Using Kinetic Network Models to Understand Folding Mechanisms of GB1 Hairpin and its Trpzip Variants

Abstract

We used Markov State Model (MSM) approaches to analyze over 9 ms of explicit-solvent simulation trajectories for GB1 hairpin (the 16-residue C-terminal domain of protein G) and its three mutants (trpzip4, trpzip5, and trpzip6) at multiple temperatures, to investigate folding thermodynamics, kinetics, and mechanisms. Our MSM results show predicted folding rates and equilibrium populations that agree well with experimental data. Furthermore, we show how MSMs constructed from combined datasets reveal mechanistic differences resulting from tryptophan mutations. While wild type and trpzip4 hairpins are predicted to be two-state folders, our MSMs predict more complicated kinetics for trpzip5 and trpzip6 due to the presence of non-native traps. We find that changes in the folding landscape can also be revealed by analyzing MSM rate perturbations, provided that metastable states are conserved.