Fig 1.
Plasma responses in COVID-19 convalescent donors.
(A)-(C) Area under the curve (AUC) values for anti-SARS-CoV-2 RBD IgG (A), IgM (B) and IgA (C). Severe donors are colored in pink, Mild in light blue, and Unexposed in light gray (named “HD1-15”, samples collected prior to November 2019). Donors’ symbols are indicated on the right side of panel (D). (D) Inhibition of RBD:ACE2 binding by donors’ plasma in ELISA. All samples were normalized to the average value of the healthy donors, that was set as zero inhibition. (E)-(G) Correlation between plasma inhibition and AUC of IgG (E), IgM (F) and IgA (G). AUC and correlations (simple linear regression) were calculated using GraphPad Prism software. Statistical analysis for panels A-D were performed using one-way ANOVA test.
Fig 2.
BCR sequence analysis in convalescent COVID-19 donors.
(A) BCR sequencing information for each donor. The total numbers of sequenced BCRs, the number of identified clones (based on similar VHJH and CDRH3), the number of cells in clones, as well as percent of clonal sequences are given for each donor. (B) Frequencies of the top 34 VHJH combinations in BCRs from 13 COVID-19 donors versus mature BCR from 7 healthy donors (yellow bars) or BCR from naïve B cells from 42 healthy donors (orange bars). -Log10 p values were calculated using Mann Whitney test with FDR correction. Over-represented VHJH combinations are marked in green, while under-represented are marked in red. p values are listed by the corresponding horizontal column. (C) Left panel: shared clone matrix presented as a heatmap and showing the numbers of clones shared between the different donors. CNT–uninfected control. Right panel: the number of shared clones within the Severe group (magenta) and the Mild group (light blue). Statistical analysis was performed using Student’s t test. (D) Comparison between the frequencies of 60 VH genes in COVID-19 infected BCR database while comparing between Severe and Mild groups of donors. Magenta, right pointing horizontal bars indicate VH genes that are over-represented in Severe donors over Mild, while light blue, left pointing horizontal bars indicate VH genes that are over-represented in Mild donors over Severe. Blue dashed lines correspond to p < 0.05 and red dashed lines corresponds to p < 0.01. p values were calculated using Mann Whitney test. The two VH genes that are significantly different between the groups are marked in magenta (E) Volcano plot showing p value as calculated in (D) and normalized fold counts for every VH segment. (F) Comparison between the frequencies of 2500 VH-VL pairs in COVID-19 infected Severe versus Mild groups of donors. Only pairs that are below p < 0.01 are indicated. Color code, configuration and p values are the same as in (E). ** p < 0.01 *** p < 0.001 **** p < 0.0001.
Fig 3.
Monoclonal antibodies cloned from donors CoV01 and CoV02.
(A) A representative flow cytometry gating strategy for staining SARS-CoV-2 RBD-specific memory B cells. (B) The frequencies of anti-SARS-CoV-2 RBD-specific memory B cells in COVID-19 donors CoV01-CoV17. No PBMCs were obtained for donor CoV18, therefore this donor was not included. Symbol code is given on the right. (C) Pie charts representing the total number of RBD-specific memory B cell sequences (heavy chains) obtained from donors CoV01 and CoV02. The numbers in the middle of the pies represent the total number of sequences, and the pink shaded slices represent B cell clonal families. The white slices represent single sequences that were not clonal. (D) Number of nucleotide substitutions in VH of the B cell sequences of CoV01 (red) and CoV02 (magenta). Statistical analysis was calculated using one-way ANOVA. (E) Amino acid length of CDRH3 from B cell sequences isolated from CoV01 (red) and CoV02 (magenta). (F) AUC in ELISA for each one of the 22 cloned mAbs against SARS-CoV-2 RBD (left) and Spike (right) proteins. Color code is given on the right of (H). (G) and (H) Degree of mAbs competition with mAb CR3022, and ACE2, as assessed by ELISA. mGO53 was used as isotype control. (I) Frequency of RBD-PE stained hACE2-expressing cells identified by flow cytometry in the presence of the 22 mAbs isolated from CoV01 and CoV02. mAbs that reduced RBD binding to hACE2 are marked with black arrows. Unlabeled RBD was used as a positive control.
Fig 4.
Neutralization of SARS-CoV-2 by mAbs.
(A) IC50, IC80, and R squared values for neutralization of pseudo-typed GFP-reporter viral particles. Color code is given below. (B) Inhibition of authentic SARS-CoV-2 infection by the 22 mAbs. Virus infected Vero E6 cells were identified by using anti-nucleocapsid antibody (GeneTex GTX135357) conjugated to AF594 after cell fixation and permeabilization. Infected cells were quantified using Image J. (C) Representative images of cells infected with authentic SARS-CoV-2 in the presence of mAbs. mGO53 is shown as a human isotype mAb control [42]. (D) Cell death induction by SARS-CoV-2 infection. Confluent Vero E6 cells were infected with SARS-CoV-2 at MOI:1 in the presence of mAbs in three concentrations as indicated. Viability of cells in five fields of view was monitored with propidium iodide every 6 h for 60 h using an Incucyte S3. (E) Syncytia formation inhibition as measured by the percentage of fusion area between HEK-293 target cells stably expressing hACE2 and HEK-293T cells transiently expressing SARS-CoV-2-Spike (MN908947.3) in the presence of 10 μg/mL of the appropriate antibody. The cells were incubated for 2 h, prior to imaging by the IncuCyte ZOOM system (Essen BioScience) at hourly intervals over 12 h. The percentage area of syncytia was calculated by subtracting the single cell area from the total field area using the following equation: . (F) AUC quantification of (E). (G) Confocal microscopy image of HEK-293T cells transiently expressing SARS-CoV-2-Spike (MN908947.3) and stained with TAU-2212, followed by incubation with FITC-conjugated anti-human secondary antibody. mGO53 serves as an isotype control [42].
Fig 5.
TAU-2230 epitope prediction using phage-displayed affinity purified peptides and RBD mutagenesis.
(A) Left panel: AUC binding in ELISA of the 16 affinity-selected phage-displayed peptides to mAb TAU-2230. Fth-1 serves as a negative control (a phage that does not display an insert peptide). The symbols of the peptides are shown in the right panel. Right panel: amino acid sequences of 16 phage-displayed affinity purified peptides. Common GSLR/K motif is marked in red. The peptides were used as input for the Mapitope algorithm [32]. (B) Predicted TAU-2230 epitope. Left panel: ACE2:RBD complex, PDB ID 6M0J [4]. SARS-CoV-2 RBD is depicted in gray surface representation, while ACE2 is in yellow cartoon representation. The TAU-2230 predicted contact residues are shown in magenta. Middle and right panels: magnified view of the ACE2:RBD interface from two angles, the predicted TAU-2230 epitope is shown in magenta on the surface of the SARS-CoV-2 RBD. The residues that are both ACE2 and predicted TAU-2230 contacts are in red. (C) The effect of RBD point mutations on ACE2bs nAbs binding in ELISA (AUC, y axis). The amino acid substitutions that were introduced in SARS-CoV-2 RBD are indicated on the bottom. RBD wild-type is shown. The color-code for the different nAbs is shown on the right. (D) Heatmap summarizing antibody binding to mutated RBD, as well as wild-type RBD. Values for each construct are normalized to mAb TAU-1109.
Fig 6.
Combination of nAbs targeting different sites are more effective in SARS-CoV-2 neutralization.
Virus-induced cell death traced for 84 h when nAbs were combined with the virus 1 h prior to addition to confluent Vero E6 cells at MOI:1. The antibody concentration for each nAb is 1 μg/ml, and is depicted by circles (A), or 10 μg/ml and is depicted by squares (B). In every panel, the six nAbs are indicated on the top of each graph. In each graph, empty circles or squares correspond to the effect when the indicated nAb was used alone. Full circles or squares correspond to the effect when the indicated nAb was used in combination with TAU-1109 (depicted in purple), with TAU-2212 (depicted in orange), or both (depicted in red). The black bold line shows a “virus only” (no nAb) control.
Fig 7.
VHJH precursors for anti-SARS-CoV-2 nAbs are frequent within COVID-19 infected and uninfected BCR repertoires.
(A) The VJ gene frequencies of anti-SARS-CoV-2 antibodies was calculated and compared to the frequency of all other VJ genes. Frequencies are shown in COVID-19 patients, in healthy matured, as well as in healthy naïve B cells. (B) The table lists the individual frequencies of every nAb IGHV/JH combination in each one of the three BCR datasets (COVID-19 infected, n = 13, COVID-19 un-infected, mature BCR, n = 7, COVID-19 un-infected naïve BCR, n = 42).