Helicase Q promotes homology-driven DNA double-strand break repair and prevents tandem duplications

Abstract

DNA double-strand breaks are a major threat to cellular survival and genetic integrity. In addition to high fidelity repair, three intrinsically mutagenic DNA break repair routes have been described, i.e. single-strand annealing (SSA), polymerase theta-mediated end-joining (TMEJ) and residual ill-defined microhomology-mediated end-joining (MMEJ) activity. Here, we identify C. elegans Helicase Q (HELQ-1) as being essential for MMEJ as well as for SSA. We also find HELQ-1 to be crucial for the synthesis-dependent strand annealing (SDSA) mode of homologous recombination (HR). Loss of HELQ-1 leads to increased genome instability: patchwork insertions arise at deletion junctions due to abortive rounds of polymerase theta activity, and tandem duplications spontaneously accumulate in genomes of helq-1 mutant animals as a result of TMEJ of abrogated HR intermediates. Our work thus implicates HELQ activity for all DSB repair modes guided by complementary base pairs and provides mechanistic insight into mutational signatures common in HR-defective cancers.

Fig. 1. Polymerase theta-independent repair of G-quadruplex-induced DSBs is driven by flanking sequence homology.

a Model explaining DSB formation and repair at replication-blocking G-quadruplex structures. b Size distribution of deletions that accumulate in the genomes of dog-1 and dog-1 polq-1 mutant animals (dog-1 data from). c Potential mechanism explaining the formation of the deletions marked in b. d Schematic diagram of four endogenous G4 loci with different degrees of flanking homology. The most prominent repeated sequences are indicated in blue. e Representative images of PCR-based analysis of the indicated G4-containing loci. Each lane contains the genomic DNA of three adult animals. Asterisks indicate stochastic deletions, which manifest as shorter than wild-type products and Δ indicates the size range of the PCR-amplified deletion products. Deletion sequences are provided as a Source Data file. f Graphic illustration of G4 deletions profiles at four endogenous G4 loci. For each locus typical G4 deletions in dog-1 and dog-1 polq-1 animals are depicted. Black bars represent homology-independent deletions; red bars represent homology-dependent events. g Histogram depicting relative deletion frequencies at the indicated G4 loci as determined by the presence of deletion bands in the PCR-based assay on single worms of dog-1 (grey bars) and dog-1 polq-1 mutant animals (black bars). Depicted frequencies are relative to the deletion frequency in dog-1 single mutants to allow the comparison of loci expressing different stochastic G4 deletion rates. The number of analysed animals is depicted on top of the bars. Gel counts are provided as a Source Data file.

Fig. 2. HELQ is essential for eMMEJ of G4-induced DSBs and a facilitator of TMEJ.

a Spectra of deletions occurring at the Qua915 locus, which is schematically depicted above, for the indicated genotype. Single deletion events are piled and sorted from top to bottom by deletion end-point relative to the G4 motif set at 0. Deletion events are colour-coded according to the following mutational classification: grey for simple deletions without apparent MH at the junction, blue for simple deletions with MH at the junction, wherein the saturation level of the blue colour increases with an increasing amount of homology identified. Deletions containing insertions are in red: bright red for insertions that can be reliably mapped to the flank of the deletion, dark red for insertions of which the origin could not be determined with certainty. Deletion sequences are provided as a Source Data file. b Spectra of deletions occurring at the Qua317 locus, which is schematically depicted above, for the indicated genotype. Colour coding is identical to (a). Deletion sequences are provided as a Source Data file. c Representative images of PCR-based analysis of the indicated G4-containing loci for the indicated genotypes. Each well contains the DNA of 10 animals. 2-Log DNA ladders (indicated by ‘M’) are used as markers for size reference. Uncropped gel pictures are provided as a Source Data file. d Proportion of deletion types at Qua915. The difference in the ratio between deletions without and with insertion were tested using the Chi-square test (**P < 0.01). e Proportion of deletion types at Qua317. The difference in ratio between deletions without and with insertion were tested using the Chi-square test (*P < 0.05). f Examples of deletions with a complex configuration of templated insertions. For each case, the insertion, as well as the left and right flank of the corresponding deletion, is depicted. Inserted nucleotide stretches that are identical to the flank of the deletion junction are underscored in both the insertion and the cognate flank. Nucleotide stretches present multiple times within one insertion are depicted in red.

Fig. 3. SSA deficiency in C. elegans lacking HELQ-1.

a Schematic representation of the single-strand annealing (SSA) reporter. A DSB can be introduced at the I-SceI recognition site by the I-SceI endonuclease. DSB repair by annealing of ~250 bp region of sequence identity placed up and downstream of the I-SceI recognition site leads to a functioning LacZ open reading frame. Small arrows depict primers used in the PCR analysis. b Representative pictures of lig-4 mutant animals carrying the reporter transgenes that were heat-shocked or mock-treated to induce I-SceI expression, followed by staining for B-galactosidase expression. c Histogram depicting the percentage of LacZ positive worms for the indicated genotype. Experiments were performed in triplicate (***P < 0.01; Chi-square test). Each dot represents the average percentage of each replicate. Error bars represent SEM. Staining quantifications are provided as a Source Data file. d Representative images of PCR-based analysis of the reporter locus at the I-SceI site for the indicated genotypes. Each well contains the DNA of one animal. Wild-type bands are expected because many cells in the animal are insensitive to heat shock-driven I-SceI expression. 2-Log DNA ladders are used as markers for size reference. e Histogram depicting the percentage of PCR samples containing a deletion. Each dot represents the percentage of deletions found in 96 samples. Experiments were performed in triplicate (***P < 0.01; Chi-square test). Error bars represent SEM. Deletion counts are provided as a Source Data file. f Spectra of deletions occurring at the I-SceI site of the SSA reporter for the indicated genotype. Single deletion events are piled and sorted from top to bottom by deletion end-point relative to the I-SceI cut site set at 0. Deletion events are colour-coded according to the following mutational classification: grey for simple deletions without apparent MH at the junction, blue for simple deletions with MH at the junction, wherein the saturation level of the blue colour increases with an increasing amount of homology identified. Deletions containing insertions are in red: bright red for insertions that can be reliably mapped to the flank of the deletion, dark red for insertions of which the origin could not be determined with certainty. Deletion sequences are provided as a Source Data file.

Fig. 4. HELQ-1 acts in synthesis-dependent strand annealing and suppresses tandem duplications.

a Models for SSA and SDSA repair of DSBs. Homologous stretches of DNA are depicted in dark blue. b Schematic representation of the SDSA reporter. A DSB can be introduced by the I-SceI endonuclease at its recognition site positioned within a corrupted GFP expression cassette. SDSA repair using a GFP segment cloned downstream of the expression cassette will restore the GFP ORF leading to GFP expressing animals. c Histograms depicting the average number of GFP expressing cells per worm for the indicated genotype. Experiments are performed in triplicate (*P < 0.05, **P < 0.01; two-sided t-tests). P value WT vs. brc-1: 0.0085. P value WT vs. helq-1: 0.0115. Each dot represents the average number of each replicate. Error bars represent SEM. Quantifications are provided as a Source Data file. d Schematic representation of the CRISPR/Cas9-induced gene correction assay. CRISPR-induced DSB at the dpy-10 locus can be repaired via end-joining or using an injected single-stranded oligo as a template. The sequence of the oligo differs from the target locus at the centre (in green). Mutation sequences are provided as a Source Data file. e Histogram depicting the different categories of repair outcomes. Insertions are depicted in yellow, simple deletions in grey, deletions containing insertion in red, outcomes of repair using the oligo are depicted in blue. The difference in ratio between ssODN-guided repair and indels were tested using Chi-square test (***P < 0.001). f Size distribution of structural variations that accumulate in the genomes of the indicated genotype (wild-type data derived from ref. ). Mutation sequences are provided as a Source Data file. g Mechanistic model for the aetiology of tandem duplications (TDs) in helq-1 mutant genetic backgrounds.