UC Berkeley News
Press Release

UC Berkeley Press Release

Scientists exploit HIV's noisy genetics to force virus into latency

– A novel strategy for taming AIDS infections has emerged from a new look at the way the human immunodeficiency virus (HIV) evades current treatments.

When HIV infects an immune system T cell, one of two things can happen: The virus can take over the T cell and produce hordes of copies, or HIV can go into hibernation in the chromosomes - a disaster waiting to happen.

A new paper by University of California, Berkeley, scientists shows that these outcomes are not decided by any new factor or protein in HIV or the host, but rather are random occurrences emerging from a "noisy" genetic circuitry.

The scientists, in an article that appeared in the July 29 issue of the journal Cell, report their studies of how HIV can use this noise to potentially establish latent infections in T cells and suggest a way to possibly use this noise to foil HIV, either by forcing all viruses to come out of hiding into the sights of anti-HIV drugs or, more likely, force all viruses into latency.

In fact, they now are experimenting with a hot new technique, RNA interference, to turn down HIV's ability to amplify the genetic noise and thus prevent HIV from taking over any cells.

"Current anti-AIDS drugs halt active replication of the virus but don't get latent copies, so HIV can reactivate into a full blown infection if you stop the drugs," said study co-author David Schaffer, professor of chemical engineering at UC Berkeley. "The positive feedback loop that amplifies this noise is a fantastic target to hit."

Schaffer's colleague and coauthor Adam Arkin, professor of bioengineering and a pioneer of quantitative systems biology, has long argued that processes inside the cell involving a few hundred molecules are inherently random and noisy, contrary to the general perception that cell processes are orderly and proceed step-by-step toward a final product. Cell behaviors only seem deterministic and orderly when biologists look at the average behavior of hundreds of thousands of molecules or cells.

"HIV and other viruses have learned to exploit the noise in the gene expression circuitry as a survival strategy," said Arkin, a Howard Hughes Medical Institute investigator at UC Berkeley and a faculty scientist at Lawrence Berkeley National Laboratory. Schaffer and Arkin are both faculty affiliates in the California Institute for Quantitative Biomedical Research (QB3).

With the ability to amplify the noise through use of biochemical feedback loops - analogous to the loud screech of feedback generated by random noise in a microphone-speaker system - they can generate large effects from a very small, noisy fluctuation in reaction rates.

Arkin showed earlier, for example that, theoretically, a virus that attacks bacteria - bacteriophage lambda - goes into latency in a random, stochastic manner. Now, he, Schaffer and their colleagues have demonstrated, both theoretically and experimentally, that the same random, noisy processes can potentially explain why HIV sometimes forms latent infections.

According to Schaffer, only one in 1 to 10 million T cells in the body becomes infected with a virus that becomes latent. But that's enough to create a reservoir of viruses that can reactivate viral replication and spread in the future. If the host's immune system is strong enough to fight off the first wave of infection, or if a patient begins taking anti-HIV drugs, these latent viruses can wait to emerge when the immune system is less healthy or a patient stops taking the drugs.

Many researchers have assumed that some protein or other factor in the host forces HIV into latency that once in every 1 million T cells. To date, no such factor has been found, though it is being sought as a potential therapeutic target.

Arkin, Schaffer and former graduate student Leor S. Weinberger created disabled AIDS viruses and conducted experiments to show that their noise theory works in the absence of any such factors. The disabled virus contained genes only for promoters that stimulate expression of HIV's genes, and a single gene, TAT, that codes for a protein that stimulates more HIV gene expression. TAT is the key to the positive feedback loop that quickly boosts production of HIV to thousands of copies.

In their experimental system, they showed that the initially low rate of TAT production depended on where in the host's chromosomes HIV inserted itself. In some cases, the positive feedback quickly revved up production of HIV's genes, while in other positions, the feedback was too weak, and HIV soon lapsed into quiescence. These quiescent viruses can, in the right circumstances, be reactivated and enter a productive cycle.

"Viruses in some T cells amplify and produce virus and others don't in a biased random way. This is likely a bet-hedging strategy that ultimately leads to the evolutionary success of this virus," Arkin said.

Computer modeling and experiments showed that this simple system could be explained entirely by random or stochastic processes depending only on where in the genome the HIV landed. No other assumptions, such as a controlling factor, were necessary.

"This illustrates the importance of stochastic fluctuations in gene expression in a mammalian system," the authors concluded in their paper.

The result with HIV is that most T cells rev up production of the virus until the cell dies and thousands of new virus copies are released. But the noise in the system helps assure that, in some cases, HIV does not rev up, and the virus remains latent in the cell until some external event - another infection, for example - awakens it.

Schaffer, whose lab created the crippled HIV for the experiment, has begun experiments with RNA interference to lower initial production of TAT so the feedback loop never gets started. RNA interference is a technique that very specifically blocks messenger RNA to prevent the production of a protein, essentially sending it to be recycled rather than to be transcribed into a protein.

If Schaffer can produce a piece of RNA that interferes with the messenger RNA destined to produce TAT, then TAT will not be able to ignite the feedback loop that produces more virus and more TAT until the virus fills the cell. RNA interference would thus force the virus to become latent.

"In cell culture with model viruses, we have been able to knock down expression of TAT by about 90 percent, which may be good enough, because you only need to nudge gene expression a little bit in one direction to significantly affect viral latency," Schaffer said. The next step is to test this against real HIV.

Support for the study came from the Defense Advanced Research Projects Agency, the National Institute of General Medical Sciences, a Career Award from the National Science Foundation to Schaffer and the Howard Hughes Medical Institute.