DUX4 expression in FSHD muscle cells: how could such a rare protein cause a myopathy?

Abstract

Facioscapulohumeral muscular dystrophy (FSHD) is one of the most frequent hereditary muscle disorders. It is linked to contractions of the D4Z4 repeat array in 4q35. We have characterized the double homeobox 4 (DUX4) gene in D4Z4 and its mRNA transcribed from the distal D4Z4 unit to a polyadenylation signal in the flanking pLAM region. It encodes a transcription factor expressed in FSHD but not healthy muscle cells which initiates a gene deregulation cascade causing differentiation defects, muscle atrophy and oxidative stress. PITX1 was the first identified DUX4 target and encodes a transcription factor involved in muscle atrophy. DUX4 was found expressed in only 1/1000 FSHD myoblasts. We have now shown it was induced upon differentiation and detected in about 1/200 myotube nuclei. The DUX4 and PITX1 proteins presented staining gradients in consecutive myonuclei which suggested a diffusion as known for other muscle nuclear proteins. Both protein half-lifes were regulated by the ubiquitin-proteasome pathway. In addition, we could immunodetect the DUX4 protein in FSHD muscle extracts. As a model, we propose the DUX4 gene is stochastically activated in a small number of FSHD myonuclei. The resulting mRNAs are translated in the cytoplasm around an activated nucleus and the DUX4 proteins diffuse to adjacent nuclei where they activate target genes such as PITX1. The PITX1 protein can further diffuse to additional myonuclei and expand the transcriptional deregulation cascade initiated by DUX4. Together the diffusion and the deregulation cascade would explain how a rare protein could cause the muscle defects observed in FSHD.

Fig. 1

Overexpressed DUX4 is detected in consecutive myonuclei of individual myotubes. Facioscapulohumeral muscular dystrophy (FSHD) myoblasts (dFSHD12) were transfected with the pCIneo-DUX4 expression vector at a low efficiency and differentiation was induced 5 hrs later. DUX4 (green) was detected by immunofluorescence with the MAb 9A12 monoclonal antibody 3 days after transfection. DUX4 was detected in myotubes containing either clusters of nuclei (A and B) or aligned nuclei (C and D). The nuclei were visualized by DAPI staining (blue). a, b, c and d correspond to enlarged fields from the left boxes.

Fig. 2

Endogenous DUX4 expression in facioscapulohumeral muscular dystrophy (FSHD) and control primary myotubes. (A) DUX4 (green) was detected by immunofluorescence with MAb 9A12 in the nuclei of disorganized (dFSHD12) and atrophic (aFSHD3) FSHD or control (CTL12) myoblasts 4 days after the induction of differentiation. DAPI (blue) was used to visualize nuclei. (B) Quantification of DUX4-positive nuclei in aFSHD3 and dFSHD12 myotubes compared with control (CTL12) myotubes 4 days after the induction of differentiation. The number of DUX4-positive nuclei was counted in 30 random fields of three independent experiments (10 fields per experiment). Two representative fields (1, 2) used for the quantification are shown in (A) for each cell line. The percentage was calculated relatively to the number of DAPI-positive nuclei, and the histogram represents the percentage means. The significance was evaluated by an anova test. **P < 0.01 was considered significant. (C) Quantification of DUX4-positive nuclei intensity in aFSHD3 and dFSHD12 myotubes compared with control (CTL12) myotubes 4 days after the induction of differentiation. The intensity of DUX4-positive nuclei was measured in 30 random fields of three independent experiments. Intensity values below the threshold (Th) are considered as null. Each value was plotted. Rectangles represent the intensity means. The significance was evaluated by an anova test. **P < 0.01 was considered significant.

Fig. 3

DUX4 is expressed in facioscapulohumeral muscular dystrophy (FSHD) myoblasts and in consecutive nuclei in FSHD myotubes. Co-immunofluorescence with MAb 9A12 (green) and a rabbit serum directed against desmin (red) on FSHD (dFSHD12) and control (CTL10) primary myotubes, 5 days after the induction of differentiation. a, b and c correspond to enlarged fields from the left boxes. Arrows indicate the most stained nuclei and the dotted arrows the intensity gradient of the DUX4 staining (D: merge panel). DAPI (blue) was used to visualize nuclei.

Fig. 4

DUX4 and PITX1 detections in facioscapulohumeral muscular dystrophy (FSHD) myotubes. DUX4 (green) and PITX1 (red) were detected by immunofluorescence with MAb 9A12 or the rabbit anti-PITX1 serum in nuclei of primary myotubes 4 days after the induction of differentiation. Colocalization of DUX4 and PITX1 staining appears yellow (Merge). DAPI (blue) was used to visualize nuclei. dFSHD13 and aFSHD1 are derived from an affected muscle (*) and dFSHD12 and aFSHD3 from a non-affected muscle. Arrows indicate positive nuclei for DUX4 or PITX1 staining, and dotted arrows indicate PITX1 or DUX4-negative nuclei.

Fig. 5

DUX4 and PITX1 stabilization in facioscapulohumeral muscular dystrophy (FSHD) myotubes treated with MG132. (A) PITX1 (red) was detected by immunofluorescence with the rabbit anti-PITX1 serum in nuclei of myotubes 4 days after the induction of differentiation (a', b'). Phase contrast microscopy was used to visualize the myotube morphology and position of the nuclei (a, b). b and b' correspond to primary myotubes treated with 25 μM MG132, a proteasome inhibitor, for the last 5 hrs in culture before fixation. Arrows indicate the most stained nuclei and the dotted arrows the intensity gradient of PITX1 staining. (B) DUX4 (green) and PITX1 (red) were detected by immunofluorescence with MAb 9A12 or the rabbit anti-PITX1 serum, respectively, in nuclei of FSHD (aFSHD1 and dFSHD12) primary myotubes 4 days after the induction of differentiation. The myotubes were treated with 25 μM MG132 as in (A). Colocalization of DUX4 and PITX1 staining appears yellow (Merge). DAPI (blue) was used to localize nuclei. Arrows indicate positive nuclei for DUX4 and PITX1 staining and dotted arrows indicate negative nuclei for DUX4 and PITX1 staining. (C) Nuclear proteins extracted from aFSHD3 primary myoblast were analysed by 12% PAGE-SDS followed by Western blotting and immunodetection with MAb 9A12 as described in Materials and Methods. Myotubes were treated with 0 or 25 μM MG132 as indicated 5 hrs before harvest. Nuclear extracts were prepared using the NE-PER kit at the proliferation state (pro) or 4 days after the induction of differentiation (diff 4). Total extracts of TE671 cells transfected with pCIneo-DUX4 were used as a positive control (C⁺). DUX4 proteolysis products observed in the absence of MG132 are shown by red braces. Ponceau red staining of the membrane was used as loading control.

Fig. 6

DUX4 expression in facioscapulohumeral muscular dystrophy (FSHD) muscle biopsies. (A) Nuclear (Nuc), cytoplasmic (Cyt) or total (Tot) protein extracts from FSHD (F11, F7, F10) or control (C1) muscle biopsies were analysed by 12% PAGE-SDS followed by Western blotting and immunodetection with MAb 9A12. The positive control (C+) is an extract of C2C12 cells transfected with pCIneo-DUX4. Ponceau red staining of the membrane was used for loading control. (B) Total extract of the F6 and F15 muscle biopsies was analysed by 2D electrophoresis (IEF-PAGE-SDS, see Materials and Methods), followed by Western blotting and an immunodetection with MAb 9A12. The characteristics of these samples are reported in Table S2. (*) FSHD biopsies derived from an affected muscle.

Fig. 7

Dynamic model of propagation and initiation of a transcriptional cascade. I: Activation and diffusion. A myotube is a multinucleated cell with a common cytoplasm in which individual nuclei can independently activate gene expression. In an facioscapulohumeral muscular dystrophy (FSHD) myotube, the DUX4 gene is activated in one given nucleus ➀. The DUX4 gene is then transcribed into an mRNA that terminates at the polyadenylation site located in the pLAM region. The mRNA is translocated into the cytoplasm domain close to the activated nucleus and it is translated, yielding several molecules of DUX4 protein ➁. The DUX4 protein that carries a nuclear localization signal (NLS) could diffuse in the cytoplasm, and be transported into several neighbouring nuclei ➂. II: Cascade initiation and amplification. In each nucleus that has imported the DUX4 protein, this transcription factor directly activates a number of genes as shown here for the PITX1 gene ➊. The PITX1 gene is thus transcribed, its mRNA is translocated into the cytoplasm domain close to the activated nuclei and translated ➋. The molecules of PITX1 protein can diffuse in the cytoplasm and, because they also carry a NLS, they will be imported into more neighbouring nuclei ➌.The transcriptional cascade initiated by DUX4 can further extend because PITX1 is also a transcription factor and targets additional genes such as TP53. At each step of this transcription cascade, the number of activated nuclei and expressed genes increases, causing an amplification of the initial trigger, i.e. DUX4 gene activation in a single nucleus. Globally, the DUX4 transcription cascade leads to muscle atrophy, inflammation, oxidative stress and decreased differentiation potential, the key features of FSHD.