key: cord-0744243-ovvhrr4v
authors: Feldman, Jared; Bals, Julia; Denis, Kerri St.; Lam, Evan C.; Hauser, Blake M.; Ronsard, Larance; Sangesland, Maya; Moreno, Thalia Bracamonte; Okonkwo, Vintus; Hartojo, Nathania; Balazs, Alejandro B.; Lingwood, Daniel; Schmidt, Aaron G.
title: Naive human B cells can neutralize SARS-CoV-2 through recognition of its receptor binding domain
date: 2021-02-10
journal: bioRxiv
DOI: 10.1101/2021.02.02.429458
sha: 0368302e46e96c1e26afad27453cdddae979eeff
doc_id: 744243
cord_uid: ovvhrr4v

Exposure to a pathogen elicits an adaptive immune response aimed to control and eradicate. This initial exposure is imprinted on the immune system, so that a subsequent encounter to the same pathogen or a variant will result in a memory recall response that is often protective. Interrogating the naive B cell repertoire in terms of both abundance and specificity to said pathogen may contribute to an understanding of how to potentially elicit protective responses. Here, we isolated naive B cells across 8 human donors, targeting the SARS-CoV-2 receptor-binding domain (RBD). Single B cell sorting, and subsequent sequence analysis, showed diverse gene usage and pairing with no apparent restriction on complementarity determining region length in either the heavy or light chains. We show that recombinantly expressed IgGs and Fabs of these germline precursors bind SARS-CoV-2 RBD. Importantly, a subset of these naive antibodies also bind SARS-CoV, an emergent variant (501Y.V2) and a potential pandemic (WIV-1) coronavirus. Furthermore, naive antibodies can also neutralize SARS-CoV-2 pseudoviruses in the absence of any somatic hypermutation, suggesting that protective immunity to coronaviruses, more broadly, may be genetically encoded. Future studies aimed at understanding the naive repertoire to other coronaviruses may ultimately reveal shared specificities that could be leveraged to develop pan-coronavirus vaccines aimed at priming encoded germline responses.

block ACE2 cell-surface binding but are on opposite sides of the RBM ( fig. S1E, F) . 11 IgGs had 157 no detectable ∆RBM binding (e.g., ab079, ab119), while 21 IgGs had reduced ELISA binding 158 relative to wild-type RBD ( Fig. 2A, B) , reflected in the reduced ∆RBM median EC50 values (Fig. 159 2C). We also identified examples of antibodies sensitive to only ∆475 (e.g., ab185) and only ∆501 160 (e.g., ab007) ( Fig. 2B and fig. S7 ). Of the binding population, there is no apparent predisposition 161 for HCDR3 length or light chain pairing (Fig. 2 D, E) . 162 163 We next asked whether these antibodies could recognize the SARS CoV-2 variant 501Y.V2 164 currently circulating in South Africa (48). This variant includes a combination of the three RBD 165 mutations: K417N, E484K, and N501Y (48) . We find that 22 antibodies (50%) had detectable 166 binding for this variant (Fig. 2B) . Interestingly one antibody, ab033, seemed to be specific to this 167 variant and did not bind SARS-CoV-2. We additionally tested the cross-reactivity of these 168 antibodies to both SARS-CoV and WIV-1, related coronaviruses previously circulating and of 169 pandemic potential (49), respectively. While the majority of antibodies (~86%) were specific to 170 the SARS-CoV-2 RBD, 5 antibodies were also cross-reactive to both SARS-CoV and WIV-1. Of 171 these cross-reactive antibodies, ab017 and ab114, derives from the same IGHV3-33 and IGVL2-172 14 paring but were isolated from different donors, suggesting a shared or public clonotype. 173

Additionally, we found that ab189 was uniquely sensitive to the glycan at residue 475; this was 174 surprising given the lack of conservation between of SARS-CoV, CoV-2 and WIV-1 RBMs, 175 despite using the same ACE2 receptor (50). Collectively, these data demonstrate that coronavirus 176 -specific naive antibodies are present in distinct donors, are of unrestricted gene usage and when 177 as expressed as IgGs, have detectable affinity to RBDs from the currently administered vaccines, 178 a circulating variant of concern, and from at least two related viruses. 179 180 To obtain binding kinetics independent of avidity effects from bivalent IgGs, 11 antibodies were 181 selected for expression as Fabs to determine monovalent binding affinity (KDs) by biolayer 182 interferometry (BLI). Using monomeric RBD as the analyte, 9 of the 11 Fabs had detectable 183 binding with KDs ranging from ~6.5 to ~75 µM; the other two remaining Fabs (ab177, ab185), 184 gave unreliable affinity measurements (i.e., >100 µM) ( fig. S8) . Notably, all Fabs had 185 characteristically fast off rates (koff). This observation is consistent for germline B cells where fast 186 off-rates are compensated by avidity due to overall BCR surface density (51); subsequent affinity 187 . CC-BY-NC-ND 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint this version posted February 10, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 gains via somatic hypermutation often result in slowing of the off-rate and is a canonical 188 mechanism of improved antigen binding (52-54). While the biologically relevant antigen affinities 189 of naive B cells are uncertain, recent studies have suggested that B cell activation and affinity 190 maturation is not restricted by immeasurably low affinity BCR interactions (55-57), thus the 191 isolated naive antibodies described here, especially those with demonstrable affinity may very well 192 be engaged and elicited upon antigen exposure. 193

We next used a SARS-CoV-2 pseudovirus assay (9) to ask whether any of the isolated antibodies 195

were capable of blocking transduction of target cells. We found that of the 36 RBD-binding 196 antibodies tested in this assay, 5 had detectable levels of neutralization (~14%) (Fig. 3A and fig.  197 S9). These neutralizing antibodies, obtained from multiple donors, have few commonalities with 198 respect to their gene usages and HCDR3 lengths (Fig. 3B ). While these naive antibodies were not 199 as potent compared to the control antibody B38, isolated from a memory B cell (29), the 200 observation, nevertheless, that the naive repertoire contains antibodies that have, at germline, the 201 potential to neutralize is noteworthy. 202

Here, we showed that isolated naive antibodies from 5 different donors can bind SARS-CoV-2 204 RBD and its variant 501Y.V2, a proportion of which are cross-reactive to SARS CoV and WIV-205 1. These data suggest that RBM-specific precursors may be represented across a substantial 206 fraction of individual human naive repertories. This observation is consistent with longitudinal 207 studies of SARS-CoV-2 infected individuals in which the majority of convalescent individuals 208 seroconverted with detectable RBD serum antibodies and neutralization titers (14, 58) . A subset 209

of these germline antibodies already shows neutralization activity, albeit weakly, against SARS-210

CoV-2 pseudoviruses. They therefore have the potential to affinity mature into more potent 211 neutralizing antibodies after vaccination or natural infection. Indeed, recent studies have 212 characterized potent RBD-directed neutralizing antibodies with a limited level of somatic 213 hypermutation (19, 21, 24, 31, 59) . The naive antibodies characterized here engage epitopes across 214 the RBM with a range of angles of approach as defined by our glycan variant probes; this is 215 consistent with other RBD-targeting antibodies characterized by epitope-mapping, deep 216 mutational scanning and structures (25, 27, 60) (Fig. 4) . Having BCRs within the naive repertoire 217 that recognize distinct or partially overlapping epitopes across the RBM may be advantageous in 218 . CC-BY-NC-ND 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint this version posted February 10, 2021. ; https://doi.org/10.1101/2021.02.02.429458 doi: bioRxiv preprint eliciting protective responses to variants of SARS-CoV-2 (61-64). Collectively, our observations 219 described here for SARS-CoV-2 are notably in contrast to other viruses such as HIV, whose 220 germline precursors targeting its receptor binding site lack demonstrable affinity, are highly 221 restricted in their gene usages and cannot at germline neutralize the virus (65-67). and were sequence confirmed (Genwiz). SARS-CoV-2 spike contained a C-terminal foldon 247 trimerization domain and HRV 3C-cleavable 6xHis and 2xStrep II tags (68). All proteins were 248 . CC-BY-NC-ND 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint this version posted February 10, 2021. defined as CD19 + /CD3 -/IgG -/IgD + /spike PE + /spike APC + /RBD + /∆RBM --, were single-cell sorted 299 using BD FACS Aria II (BD Biosciences) into 96-well plates containing lysis buffer supplemented 300 with 1% BME. Within the CD19 + /IgG -/IgD + gated cells, we also confirmed that 97% of the events 301 were CD27 negative. Plates were stored at -80 ºC for subsequent analysis. Flow cytometry data 302 was analyzed using FlowJo software version 10.7.1. 303 304 BCR Sequencing 305 . CC-BY-NC-ND 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint this version posted February 10, 2021. ; https://doi.org/10.1101/2021.02.02.429458 doi: bioRxiv preprint BCR Sequencing was carried out as described previously (32). Briefly, whole transcriptome 306 amplification (WTA) was performed on the sorted cell-lysates according to the Smart-Seq2 307 protocol (69). We then amplified heavy and light chain sequences from the WTA products utilizing 308 pools of partially degenerate pools of V region specific primers (Qiagen HotStar Taq Plus). Heavy 309 and light chain amplifications were carried out separately, with each pool containing pooled 310 primers against human IGHV and heavy chain constant region genes, or human IGLV, IGKV, and 311 light chain constant region genes. Cellular barcodes and index adapters (based on Nextera XT 312 Index Adapters, Illumina Inc.) were added using a step-out PCR method. Amplicons were then 313 pooled and sequenced using a 250x250 paired end 8x8 index reads on an Illumina Miseq System. 314

The data were then demultiplexed, heavy and light chain reads were paired, and overlapping 315 sequence reads were obtained (Panda-Seq) (70) 10,000 293T-ACE2 cells (72) in 20 µL cell media containing 15 µg/ml polybrene were added to 333 each well and incubated at 37 °C for 60-72 hrs. Following transduction, cells were lysed using a 334 previously described assay buffer (73) and shaken for 5 min prior to quantitation of luciferase 335 expression using a Spectramax L luminometer (Molecular Devices). Percent neutralization was 336 . CC-BY-NC-ND 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made 

. CC-BY-NC-ND 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint this version posted February 10, 2021. 

. CC-BY-NC-ND 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made . CC-BY-NC-ND 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint this version posted February 10, 2021. (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made . CC-BY-NC-ND 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint this version posted February 10, 2021. binding to ∆475, region 2 refers to IgGs with reduced binding to both glycan variants, and region 667 3 refers to IgGs with reduced binding to ∆501. 668 669 . CC-BY-NC-ND 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint this version posted February 10, 2021. for binding to RBD calculated from a 1:1 binding model using a global fit to all curves for each 675 Fab using vendor supplied software. B38 Fab is used as a positive control. 676 677 . CC-BY-NC-ND 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint this version posted February 10, 2021. CC-BY-NC-ND 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

The copyright holder for this preprint this version posted February 10, 2021. ; https://doi.org/10.1101/2021.02.02.429458 doi: bioRxiv preprint

Early events in B cell activation. Annual review of immunology 28

Annual review of immunology

When designing vaccines, consider 420 the starting material: the human B cell repertoire

From vaccines to memory and back

Primary exposure to SARS-CoV-2 protects against reinfection in rhesus macaques

SARS-CoV-2 infection protects against rechallenge in rhesus macaques

Single-shot Ad26 vaccine protects against SARS-CoV-2 in rhesus macaques. 429

Development of an inactivated vaccine candidate for SARS-CoV-2

COVID-19 neutralizing antibodies predict disease severity and 433 survival

Neutralizing Antibodies Correlate with Protection from SARS-CoV-2 in Humans 435 during a Fishery Vessel Outbreak with a High Attack Rate

Antibodies to SARS-CoV-2 are associated with protection against reinfection. 438 medRxiv

Maturation and persistence of the anti-SARS-CoV-2 memory B cell response. 440 bioRxiv

Evolution of antibody immunity to SARS-CoV-2

Immunological memory to SARS-CoV-2 assessed for up to 8 months after 443 infection

Functional SARS-CoV-2-specific immune memory persists after mild COVID-445 19

Safety and immunogenicity of SARS-CoV-2 mRNA-1273 vaccine in older 447 adults

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

mRNA vaccine-elicited antibodies to SARS-CoV-2 and circulating variants. 451 bioRxiv

Potent neutralizing antibodies from COVID-19 patients define multiple 453 targets of vulnerability

Human neutralizing antibodies elicited by SARS-CoV-2 infection

Longitudinal isolation of potent near-germline SARS-CoV-2-neutralizing 457 antibodies from COVID-19 patients

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

SARS-CoV-2 neutralizing antibody structures inform therapeutic strategies

Structures of Human Antibodies Bound to SARS-CoV-2 Spike Reveal 467

Mapping Neutralizing and Immunodominant Sites on the SARS-CoV-2 Spike 469

A human neutralizing antibody targets the receptor binding site of SARS-CoV-2

A noncompeting pair of human neutralizing antibodies block COVID-19 virus binding 474 to its receptor ACE2

Structural basis of a shared antibody response to SARS-CoV-2

Human B Cell Clonal Expansion and Convergent Antibody Responses to 478 SARS-CoV-2

Germline-encoded affinity for cognate antigen enables vaccine amplification 480 of a human broadly neutralizing response against influenza virus

Anti-HIV-1 B cell responses are dependent on B cell precursor frequency and 483 antigen-binding affinity

Precursor frequency and affinity determine B cell competitive fitness in 485 germinal centers, tested with germline-targeting HIV vaccine immunogens

Defining and Manipulating B cell Immunodominance Hierarchies to Elicit Broadly 488

Neutralizing Antibody Responses Against Influenza Virus

The human naive B cell repertoire contains distinct subclasses for a 490 germline-targeting HIV-1 vaccine immunogen

HIV-1 broadly neutralizing antibody precursor B cells revealed by germline-493 targeting immunogen

A generalized HIV vaccine design strategy for priming of broadly neutralizing 495 antibody responses

Engineered receptor binding domain immunogens elicit pan-coronavirus 497 neutralizing antibodies. bioRxiv

Commonality despite exceptional diversity in 499 the baseline human antibody repertoire

Human neutralizing antibodies elicited by SARS-CoV-2 infection

Extrafollicular antibody responses

Potent Neutralizing Antibodies against SARS-CoV-2 Identified by High-Throughput 505

Single-Cell Sequencing of Convalescent Patients' B Cells

High frequency of shared clonotypes in human B cell receptor repertoires

Viral receptor-binding site antibodies with diverse germline origins

Visualizing antibody affinity maturation in germinal centers

Antibody Affinity Shapes the Choice between Memory and Germinal Center B Cell 532 Fates

Salmonella infection drives promiscuous B cell activation followed by 534 extrafollicular affinity maturation

Complex antigens drive permissive clonal selection in germinal centers

Repeated cross-sectional sero-monitoring of SARS-CoV-2 in

Stereotypic neutralizing VH antibodies against SARS-CoV-2 spike protein receptor 540 binding domain in COVID-19 patients and healthy individuals

Complete Mapping of Mutations to the SARS-CoV-2 Spike Receptor-Binding 543

Domain that Escape Antibody Recognition

Comprehensive mapping of mutations to the SARS-CoV-2 receptor-binding 545 domain that affect recognition by polyclonal human serum antibodies. bioRxiv

Genetic Variants of SARS-CoV-2-What Do They Mean?

Estimated transmissibility and severity of novel SARS-CoV-2 Variant of 550 Concern 202012/01 in England. medRxiv

Emergence and rapid spread of a new severe acute respiratory syndrome-related 552 coronavirus 2 (SARS-CoV-2) lineage with multiple spike mutations in South Africa

Recombinant HIV envelope proteins fail to engage germline versions of anti-CD4bs 555 bNAbs

Sequence and structural convergence of broad and potent HIV antibodies that 557 mimic CD4 binding

Structural basis for broad and potent neutralization of HIV-1 by antibody VRC01

Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation

Preparation of single-cell RNA-seq libraries for next generation sequencing

Structural model of 606 the ∆RBM probe designed to abrogate binding to ACE2

glycosylation sites engineered onto the RBM are shown in red spheres at amino acid positions 501 608 and 475. (B) SDS-PAGE gel under reducing (R) and non-reducing (NR) conditions for monomeric 609

C) Wildtype RBD, ∆RBM and single glycan variant binding to

Wild-type RBD binding shown in blue, glycan 611 variant binding shown in red. Streptavidin-PE was used to detect the relative intensity of antigen 612 binding to cell-surface ACE2. A PBS control (gray) indicates secondary-only staining

RBM-specific antibody, B38 (left)

RBM-specific control antibody, CR3022 (right). (E) ∆RBM and ∆501 and ∆475 variants analyzed 615 by SDS-PAGE gel under reducing conditions