Determining the Epitope Dominance on the Capsid of a Serotype SAT2 Foot-and-Mouth Disease Virus by Mutational Analyses

ABSTRACT

Monoclonal-antibody (MAb)-resistant mutants were used to map antigenic sites on foot-and-mouth disease virus (FMDV), which resulted in the identification of neutralizing epitopes in the flexible βG-βH loop in VP1. For FMDV SAT2 viruses, studies have shown that at least two antigenic sites exist. By use of an infectious SAT2 cDNA clone, 10 structurally exposed and highly variable loops were identified as putative antigenic sites on the VP1, VP2, and VP3 capsid proteins of SAT2/Zimbabwe (ZIM)/7/83 (topotype II) and replaced with the corresponding regions of SAT2/Kruger National Park (KNP)/19/89 (topotype I). Virus neutralization assays using convalescent-phase antisera raised against the parental virus, SAT2/ZIM/7/83, indicated that the mutant virus containing the TQQS-to-ETPV mutation in the N-terminal part of the βG-βH loop of VP1 showed not only a significant increase in the neutralization titer but also an increase in the index of avidity to the convalescent-phase antisera. Furthermore, antigenic profiling of the epitope-replaced and parental viruses with nonneutralizing SAT2-specific MAbs led to the identification of two nonneutralizing antigenic regions. Both regions were mapped to incorporate residues 71 to 72 of VP2 as the major contact point. The binding footprint of one of the antigenic regions encompasses residues 71 to 72 and 133 to 134 of VP2 and residues 48 to 50 of VP1, and the second antigenic region encompasses residues 71 to 72 and 133 to 134 of VP2 and residues 84 to 86 and 109 to 11 of VP1. This is the first time that antigenic regions encompassing residues 71 to 72 of VP2 have been identified on the capsid of a SAT2 FMDV.

IMPORTANCE Monoclonal-antibody-resistant mutants have traditionally been used to map antigenic sites on foot-and-mouth disease virus (FMDV). However, for SAT2-type viruses, which are responsible for most of the FMD outbreaks in Africa and are the most varied of all seven serotypes, only two antigenic sites have been identified. We have followed a unique approach using an infectious SAT2 cDNA genome-length clone. Ten structurally surface-exposed, highly varied loops were identified as putative antigenic sites on the VP1, VP2, and VP3 capsid proteins of the SAT2/ZIM/7/83 virus. These regions were replaced with the corresponding regions of an antigenically disparate virus, SAT2/KNP/19/89. Antigenic profiling of the epitope-replaced and parental viruses with SAT2-specific MAbs led to the identification of two unique antibody-binding footprints on the SAT2 capsid. In this report, evidence for the structural engineering of antigenic sites of a SAT2 capsid to broaden cross-reactivity with antisera is provided.

INTRODUCTION

Genetically modified viruses provide a valuable tool for the manipulation of the biological properties of field and laboratory strains and present a promising avenue for the design of safe and effective vaccines. The modification of antigenic regions of human immunodeficiency virus (HIV) by amino acid (aa) substitutions in a recombinant virus has been used to confirm monoclonal antibody (MAb)-binding sites and the antigenic dominance of these epitopes (1). Similarly, in recent years, epitope mapping for human viruses has been performed using human recombinant antibodies; for example, two neutralizing antibodies were used to map epitopes on the influenza A H5N1 virus (2). In this study, we utilized epitope replacement in a recombinant virus to determine the epitope dominance of an important pathogen in animals, foot-and-mouth disease virus (FMDV). FMDV, the prototype member of the genus Aphthovirus in the family Picornaviridae, is a small, nonenveloped, icosahedral virus with a single-stranded, positive-sense RNA genome. The virus capsid is composed of 60 copies each of four virus-encoded structural proteins, VP1 to VP4; the capsid outer shell is comprised of VP1, VP2, and VP3, while VP4 lines the interior surface (3, 4). Although FMDV causes a clinically indistinguishable vesicular disease in cloven-hoofed animals, there are seven distinct serotypes and multiple antigenic types (5, 6). Control of FMD has been reliant on large-scale vaccinations with inactivated whole-virus vaccines (7). However, the extensive antigenic diversity within the FMDV serotypes impedes the efficacy of vaccines; therefore, the strain composition of FMD vaccines must be selected with caution (8, 9).

Due to the strong link reported between the protection of cattle against FMDV and the levels of virus-neutralizing antibodies produced following vaccination (10), it has generally been accepted that antibodies represent the major protective arm of the immune response (11, 12). The majority of FMDV-neutralizing antibodies are directed against epitopes located in the three surface-exposed capsid proteins of the virus (4, 13). The mechanism by which antibodies protect against FMDV in vivo is poorly understood; however, previous studies have indicated that escape from neutralizing antibodies may contribute to viral persistence and disease progression (14).

MAbs have been used extensively to identify several antigenic sites on the structural proteins of virions belonging to serotypes A (15–17), O (13, 18), C (19), and Asia-1 (20). Not surprisingly, these antigenic sites were located on structural protrusions on the virus surface, formed mainly by the loops connecting β-barrel structures of the three outer capsid proteins. In particular, the βG-βH loop of VP1 has been identified as immunodominant by the use of peptides (21, 22) and is found in all serotypes of FMDV (4, 13, 23). Sequencing of MAb-resistant (MAR) mutants and mapping of the topography of the mutations on the X-ray crystallographic structure of O/BFS/18/60 (O₁BFS) (4) resolved five neutralizing antigenic sites on the capsid of serotype O FMDV (13, 18). The βG-βH loop functions either independently (site 5; residue 149 of VP1 [18]) or as a discontinuous epitope that encompasses the highly exposed C terminus (Ct) of VP1, particularly residues 200 to 213. This neutralizing antigenic site has been designated site 1 and has been mapped to critical residues at positions 144, 148, 154, and 208. Site 2 involves several amino acids in the βB-βC and βE-βF loops of VP2, spanning residues 70 to 73, 75, 77 (2a), and 131 (2b). Site 3 includes residues 43 to 45 and 48, inside the βB-βC loop of VP1, while site 4 maps within the βB “knob” of VP3, with crucial residues at positions 56 and 58 to 59 (13, 19, 24).

In the case of SAT2 serotype viruses, studies involving MAR mutants revealed at least two antigenic sites. The antigenic site located in the βG-βH loop of VP1, downstream of the RGD motif, at residues 147, 148, 156, and 158 (25), residue 154 (23), and residue 159 (26) is analogous to site 1 of serotype O₁BFS (13). Residue 79 of VP2 may also play a role in forming this antigenic site; however, the role of residue 79 in site 1 remains unclear (23). The second identified antigenic site involves residue 210 at the C terminus of VP1. In addition, the importance of each of these individual neutralizing antigenic sites in SAT2 viruses is still undefined.

In this study, the role of known and predicted epitopes in the antigenicity of SAT2 viruses was investigated. Residues located in 10 of the structurally exposed loops of VP1, VP2, and VP3 were selected and mutated, and the effect of these mutations on antigenicity was measured with virus neutralization (VN) and antibody avidity assays and profiled using SAT2-specific MAbs. We present evidence of epitope dominance within the SAT2 serotype and identify two new epitopes in VP2 for SAT2 viruses. Furthermore, the results revealed the effect of the different surface-exposed mutated residues on the interaction with antibodies in sera from convalescent animals.

MATERIALS AND METHODS

Cell lines, viruses, plasmids, and bacterial strains.

Baby hamster kidney-21 cells (BHK-21, ATCC CCL-10) were maintained and propagated in Eagle's basal medium (BME; Life Technologies) as described previously (27). The SAT2 FMDV vaccine strain Zimbabwe (ZIM)/7/83 (passage history, once in bovine cells and eight times in BHK cells [B₁BHK₈]) is a bovine virus, originating from an outbreak in western Zimbabwe during 1983 (28). The SAT2 virus Kruger National Park (KNP)/19/89 (PK₁RS₂BHK₄; PK indicates pig kidney cells, and RS indicates Instituto Biologico Renal Suino cells [IB-RS-2 is a pig kidney cell line]) is a buffalo virus, originating from the KNP in South Africa during 1989. The plasmid pSAT2, a previously described genome-length infectious cDNA clone of SAT2/ZIM/7/83 (28), was used as the genetic backbone in the construction of recombinant cDNA clones harboring mutated epitopes. The virus recovered from pSAT2 is referred to as vSAT2. Escherichia coli MAX Efficiency DH5α [genotype, Fϕ80dlacZΔM15 Δ(lacZYA-argF)U169 deoR recA1 endA1 hsdR17 (r_k⁻ m_k⁺) phoA supE44 λ⁻ thi-1 gyrA96 relA1], obtained from Life Technologies, was used as the transformation host in cloning experiments.

MAb isolation.

Monoclonal antibodies (MAbs) were prepared by inoculating BALB/c mice with a blend of inactivated and purified 146S particles of SAT1/Kenya (KEN)/11/2005, SAT/2/ZIM/5/81, and SAT3/ZIM/4/81. The antigen blend comprised 40 μg of each antigen. Three doses of the antigen blend in TiterMax Gold adjuvant (Sigma-Aldrich) were administered subcutaneously 3 weeks apart, and a final dose of antigen blend in phosphate-buffered saline (1× PBS) was administered intravenously 7 days before spleen cells were harvested for fusion with murine myeloma SP2/O cells. Cell fusion and cloning of positive hybridomas were performed according to procedures standardized at The Pirbright Institute (29). MAbs were screened by enzyme-linked immunosorbent assay (ELISA) for their reactivity against homologous virus, and five SAT2/ZIM/5/81-specific MAbs were selected (mouse IgG1 isotype MAbs 1D5 [14 μg/ml], DA10 [8 μg/ml], GE11 [19 μg/ml], GD12 [15 μg/ml], GG1 [22 μg/ml]).

Epitope prediction.

The epitopes mutated in this study were predicted as described previously (30). Briefly, potential regions of antigenicity were identified based on the identification of hypervariable regions, defined as having more than 60% variable residues within a 10-amino-acid region, and positions of high entropy, i.e., the uncertainty at each amino acid position (31), within the deduced outer capsid protein sequence of SAT2 viruses. Previous studies indicated that linear amino acid sequences with high variability or residue positions with high entropy which are structurally exposed when mapped to modeled structures of the capsid proteins have the potential to be involved in the antigenicity of the virus (30). The regions selected for mutation in this study were residues 71 to 72 and 133 to 134 of VP2, residues 133 to 134 of VP3, and residues 48 to 50, 84 to 86, 109 to 111, 137 to 140, 157 to 160, 169 to 171, and 199 to 201 of VP1.

Structural modeling.

A homology model of the SAT2 capsid proteins was built using Modeler 9v3 (32) with O₁BFS coordinates (Protein Data Bank accession no. 1FOD) as the template (33). Alignments were performed with ClustalX, and modeling scripts were generated with the structural module of FunGIMS (30). Structures were visualized and the surface-exposed residues identified with PyMOL v1.1rc2pre (DeLano Scientific LLC).

Site-directed mutagenesis, subcloning, and DNA sequencing.

Site-directed mutagenesis of 10 known and putative epitopes located in the VP1, VP2, and VP3 capsid proteins of SAT2/KNP/19/89 (GenBank accession number DQ009735) was used to replace the corresponding epitopes of the genetically disparate virus SAT2/ZIM/7/83 (GenBank accession number DQ009726) using the infectious SAT2 genome-length clone. Overlap extension PCR mutagenesis was used to introduce the mutations into the pSAT2 plasmid. Briefly, each of the PCR processes involved the use of four oligonucleotides (two inner mutagenic oligonucleotides and two genome-specific oligonucleotides) and three different PCRs (34). A description of the different oligonucleotides is provided in Table 1.

TABLE 1.

Synthetic oligonucleotides used for introducing SAT2/KNP/19/89 antigenic regions into the genome-length cDNA clone of SAT2/ZIM/7/83

Mutationa or oligonucleotideb	Oligonucleotide sequence (5′ to 3′)	Orientation (purpose)	Epitope-replaced DNA clonec	Recovered virusc
Mutations
VP2-site2a F	GCTTTTTGATTGGACACCTGAAAAACCATTTGGCACGCTGTATG	Sense	pKNPS2aSAT2	vKNPS2aSAT2
VP2-site2a R	CATACAGCGTGCCAAATGGTTTTTCAGGTGTCCATCAAAAAGC	Antisense
VP2-site2b F	GTGCCGGAGCTGTGCTCGCTTCGGAACAGAGAGGAGTTTCAAC	Sense	pKNPS2bSAT2	vKNPS2bSAT2
VP2-site2b R	GTTGAAACTCCTCTCTGTTCCGAAGCGAGCACAGCTCCGGCAC	Antisense
VP3-site4 F	CACCAGGCATTGAGACTGAAAAGCTGCCCAAGACACCCGAGG	Sense	pKNPS4SAT2	NR
VP3-site4 R	CCTCGGGTGTCTTGGGCAGCTTTTCAGTCTCAATGCCTGGTG	Antisense
VP1-site1 F	CAAAGTACGCCAACATCAAACACACGCTCCCGTCTACCTTC	Sense	pKNPS1SAT2	vKNPS1SAT2
VP1-site1 R	GAAGGTAGACGGGAGCGTGTGTTTGATGTTGGCTACTTTG	Antisense
VP1-site3 F	GTTCTGACAAATAGAACCACCTTCAACGTTGACTTGATGGACAA	Sense	pKNPS3SAT2	vKNPS3SAT2
VP1-site3 R	GGTGTCCATCAAGTCAACGTTGAAGGTGGTTCTATTTGTCAGAAC	Antisense
VP1-site5 F	CAACGGTGAGTGCAAGTACGAGACGCCCGTCACTGCCATTCGCGGTGAC	Sense	pKNPS5SAT2	vKNPS5SAT2
VP1-site5 R	GTCACCGCGAATGGCAGTGACGGGCGTCTCGTACTTGCACTCACCGTTG	Antisense
VP1-DHR F	TTGCCTGCCTTGGCGACCACCGGCGCGTGTGGTGGCAGCC	Sense	pKNPDHRSAT2	vKNPDHRSAT2
VP1-DHR R	GGCTGCCACCACACGCGCCGGTGGTCGCCAAGGCAGGCAA	Antisense
VP1-NKG F	CAACCCCATGGTGTTTTCGAACAAAGGTGTCACGCGTTTTGCTG	Sense	pKNPNKGSAT2	vKNPNKGSAT2
VP1-NKG R	CAGCAAAACGGTGACACCTTTGTTCGAAAACACCATGGGGTTG	Antisense
VP1-NS F	CCACGTGACCGCCGACAACAGCGTCGACGTTTACTACCGG	Sense	pKNPNSSAT2	NR
VP1-NS R	CCGGTAGTAAACGTCGACGCTGTTGTCGGCGGTCACGTGG	Antisense
VP1-Cterm F	CTCCTCCCTGGCTACGACTATGCAAGTAGGGACAGGTTTGACA	Sense	pKNPCtSAT2	vKNPCtSAT2
VP1-Cterm R	CTGTCAAACCTGTCCCTACTTGCATAGTCGTAGCCAGGGAGGAG	Antisense
Genome-specific outer oligonucleotides
2B	GACATGTCCTCCTGCATCTG	Antisense (cDNA)
cDNA-2a	CGCCCCGGGGTTGGACTCAACGTCTCC	Antisense (P1d PCR)
P622	GCACTGACACCACGTCTAC	Sense (P1 PCR)

^aThe structural region that was targeted for mutagenesis.

^bNames of the oligonucleotides used for cDNA synthesis and amplification of the capsid-coding region.

^cThe mutated genome-length cDNA clone and recombinant virus names are derived from the KNP buffalo isolate SAT2/KNP/19/89, followed by the structural region that was mutated (superscript) and the SAT2 genetic background of SAT2/ZIM/7/83. NR, no viable virus was recovered.

^dP1, polyprotein 1.

Mutations were introduced into distinct PCR products using overlapping inner mutagenic oligonucleotides. Using pSAT2 as the template DNA, a “left” PCR was performed by using a common forward outer oligonucleotide (P622) and an antisense inner mutagenic oligonucleotide, whereas the “right” PCR was performed with a sense inner mutagenic oligonucleotide and a common outer reverse oligonucleotide (cDNA-2A). The two PCRs were performed with the TaKaRa Ex Taq PCR system and the following cycling conditions: 95°C for 20 s, 58°C for 20 s, and 68°C for 2 min (30 cycles). The two first-round PCR amplicons were purified from an agarose gel with the NucleoSpin extraction kit (Macherey-Nagel) according to the manufacturer's instructions and mixed in equimolar amounts, and the PCR was extended for 8 cycles of 95°C for 20 s and 74°C for 5 min using the Advantage 2 PCR system (Clontech). The product was then used as the template for PCR by employing the sense and antisense outer oligonucleotides detailed above. The cycling conditions for this third PCR were 95°C for 20 s and 68°C for 3 min (25 cycles).

The agarose gel-purified PCR amplicons of ca. 2.2 kb (containing part of the VP2, VP3, and VP1-2A coding region of SAT2/ZIM/7/83 with the newly introduced mutations) were digested with SspI and XmaI and then cloned into the unique SspI and XmaI restriction sites of the pSAT2 plasmid. Replacement of the SAT2/ZIM/7/83 epitopes with those of SAT2/KNP/19/89 was verified by nucleotide sequencing using genome-specific oligonucleotides and the ABI Prism BigDye Terminator cycle sequencing ready reaction kit (v3.0; Applied Biosystems). The extension products were resolved on an ABI Prism 3100 genetic analyzer (Applied Biosystems). No unintended site mutations were found, and the epitope-replaced mutant clones are indicated in Table 2.

TABLE 2.

Summary of the surface-exposed amino acid differences between the capsid proteins of FMDV SAT2/KNP/19/89 and SAT2/ZIM/7/83^a

Virus(es)	Amino acids at indicated site in capsid protein β-sheetb
	Site 2a in VP2 βB-βC	Site 2b in VP2 βE-βF	Site 4 in VP3 βE-βF	Site 3 in VP1 βB-βC	DHR site in VP1 βD-βE	NKG site in VP1 βF-βG	Site 5 in VP1 βG-βH	Site 1 in VP1 βG-βH	NS site in VP1 βH-βI	Ct site in VP1 C terminus
Parental SAT2/ZIM/7/83	TSDK	LKDR	TDRL	TAFAV	GEHER	SHNNV	YTQQST	NTKHKL	DKPV	DHADR
Parental SAT2/KNP/19/89	TPEK	LRDR	TEKL	TTFNV	GDHRR	SNKGV	YETPVT	NIKHTL	DNSV	DYASR
Epitope-replaced clones	TPEK	LRNR	TEKL	TTFNV	GDHRR	SNKGV	YETPVT	NIKHTL	DNSV	DYASR
Recovered epitope-replaced mutants	TPEK	LRNR	NR	TTFNV	GDHRR	SNKGV	YETPVT	NIKHKLc	NR	DYASR

^aThe identified surface-exposed loops of SAT2/KNP/19/89 were used to replace the corresponding epitopes of the genetically disparate virus SAT2/ZIM/7/83 using the infectious SAT2 genome-length clone pSAT2. The amino acid sequences of the epitope-replaced p^KNPSAT2 clones, as well as those of the recovered viruses, are indicated. NR, no viable virus was recovered from these epitope-replaced mutants.

^bThe amino acid residues in SAT2/ZIM/7/83 that were mutated to the corresponding region in SAT2/KNP/19/89 are indicated in bold.

^cThe mutated amino acid residue in site 1 that reverted back to the original SAT2/ZIM/7/83 sequence when the epitope-replaced virus was recovered in BHK-21 cells is underlined and italic.

In vitro RNA synthesis, transfection, and virus recovery.

The constructed epitope-replaced mutant cDNA clones and pSAT2 were linearized at the SwaI site downstream of the poly(A) tract and used as templates to synthesize RNA in vitro with the MEGAscript T7 kit (Ambion) according to the manufacturer's instructions. RNA transcripts (3 μg) were introduced into BHK-21 cells and seeded into 35-mm-diameter cell culture plates (Nunc) using Lipofectamine 2000 reagent (Life Technologies) according to the manufacturer's instructions. Transfected monolayers were incubated at 37°C with a 5% CO₂ influx for 48 h in Eagle's basal medium (BME) containing 1% (vol/vol) fetal calf serum (FCS) and 25 mM HEPES (Invitrogen). The supernatants were used to infect fresh BHK-21 monolayers and incubated at 37°C for 48 h. Viruses were subsequently harvested from infected cells by a freeze-thaw cycle and passaged four or more times on BHK-21 cells, using 10% of the supernatant of the previous passage. Viruses recovered from transfection included vSAT2, vKNP^S1SAT2, vKNP^S2aSAT2, vKNP^S2bSAT2, vKNP^S3SAT2, vKNP^S5SAT2, vKNP^DHRSAT2, vKNP^NKGSAT2, and vKNP^CtSAT2 (where superscripts indicate the structural regions that were mutated). Following the recovery of viable viruses, the external capsid region was obtained by reverse transcription (RT)-PCR, as described below, and the presence of the mutations was verified with automated sequencing.

RNA isolation, cDNA synthesis, and PCR amplification.

RNA was extracted from infected tissue culture samples with a guanidinium-based nucleic acid extraction method (35) and used as the template for cDNA synthesis. Avian myeloblastosis virus (AMV) reverse transcriptase (Promega) and the genome-specific oligonucleotide 2B (36) were used for reverse transcription, which was carried out at 42°C for 2 h. The external capsid-coding region of the epitope-replaced mutant viruses was amplified using the Expand high-fidelity PCR system (Roche Diagnostics) and flanking oligonucleotides, P622 and cDNA-2A (Table 1).

Plaque titrations.

BHK-21 cell monolayers were infected for 1 h with the parental viruses SAT2/ZIM/7/83 and SAT2/KNP/19/89, as well as with vSAT2 and the above-mentioned epitope-replaced mutant viruses. Following the addition of 2 ml of tragacanth overlay (37) and incubation for 48 h, the cell monolayers were stained with 1% (wt/vol) methylene blue. Virus titers were calculated and are expressed as the logarithm of the number of PFU/ml.

Neutralization of infectivity in cell culture.

The antigenic diversity of the epitope-replaced mutant viruses in relation to the SAT2/ZIM/7/83 and SAT2/KNP/19/89 viruses was determined with cross-neutralization assays in microplates, as described in the OIE Manual of Standards (38). BHK-21 cells were used as the indicator system in the test. Convalescent-phase bovine reference sera were prepared by intradermolingual inoculation of cattle with 10⁴ 50% tissue culture infectious doses (TCID₅₀) of either SAT2/ZIM/7/83 or SAT2/KNP/19/89. Two cattle were infected with each virus strain, after which blood was collected at 21 days postinoculation and pooled. Cattle were housed in the biosafety level 3 isolation facility at the Transboundary Animal Diseases Programme (TADP) with the approval of the Onderstepoort Veterinary Institute (OVI) Animal Ethics Committee. The endpoint titer of the serum against the homologous and heterologous viruses was calculated as the log₁₀ of the reciprocal of the last dilution of serum to neutralize 100 TCID₅₀ in 50% of the wells (39). Differences in the average neutralization titers between each of the epitope-replaced mutant viruses and the reference viruses across four independent experiments were calculated.

To determine whether the five SAT2-specific MAbs (1D5, DA10, GE11, GD12, and GG1) were able to neutralize the parental SAT2/ZIM/7/83 and SAT2/KNP/19/89 viruses, as well as the epitope-replaced mutant viruses, cross-neutralization assays were performed using BHK-21 cell cultures as described above. A minor modification to the protocol described above was that the sera were replaced with the SAT2-specific MAbs.

Virus purification.

Confluent BHK-21 cell monolayers (eight 750-cm² plastic roller bottles; Corning) were infected at a multiplicity of infection (MOI) of 1 with SAT2/ZIM/7/83, vKNP^S5SAT2, vKNP^DHRSAT2, and vKNP^NKGSAT2 in BME containing 1× antibiotic-antimycotic solution and 25 mM HEPES buffer. Following incubation for 12 to 16 h at 37°C, the cells in each roller bottle were lysed by addition of 10% (vol/vol) Triton X-100 and 0.5 M EDTA (pH >7.4). The supernatants were pooled and subjected to centrifugation at 9,800 × g for 30 min. The 146S virus particles were precipitated from the recovered supernatant with 8% (wt/vol) polyethylene glycol 8000 (PEG 8000) at 4°C for 3 h, collected by centrifugation, and suspended in TNE buffer (50 mM Tris [pH 7.5], 150 mM NaCl, 10 mM EDTA). An additional clarification step was performed by addition of 0.5 M EDTA and centrifugation at 9,800 × g for 30 min. The virus particles were purified on a 10 to 50% (wt/vol) sucrose density gradient (SDG), prepared in TNE buffer, as described previously (40). Peak fractions corresponding to 146S virion particles (extinction coefficient at 259 nm [E₂₅₉] at 1% = 78.8 M⁻¹ cm⁻¹) were pooled, and the amount of antigen (μg) was calculated (41).

Sandwich ELISA.

A sandwich ELISA was used for titration of the five SAT2-specific MAbs and to characterize the parental and eight epitope-replaced mutant viruses. MaxiSorp ELISA plates (Nunc) were coated with an optimal dilution of rabbit SAT2 antiserum in 50 mM carbonate-bicarbonate buffer (pH 9.6) and stored at 4°C overnight. A serial 2-fold dilution (1:5 to 1:40) of the parental or epitope-replaced viruses (supernatant of infected cells) in blocking buffer (0.05 M Tris, 0.15 M KCl containing 0.5% [wt/vol] milk powder) was applied to the ELISA plates. Viruses were trapped by incubation at 37°C for 1 h, after which the plates were washed with PBS containing 0.05% (vol/vol) Tween 20 (PBS-0.05%T). Twofold dilutions (1:20 to 1:80) of each of the MAbs, prepared in blocking buffer, were added and the plates incubated at 37°C for 1 h. The ELISA plates were washed with PBS-0.05%T, and horseradish peroxidase (HRP)-conjugated rabbit anti-mouse IgG (Sigma-Aldrich), diluted 1:20,000 in blocking buffer, was added. Following incubation at 37°C for 1 h and washing of the plates, the ELISA plates were developed using substrate-chromogen solution, consisting of 4 mM 3,3′,5,5′-tetramethylbenzidine (Sigma-Aldrich) in substrate buffer (0.1 M citric acid monohydrate, 0.1 M, tri-potassium citrate; pH 4.5) and 0.015% (vol/vol) H₂O₂. The color reaction was stopped after 10 min with 1 M H₂SO₄, and the absorbance values were read at 450 nm using a Labsystems Multiskan Plus photometer (38).

The binding of the MAbs to the epitope-replaced viruses was calculated as follows. The mean absorbance reading at 450 nm (A₄₅₀) for the binding of each MAb to the epitope-replaced viruses were corrected by subtracting the background value. The adjusted A₄₅₀ values for each MAb to the epitope-replaced viruses were then expressed as percentages of the mean A₄₅₀ value obtained against vSAT2.

Additionally, to distinguish between SAT2/ZIM/7/83 12S and 146S virus particles, a sandwich ELISA was performed as described above, with minor modifications. Purified 146S particles were heat treated for 30 min at 56°C. Various concentrations (100 ng to 3 ng) of heat-treated and untreated purified (representing 12S and 146S) particles were added, in triplicate, to ELISA plates coated with saturating concentrations of rabbit anti-SAT2 antiserum. The five SAT2/ZIM5/81-specific MAbs were added to the plates at a 1:40 dilution, and the plates were developed as described above.

sd A-ELISA.

The single-dilution avidity ELISA (sd A-ELISA) protocol was adapted from the work of Lavoria et al. (43). Briefly, MaxiSorp ELISA plates were coated, in duplicate, overnight at 4°C with 200 ng of purified virus in 50 mM carbonate-bicarbonate buffer (pH 9.6). The plates were washed with PBS-0.05%T and blocked at 37°C for 1.5 h with blocking buffer containing PBS, 20% (vol/vol) fetal calf serum (FCS), 0.002% (wt/vol) thimerosal, and 0.1% (wt/vol) phenol red. Following incubation, the plates were washed with wash solution (133 mM NaCl, 8.6 mM K₂HPO₄, 1.5 mM KH₂PO₄, 0.05% [vol/vol] Tween 20 in distilled H₂O). Serum samples (SAT2/ZIM/7/83 and SAT2/KNP/19/89) were diluted 1:80 in blocking buffer and added to the plates. A pool of five negative sera was used as a negative control. The plates were incubated at 37°C for 1 h and washed three times with PBS-0.05%T, and then 4 M urea in PBS was added to one plate and PBS was added to the remaining plate. Following incubation at room temperature for 15 min, the plates were washed four times with wash solution before the FMDV-specific antibodies were detected with HRP-labeled antibovine conjugate (Sigma-Aldrich), diluted 1:20,000 in blocking buffer. The colorimetric reaction was developed after incubation at 37°C for 1 h and washing of the plates as described above. The mean A₄₅₀s of samples and controls were corrected by subtracting the mean of the blank absorbance readings at 450 nm (cA₄₅₀). The avidity index (AI) was calculated as previously described (41). Briefly, AI% = (cA₄₅₀ of the sample with urea/cA₄₅₀ of the sample without urea) × 100.

Statistical analyses.

Virus neutralization titers and avidity indexes of epitope-replaced mutant viruses and the SAT2/ZIM/7/83 and SAT2/KNP/19/89 viruses with convalescent-phase bovine reference sera were compared using repeated-measures analysis of variance (ANOVA) with Bonferroni adjustment of P values for post hoc comparisons. All statistical analyses were performed using GraphPad Prism v5.03 for Windows (GraphPad Software, Inc.).

Ethics statement.

All procedures involving animals were approved by the Onderstepoort Veterinary Institute (OVI) Animal Ethics Committee according to national animal welfare standards and performed with the permission of the Department of Agriculture, Forestry and Fisheries (DAFF). For the production of the SAT2-specific MAbs, mice were housed at The Pirbright Institute and the experiment was approved through the Institute's ethical review process, in accordance with national guidelines on animal use.

RESULTS

Prediction of antigenic sites on the SAT2 FMD capsid.

A combined approach of capsid protein amino acid sequence alignments and known structural data was used to predict antigenic sites on the surfaces of SAT2 virions. A complete alignment of the deduced amino acid sequences of the capsid proteins of 23 SAT2 viruses across Africa revealed amino acid regions of high variability (30, 44) that corresponded to or were located in close proximity to previously identified epitopes on type O and A viruses (see the introduction for references). Many of the variable regions were located within flexible structural loops of the viral capsid and have been linked to poor cross-reaction in in vitro virus neutralization tests (24, 36).

A systematic analysis of the capsid proteins of SAT2/KNP/19/89 (topotype I) and SAT2/ZIM/7/83 (topotype II) revealed that the most variation, 18% (38 of 217 aa), occurred in the VP1 protein, while the VP2 and VP3 proteins varied by ca. 4% (9 of 219 aa and 8 of 222 aa, respectively). Comparison of the deduced amino acids and structurally exposed loops revealed that of the four hypervariable regions previously identified in the VP2 region, only two, i.e., positions 71 to 72 (SD→PE; βB-βC) and 133 to 134 (KD→RN; βE-βF) of VP2 (Table 2; Fig. 1), had significant surface exposure in the whole virion and were therefore chosen for this study. Only one site of VP3 was variable, i.e., residues 133 to 134 (DR→EK; βE-βF), and seven sites with variable residues were identified in VP1 (Table 2; Fig. 1). These included residues 48 to 50 (AFA→TFN; corresponds to site 3 of serotype O), 84 to 86 (EHE→DHR; βE-βF), 109 to 111 (HNN→NKG; βF-βG), 137 to 140 (TQQS→ETPV; βG-βH, corresponds to site 5 of serotype O), 157 to 160 (TKHK→IKHT; βG-βH, corresponds to site 1 of serotype O), 169 to 171 (KP→NS; βH-βI), and 199 to 201 (HAD→YAS; C terminus of VP1) (Table 2; Fig. 1). Residues 144 to 154 and 210 of VP1, all of which fall within previously identified SAT2 antigenic regions (23, 25, 26), were conserved between SAT2/ZIM/7/83 and SAT2/KNP/19/89.

FIG 1.

Locations of the surface-exposed amino acid differences between the capsids of FMDV SAT2/KNP/19/89 and SAT2/ZIM/7/83 on a ribbon protein diagram of a modeled pentamer of SAT2/ZIM/7/83 (28). The protein subunits and structural features are color coded for VP1 (cyan), VP2 (green), and VP3 (magenta). VP4 has been hidden from the structure. The pore, located at the 5-fold axis of the capsid (black pentagon), is shown in the middle of the structure. The 3-fold axis is depicted by the black triangles. The positions of amino acid changes predicted to play a role in antigenicity are shown as yellow spheres. The amino acid changes are those indicated in Table 2 for SAT2/KNP/19/89 and SAT2/ZIM/7/83.

Generation of recombinant viruses with altered surface epitopes.

To study the effects of individual epitope-replaced mutations in a defined genetic background on the antigenic dominance of SAT2 viruses, recombinant virus mutants were constructed using the infectious cDNA clone of the SAT2 virus ZIM/7/83, pSAT2 (Fig. 2). Of the 10 putative and known epitopes for SAT2 viruses selected from sequence and structure data, eight represented surface-exposed loops connecting β-β structures in the three outer capsid proteins. However, two mutations, DR→EK (residues 133 to 134 of VP3) and KP→NS (residues 169 to 171 of VP1), though they appear to have surface exposure, were somewhat obscured by adjacent structural elements and were selected on the basis of sequence heterogeneity only. The location and the electrostatic effects of these mutations on the virion surface are shown in Fig. 3. Introducing or removing the charge on the capsid surface may completely abrogate ionic interaction between antibodies and the capsid. Some of the substitutions caused an increase in the net positive charge in the VP1 protein of the derived recombinant virus (Fig. 3B and C). The EHE→DHR (residues 84 to 86 of VP1) and HNN→NKG (residues 109 to 111 of VP1) mutations had a strong effect on the local surface potential of the capsid, creating a distinct patch of surface area that was predominantly positively charged. The TQQS→ETPV (residues 137 to 140 of VP1) mutation introduced a strong negative charge at the N-terminal base of the βG-βH loop.

FIG 2.

Schematic representation of the epitope replacement strategy used to replace epitopic structures of SAT2/ZIM/7/83 with those of SAT2/KNP/19/89. The 10 known and predicted epitopic structures located in the VP1, VP2, and VP3 capsid proteins of SAT2/KNP/19/89, as well as the corresponding epitopes of the genetically disparate virus SAT2/ZIM/7/83, are indicated. Following overlap extension mutagenesis, as described in Materials and Methods, the epitope-mutated P1 regions were cloned into the SspI and XmaI sites of pSAT2, a genome-length cDNA clone of SAT2/ZIM/7/83. C-term, C terminus; S-frag, S fragment; IRES, internal ribosome entry site; 3′UTR, 3′ untranscribed region.

FIG 3.

Surface models of the crystallographic protomers of SAT2/ZIM/7/83 and the epitope-replaced mutant virus v^KNPSAT2, indicating differences in electrostatic potential. (A) The mutations and their positions on the FMDV protomer are indicated. (B) Electrostatic potential of the SAT2/ZIM/7/83 protomer. (C) Electrostatic potential of the v^KNPSAT2 protomer, containing all 10 mutations. Positively charged surfaces are shown in blue, and negatively charged surfaces are in red. The yellow ovals indicate areas of a change in the local electrostatic potential. The black pentagon and triangle show the 5- and 3-fold axes of the virion, respectively.

The engineered epitope-replaced mutant viruses, designated vKNP^S2aSAT2, vKNP^S2bSAT2, vKNP^S3SAT2, vKNP^DHRSAT2, vKNP^NKGSAT2, vKNP^S5SAT2, vKNP^S1SAT2, and vKNP^CtSAT2, were readily obtained from the infectious cDNA clones. High-titer stocks were prepared and their genetic identities confirmed by sequencing analysis. Despite numerous attempts to recover viable vKNP^S1SAT2, either the recovered viable virus corresponded to the wild-type virus (i.e., SAT2/ZIM/7/83) or only the T156→I mutation in VP1 was present (Table 2). The K159→T mutation reverted back to SAT2/ZIM/7/83 K159 (Table 2). No viruses could be recovered for mutations at positions 133 to 134 of VP3 (KNP^S4SAT2) and 169 to 171 of VP1 (KNP^NSSAT2), despite transfection of a minimum of 20 sequence-correct clones of each mutant.

Effect of the epitope-replaced mutations on plaque morphologies and infectivity titers.

The recombinant mutant viruses were initially characterized by determining whether the introduced mutations affect viral growth in BHK-21 cells. Plaque morphologies for the eight viable epitope-replaced mutant viruses, as well as the vSAT2, SAT2/ZIM/7/83, and SAT2/KNP/19/89 viruses, were compared on BHK-21 cells (Table 3). The two VP2 epitope-replaced viruses, namely, vKNP^S2aSAT2 and vKNP^S2bSAT2, the parental viruses (SAT2/ZIM/7/83 and SAT2/KNP/19/89), and four of the VP1 epitope-replaced viruses (vKNP^S1SAT2, vKNP^S5SAT2, vKNP^CtSAT2, and vKNP^DHRSAT2) all formed medium (3- to 5-mm) and large (6- to 8-mm) plaques. The virus derived from the genome-length infectious cDNA clone of SAT2/ZIM/7/83, vSAT2, as well as vKNP^S3SAT2 and vKNP^NKGSAT2, formed small (<2-mm), medium (3- to 5-mm), and large (6- to 8-mm) plaques. Notably, vKNP^S3SAT2 formed mostly small and medium plaques, with only a few large plaques being observed. vKNP^S3SAT2 and vKNP^S1SAT2 displayed approximately 3-fold-higher infectivity titers (ca. 1.6 × 10⁷ PFU/ml) than the SAT2/ZIM/7/83 parental virus, as opposed to most of the remaining epitope-replaced mutant viruses, all of which had infectivity titers (ranging from 3.2 × 10⁶ to 7.4 × 10⁴ PFU/ml) similar to that of the SAT2/ZIM/7/83 parental virus. vKNP^S1SAT2 showed a 3- to 4-fold-lower infectivity than the parental SAT2/ZIM/7/83 virus. The SAT2/KNP/19/89 parental virus had an infectivity titer of 1 × 10⁸ PFU/ml. Taken together, the results show that the introduced mutations in the VP1 βB-βC loop (residues 48 to 50) and residues 157 to 160 of the VP1 βG-βH loop increased the infectivity to BHK-21 cells. This may, at least in part, be the functional basis of adaptation to this cell line.

TABLE 3.

Parental and recombinant viruses containing the epitope-replaced mutations in the outer capsid proteins VP1 and VP2

Virus	Epitope mutation	Infectivity titer (log PFU/ml)	Passage history	Predominant plaque sizes on BHK-21 cellsa
vSAT2 (wild type)		5.4 × 106	BHK4	S-M-L
SAT2/ZIM/7/83		4.2 × 106	B1BHK8	M-L
SAT2/KNP/19/89		1.0 × 108	PK1RS2BHK4	M-L
vKNPS2aSAT2	SD→PE	3.2 × 106	BHK4	M-L
vKNPS2bSAT2	KD→RN	3.4 × 106	BHK4	M-L
vKNPS3SAT2	AFA→TFN	1.5 × 107	BHK6	S-M-L
vKNPS5SAT2	TQQS→ETPV	1.2 × 106	BHK6	M-L
vKNPS1SAT2	TKHK→IKHK	1.6 × 107	BHK4	M-L
vKNPDHRSAT2	EHE→DHR	4.6 × 106	BHK4	M-L
vKNPNKGSAT2	HNN→NKG	4.4 × 106	BHK5	S-M-L
vKNPCtSAT2	HAD→YAS	7.4 × 106	BHK5	M-L

^aPlaque sizes are defined as small (S) (<2 mm), medium (M) (<3 to 5 mm), and large (L) (6 to 8 mm).

Antigenicities of the recombinant viruses with altered surface epitopes.

The overall antigenic distance of the epitope-replaced mutant and parental viruses were examined by virus neutralization assays using antisera raised from convalescent cattle against the parental viruses SAT2/ZIM/7/83 and SAT2/KNP/19/89. The neutralization titers of the eight epitope-replaced mutant viruses and the parental virus controls are shown in Fig. 4. Except with one epitope-replaced mutant virus, no significant differences (P > 0.05) in the neutralization titers were observed between SAT2/ZIM/7/83 and any of the epitope-replaced mutants, when measured against the SAT2/ZIM/7/83 and the SAT2/KNP/19/89 sera. In contrast, the TQQS→ETPV mutation in the N-terminal part of the βG-βH loop of VP1 resulted in a significant increase (P < 0.05) of 40% in the neutralization titer with the SAT2/ZIM/7/83 serum compared to that of parental SAT2/ZIM/7/83 (Fig. 4).

FIG 4.

Antigenic profiles of the epitope-replaced mutant viruses, as indicated, and SAT2/KNP/19/89, SAT2/ZIM/783, and vSAT2 tested against SAT2 antisera (KNP/19/89 and ZIM/7/83). Convalescent-phase cattle antisera were prepared at TADP by intradermolingual inoculation of cattle with 10⁴ TCID₅₀ of SAT2/KNP/19/89 or SAT2/ZIM/7/83, and blood was collected at 21 days postinoculation. Significant differences between the parental and the epitope-replaced mutant viruses are indicated by the *. The data are means ± standard deviations (SD) of results from four independent experiments.

Next, we determined the avidity index of the convalescent-phase bovine reference sera (SAT2/ZIM/7/83 or SAT2/KNP/19/89) against the parental virus SAT2/ZIM/7/83 and three epitope-replaced viruses, namely, the vKNP^S5SAT2 mutant virus (TQQS→ETPV), vKNP^DHRSAT2, and vKNP^NKGSAT2. Higher neutralization titers were observed for these three viruses, and it was investigated whether these increased titers were a result of increased avidity. The TQQS→ETPV (site 5) mutation significantly increased (P < 0.001) the avidity index of the vKNP^S5SAT2 virus to the SAT2/ZIM/7/83 serum (Fig. 5A). The avidity index of the vKNP^DHRSAT2 and vKNP^NKGSAT2 epitope-replaced mutant viruses to the SAT2/ZIM/7/83 serum was not significantly different from that obtained for the parental virus (Fig. 5A). Interestingly, the avidity index of vKNP^S5SAT2 and vKNP^DHRSAT2 to the SAT2/KNP/19/89 serum was significantly higher (P < 0.001) than that of SAT2/ZIM/7/83 (Fig. 5B). No significant difference in the avidity indexes between the vKNP^NKGSAT2 mutant (with the SAT2/KNP/19/89 serum) and the SAT2/ZIM/7/83 virus was observed (Fig. 5B).

FIG 5.

Avidity indexes of the parental SAT2/ZIM/7/83 virus and three epitope-replaced mutant viruses. The avidity index of the parental virus, SAT2/ZIM/7/83, and those of the epitope-replaced mutant viruses vKNP^S5SAT2, vKNP^DHRSAT2, and vKNP^NKGSAT2 with the SAT2/ZIM/7/83 serum (A) and the SAT2/KNP/19/89 serum (B) are indicated. The avidity indexes of the three epitope-replaced mutant viruses were compared to that of the parental SAT2/ZIM/7/83 virus. Significant differences are indicated by ***. The data are means ± SD of results from duplicate experiments.

Antigenic profiling of epitope-replaced and parental viruses with SAT2-specific MAbs.

The epitope-replaced viruses and the parental SAT2 viruses were characterized on the basis of their reactivity to SAT2-specific MAbs. None of the MAbs were shown to neutralize the viruses in vitro, using virus neutralization assays. The binding profiles of the SAT2-specific MAbs were subsequently examined using a sandwich ELISA (Fig. 6). All five of the MAbs reacted to SAT2/ZIM/7/83 and vSAT2. However, two distinct clusters were observed with regard to the reactivities of the MAbs to SAT2/KNP/19/89 and the epitope-replaced mutant viruses. MAbs 1D5, GG1, GE11, and DA10 reacted with vSAT2, SAT2/ZIM/7/83, vKNP^S2bSAT2, vKNP^DHRSAT2, vKNP^NKGSAT2, vKNP^S1SAT2, vKNP^S3SAT2, vKNP^S5SAT2, and vKNP^CtSAT2 but not with SAT2/KNP/19/89 or vKNP^S2aSAT2. In the second cluster, MAb GD12 reacted with all eight epitope-replaced viruses as well as vSAT2, SAT2/ZIM/7/83, and SAT2/KNP/19/89, albeit with different binding reactivities. The MAb GD12 showed less than 55% reactivity to vKNP^S2aSAT2, vKNP^S2bSAT2, vKNP^DHRSAT2, vKNP^NKGSAT2, and SAT2/KNP/19/89, compared to vSAT2 and SAT2/ZIM/7/83, in the ELISA (Fig. 6). Interestingly, GD12 was the only MAb that was able to distinguish between 146S and 12S particles, as was evidenced by a decrease in the absorbance reading following heat treatment of the purified virus particles of SAT2/ZIM/7/83 (data not shown).

FIG 6.

Reactivities of the epitope-replaced mutant viruses and parental viruses with SAT2-specific MAbs. A sandwich ELISA was performed using a 1:10 dilution of the indicated viruses and a 1:40 dilution of the MAbs. The ELISA was performed in duplicate, and the results of one experiment are shown, as the same trend was observed for both ELISAs. The reactivity scales in relation to the interaction of the MAbs to the epitope-replaced mutant viruses are as follows: white boxes, 0 to 15%; half-black boxes, 15 to 55%; and black boxes, 55 to 100%.

The four MAbs that did not react to SAT2/KNP/19/89 and vKNP^S2aSAT2 could be divided into two principal binding subclusters (Fig. 6). One subcluster, consisting of MAbs GG1 and DA10, recognized vSAT2, SAT2/ZIM/7/83, and seven epitope-replaced viruses (vKNP^S2bSAT2, vKNP^DHRSAT2, vKNP^NKGSAT2, vKNP^S1SAT2, vKNP^S3SAT2, vKNP^S5SAT2, and vKNP^CtSAT2) with the same binding reactivity as in the ELISA. The second binding subcluster, consisting of MAbs 1D5 and GE11, reacted poorly with vKNP^S2bSAT2 and vKNP^S3SAT2, i.e., exhibiting less than 55% reactivity.

DISCUSSION

Little is known about the neutralizing epitopes for the three SAT serotype viruses. In this study, the role of structurally exposed loops on a SAT2 capsid in the antigenicity of the virus was investigated. Following an epitope replacement strategy, we measured the antigenic diversity of eight epitope-replaced viruses with polyclonal antisera raised against SAT2/ZIM/7/83, used as the genetic background, and SAT2/KNP/19/89, used as the epitope donor. One of these replacements significantly increased not only the neutralization titer but also the avidity index to the SAT2/ZIM/7/83 serum compared to those of the parental SAT2/ZIM/7/83 virus. Furthermore, antigenic profiling of the epitope-replaced and parental viruses with SAT2-specific MAbs identified two novel nonneutralizing epitopes, both encompassing residues 71 to 72 of VP2. Noteworthy, residues 71 to 72 of VP2 are variable in an alignment of SAT2 viruses. The first antigenic region includes, in addition to residues 71 to 72 of VP2, residues 133 to 134 of VP2 and 48 to 50 of VP1, while the second region includes residues 133 to 134 of VP2 as well as residues 84 to 86 and 109 to 111 of VP1.

A commonly used method to assess the antigenic matching of FMDV within a serotype is the comparison of VN titers. However, the role of the surface-exposed loops of the FMDV outer capsid proteins on antigenicity and their interaction with antibodies with different affinities and avidities are still obscure. The results reported here suggest that critical residues located in surface-exposed loops contribute significantly to the stability of antigen-antibody complexes and blocking of virus entry into cells, either by direct contact with the antibody or indirect contact by local distortion of side chains. It has been shown that the affinity of antigen-antibody complexes could be strongly reduced by mutations of specific side chains within the in-contact epitope (45, 46).

Each of the modified viruses used in this study had two to four critical, variable residues located on structurally exposed loops of the outer capsid proteins changed from the SAT2/ZIM/7/83 sequence to the SAT2/KNP/19/89 sequence. By making use of this strategy, it was expected that the corresponding epitopes within SAT2/ZIM/7/83 would be modified, thus abrogating antibody interaction. Studies have indicated both in FMDV (47) and in poliovirus (48) that mutations within antigenic sites completely abrogate binding with relevant virus-specific antibodies. Thus, antibodies produced against an intact or naive epitope will not recognize the mutated epitope. Crowther et al. (18) reported that a 15% decrease in VN titers of postvaccination cattle sera was observed following a single amino acid mutation. Similarly, studies using synthetic peptides have indicated that adding, removing, or changing a single amino acid within peptides alters the binding or reactivity of the peptide to an MAb (49).

The neutralization profiles of seven of the epitope-replaced mutant viruses were not significantly different than the neutralization profiles of SAT2/ZIM/7/83 and the recombinant vSAT2 virus with the SAT2/ZIM/7/83 and the SAT2/KNP/19/89 antisera. The avidity index of the HNN→NKG mutant virus with the SAT2/ZIM/7/83 and SAT2/KNP/19/89 antisera was not significantly different from that for the SAT2/ZIM/7/83 parental virus. The avidity index of the EHE→DHR mutant did not significantly increase with the SAT2/ZIM/7/83 serum; however, a significant increase was observed with the SAT2/KNP/19/89 serum. The increase in avidity index with the SAT2/KNP/19/89 serum may be due to the presence of nonneutralizing antibodies present in the antiserum that have high avidity to the mutant virus.

Noteworthy, the neutralization profile of the epitope-replaced mutant virus containing the TQQS→ETPV mutation revealed a significantly higher neutralization titer with the SAT2/ZIM/7/83 antisera. The TQQS→ETPV mutation also resulted in a significant increase in the avidity index of the SAT2/ZIM/7/83 serum to this epitope-replaced mutant virus compared to that of the parental SAT2/ZIM/7/83 virus. High avidity indexes have previously been linked to high neutralization titers (50). The increased neutralization profile seen for the TQQS→ETPV mutation may be due to the increased stability of a neutralizing epitope, as a result of the amino acids changes introduced, thus increasing the binding avidity of the neutralizing antibodies to the epitope.

In an attempt to more precisely dissect the role of each predicted SAT2 epitope in its interaction with antibodies, we measured the reactivity of each mutant virus against five MAbs in a sandwich ELISA, as opposed to the traditional generation of virus escape mutants, and identified at least two novel discontinuous epitopes. None of the SAT2-specific MAbs neutralized any of the epitope-replaced or parental viruses, indicating that these novel epitopes are nonneutralizing. The different reactivity patterns observed for the five MAbs against the epitope-replaced and parental viruses may be explained on a structural level. The failure of MAbs 1D5, GG1, GE11, and DA10 to react to SAT2/KNP/19/89 and vKNPS^2aSAT2 is likely due to the mutation of serine to proline at position 71 in the βB-βC loop of VP2 (Table 2; Fig. 7). Therefore, MAbs 1D5, GG1, GE11, and DA10 have a common interaction site encompassing residues 71 to 72 of VP2, confirming, for the first time, the role of this site as an epitope for SAT2 viruses.

FIG 7.

Ribbon protein diagram depicting the proposed binding footprint of the SAT2-specific MAbs onto the capsid protein of a modeled SAT2 pentamer. (A) The critical reactivity residue for MAbs 1D5, GG1, GE11, and DA10, encompassing residues 71 to 72 (SD→PE) of VP2, is indicated with red spheres. The other two contact points for 1D5 and GE11 are indicated with orange spheres, and the putative footprint for the last two MAbs is shown by the broken line. (B) The putative contact points for MAb GD12 are shown by the yellow spheres, and a putative footprint is indicated by the broken line. The estimated distances between the residues that make contact with the MAbs are also indicated. The black pentagon and triangle indicate the 5- and 3-fold axes of the virion, respectively.

Two of the four MAbs, i.e., 1D5 and GE11, also showed a significant reduction in reactivity (>45%) to viruses with mutations at residues 133 to 134 of VP2 (equivalent to type O site 2B) and 48 to 50 of VP1 (equivalent to type O site 3). The critical residue substitution in the βE-βF loop of VP2 is the replacement of a negatively charged aspartic residue with a weak positive asparagine at position 134 and may explain the lower reactivity of the two MAbs to vKNPS^2bSAT2 (Table 2; Fig. 7). The A48→T and A50→N (hydrophilic and bulky) substitutions in the βB-βC loop of VP1 may contribute to the lower reactivity to vKNP^S3SAT2 (Table 2; Fig. 7). Taken together, the results suggest that residues 71 to 72 of VP2 are the major contact point for 1D5 and GE11. Furthermore, the binding footprint of these MAbs includes residues 133 to 134 of VP2 and 48 to 50 of VP1. Structurally, residues 48 to 50 of VP1 and 133 to 134 of VP2 are located approximately 51 Å and 16 Å from residues 71 to 72 of VP2, respectively (Fig. 7A). The binding footprint of the MAbs GG1 and DA10 on the SAT2 capsid overlaps that of the above-mentioned MAbs at the critical residues 71 to 72 of VP2. However, MAbs GG1 and DA10 reacted similarly to the seven remaining epitope-replaced vSAT2 and SAT2/ZIM/7/83 mutant viruses and thus do not have the same binding footprint of MAbs 1D5 and GE11. Although we do not have direct structural evidence, it can be hypothesized that these two MAbs may bind to the βB-βC loop of VP2 on opposite sides of the 2-fold axis of the virus.

Our data provided evidence of a second unique MAb binding footprint, that of GD12, which encompasses residues 71 to 72 and 133 to 134 of VP2 (site 2) and residues 84 to 86 and 109 to 111 of VP1. MAb GD12 presented a different epitope specificity to SAT2 viruses, as elucidated by its ability to react to both vKNPS^2aSAT2 and SAT2/KNP/19/89, albeit with lower reactivity. Lower reactivity was also observed for viruses with amino acid substitutions at residues 84 to 86 (βE-βF) and 109 to 111 (βF-βG) of VP1 and residues 133 to 134 of VP2. Residues 71 to 72 and 133 to 134 of VP2 and 84 to 86 of VP1 are surface exposed in the length of a shallow groove formed by the interaction of VP2, VP3, and VP1 and located ca. 16 Å and 33 Å from each other (Fig. 7B). However, the role of residues 109 to 111 of VP1 in the interaction with this MAb is not as evident, as it is located in a depression at the 5-fold axis of the virion. This is the first time that this second antigenic region, incorporating residues 71 to 72 and 133 to 134 of VP2 (site 2) and residues 84 to 86 and 109 to 111 of VP1, has been described.

A nonneutralizing, conformational epitope at the N terminus of VP2 has been reported for Asia-1 and other FMDV serotypes (50, 51). Furthermore, a structural relationship between antigenic sites situated on the exposed loops of different capsid proteins has been noted previously. For Asia-1, a structural relationship exists between antigenic site 2 (residues 67, 72, 74, 77, and 79 of VP2) and antigenic site 4 (residues 58 and 59 of VP3) (52). Similarly, antigenic site 3 of serotype A encompasses residues 82 to 88 in the βB-βC loop of VP2 and residues 58 to 61 in the βB-βC loop, 136 to 139 in the βE-βF loop, and 195 in the βH-βI loop of VP3 for A10 (53). Also, the major discontinuous antigenic site of serotype C, site D, includes several loops of VP1 (subsite D1), VP2 (subsite D2), and VP3 (subsite D3) (54). This is the first time, however, that the βE-βF and βF-βG loops of VP1 are implicated as having a role in the antigenicity of FMDV. In a previous study, it was found that antibodies against discontinuous epitopes in serotypes O and A can distinguish between the 12S and 146S particles (55). These antibodies recognize conformational epitopes and thus depend on the native structure of virus particles, which is likely the case with MAb GD12 in this study.

Unlike previous reports where the identification of epitopes relied on residue changes that abrogate binding of MAbs, but with no information on how different epitopic units contribute to the interaction with the MAb, our report describes the complete binding footprint of an antibody within the FMD capsid. The complementary determining regions (CDRs) of the heavy- and light-chain variable regions of an antibody interact with an antigen, and CDR3 is the most diverse (56, 57). Unlike in mouse antibodies, the bovine Ig heavy chain has a very long CDR3, which can be more than 60 residues in length (58, 59) and may span a spatial distance of up to 45 Å. As the CDR3 region is the most heterogenous, the three-dimensional structure of the antigen-binding site is influenced by its length (60). Therefore, the structural design of engineered vaccines with increased antigenicity will need to incorporate the entire binding footprint identified in this study, i.e., all three contact points of the binding antibody.

The results reported in this paper provide evidence that directed evolution or rational engineering of antigenic sites through mutation of a few of the antigenically relevant positions may change the cross-reactivity of antibodies to a heterologous virus within a serotype of FMDV. The change in cross-reactivity may be directly linked to the avidity of antibodies to the virus. However, further structural and functional studies are necessary to better understand the structural basis of antigenic variation and the interaction of the FMDV epitopes in antibody binding.