The most common tip for avoiding malaria is simple -- don't get bitten. Simple, but difficult. Nowhere near as difficult, though, as replacing natural mosquito populations with modified strains that cannot transmit malaria. But the first steps towards this ambitious ploy may now have been taken: researchers have announced that they can modify the genes of malaria-carrying mosquitoes.

Malaria infects millions of people each year. In lieu of a viable vaccine, attention is increasingly turning to the mosquito. But control of mosquito populations is hampered by the spread of insecticide resistance and efforts to curb the disease are regularly foiled by the malaria parasite Plasmodium developing drug immunity.

People and female mosquitoes (males do not bite) play crucial roles in the life cycle of this parasite. But not all mosquitoes are willing hosts. Some 'refractory' insects do not spread malaria because their cells hold toxins that stop Plasmodium developing. The genes encoding these toxins could make other mosquitoes harmless.

But until now, there has been no reliable way of tinkering with the genes of malaria-carrying mosquitoes. The keys to genetic modification are 'jumping genes': short DNA sequences that, when injected into insect eggs, hop into target chromosomes. These mobile 'transposons' can carry other genes with them -- those for refractory toxins, for example.

Such a genetic transformation system for the fruit fly has been around for twenty years and versions have been developed for other insects, including a mosquito that spreads yellow fever1. Now Andrea Crisanti of Imperial College, London, UK, and colleagues have developed one for the mosquito Anopheles stephensi. This strain is common in India, where malaria is fairly well controlled. But the researchers believe the technology could soon be applied to Anopheles gambiae -- the major carrier of human malaria wreaking havoc in sub-Saharan Africa.

The new technology, reported in Nature2, opens the door to breeding and releasing mosquito populations that are immune to malaria -- a suggestion guaranteed to cause controversy. "The consequences and risks associated with the field release of a transgenic mosquito will require much research and very careful consideration," says Craig Coates who works on genetic modification of mosquitoes at Texas A&M University, College Station.

But even if such a move were approved, malaria immunity would be unlikely to spread through natural populations, says Professor Chris Curtis at the Malaria Centre of the London School of Hygiene and Tropical Medicine, UK. "I doubt that malaria-immune transgenic mosquitoes will ever be more than lab curiosities," Curtis says. "I think the difficulty of driving genetic changes through populations has been significantly underestimated."

A better malaria control strategy, he suggests, is breeding and releasing sterile males. Curtis has now brokered a meeting between Crisanti's group and a second UK research team that recently developed a new method of breeding large populations of sterile male fruit flies3.

Whichever long-term strategy is adopted, this GM toolkit for malaria-carrying mosquitoes will be welcomed by researchers battling a disease that was first described in 1,600 BC.