Molecular Dynamics Simulations of Liquid-Liquid Phase Separation in Biology

A newly appreciated mechanism by which biochemistry is organized in cells without membrane barriers is liquid-liquid phase separation (LLPS). RNA and protein condense into liquid droplets through LLPS and are crucial for normal physiological function, but they are also implicated in the development of several neurological diseases as well as theories about the origins of life and cells themselves. Since the means by which droplets form, maintain shape, and achieve molecular specificity are poorly understood, a major challenge is to understand how nanometer scale molecules coalesce to produce cellular bodies of diverse material properties and distinct identities. Our aim is to model the process and properties of LLPS involving mRNAs and disordered proteins using molecular dynamics (MD). Specifically, by using both low and high-resolution representations of the system, we can map sequence-specific effects onto a low-resolution setting. In the coarse-grained space, the solvent is modeled implicitly using Brownian dynamics and the molecules as bead-spring polymers. These techniques provide access to space and time-scales that are critical to the phase separated state which involves the collective behavior of thousands of molecules.