The seeds of malaria: recent evolution cultivated a deadly scourge.

The statistics are grim. The parasites that cause malaria infect 300 to 500 million people annually. As many as 3 million of these will die of the disease this year, making it humanity's deadliest infection. Nearly half the world's population lives in countries where malaria epidemics occur, and as the parasites' resistance to drugs grows, the toll is expected to steadily worsen.

How this mosquito-borne disease became the menace it is--and how it continues to get the better of both the human immune system and modern medicine--has puzzled researchers struggling to understand and control malaria. Its evolutionary relationship with people is of more than academic interest. Understanding the history of malaria and the conditions from which it arose could give scientists an edge in finding new therapies.

The single-celled rotozoan Plasmodium falciparum, the most deadly of four Plasmodium species that cause human malaria, has preyed on people as far back as the human race's evolutionary split with chimpanzees 6 to 10 million years ago. Molecular geneticists, however, are finding evidence that malaria's devastating impact is a relatively recent phenomenon. New findings also hint that the disease is more lethal today than it was just a few thousand years ago.

People have evolved a variety of defenses in response to malaria's threat, and the pace of that evolution gives researchers insight into when the threat arose.

Theorists propose that various genetic mutations among certain groups of people are linked to malaria. These mutations occur almost exclusively in regions where the disease historically has been a killer, and each seems to protect against the infection in one way or another.

Most of these malaria-protective mutations also have a downside, causing blood disorders such as sickle cell anemia and thalassemia. Natural selection would be expected to filter out these mutations or at least keep them at very low frequencies unless their effects in opposing malaria balanced out losses from genetic disease. Scientists call this process "balancing selection."

At least one such disease appears to have cropped up in the past few millennia, according to a recent study conducted by Sarah A. Tishkoff of the University of Maryland in College Park and her colleagues. This implies that malaria earlier on wasn't a sufficiently serious cause of mortality to maintain the mutation through balancing selection.

The researchers examined a series of genetic mutations widespread in regions that show endemic malaria. They appear in the gene responsible for production of an enzyme known as glucose-6-phosphate-dehydrogenase, or G6PD. The mutations can cause G6PD deficiency and result in life-threatening anemia.

The Maryland-led researchers examined the highly variable gene in 605 people from 10 African and 5 non-African populations. The group catalogued the variants, or alleles, of the gene that the participants possess.

Tishkoff's team focused on two alleles that cause G6PD deficiency. Known as Med and A-, these variants appear in Mediterranean and Africa populations, respectively. The researchers looked closely at the DNA flanking each variant.

If either mutation had occurred long ago and been maintained by balancing selection, DNA differences would have accumulated around different copies of each allele. The researchers, however, found few differences in this neighboring DNA among people sharing a variant. This suggests that little time has elapsed since Med and A- arose in the human genome.

Applying a molecular-clock techniquebased on a predicted rate at which differences accumulate, the researchers estimate the age of A- at between 3,840 and 11,760 years. The Med allele, they suggest, goes back just 1,600 to 6,640 years.

Since natural selection would maintain Med and A- alleles only under conditions of severe malaria, the timing of the alleles' appearance presumably corresponds to the rise in human mortality from the disease. That could have been within the past 12,000 years in Africa and even more recently in Europe. Earlier, Tishkoff infers, the disease must have been significantly less common, less lethal, or both.

What caused malaria's terrible transformation? Any one or a combination of several biological factors could have played a role. An evolving parasite, perhaps with novel adaptations that permit the cells to evade immune detection, might have made the disease more deadly or caused it to spread more easily among people. Changing mosquito behavior, such as becoming better adapted to feeding on people, might have increased the incidence of bites and thereby of infection. Or shifting human habits might have put an unprecedented number of people in harm's way.

Climate change, too, could have played a part by bringing mosquitoes and people closer together.

Unraveling the mystery and determining its medical relevance requires assessing the evolutionary histories of all three organisms: the human host, as Tishkoff has begun to do; the mosquito; and the malaria parasite itself.

Clues in the genome of the P. falciparum microbe suggest a recent, dramatic population expansion. In 1998, Francisco J. Ayala, Steven M. Rich, and two of their colleagues at the University of California, Irvine made a proposal based on their studies of genetic diversity. They suggested that all living P. falciparum have descended from a single ancestral strain--a "malarial Eve," as they dubbed it--dating back between 5,750 and 57,500 years.

Such a scenario, called a bottleneck or founding event, usually occurs when a subpopulation of a species branches off and evolves in a new direction. If the new subspecies is more successful than all other populations, it may completely replace them over several generations. This triumph would establish an almost complete genetic homogeneity across the species and thereby reset the species' molecular clock to zero.

Rich and Ayala suggest that just such an ascendant subpopulation of P. falciparum arose within the past few thousand to tens of thousands of years. The putative founder could have been a highly adapted parasite that was especially effective at spreading rapidly among people--and particularly deadly, as well.

Since the work of Ayala and Rich, other researchers have probed the genome of P. falciparum for the telltale genetic uniformity that indicates a past bottleneck. The preponderance of evidence supports the bottleneck model.

In one study, researchers led by Daniel Hartl, a population geneticist at the Harvard School of Public Health in Boston, examined the DNA in eight geographically diverse populations of P. falciparum. Among the specimens, the scientists catalogued the genetic variation found in 25 segments of DNA that don't contain any genes. Such segments are presumed to be unaffected by natural selection.

In the July 20 SCIENCE, Hartl's group reported little variability in these DNA segments. That uniformity suggests that although distributed from Honduras to Papua New Guinea today, the parasites diverged from a common ancestor at some time within the past 23,000 years, probably less. "Something happened perhaps 6,000 to 10,000 years ago that cleansed the variation that existed in the ancestral population," says Sarah K. Volkman of Harvard University, the paper's lead author.

Other population geneticists have reached a similar conclusion by examining the genes in the parasite's mitochondria. Those structures within cells contain their own DNA, which they pass virtually unaltered from one generation to the next. David J. Conway of the London School of Hygiene and Tropical Medicine, with researchers from five other countries, compared mitochondrial DNA among several strains of P. falciparum and one of Plasmodium reichenowi, a related parasite that causes malaria in chimpanzees.

In contrast to the extensive variation between P. falciparum and P. reichenowi, the researchers found little variation in mitochondrial DNA among different strains of P. falciparum. If, as earlier data suggest, the two Plasmodium species split 6 to 10 million years ago when chimps and humans diverged, the uniformity of mitochondrial DNA among the P. falciparum strains indicates a divergence in recent times. That modern epoch--at most the past 50,000 years, say the researchers--contains the hypothesized bottleneck, they noted in the November 2000 MOLECULAR AND BIOCHEMICAL PARASITOLOGY.

Mario Coluzzi, a leading mosquito researcher at the University La Sapienza in Rome, argues that P. falciparum's bottleneck could not have occurred by itself, but rather required simultaneous evolution in the mosquito that spreads it. Numerous species and subspecies of mosquitoes serve as carriers, or vectors, for malaria, and some can shuttle multiple species of the parasite between infected people and new hosts.

Most mosquitoes aren't adapted to feed exclusively on people or inhabit their settlements and dwellings, but Anopheles gambiae is. As such, it serves as the main vector in Africa for P. falciparum.

By living in close association with people in Africa, A. gambiae offers the parasite it carries the most reliable promise of rapid transmission to new human hosts. In some regions of Africa, individuals are typically bitten by hundreds of infected mosquitoes per year, says Andrew Spielman, who studies vector-borne diseases at Harvard University's Center for International Development and is a coauthor of Mosquito (2001, Hyperion).

Such a continual, rapid cycle of transmission, notes Coluzzi, is essential to P. falciparum's self-perpetuation. The parasite is so destructive that it can't linger long in any one host. He suspects that accelerated transmission unleashed the highly destructive strain of P. falciparum that causes the modern scourge.

At one time, conditions were much less favorable to the rapid circulation of A. gambiae among people, Coluzzi says. These insects need shallow, sunlit pools of water to lay their eggs in, and the dense forest that covered most of Africa wouldn't have provided such pools. Coluzzi's research suggests A. gambiae is highly adaptable, and he argues the mosquito could have fed on other mammals when forest-dwelling people were less accessible.

Rich thinks that under proper conditions, a "super-vector" mosquito could have replaced all contemporary A. gambiae while carrying its particular strain of P. falciparum along for the ride. He suggests that an extended period of climatic warming around 8,000 years ago might have altered the environment in a way the benefited the super-vector strain.

Agriculture is considered as another factor in malaria's spread. In 1958, anthropologist Frank Livingstone, then a professor at the University of Michigan in Ann Arbor, proposed that ecological changes accompanying the human transition to farming fostered the spread of malaria. That theory puts Africa's first farmers and the land changes they wrought behind the widening sweep of malaria's scythe.

According to Livingstone's scenario, early West African cultivators about 3,000 years ago began clearing forest to grow crops. The change produced two factors favoring A. gambiae: more sunlit pools of water in which the insects could breed and a concentrated population of people on which they could feed.

As cultivators slashed, burned, and planted swathes through the forest, A. gambiae mosquitoes could have adapted to feed exclusively on people, suggests Coluzzi.

The agricultural scenario is appealing, but the archeological evidence to back it up isn't strong. Slash-and-burn cultivation has been common throughout much of sub-Saharan Africa for at least 500 years, but there is "very scant evidence [of it] in tropical forest zones in the archeological record," says Robert Dewar, an African archeologist at the University of Connecticut in Storrs.

Not all genetic studies in P. falciparum support a recent bottleneck, either. Austin L. Hughes, a population geneticist at the University of South Carolina in Columbia, has data showing substantial variation in portions of the parasite's genome that code for certain cell-surface proteins. The host's immune system uses these proteins, called antigens, to identify and attack the parasite. The change of a single molecule, called a point mutation, in the genes that Hughes examined can alter the shape of a surface protein, disguising it from the immune system.

In a study appearing in the Sept. 7 PROCEEDINGS OF THE ROYAL SOCIETY OF LONDON B, Hughes and Federica Verra, a parasitologist and Coluzzi's colleague in Rome, found a large number of point mutations in 23 locations in the P. falciparum genome.

That great variety suggests that P. falciparum's molecular clock has been ticking for at least 300,000 years since its various populations diverged from a common progenitor. In contrast to the bottleneck model, this scenario suggests a relatively stable parasite population as old as modern humans themselves.

Hughes suggests that selective pressure on one or a handful of genes might have eliminated considerable variation along the stretches of DNA that other researchers have studied and it could have given the appearance of a bottleneck.

"The huge amount of antigenic diversity," says Verra, helps the parasite evade the immune system and makes malaria a tricky disease to tackle medically. Immunity after a malaria infection is only partial and impermanent.

Critics of Hughes' report acknowledge the considerable antigenic diversity but doubt that it indicates a long evolutionary period. It could be, Rich says, that the mutations are very ancient and have been maintained over time, even through a bottleneck. Alternatively, he says, they could have developed all of their antigenic diversity in the short time since that bottleneck. "The genes that are the target of vaccine efforts [could be] evolving very, very rapidly," he says.

Rich, who's now at Tufts University Grafton, Mass., and Ayala recently looked at three highly variable P. falciparum antigen genes. Much of the variation appears in stretches of DNA characterized by short, repeated sequences of nucleotides and consists of differing numbers of these repeats, they reported in a study published in the June 20, 2000 PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES.

Duplication and deletion of repeat sequences are thought to occur much more quickly than the point mutations that Hughes studied. Thus, Rich and Ayala infer, the high variation found in antigen-coding regions of the P. falciparum genome may reflect not an ancient origin but a relentless evolutionary arms race against the human immune system.

Understanding when and how malaria reached epidemic proportions could bear on scientists' plans for developing a vaccine and future therapies. If antigenic diversity stems from the parasite's gradual evolution along with that of people over hundreds of thousands of years, then science may be able to catch up to the parasite and figure out how to shut it down.

If, on the other hand, the antigenic diversity emerged within the last few thousand years, it probably will continue to evolve rapidly in response to the human immune system, drugs, and vaccines. "If a vaccine targets one of those repetitive genes, you're really aiming at a moving target," warns Rich.