Accelerating the Manufacture and Scale Up of Virus-Like Particle Vaccines

Developed a VLP vaccine composed of the four SARS-CoV-2 structural proteins: spike (S), envelope (E), nucleocapsid (N), and membrane (M). Determination of the optimum ratio of the four proteins is key to development of a highly productive candidate cell line. The team used a poly-piggyBac method whereby each plasmid encoding for a single structural protein and a constitutively expressed fluorescent reporter was integrated independently of one another. This successfully allowed for random numbers of integration events of each plasmid, creating a completely heterogeneous pool of edited cells. The team developed a fluorescence-activated cell sorting (FACS)-based pipeline to sort for stable SARS-CoV-2 VLP producing cells lines with varying ratios of integrated S, E, N, and M genes. They generated 27 bins fully sorted on the four proteins. The polyclonal pools and 27 partially sorted bins are available for further sorting and evaluation to identify additional vaccine candidates. Evaluated genome copy number of the integrated genes and kinetics of mRNA production in selected bins.

Constructed a cellular-scale mechanistic model useful for inducible production of VLPs in HEK293 and used the model to select which of the vaccine candidate bins were carried forward. Bin selection was based on prediction of VLP productivity from genomic integration numbers and on how informative information from a bin would be for the model itself. The team also developed bioreactor-scale mechanistic model linking the cellular scale model to bioreactor states for both a microbioreactor platform and a 3L benchtop bioreactor.