Simulations Show how HIV Drug Traps Virus in an Inactive State

Though AIDS survival is up and new cases are down, the HIV virus is still a major cause of sickness and death. Juan Perilla of the University of Delaware and his colleagues used the Anton 2 system at the Pittsburgh Supercomputing Center to understand better how HIV “matures” to its active state. They found that a class of anti-AIDS drugs may work by locking the virus in its immature form, suggesting a route toward new treatments. 

Why It’s Important

Improved AIDS treatments have increased survival, and the number of new cases is down. Still, nearly 40,000 people in the U.S. were diagnosed with an HIV infection last year, according to the U.S. Centers for Disease Control and Prevention. AIDS killed more than 6,500 Americans in 2014. It was the eighth most common cause of death for people age 25 to 34 and the ninth for those 35 to 44. We still need better treatments—and new treatments, in case the virus develops resistance to current drugs. That’s why Juan Perilla of the University of Delaware and his colleagues at the Pittsburgh Center for HIV Protein Interactions (PCHPI) and the National Institutes of Health (NIH) wanted to understand better how the virus “matures” to its active state. This conversion relies on the virus cutting its Gag protein—a large protein that’s split to create a number of viral components, including its outer coat.

“The virus has two stages, immature and mature. You can think of it as like, for instance, a pine cone, which opens up to release its seeds … There are many drugs that act on this system; at least one, which we used in this study, was known to affect maturation specifically.”—Juan Perilla, University of Delaware

When a new HIV virus particle comes out of an infected human cell, it can’t yet infect a new cell. Scientists think the virus evolved a time delay because HIV particles that didn’t wait risked re-infecting the same cell rather than finding a new cell to infect. A family of anti-HIV drugs takes advantage of this maturation step by sticking to the Gag protein and preventing it from being cut. The University of Delaware team decided to study HIV maturation in detail. They used NMR spectroscopy, which is a kind of scan that provides clues about the chemical state of a molecule; cryo-electron microscopy, which provides crude but informative pictures of large molecules; and computer simulations to check and cross-check how the virus switches from active to inactive. Their work also helped explain how the anti-maturation drug Bevirimat prevents the switch.

How PSC Helped

One big challenge for the molecular simulations was that the cutting of the Gag protein that underlies HIV maturation takes from microseconds to a millisecond—millionths to a thousandth of a second. That’s a short amount of time by human standards, but it’s very, very long for a computer simulation of a large protein molecule. On conventional supercomputers, the amount of computational time to create that much simulated time would have been prohibitive. Enter the D.E. Shaw Research (DESRES) Anton 2 system at Pittsburgh Supercomputing Center, a supercomputer purpose-built to perform such molecular dynamics simulations. Anton 2 allowed the team to run the simulations up to 100 times faster than they could have on a traditional supercomputer. PSC makes Anton 2 available to the U.S. scientific community and their international collaborators thanks to DESRES and operational funding by the National Institutes of Health.

“We were looking at this interconversion between the hidden and exposed hotspot [in Gag]. The structural changes that happen to the protein seem to be on a long timescale, microseconds to a millisecond … It would have taken, I would say, at least ten times longer to simulate [on a traditional supercomputer than on Anton 2]. Also, we used an enhanced computational technique called simulated tempering, which results in ten times more sampling.”—Juan Perilla, University of Delaware

The Anton 2 results answered a puzzle over Gag’s structure and dynamics. Previous scientific tests had disagreed on whether the critical part of the protein that gets cleaved to activate the virus is a disordered coil or an orderly “helix”—a kind of corkscrew shape. In the Anton 2 simulations, the protein switched back and forth between a coil and a helix. The anti-maturation drug in the simulation worked by locking the protein in a helix, effectively hiding that part of the protein and preventing it from being cut and activated. Even better, the group’s studies of the real protein with NMR and cryo-electron microscopy agreed with the simulation, giving the scientists much more confidence that the simulation was reproducing the protein’s natural behavior. The researchers reported their results in the journal Nature Communications in November 2017. They hope the new information will lead to new and more effective anti-maturation drugs to fight AIDS and possibly other viral diseases.