SARS-CoV-2 Fusion Peptide-Directed Antibodies Elicited by Natural Infection Mediate Broad Sarbecovirus Neutralization

Abstract

Studies have demonstrated that repeated mRNA vaccination enhances the breadth of neutralization against diverse SARS-CoV-2 variants. However, the development of antibodies capable of neutralizing across the Coronavirinae subfamily is poorly understood. In this study, we analyze serum samples to determine their neutralization breadth and potency and identify their antigenic targets. Using a cohort of older individuals and healthcare workers, we track correlates of broad neutralizing responses, including fusion peptide (FP) antibody elicitation. We find that although broadly neutralizing responses are often a result of RBD-specific antibodies, a rare subset of donors produce FP-specific broadly neutralizing responses. Interestingly, FP-specific antibodies are not observed in COVID-naive individuals irrespective of vaccination regimen, but rather, they occur following natural infection or vaccine breakthrough. This study highlights the epitope targets underpinning broadly neutralizing antibody responses to coronaviruses and suggests that existing vaccines are insufficient to promote the elicitation of FP-directed broadly neutralizing coronavirus antibodies.

Keywords: COVID-19; SARS-COV-2; broad neutralization; fusion peptide; neutralizing antibodies; stem helix.

Figure 1:. Broad neutralizers have extensive breadth across related Betacoronaviruses.

(A) Cohort of 399 donor samples divided into 4 groups by average neutralization potency (average of n=2 technical replicates) across SARS-CoV-2 WT and the BA.1 and BA.5 variants. Groups were compared using a One-Way ANOVA with Holm-Šídák correction for multiple comparisons using statistical hypothesis testing **** = P<0.0001. (B) 14 of the broadest neutralizers were subject to our high throughput neutralization assay against SARS-CoV-2, variants, SARS-like viruses, and other distantly related coronaviruses with a much higher threshold for neutralizing activity (average of n=2 technical replicates) . These donors were further subdivided into Super Potent and Broad Potent based on average neutralizing activity. (C) Neutralizing activity (average of n=2 technical replicates) of donors averaged across Beta Coronaviruses (excluding MERS). Groups were compared using a two tailed Mann-Whitney T test with P = 0.0007.

Figure 2:. Extra potent sera have the highest levels of RBD, SH and FP Antibodies.

(A) Conservation of the amino acids within the spike protein generated using consurf50 from (PDB 6XR859 https://doi.org/10.2210/pdb6xr8/pdb). The sequences fed into the algorithm include 59 coronavirus spike sequences representing all four clades. RBD = Receptor Binding Domain, NTD = N-Terminal Domain, FP = Fusion Peptide, SH = Stem Helix. (B-E) Results from ELISA assessing the antibody endpoint titers of plasma from 22 extra potent, 13 Broad/Potent, 13 Narrow/Potent, 16 Moderate and 14 Weak donors. Antibody endpoint titers (average of n=2 technical replicates) were assessed against an S2P Stabilized spike (B), Receptor Binding Domain (C), Stem Helix Peptide (D), or Fusion Peptide (E). Antibody endpoint titer represents the maximum interpolated dilution that can still bind the antigen. The endpoint titers were evaluated at a top dilution of 100-fold with seven additional 3-fold serial dilutions. The wells incubated with no plasma incubation were used as controls, and cut-off values (used for interpolation) were defined as 2.1X the OD450 of the controls. Representative data from two independent experiments is shown. Groups were compared using a One Way ANOVA with Kruskal Wallis Test and Dunn’s multiple comparisons test. (* = P<0.0332, ** = P<0.0021, ***= P<0.0002, ****= P<0.0001).

Figure 3:. RBD and fusion peptide specific antibody titers correlate with neutralizing activity.

(A) U/mL of antigen specific antibody was assessed for a cohort of broad and potent neutralizers (average of n=2 technical replicates). All U/mL were calculated using an antigen specific ELISA and were defined as the equivalent reactivity to 1 μg/mL binding of the following antibodies: RBD182 (Spike, RBD), 76E1 (FP), S2P6 (SH). (B) All donors from A were tested for binding to the Stem Helix or Fusion Peptide along with a negative control antibody (2A10) at the same concentration as the serum dilution to determine cutoffs to define background serum binding. (C-F) Log-log regression analyses were performed on neutralization versus anti-Spike IgG (C), anti-RBD IgG (D), anti-SH IgG (E), and anti-FP IgG (F). Pearson correlations were performed and R² and p values are indicated.

Figure 4:. Fusion peptide antibodies contribute to neutralization breadth.

(A) Schematic detailing our antigen depletion protocol. Briefly, his-tag or maleimide binding magnetic beads are pre-coupled to the antigen/peptide and mixed with serum. After an incubation the magnetic beads are pulled away using a magnet and serum is collected for downstream assays. Mock depletion is done by mixing serum and beads with no antigen (Created in BioRender). (B-C) Serum samples with high neutralizing activity were subject to RBD based depletion using his-tagged beads or mock depletion using beads with no RBD bound. Mock and depleted serum was tested for neutralizing activity (average of n=2 technical replicates) against SARS-CoV-2 (C) and SARS-CoV (D). (D-E) Serum samples with remaining neutralizing activity against both SARS-CoV-2 and SARS-CoV from (B-C) were subject to further depletion with the fusion peptide and tested for remaining neutralizing activity (average of n=2 technical replicates) to SARS-CoV-2 (D) and SARS-CoV (E) after double depletion.

Figure 5:. Fusion peptide antibodies develop in response to natural infection.

(A) Anti-Fusion Peptide ELISA for a cohort of COVID+ pre-vaccination timepoints split into non-hospitalized and severity of hospitalization (average of n=2 technical replicates). Limit of detection line is defined as the background binding of a negative control antibody to the FP for that experiment. A Kruskal Wallis one-way ANOVA was performed with Dunn’s correction for multiple comparisons. (B) Anti-Fusion Peptide ELISA (average of n=2 technical replicates) for pre-vaccination serum samples for donors who were naïve or who had been infected with COVID. A Mann Whitney Two tailed T-test was performed to assess significance. (C) Anti-Fusion Peptide ELISA (average of n=2 technical replicates) for post-vaccination serum samples for donors who were naïve after 2–6 vaccines or who were COVID+ at the timepoint and had received between 2–6 vaccines. A Mann Whitney Two tailed T-test was performed to assess significance. (D) Fusion peptide ELISA (average of n=2 technical replicates) was performed for a cohort of COVID naïve donors at various stages of vaccination. No significance was seen between groups in the development of FP specific antibodies. Limit of detection line is defined as the background binding of a negative control antibody to the FP for that experiment. (E) Anti-Fusion Peptide ELISA (average of n=2 technical replicates) for longitudinal timepoints of donors who had a breakthrough infection between 2nd vaccine and bivalent. Pre-Breakthrough (Pre-BT) indicates the time point closest to (1–3 months) before breakthrough infection (BT) timepoint. A Kruskal Wallis one-way ANOVA was performed with Dunn’s correction for multiple comparisons. For A-E U/mL is defined as the reactivity seen by 1 μg/mL of 76E1 antibody. Cutoffs were determined by binding of a negative control antibody (2A10).