Estimation of atomic hydrophobicities using molecular dynamics simulation of peptides

Abstract

The hydrophobic force is one of the main driving forces in protein folding and binding. However, its nature is not yet well understood and consequently there are more than 80 different scales published trying to quantify it. Most of the hydrophobicity scales are amino acid-based, but the interaction between the molecular surface of the proteins (and DNA) and surfaces they are immobilized on, e.g., on biomedical micro/nanodevices, occurs on fractions of, rather than whole amino acids. This fragmented structure of the biomolecular surface requires the derivation of atom-level hydrophobicity. Most attempts for the evaluation of atomic hydrophobicities are derived from amino acid-based values, which ignore dynamic and steric factors. This contribution reports on the Molecular Dynamics simulations that aim to overcome this simplification. The calculations examine various tripeptides in an aqueous solution and the analysis focuses on the distance of the nearest water molecules to the individual atoms in the peptides. Different environments result in a variation of average distances for similar atoms in different tripeptides. Comparison with the atomic hydrophobicities derived from the amino acid-based hydrophobicity obtained from peptide partition in water-octanol (Dgoct) and transport through the membrane interface (Dgwif) shows a similar trend to the calculated distances. The variations are likely due to the steric differences of similar types of atoms in different geometric contexts. Therefore, Molecular Dynamics simulations proved convenient for the evaluation of atomic hydrophobicities and open new research avenues. The atomic hydrophobicities can be used to design surfaces that mimic the biomolecular surfaces and therefore elicit an expected biomolecular activity from the immobilized biomolecules.