Mice
Seven-week-old male C57BL/6 mice were purchased from Orient Bio Inc. (Gyeonggi-do, Korea). All animal procedures and protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of the Samsung Biomedical Research Institute (SBRI) at Samsung Medical Center (SMC).
ALI model
Mice were anesthetized with an intraperitoneal injection with 0.1 mL/kg zolazepam/tiletamine (Zoletil 50; Virbac, Carros, France) and 0.04 mL/kg Rompun (Elanco, Greenfield, IN, USA). Mice were then intratracheally injected with 100 µL phosphate-buffered saline (PBS) or 200 µg LPS (extracted from E. coli O55:B5, Sigma-Aldrich, St. Louis, MO, USA) per mouse using a 22-gauge catheter. Additionally, 0.4 mL air was injected to evenly spread the solution throughout the lungs.
Irradiation
The day after intratracheal injection, the mice were anesthetized and irradiated with 1 Gy X-rays using a VitalBeam linear accelerator (Varian Medical Systems, Palo Alto, CA, USA). To avoid radiation exposure of other vital organs, the thorax of each mouse was placed inside a radiation field. 6 MV X-ray beams were delivered at 4 Gy/min with a source-to-surface distance of 100 cm.
H&E stain
Mouse lung tissues were isolated and fixed in 4% paraformaldehyde (Wako, Richmond, VA, USA). The fixed lung tissues were embedded in paraffin, sectioned, stained with H&E, and scored by a pathologist for five major lung injuries: neutrophils in the alveolar space, neutrophils in the interstitial space, hyaline membranes, proteinaceous debris filling the airspaces, and alveolar septal thickening. For neutrophil aggregation, three or more neutrophil clusters throughout the lungs were manually counted.
Micro-CT
To evaluate lung inflammation, a Inveon micro-CT scanner designed as an in vivo system (Siemens Medical Solutions, Knoxville, TN, USA) was used. All mice were scanned using the following settings: 360º total rotation and 360 rotation steps, 1º step size, 80 kVp and 400 µA source setting, and 400 ms exposure time per step. The pixels were then binned by a factor of 4 to obtain an effective pixel size or approximately 39.77 µm resolution. For each scan, the dataset was reconstructed with a downsampling factor of two using the Inveon Acquisition Workplace software package (IAW, Siemens Medical Solutions), implementing a modified Feldkamp filtered back-projection algorithm (Shepp-Logan filter).
Flow cytometry
Mouse lungs and BALFs were harvested and treated with ammonium-chloride-potassium [33] solution (Quality Biological, Gaithersburg, MD, USA) for red blood cell lysis. The BALFs was centrifuged at 500× g for 5 min, the cells were counted using LUNA-Ⅱ™ automated cell counter (Logos Biosystems, Gyeonggi-do, South Korea). T cells were stained with FITC anti-CD3 (145-2C11, BD Biosciences, San Diego, CA, USA), APC-Cy7 anti-CD4 (GK1.5), and BV421 anti-CD8 (53 − 6.7) antibodies. Neutrophils were stained with APC-Cy7 anti-CD45 (30-F11), PerCP-Cy5.5 anti-CD11b (M1/70), and FITC anti-Ly6G (1A8) antibodies. AMs were stained with APC-Cy7 anti-CD45, PE anti-CD64 (X54-5/7.1), BV421 anti-SiglecF (E50-2440), PE-Cy7 anti-CD11c (HL3), and PerCP-Cy5.5 anti-CD11b antibodies. The M1 and M2 were stained with PerCP-Cy5.5 anti-CD11b, APC anti-F4/80 (T45-2342), and PE-Cy7 anti-CD206 (MR6F3, eBioscience, San Diego, CA, USA) antibodies (all antibodies were diluted 1:100 and stained at 4°C for 30 min). To evaluate the population of apoptotic cells, the cells were stained with APC Annexin V (550475, BD Biosciences) and propidium iodide (PI, 51-66211E) at room temperature (21–23 ℃) for 15 min in dark. All cells were analyzed using a BD FACSLyric flow cytometer.
ELISA
For cytokine analysis, BALFs were collected from mouse lungs 48 h after LPS or PBS injection (24 h after LDRT). The BALFs was centrifuged at 500× g for 5 min, the supernatant was collected. The supernatant was collected and protein concentration was measured using a Bradford protein assay (Bio-Rad, Hercules, CA, USA). Cytokine levels were measured using mouse IL-6, TNF-α, IFN-γ, IL-10, and TGF-β ELISA kits (Invitrogen, Carlsbad, CA, USA), according to the manufacturer’s standard protocol.
Single-cell preparation of mouse lung tissue
Mouse lung tissues were isolated and digested using a gentleMACS™ Dissociator and MACS Lung Dissociation Kit (Miltenyi Biotec, Gladbach, Germany). To separate mononuclear cells, the dissociated lung cells were slowly added to Ficoll (Ficoll: cells = 3:4 mL). The solution was centrifuged at 400× g for 30 min (20 ℃, decel = 0), the intermediate zone (layer between Ficoll and plasma) was collected and DMEM (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) was added. The solution was centrifuged at 400× g for 5 min (25 ℃, decel = 9), the supernatant was removed, and the cells were resuspended in PBS.
Single-cell RNA sequencing
Mouse cell samples were prepared for scRNA-seq using the Chromium Single-Cell 3′ Reagent version 3.1 kit (PN 1000121) and Chromium Single-Cell Controller (10× Genomics, Pleasanton, CA, USA). Subsequently, the cells were loaded onto a Single-Cell G Chip (PN 1000127) for gel bead-in-emulsion generation and barcoding, targeting a recovery of 10,000 cells per sample. Reverse transcription, clean-up, and cDNA amplification were performed to isolate and amplify cDNA or a ligated sequencing adaptor for Illumina library construction, according to the manufacturer’s protocol. The libraries were constructed using the Chromium Single-Cell 3′ reagent kit v3.1 (PN 1000121), 3′ Library Construction Kit (PN-1000157) according to the manufacturer’s protocol. For sequencing, the library concentration was assessed using a Qubit 2.0 Fluorometer and dsDNA HS Assay Kit (Thermo Fisher Scientific) according to the manufacturer’s protocol. Single-Cell 3′ gene expression libraries were run on an Illumina NovaSeq 6000 system with 2×100 bp read length.
Cell filtering and quality control
Sequencing data were processed using CellRanger version 7 and aligned to the mouse reference genome (mm10). R version 4.2.2 and Seurat Version 4.3.0 were used to process CellRanger output files. Cells with less than 250 or more than 2500 expressed genes were removed, and doublets were removed using DoubletFinder.
Cell clustering and batch correction
Seurat was used to merge each sample and perform the preprocessing. The “NormalizeData” function was used to normalize and the “FindVariableFeatures” function with “vst” method was ued to find 2,000 highly variable gene (HVG) of each sample. The “ScaleData” function with “split.by = orig.ident” parameter was used to standardize the data. Before running the principal component analysis (PCA), consensus HVGs with high variability in at least two samples were identified. PCA was performed using consensus HVGs with default parameters. Harmony version 0.1.1, with default parameters, was used to remove the sample-level batch effect, and RunUMAP was used to visualize the data.
Cell type identification
Major cell types were identified by setting a resolution of 0.1 and confirmed with canonical marker genes with sensitivity > 0.5 in PanglaoDB. To identify the sub-clusters of each cell type, reanalysis (cell clustering and batch correction) was performed as described above. Cell-type-specific clusters were visualized using specific markers determined using the FindMarkers and FindAllMarkers functions of Seurat.
scRNA data analysis
To calculate the gene set enrichment scores, gene sets were downloaded from MsigDB (http://gsea-msigdb.org/gsea/msigdb/). The gsva function with “ssgsea” method from the GSVA R library was used to calculate single sample gsea scores. FeaturePlot, DotPlot, VlnPlot, and DoHeatmap from Seurat, and geom_boxplot from ggplot2 version 3.4.2 were used for visualization.
Statistical analysis
Statistical analyses were performed using the GraphPad Prism 8.0 (GraphPad Software, La Jolla, CA, USA). Comparisons between groups were performed using Student’s t-tests. Statistical significance of the scRNA-seq data was tested using the Wilcoxon rank-sum test. Statistically significant differences are shown as follows: ns > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.