A double-strand break is not a signal by itself. The cell must first detect it, then decide whether the damage is repairable. Two related kinases do the detection. ATM (ataxia-telangiectasia mutated) is recruited directly to the break by the MRN complex (MRE11–RAD50–NBS1), which binds the broken ends and holds them together. Once at the break, ATM autophosphorylates and becomes active. ATR (ataxia-telangiectasia and Rad3-related) responds to a different structure: single-stranded DNA coated by RPA, which appears when a fork stalls or when resection exposes a template strand. ATR is recruited by ATRIP and activated by TOPBP1.
Both kinases converge on the checkpoint kinases Chk2 (downstream of ATM) and Chk1 (downstream of ATR). The Chk kinases phosphorylate p53 at serine 20, which disrupts the MDM2–p53 interaction. MDM2 normally binds p53 and targets it for ubiquitin-mediated degradation; when that binding is blocked, p53 escapes degradation and accumulates. ATM also phosphorylates p53 directly at serine 15, adding a second stabilizing modification.
Accumulated p53 acts as a transcription factor. Among its targets are CDKN1A, which encodes p21, a cyclin-dependent kinase inhibitor that blocks the G1/S transition and holds the cell in G1 while repair proceeds. p53 also transcribes pro-apoptotic genes including PUMA and NOXA. The choice between arrest and death is therefore not a single switch but a balance: if repair succeeds, p53 levels fall and the cell re-enters the cycle; if damage persists, the apoptotic arm dominates. This is the decision point that determines whether a DNA-damaging drug kills the cell or merely pauses it.