Quantification and mapping of alkylation in the human genome reveal single nucleotide resolution precursors of mutational signatures

Benzo[a]pyrene (BaP) is a known human carcinogen (IARC Group 1) found in food, coal tar, as well as cigarettes and other smoke. Its diol-epoxide metabolites (Benzo[a]pyrene diol-epoxide [BPDE]) react with DNA forming DNA adducts, predominantly N2-BPDE-deoxyguanosine (N2-BPDE-dG). While the capacity of BPDEs to alkylate DNA and induce mutations is well known, little is known about how the genomic features influence the accumulation of DNA damage at a genome-wide level. To bridge this gap, we developed a single-nucleotide resolution damage sequencing method to map N2-BPDE-dG in a BPDE exposed human lung cell line, and combined this analysis with mass spectrometry to quantify the total absolute levels of the adduct in the genome. Comparing damage abundance with DNase hypersensitive sites, transcription levels, and other genome annotation data showed that although the overall adduct levels increased with increasing concentration of BPDE, the genomic distribution patterns of N2-BPDE-dG were stable and correlated with the genomic features related to chromatin state and transcriptional activities. In addition, we extracted the preferred local DNA sequence contexts for N2-BPDE-dG, i.e., its DNA damage signature, and found that it was highly similar to the mutational signatures identified from smoking-related lung cancers. These results suggest that genomic features and sequence contexts are important in shaping the landscape of DNA damage arising from chemical exposures and provide an effective strategy for linking single-nucleotide resolution damage sequencing data with cancer mutations. Sequencing of benzo[a]pyrene-diol-oxirane-dG (BPDE-dG) in naked DNA and cells that were exposed to various BPDE concentrations. Each condition includes three biological replicates.