Biological condensates enriched with intrinsically disordered proteins (IDPs) are crucial in regulating cell functions and disease. Accurate modeling of IDPs is challenging due to their conformational heterogeneity and stochastic nature. We developed a multi-scale computational model to simulate the phase separation behavior of IDPs mimicking Polymer-oligopeptide Hybrid (IPH) synthesized in our lab by grafting the interaction motifs into a dextran backbone as a minimalistic mimic for biological condensate. We simulated two types of IPH, one can undergo liquid-liquid phase separation (LLPS), named IPH*, and the other can undergo liquid-solid phase separation (LSPS), named IPH^Aβ, under physiological conditions in-vitro. They were synthesized, getting the interaction motifs from FUS protein and amyloid-β peptide, respectively.
We used all atomistic molecular dynamics (MD) simulations followed by lattice Monte Carlo (MC) simulations to demonstrate dynamic assembly in LLPS and irreversible aggregation in LSPS of IPHs. We considered an IPH to have a sticker-spacer architecture, leveraging on classical associative polymer theory. We employed MD simulations to estimate the driving forces from the interactive motifs ( as stickers). The favorable enthalpy between stickers provided the driving force for reversible binding between IPHs. We estimated the average binding energy between a pair of stickers as 102 kJ/mol and 256 kJ/mol for IPH* and IPH^Aβ using Molecular Mechanics Poisson-Boltzmann Surface Area (MMPBSA) method, respectively.
Next, we utilized MC simulations to describe the stochastic movement of IPHs under physiological conditions. IPHs are structured as coarse-grained beads of stickers and spacers and randomly placed in a three-dimensional simple cubic lattice as a canonical ensemble. Using the Metropolis-Hastings algorithm and discretized sticker energy distributions, we performed 10⁶ orders of hypothetical MC moves as self-avoiding walks on the lattice to obtain a quasi-equilibrium state. We introduced the assembly order parameter (AOP) as the ratio of the consecutive radial distribution function to estimate each system's dynamics and phase separation propensity. The results provide insights into the propensity of two IPHs to phase separately differently in liquid-like behavior and solid-like aggregation. We provide cost-effective in-silico farmwork to predict the phase separation behavior of biological condensate-inspired engineered bio-mimetics.

Mr. Kalindu Shyamawarna Fernando

Ph.D. candidate, Hong Kong University of Science &amp; Technology

Appearances

Ph.D. candidate, Hong Kong University of Science & Technology