Early non-neutralizing, afucosylated antibody responses are associated with COVID-19 severity

Abstract

A damaging inflammatory response is implicated in the pathogenesis of severe coronavirus disease 2019 (COVID-19), but mechanisms contributing to this response are unclear. In two prospective cohorts, early non-neutralizing, afucosylated immunoglobulin G (IgG) antibodies specific to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) were associated with progression from mild to more severe COVID-19. To study the biology of afucosylated IgG immune complexes, we developed an in vivo model that revealed that human IgG-Fc-gamma receptor (FcγR) interactions could regulate inflammation in the lung. Afucosylated IgG immune complexes isolated from patients with COVID-19 induced inflammatory cytokine production and robust infiltration of the lung by immune cells. In contrast to the antibody structures that were associated with disease progression, antibodies that were elicited by messenger RNA SARS-CoV-2 vaccines were highly fucosylated and enriched in sialylation, both modifications that reduce the inflammatory potential of IgG. Vaccine-elicited IgG did not promote an inflammatory lung response. These results show that human IgG-FcγR interactions regulate inflammation in the lung and define distinct lung activities mediated by the IgG that are associated with protection against, or progression to, severe COVID-19.

Fig. 1.

Low early neutralizing titers and elevated Fc afucosylation are associated with COVID-19 progression. (A) The kinetics of neutralizing antibody responses were measured over time in Cohort 1. Half-maximal SARS-CoV-2 pseudovirus neutralizing titers (pNT50) are shown at each study time point, graphed based on days of symptoms for each participant. Samples were collected at study day 0 (D0 enrollment; n=101), 5 (D5; n=50), 14 (D14; n=33), 28 (D28; n=43), month 7 (M7; n=24) and month 10 (M10; n=9). (B) Heatmaps of pNT50 data are shown for progressors (P) (Cohort 1 n=8, Cohort 2 n=7) and non-progressors (NP) at enrollment timepoint (D0). The scale ranges from dark blue (no neutralization) to red (high neutralization). (C) SARS-CoV-2 spike-binding IgG (AUC) are shown for Cohort 1 progressors (P1, solid purple), Cohort 2 progressors (P2, solid blue), a random subset of non-progressors, and historic seronegative (SN) serum samples. (D) IgG1 Fc afucosylation abundance was measured in samples from progressors and non-progressors at enrollment timepoint (D0) of Cohort 1 (purple; progressors=P1, non-progressors=NP1) and in samples from Cohort 2 (blue; progressors=P2, non-progressors=NP2). RU, relative units. (E) IgG1 Fc afucosylation abundance was measured in patients who were hospitalized with COVID-19 (H, orange; n=52) and combined outpatient progressors (P, Cohort 1 progressors: purple), Cohort 2 progressors: blue) (n=15). (F) α-1,6-Fucosyltransferase 8 (FUT8) median fluorescence intensity (MFI) was measured in total CD19⁺ B cells and in plasmablasts (PB) from progressors (n=6) relative to sex-matched non-progressors (n=6). (G) The correlation for plasmablast expression of FUT8 and the abundance of IgG1 afucosylation is shown for matched samples. Solid and open circles represent data points from progressors and non-progressors, respectively. (H) Mean receiver operating characteristic (ROC) response and the area under the curve (AUC) with its standard deviation were obtained with a support vector machine classifier (SVM) using neutralization titers and IgG1 afucosylation. (I) ROC response and the AUC with standard deviation were obtained by testing the model on an independent Cohort 2. Median values are depicted in (C to F) with a solid black line. P values in (C) were calculated using Brown-Forsythe and Welch ANOVA test with Dunnett T3 correction, P values in (D and E) were calculated using Wilcoxon rank-sum test, and P values in (F) were calculated using unpaired Student’s test with Welch’s correction. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant. Pearson’s correlation coefficient (r = -0.6002, p = 0.0391) was computed in (G).

Fig. 2.

CD16a signaling potential is elevated in the myeloid compartment of progressors. Enrollment time point PBMCs were characterized in progressors (n=14) and a randomly selected subset of non-progressors (n=18). Solid purple and blue circles represent data points from progressors within Cohort 1 and Cohort 2, respectively whereas open circles represent data points from non-progressors. The median values have been depicted with a black line. (A) Total CD16a⁺ monocyte, CD16a⁺ CD14^- non-classical monocyte (NC), and CD16a⁺ CD14⁺ intermediate monocyte (Int) frequencies are shown as percentages of total CD11c⁺ HLA-DR⁺ CD3^- CD19^- CD56^- myeloid cells. (B) CD16a expression was measured on total CD16a⁺, non-classical, and intermediate monocyte populations. Receptor expression is measured in relative units (RU) (C) Mean ROC response and the AUC with its standard deviation were obtained using random forest classifier with 6-fold cross validation in two outpatient cohorts using FcγR expression on myeloid cells. (D) Radar plots summarizing the various features of IgG1-CD16a signaling axis in progressors and non-progressors are shown. Significant differences between the two groups are indicated with asterisks in the radar plot for progressors. P values in (A and B) were calculated using unpaired t tests with Welch’s correction. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 3.

mRNA vaccination elicits high neutralizing antibody titers with Fc glycoforms distinct from infection-induced IgG phenotypes. (A) The half-maximal SARS-CoV-2 pseudovirus neutralizing titers (pNT50) in healthy adults following mRNA vaccination (yellow n=29) or in COVID-19 outpatients on study day 28 (blue n=42) are shown. PD1: post-dose 1, PD2: post-dose 2, M3: month 3. (B) Longitudinal analysis of IgG subclasses is shown for day 21, 28, or 42 post-primary vaccination (n=17). (C) SARS-CoV2 IgG1 Fc posttranslational modifications were analyzed in samples from patients hospitalized with COVID-19 (n=52), COVID-19 outpatients (day 28 n=36) and in participants who received the Pfizer BNT162b2 SARS-CoV-2 mRNA vaccine (day 28 post primary vaccination, n=16). F0: afucosylation, S: sialylation, N: bisection, GS0: galactosylation. (D) Longitudinal analysis of anti-SARS-CoV-2 IgG1 Fc afucosylation (afucFc, red line) and sialylation (sFc, blue line) is shown on day 21, 28 or 42 post-primary vaccination. The median values in (A and C) are depicted with a black line. P values in (A) were calculated using Kruskal Wallis test with Dunn’s correction, in (B) using mixed effect analysis with Geisser-Greenhouse and Tukey’s corrections, and in (C) using a two-way ANOVA and one-way ANOVA with Tukey’s correction. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Fig. 4.

Afucosylated IgG immune complexes promote immune cell infiltration and proinflammatory cytokine production in vivo. (A) Immune cells were measured in the bronchoalveolar lavage (BAL) fluid of human Fcγ receptor mice (hFcγR) that were treated with either afucosylated (F0, pool 1), normally fucosylated (F, pool 2), or vaccine-induced (Vax, pool 3) immune complexes or with spike protein alone. Immune complexes and spike protein were administered by the intratracheal route. (B) Cytokine and chemokine concentrations in the BAL of the indicated groups of mice are shown. The size of the bubble represents normalized cytokine and chemokine concentrations. P values are indicated for each soluble factor (blue: F0 versus F, red: F0 versus Vax). (C and D) Immune cell subsets (C) as well as cytokines (TNF-α, IL-6, IL-10) and chemokines (CXCL1, CCL3, CCL4) (D) were quantified in the bronchoalveolar lavage (BAL) fluid of hFcγR or hFcγR mice with a specific deletion in CD16a (CD16a^−/−) that received afucosylated (F0, pool 1) or normally fucosylated (F, pool 2) immune complexes by intratracheal administration. In (A) and (C) neutrophils were defined as Ly6G⁺ CD11b⁺ CD3^- B220^- cells and total monocytes defined as CD11b⁺ Ly6G^- MERTK^- MHC IA/IE^- CD3^- B220^- cells. (E) Frequency of Ly6G⁺ CD11b⁺ CD3^- B220^- neutrophils was measured in BAL fluid of hFcγR mice that were pre-treated with chemokine neutralizing mAbs (anti-CXCL1 and anti-CCL3) or isotype control followed by administration of afucosylated immune complexes. The median and the 95% confidence interval are shown in each graph. P values in (A to D) were calculated using a one-way ANOVA with Dunnett’s correction using n=3 mice per group for A and B and n=4 mice per group for (C to E). Data in (A to E) are representative of at least two independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant.