Neutralizing antibodies from the rare convalescent donors elicited antibody-dependent enhancement of SARS-CoV-2 variants infection

Abstract

Currently, neutralizing antibody and vaccine strategies have been developed by targeting the SARS-CoV-2 strain identified during the early phase of the pandemic. Early studies showed that the ability of SARS-CoV-2 RBD or NTD antibodies to elicit infection enhancement in vivo is still controversial. There are growing concerns that the plasma and neutralizing antibodies from convalescent patients or people receiving vaccines mediate ADE of SARS-CoV-2 variants infections in immune cells. Here, we constructed engineered double-mutant variants containing an RBD mutation and D614G in the spike (S) protein and natural epidemic variants to gain insights into the correlation between the mutations in S proteins and the ADE activities and tested whether convalescent plasma and TOP10 neutralizing antibodies in our laboratory mediated the ADE effects of these SARS-CoV-2 variants. We found that one out of 29 convalescent plasma samples caused the ADE effect of pandemic variant B.1.1.7 and that the ADE effect of wild-type SARS-CoV-2 was not detected for any of these plasma samples. Only one antibody, 55A8, from the same batch of convalescent patients mediated the ADE effects of multiple SARS-CoV-2 variants in vitro, including six double-mutant variants and four epidemic variants, suggesting that ADE activities may be closely related to the antibody itself and the SARS-CoV-2 variants' S proteins. Moreover, the ADE activity of 55A8 depended on FcγRII on immune cells, and the introduction of LALA mutations at the Fc end of 55A8 eliminated the ADE effects in vitro, indicating that 55A8^LALA may be a clinical drug used to prevent SARS-CoV-2 variants. Altogether, ADE may occur in rare convalescent patients or vaccinees with ADE-active antibodies who are then exposed to a SARS-CoV-2 variant. These data suggested that potential neutralizing antibodies may need to undergo ADE screening tests for SARS-CoV-2 variants, which should aid in the future design of effective antibody-based therapies.

Keywords: SARS-CoV-2 variants; antibody-dependent enhancement; infection-enhancing antibodies; neutralizing antibody; receptor-binding domain (RBD).

Figure 1

ADE activities of the plasma samples from the COVID-19 convalescent patients. The B.1.1.7 pseudovirus was preincubated with serially diluted plasma samples, and the mixtures were added to Daudi cells to evaluate their ability to enhance infection. Each curve represents an individual plasma sample. RLU values resulting from infection with variant pseudotyped viruses were quantified by a luminescence meter. Data for each plasma sample were obtained from a representative infectivity experiment of three replicates and presented as the mean values ± SEM.

Figure 2

Identification of the neutralizing antibodies (Nabs)-mediated enhancement of SARS-CoV-2 variant infection. (A) Assessment of the ADE activities of 10 potent NAbs against SARS-CoV-2 WT and the D614G variant. Pseudoviruses were preincubated with 250, 500, and 1,000 ng/ml of NAbs, and these mixtures were added to Daudi cells to evaluate their ability to enhance infection. (B) Assessment of ADE activities of the neutralizing antibody 55A8 for SARS-CoV-2 WT and the D614G variant. Pseudoviruses preincubated with serial dilutions of 55A8 mixtures were added to Daudi, Raji, and K562 cells to evaluate their ability to enhance infection. RLU values resulting from infection with variant pseudotyped viruses were quantified by a luminescence meter. Data for each NAb were obtained from a representative infectivity experiment of three replicates and presented as the mean values ± SEM.

Figure 3

ADE activities of the neutralizing antibody 55A8 for the high-infective double-mutant variants. The highly infectious double-mutant variants included D614G+P330S, D614G+F338L, D614G+A348T, D614G+N439K, D614G+T478I, and D614G+H519Q. Pseudoviruses were preincubated with serial dilutions of 55A8 mAbs, and these mixtures were added to Daudi, Raji, and K562 cells to evaluate their ability to enhance infection. RLU values resulting from infection with variant pseudotyped viruses were quantified by a luminescence meter. Data for each NAb were obtained from a representative infectivity experiment of three replicates and presented as the mean values ± SEM.

Figure 4

The ADE activities of 55A8 for the newly emerged variants. New epidemic SARS-CoV-2 variants (including B.1.1.7, B.1.351, B.1.617, and B.1.617.2) were preincubated with serially 55A8, and their mixtures were added to Daudi, Raji, and K562 cells to evaluate their ability to enhance infection. RLU values resulting from infection with variant pseudotyped viruses were quantified by a luminescence meter. Data for each NAb were obtained from a representative infectivity experiment of three replicates and presented as the mean values ± SEM.

Figure 5

The ADE effects mediated by the neutralizing antibody 55A8 depend on FcγRII. (A) Assessment of ADE activities of the neutralizing antibody 55A8 for the epidemic variants (including B.1.1.7 and B.1.351) with or without anti-CD32 (blocking FcɤR II). The SARS-CoV-2 B.1.1.7 and B.1.351 pseudoviruses were preincubated with 125 ng/ml and 250 ng/ml of 55A8, and their mixtures were added to Daudi cells to evaluate their ability to enhance infection. The percentage of control-represented ADE infection was quantified by luminescence meter and normalized to the reference groups without anti-CD32 produced in parallel, and the anti-CD32 was diluted to 40 ng/ml. (B) Assessment of ADE activities of 55A8^LALA for epidemic variants, including B.1.1.7 and B.1.351. Pseudoviruses were preincubated with serially diluted 55A8 or 55A8^LALA, and their mixtures were added to Daudi cells to evaluate their ability to enhance infection. RLU values resulting from infection with variant pseudotyped viruses were quantified by a luminescence meter. Data for each NAb were obtained from a representative infectivity experiment of three replicates, presented as the mean values ± SEM, and p-values were calculated via two-sided Student's t-test. *p < 0.05 vs. the 125 ng/ml of 55A8 group without blocking anti-FcγRII antibody, and ^∧p < 0.05 vs. the 500 ng/ml of 55A8 group without blocking anti-FcγRII antibody.