Structural and Dynamical Determinants of Influenza Pathogenicity and Virulence

Rommie Amaro, University of California, San Diego

Usage Details

As illustrated most recently during the 2009 "swine flu" pandemic, novel strains of influenza A may emerge from animal reservoirs and spread globally among humans well before a vaccine against them can be developed. Thus, there is keen interest in identifying and evaluating the threat posed by potential pandemic strains before they emerge. However, given the large diversity of influenza A viruses found in nature, it is not clear which potentially pre-pandemic strains to focus on, nor how safe studies of their potential for transmissibility and pathogenicity in humans can be performed. Recent highly controversial approaches to this problem have used genetic engineering and artificial selection of highly pathogenic H5N1 "bird flu" viruses to obtain mutants capable of respiratory transmission among ferrets, the preferred animal model for human influenza. There has been great concern over the possibility of these viruses escaping from the lab. A virtual, in silico "laboratory" in which to study fundamental properties of host-pathogen interaction without risk to public health will be developed with this work. The techniques are applicable to study of a wide range of mutations and genetic backgrounds due to the wealth of sequence and structural information available for the influenza virus. Furthermore, the creation of a virtual lab will expand the number of researchers who are able to investigate such phenomena, as researchers who do not have access to expensive, specially designed experimental laboratories for high risk research will be empowered by these approaches.

The work addresses a fundamental question in influenza biology: What is the mechanism by which virulence is enhanced by deletions in the stalk of the influenza neuraminidase (NA) surface protein? MD simulations of the individual influenza membrane glycoproteins hemagglutinin (HA) and neuraminidase, carried out previously in the Amaro Lab, have identified experimentally validated antiviral compounds that bind to new glycoprotein pockets never before captured by experimental techniques. This work has been remarkable in its ability to rationalize experimental data and drive experimental design in a prospective fashion. However, influenza glycoproteins are components of a much larger virion surface that collectively participates in host recognition and infection processes. To study this, atomic level model systems were built based on cryo-electron tomographic data of full viral particles; these full virus particle simulations will be the largest atomistic simulations performed to date and allow researchers to explore the surface glycoprotein distributions in realistic context. Additionally, the functional balance of HA, which binds to host-cell receptors, and NA, which abrogates receptor binding via cleavage of terminal sialic-acid residues, has not yet been fully characterized. One hypothesis is that variable HA/NA distribution patterns and mutations produce unique electrostatic properties that alter the binding kinetics of both pharmacological and endogenous ligands, thereby contributing to virulence. This hypothesis will be tested in state-of-the-art MD simulations of entire viral constructs over a one-year period.