Molecular Genetic Diagnosis of a Bethlem Myopathy Family with an Autosomal-Dominant COL6A1 Mutation, as Evidenced by Exome Sequencing.

BACKGROUND: We describe herein the application of whole exome sequencing (WES) for the molecular genetic diagnosis of a large Korean family with dominantly inherited myopathy. CASE REPORT: The affected individuals presented with slowly progressive proximal weakness and ankle contracture. They were initially diagnosed with limb-girdle muscular dystrophy (LGMD) based on clinical and pathologic features. However, WES and subsequent capillary sequencing identified a pathogenic splicing-site mutation (c.1056+1G>A) in COL6A1, which was previously reported to be an underlying cause of Bethlem myopathy. After identification of the genetic cause of the disease, careful neurologic examination revealed subtle contracture of the interphalangeal joint in the affected members, which is a characteristic sign of Bethlem myopathy. Therefore, we revised the original diagnosis from LGMD to Bethlem myopathy.
CONCLUSIONS: This is the first report of identification of COL6A1-mediated Bethlem myopathy in Korea, and indicates the utility of WES for the diagnosis of muscular dystrophy.

INTRODUCTION

Muscular dystrophy is a clinically and genetically heterogeneous inherited disorder characterized by progressive muscle weakness and wasting. A step-by-step approach with assessment of medical history, clinical examination, laboratory evaluation, muscle pathology, muscle immunoanalysis, and mutational analysis is typically used for the diagnosis of muscular dystrophy.1,2 However, this serial approach often fails to identify causative mutations due to high phenotypic and pathologic variability, small pedigrees, and the limited power of traditional linkage analyses.3 Recent advances in next-generation sequencing has made it possible to selectively sequence only the protein-coding exons of the genome, a process termed 'whole exome sequencing' (WES). The application of WES not only saves time but is also cost-effective for the identification of causative genes in Mendelian diseases.4 Therefore, WES is being increasingly adopted for the identification of causative genes in muscular dystrophy research.5,6,7,8

Herein we report a mutation in the gene encoding collagen type VI α1 (COL6A1) in a large Korean family with autosomal-dominant Bethlem myopathy that was detected using WES.

CASE REPORT

Eighteen members of a large Korean family with dominantly inherited myopathy (7 affected and 11 unaffected) were enrolled (Fig. 1A). Written informed consent to participate was obtained from all participants and from the parents of participants younger than 18 years, according to a protocol approved by the Institutional Review Board for Ewha Womans University Mokdong Hospital, Seoul, Korea.

Fig. 1

A: Pedigree of an autosomal-dominant Bethlem myopathy family. Asterisks (*) indicate individuals whose DNA was used for exome sequencing. Genotypes of COL6A1 (c.1056+1G>A) are indicated under each subject (arrow, proband; square, male; circle, female; filled, affected; not filled, unaffected; diagonal bar across symbol, deceased). B: Sequencing chromatograms of the c.1056+1G>A splicing-site mutation in COL6A1. Arrow indicates the polymorphic site. The COL6A1 mutations detected by exome sequencing were confirmed by capillary sequencing. The heterozygous c.1056+1G>A mutation was completely cosegregated with the affected individuals within this family, and was not found in a sample of 200 healthy controls.

Patients

The proband, a 38-year-old woman (III-4), presented with progressive proximal weakness and ankle contractures (Supplementary Table 1 in the online-only Data Supplementary). Her initial development after birth was reportedly normal, but she did not begin walking until she was 16 months old. She recalled that she had always been weaker than her peers. Her motor function remained relatively stable until her mid-20s. However, she experienced slowly progressive muscle weakness after the delivery of her first child at an age of 27 years. Her neck flexors [Medical Research Council (MRC) grade 3] appeared to be more damaged than her neck extensor (MRC grade 4+). The proximal muscles of her upper and lower limbs were more severely involved than the distal muscles, and her ankle joints were affected by contracture. In addition, she appeared to exhibit mild facial weakness and absent tendon reflexes in the upper and lower limbs. However, sensory examination revealed no abnormalities. Laboratory studies revealed a serum creatine kinase level of 66 IU/L (normal, <135 IU/L), and her vital capacity was 3,120 mL. Electrocardiography and echocardiography findings were normal. Nerve conduction studies and needle electromyography revealed active generalized myopathy. A muscle biopsy sample obtained from the left biceps brachii revealed nonspecific muscular dystrophic changes. Hematoxylin and eosin staining revealed variation in muscle fiber size (Fig. 2A), and modified Gomori trichrome staining revealed a few ragged red fibers (Fig. 2B). Architectural changes of disorganized intermyofibrillar networks, such as lobulated fibers, were accentuated in staining with nicotinamide adenine dinucleotide tetrazolium reductase (Fig. 2C). In addition, adenosine triphosphatase (pH 9.4) staining demonstrated a 69% predominance of type I fibers (Fig. 2D). Electron microscopy revealed many mitochondria, and immunohistochemical analyses of the muscle specimens revealed normal staining patterns for the C-terminal of dystrophin, rod domain of dystrophin, N-terminal of dystrophin, dysferlin, α-sarcoglycan, β-sarcoglycan, γ-sarcoglycan, δ-sarcoglycan, α-dystroglycan, and caveolin.

Fig. 2

Histopathologic observations of biceps brachii muscle samples taken from the proband (III-4). A: Hematoxylin and eosin staining revealed variations in muscle fiber size and some fibers with internal nuclei (×200). B: Modified Gomori trichrome staining revealed a few ragged red fibers (×200). C: Staining with nicotinamide adenine dinucleotide tetrazolium reductase revealed architectural changes of disorganized intermyofibrillar networks, such as lobulated fibers (×400). D: Adenosine triphosphatase (pH 9.4) staining demonstrated a 69% predominance of type I fibers (×100).

The other affected members of the family had similar clinical presentations (Supplementary Table 1 in the online-only Data Supplementary). They experienced very slow progressive muscle weakness and lived without significant disability until old age. Subject II-2 required aid for ambulation after the age of 50 years, while subject II-3 (61 years old) was able to walk independently. The ankle joints were affected by contracture in all seven affected patients, while the elbow joints were involved only in subject II-2. Based on the clinical and pathologic features, the family was initially diagnosed with autosomal-dominant limb-girdle muscular dystrophy.

After identification of the causative gene, all affected members underwent a second neurologic examination. This identified contracture of the interphalangeal joint-which is a characteristic sign of Bethlem myopathy-in five family members (II-2, II-3, III-4, III-6, and IV-7). These contractures were very subtle and were not found on routine neurologic examination; they were only apparent when the wrist and fingers were extended passively.

Genetic analysis

Whole exome sequencing was performed for five members of the family, including four affected members (II-3, III-4, IV-7, and IV-8) and one unaffected member (IV-6) to identify the genetic causes of the disease, following the method described by Choi et al.9 The exome sequencing data are summarized in Supplementary Table 2 (in the online-only Data Supplementary). The mean total sequencing yield was 9.3 Gbp/sample, and the coverage rate of the targeted exon regions (≥10×) was 93.56%. The average read depth of the target regions was 69.3 reads, and the average number of observed variants per sample was 92,174 SNPs and 9,321 indels. By comparing the exome data between 4 affected and 1 unaffected family members, we found that 15 functionally significant cosegregated variants (Supplementary Table 3 in the online-only Data Supplementary). Subsequent capillary sequencing analysis of control samples and other family members who were not included in the exome sequencing excluded most variants as the underlying cause of myopathy. However, a c.1056+1G>A splicing-site mutation in COL6A1 completely cosegregated with affected status within the family (Fig. 1B), and was not found in 200 healthy controls. This mutation has been reported to be the underlying cause of Bethlem myopathy.10,11,12,13 Thus, we determined that the c.1056+ 1G>A mutation in COL6A1 was the underlying cause of the disease in this family.

DISCUSSION

Whole exome sequencing of five members from a single family identified a splice donor site mutation at c.1056+1G>A of COL6A1. This mutation causes the formation of abnormal collagen VI protein by skipping of exon 14 and consequent in-frame deletion of amino acids from the triple helical domain of the α1 chain.12

Bethlem myopathy is a dominantly inherited myopathy caused by mutations in one of three genes encoding collagen type VI alpha (COL6A1, COL6A2, and COL6A3).14 The phenotype is characterized by slowly progressive proximal weakness and multiple contractures. Prominent contracture in the early stages of the disease is one of the most important clinical features in Bethlem myopathy, Emery-Dreifuss muscular dystrophy, and Ullrich congenital muscular dystrophy. Among these conditions, Bethlem myopathy demonstrates the most benign clinical course and mildest contractures.

Bethlem myopathy is often difficult to diagnose and its frequency may be underestimated for several reasons. First, mild contractures often lead to confusion in the diagnosis.15,16 In the present family, even though ankle contracture was initially detected, this is a common nonspecific finding in many other neuromuscular diseases. Contracture of the interphalangeal joint is a hallmark of Bethlem myopathy, but is often so subtle that it goes unrecognized. Second, muscle biopsy is not typically used for confirmatory diagnosis of Bethlem myopathy due to nonspecific myopathic changes and lack of detected abnormalities of collagen VI, even in immunohistochemical analyses. Both Ullrich congenital muscular dystrophy and Bethlem myopathy are collagen-IV-related myopathies. Immunohistochemistry in Ullrich congenital muscular dystrophy exhibits complete absence or unequivocal reduction of collagen VI compared to normal; it can thus be used for diagnostic purposes. However, in Bethlem myopathy, immunostaining of muscle biopsy with various collagen VI antibodies is usually normal.14 Third, identification of the causative genes by general sequencing is costly and time-consuming because it is necessary to screen all 107 exons in all 3 genes for molecular genetic diagnosis. For these reasons, careful clinical assessment and cost-effective, time-saving strategies for genetic analysis are important for the diagnosis of Bethlem myopathy.

Whole exome sequencing is a well-justified strategy for discovering the causative genes of muscular dystrophy. WES is based on next-generation sequencing, which reduces the cost and time relative to Sanger sequencing.17 In addition, WES focuses only on protein-coding regions, but it is still an effective diagnostic tool because more than 90% of the pathogenic mutations for Mendelian disorders are found in exons.4

In conclusion, we identified a COL6A1 mutation in a Korean family with Bethlem myopathy; this is the first such report in Korea. Even though the causative mutation identified in the present study has been reported previously, this work underscores the usefulness of WES for the diagnosis of muscular dystrophy.

Supplementary Table 1

Clinical and laboratory characteristics of affected individuals

Characteristics	Patients
	II-2	II-3	III-4	III-6	IV-7	IV-8	IV-9
Sex/age (yr)	F/65	M/61	F/38	F/33	F/10	F/8	F/6
Age of onset	Infant*	Infant*	Infant*	Infant*	Infant*	Infant*	Infant*
Facial weakness	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Weakness, manual testing (MRC)
Neck
Flexor	2	3	3	4-	4+	4+	4+
Extensor	4-	4+	4+	4+	4+	4+	4+
Shoulder
Abduction	3	4+	4	4	4+	4	4
Elbow
Flexion	4-	4+	4	4+	4+	4	4
Extension	4-	4+	4	4+	4+	4	4
Wrist
Flexion	4	4+	4	4+	4+	4	4
Extension	4	4+	4	4+	4+	4	4
Hip
Flexion	2	4-	4-	4-	4+	4-	4-
Abduction	2	4-	4-	4-	4+	4-	4-
Knee
Flexion	3	4	4	4	4+	4	4
Extension	3	4	4-	4-	4+	4-	4-
Ankle
Dorsiflexion	2	4	4-	3	4+	4	4
Plantarflexion	4	4	4+	4+	4+	5	4
Reflexes	Absent	Absent	Absent	Absent	Decreased	Decreased	Decreased
Contractures
Ankle joints	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Interphalangeal joints	Yes	Yes	Yes	Yes	Yes	No	No
Elbow joints	Yes	No	No	No	No	No	No
Independent ambulation	No	Yes	Yes	Yes	Yes	Yes	Yes
Scoliosis	No	No	No	No	No	No	No
Serum CK (x-fold the upper limit 0.4 of normal values)	0.4	ND	0.7	ND	2.3	2.6	ND

*All affected members showed delayed walking, at about 16 months of age.

CK: creatine kinase, MRC: Medical Research Council, ND: not determined.

Supplementary Table 2

Symptoms and treatments administered during butter oral food challenges

Items	II-3 (affected)	III-4 (affected)	IV-7 (affected)	IV-8 (affected)	IV-6 (unaffected)
Total sequencing yields (Gbp)	10.0	8.1	10.6	9.7	8.1
Target coverage (≥10×)	93.4%	93.5%	93.9%	93.4%	93.6%
Rate of mappable reads	99.6%	99.6%	99.0%	99.6%	99.2%
Mean read depth of target region	76.9	61.2	75.2	76.6	56.7
Total observed SNPs	90,706	92,910	92,970	92,720	91,566
Total observed indels	9,122	9,477	9,536	9,196	9,273
Filtering
Coding SNPs	20,436	20,947	20,832	20,842	20,630
Functionally significant SNPs*	9,397	9,635	10,780	9,592	9,462
Unreported variants†	316	459	443	482	459
Cosegregating variants‡	17

*Functionally significant variants include nonsynonymous, splicing site, frameshift, stop gain, stop loss and coding indels, †Functionally significant variants that were not reported in dbSNP135 and 1,000 g databases, ‡Nonsynonymous variants that are present in four affected samples, but not in an unaffected sample.

SNP: single nucleotide polymorphism.

Supplementary Table 3

Nonsynonymous variants present in four affected individuals (II-3, III-4, IV-7, and IV-8), but not in one unaffected individual (IV-6)

Position* (Chr: position)	Gene	RefSeq†	Nucleotide change‡	Amino acid change	dbSNP137	1,000 Genome (Oct, 2011)	Description
5: 133473771	TCF7	NM_003202.3	c.463C>A	H155N	-		Non-cosegregation
5: 140735861	PCDHGA4	NM_032053.1	c.1094delT	V365fs	-		Polymorphic
9: 139235485	GPSM1	NM_015597.4	c.1242delC	P414fs	rs145729152	0.1433	Polymorphic
10: 50535007	C10orf71	NM_199459.3	c.2106_2107ins	H702fs	rs66701434		Polymorphic
11: 46724722	ZNF408	NM_024741.2	c.581_592del	non-fs	rs77997634		Polymorphic
13: 108518686	FAM155A	NM_001080396.2	c.c217_259sub	non-fs	-		Polymorphic
14: 93154552	RIN3	NM_024832.3	c.2913_2915del	non-fs	-		Polymorphic
15: 51242102	AP4E1	NM_001252127.1	c.1171A>T	I391F	-		Non-cosegregation
15: 51397361	TNFAIP8L3	NM_207381.2	c.13C>T	R5W	rs201297277		Non-cosegregation
15: 57730706	CGNL1	NM_001252335.1	c.509C>T	S170F	-		Non-cosegregation
15: 75648553	MAN2C1	NM_006715.3	c.2894G>C	S965T	-		Non-cosegregation
16: 1535951	PTX4	NM_001013658.1	c.1411C>T	R471C	-		Polymorphic
21: 41137642	IGSF5	NM_001080444.1	c.281C>G	T94S	-		Cosegregation, polymorphic
21: 46393132	FAM207A	NM_058190.2	c.521G>C	R174P	-		Cosegregation, polymorphic
21: 47410741	COL6A1	NM_001848.2	c.1056+1G>A	Splicing	-		Cosegregation, causative

*Reference nucleotide position: The University of California, Santa Curz assembly hg19, †GenBank registration number of reference sequence, ‡Complementary DNA numbering was achieved with +1 corresponding to the A of the ATG initiation codon.