Rational Protein Design to Dissect Class I Viral Fusion Mechanisms

Abstract

Enveloped viruses utilize surface glycoproteins to facilitate membrane fusion with a host cell in order to deliver viral genetic material, a necessary early step in the infection process. In the case of class I viruses, fusion requires dramatic conformational rearrangements in the transmembrane subunit of the trimeric envelope glycoprotein to form a highly stable six-helix bundle. Current models suggest the transmembrane subunit adopts an extended intermediate conformation during the transition from pre- to post-fusion states. At this time, the fusion protein is anchored in both the host and viral membranes in an elongated state that transiently exposes N- and C-heptad repeat regions (NHR and CHR). Eventual collapse of the extended intermediate, with the CHRs folding back into hydrophobic grooves generated by the NHR coiled-coil core trimer, draws the two membranes into proximity and promotes lipid mixing and ultimately fusion pore formation. A structural and mechanistic understanding of the fusion process is of value in the design of therapeutics, as a peptide drug targeting the HIV-1 gp41 extended intermediate is an FDA-approved antiviral therapy. Here, we explore folding properties of a designed Ebola virus (EBOV) GP2 six-helix bundle mimic as well as peptide antiviral strategies for targeting the GP2 extended intermediate. We observed a pH-dependent stability in our designed GP2 alpha-helical bundles, with a greater stability at low pH. This is contrary to what we expect from globular proteins in general and from other class I fusion proteins, in which the conformational switch is generally the result of destabilization of the prefusion state at low pH. In another project, we sought to improve the potency of a previously developed C-peptide targeting the EBOV GP2 extended intermediate through two methods: cholesterol conjugation to localize the peptide to the site of fusion, and side chain-side chain crosslinking to improve binding by promoting alpha-helical secondary structure. Although both methods improved C-peptide antiviral activity, we noted a lack of specificity for the GP2 protein, indicating caution must be exercised when employing this strategy for development of antiviral therapies. We also examine a conserved structural feature of many class I fusion protein NHR coiled-coils, a heptad repeat stutter, that may play a role in pH-dependent conformational behavior observed in those proteins using an Influenza A model system. By exploring the stability of peptides corresponding to the stutter-containing region of Influenza HA2 in comparison with peptides designed to lack the stutter, we determined that the presence of a stutter in the HA2 NHR mitigates structural dependence on low pH. The stutter permits the alpha-helical formation seen in the post-fusion conformation of HA2 across a pH range; a finding that is consistent with the spring-loaded mechanism of HA2-mediated Influenza fusion. These findings contribute to the understanding of pH effects in class I viral membrane fusion proteins and provide insight into antiviral peptide design.