration was carried out as described previously 31 with slight modifications. Briefly, 100 μl of 10 μM CTX (Latoxan) was injected intramuscularly into the tibialis anterior muscle of anesthetized mice using a 29G 1/2 insulin syringe. Regenerating muscles were isolated 3, 7, and 14 days after CTX injection and immediately frozen or fixed in 4% formaldehyde in 0.1 M PBS.

In situ hybridization and immunohistochemistry (IHC)

Mouse tissue was fixed with G‐Fix (Genostaff, Tokyo, Japan), embedded in paraffin on CT‐Pro20 (Genostaff) using G‐Nox (Genostaff) as a less toxic organic solvent than xylene, and serially sectioned at a thickness of 6 μm. A pair of sections were set up as mirror images of each other: one section was put on a glass slide for ISH, and another section was reversed and put on a different glass slide for IHC.

In situ hybridization was performed with the ISH Reagent Kit (Genostaff) according to manufacturer's instructions. Tissue sections were de‐paraffined with G‐Nox and rehydrated in a graded ethanol series and PBS. The sections were then fixed with 10% NBF (10% formalin in PBS) for 15 min at RT, treated with 4 μg/ml Proteinase K (Wako Pure Chemical Industries, Osaka, Japan) in PBS for 10 min at 37°C, re‐fixed with 10% NBF for 15 min at RT, and then incubated in 0.2 N HCl for 10 min at RT; washing in PBS was performed after each step. Next, the sections were placed within a Coplin jar containing 1X G‐Wash (Genostaff), equal to 1X SSC. Hybridization was performed with probes at a concentration of 300 ng/ml in G‐Hybo‐L (Genostaff) for 16 h at 60°C. After hybridization, the sections were washed in 1X G‐Wash for 10 min at 60°C, and then in 50% formamide in 1X G‐Wash for 10 min at 60°C. The sections were then washed twice in 1X G‐Wash for 10 min at 60°C, twice in 0.1X G‐Wash for 10 min at 60°C, and twice in TBST (0.1% Tween‐20 in TBS) at RT. After treatment with 1X G‐Block (Genostaff) for 15 min at RT, the sections were incubated with anti‐DIG AP conjugate (Roche Diagnostics, Basel, Switzerland) diluted 1:2,000 with 50X G‐Block in TBST for 1 h at RT. The sections were then washed twice with TBST and incubated in 100 mM NaCl, 50 mM MgCl₂, 0.1% Tween‐20, and 100 mM Tris–HCl pH 9.5. Coloring reactions were performed using NBT/BCIP solution (Sigma‐Aldrich) overnight, followed by washing in PBS. The sections were counterstained with Kernechtrot stain solution (Muto Pure Chemicals, Tokyo, Japan) and mounted with G‐Mount (Genostaff). The sequences of the probes used are shown in Appendix Table S4.

For IHC staining, tissue sections were de‐paraffined with G‐Nox and rehydrated in a graded ethanol series and PBS. Antigen retrieval was performed by microwave treatment with citrate buffer (pH 6.0). Endogenous peroxidase was blocked with 0.3% H₂O₂ in methanol for 30 min, followed by incubation with G‐Block and Avidin/Biotin Blocking Kit (Vector Laboratories). The sections were then incubated with 1 μg/ml of anti‐MyoD rabbit pAb (Santa Cruz Biotechnology) at 4°C overnight, followed by incubation with biotin‐conjugated goat anti‐rabbit Ig (Dako, Santa Clara, CA) for 30 min at RT, after which peroxidase‐conjugated streptavidin (Nichirei, Tokyo, Japan) was added for 5 min. Peroxidase activity was visualized by diaminobenzidine (Dojindo, Tokyo, Japan), after which the samples were washed in PBS. The sections were counterstained with hematoxylin (Muto Pure Chemicals) and mounted using Malinol (Muto Pure Chemicals).

Grip strength evaluation

The grip strength test was performed as described previously 32. Briefly, 10 μM CTX diluted in 50 μl PBS was injected into the forearm muscles of mice. Five training sessions were performed during which the animals were held facing the net of the grip strength meter (Muromachi Kikai, Tokyo, Japan). After the animal grasped the net, the animal was gently pulled away from the device. The grip strength meter measured the maximal force applied before the animal released the bar. Each animal was assessed five times consecutively on a given day.

In vivo gene transfer by electroporation

In vivo electroporation was performed according to the modified method of Aihara and Miyazaki 33. Under the anesthesia, bilateral forearm muscles were injected with 100 μg of plasmid DNA (50 μl) by using a 30‐gauge needle. Square‐wave electrical pulses (200 V/cm) were applied eight times with an electrical pulse generator (Uchida Denshi, Hachioji, Japan) at a rate of one pulse per second, with each pulse being 20 ms in duration. The electrodes were a pair of stainless steel needles inserted into the forearm muscles and fixed 6 mm apart.

Statistical analysis

Data are expressed as the means ± standard error of the mean (SEM). GraphPad Prism 6.0 software (GraphPad Software Inc., La Jolla, CA) was used for statistical analysis. When only two groups were analyzed, Student's unpaired t‐test was performed to compare differences between the groups. To compare multiple groups, one‐way analysis of variance (ANOVA) followed by a Bonferroni post hoc test was used. Repeated measures‐based parameters (such as body weight or muscle strength) were analyzed using two‐way ANOVA for repeated measures followed by Bonferroni correction. P < 0.05 was considered statistically significant.

Author contributions

KS: analysis and interpretation of data, acquisition of subjects and data; YF: acquisition of subjects and data; MY: acquisition of subjects and data; MTaka: acquisition of subjects and data; YA: acquisition of subjects and data. Electron microscope analysis, TI: interpretation of data; TS: interpretation of data; MF: acquisition of subjects and data; AH: interpretation of data; YMi: acquisition of subjects and data; AW: acquisition of subjects and data, YMan: interpretation of data and discussion; NLF: interpretation of data, discussion, edited the manuscript; RI; interpretation of data; YMae: interpretation of data, discussion; CB: interpretation of data, discussion; KY: discussion, reviewed/edited the manuscript; MTake: interpretation of data, preparation of manuscript.

Conflict of interest

The authors declare that they have no conflict of interest.

Supporting information

Appendix

Expanded View Figures PDF

Source Data for Expanded View

Review Process File

EMBO Reports (2019) 20: e47957