Genomic landscape of oxidative DNA damage and repair reveals regioselective protection from mutagenesis

BACKGROUND: DNA is subject to constant chemical modification and damage, which eventually results in variable mutation rates throughout the genome. Although detailed molecular mechanisms of DNA damage and repair are well understood, damage impact and execution of repair across a genome remain poorly defined. RESULTS: To bridge the gap between our understanding of DNA repair and mutation distributions, we developed a novel method, AP-seq, capable of mapping apurinic sites and 8-oxo-7,8-dihydroguanine bases at approximately 250-bp resolution on a genome-wide scale. We directly demonstrate that the accumulation rate of apurinic sites varies widely across the genome, with hot spots acquiring many times more damage than cold spots. Unlike single nucleotide variants (SNVs) in cancers, damage burden correlates with marks for open chromatin notably H3K9ac and H3K4me2. Apurinic sites and oxidative damage are also highly enriched in transposable elements and other repetitive sequences. In contrast, we observe a reduction at chromatin loop anchors with increased damage load towards inactive compartments. Less damage is found at promoters, exons, and termination sites, but not introns, in a seemingly transcription-independent but GC content-dependent manner. Leveraging cancer genomic data, we also find locally reduced SNV rates in promoters, coding sequence, and other functional elements. CONCLUSIONS: Our study reveals that oxidative DNA damage accumulation and repair differ strongly across the genome, but culminate in a previously unappreciated mechanism that safeguards the regulatory and coding regions of genes from mutations. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s13059-018-1582-2) contains supplementary material, which is available to authorized users.