Computer model of influenza virus shows promise of universal vaccine

According to the World Health Organization, an estimated 1 billion people get the flu each year, 3-5 million become seriously ill, and up to 650,000 flu-related respiratory deaths occur. Seasonal influenza vaccines must be readjusted each year, primarily for prevalent strains. Vaccines are highly effective when they match the dominant strain. But if they don’t match, you may get little protection.

The main targets of influenza vaccines are two surface glycoproteins, hemagglutinin (HA) and neuraminidase (NA). The HA protein helps the virus bind to the host cell, while the NA protein acts like a pair of scissors that cut HA from the cell membrane, allowing the virus to replicate. Although the properties of both glycoproteins have been studied previously, a complete understanding of their movements does not exist.

Researchers at the University of California, San Diego have created the first atomic-level computer model of the H1N1 virus. This model reveals new vulnerabilities through glycoprotein ‘breathing’ and ’tilting’ movements. This work, published in ACS Central Sciencesuggesting possible strategies for the design of future vaccines and antiviral agents against influenza.

“When I first saw how dynamic these glycoproteins were, how much respiration and tilt they had, I actually thought there might be something wrong with the simulation,” said the project’s principal investigator. said Rommie Amaro, a chemistry and biochemistry professor. “Once we knew our model was correct, we realized the great potential of this discovery. can be used for

Traditionally, influenza vaccines have targeted the head of the HA protein and, based on static images, showed a tight conformation in which the protein hardly moves. and revealed respiratory motility that exposes unknown sites of immune response known as epitopes.

The findings complemented previous work by one of the paper’s co-authors, Ian A. Wilson, Hansen Professor of Structural Biology at the Scripps Research Institute. Binds to parts of proteins that appear to be unexposed. This suggests that the glycoprotein is more dynamic than previously thought, giving the antibody the opportunity to attach. I was.

The NA protein also exhibited atomic-level movements with head-tilting movements. This provided important insight to co-authors Julia Lederhofer and her Kanekiyo Masaru at the National Institute of Allergy and Infectious Diseases. When convalescent plasma, that is, plasma from patients who had recovered from influenza, was examined, antibodies were found that specifically target what is called the “dark side” of NA under the head.

Without looking at the movement of the NA protein, it was not clear how the epitope was accessed by the antibody. demonstrated an incredible range of motion that gave insight into

The H1N1 simulation created by Amaro’s team contains an enormous amount of detail, equivalent to 160 million atoms. A simulation of this size and complexity can only be run on a few select machines in the world. For this work, Amaro Labs used Titan from Oak Ridge National Laboratory. It was formerly one of the world’s largest and fastest computers.

Amaro is making the data available to other researchers who can better understand how influenza viruses move, grow and evolve. “We are mainly interested in he HA and NA, but there are many other possibilities, such as other proteins, M2 ion channels, membrane interactions, glycans,” he said. increase.

“This also paves the way for other groups to apply similar methods to other viruses. We have influenza variants, MERS, RSV, and HIV. This is just the beginning.”