A potent pan-sarbecovirus neutralizing antibody resilient to epitope diversification

Abstract

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) evolution has resulted in viral escape from clinically authorized monoclonal antibodies (mAbs), creating a need for mAbs that are resilient to epitope diversification. Broadly neutralizing coronavirus mAbs that are sufficiently potent for clinical development and retain activity despite viral evolution remain elusive. We identified a human mAb, designated VIR-7229, which targets the viral receptor-binding motif (RBM) with unprecedented cross-reactivity to all sarbecovirus clades, including non-ACE2-utilizing bat sarbecoviruses, while potently neutralizing SARS-CoV-2 variants since 2019, including the recent EG.5, BA.2.86, and JN.1. VIR-7229 tolerates extraordinary epitope variability, partly attributed to its high binding affinity, receptor molecular mimicry, and interactions with RBM backbone atoms. Consequently, VIR-7229 features a high barrier for selection of escape mutants, which are rare and associated with reduced viral fitness, underscoring its potential to be resilient to future viral evolution. VIR-7229 is a strong candidate to become a next-generation medicine.

Keywords: SARS-CoV-2; broadly neutralizing antibody; monoclonal antibody; sarbecovirus; viral antibody escape.

Figure 1.. VIR-7229 is a potent, pan-sarbecovirus neutralizing mAb.

(A) VIR-7229-mediated neutralization of pseudoviruses. Bar color denotes sarbecovirus clade, as in panel C. Horizontal lines denote cell line employed in neutralization assay: VeroE6 (gray), HEK-293T+human ACE2 (cyan), HEK-293T+bat (R. alcyone) ACE2 (yellow). PRD0038-dm refers to PRD0038 harboring the K482Y/T487W RBD mutations (SARS-CoV-2 numbering 493Y/498W), which allow for entry using human ACE2.^, SARS-CoV-1 Urbani and WIV1 experiments were run in two assay conditions. See also Data S1. See Figure S2 for neutralization mechanisms of action. (B) VIR-7229-mediated neutralization of authentic SARS-CoV-2 virus, performed with VeroE6 cells. The WA1/2020 isolate has the same S haplotype as Wuhan-Hu-1. See also Data S1. (C) Breadth of VIR-7229 and comparator mAbs binding to a yeast-displayed panel of sarbecovirus RBDs spanning the known phylogenetic diversity. Line below the graph, denoted by “ACE2,” indicates whether a sarbecovirus binds or enters cells via human ACE2 (blue), bat but not human ACE2 (yellow), no ACE2 (pink), or unknown (unk.). See Figure 4A for phylogenetic relationships and clade definitions. See Data S1 for full sequences, phylogeny, and alignment. (D) VIR-7229 Fab fragment binding affinity measured by SPR. Bar color denotes sarbecovirus clade. SARS-CoV-1 is Urbani. See also Data S1. (E) Overview of pseudovirus neutralization by comparator mAbs, colored by sarbecovirus clade. Data points within the gray bar represent neutralization not detected (ND), i.e. IC50 >10,000 ng/ml. See Figure S1 and Data S1 for data separated by strain.

Figure 2.. In vivo efficacy of VIR-7229.

Virology and clinical endpoints on day 4 after SARS-CoV-2 XBB.1.5 or JN.1 infection of Syrian hamsters prophylactically administered with VIR-7229 (hamster Fc) or 1.5 mg/kg (mpk) isotype-matched control antibody. See also Data S3. (A) Experiment outline. (B) Infectious viral lung titers for XBB.1.5 infection. ND: not detected. (C) Lung viral RNA load for XBB.1.5 infection. ND: not detected. (D) Variation in body weight relative to day 0 for XBB.1.5 infection. No-infection control from the JN.1 experiment is provided for qualitative comparison. (E) Cumulative clinical score for XBB.1.5 infection (0–4): ruffled fur, slow movements, apathy, absence of exploratory activity. No-infection control from the JN.1 experiment is provided for qualitative comparison. (F) Infectious viral lung titers for JN.1 infection. (G) Lung viral RNA load for JN.1 infection. ND: not detected. (B-G) X-axis indicates dose of VIR-7229-hmFc or 1.5 mpk isotype control. Median ± interquartile range is shown; significance is based on ANOVA non-parametric Kruskal-Wallis test followed by Dunn’s multiple comparison test, *p<0.05, **p< 0.01, *** p < 0.001.

Figure 3.. Structural basis for VIR-7229 pan-sarbecovirus neutralization.

(A) Ribbon diagram of the cryoEM structure of the BA.2.86 S ectodomain trimer (cyan, pink and gold) in complex with two VIR-7229 Fabs (purple and magenta) with N-linked glycans rendered as blue surfaces. See also Figure S3. (B) Ribbon diagram of the VIR-7229-bound XBB.1.5 RBD crystal structure. The bound S309 Fab is omitted for clarity. The N343 glycan is rendered as a blue surface. See also Figure S4. (C) XBB.1.5 RBD (cyan) with VIR-7229 epitope residues shown as sticks and colored according to the (dominant) Fab interacting chain. RBD residues 420, 453, 455, 460 and 493 interact with the VIR-7229 heavy and light chains and were colored based on the chain with which they bury the greatest surface area. (D) Zoomed-in view of select interactions formed between VIR-7229 and the XBB.1.5 RBD. Hydrogen bonds and salt bridges are indicated with black dash lines. (E) Zoomed-in view of hydrogen-bonds (black dash lines) formed between VIR-7229 and the RBD backbone. (F) Superposition of the VIR-7229-bound XBB.1.5 RBD (cyan RBD, dark purple mAb) and VIR-7229-bound EG.5 RBD (pink RBD, light purple mAb) showing accommodation of the F456L residue mutation. (G) Superposition of the VIR-7229-bound EG.5 RBD (pink RBD, dark purple mAb) and S2V29-bound BQ.1.1 RBD (gray RBD, light purple mAb). The two CDRH3 residues differing between S2V29 and VIR-7229 (V50Y and N57D) are highlighted in orange. Select residues from the VIR-7229:EG.5 RBD structure are also shown as semi-transparent surfaces colored according to the sticks. (H) Superposition of the VIR-7229-bound XBB.1.5 RBD (cyan RBD, dark pink mAb) and VIR-7229-bound BA.2.86 S (gold RBD, bright pink mAb) structures highlighting the conservation of electrostatic interactions (dashed lines) at the epitope/paratope interface due to the BA.2.86 R403K mutation. The D52 side chain is weakly resolved in the BA.2.86 S cryoEM density and was therefore not modeled.

Figure 4.. VIR-7229 epitope encompasses sarbecovirus diversity.

(A) Collapsed sarbecovirus phylogeny (left) with multiple sequence alignment of select sarbecoviruses (right) illustrating variation at VIR-7229 epitope positions. RBD numbering is relative to SARS-CoV-2. Dots indicate the SARS-CoV-2 Wuhan-Hu-1 identity. Heatmap at top of alignment illustrates extent of variation (white) or conservation (orange) across the entire sarbecovirus alignment, matched to the structural mapping in panel B. See Data S1 for full phylogeny and alignment. (B) Sarbecovirus conservation of the VIR-7229 epitope mapped to SARS-CoV-2 Wuhan-Hu-1 RBD structure (PDB 6M0J). ACE2 binding footprint is illustrated as a black outline.

Figure 5.. VIR-7229 has high tolerance for SARS-CoV-2 epitope variation.

(A-C) Complete elucidation of mutations in the Wuhan-Hu-1, BA.2, BQ.1.1, XBB.1.5, EG.5, and BA.2.86 RBDs that enable escape from VIR-7229 (A), Omi-42 (B), or SA55 (C) binding using a yeast-display deep mutational scanning method. Letter height is proportional to mutant escape. Mutations are colored by their measured impacts on ACE2-binding affinity, where lighter yellow indicates increasingly deleterious effects on receptor binding (scale bar, bottom-right). See also Figures S5 and S6. (D) Summary of ≥4.0-μs MD simulations of XBB.1.5 RBD bound to VIR-7229 or ACE2. Boxes are the number of persistent contacts at each RBD position in the VIR-7229 epitope, expressed as the fraction occupancy for each VIR-7229 or ACE2 contact across the MD simulation, added together for each RBD position. See panels A or E for RBD position annotations. Slash indicates no contact, i.e. sum of fraction occupancy <0.1. Full glycans were modeled into the RBD:ACE2 X-ray structure; some ACE2 contacts are glycan-mediated, see Figure S4F. The third row indicates RBM residues (gray boxes), defined as RBD:ACE2 protein:protein contacts within 5 Å in the X-ray structure. See also Figure S4 and Data S4. (E) Top, logoplots illustrating the frequency of amino acid variation in VIR-7229 epitope residues across human-ACE2-utilizing sarbecovirus sequences (orange) and SARS-CoV-2 sequences available on GISAID from May 8, 2024 (blue). Bottom, barplot illustrating SARS-CoV-2 mutant frequencies (log scale; residues present in the ancestral Wuhan-Hu-1 sequence are not plotted) for all mutants with >0.005% occurrence in GISAID (up to May 8, 2024). VIR-7229 neutralization of each of these mutations was validated via neutralization of single mutants introduced into XBB.1.5 and JN.1 pseudovirus (panel F) or presence of a mutation in a circulating variant that is neutralized (Figure 1A), with the latter mutations labeled with asterisk. (F) VIR-7229-mediated neutralization of SARS-CoV-2 epitope variants with >0.005% frequency in GISAID (panel E), tested on the XBB.1.5 and JN.1 backgrounds. “Reference” refers to XBB.1.5 or JN.1 with no additional amino acid substitutions. Substitutions are annotated relative to the Wuhan-Hu-1 sequence. R403K is part of the JN.1 reference sequence. See also Data S1. (G) SARS-CoV-2 conservation of the VIR-7229 epitope mapped to SARS-CoV-2 Wuhan-Hu-1 RBD structure (PDB 6M0J). ACE2 binding footprint is illustrated as a black outline.

Figure 6.. VIR-7229 exhibits a high barrier to viral escape.

(A) Serial passaging of Wuhan-Hu-1 and XBB.1.5 rVSV in the presence of mAb did not result in escape from VIR-7229, as defined by ≥20% cytopathic effect in the presence of 20 μg/mL mAb (experiment terminated after 10 and 7 passages, respectively) whereas XBB.1.5 rVSV escaped from comparator mAbs (SA55 and Omi-42) after a single passage. EG.5 rVSV escaped from VIR-7229 after two to three passages, and from a comparator mAb (SA55) after a single passage. XBB.1.5.70 rVSV escaped from VIR-7229 after two passages and from a comparator mAb (SA55) after a single passage. Two independent replicates were performed for each experiment. Figure shows RBD mutations observed after sequencing; the T941K mutation was also observed in one replicate of the XBB.1.5.70 serial passaging with VIR-7229. See also Figure S6 and Data S5. (B) Plaque-based selection of BQ.1.1 and XBB.1 rVSV escapes was performed with VIR-7229 and comparator mAb SA55. Zero escape plaques were observed in 72 independent selections for VIR-7229 whereas 31 and 35 escape plaques, respectively, were observed in 108 independent selections for SA55. Representative images from BQ.1.1 selection are shown, red arrow indicates escape plaque. See also Data S5. (C) Impact of mutations at RBD position 455 on VIR-7229 Fab fragment binding affinity measured by SPR (top) and on VIR-7229-mediated pseudovirus neutralization (bottom). EG.5+L455F is XBB.1.5.70. Colored bars correspond to mutations plotted in panel F. See also Data S1. (D) Impact of the D420N mutation on VIR-7229 Fab fragment binding affinity measured by SPR (top) and on VIR-7229-mediated pseudovirus neutralization (bottom). See also Data S1. (E) Impact of mutations at RBD position 455 on ACE2 affinity measured by SPR. Colored bars as in panel C. See also Data S1. (F) Top – SARS-CoV-2 viral activity level in U.S. wastewater, January 2023 – April 2024 (cdc.gov). Bottom – Frequency of SARS-CoV-2 S mutations as percentage of weekly sequences deposited in GISAID, January 2023 – April 2024. L455W and D420N frequencies are too low to be visible. As of May 8, 2024, >96% of L455F and >86% of L455W mutations co-occur with F456L, primarily in EG.5 and derivative strains; approximately 94% of L455S mutations are in a BA.2.86/JN.1 background. (G) Impact of L455S +/− F456L on VIR-7229 Fab fragment binding affinity measured by SPR (top) and on VIR-7229-mediated pseudovirus neutralization (bottom). See also Data S1. (H) Impact of L455S +/− F456L mutations on ACE2 affinity measured by SPR. See also Data S1.