Evolution of Reduced Co-Activator Dependence Led to Target Expansion of a Starvation Response Pathway [Cgla ChIP-seq]

In S. cerevisiae, the phosphate starvation (PHO) responsive transcription factors Pho4 and Pho2 are jointly required for induction of phosphate response genes and survival in phosphate starvation conditions. In the related human commensal and pathogen C. glabrata, Pho4 is required but Pho2 is dispensable for survival in phosphate-limited conditions and is only partially required for inducing the phosphate response genes. This reduced dependence on Pho2 evolved in C. glabrata and closely related species. Pho4 orthologs that are less dependent on Pho2 induce more genes when introduced into the S. cerevisiae background, and Pho4 in C. glabrata both binds to more sites and induces more genes with expanded functional roles compared to Pho4 in S. cerevisiae. We used Chromatin-ImmunoPrecipitation with exonucleas followed by high-throughput sequencing (BioChIP-seq) to identify the binding locations of Pho4 from both S. cerevisiae and C. glabrata in the S. cerevisiae background lacking the negative regulator Pho80, and either with or without Pho2. Overall design: We made C. glabrata strains carrying a C-terminally tagged (3x FLAG) version of the endogenous Pho4 either with or without the endogenous Pho2. We also made strains carrying an N-terminally tagged (V5) version of the endogenous Pho2. Mock strains were made to correspond to the experimental strain's genetic background, but lacking the epitope tag. All strains were grown to mid-log phase, washed and released into no phosphate media for 1 hour and harvested for ChIP analysis. We used ChIP-exo, a variation of the ChIP protocol using exonuclease to increase the resolution and signal-to-noise ratio, to query the genomic binding locations of both Pho4 and Pho2 in C. glabrata. A routinely used control for ChIP is the input chromatin for normalizing against heterogeneity in the fragmentation process. This, however, doesn't apply to the ChIP-exo technique. We use either hte mock sample instead for peak finding and generating the WIG files.