DNA break mapping reveals topoisomerase II activity genome-wide

Abstract

Genomic DNA is under constant assault by endogenous and exogenous DNA damaging agents. DNA breakage can represent a major threat to genome integrity but can also be necessary for genome function. Here we present approaches to map DNA double-strand breaks (DSBs) and single-strand breaks (SSBs) at the genome-wide scale by two methods called DSB- and SSB-Seq, respectively. We tested these methods in human colon cancer cells and validated the results using the Topoisomerase II (Top2)-poisoning agent etoposide (ETO). Our results show that the combination of ETO treatment with break-mapping techniques is a powerful method to elaborate the pattern of Top2 enzymatic activity across the genome.

Figure 1

(a) Single-strand breaks (SSBs) are labelled during nick translation using nucleotides covalently linked to digoxigenin (green circle). Genomic DNA is purified, sonicated and immunoprecipitated with anti-digoxigenin antibody (anti-DIG). Precipitated DNA is sequenced; (b) 3' tails of double-strand breaks (DSBs) are ligated to biotinylated nucleotides (orange circle) and after sonication the labelled fragments are captured on streptavidin beads (gray circle). Tails are removed from released fragments and DNA is sequenced. Red bars represent the Illumina adaptors (see Experimental Section for details). Red arrows represent the direction of sequencing.

Figure 2

DSBs (a) and SSBs (b–d) generated in presence of Top2 (a); ETO (b); Top1 (c); and DNA-damaging agents that modify the DNA termini (d). Red arrow represents successful nick translation. Stop sign represents unsuccessful nick translation. Nick translation by DNA polymerase I necessitates a 3'-OH, which is not reconstituted in case of Top1 cleavage or when the DNA termini is damaged (shown by asterisk). In these cases the principal enzymes involved in processing and repair of the ends are listed below the black arrow. TDP1, tyrosyl-DNA phosphodiesterase 1, PNKP, polynucleotide kinase 3'-phosphatase, APE1, AP endonuclease I [17]

Figure 3

DNA quantification of recovered DNAs after SSB- and DSB-Seq in the presence (+ETO) and in the absence (−ETO) of 20 μM of etoposide. The “unlabelled” samples were not incubated with digoxigenin- or biotinylated-nucleotides. The recovered DNA is normalized to the number of cells.

Figure 4

SSB (green) and DSB (blue) profiles at two representative genes (PARD6B and TTPAL) randomly selected. The data are displayed as custom tracks on the UCSC genome browser (University of California, Santa Cruz, CA, USA). The positions of the genes are indicated below the panel. The y-axis shows the number of tags per million reads and the x-axis shows the chromosome coordinates in the genome. Cells were treated with ETO (20 μM for 5 min). Due to the smaller amount of the double-stranded breaks in the genome the DSB profile looks noisier than the SSB profile. Exons are depicted as boxes and introns as lines.

Figure 5

SSBs (top panel) and DSBs (bottom panels) density across genes ranked in 4 percentiles (0%–25%–50%–75%–100%). Data represented as sequence tags per million (TPM).