The fight against malaria could eventually be transformed by releasing into disease-ridden areas genetically modified mosquitoes that cannot transmit the infection.

Scientists in America have engineered a species of mosquito which is resistant to the malaria infection. Its ability to block the infection suggests that it could come to dominate mosquito populations if released into the wild.

The findings offer the strongest suggestion yet that engineering mosquitoes to resist the parasite could help to control a disease that takes up to 2.7 million lives each year, chiefly in Africa. Malaria infects between 300 million and 500 million people each year. Only HIV/ Aids causes more deaths from infectious disease.

Large numbers of GM mosquitoes would be released in areas where malaria is common, where they would interbreed with wild ones. Over several generations, resistance should spread through the mosquito population, so that fewer insects carry malaria. However, this approach would prove controversial with environmental groups, as it would involve supplanting a naturally occurring species with a genetically engineered variant.

Critics have argued that it is difficult to be certain of the effects of introducing new genes. Even the scientists involved accept that further research is needed before any GM insects could be introduced into the wild.

Though the first GM mosquitoes were created seven years ago, they proved to be less fit than their wild counterparts. This would mean that they would quickly die out, and have no effect on malaria transmission.

But a new study, led by Mauro Marrelli, of Johns Hopkins University in Maryland, reversed this position. The GM mosquitoes express a protein called SM1 that blocks malaria infection, and a gene that makes their eyes glow red or green, allowing them to be easily distinguished from wild insects.

The scientists found that while these modified mosquitoes have no advantage when feeding on uninfected blood, they are much better adapted when blood carries the malaria parasite. Infection with the Plasmodium organism does not kill normal mosquitoes, but it does reduce breeding efficiency.

The GM mosquitoes did not suffer from this and over nine generations (several months) they grew in number to make up 70 per cent of a laboratory population, compared with 50 per cent at the outset.

"When fed on Plasmodium-infected blood, the transgenic malaria-resistant mosquitoes had a significant fitness advantage over wild-type," the scientists wrote in Proceedings of the National Academy of Sciences.

However, the species of both mosquito and malaria parasite used in the experiment are not those that are most harmful to humans. The mosquito was the Anopheles stephensi species, the main Asian vector, but the Anopheles gambiae species is more likely to infect humans, particularly in Africa where the malaria burden is worst.

The experimental parasite was Plasmodium berghei, which does not infect humans but is considered a very good laboratory model for Plasmodium falciparum, the most dangerous of the four strains that do.

A further problem is that only a very small proportion of wild mosquitoes are exposed to malaria, and the transgenic insects did not have a competitive advantage when the parasite was not present. This would slow the rate at which they might have an impact on malaria transmission.

A different approach has been adopted by a British team, led by Andrea Cristiani, of Imperial College, London. His team has developed a GM mosquito in which the males have fluorescent testicles, allowing them to be easily identified and sterilised. The goal is to introduce large numbers of sterile males, which would mate with normal females, reducing the number of eggs laid, and thus of malaria mosquitoes. As the mosquitoes are sterile, they would not transmit transferred genes into wild populations.

The deadly bite

- The most severe form of malaria is caused by the Plasmodium falciparum parasite, transmitted by the bite of the Anopheles genus, particularly Anopheles gambiae. It is spread by pregnant females

- Malaria killed people in the Fens until the 19th century

- It has been predicted that global warming may result in malaria returning to Britain

- Malaria infects between 300 million and 500 million people a year, mainly in Africa

- In sub-Saharan Africa, malaria affects mostly young children, with almost 3,000 dying every day

- Symptoms include neck stiffness, convulsions, abnormal breathing and fever of up to 40C (104F)

- Distribution of the tropical disease mainly affects developing countries. About 90 per cent of cases are in Africa

- It costs £6.8 billion a year in Africa in lost GDP. Death and disability lead to the loss of 45 million years of productive life each year

- Alexander the Great, Genghis Khan, Oliver Cromwell, Caravaggio and David Livingstone are thought to have died of it

- Those who had it but recovered include Lord Nelson, Sir Arthur Conan Doyle, Gandhi and Hemingway

Source: WHO / UN / Times database