Evaluation of S- and M-Proteins Expressed in Escherichia coli and HEK Cells for Serological Detection of Antibodies in Response to SARS-CoV-2 Infections and mRNA-Based Vaccinations

Abstract

This study investigated the IgG and IgA antibody response against recombinant S1 and receptor binding domains (RBD) of the spike (S-) protein and the membrane (M-) protein using a set of 115 serum samples collected from patients infected with SARS-CoV-2 in Germany before April 2021 using protein and peptide ELISA. As S1- and RBD-proteins expressed in Escherichia coli provided poor sensitivities in ELISA, they were replaced by proteins expressed in HEK cells. The RBD-ELISA provided a sensitivity of 90.6% (N = 85) for samples collected from patients with confirmed SARS-CoV-2 infections more than 14 days after symptom onset or a positive PCR test. In population-based controls, the specificity was 97.9% (N = 94). In contrast, the sensitivities were only 41.2% and 72.6% for M- and N-proteins, respectively, while the specificities were 88.5% and 100%, respectively. Considering also 20 samples collected during the first two weeks of symptom onset or PCR confirmation, the sensitivity of RBD- and N-protein ELISA decreased to 82.6% and 72.6%, respectively. The combination of two data sets, i.e., N- and RBD-, N- and M-, or RBD- and M-proteins increased the sensitivity to 85.8%, 77.9%, and 87.8%, respectively. Peptide mapping mostly confirmed epitopes previously reported for S1- and M-proteins, but they were only recognized by a few samples already tested positive in the corresponding protein ELISA indicating that peptide-based assays will not improve the diagnostic sensitivity.

Influence of Mutations and N-Glycosylation Sites in the Receptor-Binding Domain (RBD) and the Membrane Protein of SARS-CoV-2 Variants of Concern on Antibody Binding in ELISA

“…Based on the above observation that glycosylated RBDs expressed in HEK293S cells provide much better serological data than deglycosylated and unglycosylated RBDs, we also reconsidered the M-protein for an IgG-ELISA to see if its expression in HEK293S cells would improve the accuracy of the assay compared to the M-protein expressed in E. coli [25]. To our knowledge, glycosylation sites have not been reported for the SARS-CoV-2 M-protein [4], although in silico methods have suggested asparagine residues 5, 21,41,43,117,121,203, and 216 as potential N-glycosylation sites [42,43].…”

“…The coding sequence of RBD (C-terminal His 6 -Tag, residues 319-541, NCBI accession YP_00972439) was synthesized, codon-optimized for E. coli by GenScript Biotech BV (Leiden, The Netherlands), and cloned into the pET21b(+) vector. The RBD and M-protein were also expressed in E. coli, as described previously [25].…”

“…Subsequent transient transfection and expression of the RBD in HEK293S GnTI− and HEK293T cells were performed as described before [26]. The same procedure was applied for the expression of the M-protein in the stable HEK293S GnTI− cell line, where the medium could be collected throughout expression [25,26]. The proteins were purified by IMAC using a HisTrap TM HP column (GE Healthcare, Solingen, Germany), followed by size-exclusion chromatography (SEC).…”

“…Medium-binding microplates (Greiner Bio-One, 12xF8, PS, F-bottom) were coated in each well with RBDs expressed in-house in HEK293S GnTI− and HEK293T cells, commercial RBDs (BA.1, BA.4), or deglycosylated RBDs (75 ng) in PBS supplemented with sodium chloride (200 mmol/L), or RBDs (150 ng) expressed in E. coli (4 • C, overnight). The RBD protein ELISA was performed as previously described [25]. The same ELISA conditions, except for the addition of 0.05% Tween 20 to the Superblock blocking solution, were used for the deglycosylation study.…”

“…Previously, we noticed that wt-RBD expressed in E. coli showed a much worse sensitivity and specificity than S-RBD in the IgG ELISA [25], which was attributed to the lack of protein glycosylation and maybe improper protein folding. Many studies have shown the importance of glycosylation for infectivity and immune evasion, which might also affect antibody recognition.…”