A typical AIDS investigator tests the power of experimental drugs or vaccines in a test tube, in chimps, or even in human patients. But Robert May and Martin Nowak of Oxford University are not typical AIDS researchers. They work on a desktop computer. With that computer, the two mathematical biologists have arrived at a remarkable new theory of how the human immunodeficiency virus, or HIV, causes AIDS. Full-blown AIDS, they say, is not a gradual dismantling of a patient’s immune system by HIV; it is a catastrophic event, triggered when the infecting virus diversifies into so many different strains that the immune system is suddenly overwhelmed.
It’s not news that HIV diversifies. From the moment it enters the body, the virus replicates rapidly, and since like other retroviruses it lacks a mechanism for correcting the genetic mutations that arise during replication, it inevitably produces genetically distinct strains. But researchers have generally assumed that the dramatic escalation of viral diversity seen in HIV-infected people is a consequence of a defunct immune system. Nowak and May claim it is the cause.
They arrived at this conclusion by analyzing the HIV-infected immune system as an example of a predator-prey relationship--the type of analysis typically used to explain booms and busts in competing animal populations. In this case viruses are initially the prey; the predators are T4 cells, immune cells that proliferate in response to a specific infectious agent.
But the diversification of HIV makes this relationship more complex than that of, say, wolves and rabbits. As the virus diversifies, some of the new strains are different enough from the original virus to elude the immune system. Until T4 cells emerge that can recognize and destroy them, these escape mutants can kill any of the existing T4 cells, because the cells have an Achilles’ heel: a receptor on their surface that the virus recognizes and uses to sneak inside. In other words, the prey turns the tables on the predator. This adds up to a most peculiar kind of predator-prey system, says May.
Nowak and May fed data into their computer on the speed at which the virus generates new strains, the number of strain-specific immune cells, and other variables. Then they watched the populations ebb and flow. Again and again, new species of virus sprang up, only to be beaten back by new species-specific bands of immune cells. Then suddenly, after a few seconds of computer simulation (corresponding to years of a real HIV infection), the number of viral species crossed a threshold, and the immune system collapsed like a camel carrying one straw too many--conquered by the sheer variety of viral strains.
Observations from two HIV-infected men support the model’s results. Both men were infected in 1985. In the first man, the number of virus strains kept escalating until he developed AIDS; then diversity fell again--just as the model predicts it should once the absence of immune control allows a few rapidly replicating strains to predominate. In the second patient, the number of viral strains has yet to peak, and that patient has yet to develop AIDS.
Such radical differences in incubation periods have been one of the most puzzling features of AIDS, but they are a natural feature of the Oxford model. Escape mutants are produced in an entirely random way, explains May. So the cascade to the diversity threshold is also random. If escape mutants appear early, an HIV-infected person develops AIDS rapidly; if they appear late, the disease emerges more slowly.
Some AIDS researchers remain skeptical of the model--first, because it is only a model, and one of many at that, and second, because it has been borne out by only two patients. To corroborate the model fully, one would need to monitor viral diversity in a large number of people at frequent intervals from the point of infection until death. That is not easy to do. But if Nowak and May’s model does turn out to be correct, it would have important implications: it would suggest that early treatment with drugs that seem to reduce viral diversity, such as AZT or DDI, offers the best hope of keeping the virus at bay.