The m.5650G>A mitochondrial tRNAAla mutation is pathogenic and causes a phenotype of pure myopathy

1. Introduction 

Point mutations of mitochondrial DNA (mtDNA) are responsible for a variety of different, but occasionally overlapping clinical syndromes [1]. Tissues heavily reliant on oxidative metabolism, such as skeletal muscle, are often worst affected and myopathy is a common clinical manifestation of mitochondrial disease. In these circumstances myopathy is usually only one of a number of clinical features comprising a ‘mitochondrial syndrome’ such as MELAS or Kearns-Sayre syndrome. Isolated or ‘pure’ myopathy as the solitary manifestation of mitochondrial disease is relatively rare. Nevertheless, it has been associated with several point mutations of mtDNA in both the cytochrome b [2] and mitochondrial tRNA (mt-tRNA) genes [3], as well as recessive mutations in the nuclear-encoded TK2 gene [4]. Maternal inheritance of this pure myopathy phenotype is even less common, being limited to a few mt-tRNA mutations [5]. We report a family where the m.5650G>A mutation in mt-tRNAAla has caused an increasingly severe proximal (limb-girdle) myopathy in successive generations. We detail the phenotype associated with this mutation and present evidence from single muscle fibre analysis that confirm its pathogenicity.

2. Methods 

Standard histological and histochemical (including sequential cytochrome c oxidase (COX) and succinate dehydrogenase (SDH)) analyses were performed on frozen sections (10μm) of quadriceps muscle biopsies taken from the index case and her mother. The activities of the respiratory chain complexes and the matrix marker citrate synthase were determined as previously described [6].

Total DNA was extracted from several tissues by standard procedures. Rearrangements of mtDNA were investigated by Southern-blot analysis and long-range PCR of muscle DNA. The entire sequence of the mitochondrial genome was amplified using 36 M13-tailed oligonucleotide primer pairs with DNA isolated from skeletal muscle as template, sequenced using BigDye® Terminator v3.1 chemistries (Applied Biosystems) and directly compared to the revised Cambridge reference sequence [7].

To determine the level of heteroplasmy, a 150bp PCR product spanning the mutation site was amplified using a modified forward primer (nt 5549–5569) 5′-GCCGTCAGTAAGTTGCAATAC-3′ and a modified reverse primer (nt 5698–5676) 5′-CTGTTAAGTAAGTGTTTGTGGGT-3′ with the mismatch nucleotides shown in bold in each primer sequence; the mismatches in each primer abolish naturally-occurring DdeI sites. Two DdeI restriction sites in the wild-type PCR product cut the 150bp amplimer into three fragments of 38, 60 and 52bp. In the presence of the m.5650G>A transition, one of these sites is lost, resulting in restriction fragments of 38 and 112bp. Prior to the last cycle of PCR, 5μCi [α-32P]dCTP (3000Ci/mmol) was added. Labelled products were precipitated, digested with 10U DdeI, separated through a 12% non-denaturing polyacrylamide gel, and the radioactivity in each fragment quantified using ImageQuant software (Amersham Biosciences/GE Healthcare).

3. Results 

Histological examination of the proband’s muscle biopsy showed marked variation in fibre size, with evidence of inflammatory changes on H&E histology (Fig. 1a), and upregulation of MHC Class I antigens on immunohistochemistry (data not shown). Enzyme histochemistry revealed a marked number (>50%) of COX-deficient fibres (Fig. 1b), some of which showed evidence of subsarcolemmal mitochondrial accumulation. Respiratory chain enzyme analysis of a skeletal muscle homogenate showed a marked decrease in the activities of both complex I (20% of controls) and complex IV (15% of controls) when related to citrate synthase activity, with normal activities of complexes II and III. Muscle biopsy of her mother showed none of the histological changes seen in the index case, but the COX defect was even more severe (>80% fibres affected). Similar to the proband, respiratory chain enzyme studies revealed a combined defect involving complexes I and IV (20% and 10% of controls, respectively).

View Large Image Download to PowerPoint

Fig. 1. Muscle biopsy analysis. (a) H&E-stained section of quadriceps muscle from the index case revealing a marked variation in fibre size and the presence of inflammatory infiltrate. Some fibres clearly show central nuclei. (b) Histochemical demonstration of combined COX and succinate dehydrogenase (SDH) activities in the patient’s muscle biopsy, revealing significant numbers of COX-deficient fibres with subsarcolemmal accumulation of abnormal mitochondria. (c and d) Show H&E and COX/SDH sections of the mother’s muscle biopsy, respectively, which does not exhibit dystrophic changes but has very high levels of respiratory-deficient fibres. (e) Is a low magnification H&E-stained section of quadriceps from the index case illustrating the extent of the inflammatory cell infiltrate.

Southern blot and long-range PCR analysis excluded the presence of large-scale mtDNA rearrangements in the proband’s muscle DNA. Direct sequencing of the entire mitochondrial genome revealed a previously reported m.5650G>A transition [8] in the MTTA gene encoding mitochondrial tRNAAla (Fig. 2). PCR-RFLP analysis showed that the mutation was heteroplasmic, but present at very high levels (>95% mutant load) in all available tissues (muscle, blood, urine and buccal epithelia) with the exception of hair follicles (10%) (Fig. 2). Analysis of several tissues from her mother revealed similarly high levels of the mutation (Fig. 2), confirming maternal transmission of this mutation, whilst the grandmother harboured detectable levels only in urinary epithelial cells (2% mutant load) (Fig. 2).

View Large Image Download to PowerPoint

Fig. 2. m.5650G>A mutation analysis. Determination of the relative amounts of mutant and wild-type mtDNA in tissues from the patient, her mother and her grandmother by PCR-RFLP analysis. Lane 1, uncut sample; lanes 2 and 3, control samples; lanes 4 and 9, skeletal muscle; lanes 5, 10 and 13, circulating lymphocytes; lanes 6, 11 and 14, buccal epithelia; lanes 7 and 15, hair shafts; lanes 8, 12 and 16, urinary epithelia.

Single muscle fibre analysis was performed to determine whether the amount of mutated mtDNA correlated with the observed biochemical phenotype in individual fibres. We detected higher levels of the m.5650G>A mutation in the proband’s COX-deficient fibres (99.0±0.29% (n=9)) than in her COX-positive fibres (87.6±2.26% (n=15)) (P<0.0008, two-tailed Student’s t test), confirming segregation of the m.5650A genotype with respiratory chain dysfunction (Fig. 3a).

View Large Image Download to PowerPoint

Fig. 3. Confirmation of m.5650G>A pathogenicity. Graphical representation of the single fibre (COX-positive and COX-deficient) studies showing highest levels of the m.5650G>A mutation segregating with the biochemical defect. Schematic representation of the mt-tRNAAla cloverleaf structure, showing the position of the mutation in the aminoacyl acceptor stem and the affected C–G base pair. Also indicated is the mitochondrial alanyl-tRNA synthetase recognition site.

4. Discussion 

The m.5650G>A mutation in mt-tRNAAla has been reported once previously in a patient with cerebral autosomal dominant arteriopathy subcortical infarcts and leukoencephalopathy (CADASIL) [8]. This patient also harboured a concomitant R133C mutation in the Notch 3 gene and although myopathy was described, the phenotype was dominated by ataxia, dysarthria, memory loss and cognitive impairment. While these clinical features are characteristic of CADASIL, they are also not uncommon in mitochondrial disease, making it difficult to determine the relative contribution of each pathology to the phenotype. It is clear from the family we report here that the phenotype of the m.5650G>A mutation is confined to a pure myopathy. The mutation is clearly transmissible through the maternal lineage and appears to have arisen independently rather than as a consequence of the Notch 3 gene mutation as has been suggested possible [9].

The clinical presentation in the index case of a pure proximal muscle weakness, in the absence of any other features and with a suspicion of autosomal dominant inheritance, led to a clinical differential diagnosis of an autosomal dominant LGMD or myopathy. Amongst the defined autosomal dominant LGMDs considered in this family, LGMD1A was thought unlikely given the early presentation of the index case, while the absence of cardiac involvement in the mother and grandmother made a diagnosis of LGMD1B less promising. Creatine kinase levels were considered rather low for LGMD1C, and there were no associated features such as rippling or mounding, though this condition can be highly variable. Other subtypes of autosomal dominant LGMD are relatively poorly characterized and the responsible genes have not yet been defined. Of the autosomal dominant congenital myopathies considered, central core disease was clinically perhaps the most likely. The subsequent identification of mitochondrial disease in this family highlights the crucial role of muscle biopsy in distinguishing these differential diagnoses.

The clear clinical distinction in disease severity between mother and daughter was not reflected in either the histochemical or biochemical analyses of the muscle biopsies. Indeed, the mother’s biopsy demonstrated more COX-deficient fibres and a lower overall activity of complex IV when compared with controls. However, the dystrophic appearance of the proband’s biopsy does suggest a more aggressive disease process and direct comparison of the two biopsies must also consider the 31-year age difference between the proband and her mother. Marked dystrophic changes on muscle biopsy are an uncommon finding in mitochondrial disease, but have been observed in patients with large-scale single deletions of mtDNA and mt-tRNA point mutations, some of whom also had clinical features consistent with LGMD [10]. It has been suggested that dystrophic changes in muscle biopsy from patients with mitochondrial disease is genotype related [11]. While to some extent this may be true, the histological discrepancy we observed between mother and daughter implies that genotype alone is not a sufficient explanation and that other factors, perhaps related to inflammatory response mechanisms, are equally important.

The threshold for impairment of oxidative phosphorylation (as evidenced by COX deficiency) appears to be extremely high for this mutation, with COX-positive muscle fibres harbouring levels of up to 95% (Fig. 3a). This suggests that the m.5650G>A mutation is relatively benign, but such high mutant loads increase transmissibility and reduce the probability of favourable segregation, (although the 10% mutant load observed in hair follicle mtDNA demonstrates that this is still possible). Consequently, offspring are highly likely to inherit the mutation at levels near or above threshold and are at increased risk of disease expression. Thus, in the family we report the severity of the myopathy increases with successive generations while the age of onset decreases.

Although the nucleotide at position 5650 within mtDNA is not highly conserved across species, this region of the mt-tRNAAla acceptor stem (Fig. 3b) is recognized by mitochondrial alanyl-tRNA synthetase and serves a vital function in mt-tRNAAla aminoacylation, an essential prerequisite for protein translation [12]. The m.5650G>A mutation may disrupt this interaction between synthetase and acceptor stem, decreasing the efficiency of aminoacylation. The mutation does however fulfil other canonical criteria for pathogenicity – it is heteroplasmic, segregates with clinical disease, produces an observable biochemical defect and is absent from the control population (not present on either of the publicly available databases [13], [14]). Interestingly, a previously reported mt-tRNAAla mutation (m.5591G>A) which also caused a limb-girdle myopathy resides within the aminoacyl acceptor stem of the tRNA molecule and was also associated with a marked histocytochemical COX deficiency [3].

In conclusion, the investigation of a possible autosomal dominant LGMD or myopathy in this family with increasingly severe disease through successive generations has revealed a maternally inherited mt-tRNA point mutation. We have been able to confirm the pathogenicity of this m.5650G>A mutation and ascribe a definitive phenotype where previously this has not been possible. The pure myopathic features seen in this family illustrate the remarkable tissue specificity of some mtDNA point mutations and indicate the need to consider mitochondrial disease in any family presenting with an apparently autosomal dominant limb-girdle syndrome.