High-resolution Repli-Seq defines the temporal choreography of initiation, elongation and termination of replication in mammalian cells

Takayo Sasaki,Peiyao A. Zhao,David M. Gilbert

doi:10.1186/s13059-020-01983-8

Abstract

BackgroundDNA replication in mammalian cells occurs in a defined temporal order during S phase, known as the replication timing (RT) programme. Replication timing is developmentally regulated and correlated with chromatin conformation and local transcriptional potential. Here, we present RT profiles of unprecedented temporal resolution in two human embryonic stem cell lines, human colon carcinoma line HCT116, and mouse embryonic stem cells and their neural progenitor derivatives.ResultsFine temporal windows revealed a remarkable degree of cell-to-cell conservation in RT, particularly at the very beginning and ends of S phase, and identified 5 temporal patterns of replication in all cell types, consistent with varying degrees of initiation efficiency. Zones of replication initiation (IZs) were detected throughout S phase and interacted in 3D space preferentially with other IZs of similar firing time. Temporal transition regions were resolved into segments of uni-directional replication punctuated at specific sites by small, inefficient IZs. Sites of convergent replication were divided into sites of termination or large constant timing regions consisting of many synchronous IZs in tandem. Developmental transitions in RT occured mainly by activating or inactivating individual IZs or occasionally by altering IZ firing time, demonstrating that IZs, rather than individual origins, are the units of developmental regulation. Finally, haplotype phasing revealed numerous regions of allele-specific and allele-independent asynchronous replication. Allele-independent asynchronous replication was correlated with the presence of previously mapped common fragile sites.ConclusionsAltogether, these data provide a detailed temporal choreography of DNA replication in mammalian cells.

Highlights

DNA replication in eukaryotes proceeds in a defined temporal order known as the replication timing (RT) programme [1]
One of the primary assays for genomewide RT in mammalian cells has been E/L Repli-Seq in which cells labelled with BrdU for 10–20% of S phase are sorted into early and late S fractions and RT profiles are generated from the log2 ratio of read enrichment in the BrdU-immunoprecipitated early fraction to that in the late fraction (E/L) [8]
We previously developed a single-cell Repli-Seq method to address this problem, concluding that the RT programme is stable from cell to cell, but this method suffers from low resolution due to the limited breadth of whole genome single-cell sequencing and the single temporal snapshot obtained from each cell [16, 17]

Summary

Introduction

DNA replication in eukaryotes proceeds in a defined temporal order known as the replication timing (RT) programme [1]. TTRs are hypothesised to consist mainly of uni-directional forks, occasionally accelerated by origin firing [2, 14, 15] Measurements supporting these hypotheses derive from prior RT profiling methods that smoothed data over hundreds of kilobases (kbs) and/or relied on long metabolic labels incorporated over large stretches (> 200 kb) of DNA and so lacked the resolution to identify sites of replication initiation within CTRs and TTRs. measurements supporting these hypotheses derive from prior RT profiling methods that smoothed data over hundreds of kilobases (kbs) and/or relied on long metabolic labels incorporated over large stretches (> 200 kb) of DNA and so lacked the resolution to identify sites of replication initiation within CTRs and TTRs Since these methods average all stochastic variation in a cell population, they do not permit one to determine the degree of cell-to-cell RT variation. DNA replication in mammalian cells occurs in a defined temporal order during S phase, known as the replication timing (RT) programme. We present RT profiles of unprecedented temporal resolution in two human embryonic stem cell lines, human colon carcinoma line HCT116, and mouse embryonic stem cells and their neural progenitor derivatives

Methods

Results

Conclusion