Main

In less than three weeks after its discovery, the Omicron variant was detected in dozens of countries. The WHO classified this lineage (previously known as Pango lineage B.1.1.529) as a variant of concern (VOC) on 26 November 2021 (ref. 1). Preliminary estimates of its doubling time range between 1.2 days and 3.6 days in populations with a high rate of SARS-CoV-2 immunity2,5. Omicron is expected to supplant the currently dominant Delta lineage during the next weeks or months. Little is known about its sensitivity to the humoral immune response. Recent reports indicated that Omicron has a reduced sensitivity to certain monoclonal and polyclonal antibodies6,7,8,9,10, and CD8+ T cell epitopes that were previously characterized in other variants seem to be conserved in Omicron11.

Isolation and characterization of an Omicron variant

We isolated an Omicron variant from a nasopharyngeal swab of an unvaccinated individual who developed moderate symptoms 11 days after returning to Belgium from Egypt. The virus was amplified by one passage in Vero E6 cells. The sequences of the swab and the outgrown virus were identical, and were identified as the Omicron variant (Pango lineage BA.1, GISAID: EPI_ISL_6794907 (swab) and EPI_ISL_7413964 (outgrown)) (Fig. 1a). The spike protein contained 32 changes compared with the D614G strain (belonging to the basal B.1 lineage), which we used here as a reference, including 7 changes in the N-terminal domain (NTD), with substitutions, deletions and a 3-amino-acid insertion (A67V, Δ69–70, T95I, G142D, Δ141–143, Δ211L212I and Ins214EPE), 15 mutations in the receptor-binding domain (RBD) (G339D, S371L, S373P, S375F, K417N, N440K, G446S, S477N, T478K, E484A, Q493R, G496S, Q498R and N501Y, and Y505H), the T574K mutation, 3 mutations close to the furin cleavage site (H655Y, N679K and P681H) and 6 mutations in the S2 region (N764K, D796Y, N856K, Q954H, N969 and L981F) (Fig. 1a). This extensive constellation of changes is unique, but includes at least 11 modifications that have been observed in other lineages and VOCs or at sites that are mutated in other variants (Fig. 1a). Viral stocks were titrated using S-Fuse reporter cells and Vero cells. S-Fuse cells become GFP+ after infection, enabling rapid assessment of infectivity and the measurement of neutralizing antibody levels12,13,14. Syncytia were observed in Omicron-infected S-Fuse cells (Extended Data Fig. 1). Syncytia were smaller after infection with Omicron, relative to Delta (Extended Data Fig. 1). Future experiments will help to determine whether the fusogenic potential of Omicron is different from that of other variants15.

Fig. 1: Neutralization of SARS-CoV-2 variants Delta and Omicron by clinical and preclinical monoclonal antibodies.
figure 1

a, The mutational landscape of the Omicron spike protein. The amino acid modifications are indicated in comparison to the ancestral Wuhan-Hu-1 sequence (GenBank: NC_045512). Consensus sequences of the spike protein were built using the Sierra tool38. The Omicron sequence corresponds to the viral strain that was isolated in Belgium and used in the study (GISAID: EPI_ISL_6794907). Mutations are compared to some pre-existing VOCs and variants of interest. The filled circles indicate changes identical to Omicron. The open circles indicate different substitutions at the same position. b, Neutralization curves of monoclonal antibodies. Dose–response analysis of the neutralization by clinical or preclinical monoclonal antibodies (bamlanivimab, etesevimab, casirivimab, imdevimab, adintrevimab, cilgavimab, tixagevimab, regdanvimab and sotrovimab) and the indicated combinations (bamlanivimab + etesevimab, casirivimab + imdevimab (corresponding to monoclonal antibodies present in Ronapreve), cilgavimab + tixagevimab (corresponding to monoclonal antibodies present in Evusheld)) on Delta (blue dots) and Omicron (red dots) variants. Data are mean ± s.d. of three independent experiments. The IC50 values for each antibody are presented in Extended Data Table 1.

Source data

Phylogenetic analysis of the Omicron lineage

We inferred a global phylogeny by subsampling SARS-CoV-2 sequences that are available on the GISAID EpiCoV database. To better contextualize the isolated virus genome, we performed a focused phylogenetic analysis using all of the Omicron samples deposited on GISAID on 6 December 2021 as background (Extended Data Fig. 2). The tree topology indicates that the Omicron lineage does not directly derive from any of the previously described VOCs. The very long branch of the Omicron lineage in the time-calibrated tree (Extended Data Fig. 2) might reflect a cryptic and potentially complex evolutionary history. At the time of writing, no Omicron genomic sequences from Egypt were available on GISAID, nor do we know of any sequences of travellers that used the same planes. The isolated strain genome showed no close connection to other Belgian Omicron infections. Follow-up analyses with additional genomic data will improve the phylogenetic resolution to determine whether the patient was infected before or after returning to Belgium.

The mutational landscape of Omicron

We highlighted the 29 amino acid substitutions, the 3 amino-acid deletions and a 3-residue insertion that are present in the Omicron spike, with respect to the Wuhan strain, in a 3D model of the protein (Extended Data Fig. 3a). The 15 mutations in the RBD cluster around the trimer interface. The RBD is the target of the most potently neutralizing monoclonal antibodies against SARS-CoV-2, which have been divided into four classes depending of the location of their epitope4,16,17 (Extended Data Fig. 3b). Monoclonal antibodies in classes 1 and 2 compete for binding to human ACE2 (hACE2), whereas those from classes 3 and 4 bind away from the hACE2 interaction surface (Extended Data Fig. 3b). The epitopes of the class 2 and 3 monoclonal antibodies are exposed irrespective of the conformation of the RBD on the spike (‘up’ or ‘down’ configuration18), whereas those of classes 1 and 4 require an RBD in the up conformation. Whereas the previous VOCs displayed mutations only in the region targeted by class 1 and 2 monoclonal antibodies, Omicron mutations are located within the epitopes of all four classes of monoclonal antibodies. The mutations, insertion and deletions in the NTD might also affect the recognition of this domain by antibodies.

Neutralization of Omicron by monoclonal antibodies

We next assessed the sensitivity of Omicron to a panel of human monoclonal antibodies using the S-Fuse assay. We tested nine antibodies that are in clinical use or in development19,20,21,22,23,24,25. These monoclonal antibodies belong to the four main classes of anti-RBD antibodies4,16,17. Bamlanivimab and etesevimab (class 2 and class 1, respectively) are mixed in the Lilly cocktail. Casirivimab and imdevimab (class 1 and class 3, respectively) form the REGN-COV2 cocktail from Regeneron and Roche (Ronapreve). Cilgavimab and tixagevimab (class 2 and class 1, respectively) from AstraZeneca are also used in combination (Evusheld). Regdanvimab (Regkirona; Celltrion) is a class 1 antibody. Sotrovimab (Xevudy) by GlaxoSmithKline and Vir Biotechnology is a class 3 antibody that displays activity against diverse coronaviruses. It targets an RBD epitope outside the receptor-binding motif that includes N343-linked glycans. Adintrevimab (ADG20), developed by Adagio, binds to an epitope that is located between the class 1 and class 4 sites.

We measured the activity of the nine antibodies described above against Omicron and included the Delta variant for comparison purposes (Fig. 1b). As previously reported, bamlanivimab did not neutralize Delta14,26,27. The other antibodies neutralized Delta with a 50% inhibitory concentration (IC50) varying from 3.1 to 325 ng ml−1 (Fig. 1b and Extended Data Fig. 4). Five antibodies (bamlanivimab, etesevimab, casirivimab, imdevimab and regdanvimab) lost antiviral activity against Omicron. The four other antibodies displayed a 2.8-fold to 453-fold increase in IC50 (ranging from 403 to 8,305 ng ml−1) against Omicron. Sotrovimab was the only antibody that displayed a rather similar activity against both strains, with an IC50 of 325 and 917 ng ml−1 against Delta and Omicron, respectively. We also tested the antibodies in combination to mimic the therapeutic cocktails. Bamlanivimab + etesevimab (Lilly) or casirivimab + imdevimab (Ronapreve) were inactive against Omicron. Cilgavimab + tixagevimab (Evusheld) neutralized Omicron with an IC50 of 773 ng ml−1, corresponding to a 58-fold increase relative to Delta (Fig. 1b and Extended Data Fig. 4).

Next, using flow cytometry, we examined the binding of each monoclonal antibody to Vero cells infected with the Delta and Omicron variants (Extended Data Fig. 4). The five clinical antibodies that lost antiviral activity (bamlanivimab, etesevimab, casirivimab, imdevimab and regdanvimab) displayed a strong reduction (8-fold to 47-fold and 11-fold to 242-fold at 1 and 0.1 μg ml−1, respectively) in their binding to Omicron-infected cells compared with Delta-infected cells, as measured by the median fluorescence intensity of the signal (Extended Data Fig. 4). Cilgavimab, sotrovimab, tixagevimab and adintrevimab, which remained partly active, were less impaired in their binding to Omicron-infected cells (2-fold to 9-fold and 1.6-fold to 11-fold decrease at 1 and 0.1 μg ml−1, respectively) (Extended Data Fig. 4).

Thus, Omicron completely or partially escapes neutralization by the tested antibodies. Our results are consistent with findings reported in recent preprints7,8,10. The neutralization escape is correlated with a decrease in the binding of the antibodies to Omicron spike.

Sensitivity of Omicron to sera from vaccinees

We next examined whether vaccine-elicited antibodies neutralized Omicron. To achieve this, we randomly selected 54 individuals from a cohort established in the French city of Orléans, comprising vaccinated individuals who were not previously infected with SARS-CoV-2. The characteristics of the vaccinated individuals are shown in Extended Data Table 1. Sixteen individuals received the Pfizer two-dose vaccine regimen and 18 received the AstraZeneca two-dose vaccine regimen. Twenty individuals who were vaccinated with the Pfizer vaccine received a booster dose. We measured the potency of the sera of these individuals against the Delta and Omicron strains. We used the D614G ancestral strain (belonging to the basal B.1 lineage) as a control (Fig. 2a). We calculated the 50% effective dilution (ED50) for each combination of serum and virus. Sera were first sampled five months after the full two-dose vaccination. For the Pfizer vaccine, the levels of neutralizing antibodies were relatively low against D614G and Delta (median ED50 of neutralization of 329 and 91), reflecting the waning of the humoral response14 (Fig. 2a). We did not detect any neutralization against the Omicron variant with these sera, except one, which displayed a low antiviral activity (Fig. 2a). The percentage of sera with detectable neutralizing activity is shown in Extended Data Fig. 5.

Fig. 2: Sensitivity of the SARS-CoV-2 variants D614G, Delta and Omicron to sera from vaccinated, convalescent or infected-then-vaccinated individuals.
figure 2

Neutralization titres of the sera against the three indicated viral variants are expressed as ED50. a, Neutralizing activity of sera from individuals who were vaccinated with the AstraZeneca (n = 18; left) and Pfizer (n = 16; right) vaccines, sampled at 5 months after the second dose. b, Neutralizing activity of sera from Pfizer-vaccinated recipients sampled one month (M1) after the third injection. n = 20. The dotted line indicates the limit of detection (ED50 = 30). c, The neutralizing activity of sera from convalescent individuals (n = 16), sampled at 6 months after the onset of symptoms (left). Middle, the neutralizing activity of sera from convalescent individuals (n = 23), sampled at 12 months after the onset of symptoms. Right, the neutralizing activity of sera from individuals who were infected then vaccinated (n = 22), sampled one month after the first injection (right). In each panel, data are the mean values from two to three independent experiments. Two-sided Friedman tests with Dunn’s multiple-comparison correction was performed to compare D614G and Omicron with the Delta variant; *P < 0.05, **P < 0.01, ***P < 0.001.

Source data

A similar pattern was observed for the AstraZeneca vaccine. Five months after vaccination, the levels of antibodies neutralizing Delta were low (ED50 of 187 and 68 against D614G and Delta, respectively). No antiviral activity was detected against Omicron in 90% of the sera (Fig. 2a and Extended Data Fig. 5).

We next examined the effect of a Pfizer booster dose, administrated seven months after Pfizer vaccination. The sera were collected one month after the third dose. The booster dose enhanced neutralization titres against D614G and Delta by 39-fold and 49-fold (ED50 of 12,739 and 4,489, respectively, compared with the sampling time 5 months after the full two-dose vaccination). The sera from 1 month after the booster dose were also associated with a strong increase in the neutralization activity against Omicron (ED50 of 722) (Fig. 2b). At this time point, 100% of the tested sera displayed neutralizing activity (Extended Data Fig. 5).

Taken together, these results indicate that Omicron is poorly or not neutralized by sera of vaccinated individuals sampled five months after vaccination. The booster dose triggered detectable cross-neutralization activity against Omicron. However, even after the booster dose, the variant displayed a decrease in ED50 of 18-fold and 6-fold compared with D614G and Delta, respectively.

Sensitivity of Omicron to convalescent sera

We subsequently examined the neutralization ability of sera from convalescent individuals. We randomly selected 61 longitudinal samples from 40 donors in a cohort of infected individuals from Orléans. Individuals were diagnosed with SARS-CoV-2 infection by quantitative PCR with reverse transcription (RT–qPCR) (Extended Data Table 1b). We previously studied the potency of these sera against D614G, Alpha, Beta and Delta isolates13,14. We analysed individuals sampled at a median of 6 and 12 months (M6 and M12) after the onset of symptoms. For the D614G and Delta variants, the neutralization titres were stable or slightly decreased over time (569 and 580 for D614G, and 235 and 143 for Delta, at M6 and M12, respectively)13 (Fig. 2c). The convalescent sera barely neutralized Omicron or did not inhibit Omicron at all at these time points. Only 36% and 39% of the samples displayed neutralizing activity against Omicron at M6 and M12, respectively, whereas the majority (91–94%) were active against Delta (Extended Data Fig. 5).

Twenty-two individuals were vaccinated at M12 with a Pfizer dose. Sera sampled 1 month after vaccination showed a considerable increase in neutralizing antibody titres against the D614G and Delta variants, reaching a median ED50 of 78,162 and 33,536, respectively (Fig. 2d). These sera also neutralized Omicron, with a median ED50 of 1,466 (Fig. 2d). Thus, as shown for other variants13,28,29, a single dose of vaccine boosted cross-neutralizing antibody responses to Omicron in previously infected individuals. However, the neutralization titres were reduced by 53-fold and 23-fold compared with D614G and Delta, respectively.

Discussion

Omicron has opened a new chapter in the COVID-19 pandemic2,30. The principal concerns about this variant include its high transmissibility, as underlined by its rapid spread in different countries, and the presence of more than 55 mutations spanning the whole viral genome. Omicron contains 32 mutations in its spike protein in the NTD, RBD and in vicinity of the furin cleavage site. Some mutations were already present in other VOCs and variants of interest, and have been extensively characterized30,31,32. Owing to their position, they are expected to affect the binding of natural or therapeutic antibodies, to increase affinity to ACE2 and to enhance the fusogenic activity of the spike. Future work will help to determine how this association of mutations affects viral fitness in culture systems and their contribution to the high transmissibility of the variant.

Here we studied the cross-reactivity of clinical or preclinical monoclonal antibodies, as well as of 115 sera from vaccine recipients and long-term convalescent individuals against an infectious Omicron isolate. We report that, among nine monoclonal antibodies in clinical use or in development, six (bamlanivimab, etesevimab, casirivimab, imdevimab, tixagevimab and regdanvimab) were inactive against Omicron. Two other antibodies (cilgavimab and adintrevimab) displayed about a 20-fold increase in IC50. Sotrovimab was less affected by Omicron’s mutations, with an increase in IC50 by only threefold. We also show that Omicron was barely neutralized by sera from vaccinated individuals who were sampled 5 months after the administration of two doses of the Pfizer or AstraZeneca vaccine. Sera from convalescent individuals at 6 or 12 months after infection barely neutralized or did not detectably neutralize Omicron.

The decrease in antibody efficacy helps to explain the high number of breakthrough infections and reinfection cases, and the spread of Omicron in both non-immune and immune individuals33. There is currently no evidence of increased disease severity associated with Omicron compared with Delta, either among naive or immunized individuals. It is probable that, even if pre-existing SARS-CoV-2 antibodies may poorly prevent Omicron infection, anamnestic responses and cellular immunity will be operative to prevent severe forms of the disease34.

We further report that a booster dose of Pfizer vaccine, as well as vaccination of previously infected individuals, strongly increased overall levels of anti-SARS-CoV-2 neutralizing antibodies, well above a threshold allowing inhibition of Omicron. Affinity maturation of antibodies is known to improve the efficacy of the humoral anti-SARS-CoV-2 response over time35,36. This process helps to explain the efficacy of booster doses. However, sera with high antibody levels displayed a 6-fold to 23-fold reduction in neutralization efficacy against Omicron compared with the currently predominant Delta strain.

Potential limitations of our work include the low number of sera analysed from vaccine recipients and convalescent individuals, and the lack of characterization of cellular immunity, which is known to be more cross-reactive than the humoral response. Our results may therefore partly underestimate the residual protection offered by vaccines and previous infections against Omicron infection, in particular with regard to the severity of disease. We analysed only sera sampled one month after the booster dose, or after vaccination of infected individuals. Future work with more individuals and longer survey periods will help to characterize the duration of the humoral response against Omicron. We focused on immune responses elicited by Pfizer and AstraZeneca vaccination. It will be worth determining the potency of other vaccines against this variant.

We focused our analyses on one single viral isolate, corresponding to the archetype Omicron variant sequence (Pango BA.1 lineage). Two related lineages with additional mutations (BA.2 and BA.3) have recently emerged and are less widely spread. It will be worth comparing the behaviour of viral isolates from these more recent lineages with the main BA.1 Omicron strain.

Our results have important public health consequences regarding the use of therapeutic monoclonal antibodies and vaccines. Clinical indications of monoclonal antibodies include pre-exposure prophylaxis in individuals who are unable to mount an immune response, as well as the prevention of COVID-19 in infected individuals who are at high risk for progressing towards severe disease. Antibody-based treatment strategies need to be rapidly adapted to Omicron. Experiments in preclinical models or clinical trials are warranted to assess whether the drops in IC50 are translated into impaired clinical efficacy of the monoclonal antibodies that retain efficacy against Omicron. Most low-income countries display a low vaccination rate, a situation that probably facilitates SARS-CoV-2 spread and continuous evolution. A booster dose improves the quality and the level of the humoral immune response, and is associated with a strong protection against severe forms of the disease37. An accelerated deployment of vaccines and boosters throughout the world is necessary to counteract viral spread. Our results also suggest that there is a need to update and complete the current pharmacopoeia, in particular with regard to vaccines and monoclonal antibodies.

Methods

No statistical methods were used to predetermine sample size. The experiments were not randomized and the investigators were not blinded to allocation during experiments and outcome assessment. Our research complies with all relevant ethical regulation.

Orléans Cohort of convalescent and vaccinated individuals

Since 27 August 2020, a prospective, monocentric, longitudinal, interventional cohort clinical study enrolling 170 individuals infected with SARS-CoV-2 with different disease severities, and 59 non-infected healthy control individuals is ongoing, aiming to describe the persistence of specific and neutralizing antibodies over a 24-month period. This study was approved by the ILE DE FRANCE IV ethical committee. At enrolment, written informed consent was collected and the participants completed a questionnaire that covered sociodemographic characteristics, virological findings (SARS-CoV-2 RT–PCR results, including date of testing), clinical data (date of symptoms onset, type of symptoms and hospitalization), and data related to anti-SARS-CoV-2 vaccination if ever (brand product, and date of first and second doses). The serological status of the participants was assessed every three months. Those who underwent anti-SARS-CoV-2 vaccination had regular blood sampling after first dose of vaccine (ClinicalTrials.gov: NCT04750720). The primary outcome was the presence of antibodies against the SARS-CoV-2 spike protein, as measured using the S-Flow assay. The secondary outcome was the presence of neutralizing antibodies as measured using the S-Fuse assay. For this study, we selected 61 convalescent and 54 vaccinated participants. Some individuals were sampled multiple times. We analysed a total of 115 sera. Study participants did not receive any compensation. The characteristics of each individual from the two cohorts are presented in Supplementary Table 2. The cohorts were constituted before the occurrence of the Omicron variant.

Phylogenetic analysis

To contextualize the isolated Omicron genome, all SARS-CoV-2 sequences available on the GISAID EpiCov database as of 6 December 2021 were retrieved. A subset of complete and high-coverage sequences, as indicated in GISAID, assigned to lineages B.1.529 or BA.1 and BA.2, were randomly subsampled. This subset was included in a global SARS-CoV-2 phylogeny reconstructed with augur and visualized with auspice as implemented in the Nextstrain pipeline (https://github.com/nextstrain/ncov, version from 6 May 2021)39. Within Nextstrain, a random subsampling approach capping a maximum number of sequences per global region was used. The acknowledgment of contributing and originating laboratories for all sequences used in the analysis is provided in Supplementary Table 1.

3D representation of mutations on the surface of spike of B1.617.2 and other variants

The panels in Extended Data Fig. 3 were prepared using The PyMOL Molecular Graphics System, v.2.1 (Schrödinger). The atomic model used (Protein Data Bank: 6XR8) has previously been described40.

S-Fuse neutralization assay

U2OS-ACE2 GFP1-10 or GFP 11 cells, also termed S-Fuse cells, become GFP+ when they are productively infected by SARS-CoV-2 (refs. 12,13). Cells tested negative for mycoplasma. Cells were mixed (ratio 1:1) and plated at 8 × 103 per well in a μClear 96-well plate (Greiner Bio-One). The indicated SARS-CoV-2 strains were incubated with serially diluted monoclonal antibodies or sera for 15 min at room temperature and added to S-Fuse cells. The sera were heat-inactivated for 30 min at 56 °C before use. Then, 18 h later, cells were fixed with 2% PFA, washed and stained with Hoechst (dilution of 1:1,000, Invitrogen). Images were acquired using an Opera Phenix high-content confocal microscope (PerkinElmer). The GFP area and the number of nuclei were quantified using the Harmony software (PerkinElmer). The percentage of neutralization was calculated using the number of syncytia as value with the following formula: 100 × (1 – (value with serum – value in ‘non-infected’)/(value in ‘no serum’ – value in ‘non-infected’)). Neutralizing activity of each serum was expressed as the half maximal effective dilution (ED50). ED50 values (in µg ml−1 for monoclonal antibodies and in dilution values for sera) were calculated with a reconstructed curve using the percentage of the neutralization at the different concentrations.

Characteristics of the patient infected with Omicron

The 32-year-old woman was unvaccinated and developed moderate symptoms on 22 November 2021, 11 days after returning to Belgium from Egypt via Turkey (stop-over to switch flights, without having left the airport). She did not display any risk factor for severe COVID-19 and rapidly recovered. She transmitted the virus to her husband but not to their children. She provided informed written consent to use the swab for future studies. The nasopharyngeal swab tested positive for SARS-CoV-2 on this date. The leftover material of the sample was used in this study after performing routine diagnostics, within the context of the mandate that was provided to UZ/KU Leuven as National Reference Center (NRC) of respiratory pathogens, as described in detail in the Belgian Royal Decree of 09/02/2011.

Virus strains

The reference D614G strain (hCoV-19/France/GE1973/2020) was supplied by the National Reference Centre for Respiratory Viruses hosted by Institut Pasteur and headed by S. van der Werf. This viral strain was supplied through the European Virus Archive goes Global (Evag) platform, a project that has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement no 653316. The variant strains were isolated from nasal swabs using Vero E6 cells and amplified by one or two passages. Delta was isolated from a nasopharyngeal swab of a hospitalized patient returning from India14. The swab was provided and sequenced by the laboratory of Virology of Hopital Européen Georges Pompidou (Assistance Publique, Hopitaux de Paris). The Omicron-positive sample was cultured on Vero E6 cells as previously described41. Viral growth was confirmed by RT–qPCR three days after infection. At day 6 after infection, a cytopathic effect was detected and a full-length sequencing of the virus was performed. The Omicron strain was supplied and sequenced by the NRC UZ/KU Leuven. Both patients provided informed consent for the use of the biological materials. Titration of viral stocks was performed on Vero E6 cells, with a limiting dilution technique enabling the calculation of the median tissue culture infectious dose, or on S-Fuse cells. Viruses were sequenced directly on nasal swabs, and after one or two passages on Vero cells. The sequences were deposited on GISAID immediately after their generation (D614G: EPI_ISL_414631; Delta ID: EPI_ISL_2029113; Omicron ID: EPI_ISL_6794907.

Flow cytometry

Vero cells were infected with the indicated viral strains at a multiplicity of infection of 0.01. Two days after, cells were detached using PBS-0.1% EDTA and transferred into U-bottom 96-well plates (50,000 cells per well). Cells were then incubated for 30 min at 4 °C with the indicated monoclonal antibodies (1 or 0.1 µg ml−1) in MACS buffer (PBS, 5 g l−1 BSA, 2 mM EDTA). Cells were washed with PBS and stained using anti-IgG AF647 (1:600 dilution in MACS buffer) (Thermo Fisher Scientific). Stainings were also performed on control non-infected cells. Cells were then fixed in 4% PFA for 30 min at room temperature. Data were acquired on the Attune Nxt instrument using Attune Nxt Software v.3.2.1 (Life Technologies) and analysed using FlowJo v.10.7.1 (Becton Dickinson).

Antibodies

Four clinically available antibodies (bamlanivimab, casirivimab, etesevimab and imdevimab) were provided by CHR Orleans. The other human SARS-CoV-2 anti-RBD neutralizing antibodies (ADG20 or adintrevimab, AZD1061 (COV2-2130) or cilgavimab, AZD8895 (COV2-2196) or tixagevimab, CT-P59 or regdanvimab, LY-CoV016 (CB6) or etesevimab, LY-CoV555 or bamlanivimab, REGN10933 or casirivimab, REGN10987 or imdevimab, and VIR-7831 (S309) or sotrovimab19,20,21,22,23,24,25 were produced as follows. DNA fragments encoding their IgH and IgL variable domains were synthetized (Life Technologies, Thermo Fisher Scientific). Purified digested DNA fragments were cloned into human Igγ1- and Igκ-/Igλ-expressing vectors42 and recombinant IgG1 antibodies were produced by transient co-transfection of Freestyle 293-F suspension cells (Thermo Fisher Scientific) using the PEI-precipitation method as previously described43. IgG1 antibodies were purified by batch/gravity-flow affinity chromatography using protein G Sepharose 4 Fast Flow Beads (Cytivia) according to the manufacturer’s instructions, dialysed against PBS using Slide-A-Lyzer dialysis cassettes (Thermo Fisher Scientific), quantified using the NanoDrop 2000 instrument (Thermo Fisher Scientific), and checked for purity and quality on a silver-stained SDS–PAGE gel (3–8% Tris-Acetate Novex, Thermo Fisher Scientific). The pan-coronavirus anti-S2 non-neutralizing antibody Ab-10 was previously described13,14.

Statistical analysis

Flow cytometry data were analysed using FlowJo v.10 (TriStar). Calculations were performed using Excel 365 (Microsoft). Figures were generated using Prism 9 (GraphPad Software). Statistical analysis was conducted using GraphPad Prism 9. Statistical significance between different groups was calculated using the tests indicated in each figure legend.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this paper.