Abstract
During DNA repair, ATM-induced H2AX histone phosphorylation and MDC1 recruitment spread megabases beyond the damage site. While loop extrusion has been suggested to drive this spread, the underlying mechanism remains unclear. Herein, we provide two lines of evidence that loop extrusion is not the only driver of damage-induced γH2AX spread. First, cohesin loader NIPBL and cohesin subunit RAD21 accumulate considerably later than the phosphorylation of H2AX and MDC1 recruitment at micro-IR-induced damage. Second, auxin-induced RAD21 depletion does not affect γH2AX/MDC1 spread following micro-irradiation or DSB induction by zeocin. To determine if diffusion of activated ATM could account for the observed behavior, we measured the exchange rate and diffusion constants of ATM and MDC1 within damaged and unperturbed chromatin. Using these measurements, we introduced a quantitative model in which the freely diffusing activated ATM phosphorylates H2AX. This model faithfully describes the dynamics of ATM and subsequent γH2AX/MDC1 spread at complex DNA lesions.