Kinetics and mechanism of ionic-liquid induced protein unfolding: application to the model protein HP35

We demonstrate an approach to quantify protein unfolding times using molecular simulation in a greatly accelerated manner compared to standard MD simulations, showing up to 400 fold speed increases. The approach uses infrequent metadynamics, which has been shown to provide quantitative rates for rare events, accelerated by biasing the RMSD of the protein structure. The results are quantitatively verified against a large benchmark dataset using the model proteins chignolin and villin headpiece (HP35). Following this, we apply the algorithm to protein unfolding in ionic liquids and study the HP35 unfolding time in four different 20% (w/w) IL/water mixtures. An interesting agreement is obtained between the ordering of the anion effects and previously published experiments on anion-induced destabilization of ribonuclease A (RNase A). Additional simulations helped shed light on the molecular mechanisms that lead to accelerated unfolding in ILs that have a chaotropic or hydrophobic anion. Further simulations suggest that, in this case, the tendency of an IL to be structure forming or breaking (kosmotrope vs. chaotrope) in water is unrelated to the unfolding times. Instead, the chaotropic anions, which are also more hydrophobic, more readily bind residues from the hydrophobic core, leading to faster unfolding. This approach should be appropriate for a wide range of protein unfolding simulations in the future and aid in systematic discovery of molecular descriptors for biomolecule function in unique nanostructured environments.