Multi-scale computational investigation of functions and mechanisms of protein-RNA phase separation.

PROJECT SUMMARY/ABSTRACT Biomolecular condensates formed through phase-separation of proteins and nucleic acids are integral compo- nents of eukaryotic cells, playing key regulatory roles in various cellular processes. Recent studies have revealed a broad spectrum of dynamic behaviors in condensates, likening them to elastic gels, viscoelastic polymers, or soft, glassy materials. The dynamic and material properties of condensates impact the timing of numerous es- sential cellular processes, including the diffusion and retention of regulatory molecules, signaling, and catalytic reactions. Furthermore, the material properties of condensates change over time during aging, which can lead to irreversible transitions into solid-like pathological states such as amyloids. Although it has been shown that material properties in cells can be regulated and tuned through sequence and composition, the molecular rules governing the material properties of condensates remain poorly understood. Furthermore, there is a lack of mechanistic models that can elucidate dynamic regulatory relationships between condensate material properties and biochemical processes inside them. This proposal aims to decipher the molecular sequence grammar of condensate material properties and their impact on the diffusion and partitioning of regulatory molecules, irre- versible aging, and active biochemical reactions. By combining multi-scale computational techniques, we will create microscopically detailed models of viscoelastic properties of condensates as a function of sequence pat- terning, solvent and environmental conditions, and biomolecular composition. Subsequently, we will dissect the impact of foldable and rigid elements of biomolecules by investigating transcription factor-DNA and nucleosomal condensates with varying folding topologies and disordered linker sequences. Our detailed molecular simulations will ultimately be used to deploy physics-based machine-learning tools, enabling us to predict condensate mate- rial properties based on composition, sequence, and environmental conditions. In another proposed direction, we will utilize non-equilibrium models developed in our group to simulate condensate aging and explain the formation of pathological solid-like states with molecular-level detail. Lastly, we will use non-equilibrium particle and phase- field-based models to investigate dynamic regulatory feedback between material characteristics with active bio- chemical processes such as microtubule assembly and enzymatic reactions. Completing the proposed research program will establish sequence-material properties-function relationships, providing far-reaching insights into condensates' biological functions. Studying condensate material properties will also provide invaluable insights into human health because of the close connection of material properties to the propensity of condensate to age and form pathological states implicated in numerous neurodegenerative diseases. the acquired fundamental biophysical knowledge will enable the targeting of material properties of condensates, halting pathways toward pathological states of condensates.