key: cord-0704430-k6gntx8i
authors: Wong, Matthew K.; Liu, Jun T.; Budylowksi, Patrick; Yue, Feng Yun; Li, Zhijie; Rini, James M.; Carlyle, James R.; Zia, Amin; Ostrowski, Mario; Martin, Alberto
title: Convergent CDR3 homology amongst Spike-specific antibody responses in convalescent COVID-19 subjects receiving the BNT162b2 vaccine
date: 2022-03-05
journal: Clin Immunol
DOI: 10.1016/j.clim.2022.108963
sha: 480a7f074075febbc7f3819a014cddfa4f369806
doc_id: 704430
cord_uid: k6gntx8i

Convalescent coronavirus disease 2019 (COVID-19) subjects who receive BNT162b2 develop robust antibody responses against SARS-CoV-2. However, our understanding of the clonal B cell response pre- and post-vaccination in such individuals is limited. Here we characterized B cell phenotypes and the BCR repertoire after BNT162b2 immunization in two convalescent COVID-19 subjects. BNT162b2 stimulated many B cell clones that were under-represented during SARS-CoV-2 infection. In addition, the vaccine generated B cell clusters with >65% similarity in CDR3 V(H) and V(L) region consensus sequences both within and between subjects. This result suggests that the CDR3 region plays a dominant role adjacent to heavy and light chain V/J pairing in the recognition of the SARS-CoV-2 spike protein. Antigen-specific B cell populations with homology to published SARS-CoV-2 antibody sequences from the CoV-AbDab database were observed in both subjects. These results point towards the development of convergent antibody responses against the virus in different individuals.

Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), the beta-coronavirus responsible for coronavirus disease-2019 (COVID- 19) , has been a pressing public health concern ever since its global pandemic spread in early 2020. Rapid successes in vaccine development and distribution in the past year, including the emergency use authorization and deployment of mRNA vaccine BNT162b2 (Pfizer/BioNTech), have helped to impede the spread of the virus.

BNT162b2 is generally delivered under a two-dose prime-boost regimen, exhibiting ~95% vaccine efficacy when doses are administered 21 days apart [1, 2] . Vaccine mRNA, delivered intracellularly via a lipid nanoparticle shell as a vector, encodes for the membrane-anchored Spike protein found on the surface of SARS-CoV-2 viral particles [3, 4] . Production of this "foreign" Spike protein elicits a robust adaptive immune response, promoting the development of T cell help/cytotoxicity and B cell antibody responses against the virus that are equal to or stronger than a natural infection [2] .

For individuals infected with SARS-CoV-2, a robust immune response dominated by COVID-19 subjects, strong Spike-specific and neutralizing antibody responses appear rapidly following the initial mRNA vaccine dose, with several studies suggesting that additional booster doses may have limited immunological benefit in this subject population [8] [9] [10] [11] [12] [13] [14] [15] . This expediated protective immunity likely stems from the pre-existent SARS-CoV-2-specific B cell memory responses present in these individuals that may be driven by germinal centre formation, which produces high affinity, long-lasting memory B cells able to secrete virus-neutralizing antibodies [4, 16] .

Many studies have separately assessed the effects of both natural SARS-CoV-2 infection [17] [18] [19] [20] and prophylactic mRNA vaccination [17, [20] [21] [22] [23] on B cell populations by single-cell RNA sequencing. These next-generation sequencing technologies have facilitated the elucidation of clonality among specific B cell subsets [17, 20] , as well as the detection of sequence homologies amongst published SARS-CoV-2 clonotypes [21] within and between subject cohorts. However, there has been limited research thus far utilizing these techniques to analyze B cell responses in convalescent COVID-19 vaccinees. With the phenomenon of convalescent individuals requiring only a single dose to generate optimal protective immunity, further investigation into the transcriptomic and antibody phenotypic changes elicited by vaccination in this population is warranted.

In this study, we compared longitudinal changes in B cell populations, B cell transcriptomes, and BCR repertoires following BNT162b2 vaccination in two convalescent COVID-19 subjects. Higher levels of SARS-CoV-2 neutralization and Spike-reactive B cells were noted for both subjects after vaccination. Single-cell RNA sequencing of Spike-reactive B cells revealed that many B cell clones were activated following SARS-CoV-2 infection and BNT162b2 vaccination. Strikingly, many clones between subjects shared similar CDR3-loop amino acid sequences in the V H and V L regions post-vaccination, suggesting that clones with similar CDR3 regions are expanded by the vaccine and dominate SARS-CoV-2 Spike protein with R-10. Cells were cultured in R-10 with IL-2 and R848 (Mabtech) for 6 days and plated in Millipore MAIP plates (Sigma) pre-coated with anti-human IgG or IgA. Plates were incubated at 37°C for 16 hours; cells were washed away with D-PBS after incubation. Biotinylated SARS-CoV-2 Spike protein was added to the wells to detect Spike-specific B cells, and biotinylated anti-human IgG or IgA was added for detection of total IgG or IgA B cells. Streptavidin-HRP and TMB substrate were used for spot development. Spots were counted with ImmunoSpot Analyzer (CTL).

Subject sera were heat inactivated at 56°C for 30 minutes. Sera were diluted in serum-free media by a factor of 20, then serially diluted 6 steps at 2-fold intervals and incubated with 100 TCID50/mL of SARS-CoV-2 virus (Wuhan variant) for 1 hour at 37°C, 5% CO2. After incubation, VeroE6 cells were inoculated with the blood-virus mixture for 1 hour at 37°C, 5% CO 2 in quadruplicate. After incubation, the inoculum was removed and fresh DMEM with 2% FBS was added to the VeroE6 cells and returned to the incubator. The VeroE6 cells were incubated for 5 days and Cytopathic Events (CPE) was microscopically assessed. CPE was recorded in GraphPad Prism (GraphPad Software) and analyzed using a 4-parameter nonlinear regression to calculate the IC50. IC50 is defined as the serum concentration at which 50% of the VeroE6 cells are protected from infection.

PBMCs were thawed and treated with 100 ug/ml DNase1 for 15 minutes before staining.

Tetramers were stained alongside fixable viability dye-eFluor 780 (Thermofisher), CD3-Pacific Blue, CD19-FITC, CD27-PerCP Cy5.5, and IgD-BV605 antibodies (Biolegend or BD) for cell sorting stains. Spike-reactive B cells were sorted based on double positive tetramer staining and submitted to the Princess Margaret Genomics Centre (Toronto, ON) for single cell RNA sequencing. Single-cell library preparation and sequencing was performed using a 10X computational antibody discovery was based on the single-cell RNAseq gene expression and BCR sequencing data. Due to the low numbers of purified S-protein reactive cells, purified T cells were added as carriers to facilitate processing.

For GEX data, we used 10X genomics CellRanger (ver.3.1.0) with default settings [26] to demultiplex UMI GEM barcodes and embedded STAR [27] to align GEX data to the human genome version GRCh38-3.0.0 and used gene count and filtered barcodes as the default output of the pipeline for further analysis. We used the Seurat R package [28] as the main framework for single-cell transcriptome analysis. Cells with coverage on less than 200 genes and more than 6000 genes and cells with a mitochondrial RNA count of more than 20% were removed. We normalized cell counts using Seurat's NormalizeData with "LogNormalize" parameters and a scale factor of 10,000 and used 5000 most variant genes to perform PCA. The first 20 principal components were used to identify clusters in the UMAP space and further visualization. We used standard cell markers to identify carrier T cells (CD3E, CD4, CD8A, CD8B) and removed them from further transcriptome analysis. Specifically, we used FindIntegrationAnchors (30 dimensions, default parameters) to identify pairwise anchors between 4 samples and used

IntegrateData with (default parameters) to integrate our 4 data sets. We then scaled the data using ScaleData, computed the PCA using RunPCA, found clusters in 20 lower dimensions using FindNeighbors and FindClusters (resolution=1), and utilized RunUMAP for the visualization. To address whether S-protein enrichment altered the samples' B cell composition, we computed the ratio of the fraction of cells in each cluster normalized by the total number of cells in each sample after clustering the combined tonsillar GEX data and clustering in UMAP space.

Single-cell BCR analysis, pre-processing:

After pre-processing the single-cell BCR sequencing data, our datasets consisted of (440, 439, 31, 690) paired antibody sequences for pre-vaccination and post-vaccination of P1 and P2, respectively. We performed clonal analysis of BCR repertoires by clustering antibody sequences of all 4 samples together. We defined each pair of antibodies in the same clone if they matched for V and J genes in both chains, had equal-length CDR3 loops in both chains, and had at least an average of 90% amino-acid similarity (hamming distance) in their CDR3 regions of both the heavy and light chains.

Our publicly available reference antibody dataset for sequence homology search consisted of 2700 SARS-CoV-2 reactive antibodies in CoV-AbDab [30] . We used MAFFT to compute pairwise alignments between our BCR data to the reference dataset antibodies and selected any sequences for further analysis if they passed the sequence homology criterion of matched VH, JH, and VL genes and at least an average of 70% amino-acid similarity (hamming distance) in their CDR3 loops.

To determine the transcriptomic and phenotypic impact of BNT162b2 vaccination on convalescent COVID-19 subjects, we studied two previously infected COVID-19 subjects who received either one or two doses of BNT162b2. Both subjects were between 40-64 in age and both were symptomatic after contracting the virus; Subject 2 (P2) required hospitalization after infection ( Table 1) . Serum and PBMC samples were collected at several timepoints during the course of infection leading up to and following vaccination (Fig. 1A) . Subject 1 (P1) was followed for 10 months with samples collected only up to one week after the first dose of BNT162b2; P2 was followed for 2 months, with samples collected up to 5 days after the second immunization.

The ability of SARS-CoV-2 infection and BNT162b2 immunization to stimulate protective immunity was first assessed via serological assays. An increase in viral neutralizing capability was noted based upon a sharp increase in the IC50 sera titres in both subjects following vaccination measured using a SARS-CoV-2 live virus neutralization assay (Fig. 1B) .

Additionally, the frequency of different circulating B cell subsets was measured before and after vaccination by flow cytometry (Supplementary Fig. 1A,B) . The proportions of Spike-specific B cells (Fig. 1C ) and memory B cell Spike-specific immunoglobin frequency (Fig. 1D ) were increased after vaccination in both subjects, as measured by flow cytometry and ELISpot, respectively, compared to both pre-vaccination and a healthy control sample (flow cytometry only). This is consistent with prior studies identifying sharp increases in neutralizing antibody titres and Spike-specific IgG in double-dosed individuals, highlighting an induction of the memory response against the virus [2, 3] . Notably, P1who received one dose of BNT162b2 at the time of samplingdisplayed equal or greater antibody-specific responses compared to double-dosed P2, underscoring the potent existing immunological memory in convalescent individuals that has been repeatedly observed in similar studies [8] [9] [10] [11] [12] [13] [14] [15] .

Expansion of different memory B cell and plasmablast or plasma cell populations was also noted in Spike-specific B cells in both subjects after receiving BNT162b2 (Fig. 1E) . Both subjects displayed an increase in plasmablast/plasma (P1: 0% to 5.93%; P2: 0% to 7.84%) and activated memory (P1: 0% to 54.24%; P2: 10% to 25.49%) Spike-specific B cell populations after vaccination, emphasizing the strong recall response elicited by mRNA vaccination in convalescent COVID-19 subjects. Additionally, an increase in IgG-switched Spike-specific B cells was observed following immunization in P2 (Fig. 1F) , but this was not seen in P1 (Fig.   1F ), possibly because this subject was sampled shortly after their initial dose of the vaccine. IgAswitched Spike-specific B cells were also increased post-vaccination (Fig. 1F) , which can be linked to the simultaneous expansion of plasmablasts observed in these subjects, as these B cell subsets are known to be primarily IgA + when in steady-state circulation [19, 31] . To determine whether vaccination played a role in activating pre-existing SARS-CoV-2reactive B cell clones following infection, we performed single-cell RNA sequencing on SARS-CoV-2 Spike-specific B cells (10x Genomics) (Supplementary Fig. 1B) . We analyzed transcriptomes and B cell receptor (BCR) sequences of 451 pre-vaccine and 512 post-vaccine B cells for P1, and 30 pre-vaccine and 754 post-vaccine cells for P2 ( Fig. 2A, Supplementary Fig.   2A-B) . We integrated these 4 datasets for further analysis (Supplementary Data 1) . Seven distinct clusters of B cells were observed from the enriched populations based upon expression of typical surface markers, forming three memory (c0, c1, c3), one naïve (c5), one plasma (c6), and two plasmablast (c2, c4) cell subsets when all samples were pooled together ( Fig. 2A) .

Memory subset clusters c0, c1, and c3 were defined by high CD19, CD69, and MS4A1 expression combined with low IGHD expression (Fig. 2B) ; memory cluster c0 notably had high CD27 levels. Plasma cluster c6, plasmablast c2, and plasmablast c4 were defined by low CD19, MS4A1, and CD69, while displaying high CD27, CD38, XBP1, and IRF4 (Fig. 2B,   Supplementary Fig. 2C) ; plasmablast clusters c2 and c4 showed some PRDM1 expression while minimal expression was observed in the plasma c6 subset. Naïve cluster c5 B cells were classified primarily based on high IGHD expression (Fig. 2B) .

While vaccination resulted in expansion of plasmablast subsets in both subjects, increases in memory B cells were predominantly observed in P2 (Fig. 2C) . Differences in B cell subset expansion can be attributed to the number of doses received at the time of sampling as well as the stage of recovery -P1 was immunized only once 10 months post-symptom onset, while P2 received their second booster dose 2 months after diagnosis. Interestingly, an increase in the proportion of IgA+ B cells was also noted for both subjects following immunization (Fig. 2D) , a phenotype that was also apparent by flow cytometry, albeit at a lower frequency (Fig. 1F ).

To examine the B cell receptor (BCR) repertoire in SARS-CoV-2 Spike-reactive B cells pre-and post-vaccination, we assessed V H -J H usage in the sorted B cell population. In both subjects, we observed many BCR heavy chain gene pairs based on V H and J H usage ( Fig. 3A; light green, 0). However, following BNT162b2 vaccination, we observed selective expansion of J o u r n a l P r e -p r o o f Journal Pre-proof specific V H -J H heavy-chain gene pairs in both subjects ( Fig. 3A; darker green, 1-4) , underscoring a shift in gene segment usage following mRNA vaccination of convalescent subjects. We also observed several novel V H -J H heavy-chain gene pairs that were not observed pre-vaccination (dark green, 5) (Fig. 3A) . Additionally, there was a significant increase in non-synonymous mutations present in the BCR V-regions in several P1 subsets post-vaccination, especially within the memory and c4 plasmablast clusters (Fig. 3B) . This result corroborates prior studies that posit mRNA vaccination as a primer for somatic hypermutation in B cells via germinal centre formation [16, 17] . P1 also displayed higher non-synonymous mutation rates in almost every subset compared to P2 (Fig. 3B) , which may suggest that extended time between infection and vaccine boost (P1's 10 months vs P2's 2 months) could lead to higher affinity antibodyproducing B cells after vaccination due to increased germinal centre-mediated development of long-lived memory B cells. Collectively, these data show extensive usage of different V H -J H gene pairs in Spike-reactive B cells pre-and post-BNT162b2 vaccination, which suggests 1) that reactivity to Spike protein can be accomplished with many V H and J H elements, and 2) that a polyclonal, higher affinity B cell memory response is stimulated post-vaccination in convalescent COVID-19 subjects.

Vaccination of healthy (no prior COVID-19 history) individuals has been shown to induce the production of antibodies that undergo clonal expansion and that are homologous to known SARS-CoV-2 targeting antibodies elicited during natural viral infection [21, 32] . To determine whether these phenomena would similarly be observed in convalescent COVID-19 subjects, we compared the clonal structure of BCR repertoires of SARS-CoV-2 Spike-specific B cells in both individuals before and after vaccination. Antibodies were classified as clonal if they were matched for the V and J genes of both the heavy and light chains and had CDR3 regions of specific B cell clones in either subject, but instead appeared to result in the activation of many different B cell clones that appeared to vary from those elicited pre-vaccination. These data suggest that the clonal response to BNT162b2 immunization after SARS-CoV-2 infection is diverse and not dominated by specific clones. As vaccination of previously infected individuals should lead to specific amplification of pre-existing B cell memory cells, the apparent lack of a link between clones before and after vaccination is likely due to the high numbers of B cell clones involved in the response pre-and post-vaccination.

Interestingly, when B cells were clustered solely based on heavy and light chain CDR3 sequence similarity (>90% average amino acid similarity), several clusters were observed containing both pre-and post-vaccination clones from each subject (Fig. 4C) . Within these clusters, CDR3 heavy and light chain consensus sequences displayed high similarity (Supplementary Fig. 3 , Fig. 4A were driven by differences in the length of the CDR3 loop (average lengths shown in Supplementary Fig. 4) and that each cluster consisted of many different clones. However, some clusters (such as clusters 2 and 3 for P2; Supplementary Fig. 3B ) had clones with similar CDR3 sequences but not matching V/J regions (Supplementary Data 1C) .

This was also the case for the light chain CDR3 in P2 which were similar between cluster 1 and 2 ( Supplementary Fig. 3B ), but these two clusters use different VL and JL elements (Supplementary Data 1C). This suggests that V/J pairing was not solely responsible for CDR3 sequence similarity. CDR3 similarity was also observed between subjects, with some intersubject B cell clusters appearing when CDR3 homology was assessed at 65% amino acid similarity (Fig. 4D) . These CDR3 clusters were dominated by post-vaccine B cells in both subjects, indicating that this common CDR3 usage developed after vaccination. This result suggests that despite a lack of clonal dominance to SARS-CoV-2 Spike protein in this cohort pre-and post-vaccination, antigen specificity was maintained with similar paratopes encoded primarily by the CDR3 loops that were generated as a result of vaccination.

To assess possible similarities with antibodies from other subject cohorts and to investigate the potential functions of B cell clones from our subjects, we performed a virtual screening and homology assessment (Fig. 5A) against the publicly available CoV-AbDab J o u r n a l P r e -p r o o f Journal Pre-proof database [30] , which contains published SARS-CoV-2-specific antibodies generated after natural infection. BCRs were considered homologous to a published antibody if their V H , J H , and V L genes were matched and possessed at least 70% sequence similarity within their CDR3 loops (both heavy and light chains). A substantial number of BCRs were identified to be homologous to published antibodies, with the majority found within the plasmablast and memory B cell subsets in P1 and P2, respectively (Fig. 5B) . Both subjects harbored 27 or more antibody groups containing >4 BCRs (both clonal and non-clonal) homologous with different published antibodies after vaccination, many of which were known to be SARS-CoV-2 neutralizing, while P1 additionally had 20 antibody groups with this level of similarity even before their initial immunization (Fig. 5C) . Hundreds of other B cells in our subjects also displayed similar published clonotype homology, albeit with fewer BCRs (<4) per group. Most homologous subject antibody groups contained 4-11 members (Fig. 5D) , but one large group in P1 contained 46 BCRs that were homologous to a published monoclonal SARS-CoV-2 antibody, Ehling_mAb-47 [33] . Strikingly, seven of the largest clones (mostly IGHV4-4/IGHJ5, defined by >90% CDR3 similarity) in P1 displayed high amino acid sequence similarity to Ehling_mAb-47 in both their heavy and light chain CDR regions (five shown in Fig. 5E , four consensus sequences shown in Supplementary Fig. 3) . The discovery of similarities in clonal antigenspecific antibody sequences between our two subjects and with those of published SARS-CoV-2 antibodies from CoV-AbDab suggests that vaccination induces the development of antibody clonotypes with antigen specificity and variable region sequences that are similar to those found during natural infection. Moreover, it supports the case for convergent antibody maturation that has also been observed in uninfected mRNA vaccinees [21, 22, 32] .

Overall, these findings highlight a shift in B cell phenotype, function, and clonality following the vaccination that may favour the development of memory B cell responses that harbour similar variable region sequences across convalescent COVID-19 vaccinees.

Matching B cell heavy chain V and J genes and assessing sequence similarity in the heavy chain CDR3 (CDRH3) region are criteria typically used to identify similar B cell J o u r n a l P r e -p r o o f Journal Pre-proof clonotypes in different subject cohorts [19, 21, 22, 32, [34] [35] [36] [37] [38] . Clonotypes from both naïve BNT162b2 vaccinees [21, 22, 32] and convalescent unvaccinated individuals [19, [34] [35] [36] [37] [38] have been found to be homologous with published antibodies and/or among different individuals based on the above criteria, showing convergent development of conserved variable region genes and sequences that work to target SARS-CoV-2. Many of these clonotypes have been identified to specifically target RBD, with subsets of these antibody clones shown to be neutralizing [32, 34, 36] . Although clonality and homology of the SARS-CoV-2-specific BCR repertoire have been commonly reported in convalescent COVID-19 subjects and individuals immunized with mRNA vaccines, these assessments have not been thoroughly performed in convalescent BNT162b2 vaccinees. Here, we report that many B cell clones participate in the response to SARS-CoV-2 Spike protein as a result of SARS-CoV-2 infection and following BNT162b2 vaccination in our subject cohort. Furthermore, we noted CDR3 sequence-specific similarity between different convalescent BNT162b2 recipients (Fig. 4D) Antibody homology based on variable region gene matching and CDR3 sequence similarity was also noted with several published SARS-CoV-2 antibodies, many of which were known to have virus-neutralizing capability. In our cohort, sequence similarity with published antibodies was observed in both CDRH3 and the light chain CDR3 (CDRL3) variable regions.

Many of the clones that were homologous to Ehling_mAb-47 showed high amino acid sequence similarity especially in CDRL3, several of which were identical in these light chain sequences to the published antibody (Fig. 5E ). This contrasts with some reports that have noted CDRL3 as less conserved compared to its heavy chain counterpart, showing heterogeneity in the light chain amino acid sequences of unvaccinated convalescent subjects and naïve vaccinees [21, 35] , although this disparity may be due to differences in cohorts assessed. Overall, this remarkable similarity in CDR3 sequences strengthens the case for convergent antibody responses, which refers to the independent development of similar antibody variable region sequences targeting For B cell subsets induced following vaccination, disparity in expanded populations was evident between subjects likely due to differences in doses received and stages of recovery. We observed robust plasmablast expansion in P1, who received one dose at time of sampling, which contrasted with the maintenance of memory subsets identified in P2, who received two doses (Fig. 2C) . These results suggest that plasmablasts are primarily expanded 1 week after initial vaccination in convalescent COVID-19 subjects, but shift towards a memory B cell phenotype as early as 1 week after the booster dose; this is synchronous with observations seen in naïve double-dosed SARS-CoV-2 mRNA vaccinees [4, 17, 22] . Importantly, viral immunity was still elevated in P1 after their initial dose, as noted by high viral neutralizing capability and Spikespecific B cell frequency, indicating enhanced anti-SARS-CoV-2 immunity regardless of subsets expanded. Interestingly, one of the B cell subsets expanded post-vaccination was IgA + B cells, a population thought to primarily line mucosal sites. As mRNA vaccines are not mucosal vaccines, an increase in IgA + B cells was unexpected, but may also be linked to a simultaneous increase in plasmablast populations, which are known to be IgA+ in circulation (19, 31) . Further research into this expanded subset of cells could reveal informative details on the breadth of immunological stimulation by BNT162b2.

Overall, this study uncovered a polyclonal B cell response and high CDR3 sequence similarity in B cells following BNT162b2 vaccination in two convalescent COVID-19 subjects.

With the ongoing implementation of third booster vaccine doses to combat waning immunity in the most at-risk populations, further studies assessing the diversification of convergent antibody clonotypes and the long-term kinetics of virus-specific immune memory after subsequent doses will be crucial in the continuing control of SARS-CoV-2 and its variants. 

Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

BNT162b2 vaccine induces neutralizing antibodies and poly-specific T cells in humans

Safety and Immunogenicity of Two RNA-Based Covid-19 Vaccine Candidates

SARS-CoV-2 mRNA Vaccines: Immunological mechanism and beyond

Maturation and persistence of the anti-SARS-CoV-2 memory B cell response

Immunological memory to SARS-CoV-2 assessed for up to 8 months after infection

Distinct antibody and memory B cell responses in SARSCoV-2 naïve and recovered individuals following mRNA vaccination

Antibody response to the bnt162b2 mrna covid-19 vaccine in subjects with prior sars-cov-2 infection

Robust spike antibody responses and increased reactogenicity in seropositive individuals after a 1 single dose of SARS-CoV-2 mRNA vaccine 2 3

Effect of previous SARS-CoV-2 infection on humoral and T-cell responses to single-dose BNT162b2 vaccine

Antibody and T cell memory immune response after two doses of the BNT162b2 mRNA vaccine in older adults with and without prior SARS-CoV-2 infection

A single dose of the SARS-CoV-2 vaccine BNT162b2 elicits Fc-mediated antibody effector functions and T cell responses

Comparative kinetics of SARS-CoV-2 antispike protein RBD IgGs and neutralizing antibodies in convalescent and naïve recipients of the BNT162b2 mRNA vaccine versus COVID-19 patients

mRNA vaccination boosts cross-variant neutralizing antibodies elicited by SARS-CoV-2 infection

SARS-CoV-2 mRNA vaccine elicits a potent adaptive immune response in the absence of IFN-mediated inflammation observed in COVID-19

Single-cell multiomics analysis of the immune response in COVID-19

Rapid isolation and immune profiling of SARS-CoV-2 specific memory B cell in convalescent COVID-19 patients via LIBRA-seq

Single cell profiling of T and B cell repertoires following SARS-CoV-2 mRNA vaccine

Single-Cell Profiling of the Antigen-Specific Response to BNT162b2 SARS-CoV-2 RNA Vaccine

BNT162b2 Vaccine Induces Divergent B cell responses to SARS-CoV-2 S1 and S2

A Viral Immunoevasin Controls Innate Immunity by Targeting the Prototypical Natural Killer Cell Receptor Family

Tetramer Immunization and Selection Followed by CELLISA Screening to Generate Monoclonal Antibodies against the Mouse Cytomegalovirus m12 Immunoevasin

Massively parallel digital transcriptional profiling of single cells

STAR: Ultrafast universal RNA-seq aligner

Comprehensive Integration of Single-Cell Data

CoV-AbDab: The coronavirus antibody database

Blood-borne human plasma cells in steady state are derived from mucosal immune J o u r n a l P r e -p r o o f Journal Pre-proof responses

Convergent antibody responses to the SARS-CoV-2 spike protein in convalescent and vaccinated individuals

Single-Cell Sequencing of Plasma Cells from COVID-19 Patients Reveals Highly Expanded Clonal Lineages Produce Specific and Neutralizing Antibodies to SARS-CoV-2

Convergent antibody responses to SARS-CoV-2 in convalescent individuals

Global characterization of B cell receptor repertoire in COVID-19 patients by single-cell V(D)J sequencing

Human B Cell Clonal Expansion and Convergent Antibody Responses to SARS-CoV-2

Deep Sequencing of B Cell Receptor Repertoires From COVID-19 Patients Reveals Strong Convergent Immune Signatures

Massive surge of mRNA expression of clonal B-cell receptor in patients with COVID-19

Figure 1. Spike-specific B cell frequency and subsets are altered following vaccination in recovered SARS-CoV-2 subjects. (A) Disease and vaccination timeline for subjects in study cohort. Timepoints highlighted in blue were submitted for single cell RNA sequencing

Each sample was performed in quadruplicate. Post-vaccination timepoints are highlighted in red. (C) Representative flow cytometry graphs showing frequency of SARS-CoV-2 Spike-specific B cells from thawed peripheral blood mononuclear cells (PBMCs) stained with PE-/APC-labelled SARS-CoV-2 Spike tetramers and gated on CD3 -, CD14 -, CD16 -live CD19+ cells