Generation of Skeletal Muscle Organoids from Human Pluripotent Stem Cells to Model Myogenesis and Muscle Regeneration

In vitro organoids derived from human pluripotent stem cells (hPSCs) have been developed as essential tools to study the underlying mechanisms of human development and diseases owing to their structural and physiological similarity to corresponding organs. Despite recent advances, there are a few methodologies for three-dimensional (3D) skeletal muscle differentiation, which focus on the terminal differentiation into myofibers and investigate the potential of modeling neuromuscular disorders and muscular dystrophies. However, these methodologies cannot recapitulate the developmental processes and lack regenerative capacity. In this study, we developed a new method to differentiate hPSCs into a 3D human skeletal muscle organoid (hSkMO). This organoid model could recapitulate the myogenesis process and possesses regenerative capacities of sustainable satellite cells (SCs), which are adult muscle stem/progenitor cells capable of self-renewal and myogenic differentiation. Our 3D model demonstrated myogenesis through the sequential occurrence of multiple myogenic cell types from SCs to myocytes. Notably, we detected quiescent, non-dividing SCs throughout the hSkMO differentiation in long-term culture. They were activated and differentiated to reconstitute muscle tissue upon damage. Thus, hSkMOs can recapitulate human skeletal muscle development and regeneration and may provide a new model for studying human skeletal muscles and related diseases.


Introduction
The skeletal muscle is a post-mitotic tissue that constitutes up to 50% of human body mass. It is composed of contractile multinucleated myofibers generated by fusing differentiated muscle cells [1]. Skeletal muscle possesses an enormous regenerative ability to repair damaged tissue resulting from physiological conditions, such as exercise and aging, or diseases, such as cachexia, sarcopenia, and muscular dystrophies. A small population of 2 of 14 myogenic stem/progenitor cells, known as satellite cells (SCs), play a role in the muscle fiber regenerative process [2][3][4]. Over the past decades, multiple attempts have been made to establish an appropriate model system to investigate embryonic and adult myogenesis regulated by the fate of SCs during muscle development and regeneration [5]. Despite recent progress, the necessity to develop novel in vitro experimental approaches to model skeletal muscle regeneration remains.
Recent advances in human pluripotent stem cell (hPSC)-based 3D in vitro organoid technology have provided new opportunities to understand human development and diseases associated with various organs, including the brain, spinal cord, liver, kidneys, and pancreas [6][7][8][9][10]. Despite its success in developing a wide range of human tissues, most differentiation protocols of hPSCs into skeletal muscle cells are solely based on two-dimensional (2D) methods. These conditions are unlikely to recapitulate the cytoarchitecture of developing skeletal muscle tissue with diverse myogenic cells, including SCs and myoblasts [11][12][13]. Several studies have reported the generation of 3D skeletal muscle tissues from hPSCs and modeled neuromuscular disorders and muscular dystrophies [14,15]. However, generating skeletal muscle-specific 3D tissue that can retain muscle regenerative capacities through damaged tissue repair remains a challenge.
Here, we introduce a novel method to recapitulate the features of embryonic myogenesis by paraxial mesodermal differentiation of hPSCs into human skeletal muscle organoids (hSkMOs), exhibiting structural organization of muscle cells and the different stages of heterogeneous myogenic cells. Notably, hSkMOs produce sustainable pair box 7-positive (PAX + ) SCs in long-term cultures capable of generating muscle fibers contributing to muscle repair upon acute injury. Thus, hSkMOs may be valuable as an advanced muscle regeneration model and a viable platform for testing novel therapeutics.

Generation and Characterization of hSkMOs from hPSCs
To generate hSkMOs, we developed a stepwise, pre-patterning differentiation protocol that guides hPSCs toward the paraxial mesodermal lineage, giving rise to myogenic progenitor cells and myoblasts ( Figure 1A). First, hPSCs were dissociated into single cells and allowed to form uniformly sized (approximately 500 µm in diameter) embryoid bodies (EBs) in low-attachment V-shaped 96-well plates ( Figure 1A). In our direct induction approach, we promoted paraxial mesodermal differentiation with WNT activation (CHIR99021), bone morphogenetic protein (BMP) inhibition (LDN193189), and fibroblast growth factor 2 (FGF2) signaling to generate presomitic cells in 3D culture conditions. Upon initial exposure to exogenous small molecules and growth factors, the time-course expression of stage-specific markers was profiled using quantitative reverse transcription-polymerase chain reaction (qRT-PCR) analysis on days 0, 5, 10, and 20. The expression of pluripotency markers octamer-binding transcription factor 4 (OCT4) and NANOG rapidly decreased, and the expression of Brachyury (mesoderm marker), T-Box transcription factor 6 (TBX6), and mesogenin 1 (MSGN1; presomitic markers) significantly increased within five days ( Figure 1B). To further characterize paraxial mesodermal differentiation, we performed immunostaining on day five against TBX6, which demonstrated successful initial paraxial mesodermal commitment ( Figure 1C).
After paraxial mesodermal induction for seven days, each organoid was embedded with growth factor-reduced Matrigel to provide an environment for enhancing 3D organization and then transferred to a six-well plate on an orbital shaker, similar to our previous organoid culture system ( Figure 1A) [6,16]. The organoids were subjected to growth factor application to promote myogenic differentiation. Hepatocyte growth factor (HGF) and insulin-like growth factor 1 (IGF1) were added to the myogenic specification media, and hSkMOs were cultured until the day of analysis. The use of an orbital shaker significantly improved the viability, survival, and differentiation of hSkMOs increasing penetration rate of oxygen and nutrients into the core area of hSkMOs [17,18]. These myogenic organoids gradually grew to more than 1.5 mm in diameter by day 60 ( Figure 1D,E). They appeared round-shaped, uniformly sized, and had relatively homogenous morphology; thus reducing batch-to-batch variability in size ( Figure 1D-G). We reasoned that the myogenic progenitor markers, PAX3 and PAX7, would be expressed during myogenic specification because the qRT-PCR data demonstrated that the PAX3 expression gradually increased around day 20 ( Figure 1B). We then performed cryosections to verify the hSkMO development on day 30. We observed that 43% of PAX3 + and 63% of PAX7 + cells were detected. Intriguingly, the myogenic cells appeared as clusters, and approximately 9% of PAX7 + cells were double-positive for Ki67 at day 30 ( Figure 1H-K), demonstrating that proliferating cells are myogenic progenitors in hSkMOs. These data indicate that our in vitro hSkMO culture system recapitulates the features of embryonic skeletal muscle development through the differentiation of paraxial mesoderm into myogenic tissue with the expression of crucial marker genes. rate of oxygen and nutrients into the core area of hSkMOs [17,18]. These myogenic organoids gradually grew to more than 1.5 mm in diameter by day 60 ( Figure 1D,E). They appeared round-shaped, uniformly sized, and had relatively homogenous morphology; thus reducing batch-to-batch variability in size ( Figure 1D-G). We reasoned that the myogenic progenitor markers, PAX3 and PAX7, would be expressed during myogenic specification because the qRT-PCR data demonstrated that the PAX3 expression gradually increased around day 20 ( Figure 1B). We then performed cryosections to verify the hSkMO development on day 30. We observed that 43% of PAX3 + and 63% of PAX7 + cells were detected. Intriguingly, the myogenic cells appeared as clusters, and approximately 9% of PAX7 + cells were double-positive for Ki67 at day 30 ( Figure 1H-K), demonstrating that proliferating cells are myogenic progenitors in hSkMOs. These data indicate that our in vitro hSkMO culture system recapitulates the features of embryonic skeletal muscle development through the differentiation of paraxial mesoderm into myogenic tissue with the expression of crucial marker genes.  (J) Immunostaining of hSkMOs for PAX7 and Ki67 at day 30. Scale bar, 20 µm. (K) Quantification of the percentage of PAX7 + /Ki67 − and PAX7 + /Ki67 + ; mean ± s.e.m., n = 3. Statistical analysis was performed using one-way analysis of variance (ANOVA), followed by Tukey's multiple comparison test. ** p < 0.01, *** p < 0.001; n.s., not significant.

Identifying and Characterizing Different Types of Skeletal Muscle Stem/Progenitor Cells during Myogenesis in hSkMOs
In both embryonic and adult myogenesis, myogenic stem/progenitor cells, such as SCs, are the major cell populations involved in skeletal muscle differentiation and regeneration [5]. SCs reside in a specific muscle stem cell niche on the muscle fiber and remain quiescent until muscle damage-induced activation during acute injury or environmental conditions [2]. Transcription factors from the PAX gene family, PAX3 and PAX7, are crucial for myogenesis and SCs determination. Additionally, muscle regulatory transcription factors (MRFs) such as myoblast determination protein 1 (MYOD) and myogenin (MYOG) are critical for myogenic progenitor cell determination and terminal differentiation ( Figure 2A) [19]. To check the expression of cell type-specific markers, such as skeletal muscle stem/progenitor and differentiating cells, we performed qRT-PCR analysis on hSkMOs on days 30 and 60. PAX3 and PAX7 (SC markers) expression peaked on day 30 ( Figure 2B). The expression of MYOD (proliferating and activated SC marker) and MYOG (differentiated myocyte marker) increased over time ( Figure 2B). These data demonstrate that PAX3-and PAX7-expressing muscle stem/progenitor cells are the major population during the early stage of myogenesis and remain as resident SCs in mature skeletal muscle tissues. Activated muscle progenitor cells and differentiated myoblasts are augmented over time.
To further characterize and quantify the specific myogenic populations within the hSkMOs, we conducted immunohistochemistry analysis and observed distinct cell typespecific marker expression in hSkMOs at day 70. MYOD − /PAX7 + , MYOD + /PAX7 + , and MYOD + /Ki67 + cells accounted for 29%, 6%, and 8% of the putative quiescent, activated, and proliferating SCs, respectively. MYOD + /PAX7 − cells constituted 39% of differentiating myoblasts ( Figure 2C-F). Additionally, MYOG − /PAX7 + cells constituted 23% of putative quiescent SCs, and MYOG + /PAX7 − cells accounted for 30% of differentiated myocytes ( Figure 2G-J). Intriguingly, we observed that 8% and 6% of the MYOG + cells in hSkMOs co-expressed PAX7 and Ki67, respectively ( Figure 2G-J). Collectively, these data demonstrate that our hSkMOs recapitulated myogenesis progression and produced myogenic stem/progenitor and differentiated cells during long-term culture. days 30 and 60. PAX3 and PAX7 (SC markers) expression peaked on day 30 ( Figure 2B). The expression of MYOD (proliferating and activated SC marker) and MYOG (differentiated myocyte marker) increased over time ( Figure 2B). These data demonstrate that PAX3and PAX7-expressing muscle stem/progenitor cells are the major population during the early stage of myogenesis and remain as resident SCs in mature skeletal muscle tissues. Activated muscle progenitor cells and differentiated myoblasts are augmented over time.

Maturation of Skeletal Muscle Cells in hSkMOs
hSkMOs grew exponentially in size within 2 months, presumably by increasing myogenic stem/progenitor and mature muscle cells, and the organoid growth rate steadily diminished in size from 2 months ( Figure 1D). This result demonstrated that embryonic myogenesis for the development of skeletal muscle tissue was complete and that hSkMOs consisted of a large population of terminally differentiated myogenic cells and a small population of preserved myogenic stem/progenitor cells ( Figure 2C-J). We performed scanning electron microscopy (SEM) imaging ( Figure 3A) to further examine the cytoarchitecture of hSkMOs after 2 months. Confocal microscopic images revealed the presence of abundant myofiber-like structures ( Figure 3B-F). We speculated that the hSkMOs should contain paraxial mesodermal progenitor-derived myogenic cells and neuromesodermal progenitor-derived neurons following early embryonic mesodermal development. Thus, we performed immunohistochemical analysis to evaluate the proportion of myogenic and neuronal cells ( Figure 3C,D). We confirmed that hSkMOs consisted of a substantial proportion of TITIN + muscle cells, and microtubule-associated protein 2 (MAP2)-positive neurons were observed within a particular portion of hSkMOs ( Figure 3C). To quantify the area of skeletal muscle tissue within these hSkMOs, we measured the fluorescence coverage of TITIN and MAP2 from confocal microscopy images. The results indicated that 62% of the TITIN + skeletal muscle region and 20% of the MAP2 + neuronal region were distinctively separated ( Figure 3D). Furthermore, we analyzed the expression levels of fibro-adipogenic progenitors (FAPS) markers, platelet-derived growth factor receptor A (PDGFRα), and PR domain containing 16 (PRDM16) using qRT-PCR analysis. The expression levels of PDGFR and PRDM16 were drastically increased on day 60 (Supplementary Figure S1). ber formation and maturation of skeletal muscle tissue. To define muscle maturation during the long-term culture of the hSkMOs, we performed immunohistochemistry analysis. The results revealed that highly organized skeletal muscle exhibited numerous sarcomeric bands labeled by TITIN antibody (Figure 3E,F) as well as myotube maturation marker by myosin heavy chain (MyHC) antibody ( Figure 3G). Moreover, we observed that MYOD + and MYOG + myogenic cells coexisted within TITIN + myofibers ( Figure 3H,I). To further characterize the presence of sustainable SCs within mature hSkMOs, we quantified the amount of dormant SCs by confocal microscopy imaging. The data revealed that approximately 56%, 31%, and 5% of PAX7 + /Ki67 -putative dormant SCs existed throughout differentiation of hSkMOs at days 30, 70, and 130, respectively ( Figure 3J,K). Hence, these data demonstrate that structurally well-developed myofibers within hSkMOs expressing sarcomeric protein and terminally differentiated myocyte markers demonstrate mature skeletal muscle properties. Additionally, dormant PAX7 + SCs were sustained in hSkMOs over time, indicating the regenerative potential of our in vitro 3D skeletal muscle tissue. Skeletal muscles are composed of mature myofibers containing numerous myofibrils, and sarcomeres are the basic functional units of skeletal muscles [20]. The expression of sarcomeric proteins and the organization of sarcomeric structures should indicate myofiber formation and maturation of skeletal muscle tissue. To define muscle maturation during the long-term culture of the hSkMOs, we performed immunohistochemistry analysis. The results revealed that highly organized skeletal muscle exhibited numerous sarcomeric bands labeled by TITIN antibody (Figure 3E,F) as well as myotube maturation marker by myosin heavy chain (MyHC) antibody ( Figure 3G). Moreover, we observed that MYOD + and MYOG + myogenic cells coexisted within TITIN + myofibers ( Figure 3H,I). To further characterize the presence of sustainable SCs within mature hSkMOs, we quantified the amount of dormant SCs by confocal microscopy imaging. The data revealed that approximately 56%, 31%, and 5% of PAX7 + /Ki67 − putative dormant SCs existed throughout differentiation of hSkMOs at days 30, 70, and 130, respectively ( Figure 3J,K). Hence, these data demonstrate that structurally well-developed myofibers within hSkMOs expressing sarcomeric protein and terminally differentiated myocyte markers demonstrate mature skeletal muscle properties. Additionally, dormant PAX7 + SCs were sustained in hSkMOs over time, indicating the regenerative potential of our in vitro 3D skeletal muscle tissue.

Toxin-Induced Muscular Degeneration and Regeneration Capacity of hSkMOs
Given the presence of sustainable SCs in mature hSkMOs ( Figure 3J,K), we expected hSkMOs to have regenerative potential by activating quiescent SCs and subsequent differentiation into myocytes upon muscle damage. The cardiotoxin (CTX) injury model, which induces myolysis, has been widely used to induce skeletal muscle damage in mouse and cellular models [21,22]. Treatment with CTX in culture media induces muscle inflammation and myofiber degeneration followed by a regenerative response. The damaged myofibers are typically restored in 2-3 weeks after CTX-induced damage [21]. With this notion, we examined muscle damage in hSkMOs by treatment with CTX and the intrinsic regenerative ability of SCs ( Figure 4A). Adding CTX (0.1 µM) in the culture media induced acute muscle damage in hSkMOs. We observed a drastic reduction in PAX7 + and MYOD + cells in hSkMOs through immunohistochemistry analysis ( Figure 4B,C). Although non-dividing quiescent SCs (PAX7 + /Ki67 − ) were reduced by more than half in the CTX treatment group compared to the control group, PAX7 + /Ki67 + cells were retained ( Figure 4D,E).  Several studies have shown that other cell types, such as mesoangioblasts, vesselassociated muscle progenitors, fibroadipogenic progenitors, and macrophages, contribute to myogenesis in the SC niche [23,24]. Given the limitations of cell-type compositions for non-myogenic cells and to accelerate muscle regeneration in our hSkMOs induced by CTX treatment, we supplemented cytokines released from the other cells in the SC niche upon muscle damage. Interleukin-4 (IL-4) is secreted from non-myogenic interstitial stem cells present in adult muscle and improves skeletal muscle regeneration and maturation of myotubes [23,25]. IL-4 was added to the medium to promote muscle regeneration in CTX-injured hSkMOs for 14 days to evaluate the extent of muscle regeneration. We then immunostained the hSkMOs using antibodies against PAX7 and MYOG to check for muscle regeneration of augmented SCs and myocytes. The CTX-untreated control hSkMOs showed a considerable presence of PAX7 + SCs (32%) and MYOG + myocytes (41%) (Figure 4F,G). Intriguingly, we observed a significant increase in MYOG + myocytes (26%) in the CTXinjured hSkMOs with the treatment of IL-4 compared to that in MYOG+ myocytes (9%) in CTX-injured hSkMOs ( Figure 4F,G). Collectively, these data indicate that hSkMOs exhibited regenerative potential upon muscle damage and were able to mimic adult myogenesis promoted by cytokine treatment.

Discussion
Over the past decades, various animal models of muscle degeneration, including cytotoxin-induced injury, crush injury, and transgenic mice carrying muscular dystrophyassociated genetic mutations, have been utilized as standard muscle regeneration models for basic research. Animal models have been invaluable for studying the mechanisms of muscle regeneration regulated by SCs. However, they do not faithfully recapitulate a diverse range of disease phenotypes, and the lack of human-like pathology limits their application in preclinical research [26]. In this context, we suggested the establishment of reliable human muscle cell-based in vitro modeling systems.
In this study, we demonstrated that hPSCs could be committed to the paraxial mesodermal lineage and further differentiate into 3D hSkMOs comprising sustainable SCs and distinct myofibers with sarcomere protein expression and structural organization. Most approaches to skeletal muscle differentiation have been developed based on 2D culture systems to differentiate hPSCs into myofibers [11][12][13]. Although 2D culture systems are straightforward and valuable for in vitro modeling studies, the lack of a native microenvironment and SC niche restricts their application to model adult myogenesis and muscle regeneration [27].
In our protocol, we utilized WNT activator and BMP inhibitors at the beginning of the EB differentiation to improve paraxial mesodermal induction, combined with stimulation of FGF2 signaling upon Matrigel embedding to promote the 3D structural organization. HGF and IGF1 were later added to accelerate myogenic specification and further myofiber differentiation. Particularly, the myogenic tissues were poorly developed when hSkMOs were embedded into Matrigel either at the beginning of EB differentiation or in the later stages. Thus, we optimized the specific time for Matrigel embedding at day seven ( Figure 1A). We observed neural cells presumably derived from neuromesodermal progenitors. FGF2 was withdrawn to arrest neural lineage to achieve exclusive muscle tissue development, and HGF and IGF1 treatment were prolonged to propagate myogenic progenitors [28]. Remarkably, based on immunohistochemical analysis, we found that 62% of the proportion was skeletal muscle tissue based on measuring the TITIN + striated muscle tissue area ( Figure 3D). Within the skeletal muscle area in hSkMOs, we observed about 31-35% of PAX7 + myogenic stem/progenitor cells, 40-44% of MYOD + activated/committed myoblasts, and 38-41% of MYOG + myocytes ( Figure 2C-J). Notably, 31% of PAX7 + /Ki67 − and 29% of MYOD − /PAX7 + non-dividing quiescent SCs were found in mature hSkMOs (Figures 2 and 3). These features indicated that our hSkMOs could effectively recapitulate embryonic myogenesis and possess regenerative potential regulated by sustainable SCs. Using comprehensive immunohistochemical analysis at multiple stages, we demonstrated the presence of different myogenic cells. However, future studies using a single-cell RNA sequencing approach may be required to further characterize different types of cells existing in hSkMOs by cell type-specific comparative transcriptomic analysis [14].
In homeostatic skeletal muscles, SCs generally remain in a quiescent state [5]. In response to muscle injuries, SCs are activated and promote myogenic regenerative activity to repair damaged tissues [29,30]. Other cell types can contribute to myogenesis, and SCs are affected by non-myogenic cells that secrete several cytokines such as IL-4 and IL-13 [31,32]. In particular, IL-4 can influence the inflammatory system, promote an anti-inflammatory muscle microenvironment, and contribute to SCs differentiation; thus promoting muscle regeneration [2]. Although the de novo organoids generated from hPSCs have tremendous potential in the study of organ development and pathogenesis, they do not fully recapitulate the in vivo native microenvironment. This includes vasculature formation and cellular composition, including immune and non-myogenic cells. Upon muscle damage, various cells in the SC niche orchestrate myogenesis and promote muscle regeneration through interactions with their niche. As these hSkMOs are devoid of immune cells such as macrophages, treatment with extrinsic cytokines in the culture media was necessary to promote muscle regeneration. They demonstrated accelerated regeneration of damaged muscle tissue. The most interesting aspect of our hSkMOs is their remarkable potential to foster in vitro model systems by co-culture with other cell types or tissues and investigate their interactions. Additionally, hSkMOs will be invaluable for evaluating the efficacy of therapeutic candidates, such as small molecules, growth factors, and cytokines.
In conclusion, we successfully established a new method to generate hSkMOs that recapitulate embryonic myogenesis and demonstrate their muscle regeneration potential. Our hSkMOs could serve as a versatile platform to study aspects of human muscle biology with environmental factors and intrinsic modulation in 3D culture conditions because of the accessibility of monitoring the properties of muscle tissue. Additionally, hSkMO could provide a more feasible in vitro model system to investigate the molecular and cellular mechanisms of muscle regeneration and identify novel therapeutic candidates. hSkMOs would allow us to perturb multiple pathways by CRISPR/Cas9-mediated genetic mutations and/or chemical insults to recapitulate muscle degeneration and progressive pathological features of muscle-wasting disorders. This would facilitate further disease modeling studies and large-scale drug tests/screens for a broad array of skeletal muscular disorders, including Duchenne muscular dystrophy and sarcopenia.

Immunofluorescence Analyses Using Cryosection
For immunofluorescence analyses, hSkMOs were washed with phosphate-buffered saline (PBS) twice and then fixed with 4% paraformaldehyde (PFA) at 4 • C overnight. Fixed hSkMOs were washed with PBS then placed in PBS containing 30% sucrose solution until hSkMOs were dispersed in solution. Subsequently, hSkMOs were cryo-embedded in optimal cutting temperature (OCT) compound (Sakura Tissues Tek) and sectioned at a thickness of 16 µm using a cryostat (Leica CM1950). Cryosectioned slices were washed with PBS before permeabilization with 0.5% Triton X-100 in PBS for 1 h. The slices were blocked with a blocking buffer consisting of 10% donkey serum in PBS with 0.3% Triton X-100 and 1% bovine serum albumin (BSA) for 1 h. Primary antibodies were diluted in antibody buffer consisting of 2% donkey serum in PBS with 0.1% Triton X-100 at 4 • C overnight. The primary antibodies and their dilutions are listed in Table 1. On day two, the slices were washed with PBS containing 0.1 % Tween-20 twice every 5 min. Secondary antibodies diluted in antibody buffer were applied to the sections for 1 h at room temperature. After 1 h, the sections were washed twice for 5 min each before mounting. Alexa Fluor 488 and 594-conjugated donkey antibodies (Invitrogen) were used at 1:1000 dilution. Images were captured using a confocal microscope (Zeiss LSM 880) in Airyscan mode. Sample images were prepared using ImageJ software (NIH). Total RNA was isolated from the hSkMOs using RNAiso Plus containing 38% phenol (#9109; Takara Bio Inc., Shiga, Japan). Then, 1 µg of RNA was reverse transcribed into first-strand cDNA using the TOPscript™ RT DryMIX (dT18 plus) (Enzynomics Inc., Daejeon, Korea) and T100™ Thermal Cycler (Bio-Rad, Hercules, CA, USA) according to the manufacturer's instructions. For qRT-PCR analysis, amplification reactions were performed using SYBR Green TOPreal ™ qPCR 2X PreMIX (Enzynomics) on a CFX96™ Real-Time System, C1000™ Touch Thermal Cycler (Bio-Rad) according to the manufacturer's protocol. The primer sequences used for qRT-PCR are listed in Table 2.

Scanning Electron Microscopy (SEM)
For electron microscopy analysis, hSkMOs were fixed in glutaraldehyde (2.5%, v/v) in phosphate buffer (0.1 M; pH7.4) for 30 min and washed thrice every 10 min using 0.1 M phosphate buffer. After primary fixation, specimens were post-fixed in osmium tetroxide (1%, w/v) in phosphate buffer (0.1 M; pH7.4) for 30 min and rinsed with deionized water thrice every 10 min. Subsequently, fixed specimens were dehydrated with a series of graded ethanol from 70 to 100% for 10 min each. After dehydration, the specimens were dried at −20 • C overnight and then coated with gold (10 nm thick) for 2 min. The sputter-coated specimens were visualized for imaging analysis using an SEM (JSM7900F, JEOL, Tokyo, Japan) equipped with a cold cathode emitter set at an acceleration voltage of 5.0 kV.

Muscle Injury Procedures and Regeneration
The hSkMOs on day 50 were treated with 0.1 µM cardiotoxin (CTX) for 24 h at 37 • C. For the regeneration model, hSkMOs injured by CTX were incubated with or without IL-4 (10 ng/mL) for 14 days. All samples were subjected to immunofluorescence analysis. The detailed procedure for the muscle regeneration model of hSkMOs is shown in Figure 4A.

Quantification and Statistical Analyses
Statistical analyses were performed using GraphPad Prism software V5. "n" indicates the number of independent experiments, which were conducted at least in triplicate. The results are expressed as mean ± standard error (exact n numbers are described in the figure legends). Statistical analyses were followed by an unpaired Student's t-test when comparing two samples and by one-way ANOVA with Tukey's multiple comparison test when comparing more than two samples.