Selective ploidy ablation, a high-throughput plasmid transfer protocol, identifies new genes affecting topoisomerase I-induced DNA damage

Abstract

We have streamlined the process of transferring plasmids into any yeast strain library by developing a novel mating-based, high-throughput method called selective ploidy ablation (SPA). SPA uses a universal plasmid donor strain that contains conditional centromeres on every chromosome. The plasmid-bearing donor is mated to a recipient, followed by removal of all donor-strain chromosomes, producing a haploid strain containing the transferred plasmid. As proof of principle, we used SPA to transfer plasmids containing wild-type and mutant alleles of DNA topoisomerase I (TOP1) into the haploid yeast gene-disruption library. Overexpression of Top1 identified only one sensitive mutation, rpa34, while overexpression of top1-T(722)A allele, a camptothecin mimetic, identified 190 sensitive gene-disruption strains along with rpa34. In addition to known camptothecin-sensitive strains, this set contained mutations in genes involved in the Rpd3 histone deacetylase complex, the kinetochore, and vesicle trafficking. We further show that mutations in several ESCRT vesicle trafficking components increase Top1 levels, which is dependent on SUMO modification. These findings demonstrate the utility of the SPA technique to introduce plasmids into the haploid gene-disruption library to discover new interacting pathways.

Figure 1.

Selective ploidy ablation (SPA) mechanism and protocol. (A) Selective ploidy ablation relies on conditionally stable chromosomes. The strong galactose-inducible promoter (arrow labeled “pGAL”) was inserted adjacent to the centromere (black circle) on each of the 16 yeast chromosomes (black horizontal ovals representing two chromosome arms). The URA3 gene from Kluyveromyces lactis (open box) serves as a counter-selectable genetic marker to ensure loss of each chromosome. (B) Mating-based plasmid transfer is achieved using a universal donor containing 16 conditionally stable chromosomes (four black bars), which is transformed with an assay plasmid (circle). The recipient strain (oval with open bars for unmodified chromosomes) mates to the universal donor to make a strain heterozygous for conditionally stable and wild-type chromosomes (mix of black and open bars). Growth on galactose destabilizes the URA3-marked chromosomes so that they are lost during mitotic growth (gray bars), while the wild-type chromosomes are maintained (open bars). Counter-selection against URA3 ensures that all conditionally stable chromosomes are lost while the plasmid is maintained in the new strain. (C) Flow chart illustrating the SPA screen method. Transformation of a plasmid into the universal donor (data not shown) can be performed days or weeks prior to the SPA method. Replica pinning steps are represented as boxes with dashed lines. CuSO₄ listed at step 4 is used to induce expression from the plasmid.

Figure 2.

Screening the haploid yeast gene disruption library using DNA topoisomerase I alleles. (A) Expression plasmids for DNA topoisomerase I were constructed as single-copy CEN-based plasmids (black circle) using LEU2 as a dominant selectable marker. The control plasmid (pWJ1512) contains only the copper-inducible promoter from the yeast CUP1 gene. The wild-type TOP1 gene is cloned in front of this promoter in the pWJ1439 plasmid. The mutant allele is cloned into pWJ1440 and contains a T722 to A substitution near the active site (star) that stabilizes the covalent catalytic intermediate and causes DNA damage. (B) The result of plasmid transfer by SPA into an arrayed strain library. Each strain from a set of 384 is quadruplicated onto a 1536 grid so that the four replicate strains make a 2 × 2 square. Growth of the strains after 3 d on 100 μM Cu to induce top1 expression is shown. (C) Mutant phenotypes are scored by growth differences caused by the different plasmids. The highlighted row shows that most strains exhibit the same growth regardless of the plasmid that is transferred. The asf1▵ strain is the exception and does not grow when the mutant top1 allele is transferred. (D) Quantification of colony growth is achieved using digitized plate images (see main text). The bar graph shows the values from the average strain growth of the colonies shown in C normalized to the median growth value of the plate. (E) Data collected from two separate screens of the gene-disruption library are represented as a scatter plot with each point showing the log growth ratios for an individual strain as x and y coordinates. Points drawn on the graph are colored according to density (inset color bar shows scale). The correlation statistic comparing the combined screens of the top1-T₇₂₂A mutant is shown in the inset (R = 0.81).

Figure 3.

Area-proportional Euler diagram comparing three different screens for gene disruptions sensitive to Top1-induced DNA damage. (A) The Euler diagram (left) shows irregular polygons with areas proportional to the numbers of genes identified in three different screens. The solid outline represents genes identified in the current study. The dashed line represents the genes identified in the Deng et al. (2005) CPT screen. The dotted line indicates the genes identified in the Parsons et al. (2004) screen. The proportion of genes unique to each screen is shown as the parts of those polygons in primary colors (yellow for this study; red for Deng et al. 2005; blue for Parsons et al. 2004). Where two screens overlap, the section is indicated by a secondary color. Genes found in all three screens are shown in the middle in gray. The colored boxes on the right list the genes that populate each section of the diagram. The Euler diagram was drawn using the “Draw Euler” applet by Stirling Chow (http://theory.cs.uvic.ca/euler/DrawEuler/). (B) Three complexes identified as unique to the top1-T₇₂₂A screens are listed with average Z-scores. Bold indicates Z-scores > 2 in both screens. Genes with Z-scores > 2 that are not in bold indicate a Z-score below the cutoff in one of the screens. (ND) A log ratio could not be derived for the gene due to slow growth with the control plasmid. (†) Deletions strains identified as top1-T₇₂₂A sensitive in the manual-pin screen. (*) The deletion strain was individually grown after transformation and tested to show top1-T₇₂₂A sensitivity.

Figure 4.

SUMO modification of Top1 affects growth of vps25 cells. (A) The SUMO-TOP1 plasmid (pWJ1763) was transformed into VPS25 wild-type (W7981-2A) and vps25▵ mutant strains (W5911-1A). The wild-type TOP1 (pWJ1439) and the empty vector (pWJ1512) were also transformed as controls. Cultures were grown overnight, 10-fold serially diluted, and 5-μL drops were spotted onto plates with 0, 100, or 200 μM CuSO₄. (B) Doubling times of strains were calculated after measuring growth in liquid cultures by optical density at 600 nm. All TOP1 mutant alleles are integrated in place of the corresponding wild-type allele. In the VPS25 wild-type background, the designated strains are: TOP1 (W8402-1A), top1-KR (EJY457), top1-YF (W8646-2B), top1-KR, YF (W8643-2D). In the vps25 mutant background the designated strains are: TOP1 (W8402-1C), top1-KR (W8402-1B), top1-YF (W8646-2A), top1-KR, YF (W8643-4A).

Figure 5.

Top1 accumulates in vps25 cells, and accumulation is suppressed by removing the SUMOylation sites from Top1. (A) Fluorescence (YFP) and differential interference contrast (DIC) images are shown from wild-type (W9471-8C) and vps25▵ (W8842-5C) cells transformed with expression vectors containing YFP-TOP1 or the YFP-top1-K_65,91,92R (YFP-Top1-KR) mutant after a 4-h induction with 100 μM CuSO4. (B) Fluorescence signals were quantified using Volocity as described in the Methods. Protein levels from 50 to 120 individual cells of each strain were quantified, and the mean value for each strain was plotted in arbitrary units. Error bars indicate 95% confidence intervals. The asterisk indicates a significant difference between YFP-Top1 and YFP-Top1-KR fluorescence in a vps25▵ strain (P = 0.0004).