The WTM genes in budding yeast amplify expression of the stress-inducible gene RNR3

Abstract

Cellular responses to DNA damage and inhibited replication are evolutionarily conserved sets of pathways that are critical to preserving genome stability. To identify new participants in these responses, we undertook a screen for regulators that, when present on a high-copy vector, alter expression of a DNA damage-inducible RNR3-lacZ reporter construct in Saccharomyces cerevisiae. From this screen we isolated a plasmid encoding two closely related paralogs, WTM1 and WTM2, that greatly increases constitutive expression of RNR3-lacZ. Moderate overexpression of both genes together, or high-level expression of WTM2 alone from a constitutive promoter, upregulates RNR3-lacZ in the absence of DNA damage. Overexpressed, tagged Wtm2p is associated with the RNR3 promoter, indicating that this effect is likely direct. Further investigation reveals that Wtm2p and Wtm1p, previously described as regulators of meiotic gene expression and transcriptional silencing, amplify transcriptional induction of RNR3 in response to replication stress and modulate expression of genes encoding other RNR subunits.

Figure 1.—

High-copy WTM1 and WTM2 increase expression of both integrated and extrachromosomal RNR3-lacZ reporters, independent of Rad53p activity. (A) Strains containing either an RNR3-lacZ or a DIN7-lacZ integrated reporter and transformed with pSR54 (a high-copy plasmid containing a genomic DNA insert encoding WTM2 and WTM1) or empty vector were grown to midlog phase in synthetic medium and then incubated for 3 hr with or without 100 mm HU. β-Galactosidase activities are the means of four independent isolates with error bars and asterisks indicating standard deviations and statistically significant differences, respectively. Effect of pSR54 on activity: P < 0.001 for RNR3-lacZ, not significant for DIN7-lacZ. (B) Isogenic rad53-21 and wild-type (wt) strains containing an integrated RNR3-lacZ reporter and either a plasmid encoding both WTM1 and WTM2 (pWTM1/2) or empty vector were assayed for β-galactosidase activity in midlog phase. Values represent means of two independent isolates; P < 0.001 for effect of pWTM1/2 on both strains. (C) Cells containing a plasmid-based GAL10-CYC1-lacZ reporter and either a plasmid expressing WTM1 and WTM2 or empty vector were grown in media containing glucose (top) or galactose (bottom) and assayed for β-galactosidase activity in midlog phase. Values represent means of three and six independent isolates for strains with empty vector and a plasmid expressing WTM1 and WTM2, respectively. Effects of pWTM1/2 and empty vector on reporter activity were not significant. (D) Cells containing a plasmid-based RNR3-CYC1-lacZ reporter and either a plasmid expressing WTM1 and WTM2 or empty vector were assayed for β-galactosidase activity in midlog phase. Values represent means of four independent isolates; P < 0.001 for pWTM1/2 vs. vector. Strains used: (A) TSR2051, TSR2048, TSG125, and TSG126; (B) TSG46, TSG47, TSG48, and TSG49; (C) TSR2425 and TSR4247; (D) TSG78 and TSG80.

Figure 2.—

High-copy WTM1 and WTM2 increase expression of the native RNR3 gene. Cells containing empty vector (lane 1) or a plasmid encoding both WTM1 and WTM2 (lane 2) and grown in selective medium were harvested in midlog phase for RNA extraction and Northern blot analysis. The same blot was probed sequentially for RNR3, WTM1, WTM2, ACT1, and scR1 (a loading control). Numbers below each band indicate the mean of three independent isolates run on the same gel relative to untreated vector control and normalized to scR1; all coefficients of variation were <30%. P-values for pWTM1/2 vs. vector: RNR3, P = 0.003; WTM1, P < 0.001; WTM2, P = 0.001; ACT1, not statistically significant. Strains used: TSG41 and TSG42.

Figure 3.—

Expression levels of both WTM1 and WTM2 influence RNR3-lacZ expression. (A) Overexpression of WTM1 and WTM2 genes individually on high-copy (2μ) vectors indicates that both genes together maximally stimulate RNR3-lacZ expression. Each strain contains a single, integrated copy of RNR3-lacZ and two high-copy plasmids, either empty vector or a plasmid encoding a single WTM gene (as indicated in the table). These strains were grown to midlog phase in selective medium and assayed for β-galactosidase activity. The means and standard deviations of five independent isolates are plotted. Effect relative to empty vector: WTM1, not statistically significant; WTM2, P = 0.04; WTM1 + WTM2, P < 0.001. (B) High-level expression of WTM2 alone increases RNR3-lacZ expression. Strains containing a single, integrated copy of RNR3-lacZ and pGPD-WTM1 (a high-copy plasmid with WTM1 under control of the constitutive GPD promoter), pGPD-WTM2 (a high-copy plasmid with WTM2 under control of the constitutive GPD promoter), or a control vector were assayed for β-galactosidase activity. The means and standard deviations of three independent isolates are plotted: P < 0.001 for effect of GPD-WTM2; GPD-WTM1 effect is not statistically significant. (C) Stimulation of RNR3-lacZ by high-level WTM2 expression does not require the endogenous WTM1 gene. Wild-type or wtmΔ12 cells containing an integrated RNR3-lacZ reporter were transformed with plasmid pGPD-WTM2 and assayed for β-galactosidase activity. The means and standard deviations of four independent isolates are plotted: P < 0.001 for pGPD-WTM2; effect of wtmΔ12 mutation on pGPD-WTM2 stimulation is not statistically significant. (D) High-level expression of WTM2 alone increases RNR3 gene expression. Strains containing high-copy plasmid pGPD-WTM2 or empty vector were grown to midlog phase and their RNAs were extracted and subjected to Northern blot analysis. The same blot was probed sequentially for RNR3, WTM1, WTM2, ACT1, and scR1 (a loading control). Numbers below each band indicate the mean of three independent isolates relative to untreated vector control and normalized to scR1. Coefficients of variation were ≤23%. P-values for pGPD-WTM2 vs. vector: RNR3, P = 0.01; WTM2, P = 0.01; ACT1 is not significantly different. Strains used: (A) TSG21, TSG19, TSG23, and TSG20; (B) TSG94, TSG95, and TSR2441; (C) TSG94, TSG95, TSG97, and TSG98; (D) TSG133 and TSG134.

Figure 4.—

Highly expressed Wtm2p associates with the upstream region of RNR3. Chromatin immunoprecipitations (ChIP's) with IgG sepharose were performed on extracts of cells with a chromosomal wtm2 deletion and plasmid pGPD-WTM2-TAP expressing epitope-tagged Wtm2p. Control extracts were from cells with untagged Wtm2p, no Wtm2p, and a truncated, mutant, epitope-tagged wtm2. (A) Quantitative PCR products from one representative ChIP experiment. Input and ChIP samples were assayed using primers for the upstream region of RNR3 and for the coding regions of SNR7 and ACT1. Products fractionated by agarose gel electrophoresis and visualized by ethidium bromide staining are shown as the inverse image. (B) Bar graph of quantitative PCR results. Dilutions of PCR products were fractionated by agarose gel electrophoresis and measured by Southern analyses. The data are represented as signals from tagged full-length, wild-type Wtm2p or mutant wtm2p samples normalized to those of untagged or no Wtm2p samples. Solid and open bars represent ChIP and input samples, respectively. The mean and standard deviations of three independent isolates, each measured in two independent chromatin preparations and experiments, are plotted for wild-type Wtm2p-TAP: P < 0.01 for RNR3, SNR7, and ACT1 enrichments, each being different from the background ratio of 1.0; P = 0.002 for RNR3 enrichment being different from those of SNR7 and ACT1; there were no statistically significant differences for input samples. Coefficients of variation were 16% for RNR3 ChIP and ≤10% for all other samples. Strains used: TSR2386, TSR2387, TSG107, and TSR2451. (C) Southern blot analysis of PCR products from amplification reactions using three different amounts of starting template. Template dilutions (from left to right: 1×, 0.25×, and 0.1×) from ChIP and total input samples from strains with full-length, wild-type Wtm2p-TAP or the untagged control were amplified by PCR and then analyzed by Southern blotting with the upstream region of RNR3 as probe. Strains used: TSR2386 and TSR2387.

Figure 5.—

RNR3-lacZ induction is attenuated in wtm2 deletion mutants. Cells lacking WTM1, WTM2, or both genes were grown to midlog phase in YPD medium, incubated with or without 100 mm HU for 3 hr, and then assayed for β-galactosidase activity. Values indicate mean and standard deviation of four independent isolates: P = 0.001 and P = 0.002 for effects of wtm2Δ and wtmΔ12 vs. control on HU induction, respectively. Strains used: TSR30-15, TSG10, TSG12, TSR1257, and TSG1.

Figure 6.—

Effects of WTM1 and WTM2 dosage on RNR2 and RNR1 reporter expression. Cells containing high-copy RNR2 or RNR1 reporter plasmids were grown in selective medium to midlog phase, incubated with or without 100 mm HU, and then assayed for β-galactosidase activity. (A) Effect of high-copy WTM1/2 plasmid compared to empty vector at 4 hr incubation. Values indicate mean and standard deviation of four independent isolates; effect of pWTM1/2 is statistically significant only for a twofold increase in the RNR2 reporter in untreated cells, P = 0.03. (B) Wild-type or wtmΔ12 deletion strains were assayed for β-galactosidase activity after 3 hr incubation. Values indicate mean and standard deviation of four independent isolates: for the RNR1 reporter, P < 0.001 for effect of wtmΔ12 in both treated and untreated cells; for the RNR2 reporter, P = 0.02 for effect of wtmΔ12 on HU response. Strains used: (A) TSR2147, TSR2148, TSR2150, and TSR2151; (B) TSR2167, TSR2170, TSR2169, and TSR2172.