Are mitochondria a permanent source of reactive oxygen species?

The observation that in isolated mitochondria electrons may leak out of the respiratory chain to form superoxide radicals (O-2(.-)) has prompted the assumption that O-2(.-) formation is a compulsory by-product of respiration. Since mitochondrial O-2(.-) formation under homeostatic conditions could not be demonstrated in situ so far, conclusions drawn from isolated mitochondria must be considered with precaution. The present study reveals a link between electron deviation from the respiratory chain to oxygen and the coupling state in the presence of antimycin A. Another important factor is the analytical system applied for the detection of activated oxygen species. Due to the presence of superoxide dismutase in mitochondria, O-2(.-) release cannot be realistically determined in intact mitochondria. We therefore followed the release of the stable dismutation product H2O2 by comparing most frequently used H2O2 detection methods. The possible interaction of the detection systems with the respiratory chain was avoided by a recently developed method, which was compared with conventional methods. Irrespective of the methods applied, the substrates used for respiration and the state of respiration established, intact mitochondria could not be made to release H2O2 from dismutating O-2(.-). Although regular mitochondrial respiration is unlikely to supply single electrons for O-2(.-) formation our study does not exclude the possibility of the respiratory chain becoming a radical source under certain conditions. (C) 2000 Elsevier Science B.V. All rights reserved.