Fig 1.
Construction, purification, and characterization of MOKV G1-436*.
A) Bar diagram showing the organization of MOKV G and G1-436*. The N-glycosylation sites are indicated as well as the two fusion loops in G and the hydrophilic sequences that replace them in G1-436*. TM: transmembrane domain; IV: intraviral domain; ENK: enterokinase cleavage site; STREP: StrepTag. B) Molecular mass of MOKV G1-436* determined by SEC-MALS. C) Electron microscopy on negatively stained MOKV G1-436* at pH 7.5 and 5.5 and VSV Gth at pH 5.5. At pH 7.5, no defined molecular shape of MOKV G1-436* can be identified. At pH 5.5, elongated structures can be distinguished as shown in the gallery below the micrograph. These structures are thinner than those observed with the post-fusion Gth trimers. Scale bar: 100 nm.
Fig 2.
Structure of MOKV G1-436* compared to the post-fusion conformation of VSV and CHAV Gth and to the late intermediate (LI) conformation of CHAV Gth.
A) Left: Ribbon representation of the structure of MOKV G1-436*. Rainbow coloring from blue to red indicates the N- to C-terminal position of the residues in the model. Right: electrostatic surface potentials distribution on the solvent-accessible surface of MOKV G1-436* generated using the APBS plugin in Pymol software [55]. The acidic side of the TrD helix is surrounded by a black ellipse. B) Ribbon representation of the structures of MOKV G1-436*, of a protomer of the trimeric post-fusion conformations of VSV and CHAV Gth, and of the monomeric CHAV Gth LI colored by domains following the same color code as for MOKV G1-436*. In A) and B), the glycoproteins are aligned on their fusion domain. C) Structure of TrD and segments R4 and R5 in MOKV G1-436* in a protomer of the trimeric post-fusion conformation of VSV and CHAV Gth and in monomeric CHAV Gth LI. The arrows indicate the lateral helix that is present in the post-fusion conformation but not in LI conformation. D) Close-up view on the triad D143, Y145, and H397 of MOKV G1-436* that locks the hairpin-shaped post-fusion conformation and on their counterpart in the post-fusion protomer of VSV Gth and in CHAV Gth LI.
Table 1.
Data collections and refinement statistics.
Fig 3.
Comparison of the structure of MOKV G domains with those of VSV and CHAV G.
A) Comparison of FDs of MOKV, VSV, and CHAV G. Blue arrows indicate additional segments present in all lyssaviruses G. The dotted line represents the missing segment in MOKV G structure. Green arrowheads indicate the disulfide bridge present in vesiculoviruses G but not in lyssaviruses G. B) Comparison of TrDs and R4 segment of MOKV, VSV, and CHAV G. Note that MOKV G helix is bent, whereas it is rather straight for CHAV G and VSV G. The green arrowhead indicates the extra disulfide bridge of lyssaviruses G. C) Comparison of PHDs of MOKV, VSV, and CHAV G.
Table 2.
Domains and segments nomenclature used in the text.
Fig 4.
Electron microscopy of MOKV G1-436WT incubated with liposomes at pH 5.5.
A) MOKV G1-436WT is inserted in liposomes in a typical trimeric post-fusion conformation. The sample was stained with sodium phosphotungstate (NaPT). B) MOKV G1-436WT forms a network favoring the formation of tubular structures. Although these proteolipidic tubes are also seen after NaPT staining, they are more easily observed with uranyl acetate staining that has been used for the micrograph presented in this panel. C) Cryo-EM micrograph of tubular structures formed by MOKV G1-436WT when incubated with liposomes at pH 5.5 and corresponding power spectra (diffraction pattern).
Fig 5.
Electron microscopy of VSVΔG-eGFP pseudotyped either by MOKV G WT or MOKV G*.
A) At pH 7.5, VSV-MOKV G WT exhibits an 8-nm-dense layer of glycoproteins at its surface. B) At pH 5.5, some VSV-MOKV G WT virions exhibit 12-nm-long individualized spikes corresponding to the post-fusion trimer that stands perpendicular to the membrane (inset 1). At the surface of some other virions, MOKV G WT appears to form a fuzzy heterogeneous layer in which some spikes have the typical post-fusion shape (arrowheads), while some others are thinner and oblique to the membrane (arrows), most probably corresponding to elongated monomers (inset 2). A regular network of G can also be observed (bottom right micrograph). C) At pH 7.5, at the surface of VSV-MOKV G*, the layer of MOKV G* around the virion is less defined than at the surface of VSV-MOKV G WT. D) At pH 5.5, post-fusion trimers are observed at the surface of VSV-MOKV G* (inset and bottom left micrographs). As with VSV-MOKV G WT, a regular network of MOKV G* can be observed at the surface of some disrupted virions (bottom right micrograph).
Fig 6.
Antigenic sites or RABV positioned on MOKV G structure.
A) Location of major antigenic sites II and III, minor antigenic site a, and low pH antigenic site. Residues that are found substituted in mutants escaping neutralization by MAbs are indicated by their number and represented in ball and sticks form. They are colored in magenta for site II, in brown for site III, in cyan for minor site a, and in green for low pH antigenic site. Proline 340, which separates site III and minor site a, is also indicated. B) Model of MOKV G pre-fusion generated by superimposing MOKV G PHD, TrD, and FD domains on the corresponding domains of VSV G. The location of major antigenic sites II and III are indicated on the model. C) Surface representation of the modeled MOKV G pre-fusion state. Residues that are found substituted in mutants escaping neutralization by MAbs are indicated by their number and colored with the same color code as in A. The epitope of Mab 17D2 is colored in pink. Segment R4, which is part of the epitope and of which the pre-fusion conformation is unknown, is indicated by a pink dotted line. D) Surface representation of VSV G and modeled MOKV G pre-fusion states. The footprint of the LDL-R CR2 domain on VSV G pre-fusion conformation is in black (with key residues for binding indicated by their number). The location of residues 330 and 333, required for RABV penetration into neurons of adult mice, are also indicated in black at the surface of MOKV G. Both molecules have the same orientation.
Fig 7.
pH-sensitive molecular switches in MOKV G.
A) Position of acidic residues on the TrD long helix. B) Lateral view of the TrD and segments R4 and R5 showing the interaction between D275 and H384 stabilizing the post-fusion state of MOKV G. C) Close up view on the pair of acidic residues (D211 in PHD and E269 in R4) facing each other in the post-fusion conformation of MOKV G and on the environment of H261 (in R4), which is sandwiched by D266 (in R4) and E45 (in PHD). D) View of the TrD long helix showing the disposition of hydrophobic (in cyan) and acidic (in yellow) residues. Residues D263, E267, E274, E281, D285, and E288 are aligned on the same side of the helix of which the top view reveals the amphiphilic character. On the right, the top view indicates that the hydrophobic residues are buried in the structure and contribute to its overall stability. E) Model of the arrangement of the helices at the trimeric interface. The helices were superimposed on a GCNt coiled-coil [35, 36]. In this model, the acidic residues that are aligned on the helices point toward the trimer axis.