ABSTRACT
Alphaviruses are serious, sometimes lethal human pathogens that belong to the family Togaviridae. The structures of human Venezuelan equine encephalitis virus (VEEV), an alphavirus, in complex with two strongly neutralizing antibody Fab fragments (F5 and 3B4C-4) have been determined using a combination of cryo-electron microscopy and homology modeling. We characterize these monoclonal antibody Fab fragments, which are known to abrogate VEEV infectivity by binding to the E2 (envelope) surface glycoprotein. Both of these antibody Fab fragments cross-link the surface E2 glycoproteins and therefore probably inhibit infectivity by blocking the conformational changes that are required for making the virus fusogenic. The F5 Fab fragment cross-links E2 proteins within one trimeric spike, whereas the 3B4C-4 Fab fragment cross-links E2 proteins from neighboring spikes. Furthermore, F5 probably blocks the receptor-binding site, whereas 3B4C-4 sterically hinders the exposure of the fusion loop at the end of the E2 B-domain.
IMPORTANCE Alphaviral infections are transmitted mainly by mosquitoes. Venezuelan equine encephalitis virus (VEEV) is an alphavirus with a wide distribution across the globe. No effective vaccines exist for alphaviral infections. Therefore, a better understanding of VEEV and its associated neutralizing antibodies will help with the development of effective drugs and vaccines.
INTRODUCTION
Venezuelan equine encephalitis virus (VEEV) is a lipid-enveloped virus belonging to the alphavirus family. VEEV is an emerging disease threat with outbreaks first occurring in the 1930s in South America (1). Large epizootics/epidemics of VEEV have spread into Central America, reaching the southern parts of the United States (2). VEEV has also been developed as a biological weapon (3). Some alphaviruses, such as Sindbis virus (SINV), generally cause only mild disease symptoms in humans. However, VEEV and Chikungunya virus (CHIKV) can be lethal or permanently disabling (4, 5). Furthermore, there are currently no treatments or vaccines for alphaviral infections.
The structures of a number of alphaviruses have been determined to resolutions better than 10 Å, including the structure of VEEV, which has been determined to a 4.8-Å resolution (6–9). Alphaviruses are icosahedral with quasi-T=4 symmetry and contain a positive-sense single-stranded RNA (ssRNA) genome. A single virus particle contains 240 copies each of E1 and E2, forming 20 “i3” spikes situated about the icosahedral 3-fold axes and 60 “q3” spikes at general positions. The capsid protein is arranged as 12 pentamers about the 5-fold vertices and 30 hexamers about the icosahedral 2-fold vertices. The nucleocapsid core is completely surrounded by a lipid envelope, derived from the host membrane, into which are embedded the icosahedral array of glycoproteins (10). The external diameter of alphaviruses is ∼700 Å, with the capsid containing a copy of a single-stranded RNA genome that is about 11 kb in length and codes for nine proteins. The four nonstructural proteins (nsP1 to nsP4) are coded at the 5′ end. The 3′ end of the genome is transcribed for subsequent translation into a polyprotein precursor containing the three structural proteins PE2 (the precursor of E3 and E2), E1, and capsid. E1 consists of 442 amino acids that form three β-barrel domains (I, II, and III). E2 consists of 423 amino acids arranged into three immunoglobulin-like domains: A, B, and C. Domain A contains the receptor binding site (11), domain B is at the distal end of each spike protecting the fusion loop on DII of E1, and domain C is situated closest to the viral membrane (11). Domain B is connected to domains A and C by a β-ribbon (Fig. 1). The crystal structures of the E1 fusion homotrimer from Semliki Forest virus and the E1E2 heterodimers from CHIKV and SINV have been determined to near-atomic resolution (12–14).
FIG 1.
Structure of the E1E2 heterodimer from CHIKV (PDB 3N42). The E1 domains I, II, and II are red, yellow, and blue, respectively. The E2 domains A, B, and C are cyan, green, and magenta, respectively. The E2 β-ribbon connector is brown, and the E1 fusion loop is orange.
The E1 glycoprotein contains a hydrophobic fusion loop that is responsible for membrane fusion with an endosomal membrane. The E2 protein can bind to cellular receptors and also protects the E1 fusion loop at neutral pH. During the initial stages of infection, the host cellular receptor is recognized by the surface glycoprotein E2 (15, 16). Once the virus has been internalized, the low pH of the endosome causes the virion to undergo an irreversible conformational change, resulting in the disassociation of E2 and the formation of E1 trimers (13, 14). Upon E2 disassociation, the fusion loop then binds the endosomal membrane, followed by fusion of the viral and endosomal membranes (11). Finally, the viral genome is released into the host cytoplasm, where replication of new viral particles can begin.
Alphavirus-neutralizing monoclonal antibodies (MAbs) bind both the E1 and E2 proteins (17–20). However, in general, neutralizing anti-E2 MAbs are more potent than E1 directed MAbs (17, 20–23). For instance, in a study of CHIKV neutralizing antibodies (7), Fab fragments of CHIKV-152 were shown to bind to the E2 B domain, hindering the exposure of the fusion loop by stabilizing the viral surface, whereas Fab fragments of m10, m242, and CHK9 were found to surround the putative receptor attachment site on the E2 A domain, probably blocking attachment to host cells.
F5 is a human monoclonal antibody (hMAb) that was isolated using phage display from human bone marrow donors (17). The F5 antibody was shown to provide both pre- and postexposure protection in mice by passive transfer of this hMAb (17). The epitope of F5 has been mapped to amino acids 115 to 119 in the E2 A domain by sequencing the genes of the VEEV structural proteins to determine the position of neutralization escape mutations.
The 3B4C-4 MAb was obtained by immunization of mice with TC-83 virus and was humanized using combinatorial antibody libraries and phage display technology (18, 19, 22, 24–26). This antibody was found to be a member of a class of MAbs that bind to the tip of the E2 B domain, between residues Ser182 and Ile207 (24). Preincubation of 3B4C-4 with TC-83 blocked 90% of virus attachment to Vero cells, suggesting that the E2 B domain is also involved in cellular attachment. In addition, 3B4C-4 can neutralize infectivity after attachment to cells. Fabs of this antibody showed reduced neutralization activity in vitro and in vivo (25–27). Also, a mutation of residue Ile207 was shown to reduce replication in Aedes albopictus-derived C636 cells, implicating this region of the B domain as critical for viral infectivity (24, 28).
We report here cryo-electron microscopy (cryo-EM) structures of the attenuated VEEV TC-83 strain in complex with Fab fragments of 3B4C-4 and F5. Structures of the Fab fragments were obtained by homology modeling. Fitting of the E-protein and Fab into the cryo-EM density of the complexes suggested that neutralization of VEEV can be achieved by either cross-linking the E2 protein to inhibit the conformational changes that are required to make the virus fusogenic or to block attachment. Whereas F5 Fab fragments cross-link the E2 molecules within a spike, 3B4C-4 Fab fragments cross-link the E2 molecules in neighboring spikes.
MATERIALS AND METHODS
The attenuated VEEV TC-83 strain was derived by serial passage of the equine-virulent, epizootic Trinidad donkey strain in fetal guinea pig cells (29). Attenuation of VEEV resulted in three specific point mutations that greatly limited the viral infectivity. The most significant impact on infectivity is the point mutation of Arg120 to Thr, located on domain A of glycoprotein E2.
VEEV TC-83 propagation and purification.
BHK-15 cells were maintained in minimal essential medium and further supplemented with 3% fetal bovine serum at 37°C in the presence of 5% CO2. TC-83, an attenuated VEEV, was then added at a multiplicity of infection of 2 and allowed to attach to the cells while rocking for 1 h at room temperature. Cell debris was then removed by centrifugation at 10,000 rpm, and the virus was purified by the addition of 2 ml of a 30% sucrose cushion, followed by centrifugation at 38,000 rpm in a Ti 50.2 rotor corresponding at a force of 45,000 × g. The final step involved purification of the virus using an OptiPrep density gradient in a SW41 rotor at 38,000 rpm, followed by buffer exchange into 1× NTE buffer (0.1 M EDTA, 100 mM NaCl, 50 mM Tris-HCl) at pH 7.6.
Antibody production and purification.
Procedures for producing MAbs F5 and 3B4C-4 have been previously reported (21, 22). Fab fragments were purified from the IgG after papain digestion using a protein A column and then further purified over a Superdex 200 16/600 column.
Complex formation, cryo-EM data collection, and single-particle reconstructions.
VEEV TC-83 particles were incubated with Fabs at room temperature for 1 h and pH 7.6 at a Fab/E2 protein molar ratio of 3:1. For the prefusion complexes, the pH was subsequently lowered to 6.0 by addition of a 1/10 volume dilution with 1 M sodium citrate (pH 5.5). Then, 3-μl aliquots of each complex were flash frozen in liquid ethane on 200 mesh Cu Quantifoil grids by hand blotting. Micrographs of the frozen complexes were recorded on Kodak SO-163 films using a FEI Titan Krios electron microscope operating at 300 kV. Images were collected at a magnification of 59,000 with an electron dose of ∼25 e−/Å2. Micrographs were then digitized using a Nikon scanner with a 6.35-μm-per-pixel step size. Particles were boxed using e2boxer in the EMAN2 software package (30). The images were corrected for the contrast transfer function using CTFIT from the EMAN package (31). To initiate the orientation search, a cryo-EM reconstruction of native CHIKV-VLP was used, followed by iterative rounds of refinement. The numbers of particles used were 1,389, 1,091, 1,120, and 1,456 for Fab complexes F5 (pH 7.6 and 6.0) and 3B4C-4 (pH 7.6 and 6.0), giving final resolutions of 17, 18.3, 17.2, and 17.5 Å, respectively, as determined by calculation of the Fourier shell correlation with a 0.5 threshold, respectively.
Homology models of Fabs F5 and 3B4C-4.
By using the target sequences from Fabs F5 and 3B4C-4 (F5 [GenBank accession numbers HM047070 and HM047071] and 3B4C-4 [GenBank accession numbers DQ487205 and DQ487206]), homology models were determined from the Swiss-Model server (32, 33). The coordinates from the crystal structure of a Fab fragment to human interleukin-2 with the solvent atoms removed was used as a template (PDB 1F8T) and was aligned to each respective Fab sequence to fill in any possible gaps.
Structural analysis of Fab binding.
The crystal structure of the E1E2 heterodimer (PDB 3N42) from CHIKV was fitted into the cryo-EM map of the complexes assuming T=4 quasi-symmetry using the program EMfit (34, 35). Subsequently, the fitting was refined by independent fitting of the E1 and E2 domains without imposing T=4 quasi-symmetry. Glycoprotein E1 was split into domain I (residues 1 to 36, 132 to 168, and 273 to 293), domain II (residues 37 to 131 and 169 to 272), and domain III residues 294 to 393). E2 was divided into domain A (residues 16 to 134), domain B (residues 173 to 231), domain C (residues 269 to 342), and the β-ribbon connector (residues 7 to 15, 135 to 172, and 232 to 268) (Fig. 1). While fitting each independent domain, the N terminus of each domain was restrained to be within 3.8 Å from the C terminus of the previous domain. Similarly, the C termini were restrained to be within 3.8 Å of the N terminus of the next domain (Table 1).
TABLE 1.
Fitting of E glycoprotein and Fab structures into cryo-EM reconstructionsa
| Fab fragment | pH | sumf | % |
Angle (°) |
Position (Å) |
|||||
|---|---|---|---|---|---|---|---|---|---|---|
| Clash | −Den | θ1 | θ2 | θ3 | cx | cy | cz | |||
| F5 | ||||||||||
| E1E2 | 7.6 | 28.8 | 0.8 | 2.8 | 184.5 | 80.2 | 175.0 | 21.9 | 88.4 | 277.3 |
| 6.0 | 30.1 | 1.1 | 3.1 | 184.1 | 80.7 | 174.6 | 22.3 | 87.9 | 276.4 | |
| Fab | 7.6 | 12.3 | 32.1 | 51.8 | 200.0 | 10.3 | 200.0 | 22.7 | 94.3 | 337.8 |
| 6.0 | 11.4 | 31.5 | 52.3 | 201.2 | 10.9 | 200.7 | 22.5 | 95.2 | 337.1 | |
| 3B4C-4 | ||||||||||
| E1E2 | 7.6 | 37.6 | 0.5 | 3.5 | 185.0 | 80.5 | 178.0 | 23.6 | 88.7 | 275.7 |
| 6.0 | 38.4 | 0.8 | 3.3 | 185.8 | 81.0 | 179.2 | 23.9 | 89.2 | 274.1 | |
| Fab | 7.6 | 40.2 | 0.0 | 2.2 | 98.0 | 359.3 | 260.8 | 14.3 | 103.2 | 294.8 |
| 6.0 | 39.1 | 0.0 | 2.6 | 97.1 | 360.0 | 261.4 | 14.0 | 101.4 | 294.3 | |
sumf, average density for all atomic positions normalized to the highest density in the map (set to 100); Clash, the percentage of clashes between symmetry-related atoms; −Den, the percentage of atoms positioned outside the density. θ1, θ2, and θ3 are the Eulerian angles that rotate the coordinates from their initial to fitted positions. cx, cy, and cz (Å) are the final center positions of the molecules after fitting.
The Fab homology models were initially fitted as a rigid body assuming T=4 symmetry. However, the F5 binds close to the 3-fold symmetry axis of each glycoprotein spike. Thus, only one F5 Fab could bind to any one spike at a time, excluding binding of other Fab molecules to the two other E2 molecules in the same spike. Hence, the observed density is an average of the Fab bound to each of the three E2 A domains, with one-third occupancy at each binding site. To achieve this with the EMfit program, the clash function was turned off. As a result simultaneous fitting was obtained at all three sites with one-third occupancy (Table 1).
Accession numbers.
The homology models for Fabs F5 and 3B4C-4 have been deposited with the Protein Data Bank (www.pdb.org) (PDB codes 4uok and 4uom, respectively); cryo-EM maps of VEEV in complex with Fabs F5 and 3B4C-4 have been deposited with the Electron Microscopy Data Bank (www.emdatabank.org) under EMB codes EMD-2645 and EMD-2655, respectively.
RESULTS AND DISCUSSION
The 3B4C-4 antibody was isolated and characterized after immunization of mice with TC-83, whereas the F5 antibody was characterized as one of 11 VEEV-specific human Fabs isolated from human bone marrow donors (F5) (17–19). By using a competitive enzyme-linked immunosorbent assay against MAbs and antibody escape variants, it was shown that these antibodies recognize epitopes primarily on the A and B domains of E2. Two of these antibodies, MAbs F5 and 3B4C-4, as well their Fab fragments, had previously been shown to have a strong neutralizing activity against VEEV infection (21, 22). In agreement with the earlier determination of their epitopes on VEEV, we show that Fab fragments F5 and 3B4C-4 bind epitopes on domains A and B consistent with earlier mapping results showing sites on domains A and domain B, respectively (17, 24).
The structures of VEEV TC-83 complexed with neutralizing Fab fragments of F5 and 3B4C-4 have been determined at both pH 7.6 and pH 6.0, using a combination of cryo-EM and X-ray crystallography. The cryo-EM maps had a resolution of ∼17 Å (Fig. 2). When the VEEV-TC-83 particles were placed in a pH 6.0 environment, the particles fused and aggregated. However, when the virus was complexed with either of these Fabs at neutral pH and then placed into a pH 6.0 environment, there was no evidence for aggregation, suggesting that both of these Fabs prevented the exposure of the fusion loops. Also, the positions of the E glycoproteins and Fab fragments did not alter significantly compared to the pH 7.6 complex.
FIG 2.
Cryo-EM reconstructions of VEEV in complex with Fab fragments of neutralizing antibodies. Fab density is represented in red. (a) Overview of the VEEV-F5 reconstruction looking down a 2-fold axis. (b) Closeup view of the asymmetric unit of VEEV-F5 indicated by a black triangle. (c) Overview of VEEV–3B4C-4 reconstruction, looking down a 2-fold axis. (d) Closeup view of the icosahedral asymmetric unit of the VEEV–3B4C-4 complex indicated by a black triangle.
The MAb F5 Fab fragment complexed with VEEV.
The F5 Fab homology model was fitted into the cryo-EM difference map. The contact region between the F5 Fab and the virus is formed by residues 27 and 28 on the light-chain complementarity determining region 1 (CDRL1) and amino acids 52, 53, 56 and 57 on the heavy-chain complementary determining region 2 (CDRH2) with residues 73, 75, 80, 81, 115, 117, and 120 on domain A of one E2 molecule and the same region on domain A of a neighboring molecule (Fig. 3 and Table 2). From the heavy-chain complementary determining region 3 (CDRH3), only residue 98 makes contacts with Ser118 of E2. Since the A domains of the three E2 molecules within one spike are situated near the spike's 3-fold axis, there would be steric clashes if there were more than one Fab molecule bound to the same spike (Fig. 3a and 4a). This is also evident from the density height of the Fab molecules, which was only about one-third of the density of the virus capsid (Table 3). Since, on average, each of the three E2 molecules within a spike binds a Fab fragment one-third of the time, the fitting process was performed by turning off the clash penalty but maintaining the 3-fold symmetry of the spike. The resultant fit showed the F5 Fab fragments at the top of the i3 and q3 spikes, extending each spike in a roughly radial direction. The base of the Fab structure that contacts the virus was resolved for each of the three different Fab locations.
FIG 3.
Roadmaps of E1E2 with Fab binding sites. (a) Roadmap of E1E2 fit to the VEEV-F5 reconstruction. (b) Roadmap of E1E2 fit to the VEEV–3B4C-4 reconstruction. The white contours represent the Fab binding positions determined from their difference maps. The black triangle denotes the boundaries of the asymmetric unit. The residues colored white represent the locations on E2 where the Fab binds. The colors represent the distance from the center of the virus as defined in the color scale bar.
TABLE 2.
Contacts between Fab F5 or Fab 3B4C-4 and glycoprotein E2
| Fab fragment | Fab residuea | E2 residue | Fab CDRb loop |
|---|---|---|---|
| F5 | Ser 27 (L) | Lys 75 | CDRL1 |
| Thr 73 | |||
| Asn 28 (L) | Lys 75 | CDRL1 | |
| Tyr 33 (H) | Lys 115 | CDRH1 | |
| Thr 52 (H) | Gln 81 | CDRH2 | |
| Tyr 53 (H) | Lys 115 | CDRH2 | |
| Arg 120 | |||
| Ser 56 (H) | Gln 81 | CDRH2 | |
| His 80 | CDRH2 | ||
| Thr 57 (H) | His 80 | CDRH2 | |
| Tyr 98 (H) | Ser 118 | CDRH3 | |
| 3B4C-4c | Ser 27 (L) | Lys 223 | CDRL1 |
| Ser 180 | |||
| Leu 181 | |||
| Asn 28 (L) | Ser 177 | CDRL1 | |
| Val 179 | |||
| Ser 180 | |||
| Phe 94 (L) | Ser 184 | CDRL3 | |
| Thr 52 (H) | Gln 216 | CDRH2 | |
| Tyr 53 (H) | Thr 214 | CDRH2 | |
| CD1 | Arg 188 | ||
| CD2 | Glu 185 |
“(H)” and “(L)” represent heavy and light chains, respectively.
CDR is the complementary determining region on the Fabs.
For Fab 3B4C-4, CD1 and CD2 are the Fab constant domains.
FIG 4.
Closeup of a single spike of VEEV F5 and 3B4C-4 complexes from a Roadmap projection. (a) Radial projection of a map from a single F5 Fab bound to its epitope. The white contours represent the Fab binding location. The white dashes enclose the E2 A (orange) and B (red) domains. (b) Radial projection of a difference map from the 3B4C Fab complex. The colors represent the distance from the center of the virus as defined in the color scale bar.
TABLE 3.
Average density heights at the Cα positions (sumf) upon fitting the individual domains from the CHIKV E1E2 heterodimer and associated Fab structures at four quasiequivalent positions
| Complex, pH, and fragment | Domain | Independent domain fitting |
T=4 fitting |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | Avg | 1 | 2 | 3 | 4 | Avg | ||
| F5 complex at pH 7.6 | |||||||||||
| E1 | I | 31.7 | 29.9 | 30.9 | 26.9 | 29.9 | 29.7 | 28.2 | 28.9 | 27.9 | 28.7 |
| II | 32.9 | 33.5 | 32.9 | 34.1 | 33.4 | 30.2 | 29.9 | 30.3 | 30.4 | 30.2 | |
| III | 31.6 | 32.2 | 30.9 | 31.7 | 31.6 | 28.5 | 29.2 | 29.0 | 28.8 | 28.9 | |
| E2 | A | 28.6 | 28.4 | 30.2 | 29.1 | 29.1 | 28.2 | 27.7 | 28.6 | 28.3 | 28.2 |
| B | 29.7 | 28.7 | 31.4 | 30.7 | 30.1 | 29.1 | 29.8 | 29.5 | 29.0 | 29.4 | |
| C | 31.0 | 28.7 | 28.9 | 28.6 | 29.3 | 28.4 | 27.9 | 27.6 | 28.3 | 28.1 | |
| Fab | Variable | 12.1 | 10.5 | 13.7 | 12.9 | 12.3 | 12.1 | 10.5 | 13.1 | 12.5 | 12.1 |
| Constant | 13.2 | 11.5 | 12.2 | 13.1 | 12.5 | 13.6 | 11.5 | 12.2 | 13.1 | 12.6 | |
| F5 complex at pH 6.0 | |||||||||||
| E1 | I | 33.9 | 21.7 | 36.2 | 32.4 | 31.1 | 31.8 | 32.5 | 31.4 | 31.9 | 31.9 |
| II | 31.9 | 32.4 | 30.7 | 30.5 | 31.4 | 30.2 | 30.9 | 31.1 | 30.5 | 30.7 | |
| III | 30.8 | 31.6 | 31.4 | 31.0 | 31.2 | 28.3 | 29.2 | 29.0 | 28.7 | 28.8 | |
| E2 | A | 30.5 | 28.2 | 29.1 | 30.2 | 29.5 | 29.1 | 28.4 | 28.6 | 28.2 | 28.6 |
| B | 32.1 | 30.8 | 31.5 | 31.2 | 31.4 | 31.2 | 31.0 | 30.8 | 31.6 | 31.2 | |
| C | 29.0 | 28.4 | 29.3 | 30.0 | 29.2 | 28.7 | 27.6 | 27.8 | 29.2 | 28.3 | |
| Fab | Variable | 11.0 | 11.5 | 9.8 | 10.5 | 10.7 | 10.8 | 11.5 | 9.7 | 10.5 | 10.6 |
| Constant | 12.9 | 12.0 | 10.7 | 11.9 | 11.9 | 12.8 | 12.6 | 10.7 | 11.9 | 12.0 | |
| 3B4C-4 complex at pH 7.6 | |||||||||||
| E1 | I | 37.4 | 38.3 | 37.9 | 36.3 | 37.5 | 35.6 | 36.2 | 36.0 | 36.5 | 36.1 |
| II | 35.2 | 36.7 | 35.6 | 36.0 | 35.9 | 33.1 | 32.7 | 32.2 | 32.8 | 32.7 | |
| III | 36.2 | 35.9 | 35.3 | 36.1 | 35.9 | 34.2 | 33.8 | 33.5 | 34.1 | 33.9 | |
| E2 | A | 35.0 | 36.4 | 34.3 | 35.7 | 35.4 | 33.5 | 34.2 | 33.8 | 34.0 | 33.9 |
| B | 39.2 | 38.4 | 39.0 | 38.7 | 38.8 | 35.8 | 36.2 | 35.6 | 35.5 | 35.8 | |
| C | 37.9 | 34.3 | 39.3 | 38.9 | 37.6 | 35.2 | 34.7 | 34.8 | 35.4 | 35.0 | |
| Fab | Variable | 35.2 | 36.3 | 35.8 | 35.0 | 35.6 | 35.2 | 36.3 | 35.8 | 35.0 | 35.6 |
| Constant | 38.1 | 33.0 | 35.2 | 32.9 | 34.8 | 38.1 | 33.0 | 35.2 | 32.9 | 34.8 | |
| 3B4C-4 complex at pH 6.0 | |||||||||||
| E1 | I | 38.1 | 38.7 | 38.2 | 37.4 | 38.1 | 36.8 | 37.0 | 37.5 | 38.2 | 37.4 |
| II | 36.6 | 35.9 | 36.2 | 36.0 | 36.2 | 35.5 | 34.2 | 35.1 | 34.9 | 34.9 | |
| III | 36.8 | 36.2 | 35.8 | 35.4 | 36.1 | 34.5 | 35.2 | 34.7 | 35.3 | 34.9 | |
| E2 | A | 37.6 | 37.8 | 37.0 | 38.3 | 37.7 | 36.2 | 35.4 | 35.8 | 36.1 | 35.9 |
| B | 38.8 | 39.3 | 39.0 | 40.1 | 39.3 | 37.2 | 37.9 | 38.0 | 38.2 | 37.8 | |
| C | 37.0 | 37.3 | 38.2 | 35.6 | 37.0 | 36.2 | 35.8 | 36.4 | 36.0 | 36.1 | |
| Fab | Variable | 37.2 | 36.8 | 37.3 | 36.9 | 37.1 | 37.2 | 36.8 | 37.3 | 36.9 | 37.1 |
| Constant | 36.1 | 37.8 | 36.5 | 36.0 | 36.6 | 36.1 | 37.8 | 36.5 | 36.0 | 36.6 | |
With only one Fab bound to a trimeric spike at a time there would remain two receptor-binding sites available for cellular interactions. However, when projecting the density of a single bound Fab molecule onto the viral surface, the density of the Fab's constant domain projects onto a part of the unoccupied receptor-binding sites of the other two E2 monomers in the same spike (Fig. 4a and 5). Thus, neutralization was achieved even with only one-third of the possible E2 Fab binding sites occupied. A similar observation has been made for a mouse-neutralizing antibody of SINV, where only one-third of the E1 molecules were bound with Fab, suggesting similar eclipsing of the receptor binding sites and perhaps also stabilization of the virus preventing essential conformational changes (20).
FIG 5.
Zoomed-in version of the roadmaps in the vicinity of the VEEV-F5 contact region. Footprint of a single F5 Fab projected radially onto the viral surface. Solid white contours represent the projection of the Fab density. Dashed white lines form the boundaries of the E2 A and B domains. Light and dark yellow residues represent contacts between the Fab and E2 glycoprotein. Dark yellow residues represent the predicted epitope from studies of antibody escape mutants. The colors represent the distance from the center of the virus as defined in the color scale bar.
When the pH of the complex was lowered to pH 6.0, the position of the E1E2 heterodimer and Fab fragment did not change by more than 2 Å and 2° as determined by the EMfit program. In previous studies, it has been shown that upon lowering the pH of the native virus, the E1E2 heterodimer undergoes a large conformational change in which E2 begins to disassociate from the spike before the formation of E1 homotrimers (13). However, the present results show that the bound F5 Fabs had stopped the detachment of E2 from the interior of the spike, which would be required for the virus to become fusogenic. However, no information is currently available about the difference in the binding affinity of the Fab fragments in different pH environments.
The quality of fit of the E1E2 heterodimer structure into the cryo-EM density was significantly worse in the F5 complex than in the 3B4C-4 complex (Table 1). This suggested that the F5 Fab might have induced some structural changes onto the E2 molecules in a spike. Consistent with this observation, difference maps between the antibody-bound and native virus show that the E2 A domains contract inward by ∼5 Å upon binding of the F5 Fab. This brings the A domains closer together, allowing a single Fab molecule to cross-link adjacent A domains within each spike. Furthermore, the conformational change caused by F5 binding might also alter the surface of the virus, providing an additional inhibitor for virus infectivity.
The MAb 3B4C-4 Fab fragment complexed with VEEV.
The complex of VEEV with 3B4C-4 Fab fragments showed that Fabs bind to their expected epitopes at the tip of domain B on E2 glycoproteins (Fig. 3b and 4b). The reconstruction showed that there was one Fab bound to each of the three E2 molecules per spike. The Fab fragments extended radially outwards toward neighboring spikes at an angle of ∼15° from the spike's 3-fold axis. The constant regions of the 3B4C-4 Fab fragments could make a salt bridge with neighboring Fabs on i3 and q3 spikes between residues Arg188 and Glu 185, respectively (Fig. 2 and Table 2). Both of these residues are highly conserved among human and mouse antibodies and are on the constant domain of the heavy chain. This links the five quasi-3-fold spikes to each other around each pentameric vertex and also links each icosahedral spike to the three surrounding quasi-3-fold spikes (Fig. 6). This “locks” the spikes, preventing the E2 glycoproteins from separating from the spikes and likely stopping the virus becoming fusogenic. In contrast to lowering the pH of the virus, when lowering the pH of the Fab 3B4C-4 complex, no significant change was seen either in the density of the virus or of the Fabs. This was also the case for the complex with F5. Therefore, the Fab fragments had prevented the conformational change that occurs when the virus is uninhibited by the presence of the Fab. The density heights of the Fab molecules of the complex at low pH are similar to those of the A and C domains of the E2 glycoproteins, indicating that each Fab position is fully occupied (Table 3).
FIG 6.
Zoomed-in version of the roadmaps in the vicinity of the VEEV–3B4C-4 contact region. The footprint of 3B4C-4 contact regions projected radially onto the viral surface. Solid white contours represent the projection of the Fab density. Dashed white lines form the boundaries of the E2 A and B domains. Light and dark yellow residues represent contacts between the Fab and E2 glycoprotein. Dark yellow residues represent the predicted epitope from studies of antibody escape mutants. The colors represent the distance from the center of the virus as defined in the color scale bar.
Conclusion.
In a previous crystallographic study of the E1E2 trimer in Sindbis virus there was no observable density of the B domain (13) at a low pH, indicating that the position of the B domain was highly flexible. Furthermore, in a crystallographic study of the CHIKV E1E2 heterodimer at neutral pH, the B domain has an average temperature factor greater than 100 Å2, compared to about 50 Å2 for the rest of the structure, indicating a high flexibility of the B domain (12). However, here we have shown that the cryo-EM density of the B domain in both the F5 and 3B4C-4 complexes with VEEV is about the same as the other domains of E1 and E2 (Table 3). It is therefore evident that the binding of Fab F5 or 3B4C-4 stabilizes the B domain. This keeps the B domain fixed over the fusion loop, preventing its exposure and subsequent viral fusion.
It is apparent that both F5 and 3B4C-4 neutralize the virus by attaching to and stabilizing the trimeric spikes, which results in blocking cellular attachment and exposure of the fusion loop, respectively. It has been previously shown that 3B4C-4 also blocks virus attachment to cellular receptors (27). Similar blocking studies with F5 have not yet been performed. The F5 Fab stabilizes the virus by cross-linking the E2 monomers within a spike, whereas 3B4C-4 Fab cross-links the E2 monomers between neighboring spikes. Furthermore, F5 uses the variable domains for performing the cross-linking, whereas 3B4C-4 uses the constant domains for cross-linking.
ACKNOWLEDGMENTS
We thank T. J. Battisti, Agustin Avilla-Sakar, and Valorie Bowman for their help with the cryo-EM analyses. We are also grateful to Sheryl Kelly for administrative aid in preparing the manuscript.
This study was supported by National Institutes of Health (NIH) grants RO1 AI095366 to M.G.R., RO1 GM056279 to R.J.K., and U54 AI065357 to J.T.R. We also thank the NIH (grant DE-AC02-06CH11357) and Purdue University for their support of the EM facility.
Footnotes
Published ahead of print 11 June 2014
REFERENCES
- 1.Weaver SC, Ferro C, Barrera R, Boshell J, Navarro JC. 2004. Venezuelan equine encephalitis. Annu. Rev. Entomol. 49:141–174. 10.1146/annurev.ento.49.061802.123422 [DOI] [PubMed] [Google Scholar]
- 2.Zehmer RB, Dean PB, Sudia WD, Calisher CH, Sather GE, Parker RL. 1974. Venezuelan equine encephalitis epidemic in Texas, 1971. Health Services Reports 89:278–282. 10.2307/4595031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Hawley RJ, Eitzen EM., Jr 2001. Biological weapons: a primer for microbiologists. Annu. Rev. Microbiol. 55:235–253. 10.1146/annurev.micro.55.1.235 [DOI] [PubMed] [Google Scholar]
- 4.Powers AM, Logue CH. 2007. Changing patterns of Chikungunya virus: re-emergence of a zoonotic arbovirus. J. Gen. Virol. 88:2363–2377. 10.1099/vir.0.82858-0 [DOI] [PubMed] [Google Scholar]
- 5.Simon F, Savini H, Parola P. 2008. Chikungunya: a paradigm of emergence and globalization of vector-borne diseases. Med. Clin. N. Am. 92:1323–1343. 10.1016/j.mcna.2008.07.008 [DOI] [PubMed] [Google Scholar]
- 6.Kostyuchenko VA, Jakana J, Liu X, Haddow AD, Aung M, Weaver SC, Chiu W, Lok SM. 2011. The structure of Barmah forest virus as revealed by cryo-electron microscopy at a 6-Å resolution has detailed transmembrane protein architecture and interactions. J. Virol. 85:9327–9333. 10.1128/JVI.05015-11 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sun S, Xiang Y, Akahata W, Holdaway H, Pal P, Zhang X, Diamond MS, Nabel GJ, Rossmann MG. 2013. Structural analyses at pseudo atomic resolution of Chikungunya virus and antibodies show mechanisms of neutralization. eLife 2:e00435. 10.7554/eLife.00435 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Tang J, Jose J, Chipman P, Zhang W, Kuhn RJ, Baker TS. 2011. Molecular links between the E2 envelope glycoprotein and nucleocapsid core in Sindbis virus. J. Mol. Biol. 414:442–459. 10.1016/j.jmb.2011.09.045 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Zhang R, Hryc CF, Cong Y, Liu X, Jakana J, Gorchakov R, Baker ML, Weaver SC, Chiu W. 2011. 4.4 A cryo-EM structure of an enveloped alphavirus Venezuelan equine encephalitis virus. EMBO J. 30:3854–3863. 10.1038/emboj.2011.261 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cheng RH, Kuhn RJ, Olson NH, Rossmann MG, Choi HK, Smith TJ, Baker TS. 1995. Nucleocapsid and glycoprotein organization in an enveloped virus. Cell 80:621–630. 10.1016/0092-8674(95)90516-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Vaney MC, Duquerroy S, Rey FA. 2013. Alphavirus structure: activation for entry at the target cell surface. Curr. Opin. Virol. 3:151–158. 10.1016/j.coviro.2013.04.003 [DOI] [PubMed] [Google Scholar]
- 12.Voss JE, Vaney MC, Duquerroy S, Vonrhein C, Girard-Blanc C, Crublet E, Thompson A, Bricogne G, Rey FA. 2010. Glycoprotein organization of Chikungunya virus particles revealed by X-ray crystallography. Nature 468:709–712. 10.1038/nature09555 [DOI] [PubMed] [Google Scholar]
- 13.Li L, Jose J, Xiang Y, Kuhn RJ, Rossmann MG. 2010. Structural changes of envelope proteins during alphavirus fusion. Nature 468:705–708. 10.1038/nature09546 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gibbons DL, Vaney MC, Roussel A, Vigouroux A, Reilly B, Lepault J, Kielian M, Rey FA. 2004. Conformational change and protein-protein interactions of the fusion protein of Semliki Forest virus. Nature 427:320–325. 10.1038/nature02239 [DOI] [PubMed] [Google Scholar]
- 15.Schmaljohn AL, McClain D. 1996. Alphaviruses (Togaviridae) and flaviviruses (Flaviviridae), Chapter 54 In Baron S. (ed), Medical microbiology, 4th ed. University of Texas Medical Branch at Galveston, Galveston, TX: [PubMed] [Google Scholar]
- 16.Wahlberg JM, Garoff H. 1992. Membrane fusion process of Semliki Forest virus. I. Low pH-induced rearrangement in spike protein quaternary structure precedes virus penetration into cells. J. Cell Biol. 116:339–348 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hunt AR, Frederickson S, Maruyama T, Roehrig JT, Blair CD. 2010. The first human epitope map of the alphaviral E1 and E2 proteins reveals a new E2 epitope with significant virus neutralizing activity. PLoS Negl. Trop. Dis. 4:e739. 10.1371/journal.pntd.0000739 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Roehrig JT, Mathews JH. 1985. The neutralization site on the E2 glycoprotein of Venezuelan equine encephalomyelitis (TC-83) virus is composed of multiple conformationally stable epitopes. Virology 142:347–356. 10.1016/0042-6822(85)90343-5 [DOI] [PubMed] [Google Scholar]
- 19.Roehrig JT, Day JW, Kinney RM. 1982. Antigenic analysis of the surface glycoproteins of a Venezuelan equine encephalomyelitis virus (TC-83) using monoclonal antibodies. Virology 118:269–278. 10.1016/0042-6822(82)90346-4 [DOI] [PubMed] [Google Scholar]
- 20.Hernandez R, Paredes A, Brown DT. 2008. Sindbis virus conformational changes induced by a neutralizing anti-E1 monoclonal antibody. J. Virol. 82:5750–5760. 10.1128/JVI.02673-07 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Hunt AR, Bowen RA, Frederickson S, Maruyama T, Roehrig JT, Blair CD. 2011. Treatment of mice with human monoclonal antibody 24h after lethal aerosol challenge with virulent Venezuelan equine encephalitis virus prevents disease but not infection. Virology 414:146–152. 10.1016/j.virol.2011.03.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Hunt AR, Frederickson S, Hinkel C, Bowdish KS, Roehrig JT. 2006. A humanized murine monoclonal antibody protects mice either before or after challenge with virulent Venezuelan equine encephalomyelitis virus. J. Gen. Virol. 87:2467–2476. 10.1099/vir.0.81925-0 [DOI] [PubMed] [Google Scholar]
- 23.Roehrig JT, Hunt AR, Chang GJ, Sheik B, Bolin RA, Tsai TF, Trent DW. 1990. Identification of monoclonal antibodies capable of differentiating antigenic varieties of eastern equine encephalitis viruses. Am. J. Trop. Med. Hyg. 42:394–398 [DOI] [PubMed] [Google Scholar]
- 24.Johnson BJ, Brubaker JR, Roehrig JT, Trent DW. 1990. Variants of Venezuelan equine encephalitis virus that resist neutralization define a domain of the E2 glycoprotein. Virology 177:676–683. 10.1016/0042-6822(90)90533-W [DOI] [PubMed] [Google Scholar]
- 25.Mathews JH, Roehrig JT, Trent DW. 1985. Role of complement and the Fc portion of immunoglobulin G in immunity to Venezuelan equine encephalomyelitis virus infection with glycoprotein-specific monoclonal antibodies. J. Virol. 55:594–600 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mathews JH, Roehrig JT. 1982. Determination of the protective epitopes on the glycoproteins of Venezuelan equine encephalomyelitis virus by passive transfer of monoclonal antibodies. J. Immunol. 129:2763–2767 [PubMed] [Google Scholar]
- 27.Roehrig JT, Hunt AR, Kinney RM, Mathews JH. 1988. In vitro mechanisms of monoclonal antibody neutralization of alphaviruses. Virology 165:66–73. 10.1016/0042-6822(88)90659-9 [DOI] [PubMed] [Google Scholar]
- 28.Woodward TM, Miller BR, Beaty BJ, Trent DW, Roehrig JT. 1991. A single amino acid change in the E2 glycoprotein of Venezuelan equine encephalitis virus affects replication and dissemination in Aedes aegypti mosquitoes. J. Gen. Virol. 72(Pt 10):2431–2435 [DOI] [PubMed] [Google Scholar]
- 29.Kinney RM, Chang GJ, Tsuchiya KR, Sneider JM, Roehrig JT, Woodward TM, Trent DW. 1993. Attenuation of Venezuelan equine encephalitis virus strain TC-83 is encoded by the 5′-noncoding region and the E2 envelope glycoprotein. J. Virol. 67:1269–1277 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Tang G, Peng L, Baldwin PR, Mann DS, Jiang W, Rees I, Ludtke SJ. 2007. EMAN2: an extensible image processing suite for electron microscopy. J. Struct. Biol. 157:38–46. 10.1016/j.jsb.2006.05.009 [DOI] [PubMed] [Google Scholar]
- 31.Ludtke SJ, Baldwin PR, Chiu W. 1999. EMAN: semiautomated software for high-resolution single-particle reconstructions. J. Struct. Biol. 128:82–97. 10.1006/jsbi.1999.4174 [DOI] [PubMed] [Google Scholar]
- 32.Arnold K, Bordoli L, Kopp J, Schwede T. 2006. The SWISS-MODEL workspace: a web-based environment for protein structure homology modeling. Bioinformatics 22:195–201. 10.1093/bioinformatics/bti770 [DOI] [PubMed] [Google Scholar]
- 33.Bordoli L, Kiefer F, Arnold K, Benkert P, Battey J, Schwede T. 2009. Protein structure homology modeling using SWISS-MODEL workspace. Nat. Protoc. 4:1–13. 10.1038/nprot.2008.197 [DOI] [PubMed] [Google Scholar]
- 34.Rossmann MG, Bernal R, Pletnev SV. 2001. Combining electron microscopic with X-ray crystallographic structures. J. Struct. Biol. 136:190–200. 10.1006/jsbi.2002.4435 [DOI] [PubMed] [Google Scholar]
- 35.Rossmann MG, Morais MC, Leiman PG, Zhang W. 2005. Combining X-ray crystallography and electron microscopy. Structure 13:355–362. 10.1016/j.str.2005.01.005 [DOI] [PMC free article] [PubMed] [Google Scholar]