An accurate density functional theory based estimation of pKa values of polar residues combined with experimental data: from amino acids to minimal proteins

Abstract

We report a scheme for estimating the acid dissociation constant (pK_a) based on quantum-chemical calculations combined with a polarizable continuum model, where a parameter is determined for small reference molecules. We calculated the pK_a values of variously sized molecules ranging from an amino acid to a protein consisting of 300 atoms. This scheme enabled us to derive a semiquantitative pK_a value of specific chemical groups and discuss the influence of the surroundings on the pK_a values. As applications, we have derived the pK_a value of the side chain of an amino acid and almost reproduced the experimental value. By using our computing schemes, we showed the influence of hydrogen bonds on the pK_a values in the case of tripeptides, which decreases the pK_a value by 3.0 units for serine in comparison with those of the corresponding monopeptides. Finally, with some assumptions, we derived the pK_a values of tyrosines and serines in chignolin and a tryptophan cage. We obtained quite different pK_a values of adjacent serines in the tryptophan cage; the pK_a value of the OH group of Ser13 exposed to bulk water is 14.69, whereas that of Ser14 not exposed to bulk water is 20.80 because of the internal hydrogen bonds.