Detection and cell isolation of CV-A5
In an etiological spectrum study from Oct 2016 to Dec 2017 at Xiangyang, 146 cases of HFMD were caused by CV-A5. The virus identification was performed using rectal swabs from HFMD patients by PCR amplification and sequencing of the 5’-UTR fragments, RD and Vero cell isolation. The result indicated that CV-A5 emerged as one of the six main serotypes associated with HFMD. Four strains were isolated in RD cells and the isolation rate of CV-A5 in RD cells was low (4/146 of CV-A5 positive swabs). Three of four isolates grew at titers higher than 1 x 107 CCID50/ml and were also adapted to grow in Vero cells.
Determination of the complete genome sequence
The complete genome sequences of the three isolates were determined and identities of nucleotide sequence between each pair were higher than 99.5% by blasting analysis (Fig. 1). The genome of the representative strain CVA5-R3487/XY/CHA/2017 (CV-A5-R3487 in short) is 7,407 nucleotides in length. The single ORF is 6,576 bases long encoding a polyprotein of 2,191 amino acid residues and a stop codon. The ORF is flanked by the 5’- and 3’-untranslated regions (5’- and 3’-UTR), 747 and 81 nucleotides long, respectively. The genome is polyadenylated at the terminus of the 3’-UTR and the poly-A tail is at least 24 bases long. The sequence (18 nucleotides) of the 5’-extreme end is the prime sequence based on the sequence of prototype Swartz.
Serotyping and subgenotyping of Xiangyang isolates
Blasting analysis based on the VP1 gene showed that homologies of the Xiangyang strains with the prototype Swartz of CV-A5 were 81.6%-82.0% for nucleotides and 94.9%-95.9% for amino acids, respectively. In contrast, homologies with other serotypes were below 75% for nucleotide sequences and 85% for amino acid sequence. Based on the criterion of the molecular typing recommended by Oberste [9], Xiangyang isolates are CV-A5. Subgenotyping of the three Xiangyang strains within the CV-A5 serotypes was performed by sequence comparison of the VP1 including 16 strains representing subgenotypes A-D (Fig. 1). They clustered with genotype D strains in the phylogenetic tree and shared identities of 95.9-96.3% and 98.3%-99.3% in nucleotide and amino acid sequences, respectively. The Xiangyang isolates were most closely related to the CV-A5 isolated at Wenzhou in 2013 [27]. In contrast, they were genetically distant with other CV-A5 subgenotypes (or clades) isolated in other areas and different countries previously.
Recombination analysis
Investigation on genetic basis of the emerging CV-A5 in this epidemic was performed by recombination analysis. The complete sequences of 8 CV-A5 strains available at GenBank were used for alignment with CV-A5-R3487/XY/CHN by using MEGA6.0 software. The multi-sequences were scanned by SimPlot software (Version 3.5.1) to generate the similarity map as shown in Fig. 2. It showed that homologies with 8 CV-A5 strains decreased dramatically from the downstream 2B region to the 3’-UTR. This implied that there might be a recombination with other serotypes with a breakpoint at the 2A/2B junction. CV-A5-R3487 strain was aligned with other EV-A serotypes by Blasting. The results showed that from 2B to the 3’-UTR, it shared the highest similarity to the CV-A2/SZ/CHN/2015 strain (KX595284). As mentioned above, from the 5’-UTR to 2A region it shared the highest similarity to the CV-A5/WZ/CHN/2013 strain (KP289362.1). Taken together, the results suggested that there might be a recombinant event between CV-A5-R3487 and an imported CV-A2 strain or a local CV-A2 strain.
Previously, we reported that a Xiangyang CV-A2 strain was a recombinant which shared the high sequence identity with CV-A5 Swartz strain in P2 region [28]. Therefore, the Xiangyang CVA2-R1580/XY/CHN/2017, Shenzhen CV-A2/SZ/CHN/2015 (KX595284), CV-A5 Swartz and Xiangyang CV-A5-R3487 were used for detailed recombination analysis. As shown in Fig. 3a, the 3’-half of CV-A5-R3487 genome was derived from Shenzhen CV-A2 strain but not from Xiangyang CV-A2 strain. Further analysis (Fig. 3b) demonstrated that CV-A5-R3487 was indeed a recombinant carrying the 3’-half of genome of a CV-A2 strain with a breakpoint at the nt position 3,791 within the 2A region. Finally, up- and down-stream regions before and after the breakpoint 3,791 were aligned for sequence comparison. The result showed that the 5’-half (nt 1-3791) and 3’-half (nt 3,792-7,404) were clustered with CV-A5 and CV-A2 strains tested, respectively (Fig. 4a and Fig. 4b). The novel CV-A5 R3487 recombinant obtained the 5’-UTR, P1 and partial 2A regions and downstream partial 2A, 2B, 2C, P3 and 3’UTR regions of CV-A2 KX595284 strain. A novel CV-A5/CV-A4 recombinant emerged and caused an outbreak of HFMD in Xiangyang in 2017.