Mitotic homologous recombination maintains genomic stability and suppresses tumorigenesis

Abstract

Mitotic homologous recombination promotes genome stability through the precise repair of DNA double-strand breaks and other lesions that are encountered during normal cellular metabolism and from exogenous insults. As a result, homologous recombination repair is essential during proliferative stages in development and during somatic cell renewal in adults to protect against cell death and mutagenic outcomes from DNA damage. Mutations in mammalian genes encoding homologous recombination proteins, including BRCA1, BRCA2 and PALB2, are associated with developmental abnormalities and tumorigenesis. Recent advances have provided a clearer understanding of the connections between these proteins and of the key steps of homologous recombination and DNA strand exchange.

Figure 1. Pathways of DNA DSB repair

Double-strand breaks (DSBs) are efficiently repaired in mammalian cells by homologous recombination (HR) and non-homologous end joining (NHEJ). HR initiates with end resection, which produces a 3′ single-stranded end that can invade a homologous template to initiate repair. Alternative HR pathways can ensue from the displacement loop (D-loop) intermediate: synthesis-dependent strand annealing (SDSA) and DSB repair (DSBR). In SDSA, the newly synthesized strand is displaced to anneal to the other DNA end, resulting in a non-crossover outcome with no change to the template DNA. In DSBR, the second DNA end is ‘captured’ by the D-loop to form a double Holliday junction, which in principle can result in a non-crossover (cleavage at black or grey arrowheads) or a crossover (cleavage at black arrowheads on one side and grey arrowheads) outcome. NHEJ involves the joining of non-homologous DNA ends. It can be imprecise and lead to deletions and other mutations through numerous end-processing steps (not shown). Single-strand annealing takes place when end resection occurs at sequence repeats (arrowheads) to provide complementary single strands that anneal, giving rise to a product with a single copy of the repeat and a deletion of intervening sequences.

Figure 2. Mechanism of Hr revealed by recA–ssDNA and recA–dsDNA structures

In these structures, five RecA molecules (grey) bind to ATP (yellow) and single-stranded DNA (ssDNA; red) to form a nucleoprotein filament in which ssDNA adopts a helical conformation and is stretched relative to B-form DNA. Triplets within the ssDNA form a repeating unit with a B-form DNA-like structure, in which the bases can pair through canonical Watson–Crick hydrogen bonds. Engagement of the double-stranded DNA (magenta and blue) forms a synaptic complex that allows homology sampling, presumably by destabilizing the duplex through disruption of base stacking and pairing. Fidelity is achieved by base pairing of the invading strand (red) with the complementary strand (magenta), as RecA forms few contacts with the complementary strand. In addition, the RecA-imposed B-form DNA-like structure of the invading strand allows only canonical Watson–Crick base pairing, which does not allow pairing to mismatched bases. The new duplex (red and magenta) and the displaced strand (blue; not present in the crystal structures) are released following ATP hydrolysis. Image courtesy of Nikola Pavletich, Memorial Sloan–Kettering Cancer Center, USA.

Figure 3. Homologous templates and repair outcomes of Hr

A homologous sequence that can act as a template to repair a double-strand break (DSB) can be found on the sister chromatid, the homologous chromosome and, in the case of repeated sequences, a sequence on the same (not shown) or a different chromosome. a | Inter-sister repair is genetically silent regardless of whether the outcome is a crossover or a non-crossover. b | Inter-homologue repair can lead to local regions of loss of heterozygosity (LOH) when the outcome is a non-crossover or to LOH of entire distal regions of chromosomes when the outcome is a crossover and recombinant sister chromatids segregate in anaphase to different daughter cells. If recombinant sister chromatids end up in the same daughter cell, the chromosomes have undergone an exchange, but there is no loss of parental information (not shown). c | Inter-heterologue repair involving a crossover would in principle lead to reciprocal translocations, but they have rarely been observed. Instead, oncogenic translocations typically involve non-homologous end joining.

Figure 4. Hr protein interactions and domains

The homologous recombination proteins breast and ovarian cancer type 1 susceptibility protein (BRCA1), partner and localizer of BRCA2 (PALB2) and BRCA2 form complexes with RAD51 and are tumour suppressors. BRCA1-interacting protein 1 (BRIP1) also promotes homologous recombination and may be a tumour suppressor. Abraxas and CtBP-interacting protein (CtIP) have key roles in the repair functions of BRCA1, although they have not been identified as tumour suppressors. These proteins are involved in various steps of DNA repair, which includes damage recognition, end processing, repair protein localization (at damage-induced nuclear foci) and DNA strand exchange. BRIP1, abraxas and CtIP interact with BRCA1 through its BRCT domain. Other relevant interactions and functional domains are indicated. Arrows show protein–protein interactions. Mutually exclusive binding of the BRCA1 BRCT domains to either BRIP1, abraxas or CtIP through their phosphorylated SerXXPhe (pSerXXPhe) residues are shown (dashed arrows). Structures of defined domains with interacting peptides are shown on the right: the BRCA1 BRCT repeats with a pSerXXPhe-containing peptide (1), the PALB2 carboxy-terminal β-propeller with a small amino-terminal fragment of BRCA2 (magenta; 2), and the BRCA2 BRC4 peptide with RAD51 (3). In the BRCA1 BRCT domains, Met1775 forms the base of the recognition pocket for the Phe residue in the pSerXXPhe peptide and has been found to be mutated to Arg in breast cancers; this mutation abrogates the ability of the BRCA1 BRCT to bind pSerXXPhe peptides in vitro. In the PALB2 interaction with the BRCA2 peptide, the BRCA2 residue Trp31 is highlighted because mutations of this residue that abrogate the interaction with PALB2 have been found in breast cancers. The asterisks in PALB2 highlight the interaction of the N- and C-terminal residues of the WD40 structure; deletion of the last four amino acids (Tyr1183X; in which X denotes a stop codon), which has been found in patients, disrupts the structure of the protein to destabilize it. In the BRCA2 BRC4 peptide, the aromatic ring of Phe1524 is buried within a hydrophobic pocket of RAD51, probably mimicking the self interaction of Phe86 of RAD51 with this RAD51 pocket^,. BRC4 contains two modules that interact with RAD51; the N terminus is shown in magenta and the C terminus in blue. Image in part 1 is reproduced, with permission, from REF. © (2004) Elsevier. Image in part 2 is reproduced, with permission, from EMBO Reports REF. © (2009) Macmillan Publishers Ltd. All rights reserved. Image in part 3 is reproduced from REF. . BARD1, BRCA1-associated RING domain protein 1; DSS1, deleted in split hand/split foot protein 1; OB, oligonucleotide–oligosaccharide binding.