Mechanisms of radiation-induced chromatid breaks

Chromatid breaks are thought to result from DNA double-strand breaks (dsb) but the mechanisms are not yet understood. The early (but still prevailing) 'breakage-first' hypothesis fails to explain the large size of chromatid breaks; many of which are estimated to represent the apparent loss of between 15 and 45 Mbp (up to 30% of an average chromatid). The alternative 'exchange' hypothesis of Revell has potential for explaining the large sizes of deletions, but assumes the interaction of two lesions which therefore predicts a quadratic dependence of chromatid breaks on radiation dose. The exchange hypothesis is not tenable for mammalian cells since chromatid breaks are observed to be induced linearly with dose in both human and rodent cells. An alternative 'signal' model of chromatid breaks is outlined whereby a single dsb, occurring within a large looped chromatin domain, is signalled (possibly by molecules such as DNAPK or ATM protein) and triggers the cell to undergo a recombinational exchange, either within a chromatid or between sister chromatids. If incomplete, such recombinational exchanges would appear as chromatid breaks at metaphase. It is suggested that the large looped chromatin domains could be equivalent to one or more likely several replication 'factories' in which the DNA processing enzymes required for exchange formation would be located.