DUX4 promotes transcription of FRG2 by directly activating its promoter in facioscapulohumeral muscular dystrophy

Abstract

BackgroundThe most common form of facioscapulohumeral muscular dystrophy (FSHD) is caused by a genetic contraction of the polymorphic D4Z4 macrosatellite repeat array in the subtelomeric region of chromosome 4q. In some studies, genes centromeric to the D4Z4 repeat array have been reported to be over-expressed in FSHD, including FRG1 and FRG2, presumably due to decreased long-distance repression by the shorter array through a mechanism similar to position-effect variegation. Differential regulation of FRG1 in FSHD has never been unequivocally proven, however, FRG2 has been reproducibly shown to be induced in primary FSHD-derived muscle cells when differentiated in vitro. The molecular function of FRG2 and a possible contribution to FSHD pathology remain unclear. Recent evidence has identified the mis-expression of DUX4, located within the D4Z4 repeat unit, in skeletal muscle as the cause of FSHD. DUX4 is a double homeobox transcription factor that has been shown to be toxic when expressed in muscle cells.MethodsWe used a combination of expression analysis by qRT/PCR and RNA sequencing to determine the transcriptional activation of FRG2 and DUX4. We examined this in both differentiating control and FSHD derived muscle cell cultures or DUX4 transduced control cell lines. Next, we used ChIP-seq analysis and luciferase reporter assays to determine the potential DUX4 transactivation effect on the FRG2 promoter.ResultsWe show that DUX4 directly activates the expression of FRG2. Increased expression of FRG2 was observed following expression of DUX4 in myoblasts and fibroblasts derived from control individuals. Moreover, we identified DUX4 binding sites at the FRG2 promoter by chromatin immunoprecipitation followed by deep sequencing and confirmed the direct regulation of DUX4 on the FRG2 promoter by luciferase reporter assays. Activation of luciferase was dependent on both DUX4 expression and the presence of the DUX4 DNA binding motifs in the FRG2 promoter.ConclusionWe show that the FSHD-specific upregulation of FRG2 is a direct consequence of the activity of DUX4 protein rather than representing a regional de-repression secondary to fewer D4Z4 repeats.

DNA-binding sequence specificity of DUX4

“…It is actually two overlapping TAATs on different strands (i.e., TAATTA); therefore, we generated a luciferase reporter containing one copy of the full 30-bp sequence used in these studies. It has also recently been shown that FRG2 , a gene upregulated in FSHD, has motifs recognized by DUX4 [ 32 ]; we included the motif of greatest match to the consensus motif in our comparison: TAACCTAATTA. We also included a sequence from the MaLR repetitive element that is a near match to the consensus (TAATT G AATCA, note the middle G which deviates from the consensus) because a number of DUX4 targets have ChIP-seq peaks in nearby MaLR elements [ 11 , 33 ].…”

DUX4 Signalling in the Pathogenesis of Facioscapulohumeral Muscular Dystrophy

“…In another study, single-cell RNA sequencing of FSHD patient-derived primary myocytes revealed the presence of a non-DUX4-associated gene signature [118]. Genes located near the D4Z4 repeat array, such as FRG1 and FRG2, have also been inconsistently implicated in FSHD [119][120][121][122]. While both are direct transcriptional targets of DUX4 [120,121], the possibility exists that alterations in the chromatin state of the D4Z4 array could directly influence the transcription of these genes and affect FSHD pathogenesis.…”

Transgenic Drosophila for Investigating DUX4 and FRG1, Two Genes Associated with Facioscapulohumeral Muscular Dystrophy (FSHD)

“…Interestingly, two proposed alternative FSHD candidate genes, FRG1 (FSHD region gene 1) [ 32 ] and FRG2 (FSHD region gene 2) [ 33 ], localized proximal to the chromosome 4q35 D4Z4 array, were recently identified as direct DUX4-FL target genes [ 34 , 35 ], thus linking misexpression of these genes to the widely accepted DUX4 model of FSHD [ 19 ]. The molecular function of FRG2 remains unknown; no FRG2 protein has ever been reported and overexpression of the FRG2 mRNA in mice resulted in no apparent phenotype [ 36 ].…”