Profiling mitochondrial complex I inhibitors by combining mitochondrial morphological features and maximum common chemical substructures

Studies on the mechanism of toxins are important to reveal the etiology of Parkinson's disease (PD). Complex I inhibitors are a main group of PD toxins, and their chemical features have been intensively studied. However, the chemical structures specific to mitochondrial morphological changes are still unknown. We developed a drug profiling system that combines mitochondrial morphological quantification and chemical substructure computation, which allows us to discover chemical substructures specific to mitochondrial morphological changes. Using this system, we quantified the mitochondrial morphology induced by annonaceous acetogenins, and calculated the maximum common substructure of acetogenins inducing similar cell responses. We discovered that (1) the hydroxyl groups close to γ-lactone in annonacin may result in stronger effects on nuclear size reduction and mitochondrial aggregation; (2) bis-THF acetogenins may have less mitochondria aggregation than those of mono-THF acetogenins; (3) less hydroxyl group at the alkyl chain opposite from γ-lactone end may result in less mitochondrial fission.