Antibody Targeting of Conserved Sites of Vulnerability on the SARS-CoV-2 Spike Receptor-Binding Domain

SUMMARY

Given the continuous emergence of Severe Acute Respiratory Syndrome-Coronavirus 2 (SARS-CoV-2) variants of concern (VoC), immunotherapeutics that target conserved epitopes on the spike (S) glycoprotein have therapeutic advantages. Here we report the crystal structure of the SARS-CoV-2 S receptor-binding-domain (RBD) at 1.95 Å and describe flexibility and distinct conformations of the angiotensin converting enzyme 2 (ACE2) binding site. We identify a set of SARS-CoV-2-reactive monoclonal antibodies (mAbs) with broad RBD cross-reactivity including SARS-CoV-2 Omicron subvariants, SARS-CoV-1, and other sarbecoviruses, and determine the crystal structures of mAb-RBD complexes with Ab246 and CR3022 mAbs targeting the class IV site, WRAIR-2134, which binds the recently designated class V epitope, and WRAIR-2123, the class I ACE2 binding site. The broad reactivity of class IV and V mAbs to conserved regions of SARS-CoV-2 VoCs and other sarbecovirus provides a framework for long-term immunotherapeutic development strategies.

Graphical Abstract

eTOC Blurb

Sankhala et al. report the apo structure of the SARS-CoV-2 receptor binding domain, wherein the flexible ACE2 binding ridge adopts two distinct conformations. Crystal structures of the RBD in complex with antibodies that target conserved RBD epitopes and neutralize the recent Omicron XBB.1.5 variant are also described.

INTRODUCTION

SARS-CoV-2, the causative agent of coronavirus disease-2019 (COVID-19), is a member of the genus Betacoronavirus (BetaCoV), subgenus sarbecovirus lineage B. The emergence of SARS-CoV-2 marks the seventh coronavirus (CoV) to be isolated from humans, and the third following SARS-CoV-1 and MERS-CoV to cause severe disease¹. Despite the rapid development of effective vaccines, vaccination does not confer indefinite sterilizing immunity against SARS-CoV-2 infection. Within select populations including older adults, immunocompromised individuals, or individuals with pre-existing conditions, breakthrough infections can develop into life-threatening disease^2–4. In particular, the ongoing evolution and emergence of SARS-CoV-2 variants of concern (VoC) or variants of interest (VoI) raise serious considerations about the effectiveness of vaccines and existing monoclonal antibody (mAbs) therapies^5–12. Major VoCs reported so far include Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1), Delta (B.1.617.2), Epsilon (B.1.427 and B.1.429), and the more recently identified Omicron subvariants (BA.1, BA.2, BA.2.12.1, BA.2.75, BA.4, BA.5, BQ.1.1, XBB.1.5) which caused a worldwide surge in COVID-19 cases^13–16. The Omicron variant was first detected in November 2021 and is the most heavily mutated VoC to emerge, with >30 mutations in the spike (S) protein, 15 of which occur in the receptor binding domain (RBD)^17,18. In addition, there are 3 deletions and one 3-residue insertion in the N-terminal domain (NTD)¹⁸. Rapid and high transmissibility of Omicron, Delta, and other VoCs warrants the need for further development of broadly protective countermeasures. Consequently, there has been enhanced focus on developing immunotherapeutic strategies that can target one or more conserved epitopes on S^19–25.

To this end, a comprehensive structural knowledge of S, RBD, and their conformational states is of paramount significance and highlights the need to explore novel S and/or RBD conformations and sites of vulnerability. The S glycoprotein of SARS-CoV-2 binds the host receptor, angiotensin converting enzyme 2 (ACE2), to mediate cell entry^26,27. The S glycoprotein is a class I viral fusion protein, the critical determinant of viral host range and tissue tropism and the primary target of neutralizing immune responses²⁸. As such, most CoV vaccine candidates are based on S or its sub-components. Since the emergence of SARS-CoV-2, an unprecedented number of cryo-electron microscopy (cryo-EM) and crystal structures of the prefusion stabilized SARS-CoV-2 S glycoprotein and its receptor binding domain (RBD) have been reported^21,29–37, which provide significant information enabling structure-based vaccine design and therapeutic development.

Here, we report the high-resolution crystal structure of SARS-CoV-2 RBD wherein the ACE2-binding loop (Glu484–Asn487) on the RBD adopts two distinct conformations. To our knowledge, this is the highest resolution structure of an unliganded SARS-CoV-2 RBD to date. Additionally, we present crystal structures of a set of broadly reactive mAbs in complex with SARS-CoV-2 RBD to highlight sites of cross-reactivity and vulnerability. A better understanding of SARS-CoV-2 RBD structure, and definition of new antibody epitope specificities and functional characteristics of cross-reactive antibodies will further aid the development of antiviral countermeasures for SARS-CoV-2 and potential future CoV pandemic pathogens.

RESULTS

High resolution structure of the SARS-CoV-2 RBD reveals ACE2-ridge flexibility

The SARS-CoV-2 RBD (residues 331–532), with a C-terminal His-tag, was expressed in 293F cells and purified by NiNTA affinity and size-exclusion chromatography. Crystals diffracted to < 1.8 Å in space group P4₃2₁2 with a complete dataset to 1.95 Å (Table 1). The structure was refined to an R_free of 20.2% and R_work of 17.4% with no Ramachandran outliers. RBD residues 333–527 were clearly interpretable from the electron density map, with two conformations visible of residues Glu484–Asn487 (Figure 1A). Clear electron density for both conformations was observed with nearly equivalent occupancy but a slight preference for Form-2 (Figure 1B). Refinement of both conformations simultaneously was accomplished by assigning the entire RBD with either ACE2-binding loop Form-1 or −2 as alternate conformations A or B and refining by occupancy. The presence of dual conformations in this region is indicative of inherent structural flexibility. Consistent with this observation, higher B-factors for the loop region (Glu484 - Asn487) in comparison to the inner core of the protein were present (Figure 1C). RMSF (root mean square fluctuations) analysis using CABS flex 2.0 also demonstrated significantly higher fluctuations in this region compared to other regions of the structure (Figure S1A)³⁸. To understand this loop flexibility, we initially assessed the crystal packing to identify any clashes or stabilizing interactions that may have contributed to the two loop conformations (Figure S2). This loop does interact with adjacent crystal packing molecules, but we did not judge these to have undue influence on the two conformations. Next, we compared the two conformations to previously reported RBD structures (Table S1). Comparison to the SARS-CoV-2 RBD-ACE2 complex structure (PDB code: 6M0J)³¹ revealed that a critical ACE2-interacting residue in this loop, Phe486, binds ACE2 when in the conformation that more closely resembles Form-1, where it directly inserts into a hydrophobic pocket on ACE2 lined by residues Leu79, Met82 and Tyr83. In contrast, in the Form-2 conformation, Phe486 protrudes away from the ACE2 hydrophobic pocket (Figure 1D–E). Although the orientation of Phe486 between the two states is striking, mutagenesis of this residue (F486A) does not entirely prevent ACE2 binding³⁹, indicating that surrounding residues within the ACE2 binding loop also significantly contribute to stabilization of the RBD-ACE2 interface.

Table 1.

Crystallographic Data Collection and Refinement Statistics.

	SARS-CoV-2 RBD	CR3022 Fab	WRAIR-2134 + RBD	WRAIR-2123 + RBD	Ab246 + RBD	CR3022 RBD
PDB ID	8SGU	6W7Y	8SMT	8SMI	7U8E	8FAH
Data collection
Space group	P43212	P21	P212121	P21221	C21	P4122
Cell dimensions
a, b, c (Å)	80.6, 80.6, 161.8	52.1, 201.0, 57.0	143.2,154.4,165.3	39.3,94.4,201.3	194.8,41.6,99.0	151.2, 151.2,192.9
α, β, γ (°)	90, 90, 90	90, 109, 90	90, 90, 90	90, 90, 90	90, 107, 90	90, 90, 90
Resolution (Å)	50.0–1.95 (2.02–1.95)	50.00–3.3 (3.42–3.30)	50.0–3.15 (3.33–3.15)	50.0–3.5 (3.71–3.5)	50.0–2.3 (2.38–2.30)	50.0–4.2 (4.35–4.20)
Reflection (uni/tot)	38,164/107,541	16,019/30,025	121,509/433,549	16,572/34,713	33,571/106,829	13,814/84,711
Rsym or Rmerge	0.047 (0.793)	0.063 (0.198)	0.149 (1.434)	0.222 (0.366)	0.124 (0.328)	0.246 (1.088)
R pim	0.031 (0.598)	0.063 (0.198)	--	--	0.08 (0.221)	0.094 (0.57)
R meas	0.056 (0.998)	0.089 (0.28)	0.176 (1.69)	0.295 (0.48)	0.148 (0.397)	0.265 (1.239)
CC1/2	0.989 (0.706)	0.991 (0.935)	0.99 (0.42)	0.89 (0.83)	0.98 (0.89)	0.98 (0.47)
I / σI	18.9 (1.1)	10.2 (1.6)	7.15 (1.05)	5.1 (2.57)	9.78 (3.89)	5.57 (1.0)
Completeness (%)	96.8 (90.0)	96.5 (95.4)	99.7 (99.1)	90.1 (90.2)	96.3 (91.0)	82.2 (48.8)
Redundancy	2.8 (2.4)	1.9 (1.9)	6.8	3.4	3.2 (2.6)	6.1 (3.4)
Wilson B-factors (Å2)	20.0	68.9	76.4	43.1	27.5	65.5
Refinement
Resolution (Å)	40.3–1.95	20.0–3.3	49.15–3.16	42.75–3.5	20.0–2.3	30.0–4.2
Reflections	29,616	15,999	58,945	5,450	28,656	11,132
Rwork / Rfree* (%)	17.4/20.2	25.4/27.5	22.1/25.2	24.6/29.3	20.4/24.8	24.7/28.6
Mean B-factors (Å2)
Protein	21.2	66.7	84.0	38.8	42.0	139.7
Ligand/ion	52.7	--	109.5	51.7	--	219.4
Water	36.4	--	--	--	50.6	--
Ramachandran statistics (%)
Favored/Allowed/Outliers	93.5/6.5/0.0	90.8/8.0/1.2	95.2/4.8/0.0	88.7/11.3/0.0	92.54/5.4/2.06	93.5/6.5/0.0
R.m.s. deviations (from library)
Bond lengths (Å)	0.009	0.015	0.006	0.004	0.008	0.004
Bond angles (°)	1.14	1.52	0.94	0.76	0.99	0.78

Values in parentheses are for highest-resolution shells.

^*R_free was calculated using ~5% randomly selected reflections.

Figure 1. Crystal structure of the SARS-CoV-2 Receptor Binding Domain (RBD).

A, The SARS-CoV-2 RBD is shown in ribbon representation, glycan N343 is shown in sphere representation, with the N- and C-termini shown as green and red spheres, respectively. Two ACE2 ridge loop conformations are labeled and colored white and brown. B, Top panel, 2Fo-Fc electron density of the ACE2 binding loop, contoured at 1 σ. Bottom panel, The ACE2 binding loop, with either Forms-1 or −2 modeled into the corresponding density. C, SARS-CoV-2 RBD is shown as a ribbon cartoon colored by the temperature factor value (B factor) in a gradient from blue to red, spanning a numeric B-factor range from 6.5 to 96.5. D, The SARS-CoV-2 RBD structure (white) is overlaid with the ACE2-bound structure of SARS-CoV-2 RBD (blue) (PDB code: 6M0J). E, Top panel, zoom-in panel showing close contacts of ACE2 residues with Form-1 conformation of RBD. Bottom panel, zoom-in panel showing close contacts of ACE2 residues with Form-2 conformation.

Next, we compared the two RBD forms to > 150 previously published mAb-bound SARS-CoV-2 RBD and S2P structures with resolution < 3.5 Å, as well as all structures of SARS-CoV-2 S trimers (displaying up and/or down RBD conformations) (Table S1). In the majority of the previously described SARS-CoV-2 structures, the ACE2-binding loop of the RBD more closely resembles the Form-1 conformation (as shown in Figure S1B)^40,41. Comparison to a set of ACE2-bound RBD structures including SARS-CoV-1 (PDB code: 2AJF)⁴² showed that the equivalent RBD loop consistently adopted the Form-1 conformation (Figure S1B). Interestingly, the ACE2-binding loop in nearly all antibody-bound RBD structures, regardless of epitope class, adopted an intermediate conformation between Form-1 and Form-2. These structural details of the unliganded RBD and comparison to previously described RBD structures highlights this area of the RBD as structurally malleable. Furthermore, as our structural analysis indicates that RBD Form-2 is rarely observed when bound to ACE2 or mAb, this form could be incompatible with ACE-2 binding. Antibodies or small molecules that can prevent rearrangement to Form-1 could exert activity via this indirect ACE2-blocking mechanism.

Identification of SARS-CoV-2 reactive antibodies

The Omicron subvariants pose a serious threat to existing antibody therapies and vaccines, reinforcing the need to explore new mAbs and conserved therapeutic targets on S^18,43–45. To identify broadly cross-reactive mAbs that can target SARS-CoV-2 VoC, multiple strategies were employed. First, a set of publicly available SARS-CoV-1-^46,47 and MERS-CoV-reactive mAbs^48,49 were screened for binding to SARS-CoV-2. These studies identified two antibodies, a mouse antibody (240CD) and a human antibody (CR3022), that cross-reacted with SARS-CoV-2. CR3022 is a SARS-CoV-1 neutralizing antibody^21,50–52 identified from a human phage-display library⁴⁷, while 240CD was identified from mouse B cell hybridomas generated following inactivated SARS-CoV-1 virus immunization⁵³. Second, broadly reactive SARS-CoV-2 RBD-binding antibodies were identified using immune repertoire phage display⁵⁴. Ab246, described in this manuscript, was obtained from this screen. Third, B cells were isolated from a SARS-CoV-2 convalescent subject. Memory B cells that stained positive for S-2P or S-6P and S ferritin nanoparticle (SpFN) displaying eight S trimers were assessed⁵⁵; all antibodies with the “WRAIR-2xxx” designation that we describe herein were isolated from these efforts. SpFN is a vaccine candidate currently in clinical development (NCT04784767)^19,56. In the current study, we aimed to further characterize class V mAbs WRAIR-2057, WRAIR-2063 and WRAIR-2134 and one highly potent class I mAb, WRAIR-2123, that outperformed ~270 mAbs in neutralization titers in a separate study⁵⁷.

Next, we used a biolayer interferometry (BLI) competition binding assay to delineate the antigenic sites targeted by these mAbs (Figure 2A). This set of mAbs could be separated into three distinct groups: RBD-A, RBD-B, and RBD-C. The first antigenic group, RBD-A, included mAbs CR3022, Ab246 and 240CD which had nanomolar (nM) affinity for SARS-CoV-2 RBD (Figure S3A). We assessed binding competition between 240CD, CR3022, and Ab246 and showed that these antibodies cross-compete for binding to the SARS-CoV-2 RBD (Figure S3B). To further confirm that the epitope targeted by this group of antibodies was similar, we produced an RBD knockout mutant with a glycan sequon at position 384, and using BLI, showed that CR3022, 240CD and Ab246 binding was eliminated by the introduction of a glycan at this site (Figure S3C–D). mAb WRAIR-2134 targeted a distinct epitope on the RBD and was grouped together with WRAIR-2057 which targets the class V epitope⁵⁵ (RBD-B). WRAIR-2123 overlapped with another ACE2-binding site targeted antibody, WRAIR-2125, and fell into the 3^rd antigenic group, RBD-C⁵⁵ (Figure 2A). To evaluate the activity of these mAbs against SARS-CoV-2 variants, we assessed binding against a set of RBD proteins harboring mutations found in the circulating VoC and VoI (Figure 2B). mAbs CR3022, Ab246 and 240CD retained binding to all the VoC RBDs, except for the Omicron BA.1 RBD where CR3022, Ab246 and 240CD showed 30%, 28% and 60% reduced binding compared to the WA-1 strain, respectively. Interestingly, binding to the RBD variants, including Omicron variants, was consistently observed for mAbs WRAIR-2057, WRAIR-2063 and WRAIR-2134. In contrast, the binding of WRAIR-2123 was severely reduced except to the Delta variant. Overall, these three sets of mAbs each showed high affinity binding to the SARS-CoV-2 RBD (WA-1) (Figure S3A). Neutralization results were largely in agreement with the binding data wherein WRAIR-2123 showed ~280.0 and >5,500-fold drop in the neutralization potency against Alpha and Beta VoC, respectively (Figure 2C). Although CR3022 and Ab246 mAbs were poor neutralizers of WA-1 when compared to WRAIR-2123, they maintained consistent binding and neutralization against the VoC. Interestingly, WRAIR-2063, which we previously structurally characterized, potently cross-neutralized SARS-CoV-1 with an IC₅₀ value of 0.095 μg/mL (Figure 2C)⁵⁸.

Figure 2. Characterization of SARS-CoV-2 RBD-targeted mAbs.

A, Epitope binning of RBD-directed mAbs via a BLI-based competition assay. Values represent the % residual binding of the indicated second antibody after saturation of the antigen (RBD molecule) with the indicated first antibody. Shading from dark to light red indicates competition strength ranging from strong (0–25%), to lack thereof (>50%). Competition groups are indicated by black boxes. B, Binding of mAbs to VoC and VoI RBD mutants assessed by BLI. Heat-map shows the fold change in binding responses relative to the WA-1 RBD protein with loss and gain in binding represented in shades of red and blue, respectively. C, Neutralization activity of RBD mAbs against a panel of pseudotyped viruses representing the VOCs and SARS-CoV-1. Heat-map indicates IC50 values (μg/mL) ranging from potently neutralizing (blue), to poorly neutralizing (dark red). D, Assessment of RBD mAbs recruitment of Fc-mediated opsonization, antibody dependent complement deposition (ADCD), antibody dependent cell cytotoxicity (ADCC) and trogocytosis. R10 is a media-only control. E, ACE2 blocking activity of RBD mAbs. mAbs were assessed for their ability to block ACE2 binding to WA-1 (left) and Omicron (right) RBDs in a BLI-based assay.

In addition to neutralization activity, Fc effector functions have also been shown to play an important role in protection against SARS-CoV-2 in vivo^59–61. We investigated the ability of our RBD mAbs, all expressed as IgG1, to promote Fc effector functions. We first observed that class V RBD mAbs, inclusive of poorly neutralizing WRAIR-2134, were significantly better than other mAbs at mediating opsonization of cells expressing S at their surface (Figure 2D), a prerequisite for any Fc-effector activities against virus-infected cells. Reasonable complement recruitment (Antibody Dependent Complement Deposition, ADCD) was observed only for WRAIR-2063 antibody, other mAbs demonstrated poor ADCD activity. Interestingly, all the mAbs demonstrated very limited antibody dependent cellular cytotoxicity (ADCC) and trogocytosis. Overall, the ability of a subset of mAbs to target conserved regions on the S glycoprotein with multiple activities including ACE2-blocking and Fc-mediated Fc effector functions highlights a capacity for therapeutic and diagnostic development.

Next, we compared this set of mAbs for their relative potency in an ACE2-blocking assay (Figure 2E). SARS-CoV-2 WA-1 and Omicron BA.1 RBDs were used to measure the % inhibition by mAbs for ACE2 binding. For WA-1 RBD, WRAIR-2123 showed robust ACE2-blocking with an IC₅₀ value of 3 μg/ml, while, in contrast to CR3022, Ab246 showed slightly less robust ACE2-blocking activity with an IC₅₀ value of 10 μg/ml. All other mAbs could only partially block ACE2 binding to the WA-1 RBD. For the Omicron BA.1 RBD, WRAIR-2123 binding was completely abrogated, and no ACE2 blocking was observed, while Ab246 maintained its binding and robust ACE2 blocking activity (Figure 2E). To understand the molecular basis of RBD recognition by these mAbs, we determined crystal structures of CR3022, Ab246, WRAIR-2134, and WRAIR-2123 in complex with SARS-CoV-2 RBD.

Crystal structure of WRAIR-2134 in complex with SARS-CoV-2 RBD

Previously, we have described a set of RBD neutralizing antibodies including RBD-B antibodies (WRAIR-2057, WRAIR-2063 and WRAIR-2134) which were of particular interest because of their novel epitope (class V epitope)^55,58. To further our knowledge of the class V epitope we determined the crystal structure of WRAIR-2134 in complex with SARS-CoV-2 RBD. The crystal structure of the WRAIR-2134-RBD complex was determined to a final resolution of 3.16 Å and refined to an R_work/R_free of 0.21/0.27 (Table 1). WRAIR-2134, binds to a unique epitope located on the “side” of the RBD molecule, distal from the ACE2 binding site (Figure 3A–B). Overall, the WRAIR-2134 epitope covers a total BSA of 913.5 Ų with heavy and light chains contributing 68.6% and 31.4% of total BSA, respectively (Table S2). WRAIR-2134 recognition of SARS-CoV-2 RBD is primarily based on CDR H2, H3, CDR L1 and L3 loop interactions (Figure 3A–B and Table S6B). Heavy chain interactions form a total of 10 hydrogen bonds and 3 salt-bridges with the RBD molecule along with a set of CDR H2 and H3 hydrophobic residues that form major contacts with the RBD (Figure 3B, Table S2). The 19-aa long CDR H3 of WRAIR-2134 packs against a β-strand of the RBD while burying a total BSA of > 400 Ų. Light chain contacts are primarily mediated by CDR L1 and L3 forming 3 hydrogen bonds and 1 salt-bridge with the RBD. The WRAIR-2134 epitope is primarily centered on residues Lys462-Arg466 and β-strand residues Asn354-Arg357 on the RBD.

Figure 3. Crystal structure of WRAIR-2134 antibody in complex with SARS-CoV-2 RBD.

A, Left, crystal structure of WRAIR-2134, in complex with SARS-CoV-2 RBD. Both the RBD and WRAIR-2134 are shown in cartoon representation. RBD is colored white, while heavy and light chains of WRAIR-2134 are colored dark and light green, respectively. ACE2 binding site on the RBD is colored dark gray. Right, Epitope footprint of WRAIR-2134 is shown on the surface of the RBD and colored based on the antibody heavy and light chain colors. ACE2 epitope is highlighted in dark gray. B, Left, SARS-CoV-2 RBD is shown in surface representation with WRAIR-2134 epitope highlighted in green. Contacting residues from RBD are shown in stick representation and contributing heavy and light chain CDRs are shown and labelled. Right, buried surface area (BSA) for the heavy and light chain CDR loops are shown as a bar diagram. C, Crystal structures of WRAIR-2057-RBD and S2H97-RBD complexes are overlaid on the WRAIR-2134-RBD structure. The WRAIR-2134 epitope is colored dark green on the surface of SARS-CoV-2 RBD while WRAIR-2057 and S2H97 are shown in cartoon representation. Side and top views are shown. All three antibodies target an overlapping epitope while approaching the RBD at slightly different angles. D, WRAIR-2057 (magenta) and S2H97 (teal) epitopes are shown on the surface of SARS-CoV-2 RBD with ACE2 epitope highlighted in dark gray color. WRAIR-2134 and ION-300 epitopes are outlined in dark green and light blue colors, respectively. E, Omicron mutations are highlighted as red spheres on the surface of SARS-CoV-2 RBD. WRAIR-2134 epitope is shown in the cartoon tube representation and colored dark green. F, Sequence alignment of SARS-CoV-2 RBD with WRAIR-2134, WRAIR-2057, S2H97, ION-300 and ACE2 epitope residues highlighted in dark green, magenta, teal, blue, and gray, respectively.

Structural superimposition of WRAIR-2134-RBD complex with the WRAIR-2057-RBD complex indicates significant overlap between the two epitopes despite approaching the RBD from different angles (Figure 3C). However, the WRAIR-2057 footprint is located closer to the ACE2-binding site. In our structural characterization of WRAIR-2063, we demonstrated that it also targets the class V epitope⁵⁸ (Figure 2A). We compared the epitopes of WRAIR-2134 and WRAIR-2057 with two other previously described antibodies, ION-300 and S2H97, that target a similar site on the SARS-CoV-2 RBD^62,63. Structural superimposition of WRAIR-2134 with both the S2H97 and ION-300 mAbs demonstrated > 80% epitope overlap. WRAIR-2057 epitope exhibited 48% and 70% overlap with S2797 and ION-300 mAbs, respectively (Figure 3D).

The WRAIR-2134 antibody heavy chain is encoded by the IGHV3-33*01 germline gene, with 3 V gene-encoded residues altered by somatic hypermutation (SHM) and consists of an elongated (19-aa) CDR H3. The WRAIR-2134 antibody light chain is encoded by IGLV3-1*01, with 1 V gene-encoded residue altered by somatic hypermutation, and an 11-aa CDR L3 (Figure S4D). WRAIR-2063 and WRAIR-2134 mAbs utilize the same heavy chain V-gene, with WRAIR-2057 using a different germline gene, IGHV5-51*01, while all three mAbs all have low levels of SHMs. Interestingly, the V-genes of both S2H97 and ION-300 mAbs are also encoded by IGHV5-51*01 V-gene and have low SHMs (Figure S4D). The neutralization potency of WRAIR-2057 (IC₅₀ 0.73 μg/mL) and WRAIR-2063 (IC₅₀ 0.14 μg/mL) was similar to that of S2H97 (IC₅₀ 0.35 μg/mL) and ION-300 (IC₅₀ 2.4 μg/mL), while WRAIR-2134 has a lower neutralization titer (IC₅₀ 16.7 μg/mL) (Figure 2C). Altogether, our analysis of the class V mAb epitopes revealed recurring modes of SARS-CoV-2 recognition by mAbs encoded by multiple V-gene lineages and further elucidates this potent site of vulnerability. Despite the ever-expanding emergence of VoC including Omicron, no VoC mutations have been seen in the class V epitope, highlighting the therapeutic potential of this epitope⁶⁴ (Figure 3E), and as demonstrated by our pseudovirus neutralization data (Figure 2C), the class V RBD-targeted mAbs are the only antibodies we characterized that maintain at least some potency against Omicron subvariants.

Crystal structures of antibody WRAIR-2123 in complex with SARS-CoV-2 RBD

As part of the assessment of ~270 antibodies under the Coronavirus Immunotherapy Consortium (COVIC) collaboration, class I mAb WRAIR-2123 emerged as the most potent neutralizing mAb from RBD epitope community 1 (designation COVIC-259⁵⁷). Interestingly, the negative-stain electron microscopy studies also revealed that WRAIR-2123 approaches the SARS-CoV-2 S protein in a unique manner, bridging two adjacent S trimers, thereby preventing viral fusion (Figure S5). To understand the epitope of WRAIR-2123 mAb and the impact of Omicron mutations on its efficacy, we crystallized this mAb in complex with SARS-CoV-RBD. The structure of WRAIR-2123-RBD was determined to a final resolution of 3.5 Å (Figure 4A–B), solved by molecular replacement, and refined to an R_work/R_free of 0.22/0.32 (Table 1). WRAIR-2123, which potently neutralizes early strains of SARS-CoV-2⁵⁵ targets the ACE2 binding site with a distinct epitope (Figure 4A–C). WRAIR-2123 forms extensive interactions across the entire length of the receptor binding motif, with a BSA of 1107.6 Ų with heavy and light chains contributing 57% and 43% of the total BSA, respectively (Figure 4B, and Table S3). Heavy and light chain interactions form a total of 5 and 8 hydrogen bonds, respectively, with the light chain forming one additional salt-bridge. The WRAIR-2123 epitope, while centered on the residues Phe486, Tyr489, Gln493-Gln498, covers almost all of the ACE2 binding interface on the RBD. WRAIR-2123 light chain contacts include residues from all three CDR loops, whereas heavy chain contacts are primarily mediated by CDR H3, covering > 500 Ų of the RBD interface, with limited contributions from CDR H1 and H2. Major heavy chain contacts are mediated by a set of hydrophobic residues in the CDR H3 (Val99, Arg100C, Val100E, and Ile100F), whereas light chain interactions involve a mix of hydrophobic and electrostatic interactions (Figure 4C and Table S3). WRAIR-2123 heavy and light chains are encoded by IGHV3-30-3*01 and IGKV1-33*01 germline genes, respectively. WRAIR-2123 exhibited low levels of SHM with heavy and light chain V-genes consisting of 5 and 1 changes, respectively. With respect to VoCs, all RBD changes in the Alpha, Beta and Gamma VoCs are within the binding site of WRAIR-2123 mAb, and result in several fold drop in the neutralization potency of WRAIR-2123 mAb (Figure 2C). Similarly, many of the mutations in the Omicron RBD occur within the WRAIR-2123 epitope and eliminate binding (Figure 2B, 4D).

Figure 4. Crystal structure of WRAIR-2123 antibody in complex with SARS-CoV-2 RBD.

A, Left, crystal structure of WRAIR-2123, in complex with SARS-CoV-2 RBD. Both the RBD and WRAIR-2123 are shown in cartoon representation. RBD is colored white, while heavy and light chains of WRAIR-2123 are shown in dark and light blue colors, respectively. ACE-2 binding site on the RBD is colored dark gray. Right, Epitope footprint of WRAIR-2123 are shown on the surface of the RBD and colored based on the antibody heavy and light chain colors. ACE2 epitope is indicated by a dark gray line. B, SARS-CoV-2 RBD is shown in surface representation with WRAIR-2123 epitope highlighted in blue. Contacting residues from the RBD are shown in stick representation and heavy and light chain contacting CDRs are shown in ribbon representation. C, Buried surface area (BSA) for the heavy and light chain CDR loops are shown as a bar diagram. D, Omicron mutations are highlighted as red spheres on the surface of SARS-CoV-2 RBD. The WRAIR-2123 epitope is shown in tubular representation and colored blue. Omicron mutations that fall in the mAb epitope are colored blue and labeled. E, Sequence alignment of SARS-CoV-2 RBD with WRAIR-2123 and ACE2 epitope residues highlighted in blue and gray, respectively.

Crystal structures of antibodies Ab246 and CR3022 in complex with SARS-CoV-2 RBD

To understand the molecular basis of the different ACE2 blocking activity of CR3022- and Ab246-SARS-CoV-2 recognition, we determined the crystal structures of both mAbs in complex with SARS-CoV-2 RBD. The CR3022 heavy chain is encoded by IGHV5-51*03 and contains a 12-amino acid (aa) CDR H3 with 8 V gene-encoded residues altered by somatic hypermutation. CR3022 light chain is encoded by IGKV4-1*01 with 1 V gene-encoded residue altered by somatic hypermutation, and a 9-aa CDR L3 (Figure S4). In contrast, Ab246 heavy chain is encoded by the IGHV4-59*01 germline gene and has an elongated CDR H3 loop (19 residues), and the light chain is encoded by the IGLV5-45*02 germline gene. Ab246 is extensively mutated in comparison to the CR3022 antibody with 11 heavy chain and 15 light chain mutations. Crystals of the unliganded CR3022 antibody diffracted to 3.2 Å resolution in space group P2₁ (Table 1). Overall, the structure of the CR3022 Fab revealed a relatively flat antigen-combining site, apart from a protruding 12-aa CDR L1 loop (Figure S4A). Crystal structures of CR3022-RBD and Ab246-RBD complexes were determined to a resolution of 4.25 Å and 2.2 Å, solved by molecular replacement, and refined to an R_work/R_free of 0.25/0.29 and 0.20/0.25, respectively (Figure 5; Table 1).

Figure 5. Crystal structures of antibodies CR3022 and Ab246 in complex with SARS-CoV-2 RBD.

A, Left, crystal structure of Ab246, in complex with SARS-CoV-2 RBD. Both the RBD and Ab246 are shown in cartoon representation. RBD is colored white, while heavy and light chains of Ab246 are shown in dark blue and light blue colors, respectively. ACE2 binding site on the RBD is colored dark gray. Right, Epitope footprint of Ab246 is shown on the surface of the RBD and colored based on the antibody heavy and light chain colors. ACE2 epitope is indicated by a dark gray line. B, Left, SARS-CoV-2 RBD is shown in surface representation with Ab246 epitope highlighted in blue. Contacting residues from RBD are shown in stick representation and contributing heavy and light chain CDRs are shown and labelled. CDR loops are designated using the Kabat numbering system. Right, buried surface area (BSA) for the heavy and light chain CDR loops are shown as bar diagram. C, Left, crystal structure of CR3022 in complex with SARS-CoV-2 RBD. Both the RBD and CR3022 are shown in cartoon representation. RBD is colored white, while heavy and light chains of CR3022 are shown in dark and light orange colors, respectively. ACE2 binding site on the RBD is colored dark gray. Right, Epitope footprint of CR3022 is shown on the surface of the RBD and colored based on the antibody heavy and light chain colors. ACE2 epitope is indicated by a dark gray line. D, Left, SARS-CoV-2 RBD is shown in surface representation with CR3022 epitope highlighted in orange. Contacting residues from RBD are shown in stick representation and CDR contact loops are shown and labelled. CDR loops were assigned using the Kabat system. Right, buried surface area (BSA) for the heavy and light chain CDR loops are shown as a bar diagram. E, Binding interface of Ab246 CDR H3 loop and RBD is shown. CDR H3 is shown as dark blue ribbon with disulfide bond displayed in stick representation. F, Crystal structure of CR3022-RBD complex is overlaid with the Ab246-RBD structure. Both antibodies target an overlapping epitope while approaching at a right angle to each other. G, Omicron mutations are highlighted as red spheres on the SARS-CoV-2 RBD. CR3022 and Ab246 epitopes are shown in tubular representation and colored orange and dark blue, respectively, with overlapping regions of the two epitopes colored dark brown. The single Omicron mutated residue S375 that is in the CR3022 epitope is colored orange. H, Sequence alignment of SARS-CoV-2 RBD with CR3022 and Ab246 epitopes highlighted in orange and blue, respectively. ACE-2 epitope is highlighted in gray. Pro384, mutated to Asn, is highlighted in red.

The crystal structure of CR3022-RBD complex is in good agreement with previously reported structures^21,51. CR3022 bound to the RBD at an epitope centered on S glycoprotein residues 377–386 with a total buried surface area (BSA) of 861 Ų with heavy and light chains contributing 60% and 40% of total BSA, respectively. SARS-CoV-2 RBD recognition by CR3022 is primarily based on CDR H1-3 and CDR L1-2 interactions (Figure 5C–D, Table S5B). CR3022 binding to the RBD is centered on a continuous stretch of amino acids including a β-strand (residues Phe377-Tyr380), and a small loop connecting the strand to residues Pro384-Lys386. An unusually elongated CDR L1 loop makes a distant yet tight contact with residue Leu517 on the RBD. Comparison of the CR3022 epitope site with previously described antibody-complex structures for SARS-CoV-1 indicates that CR3022 targets a conserved epitope away from the ACE2-binding site and distinct from other SARS-CoV-1-targeting antibodies (Figure 5 and Figure S4B). The observed higher binding and neutralization potency of CR3022 against SARS-CoV-1 is due to residue 384 which is an Ala in SARS-CoV-1, and Pro in SARS-CoV-2. Ab246 targets a similar epitope on the RBD as CR3022, binding to 825.8 Ų with heavy and light chains contributing ~70.0% and ~30% of the total BSA, respectively (Figure 5A–B, and Table S4B). Ab246 recognition of SARS-CoV-2 RBD is primarily based on CDR H1, H3 and CDR L2 (Figure 5B and Table S4B). The Ab246 CDR H3 loop is noticeably longer than the CR3022 CDR H3, is stabilized by a Cys-Cys disulfide bond, and packs laterally against β-strand (residues Phe377-Tyr380) on the RBD. The presence of a Cys-Cys disulfide is likely important for the conformational stability of the CDR H3, while also interacting with the RBD with a BSA of 70 Ų (Figure 5E). Overall, for both CR3022 and AB246, contacts are mediated by a mix of hydrophobic and hydrophilic residues (Figure 5, Table S4–S5) while having no direct overlap with the ACE2 epitope (Figure 5H). In contrast to CR3022, Ab246 binds to the RBD with an altered approach angle differing by ~90°, likely explaining the difference in the ACE2-blocking activity (Figure 5F). Interestingly, none of the RBD mutations in the Alpha, Beta, Gamma, and Delta variants are within the CR3022 and Ab246 mAb epitopes. In the case of the recently emerged Omicron variant, only RBD mutant residue S375F, with a modest BSA of 23.6Ų, is part of the CR3022 epitope (none for Ab246) (Figure 5G). In agreement with the structural findings, we only observed slight loss of binding for mAbs Ab246 to the VoC Beta (B1.351) and Omicron BA.4/5 (Figure 2B). These data highlight the potential of targeting class IV conserved epitopes in the face of constant emergence of unique and highly transmissible VoCs.

mAb epitope analysis in the context of trimeric prefusion S

In order to compare this set of mAbs to previously described antibodies, we performed structural superimposition of CR3022, Ab246, WRAIR-2134, and WRAIR-2123 with representative antibodies from previously defined classes²². Based on this analysis (Figure 6A and S6), antibodies CR3022 and Ab246 are class IV type mAbs, WRAIR-2134 targets the recently identified class V epitope⁶³, while WRAIR-2123 is classified as a class I mAb. The WRAIR-2134 epitope overlaps with the recently reported WRAIR-2057, S2H97 and ION-300 antibodies (Figure 6A–B). In the context of the CoV prefusion trimeric S, the RBD generally displays two prototypical conformations, either in an “up” or “down” position, with implications for receptor binding and cell entry. To further analyze these conformations, we modeled binding of our mAbs to the closed conformation (all RBD down; PDB code: 6VXX), and 1- or 2-RBD in the up conformations (PDB codes: 7DWZ and 6X2B) (Figure 6C–D and S7)^30,40,65. Structure superimposition demonstrated that the WRAIR-2123 antibody binding is compatible with both the up and down conformations of RBD, indicating accessibility of the epitope. In contrast, epitopes for CR3022, Ab246, WRAIR-2134 and WRAIR-2057 are occluded by adjacent S protomers when the RBD is in the “down” conformation while more accessible when the S is in the open conformation, indicating the cryptic nature of these epitopes (Figure 6C–D and S7).

Figure 6. mAb class assignment and epitope accessibility.

A, Structures of WRAIR-2134-RBD, WRAIR-2057-RBD and S2H97-RBD (PDB code: 7M7W) complexes are overlaid with previously reported antibodies that represent frequently observed SARS-CoV-2 antibody classes²². RBD and representative mAbs are shown in surface representation. mAbs WRAIR-2057 and S2H97 are shown as ribbon and cartoon representation in magenta and teal colors, respectively. WRAIR-2134 epitope is shown in dark green color on the RBD surface. B, Epitopes for WRAIR-2057, S2H97 and ION-300 are outlined and colored magenta, teal and blue, respectively. C, Structural alignment of the WRAIR-2134-RBD and WRAIR-2057-RBD complexes with closed conformation of SARS-CoV-2 S-2P (all RBD down conformation, PDB code: 6ZGE). One protomer of the S trimer is displayed in surface representation, with the RBD of interest labeled, and epitopes for the WRAIR-2134 and WRAIR-2057 are shown on the surface of RBD, from a single protomer, in dark green and magenta colors, respectively. The other two protomers are shown as cartoon representation. In this overlay, the Fv region of WRAIR-2134 and both the Fv and Fc1 regions of WRAIR-2057 clash with the NTD of a neighboring protomer. D, Structure alignment of the WRAIR-2134-RBD and WRAIR-2057-RBD complexes with open conformation of SARS-CoV-2 S-2P (1-RBD up conformation, PDB code: 7BWJ). In this overlay, both Fab molecules have full access to the RBD, with no clashes observed.

Characterization of epitope conservation and cross-reactivity

Structural and phylogenetic analysis revealed that the epitopes for the CR3022, Ab246 and WRAIR-2134 mAbs were about 92%, 91% and 78% identical between SARS-CoV-2 and SARS-CoV-1, respectively. We aligned sequences from SARS-CoV-2 VoC and other representative sarbecoviruses (Figure 7A and S8) with the CR3022 and Ab246 epitopes also highly conserved within the sarbecoviruses. The WRAIR-2134 epitope is also relatively conserved among sarbecoviruses (Figure 7A). Intrigued by the sequence conservation of the epitopes for CR3022, Ab246 and WRAIR-2134, we next explored their antigen cross-reactivity. Both the class IV antibodies CR3022 and Ab246, and WRAIR-2134 demonstrated reasonable to good binding responses against both the SARS-CoV-1 and SARS-CoV-2 RBDs (Figure 7B). SARS-CoV-2 has a likely zoonotic origin and horseshoe bats have been implicated as natural reservoirs of both SARS-CoV-1 and SARS-CoV-2^66–70. As such, we also assessed antibody cross-reactivity with the S glycoproteins of two bat SARS-related-CoVs: SARSr-CoV Rs4874 and Rs4231⁷¹, which are closely related to the progenitor of SARS-CoV-1 and retain the ability to use human ACE2. Interestingly, CR3022 was able to recognize both recombinant S, while antibodies Ab246 and WRAIR-2134 demonstrated good binding response for Rs4874 but a weak response for the Rs4231 S trimer (Figure 7C). The remaining antibodies did not show binding to these bat SARS-CoV S glycoproteins.

Figure 7. Epitope conservation and mAb cross-reactivity analysis.

A, (Top) Structural and sequence analysis of the CR3022, Ab246 and WRAIR-2134 footprints across sarbecoviruses. The epitope residues are numbered according to the Wuhan reference; the strength of the interaction between the mAb and the RBD is indicated by the height and color of the histogram bars above the sequence alignment. Sequences are ordered based on their phylogenetic relationships based on a maximum likelihood phylogenetic tree derived from RBD protein sequences. (Left) The RBD structure is shown in surface representation and depicts mutations between SARS-CoV-1 and SARS-CoV-2 in red; the CR3022, Ab246, and WRAIR-2134 epitopes are outlined and labeled. B, mAb binding to sarbecovirus RBDs using BLI to assess cross-reactivity. Heat-map shows the binding responses (nm) represented in shades of purple. C, Cross-reactivity of mAbs were also assessed for binding to bat SARS-related CoV Rs4874 and Rs4231 S glycoproteins, using BLI.

To further understand the cross-reactivity of WRAIR-2134 and other class V mAbs to diverse sarbecoviruses, we overlayed their epitopes onto the SARS-CoV-2 RBD surface (Figure S8B). Our analysis revealed that antibodies WRAIR-2134, S2H97 and ION-300 target a relatively conserved area on the RBD with minimal sequence difference between SARS-CoV-2 and SARS-CoV-1 (Figure S8B). In contrast, the epitope for the WRAIR-2057 has greater sequence diversity (Figure S8A–B). This analysis agrees with the antigenic analysis showing potent cross-reactive binding of WRAIR-2134 with diverse S molecules whereas WRAIR-2057 has somewhat reduced breadth. Antibodies S2H97 and ION-300 have also previously been reported to be cross-reactive to SARS-CoV-1 RBD^62,63. Even in the context of the multiple viral VoCs including Omicron and Delta, we did not find a single residue that is mutated within the class V epitope and only one residue mutated for the class IV mAb CR3022 reinforcing the utility of these mAbs for therapeutics.

DISCUSSION

The ongoing COVID-19 pandemic has resulted in frequent emergence of SARS-CoV-2 VoCs with increased transmissibility and reduced sensitivity to existing mAb therapeutics. The Omicron subvariants highlight this issue due to the significant number of mutations and the resulting reduction in vaccine and mAb efficacy^18,43,44,72. This warrants the need to explore new sites of vulnerability on S and to develop next-generation, broadly reactive therapeutics that can target future VoCs. Further studies could also assess whether these conserved sites can undergo mutation and escape without demonstrable loss of activity, in the presence of mAbs.

Here we describe the unliganded SARS-CoV-2 RBD structure wherein the ACE2 binding ridge adopts two distinct conformational states. Structure-based alignment of SARS-CoV-2 RBD with previously reported SARS-CoV-1 and SARS-CoV-2 RBD structures, either in the context of S or mAb/ACE2-bound RBD revealed that an intermediary conformation between Form-1 and Form-2 is most favored and is the widely adopted conformational state of the RBD. This form allows ACE2 binding while Form-2 may represent a conformation incompatible with ACE2 association. Binding of ACE2 or mAbs appears to lock the RBD ACE2 ridge loop in the intermediate conformation. Identification of the alternate, presumably non-functional Form-2 of the RBD may enable therapeutic development of small molecule inhibitors or mAbs that could cause allosteric ACE2 inhibition.

We characterized a set of mAbs that are targeted to three distinct, non-overlapping sites on the SARS-CoV-2 RBD. The epitope for the potently neutralizing antibody, WRAIR-2123, overlaps with the ACE2 binding site on the RBD. Additionally, low resolution negative stain 3D reconstruction shows bridging of two S molecules by WRAIR-2123 IgG. This suggests that WRAIR-2123 likely interferes with viral fusion by this mechanism. Antibody CR3022 has been previously described and is grouped as a class IV mAb, and we characterize the additional class IV mAbs Ab246. The Ab246-RBD complex structure shows that Ab246 has a distinct approach angle for binding to the RBD, further enabling ACE2-blocking which contrasts with other class IV mAbs. WRAIR-2134 targets the class V epitope on the RBD and overlaps with recognition by WRAIR-2057, WRAIR-2063, ION-300, S2H97, and other class V mAbs. Additionally, WRAIR-2057 and WRAIR-2063 from this group also showed reasonable to good Fc effector activity. The class V antibodies are encoded by multiple germline genes, have low SHM levels, and typical CDR H3 lengths indicating that these mAbs can be readily elicited in humans due to the absence of unusual genetic features. This highlights these epitopes as particularly favorable and may represent mutation barriers that can be exploited for next-generation therapeutics, particularly when paired with class I or II antibodies which typically demonstrate greater neutralization capacity.

Structural analysis suggests that when one or more RBDs are in the down configuration (on a S trimer), class IV (Ab246 and CR3022) and class V (WRAIR-2134) epitopes are hidden by other S domains, most notably, the NTD. The presence of such “cryptic” but protective sites have also been previously described for influenza⁷³ and ebolaviruses⁷⁴. Targeting cryptic epitopes, these class IV and V antibodies may confer protection in two different ways. First, these antibodies can provide moderate ACE2 blocking, even though the class IV and class V epitopes do not directly overlap with the ACE2 binding site (Figures 3A and 3C). Second, binding to a “cryptic” epitope by these antibodies would result in significant conformational changes in the S trimer, potentially blocking the pre-to-post-fusion S structural transition needed for viral fusion^75–77. This has indeed been observed for other class V RBD-targeted mAbs^78,79.

In view of the emergence of SARS-CoV-2 variants, immunotherapeutics targeting conserved sites on S are of significant interest. Class IV (CR3022 and Ab246) and class V (WRAIR-2134, WRAIR-2057 and WRAIR-2063) mAbs described here also target conserved regions on a diverse set of sarbecoviruses. To date, SARS-CoV-2 VoCs do not contain significant mutations in the class IV and class V antigenic sites on the RBD, and such antibodies may have advantages over ACE2-binding site targeting antibodies. First, a high degree of epitope conservation may provide resistance to viral escape. In agreement with this, our binding and neutralization data demonstrated a broad cross-reactivity of class V mAbs across all SARS-CoV-2 VoCs reported to date, including Omicron XBB.1.5 (Figure 2B). In contrast, the binding and neutralization of WRAIR-2123, a class I mAb, was ablated by nearly all VoCs tested. A number of studies have also demonstrated a direct link between epitope conservation and viral escape^20–23. Second, identification and targeting of conserved sites of vulnerabilities on S may be critical for the long-range development of pan-CoV immunotherapeutics. Our phylogenetic and antigenic studies revealed that class IV and V epitopes are highly conserved across sarbecoviruses, and mAbs targeting these sites demonstrated high cross-reactivity. Early reports on Omicron neutralization also indicated that mAbs targeted to the conserved regions on the RBD are minimally impacted^18,72.

In summary, we have reported a high-resolution structure of SARS-CoV-2 RBD displaying a previously uncharacterized conformational state for the RBD ACE2-binding ridge. We present structural and biochemical characterization of a set of SARS-CoV-2 reactive antibodies. Several of these antibodies demonstrate a high degree of cross-reactivity and epitope conservation amongst sarbecoviruses that can be harnessed for CoV vaccine design and countermeasures development.

STAR METHODS

RESOURCE AVAILABILITY

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, M. Gordon Joyce (gjoyce@eidresearch.org).

Materials Availability

All reagents will be made available on request.

Data and Code Availability

• All data supporting the findings of this study are found within the paper and its Supplementary Information. X-ray structure factors and models have been deposited into the Protein Data Bank under accession numbers 8SGU, 6W7Y, 8SMT, 8SMI, 7U8E, and 8FAH.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this paper is available from the Lead Contact author upon request.

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Cell lines

Expi293F and FreeStyle 293F cells (Thermo Fisher Scientific) were maintained in Expi293 and FreeStyle^™ 293 expression medium (Thermo Fisher Scientific), respectively, in a 37°C shaker supplied with 8% CO2 and 80% humidity.

METHOD DETAILS

Production of recombinant proteins

The Shanghai Public Health Clinical Center & School of Public Health, in collaboration with the Central Hospital of Wuhan, Huazhong University of Science and Technology, the Wuhan Center for Disease Control and Prevention, the National Institute for Communicable Disease Control and Prevention, Chinese Center for Disease Control, and the University of Sydney, Sydney, Australia released the sequence of a CoV genome from a case of a respiratory disease from Wuhan on January 10^th 2020, available at recombinomics.co/topic/4351-wuhan-coronavirus-2019-ncov-sequences/. The sequence was also deposited in GenBank (accession MN908947) and GISAID (>EPI_ISL_402125). DNA encoding the SARS-CoV-2 RBD (residues 331–527) and RBD proteins from other sarbecoviruses were synthesized (Genscript Biotech Corp.) with a C-terminal His6 purification tag and cloned into a CMVR plasmid. SARS-CoV-2 mutant and/or variant RBD constructs were generated using QuikChange site-directed mutagenesis. All the RBD proteins were expressed by transient transfection in 293F cells for six days. The SARS-CoV-2 RBD-His protein was purified from cell culture supernatant using a Ni-NTA (QIAGEN Sciences Inc., Germantown, MD) affinity column. DNA encoding the S protein ectodomains (residues 1-1194) from bat SARS-related CoV isolates Rs4231 and Rs4874⁷¹ were synthesized (Genscript) with a C-terminal T4-Foldon domain or C-terminal GCN domain, respectively, followed by two factor xA cleavage sites and Strep-Tactin purification tags. Bat SARSr-CoV S genes were cloned into a modified pcDNA3.1 expression plasmid⁸³. Protein was initially expressed by transient transfection in Freestyle 293F cells for six days, then serial cloned to select stably expressing cell lines (Yan L., in submission). The Rs4231-T4 and Rs4874-GCN S proteins were purified from cell culture supernatant using a Strep-Tactin affinity column. The oligomeric structure of these S proteins was selected by size exclusion chromatography (Cytiva) using a 16/600 Superdex-200 purification column and trimeric S proteins were confirmed by Native-PAGE. SARS S-2P was produced as previously described, with Strep-Tactin affinity chromatography followed by gel filtration using a 16/600 Superdex-200 purification column. Purification purity for all S glycoproteins was assessed by SDS-PAGE.

The sequences of the CR3022 variable regions of the heavy and light chains are available in GenBank under accession numbers DQ168569 and DQ168570, respectively⁴⁷. WRAIR-2057, WRAIR-2063, WRAIR-2134 and WRAIR-2123 mAbs were sorted from peripheral blood mononuclear cells (PBMCs) from a convalescent donor⁵⁵. Mouse mAb 240CD was obtained from BEI Resources Repository, NIAID, NIH. Ab246 was identified using immune repertoire phage display⁵⁴. Antibody sequences were synthesized (Genscript) and cloned into CMVR expression vectors (NIH AIDS reagent program) between a murine IgG leader sequence (GenBank DQ407610) and the constant regions of human IgG1 (GenBank AAA02914), Igκ (GenBank AKL91145). Plasmids encoding heavy and light chains were co-transfected into Expi293F cells (ThermoFisher, Waltham, MA) according to the manufacturer’s instructions. After 5 days, antibodies were purified from cleared culture supernatants with Protein A agarose (ThermoFisher) using standard procedures, buffer exchanged into Phosphate-Buffered Saline (PBS) and quantified using calculated E and A280 measurements.

The Fab fragments were prepared by digestion of the full-length IgG using enzyme Lys-C (Roche Pharmaceutical, Inc., Basel, Switzerland). The digestion reaction was allowed to proceed for 2.5 hours at 37°C. Digestion was assessed by SDS-PAGE and upon completion, the reaction mixture was passed through protein-G beads (0.5–1 mL beads), 3 times and the final flow through was assessed by SDS-PAGE for purity. The Fab fragment was mixed with purified SARS-CoV-2 RBD, and the complex was allowed to form for 1 hour at room temperature.

Sequence information

SARS-CoV-2 RBD (signal peptide is underlined, purification tag in italics)

MDSKGSSQKGSRLLLLLVVSNLLLPQGVVGNITNLCPFGEVFNATRFASVYAWNRKRISNCVAD YSVLYNSASFSTFKCYGVSPTKLNDLCFTNVYADSFVIRGDEVRQIAPGQTGKIADYNYKLPDDF TGCVIAWNSNNLDSKVGGNYNYLYRLFRKSNLKPFERDISTEIYQAGSTPCNGVEGFNCYFPLQS YGFQPTNGVGYQPYRVVVLSFELLHAPATVCGPGSHHHHHH

CR3022 Heavy chain Fv

EVQLVQSGTEVKKPGESLKISCKGSGYGFITYWIGWVRQMPGKGLEWMGIIYPGDSETRYSPSFQ GQVTISADKSINTAYLQWSSLKASDTAIYYCAGGSGISTPMDVWGQGTTVTVSS

CR3022 Light chain Fv

DIVMTQSPDSLAVSLGERATINCKSSQSVLYSSINKNYLAWYQQKPGQPPKLLIYWASTRESGVP DRFSGSGSGTDFTLTISSLQAEDVAVYYCQQYYSTPYTFGQGTKVEIK

Ab246 Heavy chain Fv

QVQLQESGPGLVKPSETLSLTCSVSGDSIGTYYWSWIRQSPEKGLEWIGYIHHSGKTNYNPSLKS RVTMSVDTSKNQFSLKLTSVTAADTAVYYCARSPIGYCSSTKCYIDKPFDSWGQGTLVTVSS

Ab246 Light chain Fv

QAVLTQPSSLSASPGASASLTCTLRGGINVVNQRLYWYQQKPGSPPRFLLKYKSDSDNFLGSGVP SRFSGSKDASANAGILLISEVQSEDEADYYCMMWHSSAYVFGGGTKL

WRAIR-2134 Heavy chain Fv

EVQLVESGGGVVQPGRSLRLSCAASGFIFSSYGMYWVRQAPGKGLEWVAVIWYDGSNKYYAD SVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYYCARDLSSSSGWDDYYYYYGMDVWGQGTTV TVSS

WRAIR-2134 Light chain Fv

SYELTQPPSVSVSPGQTASITCSGDKLGDKYACWYQQKPGQSPVLVIYQDSKRPSGIPERFSGSNS GNTATLTISGTQAMDEADYYCQAWDSSTSYVVFGGGTKL

WRAIR-2123 Heavy chain Fv

EVQLVESGGGVVQPGRSLRLSCAASGFTFSSYAMHWVRQAPGKGLEWVAVLSYDGSNKYNAD SVKGRFTISRDNSKNTLYLQMNSLRVEDTAMYYCARDGVSVTMVRGVIGPLCDYWGQGTLVT VSS

WRAIR-2123 Light chain Fv

DIQMTQSPSSLSASVGDRVTITCQTSQDISNYLNWYQQKPGKAPKLLIYDASNLETGVPSRFSGSG SGTDFTFTISSLQPEDIATYYCQQYDNLPLTFGGGTKVDIK

X-ray Crystallography

Crystallization

SARS-CoV-2 RBD at 10 mg/mL and 5 mg/mL in PBS buffer was screened for crystallization conditions using an Art Robbins Gryphon crystallization robot, 0.2 ul drops, and a set of 1200 crystallization conditions. Crystal drops were observed using a Jan Scientific UVEX-PS with automated UV and brightfield drop imaging robot. Crystals of the SARS-CoV-2 RBD grew after 24 hours in multiple conditions from the Molecular Dimensions MIDAS crystal screen, with diffraction-quality crystals seen in following condition: 20% Jeffamine D2000, 10% Jeffamine M2005, 0.2 M NaCl, 0.1 M MES pH 5.5. CR3022 Fab was screened for crystallization at 10.0 mg/mL and 5.0 mg/mL concentrations in PBS. Diffraction quality crystals grew after 48 hours in 0.1 M Imidazole pH 6.5, 40% 2-propanol and 15% PEG 8,000. For the complexes, antibody Fabs (CR3022 Fab, Ab246 Fab, WRAIR-2134 and WRAIR-2123 Fabs) and SARS-CoV-2 RBD were mixed in 1:1 molar ratio, followed by superdex-200 size exclusion chromatography and crystallization drops were set-up at 8.0 and 4.0 mg/mL concentrations in PBS buffer as described above. Crystals used for data collection grew in the following crystallization conditions: CR3022-RBD complex: 1 M Succinic acid, 0.1 M HEPES pH 7.0 and 2% PEG MME2000; Ab246-RBD complex: 0.1 M Sodium acetate trihydrate pH 4.6, 2.0 M Ammonium sulfate; WRAIR-2134-RBD complex: 8% v/v Tacsimate pH 5.0, 20% w/v Polyethylene glycol 3,350; WRAIR-2123-RBD complex: 0.12 M alcohol mixture (1,6-Hexanediol; 1-Butanol; 1,2-Propanediol; 2-Propanol; 1,4-Butanediol; 1,3-Propanediol), 0.1 M buffer system 3 (Tris base and BICINE, pH 8.5), 50% precipitant mix 4 (25% v/v MPD; 25% PEG 1000; 25% w/v PEG 3350) and 0.1 M Manganese(II) chloride tetrahydrate.

Diffraction data collection and processing

Single crystals were transferred to mother liquor containing 22% glycerol, and cryo-cooled in liquid nitrogen prior to data collection. Diffraction data for SARS-CoV-2 RBD were collected at Advanced Photon Source (APS), Argonne National Laboratory, NE-CAT ID24-C beamline, and measured using a Dectris Eiger 16M PIXEL detector. Crystals grown in MIDAS condition B1 (20% Jeffamine D2000, 10% Jeffamine M2005, 0.2 M NaCl, 0.1M MES pH 5.5) provided the highest resolution diffraction with spots visible to 1.8 Å. A complete dataset could be processed to 1.95 Å in space group P4₁2₁2. CR3022 Fab crystals diffracted to 3.3 Å on NE-CAT ID24-C beamline. Diffraction data could be scaled in P2₁ space group with 99.9% completeness. Diffraction data for CR3022-RBD and Ab246-RBD complexes were collected on NE-CAT ID24-C and ID-19 beamlines, respectively, at Advanced Photon Source (APS). Both WRAIR-2134-RBD and WRAIR-2123-RBD complexes were collected on NE-CAT ID24-E beamline at at Advanced Photon Source (APS). Diffraction data from 24ID-C and 24ID-E beamlines were measured using a Dectris Eiger 16M PIXEL detector, whereas diffraction data from ID-19 beamline were measured using a Q315r CCD detector. Data collection statistics are reported in Table 1.

Structure solution and refinement

Phenix Xtriage was used to analyze the scaled diffraction data produced from HKL2000 and XDS. Data was analyzed for completeness, Matthew’s coefficient, twinning or pseudo-translational pathology. The structure of the SARS-CoV-2 RBD was determined by molecular replacement using Phaser and a search model of the SARS RBD (PDB ID: 2AJF, molecule C). CR3022 Fab crystal structure was determined by molecular replacement using Coxsackievirus A6 neutralizing antibody 1D5 (PDB ID: 5XS7) as a search model. The CR3022-RBD complex structure was determined by molecular replacement using the refined CR3022 and SARS-CoV-2 RBD structures as search models. The Ab246-RBD complex structure was determined by molecular replacement using heavy and light chain structure models of dmCBTAU-22.1 Fab (PDB: 6H0E). WRAIR-2134-RBD complex structure was determined by molecular replacement using heavy and light chain structure models from tremelimumab Fab (PDB: 5GGU) and 1710 Fab (PDB: 6B0W) Fab structures. Finally, WRAIR-2123-RBD complex structure was determined by molecular replacement using heavy and light chain structure models from P5A-1B6_2B Fab (PDB: 7CZU).

Refinement was carried out using Phenix refine with positional, global isotropic B factor refinement, and defined TLS groups, with iterative cycles of manual model building using COOT. Structure quality was assessed with MolProbity. For the SARS-CoV-2 RBD, refinement was performed with Buster⁸⁴ and Phenix refine. Both orientations of the ACE2-binding loop were manually built in COOT as two separate structures, then combined into a single coordinate file with structures defined as altloc A or altlocB and the occupancy of each set to 0.5. Refinement was performed with individual B factors, defined TLS groups, and by occupancy. All structure figures were generated using PyMOL (The PyMOL Molecular Graphics System, Version 2.1 Schrodinger, LLC). Interactive surfaces were analyzed using PISA (www.ebi.ac.uk/pdbe/pisa/). Software used in this work was curated by SBGrid⁸⁵. The final refinement statistics for all the structures are reported in Table 1.

Structure comparisons

Weighing epitope sites based on antigen-antibody interactions

Epitope sites correspond to antigen sites that are in contact with the antibody in the antigen-antibody complex (i.e. all sites that have non-hydrogen atoms within 4 Å of the antibody). For a given epitope site, the weight, which characterizes the interaction between the epitope site and the antibody (improved based on⁸⁶), was defined as:

$$w=\frac{1}{2}\left(\frac{n_c}{〈n_c〉}+\frac{n_{nb}}{〈n_{nb}〉}\right)$$

in which, $n_c$ is the number of contacts with the antibody (i.e. the number of non-hydrogen antibody atoms within 4 Å of the site); $n_{nb}$ is the number of neighboring antibody residues; $〈n_c〉$ is the mean number of contacts $n_c$ and $〈n_{nb}〉$ is the mean number of neighboring antibody residues $n_{nb}$ across all epitope sites. A weight of 1.0 is attributed to the average interaction across all epitope sites. Neighboring residue pairs were identified by Delaunay tetrahedralization of side-chain centers of residues (C_α is counted as a side chain atom, pairs further than 8.5 Å were excluded). Quickhull⁸⁷ was used for the tetrahedralization and Biopython PDB⁸⁸ to handle the protein structure. In the SARS-CoV-2 and SARS-CoV-1 RBD comparison, residues were considered similar for the following residues pairs: RK, RQ, KQ, QE, QN, ED, DN, TS, SA, VI, IL, LM and FY.

On Dec 22^nd, 2021, a search of the PDB revealed >150 SARS-COV-2 S structures. A systematic comparison between the SARS-CoV-2 RBD structure described herein with each of the PDB files is noted in Table S3.

Biolayer interferometry

mAb epitope groups were mapped by binding competition using Biolayer interferometry (BLI) on an Octet RED96 instrument (FortéBio). His-tagged recombinant RBD protein (30 μg/mL) in PBS was loaded onto HIS1K biosensors (FortéBio) for 250 s, to reach ~50% of the sensor maximum binding capacity. Loaded biosensors were immersed into wells containing the first competing antibody at 100 nM for 900 s to saturate all binding sites. Next, biosensors were dipped into wells containing the second antibody, in the presence of the first competing antibody (100 nM), and binding responses were measured after 900 s of association. Residual binding signal of the second antibody was expressed as a percentage of the maximum binding signal obtained in absence of the first competing antibody, ran in parallel. Antibodies were defined as competing when the binding signal of the second antibody was reduced to less than 25% of its maximum binding capacity and non-competing when binding was greater than 50%.

Antibody binding responses for the SARS-CoV-2 VoC RBDs and sarbecovirus RBDs were measured using BLI. Briefly, mAbs (30 μg/mL) in PBS were loaded onto AHC biosensors (FortéBio) for 120 s. Loaded biosensors were immersed into wells containing respective RBD proteins at 30 μg/mL to saturate all binding sites and binding responses were measured after 120 s of association.

mAbs were assessed for the ability to block ACE2 binding to SARS CoV-2 RBD. RBD (30 μg/ml diluted in PBS) was immobilized on HIS1K biosensors (FortéBio) for 180 s followed by baseline equilibration for 30 s and mAbs were allowed to bind for 180 s followed by baseline equilibration for 30 s. ACE2 protein (30 μg/ml) was then allowed to bind for 120 s. Percent inhibition (PI) of RBD binding to ACE2 by mAbs were determined using the equation: PI = 100 − (ACE2 binding in the presence of mAb) / (ACE2 binding in the absence of mAb) × 100.

Affinity kinetic interactions between SARS-CoV-2 RBD and mAbs were monitored on an Octet RED96 instrument (FortéBio). After reference subtraction, binding kinetic constants were determined from at least 4 concentrations of IgGs, by fitting the curves to a 1:1 Langmuir binding model using the Data analysis software 9.0 (FortéBio). Antibodies were loaded at 30 μg/mL onto an AHC probe for 120 s followed by baseline incubation for 30–60 s. SARS-CoV-2 RBD with an added glycan at residue 384, (30μg/mL) was assessed for binding to mAbs, loaded onto AHC biosensors (FortéBio) for 120 s followed by a baseline correction in PBS for 30 s. Loaded biosensors were then immersed into wells containing RBD (30 μg/mL) and binding responses were measured after 120 s of association. Binding with WA-1 SARS-CoV-2 RBD was used as a reference.

Neutralization

The S expression plasmid sequences for SARS-CoV-2 and SARS-CoV-1 were codon optimized and modified to remove an 18 amino acid endoplasmic reticulum retention signal in the cytoplasmic tail in the case of SARS-CoV-2, and a 28 amino acid deletion in the cytoplasmic tail in the case of SARS-CoV-1. This allowed increased S incorporation into pseudovirions (PSV) and thereby improved infectivity. Virions pseudotyped with the vesicular stomatitis virus (VSV) G protein were used as a non-specific control. SARS-CoV-2 pseudovirions (PSV) were produced by co-transfection of HEK293T/17 cells with a SARS-CoV-2 S plasmid (pcDNA3.4) and an HIV-1 NL4-3 deltaEnv luciferase reporter plasmid (NIH HRP). The SARS-CoV-2 S expression plasmid sequences were derived from Wuhan-Hu-1, Alpha variant, Beta variant, Omicron BA.1, Omicron BA.2, Omicron BQ.1.1, Omicron XBB.1.5, and the SARS-CoV-1 expression plasmid was derived from the Urbani S sequence.

Infectivity and neutralization titers were determined using ACE2-expressing HEK293 target cells (Integral Molecular) in a semi-automated assay format using robotic liquid handling (Biomek NXp Beckman Coulter). Test mAbs were serially diluted in growth medium and 25 μL/well was added to a white 96-well plate. An equal volume of diluted SARS-CoV-2 PSV was added to each well and plates were incubated for 1 hour at 37°C. Target cells were added to each well (40,000 cells/ well) and plates were incubated for an additional 48 hours. RLUs were measured with the EnVision Multimode Plate Reader (Perkin Elmer, Waltham, MA) using the Bright-Glo Luciferase Assay System (Promega Corporation, Madison, WI). Neutralization dose–response curves were fitted by nonlinear regression using the LabKey Server^®, and 50% inhibitory concentrations (IC50) are reported.

Measurements of antibody Fc effector functions

Opsonization: SARS-CoV-2 S-expressing FreeStyle 293F cells were generated by transfection with linearized plasmid encoding a codon-optimized full-length SARS-CoV-2 S protein matching the amino acid sequence of the IL1/2020 isolate (GenBank, MN988713). Stable transfectants were single-cell sorted and selected to obtain a high-level spike surface expressing clone (293F-Spike-S2A). 293F-Spike-S2A cells were incubated with mAbs diluted at 15 μg/mL for 30 min at 37 °C. Cells were washed twice and stained with anti-human IgG PE (Southern Biotech). Cells were then fixed with 4% formaldehyde solution and fluorescence was evaluated on an LSR II (BD Bioscience).

Antibody-dependent cell cytotoxicity: SARS-CoV-2 spike protein-expressing Expi293F cells were generated as described above for the opsonization assay. Stable transfectants were single-cell sorted and selected to obtain a high-level spike surface expressing clone (CEM-WA-1 Spike). CEM-Spike cells were plated at 100,000 per well in round bottom 96-well plates and incubated with mAbs diluted at 5 μg/mL for 30 min at 4 °C. Cells were washed and 200,000 Jurkat-Lucia NFAT-CD16 cells (RRID:CVCL_A7ZT; Invivogen, San Diego, CA, USA, cat. no. jktl-nfat-cd16) were added to each well in 100 μl of IMDM (Gibco, Burlington, ON, Canada, cat. no. 12440-053) 10% FBS. The cells were then centrifuge for 1 min at low speed and co-cultured for 24 h at 37 °C. 50 μl of Quanti-Luc (Invivogen, cat. no. rep-qlc2) was added to 20 μl of co-culture and luminescence was measured immediately on a luminometer (2104 Multilabel reader, PerkinElmer).

Trogocytosis: Trogocytosis was measured using a previously described assay⁸⁹. Briefly, SARS-CoV-2 spike–expressing Expi293F cells were stained with PKH26 (Sigma-Aldrich). Cells were then washed with and resuspended in R10 media. Cells were then incubated with mAbs diluted to 5 μg/mL for 30 min at 37°C. Effector peripheral blood mononuclear cells were next added to the R10 media at an effector to target (E:T) cell ratio of 50:1 and then incubated for 5 hours at 37°C. After the incubation, cells were washed, stained with live/dead aqua fixable cell stain (Life Technologies) and CD14 APC-Cy7 (clone MϕP9) for 15 min at room temperature, washed again, and fixed with 4% formaldehyde (Tousimis) for 15 min at room temperature. Fluorescence was evaluated on an LSRII flow cytometer (BD Biosciences). Trogocytosis was evaluated by measuring the PKH26 mean fluorescence intensity of the live CD14⁺ cells.

Antibody-dependent cellular phagocytosis: ADCP was measured as previously described⁹⁰. Briefly, biotinylated SARS-CoV-2 S stabilized trimer was incubated with red streptavidin-fluorescent beads (Molecular Probes) for 2 h at 37 °C. Ten μl of a 100-fold dilution of beads–protein mixture was incubated for 2 h at 37 °C with 100 μl of monoclonal antibodies diluted at 5 μg/mL before addition of THP-1 cells (20,000 cells per well; Millipore). After 19 h incubation at 37 °C, the cells were fixed with 2% formaldehyde solution and fluorescence was evaluated on a LSRII flow cytometer (BD Bioscience). The phagocytic score was calculated by multiplying the percentage of bead-positive cells by the geometric mean fluorescence intensity (MFI) of the bead-positive cells and dividing by 10⁴.

QUANTIFICATION AND STATISTICAL ANALYSIS

Statistical analyses related to the X-ray crystal structures are provided in Table 1.

Supplementary Material

2

Table S1. SARS-CoV-2 RBD structures mined from the Protein Data Bank, related to Figure 1 and Figure S1.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
CR3022 Heavy chain	47PMID: 16796401	DQ168569
CR3022 Light chain	47PMID: 16796401	DQ168570
WRAIR-2057 Heavy chain	55PMID: 34716452	MZ825503
WRAIR-2057 Light chain	55PMID: 34716452	MZ825502
WRAIR-2063 Heavy chain	55PMID: 34716452	MZ825501
WRAIR-2063 Light chain	55PMID: 34716452	MZ825500
WRAIR-2134 Heavy chain	55PMID: 34716452	MZ825487
WRAIR-2134 Light chain	55PMID: 34716452	MZ825486
WRAIR-2123 Heavy chain	55PMID: 34716452	MZ825491
WRAIR-2123 Light chain	55PMID: 34716452	MZ825490
Horseradish peroxidase (HRP)-conjugated sheep anti-mouse IgG, gamma chain specific	The Binding Site	Cat# AP272
Anti-RBD mouse mAb, 240CD	BEI Resources	NR-616
Ab246	54PMID: 30271892	N/A
Anti-human IgG PE	Southern Biotech	Cat# 2040–09
APC-Cy7 Mouse Anti-human CD14 (clone MϕP9)	BD Pharmingen	Cat# 557831
Bacterial and virus strains
Stbl3 competent cells	ThermoFisher Scientific	Cat# C737303
Top10 competent cells	ThermoFisher Scientific	Cat# C404010
SARS-CoV-2/human/USA/WA-CDC-WA1/2020	CDC	GenBank: MN985325.1
Biological samples
Chemicals, peptides, and recombinant proteins
3,5,3′5’-tetramethylbenzidine (TMB)	KPL	Cat# 5150-0021
Formaldehyde	Tousimis	Cat# 1008B
QUANTI-Luc	Invivogen	Cat# rep-qlc1
SARS-CoV-1 RBD protein	This manuscript	N/A
SARS-CoV-2 S-2P protein	80PMID: 32075877	N/A
SARS-CoV-2 RBD protein	This manuscript	N/A
SARS-CoV-2 B.1.1.7 RBD protein	This manuscript	N/A
SARS-CoV-2 B.1.351 RBD protein	This manuscript	N/A
SARS-CoV-2 B.1.617.2 RBD protein	This manuscript	N/A
SARS-CoV-2 Omicron BA.1 RBD protein	This manuscript	N/A
SARS-CoV-2 Omicron BA.2 RBD protein	This manuscript	N/A
SARS-CoV-2 Omicron BA.2.12.1 RBD protein	This manuscript	N/A
SARS-CoV-2 Omicron BA.4/5 RBD protein	This manuscript	N/A
SARS-CoV-2 Omicron BA.2.75 RBD protein	This manuscript	N/A
SARS-CoV-2 Omicron XBB.1.5 RBD protein	This manuscript	N/A
SARSr-CoV Rs4231-T4 S protein	This manuscript	N/A
SARSr-CoV Rs4874-GCN S protein	This manuscript	N/A
CoV BM48-31 RBD	This manuscript	N/A
CoV SHC014 RBD	This manuscript	N/A
CoV WIV1 RBD	This manuscript	N/A
CoV ZXC21 RBD	This manuscript	N/A
CoV BANAL-20-103 RBD	This manuscript	N/A
CoV RaTG13 RBD	This manuscript	N/A
CoV BANAL-20-247 RBD	This manuscript	N/A
CoV Rf4092 RBD	This manuscript	N/A
CoV Shaanxi2011 RBD	This manuscript	N/A
CoV HeB2013 RBD	This manuscript	N/A
CoV Rp3 RBD	This manuscript	N/A
CoV Rs_672 RBD	This manuscript	N/A
hACE2 protein	This manuscript	N/A
Imidazole	ThermoFisher Scientific	Cat# AC122020050
PBS	ThermoFisher Scientific	Cat# 10010023
Bovine serum albumin	Sigma-Aldrich	Cat# A8327
Anti-Human IgG Fc Capture (AHC) Biosensors	FortéBio	Cat# 18-5063
HIS1K sensors	ForteBio	Cat# 18-5120
AHC sensors	ForteBio	Cat# 18-5060
FreeStyle 293 Expression Medium	GIBCO	Cat# 12338002
Expi293 Expression Medium	GIBCO	Cat# A1435101
OPTI-MEM, Reduced Serum Medium	ThermoFisher Scientific	Cat# 11058021
IMDM	GIBCO	Cat# 12440-053
RPMI 1640 Medium	GIBCO	Cat# 11875093
Methyl-α -D mannopyranoside	Sigma Aldrich	Cat# 617-04-9
Glycerol	ThermoFisher Scientific	Cat# BP229-1
Lys-C	Roche	Cat# 50-100-3288
Critical commercial assays
Bright-Glo Luciferase Assay System	Promega	Cat# E2610
PKH26 Red fluorescent cell linker mini kit	Sigma-Aldrich	Cat# PKH26GL
LIVE/DEAD Fixable Aqua dead cell stain kit	Invitrogen	Cat# L34957
Deposited data
SARS-CoV-2 RBD	This manuscript	PDB ID: 8SGU
CR3022 Fab	This manuscript	PDB ID: 6W7Y
WRAIR-2134 + SARS-CoV-2 RBD	This manuscript	PDB ID: 8SMT
WRAIR-2123 + SARS-CoV-2 RBD	This manuscript	PDB ID: 8SMI
Ab246 + SARS-CoV-2 RBD	This manuscript	PDB ID: 7U8E
CR3022 + SARS-CoV-2 RBD	This manuscript	PDB ID: 8FAH
Experimental models: Cell lines
Freestyle 293F cells	ThermoFisher Scientific	Cat# R7007
Expi293F cells	ThermoFisher Scientific	Cat# A14635
HEK293T/17 cells	ATCC	Cat# CRL-11268
hACE2-expressing HEK293 cells	Integral Molecular	N/A
Spike S2A-expressing 293F cells	This manuscript	N/A
CEM-Spike-expressing Expi293F cells	This manuscript	N/A
Jurkat-Lucia NFAT-CD16 cells	Invivogen	Cat# jktl-nfat-cd16
THP-1 cells	ATCC	Cat# TIB-202
Experimental models: Organisms/strains
Oligonucleotides
See Table S6	This manuscript	N/A
Recombinant DNA
HIV-1 NL4-3 deltaEnv luciferase reporter plasmid	HIV Reagent Program	N/A
pCMVR hACE2	This manuscript	N/A
pCMVR SARS-CoV-1 RBD	This manuscript	N/A
pCMVR SARS-CoV-2 S-2P	80PMID: 32075877	N/A
pCMVR SARS-CoV-2 RBD	This manuscript	N/A
pCMVR SARS-CoV-2 B.1.1.7 RBD	This manuscript	N/A
pCMVR SARS-CoV-2 B.1.351 RBD	This manuscript	N/A
pCMVR SARS-CoV-2 B.1.617.2 RBD	This manuscript	N/A
pCMVR SARS-CoV-2 Omicron BA.1 RBD	This manuscript	N/A
pCMVR SARS-CoV-2 Omicron BA.2 RBD	This manuscript	N/A
pCMVR SARS-CoV-2 Omicron BA.2.12.1 RBD	This manuscript	N/A
pCMVR SARS-CoV-2 Omicron BA.4/5 RBD	This manuscript	N/A
pCMVR SARS-CoV-2 Omicron BA.2.75 RBD	This manuscript	N/A
pCMVR SARS-CoV-2 Omicron XBB.1.5 RBD	This manuscript	N/A
pCMVR SARSr-CoV Rs4231-T4 S	This manuscript	N/A
pCMVR SARSr-CoV Rs4874-GCN S	This manuscript	N/A
pCMVR CoV BM48-31 RBD	This manuscript	N/A
pCMVR CoV SHC014 RBD	This manuscript	N/A
pCMVR CoV WIV1 RBD	This manuscript	N/A
pCMVR CoV ZXC21 RBD	This manuscript	N/A
pCMVR CoV BANAL-20-103 RBD	This manuscript	N/A
pCMVR CoV RaTG13 RBD	This manuscript	N/A
pCMVR CoV BANAL-20-247 RBD	This manuscript	N/A
pCMVR CoV Rf4092 RBD	This manuscript	N/A
pCMVR CoV Shaanxi2011 RBD	This manuscript	N/A
pCMVR CoV HeB2013 RBD	This manuscript	N/A
pCMVR CoV Rp3 RBD	This manuscript	N/A
pCMVR CoV Rs_672 RBD	This manuscript	N/A
Software and algorithms
Protein Repair One-Stop Shop (PROSS) server	81PMID: 27425410	https://pross.weizmann.ac.il/step/prossterms/
Octet Data Analysis software	FortéBio	v11.1 https://www.sartorius.com/en/products/protein-analysis/octetbli-detection
GraphPad Prism	GraphPad	V8.0 https://www.graphpad.com/
PyMOL	Schrödinger	V2.1 https://pymol.org/2/
COOT	82PMID: 15572765	http://bernhardcl.github.io/coot/
SnapGene	Insightful Science	https://www.snapgene.com/
Other
Strep-Tactin Superflow resin	IBA Lifesciences	Cat# 2-1206-010
FluoroSpheres NeutrAvidin-labeled microspheres, red fluorescent	Invitrogen	Cat# F8770
Pierce™ Protein A Agarose	ThermoFisher Scientific	Cat# 20334
Superdex 200 increase 10/300 GL	Cytiva	Cat# 28990944
Galanthus Nivalis Lectin (GNL), Agarose bound	Vector Labs	Cat# AL-1243-5
NiNTA Resin	Thermo Fisher Scientific	Cat# 88221

HIGHLIGHTS.

• Structural characterization of two conformations of the apo RBD ACE2 binding loop.

• Identification of a set of broadly reactive mAbs that bind conserved RBD epitopes.

• Structural definition of class IV and V mAbs that target conserved RBD epitopes.