Porcine myogenesis in cloned wildtype and MYF5/MYOD/MYF6-null porcine embryo

Abstract

The pig is an important animal model increasingly used for biomedical research, particularly in transplantation strategies involving xenotransplantation or the development of human organs in pig for exotransplantation. Pigs, however, are less characterized than other animal models. In this study, we produced wildtype (WT) pig embryos via somatic cell nuclear transfer (SCNT) technology and compared them to skeletal muscle null embryos (lacking MYF5/MYOD/MYF6) at embryonic day 41, 62, and 90, critical stages of porcine myogenesis. Magnetic resonance imaging (MRI) and histological analyses revealed progressive development of skeletal muscle in WT embryos but not in null embryos whereas development of viscera progressed equally in both groups. Molecular analyses highlighted dynamic changes in myogenic gene expression and myofiber formation, demonstrating an organized progression of myogenesis in WT embryos. Morphologically, the null embryos exhibited abnormalities, including marked edema and underdeveloped limbs. MRI revealed severe skeletal abnormalities, including the absence of ribs, sternum, and associated vertebral malformations. In addition, histological analysis confirmed the complete lack of myofiber formation. Immunohistochemical analysis revealed the absence of myogenic stem cells and muscle differentiation, and RNA sequencing demonstrated that the skeletal muscle development process was entirely disrupted in null embryos. Additionally, analysis of neuromuscular junctions (NMJs) in the null embryos revealed that functional NMJ formation was absent, consistent with the lack of skeletal muscle formation. Importantly, these defects culminated in embryonic lethality after day 62 in the null embryos. We determined that the myogenic regulatory gene cascade is crucial for porcine embryo development and viability. The deletion of skeletal muscle is essential for the creating a vacant niche to allow for complementation of null porcine embryos with human induced pluripotent stem cells. Characterization of this skeletal muscle null pig model provide an important platform for engineering humanized muscle in gene-edited pigs.

Introduction

Skeletal muscle development is a highly orchestrated process that is critical for the formation and function of the musculoskeletal system¹. This complex process involves the proliferation, differentiation, and fusion of myogenic precursor cells into multinucleated myofibers, which ultimately form functional skeletal muscle. The basic-helix-loop-helix (bHLH) class of DNA-binding proteins are master regulators and include muscle-specific transcription factors MYOD, MYF5, and MYF6 (MRF4)^2,3. The role of each of these factors has been elegantly defined using targeted gene disruption technologies in mice. Primary myogenic regulatory factors (MRFs), such as MYF5 and MYOD, are essential for the determination of skeletal myoblasts, while secondary MRFs, including MYOG and MYF6, function later as differentiation factors⁴. Interestingly, mice carrying single mutations in Myf5, Myod, or Myf6 develop normal skeletal muscle with typical patterning, largely due to compensatory mechanisms within the MRF family^5,6,7. However, double mutations in Myod and Myf5 result in the complete absence of myoblasts and skeletal myofibers, demonstrating their critical roles in myogenic determination⁸. Furthermore, Kassar-Duchossoy et al. identified Myf6 as an additional determination factor in myogenesis⁹. Their findings revealed that skeletal muscle could form in Myf5/Myod double-null mice only when Myf6 expression remained intact, highlighting its essential role as both a determination and differentiation gene. Therefore, combined deletion of Myf5/Myod, and Myf6 resulted in the complete absence of skeletal muscle in mice⁹. These extensive studies in mouse models have elucidated the distinct functions of these MRFs, leading to the development of skeletal muscle knockout models^4,8,9. However, gene disruption studies targeting MRFs have not yet been conducted in pigs.

Pigs (Sus scrofa) are increasingly recognized for their anatomical and physiological similarities to humans, making them invaluable for biomedical research, especially in the context of xenotransplantation¹⁰. Despite their importance, pigs remain less well characterized than rodents. While the development of the cardiovascular system and abdominal organs in pig embryos have been extensively studied from the early to late stages of gestation^11,12, research on the embryonic development of skeletal muscles remains limited and there are no studies related to somatic cell nuclear transfer (SCNT) produced porcine embryos¹³. The lack of research on MRF functions and skeletal muscle knockout models in pigs limits our understanding of porcine myogenesis and the potential applications of pig models in translational research. Based on findings from mouse genetic studies on MRFs^5,6,7,8,9, we applied a triple knockout of MYF5/MYOD/MYF6 in pigs in the current study to eliminate all compensatory mechanisms of MRFs. This model allowed us to comprehensively investigate the consequences of eliminating the myogenic lineage and advance our understanding of skeletal muscle development in a translationally relevant large animal model.

In our previous studies, we produced pigs lacking skeletal muscle as the first step in humanizing this lineage to engineer whole mature muscle for the treatment of injury induced muscle loss, and examined these MYF5/MYOD/MYF6-null embryos up to embryonic day 35 (E35)¹⁴. In the current study, we comprehensively defined the developmental dynamics of skeletal muscle and internal organ systems in cloned wildtype (WT) porcine embryos at critical stages of myogenesis in pig. The selected time points (E41, E62, and E90), correspond to key phases of porcine myogenesis. The initial phase of skeletal muscle development, known as primary myogenesis, takes place from gestational days 30–60, a process where myoblasts initially fuse to create primary myotubes¹³. In the subsequent phase, termed secondary myogenesis, which occurs between days 54–90, myoblasts continue to fuse using primary myotubes as a scaffold to establish secondary myotubes¹³. To investigate these developmental stages, we cloned WT pig embryonic fibroblasts via SCNT technology and harvested the cloned WT embryos at E41, E62, and E90 days of gestation. Furthermore, we have extended the characterization of MYF5/MYOD/MYF6-null embryos to E62. Using age-matched cloned WT embryos as comparators, we investigated the phenotype of porcine embryos lacking three essential myogenic bHLH factors—MYF5, MYOD, and MYF6—produced via SCNT. Additionally, we evaluated the development of the neuromuscular junction (NMJ) in both WT and MYF5/MYOD/MYF6-null embryos to determine the role of the formation of skeletal muscle and expression of post-synaptic acetylcholine receptors in the course of development. Through detailed morphological and histological examinations, as well as RNA sequencing, we compared the development of these null embryos to their WT counterparts, highlighting the critical role of these MRFs in skeletal muscle formation.

The development of the MYF5/MYOD/MYF6-null pig model and its systematic characterization at defined stages not only fills the gap in porcine myogenesis research but also provides a valuable platform for producing xeno- or exogeneic organs through blastocyst complementation. This lineage (skeletal muscle) deletion strategy in these embryos creates a vacant niche for exogenous donor cells to populate and form human skeletal muscle in the absence of competition from the host.

Results

Developmental dynamics of muscle and internal organ systems in cloned WT porcine embryos

We harvested cloned WT embryos at E41, E62, and E90 days of gestation, which are important stages in porcine myogenesis¹³. The morphology of embryos was phenotypically normal, and the growth rates, determined by crown-to-rump measurements, indicated they were comparable to naturally bred porcine embryos (Fig. 1a and Supplementary Table 1)¹⁵.

Fig. 1: Developmental progression in cloned wildtype (WT) porcine embryos.

a The morphological overview of WT porcine embryos at E41, E62, and E90, demonstrating a developmental progression in physical form and size. Scale bar, 5 cm. b–d Parasagital scans of magnetic resonance imaging (MRI) at E41 (b, b′, b″), E62 (c, c′, c″), and E90 (d, d′, d″) for detailed visualization of organ development and evaluation of fetal morphometric parameters. Scale bar, 1 cm (b, b′, b″), 2 cm (c, c′,c″), and 4 cm (d, d′, d″). e–g Representative H&E staining of hindlimb sections across gestational stages E41 (e), E62 (f), and E90 (g). High magnification images of the boxed area are presented in (e′, f′, g′). Scale bar, 1 mm (e, f, g) and 250 µm (e′, f′, g′). h–j Representative Masson’s Trichrome staining demonstrates the progression of connective tissue development within the muscle across the same developmental stages E41 (h), E62 (i), and E90 (j). Boxed areas are enlarged in (h′–j′). Scale bar, 1 mm (h, i, j) and 250 µm (h′, i′, j′).

To assess the morphological development of internal organogenesis in WT embryos, we conducted magnetic resonance imaging (MRI) at E41, E62, and E90 days of gestation (Fig. 1b, c, d). At E41 days of gestation, the body of the WT embryo was clearly delineated and defined into thoracic and abdominal cavities. E41 embryos demonstrated well defined cardiac structures, including visualization of the atria, ventricles and great vessels. Lung tissue was defined with early lobulation (Fig. 1b, b′, b″). Abdominal organs that were clearly identified included the liver, left and right kidney, stomach and small intestines (Fig. 1b, b′, b″). The stomach and small intestines contained fluid intense material, indicating a swallowing reflex was present. Within the cranium, the brain parenchyma was noted; however, further anatomical differentiation was not noted (Fig. 1b″). Fluid intense cavities were noted in the cranium, representing ventricular and subarachnoid fluid although distribution was disorganized. Neurological tissue, consistent with spinal cord was visualized within the vertebral canal. The globes and lenses were identified with a typical shape and intensity, however, resolution of the various chambers of the globe was not noted. The musculoskeletal system was deemed appropriately developed with identification of the skull, vertebrae, ribs, sternum and proximal bones of the pelvic and thoracic limbs (Fig. 1b′, b″). The osseous structures from the tarsi and carpi distally were not identified with certainty. Muscles of the axial and appendicular skeleton were visualized with normal imaging features, however, individual muscles nor myotendinous junctions could be resolved. Hypointense tissue in the anatomical location of the diaphragmatic crura was visualized. A thin rim of hyperintense, subcutaneous tissue was identified, representing adipose tissue.

Organogenesis progressed appropriately, and E62 embryos had enhanced identification of the great vessels (Fig. 1c). The lung parenchyma showed increased lobulation and the trachea and bronchi were fluid-filled. The esophagus was visualized as a fluid-filled structure and a prominent junction with the stomach, likely the cardia, was appreciated. Renal parenchyma showed corticomedullary distinction (Fig. 1c′). The renal arteries and veins were clearly visualized and interlobular renal vessels were noted. The left and right adrenal glands were visualized. The gall bladder was identified as a fluid-distended cystic structure (Fig. 1c′). The small and large intestines could be resolved at this stage with fluid present through the gastrointestinal tract (Fig. 1c′). The brain could be delineated into the cerebrum, cerebellum, brain stem and pons with the brain parenchyma being homogenously hypointense lacking white matter—gray matter distinction (Fig. 1c). The lateral ventricles were identified and a moderate amount of subarachnoid fluid was present. The anterior and posterior chambers of the globes could be resolved at this stage of development (Fig. 1c″). The axial and appendicular skeleton was completely visualized with appropriate conformation (Fig. 1c″). Progression of muscular development allowed individual muscles of the thoracic and pelvic girdle and corresponding limbs to be resolved. The myotendinous junctions could not be resolved. The diaphragmatic crura could be visualized (Supplemental Fig. 1a). A thin rim of subcutaneous fat was noted in all embryos. Dental structures and the tongue were visible (Fig. 1c″ and Supplemental Fig. 1a′).

At E90, coronary arteries, the thymus, enhanced branching of the pulmonary arteries, and veins toward the lung lobules were visible. The following structures were visualized for the first time at this time point: a fluid-distended bladder (Fig. 1d″), including the fluid-filled urachus and urethra, mildly dilated ureters, and reproductive structures. The nervous system showed progressive differentiation with novel visualization of the optic nerves. The tympanic bulla was identified (Fig. 1d′). The musculoskeletal system showed progressive development and differentiation of the osseous structures of the distal limbs (Fig. 1d′, d″) and appropriate rib, sternal and vertebral development was noted (Fig. 1d). The left and right diaphragmatic crura were visualized (Supplemental Fig. 1b). The tongue and dental structures demonstrated progressive development (Supplemental Fig. 1b′). Myotendinous junctions could not be resolved.

Next, to determine muscle development, we undertook both H&E and Masson’s Trichrome stains of muscle sections at E41, E62, and E90. The transverse sections of E41 muscles showed densely packed fibers, which had central nuclei and a tubular appearance (Fig. 1e and Supplementary Fig. 2a). The formation of connective tissue surrounding muscle fibers was also detected using Masson’s Trichrome stain (Fig. 1h and Supplementary Fig. 2b). By E62, a progressively increasing formation of fibers and fascicle density were observed (Fig. 1f and Supplementary Fig. 2c). In addition, the formation of secondary fibers, which have a smaller size and surround the primary fiber appeared to form and vascular development within connective tissue septa was observed at this stage (Fig. 1f, i and Supplementary Fig. 2c, d). At E90, a greater fiber density per area than E62 were determined, and individual fascicles were well organized and delineated by connective tissues (Fig. 1g, j and Supplementary Fig. 2e, f).

The myofiber development in WT porcine embryos was dynamic

To capture the dynamic changes in myofibers of WT porcine embryos throughout various stages of embryonic and fetal development, we employed confocal microscopy and immunohistochemical techniques. WT embryos exhibited a normal and expected progression of skeletal muscle development (i.e., myogenesis). (Fig. 2 and Supplementary Fig. 3). The muscle development was based on distinct myosin heavy chain (MyHC) expression patterns and transitions, surrounded by a developing myofiber membrane that appeared consistent with normal myofiber maturation.

Fig. 2: Fiber typing in cloned wildtype (WT) porcine embryos.

a–c Immunohistochemical staining of cross sections of hindlimb from WT embryos. Fast-twitch fibers were detected using myosin heavy chain fast (yellow) and slow-twitch fibers were stained with myosin heavy chain slow (red). Nuclei were labeled with DAPI (blue). Across gestational stages, E41 (a), E62 (b), and E90 (c), representative sections with fast-twitch fibers, slow-twitch fibers, and merged images are displayed vertically. Scale bar, 100 µm. d Quantification of distribution of slow and fast-twitch fibers in E90 embryos are depicted (n = 4). Bar graphs represent mean ± s.e.m. with statistical validation using unpaired student’s t-test (***p < 0.001). e–g Immunohistochemical staining of hindlimb section with nuclei (blue), laminin (green), and embryonic myosin heavy chain (eMyHC, red). Across gestational stages E41 (e), E62 (f), and E90 (g), representative sections with laminin, embryonic MyHC and merged image are displayed vertically. Scale bar, 100 µm. h, i Quantitative analysis of muscle fiber cross-sectional area at each developmental stage (n = 3–5 each stage). The summarized cross-sectional area of muscle fibers (µm²) at each developmental stage is depicted in (h), and the distribution of muscle fiber cross-sectional areas, ranging from 10 to 400 µm², segmented into intervals of 50 µm² is presented in (i). Bar graphs represent mean ± s.e.m. with statistical validation using one-way ANOVA and Tukey’s test for average cross-sectional area within muscles (h).

Specifically, at E41 and E62 in WT embryos, the majority of myofibers expressed MyHC fast-twitch isoforms (Fig. 2a, b and Supplementary Fig. 3a, c). The increased appearance of smaller, fast-twitch myofibers at E62 appeared to represent the “splitting” of secondary myofibers from primary myofibers. By E90, this progression advanced to the stage where the formation of fascicles was visible and myofibers distinctly expressed either MyHC fast (~ 92%)- or MyHC slow (~ 8%)-twitch expressing fibers (Fig. 2c, d and Supplementary Fig. 3e). Specifically, a single, larger (average 256 µm²), MyHC slow-twitch expressing myofiber could be observed in the center of each fascicle. Each myofiber with centrally located nuclei and surrounded by smaller (average 148 µm²), MyHC fast-twitch expressing myofibers that make up the remainder of each fascicle.

Next, for further confirmation of myofiber development and maturation, we probed for laminin, a myofiber membrane marker, and embryonic MyHC (eMyHC), an isoform typically observed only during development (Fig. 2e–g and Supplementary Fig. 3b, d, and f). As expected, staining patterns in the WT embryos revealed an immature, but developing myofiber membrane (laminin), supported by robust eMyHC expression. At E62 days, both laminin and eMyHC staining were strongly expressed in and around all myofibers. By E90, while eMyHC continued to be expressed in most of the myofibers, the large, single MyHC slow-twitch myofibers in the middle of the fascicles lacked eMyHC expression. Collectively, these studies established that the cloned WT embryos displayed a well-organized but complex pattern of myofiber development and maturation.

The cross-sectional area (CSA) and distribution of a total of ~8500 myofibers with at least 2200 myofibers from each group were evaluated. The data revealed that the average overall myofiber CSA did not change (p = 0.117) between E41 (117.3 ± 16.7 µm²) and E62 (102.3 ± 37 µm²). While the primary MyHC slow-twitch myofibers were larger than the fast-twitch fibers there was not significant hypertrophy in the E90 embryos (154.3 ± 39.9 µm²; p = 0.117; Fig. 2h). These findings were further supported by a myofiber distribution analysis that categorized each myofiber into bins of 50 µm², ranging from the smallest (10–50 µm²) to the largest (> 400 µm²) myofiber diameter. The CSA distributions were unique across E41, E62, and E90 embryos. With the E42 fiber sizes were more normally distributed, the E62 demonstrated an increased number of smaller fibers, possibly supporting the splitting of secondary fibers from primary fibers (Fig. 2i and Supplementary Fig. 3g, h). The E90 fibers were noted to have a rightward shift toward larger fibers (i.e., the MyHC slow-twitch fibers) (Fig. 2i and Supplementary Fig. 3i). Overall, these data provided evidence of progressive and organized patterns of myogenesis in cloned WT embryos and supported the notion that myofiber expansion was the primary contributor to whole muscle growth during the early phase of myogenesis.

MYF5/MYOD/MYF6 deletion resulted in the lack of skeletal muscle in the porcine embryo

We produced skeletal muscle null porcine embryos by SCNT using CRISPR/Cas9 gene edited MYF5/MYOD/MYF6-null porcine fibroblasts as previously described¹⁴. As early as E41, MYF5/MYOD/MYF6-null embryos displayed an identifiable phenotype as we observed morphological abnormalities such as edema in the posterior wall and reduced limb development compared to age-matched WT embryos (Fig. 3a). Consistently, this gross phenotype was also observed in the E62 null embryo, where we noted expanded regions of edema associated with the ventral surface of the embryo and underdeveloped limbs. However, crown-to-rump measurements were not significantly different between the WT and null groups at either the E41 or E62 stage (Supplementary Tables 2 and 3). We evaluated embryos at E90 and observed that they were necrotic and resorbing (data not shown). Moreover, there was no significant increase in crown-to-rump length of the null embryos at E90 when compared to E62.

Fig. 3: Morphological and histological abnormalities in MYF5/MYOD/MYF6-null porcine embryos.

a Gross image of MYF5/MYOD/MYF6-null embryos at E41 and E62, highlighting morphological features such as edema and limb underdevelopment. Scale bar, 5 cm. b, c Parasagittal MRI scans at E41 (b, b′, b″) and E62 (c, c′, c″) displaying detailed internal structures with morphological abnormalities, rigid vertebral column development, no visible rib formation, and limb reduction. Scale bar, 1 cm (b, b′, b″) and 2 cm (c, c′, c″). d, e Representative H&E staining of hindlimb sections from MYF5/MYOD/MYF6-null embryos at E41 (d) and E62 (e), illustrating the lack of muscle fibers, replaced by edematous fluids. High magnification views of the boxed area are presented in (d′, e′). Scale bar, 1 mm (d, e), 250 µm (d′, e′). f, g Representative Masson’s Trichrome staining of sections at E41 (f) and E62 (g), showing connective tissue presence amidst the absence of muscle fibers. Boxed areas are enlarged in (f′, g′). Scale bar, 1 mm (f, g), 250 µm (f′, g′).

The null embryos had comparable degrees of organogenesis present with appropriate organ locations. Evidence of normal swallowing and renal function was appreciated by noting fluid-distended stomachs and bladders (Fig. 3c, c″) and whilst clear distinction between the thoracic and abdominal cavities could be made, the embryos lacked normal definition of the peritoneal and pleural margins. The thoracic and abdominal cavities were larger in volume compared to WT embryos with the visceral organs surrounded by edematous tissue (Fig. 3b′, c′). There was marked subcutaneous edema present (Fig. 3b). The neurological system showed comparable development to the stage-matched WT embryos. The null embryos, however, lacked normal development of the axial and appendicular musculature, with stunted limbs and absence of ribs and sternal structures (Fig. 3c, c′, c″). Cardiac and tongue muscles were normal when compared to the WT embryos (Fig. 3c′, c″).

Using H & E and Masson’s Trichrome staining techniques, we characterized the histological phenotype of skeletal muscle in the null embryos. At both E41 and E62, we determined a complete absence of muscle fiber development in the cross-section of hindlimbs and found that the regions of the muscle were replaced with edematous fluids (Fig. 3d, e and Supplementary Fig. 4a, c). Consistent with WT tissue, the appearance of connective tissue and vasculature were observed using Masson’s Trichrome staining, the degree of development, however, appeared to be delayed compared to the age-matched WT control (Fig. 3f, g and Supplementary Fig. 4b, d). Taken together, histological examination revealed that these embryos lacked all skeletal muscle of the body wall and the limbs.

Differential expression of myogenic stem cell and myogenic markers in WT and MYF5/MYOD/MYF6-null porcine embryos

The myogenic factors are crucial for the initial stages of muscle fiber formation and subsequent growth, regulating both the proliferation and differentiation of myogenic precursor cells⁴. In this study, we examined the distribution of myogenic stem cells (i.e., satellite cells) to investigate how the absence of MYF5, MYOD, and MYF6 affects myogenic lineage commitment and differentiation.

First, we examined the distribution of myogenic stem cells (i.e., satellite cells) in WT and MYF5/MYOD/MYF6-null porcine embryos. Immunohistochemical analysis was performed on cross-sections of hindlimbs stained with DAPI to mark nuclei, Laminin to outline the basement membrane, and Pax7 to identify myogenic stem cells (satellite cells). In WT embryos, at embryonic days E41, E62, and E90, Pax7⁺ cells were clearly visible (Fig. 4a–c and Supplementary Fig. 6a, c, e), indicating the presence of myogenic stem cells (satellite cells) in the muscle tissue. Conversely, sections from MYF5/MYOD/MYF6-null embryos at E41 and E62 exhibited an absence of Pax7 positive staining and a reduction in Laminin expression, consistent with the absence of the satellite cell population (Fig. 4d, e and Supplementary Fig. 6g, i). Quantitative analysis of Pax7⁺ cells confirmed an absence of satellite cells or myogenic stem cell population in MYF5/MYOD/MYF6-null samples compared to WT controls (p < 0.001; Fig. 4f).

Fig. 4: Satellite cell distribution in WT and MYF5/MYOD/MYF6-null porcine embryos.

Immunohistochemical staining of cross sections of hindlimbs labeled with DAPI (nuclei, blue), Laminin (basement membrane, red), and Pax7 (satellite cell, yellow). a–c Representative sections of WT hindlimbs across gestational stages E41 (a), E62 (b), and E90 (c) are displayed. d, e Representative hindlimb sections from MYF5/MYOD/MYF6-null embryos at E41 (d) and E62 (e), illustrating the lack of Pax7⁺ cells and diminished Laminin expression. Scale bar, 100 µm. f Quantification of Pax7⁺ cells per nuclei across the samples is depicted (n = 4). Bar graphs represent mean ± s.d. and statistical differences were analyzed by one-way ANOVA with Tukey’s test for multiple comparisons. (***p < 0.001). Immunohistochemical staining of cross sections of a hindlimb labeled with DAPI (nuclei, blue), Dystrophin (muscle membrane, red), and Myogenin (muscle differentiation marker, yellow). g–i Representative section of WT hindlimb skeletal muscle across developmental stages E41 (g), E62 (h), and E90 (i) are displayed. j, k Representative sections from MYF5/MYOD/MYF6-null hindlimbs at E41 (j) and E62 (k), showing the absence of Myogenin⁺ cells and Dystrophin. Scale bar, 100 µm. l Myogenin⁺ cells quantification across the samples is presented (n = 4). Bar graphs represent mean ± s.d. with statistical validation using one-way ANOVA and Tukey’s test (**p < 0.01, ***p < 0.001).

Similarly, cross-sections were also labeled for Dystrophin, a muscle membrane marker, and Myogenin, a muscle differentiation marker. WT embryos showed robust Myogenin positive staining across all developmental stages examined (Fig. 4g–i and Supplementary Fig. 6b, d, f), whereas MYF5/MYOD/MYF6-null samples displayed a complete lack of Myogenin and Dystrophin positive cells at both E41 and E62 (Fig. 4j, k and Supplementary Fig. 6h, j). Quantitative analysis confirmed an absence of Myogenin⁺ cells in MYF5/MYOD/MYF6-null samples compared to WT controls (p < 0.001; Fig. 4l).

Quantitative assessment of Pax7 and Myogenin positive cells indicated a developmental trajectory in WT embryos (Fig. 4f and Fig. 4l), with cell counts increasing from embryonic day E41 to E62 (p < 0.001), followed by a decline by E90 (p < 0.001 for Pax7; p = 0.003 for Myogenin). This trend suggested a peak in satellite cell activation and muscle differentiation markers during mid-gestation, with a subsequent decrease as the embryos approached full term.

Comparative transcriptomic analysis of the MYF5/MYOD/MYF6-null and WT embryo

To gain a comprehensive insight into the gene expression profile and the transcriptional network involved in skeletal muscle development, we isolated RNA at embryonic day E41 and E62 from hindlimb samples of both WT and MYF5/MYOD/MYF6-null embryos and performed bulk RNA sequencing. Comparing the gene expression profile at E41, we observed 1085 upregulated and 1100 downregulated genes in the MYF5/MYOD/MYF6-null hindlimb compared to the corresponding WT hindlimb (Supplementary Data 1). We observed that along with MYOD, MYF5, and MYF6, other key regulatory genes of skeletal muscle development including MYOG, PAX7, muscle fiber-related structural genes including MYH3, MYH13, DES, ACTG1, TNNI1, TNNI2 were undetectable in the MYF5/MYOD/MYF6-null hindlimb compared to the WT hindlimb (Fig. 5a, d). Myogenesis inhibitory genes such as TWIST1, TWIST2 as well as the embryonic limb development gene GLI2 were upregulated in the MYF5/MYOD/MYF6-null hindlimb samples compared to the WT control (Fig. 5a, d). Gene ontology (GO) pathway analysis also identified muscle organ and structure development pathways to be the most downregulated pathways in the MYF5/MYOD/MYF6-null limbs whereas the blood and vasculature developmental pathways were upregulated in the absence of skeletal muscle development program (Fig. 5b, c).

Fig. 5: Differential expression analysis between E41 MYF5/MYOD/MYF6-null and E41 WT porcine hindlimbs.

a Volcano plot of differentially expressed genes (DEGs). Red dots indicate upregulated genes in MYF5/MYOD/MYF6-null hindlimbs and blue dots indicate downregulated genes in MYF5/MYOD/MYF6-null hindlimbs. The cut-off criteria for assessing the DEGs was |log2FC|>1 and adjusted p < 0.05. b, c Gene ontology (GO) pathway analysis highlights pathways, signaling, and development-related terms of upregulated genes in MYF5/MYOD/MYF6-null hindlimbs (b) and downregulated genes in MYF5/MYOD/MYF6-null hindlimb samples (c). The x-axis represents the counts of genes in each GO term and the color scale shows the increased significance of biological processes using the over-representation test with an adjusted p < 0.05. d The heatmap represents upregulated and downregulated genes in the WT hindlimbs vs. the MYF5/MYOD/MYF6-null hindlimbs at E41. The heatmap color scheme key is provided, with red representing upregulated and blue representing downregulated genes.

A similar comparison of gene expression at E61 hindlimb samples identified a more severe phenotype with 1760 upregulated and 1599 downregulated genes in the MYF5/MYOD/MYF6-null samples compared to the corresponding WT samples (Supplementary Data 2). Similar to the E41 hindlimb samples, skeletal muscle structure and fiber development genes like MYH6, MYH7, MYH8, MYH14, TNNI2, and DES were undetectable (Fig. 6a, d). Myogenesis inhibitory genes like TWIST1 and TWIST2, immune system-related genes like IL6R, IL10 as well as NMJ development factors like AGRN were ectopically overexpressed in the MYF5/MYOD/MYF6-null limb samples. GO pathway analysis again identified muscle organ, structure, and striated muscle cell development pathways to be the top downregulated pathways in the MYF5/MYOD/MYF6-null hindlimb whereas the hematopoietic and immune system development pathways were upregulated in the absence of skeletal muscle development program at E62 (Fig. 6b, c).

Fig. 6: Differential expression analysis between E62 MYF5/MYOD/MYF6-null and E62 WT porcine hindlimbs.

a Volcano plot of differentially expressed genes (DEGs). Red dots indicate upregulated genes in MYF5/MYOD/MYF6-null hindlimbs and blue dots indicate downregulated genes in MYF5/MYOD/MYF6-null hindlimbs, the cut-off criteria for assessing the DEGs was |log2FC|>1 and adjusted p < 0.05. b, c Gene ontology (GO) pathway analysis highlights pathways, signaling, and development-related terms of upregulated genes in MYF5/MYOD/MYF6-null hindlimbs (b) and downregulated genes in MYF5/MYOD/MYF6-null hindlimbs (c). The x-axis represents the counts of genes in each GO term and the color scale shows the increased significance of biological processes using the over-representation test with an adjusted p < 0.05. d The heatmap represents upregulated genes and downregulated genes in the WT hindlimbs vs. the MYF5/MYOD/MYF6-null hindlimbs at E62. The heatmap color scheme key is provided, with red representing upregulated and blue representing downregulated genes.

Perturbed NMJ development of the neuromuscular junction in the MYF5/MYOD/MYF6-null porcine embryos

Neuromuscular junction development was broadly evaluated using both post-synaptic (acetylcholine receptors, AChR; α-BTX) and pre-synaptic (motor neuron/synaptic vesicles; 2H3/SV2) markers, in addition to the Schwann cell marker S100 (Fig. 7). As the interaction and signaling point between the nervous system and skeletal muscle, an understanding of NMJ development in both the cloned WT and MYF5/MYOD/MYF6-null embryo is necessary to completely inform our knowledge of myogenic stages. Consistent with findings of myofiber development, NMJ formation was present in E41 WT embryos (Fig. 7a), though sparse and unorganized. Numerous AChR clusters were present that did not yet have an associated pre-synaptic terminal; however, there were motor nerves throughout the tissue with clear Schwann cell development. In contrast, the E41 MYF5/MYOD/MYF6-null hindlimb samples displayed little to no AChR clusters (Fig. 7b), as expected based on the histology. There was gross nerve development present throughout the E41 MYF5/MYOD/MYF6-null hindlimb tissue, but interestingly, Schwann cell development appeared to be absent compared to their WT counterparts. Mature Schwann cells wrap around the axon and produce myelin; thus, it appeared the Schwann cells in E41 WT hindlimb samples were still immature due to their spot-like presentation along the nerve, termed radial sorting¹⁶.

Fig. 7: Differential innervation of neuromuscular junctions (NMJs) at the developmental time points.

The postsynaptic staining (red) indicates acetylcholine receptors using fluorescently tagged α-bungarotoxin. The presynaptic staining (green) displays the motor nerve and synaptic terminal using an antibody cocktail of neurofilament and synaptic vesicle glycoprotein 2. The Schwann cells (purple) are targeted by an anti-S100 antibody. Nuclei are stained with DAPI (blue). Across the row, individual maximal projection images for each fluorophore are presented, as well as the merged image. a–e Each row represents a different developmental stage of WT and MYF5/MYOD/MYF6-null hindlimbs. E41 WT hindlimbs (a), E41 MYF5/MYOD/MYF6-null hindlimbs (b), E62 WT hindlimbs (c), E62 MYF5/MYOD/MYF6-null hindlimbs (d), and E90 WT hindlimbs. Scale bar, 100 µm.

Overlap between pre- and post-synaptic structures (i.e., the nervous and skeletal muscle components respectively), indicating NMJ formation, was more consistently observed in E62 WT hindlimb samples (Fig. 7c) compared to E41 WT hindlimb samples, though clusters of unoccupied AChRs still remained. Terminal Schwann cells, a type of non-myelinating Schwann cell supportive of NMJ development and maintenance, were identifiable at a number of NMJs in the E62 WT hindlimb tissue. Conversely, similar to E41 MYF5/MYOD/MYF6-null hindlimb tissue, the E62 MYF5/MYOD/MYF6-null hindlimb tissue had little to no AChR staining present (Fig. 7d), and no areas of overlap that would indicate NMJ formation. The Schwann cell marker, S100, that was minimally expressed in the E41 MYF5/MYOD/MYF6-null, was clearly present in the E62 null, suggesting MYF5/MYOD/MYF6 gene deletions may delay Schwann cell development. Within the E90 WT skeletal muscle (Fig. 7e), nerve patterning was more organized and AChR clusters were starting to assume a more mature structural shape. Most AChR clusters were innervated by a pre-synaptic site and there also appeared to be terminal Schwann cells located adjacent to most NMJs.

Discussion

In these studies, we comprehensively examined the development and skeletal muscle myogenesis of WT pig at mid-gestational stages (E41, E62, and E90). In addition, we investigated the phenotype of porcine embryos lacking three myogenic bHLH factors, MYF5, MYOD, and MYF6. Importantly, we evaluated development in porcine embryos and fetal pigs which were generated via SCNT. Given the increasing use of pigs generated in this manner as well as the cloning defects that can occur, the definition of developmental progression in cloned pigs provides essential information to the field. We performed these studies using MRI which offers novel insights of organogenesis over time in both WT and skeletal muscle null embryos and fetal pigs. We focused, particularly, on skeletal muscle development using multiple techniques including histology, morphology, immunohistochemistry, and bulk RNA-sequencing to define the molecular cascades that are relevant to myogenesis and myofiber formation in swine. Lastly, we describe an embryonic lethal phenotype in pig which is more severe than that defined in mice.

Establishing disease models in pigs is important and valuable given their anatomical and physiological similarities to humans. In the current study, we examined the normal developmental profile/stages of the SCNT-produced porcine embryo with a particular emphasis on skeletal muscle. The gestational period and physiological similarities to humans pig is 113–115 days¹⁷. Our gross observations indicated a rapid growth period between E41 and E90. MRI results of WT porcine embryos at defined embryonic ages (E41, E62, and E90) supported and verified the previously established developmental process of various organs. Notably, distinct morphologies of the kidney, liver, spleen, stomach, small intestine¹², as well as the heart¹¹ were observed as early as E41. Furthermore, there was a noticeable progression in both size and structure complexity by E90 (Fig. 1b–d). Our MRI analysis demonstrated that brain development was rapid at approximately day 60 of gestation as previously reported in naturally bred swine^18,19. At E41, brain parenchyma appeared hypointense, lacking clear differentiation into the cerebellum, cerebrum, brainstem, and pons. However, by E62, a pivotal developmental milestone was reached, marked by the delineation of these brain regions, despite the persistence of homogenous hypointensity and the absence of white-matter/gray-matter distinction. These results support the onset of rapid brain growth and structural organization. These observations support the conclusion that myogenesis in the cloned porcine embryos is similar to that observed in naturally bred animals²⁰ and also provide additional insights into organ development in porcine embryos produced by SCNT.

Our studies revealed the sequence and pattern of muscle development in E41, E62, and E90 porcine embryos produced by SCNT (Fig. 1e–j and Supplementary Fig. 2). Immunohistochemically, we observed a typical progression of myofiber development characterized by distinct MyHC expression patterns and the maturation of myofibers, which underwent a process known as “splitting” to form secondary myofibers. Our study demonstrated that myofibers initially expressed fast MyHC at earlier embryonic and developmental stages, with a transition to the expression of slow MyHC (i.e., primary fibers) as the myofibers mature (Fig. 2a–d). Myofiber cross-sectional area analysis further confirmed the complex yet organized progression of myofiber development and maturation, highlighting significant hypertrophy by gestational day 90, especially in the primary MyHC slow-expressing fibers (Fig. 2e–h). Overall, these findings suggest that myofiber expansion plays a key role in early muscle growth, followed by a shift towards hypertrophy in later stages of development, thus providing valuable insights into the mechanisms of myogenesis in the cloned, developing pig.

In the current study, we produced porcine embryos carrying null mutants in MYF5, MYOD, and MYF6, termed as MYF5/MYOD/MYF6-null embryos using SCNT (Fig. 3). MYF5/MYOD/MYF6-null porcine embryos had grossly abnormal phenotypes as early as the E41 stage (Fig. 3a), which were comparable to the phenotype of Myog- or Myf5/Myod double-mutant mice^8,21,22. Myog-disrupted mice displayed abnormal curvature of the spine, thickening in the dorsal neck area, and reduced musculature in the limbs²¹. Accretion of fluid under the skin was detected in Myog-null mice²². Myf5/Myod double-mutant mice exhibited arched spines and thin limbs⁸. Consistent with previous findings in mice, MYF5/MYOD/MYF6-null porcine embryos displayed a curved spine, severe peripheral edema, and reduced limb development compared to age-matched WT embryos (Fig. 3a). We further confirmed that the MYF5/MYOD/MYF6-null porcine embryos exhibited pronounced musculoskeletal anomalies by MRI. We observed a significant curvature of the spine and severe edema in various regions, indicative of profound developmental disruptions (Fig. 3b, c). Importantly, we concluded that MYF5/MYOD/MYF6-null embryos stopped growth progression after E62 by crown to rump length measurement, indicating embryonic lethality (Supplementary Table. 3). These measurements, however, might have been affected by the anatomical distortion that the embryos exhibited. This lethal phenotype in the porcine embryo was more severe compared to the perinatal lethality observed in mice lacking Myf5 alone, and both Myf5 and Myod^5,8.

Extending these findings, our MRI analysis provided additional information on musculoskeletal developmental abnormalities of the MYF5/MYOD/MYF6-null porcine embryos (see Fig. 3b, c). Despite appropriate organogenesis of the thoracic and abdominal organs; abnormal development in the axial/appendicular systems and spinal scoliosis were observed. Furthermore, the absence of a sternum and ribs provided a plausible explanation for the observed embryonic lethality post-E62^5,8. These results support the hypothesis that respiratory failure due to severe rib defects or the absence of diaphragmatic muscle could be a leading cause of lethality in fetuses lacking skeletal muscle. It is believed that rib defects are a trait caused by disruption of the Myf5 and Myf6 genes in mice^4,5,8. Myf6 as well as Myf5 have been discovered to be important in rib formation in mice^23,24. Rib defects observed in MRF-disrupted mice are interesting and indicative of the interaction between the myotome and sclerotome which plays a significant role in the growth and patterning of ribs²³. Our findings not only reinforced the critical role of MRFs in musculoskeletal development but also provided new insights into the severity of skeletal abnormalities and their potential implications in development and viability in the pig.

Previous studies have shown that the deletion of MRFs such as MYF5 and MYOD drastically altered the developmental trajectory of satellite cells during embryogenesis. For instance, in the absence of Myod, satellite cells fail to properly commit to the myogenic lineage, resulting in impaired activation and proliferation²⁵. Similarly, in Myf5-null mice, muscle progenitor cells undergo normal epithelial-mesenchymal transition but exhibit defective migration, disrupting proper muscle development²⁶. Furthermore, in Myod/Myf5 double knockout mice, the myogenic programming of satellite cells is significantly disrupted, leading to profound defects in muscle formation²⁷. These findings highlight the critical role of MRFs not only in maintaining the population size of satellite cells but also in ensuring their normal progression through the myogenic program. In the current study, the absence of these key transcription factors, MYF5/MYOD/MYF6, led to an absence in Pax7⁺ satellite cells, the muscle differentiation marker, MYOG, and structural proteins when compared to WT (Fig. 4). Unlike findings in Myod single or Myf5/Myod double knockout mice, where satellite cells remain in an inactive state, our results demonstrate the complete depletion of satellite cells in MYF5/MYOD/MYF6-null porcine embryos. This finding suggests that the combined loss of all three MRFs results in more severe disruptions to the myogenic lineage. Further studies are needed to elucidate the mechanisms driving this complete loss of satellite cells and to determine whether it is due to distinct mechanisms specific to the triple knockout or represents a phenomenon unique to pigs, differing from what has been observed in mice.

The comparative transcriptomic analysis between MYF5/MYOD/MYF6-null and WT embryos at E41 and E62 reveals significant insights into the regulatory mechanisms underlying skeletal muscle development (Figs. 5 and 6). At E41, the absence of MYF5/MYOD/MYF6 resulted in the dysregulation of a significant number of genes, including the downregulation of critical myogenesis-related genes such as MYOG and PAX7, as well as structural genes. Conversely, myogenesis inhibitory genes TWIST1 and TWIST2 were upregulated. By E62, the gene dysregulation reflected a more pronounced disruption in skeletal muscle development. Notably, vasculature related genes, immune system-related genes and NMJ development factors were ectopically overexpressed. We believe that these results indicate a compensatory response to the absence of muscle fiber development and were consistent with the edematous and hemorrhagic regions that we observed in the null embryos. These findings are similar to a previous study in mice that also demonstrated severe skeletal muscle defects and compensatory overexpression of non-muscle genes in the absence of key myogenic factors²⁸. While global gene expression studies focused on adult porcine skeletal muscle exist, such as the identification of RNA editing sites across various tissues²⁹, comprehensive analyses of pig muscle tissues through DNA methylation/gene expression mapping³⁰, and transcriptomic analyses studies focusing specifically on porcine myogenesis, especially in the context of gene knockouts like MYF5/MYOD/MYF6, are still limited. Furthermore, our studies uniquely evaluate the developmental progression of cloned porcine WT and MYF5/MYOD/MYF6-null embryos produced via SCNT, specifically at E41 and E62. This emphasizes the novelty of our study in providing crucial insights into the molecular mechanisms of muscle development in cloned porcine embryos.

Developmental formation of the NMJ is a well-orchestrated process involving the clustering of AChRs and the alignment of motor neurons, and is critical for muscle function^31,32. In WT embryos, we observed a progressive maturation of the NMJ which appeared complete by day 90. This process was significantly disrupted in MYF5/MYOD/MYF6-null embryos Fig. 7b. At gestational day 41, these null embryos exhibit an absence of AChR clustering and Schwann cell development, suggesting severe developmental arrest. Consistent with our results, various studies have demonstrated that disruptions in MRFs influence NMJ development. For instance, Myod deletion in mice leads to significant alterations in NMJ organization and muscle contractile properties, highlighting the unique roles of Myod in both developing and adult skeletal muscle²⁵. Mice lacking both Myf5 and Myod leads to complete ablation of motor neurons³³. Our bulk RNA-seq results indicate that in null embryos, there is a notable absence of LRP4 and MUSK (Figs. 5 and 6) which are known to regulate NMJ development^34,35. In contrast, the expression of another NMJ regulatory gene, ARGN³⁶, was maintained in the null. We hypothesize that MRFs do not directly participate in NMJ development but that the absence of MRFs leads to a lack of muscle, which in turn causes a deficiency of post-synaptic elements necessary for NMJ formation and maintenance. Interestingly, Schwann cells were absent in E41 MYF5/MYOD/MYF6-null embryos, but were present by E62 (see Fig. 7d).

WT cloned porcine embryos showed normal skeletal muscle progression, while MYF5/MYOD/MYF6-null porcine embryos exhibited severe musculoskeletal deficiencies, including the absence of satellite cells, disrupted NMJ formation, and embryonic lethality post-E62. In conclusion, our findings provide critical insights into the essential roles of MRFs in skeletal muscle development and viability. Future research should focus on elucidating the mechanisms driving the observed satellite cell depletion and NMJ disruptions in MYF5/MYOD/MYF6-null pigs. Despite these challenges, our study demonstrates the significant utility of the MYF5/MYOD/MYF6-null pig model. The development and comprehensive analysis of WT and skeletal muscle null porcine embryos and fetal pigs advances our understanding of porcine myogenesis. In addition, these studies reveal the process of myogenesis in cloned embryos and highlight the potential of these genetically modified pig models in advancing regenerative medicine and exotransplantation science. Specifically, the MYF5/MYOD/MYF6-null pig model provides a unique platform for exogeneic donor muscle generation, creating a vacant niche devoid of skeletal muscle where interspecific cells can populate and proliferate without competition. This system holds promise for producing human skeletal muscle, offering a potential therapeutic solution for VML. Furthermore, it supports the isolation of human myogenic stem cells, providing a critical resource for cell transplantation therapies and expanding the scope of regenerative medicine.

Methods

Animal assurance

Experimental studies were approved by the Institutional Review Board, Institutional Animal Care and Use Committee, and Stem Cell Research Oversight panels at the University of Minnesota and Collagen Solutions, Inc. We have complied with all relevant ethical regulations for animal use.

Fetal porcine fibroblast collection, culture and generation of the MYF5/MYOD/MYF6 gene-edited fibroblasts

The procedures for fetal porcine embryo harvest, fibroblast isolation and culture, and the CRISPR/Cas9-mediated generation of MYF5/MYOD/MYF6-null fibroblasts performed as previously described¹⁴.

Generation of wildtype (WT) and MYF5/MYOD/MYF6-null embryos

Wildtype and MYF5/MYOD/MYF6-null embryos were generated by SCNT as previously described¹⁴. Briefly, pig primary oocyte complexes aspirated from ovarian follicles were cultured in vitro for 40 to 42 h in medium-199 (Corning, 10-060-CV) supplemented with a series of growth factors and hormones at 38.5 °C under 5% CO₂. Mature oocytes were enucleated in a manipulation medium by aspirating the polar body and metaphase II (MII) spindles, visualized under an Oosight Spindle Viewer (Hamilton Thorne). MYF5/MYOD/MYF6-null or WT fibroblasts, prepared in advance, were introduced into the perivitelline space of each enucleated oocyte, and fusion was induced by electrofusion. Activated embryos were cultured in vitro using PZM-MU2 media, supplemented initially with the histone deacetylase inhibitor, Scriptaid, to enhance epigenetic reprogramming. Transfer of generated wildtype and MYF5/MYOD/MYF6-null embryos into oestrus synchronized gilts was performed as described previously³⁷. Gilts were Sus Domesticus, female of approximately 6 months of age at time of embryo transfer. Embryos were harvested at the gestational ages noted.

Harvest of porcine embryos and tissue collection

Fetal porcine embryos were harvested at E41, E62, and E90 days of gestation to capture important developmental milestones in both the WT and MYF5/MYOD/MYF6-null embryos. The procedure involved anesthetizing the surrogate gilts, surgically removing the uterus, and individually extracting, cleaning, measuring, and imaging the embryos. Muscle samples from the hindlimbs were collected for histological examination and gene expression studies. Samples for IHC were either fixed in 4% paraformaldehyde or fresh frozen on tragacanth gum and isopentane cooled in liquid nitrogen, and samples for gene expression were preserved in RNAlater solution ThermoFisher #AM7021) until RNA isolation.

Magnetic resonance imaging

Selected porcine embryo specimens were imaged using a 3 T MRI scanner (MAGNETOM Prisma; Siemens Healthcare; Erlangen, Germany) and a vendor-provided 16-channel hand/wrist coil. The following imaging sequences and typical parameters were sagittally acquired to provide a range of image contrasts: (1) 3D FISP (resolution = 0.31 × 0.31 × 0.31 mm³; TR/TE = 12.0/5.0 ms; flip angle = 40°; BW = 230 Hz/px; scan time = 6 min); (2) 3D VIBE (resolution = 0.31 × 0.31 × 0.40 mm³; TR/TE = 12.2/6.2 ms; flip angle = 10°; BW = 190 Hz/px; water excitation; scan time = 6 min); (3) T2-weighted 3D SPACE (resolution = 0.71 × 0.71 × 0.71 mm³; TR/TE = 3200/410 ms; turbo factor = 282; BW = 723 Hz/px; fat sat; scan time = 2.5 min); (4) PD-weighted 3D SPACE (resolution = 0.39 × 0.39 × 0.39 mm³; TR/TE = 1200/25 ms; turbo factor = 34; BW = 383 Hz/px; fat sat; scan time = 8.5 min); and (5) 2D TIRM (resolution = 0.26 × 0.26 mm²; slices = 20; slice thickness/gap = 1.0/0.1 mm; TR/TE = 4000/29 ms; TI = 220 ms; flip angle = 150°; turbo factor = 9; BW = 200 Hz/px; scan time = 6 min). Representative MRI slices were viewed and exported using RadiAnt DICOM viewer. The MRI images were evaluated by a board-certified veterinary radiologist to summarize morphological features and abnormalities for each specimen.

Histology of skeletal muscle sections

Serial 8 µm cross sections of muscle and hindlimb samples were cut using a Leica cryostat set to −20 °C. Sections were mounted on Superfrost slides and air dried for at least 10 min before proceeding with histological or immunohistochemical staining. Broad qualitative analysis of muscle tissue was obtained using Masson’s trichome (Thermo Scientific Richard-Allan Scientific Chromaview Kit; 87019) and Hematoxylin and Eosin staining following manufacturer instructions. Brightfield imaging was performed at a magnification of 20× (0.75 NA) using the Huron TissueScope LE, a high-resolution digital slide scanning microscope (Huron Digital Pathology, Ontario Canada). For fluorescent images, samples were fixed in 4% paraformaldehyde, washed, and incubated in 10% normal goat serum prior to primary antibody incubation overnight at 4 °C. Samples were then washed and incubated with the appropriate corresponding secondary antibodies and counterstained (e.g., DAPI) for at least 1 h at room temperature before being mounted with coverslips and imaged. In total 5–10 randomly selected areas of the tissue were imaged, providing an overall representation of the tissue. Imaging was conducted using a Nikon C2 automated upright laser scanning confocal microscope equipped with an NIR Apo 40 × 0.80 W DIC oil immersion objective and dual GaSP detectors (Nikon Instruments Inc., Melville, NY). Image sampling was determined using nyquist calculations with a pixel size set to 0.16 μm and 1024 × 1024 dimensions. The step size was set to match the empirically-determined Z-axis resolution (1 μm). In all cases, tissue samples were stained in batches to minimize discrepancies between samples.

Immunohistochemical reagents: primary antibodies

Pax7 (Pax7; Developmental Studies Hybridoma Bank; 5 µg/ml with Alexa Fluor 647 Goat anti-mouse IgG1), eMyHC (F1.652; Developmental Studies Hybridoma Bank; 5 µg/ml with Alexa Fluor 546 Goat anti-mouse IgG1), MyHC all fast (F-59; Developmental Studies Hybridoma Bank; 5 µg/ml with Alexa Fluor 647 Goat anti-mouse IgG1), MyHC slow (BA-F8; Developmental Studies Hybridoma Bank; 5 µg/ml with Alexa Fluor 546 Goat anti-mouse IgG2B), Myogenin (PCRP-MYOG-1C5; Developmental Studies Hybridoma Bank; Secondary antibodies: 5 µg/ml with Alexa Fluor 647 Goat anti-mouse IgG2B), Laminin (L9393; Sigma Aldrich; 1 µg/ml with Alexa Fluor 546 goat anti-rabbit IgG), Dystrophin (15277; abcam; 1:200 with Alexa Fluor 546 goat anti-mouse IgG). Nuclei Hoechst 33342 Dye 20 mM Solution (62249; Pierce; 1:1000) or DAPI (D21490; Invitrogen; 1:1000). Mounting Media: VectaShield (H-1000, Vector Labs) or Fluoro-gel (50-247-04, Electron Microscopy Sciences).

Neuromuscular junction (NMJ) analysis

Muscles were sectioned at 100 µm and pinned into a Sylgard® coated dish. Sections were then placed in the following sequence of solutions: 4% paraformaldehyde for ~45 min; washed with PBS; α-Bungarotoxin, Alexa Fluor 555 conjugate (Invitrogen, B35451, 1.33 µg/ml in PBS) for ~45 min; washed with PBS; 1% Triton X-100-PBS for 90 min; 10% goat serum in 0.5% Triton X-100-PBS for 30 min. The sections were then placed in the following primary antibody solution overnight at 4 °C: ready-to-use rabbit anti-S100 (1:200; Agilent Dako Omnis, GA50461-2), mouse anti-neurofilament (DSHB, 2H3, 0.25 µg/ml) and mouse anti-synaptic vesicles (DSHB, SV2, 0.9 µg/ml), in 5% goat serum. Sections were washed in PBS 6 × 10 min before being placed in the following secondary antibody solution (all in PBS) for 2 h at room temperature: Alexa Fluor 488 goat anti-mouse IgG1 (5 µg/ml), Alexa Fluor 647 goat anti-rabbit IgG (5 µg/ml), and DAPI (1 µg/ml). Before imaging, sections were unpinned from the dish and transferred to a microscope slide (Fisherbrand Superfrost Plus, 12-550-15). A drop of mounting media (Invitrogen ProLong Diamond Antifade, P36970) was placed onto each section and a #1.5H coverslip (Thorlabs, CG15KH) was pressed and held firmly over the sections. Slides were kept at room temperature and protected from light for 2–3 days before imaging to achieve optimal refractive index. For evaluation of the NMJ, images were obtained using a Nikon C2 microscope equipped with a PlanFluor 40X oil objective (1.30 NA, DIC N2), galvano scanner, and dual GaSP detectors (Nikon Instruments Inc., Melville, NY).

Confocal imaging quantification

Quantitative confocal analysis was used for fast and slow-twitch fiber–type identification, cross-sectional area measurements, and counting of nuclei and nuclear-expressing transcription factors (e.g., Pax7 and myogenin). This was performed by converting images into 8-bit confocal stacks using Fiji (ImageJ) imaging software and by setting an image intensity threshold using the otsu algorithm to identify positively stained regions of the tissue. The analysis settings were adjusted to include a minimal size requirement of 5 µm² for nuclear markers to eliminate off-target staining of debris. For the myofiber cross-sectional area, data were manually obtained using Nikon Elements AR software. Maximal intensity projections of each image were used to measure the area of each myofiber, with a minimal size requirement of 10 µm²

RNA isolation and quality assessment

Total RNA was purified using RNeasy mini kit (Qiagen) and quantified using a fluorimetric RiboGreen assay. Total RNA integrity is assessed using capillary electrophoresis (e.g., Agilent BioAnalyzer 2100), generating an RNA Integrity Number (RIN). For samples to pass the initial QC step, they needed to have more than 500 ng and have a RIN of 8 or greater. Total RNA samples were then converted to Illumina sequencing libraries.

Bulk RNA-Seq library creation

Total RNA samples were converted to Illumina sequencing libraries using Illumina’s TruSeq RNA Sample Preparation Kit (Cat. # RS-122-2001 or RS-122-2002) or Stranded mRNA Sample Preparation Kit (Cat. # RS-122-2101). (Please see www.illumina.com for a detailed list of kit contents and methods). The mRNA from a normalized input mass of total RNA was isolated using oligo-dT coated magnetic beads, fragmented and then reverse transcribed into cDNA. The cDNA was blunt-ended, A-tailed and indexed by ligating molecularly barcoded adaptors. Libraries were amplified using 15 cycles of PCR. Final library size distribution was validated using capillary electrophoresis and quantified using fluorimetry (PicoGreen) and via Q-PCR. Indexed libraries were then normalized, pooled and size-selected to 320 bp (tight) using the PippinHT instrument.

Cluster generation, sequencing, and primary analysis

Pooled libraries were denatured and diluted to the appropriate clustering concentration. Denatured and diluted libraries were loaded onto the NextSeq 2000 cartridge and clustering occurs onboard the instrument. Once clustering was completed, sequencing immediately commenced using Illumina’s 2-color SBS chemistry. Upon completion of read 1, an index read 1 of varying length was performed depending on the library kit used. If libraries were dual indexed, a second index read was performed. Finally, the library fragments were re-synthesized in reverse orientation and sequenced from the opposite end of the read 1 fragment to produce the paired end read 2. Base call (.bcl) files for each cycle of sequencing were generated by Illumina Real Time Analysis software. The base call files and run folders were streamed to servers maintained at the Minnesota Supercomputing Institute. Primary analysis and de-multiplexing were performed using Illumina’s bcl2fastq v2.20. The end result of the bcl2fastq workflow was de-multiplexed FASTQ files.

Bulk RNA-sequencing analysis

The sequencing reads were aligned to the Sus scrofa (pig) genome assembly Sscrofa11.1 (susScr11), submitted by the Swine Genome Sequencing Consortium, using kallisto (v0.48.0) with the following parameters: −t 8 −b 100. Subsequently, we employed tximport (v1.26.0) to determine the read count for each transcript and quantified transcript abundance as transcripts per kilobase per million reads mapped (TPM). The transcriptome index was generated from the Ensemble release 108 FASTA file for Sus scrofa (Sscrofa11.1 genome build). Although a GTF file was not directly used in the pipeline, the transcript-to-gene mapping required for downstream analysis was obtained from Ensemble’s biomaRt database, aligned with the Ensemble release 108 transcript sequence used to build index. We summed the read counts and TPM of all alternative splicing transcripts of a gene to obtain gene expression levels. To identify statistically significant differences in gene abundance between the WT and MYF5/MYOD/MYF6-null embryos, as well as differences among WT embryos at E41 and E62, a Wald test (corrected for multiple testing using the Benjamini and Hochberg method) was applied using DESeq2 (v1.38.3). The test was conducted for a fold change difference greater than 2× (log2 fold change threshold of 1), rejecting the null hypothesis when the adjusted p-value was less than 0.05, followed by differential expression analysis. Gene Ontology (GO) enrichment analysis for biological process (BP) terms was then performed with clusterProfiler (v4.7.0), using a p-value-adjusted false discovery rate < 0.05.

Statistical analysis

At least 5 fields were randomly chosen and quantified. Statistical analyses were performed using unpaired student’s t-test or one-way ANOVA with Tukey’s test. Analyses were performed using JMP Pro (version 14.2 SAS Institute, Cary, 220 NC, USA) and figures were generated using Graphpad Prism (version 9.3.1). Data are presented as mean ± s.e.m. or s.d. Results were considered statistically significant at a p-value < 0.05.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.