Glycoprofile Comparison of the SARS-CoV-2 Spike Proteins Expressed in CHO and HEK Cell Lines

Coronavirus SARS-CoV-2 spike protein remains a key focus of research due to a continued need for diagnostic and therapeutic tools to monitor and respond to new variants. Glycosylation of the spike protein is critical for the protein’s functions in viral attachment and host cell entry. For scalable and cost-effective production of the spike protein, expression system-driven divergence in glycosylation patterns on recombinant spike proteins needs to be fully understood. This study assessed the N-glycosylation profiles of a full-length trimeric spike protein expressed in either Human Embryonic Kidney (HEK Expi293F) or Chinese Hamster Ovary (CHO-S) cells. Glycopeptide analysis was performed using a tandem mass spectrometry workflow and BioPharma \({\hbox {Finder}}^{\textrm{TM}}\) ${\text{Finder}}^{\text{TM}}$ incorporating HEK and CHO glycan databases for protein characterisation. The results outline important differences in the variety and types of N-glycan generated by the two cell lines across the 22 known N-glycosylation sites of the spike protein. A notable increase in terminal sialylation, as well as the presence of the potentially immunogenic N-glycolylneuraminic acid at a functionally key N-glycosylation site, was observed in the CHO-S derived spike protein. With the potential for the relatively vast and more complex CHO glycan repertoire (182 glycans relative to 39 human glycans) to produce functional implications with CHO-S expressed spike protein, this study adds valuable knowledge to aid Quality by Design approaches and enable Multi Attribute Monitoring of specific N-glycosylation sites for proteoform analyses. This can further inform antigen development with future variants in order to devise updated diagnostic tests and therapeutic vaccine designs.