![]() |
|
![]() |
|
![]() |
|
![]() |
![]() ![]() |
  |
Methods. Randomness of DNA double strand break induction and misrejoining is analyzed and adopted as a working hypothesis. Proximity effects are approximated by using interaction sites. Model results, involving two adjustable parameters, are calculated using CAS (chromosome aberration simulator) Monte Carlo computer software. CAS emulates the specific painting protocol and aberration classification system used in an experiment, allowing very detailed tests of the models.
Results. To reasonable approximation, breakage-and-reunion model predictions on aberration types and dose-response relations are consistent with low-LET FISH results, including two large, elaborate, one-colour data sets. The recombinational-repair model, however, has three problems: it underpredicts by about 10-fold the frequency of the visibly complex aberration patterns most commonly observed with one-colour FISH; it is inconsistent with some observed multi-colour FISH patterns; and in at least one case it gives a dose-response relation for apparently simple aberrations which is, despite the admixture of complex aberrations, sub-linear due to distortion.
Conclusions. The random breakage-and-reunion model gives comprehensive baseline predictions, sufficiently accurate for organizing experimental results coherently. Data on visibly complex aberrations speak strongly against random recombinational repair being the dominant aberration formation pathway at low LET doses of several Gy or more.
Indexing Phrases. Ionizing radiation; DNA double strand breaks; chromosome aberration models; computer simulations; dose-response relations; recombinational repair.
Correspondence to: R.K. Sachs, Dept. Mathematics, Evans Hall, UCB, Berkeley, CA 94720. Voice: 510-642-4384; Fax: 510-642-6726.
email: sachs@math.berkeley.edu;
/sachs/