Mitochondrial mutations in mammalian aging: an over- hasty about-turn?

The very low abundance of mitochondrial DNA (mtDNA) mutations in nearly all mammalian tissues even in old age has led most mitochondriologists to reject the idea that such mutations might have a causal role in aging, despite the strong circumstantial (e.g. interspecies) evidence that they do have such a role, the promulgation since 1998 of two detailed mechanisms whereby low levels of mtDNA mutations could be harmful, and the report of a transgenic mouse with cardiomyopathy apparently caused by artificially high levels of mtDNA mutations in the heart. A recent report of a mouse with ubiquitously accelerated accumulation of mtDNA mutations and an array of phenotypes reminiscent of aging has abruptly overturned this consensus, with not only the authors but also many other expert commentators suggesting that the mtDNA mutation theory of aging has risen from the ashes. However, there are compelling reasons to doubt the relevance of this mouse to normal mammalian aging and thus to seek further testing of specific mechanistic hypotheses for how mtDNA mutations could cause age-related dysfunction. It has been over 30 years since Harman extended his original free radical theory of aging with the idea that mitochondria might be the main victims, as well as perpetrators, of free radical damage, and specifically that the mitochondrial DNA (mtDNA) might be the component whose vulnerability to such attack mediates the role of free radical reactions in limiting lifespan. This hypothesis has received enormous attention within biogerontology over the years, especially after it was more successfully publicised 17 years later by Linnane and colleagues. However, mitochondriologists (and, increasingly, also biogerontologists) have become less enthusiastic about this theory since a decade or so ago, as it became clear that the abundance of mutant mtDNA remains extremely low – below 0.1% – throughout life in almost all tissues. Attempts to revive the theory with proposals for how the focal distribution of mutant mtDNA could allow low levels to be pathogenic were generally met with skepticism on account of their complexity as perceived (perhaps not altogether justifiably) by the mitochondriology community. However, the possible role of mtDNA mutations in mammalian aging remains attractive in view of the clear correlation between rate of aging and rate of accumulation of mtDNA damage that is seen across species and also between calorie-restricted and ad lib-fed rodents. In 2000, Zassenhaus’s group reported that elevated levels of mtDNA mutations in the heart, caused by introduction of a transgenic mtDNA polymerase carrying a mutation in its proof-reading domain, induced severe cardiomyopathy in mice. Remarkably little attention has been paid to this work, despite the group’s subsequent confirmation that this was not an effect of the promoter, as had been suggested after expression of GFP under this promoter was reported to be toxic. Evidently the community was still not ready to entertain the idea that rare mtDNA mutations could be harmful. This all changed a few months ago, when Trifunovic et al. published on a mouse also expressing a proofreading-deficient mtDNA polymerase, but this time in all tissues. These mice exhibited elevated (3to 5-fold) mtDNA mutation levels in numerous tissues (all examined), and a plethora of phenotypes associated with aging appeared prematurely (Table 1); life expectancy was about a year. The publication was accompanied by a commentary by two highly respected biogerontologists, George Martin and Larry Loeb, and attracted considerable attention from the popular science press, in which quoted biogerontologists were more or less unanimous in their opinion that, while a