Process development and scale-up optimization of the SARS-CoV-2 receptor binding domain-based vaccine candidate, RBD219-N1C1

Abstract

A SARS-CoV-2 RBD219-N1C1 (RBD219-N1C1) recombinant protein antigen formulated on Alhydrogel® has recently been shown to elicit a robust neutralizing antibody response against SARS-CoV-2 pseudovirus in mice. The antigen has been produced under current good manufacturing practices (cGMPs) and is now in clinical testing. Here, we report on process development and scale-up optimization for upstream fermentation and downstream purification of the antigen. This includes production at the 1-L and 5-L scales in the yeast, Pichia pastoris, and the comparison of three different chromatographic purification methods. This culminated in the selection of a process to produce RBD219-N1C1 with a yield of >400 mg per liter of fermentation with >92% purity and >39% target product recovery after purification. In addition, we show the results from analytical studies, including SEC-HPLC, DLS, and an ACE2 receptor binding assay that were performed to characterize the purified proteins to select the best purification process. Finally, we propose an optimized upstream fermentation and downstream purification process that generates quality RBD219-N1C1 protein antigen and is fully scalable at a low cost. KEY POINTS: • Yeast fermentation conditions for a recombinant COVID-19 vaccine were determined. • Three purification protocols for a COVID-19 vaccine antigen were compared. • Reproducibility of a scalable, low-cost process for a COVID-19 vaccine was shown. Graphical abstract.

Keywords: COVID-19; Fermentation; Pichia pastoris; Purification; Spike protein.

Fig. 1

Fermentation (a) and purification (b) flow diagrams. Three purification processes performed are shown in different colors. The color scheme remains consistent throughout all figures. UFDF, ultrafiltration and diafiltration; HIC, hydrophobic interaction chromatography; SEC, size exclusion chromatography; TFF, tangential flow filtration; CEX, cation exchange chromatography; AEX, anion exchange chromatography

Fig. 2

Timepoint SDS-PAGE analysis of pre- and post-induction fermentation samples of the lockdown process (run 5). PI: pre-induction; D1, D2, D3: days 1–3 after induction. The arrow shows RBD219-N1C1 in the fermentation supernatant after induction

Fig. 3

In-process sample comparison from three processes. Yield, step recovery, overall recovery, and purity are shown as an average ± SD calculated from two independent gels that are shown in the table (left) and a representative gel stained with Coomassie blue that is shown (right) from process 1 (a), process 2 (b), and process 3 (c). FS, fermentation supernatant; HIC, hydrophobic interaction chromatography; UFDF, ultrafiltration and diafiltration; SEC, size exclusion chromatography; AEX, anion exchange chromatography; CEX, cation exchange chromatography

Fig. 4

Characterization of purified RBD219-N1C1 proteins from three processes. Purified proteins were analyzed by SDS-PAGE with Coomassie blue stain (a) and Western blot with a monoclonal anti-SARS-CoV-2 spike antibody (b). Size and aggregate evaluation by SEC-HPLC (c). Hydrodynamic radius and size in solution measured by dynamic light scattering (d). Averages ± SD are shown from four independent measurements

Fig. 5

Impurity evaluation of the purified RBD219-N1C1 proteins from three processes. Unpurified (FS) and purified RBD219-N1C1 in reduced SDS-PAGE with Coomassie blue stain (a) and with Western blot using anti-P. pastoris HCP antibody (b). Measured P. pastoris HCP content by quantitative ELISA (c) and endotoxin levels (d) are shown

Fig. 6

Binding ability of the purified RBD219-N1C1 from three processes to a recombinant human ACE2 receptor