Genetic screen for suppressors of increased silencing in rpd3 mutants in Saccharomyces cerevisiae identifies a potential role for H3K4 methylation

Abstract

Several studies have identified the paradoxical phenotype of increased heterochromatic gene silencing at specific loci that results from deletion or mutation of the histone deacetylase (HDAC) gene RPD3. To further understand this phenomenon, we conducted a genetic screen for suppressors of this extended silencing phenotype at the HMR locus in Saccharomyces cerevisiae. Most of the mutations that suppressed extended HMR silencing in rpd3 mutants without completely abolishing silencing were identified in the histone H3 lysine 4 methylation (H3K4me) pathway, specifically in SET1, BRE1, and BRE2. These second-site mutations retained normal HMR silencing, therefore, appear to be specific for the rpd3Δ extended silencing phenotype. As an initial assessment of the role of H3K4 methylation in extended silencing, we rule out some of the known mechanisms of Set1p/H3K4me mediated gene repression by HST1, HOS2, and HST3 encoded HDACs. Interestingly, we demonstrate that the RNA Polymerase III complex remains bound and active at the HMR-tDNA in rpd3 mutants despite silencing extending beyond the normal barrier. We discuss these results as they relate to the interplay among different chromatin-modifying enzyme functions and the importance of further study of this enigmatic phenomenon.

Keywords: BRE1; BRE2; RPD3; SET1; Saccharomyces cerevisiae; COMPASS; chromatin; histone modifications; silencing.

Figure 1

Schematic diagram of screen to identify suppressors of increased silencing in rpd3 mutants. (A) MATα strains with a defective native ade2-1 allele and containing functional ADE2 integrated downstream of the HMR-tDNA barrier express ADE2 and grow as white colonies. (B) Deletion of RPD3 results in increased extended silencing through the barrier that represses ADE2 expression and results in red colonies on minimal media containing suboptimal amounts of adenine. (C) LEU2-marked transposon mutagenesis of rpd3Δ strains identified rare Leu+ white colonies on media lacking leucine and containing suboptimal adenine. (D) White colonies were screened in a mating assay to identify and exclude unwanted nonmating isolates that have completely lost silencing (e.g., those with transposon insertions in the SIR genes).

Figure 2

Representative primary Leu+ isolates from mutagenesis of HMR-ADE2 rpd3Δ strains were restreaked multiple times for single colonies to verify stable propagation of the white colony phenotype. A single isolate containing the insertion in NGG1 gave rise to a variegated light pink phenotype. Additionally, isolates DDY5625 (set1) and DDY5692 (bre2) gave rise to rare very light pink colonies upon multiple restreaks of white colonies. DDY5640 and DDY5641 are two randomly selected Leu+ nonmating white colony isolates, and both contained different transposon insertions in the SIR4 gene consistent with a complete loss of silencing.

Figure 3

Colony color and mating phenotype of primary transposon mutants and haploid segregants after backcrossing to strain DDY814. (A) For each cross, the suppression phenotype bred true, as all double mutant segregants (Leu+ Kan+) produced white colonies, and segregants containing only the rpd3Δ::KanMX allele produced red colonies. Selected isolates were streaked on media lacking leucine and containing suboptimal adenine as in Figure 2. (B) The same his3 isolates were patched to YPD plates and after overnight growth were replica plated to a lawn of strain DDY20 (his4) on yeast minimal media lacking histidine. Each strain mated, confirming normal HMR silencing is not lost due to the second mutations. Parent strains DDY3133 and DDY2973 were also confirmed as mating, and DDY5641 containing a transposon insertion in SIR4 was included as a nonmating control.

Figure 4

Confirmation of suppression phenotypes by direct knockout of candidate genes. Parent strain DDY2973 was transformed with an LEU2 fragment containing homology to the flanking regions of each designated gene and plated on media lacking leucine and containing suboptimal adenine. White colonies were confirmed for the deletions by PCR at both ends of the deleted gene-LEU2 junctions. Confirmed isolates were restreaked as in Figures 2 and 3, and one red colony from the primary BRE1 transformation plate that tested negative for the deletion was included as an additional red control.

Figure 5

Alternate reporter assay for suppression of extended silencing phenotype. (A) Known and predicted phenotypes for each reporter strain are depicted as in Figure 1. Silencing of a1 in the MATα background allows normal mating and growth on minimal media lacking histidine, while expression of a1 results in a nonmating phenotype. (B) MATα his3 reporter strains were patched onto YPD plates and grown overnight, then replica plated onto a lawn of strain DDY20 (MATa his4) on YMD plates lacking histidine. Only mated diploid cells complemented for histidine biosynthesis grow. DDY282 lacks the barrier tDNA sequence, so a1 is silenced in this positive control for mating. Insertion of the barrier in DDY277 blocks propagation of silencing, allowing expression of a1 in the MATα cells to impart a nonmating phenotype. DDY5681 is identical to DDY277 except for the deletion of RPD3, resulting in the spread of silencing through the barrier to silence a1 and allow mating. The transposon insertion in BRE2 in DDY5679 suppresses the extended silencing caused by the deletion of RPD3, resulting in the expression of a1 and the nonmating phenotype.

Figure 6

Deletion of known effectors of H3K4me mediated gene repression do not affect extended rpd3Δ silencing. Strains were constructed in the rpd3Δ background to have deletions of HDAC genes HST3, HST1, or both HST1 and HOS2. Since hst1Δhos2Δ strains grew slowly, patches of ∼5 mm diameter of each strain were made to obtain comparable growth to verify the red phenotype (lower panels).

Figure 7

Binding of RNA Polymerase III transcription complex and HMR-tRNA expression are not affected by increased rpd3Δ silencing. (A) Chromatin immunoprecipitation was performed using anti-FLAG antibody in wild type and rpd3Δ strains containing FLAG-tagged Brf1p. PCR primer sets distal to (B and D) and overlapping the HMR-tDNA (C) were used to determine Pol III complex formation. Primer set A surrounds a tDNA distal to the HMR domain and was used as a positive control. A strain lacking the FLAG epitope on Brf1p was used as a negative control. (B) Northern blot analysis of a marked HMR-tDNA confirms its expression in rpd3Δ mutants. The HMR-tRNA transcript is detected using a complementary oligonucleotide probe specific to the 19 base pair extension (DDO-767, Table 3) to distinguish it from transcripts emanating the other seven copies of this tRNA^Thr isoacceptor. Strains deleted for the HMR-tDNA show no signal, confirming the specificity of the assay. The blot was stripped and reprobed with a bulk tRNA^Thr 76 base oligonucleotide probe (complementary to the tRNA^Thr(AGU)C, SGD YNCC0014W final processed transcript) as a loading control. Strains used are (left to right) DDY3401, 3402, 3403, 3404, 3396, 3398, 465, and 466.