Adr1 and Cat8 mediate coactivator recruitment and chromatin remodeling at glucose-regulated genes

Abstract

Background: Adr1 and Cat8 co-regulate numerous glucose-repressed genes in S. cerevisiae, presenting a unique opportunity to explore their individual roles in coactivator recruitment, chromatin remodeling, and transcription.

Methodology/principal findings: We determined the individual contributions of Cat8 and Adr1 on the expression of a cohort of glucose-repressed genes and found three broad categories: genes that need both activators for full derepression, genes that rely mostly on Cat8 and genes that require only Adr1. Through combined expression and recruitment data, along with analysis of chromatin remodeling at two of these genes, ADH2 and FBP1, we clarified how these activators achieve this wide range of co-regulation. We find that Adr1 and Cat8 are not intrinsically different in their abilities to recruit coactivators but rather, promoter context appears to dictate which activator is responsible for recruitment to specific genes. These promoter-specific contributions are also apparent in the chromatin remodeling that accompanies derepression: ADH2 requires both Adr1 and Cat8, whereas, at FBP1, significant remodeling occurs with Cat8 alone. Although over-expression of Adr1 can compensate for loss of Cat8 at many genes in terms of both activation and chromatin remodeling, this over-expression cannot complement all of the cat8Delta phenotypes.

Conclusions/significance: Thus, at many of the glucose-repressed genes, Cat8 and Adr1 appear to have interchangeable roles and promoter architecture may dictate the roles of these activators.

Figure 1. Recruitment profiles under derepressed conditions.

A–G: ChIP analysis for coactivators/general transcription machinery at the indicated promoter regions in a wildtype strain (black), adr1Δ (pink), cat8Δ (blue) and adr1Δcat8Δ (green). Binding is expressed as the percent of the wildtype derepressed value (set to 100%) after normalization to the TEL negative control locus. Error bars represent the standard deviation of biological replicates (two or more). Data is shown at 4 hours of derepression. H: ChIP analysis for Taf1 in repressing (grey bars) and derepressing (four hours, black bars) conditions. Error bars represent technical replicates. Inset: ChIP analysis by PCR at for Taf1 at ACT1. The protein(s) assayed for in each case is listed in Table 2.

Figure 2. Differential roles for Adr1 and Cat8 in chromatin remodeling at ADH2 and FBP1.

NuSA results are displayed as the amount of relative protection, after normalization to the well-positioned nucleosome at CEN3. The position of each amplicon (referenced to the middle of each amplicon) within the promoter is shown on the x-axis, with approximate location of nucleosomes shown. (A) and (C) are the results at ADH2, (B) and (D) are at FBP1. (A) and (B) compare NuSA results of Δadr1 (pink) and Δcat8 (green) in derepressed conditions to a wildtype strain either in repressed (dark blue) or derepressed (light blue) conditions. (C) and (D) compare NuSA results between over-expression of Adr1 with (red) or without Cat8 (blue) in derepressed conditions.

Figure 3. Adr1 and Cat8 play different roles at ADH2 and FBP1.

Comparison between the ADH2 (A and C, orange) and FBP1 (B and D, blue) promoters in either adr1Δ (A and B) or a cat8Δ (C and D). Combined expression, recruitment and chromatin remodeling data at 4 hours of derepression are shown. Values in parenthesis indicate the percent binding of the wild-type derepressed value, after normalization (an asterisk represents previously published data [9]). Gene activation is reported as the percent of the wild-type derepressed level. Binding sites are shown at their approximate locations along the promoter (Cat8 in green, Adr1 in pink, and TATA box in blue). Nucleosome positions under derepressed conditions are depicted as shaded ovals on the promoter, and the degree of chromatin remodeling estimated as a percent of the remodeling observed in a wild-type strain. Locations of coactivators are not intended to reflect interactions.