Natural selection is not natural perfection. Time and again, biologists have discovered traits that are both beneficial and harmful. Perhaps the most famous example is the devastating disorder known as sickle-cell anemia. To get sickle-cell anemia, you have to inherit two faulty copies of a gene that helps build hemoglobin, the molecule that traps oxygen in red blood cells. In this condition, hemoglobin can't hold its shape if it's not clamped around oxygen. Without it, the defective hemoglobin collapses into needle-shaped clumps, which then turn the cell itself into a sickle shape. The sickle cells snag in small capillaries, and the blood can no longer supply as much oxygen to the body. People who inherit only one copy of this defective gene can get by on the hemoglobin made by the remaining normal copy. But people who get two copies of the bad gene make nothing but defective hemoglobin, and they're usually dead by the time they're thirty. A person who dies of sickle cell anemia is less likely to pass on the defective gene, and that means that the disease should be exceedingly rare. But it's not--one in 400 American blacks has sickle cell anemia, and one in ten carries a single copy of the defective gene.

In the 1940s, scientists discovered what keeps sickle-cell anemia so common: the defective gene provides protection from malaria. Carrying a single copy of the gene reduces a person's chance of getting severe malaria by a factor of ten. Malaria is caused by a single-celled parasite called Plasmoidum carried by mosquitoes. Normally it feeds on hemoglobin, but a red blood cell deformed by the sickle-cell gene somehow becomes a miserable home for the parasite. The needle-shaped clumps of hemoglobin may be able to spear the parasites, or the deformed cells may not be able to pump in potassium, an element essential to Plasmodium. It's also possible that it is easier for the immune system to recognize infected blood cells when they are deformed by sickle cell anemia.

Malaria wreaks colossal damage in many parts of the world. Today it kills over a million people a year, mostly children, and it has been plaguing our species for thousands of years. Carrying a single copy of the sickle-cell gene boosts the odds that people can have children in malaria-prone regions. Unfortunately, when two people who carry the gene have children together, there's a one-in-four chance that each child will get both copies of the gene. Over many generations, the advantage of having one copy of the gene outweighs the disadvantage of having two—at least in populations that have endured centuries of malaria.

In the decades since the discovery of the sickle-cell trade-off, scientists have discovered that several other defenses to malaria have evolved where the disease is a high risk—in Africa, the Mediterranean, Southeast Asia, and New Guinea. And many of these adaptations come with drawbacks of their own. Now a new study offers evidence of yet another mixed blessing: one defense against malaria may make people prone to alcoholism.

The discovery of this defense sprang up in an unexpected place: on the tongue. In recent years scientists have deciphering the molecular biology of how we taste. They have pinpointed several of the genes that produce receptors on taste bud cells. They've also reconstructed the structure of the receptors, and have even discovered some of the molecules that locked onto them. And scientists have also been reconstructing the evolution of those taste receptors. It turns out that they're the product of a complicated history. Taste receptor genes can get accidentally duplicated, and mutations to the new copies can cause them to grab different molecules. This growing diversity of receptors can let animals perceive a growing diversity of tastes—in some cases tastes of dangerous toxins in foods.

Compared to other primates, humans and chimpanzees have a relatively bad sense of taste—perhaps because we eat meat and fruits, as opposed to leaves and other plant material that's loaded with dangerous foods. But scientists have identified a couple taste receptors that have experienced a significant amount of natural selection in the human lineage. This summer a team of scientists reported the discovery of one of these highly evolved genes, known as TAS2R16. The evidence indicates that the receptor causes a feeling of bitterness in response to compounds called beta-glucopyranosides, which plants and insects produce to protect themselves against predators. If these compounds get into a person's intestines, they produce cyanide as they are broken down. Avoiding beta-glucopyranosides thanks to a bitter taste may keep people healthy, and thus be favored by natural selection. The researchers found that an ancestral version of the receptor was replaced by newer versions on many occasions, beginning over 80,000 years ago. The newer versions produce a nastier taste to the beta-glucopyranosides.

But the researchers discovered a peculiar exception to this rule. Some populations in Africa had unusually high levels of the ancestral version of the gene. These populations also turn out to be at very high risk of malaria.

Why would malaria favor a weaker sense of bitterness? One possibility is that beta-glucopyranosides can fight the parasite that causes the disease. Cyanide isn't just bad for people, but for Plasmodium as well. It can even trigger a sickle-cell-like condition in red blood cells. When malaria poses a major risk, the danger of eating poisons may be offset by their protection against the disease.

TAS2R16's intriguing history prompted scientists to look for other conditions with which it might be associated. Previous research had found a genetic disposition towards alcoholism, although the scientists could only link the diseases to a large chunk of chromosome seven. In a new study in press at the American Journal of Human Genetics, researchers now report that this region contains TAS2R16. The scientists zeroed in on the taste receptor gene, comparing the versions carried by alcoholics and their relatives (2310 people were studied all told).

The researchers discovered that people who carried the low-bitterness version of the gene were at a significantly increased risk of alcoholism. They also found that this gene was rare in European-Americans in their study, but 45% of the African-Americans carried it. Based on these results, the scientists suggest that a weak sense of bitterness not only provided protection against malaria, but also changes the taste of alcohol. Other versions of TAS2R16 may give alcoholic drinks a nasty taste. When these drinks don't taste as bad, it may be easier for people to develop alcoholism. This study by no means slight the complexity of alcoholism--genes probably only account for half of the variation in people's risk, and those genes probably all have their own complex evolutionary history. But it's further evidence of the many evolutionary trade-offs that have shaped our genomes, and our lives.