Skip Navigation

Unfit viruses cause worse disease

August 1, 2007 By Ewen Callaway This article courtesy of Nature News.

Computer model pins AIDS on quick-and-dirty strains of HIV.

The strains of HIV that produce most progeny might not be the ones that cause AIDS. According to a new computer model, HIV must evolve to actually become less productive before it causes disease.

The model predicts that, for AIDS to develop, viral strains that kill host cells quickly must drive out those that spend a long time in the host cell just churning out copies of themselves.

The prevailing wisdom is that slow-killing, highly productive viruses will predominate in a patient. But Dominik Wodarz of the University of California in Irvine and his colleague David N. Levy at New York University challenge this wisdom by considering a newly discovered and controversial property of HIV called co-infection, where more than one strain of HIV can infect the cells of the immune system.

Competing to survive

HIV targets a protein called CD4 on the surface of immune-system cells and uses it to gain entry into the cell. It was thought that once a virus was inside the cell it switched off CD4 production, keeping other HIV strains out.

The fittest virus is not virulent enough to cause AIDS
Dominik Wodarz
University of California, Irvine
But this may not be the case, says Wodarz: "It turns out the virus takes its time to downregulate CD4. During this time more viruses can co-infect the cell."

When two viruses infect a cell, the fast-killing one can do its damage before the slower, more productive, virus has had a chance to get comfortable, wiping out the slow strain's advantage.

In Wodarz and Levy's model, clinical AIDS occurred only when immune-system cells were co-infected at a high rate — and the fast-killing virus strains won out. Without co-infection, not enough CD4-positive cells die to cause AIDS.

"The fittest virus is not virulent enough to cause AIDS," says Wodarz.

Food for thought

"It's certainly a provocative paper," says John Mittler, an AIDS modeller at the University of Washington in Seattle. The authors make a lot of assumptions that may or may not be true, he says. For example, the prevalence of co-infection is unknown — Mittler thinks it occurs in only a minuscule proportion of HIV-infected cells.

The model will get people thinking, Mittler says. "I hope it has the effect of causing people to look at the evolutionary effects of having high levels of co-infection," he adds.

Some monkeys and apes harbour a related virus called SIV, but do not succumb to AIDS. This might be because they block co-infection, says Wodarz. A possible treatment could involve doing the same in humans, he says.


  1. Wodarz, D. & Levy, D. N. Proc. Roy. Soc. B doi:10.1098.rspd.2007.0413 (2007).


Need Assistance?

If you need help or have a question please use the links below to help resolve your problem.