Human subjects
A total of 4 patients with COVID-19 including 3 convalescent cases and one acute case were recruited between February and May 2020. All patients were confirmed by reverse-transcription polymerase chain reaction (RT-PCR) as described previously 28. Clinical and laboratory findings were entered into a predesigned database. Written informed consent was obtained from all patients. This study was approved by the Institutional Review Board of The University of Hong Kong/Hospital Authority Hong Kong West Cluster, the Hong Kong East Cluster Research Ethics Committee, and the Kowloon West Cluster Research Ethics Committee (UW 13-265, HKECREC-2018-068, KW/EX-20-038[144-26]).
Syrian hamsters
The animal experimental plan was approved by the Committee on the Use of Live Animals in Teaching and Research (CULATR 5359-20) of the University of Hong Kong (HKU). Male and female golden Syrian hamsters (Mesocricetus auratus) (aged 6–10 weeks) were purchased from the Chinese University of Hong Kong Laboratory Animal Service Centre through the HKU Laboratory Animal Unit (LAU). The animals were kept in Biosafety Level-2 housing and given access to standard pellet feed and water ad libitum following LAU’s standard operational procedures (SOPs). The viral challenge experiments were then conducted in our Biosafety Level-3 animal facility following SOPs strictly, with strict adherence to SOPs
Cell lines
HEK293T cells, HEK293T-hACE2 cells Vero-E6 cells, HK2 cells and Vero-E6-TMPRSS2 cells were maintained in DMEM containing 10% FBS, 2 mM L-glutamine, 100 U/mL/mL penicillin and incubated at 37 ℃ in a 5% CO2 setting 62. Expi293FTM cells were cultured in Expi293TM Expression Medium (Thermo Fisher Scientific) at 37 ℃ in an incubator with 80% relative humidity and a 5% CO2 setting on an orbital shaker platform at 125 ±5 rpm/min (New Brunswick innova™ 2100) according to the manufacturer’s instructions.
ELISA analysis of plasma and antibody binding to RBD and trimeric spike
The recombinant RBD and trimeric spike proteins derived from SARS-CoV-2 (Sino Biological) were diluted to final concentrations of 1 μg/mL/mL, then coated onto 96-well plates (Corning 3690) and incubated at 4 °C overnight. Plates were washed with PBS-T (PBS containing 0.05% Tween-20) and blocked with blocking buffer (PBS containing 5% skim milk or 1% BSA) at 37 °C for 1 h. Serially diluted plasma samples or isolated monoclonal antibodies were added to the plates and incubated at 37 °C for 1 h. Wells were then incubated with a secondary goat anti-human IgG labelled with horseradish peroxidase (HRP) (Invitrogen) or with a rabbit polyclonal anti-human IgA alpha-chain labelled with HRP (Abcam) and TMB substrate (SIGMA). Optical density (OD) at 450 nm was measured by a spectrophotometer. Serially diluted plasma from healthy individuals or previously published monoclonal antibodies against HIV-1 (VRC01) were used as negative controls.
Isolation of RBD-specific IgG+ single memory B cells by FACS
RBD-specific single B cells were sorted as previously described 75. In brief, PBMCs from infected individuals were collected and incubated with an antibody cocktail and a His-tagged RBD protein for identification of RBD-specific B cells. The cocktail consisted of the Zombie viability dye (Biolegend), CD19-Percp-Cy5.5, CD3-Pacific Blue, CD14-Pacific Blue, CD56-Pacific Blue, IgM-Pacific Blue, IgD- Pacific Blue , IgG-PE, CD27-PE-Cy7 (BD Biosciences) and the recombinant RBD-His described above. Two consecutive staining steps were conducted: the first one used an antibody and RBD cocktail incubation of 30 min at 4 °C; the second staining involved staining with anti-His-APC and anti-His-FITC antibodies (Abcam) at 4 °C for 30 min to detect the His tag of the RBD. The stained cells were washed and resuspended in PBS containing 2% FBS before being strained through a 70-μm cell mesh filter (BD Biosciences). RBD-specific single B cells were gated as CD19+CD27+CD3-CD14-CD56-IgM-IgD-IgG+RBD+ and sorted into 96-well PCR plates containing 10 μL of RNAase-inhibiting RT-PCR catch buffer (1M Tris-HCl pH 8.0, RNase inhibitor, DEPC-treated water). Plates were then snap-frozen on dry ice and stored at −80 °C until the reverse transcription reaction.
Single B cell RT-PCR and antibody cloning
Single memory B cells isolated from PBMCs of infected patients were cloned as previously described 76. Briefly, one-step RT-PCR was performed on sorted single memory B cell with a gene specific primer mix, followed by nested PCR amplifications and sequencing using the heavy chain and light chain specific primers. Cloning PCR was then performed using heavy chain and light chain specific primers containing specific restriction enzyme cutting sites (heavy chain, 5′-AgeI/3′-SalI; kappa chain, 5′-AgeI/3′-BsiWI). The PCR products were purified and cloned into the backbone of antibody expression vectors containing the constant regions of human Igγ1 or Igα1 and Igα2. The Igα1 and Igα2 vectors were purchased from InvivoGen (pfusess-hcha1 for IgA1 and pfusess-hcha2m1 for IgA2). The constructed plasmids containing paired heavy and light chain expression cassettes were co-transfected into 293T cells (ATCC) grown in 6-well plates. Antigen-specific ELISA and pseudovirus-based neutralization assays were used to analyze the binding capacity to SARS-CoV-2 RBD and the neutralization capacity of transfected culture supernatants, respectively.
Genetic analysis of the BCR repertoire
Heavy chain and light chain germline assignment, framework region annotation, determination of somatic hypermutation (SHM) levels (in nucleotides) and CDR loop lengths (in amino acids) were performed with the aid of the IMGT/HighV-QUEST software tool suite (www.imgt.org/HighV-QUEST). Sequences were aligned using Clustal W in the BioEdit sequence analysis package (Version 7.2). Antibody clonotypes were defined as a set of sequences that share genetic V and J regions as well as an identical CDR3.
Antibody production and purification
The paired antibody VH/VL chains were cloned into Igγ, Igα1 or Igα2 and Ig expression vectors using T4 ligase (NEB). For production of IgG and monomeric IgA, the plasmids with paired heavy chain (IgG, IgA1, IgA2) and light chain genes were co-transfected into Expi293TM expression system (Thermo Fisher Scientific) following the manufacturer’s protocol to produce recombinant monoclonal antibodies. For dIgA antibody production, plasmids of paired heavy chain (IgA1, IgA2) and kappa light chain together with a J chain were co-transfected into Expi293TM expression system (Thermo Fisher Scientific) at the ratio of 1:1:1 following the manufacturer’s instructions. Antibodies produced from cell culture supernatants were purified immediately by affinity chromatography using recombinant Protein G-Agarose (Thermo Fisher Scientific) or CaptureSelectTM IgA Affinity Matrix (Thermo Fisher Scientific) according to the manufacturer’s instructions, to purify IgG and IgA, respectively. The purified antibodies were concentrated by an Amicon ultracentrifuge filter device (molecular weight cut-off 10 kDa; Millipore) to a volume of 0.2 mL in PBS (Life Technologies), and then stored at 4 °C or -80 °C for further characterization.
Size exclusion chromatography
The prepacked HiLoad 26/60 SuperdexTM 200pg (code No. 17-1071-01, Cytiva) column was installed onto the Amersham Biosciences AKTA FPLC system. After column equilibration with 2 column volumes (CV) of PBS, the concentrated IgA antibodies were applied onto the column using a 500-ul loop at a flow rate of 2 mL/min. Dimers of IgA1 or IgA2 were separated from monomers upon washing with 2 CV of PBS. The milli-absorbance unit at OD280nm was recorded during the washing process. 2 mL-fractions were collected, pooled, concentrated and evaluated by western blot using mouse anti-IGJ monoclonal antibody [KT109] (Abcam) and rabbit anti-human IgA alpha chain antibody (Abcam).
Pseudovirus-based neutralization assay
The neutralizing activity of NAbs was determined using a pseudotype-based neutralization assay as we previously described 77. Briefly, The pseudovirus was generated by co-transfection of 293T cells with pVax-1-S-COVID19 and pNL4-3Luc_Env_Vpr, carrying the optimized spike (S) gene (QHR63250) and a human immunodeficiency virus type 1 backbone, respectively 77. Viral supernatant was collected at 48 h post-transfection and frozen at -80 °C until use. The serially diluted monoclonal antibodies or sera were incubated with 200 TCID50 of pseudovirus at 37 °C for 1 hour. The antibody-virus mixtures were subsequently added to pre-seeded HEK 293T-ACE2 cells. 48 hours later, infected cells were lysed to measure luciferase activity using a commercial kit (Promega, Madison, WI). Half-maximal (IC50) or 90% (IC90) inhibitory concentrations of the evaluated antibody were determined by inhibitor vs. normalized response -- 4 Variable slope using GraphPad Prism 6 or later (GraphPad Software Inc.).
Neutralization activity of monoclonal antibodies against authentic SARS-CoV-2
The SARS-CoV-2 focus reduction neutralization test (FRNT) was performed in a certified Biosafety level 3 laboratory. Neutralization assays against live SARS-CoV-2 were conducted using a clinical isolate (HKU-001a strain, GenBank accession no: MT230904.1) previously obtained from a nasopharyngeal swab from an infected patient 78. The tested antibodies were serially diluted, mixed with 50 μL of SARS-CoV-2 (1×103 focus forming unit/mL, FFU/mL) in 96-well plates, and incubated for 1 hour at 37°C. Mixtures were then transferred to 96-well plates pre-seeded with 1×104/well Vero E6 cells and incubated at 37°C for 24 hours. The culture medium was then removed and the plates were air-dried in a biosafety cabinet (BSC) for 20 mins. Cells were then fixed with a 4% paraformaldehyde solution for 30 min and air-dried in the BSC again. Cells were further permeabilized with 0.2% Triton X-100 and incubated with cross-reactive rabbit sera anti-SARS-CoV-2-N for 1 hour at RT before adding an Alexa Fluor 488 goat anti-rabbit IgG (H+L) cross-adsorbed secondary antibody (Life Technologies). The fluorescence density of SARS-CoV-2 infected cells were scanned using a Sapphire Biomolecular Imager (Azure Biosystems) and the neutralization effects were then quantified using Fiji software (NIH).
Antibody binding kinetics, and competition with the ACE2 receptor measured by Surface Plamon Resonance (SPR)
The binding kinetics and affinity of recombinant monoclonal antibodies for the SARS-CoV-2 spike protein (ACROBiosystems) were analysed by SPR (Biacore 8K, GE Healthcare). Specifically, the spike protein was covalently immobilized to a CM5 sensor chip via amine groups in 10mM sodium acetate buffer (pH 5.0) for a final RU around 500. SPR assays were run at a flow rate of 30 mL/min in HEPES buffer. For conventional kinetic/dose-response, serial dilutions of monoclonal antibodies were injected across the spike protein surface for 180s, followed by a 600s dissociation phase using a multi-cycle method. Remaining analytes were removed in the surface regeneration step with the injection of 10 mM glycine-HCl (pH 2.0) for 2×30s at a flow rate of 30 µl/min. Kinetic analysis of each reference subtracted injection series was performed using the Biacore Insight Evaluation Software (GE Healthcare). All sensorgram series were fit to a 1:1 (Langmuir) binding model of interaction. Before evaluating the competition between antibodies and the human ACE2 peptidase domain, both the saturating binding concentrations of antibodies and of the ACE2 protein (ACROBiosystems) for the immobilized SARS-CoV-2 spike protein were determined separately. In the competitive assay, antibodies at the saturating concentration were injected onto the chip with immobilized spike protein for 120s until binding steady-state was reached. ACE2 protein also used at the saturating concentration was then injected for 120s, followed by another 120s of injection of antibody to ensure a saturation of the binding reaction against the immobilized spike protein. The differences in response units between ACE2 injection alone and prior antibody incubation reflect the antibodies’ competitive ability against ACE2 binding to the spike protein.
Hamster experiments
In vivo evaluation of monoclonal antibody B8-IgG1, B8-mIgA1, B8-mIgA2, B8-dIgA1, B8-dIgA2 in the established golden Syrian hamster model of SARS-CoV-2 infection was performed as described previously, with slight modifications 48. Approval was obtained from the University of Hong Kong (HKU) Committee on the Use of Live Animals in Teaching and Research. Briefly, 6-8-week-old male and female hamsters were housing with access to standard pellet feed and water ad libitum until live virus challenge in the BSL-3 animal facility. The hamsters were randomized from different litters into experimental groups. Experiments were performed in compliance with the relevant ethical regulations 48. For prophylaxis studies, 24 hours before live virus challenge, three groups of hamsters were intraperitoneally or intranasally administered with one dose of test antibody in phosphate-buffered saline (PBS) at the indicated dose. At day 0, each hamster was intranasally inoculated with a challenge dose of 100 μL of Dulbecco’s Modified Eagle Medium containing 105 PFU of SARS-CoV-2 (HKU-001a strain, GenBank accession no: MT230904.1) under anaesthesia with intraperitoneal ketamine (200 mg/kg) and xylazine (10 mg/kg). For pre-treatment study, each hamster received one 1.5 mg/kg dose of intraperitoneal B8-IgG1 at 24 , 48, 72 hours (n=4 per group) after virus challenge. The hamsters were monitored twice daily for clinical signs of disease. Syrian hamsters typically clear virus within one week after SARS-CoV-2 infection. Accordingly, animals were sacrificed for analysis at day 4 after virus challenge with high viral loads 48. Half the nasal turbinate, trachea, and lung tissues were used for viral load determination by quantitative SARS-CoV-2-specific RdRp/Hel RT-qPCR assay 28 and infectious virus titration by plaque assay 48.
Cryo-EM sample preparation and data acquisition
The purified SARS-CoV-2 S-B8 protein complexes were concentrated before being applied to the grids. Aliquots (4 μL) of the protein complex were placed on glow-discharged holey carbon grids (Quantifoil Au R1.2/1.3, 300 mesh). The grids were blotted and flash-frozen in liquid ethane cooled by liquid nitrogen with a Vitrobot apparatus (Mark IV, ThermoFisher Scientific). The grids sample quality was verified with an FEI Talos Arctica 200-kV electron microscope (Thermo Fisher Scientific). The verified grids with optimal ice thickness and particle density were transferred to a Titan Krios operating at 300 kV and equipped with a Cs corrector, a Gatan K3 Summit detector (Gatan Inc.) and a GIF Quantum energy filter (slit width 20 eV). Micrographs were recorded in the super-resolution mode with a calibrated pixel size of 0.54895 Å. Each movie has a total accumulated exposure of 50 e−/Å2 fractionated in 32 frames. The final image was binned 2-fold to a pixel size of 1.0979 Å. AutoEMation was used for the fully automated data collection. The defocus value of each image, which was set from -1.0 to -2.0 μm during data collection, was determined by Gctf. Data collection statistics are summarized in Supplementary Table 11.
Cryo-EM data processing
The procedure for image processing of SARS-CoV-2 S-B8 complex is presented in Supplementary Fig. 2. In brief, Motion Correction (MotionCo2), CTF-estimation (GCTF) and non-templated particle picking (Gautomatch, http://www.mrc-lmb.cam.ac.uk/kzhang/) were automatically executed by the TsinghuaTitan.py program (developed by Dr. fang Yang). Sequential data processing was carried out on RELION 3.0 and RELION 3.1. Initially, 2,436,776 particles were auto-picked by Gautomatch or RELION 3.0 from 4213 micrographs. After several 2D classifications, 1,451,176 particles were selected and applied for 3D classification with one class. Two different states were obtained after further 3D classification: 3 RBD in up conformation bound with B8 Fab (3u), and 2 up RBDs and 1 down RBD with each bound to a B8 Fab (2u1d). 616,799 particles for the 2u1d state and 351,095 particles for the 3u state were subjected to 3D auto-refinement, yielding final resolutions at 3.21 Å and 3.06 Å, respectively. Further CTF refinement and Bayesian polishing improved the resolution to 2.65 Å (2u1d, C1 symmetry) and 2.67 Å (3u, C3 symmetry) with better map quality. To improve the RBD-B8 portion map density, focused local search classification was applied for each RBD-B8 portion with an adapted soft mask. The best classes for each RBD-Fab portion were selected and yielded a final resolution at 3.56 Å (RBD-Fab1, up), 3.34 Å (RBD-Fab2, up), 3.69 Å (RBD-Fab3, down), 3.87 Å (RBD-Fab3, up) from 479,305, 508,653, 656,429, and 136,482 particles, respectively. Further CTF refinement and Bayesian polishing improved the resolution of RBD-Fab2 to 3.11 Å with better map quality. RBD-Fab maps were fitted onto the whole structure map using Chimera, then combined using PHENIX combine_focused_maps. The reported resolutions were estimated with the gold-standard Fourier shell correlation (FSC) cutoff of 0.143 criterion. Data processing statistics are summarized in Supplementary Table 11.
Model building and structure refinement
The spike model (PDB code: 6VSB) and the initial model of the B8 Fab generated by SWISS-Model were fitted into the EM density map, and further manually adjusted with Coot. Glocusides were built manually with carbohydrate tool in Coot. The atomic models were refinement using Phenix in real space with secondary structure and geometry restraints. The final structures were validated using Phenix.molprobity. UCSF Chimera, ChimeraX and PyMol were used for map segmentation and figure generation. Model refinement statistics are summarized in Supplementary Table 11.
SARS-CoV-2 infection of reconstructed human nasal epithelia
MucilAirTM, corresponding to reconstructed human nasal epithelium cultures differentiated in vitro for at least 4 weeks, were purchased from Epithelix (Saint-Julien-en-Genevois, France). The cultures were generated from pooled nasal tissues obtained from 14 human adult donors. Cultures were maintained in air/liquid interface (ALI) conditions in transwells with 700 µL of MucilAirTM medium (Epithelix) in the basal compartment, and then kept at 37 °C under a 5% CO2 atmosphere. SARS-CoV-2 infection was performed as previously described 50. Briefly, the apical side of ALI cultures was washed 20 min at 37 °C in MucilairTM medium to remove mucus. Cells were then incubated with 104 plaque-forming units (PFU) of the isolate BetaCoV/France/IDF00372/2020 (EVAg collection, Ref-SKU: 014V-03890; kindly provided by S. Van der Werf). The viral input was diluted in DMEM medium to a final volume 100 µL, and then left on the apical side for 4 h at 37 °C. Control wells were mock treated with DMEM medium (Gibco) for the same duration. Viral inputs were removed by washing twice with 200 µL of PBS (5 min at 37 °C) and once with 200 µL MucilairTM medium (20 min at 37 °C). The basal medium was replaced every 2-3 days. Apical supernatants were harvested every 2-3 days by adding 200 µL of MucilairTM medium on the apical side, with an incubation of 20 min at 37 °C prior to collection. For IgA treatment, cultures were washed once and then pretreated with antibodies added to the apical compartment for 1 h in 50µL. Viral input was then directly added to reach a final volume of 100 µL. The antibodies were added again at day 2 d.p.i. in the apical compartment during an apical wash (20 min at 37 °C). To test the effect of dIgA treatment in the presence of mucus, dIgA were added directly to the apical compartment of MucilAirTM cultures without an initial wash. After IgA treatment for 1h, the virus was added directly to the IgA/mucus mixture and left on the apical side for 4h at 37°C. After viral inoculation, a single brief wash was made to remove the viral input while limiting mucus loss. The cultures were then maintained as in the no-mucus condition.
Viral RNA quantification in reconstructed human nasal epithelia
Apical supernatants were collected, stored at -80 °C until thawing and were then diluted 4-fold in PBS in a 96-well plate. Diluted supernatants were inactivated for 20 min at 80 °C. For SARS-CoV-2 RNA quantification, 1 µL of diluted supernatant was added to 4 µL of PCR reaction mix. PCR was carried out in 384-well plates using the Luna Universal Probe One-Step RT-qPCR Kit (New England Biolabs) with SARS-CoV-2 NP-specific primers (Forward 5'-TAA TCA GAC AAG GAA CTG ATT A-3'; Reverse 5'-CGA AGG TGT GAC TTC CAT G-3') on a QuantStudio 6 Flex thermocycler (Applied Biosystems). A standard curve was established in parallel using purified SARS-CoV-2 viral RNA.
Histopathology and immunofluorescence (IF) staining
The lung and nasal turbinate tissues collected at necropsy were fixed in zinc formalin and then processed into paraffin-embedded tissue blocks. The tissue sections (4 μm) were stained with hematoxylin and eosin (H&E) for light microscopy examination as previously described with modifications 47. For identification and localization of SARS-CoV-2 nucleocapsid protein (NP) in organ tissues, immunofluorescence staining was performed on deparaffinized and rehydrated tissue sections using a rabbit anti-SARS-CoV-2-NP protein antibody together with an AF488-conjugated anti-rabbit IgG (Jackson ImmunoResearch, PA, USA). Briefly, the tissue sections were first treated with antigen unmasking solution (Vector Laboratories) in a pressure cooker. After blocking with 0.1% Sudan black B for 15 min and 1% bovine serum albumin (BSA)/PBS at RT for 30 min, the primary rabbit anti-SARS-CoV-2-NP antibody (1:4000 dilution with 1% BSA/PBS) was incubated at 4°C overnight. This step was followed by incubation with a FITC-conjugated donkey anti-rabbit IgG (Jackson ImmunoResearch) for 30 min and the sections were then mounted in medium with 4’,6-diamidino-2-phenylindole (DAPI). For identification of DC-SIGN expression, we stained the NT slices with rabbit anti-DC-SIGN primary antibody (Abcam) and Alexa Fluor 488 goat anti-rabbit IgG (H+L) cross-adsorbed secondary antibody (Life Technologies) according to the manufacturer’s instructions. For identification of ACE2 expression, the goat anti-ACE2 primary antibody (R&D) and Alexa Fluor 568 donkey anti-goat IgG (H+L) secondary antibodies (Invitrogen) according to the manufacturer’s instructions. All tissue sections were examined, and the fluorescence images and whole section scanning were captured using 5×, 10× and 20× objectives with Carl Zeiss LSM 980. NP+ cells per field were quantified based on the mean fluorescence intensity (MFI) using the ZEN BLACK 3.0 and ImageJ (NIH).
Effects of B8-dIgA on SARS-CoV-2 infection in HK2 cells
HK2 cells were seeded into 24-well plates at the 40-50% confluency and cultured overnight. The B8-dIgA or control dIgA at the concentration of 1, 10, 100, 1000 ng/ml/mL and then mixed with SARS-CoV-2 (1:10 TCID50) and incubated for 1 hour at room temperature. The antibody/virus mixture was then added to HK2 cells after the cell culture medium was removed and washed with PBS once and incubated for 1 hour at 37°C. The infectious medium was replaced with fresh medium containing respective concentration of antibody after washing 3 times with PBS. 24 h later, the infected cells were imaged under fluorescence microscope after staining with AF488-conjugated anti-SARS-CoV-2 NP antibody. Alternatively, the infected cells were lysed and blotted for SARS-CoV-2 NP protein to determine the extent of infection. Tubulin was blotted as the internal control.
B8-dIgA mediated enhancement via CD209
HEK293T cells were seeded into 10-cm dish at 40% confluency and cultured overnight. The HEK293T cells were transfected with human CD209 (Sino Biological) at 70%-90% confluency. The expression of CD209 was measured by flow cytometry. The transfected HEK293T-CD209 cells were seeded into 96-well plates with 2.4×104 cells per well and cultured overnight. The HEK293T-CD209 cells were pre-treated with 10 ng/ml/mL of B8-dIgA or control dIgA and incubated for 6 h prior SARS-CoV-2 infection (MOI=0.05). 24 h later, cells were then fixed with 4% paraformaldehyde solution for 30 min and air-dried in the BSC. Cells were further permeabilized with 0.2% Triton X-100 and incubated with cross-reactive rabbit sera anti-SARS-CoV-2-N for 1 hour at RT before adding Alexa Fluor 488 goat anti-rabbit IgG (H+L) cross-adsorbed secondary antibody (Life Technologies). The fluorescence density of SARS-CoV-2 infected cells was acquired using a Sapphire Biomolecular Imager (Azure Biosystems) and then the MFI of four randomly selected areas of each sample was quantified using Fiji software (NIH).
Effects of B8 antibodies on SARS-CoV-2 mediated cell-cell fusion
Vero-E6 TMPRSS2 cells were seeded into 48-well plates and cultured overnight. After treatment with B8 antibodies at the dose of 3000 ng/ml/mL for 1 hour, HEK293T cells transfected with SARS-CoV-2 spike-GFP were added into the treated Vero-E6 TMPRSS2 cells and co-cultured for 48 hours. The cell-cell fusion between Vero-E6 TMPRSS2 and HEK293T-Spike-GFP was then determined under a fluorescence microscope (Nikon ELIPSE) and the images of randomly selected region were captured using 4× and 10× objectives using the Nikon software.
Re-analysis of published nasal brushing single-cell data
The preprocessed scRNA-seq data from nasal brushing samples of 2 healthy controls and 4 COVID-19 patients were downloaded from Gene Expression Omnibus (GEO) database with accession numbers GSE171488 and GSE164547. Quality control metrics were consistent with the original article [PMID: 34003804] and performed based on the R package Seurat (version 4.0.3) [PMID: 34062119]. Harmony [PMID: 31740819] was used to integrate the samples based on the top 4000 most variable genes obtained with the FindVariableFeatures() function in Seurat. CD14+ (monocyte) cells were extracted for further analysis. The annotation of the cell type was performed by manually checking the marker genes of each cluster identified from the FindAllMarkers() function in Seurat.
Quantification and statistical analysis
Statistical analysis was performed using PRISM 6.0 or later. Ordinary one-way ANOVA and multiple comparisons were used to compare group means and differences between multiple groups. Unpaired Student's t tests were used to compare group means between two groups only. A P-value <0.05 was considered significant. The the number of independent experiments performed, the number of animals in each group, and the specific details of statistical tests are reported in the figure legends and the Methods section.