key: cord-1016539-9zu4gkx6
authors: Adeniji, Opeyemi S.; Giron, Leila B.; Zilberstein, Netanel F; Shaikh, Maliha W.; Balk, Robert A; Moy, James N; Forsyth, Christopher B.; Keshavarzian, Ali; Landay, Alan; Abdel-Mohsen, Mohamed
title: COVID-19 Severity Is Associated with Differential Antibody Fc-mediated Innate Immune Functions
date: 2021-01-11
journal: bioRxiv
DOI: 10.1101/2021.01.11.426209
sha: 71cb2c46ea3d365932813a786155e3c0c894616a
doc_id: 1016539
cord_uid: 9zu4gkx6

Beyond neutralization, antibodies elicit several innate immune functions including complement deposition (ADCD), phagocytosis (ADCP), and cytotoxicity (ADCC). These functions can be both beneficial (by clearing pathogens) and/or detrimental (by inducing inflammation). We tested the possibility that qualitative differences in SARS-CoV-2 specific antibody-mediated innate immune functions contribute to Coronavirus disease 2019 (COVID-19) severity. We found that antibodies from hospitalized COVID-19 patients elicited higher ADCD but lower ADCP compared to antibodies from non-hospitalized COVID-19 patients. Consistently, higher ADCD was associated with higher systemic inflammation during COVID-19. Our study points to qualitative, differential features of anti-SARS-CoV-2 antibodies as potential contributors to COVID-19 severity.

While most SARS-CoV-2 infected individuals exhibit asymptomatic or mild respiratory 51 tract infection, a significant population face severe symptoms requiring hospitalization [1, 52 2] . Hospitalized Coronavirus Disease 2019 (COVID-19) is associated with a state of hyper-53 inflammation and increased complement activation [1, [3] [4] [5] . However, the mechanisms that 54 contribute to this hyper-inflammation are not fully clear. 55

Higher titers of SARS-CoV-2 specific neutralizing antibodies have been associated with 56 higher COVID-19 severity [6, 7] . However, beyond neutralization, antibodies can elicit an 57 array of Fc-mediated innate immune functions such as antibody-dependent complement 58 deposition (ADCD), antibody-dependent cellular phagocytosis (ADCP), and antibody-59 5 cells. 2.5 x 10 4 S1-or RBD-pulsed (ACE2)-CHO cells were mixed with 20 µg IgG and 109 incubated for 30 min at 37˚C. 1 x 10 5 human NK cells isolated from PBMC using EasySep 110 human NK cell isolation kit (Stem Cell Technologies) were added to wells in the presence 111 of CD107a PE (BD) and Golgi stop (BD). Pulsed (ACE2)-CHO immune complexes were 112 then added to the wells, mixed, pelleted, and incubated at 37˚C for 16 h. Post incubation, 113 NK cell activation was detected by staining for CD56 PerCP-Cy5.5 (BD), IFN-γ (BD), 114 TNF (BioLegend) and acquired via flow cytometry. Data are reported as the percentage of 115 cells positive for the marker as indicated. 116

Measurement of S1-and RBD-specific antibody titers. S1-and RBD-specific antibody 117 titers were measured in the plasma of the 80 individuals using MSD V-PLEX multiplex 118 assay (Meso Scale Diagnostic). 119 Plasma levels of GM-CSF, IFN-β,  120   IFN-γ, IL-10, IL-13, IL-1β, IL-33, IL-4, IL-6, TNF-α, Fractalkine, IL-12p70, IL-2, IL-21,  121   IL-22, IL-23, IP-10, MCP-2, MIP-1α, SDF-1a, IFN-α2a, IL-12/IL-23p40, and IL-15 were  122 determined using customized MSD U-PLEX multiplex assay (Meso Scale Diagnostic). 123

Plasma levels of C-Reactive Protein (CRP), Galectin-1, Galectin-3, Galectin-9, Growth 124 Differentiation Factor-15 (GDF-15), soluble CD14 (sCD14), soluble CD163 (sCD163), 125 LPS Binding Protein (LBP), and FABP2/I-FABP were measured using ELISA kits (R&D 126 Systems). The plasma level of zonulin was measured using an ELISA kit from 127

MyBiosorce. Levels of occludin were measured by ELISA (Biomatik). Plasma levels of 128 Reg3A were measured by ELISA (RayBiotech). β-glucan detection in plasma was 129 performed using Limulus Amebocyte Lysate (LAL) assay (Glucatell Kit, CapeCod). 

Severe COVID-19 is associated with higher ADCD and lower ADCP compared to mild 136 COVID-19. IgG from hospitalized COVID-19 patients elicited significantly higher ADCD 137 6 against S1-and RBD-coated target cells compared to IgG from non-hospitalized 138 individuals (Figure 1a, b) . In contrast, a higher fraction of IgG from non-hospitalized 139 COVID-19 patients elicited ADCP activity above background (maximum value from the 140 SARS-CoV-2 negative group) compared to hospitalized patients (Figure 1c-e) . S1-specific 141 antibodies from hospitalized individuals induced significantly higher NK cell 142 degranulation and intracellular cytokine production (a surrogate of ADCC) than did S1-143 specific antibodies from non-hospitalized individuals (Fig. 1f, g) . In contrast, RBD-144 specific antibodies from non-hospitalized individuals induced significantly higher ADCC 145 surrogates than did RBD antibodies from hospitalized individuals (Fig. 1h, i) . These data 146 suggest that hospitalized COVID-19 is associated with differential, qualitative features of 147 SARS-CoV-2-specific antibodies. In particular, hospitalized COVID-19 is associated with 148 a higher ability of antibodies to elicit complement deposition and a lower ability to elicit 149 phagocytosis. These data are compatible with reports suggesting that complement immune 150 system plays a significant negative role in Coronavirus disease pathogenesis [9-11]. 151

sought to examine whether the differential levels of ADCD, ADCP, and ADCC between 153 hospitalized and non-hospitalized patients can be simply explained by the titers of anti-S1 154 and anti-RBD antibodies. There was no significant difference in the titer of S1-specific 155 antibodies between hospitalized and non-hospitalized patients in our cohort (Fig. 2a) . 156

When we examined S1-specific antibody titers in patients having ADCD, ADCP, or ADCC 157 values below or above background levels (defined as the maximum values observed in the 158 SARS-CoV-2 negative group), we found no difference in ADCD or ADCC but a 159 significant difference in ADCP ( Fig. 2b-d) . Similar results were observed with RBD-160 specific antibodies ( Fig. 2e-h) . These data indicate that the quantities of S1-or RBD-161 specific antibodies could explain the differences in ADCP activity between hospitalized 162 and non-hospitalized patients (the higher the antibody titer, the higher the ADCP activity) 163 but cannot fully explain the differences in ADCD or ADCC activity between hospitalized 164 and non-hospitalized patients, suggesting that qualitative rather than quantitative properties 165 of anti-SARS-CoV-2 antibodies associate with disease severity. 166

Anti-SARS-CoV-2 ADCD associates with higher inflammation. Given that complement 167 activation can significantly contribute to inflammation-mediated tissue damage, including 7 during COVID-19 [12, 13], we examined the relationship between the ADCD property of 169 IgGs (as well as ADCP and ADCC) and an array of 39 plasma markers of systemic 170 inflammation and immune activation measured using ELISA and multiplex cytokine arrays 171 (a list in Supplementary Table 2 ). As expected, higher S1-and RBD-specific ADCD was 172 associated with higher levels of several markers of inflammation and immune activation 173 (Fig. 2f) . In contrast, higher levels of S1-and RBD-specific ADCP were associated with 174 lower levels of several inflammatory markers (Fig. 2f) . Finally, ADCC exhibited an 175 antigen-specific relationship with inflammation: anti-S1 specific ADCC exhibited negative 176 correlations with inflammation, whereas anti-RBD exhibited positive correlations with 177 inflammation (Fig. 2f) . These data further point to the potentially detrimental effects of 178 ADCD on COVID-19 pathogenesis. 179

Future studies will be needed to examine other antigen-specific antibodies including anti-180 N antibodies, as well as different antibody classes such as IgA and IgM. These studies will 181 also need to examine the mechanisms that underlie these differential, qualitative features. be needed. Our sample size did not allow for addressing potential confounders. 187 Independent test sets from large cohorts from varying geographic and demographic settings 188 will be needed in future studies. Finally, it will be important to understand how these 189 antibody features impact the function of vaccine-mediated antibodies as well as therapeutic 190 antibodies in order to achieve an optimal balance between neutralization and innate 191 immune functions without inciting potential side effects. 192

In summary, our data suggest that differential, qualitative Fc-mediated antibody effector 193 properties of anti-SARS-CoV-2 antibodies associate with COVID-19 severity. In 10 7 S1-speci c antibodies pg/ml plasma ns S1-specific ADCC (% CD107 + IFNγ + NK cells) e g h f

Clinical Characteristics of Coronavirus Disease 2019 in 262 China

Mild or Moderate Covid-19

Imbalanced Host Response to 265 SARS-CoV-2 Drives Development of COVID-19

Clinical and immunological features of severe and 267 moderate coronavirus disease 2019

COVID-19: consider cytokine storm syndromes 269 and immunosuppression

Low serum neutralizing anti-SARS-CoV-2 S 271 antibody levels in mildly affected COVID-19 convalescent patients revealed by two 272 different detection methods

Deep immune profiling of COVID-19 patients 274 reveals distinct immunotypes with therapeutic implications

Beyond binding: antibody effector functions 276 in infectious diseases

Complement as a target in COVID-280 19?

Complement activation in the disease 283 course of COVID-19 and its effects on clinical outcomes

Systemic complement activation is associated 285 with respiratory failure in COVID-19 hospitalized patients

Optimization and qualification of an Fc Array assay 288 for assessments of antibodies against HIV-1/SIV

Proinflammatory IgG Fc structures in 290 patients with severe COVID-19