Novel antibody binding determinants on the capsid surface of serotype O foot-and-mouth disease virus

Abstract

Five neutralizing antigenic sites have been described for serotype O foot-and-mouth disease viruses (FMDV) based on monoclonal antibody (mAb) escape mutant studies. However, a mutant virus selected to escape neutralization of mAb binding at all five sites was previously shown to confer complete cross-protection with the parental virus in guinea pig challenge studies, suggesting that amino acid residues outside the mAb binding sites contribute to antibody-mediated in vivo neutralization of FMDV. Comparison of the ability of bovine antisera to neutralize a panel of serotype O FMDV identified three novel putative sites at VP2-74, VP2-191 and VP3-85, where amino acid substitutions correlated with changes in sero-reactivity. The impact of these positions was tested using site-directed mutagenesis to effect substitutions at critical amino acid residues within an infectious copy of FMDV O1 Kaufbeuren (O1K). Recovered viruses containing additional mutations at VP2-74 and VP2-191 exhibited greater resistance to neutralization with both O1K guinea pig and O BFS bovine antisera than a virus that was engineered to include only mutations at the five known antigenic sites. The changes at VP2-74 and VP3-85 are adjacent to critical amino acids that define antigenic sites 2 and 4, respectively. However VP2-191 (17 Å away from VP2-72), located at the threefold axis and more distant from previously identified antigenic sites, exhibited the most profound effect. These findings extend our knowledge of the surface features of the FMDV capsid known to elicit neutralizing antibodies, and will improve our strategies for vaccine strain selection and rational vaccine design.

Fig. 1.

(a) O1K reduced structure (cartoon) showing position of critical amino acid residues of five neutralizing antigenic sites in blue. Residue VP2-188 shown in orange and VP2-78 shown in purple have been identified using bovine mAbs (Barnett et al., 1998). Additional residues mutated in this study are shown in magenta. (b) Three-dimensional structure (external surface) of the O1K (reduced) showing changes in antigenic sites 2 and 4 within the circle. Changes in amino acids forming the five neutralizing antigenic sites (5M) in the central protomer (shown in white) are shown in blue. Amino acids at position VP3-85 (close to VP3-58 that defines antigenic site 4), VP2-191 (located at the threefold axis) and VP2-74 (close to VP2-72 that defines antigenic site 2) are shown in magenta, VP2-188 in orange and VP2-78 in purple.

Fig. 2.

(a) Reactivity profile (ELISA) of rO1K-wt virus and its derivatives with neutralizing antigenic site-specific O Lausanne murine mAbs. Sucrose gradient-purified 146S antigens were used to ensure all the mutant viruses had an equivalent amount of antigen in the test. (b) ELISA results of individual and pooled guinea pig hyperimmune sera against rO1K-wt virus and its derivatives. Sucrose gradient-purified 146S antigens were used to ensure all the mutant viruses had an equivalent amount of antigen in the test. (c) Reactivity profile (ELISA) of the 5M virus when compared to the parent rO1K-wt virus using neutralizing antigenic site-specific O Manisa murine mAbs (SA85, SA107, SA113 and SA127) and O Lausanne bovine mAbs (C2, C96 and MH5). Sucrose gradient-purified antigens were used to ensure all the mutant viruses had an equivalent amount of antigen in the test.

Fig. 3.

(a) Percentage reduction in VN titre of the single and double mutant viruses when compared to the parent rO1K-wt virus using guinea pig antisera. Two-dimensional VN testing was carried out using post-vaccinal sera raised in guinea pigs against rO1K-wt vaccine antigen. * and ** indicate significant difference (to 5M virus) at P<0.05 and P<0.01, respectively. (b) Percentage reduction in VN titre of the single and double mutant viruses when compared to the parent rO1K-wt virus using bovine antisera. Two-dimensional VN testing was carried out using post-vaccinal sera raised in bovine against a similar type O antigen (O BFS). * and ** indicate significant difference (to 5M virus) at P<0.05 and P<0.01, respectively.

Fig. 4.

(a) Percentage reduction in VN titre of the rO1K-VP2-191virus when compared to the parent rO1K-wt virus. Two-dimensional VN testing was carried out using the guinea pig antisera raised against rO1K-wt viral antigen. ** indicates significant difference at P<0.01. (b) Reactivity profile (ELISA) of rO1K-VP2-191 virus with neutralizing antigenic site-specific O Lausanne murine mAbs (B2, D9, C6, C8, EH9 and OC3), neutralizing antigenic site 2-specific O Manisa murine mAbs (SA85, SA107, SA113 and SA127) and O Lausanne bovine mAbs (C2 and C96). Sucrose gradient-purified 146S antigens were used to ensure all the mutant viruses had an equivalent amount of antigen in the test.

Fig. 5.

Reactivity profile (ELISA) of the single and double mutant viruses when compared to the parent rO1K-wt virus using site 2-specific O Lausanne bovine mAbs. Sucrose gradient-purified 146S antigens were used to ensure all the mutant viruses had an equivalent amount of antigen in the test; an O BFS post-vaccinal serum was used as a positive control.

Fig. 6.

Percentage reduction in VN titre (a) and the associated r₁ values (b) of various serotypes O, A and Asia 1 field isolates in comparison to the rO1K-wt virus. Two-dimensional VN testing was carried out using the guinea pig antisera raised against rO1K-wt viral antigen.