Detection and characterization of replication origins defined by DNA polymerase epsilon

Abstract

AbstractHuman origins of replication (ORI) are recognized by the cellular machinery through mechanisms which are still poorly understood. The process of DNA replication is highly conserved, but the location of ORI may vary considerably from one round of replication to the next, and depends on many factors, making them difficult to map in the genome. We investigated the possibility that at genomic scale, the mutator phenotype associated with DNA polymerase ε exonuclease domain mutation could identify genomic positions of replication origins. Here we report the genome-wide localization of DNA replication origins in POLE-mutated tumors using whole genome sequencing data from The Cancer Genome Atlas project. We show that mutation-based detection of replication origins allows to identify constitutive origins shared between various individuals.We developed a novel method to compare two or more sets of genomic coordinates based on Smith–Waterman-like dynamic programming, which we used to compare replication origin positions obtained using multiple different methods. The comparison allowed us to define a consensus set of replication origins, identified consistently by multiple ORI detection methods. Many DNA features co-localized with ORI, including chromatin loop anchors, G-quadruplexes, S/MARs and CpGs. Among all features, the H2A.Z histone exhibited the most significant association. Our results show that mutation-based detection of replication origins is a viable approach to determining their location and associated sequence features.