The effect of spike mutations on SARS-CoV-2 neutralization

Abstract

Multiple severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccines show protective efficacy, which is most likely mediated by neutralizing antibodies recognizing the viral entry protein, spike. Because new SARS-CoV-2 variants are emerging rapidly, as exemplified by the B.1.1.7, B.1.351, and P.1 lineages, it is critical to understand whether antibody responses induced by infection with the original SARS-CoV-2 virus or current vaccines remain effective. In this study, we evaluate neutralization of a series of mutated spike pseudotypes based on divergence from SARS-CoV and then compare neutralization of the B.1.1.7 spike pseudotype and individual mutations. Spike-specific monoclonal antibody neutralization is reduced dramatically; in contrast, polyclonal antibodies from individuals infected in early 2020 remain active against most mutated spike pseudotypes, but potency is reduced in a minority of samples. This work highlights that changes in SARS-CoV-2 spike can alter neutralization sensitivity and underlines the need for effective real-time monitoring of emerging mutations and their effect on vaccine efficacy.

Keywords: B.1.1.7; SARS-CoV-2; antibodies; immune escape; neutralization; serology; variant.

Graphical abstract

Figure 1

Mutating amino acids in SARS-CoV-2 spike to match SARS-CoV decreases mAb neutralization (A) The indicated mAbs were assessed by pseudotyped neutralization assay. Data are representative of three independent repeats. The horizontal dotted line in each graph indicates 50% neutralization. (B) IC₅₀ values for each mAb against the mutant SARS-CoV-2 pseudotyped viruses indicated in the left column. The previously established binding cluster and binding to RBD for each mAb are indicated above each column. See also Figure S1.

Figure 2

Neutralization by serum is affected less adversely by SARS-CoV amino acid substitutions in SARS-CoV-2 spike (A) Representation of the SARS CoV-2 spike trimer (blue) in complex with ACE-2 (pink) (PDB: 7DF4). The magnified image shows mutated amino acid side chains at residues of interest. (B) Thirty-six serum samples were assessed by pseudotyped neutralization assay. Average ID₅₀ values for 3 independent repeats are linked by horizontal bars for each individual sample. (C) Fold decrease in IC₅₀ values for each mAb against each mutant pseudotype relative to the SARS-CoV-2 wild-type pseudotype. Competitive binding clusters of each mAb that loses more than 3-fold neutralization activity are labeled. (D) The y axis shows the fold decrease in ID₅₀ values for each serum sample against each mutant pseudotype relative to the SARS-CoV-2 wild-type pseudotype; the group of affected individuals is indicated above each graph. (C and D) The dotted horizontal lines indicate a 3-fold drop in neutralization potency.

Figure 3

Serum responses following severe COVID-19 have greater polyclonality but less efficient neutralization (A) The y axis shows spike S1 subunit semiquantitative titers measured by ELISA (STAR Methods). (B) The y axis shows ID₅₀ values measured by pseudotyped neutralization assay. (C and D) ID₅₀ values for serum samples versus the corresponding S1 IgG binding titer. The relative ranking of neutralization titers is indicated in the graph. (E) Concentrations of S1-specific serum IgG (picograms) at ID₅₀ dilutions were calculated using the IgG titers quantified via semiquantitative ELISA. Data for (A), (B), and (E) were analyzed by a non-parametric Mann-Whitney U test; ∗∗∗∗p < 0.05. Data were measured in duplicate. Mild and severe illness groups are defined in STAR Methods. See also Figure S2.

Figure 4

Variant B.1.1.7 SARS-CoV-2 spike effect on mAb and serum neutralization (A) The indicated mAbs were assessed by pseudotype neutralization assay. Data are representative of three independent repeats. The horizontal dotted line in each graph indicates 50% neutralization. (B) Thirty-six serum samples (mild illness, left; severe illness, right) were assessed by pseudotype neutralization assay. ID₅₀ values are linked by horizontal bars for each individual sample. (C) Fold decrease in average ID₅₀ values from 3 repeats for each serum sample against each mutant pseudotype versus D614G. The dotted horizontal line indicates a 3-fold drop in neutralization potency.