JOURNAL OF CLINICAL MICROBIOLOGY, Feb. 2010, p. 619–622 Vol. 48, No. 20095-1137/10/$12.00 doi:10.1128/JCM.02338-09Copyright © 2010, American Society for Microbiology. All Rights Reserved.

Molecular Evidence of Persistent Epidemic and Evolution ofSubgenotype B1 Coxsackievirus A16-Associated Hand,

Foot, and Mouth Disease in China�†Yong Zhang,1 Dongyan Wang,1 Dongmei Yan,1 Shuangli Zhu,1 Jianfeng Liu,2 Haiyan Wang,3

Shengcang Zhao,4 Deshan Yu,2 Lijuan Nan,5 Junjing An,1 Li Chen,1 Hongqiu An,1Aiqiang Xu,3* and Wenbo Xu1*

WHO WPRO Regional Polio Reference Laboratory and State Key Laboratory for Molecular Virology & Genetic Engineering,Institute for Viral Disease Control and Prevention, Chinese Center for Disease Control and Prevention, No. 27, Nanwei Road,

Xuanwu District, Beijing 100050, People’s Republic of China1; Gansu Center for Disease Control and Prevention,Lanzhou 730000, People’s Republic of China2; Shandong Center for Disease Control and Prevention, No. 16992,

Jingshi Road, Jinan 250014, People’s Republic of China3; Qinghai Center for Disease Control andPrevention, Xining 810007, People’s Republic of China4; and Inner Mongolia Center for

Disease Control and Prevention, Hohhot 010020, People’s Republic of China5

Received 30 November 2009/Accepted 3 December 2009

The molecular epidemiology of CVA16 in China between 1999 and 2008 reflects a pattern of endemiccocirculation of clusters B1a and B1b within subgenotype B1 viruses. The annual evolution rate of CVA16 wasestimated as approximately 0.91 � 10�2 substitutions per synonymous nucleotide/year and is slightly lowerthan that of HEV71.

Coxsackievirus A16 (CVA16) and human enterovirus 71(HEV71) are the two major causative agents of hand, foot, andmouth disease (HFMD) (12, 17). In recent years, numerouslarge outbreaks of HEV71-associated HFMD, which were of-ten accompanied by severe complications, including death, in-creased research interest in HEV71 strains (1, 14, 16). Incontrast, little interest was paid to CVA16 strains becauseCVA16-associated HFMD was usually mild and benign (2, 12).Although CVA16 is genetically most closely related to HEV71,the genetic diversity and molecular evolution of CVA16, unlikethose of HEV71, have not been fully described (5, 7, 12, 13).Cocirculation of CVA16 and HEV71 has been proven to havecontributed to the serious outbreaks of HFMD that have oc-curred in China since 2007 (17); therefore, the genetic vari-ability and the evolution of CVA16 were determined in thisstudy.

The 42 CVA16 strains evaluated in this study were isolatedfrom HFMD patients from different geographical locations inthe Shandong, Gansu, Inner Mongolia, and Qinghai provincesof China between 2007 and 2008 (see supplemental data). Toinvestigate the molecular epidemiology of CVA16 in mainlandChina, 24 additional Chinese CVA16 sequences found in Bei-

jing and Guangdong provinces between 1999 and 2005 and 35international CVA16 sequences (obtained from the GenBankdatabase) were also analyzed.

The complete VP1/capsid sequences of the CVA16 strains wereobtained as previously described, using in-house primers thatflanked the VP1 region (13, 17): CVA16-VP1-S, 5�-ATTGGTGCTCCCACTACAGC-3� (nucleotides 2335 to 2354, relative tostrain CVA16/G-10), and CVA16-VP1-A, 5�-GCTGTCCTCCCACACAAGAT-3� (nucleotides 3426 to 3445, relative to strainCVA16/G-10).

A total of 66 Chinese CVA16 sequences were divided intothree lineages on the basis of phylogenetic analysis (Fig. 1). A6.5 to 8.1% nucleotide divergence was found among thesethree lineages, suggesting that the CVA16-associated HFMDoutbreaks in China were a result of the coincident circulationof three genetically distinct viruses.

To determine the molecular epidemiology of ChineseCVA16 strains associated with HFMD epidemics, a phyloge-netic dendrogram was constructed with 21 Chinese CVA16sequences (randomly selected on the basis of their geneticrelationships) that circulated during the period 1999–2008 inaddition to the 35 international CVA16 sequences that repre-sented two known genotypes (A and B) (13) (Fig. 2).

As in a previous study (13), all CVA16 strains could begrouped into genotypes A and B. The prototype G-10 straindiffered from the other strains by 27.5 to 30.2% and clusteredseparately from all other CVA16 strains, including ChineseCVA16 strains, which clearly belonged to genotype B. Thisfinding was based on the fact that the genetic variation be-tween all other CVA16 strains was less than 13.5%. The se-quences in genotype B could be further divided into B1 and B2subgenotypes with a bootstrap support of 100% (Fig. 2). Chi-nese CVA16 strains isolated between 1999 and 2008 and themajority of international CVA16 strains isolated between 1997

* Corresponding author. Mailing address for A. Xu: Shandong Cen-ter for Disease Control and Prevention, No. 16992, Jingshi Road, Jinan250014, People’s Republic of China. Phone: 0086-531-82679606. Fax:0086-531-82679620. E-mail: aqxuepi@163.com. Mailing address forW. Xu: WHO WPRO Regional Polio Reference Laboratory and StateKey Laboratory for Molecular Virology & Genetic Engineering,Institute for Viral Disease Control and Prevention, Chinese Center forDisease Control and Prevention, No. 27, Nanwei Road, XuanwuDistrict, Beijing 100050, People’s Republic of China. Phone and fax:0086-10-63028480. E-mail: wenbo_xu1@yahoo.com.cn.

† Supplemental material for this article may be found at http://jcm.asm.org/.

� Published ahead of print on 16 December 2009.

619

on October 3, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

and 2007 formed subgenotype B1, and the 9 CVA16 strainsisolated from Japan and Malaysia between 1981 and 2000formed subgenotype B2. The nucleotide divergence betweensubgenotypes B1 and B2 was 11.8%.

Phylogenetic classification based on the complete VP1 re-gion (891 bp) of HEV71 has proved to be useful in trackinggenotypes of HEV71-associated HFMD over different tempo-ral and geographical outbreaks (1, 3, 8, 14, 17). Previous stud-ies of CVA16 genetic diversity by Li et al. (12) and Iwai et al.(7) showed that CVA16 strains could be divided into three

different clusters, called A, B, and C. However, clusters B andC in their studies correspond to subgenotypes B2 and B1,respectively, in our study, in which a difference of at least 15%in the complete VP1 region of CVA16 strains was used todistinguish genotypes (13). Thus, their B and C clusters shouldbe combined into one genotype.

Subgenotype B1 could be further divided into clusters B1a,B1b, and, possibly, B1c. The nucleotide divergence betweenclusters B1a and B1b was 6.5%. All Chinese strains isolatedbetween 1999 and 2008 belonged to clusters B1a and B1b,

FIG. 1. Phylogenetic dendrogram constructed by using the maximum-likelihood method implemented in the PHYLO_WIN program, version2.0 (4), based on the alignment of the complete VP1 gene sequences of 66 CVA16 strains (from HFMD patients in the Shandong, Gansu, InnerMongolia, Qinghai, Beijing, and Guangdong provinces of China) and 1 CVA16 prototype strain (G-10). The bootstrap values from 1,000pseudoreplicates for major lineages within the tree are shown as percentages. Twelve Chinese CVA16 strains, which were isolated from 12 HFMDcases during 2007–2008 and are indicated by filled circles (●), and 9 strains, which were isolated from 9 HFMD cases in two provinces during1999–2005 and are indicated by a triangle (Œ), were selected for further molecular epidemiological analysis. Abbreviations of provinces of China:GS, Gansu; NM, Inner Mongolia; SD, Shandong; QH, Qinghai; BJ, Beijing; GD, Guangdong.

620 NOTES J. CLIN. MICROBIOL.

on October 3, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

which could also be found in Taiwan, Malaysia, Thailand,Australia, Vietnam, and Saudi Arabia during the same period.This indicated that Chinese strains coevolved and cocirculatedwith those from neighboring countries. Only one strain fromMalaysia (SB16087/MAL/2005) belonged to the probable sub-genotype B1c, and the nucleotide divergences between clustersB1c and B1a and clusters B1c and B1b were 7.0% and 6.8%,respectively. Close relatives of this strain have not been foundamong the known CVA16 strains. This probably reflects thelack of additional strains from Malaysia and its surrounding

countries and may represent a surveillance gap in these coun-tries.

Because of the lack of a true “ancestor” strain and theapparent presence of multiple lineages, CVA16 sequenceswere selected on the basis of their genetic relationships inorder to estimate the evolutionary rate (substitutions per nu-cleotide per year [Fig. 2]) (1, 9). Using the oldest strain in eachsubgenotype/cluster as a reference, the evolutionary distanceswere calculated by pairwise comparison according to theKimura two-parameter method of the MEGA program (11,

FIG. 2. Phylogenetic dendrogram constructed by using the maximum-likelihood method implemented in the PHYLO_WIN program, version2.0 (4), based on the alignment of the complete VP1 gene sequences of 21 representative Chinese CVA16 strains and other international CVA16strains of known genotypes listed in the supplementary material. The bootstrap values from 1,000 pseudoreplicates for major lineages within thetree are shown as percentages. The prototype HEV71 strain (BrCr) was used as an outgroup. Strains indicated by filled circles (●) are CVA16strains isolated from HFMD cases during 2007–2008. Strains indicated by a triangle (Œ) are CVA16 strains isolated from HFMD cases during1999–2005. Bootstrap values greater than 80% were considered to be statistically significant for grouping. A difference of at least 15% in thecomplete VP1 region of CVA16 strains was used to distinguish genotypes (13). Abbreviations of provinces of China: GS, Gansu; NM, InnerMongolia; SD, Shandong; QH, Qinghai; BJ, Beijing; GD, Guangdong.

VOL. 48, 2010 NOTES 621

on October 3, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

15). Both Ks (synonymous substitutions per synonymous site)and Kt (all substitutions per site) values were analyzed, and theevolutionary rate was calculated by linear regression of thegenetic distance from the oldest strain versus the year of iso-lation.

The approximate evolution rate of CVA16 was estimatedfrom the differences in VP1/capsid sequences and the date ofspecimen collection among strains within cluster B1a andstrains within subgenotype B2 (Fig. 2 and Table 1). The aver-age evolution rates for the cluster B1a and the subgenotype B2were 0.91 � 10�2 (Ks estimate) and 2.49 � 10�3 (Kt estimate),respectively (Table 1), which were slightly lower than the ratescalculated for HEV71 (1.35 � 10�2) (1) and poliovirus type 1(3.36 � 10�2), which is similar to HEV71 (10), althoughCVA16 strains are broadly distributed geographically.

The evolution of HEV71 and that of CVA16 were similar inthe sense that their prototype strains were the sole members ofgenotype A; eventually, they all gave way to modern strainsdue to molecular evolution. HEV71 was first identified in 1970in the United States, after which it appeared as two cocircu-lated genotypes (genotype B, subgenotypes B1 to B5; genotypeC, subgenotypes C1 to C5) during a 40-year course of evolution(1, 6, 14, 17). In contrast, the evolutionary rate of CVA16 isrelatively slow in light of the fact that CVA16 was first identi-fied in 1951 in South Africa and that all CVA16 strains, exceptthe prototype strain, formed a single genotype (genotype B)containing two subgenotypes (B1 to B2) as a result of approx-imately 60 years of evolution.

In conclusion, our study revealed that the molecular epide-miology of CVA16 in China during the years 1999–2008 re-flects a pattern of endemic circulation of clusters B1a and B1bwithin subgenotype B1 viruses; however, in comparison withHEV71, the evolutionary rate of CVA16 is relatively lower.

This study was supported by the National Natural Science Founda-tion of China (project no. 30900063) and the National Key TechnologyR&D Program of China (project no. 2008BAI56B00, 2009ZX1004-201, and 2009ZX1004-202).

We acknowledge all laboratories that isolated the viruses necessaryfor this study.

REFERENCES

1. Brown, B. A., M. S. Oberste, J. P. Alexander, Jr., M. L. Kennett, and M. A.Pallansch. 1999. Molecular epidemiology and evolution of enterovirus 71strains isolated from 1970 to 1998. J. Virol. 73:9969–9975.

2. Chang, L. Y., T. Y. Lin, Y. C. Huang, K. C. Tsao, S. R. Shih, M. L. Kuo, H. C.Ning, P. W. Chung, and C. M. Kang. 1999. Comparison of enterovirus 71 andcoxsackie-virus A16 clinical illnesses during the Taiwan enterovirus epi-demic, 1998. Pediatr. Infect. Dis. J. 18:1092–1096.

3. Diedrich, S., A. Weinbrecht, and E. Schreier. 2009. Seroprevalence andmolecular epidemiology of enterovirus 71 in Germany. Arch. Virol. 154:1139–1142.

4. Galtier, N., M. Gouy, and C. Gautier. 1996. SEAVIEW and PHYLO_WIN:two graphic tools for sequence alignment and molecular phylogeny. Comput.Appl. Biosci. 12:543–548.

5. Hosoya, M., Y. Kawasaki, M. Sato, K. Honzumi, A. Hayashi, T. Hiroshima,H. Ishiko, K. Kato, and H. Suzuki. 2007. Genetic diversity of coxsackievirusA16 associated with hand, foot, and mouth disease epidemics in Japan from1983 to 2003. J. Clin. Microbiol. 45:112–120.

6. Huang, Y. P., T. L. Lin, C. Y. Kuo, M. W. Lin, C. Y. Yao, H. W. Liao, L. C.Hsu, C. F. Yang, J. Y. Yang, P. J. Chen, and H. S. Wu. 2008. The circulationof subgenogroups B5 and C5 of enterovirus 71 in Taiwan from 2006 to 2007.Virus Res. 137:206–212.

7. Iwai, M., A. Masaki, S. Hasegawa, M. Obara, E. Horimoto, K. Nakamura, Y.Tanaka, K. Endo, K. Tanaka, J. Ueda, K. Shiraki, T. Kurata, and T.Takizawa. 2009. Genetic changes of coxsackievirus A16 and enterovirus 71isolated from hand, foot, and mouth disease patients in Toyama, Japanbetween 1981 and 2007. Jpn. J. Infect. Dis. 62:254–259.

8. Jee, Y. M., D. S. Cheon, K. Kim, J. H. Cho, Y. S. Chung, J. Lee, S. H. Lee,K. S. Park, J. H. Lee, E. C. Kim, H. J. Chung, D. S. Kim, J. D. Yoon, andH. W. Cho. 2003. Genetic analysis of the VP1 region of human enterovirus71 strains isolated in Korea during 2000. Arch. Virol. 148:1735–1746.

9. Jorba, J., R. Campagnoli, L. De, and O. Kew. 2008. Calibration of multiplepoliovirus molecular clocks covering an extended evolutionary range. J. Vi-rol. 82:4429–4440.

10. Kew, O. M., R. W. Sutter, B. K. Nottay, M. J. McDonough, D. R. Prevots, L.Quick, and M. A. Pallansch. 1998. Prolonged replication of a type 1 vaccine-derived poliovirus in an immunodeficient patient. J. Clin. Microbiol. 36:2893–2899.

11. Kumar, S., M. Nei, J. Dudley, and K. Tamura. 2008. MEGA: a biologist-centric software for evolutionary analysis of DNA and protein sequences.Brief. Bioinform. 9:299–306.

12. Li, L., Y. He, H. Yang, J. Zhu, X. Xu, J. Dong, Y. Zhu, and Q. Jin. 2005.Genetic characteristics of human enterovirus 71 and coxsackievirus A16circulating from 1999 to 2004 in Shenzhen, People’s Republic of China.J. Clin. Microbiol. 43:3835–3839.

13. Perera, D., M. A. Yusof, Y. Podin, M. H. Ooi, N. T. Thao, K. K. Wong, A.Zaki, K. B. Chua, Y. A. Malik, P. V. Tu, N. T. Tien, P. Puthavathana, P. C.McMinn, and M. J. Cardosa. 2007. Molecular phylogeny of modern cox-sackievirus A16. Arch. Virol. 152:1201–1208.

14. Shimizu, H., A. Utama, N. Onnimala, C. Li, Z. Li-Bi, M. Yu-Jie, Y. Pong-suwanna, and T. Miyamura. 2004. Molecular epidemiology of enterovirus 71infection in the Western Pacific Region. Pediatr. Int. 46:231–235.

15. Tamura, K., J. Dudley, M. Nei, and S. Kumar. 2007. MEGA4: MolecularEvolutionary Genetics Analysis (MEGA) software version 4.0. Mol. Biol.Evol. 24:1596–1599.

16. Wang, J. R., Y. C. Tuan, H. P. Tsai, J. J. Yan, C. C. Liu, and I. J. Su. 2002.Change of major genotype of enterovirus 71 in outbreaks of hand-foot-and-mouth disease in Taiwan between 1998 and 2000. J. Clin. Microbiol.40:10–15.

17. Zhang, Y., X. J. Tan, H. Y. Wang, D. M. Yan, S. L. Zhu, D. Y. Wang, F. Ji,X. J. Wang, Y. J. Gao, L. Chen, H. Q. An, D. X. Li, S. W. Wang, A. Q. Xu,Z. J. Wang, and W. B. Xu. 2009. An outbreak of hand, foot, and mouthdisease associated with subgenotype C4 of human enterovirus 71 in Shan-dong, China. J. Clin. Virol. 44:262–267.

TABLE 1. Estimation of the nucleotide substitution rate in the VP1capsid region of CVA16

Data setSubstitution rate (substitutions per nucleotide per yr)a

Ks R2 Kt R2

Cluster B1ab 1.25 � 10�2 0.756 3.75 � 10�3 0.600Subgenotype B2c 1.19 � 10�2 0.973 3.41 � 10�3 0.976Avgd 0.91 � 10�2 2.49 � 10�3

a Ks, synonymous substitutions; Kt, total substitutions; R2, linear regressioncoefficient.

b Twenty-six strains, 1997–2008.c Nine strains, 1981–2000.d Average for cluster B1a and subgenotype B2.

622 NOTES J. CLIN. MICROBIOL.

on October 3, 2020 by guest

http://jcm.asm

.org/D

ownloaded from