SARS-CoV-2 elicits robust adaptive immune responses regardless of disease severity

Abstract

Background: The SARS-CoV-2 pandemic currently prevails worldwide. To understand the immunological signature of SARS-CoV-2 infections and aid the search and evaluation of new treatment modalities and vaccines, comprehensive characterization of adaptive immune responses towards SARS-CoV-2 is needed.

Methods: We included 203 recovered SARS-CoV-2 infected patients in Denmark between April 3^rd and July 9^th 2020, at least 14 days after COVID-19 symptom recovery. The participants had experienced a range of disease severities from asymptomatic to severe. We collected plasma, serum and PBMC's for analysis of SARS-CoV-2 specific antibody response by Meso Scale analysis including other coronavirus strains, ACE2 competition, IgA ELISA, pseudovirus neutralization capacity, and dextramer flow cytometry analysis of CD8⁺ T cells. The immunological outcomes were compared amongst severity groups within the cohort, and 10 pre-pandemic SARS-CoV-2 negative controls.

Findings: We report broad serological profiles within the cohort, detecting antibody binding to other human coronaviruses. 202(>99%) participants had SARS-CoV-2 specific antibodies, with SARS-CoV-2 neutralization and spike-ACE2 receptor interaction blocking observed in 193(95%) individuals. A significant positive correlation (r=0.7804) between spike-ACE2 blocking antibody titers and neutralization potency was observed. Further, SARS-CoV-2 specific CD8⁺ T-cell responses were clear and quantifiable in 95 of 106(90%) HLA-A2⁺ individuals.

Interpretation: The viral surface spike protein was identified as the dominant target for both neutralizing antibodies and CD8⁺ T-cell responses. Overall, the majority of patients had robust adaptive immune responses, regardless of their disease severity.

Funding: This study was supported by the Danish Ministry for Research and Education (grant# 0238-00001B) and The Danish Innovation Fund (grant# 0208-00018B).

Keywords: Adaptive; Antibody; Asymptomatic; CD8(+) T-cell; COVID-19; Corona; Immune response; SARS-CoV-2; Severe; Virus.

Fig. 1

Extensive IgG and IgA presence with multiple SARS-CoV-2 antigens. a+b) Serum IgG levels for all individuals (n=203) and 10 pre-pandemic healthy controls. IgG was detected against SARS-CoV-2 Spike, RBD (receptor binding domain), NTD (N-terminal domain), nucleocapsid and non-SARS-CoV-2 spike proteins of other corona virus. Data are blank-corrected electro chemiluminescent signal measured by MSD multiplex serology assays. c) Serum IgA levels for all individuals (n=203) and eight pre-pandemic healthy controls, measured by ELISA. IgA is shown as a ratio against a standard calibrator. d+e) Distribution of IgG volumes between each disease severity group, for both SARS-CoV-2 spike (d) * = p=0.0416 and nucleocapsid (e) * = p=0.0271. Data are blank-corrected electro chemiluminescent signal measured by MSD multiplex serology assays. Scatter plots with individual data points are shown with median (wide line) and interquartile range (narrow lines). Statistical comparison between groups were done by Mann-Whitney test. n.s = not significant, **** = p<0.0001.

Fig. 2

SARS-CoV-2 neutralization capacity correlates with disease severity. a) Representative neutralization curves for control ID308, and individuals ID54, ID194, and ID203, quantified as eGFP⁺ cells by flow cytometry. Control plasma was unable to neutralize below a 50% infection rate, where SARS-CoV-2 recovered patients accomplish 100% neutralization at the lowest plasma dilution. X-axis shows the log10 transformed patient plasma dilution, from 1:25 – 1:1,953,125. Error bars represent mean and s.e.m. of duplicate determinations. Three-parameter non-linear fit is plotted. b) IC50 values calculated from neutralization curves, graphed from lowest (left) – highest (right) within the cohort. Error bars show 95% confidence interval. Nine individuals unable to neutralize 100% are represented with the value zero on the y-axis far left, n = 202. c) Distribution of IC50 values between disease severities. *** = p=0.0008. Scatter plot with individual data points shown with median (wide line) and interquartile range (narrow lines). Statistical comparison were by Mann-Whitney test. **** = p < 0.0001, n = 193.

Fig. 3

SARS-CoV-2 antibody quantification by ACE2 competition assay. a) Schematic drawing of the MSD ACE2 competition assay. Spike-specific serum antibodies bind to their respective epitopes, blocking SULFO-Tag conjugated ACE2. Antibody concentration in ng/ml is calculated based on internal standard antibody blocking ACE2 binding. b) Serum ACE2 blocking antibody levels detected against SARS-CoV-2 Spike and RBD, and SARS-CoV-1 spike proteins. *** = p=0.001. Scatter plot with individual data points shown with median (wide line) and interquartile range (narrow lines). Statistical comparison by Mann-Whitney test. **** = p<0.0001, n = 203. c) Distribution of SARS-CoV-2 spike specific ACE2 blocking antibodies between disease severity groups. Scatter plot with individual data points shown with median (wide line) and interquartile range (narrow lines). Statistical comparison by Mann-Whitney U test. **** = p < 0.0001, n = 203. d) Correlation analysis of pseudotype virus neutralization IC50 values and the quantity of SARS-CoV-2 spike specific ACE2 blocking antibodies. Correlation by Spearman's rank coefficient, p < 0.0001. n = 193.

Fig. 4

The breadth of immunological response shifts in conjunction with neutralization capacity. Presentation of all IC50 values listed from lowest (left) to highest (right) with a heatmap representing the individuals corresponding relative IgG levels and ACE2 blocking antibody quantities collected through MSD analysis. The normalization of variables within each measured immunological parameter was performed by assigning the highest values to one (bright yellow) and the lowest value to zero (dark blue). n=202.

Fig. 5

Characterization of CD8+ T-cell responses towards SARS- CoV-2 in HLA-A2+ individuals. a) Overview of HLA-A2⁺ epitope location within the SARS-CoV-2 proteins. b). Epitope sequence and individual dextramer signal gating strategy on CD8⁺ T cells, with the percentage of recognition within the cohort shown for each. Full gating strategy is displayed in S1 Fig 2. c) The frequency of SARS-CoV-2 responsive CD8⁺ T-cells for each epitope. Scatter plot with individual data points shown with median (wide line) and interquartile range (narrow lines). n = 106 d) Breadth of CD8⁺ T-cell responses shown as the cumulative number of CD8⁺ T-cell epitopes targeted by patients. Percentage equivalents of patient numbers are indicated on top of the bars for each cumulative group. n = 106 e) Distribution of the cumulative CD8⁺ T-cell responses in HLA-A2⁺ individuals, between the disease severity groups. Error bars show median (wide line) and interquartile range (narrow lines). n=106. 10% of individuals had no detectable CD8⁺ T-cell epitope response, and are not shown on the graph but were included in statistical tests. Statistical comparison by Mann-Whitney test. n.s. = p > 0.05.