Antisense Oligonucleotides Used to Target the DUX4 mRNA as Therapeutic Approaches in FaciosScapuloHumeral Muscular Dystrophy (FSHD)

Abstract

FacioScapuloHumeral muscular Dystrophy (FSHD) is one of the most prevalent hereditary myopathies and is generally characterized by progressive muscle atrophy affecting the face, scapular fixators; upper arms and distal lower legs. The FSHD locus maps to a macrosatellite D4Z4 repeat array on chromosome 4q35. Each D4Z4 unit contains a DUX4 gene; the most distal of which is flanked by a polyadenylation site on FSHD-permissive alleles, which allows for production of stable DUX4 mRNAs. In addition, an open chromatin structure is required for DUX4 gene transcription. FSHD thus results from a gain of function of the toxic DUX4 protein that normally is only expressed in germ line and stem cells. Therapeutic strategies are emerging that aim to decrease DUX4 expression or toxicity in FSHD muscle cells. We review here the heterogeneity of DUX4 mRNAs observed in muscle and stem cells; and the use of antisense oligonucleotides (AOs) targeting the DUX4 mRNA to interfere either with transcript cleavage/polyadenylation or intron splicing. We show in primary cultures that DUX4-targeted AOs suppress the atrophic FSHD myotube phenotype; but do not improve the disorganized FSHD myotube phenotype which could be caused by DUX4c over-expression. Thus; DUX4c might constitute another therapeutic target in FSHD.

Keywords: double homeobox; myopathy; polyadenylation; primary myoblasts; splicing interference.

Figure 1

DUX4 mRNA variants. Schematic representation of the D4Z4 repeat array (Genbank # AF117653), the last D4Z4 unit and the adjacent pLAM region (Genbank #U7746). The DUX4 ORF is contained in the first exon. Two polyadenylation signals (PAS) were reported, one in exon 3 and one in exon 7 [19,23]. All the mRNAs reported to date or identified in our group and not published yet are represented as detected in primary FSHD myoblasts/tubes [19,22], in immortalized myotubes from clones of a mosaic individual, in human Mesenchymal Stromal cells [70] and in germline and testis [23]. We observed two different transcription start sites (TSS) and different splice donor sites for intron II. These mRNAs derived exclusively from chromosome 4. The short mRNA (*) found in FSHD immortalized myotubes is discussed in the text.

Figure 2

D4Z4 repeat array on chromosome 4 and antisense oligonucleotide positions. Top: Each D4Z4 repeated unit contains a promoter and an open reading frame (ORF) for DUX4 in the first exon (E1) and a short non-coding exon 2 (E2). On a 4qA permissive allele, the last D4Z4 unit is extended by a pLAM region providing intron I, an untranslated exon 3 (E3) and a polyadenylation signal (PAS) allowing for DUX4 mRNA stabilization and subsequent translation in DUX4 protein. Healthy individuals present 11–100 D4Z4 units while patients with FSHD1 only have 1–10 units. Bottom: Positions of the antisense oligonucleotides (AOs) described in this review. They are all designed for splicing interference except for PAS that affects mRNA polyadenylation.

Figure 3

Treatment of FSHD myoblasts with antisense oligonucleotides against DUX4 prevents the formation of myotubes with the atrophic phenotype. (A) Primary healthy or FSHD myoblasts were transfected with either the negative control AO (nc-AO) or the indicated AOs targeting DUX4 mRNAs (AO positions shown in Figure 2), and differentiation was induced 4 h later. After 8 days, cells were observed under white light before processing to detect troponin T through immunofluorescence (green). The nuclei were stained with DAPI; (B) The efficacy of the AOs used in (A) and two additional ones was evaluated based on the proportion of atrophic myotubes (caused by DUX4 expression, and observed in the ncAO-treated cells) in the culture. Myotubes were counted from at least 10 random fields: those with a width <20 µm were considered “atrophic”. The percentage of atrophic to total myotubes is expressed as the mean ± SD; (C) DUX4 was detected by immunofluorescence (MAb 9A12; not shown). The number of DUX4-positive nuclei was counted in 10 fields in healthy control or FSHD myoblast cultures following treatment with the indicated AOs. The percentage of DUX4-positive nuclei among total nuclei (DAPI staining) was calculated and is reported in the graph. The significance of the differences between experiments with each individual AO compared to the ncAO, was evaluated using Student’s t-test. *** p-value < 0.001.

Figure 4

DUX4 inhibition prevents the formation of atrophic, but not disorganized FSHD myotubes. aFSHD and dFSHD primary myoblasts were transfected with either DUX4 siRNA or a non-targeted (nt) siRNA as indicated and switched to differentiation medium. Eight days later, the myotube morphology was highlighted through immunofluorescence staining of troponin T (green) and nuclei with DAPI (blue). Bottom panels: Magnified boxed regions. Stars indicate the accumulation of troponin T near clusters of nuclei. Scale bars: 50 µm.

Figure 5

Evaluation of DUX4-AO efficiency on endogenous DUX4 and target mRNA expression in FSHD primary myotubes. 10⁵ primary myoblasts (control, aFSHD and dFSHD) were seeded in 35 mm culture dishes. The next day, cells were transfected with either the negative control AO (nc-AO, 600 nM) or AOs pLAM2A(−7 + 18) and pLAM3A(−12 + 13) either alone (lanes 2A or 3A) or in a cocktail (lane 2A/3A) and pLAM1D(+7 − 18) (lane 1D) at previously determined optimal concentrations. Differentiation was induced 4 h after transfection and the cells were harvested 72 h later. (A) Top: Total RNAs of dFSHD myotubes were extracted and submitted to RT-PCR with primers we had previously shown to be specific of the DUX4 full length ORF [19]. The RT-PCR products were separated by electrophoresis on an agarose gel and stained with ethidium bromide. The controls were total RNAs of C2C12 cells transfected with the pGEM plasmid either without insert (negative control) or with a genomic fragment containing 2 D4Z4 units [18] (pGEM42). The experiment was done in the presence (RT+) or absence (RT-) of retrotranscriptase to demonstrate the products did not result from amplification of contaminating genomic DNA. Bottom: Total RNAs of aFSHD myotubes were extracted and submitted to reverse transcription (RT) and amplification by qPCR with DUX4-specific primers. The relative abundance was calculated using Ribosomal Protein Lateral Stalk Subunit P0 (RPLPO) as a reference for cDNA input and following M. Pfaffl’s guidelines [68,69]. The data are presented as fold change in DUX4 mRNA abundance with or without AO treatment; (B) Fold change in TRIM43 mRNA abundance after treatment with an AO or siRNA against DUX4 mRNA. The RT-qPCR was performed as described [67].

Figure 6

Evaluation of DUX4 mRNA expression in mouse tibialis anterior muscles co-injected with rAAV-D4Z4/pLAM virus and either pLAM3A or control vivo-PMO. The TAs of the mice were injected with 1E11 DRP of D4Z4/pLAM AAV virus and 10 µg of vivo-PMO per leg. The mice were sacrificed 10 days post-injection, and total RNAs were extracted with Trizol. Reverse transcription was performed on 800 ng of DNase-treated total RNA with the 3’ adaptor of the RLM-RACE kit (Ambion) and 2 μL of the resulting cDNA were amplified by nested PCR for 50 cycles in total. The RT-PCR products were analysed by electrophoresis on a 1.5% agarose gel and stained with ethidium bromide.