Cell line
Vero cells (E6 lineage, African Green monkey kidney epithelial cells) were obtained from the American Type Culture Collection (referred as VERO C1008 [Vero 76, clone E6, Vero E6], CRL-1586). Vero E6 cells were cultured in Dulbecco’s minimal essential medium (DMEM, Eurobio Scientific) supplemented with 5 % fetal calf serum (FCS, Eurobio Scientific), 100 IU/ml penicillin,100 mg/ml streptomycin at 37°C.
Study design
The EVLP part of the study (Fig. 1A) was approved by the Agence de la Biomédecine and by the French “Ministère de l’Enseignement Supérieur et de la Recherche, Direction Générale de la Recherche et de l’Innovation” under the number 2020-007, as well as by the ethic committee of the Foch Hospital (IRB00012437). The five donor lungs used in this study were from donation after brain death and were determined to be unsuitable for transplantation (Additional file 1. Anamnesis). Donor lung retrieval was carried out according to current clinical practice using Perfadex (Xvivo Perfusion, Göteborg, Sweden) flush preservation. After transportation at 4°C, lungs were processed to EVLP that was conducted according to the Toronto protocol for 10 h 31 in the BSL3 facilities of the Molecular Virology and Immunology laboratory in Jouy en Josas in the case of the SARS-CoV-2 infection in EVLP, and at the Foch Hospital, Suresnes, in the case of the control EVLP. For SARS-CoV-2 infection in EVLP (3 donor lungs, donors 1, 2, 3), the right lung was connected to the circuit that was filled with 1.5 L of Steen solution supplemented with 1 g methylprednisolone, 1.5 g cefuroxime, and 7500 UI heparin. A flow rate at 24% of the theoretical cardiac output was applied at normothermia. The lung was ventilated at 4 ml/kg of donor body weight with a standard ICU-type ventilator equipped with a connected nebulizer with vibrating meshes (Aerogen®), for the delivery of SARS-CoV-2 (see “Viruses” section). The Steen perfusate (100 ml) was replaced with fresh one every 2 h. The left lung was placed at 4°C and used for sampling before EVLP, designated as 0 h (see sampling section). For control EVLP experiments (2 donor lungs, donors 4 and 5), the whole lung was connected to the circuit that was filled with 1.5 L of Steen solution supplemented with 1 g methylprednisolone, 1.5 g cefuroxime, and 7500 UI heparin. A flow rate at 40% of the theoretical cardiac output was applied at normothermia. The Steen perfusate (100 ml) was replaced with fresh one every 2 h. Views of the general experimental set up are shown in Additional file 2. Experimental set-up. This part of the study has been conducted between September 2020 and May 2021. Donors 1, 2, 3 were negative for SARS-CoV-2 virus, not vaccinated and without history of SARS-CoV-2 infection. Anti-viral IgGs were undetectable in their broncho-alveolar lavages (Additional file 3. ELISAIgG). The lungs from donors 4 and 5 were included in a previous study of our group 32.
The lung isolated monocyte/macrophage part of the study (Fig. 1B) was declared as a “dossier de Conservation et de préparation à des fins scientifiques D’Eléments du Corps Humain”: (CODECOH ) DC N° DC-2020-3981(1). Experiments on human tissues were approved by the regional investigational review board (Comité de Protection des Personnes Île de France VIII, Boulogne-Billancourt, France). Lung tissue samples were obtained from seven patients undergoing surgical resection for lung carcinoma at the Foch Hospital in Suresnes (France) (Additional file 1. Anamnesis). In line with the French legislation on clinical research and as approved by the investigational review board, all the patients gave their informed consent for the use of resected lung tissue forresearch.
Viruses
The SARS-CoV-2 BetaCoV/France/IDF0372/2020 strain (passage 2 on Vero E6 cells), clade 19A, was obtained from the Pasteur Institute Paris, and has been isolated from a patient back from Hubei (China) at the Bichat hospital. The whole sequence is available under the accession number EPI_ISL_410720 (GISAID ID) and belongs to the original Wuhan lineage 33, that will be designated as WL (for Wuhan Lineage) in this paper. For the nebulization of donor 1’s lung (Fig. 1A), the virus stock (passage 4) was produced by amplifications in Vero E6 cells at an initial multiplicity of infection of 0.03 for 3-4 days at 33° C in DMEM, centrifuged at 4000g for 10 min 4°C, aliquoted and stored at −80 °C before titration (titer 2x107 PFU/ml). For infection of monocytes/macrophages with WL (Fig. 1B), a second WL virus stock was similarly produced (titer 106 PFU/ml). The SARS-CoV-2 viral strains UCN1 and UCN15 were isolated in March 2020, during the course of the active epidemic, from nasopharyngeal flocked swabs obtained at the University Hospital of Caen, Normandy, France as described in 34. The spike nucleotide sequences of UCN1 and UCN15 strains as well as the WL sequence are provided in Additional file 4. Viral sequences and they show D614G mutations in both UCN1 and UCN15 strains. UCN1 and UCN15 will be designated as D614G-a and D614G-b in the rest of the paper. For the nebulization of D614G-a in donor 2, a passage 1 was used (titer 5.5x106 PFU/ml) and for infection of monocytes/macrophages, a passage 2 was used (titer 106 PFU/ml). D614G-b was only used for nebulization (donor 3) and was produced by polyethylene glycol concentration of a passage 2 production (titer 108 PFU/ml), following a published protocol 35.
Viral titration
For determination of tissue culture infective dose 50 (TCID50)/ml, Vero E6 cells (2x104/well in 96-well plates) were inoculated with tenfold serial dilutions of viral suspension, cultured at 33°C under 5% CO2 and after 96 h, the cells were fixed with 3.7% paraformaldehyde and stained with 0.2% crystal violet. The cytopathic effect was evaluated and the viral titers were expressed as 50% tissue culture infective dose per ml (TCID50/ml). For determination of particle forming units (PFU)/ml in a plaque assay, Vero E6 cells (3x105/well in 12-well plates) were inoculated with tenfold serial dilutions of viral supernatants for 1 h, overlaid with 1.2% cellulose microcrystalline (Avicel, FMC BioPolymer) in Minimum Essential Medium Eagle + 2% FCS, cultured at 33°C under 5% CO2 for 5 days. Following fixation and crystal violet staining, the number of plaques were counted, and the numbers of PFUs/ml were calculated by taking into account the dilution factor.
Nebulization
Viral nebulization was done using a nebulizer with vibrating meshes, see the “study design” section. The SARS-CoV-2 preparations were adjusted to 6.5 ml in DMEM and the total PFU amounts delivered were 1.2x108 PFUs for WL (donor 1), 3.3x107 PFUs for D614G-a (donor 2) and 107 PFUs for D614G-b (donor 3) that were nebulized in the lungs for about 20 min. The potential effect of nebulization on viral infectivity was evaluated as follows: 106 PFUs of WL were diluted in 6.5 ml DMEM and used for a control nebulization delivered to a 50 ml tube instead of to a lung. Three viral suspension samples were harvested before and after 30 min nebulization for viral detection using viral titration (TCID50) and RT-qPCR.
Sampling during the EVLP part
Lung biopsies (2 g each) were taken from similar lung zones from the upper lobes (wedges) before EVLP (0 h), and after 10 h EVLP, cut in 4 pieces, placed in 10 ml HypoThermosol® (STEMCELL Technologies, Vancouver, Canada) and kept on ice for 10 to max 24 h. This process was shown to maintain lung tissue stability up to 72 h for scRNA-seq 36. In addition, 3 small biopsies (100 mg each) were placed in RNAlater at 0 h, 30 min, 5 h and 10 h after the end of nebulization, for RNA extractions. Perfusion liquid was collected every hour during EVLP and stored at -80°C. A broncho-alveolar lavage (BAL) was done with 100 ml PBS without Ca2+ and Mg2+ in the left lung (un-infected control). The aspirated BAL was spun for 10 min at 470g at 4°C and the supernatant was frozen at -80°C.
Anti-SARS-CoV2 IgG detection
The BALs were thawed, spun at 10000g for 5 min, diluted 1:4 and assayed in duplicates using the SARS-CoV-2 Spike Protein IgG ELISA Kit from Elabscience (Houston, USA) as recommended.
Lung cell isolation
For the EVLP part of the study, the lung tissue from biopsies kept in HypoThermosol was minced finely with scissors, placed in Multi Tissue Dissociation Kit 1 solution as recommended by the manufacturer (Miltenyi Biotec, Bergisch Gladbach, Germany), and incubated at 37°C for 45 min in gentle agitation. The minced preparation was crushed on nylon mesh (1 mm) and filtered through successive nylon filters (500 µm, 100 µm, 40 µm). The cell suspension (referred as total cells) was washed in PBS (470 g, 15 min), processed to erythrocyte lysis, resuspended in RPMI + 2 % FCS, filtered twice on 40 µm and counted by 3 independent measurements with a counting chamber. For donors 2 and 3, both at 0 h and 10 h, MHC class 2 positive cells (referred as MHC class 2pos) were enriched using the anti-human HLA-DR bead kit from Miltenyi Biotech (ref 130-046-101) and MS separation columns (ref 130-042-201), following the manufacturer recommendations. Briefly, 10x106 purified lung cells were incubated with 20 µl beads + 80 µl PBS-EDTA-BSA buffer for 15 min, filtered on 40 µm, and loaded on MS column for positive selection. The total cells and HLA-DR-positively-selected cells were counted using a counting chamber and checked for viability using trypan blue, and showed over 90% viability.
For the isolated monocyte/macrophage part of the study, lung tissue was obtained from lobectomy at distance from tumors with the minimal possible duration of ischemia in the operating room. The lung tissue was immediately processed in the laboratory. Six grams of lung were minced and incubated for 45 minutes at 37°C on a rotary shaker in RPMI 1640 supplemented with 100 IU/ml penicillin,100 mg/ml streptomycin, 2 mM L-glutamine and 10% FCS containing 3 mg/ml collagenase D, 0.25 mg/ml Dnase I (Sigma-Aldrich) and 0.7 mg/ml dispase II (Gibco®, ThermoFisher Scientific, St Aubin, France). The minced preparation was crushed and filtered on a nylon mesh (1 mm diameter) and filtered through successive cell strainers (500 µm, 100 µm, 40 µm). Red blood cells were lysed with erythrocytes lysis buffer. After a wash in PBS, about 108 cells (all instances over 90% viability) were kept in 10% FCS overnight at 4°C before being used for staining, analysis or sorting.
Monocyte/macrophage cell sorting and culture with SARS-CoV-2
Isolated human lung cells were stained with fluorescently labelled mAbs following a 15 min incubation at 4°C with Fc block (1:4 ratio, Miltenyi Biotech). We used a mAb combination that was previously documented to identify lung monocytes and macrophage subsets 37. The mAbs were diluted in RPMI containing 5% horse serum and 1% Hepes to a final dilution or concentration recommended by the manufacturers: anti-human CD45-FITC (clone HI30, 1/20), anti-murine CD11b-APC/Cy7 (clone M1/70, 5 µg/ml), anti-human CD206-APC (clone 19.2, 1/5), anti-human CD14-PE (clone TUK4, 5 µg/ml), anti-human CD16-Alexa700 (clone 3G8, 5 µg/ml), anti-human CD163-PerCp/Cy5.5 (clone GHI/61, 1/20), anti-human CD169-BV605 (clone 7-239, 1/20), anti-human CD43-PerCp/Cy5.5 (1G10, 1/20, BD Bioscience). For each mAb, a labelled isotype-matched control was used and the specificity of the labeling was controlled using the fluorescence minus one method. Dead cells were excluded by DAPI staining. The cell staining results were analyzed using FlowJo 10.8.1 software. The cell subsets were sorted from an initial number of 50-120x106 cells using the “purity 1-2” mode of the CytoFLEX SRT sorter (Beckman Coulter). After centrifugation, the sorted cells were resuspended in X-VIVO 15 serum-free medium (Lonza), 100 U/ml penicillin and 1 µg/ml streptomycin and 5 x 104 cells were plated per well in a 96-well plate. Cells were incubated at 33°C, 5% CO2, with WL and D614G SARS-CoV-2 viruses at 0.1 and 0.001 MOIs, in duplicates, for 24 h (and for 48 h for viral detection). MOI 0.1 corresponds to a classical dose used in the literature, and MOI 0.001 corresponds to an estimate of the viral exposure per cell used in the EVLP part of the study (about 108 PFUs for 1011 cells in the human lung 38). In parallel the virus inoculum was checked for infectivity on Vero E6 cells. Before and at the end of the incubation, the cells were observed on a ZOETM fluorescent imager (BIO-RAD). The culture supernatants were collected at 2 h, and 24 h (48 h in some cases) and stored at -80°C. Additional experiments were conducted in order to analyze IFN gene expression in the four monocyte/macrophages subsets. In that case, 105 cells were plated per well in a 96-well plate in X-VIVO 15 serum-free medium (Lonza), 100 U/ml penicillin and 1 µg/ml streptomycin and incubated at 33°C, 5% CO2, with medium alone or with D614G SARS-CoV-2 at 0.1 MOI or 0.001 MOIs, for 10 h.
Cytokine detection
The supernatants of monocyte/macrophage cultures were assessed for detection of CCL2, CCL3, CCL4, CXCL8, CXCL10, TNFα, IL-1β, IL-1RA, IL-6, IL-10, IL-18 and IFN-a with a Human ProcartaPlexTM Mix&Match12-plex (ThermoFischer Scientific, Waltham, MA) using a MagPix instrument (Luminex, Austin, TX) and the data were analyzed with the Bio-Plex Manager software (Bio-Rad, Hercules, CA). The detection limit for each cytokine was established from the lowest consistent calculated data by the BioPlex Manager software.
Cell death analysis
After collecting the culture supernatant, the monocyte/macrophage cultures were incubated at 33°C for 15 min with 200 µl of 2 µg/ml Sytox Green Nucleic Acid Stain (Invitrogen) diluted in HBSS. Images of wells incubated with Sytox or control solution were captured with a ZOE fluorescent imager and analyzed with the Image J software for determining the percentage of dead cells that stained with Sytox.
Viral detection using RT-qPCR
SARS-CoV-2 RNA was detected using RT-qPCR from lung tissue or putatively infected suspensions (perfusion and nebulization liquids). Lung biopsies in RNAlater (100 mg) were placed in Trizol, homogenized with 1.4 mm ceramic beads in a Precellys 24 bead grinder homogenizer (Bertin Technologies), and purified using the NucleoSpin RNA kit that includes a DNAse digestion step (Macherey-Nagel, ref 740955.250). In the case of infected solutions, viral RNA was extracted from 100 µL using the NucleoSpin® RNA Virus kit from Macherey- Nagel kit (ref 740956.250). The RT-qPCR for the E gene detection was done on 500 ng RNA from lung tissue in 25 µl final reaction volume and on 2 µl eluate of viral RNA from solutions in 10 µl final reaction volume, using the Superscript™ III Platinum® One-Step qRT-PCR from In vitrogen (ref 11732-088) with forward primer E_Sarbeco-F1 ACAGGTACGTTAATAGTTAATAGCGT, reverse primer E_Sarbeco-R2 ATATTGCAGCAGTACGCACACA andfluorescent probe E_Sarbeco-P1 FAM-ACACTAGCCATCCTTACTGCGCTTCG-Tamra (Sigma-Aldrich, Merck) 39. In the case of viral RNA detection from lung tissue, a SARS-CoV-2 RNA calibration curve of a previously titrated WL virus preparation was made with 10-fold dilutions from 2000 to 0.2 PFU per reaction and PFU equivalent per 100 mg were calculated. Negative controls included control lung tissue RNA and H2O. In the case of viral RNA detection from suspension, a SARS-CoV-2 RNA calibration curve of a previously titrated WL virus preparation was made with 10-fold dilutions from 500 to 0.05 TCID50 equivalent per reaction and TCID50 equivalent per ml were calculated. Negative controls included elution from the RNA extraction performed on control solutions and H2O. The cycling involved the following steps: reversetranscription at 55 °C for 20 min, denaturation at 95 ◦C for 3 min, amplification 50 cycles at 95 °C for 15 s and 58 °C for 30 s. A Cq value of 40 was attributed in cases of absence of detection (NA from the machine).The reactions were carried out in a CFX ConnectTM light cycler (BIO-RAD).
IFN gene expression detection using RT-qPCR
Total mRNA from the SARS-CoV-2 exposed monocyte/macrophage subsets were extracted using the Arcturus (PicoPure™ RNA kit-ThermoFisher Scientific) and quantified by Qubit™ RNA high sensitivity kit (Invitrogen™, Fisher Scientific SAS, Illkirch, France). RNA (32 ng) was reverse-transcribed using random primers and the Multiscribe reverse transcriptase (Applied Biosystem, ThermoFisher Scientific). Quantitative real-time PCR was carried out on 1:4 of the RT reaction with 300 nM primers in a final reaction volume of 25 µl of 1 X SYBR Green PCR Master Mix (Applied Biosystem, ThermoFisher Scientific). The used the KiCqStartTM predesigned primer pairs H_IFNA1_1, H_IFNB1_1, IFNG, H_RPS18_1 from Sigma-Aldrich, Merck, Darmstadt, Germany. PCR cycling conditions were 95°C for 30 sec, linked to 40 cycles of 95°C for 5 sec and 60°C for 30 sec. Real-time qPCR data were collected by the Bio-Rad CFX Maestro system (Bio-Rad Laboratories Inc, Marne-la-Coquette, France) and expression of the different genes relatively to RPS18 and normalized to an internal calibrator (arbitrary units) were calculated by the 2-DDCt method.
scRNA-seq and preprocessing of sequencing data
Single-cell suspensions were generated from 14 samples of the EVLP part of the study Additional file 5. Filtration results. For each sample, 2 x 104 cells were loaded onto the 10x Chromium to produce sequencing libraries, which were processed according to methods provided by 10x Genomics (v3 Chemistry). Cell cDNA was sequenced using the Truseq Illumina Stranded protocol and the Illumina NextSeq 550 sequencing machine (> 3x108 reads/sample). The reads were aligned with Cell Ranger v3.1.0 on the human genome using the GRCh38 assembly and the GTF file downloaded from Ensembl release 101, and on the genome sequence of the virus used for each infection. The 14 samples’ sequencing results were pre-processed and normalized using Seurat v4.3.0. Cells expressing less than 1,000 genes were removed. Dead or lysed cells were excluded by removing cells with a percentage of mitochondrial genes above a threshold calculated using the Scater package (median percentage of mitochondrial genes across all individual cells + 3 median absolute deviations). Filtration of the data also included the removal of doublets using Scrublet (https://github.com/AllonKleinLab/scrublet 40, expected doublet rate set to 0.08), see Additional file 5. Filtration results for the results of the filtering procedure. The scRNA-seq raw data of donors 4 and 5 were included in a previous study of our group 32, however they were reprocessed together with the other data sets for the clustering strategies and definition of cell identities in the subsequent steps.
Clustering strategies
We followed a similar workflow as the one used in our previous work 32. Specifically, in order to correct for the donor and time effect, we integrated the 14 scRNA-seq samples (Additional file 5. Filtration results) using the FindIntegrationAnchors and IntegrateData functions in Seurat and we produced an “initial integrated UMAP” with 86,253 cells (Additional file 6. Cell identity determination). Parameters of the dimensionality reduction and graph-based clustering were adjusted (k.param = 12, resolution = 0.6) to obtain 22 clusters. Note that the anchor integration procedure was used only for integration of the 14 samples and assignment of cells to clusters. The whole dataset (86,253 cells) was subsetted in sub-objects per donor (5 sub-objects, donor 1 to 5). All downstream analyses used the expression values normalized for each donor separately, in order to avoid using the data transformed upon integration.
Definition of cell identities in the scRNA-seq data sets
The original (cell x gene) matrices were pre-processed and normalized separately for each donor as described above and cells were assigned the cluster number of the “initial integrated UMAP”. Each donor data set was then analyzed with Azimuth, an automated reference-based algorithm for single-cell annotation (https://azimuth.hubmapconsortium.org/, version 2.0.0), using the Human Lung Cell Atlas as a core consensus reference model which encompasses 584,884 human cells of the lung and nose. The finest level of annotation was used. Cells presenting Azimuth annotation scores below 0.6 were discarded. We proceeded to a grouping of “close cell subtypes” in the cases of the B cells (B cells, plasma cells), stromal cells (smooth muscle, adventitial, peri-bronchial and alveolar fibroblasts), alveolar macrophages (alveolar macrophages, alveolar macrophage CCL3+, alveolar macrophage MT-positive, alveolar macrophage proliferating), blood endothelial cells (arterial, aerocyte capillary, general capillary, venous systemic, venous pulmonary), and lymphatic endothelial cells (lymphatic, differentiating, lymphatic mature, lymphatic proliferating). Each cell was associated with both a cluster number and a cell identity. In order to generate robust downstream analyses related to SARS-CoV-2 exposure, the cell identities representing less than 10% of a cluster in donors 1, 2 and 3 (EVLP with virus) and cells belonging to an identity/cluster not shared between donors 1, 2 and 3, were discarded for the subsequent analyses (Additional file 6. Cell identity determination). The cells selected with this Azimuth annotation-based process (66,737 cells) were projected onto the “integrated UMAP-filtered”, illustrated in Fig. 3A. Cluster C20 is absent in this “integrated UMAP-filtered” compared to in the “initial integrated UMAP” because C20 was discarded due to the heterogeneous cell identities found in this cluster. The contribution of each donor to the “integrated UMAP-filtered” is shown in Additional file 7. Representation of donors. The Azimuth-based cell identities (17 in total) that were kept for downstream analyses are: AMs, MoMacs, cMos, ncMos, DC2, mast cells, stromal cells, CD4+ T cells, CD8+ T cells, NK cells, B cells, blood endothelial cells, AT1s, AT2s, Transitional Club/AT2s, Club (non-nasal), Ciliated (non-nasal). As several Azimuth-based identities were found in several clusters, we considered the final identities on the basis of Azimuth identity and cluster belonging, leading to 28 identities. The representation of the 28 identities in each donor and timing is provided Additional file 8. Identities per donor per timing. The top markers of the 28 analyzed cell identities found in the “integrated UMAP-filtered” were extracted separately for each donor, using the normalized expression values of each dataset before integration (top marker genes per cell identity versus the other identities (minimal log2FC≥0.25, Bonferroni adjusted p-value≤0.05)). The intersection of the top marker lists was then computed for each cell identity, to keep only common markers to all donors, ranked in a decreasing order using the lowest gene expression ratio among the donors (Additional file 9. Top expressed markers). An interactive viewer for visualizing the cells from the different donors, their cluster belonging, their identities and gene expression is available at https://applisweb.vim.inrae.fr/ICARE/.
Statistics
The cytokine data were analyzed with R and were Log10-transformed. A Shapiro test was used to evaluate the normality of the data distribution in each group and timing. When the data did not pass the normality test, a non-parametric paired Wilcoxon test was used to compare the data between 2 groups. Alternatively, a paired t-test was used upon equal variance evaluation. The statistics of the genomic data are reported in the dedicated paragraph. The Fisher’s exact test was used to calculate the probability of finding viral reads in AMs and/or MoMacs among the cells from infected lungs.