The idea is compelling: malaria kills more than a million children every year in Africa alone. It is also realistic: naturally occurring bacteria, which produce toxins that attack mosquito larvae, can now be made more potent by genetic engineering. And the need is urgent: mosquitoes are becoming increasingly resistant to pesticides, while the malarial parasites they carry are developing ever greater resistance to antimalarial drugs.
Yet there are anxieties about the idea of releasing a microbe deliberately designed to destroy another living species which could not be recalled if anything went wrong. Exploratory experiments in the laboratory beforehand could never simulate conditions in nature. Although such tests, as well as dedicated thinking, would indicate how such a microbe would behave "out there", they could not deliver certainty.
Writing in Parasitology Today, Alan Porter of the National University of Singapore argues that there is nevertheless a strong incentive to develop the new approach. It comes not only from the burgeoning resistance of both mosquitoes and malarial parasites, but also from the ecological need to phase out chemical insecticides.
Malaria is only one of several targets for attack by genetically engineered bacteria. Others include dengue fever and filariasis. They, too, are debilitating and potentially fatal, are carried by mosquitoes and are on the rise in many tropical and subtropical areas.
Some reassurance about the safety of deploying a microbe to combat a nuisance species comes from the fact that this strategy (known as biological control, in contrast to the chemical control of pests) is far from new. For more than 30 years, farmers have dusted plant leaves with Bacillus thuringiensis spores to kill various caterpillars. B thuringiensis does this in nature anyway. All scientists have done is to enhance its lethal capacity by, for example, preparing it in highly concentrated doses.
The World Health Organisation has used B thuringiensis with dramatic success in eradicating river blindness almost totally from 11 countries in West Africa. The bacterium, which works by producing toxins that are poisonous to certain insects, is a powerful weapon to exterminate the larvae of blackflies that carry the river blindness parasite.
B thuringiensis, and the closely related B sphaericus, are also effective against certain mosquitoes, even when applied in extremely low concentrations. In contrast to chemical pesticides, they are thought to be entirely safe for animals and the environment, and they do not affect insects other than the species against which they are targeted.
Unfortunately, their ability to kill the anopheles mosquitoes that transmit malaria is not impressive, and they also have other drawbacks. When sprayed on to water, they quickly drift to the bottom. This is a major snag, as the anopheles mosquito feeds on the surface. The microbes are also sensitive to ultraviolet light from the sun, which reduces their potency.
This is where genetic engineering comes in. Porter describes several lines of research designed to enhance the power of bacteria as weapons against mosquitoes carrying malarial and other parasites. One approach is to give bacteria the capacity to manufacture toxins they do not otherwise produce, by programming them with the genes that make those toxins in other bacteria. Thus B sphaericus can be empowered to destroy one particular species of mosquito, which it does not normally attack, by introducing into it the genes responsible for toxin manufacture in B thuringiensis.
Researchers at the National University of Singapore also hope to overcome the problems that would at present severely reduce the value of B sphaericus and B thuringiensis under field conditions. One tactic is to insert the toxin genes into a bacterium relatively insensitive to ultraviolet light. Another, perhaps in combination, is to engineer a toxin-carrying microbe that does not sink to the bottom when sprayed on water.
While investigators have made some progress by moving genes between different species of bacillus, the most promising avenue may be to place the toxin- producing genes into unrelated bacteria that have other desirable characteristics. Caulobacters, for example, live near the surface of lakes and swamps, so are resistant to ultraviolet light. Moreover, toxin genes from bacillus can be inserted into caulobacters' DNA.
Experiments conducted thus far suggest that the engineered microbes produce only small quantities of the toxins, although this needs to be offset against their much greater persistence at the water surface.
With more than 400 million people living in highly malarious areas of the world, and existing methods of controlling the disease faltering seriously, the need for a new approach is obvious. Past experience with B thuringiensis also suggests that mosquito-killing bacteria with similar toxins can be deployed safely.
Yet key concerns remain. Will toxin genes transfer to other species? If so, with what consequences? And will the genes be inactivated anyway, after the toxins have done their work, since they will be of no lasting value to the microbes carrying them? Only when such questions have been answered satisfactorily can this potentially highly beneficient technology be applied with confidence.