Immune Dysregulation in Acute SARS-CoV-2 Infection

Abstract

Introduction: Neutralizing antibodies have been shown to develop rapidly following SARS-CoV-2 infection, specifically against spike (S) protein, where cytokine release and production is understood to drive the humoral immune response during acute infection. Thus, we evaluated the quantity and function of antibodies across disease severities and analyzed the associated inflammatory and coagulation pathways to identify acute markers that correlate with antibody response following infection.

Methods: Blood samples were collected from patients at time of diagnostic SARS-CoV-2 PCR testing between March 2020-November 2020. Plasma samples were analyzed using the MesoScale Discovery (MSD) Platform using the COVID-19 Serology Kit and U-Plex 8 analyte multiplex plate to measure anti-alpha and beta coronavirus antibody concentration and ACE2 blocking function, as well as plasma cytokines.

Results: A total of 230 (181 unique patients) samples were analyzed across the 5 COVID-19 disease severities. We found that antibody quantity directly correlated with functional ability to block virus binding to membrane-bound ACE2, where a lower SARS-CoV-2 anti-spike/anti-RBD response corresponded with a lower antibody blocking potential compared to higher antibody response (anti-S1 r = 0.884, P < 0.001; anti-RBD r = 0.75, P < 0.001). Across all the soluble proinflammatory markers we examined, ICAM, IL-1β, IL-4, IL-6, TNFα, and Syndecan showed a statistically significant positive correlation between cytokine or epithelial marker and antibody quantity regardless of COVID-19 disease severity. Analysis of autoantibodies against type 1 interferon was not shown to be statistically significant between disease severity groups.

Conclusion: Previous studies have shown that proinflammatory markers, including IL-6, IL-8, IL-1β, and TNFα, are significant predictors of COVID-19 disease severity, regardless of demographics or comorbidities. Our study demonstrated that not only are these proinflammatory markers, as well as IL-4, ICAM, and Syndecan, correlative of disease severity, they are also correlative of antibody quantity and quality following SARS-CoV-2 exposure.

Keywords: SARS-CoV-2; adaptive immunity; autoantibodies; cytokines; spike antibody.

Figure 1.

Quantitative antibody response by disease severity. Dot plot representation of SARS-CoV-2 antibody concentration against spike (A), RBD (B), nucleocapsid (C), and N-terminal Domain of the nucleoprotein (NTD, D) across disease cohorts. Dot plot representation of samples for antibody cross-reactivity against the spike protein of SARS-CoV-1 (E), along with other alpha and beta coronaviruses (F-I) across disease severity. A static cutoff of 1000 AU/mL (anti-spike), 800 AU/mL (anti-RBD), and 5000 AU/mL (anti-N), depicted by the gray dashed line, was used to determine positive antibody response. SARS-CoV-2 PCR-positive mild, moderate, and severe disease cohorts had increased antibody titers against CoV-1, with increasing antibody concentration by increased disease severity (P < 0.001). HKU1 and OC43 anti-S concentrations were statistically significantly increased in severe disease groups, compared to mild and moderate disease groups (P < 0.001). *: P < 0.001

Figure 2.

Analysis of ACE2 inhibition by anti-S1(A) and anti-RBD (B) IgG antibodies by disease severity. Quantitative plasma anti-S had a correlation of r = 0.885 (P < 0.001) and anti-RBD had a correlation of r = 0.75 (P < 0.001) to ACE2 neutralization.

Figure 3.

Inflammatory markers across disease severity. Dot plot representation of cytokine, coagulation factors, cell surface markers, cardiovascular and apoptotic factors, across disease severity cohorts. IFNγ, TNFα, IL-1β, IL-4, IL-6, IL-10, IP-10, ICAM, Syndecan, Cathepsin-L, and D-Dimer statistically significantly differed between PCR-positive non-hospitalized (mild disease) and PCR-positive hospitalized (moderate and severe disease) cohorts (*:P < 0.05). D-Dimer also significantly differed among hospitalized moderate and severe COVID-19 disease groups (P < 0.05).

Figure 4.

Inflammatory dysregulation between cytokine microenvironment and antibody production. Correlation analysis of inflammatory marker plasma concentration and anti-spike antibody production across disease severity. Across all soluble proinflammatory markers, 6 showed a statistically significant positive correlation between cytokine or epithelial marker and antibody quantity regardless of COVID-19 disease severity, including ICAM (P < 0.001), IL-1β (P = 0.0233), IL-4 (P < 0.001), IL-6 (P = 0.037), TNFα (P < 0.001), and Syndecan (P < 0.001).

Figure 5.

Proinflammatory dysregulation between cytokine microenvironment and antibody neutralization capacity. Correlation analysis of inflammatory marker plasma concentration and anti-spike antibody ACE2 neutralization disease severity. Elevations in ICAM (P < 0.0001), IL-1β (P = 0.0233), IL-4 (P < 0.0001), IL-6 (P = 0.0001), TNFα (P = 0.001), and Syndecan (P = 0.00) all showed statistically significant correlations to antibody neutralization capacity amongst PCR-positive COVID-19 disease cohorts.

Figure 6.

Prevalence of anti-interferon antibodies in a subset of patients. Plasma IgG antibodies were profiled with a 58-plex bead-based protein array containing cytokines and chemokines. Example distributions of autoantibodies against interferons in healthy controls (HC, n = 17), PCR-negative mild patients (n = 31), PCR-positive mild patients (n = 27), and PCR-positive moderate/severe patients (n = 60) are shown. MFI is shown on the y-axis. A static cutoff of 3000 MFI, depicted by the gray dashed line, was used to determine positive anti-interferon response.