High molecular weight genomic DNA was isolated from cells lysed in the presence of SDS and proteinase K extracted twice with phenol-chloroform, and precipitated with ethanol in presence of ammonium acetate. Samples were subjected to nick-translation by DNA polymerase I in the presence of deoxynucleotide triphosphates including digoxigenin labeled dUTP. To restrict labeling to a small patch of DNA in order to increase the resolution of mapping, dideoxynucleotides were included in the reaction to inhibit excessive chain elongation by DNA polymerase I. As a control for the labeling, samples were also nick-translated without digoxigenin-labled nucleotides. DNA was sheared by sonication and labeled fragments were immunoprecipitated with anti-digoxigenin antibody, purified and subjected to Illumina library preparation and sequencing (Figure 1a).

After isolating high molecular weight genomic DNA as above, the double-stranded DNA ends were 3'-end tailed with TdT in the presence of biotinylated nucleotides. Finally, samples were sonicated, and biotinylated DNA was streptavidin-selected. In parallel, as a negative-control for labeling and selection, 3'-tailing was performed in the absence of biotinylated nucleotides. To remove the biotinylated-tails, samples were treated with S-1 nuclease; the resulting DNA was purified and subjected to Illumina library preparation and sequencing (Figure 1b).

To demonstrate the specificity and sensitivity of SSBs and DSBs capture, we performed a pilot experiment on HCT116 cells briefly treated with ETO. The drug stabilizes the covalent Top2-cleaved DNA complex that is a transient intermediate in the catalytic cycle of the enzyme. Because Top2 is a homodimer with each monomer cleaving a single DNA strand, denaturation of topoisomerase in the complex with DNA by SDS (Figure 2a) could result in a permanent DSB [15]. Since the drug inhibits each Top2 subunit independently, contrary to conventional expectation, ETO treatment generates mostly SSBs, to the detriment of DSBs (Figure 2b). SSBs occur when only one Top2 monomer is inhibited, leaving the other subunit free to rejoin the DNA ends [16]. Top2 performs DNA cleavage via tyrosyl-active site residues, establishing a covalent phosphotyrosyl bond linking the enzyme to the 5'-terminus of the DNA. It also generates a 3'-hydroxyl moiety on the opposite terminus of the cleaved strand, which is a proper substrate for nick-translation. Consequently, SSB-Seq in parallel with ETO treatment is highly specific for Top2 activity. This procedure does not detect SSBs resulting from Topoisomerase I (Top1) activity or SSBs bearing damaged DNA termini unless processed by DNA-repairing enzymes (Figure 2d) [17]. Within the Top1–DNA cleavable complex, the enzyme is covalently bound with the phosphate to the 3' end of DNA (Figure 2c). To detect these SSBs it would be necessary to reconstitute a free 3'-OH terminus to serve as a substrate for DNA polymerase I. Therefore, even if Top1 activity were indirectly altered in cells due to Top2 inhibition, the SSBs detected in our experiments, reflect only catalytically engaged Top2.

To validate our experimental approach, we briefly exposed cells to a low dose of ETO, labelled SSBs and DSBs and quantified the DNA recovered after antibody/streptavidin selection. As expected, cells exposed to the drug accumulate more SSBs, compared to the untreated sample (Figure 3). On the contrary, the amount of DBSs was lower than the untreated control.

After validation, we next applied the SSB- and DSB-Seq to map Top2 activity genome-wide. Top2 is functional in a wide range of biological contexts such as DNA replication, chromosome segregation and transcription [18]. The twin supercoiling model [19] provides an important rationale for the Top2-requirement during transcriptional elongation where transcribing RNA polymerase twists the DNA through its active site, driving torsional stress ahead and behind it. However, how and where Top2 works during transcription remains poorly investigated. The relationship between transcription and Top2-associated breaks was revealed by deep-sequencing the DNA from SSB-Seq and DSB-Seq and showed that DNA breaks displayed a striking preference for the 5'-end of the genes (Figure 4). As expected, we observed a similar pattern for DSBs and SSBs, indicating that during genomic DNA preparation the main pool of DNA double-stranded breaks results from the enzymatic activity of Top2.

As postulated in the twin supercoiled domain model, the increase in gene expression should increase the level of transcription-generated torsional stress and require more Top2. Therefore we computed the break-density at genes with high, medium and low output, as measured by microarray experiments [20]. The analysis revealed that Top2-breaks were most enriched at transcriptional start sites (TSS) compared to the rest of the gene. Notably, genes with elevated expression exhibited higher frequency of Top2-breakage than did lower expressed or silent genes (Figure 5). These observations complement and extend a recent work showing for a small set of the genes that Top2 preferentially binds promoters of the highly expressed genes [21]. Overall these finding suggest that Top2 exerts its enzymatic activity mainly at TSS in a manner dependent on the level of promoter output. These results also indicate that the sites of breakage in standard DNA preparations may reflect Top2 halted in mid-catalytic cycle at the time of cell lysis.

DNA breakage is intrinsically linked to the genome biology. Breaks can be hazardous for genome stability but also a means to alter the arrangement, conformation and topology of chromatin [14]. Our methods provide a useful tool to map SSBs and DSBs in different cells and physiological or pathological conditions, on a genome-wide scale. Although past research, mainly based on chromatin immunoprecipitation (ChIP) approach, has provided some information about Top2 localization [21], these studies cannot discriminate binding from activity. Previous works shows that Top2 activity is favored by the crossing of DNA segments [22], which likely occurs at promoters where plectonemes form in negatively supercoiled DNA unbuffered by chromatin rearrangement [23]. Accordingly, we showed that the majority of Top2-associated breaks occur at promoters where high negative torsional stress accumulates in the wake of RNA polymerase movement [21].

Human colon cancer cells HCT116 were grown in DMEM supplemented with 10% heat-inactivated FCS. Etoposide (Sigma, St. Louis, MO, USA) treatment was performed on exponentially growing cells at 37 °C for 20 min with a final concentration of 20 μM.

To prepare high molecular weight DNA, 8 × 10⁷ cells were washed twice with ice-cold PBS and lysed with 10 mL of lysis buffer (10 mM Tris–HCl pH 8.0, 100 mM EDTA pH 8.0, 0.5% SDS). Lysates were collected, digested overnight with 200 μg/mL of proteinase K (Roche, Indianapolis, IN, USA) at 52 °C and DNA was extracted twice with phenol, once with phenol chloroform, and then ethanol precipitated in the presence of 2 M ammonium acetate (Sigma). The sample was then treated with 5 μg of pancreatic RNase (Roche) for 1 h at 37 °C, adjusted to 0.5% SDS and incubated for 1 h at 55 °C with proteinase K (200 μg/mL). DNA was extracted twice with phenol and precipitated with ethanol in the presence of 2 M ammonium acetate. Finally, the pellet was resuspended in 1 mL of TE (10 mM Tris–HCl, 1 mM EDTA pH 8.0) and incubated at room temperature for 12 h with gentle rotation. DNA size was determined by gel electrophoresis, and its concentration was measured with a NanoDrop ND-1000 spectrophotometer (Thermo Scientific, Rochester, NY, USA).

SSBs were labeled by nick translation. 500 μg of DNA was incubated for 40 s at 16 °C with a mixture of 200 μM of dATP, dGTP, dCTP and 20 μM of digoxigenin-11-dUTP (Roche), 117 μM of ddATP, ddGTP, ddCTP (Roche) and 1000 units of Escherichia coli DNA Polymerase I (New England Biolabs, Ipswich, MA, USA). As a control for labeling, 500 μg of DNA was incubated with the same reagents except digoxigenin-11-dUTP was substituted by 20 μM of dTTP. The reactions were stopped with 50 μM EDTA and DNA purified with phenol, precipitated twice in the presence of ethanol and 2.5 M ammonium acetate, and DNA was sheared by sonication to an average fragment size of 200–400 bp. Sonication was performed with an ultrasonic sonicator Bioruptor (Diagenode, Denville, NJ, USA) by pulsing 30 times for 30 s at medium power and incubating on ice for 30 s between each pulse. The samples were incubated at 4 °C overnight with 10 μg of Anti-digoxigenin antibody (Roche) with gentle rotation. Immunocomplexes were recovered by addition of 60 μL of Protein G-Sepharose beads (Roche) and incubated for 4 h at 4 °C. The beads were washed once with PBS, three times with NP-40 buffer (20 mM Tris–HCl pH 8.0, 137 mM NaCl, 10% Glycerol, 1% NP-40, 2 mM EDTA pH 8.0); twice with TE (10 mM Tris–HCl, 1 mM EDTA pH 8.0); and finally resuspended in 200 μL of TE. Each wash was performed for 10 min by gentle agitation followed by 4 min of centrifugation at 1500× g. The pellets were adjusted to 0.5% SDS and digested with proteinase K (200 μg/mL) at 65 °C overnight. DNA was purified using QIAquick PCR Purification Kit Protocol (QIAGEN, Valencia, CA, USA) according to the manufacturer’s instructions and quantified.

This method was modified from [24]. To label DSBs 500 μg of DNA was biotinylated by 3'-end tailing reaction in 3 mL of TdT buffer (Roche) with 24,000 U TdT (Roche), 0.5 mM dCTP, and 0.02 mM Biotin-16-dUTP (Roche) at 37 °C for 30 min. Reactions were stopped by adding EDTA to a 20 mM final concentration. As a control of labeling, 500 μg of DNA was incubated with the same reagents substituting TTP for Biotin-16-dUTP. To remove unincorporated biotin, samples were extracted with phenol chloroform, precipitated twice with 2 M ammonium acetate and ethanol and dissolved in 200 mL of TE buffer. Biotinylated DNA was sonicated, as above, to generate 200–400 bp DNA fragments. Then, biotinylated fragments were captured with streptavidin-coated beads by using Dynabeads kilobaseBINDER Kit (Invitrogen Dynal, Oslo, Norway) according to the manufacturer’s protocol. After 4 washes at 50° and 4 washes at room temperature with 10 mM Tris–HCl pH 7.5, 1.0 mM EDTA, and 2.0 M NaCl, biotin-streptavidin complexes were disrupted by incubating them in 10 mM Tris–HCl pH 7.5, 1 mM EDTA, 1 M NaCl, and 2 M β-mercaptoethanol at 75 °C for 4 h. Free DNA fragments were purified with a QIAquick PCR Purification Kit (QIAGEN). To remove biotinylated tails from DNA, samples were incubated with 30 U of S1 nuclease (Fermentas, Clen Burnie, MD, USA) in 110 μL of recommended buffer for 30 min at 37 °C. DNA was purified with a QIAquick PCR Purification Kit (QIAGEN).

The Epicentre DNA END-Repair kit (Epicentre Biotechnologies, Madison, WI, USA) was used to generate blunt-ended DNA. DNA was incubated for 45 min at room temperature with a mixture of End repair buffer (33 mM Tris–acetate pH 7, 66 mM potassium acetate, 10 mM magnesium acetate, 0.5 mM DTT), 0.25 mM of each dNTPs, 1 mM ATP, and 1 μL End-Repair Enzyme mix (T4 DNA polymerase + T4 PNK). After purification, the blunt-ended DNA was treated with 15 units of Klenow (exo-) for 30 min at 37 °C in the presence of 0.2 mM dATP to generate a protruding 3'-A base used for adaptor ligation. Illumina adapter was ligated to the end of DNA fragments by incubating with 0.1 μL Adaptor oligo mix and 1000 units of T4 DNA ligase at room temperature for 30 min. After one step of DNA purification using QIAquick PCR Purification Kit (QIAGEN) the DNA was eluted. A size selection of the adapter ligated DNA was performed through 2% E-Gel (Invitrogen, Carlsbad, CA, USA) electrophoresis. The gel slice, around the 200–400 bp region was excised, then DNA was extracted using the MinElute gel extraction kit (QIAGEN) in a final volume of 12 μL elution buffer. The DNA was then amplified for 18 cycles using Illumina primers (Fw: 5'-ACACTCTTTCCCTACACGACGC-3'/Rv: 5'-CAAGCAGAAGACGGCATACGAGC-3') according to the following protocol: 98 °C for 30 s; 65 °C for 30 s; 72 °C for 30 s. The PCR product was obtained by excising 220–500 bps DNA from a 2.5% agarose gel and purifying it through a Qiagen gel extraction kit (QIAGEN). The purified DNA was used directly for cluster generation and sequencing analysis using the Illumina Genome Analyzer (Illumina, San Diego, USA) following manufacturer protocols.

The oligo microarray data (Affymetrix Human Genome U133 Plus 2.0 Array) of total RNA from HCT116 cells (four replicates) were retrieved from the NCBI Gene Expression Omnibus (National Center for Biotechnology Information, Bethesda, MD, USA) under the GEO accession number GSE7161. CEL files were analyzed using R environment (Free Software Foundation, Boston, MA, USA) with Bioconductor package (Fred Hutchinson Cancer Research Center, Seattle, WA, USA). We used a GeneChip robust multi-array analysis, gcrma, method for background correction and normalization of the raw probe measurements to attain expression values. The maximum expression of Affymetrix U133 Plus 2.0 identifiers (Affymetrix, Santa Clara, CA, USA) mapped to human genes annotated in Ensembl 60 [25] was chosen as the gene expression value. Genes were split into quartile groups based on their expression values.

Sequencing data were preprocessed using the Illumina Analysis Pipeline (image analysis and base calling, Illumina, San Diego, CA, USA). The short reads of length 36 bp were aligned to the human genome (hg19) using Bowtie 2 program (version 2.2.2) [26]. To minimize potential PCR bias redundant reads were removed. The profiles of average distribution of sequencing reads along normalized protein-coding gene bodies were generated using ngs.plot (version 2.0.8) [27].