Both the double homeodomain DUX4 and DUX4c proteins interact with the multifunctional protein C1qBP in committed myogenic progenitors around the time of fusion

DUX4 and DUX4C (also named DUX4-Like 9) are genes encoding similar proteins that share identical homeodomains. Members of DUX4-Like gene family have been understudied as they are considered pseudogenes, as previously was the multicopy DUX4 gene, before its identification as the causal gene for facioscapulohumeral muscular dystrophy (FSHD) and the recent discovery of its physiological role in embryonic development. We showed that DUX4c, in contrast to DUX4, is normally expressed in skeletal muscle cells, and that DUX4c gain- or loss-of-function impacts their differentiation. In proliferating myoblasts, DUX4c is located in the nuclei. However, after the differentiation induction, DUX4c is also detected either around nuclei cluster (observed around the time of fusion) or in regions, near the plasma membrane, that look set to fuse. In these areas, DUX4c sporadically interacts with specific RNA-binding proteins (RBP) involved in muscle differentiation. The fusion event is crucial for muscle regeneration to form mature multinucleated myofibers. In FSHD primary myotubes, the nuclei clusters are significantly numerous and bigger than in healthy cells. Moreover, only an antisense agent specifically targeting DUX4c, not DUX4, is able to decrease their numbers. In the skeletal muscle tissue isolated from patients affected by FSHD, both DUX4c and DUX4 were immunodetected in abnormal shaped myofibers that also stained for regeneration markers. Some couples of muscle cells locally exhibited, near the sarcolemma, DUX4c-positive areas that were very close to each other but in distinct cells. MYOD1 or intense desmin staining co-detection in these cells supported an imminent muscle cell fusion. By co-immunoprecipitation coupled to mass spectrometry analyses, we found that the RBP Complement component 1 Q subcomponent-binding protein (C1qBP) is the major DUX4c, but also DUX4, interactor. This suggesting a DUX4/DUX4c competitive interaction for C1qBP in FSHD muscle cells. Using proximity ligation assay in FSHD muscle sections, we demonstrated DUX4c-C1qBP and the pathological DUX4-C1qBP interactions in activated muscle progenitor cells. The abnormal morphology of FSHD myofibers, along with our other findings, suggests that muscle cell fusion occurs more slowly than usual resulting in a delayed or incomplete muscle regeneration. In agreement, DUX4 overexpression impacts myoblast fusion in cell cultures. Furthermore, the membrane defects, described in FSHD muscle cells and in myofibers from DUX4 mouse models, might result from fusion anomalies. We propose that DUX4 abnormal re-expression in myogenic progenitors prevents them from committing to proper regeneration by competing with the physiological role of DUX4c. This hypothesis could explain why skeletal muscle is particularly sensitive to DUX4 toxicity.