Histone H3 lysine 4 hypermethylation prevents aberrant nucleosome remodeling at the PHO5 promoter

Abstract

Recent studies have highlighted the histone H3K4 methylation (H3K4me)-dependent transcriptional repression in Saccharomyces cerevisiae; however, the underlying mechanism remains inexplicit. Here, we report that H3K4me inhibits the basal PHO5 transcription under high-phosphate conditions by suppressing nucleosome disassembly at the promoter. We found that derepression of the PHO5 promoter by SET1 deletion resulted in a labile chromatin structure, allowing more binding of RNA polymerase II (Pol II) but not the transactivators Pho2 and Pho4. We further showed that Pho23 and Cti6, two plant homeodomain (PHD)-containing proteins, cooperatively anchored the large Rpd3 (Rpd3L) complex to the H3K4-methylated PHO5 promoter. The deacetylation activity of Rpd3 on histone H3 was required for the function of Set1 at the PHO5 promoter. Taken together, our data suggest that Set1-mediated H3K4me suppresses nucleosome remodeling at the PHO5 promoter so as to reduce basal transcription of PHO5 under repressive conditions. We propose that the restriction of aberrant nucleosome remodeling contributes to strict control of gene transcription by the transactivators.

Fig. 1.

Set1-mediated H3K4 trimethylation associates with the PHO5 promoter and represses PHO5 transcription. (A) Schematic of the nucleosomal organization of the repressed promoter. The primer sets used for ChIP analysis are listed in the text. The light-gray ovals indicate nucleosomes that are removed during activation of PHO5, while the dark-gray ovals indicate nucleosomes that are not removed. (B) ChIP analysis of H3K4me3 at the promoter of PHO5. Yeast cells grown in P_i⁺ medium were shifted to phosphate-depleted medium, and the cells were harvested for ChIP at timed intervals. The H3K4me3 was corrected by histone H4 occupancy. The IP (% Input) at the PHO5 promoter and the subtelomeric region of chromosome VI-R are shown. (C and D) Analysis of PHO5 induction in wild-type and set1Δ cells. (C) Relative mRNA level of PHO5 at timed intervals after P_i withdrawal, normalized to ACT1. (D) The difference between the beginning and ending levels was set as 100%, and the values at the indicated times were determined. (E) Analysis of PHO5 induction in wild-type and H3K4A mutant cells. mRNA analysis was performed as described for panel C. (F) ChIP analysis of myc-tagged Rpb3 at PHO5. The corresponding regions of the PCR products are shown in panel A. Error bars represent standard errors of the means from three trials.

Fig. 2.

Set1-mediated H3K4me does not affect the recruitment of Pho2 and Pho4 onto the PHO5 promoter. (A) Schematic of the regulatory circuit for phosphatase genes (46). (B) mRNA analysis of PHO5 in cells from different isogenic strains. Yeast cells were cultured in either P_i⁺ medium (top) or P_i⁻ medium (bottom). (C) ChIP analysis of H3K4me3 at the PHO5 promoter. The conditions and the isogenic strains are indicated. ChIP experiments were performed as described for Fig. 1B. (D) Confocal images of the immunolocalization of myc-tagged Pho4 in wild-type and set1Δ cells. Pho4-myc was stained with mouse anti-Myc monoclonal antibody and detected with a Cy3-conjugated secondary antibody. DNA is stained with DAPI (4′,6-diamidino-2-phenylindole). (E and F) ChIP analysis of myc-tagged Pho4 and Pho2 level at the PHO5 promoter region in wild-type and set1Δ cells. Experiments were performed as described for Fig. 1B. Error bars represent standard errors of the means from three trials.

Fig. 3.

H3K4 methylation suppresses nucleosome remodeling at the PHO5 promoter. (A and B) ChIP analysis of histone H4 level at the promoter (A) and the 3′ ORF (B) of PHO5 in wild-type and set1Δ cells upon phosphate starvation. ChIP assays were performed as described for Fig. 1B. (C and D) ChIP analysis of myc-tagged Spt6 at the promoter (C) and the 3′ ORF (D) of PHO5 in wild-type and set1Δ cells upon phosphate starvation. ChIP assays were performed as described for Fig. 1B. (E) Schematic of the extrachromosomal copy of HHT1-HHF1 expressing H3 and HA-tagged H4 (H4-HA) under the control of the GAL1 promoter. (F) Schematic of the experimental procedures used to obtain results shown in panels G through I. (G to I) ChIP of H4-HA at the PHO5 promoter (G), the 3′ ORF (H), and the ACT1 ORF (I) after galactose induction and P_i withdrawal. Error bars represent standard errors of the means from three trials.

Fig. 4.

SET1 deletion leads to a more accessible chromatin structure at the PHO5 promoter. (A) Schematic of the recognition sites of the restriction enzymes used in nuclease assays and Southern blotting. (B and C) Restriction enzyme digestion of yeast nuclei. Nuclei were digested with 5 U or 50 U ClaI (B) or HhaI (C) for 120 min. To monitor the extent of cleavage, DNA was isolated, cleaved with HaeIII, separated by a 1.5% agarose gel, blotted, and hybridized with the probe indicated in panel A. Histograms represent the percentages of enzyme cleavage, quantified by phosphorimager. (D) MNase digestion of yeast nuclei. Nuclei were digested with increasing amounts of MNase for 15 min. Southern blotting was performed as described for panel B, except that the DNA was cleaved with ApaI. Black arrows denote sensitive sites representing the internucleosomal regions. Positioned nucleosomes at the PHO5 promoter are represented (ovals) on the right.

Fig. 5.

H3K4me recruits the Rpd3L complex for PHO5 repression. (A) ChIP analysis of myc-tagged Rpd3 at the PHO5 promoter in wild-type and set1Δ cells. Yeast cells were grown in P_i⁺ medium. (B) Analysis of PHO5 induction in wild-type, rpd3Δ, and rpd3Δ set1Δ cells. PHO5 mRNA was analyzed as described for Fig. 1C. (C) mRNA analysis of PHO5 in wild-type and Rpd3 complex mutant cells. Yeast was grown in P_i⁺ medium. The wild-type level of PHO5 was set as 1. The log₂ values above 0 indicate increased transcription of PHO5. (D) ChIP analysis of myc-tagged Rpd3 at the PHO5 promoter in wild-type and mutant cells. ChIP was performed as described for panel A. Error bars represent standard errors of the means from three trials. (E) Pulldown assay of the PHD of either Cti6 or Pho23 with H3 N-terminal peptides. Agarose beads conjugated by peptides un-, mono-, di-, or trimethylated at H3K4 were incubated with equal amounts of recombinant proteins as indicated on the right. The peptide-bound proteins were detected by Western blotting with anti-GST antibody. “Mock” indicates a no-peptide control. Coomassie blue-stained peptides are shown at the bottom. (F) Pulldown assay of whole-cell extracts with H3 N-terminal peptides. The whole-cell extracts were derived from yeasts with the genetic backgrounds indicated on the left. The bound Rpd3-myc was detected with anti-myc antibody.

Fig. 6.

Set1 suppresses nucleosome disassembly by Rpd3-dependent H3 deacetylation. (A) ChIP analysis of histone H4 level at the promoter region of PHO5 in wild-type and rpd3 mutant cells upon phosphate starvation. ChIP assays were performed as described for Fig. 3A. (B and C) ChIP analysis of AcH3 (B) and AcH4 (C) at the promoter region of PHO5 in wild-type and mutant cells upon phosphate starvation. ChIP assays were performed as described for Fig. 1B. The IP (% Input) values of histone acetylation were then corrected by histone H4 density. Error bars represent standard errors of the means from three trials.

Fig. 7.

Predicted model for the regulation of nucleosome remodeling at the PHO5 promoter. (A) Under P_i⁺ conditions, the four nucleosomes at the PHO5 promoter are well positioned, inaccessible to chromatin remodelers or coactivators. Histone H4 (shown as red sector) N-terminal lysines are acetylated by Pho2-recruited NuA4. H4 acetylation is required but not sufficient for initiation of promoter chromatin remodeling. Histone H3 (shown as blue sector) is hypermethylated at K4, which recruits the Rpd3L complex to maintain a hypoacetylated state of H3 lysines. At this stage, there is only basal PHO5 transcription. (B) Upon P_i withdrawal, histone modifiers and chromatin remodelers are targeted to the PHO5 promoter by the relocalized Pho4. Remodeling of the promoter nucleosomes requires acetylation of histone H3. Once acetylated, histone H3 is rapidly removed from the PHO5 promoter; otherwise, it will be deacetylated by the Rpd3L complex. Therefore, the relative acetylation level at the PHO5 promoter is essentially stable during the induction process. At this intermediate stage, the PHO5 transcript is being accumulated but still at a low level. (C) After growth in P_i⁻ medium, nucleosomes at the PHO5 promoter are completely unfolded, allowing efficient action of the transcriptional machinery. At this stage, Set1-mediated H3K4me is dispensable. (D) When SET1 is deleted, the promoter becomes partially accessible to chromatin remodelers even under P_i⁺ conditions. Loss of Rpd3 increases histone H3 acetylation and thereby elevates basal chromatin remodeling. The four nucleosomes are dynamically disassembled and reassembled (shown as dashed circles). Such a labile promoter structure results in high basal transcription of PHO5 even under repressed conditions.