A multiplexed, systems-based approach for prediction of antibody neutralization breadth for soluble human receptors

Figure 1.. Multiplex binding inhibition of SARS-CoV-2 spike variants and the human receptor ACE2. Binding inhibition curve fits for pre-COVID-19 (dark gray), vaccine-immune (blue), and hybrid-immune (red) serum samples to an array of SARS-CoV-2 spike variant trimers; Ebola GP was used as a negative ACE2 binding control. Each colored line represents the binding inhibition of spike to human ACE2 at the indicated serum dilution (x axis). The y axis represents the % inhibition. Individual samples were done in technical replicates, and mean values for each dilution were used to generate the inhibition curve. The color legend is shown at the top. All analytes were measured simultaneously.

Figure 2.. Quantified systems-predicted neutralizing antibody (SNAb) titers to SARS-CoV-2 spikes show exposure and waning sensitive dynamics. (A) Reciprocal serum dilutions for 50% ACE2 binding inhibition (SNAb titer) were quantified for the SARS-CoV-2 WT spike trimer. Shown on the x axis are the vaccine doses received and the timeframe after the most recent vaccination. Shown on the y axis is the reciprocal SNAb titer in log10. Each circle represents the SNAb titer of an individual participant (mean of technical replicates); in blue are vaccine-only participants, in red are hybrid participants, and in dark gray is pre-COVID-19 serum. The median of each group at each time point is shown in a horizontal black bar. (B–K) Same as (A), but for the indicated SARS-CoV-2 variants (all trimers). (L) Same as (A), but for the negative control Ebola GP. Shown on the right is the color legend.

Figure 3.. Comparison of SNAb and pseudovirus NT50 assays. (A) Correlation between SNAb titers and reported NT50 titers for SARS-CoV-2 WT spike. Pseudovirus NT50 values were correlated with the calculated SNAb titers for vaccine- and hybrid-immune specimens for WT spike; pre-COVID-19 immune serum was not included in this correlation as neither assay had any detectable titers to WT spike. The Pearson correlation coefficient (R) and corresponding Bonferroni-corrected P value for the 2 neutralization outputs are shown in the upper left. On the x axis is the SNAb WT spike titer, and on the y axis is the NT50 WT spike titer; all datapoints for each immune group are shown independent of the time since immunization. Shown is the best-fit correlation line (solid blue line) along with the 95% CIs (shaded gray region). (B–E) Same as (A), but correlating SNAb and NT50 values for SARS-CoV-2 Alpha spike variant (B), Beta spike variant (C), Delta spike variant (D), and Omicron BA.1 spike variant (E). (F) Bland–Altman method agreement analysis for NT50 and SNAb titers to WT spike. Shown on the x axis is the mean between NT50 and SNAb titers for individual participants, colored by immune type. Shown on the…

Figure 4.. Systems serology binding assays to spike variants and controls. (A) Antibody binding levels to the indicated antigens from human serum samples of pre-COVID-19 pandemic (dark gray), vaccine-immune (blue column), and hybrid-immune (red). Each group was arranged in increasing pseudovirus NT50 values to WT spike (top row). All binding levels were log10 transformed and each combination of isotype and antigen were independently z-scored to standardize across analytes. The order of targets is shown on the left callout boxes; the antibody feature is shown on the right. Each column represents one participant. (B) SNAb levels for the groups were similarly shown for each immune group. The order of SNAb targets is shown on the left callout box. All SNAb levels were standardized through z-scoring. Color legend is shown on the top, and is the same as (A). All specimens were arranged in the same order as (A), and each column represents one participant.

Figure 5.. Identification of antibody features correlating with WT spike SNAb titers through systems serology. (A) Pearson correlation coefficients (R) for all antibody features against WT spike SNAb titers were plotted using a volcano plot for 3-dose vaccine recipients 114 days after the third dose. Shown on the x axis is the R value to WT spike SNAb titers, ranging from −1 to 1. Shown on the y-axis is the FDR-adjusted P value in -log10. Thresholds for significance were an |R| ≥ 0.75 and an FDR P < 0.05 (dashed blue lines). All features passing these thresholds were highlighted and labeled; all other features were left gray and not labeled. Three antigen encounters, either all by vaccine or hybrid immunity, followed by a 2-dose mRNA vaccine series, were used to build the correlation. (B) Same as (A), but for 3-dose vaccine recipients 148 days after the third dose. Other time points could not be assessed due to low n or imbalance in antigen encounters. (C) Same as (A), but for hybrid-immune participants 105 days after receiving their second mRNA vaccine dose (3 total antigen encounters). (D) Same as (C), but for hybrid-immune participants 223 days after receiving their second mRNA vaccine…

Figure 6.. Identification of antibody features correlating with composite SNAb titers through systems serology. (A) Pearson correlation coefficients (R) for all antibody features against composite spike SNAb titers were plotted using a volcano plot for 3-dose vaccine recipients 114 days after the third dose. Shown on the x axis is the R value to composite spike SNAb titers (sum of all spike SNAb titers), ranging from −1 to 1. Shown on the y axis is the FDR-adjusted P value in -log10. Thresholds for significance were an |R| ≥ 0.75 and an FDR P value <0.05 (dashed blue lines). All features passing these thresholds were highlighted and labeled; all other features were left gray and not labeled. Three antigen encounters, either all by vaccine or hybrid immunity followed by a 2-dose mRNA vaccine series, were used to build the correlation. (B) Same as (A), but for 3-dose vaccine recipients 148 days after the third dose to composite SNAb titers. Other time points could not be assessed due to low n or imbalance in antigen encounters. (C) Same as (A), but for hybrid-immune participants 105 days after receiving their second mRNA vaccine dose to composite SNAb titers (3 total antigen encounters). (D)…