Structural instability of mitochondrial DNA (mtDNA), consisting either of large-scale rearrangements, tissue-specific depletion or both, is a major cause of mitochondrial dysfunction and disease in humans (Zeviani and Di Donato, 2004). Almost 20 years have elapsed since the discovery that single, large-scale deletions (Holt et al. , 1988) across short direct repeats (Mita et al. , 1990) occur and become clonally expanded (Mita et al. , 1989) in the mtDNA of patients with the sporadic form of KearnsâSayre syndrome (Zeviani et al. , 1988) or its milder variant, progressive external ophthalmoplegia (PEO; Moraes et al. , 1989). Shortly thereafter, the accumulation of multiple mtDNA-deleted species was observed in families with recurrent cases of PEO transmitted as an autosomal dominant trait (Zeviani et al. , 1989). Mendelian inheritance of mtDNA mutations appeared at first to be a contradiction, since this genome is transmitted in a strictly maternal fashion; but this was explained as the consequence of a dominant mutation in a nuclear gene affecting the structural integrity of mtDNA (Zeviani et al. , 1990). This hypothesis proved to be correct, as it led to the discovery of at least four genes responsible for familial PEO associated with multiple mtDNA deletions. These genes encode ANT1, the muscle-specific isoform of the mitochondrial adenine nucleotide translocator (Kaukonen et al. , 2000); Twinkle, a mtDNA helicase (Spelbrink et al. , 2001); and pol ÎłA (van Goethem et al. , 2001) and B (Longley et al. , 2006); the subunits of the mtDNA-specific polymerase holoenzyme, pol Îł. Mutations in pol ÎłA, the larger catalytic subunit of the pol Îł holoenzyme, are indeed responsible for a much wider spectrum of clinical presentations (Horvath et al. , 2006), including not only dominant but also recessive PEO (Lamantea et al. , 2002), juvenile-onset spino-cerebellar ataxia and epilepsy (Tzoulis et âŠ