G C A T T A C G G C A T genes Review Does DNA Methylation Matter in FSHD? Valentina Salsi 1, Frédérique Magdinier 2 and Rossella Tupler 3,4,5,6,* 1 Department of Life Sciences, University of Modena and Reggio Emilia, 4, 41121 Modena, Italy; [email protected] 2 Aix Marseille Univ, INSERM, MMG, U 1251, 13005 Marseille, France; [email protected] 3 Department of Biomedical, Metabolic and Neural Sciences, University of Modena and Reggio Emilia, 4, 41121 Modena, Italy 4 Center for Neuroscience and Neurotechnology, University of Modena and Reggio Emilia, 4, 41121 Modena, Italy 5 Department of Molecular Cell and Cancer Biology, University of Massachusetts Medical School, Worcester, MA 01003, USA 6 Li Weibo Institute for Rare Diseases Research at the University of Massachusetts Medical School, Worcester, MA 01003, USA * Correspondence: [email protected]; Tel.: +3905-9205-5414 Received: 23 January 2020; Accepted: 25 February 2020; Published: 28 February 2020 Abstract: Facioscapulohumeral muscular dystrophy (FSHD) has been associated with the genetic and epigenetic molecular features of the CpG-rich D4Z4 repeat tandem array at 4q35. Reduced DNA methylation of D4Z4 repeats is considered part of the FSHD mechanism and has been proposed as a reliable marker in the FSHD diagnostic procedure. We considered the assessment of D4Z4 DNA methylation status conducted on distinct cohorts using different methodologies. On the basis of the reported results we conclude that the percentage of DNA methylation detected at D4Z4 does not correlate with the disease status. Overall, data suggest that in the case of FSHD1, D4Z4 hypomethylation is a consequence of the chromatin structure present in the contracted allele, rather than a proxy of its function. Besides, CpG methylation at D4Z4 DNA is reduced in patients presenting diseases unrelated to muscle progressive wasting, like Bosma Arhinia and Microphthalmia syndrome, a developmental disorder, as well as ICF syndrome. Consistent with these observations, the analysis of epigenetic reprogramming at the D4Z4 locus in human embryonic and induced pluripotent stem cells indicate that other mechanisms, independent from the repeat number, are involved in the control of the epigenetic structure at D4Z4. Keywords: FSHD1; FSHD2; D4Z4 macrosatellite; DNA methylation; epigenetics 1. Introduction Facioscapulohumeral muscular dystrophy (FSHD, OMIM#158900) is the third most common myopathy with a reported prevalence ranging between 1 in 8333 [1] and 1 in 20,000 [2]. FSHD is characterized by insidious onset and progressive wasting of a highly selective set of muscle groups, the facial, limb girdle, and foot extensor muscles, and by a great variability of clinical expression among patients and within families [3,4]. The disease appears significantly earlier in males [5–7] and this determines male patients to be in higher number and more severely affected than females [5,6,8–10]. In some families, individuals affected by FSHD can be found only in one generation [5,7,11,12]. These differences are striking in discordant monozygotic twins [13–15]. Genetically, FSHD has been considered a Mendelian disease with autosomal dominant inheritance twins [16–18]. On this basis, the FSHD genetic locus was mapped on chromosome 4q35 by genetic linkage analysis [16,18,19] and associated with rearrangements of an array of tandemly repeated 3.3 kb Genes 2020, 11, 258; doi:10.3390/genes11030258 www.mdpi.com/journal/genes Genes 2020, 11, 258 2 of 19 Genes 2020, 11, 258 2 of 18 segmentsby genetic (D4Z4) linkage [20]. FSHDanalysis is [16,18,19] the only and human associated disease with causally rearrangements linked to of copy an array number of tandemly variation of macrosatelliterepeated DNA3.3 kb elementssegments [(D4Z4)21]. The [20]. number FSHD ofis the D4Z4 only repeats human varies disease from causally 11 to linked 100 in to the copy general population,number whereas variation 10 of repeats macrosatellite or fewer DNA are usuallyelements found [21]. The in sporadic number of and D4Z4 familial repeats FSHD varies patients. from 11 As a generalto rule, 100 in alleles the general composed population, by 11–100 whereas copies 10 of repeats D4Z4 repeats or fewer constitute are usually the found normal in size sporadic range and of D4Z4 alleles,familial whereas FSHD alleles patients. with 8As or fewera general D4Z4 rule, repeats alleles are composed considered by 11–100 diagnostic copies for of the D4Z4 disease repeats [22 ,23]. Theconstitute routine the DNA normal molecular size range testing,of D4Z4 alleles, based whereas on the alleles identification with 8 or offewer D4Z4 D4Z4 arrays repeats with are less than 10considered units at diagnostic 4q35, has for been the considereddisease [22,23]. highly sensitive and specific [23,24]. However, several The routine DNA molecular testing, based on the identification of D4Z4 arrays with less than 10 genotype–phenotype studies have shown that D4Z4 reduction (4–8 reduced units, RU) is also present units at 4q35, has been considered highly sensitive and specific [23,24]. However, several genotype– in 3%phenotype of the general studies population have shown [that25– 28D4Z4] and reduction in some (4–8 cases reduced is associated units, RU) is with also distinctpresent in myopathic 3% of phenotypesthe general not reminiscentpopulation [25–28] of FSHD and in [29 some,30]. cases Moreover, is associated 10% with of FSHD distinct patients myopathic carry phenotypes D4Z4 alleles of borderlinenot reminiscent size (9–10 of FSHD RU), [29,30]. which Moreover, are also found 10% of in FSHD healthy patients people carry or D4Z4 in subjects alleles of with borderline a different myopathysize (9–10 [31] RU), (Ricci which et al., are submitted). also found in Finally, healthy people 5–10% or of in FSHD subjects patients with a different carry D4Z4 myopathy arrays [31] of size within(Ricci the range et al., submitted). of the general Finally, population 5–10% of (11FSHD RU patients or more) carry on D4Z4 both arrays chromosomes of size within 4q [ the32, 33range]. These subjectsof the represent general apopulation second form (11 RU of disease,or more) FSHD2.on both chromosomes Therefore, the 4q [32,33]. number These of D4Z4 subjects RU represent at 4q35 does not pera sesecond characterize form of FSHDdisease, (Figure FSHD2.1). Therefore, the number of D4Z4 RU at 4q35 does not per se characterize FSHD (Figure 1). As a matter of fact, genotype–phenotype studies have shown a large spectrum of clinical As a matter of fact, genotype–phenotype studies have shown a large spectrum of clinical phenotypes in myopathic subjects [29–31], carrying a D4Z4 reduced allele as well as reduced penetrance phenotypes in myopathic subjects [29–31], carrying a D4Z4 reduced allele as well as reduced amongpenetrance relatives carryingamong relatives the same carrying D4Z4 the reduced same D4Z4 allele reduced [7,9,10,12 allele,25, 34[7,9,10,12,25,34,35].,35]. All this has All relevant this has eff ects on clinicalrelevant practice, effects complicatingon clinical practice, diagnosis, complicating prognosis, diagnosis, and geneticprognosis, counseling. and genetic counseling. FigureFigure 1. Molecular 1. Molecular complexity complexity in facioscapulohumeral in facioscapulohumeral muscular muscular dystrophy dystrophy (FSHD). (FSHD). D4Z4 contractions D4Z4 are associatedcontractions with are a associated permissive with 4qA a permissive genotype 4qA that genotype involves that the involves aberrant the aberrant expression expression of the ofDUX4 retrogenethe DUX4 and is retrogene responsible and foris responsible FSHD1, but for also FSHD1, occurs but also in 1.3–2% occurs ofin 1.3–2% the normal of the population. normal population. 4qA is also found4qA in cases is also presenting found in complex cases presenting phenotypes complex and asphenotypes well as in and 30–50% as well of healthyas in 30–50% relatives. of healthy Mutations in SMCHD1relatives.or MutationsDNMT3B ingenes SMCHD1 have or been DNMT3B associated genes with have FSHD2been associated and are with responsible FSHD2 forand ICFare and BAMSresponsible syndromes. for Parents ICF and of BAMS ICF patients syndromes. are Parents heterozygous of ICF patients for DNMT3B are heterozygouspathogenic for variants DNMT3B but do pathogenic variants but do not show any sign of muscular dystrophy. Other myopathic patients carry not show any sign of muscular dystrophy. Other myopathic patients carry SMCHD1 mutations. Grey SMCHD1 mutations. Grey dot = methyl CpG; white dot = unmethylated CpG. dot = methyl CpG; white dot = unmethylated CpG. 2. Molecular Features and the Epigenetic Model for FSHD 2. Molecular Features and the Epigenetic Model for FSHD The D4Z4 repeat is part of a family of 3.3 kb sequences frequently found in heterochromatic Theregions, D4Z4 such repeat as the is partshort ofarms a family of the ofacrocentric 3.3 kb sequences chromosomes frequently [36], and found a nearly in identical heterochromatic and regions,equally such polymorphic as the short armsD4Z4 ofarray the reside acrocentric at the sub chromosomes‐telomere of chromosome [36], and a nearly 10q [37,38]. identical and equally polymorphicD4Z4 D4Z4 is a CpG array‐rich reside (73%) at macrosatellite the sub-telomere DNA element of chromosome encompassing 10q [more37,38 than]. 16 nucleosomes D4Z4and containing is a CpG-rich multiple (73%) repeat macrosatellite sequences normally DNA element associated encompassing with heterochromatin more than [36]. 16 The nucleosomes D4Z4 and containingrepeat