Materials & Methods

Cell line and culture media

The suspension adapted Vero cell line was provided by the National Research Council (NRC) of Canada, Montreal, Canada, and its adaptation process has been described previously (Shen et al., 2019). For routine passaging, the cells were harvested by centrifugation for 5 minutes at 500 g and the cell pellet was resuspended in fresh medium to a seeding cell density of 2.5–5 × 105 cells/mL in 125 mL polycarbonate shake flasks (TriForest Enterprises, USA) and maintained at 37 °C, 135 rpm and 5% CO2 in an humified Multitron orbital shaker (Infors HT, Switzerland). The cells were cultivated in 20 mL working volume of either IHM03 medium, provided by the NRC, or in MDXK medium (Xell AG, Germany), supplemented with 4 mM GlutaMAX (Thermo Fisher Scientific, USA).

Viruses

Origin and viral seed stock amplification of rVSV-ZEBOV and rVSV-B6-A74Env(PN6)/SIVtm (hereafter referred to as rVSV-HIV) have been described previously (Kiesslich et al., 2020) (Mangion et al., 2020).
The recombinant COVID-19 vaccine candidate rVSVInd-msp -SF-Gtc was constructed as follows: Codon-optimized full-length spike protein gene of SARS-CoV-2 (GenBank: JX869059.2) was purchased from Genscript USA Inc (Pscataway, NJ, USA), cloned into an avirulent Indiana serotype of vesicular stomatitis virus [VSVInd(GML)] as has been described previously (Kim et al., 2015). The honeybee melittin signal peptide (msp ) was inserted at the NH2-terminus and the transmembrane domain and cytoplasmic tail (Gtc) of the S protein was substituted by the Gtc of VSVInd G protein at the COOH-terminus. In addition, the VSV intergenic sequences were added in front of the spike protein gene in order to provide the transcription termination signal, polyadenylation signal, and the transcription re-initiation signal. The modified SARS-CoV-2 spike protein gene was inserted into the G and L gene junction of the VSVInd(GML) at Pme I and Mlu I sites. Recombinant rVSVInd-msp -SF-Gtc virus was recovered by VSV reverse genetics as has been described previously (Kim et al., 2015). The recombinant virus was purified by three consecutive plaque picking, and a stock virus was prepared by infecting BHK21 cells.

Shake flask virus studies

For virus infection studies in shake flasks, suspension adapted Vero cells were harvested and seeded at the indicated cell density in fresh medium. Vero cells were infected with rVSV at the indicated MOI and the temperature was shifted to either 34 °C (rVSV-ZEBOV, rVSV-HIV) or 31 °C (rVSVInd-msp -SF-Gtc ). Samples for virus titration were centrifuged for 5 minutes at 1200 × g to remove cellular debris, aliquoted and stored at −80 °C.

Bioreactor cultures

Bioreactor cultures were performed in a 1 L bioreactor (Applikon Biotechnology, The Netherlands) equipped with a marine impeller, pH sensor, temperature sensor, and dissolved oxygen (DO) concentration sensor. MDXK medium was supplemented 4 mM L‑glutamine (GE Healthcare, USA) instead of GlutaMAX to enable monitoring of L‑glutamine consumption. Vero cell seed cultures were grown in progressively larger polycarbonate shake flasks (TriForest Enterprises, USA), harvested by centrifugation and resuspended in fresh medium before inoculation. The bioreactor was inoculated at a cell density of 2.5 × 105 cells/mL in 850 mL working volume. The culture was agitated at 100 rpm and kept at 37 °C. The DO concentration was kept at 50 % air-saturation by continuous surface aeration of 5 mL/min air and injection of pure oxygen through the sparger when required. The pH was set to 7.2 and regulated by injection of CO2 into the headspace or addition of NaHCO3 (90 g/L) (Sigma, USA). Samples were taken once or twice daily, depending on the progress of the culture, to subsequently determine viable cell density, metabolite concentration and virus titer. Samples for metabolite analysis and virus titration were centrifuged for 5 minutes at 1200 × g to remove cellular debris, aliquoted and stored at −80 °C.
For virus production, Vero cells were infected with rVSV at an MOI of 0.01 once the targeted cell density was reached and the temperature was shifted to 34 °C or 31 °C, respectively, during the virus production phase. The glucose concentration was estimated once daily and if required adjusted to 2 g/L by feeding glucose (Sigma) concentrate (180 g/L). In addition, L-glutamine was maintained at a minimum concentration of 2 mM.

Analytical Methods

Vero cell concentration and viability were determined via the Vi-CELL XR cell counter (Beckman Coulter, USA). The Median Tissue Culture Infectious Dose (TCID50) assay and digital PCR assay used in this work to quantify the infectious titer and the number of viral genomes, respectively, has been described previously (Gélinas et al., 2020; Kiesslich et al., 2020). For rVSVInd-msp -SF-Gtc , the TCID50 plates were incubated at 31 °C, due to the temperature sensitivity of this construct.

Metabolite analysis

During cell culture cultivations, the glucose concentration was estimated using the D-Fructose/D-Glucose Assay Kit (Megazyme, Ireland). More extensive metabolite analysis was performed offline via Bioprofile 400 (Nova Biomedical, USA) from samples that were stored at −80 °C.