(39f) Predicting Hydrogen-Deuterium Exchange Rates in Proteins Using Molecular Dynamics Simulations

Hydrogen-deuterium exchange (HDX) is a novel experimental method that uses exchange of hydrogen for deuterium as a probe for studying dynamics in proteins because certain hydrogen atoms in proteins, especially in highly flexible regions, are more prone to exchange by deuterium in solution than those in rigid structural motifs. Although HDX combined with mass spectrometry (HDX-MS) analysis can provide a resolution of deuterium incorporation at the level of peptide fragments, there is no direct route to obtaining residue-level exchange rates and hence their contribution to overall flexibility. Among many modeling approaches to predicting such residue-level exchange rates, molecular dynamics (MD) simulations are emerging as a highly useful tool. However, a large number of studies in the literature have used significantly variable criteria (solvent-accessible surface area, number of water molecules near proteins, hydrogen-bonds in the vicinity of water molecules or polar atoms) for predicting exchange rates and no generally applicable criteria have been discovered yet. In this work, we present a comprehensive analysis on prediction of HDX rates in three closely-related signaling proteins by carrying out multiple microsecond-long classical MD simulations. Specifically, we discover novel correlations between known criteria and the protection factor of each amino-acid by optimizing parameters using experimentally measured rates. These methods reveal dramatically different patterns of flexibility in these proteins that cannot inferred by their sequences or structures alone and suggest new ways of targeting them using small molecules. Our results have wide implications for studies on dynamics of other proteins, especially when unique modes of flexibility can be exploited to gain control of protein function via direct or allosteric inhibition.