Enhanced variant neutralization through glycan masking of SARS-CoV-2 XBB1.5 RBD

Abstract

SARS-CoV-2 continues to evolve antigenically under the immune pressure exerted by both natural infection and vaccination. As new variants emerge, we face the recurring challenge of updating vaccines at significant financial cost to maintain their efficacy. To address this, novel strategies are needed to enhance the breadth of protection offered by vaccines or, at a minimum, extend their effectiveness over time. One such strategy is antigen modification. In this study, we introduce a glycosylation site into a binding but non-neutralizing epitope within the SARS-CoV-2 XBB.1.5 receptor binding domain (RBD) to redirect the immune response towards more potent neutralizing epitopes. Immunization of mice with this modified antigen via the mRNA vaccine platform resulted in a dramatic increase in neutralizing antibodies compared to the wild-type XBB.1.5 RBD, showing superior protection against a range of SARS-CoV-2 Omicron variants, from BA.2 to JN.1. Our findings reinforce the power of the glycan masking approach, which in combination with the now well-established mRNA vaccine platform can contribute to broader and better vaccines.

Keywords: Glycan masking; SARS-CoV-2; neutralising antibodies; next generation vaccines; receptor binding domain.

Figure 1.

(A) Immunization schematic showing dates of immunizations and bleeds (PB, pre-bleed, BL1, bleed 1, BL2, bleed 2, TB, terminal bleed. IM1/IM2, immunization 1 or 2). (B, C) cartoon diagram of the SARS-CoV-2 XBB.1.5 RBD in complex with ACE-2 (pink). Glycosylation sites shown in green for XBB.1.5 (B) and the glycan masked construct (C). (D) Visualization of the Log₁₀IC₅₀ values obtained through pseudotype virus microneutralization assay with guinea pig serum after two immunizations with mRNA vaccines. The constructs against the different variants is given using box-and-whisker plots. A regression line (method = LOESS) with shaded confidence interval was added. The colours for the graph indicate the three constructs: Vehicle/naïve (black), XBB.1.5/WT (pink), XBB.1.5_M7 (blue). Mann–Whitney U-tests were performed to determine significance (p < 0.05 indicated by a single asterisk (*)). The graph was produced using the ggplot2 package in R. (E) A scatterplot of the Log₁₀IC₅₀ values and percentage identity between the variant RBDs and the different constructs is given. The different variants are indicated by distinct symbols as given by the legend. XBB.1.5/WT is given in pink, while XBB.1.5_M7 is given in blue. The different rectangles indicate the Log₁₀IC₅₀ values for both constructs for BA.2.86 (purple), JN.1 (red), and XBB.1.5 (black). A dashed dark yellow line indicates the separation between pre-Omicron variants and subsequent Omicron variants. The ellipse indicates the 95% confidence level. The graph was produced using the ggplot2 package in R. (F) A regression tree indicating the Log₁₀IC₅₀ value for the constructs when accounting for percentage identity. The regression tree model was produced using the R package rpart, while the plot was created using rpart.plot and DiagrammeR. The rules for the regression tree are as follows: [Table: see text]