Mitochondrial ROS production correlates with, but does not directly regulate lifespan in Drosophila.

The Mitochondrial Free Radical Theory of Aging (MFRTA) is currently one of the most widely accepted theories used to explain aging. From MFRTA three basic predictions can be made: long-lived individuals or species should produce fewer mitochondrial Reactive Oxygen Species (mtROS) than short-lived individuals or species; a decrease in mtROS production will increase lifespan; and an increase in mtROS production will decrease lifespan. It is possible to add a further fourth prediction: if ROS is controlling longevity separating these parameters through selection would be impossible. These predictions have been tested in Drosophila melanogaster. Firstly, we studied levels of mtROS production and lifespan of three wild-type strains of Drosophila, Oregon R, Canton S and Dahomey. Oregon R flies live the longest and produce significantly fewer mtROS than both Canton S and Dahomey. These results are therefore in accordance with the first prediction. A new transgenic Drosophila model expressing the Ciona intestinalis Alternative Oxidase (AOX) was used to test the second prediction. In fungi and plants, AOX expression regulates both free radical production and lifespan. In Drosophila, AOX expression decreases mtROS production, but does not increase lifespan. This result contradicts the second prediction of MFRTA. The third prediction was tested in flies mutant for the gene dj-1beta. These flies are characterized by an age-associated decline in locomotor function and increased levels of mtROS production. Nevertheless, dj-1beta mutant flies do not display decreased lifespan, which again is in contradiction with MFRTA. In our final experiment we utilized flies with DAH mitochondrial DNA in an OR nuclear background, and OR mitochondrial DNA in DAH nuclear background. From this, Mitochondrial DNA does not control free radical production, but it does determine longevity of females independently of mtROS production. In summary, these results do not systematically support the predictions of the MFRTA. Accordingly, MFRTA should be revised to accommodate these findings.