Identification and Analysis of Monoclonal Antibodies with Neutralizing Activity against Diverse SARS-CoV-2 Variants

Abstract

We identified neutralizing monoclonal antibodies against severe acute respiratory syndrome-coronavirus 2 (SARS-CoV-2) variants (including Omicron variants BA.5 and BA.2.75) from individuals who received two doses of mRNA vaccination after they had been infected with the D614G virus. We named them MO1, MO2, and MO3. Among them, MO1 showed particularly high neutralizing activity against authentic variants: D614G, Delta, BA.1, BA.1.1, BA.2, BA.2.75, and BA.5. Furthermore, MO1 suppressed BA.5 infection in hamsters. A structural analysis revealed that MO1 binds to the conserved epitope of seven variants, including Omicron variants BA.5 and BA.2.75, in the receptor-binding domain of the spike protein. MO1 targets an epitope conserved among Omicron variants BA.1, BA.2, and BA.5 in a unique binding mode. Our findings confirm that D614G-derived vaccination can induce neutralizing antibodies that recognize the epitopes conserved among the SARS-CoV-2 variants. IMPORTANCE Omicron variants of SARS-CoV-2 acquired escape ability from host immunity and authorized antibody therapeutics and thereby have been spreading worldwide. We reported that patients infected with an early SARS-CoV-2 variant, D614G, and who received subsequent two-dose mRNA vaccination have high neutralizing antibody titer against Omicron lineages. It was speculated that the patients have neutralizing antibodies broadly effective against SARS-CoV-2 variants by targeting common epitopes. Here, we explored human monoclonal antibodies from B cells of the patients. One of the monoclonal antibodies, named MO1, showed high potency against broad SARS-CoV-2 variants including BA.2.75 and BA.5 variants. The results prove that monoclonal antibodies that have common neutralizing epitopes among several Omicrons were produced in patients infected with D614G and who received mRNA vaccination.

Keywords: Omicron variants; broad neutralizing activity; common epitope; cryoelectron microscopy; human monoclonal antibody; receptor-binding domain; severe acute respiratory syndrome-coronavirus 2 (SARS-CoV-2); spike; vaccine.

FIG 1

Flow chart of experimental procedures for the neutralizing antibody analysis in this study. (1) PBMCs isolated from three patients who showed high neutralizing activity for Omicron BA.1 (Table 1), were selected, mixed, and used for the study. (2) Memory B cells were isolated from the PBMCs, sorted by fluorescently labeled spike antigen, and then dispensed as single cells. (3) Antibody variable region genes amplified from each single cell were used to express fragment antigen binding (Fab) according to the protocol of the Ecobody technology (9). Each Fab was tested for reactivity against spike antigens by ELISA, and the top 10 candidates were selected based on the results. (4) The antibody variable region genes were used to construct recombinant IgG antibody expression plasmids. Each MAb secreted from culture cells transfected by the plasmids was purified for subsequent analysis. (5) 10 MAbs were subjected to the neutralization assay against the SARS-CoV-2 Omicron BA.1, and were labeled MO1 to MO10 according to the potency. Only three antibodies (MO1, MO2, and MO3) showed neutralizing activity against BA.1; therefore, these 3 antibodies were further analyzed. (6) Reactivities of MO1, MO2, and MO3 to spike antigen from several variants were evaluated by ELISA. (7) The neutralizing activities of MO1, MO2, and MO3 were quantitatively evaluated by a plaque reduction neutralizing test. (8) The affinity of MO1 and MO2 to the spike receptor binding domain (RBD) was analyzed by biolayer interferometry assay. (9) The best neutralizing antibody, MO1, was subjected to cryo-EM to analyze the binding mode to the spike antigen. (10) Spike residues at the MO1-binding site were examined for their importance by ELISA using mutant spike proteins. (11) Animal challenge experiment was performed to evaluate the neutralizing activities of MO1, MO2, and MO3 against Omicron BA.5 in a hamster infection model.

FIG 2

Identification of broadly neutralizing MAbs against SARS-CoV-2 variants. (A) The binding of three MAbs to the SARS-CoV-2 spike ectodomains of the D614G, Delta, BA.1, BA.2, BA.5, and BA.2.75 variants as revealed by respective ELISAs. (B) The neutralizing activity of MAbs MO1, MO2, and MO3 against D614G, Delta, BA.1, BA.1.1, BA.2, BA.5, BA.2.75, BQ.1.1, or XBB.1 as evaluated by the plaque reduction neutralization test (PRNT). (C) The 50% inhibitory concentrations (IC₅₀) of MAbs MO1, MO2, and MO3 against the SARS-CoV-2 variants calculated from the above neutralization data (B) are shown. (D) A presentation of plaque reduction in MO1’s PRNT test against BA.5.

FIG 3

Analysis of the affinity between the three MAbs and spike antigens by biolayer interferometry (BLI). (A) The sensorgram for the BA.2 spike RBD’s binding to the MAb MO1 or MO2. Dashed lines: the fitting curves. (B) The same BLI analysis as in panel A between the BA.5 spike RBD and MO1. (C) Summary of the BLI kinetics evaluated from the curve fitting. (D) Competition between MAbs and human ACE. A nonneutralizing antibody MO7 was used as a competition-negative control.

FIG 4

The binding mode of MAb MO1 with the BA.1 spike trimer. (A and B) The cryo-EM density map (A) and of the atomic model (B) of the MO1 Fab and prefusion BA.1 spike trimer. The coloring is as follows. MO1 heavy chain, orange; light chain, yellow; spike RBD, cyan; spike NTD, and S2, green. (C) Map of the Omicron BA.1 or BA.2 mutation sites on the RBD viewed from the MO1 binding site. The Omicron BA.5 mutations F486V and L452R, and Omicron BA.1.1 mutation R346K are also indicated. (D) The residues involved in the MO1 footprint are yellow. The view is the same as that in panel C. (E) The MO1 binding mode on the RBD. The left image is the same view as in panel C and D. (F to I) Notable interactions between MO1 and the RBD at the interface. Three-dimensional arrangements of the protein residues (top) and their schematic illustrations (bottom) are shown.

FIG 5

MO1 recognized R346 and N448 as key epitopes. (A) Binding of MO1 to T345, R346, K440, D442, K444, V445, N448, N450, and Y451 was analyzed by ELISA using site-specific alanine mutations in the BA.1 spike ectodomain. (B) Summarized results of site-specific alanine mutations in the BA.1 spike ectodomain.

FIG 6

(A) Schematic representation of the evaluation of the toxicity and efficacy of monoclonal antibodies in vivo. (B) The weight changes of hamsters after MO1 administration (n = 5) or mock treatment (n = 5). The averages of weight changes were plotted with symbols, and error bars represent SDs. Two-way ANOVA analysis was performed for statistical analysis (n.s.: not significant). (C) The weight changes after challenge with SARS-CoV-2 omicron subvariant BA.5 to hamsters administered monoclonal antibodies (n = 10). The average weight changes are plotted with symbols, and error bars represent SDs. Two-way ANOVA analysis was performed for statistical analysis to compare to mock-treated hamsters (n = 5, identical to the data shown in Fig. 6B; n.s., not significant; ***, P < 0.001; ****, P < 0.0001). (D and E) Infectious virus titers in lung homogenates (D) and nasal wash specimens (E) at 4 days postinfection. The individual viral titers are represented with symbols. Medians are shown with bars and 95% CIs are indicated with error bars. The limit of detections was 101.5 PFU/g (D) and 101.3 PFU/mL (E), respectively. One-way ANOVA analysis was used for statistical analysis after logarithmic conversion, and Tukey’s test was used as a post hoc analysis (n.s., not significant; *, P < 0.05; **, P < 0.01; ****, P < 0.0001).