ATM Promotes RAD51-Mediated Meiotic DSB Repair by Inter-Sister-Chromatid Recombination in Arabidopsis

Abstract

Meiotic recombination ensures accurate homologous chromosome segregation during meiosis and generates novel allelic combinations among gametes. During meiosis, DNA double strand breaks (DSBs) are generated to facilitate recombination. To maintain genome integrity, meiotic DSBs must be repaired using appropriate DNA templates. Although the DNA damage response protein kinase Ataxia-telangiectasia mutated (ATM) has been shown to be involved in meiotic recombination in Arabidopsis, its mechanistic role is still unclear. In this study, we performed cytological analysis in Arabidopsis atm mutant, we show that there are fewer γH2AX foci, but more RAD51 and DMC1 foci on atm meiotic chromosomes. Furthermore, we observed an increase in meiotic Type I crossovers (COs) in atm. Our genetic analysis shows that the meiotic phenotype of atm rad51 double mutants is similar to the rad51 single mutant. Whereas, the atm dmc1 double mutant has a more severe chromosome fragmentation phenotype compared to both single mutants, suggesting that ATM functions in concert with RAD51, but in parallel to DMC1. Lastly, we show that atm asy1 double mutants also have more severe meiotic recombination defects. These data lead us to propose a model wherein ATM promotes RAD51-mediated meiotic DSB repair by inter-sister-chromatid (IS) recombination in Arabidopsis.

Keywords: ATM; DSBs; RAD51; inter-sister chromatids; meiosis; recombination.

FIGURE 1

Phenotypic analysis of a mutant allele with fertility defects. (A) Plant fertility phenotype of wild type (Ler) and line 184 mutant. Red arrow indicates short silique of the mutant. Bar = 1 cm. (B) Morphology of open flower of wild type (Ler) and line 184 mutant. The mutant has shorter filaments and nearly no pollen grains could be observed on the stigma. (C) Pollen viability (Alexander staining) in wild type and line 184 mutant anthers. Bar = 100 μm. (D) Morphology of microspore tetrad (stained with toluidine blue). Normal meiosis produce tetrad contains four microspores whereas the mutant produce polyads contain more than four microspores. Bar = 10 μm. (E) Chromosome morphology of male meiosis in wild type and line 184. Bar = 5 μm.

FIGURE 2

Identification of the mutated gene in the mutant. (A) The fine-linkage map generated by analyzing 143 line 184 × Col-0 F2 segregating plants. The recombinant numbers are given between markers. (B) Diagram of ATM gene, the triangle showing the position of 766 bp insertion. All the colored rectangles represent an exon and each color indicates gene coding regions of corresponding protein domains. (C) RT-PCR detection of ATM transcripts. Six pairs of primers (30 cycles of PCR) were used to analyze the transcript and the location of these primer pairs was shown in (B). The red arrow indicated the abnormal transcription products of ATM detected in atm-5. Amplification of the ACTIN cDNA has been used as a control (25 cycles of PCR).

FIGURE 3

Immunolocalization of γH2AX in male meiocytes of different alleles. (A) γH2AX localization on zygotene stage chromosomes of wild type, atm-2, atm-5, atr-2, and atm-5 atr-2 respectively. (B) γH2AX localization on pachytene stage chromosomes of wild type, atm-2, atm-5, atr-2, and atm-5 atr-2 respectively. (C) Number of γH2AX foci per cell in zygotene stage in different alleles: wild type (204.07 ± 10.64, n = 30), atm-2 (146.84 ± 18.55, n = 32, P = 1.40E-11), atm-5 (152.3 ± 6.59, n = 30, P = 2.91E-11), atr-2 (208.48 ± 24.83, n = 33, P = 4.36E-1) and atm-5 atr-2 (21.1 ± 6.58, n = 29, P = 4.31E-11). (D) Number of γH2AX foci per cell in pachytene stage in different alleles: wild type (80.2 ± 10.31, n = 30), atm-2 (47.31 ± 10.02, n = 31, P = 6.27E-11), atm-5 (51.07 ± 5.33, n = 30, P = 2.93E-11), atr-2 (72.44 ± 15.66, n = 33, P = 1.86E-2) and atm-5 atr-2 (18.64 ± 5.86, n = 36, P = 3.64E-12). Bar = 5 μm. NS: P > 0.05, ***P < 0.001 (Wilcoxon Rank Sum test). Cells counted in each allele are from at least 30 plants. The atm-2 and atm-5 are two independent atm mutant alleles.

FIGURE 4

Immunolocalization of RAD51 and DMC1 in male meiocytes of wild type and atm mutants. (A) RAD51 localization on zygotene stage chromosomes of wild type, atm-2 and atm-5 respectively. (B) RAD51 localization in pachytene stage chromosomes of wild type, atm-2 and atm-5 respectively. (C) Number of RAD51 foci per cell of zygotene (WT: 168 ± 21.09, n = 27; atm-2: 279.15 ± 52.72, n = 40, P = 8.91E-12; atm-5: 285.67 ± 45.57, n = 39, P = 1.55E-11) and pachytene (WT: 32.15 ± 6.14, n = 47; atm-2: 58.84 ± 10.17, n = 43, P = 2.15E-15; atm-5: 53.6 ± 8.6, n = 32, P = 5.55E-13) stage cells in different alleles. (D) DMC1 localization on zygotene stage chromosomes of wild type, atm-2 and atm-5 respectively. (E) DMC1 localization in pachytene stage chromosomes of wild type, atm-2 (and atm-5 respectively. (F) Number of DMC1 foci per cell of zygotene (WT: 196.34 ± 49.39, n = 35; atm-2: 223.34 ± 28.02, n = 32, P = 3.38E-3; atm-5: 236.17 ± 42.17, n = 29, P = 4.97E-4) and pachytene (WT: 49.79 ± 8.61, n = 24; atm-2: 55.48 ± 8.77, n = 27, P = 3.35E-2; atm-5: 54.4 ± 9.76, n = 32, P = 6.24E-2) stage cells in different alleles. Bar = 5 μm. NS: P > 0.05 ***P < 0.001, **P < 0.01 (Wilcoxon Rank Sum test). Cells counted in each allele are from at least 30 plants. The atm-2 and atm-5 are two independent atm mutant alleles.)

FIGURE 5

Number of chiasmata in wild type and atm mutants. (A) Bivalents in metaphase I meiocytes in wild type and atm mutants. The number of chiasmata of each chromosome was marked aside. (B) Number of chiasmata per cell in wild type (10.85 ± 1.18, n = 20) and atm mutants (atm-2: 12 ± 1.02, n = 25, P = 2.72E-3; atm-5: 12.01 ± 1.37, n = 21, P = 7.76E-3, Wilcoxon Rank Sum test). For each allele, metaphase I cells used for chiasmata counting are from about 30 plants. The atm-2 and atm-5 are two independent atm mutant alleles. (C) Immunolocalization of HEI10 (green dots) on pachytene (atm-2: 11.3 ± 1.82, n = 113, P = 7.59E-13; atm-5: 11.05 ± 1.47, n = 65, P = 3.25E-09) and diakinesis (atm-2: 11.25 ± 1.49, n = 110, P = 2.18E-04; atm-5: 10.94 ± 1.41, n = 48, P = 7.17E-03) stage chromosomes (magenta) in wild type and atm mutants. Bar = 5 μm. (D) Number of HEI10 foci in wild type and atm mutants. Cells counted in each allele are from at least 30 plants. The atm-2 and atm-5 are two independent atm mutant alleles.

FIGURE 6

Chromosome morphology analysis of male meiocytes in different alleles. Wild type (WT) (A1–E1), atm-5 (A2–E2), rad51-3 (A3–E3), atm-5 rad51-3 (A4–E4), dmc1-3 (A5–E5), atm-5 dmc1-3 (A6–E6), asy1 (A7–E7), and atm-5 asy1 (A8–E8) chromosome morphologies at zygotene, pachytene, metaphase I, anaphase I, and anaphase II. Centromere signals are in red. Bar = 5 μm.

FIGURE 7

Quantification of bivalents and chromosome fragments in different alleles. (A) Quantification of bivalent and univalent in different alleles. The number of bivalents or univalents in each allele was counted in the metaphase I cells. For those metaphase I cells contain entangled chromosomes were classified as entangled. For those metaphase I cells contain both entangled chromosomes and univalents were classified as univalents and entangled. For each allele, the counted cells are from about 20 plants. (B) Quantification of chromosome fragmentation levels in different alleles. For each allele, the anaphase II and telophase II cells were observed after DAPI stained chromosome spreads and centromeric FISH. DAPI stained bodies without centromere signal were taken as chromosome fragments. Chromosome fragmentation levels are divided into four categories: 0 (no chromosome fragmentation), 1–5 (1–5 fragments per cell), 6–10 (6–10 fragments per cell) and >10 (more than 10 fragments per cell). For each allele, the counted cells are from about 12 plants.

FIGURE 8

Model of ATM promote RAD51 mediated inter-sister meiotic DSB repair. After meiotic DSBs are formed, it will be processed to generated the 3’ ssDNAs (A). With the help of RAD51 and DMC1, some of the ssDNAs will invade homologous chromosomes to seek template for repair (A,B). Those DSBs repaired using homologous chromosomes will generate COs or NCOs while NCOs are in the majority (B,C). Whereas some of the DSBs are repaired through inter-sister recombination which is mediated by RAD51 (B,C). Thus, we proposed that ATM could promote this RAD51-mediated inter-sister repair process and it is possible that ATM regulate this process through phosphorylating a potential protein X (C). Thus, when ATM is absent, this inter-sister repair pathway was impaired and leads to some unrepaired meiotic DSBs which eventually result in the mild chromosome fragmentations observed in atm mutants.