Dissecting the contributions of phase separation to yeast transcription .

Summary Transcription requires the coordinated recruitment of transcription factors, coactivators, and regulators to specific DNA elements with high spatial and temporal regulation. Phase separation has recently been implicated in the regulated assembly of the transcriptional machinery, but it remains unclear to which extent and how phase

separation contributes to transcriptional activity. This is a bottleneck in our current understanding given the importance of this fundamental biological process in development and its dysregulation in disease processes. The small size of transcriptional assemblies is a technical challenge that we will circumvent here by employing

mutagenesis. Another challenge is the limited quantitative comparison of alternative models in the current literature. Importantly, the existence of multivalent interactions does not necessitate that they give rise to function via phase separation. E.g., multiple motifs in the yeast transcription factor Gcn4 bind to multiple binding sites in

its cognate mediator subunit Med15, generating multivalent, dynamic interactions that give rise to non- stoichiometric, soluble complexes below the threshold concentration for phase separation. Thus, we hypothesize that soluble complexes and condensates are mediated by the same multivalent interactions and serve similar

functions. We further hypothesize that transcriptional condensates have additional emerging properties that can enhance or suppress transcriptional functions. Here, we propose to leverage a conceptual phase separation framework recently developed by PI Mittag to generate separation-of-function mutants. Her rigorous approaches

for quantifying driving force for phase separation will be combined with extensive expertise on transcription and its regulation by PI Ansari. We will perform the proposed studies in yeast because function can be analyzed in vivo, and activity compared with well-defined in vitro readouts. We will ask three main questions: 1) How do

transcription factors, coactivators and DNA phase separate together, and how do multivalent interactions and solubility contribute to their phase behavior? 2) Is transcriptional activity of the yeast transcription factor Gcn4 driven by small complexes or by phase-separated condensates? 3) Given that condensates have emergent

properties that small complexes lack, how does the ability to include or exclude regulatory factors from condensates and the ability to percolate and alter chromatin structure influence transcriptional function? The overall impact of this project will stem from the careful quantitative and conceptual characterization of the phase

behavior of transcription factors with downstream factors. The rigorous comparison of alternate models will reveal the role of phase separation in transcription and which functions of transcriptional condensates go beyond those of small complexes. The proposed work is of high relevance given the interest in targeting transcriptional

condensates therapeutically in diseases.