Chromatin disassembly from the PHO5 promoter is essential for the recruitment of the general transcription machinery and coactivators

Abstract

The disassembly of promoter nucleosomes appears to be a general property of highly transcribed eukaryotic genes. We have previously shown that the disassembly of chromatin from the promoters of the Saccharomyces cerevisiae PHO5 and PHO8 genes, mediated by the histone chaperone anti-silencing function 1 (Asf1), is essential for transcriptional activation upon phosphate depletion. This mechanism of transcriptional regulation is shared with the ADY2 and ADH2 genes upon glucose removal. Promoter chromatin disassembly by Asf1 is required for recruitment of TBP and RNA polymerase II, but not the Pho4 and Pho2 activators. Furthermore, accumulation of SWI/SNF and SAGA at the PHO5 promoter requires promoter chromatin disassembly. By contrast, the requirement for SWI/SNF and SAGA to facilitate Pho4 activator recruitment to the nucleosome-buried binding site in the PHO5 promoter occurs prior to chromatin disassembly and is distinct from the stable recruitment of SWI/SNF and SAGA that occurs after chromatin disassembly.

FIG. 1.

Asf1 mediates nucleosome disassembly at the ADH2 and ADY2 promoters. (A) ChIP analysis of histone H3 levels over the ADH2 TATA region. Histone H3 levels were measured by ChIP analysis in wild-type (WT) (JKT0010) and asf1Δ (JKT0018) strains growing in 2% glucose-containing media (+Glu) and 0.05% glucose-containing media (−Glu). As a control, primer sets were also used for the TELVIR region. (B) Quantitation of histone H3 ChIP analyses at the ADH2 TATA. (C) Quantitation of ADH2 transcript levels. Shown is quantitative RT-PCR analysis of ADH2 transcript levels in wild-type and asf1Δ strains grown as described for panel A. (D) ChIP analysis of histone levels over the ADY2 TATA region as described for panel A. (E) Quantitation of histone H3 ChIP as described for panel B, with primers directed against the TATA box region of ADY2. (F) Quantitation of ADY2 transcript levels as described for panel C. All histograms shown present the mean and standard deviation of three or more independent experiments.

FIG. 2.

Pho2 and Pho4 recruitment is not impaired in asf1 mutants. (A) Schematic of the PHO5 promoter region. The light gray circles represent nucleosomes that are disassembled upon transcriptional activation. Indicated are the locations of regions that are PCR amplified in the ChIP analyses. ORF, open reading frame. (B) ChIP analyses were performed in wild-type (WT) (MAY0068) and asf1Δ (MAY0074) strains growing in phosphate-containing media and then shifted to a phosphate-depleted medium for the indicated times. (C) Quantitation of ChIP analysis of Pho4 levels over the PHO5 UAS region, using Pho4:HA, normalized to the input and relative to the TELVIR control region. (D) Quantitation of ChIP analysis of Pho2 levels over the PHO5 UAS region, normalized as described above. (E) ChIP analyses using the same strains as panels C and D at the PHO5 UAS and the GAL1 UAS under repressing conditions (+P_i, +Glu), but including an unrelated antibody (myc) as a control. Note that the y axis is a log scale to allow comparison of the ChIP antibodies to the control antibody. All histograms shown present the mean and standard deviation of three independent experiments.

FIG. 3.

Asf1-mediated nucleosome disassembly is required for TBP and Pol II recruitment to the PHO5 promoter. (A) ChIP analysis of Pol II, histone H3, and TBP levels over the PHO5 promoter, measured by ChIP analysis in wild-type (WT) (LKY0069) and asf1Δ (MAY0063) strains growing in phosphate-containing media and then shifted to phosphate-depleted media for the indicated times. As a control, primer sets were used for a region in TELVIR. (B to D) Quantitation of ChIP analysis of TBP recruitment (B), Pol II recruitment (C), and histone H3 disassembly (D) over the PHO5 TATA region, normalized to input and control regions as described in the legend to Fig. 2. (E) ChIP analyses using the same strains as in panels B to D at the PHO5 TATA and GAL1 TATA under repressing conditions (+P_i, +Glu) but including an unrelated antibody (myc) as a control. Note that the y axis is a log scale to allow comparison of the ChIP antibodies to the control antibody. All histograms shown present the mean and standard deviation of three independent experiments.

FIG. 4.

Asf1-mediated nucleosome disassembly is required for stable Gcn5 and Snf2 recruitment. (A) SWI/SNF and SAGA are essential for PHO5 activation following phosphate depletion. Activation of the PHO5 gene was measured by acid phosphatase activity determinations in strain BY4741 (wild type [WT]) and the isogenic asf1Δ, gcn5Δ, spt7Δ, spt3Δ, and snf2Δ versions of BY4741 at the indicated times after transfer to phosphate-depleted media. (B) ChIP analysis of Gcn5/MYC and Snf2 levels at the PHO5 promoter. Gcn5 and Snf2 levels were measured by ChIP analysis in WT (Z1466) and asf1Δ (JCY0011) strains growing in phosphate-containing media and then shifted to a phosphate-depleted medium for the indicated times. (C) Quantitation of ChIP analysis of Gcn5/MYC and Snf2 levels over the PHO5 UAS region, normalized to input and the control region as described in the legend to Fig. 2. (D) ChIP analyses using the same strains as for panels B and C at the PHO5 UAS and the GAL1 UAS under repressing conditions (+P_i, +Glu) but including an unrelated antibody (HA) as a control. The large error bars reflect the extremely low signal that is essentially measuring background and is at the limit of quantitative detection for the real-time PCR analysis. (E) Asf1-mediated chromatin disassembly is required for activation by Pho4-VP16. Acid phosphatase levels were measured in strains BY4741 (WT), BY4741asf1Δ (asf1Δ), BY4741pho4Δ (pho4Δ), MAY1000 (pho4Δ Pho4VP16), and MAY10001 (asf1Δ pho4Δ Pho4VP16) at the indicated times after transfer to phosphate-depleted media to provide a measurement of PHO5 induction. All histograms shown present the mean and standard deviation of three independent experiments.

FIG. 5.

SWI/SNF and SAGA are required for chromatin disassembly to allow Pho4 recruitment prior to their stable recruitment. (A) Chromatin disassembly from nucleosome −1 of the PHO5 promoter in SAGA and SWI/SNF mutants. Histone occupancy at nuc −1 was determined by H3 ChIP analysis in strain BY4741 (wild type [WT]) and the isogenic asf1Δ, gcn5Δ, spt7Δ, spt3Δ, and snf2Δ versions of BY4741 at the indicated times after transfer to phosphate-depleted media. The data were normalized as described in the legend to Fig. 2. (B) Chromatin disassembly from nucleosome −2 of the PHO5 promoter in SAGA and SWI/SNF mutants, performed as described for panel A. (C) SWI/SNF and SAGA contribute to Pho4 recruitment to the PHO5 promoter. Pho4 occupancy at the PHO5 UAS was determined by ChIP analysis in strain JLY098 (WT) and isogenic SKW0015 (gcn5Δ) and SKW0025 (snf2Δ) strains at the indicated times after transfer to phosphate-depleted media. An unrelated antibody (myc) was used as a control. The data were normalized as described in the legend to Fig. 2. (D) Pho2 levels are not affected by deletion of SNF2 or GCN5. Western blot analysis in strain BY4741 (WT) and the isogenic gcn5Δ and snf2Δ versions of BY4741. Amido black was included as a loading control. All histograms shown present the mean and standard deviation of three independent experiments.

FIG. 6.

Stepwise model for transcriptional activation by chromatin disassembly. At the top is shown a schematic depicting the repressed PHO5 promoter. The binding sites for Pho4 and Pho2 are shown by red boxes. (Step 1) Upon phosphate depletion (activating conditions), the activators bind the DNA with the assistance of SWI/SNF and SAGA in wild-type strains, but in the absence of stable recruitment of SWI/SNF and SAGA to the promoter. (Step 2) Following activator recruitment, the promoter nucleosomes are disassembled by Asf1. (Step 3) Chromatin disassembly allows the recruitment of the general transcriptional machinery and stable recruitment of SAGA and SWI/SNF. However, upon phosphate depletion in asf1 mutants, nucleosome disassembly does not occur and SWI/SNF and/or SAGA can promote Pho4 binding to the promoter (step 1) but are unable to interact with the activator Pho4 to enable their recruitment to the PHO5 promoter (step 3).