On sunny days I often slip out of the lab and visit a nearby cafe, where I am greeted by the aroma of freshly brewed coffee and oven-baked pastries. Within a few minutes, however, these odours become less noticeable. They're not gone, but my ability to perceive them has somehow diminished. My typical order, a large coffee with a shot of espresso, no longer provides the nervous jolt it once did — an unfortunate consequence of my caffeine habit. Staring into the bright green lights above makes the white pages of Nature appear slightly orange. These familiar experiences are all manifestations of desensitization — the underlying mechanisms of which have more in common than might be expected.

Caffeine, odours and light, as well as most hormones and neurotransmitters, act on a family of cell-surface receptors that span the cell's membrane seven times. Several of these 'serpentine' receptors bind to useful drugs (such as β-blockers and antihistamines) and also to some illicit ones (such as LSD and cannabis). Even in the simplest eukaryotic organisms, related receptors respond to mating pheromones, chemoattractants and nutritional signals. In each case, receptor stimulation leads to the activation of a G protein, which binds to the nucleotide GTP and transmits a signal to an effector enzyme inside the cell. Signalling persists until GTP is converted to GDP, when it stops.

In general, desensitization is characterized by feedback inhibition after a delay. Any external stimulus that triggers a response inside the cell also triggers a process designed to inhibit the response to further exposure to the same stimulus. Desensitization can occur on the order of seconds (to a flash of light), minutes (to odours) or even days (to caffeine).

Receptors have long been considered as likely targets of desensitization, a logical hypothesis considering their role as the 'gatekeepers' of the cell. Prolonged stimulation of most receptors leads to rapid phosphorylation, physical uncoupling from the G protein and internalization. The initial phosphorylation event can result from conformational changes in the agonist-bound receptor, because this also makes it a better substrate for some kinases. For instance, activation and phosphorylation of the green photoreceptors can temporarily alter colour perception, making the world appear less green and more orange. Alternatively, phosphorylation can result from activation of a downstream effector kinase that has among its substrates the activating receptor. Long-term desensitization can occur through increased expression of the relevant kinase, or reduced expression of the receptor. There is ample precedent for each of these suggestions.

Desensitization can also occur at later stages of the pathway — G proteins and effectors are also targets of desensitization. About five years ago, a family of proteins, known as RGS (for 'regulators of G-protein signalling') proteins, was discovered in yeast and was subsequently shown to promote desensitization by accelerating GTP hydrolysis by G proteins, thereby rapidly inactivating the signal. In at least some cases, RGS expression is strongly induced by activation of the pathway, as is typical of feedback-inhibition loops.

More recently, attention has turned to other types of protein modification besides phosphorylation. Some common biochemical modifications, such as glycosylation and myristoylation, are fairly static and are therefore unlikely to have a role in desensitization. The activation of some receptors, RGS proteins and at least one effector kinase result in ubiquitination of these proteins — the covalent attachment of multiple ubiquitin polypeptides to the substrate protein. These proteins are usually degraded quickly by a complex of proteases called the 26S proteasome. In yeast, ubiquitination of the receptor requires prior phosphorylation. This situation is unusual, however, in that the yeast receptor is mono-ubiquitinated rather than poly-ubiquitinated, and is delivered to the vacuole compartment rather than to the proteasome.

Another surprise is the discovery that the downstream kinase Ste7 is also ubiquitinated in response to a chronic stimulus. Ste7, a member of the mitogen-activated protein kinase (MAPK) kinase family, is thought to be a limiting component of the MAPK cascade of reactions. There is now good evidence that the ubiquitination of Ste7 is part of a feedback-inhibition loop that leads to MAPK desensitization. So, whereas attention in the past has focused on desensitization through receptor phosphorylation, we can now consider a new paradigm of desensitization of G proteins and effector enzymes, through modifications other than phosphorylation.

History shows that breakthroughs in our understanding of cell regulation are often followed by an increased understanding of counter-regulatory events. Consider the tandem discoveries of protein kinases and phosphatases, or oncogenes and tumour-suppressor genes. The discovery of RGS proteins, acting in opposition to cell-surface receptors, is another example. What will be next? I believe that the current focus on cell regulation by ubiquitination will soon shift to mechanisms of protein de-ubiquitination. In yeast, there are 16 ubiquitin-processing proteases. This large number suggests that there are highly specific functions for each family member. Deletion of the gene for just one such protease, UBP3, enhances the pheromone response in yeast and promotes pheromone-dependent ubiquitination of the effector kinase Ste7. Perhaps other family members will have similar, pathway-specific effects on cell regulation.

The existence of various neurotransmitters, receptors, G proteins and effector enzymes has long been part of college textbooks and popular reviews. Feedback inhibition is at last gaining a similar degree of recognition, which (as befits the subject) has occurred after a period of prolonged activity and a suitable delay.

FURTHER READING

Wang, Y. & Dohlman, H. G. J. Biol. Chem. 277, 15766–15772 (2002).

Shenoy, S. K., McDonald, P. H., Kohout, T. A. & Lefkowitz, R. J. Science 294, 1307–1313 (2001).

Hicke, L. Nature Rev. Mol. Cell Biol. 2, 195–201 (2001).

Ross, E. M. & Wilkie, T. M. Annu. Rev. Biochem. 69, 795–827 (2000).