Fighting HIV with HIV: in its zeal to keep competing viruses out of a cell it controls, the AIDS virus may have exposed its own vulnerability

Central to Baltimore's idea of intracellular immunization is a phenomenon known as viral interference. After one virus invades a host, any other virus becomes its competitor. The established virus best serves its own interests by excluding any would-be newcomers. Retroviruses such as HIV could minimize competition essentially, by locking the door behind them, once they enter the host cell. To do that, their proviral DNA could direct the host cell to make enough receptor-binding protein to fill, say, all the CD4 receptor sites on the surface of the host cell. Blocking those sites would prevent the proteins of other HIV particles from attaching to the binding sites [see illustration on page 44].

[ILLUSTRATION OMITTED]

Baltimore's idea is that an intentionally damaged or incomplete HIV provirus might be engineered that can produce receptor-blocking proteins without doing anything else to affect the host cell. The virus that gave rise to such a provirus could serve as a vaccine: like a dog in a manger, it would occupy a cell it could not use, but it would still keep out other viruses--in particular, undamaged, lethal HIV.

The strategy is not as far-fetched as it might seem. As I suggested earlier, proviral DNA from several different retroviruses comprises roughly 8 percent of the human genome. Fortunately, large chunks of the genetic material from the original retroviruses are missing from those proviral copies, and so the proviruses cannot make infectious viruses. Comparative genetic methods suggest these proviruses entered our primate lineage during a burst of retroviral infections about 30 million years ago, about the time the old and new world monkeys were diverging. What led to the infections then, or accounts for the relative lack of such activity ever since, is not known.

Nevertheless, these legacy chunks of proviral DNA are still partly functional within the human genome. Although they are too crippled to instruct human cells to make new retrovirus, they may block other retroviruses--at least retroviruses from the same family as the one that originally gave rise to the proviral DNA--from reinfecting our cells. Unfortunately, none of the proviral DNA is closely enough related to HIV to interfere with the HIV binding site, and so we as a species remain susceptible to the AIDS virus.

As with almost every evolutionary story, this one, too, has elements of conjecture and plausibility, but no definitive proof. It is clearly impossible for anyone to go back 30 million years and ask exactly what molecular events led these defective viral genomes to be incorporated inside our own. Yet in the late 1980s a team of pathologists led by Murray B. Gardner of the University of California, Davis, discovered that a group of mice had recently developed immunity to a retroviral infection in a way that probably mirrors what happened to our primate ancestors. The mice were undergoing what primates underwent 30 million years ago--and were showing, incidentally, that not every problem of human medicine can be answered by studying human beings.