The onset and spread of EV-71 is rapid among infants and the strain varies according to the geographical location and season . After a quick diagnosis of the strain type, large amounts of vaccine should be produced for use in prophylactic measures. The particular use of inactivated EV-71 particles as a potent vaccine requires high quantities of the virus to be produced, inactivated and purified as rapidly as possible. A literature survey [3, 27] showed that EV-71 is preferably purified by conventional techniques such as ultracentrifugation or precipitation followed by sucrose gradient/caesium chloride centrifugation or size-exclusion chromatography, which are laborious and time consuming processes.
A serum-free cell-based EV-71 vaccine candidate was proposed by Chou et al. , in which a combination of sucrose-gradient ultracentrifugation and/or gel-filtration liquid chromatography purifications was studied. They employed a pilot liquid chromatography system using Sepharose Fast Flow 6 gel. Even though the procedure yielded infectious viral particles of high purity, the overall yield was only 7–10% as determined by a VP2-based quantitative ELISA. They further demonstrated that the purified, formalin-inactivated vaccine candidate was stable and could induce strong virus-neutralizing antibody responses in mice, rats, rabbits, and non-human primates. Production of bulk EV-71 vaccines in roller bottles/bioreactor using a serum-free media and its purification by Sepharose Fast Flow 6 gel was reported by Liu et al.  where the average recovery was about 50%, as is the case with our results with a slight (5%) increase in recovery. A 50% recovery of EV-71 during the purification by gel-filtration chromatography was also reported by Chang et al.  but they suggested that the chromatographic process needs to be improved due to the inconsistent results obtained.
Thus, for the bulk production of different strains of EV-71 as vaccine candidates, a number of factors should be considered. This includes the formulation of a growth medium that supports the production of a high virus titre combined with minimal downstream steps, two factors that are critical for economic viability. During infection of RD cells, we used serum-free DMEM to minimize binding of serum proteins during the purification process.
Ion-exchange columns are extensively used in downstream applications due to their efficiency and cost-effectiveness. Anion exchange columns are widely used as the first step in the purification of viruses and viral vectors because of their enhanced binding of viruses instead of contaminating proteins . This might be the reason that the majority of the contaminating proteins are easily eluted in lower salt concentrations in the earlier steps of purification.
Monoliths are highly interconnected porous structures, made up of different chemistries, currently being used for purification of large biomolecules such as viruses and bacteriophages. They are considered advantageous over other columns as they have properties unaffected by flow, shorter separation times and a high binding capacity . Moreover, the purification protocol is simple and the total time taken can be shortened from several days to a few hours without compromising purity. The method developed from a disk/tube monolithic column can be easily scaled-up to litres for bulk purification of biomolecules . Recently, CIM DEAE monolithic columns have been used for the purification of many viruses and bacteriophages including Tomato mosaic virus [16, 30], filamentous potato virus , Pseudomonas phage LUZ19 , Rubella virus , Staphylococcus phage ISP , viral vectors such as Canine Adenoviral Vectors , and Lentiviral Vectors .
Presence of three bands (VP1, VP2 and VP3) suggested of viral particles and similar results were presented by Liu et al.  during the production and purification of EV-71 viral particles in a serum free bioreactor system. Western blotting with anti-VP2 monoclonal antibody showed two bands (VP0 and VP2), as reported by Liu et al. using the commercial monoclonal antibody mAb 979 . Our method of purification yielded some empty particles when analysed by TEM (Figure 5). The antigenicity of the empty particles may be different from native virus but they only represent a minority of the particles. Hence their presence is unlikely to have a major effect on the overall vaccine efficacy of the purified virus.