Micro and macro models of the sweet receptor.

The study of the structure–activity relationship (SAR) of sweet molecules has been traditionally based on indirect analyses that have led to several models of the receptor active site, consistent with the shape of many conformationally rigid sweeteners—see, for example, Temussi et al. (1991) and the references therein. Most sweeteners are low mol. wt compounds but a few sweet proteins are also known. Only recently it has been demonstrated that small mol. wt sweeteners and sweet macromolecules interact with the same T1R2–T1R3 receptor (Li et al., 2002), inferring that there is just one sweet taste receptor. However, it is not easy to understand (and show) how low mol. wt sweet compounds and sweet proteins can activate the same receptor. The three-dimensional structure of the receptor has not yet been elucidated, but it is possible to build homology models to shed light on the SAR of sweet molecules. The sweet taste receptor (SR) is a G protein coupled receptor similar to the dimeric metabotropic glutamate m1-LBR receptor (Margolskee, 2002). The similarity between the sequences of the two chains of the T1R2–T1R3 receptor and that of the single chain of the homodimer of the m1-LBR mGlu receptor is sufficient to allow homology modelling and to assume that it has the same general features. Existing three-dimensional SR models (Margolskee, 2002; Temussi, 2002) have been built using, as template, the N-terminal domain of mGluR1 whose crystal structure is available (Kunishima et al., 2000).