Clustering of radiation-produced breaks along chromosomes: modelling the effects on chromosome aberrations

For high-LET radiations, and perhaps even for hard X-rays, DNA double-strand breaks (dsb) are clustered nonrandomly along chromosomes; disproportionately, many inter-dsb segments are less than a few Mbp (10(6) base pairs). The implications of such dsb clustering for chromosome aberrations are analysed.

Chromosome segments between different dsb within one dsb cluster are assumed too small to detect in the aberration assay. Enumeration or Monte-Carlo computer simulations are used to compute the relative frequencies of many observable aberration patterns: apparently simple or visibly complex. The theoretical predictions are compared with X-ray data for human fibroblasts, involving painted chromosomes 1, 2, 4, 5, 7 or 13.

Surprisingly, cryptic dsb multiplicity does not affect the frequency ratios predicted for aberration patterns by a random breakage-and-rejoining model. The model is generally consistent with current data on many different types of aberrations, whether or not dsb usually occur in cryptic clusters. For a Revell-type exchange model, however, the predictions do depend on clustering configurations; they gradually approach the predictions of the breakage-and-rejoining model as average cluster multiplicity increases. The model is consistent with the data, for example with the ratio of visibly complex to apparently simple aberrations, only if there is considerable dsb clustering even at low-LET, with approximately 1.5 or more reactive dsb per cluster on average.