The gustatory system of mammals can sense four basic taste qualities, bitter, sweet, salty and sour, as well as umami, the taste of glutamate. Previous studies suggested that the detection of bitter and sweet tastants by taste receptor cells in the mouth is likely to involve G-protein-coupled receptors. Although two putative G-protein-coupled bitter/sweet taste receptors have been identified, the chemical diversity of bitter and sweet compounds leads one to expect that there is a larger number of different receptors. Here we report the identification of a family of candidate taste receptors (the TRBs) that are members of the G-protein-coupled receptor superfamily and that are specifically expressed by taste receptor cells. A cluster of genes encoding human TRBs is located adjacent to a Prp gene locus, which in mouse is tightly linked to the SOA genetic locus that is involved in detecting the bitter compound sucrose octaacetate. Another TRB gene is found on a human contig assigned to chromosome 5p15, the location of a genetic locus (PROP) that controls the detection of the bitter compound 6-n-propyl-2-thiouracil in humans.

The mechanisms underlying sweet taste in mammals have been elusive. Although numerous studies have implicated G proteins in sweet taste detection, the expected G protein-coupled receptors have not been found. Here we describe a candidate taste receptor gene, T1r3, that is located at or near the mouse Sac locus, a genetic locus that controls the detection of certain sweet tastants. T1R3 differs in amino acid sequence in mouse strains with different Sac phenotypes ('tasters' versus 'nontasters'). In addition, a perfect correlation exists between two different T1r3 alleles and Sac phenotypes in recombinant inbred mouse strains. The T1r3 gene is expressed in a subset of taste cells in circumvallate, foliate and fungiform taste papillae. In circumvallate and foliate papillae, most T1r3-expressing cells also express a gene encoding a related receptor, T1R2, raising the possibility that these cells recognize more than one ligand, or that the two receptors function as heterodimers.

The sense of taste informs the body about the quality of ingested foods. Tastant-mediated signals are generated by a rise in free intracellular calcium levels ([Ca(2+)]i) in the taste bud cells and then are transferred to the gustatory area of brain via connections between the gustatory nerves (chorda tympani and glossopharyngeal nerves) and the nucleus of solitary tract in the brain stem. We have recently shown that lingual CD36 contributes to fat preference and early digestive secretions in the mouse. We show here that 1) the induction of an increase in [Ca(2+)]i by linoleic acid is CD36-dependent in taste receptor cells, 2) the spontaneous preference for or conversely conditioned aversion to linoleic acid requires intact gustatory nerves, and 3) the activation of gustatory neurons in the nucleus of the solitary tract elicited by a linoleic acid deposition on the tongue in wild-type mice cannot be reproduced in CD36-null animals. We conclude that the CD36-mediated perception of long-chain fatty acids involves the gustatory pathway, suggesting that the mouse may have a "taste" for fatty foods. This system would constitute a potential physiological advantage under conditions of food scarcity by leading the mouse to select and absorb fatty foods. However, it might also lead to a risk of obesity and associated diseases in a context of constantly abundant food.