(62f) Nuclear Organization and Genome Regulation By RNA-Dependent Phase Separation

Precise gene expression is essential for cell-type specification, development, and physiology. In multi-cellular organisms such as mammals, this regulation is dependent on the coordinated organization of many DNA, RNA, and protein species, which are often present at far distances from each other in a crowded nuclear milieu. Recently, phase separation has emerged as a key mechanism contributing to organization of multiple nuclear compartments such as the nucleolus, transcriptional condensates, and nuclear speckles. The interplay between these nuclear compartments that form by phase separation, gene activity, and newly synthesized RNA is not well understood. To investigate this interplay, we develop computer simulations based on statistical physics models, which are tested and refined in collaboration with experiments in vitro and in cells. We first find that different RNA species and transcriptional components undergo phase separation in a manner consistent with complex coacervation of charged poly-ions. We identify that diffusion is limiting under cellular conditions, and thus synthesis of RNA can either assemble (low rates, such as at non-coding or regulatory elements) or dissolve (high rates, such as during productive gene expression) phase-separated compartments. We argue that this provides a model of dual feedback by RNA levels and may contribute to self-limited regulation of gene expression. Model predictions that connect the effect of perturbations in RNA synthesis to variations in condensate size and lifetime are verified by through live-cell imaging as well as gene-reporter assays in cells. We then extend the model to investigate how spatially localized gene activity (for e.g. through clustering on the genome) influences condensate properties. Through extensive simulations, we identify a unified framework that describes diverse phenomena in nuclear organization that often lack direct mechanistic bases - including condensate flow, nucleation by RNA synthesis, nucleolar vacuoles, and aspherical morphologies. Overall, our model implicates gene activity (RNA synthesis) as a potent regulatory axis of gene expression and nuclear organization.