Scientists have designed a molecule that reduced the "reservoir" of HIV-infected cells hiding in mice and say the strategy could be developed into a new treatment or possibly even a cure for the disease.

The molecule, which mimics a protein found on the surface of immune cells, also significantly delayed the resurgence of the virus after the rodents discontinued antiretroviral therapy, the team reported Thursday in Cell Host & Microbe.

This molecule helps the immune system combat the virus more effectively than the immune system would be capable of doing on its own, said Priti Kumar, director of the graduate training program in virology at the Yale School of Medicine and last author of the study.

"You're only using immune components which are already there in HIV-positive subjects," she said. "This would be a way to get rid of reservoir cells; this would be one approach toward getting a cure for HIV."

There are about 38 million people around the world living with HIV, the virus that causes AIDS. Antiretroviral therapy can dramatically reduce the amount of virus in a person's blood but doesn't eradicate the infection. If a person stops treatment, the virus returns in full force within several weeks, Kumar said.

When HIV enters the body, the immune system homes in on a protein in the virus's fatty envelope that helps it infect white blood cells called helper T cells. This viral protein binds to a different kind of protein, known as CD4, on the surface of the T cells. 

So-called neutralizing antibodies can thwart the virus by attaching to the envelope protein so it can't bind to CD4. However, HIV mutates very quickly to prevent these antibodies from learning to recognize and latch on to its envelope.

When HIV binds to CD4, Kumar said, its envelope opens up and exposes parts of the viral protein that would otherwise be hidden. This form of the envelope protein attracts the attention of another kind of antibody known as non-neutralizing antibodies. Although they can't save the cell from infection, the antibodies do recruit white blood cells called natural killer cells to destroy the infected cell.

However, HIV has another trick to evade these antibodies: Once the virus enters the cell, it makes proteins that cause all the CD4 molecules and any attached virus, to be withdrawn from the cell surface. As a result, the vulnerable form of the envelope protein is no longer available to be spotted by antibodies.

"It's a very nice mechanism by which the virus can hide," Kumar said.

Most of the infected helper T cells die, but a few survive and become a type of long-lived immune cell known as memory T cells. HIV becomes part of the genetic code of these cells. 

Antiretroviral therapy doesn't kill the virus but prevents it from making more copies of itself. If treatment is discontinued, the virus hidden in the memory T cells begins replicating again. 

To target this reservoir of HIV, Kumar and her colleagues designed a protein that resembles CD4. This molecule can bind to the virus's envelope protein and cause it to adopt the form that the non-neutralizing antibodies can recognize and flag for destruction.

For its experiments, the team examined HIV-infected mice that were genetically engineered to produce natural killer cells like those found in humans. The CD4-mimicking molecules as well as donated plasma that contained antibodies from people with HIV were administered to the rodents. This led to a significant decrease in the number of reservoir cells with HIV that Kumar and her colleagues could detect in the spleen and other tissues of the mice.  

The protein-and-plasma cocktail also lengthened the time it took for the virus to return to high levels after the mice stopped receiving antiretroviral therapy. Rather than the typical two weeks, it took the virus 49 to 60 days to rebound in mice that had been dosed with the protein, Kumar said.

In another experiment, she and her colleagues administered the cocktail to mice that had been injected with HIV-infected T cells from the same person who'd donated the plasma. This provided a more representative picture of what would likely happen in a person, because the strain of HIV that the antibodies in the plasma would recognize matched the one that infected the rodents.

"To our absolute delight, we found that many of these mice in fact did not rebound at all, indicating that it's quite likely that we have completely destroyed the reservoir in the mice," Kumar said.

However, given the short life span of the mice, she cautioned, "It's difficult to say whether they were completely cured."

In addition to the natural killer cells recruited by the antibodies, another group of immune cells called killer T cells also seemed to play an important role in controlling the virus. When the researchers gave the mice a "depleting antibody" that eliminated their killer T cells, the virus rebounded much more quickly.  

Unlike natural killer cells, killer T cells patrol for and recognize infected cells without help from antibodies. Kumar suspects that the protein she and her team gave the mice reduced the HIV reservoir to low-enough levels that it could be controlled by the killer T cells.

These cells "will do surveillance, and when a cell becomes positive for HIV, will clear it out," Kumar said. "You don't really need to go weeding out every reservoir cell as long as there are killer cells that can recognize these cells when they 'turn on' HIV replication."

The one place where the researchers didn't see a major reduction in HIV-infected reservoir cells was the bone marrow. The team has since tweaked its CD4-mimicking molecule to better penetrate this area and next plans to test it in primates.

Kumar envisions that molecule eventually being used in brief treatments rather than continuously.

"It could be in short bursts or short phases," she said. "It's not something like antiretroviral therapy, where you need to maintain it forever."

The study, "Modulating HIV-1 envelope glycoprotein conformation to decrease the HIV-1 reservoir," published May 20 in Cell Host & Microbe, was authored by Jyothi K. Rajashekar, Jagadish Beloor, Liang Shan, Dietmar Herndler-Brandstetter, Irfan Ullah, Kelly Symmes, Andrew Peric, Richard A. Flavell and Priti Kumar, Yale University School of Medicine; Jonathan Richard, Jérémie Prévost, Guillaume Beaudoin-Bussières and Daniel E. Kaufmann, Centre de Recherche du CHUM and Université de Montréal; Sai Priya Anand, Centre de Recherche du CHUM and McGill University Montreal; Gabrielle Gendron-Lepage, Halima Medjahed, Catherine Bourassa and Fleur Gaudette, Centre de Recherche du CHUM; Andrés Finzi, Centre de Recherche du CHUM, Université de Montréal and McGill University Montreal; Emily Lindemuth, Frederic Bibollet-Ruche, Jun Park, Hung-Ching Chen, Beatrice H. Hahn and Amos B. Smith, University of Pennsylvania; Joseph Sodroski, Consortium for HIV/AIDS Vaccine Development (CHAVD), Dana-Farber Cancer Institute and Harvard Medical School, and Harvard School of Public Health; and Marzena Pazgier, Uniformed Services University of the Health Sciences.