May 12, 2011 (CIDRAP News) In detailing a new process that might someday speed the development of antivirals and other disease-fighting tools, researchers said today that they can design protein-protein interactions from scratch with a computer and bind them to the surface of the 1918 pandemic influenza virus.
They focused on a conserved patch on the surface of the virus' hemagglutinin (HA) stem, a region found to be conserved among all group 1 flu strains (H1, H2, H5, H6, H8, H9, H11, H12, H13, and H16 types). The group from the University of Washington in Seattle and Scripps Research Institute in La Jolla, Calif., described their findings today in Science.
Currently, researchers use animal immune systems to generate antibodies or they must screen "libraries" of proteins to identify binding candidates.
Dr. David Baker, lead author of the study and professor of biochemistry at the University of Washington, said in a Science podcast about the study that researchers developed complex algorithms that contain everything known about protein interactions. "It does take a lot of computing power," he said.
The group used the region on the 1918 pandemic H1N1 flu virus, because making a binder for the HA surface could be useful. "In a new flu outbreak, it would be very important to bind and block."
He said the computerized process initially yielded about 80 different protein candidates. The group used a yeast display process, which can test recombination without going through a gene-cloning step, to test each of the designs and found two that bound to the surface. They used x-ray crystallography to validate the designs.
Baker said the process first involves finding "footholds," then engineering complimentary scaffold proteins to anchor the protein interactions to the surface of the HA stem.
In an editorial on the new findings, which appears in the same issue of Science, Dr. Bryan Der and Dr. Brian Kuhlman, both biochemistry researchers from the University of North Carolina at Chapel Hill, said the work shows how far scientists have come in predicting protein-protein interactions.
They wrote that Baker and his colleagues have "solved an enormously complicated jigsaw puzzle" by finding amino acid sequences that can "dock" to targeted protein surfaces and have side chain and backbone atoms that fit together. The software task involved the work about 20 groups worldwide and about 100,000 hours of computing time, they noted.
Der and Kuhlman said that the computerized process has an edge over other techniques, because it allows precise control over binding location. "This specificity is of utmost importance in efforts to develop drugs that target flu's hemagglutinin protein," they added.
Zeroing in on the stem region allows researchers to develop drugs across broader subtypes and prevent resistance, they wrote.
Computerized protein interface design is in the early stages, and it's not yet known if the method will be useful for designing binders for other surfaces or structures, and that further work to understand protein interaction will continue in the decades ahead. "The impacts of rational design and manipulation of protein-protein interactions can hardly be overstated," they concluded.
Baker, in the podcast, said another question still to be answered is whether the engineered proteins will block infectivity. "But our findings suggest that they would," he said.
The process could be used to generate designs that disrupt pathogen entry of other flu strains or bacterial pathogens, Baker said. "There's a huge amount to be done."
Also, the researchers hope to improve the computational method. For example, Baker said the initial protein needed to be optimized before it bound tightly to the virus. He also said the group hopes to fine-tune the computer program to yield fewer and more specific proteins, rather than large groups that need to be screened.
"There's a lot of room for improvement in the computational methods, and we're working hard on that now," he said.
Fleishman SJ, Whitehead TA, Ekiert DC, et al. Computational design of proteins targeting the conserved stem region of influenza hemagglutinin. Science 2011 May 13;332(6031):816-21 [Abstract]
Der BS, Kuhlman B. From computational design to a protein that binds. (Editorial) Science 2011 May 13;332(6031):801-2 [Summary]
May 12 Science podcast