Agent-based modeling of neuronal mitochondrial dynamics using intrinsic variables of individual mitochondria

Mitochondrial networks undergo remodeling to regulate form and function. The dynamic nature of mitochondria is maintained by the dueling processes of mitochondrial fission and fusion. Dysfunctional mitochondrial dynamics have been linked to debilitating diseases and injuries, suggesting mitochondrial dynamics as a promising therapeutic target. Increasing our understanding of the factors influencing mitochondrial dynamics will help inform therapeutic development. Utilizing live imaging of primary neurons, we analyzed how intrinsic properties of individual mitochondria influence their behavior. We found that size, shape, mitochondrial membrane potential, and protein oxidation predict mitochondrial fission and fusion. We constructed an agent-based model of mitochondrial dynamics, the mitochondrial dynamics simulation (MiDyS). In silico experiments of neuronal ischemia/reperfusion injury and antioxidant treatment illustrate the utility of MiDyS for testing hypothesized mechanisms of injury progression and evaluating therapeutic strategies. We present MiDyS as a framework for leveraging in silico experimentation to inform and improve the design of therapeutic trials.