Research ArticleImmunologyVirology Open Access | 10.1172/jci.insight.197773

Natalia A. Kuzmina,1,2 Sivakumar Periasamy,1,2 Kritika Kedarinath,1,2 Keziah Hernandez,1,2 Caroline Atyeo,3 S. Moses Dennison,4 Kan Li,4 Daniel Bedinger,5 Sharon L. Schendel,6 Georgia D. Tomaras,4 Hanif Ali,7 Galit Alter,3 Erica Ollmann Saphire,6,8 and Alexander Bukreyev1,2,9,10

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Kuzmina, N. in: PubMed | Google Scholar

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Periasamy, S. in: PubMed | Google Scholar

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Kedarinath, K. in: PubMed | Google Scholar

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Hernandez, K. in: PubMed | Google Scholar

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Atyeo, C. in: PubMed | Google Scholar |

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Dennison, S. in: PubMed | Google Scholar

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Li, K. in: PubMed | Google Scholar

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Bedinger, D. in: PubMed | Google Scholar

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Schendel, S. in: PubMed | Google Scholar

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Tomaras, G. in: PubMed | Google Scholar |

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Ali, H. in: PubMed | Google Scholar

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Alter, G. in: PubMed | Google Scholar |

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Saphire, E. in: PubMed | Google Scholar

1Department of Pathology, University of Texas Medical Branch (UTMB), Galveston, Texas, USA.

2Galveston National Laboratory, Galveston, Texas, USA.

3Ragon Institute of MGH, MIT, and Harvard, Cambridge, Massachusetts, USA.

4Center for Human Systems Immunology, Departments of Surgery, Immunology, and Molecular Genetics and Microbiology and Duke Human Vaccine Institute, Duke University, Durham, North Carolina, USA.

5Carterra Inc. Salt Lake City, Utah, USA.

6Center for Infectious Disease and Vaccine Research, La Jolla Institute for Immunology, La Jolla, California, USA.

7Quadrucept Bio Ltd, Kemp House, London, United Kingdom.

8Department of Medicine, University of California San Diego, La Jolla, California, USA.

9Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, Texas, USA.

10Center for Biodefense and Emerging Viral Infections, University of Texas Medical Branch, Galveston, Texas, USA.

Address correspondence to: Alexander Bukreyev, University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555, USA. Email: alexander.bukreyev@utmb.edu.

Find articles by Bukreyev, A. in: PubMed | Google Scholar |

Published February 23, 2026 - More info

Published in Volume 11, Issue 4 on February 23, 2026
JCI Insight. 2026;11(4):e197773. https://doi.org/10.1172/jci.insight.197773.
© 2026 Kuzmina et al. This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Published February 23, 2026 - Version history
Received: July 10, 2025; Accepted: January 2, 2026 View PDF Abstract

Antibody-dependent enhancement (ADE) of infection is a well-described phenomenon for several viruses, including dengue, Ebola, respiratory syncytial virus, and HIV. ADE occurs when virus-antibody complexes engage Fc receptors (FcRs) and virus-specific receptors, enhancing infection under conditions of incomplete neutralization. The Coronavirus Immunotherapeutic Consortium (CoVIC) assembled a comprehensive dataset of functional properties for over 400 mAbs, enabling direct comparison of neutralization, Fc-mediated functions, receptor binding, and infection of immune cells. Infection rates in most primary human immune cell types were low, with modest increases observed for some mAbs. In contrast, macrophages were more susceptible to SARS-CoV-2 and exhibited substantial ADE with select mAbs. ADE was completely inhibited by FcR blockade and significantly reduced by antibody- or ceftazidime-mediated blocking of angiotensin-converting enzyme 2 (ACE2). Neutralization potency did not correlate with ADE, as both strongly and weakly neutralizing antibodies induced enhancement. Instead, ADE magnitude depended on an antibody’s ability to block spike protein binding to ACE2. Importantly, ADE resulted in productive infection with release of infectious virus. Evaluation of antibodies against the BA.1 (Omicron) variant revealed reduced or lost ADE for most mAbs, with increased ADE observed for several mAbs relative to the USA-WA1/2020 strain.

Introduction

The Coronavirus Immunotherapeutic Consortium (CoVIC) comprising 56 partners across the world was created for global partnership to accelerate discovery, optimization, and delivery of life-saving antibody-based therapeutics against SARS-CoV-2. The CoVIC alliance managed contribution of multiple antibodies and comparative investigation of their binding to the target, neutralizing activities, binding to Fc receptors (FcRs), Fc-mediated effects, epitopes, and protective efficacy in animal models (1), resulting in the creation of a database (2) allowing a direct comparison of a large number of mAbs.

Antibody dependent enhancement (ADE) of infection is a well-known phenomenon described for multiple viruses: dengue (3, 4), respiratory syncytial virus (5), influenza (6, 7), and Ebola (8–10). As shown for many viral infections, antibodies may enable viral entry into FcγR-bearing cells, bypassing specific receptor-mediated entry, and can amplify the number of virions entering cells or enhance infection by involving immune mechanisms (11). If these antibodies are unable to neutralize viruses, this increased entry leads to an enhancement of a productive infection (11). ADE was documented for several coronaviruses as well: Middle East respiratory syndrome–related coronavirus (MERS) (12), SARS-CoV (13–16), and feline infectious peritonitis virus (FIPV) (17–20), but it never appeared to be as strong as was described for dengue virus. Dengue differs from other viruses because it targets monocytes, macrophages, and DCs and can produce progeny virus in these cells (4). Unlike dengue, the preexisting immunity gained from previous exposure with other coronaviruses does not promote ADE during infection with SARS-CoV-2 (21), suggesting that SARS-CoV-2, SARS-CoV, and MERS-CoV have a significant antigenic distance, and the observed cross-reactivity with SARS-CoV and MERS-CoV spike protein (22) does not contribute to the enhancement. SARS-CoV-2 uses angiotensin-converting enzyme 2 (ACE2) as a receptor to infect host cells (23, 24). ACE2 expression in primary immune cells varies broadly; macrophages demonstrate detectable levels of ACE2 (25–27). Compared with primary macrophages, THP-1 cells show lower baseline ACE2 expression, but their response to stimuli mirrors primary cells (28–30). After recognition of ACE2 by the spike protein on the surface of SARS-CoV-2 and membrane fusion, virus enters target cells (24, 31). Though SARS-CoV uses the ACE-2 receptor for entry as well, the binding affinity of SARS-CoV-2 to this receptor is up to 20-fold higher compared with SARS-CoV, making it pivotal for higher infectivity of SARS-CoV-2 compared with other coronaviruses (32). The receptor binding domain (RBD) of the spike protein is a key target for neutralizing antibodies (33) and is highly variable between individual coronaviruses; in addition, its mutations allow the virus to evade the antibody response (34). RBD plays a central role in this entry. Blocking of binding of the spike protein to ACE2 is an effective way to inhibit the infection of target cells by SARS-CoV-2 (35, 36).

For SARS-CoV-2, at least 2 case reports described ADE that possibly occurred due to the infusion of mAbs and led to an acute boost in the severity of COVID-19 pneumonia (37, 38). Importantly, 2 mAbs used for treatment of SARS-CoV-2, casirivimab and imdevimab, were found to induce ADE in vitro (39). Recent reports have demonstrated that some plasma samples from patients with COVID-19 can enhance SARS-CoV-2 infection only in cells expressing both FcR and ACE2 (40, 41), and SARS-CoV-2 may promote ACE2 expression in immune cells (42–45). Moreover, ACE2 may act as a secondary receptor required for antibody- and FcγR-mediated elevated entry of SARS-CoV-2 (46). This study aimed to investigate the relative contributions of FcγRs, ACE2, neutralization potencies, and epitope specificity in ADE. We used a large and diverse panel of antibodies to understand how these factors influence ADE and to identify the most significant factors contributing to ADE. We show that ADE caused by these antibodies can be inhibited by blocking either the FcγR or ACE-2, suggesting a dual mechanism of the enhancement.

Results

SARS-CoV-2 ADE is mediated by both neutralizing and non-neutralizing antibodies in a dose-dependent manner. CoVIC has analyzed a panel of 407 mAbs in multiple in vitro and in vivo assays and comparative data has been uploaded into a publicly accessible database (https://covic.lji.org). To monitor viral infection, we used a recombinant SARS-CoV-2 expressing neon green (SARS-CoV-2-mNG), which allows quantification of virus-infected cells based on fluorescence of neon green intracellularly produced from virus-encoded mRNA (47). We measured neutralizing potencies and ADE at various concentrations of mAbs in vitro (Figure 1A) and analyzed their properties, available from the database, which can affect their abilities to decrease or increase viral infection. For this work, 364 IgG1 antibodies from the CoVIC panel were selected, the majority of which were nonmodified mAbs. For comparative purposes, we also included several engineered multivalent antibodies (Supplemental Figure 1; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.197773DS1). Their binding magnitude to various FcRs were comparable with the entire analyzed dataset (Supplemental Figure 2), which is shown in https://covicdb.lji.org/ The monocytic cell line THP-1 was treated with a panel of 364 selected mAbs at concentrations of 10, 1, or 0.1 μg/mL, and SARS-CoV-2-mNG was added at MOI 3 PFU per cell. Cells were incubated for 48 hours and the numbers of infected (mNG+) cells were quantified by flow cytometry (Figure 1, B–D). We observed dose-dependent enhancement of infection (ADE), with a maximum at concentration of 1 μg/mL for the most mAbs, and a typical dependence of ADE magnitude on mAb concentration: the increase of ADE as concentration decreased from 10 μg/mL to 1 μg/mL, and then ADE subsequently decreased at lower concentrations. In contrast, no ADE was observed with the VIC16 mAb specific for Ebola virus (48) (Supplemental Figure 3). In a separate experiment, the viability of infected cells treated with representative SARS-CoV-2 mAbs or an irrelevant control antibody was confirmed (Supplemental Figure 4). Furthermore, we demonstrated that 2 representative mAbs that caused strong ADE (CoVIC-27 and CoVIC-113), 2 representative mAbs that caused no ADE (CoVIC-41 and CoVIC-147), and an irrelevant antibody bound FcRs on THP-1 cells with a similar efficiency (Supplemental Figure 5). These data suggest that ADE occurs under conditions of incomplete virus neutralization. Because of the drop of enhancement at the lowest concentration evaluated — 0.01 μg/mL in preliminary experiments — this concentration was excluded from further evaluation. The mean increases compared with no antibody control based on all mAbs tested were 429%, 3,814%, and 2,319% at 10, 1, and 0.1 μg/mL, respectively (P = 0.0001, 1-way ANOVA). The percentages of infected cells or the extent of ADE caused by mAbs at 1 μg/mL were considered the most representative and used for comparison of mAbs. Unexpectedly, we observed a lack of correlation between the neutralization potencies of mAbs (IC50) and their ability to mediate ADE (percentage enhancement compared with no antibody control), R2 = 0.0098. Moreover, some very potent antibodies caused strong ADE (Figure 1, E and F).

Figure 1

Both neutralizing and non-neutralizing mAbs increase susceptibility of THP-1 cells to SARS-CoV-2. (A) Schematic of virus neuralization assay. (B) Schematic of assessment of ADE. (C) Various mAbs (indicated at right) demonstrate different extent of ADE (tested at concentration of 1 μg/mL): no or low (<2% infected cells); moderate (2%–20% infected cells), and strong (>20% infected cells). CC12.3 and CC12.14 served as control antibodies for the CoVIC panel (1). (D) ADE at various mAb concentrations in THP-1 cells: no or low ADE, moderate ADE, strong ADE. For most of the mAbs, the greatest enhancement was observed at 1 μg/mL. The antibodies are indicated at the right. (E) Heatmap of tested mAbs ranked by IC50 and the extent of ADE. (F) Absence of a direct correlation between neutralization and ADE, R2 = 0.0097 for simple linear regression, IC50 in ng/mL and ADE mediated by antibodies used at the dose 1 mg/mL, percentage of infection with no mAb control.

SARS-CoV-2 ADE is preferentially mediated by antibodies that inhibit spike protein binding to ACE2 receptors. We distinguished 2 sets of antibodies by their effectiveness in obstructing viral binding to ACE2 receptors: those with low (<50%) and with high (>50%) blocking capabilities, based on the inhibition measured by biolayer interferometry (49). The proportions of antibodies mediating low ADE (<10-fold compared with no mAb control or 2% infected cells), moderate ADE (10–100-fold, or 2%–20% infected cells), or high ADE (>100-fold, or >20% infected cells) varied across these 2 groups (Figure 2, A and B). Antibodies that were less effective in blocking binding of RBD to ACE2 demonstrated low ADE in most cases. In contrast, many of the antibodies that significantly reduced binding of RBD to ACE2 were associated with strong ADE. This difference between the antibody groups was significant (χ2 test, P < 0.0001), indicating that the ability of antibodies to block spike binding to ACE2 receptors is implicated in ADE.

Figure 2

The extent of ADE depends on binding to ACE2 and Fcγ receptors. (A) Schematic of the experiments. (B) Proportion of mAbs causing low, moderate, and strong ADE differs among mAbs that block viral binding to ACE2 weakly versus strongly. (C) Blocking of ACE2 receptors with indicated polyclonal antibodies (10 μg/mL), chemical inhibitors ceftazidime (400 μM), dalbavancin (25 μM), or FcγRI or FcγRIIa receptors with anti CD32 or CD64 antibodies significantly reduces ADE mediated by CoVIC-58. Relative percentages of infected cells normalized to no antibody control. Tukey’s multiple-comparison test. (D) Effects of FcR modifications on ADE tested with the CR3022 mAb. VIC16, which is an IgG1 specific for Ebola virus glycoprotein (48), included as isotype control. Dunnett’s multiple-comparison test. NS, not significant. (E) Effects of blocking of ACE2 receptors and lysosomal cathepsin inhibitor E64d on ADE. (F) Heatmap of the relative effects of each treatment on ADE caused by individual mAbs assessed by percentage of infected cells. Treatment with ceftazidime or anti-ACE2 polyclonal antibodies and endosomal inhibitors has a cumulative effect, suggesting distinct mechanism of ADE inhibition.

The magnitude of ADE depends on binding to ACE2 and Fcγ receptors. To investigate the involvement of ACE2, which is the receptor of SARS-CoV-2 in ADE mediated by the virus, we utilized THP-1 cells (14, 50). The THP-1 cells used in our experiments demonstrated detectable expression of ACE2 (Supplemental Figure 6). We blocked binding of the antibodies to ACE2 by either ceftazidime, which directly interacts with the S-RBD (35) at 400 μM, or by mono- and polyclonal ACE2 blocking antibodies at 10 μg/mL for 1 hour. The cells were inoculated with SARS-CoV-2 at MOI 3 PFU per cell and incubated for 24 hours. The optimal concentration of ceftazidime was identified in preliminary experiments. We observed a significant decrease of infection when ACE2 receptors were blocked with ceftazidime or with mono- and polyclonal antibodies against ACE2 (Figure 2C). Next, we incubated THP-1 cells with anti-CD32 (FcγIIa) and anti-CD64 (FcγI) mAbs for 1 hour to block FcγRs, added CoVIC-58 mAb at 1 μg/mL, inoculated the cells with SARS-CoV-2 at MOI 3 PFU per cell, and incubated the cells for 24 hours. Blocking of FcγRIIa or FcγRI reduced ADE; in the latter case, the effect was more pronounced, which can be explained by a higher affinity of FcγRI compared with FcγRIIa (51). The partial inhibition of ADE by FcγR-blocking antibodies alone suggests that FcγRs are not the sole mediators of ADE. The complete abrogation of ADE observed with the combined use of ceftazidime and anti-FcγR antibodies supports the involvement of ACE2 in the enhancement mechanism (Figure 2C).

To further investigate the significance of FcRs in ADE, we evaluated antibodies with mutations within their Fc domain: N297Q, which completely eliminates the antibody’s ability to interact with the FcRs (52) and S239D/I332E (SDIE), which enhances Fc functions (53) (Figure 2D). We selected the CR3022 mAb, which is specific for SARS-CoV and capable of binding to but not neutralizing SARS-CoV-2. At 1 μg/mL, CR3022 in the form of IgG1 caused low ADE, which was completely abrogated by the mutation N297Q but significantly increased when the antibody version with the SDIE mutation was used. Importantly, in our previous studies, we demonstrated that this Fc modification exacerbated the severity of COVID-19 in antibody-treated mice, which was evidenced by increased viral load in the lungs and a greater loss of weight (54).

Cathepsin L plays a key role in SARS-CoV-2 infection in humans and is required for virus cell entry (55). To test whether the cysteine protease cathepsin L affects ADE, we administered the E64d inhibitor to THP-1 cells (Figure 2E). We observed a considerable decrease in the infected cell count across all tested enhancing mAbs, suggesting that cathepsin cleavage is required for effective ADE. The concurrent treatment with ACE2 and E64d inhibitors demonstrated a cumulative effect, indicating that these inhibitors suppress virus entry through distinct mechanisms (Figure 2F).

Epitope specificity has only a limited effect on the magnitude of ADE. In our previous study, antibodies that bind to soluble RBD were grouped into 7 primary epitope groups, known as RBD-1 through RBD-7. These groups were further divided into subgroups according to the level of pairwise competition with other subgroups (1, 56). A set of mAbs, each specific to 1 of the 7 epitope groups, was chosen and evaluated for their capacity to facilitate ADE (Figure 3A). Antibodies targeting epitope groups 5 and 6 exhibited markedly different ADE profiles. Most mAbs from group 5 showed minimal ADE, despite their interaction with the RBD’s outer surface (1). Conversely, mAbs that target epitope group 6, which attach to the RBD’s inner surface, displayed high ADE (Figure 3A). These epitope groups also presented a consistent pattern at the subgroup level in their capacity to enhance ADE (Figure 3, B and C). However, for other epitope groups, no clear link was found between the specific groups and the extent of ADE, suggesting that the exact location of the binding site may be not a major factor in ADE. Next, we tested the neutralizing activities of the mAbs (Figure 3D), which suggested that the pronounced ADE observed in mAbs belonging to epitope group 6 could be due to their reduced neutralizing effectiveness. Although the majority of group 5 mAbs, which weakly block binding of RBD to ACE2, were also found to have low neutralizing activity, those that effectively block binding of RBD to ACE2 exhibited potent neutralizing capabilities. Remarkably, mAbs belonging to subgroups 5a and 5b, which have low neutralizing activity, also have a low affinity to the spike protein; these mAbs nonetheless demonstrate good protection in vivo (56). These antibodies are able to cross-link adjacent spikes together (1), and their enhanced in vivo protection despite moderate in vitro neutralization may be linked to this cross-linking ability.

Figure 3

The extent of ADE depends on binding to ACE2 and Fcγ receptors and less on epitope groups. (A) Proportion of mAbs from different epitope groups mediating low, moderate, and high ADE based on percentage relative to no mAb control. (B) Percentages of infected THP-1 cells treated with the selected mAbs of various epitope groups (shown under the graphs) and subgroups: 1, 2a, 2b, 2c, 2d, 3a, 3b, 4a, 4b, 4c, 5a, 5b, 5c, 5d, 6a, 6b, 7a, 7b, and 7c at 1 μg/mL. (C) Proportions of mAbs mediating low, moderate, and high ADE within epitope groups and subgroups described for B. (D) Proportion of mAbs from different epitope groups that cause no, moderate, or strong neutralization based on IC50. (E) Multivariable analysis reveals distinct clusters of mAbs with low, moderate, or strong ADE based on their ability to block ACE2 receptors, the FcγRIIa activity, neutralization IC50, ADE, and ADMP. (F) A loadings plot showing a relationship between ADE and ACE2, and between ADMP and FcγR activities. (G) Spearman’s correlation matrix of variables used for the principal component analysis.

Out of the 329 mAbs with known binding to RBD or N-terminal domain (NTD) tested for ADE, 131 mAbs bind exclusively to RBD, 154 mAbs bind to both RBD and NTD, and 44 mAbs bind solely to NTD (49) (https://covicdb.lji.org). Interestingly, about half of antibodies binding to RBD alone or to both RBD and NTD induced high ADE, showing no significant difference between these 2 groups. However, all antibodies specific to NTD alone did not promote ADE in contrast to antibodies specific to RBD (P < 0.0001, Fisher’s exact test). Importantly, all these NTD binders did not block binding to ACE2.

We next determined the contribution of individual mAb properties including binding to FcγR2a assessed by flow cytometry, blocking of RBD binding to ACE2, neutralization potencies (IC50), and antibody-dependent monocyte phagocytoses (ADMP) in ADE. The properties of mAbs were taken from the CoVIC database (https://covicdb.lji.org). Principal component analysis revealed clusters of mAbs with strong, moderate, and low ADE (Figure 3, E and F). Interestingly, ADE was strongly associated with the capability of mAbs to block RBD binding to ACE2 receptors. As expected, ADMP was strongly associated with binding to FcγR2a, and no correlation was observed between neutralization and binding to FcγR2a or ADMP. Furthermore, we found moderate negative correlation between neutralization and blocking of binding to ACE2 (Figure 3G).

SARS-CoV-2 antigenic drift does not increase the overall ability of the antibodies to cause ADE. To investigate the effect of SARS-CoV-2 antigenic drift on the viral ability to cause ADE, 139 mAbs from the CoVIC panel were tested for ADE with the recombinant USA-WA1/2020, in which the spike protein was replaced with the counterpart from the Omicron BA.1 strain, which is hereinafter referred as the BA.1 strain (Figure 4A). The proportions of mAbs that cause high and moderate ADE with the original USA-WA1/2020 strain were strongly reduced with the BA.1 strain, with a significant number of mAbs (64 in total) that no longer facilitated ADE. The fractions of mAbs that exhibited either low or moderate ADE were similar, at 74% for USA-WA1/2020 and 84% for BA.1 (Figure 4B), with a decreasing number of mAbs showing moderate ADE for BA.1. Only 8% of the tested mAbs increased their ability to mediate ADE with BA.1 compared with the original USA-WA1/2020 strain (Figure 4C). The changes in the ability to promote ADE with the BA.1 variant were observed within all epitope groups. We noted that among the mAbs that lost the ability to mediate ADE, the members with the biggest changes belonged to epitope groups 2 and 6 (Figure 4D). The greatest proportion of mAbs that maintained strong ADE or gained ADE belonged to epitope group 7. Overall, these data suggest that the tested BA.1 strain of SARS-CoV-2 does not have an increased ability to cause ADE with the CoVIC mAbs. Moreover, 64% of mAbs completely lost their ability to mediate ADE when tested with the BA.1 strain. The observed reduction of ADE could be due to the lesser antigenic match between the virus and the mAbs. It is possible that mAbs