key: cord-0788019-1xxjsn9d
authors: Liu, Lihong; Iketani, Sho; Guo, Yicheng; Casner, Ryan G.; Reddem, Eswar R.; Nair, Manoj S.; Yu, Jian; Chan, Jasper F-W.; Wang, Maple; Cerutti, Gabriele; Li, Zhiteng; Castagna, Candace D.; Corredor, Laura; Chu, Hin; Yuan, Shuofeng; Poon, Vincent Kwok-Man; Chan, Chris Chun-Sing; Chen, Zhiwei; Luo, Yang; Cunningham, Marcus; Chavez, Alejandro; Yin, Michael T.; Perlin, David S.; Tsuji, Moriya; Yuen, Kwok-Yung; Kwong, Peter D.; Sheng, Zizhang; Huang, Yaoxing; Shapiro, Lawrence; Ho, David D.
title: Isolation and comparative analysis of antibodies that broadly neutralize sarbecoviruses
date: 2021-12-14
journal: bioRxiv
DOI: 10.1101/2021.12.11.472236
sha: 48f4ca4ce53910315fc64d4934706292e3c1c8d1
doc_id: 788019
cord_uid: 1xxjsn9d

The devastation caused by SARS-CoV-2 has made clear the importance of pandemic preparedness. To address future zoonotic outbreaks due to related viruses in the sarbecovirus subgenus, we identified a human monoclonal antibody, 10-40, that neutralized or bound all sarbecoviruses tested in vitro and protected against SARS-CoV-2 and SARS-CoV in vivo. Comparative studies with other receptor-binding domain (RBD)-directed antibodies showed 10-40 to have the greatest breadth against sarbecoviruses and thus its promise as an agent for pandemic preparedness. Moreover, structural analyses on 10-40 and similar antibodies not only defined an epitope cluster in the inner face of the RBD that is well conserved among sarbecoviruses, but also uncovered a new antibody class with a common CDRH3 motif. Our analyses also suggested that elicitation of this class of antibodies may not be overly difficult, an observation that bodes well for the development of a pan-sarbecovirus vaccine. One sentence summary A monoclonal antibody that neutralizes or binds all sarbecoviruses tested and represents a reproducible antibody class.

The devastation caused by SARS-CoV-2 has made clear the importance of pandemic 49 preparedness. To address future zoonotic outbreaks due to related viruses in the 50 sarbecovirus subgenus, we identified a human monoclonal antibody, 10-40, that 51

neutralized or bound all sarbecoviruses tested in vitro and protected against SARS-52

CoV-2 and SARS-CoV in vivo. Comparative studies with other receptor-binding domain 53 (RBD)-directed antibodies showed 10-40 to have the greatest breadth against 54 sarbecoviruses and thus its promise as an agent for pandemic preparedness. Moreover, 55 structural analyses on 10-40 and similar antibodies not only defined an epitope cluster 56 in the inner face of the RBD that is well conserved among sarbecoviruses, but also 57 uncovered a new antibody class with a common CDRH3 motif. Our analyses also 58 suggested that elicitation of this class of antibodies may not be overly difficult, an 59 observation that bodes well for the development of a pan-sarbecovirus vaccine.

The COVID-19 pandemic is caused by the severe acute respiratory syndrome 63 coronavirus 2 (SARS-CoV-2) that has infected >245 million people and resulted in >5 64 million deaths (1). Multiple variants of this virus have emerged, including some capable 65 of increased transmission or antibody evasion (2, 3). Furthermore, the threat of 66 continued zoonotic spillovers warrants the development of interventions that could 67 broadly combat animal coronaviruses with pandemic potential (4). 68

Numerous anti-SARS-CoV-2 monoclonal antibodies (mAbs) have been isolated 69 and characterized, with several demonstrating clinical utility (5, 6). Some have been 70 reported to possess broadly neutralizing activity against not only SARS-CoV-2 but also 71 other sarbecoviruses (7-15), a viral subgenus containing both SARS-CoV-2 and SARS-72 CoV (16). Such mAbs could serve as a therapeutic adjunct for the current pandemic as 73 well as a useful agent in addressing future zoonoses due to a sarbecovirus. We now 74 report the isolation of three mAbs that broadly neutralize sarbecoviruses. In addition, we 75 describe virological and structural findings from a comprehensive comparative analysis 76 of these mAbs together with those previously reported to have broad activity. The 77 information provided herein could aid the development of pan-sarbecovirus antibodies 78 and vaccines. 79

To isolate mAbs with the desired neutralization breadth, we screened sera from 80 convalescent COVID-19 patients for neutralizing activity against a panel of variant 81 viruses. Serum from Patient 10 and Patient 11 potently neutralized all SARS-CoV-2 82 variants tested as well as SARS-CoV, albeit weakly (Fig. S1) . We then sorted for B.1.351 spike trimer-specific memory B cells from the blood of both patients, followed 84 by single-cell RNA-sequencing to determine the paired heavy and light chain sequences 85 of each mAb (Fig. S2 ) (17). A total of 58 mAbs were isolated and characterized. 86

Three mAbs, 10-40, 10-28, and 11-11, were found to bind to SARS-CoV-2 spikes 87 of variants D614G and B.1.351 as well as the SARS-CoV spike (Fig. 1A) . All three 88 antibodies recognized epitopes within the receptor binding domain (RBD) (Fig. 1A) and 89 inhibited the binding of soluble human ACE2 receptor to the spike (Fig. 1B) . Epitope 90 mapping by competition ELISA was carried out on these 3 antibodies along with a panel 91 of 9 RBD-specific mAbs reported to have breadth against sarbecoviruses, including 92 DH1047 (7), S2X259 (8), REGN10985 (9), ADG-2 (10), 2-36 (11), COVA1-16 (12), 93 CR3022 (13), S2H97 (14), and S309, also known as sotrovimab (15). 10-40, 10-28, and 94 11-11 fell into one competition group with 7 other mAbs ( Fig. 1C and S3 ) that are 95 known to recognize an inner face of RBD when it is in the "up" position (7-12). The 96 epitope of S2H97 was partially overlapping whereas that of S309 was discrete, not 97 surprisingly since the latter is directed to an epitope on the outer face on RBD (14, 15). 98

The binding affinities of this panel of mAbs to SARS-CoV-2 and SARS-CoV spikes were 99 measured by surface plasmon resonance and summarized in Fig. S4 . 100

Genetically, 10-40, 10-28, and 11-11 utilized IGHV4-39*01, IGHV3-30*18, and 101 IGHV4-31*03 heavy-chain V genes with CDRH3 lengths of 22, 13, and 21 amino acids, 102 respectively. The light chains of 10-40, 10-28, and 11-11 were derived from IGLV6-103 57*01, IGKV1-39*01, and IGLV1-40*01, respectively (Fig. S5A) . All three antibodies 104 had low levels of somatic hypermutation (Fig. S5, A 

We then comprehensively compared the virus-neutralizing potency and breadth 106 of 10-40, 10-28, and 11-11 to other RBD-directed mAbs with known activity against 107 other sarbecoviruses. First, each antibody was evaluated against SARS-CoV-2 variants 108 in neutralization assays using both VSVΔG-pseudotyped viruses and authentic viruses. There are more sarbecoviruses found in bats in Africa/Europe or Asia (Fig. S9A ) 123 that do not use human ACE2 as receptor (16, 18). Since their target cells are unknown, 124 performing virus-neutralization assay is not readily feasible. We therefore examined the 125 binding profiles of this panel of mAbs to RBD proteins derived from six sarbecoviruses 126 outside of SARS-CoV-2 and SARS-CoV sublineages (Figs. 2D and S9B) . 10-40, 10-28, 127

and S2H97 bound all RBDs tested, whereas DH1047 did not recognize the RBD of Rf1. 128

The remaining mAbs bound only a subset of the RBDs. In particular, S2X259, 129 REGN10985, ADG-2, and S309 did not recognize, largely, the RBD of Asian bat 130 sarbecoviruses. Similarly, we assessed the binding of this panel of mAbs to three spikes 131 of Asian bat sarbecoviruses (Fig. S9A) as expressed on the surface of transfected cells. 132 10-40, 10-28, 2-36, and DH1047 exhibited breadth, but other mAbs were largely non-133 reactive (Figs. 2E and S10) . 134

In summary, the totality of findings in Fig. 2 shows that 10-40 can neutralize or 135 bind every sarbecovirus we have studied, and it appears to be an RBD-directed mAb 136 with the greatest breadth against sarbecoviruses known to date. 137

To investigate the nature of antibody-spike interactions for 10-40, 10-28, and 11-138 11, we determined the cryo-EM structures for Fabs of these mAbs in complex with S2P-139 prefusion-stabilized spike proteins from SARS-CoV-2 WA-1 and B.1.351 strains. 140

Interestingly, a greater degree of spike disassembly was observed for all WA-1 141 complexes, and higher quality cryo-EM maps were obtained for all three B.1.351 142 complexes. A single predominant population was observed where three Fabs were 143 bound per spike in a 3-RBD-up conformation (Fig. 3A , S11, S12, S13, and Table S1 ). 144

For 11-11, an additional class of two Fabs bound with 2-RBD up was also observed. 145

The 10-40 complex reconstruction reached 3.5 Å global resolution, but local refinement 146 of the RBD + Fab maps did not surpass 4 Å resolution. To resolve the interfaces, we 147 determined crystal structures of the 10-40 and 10-28 Fab:RBD complexes, as 148 elaborated in paragraphs below. Both the crystal structures of 10-40 and 10-28 fitted 149 nearly perfectly in the cryoEM reconstruction density (Fig. S11 and S12) . For 11-11, a 150 homology model was built and docked into the map (Fig. S13) , which showed this antibody recognizes RBD in the same way as S2X259 (8), with both sharing a similar 152 light chain and CDRH3 motif. 153

To visualize the epitopes of 10-40, 10-28, and 11-11 on the SARS-CoV-2 WA1 154 RBD at higher resolution, we determined the structures of Fab in complex with RBD 155 using X-ray crystallography, with suitable crystals obtained for 10-40 and 10-28 (Table  156 S2). The 10-40 crystal structure at 1.5 Å resolution revealed recognition of an epitope 157 that is highly conserved among sarbecoviruses on the inner face of RBD (Figs. 3B and  158   3C ). 10-40 used three of its six CDR loops: H1, H3, and L2, to interact with an epitope 159 consisting of a loop on RBD (residues 377-385) and extended toward the RBD ridge 160 near the ACE2 binding site. 10-40 also established extensive polar contacts and 161 hydrophobic interactions with RBD residues (Fig. 3B) . 162

For the 10-28 crystal structure (Fig. S14A) , antibody-RBD side-chain interactions 163 were well defined at 3.2 Å resolution. Interactions were mediated by 10-28 CDR loops 164 H1, H3, L1, and L3, which predominantly contacted residues formed a total of five hydrogen bonds (Fig. S14B) . 168

All three antibodies recognized a region on the inner side of RBD that is hidden 169 in the RBD-down conformation of spike. Thus, they can only recognize RBD in the up 170 conformation. The 10-40 epitope is similar to the 'class 4' antibody epitope previously 171 defined for COVA1-16, C022 and 2-36 (Fig. S15A) (19) . Superposition of the 10-40 Fab 172 with the ACE2-RBD complex (6M0J) showed that antibody binding places the VL 173 domain of the antibody in a position that would clash with ACE2 (Fig. 3D) , consistent with the experimental data showing inhibition of ACE2 binding (Fig. 1B) . In addition, 175 these mAbs all targeted the same exposed β -sheet on RBD from a similar angle. 176 S2X259 and DH1047 also recognized the same β -sheet, but from a different angle, 177 similar to 11-11 (Fig. S15B) . The epitope of 10-28 also included a portion of this β -sheet 178 although the focus was lower on the RBD, similar to CR3022 (Fig. S15C) . 179

Given the broad recognition of 10-40 for sarbecoviruses, we carefully analyzed 180 the amino acids that form its epitope and noted their remarkable conservation among 181 sarbecoviruses (Figs. 3C, S16A, and S16B), suggesting that this RBD region must 182 have been subjected to strong functional constraints during evolution of this subgenus. 183

Residues 377-379 in a β -sheet was specifically targeted by the CDRH3 of 10-40 via 184 multiple hydrogen bonds (Fig. 3E) . Interestingly, COVA1-16, 2-36, and C022 interacted 185 quite similarly with the same residues. We also noticed that these three mAbs, together 186 with 10-40, contact this particular β -sheet through a 'YYDRSGY' motif originating from 187 IGHD3-22 (Fig. 3F) , a D gene that is frequently used by antibodies in the human 188 repertoire (Fig. S17) . This motif contains a Ser-to-Arg substitution, which formed 189 hydrogen bonds with 369Y and 371S on RBD (Fig. 3E) . Importantly, we believe these 190 structural similarities define 10-40, 2-36, C022, and COVA1-16 as members of a new 191 antibody class, each using a shared mode of heavy-chain binding to RBDs of 192 sarbecoviruses (Fig. 3F) . As these four mAbs use diverse heavy-chain V genes and 193 light-chain recombinations (Fig. S18) and show low-level somatic hypermutation (Fig.  194   S5) , the elicitation of this class of antibodies may not be overly difficult. This observation 195 bodes well for the development of a pan-sarbecovirus vaccine.

Finally, we evaluated the in vivo protective efficacy of 10-40 by challenging wild-197 type mice with a mouse-adapted SARS-CoV-2 strain, MA10 (20), or hACE2-transgenic 198 mice with SARS-CoV (21). The RBD mutations in MA10 (Q493K, Q498Y, and P499T) 199 mapped outside of the 10-40 epitope (Fig. S19A) and did not strongly affect the 200 neutralizing activity of 10-40 in vitro (Fig. S19B) . We then performed a prevention 201 experiment (Fig. 4A) , administering 10-40 or an anti-HIV-1 control mAb 24 hours before 202 the mice were challenged intranasally with MA10. Compared to the control group, 203 significant weight loss was prevented (10 mg/kg) or mitigated (2 mg/kg) by 10-40 204 administration (Fig. 4B) . In mice given the control antibody, high levels of infectious 205 virus were observed in the lungs (>10 5 TCID50/g lung), while little (≤10 4 TCID50/g lung) 206 or no infectious virus was found in mice given 10-40 at 2 mg/kg or 10 mg/kg, 207 respectively (Fig. 4C ). An analogous prevention experiment was conducted against 208 SARS-CoV in hACE2-transgenic mice (Fig. 4D) . Weight loss was again prevented by 209 10-40 administration (Fig. 4E) , and levels of SARS-CoV were markedly reduced in the 210 lungs of mice pre-treated with 10-40 (Fig. 4F) subgenus that is presently harbored by bats and other animals (4, 16, 22) . A pan-216 sarbecovirus neutralizing mAb and/or vaccine could be useful interventions to contain 217 another outbreak. We have identified a human mAb, 10-40, that suits this need. While 218 RBD-directed mAbs that neutralize sarbecoviruses have been reported (7-15), 10-40 shows the greatest breadth in our comprehensive comparative analysis, followed 220 closely by DH1047 (7) (Fig. 2) . 10-40 neutralizes or binds to every sarbecovirus we 221 have tested, regardless of their usage of ACE2 as receptor. It has the requisite potency 222 against SARS-CoV-2 in vitro and in vivo (Fig. 4) ; interestingly, its potency against other 223 sarbecoviruses is even better (Fig. 2) , despite being isolated from a COVID-19 patient. 224

Very recently, three sarbecoviruses closest genetically to SARS-CoV-2 have been 225 identified in bats in Laos (22), along with two more in a different sublineage (Fig. S16) . 226

While not empirically tested, we note that they are likely to be susceptible to 10-40 227 because the key amino acids that would form this RBD epitope are identical to those in 228 either GD-Pangolin or RmYN02 (Fig. S16) , both of which are neutralized or bound by 229 10-40 (Figs. 2C-E) . We believe 10-40 is a promising candidate for pandemic 230 preparedness. 231

The pursuit of a pan-sarbecovirus vaccine is already underway, including 232 strategies that specifically target the stem helix in the S2 region of spike (23) or 233 conserved elements on RBD. Efforts directed to the latter have already shown promise 234 (24-26). The structural differences in the epitopes recognized by 10-40 and DH1047 235 (Figs. 3C and S15) versus epitopes recognized by mAbs (such as ADG-2 or S2X259) 236 with lower breadth against sarbecoviruses (Fig. S15 ) could be informative in focusing 237 the antibody response to certain conserved residues on the inner face of RBD. The 238 epitopes of 10-40 and DH1047 are substantially overlapping, but their angles of 239 approach are different (Fig. S15) , suggesting that this site could be attacked in more 240 than one way. Moreover, by genetically comparing 10-40 with other antibodies, we have 241 uncovered a unique class of mAbs that target a particular β -sheet in RBD through a 242 common CDRH3 motif (Fig. 3E-F) . That this class of mAbs could use multiple heavy 243 and light chain V genes (Fig. S18) and without extensive somatic hypermutation (Fig.  244   S5) is certainly welcome news for the development of a pan-sarbecovirus vaccine 245 targeting the RBD. 246 

An interactive web-based dashboard to track 249 COVID-19 in real time

Antibody resistance of SARS-CoV-2 variants B.1.351 and B.1.1.7. 251

Estimated transmissibility and impact of SARS-CoV-2 lineage 253

B.1.1.7 in England

Disease X: 255 accelerating the development of medical countermeasures for the next pandemic

Bamlanivimab plus Etesevimab in Mild or Moderate Covid-19

REGN-COV2, a Neutralizing Antibody Cocktail, in 260 Outpatients with Covid-19

In vitro and in vivo functions of SARS-CoV-2 infection-enhancing and 262 neutralizing antibodies

Broad sarbecovirus neutralization by a human monoclonal 264 antibody

The monoclonal antibody combination REGEN-COV protects 266 against SARS-CoV-2 mutational escape in preclinical and human studies

Broad and potent activity against SARS-like viruses by an 269 engineered human monoclonal antibody

A monoclonal antibody that neutralizes SARS-CoV-2 variants

Cross-Neutralization of a SARS-CoV-2 Antibody to a Functionally 273

A highly conserved cryptic epitope in the receptor binding 275 domains of SARS-CoV-2 and SARS-CoV

SARS-CoV-2 RBD antibodies that maximize breadth and 277 resistance to escape

Cross-neutralization of SARS-CoV-2 by a human monoclonal 279 SARS-CoV antibody

Evolutionary origins of the SARS-CoV-2 sarbecovirus lineage 281 responsible for the COVID-19 pandemic

Potent neutralizing antibodies against multiple epitopes on SARS-283

CoV-2 spike

The evolutionary history of ACE2 usage within the coronavirus 285 subgenus Sarbecovirus

SARS-CoV-2 neutralizing antibody structures inform 287 therapeutic strategies

A Mouse-Adapted SARS-CoV-2 Induces Acute Lung Injury and 289

Lethal infection of K18-hACE2 mice infected with severe 291 acute respiratory syndrome coronavirus

Coronaviruses with a SARS-CoV-2-like receptor-binding 293 domain allowing ACE2-mediated entry into human cells isolated from bats of 294 Indochinese peninsula

Epitope-resolved profiling of the SARS-CoV-2 antibody 296 response identifies cross-reactivity with endemic human coronaviruses

Mosaic nanoparticles elicit cross-reactive immune responses 299 to zoonotic coronaviruses in mice

Chimeric spike mRNA vaccines protect against 301

Sarbecovirus challenge in mice

Elicitation of broadly protective sarbecovirus immunity by 303 receptor-binding domain nanoparticle vaccines

We are grateful to A. Cohen