Immunology

Developing a Common Influenza Vaccine

One of the most successful pathogens ever, the influenza virus is easily spread and potentially deadly. Waves of pandemics arise when the virus mutates the hemagglutinin (H) protein in its outer coat, producing new strains that bypass the immunity to previous versions. Currently, vaccines must be given every year and tailored to the prevailing H strain to be effective. For a long time, the goal of a common flu vaccine, a one-time shot that would silence all strains, has been elusive.

Wayne Marasco, MD, PhD, and Jianhua Sui, MD, PhD, a fellow in his laboratory in the Department of Cancer Immunology and AIDS, are bringing that goal closer to reality. In recent work, Marasco's lab screened a library of 27 billion different antibodies, developed previously in his laboratory, with the hemagglutinin protein from the highly pathogenic avian influenza virus or "bird flu." He found ten antibodies that potently inhibited infection in mice. Unlike most previously isolated antibodies, the new ones did not block the attachment of the virus to the cell, but instead stopped its internalization. Even more surprising, the antibodies blocked the infectivity of other influenza variants that have a different hemagglutinin subtype, including the 1918 pandemic flu virus.

The H5 protein is composed of three monomers, depicted in green, blue, and yellow. The highly mutable H5 mushroom cap region, where most antiinfluenza antibodies bind, is shown at the top of the structure; the highly conserved stem region is at the bottom. The F10 antibody (red) binds to a pocket within the H5 stem region that is critical in the large structural reorganizations required for postattachment membrane fusion. When bound by F10, these structural reorganizations cannot occur, and thus viral entry into the host cell is prevented. 

Cross-reactivity between different hemagglutinin subtypes was a very unusual finding. To understand how that could happen, Marasco turned to Robert Liddington, PhD, a crystallographer at the Burnham Institute for Medical Research. When he analyzed the crystal structure of the antibody-hemagglutinin complex, Liddington found that the antibody bound to an unusual, and usually hidden, region of the hemagglutinin protein. Knowing the exact spot recognized by the antibodies, the researchers then compared that to other hemagglutinin proteins whose sequences were available in public databases. "That was the moment we realized that this was a highly conserved region present in all influenza A viruses," Marasco says. "For the next several months, we worked closely with the Centers for Disease Control (CDC) to prove that what looked like a broadly cross-reactive antibody in silico was confirmed by bio logical data."

Marasco's work, published in February 2009, was especially timely with the news of the H1N1 (swine) flu breaking shortly after. In April, he was on vacation when he got a phone call from the CDC. "It was about the new outbreak," he said. "It [the H1N1 virus] happened to be of the same class that our antibodies recognize. Can you imagine? How often in one's research career does something like that happen, a new pandemic emerging on the heels of a major therapeutic advance in the field?"

While it is still too early to say definitely that Marasco's antibodies provide protection against the new H1N1 pandemic strain in vivo (those experiments are underway), they do block infection in tissue culture studies. Meanwhile, Dana-Farber is in the process of licensing the finding to a major pharmaceutical company, and the NIH is supporting Marasco's lab to develop the antibodies and wants to stockpile them in case of a pandemic outbreak.