Reactive Oxygen Species (ROS) are generated during several electron transfer reactions in vivo, but primarily in the mitochondrion. Although cells have developed scavenging systems to help cope with the deleterious effects of ROS, when ROS levels inundate the cellular antioxidant system, the cells assume a state of "oxidative stress". A large body of in vitro studies have suggested that while a non-selective damage to macromolecules may occur when cells are under oxidative stress, oxidative stress is frequently associated with the activation of specific signal transduction pathways which act in concert to determine the overall outcome. Unfortunately, most of these in vitro studies have relied on the use of inhibitors of mitochondrial function, which may not be specific to the mitochondrion, making it necessary to examine the oxidative stress response in vivo.

I have addressed some of the lingering questions about the oxidative stress response in the genetically tractable model system—Drosophila melanogaster. First, I examined ROS production in real time, and found that while mutations in mitochondrial complex I proteins trigger increased ROS production, cells harboring mutations in some complex IV proteins produce lower levels of ROS than surrounding wildtype cells. Further analyses showed that although both complex I and IV mutants impair cell proliferation, they activate distinct signal transduction pathways to elicit the cell cycle arrest phenotype: complex I mutants activate a specific signaling cascade initiated by ROS and transduced by the JNK pathway and the p27 homolog, Dacapo; while complex IV mutants utilize AMP and a pathway involving AMPK, p53 and Cyclin E. In addition, I found that while sub-lethal levels of ROS function as signaling molecules to modulate cell cycle progression, they do not affect differentiation or patterning of photoreceptors.

Finally, I report herein that ROS can be transported over a few cell diameters to signal at sites remote from the source of generation; and that p53 confers tolerance to moderate levels of ROS in vivo. Taken together, this dissertation establishes that mitochondrially generated ROS can function both intracellularly and intercellularly as signaling molecules; and paves the way for characterizing additional aspects of the oxidative stress response in Drosophila.