Novel high-throughput approaches to discover and dissect human transcriptional protein interactions

PROJECT SUMMARY/ABSTRACT Understanding transcription factors in human cells is essential for insights into development, health, and disease. However, our understanding of the protein-protein interactions driving transcriptional regulation remains incomplete. While high-throughput screens have identified hundreds of effector domains within transcription factors, their underlying molecular mechanisms remain largely unknown. We lack a comprehensive molecular understanding of which of the >1000 effector domains interact with which of the >700 cofactors, the specificity of these interactions, and how this specificity quantitatively influences gene regulation. This information is crucial for understanding basic principles of transcription and could also stimulate development of novel therapeutics capable of targeting these biologically essential but historically undruggable interactions. Current high-throughput interaction screens, capable of measuring many-by-many candidate interactors, employ transcription as a readout (e.g. two-hybrid approaches), making them ill-suited for studying transcription- associated networks. Furthermore, most methods for generating quantitative measurements of direct interactions (e.g. Kds, kons, koffs) are low-throughput and labor-intensive, involving extensive cloning, purification, and testing, which can take months if not years, severely limiting our ability to quantify the direct connections within these extensive protein-protein interaction networks. As a result, we have a very limited picture of how these interactions contribute to transcription. Here we will overcome these challenges by leveraging our expertise in development of high-throughput approaches, biophysics, and quantitative cell biology to first, develop a novel cell-based technology to generate a comprehensive matrix of 106 proximities between human transcriptional effector domains and the cofactors they recruit to regulate transcription (Aim 1). Next, we will determine the degree of specificity between these interactions using a high-throughput microfluidic technology we recently developed to quantitively measure the strength and timing of direct interactions (e.g. kon, koff, Kd) (Aim 2). Last, we will move beyond simple explanations of our observations towards predictive models to quantify how the specificity of interactions determines the strength and timing of gene regulation (Aim 3). The proposed research will generate a novel technology for large-scale mapping of many-by-many protein interactions with a wide dynamic range that will have immediate broad application to other biological systems beyond transcription (e.g. cell signaling networks). In addition to this technological innovation, by mapping the proximity network between effector domains and cofactors, quantifying their direct interactions, and linking these interaction specificities, strengths, and kinetics to gene regulatory dynamics, we will reveal new principles of transcription factor-mediated gene regulation.