Novel Method For High-Resolution Detection of Double-Strand DNA Breaks
Principle Investigator: Malgorzata Rowicka-Kudlicka, PhD.
DNA double-strand breaks (DSBs) are caused by physical and chemical agents or replication fork collapse, and physiologically occur during apoptosis, meiotic crossing-over, and gene rearrangements. Despite the extensive knowledge on their sensing and repair, the precise distribution of DSBs in different cells and conditions remains obscure. No method has been able so far to map DSBs with high resolution (~100b). Megabase-long cytogenetic bands that break upon partial inhibition of DNA synthesis – common fragile sites (CFSs) – have been identified on multiple chromosomes by G-staining of metaphase spreads, but the exact breakpoints are not known due to poor resolution of the method.
ChIP-on-chip with antibodies against the phosphorylated histone H2A.X ( H2A.X), a well defined DSB marker, has been used to map fragile sites in S.cerevisiae. Despite the great resolution improvement over G-banding, the applicability of this method is limited by the fact that phosphorylation of H2A.X can spread many kilobases away from a DSB, and mark structures other than breaks, such as sex and Barr bodies. Therefore, more resolved and specific genome-wide methods are needed to gain insights in the biology of DNA double-strand breaks in different cell types and conditions.
Recently, we developed BLESS method to map DSBs. Our method was comprehensively validated and is suitable for genome-wide mapping of DSBs in various cells and experimental conditions with resolution at least three orders of magnitude better than cytogenetical methods. However, in initial experiments, due to low coverage of sequencing and lack of model-based, sensitive methods for detection of DSBs, we detected only 60 medium-resolution (several kb) and 191 low-resolution (several Mb) breaking hotspots. Here, we propose to improve sensitivity of the experimental method and develop data analysis approach and tools that would allow us to detect thousands of ultra-high-resolution (less than 100b) DSBs using affordable sequencing capacity of no more than one sequencing lane per sample.