The ERCC1/XPF endonuclease is required for efficient single-strand annealing and gene conversion in mammalian cells

Abstract

The mammalian ERCC1-XPF endonuclease has a suggested role in the repair of DNA double-strand breaks (DSB) by single-strand annealing (SSA). Here, we investigated the role of ERCC1 in homologous recombination in mammalian cells, and confirm a role of ERCC1 in SSA. Interestingly, we also report an unexpected role for ERCC1 in gene conversion. This provides support that gene conversion in mammalian somatic cells is carried out through synthesis-dependent strand annealing, rather than through a double Holliday Junction mechanism. Moreover, we find low frequencies of SSA and gene conversion in G1-arrested cells, suggesting that SSA is not a frequent DSB repair pathway in G1-arrested mammalian cells, even in the presence of perfect repeats. Furthermore, we find that SSA is not influenced by inhibition of CDK2 (using Roscovitine), ATM (using Caffeine and KU55933), Chk1 (using CEP-3891) or DNA-PK (using NU7026).

Figure 1.

Homologous recombination is induced >100-fold by a single DSB in ERCC1-defective UV4DR7 cells. (A) Structure of the DRneo recombination substrate containing two non-functional copies of the neo^R gene. A functional neo^R gene can be produced by SSA, SCE or gene conversion upon induction of a DSB following expression of the I-SceI restriction endonuclease. Recombination to a functional neo^R gene using SSA or SCE results in loss of the hyg^R gene, while the hyg^R gene is preserved following recombination using gene conversion. (B) Southern blot analysis performed on UV4DR7 to confirm the intact DRneo structure in the cell line. The S2neo was used as probe. (C) Recombination frequency in UV4DR7 cells following induction of a DSB using transient transfection with the pCMV3nlsI-SceI vector. The average and standard deviation of at least three experiments is depicted.

Figure 2.

ERCC1 is involved in SSA and gene conversion. (A) Western blot analysis showing the expression of ERCC1 protein in individual blasticidin resistant UV4DR7 clones transfected with pEF6-V5-His-ERCC1 vector using a monoclonal ERCC1 antibody and anti-β-actin antibody as a loading control. (B) Surviving fraction of wild-type AA8, UV4DR7, PEF7 and ERCC1-complemented UV4DR7 clones ERCC1.17 and ERCC1.21 following exposure to UV light. The average (symbol) and standard deviation (bars) of at least three experiments is depicted. (C) Recombination frequency to G418 or G418 + Hygromycin (Hyg) resistance following transient transfection of the pCMVI-SceI3xnls vector. The level of resistance to G418 alone reflects a recombination event using either gene conversion or SSA (blue), while frequencies of gene conversion alone are indicated by G418 + Hyg resistance (yellow). The average and standard deviation of at least three experiments is depicted. (D) Southern blot analysis using S2neo as probe on XhoI and HindIII digested DNA isolated from individual recombinant G418^R clones. The subsequently found resistance to Hygromycin is indicated.

Figure 3.

SSA is suppressed in the G1 phase of the cell cycle. (A) Cell cycle profiles of ERCC1.17 and UV4DR7 cells arrested in the G1 phase by SFM or CI. After 48 h arrest, cells were transfected with either pCMV3xnlsI-SceI or pEGFP-C2 vector for 5 h and then maintained arrested for an additional 19 h before the recombination or GFP expression assays were performed. (B) Homologous recombination induced by an I-SceI-induced DSB in the DRneo construct in asynchronous and G1-arrested cells. The average and standard deviation of at least three experiments is depicted. All values have been corrected for reduced GFP expression in arrested cells. Blue bars indicate gene conversion + SSA (resistance to G418 alone) and yellow bars gene conversion (resistance to both G418 + Hyg).

Figure 4.

SSA of a single DSB is independent of inhibition of ATM, DNA-PK, CDK and Chk1 activity. Homologous recombination induced by an I-SceI-induced DSB in the DRneo construct in the ERCC1.17 cell line, co-treated with 2 mM caffeine, 10 µM KU55933, 10 µM NU7026, 25 µM Roscovitine or 500 nM CEP-3891 throughout transfections and 24 h recovery period prior to the recombination assay, using selection with (A) G418 alone (gene conversion + SSA) or (B) G418+Hyg (gene conversion only). The average and standard deviation of at least three experiments is depicted.

Figure 5.

The role of ERCC1/XPF in gene conversion supports the synthesis-dependent strand annealing model for repair of 2-ended DSBs. A resected 2-ended DSB may be repaired through SSA between two repeat sequences and the non-homologous 3′ DNA ends trimmed away by the ERCC1/XPF endonuclease. A 2-ended DSB may also be repaired by homologous recombination using the intact sister chromatid as donor. There are two major models for repair of 2-ended DSBs by homologous recombination that result in gene conversion: the double-Holliday Junction model (41) and SDSA model (42). Here, we show that ERCC1/XPF is involved in the repair of 2-ended DSBs using gene conversion. A non-homologous 3′ DNA end substrate for ERCC1/XPF is only produced in the SDSA model. Thus, our data suggest that SDSA is the preferred repair pathway for 2-ended DSBs. A third alternative D-loop nicking model for 2-ended DSB repair has recently been proposed for meiotic recombination (45), and is illustrated here to envision how ERCC1/XPF may potentially be involved in such pathway. However, there is so far no evidence for this pathway for mitotic DSB repair. Only non-cross-over resolutions of Holliday Junctions are illustrated, as these are strongly favoured in mitotic mammalian cells (22,48).