PMC

Altered distribution of Gcn5 on individual long and short genes in response to KCl induced stress. Examples of Gcn5-myc enrichment in the absence (blue curve) and presence of KCl (red curve) on individual long genes (left panels) and short genes (right panels) are shown. The data plotted for individual genes is taken from the data used for average gene analysis in Fig. 5. Gene length is indicated in parenthesis.

A subset of KCl response genes are general environmental stress response genes in S. cerevisiae. (A) Scatter plot showing the expression levels of genes in the absence (YPD) and presence of KCl (YPD+KCl). The diagonal lines indicate a 2-fold change in expression. (B) A subset of KCl response genes is common environmental stress response genes. Venn diagram of genes differentially regulated during the KCl stress response (left) and genes previously [] defined as environmental stress response genes (right).

Gcn5 differs in gene regulation and genome-wide localisation in S. cerevisiae and S. pombe during KCl stress adaptation. (A) Gcn5 regulates different sets of KCl responsive genes in S. cerevisiae and S. pombe. The Venn diagram shows the relationship between gene sets identified in S. cerevisiae and S. pombe. (B) The altered Gcn5-myc gene distribution pattern between normal and KCl-stress conditions in S. cerevisiae is not observed in S. pombe. The average gene distribution of Gcn5-myc is shown for genes that are associated with Gcn5 in both the absence and presence of KCl stress. The distribution of Gcn5 is shown for conditions in the absence (blue line) and presence of KCl (red line) for S. cerevisiae (average of 383 genes, upper panel) and S. pombe (average of 504 genes, lower panel), respectively.

Genome-wide re-distribution of Gcn5 when cells are exposed to KCl stress. Average gene analysis shows the average gene distribution of Gcn5-myc enrichment in the absence (left panel) and presence of KCl (right panel). (A) Average Gcn5-myc localisation is shown for groups of up-regulated, down-regulated and unchanged genes. (B) Gcn5-myc is preferentially enriched in highly expressed genes in the absence and presence of KCl stress. Genes are grouped according to their expression level as described in Methods section. (C) Gcn5-myc is preferentially enriched on long genes under KCl stress but is under-represented on such genes in the absence of stress. Genes are grouped according to their ORF length. (D) Gcn5 redistributes between groups of genes with different Gcn5-binding patterns in response to KCl stress. The graphs show the average distribution of Gcn5 in 5 different K-means clusters that were selected to reveal gene groups with different patterns of Gcn5 localisation throughout the genes. The shift of Gcn5 association between groups in response to KCl stress is shown.

Gcn5 is associated with the transcribed region of the FLO8 gene and causes its repression during KCl stress. (A) Verification of FLO8 regulation and binding by Gcn5. The relative expression level of FLO8 gene in wild type and gcn5Δ cells under KCl stress conditions is shown (left panel) in relation to a control gene, YDL212W. Middle panel: tiling array data visualized by IGB showing relative enriched association of Gcn5-myc with FLO8 in immunoprecipiated material normalized against input material. The arrow shows the direction of transcription. A and B represent the location of the amplified binding region and control region used for semi quantitative PCR for ChIP, Right panel: semi quantitative PCR validation of Gcn5 association with FLO8 showing the level of Gcn5-myc associated chromatin in relation to the input for regions A and B as designated in the middle panel. (B) Diagram showing how deletion of Gcn5 causes indirect regulation of FLO8 target genes. FLO8Δ vs wild type microarray data are from ([] GSE4654).

Identification of Gcn5 dependent KCl stress response genes in S. cerevisiae. (A) The Gcn5 dependency of global gene expression increases during KCl induced stress. Graph shows the frequency distribution (y-axis, arbitrary units) of the Gcn5 dependency of genes (x-axis) in the absence (black curve) and presence (red curve) of KCl. The Gcn5 dependency for each gene is expressed as the ratio between the mean transcript level in the gcn5Δ strain in relation to the mean transcript level in the wild type strain. (B) Diagram illustrating how Gcn5 dependent KCl response genes were defined. The selected genes are differentially regulated in wild type cells during KCl stress conditions in a fashion that requires Gcn5 when gcn5Δ is compared to wild type during KCl adaptation. Selection criteria are described in the Methods section. The number for of genes in each group is indicated. The heat map shows the fold change values for these genes in the two data sets used to identify them, +KCl/-KCl (wt) and gcn5Δ/wild type (gcn5Δ).

Gcn5 is required for several common stress responses in divergent yeast species and defects in its conserved histone acetyltransferase domain cause stress sensitivity. (A) Gcn5 is required for efficient growth under a common set of environmental stress conditions in evolutionarily divergent yeast species. The figure shows serial dilutions of wild type (wt) or mutant (gcn5Δ) yeast cells spotted on normal growth media (control) or media containing stress inducing levels of KCl, CaCl₂, Calcoflour White, MnCl₂ and caffeine. (B) The domain structure of Gcn5 showing the highly conserved histone acetyltransferase (HAT). The diagram shows the HAT, the Ada2-interaction domain (ADA2) and the Bromo domain (Bromo) domains defined previously for the S. cerevisiae protein []. Residue numbers equivalent to domain boundaries are shown. The extent of domain identity between S. cerevisiae and S. pombe is shown (%). The approximate position of each mutant triple-alanine substitution mutant, starting at residues 126, 129 and 132 respectively, is shown. (C) The HAT activity of Gcn5 is required for Gcn5 dependent stress responses. Cell plating assay performed as in Fig. 1A showing S. cerevisiae gcn5Δ cells carrying pRS316 plasmids expressing either wild type GCN5 or GCN5 with substitution mutations in the HAT domain as indicated in (B). Wild type (wt) and GCN5^- (gcn5Δ) cells were used as controls.