PrimPol-dependent single-stranded gap formation mediates homologous recombination at bulky DNA adducts

Ann Liza Piberger,Eva Petermann,Richard D W Kelly,Joanna R Morris,Aidan J Doherty,Akhil Bowry,Helen E Bryant,Juan Méndez,Daniel González-Acosta,Laura J Bailey,Alexandra K Walker

doi:10.1038/s41467-020-19570-7

Abstract

Stalled replication forks can be restarted and repaired by RAD51-mediated homologous recombination (HR), but HR can also perform post-replicative repair after bypass of the obstacle. Bulky DNA adducts are important replication-blocking lesions, but it is unknown whether they activate HR at stalled forks or behind ongoing forks. Using mainly BPDE-DNA adducts as model lesions, we show that HR induced by bulky adducts in mammalian cells predominantly occurs at post-replicative gaps formed by the DNA/RNA primase PrimPol. RAD51 recruitment under these conditions does not result from fork stalling, but rather occurs at gaps formed by PrimPol re-priming and resection by MRE11 and EXO1. In contrast, RAD51 loading at double-strand breaks does not require PrimPol. At bulky adducts, PrimPol promotes sister chromatid exchange and genetic recombination. Our data support that HR at bulky adducts in mammalian cells involves post-replicative gap repair and define a role for PrimPol in HR-mediated DNA damage tolerance.

Highlights

Stalled replication forks can be restarted and repaired by RAD51-mediated homologous recombination (HR), but HR can perform post-replicative repair after bypass of the obstacle
Damaged DNA can be bypassed either through error-prone translesion synthesis (TLS) that is promoted by proliferating cell nuclear antigen (PCNA) mono-ubiquitination, or through an error-free damage tolerance pathway that is promoted by PCNA polyubiquitination and uses recombination proteins
Using low concentrations of BPDE that are non-toxic and recombinogenic, we report that RAD51 foci formation and HR activation in response to bulky DNA adducts can occur independently of replication fork stalling or -collapse

Summary

Introduction

Stalled replication forks can be restarted and repaired by RAD51-mediated homologous recombination (HR), but HR can perform post-replicative repair after bypass of the obstacle. Damaged DNA can be bypassed either through error-prone translesion synthesis (TLS) that is promoted by proliferating cell nuclear antigen (PCNA) mono-ubiquitination, or through an error-free damage tolerance pathway that is promoted by PCNA polyubiquitination and uses recombination proteins This last pathway requires RAD51 and involves template switching to the undamaged sister chromatid. There is evidence that re-priming and post-replicative gap filling is preferred over fork reversal, and this pathway has been shown to involve a double Holliday junction-like intermediate[16,17] It has been proposed since the 1970s that re-priming at DNA lesions can lead to recombination in mammalian cells, fork reversal has been reported to be frequent in mammalian backgrounds[4], and may be preferred over re-priming. A longstanding and important mechanistic question is: where is RAD51 most important for replication bypass of bulky lesions, at the fork, or behind the fork? Does the observed HR activity at bulky lesions occur at stalled forks, collapsed forks or post-replicative gaps, and what are the molecular mechanisms involved?

Methods

Results

Conclusion