Antibodies
The human antibodies studied in this paper were isolated from 39 SARS-CoV-2 convalescent patient plasma in China and the original clinical studies to obtain blood samples after written informed consent were previously described29 and had been approved by the Ethics Board of ChongQing Medical University. The antibodies were isolated using flow sorting for isolation and cloning of single antigen-specific B cells and the antibody variable genes encoding monoclonal antibodies29.
Recombinant antibody production and purification
A pair of plasmids separately expressing the heavy- and light- chain of antibodies were transiently co-transfected into Expi293™ cells (Catalog No. A14528, ThermoFisher) with ExpiFectamine™ 293 Reagent. Then the cells were cultured in shaker incubator at 120 rpm and 8% CO2 at 37℃. After 7 days, the supernatants with the secretion of antibodies were collected and captured by protein G Sepharose (GE Healthcare). The bound antibodies on the Sepharose were eluted and dialyzed into phosphate-buffered saline (PBS). The purified antibodies were used in following binding and neutralization analyses.
Sequence analysis of antigen-specific mAb
IMGT/V-QUEST (http://www.imgt.org/ IMGT_vquest /vquest) and IgBLAST (https://www.ncbi.nlm.nih.gov/igblast/), MIXCR (https://mixcr.r eadthedocs.io/en/master/) and VDJtools (https://vdjtools-doc.readthedocs.io/en/master/overlap.html) tools were used to do the variable region analysis and annotation for each antibody clone.
The competition with ACE2 measured by SPR
To determine the competition with the human ACE2 peptidase domain, SARS-CoV-2 S-RBD was coated on a CM5 sensor chip via amine group for a final RU around 250. Antibodies (20 μg/mL) were injected onto the chip until binding steady-state was reached. ACE2 (20 μg/mL) was then injected for 60 seconds. Blocking efficacy was determined by comparison of response units with and without prior antibody incubation.
Competitive ELISA
For the competitive ELISA used in epitope mapping of mAbs, 2 μg/mL recombinant S-RBD-his (Sino Biological, Beijing, China) was added in 384-well plates and incubated at 4 °C overnight. 50 μg/mL mAbs per well were added. The plates were incubated at 37°C for 1 h and then washed. Biotinylation of mAbs (the 20 neutralizing Abs and 81A11, previously reported SARS-CoV CR302232 were performed using the EZ-link NHS-PEO Solid Phase Biotinylation Kit (Pierce) according to the manufacturer’s protocol and purified using MINI Dialysis Unit (ThermoFisher, 69576). 500 ng/mL biotinylated mAbs were added to each well, and the plates were incubated at 37°C for 1 h. ALP-conjugated Streptavidin (Mabtech, Sweden, 3310-10) was added at 1:1000, followed by an incubation of 30 min at 37°C. Quantification step was the same as described above.
Peptide ELISA
Peptide ELISA was performed with synthesized peptides overlapping with 5 amino acids (Genescripts, Wuhan, China). These peptides were tethered by N-terminal biotinylated linker peptides (biotin-ahx), except for the first peptide at the N-terminus, whose biotin was linked to the C terminus instead. The RBD9-1 amino acid residues were selected and mutated to alanine and synthesized by Genescripts (Wuhan, China). 50 μL synthesized peptide was added to the streptavidin-coated 384-well plate in duplets to make a final concentration of 5 μg/mL. The plates were incubated for 2 hrs at RT. After washing, the plates were blocked with Protein-Free Blocking Buffer (Pierce, USA, 37573) at RT for 1 h and incubated with 10 μg/mL testing mAbs at RT for another 1 h. Reacted mAbs were detected using ALP-conjugated Goat F(ab')2 Anti-Human (IgG (Fab')2) secondary antibody (Abcam, ab98532, 1:2000) for 30 min at RT, followed with quantification detection.
For the ACE2 competitive peptide ELISA, 5 μg/mL synthesized RBD9-1 was immobilized on the streptavidin-coated 384-well plate at RT for 2 hrs. After washing with Protein-Free Blocking Buffer, the plates were blocked with this blocking buffer. Next, serial diluted 58G6 (20-0.625 μg/mL) in 50 μL of the blocking buffer were added into plate and the plates were incubated at RT for 1 h. Then, the plate incubated with 2 μg/mL ACE2 at RT for another 1 h. The ELISA plates were washed 4 times by blocking buffer and 50 μL Goat F(ab')2 Anti-Human (IgG (Fab')2) secondary antibody conjugated with ALP (Abcam, ab98532, 1:2000) was incubated with the plate at RT for 30 min. The plate was washed and followed with quantification detection.
Production of S protein pseudovirus
pVSVG expressing SARS-CoV-2 spike was constructed as previously described33. The packaging plasmid (VSV-G pseudotyped ΔG-luciferase) encoding either SARS-CoV-2 S, 501Y.V1 S and chimeric construct including 501Y.V2 S-RBD and SD614G was generated. HEK293T cells were grown to 80% confluency before transfection with VSV-G pseudotyped ΔG-luciferase, pWPXL and pSPAX2. Cells were cultured overnight at 37°C with 5% CO2. DMEM supplemented with 5% fetal bovine serum and 100 I.U./mL of penicillin and 100 μg/mL of streptomycin were added to the inoculated cells, which were cultured overnight for 72 hrs. The supernatant was harvested, filtered by 0.45 μm filter and centrifugated at 300 g for 10 mins to collect the supernatant and then aliquoted and storied at -80°C.
Pseudovirus neutralization assay
50 μL serially diluted mAbs were incubated with the same volume of the HEK293T cell supernatants containing pseudovirus for 1 h at 37 °C. These pseudovirus-antibody mixtures were added to ACE2 expressing HEK293T cells (HEK293T/ACE2). After 72 hrs, the luciferase activities of infected HEK293T/ACE2 cells were detected by the Bright-Luciferase Reporter Assay System (Promega, E2650). The IC50 of the evaluated mAbs were tested by the Varioskan LUX Microplate Spectrophotometer (Thermo Fisher), and calculated by a four-parameter logistic regression using GraphPad Prism 8.0.
Authentic SARS-CoV-2 neutralization assay
This assay was performed in a biosafety level 3 laboratory of Fudan University. Serially diluted mAbs were incubated with authentic SARS-CoV-2 (nCoV-SH01, GenBank: MT121215.1, 100 TCID50) for 1 h at 37°C. After the incubation, the mixtures were then transferred into 96-well plates, which were seeded with Vero E6 cells. The plates were kept at 37°C for 48 hrs and each well was visually assessed for CPE. And the supernatant viral RNA load of each well was quantified by RT-qPCR.
For RT-qPCR, the viral RNA was extracted from the collected supernatant using Trizol LS (Invitrogen) and used as templates for the RT-qPCR analysis by Verso 1-Step RT-qPCR Kit (Thermo Scientific) following the manufacturer’s instructions. PCR primers targeting SARS-CoV-2 N gene (nt 608-706) were as followed, forward: 5’-GGGGAACTTCTCCTGCTAGAAT-3’, and reverse: 5’-CAGACATTTTGCTCTCAAGCTG-3’. RT-qPCR was performed using the LightCycler 480 II PCR System (Roche) with the following program: 50°C 15 min; 95°C 15 min; 40 cycles of 95°C 15 sec, 50°C 30 sec, 72°C 30 sec.
Western blot analysis
The recombinant S-RBD protein was mixed with 5× loading buffer (Beyotime, Shanghai, China) and denatured for 5 min at 100°C. Denatured proteins (200 ng) were subjected to electrophoresis with 10% SDS-polyacrylamide gel and then transferred to PVDF membranes. After blocking by skim milk (Biofroxx), the membranes were incubated at 4 °C overnight, with the purified mAbs as primary Abs. The next days, the membranes were washed with TBST and incubated with HRP-conjugated Goat-anti-human Fc antibody (Abcam, ab99759, 1:10000) for 1 h at RT. The membranes were examined on Bio-rad ChemiDoc Imaging System (Bio-rad).
Protein expression and purification
To express the prefusion spike ectodomain, the gene encoding residues 1-1208 of SARS-CoV-2 spike (GenBank: MN908947.3) with a C-terminal T4 fibritin trimerization motif, an HRV-3C protease cleavage site, a Twin-Strep-tag and an 8×His-tag was synthesized, and cloned into the mammalian expression vector pcDNA3.1, which was a kind gift from L. Sun at Fudan University, China. The gene of spike protein was constructed with proline substitutions at residues 986 and 987, a “GSAS” instead of “RRAR” at the furin cleavage site (residues 682-685) according to Jason S. McLellan’s research12.
Expi293 cells (Thermo Fisher Scientific, USA) cultured in Freestyle 293 Expression Medium (Thermo Fisher Scientific, USA) were maintained at 37 °C. Cells were diluted to a density of 2.5 × 106 to 3 × 106 cells per ml before transfection. For protein production, 1.2 mg DNA was mixed with 3 mg polyethyleneimine in 30 mL Freestyle 293 Expression Medium, incubated for 20 min, then added to 1000 mL of cells 34. Transfected cells were cultured at 35 °C, and the cell culture supernatant was collected at day 4 to day 5.
Protein was purified from filtered cell supernatants using Strep-Tactin resin (IBA) before being subjected to additional purification by gel filtration chromatography using a Superose 6 10/300 column (GE Healthcare, USA) in 1×PBS, pH 7.4 (Extended Data Fig. 5a, b).
Cryo-EM sample preparation and data collection
Purified SARS-CoV-2 spike was diluted to a concentration of 1.5 mg/mL in PBS, pH 7.4. 5 μL of purified SARS-CoV-2 S was mixed with 1 μL of 58G6 Fab fragments at 2 mg/ml in PBS and incubated for 30 min on ice. A 3 μL aliquot of the mixture (added with 0.01% DDM) was applied onto an H2/O2 glow-discharged, 300-mesh Quantifoil R1.2/1.3 grid (Quantifoil, Micro Tools GmbH, Germany). The grid was then blotted for 3.0 s with a blot force of -1 at 8 °C and 100% humidity and plunge-frozen in liquid ethane using a Vitrobot (Thermo Fisher Scientific, USA). Cryo-EM data sets were collected at a 300 kV Titan Krios microscope (Thermo Fisher Scientific, USA) equipped with a K3 detector (Gatan, USA). The exposure time was set to 2.4 s with a total accumulated dose of 60 electrons per Å2, which yields a final pixel size of 0.82 Å. 2605 micrographs were collected in a single session with a defocus range comprised between 1.0 and 2.8 μm using SerialEM. The sample preparation and data collection for the SARS-CoV-2 S-13G9 Fab complex were in accordance with the SARS-CoV-2 S-58G6 Fab complex. The statistics of cryo-EM data collection can be found in Extended Data Table 2.
Cryo-EM data processing
All dose-fractioned images were motion-corrected and dose-weighted by MotionCorr2 software35 and their contrast transfer functions were estimated by cryoSPARC patch CTF estimation36. For the dataset of SARS-CoV-2 S-58G6 Fab complex, a total of 1,255,599 particles were auto-picked using the template picker and 820,872 raw particles were extracted with a box size of 512 pixels in cryoSPARC36. The following 2D, 3D classifications, and refinements were all performed in cryoSPARC. 237,062 particles were selected after two rounds of 2D classification, and these particles were used to do Ab-Initio reconstruction in six classes. Then these six classes were used as 3D volume templates for heterogeneous refinement with all selected particles, with 108,020 particles converged into the SARS-CoV-2 S-58G6 Fab class. Next, this particle set was used to perform non-uniform refinement, yielding a resolution of 3.56 Å.
For the dataset of SARS-CoV-2 S-13G9 Fab complex, a total of 445,137 particles were auto-picked using the template picker and 266,357 raw particles were extracted with a box size of 512 pixels in cryoSPARC. The following 2D, 3D classifications, and refinements were all performed in cryoSPARC. 70,519 particles were selected after two rounds of 2D classification, and these particles were used to do Ab-Initio reconstruction in six classes. Then these 6 classes were used as 3D volume templates for heterogeneous refinement with all selected particles, with 52,880 particles converged into the SARS-CoV-2 S-13G9 Fab class. Next, this particle set was used to perform non-uniform refinement, yielding a resolution of 3.92 Å.
Local resolution estimation, filtering, and sharpening were also carried out using cryoSPARC. The full cryo-EM data processing workflow is described in Extended Data Fig. 5 and the model refinement statistics can be found in Extended Data Table 2.
Model Building and Refinement
To build the structures of the SARS-CoV-2 S-58G6 Fab and spike-13G9 Fab complexes, the structure of the SARS-CoV-2 S glycoprotein in complex with the C105 neutralizing antibody Fab fragment28 (PDB: 6XCN) was placed and rigid-body fitted into the cryo-EM electron density maps using UCSF Chimera37, respectively. Both of the 58G6 and 13G9 Fab models were first predicted using Phyre230 and then manually built in Coot 0.938 with the guidance of the cryo-EM electron density maps, and overall real-space refinements were performed using Phenix 1.1839. The data validation statistics are shown in Extended Data Table 2.
Creation of Figures
Figures of molecular structures were generated using PyMOL40 and UCSF ChimeraX41.
Data analysis
Data are shown as mean ± SEM. Two-group comparisons were performed by Student’s t-test. The difference was considered significant if p < 0.05.