First-Principles-Based Force Field for the Interaction of Proteins with Au(100)(5 x 1): An Extension of GoIP-CHARMM

Noncovalent recognition between peptides and inorganic materials is an established phenomenon. Key to exploiting these interactions in a wide range of materials selfassembly applications would be to harness the facet-selective control of peptide binding onto these materials. Fundamental understanding of what drives facet-selectivity in peptide binding is developing, but as yet is not sufficient to enable design of predictable facet-specific sequences. Computational simulation of the aqueous peptide gold interface, commonly used to understand the mechanisms driving adsorption at an atomic level, has thus far neglected the role that surface reconstruction might play in facet specificity. Here the polarizable Go1PC1-IARMM suite of force fields is extended to include the reconstructed Au(100) surface. The force field, compatible with the bio-organic force field CHARMM, is parametrized using first-principles data. Our extended force field is tailored to reproduce the heterogeneity of weak chemisorbing N and S species to specific locations in the Au(100)(5 x 1) surface identified from the firstprinciples calculations. We apply our new model to predict and compare the three-dimensional structure of liquid water at Au(111), Au(100)(1 x 1), and Au(100)(S x 1) interfaces. Using molecular dynamics simulations, we predict an increased likelihood for water-mediated peptide adsorption at the aqueous Au(100)(1 X 1) interface compared with the Au(100)(5 x 1) interface. Therefore, our findings suggest that peptide binding can discriminate between the native and reconstructed Au(100) interfaces and that the role of reconstruction on binding at the Au(100) interface should not be neglected.