Joint hypermobility is a common, mostly benign, finding in the general population. In a subset of individuals, however, it causes a range of clinical problems, mainly affecting the musculoskeletal system. Joint hypermobility often appears as a familial trait and is shared by several heritable connective tissue disorders, including the hypermobility subtype of the Ehlers-Danlos syndrome (EDS-HT) or benign joint hypermobility syndrome (BJHS). These hereditary conditions provide unique models for the study of the genetic basis of joint hypermobility. Nevertheless, these studies are largely hampered by the great variability in clinical presentation and the often vague mode of inheritance in many families. Here, we performed a genome-wide linkage scan in a unique three-generation family with an autosomal dominant EDS-HT phenotype and identified a linkage interval on chromosome 8p22-8p21.1, with a maximum two-point LOD score of 4.73. Subsequent whole exome sequencing revealed the presence of a unique missense variant in the
Joint hypermobility is a common finding in the general population with epidemiologic studies showing its presence in over 10% of Caucasians. It is more prevalent among Asians and Africans than Caucasians, with women affected more often than men, and it is usually maximal at birth, decreasing with age [
Families in which the EDS-HT or BJHS phenotype is transmitted in a clear autosomal dominant fashion provide unique genetic models for the study of genes and molecular pathways that are involved in the pathogenesis of joint hypermobility. However, only a limited number of clinically well-defined large families with EDS-HT or BJHS are available for informative genetic linkage studies. In the past, sporadic defects have been identified in the genes encoding fibrillar types I, III, and V collagen, but in the large majority of EDS-HT patients, molecular studies have not revealed causal defects in any of these genes. Ultrastructural studies in patients with EDS-HT have revealed abnormalities in both morphology and diameter of the collagen fibrils, suggesting that impaired collagen fibrillogenesis plays a central role in its pathogenesis [
We present the identification of a new genetic locus for EDS-HT in a large three-generation Belgian family through genome-wide linkage analysis. In addition to sequencing a number of interesting positional candidate genes, we applied whole exome sequencing for two affected individuals in an attempt to unravel the underlying genetic cause within this candidate locus.
We examined a large three-generation family from Belgian descent with EDS-HT. Affected family members presented generalized joint hypermobility with related musculoskeletal problems, in association with a mild dermal phenotype, including soft skin, mild atrophic scarring, and easy bruising. Thirty-four family members were examined after full informed consent was obtained in accordance with requirements of the Ethics Committee of the Ghent University Hospital. Clinical history was taken from all individuals and clinical examination was performed at two different time points (time 0 and after 24 months) by the corresponding author. In addition, another clinical geneticist examined all individuals independently. Individuals were classified as being “affected” if they fulfilled the Villefranche criteria for EDS-HT [
Clinical characteristics of affected and unaffected family members.
Age |
Beighton score (/9) | Arthralgia | (Sub) luxation | Cutaneous | Tendonitis, muscle cramps | Other | Status | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
CMPS >4 joints | CMPS 1–3 joints | Easy bruising | Soft skin | Skin |
Atrophic scarring/dilated scars | Striae | |||||||
I.2 | 75 | 5 | + | Hip, TMJ, and sternoclavicular joint | + | + | − | + | − | Hernia umbilicalis, postoperative haemorrhages, and fragility of internal organs | A | ||
|
|||||||||||||
II.1 | 47 | 2 | + | Ankles | − | − | − | + | + | − | U | ||
|
|||||||||||||
II.2 | 46 | 0 | − | − | − | − | − | − | − | − | NA | ||
|
|||||||||||||
II.3 | 45 | 6 | + | + | + | − | − | + | + | Osteopenia, keratoconus, and muscle ruptures | A | ||
|
|||||||||||||
II.4 | 44 | 5 | + | Fingers, TMJ | + | + | + | + | + | + | Discus hernia | A | |
|
|||||||||||||
II.5 | 43 | 9 | + | Generalized | + | + | − | + | + | + | A | ||
|
|||||||||||||
II.6 | 42 | 5 | + | Generalized | + | + | − | − | + | + | A | ||
|
|||||||||||||
II.7 | 36 | 3 | + | Shoulders, wrists, and patella | + | + | + | + | + | + | A | ||
|
|||||||||||||
II.8 | 34 | 2 | − | − | +/− | − | − | + | − | U | |||
|
|||||||||||||
II.9 | 32 | 3 | − | − | Shoulder | − | − | − | ++ | + | − | U | |
|
|||||||||||||
III.1 | 24 | 0 | − | − | − | − | − | − | − | − | NA | ||
|
|||||||||||||
III.2 | 19 | 0 | − | − | − | − | − | − | − | − | NA | ||
|
|||||||||||||
III.3 | 19 | 0 | − | − | − | − | − | − | − | − | NA | ||
|
|||||||||||||
III.4 | 17 | 2 | − | + | Patella | + | − | − | − | − | − | U | |
|
|||||||||||||
III.5 | 15 | 2 | − | − | − | − | − | − | − | − | NA | ||
|
|||||||||||||
III.6 | 14 | 7 | − | Fingers, ankles | + | + | + | + | − | − | A | ||
|
|||||||||||||
III.7 | 10 | 2 | − | − | − | − | − | − | − | NA | |||
|
|||||||||||||
III.8 | 9 | 9 | + | Wrist, toes | + | + | + | − | − | − | Transparent skin | A | |
|
|||||||||||||
III.9 | 6 | 2 | − | − | − | − | − | − | − | NA | |||
|
|||||||||||||
III.10 | 18 | 9 | + | Fingers, patella | + | + | − | + | + | + | A | ||
|
|||||||||||||
III.11 | 11 | 6 | + | + | + | − | − | − | − | A | |||
|
|||||||||||||
III.12 | 9 | 7 | + | + | + | − | − | − | − | A | |||
|
|||||||||||||
III.13 | 7 | 2 | − | − | +/− | +/− | − | − | − | − | U | ||
|
|||||||||||||
III.14 | 11 | 9 | + | Ankles | + | + | + | − | − | − | A | ||
|
|||||||||||||
III.15 | 11 | 2 | − | − | − | − | − | − | − | NA | |||
|
|||||||||||||
III.16 | 7 | 6 | − | − | − | + | − | + | − | − | A |
CMSP: chronic musculoskeletal pain for >3 months.
TMJ: temporomandibular joint.
A: “affected,” NA: “not affected,” and U: “unknown.”
Blood and/or skin biopsy specimens were obtained and RNA and genomic DNA (gDNA) were isolated. Total RNA was isolated from cultured skin fibroblasts by Trizol and treated with RNase-free DNase (Life Technologies Europe, Ghent, Belgium). For the conversion to cDNA, Moloney murine leukemia virus reverse transcriptase was used in combination with random hexanucleotide primers (Invitrogen, Life Technologies Europe). gDNA was isolated from fibroblast cultures by the Easy-DNA kit (Invitrogen) or extracted from blood samples using the Qiaquick kit (Qiagen, Hilden, Germany).
Fibroblast cultures were established from a skin biopsy from the proband (II.5) and an affected sister (II.6). At confluency, cells were labelled with 14C-proline as described earlier [
As an initial step towards defining the genetic basis of EDS-HT in this family, linkage to a number of candidate genes, including the genes encoding types I, III, and V collagen and tenascin-X (
Pedigree of the three-generation Belgian family. The genotypes of 5 markers within and surrounding the linked region on chromosome 8 are depicted. The haplotype cosegregating with the EDS-HT phenotype is indicated with a black bar. An arrow indicates the proband. Individuals included in the initial genome-wide linkage screen are indicated with an asterisk. Affected and unaffected individuals are indicated with a black or white pictogram, respectively, whereas a grey pictogram with a question mark points to a patient for whom the phenotype was not clear (unknown).
A genome-wide linkage scan was performed using gDNA from the same 14 individuals (Figure
After the initial assessment of a linked chromosomal region in 14 individuals, gDNA of these individuals and of 13 additional members was further analyzed with four additional microsatellite markers. These markers were chosen from the Marshfield and Généthon genetic map for fine genetic mapping and identifying the critical interval. Order and distance between the markers were obtained from the Généthon genetic database: D8S254, 2.28 cM, D8S261, 4.51 cM, D8S258, 8.5 cM, D8S1771, 5.99 cM, and D8S1820. Haplotypes were constructed, assuming the minimal number of recombination, by tracing the segregation of alleles in the family.
The presence of a causal mutation in one of the genes associated with autosomal dominant forms of EDS was excluded by means of mutation screening of
For mutation analysis of the positional candidate genes
Gene prioritization was performed with the web-based algorithms Endeavour [
Whole exome sequencing (WES) was performed for two affected individuals (II.5 and II.6) by Aros Applied Biotechnology AS (Aarhus, Denmark). Exome capture was performed using the TruSeq Exome Enrichment kit (Illumina, San Diego, CA, USA) and sequencing was carried out on the Illumina HiSeq 2000 platform with paired-end 100-bp reads. The CLC Genomics workbench version 6.0.4 (CLCBio, Aarhus, Denmark) software was used for read mapping against the human genome reference sequence (NCBI, GRCh37/hg19) followed by duplicate read removal and coverage analysis for all regions enriched with the TruSeq Exome Enrichment kit. Single nucleotide variants and small insertions and deletions were called using the quality-based variant calling and subsequently annotated prior to exporting the resulting variant lists for filtering. The filter strategy focused on retaining heterozygous variants (variant allele frequency between 25% and 75%) shared between both sisters located within the coding sequence or flanking intronic regions (+/−20 bp) of the linkage region. Candidate variants were evaluated using the mutation interpretation software Alamut version 2.2.1 or Alamut HT version 1.1.5 (Interactive Biosoftware, Rouen, France) and segregation analysis was performed for selected variants.
In order to evaluate copy number alterations within the identified region, array comparative genomic hybridization (aCGH) was performed on individuals II.2, II.9, II.11, II.15, III.2, III.6, III.10, and III.14 (Figure
Four additional families were analyzed for linkage to the candidate locus identified in the initial family. These families were examined either by one of the authors (Anne De Paepe or Fransiska Malfait) or by clinical geneticists familiar with the disorder. All families were considered to be affected by EDS-HT, based on the presence of generalized joint hypermobility and related complications of dislocations and chronic pain, in combination with a mild dermal phenotype of soft and/or mildly hyperextensible skin, mild atrophic scarring, and easy bruising.
The index patient (Figure
After clinical evaluation, 13 individuals were scored as “affected” and 9 as “unaffected.” Affected individuals showed typical features of EDS-HT with joint hypermobility (Beighton score > 5/9), chronic musculoskeletal pain, and, especially in the adults, repetitive dislocations of one or more joints. They all presented easy bruising, and most of them had a soft skin with dilated scars. Many presented striae atrophicae, especially over the thighs and the abdomen. None of them had a marfanoid habitus. Five individuals with mild joint hypermobility and some skin features were scored as “unknown” (II.1, II.8, II.9, III.4, and III.13). A summary of the clinical data is provided in Table
SDS-PAGE of metabolically labelled types I, III, and V (pro)collagen secreted by and retained in fibroblasts of individual II.5 did not show qualitative or quantitative abnormalities in electrophoretic mobility (data not shown).
Segregation analysis for candidate genes
A systemic genome scan was performed using gDNA of 14 family members in order to identify a predisposition locus for EDS-HT in this family. In the initial genome-wide linkage search, marker D8S258 generated the highest two-point LOD score of 2.5 at theta
Two-point LOD scores for microsatellite markers on the 8p22-8p21.1 region for a given
Marker | 0 | 0.05 | 0.1 | 0.15 | 0.2 | 0.25 | 0.3 | 0.35 | 0.4 | 0.45 |
---|---|---|---|---|---|---|---|---|---|---|
D8S254 |
0.21 |
|
2.49 | 2.32 | 2.08 | 1.79 | 1.46 | 1.10 | 0.71 | 0.34 |
D8S261 |
|
1.73 | 1.59 | 1.43 | 1.27 | 1.10 | 0.91 | 0.71 | 0.49 | 0.26 |
D8S258 |
|
4.35 | 3.95 | 3.53 | 3.09 | 2.62 | 2.12 | 1.58 | 1.02 | 0.47 |
D8S1771 |
2.04 |
|
3.26 | 3.01 | 2.69 | 2.32 | 1.89 | 1.43 | 0.92 | 0.42 |
D8S1820 |
|
1.20 |
|
1.24 | 1.12 | 0.96 | 0.78 | 0.57 | 0.35 | 0.15 |
Copy number profiling was performed using custom aCGH, thereby focusing on the gene-containing regions of the candidate interval. This did not reveal deletions or duplications segregating with the disease phenotype.
Ranking of the candidate genes according to different gene prioritization tools with functionally interesting candidates indicated in bold.
Endeavour |
Suspects | G2D | GeneDistiller | ToppGene | |
---|---|---|---|---|---|
1 |
|
|
|
|
|
2 |
|
|
|
|
|
3 |
|
|
|
|
|
4 |
|
|
|
|
|
5 |
|
|
|
|
|
6 |
|
|
|
|
|
7 |
|
|
|
|
|
8 |
|
|
|
|
|
9 |
|
|
|
|
|
10 |
|
|
|
|
|
11 |
|
|
|
|
|
12 |
|
|
|
|
|
13 |
|
|
|
|
|
14 |
|
|
|
|
|
15 |
|
|
|
|
|
16 |
|
|
|
|
|
17 |
|
|
|
|
|
18 |
|
|
|
|
|
19 |
|
|
|
|
|
20 |
|
|
|
|
|
WES performed for the proband (II.5) and an affected sister (II.6) generated over 132 million reads for each, of which, respectively, 80.4% and 81.2% mapped uniquely against the human reference sequence. Approximately 93% of the enriched TruSeq target regions were covered at least 10-fold in both patients and about 90% of these regions were covered at least 20-fold. An average read depth of 93.7x was achieved for the proband (II.5) and 94.5x for her sister (II.6). Within the identified candidate region, 85.5% of the targeted regions were covered at least 20-fold in both patients whereas 3.4% and 3.6% of these targets were partially (0.6% and 0.9%) or completely (2.8% and 2.7%) uncovered for II.5 and II.6, respectively.
Single nucleotide variants and small insertions and deletions were called using the CLC Genomics workbench. Given the autosomal dominant inheritance pattern, only variants that were heterozygous (with a variant allele frequency between 25% and 75%) in both affected individuals and located within the candidate region on chromosome 8 were considered. This resulted in a total of 300 shared heterozygous variants, which were narrowed down to 11 by selecting variants that result in nonsynonymous amino acid changes, the introduction of a premature termination codon, frameshifts, or (predicted) splice sites changes with an allele frequency below 0.1 in dbSNP137. Comparison with 54 in-house exomes showed that the majority of these variants occurred multiple times in patients suffering from other inherited conditions, thereby omitting these for further analysis. Combined with the exclusion of variants that were predicted to be benign by all four
Overview of the
cDNA | Protein | SIFT | PolyPhen-2 | Align | Mutation |
---|---|---|---|---|---|
GVGD | Taster | ||||
c.633C>A | p.(His211Gln) | Tolerated | Probably damaging | C0 | Disease-causing |
c.49C>G | p.(His17Asp) | Tolerated | Probably damaging | C0 | Disease-causing |
c.1585C>T | p.(Arg529Trp) | Deleterious | Probably damaging | C0 | Disease-causing |
c.749C>A | p.(Ser250 |
— | — | — | — |
Genetic studies in EDS-HT and BJHS are scarce because they are hampered by several difficulties. First, the diagnosis in an individual is not always obvious. Since joint hypermobility is common in the general population, it is often difficult to make a clear distinction between those individuals at the upper end of the normal “joint mobility spectrum” and those whose increased mobility reflects the presence of a connective tissue disorder. For the vast majority of the EDS-HT patients, the underlying molecular defect remains unknown. Therefore, the diagnosis is entirely based on correct assessment of the clinical phenotype as no pathognomonic, radiographic, biochemical, ultrastructural, or other abnormalities are available to confirm this. The joint hypermobility may decrease or even completely disappear with increasing age. The extra-articular symptoms such as the skin hyperextensibility, abnormal scarring, or other signs of connective tissue fragility may be very subtle. Second, the inheritance pattern in a family is not always straightforward because of variable penetrance and phenotypic expression. Joint hypermobility may be influenced by specific environmental factors, such as hormonal status and degree of activity, and by different, yet to be identified, genetic factors, such as polymorphic genetic determinants in genes encoding proteins of the extracellular matrix. Large families suitable for informative linkage studies are therefore scarce.
This report presents the first successful linkage study in a unique three-generation Belgian family in which EDS-HT segregates with autosomal dominant inheritance pattern. In order to make the clinical assessment in this family as accurate as possible, the same physician (Fransiska Malfait) was responsible for repeated clinical evaluation and history taking. In addition to assessment of joint hypermobility and musculoskeletal problems, a number of clinical signs, indicative of generalized connective tissue fragility, were registered. These were helpful for determining the phenotype in individuals who were ambiguous on the joint phenotype alone. After exclusion of a number of candidate genes, a suggestive candidate locus was identified, which segregates with the EDS-HT phenotype. The highest two-point LOD score of 4.73 was obtained for marker D8S258, highly suggestive for linkage between the phenotype and this locus. A haplotype of marker alleles segregated with the phenotype and critical meiotic recombinants placed the EDS-HT predisposition gene within an 8.8 Mb region on chromosome 8p22-8p21.1. Moreover, the absence of linkage for the identified locus in several other EDS-HT families underscores the genetic heterogeneity of the disorder.
Four of the protein-coding genes located within this region were considered potentially promising candidate genes for EDS-HT.
In a next step, WES was performed for two affected sisters. The applied variant filtering strategy narrowed the number of shared heterozygous alterations down to a single missense variant in the
It should also be noted that, as the majority of disease-causing mutations reside within the protein-coding part of the genome, our study focused on the exonic part of the candidate interval, thereby omitting the interrogation of important noncoding regions, such as (deep) intronic regions, promoter sequences, and other regulatory and conserved sequences that reside within the candidate interval and which can also contain disease-causing alterations. In addition, the WES technology is also subjected to other limitations, such as the lack of a (small) proportion of genes/exons on the enrichment platform as well as the inefficient capturing and/or incomplete sequencing of some regions. Collectively, these issues can lead to false negative results and can mask the true underlying cause.
Efforts towards the identification of genes underlying joint hypermobility syndromes such as EDS-HT and BJHS are important in order to gain better insights into the clinical phenotypes, their natural history, and the underlying pathogenic basis. Families with clear autosomal dominant inheritance of EDS-HT probably represent rare forms of the condition, but elucidation of the causal gene defect in these families may allow identification of new genes and/or genetic pathways that are involved in EDS and joint hypermobility. This will improve early recognition and diagnosis of joint hypermobility and lead to a more logical classification of the hypermobility syndromes. For patients who are suffering from these chronic and painful conditions, accurate clinical diagnosis, confirmation of diagnosis by biochemical and/or molecular testing, better understanding of the pathogenic basis, and well-adjusted therapeutic follow-up by medical and paramedical staff are important factors that will help them cope with this condition and prevent feelings of frustration and depression. Understanding the genetic basis of joint hypermobility syndromes will moreover enhance our understanding of normal connective tissue biology and homeostasis and of mechanisms underlying EDS and joint mobility.
The authors declare that there is no conflict of interests regarding the publication of this paper.
The authors would like to thank the families for their cooperation. The authors are grateful to Karen Wettinck, Petra Van Acker, and Geert De Vos, for technical assistance, and Bart Loeys, for clinical evaluation. Delfien Syx and Fransiska Malfait are fellows of the Fund for Scientific Research Flanders (FWO). This work was supported by a Methusalem Grant 08/01M01108 from the Ghent University, Grant G.0171.05 from the Fund for Scientific Research (FWO), Flanders, Belgium, and a grant from Association Française des Syndromes d’Ehlers-Danlos (AFSED), all to Anne De Paepe.