Generation and initial characterization of skeletal muscle Dystrophin knockout mice
We performed CRISPR-targeting of the mouse Dmd exon 52 locus and inserted two loxP sequences in the intronic region immediately flanking the exon (Fig. 1A). Upon Cre-mediated excision of Dmd exon 52, the resulting DNA sequence contains an out-of-frame deletion that impairs the stability of the Dp427, Dp260, and Dp140 transcripts as has been previously shown in the mdx52 mouse. We mated female Dmd mice with male constitutive human skeletal actin (HSA)-Cre to ablate expression of the large Dp427m protein within the skeletal myofibers (Fig. 1B). Quantitative western blot assessments of Dp427m protein levels in the Dmd:HSA KO TA and soleus muscles showed an average of <5% remaining Dp427m protein in the TA and the soleus muscles (Fig. 1C). This was further confirmed using immunofluorescent antibody staining in these muscles showing severely reduced levels of the large Dp427m protein (Supplemental Fig. 1A). We compared our conditional dystrophin knockout mouse model to the mdx mouse that has well-characterized dystrophic histological, molecular, and physiological muscle symptoms. The resulting Dmd:HSA-Cre+ (Dmd:HSA KO) mice have no overt pathologies at birth, but similar to many mdx strains develop a progressive muscle degeneration starting between 4-6 months of age (Fig. 1D). Dmd:HSA KO diaphragm muscle revealed hallmark muscle degeneration, increased centralized myonuclei, and the presence of collagen deposits compared to Cre- controls (Fig. 1D). The dystrophin-negative myofibers were intact but showed pockets of degeneration and regeneration as evident via fibrotic and collagen-specific histochemical staining (Fig. 1D). Cross-sectional area (CSA) of Dmd:HSA KO TA myofibers showed an increased frequency of smaller myofibers compared to Cre- controls and similar to those found in our mdx mice (Fig. 1E). Analysis of both fast twitch Tibialis anterior (TA) and slow twitch (soleus) muscles from Dmd:HSA KO mice showed increased centralized myonuclei, large areas of fibrosis, and increased collagen production throughout the myofibers (Fig. 1F, G). Hydroxyproline content within the skeletal muscles of mdx mice is a well-described hallmark of fibrosis or excessive collagen in myofibers. Hydroxyproline levels were elevated in the Dmd:HSA KO TA muscles at similar levels to the mdx mice compared to both WT and Cre negative controls (Fig. 1H). Increased damaged myofibers in the Dmd:HSA KO mice were similar to that in mdx mouse TA muscles via immunofluorescent imaging (Supplemental Fig. 1B). These findings implicated significant skeletal muscle histopathological remodeling that occurs in Dmd:HSA KO mice similar to those observed in mdx strains.
We performed a muscle functional analysis of the Dmd:HSA KO adult mice by performing an assessment of muscle performance following TREAT-NMD standard operating procedures (SOPs) for mdx mice. Dmd:HSA KO mice showed a trending increased rotarod latency to fall, significantly reduced open field test (OFT) total distance travelled, and impaired forelimb grip strength (Fig. 2A-C). All of these locomotor and muscle strength metrics were comparable to those of aged-matched mdx mice. Dystrophin-deficient mice exhibit impaired treadmill running performance, partly due to the absence of sarcolemma-associated neuronal NOS (nNOS) binding to Dp427 at the spectrin-like repeats encoded by Dmd exons 45-48. We performed a standardized treadmill downhill running assessment of age-matched male WT, mdx, Dmd (Cre negative), and Dmd:HSA KO mice. The Dmd:HSA KO and mdx mice performed similarly in showing quicker times to exhaustion, reduced overall distance travelled, and reduced velocity compared to internal cohort control mice (Fig. 2D-F). These findings are consistent with the requirement of NOS-binding to skeletal muscle dystrophin as essential for skeletal muscle exercise and running performance. We assessed physiological force in EDL muscles from our Dmd:HSA KO mice to determine if there were functional deficits in the peak force and resistance to high mechanical strain. We observed no wet weight differences in the EDL muscles from the Dmd:HSA KO mice (Fig. 3A). EDL's from these mice attained similar peak force as EDL's from the Dmd:HSA Cre-controls (Fig. 3B). This reduction in force per unit physiological cross-sectional area of muscle (Fig. 3C) indicates a reduction in the quality of the Dmd:HSA KO EDL. Eccentric contractions reduced force of muscles from both groups of mice but this response was exacerbated in the Dmd:HSA KO mice (Fig. 3D, E). These results suggest that skeletal muscle dystrophin loss has a moderate reduction in myofiber force output and resistance to mechanical strain in comparison with that of mdx mice.
We evaluated the impact of dystrophin myofiber loss on the maintenance of muscle satellite cells (MuSCs) with respect to their numbers, activation, and ability to regenerate muscle following injury. Dystrophin expression within MuSCs is essential for regulating MuSC polarity, activation, and the MuSC niche. In mdx mice, MuSCs are impaired in muscle cell fate, cell death processes, Pax7+ expression, and regenerative capacity following injury. We investigated the impact of dystrophin myofiber ablation by evaluating the number of Pax7+ MuSCs in our Dmd:HSA KO mice compared to control mice. The uninjured adult Dmd:HSA KO mice had reduced total numbers of MuSCs when compared to Cre negative control mice (Supplemental Figs. 2A and 2B). These results implicate the role of myofiber dystrophin in the maintenance of MuSC viability. We performed a cardiotoxin (ctx)-induced injury to the TA muscles of the Dmd:HSA KO and control Dmd mice to evaluate the regenerative capacity of the skeletal muscle in the absence of the Dp427m protein. The Dmd:HSA KO mice showed significant fibrosis, collagen, and necrotic regions within the muscle that was comparable to the mdx strain (Supplemental Fig. 3A). Immunofluorescent analysis of the myofibers from the Dmd:HSA KO mice showed elevated levels of centralized myonuclei after 14 and 21 days post ctx-injury (Supplemental Fig. 3B). These results demonstrate the absolute requirement for the dystrophin myofiber expression is essential for the maintenance of normal numbers of muscle satellite cells. It remains unclear if dystrophin expression within MuSCs can fully compensate for dystrophin-loss within the myofiber; however our findings demonstrate that dystrophin myofiber expression is essential for maintaining MuSC homeostasis within the skeletal muscle.
We performed RNA-sequencing on TA muscles of the Dmd:HSA KO and Cre- control mice to identify transcriptomic changes and signaling pathways disrupted by the loss of skeletal muscle Dystrophin. This analysis revealed 1804 differentially expressed genes (DEGs; FDR < 0.05, |logFC | >= 1), including 640 downregulated and 1,164 upregulated transcripts (Fig. 4A). Gene Ontology (GO) analysis of the upregulated transcripts indicated that the most significantly enriched biological processes were inflammatory response (FDR = 8.32 × 10), innate immune response (FDR = 4.81 × 10) and collagen fibril organization (FDR = 3.54 × 10) (Supplemental Fig. 4A). Ingenuity Pathway Analysis further highlighted disruption of inflammatory cytokine transcripts and extracellular matrix (ECM) factors, including multiple collagen isoforms (Supplemental Fig. 4B). Among the DEGs, Nos2 (iNos) was downregulated, whereas key immune-responsive cytokines such as Tgfb1 were upregulated, reflecting disruption of key cell-cell signaling between skeletal myofibers and resident immune cells (Fig. 4B-D). Consistent with the transcriptomic changes, protein analysis of iNOS and TGFB1 reflected these alterations, confirming that the observed gene expression changes translate to the protein level (Fig. 4B-D). To contextualize these findings, we compared cytokine and ECM-related transcriptomic changes in our Dmd:HSA KO muscles with published data from the CRISPR-generated Dmd exon 52 global knockout (Dmd ΔEx52) mouse (Chemello et al., 2020), which also exhibited disruption of muscle fibers and transcriptional pathways. Both datasets revealed significant alterations in cytokine signaling and ECM transcripts (Fig. 4E, F). These transcriptomic analysis confirms and delineates an essential transcriptomic regulation of key signaling pathways that are known regulators of dystrophin pathophysiology.
We next sought to evaluate the role of skeletal muscle dystrophin in the regulation of adult skeletal muscle maintenance and regeneration. Subsequently, we mated our conditional Dmd mice to the HSA-MerCreMer tamoxifen-inducible skeletal myofiber Cre line to temporally ablate Dp427m expression in the adult mice (Fig. 5A). Two months after tamoxifen dosing, we observed 25-30% reduction of total Dp427m protein within the TA skeletal muscles of Dmd:HSA-MerCreMer+ ( + Tam) of Dmd:HSA-MerCreMer+ (-Tam) controls (Fig. 5B). We observed a similar 20-30% decrease of total Dp427m protein within the gastrocnemius and soleus muscles in the Dmd:HSA-MerCreMer+ ( + Tam) mice (Fig. 5B). These results are similar to the other reported conditional Dmd mice that showed partial reduction of Dp427mprotein and highlights the stability of the large dystrophin protein after 2 months of protein turnover. A histological examination of the Dmd:HSA-MerCreMer+ ( + Tam) mice showed minimal myofiber degradation and small pockets of regenerating myofibers compared to controls (Fig. 5C). Overall myofiber cross-sectional area (CSA) and minimal Feret diameter analysis revealed smaller clusters of myofibers in the Dmd:HSA-MerCreMer+ ( + Tam) mice compared to no tamoxifen mock controls (Fig. 5D, E). These findings implicate the significant stability of the large dystrophin Dp427m protein and implicate protein turnover as an essential part of dystrophin-restoration therapies.
Previous studies from our lab and others in various mdx strains demonstrated that the Dp427 cerebellar dystrophin isoform was essential for normal neurobehavior and social development. We mated the conditional Dmd mice to L7/Pcp2-Cre (Dmd:Pcp2 KO) mice to ablate the Dp427 protein in Purkinje neurons of the cerebellum (Fig. 6A). Isolated cerebellums from adult Dmd:Pcp2 KO mice showed loss of approximately 65% of cerebellar dystrophin protein compared to Cre- controls (Fig. 6B; Supplemental Fig. 5). We performed a series of neurobehavioral assessments in the Dmd:Pcp2 KO mice including social approach and social novelty between novel and familiar objects and mice (Fig. 6C). Dmd:Pcp2 KO showed impaired social novelty in object recognition compared to Dmd Cre- control mice (Fig. 6D, E). We performed a similar three-chamber assessment (novel mouse versus familiar mouse) and the Dmd:Pcp2 KO mice showed no preference for either mouse as compared to controls suggesting that social behavior defects are linked to cerebellar dystrophin protein expression. We performed both T- and Y-maze assessments in the Dmd:Pcp2 KO mice and found no differences in T-maze alternations compared to Cre- control mice (Fig. 6F). However, in the Y-maze assessment, Dmd:Pcp2 KO mice had reduced alternations without a change in total arm visits which together suggest reduced spatial working memory compared to Cre- controls (Fig. 6G). Memory tests were also conducted on the Dmd:Pcp2 KO mice that displayed increased time to find the platform in the Morris water maze (Fig. 6H). These findings support an essential role for cerebellar Purkinje cell expression of dystrophin and demonstrate a functional role for Dp427 in the regulation of neurobehavior in DMD.