Skeletal muscle regeneration is attenuated in global Fn14-KO mice

We first investigated how the levels of Fn14 are regulated in the regenerating skeletal muscle of adult mice. One side tibialis anterior (TA) muscle of 8-wk-old WT mice was injured by intramuscular injection of 1.2% BaCl₂ solution, whereas the contralateral TA muscle was used as a control. To evaluate the initial phases of skeletal muscle regeneration, the TA muscle was isolated 5 d post-injury and analyzed by performing quantitative real-time PCR (qRT–PCR) and Western blot. Consistent with published reports (Girgenrath et al, 2006; Mittal et al, 2010b), we found that mRNA and protein levels of Fn14 were significantly increased in the injured TA muscle compared with the contralateral uninjured muscle (Fig 1A and B). To understand the role of Fn14 in skeletal muscle regeneration, we first employed WT and whole-body Fn14-KO mice. The TA muscle of the mice was injured by intramuscular injection of 1.2% BaCl₂ solution, and muscle regeneration was evaluated 5 d later. There was no significant difference in wet weight of corresponding uninjured and injured TA muscles normalized by body weight between WT and Fn14-KO mice (Fig S1A). We next generated TA muscle sections and performed hematoxylin and eosin (H&E) staining. In agreement with a previously published report (Girgenrath et al, 2006), we found that muscle regeneration was considerably reduced in Fn14-KO mice compared with WT mice (Fig 1C). The average cross-sectional area (CSA) of centronucleated myofibers was significantly lower in the 5d-injured TA muscle of Fn14-KO mice compared with WT mice (Fig 1D and E). We also performed immunostaining for embryonic isoform of myosin heavy chain (eMyHC) protein that is expressed only in newly formed myofibers (Fig S1B). This analysis showed that the percentage of eMyHC⁺ myofibers containing two or more centrally located nuclei was significantly reduced in the 5d-injured TA muscle of Fn14-KO mice compared with WT mice (Fig 1F). However, there was no difference in the total number of eMyHC⁺ myofibers (Fig S1C). We also measured protein levels of various markers of muscle regeneration, such as MyoD, myogenin, and eMyHC. Results showed that there was no significant difference in the protein levels of MyoD between the two groups. However, protein levels of myogenin were significantly reduced in the injured TA muscle of Fn14-KO mice compared with WT. Intriguingly, we found a significant increase in the levels of eMyHC protein in the injured TA muscle of Fn14-KO mice compared with WT mice (Fig S1D and E). Because satellite cells play an indispensable role in muscle regeneration (Yin et al, 2013), we next investigated whether genetic ablation of Fn14 affects abundance of satellite cells in the skeletal muscle. Transverse sections generated from uninjured and 5d-injured TA muscles of WT and Fn14-KO mice were immunostained for Pax7 (a marker for satellite cells) and laminin (to mark myofiber boundaries) proteins, whereas nuclei were counterstained with DAPI (Fig 1G). There was no significant difference in the number of Pax7⁺ cells in uninjured or injured TA muscles of WT and Fn14-KO mice (Fig 1H). Moreover, protein levels of Pax7 were comparable in the 5d-injured TA muscle of WT and Fn14-KO mice (Fig S1C and D). Western blot confirmed that Fn14 protein was absent in the 5d-injured TA muscle of Fn14-KO mice (Fig 1I). These results suggest that constitutive deletion of Fn14 attenuates skeletal muscle regeneration in adult mice.

Myoblast-specific deletion of Fn14 inhibits muscle regeneration

A previously published report suggests that attenuation in muscle regeneration in whole-body Fn14-KO mice could be attributed to the diminished inflammatory response in the skeletal muscle in response to injury (Girgenrath et al, 2006). However, the cell-autonomous role of Fn14 in skeletal muscle regeneration remained unknown. We investigated whether targeted ablation of Fn14 in differentiated myofibers or myoblasts affects skeletal muscle regeneration in adult mice. To generate myofiber-specific Fn14-KO mice, we employed muscle creatine kinase (MCK)-Cre mice. Floxed Fn14 (henceforth Fn14^fl/fl) mice were crossed with MCK-Cre mice to generate Fn14^fl/fl; MCK-Cre (henceforth Fn14^mKO) and littermate Fn14^fl/fl mice, as described previously (Tomaz da Silva et al, 2022). The TA muscle of 8-wk-old Fn14^fl/fl and Fn14^mKO mice was injured by intramuscular injection of 1.2% BaCl₂ solution, and muscle regeneration was studied at day 5 post-injury. There was no significant difference in overall body weight of Fn14^fl/fl and Fn14^mKO mice (Fig S2A). Moreover, there was no difference in wet weight of the corresponding uninjured or 5d-injured TA muscle normalized by body weight between Fn14^fl/fl and Fn14^mKO mice (Fig S2B). We next generated TA muscle transverse sections followed by performing H&E staining and morphometric analysis (Fig S2C). There was no significant difference in the average myofiber CSA or the percentage of myofibers containing two or more centrally located nuclei in the 5d-injured TA muscle of the two groups (Fig S2D–F). Furthermore, anti-eMyHC staining showed no difference in the expression of eMyHC in the regenerating TA muscle of Fn14^fl/fl and Fn14^mKO mice (Fig S2G), suggesting that myofiber-specific deletion of Fn14 does not affect skeletal muscle regeneration in adult mice.

We next sought to investigate whether myoblast-specific ablation of Fn14 affects skeletal muscle regeneration in adult mice. Fn14^fl/fl mice were crossed with Myod1-Cre mice to generate myoblast-specific Fn14-KO (Fn14^fl/fl;Myod1-Cre; henceforth Fn14^myoKO) and littermate Fn14^fl/fl mice. Next, the TA muscle of 8-wk-old littermate Fn14^fl/fl and Fn14^myoKO mice was injured by intramuscular injection of 1.2% BaCl₂ solution. Muscle regeneration was evaluated at day 5 or 14 post-injury. There was no significant difference in overall body weight or uninjured TA muscle weight normalized by body weight of Fn14^fl/fl and Fn14^myoKO mice (Fig 2A and B). However, on day 5 post-injury, there was a significant decrease in TA muscle wet weight normalized by body weight in both groups. Interestingly, wet weight of the 5d-injured TA muscle was significantly lower in Fn14^myoKO mice compared with Fn14^fl/fl mice (Fig 2B). In addition, there was a significant increase in TA muscle weight normalized by body weight, on day 14 post-injury, only in Fn14^fl/fl mice compared with the uninjured muscle (Fig 2C). We next generated TA muscle transverse sections and performed H&E staining (Fig 2D). Intriguingly, the size of regenerating myofibers was significantly reduced and there was an increase in mononuclear cell infiltration between myofibers in the TA muscle of Fn14^myoKO mice compared with Fn14^fl/fl mice at both day 5 and day 14 post-injury (Fig 2D–G). In addition, there was a significant reduction in the percentage of myofibers containing two or more centrally located nuclei in the 5d-injured TA muscle of Fn14^myoKO mice compared with corresponding Fn14^fl/fl mice (Fig 2H). Myod1-Cre is a knock-in line in which Cre recombinase cDNA has been inserted at one of the alleles of the Myod1 gene (Chen et al, 2005). Because Myod1 is a critical regulator of myogenesis, we investigated whether the lack of one allele of Myod1 contributes to the observed phenotype in myoblast-specific Fn14-KO mice. To address this issue, we compared muscle regeneration defects in Fn14^myoKO mice with the mice that have the Myod1-Cre allele but heterozygous for the floxed Fn14 allele (i.e., Fn14^fl/wt; Myod1-Cre). Results showed that there was a significant deficit in muscle regeneration in Fn14^myoKO mice compared with corresponding Fn14^fl/wt; Myod1-Cre mice at day 5 post-injury, confirming that myoblast Fn14 is essential for muscle regeneration (Fig S3A and B).

Two consecutive injuries carried out 3 or 4 wk apart, with enough time elapse allowing the regeneration of the muscle after the first injury, is another model to study muscle progenitor cell function, maintenance, or depletion (Hardy et al, 2016; Roy et al, 2021). Therefore, we next examined muscle regeneration in Fn14^fl/fl and Fn14^myoKO mice after performing double injury. On day 21 after the first injury, the TA muscle was injured again by intramuscular injection of 1.2% BaCl₂ solution. At day 5 after the second injury, the muscle was isolated and analyzed by performing H&E staining (Fig 2I). Our analysis showed that the average myofiber CSA and the percentage of myofibers containing two or more centrally located nuclei were significantly reduced in Fn14^myoKO mice compared with Fn14^fl/fl mice upon double injury (Fig 2J and K). Taken together, these results suggest that Fn14-mediated signaling in myoblasts but not in myofibers is essential for efficient muscle regeneration.

Myoblast-specific ablation of Fn14 reduces myofiber formation

To understand the mechanisms by which myoblast Fn14 regulates skeletal muscle regeneration, we studied the expression of early markers of muscle regeneration. We first performed immunostaining for eMyHC and laminin proteins on 5d-injured TA muscle sections. Nuclei were counterstained with DAPI (Fig 3A). Results showed that the percentage of eMyHC-positive myofibers containing two or more centrally located nuclei was significantly reduced in Fn14^myoKO mice compared with Fn14^fl/fl mice (Fig 3B). In a parallel experiment, uninjured and 5d-injured TA muscles of Fn14^fl/fl and Fn14^myoKO mice were analyzed by performing qRT–PCR and Western blot. There was a significant increase in the mRNA and protein levels of eMyHC, MyoD, and myogenin in the injured TA muscle compared with the contralateral uninjured muscle in both groups (Fig 3C–E). However, mRNA levels of eMyHC, MyoD, and myogenin, as well as protein levels of myogenin, were significantly higher in injured TA muscle of Fn14^myoKO mice compared with Fn14^fl/fl mice (Fig 3C–E). To determine whether myoblast-specific deletion of Fn14 affects the number of satellite cells, transverse sections generated from the 5d-injured TA muscle of Fn14^fl/fl and Fn14^myoKO mice were immunostained for Pax7 and laminin proteins. Nuclei were identified by staining with DAPI (Fig 3F). Interestingly, the number of Pax7⁺ cells per unit area was significantly higher in the injured TA muscle of Fn14^myoKO mice compared with Fn14^fl/fl mice (Fig 3G). In addition, mRNA levels of Pax7 were significantly increased in the 5d-injured TA muscle of Fn14^myoKO mice compared with Fn14^fl/fl mice (Fig 3H). These results suggest that although deletion of Fn14 in myoblasts reduces myofiber formation, it does not impede the expression of various MRFs and abundance of satellite cells in the regenerating skeletal muscle of adult mice.

Fn14 promotes myoblast fusion in regenerating myofibers

Because ablation of Fn14 in myoblasts does not reduce the number of satellite cells or the expression of various MRFs in the regenerating muscle, we next investigated whether Fn14 has any role in regulating myoblast fusion during muscle regeneration. For this experiment, the TA muscle of 8-wk-old Fn14^fl/fl and Fn14^myoKO mice was injured by intramuscular injection of 1.2% BaCl₂ solution. After 72 h, the mice were given a single intraperitoneal injection of EdU and the TA muscle was isolated 11 d later. Finally, transverse muscle sections were generated and processed for the detection of EdU⁺ myonuclei (Fig 4A). Results showed that the number of EdU⁺ nuclei per myofiber was significantly reduced in the TA muscle of Fn14^myoKO mice compared with Fn14^fl/fl mice (Fig 4B). In addition, the percentage of myofibers containing two or more EdU⁺ nuclei was also significantly reduced in the TA muscle of Fn14^myoKO mice compared with Fn14^fl/fl mice (Fig 4C). Impairment in myoblast fusion results in the persistent expression of eMyHC in regenerating myofibers. We next performed immunostaining for eMyHC protein in the TA muscle at day 14 after injury (Fig 4D). There were almost no eMyHC⁺ myofibers in the 14d-injured TA muscle of Fn14^fl/fl mice suggesting normal progression of muscle regeneration. In contrast, small-sized eMyHC⁺ myofibers were in abundance in the 14d-injured TA muscle of Fn14^myoKO mice (Fig 4D and E).

Myoblast fusion exhibits its peak between 4 and 7 d after muscle injury followed by maturation and functional recovery phases (Forcina et al, 2020). To understand the molecular mechanisms by which Fn14 regulates myoblast fusion in vivo, we also measured transcript levels of a few profusion molecules in the 5d-injured TA muscle of Fn14^fl/fl and Fn14^myoKO mice. Results showed that mRNA levels of M-cadherin, caveolin-3, β1D-integrin, and myomerger-L were significantly reduced in the 5d-injured TA muscle of Fn14^myoKO mice compared with Fn14^fl/fl mice. In contrast, there was no significant difference in mRNA levels of N-cadherin, myoferlin, myomaker, ADAM12, and myomerger-S in the 5d-injured TA muscle of Fn14^fl/fl and Fn14^myoKO mice (Fig 4F). These results suggest that Fn14 mediates myoblast fusion during muscle regeneration in adult mice.

Fn14 coordinates the differentiation and fusion of cultured myoblasts

To investigate the role and mechanisms of action of Fn14 during myogenesis, we established primary myoblast cultures from WT and whole-body Fn14-KO mice. We used myoblasts isolated from whole-body Fn14-KO mice rather than Fn14^myoKO mice to mitigate any potential effect of the loss of the expression of Myod1 from one allele on myogenesis. Using RNA-Seq, we were able to define the effect of Fn14 deficiency on global gene expression in cultured myoblasts. The loss of Fn14 resulted in the differential expression of 1,383 genes by over 1.5-fold (computed using a FDR of P < 0.05) in Fn14-KO versus WT myoblasts. Of the 1,383 genes, 499 genes were down-regulated, whereas 884 genes were up-regulated in Fn14-KO compared with WT myoblasts (Fig 5A). Functional enrichment analysis of the differentially expressed genes using gene ontology (GO) annotations showed that several gene sets related to morphogenesis, migration, motility, and movement of subcellular components were down-regulated in Fn14-KO compared with WT cultures. In contrast, the gene sets involved in muscle cell development, differentiation, and regulation of immune response were up-regulated in Fn14-KO cultures (Fig 5B). Similar results were obtained when up-regulated and down-regulated gene sets were evaluated using the Metascape gene annotation and analysis tool (Fig S4). Interestingly, RNA-Seq analysis showed the higher expression of myogenin in Fn14-KO cultures compared with WT cultures in a normal growth medium (Fig 5C). We next investigated how the loss of Fn14 affects myogenic differentiation and myotube formation in vitro. WT and Fn14-KO myoblasts were incubated in DM for 6, 24, or 48 h, and the cultures were immunostained for myosin heavy chain (MyHC), a marker of muscle differentiation (Fig 5D). Results showed that the number of MyHC⁺-mononucleated cells was significantly higher in Fn14-KO cultures compared with WT cultures at all the time points after addition of DM (Fig 5E). Importantly, the proportion of myotubes containing 2–4 nuclei was significantly increased, whereas those containing 5–10 or more than 10 nuclei were significantly reduced in Fn14-KO cultures compared with corresponding WT cultures at 24 or 48 h of incubation in differentiation medium (Fig 5F). To understand whether Fn14 affects myogenic differentiation, we measured protein levels of myogenin and MyHC at different time points after addition of DM. Consistent with RNA-Seq results, there was a slight increase in the levels of myogenin and MyHC in Fn14-KO cultures compared with WT cultures at 0 and 6 h after addition of DM. However, protein levels of both myogenin and MyHC were comparable in WT and Fn14-KO cultures at 24, 48, and 72 h after addition of DM (Figs 5G and S5). These results suggest that the loss of Fn14 leads to precocious differentiation of myoblasts, but myotube formation is diminished because of the inhibition of myoblast fusion.

There are published reports suggesting that the Fn14 receptor can function independent of its ligand TWEAK (Winkles, 2008; Burkly, 2014; Tajrishi et al, 2014). Because deletion of Fn14 inhibits myotube formation, we next sought to determine whether the forced expression of Fn14 in myoblasts is sufficient to improve myotube formation after induction of differentiation. Primary myoblasts were transduced with the retrovirus expressing EGFP or Fn14 protein. The cells were then incubated in DM for 48 h, and the myotube formation was examined by performing anti-MyHC and DAPI staining (Fig 5H). Interestingly, the overexpression of Fn14 improved myotube formation. The average myotube diameter was significantly higher in Fn14-overexpressing cultures compared with control cultures (Fig 5I). We also investigated whether the forced expression of Fn14 affects the expression of various MRFs and MyHC. Although there was no difference in MyoD, the levels of myogenin were significantly higher in Fn14-overexpressing cultures compared with controls in the growth medium. However, the levels of MyoD or myogenin were comparable between control and Fn14-overexpressing cultures after 48 h of incubation in DM. Intriguingly, we found that the levels of MyHC were significantly higher in Fn14-overexpressing cultures compared with controls at 48 h of addition of DM, suggesting that the overexpression of Fn14 augments myogenic differentiation (Fig S6A and B). The p38 MAPK signaling and Akt signaling positively regulate initial stages of myogenic differentiation (Perdiguero et al, 2007; Gardner et al, 2012; Brennan et al, 2021). Therefore, we investigated whether the forced expression of Fn14 affects activation of p38 MAPK or Akt in cultured myoblasts. Results showed that the levels of phosphorylated p38 (p-p38) and phosphorylated Akt (p-Akt) were significantly higher in Fn14-overexpressing cultures compared with control cultures incubated in the normal growth medium. However, the levels of p-p38 or p-Akt were comparable between control and Fn14-overexpressing (Fn14-OE) myoblast cultures at 48 h of incubation in DM (Fig S6C–E). Western blot analysis confirmed increased levels of Fn14 in cultures transduced with the Fn14 retrovirus (Fig S6A and B). Altogether, these results suggest that Fn14 promotes myotube formation through stimulating myoblast fusion and fine-tuning myogenic differentiation.

Fn14 regulates the expression of components of calcium-dependent signaling

To understand the signaling mechanisms by which Fn14 promotes myotube formation, we measured the phosphorylation of various signaling molecules that are known to promote myoblast fusion. WT and Fn14-KO myoblasts were incubated in DM for different time periods, and the cell extracts prepared were analyzed by performing Western blot. Results showed that there was no difference in the levels of phosphorylated ERK1/2 and ERK5 between WT and Fn14-KO cultures. Furthermore, the levels of phosphorylated p65 (a marker of canonical NF-κB signaling) and relative levels of p100/p52 (a marker of non-canonical NF-κB signaling) were comparable in WT and Fn14-KO cultures at different time points after addition of DM (Fig 6A). Similarly, there was no significant difference in the levels of phosphorylated ERK1/2 and phosphorylated p65, or in relative amounts of p100/p52 protein in the 5d-injured TA muscle of Fn14^fl/fl and Fn14^myoKO mice (Fig 6B and C). Published reports suggest that the calcium (Ca²⁺)-dependent signaling pathway also plays an important role in myoblast fusion during myogenesis (Constantin et al, 1996; Bijlenga et al, 2000; Hindi et al, 2013). Interestingly, the levels of NFATc2 protein, a component of Ca²⁺–calcineurin–NFAT pathway, were considerably reduced in Fn14-KO cultures compared with WT cultures (Fig 6A). Moreover, transcript levels of various regulators of calcium-dependent signaling, such as calpain-3, calsequestrin-1, calsequestrin-2, RyR1, and CamKIIb, but not calmodulin1, were significantly reduced in Fn14-KO myoblast cultures compared with WT cultures (Fig 6D). We also studied the expression of these molecules in the regenerating muscle of Fn14^fl/fl and Fn14^myoKO mice. Results showed that there was a significant reduction in the mRNA levels of calpain-3, calsequestrin-1, calsequestrin-2, and CamKIIb in the 5d-injured TA muscle of Fn14^myoKO mice compared with Fn14^fl/fl mice (Fig 6E). Collectively, these results suggest that Fn14-mediated signaling regulates the expression of various molecules involved in the regulation of calcium-dependent pathways.

Fn14 stimulates Wnt signaling in myoblasts

Our RNA-Seq analysis showed that the expression of a few components of Wnt signaling is differentially regulated between WT and Fn14-KO cultured myoblasts (Fig 7A). Interestingly, canonical Wnt signaling also plays a major role in myoblast fusion (Abmayr & Pavlath, 2012; Hindi et al, 2013). To understand the impact of Fn14 on Wnt signaling, we measured the levels of Wnt3a protein in WT and Fn14-KO cultures at different time points after addition of DM. Intriguingly, the protein levels of Wnt3a were increased in WT cultures but not in Fn14-KO cultures after addition of DM (Fig 7B). Phosphorylation of GSK-3β at Ser9 residue leads to its inactivation, which is required for activation of canonical Wnt signaling (Tejeda-Munoz & Robles-Flores, 2015). We found that the levels of phosphorylated GSK-3β (Ser9) were reduced in Fn14-KO cultures compared with WT cultures (Fig 7B). We previously reported that MyD88 protein regulates Wnt signaling to promote myoblast fusion during myogenesis (Hindi et al, 2017b). Accordingly, we found that the levels of MyD88 were also reduced in differentiating Fn14-KO cultures compared with WT cultures (Fig 7B). By performing qRT–PCR, we next investigated how the expression of various components of Wnt signaling is regulated in the regenerating skeletal muscle of Fn14^myoKO mice. Interestingly, the mRNA levels of Wnt ligands Wnt4, Wnt5a, and Wnt11; Wnt receptor Frizzled-4 (Fzd4); and Wnt target gene Axin-2 were found to be significantly reduced in the 5d-injured TA muscle of Fn14^myoKO mice compared with the corresponding injured TA muscle of Fn14^fl/fl mice (Fig 7C–E).

We also studied the effects of the overexpression of Fn14 on the expression of components of Wnt signaling in myoblasts. Primary WT myoblasts were transduced with control or Fn14-expressing retrovirus followed by performing qRT–PCR. Interestingly, the mRNA levels of Wnt4, Wnt5a, Fzd1, Fzd2, Fzd4, and Axin-2 were significantly increased in Fn14-overexpressing myoblasts compared with corresponding controls (Fig 7F–H). These results suggest that Fn14-mediated signaling cross-talks with Wnt signaling during myogenesis.

Forced activation of Wnt signaling rescues fusion defects in Fn14-KO cultures

Because the loss of Fn14 results in the inhibition of Wnt signaling, we next investigated whether forced activation of Wnt signaling improves fusion in Fn14-KO cultures. For this experiment, WT and Fn14-KO myoblasts were incubated in DM with or without recombinant Wnt3a protein for 48 h. In parallel, the cultures were also treated with LiCl, which activates canonical Wnt signaling through inhibiting GSK-3β (Lindsley et al, 2006). Finally, the cultures were fixed and the myotube formation was evaluated by immunostaining for MyHC protein (Fig 8A). Interestingly, addition of Wnt3a protein or LiCl significantly improved myotube formation in Fn14-KO cultures. Quantitative analysis confirmed that Wnt3a and LiCl significantly increased the proportion of myotubes containing 10 or more nuclei (Fig 8B). Furthermore, the average diameter of myotubes in Fn14-KO cultures was significantly higher in cultures treated with Wnt3a protein compared with corresponding controls (Fig 8C). These results suggest that Fn14 promotes myotube formation, at least in part, through activation of canonical Wnt signaling.

Animals

C57BL/6 mice were purchased from Jackson Labs. Whole-body Fn14-knockout (i.e., Fn14-KO) mice have been previously described (Girgenrath et al, 2006). Floxed Fn14 (i.e., Fn14^fl/fl) mice were generated by Taconic by inserting loxP sites upstream of exon 2 and downstream of exon 4 of Fn14 as described previously (Gupta et al, 2021). Myofiber-specific Fn14-KO (henceforth Fn14^mKO) mice were generated by crossing Fn14^fl/fl mice with MCK-Cre (strain: B6.FVB(129S4)-Tg(Ckmm-cre)5Khn/J; Jackson Laboratory) mice as described previously (Tomaz da Silva et al, 2022). Myoblast-specific Fn14-KO (Fn14^myoKO) mice were generated by crossing Fn14^fl/fl mice with Myod1-Cre (Jax strain: FVB.Cg-Myod1 ^{tm2.1(icre)Glh}/J) mice. All mice were in the C57BL/6 background, and their genotype was determined by PCR from tail DNA. All the experiments were performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All the animals were handled according to an approved Institutional Animal Care and Use Committee protocol (PROTO201900043) of the University of Houston. All surgeries were performed under anesthesia, and every effort was made to minimize suffering.

Skeletal muscle injury and in vivo myoblast fusion assay

The TA muscle of adult mice was injected 50 μl of 1.2% BaCl₂ (Sigma Chemical Co.) dissolved in saline to induce necrotic injury as described previously (Ogura et al, 2015; Hindi and Kumar, 2016). The TA muscle was collected at different time points after injury from euthanized mice for biochemical and morphometric analysis. To study myoblast fusion in vivo, the mice were given an intraperitoneal injection of EdU (4 μg per gram body weight) at day 3 after intramuscular injection of 1.2% BaCl₂ into the TA muscle. On day 11 after EdU injection, the TA muscle was isolated and sectioned in a microtome cryostat. The sections were subsequently immunostained with anti-Laminin for marking boundaries of myofibers and processed for the detection of EdU⁺ nuclei similar to as described previously (Hindi et al, 2017a). The EdU⁺ nuclei on muscle sections were detected as instructed in the Click-iT EdU Alexa Fluor 488 Imaging Kit (Invitrogen). Finally, images were captured and the number of intramyofiber EdU⁺ myonuclei/myofiber, percentage of 2 or more EdU⁺ centrally nucleated fibers, and percentage of EdU⁺ myonuclei/total nuclei were quantified using NIH ImageJ software. To reduce variations, three to four different sections from the mid-belly of each muscle were included for analysis.

Histology and morphometric analysis

The TA muscle was isolated from mice, snap-frozen in liquid nitrogen, and sectioned with a microtome cryostat. For the assessment of muscle morphology and quantification of the fiber CSA, 7-μm-thick transverse sections of the TA muscle were stained with H&E. The sections were examined under an Eclipse TE2000-U microscope (Nikon). For quantitative analysis, the CSA of myofibers was analyzed in H&E-stained TA muscle sections using NIH ImageJ software. For each muscle, the distribution of the fiber CSA was calculated by analyzing approximately 300 myofibers.

Immunohistochemistry

For immunohistochemistry studies, frozen TA or plantaris muscle sections were fixed in acetone or 4% PFA in PBS, blocked in 2% BSA in PBS for 1 h, and incubated with mouse anti-Pax7, mouse anti-eMyHC, and rabbit anti-laminin, in blocking solution at 4°C overnight under humidified conditions. The sections were washed briefly with PBS before incubation with goat anti-mouse Alexa Fluor 594 and goat anti-rabbit Alexa Fluor 488 secondary antibodies for 1 h at room temperature and then washed three times for 15 min with PBS. Nuclei were counterstained with DAPI. The slides were mounted using the fluorescence medium (Vector Laboratories) and visualized at room temperature on a Nikon Eclipse TE2000-U microscope (Nikon), a digital camera (Nikon Digital Sight DS-Fi1), and NIS-Elements BR 3.00 software (Nikon). Image levels were equally adjusted using Photoshop CS6 software (Adobe).

Myoblast isolation, culturing, and fusion assay

Primary myoblasts were prepared from the hindlimbs of 8-wk-old male or female mice following a protocol as described previously (Hindi et al, 2017a, 2017b). Briefly, hindlimb muscles were collected and minced/digested with collagenase type II in DMEM. Muscle tissues were mechanically scraped, centrifuged, and resuspended in DMEM, and cell mixture was filtered and incubated at 37°C for 72 min. After that, cells were purified by doing preplating to assure myoblast purification. To induce differentiation, the cells were incubated in a differentiation medium (DM; 2% horse serum in DMEM). Cultures were fixed with 4% PFA in PBS for 15 min at room temperature and then permeabilized with 0.1% Triton X-100 in PBS for 5–8 min. Cells were blocked with 2% BSA in PBS and incubated with mouse anti-MyHC (MF-20) overnight at 4°C and goat anti-mouse Alexa Fluor 568 at room temperature for 1 h. Nuclei were counterstained with DAPI for 3 min. Stained cells were photographed and analyzed using a fluorescent inverted microscope (Nikon Eclipse TE2000-U), a digital camera (Digital Sight DS-Fi1), and Elements BR 3.00 software (Nikon). To measure fusion efficiency, more than 100 of MyHC-positive myotubes containing 2 or more nuclei were counted. More than five field images were analyzed per experimental group. To measure the average diameter of myotubes, 100–120 myotubes per group were included. For consistency, diameters were measured at the midpoint along with the length of the MyHC⁺ myotubes. The myotube diameter was measured using NIH ImageJ software. Results presented are from four to five independent experiments. Image levels were equally adjusted using Photoshop CS6 software (Adobe).

Generation and use of retroviruses

A pBABE-Puro empty vector and a pBABE-Puro-EGFP plasmid were purchased from Addgene. Mouse Fn14 cDNA was isolated and ligated at BamHI and SalI sites in the pBABE-Puro plasmid. The integrity of the cDNA was confirmed by performing DNA sequencing. About 5 × 10⁶ Platinum-E packaging cells (Cell Biolabs, Inc.) were transfected with 5 μg of pBABE-Puro-Fn14 or pBABE-Puro-EGFP using FuGENE-HD (Promega). After 24 h of transfection, the medium was replaced with 10% FBS. 48 h after transfection, viral supernatants were collected, filtered through 0.45-micron filters, and then added to primary myoblasts in growth media containing 10 μg/ml polybrene. After two successive retroviral infections, cells were grown for 48 h and selected using 1.2 μg/ml puromycin for 2 wk.

RNA isolation and qRT–PCR

RNA isolation and qRT–PCR were performed following a standard protocol as described previously. In brief, total RNA was extracted from uninjured and injured TA muscles of mice or cultured myoblasts using TRIzol reagent (Thermo Fisher Scientific) and RNeasy Mini Kit (QIAGEN) according to the manufacturers’ protocols. First-strand cDNA for PCR analyses was made with a commercially available kit (iScript cDNA Synthesis Kit; Bio-Rad Laboratories). The quantification of mRNA expression was performed using the SYBR Green dye (Bio-Rad SsoAdvanced Universal SYBR Green Supermix) method on a sequence detection system (CFX384 Touch Real-Time PCR Detection System; Bio-Rad Laboratories). The sequence of the primers is described in Table S1. Data normalization was accomplished with the endogenous control (β-actin), and the normalized values were subjected to a 2^−ΔΔCt formula to calculate the fold change between control and experimental groups.

Western blot

The relative amount of various proteins was determined by performing Western blot following a standard protocol. In brief, TA muscles of mice or cultured primary myoblasts were washed with PBS and homogenized in lysis buffer (50 mM Tris-Cl [pH 8.0], 200 mM NaCl, 50 mM NaF, 1 mM dithiothreitol, 1 mM sodium orthovanadate, 0.3% IGEPAL, and protease inhibitors). ∼100 μg protein was resolved on each lane on 8–12% SDS–PAGE, transferred onto a nitrocellulose membrane, and probed using the specific primary antibody (Table S2). Bound antibodies were detected by secondary antibodies conjugated to horseradish peroxidase (Cell Signaling Technology). Signal detection was performed by an enhanced chemiluminescence detection reagent (Bio-Rad). Approximate molecular masses were determined by comparison with the migration of prestained protein standards (Bio-Rad). Uncropped gel images are presented as source data.

RNA-sequencing and data analyses

Total RNA from WT and Fn14-KO cultures was extracted using TRIzol reagent (Thermo Fisher Scientific) using RNeasy Mini Kit (QIAGEN) according to the manufacturers’ protocols. The mRNA-Seq library was prepared using poly (A)-tailed enriched mRNA at the UT Cancer Genomics Center using the KAPA mRNA HyperPrep Kit protocol (KK8581; Roche, Holding AG) and the KAPA Unique Dual-indexed Adapter kit (KK8727; Roche). The Illumina NextSeq 550 was used to produce 75 base paired-end mRNA-Seq data at an average read depth of ∼38 M reads/sample. RNA-Seq fastq data were processed using CLC Genomics Workbench 20 (QIAGEN). Illumina sequencing adapters were trimmed, and reads were aligned to the mouse reference genome RefSeq GRCm39.105 from the Biomedical Genomics Analysis Plugin 20.0.1 (QIAGEN). Normalization of RNA-Seq data was performed using the trimmed mean of M-values. Genes with fold change (FC) ≥ ǀ1.5ǀ (or log₂FC ≥ ǀ0.5ǀ) and FDR <0.05 were assigned as differentially expressed genes and represented in a volcano plot using the ggplot function in R software (v 4.2.2). Over-representation analysis was performed with a hypergeometric test using WebGestalt (v 0.4.3) (Benjamini & Hochberg, 1995). Gene ontology (GO) biological processes associated with the up-regulated and down-regulated genes were identified with a FDR cutoff value of 0.05. Network enrichment analysis was performed using the Metascape gene annotation and analysis tool (metascape.org) as described previously (Zhou et al, 2019). Heatmaps were generated using the heatmap.2 function (Gu & Hubschmann, 2022) using z-scores calculated based on transcripts per million (TPM) values. TPM values were converted to log (TPM+1) to handle zero values. Genes involved in specific pathways were manually selected for heatmap expression plots. The mean TPM values for WT cultures in the growth medium are provided in Table S3. All the raw data files can be found on the NCBI SRA repository using the accession code PRJNA999372.

Statistical analyses and experimental design

The sample size was calculated using power analysis methods for a priori determination based on the s.d., and the effect size was previously obtained using the experimental procedures employed in the study. For animal studies, we calculated the minimal sample size for each group as eight animals. Considering a likely drop-off effect of 10%, we set the sample size of each group of six mice. For some experiments, three to four animals were sufficient to obtain statistical significant differences. Animals of the same sex and the same age were employed to minimize physiological variability and to reduce s.d. from the mean. The exclusion criteria for animals were established in consultation with a veterinarian and experimental outcomes. In case of death, skin injury, ulceration, sickness, or weight loss of >10%, the animal was excluded from analysis. Tissue samples were excluded in cases such as freeze artifacts on histological sections or failure in extraction of RNA or protein of suitable quality and quantity. We included animals from different breeding cages by random allocation to the different experimental groups. All animal experiments were blinded using number codes until the final data analyses were performed. Statistical tests were used as described in Figure legends. Results are expressed as the mean + SEM. Statistical analyses used a two-tailed t test or a two-way ANOVA followed by Tukey’s multiple-comparison test to compare quantitative data populations with normal distribution and equal variance. A value of P ≤ 0.05 was considered statistically significant unless otherwise specified.