Genotype - phenotype
Deletion mechanism
Chromatin structure
Animal models
Positions available
Student training
Selected reading
Useful links
Contact us


Last updated:

July 2001







Facioscapulohumeral muscular dystrophy



Clinical synopsis

Autosomal dominant facioscapulohumeral muscular dystrophy (FSHD, MIM *158900) is the third most common inherited neuromuscular disorder with an incidence of 1:20.000. The disease is characterized by a progressive weakness and atrophy of facial and shoulder girdle muscles and a gradual spread, variable in time, to abdominal and foot-extensor muscles followed by clinical involvement of upper arm and pelvic girdle muscles. Also, mild, often subclinical, sensorineural deafness and a retinal vasculopathy have been recognized as part of the clinical picture. More recently, attention has been drawn to the possible absence of facial weakness and pelvic girdle onset in some cases. The age at onset is generally in the second decade of life and the disease displays large inter- and intrafamilial variability: approximately 20% of gene carriers is asymptomatic while an equal percentage becomes wheelchair-bound.


Genetic mutation

FSHD is caused by partial deletion of a subtelomeric repeat array on chromosome 4q. This polymorphic array consists of 11>150 repeated units (D4Z4), each 3.3kb in size, and is located approximately 25-40kb proximal to the hexameric telomere repeat itself. Patients carry, due to the deletion of an integral number of repeated elements, arrays of 1-10 units . 

The subtelomere of chromosome 4q is duplicated to 10q where, as a consequence, a highly homologous polymorphic repeat array resides. In contrast to chromosome 4q, the homologous array on chromosome 10q may vary between 1 and >150 units without pathological consequences. Due to the plastic nature of subtelomeric domains, exchanged repeat units are frequently found on both chromosomes (approximately 20% of chromosomes in the Dutch population). Also, hybrid repeat arrays consisting of units derived from chromosomes 4 and 10 (4-type and 10-type, respectively) are readily identified. Despite this recombinogenic behavior, only small arrays on chromosome 4q are associated with FSHD.

The sequence homology between the distal ends of chromosomes 4q and 10q varies between 95-98% and extends 42kb proximal to the D4Z4 repeat array. Two consistent nucleotide changes within the D4Z4 unit and its homologue on chromosome 10 generate chromosome-specific restriction sites in both units allowing full analysis of the behavior of these arrays.


Mutational mechanism

In the last few years, detailed analysis of these repeat arrays in de novo and familial FSHD cases by pulsed field gel electrophoresis (PFGE) has contributed significantly to our knowledge on genotype-phenotype relationships and on the genesis of these deletions. An inverse relationship has been established between the residual length of the D4Z4 repeat and the age at onset and severity of the disease. Large FSHD alleles are usually identified in familial FSHD cases, while short FSHD alleles are more often associated with more severe, mostly sporadic cases. A survey of Dutch de novo FSHD kindreds demonstrated that the partial D4Z4 repeat array deletion occurs in 40% mitotically resulting in somatic mosaicism for the deletion. Somatic mosaicism was identified in clinically unaffected parents and de novo patients. In line with the previous finding that females are in general less severely affected than males, mosaic females are usually the clinically unaffected parent of a non-mosaic de novo patient, while mosaic males are more often affected by the disease. This survey also for the first time provided insight in the mutational mechanism by demonstrating that the presence of supernumerary 4-type repeat units on chromosome 10q predispose for a deletion on chromosome 4. Thus, partial D4Z4 deletions likely occur by interchromosomal repeat exchanges between identical repeat arrays.


Pathogenic mechanism

Assuming that disruption of a gene within the D4Z4 repeat array is the cause for FSHD, the D4Z4 unit itself has been scrutinized for expressed sequences over the past decade. It contains a putative open reading frame (DUX4) encoding a double homeobox domain, which is preceded by a strong potential promoter. These elements are also contained within the homologous unit on chromosome 10. Despite numerous efforts, transcription of this putative gene has never been demonstrated. Also, subtle nucleotide changes between individual D4Z4 repeat units resulting in variable open reading frames and the lack of a polyadenylation signal, which has never been observed for homeobox genes, challenge a transcriptional role of the potential DUX4 gene.

Perhaps the strongest argument against a direct transcriptional role of DUX4 in FSHD pathogenesis is the consistent linkage of the disease with chromosome 4. Short arrays on chromosome 10, even consisting (in part) of 4-type units and thus carrying DUX4 coding sequences are non-pathogenic.

Other sequences within the D4Z4 unit, its high GC content, and its spreading to mostly pericentromeric regions of the genome, has put forward the hypothesis that the D4Z4 repeat is a structural heterochromatic element within the subtelomere of chromosome 4q and that FSHD is caused by a position effect variegation mechanism in which partial deletion of the D4Z4 repeat array causes a transcriptional deregulation of the subtelomeric domain of chromosome 4q. This array element could prevent spreading of the heterochromatic sequence domain distal to the repeat into more gene-rich sequences proximal to the repeat. In this scenario, contraction of the repeat to sizes beneath a critical length may undermine this function thereby causing transcriptional deregulation of genes nearby. Consequently, the region proximal to the D4Z4 repeat array has been scrutinized for functional genes of which FRG1 and FRG2, both isolated in our lab, are closest by. Recently, we sequenced the distal 150kb of chromosome 10q likewise to identify similarities and discrepancies between both chromosome ends.


Perhaps in contrast to the expectations, and possibly in conflict with a ‘simple view’ in which partial reduction of the D4Z4 repeat array causes the deregulation of nearby genes, the distal end of chromosome 4q is devoid of genes and contains repetitive and dispersed sequences over megabases of sequence while already 100kb proximal to the homologous repeat array on chromosome 10, the sequence is chromosome-specific with a high density of genes. Thus, if the transcription regulation of some genes is sensitive to the length of the D4Z4 repeat array, repeat array reductions on chromosome 10q would be more likely associated with disease than chromosome 4, simply because it harbors far more potential target genes in its close vicinity. Nevertheless, genes identified in the distal 5-10Mb of chromosome 4q seem often to be transcriptionally upregulated in FSHD patients, whether primary or secondary to the deletion. For none of these genes however, a direct relationship with FSHD pathogenesis has been established.