Serological Test to Determine Exposure to SARS-CoV-2: ELISA Based on the Receptor-Binding Domain of the Spike Protein (S-RBDN318-V510) Expressed in Escherichia coli

Abstract

Massive worldwide serological testing for SARS-CoV-2 is needed to determine the extent of virus exposure in a particular region, the ratio of symptomatic to asymptomatic infected persons, and the duration and extent of immunity after infection. To achieve this, the development and production of reliable and cost-effective SARS-CoV-2 antigens is critical. We report the bacterial production of the peptide S-RBD_N318-V510, which contains the receptor-binding domain of the SARS-CoV-2 spike protein (region of 193 amino acid residues from asparagine-318 to valine-510) of the SARS-CoV-2 spike protein. We purified this peptide using a straightforward approach involving bacterial lysis, his-tag-mediated affinity chromatography, and imidazole-assisted refolding. The antigen performances of S-RBD_N318-V510 and a commercial full-length spike protein were compared in ELISAs. In direct ELISAs, where the antigen was directly bound to the ELISA surface, both antigens discriminated sera from non-exposed and exposed individuals. However, the discriminating resolution was better in ELISAs that used the full-spike antigen than the S-RBD_N318-V510. Attachment of the antigens to the ELISA surface using a layer of anti-histidine antibodies gave equivalent resolution for both S-RBD_N318-V510 and the full-length spike protein. Results demonstrate that ELISA-functional SARS-CoV-2 antigens can be produced in bacterial cultures, and that S-RBD_N318-V510 may represent a cost-effective alternative to the use of structurally more complex antigens in serological COVID-19 testing.

Keywords: COVID-19; ELISA; Escherichia coli; SARS-CoV-2; antigen; receptor binding domain; serological testing; spike.

Figure 1

Expression of S-RBD_N318-V510 in Escherichia coli. (a) Schematic representation of the sequence used to produce the MFH8-RBD_Spike-HisTag protein (S-RBD_N318-V510). This expression cassette was inserted into the (b) pFH-RBD SARS-CoV-2 plasmid for expression in E. coli. (c) Molecular 3D structure of the S-RBD_N318-V510 protein, as predicted by molecular structure simulations.

Figure 2

Contrast of two versions of ELISAs for identification of anti-spike SARS-CoV-2 antibodies. Comparison of a commercial full-length spike protein from SARS-CoV-2 (red circles) and the S-RBD_N318-V510 protein (blue circles) produced in Escherichia coli as antigens in ELISA experiments performed to identify anti-spike SARS-CoV-2 antibodies. The antigen was (a) directly bound to the surface of 96-well plates, or (b) bound through a layer of anti-histidine antibodies (red Ys). In both cases, the specific attachment of anti-spike SARS-CoV-2 antibodies (blue Ys) was revealed by binding of anti-human heavy-chain antibodies functionalized with horseradish peroxidase (yellow Ys). Comparison of absorbance readings for (c) direct ELISAs or (d) anti-histidine-mediated ELISAs. A commercial full-length spike (red bars) or the S-RBD_N318-V510 protein (blue bars) was used as the antigen. Commercially available anti-spike SARS-CoV-2 antibodies were used as reactants. Negative controls consisted of the S-RBD_N318-V510 protein either deficiently folded in glycerol and phosphate-buffered saline (PBS) (green) or properly folded S-RBD_N318-V510 antigen in a highly diluted solution (1:20,000) of commercial anti-spike S1 antibodies (yellow). A blank control, where only PBS is added, was also included (absorbance reading ~0.0). (e) The specific binding between S-RBD_N318-V510 and the angiotensin-converting enzyme 2 (ACE2) was determined in experiments with Caco-2 cell monolayers. (f) Absorbance reading in ELISA experiments where Caco-2 cell monolayers were incubated with different solutions containing commercial spike protein (red bars) or varying concentrations of the S-RBD_N318-V510 protein (blue bars). The retention of the antigen on the surface of the Caco-2 cells was evidenced by the subsequent addition of commercial rabbit anti-spike pAb and donkey anti-rabbit-HRP pAb solutions, and the detection of the HRP (as described in Section 2, Materials and Methods). Negative controls containing albumin (yellow bars), recombinant glycoprotein (GP) from the Ebola virus (gray bars), and only anti-spike rabbit pAb (pink bars) were included. Letters indicate statistically different groups (p < 0.05; Tukey’s test).

Figure 3

Binding of antibodies from human sera to S-RBD_N318-V510 using direct and anti-histidine-mediated ELISAs. (a) Binding of antibodies from human sera, measured as absorbance readings, in direct ELISA experiments that used a commercial full-length spike (red bars) or the S-RBD_N318-V510 protein (blue bars) as antigens. Serum samples were obtained from non-exposed volunteers (FLU X; collected during pandemic Influenza A/H1N1/2009) and from volunteers possibly exposed to SARS-CoV-2 (COVID X; collected during pandemic COVID-19). NEG bars indicate the average absorbance reading exhibited by serum samples from non-exposed volunteers. (b) Binding of antibodies from human sera, measured as absorbance readings, in anti-histidine mediated ELISA experiments that used a commercial full-length spike (red bars) or the S-RBD_N318-V510 protein (blue bars) as antigens. Serum samples were obtained from non-exposed volunteers (FLU X; collected during pandemic Influenza A/H1N1/2009) and from volunteers possibly exposed to SARS-CoV-2 (COVID X; collected during pandemic COVID-19). NEG bars indicate the average absorbance reading exhibited by serum samples from non-exposed volunteers. (c) Graphic analysis of the correlation between titers obtained in anti-histidine-mediated ELISAs that used the commercial full-length spike (red bars) or the S-RBD_N318-V510 protein (blue bars) as antigens. All experiments were conducted using 1:100 serum dilutions. (d,e) Binding of S-RBDN_318-V510 serum samples donated by convalescent patients confirmed as COVID-19 (+) by RT-qPCR. Normalized absorbance readings related to the binding affinity of the S-RBD_N318-V510 protein to human sera antibodies from COVID-19 convalescent patients for (d) direct, and (e) anti-histidine mediated ELISAs. All absorbance readings were normalized by the average absorbance reading exhibited by samples from non-exposed individuals. Normalized readings higher than the threshold value of 1.11 (for direct ELISAs) or 1.12 (for anti-histidine-mediated ELISAs) are indicated in blue. Absorbance readings from blanks (phosphate-buffered saline only) are indicated in yellow as a reference. Experiments were conducted using 1:100 serum dilutions. (f) Absorbance readings related to the binding affinity of the S-RBD_N318-V510 protein to sera antibodies from mice subcutaneously inoculated with S-RBD_N318-V510 14 days before (blue bars) and to control animals (injected only with vehicle (i.e., HEPES; gray bar). Absorbance readings related to the complete set of mice before injection (at day 0; white bar). Letters indicate statistically different groups (p < 0.05; Tukey’s test).