HARVARD GAZETTE ARCHIVES

A dengue virus envelope protein trimer just after fusion of viral and cellular lipid membranes. A fusion loop (orange) anchors the protein in the membrane. Viral contents (orange) enter the cell (green) via the fusion pore. Membrane fusion is a key event in the entry of a protein-coated virus into a host cell. The structure gives a direct view of fusion peptide loops as they insert into the target membrane, and points to a possible fusion mechanism and strategies for inhibiting viral entry into host cells.

The research involved a major category of viruses known as "enveloped" viruses, so called because of their fatty outer membrane. Class 1 enveloped viruses include influenza and HIV; the new research focuses on class 2 enveloped viruses, responsible for causing dengue fever, West Nile fever, hepatitis C, tick-borne encephalitis, Japanese encephalitis, and other lesser-known diseases.

"Many of these are emerging infections," notes Stephen Harrison, a Howard Hughes Medical Institute investigator and chief of the laboratory of molecular medicine at Children's Hospital, who is the senior investigator on the study. "Infectious disease is a moving target, and understanding the mechanism of viral entry is one of the ways that we can be forearmed against these viruses."

Enveloped viruses infect cells through a series of steps, but the key final step is fusion of the virus's envelope, or membrane, with the membrane of the cell being attacked. Membrane fusion opens the way for the virus to release its genes into the interior of the cell, allowing the virus to reproduce and infection to spread. How fusion occurs is partially known for class 1 enveloped viruses like HIV and influenza, but has been poorly understood for class 2 viruses.

"We knew certain outlines of the fusion mechanism from previous work on the flu and AIDS viruses," says Harrison, also a professor of biological chemistry and molecular pharmacology and director of the Center for Molecular and Cellular Dynamics at Harvard Medical School. "This gets us a lot further. The better we understand membrane fusion, the more flexibly we can think about it as a therapeutic target."

Led by Yorgo Modis, a structural biologist and postdoctoral fellow in Harrison's laboratory at Children's Hospital, the researchers used X-ray crystallography to study a key envelope protein that sits on the membrane of the dengue virus. By aiming an X-ray beam through a crystallized form of the protein, they obtained three-dimensional images, precise down to the atom, showing how a shape change in the protein causes fusion to happen.

"We managed to determine the crystal structure of the envelope protein before and after the membrane fusion event," explains Modis. Before fusion, the protein has an elongated shape. During fusion, it inserts one of its ends into the cell's membrane, while the other end remains anchored in the viral membrane. "The protein then folds in half - jackknifes on itself," says Modis. "This brings the two ends together and forces the cell membranes together, causing the membranes to fuse."

The work yields two promising drug or vaccine targets for inhibiting viral entry, according to Modis. The first strategy would target the envelope protein in its elongated, prefusion form. Here, the protein has a small pocketlike structure in the "hinge" region where the molecule folds in half. Drugs that bind tightly in this pocket could potentially block the hinge motion. "In effect we would prevent the shape change by putting a wedge or lock between the two parts of the molecule that need to move," Modis explains.

A second possible strategy would be deployed later in the fusion process, targeting the folded, postfusion form of the envelope protein. This form has a small piece called the "stem" that folds back onto the main body of the protein during the final stage of membrane fusion. An agent that "looks" like the stem could potentially block the completion of fusion by binding to the main body of the protein, preventing the stem from folding back.

These two strategies - on which Modis and Harrison hold a provisional patent - could be adapted for use with many other class 2 enveloped viruses, the researchers say. "These viruses have very similar envelope proteins that undergo a very similar series of structural changes during cell entry," Harrison says. "For thinking about how to target these proteins with the appropriate therapeutics, we now have a model for a whole family of viruses."