Analysis of Serological Biomarkers of SARS-CoV-2 Infection in Convalescent Samples From Severe, Moderate and Mild COVID-19 Cases

Abstract

Precision monitoring of antibody responses during the COVID-19 pandemic is increasingly important during large scale vaccine rollout and rise in prevalence of Severe Acute Respiratory Syndrome-related Coronavirus-2 (SARS-CoV-2) variants of concern (VOC). Equally important is defining Correlates of Protection (CoP) for SARS-CoV-2 infection and COVID-19 disease. Data from epidemiological studies and vaccine trials identified virus neutralising antibodies (Nab) and SARS-CoV-2 antigen-specific (notably RBD and S) binding antibodies as candidate CoP. In this study, we used the World Health Organisation (WHO) international standard to benchmark neutralising antibody responses and a large panel of binding antibody assays to compare convalescent sera obtained from: a) COVID-19 patients; b) SARS-CoV-2 seropositive healthcare workers (HCW) and c) seronegative HCW. The ultimate aim of this study is to identify biomarkers of humoral immunity that could be used to differentiate severe from mild or asymptomatic SARS-CoV-2 infections. Some of these biomarkers could be used to define CoP in further serological studies using samples from vaccination breakthrough and/or re-infection cases. Whenever suitable, the antibody levels of the samples studied were expressed in International Units (IU) for virus neutralisation assays or in Binding Antibody Units (BAU) for ELISA tests. In this work we used commercial and non-commercial antibody binding assays; a lateral flow test for detection of SARS-CoV-2-specific IgG/IgM; a high throughput multiplexed particle flow cytometry assay for SARS-CoV-2 Spike (S), Nucleocapsid (N) and Receptor Binding Domain (RBD) proteins); a multiplex antigen semi-automated immuno-blotting assay measuring IgM, IgA and IgG; a pseudotyped microneutralisation test (pMN) and an electroporation-dependent neutralisation assay (EDNA). Our results indicate that overall, severe COVID-19 patients showed statistically significantly higher levels of SARS-CoV-2-specific neutralising antibodies (average 1029 IU/ml) than those observed in seropositive HCW with mild or asymptomatic infections (379 IU/ml) and that clinical severity scoring, based on WHO guidelines was tightly correlated with neutralisation and RBD/S antibodies. In addition, there was a positive correlation between severity, N-antibody assays and intracellular virus neutralisation.

Keywords: Antibodies; COVID-19; COVID-19 immune response; Correlates of Protection; SARS-CoV-2; Serological biomarkers; WHO International Standard.

Figure 1

Comparison of internal calibrants and International Standard for neutralization and binding assays. The IC₅₀ from these curves were used to calculate the (A) International units (IU); (B) RBD-specific Binding Antibody Units (BAU); and (C) N-specific Binding Antibody Units (BAU). Calibrants were run four times at each dilution for pMN and twice for ELISA tests.

Figure 2

(A) Correlation plot showing pairwise Spearman’s rank correlation between assays. Darker and larger points indicate stronger correlations. Blue indicates positive and red indicates negative correlations. Assays are ordered by hierarchical clustering so that assays with similar relationships are together. (B) Correlation plot showing pairwise Spearman’s rank correlation between clinical severity, pMN and S-specific antibody assays. Darker and larger points indicate stronger correlations. Blue indicates positive and red indicates negative correlations. Assays are ordered by hierarchical clustering so that assays with similar relationships are together. (C) Correlation plot showing pairwise Spearman’s rank correlation between clinical severity, intracellular neutralization and N-specific antibody assays. Darker and larger points indicate stronger correlations. Blue indicates positive and red indicates negative correlations. Assays are ordered by hierarchical clustering so that assays with similar relationships are together.

Figure 3

pMN virus neutralisation in International Units by cohort and COVID-19 severity. (A) Boxplot showing the difference between cohorts. (B) Scatterplot showing neutralisation against disease severity. The dotted line shows the 95% upper CI calculated from pre-pandemic sera, 5.9 International Units.

Figure 4

IgG, IgA and IgM responses against Spike, RBD, S1, S2 and N antigens of SARS-CoV-2. The image displays the Median Chemiluminescence Intensities of antigen-specific IgG (top), IgA (middle) and IgM (bottom) of seronegative HCWs (left), seropositive HCWs (middle) and COVID-19 patients (right). A panel of pre-pandemic sera was used to calculate the negative cut-off value for each antigen (mean + 2STD) and then subtracted from the chemiluminescent signal of each sample.

Figure 5

Representative electropherograms of IgG, IgA and IgM antibody responses to Spike, RBD, S1, S2 and N antigens of SARS-CoV-2 in Health Care Workers and Patients. Thefigure shows antigen-specific IgG (A, D, G), IgA (B, E, H) and IgM (C, F, I) antibody reactivities of seropositive HCWs (left panel), serongative HCWs (middle panel), and patients (right panel).

Figure 6

Intracellular neutralisation data from EDNA assay. The results are expressed in genome copies relative to 18S and percentage normalised to PBS. Panel (A) depicts median values of EDNA results of the three cohorts, expressed as Fold neutralisation relative to PBS; Panels (B–D) correspond to individual EDNA results of patients, seropositive HCW and seronegative HCW respectively.