Paper BI-TuP11
Molecular Dynamics Parameterization for Electrostatic Interactions between Proteins and Biomaterial Surfaces

Tuesday, October 16, 2007, 6:00 pm, Room 4C

Since protein-biomaterial interactions govern the biocompatibility of implanted materials, controlling biocompatibility through material design must begin with the study of protein-biomaterial interactions at the atomic level. All-atom molecular dynamics simulation provides an excellent approach to investigate this type of problem. However, current molecular dynamics simulation methods and parameters are generally not designed to accommodate the unique types of atomic interactions that exist for the case of a protein interacting with a functionalized surface. To address this problem, we have begun adapting the molecular modeling community’s range of tools to develop a set that is specifically designed for the simulation of the adsorption behavior of proteins to functionalized surfaces. Protein adsorption behavior is predominantly governed by nonbonded interactions, with electrostatic effects representing the strongest type of these interactions and the type that is most difficult to accurately represent. In an effort to establish the most appropriate method of treating electrostatic interactions for the simulation of adsorption processes, we are evaluating the calculated differences in ion distribution over a charged surface using a variety of nonbonded interaction techniques. Our 4.5 x 4.3 x 10.0 nm³ model system is comprised of a 150 mM NaCl aqueous solution with TIP3P water over a 50% deprotonated COOH-SAM surface (pK_a = 7.4) with Na⁺ counterions. Nanosecond-scale molecular dynamics simulations are then conducted to model the structure of the electric double layer over the surface using a series of different methods to represent the electrostatic interactions of the system, including particle-mesh Ewald, radial cutoffs, isotropic periodic sum, and anisotropic periodic sum methods. The results of these simulations are then compared to the analytical solution of the ion distribution based on the Poisson-Boltzmann equation to gauge the accuracy of each of the different simulation methods. Within each different method, the parameters defining the limits of atomic interactions, such as interaction cutoff distances in the case of radial cutoffs, have been varied to establish a balance between computational cost and simulation accuracy. The results of this study will establish the most efficient and accurate method for the representation of nonbonded electrostatic interactions for the simulation of protein-surface interactions.