A review of the bystander effect and its implications for low-dose exposure.

Current models for the interaction between ionising radiation and living cells or tissues are based on direct genetic damage produced by energy deposition in cellular DNA. An important observation which has questioned this basic assumption is the radiation-induced bystander response, in which cells which have not been directly targeted respond if their neighbours have been exposed. This response predominates at low doses of relevance to radiation risk analysis (<0.2 Gy) and therefore needs to be fully characterised. The development of microbeams, which allow individual cells within populations to be targeted with precise doses of radiation, has provided a useful tool for quantifying this response. The authors' studies have targeted individual human and mouse cells with counted protons and helium ions and monitored neighbouring cells for the production of bystander responses. Bystander responses have been measured after exposures as low as a single proton or helium ion delivered to an individual cell. An important aspect is that these responses saturate with increasing dose to the single target cell, thus the relative roles of direct and indirect (non-targeted) responses change with dose. Studies with multicellular, tissue-based models are providing evidence that bystander responses may have a complex phenotype involving multiple pathways and the overall response may be a balance between multiple signalling processes and responses to radiation exposure. Current models for radiation risk assume a linear non-threshold response and have generally been extrapolated from high-dose exposures. The involvement of competing processes at low doses may have important consequences for understanding the effects of low-dose exposure.