Sensitive detection of SARS-CoV-2 seroconversion by flow cytometry reveals the presence of nucleoprotein-reactive antibodies in Covid-19-naive individuals

We have developed a novel multiplexed flow cytometric bead array (C19BA) for the detection of SARS-CoV-2 seroconversion that allows sensitive identification of IgG and IgM antibodies against three immunogenic proteins: the spike receptor-binding domain (RBD), the spike protein subunit 1 (S1) and the nucleoprotein (N) simultaneously. This assay is more sensitive than ELISA, and the combination of three antigens allows for the interrogation of full seroconversion. Importantly, we have detected N-reactive antibodies in COVID-19-negative individuals.


MAIN TEXT
The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) global spread has resulted in an ongoing pandemic 1  The S protein contains the receptor-binding domain (RBD) that binds to its cognate receptor angiotensin converting enzyme 2 (ACE2) expressed by host cells 3,4 . The S protein is comprised of two subunits: S1 and S2. S1 includes the RBD domain and its sequence is specific for SARS-CoV-2, often generating neutralising antibodies in seropositive individuals 5 .
Given their specificity, both RBD and S are considered ideal for serology assays 6-8 , especially in the form of recombinant proteins produced in mammalian cell systems that reflect a physiological glycosylation pattern. In general, ELISAs have an acceptable specificity and sensitivity profile, but have important limitations including cost, time and throughput. ELISAs require optimisation and the use of individual plates for each antigen or antibody to be tested.
Moreover, the antigen is immobilised to the plate, which can hide epitopes or increase the background noise. For these reasons, ELISAs are not well suited to detect low antibody titers and often give undetermined values that are close to the cut-off with difficult interpretation of results.
We have developed a novel flow cytometry assay based on multiplexed microbeads with different intrinsic fluorescence intensities coated with different viral antigens. The coupling was . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 29, 2020. . https://doi.org/10.1101/2020.07.28.20162941 doi: medRxiv preprint based on the interaction between microbeads functionalized with streptavidin and proteins tagged with a unique terminal biotin, which allows for the orientation of the antigen on the surface of the bead. The bead array (C19BA) is incubated with serum samples to allow the binding of anti-SARS-CoV-2 antibodies and then stained with anti-IgG and anti-IgM secondary antibodies labelled with different fluorochromes (Fig. 1a and Supplementary Fig.1). In order to fully assess the specific seroconversion against SARS-CoV-2, we have chosen RBD, S1 and N as target antigens. A previously developed ELISA includes a S1 confirmatory assay after positivity against RBD 6 . In our method, the redundancy of RBD as a sequence included in S1 allows for the confirmation of intra-assay specificity on different microbeads simultaneously.
The N protein was also included in the assay because of its immunogenicity. N is predicted to be less specific for SARS-CoV-2 based on the analysis of the sequence alignment with other coronavirus family members ( Supplementary Fig.2). We reasoned that fully seroconverted individuals would present antibodies against the three chosen antigens. We first tested the ability of this assay to identify recombinant IgG antibodies against RBD and N. As can be seen on Fig. 1b, the microbead array clearly identified the binding of these antibodies. Importantly, the sensitivity of C19BA was superior to ELISA (Fig. 1c) when their performance was compared in serial dilutions of recombinant anti-RBD and anti-N IgG antibodies. C19BA presented a better linear range and identified low antibody concentrations that were not detected by ELISA.
We then applied C19BA to interrogate serum samples from a cohort of 43 individuals who tested positive by PCR for SARS-CoV-2 infection that were obtained at the time of hospital admission (COVID cohort). As a control, sera from 50 individuals collected before the pandemic (2018-2019) were analysed (preCOVID cohort). Both ELISA and C19BA were able to discriminate both cohorts based on the presence of IgG and IgM antibodies against RBD, S1, and N. At this early stage of infection, not all samples from the COVID cohort presented reactivity against viral antigens ( Supplementary Fig.3). Serial dilutions of 10 COVID and 10 preCOVID samples were performed to further compare the sensitivity of C19BA versus ELISA.
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 29, 2020. . https://doi.org/10.1101/2020.07.28.20162941 doi: medRxiv preprint Fig. 2a shows the dilution curves corresponding to the presence of IgG antibodies by these two methods. C19BA was superior to ELISA separating both cohorts and identifying the presence of antibodies at lower concentrations. This was confirmed by plotting the area under the curve (AUC) and determining its statistical significance (Fig. 2b). The titers of IgM antibodies were lower than IgG as measured by both methods (Supplementary Fig. 4), in line with previous studies 9 .
Importantly, our assay identified the presence of N-reactive IgG antibodies in preCOVID samples, although in general at lower titers than in the COVID cohort ( Fig. 2a,b). Fig. 2c shows representative dot plot profiles corresponding to preCOVID and COVID samples. While the reactivity against RBD and S1 was specific for COVID samples, cross-reactivity against N was observed in some preCOVID individuals that contained IgG, but not IgM antibodies. This fact raises concerns about the specificity of the use of N protein for serological assays, given its high homology with N proteins of other coronavirus. In the COVID cohort, several samples that presented full seroconversion against RBD, S1 and N for both IgG and IgM were identified. A minority of COVID samples presented only N-reactive IgG and IgM antibodies, testing negative for RBD or S1. This suggests that these individuals mounted secondary In summary, we have developed a novel multiplexed method with higher sensitivity than traditional serology assays, using a triple combination of antigens that exploits the specificity of the Spike and RBD together with the less-specific N protein for the detection of antibodies against SARS-CoV-2.
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 29, 2020.  is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 29, 2020.

Data analysis
To calculate the sample ODs and gMFIs, the values corresponding to the negative controls were substracted from all samples. For values below 0.11, the OD or gMFI values were set at 0.11 in order to calculate the AUC and generate the graphs. Statistical analyses comparing . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 29, 2020. . https://doi.org/10.1101/2020.07.28.20162941 doi: medRxiv preprint the preCOVID and COVID cohorts were performed with an unpaired two-tailed Student's ttest. Data were analyzed using Prism 8 (GraphPad). . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 29, 2020. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 29, 2020. . https://doi.org/10.1101/2020.07.28.20162941 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted July 29, 2020. . https://doi.org/10.1101/2020.07.28.20162941 doi: medRxiv preprint and N) and mixed and incubated with pre-diluted serum samples together with control beads.
After incubation, microbeads are washed and stained with anti-human IgG and IgM secondary antibodies, washed and acquired on a flow cytometer for downstream analysis. b, Representative dot plots showing unstained beads corresponding to each control or antigen (left), and the specificity of the staining pattern of recombinant anti-RBD IgG (middle) or anti-N IgG (right) antibodies. c, Comparison of the titration of anti-RBD (blue and green) and anti-N (orange) IgG antibodies against recombinant RBD, S1 and N proteins by C19BA (top) and ELISA (bottom). The mean value of two replicates is shown, error bars represent SD.
. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 29, 2020. . https://doi.org/10.1101/2020.07.28.20162941 doi: medRxiv preprint and ELISA (bottom). Statistical analyses were performed using an unpaired two-tailed Student's t-test (*** represent p<0.001). Horizontal lines represent median values. c, Dot plots . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted July 29, 2020. . https://doi.org/10.1101/2020.07.28.20162941 doi: medRxiv preprint