Summary of CVB4 data
During the 2011 HFMD outbreak in Shandong Province of China, 178 children were infected, including 25 severe cases (14.0%). Most children (76.4%) were under 6 years of age. In specimens from infected children, 122 (68.5%) enterovirus strains were detected in the laboratory. Among these, 31 (17.4%) were coxsackievirus B strains and 21(11.8%) were CVB4 (isolated from different cases), accounting for 67.7% of all detected Coxsackievirus B strains (Fig. 1).
Among 197 specimens from healthy children, 136 (69.0%) were enterovirus-positive of which 20 (10.2%) were identified as Coxsackievirus group B. Additionally, 14 strains were identified as CVB4, accounting for 70.0% of Coxsackievirus B detected in the study (Fig. 1).
Global CVB4 genotype distribution
In addition to 35 entire VP1 sequences of CVB4 obtained in Shandong Province in 2010 and 2011, 209 CVB4 sequences containing VP1 region were obtained from GenBank. Among the sequences obtained in GenBank, 29 did not meet quality control thresholds and 180 sequences were used for analyses. These isolates were from Australia (N = 10), Canada (N = 1), China (N = 35), Denmark (N = 17), Finland (N = 20), France (N = 12), India (N = 7), Italy (N = 1), Latvia (N = 1), Madagascar (N = 20), Netherlands (N = 32), Pakistan (N = 3), Poland (N = 4), Romania (N = 2), Russia (N = 2), Turkey (N = 1), United Kingdom (N = 1), United States (N = 9), and unknown areas (N = 2), representing a wide distribution (Table 1). A phylogenetic tree based on all entire VP1 sequences showed that CVB4 can be divided into genotypes A, B, C, D, and E (Fig. 2). Genotype A included the prototype strain (X05690/JVB/New York/US/1951) and the other two CVB4 strains (GenBank accessions S39291 and DQ480420). Genotype B–D contain strains from several countries other than China. Genotype E consisted of CVB4 strains isolated from China (N = 70) and Australia (N = 3). CVB4 strains isolated from China were all classified as genotype E, including the new sequences obtained in this study. CVB4 in China formed essentially an independent lineage in the phylogeny. Genotype E was the most prevalent CVB4 genotype in China. Basic data for each genotype are summarized in Table 1.
Table 1
Nucleotide sequence similarity, amino acid sequence similarity, and basic information for entire VP1 region sequences of CVB4 genotypes.
Genotype
|
GenBank accession number
|
Identity in VP1
|
Isolated country
|
Year
|
Intra-group distances(%)
|
Nucleotide
sequences(%)
|
Amino acid
sequences(%)
|
A
|
X05690, S39291, DQ480420
|
99.4–99.7
|
99.2–100.0
|
USA, Italy
|
1951,2006
|
0.3
|
B
|
AF160048- AF160073,
GU142875
KC558560- KC558573,
|
83.3–100.0
|
96.4–100.0
|
Australia, Canada, Denmark, Finland, Netherlands, USA
|
1959–1997
|
11.8
|
C
|
AF159998- AF160007,
|
86.0-99.8
|
97.5–100.0
|
Finland, France, Netherlands
|
1964–1996
|
11.5
|
D
|
AF160008-AF160047, AY919465, FJ868288, FJ868289, FJ868330-FJ868332, GU142877, HF948089, HF948090,JX513577, JX513579-JX513581, JX513583, KF177118,KF412921, KR232695, KR232711, KR232712, KR232730- KR232732, KR232739, KR232743, KR232751, KR232758, KR232760, KR232761, KR232763, KR232775, KR232776, KR232778, KR232782, KR232792, KR232793, KR232796, KU189236, KU189240, KU189242, KU189251, KU841463, KU841464, KY369904, LS451289, MG451808, MG845888,MG845889, MK044541
|
83.9–100.0
|
96.4–100.0
|
Australia, Finland, France, India, Latvia, Madagascar, Netherlands, Pakistan, Poland, Romania, Russia, Turkey, UK, USA
|
1977–2016
|
11.5
|
E
|
JF794741, JN016524, JX308222, KC867091-KC867093,
KF781519-KF781525, KJ933328, KJ933329, KM890276, KP289433, KU566507, KX752784, LC412980, MF179585-MF179588, MF422558-MF422563, MF678139, MF678300, MF678347, MG922519-MG922523
|
88.7–100.0
|
96.8–100.0
|
Australia, China
|
2007–2017
|
8.0
|
Evolutionary dynamics of global CVB4
With an uncorrelated relaxed molecular clock model (exponential) and exponential growth tree prior, MCC trees based on sequences of 200 globally distributed CVB4 were obtained using BEAST (Fig. 3). The average nucleotide substitution rate for the VP1 coding region in all CVB4 strains worldwide was 4.82 × 10− 3 substitutions per site per year (95% highest posterior density (HPD): 4.22 × 10− 3, 5.46 × 10− 3). The estimated coefficient of variation was 0.8925 (95% HPD: 0.8151, 0.9709), supporting the use of the uncorrelated relaxed molecular clock. The topological structure of the MCC tree constructed using BEAST was nearly identical to that of the ML tree constructed using RAxML. Global CVB4 strains isolated since 1930 formed two branches. Branch 1 contained genotype A arising in 1950 (including the prototype strain X05690/JVB/New York/US/1951) and genotype E arising in 2001. This branch included mainly Chinese and Australian strains. Branch 2 included genotype B (differentiated in 1951), genotype C (differentiated in 1959), and genotype D (differentiated in 1970). Isolates on this branch were geographically widespread (Oceania, North America, Asia, Europe, and Africa). CVB4 strains were continuously detected in many European countries, including Finland, France, and the Netherlands, since 1951, indicating that the strain was responsible for many epidemics with a long history in Europe. Based on the MCC tree and previous studies, CVB4 was prevalent in Madagascar in 2011, with 20 strains. Seven CVB4 strains were detected in children with diarrhea in India from 2005 to 2011 [39]. Additionally, 21 CVB4 strains were detected from specimens from the HFMD outbreak in Shandong Province in our study, indicating that CVB4 is prevalent worldwide.
According to a Bayesian skyline plot of the entire CVB4 VP1 sequence (Fig. 3), the global population diversity of CVB4 exhibited five changes over time. The first change was detected in 1945–1984, during which diversity increased. In 1945–1978, the growth rate was slow, and genotypes A, B, C, and D formed. During this period, CVB4 was mainly prevalent in the Netherlands and the United States. In 1979–1984, genotype B increased rapidly and appeared in Europe, Canada, and Australia. The second major change occurred in 1985–1989, when population diversity began to decline, possibly due to a global decline in the prevalence of genotype D. The third change occurred in 1990–1991, when the number of populations increased briefly, probably due to the global spread of genotypes B and D. In the fourth stage, 1992–2010, the population of CVB4 showed a slow decline, followed by a rapid decline. Isolates mainly belonged to genotype D during the first half of this stage and to genotype E, mainly in China, in the second half of the stage. In the fifth stage, beginning in 2011, the diversity of CVB4 increased rapidly and then remained steady. During this period, there were many epidemics and outbreaks of CVB4, which may be related to the rapid increase in diversity. Gradual stabilization since then may be due to a decrease in the quantity of CVB4 found in disease surveillance.
Phylogenetic analysis of CVB4 in China
A total of 70 CVB4 strains isolated from China were analyzed using BEAST with strict molecular clock model and Bayesian skyline tree prior (Fig. 4). The average nucleotide substitution rate of the CVB4 VP1 coding region in China was 5.20 × 10− 3 substitutions per site per year (95% HPD: 3.69 × 10− 3, 6.78 × 10− 3), which was slightly faster than that of the VP1 coding region in global CVB4 strains. The CVB4 strains isolated in China arise in 1995 and were detected between 2007 and 2014. The areas of monitoring covered Shandong Province, Inner Mongolia Autonomous Region, Fujian Province, Yunnan Province, and other regions in China. According to the Bayesian skyline plot of the entire VP1 sequence of CVB4, the change in CVB4 diversity in China can be divided into two stages. The decline before 2010 and rapid increase in diversity after 2011 suggest that the outbreak of HFMD in Shandong Province was an important determinant of the increase in CVB4 diversity in China.
The MCC tree showed that CVB4 strains isolated from the HFMD outbreak in Shandong Province in 2011 and a portion of CVB4 strains isolated from healthy individuals in 2010 diverged from a common ancestor (with a posterior probability of up to 0.89). The nucleotide sequence similarity between VP1 sequences of CVB4 in Shandong in 2011 was 99.6–100% and the amino acid sequence similarity was 100%. The nucleotide sequence similarity in comparison with CVB4 strains isolated in Shandong Province in 2010 was 95.7–99.4% and amino acid sequence similarity was 97.8–98.9%. The nucleotide sequence similarity with other CVB4 isolates from China was 90.9–98.8% and amino acid sequence similarity was 97.5–99.2%.