Histone H4 lysine 20 of Saccharomyces cerevisiae is monomethylated and functions in subtelomeric silencing

Abstract

Histones undergo post-translational modifications that are linked to important biological processes. Previous studies have indicated that lysine methylation correlating with closed or repressive chromatin is absent in the budding yeast Saccharomyces cerevisiae, including at H4 lysine 20 (K20). Here we provide functional evidence for H4 K20 monomethylation (K20me1) in budding yeast. H4 K20me1 is detectable on endogenous H4 by western analysis using methyl-specific antibodies, and the signal is abrogated by H4 K20 substitutions and by competition with H4 K20me1 peptides. Using chromatin immunoprecipitation, we show that H4 K20me1 levels are highest at heterochromatic locations, including subtelomeres, the silent mating type locus, and rDNA repeats, and lowest at centromeres within euchromatin. Further, an H4 K20A substitution strongly reduced heterochromatic reporter silencing at telomeres and the silent mating type locus and led to an increase in subtelomeric endogenous gene expression. The correlation between the location of H4 K20me1 and the effect of the H4 K20A substitution suggests that this modification plays a repressive function. Our findings reveal the first negative regulatory histone methylation in budding yeast and indicate that H4 K20me1 is evolutionarily conserved from simple to complex eukaryotes.

Figure 1

H4 K20me1 detectable in budding yeast. (A) Peptides matching the first 30 amino acids of H4 with or without acetylated lysine 16 or monomethylated lysine 20 were spotted onto PVDF and probed with antibodies. Peptides, peptide amounts, and antibodies are respectively indicated to the left, top, and bottom of the blots. (B) H4 K20me1 is detectable in whole-cell extracts (WCEs) from strains of different genetic backgrounds, mating types, and ploidies, is detectable whether H4 genes are present in the genome or on a plasmid, and is abrogated by an H4 K20R substitution. WCEs were analyzed on polyacrylamide gels, transferred to PVDF, and probed with antibodies. H4 levels are a loading control. Calf thymus H4 is a positive antibody control. Antibodies are indicated to the left of each blot. Strains, genetic backgrounds, and mating types are indicated above each lane. Strains FY1716, YKI071, and YWD193 have H3/H4 genes present only on a plasmid. (C) To confirm that the H4 K20me1 signal is H4 and not a similarly migrating protein, WCEs with untagged H4, FLAG-H4, or FLAG-H4 K20R were subjected to FLAG-affinity purification. WCEs and elutions were analyzed by western analyses. Antibodies are indicated to the left of each blot. The FLAG tag and K20R substitution are indicated above each lane. An asterisk indicates a non-specific band that obscures FLAG-H4 in WCEs. Arrows indicate untagged H4 in WCEs and FLAG-H4 in eluates. (D & E) H4 K20me1 western signals preferentially competed by monomethylated rather than unmethylated H4 peptides. WCEs were analyzed on polyacrylamide gels, transferred to PVDF, and probed with anti-H4 K20me1 antibodies (1mg/mL) that were pre-incubated with no peptides or H4 peptides (100ng/mL or 1ng/mL in D, 100ng/mL in E) that were or were not monomethylated at lysine 20. H4 levels are a loading control. Antibodies are indicated to the left of each blot. Strains, peptides, and peptide concentrations are indicated above each blot. Arrow indicates H4 K20me1. Asterisk indicates a non-specific band.

Figure 2

H4 K20me1 detectable with Millipore antibody 04–735. (A, B) Peptide dot blots as in figure 1A. (C) Substitutions of modifiable histone residues do not affect H4 K20me1 levels and H4 K20 substitutions do not affect H4 K16ac levels. WCEs from yeast with wild-type or mutant histones were analyzed by western analyses. Antibodies and histone mutations are respectively indicated to the left of and above each blot. Asterisk indicates a non-specific band present on blot probed for H4 K16ac. (D) Western blots with peptide competitions as in 1D, E.

Figure 3

Deleting and overexpressing candidate genes does not change H4 K20me1 levels. (A, B) FLAG-H4 was purified by FLAG-affinity purification from strains with candidate HMTs deleted. WCEs and FLAG elutions were analyzed by western analysis. Antibodies and strains are respectively indicated to the left of and above each blot. FLAG levels are a loading control. (C) WCEs from strains with candidate HMTs overexpressed from a plasmid were analyzed by western analysis. Antibodies and strains are respectively indicated to the left of and above each blot. H4 levels are a loading control. An H4 K20R strain serves as a negative antibody control.

Figure 4

ChIP-qPCR analysis of H4 K20me1. Sonicated chromatin from yeast with WT or K20R histone H4 were immunoprecipitated by antibodies against H4 K20me1 or total H4 and co-purified DNA was analyzed by qPCR. Bars indicate H4 K20me1 levels normalized to total H4 levels (A) or no antibody or H4 K20me1 levels normalized to input levels (B). Genomic locations are indicated below each pair of bars. (C) Comparison of H3 K4me3 versus H4 K20me1 levels relative to H4 levels at several loci. (D) H4 K20me1 levels relative to H4 levels across several regions: ACT1 ORF, RCK2-YEF3 region, and YLR454w region. Horizontal bars and arrows underneath graph represent genome and genes respectively. Position of bar graph bars above genes indicates locations of checked regions. A, B, and D represent means and standard error of the mean (S.E.M.) of three experiments.

Figure 4

ChIP-qPCR analysis of H4 K20me1. Sonicated chromatin from yeast with WT or K20R histone H4 were immunoprecipitated by antibodies against H4 K20me1 or total H4 and co-purified DNA was analyzed by qPCR. Bars indicate H4 K20me1 levels normalized to total H4 levels (A) or no antibody or H4 K20me1 levels normalized to input levels (B). Genomic locations are indicated below each pair of bars. (C) Comparison of H3 K4me3 versus H4 K20me1 levels relative to H4 levels at several loci. (D) H4 K20me1 levels relative to H4 levels across several regions: ACT1 ORF, RCK2-YEF3 region, and YLR454w region. Horizontal bars and arrows underneath graph represent genome and genes respectively. Position of bar graph bars above genes indicates locations of checked regions. A, B, and D represent means and standard error of the mean (S.E.M.) of three experiments.

Figure 5

Phenotype analysis of H4 K20me1. (A) Dilutions of hml∷URA3 ADE2-TEL-VR yeast with wild-type or mutant histones were grown on synthetic complete (SC), SC-URA, or SC+5FOA media. Media and strains are indicated above and to the right of each plate respectively. Increased growth on SC-URA media and decreased growth on SC+5FOA media compared to SC media indicate increased URA3 expression. White and red yeast coloration respectively indicates ADE2 expression and silencing. (B) H4 K20me1 levels at and expression of genes proximal and distal to telomere 7L were determined by ChIP of H4 K20me1 and total H4 and by qPCR of harvested RNAs. (top) ChIP was as in Figure 4. H4 K20me1 levels per total H4 levels are shown with means and standard errors of the mean. ChIP locations residing along chromosome 7L are indicated in the map by black triangles while above numbers indicate which locations correspond to which bars in the above graph. Locations 2–5 were compared to locations 6–8 using a Wilcoxon rank sum test and generated the indicated P-value. (bottom) RNA was extracted from yeast with WT, K20R, or K20A histone H4. mRNA levels of genes proximal and distal to telomere 7L were determined by qPCR and normalized to ACT1 mRNA. Means and standard error of the means are shown. WT and H4 K20A samples were compared using a T-test and generated the indicated P-values. (C) H4 K20me1 levels are lower in replicatively older than younger yeast. WCEs from replicatively young or old yeast were analyzed by western analyses. Antibodies are indicated to the left of each blot. Samples and strains are indicated above each lane. Average numbers of bud scars (cell divisions) for each yeast sample are indicated in parentheses. H3 levels are a loading control.

Figure 5

Phenotype analysis of H4 K20me1. (A) Dilutions of hml∷URA3 ADE2-TEL-VR yeast with wild-type or mutant histones were grown on synthetic complete (SC), SC-URA, or SC+5FOA media. Media and strains are indicated above and to the right of each plate respectively. Increased growth on SC-URA media and decreased growth on SC+5FOA media compared to SC media indicate increased URA3 expression. White and red yeast coloration respectively indicates ADE2 expression and silencing. (B) H4 K20me1 levels at and expression of genes proximal and distal to telomere 7L were determined by ChIP of H4 K20me1 and total H4 and by qPCR of harvested RNAs. (top) ChIP was as in Figure 4. H4 K20me1 levels per total H4 levels are shown with means and standard errors of the mean. ChIP locations residing along chromosome 7L are indicated in the map by black triangles while above numbers indicate which locations correspond to which bars in the above graph. Locations 2–5 were compared to locations 6–8 using a Wilcoxon rank sum test and generated the indicated P-value. (bottom) RNA was extracted from yeast with WT, K20R, or K20A histone H4. mRNA levels of genes proximal and distal to telomere 7L were determined by qPCR and normalized to ACT1 mRNA. Means and standard error of the means are shown. WT and H4 K20A samples were compared using a T-test and generated the indicated P-values. (C) H4 K20me1 levels are lower in replicatively older than younger yeast. WCEs from replicatively young or old yeast were analyzed by western analyses. Antibodies are indicated to the left of each blot. Samples and strains are indicated above each lane. Average numbers of bud scars (cell divisions) for each yeast sample are indicated in parentheses. H3 levels are a loading control.

Figure 5

Phenotype analysis of H4 K20me1. (A) Dilutions of hml∷URA3 ADE2-TEL-VR yeast with wild-type or mutant histones were grown on synthetic complete (SC), SC-URA, or SC+5FOA media. Media and strains are indicated above and to the right of each plate respectively. Increased growth on SC-URA media and decreased growth on SC+5FOA media compared to SC media indicate increased URA3 expression. White and red yeast coloration respectively indicates ADE2 expression and silencing. (B) H4 K20me1 levels at and expression of genes proximal and distal to telomere 7L were determined by ChIP of H4 K20me1 and total H4 and by qPCR of harvested RNAs. (top) ChIP was as in Figure 4. H4 K20me1 levels per total H4 levels are shown with means and standard errors of the mean. ChIP locations residing along chromosome 7L are indicated in the map by black triangles while above numbers indicate which locations correspond to which bars in the above graph. Locations 2–5 were compared to locations 6–8 using a Wilcoxon rank sum test and generated the indicated P-value. (bottom) RNA was extracted from yeast with WT, K20R, or K20A histone H4. mRNA levels of genes proximal and distal to telomere 7L were determined by qPCR and normalized to ACT1 mRNA. Means and standard error of the means are shown. WT and H4 K20A samples were compared using a T-test and generated the indicated P-values. (C) H4 K20me1 levels are lower in replicatively older than younger yeast. WCEs from replicatively young or old yeast were analyzed by western analyses. Antibodies are indicated to the left of each blot. Samples and strains are indicated above each lane. Average numbers of bud scars (cell divisions) for each yeast sample are indicated in parentheses. H3 levels are a loading control.