S1 Appendix -

<p><b>Fig A. Sorting of FMDV-specific B cells using different bait antigens by fluorescence-activated cell sorting (FACS)</b>. <b>(A)</b> Distribution and proportion of FMDV O serotype (O/18074) of specific B cells in the porcine peripheral blood mononuclear cells (PBMCs) identified by flow cytometry. Gate 1 (P1) was selected to exclude cell debris, with lower values of SSC-A and FSC-A, and further analyzed singlets in gate P2 based on diagonal streak of the FSC-A and FSC-H plot. The class-switched (IgG<sup>+</sup>) B cells in gate P3 were used to check the distribution of FMDV-specific cells, in the presence of FMDV serotype O (O/18074) bait antigen. Appropriately one million counts were collected to analyze the proportion of O/18074-binding B cells in PBMCs. <b>(B)</b> Sorting of the FMDV A serotype (A/AF72) or FMDV O serotype (O/18074) specific B from the enriched B cells population using FACS. After excluding cell debris to gate singlets in P2, the class-switched B cells being IgM<sup>-</sup>CD14<sup>-</sup>CD3<sup>-</sup>CD335<sup>-</sup> population were gated in P3 to sort the FMDV-specific cells in the presence of bait antigen A/AF72 and O/18074, respectively. <b>(C)</b> Determination of the purity of the sorted FMDV-specific B cells by flow cytometry. After sorting, the purified B cells were reloaded and one thousand counts were collected to check the proportion of O/18074-binding B cells. <b>Fig B. The phenotype and constitution of FMDV-binding cells revealed by scRNA-seq transcripts. (A)</b> UMAP plot of unsupervised clustering of FMDV-binding cells, comprising of majority B cells, as well as other minimal cell populations such as monocytes, T cell, dendritic cells (DCs) and unidentified cells. <b>(B)</b> Heatmap of differentially expressed genes in B cells, Monocytes, T cells, DCc, and others. <b>Fig C. The heterogeneity of porcine memory B cells revealed by pairing analysis of BCR and transcripts. (A)</b> Heatmap of differentially expressed genes between memory B cells and PBs. <b>(B-F)</b> UMAP plot of unsupervised clustering of porcine memory B cells, comprising of six clusters in <b>B</b>. According to the germline gene usages of VL, the memory B cells were separated into kappa V1-expressing B cells (pairing with cluster 3 in <b>C</b>), kappa V2-expressing B cells (pairing with cluster 2 in <b>D</b>), lambda V3-expressing B cells (pairing with cluster 4 in <b>E</b>) and lambda V8-expressing B cells (pairing with clusters 0, 1 and 5 in <b>F</b>). <b>Fig D. Analysis of the relative abundance and density of BCR clonotypes for serotype O and A specific BCR repertoires.</b> (A) The proportion of BCR clonotypes was shown in O/18074-specific and A/AF72-specific repertoires respectively. (B) The number of the clonotypes with different frequency indicated the difference in abundance between the two repertoires. <b>Fig E. Analysis of the SHMs of porcine BCR repertoires. (A)</b> The SHM difference of porcine VH between memory B cells and plasmablasts in FMDV serotype O-specific repertoire. <b>(B)</b> The SHM difference of porcine VH between memory B cells and plasmablasts in FMDV serotype A-specific repertoire. <b>(C-F)</b> The difference in SHM of each isotype antibodies between the serotype O and A specific BCR repertoires. The statistical analysis was performed using non-parametric Mann-Whitney test in R program and showed the medians in violin plot. <b>(G)</b> The difference in SHM among the isotypes (IgA, IgD, IgE and IgG) antibodies within each serotype O and A specific BCR repertoire. The statistical analysis was performed using Kruskal-Wallis chi-square test in R program and showed the medians in violin plot. P<0.05 indicates a significant difference between two samples. P<0.01 indicates a very significant difference between two samples. P<0.001 indicates an extremely significant difference between two samples. NS indicates no significant difference. Fig F. Identification of the reactivity of porcine mAbs with FMDV serotype O and A using indirect immunofluorescence assay (IFA). (A, B) BHK-21 cells infected respectively with the O/18074 strain (A) or A/AF72 strain (B), and the working concentration of the tested porcine mAbs was 5μg/ml, followed by incubation with rabbit anti-pig FITC (diluted 1:200 in PBS). The cells were observed under an FL Imaging System (Life Technology, USA). The experiments were independently conducted in triplicate. Fig G. Identification of the reactivity of porcine mAbs with FMDV serotype O and A using enzyme-linked immunosorbent assay (ELISA). (A, B) The ELISA plates were respectively coated with inactivated 146S antigen of O/Tibet/99 (A) or A/AF72 (B), and then probed with different concentrations of 0–40 μg/ml of the tested mAbs, followed by probing with HRP-conjugated goat anti-porcine IgG. Color was developed by adding 50 μl of TMB substrate (Pierce, Life Technology) for 10 min at room temperature. The process was stopped by adding equal volumes of 1 M H<sub>2</sub>SO<sub>4</sub>. Optical density at 405 nm (OD<sub>450</sub>) was measured on a microplate reader (BioRad). The experiments were independently conducted in triplicate. Fig H. Analysis of sequence conservation of VP1 GH-loop of available FMDV strains in serotypes O, A and Asia1. The full VP1 amino acids sequences of available FMDV serotype O, A and Asia1 were downloaded from national center for biotechnology information (NCBI) as of June 30, 2023. The sequence logo of VP1 G-H loop of FMDV serotype O (numbers of full VP1 sequences = 6317), A (numbers of full VP1 sequences = 2352) and Asia1 (numbers of full VP1 sequences = 812). The key antigenic determinants on GH-loop that recognized by porcine bnAbs were marked with bold black line. The integrin receptor (αvβ6) recognition motif, “RGD”, was framed with rectangles and marked with yellow. Fig I. The resolution of the cryo-EM reconstruction complex. (A, B) Fourier shell correlation (FSC) of FMDV-O18-POA2 complex (A) or FMDV-AWH-POA2 complex (B). Fig J. Identification of the rescued single-substitution mutants by plaque formation assay. (A) The wild-type (O/18074) and rescued mutants (VP2 D68A, VP2 T71A, VP2 N72A, VP2 R77A, VP2 P195A, VP3 D69A, VP3 S70A and VP3 D195A) formed in BHK-21 cells, and the sizes were correlated to the CPE patterns. (B) The wild-type (A/WH/CHA/09 strain) and rescued mutants (VP2 H65A, VP2 D68A, VP2 T71A, VP2 P195A, VP3 S70A) formed in BHK-21 cells, and the sizes were correlated to the CPE patterns. Fig K. Neutralization mechanism of porcine bnAbs against FMDV serotype O and A. The inhibition effect of porcine bnAbs on viral attachment to BHK-21 cells was determined by IFA. Different amounts of pOA-1, pOA-2 or pOA-13 were respectively mixed with FMDV serotype O (O/HN/CHA/93 strain) (A) or serotype A (A/WH/CHA/09 strain) (B) at 37°C for 1 h, and the mixtures were loaded on BHK-21 cells at 4°C for 1 h to allow virus attachment, then washed 3 times with cold PBS to remove unbound virus. The viruses were detected by IFA. The experiments were independently conducted in triplicate. Fig L. Binding modes of FMDV integrin receptor and antibody. Binding modes of FMDV integrin receptor (avβ6) and bnAb pOA-2. (A) Superposition of FMDV- avβ6 with FMDV-O18-POA2 (A) and FMDV-AWH-POA2 (B). VP1, VP2, VP3 and VP4 of the protomer are shown in blue, green, red and yellow, respectively. The av and β6 chains of integrin (avβ6) and pOA-2 are drawn in cartoon representation and colored in cyan, limon and magenta, respectively. Black dashed circles show significant clashes between antibody (pOA-2) and integrin receptor. Fig M. Effect of porcine bnAbs on virus at the post-attachment stage. The BHK-21 cells were respectively incubated with FMDV serotype O (O/HN/CHA/93 strain) (A, C, E, G) or serotype A (A/WH/CHA/09 strain) (B, D, F, H) at 4°C for 1 h. Subsequently, the cells were treated with different amounts of pOA-1, pOA-2 or pOA-13 at 37°C for 1h, the cells were washed with cold PBS to remove unbound virus, then the cells were further cultured for a duration of 4h. The viruses were quantified through detecting VP1 protein by IFA (A, B) and Western blotting (C, D), 3D gene by qRT-PCR (E, F) and the numbers of PFUs by plaque phenotypes assay (G, H). The experiments were independently conducted in triplicate. The data differences between conditions with virus only and different bnAb treatments were assessed using unpaired T-test (Holm-Sidak method, α = 0.05) in GraphPad Prism 7.2. *** Indicates an extremely significant difference at P<0.001. ** Indicates a very significant difference at P<0.01. * Indicates a significant difference at P<0.05. NS indicates no significant difference. Fig N. The recognized antigen structures of porcine cross-serotype bnAbs against FMDV. Footprints of two distinct cross-serotype antigenic sites on surface of two protomers from two adjacent pentamers of FMDV. One protomer comprising of VP1, VP2, VP3 and VP4 was circled in grey line. Cross-serotype antigenic site 1 that consisted of VP1 143, 145–148 and 151 position residues was marked in pink. Cross-serotype antigenic site 2 that consisted of VP2 65, 68, 71, 72, 77 and 195 position residues on one protomer and VP3 68, 69, 70 and 195 position residues on another protomer, was marked in orange. Table A. Porcine single B cell sequencing primers used for V(D)J amplifications. Table B. The interactive residues between the integrin receptor (alpha V beta 6) with VP1 G-H loop on FMDV. Table C. Porcine broad neutralizing mAb escape mutants. Table D. Interface identification and interaction analysis of pOA-2 with FMDV O/18074 by PISA Program. Table E. Interface ide