Coarse-grained intrinsically disordered protein simulations with explicit ions parameterized by salting-out constants.

Abstract

Implicit solvent coarse-grained models have emerged as an effective method for simulating intrinsically disordered proteins (IDPs) due to the long length and time-scale need to effectively simulate their equilibrium behavior. These coarse-grained methods typically rely on Debye-Hückel theory to explain the effects of salt in the simulations. However, Debye-Hückel electrostatics are only accurate over a limited range of ionic strengths and are also unable to capture ion-type specific effects, typically called Hofmeister effects that arise at higher salt concentrations. To address this, we developed an explicit ion model of potassium chloride that can interface with the otherwise implicit solvent hydrophobicity scale model for intrinsically disordered protein. Parameters for the explicit ion model were determined by matching the salting-out constants of several small peptide-like molecules. Salting-out constants were determined in the simulation by creating a system with an ion concentration gradient via an external potential and performing umbrella sampling along the concentration gradient. The remaining amino acids for which there is no data for the salting-out constant are determined using a scaled Kapcha-Rosky hydrophobicity scale determined from the umbrella sampling simulations. The parameters are validated against a set FRET data from a diverse set of linker peptides at several salt concentrations. This explicit ion model is able to capture the changes in protein expansion as a function of salt concentration at concentration above 0.5 M where Debye-Hückel screened electrostatics predict no change in protein expansion.