(169am) Evaluation of Effective Dielectric Constants: Implications for Modeling Protein-DNA Liquid-Liquid Phase Separation

Electrostatic interactions are a fundamental force governing the behavior of biological macromolecules such as proteins and nucleic acids. Therefore, accounting for electrostatic interactions correctly is crucial for understanding mechanisms of biological processes. In general, the existing CG protein/DNA models use the implicit solvent potential based on Debye-Hückel theory, where dielectric permittivity corresponds to water (D ≈ 80 at 300K). However, in the vicinity of the protein/DNA, this dielectric is a factor of ≈ 2 − 10 too large and thus protein-DNA interactions are underestimated. Further, protein-DNA interactions are known to be mediated by water, a phenomenon excluded from the models by the use of implicit solvent (i.e. Langevin dynamics). This suggests the need to account for spatial variation. Seeking to use the existing protein/DNA CG models with minimal modification, this research project aims to evaluate the impact of effective dielectric constants by incorporating a distance, temperature, or concentration-dependent dielectric constants and provide recommendations for the appropriate dielectric constants for the calculation of electrostatic interactions. Using prototypical protein-DNA systems that undergo liquid-liquid phase separation (LLPS), we access the validity of these models by comparing their predictive capability of reproducing local structure properties and binding affinities. Overall, this research holds promise and can be used to investigate protein−DNA complexes, enabling a mechanistic understanding of various biological and engineering applications.