International Journal of Vaccine Research

Development of enterovirus A71 (EV-A71) vaccine to prevent severe symptoms or even fatalities caused by EV-A71 infection is very much interested worldwide. In this study, high titers of EV-A71 could be produced with Vero cell cultures either by roller bottle or small scale bioreactor system. To inactivate the virus, we found that appropriate condition to inactivate the virus completely was 0.01% formaldehyde, 37oC for 24 hr, while at 4oC either with 0.01% or 0.02% formaldehyde, the virus still existed after 240 hr incubation. We developed purification process of the inactivated EV-A71 by size exclusion with the techniques of 100 kDa tangential flow filtration (TFF), subsequent 10% - 50% sucrose density gradient ultracentrifugation, and washing by ultracentrifugation with 30% sucrose, and characterized the purity by parameters as follows. Comparable to initial amounts before taken to purification by these methods, residual total proteins were decreased as 44.6%, 0.9% and 0.3%, respectively, bovine serum albumin was 58.8%, 0.04% and 0.01%, respectively, Vero cell proteins was 53.0%, 0.58% and 0.03%, respectively, while Vero cell DNA which was monitored after 10% - 50% sucrose density gradient ultracentrifugation, and washing by ultracentrifugation were 0.06% and 0.008%, respectively, however, the recovery of VP1 antigen was only 2.1%. To evaluate the purity according to standard requirements, a predicted human dose of 1.0 μg/0.5 ml total proteins of purified inactivated EV-A71 was set. By this, some revealed 180 – 900 doses per purification run. All parameters met the requirements as the maximum amount of total proteins was only 1.0 μg, BSA was 5.7 ng, and Vero cell DNA was 1.03 ng per dose, while Vero cell proteins were 0.126 μg/ml. Moreover, the amount of VP1 antigen ranged 233 – 599 ng per dose. This purification process demonstrated high purity of inactivated EV-A71 that should be investigated further as an EV-A71 vaccine candidate.

Keywords: Enterovirus A71 vaccine; EV-A71 vaccine; inactivated EV-A71 vaccine; vaccine purification; EV-A71 vaccine purification;

Hand, foot and mouth disease (HFMD) is a common virus illness in infants and children caused by enteroviruses, and enterovirus A71 (EV-A71) is known to be the cause of the disease with severe morbidity and mortality. EV-A71 infections, although present in most countries, the largest outbreaks of disease have been reported in many countries in the Asia-Pacific region, and since 1997, EV-A71 infections have been a major public health burden and epidemiologic concern in the Asia-Pacific Region [1, 2, 3]. Severe manifestations caused by EV-A71 are neurological symptoms which range from aseptic meningitis to acute flaccid paralysis and brainstem encephalitis, and association with severe pulmonary edema and shock in many cases. Due to the lack of antiviral drug against EV-A71 to treat the cases of severe manifestations, therefore developing the vaccine to prevent severe symptoms or even fatalities caused by this virus is very much interested worldwide. There have been several types of the vaccine being developed. Each type has its own advantages and disadvantages, however, the inactivated EV-A71 vaccine is considered the safest viral vaccine, as there will be no reversion to an infectious virus, therefore, we were interested to develop the inactivated EV-A71 virus as the vaccine candidate. Since Vero cells are susceptible for EV-A71 growth, and adherent Vero cells can be easily scalable for further larger virus production in bioreactor systems, in this study we thus used Vero cell cultures to propagate the virus in roller bottles and also smaller scale bioreactor system to provide product viruses for studying purification process.

From the cell cultures, plenty of bimolecular rather than the virus product were generally generated from host cells, and since culture medium was added as well as serum for optimal growth of mammalian cells as it is a source of nutrients, hormones and growth factors, and also facilitate the attachment and spreading of cells, and provide protection against mechanical damage and shear forces, various proteins, lipids, nucleic acids, and other bimolecular as well as bovine proteins from serum were removed intensively from the virus product [4, 5]. In addition, the residual DNA from Vero cells has been proofed positive to carcinogenicity test after over 170 passages, thus their genomic DNA carried in vaccines could be oncogenic and hence poses a potential risk for human carcinogenesis [6, 7 and 8], therefore, strictly removal of Vero cell DNA was needed. Since tangential-flow filtration is an efficient method for concentration and separation of bimolecular by retaining product molecules and removing unwanted molecules through membrane by size exclusion, and density gradient centrifugation can be used for purification of bimolecular on the basis of size and density through a gradient of viscous liquid, such as sucrose or glycerol, and when centrifuged, with gravity, virus particles will migrate until reaching their densities. We thus developed a purification process based on such techniques.

Recently, we have described elsewhere that our inactivated EVA71 vaccine candidate produced from Vero cells and purified the inactivated virus similarly as described in the present report, gave high neutralizing antibodies in mice [9]. In the present study, we achieved small scale process for producing the inactivated EV-A71 vaccine candidate including the condition for virus inactivation, down-stream purification based on tangential-flow filtration and sucrose gradient ultracentrifugation, and characterization of product purity in various parameters. Therefore, these results provided valuable relevant information for cell-based EV-A71 vaccine development for further investigations.

The sub-genotype C4 EV-A71, strain THA-08-29961 isolated from a fatal case with severe hand-foot and mouth disease in Thailand in 2008 (Kindly provided by Guntapong R, et al, National Institute of Health, Department of Medical Sciences, Thailand) was used to prepare the virus stock by propagation in 90% Rhabdomyosarcoma (RD, ATCC: CCL-136) confluent cell monolayer in modified Eagle’s medium (MEM) with 2% fetal bovine serum (FBS) as described elsewhere to prepare as the target EV-A71 [10].

Viral titration was performed by the plaque assay. Confluent monolayers of RD cells were prepared in 24-well plates (2x105cells/well). The cells were infected with serial dilutions of viral suspensions, overlaid with 1.5% agarose gel in the culture medium DMEM + 2% fetal bovine serum (FBS), and incubated at 37°C for 3 days. To visualize the plaques, stain the gel with crystal violet, and the viral titer was estimated in pfu/ml by the plaque assay [11].

To propagate the EV-A71 in a roller bottle, 2.0 x 107 Vero cells (CCL-81, ATCC, VA, USA) were seeded in a 850 cm2 roller bottle containing the culture medium EMEM + 10% FBS until cell monolayer was performed. EV-A71 was produced by inoculating the EV-A71 seed onto Vero cell monolayer at a multiplicity of infection (MOI) of 10−5 with EMEM + 2% FBS, and bottles were rotated at 0.33 rpm at 37oC on a roller bottle rack as previously described [9]. EV-A71 was collected from the culture supernatant of each roller bottle at day 4 post infection and determined by the plaque assay [11]. The cells were lysed by freezing-thawing at -80oC and 37oC for 3 times, and the lysate was collected by centrifugation at 2,000 RPM at 4oC for 20 min. Cell debris were removed by filtration through a 0.65 μm membrane (Sartorius Stedim Biotech, USA).

To propagate the virus in a small scale bioreactor, 1.2 x 108 Vero cells were grown on Micro carrier beads (Cytodex-1, Merck) in a spinner flask (Merck), and stirred at 30 rpm in 37°C at 5% CO2. The mixture of cells and micro carriers was transferred to 2.0 L EMEM (Thermo Fisher Scientific) containing 10% fetal bovine serum (Sigma Aldrich) in a 5-L bioreactor (Biostat A, Sartorius) operated under the conditions of 40 rpm, 37°C, pH 7.4, and partial pressure of oxygen (pO2) of 50% air saturation. After day 3, when the cell density exceeded around 1.0 × 106 cells/ml, the culture medium was removed, and 0.0001 MOI EV-A71 was inoculated to the cell monolayers formed on the outer surface of micro carriers for 5 hr, then the bioreactor was filled with the medium up to 2.0 L and continuously operated at 40 rpm, 37°C, pH 7.4 and 50% air saturation and incubated further for another 3 days and then harvested. The virus growth was observed by 95% of cells exhibited the cytopathic effect (CPE) under a converted microscope, and determined by the plaque assay [11], then the virus was harvested similarly as those cultured by roller bottles.

Formaldehyde solution (37%) (Merck) was added into 30.0 ml culture medium EMEM containing 1.8 x 107pfu/ml EV-A71 as 0, 0.01% or 0.02% final concentration, and each condition was incubated at 37oC for 0, 24, 48 and 72 hr, or at 4oC for 0, 48, 96, 168, and 240 hr. After each time interval, the suspensions were determined residual EV-A71 by the plaque assay [11].

Before EV-A71 purification, the culture medium supernatant was harvested by centrifugation and cell debris removal by filtration through 0.65 μ as described earlier. The virus was then inactivated with the optimal condition studied earlier as 0.01% formaldehyde at 37oC for 24 hr.

The inactivated EV-A71 was then concentrated, and some unwanted proteins, lipids, nucleic acids and salts were removed by using a 100 kDa cut-off tangential flow filtration (TFF) membrane cassette (Sartorius Stedim Biotech, USA) by filtering the inactivated EV-A71 through the TFF apparatus against PBS pH 7.4 of at least 5 times higher volume of the inactivated virus suspension, and the retentive was collected.

The inactivated EV-A71 was then purified further with 2 volumes of continuous 10% - 50% sucrose density gradient ultracentrifugation by mixing 10% with 50% sucrose solution in PBS pH 7.4 into an ultracentrifuge tube using a gradient former (Model 385, Bio-Rad), the inactivated EV-A71 was then layered over the gradient and centrifuged at 36,000 rpm, 4oc for 3 hr using a zonal rotor in an ultracentrifuge (CP-NX Series, Hitachi, Japan). After centrifugation, 1.0 ml each fraction was collected from the top until the pellet at the bottom of the tube. Further optional was washing step by centrifugation the pellet with 30% sucrose at 36,000 rpm, 4oC for 90 min, and the pellet collected was dissolved with PBS pH 7.4. The fractions and pellet collected were checked for total protein by the Bradford assay as described by the manufacturer (Bio-Rad), and the EV-A71 VP1 protein by quantitative ELISA as described further.

Anti-EV71 VP1 mouse polyclonal antibody (GeneTex, USA) was diluted to be 0.5 μg/ml and coated onto microtiter plates at 4oC overnight. One hundred μl EV-A71 samples diluted as 1:10 to 1:10,000 as appropriated were added onto the coated plates and incubated for 1 hr, and 1 μg/ml anti-EV71 VP1 mouse monoclonal antibody (Abnova, USA) was then added into the wells and incubated for 1 hr. To determine the reaction, 1:10,000 anti-mouse IgG antibody-horse radish peroxidase conjugate (KPL, USA), and 3, 3’, 5, 5’-tetramethylbenzidine (TMB) solution (KPL, USA) were added. In this assay, diluting the reagents, blocking, washing the reaction mixtures, and removing nonspecific binding agents from each step were done by using the solutions and buffers supplied by KPL, USA, which the procedures were as described by the manufacturer. To estimate the amount of VP1 in the samples, the standard curve of VP1 protein was established by replacing the sample with various concentrations of VP1 (EV71) protein (Immune Technology Corp., USA) between 2.5 and 500 ng/ml.

Various dilutions of residual Vero cell proteins in the inactivated EV-A71 samples as non-diluted, 1:10 or 1:100 as appropriated in PBS pH 7.4 were determined using the Vero Host Cell Proteins ELISA kit (Cygnus Technologies Inc., North Carolina, USA) by the procedure described by the manufacturer. The amount of Vero cell proteins was estimated from the standard curve generated by 0 - 200.0 ng/ml standard Vero cell proteins (Cygnus Technologies Inc., North Carolina, and USA). To determine residual Vero cell proteins in the samples either after 10% - 50% sucrose density gradient ultracentrifugation, or after the optional washing step.

To determine residual Vero cell DNA in the inactivated EVA71 samples either after 10% - 50% sucrose density gradient ultracentrifugation, or after the optional washing step, the DNA sample was purified by using QIAamp DNA Mini (Qiagen, Germany) as described by the manufacturer. Residual Vero cell DNA was then detected by real-time PCR using resDNASEQ® Vero Residual DNA Quantization System (Life Technologies, USA). The method performed was described by the manufacturer, and the control Vero cell DNA, primers and all other reagents as the assay mix were supplied in the kit.

To determine residual bovine serum albumin (BSA) in the inactivated EV-A71 samples was by using BSA ELISA kit (Cygnus Technologies Inc., North Carolina, USA) according to the kit instructions. BSA concentration was estimated from the standard curve generated by the standard BSA 0, 0.5, 2.0, 8.0, 32.0 ng/ml provided by the kit.

SDS-PAGE and Western blot analyses of the EV-A71 antigens from Vero cell cultures were performed according to the protocols reported previously by Liu CC, et al. [12], and molecular weight markers (PageRuler Prestained Protein Ladder, Thermo Scientific) were also run simultaneously. For immunoblotting, the proteins were directly electro-transferred onto the PVDF membrane. Each membrane was incubated with PBS pH 7.4 containing diluted (1∶1000) either EV-A71-specific monoclonal antibodies against VP1 or VP2 antigen. Antibodies were bound for 2 hr at room temperature. Binding of the respective antibodies to the viral proteins was detected by adding 2.0 ml phosphate buffer saline (PBS) containing a horseradish peroxidase (HRP)- conjugated anti-mouse secondary antibody (Thermo Fischer Scientific) at a dilution of 1:10,000. After 1-hr incubation at room temperature, the membrane was washed six times with the assay buffer and blotted dry. The protein bands were revealed by adding TMB substrate solution (KPL).

When the EV-A71 THA-08-29961 strain was cultured in Vero cell monolayers either those attached on the inner surface of 850 cm2 roller bottles, or on the outer surface of the micro carrier beads in small scale bioreactor system, Vero cells and the virus could be produced extensively, obviously, very larger amount of the cells and the virus were revealed by the bioreactor system than by the roller bottles due to the culture volumes and the amount of starter cells. By SDS PAGE which all proteins could be stained, both lanes 2 and 3 in Figure 1 loaded with inactivated EV-A71 THA-08—29961 produced from Vero cell culture, had many protein bands of different sizes observed (figure not shown) which meant that the purification methods used could not remove all other proteins and leave only the viral antigens needed. However, to determine the antigens of EV-A71 particles, Western blot analysis was performed further. When the purified proteins from EV-A71 cultures were stained either with VP1- specific or VP2-specific monoclonal antibodies, protein bands were seen positive and the sizes of VP1 was estimated around 36 kDa as shown in Figure 1 A, while there were two bands against VP2-specific monoclonal antibody of VP0 and VP2 found at around 38 kDa and 28 kDa, respectively, as shown in Figure 1 B. Moreover, in B some aggregated proteins of VP2 or VP0 were also found larger than 60 kDa or more which might be the particle of VP4 + VP2 +VP3 [12]. Figure 1.

As shown in Table 1, from equal initial EV-A71 titer of 1.8 x 107 pfu/ml started, at 37oC either with 0.01% or 0.02% formaldehyde treated for 24 hr, whole of the virus was inactivated completely as zero titer was shown, while without formaldehyde at this temperature, after 72 hr the virus titer was still left to 1.5 x 106 pfu/ml, although the titers were gradually decreased as the time passed. In contrary, at 4oC, when 2.9 x 107 pfu/ml initial titer was treated with 0.01% or 0.02% formaldehyde, up until 240 hr (Day 10) the virus still existed as 1.0 x 105 and 8.8 x 103 pfu/ ml, respectively, while without formaldehyde at 4oC, by longer incubation time for 240 hr, high virus amount (1.5 x 107 pfu/ml) were still left.

Before monitoring the efficiency of purification process of inactivated EV-A71 developed in this study, it was proofed that the yield of EV-A71 among the very first steps of the culture harvest before taken to virus inactivation was quite consistent. As shown in Table 2, EV-A71 of different culture batches after 0.65 μ ultra filtration yielded 97.4% mean virus amount to those of the harvests, thus the initial virus amount taken to be purified was approximately unchanged before inactivation. Therefore, monitoring of purity of the inactivated virus product was assumed to start at the step of virus inactivation afterwards. Since parameters to monitor impurity of the inactivated EV-A71 in this study were total proteins, Vero cell proteins and DNA, and BSA, the amount of each parameter was quantified as following results. As seen in Table 3 as from several culture samples were shown, after virus inactivation and then consecutive purification methods performed, initially, by 100 kDa tangential flow filtration (TFF) residual total proteins were 34.0% - 50.3% (44.6% mean) comparable to those of the inactivation step, while by further 10% - 50% sucrose density gradient ultracentrifugation, it removed more total proteins to 0.3% - 1.2% (0.9% mean), and if washing step by centrifugation with 30% sucrose was done, remaining total proteins were 0.1% - 0.6% (0.3% mean).

For removal of Vero cell proteins, from five inactivated EVA71 samples taken, by 100 kDa TFF, Vero cell proteins were eliminated from the inactivated virus as 33.1% - 88.6% (53.0% mean), while consecutive 10% - 50% sucrose density gradient ultracentrifugation removed much greater amounts of Vero cell proteins as only 0.02% - 1.5% (0.58% mean) were left, and further washing by 30% sucrose removed the host proteins more as 0.01% - 0.09% (0.03% mean) were left as shown in Table 4.

To remove BSA, as shown in Table 5, by 100 kDa TFF BSA was removed as 43.4% - 81.4% (58.8% mean) left, after 10% - 50% sucrose density gradient ultracentrifugation, from two runs performed, residual BSA was left as 0.02% - 0.05% (0.04% mean), but if washing with 30% sucrose centrifugation was further done, no BSA was found or only 0.01% maximum left.

All Vero cell DNAs were also aimed to remove. By the purification methods described, as comparable to the amounts of Vero cell DNA at the step of virus inactivation, at the end of purification either by 10% - 50% sucrose density gradient ultracentrifugation or further washing by ultracentrifugation with 30% sucrose, residual Vero cell DNA was left only 0.06% and 0.008%, respectively, as shown in Table 6.

In this study, to search for which fraction that most of the target virus accumulated during 10% - 50% sucrose gradient ultracentrifugation, we set the same experiment with live EV-A71 viruses to determine the virus titers to lead to fractions of virus accumulation. Equal fifty fractions were collected after the run and determined total proteins versus the virus titers. As shown in Figure 2, the curve of total proteins was higher as higher fractions obtained which the highest concentrations were steady between fractions 24 – 29, after that it declined until fraction 47, then total protein concentrations were increased much again at fractions 49 – 50, while the EV-A71 virus titers were depleted from the first fractions but accumulated only in the two last fractions of 49 and 50. This experiment guided that the accumulation of inactivated EV-A71 might also be at the fractions 49 – 50, therefore, these two fractions were determined for the recovery of inactivated EV-A71 which was monitored by measuring the amount of the VP1 antigen of EV-A71 by quantitative ELISA as described. By the purification methods used, from four different experiments, at the end, quite small amount of VP1 was left as % recovery was ranged 0.85% - 3.2% (2.1% mean) after 10% - 50% sucrose density ultracentrifugation comparable to those found at the step of virus inactivation as shown in Table 7, while after washing with 30% sucrose, a bit less % recovery was found (data not shown). Figure 2.

As reported by Promkhatkaew D, et al [9], we have investigated immunogenicity in BALB/c mice of the inactivated EV-A71 purified accordingly to the procedures used in this study, by varying the immunogen as 1.0 or 2.5 μg total proteins which high neutralizing antibodies against the homologous EV-A71 virus have been shown, therefore, safety and immunogenicity in other animal models and in human were expected to study further if it is applicable. Consequently, examples of purified inactivated EV-A71 total proteins performed accordingly as described were elucidated for design of a human vaccine dose and evaluation of vaccine candidate purity. In the Table 8, if 1.0 μg total proteins per 0.5 ml volume was designed to be a human dose, thus total numbers of human doses were estimated accordingly to total proteins obtained ranging 180 – 900 doses per purification run and culture volume. After purification, due to the yields of VP1 left in each sample, one dose might contain 233 -599 ng VP1 antigen. To evaluate purity, residual BSA, Vero cell DNA, and Vero cell proteins were 0.41 – 5.7 ng, 0.04 – 1.03 ng, and 0.054 – 0.126 μg/ml per human dose, respectively.

In this study, EV-A71 could be propagated extensively in Vero cell cultures both by roller bottles and small scale bioreactor system. Types of cultivation were dependent on amount of the virus needed to handle for purification, and in order to study preliminarily a bioreactor scale upstream process. The viruses obtained by both cultivation types exhibited similar characteristics as determined by plaque assay, and with specific anti-EV-A71 VP1 monoclonal antibody by quantitative ELISA and anti-EV-A71 VP1 or VP2 monoclonal antibody by Western blot. Since no infectivity was detected in the EV-A71 samples taken for 24 hr after exposure to 0.01% formaldehyde at 37oC, this condition was selected to do virus inactivation throughout the study, whereas, by the temperature of 4oC for 240 hr (10 days) the virus was still detected, even after 30 days (data not shown). Additionally, there has also been some evidence reported that low temperature of 4oC bore a high risk of incomplete inactivation, while higher temperatures were more efficient [13], moreover some has reported the inactivation of EV-A71 at 37oC but with longer time for 3 days [12]. Although at 37oC, we did not vary the incubation time less than 24 hr, but this time interval seemed to be convenient to do and it certainly gave complete EV-A71 inactivation. The present formaldehyde inactivation condition did not modify the EV-A71 antigens since the VP1 antigen could be detected by quantitative ELISA, and VP1 and VP2 antigens by Western-blot as shown in this study, and moreover, our inactivated virus processed similarly could induce neutralizing antibodies in mice against EV-A71 virus as reported elsewhere [9].

The EV-A71 taken to use for study of purification either from roller bottles or bioreactor system was in the harvests of culture media since these were easier to obtain and the amounts of the virus released from Vero cells were quite high enough. After removing all cell debris by filtration with 0.65 μ membrane which was the step prior to virus inactivation where the virus mean titer was as high as 97.4% comparable to those of the culture supernatant, thus the step of virus inactivation was subjected to an initial step for monitoring degrees of purification. In this study TFF with 100 kDa membrane was selected to primarily remove molecules like most of proteins, nucleic acids, lipids and other cellular materials having the size smaller than 100kDa in the permeate, and expected to retain the inactivated virus in the retentive of the runs. We found that this method was effective as abundant total proteins were removed as the mean residual total proteins was 44.6%, and much more total proteins could be removed by further methods of 10% - 50% sucrose gradient ultracentrifugation and washing with 30% sucrose, as 0.9% and 0.3% total proteins left, respectively. Moreover, from the experiments, the process also removed significant amounts of BSA, and the proteins and DNA of Vero host cell.

Since in some other studies, 1.0 μg and 0.25 μg total proteins, as relatively as 640 U and 160 U, respectively, of Vigoo inactivated EV71 vaccine have been used as doses for immunization in adults and children to study tolerability and immunogenicity in China, that no serious adverse event was observed after two doses, and gave overall efficacy as 94.8% during 2-year follow up, we might hypothesize the same amount of 1.0 μg total proteins for a human dose in order to evaluate the purity of our vaccine candidate [14,15] . As in Table 8, if our human dose was of 1.0 μg total proteins, thus it would certainly not be greater than 10 μg per human dose as specified by some WHO Technical Report Series [16]. For residual BSA which should not be greater than 50 ng per human dose [17], our purified inactivated virus met the requirement since the maximum amount was only 5.7 ng/dose estimated. Moreover, the left of Vero cell DNA of which the higher was 1.03 ng/dose shown was within the requirement of WHO as shall be less than 10 ng per human dose [17]. Similarly, residual Vero cell proteins which were maximally found 0.126 μg/ml was less than 0.32 μg/ml found in the inactivated EV71 vaccine of Liang Z, et al that have claimed to met the specifications of Pharmacopoeia of the People’s Republic of China, 2005, and WHO Recommendations for the Preparation, Characterization and Establishment of International and other Biological Reference Standards (Revised 2nd edn) [18].

For the VP1 amounts shown in Table 7, those data of VP1 could not be brought to summarize in Table 8 since the amounts of respective total proteins of such samples were not available to interpret VP1/dose, however, in Table 7, % recovery comparable to initial VP1 amounts of inactivation step were revealed. Actually, although we only determined the quantity of VP1, other EV-A71 structural capsid proteins like VP2, VP3 and VP4 might also exist in the product, where not only in VP1 but various neutralization epitopes have been reported in VP2 and VP4 [19]. By 10% - 50% sucrose gradient ultracentrifugation performed in our study, most of total proteins were accumulated at around 25% - 30% sucrose, while most of live EV-A71 were shown sedimented at around 45% -50% sucrose as shown in Figure 2, however, since we did not monitor EV-A71 antigens throughout all fractions collected, we could not proof whether there were some viral antigens which had smaller molecular sizes deposited along different densities or not.

However, the inactivated EV-A71 processed as described has already been studied immunogenicity in mice, of which two and three injections of this candidate with and without alum have shown desirable effective neutralizing antibodies [9]. In conclusion, this study presented primarily downstream process in smaller scale with satisfaction of purity of the inactivated virus, however, it seemed revealing very low recovery of the antigen detected, this might need more studies to reveal easier or less method for purification but should still reveal similar or higher product yield with standard purity. Since we used the techniques to separate bimolecular based on sizes, therefore if the product was still an intact particle which might be composed of sixty copies of VP1+VP2+VP3+VP4 capsid proteins, the size of the antigens might be larger that could not pass the TFF membrane, but if the particle was disaggregated presenting smaller size capsid molecules ranges 8 – 59 kDa [12], those molecules might pass through the membrane used, and be accumulated at other sucrose densities unexpected to collect by the 10% - 50% sucrose gradient used. Consequently, we predicted to use 1.0 μg total proteins to be a human dose, thus this might contain 233 – 599 ng VP1 antigen/dose which was quite in high amount. Whether this dose would appropriate to induce protective neutralizing immune response or even safety to adults and children would have to be investigated in animals prior to human further.

We were grateful to Mr. Ratigorn Guntapong and Ms. Ratana Tacharoenmuang, National Institute of Health, Department of Medical Sciences, Ministry of Public Health for kindly provision us the enterovirus A71 strain THA-08-29961 used in this study.

1. McMinn P, Lindsay K, Perera D, Chan HM, Chan KP, Cardosa MJ. Phylogenetic analysis of enterovirus 71 strains isolated during linked epidemics in Malaysia, Singapore, and Western Australia. J Virol. 2001;75(16):7732-7738.
2. Cardosa MJ, Perera D, Brown BA, Cheon D, Chan HM, Chan KP, et al. Molecular epidemiology of human enterovirus 71 strains and recent outbreaks in the Asia-Pacific region: comparative analysis of the VP1 and VP4 genes. Emerg Infect Dis. 2003; 9(4):462–468.doi: 10.3201/eid0904.020395
3. Shimizu H, Utama A, Onnimala N, Chen L, Zhang LB, Ma YJ, et al. Molecular epidemiology of enterovirus 71 infection in the Western Pacific Region. Pediatrics International. 2004;46(2):231-235. Doi: 10.1046/j.1442-200x.2004.01868.x
4. Croughan MS, Hamel JF, Wang DI. Hydrodynamic effects on animal cells grown in microcarrier cultures. Biotechnol Bioeng. 2000;67(6):841–852.
5. van der Pol L and Tramper J. Shear sensitivity of animal cells from a culture-medium perspective. Trends Biotechnol 1998;16(8):323–328.
6. WHO Expert Committee on Biological Standardization. World Health Organ Tech Rep Ser 1998; 872: i-vi, 1-90.
7. Cao S, Dong G, Tang J, Li J, Liu J, Shi L, et al. Development of a Vero cell DNA reference standard for  residual DNA measurement in China. Hum Vaccine Immunother 2013; 9(2):413-419.
8. Ferguson M and Schild GC. A single-radial-immunodiffusion technique for the assay of rabies glycoprotein antigen: application for potency tests of vaccines against rabies. J Gen Virol 1982; 59:197–201.
9. Promkhatkaew D, Pinyosukhee N, Wichajarn R,  Thongdeecharoen W, Manoch Posung M, Tuntigumthon S, et al. Neutralizing Antibodies of Inactivated Thai Enterovirus A71 Strain in Mice for Development of Enterovirus A71 Vaccine. Int J Vaccine Res. 2019; 4(1):1-11. Doi: 10.15226/2473-2176/4/1/00130
10. Solomon T, Lewthwaite P, Perera D, Cardosa MJ, McMinn P, Ooi MH. Virology, epidemiology, pathogenesis, and control of enterovirus 71. Lancet Infect Dis 2010; 10(11):778-790. Doi: 10.1016/S1473-3099(10)70194-8
11. Baer A and Hall KK. Viral concentration determination through plaque assays: using traditional and novel overlay systems. J Vis Exp 2014;(93):e52065. Doi: 10.3791/ 52065
12. Liu CC, Guo MS, Lin FH, Hsiao KN, Chang KH, Chou AH, et al. Purification and characterization of enterovirus 71 viral particles produced from Vero cells grown in a serum-free microcarrier bioreactor system. PLoS One 2011;6(5):e20005. Doi: 10. 1371/journal.pone.0020005
13. Moller L, Schünadel L, Nitsche A, Schwebke I, Hanisch M, Laue M. Evaluation of Virus Inactivation by Formaldehyde to Enhance Biosafety of Diagnostic Electron Microscopy. Viruses. 2015; 7(2): 666–679. Doi:10.3390/v7020666
14. Fan-Yue Meng, Jing-Xin Li, Xiu-Ling Li, Kai Chu, Yun-Tao Zhang, Hong Ji, et al. Tolerability and immunogenicity of an inactivated enterovirus 71 vaccine in Chinese healthy adults and children. Human Vaccines & Immunotherapeutics. 2012;8(5);668–674. Doi: 10.4161/hv.19521
15. Wei M, Meng F, Wang S, Li J, Zhang Y, Mao Q, et al. 2-Year Efficacy, Immunogenicity, and Safety of Vigoo Enterovirus 71 Vaccine in Healthy Chinese Children: A Randomized Open-Label Study. J Infect Dis. 2017; 215(1):56-63. Doi: 10. 1093/infdis/jiw502
16. Recommendations for the production and control of poliomyelitis vaccine (inactivated) WHO Technical Report Series, No. 910, 2002.
17. Recommendations for inactivated rabies vaccine for human use produced in cell substrates and embryonated eggs. WHO Technical Report Series No 941, 2007.
18. Lianga Z, Mao Q, Gao Q , Li X,  Dong C, Xiang Yu X, et al. Establishing China’s national standards of antigen content and neutralizing antibody responses for evaluation of enterovirus 71 (EV71) vaccines. Vaccine. 2011; 29(52):9668–9674. Doi:10. 1016/j.vaccine.2011.10.018
19. Yuan J, Shen L, Wu J, Zou X, Gu J, Chen J, et al. Enterovirus A71 Proteins: Structure and Function. Front Microbiol. 2018; 9:286. Doi: 10.3389/fmicb.2018.00286