Complex damage, low doses and Bystander effects

Ionizing radiations of all types can produce a wide array of biological effects that overlap with those produced by many other genotoxic agents. Ionizing radiation, however, has unique features that sets it apart and these features are likely to dominate its modes of action and consequences, especially at low doses. Radiation insult is always in the form of highly structured 'tracks' along the paths of moving charged particles. This feature largely determines the spectrum of initial DNA damage produced in cells or people, the repairability of the damage by cellular processes, and its spatial and temporal distribution in the irradiated material. At the DNA level a substantial proportion of the initial damage is clustered over a few base pairs within a few nanometres of the track, thereby forming local complex damage consisting of several strand breaks and/or damaged bases. Simple double-strand breaks are in the minority compared to more complex combinations, even from so-called sparsely-ionizing (or low-LET) radiations. At low doses, the nature of the radiation tracks determines also the distributions of the primary complex lesions over the larger subcellular, cellular and tissue scales. At exposure levels of natural background radiation and also in most situations of occupational or diagnostic medical exposure, cells are traversed by single isolated individual tracks, so it is the biological capabilities of these single tracks that must determine the probability of harmful effects, if any. Thus, questions relating to the linear no-threshold hypothesis versus thresholds, hormesis, hypersensitivity, and so on, reduce at the low-dose end to understanding the capabilities and consequences of a single track, or a small number, and the persistence and range of its influence in tissue. Much is now understood about the nature and consequences of immediate 'targeted' radiation damage in cells from radiation tracks in them and near their DNA. However, over the last decade, there has been a wealth of new data showing that there are also 'untargeted' intracellular and extracellular processes that can be observed as high frequencies of delayed cellular effects (eg genomic instability) or effects in cells that have not themselves been irradiated ('bystander' effects). In considering risks from low doses and/or low dose rates of radiation, what are the relative contributions from untargeted mechanisms and how should this balance be applied to guide extrapolations of robust epidemiological data to the low exposure levels of practical relevance?