Nature Neuroscience

The conventional wisdom, repeated in many textbooks, is that there are four primary tastes: salty, sour, sweet and bitter. Complex flavors are generally supposed to arise from a combination of these primary tastes, with the more subtle effects arising from olfaction.

There is a fifth taste, however, which is much less widely recognized. This is 'umami', better known in the west as monsodium glutamate (MSG). Umami was first identified as a taste in 1908 by Kikunae Ikeda of the Tokyo Imperial University. Ikeda, having been struck by the distinctive flavor of seaweed broth, isolated the molecule responsible for the flavor and showed that it was glutamate. Although taste researchers have known about Ikeda's work for decades, it is only recently that umami has gradually gained wider public recognition, probably because of the increasing popularity of Asian food.

It makes good sense that animals should have evolved the ability to taste glutamate. As the most abundant amino acid, glutamate is present in many protein-containing foods, including meat, seafood and aged cheese. Glutamate is also used, in much smaller quantities, as a neurotransmitter in the brain, and neurons have a variety of receptors to detect its presence. In principle, these receptors could also be used to detect glutamate in food, and one of them (a protein called mGluR4) is known to be present in the taste buds of the tongue, making it a plausible candidate for the umami taste receptor. The problem with this idea, however, was that mGluR4 is very sensitive to glutamate, so if it were the taste receptor, one would predict that the tiniest trace of glutamate in food would give an overwhelming taste of umami, which clearly does not happen.

Now, a team of scientists from the University of Miami have solved the puzzle and identified the receptor for umami. The molecule they describe is a modified form of mGluR4, in which the end of the molecule is missing. The strong binding of glutamate to mGluR4 requires this terminal region, and so its absence explains why the truncated form of mGluR4 is less sensitive to glutamate. The authors confirm that the truncated molecule, which they call 'taste-mGluR4', has all the properties that one would predict of an umami receptor. Most importantly, they show that it responds to glutamate at the same concentrations at which glutamate can be tasted, and that chemicals that mimic the taste of glutamate also activate the receptor.

Now the hunt is on to find the receptors for sweet and bitter, which are still not known. Meanwhile, the identification of a receptor for umami is likely to strengthen its claim to recognition as a fifth primary taste, on an equal footing with the four that are better known.

This work is discussed in an accompanying News & Views article by Bernd Lindemann.

Nature Neuroscience

Neurodegenerative diseases (of which Alzheimer's is the most common) constitute a serious public health problem in most industrialized countries, but they are also among the most difficult diseases to study. The symptoms usually develop slowly over many years, and often there is no way to predict in advance who will get the disease. Moreover, it is not normally feasible to take brain biopsies from human patients, making it very difficult to monitor the course of events or to identify the early events that trigger the later disease symptoms.

The best prospect for understanding the cause of these diseases, therefore, is to use animal models, particularly strains of mutant mice that have been genetically engineered to replicate the human disease symptoms. Huda Zoghbi and colleagues at Baylor College of Medicine and the University of Minnesota have adopted this approach to study spinocerebellar ataxia type 1 (SCA1), a rare inherited disease for which a good mouse model is now available.

SCA1 occurs in patients who inherit a mutant form of a protein called ataxin-1. Expression of the defective protein causes the loss of neurons in various brain regions (particularly the cerebellum), which leads to uncoordinated movements and dementia by the third or fourth decade of life. The mutant mice studied by Zoghbi and colleagues show similar clinical features to human patients. The mutant ataxin protein is switched on in the mouse brain soon after birth, but the first signs of symptoms do not occur until several weeks later. Zoghbi and colleagues used these mice to identify other brain genes whose expression was affected by ataxin-1, and which might therefore be responsible for causing neurons to die.

The authors identified a number of genes whose expression is decreased by the mutant ataxin, both in the mice and in a rare human patient. Most of the observed changes began long before the onset of symptoms, suggesting that they may reflect the earliest events in the disease process. Moreover, the results provide an tantalizing hint at the cause of neuronal death; the identities of the genes affected point toward a phenomenon called 'excitotoxicity', in which neurons die after receiving too much stimulation. It should be possible to test this idea in future studies on the mutant mice; if it turns out to be correct, it may provide the basis for new approaches to the prevention of both SCA1 and other neurodegenerative diseases.

This work is discussed in an accompanying News & Views article by Robert Nussbaum and Georg Auburger.

Nature Neuroscience

Children learn early to recognize the faces of their parents, and most of us are much better at identifying faces than other objects. For example, we may pass a person on the street and recognize him or her the next day, whereas we are less likely to register the individual features of cars or chairs. Watching faces is known to activate a small region of the visual cortex called the fusiform face area (FFA). Now a study by Isabelle Gauthier and colleagues suggests that the FFA may not be selective for faces alone, but that recruitment of this area could reflect expertise with discriminating among members of a particular category (species of birds, models of cars, faces of people). The authors used functional magnetic resonance imaging (fMRI) to image the brains of car and bird experts (who had many years of experience recognizing models or species), while these subjects viewed faces, familiar objects, cars and birds. They find that there is a strong correlation between relative expertise and activation of the FFA, in that the car experts recruit the FFA when viewing cars, as do bird-watchers when viewing birds. Familiarity and expertise, rather than faces, seem to be the critical feature in activating what was thought to be a face-selective area.