Replicative segregation of mitochondrial DNA (mtDNA) can produce large differences in the proportions of wild-type and mutant mtDNAs in different cell types of patients with mitochondrial encephalomyopathy. This is particularly striking in the skeletal muscle of patients with Kearns-Sayre syndrome (KSS), a sporadic disease associated with large-scale mtDNA deletions, and in sporadic patients with tRNA point mutations. Although the skeletal muscle fibres of these patients invariably contain a large proportion of mutant mtDNAs, mutant mtDNAs are rare or undetectable in satellite cells cultured from the same muscle biopsy specimens. Since satellite cells are responsible for muscle fibre regeneration, restoration of the wild-type mtDNA genotype might be achieved in these patients by encouraging muscle regeneration. To test this concept, we re-biopsied a patient with a KSS phenotype and a mtDNA point mutation in the tRNAleu(CUN) gene and analysed muscle fibres regenerating at the site of the original muscle biopsy. Regenerating fibres were identified by morphological criteria and by expression of neural cell adhesion molecule (NCAM). All such fibers were positive for cytochrome c oxidase (COX) activity by cytochemistry and essentially homoplasmic for wild-type mtDNA, while the majority of non-regenerating fibres were COX-negative and contained predominantly mutant mtDNAs. These results demonstrate that it may be possible to improve muscle function in similar patients by methods that promote satellite cell incorporation into existing myofibres.
Mutations in mitochondrial DNA (mtDNA) are responsible for a wide variety of diseases with prominent neurological or neuromuscular phenotypes (1 -3 ). Most somatic cells contain hundreds to thousands of copies of mtDNA and, in the majority of patients with mitochondrial DNA diseases, two different populations of mtDNA co-exist, wild-type and mutant, a condition known as mtDNA heteroplasmy. The proportions of wild-type and mutant genomes can vary greatly among tissues of patients, as a result of replicative segregation of mtDNA, and this is, in part, responsible for the tissue-specific features of these diseases. The most striking examples of this phenomenon are seen in sporadic patients with mitochondrial myopathies who have apparently inherited new germline mutations. This includes patients with large-scale mtDNA deletions and Kearns-Sayre syndrome (KSS) (4 ) and patients with myopathies associated with tRNA point mutations (5 -8 ) or microdeletions in protein-coding genes (9 ). In these patients, mutant mtDNAs are abundant (usually predominant) in affected skeletal muscles but rare or undetectable in unaffected tissues like peripheral blood cells or fibroblasts.
These disorders are slowly progressive and this is thought to be due to an increase in the proportion of mutant mtDNAs in skeletal muscle with age. There is direct evidence for such an age-related increase in the case of large-scale mtDNA deletions (10 ) and for a tRNAleu(CUN) point mutation (8 ), and suggestive evidence for other tRNA mutations (11 ). This shift in mtDNA genotype is associated with an increased proportion of cytochrome c oxidase (COX)-negative muscle fibres and a worsening clinical course. The molecular mechanism responsible for selection of mutant mtDNAs in skeletal muscle remains unknown; however, it is probably related to a futile attempt by the cell to restore the required oxidative phosphorylation function.
As a result of this selection, an increasingly large difference develops between the mtDNA genotype of mature muscle fibres and the satellite cell population. Satellite cells are mononuclear, committed myogenic cells that remain dormant for most of their lives, but are reactivated as needed for muscle growth and repair (12 ). As in other unaffected cell types, mutant mtDNAs are often undetectable in satellite cells cultured from affected muscles of these patients (7 ,8 ,13 ,14 ). This is the case primarily in patients with new germline mtDNA mutations in whom the overall load of mutant mtDNAs would be expected to be relatively small at birth (7 ). On the basis of these observations, we raised the possibility of elimination of the mutant mtDNA genotype from skeletal muscle by encouraging regeneration of mature muscle fibres from satellite cells (7 ). We tested this idea in a sporadic patient with a tRNA mutation and a KSS phenotype, and demonstrate that regenerating muscle fibres are positive for COX activity and essentially homoplasmic for wild-type mtDNA.
To test whether muscle fibre regeneration was associated with a shift in mtDNA genotype and a corresponding cytochemical phenotype, we analysed regenerating muscle fibres in a patient with a KSS phenotype associated with a heteroplasmic G -> A substitution at position 12 315 in the tRNAleu(CUN) gene. Previous investigations showed almost complete segregation of mutant mtDNAs in mature muscle fibres. In two independent biopsies from the right and left biceps, mutant mtDNAs constituted 94% of total mtDNAs in mature muscle fibres but were undetectable in satellite cells cultured from the same biopsy specimens (7 ). The right biceps muscle was re-biopsied 3 weeks after the original biopsy to analyse regenerating fibres that arose as a result of traumatic muscle fibre necrosis at the site of the original biopsy. These fibres could be recognized by their morphological appearance (small, round with large myonuclei and prominent nucleoli) and by their immunofluorescence with anti-neural cell adhesion molecule (NCAM) antibodies (Fig. 1 C). All such fibres stained positive for COX activity (Fig. 1 D). While the COX staining was less intense than in fully differentiated (large, polygonal) COX-positive fibres (Fig. 1 B), this pattern is typical of regenerating fibres in general.
This study demonstrates the virtual complete reversal of a pathogenic mtDNA genotype in regenerating skeletal muscle fibres in a patient with a tRNA point mutation and a KSS phenotype, a result predicted from the skewed distribution of mutant mtDNAs in the mature muscle fibres and satellite cells. A very similar result has been reported recently in a patient with a point mutation at position 12 320 in tRNAleu(CUN), where it was demonstrated that mutant mtDNAs were rare or undetectable in regenerated fibres 3 weeks after chemically induced muscle necrosis (15 ). Although mutant mtDNAs were not detected in regenerating fibres in our study, we cannot exclude the possibility that they are present below the detection theshold of the assay, which is ~1%. Any remaining mutant mtDNAs would be expected to increase in frequency with time, but it is likely that the rate of increase would be similar to the clinical course of the disease, which has been slowly progressive over decades.
The patient tested in this study and in (15 ) have mtDNA mutations that have not been reported in other pedigrees; however, the results should apply to any patient in whom there is a large difference in the proportion of mtDNA mutants between mature muscle fibres and the satellite cell population. This might include not only sporadic patients with heteroplasmic, primary mtDNA mutations, but also patients with autosomal dominant chronic progressive external ophthalmoplegia in whom multiple different mtDNA deletions are present in mature muscle fibres, but not in satellite cells (16 ). Although the molecular basis for this nuclear defect remains unknown, it is associated with the age-dependent expansion of single, clonal mtDNA deletions in individual myofibres (17 ). By contrast, patients who have inherited mtDNA point mutations from heteroplasmic carrier mothers generally have relatively high proportions of mutant mtDNAs in the satellite cell population (18 ,19 ), and in these cases muscle regeneration would have little impact on the mtDNA genotype.
The regenerating fibres we analysed by PCR could be clearly identified on morphological grounds and by NCAM expression. NCAM is expressed in developing and regenerating muscle fibres but, with the exception of the extraocular muscles (20 ), not in mature muscle fibres (21 ). We did, however, observe a small proportion of NCAM-positive fibres (<5%) in the original muscle biopsy whose presence cannot be attributed to the traumatic injury induced by the biopsy procedure. A recent report of NCAM expression in patients with mitochondrial myopathy suggested that NCAM expression may be present in (non-regenerating?) ragged-red fibres (22 ). This is contrary to what we observed in the muscle of the patient in this study. The vast majority of NCAM-positive fibres did not show succinate dehydrogenase (SDH) overexpression (a sensitive indicator of mitochondrial proliferation) and were COX-positive. The reason for the discrepancy between the two studies is not clear. In any case, the strong positive correlation between NCAM expression and COX activity that we observed is probably not a chance association. It is possible that a small population of muscle fibres is undergoing a process of naturally occurring necrosis and repair at all times and that this is reflected in the expression of NCAM. It may be that such a situation would occur most frequently in ragged-red fibres in which abnormal mitochondria proliferate to such an extent that they displace myofibres, making these segments particularly vulnerable to mechanical damage and necrosis.
Significantly, in several large NCAM-positive fibres, COX activity was not distributed uniformly throughout the cross-sectional area (data not shown), suggestive of a junction between a regenerating segment and a previously COX-negative intact segment. This would be consistent with the heteroplasmic mtDNA genotype that we found in one NCAM- and COX-positive fibre.
How might the results of this study be translated into therapeutic practice? Although myotoxic agents such as the local anaesthetic bupivacaine (15 ,23 ) and the snake toxin notexin (24 ) have been used in experimental systems to induce extensive necrosis and subsequent muscle regeneration, it is unlikely that such a dramatic intervention would be clinically justified. A more realistic approach might involve a programme of strength training and/or an exercise programme that included abundant eccentric muscle contractions. Inducing muscle hypertrophy through strength training may help to change the proportion of wild-type to mutant mtDNAs, but the small number of mitochondria (and mtDNAs) in the satellite cell cytoplasm that would be incorporated into existing myofibres during this process may not alone be sufficient to significantly alter the ratio of mutant:wild-type mtDNAs. Eccentric contractions (muscle contractions at full muscle extension) are known to produce microscopic muscle damage that is repaired by satellite cells (25 ). If the limited damage induced by such an exercise occurred preferentially in ragged-red fibres, as suggested above, then it might be possible to reverse the inexorable increase in the proportion of mutant mtDNAs that appears to result from normal aerobic muscle activity. Certainly these approaches merit testing in restricted muscle groups in patients with an appropriate mtDNA genotype.
The patient was a 54 year old Italian-Canadian man who presented with a KSS-like syndrome. The clinical features of the patient have been described previously (7 ).
Single muscle fibres were dissected with tungsten needles from a 40 µM cross-section of muscle that had been reacted for SDH. Regenerating fibres could be identified easily on adjacent sections stained for COX, SDH or tested for anti-NCAM reactivity by immunofluorescence. Numerous small, round NCAM-positive fibres were present adjacent to the scar formed at the site of the previous biopsy. DNA was isolated from the single fibres directly in the tube used for PCR amplification by extraction in KOH/dithiothreitol (DTT) for 10 min at 65°C (26 ), and a PCR test incorporating a mismatch in the reverse primer was used to analyse the mtDNA genotype (7 ).
Unfixed cryostat sections (6 µM) were reacted with the anti-human NCAM monoclonal antibody 5.1H1.11 (27 ). Supernatant from the hybridoma cells secreting the 5.1H11 antibody was used neat as the primary antibody. The same results were obtained using a commercial anti-NCAM antibody (Leu 19, Becton Dickinson). Biotinylated horse anti-mouse IgGs (Vector Laboratories) were used as secondary antibodies at 1:100. Immunoreactivity was visualized using streptavidin-conjugated Cy3 (1:1000) (Jackson Immunoresearch Laboratories). COX and SDH activity was assessed as described in (27 ).
This work was supported by grants from the MRC and MDAC to EAS. E.A.S. is an MNI Killam Scholar.
Larsson, N.G, Holme, E., Kristiansson, B., Oldfors, A. and Tulinius, M. (1990) Progressive increase of the mutated mitochondrial DNA fraction in Kearns-Sayre syndrome. Pediatr. Res., 28, 131-136.
Human Molecular Genetics
Pages
Introduction
Results
Discussion
Materials And Methods
Patient
Single fibre PCR analysis
Immunofluorescence and cytochemistry
Acknowledgements
References
REFERENCES
This page is maintained by OUP admin. Last updated Sat Oct 18 13:29:50 BST 1997. Part of the OUP Journals World Wide Web service.
Copyright
Oxford University Press, 1997