Sir2 deacetylates histone H3 lysine 56 to regulate telomeric heterochromatin structure in yeast

Abstract

At telomeric heterochromatin in yeast, the Sir protein complex spreads from Rap1 sites to silence adjacent genes. This cascade is believed to occur when Sir2, an NAD(+)-dependent enzyme, deacetylates histone H3 and H4 N termini, in particular histone H4 K16, enabling more Sir protein binding. Lysine 56 of histone H3 is located at the entry-exit points of the DNA superhelix surrounding the nucleosome, where it may control DNA compaction. We have found that K56 substitutions disrupt silencing severely without decreasing Sir protein binding at the telomere. Our in vitro and in vivo data indicate that Sir2 deacetylates K56 directly in telomeric heterochromatin to compact chromatin and prevent access to RNA polymerase and ectopic bacterial dam methylase. Since the spread of Sir proteins is necessary but not sufficient for silencing, we propose that silencing occurs when Sir2 deacetylates H3 K56 to close the nucleosomal entry-exit gates, enabling compaction of heterochromatin.

Figure 1. Lysine 56 of histone H3 is essential for telomeric silencing in yeast

(A) Wild type and K56 substitution strains were assayed for silencing of a telomeric URA3 gene by monitoring cell growth on the 5-FOA plate. Tenfold serial dilutions of WT, K56G, K56Q and K56R cells were spotted onto SD-Trp⁻ plate lacking or containing 0.1% 5-FOA. The fraction of viable cells on the 5-FOA plates was determined as relative to the non-5-FOA plates. Averages of three independent experiments were graphed logarithmically. Position and transcriptional orientation of the telomeric URA3 gene were schematically represented under the graph. (B) Expression of a natural telomeric gene YFR057W was examined by quantitative RT-PCR in wild type and K56 substitution cells. Relative level of YFR057W mRNA in each strain was calculated by first normalizing to the internal control, SCR1, and then to that observed in the WT cells, which has been set to 1. Right panel shows a representative gel. Position and transcriptional orientation of the YFR057W gene were shown under the graph. Values are averages of three independent experiments with error bars shown for standard deviations.

Figure 2. Lysine 56 of histone H3 is not required for H4 K16 acetylation or Sir2 binding at telomere

(A) H4 K16 acetylation state and (B) Sir2 binding were assayed by ChIP using antibodies specific for the acetylated H4 K16 and Sir2 protein in WT and K56 substitution strains. Both H4 K16 acetylation and Sir2 binding data were normalized to an internal control (SPS2) and the input DNA. H4 K16 acetylation data were further normalized to histone H3 level, which was examined using an antibody specific to the histone H3 C- terminus. Generally, Sir2 binding at the telomeric silent region and the adjacent euchromatic region doesn't decrease in K56 substitution mutants (K56Q and K56R) as compared to WT cells. H4 K16 acetylation state at this 20-kb region in the mutants is also comparable to the WT strain. Gene map under the graphs shows the positions of fragments amplified in PCR. The results are averages of three independent ChIPs with error bars shown for standard deviations.

Figure 3. H3 K56 is hypo-acetylated at telomeres and hyper-acetylated at active genes

ChIP DNA of K56 acetylation and H3 C-terminus antibodies and input were amplified, fragmented, labeled and hybridized to GeneChip S.cerevisiae Tiling 1.0R Array. Moving averages of K56 acetylation, histone H3 and K56 / H3 data were plotted against (A) the distance from telomere end (window size, 20; step size, 500 bp) or (B) gene transcription rate (window size, 100; step size, 1).

Figure 4. Sir2 is required for deacetylating H3 K56 in vivo at telomere and HM loci

Relative H3 K56 acetylation and histone H3 levels were determined by ChIP assays using K56 acetylation and H3 C-terminus specific antibodies and normalized to an internal control (SPS2) and the input DNA. The relative level of K56 acetylation was further normalized to that of histone H3. K56 acetylation increases dramatically at telomere, E and I silencer of HML and HMR upon SIR2 deletion. In hst3Δ hst4Δ double deletion strain, H3 K56 acetylation increases in parallel at both the heterochromatic and the euchromatic regions, resulting in that the relative levels of K56 acetylation at silent loci to the internal control SPS2 are comparable to the wild type strain. Right panel shows representative gels. Values are averages of three independent experiments with error bars shown for standard deviations.

Figure 5. Sir2 deacetylates H3 K56 in vitro

(A) In vitro deacetylation assays were performed by incubating recombinant yeast Sir2 and various acetylated peptides in the presence and absence of NAD⁺. MALDI-TOF mass spectrometry chromatograms of the reaction mixture of (1) H3 K56 acetylated peptide without NAD⁺ or (2) with 1 mM NAD⁺, (3) H4 K16 acetylated peptide without NAD⁺ or (4) with 1 mM NAD⁺, (5) H4 K12 acetylated peptide without NAD⁺ or (6) with 1 mM NAD⁺, (7) H4 K16 acetylated peptide with vector control (pET) and (8) H3 K56 acetylated peptide with vector control (pET) were shown respectively. (B) Full length wild type core histones were incubated with recombinant yeast Sir2 in the presence and absence of 1 mM NAD⁺. A vector control experiment (pET) was performed in parallel. Reaction mixtures were resolved by 15% SDS-PAGE and probed with anti-H3 K56 Ac, anti-H4 K16 Ac and anti-H4 K12 Ac antibodies in the western blots. The bottom panel shows Coomassie Blue stained gel as a loading control.

Figure 6. Telomeric chromatin is accessible to dam methylase in K56 substitution mutants

Yeast genomic DNA was isolated from wild type (WT), K56G, K56Q and K56R cells expressing the E.coli dam methyltransferase. DNA samples were first digested with Nde I (lanes −) to yield a 870 bp telomeric fragment (fragment A), then a fraction of Nde I digested DNA was further cleaved with Dpn I or Mbo I or Sau3 AI (lanes +). Enzyme digested DNA samples were subjected to southern blot analysis using a 544 bp telomere VI-R DNA probe (indicated by a solid box in the gene map under the graphs). Percentile of enzyme cleavage in each strain was calculated by dividing the sum of the intensities of B and C bands by the sum of the intensities of all the bands in lane + ( (B+C)/(A+B+C) ). Right panel shows representative southern blots. Position of the telomeric southern blot region was shown under the graph. D, Dpn I, Mbo I and Sau3 AI site; N, Nde I site.