Molecular deconvolution of the neutralizing antibodies induced by an inactivated SARS-CoV-2 virus vaccine

The rapid emergence and persistence of the pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has had enormous impacts on global health and the economy. Effective vaccines against SARS-CoV-2 are urgently needed to control the coronavirus disease 2019 (COVID-19) pandemic, and multiple vaccines have been found to be efficacious in preventing symptomatic COVID-19 (Polack et al., 2020; Wu et al., 2020; Jones and Roy, 2021). We have developed a traditional beta-propiolactone-inactivated aluminum hydroxide-adjuvanted whole-virion SARSCoV-2 vaccine (BBIBP-CorV), which elicited protective immune responses in clinical trials (Wang et al., 2020; Xia et al., 2021). The vaccine has been granted conditional approvals or emergency use authorizations (EUAs) in China and other countries. The spike protein (S protein) is the main target of the humoral response during SARS-CoV-2 vaccination. The S protein is located on the surface of the SARS-CoV-2 virion, is involved in the entry step of virus infection and consists of two subdomains: the N-terminal S1 domain, which contains the N-terminal domains (NTDs) and the receptor-binding domain (RBD) that recognizes the host cell receptor angiotensin-converting enzyme 2 (ACE2), and the S2 domain responsible for fusion between the virus and cell membranes. Philip J. M. Brouwer isolated monoclonal antibodies from three convalescent COVID-19 patients using a SARSCoV-2 spike protein and revealed that the SARS-CoV-2 spike protein contains multiple distinct antigenic sites, which could provide guidance for vaccine design (Brouwer et al., 2020). The serological response after viral infection or vaccination is composed of a mixture of antibodies against different antigenic domains of the virus. Currently, serological assays are used to monitor the antibody response following vaccination (Anderson et al., 2020; Wang et al., 2020). Molecular deconvolution of the antibody repertoire after vaccination could provide a more complete understanding of the effectiveness and mechanism of the vaccines than conventional methods. Cloning of individual B cells isolated by fluorescence-activated cell sorting (FACS) has been used extensively to discover neutralizing antibodies from convalescent patients who have recovered from infections (Wen et al., 2020). Potent neutralizing antibodies that bind to the S protein of SARS-CoV-2 have been identified using these methods (Ju et al., 2020). SARS-CoV-2-neutralizing antibodies were also discovered by single-cell VDJ sequencing of antigen-enriched B cells from convalescent patients (Cao et al., 2020). The single-cell sequencing method allows simultaneous acquisition of B cell receptor (BCR) sequences and transcriptomic information, with the cognate heavy and light chains of antibodies determined bioinformatically. The selected antibodies need to be synthesized and expressed for further characterization, which is well suited for fast antibody identification and development. Recently, a microfluidics-based technology was developed to physically link the variable region of the heavy chain (VH) and variable region of the light chain (VL) from the same B cell (Wang et al., 2018). The resulting natively paired VH:VL antibody library can be directly screened using phage display or yeast display to isolate antibody clones specific to different antigens (Lerner, 2006). This method has been used to discover antiinfection antibodies, including broadly neutralizing antibodies (bNAbs) specific to HIV-1, Ebola virus and influenza virus (Rajan et al., 2018; Wang et al., 2018). In addition, as complete sets of VH and VL genes are preserved in their natural pairing, this method is well suited for characterization of the immune repertoire. Two individuals (Table. S1) with no prior SARS-CoV-2 infection history were vaccinated with the two-dose SARS-CoV-2 vaccine BBIBPCorV, and blood was collected two months after the 2nd dose of vaccine (Fig. 1A). Plasma from both donors demonstrated strong binding to the SARS-CoV-2 S protein and effective neutralizing activity against 2 strains of live SARS-CoV-2 (Fig. 1B and 1C). We investigated the B cell response to vaccination by sequencing heavy chain variable regions of antibodies. The mean somatic hypermutation (SHM, or germline divergence) of donor 1 and donor 2 was 4.68% and 5.69%, respectively (Fig. 1D). The SHM of naïve donors was 3.25%, which was

convalescent patients were assigned to VH1 to VH4, we designed primers specifically targeting VH1 to VH4 to increase the specificity of amplification. Therefore, this study analyzed only VH1-4-derived antibodies.
For nested PCR, the product from the last step was first electrophoresed on a 1.7% agarose gel, and then the region between 800 bp and 1200 bp was excised and purified using NucleoSpin Gel and a PCR Clean-up Kit (Macherey-Nagel). Nested PCR was performed using new primer sets designed to anneal to the FR1 region of VH and CL to increase the specificity and add restriction sites for cloning into phagemid. PCR products were run on a 1.2% agarose gel, and the band at 900 bp was excised and purified using NucleoSpin Gel and a PCR Clean-up Kit (Macherey-Nagel).

Phage display library construction and preparation
The DNA amplicons from nested PCR were digested and ligated into pCGMT3 phagemid and transformed by electroporation into E. coli XL1-Blue. After transformation, prewarmed SOC medium was added, and the culture was shaken for 1 h at 37°C, after which 2×YT medium containing 2% glucose, 50 µg/mL carbenicillin and 10 µg/mL tetracycline was added, and the culture was then shaken for an additional hour. Then, helper phage VCMS13 was added, and the culture was shaken for an additional 2 h. Kanamycin was added to 70 µg/mL, and the culture was incubated at 30°C overnight.

Phage display to isolate SARS-CoV-2 S protein binding antibodies.
The phage library was incubated with biotinylated SARS-CoV-2 S protein (Acro Biosystems) for 2 h at room temperature, and the phage-antigen complex was captured by Dynabeads M280 (Life Technologies). Bound phage was eluted using glycine-HCl (pH 2.2) for 10 min at room temperature, and the pH of the eluate was adjusted to pH 7.5 (neutralized) with Tris-HCl (pH 8.0). The eluted phage was used to infect XLI-Blue cells at OD600 = 0.5 for 30 min at 37°C, after which 50 mL of 2×YT medium containing 2% glucose, 50 µg/mL carbenicillin and 10 µg/mL tetracycline was added; the culture was then shaken for 1 h at 37°C. Further growth, phage preparation, and panning were repeated as described above. After 2-3 rounds of panning, individual phage clones were amplified for phage ELISA. For ELISA, 100 ng NTD (Acro Biosystems), RBD or S2 domain (Acro Biosystems) protein was immobilized on 96-well plates overnight and blocked for 40 min with 4% M-PBST (PBS+4% skim milk). Following washing with PBST (PBS + 0.05% Tween-20), diluted phage supernatant was added, and phage bound to the coated antigens was detected using an anti-M13-HRP antibody (1:5000) and ABTS as substrate.

Antibody expression and purification
Phagemid DNA from positive clones was isolated and sequenced. Antibodies were cloned into mammalian expression vectors. The antibodies were expressed transiently in HEK293F cells. Cells were transfected at a density of 0.8-1.2 million cells/mL by addition of a mix of PEIMAX (1 μg/μL) with expression plasmids. Supernatants were harvested at four days post transfection. Antibodies were purified by affinity purification using a Protein A bead (Merck) suspension according to the manufacturer's protocol for gravity flow. These monoclonal antibodies were eluted with 50 mM glycine-HCl pH 2.7 and neutralized with 1 M Tris/HCl pH 8.5. Protein eluates were concentrated, and the buffer was exchanged to PBS using a protein ultrafiltration tube (Millipore) with a 30 kDa molecular weight cutoff. Protein concentrations were determined by BCA assay.

ACE2 competition assay
HEK293T cells were transfected with the extracellular domain of the ACE2-expressing plasmid. HEK293T cells displayed ACE2 at 4 h after transfection. Before assaying competition with ACE2, the biotinylated spike RBD protein was blocked by Fab or BSA (1:10) at 37°C for 1.5 h. Then, the cells were stained with the mixture at 4°C for 1 h. After washing three times with ice-cold FACS buffer, the cells were incubated with SA-PE (1:200 dilution, Thermo Fisher) and incubated at RT for 30 min. After the final wash, the cells were analyzed by flow cytometry (LSR Fortessa, BD).

Measurement of antibody affinity for S proteins
The binding kinetics of anti-RBD antibodies were determined by biolayer interferometry on a ForteBio Octet RED96 instrument (Sartorius AG). In brief, his-tag RBD or S1 protein (wild type or mutant) (Tab. S4) was immobilized onto a Ni-NTA biosensor (Forte Bio cat no 18-5019) at 1 µg/mL for 180 s, and biosensors were then equilibrated for 180 s in PBST (PBS+0.05% Tween). The association of serial dilutions of antibodies was measured over 420 s, followed by 600 s dissociation in PBST. All the data were analyzed by "Data Analysis 10.0" from ForteBio (Sartorius AG), and the curve fitting model was 1:1 for both association and dissociation. Reference samples were subtracted to correct for nonspecific baseline drift.

Epitope binning of antibodies
The Avi-RBD-immobilized SA biosensors were incubated with the first antibodies at 200 nM for 600 s to equilibrate and then submerged in 200 nM second antibodies for 600 s after 180 s baseline in PBST, followed by running with reference samples and buffer. The matrix data were exported by "Data Analysis 10.0" from ForteBio (Sartorius AG) for external analysis of the self-binding signal threshold. The nonoverlapping epitopes were indicated by a ≥100% signal threshold compared with the self-activity signal.       to the reference antibodies and sequence divergence (%, x axis) from their putative germline genes. Color coding indicates sequence density on the 2D plot. The six neutralizing antibody lineages are visible as a distinct island of sequences with higher identity. The clones with germline divergence above 40% accounted for only less than 0.03% of total clones (367/3543964, 330/1916964 and 210/766607 for donor 1, donor 2 and the naïve donor, respectively).

Figures S5-8: Plot curves of antibody binning for RBD and S1 protein mutants.
The title of each plot shows the antibody ID/mutant ID and KD (M) value of the corresponding antibody against the mutant.