Extremely potent human monoclonal antibodies from COVID-19 convalescent patients

Abstract

Human monoclonal antibodies are safe, preventive, and therapeutic tools that can be rapidly developed to help restore the massive health and economic disruption caused by the coronavirus disease 2019 (COVID-19) pandemic. By single-cell sorting 4,277 SARS-CoV-2 spike protein-specific memory B cells from 14 COVID-19 survivors, 453 neutralizing antibodies were identified. The most potent neutralizing antibodies recognized the spike protein receptor-binding domain, followed in potency by antibodies that recognize the S1 domain, the spike protein trimer, and the S2 subunit. Only 1.4% of them neutralized the authentic virus with a potency of 1-10 ng/mL. The most potent monoclonal antibody, engineered to reduce the risk of antibody-dependent enhancement and prolong half-life, neutralized the authentic wild-type virus and emerging variants containing D614G, E484K, and N501Y substitutions. Prophylactic and therapeutic efficacy in the hamster model was observed at 0.25 and 4 mg/kg respectively in absence of Fc functions.

Keywords: COVID-19; Fc functions absence; SARS-CoV-2; emerging variants; monoclonal antibodies; prophylaxis; therapy.

Graphical abstract

Figure 1

Workflow and timeline for SARS-CoV-2 neutralizing antibodies identification The overall scheme shows three different phases for the identification of SARS-CoV-2 neutralizing antibodies (nAbs). Phase 1 consisted in the enrolment of COVID-19 patients (n = 14) from which PBMCs were isolated. Memory B cells were single-cell sorted (n = 4,277), and after 2 weeks of incubation, antibodies were screened for their binding specificity against the S protein trimer and S1/S2 domains. Once S protein-specific monoclonal antibodies (mAbs) were identified (n = 1,731) phase 2 started. All specific mAbs were tested in vitro to evaluate their neutralization activity against the authentic SARS-CoV-2 virus, and 453 nAbs were identified. nAbs showing different binding profiles on the S protein surface were selected for further functional characterization and to identify different neutralizing regions on the antigen. Phase 3 starts with the characterization of the heavy and light chain sequences of selected mAbs (n = 14) and the engineering of the Fc portion of three most promising candidates. The latter were also selected for structural analyses that allowed the identification of the neutralizing epitopes on the S protein. Finally, the most potent antibody was tested for its prophylactic and therapeutic effect in a golden Syrian hamster model of SARS-CoV-2 infection.

Figure S1

Gating strategy for single-cell sorting and monoclonal antibodies screening for S protein S1 + S2 subunits binding and neutralization of binding (NoB) activity, related to Figure 2 (A) Starting from top left to the right panel, the gating strategy shows: Live/Dead; Morphology; CD19⁺ B cells; CD19⁺CD27⁺IgD^-; CD19⁺CD27⁺IgD^-IgM^-; CD19⁺CD27⁺IgD^-IgM^-S-protein⁺ B cells. (B) The graph shows supernatants tested for binding to the SARS-CoV-2 S-protein S1 + S2 subunits. Threshold of positivity has been set as two times the value of the blank (dotted line). Darker dots represent mAbs which bind to the S1 + S2 while light yellow dots represent mAbs which do not bind. (B) The graph shows supernatants tested by NoB assay. Threshold of positivity has been set as 50% of binding neutralization (dotted line). Dark blue dots represent mAbs able to neutralize the binding between SARS-CoV-2 and receptors on Vero E6 cells, while light blue dots represent non-neutralizing mAbs.

Figure 2

Identification of SARS-CoV-2 S protein-specific nAbs (A) The graph shows supernatants tested for binding to the SARS-CoV-2 S-protein stabilized in its prefusion conformation. Threshold of positivity has been set as two times the value of the blank (dotted line). Red dots represent mAbs that bind to the S protein, while pink dots represent mAbs that do not bind. (B) The bar graph shows the percentage of non-neutralizing (gray), partially neutralizing (pale yellow), and neutralizing antibodies (dark red) identified per each donor. The total number (n) of antibodies tested per individual is shown on top of each bar. (C) The graph shows the neutralization potency of each nAb tested once expressed as recombinant full-length IgG1. Dashed lines show different ranges of neutralization potency (500, 100, and 10 ng/mL). Dots were colored based on their neutralization potency and were classified as weakly neutralizing (>500 ng/mL; pale orange), medium neutralizing (100–500 ng/mL; orange), highly neutralizing (10–100 ng/mL; dark orange), and extremely neutralizing (1–10 ng/mL; dark red). The total number (n) of antibodies tested per individual is shown on top of each graph. A COVID-19 convalescent plasma and an unrelated plasma were used as positive and negative control, respectively, in all the assays.

Figure S2

Characterization and distribution of SARS-CoV-2 S protein-specific nAbs, related to Figure 2 (A) The bar graph shows the distribution of nAbs binding to different S-protein domains. In dark red, light blue and gray are shown antibodies binding to the S1-domain, S2-domain and S-protein trimer respectively. The total number (n) of antibodies tested per individual is shown on top of each bar. (B) The bar graph shows the distribution of nAbs with different neutralization potencies. nAbs were classified as weakly neutralizing (> 500 ng/mL; pale orange), medium neutralizing (100 – 500 ng/mL; orange), highly neutralizing (10 – 100 ng/mL; dark orange) and extremely neutralizing (1 – 10 ng/mL; dark red). The total number (n) of antibodies tested per individual is shown on top of each bar.

Figure 3

Functional characterization of potent SARS-CoV-2 S protein-specific nAbs (A–C) Graphs show binding curves to the S protein in its trimeric conformation, S1 domain, and S2 domain. Mean ± SD of technical triplicates are shown. Dashed lines represent the threshold of positivity. (D–F) Neutralization curves for selected antibodies were shown as percentage of viral neutralization against the authentic SARS-CoV-2 wild type (D), D614G variant (E), and the emerging variant B.1.1.7 (F). Data are representative of technical triplicates. A neutralizing COVID-19 convalescent plasma and an unrelated plasma were used as positive and negative control, respectively. (G–I) Neutralization potency of 14 selected antibodies against the authentic SARS-CoV-2 wild type (G), D614G variant (H), and the emerging variant B.1.1.7 (I). Dashed lines show different ranges of neutralization potency (500, 100, and 10 ng/mL). In all graphs, selected antibodies are shown in dark red, pink, gray, and light blue based on their ability to recognize the SARS-CoV-2 S1 RBD, S1 domain, S protein trimer only, and S2 domain, respectively.

Figure S3

Binding to S protein receptor binding domain (RBD) and NoB activity of S1-RBD antibodies, related to Figure 3 (A) Histograms show the ability of selected antibodies to bind the S-protein RBD. Gray histograms represent the negative control while colored histograms show tested antibodies. Percentage of positive and negative populations are denoted on each graph. (B) Neutralization of binding (NoB) curves for S1-RBD specific antibodies are shown as percentage of reduction of signal emitted by a fluorescently labled S-protein incubated with Vero E6 cells. Mean ± SD of technical duplicates are shown. Dashed lines represent the threshold of positivity; A neutralizing COVID-19 convalescent plasma and an unrelated plasma were used as positive and negative control, respectively.

Figure S4

Binding kinetics of SARS-CoV-2 nAbs to the S protein antigen, related to Figure 3 Representative binding curves of selected antibodies to SARS-CoV-2 S-protein trimer. Different curve colors define the spike concentration used in the experiment. Kon, Koff and KD are denoted on each graph.

Figure S5

Neutralization activity of selected nAbs against SARS-CoV-2, SARS-CoV, and MERS-CoV pseudotypes, related to Figure 3 (A–D) Graphs show the neutralizing activities of 14 selected nAbs with different SARS-CoV-2 S-protein binding profiles against SARS-CoV-2, SARS-CoV-2 D614G, SARS-CoV and MERS-CoV pseudotypes respectively. Dashed lines represent the threshold of positivity. Mean ± SD of technical duplicates are shown. In all graphs selected antibodies are shown in dark red, pink, gray and light blue based on their ability to recognize the SARS-CoV-2 S1-RBD, S1-domain, S-protein trimer only and S2-domain respectively.

Figure 4

Identification of four different sites of pathogen vulnerability on the S protein surface (A) Representative cytometer peaks per each of the four antibody groups are shown. Positive (beads conjugated with only primary labeled antibody) and negative (un-conjugated beads) controls are shown as green and red peaks, respectively. Competing and not-competing nAbs are shown in blue and gray peaks, respectively. (B) The heatmap shows the competition matrix observed among the 14 nAbs tested. Threshold of competition was set at 50% of fluorescent signal reduction. A speculative representation of the vulnerability sites is shown on the S protein surface.

Figure 5

Heavy and light chain analyses of selected nAbs (A and B) Bar graphs show the heavy and light chains usage for neutralizing antibodies against SARS-CoV-2 in the public repertoire compared to the antibodies identified in this study. Our and public antibodies are shown in dark and light colors, respectively. (C and D) The heavy and light chain percentage of identity to the inferred germline and amino acidic CDR3 length are shown as violin and distribution plot, respectively. (E) The heatmap shows the frequency of heavy and light chain pairing for SARS-CoV-2 neutralizing human mAbs already published. The number within the heatmap cells represent the amount of nAbs described in this manuscript showing already published (colored cells) or novel heavy and light chain rearrangements (blank cells).

Figure S6

Characterization of Fc-engineered candidate nAbs, related to Figure 7 (A) the graph shows binding curves of J08, I14 and F05 MUT and WT to the FcγR2A. (B and C) graphs show binding curves of J08, I14 and F05 MUT and WT to the FcRn at pH 6.2 (B) and 7.4 (C). (D and E) Graphs show the ADNP and ADNK induced by J08, I14 and F05 MUT and WT versions; all the experiments were run as technical duplicates. In every experiment a control antibody (CR3022) and an unrelated protein were used as positive and negative control respectively. (F–H) Graphs show binding curves to the S-protein in its trimeric conformation, S1-domain and S2-domain. Mean of technical triplicates are shown. (I–K) Neutralization curves against the authentic SARS-CoV-2 wild type, the D614G variant and the B.1.1.7 emerging variant for J08-MUT, I14-MUT and F05-MUT shown in blue, green and red respectively. Data are representative of technical triplicates.

Figure S7

Autoreactivity assessment of selected SARS-CoV-2 candidate nAbs, related to Figure 7 (A) Schematic representation of the indirect immunofluorescent assay for the screening of autoreactive nAb. (B) Single figures show the fluorescent signal detected per each sample tested in this assay. Positive and negative controls were used at three different dilutions (1:1, 1:10 and 1:100). Three candidate nAbs were incubated on HEp-2 cells at a concentration of 100 μg/mL. Representative pictures of the scoring system are shown. Autoreactive samples are highlighted in pink. 250 nm scale bar is shown.

Figure 6

EM epitope mapping of RBD mAbs (A) Negative stain for J08, I14, and F05 in complex with the S protein. 200 nm scale bar is shown. (B) Figures show the binding of J08 (blue), I14 (green), and F05 (red) to the SARS-CoV-2 S protein RBD.

Figure 7

Prophylactic and therapeutic efficacy of J08-MUT in the golden Syrian hamster model of SARS-CoV-2 infection (A) Schematic representation and timelines of prophylactic and therapeutic studies performed in golden Syrian hamster. (B and C) The figure shows the prophylactic impact of J08-MUT at three different concentrations (4, 1, and 0.25 mg/kg) (B) on body weight loss change (C). The figure shows the therapeutic impact of J08-MUT at 4 mg/kg on body weight loss change. Mean ± SD are denoted in the graphs. (D–F) The figures show the lung viral titer at day 3 (D) and the detection of human antibodies in hamster sera at day 3 (E) and day 8 (F) in the prophylactic study. Mean ± SD of technical triplicates are shown. (G–I) The figures show the lung viral titer at day 3 (G) and the detection of human antibodies in hamster sera at day 3 (H) and day 12 (I) in the therapeutic study. Mean ± SD of technical triplicates are shown. Statistical differences were calculated with two-way analysis of variance (ANOVA) for body weight change and with a nonparametric Mann–Whitney t test for the lung viral titer. Significances are shown as ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, and ^∗∗∗∗p < 0.0001.