Loading Events
  • This event has passed.

University of Minnesota Seminars–Multiscale modelling approaches for probing the formation of biomolecular condensates via liquid–liquid phase separation

Jerelle A. Joseph

Cavendish Laboratory, Physics Department
University of Cambridge

When we consider intracellular organisation, we often picture different regions of the cell forming tiny membrane-bound compartments, like ribosomes and mitochondria. However, there are several important micro-environments inside the cell that are not enclosed by membranes. For decades, scientists have wondered how do these “membrane-less” organelles arise, and how are their structures maintained without physical membranes? Over the last two decades, ground-breaking experiments have proposed the transformative paradigm of liquid-liquid phase separation (LLPS), which suggests that the physical chemistry of phase-separation of multicomponent mixtures sustains such micro-environments: Analogous to the separation of oil and water into distinct liquid phases, macromolecules in the cytoplasm and nucleoplasm exhibit weak attractive interactions with each other that drive them to condense and then undergo LLPS, forming “oil-like droplets” inside the cells (formally termed biomolecular condensates). The formation of biomolecular condensates via LLPS provides a mechanism for spatiotemporal control of vital cellular processes, including RNA processing and stress signalling. Furthermore, aberrant LLPS has been implicated in several age-related disorders. Hence, elucidating the precise molecular interactions that sustain LLPS, as well as the physical determinants governing the composition, structural, and kinetic properties of biomolecular condensates, is now an active area of research. Resolving the behaviour of biomolecules inside liquid droplets and describing the phenomenon leading to their condensation is challenging both experimentally and computationally. Impressive advances in experimental techniques, particularly microfluidic analysis methods and single-molecule fluorescence microscopy, hold great potential for characterising biomolecular condensates and visualising molecules within liquid droplets. The availability of such techniques necessitates the development of computational models to determine the underlying physical mechanisms leading to the observed structural and dynamical features. From a modelling point of view, the computational costs of simulating large numbers of biomolecules for the long timescales involved in LLPS would be prohibitively expensive. In this talk, I will discuss (1) the development of a minimal coarse-grained model that allows for simulating thousands of interacting biomolecules in an efficient manner and reveals fundamental connections between the stability and composition of multicomponent condensates and critical points of pure systems. (2) I will then describe the use of atomistic potential of mean force calculations and residue-level coarse-grained simulations to shed light on the molecular grammar underlying LLPS and account for re-entrant phase transitions observed experimentally. (3) Finally, I will describe a novel multiscale strategy that probes atomistic condensates from pre-equilibrated coarse-grained ones and rationalizes trends in stability against fusion of condensates—quantified experimentally via zeta potential measurements at single-droplet resolution.

Seminars are open to alumni, friends of the Department, and the general public.