Fig 1.
Antibody neutralization activity and breadth against SARS-CoV-2.
A, Illustration of the architectures of Fab, IgG-14, and IgM-14. B, Construction of spike variants. L, 5’ lead sequence; ORF, open reading frame; E, Envelope; M, Membrane; N, nucleocapsid; UTR: untranslated region; mNG, mNeonGreen. C, Summary of neutralization activities of IgM-14 and IgG-14 against naturally emerged variants. *, data from our previous study. #, Fold changes in EC50 relative to the parental USA-WA1/2020 strain. D, Scatter plot comparing EC50 fold changes of IgM-14 and IgG-14 across SARS-CoV-2 variants, normalized to the parental strain. Each point represents a variant. Blue, variants carrying the E484K/Q mutation; gray, variants without the E484K/Q mutation; orange, Omicron strains. E, Scheme of testing IgM-14’s antiviral activity in human airway epithelial (HAE). F-K, antiviral efficacy of IgM-14 against SARS-CoV-2 in HAE cultures. Fluorescent microscopy of HAE infected with mNG BA.1-spike (F) or Delta-spike (I) variants after IgM-14 treatment. Intracellular viral RNA on day 3 after mNG BA.1-spike (G) or Delta-spike (J) infection. Extracellular viral RNA on day 3 after mNG BA.1-spike (H) or Delta-spike (K) infection. Geometric means with 95% confidence intervals (CI) are shown in panels G-H and J-K. One-way ANOVA with Dunn’s multiple comparison corrections was used for statistical analysis. ***, p < 0.001; ****, p < 0.0001.
Fig 2.
Characterization of IgM-14-resistant variants.
A, Diagram of the process for selecting IgM-14-resistance in VeroE6 cells. Four independent selections (SL1-4) started at passage 1 (P1) and were passaged 6 times. SL4 was aborted at the second passage. B, Spike mutations occurred in P6. EC50s of IgM-14 against P6 viruses (SL1-3) are shown. *, intact and deletion sequences observed in the selection. C, Construction of mNG SARS-CoV-2 with spike mutations. D, Neutralization curves of IgM-14 and IgG-14 against WT or mutant mNG SARS-CoV-2. E, Summary of the EC50 values derived from panel D. F, Plaque morphologies of WT or mutant mNG SARS-CoV-2 on VeroE6 cells. Growth kinetics of G476D (G) or F486S (H) versus WT mNG SARS-CoV-2 in VeroE6, A549-hACE2, and human airway epithelial (HAE) cultures. Two-way ANOVA with Šidák’s multiple comparison corrections was used for statistical analysis. **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. I, Relative expression of RBD mutants to the WT RBD. Error bars indicate variation across three independent experiments. J, BLI analysis of the binding of WT and mutant RBDs to human ACE2. Association rate (kon), dissociation rate (koff), affinity constant (KD), and R2 values of curve fitting are indicated. K, ELISA analysis of IgM-14 and IgG-14 binding to WT (solid circles), G476D (red circles), and F486S (blue circles) RBDs. Error bars indicate variation across three independent experiments.
Fig 3.
Five modes of Fab-14 binding to D614G spike trimer.
A, Cryo-EM density of the Fab-14 and spike trimer complex in Mode I. The primary and secondary binding sites are indicated by orange and red circles, respectively. The down RBD involved in the primary binding site is shown in yellow, while the down RBD involved in the secondary binding site is shown in blue. The heavy and light chains of Fab-14 are colored pale cyan and light gray, respectively. B, Cryo-EM density of the Fab-14 and spike trimer complex in Mode II. The two down RBDs are shown in yellow and blue, respectively. The up RBD is shown in green. C, Cryo-EM density of the Fab-14 and spike trimer complex in Mode III. D, Cryo-EM density of the Fab-14 and spike trimer complex in Mode IV. The up RBDs are shown in yellow and green, respectively. The down RBD is shown in blue. E, Cryo-EM density of the Fab-14 and spike trimer complex in Mode V. F, local refined structures of primary and secondary binding sites in the bipartite binding mode. The paperclip motif in the primary binding site is shown in orange. The loop 331-335 in the secondary binding site is shown in dark blue.
Fig 4.
Structural analysis of the primary binding site.
A, overview of the primary binding site. B, Zoomed-in view of the interaction between HCDR2 and the RBD paperclip motif. Pink dashed lines indicate salt bridges, purple dashed lines show hydrogen bonds. C, Zoomed-in view of the interaction between Fab-14 light chain and RBD residues G476-P479. D, Superposition of Fab-14 and ACE-2 on the same up RBD. The surface of ACE2 and Fab-14 are shown in pink and blue, respectively. E, MM/PBSA analysis of changes in RBD/Fab-14 complex binding free energy changes caused by individual mutations. Data shows the mean ± standard deviations. F, Sequence alignment of the primary binding site between the parental SARS-CoV-2 (USA-WA1/2020) and Omicron variants. The GISAIDs of BA.1, BA.2, BA.3, and JN.1 spikes were EPI_ISL_6640916, EPI_ISL_6795834.2, EPI_ISL_7605591, and EPI_ISL _18237538, respectively.
Fig 5.
Analysis of the secondary binding site.
A, Structural analysis of the secondary binding site. B, MM/PBSA analysis of changes in RBD/Fab-14 complex binding free energy changes caused by indicated mutations. C-D, Neutralization curves of IgG-14 (C) or IgM-14 (D) against mNG SARS-CoV-2 D614G (reference) and mNG SARS-CoV-2 D614G+N331Q + L335A (N331Q+L335A). E-F, Neutralization curves of IgG-14 (E) or IgM-14 (F) against mNG SARS-CoV-2 D614G+E484K (E484K) and mNG SARS-CoV-2 D614G+E484K + N331Q+L335A (E484K+N331Q + L335A). Means and standard deviation from three independent experiments are shown in panels C-F. G, Summary of EC50 values of IgG-14 or IgM-14 against indicated SARS-CoV-2 mutants. H, A model of IgM-14 complexed with SARS-CoV-2 spike trimer shows different modes of spike-IgM-14 interactions. The coordinates of IgM were taken from a published report41 by Chen, Q. et al. and adjusted by rigid body movement of the Fab arms while preserving the integrity of the Fab tethers.