Identification of transcription factor targets by phenotypic activation and microarray expression profiling in yeast.

A major obstacle to identify physiological transcriptional targets is that the conditions that induce the majority of yeast transcription factors (TFs) are unknown. Microarray analyses of deletion mutants indicate that most TFs are inactive under standard growth conditions. To overcome this, we screened an ordered array of yeast open reading frames (ORFs) to identify TFs that confer reduced fitness upon overexpression, suggesting that overexpression results in an activated state (phenotypic activation). Approximately one-third of all yeast TFs exhibited this phenotype. Here, we describe in detail our methodology to characterize these TF overexpression strains including microarray expression profiling, data analysis, and motif searching. Our analyses show that in many cases, the differentially regulated genes correspond to physiological functions and known targets of well-characterized TFs. The expected binding sites of several TFs were also identified in the promoters of these genes. Moreover, novel DNA-binding sequences and putative targets were identified for less-characterized TFs. These results demonstrate that phenotypic activation is an effective approach to rapidly characterize TFs on a large scale, which should also be feasible in other organisms.