RAD51: Beyond the break

Abstract

As the primary catalyst of homologous recombination (HR) in vertebrates, RAD51 has been extensively studied in the context of repair of double-stranded DNA breaks (DSBs). With recent advances in the understanding of RAD51 function extending beyond DSBs, the importance of RAD51 throughout DNA metabolism has become increasingly clear. Here we review the suggested roles of RAD51 beyond HR, specifically focusing on their interplay with DNA replication and the maintenance of genomic stability, in which RAD51 function emerges as a double-edged sword.

Keywords: Double-stranded DNA breaks; Fork protection; Homologous recombination; RAD51; Replicative stress.

Fig. 1

Roles of RAD51 at stressed replication forks. Simplified depiction of the mechanisms by which RAD51 protects the stalled replication fork (RF). RAD51 recruitment to replication forks mediates fork reversal, as well as fork protection from breakage, either from mechanical force or nucleolytic activity (A). This initial process is independent of long RAD51 filament formation [51,80], although the detailed mechanisms remain elusive. Subsequently, RAD51 forms a stable nucleoprotein filament on the regressed arm of the ‘chicken foot’ structure to protect the newly formed DNA end from nucleolytic attack, for example by MRE11. This allows for the stabilisation of the stalled fork until fork convergence (B) or removal of the replication-obstructing lesion ahead of the fork (C). The regressed fork also offers the opportunity for 3′ end extension using the opposite strand as template, hence allowing for lesion bypass (D). Fork reversal prevents excessive repriming of DNA synthesis (E), which leads to the accumulation of ssDNA behind the fork. Remaining ssDNA can be repaired by RAD51-mediated template-switching, followed by gap-filling (F). In the absence of RAD51 stabilisation at the regressed arm the reversed fork may be targeted by MUS81 or other structure-specific nucleases, leading to fork breakage followed by break-induced replication (BIR). Strand-invasion past the replication-obstructing lesion could be enabled by 3′ extension at the reversed fork, as depicted, or alternatively rely on (partial) homology ahead of the fork (G).

Fig. 2

Mechanisms promoting RAD51 recruitment. Simplified depiction of mechanisms that promote RAD51 recruitment to sites of DSBs (A) and stalled replication forks (B–E). RAD51 recruitment to sites of DSBs is largely mediated by the BRCA1-PALB2-BRCA2 complex, which is initiated by BRCA1 recruitment to DNA damage markers, such as S130-phosphorylated and/or ubiquitinated H2AX (A). RAD51 recruitment at stalled replication forks may be mediated by the MMS22L-TONSL complex (B) or the BARD1-BRCA1 complex (C), both of which bind the histone H4 non-methylated at K20 (H4K20me0) that is present in newly synthesised DNA. RAD51 directly binds MMS22 L and a Pin1-generated BRCA1 isomer, providing a constitutive RAD51 loading mechanism in S-phase when H4K20 remains unmethylated. Upon replicative stress the ATAD5-RFC complex is recruited to replication forks, where it removes PCNA and facilitates RAD51 loading to replication forks (D). The CST complex, an RPA-like complex composed of CTC1, STN1 and TEN1, is enriched at telomeres and GC-rich repeats and recruits DNA polymerase Pol α-primase. Replicative stress triggers its interaction with RAD51, promoting the loading of RAD51 (E). RAD51 S14 is targeted by PLK1 transiently after DNA damage or during late G2-M. This is followed by T13 phosphorylation by CK2, which creates a binding motif for the NBS1 FHA domain of the MRE11-RAD50-NBS1 complex, which travels together with the replication machinery. The direct interaction of T13-phosphorylated RAD51 with NBS1 hence mediates RAD51 recruitment to stalled forks (F).

Fig. 3

Overview of difficult-to-replicate regions. Simplified depiction of chromosomal regions that are ‘difficult-to-replicate’ and vulnerable to replicative stress. Proposed sources of replicative stress are also indicated. Common fragile sites (CFSs) often associate with long genes where stable association of the transcriptional machinery and RNA-DNA hybrids (R-loops) are found. Sparse replication origins within these genes also contribute to their vulnerability. Telomeres are composed of a GC-rich repetitive sequence, and G4 DNA structures and heterochromatic sub-telomeric regions exacerbate the challenge in replicating these regions. In the absence of a distal origin, fork convergence cannot overcome replication stalling. Centromeres and pericentromeric regions, spanning millions of base pairs in mammalian genomes, also comprise repetitive sequences. Non-B form DNA structures, R-loops and active transcription are found at centromeres, while pericentromeres are largely heterochromatic. The presence of origins within these regions remains unclear.