Abstract
Objective: To establish the usefulness of a molecular diagnostic protocol for the autosomal dominant disease facioscapulohumeral dystrophy (FSHD).
Background: The genetic defect underlying the majority of cases is a deletion on chromosome 4q35 that is not associated with the coding sequence of any known gene. Molecular diagnosis of FSHD involves the visualization of this deletion as a “small” EcoRI restriction fragment. However, molecular diagnostics are complicated because of the homology of the telomeric regions of chromosomes 4q and 10q; the homologous 10q26 EcoRI fragments are also detected, and can fall into the size range considered to be diagnostic for FSHD. It is therefore important to distinguish the 4q35 and 10q26 EcoRI fragments, taking advantage of the presence of additional restriction sites (BlnI) in the alleles of chromosome 10q origin.
Methods: Paired digests of genomic DNA (EcoRI only and EcoRI/BlnI double digest), followed by pulsed field gel electrophoresis (PFGE), were used to establish the molecular diagnosis of FSHD in 82 unrelated index cases (46 familial, 24 proven sporadic with de novo mutations, and 12 with uncertain family history).
Results: In all cases fulfilling FSHD diagnostic criteria, a 4q35 EcoRI allele size of ≤38 kb was present. The smallest 4q35 EcoRI allele in 205 normal control subjects was 41 kb. EcoRI alleles ≤38 kb of chromosome 10q26 origin were present in 11.2% of this control group. In problematic cases, it was possible to resolve the diagnostic question.
Conclusions: The combination of double digestion with EcoRI and BlnI followed by PFGE is the most reliable molecular protocol for distinguishing patients with FSHD.
Facioscapulohumeral dystrophy (FSHD) is one of the most common muscular dystrophies, with a prevalence of \f1:20,000.1,2 Onset is variable, from infancy to middle age. Clinical signs are present in 95% of affected individuals by age 20, although many may be unaware of the symptoms. Inheritance is autosomal dominant, but sporadic cases are frequently seen. The disease is slowly progressive, and severity is variable. Major disability eventually requires a wheelchair in \f20% of patients. Typically, the face is involved first, followed by the shoulder girdle and then the pelvic muscles. The muscle involvement is often asymmetric, and progression may appear to occur in a stepwise fashion.
The full pathogenic mechanism of causation in FSHD is not known, but there has been progress in determining the genetic mechanism. Genetic linkage studies located the genetic defect to the terminal region of chromosome 4 at 4q35.3,4 Probe p13E-11, which hybridized to this region, detected small EcoRI digestion fragments in patients with FSHD.5,6 The fragments ranged in size from 14 to 28 kb, the size in normal patients being 30 to 300 kb. However, short EcoRI fragments could also occasionally be found in normal persons. In one instance, a family with FSHD and linkage to 4q35 had no small fragment segregating with the disease.7
The polymorphic EcoRI fragment detected by p13E-11 consists almost entirely of a 3.3-kb tandem repeat sequence termed D4Z4, delineated by KpnI restriction sites. The variability in fragment size was demonstrated to be caused by the deletion of an integral number of D4Z4 repeats.8 Each 3.3-kb D4Z4 repeat contains two homeodomains (i.e., DNA binding protein domains that regulate developmental gene expression), and two repetitive sequences LSau and hhspm3, but has not been shown to encode a protein.9
The p13E-11 probe has been shown to identify alleles from two distinct chromosomal loci, chromosome 4q35 and chromosome 10q 26,10 as well as a 9.5-kb fragment from the Y chromosome. The p13E11 fragments of 4q35 or 10q26 origin sometimes overlap in size, compromising the utility of this hybridization as a diagnostic test for FSHD. Subsequently, it was established that each of the 10q26 repeat sequences contain a restriction site for BlnI, while there is only a single BlnI site on the proximal portion of the chromosome 4q35 EcoRI fragments.11 The result is complete digestion of the 10q p13E11 alleles on double digestion with EcoRI and BlnI to give fragments of 3.3 kb, and a reduction in size of the 4q alleles by \f3 kb, allowing a distinction between the fragments of chromosome 4q35 and 10q26 origin, and an improvement in the diagnostic potential of this technique. Figure 1 shows the genomic organization of 4q35 and 10q26.
Figure 1. Hierarchical genetic organization of the telomeric regions of chromosomes 4q and 10q. Relative positions of the FRG1 and TUB4Q genes, D4Z4 repeats, and p13E-11 probe are shown. Normal individuals have \f12–100 copies of D4Z4 on chromosome 4. Patients with facioscapulohumeral dystrophy (FSHD) have <12 copies. B = BlnI, E = EcoRI, K = KpnI restriction sites.
Methods.
Patients with FSHD, familial and sporadic, and normal control subjects were investigated for size of the D4Z4 repeat region by paired digestion of genomic DNA (EcoRI only and EcoRI/BlnI double digest) followed by pulsed field gel electrophoresis (PFGE) and labeling with probe p13E-11.
Clinical assessment.
Established diagnostic criteria for FSHD were used (table).12
Diagnostic criteria for facioscapulohumeral dystrophy
Familial patients (n = 46) were defined as those in whom clear transmission of FSHD was documented. In these families, the propositus and one other member of the kindred were evaluated with EcoRI/BlnI double digest p13E-11 analysis. Whenever possible, the second member to be evaluated was from the preceding or succeeding generation of the kindred in relation to the propositus.
Sporadic cases (n = 24) were defined as those in which both parents of the propositus had been clinically examined and were unaffected for FSHD. In those cases, DNA from the affected offspring and both parents was evaluated by EcoRI/BlnI double digest p13E-11 analysis.
Normal control subjects (n = 205) were unaffected spouses marrying into our FSHD kindreds, and other unrelated individuals with no evidence of neurologic disease. Control subjects were evaluated with EcoRI and p13E-11. Any individual showing an allele <50 kb on hybridization with p13E-11 was evaluated with the EcoRI/BlnI double digest to determine chromosomal origin of that allele.
Uncertain inheritance (n = 12) of FSHD included cases in which the mode of heredity could not be clearly established, adoption, and instances in which obtaining clinical evaluation or DNA from both parents was not possible.
Molecular genetic investigation.
Genomic DNA was obtained from leukocytes. Five micrograms DNA was digested with 20 U of EcoRI (Boehringer, Mannheim, Germany) at 37 °C for 18 hours. In the double digests, 20 U of BlnI (Boehringer) was also added to the reaction. The digested DNA was loaded in a 1% 0.5 × TBE agarose gel for PFGE using the CHEF-DRIII system (BioRad, Hercules, CA). Whenever double digests were analyzed, samples were loaded in a pairwise fashion with the EcoRI single digest directly adjacent to the EcoRI/BlnI double digest, so that the 3-kb shift in the 4q35 alleles could be identified. High molecular weight marker (BRL; Life Technologies, Rockville, MD) and 5-kb ladder (BioRad) were used as molecular weight standards. Electrophoresis was conducted to optimize separation between 5 kb and 50 kb (initial switch time 0.5 seconds, final switch time 3 seconds, 6 V/cm, included angle 120°, run time 15 hours, and temperature 14 °C).
DNA was transferred to Hybond-N nylon membrane (Amersham, Arlington Heights, IL) by Southern blotting. Probe p13E-11 was labeled with the RTG labeling system (Pharmacia, Piscataway, NJ) using a 32P dCTP. The membranes were hybridized at 65 °C for 24 hours with p13E-1113 and washed to a stringency of 1.0 × SSC/0.1% sodium dodecyl sulfate at 65 °C. The membranes were then exposed to BiomaxMS film (Eastman Kodak, Rochester, NY) at −70 °C with intensifying screens, the films were developed, and the allele size was read.
Results.
We applied the paired EcoRI/EcoRI + BlnI digestion technique to the molecular analysis of the D4Z4 repeats in 82 unrelated index cases (46 familial, 24 truly sporadic with de novo mutations, and 12 with uncertain family history). Our final analysis included a total of 113 affected individuals, representing 31 familial pairs plus 51 unrelated singleton cases: 52 women and 61 men, age 11 to 88 years. Figure 2 shows a representative example of this analysis.
Figure 2. EcoRI, EcoRI/BlnI double digest in normal subject (A,B) and patient with facioscapulohumeral dystrophy (FSHD) (C,D). (A,C) EcoRI single digest. (B,D) EcoRI/BlnI double digest. The photographs are of pulsed field gel electrophoresis (PFGE) fragments, with sizes as indicated. Fragments >200 kb and <2 kb are not resolved. The corresponding diagrams below illustrate the chromosome 4q alleles (dark) and 10q alleles (light). The region of probe p13E-11 is indicated. With double digestion, alleles of chromosome 10 origin are no longer visible on the gel. (A) Small 38-kb EcoRI fragment that might be associated with the FSHD clinical phenotype is not present on double digest, indicating 10q origin and a normal individual. (C) Small 33-kb EcoRI fragment is reduced to 30 kb on double digest, indicating 4q origin and diagnostic of FSHD. The 103-kb normal size allele in C is fully absent on double digest, indicating 10q origin.
All patients with a clinical diagnosis of FSHD had a distinct reduction in the 3.3-kb D4Z4 repeat number, with a 4q35 EcoRI/p13E-11 allele ≤38 kb. Two hundred five normal control subjects had allele sizes ≥41 kb. Thus, there was no overlap between the FSHD diagnostic fragment size and normal alleles of chromosome 4 origin in our data set. An EcoRI/p13E-11 allele of chromosome 10 origin ≤38 kb was present in 23/205 (11.2%) of this control population. No change in fragment size was observed between generations in 31 familial pairs. In the index cases, 24/78 (31%) represented new mutations, and two affected siblings in one family with unaffected parents reflected germ line mosaicism.14 Figure 3 summarizes the molecular analysis of D4Z4 repeats in patient and control populations.
Figure 3. Size of D4Z4 containing allele in facioscapulohumeral dystrophy (FSHD) patients (C,D,E) and normal (unaffected) control subjects (A,B). With EcoRI single digestion (A), small fragments ≤38 kb are observed in normal individuals, within the range of EcoRI fragments found in patients with FSHD. This limits its utility as a diagnostic test. Double digestion with EcoRI/BlnI (B) removes the short fragments (of 10q origin); notice the clear distinction in size between the 4q fragments in FSHD patients (≤38 kb) and those in control subjects (≥45 kb). In the index cases of familial FSHD, the allele (mean = 27 kb, ± 6.0 SEM) is significantly larger than in the sporadic cases (mean = 19 kb, ± 1.7 SEM), p < 0.001 (t-test).
In one family, initial investigation with field inversion gel electrophoresis (FIGE) demonstrated small fragments of \f28 kb in both affected and unaffected family members, suggesting that FSHD was not linked to the p13E11 locus. With PFGE, the resolution was improved. On single EcoRI digest, fragments of 28 kb and 25 kb were detected in the affected mother and daughter, and a fragment of 25 kb was detected in the unaffected sibling. Linkage of FSHD to the 28-kb allele was further clarified by double digest (EcoRI + BlnI); the 28-kb fragment reduced to 25 kb, indicating chromosome 4q origin, and the 25-kb fragment was completely digested, indicating the chromosome 10q nonpathogenic origin of the short fragment in the unaffected family members. Figure 4 demonstrates clear resolution of p13E11 alleles in this family, which had comigrating chromosome 4q35 and chromosome 10q26 alleles according to earlier techniques.
Figure 4. Clarification of family with suspected recombination event. Initial investigation with field inversion gel electrophoresis (FIGE) demonstrated small fragments of \f28 kb in both affected and unaffected individuals, suggesting that facioscapulohumeral dystrophy (FSHD) was not linked to this allele. With pulsed field gel electrophoresis (PFGE), the resolution is improved, and on single EcoRI digest, fragments of 28 kb and 25 kb were detected in the mother and affected daughter, and 25 kb in the unaffected sibling. Linkage of FSHD to the 28-kb allele was further clarified by double digestion with EcoRI/BlnI. The 28-kb EcoRI fragment reduced to 25 kb, indicating chromosome 4q origin, and the 25-kb EcoRI fragment was completely digested, indicating its chromosome 10q (i.e., nonpathogenic) origin.
Discussion.
The EcoRI/BlnI double digest improves the sensitivity and specificity of this diagnostic test for FSHD. In our series, all FSHD patients studied had chromosome 4q35/p13E11 alleles ≤38 kb, and unaffected individuals, ≥41 kb. Another series, from England and Wales, suggested a sensitivity of 94.6% at a diagnostic threshold of <34 kb because of apparently affected individuals with alleles >35 kb, but specificity remained at or very near 100%.15 The smallest fragments detected in their control population of 200 were 38 ± 2 kb in 1, 40 kb in 2, and 41 kb in 1. Of the 130 unrelated FSHD patients studied, 7 had alleles >35 kb. Three of them had alleles of 38 kb, within the diagnostic size range indicated in our study; 2 had alleles of 42 kb, 1 of 48 kb, and 1 of >48 kb. In a Dutch study of FSHD index cases,16 46 of 160 individuals had EcoRI/p13E-11 fragments >35 kb, but almost all had clinically uncertain FSHD and had been referred for exclusion of the diagnosis. In the remaining 114 index cases, EcoRI digestion yielded a p13E-11 fragment <35 kb, and further double digest analysis with BlnI indicated that 102 of those 114 index patients had alleles <35 kb, which were of chromosome 4q origin. Of the 12 cases that appeared not to be of chromosome 4q origin, 2 were thought to be clinically definite FSHD, and on further analysis with PFGE, those patients all had small hybrid 4q35 fragments. Such hybrid fragments, resulting from exchanges between homologous, subtelomeric regions of 4q35 and 10q26, have been previously reported and may be as frequent as 20% in the normal population.16,17 We have observed only one instance of such an exchange, in a non–disease-associated allele.
One of our patients with an uncertain diagnosis of FSHD had the diagnosis confirmed, having a fragment size of 33 kb. He was thought clinically to have a scapuloperoneal syndrome rather than FSHD, the facial weakness being only slight. He had no clear family history of muscular dystrophy, but symptoms attributed to rupture of the long head of the biceps occurred at 30 years of age, and muscle biopsy and EMG showed myopathic features. By age 57, he had disabling weakness of the arms and legs, with inability to bury his eyelashes. Another patient with autosomal dominant scapuloperoneal syndrome was investigated because of slight asymmetry of the lips on pouting and pursing. This was originally thought to be probably of no diagnostic significance, and the absence of a 4q35 EcoRI/p13E-11 allele ≤38 kb confirmed the suspicion that his condition was not FSHD.
In the index cases of sporadic FSHD, the EcoRI/p13E-11 allele size (mean 18.8 kb ± 1.1 SEM) is significantly smaller than in familial FSHD (mean 27.1 kb ± 0.8 SEM), p < 0.001 (Student’s t-test). This difference has been observed previously15 and raises questions regarding the production of these mutations. Because of the known correlation between the size of the D4Z4 deletion and clinical severity,18,19 our results also suggest that sporadic patients with a milder form of the disease are not being clinically detected and diagnosed.
PFGE improves the resolution of fragment size and has allowed larger sizes of alleles to be recognized in FSHD. The earlier limits of 25–28 kb reflect the limits of resolution of the FIGE used. One of our families was previously reported as not showing linkage to the EcoRI/p13E-11 allele. The family showed linkage to other FSHD markers on chromosome 4q35, but small EcoRI/p13E-11 alleles of \f28 kb were demonstrated in both affected and unaffected individuals. It was suggested at the time that the EcoRI fragment might have been larger than the 28-kb resolution of the FIGE used, or it was too small to be detected, or the p13E-11 sequence was deleted.7 On PFGE, the resolution improved, and hence it has been possible to confirm linkage of FSHD with a 28-kb EcoRI digestion fragment as well as to identify a 25-kb fragment that was completely digested with EcoRI + BlnI, indicating its 10q, nonpathogenic origin in the unaffected family members. In our patient series, molecular diagnosis was not complicated by confounders such as genetic heterogeneity,20 or by complete or partial translocation of D4Z4 repeat units between chromosome 10q26 and chromosome 4q35, which occur at relatively high frequency.16 A new digestion protocol that uses the enzyme Tru9I in place of EcoRI may be useful in resolving mosaicism in the D4Z4 locus.21
Molecular testing may be indicated in a clinical setting to establish a diagnosis of FSHD. It is also appropriate in the experimental setting in order to include appropriate patients in clinical trials (for example, it is applied in the current large experimental therapeutic investigation of albuterol in FSHD),22 and hopefully in the subsequent application of therapy to patients. In investigating an affected individual, an EcoRI/p13E-11 allele of ≤38 kb, which is reduced by 3 kb on digestion with BlnI, is diagnostic. It should be possible to screen initially with EcoRI/p13E-11, using PFGE, and proceed to additional BlnI digestion only in DNA samples with alleles ≤38 kb. If the p13E11 alleles are ≥41 kb, FSHD is very unlikely, but further clinical assessment and molecular investigation may help clarify this.
Acknowledgments
Acknowledgment
This study was conducted in conjunction with the collaborative FSH-DY group of the University of Rochester (R.C. Griggs, principal investigator) and Ohio State University (J.R. Mendell, principal investigator). The authors thank Dr. B. Weiffenbach for helpful discussions, Prof. Dr. R. Frants for generously providing probe p13E-11, and all the patients with FSHD and their families who have helped with this work.
Footnotes
Dr. Orrell is currently affiliated with the Department of Clinical Neurosciences, Royal Free and University College Medical School of University College London, UK.
Funding and support provided by NIH grants NS22099, IR13NS35435, and General Clinical Research Center grant 5MO1RR00044; Muscular Dystrophy Association; New York State Education Department C960284; Wayne C. Gorell Jr Molecular Biology Laboratory, University of Rochester; and the Saunders Foundation. Dr. Orrell is a Medical Research Council (UK) Traveling Fellow.
References
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Hemiplegic Migraine Associated With PRRT2 Variations A Clinical and Genetic Study
Dr. Robert Shapiro and Dr. Amynah Pradhan