Modeling of peptide adsorption interactions with a poly(lactic acid) surface.

The biocompatibility of implanted materials and devices is governed by the conformation, orientation, and composition of the layer of proteins that adsorb to the surface of the material immediately upon implantation, so an understanding of this adsorbed protein layer is essential to the rigorous and methodical design of implant materials. In this study, novel molecular dynamics techniques were employed in order to determine the change in free energy for the adsorption of a solvated nine-residue peptide (GGGG-K-GGGG) to a crystalline polylactide surface in an effort to elucidate the fundamental mechanisms that govern protein adsorption. This system, like many others, involves two distinct types of sampling problems: a spatial sampling problem, which arises due to entropic effects creating barriers in the free energy profile, and a conformational sampling problem, which occurs due to barriers in the potential energy landscape. In a two-step process that addresses each sampling problem in turn, the technique of biased replica exchange molecular dynamics was refined and applied in order to overcome these sampling problems and, using the information available at the atomic level of detail afforded by molecular simulation, both quantify and characterize the interactions between the peptide and a relevant biomaterial surface. The results from these simulations predict a fairly strong adsorption response with an adsorption free energy of -2.5 +/- 0.6 kcal/mol (mean +/- 95% confidence interval), with adsorption primarily due to hydrophobic interactions between the nonpolar groups of the peptide and the PLA surface. As part of a larger and ongoing effort involving both simulation and experimental investigations, this work contributes to the goal of transforming the engineering of biomaterials from one dominated by trial-and-error to one which is guided by an atomic-level understanding of the interactions that occur at the tissue-biomaterial interface.