Structural basis of SARS-CoV-2 Omicron immune evasion and receptor engagement

The SARS-CoV-2 Omicron variant of concern evades antibody-mediated immunity that comes from vaccination or infection with earlier variants due to accumulation of numerous spike mutations. To understand the Omicron antigenic shift, we determined cryo-electron microscopy and X-ray crystal structures of the spike protein and the receptor-binding domain bound to the broadly neutralizing sarbecovirus monoclonal antibody (mAb) S309 (the parent mAb of sotrovimab) and to the human ACE2 receptor. We provide a blueprint for understanding the marked reduction of binding of other therapeutic mAbs that leads to dampened neutralizing activity. Remodeling of interactions between the Omicron receptor-binding domain and human ACE2 likely explains the enhanced affinity for the host receptor relative to the ancestral virus.

Recombinant protein production SARS-CoV-2 RBD proteins for SPR binding assays (residues 328-531 of S protein from GenBank NC_045512.2with N-terminal signal peptide and C-terminal thrombin cleavage site-TwinStrep-8xHis-tag) were expressed in Expi293F (Thermo Fisher Scientific) cells at 37°C and 8% CO2.Transfections were performed using the ExpiFectamine 293 Transfection Kit (Thermo Fisher Scientific).Cell culture supernatants were collected three days after transfection and supplemented with 10x PBS to a final concentration of 2.5x PBS (342.5 mM NaCl, 6.75 mM KCl and 29.75 mM phosphates).RBDs were purified using cobalt-based immobilized metal affinity chromatography followed by buffer exchange into PBS using a HiPrep 26/10 desalting column (Cytiva) for Wuhan-Hu-1 protein, or a Superdex 200 Increase 10/300 GL column (Cytiva) for Omicron protein.SARS-CoV-2 Omicron RBD for crystallization (residues 328-531, with N-terminal signal peptide and 'ETGT', and C-terminal 8xHis-tag) was expressed similarly as described above in the presence of 10 µM kifunensine.Cell culture supernatant was collected four days after transfection and supplemented with 10x PBS to a final concentration of 2.5x PBS.Protein was purified using a HiTrap TALON crude cartridge followed by size exclusion chromatography using a Superdex 200 Increase 10/300 GL column (Cytiva).ACE2 for crystallization (residues 19-615 from Uniprot Q9BYF1 with a C-terminal thrombin cleavage site-TwinStrep-10xHis-GGG-tag, and N-terminal signal peptide) was expressed in ExpiCHO cells in the presence of 10 µM kifunensine at 37°C and 8% CO2.Transfection was performed using the ExpiCHO transfection kit (Thermo Fisher Scientific).Cell culture supernatant was collected eight days after transfection and supplemented to a final concentration of 80 mM Tris-HCl pH 8.0, 100 mM NaCl, and then incubated with BioLock (IBA GmbH) solution.hACE2 was purified using a 5 mL StrepTrap HP column (Cytiva) followed by size exclusion chromatography using a Superdex 200 Increase 10/300 GL column (Cytiva) pre-equilibrated in PBS.

SPR binding measurements of IgG
SPR binding measurements were performed using a Biacore T200 instrument with a CM5 sensor chip covalently immobilized with Cytiva Human Antibody Capture Kit (RBD binding) or Cytiva His Capture Kit (S binding).Running buffer was Cytiva HBS-EP+ (pH 7.4).All measurements were performed at 25 °C.RBD analyte concentrations were 3.1, 12.5, and 50 nM, run as singlecycle kinetics.IgG analyte concentrations for S binding experiments were 11, 33, 100, and 300 nM, run as single-cycle kinetics.For RBD binding, double reference-subtracted data were fit using Biacore T200 Evaluation software (version 3.1) to a 1:1 binding model except two datasets (LY-CoV016 and REGN10987 binding to Wuhan-Hu-1 RBD) were fit to a Heterogeneous Ligand model ("biphasic fit") due to an artefactual kinetic phase with very slow dissociation that often arises when RBD is an analyte.For the Heterogeneous Ligand fits, the lower affinity of the two KD values reported by the fit is taken as the KD (the two KD values are separated by at least two orders of magnitude).Binding of the Omicron RBD to mAbs has been carried out once (n=1) except for S309 which has been performed three times (n=3).Wuhan-Hu-1 RBD binding data are representative of at least two (n=2) for S309, CT-P59, LY-CoV555, LY-CoV16, REGN10987 and REGN10933 mAbs and once for all others (n=1).S binding data were evaluated qualitatively, rather than quantitatively, due to the avidity of the bivalent analyte precluding the use of a 1:1 binding model, and due to the anticipated differential impact of avidity on binding in the SPR experiment as compared to the biological context (e.g. S density and orientation on the CM5 matrix of the chip surface likely does not recapitulate S density and orientation on the surface of a virus or infected cell).

Crystallization, data collection, structure determination, and analysis
Prior to forming the SARS-CoV-2 Omicron RBD-ACE2-S304-S309 complex, recombinant SARS-CoV-2 Omicron RBD was digested with EndoH (New England Biolabs, 21 units/µg RBD) at 4°C overnight.Recombinant hACE2 protein was digested using EndoH (New England Biolabs, 25 units/µg ACE2) and thrombin (Sigma-Aldrich, 1 unit/75 µg ACE2) at 4°C overnight.RBD was mixed with a 1.2-fold molar excess of EndoH deglycosylated hACE2, and a 1.1-fold molar excess of S304 Fab and S309 Fab.The complex was purified on a Superdex 200 10/300 GL column preequilibrated with 20 mM Tris-HCl pH 7.5, 150 mM NaCl.Crystals of the SARS-CoV-2 Omicron RBD-hACE2-S304-S309 complex were obtained at 20°C by sitting drop vapor diffusion.A total of 200 nL of the complex at 6.5 mg/mL were mixed with 200 nL mother liquor solution containing 0.1 M NDSB-256, 20% v/v ethylene glycol, 10% w/v PEG 8000, and 0.1 M Tris (base)/bicine pH 8.5.Data were collected at the Molecular Biology Consortium beamline 4.2.2 at the Advanced Light Source synchrotron facility in Berkeley, CA and processed with the XDS software package yielding a final dataset of 2.85 Å in space group P21.The SARS-CoV-2 Omicron RBD-hACE2-S304-S309 complex structure was solved by molecular replacement using Phaser (76) from starting models consisting of RBD-S304-S309 (PDB: 7JX3) and hACE2 (PDB: 6m0j).Several subsequent rounds of model building and refinement were performed using Coot (67), ISOLDE (71), Refmac5 (77), Phenix (70) and MOE (https://www.chemcomp.com),to arrive at a final model of the quaternary complex.a Measured dissociation is beyond the limit of detection.Reporting KD as an upper bound b Rmax fixed to Rmax from Wuhan-Hu-1 fit (without constraint, Rmax from Omicron fit was implausibly much lower than Rmax for corresponding Wuhan-Hu-1 binding) c Values for the kinetic phase attributed to real binding (vs an artifact) are highlighted in bold

Figure S1 .
Figure S1.CryoEM data processing and validation.(A) Representative electron micrograph (left, scale bar: 100 nm) and 2D class averages (right) are shown for the indicated particles embedded in vitreous ice.(B-F) Gold-standard Fourier shell correlation curves with the 0.143 cutoff indicated by a horizontal blue line (top) and unsharpened maps colored by local resolution calculated using cryoSPARC (bottom) for whole reconstructions (B, C, and D) and the locally refined reconstructions of NTD-or RBD-bound Fab variable domains (E and F).

Figure S2 .
Figure S2.Density resolving the Omicron S mutations.The sharpened cryoEM map (NTD and

Figure S3 .
Figure S3.Comparison of Omicron and Wuhan-Hu-1 NTDs.Omicron mutated residues shown in red as sticks.The glycans are shown as dark blue surfaces.

Fig
Fig S4.SARS-CoV-2 Omicron mutations promote escape from clinical mAb binding in the context of the full S ectodomain trimer.Binding of clinical-stage IgG to surface-immobilized S ectodomain, either Wuhan-Hu-1 (gray line) or Omicron (red line), was evaluated using singlecycle kinetics surface plasmon resonance (SPR) measurements.White and gray stripes are association and dissociation phases, respectively.Data reflect binding avidity due to the bivalent IgG in the mobile phase, and due to the required high density of S attached to the sensor chip matrix.

Figure S5 .
Figure S5.Structural basis for S309 binding to the Omicron, Kappa and Wuhan-Hu-1 RBD.Zoomed-in view of the cryoEM structure of the S309 Fab fragment (black/grey) bound to the Omicron, Kappa (2) or Wuhan-Hu-1 (12) RBD (blue ribbon) with Omicron mutated residues shown in red as sticks.The N343 glycan is shown as sticks.

Fig S6 .
Fig S6.Comparison of the region comprising residues 366 to 375 across structures, and their impact on S2X35 binding.(A-D) Zoomed-in view of the Omicron RBD from the cryoEM structure (A) and crystal structure (B), as well as the Wuhan-1 RBD (C, PDB 6m0j) (42), and Wuhan-1 RBD bound to S2X35 (D, PDB 7r6w) (24).(E) Overlay of the RBD from (B) with S2X35 from (D) showing clashing between S373P and K55 (asterisk).(F) Binding of the Wuhan-Hu-1 (gray line) or Omicron (red line) RBD to S2X35 mAb was evaluated using surface