Last year the worldwide community of HIV researchers had a major disappointment—its leading potential vaccine, so promising in lab tests, had fallen flat in human trials. Today there is better news: in the new issue of Cell, researchers from Harvard Medical School and the Texas Tech University Health Sciences Center announce that a technique called RNA interference can dramatically suppress HIV's spread not just in a petri dish but also in mice carrying human immune cells. The findings suggest a new mechanism for treating HIV with drugs—and perhaps also preventing it with a vaccine. NEWSWEEK's Mary Carmichael spoke with Harvard's Priti Kumar, one of the study's leaders. Excerpts:
NEWSWEEK: RNA interference is certainly promising—its discoverers won the Nobel Prize two years ago. Can you explain to me how it works, both in general and when it's used against HIV?
Priti Kumar: RNA interference is a mechanism where short molecules of ribonucleic acid called siRNA are introduced into cells. They can attach to very specific strands of mRNA, the chemical that DNA uses to [express genes], and prevent the mRNA from doing its job [by destroying it]. That stops gene expression from taking place. These short molecules of RNA were discovered as an innate defensive mechanism that plants were using against viruses. They are now known to function in mammalian cells, too.
Scientists had the idea of using them against HIV a few years ago, and they proved that the RNA molecules could stop the virus in a lab setting.
Yes, [the laboratory of Kumar's collaborator Premlata Shankar] was one of the first to discover that targeting HIV with RNA interference could lead to the virus' destruction—replication couldn't take place, and thus the spread of HIV from one cell to another could be totally stopped. The challenge then was how to introduce the siRNA into the T cells, the human cells that are targeted by HIV. Once the virus infects the body, there's a huge decline in these T cells, and we suspected that having the short molecules of RNA ready in the T cells could prevent that. The challenge was to get them into those specific T cells in the first place.
So how did you do it?
We attached the short RNA molecules to antibodies that are attracted to a protein found only on the outside of T cells. Binding an antibody to this particular protein doesn't seem to affect the T cell's normal activity at all. And we had to be very careful about that—we didn't want the T cells to get activated or suppressed. But it appears that engaging this molecule on the cells does not have any debilitative consequences; it just delivers the RNA into the inside of the cells.
Once the RNA is inside the cell, what does it do?
It stops three key genes … from working. One of these is a human gene, CCR5, a molecule of choice for preventing the virus' entry into cells. But we also had short molecules of RNA targeting two other HIV genes to destroy the viral RNA if it gains entry into cells. One of the major problems with current HIV drugs, the antiretrovirals, is that the virus can develop into mutated strains that become resistant to the drug. You could also have this happen if you used a single type of siRNA—that could lead to the development of mutated, resistant strains of the virus too. But if you can target multiple genes at the same time, the probability of resistant strains coming up is very low.
So you're targeting the other two genes as a backup. Let's zoom out from the cell for a minute—what happened when you actually tried this out in mice?
Well, one of the other challenges in studying HIV infection is the lack of good animal models. HIV can infect only human T cells. Even with primate models, you are forced to use simian immunodeficiency virus instead, which is closely related to HIV, or artificial viruses that contain parts of both SIV and HIV. But we had a group of immunodeficient mice to study. They lack their own immune system, so you can introduce human T cells into these mice and they don't regard those cells as foreign or reject them.
Essentially, we [used human stem cells] to make a mouse continuously produce all the immune cells of a human while all its other cells were of mouse origin. And when these mice are infected with HIV, you start seeing the decline in T cells afterwards, just like you would in the human body. So in the control, mice that were not given the short RNA molecules, we infected them with HIV, and they showed a very rapid decline in T cells about a week after infection. They were at almost negligible levels of T cells two weeks after infection with HIV. But in the mice who received the short RNA molecules, the T cell levels were very similar to healthy mice. The siRNA molecules were able to totally control the infection.
These mice were given the RNA before they were infected with the virus? Like a vaccine?
Yes, it was being used as a preventive drug. But we also wanted to look at what happened if it was administered post-infection. In that case, we created mice with T cells that had already been exposed to HIV. In these, we saw very high viral loads at first. But once they received the RNA molecules, they were able to control the virus. When we administered the RNA weekly, we found absolutely no drop in T cells. Two of the mice did not show any viral load at all.
No viral load? Does that mean you cured these mice of HIV?
I wouldn't say it's a cure. The problem is that these RNA molecules are transient. They stay in the system only for a short period of time, so you'd have to administer them repeatedly to keep the virus at bay.
How long did the effect last?
In the mouse, we could pick up the presence of the RNA molecules for about nine days. By the ninth day we saw a decline. But that's just for mice.
You wouldn't expect a similar nine-day effectiveness in humans?
Right.
Do you worry that, like the most recent HIV vaccine candidate, this technique might work in lab animals but fail in humans?
Definitely. Several things have to be addressed before we take it into higher animal models. For instance, the antibody we're using right now to deliver the RNA molecules is a mouse antibody. It does not cause any adverse effects in mice. But you cannot administer it to humans repeatedly without any kind of immune consequences. The second thing is the amount of RNA we deliver. Right now, we need at least two antibody molecules to carry in one molecule of RNA into the T cells. That means we can't deliver a large amount of RNA. That is not something you can stick with in humans. We'd have to attach the RNA to another type of molecule instead.
So what's the next step?
We have to go large-scale with this. We'll have to improve the efficiency of delivering the RNA to the T cells. And we are trying to move into higher animals.
Like humans?
Like chimpanzees. It's a very long way from getting into [human] clinical therapy.