Identifying diagnostic DNA methylation profiles for facioscapulohumeral muscular dystrophy in blood and saliva using bisulfite sequencing
© Jones et al.; licensee BioMed Central Ltd. 2014
Received: 10 July 2014
Accepted: 17 October 2014
Published: 29 October 2014
Facioscapulohumeral muscular dystrophy (FSHD) is linked to chromatin relaxation due to epigenetic changes at the 4q35 D4Z4 macrosatellite array. Molecular diagnostic criteria for FSHD are complex and involve analysis of high molecular weight (HMW) genomic DNA isolated from lymphocytes, followed by multiple restriction digestions, pulse-field gel electrophoresis (PFGE), and Southern blotting. A subject is genetically diagnosed as FSHD1 if one of the 4q alleles shows a contraction in the D4Z4 array to below 11 repeats, while maintaining at least 1 repeat, and the contraction is in cis with a disease-permissive A-type subtelomere. FSHD2 is contraction-independent and cannot be diagnosed or excluded by this common genetic diagnostic procedure. However, FSHD1 and FSHD2 are linked by epigenetic deregulation, assayed as DNA hypomethylation, of the D4Z4 array on FSHD-permissive alleles. We have developed a PCR-based assay that identifies the epigenetic signature for both types of FSHD, distinguishing FSHD1 from FSHD2, and can be performed on genomic DNA isolated from blood, saliva, or cultured cells.
Samples were obtained from healthy controls or patients clinically diagnosed with FSHD, and include both FSHD1 and FSHD2. The genomic DNAs were subjected to bisulfite sequencing analysis for the distal 4q D4Z4 repeat with an A-type subtelomere and the DUX4 5’ promoter region. We compared genomic DNA isolated from saliva and blood from the same individuals and found similar epigenetic signatures. DNA hypomethylation was restricted to the contracted 4qA chromosome in FSHD1 patients while healthy control subjects were hypermethylated. Candidates for FSHD2 showed extreme DNA hypomethylation on the 4qA DUX4 gene body as well as all analyzed DUX4 5’ sequences. Importantly, our assay does not amplify the D4Z4 arrays with non-permissive B-type subtelomeres and accurately excludes the arrays with non-permissive A-type subtelomeres.
We have developed an assay to identify changes in DNA methylation on the pathogenic distal 4q D4Z4 repeat. We show that the DNA methylation profile of saliva reflects FSHD status. This assay can distinguish FSHD from healthy controls, differentiate FSHD1 from FSHD2, does not require HMW genomic DNA or PFGE, and can be performed on either cultured cells, tissue, blood, or saliva samples.
KeywordsBisulfite sequencing DNA methylation D4Z4 Disease diagnostics DUX4 Epigenetics Facioscapulohumeral muscular dystrophy FSHD1 FSHD2
The typical genetic diagnosis for FSHD1 is complex [12, 22]. It first requires careful isolation of 40 to 50 μg of very high molecular weight (HMW) DNA from peripheral blood mononuclear cells (PBMCs) obtained from fresh blood samples . The purified genomic DNA is then embedded in agarose for in-gel digestion with combinations of several restriction enzymes. The agarose-DNA plugs are subjected to PFGE, Southern blotting, and hybridizations with DNA probes: the p13E-11 probe to identify the size of each 4q35 and 10q26 array , and probes for the generally permissive A-type subtelomere and the non-permissive B-type subtelomere to identify the haplotype of 4q35 and 10q26 chromosomes, respectively. Recently, an alternative fluorescent cell-based technique, termed molecular combing, was developed to identify an FSHD1 deletion on a 4qA chromosome . The additional information one obtains from these assays includes interchromosomal rearrangements and potential somatic mosaicism; however, these assays are incapable of identifying a functional DUX4-fl PAS [12, 22, 25, 26]. The ~5% of clinical FSHD patients that do not have an FSHD1-sized pathogenic 4qA allele are candidates for FSHD2, but neither assay can identify these individuals as FSHD2 as opposed to another myopathy with similar clinical symptoms. Sequencing the SMCHD1 gene for known FSHD2 mutations in candidates with permissive 4A-type subtelomeres will identify many, but not all, FSHD2 subjects [21, 27, 28].
Results and discussion
Development of a combined distal 4qA-specific and 4q/10q 5’ D4Z4 DNA methylation assay
To complement the distal D4Z4 methylation analysis and provide the context for both 4q35 D4Z4 arrays that is important for the determination of FSHD2 status, we designed a third BSS analysis upstream of the DUX4 open reading frame (referred to as the DUX4 5’ BSS assay). This assay analyzes the methylation status of 59 CpGs preferentially in 4q35 D4Z4 RUs but also in 10q35 D4Z4 RUs (Figures 2 and Additional file 1: Figure S1E). This DUX4 5’ region can be amplified from all 4q35 and 10q26 D4Z4 RUs, does not amplify homologous D4Z4s elsewhere in the genome , and encompasses a putative CTCF binding site and the DR1 region found to be hypomethylated in all 4q and 10q D4Z4 RUs in FSHD2 cells [29, 31]. As anticipated, all seven of the BS-converted genomic DNAs were successfully amplified using this protocol (Figure 3, lower panel), validating the integrity of the BS-converted DNAs from the two healthy subjects. Analysis of the DUX4 5’ BSS products revealed that all 59 of the CpGs predicted by the reference sequence were accounted for in all sequences in this assay, confirming that these sequences were derived from 4q/10q D4Z4 RUs, which characteristically have very few polymorphisms, and not from homologous D4Z4s located elsewhere in the genome that contain numerous sequence polymorphisms . Thus, combining the DUX4 5’ BSS and 4qA/4qA-L BSS assays provides a detailed analysis of the DNA methylation status of the pathogenic distal 4qA D4Z4 RU in the context of overall 4q/10q D4Z4 DNA methylation.
Characterization of healthy and FSHD1 DNA methylation patterns in the distal D4Z4 repeat unit using blood and saliva
Epigenetic marks often show tissue specificity; thus, it is very important to carefully examine and compare each locus of interest when performing epigenetic studies on genomic DNAs isolated from different tissue sources . Since FSHD is a myopathy and the pathogenic DUX4 mRNA is expressed predominantly in skeletal muscle [1, 33], the epigenetic status of myocytes is of particular interest. However, muscle biopsies require participants to visit a hospital or clinic, and can be expensive, painful, and difficult to obtain from FSHD patients of any age already exhibiting muscle atrophy. Fortunately, in FSHD1 and FSHD2, the DNA methylation status of the 4q35 D4Z4 is similar between PBMCs and myogenic cells . For example, in FSHD1, the proximal repeats of the D4Z4 array on the contracted 4q35 allele are significantly hypomethylated in both PBMCs and myogenic cells compared to the non-contracted allele or healthy controls . In order to assess the DNA methylation status of the pathogenic distal 4q35 D4Z4 repeat, we used our 4qA and 4qA-L BSS assays to analyze the distal D4Z4 in PBMCs from FSHD1 patients and healthy first-degree relatives. In addition, we are interested in analyzing the epigenetic signatures of large numbers of family members over time, including healthy individuals, some of whom may be identified as potential asymptomatic carriers. Therefore, in addition to testing our assay on genomic DNA isolated from PBMCs, we performed our analysis on saliva samples obtained from the same subjects for a comparison. The advantage of saliva samples is that they can be collected without additional help, there is no needle injection, and collection kits can be mailed to subjects who have undergone informed consent, with the stable 2 mL sample returned by standard mail. This type of testing would be particularly useful for children and in communities or countries where access to a phlebotomist is limiting or relatively expensive and/or standard genetic testing by PFGE or molecular combing is cost-prohibitive or unavailable.
To avoid diluting the signature of FSHD1 by averaging with the methylation levels of the non-contracted array, we use the first quartile (Q1) of the methylation percent of all analyzed chromosomes as a summary statistic. This corresponds to dividing all sequences into two groups based on methylation percentage, and taking the median value of only those sequences in the lower group. If the total number of sequences is odd, there is the issue of whether to include the central sequence in the lower group or not before taking the median; to give it half weight we compute the median both ways, then take the arithmetic average. This corresponds to the R function quantiles with type = 5.
In a 4qA/A FSHD1 subject for whom all chromosomes with the contracted array have lower 4qA BSS methylation than any chromosomes with the non-contracted array, Q1 gives an estimate of the median 4qA methylation of just the contracted array. With n = 10 sequences analyzed, there is a 5.4% chance that more than 75% will arise from the non-contracted allele due to random sampling, so Q1 will not be an accurate reflection of the contracted allele; increasing n to 18 reduces the probability of this sort of failure to 1.5%.
Note, however, that if there is any overlap in methylation levels between alleles (as may be expected in healthy controls, FSHD2 subjects, and, potentially, in some FSHD1 subjects as well), then the half of analyzed sequences with lower methylation need not arise from a single allele, and Q1 underestimates the median methylation of any one allele. In the extreme case of no difference in methylation distributions between two 4qA alleles, or of 4qA/4qB genotypes (in which all sequences arise from a single allele), Q1 instead is an estimate of the lower quartile of methylation of one allele, rather than the median. This bias is tolerable for the present application, so for simplicity we use Q1 (Additional file 2: Table S1) as a summary statistic uniformly for all samples, without requiring the genotype to be known; we have also developed a mixture-model based statistical approach that aims to mitigate this bias [unpublished observations by Jones et al. 2014].
As shown in Figure 4A and Additional file 2: Table S1, the distal 4qA D4Z4 was dramatically hypomethylated in both blood and saliva samples for subjects 75194 (Q1 = 21.4% methylated, PBMCs; Q1 = 10.7% methylated, saliva) and 75204 (Q1 = 7.1% methylated, PBMCs; Q1 = 8.9% methylated, saliva), and was hypermethylated in both blood and saliva of subject 75195 (Q1 = 87.5% methylated, PBMCs; Q1 = 89.3% methylated, saliva). The 4qA-L BSS analysis indicated that the A-L haplotype was only present in subject 75194 and this allele was hypermethylated (Q1 = 70.8% methylated, PBMCs; Q1 = 80.0% methylated, saliva). Neither of these 4qA-specific BS PCRs produced a product from either the PBMCs or saliva of subject 75205, indicating that this subject lacked any 4qA alleles. Based on this analysis we predicted that subjects 75194 and 75204 were FSHD patients, and subjects 75195 and 75205 were healthy controls.
BSS analysis of the DUX4 5’ promoter region is more complex (Figure 2). This analysis was designed to preferentially detect all 4q D4Z4s regardless of haplotype from both the contracted and non-contracted 4q chromosome arrays. Because in FSHD1 only the contracted chromosome 4 D4Z4 is hypomethylated , the observed proportion of hypomethylated sequences is expected to depend on the number of D4Z4 RUs in the contracted 4q array relative to the number of D4Z4 RUs on the non-contracted 4q array, together with chromosome 4q-type RUs on hybrid chromosome 10s, if present. In addition, preference for the 4q D4Z4 is based on a conserved 4q-specific polymorphism at the 3’ terminal base of a BSS PCR primer; however, since this relies on a single base polymorphism, there is the potential that a fraction of 10q-derived D4Z4 sequences could be amplified. In fact, sequence analysis of the DUX4 5’ BS PCR products identified both 4qA-specific polymorphisms and 10q-specific polymorphisms, indicating that although the reaction has a preference for 4q, it does not preclude amplification of some 10q array RUs. Fortunately, this does not adversely affect our analysis. For healthy controls, we anticipate that the vast majority of the analyzed chromosomes will show D4Z4 hypermethylation (>35% methylation) regardless of origin. By contrast, FSHD1 subjects should contain a combination of hypermethylated (from D4Z4 RUs residing in the non-contracted 4q array and both 10q arrays) and hypomethylated (from the D4Z4 RUs residing in the contracted 4q array) sequences with a clear minority of the analyzed D4Z4 RUs being hypomethylated; FSHD2 subjects should be hypomethylated (~ <25% methylation) on most sequences analyzed. Thus, the DUX4 5’ BSS assay is expected to be less sensitive than the 4qA and 4qA-L BSS assays in distinguishing FSHD1 from healthy controls; however, this assay should support those results, and would clearly distinguish FSHD2 from FSHD1 or healthy controls. Therefore, to more accurately distinguish FSHD1 from FSHD2 we use the mean percent methylation of each sample for comparison (Additional file 2: Table S1).The DUX4 5’ BSS analysis was tested on the same eight genomic DNA samples as above (Figure 4B). As with the 4qA BSS assay, DUX4 5’ BS products from subjects 75195 (91.0% methylation mean, PBMCs; 94.4% methylation mean, saliva) and 75205 (71% methylation mean, PBMCs; 63.9% methylation mean, saliva) showed pronounced DNA hypermethylation in both PBMCs and saliva, suggesting that these two subjects were healthy controls. Subjects 75194 (47.8% methylation mean, PBMCs; 59.7% methylation mean, saliva) and 75204 (50.8% methylation mean, PBMCs; 59.3% methylation mean, saliva) showed less methylation than the putative controls but more methylation, on average, than found for these samples in the 4qA BSS analysis. However, in accordance with our predictions for FSHD1, these subjects contained a mixture of hypermethylated and hypomethylated DNA, resulting in a wide range of DNA methylation density per analyzed chromosome that reached much lower in subjects 75194 (5.1 to 78.0% methylation, PBMCs; 5.1 to 81.4% methylation, saliva) and 75204 (6.8 to 91.5% methylation, PBMCs; 5.1 to 88.1% methylation, saliva) compared with 75195 (72.9 to 100% methylation, PBMCs; 78.0 to 100% methylation, saliva) and 75205 (40.7 to 88.1% methylation, PBMCs; 32.2 to 86.4% methylation, saliva). This data indicates that subjects 75194 and 75204 are FSHD1 and not FSHD2 patients, while subjects 75195 and 75205 are healthy controls. In each case, the genomic DNAs isolated from PBMCs and saliva samples produced similar BSS results for each subject.
Upon final analysis, subjects 75194 and 75204 exhibited D4Z4 hypomethylation detected by the 4qA BSS analysis (Q1 < 25% methylated), indicative of FSHD, and by the DUX4 5’ BSS analysis they were clearly not FSHD2 (see below) and were thus predicted to be two FSHD1 patients. In fact, subjects 75194 and 75204 indeed had positive genetic tests for FSHD1. Importantly, subject 75204 (34 kb Eco RI/Bln I fragment corresponding to 9 D4Z4 RUs) and subject 75194 (27 kb Eco RI/Bln I fragment corresponding to 7 D4Z4 RUs) were both in the high end of the genetic FSHD1 contraction range, yet both were still accurately identified as FSHD1 by our analysis highlighting the sensitivity of these assays. Similarly, subjects 75195 and 75205, displaying hypermethylation at D4Z4 of all analyzed sequences by both the 4qA BSS and the DUX4 5’ BSS methods, were accurately determined to be healthy controls. With respect to the distal 4qA BSS analysis, subject 75195 was accurately identified from both blood and saliva genomic DNA as a healthy control, while control subject 75205 was accurately determined to lack a 4qA allele at either chromosome 4 (see below).
Overall, genomic DNAs isolated from blood and saliva provided similar epigenetic profiles of the FSHD-associated D4Z4 array in FSHD1 affected patients and healthy first-degree relatives. This test analysis confirmed the specificity of the 4qA BSS and 4qA-L BSS protocols for 4qA alleles over 10qA alleles or 4qB alleles. In addition, we have applied this analysis to myogenic cells or PBMCs from an additional 20 subjects having a clinical and genetic diagnosis of FSHD1 and 10 subjects confirmed as healthy unaffected. The simple cutoff of Q1 < 30% for 4qA and 4qA-L methylation accurately classified 19 of the 20 FSHD subjects and 9 of the 10 healthy controls (P = 7 × 10-6 by Fisher’s Exact Test); the one false positive was the only sample in the intermediate zone of 25% < Q1 < 35% [unpublished observations by Jones et al., 2014]. We conclude that the described BSS analysis can readily identify FSHD1 hypomethylation, is suitable for epigenetic analysis of the D4Z4 array in both FSHD1 and healthy subjects, and that saliva samples are comparable to PBMCs in terms of providing suitable genomic DNA for DNA methylation analysis of the 4q35 D4Z4.
Identification of the FSHD2 DNA hypomethylation signature
Identification and elimination of the rare 10A176T and 4A166 non-permissive haplotypes from BSS analysis
It is important to keep in mind that the majority of analyzed chromosomes in FSHD and healthy subjects will have chromosomes with standard 4qA (44%, including 4qA-L), 4qB (50%), and 10qA (91%) haplotypes; however, there are some important exceptions to consider . Two of them are the rare, non-permissive 10A176T and 4A166 haplotypes, neither of which is identified by current standard diagnostic testing . Since D4Z4 arrays of 10A176T have chromosome 4-like resistance to digestion with Bln-I, the enzyme used to distinguish chromosome 4 arrays from chromosome 10 arrays, this chromosome 10 haplotype can be misidentified as chromosome 4 by PFGE analysis and 4A166 linked arrays are indistinguishable from permissive 4qA arrays using PFGE. Thus, the presence of 10A176T or 4A166 can complicate genetic diagnosis and epigenetic analyses, particularly when these haplotypes are associated with a short D4Z4 array. Since the prevalence of 10A176T and 4A166 in the European population are ~2.5% and ~4.1%, respectively, it is to be expected that ~1 out of 15 FSHD patients, healthy control subjects, and even patients with other myopathies will carry one of these potentially confusing haplotypes . Fortunately, the 10A176T and 4A166 alleles have several distinguishing polymorphisms and can be identified by PCR haplotyping of genomic DNA . However, for our diagnostic purposes as well as epigenetic analyses, it is important to know if our 4qA and 4qA-L BSS assays can identify and/or eliminate these non-permissive 10A176T or 4A166 haplotypes from the BSS analysis.
Combined analysis and epigenetic diagnosis of FSHD
Even a small false positive rate (e.g., 1%) can result in poor positive predictive value when applied to populations in which FSHD prevalence is smaller still (such as the general population). However, since individuals with a variety of non-FSHD muscular dystrophies have D4Z4 methylation-levels similar to healthy controls , our assay can be used as a differential diagnostic between FSHD and other diseases when applied to patients with clinical characteristics consistent with FSHD. In addition, all of the samples from FSHD1 subjects that were tested with the DUX4 5’ BSS assay showed DNA methylation levels above 25%, consistent with an FSHD1 diagnosis and not FSHD2. Conversely, all FSHD2 subjects showed DNA methylation levels well below 25% in both the DUX4 5’ and 4qA BSS assays, providing clear evidence for FSHD2 as opposed to FSHD1. However, while this assay is specific for the generally FSHD permissive 4qA allele, as with standard FSHD1 testing by PFGE or molecular combing , it does not positively identify a functional DUX4 PAS, which is required of a truly permissive 4qA allele. We conclude that the combination of these two assays used for individuals with clinical symptoms of FSHD is diagnostic for FSHD1 and FSHD2 (Figure 6).
We have developed a PCR-based technique to identify and distinguish all forms of FSHD from DNA methylation profiles in blood, saliva, or fibroblasts. The combination of two BSS assays allows the analysis of the DNA methylation profile of a portion of the distal 4q35 D4Z4 RU associated with all forms of FSHD. These assays are specific for 4q chromosomes with the FSHD-associated A-type subtelomere and do not amplify D4Z4 sequence from B-type subtelomeres. Sequences from non-permissive 10qA (including 10qA176T) and 4A166 are not amplified in most assays and, if present (a sign of PCR primer degradation), are readily removed from analysis. The DNA methylation profiles produced by this assay clearly distinguish between FSHD and healthy subjects (Figure 8). We also describe a companion BSS assay that analyzes the DNA methylation status of a region 5’ of the DUX4 gene that is present on all 4q35 and 10q26 D4Z4 repeats. Utilizing the three BSS assays in combination discloses the DNA methylation status of the distal D4Z4 in the context of overall 4q35 D4Z4 DNA methylation. Therefore, in addition to determining contracted 4qA-specific DNA hypomethylation characteristic of FSHD1 and overall D4Z4 hypermethylation in healthy controls, this assay identifies FSHD2-specific DNA hypomethylation signatures on the 4qA allele and clearly distinguishes them from FSHD1 signatures (Figure 8). Importantly, this analysis does not require HMW genomic DNA and can be performed on genomic DNAs isolated from blood or saliva, producing similar results. Additionally, the protocols can readily be modified with bar-coded oligonucleotide primers such that data acquisition and analysis can be performed using next-generation sequencing technology.
Subjects and methods
The University of Massachusetts Medical School Institutional Review Board approved this study; participants provided written informed consent. Patients 75194, 75204, and RB19518 were clinically diagnosed as FSHD. Patients 75194 and 75204 each had a positive genetic test for FSHD1 and RB19518 had a negative genetic test for FSHD1. Subjects 75205 (healthy relative of 75204) and 75195 (healthy relative of 75194) were clinically unaffected. The FSHD2 family cohort (1090) was previously described  and contains a mutation in the SMCHD1 gene that segregates with disease. Myogenic cells for cohort 27 were obtained from the previously described Wellstone Center cell repository housed at the University of Massachusetts Medical School [6, 35].
Sample collection and DNA preparation
Saliva samples (2 mL) were collected from subjects using the DNAgenotek Oragene Discover (ORG-500) DNA collection kit and genomic DNAs were isolated using the manufacturer’s recommended protocol. Genomic DNAs from blood samples were isolated using the Qiagen Puregene DNA isolation kit using the recommended protocol.
DNA methylation analysis
DNA methylation was analyzed by BSS assay. BS conversion was performed on 1 μg of genomic DNA using the EpiTect Bisulfite Kit (Qiagen) as per the manufacturer’s instructions, and 200 ng of converted genomic DNA was used per PCR. For the 4qA BSS analysis, converted DNA was amplified by nested PCR using oligonucleotide primers and thermocycling conditions that amplify 4qA but not 4qB; the initial PCR was performed with oligonucleotide primers BSS1438F (5’-GTTTTGTTGGAGGAGTTTTAGGA) and BSS3742R (5’-AACATTCAACCAAAATTTCACRAAA) and then followed by nested PCR with oligonucleotide primers BSS1438F and BSS3626R (5’-AACAAAAATATACTTTTAACCRCCAAAAA) using 10% of the first PCR product as template. Polymorphic nucleotide changes that preferentially amplify the 4A subtelomeric region are underlined. The BSS3742R sequence does not exist in 4B or 10B and utilizes a polymorphic change at bp 7946 in FJ439133 to eliminate 10A166, and BSS3626R utilizes polymorphic changes at bp 7827 in FJ439133 to eliminate 10A, 4B, and 10B . All PCRs were performed using GoTaq Hot Start Polymerase (Promega) as follows: 94°C for 2 min, 25 cycles of 94°C for 15 sec, 58°C for 20 sec, and 72°C for 50 sec, followed by a final extension at 72°C for 10 min. The 593-bp PCR product spans the end of full-length DUX4 exon 1 to the beginning of DUX4 exon 3, therefore allowing specific analysis of the methylation status of the most distal 4qA D4Z4 repeat, which contains 57 CpGs (Additional file 1: Figure S1A). For the 4qA-L BSS analysis, converted DNA was similarly amplified by nested PCR. The initial PCR was performed with oligonucleotide primers BSS4qALF (5’-TTATTTATGAAGGGGTGGAGTTTGTT) and BSS3742R, and then followed by nested PCR with oligonucleotide primers 4qALF and BSS3626R using 10% of the first PCR product as template. All PCRs were performed using GoTaq Hot Start Polymerase (Promega) as follows: 94°C for 2 min, 25 cycles of 94°C for 15 sec, 58°C for 20 sec, and 72°C for 30 sec followed by a final extension at 72°C for 10 min. The 354-bp PCR product spans the 3’ end of the extended 4qA-L D4Z4 repeat to the beginning of DUX4 exon 3, therefore allowing specific analysis of the methylation status of the most distal 4qA D4Z4 repeat sequence, which contains 30 CpGs (Additional file 1: Figure S1D). When no PCR product was obtained with either the 4qA- or 4qA-L-specific BS PCRs, DNA methylation status of same distal D4Z4 region was analyzed using primer BSS3702R (5’-AAAACCAACRAACTCCCTTACAC) instead of BSS3626R. BSS3702R amplifies distal D4Z4 from both 10A and 4A. For the DUX4 5’ region, BS-converted DNA was amplified by nested PCR as described above. The initial PCR was performed with oligonucleotide primers BSS167F (5’-TTTTGGGTTGGGTGGAGATTTT) and BSS1036R (5’-AACACCRTACCRAACTTACACCCTT), and then followed by nested PCR with oligonucleotide primers BSS475F (5’-TTAGGAGGGAGGGAGGGAGGTAG) and BSS1036R. A polymorphic nucleotide change at bp 6748 in FJ439133 (underlined) was used to preferentially amplify the 4A subtelomeric region. This 578-bp PCR product contains 61 CpGs to preferentially analyze the methylation status of the DUX4 5’ region of chromosome 4-type D4Z4 repeats (Additional file 1: Figure S1E).
All BS PCR products were cloned into the pGEM-T Easy Vector system I (Promega) for sequencing analysis. At least 10 clones were sequenced for each subject and their methylation status was analyzed using web-based analysis software BISMA (http://biochem.jacobs-university.de/BDPC/BISMA/)  with the default parameters. Default parameters have a lower threshold of 90% identity to the reference sequence, a lower threshold of BS conversion rate of 95%, and remove identical sequences derived from the same genomic template based on conversion artifacts. To remove PCR amplification bias, 1 CpG in BSS3626R primer and 2 CpGs in BSS1036R primer were removed from the analysis; therefore, a total of 56 CpGs, 30 CpGs, and 59 CpGs were analyzed for the 4qA, 4qA-L, and DUX4 5’ region, respectively. The “R” designation in primer sequences represents a purine (A or G).
Detection of 10A176T haplotype
BSS analysis using our 4qA-specific BSS primers and conditions does not amplify 10A176T alleles and will eliminate 10A176T from analysis. To confirm a 10A176T haplotype or analyze its DNA methylation status, oligonucleotide primer BSS3626R was replaced with BSS3702R. The bases corresponding to the 55th CpG in the 4qA BSS fragment are “TA” in 10A176T alleles due to the G7820A polymorphic change, and the C7808A polymorphism can be identified as an “A” instead of a “T” at this position in the BS-converted 10A176T .
Detailed genotyping of 4q chromosomes
Standard genomic PCR was performed on non-converted DNA to identify the 4qA, 4qA-L, and 4qB chromosomes as previously described .
- BS PCR:
Facioscapulohumeral muscular dystrophy
High molecular weight
Peripheral blood mononuclear cells
Polymerase chain reaction
Pulse-field gel electrophoresis
The authors thank the participating subjects and their families, and Drs. Kathryn R. Wagner and Genila Bibat (Kennedy-Krieger Institute and Johns Hopkins School of Medicine) for initially collecting the clinical samples for the Wellstone Center repository , Dr. Jennifer C.J. Chen for establishing the Wellstone Center cell cultures, and Dr. Charis L. Himeda for helpful comments on the manuscript. This work was funded by grant R01AR062587 from the National Institute of Arthritis, Musculoskeletal and Skin Diseases NIH awarded to PLJ. ODK is supported by the Senator Paul D. Wellstone Muscular Dystrophy Cooperative Research Center grant U54HD060848 from the Eunice Kennedy Shriver National Institute for Child Health and Human Development. TIJ was supported by grant MDA216652 from the Muscular Dystrophy Association. CY was supported by the Chinese Scholarship Council (CSC) File No. 201306760041 and grant R01AR062587 (PLJ). PCS was supported through the auspices of Dr. H. Robert Horvitz, an Investigator at the Howard Hughes Medical Institute in the Department of Biology at the Massachusetts Institute of Technology. In addition, we thank the Chris Carrino Foundation for FSHD and the FSH Society for their continued support. The funding agencies played no role in the design, collection, analysis, or interpretation of data, or in the decision to publish this study.
- Padberg GW: [thesis]. Facioscapulohumeral Disease. 1982, Leiden, The Netherlands: Leiden UniversityGoogle Scholar
- Tawil R, Van Der Maarel SM: Facioscapulohumeral muscular dystrophy. Muscle Nerve. 2006, 34: 1-15. 10.1002/mus.20522.View ArticlePubMedGoogle Scholar
- Deenen JC, Arnts H, van der Maarel SM, Padberg GW, Verschuuren JJ, Bakker E, Weinreich SS, Verbeek AL, van Engelen BG: Population-based incidence and prevalence of facioscapulohumeral dystrophy. Neurology. 2014, 83 (12): 1056-1059. 10.1212/WNL.0000000000000797.PubMed CentralView ArticlePubMedGoogle Scholar
- Brouwer OF, Padberg GW, Wijmenga C, Frants RR: Facioscapulohumeral muscular dystrophy in early childhood. Arch Neurol. 1994, 51: 387-394. 10.1001/archneur.1994.00540160085011.View ArticlePubMedGoogle Scholar
- Chen TH, Lai YH, Lee PL, Hsu JH, Goto K, Hayashi YK, Nishino I, Lin CW, Shih HH, Huang CC, Liang WC, Wang WF, Jong YJ: Infantile facioscapulohumeral muscular dystrophy revisited: Expansion of clinical phenotypes in patients with a very short EcoRI fragment. Neuromuscul Disord. 2013, 23 (4): 298-305. 10.1016/j.nmd.2013.01.005.View ArticlePubMedGoogle Scholar
- Jones TI, Chen JC, Rahimov F, Homma S, Arashiro P, Beermann ML, King OD, Miller JB, Kunkel LM, Emerson CP, Wagner KR, Jones PL: Facioscapulohumeral muscular dystrophy family studies of DUX4 expression: evidence for disease modifiers and a quantitative model of pathogenesis. Hum Mol Genet. 2012, 21: 4419-4430. 10.1093/hmg/dds284.PubMed CentralView ArticlePubMedGoogle Scholar
- Ricci G, Scionti I, Sera F, Govi M, D’Amico R, Frambolli I, Mele F, Filosto M, Vercelli L, Ruggiero L, Berardinelli A, Angelini C, Antonini G, Bucci E, Cao M, Daolio J, Di Muzio A, Di Leo R, Galluzzi G, Iannaccone E, Maggi L, Maruotti V, Moggio M, Mongini T, Morandi L, Nikolic A, Pastorello E, Ricci E, Rodolico C, Santoro L: Large scale genotype-phenotype analyses indicate that novel prognostic tools are required for families with facioscapulohumeral muscular dystrophy. Brain. 2013, 136: 3408-3417. 10.1093/brain/awt226.PubMed CentralView ArticlePubMedGoogle Scholar
- Schaap M, Lemmers RJ, Maassen R, van der Vliet PJ, Hoogerheide LF, van Dijk HK, Basturk N, de Knijff P, van der Maarel SM: Genome-wide analysis of macrosatellite repeat copy number variation in worldwide populations: evidence for differences and commonalities in size distributions and size restrictions. BMC Genomics. 2013, 14: 143-10.1186/1471-2164-14-143.PubMed CentralView ArticlePubMedGoogle Scholar
- Wijmenga C, Frants RR, Brouwer OF, Moerer P, Weber JL, Padberg GW: Location of facioscapulohumeral muscular dystrophy gene on chromosome 4. Lancet. 1990, 336: 651-653. 10.1016/0140-6736(90)92148-B.View ArticlePubMedGoogle Scholar
- Wijmenga C, Hewitt JE, Sandkuijl LA, Clark LN, Wright TJ, Dauwerse HG, Gruter AM, Hofker MH, Moerer P, Williamson R, van Ommen GJB, Padberg GW, Frants RR: Chromosome 4q DNA rearrangements associated with facioscapulohumeral muscular dystrophy. Nat Genet. 1992, 2: 26-30. 10.1038/ng0992-26.View ArticlePubMedGoogle Scholar
- van Deutekom JC, Wijmenga C, van Tienhoven EA, Gruter AM, Hewitt JE, Padberg GW, van Ommen GJ, Hofker MH, Frants RR: FSHD associated DNA rearrangements are due to deletions of integral copies of a 3.2 kb tandemly repeated unit. Hum Mol Genet. 1993, 2: 2037-2042. 10.1093/hmg/2.12.2037.View ArticlePubMedGoogle Scholar
- Lemmers RJ, de Kievit P, Sandkuijl L, Padberg GW, van Ommen GJ, Frants RR, van der Maarel SM: Facioscapulohumeral muscular dystrophy is uniquely associated with one of the two variants of the 4q subtelomere. Nat Genet. 2002, 32: 235-236. 10.1038/ng999.View ArticlePubMedGoogle Scholar
- Lemmers RJ, Wohlgemuth M, Frants RR, Padberg GW, Morava E, van der Maarel SM: Contractions of D4Z4 on 4qB subtelomeres do not cause facioscapulohumeral muscular dystrophy. Am J Hum Genet. 2004, 75: 1124-1130. 10.1086/426035.PubMed CentralView ArticlePubMedGoogle Scholar
- Lemmers RJ, Wohlgemuth M, van der Gaag KJ, van der Vliet PJ, van Teijlingen CM, de Knijff P, Padberg GW, Frants RR, van der Maarel SM: Specific sequence variations within the 4q35 region are associated with facioscapulohumeral muscular dystrophy. Am J Hum Genet. 2007, 81: 884-894. 10.1086/521986.PubMed CentralView ArticlePubMedGoogle Scholar
- Lemmers RJ, van der Vliet PJ, Klooster R, Sacconi S, Camano P, Dauwerse JG, Snider L, Straasheijm KR, van Ommen GJ, Padberg GW, Miller DG, Tapscott SJ, Tawil R, Frants RR, van der Maarel SM: A unifying genetic model for facioscapulohumeral muscular dystrophy. Science. 2010, 329: 1650-1653. 10.1126/science.1189044.View ArticlePubMedGoogle Scholar
- de Greef JC, Lemmers RJ, Camano P, Day JW, Sacconi S, Dunand M, van Engelen BG, Kiuru-Enari S, Padberg GW, Rosa AL, Desnuelle C, Spuler S, Tarnopolsky M, Venance SL, Frants RR, van der Maarel SM, Tawil R: Clinical features of facioscapulohumeral muscular dystrophy 2. Neurology. 2010, 75: 1548-1554. 10.1212/WNL.0b013e3181f96175.PubMed CentralView ArticlePubMedGoogle Scholar
- van Overveld PG, Lemmers RJ, Sandkuijl LA, Enthoven L, Winokur ST, Bakels F, Padberg GW, van Ommen GJ, Frants RR, van der Maarel SM: Hypomethylation of D4Z4 in 4q-linked and non-4q-linked facioscapulohumeral muscular dystrophy. Nat Genet. 2003, 35: 315-317. 10.1038/ng1262.View ArticlePubMedGoogle Scholar
- Lemmers RJ, van der Vliet PJ, van der Gaag KJ, Zuniga S, Frants RR, de Knijff P, van der Maarel SM: Worldwide population analysis of the 4q and 10q subtelomeres identifies only four discrete interchromosomal sequence transfers in human evolution. Am J Hum Genet. 2010, 86: 364-377. 10.1016/j.ajhg.2010.01.035.PubMed CentralView ArticlePubMedGoogle Scholar
- de Greef JC, Lemmers RJ, van Engelen BG, Sacconi S, Venance SL, Frants RR, Tawil R, van der Maarel SM: Common epigenetic changes of D4Z4 in contraction-dependent and contraction-independent FSHD. Hum Mutat. 2009, 30: 1449-1459. 10.1002/humu.21091.View ArticlePubMedGoogle Scholar
- van Overveld PG, Enthoven L, Ricci E, Rossi M, Felicetti L, Jeanpierre M, Winokur ST, Frants RR, Padberg GW, van der Maarel SM: Variable hypomethylation of D4Z4 in facioscapulohumeral muscular dystrophy. Ann Neurol. 2005, 58: 569-576. 10.1002/ana.20625.View ArticlePubMedGoogle Scholar
- Lemmers RJ, Tawil R, Petek LM, Balog J, Block GJ, Santen GW, Amell AM, van der Vliet PJ, Almomani R, Straasheijm KR, Krom YD, Klooster R, Sun Y, den Dunnen JT, Helmer Q, Donlin-Smith CM, Padberg GW, van Engelen BG, de Greef JC, Aartsma-Rus AM, Frants RR, de Visser M, Desnuelle C, Sacconi S, Filippova GN, Bakker B, Bamshad MJ, Tapscott SJ, Miller DG, van der Maarel SM: Digenic inheritance of an SMCHD1 mutation and an FSHD-permissive D4Z4 allele causes facioscapulohumeral muscular dystrophy type 2. Nat Genet. 2012, 44: 1370-1374. 10.1038/ng.2454.PubMed CentralView ArticlePubMedGoogle Scholar
- Lemmers RJL, de Kievit P, van Geel M, van der Wielen MJ, Bakker E, Padberg GW, Frants RR, van der Maarel SM: Complete allele information in the diagnosis of facioscapulohumeral muscular dystrophy by triple DNA analysis. Ann Neurol. 2001, 50: 816-819. 10.1002/ana.10057.View ArticlePubMedGoogle Scholar
- Wijmenga C, Frants RR, Hewitt JE, van Deutekom JC, van Geel M, Wright TJ, Padberg GW, Hofker MH, van Ommen GJ: Molecular genetics of facioscapulohumeral muscular dystrophy. Neuromuscul Disord. 1993, 3: 487-491. 10.1016/0960-8966(93)90102-P.View ArticlePubMedGoogle Scholar
- Nguyen K, Walrafen P, Bernard R, Attarian S, Chaix C, Vovan C, Renard E, Dufrane N, Pouget J, Vannier A, Bensimon A, Lévy N: Molecular combing reveals allelic combinations in facioscapulohumeral dystrophy. Ann Neurol. 2011, 70: 627-633. 10.1002/ana.22513.View ArticlePubMedGoogle Scholar
- Lemmers RJ, van der Maarel SM, van Deutekom JC, van der Wielen MJ, Deidda G, Dauwerse HG, Hewitt J, Hofker M, Bakker E, Padberg GW, Frants RR: Inter- and intrachromosomal sub-telomeric rearrangements on 4q35: implications for facioscapulohumeral muscular dystrophy (FSHD) aetiology and diagnosis. Hum Mol Genet. 1998, 7: 1207-1214. 10.1093/hmg/7.8.1207.View ArticlePubMedGoogle Scholar
- Lemmers RJ, van der Wielen MJ, Bakker E, Padberg GW, Frants RR, van der Maarel SM: Somatic mosaicism in FSHD often goes undetected. Ann Neurol. 2004, 55: 845-850. 10.1002/ana.20106.View ArticlePubMedGoogle Scholar
- Mitsuhashi S, Boyden SE, Estrella EA, Jones TI, Rahimov F, Yu TW, Darras BT, Amato AA, Folkerth RD, Jones PL, Kunkel LM, Kang PB: Exome sequencing identifies a novel SMCHD1 mutation in facioscapulohumeral muscular dystrophy 2. Neuromuscul Disord. 2013, 23: 975-980. 10.1016/j.nmd.2013.08.009.View ArticlePubMedGoogle Scholar
- Winston J, Duerden L, Mort M, Frayling IM, Rogers MT, Upadhyaya M: Identification of two novel SMCHD1 sequence variants in families with FSHD-like muscular dystrophy. Eur J Hum Genet. 2014, In pressGoogle Scholar
- Hartweck LM, Anderson LJ, Lemmers RJ, Dandapat A, Toso EA, Dalton JC, Tawil R, Day JW, van der Maarel SM, Kyba M: A focal domain of extreme demethylation within D4Z4 in FSHD2. Neurology. 2013, 80: 392-399. 10.1212/WNL.0b013e31827f075c.PubMed CentralView ArticlePubMedGoogle Scholar
- Zeng W, Chen YY, Newkirk DA, Wu B, Balog J, Kong X, Ball AR, Zanotti S, Tawil R, Hashimoto N, Mortazavi A, van der Maarel SM, Yokomori K: Genetic and epigenetic characteristics of FSHD-associated 4q and 10q D4Z4 that are distinct from non-4q/10q D4Z4 homologs. Hum Mutat. 2014, 35: 998-1010.PubMed CentralView ArticlePubMedGoogle Scholar
- Ottaviani A, Schluth-Bolard C, Gilson E, Magdinier F: D4Z4 as a prototype of CTCF and lamins-dependent insulator in human cells. Nucleus. 2010, 1: 30-36. 10.4161/nucl.1.1.10799.PubMed CentralView ArticlePubMedGoogle Scholar
- Wu HC, Wang Q, Chung WK, Andrulis IL, Daly MB, John EM, Keegan TH, Knight J, Bradbury AR, Kappil MA, Gurvich I, Santella RM, Terry MB: Correlation of DNA methylation levels in blood and saliva DNA in young girls of the LEGACY study. Epigenetics. 2014, 9 (7): 929-933. 10.4161/epi.28902.PubMed CentralView ArticlePubMedGoogle Scholar
- Himeda CL, Debarnot C, Homma S, Beermann ML, Miller JB, Jones PL, Jones TI: Myogenic enhancers regulate expression of the facioscapulohumeral muscular dystrophy associated DUX4 gene. Mol Cell Biol. 2014, 34: 1942-1955. 10.1128/MCB.00149-14.PubMed CentralView ArticlePubMedGoogle Scholar
- Gaillard MC, Roche S, Dion C, Tasmadjian A, Bouget G, Salort-Campana E, Vovan C, Chaix C, Broucqsault N, Morere J, Puppo F, Bartoli M, Levy N, Bernard R, Attarian S, Nguyen K, Magdinier F: Differential DNA methylation of the D4Z4 repeat in patients with FSHD and asymptomatic carriers. Neurology. 2014, 83 (8): 733-742. 10.1212/WNL.0000000000000708.View ArticlePubMedGoogle Scholar
- Homma S, Chen JC, Rahimov F, Beermann ML, Hanger K, Bibat GM, Wagner KR, Kunkel LM, Emerson CP, Miller JB: A unique library of myogenic cells from facioscapulohumeral muscular dystrophy subjects and unaffected relatives: family, disease and cell function. Eur J Hum Genet. 2012, 20: 404-410. 10.1038/ejhg.2011.213.PubMed CentralView ArticlePubMedGoogle Scholar
- Rohde C, Zhang Y, Reinhardt R, Jeltsch A: BISMA–fast and accurate bisulfite sequencing data analysis of individual clones from unique and repetitive sequences. BMC Bioinformatics. 2010, 11: 230-10.1186/1471-2105-11-230.PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.