Introduction

The group B coxsackieviruses (CVB) are members of the species Human enterovirus B, the genus Enterovirus and the family Picornaviridae. They are nonenveloped, positive-sense, single-stranded RNA viruses with six serotypes (CVB1 to CVB6), which cause a broad spectrum of diseases, ranging from minor febrile illness to severe infections such as aseptic meningitis and myocarditis [1]. In addition, CVBs were also reported to play a role in the development of type 1 diabetes [2, 3]. CVB3 and CVB5 frequently cause outbreaks of aseptic meningitis and hand-foot-and-mouth disease (HFMD) [4–7]. CVB1, CVB2, CVB4, and CVB6 outbreaks are rare. CVB2 has been implicated in 1.5–6.0 % of the reported surveillance for known enterovirus serotypes [8] and is also one of the 15 most frequently isolated enterovirus serotypes in the United States [9]. Since 2008, hand-foot-and-mouth disease (HFMD) has been a serious public health threat in China. HFMD is primarily caused by enterovirus 71 (EV-A71), coxsackievirus A16 (CV-A16) and CV-A6 [10], but other CVBs such as CVB1 and CVB4 have frequently been detected [11].

At present, only four entire genomes of CVB2 are available in GenBank, including the prototype Ohio-1/AF085363 strain isolated from a patient with “summer grippe” in 1950 [12], KOR 04-279/EF174469 isolated from a patient with acute myocarditis in 2004, KOR 04-243/EF174468 isolated from a patient with aseptic meningitis in 2004 and BCN314/KM386639 (except for 55 nucleotides at the 5′-end of the genome) isolated from nasopharyngeal aspirates from children in 2007.

To the best of our knowledge, the RW41-2/YN/CHN/2012 strain was first isolated from a two-year and three-month-old boy hospitalized with clinical manifestations of vesicles on his hands, feet and mouth, along with fever, irritability and vomiting (no myocarditis) in Yunnan, China. A stool specimen was collected and processed according to standard protocols recommended by the World Health Organization [13]. The samples were then inoculated into the following three cell lines: RD (human rhabdomyosarcoma cells), KMB17 (human embryonic lung diploid fibroblasts), and A549 (human lung cancer cells) [14]. An enterovirus-like cytopathic effect (CPE) was only observed in RD cells. The sample was passaged three times in RD. Viral RNA was extracted using QIAamp Viral RNA Mini Kit (QIAGEN, Valencia, CA, USA). RT-PCR was performed using a PrimeScriptTM one-step RT-PCR Kit Ver. 2 (TakaRa, Shiga, Japan) by ‘‘primer-walking’’ strategy [14]. The sequence analyses were performed using an ABI 3730XL automatic sequencer by Sangon (Shanghai, China). The entire genome sequence of CVB2 RW41-2/YN/CHN/2012 strain was determined and its molecular characterization was compared with other human enterovirus B (HEV-B) serotypes based on sequences available in GenBank. The sequence was deposited in the GenBank database (accession number: KX499536).

The entire genome sequence of the RW41-2/YN/CHN/2012 strain consisted of 7,405 nucleotides. The genome consisted of a 5′ UTR of 742 nucleotides followed by an open reading frame (ORF) that encoded a polyprotein of 2,187 amino acids and a 3′ UTR of 102 nucleotides. Compared with the CVB2 prototype strain Ohio-1 and other CVB2 strains, the nucleotide sequence was more variable (82.1 % to 90.3 %), however the amino acid sequence was highly conserved (96.4 % to 97.3 %) (Table 1). The RW41-2/YN/CHN/2012 strain was most homologous to the BCH314 strain (90.3 %), which was isolated from Beijing, China in 2007.

Table 1 Nucleotide and amino acid identity between Ohio-1, RW41-2/YN/CHN/2012 and other CVB2 strains in sequenced genomic regions
Full size table

Phylogenetic analyses based on the complete VP1, P1, P2, P3, and 3D regions of the strain, all CVB2 strains available in the GenBank database, as well as nine HEV-B strains (CVB4/Tuscany/DQ480420, CVB5/Faulkner/AF114383, CV-A9/1-D7-CA9/KT353721, EV-B84/AFP452/GD/CHN/2004/KP262053, EV-80/HZ01/SD/CHN/2004/JX644073, EV-B75/102/SD/CHN/97/KF874627, CVB1Nm/EU147493, CVB6/Schmitt/AF105342, and CVB3/Nancy/JX312064), were carried out. Phylogenetic trees were constructed by the neighbor-joining method using the MEGA program (version 6.06). Once aligned, similarity plot and boot-scanning analyses were performed using the SimPlot program (version 3.5.1). According to phylogenetic and homogeneity analysis for the complete VP1 gene, all of the CVB2 isolates separated into four distinct genogroups (A–D) with some temporal and regional sub-clustering. Chinese strains, isolated between 2007 and 2014, belonged to genogroup D, while other Chinese strains isolated between 1987 and 2003 belonged to genogroup B (Figure 1). The RW41-2/YN/CHN/2012 strain was the most homologous to the AYTY11001/HN/CHN/2011 strain (96.9 %), which was isolated from Henan, China in 2011. In the P1 and P2 genome regions, RW41-2/YN/CHN/2012 was similar to other CVB2 strains (including the BCH314 strain). The P3 region (including 3D) was more related to other HEV-B members (Figure 2), suggesting genetic rearrangement between these strains. In addition, our similarity plot and boot-scanning analyses of the genomic sequences of the Yunnan RW41-2/YN/CHN/2012 strain and other HEV-B strains showed several recombination sites (Supplementary Figure 1). In the nonstructural coding regions, recombination events were observed between the RW41-2/YN/CHN/2012 strain and other HEV-B strains, such as the EV-B75/102/SD/CHN/97/KT353721 (Shangdong, China), EV-80/HZ01/SD/CHN/2004/ JX644073 (Shangdong, China) and EV-B84/AFP452/GD/CHN/2004/KP262053 (Guandong, China), with support values of more than 85 %. Mosaic recombination of HEVs often occurs in the nonstructural region [15, 16]. Obviously, recombination of the RW41-2/YN/CHN/2012 strain only occurred with HEV-B species in the nonstructural region. All of these strains were isolated from the same country – China. Moreover, Yunnan is a leading tourism province, attracting hundreds of thousands of visitors, which might provide plentiful opportunity for these strains to mingle and recombine.

Fig. 1
figure 1

Phylogenetic analysis of CVB2 strains, based on the 847 nt of VP1, generated by the neighbor-joining algorithm implemented in MEGA (version 6.1) using the Kimura two-parameter substitution model and 1,000 bootstrap pseudo-replicates. ▲ represents strains isolated in this investigation. ● represents strains isolated from other provinces of China

Full size image
Fig. 2
figure 2

Phylogenetic analysis of the 14 CVB strains, based on the P1, P2, P3, and 3D coding sequences, generated by the neighbor-joining algorithm implemented in MEGA (version 6.1) using the Kimura two-parameter substitution model and 1,000 bootstrap pseudo-replicates. Only strong bootstrap values (>75 %) are shown. ▲ represents strains isolated in this investigation. ● represents CVB2 strains

Full size image

The average VP1 nucleotide and amino acid sequence divergence between the CVB2 genogroups A, B, C, and D were 14.3–19.2 % and 1.4–3.5 %, respectively (Supplementary Table 2). In the VP1 gene, the homologies of the nucleotide and amino acid sequences of the RW41-2/YN/CHN/2012 strain were 81.4 %–96.9 % and 97.2 %–99.3 %, respectively, compared with other Chinese strains. The VP1 protein of human enterovirus (HEV) is the major neutralization determinant [17]. This indicated the antigenic determinant was conserved for CVB2. The low frequency of CVB2 outbreaks may be related to the highly conserved amino acid sequence and a low evolution rate in VP1 [8].

All six serotypes of CVB can use the coxsackie and adenovirus receptor (CAR) for cell attachment and entry [18–20]. However, CVB1, 3, 5 and 6 also use decay-accelerating factor (DAF), and can replicate and cause cytolysis in RD cells, a CAR-deficient cell line. The growth in RD cells has been associated with the use of DAF as an attachment molecule [21]. A previous study has found that a CVB2 variant causing cytolysis in RD cells had three amino acid changes: one (Thr118 → Met) situated in the BC region of VP1, another (Gln164 → Lys) which is likely to be responsible for conformational changes in the VP1 enabling binding to DAF, and the third (Lys185 → Arg) in 2C [22], which has RNA-binding, NTPase, and cysteine-rich sequence motifs that are highly conserved. In addition, Gullberg et al. [23] identified that a single amino acid change (Gln164 → Lys) was responsible for the cytolytic RD phenotype. However, the virus-host interaction was not affected by anti-DAF antibodies [20] and these mutations do not occur in the RW41-2/YN/CHN/2012 strain, which can cause cytolysis in RD cells. Thus, in addition to CAR and DAF, a third receptor may be used during CVB2 infection [22] and the function of these mutations needs further confirmation.