key: cord-0937769-v6sphtnx
authors: Li, Wenwei; Chen, Yaozong; Prévost, Jérémie; Ullah, Irfan; Lu, Maolin; Gong, Shang Yu; Tauzin, Alexandra; Gasser, Romain; Vézina, Dani; Anand, Sai Priya; Goyette, Guillaume; Chaterjee, Debashree; Ding, Shilei; Tolbert, William D.; Grunst, Michael W.; Bo, Yuxia; Zhang, Shijian; Richard, Jonathan; Zhou, Fei; Huang, Rick K.; Esser, Lothar; Zeher, Allison; Côté, Marceline; Kumar, Priti; Sodroski, Joseph; Xia, Di; Uchil, Pradeep D.; Pazgier, Marzena; Finzi, Andrés; Mothes, Walther
title: Structural Basis and Mode of Action for Two Broadly Neutralizing Antibodies Against SARS-CoV-2 Emerging Variants of Concern
date: 2021-12-15
journal: Cell Rep
DOI: 10.1016/j.celrep.2021.110210
sha: 5549fdb3ffed06a1c1cca2910ba3fcc1ea7586b0
doc_id: 937769
cord_uid: v6sphtnx

Emerging variants of concern for the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) can transmit more efficiently and partially evade protective immune responses, thus necessitating continued refinement of antibody therapies and immunogen design. Here we elucidate the structural basis and mode of action for two potent SARS-CoV-2 Spike (S) neutralizing monoclonal antibodies CV3-1 and CV3-25 that remain effective against emerging variants of concern in vitro and in vivo. CV3-1 binds to the (485-GFN-487) loop within the receptor-binding domain (RBD) in the “RBD-up” position and triggers potent shedding of the S1 subunit. In contrast, CV3-25 inhibits membrane fusion by binding to an epitope in the stem helix region of the S2 subunit that is highly conserved among β-coronaviruses. Thus, vaccine immunogen designs that incorporate the conserved regions in RBD and stem helix region are candidates to elicit pan-coronavirus protective immune responses.

Given that soluble S trimers are truncated, lack the transmembrane region, and feature a T4 284 foldon, we reasoned that cryoET of native spike proteins embedded into virus particles could 285 provide more insight into CV3-25's epitope. We used cryoET followed by subtomogram 286 averaging of ~7000 prefusion spikes to examine CV3-25 binding to SB.1.1.7. Subclassification 287 revealed that about half of S had two CV3-25 Fabs bound to the stem of S2, and the other half 288 had only one CV3-25 Fab bound ( Figure S5 ). We further aligned the subtomograms with a 289 mask for two CV3-25 Fabs to arrive at ~10 Å resolution map. This structure places the CV3-25 290 epitope within the connecting domain (CD) of the stem helix ( Figure 5A To gain atomic insight, we screened S2 peptides for binding to CV3-25 with the goal of isolating 304 peptides suitable for X-ray crystallography. The first insight that CV3-25 binds a linear peptide 305 was gained from Western blotting following SDS-PAGE. CV3-25 was clearly able to bind to S2 306 as well as the S2-containing S precursor under fully denaturing conditions and independently of 307 N-linked glycans ( Figure S6A , B). We then tested a set of peptides (15-mer) spanning the entire 308 S2 subunit including the connecting domain and performed two rounds of ELISA to identify 309 peptides capable of binding CV3-25 ( Figure 5C -E). The identified peptides (#288 and #289) 310

were also tested in competition assays and the binding was quantified using SPR assays 311 To obtain molecular insight into CV3-25 interaction with the S2 stem peptide, we determined the 318 co-crystal structure of CV3-25 Fab with a synthetic peptide spanning residues 1140-1165 319 (26mer) of SARS-CoV-2 spike. The structure was solved to 2.1 Å resolution and allowed us to 320 resolve 20 of the 26 residues in relation to the Fab paratope ( Figure 6 , Figure S7 , Table S2 ). 321

When bound to CV3-25 Fab, the peptide adopted a bent conformation with the N-terminal half 322 of the peptide (residues 1146-DSFKEELDKYFK-1157) forming an α-helix and the C-terminal 323 half a random coil (residues 1158-NHTSPDVD-1165) with a bend of ~95˚ between the two 324 ( Figure 6A ). This bent conformation fit well with the long complementary determining region 325 (CDR) H3 loop of the Fab (16 aa long) that is stabilized by extensive H-bonds, salt bridges and 326 intra-molecular π-π stacking between residues Y 1155 and H 1159 of the peptide ( Figure 6B -D and 327 Figure S7 ). A rare CDR H3 disulfide bond between residues C 99 and C 100D also stabilizes the 328 CDR H3 hairpin that tightly associates with the S2 peptide random coil. Interestingly, the S2 329 stem region recognized by CV3-25 is conserved among the B lineage of β-coronaviruses 330 ( Figure 6F ), with several key epitope residues also conserved among A, C, and D lineages. 331

Indeed, CV3-25 is able to recognize the stem helix peptide derived from MERS Spike (Hurlburt 332 et al., 2021) . This suggests that CV3-25 displays cross-reactivity with coronaviruses beyond the 333 B-lineage (Jennewein et al., 2021; Ullah et al., 2021) . Furthermore, although crystallographic 334 analyses confirm residues 1149 to 1165 of the S2 stem to interact with CV3-25 ( Figure 6D ), 335 SPR analyses using S2 peptide truncations indicate that CV3-25 may also interact with residues 336 following D 1165 , the terminal S2 residue used in crystallographic studies. These contacts could 337 be mediated by the light chain of CV3-25 that is positioned to accommodate the C-terminal 338 extension of the peptide ( Figure 6A ). Additionally, a potential N-linked glycosylation site (PNGS) 339 is present in the CV3-25 epitope (N1158) but the residue is positioned in a way that the glycan 340 protrudes on the opposite side of the stem helix. Removal of the N1158 PNGS using different 341 substitution did not affect CV3-25 binding, which is in line with its glycan independence ( Figure  342 S6B, F-G). Of note, the S2 recognition site and angle of approach of CV3-25 differentiate it from 343 B6, an anti-MERS-CoV cross-reactive S2-binding Nab, as well as two known human NAbs 344 specific for SARS-CoV-2 stem-helix, CC40. Here we describe the structures and mode of action of two potent anti-SARS-CoV-2 Spike 377 antibodies. Both antibodies remained effective against emerging variants of concern and 378 therefore were prime candidates to elucidate mode of action and identify epitopes with pan-379 coronavirus activity. CV3-1 stabilized the receptor-binding domain (RBD) in the "RBD-up" 380 conformation and triggered potent shedding of S1. The ability of CV3-1 to neutralize variants of 381 concern correlated with its ability to shed S1 and inactivate S. In contrast, CV3-25 bound to a 382 highly conserved epitope in the stem helix in the S2 subunit and inhibited membrane fusion. We 383 believe that both epitopes of these two antibodies are of interest for passive and active 384 immunization strategies against emerging variants. 385

The cryoET structure of CV3-1 to S suggested binding to the 485-GFN-487 loop of RBD, an 386 interpretation confirmed by mutagenesis. While mutations in these positions abrogate the 387 binding of CV3-1 to S, they are rarely observed among circulating strains, suggesting that they 388 are associated with a high fitness cost likely due to their importance in ACE2 interaction, in all the way to the S2 stem region. This allosteric signaling likely weakens the S1-S2 interface 395 leading to the observed shedding of S1. While CV3-1 is specific against SARS-CoV-2, it 396 remained active against all tested variants of concern and variants of interest and protected 397 K18-hACE2 transgenic mice from lethal challenges using the B.1.351 variant of concern. The 398 One of the most exciting aspects of CV3-25 is its linear peptide epitope, which offers easy 435 access to exploration of its potential as an immunogen. As the structure of the native S on the 436 surface of virus particles revealed, access of CV3-25 is hindered by the need for rotation of the 437 stem helix. However, such conformational readjustment is not needed for an immunogen. As 438 such, eliciting antibodies targeting this S2 stem epitope using peptide or scaffold-presented 439 peptide immunogens is predicted to be easier than when the entire spike trimer is the antigen. 440

The potential of the CV3-25 epitope described herein should be explored as a candidate 441 immunogen for vaccines that could be effective against all emerging variants and possibly 442 exhibit pan-coronavirus efficacy. 443 444

Our study suggests that neutralizing antibodies of SARS-CoV-2 are also inhibitors of virus entry, 446 e.g. by causing shedding of S1 or inhibiting membrane fusion, a hypothesis that requires future 447 studies. We observed that potent neutralization of SARS-CoV-2 emerging variants of concern 448 by CV3-1 correlated with shedding of S1. A detailed understanding of how CV3-1 activates and 449 sheds S1 requires higher resolution structures, and comparative studies among RBD-targeting 450 antibodies are needed to understand what antibodies trigger S1 shedding. 451

We arrived at a model for CV3-25 bound to the S2 stem helix of spike by flexible fitting of the X-452 ray structure of peptide bound to CV3-25 and 6XR8 into our cryoET density map. 

Reduced sensitivity of SARS-CoV-2 709 variant Delta to antibody neutralization

Safety and Efficacy of the BNT162b2 712 mRNA Covid-19 Vaccine

Choice of data-collection parameters based on 714 statistic modelling

The great escape? SARS-CoV-2 variants evading neutralizing 716 responses

Cross-Sectional Evaluation of Humoral 719 Responses against SARS-CoV-2 Spike

cryoSPARC: algorithms for 721 rapid unsupervised cryo-EM structure determination

Broad and potent activity against SARS-like viruses by an engineered human 725 monoclonal antibody

Convergent antibody responses to SARS-728

CoV-2 in convalescent individuals

Alignment of cryo-EM movies of individual particles 730 by optimization of image translations

Safety and Efficacy of Single-Dose 733

S Vaccine against Covid-19

Interim Results of a Phase 1-2a Trial 736 of Ad26.COV2.S Covid-19 Vaccine

FreeStyle 293-F (Thermo Fisher) cells were grown to a density of 1x10 6 cells/mL at 37°C with 8% 1241 CO2 with regular 135 rpm agitation. A plasmid encoding for non-cleavable

2020) -a 1243 gift from Dr. Jason S. McLellan) with a removable C-terminal twin-strep tag was transfected into 1244 cells with EndoFectin Max (GeneCopoeia) using the manufacturer's protocol. One-week post-1245 transfection, the clarified supernatant was purified on strep-tactin resin (IBA) followed by size-1246 exclusion chromatography on a Superose 6 10/300 column

Sigma Aldrich) digestion overnight at 4 °C and the 1249 uncleaved protein was removed by passage over Ni-NTA resin. The cleaved protein was further 1250 purified on a Superose 6 10/300 column in SEC buffer. Alternatively, cells were transfected with 1251 a plasmid coding for SARS-CoV-2 RBD or ACE2-Fc and were purified on Ni-NTA resin 1252 (Invitrogen) or Protein A resin (Cytiva), respectively. Protein purity was confirmed by SDS-1253 PAGE

Expression plasmids encoding the heavy and light chains of CV3-1 IgG or CV3-25 IgG were 1255 transiently transfected into Expi293F cells (Thermo Fisher) with ExpiFectamine 293 transfection 1256 reagent using the manufacturer's protocol (Thermo Fisher)

Fab was separated from Fc and uncleaved IgG by passage over protein A resin 1260 followed by size-exclusion chromatography on a Superose 6 10/300 column before being used 1261 in SPR binding, X-Ray crystallography or Cryo-EM experiments

Surface Plasmon Resonance

For the kinetic binding measurements of S2 peptides #289 (15-mer), #289 (11-1268 mer) and the 26mer (1140-1165) to CV3-25, ~5800 RU of CV3-25 IgG was first immobilized on 1269 a protein A chip (Cytiva) and 2-fold serial dilutions of the S2 peptides were then injected with 1270 concentrations ranging from 6.25 to 200 nM. After each cycle the protein A sensor chip was 1271 regenerated with 0.1 M Glycine pH 2.0. CV3-1 IgG was used as a negative control. All 1272 sensorgrams were corrected by subtraction of the corresponding blank channel in addition to 1273 the buffer background and the kinetic constant determined using a 1:1 Langmuir model with the 1274 BIAevaluation software (GE Healthcare)

Sample Preparation and Data Collection 1277

HexaPro spike (293F produced) was incubated with 20-1278 fold excess of CV3-25 Fab overnight at 4ºC before purification on a Superose 6 300/10 GL 1279 column (GE Healthcare). The complex peak was harvested, concentrated to about 0.5 mg/mL in 1280 SEC buffer and immediately used for CryoEM grid preparation. 3μL of protein was deposited on 1281 a holey copper grids (QUANTIFOIL R 1.2/1.3, 200 mesh, EMS) which had been glow-1282 discharged for 30s at 15 mA

Thermo Fisher) with a blot time of 2-4 s and the blot force of 20 at 4°C and 95% 1284 humidity

300kV Titan Krios electron microscope, 1286 equipped with a Gatan K2-BioQuantum Image filter camera system (Thermo Fisher and Gatan 1287

50-frame image stacks 1288 were collected at a magnification of 165,000x

CryoEM Data Processing, Model building and Analysis 1291

Motion correction, CTF estimation, particle picking, curation and extraction, 2D classification, ab 1292 initio model reconstruction, volume refinements and local resolution estimation were carried out 1293 in cryoSPARC

Model-to-map fitting cross correlation 1301 and figures generation were carried out in USCF Chimera

2021) and PyMOL (The PyMOL Molecular Graphics 1303

The complete cryoEM data processing workflow is 1304 in Figure S2 and statistics of data collection, reconstruction and refinement is described 1305 in Table S3

Crystallization and Structure Determination of CV3-25 with S2 Stem Peptide 1307

10 mg/mL of CV3-25 1308 was mixed with synthetic S2 peptide spanning residues 1153-1163, 1153-1167 or 1140-1165 1309 (26mer) in a 1:10 molar ratio of Fab to peptide. Crystal screening of Fab-peptide complexes 1310 were performed using the vapor-diffusion hanging drop method using the sparse matrix 1311 crystallization screens ProPlex (Molecular Dimensions)

Crystals were 1315 snap-frozen in the crystallization condition supplemented with 20% 2-methyl-2, 4-pentanediol 1316 (MPD) as the cryoprotectant. X-ray diffraction data was collected at the SSRL beamline 9-2 and 1317 was processed with HKL3000

Iterative cycles of model building and refinement were done in 1320

Structural analysis and figure generation were 1321 performed in PyMOL and ChimeraX. Fab-peptide interface and buried surface area were 1322 determined in PISA (Krissinel and Henrick

Flow Cytometry Analysis of Cell-Surface Staining

At 48h post transfection, 293T cells were stained with anti-Spike monoclonal 1328 antibodies CV3-25, CV3-1 (5 µg/mL) or using the ACE2-Fc chimeric protein (20 µg/mL) for 45 1329 min at 37°C. Alternatively, to determine the Hill coefficients

Abs (Invitrogen) were used as secondary 1332 antibodies to stain cells for 30 min at room temperature. The percentage of transfected cells 1333 (GFP+ cells) was determined by gating the living cell population based on the basis of viability 1334 dye staining (Aqua Vivid, Invitrogen)

Hill coefficient 1336 analyses were done using GraphPad Prism version 9.1.0 (GraphPad). Alternatively, for peptide 1337 epitope competition assay, CV3-25 (5µg/mL) was pre-incubated in presence of increasing 1338 concentrations of peptide #288 (1149-KEELDKYFKNHTSPD-1163), peptide #289 (1153-1339 DKYFKNHTSPDVDLG-1167), a shorter version of peptide #289 (1153-DKYFKNHTSPD-1163) 1340 or a scramble version of the peptide #289 (DHDTKFLNYDPVGKS)

Viral Neutralization Assay

NIH AIDS Reagent Program) and a plasmid encoding 1346 for SARS-CoV-2 Spike at a ratio of 5:4. Two days post-transfection, cell supernatants were 1347 harvested and stored at -80°C until further use. 293T-ACE2 target cells were seeded at a 1348 density of 1×10 4 cells/well in 96-well luminometer-compatible tissue culture plates (Perkin Elmer) 1349 24h before infection. To measure virus neutralization, recombinant viruses in a final volume of 1350 100 μL were incubated with increasing concentrations of CV3-1 or CV3-25 (0.01 to 10 µg/mL) 1351 1h at 37°C and were then added to the target cells followed by incubation for 48h at 37°C; 1352 cells were lysed by the addition of 30 μL of passive lysis buffer (Promega) followed by one 1353 freeze-thaw cycle. An LB942 TriStar luminometer (Berthold Technologies) was used to measure 1354 the luciferase activity of each well after the addition of 100 μL of luciferin buffer (15 mM MgSO4, 1355 15 mM KH2PO4

Radioactive Labeling and Immunoprecipitation 1363

For pulse-labeling experiments, 5 × 10 5 293T cells were transfected by the calcium phosphate 1364 method with SARS-CoV-2 Spike expressors. One day after transfection, cells were 1365 metabolically labeled for 16 h with 100 μCi/ml [ 35 S]methionine-cysteine ([ 35 S] protein labeling 1366 mix; Perkin-Elmer) in Dulbecco's modified Eagle's medium lacking methionine and cysteine and 1367 supplemented with 10% of dialyzed fetal bovine serum and 1X GlutaMAX TM

Cells were subsequently lysed in radioimmunoprecipitation assay (RIPA) buffer (140 1369 mM NaCl, 8 mM Na2HPO4

Precipitation of radiolabeled SARS-CoV-2 Spike glycoproteins from cell lysates or supernatant 1372 was performed with CV3-25 in combination with a polyclonal rabbit antiserum raised against 1373 SARS-CoV-2 RBD

SARS-CoV-2 Spike peptide ELISA (enzyme-linked immunosorbent assay) The SARS-CoV-2

Peptides covering the entire SARS-CoV-2 S2 sequence with a length of 15 residues (15-mer) 1379 and an overhang of 4 residues were purchased from JPT Peptide Technologies. Briefly

CoV-2 S2 peptide pools or individual peptides (1 μg/ml), or bovine serum albumin (BSA) (1 1381 μg/ml) as a negative control

Nunc) overnight at 4 °C. Coated wells were subsequently blocked with blocking buffer

CV3-1385 25 mAb (50 ng/ml) was prepared in a diluted solution of blocking buffer (0.1 % BSA) and 1386 incubated with the peptide-coated wells for 90 minutes at room temperature. Plates were 1387 washed four times with washing buffer followed by incubation with HRP-conjugated anti-IgG 1388 secondary Abs (Invitrogen) (diluted in a diluted solution of blocking buffer [0.4% BSA]) for 1 hour 1389 at room temperature, followed by four washes. HRP enzyme activity was determined after the 1390 addition of a 1:1 mix of Western Lightning oxidizing and luminol reagents (Perkin Elmer Life 1391 Sciences). Light emission was measured with a LB942 TriStar luminomete

Western Blotting

293T-S cells were seeded in 6-well plates at a density of 1x10 6 cells per 1396 well on day 0. On day 1, cells were either induced with 1 µg/ml doxycycline or mock treated as a 1397 control

Cell lysates were subjected to Western blotting using the 1399 CV3-1 or CV3-25 antibodies; mouse anti-S1 antibody (Sino Biological) and rabbit anti-S2 1400 antibody (Sino Biological) were used as controls. The Western blots were developed with 1401 horseradish peroxidase (HRP)-conjugated secondary antibodies (anti-human IgG

To evaluate antibody recognition of S glycoproteins 1403 lacking N-linked glycans, 293T-S cells expressing the wild-type SARS-CoV-2 S glycoprotein 1404 were lysed with lysis buffer, as described above. Lysates were treated with PNGase F (NEB) 1405 following the manufacturer's instructions or mock treated as a control

Quantification and Statistical Analysis

To obtain statistical significance for survival curves, grouped data were 1411 compared by log-rank (Mantel-Cox) test. To obtain statistical significance for grouped data we 1412 employed one-way ANOVA with a Holm-Sidak post-test (Figure 1A and B, 3F and G) or 1413 2-way ANOVA followed by Tukey's multiple comparison tests (Figure 1D)

Schematics 1417

Schematics for showing experimental design in figures were created with BioRender.com. 1418 1419 Supplementary Information

Flexible Fitting of CV3-1 Bound SARS-CoV-2 Spike cyroET Structure with 3-RBD-1421 down atomic model (PDB 7LWS)

Flexible Fitting of two CV3-25 Bound SARS-CoV-2 Spike cyroET Structure with 1423 prefusion S (PDB 6XR8) superimposed with two CV3-25 models at the stem helix

Grouped data in (D) were analyzed by 2-way ANOVA followed by Tukey's multiple comparison 875 tests. Statistical significance for group comparisons to isotype control are shown in black and for 876 those to CV3-25 GASDALIE are shown in blue. * , p < 0.05; * * , p < 0.01; * * * , p < 0.001; * * * * , p 877 Further information and requests for resources and reagents should be directed to and will be 1035 fulfilled by the lead contact, Walther Mothes (walther.mothes@yale.edu). 1036

All reagents generated in this study will be made available by the lead contact with a completed 1038Materials Transfer Agreement. 1039 

Lentiviral particles were collected and clarified by low-speed spinning (1500g for 5 min) twice, 1108 then pelleted by ultracentrifugation (130,000g for 2 hour) once and resuspended in PBS buffer. 1109 6 nm gold tracer was added to the concentrated S-decorated HIV-1 lentivirus at 1:3 ratio, and 5 1110 µl of the mixture was placed onto freshly glow discharged holey carbon grids (R 2/1, Quantifoil) 1111 for 1 min. Grids were blotted with filter paper, and plunge frozen into liquid ethane by a 1112 homemade gravity-driven plunger apparatus. Frozen grids were stored in liquid nitrogen until 1113imaging. 1114

Cryo-grids were imaged on a cryo-transmission electron microscope (Titan Krios, Thermo 1116Fisher Scientific) operated at 300 kV, using a Gatan K3 direct electron detector in counting 1117 mode with a 20 eV energy slit. Tomographic tilt series between -60˚ and +60˚ were collected by 1118using Mastronarde and Held, 2017) with increments of 3˚. The nominal magnification was 64,000 X, 1120giving a pixel size of 1.346 Å on the specimen. The raw images were collected from single-axis 1121 J o u r n a l P r e -p r o o f tilt series with accumulative dose of ~120e per Å 2 . The defocus range was -2 to -6 µm and 9 1122 frames were saved for each tilt angle. Detailed data acquisition parameters are summarized in 1123  Table S1 . 1124Frames were motion-corrected using Motioncorr2 (Zheng et al., 2017) to generate drift-1125 corrected stack files, which were aligned using gold fiducial makers by IMOD/etomo 1126 (Mastronarde and Held, 2017). The contrast transfer function (CTF) was measured by the 1127 ctfplotter package within IMOD. Tilt stacks were CTF-corrected by ctfphaseflip within IMOD. 1128Tomograms were reconstructed by weighted back projection and tomographic slices were 1129 visualized with IMOD. 1130

For the CV3-1 sample, all spikes were manually picked. Euler angles were determined based 1132 on the vector between two points, one on the head of the spike and the other on the membrane 1133where the spike locates. For CV3-25 and unliganded samples, a low-pass filtered (30Å) 1134 structure from previous aligned S structure was used as the template for template matching 1135 search in 8 x binned tomograms. Subtomograms were extracted for initial alignment. After this 1136 alignment, particles with cross-correlation coefficients (CCC) below 0.25 were removed. Visual 1137 inspection of the tomograms in IMOD confirmed that the rest of the subtomograms 1138 corresponded to S trimers on the viral surface. Particles that had tilted by more than 90° relative 1139 to their perpendicular positions to the viral surface were excluded. Subsequent processing was 1140 performed by using I3 (Winkler, 2007) with 2 x and 4 x binned tomograms. 1141All the density maps were segmented in the UCSF Chimera 

For all in vivo experiments, the 6 to 8 weeks male and female mice were intranasally challenged 1147with 1 x 10 5 FFU in 25-30 µL volume under anesthesia (0.5 -5 % isoflurane delivered using 1148precision Dräger vaporizer with oxygen flow rate of 1 L/min). For NAb treatment using 1149prophylaxis regimen, mice were treated with 250 µg (12.5 mg/kg body weight) of indicated 1150 antibodies (CV3-1 or CV3-25 GASDALIE) via intraperitoneal injection (i. introduced into designed positions in the S1 subunit on the plasmid encoding SB.1.1.7, pCMV-1217 SB.1.1.7. Plasmids pCMV-SB.1.1.7, dual-tagged pCMV-SB.1.1.7 Q3-1 A4-1, and pCMV delta R8.2 1218were transfected into 293T cells at a ratio of 20:1:21. Using this very diluted ratio of tagged-S vs. 1219 wildtype S, for the virus particles containing tagged S, more than 95 % S trimers will have one 1220 dual-tagged protomer and two wildtype protomers within a trimer. Using this strategy, we 1221 generated lentiviral particles with an average of one dual-tagged S protomer for conjugating 1222FRET-paired fluorophores among predominantly wildtype S trimers presented on lentivirus 1223surface. Viral particles were harvested 40 h post-transfection, filtered with a 0.45 μm pore size 1224 filter, and partially purified using ultra-centrifugation at 25,000 rpm for 2 h through a 15 % 1225 sucrose cushion made in PBS. Then the particles were re-suspended in 50 mM pH 7.5 HEPES 1226buffer, labeled with self-healing Cy3 and Cy5 derivatives (LD555-CDand LD650-CoA, 1227 respectively) and purified through an Optiprep TM (Sigma Aldrich) gradient as previously Highlights  Antibodies CV3-1 and CV3-25 neutralize emerging SARS-CoV-2 variants  CV3-1 binds to 485-GFN-487 Loop of RBD on spike and triggers S1 shedding  CV3-25 binds the stem helix of S2 and inhibits membrane fusion  Conserved epitopes are candidates for pan-coronavirus vaccines

Li et al. elucidate the structural basis and mode of action for two potent anti-Spike neutralizing monoclonal antibodies that remain effective against SARS-CoV-2 emerging variants of concern. Vaccine immunogen designs based on both conserved epitopes are candidates to elicit pancoronavirus protective immune responses.