key: cord-0306212-pbnces6i
authors: Reincke, S Momsen; Yuan, Meng; Kornau, Hans-Christian; Corman, Victor M; van Hoof, Scott; Sánchez-Sendin, Elisa; Ramberger, Melanie; Yu, Wenli; Hua, Yuanzi; Tien, Henry; Schmidt, Marie Luisa; Schwarz, Tatjana; Jeworowski, Lara Maria; Brandl, Sarah E; Rasmussen, Helle Foverskov; Homeyer, Marie A; Stöffler, Laura; Barner, Martin; Kunkel, Désirée; Huo, Shufan; Horler, Johannes; von Wardenburg, Niels; Kroidl, Inge; Eser, Tabea M; Wieser, Andreas; Geldmacher, Christof; Hoelscher, Michael; Gänzer, Hannes; Weiss, Günter; Schmitz, Dietmar; Drosten, Christian; Prüss, Harald; Wilson, Ian A.; Kreye, Jakob
title: SARS-CoV-2 Beta variant infection elicits potent lineage-specific and cross-reactive antibodies
date: 2021-09-30
journal: bioRxiv
DOI: 10.1101/2021.09.30.462420
sha: 220af73258359647507b9d3e11aa4c9861e16cac
doc_id: 306212
cord_uid: pbnces6i

SARS-CoV-2 Beta variant of concern (VOC) resists neutralization by major classes of antibodies from non-VOC COVID-19 patients and vaccinated individuals. Here, serum of Beta variant infected patients revealed reduced cross-neutralization of non-VOC virus. From these patients, we isolated Beta-specific and cross-reactive receptor-binding domain (RBD) antibodies. The Beta-specificity results from recruitment of novel VOC-specific clonotypes and accommodation of VOC-defining amino acids into a major non-VOC antibody class that is normally sensitive to these mutations. The Beta-elicited cross-reactive antibodies share genetic and structural features with non-VOC-elicited antibodies, including a public VH1-58 clonotype targeting the RBD ridge independent of VOC mutations. These findings advance our understanding of the antibody response to SARS-CoV-2 shaped by antigenic drift with implications for design of next-generation vaccines and therapeutics. One sentence summary SARS-CoV-2 Beta variant elicits lineage-specific antibodies and antibodies with neutralizing breadth against wild-type virus and VOCs.

In the course of the COVID-19 pandemic, multiple SARS-CoV-2 lineages have emerged including lineages defined as variants of concern (VOC), such as Alpha (also known as lineage B.1.1.7), Beta (B.1.351), Gamma (P.1) and Delta (B.1.617.2). VOCs are associated with increased transmissibility, virulence and/or resistance to neutralization by sera from vaccinated individuals and convalescent COVID-19 patients who were infected with the original, non-VOC strain (1-7). These distinct lineages carry a variety of mutations in the spike protein, several of which are within the receptor binding domain (RBD), especially at residues K417, L452, T478, E484, and N501. Some mutations like N501Y are associated with enhanced binding to angiotensin-converting enzyme 2 (ACE2), largely driving the global spread of VOCs incorpororating these mutations (2) . However, with increasing immunity either through natural infection or vaccination, antibody escape might become more relevant in emerging VOCs. Many studies have investigated RBD antibodies in COVID-19 patients prior to identification of SARS-CoV-2 variants, and we refer to these as non-VOC antibodies. Non-VOC RBD antibodies revealed a preferential response towards distinct epitopes with enriched recruitment of particular antibody germline genes, where the most prominent were VH3-53 and closely related VH3-66, as well as VH1-2 (8, 9) . Structural and functional classification of non-VOC RBD mAbs has demonstrated that mAbs from these three enriched germline genes form two major classes of receptor-binding site (RBS) mAbs whose binding and neutralizing activity depends on either K417 and/or E484 (9, 10) . Mutations at these key residues (K417N and E484K) and at N501Y are hallmarks of the Beta variant (2) , and largely account for the reduced neutralizing activity of sera from vaccinated individuals and convalescent COVID-19 patients against this VOC (1- 6, 11, 12) . These key mutations also occur in further sublineages including an Alpha variant strain carrying E484K and a Delta variant strain carrying K417N (Delta Plus). Of all current VOCs, the Beta variant appears to be most resistant to neutralization from non-VOC sera, suggesting conspicuous differences in its antigenicity (13) . However, little is known about the antibody response elicited by Beta variant infection. For example, it is unknown if antibodies targeting the RBD of the Beta variant (RBD Beta) share the preferential recruitment of particular germline genes with non-VOC antibodies, and whether VOC-defining mutations K417N and E484K could be accommodated in the canonical binding modes of public antibody classes like VH3-53/VH3-66 antibodies. Investigating Beta variant induced immunity can therefore bolster efforts to monitor and prevent the spread of SARS-CoV-2 by informing vaccine design in the context of the ongoing antigenic drift. Thus, we set out to explore genetic, functional and structural features of the antibody response against RBD Beta compared to non-VOC RBD.

We identified 40 individuals infected with the SARS-CoV-2 Beta variant from three metropolitan areas in Germany and Austria (table S1). Serum from these patients was collected 38.6 ± 19.2 days after their first positive SARS-CoV-2 RT-PCR test. The patients' IgG bound to non-VOC nucleocapsid protein and/or non-VOC spike protein in 37 of 40 patients with stronger reactivity to RBD Beta than to non-VOC RBD (fig. S1A). The VOC patients' sera also inhibited ACE2 binding to RBD Beta to a greater extent than to non-VOC RBD (fig. S1B, table S1), indicating the presence of highly effective RBD antibodies after Beta variant infection. Reactivity to non-VOC spike S1 antigen was confirmed in an additional commercially available ELISA; however, only 23 Collectively, these data show that sera from Beta-infected patients exhibit reduced cross-reactivity and cross-neutralization to non-VOC SARS-CoV-2, impacting diagnostic antibody testing and assessment of antibody levels when using non-VOC antigens.

To investigate this difference in reactivity between RBD Beta and non-VOC RBD on the level of mAbs elicited by SARS-CoV-2 Beta variant infection, we isolated CD19 + CD27 + memory B cells from the peripheral blood of 12 donors in our cohort via fluorescence activated cell sorting using a recombinant RBD Beta (K417N/E484K/N501Y) probe ( fig. S2A ). Frequencies of RBD-double-positive memory B cells ranged from 0.007% to 0.1% ( fig. S2B ). Using single-cell Ig gene sequencing (15, 16) , we derived 289 pairs of functional heavy (IGH) and light (IGL) chain sequences from IgG mAbs (table S2) .

Sequence analysis showed enrichment of certain VH genes compared to mAbs derived from healthy, non-infected individuals, including VH1-58, VH3-30, VH4-39 and VH3-53, illustrating a preferential recruitment of certain VH genes ( Fig. 2A) and VH-JH gene combinations ( fig. S3A ). For some genes like VH1-58 and VH3-53, enrichment has previously been identified in CoV-AbDab, a database of published non-VOC SARS-CoV-2 mAbs (9, 17) . We here confirmed this finding for all human RBD mAbs in this database ( Fig. 2A) . Consistent with previous reports from non-VOC SARS-CoV-2

infections (18) (19) (20) , the number of somatic hypermutations (SHM) in the IGH and IGL chains was generally low in mAbs derived from our cohort ( fig. S3B ). Together, these findings argue for conservation of certain antibody sequence features between antibody responses in different donors and between antibody responses elicited against SARS-CoV-2 Beta variant and non-VOC viruses. Hence, we compared antibody sequences after Beta infection to all previously published non-VOC RBD antibodies and identified several clonotypes shared between both datasets (Fig. 2B) , some of which were present in multiple patients of our study (Fig. 2C ). Taken together, these results demonstrate that a subset of the antibodies to RBD Beta and non-VOC RBD converge on recruitment of specific germline genes.

However, other gene enrichments found in our study like VH4-39 have not been identified within the CoV-AbDab mAbs (9) (Fig. 2A) , exemplifying concurrent divergence in the antibody response to the different RBDs. Strikingly, VH1-2, one of the most common genes contributing to the RBD antibody response to non-VOC SARS-CoV-2, was virtually absent in our dataset of Beta variant elicited mAbs ( Fig. 2A and table S2 ), in line with our previous predictions about the effect of Beta variant mutations on VH1-2 binding and neutralization (9) . VH3-53/VH3-66 antibodies have previously been shown to bind to non-VOC RBD in two canonical binding modes, which involve residues K417 and E484, respectively; binding and neutralization of these antibodies are strongly affected by the K417N and E484K mutations in RBD Beta (9, 21) . We therefore hypothesized a similarly reduced recruitment of VH3-53/VH3-66 mAbs after Beta variant infection. Surprisingly, we identified 15 VH3-53/VH3-66 mAbs, albeit at a reduced frequency compared to the CoV-AbDab dataset (4.7% vs. 19.4%), but still at an increased frequency compared to healthy donors ( Fig. 2A) , thus indicating either a non-canonical binding mode or accommodation of these mutations into the known binding modes.

To determine the binding properties of antibodies elicited by SARS-CoV-2 Beta , we selected mAbs for expression and further characterization based on the following criteria: (i) mAbs that are clonally expanded within one patient, (ii) mAbs of clonotypes present in several patients in our dataset to decipher the shared antibody response to RBD Beta, (iii) mAbs of clonotypes found both in our and CoV-AbDab datasets to potentially identify cross-reactive mAbs, (iv) VH3-53/VH3-66 mAbs to elucidate the unexpected recurrent shared antibody response against RBD Beta, (v) mAbs of VH genes with strongest enrichment in our dataset, including VH4-39 and VH1-58. We identified 81 mAbs with strong binding to RBD Beta (table S3) , as defined by detectable binding at 10 ng/ml. Of those, a majority (44 in total) revealed comparable binding to non-VOC RBD and were considered cross-reactive mAbs, whereas 37 mAbs did not bind non-VOC RBD at 10 ng/ml and were considered RBD Beta-specific.

We aimed to determine the residues that define the mAb binding selectivity for the 37 RBD Beta-specific mAbs, and performed ELISAs with single mutant constructs of RBD Beta and non-VOC RBD. For all three Beta-defining RBD mutations (K417N, E484K and N501Y), we identified mAbs with RBD binding that were dependent on a single residue, with a larger fraction of binders that were dependent on K484 (12) and Y501 (11) than N417 (3). The RBD Beta specificity of other mAbs was dependent on multiple residues (Fig. 3A) . 26 of the RBD Beta-specific mAbs (70.2%) neutralized the authentic SARS-CoV-2 Beta isolate, with representation in all of the above-mentioned specificity categories (Fig. 3A) .

RBD Beta-specific mAbs were encoded by a broad variety of VH genes (Fig. 3A and table S2) .

Interestingly, all nine VH4-39 mAbs with RBD Beta specificity from three different patients were Y501dependent, comprising 81.8% of the mAbs in this category. This finding suggests a common binding mode of these mAbs that depends on Y501, a mutation that is present in RBD Beta, Alpha and Gamma, but not Delta, and may explain the frequent use of VH4-39 in mAbs to RBD Beta ( Fig. 2A) . VH4-39 Y501-dependent mAbs revealed few SHM in VH genes but no uniform pattern in other sequence features ( fig. S4A ). Although all VH4-39 RBD Beta-specific mAbs bind to a Y501-dependent epitope, their neutralization activity against authentic Beta virus showed noticeable differences (IC 50 ranging from 5.2 to 947 ng/ml, fig. S4B ). Surface plasmon resonance measurements of these mAbs to RBD Beta Furthermore, we identified three VH3-53/VH3-66 mAbs with RBD Beta specificity that all showed neutralizing activity. To determine whether this RBD Beta specificity results from a non-canonical binding mode or accommodation of the Beta variant-defining mutations in one of the two main VH3-53/VH3-66 mAb binding modes, we determined a crystal structure of VH3-53 antibody CS23 in complex with RBD Beta. Previously, we and others found that VH3-53/VH3-66 mAbs with short CDRs H3 (<15 amino acids) target the RBS of non-VOC RBD via a canonical mode (10, (22) (23) (24) (25) that is highly sensitive to the K417N mutation (9) . CS23 contains a short CDR H3 with only ten amino acids and is specific to N417 RBDs including RBD Beta (Fig. 3A) . Perhaps unexpectedly, CS23 binds to RBD Beta in the canonical mode, with a nearly identical approach angle compared to non-VOC VH3-53 antibody CC12.3 (24) (Fig. 3B) . We also previously determined that the CDR H1 33 NY 34 and H2 53 SGGS 56 motifs of VH3-53/VH3-66 mAbs are critical for RBD recognition (24) . Here we find that CS23 retains these motifs and they interact with the RBD in the same way (Fig. 3 , C and D). Residues in CDR H3 usually interact with K417 and thus confer specificity to the non-VOC RBD (9) . For example, V H D97 of CC12.1 forms a salt bridge with the outward facing RBD-K417, whereas V H F99 and V H G97 of CC12.3 interact with K417 through cation-π and hydrogen bonds (H-bonds), respectively (Fig. 3 , E and F).

Instead, in RBD Beta, the shorter N417 flips inward and H-bonds with RBD-E406 and Q409 (Fig. 4G (table S2) . To elucidate the structural basis of this public pan-VOC clonotype, we determined crystal structures of CS44 and CV07-287, a mAb of the same clonotype that was isolated from a non-VOC infected individual (19) , in complex with RBD Beta and non-VOC RBD respectively (Fig. 4C) . We compared the structures of CS44 and CV07-287 with other published VH1-58 antibodies, including COVOX-253 (27), S2E12 (28), A23-58.1, and B1-182.1 (26) . These antibodies all target the RBD in the same binding mode (Fig. 4C) , which suggests that this public clonotype is structurally conserved. The dominant interaction of VH1-58 antibodies is with the RBD ridge region infection. This issue could be addressed by the use of VOC-based antigens, but these data challenge the concept of defining a universal threshold for protective antibody titers or threshold-based reasoning for booster vaccination. Furthermore, on the monoclonal level, we show that RBD mAbs with Beta-specificity are frequent after Beta infection, similar to the frequency of RBD mAbs that do not react with RBD Beta after non-VOC infection or vaccination (2) . Antibodies with Beta-specificity include novel VOC-specific public clonotypes such as Y501-dependent VH4-39 mAbs, and others in the major non-VOC antibody class encoded by VH3-53/VH3-66 genes. Surprisingly, VOC-defining mutations in the Beta variant can be accommodated by local conformational changes and mutations in these VH3-53/VH3-66 mAbs that enable canonical mode binding. Moreover, a subset of RBD Beta elicited mAbs was cross-reactive to non-VOC RBD and also against the other VOCs investigated including the Delta variant, with multiple of these mAbs revealing potent cross-neutralization. VH1-58 antibodies form a clonotype with both ultra-high potency and high resistance to currently circulating VOCs, including variants of concern Alpha, Beta, Gamma and Delta (26) . Here, we show that pan-VOC VH1-58 antibodies are also frequently elicited by the Beta variant, and target the S protein in a nearly identical binding mode Bank under accession codes 7S5P, 7S5Q and 7S5R. The amino acid sequences of the antibodies described in this study can be found in table S3. All requests for materials including antibodies, viruses, plasmids and proteins generated in this study should be directed to the corresponding authors. Materials will be made available for non-commercial usage.

Figs. S1 to S6

Tables S1 to S5

References 32-44 His-tagged recombinant RBD Beta protein was produced in HEK cells (ACROBiosystems, SPD-C52Hp) and covalently labeled using CruzFluor488 (Santa Cruz Biotechnology, sc-362617)

according to the manufacturer's instructions. Separately, the same protein was labeled using its His-tag by incubating the antigen with an Alexa Fluor 647-conjugated anti-His-antibody (R&D Systems, IC0501R) for 30 minutes at room temperature at a 2:1 ratio (RBD molecules:IgG molecules). Ovalbumin (Sigma, A5503) was covalently labeled with PE-Cy7 (abcam, ab102903)

according to the manufacturer's instructions.

Using 

The mAbs were generated following our established protocols (19) Second, the presence of SARS-CoV-2 S1-specific antibodies was analyzed using a commercially available anti-SARS-CoV-2-S1 IgG ELISA (EUROIMMUN Medizinische Labordiagnostika AG, Lübeck, Germany) according to the manufacturer's instructions. Serum samples were diluted 1:101. The optical density (OD) at 450nm was measured and OD ratios were calculated by dividing this value by the OD of the kit-included calibrator.

Additionally, we applied a modified solid phase immunoassay (Seramun Diagnostica GmbH, Heidesee, Germany) as described above, which additionally contained SARS-CoV-2 RBD-VOCs (InVivo BioTech Services GmbH, Hennigsdorf, Germany).

Neutralizing capacity of patients' sera was assessed by a surrogate virus neutralization test 

Binding of mAbs to SARS-CoV-2 spike RBD or variants thereof was detected by ELISA as previously described (19) . 

To assess the neutralizing activity of serum samples and SARS-CoV-2 mAbs, we performed plaque reduction neutralization tests (PRNT) as described (14) 

Competition assays were performed by biolayer interferometry (BLI) using an Octet Red instrument (FortéBio). IgGs were diluted with kinetic buffer (1x PBS, pH 7.4, 0.01% BSA and 0.002% Tween 20). After His-tagged RBD Beta was immobilized on anti-Penta His BLI sensors, sensors were first dipped into CS82 IgG (50 μg/ml), and then dipped into indicated IgG antibodies (12.5 μg/ml). Three replicates were performed for each BLI experiment.

Area under the curve (AUC) calculations in Fig. 1 and all statistical analyses were performed using GraphPad Prism (9.2.0). 

Antibody evasion by the P.1 strain of SARS-CoV-2

Antibody resistance of SARS-CoV-2 variants B.1.351 and B.1.1.7

SARS-CoV-2 variants B.1.351 and P.1 escape from neutralizing antibodies

SARS-CoV-2 variant B.1.617 is resistant to bamlanivimab and evades antibodies induced by infection and vaccination

Sensitivity of infectious SARS-CoV-2 B.1.1.7 and B.1.351 variants to neutralizing antibodies

Evidence of escape of SARS-CoV-2 variant B.1.351 from natural and vaccineinduced sera

Limited neutralization of authentic SARS-CoV-2 variants carrying E484K in vitro

Modular basis for potent SARS-CoV-2 neutralization by a prevalent VH1-2-derived antibody class

Structural and functional ramifications of antigenic drift in recent SARS-CoV-2 variants

SARS-CoV-2 neutralizing antibody structures inform therapeutic strategies

Impact of circulating SARS-CoV-2 variants on mRNA vaccine-induced immunity in uninfected and previously infected individuals. medRxiv

SARS-CoV-2 variants of concern partially escape humoral but not T-cell responses in COVID-19 convalescent donors and vaccinees

COVID-19 vaccines: Keeping pace with SARS-CoV-2 variants

Virological assessment of hospitalized patients with COVID-2019

Human cerebrospinal fluid monoclonal N-methyl-D-aspartate receptor autoantibodies are sufficient for encephalitis pathogenesis

Efficient generation of monoclonal antibodies from single human B cells by single cell RT-PCR and expression vector cloning

CoV-AbDab: the coronavirus antibody database

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

A therapeutic non-self-reactive SARS-CoV-2 antibody protects from lung pathology in a COVID-19 hamster model

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

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

Structures of human antibodies bound to SARS-CoV-2 spike reveal common epitopes and recurrent features of antibodies

An alternative binding mode of IGHV3-53 antibodies to the SARS-CoV-2 receptor binding domain

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

Recognition of the SARS-CoV-2 receptor binding domain by neutralizing antibodies

Ultrapotent antibodies against diverse and highly transmissible SARS-CoV-2 variants

We thank all study participants who devoted samples and time to our research, Dr. Kim Stahlberg for patient recruitment, Stefanie Bandura, Matthias Sillmann, Doreen Brandl, Patricia Tscheak, and Sabine Engl for excellent technical assistance, and Dr. Marcel A Müller and Dr. Daniela Niemeyer for support