A Continuum Model of Cell Fate in the Response of Human Cells to Ionizing Radiation

Abstract

In response to ionizing radiation induced DNA damage, human cells can either recover and resume proliferation or activate anti-proliferative programs such as cell death and cellular senescence, a state characterized by the long-term enforcement of cell cycle arrest and the loss of recovery potential. Knowledge in this field has been primarily constructed on the basis of experimental approaches that rely on averaging or static snapshots of cell populations. While invaluable insights have been gained regarding the mechanistic details of the molecular circuitry linking DNA damage sensing and downstream cellular programs, such approaches are limited in their ability to account for heterogeneity in the long-term phenotypic outcomes of DNA damage, particularly in the temporal dimension of this cell fate decision-making process. Here, we used live cell imaging to quantify the number and timing of division events that individual human cells go through in the course of one week after acute exposure to ionizion radiation. Such single cell division profiles unmasked heterogeneity both in the long-term maintenance of cell cycle arrest in senescing cells, and in the timescale of cell cycle re-entry of recovering cells. We first present our findings on the molecular mechanisms underlying escape from cell cycle arrest in the presence of unrepaired DNA damage. Using fluorescent reporters for p53 and p21, which trigger cell cycle arrest upon DNA damage, we identified two levels at which quantitative variation in these proteins contributes to escape from cell cycle arrest: (i) low historical averages of p53 and p21 prime cells to eventually escape the arrested state; and (ii) a transient decrease in p53 levels precedes the onset of cell cycle re-entry. Using mathematical modeling, we show that fluctuations in p53 can be amplified into a sharp switch between p21 and CDK2, and consequent cell cycle re-entry. Taken together, our work revealed that p53 and p21 dynamically contribute to the active maintenance of the arrested state, the homeostasis of which can be broken through local amplification of noisy DNA damage signaling. We next discuss our findings on the consequences of heterogeneity in cell cycle arrest duration in the context of combination cancer therapy. We focused on the interaction between acute exposure to ionizing radiation and transient treatment with the Eg5 kinesin inhibitor STLC, which selectively targets mitotic cells. DNA damage antagonized STLC treatment through establishment of cell cycle arrest, leading to a non-monotonic dose response curve in which an intermediate dose of damage optimized cell viability. Optimal damage dose shifted as a function of STLC treatment duration. We aim to understand the extent to which input-ouput relationships in DNA damage signaling can modulate arrest-exit time distributions and shape the response of human cells to dynamic therapeutic strategies. Our investigations on the long-term fate of human cells after exposure to ionizing radiation revealed a phenotypic continuum in which the discreteness of arrested and cycling states is blurred out when taking into consideration whole histories of individual cell behavior. A quantitative account of DNA damage signaling will provide a framework to understand the way human cells regulate specific properties of this cell fate continuum, including arrest duration and recovery potential, and will guide future interventions aimed at optimizing DNA damage-based cancer therapies.