Molecular simulation of biomaterials and biomolecules at the solid-liquid interface

Abstract

Biomaterials and biomineralization have been successfully utilized in a broad variety of technical applications. Properties of natural biopolymers, such as the ability to control the nucleation, growth, and organization of crystals, have been extended to a much wider array of technologically applicable materials through combinatorial selection techniques. However, detailed mechanisms of peptide adsorption on inorganic surfaces have largely escaped characterization. This knowledge would open new routes for the rational design of nanostructures and composite biomaterials. The development of accurate and computationally efficient methods for the simulation of biopolymer-inorganic surface adsorption could provide a more detailed understanding of adsorption mechanisms. While simple models involving reduced solvent representations and polymer flexibility have found some success in limited applications, robust performance for systems of varying size and composition can generally be expected only through accurate inclusion of these key details. Fully atomistic representations of biopolymer and surface are necessary for detailed molecular recognition, while polymer flexibility is required to capture structural rearrangement and the resulting free energy contributions. Finally, electrostatic interactions between the adsorbing biopolymer and inorganic surface, as well as interactions of the polymer and surface with the surrounding solvent environment will play a dominant role in the adsorption process, and an accurate representation of the solvated system is inherently necessary. Computational efficiency can be increased through the application of implicit solvent models, which replace the numerous solvent molecules with a continuum dielectric, and seek to capture the average effects of the statistical solvent environment. The Poisson-Boltzmann model represents the most rigorous treatment of implicit solvent.

(cont.) This model, however, requires the relatively expensive solution of a second order elliptical differential equation over the space of the system. Here, a method is presented which reduces the scale at which the Poisson-Boltzmann equation must be solved. However, even when combined with an efficient multi-grid solver, the Poisson-Boltzmann model represents a significant computational cost. The modified Generalized Born model, GBr6, based on an approximation to the Poisson-Boltzmann model, offers a computationally efficient alternative. Generalized Born models, however, are often inaccurate in the case of charges positioned near an extended dielectric interface, which is precisely the system we wish to investigate. Here, an analytical integration of the approximate electric displacement is used to calculate Born radii, and tested in application to surface adsorption studies. Replica-exchange Monte Carlo simulations with modified Generalized Born implicit solvent environment is then used to study the adsorption mechanism of a set of rationally designed sapphire-binding peptides. Modulation of binding affinity is predicted to depend on multiple interactions between basic amino acids and the negatively charged sapphire surface. The proximity of charged residues to one another as well as the conformational ability of each peptide to present functional groups towards the surface are shown to control the relative binding affinities.