The Taste of Molecules: Bitter-Sweet Discoveries

Taste is a major driver for food choice and consumption. The attractive sweet and umami taste modalities are thought to signal presence of nutritious calorie- and protein- rich foods respectively, while the aversive bitter taste is thought to protect against consumption of poisonous food. These basic taste modalities are mediated by G-protein coupled receptors (GPCRs), which are expressed also in extra-oral tissues, including heart, lungs and gut and are likely to have endogenous ligands and physiological roles beyond food quality assessment.

Several approaches to identifying novel bitter and sweet compounds, including ligand-based, structure-based and machine learning techniques will be presented. Specifically, we developed a machine-learning (decision trees-based) tool BitterPredict that correctly classifies 80% of the compounds in the hold-out test set and in three independent external sets, into bitter and non-bitter compounds¹. We showed that only 60% of the toxic compounds are known or predicted to be bitter. This is similar to the predicted abundance of bitter molecules among FDA-approved drugs and lower than in natural compounds. Our results suggest existence of many non-bitter toxic compounds, as well as negligible toxicity of many bitter molecules².

Indeed, bitterness is a common off-flavor of non-caloric sweeteners. Computer-aided sweeteners discovery will be discussed in view of current challenges in food and nutrition.

(1) Dagan-Wiener, A.; Nissim, I.; Abu, N. B.; Borgonovo, G.; Bassoli, A.; Niv, M. Y. Bitter or not? BitterPredict, a tool for predicting taste from chemical structure. Scientific Reports 2017.

(2) Nissim, I.; Dagan-Wiener, A.; Niv, M. Y. The taste of toxicity: A quantitative analysis of bitter and toxic molecules. IUBMB Life, in press.