Topoisomerase 3alpha and RMI1 suppress somatic crossovers and are essential for resolution of meiotic recombination intermediates in Arabidopsis thaliana

Abstract

Topoisomerases are enzymes with crucial functions in DNA metabolism. They are ubiquitously present in prokaryotes and eukaryotes and modify the steady-state level of DNA supercoiling. Biochemical analyses indicate that Topoisomerase 3alpha (TOP3alpha) functions together with a RecQ DNA helicase and a third partner, RMI1/BLAP75, in the resolution step of homologous recombination in a process called Holliday Junction dissolution in eukaryotes. Apart from that, little is known about the role of TOP3alpha in higher eukaryotes, as knockout mutants show early lethality or strong developmental defects. Using a hypomorphic insertion mutant of Arabidopsis thaliana (top3alpha-2), which is viable but completely sterile, we were able to define three different functions of the protein in mitosis and meiosis. The top3alpha-2 line exhibits fragmented chromosomes during mitosis and sensitivity to camptothecin, suggesting an important role in chromosome segregation partly overlapping with that of type IB topoisomerases. Furthermore, AtTOP3alpha, together with AtRECQ4A and AtRMI1, is involved in the suppression of crossover recombination in somatic cells as well as DNA repair in both mammals and A. thaliana. Surprisingly, AtTOP3alpha is also essential for meiosis. The phenotype of chromosome fragmentation, bridges, and telophase I arrest can be suppressed by AtSPO11 and AtRAD51 mutations, indicating that the protein is required for the resolution of recombination intermediates. As Atrmi1 mutants have a similar meiotic phenotype to Attop3alpha mutants, both proteins seem to be involved in a mechanism safeguarding the entangling of homologous chromosomes during meiosis. The requirement of AtTOP3alpha and AtRMI1 in a late step of meiotic recombination strongly hints at the possibility that the dissolution of double Holliday Junctions via a hemicatenane intermediate is indeed an indispensable step of meiotic recombination.

Figure 1. Molecular analysis of the T-DNA insertion lines.

(A) The location of the T-DNA insertion in lines top3α-1 and 2 is depicted. The schematically drawn TOP3α gene (At5g63920) contains 24 exons (white boxes). The T-DNA insertions interrupted the gene either in intron 16 (top3α-1) or in intron 11 (top3α-2) (B) Scheme of genomic changes of rmi1-1 and 2 mutants. AtRMI1 (At5g63540) contained eight exons, rmi1-1 showed a 2 kb deletion including exon 1-5 and in rmi1-2 the T-DNA is inserted into the 5th exon. Primers used are indicated by numbered or labelled arrows. The total length of the original or truncated gene (in case of rmi1-1) is given on the right. The genomic sequences adjacent to each T-DNA insertion or deletion locus were determined by PCR and sequencing, and are depicted below the schematic drawings. The genomic sequences are shown in bold and the respective left or right border sequences are in italics and underlined. Abbreviations: LB = left border; RB = right border; FIL = filler; DEL = deletion.

Figure 2. Phenotypes of the top3α-1 and 2 mutant lines and the complemented plants.

(A) Upper panel: Typical example of a homozygous top3α-1 seedling on agar plates shown seven days after germination. A normally growing heterozygous seedling is shown on the left for comparison. Lower panel: Examples of wild type (left), heterozygous (middle) and homozygous (right) top3α-2 plants after four weeks grown in soil. The heterozygous plant appeared absolutely normal, whereas the homozygous plant was dwarfed and showed malformed or fasciated organs. (B) Upper panel: Wild type Col-0, heterozygous and homozygous top3α-2 mutant plants each containing the complementation construct (c+). Plants were transferred into soil after three weeks being on gentamycin selection plates. All three types of successfully transformed plants appeared normal and yielded comparable amounts of seeds. Lower panel: Exemplary siliques of the plants transformed with the complementation construct in different genetic backgrounds (wild type, heterozygous or homozygous for the respective T-DNA insertion) and from the non-transformed top3α-2 line are shown.

Figure 3. Mutagen sensitivity assays using MMS, cisplatin or camptothecin.

(A to C) Fresh weight of five plantlets from each mutagen concentration was determined and expressed in percentage. The percentages were calculated from the relation of fresh weight of each line at a given mutagen concentration to the fresh weight of the same line without mutagen. Each assay was performed at least five times as described. Mean values and standard deviations are given. All three lines (recq4A-4, top3α-2 and rmi1-2) were sensitive to MMS and cisplatin, whereas only top3α-2 showed a significant sensitivity towards camptothecin (panel C). Top3α-2 also showed a more elevated MMS sensitivity than recq4A-4 or rmi1-2 (panel A). ppm = parts per million.

Figure 4. Recombination frequencies of untreated and bleomycin treated T-DNA insertion lines.

Mean value of at least five independent assays using the IC9C background line in which restoration of the functional GUS-gene was only possible by intermolecular recombination. All three T-DNA insertion lines (recq4A-4, top3α-2 and rmi1-1) showed a five to seven-fold enhanced basic level of crossovers. The absolute numbers of blue sectors after induction of DSBs by bleomycin were more or less the same in all three lines. This indicates that the RTR complex is not involved in intermolecular DSB repair but rather in replication-dependent recombination.

Figure 5. DAPI staining of different meiotic stages during pollen development in the mutant lines.

Wild type meiosis (A1–7) is characterized by synapsis of homologous chromosomes and the formation of bivalents during the pachytene stage of prophase (A1). During diakinesis, the chromosome pairs condensed for the first meiotic division (A2). At metaphase I, homologues assembled at the metaphase plate (A3). During anaphase I, the homologous chromosomes separated to the poles (A4) and decondensed in telophase I (A5). During anaphase II, the chromatids slightly condensed and separated (A6), followed by another decondensation phase in telophase II and ended up in tetrads (A7). The mutant lines top3α-2 and rmi1-1 both exhibited chromosome fragmentation became visible in diakinesis and ending in telophase I (B2 to B5 and E2 to E5). Furthermore, both lines did not progress to meiosis II and showed an arrest in telophase I (B5 and E5). In the rmi1-1 mutants, massive amounts of stainable DNA fragments appeared stuck at the metaphase plate (compare B4 to E4). Double mutants of top3α-2 and two genes involved in early steps of meiotic DSB repair (Atspo11-2 and Atrad51) showed the respective phenotype of the background mutant in both cases, indicating that top3α-2 is involved in the latter steps of DSB repair (C1 to 7 and D1 to 7). The arrest in telophase I was released in spo11-2 or rad51 backgrounds, respectively.

Figure 6. Examples of anaphase I in the mutant lines visualized by DAPI staining of developing pollen.

Several anaphase I chromosome preparations of top3α-2 and rmi1-1 mutant lines are shown in comparison to the double mutants of top3α-1 or 2 and recq4A-4.

Figure 7. Three different pathways to remove aberrant DNA structures in which AtTOP3α is involved.

In replication repair, the RTR complex was involved in the suppression of CO recombination by removing recombinogenic DNA structures. In mitotic division, only TOP3α was necessary for proper segregation of the sister chromatids (SC). Neither RECQ4A nor RMI1 were required for this segregation. During meiotic recombination, both TOP3α and RMI1 were essential for the resolution of homologous chromosomes (HC).