Recruitment of Rpd3 to the telomere depends on the protein arginine methyltransferase Hmt1

Abstract

In the yeast Saccharomyces cerevisiae, the establishment and maintenance of silent chromatin at the telomere requires a delicate balance between opposing activities of histone modifying enzymes. Previously, we demonstrated that the protein arginine methyltransferase Hmt1 plays a role in the formation of yeast silent chromatin. To better understand the nature of the Hmt1 interactions that contribute to this phenomenon, we carried out a systematic reverse genetic screen using a null allele of HMT1 and the synthetic genetic array (SGA) methodology. This screen revealed interactions between HMT1 and genes encoding components of the histone deacetylase complex Rpd3L (large). A double mutant carrying both RPD3 and HMT1 deletions display increased telomeric silencing and Sir2 occupancy at the telomeric boundary regions, when comparing to a single mutant carrying Hmt1-deletion only. However, the dual rpd3/hmt1-null mutant behaves like the rpd3-null single mutant with respect to silencing behavior, indicating that RPD3 is epistatic to HMT1. Mutants lacking either Hmt1 or its catalytic activity display an increase in the recruitment of histone deacetylase Rpd3 to the telomeric boundary regions. Moreover, in such loss-of-function mutants the levels of acetylated H4K5, which is a substrate of Rpd3, are altered at the telomeric boundary regions. In contrast, the level of acetylated H4K16, a target of the histone deacetylase Sir2, was increased in these regions. Interestingly, mutants lacking either Rpd3 or Sir2 display various levels of reduction in dimethylated H4R3 at these telomeric boundary regions. Together, these data provide insight into the mechanism whereby Hmt1 promotes the proper establishment and maintenance of silent chromatin at the telomeres.

Figure 1. Synthetic Genetic Array Analysis of HMT1.

A) HMT1 genetic interaction network. Physical interactions between HMT1 and all of its genetic interactors identified from this study using the Synthetic Genetic Array methodology. The product of the gene identified is indicated within the relevant circle. The blue line represents a previously identified physical interaction. The complete genetic interaction network was created using Cytoscape, and the physical interaction data were obtained from the Saccharomyces Genome Database , . B) GO terms for the Δhmt1 SGA interaction data set reveals enrichment for components of the Rpd3(L) complex . Overrepresented GO terms and corresponding adjusted p-values were determined using FuncAssociate 2.0. GO term attributes are listed based on ascending adjusted p-values , .

Figure 2. Genetic interactions between HMT1 and genes encoding Rpd3L complex components.

A) Spot assay of 10-fold serially diluted haploid cells with single Rpd3 complex component deletion (+Kan) or double mutant selection (+Kan+Nat) resulting from mating with either HMT1 or Δhmt1 query strains. B) Growth of haploid, double-drug resistant strains produced from matings between strains deleted for genes encoding components of the Rpd3L complex, or components specific for Rpd3S complex (Δeaf3 and Δrco1), with either HMT1 or Δhmt1 query strains. C) Tabulated results of growth differences for Rpd3 complex components from genome-wide SGA screen.

Figure 3. Rpd3 occupancy at telomeric boundary regions is increased in the Hmt1 loss-of-function mutants.

A) Schematic representation of telomere VI-R and the quantitative PCR primer sets (A–E) for loci examined by ChIP in this study. B) Rpd3 occupancy across the telomeric boundary region in Hmt1 loss-of-function mutants. ChIP was performed using anti-Myc antibody to immunoprecipitate myc-tagged Rpd3 from wild-type, Δhmt1, and hmt1(G68R) cells. Bars represent the experimental signal normalized to signal from a non-transcribed intergenic region (“(–) Ctrl”). Error bars represent standard deviation of three biological samples (n = 3) per genotype, and asterisks denote p-value of 0.05 by Student’s t-Test.

Figure 4. Epistatic analysis of silencing in Hmt1 and Rpd3 mutants.

A) Telomeric Silencing Assay comparing Δhmt1/Δrpd3 to either Δhmt1 or Δrpd3 single mutant. B) Sir2 occupancy across the telomeric boundary region in Δhmt1, Δrpd3, or Δhmt1/Δrpd3 mutants. ChIP was performed using anti-Sir2 antibody to immunoprecipitate Sir2 from Δhmt1, Δrpd3, and Δhmt1/Δrpd3 cells. Primer sets are the same as in Fig. 3. Bars represent the experimental signal normalized to the GAL1 ORF. Error bars represent standard deviation of three biological samples (n = 3) per genotype, and asterisks denote p-value of 0.05 by Student’s t-Test.

Figure 5. Hmt1 loss-of-function mutants display various changes in the occupancy of acetyl-K5 and -K16 of histone H4 at telomeric boundary regions.

Directed ChIPs using anti-acetyl-H4K5 antibody (part A), or anti-acetyl-H4K16 antibody (part B) in wild-type, Δhmt1, and hmt1(G68R) cells. Primer sets used for this analysis were the same as those used in Fig. 3. Bars represent the experimental signals normalized to signal for the highly transcribed control, ACT1. Error bars represent standard deviation of three biological samples (n = 3) per genotype, and asterisks denote p-value of 0.05 by Student’s t-Test.

Figure 6. The effects of Rpd3 and Sir2 on H4R3 methylation.

Directed ChIP was performed using anti-dimethyl-H4R3 (H4R3me2) antibody in wild-type, Δrpd3 (part A), and Δsir2 (part B) cells. Primer sets used for this analysis were the same as in Fig. 3. Bars represent the experimental signals normalized to the highly transcribed control, ACT1. Error bars represent standard deviation of three biological samples (n = 3) per genotype, and asterisks denote p-value of 0.05 by Student’s t-Test.