3.5 Quality of vaccines
After cultivation, bacteria were harvested as previously described and samples were analyzed according to standardized parameters (Gonçalves et al., 2014). They all met the established Specification Criteria for Acceptance (data not shown), including the parameter of the percentage of soluble protein, which should be less than 15% (Lu et al., 2010a), indicating that bacterial death was mediated by BPL rather than by starvation. We also analyzed the protein profile of each lot produced using different fermentation strategies by Western blot (Figure 5). A panel of antibodies induced against potential vaccine candidates such as PspA, PdT and PppA, or IL-17-inducing proteins, such as SP0785, SP2070, SP2145 and SP1572, or even anti-PWCV sera were used. We observed that all lots presented similar amounts of specific proteins, with the expected size, independently of which process was used, and all were comparable to our standard, the original vaccine lot produced in 60 L in a batch process (Gonçalves et al., 2014).
In order to verify if different fermentation/downstream strategies would produce effective vaccines, different preparations were evaluated for the induction of IgG antibodies and IL-17A production in immunized animals. Mice immunized with PWCV from lots prepared by any of the fermentation strategies were equally immunogenic, producing high titers of IgG and IL-17A (Figure 6A and B). In fact, PWCV obtained from the perfusion-batch integrated to cell separation induced statistically higher antibodies and IL-17A titers than the standard vaccine, the cGMP lot produced at 60 L (Gonçalves et al., 2014) (Figure 6A and B, 2 and 6). This may be somehow due to a higher quality of the vaccine produced in the integrated process, as 100 µg of total protein was given as vaccine dose to all animals. One hypothesis is that the constant removal of inhibitory metabolites throughout the integrated process led to lower acetate and lactate concentration at the end of the culture, which would be beneficial not only for the cell growth, but also for the production of important antigens. On the other hand, metabolites accumulated in batch and fed-batch processes, probably affecting antigen synthesis besides inhibiting cell growth. Potency of these vaccines was evaluated by challenge in the fatal aspiration model. All lots independently of the fermentation strategy protected mice against challenge (Figure 6C), indicating that the process intensification did not impact the quality of vaccine.
The results of different downstream processes and storage temperatures evaluation were also presented in Figure 6. Again, all lots were equally immunogenic and induced high IgG and IL-17A titers (Figure 6D and E). The lot obtained from the fed-batch process was the only one that induced higher IL-17A titers than the positive control (cGMP lot produced at 60 L) (Figure 6E, 2 and 7). Vaccine produced in batch process with bacteria harvested by microfiltration induced statistically higher IL-17A titers than the lot using the same fermentation strategy and bacteria harvested by centrifugation (Figure 6E, 10 and 11). It is worth noting that, after heating for BPL degradation, the final product obtained using centrifugation for cell separation was not as homogenous as before heating, whereas the same product obtained by microfiltration was homogenous before and after heating for BPL degradation. When centrifugation was applied, small particles or clumps were observed after heating for BPL degradation. These clumps were not present during the washing steps and after cell inactivation period, 30 h at 4 °C. We conclude that this phenomenon was only observed when bacteria were harvested by centrifugation, thus this methodology may compromise the quality of the final product. Nevertheless, all lots protected > 90% of immunized mice (Figure 6F).
4. Conclusion
In our studies, perfusion-batch with cell recycling integrated to the cell separation process was the best promising strategy for the production of PWCV, producing 3-fold higher biomass than batch and fed-batch processes and 4-fold greater number of doses than the previously described batch process, in which cells were harvested at OD 6.0. The perfusion-batch strategy supported the growth of S. pneumoniae RM200 by removing inhibitory metabolites from the culture and supplying nutrients. The integration of the perfusion with cell separation system could be a cost-effective alternative to produce high amounts of PWCV doses, using the same space and equipment as batch or fed-batch cultures, also diminishing the auxiliary time. Therefore, the process intensification achieved in this study has high potential for scale up and bulk production of PWCV, as well as of other whole-cell vaccines, which could be especially important to attend an epidemiological emergency, when a high number of vaccine doses are needed in a very short period of time. Moreover, the potential impact of process intensification was carefully evaluated with respect to the quality of the final product and no quality differences among the vaccines were observed. Therefore, the perfusion-batch cultivation with cell recycling integrated to the cell separation process should be explored for large-scale production of PWCV for human immunization.