Fig 1.
Delta and Beta spikes contain mutations in multiple domains.
(A,B) Mutations in the Delta (A) or Beta (B) spikes relative to Wuhan-Hu-1 [6,26]. Sites where mutations occur in the spike ectodomain are highlighted in dark blue on the Wuhan-Hu-1 one-RBD open spike trimer (PDB 6ZGG) [63]. The surface of one spike monomer is shown; the other two protomers are transparent. Visualization of linear spike sequence modified from https://covdb.stanford.edu/sierra/sars2/by-patterns/. Scissors icon by Mario Verduzco from NounProject.com.
Fig 2.
SARS-CoV-2 infection or vaccination elicits a neutralizing antibody response highly focused on the RBD.
(A) Plasma neutralization (neutralization titer 50%, NT50) against lentiviral particles pseudotyped with the homologous spike of exposure (in parentheses) following either mock depletion or depletion of RBD-binding antibodies for individuals who had primary infections with early 2020, Beta, or Delta viruses (and no prior SARS-CoV-2 infections or vaccinations), or individuals who completed the two-dose series of the mRNA vaccines BNT162b2 or mRNA-1273 (labeled 2x BTN162b2 or 2x mRNA-1273). (B) Plasma neutralization following either mock depletion or depletion of RBD-binding antibodies for individuals who were vaccinated 2x with BNT162b2 or who had Delta breakthrough infections after 2x mRNA vaccination. Plasma samples were depleted of D614G RBD-binding antibodies and tested for neutralization against D614G spike (left), or depleted of Delta RBD-binding antibodies and tested for neutralization against Delta spike (right). In all cases, the RBD and spike from the same viral variant were used in both the depletions and neutralization assays. Note that the same 2x BNT162b2 data are shown in the right panel of (A) and in the left panel of (B) to facilitate visual comparison. The neutralizing titers for the early 2020 plasmas were first published in [23], the Beta plasmas in [12], and the 2x mRNA-1273 plasmas in [29]. The assays for the primary Delta infection, Delta breakthrough infection, and 2x BNT162b2, with both D614G and Delta RBD and spike were newly performed in this study. ELISAs that confirm depletion of RBD-binding antibodies performed in this study are in S1 Fig. The results new to this study are plotted by individual serum or plasma in S2 Fig. Full neutralization curves are shown in S7 Fig.
Fig 3.
Complete antibody-escape maps for plasmas from individuals convalescent of primary Delta infections.
(A) The surface of the Wuhan-Hu-1 RBD is colored according to antibody epitope per the Barnes classification scheme [37], with the L452R and T478K mutations in the Delta variant highlighted in yellow (and labeled with red text in panel B). We define L452R to be in the class 3 epitope, and T478K in the class 1 epitope. (B) Complete maps of mutations that reduce binding of plasma antibodies to the Delta RBD from individuals convalescent of primary Delta infection. Sites of strong antibody escape (see Methods) for any of the 8 plasmas are highlighted with pink in the line plots at left and shown in the logo plots at right. Sites 452, 477, and 501 are included despite not being sites of strong escape due to their high frequency in circulating viral isolates. Site 408 is included to facilitate comparison to Delta breakthrough infection-elicited plasmas (Fig 5). Interactive versions of logo plots and structural visualizations are at https://jbloomlab.github.io/SARS-CoV-2-RBD_Delta/. Correlations between independent library replicates are in S5 Fig, and all escape scores are in S3 Data and online at https://github.com/jbloomlab/SARS-CoV-2-RBD_Delta/blob/main/results/supp_data/aggregate_raw_data.csv.
Fig 4.
Antibody-escape maps for individuals infected with early 2020 viruses, Beta, or Delta.
(A) Site-total antibody escape averaged across each group of plasmas for early 2020, Beta, and Delta convalescent plasmas, with key epitope regions highlighted (key in panel C). (B) Site-total antibody escape averaged across each group mapped to the Wuhan-Hu-1 RBD surface (PDB 6M0J, [64]), with red indicating the site with the most antibody escape and white indicating sites with no escape. Key antibody-escape sites are labeled, with site labels colored according to epitope. (C) Multidimensional scaling projection with the distance between points representing the dissimilarity between antibody-escape maps for individual early 2020, primary Beta, and primary Delta infection convalescent plasmas, with monoclonal antibody escape maps of the 4 classes used as anchors. In all panels, the convalescent plasmas were mapped against RBD mutant libraries corresponding to the viral variant of exposure. The antibody-escape maps for early 2020 infections were first reported in [23,24], the Beta infections in [12], and the monoclonal antibodies in [22,24,42–44]. The antibody-escape maps for primary Delta infections are new to this study. The complete antibody-escape scores are in S3 Data and online at https://github.com/jbloomlab/SARS-CoV-2-RBD_Delta/blob/main/results/supp_data/aggregate_raw_data.csv.
Fig 5.
Delta breakthrough infection after 2x mRNA vaccination elicits an antibody response focused on the class 1 and 2 epitopes.
(A) Complete maps of mutations that reduce binding of plasma antibodies to the Delta RBD from individuals convalescent of Delta breakthrough infection. The same sites as in Fig 3B are shown in the logo plots and highlighted in pink in the line plots. Logo plots for both cohorts are shown side-by-side in S6 Fig. (B) Site-total antibody escape averaged across each group of plasmas for primary Delta convalescent plasmas, Delta-breakthrough convalescent plasmas, and previously published mRNA-1273-vaccinated sera [29], with key epitope regions highlighted (key in panel D). (C) Site-total antibody escape averaged across each group mapped to the RBD surface, with red indicating the site with the most antibody escape and white indicating sites with no escape. The Delta primary infection and Delta breakthrough plasmas were mapped against the Delta RBD libraries, and the mRNA-1273 sera were mapped against the Wuhan-Hu-1 RBD libraries. Site labels are colored by epitope. (D) Multidimensional scaling projection with the distance between points representing the dissimilarity between antibody-escape maps, comparing primary Delta infections, Delta breakthrough after 2x mRNA vaccination convalescent plasmas, and mRNA-1273-vaccinated sera with monoclonal antibody escape maps of the 4 classes used as anchors, first reported in [22,24,42–44]. The antibody-escape maps for mRNA-1273 vaccine-elicited sera were first reported in [29], and primary Delta and Delta breakthrough infections are new to this study. The primary Delta infection escape maps are replicated from Fig 4. Interactive versions of logo plots and structural visualizations are at https://jbloomlab.github.io/SARS-CoV-2-RBD_Delta/. The complete antibody-escape scores are in S3 Data and online at https://github.com/jbloomlab/SARS-CoV-2-RBD_Delta/blob/main/results/supp_data/aggregate_raw_data.csv.
Fig 6.
Effect of mutations on neutralization by antibodies elicited by 2x BNT162b2 vaccination, Delta breakthrough after 2x mRNA vaccination, or primary Delta infection.
(A) Neutralization titer 50% (NT50) of spike-pseudotyped lentiviral particles by sera or plasmas from individuals with various exposure histories. The leftmost spike is D614G, which was circulating in 2020. The point mutations and depletion of RBD-binding antibodies are in the indicated spike background (D614G or Delta). Bars represent geometric mean. Dashed line indicates assay limit of detection. (B) Fold-decrease in neutralization by sera or plasmas from individuals with various exposure histories against K417N (left) or E484K (right) in the Delta and D614G spikes, compared to the homologous spike (dashed line at 1). Points represent the geometric mean across cohorts (NT50s plotted in (A)). Error bars represent geometric standard deviation. The neutralization assays with 2x BNT162b2, Delta breakthrough, and Delta primary infection were newly performed in this study, and assays with 2x mRNA-1273 and early 2020 plasmas were performed in [12,29] and are reanalyzed here. Full neutralization curves are in S7 Fig and numerical NT50s are in S1 Data and online at https://github.com/jbloomlab/SARS-CoV-2-RBD_Delta/blob/main/experimental_data/results/neut_titers/combined_neut_titers.csv.