H1N1. H5N1. H7N9. The influenza virus comes in many varieties and goes by many names. Its diversity is also its greatest asset: the ability to change two proteins on its surface—H and N in shorthand—lets it evade the immune system and complicates the work of drug- and vaccinemakers. Now, researchers report that they’ve used computers to help them target a region of the virus that rarely changes, and they’ve designed a small protein that renders the virus unable to infect cells and cause disease, at least in mice.
“It’s a new way to rationally develop antiviral drugs,” says veterinary microbiologist Jürgen Richt at Kansas State University, Manhattan, who was not involved in the study. “It’s a nice paper and I think they’re doing good work, but as always there is more to do.”
Today in PLOS Pathogens, scientists report that they’ve designed small molecules to target a particularly vulnerable region of hemagglutinin, one of the two proteins that jut from the viral surface like a flower. They home in on the stem of the flower, a portion of the protein that remains conserved between varied strains of the virus. Scientists previously have shown that antibodies directed at the stem stop a diverse array of influenza viruses from causing an infection, and some are being developed as treatments. But the drug might have an advantage because mouse studies have found that monoclonal antibodies need help from other arms of the immune system that sometimes are compromised in the elderly—the group most vulnerable to influenza virus. Small molecule drugs are also far easier—and cheaper—to manufacture than monoclonal antibodies.
The researchers began with a small protein called HB36.5, which is known to bind to influenza’s hemagglutinin. Using a combination of laboratory assays and computer algorithms, the team tested various mutations in HB36.5, looking for single amino acid changes that would increase how tightly the protein bound to a diverse group of hemagglutinins. Eventually the tests converged on a protein with nine mutations, which the researchers dubbed HB36.6.
The scientists gave HB36.6 to a group of mice both before and after infecting them with a lethal dose of flu virus. When given prophylactically, all of the mice survived and lost less weight than control animals infected with the same virus. When administered after infection, the protein was less effective, but still significantly reduced death and weight loss. Further testing showed that HB36.6 shuts down the virus without help from the immune system.
Deborah Fuller, a microbiologist at the University of Washington, Seattle, who led the study, notes that HB36.6 is still a very long way from the pharmacy shelf. “This is a proof of concept,” Fuller says. “Everything works in mice, it seems. They’re not always the best predictor of what’s going to work in humans.” Still, Fuller and her team hope to move into clinical trials soon.
Even if HB36.6 doesn’t prove its worth in human studies, the researchers contend it opens the door to a new class of computer-generated antiviral drugs. Aside from operating independent of the immune system, the approach may make it more difficult for the virus to become resistant to drugs. Like the combination therapies used to treat HIV, scientists could design a suite of slightly different proteins that targets the hemagglutinin stem region, making it more difficult for the virus to dodge the drug with a single mutation.
Richt agrees the results are encouraging, but points out that HB36.6 may only have limited usefulness in the real world. Flu symptoms typically occur a few days after infection, and HB36.6 lost about half its effectiveness when administered 24 hours after infection in mice. “If we have a pandemic and everybody is nervous, and they want to take it as a prophylactic, that might be ok,” Richt says. But the ideal flu drug would work against all strains when symptoms surface, which remains a tall order.