The induction of DNA strand breaks by fission neutrons was studied in aqueous plasmid (pBR322) DNA under aerobic conditions for a wide range of hydroxyl radical ( ${}^{\bullet}{\rm OH}$ ) scavenger concentrations and was compared to the induction of strand breaks by 60 Co γ rays. Strand breaks were measured using agarose gel electrophoresis coupled with sensitive ${}^{32}{\rm P}\text{-}\text{based}$ phosphor imaging. Yields are reported for DNA single-strand breaks (SSBs) and double-strand breaks formed linearly with dose (αDSBs). The fraction of αDSBs that were dependent on the multiply damaged site (MDS) or clustered damage mechanism was also calculated using a model. G values for SSBs and αDSBs declined with increasing ${}^{\bullet}{\rm OH}$ scavenging capacity. However, with increasing ${}^{\bullet}{\rm OH}$ scavenging capacities, the decrease in yields of strand breaks for fission neutrons was not as pronounced as for γ rays. The percentage of αDSBs for γ rays was dependent on ${}^{\bullet}{\rm OH}$ scavenging capacity, appearing negligible at low scavenging capacities but increasing at higher scavenging capacities. In contrast, fission neutrons induced high percentages of αDSBs that were approximately independent of ${}^{\bullet}{\rm OH}$ scavenging capacity. The levels of αDSBs formed by the MDS mechanism after exposure to fission neutrons are consistent with the expected distinctive features of high-LET energy deposition events and track structure. The results also confirm observations made by others that even for low-LET radiation, the MDS mechanism contributes significantly to DNA damage at cell-like scavenging conditions.

Although double-strand breaks have long been recognized as an important type of DNA lesion, it is well established that this broad class of damage does not correlate well with indicators of the effectiveness of radiation at the cellular level. Assays of double-strand breaks do not distinguish the degree of complexity or clustering of singly damaged sites produced in a single energy deposition event, which is currently hypothesized to be key to understanding cellular end points. As a step toward this understanding, double-strand breaks that are formed proportionally to dose in plasmid DNA are analyzed from the mechanistic aspect to evaluate the yield that arises from multiply damaged sites as hypothesized by Ward (Prog. Nucleic Acid Res. Mol. Biol. 35, 95-125, 1988) and Goodhead (Int. J. Radiat. Biol. 65, 7-17, 1994) as opposed to the yield that arises from single hydroxyl radicals as hypothesized by Siddiqi and Bothe (Radiat. Res. 112, 449-463, 1987). For low-LET radiation such as γ rays, the importance of multiply damaged sites is shown to increase with the solution's hydroxyl radical scavenging capacity. For moderately high-LET radiation such as 100 keV/μm helium ions, a much different behavior is observed. In this case, a large fraction of double-strand breaks are formed as a result of multiply damaged sites over a broad range of scavenging conditions. Results also indicate that the RBE for common cellular end points correlates more closely with the RBE for multiply damaged sites than with the RBE for total double-strand breaks over a range of LET up to at least 100 keV/μm.