Enhanced evasion of neutralizing antibody response by Omicron XBB.1.5, CH.1.1, and CA.3.1 variants

Abstract

Omicron subvariants continuingly challenge current vaccination strategies. Here, we demonstrate nearly complete escape of the XBB.1.5, CH.1.1, and CA.3.1 variants from neutralizing antibodies stimulated by three doses of mRNA vaccine or by BA.4/5 wave infection, but neutralization is rescued by a BA.5-containing bivalent booster. CH.1.1 and CA.3.1 show strong immune escape from monoclonal antibody S309. Additionally, XBB.1.5, CH.1.1, and CA.3.1 spike proteins exhibit increased fusogenicity and enhanced processing compared with BA.2. Homology modeling reveals the key roles of G252V and F486P in the neutralization resistance of XBB.1.5, with F486P also enhancing receptor binding. Further, K444T/M and L452R in CH.1.1 and CA.3.1 likely drive escape from class II neutralizing antibodies, whereas R346T and G339H mutations could confer the strong neutralization resistance of these two subvariants to S309-like antibodies. Overall, our results support the need for administration of the bivalent mRNA vaccine and continued surveillance of Omicron subvariants.

Keywords: CA.3.1; CH.1.1; CP: Immunology; CP: Microbiology; Omicron subvariants; XBB.1.5; fusogenicity; immune evasion; mAb S309; neutralizing antibody.

Figure 1.. Distribution and infectivity of emerging Omicron subvariants XBB.1.5, CH.1.1, and CA.3.1

(A) Schematic depiction of the relationships between different Omicron subvariants, with key lineage-defining amino acid mutations for each displayed. (B) Distribution of recently emerged Omicron subvariants in the United States starting in early October of 2022 through the beginning of January 2023. Data were collected from the Centers for Disease Control and Prevention and plotted using Prism software. (C and D) Infectivity of pseudotyped lentiviruses carrying each of the indicated S proteins of the Omicron subvariants were determined in (C) HEK293T cells overexpressing human ACE2 and (D) human lung epithelia-derived CaLu-3 cells. Bars in (C) and (D) represent means ± standard deviation from three biological replicates of one typical experiment. Significance relative to D614G was determined using unpaired two-sided Student’s t tests (n = 3). p values are displayed as ns p > 0.05. The fold change in the mean viral titer of Omicron subvariants was calculated relative to that of D614G.

Figure 2.. Neutralization of Omicron XBB.1.5, CH.1.1, and CA.3.1 subvariants by sera of bivalent or monovalent mRNA vaccinated healthcare workers (HCWs) and BA.4/5 wave infection

Neutralizing antibody titers were determined using lentiviral pseudotypes carrying each of the indicated S proteins of the Omicron subvariants. They were compared against BA.4/5 and/or respective parental Omicron subvariants as specified in the text. The cohorts included sera from 14 HCWs that received 3 monovalent doses of mRNA vaccine plus a dose of bivalent mRNA vaccine (n = 14) (A and D), 15 sera from HCWs that only received three doses of monovalent mRNA vaccine (B and E), and 20 sera from BA.4/5-wave SARS-CoV-2-infected first responders and household contacts that tested positive during the BA.4/5 wave of infection in Columbus, Ohio (C and F). Bars represent geometric means with 95% confidence intervals. Geometric mean NT₅₀ values are displayed for each subvariant on the top. Statistical significance was determined using log₁₀-transformed NT₅₀ values to better approximate normality. Comparisons between multiple groups were made using a one-way ANOVA with Bonferroni post-test. Comparisons between two groups were performed using a paired, two-tailed Student’s t test with Welch’s correction. Dashed lines indicate the threshold of detection (80 for monovalent and bivalent mRNA vaccinees and 40 for BA.4/5 infection cohort). p values are displayed as ns p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Heatmaps in (D)–(F) depict neutralizing antibody titers by each individual against each Omicron subvariant tested. Asterisk in (D) indicates the individual infected by SARSCoV-2 within 6 months before the sera sample collection, and asterisk in (F) indicates the individuals who had received three doses of mRNA vaccine before infection.

Figure 3.. Syncytia formation, cell surface expression, and S processing of Omicron XBB.1.5, CH.1.1, and CA.3.1 subvariants

(A and B) Syncytia-forming activity. HEK293T-ACE2 cells were co-transfected with Omicron subvariant S proteins and GFP and incubated for 30 h before (A) imaging and (B) quantifying syncytia. D614G and no S serve as positive and negative controls, respectively. Comparisons in the extent of syncytia for each variant were made against D614G, with p values indicating statistical significance. Similar results were obtained by using CaLu3 cell as target (see data in Figure S4). (C and D) Cell surface expression of S proteins. HEK293T cells used for production of pseudotyped lentiviral vectors bearing S proteins (Figures 1 and 2) from Omicron subvariants were fixed and surface stained for S with an anti-S1 specific antibody T62 followed by flow cytometric analyses. (C) Histogram plots of anti-S1 signals in transfected cells and (D) calculated relative mean fluorescence intensities of each subvariant by setting the value of D614G as 1. (E) S expression and processing. HEK293T cells used to produce pseudotyped vectors were lysed and probed with anti-S1, anti-S2, anti-GAPDH (loading control), or anti-p24 (HIV capsid, transfection control) antibodies; the signal for anti-GAPDH was from reblotting the membrane of anti-S1, and the signal for anti-S2 was from reblotting the membrane of anti-p24. S processing was quantified using NIH ImageJ used to determine an S1/S or S2/S ratio and normalized to D614G (D614G = 1.0). Bars in (B) and (D) represent means ± standard error. Dots represent three biological replicates from one typical experiment. Significance relative to D614G was determined using a one-way repeated measures ANOVA with Bonferroni’s multiple testing correction (n = 3). p values are displayed as ns p > 0.05 and ****p < 0.0001.

Figure 4.. Homology modeling of key mutations in XBB.1.5, CH.1.1, and CA.3.1

(A) Structures of S receptor-binding domain (RBD)-ACE2 binding interface shown as ribbons. (B) Structure of RBD with class I antibody AZD8895. The recognition focuses on residue F486, with multiple antibody residues forming a surrounding hydrophobic cage, whereas this interaction is abolished by F486S/P mutation. (C) Structures of an immune-dominant region of S N-terminal domain (NTD) with a representative antibody COVOX-159. The nAb recognition on residue G252 is abolished by G252V mutation through creating a steric hindrance (shown as red plates). (D) Residues K444 and L452 are located within acommon epitope site of class II RBD-targeting neutralizing antibodies represented as green surface. (E) Antibody S309 epitope and sequence diversity.(Top) S protein sequence (330–441) with residues of antibody S309 epitope highlighted in green, and mutation hotspots in bold font, (left) amino acid variation at residues 339, 346, and 368 among different Omicron subvariants, (middle) R346T abolishes a salt bridge and a hydrogen bond; and (right) G/D339H interferes with the S309 recognition of glycan-N343 while L368I stabilizes it.