Researchers at Harvard and the University of California, San Diego have discovered that mutations in the HIV virus work together to induce drug resistance—a finding that may offer new leads in HIV drug research and therapy.
Resistance arises when mutations in the HIV genome lead to changes in the viral protein structure, thus creating unfavorable conditions for drug binding. As a result of the ease with which the virus mutates, HIV patients eventually develop resistance to all antiviral drugs.
“If we can find the structural mechanism for the HIV virus’s drug resistance, we can see how those mutations will react with the drug molecules and better design drug therapy,” said Jing Zhang, a postdoctoral fellow at Harvard and the lead author of the paper.
Researchers have already identified 99 sites on viral proteins that are associated with resistance to HIV drugs, but inadequate data and the lack of advanced calculation techniques have hindered further research.
The Harvard researchers applied an innovative Bayesian modeling method to gene sequences from both untreated HIV patients and those treated with drugs. Having ran a comprehensive computer search for over a month, the researchers looked for patterns of mutation that were more prevalent among the treated patients—the more common the mutation, the greater the probability of it being responsible for the drug resistance, according to Jun Liu, a professor of statistics at the Harvard School of Public Health and one of the authors of the paper.
The statistical, rather than empirical, approach yielded some remarkable results, Liu said, citing as an example the combination of mutations found at sites 46, 54, and 82 in the HIV genome.
Liu said that analysis found that mutations at solely sites 46 or 82 lead to weak resistance to Indinavir, an HIV protease inhibitor, but mutations at site 54 alone have no effect. More surprisingly, mutations at sites 54 and 45 actually made the virus even more susceptible to the drug.
After the Harvard researchers found possible combinations of mutation site interactions, scientists at UC San Diego digitally simulated the effects of each mutation on the protein-drug binding energies.
"The patterns of collaboration between mutations are very complex," Zhang said. “The fact that our statistical methods were actually testable and could be verified biophysically with computer modeling was one of the main reasons for our successful work.”
This knowledge could one day help doctors better design drug therapies by allowing them to sequence patients’ viruses and predict which drugs the HIV strain would be the most likely to become resistant to in advance, according to Zhang. Such an application could make treatments much more cost-effective, she added.
“It’s very encouraging that we can identify patterns of resistance so that we can understand something about why these patterns exist,” said Daniel R. Kuritzkes, a professor of medicine at HMS who has conducted extensive research on mechanisms and clinical significance of HIV drug resistance.
Describing the approach of the study as “putting a story together,” Kuritzkes said it may be possible to eventually design drugs with molecular structures that accommodate those mutations.
For the study, which was published on Monday, HIV sequences from both treated and untreated patients were obtained from the HIV Drug Resistance Database, an open access database of reverse transcriptase and protease sequences maintained by Stanford University.
—Staff writer Helen X. Yang can be reached at email@example.com.