Comparative mechanistic study of RPE cell death induced by different oxidative stresses.

Oxidative stress is hypothesized to drive the progression of age-related macular degeneration (AMD). Retinal pigment epithelial (RPE) cell layer is important for supporting the function of retina and is particularly susceptible to oxidative stress-induced cell death. How RPE cells die in AMD, especially in geographic atrophy (GA), a late stage of dry AMD, is still controversial. The goal of this study is to compare the features and mechanisms of RPE cell death induced by different oxidative stresses, to identify potential universal therapeutic targets for GA. RPE cell death was induced both in vitro and ex vivo by 4-Hydroxynonenal (4-HNE), a major product of lipid peroxidation, sodium iodate (NaIO3) that has been widely used to model RPE cell death in dry AMD, a ferroptosis inducer RAS-selective lethal 3 (RSL3) or a necroptosis inducer shikonin. We found that RPE necroptosis and ferroptosis show common and distinct features. Common features include receptor-interacting protein kinase (RIPK)1/RIPK3 activation and lipid reactive oxygen species (ROS) accumulation, although lipid ROS accumulation is much milder during necroptosis. This supports cross talk between RPE ferroptosis and necroptosis pathways and is consistent with the rescue of RPE necroptosis and ferroptosis by RIPK1 inhibitor Necrostatin-1 (Nec-1) or in Ripk3-/- RPE explants. Distinct feature includes activated mixed lineage kinase domain like pseudokinase (MLKL) that is translocated to the cell membrane during necroptosis, which is not happening in ferroptosis. This is consistent with the failure to rescue RPE ferroptosis by MLKL inhibitor necrosulfonamide (NSA) or in Mlkl-/- RPE explants. Using this framework, we found that 4-HNE and NaIO3 induced RPE cell death likely through necroptosis based on the molecular features and the rescuing effect by multiple inhibitors. Our studies suggest that multiple markers and inhibitors are required to distinguish RPE necroptosis and ferroptosis, and that necroptosis inhibitor Nec-1 could be a potential therapeutic compound for GA since it inhibits RIPK1/RIPK3 activation and lipid ROS accumulation occurred in both necroptosis and ferroptosis pathways.