Reactive oxygen species (ROS) are molecules that contain oxygen with reactivity greater than the ground-state molecular oxygen. Superoxide radicals (O₂^·–) and hydrogen peroxide (H₂O₂) are ROS produced in cells during normal aerobic metabolism. High levels of these ROS are known to induce cell damage and cell death. However, cells contain large amounts of antioxidant enzymes, such as superoxide dismutases (SOD) and glutathione peroxidase (GPx), providing a defense system to prevent and repair the damage by converting O₂^·– to H₂O₂, and finally to H₂O.

Due to its involvement in pathophysiological processes, the interest is focused on investigating the metabolism of H₂O₂. Understanding the complex cellular enzymatic processes involved in H₂O₂ removal is challenging. It is problematic to experimentally detect most species in vivo because of their low concentrations and short half lives. In addition, kinetic observations from in vitro experiments done in dilute aqueous buffer do not always hold in the complex milieu of the cell. A mathematical model, however, can works as a heuristic tool to study the complex system describing the metabolism of H₂O₂.

In previous works, lumped parameter models were established for studying mitochondrial H₂O₂ level. Mitochondria were chosen as the venue because ROS (mainly O₂^·– in this case) primarily comes from the mitochondrial respiratory chain. Using carefully chosen sets of physiological species concentrations and reaction rate constants obtained from literatures, simulated results from those models were able to show correlations to reported biological observations. However, spatial dependency was not considered and cellular average concentrations were used in the lumped parameter models. Therefore, gradients in concentrations exist within a mitochondrion, which might lead to local dominance of the rate of change of H₂O₂, were not captured.

The current work is an extension of the previous kinetic models with spatial dependency included. A three-dimensional kinetic model of mitochondrial redox system is proposed. Reactions from the generation of O₂^·–, dismutation of O₂^·– to H₂O₂, and removal of H₂O₂ are modeled for various compartments (intermembrane space, inner membrane, matrix) of the mitochondria. By simulating in vivo physiological state, this modeling is aimed to approximate mitochondrial ROS concentrations and provide insights for better understanding the dynamic links between these species.