Investigating the nature of chromatid breaks produced by restriction endonucleases.

It is a basic assumption of the breakage-and-reunion theory that the majority of open chromatid breaks seen at metaphase are the residue of unrejoined primary breaks that have neither restituted nor rejoined illegitimately to form exchange aberrations. If Chinese hamster chromosomes with BrdU sister-chromatid differentiation are irradiated, and chromatid aberrations scored from G2 cells, some 15-20% of open breaks show a colour-jump at the point of discontinuity, indicating a two-lesion intrachange origin. Since we see complete forms of several intrachanges whose incomplete forms will also look like breaks, but devoid of a colour-jump, it appears that a substantial proportion of observed breaks are intrachange derived. Experiments to date show that the colour-jump proportion is constant, irrespective of radiation dose, radiation quality, BrdU concentration and hamster cell origin. It is the same for the very low "spontaneous' breaks found in control samples. Restriction endonucleases (RE) can be introduced into cells by various poration methods, and are highly efficient at producing all types of aberrations. This is taken as strong evidence that DNA dsb are significant lesions triggering aberrations. One might anticipate, therefore, that observed breaks will be predominantly unrejoined dsb, and the proportion of colour-jump break correspondingly low. We tested this supposition using three RE; Alu 1, a blunt-end cutter, Sau3A 1, a cohesive-end cutter, both with a short life-time in vivo, and Mbo 1, an isoschizomer of Sau3A 1, which has a long cutting life-time in vivo. Although there were differences in absolute yields of breaks, and of relative frequencies of aberration types recovered, the proportion of colour-jump breaks was as high as that in a parallel X-ray experiment, and fell well within the range encountered in all our previous experiments. It is difficult to reconcile this universal constancy of colour-jump breaks with the expectations of breakage-and-reunion theory, where the occurrence of two-break events must be a treatment variable. Rather, our results suggest that most open breaks are secondary, resulting from a regular intrachange processing mechanism.