Introduction

Foot-and-mouth disease (FMD) is one of the most highly contagious diseases in cloven-hoofed animals, particularly in cattle, pigs, and sheep. FMD is caused by the FMD virus (FMDV) that is a positive-stranded RNA virus belonging to the Aphthovirus genus in the family Picornaviridae. Seven serotypes (A, O, C, Asia 1, and South African Territories 1, 2, and 3) of FMDV have been identified serologically, and multiple subtypes of each serotype of FMDV have been discovered [1, 2].

The FMDV RNA genome is approximately 8.5 kb in length and contains a single long open reading frame (ORF) encoding a polypeptide that is processed into mature proteins by virally encoded proteinases [3, 4]. These mature proteins include four structural proteins (VP1, VP2, VP3, and VP4) and eight non-structural proteins (Lpro, 2A, 2B, 2C, 3A, 3B, 3Cpro, and 3D) [5–7]. The 5′ end of RNA is covalently bound to a 23–24 amino-acid residue genome-linked protein (3B, also known as VPg) [7, 8]. The FMDV genome also contains a 5′ untranslated region (UTR) and a 3′ UTR [9].

Reverse genetics is critical in the study of RNA viruses. The availability of full-length FMDV cDNA clones offers an opportunity for analysis and modification of viral genomes at the molecular level, and is helpful for research on virus replication, pathogenesis, and vaccine development [10]. Previous studies have successfully generated the full-length infectious cDNA clones of serotype O and A of FMDV [11–14]. Infectious cDNA clone of serotype Asia 1 FMDV was seldom reported. Chinese serotype Asia1 FMDV was first isolated in Baoshan City, Yunnan Province, in 1958 (Asia1/YNBS/58). A new Jiangsu lineage of Asia1/JS/CHA/05, which is different from Asia1/YNBS/58, was isolated from cattle hosts during an FMD outbreak in Jiangsu Province of China in 2005 [15]. The strain of As1/CHA/05 FMDV has 98.3% similarity in nucleotide sequence with As1/JS/CHA/05. We have generated a full-length cDNA clone of FMDV As1/CHA/05 [16]. However, little is known about the molecular basis of viral infectivity and species specificity of As1/CHA/05 FMDV.

While the first half of the 153 amino-acids of 3A is highly conserved, the second half varies among FMDV strains [17]. Beard and Mason [11] have shown that a deletion of amino-acids 93-102 of 3A protein is associated with an inability of a Taiwanese strain of FMDV (OTai) isolated from swine to cause disease in bovines [18]. Interestingly, the similar mutant of OTai is found in egg-adapted derivatives (O1C-O/E and C3R-O/E) of FMDV and has been used as vaccines in South America [18–21]. However, the deletion in 3A protein of these strains is different from other reference viruses. The OTai contains numerous mutations located between codons 128 and 147 [11]. There are mutations in the capsid proteins and precursor polypeptide 3ABCD (P3) of both attenuated strains O1C-O/E and C3R-O/E. In addition, there are 57 (sequence of 85-103) and 60 (sequence of 88-107) nucleotide deletions in the 3A region of the attenuated strains, which were longer than the deletion in OTai [19]. Hence, analysis of the precise changes involved in the attenuated phenotype in cattle should be addressed through the use of infectious cDNA clones. Natural mutations in the 3A protein are often detected in serotype O FMDV. However, the impact of the deletion at position 93-102 of 3A on the infectivity, immunoreactivity, and species specificity of the serotype Asia 1 FMDV has not been explored.

In this study, the recombinant virus (rvAs-3A10D) carrying the deletion at position 93-102 of 3A of the As1/CHA/05 strain was generated, and its immunoreactivity, infectivity, in vivo virulence, and host-cell tropism were compared with that of the wild-type of virus. The rvAs-3A10D was unique as its parent wide-type virus was isolated from cattle and it also was different from the attenuated strains O1C-O/E and C3R-O/E. We found that the mutant virus retained immunoreactivity, infectivity, and in vivo virulence similar to that of the wild-type of virus. Although the mutant virus was capable of replication in the baby hamster kidney cell line (BHK-21) and porcine kidney cell line (PK-15), it had significantly lower ability to replicate in calf kidney cells. Therefore, the position 93-102 of 3A protein is required for effective replication of this Asia 1 FMDV strain in calf cells.

Materials and methods

Cell line, virus, and plasmids

Baby hamster kidney cell line (BHK-21) and porcine kidney cell line (PK-15) were maintained in Dulbecco’s Modified Eagle Medium (DMEM, Gibco, USA) containing 10% fetal calf serum (FCS). The parental FMDV As1/CHA/05 strain was isolated from cattle in China during the 2005 outbreak. Plasmid pBSAs containing the full-length cDNA of FMDV As1/CHA/05 (Fig. 1a) and the recombinant virus (termed rvAs) derived from pBSAs were prepared, as described previously [16].

Fig. 1
figure 1

Schematic diagram of the strategy for construction of cDNA clone. a The full-length As1/CHA/05 virus cDNA clone, pBSAs. A T7 RNA polymerase promoter was incorporated upstream of the 5′ terminus to facilitate transcription. In addition, EcoRV sites were incorporated both at the 3′ and 5′ termini to allow linearization of the sequence. b To delete the amino acids 93 to 102 of 3A, 3AL (about 1520 bp), and 3AR (about 522 bp) were amplified from pBSAs by PCR. c The fusion PCR product contains the mutant 3A. d The full-length 3A was replaced by the mutant 3A fragment by restriction endonuclease digestion and linkage. e The 5′ and 3′ termini of cDNA contained in the clone pBSAS-3A10D

Full size image

Construction of mutant genome-length cDNA of FMDV As1/CHA/05

Two DNA fragments of 3AL and 3AR were amplified from the plasmid pBSAs by PCR, using pfu DNA polymerase and the primers (L1/L2 and R1/R2, Table 1, Fig. 1b). Subsequently, these two DNA fragments were linked by fusion PCR with the primers of L1/R2 to generate a DNA fragment with a 30-bp deletion between the L2 and R1 primers (Fig. 1c). The mutant fragment was cloned into the pBSAs with unique restriction sites (EcoRT22 I and Sma I) to substitute the full-length of the 3A region. After transformation to competent Escherichia coli JM109 cells, the recombinant plasmids were screened and characterized by restriction endonuclease analysis and sequencing. The recombinant plasmid was named as pBSAs-3A10D (Fig. 1d). In addition, this cDNA clone contained a modified 5′ end characterized by a EcoRV site upstream of a T7 polymerase promoter fused to the 5′ terminus of the coding strand and a modified 3′ end characterized by a EcoRV site immediately following the 3′ terminal poly(A) (Fig. 1e).

Table 1 The sequences of primers
Full size table

In vitro RNA synthesis

The pBSAs-3A10D was digested with EcoRV to generate a 8.2 kb DNA fragment containing the T7 promoter at the 5-terminus and the poly(A) tract at the 3-terminus. After purified by the gel clean-up system (Promega, USA), this DNA fragment was used as a template for in vitro transcription using the T7 RiboMAXTM express large scale RNA production system (Promega), according to the manufacturer’s instruction. After digesting with DNase, the synthesized RNA was extracted with phenol/chloroform and precipitated in ethanol. The RNA concentrations were determined spectrometrically.

Generation of viruses

BHK-21 cells (2 × 105/ml) were cultured in six-well plates, and at 80% confluency, the cells were transfected with 5 μg of in vitro transcribed mutant RNA mixed with 10 μl DMRIE-C regent for 48 h, according to the manufacture’s protocol (Invitrogen). The supernatants of the cultured cells were harvested and passed in BHK-21 cells up to the appearance of cytopathic effect (CPE) for five passages. The cells were harvested from each passage and virions in the transfected cells were purified by freeze-thawing of the cells and centrifugation. The viral titers were characterized by CPE, and the virus stocks of rvAs-3A10D were aliquoted for subsequent experiments.

Characterization of rvAs-3A10D

Monolayer BHK-21 cells were infected with rvAs-3A10D or parental virus for 12 h, respectively. Their supernatants were harvested and centrifuged at 6000g and 4°C for 30 min. The supernatants were further ultra-centrifuged at 125000g and 4°C for 3 h through a 25% sucrose cushion in NTE (NaCl 100 mM, Tris–HCl 10 mM, EDTA 1 mM, pH 7.4) to pellet the virions [12]. The virions were re-suspended in NTE buffer and reacted with anti-FMDV type Asia 1 monoclonal antibody (mAb 3E11 with unknown epitope specificity, 1:100, Haerbin Veterinary Research Institute, PR China) at 37°C for 2 h, followed by centrifugation at 8000g and 4°C for 30 min. The precipitates were re-suspended in PBS and loaded onto carbon-shadowed, formvar-coated grids, followed by staining with 1% phosphotungstate pH 7.0 (Sigma) and examining under a JEM-1200EX transmission electron microscope. The rescued virus, rvAs-3A10D, was reverse-transcribed into cDNA by RT–PCR using the specific primers and characterized by sequencing.

The infected cells were washed three times with PBS and fixed in 4% paraformaldehyde for 15 min. The cells were probed with anti-FMDV mAb 3E11 (1:2000) or isotype control at 37°C for 1 h. After washing, the bound antibodies were detected with FITC-conjugated goat anti-mouse IgG (1:2000, Sigama) for 1 h. The specific staining of the cells was examined under a fluorescence microscope (Nikon).

Plaque assays

Monolayer of BHK-21 cells, PK-15, and calf kidney cells in 12-well tissue culture plates were infected with 0.5 ml of serial 10-fold diluted parental virus, rvAs and rvAs-3A10D for 1 h, respectively. After removing the unabsorbed virus, the cells were cultured with 1 ml of Eagle’s minimal essential media containing 5% FCS and 1% agar (Promega). The cells were cultured at 37°C in 5% CO2 for 48 h and stained with 0.01% neutral red. The formed plaques were quantified.

Virus replication in kidney cells

The primary calf kidney cells, PK-15, and BHK-21 cells were used to determine the replication of viruses in vitro. The primary calf kidney cells, PK-15, and BHK-21 cells were infected with serial 10-fold dilutions of parental virus, rvAs and rvAs-3A10D, respectively (starting with 107 PFU/ml, determined in BHK-21 cells) for 2 days. The CPE of individual types of viruses in each type of cells was measured and used for the calculation of the 50% tissue culture infectious dose(s) (TCID50) [22].

BHK-21, PK-15 and primary calf kidney cells at 106/well were cultured in six-well plates overnight and infected with parental virus, rvAs and rvAs-3A10D at a MOI of 0.01 for multicycle growth or 5 for one-step growth, respectively, for the varying periods. The contents of virions were determined by TCID50 assays on BHK-21 cells.

Virulence in mice

Sucking mice at 2 days of age were inoculated subcutaneously with different doses of rvAs-3AD10 or parental virus in 0.2 ml PBS containing 1% FBS for 72 h (5 mice per group). The mice were monitored and their deaths were recorded. The lethal dose50 (LD50) of each type of virus was determined, as described previously [22]. The experimental protocols were approved by the Animal Research Protection Committee of the Northeast Agriculture University (China).

Results

Generation of FMDV deletion mutants

In order to generate a recombinant FMDV deletion mutant virus, two DNA fragments in the 3A region were amplified from the plasmid of pBSAs by PCR, fused, and replaced the related region in the wild-type virus in the plasmid of pBSAs to generate a recombinant plasmid of pBSAS-3A10D containing a 30-bp deletion corresponding to amino-acid positions 93-102 of 3A (Fig. 1). After DNA sequencing, the linearized DNA fragments of the pBSAS-3A10D were transcribed in vitro into RNA. Subsequently, the resulting RNA was transfected into BHK-21 cells and the supernatants were harvested for continual passages. In the third passage, the cells infected with rvAs-3A10D developed typical cytopathy (Fig. 2a). The recombinant rvAs-3A10D was recovered from the collected cells and reached a titer of 106.9 TCID50/50 μl.

Fig. 2
figure 2

Characterization of the recombinant rvAs-3A10D virus. BHK-21 cells were infected with, or without, rvAs-3A10D or wild-type virus, respectively. a The CPE was caused by the rvAs-3A10D. b The CPE was caused by the parental virus. c Uninfected BHK-21 cells. The cells were probed with anti-FMDV serotype Asia 1-specific mAb and the bound antibodies were detected with FITC-anti-mouse IgG, followed by examining under a fluorescent microscope. d BHK-21 cells infected with the rvAs-3A10D. e BHK-21 cells infected with the parental virus. f Uninfected BHK-21 cells. Immuno-electron micrograph of the rvAs-3A10D. g The rvAs-3A10D. h The parental virus (bar represents 50 nm)

Full size image

Characterization of the recombinant rvAs-3A10D virus

In order to characterize the mutant virus, total RNA was extracted from the mutant and wild-type virions, and reverse-transcribed into cDNA, respectively. Subsequently, a fragment (about 680 bp) containing the deletion region (position 93-102 of 3A) was amplified by PCR. The sequence of rvAs-3A10D virus was confirmed by sequencing (data not shown), demonstrating that the recombinant rvAs-3A10D virus carried the deletion and was stable during in vitro passages in BHK-21 cells.

In order to determine the immunoreactivity of the mutant virus, monolayer BHK-21 cells were infected with rvAs-3A10D or parental virus for 10 h, and probed with anti-viral mAb by indirect immunofluorescence assays (Fig. 2d–e). There was no significant difference in the frequency and intensity of FITC-positive cells, demonstrating that the deletion in the 3A gene did not affect the expression of antigen recognized by this mAb. Furthermore, the rvAs-3A10D and parental virus were reacted with anti-viral mAb, and the immunocomplex was examined under an electron microscope (Fig. 2g). The rvAs-3A10D virions were spherical with a diameter of about 25 nm and indistinguishable from the wild-type virions. These data indicated the deletion in these aa in 3A protein did not affect the morphology of FMDV particles.

Infectivity of the mutant virus

In order to determine the infectivity of the mutant virus, calf kidney, BHK-21, and PK15 cells were infected with different doses of recombinant rvAs-3A10D, rvAs and parent wide-type of virus and the CPE of individual types of virus in each type of cells were characterized. The plaque size and morphology induced by recombinant rvAs-3A10D and rvAs were similar to those induced by parental virus in BHK-21 and PK-15 cells. The virus titers were at 3 × 107 to 6 × 107 PFU/ml in BHK-21 cells and 1 × 107 to 4 × 107 PFU/ml in PK-15 cells. Interestingly, while parental virus and rvAs were able to form plaques in these cells with the titers of 1 × 106 to 3 × 106 PFU/ml the mutant virus rvAs-3A10D failed to form plaque on calf kidney cells.

The rvAs-3A10D replicates in pig, but not in calf kidney cells

In order to further determine the infectivity in vitro, BHK-21, PK-15, and primary calf kidney cells were infected with different doses of rvAs-3A10D, rvAs or wild-type virus for 48 h, and the CPE of the cells were monitored. The lowest dose of the virus able to cause complete CPE was calculated as TCID50 in Fig. 3. All of the viruses tested showed similar ability to cause CPE in BHK-21 cells and PK-15 cells at 48 h post-inoculation. Furthermore, the parental virus (As1/CHA/05) and rvAs expressing full-length 3A replicated well in primary calf kidney cells, whereas the mutant rvAs-3A10D was unable to cause CPE at any of the dilutions tested. These data indicated that a full-length 3A was required for the effective replication of As1/CHA/05 in calf-derived kidney cells.

Fig. 3
figure 3

The viral replication in different types of cells. BHK-21, PK-15, and the prepared calf primary kidney cells were infected with 107 PFU/ml of rvAs-3A10D, rvAs, or its wild-type virus for 48 h, respectively. The cells were harvested and the contents of each type of virions were determined by TCID50 assay. TCID50 values were determined. Data are expressed as mean ± SEM of log TCID50/50 μl for each type of virus from three independent experiments

Full size image

Growth kinetics of the recovered virus

In order to characterize the viral replication, rvAs-3A10D, rvAs, and parental virus were compared simultaneously in BHK-21, PK-15, and calf kidney cells. The viral replications were determined by measuring the viral titers by TCID50 (Fig. 4). Apparently, the recombinant rvAs-3A10D, rvAs, and the parent viruses displayed similar replication kinetics in BHK-21 cells and PK-15 cells. While parental wide-type and rvAs replicated well with the tiers of about 106 TCID50/ml in calf kidney cells the rvAs-3A10D failed to replicate in the cells because the viral titer was below the detection limit of 100.5 TCID50/50 μl. Furthermore, analysis of multicycle growth curves of different viruses revealed that the replication kinetics of parental virus and rvAs were similar, which was faster than that of rvAs-3A10D at earlier time points, but was not significant difference in the later time points in in BHK-21 cells or PK-15 cells. Interestingly, while a typical multi-cycle growth curve was observed in the parental virus and rvAs-infected calf kidney cells, there was little virus detected in the rvAs-3A10D-infected calf kidney cells. Therefore, the deletion of 10 aa in 3A interfered with the viral replication in bovine cells, but did not in BHK-21 and PK-15 cells.

Fig. 4
figure 4

One-step growth and multicycle replication curve of recombinant viruses. The indicated cells were infected with rvAs-3A10D, rvAs, or wild-type virus at a MOI of 5 (one-step growth curve; a) or 0.01 (multicycle growth; b). The cells were harvested at the indicated time points and the contents of virions were determined by TCID50 assays. Data are expressed as mean of log TCID50/50 μl for each type of virus at different times post-infection from three independent experiments. Intragroup variations were less than 20% of mean (data not shown) and experimental and control viruses were simultaneously analyzed

Full size image

Mouse virulence studies

In order to determine the virulence of rv-3A10D, sucking mice at 2 days of age were infected subcutaneously with different doses of rv-3A10D or wild-type viruses (5 mice per group). The mice were monitored longitudinally. The first mouse infected with wild-type virus displayed signs of hind leg paralysis at 18 h post-infection and died at 30 h post-infection. Similarly, the first mouse infected with the rvAs-3A10D developed signs at 20 h post-infection and died at 33 h post-infection. The LD50 values of the rvAs-3A10D and wild-type viruses were 10−6.50 and 10−6.77, respectively. Apparently, both rvAs-3A10D and wide-type viruses had similar virulence in mice.

Discussion

The Asia 1 was the main epidemic serotype of FMDV, responsible for the outbreak of FMD in China in 2005. This severe FMD outbreak caused by Asia 1 FMDV mainly occurred in cattle. Our previous study has successfully established an infectious cDNA clone of As1/CHA/05 containing the full-sequence 3A [16]. The 3 A protein has been shown to be crucial for host range and pathogenicity of FMDV and A single amino acid substitution in 3A can mediate adaptation of FMDV to the guinea pig [23]. Furthermore, mutations in 3A may alter the host range and virulence of FMDV and poliovirus in vivo and in cell culture. The 19- and 20-amino-acid deletions in 3A of FMDV O1 Campos and C3 Resende are associated with attenuated virulence in cattle, respectively [19]. In addition, an overlapping 10-amino-acid deletion and multiple amino acid mutations are associated with the loss of virulence of FMDV O/TAW/97 (OTai) in cattle and with impaired plaque formation by this virus in bovine kidney cells [11]. We studied the impact of 10-amino-acid deletion in 3A on the replication, infectivity, and immunoreactivity of the As1/CHA/05 strain, a unique strain of FMDV different from previous strains reported. First, we generated the mutant virus with a precise deletion of 10-amino-acid in 3A with which, we found that the mutant rvAs-3A10D could be efficiently rescued from BHK-21 cells, similar to the recombinant wide-type of rvAs. Currently, there is no report about the natural mutant of Asia 1 FMDV with the deletion of 93-102 in 3A. The successful generation of mutant rvAs-3A10D may provide a basis for further research on the relationship between the 3A protein and Asia 1 FMDV host range.

Characterization of the immunoreactivity of rvAs-3A10D revealed that the mutant rvAs-3A10D like wide-type of virus, was recognized by mAb against the virus, indicating that the mutation of 10-aa in 3A did not alter its immunoreactivity with the mAb. Furthermore, the number, size, and morphology of formed plaques were similar in the BHK-21 and PK15 cells that had been infected with the mutant rvAs-3A10D, rvAs or its parental wide-type of virus. In addition, the rvAs-3A10D replicated well in hamster and pig kidney cells with similar kinetics with the control rvAs and parental wide-type of virus in hamster and pig kidney cells. Moreover, the mutant rvAs-3A10D and its parent wide-type of virus displayed similar virulence in mice. These data indicated that the mutated 10-aa in 3A were not necessary for the viral plaque formation and replication in these two species of kidney cells and for the virulence to baby mice. The 3A protein may function in supporting viral transmission and pathogenesis in calf cells [18]. Interestingly, while the recombinant rvAs and its parental wide-type of virus replicated well and formed plaques in calf kidney cells the rvAs-3A10D failed to replicate in calf kidney cells in vitro. Evidentially, the virus titer of the rvAs-3A10D-infected calf kidney cells was less than the detection limit of 100.5 TCID50/50 μl. Apparently, the 10-aa in 3A is crucial for the replication of the As1/CHA/05 strain in calf kidney cells in vitro.

Study of the 3A function of poliovirus, a distantly related picornavirus, has shown that the 3A protein and its precursors regulate the virus-specific RNA synthesis [23]. First, the 3A can serves to anchor 3B (VPg) in the membranes through its single hydrophobic domain (residues 53–81) and support RNA synthesis [24–26]. Second, the 3AB has RNA-binding activity and can be associated with the 59 cloverleaf of poliovirus RNA and precursor 3CD to form a ribonucleoprotein complex, which appears to be essential for the poliovirus RNA synthesis [25, 27, 28]. We found that the mutant rvAs-3A10D failed to replicate in calf kidney cells. It is possible that the deleted 10-aa in 3A changed conformational structure, leading to the loss of its ability to support the viral RNA synthesis in calf kidney cells. Alternatively, the deleted 10-aa in 3A may be necessary component to regulate and support the viral RNA synthesis. We are interested in further investigating the precise mechanisms by which deletion of the 10-aa in 3A results in the failure of the As1/CHA/05 strain to replicate in calf kidney cells.

Notably, there are many factors that regulate the virulence and species specificity of FMDV. An investigation has shown non-structural protein 2C, 3A, and VPg as key determinants for modulating cytopathology in cell culture [29]. In fact, the replacements either in 2C or 3A are associated with altered virulence, cell tropism or host range in several picornaviruses [23, 30–34]. We observed the mutation with 10-aa deletion in 3A dramatically reduced the ability of the mutant rvAs-3A10D to replicate, form plaques and induce CPE in calf kidney cells. Whether other factors contribute to the attenuation of the mutant rvAs-3A10D in calf kidney cells remains further to be determined. In this present study, though the mutant virus rvAs-3A10D poorly replicated in bovine origin cells, determination of the impact of the 10-aa deletion in 3A on the virulence and species specificity of rvAs-3A10D in vivo is warranted.

In summary, we generated a 3A deletion infectious cDNA clone of Asia 1 FMDV and found that the mutant rvAs-3A10D retained immunoreactivity, morphology, infectivity, and replication, similar to its parent virus in hamster and pig kidney cells in vitro. However, the mutant rvAs-3A10D failed to replicate in bovine kidney cells in vitro, indicating that the complete 3A protein was necessary for viral replication in bovine cells. The availability of mutant rvAs-3A10D and its infectious cDNA clone may provide a basis for the design of new vaccines for controlling FMD in cattle and helping in determining the molecular mechanisms underlying the functions of viral gene products in infection and replication of serotype Asia 1 FMDV.