Reducing Malaria to its Constituent Parts

LOOKING AT THE CULPRIT

P. falciparum's 23-megabase, adenine-thymine (A-T) rich genome posed a major sequencing challenge. Some initially questioned the feasibility, according to the paper's lead author, Malcolm Gardner, because the parasite's DNA was also unstable in Escherichia coli. "We weren't absolutely convinced ourselves," says Gardner, an investigator at The Institute for Genomic Research (TIGR).

Before committing money, a projected $30 million (US) at the time, funders supported proof-of-principle pilot studies on two of P. falciparum's 14 chromosomes. TIGR sequenced chromosome 2; the Sanger Institute in the United Kingdom sequenced chromosome 3. To address the genome's challenges, researchers used smaller insert libraries and revised gap-closure techniques to accommodate the high A-T content within introns and intergenic regions. This necessitated having researchers do frequent, laborious, and time-consuming gap closures. Researchers also adjusted the annotation and assembly software to deal with the genome's A-T profusion.

Once chromosomes 2 and 3 were underway,³⁴ and the technological and biological hurdles cleared, the project went forward. Investigators used a chromosome-by-chromosome strategy; each center was assigned different chromosomes. "It would've been a very large whole-genome shotgun project for the time," says Gardner, adding that back then it was five to eight times bigger than any shotgun project yet completed. "It seems ridiculous to talk about it like that these days when they can do mammalian genomes by a shotgun process."

Alas, political challenges loomed. As with the Human Genome Project, concerns about access to sequence data surfaced early on. Researchers doing the sequencing feared that chunks of data, if released, would be poached and published by others in incomplete form. Meanwhile, researchers outside of the project wanted immediate access to the genomic data to expedite ongoing research projects.

"There was a fair amount of rancor on the part of community scientists, who felt, 'We've all supported the investment of a lot of money in this project, and now they're generating data and they won't let us see it,"' says David Roos, director of the Genomics Institute at the University of Pennsylvania. Time and money might be wasted on sequencing individual genes that had already been sequenced but not yet released by the genome project. Roos and others addressed these issues by starting the Plasmodium database (PlasmoDB), a centralized data depository.

Access revealed early clues about the parasite's metabolism and pathogenicity. Prior to the genome's completion, researchers identified novel classes of genes involved in antigenic variation, the process by which the parasite evades the host immune system. Researchers also discovered pathways associated with a chloroplast-like organelle called the apicoplast, including an unusual pathway for isoprenoid biosynthesis. Clinical trials targeting this pathway, which is crucial for parasite survival, are now underway.

STARTING ON THE MOSQUITO

Courtesy of CDC/Edwin P. Ewing, Jr., M.D.

Placental tissue revealing the presence of Plasmodium falciparum.

Started several years after the P. falciparum project, A. gambiae avoided many of the former's technological and political hurdles. Indeed, almost para-doxically, if the P. falciparum genome efforts had been started two years later, it may have finished sooner. But the success of the project helped bring down costs and encouraged funders to support other parasite sequencing projects, notes Neil Hall, a P. falciparum study coauthor and assistant investigator at TIGR.

Prior to starting the whole-genome shotgun of the 278-megabase A. gambiae genome in August 2001, investigators had already begun to acquire hints as to gene function. In the 1990s, mosquito researchers made bacterial artificial chromosome (BAC) libraries, did expressed sequence tag (EST) sequencing, and isolated genetic loci responsible for certain traits, such as resistance to insecticides. This preliminary work proved invaluable when assembling the whole genome.

Though more than 10 times larger than P. falciparum, the $9 million A. gambiae genome project – from DNA isolation, to sequencing, assembly, gene annotation, and preliminary analysis – took less than a year. "It's probably the fastest whole-genome project ever done," says lead author Robert Holt, now head of sequencing at the Michael Smith Genome Sciences Center, Vancouver. The mosquito sequence, unlike the P. falciparum, however, is still considered a draft rather than a finished genome.

One feature of the mosquito genome proved to be both a benefit and a drawback: its variation. To get sufficient DNA, investigators ground up about 100 mosquitoes, but the strain they used wasn't fully inbred, thus causing significant genetic heterogeneity. As a result, assembly was tricky; for about 7% of the genome, researchers had two versions of the same chromosomal region. This variation had an upside, however. Researchers were able to identify potentially important genomic variants, or single nucleotide polymorphisms (SNPs), about 450,000 of them. These SNPs may serve as markers for genes underlying specific traits such as pesticide resistance, or for classifying wild mosquitoes according to their subspecies.

Holt and colleagues have begun a large-scale study of wild mosquito populations. They'll collect isolates throughout Africa and sequence small chromosome fragments at regular intervals across the genome to estimate overall variation in the population.

Tracking species and subspecies differences could provide insight into how malaria is transmitted. It might help explain, for example, the genetic basis of A. gambiae's peculiar preference for human blood meals, perhaps a result of its evolving alongside humans for thousands of years. A group at Yale recently characterized and isolated an A. gambiae odor receptor that responded specifically to a compound in human sweat.⁵

The genome also helped investigators isolate three metabolic enzyme families associated with insecticide resistance: carboxyl esterases, cytochrome p-450s, and glutathione S-transferases. One practical implication: If scientists can track mosquitoes and identify those that are resistant, they can avoid spraying, say, DDT on DDT-resistant mosquitoes.

Researchers also recently identified mosquito genes required for the development of the parasite within. The proteins TEP1 and LRIM1 helped the mosquito's immune system fight the parasite⁶; the proteins CTL4 and CLMA2, discovered directly from the genome data, helped to protect the parasite.⁷ "We're now trying to understand ... whether they bind to the parasite, for example, mediating its killing or its protection, or whether it's an indirect kind of reaction," says European Molecular Biology Laboratory staff scientist George Christophides, a coauthor of the latter paper. Future work may uncover parasite- and vector-protein interactions.

OTHER SEQUENCING PROJECTS

Meanwhile, more sequencing projects are underway. The sequencing of the mosquito that causes yellow fever, Aedes aegypti, has begun, and there have been preliminary discussions about sequencing the common house mosquito, Culex pipiens, the vector for West Nile virus. TIGR is sequencing Plasmodium vivax, a less deadly but more widespread species; Sanger has sequenced other malaria parasites found in rodents, chickens, and monkeys. Comparing genomes should help investigators home in on the vectors' and the parasites' Achilles' heels.

Despite many scientific strides, the public health problem remains massive. "We realize that today the genome sequence [of P. falciparum] is doing absolutely nothing for people who have malaria," says Gardner. "But I think it's provided new avenues of research, and it's provided some novel targets."

Eugene Russo erusso@the-scientist.com is a freelance writer in Takoma Park, Md.