Sickle cell anemia is one of the textbook classics of how disease-causing alleles can remain in the population. Caused by a mutation in the hemoglobin chain, its effects depend entirely on whether a carrier has one copy or two: While a single copy protects its carriers against malaria, two copies give the carrier sickle cell anemia.

Sickle cell anemia results from a single amino acid substitution in the hemoglobin molecule, the oxygen-carrying molecule in red blood cells. Regular hemoglobin is known as hemoglobin A, whereas sickle cell hemoglobin is hemoglobin S.

Then there's hemoglobin C, which results from a different amino acid substitution at the same point in the hemoglobin molecule.

"C-type hemoglobin has been known for 50 years," said Thomas Wellems, head of the malaria genetics section at the National Institute of Allergy and Infectious Diseases. Early studies done in southern parts of Africa did not clearly show a protective effect of hemoglobin C, but when Wellems and his colleagues were doing research on drug resistance in the Sahel zone, a West African region just south of the Sahara, they noticed that the population there had an extremely high rate of hemoglobin C.

"One of the big problems in West Africa is the cerebral form of malaria, where the parasites actually line the blood vessels of the brain," Wellems said. Once the researchers had noticed the high rates of C-type hemoglobin in that region, further studies found that hemoglobin C was protecting children against cerebral malaria with about 80 percent efficacy.

Wellems noted that not every mosquito bite will transmit malaria, and not every malaria case will be severe.

"There is an entire range of disease - from mild headache to catastrophic illness and death," he said. Having a copy of hemoglobin C specifically will protect against severe disease: "The infection rates were the same, and mild malaria was the same, but AC children were protected against cerebral malaria." The distribution and relative frequencies of hemoglobin A, C and S suggested that in the Sahel zone, it might be more protective to have one copy of hemoglobin C than hemoglobin S, though Wellems noted that the idea is by no means universally accepted.

In the June 23, 2005, issue of Nature, Wellems and his colleagues from the National Institutes of Health in Bethesda, Md.; the University of Calgary in Alberta; University of Bamako in Mali; and Case Western Reserve University in Cleveland, reported on the mechanism of hemoglobin C protection.

Initially, the researchers had hypothesized that something about the AC genotype might slow the growth of the malaria parasite; but "when we brought cells to the lab and began to investigate with precise techniques, we found absolutely no difference in the ability of parasites to grow. So we had a mystery," Wellems said.

Next, the scientists investigated whether the hemoglobin C was affecting the way that infected red blood cells interacted with the rest of the bloodstream. After infection, the malaria parasite modifies the red blood cells to stick to other red blood cells, as well as blood vessels.

Wellems said that "this sequestration and accumulation is very important from the parasite's point of view, because they escape being swept up into the spleen that way," which normally filters out and destroys infected cells. The researchers found that a critical protein that the parasite uses for those attachments, Plasmodium falciparum erythrocyte membrane protein one, or PfEMP-1, was abnormal in two ways in the AC cells: Its overall expression was reduced, and its cell-surface clusters were formed differently in cells containing hemoglobin C. Wellems likened the difference in clusters to being "like a shower mat - lots of small suction cups will make the mat stick better than a few large ones." Likewise, red blood cells containing hemoglobin C tended to have fewer surface protrusions with PfEMP-1, known as knobs.

How hemoglobin C, which is an intracellular protein, should be able to affect those changes at the surface, was the next part of the puzzle. Hemoglobin C is slightly less stable and more prone to oxidation than hemoglobin A; the oxidation products affect the cytoskeleton of the red cell membrane, and that in turn affects Plasmodium falciparum's ability to make the PfEMP-1-containing knobs that allow it to hide from the spleen.

"What made the paper exciting to the referees is that it points to an even more prominent role for PfEMP-1," Wellems said. But that importance is not unmitigated good news for drug developers, given the protein's other features.

The PfEMP-1 protein "is hugely antigenically variable," Wellems said. "Each parasite genome carries 50 to 100 copies, but as it grows, only one is expressed." So far, so good, but as the parasite multiples in the bloodstream, that one copy is not the same in daughter cells as in the mother, condemning the immune system to a permanent game of catch-up.

Wellems called PfEMP-1 an important target, but also noted that "something so variable is going to be a mighty tough vaccine target - as tough as the AIDS virus."