(770b) Using Parallel Bias Metadynamics to Efficiently Explore Peptide Folding over a Range of Cytostolic Salinities
Protein folding in vivo is greatly influenced by cytostolic conditions which affect folding kinetics and observed stable conformations. To determine the effects of ion species and concentration on protein structures, we have conducted a set of enhanced molecular dynamics (MD) simulations using a recently developed method known as parallel bias metadynamics (PBMetaD). Metadynamics explores protein conformations in a systematic way, allowing us to calculate free energy landscapes of protein folding. Within MD simulations, many ion models have been parameterized to reproduce the thermodynamics of ion solvation. Early parameterization of monovalent alkali and halogen ions for use in biophysical simulations exhibited salt crystallization at concentrations well below solubility limits. Improved (Joung) ion parameterization has greatly improved this prior shortcoming. However, little attention has yet been paid to ion-protein interactions. We have explored the folding landscape of two model peptides, namely GB1 and trp-cage in 0, 150, 300 and 1000 mM solutions of NaCl and KCl. Two protein models were also tested, namely CHARMM36 and AMBER99SB*-ILDN. Simulations demonstrate that the folding free energy of these model peptides is greatly affected by force selection and ion concentration. Ion-protein interactions are greatly affected by the solution's ionic strength and model parameters. Based upon our observations, we recommend a set of parameters under which protein simulations should be conducted. We foresee applications of this work in the study of proteins and enzymes in hypertonic and hypotonic solutions as well as in the design of solutions to stabilize specific protein conformational states.