Orchestrating the nucleases involved in DNA interstrand cross-link (ICL) repair

Abstract

DNA interstrand cross-links (ICLs) pose a significant threat to genomic and cellular integrity by blocking essential cellular processes, including replication and transcription. In mammalian cells, much ICL repair occurs in association with DNA replication during S phase, following the stalling of a replication fork at the block caused by an ICL lesion. Here, we review recent work showing that the XPF-ERCC1 endonuclease and the hSNM1A exonuclease act in the same pathway, together with SLX4, to initiate ICL repair, with the MUS81-EME1 fork incision activity becoming important in the absence of the XPF-SNM1A-SLX4-dependent pathway. Another nuclease, the Fanconi anemia-associated nuclease (FAN1), has recently been implicated in the repair of ICLs, and we discuss the possible ways in which the activities of different nucleases at the ICL-stalled replication fork may be coordinated. In relation to this, we briefly speculate on the possible role of SLX4, which contains XPF and MUS81- interacting domains, in the coordination of ICL repair nucleases.

Acknowledgments

This work was supported by a Cancer Research UK Programme Grant to P.Mc, and a New Zealand Top Achiever Scholarship to A.T.W.

Figures and Tables

Figure 1 Existing models for replication coupled ICL repair in mammalian cells. (A) Model for the repair of ICLs during S phase in mammalian cells as adapted from Niedernhofer et al.17 (B) Converging fork model of ICL repair, as based on Knipscheer et al. and raschle et al.18,19

Figure 2 Nuclease activity of structure-specific endonucleases implicated in ICL repair. The substrate preference for human (A) MUS81-EME1, (B) XPF-ERCC1, (C) SLX1–SLX4 and (D) FAN1 is summarized. The region of cleavage is indicated by a red arrow in the DNA structure. The preference of cleaving one DNA substrate over another is indicated by chevron arrows, commas denote similar levels of incision efficiency. Figure adapted from Ciccia et al., Svendsen et al. and Smogorzewska et al.33–35

Figure 3 Translesion synthesis is more efficient following trimming of the tethered oligonucleotide. (A) Increasing efficiency of bypass synthesis/TLS as the cross-linked oligonucleotide is shortened. (B) the hSNM1A exonuclease is able to digest DNA past a cross-linked base, as described by our recent study in reference 48.

Figure 4 Role of structure-specific endonucleases in ICL repair. (A) the initial steps of the models for context-dependent ICL repair during replication, discussed in our recent study in reference 48, and in this review. The initial incisions could be on either the (i) leading or (ii) lagging strand or (iii) two replication forks converge on an ICL, more likely during late S phase or in the absence of the XPF-ERCC1/hSNM1A pathway. (B) Potential roles for the recently identified FAN1 nuclease in ICL repair, based on models proposed by Kratz et al., Mackay et al. and Smogorzewska et al.31–33 (i) when two replication forks converge, FAN1 could be involved in early endonucleolytic incision of the replication forks, in the removal of the cross-linked oligonuclotide from the second strand or in later steps required for HR, including ssDNA resection or 5′-flap incisions during HJ resolution. (ii) In a single stalled replication fork model, FAN1 may also function in later steps required for HR, including ssDNA resection or D-loop incision.