Genetic characterization of the cell-adapted PanAsia strain of foot-and-mouth disease virus O/Fujian/CHA/5/99 isolated from swine

FMDV O/Fujian/CHA/5/99 strain of the 6th passage producing obvious cytopathic effect (CPE) was adapted to BHK-21 cells, and formed clear large plaque. However, the FMDV bovine-isolated O/Tibet/CHA/1/99 strain formed small plaque shaped a fringe of snowflakes (Fig. 1). The virus titres of O/Fujian/CHA/5/99 (1.5 × 10⁷ PFU/ml) was no significant different from O/Tibet/CHA/1/99 (2.0 × 10⁷ PFU/ml). However, the pathogenicity in suckling mice of O/Fujian/CHA/5/99 was distinguishable from that estimated with O/Tibet/CHA/1/99. The LD₅₀ value was 10^-8.3 for O/Fujian/CHA/5/99 compared to 10^-7.0 for O/Tibet/CHA/1/99.

The genome sequence of the O/Fujian/CHA/5/99 strain is 8,172 nt (excluding the poly(C) tract and the poly(A) tail) in length including a 1,081-nt 5'-UTR which is divided into S (366 nt), PKs (219 nt), cre (54 nt), and IRES (442 nt), a 6,999-nt ORF that encodes 2,332 amino acids terminating at a "TAA" stop codon, and a 92-nt 3'-UTR. All sequences were unique and comprised the complete genome, excluding 36 primer orderly determined nucleotides [22 nt (S+ primer) at the 5' end of the viral genome, 7 and 8 nt (Pan/S-, Pan/I+ primers) on either side of the poly(C) tract] (Table 1). The full-length genomic sequence of FMDV O/Tibet/CHA/1/99 strain has been determined and submitted to GenBank (accession NO, AF506822) by Zhang et al (2003) [19].

A detailed examination of the mutations in the whole genome of the O/Fujian/CHA/5/99 strain was based on multiple sequence alignments (Table 2). The S-fragment is 366 nucleotides in length at the 5' terminus of the viral genome, which is predicted to form a large hairpin structure. Nucleotide transitions and deletions were found at positions T82C, T84C, C105T, C119T, T138C, A139G, T145C, T147C, C155T, C160T, C182T (peak-loop), T222C, C238T, C280T, T288C, T327C, C345T, and T199, A200 in O/Fujian/CHA/5/99, which compared to reference strains of PanAsia lineage. Downstream of the poly(C) tract, there is a stretch sequence of highly tolerant to changes, containing four PKs in structure of O/Fujian/CHA/5/99 (positions -1 to +218) for the maintenance of biological function. Substitutions were observed at positions T26C, A52T and T114C; A51G, C121T (including O/AS/SKR/2002); T132C and T193C (including O/JPN/2000) in O/Fujian/CHA/5/99. Notably, a 43-nt deletion started at postion 53 downstream of the poly(C) tract in O/YUN/TAW/97 strain was determined [20], resulting in the pseudoknot 2 deletion. The conserved AAACA sequence in cre is required for viral RNA genome replication, while A30G, T33C distinctively located at this hairpin loop of O/Fujian/CHA/5/99 and O/AS/SKR/2002. The IRES element consists of a five structural domains, where several conserved motifs were identified [8]. In addition, the formation of a helical structure around positions 67 (G) and 275 (C) located at the base of domain 3 is needed for efficient internal initial of FMDV RNA translation [21]. Here, the substitution of C275T in O/YUN/TAW/97, O/AS/SKR/2002 and O/Fujian/CHA/5/99 strains, may induce a reorganization of the whole element with important consequences for IRES function in cattle? The other variations were present at positions T55C (domain 2), C228T (domain 3), T312C and C389T (domain 4), T423C, C436T, T437C, and A428C, 442A in O/Fujian/CHA/5/99.

The leader (L) protein, a member of the papain-like cysteine proteinase, is located at 5' end of the ORF and contains two in-frame initiation codons (84 nt in distance, Lab/Lb), that cleaves itself from the viral polyprotein [22] acting as a trans-proteinase and initiation factor eIF4G at G₄₇₉/R₄₈₀ resulting the shut-off of host protein synthesis [23]. 51 D, 148 H and 164 D were the active site residues, by playing a essential role in substrate binding [24, 25]. It's also an important determinant of virulence in animals [26]. The amino acid sequence identities of O/Fujian/CHA/5/99 with reference PanAsia strains and O/YUN/TAW/97 was 92.0%-94.5% and 88.1%, respectively. The variable substitutions appeared in three distinct regions (A25T, Q26R at N-terminus; T55A, F68L, Y73 S, P75 H and D81 S on the C-terminal side of 51C; K144Q and Q146 H on the N-terminal side of 148H). VP4 is the most conserved FMDV protein. There was 100% homology in amino acid sequence between O/Fujian/CHA/5/99 and reference PanAsia strains. The amino acid sequence alignments of VP2 and VP3 showed that the specific substitutions of O/Fujian/CHA/5/99 existed at the residues E136G, K175R and F214L in VP2, and A174 S in VP3, respectively (Table 2). The 136th in VP2 and 174th of VP3 are very close to their respective heparin interacting regions (residues 134th, 135th in VP2, and 173rd in VP3, respectively). A phylogenetic tree was generated from VP1 nucleotide sequence alignment of 16 FMDV which caused outbreaks of FMD in EA and the Far East in 1997-2002 (Fig. 2). The identity data of VP1 showed that the O/Fujian/CHA/5/99 strain is clustered in the PanAsia lineage and closely related to O/AS/SKR/2002 (94.1% similarity). Furthermore, comparison of the amino acid sequences in VP1 of O/Fujian/CHA/5/99 and O/Tibet/CHA/1/99 indicated that 9 substitutions were found at residues Y72C, T96A, G137 D, T142A, A152T, Q153P, I168V, A199T and L212 S of O/Tibet/CHA/5/99 (Table 2). Most of these substitutions were present in C-terminal segment of VP1, in particular in G-H loop (antigenic site 1). The important integrin-binding RGD motif (145-147 residues), RGD+1, RGD+2 and RGD+4 were conserved for virus reception and pathogenesis in these FMDV strains (Fig. 2).

In non-structural protein regions, we also found the highest degree of sequence conservation in 2A, 2B, 2C, 3B, 3C and 3 D that it was predicted probably due to their functions or interaction with host factors. The characteristic amino acid mutations occurred at residues S5A, K48R in 2B; K64E, V92A, I241T, S312N in 2C; K18R in 3B1; R196K in 3C; and H27Y, K42Q, G62E, N63 D, T98I, Q210R, R234K and R440W in 3 D of O/Fujian/CHA/5/99, respectively (Table 2). Comparison of 3A protein sequences showed that O/Fujian/CHA/5/99 contains a full-length 3A coding region, whereas the 93-102 amino acid deletions harbored in 3A of O/YUN/TAW/97 (Fig. 3). I72 M was present in transmembrane domain (positions 60-76) as previously described [14]. The other 14 amino acid substitutions were identified at positions I3V; H31C and I34V; I42V and E57D; M85T, A89V and N91D; I94T and T100A; E108G, N112 S, K144E and E148G in O/Fujian/CHA/5/99 (Table 2), which predicted to undergo positive selection of viral evolution. These data suggested that the variability of 3A may be highly informative for molecular epidemiological studies.

The 3'-UTR of O/Fujian/CHA/5/99, a region of 92 nt with high tolerant changes (72.8%-95.7% similarity) following the ORF termination codon, contains a "Y" shape of RNA which is required for its function, where the nucleotide changes of C32T and A91G were observed (Table 2).

We have identified that O/Fujian/CHA/5/99 and O/Tibet/CHA/1/99 were differed in the amino acid sequence of VP2, VP3 and VP1 (Table 2). By using the atomic coordinates obtained by X-ray crystallography of FMDV O₁BFS, six mutations which are clustered the position occupied by the G-H loop of VP1 fixed in the capsid of O/Fujian/CHA/5/99 were determined (K175R in VP2, A174 S in VP3, G137 D, T142A, A152T and Q153P in VP1). E136G substitution in VP2 was measured close to antigenic site 2 within E-F loop; Y72C, T96A and A199T substitutions in VP1 are located at βD-strand, E-F loop and antigenic site 1 of C-terminus, respectively. In addition, G72 within B-C loop (antigenic site 2) of VP2 is fixed in O/Tibet/CHA/1/99, and I168 of VP1 mapped in H-I loop (Fig. 4). As stated previously [18, 27, 28], these substitutions around heparin binding sites and antigenic site 1 on the viral capsid may influence plaque size and the pathogenicity to suckling mice.

Baby hamster kidney (BHK-21) cells were maintained at 37°C in Dulbecco's modified Eagle's medium (DMEM, Gibco) containing 10% fetal bovine serum (FBS, Hyclone). The sample of vesicles on hoof was collected from swine, which showed clinical symptoms of FMD, in Fujian province of China (OIE, May 1999). The grinding suspension (1/10, w/v) was prepared in phosphate buffered saline (PBS) containing the antibiotics penicillin (100 U/ml) and streptomycin (0.1 mg/ml), overnight at 4°C, clarified by centrifugation at 2,000 × g for 10 min, sterilized by using 0.45 μm filter unit (Millipore), and the virus was propagated on BHK-21 cells as described previously [46]. The isolated virus adapted to BHK-21 cells was designated O/Fujian/CHA/5/99 strain. The FMDV O/Tibet/CHA/1/99 strain isolated from bovine in Tibet of China was used in this work and conserved in national foot-and-mouth disease reference laboratory.

Confluent BHK-21 cell cultures in 6-well plates were prepared for plaque-forming assay. The serial 10-fold dilutions of viruses were inoculated 200 μl per well. After 1 h of incubation at 37°C in 5% CO2, 2 ml overlay medium containing 0.6% Gum and 1% FBS was added and cultured for 48 h under the same conditions. Subsequently, the BHK-21 cells were washed three times with PBS (pH 7.5), then fixed with cold acetone/methanol for 20 min at -20°C, and stained with 0.2% crystal violet for 30 min at room temperature. Finally, we were able to observe plaque morphology and calculate virus titres by plaque-forming units (PFU) from the infected BHK-21 cell cultures.

Serial ten-fold diluted viruses were prepared in DMEM containing 2% FBS, and the pathogenicities were titrated by intraperitoneal inoculation of 3-day-old suckling-mice in groups of five animals each with 0.2 ml of virus dilutions. The suckling mice were observed for 72 h after infection and the 50% lethal dose (LD₅₀) was determined by the method of Reed and Muench (1938) [47].

Rneasy Mini Kit (Qiagen) RNA extraction was performed as the manufacture's protocol. 4 μl 5 × AMV buffer, 4 μl 10 mM dNTP, 10 μl RNA, 1 μl 50 pmol/L anti-sense genome specific primers (Pan/S-, L3, NK61, P222, Pan/205, and Dnn-, respectively; Table 1), 1 μl AMV (TaKaRa) mix was incubated at 42°C for 1 h and then on ice for 3 min. After completion of the reverse transcript (RT) reaction, six overlapping PCR fragments covering the viral genome were amplified from each sample by using specific primer sets (set 1, S+ and Pan/S-; set 2, Pan/I+ and L3; set 3, Pan/204 and NK61; set 4, P211 and P222; set 5, Pan/201 and Pan/205; set 6, D3+ and Dnn-; Table 1) with LA Taq DNA polymerase (TaKaRa). The target PCR products were cleaned up using Wizard^® Gel and PCR Clean-Up System (Promega) and cloned into pGEM^®-T Vector (Promega). Cycle sequencing reaction were performed with fluorescent BigDye chain terminators (Applied Biosystems), followed by resolution on an ABI Prism 310 genetic analyzer (Applied Biosystems).

The complete genetic sequences were assembled using SeqMan (DNAStar). Multiple sequence alignment was analyzed using MegAlign (DNAStar) to construct a phylogenetic tree. MegAlign was also used for the genomic analysis of nucleotide mutations in 5'-UTR and 3'-UTR, and amino acid substitutions in leader (L) protein, structural proteins and non-structural proteins. The atomic coordinates of FMDV crystallized for O₁BFS [48‐51] were used to model the conformations, and the structures of FMDV O/Tibet/CHA/1/99 and O/Fujian/CHA/5/99 strains were optimized by placing substituted amino acids which exist in the external surface of capsid, in their standard conformations.