Amino acids are essential for life: they form the building blocks of the proteins that our genetic code encrypts. And thanks to nature's ingenuity, mammals have evolved a partiality for the flavour of these crucial components; some taste bitter, others sweet and some simply delicious, or 'umami' — a taste that is associated in Western countries with Chinese takeaway meals. But what are the molecular players that underlie this perceptual diversity?

In this issue (Nature 416, 199–202; 2002), Greg Nelson and colleagues begin to solve the conundrum. They have characterized a mammalian amino-acid taste receptor and find that, surprisingly, this one receptor responds to several of the 20 l-amino acids that are commonly found in proteins. In another clever move by nature, the receptor is entirely unresponsive to their mirror images — the biologically less significant d-amino acids.

The amino-acid receptor is located on the surface of taste cells and is coupled to well-known molecular switches called G proteins. In fact, it consists of two different, taste-specific G-protein-coupled receptors (GPCRs), T1R1 and T1R3. Nelson et al. developed a cell-based assay to identify candidate taste receptors, and found that cells expressing both T1R1 and T1R3, together with a 'promiscuous' G protein that causes the release of calcium from internal cellular stores, responded to the application of different l-amino acids with an increase in intracellular calcium concentration. Moreover, a substance that enhances the tongue's response to l-amino acids — inosine monophosphate — increased the response of the T1R1+3 receptor in vitro and in vivo.

Interestingly, when T1R3 is combined with a different taste-specific GPCR protein called T1R2, it forms a functional 'sweet receptor' (G. Nelson et al. Cell 106, 381–390; 2000). The authors now show that this receptor is activated by a range of sweet-tasting d-amino acids. So, depending on the identity of its partner protein, T1R3 may contribute to the sensation of different tastes.

While carrying out this work, Nelson et al. also found a molecular explanation for the taste preferences of different species. For example, the human T1R2+3 sweet receptor is much more responsive than its rodent equivalent to artificial sweeteners such as aspartame and cyclamate. And the human T1R1+3 amino-acid receptor is more sensitive to monosodium glutamate than to any of the amino acids. This suggests that activation of the human T1R1+3 receptor gives rise to the sensation of umami, perhaps in concert with another candidate receptor for monosodium glutamate that was described two years ago by N. Chaudhari and colleagues (Nature Neurosci. 3, 113–119; 2000).

But the mystery of how different amino acids can activate just one receptor, yet give rise to a whole spectrum of taste sensations, remains to be solved. If our sense of smell is any precedent, the answer will emerge not only by working out which tastes are detected by which receptors, but also by understanding the complex network of nerves that sends these signals to the brain.