(497a) Detailed Molecular Models of Interfacial Proteins from Sum Frequency Generation Spectroscopy

A wealth of structural information exists for solution- and crystal-state proteins due to the wide application of NMR and x-ray crystallography. Unfortunately, these techniques are unable to specifically probe interface-bound protein structures. Thus, other strategies must be applied to obtain information about a protein’s interfacial conformational ensemble. Sum frequency generation (SFG) spectroscopy is one such experimental technique that probes interfaces exclusively while ignoring material in the bulk. Much like the related techniques of IR and Raman spectroscopy, SFG in its most common form provides secondary structure data as a spectral line, but it lacks atomic-level structural detail. Molecular dynamics (MD) simulations of hypothetical interfacial proteins can provide the detail that is missing in SFG experiments. Theoretical SFG spectra can then calculated from MD structures and compared directly to experiment. Because an interface can greatly affect a protein’s structure, special care must be taken to consider timescales of protein configurational changes when simulating an adsorption process using MD. In this work, a theoretically rigorous method based on exhaustive sampling using metadynamics has been developed for simulating proteins at interfaces. This method overcomes the common problem of simulation timescales that are insufficient to sample an interfacial protein’s full conformational ensemble. Clusters of conformations from MD simulations are visualized and assigned probabilities based upon simulation results. Best practices for selecting protein and water force fields will be discussed. Theoretical results are directly compared to four experimental model peptide systems containing small secondary structure elements. The diversity of theoretical spectra from samples within a single MD simulation and from a survey of the Protein Data Bank will be presented to demonstrate the unique ability of SFG coupled with MD simulation and theory to provide molecular-level insight into protein-interface interactions.