UCMD model mice were generated as previously described (Noguchi et al., 2017). In order to distinguish these mice from conventional Col6a1 ^−/− mice, we call them Col6a1 ^KO mice. Col6a1 ^KO mice were backcrossed to achieve the NSG (NOD acid gamma) background. After establishing the Col6a1 ^HETERO /NSG mice line, we crossed the mice again to produce Col6a1 ^KO /NSG mice. Because Col6a1 ^KO /NSG mice are hard to conceive, the offspring were maintained by in vitro fertilization (IVF) with ICR surrogate mothers. We used 3–8 generations of the Col6a1 ^KO /NSG strain. The mice were kept under a 12/12 h light-dark cycle with easy and comfortable access to food and water. The pups were weaned at 3 weeks of age. The genotype of the offspring confirmed no contamination of Col6a1 ^HETERO /NSG mice. WT/NSG mice (NOD.Cg-Prkdc ^scid Il2rg ^tm1Wjl /Szj) were purchased from Charles River Laboratory and kept under the same environment as Col6a1 ^KO /NSG mice.

The differentiation of 201B7 iPSCs to iMSCs via iPSC-derived neural crest cells (iNCCs) was performed as previously published (Fukuta et al., 2014). Luciferase-expressing piggyBac vector (pPV-EF1a-Luc2-iP-A) (Xu et al., 2019) and piggyBac transposase plasmids (pHL-EF1a-hcPBase-A) were introduced into the iNCCs by electroporation with a NEPA21 electroporator (Nepa Gene, Co., Ltd., Chiba, Japan), and the transfected cells were selected with 0.5% puromycin over 7 days. Luciferase-expressing iNCCs were maintained by passaging after 90% confluency or kept frozen in cryopreservation liquid (Bambanker, Nippon Genetics Co., Ltd., CS-02-001) as a stock. The iNCCs were induced to iMSCs just before use. Only passage 5 or 6 iMSCs were used for the transplantation.

Only male pups were used for analysis throughout the study. For neonatal i. p. transplantation, 5 × 10⁶ iMSCs in 20 μl phosphate-buffered saline (PBS) were injected intraperitoneally using a 30G insulin syringe (BD, No. 326668). All the mice analyzed at 8 weeks had a boost i. p. transplantation at 4 weeks, in which 5 × 10⁶ iMSCs in 100 μl PBS were injected using a 23G 1 ml syringe (Terumo, SS-01T2613).

Mice were anesthetized with isoflurane, and the luminescence was imaged with an IVIS imaging system (Parkin Elmer) 10 min after the administration of 150 mg/kg d-luciferin substrate (Summit Pharmaceutical International Corporation, XLF-1) in PBS. The luminescence intensity was measured, and the region of interest (ROI) was set to cover all the signals.

cDNA used for luciferase detection was treated as follows. After the removal of DNA contamination with DNase I (Invitrogen, 18068-015), 1 μg mRNA was reverse transcribed with SuperScript III using a random hexamer as instructed (Invitrogen, 18080-051). The PCR products were electrophoresed on a 1% agarose gel for 30 min to detect amplified fragments of luciferase and imaged under a FAS-IV gel imaging device (Nippon Genetics, Co., Ltd.). Positive control samples were dissected from a teratoma that had been formed 8 weeks after the subcutaneous injection of luciferase-integrated iPSCs into the TA muscle of WT/NSG mice. Negative control samples were extracted from the TA muscle of Col6a1 ^KO /NSG mice.

Mice were euthanized by exposure to gradually increasing concentrations of carbon dioxide. Dissected TA and quadriceps were stuck onto the pedestal by tragacanth gum (Wako, 206-02242) on a cork and frozen by dipping into 2-methylbutane (Wako, 166-00615) for 45–60 s at melting point temperature (−156°C). Freshly isolated diaphragms were embedded in OCT compound (Sakura Finetek, 4583, USA) and directly frozen within 2-methylbutane. Frozen muscles were sliced at 12-μm thickness with a Cryostat (Leica, CM1850) and attached to the slide. The sample on the slide was kept at −80°C for the following immunostaining.

The frozen sample on the slide was fixed in 4% paraformaldehyde for 20 min and washed twice with PBS. After 1-h incubation with Blocking one (Nacalai tesque, 03953-95), the primary antibody in immunoreaction enhancer solution (Can get signal, TOYOBO, NKB-601) was applied to the sample overnight at 4°C. The sample was washed with 0.1% Triton-X in PBS three times for 10 min each. The secondary antibody was applied for 1 h at room temperature, and the sample was washed three times. Images were obtained under a confocal microscope (Zeiss, LSM710). The primary and secondary antibodies used for the immunostaining and the dilution ratios are described in Supplementary Table S1.

TUNEL positive cells were stained with an Apoptag® plus peroxidase in situ apoptosis detection kit (Millipore, S7101) following the manufacturer’s instructions. After the TUNEL staining, an additional course of immunostaining for MYH3 was performed as described above. The fixation and blocking steps were skipped for the second staining.

We planned three different functional tests with the same mice. All tests were conducted within 1 week in the same order: muscle force first, grip test second, and rotarod test third. Non-transplanted and transplanted mice were the siblings of the same surrogate ICR mothers and were randomly selected into either group. They grew up together until weaning at 3 weeks. After weaning, the mice were divided into same-sized cages with five mice each.

Numerical data with three groups were analyzed with a one way-ANOVA followed by the Tukey-Kramer method. The student’s t-test was performed between two independent groups. *p < 0.05, **p < 0.01, and ***p < 0.001 are labeled in the figures. ^† was used when p-values were <0.1, indicating that the two groups tended to have a difference, but it was not statistically different.

All animal experiments were performed in accordance with the guidelines of the animal experiment committee and recombinant DNA experiment committee at Kyoto University. The study design was reviewed and accepted by the animal experiment committee in advance of conducting all the experiments (No.17-95-2).

First, we confirmed that the donor cells were positive for CD73, CD44, and CD105 and negative for CD45 and HLA-DR (Supplementary Figure S1A). The expression of collagen VI was also confirmed (Supplementary Figure S1B). Next, we evaluated the potential of human iMSCs to deliver collagen VI to the skeletal muscles in Col6a1 ^KO /NSG mice after neonatal i. p. injection. Two days after birth, 5 × 10⁶ luciferase-expressing iMSCs were transplanted to the peritoneal cavity, and their distribution was observed with an IVIS imaging system. The transplanted cells were widely delivered including to the lower and upper extremities after 24 h (Figure 1A). The total intensity of the luciferase signal decreased with time but was detectable for 28 days (Figure 1B). iMSCs were seen in major organs except for the brain based on luciferase transcript expressions at 7 days after the transplantation (Figure 1C). Fewer iMSCs were detected at 28 days.

Histological analysis revealed that the iMSCs had reached the interstitial space of the quadriceps and diaphragm 1 week after the transplantation (Figure 1D). The expression of restored collagen VI was also confirmed by Western blotting in the quadriceps and diaphragm after 2 weeks (Figure 1E). Collagen VI expression in the tissues suggested that iMSCs migrated to the skeletal muscle and secreted collagen VI locally.

To prove that iMSCs can migrate throughout the body via blood vessels, we performed intravenous (i.v.) transplantation from the facial vein 1 day after birth. The transplanted cells distributed to the whole body including the quadriceps and diaphragm, where they secreted collagen VI (Supplementary Figures S2A,B). Collagen VI secretion was sustained for more than 28 days in the limb muscles (Supplementary Figure S2C). Although the i.v. transplantation had comparable results with the i. p. transplantation, we chose to perform further experiments with i. p. transplantation because i. v. transplantation was technically difficult. Overall, the above findings indicate that transplanted iMSCs can deliver molecules into organs including skeletal muscle, suggesting a potential for treating systemic diseases.

Next, we searched for indicators to evaluate the effects of the transplantation on skeletal muscle. At 4 weeks, there was no positive effect on muscle weight after the transplantation (Supplementary Figure S3A), although the engraftment of donor cells and expression of collagen VI in the quadriceps and diaphragm were confirmed (Supplementary Figure S3B). A histogram of the diameters of the quadriceps revealed that the myofiber size was smaller in Col6a1 ^KO /NSG mice than in wild type (WT)/NSG mice, and its distribution shifted and became larger in the transplanted mice (Supplementary Figure S3C). The percentage of small myofibers and large myofibers decreased and increased, respectively (Supplementary Figure S3D), and the average size of the myofibers was increased after the transplantation (Supplementary Figure S3E). The area of the myofibers tended to increase (p = 0.064) (Supplementary Figures S3F,G).

When a boost i. p. injection of 5 × 10⁶ iMSCs was added at 4 weeks, the therapeutic effects of the transplantation were more clearly demonstrated at 8 weeks than at 4 weeks. Donor cells after the second i. p. transplantation were dominantly observed in the interstitium, suggesting that the transplanted cells were extravasated from the blood vessels to the quadriceps (Supplemental Figure S4A). An IVIS study confirmed similar iMSC dynamics with neonatal transplantation, with the period of the luciferase signal ranging from 2 to 4 weeks (Supplementary Figure S4B). Muscle weight at 8 weeks was increased in the transplanted mice compared to the non-transplanted mice, although BW was not significantly changed (Figure 2A). Donor cells were much less detected than at 4 weeks, but collagen VI expression had dispersed and remained in the quadriceps and diaphragm of the transplanted mice (Figure 2B). Histological analysis of the quadriceps showed the diameter and area of myofibers were increased with the transplantation (Figures 2C,D).

Fibrosis in skeletal muscle is abundant in adult Col6a1 ^KO mice (Noguchi et al., 2017). On the other hand, one of the concerns of injecting iMSCs is that iMSCs may induce fibrosis depending on their microenvironment because mesenchymal progenitor cells in skeletal muscle can be a source of both fat accumulation and fibrosis in diseased mice (Uezumi et al., 2011). To understand the impact of the iMSC transplantation on fibrosis formation, we performed several studies. Sirius red staining showed that the fibrotic change in the quadriceps was more prominent in non-transplanted Col6a1 ^KO /NSG mice than in WT/NSG mice but was ameliorated with the transplantation (Supplementary Figure S5A). Quantitative analysis of the fibrotic area also showed suppressed fibrosis in the transplanted mice (Figure 2E). The gene expression of Postn, a matricellular protein that contributes to fibrosis (Kanaoka et al., 2018), was suppressed in the diaphragm of transplanted mice (Supplementary Figure S5B), and reduced Periostin expression in the transplanted mice was confirmed by immunostaining (Supplementary Figure S5C).

In sum, iMSC-transplanted mice showed an increased myofiber size at 4 weeks. When the boost transplantation was added at 4 weeks, the muscle weight and myofiber size were increased at 8 weeks without any enhancement of fibrosis.

The expression of collagen VI was measured by Western blotting and immunofluorescence staining in the quadriceps and diaphragm. The expression level of collagen VI in the Western blotting was detected in all samples at 4 weeks, but undetectable in one of the four samples at 8 weeks (Supplementary Figure S6A). The average expression level in the quadriceps sample of the transplanted mice was 0.48% at 4 weeks and 0.24% at 8 weeks that of WT mice at 4 weeks. In the diaphragm, the average percentages were 1.2 and 0.88%, respectively (Supplementary Figure S6B).

The collagen VI-rescued area based on immunofluorescence staining was also calculated (Supplementary Figures S7A, S8A). The collagen VI expression gradually increased after neonatal transplantation, reaching a peak at around 4–5 weeks, and decreased thereafter in the quadriceps even with boost transplantation at 4 weeks (Supplementary Figure S7B). The collagen VI-restored area at 4 weeks in the quadriceps was 5.87%, which is equivalent to 16.3% of the collagen VI positive areas in 4-week-old WT/NSG mice (Supplementary Figures S7B,C). The average collagen VI-rescued area at 8 weeks in the quadriceps was 3.45% (equivalent to 11.0% in 8-week-old WT/NSG mice) after neonatal and boost iMSC transplantation at 4 weeks, but 1.03% (equivalent to 3.31% in 8-week-old WT/NSG mice) when only neonatal transplantation was performed (Supplementary Figures 7B,C). Boost transplantation at 4 weeks slowed the rate at which collagen VI degraded, but it did not maintain the collagen VI-rescued area at 4 weeks. The donor cells in the quadriceps decreased with time and were rarely detected after 8 weeks (Supplementary Figure S7d).

As for the diaphragm, the average collagen VI-rescued area was 5.7% at 4 weeks (equivalent to 22.2% in 4-week-old WT/NSG mice) and 3.6% at 8 weeks (10.4% in 8-week-old WT/NSG mice) (Supplementary Figures 8B,C). The number of donor cells in the diaphragm showed a similar pattern as in the quadriceps and decreased with time (Supplementary Figure S8D).

Collagen VI expression was barely observed in the quadriceps at 20 weeks following transplantations at the neonatal stage and boost transplantation at 4 weeks (Supplementary Figure S10A). The longest period the donor cells could engraft was not confirmed, but some donor cells were detected in the quadriceps at 20 weeks (Supplementary Figure S9B). No tumorigenesis was confirmed at 20 weeks in any of the 10 mice that received the two transplantations (neonatal and 4-week boost) (data not shown).

We then focused on the impact of iMSC transplantation on MYH3+ myofibers. MYH3, an embryonic myosin heavy chain, is the isoform expressed in regenerating fibers after injury or the early stage of muscle development (Schiaffino et al., 2015) and generally detected for 2–3 weeks in newly regenerating myofibers (Jerkovic et al., 1997; Kalhovde et al., 2005). The MYH3 protein expression level gradually decreases and switches to adult-type myosin heavy chain (Li et al., 2013; Yoshimoto et al., 2020). The position change of nuclei is another hallmark of muscle development (Roman and Gomes, 2018). During development, nuclei are spread longitudinally in the center of the myofiber and migrate peripherally toward the end part of the muscle (Roman and Gomes, 2018), which can be observed as a multinuclear myofiber in the section. In a previous study using muscle biopsies from UCMD patients, MYH3 positive regenerating muscle fibers were limited to small diameters (Higuchi et al., 2003), indicating that the impairment of muscle regeneration is one of the phenotypes of UCMD.

At 4 weeks, we observed the area where MYH3+ myofibers were aggregated, which was consistent with the area where collagen VI was expressed in transplanted mice, while MYH3+ myofibers in non-transplanted mice solely existed among mature myofibers (Figure 3A). In addition, myofibers weakly positive for MYH3 observed in the transplanted mice grew (Figure 3A, arrowheads). Therefore, we analyzed the correlation of the fluorescence intensity and size of MYH3+ myofibers to evaluate the difference in muscle regeneration between transplanted and non-transplanted mice. The number of large, weakly positive MYH3+ myofibers was increased in transplanted mice, whereas smaller, strongly positive MYH3+ myofibers were observed in non-transplanted mice (Figure 3B). The number of myofibers with multiple nuclei as well as the ratio of multinucleated to single nucleated myofibers were increased in transplanted mice (Figure 3C).

Next, in order to reveal the impact of iMSC transplantation on muscle stem (satellite) cells, we counted the number of Pax7+ and MyoD + nuclei in the quadriceps (Figure 4A). Pax7 is a marker of satellite cells, and Pax7+MyoD-cells are in the quiescent state (Gnocchi et al., 2009). On the other hand, MyoD is one of the earliest markers of myoblasts (Weintraub, 1993). Pax7+MyoD + cells are active satellite cells during the process of cell division, and MyoD + Pax7-cells are committed myoblasts (Chal and Pourquié, 2017). Muscle regeneration was more active at 4 weeks than at 8 weeks, during which time the total number of positive nuclei for these regeneration markers decreased in both WT/NSG and Col6a1 ^KO /NSG mice (Figure 4B). A difference in the number of MyoD + Pax7-cells was notable at 8 weeks between non-transplanted and transplanted mice. MyoD + Pax7-cells were upregulated in the non-transplanted mice up to 4 weeks, but thereafter decreased to a level close to WT/NSG mice at 8 weeks. Contrastingly, the count of MyoD + Pax7-cells in the transplanted mice decreased at a slower rate at 8 weeks than in the non-transplanted mice (Figures 4B,C). The total number of myofibers in TA muscle increased in transplanted mice as well as in WT/NSG mice, while it was decreased in non-transplanted mice from 4 to 8 weeks (Figure 4D).

The dynamics of MYH3+ regenerating myofibers followed the dynamics of MyoD + cells. While MYH3+ myofibers were rarely observed in WT/NSG mice at 4 and 8 weeks, they were scattered in both transplanted and non-transplanted Col6a1 ^KO /NSG mice at 4 weeks. However, at 8 weeks, larger MYH3+ myofibers were present in the transplanted mice but rarely detected in the non-transplanted mice (Figure 4E). The number of MYH3+ myofibers was not different between the transplanted and non-transplanted mice at 4 weeks, but higher in the transplanted mice at 8 weeks (Figure 4F).

In sum, the dynamics of myogenic cells and MYH3+ myofibers indicated that iMSC transplantation retained muscle regenerative capacity over longer periods in Col6a1 ^KO /NSG mice.

Ultrastructural alterations of mitochondria and spontaneous apoptosis are observed in the muscles of Col6a1 ^−/− mice (Irwin et al., 2003; Palma et al., 2009) and UCMD patients (Angelin et al., 2007). Mitochondria are involved in the pathogenesis of UCMD, as mitochondrial defects are common findings irrespective of the genetic mutation locus or mode of inheritance, but additional factors are involved in the susceptibility to apoptosis or regeneration (Angelin et al., 2007). Therefore, to elucidate the mechanism of the therapeutic effects of iMSC transplantation, we investigated mitochondria and apoptosis in the muscles of Col6a1 ^KO /NSG mice.

Regarding the morphology of the mitochondria, multiple abnormal mitochondria were often detected in the quadriceps of Col6a1 ^KO /NSG mice (Figure 5A), but they were hardly observed in transplanted mice (Figure 5B). Moreover, some images of the transplanted mice were consistent with the formation of the autophagosome (Figure 5B), which was not the case in non-transplanted mice. The number of abnormal mitochondria was reduced in the quadriceps (Figure 5C). Abnormal mitochondria were observed in the diaphragm of both non-transplanted and transplanted mice (Figure 5D), but they were fewer in the transplanted group (Figure 5E).

Apoptotic nuclei were counted in the TA muscles and diaphragm (Figures 5F,H). The number of apoptotic cells tended to be reduced in the TA muscles of transplanted mice (Figure 5G), and less apoptosis was identified in the diaphragm of the same group (Figure 5I). MYH3+ myofibers were simultaneously stained to evaluate the correlation between apoptosis and regenerating myofibers. Some TUNEL positive nuclei were observed in the MYH3+ myofibers in the TA muscles of both non-transplanted and transplanted mice at 4 weeks (Figure 5F), possibly reflecting the specific pathology of UCMD muscles, in which new myofibers during regeneration fail to grow and mature, undergoing apoptosis instead (Paco et al., 2012). On the contrary, TUNEL + MYH3+ myofibers were not detected in the transplanted mice at 8 weeks (data not shown) or at 4 weeks in the diaphragm (Figure 5H). Single-stranded DNA-positive nuclei were examined as an alternative marker for apoptosis. A similar distribution of single-stranded DNA positive nuclei was confirmed in the quadriceps and diaphragm of both non-transplanted and transplanted mice (Supplementary Figure S10).

These results confirmed abnormal mitochondria and apoptosis in Col6a1 ^KO/NSG mice at 4 and 8 weeks and that iMSC transplantation reduced the number of dysmorphic mitochondria and apoptotic cells.

According to a previous report, 14-week-old Col6a1 ^KO mice showed functional impairment in grip power, voluntary running, and muscle force (Noguchi et al., 2017). We compared the difference in motor functions in 8-week-old mice to evaluate if the transplantation therapy was functionally beneficial. Care was taken to minimize any environmental factors that might affect the results of the function tests (see Methods). Muscle force was measured with isometric contraction by both sides of the gastrocnemius muscle and was increased with iMSC transplantation (Figure 6A). Grip power was improved by 35% with the iMSC transplantation (Figure 6B). The average running time in the rotarod test was increased in the transplanted mice compared to the non-transplanted mice (Figure 6C). The percentage of mice that completed a 5-min run was 2.6% (1/30) in the non-transplanted group and 9.5% (4/42) in the transplanted group.