Structural and functional mapping of protective human monoclonal antibodies against enterovirus A71

Figures

• 

  Fig. 1.. Neutralizing activities of anti–EV-A71 mAbs against recently circulating isolates (2019–2020 and 2023). Neutralizing (NT) activity of the antibodies was assessed using a cytopathic effect–based neutralization assay. Neutralizing concentrations are shown as a brown color gradient and expressed in microgram/milliliter. Genotype B5 EV-A71 virus that caused HFMD outbreaks in 2019–2020 (strains 5041 and 5005) and 2023 (strain 12350) in Taiwan were tested in the assay. Representative genotype C4 (strain 11-96017), genotype B5 (strain 12-50929), and coxsackievirus A16 (strain 4126) were also included. Convalescent serum that was collected from a patient with HFMD infected by EV-A71 in 2019 and anti–SARS-CoV-2 receptor-binding domain human IgG mAb FD-11A were used as controls. The neutralizing titer of the serum is expressed as the reciprocal of the dilution factor.

• 

  Fig. 2.. Cryo-EM reconstructions of EV-A71/Fab complexes and mapping of antibody footprints. (A) Cryo-EM reconstructions of EV-A71/Fab complexes. Genotypes B2 (strain MS742387), B5 (strain 12-96015), and C4 (strain FY/AH/CHN/2008) of EV-A71 were used for complex formation. The reconstructions are rainbow colored based on the distance (Å) of the surface from the center of the particle. (B) Schematic roadmap showing the footprints of anti–EV-A71 Fabs on the surface of the viral capsid in an asymmetric unit. Footprints of canyon binders are colored in blue, while two- and threefold axis binders are colored in purple and red, respectively. The footprint of the SCARB2 receptor is depicted in light brown. Footprints are drawn with some transparency, so that the colors are more intense in overlapping regions. The canyon region is outlined in green. (C) Heatmap showing the degree of epitope overlap. Numbers indicate the number of residues shared between Fab footprints. Green indicates no overlap, while red indicates major overlap. (D) Surface representation of the EV-A71 capsid with VP1, VP2, and VP3 colored in blue, green, and red, respectively. The positions of some fivefold, three, and twofold…

• 

  Fig. 3.. Varieties of canyon access by neutralizing antibodies. Different binding patterns of canyon-binding antibodies with EV-A71 are shown. (A) Fab 16-2-12D, representative of most canyon binders, inserts its CDR-H3 loop deeply into the bottom of the canyon region. (B) Fab 16-3-10B uses its extended CDR-L1 loop rather than CDR-H3 to reach into the canyon. (C) Fab 17-2-2B does not penetrate the canyon but engages along the southern wall of the canyon. Complexes were formed with EV-A71 B2 (strain MS742387) for 16-2-12D, C4 (strain FY/AH/CHN/2008) for 16-3-10B and B5 (strain 12-96015) for 17-2-2B. In the left panels, the EV-A71 protomer is shown as a gray surface, and the Fab is shown as a cartoon, with heavy and light chains colored in blue and gray, respectively. Contact areas between virus and Fab are highlighted in red for distances ≤4.0 Å and in yellow for >4.0 and ≤9.0 Å. In the right panels, VP1, VP2, and VP3 of EV-A71 are colored in light blue, green, and red, respectively. Heavy and light chains of the Fab are colored in dark blue and gray, respectively. Residues that involve the Fab-virus interactions and their neighboring residues are further indicated.

• 

  Fig. 4.. Clashes between the SCARB2 receptor and neutralizing antibodies. (A) Canyon-binding antibodies 17-2-2B and 16-3-10B, representative of the majority of canyon binders, overlap with the SCARB2 receptor. Yellow and blue ovals highlight the regions of antibody-receptor clash. The EV-A71 virion is shown as a gray surface, and the Fab is shown as a cartoon, with 17-2-2B, 16-3-10B, and SCARB2 colored in blue, purple, and red, respectively. (B) Threefold axis-binding antibodies sterically block SCARB2 engagement. SCARB2 (A) and (B) denote two receptor molecules bound to neighboring protomers. The white triangle shows the approximate position of the threefold icosahedral symmetry axis. Yellow and blue ovals highlight the regions of antibody-receptor clash. EV-A71 is shown as a gray surface and the Fab is shown as a cartoon, with 16-3-4D, 34-1-6D, and SCARB2 colored in cyan, green, and red, respectively. (C) Constant domain of the heavy chain of antibody 16-2-11B clashes with SCARB2, as indicated by the gray oval. EV-A71 is shown as a gray surface. 16-11B heavy chain, light chain, and SCARB2 are colored in pink, cyan, and red, respectively.

• 

  Fig. 5.. Empty particles of EV-A71 induced by Fabs 16-2-12D, 16-3-3C, and 16-2-2D. (A and C) EV-A71 genotype B2 (MS742387) mixed with Fab 16-2-12D. (B) EV-A71 genotype C4 (FY/AH/CHN/2008) mixed with Fab 16-2-12D. (D and E) EV-A71 genotype B5 (12-96015) mixed with Fab 16-3-3C. (F) EV-A71 genotype B5 (12-96015) mixed with Fab 16-2-2D. Molar ratio of Fab: EV-A71 particle in (A), (B), (D), and (F) is ∼200:1 and in (C) and (E) is ∼90:1. Empty particles can be observed in all of these micrographs except (C). (G) Cryo-EM reconstruction of EV-A71 empty particle complexed with Fab 16-3-3C. The map is shown at low contour level so the weak density of Fab 16-3-3C can be seen. Density for the EV-A71 empty particle in this complex is shown in (H) at high contour level, revealing the gaps in the capsid characteristic of the expanded particle. The maps are rainbow colored in the same way as in Fig. 2A.

• 

  Fig. 6.. Molecular mechanism of escape mutations. (A) Substitution VP1 S283F confers escape of genotype C4 EV-A71 from antibody 16-3-10B. The large phenylalanine side chain (shown as cyan semitransparent sticks) introduced reduces the complementarity with antibody 16-3-10B. (B and C) Substitution VP3 E81G/K confers escape of genotype B5 EV-A71 from antibodies 16-3-4D (B) and 34-1-6D (C). Although residue VP3 E81 is not directly involved in antibody binding, it stabilizes a nearby loop (shown thicker) that is targeted by antibodies 16-3-4D and 34-1-6D. The observed escape mutations VP3 E81G/K destabilize this loop, interfering with antibody binding. (D) Substitution VP1 T232A confers escape of EV-A71 from antibodies 16-2-8C, 16-2-9D and 16-2-12D. In the wild type, VP1 T232 interacts with VP1 D110, making the side chain of VP1 D110 optimally interacts with antibody 16-2-8C, 16-2-9D, or 16-2-12D. The mutation VP1 T232A disrupts the interaction between VP1 232 and VP1 D110, increasing flexibility of the side chain of VP1 D110 and leading to escape from antibodies. (E) Substitution VP1 D164E confers escape of genotype B5 EV-A71 from antibody 16-3-3C. The longer side chain of glutamic acid…

• 

  Fig. 7.. In vivo protection by neutralizing anti–EV-A71 mAbs in the human SCARB2 mouse model. (A) Prophylactic and therapeutic efficacy of antibodies (10 mg/kg) against genotype C4 EV-A71. A single dose of antibody or control was administered intraperitoneally 24 hours before or after a lethal challenge of genotype C4 EV-A71 12-73 (n = 4 per group). (B) Prophylactic and therapeutic efficacy of antibodies (10 mg/kg) against genotype B5 EV-A71. A single dose of antibody or control was administered intraperitoneally 24 hours before or after a lethal challenge of genotype B5 EV-A71 13-50144 (n = 4 per group). Anti-influenza B hemagglutinin IgG mAb (BN-2B) was used as a human IgG control. Weight changes and survival rate were monitored following infection. The weight data represents the means ± SD days post infection (d.p.i.).