Increasing the Accuracy in All-Atom Simulations of Intrinsically Disordered Proteins based on the Absinth Model

Abstract

Sequence-specific conformational preferences of intrinsically disordered proteins (IDPs) are governed by the overall charge content, the linear patterning of oppositely charged residues as well as proline and charged residues. Physical insights generated using the ABSINTH implicit solvation model and forcefield paradigm assume fixed charge states of +1e for Lys and Arg, −1e for Asp, and Glu, and electroneutrality for His. However, the sequences of IDPs and putative IDPs encode rich possibilities for alterations in sequence-to-conformation relationships via charge regulation. This refers to protonation / deprotonation of titratable groups and / or the addition or deletion of charges via post-translational modifications. We have developed a general strategy to capture the effects of charge regulation, specifically pKa shifts, within the ABSINTH framework. This is achieved using a generalized thermodynamic integration approach. The effects of pKa shifts on conformational equilibria and reversible order-to-disorder transitions are calibrated using data from spectroscopic and potentiometric measurements. Additionally, we have uncovered the context dependence of reference free energies of solvation to account for non-additive contributions that are missing when we coopt data for free energies of solvation of model compounds. We are also refining the original forcefield parameters utilizing experimental results on a diversity of IDPs. These refinements are performed using an entropy maximization procedure that optimizes the agreement between simulated and experimentally derived conformational ensembles. Taken together, the generalizations contribute to improved accuracy in simulations of IDPs using the ABSINTH model.