Stochastic Biophysics of Cellular Radiosensitivity: From Molecular Noise and Repair Kinetics to Evolutionary Demographics

Abstract

Radiation-induced DNA double-strand breaks (DSBs) drive cellular mortality, mutagenesis, and evolutionary bottlenecks. Classical phenomenological models such as the Linear-Quadratic (LQ) framework predict macroscopic survival but obscure the single-cell stochasticity that governs rare outcomes, including tumor recurrence and radioresistant persistence. We develop a stochastic differential equation framework describing DSB induction and repair as a Feller square-root process. Exact closed-form expressions for the foci mean and variance enable efficient likelihood-based inference of repair kinetics and effective molecular noise from single-cell $\gamma$-H2AX data without repeated Monte Carlo simulation. Coupling these kinetics to a cumulative damage hazard through the Feynman-Kac formalism links microscopic damage dynamics to macroscopic survival and provides a mechanistic mapping to the classical LQ form. Sensitivity analysis further shows that physical damage induction acts additively, whereas survival depends nonlinearly on repair rate, damage hazard, and stochastic noise. Increased noise enhances population survival by broadening the distribution of cumulative damage, consistent with Jensen's inequality and a bet-hedging-like survival advantage for cells experiencing transiently low damage loads. This framework therefore connects stochastic single-cell biophysics with population-level radiosensitivity and provides a tractable basis for studying how cellular heterogeneity shapes survival under radiation stress.