Table of Contents:
Influenza-A-viruses cause highly contagious respiratory infections in different species against which vaccination provides the most effective protection. Presently the majority of influenza vaccines are still produced in embryonated chicken eggs but the development of cell culture-based systems is increasing (e.g. MDCK- and Vero-cells). The producible antigen yield that is highly dependent on the seed virus growth capacities is the major limiting factor for sufficient vaccine supply. In particular for emerging pandemic viruses the recoverable amounts of antigen are often very low (such as those of the H1N1 pandemic of 2009 and potentially pandemic H5N1 isolates).
The focus of this work was to generate higher antigen yields by using an alternative acceptor strain for the generation of the reassortant seed viruses to be used for vaccine production. Usually the human H1N1-subtype isolate A/Puerto Rico/8/1934 (PR8) is the recommended high-yield acceptor strain for the integration of the surface antigen proteins of circulating viruses. In this study the suitability of the non-pathogenic, attenuated variant FPVmut of the avian influenza strain A/chicken/Rostock/1934 (H7N1) to serve as an alternative background strain was explored. For this purpose recombinant reassortant viruses were generated in both acceptor strain backgrounds that contained the antigenic surface proteins hemagglutinin and neuraminidase (HA and NA) of selected seasonal and (potentially) pandemic donor viruses (subtypes H1N1, H1N1pdm, H3N2, H5N1, H9N2). Further, due to its established function in promoting viral particle formation the potential influence of the incorporation of the HA- and NA-homologous M1 protein was analyzed.
Both acceptor strains, FPVmut and PR8, produced almost identical viral yields when propagated in embryonated chicken eggs, MDCK- and Vero-cells. However, we found that the onset of virus production and growth to high titers in MDCK cells was faster for FPVmut than for PR8. Both acceptor strains were equally suitable for the generation of recombinant reassortant viruses applying a “reverse genetics” approach.
For reassortants carrying HA and NA from an avian donor strain a more rapid and pronounced cell-to-cell spread was detectable in the FPVmut acceptor background. No significant differences were observed in the NA yields or the virus morphology depending on the background strain or the genetic composition.
When using a potentially pandemic avian H5N1 strain as HA and NA donor a significant increase in the antigen yield was obtained in chicken eggs with FPVmut-reassortants. Both, higher average and maximum HA-titers and higher absolute amounts of HA-antigen were reached in the allantoic fluid of infected eggs in the FPVmut-background. This finding was even more pronounced in the Vero cell production system where the HA-antigen yields achievable with the FPVmut-reassortants were threefold higher than those of the PR8-reassortants. For the H9N2 isolate the incorporation of the surface proteins homologous M gene segment within the FPVmut genetic background resulted in an 1,5-fold increase in antigen yield. Similarly, for the H1N1pdm09 donor virus the highest individual values of the HA yields were measured in the PR8 background with the homologous M gene segment.
Taken together, these data show that depending on the donor virus the use of the alternative acceptor strain FPVmut and/or the integration of the homologous M1 protein are effective approaches to optimize the composition of influenza vaccine seed viruses to achieve higher HA antigen yields. Therefore, both approaches could be of good value for the production of future seasonal and pandemic vaccines.