Digenic inheritance of an SMCHD1 mutation and an FSHD-permissive D4Z4 allele causes facioscapulohumeral muscular dystrophy type 2

Abstract

Facioscapulohumeral dystrophy (FSHD) is characterized by chromatin relaxation of the D4Z4 macrosatellite array on chromosome 4 and expression of the D4Z4-encoded DUX4 gene in skeletal muscle. The more common form, autosomal dominant FSHD1, is caused by contraction of the D4Z4 array, whereas the genetic determinants and inheritance of D4Z4 array contraction-independent FSHD2 are unclear. Here, we show that mutations in SMCHD1 (encoding structural maintenance of chromosomes flexible hinge domain containing 1) on chromosome 18 reduce SMCHD1 protein levels and segregate with genome-wide D4Z4 CpG hypomethylation in human kindreds. FSHD2 occurs in individuals who inherited both the SMCHD1 mutation and a normal-sized D4Z4 array on a chromosome 4 haplotype permissive for DUX4 expression. Reducing SMCHD1 levels in skeletal muscle results in D4Z4 contraction-independent DUX4 expression. Our study identifies SMCHD1 as an epigenetic modifier of the D4Z4 metastable epiallele and as a causal genetic determinant of FSHD2 and possibly other human diseases subject to epigenetic regulation.

Fig. 1. D4Z4 methylation test and FSHD2 families

(a) FseI methylation values of 72 control, 93 FSHD1 and 53 FSHD2 gDNA samples. Error bar represents standard deviation. FSHD2 patients are significantly hypomethylated by this test compared to controls and FSHD1 patients (*: p<0.005). (b) Pedigrees of FSHD2 families. For each individual in the upper box their ID, their FseI methylation level (%) and whether they carry a SMCHD1 mutation (SMC: grey) or not (CTR: white), is indicated. Also indicated in the lower two boxes are the lengths of both D4Z4 arrays on chromosomes 4 in units (U). Permissive alleles, typically A alleles based on a polymorphism distal to the repeat are indicated in grey boxes. B alleles, which are non-permissive alleles are indicated in white boxes. Some less common subtypes of the A allele are considered to be non-permissive, these are marked with an # and colored white (Rf399 and Rf739). Note the independent segregation of D4Z4 hypomethylation and FSHD-permissive alleles. Only in those individuals in whom a permissive allele combines with D4Z4 hypomethylation (<25%) was FSHD diagnosed, while D4Z4 hypomethylated individuals carrying non-permissive alleles were unaffected by FSHD. Individuals selected for whole exome sequencing (upper 7 pedigrees) are indicated by asterisks. SMC# indicates coding synonymous SNP identified in Rf854. Color key is shown in the figure.

Fig. 2. FSHD2 families with SMCHD1 mutations

(a) Western blot analysis of fibroblast cultures of 6 controls (C) and 8 individuals carrying a SMCHD1 mutation (S). Sample identifiers refer to pedigrees in Fig. 1b and S6# denotes FSHD2 patient with only a synonymous coding SNP. (b) Bar diagram of ChIP analysis showing binding of SMCHD1 to D4Z4 but not to GAPDH (left panel) and reduced levels of SMCHD1 binding to D4Z4 (right panel) in FSHD2 patient 2305 from family Rf683 (Fig. 1b). Error bars represent +/− 1 standard deviation of duplicate experiments.

Fig. 3. SMCHD1 haploinsufficiency results in DUX4 expression in normal human myoblasts

(a) Short hairpin RNAs against different regions of SMCHD1 are effective in reducing the production of SMCHD1 in normal human primary myoblasts on RNA and protein levels. Numbers below the graph and above the gel lanes indicate the regions within the SMCHD1 transcript that are homologous to the indicated shRNA. SMCHD1 mRNA levels were quantified by qRT-PCR and l normalized to RNAse P transcripts in a multiplexed reaction. Normalized SMCHD1 levels are shown as a percentage of the levels found in the same cells treated with a vector expressing a scrambled sequence. Error bars show the standard deviation of the mean of three separate reactions. (b) Western blot of protein samples from the same cultures described in a normalized to tubulin. (c) Semi-quantitative RT-PCR analysis of DUX4 in cells deficient for SMCHD1. GAPDH was amplified to demonstrate RNA integrity. (d) Examples of DUX4 immuno-reactive nuclei observed in tubes where SMCHD1 levels were reduced using shRNA 4103 or 6051. Myotubes are shown with nuclei labeled blue with DAPI and DUX4 (red). GFP fluorescence produced from the lentivirus vector expressing the shRNAs is also shown. Depicted scale bars are 50 μm in length. (e) AON-mediated exon skipping of SMCHD1 exon 36 in normal human myoblasts 2333 and 2435. The mutation in family Rf1014 results in skipping of exon 36. Cells were treated with AONs designed to reproduce this skipping, and primers homologous to flanking exons (shown above each gel) were used to evaluate the proportion of exon-skipped transcripts. The 184-bp fragment is produced when exon 36 is skipped. DUX4 expression from the same cells is shown below each panel of SMCHD1 exon analysis. Asterisk marks low DUX4 expression levels consistent with inefficient SMCHD1 exon skipping levels. Results are also shown for myotube RNA of affected individuals from both families. An AON targeting exon 50 of the DMD gene was used as a negative control. (f) Similar as panel e, AON-mediated exon skipping of SMCHD1 exon 29 in normal human myoblasts 2333 and 2435. The mutation in family Rf649 results in skipping of exon 29, giving rise to the 124-bp fragment.