Activation of Fn14 signaling in satellite cells during muscle regeneration. We first investigated how the expression of Fn14 is regulated during muscle regeneration in adult mice. Using a published single-cell RNA-seq (scRNA-seq) dataset (NCBI GEO GSE143435) (31), we analyzed the expression of the Fn14 gene (Tnfrsf12a) in satellite cells at different time points after injury. In this study, the tibialis anterior (TA) muscle of adult C57BL/6J mice was injured by a single intramuscular injection of notexin. The TA muscle was collected at various time points (0 [uninjured], 2, 5, and 7 days) followed by preparation of single-cell suspensions and scRNA-seq (31). Our analysis showed that Fn14 is expressed in satellite cells of uninjured TA muscle. Intriguingly, the mRNA levels of Fn14 were increased in satellite cells on day 2 and persisted through days 5 and 7 after injury. By contrast, the mRNA levels of TWEAK (encoded by Tnfsf12) in satellite cells were low in uninjured muscle and did not show any increase after muscle injury (Figure 1A).

Expression of Fn14 in satellite cells. (A) Violin plots showing gene expression of Fn14 (gene: Tnfrsf12a) and TWEAK (gene: Tnfsf12) in satellite cells at different time points after muscle injury in mice analyzed from a publicly available scRNA-seq dataset (GSE143435). (B) Single cells isolated from uninjured and injured TA muscle of WT mice were analyzed by FACS for the expression of α7-integrin and Fn14 protein. Representative scatter plots of FACS-based analysis demonstrate presence of Fn14+ cells among α7-integrin+ population (lower panel). (C) Satellite cells isolated from TA muscle of WT mice on day 0, 5, or 21 after injury were analyzed for mRNA levels of Fn14 by performing qPCR. n = 3 mice in each group. Data are presented as mean ± SEM and were analyzed by 1-way ANOVA followed by Tukey’s multiple-comparison test. #P ≤ 0.05, values significantly different from uninjured muscle (day 0). *P ≤ 0.05, values significantly different from 5-day-injured TA muscle. (D) Representative photomicrographs showing expression of Fn14 in Pax7+ cells in cultured mouse primary myoblasts. Scale bars: 50 μm.

We also studied the expression of Fn14 in satellite cells of uninjured and regenerating skeletal muscle of mice. To induce regeneration, the right-side TA muscle of 8-week-old male C57BL/6J mice was injured by intramuscular injection of 1.2% BaCl2 solution, whereas the contralateral TA muscle served as uninjured control. The muscle tissue was collected on day 5 after injury and Fn14 expression in satellite cells was analyzed by a FACS-based approach using a well-established panel of cell surface markers: Lin– (CD31–CD45–Sca-1–TER119–) and α7-integrin+ (15). Our analysis showed that Fn14 was present in the α7-integrin+ cells of both uninjured and injured TA muscles of mice (Figure 1B). We also investigated how the expression of Fn14 in satellite cells is regulated on day 21, a time point when muscle regeneration is complete, and satellite cells begin to return to quiescence. For this experiment, the TA muscle of adult C57BL/6J mice was injured for 5 or 21 days, followed by isolation of satellite cells by a magnetic-activated cell sorting (MACS) approach and performing quantitative PCR (qPCR). Results showed that the mRNA levels of Fn14 were significantly higher in satellite cells of 5-day-injured TA muscle compared with uninjured muscle. Interestingly, the mRNA levels of Fn14 in satellite cells were significantly lower in 21-day-injured TA muscle compared with both uninjured and 5-day-injured TA muscle (Figure 1C). These results suggest a transient increase in the expression of Fn14 in satellite cells during muscle regeneration.

Satellite cells are rapidly activated (express MyoD) and transition to the progenitor stage after isolation and culturing. By performing immunostaining for Fn14 and Pax7 (a marker of satellite cells) proteins, we investigated the expression of Fn14 in myoblast cultures established from hind limb muscle of C57BL/6J mice. Consistent with the scRNA-seq analysis, Fn14 protein was present in Pax7+ cells in myoblast cultures (Figure 1D). These results demonstrate that Fn14 is expressed in muscle progenitor cells both in vivo and in vitro.

Inducible ablation of Fn14 in satellite cells inhibits muscle regeneration in adult mice. We investigated whether the satellite cell–specific deletion of Fn14 affects skeletal muscle regeneration in adult mice. Floxed Fn14 (henceforth Fn14fl/fl) mice were crossed with tamoxifen-inducible satellite cell-specific Cre mice (Pax7-CreERT2) to generate satellite cell–specific inducible Fn14-KO mice (i.e., Fn14fl/fl;Pax7-CreERT2, henceforth Fn14scKO) and littermate Fn14fl/fl mice. To induce Fn14 deletion, 7-week-old male Fn14scKO mice were given daily intraperitoneal (i.p.) injections of tamoxifen (75 mg/kg body weight) for 4 consecutive days. The mice were kept on a tamoxifen-containing chow for the entire duration of the study. Littermate male Fn14fl/fl were also treated with tamoxifen and served as controls. Next, TA muscle on one side of Fn14fl/fl and Fn14scKO mice was injected with 1.2% BaCl2 solution, whereas the contralateral uninjured TA muscle served as a control. Muscle regeneration was evaluated on days 5, 14, and 21 after injury (Figure 2A). To confirm the inactivation of Fn14 in satellite cells of Fn14scKO mice, we performed FACS and qPCR analyses. Results showed that the proportion of Fn14+α7-integrin+ cells and the mRNA levels of Fn14 in purified satellite cells were significantly reduced in 5-day-injured TA muscle of Fn14scKO mice compared with the corresponding muscle of Fn14fl/fl mice (Supplemental Figure 1, A–C; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.187825DS1).

Satellite cell–specific ablation of Fn14 inhibits muscle regeneration. (A) Schematic representation of the experimental design. (B) Body weight of Fn14fl/fl and Fn14scKO mice. (C) Uninjured and 5-day-injured TA muscle wet weight normalized by body weight (BW) of Fn14fl/fl and Fn14scKO mice. (D) Uninjured and 14-day-injured TA muscle wet weight normalized by BW of Fn14fl/fl and Fn14scKO mice. (E) Representative photomicrographs of H&E-stained transverse sections of TA muscle of Fn14fl/fl and Fn14scKO mice at indicated time points after injury. Scale bars: 50 μm. (F–H) Quantitative analysis of average myofiber cross-sectional area (CSA) in TA muscle of Fn14fl/fl and Fn14scKO mice on (F) day 0, (G) day 5, and (H) day 14 after injury. (I) Representative photomicrographs of transverse sections of 5-day-injured TA muscle of Fn14fl/fl and Fn14scKO mice after immunostaining for eMyHC and laminin protein. Nuclei were identified by staining with DAPI. Scale bars: 50 μm. (J and K) Quantification of (J) average CSA of eMyHC+ laminin+ myofibers, and (K) percentage of eMyHC+ laminin+ myofibers containing 2 or more centrally located nuclei in 5-day-injured TA muscle of Fn14fl/fl and Fn14scKO mice. n = 3–6 mice in each group. All data are presented as mean ± SEM. *P ≤ 0.05, values significantly different from contralateral uninjured muscle of Fn14fl/fl or Fn14scKO mice. #P ≤ 0.05, values significantly different from corresponding 5-day- or 14-day-injured TA muscle of Fn14fl/fl mice analyzed by 2-way ANOVA followed by Tukey’s multiple-comparison test. @P ≤ 0.05, values significantly different from corresponding Fn14fl/fl mice analyzed by unpaired Student’s t test.

There was no significant difference in body weight or wet weight of muscle normalized by body weight between Fn14fl/fl and Fn14scKO mice under naive conditions (Figure 2, B–D). However, there was a significant decrease in muscle wet weight normalized by body weight in Fn14scKO mice compared with Fn14fl/fl mice on days 5 and 14 after injury (Figure 2, C and D). Next, transverse sections of TA muscle were generated, followed by performing hematoxylin and eosin (H&E) staining and morphometric analysis. There was no significant difference in the myofiber cross-sectional area (CSA) in TA muscle of Fn14fl/fl and Fn14scKO mice in naive muscle (Figure 2, E and F). In contrast, muscle regeneration was significantly impaired in Fn14scKO mice compared with their littermate Fn14fl/fl mice. At both days 5 and 14 after injury, there was a significant decrease in myofiber CSA of Fn14scKO mice compared with Fn14fl/fl mice (Figure 2, E, G, and H). Notably, defects in muscle regeneration (e.g., reduced myofiber CSA and presence of cellular infiltrate) in Fn14scKO mice persisted even on day 21 after injury (Figure 2E). In a separate experiment, we also investigated whether the deletion of only one allele of the Fn14 gene in satellite cells contributes to the observed phenotype in Fn14scKO mice. We compared muscle regeneration defects in male Fn14fl/wt;Pax7-CreERT2 (Fn14fl/scKO) mice with male Fn14scKO mice. There was a significant deficit in muscle regeneration in Fn14scKO mice compared with the corresponding Fn14fl/scKO or Fn14fl/fl mice on day 5 after injury (Supplemental Figure 1, D–F).

Deficits in muscle regeneration in Fn14scKO mice were also evidenced by immunostaining of 5-day-injured TA muscle sections for embryonic isoform of MyHC (eMyHC), which is expressed in newly formed myofibers (Figure 2I). Average CSA of eMyHC+ myofibers and the proportion of eMyHC+ myofibers containing 2 or more centrally located nuclei were significantly reduced in 5-day-injured TA muscle of Fn14scKO mice compared with Fn14fl/fl mice (Figure 2, J and K). We also investigated the expression of eMyHC in regenerating myofibers on day 14 and 21 after injury (Supplemental Figure 2, A–C). There were almost no eMyHC+ myofibers in the TA muscle of Fn14fl/fl mice on day 14 or 21 after injury, suggesting normal progression of muscle regeneration. In contrast, small-sized eMyHC+ myofibers were abundant in the muscle of Fn14scKO mice on day 14 after injury (Supplemental Figure 2, A and B). The presence of eMyHC+ myofibers in TA muscle sections of Fn14scKO mice was even prolonged to day 21 after injury (Supplemental Figure 2C). These results suggest that satellite cell–specific inducible deletion of Fn14 inhibits skeletal muscle regeneration in adult mice.

To understand the role of Fn14 in the regulation of the abundance of satellite cells, uninjured and 5-day-injured TA muscle sections of Fn14fl/fl and Fn14scKO mice were immunostained for Pax7 and laminin protein (Figure 3A). There was no significant difference in the number of satellite cells in uninjured TA muscle of Fn14fl/fl and Fn14scKO mice (Supplemental Figure 3, A and B). However, the number of satellite cells per unit area was significantly reduced in the TA muscle of Fn14scKO mice compared with Fn14fl/fl mice on days 5 and 14 after injury (Figure 3, A–C). Consistent with immunohistochemistry results, FACS analysis also showed that there was no significant difference in the proportion of satellite cells in uninjured TA muscle of Fn14fl/fl and Fn14scKO mice (Supplemental Figure 3, C and D). In contrast, the proportion of satellite cells in Fn14scKO mice was significantly reduced in 5-day-injured TA muscle of Fn14scKO mice compared with Fn14fl/fl mice (Figure 3, D and E). Moreover, the mRNA levels of Pax7 were also significantly reduced in 5-day-injured TA muscle of Fn14scKO mice compared with Fn14fl/fl mice (Figure 3F). Finally, the mRNA levels of Myod1, Myh3 (eMyHC), and Myog (myogenin) were significantly reduced in injured TA muscle of Fn14scKO mice compared with Fn14fl/fl mice (Figure 3G). These results suggest that targeted inducible deletion of Fn14 reduces the number of satellite cells and attenuates regeneration of injured skeletal muscle in adult mice.

Fn14 regulates satellite cell number in regenerating skeletal muscle. (A) Representative photomicrographs of 5-day-injured TA muscle sections of Fn14fl/fl and Fn14scKO mice after immunostaining for Pax7 (red) and laminin (green) protein. Scale bars: 50 μm. White arrows point to Pax7+ satellite cells. (B and C) Average number of Pax7+ cells per unit area in (B) 5-day-injured and (C) 14-day-injured TA muscle of Fn14fl/fl and Fn14scKO mice. (D) Representative FACS dot plots demonstrating the percentage of α7-integrin+ cells in 5-day-injured TA muscle of Fn14fl/fl and Fn14scKO mice. (E) Quantification of α7-integrin+ satellite cells in 5-day-injured TA muscle of Fn14fl/fl and Fn14scKO mice assayed by FACS. (F and G) Relative mRNA levels of (F) Pax7 and (G) Myod1, Myh3, and Myog in 5-day-injured TA muscle of Fn14fl/fl and Fn14scKO mice assayed by qPCR. n = 3–4 mice in each group. All data are presented as mean ± SEM. #P ≤ 0.05, values significantly different from corresponding muscle of Fn14fl/fl mice analyzed by unpaired Student’s t test.

Fn14 is required for the proliferation of satellite cells. While the tamoxifen-inducible approach, using a Cre-ERT2 system, is considered highly efficient for excising floxed alleles, it does not always result in a complete deletion of the target gene due to various factors, including incomplete recombination efficiency that results in a mixed population of cells with varying levels of gene expression. Therefore, for in vitro experiments, we employed primary myoblast cultures established from skeletal muscle of germline Fn14-KO mice that ensures complete and sustained ablation of Fn14 protein in all myogenic cells. We first performed bulk RNA-seq using primary myoblast cultures established from hind limb muscle of WT and Fn14-KO mice. Functional enrichment analysis of the differentially expressed genes (DEGs) using gene ontology (GO) annotations showed that several gene sets related to cell projection morphogenesis, skeletal system development, muscle cell proliferation, cell-cell signaling, and positive regulation of stem cell proliferation were downregulated in Fn14-KO compared with WT cultures. By contrast, the gene sets involved in the regulation of myoblast and myotube differentiation, negative regulation of cell proliferation, and tyrosine phosphorylation of STAT protein were upregulated in Fn14-KO cultures (Figure 4A). Further analysis of DEGs showed that gene expression of many molecules associated with cell proliferation and migration were downregulated in Fn14-KO cultures compared with WT cultures (Figure 4B and Supplemental Figure 4A). To validate the role of Fn14 in the proliferation of myogenic cells, WT and Fn14-KO cells were incubated in growth medium for 48 hours and pulsed with EdU for 1 hour, followed by detection of EdU+ cells (Figure 4C). There was a significant reduction in the proportion of EdU+ cells in Fn14-KO cultures compared with WT cultures (Figure 4D). In another experiment, myoblast cultures were also stained with anti-Pax7 and EdU (Supplemental Figure 4B) followed by quantification of Pax7+EdU+ cells. Results showed that the proportion of Pax7+EdU+ cells was significantly reduced in Fn14-KO cultures compared with WT cultures (Supplemental Figure 4C). Lactate dehydrogenase (LDH) is a stable enzyme present in all cells and released following plasma membrane damage and cell death (32). Our analysis showed that there was no significant difference in the levels of LDH in culture supernatants of WT and Fn14-KO myoblast cultures, suggesting that the deficiency of Fn14 does not affect the survival of myogenic cells (Supplemental Figure 4D).

Fn14 mediates the activation and proliferation of satellite cells. (A) Gene ontology (GO) biological processes associated with downregulated and upregulated genes. (B) Heatmap representing selected genes involved in the regulation of cell proliferation in cultured WT and Fn14-KO myogenic cells generated after analysis of RNA-seq dataset. (C) WT and Fn14-KO myogenic cultures were pulse labeled with EdU for 60 minutes. Representative images of the cultures after detection of EdU and Hoechst staining (nuclei detection). Scale bars: 50 μm. (D) Quantification of percentage of EdU+ cells in WT and Fn14-KO cultures. n = 3 biological replicates in each group. (E) TA muscle of Fn14fl/fl and Fn14scKO mice was injured by intramuscular injection of 1.2% BaCl2 solution. After 3 days, the mice were given an intraperitoneal injection of EdU and 11 days later TA muscles were collected and transverse muscle sections were generated and stained to detect EdU, anti-laminin, and nuclei. Representative photomicrographs after EdU, laminin, and DAPI staining are presented here. Scale bars: 50 μm. (F) Quantification of the EdU+ nuclei per unit area. n = 4 mice in each group. (G) Single myofibers were isolated from EDL muscle of Fn14fl/fl and Fn14scKO mice. After 48 hours of culturing, the myofibers were pulse-labeled with EdU for 60 minutes. Representative merged images of EdU+ and DAPI-stained myofibers are presented here. Scale bars: 100 μm. (H) Quantification of number of EdU+ cells per myofiber. n = 3 mice in each group. Analysis was done using 15–20 myofibers for each mouse. All data are presented as mean ± SEM. #P ≤ 0.05, values significantly different from WT myoblast, or corresponding muscle of Fn14fl/fl mice, analyzed by unpaired Student’s t test.

We next sought to investigate the role of Fn14 in the proliferation of satellite cells in response to muscle injury in vivo. For this experiment, TA muscle of 8-week-old male Fn14fl/fl and Fn14scKO mice was injured by intramuscular injection of 1.2% BaCl2 solution. The mice were given a single i.p. injection of EdU on day 3 after injury and the TA muscle was isolated on day 14 after injury and processed for detection of EdU, DAPI staining, and immunostaining for laminin protein (Figure 4E). Quantitative analysis showed that the number of EdU+ nuclei per unit area (mm2) was significantly reduced in TA muscle of Fn14scKO mice compared with Fn14fl/fl mice (Figure 4F). To further validate the role of Fn14 in satellite cell proliferation, we also established single myofiber cultures from extensor digitorum longus (EDL) muscle of male Fn14fl/fl and Fn14scKO mice and the myofiber-associated satellite cells were analyzed after 48 hours of culturing. Consistent with the in vitro and in vivo results, we found a significant reduction in the number of EdU+ cells per myofiber in cultures from Fn14scKO mice compared with Fn14fl/fl mice (Figure 4, G and H). Altogether, these results suggest that Fn14 is essential for the proliferation of muscle progenitor cells.

Fn14 is required for satellite cell self-renewal. In addition to generating sufficient numbers of myoblasts during skeletal muscle regeneration, activated satellite cells play a crucial role in maintaining their own reserve pool via self-renewal, which is essential for sustaining continuous muscle regeneration throughout the lifespan of the organism (33). To investigate the role of Fn14 in satellite cell self-renewal, we employed a re-injury model, involving 2 injuries separated by 3–4 weeks (30, 34, 35). Twenty-one days after the first injury, the TA muscle of Fn14fl/fl and Fn14scKO mice was re-injured using intramuscular injection of 1.2% BaCl2 solution. The TA muscle was isolated on day 5 after the second injury and analyzed by performing H&E staining, Sirius red staining, and immunostaining for Pax7 and laminin protein (Figure 5, A and D). There was a significant reduction in the proportion of myofibers containing 2 or more nuclei in TA muscle of Fn14scKO mice compared with Fn14fl/fl mice (Figure 5B). Additionally, there was an increase in connective tissue in Fn14scKO mice compared with Fn14fl/fl mice assessed by Sirius red staining (Figure 5C). Remarkably, the number of Pax7+ cells per unit area (mm2) was significantly reduced in TA muscle of Fn14scKO mice compared with Fn14fl/fl mice after double injury (Figure 5E).

Fn14 promotes satellite cell self-renewal. (A) TA muscle of Fn14fl/fl and Fn14scKO mice was injured by intramuscular injection of 1.2% BaCl2 solution. After 21 days, the same TA muscle was injured again, and the muscle was isolated 5 days later. Representative photomicrographs of H&E- and Sirius red–stained TA muscle sections. Scale bars: 50 μm. Quantification of (B) percentage of myofibers containing 2 or more centrally located nuclei and (C) percentage of total area stained with Sirius red. (D) Representative photomicrograph and (E) quantification of Pax7+ cells per unit area in double-injured TA muscle sections of Fn14fl/fl and Fn14scKO mice after immunostaining for Pax7 (red) and laminin (green) protein. Nuclei were identified by staining with DAPI. Scale bars: 50 μm. n = 3–4 mice in each group. (F) Heatmap of selected genes associated with stem cell population maintenance in WT and Fn14-KO cultures generated after analysis of RNA-seq dataset. (G) Immunoblots and (H) densitometry analysis showing levels of cleaved Notch1 and total Notch1 protein in WT and Fn14-KO cultures. n = 3 biological replicates in each group. (I) Relative mRNA levels of Notch receptors (Notch1, Notch2, and Notch3), Notch ligands (Jagged1, Jagged2, Dll1, and Dll4), and Notch targets (Hes1, Hes6, Heyl, and Hey1) in 5-day-injured TA muscle of Fn14fl/fl and Fn14scKO mice. n = 4 mice in each group. All data are presented as mean ± SEM. #P ≤ 0.05, values significantly different from corresponding muscle of Fn14fl/fl mice analyzed by unpaired Student’s t test.

The GO term analysis of DEGs revealed a significant downregulation in the gene expression of molecules involved in cell-cell signaling in Fn14-KO cultures (Figure 4A). Notch is an important cell-cell signaling pathway that is essential for satellite cell self-renewal and proliferation during regenerative myogenesis (12, 36–38). Indeed, our RNA-seq analysis showed that gene expression of several molecules associated with the maintenance of stem cell population, including components of Notch signaling, was downregulated in primary myoblast cultures of Fn14-KO mice compared with WT mice (Figure 5F). Moreover, levels of cleaved Notch1 protein were significantly reduced in Fn14-KO cultures compared with WT cultures, indicating an inhibition in Notch signaling in Fn14-KO cultures (Figure 5, G and H). Furthermore, qPCR analysis showed that there was a significant reduction in the mRNA levels of Notch receptors (Notch1 and Notch3), Notch ligands (Jagged2 and Dll1), and Notch target genes (Hes1, Hes6, and HeyL) in 5-day-injured TA muscle of Fn14scKO mice compared with the corresponding muscle of Fn14fl/fl mice (Figure 5I). These results suggest that targeted deletion of Fn14 inhibits self-renewal of satellite cells.

Fn14 prevents precocious differentiation of satellite cells. GO analysis of DEGs in the RNA-seq experiment demonstrated significant upregulation of pathways related to myoblast and myotube differentiation (Figure 4A). Further analysis of DEGs showed that mRNA levels of several molecules associated with muscle differentiation, including myogenin and Myf6, were significantly upregulated in Fn14-KO cultures compared with WT cultures (Figure 6A). Based on the expression of Pax7 and MyoD, the progeny of satellite cell can be classified as self-renewing (Pax7+MyoD–), activated/proliferating (Pax7+MyoD+), or differentiating (Pax7–MyoD+) (35, 39). To directly assess the role of Fn14 in myogenic differentiation, equal numbers of WT and Fn14-KO primary myoblasts were seeded in culture plates and incubated in growth medium for 24 hours followed by immunostaining for Pax7 and MyoD proteins. Results showed that the proportion of Pax7+MyoD+ cells was significantly decreased, whereas the proportion of Pax7–MyoD+ cells was significantly increased, in Fn14-KO cultures compared with WT cultures (Supplemental Figure 4, E–G). Furthermore, Western blot analysis showed a significant decrease in the protein levels of Pax7 and an increase in the levels of myogenin in Fn14-KO cultures compared with WT cultures (Figure 6, B and C).

Deletion of Fn14 leads to precocious differentiation of satellite cells. (A) Heatmap representing selected genes associated with muscle cell differentiation in WT and Fn14-KO myogenic cultures generated after analysis of RNA-seq dataset. (B) Immunoblots and (C) densitometry analysis of protein levels of Pax7, myogenin, Fn14, and unrelated protein GAPDH in WT and Fn14-KO cultures. n = 3 biological replicates in each group. (D) Representative individually stained and merged images of 0-hour and 72-hour cultured myofibers from Fn14fl/fl and Fn14scKO mice. Scale bars: 100 μm. Quantification of number of (E) Pax7+ cells, and (F) MyoD+ cells per myofiber at 0 hours. Quantification of (G) number of clusters per myofiber, percentage of (H) Pax7+MyoD– (self-renewing), (I) Pax7+MyoD+ (activated/proliferating), and (J) Pax7–MyoD+ (differentiating) cells per myofiber following 72 hours of culturing. n = 3 mice in each group. Analysis was done using 15–20 myofibers for each mouse at each time point. All data are presented as mean ± SEM. #P ≤ 0.05, values significantly different from WT myoblast, or corresponding muscle of Fn14fl/fl mice analyzed by unpaired Student’s t test.

We next established single myofiber cultures from the EDL muscle of male Fn14fl/fl and Fn14scKO mice and performed anti-Pax7 and anti-MyoD staining at 0 hours and 72 hours of culturing (Figure 6D). While there was no significant difference in the number of Pax7+ cells per myofiber, a significant increase in the number of MyoD+ cells per myofiber was observed in the freshly isolated myofibers of Fn14scKO mice compared with Fn14fl/fl mice (Figure 6, E and F). After 72 hours in suspension culture, the number of clusters per myofiber was significantly reduced in myofiber cultures prepared from Fn14scKO mice compared with Fn14fl/fl mice (Figure 6G). Importantly, there was a significant decrease in the proportion of Pax7+MyoD– (self-renewing) and Pax7+MyoD+ (activated) cells along with a significant increase in the proportion of Pax7–MyoD+ (differentiating) cells per myofiber in Fn14scKO cultures compared with Fn14fl/fl cultures (Figure 6, H–J). Collectively, these results suggest that the absence of Fn14 leads to premature differentiation of muscle progenitor cells.

Deletion of Fn14 leads to hyperactivation of STAT signaling. JAK/STAT signaling plays an important role in the regulation of satellite cell function and muscle regeneration (20, 21). Interestingly, GO term analysis of DEGs from the RNA-seq experiments showed a significant deregulation of the molecules involved in the regulation of tyrosine phosphorylation of STAT protein and JAK/STAT signaling in Fn14-KO cultures (Figure 4A and Figure 7A). Interestingly, Western blot analysis showed a dramatic increase in phosphorylation of STAT3 protein at S727 and Y705 residues without having any effect on total levels of STAT3 protein in 5-day-injured TA muscle of Fn14scKO mice compared with the corresponding muscle of Fn14fl/fl mice (Figure 7B and Supplemental Figure 5A). There was also a small but significant increase in the levels of p-STAT2 in 5-day-injured TA muscle of Fn14scKO mice compared with Fn14fl/fl mice. Interestingly, levels of total STAT2 and STAT5 proteins were also significantly higher in 5-day-injured TA muscle of Fn14scKO mice compared with Fn14fl/fl mice. In contrast, there was no difference in the levels of p-STAT5 or total STAT3 protein in injured TA muscle of Fn14fl/fl and Fn14scKO mice (Figure 7B and Supplemental Figure 5A). Similar to the in vivo results, we also found an increase in the levels of p-STAT3 protein in Fn14-KO cultures compared with WT cultures of myogenic cells (Figure 7C and Supplemental Figure 5B). We next sought to determine whether activated STAT3 has any role in the proliferation and differentiation of Fn14-KO myogenic cells. Primary myoblasts prepared from WT and Fn14-KO mice were transfected with control or STAT3 siRNA. After 24 hours of transfection, cellular proliferation was evaluated by pulse labeling of the cells with EdU for 60 minutes (Figure 7D). Results showed that the siRNA-mediated knockdown of STAT3 significantly increased the proportion of EdU+ cells in Fn14-KO cultures compared with cultures transfected with control siRNA (Figure 7E). By performing Western blotting, we also investigated the effect of silencing of STAT3 on the protein levels of Pax7 and myogenin in WT and Fn14-KO cultures. While there was no significant effect on the levels of Pax7 protein, knockdown of STAT3 significantly reduced the levels of myogenin protein in Fn14-KO cultures (Figure 7, F and G). Interestingly, knockdown of STAT3 increased levels of Fn14 protein, suggesting a feedback mechanism between Fn14 and STAT3 signaling (Figure 7, F and G). These results suggest that the upregulation of STAT3 signaling contributes to reduced proliferation and precocious differentiation of Fn14-deficient satellite cells.

Fn14 regulates STAT signaling in myogenic cells. (A) Heatmap of selected genes associated with positive regulation of JAK/STAT pathway, target genes, and inhibitors of JAK/STAT signaling in WT and Fn14-KO cultures generated after analysis of RNA-seq dataset. (B) Immunoblots showing protein levels of phosphorylated and total levels of STAT2, STAT3, and STAT5 protein in uninjured and 5-day-injured TA muscle of Fn14fl/fl and Fn14scKO mice. n = 3–4 mice in each group. (C) Immunoblots show protein levels of phosphorylated and total STAT3 and Fn14 protein in WT and Fn14-KO cultures. (D) WT and Fn14-KO myoblasts were transfected with control and STAT3 siRNA. After 24 hours, the cells were pulse labeled with EdU for 60 minutes and analyzed for EdU incorporation. Representative photomicrographs are presented here. Scale bars: 100 μm. (E) Quantification of percentage of EdU+Hoechst33258+ cells in WT and Fn14-KO cultures transfected with control or STAT3 siRNA. n = 3 biological replicates in each group. (F) Representative immunoblots and (G) densitometry analysis showing protein levels of Pax7, myogenin, STAT3, and Fn14 in WT and Fn14-KO myogenic cultures transfected with control or STAT3 siRNA. n = 6 biological replicates in each group. All data are presented as mean ± SEM. *P ≤ 0.05, values significantly different from corresponding cultures transfected with control siRNA. #P ≤ 0.05, values significantly different from corresponding WT cultures analyzed by 2-way ANOVA followed by Tukey’s multiple-comparison test.

Deletion of Fn14 in satellite cells attenuates regeneration and exacerbates myopathy in dystrophic muscle of mdx mice. Dystrophic muscle of mdx and D2.mdx mice (mouse models of DMD) undergo repetitive cycles of degeneration and regeneration, making them a valuable model for studying muscle repair in response to chronic injury (40, 41). We first analyzed a published scRNA-seq dataset (NCBI GEO GSE213925) (42) to determine how the gene expression of Fn14 is regulated in muscle stem cells of dystrophic muscle of mdx and D2.mdx mice. Interestingly, transcript levels of Fn14, TWEAK, and various components of canonical and noncanonical NF-κB signaling were reduced in satellite cells of mdx and D2.mdx mice compared with corresponding WT mice (Figure 8A). To understand the role of satellite cell Fn14 signaling in the progression of the dystrophic phenotype, we crossed Fn14scKO mice with mdx mice to generate littermate mdx;Fn14scKO and mdx;Fn14fl/fl mice. At the age of 3 weeks, male mdx;Fn14fl/fl and mdx;Fn14scKO mice were given daily i.p. injections of tamoxifen for 4 consecutive days to induce Cre-mediated recombination and were kept on a tamoxifen-containing chow for the entire duration of the study. The mice were analyzed at the age of 7 weeks. Deletion of Fn14 led to a significant reduction in the size and body weight of mdx;Fn14scKO mice compared with mdx;Fn14fl/fl mice (Figure 8, B and C). Moreover, wet weight of TA, gastrocnemius (GA), and quadriceps (QUAD) muscles normalized by body weight was significantly reduced in mdx;Fn14scKO mice compared with mdx;Fn14fl/fl mice (Figure 8D). Analysis of H&E-stained GA muscle sections showed signs of exacerbated myopathy marked by a significant decrease in the number of centronucleated fibers in mdx;Fn14scKO mice compared with mdx;Fn14fl/fl mice (Figure 8, E and F). Moreover, the average CSA of eMyHC+ myofibers and the number of Pax7+ cells per unit area were also significantly reduced in mdx;Fn14scKO mice compared with mdx;Fn14fl/fl mice (Figure 8, E, G, and H).

Satellite cell–specific ablation of Fn14 exacerbates dystrophic phenotype in mdx mice. (A) Heatmaps showing expression of Fn14 (gene: Tnfrsf12a), TWEAK (gene: Tnfsf12), Nfkb1, Nfkb2, Rela, Relb, and Nfkbia in muscle stem cells of WT and mdx, and WT and D2.mdx mice analyzed from a publicly available single-cell dataset (GSE213925). (B) Representative pictures of 7-week-old mdx;Fn14fl/fl and mdx;Fn14scKO mice. Quantitative analysis of (C) body weight, and (D) wet weight of tibialis anterior (TA), gastrocnemius (GA), and quadriceps (QUAD) muscle normalized by BW of corresponding mice. (E) Representative photomicrographs of GA muscle sections of 7-week-old mdx;Fn14fl/fl and mdx;Fn14scKO mice after H&E staining, anti-eMyHC staining, or anti-Pax7/anti-laminin/DAPI staining. White arrows point to satellite cells in muscle sections. Scale bars: 50 μm. Quantification of (F) number of centronucleated fibers (CNF) per field, (G) average CSA of eMyHC+ myofibers, and (H) average number of Pax7+ cells per unit area. n = 3–6 mice in each group. All data are presented as mean ± SEM. #P ≤ 0.05, values significantly different from mdx;Fn14fl/fl mice analyzed by unpaired Student’s t test.

To further evaluate the role of Fn14 in satellite cell function, we performed myoblast transplantation and studied their engraftment in skeletal muscle of mdx mice (43, 44). Cultured myoblasts prepared from WT mice were stably transduced with control or Fn14-overexpressing (Fn14OE) retrovirus. Western blot analysis confirmed overexpression of Fn14 in myoblasts transduced with Fn14OE retrovirus (Figure 9A). Next, TA muscle of mdx mice was injured by intramuscular injection of 1.2% BaCl2 solution. After 24 hours, the TA muscle was injected with control or Fn14OE myoblasts. After 4 weeks, the TA muscle was isolated and analyzed. While there was no significant difference in wet weight of TA muscle injected with control or Fn14OE myoblasts (Figure 9B), there was a significant increase in the average CSA of dystrophin+ myofibers in the TA muscle injected with Fn14OE myoblasts compared with that injected with control myoblasts (Figure 9, C and D). Collectively, these results suggest that targeted ablation of Fn14 exacerbates myopathy in mdx mice potentially due to impairment in muscle regeneration and overexpression of Fn14 in exogenous myoblasts enhances their engraftment in dystrophic muscle of mdx mice.

Overexpression of Fn14 improves engraftment of myoblasts into dystrophic muscle. (A) Western blot demonstrating levels of Fn14 protein in myoblasts transduced with EGFP (control) or Fn14-overexpressing (Fn14OE) retrovirus. (B) TA muscle of 8-week-old mdx mice was injured by intramuscular injection of 1.2% BaCl2 solution. After 24 hours, the muscle was injected with 1 × 106 primary myoblasts stably transduced with control or Fn14OE retrovirus. After 28 days, the TA muscle was isolated and weighed. (C) Representative anti-dystrophin–stained images of TA muscle sections of mdx mice transplanted with control or Fn14OE myoblasts. Scale bars: 50 μm. (D) Quantitative analysis of average CSA of dystrophin+ myofibers in TA muscle of mdx mice transplanted with control or Fn14OE myoblasts. n = 3–4 mice in each group. All data are presented as mean ± SEM. *P ≤ 0.05, values significantly different from TA muscle injected with control myoblasts by unpaired Student’s t test.