An integrated model of population growth saturation and basal ROS levels predicts cellular ferroptosis sensitivity

Ferroptosis is a form of regulated cell death characterized by iron-dependent lipid peroxidation and membrane rupture. While cellular populations reaching confluence are known to have limited sensitivity to ferroptosis, an understanding of the interplay between growth dynamics, ROS levels and ferroptosis is currently lacking. Here we use live-cell imaging coupled to ROS tracing to reveal a feedback loop between population growth and ferroptotic cell death. Starting out from the observation that the cellular proliferation rate declines with increased cellular density, we find that ROS levels also decline with increasing cellular density. Low ROS levels make cells insensitive to ferroptosis, which in turn enables population growth. This feedback produces two steady states: (i) a ferroptosis-insensitive state characterized by slow growth, low levels of ROS and low rates of cell death and (ii) a ferroptosis-sensitive state characterized by rapid growth, ROS accumulation, and high rates of ferroptosis. Interestingly, triggering effective ferroptosis by interfering with GPX4 activity is directly linked with this mechanism. On the other hand, keeping cell numbers and drug concentration/cell constant while restricting growth space led to reduced proliferation, reduced ROS and decreased ferroptotic cell death. Importantly, ferroptosis resistance at high cellular confluency could be broken by increasing cellular ROS and lipid ROS through a galactose-promoted OXPHOS switch. A mathematical model of the feedback mechanism predicts the long-term fate of populations as well as their ferroptosis sensitivity when external conditions impacting cell proliferation rates, ROS, or both are changed.