As an aberrant base in DNA, uracil is generated by either deoxyuridine (dU) misincorporation or cytosine deamination, and involved in multiple physiological and pathological processes. Genome-wide profiles of uracil are important for study of these processes. Current methods for whole-genome mapping of uracil all rely on uracil-DNA N-glycosylase (UNG) and are limited in resolution, specificity, and/or sensitivity. Here, we developed a UdgX cross-linking and polymerase stalling sequencing (“Ucaps-seq”) method to detect dU at single-nucleotide resolution. First, the specificity of Ucaps-seq was confirmed on synthetic DNA. Then the effectiveness of the approach was verified on two genomes from different sources. Ucaps-seq not only identified the enrichment of dU at dT sites in pemetrexed-treated cancer cells with globally elevated uracil but also detected dU at dC sites within the “WRC” motif in activated B cells which have increased dU in specific regions. Finally, Ucaps-seq was utilized to detect dU introduced by the cytosine base editor (nCas9-APOBEC) and identified a novel off-target site in cellular context. In conclusion, Ucaps-seq is a powerful tool with many potential applications, especially in evaluation of base editing fidelity.

Protein–DNA damage interactions are critical for understanding the mechanism of DNA repair and damage response. However, due to the relatively random distributions of UV-induced damage and other DNA bulky adducts, it is challenging to measure the interactions between proteins and these lesions across the genome. To address this issue, we developed a new method named Protein-Associated DNA Damage Sequencing (PADD-seq) that uses Damage-seq to detect damage distribution in chromatin immunoprecipitation-enriched DNA fragments. It is possible to delineate genome-wide protein–DNA damage interactions at base resolution with this strategy. Using PADD-seq, we observed that RNA polymerase II (Pol II) was blocked by UV-induced damage on template strands, and the interaction declined within 2 h in transcription-coupled repair-proficient cells. On the other hand, Pol II was clearly restrained at damage sites in the absence of the transcription–repair coupling factor CSB during the same time course. Furthermore, we used PADD-seq to examine local changes in H3 acetylation at lysine 9 (H3K9ac) around cisplatin-induced damage, demonstrating the method's broad utility. In conclusion, this new method provides a powerful tool for monitoring the dynamics of protein–DNA damage interaction at the genomic level, and it encourages comprehensive research into DNA repair and damage response.