« Evolution of color vision: transgenic mice see red | Main | Darwinian agriculture I »

Can a selfish gene stop malaria?

A bird that risks her life to lead a fox away from her chicks may be influenced by a "selfish gene" (Dawkins, 1976). Genes can't think, of course. However, a gene causing behavior that risks the loss of one copy of itself (in the mother) will become more common over time, if this same behavior often saves more than one copy of itself (in the chicks). The gene can be considered "selfish", in the sense that the welfare of the mother, her species, or the whole ecosystem only indirectly affect the gene's spread. It's as if each gene were at war with rivals (other versions of the gene, or alleles) for its place on the chromosome.

The selfish gene concept is now being used to design new methods to control the spread of disease. Mosquitoes that resist infection by the malaria parasite can be made by genetic engineering. Unfortunately, the small benefit (to a mosquito) of resistance to this parasite is probably not enough for resistant mosquitoes to take over in the wild, because most of the animals they bite aren't infected. (It would be nice if the laws of nature always favored human welfare, but they don't.)

How can we make such beneficial genes spread through mosquito populations? This week's paper, "A Synthetic Maternal-Effect Selfish Genetic Element Drives Population Replacement in Drosophila" by Chun-Hong Chen and colleagues at Cal Tech and UCLA, published on-line in Science, demonstrates one interesting approach.

The proposed method would use a cluster of genes that are tightly linked (close together on the chromosome), so that they would usually be inherited together. One gene, if present in the mother, tends to kill her babies at the embryo stage. Call this gene K, with k being the nonkilling variant. A second gene, if inherited from the mother, makes the babies resistant to killing by the mother's K gene. Call this second gene R, with r the nonresistant variant. The third gene (M) is the one that prevents infection by the malaria parasite.

Suppose a female mosquito has genotype Kk (two different versions of the K gene from her two parents). Because the genes are so closely linked, she would also probably have genotype Rr (and Mm). Roughly half of her eggs would receive the r gene from her. The resulting embryos would all die (unless they happened to receive an R gene from their father), because they wouldn't be resistant to the mother's K-gene activity. But half of her eggs would get the R gene from her and survive. Because these would also have the M gene (due to linkage), her only surviving offspring would all be resistant to infection by the malaria parasite.

Chen et al. cite theoretical analyses showing that linked genes with these properties would take over an insect population, so long as they are "introduced into the population above a threshold frequency, determined by any associated fitness cost." Losing half of ones offspring seems like a major fitness cost, but this may not be the case. If a mosquito lays a bunch of eggs in a puddle with only enough food for 100 baby mosquitoes, it may not matter whether she lays 1000 eggs or 2000.

So much for the theory. To test this approach, Chen et al. made fruit flies (Drosophila) with a maternally expressed gene (encoding an interfering RNA, not a protein, in this case) that prevents hatching in normal ("wild-type") embryos. They linked this to a mutant version of the interfering RNA's target. Presence of this mutant gene allowed normal embryo development even in the presence of the interfering RNA.

They started with a mix of wild-type and transgenic genotypes, and allowed the Drosophila populations to evolve in cages. As predicted by "selfish-gene" theory, the wild-type genes disappeared from the population after 10-12 generations, in replicated experiments.

How soon will we see the deliberate release of genetically engineered mosquitoes bearing selfishly propagating genes linked to malaria resistance? This question really has two parts: would it work, and would (politically powerful elites in) malaria-infested countries accept this approach?

Possible technical problems include mutations in the malaria-parasite-killing gene. Because this gene is not very beneficial to the mosquitoes (partly because most of the animals they bite are malaria-free), we could end up with a population of mosquitoes that have the interfering RNA "poison" and its "antidote", but which still transmit malaria. This problem and possible solutions are discussed in the paper.

There is also wide-spread public suspicion about genetic engineering. This is, in my opinion, partly a reaction to the reckless arrogance of some early proponents of genetic engineering ("don't worry, herbicide-resistance genes won't spread to, or evolve in, weeds!"), and partly irrational public hostility to unfamiliar science and technology.

There are other possible ways to guide the evolution of the malaria parasite and/or mosquitoes in ways that reduce the prevalence and severity of malaria. As mentioned in a previous entry, window screens favor the evolution of reduced virulence in the malaria parasite, because only people with mild cases go outside where they can be bitten. Indiscriminate spraying of insecticides selects for resistant mosquitoes, but spraying only the inside walls of houses selects for mosquitoes that stay outside. Of course, this should only be considered with insecticides that have low toxicity to humans, such as, I hope, the one we were all sprayed with last year when our plane landed in Beijing.

I wouldn't rule out a genetic engineering approach, however. Malaria kills millions and makes millions more sick every year. The important thing is that any such program be evaluated and discussed by a wide range of experts (including disease ecologists and evolutionary biologists) before deciding whether the benefits of transgenic selfish-gene mosquitoes will exceed the risks. Then perhaps it could be tested first on a remote island. Before tourism boards around the world clamor to be the test site, remember that the introduced gene wouldn't stop mosquitoes from biting tourists, only from giving them malaria.

Selfish genes have been used for other practical purposes in the past. For example, some plant genes that are inherited only in seed, not in pollen, suppress pollen production, freeing up resources for more seeds (Trends in Ecology and Evolution 10:412). The resulting "male-sterile" plants have been useful in plant breeding. However, reliance on only one such genetic system for maize (corn) breeding led to widespread damage by Southern Corn Leaf Blight. There's a lesson here about not putting all your eggs in the the same basket, but it doesn't necessarily indicate that practical applications of selfish gene theory all involve similar risks.

Other recent discoveries:

Dinosaur extinction wasn't so important to the diversification of mammals.

Experimental evolution of robots shows selection at the level of groups, rather than individuals, favors evolution of within-group cooperation. We already knew this, from experiments with chickens, for example (Poultry Science 75:447; thanks to Richard Dawkins for calling my attention to this paper). The question is whether the right kind of group selection happens in the natural world often enough to overcome the evolutionary effects of within-group competition, especially when migration between groups is possible. Probably not.

Two papers in Nature show how experiments with microbes can be used to explore roles of competition, predation, and migration in evolutionary diversification.


I find the analysis of selfish genetic elements fascinating. Have you read the new book "Genes in Conflict: The Biology of Selfish Genetic Elements", by Austin Burt and Robert Trivers?


You are obviously quite well-read. No, I haven't read that book, just an article or two on the subject.

Did the article discuss the possibility of compensatory mutations outside of the linked group of genes that might counteract the lethal ones?

Not as I read it, but they do discuss other possible failure modes not in my summary. "Therefore, it is important that strategies be available for removal of an element from the population, followed by its replacement with a new element." You might want to read the paper.

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Type the characters you see in the picture above.