Adult C57BL/6 WT mice (8–16 weeks old, from Jackson Laboratory or Japan SLC) and adult CAG-EGFP mice (8–16 weeks old, from Japan SLC) were used as host and donor mice, respectively, for transplantation experiments. These mice were fed normal chow and water under standard lighting conditions (12 h light/dark cycle) at 23°C–25°C.

The extensor digitorum longus (EDL) was isolated from CAG-EGFP mice, and single muscle fibers were digested using type I collagenase as previously described (Furuichi et al., 2021). For culturing satellite cells, the collagenase concentration was 0.25% (w/v), and the incubation time was 2 h at 37°C. Single muscle fibers were collected under a stereomicroscope; cells and debris other than muscle fibers were removed. The muscle fibers were pooled and incubated with Accutase (Innovative Cell Technologies, SAN, United States) for 10 min at room temperature and cultured on Matrigel-coated dishes. Myoblasts derived from satellite cells were cultured in growth medium (No glucose DMEM supplemented with 30% (v/v) FBS (NICHIREI BIOSCIENCES INC., Tokyo, Japan), 1% (v/v) GlutaMAX, 1% (v/v) chicken embryo extract, 10 ng/mL bFGF and 1% (v/v) penicillin–streptomycin) at 37°C with 5% CO₂ (Furuichi et al., 2021). Myoblasts were expanded 7–8 days for the transplantation study.

Myoblasts derived from green fluorescent protein (GFP) mice were removed from culturing dishes with 0.25% trypsin-EDTA (Gibco) and centrifuged at 1,000 rpm for 5 min before the supernatant was discarded. The cell pellet was then slowly suspended in growth medium, and a portion of the suspension was used for cell counting (OneCell, Japan). A sample (1.0 × 10⁵) of cells was removed from the cell suspension and centrifuged at 500 g for 5 min and mixed with Matrigel (final concentration was 1.0 × 10⁵ cells per 50 µL with 0 mg/mL, 0.5 mg/mL, 2.5 mg/mL, and 5.0 mg/mL Matrigel) or without Matrigel (only DMEM). The prepared cell suspension was injected into the tibialis anterior (TA) muscle of WT mice anesthetized with 7.5% Domitor, 8% midazolam, 10% butorphanol, and 74.5% saline (0.05 mL per 10 g mouse body weight). Anesthetized mice were shaved around the TA muscle. A 26 gauge (G) plastic syringe was injected from the bottom of the TA near the distal tendon to the top of the muscle and held for 10 s. The starting point of each needle injection was 2 mm above the TA distal tendon. The needle was pushed forward straight to the side of the knee within the TA muscle. The insertion of the needle was always stopped before the end of the muscle. The injection depth was approximately 500 µm deeper than the TA muscle surfaces. Then, 50 µL of the prepared cell suspension was injected over 1 min, and the syringe was removed. After confirming that the injected cell suspension did not leak out, the mice were awaked by an anti-anesthetic (5 mg/mL Antisedan in saline, ZENOAQ, Japan), using the same dose as what was used to anesthetize them. To determine the optimal needle and syringe, a 26 G plastic syringe (JMS Co., Ltd., Japan), a 27 G (TERUMO CORPORATION, Japan), a 29 G (Becton, Dickinson and Company, USA), or a 26 G glass syringe (Hamilton, GL Science, Japan) was used. For muscle injury, TA muscles of the anesthetized WT mice were injected with 100 µL of cardiotoxin at a concentration of 10 µM (Latoxan, France) 1 day prior to transplantation. Three or 6 weeks later, the transplanted TA muscles were dissected and embedded in Tissue-Tek O.C.T. compound (Sakura) to fix in isopentane cooled with liquid nitrogen. The frozen muscles were stored at −80°C until they were cryosectioned.

The frozen TA muscles were sectioned (8–10 µm) using a cryostat (Leica). The sections were immediately observed to acquire their GFP signal using a KEYENCE BZ-X710/810. For immunostaining, the sections were fixed in 4% paraformaldehyde or acetone for 10 min and blocked with 20% goat serum in PBS for 30 min at room temperature. The sections were then incubated with primary antibody (Laminin-α2; 1:100, Enzo Life Science) in blocking buffer at 4°C overnight before being washed with PBS and labeled with fluorescence-conjugated secondary antibody (Alexa594-anti Rat, Invitrogen) for 1 h at room temperature. The sections were mounted using mounting medium with DAPI (Vector Laboratory). Laminin and DAPI were visualized using a KEYENCE BZ-X710/810, and the images were merged with the GFP images (Supplementary Figure S1). The merged images were analyzed by ImageJ (for counting GFP positive fibers) or KEYENCE BZ-X series Image Analysis Software (for detecting GFP positive area, GFP fluorescence intensity, and fibrosis area).

To stain the α-Bungarotoxin protein, the sections were blocked with 3% BSA, 5% goat serum, and 0.1% Triton X-100 in PBS and incubated with Laminin-α2 primary antibody. The following day, the sections were reacted with α-Bungarotoxin fluorescence-conjugated antibody (1:500, Alexa594, Invitrogen) and Laminin-α2 secondary antibody for 2 h at room temperature and mounted with DAPI.

To stain the neurofilament protein, the sections were blocked with the same solution of α-Bungarotoxin staining and incubated with ready-to-use anti-Neurofilament (Dako) at 4°C overnight. After washing to remove excess unbound primary antibody, the sections were reacted with fluorescence-conjugated secondary antibody (1:500, Alexa594-anti-Mouse, Invitrogen) for 2 h at room temperature and mounted with DAPI.

CD31 staining, the sections were fixed in acetone for 10 min at −20°C and blocked with 5% BSA, 1% goat serum, and 0.05% Triton X-100 in PBS for 1 h at room temperature. The sections were then incubated with primary antibody (CD31; 60 ng, BioLegend) in blocking buffer for 2 h at room temperature. After washing to remove excess unbound primary antibody, fluorescence-conjugated secondary antibody (1:200, Alexa594-anti-Rat, Invitrogen) was added for 2 h at room temperature, followed by mounting with DAPI.

For hematoxylin and eosin (HE) staining, muscle cryosections were washed in PBS for 3 min and stained in hematoxylin solution (Cosmo Bio, Tokyo, Japan) for 4 min. The sections were washed in running water for 1 min and stained in eosin solution (Cosmo Bio, Tokyo, Japan) for 5 min before being washed in distilled water for 2 min. The sections were incubated in 70%, 80%, 90%, and 100% ethanol for 1 min, 2 min, 3 min, and 5 min and then mounted with a hydrophobic mounting medium.

For Sirius Red staining, muscle cryosections were fixed in 4% PFA for 10 min and washed twice in 100% ethanol for 5 min and 15 min. After washing, the sections were dried for 30 min, washed in distilled water for 1 min, and incubated in Sirius Red solution (1.3% picric acid; Wako, Osaka, Japan, 0.1% Direct Red80; Sigma, St. Louis, USA) for 45 min. The sections were washed in 0.5% acetic acid for 5 min and distilled water for 1 min before being mounted with a hydrophobic mounting medium. The percentage of collagen area in the whole TA (%) was evaluated using the formula {collagen-positive area (µm²)/whole TA area (µm²)} × 100. The region of interest (ROI) was defined as the square that enclosed all GFP + fibers within the cross section, where each corner of the square was set to coincide with the outermost GFP fibers (Supplementary Figure S2). The collagen area in the ROI (%) is evaluated using the formula {collagen-positive area in ROI (µm²)/TA area in ROI (µm²)} × 100.

Total RNA was extracted from skeletal muscle tissue using TRIzol reagent (Invitrogen). RNA was transcribed into cDNA via a standard reverse transcriptase reaction using the PrimeScript^TM first-strand cDNA Synthesis Kit (Takara, Shiga, Japan), following the manufacturer’s protocol. These cDNAs were analyzed with quantitative real-time PCR (qPCR). qPCR was performed on a 96-well CFX Connect Real-Time PCR Detection System (Bio-Rad Laboratory, Tokyo, Japan) using a THUNDER-BIRD Probe qPCR Mix (TOYOBO, Osaka, Japan). The mRNA levels of GFP or TNF-α were normalized to that of the housekeeping gene TATA-binding protein (Tbp). Primer sequences 5′-3′ for this study were as follows: GFP (forward; ACG​TAA​ACG​GCC​ACA​AGT, reverse; AAG​TCG​TGC​TGC​TTC​ATG​TG); TNF-α (forward; ATGGGGGGCTTCCAGAA, reverse; CCT​TTG​GGG​ACC​GAT​CA); Tbp (forward; AAT​GAC​TCC​TAT​GAC​CCC​TAT​CAC, reverse; AGG​TCA​AGT​TTA​CAG​CCA​AGA​TTC). The mRNA content of both GFP and Tbp was calculated from the cycle threshold values using a standard curve, and the ratio between GFP and Tbp was calculated. Primers were synthesized by Eurofins Genomics Co., Ltd. (Tokyo, Japan).

Under anesthesia, as described above, the TA muscle was exposed and isolated at its distal tendon. The mouse was placed on the mouse muscle tension measurement system (Uchida Denshi, Tokyo, Japan), and the mouse knee was fixed with forceps. The distal TA tendon was attached to a force transducer, and a pair of platinum-coated electrodes was placed on the common peroneal nerve for electrical stimulation with rectangular unipolar 1.0 voltage pulses of 0.2 m duration at a frequency of 250 Hz. The TA muscle was maintained at optimal length, and the maximal tetanic force was determined from stimulation with the above signal for 10 s. The maximum voltage in the first contraction was determined as the contraction force (Supplementary Figure S3).

All values are shown as the mean ± standard error (s.e.m.). Two-tailed unpaired Student’s t-tests were used to compare the two groups. For multiple comparisons, data were analyzed using a one-way ANOVA or two-way ANOVA followed by the Tukey post hoc test or Dunnett’s test. The significance was set to p< 0.05. Sample numbers n in all experiments were independent replicates.

In tissue, cells are covered by an ECM that preserves their shape and function; the ECM allows myoblasts to bind to the dish or each other for myogenesis, including proliferation and differentiation (Chaturvedi et al., 2015; Penton et al., 2016; Thomas et al., 2015). Myoblasts derived from GFP mice and cultured for 1 week were detached from the culture dishes using trypsin for transplantation. The myoblasts were separated into aliquots of 1.0 × 10⁵ cells and mixed with Matrigel, an ECM component, at a concentration of 0.5 mg/mL. We transplanted myoblasts into the TA muscle of WT mice with or without Matrigel. We also compared what happened when cells in either condition were transplanted into intact muscle or muscle with a CTX-induced injury. Twenty-one days later, the transplanted TA muscle was dissected and analyzed (Figure 1A).

As reported in previous studies, GFP + myofibers were observed in small numbers in pre-injured muscles but rarely in intact muscles, suggesting that myoblasts did not engraft into intact muscle (Figure 1B). However, the number of GFP + myofibers was increased when myoblasts mixed with 0.5 mg/mL Matrigel were transplanted into intact muscle as well as under pre-injury conditions (Figures 1B, C), indicating that ECM components such as Matrigel allowed myoblasts to engraft into muscle without injury. Moreover, when myoblasts mixed with 0.5 mg/mL Matrigel were implanted into intact muscle, GFP + myofibers with central nuclei, considered immature, were observed. Such immature GFP + myofibers were brighter than matured cells (Figure 1B). Note that in intact muscle, the number of GFP + myofibers was increased by adding Matrigel (Figures 1B,C); meanwhile, Matrigel had no effect on myoblast engraftment when the muscle was pre-injured (Figures 1B, C). In general, the number of GFP + fibers is interpreted as reflecting the engraftment efficiency of transplanted cells, but this could be an underestimate if the transplanted cells cluster together to form a single myofiber. Therefore, we measured the intensity of the GFP signal, which reflects the amount of GFP myoblasts, and found that the GFP intensity was significantly stronger when transplantation was carried out with Matrigel in intact muscle than in other conditions (Figure 1D). To validate the accuracy of our measurement of the amount of successfully fused GFP myoblasts, we measured the GFP mRNA expression level in intact conditions by qPCR and confirmed that the expression of GFP mRNA was higher with Matrigel than without Matrigel (Figure 1E). From these results, we concluded that ECM components enhanced myoblast engraftment into intact muscle and that this effect was specific to the intact condition.

Transplantation of allogeneic cells or injection of Matrigel may induce an immune response in host tissues. To evaluate an immune response caused by myoblast transplantation, we measured the expression of TNF-α mRNA, which is a well-known immune marker, in skeletal muscles. We confirmed that muscle injury induced by CTX injection increased TNF-α mRNA expression compared to basal control (Supplementary Figures S4A, B). Our data showed that TNF-α expression level was not increased with myoblast injection compared to injection of only saline, suggesting that the transplantation of myoblasts mixed with Matrigel had little effect on the immune response (Supplementary Figures S4C, D).

Because cells were damaged as they passed through the syringe needle, we examined whether the syringe used for cell injection affected engraftment efficiency. Previously, many types of syringes were used for muscle cell transplantation; materials were mainly plastic or glass, and the needle diameters were different among researchers (Lorant et al., 2018; Arpke and Kyba, 2016; Hashimoto et al., 2004; Elhussieny et al., 2021; Parker and Tapscott, 2013; Motohashi et al., 2014; Praud et al., 2018). For example, glass can have a charge on the surface (Behrens and Grier, 2001), and this feature could affect cell viability, but it was unclear which materials were useful for myoblast transplantation. Moreover, the size of myoblasts was larger than that of satellite cells, so they may be more damaged by passage through a needle than satellite cells.

First, we investigated whether plastic or glass is more suitable for transplanting myoblasts. We used plastic or glass syringes and transplanted 1.0 × 10⁵ myoblasts into intact TA muscle with 0.5 mg/mL Matrigel following the same schedule shown in Figure 1A. The number of GFP + myofibers was significantly reduced when a glass syringe was used (Figures 2A,B), suggesting that a plastic syringe is better for myoblast injection. This is supported by previous studies demonstrating that cells can adhere to glass cannulas (Torres et al., 2015; Amer et al., 2017), which may reduce the number of cells that completely transit a glass syringe.

Skuk et al. (2014) studied the relationship between the needle size and the engraftment efficiency of transplanted myoblasts in monkey models. The authors reported that a 27 G needle provided the best balance between minimizing host muscle injury and preserving the viability of donor-transplanted cells. However, in the mouse model, the optimal needle size is unclear. We explored the best needle size for myoblast transplantation using different needles, including 23 G, 26 G, and 29 G. We evaluated transplant efficiency after 21 days in the same way as in Figure 1. There was a trend of a higher number of GFP + fibers using a 26 G needle. However, the difference among the groups was not significantly different (one-way ANOVA p = 0.074, Figures 2C, D). On the other hand, the GFP + area was higher using a 26 G needle than with a 29 G needle, but there was no significant difference between 26 G and 23 G (Figure 2E). Therefore, the 26 G plastic syringe was deemed suitable for myoblast transplantation into intact muscle.

We examined whether further increasing the concentration of Matrigel would enhance the efficiency of myoblast engraftment into intact muscle. We transplanted 1.0 × 10⁵ myoblasts into intact TA muscle mixed with 0 mg/mL, 0.5 mg/mL, 2.5 mg/mL, and 5.0 mg/mL Matrigel. Twenty-one days after transplantation, the TA muscle was dissected and analyzed. The number of GFP + myofibers was higher in the 2.5 mg/mL and 5.0 mg/mL Matrigel concentrations than in the 0 mg/mL sample (Figures 3A,D). However, these high concentration conditions increased the non-myofiber area that was Laminin-α2 positive (Figure 3A). We prepared the cross section to confirm the morphology with HE staining and observed that the non-myofiber area showed signs of fibrosis (Figure 3B). Next, the transplanted muscle sections were stained with Sirius Red, which stains collagen in red and myofiber in yellow. This experiment revealed that the Laminin-α2 positive area matched the collagen-positive area, indicating that fibrosis occurred with the 2.5 mg/mL and 5.0 mg/mL Matrigel injections but not with the 0 mg/mL and 0.5 mg/mL injections (Figures 3A–C, E). Although myoblast transplantation with more Matrigel (especially 2.5 mg/mL) showed high engraftment efficiency, 2.5 mg/mL Matrigel caused fibrosis. If the fibrosis caused by the Matrigel injection could be eliminated, myoblast transplantation with 2.5 mg/mL Matrigel would be a better method to achieve high engraftment efficiency. We also examined the effect of injecting a high concentration (2.5 mg/mL) of Matrigel on the immune response (Supplementary Figures S4E, F). The expression of TNF-α was not changed following the injection of saline and 2.5 mg/mL Matrigel (Supplementary Figure S4F).

We examined whether increasing the number of transplanted cells from 1.0 × 10⁵ to 1.0 × 10⁶ could reduce the collagen area by stacking the transplanted cells. To do this, 1.0 × 10⁵ cells, and 10 times that number, 1.0 × 10⁶ cells, were mixed with 0.5 mg/mL or 2.5 mg/mL Matrigel and transplanted into TA muscle. We found that the number of GFP + fibers was observed to exceed 1,000 fibers when 1.0 × 10⁶ cells were transplanted with 2.5 mg/mL Matrigel, which was 10 times higher than that of the 1.0 × 10⁵ cells transplanted with 0.5 mg/mL Matrigel (Figures 4A, B). The GFP + area using 1.0 × 10⁶ cells in 2.5 mg/mL Matrigel was higher than in the other conditions, occupying almost 10% of the total TA area (Figures 4A, C; Supplementary Figure S1). The number and area of GFP + fibers were also found to increase with more transplanted cells (Figures 4A, C). We also analyzed the longitudinal cryosections, and they showed that the transplanted myoblasts were extensively engrafted in whole TA muscle, and the amount of the engrafted cells was slightly sparse in different locations (Supplementary Figures S5A, B). The GFP + fibers were formed in the same orientation as existing myofibers.

As shown in Figure 3, high-concentration Matrigel (2.5 mg/mL) increased the area where fibrosis was seen despite the increase in the engraftment of transplanted cells (Figure 4D). By transplanting 10 times as many cells (1.0 × 10⁶ cells), however, cells could invade the collagen area, suppressing the collagen deposition (Figure 4D). We quantified the collagen area over the whole TA cross section but failed to detect a significant difference between the conditions (Figure 4E). However, this analysis underestimates the change in collagen area due to Matrigel injection because it includes intrinsic collagen that is scattered throughout the muscle sections, even where myoblasts and Matrigel have not spread. Thus, we determined a region of interest (ROI), defining it as the area over which injected myoblasts and Matrigel have infiltrated and recalculated the collagen area in the ROI, as described in the Methods section (Supplementary Figure S2). Although the collagen area in the ROI increased when 1.0 × 10⁵ myoblasts were transplanted with 2.5 mg/mL Matrigel compared to 0.5 mg/mL (Figure 4F), there was no significant difference between 1.0 × 10⁵ myoblasts mixed with 0.5 mg/mL Matrigel and 1.0 × 10⁶ myoblasts transplanted with 2.5 mg/mL. Although no significant difference was observed by increasing cell number in the condition of 2.5 mg/mL Matrigel, our data imply that the increasing number of transplanted cells prevented the collagen accumulation caused by injecting high concentrations of Matrigel. We propose that 1.0 × 10⁶ myoblasts transplanted with 2.5 mg/mL Matrigel will engraft efficiently and will suppress fibrosis.

We also analyzed results 6 weeks after myoblast transplantation, which was twice the duration of the previous experiment. We transplanted 1.0 × 10⁵ or 5.0 × 10⁵ myoblasts mixed with 0.5 mg/mL or 2.5 mg/mL Matrigel into intact TA muscle and successfully detected the engraftment of myoblasts in all conditions (Supplementary Figure S6). When 1.0 × 10⁵ transplanted myoblasts were mixed with 0.5 mg/mL Matrigel, we observed 288.0 GFP + fibers, which was three times more than the count of 96.0 in the same conditions at 3 weeks (Supplementary Figure S6C, Figure 4B). Additionally, the area of GFP + as a percentage of the total TA area was also three times higher at 6 weeks than at 3 weeks (4.95% vs. 1.64%) (Supplementary Figure S6D, Figure 4C). Although Pavlath et al. (1994) reported that engrafted myofibers were reduced depending on the periods after injection, our data suggest that the engrafted myofibers can survive at least 6 weeks after transplantation.

As shown in Figure 4B, 1.0 × 10⁶ myoblasts transplanted with 2.5 mg/mL Matrigel generated more than 1,000 myofibers in TA muscle. We examined whether muscle weight and strength could be increased by myoblast transplantation in this manner. A sample of 1.0 × 10⁶ cells of myoblasts isolated from a GFP mouse was mixed with 2.5 mg/mL Matrigel and transplanted into one intact TA muscle of a WT mouse (shown as the Transplant condition), while the other TA was injected with only 2.5 mg/mL Matrigel as a control (Figure 5A). After transplantation, the TA with transplanted cells showed GFP + fibers (Figure 5B) and tended to increase the weight of the TA muscle (Figure 5C). Furthermore, the TA muscle weight normalized by body weight was increased by myoblast transplantation (Figure 5D). We demonstrated that the transplantation of myoblasts into intact muscle tissue could increase muscle weight by 10%. We also measured the strength of the transplanted TA muscle using the mouse muscle strength measurement system (Supplementary Figure S3). We found that the TA contraction force was not changed by myoblast transplantation, and the force normalized to body weight was the same (Figures 5E, F). Staining for α-Bungarotoxin and Neurofilament, which are expressed in the synapse structure, revealed that both markers were present in the GFP + fibers sample (Supplementary Figure S7). Moreover, we observed that CD31 expressed in endothelial cells was colocalized with GFP + fibers, suggesting that the vascularization occurred in the transplantation area (Supplementary Figure S8). Thus, our myoblast transplantation method did not show augmentation of muscle strength even if myofibers formed by transplanted cells had synapses and capillaries.