FLASH-RT does not affect chromosome translocations and junction structures beyond that of CONV-RT dose-rates

The molecular and cellular mechanisms driving the enhanced therapeutic ratio of ultra-high dose-rate radiotherapy (FLASH-RT) over slower conventional (CONV) ionizing radiation (IR) dose-rate are not known. However, attenuated DNA damage and transient oxygen depletion are among a number of proposed models. Here, we tested whether FLASH-IR under physioxic (4% O2) and hypoxic conditions (=2% O2) attenuates detection of genome-wide translocations relative to CONV dose rates and whether any differences identified revert under normoxic (21% O2) conditions. We employed high-throughput rejoin and genome-wide translocation sequencing (HTGTS-JoinT-seq), usingS. pyogenesandS. aureusCas9 “bait” DNA double strand breaks (DSBs), to measure differences in proximal deletions, inversions, and excision circles and their translocation to “prey” genome-wide DSBs generated by CONV (0.08-0.13Gy/s) and FLASH (1x102-5x106Gy/s) dose rates, under varying IR doses and oxygen tensions. Normoxic and physioxic IR exposure in HEK293T cells increased translocations at the cost of decreasing proximal repair but were indistinguishable between CONV and FLASH dose-rates. Although decreased numbers of translocations were recovered as cells transitioned to hypoxic (<2%) conditions, the combined decrease in oxygen tension with IR dose-rate modulation did not reveal significant differences in the level of translocations nor in junction structures, which, for proximal repair, were increased in direct and short microhomologies. We conclude that irrespective of oxygen tension, FLASH IR dose rates produce translocations and junction structures at levels that are indistinguishable from CONV dose rates. Overall design: We employed saCas9:RAG1.1 and spCas9:RAG1B gRNA bait DNA double strand break to map translocation and rejoining outcomes post raditaion treatment (CONV-RT and FLASH-RT) at various oxygen tensions (Normoxia or Hypoxia). Each experimental parameter was tested in triplicate.