Mechanism of Filovirus Entry: Disruption and Stabilization of the GP2 Post-Fusion Conformation

Abstract

My thesis work focuses on the mechanisms that allow filoviruses to enter human cells. The Filoviridae family consists of both ebolaviruses and marburgviruses. In order for enveloped viruses to enter the cell, they must fuse the viral and host membranes. Filoviruses enter via the host endocytic pathway and, once inside the endosome, conformational changes in the surface glycoproteins GP1 and GP2 mediate membrane fusion. During these conformational changes, the fusogenic protein GP2 forms a thermostable alpha-helical bundle to overcome the kinetic barrier of membrane fusion. Inhibiting the formation of the alpha-helical bundle prevents the virus from entering the cell; peptides and small molecules have been designed that arrest these conformational changes in other viruses. We have reported that delivering a peptide derived from the alpha-helical segment of GP2 to endosomal compartments with a protein transduction domain (PTD) derived from HIV trans-activator of transcription (Tat) results in peptide-mediated inhibition of viral entry (we refer to this peptide as 'Tat-Ebo'). We demonstrate that the antiviral activity of the peptide is both sequence specific and dependent on endosomal localization.;We have also designed two monomeric alpha-helical bundle proteins (Ebo6H and Ebo5H) modeled after the alpha-helical region of ebolavirus GP2 in order to gain further insights into filovirus fusion proteins. These proteins are highly thermostable below pH 5.3 and the melting midpoint, Tm, drastically decreases as the pH increases (DeltaTm of 37°C over 0.8 pH units for Ebo6H). The pH range in which Ebo5H and Ebo6H are most stable is consistent with the pH found in the late endosome, suggesting a mechanism for conformational control that is dependent on environmental pH. Examination of the crystal structure of GP2 revealed oxygen atoms on anionic residues side chains within 3 A of one another in the folded state. We hypothesized that electrostatic repulsion between these anionic residues was responsible for the destabilization of the alpha-helical bundle at neutral pH. We created an Ebo6H variant in which four of these anionic residues were mutated to neutral analogs and the protein was no longer highly thermostable at low pH (DeltaTm = 80°C at pH 4.7). Finally, we purified and characterized the marburgvirus GP2 ectodomain (MarV-GP2) and demonstrated that the alpha-helical bundle of this protein was also stabilized by acidic pH. Inspection of a structural model of MarV-GP2 revealed sets of anionic residues in close proximity. We prepared variants in which two glutamate residues were mutated to glutamine and these proteins had higher thermostability at neutral pH (DeltaTm ≥20°C). These results represent the first case of an alpha-helical bundle derived from a viral fusion protein displaying increased stability at acidic pH. These findings provide thermodynamic insights into the mechanism of filovirus fusion and complement the current model of fusion.