Intramitochondrial hydrogen peroxide production induces a p53-dependent G0/G1 cell cycle arrest in RPE1 cells

Abstract

Hydrogen peroxide (H2O2) is an important molecule in both physiological and pathological signaling. Cellular antioxidant systems limit the diffusion of H2O2 to a few microns. Targetable and inducible H2O2 production is thus needed to study local H2O2 signaling. For this purpose, D-amino acid oxidase (DAAO) can be used. This targetable yeast enzyme produces H2O2 from D-amino acids, which do not occur in most mammalian cells. In order to study local H2O2 signaling using DAAO model systems that mimic H2O2 production from different sites, these model systems should have comparable levels of DAAO activity. Otherwise, observed differences in phenotype might be caused by varying levels of H2O2 production and not by different localizations of DAAO. To quantify DAAO activity, fluorescent H2O2 sensors like HyPer7 are commonly used. However, these sensors are also influenced by the local reductive capacity, which makes comparisons of DAAO activity across different model systems difficult. We have developed a DAAO activity assay that is not influenced by the local reductive capacity. Since H2O2 production by DAAO consumes an equimolar amount of oxygen, DAAO activity can be determined by measuring the oxygen consumption of cells upon the addition of D-amino acid in a Seahorse XF Analyzer. By enabling unbiased measurements of DAAO activity across different model systems, our novel method advances the study of local H2O2 production and signaling using DAAO. We have used this method to quantify DAAO activity in monoclonal RPE1-hTERT cell lines that express the enzyme at various subcellular sites, in order to select cell lines with similar DAAO activity. Using these cell lines, we found that H2O2 production in the mitochondrial matrix and intermembrane space, but not mitochondrial H2O2 release, induces a p53- and p21-dependent cell cycle arrest in G0/G1.