DNA-binding site II is required for RAD51 recombinogenic activity in Arabidopsis thaliana

Abstract

Homologous recombination is a major pathway for the repair of DNA double strand breaks, essential both to maintain genomic integrity and to generate genetic diversity. Mechanistically, homologous recombination involves the use of a homologous DNA molecule as a template to repair the break. In eukaryotes, the search for and invasion of the homologous DNA molecule is carried out by two recombinases, RAD51 in somatic cells and RAD51 and DMC1 in meiotic cells. During recombination, the recombinases bind overhanging single-stranded DNA ends to form a nucleoprotein filament, which is the active species in promoting DNA invasion and strand exchange. RAD51 and DMC1 carry two major DNA-binding sites-essential for nucleofilament formation and DNA strand exchange, respectively. Here, we show that the function of RAD51 DNA-binding site II is conserved in the plant, Arabidopsis. Mutation of three key amino acids in site II does not affect RAD51 nucleofilament formation but inhibits its recombinogenic activity, analogous to results from studies of the yeast and human proteins. We further confirm that recombinogenic function of RAD51 DNA-binding site II is not required for meiotic double-strand break repair when DMC1 is present. The Arabidopsis AtRAD51-II3A separation of function mutant shows a dominant negative phenotype, pointing to distinct biochemical properties of eukaryotic RAD51 proteins.

Figure 1.. AtRAD51 schematic structure and sequence.

(A) Schematic representation of the domain structure of AtRAD51. The three essential amino acids mutated into alanine in the AtRAD51-II3A are displayed in red (R133, R306, and K316). R133 is located in the Walker A domain, whereas R306A and K316A are located in the DNA-binding site II. (B) Full alignment of Saccharomyces cerevisiae, Homo sapiens, and Arabidopsis thaliana RAD51 amino acid sequence. Mutated amino acids in the AtRAD51-II3A are written in blue. Walker (A and B) domains are outlined in magenta and DNA-binding sites (I and II) are outlined in blue. Conserved amino acids are highlighted in red, semi-conserved amino acids written in red, and non-conserved aa written in black.

Figure 2.. 3D Structure model of AtRAD51.

(A) Ribbon representation of the AlphaFold structure prediction of Arabidopsis RAD51 (AF-P94102-F1 model, top) and the corresponding Predicted Aligned Error output (bottom). The three essential amino acids in site II are labelled in magenta, and their position and name is written next to them (in magenta). In the Predicted Aligned Error plot, dark green tile corresponds to good prediction (low expected position error in Ångströms), whereas light green indicates low prediction (high error). (B, C) Superimposition of the AlphaFold proposed model structure of AtRAD51 and the crystalized structure of (B) S. cerevisiae (Turquoise; PDB ID:1szp) or (C) Human (Cyan; PDB ID:5h1c) RAD51. A close-up view of the region comprising the site II essential amino acids is displayed. (B, C) Amino acids are shown in magenta for AtRAD51, turquoise (black font) for ScRad51 (B), and cyan (black font) for human RAD51 (C). For better visualization of the site II essential amino acids, model structures in the close-up views have been slightly rotated and some residues hidden, compare to the views in the square.

Figure 3.. AtRAD51-II3A focus formation in somatic cells.

(A, B, C) Immunolocalization of RAD51 in root tip nuclei of untreated seedlings (A), or seedlings 2 h (B) or 8 h (C) after treatment with 30 μM mitomycin C (MMC). Experiments were conducted on two independent transgenic rad51 lines carrying RAD51-II3A (named T1-1 and T1-2). Scale bars: 5 μm. (D) Number of RAD51 foci per nucleus in transgenic rad51_RAD51-II3A lines compared with WT before or 2 and 8 h after treatment with 30 μM MMC. Data are presented as mean ± SD. n indicates number of cells analyzed. Kruskal–Wallis test; **P-value < 0.01, ****P-value < 0.0001. (E) Percentage of cells with 0, 1–2, 3–10, 11–20, and >20 RAD51 foci is shown for each genotype without or after MMC treatment. (D) Number of cells analyzed is the same as in (D). Source data are available for this figure.

Figure S1.. Mean and sum intensity per RAD51 focus.

(A) Mean intensity per RAD51 focus before and 2 or 8 h after treatment with 30 μM mitomycin C. (B) Sum intensity per RAD51 focus before and 2 or 8 h after treatment with 30 μM mitomycin C. Kruskal–Wallis test; **P-value < 0.01, ***P-value < 0.001, ****P-value < 0.0001. Source data are available for this figure.

Figure 4.. AtRAD51-II3A is defective in double-strand break repair and HR in somatic cells.

(A) Pictures of 2-wk-old seedlings grown without (left) or with (middle and right pictures) 30 μM mitomycin C. (B) Fraction of sensitive plants was estimated based on the number of true leaves per seedling. Seedlings with three or less true leaves were considered sensitive to DNA damage. Bars are mean ± SEM of at least three independent experiments (except for rad51_RAD51g for which two replicates were performed) with 20–24 seedlings per genotype per experiment. Error bar for T1-1 is not visible because all replicates exhibited 100% of sensitive plants. Two-way ANOVA test; ****P-value < 0.0001, n.s. not significant. (C) Quantification of spontaneous somatic HR events using the IU.GUS reporter system. HR events were quantified as the number of blue spots per seedling. 52–62 seedlings were analyzed per genotype, and two to three biological replicates were performed per genotype. Bars indicate mean ± SEM. Kruskal–Wallis test; ****P-value < 0.0001. Source data are available for this figure.

Figure S2.. Dominant negative effect of AtRAD51-II3A in WT background.

(A) Pictures of 2-wk-old seedlings grown with 30 μM mitomycin C. (B) Fraction of sensitive plants was estimated based on the number of true leaves per seedling. Seedlings with three or less true leaves were considered sensitive to DNA damage. Data from Fig 4 has been used for controls (WT and RAD51g). 22 seedlings were analyzed for each RAD51+/+_AtRAD51-II3A line. Four independent RAD51+/+_AtRAD51-II3A lines were tested and are shown. Bars are mean ± SEM of 1 experiment with 20–24 seedlings per genotype. Error bar for RAD51+/+_AtRAD51-II3A-1 is not visible because all seedlings were not sensitive to DNA damage. Two-way ANOVA test; *P-value < 0.05, ****P-value < 0.0001, n.s., not significant. Source data are available for this figure.

Figure 5.. AtRAD51-II3A restores fertility and meiotic progression of rad51 mutant.

(A) Comparison of fertility based on the number of seeds per silique in WT, rad51, rad51 complemented with RAD51 genomic sequence (RAD51g) and rad51_AtRAD51-II3A lines. Each dot represents the mean number of seeds per silique for one plant, obtained by counting at least 12 siliques. Five to six plants were analyzed per genotype. No significant difference was measured between WT, rad51_RAD51g, rad51_AtRAD51-II3A T1-1, and T1-2. Kruskal–Wallis test. (B) RAD51/ASY1 co-immunolocalization on late prophase I–staged meiocytes. Scale bars: 5 μm. (C) DAPI-stained chromosome spread of male meiocytes at late prophase I, metaphase I, anaphase I, and tetrad stage. Scale bars for each stage: 10 μm. Source data are available for this figure.

Figure 6.. AtRAD51-II3A does not repair meiotic double-strand break.

DAPI-stained chromosome spread of male meiocytes at late prophase I, metaphase I, anaphase I, and tetrad stage. Plants expressing AtRAD51-II3A show massive chromosome fragmentation. Percentage of cells showing chromosome fragmentation and number of cells analyzed is indicated on the right. Scale bars for each stage: 10 μm.