General pseudoreceptor model for sweet compounds: a semiquantitative prediction of binding affinity for sweet-tasting molecules.

The chemical structures of sweet compounds are very different, ranging from sugars to amino acids and peptides or other compounds such as saccharin. The biological mechanism underlying the generation of sweet taste is still unknown, although in the past few years much research has provided evidence for the existence of a true chemoreception process, mediated by receptor proteins on the taste buds. In particular, the initial step of the process involves the reversible binding of the sweet compounds to their receptor(s). In this work, we have investigated this binding via a pseudoreceptor model, which has been developed using a training set of 24 compounds belonging to different families including sugars, peptides, and other intensive sweeteners. This model provided a correlation coefficient (r(2)) of 0.985 between the calculated and the experimental free energies of binding, which are related to the molar relative sweetness, for the training set and is able to predict semiquantitatively free energies of ligand binding for an independent set of five test ligand molecules within 0.3-2.1 kcal mol(-1) of the experimental values.