Radiation-induced breakpoint misrejoining in human chromosomes: random or non-random?

Purpose: To investigate whether radiation-induced misrejoining of chromosome breakpoints is randomly or non-randomly distributed throughout the human genome. Materials and methods: Data were combined from as many published cytogenetic studies as possible. The percentage of radiation-induced breaks per megabase (Mb) of DNA between all human chromosomes was calculated, and the observed and expected numbers of breakpoints based on DNA content between and within chromosomes were compared. Results: A DNA-proportional distribution of breakpoints in 14 autosomes and a statistically significant deviation from proportionality in the other eight autosomes and the sex chromosomes was found. Regression analysis showed no significant change in breakpoint frequency per Mb of DNA relative to autosome size. Analysis between chromosome arms showed a non-random distribution of induced breakpoints within certain autosomes, particularly the acrocentrics. In cases of non-random distributions, a prevalence of events was found at heterochromatic regions and/or telomeres, and a clustering of breakpoints was found near the centromeres of many chromosomes. Conclusions: There is an approximately linear proportionality between autosomal DNA content and observed breakpoint number, suggesting that subsets of autosomes can be used to estimate accurately the overall genomic frequency of misrejoined breakpoints contingent upon a carefully selected subset. However, this conclusion may not apply to the sex chromosomes. The results also support the influence of chromatin organization and/or preferential DNA repair/misrejoining on the distribution of induced breakpoints. However, these effects are not sufficient at a global level to dismiss the value of cytogenetic analysis using a genome subset for biodosimetry.