Genome-wide recruitment profiling of transcription factor Crz1 in response to high pH stress

Abstract

Background: Exposure of the budding Saccharomyces cerevisiae to an alkaline environment produces a robust transcriptional response involving hundreds of genes. Part of this response is triggered by an almost immediate burst of calcium that activates the Ser/Thr protein phosphatase calcineurin. Activated calcineurin dephosphorylates the transcription factor (TF) Crz1, which moves to the nucleus and binds to calcineurin/Crz1 responsive gene promoters. In this work we present a genome-wide study of the binding of Crz1 to gene promoters in response to high pH stress.

Results: Environmental alkalinization promoted a time-dependent recruitment of Crz1 to 152 intergenic regions, the vast majority between 1 and 5 min upon stress onset. Positional evaluation of the genomic coordinates combined with existing transcriptional studies allowed identifying 140 genes likely responsive to Crz1 regulation. Gene Ontology analysis confirmed the relevant impact of calcineurin/Crz1 on a set of genes involved in glucose utilization, and uncovered novel targets, such as genes responsible for trehalose metabolism. We also identified over a dozen of genes encoding TFs that are likely under the control of Crz1, suggesting a possible mechanism for amplification of the signal at the transcription level. Further analysis of the binding sites allowed refining the consensus sequence for Crz1 binding to gene promoters and the effect of chromatin accessibility in the timing of Crz1 recruitment to promoters.

Conclusions: The present work defines at the genomic-wide level the kinetics of binding of Crz1 to gene promoters in response to alkaline stress, confirms diverse previously known Crz1 targets and identifies many putative novel ones. Because of the relevance of calcineurin/Crz1 in signaling diverse stress conditions, our data will contribute to understand the transcriptional response in other circumstances that also involve calcium signaling, such as exposition to sexual pheromones or saline stress.

Keywords: Alkaline pH stress; Calcineurin signaling; Consensus binding sequence; Crz1 recruitment; Saccharomyces cerevisiae.

Fig. 1

Kinetics of Crz1 recruitment to gene promoters upon high pH stress. a Time-course of accumulation of Crz1 at intergenic regions. The time-point at which the higher number of reads was accumulated at the peak region was identified for the 100 intergenic regions for which a given gene could be assigned (filled bars) and for the 52 divergent promoters (empty bars). b The ratio of the number of reads at the peak of each intergenic region at a given time and at time zero was calculated (fold-change) and the resulting values (expressed as log2) clustered using Cluster 3.0 (Euclidean distance, complete linkage) and represented with Java Treeview v1.1.6. The positions in the cluster of promoters corresponding to genes known to respond to calcineurin/Crz1 activation or for which evidence for in vitro or in vivo binding of Crz1 exists are represented at the right. c The time-course distribution of the number of reads at the Crz1 peak of recruitment is shown for diverse promoters of genes involved in glucose transport and phosphorylation

Fig. 2

Influence of the calcineurin pathway in the transcriptional response of diverse genes recruiting Crz1 upon high pH stress. The expression levels of the indicated genes in the calcineurin-deficient cnb1 strain (open bars) and the crz1 mutant (closed bars) are plotted as fold-change vs the wild type strain. For PHO89 the values are the mean from references [2, 4, 18]; for ENA1, from references [2, 4, 23, 39]; and for HXT2 and ALD6 from references [4, 17]. For the rest of the genes, data is taken from reference [4]. na, not available

Fig. 3

Gene ontology analysis of categorized gene promoters. a The 140 gene promoters selected (see text) were classified in “Strong” (>10-fold increase over the genome average reads), “Moderate” (5 to 10-fold increase), and “Weak” (2 to 5-fold change). b Each group of genes was subjected to Gene Ontology analysis and terms with p-value <0.01 were selected. Only the highest scores are presented in the Table. NSOT, No Significant Ontology Term can be assigned to this gene subset. c Genes were grouped according to the potency of Crz1 recruitment and the corresponding transcriptional data in response to alkaline pH stress (fold-change) were extracted from five independent datasets and presented as box plots for each dataset. Because the large dispersion of higher values and to facilitate comparison between plots the upper whisker is not shown fully in some graphs. Datasets used are: Casado et al. [34]; Casamayor et al. [35]; Causton et al. [1]; Serrano et al. [33]; and Viladevall et al. [4]. Data correspond in all cases to the response after 10 min of exposure to pH 8.0, except in reference [33], which correspond to 15 min to pH 8.2. Only genes with valid data in at least 2 datasets were used for the plots

Fig. 4

Positioning of Crz1 in promoters with strong (blue line), moderate (green line) or weak (red line) response. The fold-change with respect time 0 for each 50 bp bin of the region spanning 1 kb from the reported TSS were normalized as percentages of the fold-change in this region, aligned by position with respect to the TSS and the average for each 50 bp bin was calculated and plotted

Fig. 5

Identification of Crz1 binding motifs in vivo. a Web logo and PWM obtained for the putative CDREs of all (131) selected intergenic regions showing Crz1 binding (right) in comparison with the reported in vitro consensus [29] (left). b Web logo and PWM obtained for the putative CDREs of strong, moderate and weak promoters subsets (see Methods section)

Fig. 6

Chromatin accessibility and kinetics of Crz1 binding. Black bars correspond to the percentage of promoters where the putative CDRE was found to lie at NFRs. Empty bars and gray bars correspond to the percentage of promoters with enrichment in H3K4me3 and H3K4ac at the Crz1 recruitment peak, respectively. The data for nucleosome positioning and core histone modifications were taken from [31, 32], respectively. “Early response” includes those genes peaking at 1 or 2 min after stress, whereas “Late response” comprises those peaking at 5 or 20 min. Significance of the differences was determined using the Fisher’s exact test. *, p = 0.020; **, p = 0.006; ns, not significant (p = 0.391)