Global landscape of replicative DNA polymerase usage in the human genome

Abstract

The division of labour among DNA polymerase underlies the accuracy and efficiency of replication. However, the roles of replicative polymerases have not been directly established in human cells. We developed polymerase usage sequence (Pu-seq) in HCT116 cells and mapped Polε and Polα usage genome wide. The polymerase usage profiles show Polε synthesises the leading strand and Polα contributes mainly to lagging strand synthesis. Combining the Polε and Polα profiles, we accurately predict the genome-wide pattern of fork directionality plus zones of replication initiation and termination. We confirm that transcriptional activity contributes to the pattern of initiation and termination and, by separately analysing the effect of transcription on both co-directional and converging forks, demonstrate that coupled DNA synthesis of leading and lagging strands in both co- directional and convergent forks is compromised by transcription. Polymerase uncoupling is particularly evident in the vicinity of large genes, including the two most unstable common fragile sites, FRA3B and FRA3D, thus linking transcription-induced polymerase uncoupling to chromosomal instability.

Strand-resolved mutagenicity of DNA damage and repair

“…Given that the leading and lagging strands are synthesised by distinct cellular machinery 4,7,[27][28][29] , the observed consistency in the rate and spectrum of mutations generated by DEN lesions is unexpected. These results point to a shared mechanism of lesion bypass, strongly suggesting the same translesion polymerases are recruited by both the leading and lagging strands replication machineries.…”

Ubiquitin ligase TRAIP plays an essential role during the S-phase of unperturbed cell cycle in the resolution of DNA replication – transcription conflicts

“…As per our previous data, the ChIP-Seq detected overall increases to the γ-H2AX signal detected following auxin (IAA) treatment in both repeats (example in Figure 4A). We first checked whether the position of γ-H2AX signal correlated with known DNA replication features: replication initiation sites and replication termination zones (both mapped in HCT116 cells by alternative techniques by 33 and 32 ) and common fragile sites (CFS) 42 . However, no correlation could be found in either of the repeats (Figure 4B and Supplementary Figure 8A).…”

“…We then identified the closest replication origin, as assigned by Koyanagi et al . to the γ-H2AX peak 32 , determining also whether the replication fork will reach the TSS in a codirectional or head-to-head orientation with the RNA Pol II. We found that γ-H2AX peaks are close to their nearest origin, on average only 29 kb away (Figure 4E).…”

“…TRAIP was immunodepleted from Xenopus egg extract to less than 10% as judged by immunoblotting of depleted extract (top). The ability of TRAIP-depleted extract to synthesise DNA was established by measuring α 32 P-dATP incorporation into newly synthesised DNA and compared with replication of non-specific IgG-depleted extract. (B) No increase in H2AX phosphorylation during DNA replication is detected upon TRAIP depletion.…”

“…The bedtools intersect intervals function on the online platform Galaxy was used to identify the conserved γ-H2AX ChIP-Seq peaks in the two repeats. The distance between the γ-H2AX ChIP-Seq peak or the TSS with lowest γ-H2AX fold change +IAA/-IAA was calculated using the position of the origins of replication in HCT116 cells provided by Dr Daigaku 32 . The PRO-Seq dataset was used to determine the reciprocal direction between the oncoming replication fork and gene transcription (head-to-head or codirectional), as well as whether the TSS was the first, second or third and above transcribed gene encountered.…”