The Cross-Neutralizing Activity of Enterovirus 71 Subgenotype C4 Vaccines in Healthy Chinese Infants and Children

Background EV71 is one of major etiologic causes of hand-foot-mouth disease (HFMD) and leads to severe neurological complications in young children and infants. Recently inactivated EV71 vaccines have been developed by five manufactures and clinically show good safety and immunogenicity. However, the cross-neutralizing activity of these vaccines remains unclear, and is of particular interest because RNA recombination is seen more frequently in EV71 epidemics. Methodology/Principal Findings In this post-hoc study, sera from a subset of 119 infants and children in two clinical trials of EV71 subgenotype C4 vaccines (ClinicalTrials.gov Identifier: NCT01313715 and NCT01273246), were detected for neutralizing antibody (NTAb) titres with sera from infected patients as controls. Cytopathogenic effect method was employed to test NTAb against EV71 subgenotype B4, B5, C2, C4 and C5, which were prominent epidemic strains worldwide over the past decade. To validate the accuracy of the results, ELISpot assay was employed in parallel to detect NTAb in all the post-vaccine sera. After two-dose vaccination, 49 out of 53 participants in initially seronegative group and 52 out of 53 participants in initially seropositive group showed less than 4-fold differences in NTAb titers against five EV71 strains, whereas corresponding values among sera from pediatric patients recovering from EV71-induced HFMD and subclinically infected participants were 8/8 and 41/43, respectively. The geometric mean titers of participants against five subgenotypes EV71 all grew significantly after vaccinations, irrespective of the baseline NTAb titer. The relative fold increase in antibody titers (NTAb-FI) against B4, B5, C2, and C5 displayed a positive correlation to the NTAb-FI against C4. Conclusions/Significance The results demonstrated broad cross-neutralizing activity induced by two C4 EV71 vaccines in healthy Chinese infants and children. However, the degree of induced cross-protective immunity, and the potential escape evolution for EV71 still need to be monitored and researched in future for these new vaccines.


Introduction
Enterovirus 71 (EV71), a member of the Enterovirus genus in the Picornaviridae family, is a highly infectious agent that causes handfoot-mouth disease (HFMD), herpangina, aseptic meningitis, encephalitis, and pulmonary edema in humans [1][2][3][4][5]. The past ten years have witnessed an increase in the severity of HFMD onset, the incidence of severe HFMD cases, and the number of mortalities in the West Pacific regions [6][7][8][9]. EV71 is now considered as the most dangerous neurotropic enterovirus of the post-polio era [8][9]. In order to prevent and control EV71associated epidemics, researchers in Mainland China, Chinese Taiwan, and Singapore have developed five inactivated EV71 vaccines, employing one EV71 strain of subgenotype C4, B3 or B4, respectively [10][11][12][13][14]. Candidates which induced high NTAb titers and exert protective effects in animals have entered clinical trials [14]. The clinical trial results also suggest that these vaccines have good safety, and satisfactory immunogenicity when tested by EV71 strains which belong to the same subgenotypes with their vaccine strains [13,15]. However, EV71 was classfied 11 subgenotypes (A, B1-B5, and C1-C5). In recent years, numerous large outbreaks of HFMD caused by different subgenotype of EV71 have occurred in Eastern and Southeastern Asian countries [3,[16][17][18][19][20]. By now, cross-protection against other genotypes and subgenotypes EV71 for EV71 vaccines in human has yet been elucidated, which is a key issue concerning the practical protective effects of the vaccine.
In our post-hoc study, serum samples from a subset of 119 participants (aged 6 months to 11 years) in two clinical trials of inactivated EV71 (subgenotype C4) vaccines (ClinicalTrials.gov Identifier: NCT01313715 and NCT01273246) [21,22] were detected for neutralizing antibody (NTAb) titers. Cytopathogenic effect (CPE) method was employed for B4, B5, C2, C4 and C5 subgenotypes EV71 strains, which were prominent epidemic strains worldwide over the past decade [20]. Our results suggest that inactivated EV71 vaccines derived from subgenotype C4 have a broad cross-neutralizing activity in Chinese infants and children.

Serum samples from EV71 vaccine clinical trial participants
Serum samples were acquired from a subset of 72 participants (aged 6 months to 5 years) in a clinical trial with a Clinical-Trials.gov Identifier of NCT01313715; and from 47 participants (aged from 6 months to 11 years) in a clinical trial with a ClinicalTrials.gov Identifier of NCT01273246 (Table 1) [21,22]. Both trials also included samples from healthy participants with no HFMD history. Participants received the inactivated subgenotype C4 EV71 vaccines (vaccine A and vaccine B) on day 0 and day 28; these vaccines were developed by Sinovac Biotech Co., Ltd or Bejing Vigoo Biological Co., Ltd with different viral strains (H07 and FY7VP5/AH/CHN/2008), cell culture system (cell factories and bioreator system), production process, and vaccine dosage [15,21,22]. The samples were sequentially selected based on the following criteria: the participant had a day 56 post-vaccination titer against EV71 subgenotype C4 of .1:8 [21,22], and a serum sample (collected before vaccination and/or 28 days after two doses) residual volume greater than 1.0 ml (to be used for crossneutralizing antibody tests). Written informed consent was received from donors' guardians.

Serum samples from HFMD pediatric patients
In the region of Guangxi Province of China, eight EV71infected HFMD patients were diagnosed during the outbreak of HFMD in 2010-2011. After the study was approved by the Guangxi CDC Ethics Committee, serum samples were collected from these patients 28 days after the onset of the disease, to serve as serum samples from HFMD recovery patients.
All serum samples were stored at 220uC before use.

Cells and Virus Strains
Rhabdomyosarcoma cells (RD cells: ATCC, CCL-136) were cultured in MEM solution (GIBCOL; USA) supplemented with 10% fresh bovine calf serum (GIBCOL; USA), 2 mM L-glutamine (GIBCOL; USA), 100 IU/ml penicillin and streptomycin (GIBCOL; USA). EV71 clinical isolates C2, C4, C5, B4, and B5 were kindly provided by National Institute of Diagnostics and Vaccine Development in Infectious Disease, School of Life Science, Xiamen University, China; and by Graduate Institute of Clinical Medicine, National Taiwan University and Hospital, Taiwan ( Figure S1). These EV71 clinical isolates were adapted for growth in RD cells. Virus titers ranged from 10 7 to 10 8 TCID 50 / ml. The virus strains were proliferated in RD cells. The infected RD cells were frozen and thawed three times to release the virus. After centrifuging at 2006g for 10 min to remove cell debris, the supernatant was collected and stored at 270uC.

CPE assays for the detection of NTAb against different EV71 genotypes
Classical CPE assay was employed to measure NTAb against EV71 genotypes [11,21,22]. Blood samples were inactivated at 56uC for 30 minutes, serially diluted two-fold from 1:8 and mixed with equal volumes of TCID 50 of a EV71 strain. The mixture was dispensed into a 96-well microplate and incubated at 37uC for 2 hours. RD cells (1-2610 5 cells/mL) were added to the mixture. The plates were then placed in a CO 2 incubator at 35uC for seven days. CPE was observed by microscopy. EV71 national standards were included in each test as a control for the reproducibility of the results [23]. Neutralizing antibody titers of EV71 were defined as the dilution rate showing 50% inhibition of the CPE. NTAb titers equal to or greater than 1:8 were defined as seropositive [11,[21][22][23][24].

Enzyme-linked Immunospot (ELISpot) Assay for the measurement of NTAb against different EV71 genotypes
In order to validate the accuracy of our cross-protection results acquired using the CPE NTAb detection method, a microscale ELISpot assay was employed simultaneously for NTAb detection. This technique has been used previously for respiratory syncytial virus and HCMV [25,26]. The ELISpot neutralization assay uses a monoclonal antibody against the viral capsid protein VP1 to detect EV71-infected cells. The same cells and EV71 virus strains of five subgenotypes with CPE method were used. After immunoperoxidase staining, infected cells were counted by an automated ELISpot analyzer. Neutralizing antibodies can suppress viral infection, leading to a reduction in the number of infected cells, which is reflected by a reduction in the number of spots. The neutralization titers were read as the highest dilution that completely inhibited over 50% of the number of spots. Results from the ELISpot neutralization assay were consistent with those from the classical CPE-based neutralization assay [27].

Statistical Analysis
Curve fitting and ID 50 value determination were performed with Prism 5 software (GraphPad Software, Inc.). Seropositive rates were compared by chi-square test. Statistical analysis of the geometric mean titers (GMTs) was done by SPSS 10.0 software, after subjecting the data to a log 2 transformation. This transformation was effective in stabilizing the dispersion and rendered the variances independent of the means. If the titers of neutralizing antibodies were negative, they were assumed to be 1:4 for calculation purposes. A paired t-test was performed, with p,0.05 considered statistically significant. Seroconversion is defined as pre-vaccination titre less than 1:8 and post-vaccination titre 1:32 or more, or pre-vaccination titre 1:8 or more and at least four-fold increase post-vaccination.

Participants
Serum samples from 119 participants who received two doses of inactivated EV71 vaccine on days 0 and 28 were analyzed in this study. The demographic characteristics of these blood donors are shown in Table 1. The donors were stratified into two subgroups according to their baseline EV71 NTAb level pre-vaccination [21,22]. The participants with baseline NTAb titer ,1:8 were categorized as the initially seroneg'ative group who were considered to be uninfected case, while the remaining participants with NTAb titer $1:8 were categorized as the initially seropositive group who were considered to be subclinically infected cases because they are all children without the history of HFMD.
This study also utilized the recovery-period sera of eight EV71infected HFMD patients (aged from 20 months to 73 months) in Guangxi Province from 2010-2011 (Table S1). These patients were categorized as the EV71 patient group (considered to be clinical infected cases, Table 1).

The NTAb response of participants after immunized with two C4 EV71 vaccines
Among 119 participants, serum samples from 83 participants were paired from pre-vaccination and 28 days after two-dose vaccination. NTAb titer of each paired serum sample was measured by CPE method with EV71 subgenotype B4, B5, C2, C4 and C5, respectively. As shown in Table 2, NTAb GMTs of the participants against 5 different subgenotypes EV71 all grew significantly after vaccination, irrespective of the baseline NTAb titers (all, p,0.0001). Among initially seronegative participants, the seroconversion rates of participants immunized with vaccine A were all 100% for 5 different subgenotypes and of participants immunized with vaccine B were 96.6-100% for 5 different subgenotypes; among initially seropositive participants who are immunized with vaccine A, the seroconversion rates were 56.7-73.3% for 5 different subgenotypes. There were no significant differences in seroconversion rate among the five EV71 subgenotypes in both initially seronegative participants and initially seropositive participants after vaccination ( Table 2). The fold increases in geometric mean titers (GMFIs) were 64.0-91.6, 44.3-57.7 and 5.4-6.8 fold in above groups after vaccination. No significant differences in GMFIs among the five EV71 subgenotypes were found in all groups (Table 2). However, significant NTAb titer increases against five different genotypes were observed in initially seronegative infants and children, compared with initially seropositive infants and children (p,0.05). All these results indicated that NTAb titer induced by C4 EV71 vaccine which was tested with EV71 strains of five genotypes was similar.
To further analyze the NTAb response against different EV71 subgenotypes induced by C4 EV71 vaccine, We made up the correlation between the fold increase of NTAb titers (NTAb-FI) against C4 subgenotype with that of the other four subgenotypes (Figure 1). The results showed that in the initially seronegative group, NTAb-FI against the other four genotypes were significantly correlated with that against C4 (p,0.0001 in both cases, r 2 in the range of 0.7636-0.8697 and 0.6827-0.9050, respectively). Among initially seropositive infants and children, the r 2 value was between 0.5977 and 0.7894, with p,0.0001 in both cases. The NTAb-FI against B4, B5, C2 and C5 showed a positive correlation with that against C4 ( Figure 1) among children and infants vaccinated with two C4 genotype EV71 vaccines. Correlation in NTAb-FI between C5 and C4 was indicated by the r 2 values of 0.9050, 0.8697, and 0.7894, which were the highest among other subgenotypes. These results suggested that compared to other genotypes, the antigenic properties of C5 are similar to those of C4, which is consistent with the results from EV71 VP1 evolutionary genetics analysis ( Figure S1).

The cross-neutralizing effect of sero-antibodies induced by two EV71 vaccines was tested against EV71 strains of five subgenotypes
In this part of study, to analyze the serum cross-neutralizing capacity of participants immunized with C4 vaccines, the sera of initially seronegative participants and initially seropositive participants were tested with 5 different genotypes strains and compared with data of sera from subclinically infected participants and clinically infected patients. The fold difference of each serum sample was calculated as highest NTAb titer/lowest NTAb titer among 5 EV71 subgenotypes. The fold difference was used to analyze the cross-neutralizing capacity of sera with different types, which was a common index for evaluating the cross-neutralizing capacity of sera [35,36]. As Figure 2 shown, in the initially seronegative group, there are 24 out of 24 cases showing less than 4-fold differences for vaccine A, and 25 out of 29 cases showing less than 4-fold differences for vaccine B. In the initially seropositive group, there are 22 out of 23 cases showing less than 4-fold differences for vaccine A, and 30 out of 30 cases showing less than 4-fold differences for vaccine B. Forty one out of 43 subclinically infected infants and children showed less than 4-fold difference. Eight out of 8 clinically infected patients showed less than 4-fold differences. Chi-square analysis showed that the percentage of cases with NTAb titer changes #4 fold was not significantly varied among above six groups (x 2 = 2.04, p.0.05). Four-fold difference rate for the post-vaccination sera was found no significant difference between initially seronegative and seropositive infants and children (x2 = 1.89, p.0.05). In addition, 4-fold difference rate between vaccine A and vaccine B was found no significant difference in their EV71 NTAb titers (x2 = 1.26, N: number of total participants included in the analyses. Data are GMT (95% CI) or GMFI (95% CI). GMT, geometric mean titer; GMFI, fold increase in geometric mean titers. Seroconversion is defined as pre-vaccination titre less than 1:8 and post-vaccination titre 1:32 or more, or pre-vaccination titre 1:8 or more and at least four-fold increase post-vaccination. a Negative antibody titers (i.e., ,1:8) were assigned a value of 1:4 for calculation purposes.
# Seroconversion rate was significantly higher in the initially seronegative infants and children than the initially seropositive infants and children after vaccination (p,0.05 in both cases).
*The GMTs of neutralizing antibodies against five EV71 subgenotypes increase significantly after vaccination (p,0.0001 in both cases

Comparison of the results obtained from the CPE and ELISpot assays
In order to verify the accuracy of the CPE results, we used China's national NTAb standard as a quality control in each CPE assay 23 , and ELISpot, a high-throughput detection method, to measure NTAb titers against different EV71 genotypes in all postvaccination serum samples. The correlation analysis was performed for all samples tested by both the ELISpot and CPE methods with EV71 B4, B5, C2, C4 and C5. The results shown that higher NTAb titer were found in ELISpot assay than in CPE assay for every subgenotype EV71. However, the r 2 between two assay were 0.9037, 0.9210, 0.8296, 0.9121 and 0.9156 against B4, B5, C2, C4 and C5 subgenotypes, respectively. There were significant relativity between the two assays for each subgenotype EV71(p,0.0001 in both cases; Figure 3). The results demonstrated that there was a good relativity between the two assays, which support the reliablity of the results in this study.

Discussion
EV71 belongs to Picornaviridae, a family of virus that is prone to genetic mutations. Recombinations can occur among EV71 strains of different genotypes, and occasionally even between EV71 and other enterovirus, such as CA16. These events lead to constant changes in the predominant strain in EV71 epidemics [18]. Genotype C2 was the predominant strain of the large HFMD outbreak in Taiwan in 1998 [6]. After 2000, B4 and B5 gradually replaced C2 and C3 as the predominant strains in Taiwan [28,29]. In Malaysia, EV71 strains C1, C2, B3 and B4, with B3 being the most prominent strain, were detected in 1997 [30]. Between 2000 and 2003, B4 and B5 replaced B3 to become the predominant strains [31]; all of which were then gradually replaced by C1. Epidemiological data from Singapore, Japan and additional regions have all confirmed this trend [32]. However, in China, ever since the first EV71 strain was isolated in 1997, C4 has remained to be the predominant strain [3].
A study of the evolutionary genetics of predominant EV71 strains in the past 40 years demonstrated that the phylogeny of EV71 is characterized by temporal structure with continuous strain replacement over time. This characteristic renders EV71 different from other enteroviruses, notably poliovirus, but similar to human influenza virus, which is a classic example of an antigenically variable pathogen [18]. It remains unclear whether genetic mutations of EV71 lead to alterations in antigenicity like those seen in influenza virus. Presently, little is known about the effect of genotype change on the antigenicity of EV71 [18]. Whether the immune response induced by one vaccine strain can effectively and adequately neutralize other strains is another key factor to consider for evaluating the safety and immunogenicity of these vaccines. To date, EV71 vaccines developed by mainland China, Chinese Taiwan, and Singapore are all vaccines of a single genotype. Thus, studying the cross-protective activity is of great importance for these vaccines.
Some studies have focused on the cross-neutralizing of animal immunized sera,the data demonstrated certain cross-protective activity against different EV71 strains. However, the crossprotection capacity in these animal studies shown more difference and were confusing [33,34]. And no study has focused on the cross-neutralizing activity of vaccinated human sera, especially those of infants and children, who are the target population, yet such study, could be particularly important for the understanding of the cross-neutralizing immunity elicited by vaccines composed of a single genotype of EV71 strain. In our study, the serum samples from a subset of 119 participants, aged 6 months to 11 years, in two C4 genotype EV71 vaccine clinical trials were used to detect NTAb against EV71 subgenotypes B4, B5, C2, C4 and C5. In order to validate the accuracy of the CPE results, ELISpot was employed in parallel to detect NTAb in all the post-vaccine sera. In both clinical trials, no significant EV71 epidemic was observed at the trial sites; and the EV71 NTAb titer of placebo groups, which were inoculated with aluminium adjuvant, showed no significant increase during the study period [21,22].
These results showed that, irrespective of the baseline NTAb titres being negative or positive before vaccination, the seroantibodies of infants and children, who were vaccinated with C4genotype EV71 vaccines A or B, have broad cross-neutralizing activity against EV71 strains C2, C4, C5, B4 and B5. Forty nine out of 53 participants in initially seronegative group and 52 out of 53 participants in initially seropositive group showed less than 4fold difference in neutralizing titer after two doses of vaccination, which are close to pediatric patients recovering from EV71induced HFMD and subclinical-infected infants and children. We further examined paired serum samples from 83 healthy participants before and after vaccination. Results from these samples showed that GMTs of participants against different subgenotypes of EV71 grew significantly after vaccination, irrespective of the baseline NTAb titer being negative or positive. The NTAb-FI against B4, B5, C2, and C5 showed positive correlation to NTAb-FI against C4 with p values. It is surprised and comforted that better cross-protection capacity and smaller neutralizing-antibody differences of human sera were shown than that of animal sera against EV71 with different subgenotype [33,34].
This study used 43 serum samples of subclinically infected infants and children and 8 serum samples of pediatric patients recovering from EV71-induced HFMD as controls. These samples and the serum samples of vaccinated participants were examined for cross-protective activity against different EV71 strains. The results showed that more than 95.3% serum samples of infected infants and children elicited homogenous cross-protective activity. These results agree with those reported by Mizuta and Huang [35,36]. Mizuta, et al. analyzed the cross-neutralizing ability of sera from 83 residents aged 1 to 60+ years in Japan. Their study suggested differences in the serum NTAb titer of residents against six subgenogroups (B2, B4, B5, C1, C2, and C4) were mostly within 4-fold [35]. Huang, et al. examined the cross-neutralizing ability of sera from 25 young Taiwanese children, who were infected by EV71 genotype C2, B4, C4, B5, or C4 from 1998-2010. Their study reported that differences in the serum NTAb infected by genotype C EV71 against all different EV71 subgenotypes, except genotype A, which is the first EV71 genotpye discovered in 1969, were mostly within 4-fold [36]. These results indicated that upon natural infection with EV71 genotype C, humans are able to generate serum antibody with high crossprotective activity against EV71 strains. At present, a highly conserved epitope of VP1, containing amino acids 211-224 (FGEHKQEKDLEYGAC), in the GH loop of the VP1 protein, was confirmed as the dominant linear neutralizing epitope [37]. This epitope was also present in three EV71 vaccine strains of subgenotype C4 and five EV71 detection strains in this study. These results suggesting that although the EV71 genotype defined based on the VP1 sequence was changeable, the neutralizing sites, which perhaps significantly affect the protective activity in humans, could be highly conserved. Evasion from the immune response due to antigenic shift therefore, may not be the main cause of periodic EV71 epidemics. In contrast, newborn accumulation in the epidemic regions, and a decrease of antibody levels in the population, could be the major causes for large-scale EV71 epidemics. This finding is consistent with the results of an epidemiological study conducted in Taiwan that found the positive rate of EV71 NTAb in the examined population was inversely correlated with the incidence of severe EV71 cases and EV71associated mortality [24].
In this study, EV71strains of genotype B4, B5, C2, C4 and C5 were selected to evaluate the cross-neutralizing effect of serum samples. Genotype C2 and C4 were the predominant strain of the wide HFMD outbreak in Taiwan in 1998 and in mainland china, which also played important role in EV71 epidemic [36,38]; while B4, B5 and C5 were the prominent strain in the worldwide epidemic over the past decade [31,39]. Evaluation study of the cross-protective ability of EV71 vaccine using these five genotypes of EV71 is thus more representative, and has better accordance with the EV71 epidemic situation. However, some researchers reported that current genotyping of EV71 does not reflect their antigenicity [40]. For instance, the EV71 C2-like genotype is a special and new subgenogroup of EV71 that was identified in Taiwan in 2008. EV71 C2-like genotype showed significant differences in antigenicity compared to other genotypes EV71 [41]. Therefore, further investigation is needed to identify the Figure 3. Correlation analyses of the NTAb titers against B4, B5, C2, C4 and C5 by CPE and ELISpot assays. Note: The NTAb titers against B4, B5, C2, C4, and C5, quantified by CPE assays, showed a statistically significant correlation with those detected by ELISopt assays (r 2 $0.8296, p,0.0001 in both cases). doi:10.1371/journal.pone.0079599.g003 The Cross-Neutralizing Activity of EV71 Vaccines PLOS ONE | www.plosone.org cause of the antigenicity change in this specific strain or subgenotype.
In the present study, we first used sera from healthy infants and children to assay the cross-protective activity of two EV71 C4 inactivated vaccines. These data indicated that EV71 vaccination of subgenotype C4 provided the broad and homogeneous protection against different subgenotypes of EV71 in health children and infants, similar to the outcome of natural infection. Altogether, our study demonstrated that the two EV71 inactivated vaccines produced from two different C4 strains, using different manufacturing processes, showed similar cross-protective activity, thereby providing the evidence supporting the broad immunoprotection of EV71-inactivated vaccines of C4 genotype. As newly designed vaccines, the degree of cross-protective immunity elicited, as well as the potential for EV71 to evolve to escape neutralization, remain to be investigated in clinical trials and monitored after market release. Further study of EV71 antigenicity and genetic evolution is warranted. Moreover, when the EV71 vaccine is inoculated simultaneously or at short-intervals with other vaccines, mutual interference or enhancement among them is worth exploring.