Cattle antibodies identify a cross-serotype broadly neutralising foot-and-mouth disease virus epitope

Fig. 1. Cattle vaccination regime and outputs. a Schematic representation of the cattle vaccination regime. Timeline of vaccinations from day 0 (prime vaccination) until day 105 (end point) post prime vaccination (dpv) is shown. Vaccines were administered every 21 days with the following antigens: prime and boost with O1M, followed by heterologous boosts with A22, Asia1S and SAT2S antigens (dotted lines). Blood sampling time points are indicated by red diamonds. b Sera VNT and LPBE titres during the sequential heterologous vaccination timeline. Average titres are shown for viruses of O1M, A22, Asia1S and SAT2S serotypes for each sampling day during the 105 dpv study. Results are expressed as the group mean (4 animals assessed in duplicate). Standard errors of the mean are shown. Vaccination timepoints are indicated by dotted lines. c Kinetics of FMDV-specific and total IgG secreting cells in PBMCs during the sequential heterologous vaccination study. Average numbers of serotype-specific antibody secreting B cells (ASC) are shown per 106 PBMC, for O1M, A22, Asia1S, and SAT2Z serotypes VLPs at each sampling day along the 105 dpv. Total IgG ASCs are also indicated in purple. Results are…

Fig. 2. Identification and characterisation of the tri-serotype specific antibodies by biolayer interferometry binding and VNT. a Binding response of the 50 recombinantly expressed bovine IgG1 antibodies assessed by BLI. Response values (in nm) to O1M, A22, Asia1S, and SAT2E VLPs are plotted for each antibody. Green symbols indicate positive control samples from serotype-specific sera, and red symbols indicate the responses of B4, an RSV-specific bovine IgG1, as a negative control. Tri-serotype specific antibody responses are shown by solid black symbols. Responses over 0.1 nm indicate binding of the antibodies to the VLP-coated biosensor. b Venn diagram showing the relationships between antibodies exhibiting a binding response above 0.1 nm assessed by BLI, for O1M, A22, and Asia1S. Three antibodies were binders to all three FMDV serotypes. c Off-rate and neutralisation potency of the three tri-serotype specific antibodies to three FMDV serotypes. Antigen-antibody interactions were measured by BLI (black circles) for O1M, A22, and Asia1S VLPs. Minimum antibody concentrations required to neutralise O1M, A22, and Asia1S viruses were assessed by VNT (pink squares). Values are averages of…

Fig. 3. Broad topotype coverage of tri-serotype specific antibodies assessed by ELISA and VNT. Phylogenetic tree of FMDV topotypes tested by ELISA and/or VNT for binding to/neutralisation by Ab17, Ab34, and Ab49. Effect was categorised as positive (pink), negative (blue), or intermediate (green) based on ELISA absorbance at 450 nm or on the minimum antibody concentration (nM) required for virus neutralisation. FMDV isolates are arranged phylogenetically according to full-length VP1 sequences; samples not tested in either assay are indicated in white.

Fig. 4. Crystal structures of Fabs of the three tri-serotype specific antibodies. a–c Overall structures of Fab 17, Fab 34, and Fab 49, respectively. Heavy chains are coloured in red, light chains in blue. Overlap of the VL (d) and VH (e) domains of the three Fabs complementarity determining regions L1, L2, L3 and H1, H2, H3 are shown for VL and VH, respectively. Fab 17 is in red, Fab 34 in cyan and Fab 49 in grey. Disulphide bonds in CDR-H3 are shown as sticks in yellow in (a–e). f Views onto the paratopes of Fabs 17, 34 and 49 coloured by electrostatic charge. The common groove is marked by a dashed line.

Fig. 5. Tri-serotype specific Fab epitope mapping using cryo-EM. a Complex of O1M VLPs with Fab34, density with an applied B-factor of 500 Å. For one penton, the VP1 G-H loops (N and C terminal ends) are circled and the VP1 C-termini are indicated by stars. Pentamer, triangle, and oval symbols correspond to the 5-, 3-, and 2-fold axes respectively for one monomeric unit. b Globally refined maps of O1M VLPs complexed with each of the tri-serotype specific Fabs with the model shown for one monomer coloured by capsid protein (VP1 blue, VP2 green, VP3 pink). The C-terminus of VP1 is indicated by a star and the position of the G-H loop on adjacent protomers is shown by circles in each panel. c Side view of a single Fab density from the B-factor ‘blurred’ cryo-EM map of the Fab34-O1M complex with capsid protein models coloured blue, green, pink for VP1, 2, 3 and overlay of integrin model, shown in red and gold. The C-terminus of VP1 is marked as a star and the VP1 G-H loop termini are circled.

Fig. 6. Tri-serotype specific antibodies target a conserved linear epitope at the VP1 C-terminus. a ELISA reactivity of Ab17, 34, 49 and negative control Ab117 with O1M peptides designed to broadly cover the VP1 C terminus region. Alignment of these peptides with each other is also shown. b Sequence logo indicating the frequency of each amino acid at positions 203–210 of VP1 in the viruses bound (top) and not bound (bottom) by the tri-serotype specific antibodies. c Average RUMax for SPR peptide binding with Ab17, 34, and 49. Ab117 was included as a negative control. Each block represents the percentage of RUMax derived from three independent couplings on the same chip. Each line is a peptide from the virus strain indicated, with its sequence shown relative to that of O/Tan used as the reference. Amino acids highlighted in blue are highly conserved across all viruses bound and neutralised by the tri-serotype specific antibodies.

Fig. 7. Binding of C-terminal VP1 peptide to the tri-serotype specific Fabs. a Summary of ELISA results with a VP1 C-terminus sequence alignment (corresponding to residues 200–211 in O1M) for each of the listed viruses. The rows are grouped by the ELISA signal, which represents the average for the three antibodies across all viruses (except where a specific topotype is indicated in which case the average for the three antibodies for that topotype is shown). Standard deviation (SD) is also shown. For O, A, Asia1 and SAT2 the sequence displayed is the most prevalent within the viruses tested. For SAT1 and SAT3 sequences are more diverse with no single prevalent sequence present. Instead, the most prevalent amino acid for each position is shown and variability at position 204 is highlighted. b Overlay of two peptide/Fab34 structures determined by co-crystallisation, the 7-mer peptide complex is in blue, the 18-mer complex in green and the apo Fab model is in orange. c The left panel shows the Fab electrostatic surface for the 18-mer peptide/Fab34 complex structure. In the two panels to the right, this peptide structure has been mapped onto the two other Fabs by superimposing the variable…

Fig. 8. Details of peptide binding interactions with Fab34. a Structures for 18-mer (right) and 7-mer (left) peptides viewed from above the peptide binding groove. The light chain is in blue, heavy chain in red, and peptide in green. H-bonds determined in Chimera88 are shown as dashes. Heavy and light chain CDRs are labelled (H3, L1-3). b Ligplot+89 representation of the 7-mer and 18-mer (left and right, respectively) interactions. Heavy chain residue labels are highlighted in red, those of the light chain in blue. Note key H bond interactions between K210 and CDR-L1 D51 and residues V206 and K204 with the C-terminal portion of CDR-H3.