Characterization of human tibrovirus envelope glycoproteins

Abstract

Tibroviruses are novel rhabdoviruses detected in humans, cattle, and arthropods. Four tibroviruses are known to infect humans: Bas-Congo virus (BASV), Ekpoma virus 1 (EKV-1), Ekpoma virus 2, and Mundri virus. However, since none of them has been isolated, their biological properties are largely unknown. We aimed to characterize the human tibrovirus glycoprotein (G), which likely plays a pivotal role in viral tropism and pathogenicity. Human tibrovirus Gs were found to share some primary structures and display 14 conserved cysteine residues, although their overall amino acid homology was low (29%-48%). Multiple potential glycosylation sites were found on the G molecules, and endoglycosidase H- and peptide-N-glycosidase F-sensitive glycosylation was confirmed. AlphaFold-predicted three-dimensional (3D) structures of human tibrovirus Gs were overall similar. Membrane fusion mediated by these tibrovirus Gs was induced by acidic pH. The low pH-induced conformational change that triggers fusion was reversible. Virus-like particles (VLPs) were produced by transient expression of Gs in cultured cells and used to produce mouse antisera. Using vesicular stomatitis Indiana virus pseudotyped with Gs, we found that the antisera to the respective tibrovirus VLPs showed limited cross-neutralizing activity. It was also found that human C-type lectins and T-cell immunoglobulin mucin 1 acted as attachment factors for G-mediated entry into cells. Interestingly, BASV-G showed the highest ability to utilize these molecules. The viruses infected a wide range of cell lines with preferential tropism for human-derived cells whereas the preference of EKV-1 was unique compared with the other human tibroviruses. These findings provide fundamental information to understand the biological properties of the human tibroviruses.

Importance: Human tibroviruses are poorly characterized emerging rhabdoviruses associated with either asymptomatic infection or severe disease with a case fatality rate as high as 60% in humans. However, the extent and burden of human infection as well as factors behind differences in infection outcomes are largely unknown. In this study, we characterized human tibrovirus glycoproteins, which play a key role in virus-host interactions, mainly focusing on their structural and antigenic differences and cellular tropism. Our results provide critical information for understanding the biological properties of these novel viruses and for developing appropriate preparedness interventions such as diagnostic tools, vaccines, and effective therapies.

Keywords: Bas-Congo virus; Ekpoma virus 1; Ekpoma virus 2; Mundri virus; Tibrogargan virus; glycoprotein; tibrovirus.

Fig 1

Multiple alignment of human tibrovirus-G, TIBV-G, and VSIV-G sequences. A. VSIV-G domain map is depicted. Green: DI (lateral domain); yellow: DII (trimerization domain); red: DIII (pleckstrin homology domain); purple: DIV (fusion domain). Animo acid numbering is based on VSIV without its signal peptide. B. Complete amino acid sequences of BASV (#JX297815.1), MUNV (#OM320812.1), EKV-1 (#NC_038282.1), EKV-2 (#NC_038283.1), TIBV (#NC_020804.1), and VSIV (#NP_041715.1) were aligned using the Clustal Omega multiple-sequence alignment tool from EMBL-EB (https://www.ebi.ac.uk/Tools/msa/clustalo/), and the output was edited using Jalview 2.11.2.7. Transmembrane domains and signal peptides were predicted by DeepTMHMM (https://dtu.biolib.com/DeepTMHMM) and SOSUI (https://harrier.nagahama-i-bio.ac.jp/sosui/mobile/). N-Terminal signal peptides, bipartite fusion loop motifs, and transmembrane domains are framed in green, black, and red, respectively. The percent identity is indicated by shades of dark-blue, medium-blue, light-blue, and none, which indicate 80%–100%, 60%–80%, 40%–60%, and <40%, respectively. NetNGlyc-1.0 and NetOGlyc-4.0 were used to predict potential N-glycosylation and O-glycosylation sites, which are framed in yellow and brown, respectively. Conserved cysteine residues are indicated by pink asterisks and labeled as previously described (18), while the additional highly conserved cysteine residue compared with most animal rhabdoviruses is framed in pink. Pink daggers (†) represent cysteine residues shared among VSIV, MUNV, EKV-2, and TIBV. The conserved rhabdovirus N-glycosylation site (19) is marked with a blue asterisk on VSIV-G. Each domain of G is indicated as colored bars above the sequences.

Fig 2

3D structure and domain organization of VSIV and tibrovirus Gs. Both prefusion (A) and postfusion structures (B) are represented with monomers colored by domains. VSIV-G was used as a custom template for tibrovirus G modeling with Alphafold2-multimer as described in Materials and Methods. A VSIV-G domain map is depicted below panels. Green: DI (lateral domain); yellow: DII (trimerization domain); red: DIII (pleckstrin homology domain); purple: DIV (fusion domain). Animo acid numbering is based on VSIV without its signal peptide.

Fig 3

Purification of VLPs and G incorporation into VLPs and pseudotyped VSIVs. VLPs were produced from Expi293F cells transfected with pCAGGS encoding tibrovirus Gs and VSIV G with hemagglutinin (HA) tag. The culture supernatants were fractioned through 20%–50% sucrose gradient centrifugation as described in Materials and Methods. Each fraction of BASV-G (A), MUNV-G (B), EKV-1-G (C), EKV-2-G (D), TIBV-G (E), and VSIV-G (F) was analyzed by western blotting. Fractions 5–8 for each tibrovirus G were pooled, layered onto 20% sucrose cushions, and sedimented to obtain purified VLPs. Panels G and I show G incorporation into purified VLPs and G expression in cell lysates, respectively. G incorporation into VSIV particles and G expression in cell lysates were also analyzed in western blotting (H and J, respectively).

Fig 4

Transmission electron microscopy (EM) of VLPs. VLPs were produced from Expi293F cells transfected with pCAGGS expressing BASV-G (A, B, and H–K), VSIV-G (C and L), MUNV-G (D and M), EKV-1-G (E and N), EKV-2-G (F and O), or TIBV-G (G and P). Culture supernatants were harvested 96 hours after transfection, subjected to sucrose gradient purification (A) or freshly prepared for EM analyses (B–P), fixed, and stained as described in Materials and Methods. Panel A shows the typical shape of BASV-G VLPs. The boxed area at the lower right corner of panel A shows a higher magnification image of a BASV-G VLP. In panels H–P, an anti-BASV-G monoclonal antibody was used for immunogold staining. The arrowheads indicate gold particles. The scale bar in panels B–I indicate 100 nm.

Fig 5

Enzymatic deglycosylation of tibrovirus and VSIV Gs. Concentrated particles of VSIV pseudotyped with HA-tagged BASV-G, MUNV-G, EKV-1-G, EKV-2-G, TIBV-G, and VSIV-G were undigested or digested with Endo H or PNGase F and analyzed in western blotting as described in Materials and Methods.

Fig 6

Fusion activity of tibrovirus Gs at different pH. Panel A shows pH-dependent cell-to-cell fusion. Vero E6-EGFP-HiBiT cells were transfected with pCAGGS expressing BASV-G, MUNV-G, EKV-1-G, EKV-2-G, TIBV-G, VSIV-G, and influenza virus HA. Then, Vero E6-mCherry-LgBiT cells were added and allowed to fuse with Vero E6-EGFP-HiBiT cells at the indicated pH. Nano luciferase activity was measured using a GloMax Discover Microplate Reader. The luminescence values of MOCK-transfected cells were deducted from those of glycoprotein-transfected cells. Representative data from a total of four independent experiments are shown. Panel B shows the reversible pH-dependency of tibrovirus G-mediated entry into cells. VSVΔG*BASV-G, VSVΔG*MUNV-G, VSVΔG*EKV-1-G, VSVΔG*EKV-2-G, VSVΔG*TIBV-G, VSVΔG*VSIV-G, and VSVΔG*HA/NA were exposed to the indicated pH for 30 minutes before neutralization and inoculation to Huh-7 cells. At 20 hours postinoculation, GFP-expressing cells were counted, and the percentage of infected cells was determined based on the GFP count in the cells infected with neutral pH-treated virus. The experiment was performed in triplicate, and averages and standard deviations are shown.

Fig 7

Neutralizing activities of VLP-immunized mouse antisera. Fivefold dilutions (1:100–1:1,562,500) of mouse antisera to BASV-G (A), MUNV-G (B), EKV-1-G (C), EKV-2-G (D), TIBV-G (E), and EBOV GP (F) were tested for neutralizing antibodies to VSVΔG*BASV-G, VSVΔG*MUNV-G, VSVΔG*EKV-1-G, VSVΔG*EKV-2-G, VSVΔG*TIBV-G, and VSVΔG*EBOV-GP. Three mice were used for each virus, and averages and standard deviations of at least four independent experiments for each mouse are shown.

Fig 8

Infectivity of VSIVs pseudotyped with tibrovirus Gs in cell lines expressing viral attachment factors. VSVΔG*BASV-G, VSVΔG*MUNV-G, VSVΔG*EKV-1-G, VSVΔG*EKV-2-G, VSVΔG*TIBV-G, VSVΔG*EBOV-GP, and VSVΔG*VSIV-G were inoculated onto control K562 cells and K562 cells expressing DC-SIGN (A) and hMGL (B), respectively. These viruses were also inoculated onto control human embryonic kidney (HEK) 293T cells and HEK293T cells expressing human T-cell immunoglobulin mucin 1 (hTIM-1) (C). The relative percentages of infectivity were calculated by setting the number of infected control cells to 100%. Each experiment was performed four times, and the averages and standard deviations are shown. Student’s t-test was used to compare relative infectivity among viruses. *P < 0.05, **P < 0.01, and ***P < 0.001; † means significantly higher than all the other viruses [P < 0.05 in panel B; P < 0.001 in panel C, except VSVΔG*MUNV-G (P < 0.05)].

Fig 9

Infectivity of VSIVs pseudotyped with tibrovirus Gs in various cell lines. VSVΔG*BASV-G, VSVΔG*MUNV-G, VSVΔG*EKV-1-G, VSVΔG*EKV-2-G, VSVΔG*TIBV-G, VSVΔG*EBOV-GP, and VSVΔG*VSIV-G were inoculated onto 15 mammalian cell lines derived from various species. Panel A shows IU of each virus in the cell lines while panel B shows relative infectivity after setting each IU value in VeroE6 cells to 1. Each experiment was performed four times. Averages and standard deviations are shown.