C-Terminal Tyrosine Residues Modulate the Fusion Activity of the Hendra Virus Fusion Protein

Paramyxoviruses are enveloped, negative-stranded RNA viruses which enter cells by fusion with a cellular membrane (1). Membrane fusion is promoted by specific viral glycoproteins, which mediate both attachment of the virus to target cells, and subsequent fusion of the viral and cellular lipid bilayers. Most paramyxoviruses require both an attachment (HN, H, or G) and a fusion (F) protein to enter cells, with fusion generally occurring at neutral pH (1, 2). The Hendra and Nipah viruses are recently emerged, highly pathogenic zoonotic paramyxoviruses, classified as biosafety level four pathogens due to their ability to infect humans, high mortality rates, the absence of treatments or vaccines and the possibility of human to human transmission (3). The attachment (G) and the fusion (F) proteins are both required for Hendra and Nipah membrane fusion and viral entry. Hendra F is a 546 amino acid type I integral membrane protein, which folds as a homotrimer and is post-translationally modified by the addition of carbohydrate chains (4). Similar to other class I fusion proteins, Hendra F contains a fusion peptide (FP), two heptad repeat regions (HRA and HRB), a transmembrane domain (TM), and a 28 amino acid cytoplasmic tail (CT) (Fig. 1). Figure 1 Schematic representation of the Hendra virus fusion protein Paramyxovirus F proteins require proteolytic processing of the F0 inactive precursor to the fusogenically active F1+F2 form. This cleavage event positions the FP at the N-terminus of the newly-formed F1 subunit. While the majority of paramyxovirus F proteins are cleaved by furin within the trans-Golgi network (5–8), processing of Hendra and Nipah F requires endocytic recycling (9–13) and cleavage by the endosomal/lysosomal protease cathepsin L (10, 11). A YXXΦ endocytosis motif within the F CT (where X represents any amino acid and Φ a hydrophobic amino acid), is critical for Hendra and Nipah F endocytosis and cathepin L proteolytic processing (9–13). Following cathepsin L cleavage, Hendra and Nipah F are recycled to the cell surface, where they are thought to interact with the attachment protein, G (10, 14, 15). G binding to the cellular receptors ephrinB2 and/or B3 (16, 17) is hypothesized to disrupt the F-G interaction, leading to triggering of membrane fusion by F (15, 18–20). Mutations that inhibit the F-G interaction also inhibit membrane fusion (18), and F-G avidity is inversely proportional with fusion activity (19, 20). Once F is triggered, the FP inserts into the target membrane and the HRB and HRA domains form a six helix bundle complex, which is hypothesized to provide the energy necessary for membrane merger. (21). The crystal structures of the prefusion form of PIV5 F (22) and of the postfusion forms of Newcastle Disease Virus (NDV) and human parainfluenza virus 3 (HPIV3) F (23–25) have been solved, greatly contributing to understanding of the refolding events that take place in the ectodomain of paramyxovirus fusion proteins. However, the crystal structures lack the membrane-interacting TM domain and the CT of F, two regions that have also been shown to play important roles in the fusion process (2, 26, 27). To aid in our investigation of the function and intracellular trafficking of Hendra F and G, we added a HA epitope tag (YPYDVPDYA) to the CT of Hendra F and Hendra G, as no monoclonal antibodies were available. Surprisingly, addition of the HA tag to Hendra F significantly increased membrane fusion activity. To examine if extending the Hendra F CT stimulates fusion activity in a sequence independent manner, a Myc epitope tag was added (EQKLISEEDL). No increase in fusion was observed, suggesting a sequence-specific effect of the HA tag. Further mutational analysis demonstrated that the three tyrosine residues present in the HA tag are critical for the observed fusion stimulation. Finally, to determine if HA tag-stimulated fusion is a general effect, the HA tag was added to the CT tail of parainfluenza virus 5 (PIV5) F protein. Fusion stimulation was not observed, and overall protein expression was significantly reduced. These results indicate that the C-terminal end of the Hendra F CT modulates fusion in a sequence specific manner, with tyrosine residues in this region specifically enhancing membrane fusion activity.