Antibody Response against SARS-CoV-2 and Seasonal Coronaviruses in Nonhospitalized COVID-19 Patients

Abstract

The majority of infections with SARS-CoV-2 are asymptomatic or mild without the necessity of hospitalization. It is of importance to reveal if these patients develop an antibody response against SARS-CoV-2 and to define which antibodies confer virus neutralization. We conducted a comprehensive serological survey of 49 patients with a mild course of disease and quantified neutralizing antibody responses against a clinical SARS-CoV-2 isolate employing human cells as targets. Four patients (8%), even though symptomatic, did not develop antibodies against SARS-CoV-2, and two other patients (4%) were positive in only one of the six serological assays employed. For the remaining 88%, antibody response against the S protein correlated with serum neutralization whereas antibodies against the nucleocapsid were poor predictors of virus neutralization. None of the sera enhanced infection of human cells with SARS-CoV-2 at any dilution, arguing against antibody-dependent enhancement of infection in our system. Regarding neutralization, only six patients (12%) could be classified as high neutralizers. Furthermore, sera from several individuals with fairly high antibody levels had only poor neutralizing activity. In addition, employing a novel serological Western blot system to characterize antibody responses against seasonal coronaviruses, we found that antibodies against the seasonal coronavirus 229E might contribute to SARS-CoV-2 neutralization. Altogether, we show that there is a wide breadth of antibody responses against SARS-CoV-2 in patients that differentially correlate with virus neutralization. This highlights the difficulty to define reliable surrogate markers for immunity against SARS-CoV-2.IMPORTANCE There is strong interest in the nature of the neutralizing antibody response against SARS-CoV-2 in infected individuals. For vaccine development, it is especially important which antibodies confer protection against SARS-CoV-2, if there is a phenomenon called antibody-dependent enhancement (ADE) of infection, and if there is cross-protection by antibodies directed against seasonal coronaviruses. We addressed these questions and found in accordance with other studies that neutralization is mediated mainly by antibodies directed against the spike protein of SARS-CoV-2 in general and the receptor binding site in particular. In our test system, utilizing human cells for infection experiments, we did not detect ADE. However, using a novel diagnostic test we found that antibodies against the coronavirus 229E might be involved in cross-protection to SARS-CoV-2.

Keywords: COVID-19; SARS-CoV-2; antibody-dependent enhancement (ADE); neutralizing antibodies; seasonal coronaviruses.

FIG 1

Neutralization of SARS-CoV-2 by sera of COVID-19 convalescent patients. (a) Experimental layout of the two neutralization assays employed using the clinical isolate (SARS-CoV-2_Tü1) and the green fluorescent virus (SARS-CoV-2_mNG). (b) Primary data showing results of both neutralization assays using one patient serum as an example (S28). In the upper row of each set, the total amount of cells for each well of the 2-fold serial dilution of sera is shown, as DAPI⁺/Hoechst⁺, respectively. In the lower row of each set, infected cells are visualized, indicated as Alexa 594⁺/mNG⁺ cells, respectively. (c) Neutralization curves of five representative sera measured by both assays. The graphs show the nonlinear regression fitting calculated for five patients who displayed different neutralization capacity: no, poor, low, medium, and high neutralization. The VNT₅₀ for each patient is shown next to each curve. (d) Correlation analysis of VNT₅₀ measured by both assays (n = 49). Correlation is calculated as Pearson’s r. (e) Percentage of patients classified according to the VNT₅₀ using SARS-CoV-2_Tü1. The titers used to classify the sera are shown below the columns: <20, 20 to 200, 201 to 400, 401 to 1,000, and 1,000 to 2,560. Above the columns is shown the percentage of sera that correspond to each category.

FIG 2

Association of patient characteristics with serum VNT₅₀. The VNT₅₀ of each patient serum was associated with the individual date of the positive SARS-CoV-2 qRT-PCR diagnostic test to blood sampling (a), the age of the patient (b), the gender (c), and the number of symptoms reported (d). Statistical analyses were done with an unpaired two-tailed Student t test; correlations are calculated as Pearson’s r. See detailed patient characteristics in Table S1.

FIG 3

Correlation of serological parameters with serum VNT₅₀. The VNT₅₀ of each patient serum was correlated with the value of SARS-CoV-2 S-specific IgGs measured by the Euroimmun ELISA (a), the relative quantitative value of SARS-CoV-2 S-RBD-specific IgGs measured by the Mediagnost ELISA (b), the relative quantitative value of SARS-CoV-2 S-RBD-specific IgAs measured by the Mediagnost ELISA (c), the relative quantitative value of SARS-CoV-2 S-RBD-specific IgMs measured by the Mediagnost ELISA (d), and the relative quantitative value of SARS-CoV-2 NC-specific IgGs measured by the Roche ECLIA (e). Dotted lines indicate the assay thresholds. Correlations are calculated as Pearson’s r.

FIG 4

Correlation of antibodies against seasonal coronaviruses with serum VNT₅₀. The VNT₅₀ of each patient serum was correlated with the relative quantitative value of a throughput diagnostic Western blot detection system measuring CoV-specific IgG against SARS-CoV-2 (a), CoV-229E (b), CoV-OC43 (c), or CoV-NL63 (d). Dotted lines indicate the respective assay thresholds defined as positive. Correlations are calculated as Pearson’s r.