Fig 1.
LALA-PG substitutions eliminate antibody binding to mouse FcγRI, II, III, and IV for both DH1052 and DH1050.1, without altering binding to SARS-CoV-2 Spike.
(A) Antibody engineering schematic depicting wildtype (allotype G1m17) versus Fc-function knockout antibodies (LALA- PG substitutions: L234A, L235A, P329G). NTD-directed non-nAb DH1052 and nAb DH1050.1 are produced in both versions. Color scheme for each antibody is the same throughout A-J. (B) ELISA binding of G1m17 and LALA-PG antibodies to their cognate antigen SARS-CoV-2 Spike_D614G versus negative control antigen HIV-1 envelope. Binding response is measured as area under the log transformed curve (logAUC). Serum from a nonhuman primate vaccinated with NTD was used as the positive control, and CH65 was used as the negative control antibody. (C-F) DH1052 G1m17 versus LALA-PG binding to immobilized mouse FcγRI, II, III, and IV measured via surface plasmon resonance (SPR). (G-J) DH1050.1 G1m17 versus LALA-PG binding to mouse immobilized FcγRI, II, III, and IV measured by SPR.
Fig 2.
NTD antibodies with enhanced FcγR binding show increased antibody-dependent cellular cytotoxicity (ADCC).
(A) Antibody engineering schematic depicting wildtype (allotype G1m17) versus Fc-function enhanced antibodies (DLE, sometimes called DLE3, substitutions: S239D, A330L, I332E). (B-E) DH1052 G1m17 and DLE or (F-I) DH1050.1 G1m17 and DLE binding to immobilized mouse FcγRI, II, III, and IV measured by SPR. Antibodies were titrated to determine binding kinetics. Kinetic measurements are reported underneath the respective graph as an average of two independent experiments. Fast off-rates precluded the calculation of binding kinetics for FcγR II and III. (J) Natural killer (NK) cell-mediated antibody-dependent cellular cytotoxicity of 293T cells expressing SARS-CoV-2 WA-1. Titers are shown as % NK cells expressing the degranulation marker CD107a. Each antibody was tested at 2, 8, and 32 μg/mL and is shown ordered in the graph from lowest to highest concentration.
Fig 3.
Modification of sites 211–214 in the Spike NTD eliminates binding by NTD non-nAbs.
(A) Three-dimensional reconstruction by negative stain electron microscopy of DH1052 Fab (orange) in complex with SARS-CoV-2 Spike 2P (gray) (Electron microscopy data bank accession number: EMD-44644). In the enlargement, the density corresponding to the spike has been rigidly fit with a spike model (PDB 7QUS) shown in ribbon diagram. The Spike model is colored red where the NTD loops are most proximal to the Fab. NTD loops proximal to the putative DH1052 antigen combining site (red) were mutated as shown in B. (B) Sequence modifications for each mutant NTD tested. (C) ELISA binding results of a neutralizing NTD antibody panel (DH1048-DH1051, left) and a non-neutralizing NTD antibody panel (DH1052-DH1056, right) against mutant antigen candidates. NTD_ADEm3 combined the three mutations of NTD_ADEm1(a-c). Serum from a nonhuman primate vaccinated with NTD was used as the positive control, anti-influenza antibody CH65 was used as the negative control antibody, and HIV Env was used as the negative control antigen.
Fig 4.
LALA-PG substitutions eliminate or severely attenuate antibody-dependent cellular phagocytosis (ADCP).
(A-C) ADCP activity against NTD and mutant NTD_ADEm3 for G1m17, LALA-PG and DLE versions of DH1052 was determined in a THP-1 cell-based assay. Values shown are the mean of two technical replicates. ADCP score is a ratio of the fluorescence of the test result to the no antibody control (PBS). (D) A comparison of ADCP activity of WT NTD for all versions of DH1052 is shown. (E-G) ADCP of NTD and NTD_ADEm3 by G1m17, LALA-PG and DLE versions of DH1050.1 are shown. (H) A comparison of ADCP of Wuhan-Hu-1 NTD for all DH1050.1 versions is shown.
Fig 5.
Fc knockout substitutions eliminated NTD non-nAb protection, and Fc enhancement increased NTD non-nAb protection.
(A) Study design. BALB/c mice were passively infused at -12 hours and challenged with SARS-CoV-2 MA10 virus at 0 hours. Weights for each animal were collected each day of the experiment and lung tissue was harvested four days post- infection (DPI) to measure infectious viral titers and assign the gross lung discoloration (GLD) score. N = 10 mice per test group; n = 25 isotype control mice; n = 5 uninfected mice. (B) Percent weight loss in mice administered DH1052 G1m17, DH1050.1 G1m17, or Isotype control antibody. Uninfected mice were included as negative controls for weight loss. (C-E) Non-nAb DH1052 and (F-H) nAb DH1050.1 protection against infection and disease. Protection was assessed by (C,F) lung viral titers quantified as the log(PFU/mL), (D,G) weight loss each day post-infection (DPI) expressed as % original weight, and (E,H) median gross lung discoloration scores. All statistical comparisons were calculated using exact Wilcoxon rank sum tests using an alpha level of 0.05 (* p<0.05; ** p<0.01; *** p<0.001).
Fig 6.
A distinct proinflammatory, antiviral cytokine response was elicited in mice after passive immunization with Fc knockout LALA-PG antibodies and challenge.
Clarified lung homogenates obtained from mice day 4 post-infection were analyzed via Luminex multiplex assay for concentrations of 26 selected cytokines as labeled. The resulting concentrations (normalized for total homogenate protein) were compared to normalized concentrations in infected isotype control mice via Wilcoxon test. The fold change and p-value of each comparison between NTD antibody test group and isotype control are shown. (A) Comparisons between isotype control and the wildtype and Fc knockout DH1052 antibody versions. (B) Comparisons between isotype control and the wildtype and Fc knockout DH1050.1 antibody versions. (C) Overlay of all test groups as compared to isotype control. (D) Summary heatmap of mean normalized cytokine concentrations measured for each NTD antibody test group, infected isotype control, and the uninfected group of mice.
Fig 7.
NTD non-nAb therapeutic administration reduces infectious viral burden and lung discoloration in an Fc-dependent manner but does not prevent weight loss.
(A) Study design. BALB/c mice were challenged with SARS-CoV-2 MA10 virus at -12 hours and passively infused with antibody at 0 hours. Weight loss, infectious virus in the lung, and gross lung discoloration (GLD) were assessed for each animal four days post infection as in Fig 5. N = 9–10 mice per test group; n = 5 isotype control mice; n = 5 uninfected mice. (B) Percent weight loss in mice therapeutically administered DH1052 G1m17, DH1050.1 G1m17, or Isotype control antibody. Uninfected mice were included as negative controls for weight loss. (C-E) Non-nAb DH1052 and (F-H) nAb DH1050.1 amelioration of infection and disease. Therapeutic benefit was assessed by (C,F) lung viral titers quantified as the log(PFU/mL), (D,G) weight loss each day post-infection (DPI) expressed as % original weight, and (E,H) median gross lung discoloration scores. All statistical comparisons were calculated using exact Wilcoxon rank sum tests using an alpha level of 0.05 (* p<0.05; ** p<0.01; *** p<0.001).