Molecular epidemiology of Coxsackievirus A16 strains isolated from children in Yamagata, Japan between 1988 and 2011



Katsumi Mizuta, Department of Microbiology, Yamagata Prefectural Institute of Public Health, Tokamachi 1-6-6, Yamagata 990-0031 Japan.

Tel: +81 23 627 1373; fax: +81 23 641 7486; email:


To clarify the longitudinal molecular epidemiology of coxsackievirus A16, phylogenetic analysis based on the VP1 region of 220 isolates in Yamagata, Japan was performed. The resultant phylogenetic tree indicates that the Yamagata isolates and reference strains can be readily genotyped into three genogroups, and 0, 12 and 208 isolates belonged to the first, second, and third genogroups, respectively. The first genogroup includes only the prototype strain, the second strains that had disappeared by the end of the 20th century and the third comprises those that have been circulating since then in local communities, such as Yamagata.

List of Abbreviations

coxsackievirus A


enterovirus 71


hand-foot-and-mouth disease

Coxsackievirus A16 is a small, non-enveloped virus belonging to the genus Enterovirus within the family Picornaviridae. Together with EV71 and CVA10, CVA16 is known to cause HFMD [1]. Although HFMD typically manifests as a benign and self-limiting rash in children, concern about EV71 infections has heightened because of the increased frequency of neuropathogenicity observed since the last decade of the 20th century (mainly in the Asia-Pacific region) [1, 2]. This has resulted in an increase in the number of epidemiological and preventive or therapeutic studies related to EV71. Several epidemiological studies have clarified that EV71 is an actively circulating agent capable of rapid evolution into 12 subgenogroups (A, B0–5, C1–5) [2-6]. However, no direct correlation between neurovirulence and EV71 genogroup has been observed. Several vaccine candidates for prevention of EV71 have been developed [7-9]. In contrast, there have been few studies on CVA16 [10-15], as with EV71 infections, the availability of relevant data was limited before a number of severe outbreaks during the 1990 s. However, because several severe cases of CVA16 have been reported recently and, as a consequence, development of a vaccine started [16-21], molecular study of the complete sequence of the VP1 region, which plays an important role in characterizing antigenicity [22], has become more important. We previously reported the molecular epidemiology, based on sequence analysis and genotyping, of EV71 strains isolated from children in Yamagata, Japan between 1998 and 2007, [4, 23]. In this study, we analyzed the VP1 region of the CVA16 strains isolated between 1988 and 2011 in Yamagata to clarify the molecular epidemiology of CVA16.

We performed virus isolation from nasopharyngeal samples from children with HFMD or respiratory illnesses and subsequently identified them by a microplate method as described previously [24]. Mainly using HEF, Vero, Vero E6 and GMK cell lines, CVA16 strains were isolated at the Virus Research Center, National Hospital Organization, Sendai Medical Center, Sendai, Japan between 1988 and 1998, and at the Department of Microbiology, Yamagata Prefectural Institute of Public Health between 1997 and 2011. We succeeded in neutralizing all isolates by using a single polyclonal antiserum against CVA16 for identification. We carried out sequence analysis of the complete VP1 region (891 nucleotides) of the CVA16 isolates using the same procedure as described previously for EV71 [4], except that we used the published primers 159, 222 and CVA16R-3400 [3, 10, 25], and our original primers CVA16YFor (ATgggATATgCCCAACTACg) and CVA16YRev (CAgTCTgCCAAgCAAATgAA). We registered sequence data for the isolates from Yamagata with GenBank (accession numbers AB634286–AB634452 and AB771956–AB772007). We analyzed sequence data using CLUSTAL W (version 1.83), and constructed a phylogenetic tree by the neighbor-joining method using the same software [26].

We isolated 293 CVA16 strains in Yamagata between 1988 and 2011. Figure 1 shows the number of CVA16 strains isolated by month. Of the 293 strains, we sequenced 220. Of the 108 strains isolated between 1988 and 1997, we did not sequence 73 because unfortunately these isolates had not been stored in a freezer. However, we did sequence all 185 strains isolated between 1998 and 2011. Figure 2 depicts the resultant phylogenetic tree, in which three distinct genogroups are distinguishable. Figure 1 also shows the number of strains from each genogroup (based on Fig. 2) by month.

Figure 1.

Monthly distribution and genogroups of CVA16 strains isolated in Yamagata, Japan between 1988 and 2011. The genogroups were classified according to the phylogenetic analysis shown in Figure 2.

Figure 2.

A phylogenetic tree for the complete (891 nucleotide) VP1 region of CVA16 representative strains isolated in Yamagata, Japan between 1988 and 2011 together with reference strains and GenBank accession numbers for reference strains. The numbers of Yamagata isolates having 99–100% sequence similarities with the representative Yamagata strains are shown in parentheses. The first genogroup was previously established as Genogroup A (A) and the second and third genogroup strains have been reported as B or B lineage 1 or B2, and C or B lineage 2 or B1, respectively [11, 13-15]. Branch lengths are proportional to the number of nucleotide differences. Only significant bootstrap values are shown. The marker denotes a measurement of relative phylogenetic distance. •, Yamagata strains; ○, China strains; □, reference strains subgenotyped as C1 by Iwai et al. [13]; ◊, reference strains subgenotyped as C2 by Iwai et al. [13]; ▵, reference strains subgenotyped as C3 by Iwai et al. [13]; ★, reference strains subgenotyped as B1a by Zhang et al. [14];★★, reference strains subgenotyped as B1b by Zhang et al. [14]; ★★★, reference strains subgenotyped as B1c by Zhang et al. [14].

No strains from the first genogroup had been isolated in Yamagata; however, 12 Yamagata isolates obtained between 1988 and 1998 and represented by Y88-5375 belonged to the second genogroup (Figs 1, 2). All other strains isolated between 1995 and 2011 belonged to the third genogroup (Figs 1, 2); in particular, we analyzed all isolates from the year 2000 onwards and they all belonged to the third genogroup.

The nucleotide sequence similarities among all the Yamagata isolates were 87–100%. The nucleotide sequence similarities among the same genogroup strains (shown in Fig. 2) were 92–100% of the second genogroup and 89–100% of the third genogroup, whereas the similarities between the strains of the second and third genogroup were 87–92%. When we compared 297 amino acids of the VP1 region, there were 0–2 amino acid differences (99–100% similarities) among the second genogroup strains and 0–7 amino acid differences (98–100% similarities) among the third genogroup strains. There were 0–5 amino acid differences (98–100% similarities) between the second and third genogroup strains. Surprisingly, we found identical amino acid sequences between the second and third genogroup strains, such as Y92-2389 and Y95-2096. There were no specific amino acid changes that differentiated the two genogroups and no accumulation of amino acid changes at specific positions, with only a few exceptions.

Some molecular epidemiological studies were previously carried out based on VP1 and/or VP4 [11, 12, 15]; they revealed no major discrepancies [11]. However, no standardized nomenclature for CVA16 subgenogroups has yet been established. As shown in Figure 2, three genogroups are distinguishable. Based on the VP1 region, genogroups A, B and C, with the subgenogroups B lineage 1–2, C1–3, B1a–c, and B2, have been described in published reports [11, 13-15] (Table 1). All authors, including us, agree that the first genogroup includes only the prototype strain [11, 13-15]. Thus, there is no question that genogroup A is an independent group. It is likely that the B strains reported by Iwai et al. [13], B lineage 1 strains reported by Perera et al. [11] and Li et al. [15], and B2 strains reported by Zhang et al. [14], together with the 12 Yamagata isolates, all belong to the same genogroup (Table 1 and Fig. 2). Thus, we can recognize this genogroup as a second independent genogroup. Previous studies, and our data, revealed that the strains in this genogroup had disappeared before the start of the 21st century [11, 13, 14]. Most of the Yamagata strains belong to a third genogroup, which includes C1–3, B lineage 2 and B1a–c strains (Table 1 and Fig. 2). Figure 2 shows that the reference strains of subgenogroups C1, C2 and C3 reported by Iwai et al. [13] and B1a, B1b and B1c reported by Zhang et al. [14] do not always branch together among the same subgenogroups, contrary to what was shown in their studies. It is difficult to subgenotype their strains together with the Yamagata strains using their nomenclature. Thus, the subgenotyping of the third genogroup strains is inconsistent, the subgenogroup of each isolate changing according to the strains used to construct the phylogenetic tree. These findings suggest that, at this stage, CVA16 strains can be roughly classified into three genogroups without subgenotyping. Based on VP4 sequence analysis, Li et al. previously proposed this idea of three genetic lineages [10].

Table 1. Genogroups and subgenogroups of CVA16 reference strains based on VP1 and VP4 regions
StrainGenBank Accession No.Genogroup/SubgenogroupProposed Genogroups based on VP1 analysis in this study
Based on VP1Based on VP4
Iwai et al. [13]Perera et al. [11]Li et al. [15]Zhang et al. [14]Li et al. [10]
  • Unclassified.
G-10/South Africa/51U05876AAAAAFirst genogroup (or lineage) ⇒ established as “Genogroup A”, which includes prototype strain only
S10432/SAR/98AM292455B B lineage1B2 Second genogroup ⇒ “Genogroup B”, which disappeared by the end of the 20th century
24/Toyama/1981AB465366B  B2  
107/Toyama/1990AB465369B  B2  
379/Toyama/1984AB465367B  B2  
S70382/SAR/1998AM292461 B lineage1B lineage1B2  
shzh02-14AY895110C1B lineage2  CThird genogroup ⇒ “Genotype C” without subgenotyping, which has been circulating recently
120/Toyama/2003AB465399C1  B1b  
188/Toyama/2000AB465372C1  B1b  
SB16087/SAR/2005AM292476C1B lineage2 B1c  
CNS32874/SAR/03AM292445C2B lineage2    
S10051/SAR/98AM292454C2B lineage2 B1a  
927/Toyama/1998AB465371C2  B1a  
shzh02-124AY895094C2B lineage2 B1aC 
shzh04-J52AY895093C2B lineage2 B1a  
S114131/SAR/05AM292463C2B lineage2    
CNS11062/SAR/01AM292444C2B lineage2    
MY823-3/SAR/97AM292433C2B lineage2 B1a  
521/VNM/05AM292436C3B lineage2    
0033/AUS/05AM292435C3B lineage2B lineage2B1a  
UM16809AM292483 B lineage2B lineage2B1a  
S5JF317960  B lineage2   
S12JF317961  B lineage2   
S176JF317966  B lineage2   
S358JF317970  B lineage2   
S388JF317972  B lineage2   
shzh99-83AY821797    C 
shzh00-1AY790926    C 
shzh01-58AY895101    C 
shzh01-66AY895105    C 
shzh01-77AY895097    C 
shzh02-12AY895098    C 
shzh02-16AY895100    C 
shzh02-38AY821795    C 
shzh02-75AY895102    C 
shzh02-78AY895108    C 
shzh02-93AY895107    C 
shzh03-10AY895095    C 

Because our hypothesis was that CVA16 strains circulate widely throughout regions in the same manner as EV71, we were surprised at our findings [4]. Zhang et al. reported that in China the molecular epidemiology of CVA16 reflects a pattern of endemic circulation, with a relatively slower evolutionary rate than that of EV71 [14]. From this perspective, Figure 2, which shows that, with a very few exceptions, the Yamagata and China strains branched and evolved independently, does not contradict the endemicity of CVA16 in local communities. This is in contrast to EV71, which circulates widely throughout the Asia-Pacific region [4]. In view of antigenicity, the facts that we succeeded in neutralizing all isolates using a single polyclonal antiserum against CVA16 for identification and that we found identical amino acid sequences of the VP1 region in the second and third genogroup strains suggest its stability. This is also in contrast to EV71; we previously demonstrated that genogroup C strains of EV71 are more difficult to neutralize than genogroup B strains [23]. However, we are not yet sure whether such epidemiological differences between CVA16 and EV71 infections affect the clinical course.

In conclusion, phylogenetic analysis of both CVA16 and EV71 should be carried out based on the complete VP1 region when clarifying the epidemiology of CVA16 and developing control measures such as vaccines. Standardized nomenclature for the genotyping of CVA16 is not yet established: as shown in Table 1 and Figure 2, we propose that for now the CVA16 strains be roughly classified based on analysis of the VP1 region into three genogroups, without subgenotyping. The first genogroup, currently known as Genogroup A, includes only the prototype strain, which disappeared a long time ago. The strains belonging to the second genogroup, which we propose here as a new Genogroup B (Table 1 and Fig. 2), had all disappeared by the end of the 20th century. In contrast, since that time the ones belonging to the third genogroup, which we propose here as a new Genogroup C (Table 1 and Fig. 2), have definitely been actively circulating in Japan and possibly in other areas.


We thank the medical staff and people of Yamagata Prefecture for their collaboration in specimen collection for the surveillance of viral infectious diseases.


All authors declare they have no conflict of interests.