A CCG expansion in ABCD3 causes oculopharyngodistal myopathy in individuals of European ancestry

Abstract

Oculopharyngodistal myopathy (OPDM) is an inherited myopathy manifesting with ptosis, dysphagia and distal weakness. Pathologically it is characterised by rimmed vacuoles and intranuclear inclusions on muscle biopsy. In recent years CGG • CCG repeat expansion in four different genes were identified in OPDM individuals in Asian populations. None of these have been found in affected individuals of non-Asian ancestry. In this study we describe the identification of CCG expansions in ABCD3, ranging from 118 to 694 repeats, in 35 affected individuals across eight unrelated OPDM families of European ancestry. ABCD3 transcript appears upregulated in fibroblasts and skeletal muscle from OPDM individuals, suggesting a potential role of over-expression of CCG repeat containing ABCD3 transcript in progressive skeletal muscle degeneration. The study provides further evidence of the role of non-coding repeat expansions in unsolved neuromuscular diseases and strengthens the association between the CGG • CCG repeat motif and a specific pattern of muscle weakness.

Fig. 1. European and Australian OPDM pedigrees.

Pedigrees of the eight OPDM-affected families included in this study. Obligate carriers are indicated and “?” denotes individuals in whom the affection status remains unknown.

Fig. 2. Identification of an expanded CCG repeat within the 5′-UTR of ABCD3 in OPDM.

A srWGS data (AUS1-V:3) in IGV showing ~50% of reads mapped to the ABCD3 promoter contain expanded repetitive sequences. B Identification of ABCD3 CCG expansion causing OPDM in srWGS through Genomics England 100,000 Genome Project. C Outputs from the Bionano Access® Software showing the presence of an expansion within the 5′-UTR of ABCD3 in an OPDM individual relative to a healthy control. D Electropherogram of repeat-primed PCR. The upper panel showed no-repeat expansion in the unaffected individuals, while the lower panel showed a saw-tooth tail pattern of the repeat expansion in an affected OPDM individual (FR2-II:1).

Fig. 3. Targeted long-read sequencing and analysis of the ABCD3 repeat expansion.

A Long-read genotyping of the ABCD3 5′-UTR STR expansion showed heterozygous pure CCG expansions in all affected individuals (n = 19). The wild-type alleles in affected individuals and two unaffected relatives contained 7 x CCG repeats. Unaffected individuals are denoted by an *. Individual IDs for affected males are bolded. B Repeat expansions are larger and more variable in size in affected females (n = 11) compared to affected males (n = 8). Repeat expansion sizes in males vs females were compared with two-sided t test and data are presented as mean values ±SEM. C The shared haplotype region of 560 kb, identified by 15 SNPs between the vertical black bars. The red line indicates the position of the ABCD3 repeat expansion. For comparison, we show in the bottom row the most common haplotype found in controls (18% of alleles). Haplotypes were obtained by phasing genotype data (WGS and WES) with SHAPEITv4. Since the rightmost SNP in the shared haplotype is only available for WGS samples, the four WES samples have been greyed out. D Distribution of the repeat count estimation by EH, for carriers of the haplotype (N = 96) and non-carriers (N = 34,583) in the Rare Disease cohort. Whiskers are based on the standard 1.5 IQR value. For repeat haplotype carriers: min = 7, min whisker = 23, Q1 = 27, median = 30, Q3 = 34, max whisker = 42, max = 109. In non-carriers: min = 6, min whisker = 7, Q1 = 7, median = 7, Q3 = 7, max whisker = 7, max = 93.

Fig. 4. OPDM individuals present with ptosis, and frequently progress to develop weakness of the facial muscles and muscle atrophy and weakness of the distal lower and upper limbs.

Clinical images of AUS2-IV:2 at 67 years of age (A), AUS1-II:3 (B), AUS3-II:1 at 71 years of age (C), AUS1-IV:3 at 77 years of age (D) and UK2-III:1 at the age of 52 years of age (J) showing ptosis and marked facial weakness. Muscle atrophy of the distal limbs is seen in the hands (E) of AUS2-IV:2 (67 years old) and individual AUS1-II:3 developed contractures (F). Wasting of the muscles of the lower legs is also seen in AUS3-II:1 (71 years old) (G) and UK1-II:1 at 57 years of age (K). Mild ptosis is seen in younger affected individuals: AUS1-V:3 at 50 years of age (H) and AUS1-V:4 at 48 years of age (I). Lower limbs muscle MRI for case UK2-III-2. L Thighs are mostly preserved with some mild fatty infiltration predominantly affecting the biceps femoris (*). M Calf muscles are more severely affected. There is a marked predominant involvement of the posterior compartment including soleus and gastrocnemius muscles.

Fig. 5. Muscle pathology in ABCD3-related OPDM.

Haematoxylin and eosin staining (A–C) and modified Gomori trichrome staining (D–F) show rimmed vacuoles, internally localised myonuclei, and variation in myofibre size. Increased fibrotic tissue is also evident. Inclusions also stain with AMP deaminase (G). Subsarcolemmal myeloid bodies and autophagic vacuoles are evident on electron microscopy (H). I Strong p62-positivity (red) observed with immunofluorescence in occasional myofibres on a muscle section of the proband from AUS3, AUS3-II:1. No intranuclear p62 inclusions were evident on muscle biopsy. Myonuclei are stained with Hoechst (blue). Individuals: AUS2-IV:3 (A), FR3-II:1 (B), AUS3-II:1 (C), AUS1-II:3 (D), AUS1-III:3 (E), FR2-II:2 (cricopharyngeal muscle, F), AUS2-IV:2 (G, H). Immunohistochemistry to label p62 (L–N) showed intranuclear aggregates in fibroblasts (L), and exocrine glands (M, N) from OPDM individuals UK1-II:1 and UK2-III:2 but not in control skin (K). FFPE slide from post-mortem brain of an individual with dementia showing a strong p62 neuronal intranuclear signal used as positive control for the staining (J). O p62-positive (red) intranuclear inclusion in cultured primary skin fibroblast from an OPDM individual (AUS1-IV:3). Scale bars: 10 µm (O), 20 µm (G, I), 50 µm (A, B, D–F, K–N), 100 µm (C, J). Stainings were carried out once on patients’ samples with appropriate controls according to standard practice and histopathology procedures in an ISO15189 accredited laboratory.

Fig. 6. ABCD3 is overexpressed in OPDM skeletal muscle and forms nuclear and cytoplasmic foci in OPDM-derived fibroblasts and OPDM skeletal muscle.

A Analysis of skeletal muscle RNA-seq showed that ABCD3 appears to be overexpressed in OPDM skeletal muscle compared to healthy controls. The normalised counts were obtained from the OUTRIDER results table. B Detection of ABCD3 transcript in fibroblasts and muscle tissue by HCR RNA FISH. In the OPDM line (top) and muscle (bottom), the intensity of ABCD3 signal (in red) is increased and tends to aggregate in nuclear foci in myonuclei. C Increased ABCD3 nuclear expression in patient myonuclei (n = 33) compared to control myonuclei (n = 46, p < 0.0001). Scale bars: 10 μM. HCR RNA FISH experiments were repeated independently twice with similar results.