key: cord-1048499-khzalpon
authors: Kreye, Jakob; Reincke, S. Momsen; Kornau, Hans-Christian; Sánchez-Sendin, Elisa; Corman, Victor Max; Liu, Hejun; Yuan, Meng; Wu, Nicholas C.; Zhu, Xueyong; Lee, Chang-Chun D.; Trimpert, Jakob; Höltje, Markus; Dietert, Kristina; Stöffler, Laura; von Wardenburg, Niels; van Hoof, Scott; Homeyer, Marie A.; Hoffmann, Julius; Abdelgawad, Azza; Gruber, Achim D.; Bertzbach, Luca D.; Vladimirova, Daria; Li, Lucie Y.; Barthel, Paula Charlotte; Skriner, Karl; Hocke, Andreas C.; Hippenstiel, Stefan; Witzenrath, Martin; Suttorp, Norbert; Kurth, Florian; Franke, Christiana; Endres, Matthias; Schmitz, Dietmar; Jeworowski, Lara Maria; Richter, Anja; Schmidt, Marie Luisa; Schwarz, Tatjana; Müller, Marcel Alexander; Drosten, Christian; Wendisch, Daniel; Sander, Leif E.; Osterrieder, Nikolaus; Wilson, Ian A.; Prüss, Harald
title: A therapeutic non-self-reactive SARS-CoV-2 antibody protects from lung pathology in a COVID-19 hamster model
date: 2020-09-23
journal: Cell
DOI: 10.1016/j.cell.2020.09.049
sha: bcbf2537f6b1ba293704f700af0393c14a39ae5c
doc_id: 1048499
cord_uid: khzalpon

The emergence of SARS-CoV-2 led to pandemic spread of coronavirus disease 2019 (COVID-19), manifesting with respiratory symptoms and multi-organ dysfunction. Detailed characterization of virus-neutralizing antibodies and target epitopes is needed to understand COVID-19 pathophysiology and guide immunization strategies. Among 598 human monoclonal antibodies (mAbs) from ten COVID-19 patients, we identified 40 strongly neutralizing mAbs. The most potent mAb CV07-209 neutralized authentic SARS-CoV-2 with IC50 of 3.1 ng/ml. Crystal structures of two mAbs in complex with the SARS-CoV-2 receptor-binding domain at 2.55 and 2.70 Å revealed a direct block of ACE2 attachment. Interestingly, some of the near-germline SARS-CoV-2 neutralizing mAbs reacted with mammalian self-antigens. Prophylactic and therapeutic application of CV07-209 protected hamsters from SARS-CoV-2 infection, weight loss and lung pathology. Our results show that non-self-reactive virus-neutralizing mAbs elicited during SARS-CoV-2 infection are a promising therapeutic strategy.

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) started emerging in humans in late 2019, and rapidly spread to a pandemic with millions of cases worldwide. SARS-CoV-2 infections cause coronavirus disease 2019 (COVID-19) with severe respiratory symptoms, but also pathological inflammation and multi-organ dysfunction, including acute respiratory distress syndrome, cardiovascular events, coagulopathies and neurological symptoms (Helms et al., 2020; Zhou et al., 2020; . Some aspects of the diverse clinical manifestations may result from a hyperinflammatory response, as suggested by reduced mortality in hospitalized COVID-19 patients under dexamethasone therapy (Horby et al., 2020) .

Understanding the immune response to SARS-CoV-2 therefore is of utmost importance. Multiple recombinant SARS-CoV-2 mAbs from convalescent patients have been reported (Brouwer et al., 2020; Cao et al., 2020; Ju et al., 2020; Kreer et al., 2020; Robbiani et al., 2020; Rogers et al., 2020; Wec et al., 2020) . mAbs targeting the receptor-binding domain (RBD) of the viral spike protein S1 can compete with its binding to human angiotensin converting enzyme 2 (ACE2) and prevent viral entry and subsequent replication (Cao et al., 2020; Ju et al., 2020; Walls et al., 2020) . Potent virus neutralizing mAbs that were isolated from diverse variable immunoglobulin (Ig) genes typically carry low levels of somatic hypermutations (SHM). Several of these neutralizing mAbs selected for in vitro efficacy showed prophylactic or therapeutic potential in animal models (Cao et al., 2020; Liu et al., 2020; Rogers et al., 2020; Zost et al., 2020) . The low number of SHM suggests limited affinity-maturation in germinal centers compatible with an acute infection. Near-germline mAbs usually constitute the first line of defense to pathogens, but carry the risk of self-reactivity to autoantigens (Lerner, 2016; Liao et al., 2011; Zhou et al., 2007) . Although critical for the therapeutic use in humans, potential tissue-reactivity of near-germline SARS-CoV-2 antibodies has not been examined so far.

Here, we systematically selected 18 strongly neutralizing mAbs out of 598 antibodies from 10 COVID-19 patients by characterization of their biophysical properties, authentic SARS-CoV-2 neutralization, and exclusion of off-target binding to murine tissue. Additionally, we solved two crystal structures of J o u r n a l P r e -p r o o f

We first characterized the B cell response in COVID-19 using single-cell Ig gene sequencing of human mAbs ( Figure 1A ). From ten COVID-19 patients with serum antibodies to the S1 subunit of the SARS-CoV-2 spike protein ( Figure S1A , Table S1), we isolated two populations of single cells from peripheral blood mononuclear cells with fluorescence-activated cell sorting (FACS): CD19 + CD27 + CD38 + antibody secreting cells (ASC) reflecting the overall humoral immune response and SARS-CoV-2-S1-labeled CD19 + CD27 + memory B cells (S1-MBC) for characterization of antigen-specific responses ( Figure S1B and S1C). We obtained 598 functional paired heavy and light chain Ig sequences (Table S2 ). Of 432 recombinantly expressed mAbs, 122 were reactive to SARS-CoV-2-S1 (S1+), with a frequency of 0.0-18.2% (median 7.1%) within ASC and 16.7-84.1% (median 67.1%) within S1-MBC ( Figure 1B and 1C ).

Binding to S1 did not depend on affinity maturation as measured by the number of SHM ( Figure 1D ).

Compared to mAbs not reactive to SARS-CoV-2-S1, S1+ mAbs had less SHM, but equal lengths for both their light and heavy chain complementarity-determining region 3 (CDR3) (Figure S1D, S1E and S1F).

Within the ASC and S1-MBC population, 45.0% and 90.2% of S1+ mAbs bound the RBD, respectively ( Figure S1G ). S1+ mAbs were enriched in certain Ig genes including VH1-2, VH3-53, VH3-66, VK1-33 and VL2-14 ( Figure S2 ). We identified both clonally related antibody clones within patients and public and shared S1+ clonotypes from multiple patients ( Figure S3A and S3B). Some public or shared clonotypes had been previously reported, such as IGHV3-53 and IGHV3-66 ( Figure S3D ) (Cao et al., 2020; Yuan et al., 2020a) , while others were newly identified, such as IGHV3-11 ( Figure S3C ).

We next determined the mAbs with the highest capacity to neutralize SARS-CoV-2 in plaque reduction neutralization tests (PRNT) using authentic virus (Munich isolate 984) (Wolfel et al., 2020) . Of 87 mAbs strongly binding to RBD, 40 showed virus neutralization with a half-maximal inhibitory concentration (IC 50 ) J o u r n a l P r e -p r o o f ≤250 ng/ml and were considered neutralizing antibodies ( Figure 1A , Table S2 ), from which 18 were selected for further characterization (Table S3 ). The antibodies bound to RBD with a half-maximal effective concentration (EC 50 ) of 3.8-14.2 ng/ml ( Figure 1E ) and an equilibrium dissociation constant (K D ) of 6.0 pM to 1.1 nM ( Figure S4 , Table S3 ), thereby neutralizing SARS-CoV-2 with an IC 50 of 3.1-172 ng/ml ( Figure 1F , Table S3 ). The antibody with the highest apparent affinity, CV07-209, was also the strongest neutralizer ( Figure 1G ). We hypothesized that the differences in neutralizing capacity relate to different interactions with the ACE2 binding site. Indeed, the strongest neutralizing mAbs CV07-209 and CV07-250 reduced ACE2 binding to RBD to 12.4% and 58.3%, respectively. Other Top-18 mAbs including CV07-270 interfered only weakly with ACE2 binding ( Figure S5A ).

The spike proteins of SARS-CoV-2 and SARS-CoV share more than 70% amino acid sequence identity, whereas sequence identity between SARS-CoV-2 and MERS-CoV and other endemic coronaviruses is significantly lower . To analyze potential cross-reactivity of mAbs to other coronaviruses, we tested for binding of the Top-18 mAbs to the RBD of SARS-CoV, MERS-CoV, and the human endemic coronaviruses 229-E, NL63, HKU1 and OC43. CV38-142 detected the RBD of both SARS-CoV-2 and SARS-CoV, whereas no other mAb was cross-reactive to additional coronaviruses ( Figure S5C and S5D). To further characterize the epitope of neutralizing mAbs, we performed ELISAbased epitope binning experiments using biotinylated antibodies. Co-applications of paired mAbs showed competition of most neutralizing antibodies for RBD binding ( Figure S5B ). As an exception, SARS-CoV cross-reactive CV38-142 bound RBD irrespective of the presence of other mAbs, suggesting an independent and conserved target epitope ( Figure S5B ).

Many SARS-CoV-2 neutralizing mAbs carry few SHM or are in germline configuration ( Figure 1D ) (Ju et al., 2020; Kreer et al., 2020) . Such antibodies close to germline might be reactive to more than one target (Zhou et al., 2007) . Prompted by the abundance of near-germline SARS-CoV-2 antibodies and to exclude J o u r n a l P r e -p r o o f potential side effects of mAb treatment, we next analyzed whether SARS-CoV-2 antibodies can bind to self-antigens. Therefore, we tested the binding of S1-mAbs to unfixed murine tissues. Surprisingly, four of the Top-18 potent SARS-CoV-2 neutralizing mAbs showed anatomically distinct tissue reactivities ( Figure 2 , Table   S3 ). CV07-200 intensively stained brain sections in the hippocampal formation, olfactory bulb, cerebral cortex and basal ganglia ( Figure 2A ). CV07-222 also bound to brain tissue, as well as to smooth muscle ( Figure 2B ). CV07-255 and CV07-270 were reactive to smooth muscle from sections of lung, heart, kidney and colon, but not liver ( Figure 2C and 2D, Table S3 ). None of the Top-18 mAbs bound to HEp-2 cells, cardiolipin or beta-2 microglobulin as established polyreactivity-related antigens (Jardine et al., 2016) ( Figure S5E ).

Diffraction-quality crystals were obtained for SARS-CoV-2 RBD complexed with two individual neutralizing mAbs, CV07-250 and CV07-270, which have notable differences in the number of SHM, extent of ACE2 competition and binding to murine tissue. CV07-250 (IC 50 =3.5 ng/ml) had 33 SHM (17/16 on heavy and light chain, respectively), strongly reduced ACE2 binding and showed no binding to murine tissue. In contrast, CV07-270 (IC 50 = 82.3 ng/ml) had only 2 SHM (2/0), did not reduce ACE2 binding in our assay, and showed binding to smooth muscle tissue. Using X-ray crystallography, we determined structures of CV07-250 and CV07-270 in complex with SARS-CoV-2 RBD to resolutions of 2.55 and 2.70 Å, respectively (Figure 3, Table S4 , Table S5 ).

The binding mode of CV07-250 to RBD is unusual in that it is dominated by the light chain ( Figure 3A and 3D), whereas in CV07-270, the heavy chain dominates as frequently found in other antibodies ( Figure 3B and 3E). Upon interaction with the RBD, CV07-250 has a buried surface area (BSA) of 399 Å 2 and 559 Å 2 on the heavy and light chains, respectively, compared to 714 Å 2 and 111 Å 2 in CV07-270. CV07-250 uses CDR H1, H3, L1, L3, and framework region 3 (LFR3) for RBD interaction ( Figure 3D and Figure The epitope of CV07-250 completely overlaps with the ACE2 binding site with a similar angle of approach as ACE2 ( Figure 3A , 3C, 4G and 4I). In contrast, the CV07-270 epitope only partially overlaps with the ACE2 binding site and the antibody approaches the RBD from a different angle compared to CV07-250 and ACE2 ( Figure 3B , 3C, 4H, 4I), explaining differences in ACE2 competition. Although CV07-250 and CV07-270 both contact 25 epitope residues, only seven residues are shared (G446/G447/E484/G485/Q493/S494/Q498). Furthermore, CV07-270 binds to a similar epitope as SARS-CoV-2 neutralizing antibody P2B-2F6 (Ju et al., 2020) with a similar angle of approach ( Figure S5F ). In fact, 18 out of 20 residues in the P2B-2F6 epitope overlap with the CV07-270 epitope, although CV07-270

and P2B-2F6 are encoded by different germline genes for both heavy and light chains. Thus, these two mAbs represent antibodies encoded by different germline genes that bind to the same epitope in the RBD with near-identical binding modes and approach angles. This structural convergence is also encouraging for targeting this highly immunogenic epitope for vaccine development.

Interestingly, CV07-250 was isolated 19 days after symptom onset, but already acquired 33 SHM, the highest number among all S1+ MBCs ( Figure S1D ). Some non-germline amino acids are not directly involved in RBD binding, including all five SHMs on CDR H2 ( Figure S6 ). This observation suggests that CV07-250 could have been initially affinity-matured against a different antigen.

Next, we selected mAb CV07-209 for evaluation of in vivo efficacy based on its high capacity to neutralize SARS-CoV-2 and the absence of reactivity to mammalian tissue. We used the hamster model of COVID-19, as it is characterized by rapid weight loss and severe lung pathology (Osterrieder et al., 2020) . In this experimental set-up, hamsters were intranasally infected with authentic SARS-CoV-2. Nine hamsters per group received either a prophylactic application of CV07-209 24 hours before viral challenge, or a J o u r n a l P r e -p r o o f therapeutic application of CV07-209 or control antibody mGO53 two hours after viral challenge ( Figure   5A ).

Hamsters under control mAb treatment lost 5.5±4.4% (mean±SD) of body weight, whereas those that received mAb CV07-209 as a therapeutic or prophylactic single dose gained 2.2±3.4% or 4.8±3.4% weight after 5 days post-infection (dpi), respectively. Mean body weights gradually converged in the animals followed up until 13 dpi, reflecting the recovery of control-treated hamsters from SARS-CoV-2 infection ( Figure 5B ).

To investigate the presence of SARS-CoV-2 in the lungs, we measured functional SARS-CoV-2 particles from lung tissue homogenates. Plaque forming units were below the detection threshold for all animals in the prophylactic and in 2 of 3 in the treatment group at 3 and 5 dpi ( Figure 5C and 5D). qPCR measurements of lung viral genomic RNA copies revealed a 4-5 and 3-4 log reduction at both time points in the prophylactic and therapeutic group, indicating a drastic decrease of SARS-CoV-2 particles in lungs after CV07-209 application. Reduced virus replication and cell infection was confirmed by lowered detection of subgenomic viral RNA ( Figure 5C and 5D). However, genomic and subgenomic RNA levels from nasal washes and laryngeal swaps were similar between all groups, indicating virus replication in the upper airways ( Figure 5C and 5D).

Additionally, we performed histopathological analyses of infected hamsters. As expected, all lungs from control-treated animals sacrificed at 3 dpi revealed typical histopathological signs of necro-suppurative pneumonia with suppurative bronchitis, necrosis of bronchial epithelial cells and endothelialitis ( Figure 6A ).

At 5 dpi, control-treated animals showed marked bronchial hyperplasia, severe interstitial pneumonia with marked type II alveolar epithelial cell hyperplasia and endothelialitis ( Figure 6D ). In contrast, animals receiving prophylactic treatment with CV07-209 showed no signs of pneumonia, bronchitis, necrosis of bronchial epithelial cells, or endothelialitis at 3 dpi. A mild interstitial pneumonia with mild type II alveolar epithelial cell hyperplasia became apparent 5 dpi. Animals receiving therapeutic treatment with CV07-209 also showed a marked reduction of histopathological signs of COVID-19 pathology, although at both time points one out of three animals showed mild bronchopulmonary pathology with signs of interstitial J o u r n a l P r e -p r o o f pneumonia and endothelialitis. These qualitative findings were mirrored in the reduction of the bronchitis and edema scores ( Figure 6B , 6E and Table S6 ).

To confirm the absence of viral particles under CV07-209 treatment, we performed in-situ hybridization of viral RNA at 3 dpi. No viral RNA was detectable in the prophylactic group, whereas all animals in the control group and one in the therapeutic group revealed intensive staining of viral RNA in proximity of bronchial epithelial cells ( Figure 6C ). Taken together, these findings show that systemic application of SARS-CoV-2 neutralizing mAb CV07-209 protects hamsters from COVID-19 lung pathology and weight loss in both the prophylactic and the therapeutic setting.

Driven by the pandemic spread of COVID-19 in early 2020, numerous groups have reported the isolation, characterization, structural analysis and animal model application of SARS-CoV-2 neutralizing mAbs Brouwer et al., 2020; Cao et al., 2020; Ju et al., 2020; Kreer et al., 2020; Robbiani et al., 2020; Rogers et al., 2020; Wec et al., 2020) . In many places, our work confirms previous results, including the observation of a shared antibody response against the SARS-CoV-2 spike protein, the identification of ACE2 blocking as an important mechanism of virus neutralization, the isolation of highaffinity near-germline antibodies, and the in vivo efficacy of prophylactic mAb applications. Additionally, our results add several findings to the growing knowledge on the humoral immune response in SARS-CoV-2

infections.

First, we provide two structures of neutralizing mAbs identified in this study binding to the RBD of SARS-CoV-2 at resolutions of 2.55 and 2.70 Å, allowing detailed characterization of the target epitopes and the SARS-CoV-2 neutralization mechanism of these two mAbs. SARS-CoV-2 mAbs can compete with ACE2 binding and exert neutralizing activity by inhibiting viral particle binding to host cells Brouwer et al., 2020; Cao et al., 2020; Ju et al., 2020; Kreer et al., 2020; Robbiani et al., 2020; Rogers et al., 2020; Wec et al., 2020) , a key mechanism previously identified in SARS-CoV neutralizing antibodies (Prabakaran et al., 2006; ter Meulen et al., 2006) . Steric hindrance of mAbs blocking ACE2 binding to the J o u r n a l P r e -p r o o f RBD provides one mechanistic explanation of virus neutralization Cao et al., 2020; Wu et al., 2020) . CV07-250 clearly belongs to this category of antibodies, as its epitope lies within the ACE2 binding site and it approaches the RBD from a similar angle as ACE2. In contrast, the epitope of CV07-270 only partially overlaps with the ACE2 binding site and approaches the RBD ridge from a different angle. In line with these findings, competition of CV07-270 with ACE2 binding as detected by ELISA was very weak. Its mechanism of virus neutralization therefore remains elusive. Of note, there have been reports of neutralizing antibodies targeting epitopes distant to the ACE2 binding site (Chi et al., 2020) . Future research will need to clarify if additional mechanisms like triggering conformational changes in the spike protein upon antibody binding contribute to virus neutralization, as reported for SARS-CoV (Walls et al., 2019) .

Secondly, the majority of our SARS-CoV-2 mAbs are close to germline configuration, supporting previous studies (Kreer et al., 2020; Robbiani et al., 2020) . Binding of some antibodies to HEp-2 cells was reported before (Kreer et al., 2020) , a finding we could confirm in our cohort. Given the increased probability of auto-reactivity of near-germline antibodies, we additionally examined for reactivity of SARS-CoV-2 mAbs with unfixed murine tissue, allowing the detection of reactivity to potential self-antigens in their natural conformation. Indeed, we found that a fraction of SARS-CoV-2 neutralizing antibodies also bound to brain, lung, heart, kidney or gut expressed epitopes. Such reactivity with host antigens should ideally be prevented by immunological tolerance mechanisms, but complete exclusion of such antibodies would generate "holes" in the antibody repertoire. In fact, HIV utilizes epitopes shared by its envelope and mammalian self-antigens, thus harnessing immunological tolerance to impair anti-HIV antibody responses (Yang et al., 2013) and impeding successful vaccination (Jardine et al., 2016) . To defy viral escape in HIV, but similarly COVID-19, anergic strongly self-reactive B cells likely enter germinal centers and undergo clonal redemption to mutate away from self-reactivity, while retaining HIV or SARS-CoV-2 binding (Reed et al., 2016) . Interestingly, longitudinal analysis of mAbs in COVID-19 showed that the number of SHM in SARS-CoV-2-neutralizing antibodies only marginally increased over time (Kreer et al., 2020) . This finding suggests that the self-reactivity observed in this study may not be limited to mAbs of the early humoral loss, but also to markedly reduced lung pathology. While the findings confirm the efficacy of prophylactic mAb administration as described by other groups in mice, hamsters and rhesus macaques (Cao et al., 2020; Liu et al., 2020; Rogers et al., 2020; Zost et al., 2020) , our work also demonstrates the efficacy of post-exposure treatment in hamsters leading to viral clearance, clinical remission and prevention of lung injury. We provide detailed insights into the lung pathology of SARS-CoV-2 infected hamster at multiple times of the disease course including the regeneration phase. It complements two very recent demonstrations of a therapeutic effect of mAbs in a hamster model of COVID-19 (Baum et al., 2020; . These data expand the growing knowledge on post-exposure treatment from transgenic hACE2 mice (Cao et al., 2020 ) and a mouse model using adenovector delivery of human ACE2 before viral challenge . Collectively, our results indicate that mAb treatment can be fine-tuned for exclusion of self-reactivity with mammalian tissues and that mAb administration can also be efficacious after the infection, which will be the prevailing setting in COVID-19 patients.

While our study confirms the potential of therapeutic mAb applications for the treatment of COVID-19, the interpretation of the data is limited to a first exploration of a short window between viral infection and antibody administration. Although our paradigm mimics the relevant scenario of immediate post-exposure treatment, we cannot conclude whether the therapeutic benefit can also be translated into the more common clinical setting of treatment at heterogenous timepoints after symptoms have occurred. For this, follow-up studies will have to focus on delayed mAb application after symptom onset.

Also, we here describe the reactivity of SARS-CoV-2 mAbs to self-antigens from different tissues. These findings obligate attention, but simultaneously careful interpretation and require thorough investigations to provide better understanding of their functional relevance beyond the observed binding. Amongst others, this includes the identification of non-viral target antigens, functional in vitro studies and in vivo models.

The self-reactive mAbs identified in this study derived from patients without severe extra-pulmonary symptoms. To address a possible connection between self-reactive antibodies and the diverse clinical manifestations in COVID-19, the expression and characterization of mAbs from patients with such disease courses will be needed.

We thank Stefanie Bandura, Matthias Sillmann and Doreen Brandl for excellent technical assistance, Christian Meisel for performing a cardiolipin ELISA and Martin Barner for assistance in generating the circos plot in Figure S4B . (B) Normalized binding to S1 of SARS-CoV-2 for mAbs isolated from antibody secreting cells (▼; blue = S1-binding, grey = not S1-binding). (OD=optical density in ELISA) (C) Normalized binding to S1 of SARS-CoV-2 for mAbs isolated from S1-stained memory B cells (▲; colors like in (B)) (D) S1-binding plotted against the number of somatic hypermutations (SHM) for all S1-reactive mAbs.

(E) Concentration-dependent binding of Top-18 SARS-CoV-2 mAbs to the RBD of S1 (mean±SD from two wells of one experiment).

(F) Concentration-dependent neutralization of authentic SARS-CoV-2 plaque formation by Top-18 mAbs (mean±SD from two independent measurements).

(G) Apparent affinities of mAbs to RBDs (KD determined by surface plasmon resonance) plotted against IC50 of authentic SARS-CoV-2 neutralization.

See also Figures S1, S2, S3, S4, S5 and Tables S1, S2, S3. (E) VH W100h and VH W100k on CDR H3 of CV07-270 make π-π stacking interactions with Y449. VH W100k is also stabilized by a π-π stacking interaction with VL Y49.

(F) VH R100g on CDR H3 of CV07-270 forms an electrostatic interaction with RBD E484 as well as a π-cation interaction with RBD neutralizing mAb CV07-209 or control antibody (mean±SEM from n = 9 animals per group from day -1 to 3, n = 6 from days 4 to 5; n = 3 from days 6 to 13; mixed-effects model with posthoc Dunnett's multiple tests in comparison to control group; significance levels shown as * (p<0.05), ** (p<0.01), *** (p<0.001), **** (p<0.0001), or not shown when not significant). See also Figure 6 and Table S6 . (E) Bronchitis and edema score at 5 dpi. Bars indicate mean.

See also Figure 5 and Table S6 . (B-C) A representative flow cytometry plot from patient CV38 indicating gating on (B) CD19 + CD27 + antibody-secreting cells (ASC) and (C) SARS-CoV-2-S1-stained memory B cells (S1-MBC). Cells were pre-gated on live CD19 + B cells.

(D) Comparison of somatic hypermutation (SHM) count within immunoglobulin V genes combined from heavy and light chains of S1-reactive (S1+, blue) and non-S1-reactive (S1-, grey) mAbs. Statistical significance was determined using a Kruskal-Wallis test with Dunn's multiple comparison test. (ASC: n = 20 S1+, n = 260 S1-; S1-MBC: n = 102 S1+, n = 50 S1-, n-values represent number of mAbs). All expressed mAbs are displayed. Each triangle represents one mAb, isolated from an ASC (▼) or a S1-MBC (▲). Bars indicate mean.

(E-F) Length comparison of complementarity-determining region (CDR) 3 amino acid sequences between S1+ and S1-mAbs within (E) heavy and (F) light chains. Bars indicate mean. Symbols and colors have the same meaning as in (D).

(G) Frequency of RBD-binder (S1+RBD+) and non-RBD-binder (S1+RBD-) relative to all expressed mAbs (upper lanes) and relative to S1+ mAbs (lower lanes). Figure S2 . Comparison of variable gene usage. Related to Figure 1 and Table S2 .

Comparison of gene usage between SARS-CoV-2-S1-reactive (S1+) and non-reactive (S1-) mAbs is shown for immunoglobulin ( and S1-stained memory B cells (S1-MBC). mAbs were considered S1-reactive (S1+) or non-S1-reactive (S1-) based on SARS-CoV-2-S1 ELISA measurements. Antibodies were considered to be clonally expanded when they were isolated from multiple cells.

(B) Circos plot displays all isolated mAbs from ten donors. Interconnecting lines indicate relationship between mAbs that share the same V and J gene on both Ig heavy and light chain. Such public or shared clonotypes in which more than 50% of mAbs are S1reactive are represented as colored lines. Small black angles at the outer circle border indicate expanded clones within the respective donor.

J o u r n a l P r e -p r o o f (C) Properties of public clonotypes from S1+ mAbs according to the colors used in (B) with sequence similarities between mAbs isolated from different donors, also within CDR3.

(D) Public or common antibody response using VH3-53 and VH3-66 genes.

IGHV, IGHJ IGKV, IGKJ, IGLV, IGLJ = V (variable) and J (joining) genes of immunoglobulin heavy, kappa, lambda chains; CDR = complementarity-determining region; n.exp. = not expressed. Figure S4 . Binding kinetic measurements of mAbs to RBD. Related to Figure 1 and Table S3 .

Binding kinetics of mAbs to RBD were modeled (black) from multi-cycle surface plasmon resonance (SPR) measurements (blue, purple, orange). Fitted monovalent analyte model is shown. For CV07-200, neither a bivalent nor a monovalent analyte model described the data accurately (no model is shown). Three out of the 18 selected mAbs for detailed characterization (Top-18) were not analyzed using multi-cycle-kinetics: CV07-270 was excluded as it interacted with the anti-mouse IgG reference surface on initial qualitative measurements. CV07-255 and CV-X2-106 were not analyzed since they showed biphasic binding kinetics and relatively fast dissociation rates in initial qualitative measurements. Non-neutralizing CV03-191, a mAb not included in the Top-18 mAbs, was included in the multicycle experiments as it has the same clonotype as strongly neutralizing CV07-209 ( Figure S4C ). All measurements are performed by using a serial 2-fold dilution of mAbs on reversibly immobilized SARS-CoV-2-S1 RBD-mFc. Table S3 . 

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Jakob Kreye (jakob.kreye@dzne.de).

All requests for materials including antibodies, viruses, plasmids and proteins generated in this study should be directed to the Lead Contact author. Materials will be made available under a Material Transfer Agreement (MTA) for non-commercial usage. 

The patients have given written informed consent and analyses were approved by the Institutional Review

Board of Charité -Universitätsmedizin Berlin, corporate member of Freie Universität Berlin, Humboldt-Universität Berlin, and Berlin Institute of Health, Berlin. All patients in this study were tested positive for SARS-CoV-2 infection by RT-PCR. Most patients belong to a prospective COVID-19 cohort (Kurth et al., 2020) . Patient characteristics are described in Table S1 .

The animal experiment was approved by the Landesamt für Gesundheit und Soziales in Berlin, Germany Twenty-seven six-week old female and male golden Syrian hamsters (Mesocricetus auratus; outbred hamster strain RjHan:AURA, Janvier Labs) were kept in groups of 3 animals in enriched, individually ventilated cages. The animals had ad libitum access to food and water and were allowed to acclimate to these conditions for seven days prior to prophylactic treatment and infection. Cage temperatures and relative humidity were recorded daily and ranged from 22-24°C and 40-55%, respectively.

Recombinant SARS-CoV-2-S1 protein produced in HEK cells (Creative Diagnostics, DAGC091) was covalently labeled using CruzFluor647 (Santa Cruz Biotechnology, sc-362620) according to the manufacturer's instructions.

Using fluorescence-activated cell sorting we sorted viable single cells from freshly isolated peripheral blood mononuclear cells (PBMCs as 7AAD -CD19 + CD27 + CD38 + antibody-secreting cells (ASCs) or SARS-CoV2-S1-enriched 7AAD -CD19 + CD27 + memory B cells ( 

Monoclonal antibodies were generated following established protocols (Kornau et al., 2020; Kreye et al., 2016; Tiller et al., 2008) with modifications as mentioned. We used a nested PCR strategy to amplify variable domains of immunoglobulin heavy and light chain genes from single cell cDNA and analyzed sequences with aBASE module of customized Brain Antibody Sequence Evaluation (BASE) software (Reincke et al., 2020) . Pairs of functional Ig genes were PCR-amplified using specific primers with Q5 

Screening for SARS-CoV-2-specific mAbs was done by using anti-SARS-CoV-2-S1 IgG ELISAs (EUROIMMUN Medizinische Labordiagnostika AG) according to the manufacturer's protocol. mAb containing cell culture supernatants were pre-diluted 1:5, patient sera 1:100. Optical density (OD) ratios were calculated by dividing the OD at 450 nm by the OD of the calibrator included in the kit. OD ratios of 0.5 were considered reactive.

Binding to the receptor-binding domain (RBD) of S1 was tested in an ELISA. To this end, a fusion protein (RBD-Fc) of the signal peptide of the NMDA receptor subunit GluN1, the RBD-SD1 part of SARS-CoV2-S1

(amino acids 319-591) and the constant region of rabbit IgG1 heavy chain (Fc) was expressed in HEK293T cells and immobilized onto 96-well plates from cell culture supernatant via anti-rabbit IgG (Dianova, 711-005-152) antibodies. Then, human mAbs were applied and detected using horseradish peroxidase (HRP)-conjugated anti-human IgG (Dianova, 709-035-149) and the HRP substrate 1-step Ultra TMB (Thermo Fisher Scientific, Waltham, MA). All S1+ mAbs were screened at a human IgG concentration of 10 ng/ml to detect strong RBD binders and the ones negative at this concentration were re-evaluated for RBD reactivity using a 1:5 dilution of the cell culture supernatants. To test for specificity within the coronavirus family, we expressed and immobilized Fc fusion proteins of the RBD-SD1 regions of SARS-CoV, MERS-CoV and the endemic human coronaviruses HCoV-229E, HCoV-NL63, HCoV-HKU1, and HCoV-OC43 and applied mAbs at 1 µg/ml. The presence of immobilized antigens was confirmed by incubation with HRP-conjugated anti-rabbit IgG (Dianova, 711-036-152). Assays for concentrationdependent RBD binding (Fig. 1E) were developed using 1-step Slow TMB (Thermo Fisher Scientific). EC 50 was determined from non-linear regression models using Graph Pad Prism 8.

To evaluate the ability of mAbs to interfere with the binding of ACE2 to SARS-CoV-2 RBD, we expressed ACE2-HA, a fusion protein of the extracellular region of human ACE2 (amino acids 1-615) followed by a Recovery rate of IgG ranged from 60-100%. RBD-Fc captured on ELISA plates was incubated with mAbs at 10 µg/ml for 15 minutes. Then, one volume of biotinylated mcAbs at 100 ng/ml was added and the mixture incubated for additional 15 minutes, followed by detection using HRP-conjugated streptavidin (Roche Diagnostics) and 1-step Ultra TMB. Background by the HRP-conjugated detection antibodies alone was subtracted from all absorbance values.

Antibodies which share same V and J gene on both Ig heavy and light chain are considered to be one clonotype. Such clonotypes are considered public if they are identified in different patients. After identification of public clonotypes, they were plotted in a Circos plot using the R package circlize (Gu et al., 2014) .

To identify the most potent SARS-CoV-2 neutralizing mAb, all 122 S1-reactive mAbs were screened for binding to RBD. 87 were defined as strongly binding to RBD (defined as detectable binding at 10 ng/ml in an RBD ELISA) and then assessed for neutralization of authentic SARS-CoV-2 at 25 and 250 ng/ml using mAb-containing cell culture supernatants. Antibodies were further selected (i) as the strongest neutralizing mAb of the respective donor and / or (ii) with an estimated IC 50 of 25 ng/ml or below and / or (iii) with an estimated IC 90 of 250 ng/ml or below. These were defined as the 18 most potent antibodies (Top-18) and expressed as purified antibodies for detailed biophysical characterization.

The antigen (SARS-CoV-2 S protein-RBD-mFc, Accrobiosystems) was reversibly immobilized on a C1

sensor chip via anti-mouse IgG. Purified mAbs were injected at different concentrations in a buffer consisting of 10 mM HEPES pH 7.4, 150 mM NaCl, 3 mM EDTA, 0.05% Tween 20. CV-X1-126 and CV38-139 were analyzed in a buffer containing 400 mM NaCl as there was a slight upward drift at the beginning of the dissociation phase due to non-specific binding of to the reference flow. Multi-cycle-kinetics analyses were performed in duplicates except for non-neutralizing CV03-191. K a , K d and K D -values were determined using a monovalent analyte model. Recordings were performed on a Biacore T200 instrument at 25°C.

To detect neutralizing activity of SARS-CoV-2-specific mAbs, plaque reduction neutralization tests (PRNT)

were done as described before (Wolfel et al., 2020) . Briefly, Vero E6 cells (1.6 x10 5 cells/well) were seeded in 24-well plates and incubated overnight. For each dilution step, mAbs were diluted in OptiPro and mixed 1:1 with 200 µl virus (Munich isolate 984) (Wolfel et al., 2020) solution containing 100 plaque forming units. The 400 µl mAb-virus solution was vortexed gently and incubated at 37°C for 1 hour. Each 24-well was incubated with 200 µl mAb-virus solution. After 1 hour at 37°, the supernatants were discarded and cells were washed once with PBS and supplemented with 1.2% Avicel solution in DMEM.

After 3 days at 37°C, the supernatants were removed and the 24-well plates were fixed and inactivated using a 6% formaldehyde/PBS solution and stained with crystal violet. All dilutions were tested in duplicates. For PRNT-screening mAb dilutions of 25 and 250 ng of IgG/ml were assessed. IC 50 was determined from non-linear regression models using Graph Pad Prism 8.

Recombinant spike protein-based immunofluorescence assays were done as previously described (Buchholz et al., 2013; Corman et al., 2020; Wolfel et al., 2020) . Briefly, VeroB4 cells were transfected with previously described pCG1 plasmids encoding SARS-CoV-2, MERS-CoV, HCoV-NL63, -229E, -OC43, and -HKU1 spike proteins (Buchholz et al., 2013; Hoffmann et al., 2020) . For transfection, Fugene HD J o u r n a l P r e -p r o o f (Roche) was used in a Fugene to DNA ratio of 3:1. After 24 hours, transfected as well as untransfected VeroB4 cells were harvested and resuspended in DMEM/10% FCS to achieve a cell density of 2.5x10 5 cells/ml each. Transfected and untransfected VeroB4 cells were mixed 1:1 and 50 µl of the cell suspension was applied to each incubation field of a multitest cover slide (Dunn Labortechnik). The multitest cover slides were incubated for 6 hours before they were washed with PBS and fixed with icecold acetone/methanol (ratio 1:1) for 10 minutes. For the immunofluorescence test, the incubation fields were blocked with 5% non-fat dry milk in PBS/0.2% Tween for 60 minutes. Purified mAbs were diluted in EUROIMMUN sample buffer to a concentration of 5 µg/ml and 30 µl of the dilution was applied per incubation field. After 1 hour at room temperature, cover slides were washed 3 times for 5 minutes with PBS/0.2% Tween. Secondary detection was done using a 1:200 dilution of a goat-anti human IgG-Alexa488 (Dianova). After 30 minutes at room temperature, slides were washed 3 times for 5 minutes and rinsed with water. Slides were mounted using DAPI prolonged mounting medium (FisherScientific).

The coding sequence for receptor binding domain (RBD; residues 319-541) of the SARS-CoV-2 spike (S) protein was synthesized and cloned into a customized pFastBac vector (Ekiert et al., 2011) , which is Builder (Schritt et al., 2019) . Iterative model building and refinement were carried out in COOT (Emsley and Cowtan, 2004) and PHENIX (Adams et al., 2010) , respectively. In the CV07-250 + RBD structure, residues 319-337, 357-366, 371-374, 383-396, 516-541 were not modeled due to paucity of electron density. The N and C terminal regions are normally disordered in the SARS CoV-2 RBD structures. These flexible regions are not involved in any other contacts, including crystal lattice contacts, and are on the opposite side of the RBD to the epitope, which is well ordered. In the CV07-270 + RBD structure, Fab residues in a region of the heavy-chain constant domain also have greater mobility as commonly found in Fabs. Likewise, the N and C-terminal residues of 319-333 and 528-541 in both RBD molecules of the asymmetric unit are disordered. Epitope and paratope residues, as well as their interactions, were identified by accessing PDBePISA (Krissinel and Henrick, 2007) at the European Bioinformatics Institute (https://www.ebi.ac.uk/pdbe/prot_int/pistart.html).

Preparations of brain, lung, heart, liver, kidney and gut from 8-12 weeks old C57BL/6J mice were frozen in -50°C 2-methylbutane, cut on a cryostat in 20 µm sections and mounted on glass slides. For tissue reactivity screening according to established protocols (Kreye et al., 2016) , thawed unfixed tissue slices were rinsed with PBS then blocked with blocking solution (PBS supplemented with 2% Bovine Serum Albumin (Roth) and 5% Normal Goat Serum (Abcam)) for 1 hour at room temperature before incubation of mAbs at 5 µg/ml overnight at 4°C. After three PBS washing steps, goat anti-human IgG-Alexa Fluor 488 (Dianova, 109-545-003) diluted in blocking solution was applied for 2 hours at room temperature before additional three washes and mounting using DAPI-containing Fluoroshield. Staining was examined under an inverted fluorescence microscope (Olympus CKX41, Leica DMI6000) or confocal device (Leica TCS SL). For co-staining, tissue was processed as above, but sections were fixed with 4% PFA in PBS for 10 minutes at room temperature before blocking. For co-staining, the following antibodies were used: mouse J o u r n a l P r e -p r o o f

HEp-2 cell reactivity was investigated using the NOVA Lite HEp-2 ANA Kit (Inova Diagnostics) according to the manufacturer's instructions using mAb containing culture supernatant (screening of all S1+ mAbs)

or purified mAbs at 50 µg/ml (polyreactivity testing of CV07-200, CV07-209, CV07-222, CV07-255, CV07-270 and CV38-148) and examined under an inverted fluorescence microscope.

Purified mAbs were screened for reactivity against cardiolipin and beta-2 microglobulin at 50 µg/ml using routine laboratory ELISAs kindly performed by Christian Meisel (Labor Berlin).

Virus stocks for animal experiments were prepared from the previously published SARS-CoV-2 München isolate (Wolfel et al., 2020) . Viruses were propagated on Vero E6 cells (ATCC CRL-1586) in minimal essential medium (MEM; PAN Biotech) supplemented with 10% fetal bovine serum (PAN Biotech) 100 IU/ml Penicillin G and 100 µg/ml Streptomycin (Carl Roth). Stocks were stored at -80°C prior to experimental infections.

For the SARS-CoV-2 challenge experiments, hamsters were randomly distributed into three groups: In the first group (prophylaxis group), animals received an intraperitoneal (i.p.) injection of 18 mg per kg bodyweight of SARS-CoV-2 neutralizing mAb CV07-209 24 hours prior to infection. In the second and third group (treatment and control group, respectively), animals were given the identical mAb amount two hours after infection, either with 18 mg/kg of CV07-209 (treatment group) or with 20 mg/kg of non-reactive isotype-matched mGO53 (control group). Hamsters were infected intranasally with 1x10 5 PFU SARS-CoV-2 diluted in minimal essential medium (MEM; PAN Biotech) to a final volume of 60 µl as previously described (Osterrieder et al., 2020) .

On days 2, 5 and 13 post infection, three hamsters of each group were euthanized by exsanguination under general anesthesia employing a protocol developed specifically for hamsters and consisting of 0.15 mg/kg medetomidine, 2 mg/kg midazolam and 2.5 mg/kg butorphanol applied as a single intramuscular J o u r n a l P r e -p r o o f injection of 200 µl (Nakamura et al., 2017) . Nasal washes, tracheal swabs, and lungs (left and right) were collected for histopathological examinations and/or virus titrations and RT-qPCR. Body weights were recorded daily and clinical signs of all animals were monitored twice daily throughout the experiment.

For histopathological examination and in situ hybridization (ISH) of lung tissues, the left lung lobe was carefully removed and immersed in fixative solution (4% formalin, pH 7.0) for 48 hours. Lungs were then embedded in paraffin and cut in 2 µm sections. For histopathology, slides were stained with hematoxylin and eosin (HE) after dewaxing in xylene and rehydration in decreasing ethanol concentrations. Lung sections were microscopically evaluated in a blinded fashion by board-certified veterinary pathologists to assess the character and severity of pathologic lesions using lung-specific inflammation scoring parameters (Dietert et al., 2017) as previously described for SARS-Cov2 infection in hamsters Osterrieder et al., 2020) . These parameters included severity of interstitial pneumonia, immune cell infiltration by neutrophils, macrophages, and lymphocytes, bronchitis, epithelial necrosis of bronchi and alveoli, hyperplasia of bronchial epithelial cells and type II-alveolar epithelial cells, endothelialitis, perivascular lymphocytic cuffing, as well as alveolar edema and perivascular edema (Table S6 ). The following parameters were evaluated to assess three different scores: (1) the bronchitis score that includes severity of bronchial inflammation and epithelial cell necrosis of bronchi ( Figure 6B and 6E), (2) the regeneration score including hyperplasia of bronchial epithelial cells and type-II-alveolar epithelial cells, and (3) the edema score including alveolar and perivascular edema ( Figure 6B and 6E).

ISH was performed as reported previously (Erickson et al., 2020; Osterrieder et al., 2020) using the ViewRNA™ ISH Tissue Assay Kit (Invitrogen by Thermo Fisher Scientific) following the manufacturer's instructions with the minor following adjustments. Probes for the detection of the Nucleoprotein (N) gene RNA of SARS-CoV-2 (NCBI database NC_045512.2, nucleotides 28,274 to 9,533, assay ID: VPNKRHM), and the murine housekeeping gene eukaryotic translation elongation factor-1α (EF1a; assay ID: VB1-14428-VT, Affymetrix, Inc.), that shares 95% sequence identity with the Syrian hamster, were designed.

Lung sections (2 µm thickness) on adhesive glass slides were dewaxed in xylol and dehydrated in ethanol.

Tissues were incubated at 95°C for 10 minutes with subsequent protease digestion for 20 minutes.

Sections were fixed with 4% paraformaldehyde in PBS (Alfa Aesar, Thermo Fisher) and hybridized with the probes. Amplifier and label probe hybridizations were performed according to the manufacturer's instructions using fast red as the chromogen, followed by counterstaining with hematoxylin for 45 s, washing in tap water for 5 minutes, and mounting with Roti®-Mount Fluor-Care DAPI (4, 6-diaminidino-2phenylindole; Carl Roth). An irrelevant probe for the detection of pneumolysin was used as a control for sequence-specific binding. HE-stained and ISH slides were analyzed and images taken using a BX41 microscope (Olympus) with a DP80 Microscope Digital Camera and the cellSens™ Imaging Software, Version 1.18 (Olympus).

To determine virus titers from 25 mg lung tissue, tissue homogenates were serially diluted and titrated on Vero E6 cells in 12-well-plates. Three days later, cells were formalin-fixed, stained with crystal violet and plaques were counted. RNA was extracted from homogenized lungs, nasal washes and tracheal swabs Detection of subgenomic RNA (sgRNA) was done using by oligonucleotides targeting the leader transcriptional regulatory sequence and by oligonucleotides targeting regions within the E gene, as described previously Wolfel et al., 2020) : Table S3 ) were determined from non-linear regression models using Graph Pad Prism 8.4. Binding kinetics of mAbs to RBD were modeled from multi-cycle surface plasmon resonance measurements ( Figure S4 ) using the Biacore T200 software, version 3.2. For clonality analysis mAbs from an identical donor were identified as clones when sharing the same V and J gene on both heavy and light chain and showing similarities within amino acid sequences of both CDR3. S1 reactivity was determined as a normalized optical density (OD) measured by SARS-CoV-2-S1 ELISA, with values above 0.5 considered as positive. From S1+ mAbs specificity to the RBD was determined using an RBD-ELISA and noted as negative (-), positive (+) or strongly positive (++), when detectable at 10 ng/ml. All strongly positive RBD+ mAbs were screened for neutralization of authentic SARS-CoV-2 (Fig. 1A) , from which the IC50 was estimated.

Abbreviations: ASC = antibody-secreting cell, S1-MBC = S1-SARS-CoV2-enriched memory B cell, Ig = immunoglobulin, HC = heavy chain, KC = kappa chain, LC = chain, IGHV/IGHJ = immunoglobulin heavy chain variable/joining gene, IGKV/IGKJ = immunoglobulin kappa chain variable/joining gene, IGLV/IGLJ = immunoglobulin lambda chain variable/joining gene, CDR = complementarity-determining region, SHM = somatic hypermutations, n.exp. = not expressed, n. t. = not tested.

PHENIX: a comprehensive Python-based system for macromolecular structure solution

Locking the elbow: Improved Antibody Fab Fragments as Chaperones for Structure Determination

Structures of Human Antibodies Bound to SARS-CoV-2 Spike Reveal Common Epitopes and Recurrent Features of Antibodies

REGN-COV2 antibody cocktail prevents and treats SARS-CoV-2 infection in rhesus macaques and hamsters. bioRxiv

Contact investigation of a case of human novel coronavirus infection treated in a German hospital

Potent neutralizing antibodies against SARS-CoV-2 identified by high-throughput single-cell sequencing of convalescent patients' B cells

Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR

Spectrum of pathogen-and model-specific histopathologies in mouse models of acute pneumonia

A highly conserved neutralizing epitope on group 2 influenza A viruses

Coot: model-building tools for molecular graphics

The Family of Chloride Channel Regulator, Calcium-activated Proteins in the Feline Respiratory Tract: A Comparative Perspective on Airway Diseases in Man and Animal Models

Standardization of Reporting Criteria for Lung Pathology in SARS-CoV-2 Infected Hamsters -What Matters?

circlize implements and enhances circular visualization in R

Neurologic Features in Severe SARS-CoV-2 Infection

SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Dexamethasone in Hospitalized Patients with Covid-19 -Preliminary Report

Minimally Mutated HIV-1 Broadly Neutralizing Antibodies to Guide Reductionist Vaccine Design

Human neutralizing antibodies elicited by SARS-CoV-2 infection

Human Cerebrospinal Fluid Monoclonal LGI1 Autoantibodies Increase Neuronal Excitability

Longitudinal Isolation of Potent Near-Germline SARS-CoV-2-Neutralizing Antibodies from COVID-19 Patients

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

Inference of macromolecular assemblies from crystalline state

Studying the pathophysiology of coronavirus disease 2019: a protocol for the Berlin prospective COVID-19 patient cohort (Pa-COVID-19)

Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor

Combinatorial antibody libraries: new advances, new immunological insights

High potency of a bivalent human VH domain in SARS-CoV-2 animal models

Initial antibodies binding to HIV-1 gp41 in acutely infected subjects are polyreactive and highly mutated

Potent neutralizing antibodies directed to multiple epitopes on SARS-CoV-2 spike

Phaser crystallographic software

HKL-3000: the integration of data reduction and structure solution--from diffraction images to an initial model in minutes

Effects of a mixture of medetomidine, midazolam and butorphanol on anesthesia and blood biochemistry and the antagonizing action of atipamezole in hamsters

Age-Dependent Progression of SARS-CoV-2 Infection in Syrian Hamsters

Structure of severe acute respiratory syndrome coronavirus receptor-binding domain complexed with neutralizing antibody

Clonal redemption of autoantibodies by somatic hypermutation away from self-reactivity during human immunization

Brain Antibody Sequence Evaluation (BASE): an easyto-use software for complete data analysis in single cell immunoglobulin cloning

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

Isolation of potent SARS-CoV-2 neutralizing antibodies and protection from disease in a small animal model

Repertoire Builder: high-throughput structural modeling of B and T cell receptors

Human monoclonal antibody combination against SARS coronavirus: synergy and coverage of escape mutants

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

Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein

Unexpected Receptor Functional Mimicry Elucidates Activation of Coronavirus Fusion

Broad neutralization of SARS-related viruses by human monoclonal antibodies

Virological assessment of hospitalized patients with COVID-2019

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

Identification of autoantigens recognized by the 2F5 and 4E10 broadly neutralizing HIV-1 antibodies

A highly conserved cryptic epitope in the receptor binding domains of SARS-CoV-2 and SARS-CoV

A pneumonia outbreak associated with a new coronavirus of probable bat origin

Properties and function of polyreactive antibodies and polyreactive antigen-binding B cells

A Novel Coronavirus from Patients with Pneumonia in China

Validation of assays to monitor immune responses in the Syrian golden hamster (Mesocricetus auratus)

Potently neutralizing and protective human antibodies against SARS-CoV-2

HIGHLIGHTS • Characterization of potent human monoclonal SARS-CoV-2 neutralizing antibodies • Some SARS-CoV-2 antibodies reacted with mammalian self-antigens in different organs • Crystal structures of two antibodies in complex with SARS-CoV-2 RBD at 2.55/2.70 Å • Post-exposure antibody treatment protected from lung damage in infected hamsters