DNA Topoisomerases maintain promoters in a state competent for transcriptional activation in Saccharomyces cerevisiae

Abstract

To investigate the role of DNA topoisomerases in transcription, we have studied global gene expression in Saccharomyces cerevisiae cells deficient for topoisomerases I and II and performed single-gene analyses to support our findings. The genome-wide studies show a general transcriptional down-regulation upon lack of the enzymes, which correlates with gene activity but not gene length. Furthermore, our data reveal a distinct subclass of genes with a strong requirement for topoisomerases. These genes are characterized by high transcriptional plasticity, chromatin regulation, TATA box presence, and enrichment of a nucleosome at a critical position in the promoter region, in line with a repressible/inducible mode of regulation. Single-gene studies with a range of genes belonging to this group demonstrate that topoisomerases play an important role during activation of these genes. Subsequent in-depth analysis of the inducible PHO5 gene reveals that topoisomerases are essential for binding of the Pho4p transcription factor to the PHO5 promoter, which is required for promoter nucleosome removal during activation. In contrast, topoisomerases are dispensable for constitutive transcription initiation and elongation of PHO5, as well as the nuclear entrance of Pho4p. Finally, we provide evidence that topoisomerases are required to maintain the PHO5 promoter in a superhelical state, which is competent for proper activation. In conclusion, our results reveal a hitherto unknown function of topoisomerases during transcriptional activation of genes with a repressible/inducible mode of regulation.

Figure 1. Global reduction in mRNA levels occurs due to lack of topoisomerases I and II.

(A) Experimental setup showing the timing of G1 arrest by α-factor (α) and inhibition of Top2p at 37°C. (B) Distribution of gene expression changes (between mutant and wild-type) in topoisomerase single and double mutants. (C) Relative mRNA levels were calculated using the total microarray signal intensities in mutants and wild-type. mRNA levels in wild-type were set to 100%. Error bars represent ± one standard deviation from biological triplicates. (D) Percentage of genes up- and down-regulated 2-fold or more (open and filled columns, respectively). Error bars represent ± one standard error of the means from biological triplicates.

Figure 2. Transcriptional activity and not transcript length reflects topoisomerase dependency.

(A) top1Δ, top2ts, and top1Δtop2ts gene expression changes plotted against mRNA abundance in wild-type cells as a 200 gene moving average. (B) top1Δtop2ts gene expression changes plotted against transcriptional activity in wild-type cells as a 200 gene moving average. (C) top1Δtop2ts gene expression changes plotted against transcript length as a 200 gene moving average. Nt, nucleotides.

Figure 3. Genes de-regulated in top1Δtop2ts have high transcriptional plasticity and are chromatin-regulated.

(A) Transcriptional plasticity plotted against top1Δtop2ts gene expression changes (200-gene moving average). (B) Sensitivity to chromatin regulation plotted against top1Δtop2ts gene expression changes (200-gene moving average). (C) Left panel, the average nucleosome-binding pattern around the transcription start site (TSS) was compared between groups of the 100 most unaffected, the 100 most up-regulated, the 100 most down-regulated genes in top1Δtop2ts, and the average pattern for all genes in the yeast genome . Nucleosome-free region (NFR) is highlighted in yellow. Right panel, statistical analysis of nucleosome occupancy in the NFR displayed by a box plot. P-values were calculated by an unpaired, two-sample t-test assuming equal variances.

Figure 4. Topoisomerases are required for transcriptional induction of a range of inducible genes.

(A) Experimental setup. α indicates α-factor. (B) Time-course experiments of induced gene expression in wild-type and top1Δtop2ts cells. The mRNA levels of the indicated genes were quantified by qPCR at the indicated time points after transfer of cells to inducible conditions and normalized to the mRNA level obtained in the wild-type at the latest time point (set to 100%). Averages from two individual experiments are shown with error bars representing ± one standard deviation. Numbers indicate the mean fold increase in wild-type cells at the latest time point.

Figure 5. Topoisomerases are dispensable for transcription once the PHO5 promoter is activated, and they are not required during PHO5 inactivation.

(A) Time-course experiments of PHO5 transcription in wild-type, pho80Δ, and pho80Δtop1Δtop2ts cells after transfer from high phosphate to phosphate-free conditions. Upper panel, experimental setup. Lower panel, the PHO5 mRNA levels were quantified at the indicated time points, normalized to the wild-type level at the 0 min time point (set to 1) and presented on a log2-scale. Number indicates the mean fold increase in wild-type cells at the latest time point. (B) Time course experiment of PHO5 transcriptional inactivation in wild-type and top1Δtop2ts cells. Upper panel, experimental setup. Lower panel, the PHO5 mRNA levels were quantified at the indicated time points and normalized to the level at the 0 min time point (set to 100%). In A and B the averages from three and two individual experiments, respectively, are shown. Error bars represent ± one standard deviation. α indicates α-factor and +P_i and -P_i indicate high and no phosphate, respectively.

Figure 6. Changes in global DNA supercoiling levels affect PHO5 transcription.

(A) Time-course experiment of PHO5 transcriptional activation in wild-type, top1Δ, top2ts, and top1Δtop2ts cells. The experimental setup was as described for Figure 5A. The quantified PHO5 mRNA levels were normalized to the wild-type level at the 180 min time point (set to 100%). (B) Time course of PHO5 transcriptional activation in wild-type cells with and without expression of TopA from a high-copy YEp-TopA plasmid . The experimental setup was as described for Figure 5A except for TopA expression. The PHO5 mRNA levels were quantified at the indicated time points, normalized to the mRNA level at the 0 min time point (set to 1) and presented on a log2-scale. In A and B averages from three individual experiments are shown with error bars representing ± one standard deviation. Numbers indicate the mean fold increase in wild-type cells at the latest time point.

Figure 7. Topoisomerases are essential for binding of the Pho4p transcription factor.

(A) Nucleosome occupancy profile of the PHO5 promoter . Positioned nucleosomes (yellow), which are removed upon PHO5 induction, are denoted −1 to −4 relative to the transcription start site (TSS). Red box indicates TATA box and green boxes indicate upstream activating sequences, UAS1 and UAS2, which both contain a Pho4p binding site. Black and grey arrows indicate the primers used in the H3 and Pho4-13xcMyc ChIP experiments, respectively. (B) Time course ChIP analysis of nucleosome removal from the PHO5 promoter in wild-type and top1Δtop2ts cells following transcriptional activation. Experimental setup was as described for Figure 5A. The plot depicts average levels of histone H3 in nucleosome regions −1 to −2 and −3 to −4, respectively, in the PHO5 promoter. H3 binding levels were normalized relative to the binding under un-induced conditions at the 0 min time point (set to 1). (C) Localization of Pho4-GFP was investigated by fluorescence microscopy in wild-type and top1Δtop2ts cells grown under high phosphate conditions (+P_i) at 25°C, and 90 and 180 min after shifting cells to phosphate-free medium (−P_i) at 37°C for inhibition of Top2p. The experimental setup was as described for Figure 5A. Differential interference contrast (DIC) and fluorescence (GFP) images of representative cells are shown, and the nuclei are indicated by DAPI (4′,6-diamidino-2-phenylindole) staining of DNA. Scale bars represent 2 µm. (D) Time course ChIP analyses of Pho4-13xcMyc recruitment kinetics in the PHO5 promoter in wild-type and top1Δtop2ts cells following transcriptional activation. The experimental setup was as described for Figure 5A. The plots depict levels of Pho4-13xcMyc binding to UAS1 (left panel) and UAS2 (right panel) in the PHO5 promoter, although the resolution of the ChIP assay may be insufficient to discriminate between the two sites, and Pho4p binding at the low affinity UAS1 site may be below the detection threshold of the assay. Pho4-13xcMyc binding levels were normalized relative to the binding under un-induced conditions at the 0 min time point (set to 1). In B and D averages from three individual experiments are shown and error bars represent ± one standard deviation.