Facioscapulohumeral Muscular Dystrophy 1

A number sign (#) is used with this entry because facioscapulohumeral muscular dystrophy-1 (FSHD1) is associated with contraction of the D4Z4 macrosatellite repeat (see 606009) in the subtelomeric region of chromosome 4q35.

In unaffected individuals, the D4Z4 array consists of 11 to 150 repeat units (corresponding to EcoRI fragments of 41 to 350 kb), whereas FSHD patients have contraction of the repeat units from 1 to 10 (corresponding to EcoRI fragments of 10 to 35 kb). Borderline fragment sizes (35 to 40 kb) must be interpreted with caution (summary by Mostacciuolo et al., 2009).

The genetics of FSHD is, however, complex, and detection of a D4Z4-reduced allele may not be sufficient for diagnosis (see DIAGNOSIS below and Scionti et al., 2012).

Description

Facioscapulohumeral muscular dystrophy is the third most common hereditary disease of muscle after Duchenne (DMD; 310200) and myotonic (160900) dystrophy. It is a highly variable disorder with weakness appearing from infancy to late life but typically in the second decade. In general, the disease initially involves the face and the scapulae followed by the foot dorsiflexors and the hip girdles. Typical features are striking asymmetry of muscle involvement from side to side and sparing of bulbar extraocular and respiratory muscles (Tawil et al., 1998).

Richards et al. (2012) provided a detailed review of FSHD.

See also FSHD2 (158901), which is phenotypically indistinguishable from FSHD1 but not associated with contraction of the D4Z4 microsatellite repeat. Evidence suggests, however, that epigenetic changes in this region are associated with both FSHD1 and FSHD2 (Zeng et al., 2009).

Clinical Features

Justin-Besancon et al. (1964) added 3 affected generations to the 4 described by Landouzy and Dejerine (1885) and gave autopsy findings in 1 of the original patients who died at age 86 years. Some cases show congenital absence of part or all of certain muscles such as a pectoral muscle. The relationship of the congenital defect of muscle to the dystrophy is unclear. Tyler and Stephens (1950) and Tyler (1953) reported 17 families. In 1 kindred 150 members were affected over 6 generations. A girl, whose face alone was affected at age 9 when examined by Landouzy and Dejerine (1885), did not develop weakness of the arms until age 60 and of the legs until age 70, and survived to age 85 years. In her family, affected members were distributed through 8 generations.

Meyerson et al. (1984) reported sensorineural hearing loss in 2 sibs with FSH muscular dystrophy which affected other members of the family in typical manner. The studies of Brouwer et al. (1991) suggest that a 'change of hearing function is part of the disease and may lead to severe hearing loss in some patients.' Deafness and abnormalities of retinal vessels (see later) are probably integral parts of the disorder.

Sayli et al. (1984) found at least 53 affected persons in a Turkish kindred originating in the village of Cullar. Initial signs and symptoms seemed to appear early in infancy in many. The disorder progressed slowly without interfering significantly with survival and reproduction. Symptoms first involved the face, upper arms, and shoulder muscles. Creatine kinase levels were 1.5 to 2 times normal. Many of the affected persons were identified on examination; only 13 reported complaints and their mean age was 40.1 years. After the kindred reported by Tyler and Stephens (1950), this is the most extensively affected family studied to date. Awerbuch et al. (1990) found that the Beevor sign was present in 27 of 30 patients with FSHD but absent in all 40 patients with other neuromuscular disorders. They concluded that it is a common finding in FSHD patients even before functional weakness of abdominal wall muscles is apparent. Because of weakness of the lower rectus abdominis muscles, the umbilicus moves upward when the subject in the supine position raises his or her head, producing the Beevor sign. (This sign was originally proposed by English neurologist Charles E. Beevor as an indication of the level of involvement in spinal cord lesions.) Bodensteiner and Schochet (1986) suggested the supraspinatus muscle as the site of choice for biopsy in this disorder. Reardon et al. (1991) pointed out the difficulties in some cases in distinguishing FSHD from the Becker type of muscular dystrophy (300376). Calf hypertrophy, although rare, has been reported in FSHD. Jardine et al. (1993) reexamined the 3 affected members in the family reported by Reardon et al. (1991) and demonstrated that they had a rearrangement in the region of the FSHD gene as indicated by studies with probe p13E-11.

Shen and Madsen (1991) described symptomatic atrial tachycardia, for which an antitachycardic pacing device was implanted, in a 44-year-old woman. Her severe scapular and shoulder weakness led to recurrent dislodgment of the atrial pacemaker lead.

According to Bailey et al. (1986), infantile facioscapulohumeral muscular dystrophy was recognized by Duchenne (1862) and differentiated from pseudohypertrophic muscular dystrophy over 20 years before the description of the usual form of facioscapulohumeral muscular dystrophy by Landouzy and Dejerine (1885). Whereas FSHD is generally a benign, slowly progressive myopathy that begins in late childhood or adolescence and leads to disability only late in its course, occasional families contain individuals with a severe infantile form of the disorder who have 1 asymptomatic or minimally affected parent. Bailey et al. (1986) reported a family in which the severe infantile presentation predominated. They suggested 'that the gene coding for this disorder may be different from that responsible for conventional facioscapulohumeral muscular dystrophy.' The genes for the 2 forms of the disease may, of course, be allelic.

In 56 of 75 persons with clinical or genetic evidence of FSH muscular dystrophy, Fitzsimons et al. (1987) found peripheral retinal capillary abnormalities including telangiectasia, closure, leakage, and microaneurysm formation. They were prompted to do this study by the occasional reports of exudative retinal detachment and deafness with this disorder. The study included 1 FSH family in which the proband was treated for exudative retinopathy and 13 other members had retinal telangiectasia. There were 8 cases, including 3 parents of apparently 'sporadic' FSH cases, in which fluorescein angiography 'confirmed the abnormal genotype, even though clinical examination of skeletal muscle revealed no clear abnormality.' Fitzsimons et al. (1987) concluded that retinal capillary abnormalities are an integral part of FSH muscular dystrophy and raised the question as to whether analogous capillary abnormalities may be implicated in the pathogenesis of the muscle disease. Padberg et al. (1992) came to similar conclusions on the basis of studies using fluorescein retinal angiography in 32 patients from 19 families. In 16 of the 32, representing 11 families, retinal capillary changes were found consisting of telangiectasia, microaneurysms, vessel occlusions, and small exudates and hemorrhages in the macular as well as in the peripheral retina. In addition, tone audiometry was performed, with the finding that 20 sibs from 14 families had some degree of high-tone deafness. Similar findings were observed in 8 sporadic cases: 4 had retinal vasculopathy and 5 had high-tone deafness. Combined with linkage data, these observations demonstrated that retinal vasculopathy and high-tone sensorineural deafness are part of the clinical picture of FSHD and are no grounds for assuming genetic heterogeneity.

Small (1968) described 4 sibs with facioscapulohumeral dystrophy and bilateral retinal exudative telangiectasia, labeled Coats disease (see 300216). Neurosensory deafness and mental retardation were present in all 4. Taylor et al. (1982) and Voit et al. (1986) described the same association. Yasukohchi et al. (1988) described a brother and sister with facioscapulohumeral dystrophy. The brother, aged 13 years, also had sensorineural hearing loss, marked tortuosity of retinal arterioles, early onset and progression of severe restrictive pulmonary dysfunction, and cor pulmonale. The 8-year-old sister had only muscle manifestations. Bilateral sensorineural hearing loss in the high frequency range was described in the above patients. In some, the hearing loss was clearly progressive and with time tended to involve lower frequencies (Voit et al., 1986). As noted from the cases cited, autosomal dominant inheritance was not always clear; recessive inheritance was possible and might point to this being an entity separate from FSHD. Brouwer et al. (1991) performed screening audiometry in 56 patients with autosomal dominant FSHD and in 72 healthy family members, and found that the difference in hearing levels between 4,000 Hz and 6,000 Hz was significantly greater in FSHD patients than in their unaffected relatives. This led them to conclude that a change in hearing function is part of the disease and may lead to severe hearing loss in some patients. Gieron et al. (1985) described a mother and 3 children with FSHD, sensorineural hearing loss, and marked tortuosity of retinal vessels. The deafness, which varied from mild to moderate, was bilateral and early in onset. Audiologic studies indicated the cochlea as the site of the abnormality. Matsuzaka et al. (1986) reported a sporadic case of this constellation of manifestations plus mental retardation and suggested that these cases constitute a nosologically specific form of FSHD.

Shields et al. (2007) noted that retinal telangiectasia compatible with Coats disease (300216) can be an extramuscular manifestation of FSHD but that most affected patients have asymptomatic retinal telangiectasia found at ocular screening after diagnosis of FSHD. They described a young child who had advanced eye findings of unilateral neovascular glaucoma from bilateral retinal telangiectasia 3 years before FSHD became apparent.

Miura et al. (1998) reported 2 sporadic cases of early-onset scapulohumeral muscular dystrophy with mental retardation and epilepsy in unrelated, severely affected females. In both cases, Southern blot analysis of the EcoRI-digested genomic DNA, using 2 probes, detected 10-kb EcoRI fragments, the shortest reported to that time. Patient 1 showed infantile spasms at the age of 4 months and localization-related epilepsy at the age of 2.5 years. Muscular atrophy in the face, shoulder girdle, and upper arms was observed from the age of 4 years. In patient 2, lack of facial expression was noticed since the age of 1 year, and at 4 years she was noted to have loss of upward gaze bilaterally. She developed localization-related epilepsy at the age of 9 years. From the age of 10 years, weakness of the lower limbs progressed and she became wheelchair-bound at the age of 14 years. She had moderate sensorineural hearing loss, a loss of upward gaze bilaterally, and tongue atrophy. Their IQs were 33 and 45, respectively. Miura et al. (1998) suggested that mental retardation and epilepsy may be part of the clinical spectrum of FSHD, especially in very early-onset patients with large deletions.

Among 151 Japanese patients with FSHD, Yamanaka et al. (2001) reported 7 patients with tongue atrophy with abnormal tongue MRI findings (disorganized architecture) and typical myogenic patterns of electromyography. All patients were classified as having early-onset FSHD with large gene deletions within the 4q35 gene region.

Krasnianski et al. (2003) described atypical features in 6 of 41 patients with FSHD and the 4q35 deletion. Three patients from 1 family showed the typical phenotype with the additional feature of chronic progressive external ophthalmoplegia. Three patients, 2 from 1 family, showed sparing of the facial muscles, and 2 of these patients had severe, diffuse myalgia. There was no correlation between the atypical features and the DNA fragment size due to the deletion.

Wohlgemuth et al. (2006) found that 10 of 87 individuals with FSHD had signs of weakness in the jaw, lips, or tongue. Oropharyngeal evaluation in 8 of these patients detected mild to moderate swallowing abnormalities in 7 patients and tongue atrophy in 6.

See 182970 for a form of spinal muscular atrophy simulating FSH muscular dystrophy.

Pathologic Features

In a 20-year-old woman who had inherited FSHD from her father and who also had an affected brother, Slipetz et al. (1991) found that muscle biopsy showed fiber atrophy with patchy staining for oxidative enzymes; that electron microscopy of liver showed many enlarged mitochondria with paracrystalline inclusions; and that skin fibroblasts showed decreased oxidation of the respiratory substrates alanine and succinate, suggesting deficiency of complex III of the electron-transport chain. Cytochrome c oxidase activity (complex IV) was normal. Biochemical analysis of liver supported the fibroblast data, since succinate oxidase activity (electron-transport activity through complexes II-IV) was reduced and complex IV activity was normal. Cytochrome b, a component of complex III, was undetectable in liver, although typical peaks were found for other cytochromes. Southern blot analysis of fibroblast mtDNA showed no major deletions or rearrangements.

Using confocal microscopy, Reed et al. (2006) found that some of the structures at the sarcolemma in FSHD skeletal muscle biopsies were misaligned with respect to the underlying contractile apparatus. Electron microscopy showed a significant increase in the distance between the sarcolemma and the nearest myofibrils, from less than 100 nm in controls to 550 nm in FSHD. Reed et al. (2006) concluded that the pathophysiology of FSHD includes novel changes in the organization of the sarcolemma and the subsarcolemmal membrane cytoskeleton.

Mapping

Padberg et al. (1984) found possible linkage to Gm (147100) (maximum lod score = 1.428 at theta = 0.2). Padberg et al. (1988) followed up on this observation by testing linkage with D14S1 (107750). They excluded linkage, thus suggesting that the FSH muscular dystrophy locus is not situated on the distal part of the long arm of chromosome 14. Sarfarazi et al. (1989) presented exclusion data, resulting from an international collaboration, showing that more than 85% of the human genome has been excluded as a likely location for the FSHD gene. Chromosomes 3, 5, 10, 11, 15, and 19 remained largely unexcluded. Lucotte et al. (1989) excluded 3 chromosomal regions, including proximal 19q. Lunt et al. (1989) presented linkage data for 22 different DNA markers in 24 families with FSHD. They achieved a maximum lod score of 1.87 at a theta of 0.15 with anonymous marker D18S3, which has been localized to 18p11.3. In a clinically homogeneous group of families, Jacobsen et al. (1990) excluded chromosomes 1, 2, 5, 7, 10, and 16 as the site of the FMD mutation. They estimated that 23% of the human genome was excluded.

In a study of 10 Dutch families, Wijmenga et al. (1990) found linkage of FSH muscular dystrophy to a microsatellite marker Mfd22 (D4S171) on chromosome 4; maximum lod score = 6.34 at theta = 0.13. Only 1 family was uninformative for this marker. No evidence of heterogeneity was found. From the map location of the marker, it appeared that the FSHD gene was located near the distal end of 4q. Siddique et al. (1989) studied 2 multigeneration families with the neurogenic form of facioscapulohumeral disease considered to be a form of spinal muscular atrophy. Possible linkage to MNS on 4q was found.

In a linkage study of FSHD families, Upadhyaya et al. (1991) found 3 recombinants in a total of 140 meioses for linkage with a hypervariable DNA probe on 4q (pH30, locus D4S139), giving a maximum lod score of 36.77 at a recombination fraction of 0.02. Wijmenga et al. (1991) found a VNTR locus, D4S139, to be much more closely linked to FSHD than to D4S171. This marker was mapped to 4q35-qter by in situ hybridization. One small family yielded a negative lod score for D4S139, but otherwise there was no evidence of genetic heterogeneity. In 9 informative families, they found a maximum lod score of 17.28 at theta = 0.027.

Mathews et al. (1991) defined the location of FSHD as 4q35. They used a probe derived from the breakpoint of an X;4 translocation with a breakpoint at 4q35. The patient had facial weakness at age 4 years. A peak lod score of 8.23 at theta = 0.00 was found with a 4q35 probe, D4S139. Reporting for a consortium, Sarfarazi et al. (1992) defined the relationship of the FSHD locus to markers in the 4q35 region in 65 families containing a total of 504 affected persons. No evidence of heterogeneity was found in the initial study; however, after completion of the analysis, 1 large family that might show heterogeneity was identified. In view of this and the fact that all of the linked markers were on the centromeric side of the FSHD locus, Sarfarazi et al. (1992) recommended that these markers not yet be used in clinical applications. Upadhyaya et al. (1992) found that FSHD was situated telomeric to all 4 DNA markers that they used in a series of 23 families. However, Wijmenga et al. (1992) found at least one recombination event which suggested that 1 of the 4 markers, D4S139, might lie distal to FSHD. In a study of 4 families with FSHD, Mathews et al. (1992) found close linkage to three 4q35 probes, D4S163, D4S139, and D4S171. Two of their families included males with a rapidly progressive muscle disease that had been diagnosed, on the basis of clinical features, as Duchenne muscular dystrophy. One of these males was available for linkage study and shared the haplotype of his FSHD-affected cousin and aunt.

Mills et al. (1992) generated a fine-structured genetic map of the distal long arm of chromosome 4 by a combination of classical RFLPs and PCR-based polymorphisms including CA repeats and single-strand conformation polymorphisms (SSCP). Gilbert et al. (1992) and Weiffenbach et al. (1992) demonstrated that both D4S139 and D4S163 are closely linked to FSHD, with D4S139 being the closest proximal marker to FSHD. No telomeric marker to FSHD had been demonstrated.

Gilbert et al. (1993) found evidence for heterogeneity in FSHD. In linkage studies, 5 of 7 families gave a posterior probability of more than 95% of being of the linked type, while 2 families appeared unlinked to that region of distal 4q. Affected members of the 2 unlinked families met the clinical criteria for the diagnosis of FSHD, including facial weakness, clavicular flattening, scapula winging, proximal muscle weakness, and myopathic changes on muscle biopsy without inflammatory or mitochondrial pathology. See 158901.

Molecular Genetics

All patients with a confirmed diagnosis of FSHD and for whom detailed molecular studies have been performed carry a chromosomal rearrangement within the subtelomeric region of 4q (4q35). This subtelomeric region is composed mainly of a polymorphic repeat structure consisting of 3.3-kb repeated elements, designated D4Z4 (see 606009). The number of repeat units varies from 10 to more than 100 in the population, and, in FSHD patients, an allele of 1 to 10 residual units is observed because of the deletion of an integral number of these units (Wijmenga et al., 1992; Hewitt et al., 1994).

D4Z4 Macrosatellite Repeat

For a full discussion of the D4Z4 macrosatellite repeat, see 606009.

Lemmers et al. (2004) summarized the relationship of the D4Z4 repeat to FSHD. The polymorphic D4Z4 repeat is highly recombinogenic, since somatic mosaicism for rearrangement of D4Z4 is found in as much as 3% of the general population (van Overveld et al., 2000). The D4Z4 repeat consists of identical units defined by the restriction enzyme KpnI, each 3.3 kb, ordered in a head-to-tail fashion, and varying between 11 and 100 units on 'healthy' chromosomes (van Deutekom et al., 1993). Patients with FSHD carry a repeat of 1 to 10 units on one of their chromosomes 4 (Wijmenga et al., 1992). A rough and inverse correlation has been observed between the severity and age at onset of the disease and the residual repeat unit number.

In a review of genetic disorders associated with aberrant chromatin structure, Bickmore and van der Maarel (2003) noted that FSHD represents a potential example of gene activation through loss of repression complexes.

A review of the D4Z4 repeat-mediated pathogenesis of FSHD was provided by van der Maarel and Frants (2005). They pointed out that in contrast to most monogenic disorders, in which the genetic lesion typically affects the structure or function of a specific disease gene, evidence suggested that FSHD is caused by a complex epigenetic mechanism involving the contraction of a subtelomeric macrosatellite repeat. It is likely not the structure but rather the (spatiotemporal-restricted) transcriptional control of one or more disease genes that is perturbed in FSHD as a result of repeat-contraction-mediated chromatin alterations. Their review focused on the cause and consequence of the repeat-array contraction. Van der Maarel and Frants (2005) designated FSHD a macrosatellite repeat-contraction disease.

Van Overveld et al. (2003) showed that contraction of the D4Z4 repeat array causes marked hypomethylation of the contracted D4Z4 allele in individuals with FSHD1. Individuals with FSHD clinically identical to other cases but with an unaltered D4Z4 (FSHD2; 158901) also have hypomethylation of D4Z4. These results strongly suggested that hypomethylation of D4Z4 is a key event in the cascade of epigenetic events causing FSHD1.

Using chromatin immunoprecipitation (ChIP) in HeLa cells, Zeng et al. (2009) found SUV39H1 (300254)-mediated trimethylation of histone H3 (see 602810) at lysine-9 (H3K9), as well as trimethylation at H3 at lysine-27 (H3K27), both at D4Z4, representing transcriptionally repressive heterochromatin. There was also H3K4 dimethylation and H3 acetylation at proximal D4Z4 repeat regions, marking transcriptionally permissive euchromatin. The methylation signal at H3K9, at both the 4q and the 10q locus, was significantly decreased in cell lines derived from patients with FSHD1 (myoblasts and fibroblasts) and FSHD2 (fibroblasts) compared to controls. Contraction of D4Z4 at 1 allele showed a dominant effect on methylation of H3K9 at the other allele, as well as at the 10q locus, suggesting a spreading effect of histone modification. DNA hypomethylation was not observed in FSHD cells, and the decrease in H3K9 methylation was not observed in cells from patients with other forms of muscular dystrophy. Immunoprecipitation studies showed that loss of methylation at H3K9 interrupted binding of CBX3 (604477) and the cohesin complex (see, e.g., SCC1, 606462) at this region. Zeng et al. (2009) hypothesized that loss of H3K9 methylation, and thus loss of CBX3 and cohesion, results in the disruption of chromatin regulation, thereby causing abnormal derepression of distant target genes that leads to the dystrophic phenotype specific to muscle tissue.

DUX4 Expression

Dmitriev et al. (2008) noted that the DUX4 gene (606009) found within the D4Z4 repeat was initially considered to be nonfunctional because it lacked introns and polyadenylation signals. Furthermore, expression of DUX4 could not be detected despite the efforts of several groups using various methods including screening of cDNA libraries, RT-PCR, microarray, and Pol II ChIP. Subsequently, work by Dixit et al. (2007) and Clapp et al. (2007) confirmed expression of DUX4 in human and mouse, respectively, and showed conservation of the DUX4 ORF for more than 100 million years.

Bosnakovski et al. (2008) conditionally expressed cDNAs for FSHD candidate genes within the D4Z4 repeat, DUX4, FRG1 (601278), FRG2 (609032), and ANT1 (SLC25A4; 103220), in mouse C2C12 myoblasts at both high and low expression levels and found that only DUX4 was overtly toxic, as indicated by cellular ATP content, morphologic changes, and apoptosis. DUX4 showed variable toxicity when expressed in mouse fibroblasts or embryoid bodies. DUX4 localized to C2C12 cell nuclei within 2 hours of induction. Microarray analysis revealed altered expression in a broad range of genes, with greatest changes in those involved in growth and development and signal transduction. Expression of Myod was also downregulated at an early time point. Oxidative stress and heat shock genes were downregulated at later time points, suggesting that they may be secondary targets. The DUX4 homeodomains are most similar to those of PAX3 (606597) and PAX7 (167410), and overexpression of these genes rescued viability and proliferation in DUX4-expressing C2C12 cells. Bosnakovski et al. (2008) concluded that DUX4 may cause FSHD by interfering with normal PAX3 or PAX7 function in muscle satellite cells.

Lemmers et al. (2010) showed that FSHD patients carry specific single-nucleotide polymorphisms in the chromosomal region distal to the last D4Z4 repeat. This FSHD-predisposing configuration creates a canonic polyadenylation signal for transcripts derived from DUX4, a double homeobox gene that straddles the last repeat unit and the adjacent sequence. Transfection studies revealed that DUX4 transcripts are efficiently polyadenylated and are more stable when expressed from permissive chromosomes. This particular chromosomal setting containing the pathogenic sequences was named 4APAS (4A polyadenylation signal; Scionti et al., 2012). Lemmers et al. (2010) concluded that their findings suggested that FSHD arises through a toxic gain of function attributable to the stabilized distal DUX4 transcript.

By RT-PCR, Snider et al. (2010) found that full-length DUX4 (DUX4-fl) was aberrantly expressed in 5 of 10 FSHD muscle biopsy specimens, but not in normal muscle specimens. PCR analysis of myoblast cultures of varying pool size, and immunohistochemical analysis of cultured myoblasts revealed that only about 0.1% of FSHD muscle nuclei expressed a relatively abundant amount of DUX4-fl mRNA and protein. DUX4-fl was also aberrantly expressed in differentiated FSHD embryoid bodies. A short variant of DUX4 (DUX4-s) was expressed in both FSHD and control tissues and cells. Aberrant DUX4-fl was often associated with apoptotic changes, and in all cases was associated with the FSHD permissive 4qA haplotype (see below). Snider et al. (2010) concluded that repressive chromatin associated with D4Z4 in differentiated cells may facilitate usage of the noncanonical splice donor site to generate DUX4-s, and the more permissive chromatin in FSHD may favor polymerase progression through to the consensus splice donor and generate DUX4-fl.

Wallace et al. (2011) showed that apoptotic changes observed in DUX4-expressing HEK293 cells were eliminated by inactivating mutation of the first DNA-binding homeobox domain of DUX4. The development of lesions in DUX4-injected mouse muscle was also abrogated by mutation of the DUX4 homeobox domain. Pharmacologic inhibition of p53 (TP53; 191170) mitigated DUX4 toxicity in HEK293 cells, and muscle from p53-null mice were resistant to DUX4-induced damage. Wallace et al. (2011) concluded that DUX4-induced myopathy is dependent on p53-induced apoptosis.

4qA and 4qB Polymorphic Segment

Human 4qter and 10qter share a high degree of similarity, including the D4Z4 repeat array; however, contractions affecting the 10qter repeat are nonpathogenic. Van Geel et al. (2002) detected a polymorphic segment of 10 kb directly distal to D4Z4, which they called alleles 4qA and 4qB. Lemmers et al. (2002) reported that although the 2 alleles are equally common in the general population, FSHD is associated solely with the 4qA allele. They suggested that this was the first example of an intrinsically benign subtelomeric polymorphism predisposing to the development of human disease.

Lemmers et al. (2004) concluded that contractions of D4Z4 on 4qB subtelomeres do not cause FSHD. The 2 allelic variants of 4q, 4qA and 4qB, exist in the region distal to D4Z4. Although both variants are almost equally present in the population, FSHD is associated exclusively with the 4qA allele. Lemmers et al. (2004) identified 3 families with FSHD in which each proband carried 2 FSHD-sized alleles and was heterozygous for the 4qA/4qB polymorphism. Segregation analysis demonstrated that FSHD-sized 4qB alleles are not associated with disease, since these were present in unaffected family members. Thus, in addition to a contraction of D4Z4, additional cis-acting elements on 4qA may be required for the development of FSHD. Alternatively, 4qB subtelomeres may contain elements that prevent FSHD pathogenesis.

Lemmers et al. (2007) hypothesized that allele-specific sequence differences among 4qA, 4qB, and 10q alleles underlie the 4qA specificity of FSHD. By examining sequence variations in the FSHD locus, they demonstrated that the subtelomeric domain of chromosome 4q can be subdivided into 9 distinct haplotypes, of which 3 carry the distal 4qA variation. They showed that repeat contractions in 2 of the 9 haplotypes, 1 of which is a 4qA haplotype, are not associated with FSHD. Lemmers et al. (2007) showed that each of these haplotypes has its unique sequence signature, and proposed that specific SNPs in the disease haplotype are essential for the development of FSHD.

Thomas et al. (2007) measured the frequency of 4qA-defined and 4qB-defined subtelomeric sequences in 164 unrelated patients with FSHD from the UK and Turkey, all known to have large D4Z4 deletions. An almost complete association (162 of 164 patients) was found between large D4Z4 repeat array deletions located on 4qA-defined 4qter subtelomeres and disease expression. DNA samples from 50 controls displayed equivalent frequencies for 4qA and 4qB markers, as did the normal chromosome 4 in all 65 patients studied. The 4qA and 4qB probes failed to hybridize in 2 patients, confirming the presence of an additional rare type of 4qter subtelomeric sequence in humans.

Wang et al. (2011) provided evidence that the 4qB-associated D4Z4 contraction is not pathogenic in Chinese individuals. Molecular reexamination of 3 unrelated Chinese patients originally diagnosed with FSHD on the basis of a D4Z4 contraction showed that all were nonpathogenic 4qB variants. All 3 patients were found to have different disorders with similar phenotypes, including LGMD2A (253600), LGMD2E (604286), and DM1 (160900), respectively. A fourth Chinese patient originally diagnosed with FSHD was found to have a pathogenic 18-kb D4Z4 4qA contraction that she inherited from her symptomatic mother. Her 2 daughters carried a paternally inherited 24-kb nonpathogenic contraction that was found to be a mixture of 4q and 10q. The daughters were thus considered unaffected, whereas previously they had been misdiagnosed as asymptomatic cases. Wang et al. (2011) concluded that the D4Z4 repeat length analysis alone is insufficient for the diagnosis of FSHD, and should be accompanied by 4qA/4qB variant determination.

Expression of Other Genes within the FSHD Candidate Region

Van Deutekom et al. (1996) identified a novel gene, which they referred to as FRG1 (601278), that mapped 100 kb centromeric of the repeated units on chromosome 4q35 that are deleted in FSHD. They identified a polymorphism in exon 1 of this gene and used RT-PCR to amplify reverse transcribed mRNA from lymphocytes and muscle biopsies of patients and controls. These studies indicated that both alleles were transcribed and gave no evidence of 'position effect' variegation leading to repression of allelic transcription.

Gabellini et al. (2002) found that in FSHD muscle, genes located upstream of D4Z4 on 4q35, including FRG1, FRG2 (609032), and ANT1 (103220), are inappropriately overexpressed. They showed that an element within D4Z4 specifically binds a multiprotein complex consisting of transcriptional repressor YY1 (600013), HMGB2 (163906), and nucleolin (NCL; 164035). This multiprotein complex binds D4Z4 in vitro and in vivo and mediates transcriptional repression of 4q35 genes. Gabellini et al. (2002) proposed that deletion of D4Z4 leads to the inappropriate transcriptional derepression of 4q35 genes, resulting in disease. In normal individuals, the presence of a threshold number of D4Z4 repeats leads to repression of 4q35 genes by virtue of the DNA-bound multiprotein complex that actively suppresses gene expression. In FSHD patients, deletion of an integral number of D4Z4 repeats reduces the number of bound repressor complexes and consequently decreases or abolishes transcriptional repression of 4q35 genes.

By oligonucleotide microarrays, Winokur et al. (2003) compared FSHD expression profiles with those from normal muscle and DMD and LGMD2D (608099). Several genes whose expression was altered in an FSHD-specific and highly significant manner are involved in myogenic differentiation, suggesting a partial block in the normal differentiation program. Many of the transcripts affected in FSHD were direct targets of the transcription factor MYOD1 (159970). Additional misexpressed genes confirmed a diminished capacity to buffer oxidative stress, as demonstrated in FSHD myoblasts. This enhanced vulnerability of proliferative stage myoblasts to reactive oxygen species was also disease-specific, further implicating a defect in FSHD muscle satellite cells. None of the genes localizing to the FSHD region at 4q35 were found to exhibit a significantly altered pattern of expression in FSHD muscle. Winokur et al. (2003) hypothesized that disruptions in FSHD myogenesis and oxidative capacity may not arise from a position effect mechanism, as has been previously suggested, but rather from a global effect on gene regulation.

Jiang et al. (2003) found that H4 acetylation levels of a nonrepeated region adjacent to the 4q35 and 10q26 D4Z4 arrays in normal and FSHD lymphoid cells were like those in unexpressed euchromatin, rather than like constitutive heterochromatin. The control and FSHD cells also displayed similar H4 hyperacetylation (like that of expressed genes) at the 5-prime regions of 4q35 candidate genes FRG1 (601278) and ANT1. There was no position-dependent increase in transcript levels from these genes in FSHD skeletal muscle samples compared with controls. Jiang et al. (2003) proposed a model for FSHD in which differential long-distance cis looping depends upon the presence of a 4q35 D4Z4 array with less than a threshold number of copies of the 3.3-kb repeat.

Perini and Tupler (2006) suggested that FSHD might be considered a useful model for the study of position effect in humans. D4Z4 deletion might result in stochastic variation in gene expression in muscle cells and explain the asymmetric involvement of muscles, the great variability of clinical expression between and within families, and the apparent threshold effect whereby there is a requirement for the deletion of a certain number of copies of D4Z4 to develop FSHD.

Osborne et al. (2007) detected no change in expression of the FRG1, FRG2, or ANT1 genes in muscle biopsies from 19 FSHD patients compared to controls. Further studies of the 8-Mb region proximal to the D4Z4 array showed no significant changes in gene expression, no evidence of a position effect, and no evidence of unequal allele-specific expression. However, microarray analysis of global gene expression in FSHD muscle identified 11 upregulated genes with a role in vascular smooth muscle or endothelial cells, suggesting a possible link between muscular dystrophy and vasculopathy in FSHD.

Davidovic et al. (2008) found that myoblasts isolated from patients with FSHD showed an abnormal pattern of expression of isoforms of the FXR1P (600819) gene compared to controls. FXR1P encodes an RNA-binding protein involved in the metabolism of muscle-specific-mRNAs during myogenesis. The altered pattern of FXR1P expression was due to a specific reduced stability of muscle-specific FXR1 mRNA variants. The findings suggested that the molecular basis of FSHD not only involves splicing alterations, but may also involve a deregulation of mRNA stability.

Using RNA-DNA FISH, Masny et al. (2010) found no change in gene transcription of 16 genes in cis in the 4q35 region in nuclei of differentiated myotubes derived from patients with FSHD compared to differentiated myotubes from controls. In particular, there was no change in expression in the FRG1, FRG2, or ANT1 genes, which had previously been implicated.

Dmitriev et al. (2011) showed that KLF15 (606465) bound an enhancer element within the D4Z4 repeat unit. Binding of KLF15 to 2 sites within the D4Z4 enhancer drove expression of FRG2 and DUX4C (DUX4L9; 615581), which are located over 40 kb centromeric to the D4Z4 repeat array. KLF15 expression was upregulated following differentiation of normal human myoblasts and following expression of MYOD, and it was upregulated in FSHD myoblasts, myotubes, and muscle biopsies. FSHD cells also showed upregulated expression of MYOD and the KLF15 target gene PPARG (601487), in addition to DUX4C and FRG2. Dmitriev et al. (2011) concluded that MYOD-dependent KLF15 expression is involved in partial activation of the differentiation program in FSHD myoblasts.

Associations Pending Confirmation

For discussion of a possible association between FSHD and variation in the FAT1 gene, see 600976.

Genotype/Phenotype Correlations

As indicated earlier, FSHD is associated with a short (less than 35 kb) EcoRI/BlnI fragment resulting from deletion of an integral number of units of a 3.3-kb repeat located at 4q35. Vitelli et al. (1999) determined fragment sizes separated by pulsed field gel electrophoresis in a patient with an apparently sporadic case of FSHD and in his healthy family members. A 38-kb fragment was detected in the proband, in his older brother, and in their father. This finding prompted a clinical reevaluation of the father and brother. A subclinical phenotype restricted to abdominal muscle weakness was detected, and serum creatine kinase values were found to be elevated in both. Thus, whereas in healthy individuals the size of the 4q35 polymorphic fragment varies from 48 to 300 kb, and most patients with FSHD show fragments less than 35 kb, fragment sizes between 35 and 48 kb must be interpreted with caution. An inverse correlation between fragment size and severity was described by Lunt et al. (1995) and Tawil et al. (1996). In familial cases, although the size of the inherited fragment remains constant, FSHD seems to become more severe with each generation (anticipation), according to the findings of Griggs et al. (1993), Lunt et al. (1995), Nakagawa et al. (1996), and Tawil et al. (1996).

Wohlgemuth et al. (2003) reported 2 unrelated families with FSHD in which the probands were compound heterozygous for 2 FSHD-sized alleles: a severely affected woman had chromosome 4-type arrays of 17 and 24 kb, and in the other family, a man had arrays of 33 and 36 kb. All of these alleles resided on 4qA. In the first family, 1 unaffected member had the 24-kb allele and 1 affected member had the 17-kb allele; in the second family, 3 unaffected children of the proband carried either the 33-kb allele or the 36-kb allele. Wohlgemuth et al. (2003) noted that the findings showed that having 2 disease alleles is not lethal, and proposed that the phenotype in both probands reflected a dosage effect.

In 21 FSHD1 patients who had 1 translocated chromosome 10-type array on chromosome 4, referred to as 'monosomic,' van Overveld et al. (2005) found D4Z4 hypomethylation compared to monosomic controls. Further analysis delineated 2 classes of clinical severity: patients with repeat sizes of 10 to 20 kb were severely affected and showed pronounced hypomethylation, whereas patients with repeat sizes of 20 to 31 kb showed variation in clinical severity and in hypomethylation. However, the authors could not establish a linear relationship between methylation and disease severity.

Sacconi et al. (2013) found that the SMCHD1 gene (614982), mutation in which causes FSHD2, is a modifier of disease severity in families affected by FSHD1. Three unrelated families with intrafamilial clinical variability of the disorder were studied. In 1 family, a mildly affected man with FSHD1 carried a 9-unit D4Z4 repeat on a 4A allele with no SMCHD1 mutations, whereas his mildly affected wife carried a SMCHD1 mutation (T527M; 614982.0006) on a normal-sized 4A allele, consistent with FSHD2. Their more severely affected son and grandson each carried the 9-unit D4Z4 repeat on a 4A allele as well as the T527M SMCHD1 mutation, consistent with having both FSHD1 and FSHD2. In a second family, a man with a severe early-onset phenotype had both a 9-unit D4Z4 repeat on a 4A permissive allele and a mutation in the SMCHD1 gene. Each of his children, who had milder symptoms, inherited 1 of the genetic defects. In a third family, a man with a severe phenotype was also found to carry a 9-unit D4Z4 repeat on a 4A permissive allele with a SMCHD1 mutation. No information from his parents was available. Transduction of SMCHD1 shRNA into FSHD1 myotubes caused increased levels of DUX4 mRNA as well as transcriptional activation of known DUX4 target genes. These findings were consistent with further chromatin relaxation of the contracted FSHD1 repeat upon knockdown of SMCHD1. Sacconi et al. (2013) concluded that FSHD1 and FSHD2 share a common pathophysiologic pathway converging on transcriptional derepression of DUX4 in skeletal muscle.

Inheritance

Morton and Chung (1959) estimated the frequency to be about 2 per million living persons, with a frequency of about 4 persons destined to develop the trait in each million births. Fertility is little reduced and the mutation rate is not more than 5 per 10 million gametes. Lunt et al. (1989) estimated that the penetrance of the FSHD gene is less than 5% for ages 0 to 4 years, 21% for ages 5 to 9, 58% for ages 10 to 14, 86% for ages 15 to 19, and 95% for age 20 years and over.

Lunt and Harper (1991) studied the families of 41 probands. They found 6 isolated cases that might represent new mutation. Although penetrance reaches 95% by age 20, one-third of heterozygotes over age 40 are mildly affected; 19% over age 40 require wheelchairs. Most patients develop significant lower limb weakness--a fact that is not reflected in the term 'facioscapulohumeral.' Distribution of weakness, severity, age of onset, and serum creatine kinase levels varied among subjects, but provided no clinical evidence for genetic heterogeneity in a comparison of the 11 largest families. Genetic homogeneity, including subjects previously diagnosed with FSH-type spinal muscular atrophy, is strongly supported by the genetic linkage data. They estimated the minimum prevalence in Wales of 2 per 100,000. In this study only 2 of 113 affected persons from the 11 largest families showed no detectable facial weakness; one of these was an obligate carrier, aged 42 years, with complete nonpenetrance. Lunt and Harper (1991) concluded that there is a dominantly inherited scapulohumeral or scapuloperoneal syndrome genetically distinct from FSHD that does not have facial weakness as a feature.

Using the EcoRI polymorphism demonstrated by Wijmenga et al. (1992), Weiffenbach et al. (1993) found 12 recombinants in 5 families, giving a recombination fraction of 0.05 between this marker and the disease. Two families with apparent germline mosaicism were also identified. In these instances, 2 offspring were affected but both parents were clinically normal.

In a study of 34 Brazilian FSHD families, Zatz et al. (1995) concluded that at least one-third of cases may represent new mutations. Somatic mosaicism, furthermore, may not be rare. Biologic fitness was reduced to the 0.6 to 0.82 range, with no difference in sexes. The age at onset of clinical signs, as well as the age at ascertainment, in patients from multigenerational families suggested that anticipation occurs for FSHD.

Zatz et al. (1998) extended their study to 52 families, including 172 patients (104 males and 68 females). Among 273 individuals, 131 (67 males and 64 females) were shown to carry an EcoRI fragment smaller than 35 kb. Of these 131, 114 were examined. The excess of affected males was explained by a greater proportion of asymptomatic females and significantly greater number of affected sons than daughters of asymptomatic mothers. The penetrance by age 30 was 95% for males but only 69% for females. New mutations occurred more frequently in females than in males among somatic/germinal mosaic cases. Severely affected cases were more commonly the result of new mutations or mutations transmitted through maternal lines, including through mosaic mothers. Zatz et al. (1998) noted the implications of these results for prenatal and prognostic counseling depending on the gender of the affected patient.

Tupler et al. (1998) reported monozygotic male twins with FSHD carrying an identical de novo p13E-11 EcoRI fragment of paternal origin. This had arisen in paternal gametogenesis or postzygotically in the paternal chromosome 4 before twinning. A mutated fragment was subsequently transmitted to one of the affected male children. The twins showed great difference in clinical expression with one being almost asymptomatic and the other severely affected. Medical history was the same with the exception of an antirabies vaccination performed at the age of 5 in the more severely affected twin. Tupler et al. (1998) hypothesized that the vaccination might have triggered an inflammatory immune reaction contributory to the more severe phenotype.

Van der Maarel et al. (2000) surveyed 35 de novo FSHD families and found somatic mosaicism in 40% of cases, in either the patient or an asymptomatic parent. Mosaic males were typically affected; mosaic females were more often the unaffected parent of a nonmosaic de novo patient. A genotypic-severity score, composed of the residual repeat size and the degree of somatic mosaicism, yielded a consistent relationship with severity and age at onset of disease. Mosaic females had a higher proportion of somatic mosaicism than did mosaic males. The somatic mosaicism suggested a mainly mitotic origin.

Almost half of de novo FSHD cases originate from a mitotic change leading to somatic mosaicism for the FSHD allele. Lemmers et al. (2004) found that 9 of 37 (24%) patients with presumed de novo FSHD were somatic mosaic, with contraction of D4Z4 identified in 40 to 90% cells. In 7 cases (19%), 1 of the minimally affected or asymptomatic parents was a mosaic 'carrier,' with 10 to 50% affected cells. The authors