The topoisomerase 3α zinc-finger domain T1 of Arabidopsis thaliana is required for targeting the enzyme activity to Holliday junction-like DNA repair intermediates

Abstract

Topoisomerase 3α, a class I topoisomerase, consists of a TOPRIM domain, an active centre and a variable number of zinc-finger domains (ZFDs) at the C-terminus, in multicellular organisms. Whereas the functions of the TOPRIM domain and the active centre are known, the specific role of the ZFDs is still obscure. In contrast to mammals where a knockout of TOP3α leads to lethality, we found that CRISPR/Cas induced mutants in Arabidopsis are viable but show growth retardation and meiotic defects, which can be reversed by the expression of the complete protein. However, complementation with AtTOP3α missing either the TOPRIM-domain or carrying a mutation of the catalytic tyrosine of the active centre leads to embryo lethality. Surprisingly, this phenotype can be overcome by the simultaneous removal of the ZFDs from the protein. In combination with a mutation of the nuclease AtMUS81, the TOP3α knockout proved to be also embryo lethal. Here, expression of TOP3α without ZFDs, and in particular without the conserved ZFD T1, leads to only a partly complementation in root growth-in contrast to the complete protein, that restores root length to mus81-1 mutant level. Expressing the E. coli resolvase RusA in this background, which is able to process Holliday junction (HJ)-like recombination intermediates, we could rescue this root growth defect. Considering all these results, we conclude that the ZFD T1 is specifically required for targeting the topoisomerase activity to HJ like recombination intermediates to enable their processing. In the case of an inactivated enzyme, this leads to cell death due to the masking of these intermediates, hindering their resolution by MUS81.

Fig 1. Attop3α mutants exhibit a viable but growth restricted phenotype.

(A) AtTOP3α gene and domain structure. The TOP3α gene comprises 6959 bp in 24 exons (boxes). T-DNA insertions in top3A-1 and top3A-2 mutant lines are located in intron 15 and 11, respectively. For CRISPR/Cas9 mediated mutagenesis, target sequences in exon 1 were used (bolts). The TOP3α protein harbours six domains, a TOPRIM domain, the central domain with the catalytic tyrosine (star) and four C-terminal zinc-finger domains. (B) Attop3α mutant phenotypes. Two-week-old seedlings and six-week-old plants of the T-DNA insertion mutant top3A-2 and four CRISPR/Cas9 induced mutant lines top3A-3, top3A-4, top3A-5 and top3A-6 are compared to wild type (WT) plants of the same age. All mutant lines depict an identical growth restricted phenotype. (C) Root length and cell death analysis in top3α mutant lines. Root length of ten days old top3α mutant lines compared to the wild type (WT) was determined in three independent assays and mean values with standard deviation (error bars) were calculated. All top3α mutant lines show a significant reduction of root length compared to WT plants. Thereby, root length among individual mutant lines was comparably reduced. Statistical differences to the WT were calculated using a two-tailed t-test with unequal variances: *** p < 0.001. Dead cells in the root meristem were visualized with propidium iodide staining of five days old plant roots. While no cell death was visible in WT roots, top3α mutant lines exhibited a vast number of dead cells.

Fig 2. Complementation analyses in top3α mutants.

(A) Schematic structure of TOP3α protein variants used for complementation analyses. In TOP3α-ΔTOPRIM the TOPRIM domain was deleted. The protein variant TOP3α-Y342F harbours an amino acid substitution, replacing the catalytic tyrosine (red star) with a phenylalanine (black star). In TOP3α-Central both the TOPRIM domain and the C-terminus including all zinc-finger domains were deleted. In the protein variants TOP3α-ΔZnFT1/ΔZnFCCHC1/ΔZnFGRF/ΔZnFCCHC2 the individual zinc-finger domains were deleted. The variant TOP3α-N-Term consists of the N-terminus of TOP3α without C-terminal zinc-finger domains. (B) Full complementation of top3A-2 growth phenotype by expression of TOP3α. Two-week-old seedlings and six-week-old plants from three individual top3A-2::TOP3α complementation lines are compared to top3A-2 mutants and wild type (WT) plants. The characteristic growth defects of top3A-2 could be fully complemented by expression of TOP3α in all three complementation lines, leading to a growth phenotype indistinguishable to WT plants. (C) Embryo development in top3A-2 +/- ::TOP3α-Y342F/ΔTOPRIM. Depicted are representative embryos of exemplary top3A-2 +/- ::TOP3α-Y342F and top3A-2 +/- ::TOP3α-ΔTOPRIM lines compared to wild type (WT) and top3A-2 +/- embryos. All lines showed a complete embryo development leading to mature embryos. In heterozygous top3A-2 complementation lines, a Χ² test confirmed a ratio of ¼ seeds with lacking or deformed embryos, corresponding to the amount of homozygous top3A-2 mutants containing the complementation construct. (D) Growth phenotype of top3A-2::TOP3α-Central lines. Two-week-old plantlets of three individual top3A-2::TOP3α-Central complementation lines are compared to top3A-2 mutants and wild type (WT) plants. While top3A-2 mutant lines exhibit characteristic growth defects with dark, deformed leaves, expression of TOP3α-Central in this line leads to an enhanced growth defect. Plants feature only the cotyledons that are deformed and no roots are formed.

Fig 3. Growth phenotype of top3A-2::TOP3α-N-Term lines.

Two-week-old plantlets and five-week-old plants of four individual top3A-2::TOP3α-N-Term complementation lines are compared to top3A-2 mutants and wild type (WT) plants. The growth defects of top3A-2 mutant plants could be complemented completely by expression of TOP3α-N-Term leading to plants indistinguishable from the WT.

Fig 4. Root length analysis of top3A-2 complementation lines in mus81-1 mutant background.

Root length of ten-day-old top3A-2 complementation lines in mus81-1 background, wild type (WT) plants and mus81-1 mutants was determined in four independent assays and mean values with standard deviation (error bars) were calculated. In mus81-1 mutants, a significantly reduced root length compared to the WT was determined. Compared to mus81-1, both top3A-2 mus81-1::TOP3α-N-Term and top3A-2 mus81-1::TOP3α-ΔZnFT1 exhibit a significantly reduced root length while this was not the case for the further complementation lines. Statistical differences were calculated using a two-tailed t-test with unequal variances: * p < 0.05, ** p < 0.01, *** p < 0.001.

Fig 5. Rescue of growth defects by heterologous expression of RusA.

(A) Schematic illustration of the pUbi-SV40-RusA-co construct. The construct is flanked by the two border sequences (left border, LB; right border RB) for T-DNA integration. For Arabidopsis thaliana codon optimized RusA was equipped with a SV40 nuclear localization signal at the N-terminus. For expression of the RusA construct, a Ubiquitin4-2 (Ubi) promotor from Petroselinum crispum and 35S terminator (T) from cauliflower mosaic virus were used. A phosphinotricin (PPT) resistance cassette with promotor (P) and terminator (T) served for plant selection. (B) Heterologous expression of RusA in wild type (WT) plants. Three-week-old plantlets and six-week-old plants of three independent WT lines harbouring the RusA construct are compared to WT plants. The WT lines expressing the RusA construct do not display any differences in growth compared to WT plants. Scale bar = 2 cm. (C) Heterologous expression of RusA in recq4A-4 mus81-1 mutants. Three-week-old plantlets and six-week-old plants of three independent recq4A-4 mus81-1 lines containing the RusA construct are compared to the original recq4A-4 mus81-1 mutant. While recq4A-4 mus81-1 mutants depict a synthetic lethal phenotype, where plants are able to germinate but die after the formation of several deformed leaves, the expression of RusA in this line leads to viable plants. All recq4A-4 mus81-1 lines expressing RusA show a viable and fertile, but growth restricted phenotype. Scale bar = 2 cm.

Fig 6. Rescue of reduced root length in top3A-2 mus81-1 complementation lines lacking zinc-finger T1 by expression of RusA.

Root length of ten-day-old plants from each three top3A-2 mus81-1::TOP3α-N-Term/ΔZnFT1 lines independently harbouring the RusA rescue construct was determined in comparison to the original top3A-2 mus81-1 complementation lines, the mus81-1 mutant and wild type (WT) plants. Four independent assays were performed and mean values with standard deviation (error bars) were calculated. In the mus81-1 mutant, a significantly reduced root length compared to the WT could be observed. For top3A-2 mus81-1::TOP3α-N-Term/ΔZnFT1 lines a further statistically significant reduction of root length was shown. The additional expression of RusA in these lines restores root length to the same extent as in the mus81-1 mutant in all analysed lines. Statistical differences were calculated using a two-tailed t-test with unequal variances: * p < 0.05, ** p < 0.01, *** p < 0.001.