Autosomal recessive Bethlem myopathy: a 30-year journey

Dear Sirs,

Collagen VI-related myopathies (COL6-RMs) represent a wide spectrum of muscle dystrophies ranging from the more severe Ullrich Congenital Muscular Dystrophy (UCMD) and the milder Bethlem myopathy (BM), including phenotypes of intermediate severity [1, 2].

UCMD usually presents at birth or within the first few months of life with hypotonia and proximal muscular weakness associated with distal joint laxity [2]; contractures and skeletal deformities are often seen in more proximal joints. Commonly, there is a delay in the acquisition of motor skills and a loss of independent walking in the second decade of life, although more severe patients never acquire independent walking. A progressive decline in pulmonary function occurs early in disease course, quite typically when the patient is still able to walk independently. Conversely, BM is at the mild end of the spectrum. The clinical features are not much different from those seen in UCMD, but the onset is during childhood, the symptoms are milder, and the disease progression is slower [1]. Patients usually acquire independent walking, but the combination of muscle weakness and contractures may require the use of walking aids after 50 years of age. The occurrence of respiratory insufficiency is variable [2].

The COL6-RMs can be caused by dominant and recessive autosomal mutations in the three collagen VI genes (COL6A1, COL6A2, and COL6A3) [2, 3]. Collagen VI is an extracellular matrix protein that associates closely with the basement membrane of skeletal muscles and links it to the surrounding extracellular matrix. This protein is made up of three collagen VI chains, α1(VI), α2(VI), and α3(VI) encoded by the COL6A1, COL6A2, and COL6A3 genes, respectively. These three collagen VI chains form a monomer constituted of two globular regions linked by a triple-helix domain (THD) that consists of Gly-Xaa-Yaa amino acid repeat sequences. Subsequently, the monomers assemble into dimers and then into tetramers, which associate in an end-to-end fashion to form the final microfilament network [3].

The correct folding of collagen VI is critical and mutations that counteract this process influence on the clinical severity of the disease. For instance, BM is mostly caused by autosomal dominant mutation, found in the N-terminal end of the THD in close proximity to single cysteine residues that are responsible for the correct collagen VI assembly. Nevertheless, autosomal recessive inheritance has been reported in less severe cases, clinically defined, as BM [1]. UCMD and intermediate COL6-RMs are both recessively as well as dominantly. Moreover, it is now evident that de novo dominant mutations in all three collagen VI genes are responsible for a notable number of sporadic UCMD cases [2]. In autosomal recessive cases, often the mutations involve the C-terminal of THD or the C-terminal domain, which disrupt the initial formation of monomers preventing the assembly process [2, 4].

The rarity of these diseases and the substantial variability of genetic transmission make the diagnosis challenging. We reported clinical and genetic findings in a family with three members affected by autosomal recessive COL6A2 muscular dystrophy, describing all the pitfalls that complicated the achievement of the correct diagnosis.

The proband (III:2) was visited for the first time in our neuromuscular center at the age of 24 years (Fig. 1A). She reported a slowly progressive muscular weakness, mostly involving lower limbs, started some years before. She had been already visited in a peripheral hospital and then referred to our Institute. An accurate collection of the medical history revealed a disease presentation at around 6 years of age with weakness of pelvic girdle muscles; her parents, indeed, reported that she was not able to run like her peers at school and she showed clumsiness in climbing stairs or standing up from the floor. No delay in motor milestone achievement was reported. Neurological examination showed a diffuse weakness, prominent in pelvic, distal upper limb and axial muscles, with sparing of the cranial district. The patient was able to stand up from a chair without using her hands, although needing a broad anteflexion of the trunk. Gowers’s sign was positive. No respiratory involvement was reported. Scoliosis, joint contractures, or distal joints hyperlaxity were absent. Skin examination did not reveal the presence of keloids. Serum creatine kinase (CK) was in the normal range.

Fig. 1.

Family pedigree and electropherograms. A Family pedigree with black symbols indicating the affected members. The arrow indicates the proband. B Electropherograms in the proband, in the mother and the father. The red arrows indicate the variants

According to her family history (Fig. 1A), two sisters and one brother were asymptomatic. The remaining two sisters (III:4 and III:6) of 20 and 14 years of age, respectively, showed a similar but milder phenotype, with a sparing of proximal muscles of the lower limbs. The disease onset was during childhood (III:6) or later in the adolescence (III:4), characterized in both cases by difficulties in running, climbing stairs and rising from the floor. Serum CK was mildly increased (300 mg/dL) only in the younger sister (III:6). The father was asymptomatic. The mother (II:2), 48 years of age, was affected by a Myasthenia Gravis associated with anti-acetylcholine receptor antibody and confirmed by the repetitive nerve stimulation test; she was administered with pyridostigmine and prednisone. The disease onset had been approximately 20 years before the consultation in our center, characterized by muscular fatigability. Neurological examination revealed a mild weakness of tongue and cranial muscles (orbicularis oris and orbicularis oculi), diplopia after fixation and a mild weakness of axial muscles and scapular and pelvic girdles. Considering some similarities with the neuromuscular examination of the daughters, the mother’s family history was investigated: she reported an asymptomatic brother, but, unfortunately, she could not provide clinical information about her parents.

All symptomatic members of the family underwent a muscle biopsy that showed a variability of fiber diameter, some internalized nuclei, and rare fiber splittings for III:2, III:4, and III:6 (Fig. 2A). The alterations found in the mother’s muscular biopsy were mild, compatible with age and a long-term myasthenia gravis.

Fig. 2.

Immunofluorescence staining. A NADH and H&E staining in the proband’s muscle tissue (10X). B Skin fibroblasts treatment with ascorbic acid. The proband’s cells (III:2) presented the absence of secreted protein and collagen VI network; collagen staining was observable only in the cell’s cytoplasm as in NT well (63X). In her mother (II:2) and in the asymptomatic brother (III:3), a collagen VI fine network was visible in the extracellular matrix

Muscle computed tomography showed hypotrophy of the scapular and thigh muscles, associated with fatty infiltration of paraspinal muscles (III:2, III:4, III:6), hamstrings (III:2, III:4), and quadriceps femoris (III:2). Immunohistochemical analyses on muscle tissues in all affected members resulted normal, except for slight nonspecific reduction of α-dystroglycan levels in III:2. Therefore, we proceeded with Sanger sequencing for some of the dystroglycanopathies-related genes (as well as FKRP and ISPD), without finding pathogenic/likely pathogenic variants.

Next-Generation Sequencing (NGS) custom panel was then performed on the proband and on one of her affected sisters (III:6). This panel consisted of 40 amplicon-based targeted genes known to be associated with late-onset muscular dystrophies, such as limb girdle dystrophies, and distal and proximal myopathies. NGS panel did not identify pathogenic mutations, but only three variants of unknown significance (VUS), shared by the two sisters. Specifically, a variant was found in FLNC gene and two in SBF1 gene. Mutations in FLNC gene are usually associated with distal/myofibrillar myopathies with severe cardiomyopathy, with clinical and histological features different from those detected in our patients. Pathogenicity of the VUS in SBF1 gene was instead excluded by the segregation analysis that found both in a healthy family member (III:5).

Given the inconclusive result of the previous investigations, carried out between the early 1990s and 2012, a whole exome sequencing was performed for the patients III:2, III:4, and III:6, after 20 years from the proband’s first neurological examination in our neuromuscular center. This analysis highlighted several common VUS in the three affected sisters. Among these, two missense variants in the COL6A2 gene were investigated. The first variant c.1514G > A, classified by in silico predictors (such as SIFT, Mutation Taster, MutPred, etc.) as pathogenic, causes the substitution of a glycine with aspartate in position 505, involving a highly preserved residue belonging to a protein’s domain essential for the correct dimer and tetramer assembly. The second variant c.2877G > C leads to the substitution of glutamic acid in position 959 with an aspartate; this substitution is predicted to be tolerated despite being in a fairly preserved residue. The segregation analysis showed that the mother (II:2) was a carrier of the VUS involving the 505 residue, while the father (II:3) was a carrier of the VUS affecting the 959 residue (Fig. 1B). The asymptomatic siblings had none of the two variants or only the one of paternal origin.

As previously described, mutations in the genes coding for collagen VI α chains interfere with tetramers’ formation, causing a reduction or a lack of microfilament network expression in the extracellular matrix. This can be evaluated treating the patient’s cultured skin fibroblasts with ascorbic acid to allow hydroxylation and secretion of collagen molecules. Cultured fibroblasts from our proband (III:2) presented the absence of secreted protein and collagen VI network; collagen staining was observable only in the cell’s cytoplasm. Conversely, in the mother (II:2) and in her asymptomatic brother (III:3), a collagen VI fine network was visible in the extracellular matrix (Fig. 2B).

These findings confirm the autosomal recessive inheritance of this COL6-related disease, associated with a relatively mild phenotype, therefore consistent with a BM. The mother (II:2) was thus not affected by the same disease of her daughters, being only a carrier of one of the COL6A2 gene variants. Clinical and respiratory data of the three patients at the last follow-up are reported in Table 1.

Table 1.

Clinical and respiratory data of the affected patients, proband (III:2) and two sisters (III:4 and III:6), at the last neurological follow-up

	III:2	III:4	III:6
Age of disease onset (y)	6	13	9
Age at last visit (y)	55	51	45
Genotype	c.1514G > Ap.Gly505Asp c.2877G > Cp.Glu959Asp	c.1514G > Ap.Gly505Asp c.2877G > Cp.Glu959Asp	c.1514G > Ap.Gly505Asp c.2877G > Cp.Glu959Asp
Cranial/bulbar involvement	None	None	None
Muscle strength testing	HF: MRC 3/5 PUL: MRC 3/5 DUL: MRC 3/5 PLL: MRC 3/5 DLL: MRC 3/5	HF: MRC 3/5 PUL: MRC 4/5 DUL: MRC 3/5 PLL: MRC 4/5 DLL: MRC 4/5	HF: MRC 2/5 PUL: MRC 4/5 DUL: MRC 3/5 PLL: MRC 3/5 DLL: MRC 4/5
Independent walking	Yes	Yes	Yes
Stair climbing	Handrail support	Handrail support	Handrail support
Get up from the chair	Using hands	Without support	Without support
Able to run	No	No	No
WGM score	3/10	3/10	3/10
Joint contractures	None	None	None
Distal joints hyperlaxity	Elbows (mild)	None	None
Keloids	None	None	None
FVC	98%	107%	97%

HF head flexors muscles, PUL proximal upper limbs muscles, DUL distal upper limbs muscles, PLL proximal lower limbs muscles, DLL distal lower limbs muscles, WGM Walton Gardner-Medwin scale, FVC forced vital capacity

COL6A2-related muscular dystrophies are characterized by a variable pattern of muscular weakness. Some clinical signs like joint contractures and distal joints hyperlaxity or early respiratory involvement can help the diagnosis. Muscular imaging can show a unique pattern of fibro-fatty substitution at thigh level, defined as outside-in progression [5]. Conversely, histological findings can vary widely from isolated myofiber atrophy to dystrophic changes, with a degree of severity that does not always correlate with overall clinical involvement [6, 7]. In our family, the affected members did not show clinical signs specific of collagenopathy, like distal joint laxity and proximal joints contractures. The phenotype was characterized by slowly progressive muscular impairment, presenting in childhood or adolescence, even later than the reported cases, with maintenance of walking ability throughout the entire follow-up; moreover, no respiratory involvement was found. In addition, the typical outside-in pattern at thigh muscle imaging was not evident and muscular biopsy showed only minimal dystrophic changes. Overall, our patients’ phenotype was not clearly suggestive of autosomal recessive COL6A2, in terms of disease severity, suggesting that the disease spectrum of the clinical spectrum of COL6-RMs is broader than expected and should always taken into consideration in the differential diagnosis of myopathies. Clinical data reported in the literature [1, 8–13] show indeed a high frequency of joints contractures in these patients (23/26 pts), associated with distal joint hyperlaxity in around half of cases (6/13 pts); respiratory involvement is variable (7/22 pts), and generally, mild and disease onset is during infancy or adolescence in less than 50% of patients (11/26 pts). Including also patients with autosomal recessive COL6A2 BM from larger cohorts of COL6-RMs patients [14, 15], it is clearly evident that walking ability is usually preserved (49/51), although supports to walk may become necessary in adult age. Our cases suggest that milder autosomal recessive COL6A2 cases are probably underestimated and misdiagnosed.

The atypical phenotype and the nonspecific radiological and histological pattern, directed us toward several unsuccessful analyses. In the end, only the whole exome sequencing was diriment, evidencing for the proband and the two symptomatic sisters two VUS (c.1514G > A; c.2877G > C), in the COL6A2 gene, never reported before. Only the first one (maternally inherited) was indicated as pathogenic by in silico predictors. These data, together with symptoms reported by the mother, first directed the investigations toward an autosomal dominant pathology. The analysis on skin fibroblasts treated with ascorbic acid was thus essential to clarify the autosomal recessive inheritance of this BM, confirming that also the second VUS, absent in the mother, was pathogenic. Hence, mother’s symptomatology was a further relevant confounding factor; we concluded that her phenotype was just a consequence of a long-term myasthenia gravis.

The diagnostic process in this family started around 30 years ago and the correct diagnosis was reached with 22 years of delay, with relevant patient journey. Today wide genetic panels are commonly used in clinical practice, increasing the possibility of early diagnosis for these rare and clinically heterogeneous muscular diseases. However, even NGS results may be not definitive; in the context of collagen VI disorders, analysis of collagen VI on fibroblasts obtained by the skin biopsy still represents a crucial tool in atypical cases to reach genetic diagnosis.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

COL6-RMs

Collagen VI-related myopathies

UCMD

Ullrich Congenital Muscular Dystrophy

BM

Bethlem myopathy

THD

Triple-helix domain

CK

Creatine kinase

NGS

Next-Generation Sequencing

VUS

Variants of unknown significance

HF

Head flexors muscles

PUL

Proximal upper limbs muscles

DUL

Distal upper limbs muscles

PLL

Proximal lower limbs muscles

DLL

Distal lower limbs muscles

WGM

Walton Gardner-Medwin scale

FVC

Forced vital capacity

Author contributions

AG: acquisition, analysis and interpretation of data, writing, review and editing of the original draft. GR: acquisition, analysis and interpretation of data, writing, review and editing of the original draft. AR: conceptualization, review and editing of the original draft. EI: acquisition of data. FS: investigation. LNG: investigation. AC: acquisition of data. LM: resources, analysis and interpretation of data, review and editing of the original draft. SG: conceptualization and design of the work, analysis and interpretation of data, supervision, writing, review and editing of the original draft.

Funding

The Italian Ministry of Health (RRC) partially supported this work.

Availability of data and materials

The data that support the findings of this study are not openly available due to reasons of sensitivity and are available from the corresponding author upon reasonable request.

Declarations

Conflicts of interest

All authors have no relevant financial or non-financial interests to disclose related to this work. However, L. M. received funding for travel, meeting attendance, and advisory board participation from Sanofi Genzyme, Roche, Biogen, Amicus Therapeutics, Alexion Pharmaceuticals, Janssen, UCB, Lupin, and Argenx. He also received funding for fellowship from Biogen and Alexion Pharmaceuticals.

Ethics approval and consent to participate

Ethical approval was not required for this study as it is a retrospective analysis conducted within the routine clinical workflow. All data were collected as part of standard clinical care and were anonymized before analysis, in compliance with the current version of the Declaration of Helsinki as well as all national legal and regulatory requirements. Patients provided informed written consent.

Consent for publication

Patients provided written consent for publication.