Overexpression of the double homeodomain protein DUX4c interferes with myofibrillogenesis and induces clustering of myonuclei

Abstract

Background: Facioscapulohumeral muscular dystrophy (FSHD) is associated with DNA hypomethylation at the 4q35 D4Z4 repeat array. Both the causal gene DUX4 and its homolog DUX4c are induced. DUX4c is immunodetected in every myonucleus of proliferative cells, while DUX4 is present in only 1/1000 of myonuclei where it initiates a gene deregulation cascade. FSHD primary myoblasts differentiate into either atrophic or disorganized myotubes. DUX4 expression induces atrophic myotubes and associated FSHD markers. Although DUX4 silencing normalizes the FSHD atrophic myotube phenotype, this is not the case for the disorganized phenotype. DUX4c overexpression increases the proliferation rate of human TE671 rhabdomyosarcoma cells and inhibits their differentiation, suggesting a normal role during muscle differentiation.

Methods: By gain- and loss-of-function experiments in primary human muscle cells, we studied the DUX4c impact on proliferation, differentiation, myotube morphology, and FSHD markers.

Results: In primary myoblasts, DUX4c overexpression increased the staining intensity of KI67 (a proliferation marker) in adjacent cells and delayed differentiation. In differentiating cells, DUX4c overexpression led to the expression of some FSHD markers including β-catenin and to the formation of disorganized myotubes presenting large clusters of nuclei and cytoskeletal defects. These were more severe when DUX4c was expressed before the cytoskeleton reorganized and myofibrils assembled. In addition, endogenous DUX4c was detected at a higher level in FSHD myotubes presenting abnormal clusters of nuclei and cytoskeletal disorganization. We found that the disorganized FSHD myotube phenotype could be rescued by silencing of DUX4c, not DUX4.

Conclusion: Excess DUX4c could disturb cytoskeletal organization and nuclear distribution in FSHD myotubes. We suggest that DUX4c up-regulation could contribute to DUX4 toxicity in the muscle fibers by favoring the clustering of myonuclei and therefore facilitating DUX4 diffusion among them. Defining DUX4c functions in the healthy skeletal muscle should help to design new targeted FSHD therapy by DUX4 or DUX4c inhibition without suppressing DUX4c normal function.

Keywords: Cytoskeleton; Differentiation; Disorganized myotubes; Myonuclear clustering; Proliferation; β-catenin.

Fig. 1

DUX4c overexpression induces misdistribution of troponin T and abnormal distribution of myonuclei. Healthy human primary myotubes were transfected with pCIneo-DUX4c or the empty pCIneo expression vector and switched to differentiation medium. a Five days later, troponin T was detected by immunofluorescence (green), and nuclei were labeled with DAPI (blue). Scale bar: 20 μM. b A 20 μg sample of total protein extracted from the same cells was separated via electrophoresis (PAGE-SDS), then transferred to a Western blot, immunodetected with the indicated primary antibodies and appropriate secondary antibodies coupled to HRP, and revealed through chemiluminescence

Fig. 2

DUX4c overexpression induces the formation of disorganized myotubes with large nuclear clusters. Healthy primary myoblasts were transfected with pAC1M2-DUX4c and differentiated 48 h later. After the first myotubes formed (day 3), DUX4c expression was induced by the addition of 100 ng or 1 μg of doxycycline to the culture medium. At day 8, myotubes were fixed, and DUX4c (red) and troponin T (green) were immunodetected. Nuclei were labeled with DAPI. A cluster of nuclei (> 5 nuclei) without endogenous DUX4c labeling is boxed (0 ng Dox). Clusters of nuclei presenting a ring-like structure are circled. Stars indicate cytoplasmic DUX4c partially co-localizing with troponin T. Scale bar: 30 μm

Fig. 3

DUX4c overexpression induces abnormal clusters of myonuclei and misdistribution of β-catenin. a FSHD and healthy primary myoblasts were transfected with the empty pCIneo or pCIneo-DUX4c expression vector, and differentiation was induced 24 h later. Immunofluorescence detection with antibodies against β-catenin (green) or DUX4c (red) was performed at day 3. Nuclei were labeled with DAPI. Scale bar: 20 μm. b Representative field from each culture (white light). Scale bar: 50 μm. c The nuclei per cluster were counted in each culture (10 fields per culture) and are presented as the mean ± SD. The proportion of nuclei present per cluster is shown in grayscale. Nuclear clusters with more than 16 nuclei were seldom found in healthy myotubes; the mean is therefore close to zero. ***p < 0.001

Fig. 4

dFSHD myotubes present abnormal clusters of nuclei overexpressing DUX4c as well as cytoplasmic DUX4c and α-tubulin disorganization. Healthy and FSHD primary myoblasts were differentiated, then fixed 6 days later, and DUX4c (red) and α-tubulin (green) were detected by immunofluorescence. A large cluster of myonuclei (> 16, indicated by *) can be observed, while such clusters were very seldom found in healthy cultures. Scale bars: 100 μm

Figure 5

DUX4c gain and loss of function affect cell proliferation. a Representative fields of healthy myoblasts transfected with the indicated pCIneo expression vectors, fixed 24 h later, and processed for immunofluorescence detection of DUX4c (red) and KI67 (green). Arrows point to KI67-positive cells. b The percentage of perinucleolar and nucleoplasmic KI67-positive nuclei on total nuclei was counted in myoblasts showed in a. and represented as the mean ± SD. Either 10 or 15 fields were counted for myoblasts transfected either with pCIneo or pCIneo-DUX4c, respectively (total number of nuclei: 685 and 433, respectively). In addition, the perinucleolar and nucleoplasmic KI67 intensity were measured in each field. The total perinucleolar or nucleoplasmic KI67 intensity (all fields) was divided by the total number of KI67-positive nuclei (either perinucleolar or nucleoplasmic, respectively), and represented as the mean ± SD. The number of nucleoplasmic KI67-positive nuclei was 35 in the healthy culture and 31 in the FSHD culture. For the significance, a Mann-Whitney test was applied. **p < 0.01 (c). Magnification of DUX4c-expressing cells with partial co-localization of nucleolar KI67. d The number of KI67-positive nuclei was counted in 10 fields of healthy and FSHD myoblasts. The percentage of KI67-positive nuclei is represented as the mean ± SD. ***p < 0.001 was considered significant. e Representative fields of FSHD myoblasts transfected with the indicated siRNAs, then fixed 24 h later, and processed for the immunofluorescence detection of KI67. Nuclei were stained with DAPI. Scale bar: 30 μm

Fig. 6

RNA interference against DUX4c suppresses the disorganized myotube phenotype. dFSHD primary myoblasts were transfected with the indicated siRNAs and differentiated. a At day 8, microscope pictures were obtained under white light. b The cells were then fixed, and troponin T (green) was detected by immunofluorescence. Nuclei were stained with DAPI. Scale bar: 30 μm. c The average (and SD) numbers of nuclei per cluster and clusters per field were counted (10 fields per culture). The proportion of nuclei per cluster is shown in grayscale. ***p < 0.001 was considered significant