Insertion site of FLAG on foot-and-mouth disease virus VP1 G-H loop affects immunogenicity of FLAG

ZHU Yuan-yuan , ZOU Xing-qi, BAO Hui-fang, SUN Pu, MA Xue-qing, LIU Zai-xin, FAN Hong-jie ,ZHAO Qi-zu

1 MOE Joint International Research Laboratory of Animal Health and Food Safety/College of Veterinary Medicine, Nanjing Agricultural University, Nanjing 210095, P.R.China

2 China Institute of Veterinary Drug Control, Beijing 100081, P.R.China

3 Lanzhou Veterinary Research Institute, China Academy of Agricultrual Sciences, Lanzhou 730046, P.R.China

4 Jiangsu Co-Innovation Center for the Prevention and Control of Important Animal Infectious Diseases and Zoonoses, Yangzhou University, Yangzhou 225009, P.R.China

1. Introduction

Foot-and-mouth disease (FMD) is a highly contagious and fatal viral disease caused by the etiological agent FMD virus (FMDV), which affects cloven-hoofed wild and domestic animals, including cattle, goat, sheep and pigs.FMDV is a member of the genus Aphthovirus in the family Picornaviridae and is categorized into seven serotypes,including numerous genotypes within each serotype(Grubman and Baxt 2004).

The virus particle (25–30 nm in diameter) contains a single strand of sense RNA encoding the viral genome and has an icosahedral capsid without an envelope. A polyprotein is translated from the genome and further cleaved by host and virus proteases into at least twelve mature viral proteins, including four structural proteins (VP1,VP2, VP3 and VP4) and eight non-structural proteins (Lpro,2A, 2B, 2C, 3A, 3B, 3C and 3D) (Mason et al. 2003). The FMDV capsid is composed of VP1, VP2, VP3 and VP4, with VP1–3 exposed on the virion surface and VP4 located at the inner surface of the capsid. The surface structural proteins VP1–3 share a similar structure consisting of eight chainlike beta sheet barrels linked by surface loops. The surface loops protrude from the capsid and contain important viral epitopes. Among them, the G-H loop between VP1 residues 140 and 160, linking βG and βH, has a highly conserved arginine-glycine-aspartic acid (RGD) peptide motif, which recognizes an integrin ligand and is the most variable region and one of the most significant antigenic epitopes of FMDV(Bittle et al. 1982; Jackson et al. 1997; Neff et al. 1998;Mahapatra et al. 2012). The RGD motif plays a critical role in eliciting host protective immune responses (Rieder et al.1994; Mateu et al. 1995; Fischer et al. 2003).

The G-H loop has considerable resilience, ensuring reliable recognition between ligand and receptor (Acharya et al. 1989; Logan et al. 1993). Additionally, cell-adapted FMDV is able to use non-integrin-mediated mechanisms to enter cells, for instance by using heparan sulfate (HS),a cell surface glycosaminoglycan (GAG) and certain other receptors (Jackson et al.1996; Mandl et al.2001). Reverse genetics technology, peptide-mediated inhibition of plaque formation and sequencing have been frequently employed to reveal novel cell-binding sites of FMDV (Baranowski et al.2000; Zhao et al. 2003).

The insertion or replacement of amino acids around the G-H loop is tolerated in FMDV and may thus be used as a strategy to study FMDV structure, antigenicity and the dynamics of virus receptor binding (Baranowski et al. 2001;Seago et al. 2012; Wang et al. 2012; Lawrence et al. 2013;Yang et al. 2015). Certain insertion positions (139/140,150/151, 154/155, 155/156, 160/161, and 163/164) in the G-H loop may affect whether a live recombinant virus is generated, an epitope is displayed or a virus-antibody interaction occurs (Wang et al. 2012). The successes and failures of previous experiments indicate that it is critical to maintain virus structural integrity when inserting an exogenous epitope into the surface of an FMDV structural protein. However, it remains to be determined how factors such as insertion position affect interactions among the virus, cells and host immune system (Wang et al. 2012;Lawrence et al. 2013). His, hemagglutinin (HA) or FLAG tags inserted into the FMDV G-H loop may yield live viruses that are useful for developing a differentiating infected from vaccinated animals (DIVA) vaccine in the future.

Here, an infectious clone was established from a type O swine FMDV isolate with an 8-aa FLAG marker(DYKDDDDK) inserted upstream (–4) 140/141 (R/V)or downstream (+10) 157/158 (R/A) of the RGD motif.The tagged viruses vFLAG-O/CHA/90 and vO/CHA/90-FLAG were compared to the parental virus in a series of experiments to investigate how different FLAG insertion positions affect FLAG epitope display, virus virulence, and immunogenicity. Comparing to previous study, RGD–4 was an appropriate and novel inserting site which could tolerate an exogenous gene in the swine FMDV O/CHA/90 strain,vaccination with vFLAG-O/CHA/90 could induce both FLAG-antibody and FMD-antibody in vivo. This work provided new research tools to investigate the pathogenic mechanism of pig-adapted type O FMDV.

2. Materials and methods

2.1. Virus, cell lines and plaque assay

O/CHA/90 virus was isolated from a pig with clinical signs of infection in Lanzhou Veterinary Research Institute, Gansu,China (Zhao et al. 2003). The virus was passaged twice in BHK-21 cells and cultured in Eagle’s Minimal Essential Medium (MEM) containing 8% fetal calf serum (Gibco,Australia) at 37°C with 5% CO2. The Chinese hamster ovary (CHO) cell lines CHO-K1 (CCL-61TM, HS-expressing),CHO-618 (CRL-2241TM, lacking integrin and HS), CHO-677(CRL-2244TM, low level of HS alone) and CHO-745(CRL-2242TM, low levels of both HS and chondroitin sulfate)were obtained from the American Type Culture Collection(ATCC, Manassas, VA, USA) and cultured in Ham F-12 medium with 10% fetal calf serum. Plaque assays were performed as previously described (Zhao et al. 2003).

2.2. Antibodies and synthetic peptides

Hyper-immune serum to O/CHA/90 strain was collected from infected swine one month after recovery. An antibody titer of 1:1 024 was measured by liquid-phase competitive enzyme linked immunosorbent assay (ELISA). A rabbit anti-FMDV polyclonal antibody and ELISA Kit (no. 2014120901-3)were prepared in Lanzhou Veterinary Institute, Chinese Academy of Agricultural Sciences; a mouse anti-FLAG monoclonal antibody, anti-mouse IgG (conjugated to Texas Red), and anti-rabbit IgG-FITC were all purchased from Sigma (Sigma-Aldrich, St. Louis, MO, USA). Peptide1(142RVSNVRGDLQVLAQK157), corresponding to the parental FMDV G-H loop amino acid positions 142–157, peptide KGE (142RVSNVKGELQVLAQK157), in which the integrinbinding RGD was replaced with KGE (Zhao et al. 2003),and peptide2 (143VSNVRGDLQVLAQKA158), in which the N- and C-terminal amino acids differed from those of peptide1, were synthesized by Sangon Biotech Company(Shanghai, China).

2.3. RNA isolation, PCR amplification and sequencing

Total RNA was extracted from virus-infected cell cultures using a Qiagen RNA Extraction Kit (Qiagen, Hilden,Germany) and reverse-transcribed into cDNA using Superscript III reverse transcriptase (Life Technology,Gaithersburg, MD, USA) with a random primer. Target sequences were synthesized using PFU Ultra (Stratagene,La Jolla, California, USA) and specific oligonucleotide primers (Table 1). The PCR products were sequenced using selected primers (Table 1) and asymmetric amplification with Big-Dye terminators, followed by resolution on an ABI 3700 or 3100 sequencer by Sangon Biotech Company.

2.4. Construction of O/CHA/90 strain full-length infectious cDNA clones and recombinant FLAG tags

Four fragments of O/CHA/90 genes were amplified from O/CHA/90 cDNA, cloned into pMD18T (Promega, Shanghai,China) and designated pA, pB, pD, and p(E+C) (shown in Fig. 1-B). The pGEM-5Zf(+) vector (Promega) was digested with BsaAI and NsiI (New England Bio-Labs, Ipswich,Massachusetts, USA) and then ligated with a synthetic oligonucleotide cDNA fragment containing SphI, XmaI,XbaI, EagI and SwaI sites to generate the plasmid vector p5MCS. Nucleotide fragments were synthesized and then inserted into p5MCS, resulting in pRibozyme-MCS, which contained a T7 promoter, FMDV 5´ sequences (shown in Fig. 1-A), a hammerhead (HH) ribozyme sequence (SphI/XmaI), a hepatitis delta virus (HDV) ribozyme sequence, and a T7 terminator sequence (EagI/SwaI). Then, pB (including the 5´ S fragment, poly(C15) and part of an internal ribozyme entry site (IRES) was digested with XmaI/XbaI and inserted into pRibozyme-MCS, which resulted in pRibozyme-MCSAB. Next, pRibozyme-MCS-AB and p(E+C) were digested with XbaI/EagI and XbaI/NotI, respectively, and ligated to obtain pRibozyme-MCS-ABC. Finally, p(E+C) and pD were digested with XbaI and XmaI, respectively, pRibozyme-MCS-ABC was digested with XbaI, and the three resulting fragments were ligated and cloned to obtain a full-length cDNA infectious clone called pO/CHA/90 (Fig. 1). Four pairs of primers (p11–14 in Table 1) covering the P1 region(as shown in Fig. 1-A–D) of FMDV were introduced into the SspI and SgrAI restriction sites to permit insertion of the FLAG (DYKDDDDK) marker into the VP1 RGD motif upstream (–4) 140/141 and downstream (+10) 157/158 of the G-H loop using overlapping PCR. The obtained chimeric plasmids were designated pFLAG-O/CHA/90 and pO/CHA/90-FLAG (Fig. 1).

2.5. Virus rescue, passage and performance

The infectious clone plasmids pFLAG-O/CHA/90, pO/CHA/90-FLAG and pO/CHA/90 were linearized with SwaI and used as templates for RNA transcription in vitro using a MEGAscriptT7 Kit (Ambion, Austin, Texas, USA). The purified RNA transcripts were then electroporated into BHK-21 cells for viral rescue with the following parameters:5 μg of RNA per transfection in 200 μL of cells at a density of 5×106cell mL–1and electroporation with 250 V and 750 μF.The electroporated cells were then incubated until acytopathic effect (CPE) was observed. Rescued viruses were harvested via freeze-thaw cycles and labeled F0. F1to F16viruses were obtained via passaging by inoculating an Fnvirus into confluent BHK-21 cells and incubating until a 95%CPE was observed. Cell supernatants containing rescued viruses were labeled Fn+1passage and then frozen at −70°C.

Table 1 Primers for the foot-and-mouth disease virus (FMDV) O/CHA/90 strain and FLAG marker viruses

Fig. 1 Schematic diagram of the construction of O/CHA/90 strain and two FLAG marker viruses. A, the basic gene constituents of O/CHA/90 RNA. B, positions of primers synthesized to construct the full-length infectious clone O/CHA/90 strain. C, key restriction enzyme sites and components in pO/CHA/90 and the chimeric plasmids. D, primers and design of the FLAG marker virus. As indicated, the FLAG (DYKDDDDK) marker was inserted into the VP1 G-H loop upstream (–4) 140/141 and downstream (+10)157/158 of the RGD motifs. The RGD motif and FLAG epitope are shown in bold., FLAG insertion position.

The F6viruses of O/CHA/90, FLAG-O/CHA/90, and O/CHA/90-FLAG (2×108.0TCID50(50% tissue culture infective dose)) were inoculated subcutaneously into suckling mice (four mice per group) for three continuous passages according to previously described procedures(Skinner 1951; Salguero et al. 2005). At each passage, the mice died 16–20 h post infection, and the carcasses without skins were ground into homogenates that were used for the next-passage inoculation of new mice. The viruses obtained with each mouse passage were labeled M1, M2 and M3.

2.6. lmmunofluorescence microscopy

A total of 200 μL of 10-fold diluted viral supernatant(O/CHA/90, FLAG-O/CHA/90, and O/CHA/90-FLAG) was added to confluent BHK-21 cells grown on glass coverslips and incubated at 37°C for 1 h. After washing three times with pH 6.0 PBS, the coverslips were overlaid with 0.6%Tragacanth Gum (ICN Biomedicals, Inc., California, USA)and incubated for 4–5 h followed by 4% paraformaldehyde(Sigma-Aldrich, St. Louis, USA) fixation for 0.5 h and 0.2%Triton-X-100 permeabilization. The permeabilized cells were then stained with mouse-anti-FLAG mAb and rabbit-anti-FMDV polyclonal antibody, followed by goat-anti-mouse-AF594 and goat-anti-rabbit-FITC, respectively. Stained coverslips were observed using an Olympus Confocal Microscope (Japan) with a 40× oil immersion objective.

2.7. Immunization of animals and antibody detection

Animals (mice and pigs) used in the present study were housed in an air-conditioned, air filtered, biosafety level III facility in Lanzhou Veterinary Research Institute. The experimental protocol was approved by the Lanzhou Veterinary Research Institute Ethics Committee in Animal Experimentation.

The viruses (O/CHA/90, FLAG-O/CHA/90, and O/CHA/90-FLAG) were collected, BEI (binary ethylenimine)-inactivated, precipitated using 7% PEG, and then quantified via 146S sucrose density-gradient centrifugation to prepare virus antigens for animal vaccination (Barteling and Meloen 1974). Inocula (considered experimental vaccines) were prepared by emulsifying each virus antigen (20 μg mL–1146S) in pH 7.6 PBS with 50% (w/w; 1 mL) Montanide ISA 206 VG (Seppic (Shanghai) Chemical Specialties Co., Ltd.,Shanghai, China).

For the mouse experiment, 12 BALB/C mice (9–10 weeks old, 20 g) were divided into three groups. Each group was inoculated with one of the three vaccines subcutaneously at days 0 and 14 (5 μg per time point). At 28 days post inoculation, the mice were sacrificed, and serum samples were collected for antibody testing.

For the pig experiment, 12 pigs (weighing 20–25 kg,sero-negative for FMDV antibodies) were divided into four groups. One group served as a sham vaccination control,and the other three groups were intramuscularly vaccinated with each of the three virus vaccines (O/CHA/90, FLAG-O/CHA/90, and O/CHA/90-FLAG) at days 0 and 21 (10 μg/injection/pig). Serum samples were collected at day 49 for antibody testing. Furthermore, all pigs were challenged at day 49 with 1 000×ID50(50% infective dose) O/CHA/90 and were then observed for 10 days continuously.

Mice and pig serum samples were tested for the presence of anti-FMDV antibodies using a liquid-phase competitive ELISA Kit, as recommended by the OIE Manual (OIE 2016).The presence of anti-FLAG antibodies was assessed in an indirect ELISA assay. Nunc Maxisorp ELISA plates(eBioscience, California, USA) were coated with aminoterminal Met-3×Flag (Sigma-Aldrich, St. Louis, USA) at a concentration of 1 μg mL–1diluted in 0.05 mol L–1pH 8.6 carbonate/bicarbonate buffer (Sigma-Aldrich, St. Louis,USA). The coated plates were washed with PBST five times and then incubated at 37°C for 1 h; serial dilutions of serum from 1:4 to 1:1 024 were added to the wells in triplicate. After washing, goat-anti-mouse-HRP (1:2 000, Sigma) or rabbitanti-pig-HRP (1:2 000, Sigma) was added, and the plates were incubated at 37°C for 1 h. Plates were then washed,and the reaction was stopped with TMB blue stop buffer.A492 values were read on a Synergy 2 Multi Detection microplate reader (SpectraMax190, USA).

2.8. Statistical analysis

Data were statistically processed using GraphPad Prism 5.0(GraphPad Software, Inc., San Diego, CA, USA) for analysis of variance (one-way ANOVA). The data are represented as the means with the standard deviation (SD) of three independent experiments. The results with P≤0.05 were considered significant.

3. Results

3.1. Generation of FLAG-tagged FMDV mutants

Full-length cDNA infectious clones of pO/CHA/90, pFLAG-O/CHA/90, and pO/CHA/90-FLAG were constructed (Fig. 1).The pO/CHA/90 vector included an HH ribozyme and an HDV ribozyme at both ends of the FMDV genome to ensure precise ending of transcription/translation, as the rescued virus G-H loop was highly flexible and exhibited considerable length and sequence variation between different serotypes. Leucine residues L148 and L151 at RGD+1 and RGD+4 were key for interactions with integrin in most variants. Moreover, the integrity of the helical structure that followed the RGD was required due to its high-affinity binding to integrin (Burman et al. 2006; Dicara et al. 2008).After comparing different strains of different serotypes, the regions upstream and downstream of the integrin-binding sites (RGD–4 and RGD+10) were considered variable.

DNA encoding a FLAG tag was inserted into the swine FMDV Cathay topotype strain O/CHA/90 using a PCR-based method, and then viruses were rescued and examined for their growth characteristics in BHK-21 cells. All passaged viruses produced apparent CPE in BHK-21 cells. Similar plaque morphologies were produced in BHK-21 cells by the parental O/CHA/90 virus and tagged viruses (FLAG-O/CHA/90 and O/CHA/90-FLAG) (Fig. 2). The passaged F6 virus titers of O/CHA/90, vFLAG-O/CHA/90 and vO/CHA/90-FLAG were 108.57, 108.33and 108.0TCID50mL–1, respectively,indicating no obvious differences in titers. As shown in Fig. 2, the step viral growth curve indicated that the three viruses had similar growth characteristics.

Fig. 2 Plaque morphology and growth curve of virus in BHK-21 cells at 36 hours post infection (hpi). A, the viruses were inoculated into cells with 30 min for absorption and were then discarded before cells were washed three times and covered with gum. Plaque assays were performed using an overlay and crystal violet staining; the resultant plaques were observed and recorded at 36 hpi. B, cells were incubated with Minimal Essential Medium (MEM)-diluted virus supernatant at an multiplicity of infection (MOI) of 1 for 1 h, followed by aspiration of viral supernatant and washing three times with MEM, then replenishing with MEM for culturing. Cells and supernatant were harvested every 1.5 h until CPEs were observed in all cells. Viral titers were determined based on TCID50 (50% tissue culture infective dose) mL–1 and are given as the average of three replicates. Data are means±SD.

Table 2 Amino acid changes in the P1 region of O/CHA/90 and FLAG marker viruses

Individual P1 FMDV stocks were sequenced across the site of insertion, and the FLAG insertions in the tagged viruses were stable in different passaged viruses. However,amino acid variations in other regions were detected in tagged viruses after passaging. Three mutations were found in vFLAG-O/CHA/90 only (Table 2): VP3 F54L, VP1 N142S (upstream of the RGD motif of the G-H loop), and D188G (downstream of the RGD motif). Only one mutation was found in vO/CHA/90-FLAG: VP3 M86L. The VP1 E83K mutation was found in all three passaged viruses (Table 2).Amino acid variations previously reported to occur after BHK-21 cell passages (Zhao et al. 2003) were not observed in passaged infectious clone rescued vO/CHA/90 viruses.Taken together, these results suggested that, compared to the parental vO/CHA/90 strain, inserting the FLAG sequence into 140/141 (R/V, RGD–4) or 157/158 (R/A, RGD+10)did not alter viral growth characteristics. However, either insertion may lead to compensatory mutations in the P1 region during passaging (Table 2).

3.2. Tagged viruses were stable after passage in suckling mice

To confirm the in vivo stability of the FLAG marker, stably adapted F6 viruses were inoculated (in parallel) into suckling mice for three continuous passages. All viruses were pathogenic to suckling mice. The amino acid sequence of the viral P1 region showed that the FLAG epitope was preserved in suckling mice without any mutations.Additionally, the viral loads of M3O/CHA/90, vFLAG-O/CHA/90 and vO/CHA/90-FLAG were 5×106.75, 5×107, and 5×106.5LD50(50% lethal dose) mL–1, respectively, suggesting the three viruses had similar virulence. Thus, we concluded that the exogenous FLAG marker insertion was passaged stably in suckling mice, and the insertion did not affect the pathogenicity of these viruses.

3.3. FLAG insertion did not interfere with RGD motif recognition of the cell surface integrin receptor

The rescued viruses vFLAG-O/CHA/90, vO/CHA/90-FLAG and vO/CHA/90 were inoculated into the CHO-K1 (HS-expressing), CHO-618 (lacking integrin and HS-expressing),CHO-745 (low levels of HS and chondroitin sulfate) and CHO-677 (low levels of HS alone) cell lines, and no plaque formation was observed. Based on these results, these viruses did not infect cells in the absence of the integrin receptor, even in the presence of HS or a cell surface glycosaminoglycan (GAG). Virus sequencing ruled out the occurrence of vac-O/CHA/90 mutations (VP1 E83K),which were reported in a previous study (Zhao et al. 2003).Therefore, we concluded that these marker viruses did not infect cells through new types of receptors. Collectively,these data suggest that the insertion of FLAG into the FMDV VP1 G-H loop either upstream or downstream of RGD motifs had no effect on the ability of the virus to use the RGD integrin as a receptor to infect cells. A peptide blocking assay was performed for confirmation. As shown in Fig. 3, both RGD peptides 1 and 2 inhibited the replication of tagged viruses and parental viruses to varying degrees;however, the KGE control peptide did not impede viral growth, indicating that the tagged viruses recognized the integrin receptor of BHK-21 cells via the RGD motif of the VP1 G-H loop. Intriguingly, FLAG-O/CHA/90 demonstrated a similar receptor binding capacity to that of the parental virus O/CHA/90, while replication of the other FLAG-tagged virus was less affected by RGD peptides (Fig. 3), implying that different FLAG insertion sites do affect the interaction between RGD and integrin to a certain degree.

3.4. Epitope tag expression analysis

To validate the antigenicity of the tagged viruses, confocal microscopy was performed to test whether FLAG-tagged virus on cells was recognized by an anti-FLAG Ab and an anti-FMDV Ab. The parental virus was not detected by the anti-FLAG Ab (data not shown). Western blot analysis also gave positive results for the tagged viruses with both a mouse anti-FLAG mAb and swine convalescent serum,but there was only an apparent reaction between parental virus and convalescent serum (data not shown).

Fig. 3 Inhibition of foot-and-mouth disease virus (FMDV) replication by two RGD peptides and one KGE peptide. Monolayers of BHK-21 cells were allowed to react with diluted synthetic peptides for 1 h, and then virus at 50 PFU (plague forming unit) was added and incubated. After washing away unabsorbed virus with pH 6.0 PBS, cell monolayers were overlaid with gum, and plaques were counted after staining 36 h later. Data are means±SD.

Immunoprecipitation of the viruses resulted in a band of approximately 30 kDa, which was interpreted to be VP1.After acetone precipitation, the purified progeny viruses were tested for antigenicity with an anti-FLAG mAb and swine convalescent serum using SDS-PAGE. The anti-FLAG mAb reacted only with viruses containing the FLAG insertion, while the swine convalescent serum reacted with both marker viruses and parental virus. This corroborated the conclusion that the FLAG epitope was properly displayed and reacted with anti-FLAG antibodies.

3.5. Assessment of the ability of tagged viruses to induce anti-FLAG antibodies as markers in inoculated mice and pigs

To determine the immunogenicity of the FLAG-tagged viruses, we performed ELISAs to assay the antibodies elicited against FMDV and FLAG after prime and boost vaccinations of mice and pigs. FLAG-tagged viruses generated high titers (log10) of anti-FMDV antibodies (Fig. 4),similar to those elicited by the parental virus (P＞0.05):vO/CHA/90 titers were 2.775±0.15 in mice and 2.4±0.173 in pigs, vFLAG-O/CHA/90 titers were 2.625±0.287 in mice and 2.175±0.45 in pigs, and vO/CHA/90-FLAG titers were 2.4±0.173 in mice and 2.55±0.346 in pigs. However, no anti-FLAG antibodies were elicited by either vO/CHA/90 or vO/CHA/90-FLAG (Table 3). These results suggested that the FLAG insertion position had no effect on FMDV antigenic epitope presentation but did affect the immunogenicity of the inserted FLAG antigen itself. Vaccination with any of the three viruses (O/CHA/90, FLAG-O/CHA/90, and O/CHA/90-FLAG) protected pigs against challenge with the parental virus O/CHO/90 strain.

4. Discussion

Fig. 4 Detection of antibodies against foot-and-mouth disease virus (FMDV) and FLAG at 49 days post infection (dpi) in pigs.Left, anti-FMDV antibody results obtained using a liquid-phase competitive ELISA kit. Right, indirect ELISA results to determine anti-FLAG antibody titers. The OD value of each well was read at 492 nm. Data are means±SD. ＊ means significant difference.

Table 3 Antibody detection of FLAG and foot-and-mouth disease virus (FMDV) post-inoculation in mice

The FMDV serotype O Cathay topotype O/CHA/90 strain,which was the major pig FMDV vaccine strain in China (Zhao et al. 2003), is selected to construct an infectious clone. This platform represented an important research tool to study the pathogenesis of serotype O porcinophilic FMDV and would be helpful in developing a DIVA vaccine for disease control. Serotype O was the most widespread serotype of FMDV with more than 10 different topotypes co-circulating worldwide, including the Cathay topotype, which was first isolated from pigs in Hong Kong in the 1970s. This topotype had continued to cause epidemics in swine populations in Southeast Asian countries and Taiwan of China (Knowles and Samuel 2003). The Cathay topotype O/CHA/90 strain,which was isolated from the Southern China border region,was used to prepare an inactivated FMD vaccine for pigs in mainland China (Zhao et al. 2003). The most notorious FMD epidemic caused by the Cathay topotype was the 1997 swine FMD outbreak in Taiwan (Yang et al. 1999; Lee et al.2009). The virus was noted for its atypical porcinophilic phenotype and a 10-aa deletion in protein 3A (Dunn and Donaldson 1997; Beard and Mason 2000; Knowles et al.2001). As a result, this type of FMDV had been intensively studied to enable disease prevention and candidate vaccine development (Alexandersen and Donaldson 2002; Orsel et al. 2007; Park et al. 2014).

The G-H loop tolerated the insertion of exogenous genes or motifs to obtain chimeric viruses, but the insertion site and size of the insertion appeared to affect viral viability.Structural studies of FMDV had indicated that, unlike the virions of other picornaviruses, the FMDV virion had a smooth surface with small loop-shaped bumps, such as the G-H loop of the VP1 protein (Acharya et al. 1989).Cleaving the viral G-H loop with trypsin led to the complete suppression of virus infectivity and a drop in immunogenicity(Meloen et al. 1983; McCullough et al. 1987). A synthetic G-H loop peptide induced a neutralizing humoral response in animals (Wang et al. 2001; Rodriguez et al. 2003). Partial deletion of the G-H loop attenuated the virus and still induced protective immunity. Replacing the G-H loop upstream of the RGD motif residues with a FLAG marker epitope resulted in a chimeric marker virus that was useful for studying viral pathogenesis (Lawrence et al. 2013). The insertion of His, HA or FLAG tags into the FMDV G-H loop yielded live virus. Similarly, insertion of the serotype O epitope of the G-H loop into serotype Asia 1 virus had been performed to study the effects on viral replication and neutralization phenotypes (Wang et al. 2012). In these studies, live viruses had been successfully obtained, regardless of the status of the sequences upstream and downstream of the RGD motif. However, insert positions and sizes had differed,affecting viral rescue as well as the ability of the rescued virus to react with or recognize specific antibodies and cell receptors. Furthermore, the size of the insertion appeared to affect viral viability. The insertion of twelve-amino-acid type O epitopes into the type Asia 1 virus downstream of the RGD motif resulted in non-viable virus, while the insertion of a 8-aa HA or eight-amino-acid O1K strain epitope led to live virus rescue (Seago et al. 2012; Wang et al. 2012).Nevertheless, the effects of virus serotype differences were unable to be ruled out. In chimeric viruses, exogenous epitopes reacted with specific antibodies, but it was not clear whether the epitopes induced antibodies in vivo. To maintain capsid structural integrity and viral infectivity and to ensure epitope tag accessibility(Acharya et al. 1989; Parry et al. 1990; Logan et al. 1993; Curry et al. 1996), sites in the G-H loop upstream and downstream of the integrin-binding sites (RGD–4 and RGD+10, respectively) of the structural protein VP1 were selected. These sites were found to be variable based on the comparison of different strains from different serotypes. Rescued tagged vFLAG-O/CHA/90 and vO/CHA/90-FLAG were similar to the parental virus in terms of BHK cell passage stability, CPE, virulence, plaque morphology and replication dynamics (Fig. 2). The tagged viruses also demonstrated similar levels of pathogenicity in suckling mice and were passaged stably. These results confirmed earlier findings regarding the resilience of the G-H loop to insertions because tagged viruses with 8-aa insertions at sites RGD–4 or RGD+10 retained viral replication capability in cell culture and suckling mice, as demonstrated in this study. Similar to previous studies (Seago et al. 2012; Wang et al. 2012; Lawrence et al. 2013), the tagged viruses reacted in vitro with anti-FLAG antibodies or anti-FMDV antibodies(data not shown), confirming their antigenicity.

FMDV mediated cell entry by interacting with an integrin receptor on the cell surface via the virus’ RGD motif, as elucidated with several different well-studied strain-specific G-H loop monoclonal antibodies (mAbs) (Verdaguer et al.1999; Jackson et al. 2003; Wang et al. 2011). Cell-adapted FMDV variants that lacked the RGD motif-infected cells using HS or a cell surface glycosaminoglycan (GAG) (Sa-Carvalho et al. 1997; Fry et al. 1999); additionally, FMDV reportedly adapted to cells in culture utilizing other pathways (Martinez et al. 1991; Baxt and Mason 1995). Thus, there were at least three different mechanisms for cell recognition by FMDV.For our tagged viruses, the presence of the complete P1 sequence resulted in the expression of mixtures of the original E codon (GAG) and K codon (AAG) at position 83 of VP1 by vO/CHA/90-FLAG, while vFLAG-O/CHA/90 only exhibited a K at this position, which represented a change from the original plasmid (Table 2). This phenomenon was again verified via recombinant virus adaptation in the BHK cell culture system (Baranowski et al. 2000; Zhao et al.2003; Maree et al. 2010) and increased the ability of the virus to bind to HS under physiological salt conditions. We predicted that tagged virus utilized integrin-mediated or HS-mediated cell entry (Burman et al. 2006; Dicara et al.2008), preserving accessibility to the major antigenic site involved in virus neutralization. Tagged viruses were unable to infect any of the CHO cell lines (including CHO-K1, CHO-618, CHO-745 and CHO-677); thus, it appeared that the tagged viruses infected cells via the integrin receptor. To obtain additional evidence, a peptide blocking assay was performed. Two synthetic RGD peptides representing the VP1 G-H loop of the vO/CHA/90 virus strongly inhibited infection by the tagged viruses in BHK-21 cells, while the KGE control peptide (RGD→KGE) demonstrated only low inhibition of parental virus infection. Taken together, these results confirmed that FLAG insertion did not interfere with RGD motif recognition by the cell surface integrin receptor,which was consistent with a previous report (Seago et al.2012). Interestingly, when inhibited by peptide1 or peptide2,vFLAG-O/CHA/90 exhibited differential receptor binding compared to vO/CHA/90-FLAG (Fig. 3), indicating that FLAG tag insertion at different positions in the RGD motif impacted the interaction between RGD and integrin or the display of viral structures.

Inserting site was crucial to develop a marker FMDV vaccine that could induce both marker-antibody and FMD-antibody in vivo. Numerous studies attempted to solve this issue on different types of viruses, but no proper site has been found on pig-adapted FMD so far. Here, two inserting sites (RGD–4 and RGD+10, respectively) on pig-adapted O/CHA/90 FMD were analyzed, and RGD–4 was found to be a good inserting site after a series of comparison study.A FLAG tag was stably expressed over serial passages in BHK-21 cells and suckling mice. The virulence and characteristics of tagged viruses were similar to those of the parental virus. To investigate these tagged serotype O pig-adapted viruses for candidate vaccine production, the immune effects mediated by these viruses in BALB/c mice and pigs were evaluated. As previously reported (Seago et al. 2012; Wang et al. 2012; Lawrence et al. 2013), FLAG tag insertion at different positions in the G-H loop did not affect viral viability or the reaction between the virus and convalescent serum but did affect marker epitope display and subsequently interfered with immune response induction. In one report, a tagged classical swine fever virus,FlagT4v, induced a strong anti-FLAG antibody response in pigs (Holinka et al. 2009). In another report describing a marker virus with G-H loop amino acids replaced by FLAG, virus-infected animals failed to generate detectable antibodies against the marker epitope (Lawrence et al.2013). Recently, a FLAG tag and a His tag were inserted in the VP1 G-H loop downstream of the RGD motif+9, and anti-FLAG antibodies were detected in vaccinated mouse serum(Yang et al. 2015). Currently, an inactivated FMDV vaccine emulsified with adjuvant represents the major preventive strategy to control this disease in the field. In this study,BALB/c mice and pigs were vaccinated with inactivated marker viruses; subsequent immune screening detected an anti-FMDV humoral response in both species, but anti-FLAG antibodies were only found in vFLAG-O/CHA/90(RGD–4)-vaccinated animals. Thus, different insertion positions may alter the structure of the surrounding amino acids and subsequently affect the display and immunogenicity of the inserted epitope. Nevertheless, FLAG insertion position did not affect viral antigenic epitope display, and the insertion itself was stable over several passages. Based on these data, the tagged virus with FLAG inserted upstream of the RGD motif possesses sufficient resilience to properly display both the original FMDV antigenic epitopes and the inserted epitopes without interrupting virus replication or hindering the immunogenicity of the inserted marker. After several passages, amino acid mutations were detected in both tagged viruses and the parental virus in the upstream and downstream RGD motifs. The only mutation found in these passaged viruses was the VP1 E83K mutation, which was repeatedly observed in BHK-21-adapted FMDV. In the FLAG-tagged viruses, 3-aa mutations were detected in vFLAG-O/CHA/90 only (VP3 F54L, VP1 N142S and D188G), while one mutation was found in vO/CHA/90-FLAG only (VP3 M86L). These mutations did not directly interact with the G-H loop or the FLAG epitope (Fig. 5) but did affect viral structure and antigenic variation, which may impact viral antigenic epitope display and antibody induction by chimeric viruses (Guex and Peitsch 1997; Grell et al. 2006).Additionally, the X-ray crystallography-derived structures of VP1-3 showed viral structural difference between tagged virus. Maybe the 3-aa were beneficial for FLAG epitope display. It remains unclear whether it was these mutations that led to differential FLAG immunogenicity or the different insertion positions driving compensatory mutations.

Fig. 5 Structural representation of O/CHA/90 and the tagged promoter without VP4. A, O/CHA/90. B, vFLAG-O/CHA/90.C, vO/CHA/90-FLAG. Models were obtained using the SWISS-MODEL program (Guex and Peitsch 1997), available at http://www.expasy.ch, and superimposed onto the X-ray crystallography-derived structure of O1BFS (PDB accession no. 1FOD). Using cartoons derived from PyMol (Grell et al. 2006), VP1 (blue), VP2 (green), and VP3 (red), the G-H loop (yellow), the RGD motif(pink), and the FLAG epitope (white) are shown. The amino acid mutations in VP3 (54, 86) and VP1 (142, 188) are shown using space-filling representations. The pink ball and arrows represent differences among viruses.

5. Conclusion

Insertion of the FLAG epitope upstream or downstream of the RGD motifs (RGD–4 or RGD+10) of the G-H loop in type O Cathay topotype FMDV resulted in successful rescue of the marker viruses vFLAG-O/CHA/90 and vO/CHA/90-FLAG. The tagged viruses exhibited similar infectivity in BHK-21 cells and virulence in suckling mice comparable to the characteristics of the parental virus. Moreover, the tagged viruses showed reactivity to both an anti-FMDV VP1 antibody and an anti-FLAG antibody, and their ability to infect cells via integrin recognition remained unchanged.However, only vaccination with vFLAG-O/CHA/90 (RGD–4)induced anti-FLAG antibody production in mice and pigs,suggesting that, the G-H loop upstream (–4) of the RGD motif was a novel and promising insertion site. Additionally,the 3-aa mutations in the structural protein P1 may be beneficial for FLAG epitope display.

Acknowledgements

This study was supported by the National Key Research and Development Program of China (2016YFD0501500)and the Special Fund for Agro-scientific Research in the Public Interest, China (201303046).

Acharya R, Fry E, Stuart D, Fox G, Rowlands D, Brown F. 1989.The three-dimensional structure of foot-and-mouth disease virus at 2.9 A resolution. Nature, 337, 709–716.

Alexandersen S, Donaldson A I. 2002. Further studies to quantify the dose of natural aerosols of foot-and-mouth disease virus for pigs. Epidemiology and Infection, 128,313–323.

Baranowski E, Ruiz-Jarabo C M, Lim F, Domingo E. 2001. Footand-mouth disease virus lacking the VP1 G-H loop: The mutant spectrum uncovers interactions among antigenic sites for fitness gain. Virology, 288, 192–202.

Baranowski E, Ruiz-Jarabo C M, Sevilla N, Andreu D, Beck E, Domingo E. 2000. Cell recognition by foot-and-mouth disease virus that lacks the RGD integrin-binding motif:Flexibility in aphthovirus receptor usage. Journal of Virololy,74, 1641–1647.

Barteling S J, Meloen R H. 1974. A simple method for the quantification of 140S particles of foot-and-mouth disease virus (FMDV). Arch Gesamte Virusforsch, 45, 362–364.

Baxt B, Mason P W. 1995. Foot-and-mouth disease virus undergoes restricted replication in macrophage cell cultures following Fc receptor-mediated adsorption. Virology, 207,503–509.

Beard C W, Mason P W. 2000. Genetic determinants of altered virulence of Taiwanese foot-and-mouth disease virus.Journal of Virololy, 74, 987–991.

Bittle J L, Houghten R A, Alexander H, Shinnick T M, Sutcliffe J G, Lerner R A, Rowlands D J, Brown F. 1982. Protection against foot-and-mouth disease by immunization with a chemically synthesized peptide predicted from the viral nucleotide sequence. Nature, 298, 30–33.

Burman A, Clark S, Abrescia N G, Fry E E, Stuart D I, Jackson T. 2006. Specificity of the VP1 GH loop of foot-and-mouth disease virus for alpha v integrins. Journal of Virololy, 80,9798–9810.

Curry S, Fry E, Blakemore W, Abu-Ghazaleh R, Jackson T,King A, Lea S, Newman J, Rowlands D, Stuart D. 1996.Perturbations in the surface structure of A22 Iraq foot-andmouth disease virus accompanying coupled changes in host cell specificity and antigenicity. Structure, 4, 135–145.

Dicara D, Burman A, Clark S, Berryman S, Howard M J, Hart I R, Marshall J F, Jackson T. 2008. Foot-and-mouth disease virus forms a highly stable, EDTA-resistant complex with its principal receptor, integrin alphavbeta6: Implications for infectiousness. Journal of Virololy, 82, 1537–1546.

Dunn C S, Donaldson A I. 1997. Natural adaption to pigs of a Taiwanese isolate of foot-and-mouth disease virus.Veterinary Record, 141, 174–175.

Fischer D, Rood D, Barrette R W, Zuwallack A, Kramer E, Brown F, Silbart L K. 2003. Intranasal immunization of guinea pigs with an immunodominant foot-and-mouth disease virus peptide conjugate induces mucosal and humoral antibodies and protection against challenge. Journal of Virololy, 77,7486–7491.

Fry E E, Lea S M, Jackson T, Newman J W, Ellard F M,Blakemore W E, Abu-Ghazaleh R, Samuel A, King A M,Stuart D I. 1999. The structure and function of a foot-andmouth disease virus-oligosaccharide receptor complex.The EMBO Journal, 18, 543–554.

Grell L, Parkin C, Slatest L, Craig P A. 2006. EZ-Viz, a tool for simplifying molecular viewing in PyMOL. Biocheministry of Molecular Biology Education, 34, 402–407.

Grubman M J, Baxt B. 2004. Foot-and-mouth disease. Clinical Microbiology Reviews, 17, 465–493.

Guex N, Peitsch M C. 1997. SWISS-MODEL and the Swiss-PdbViewer: An environment for comparative protein modeling. Electrophoresis, 18, 2714–2723.

Holinka L G, Fernandez-Sainz I, O’Donnell V, Prarat M V, Gladue D P, Lu Z, Risatti G R, Borca M V. 2009.Development of a live attenuated antigenic marker classical swine fever vaccine. Virology, 384, 106–113.

Jackson T, Ellard F M, Ghazaleh R A, Brookes S M, Blakemore W E, Corteyn A H, Stuart D I, Newman J W, King A M.1996. Efficient infection of cells in culture by type O footand-mouth disease virus requires binding to cell surface heparan sulfate. Journal of Virololy, 70, 5282–5287.

Jackson T, King A M, Stuart D I, Fry E. 2003. Structure and receptor binding. Virus Research, 91, 33–46.

Jackson T, Sharma A, Ghazaleh R A, Blakemore W E, Ellard F M, Simmons D L, Newman J W, Stuart D I, King A M. 1997.Arginine-glycine-aspartic acid-specific binding by foot-andmouth disease viruses to the purified integrin alpha(v)beta3 in vitro. Journal of Virololy, 71, 8357–8361.

Knowles N J, Davies P R, Henry T, O’Donnell V, Pacheco J M,Mason P W. 2001. Emergence in Asia of foot-and-mouth disease viruses with altered host range: Characterization of alterations in the 3A protein. Journal of Virololy, 75,1551–1556.

Knowles N J, Samuel A R. 2003. Molecular epidemiology of foot-and-mouth disease virus. Virus Research, 91, 65–80.

Lawrence P, Pacheco J M, Uddowla S, Hollister J, Kotecha A, Fry E, Rieder E. 2013. Foot-and-mouth disease virus(FMDV) with a stable FLAG epitope in the VP1 G-H loop as a new tool for studying FMDV pathogenesis. Virology,436, 150–161.

Lee S H, Jong M H, Huang T S, Lin Y L, Wong M L, Liu C I, Chang T J. 2009. Pathology and viral distributions of the porcinophilic foot-and-mouth disease virus strain(O/Taiwan/97) in experimentally infected pigs.Transboundary and Emerging Diseases, 56, 189–201.

Logan D, Abu-Ghazaleh R, Blakemore W, Curry S, Jackson T,King A, Lea S, Lewis R, Newman J, Parry N, Rowlands D,Stuart D, Fry E. 1993. Structure of a major immunogenic site on foot-and-mouth disease virus. Nature, 362, 566–568.

Mahapatra M, Hamblin P, Paton D J. 2012. Foot-and-mouth disease virus epitope dominance in the antibody response of vaccinated animals. Journal of General Virology, 93,488–493.

Mandl C W, Kroschewski H, Allison S L, Kofler R, Holzmann H, Meixner T, Heinz F X. 2001. Adaptation of tick-borne encephalitis virus to BHK-21 cells results in the formation of multiple heparan sulfate binding sites in the envelope protein and attenuation in vivo. Journal of Virololy, 75, 5627–5637.

Maree F F, Blignaut B, de Beer T A, Visser N, Rieder E A. 2010.Mapping of amino acid residues responsible for adhesion of cell culture-adapted foot-and-mouth disease SAT type viruses. Virus Research, 153, 82–91.

Martinez M A, Hernandez J, Piccone M E, Palma E L, Domingo E, Knowles N, Mateu M G. 1991. Two mechanisms of antigenic diversification of foot-and-mouth disease virus.Virology, 184, 695–706.

Mason P W, Grubman M J, Baxt B. 2003. Molecular basis of pathogenesis of FMDV. Virus Research, 91, 9–32.

Mateu M G, Camarero J A, Giralt E, Andreu D, Domingo E.1995. Direct evaluation of the immunodominance of a major antigenic site of foot-and-mouth disease virus in a natural host. Virology, 206, 298–306.

McCullough K C, Crowther J R, Carpenter W C, Brocchi E,Capucci L, De Simone F, Xie Q, McCahon D. 1987. Epitopes on foot-and-mouth disease virus particles. I. Topology.Virology, 157, 516–525.

Meloen R H, Briaire J, Woortmeyer R J, van Zaane D. 1983.The main antigenic determinant detected by neutralizing monoclonal antibodies on the intact foot-and-mouth disease virus particle is absent from isolated VPI. Journal of General Virology, 64, 1193–1198.

Neff S, Sa-Carvalho D, Rieder E, Mason P W, Blystone S D,Brown E J, Baxt B. 1998. Foot-and-mouth disease virus virulent for cattle utilizes the integrin alpha(v)beta3 as its receptor. Journal of Virololy, 72, 3587–3594.

OIE (World Organisation for Animal Health). 2016. Chapter 2.1.8. Foot and mouth disease (infection with foot and mouth disease virus). In: Manual of Diagnostic Tests and Vaccines for Terrestrial Animals 2016. NB: version adopted in May 2017. http://www.oie.int/fileadmin/Home/eng/Health_standards/tahm/2.01.08_FMD.pdf

Orsel K, de Jong M C, Bouma A, Stegeman J A, Dekker A.2007. Foot and mouth disease virus transmission among vaccinated pigs after exposure to virus shedding pigs.Vaccine, 25, 6381–6391.

Park M E, Lee S Y, Kim R H, Ko M K, Lee K N, Kim S M, Kim B K, Lee J S, Kim B, Park J H. 2014. Enhanced immune responses of foot-and-mouth disease vaccine using new oil/gel adjuvant mixtures in pigs and goats. Vaccine, 32,5221–5227.

Parry N, Fox G, Rowlands D, Brown F, Fry E, Acharya R, Logan D, Stuart D. 1990. Structural and serological evidence for a novel mechanism of antigenic variation in foot-and-mouth disease virus. Nature, 347, 569–572.

Rieder E, Baxt B, Lubroth J, Mason P W. 1994. Vaccines prepared from chimeras of foot-and-mouth disease virus(FMDV) induce neutralizing antibodies and protective immunity to multiple serotypes of FMDV. Journal of Virololy,68, 7092–7098.

Rodriguez L L, Barrera J, Kramer E, Lubroth J, Brown F, Golde W T. 2003. A synthetic peptide containing the consensus sequence of the G-H loop region of foot-and-mouth disease virus type-O VP1 and a promiscuous T-helper epitope induces peptide-specific antibodies but fails to protect cattle against viral challenge. Vaccine, 21, 3751–3756.

Sa-Carvalho D, Rieder E, Baxt B, Rodarte R, Tanuri A, Mason P W. 1997. Tissue culture adaptation of foot-and-mouth disease virus selects viruses that bind to heparin and are attenuated in cattle. Journal of Virololy, 71, 5115–5123.

Salguero F J, Sanchez-Martin M A, Diaz-San Segundo F, de Avila A, Sevilla N. 2005. Foot-and-mouth disease virus(FMDV) causes an acute disease that can be lethal for adult laboratory mice. Virology, 332, 384–396.

Seago J, Jackson T, Doel C, Fry E, Stuart D, Harmsen M M,Charleston B, Juleff N. 2012. Characterization of epitopetagged foot-and-mouth disease virus. Journal of General Virology, 93, 2371–2381.

Skinner H H. 1951. Propagation of strains of foot-and-mouth disease virus in unweaned white mice. Proceeding of the Royal Society of Medicine, 44, 1041–1044.

Verdaguer N, Schoehn G, Ochoa W F, Fita I, Brookes S, King A, Domingo E, Mateu M G, Stuart D, Hewat E A. 1999.Flexibility of the major antigenic loop of foot-and-mouth disease virus bound to a Fab fragment of a neutralising antibody: Structure and neutralisation. Virology, 255,260–268.

Wang C Y, Chang T Y, Walfield A M, Ye J, Shen M, Zhang M L, Lubroth J, Chen S P, Li M C, Lin Y L, Jong M H, Yang P C, Chyr N, Kramer E, Brown F. 2001. Synthetic peptidebased vaccine and diagnostic system for effective control of FMD. Biologicals, 29, 221–228.

Wang H, Xue M, Yang D, Zhou G, Wu D, Yu L. 2012. Insertion of type O-conserved neutralizing epitope into the foot-andmouth disease virus type Asia1 VP1 G-H loop: Effect on viral replication and neutralization phenotype. Journal of General Virology, 93, 1442–1448.

Wang H, Zhao L, Li W, Zhou G, Yu L. 2011. Identification of a conformational epitope on the VP1 G-H Loop of type Asia1 foot-and-mouth disease virus defined by a protective monoclonal antibody. Veterinary Microbiology,148, 189–199.

Yang B, Yang F, Zhang Y, Liu H, Jin Y, Cao W, Zhu Z, Zheng H, Yin H. 2015. The rescue and evaluation of FLAG and HIS epitope-tagged Asia 1 type foot-and-mouth disease viruses. Virus Research, 213, 246–254.

Yang P C, Chu R M, Chung W B, Sung H T. 1999.Epidemiological characteristics and financial costs of the 1997 foot-and-mouth disease epidemic in Taiwan.Veterinary Record, 145, 731–734.

Zhao Q, Pacheco J M, Mason P W. 2003. Evaluation of genetically engineered derivatives of a Chinese strain of foot-and-mouth disease virus reveals a novel cell-binding site which functions in cell culture and in animals. Journal of Virololy, 77, 3269–3280.