Viruses
Various SARS-CoV-2 strains, hCoV-19/Japan/QHN002/2021 (Alpha), hCoV-19/Japan/TY8-612/2021 (Beta), hCoV-19/Japan/TY7-503/2021 (Gamma), hCoV-19/Japan/TY11-927/2021 (Delta), and 2019-nCoV/Japan/TY38-873/2021 (Omicron), isolated at the National Institute of Infectious Diseases (Japan), were used in this study. We primarily used the Delta variant for our experiments because it was the predominant circulating strain during our study and has been associated with severe vascular complications35. SARS-CoV-2 was propagated in VeroE6/TMPRSS2 cells. The virus stock was generated from the supernatant of VeroE6/TMPRSS2 cells infected with SARS-CoV-2 at a multiplicity of infection of 0.1 and harvested 2 days after infection. The viral titer was determined using a plaque assay.
Plaque formation assay
Vero/TMPRSS2 (JCRB1818) was obtained from JCRB Cell Bank in National Institutes of Biomedical Innovation, Health and Nutrition. VeroE6/TMPRSS2 cells were seeded into 24-well plates (80,000 cells/well) at 37 °C in 5% CO2 overnight. The supernatants were serially diluted using the inoculated medium and incubated for 2 h. Next, the culture medium was removed, fresh medium containing 1% methylcellulose (1.5 mL) was added, and the cells were cultured for an additional 3 days. Finally, the cells were fixed with 4% paraformaldehyde in PBS (Nacalai Tesque, Inc. Kyoto, Japan), and the plaques were visualized using Giemsa’s azur-eosin-methylene blue solution (#109204, Merck Millipore, Darmstadt, Germany).
Reagents
The 1-oleoyl-LPA was purchased from Avanti Polar Lipids (Alabaster, AL, USA). LPA stock solution (10 mM) was prepared using 50% ethanol and stored at − 20 °C before use. LPA working solution was prepared by diluting the 50% stock solution in PBS at a ratio of 18:1. LPA (10 mg/mL) was used for study. For control experiments, 5% ethanol was used as the vehicle.
Cell culture
Human Coronary Artery Endothelial Cell (HCAEC), HMVEC-Lung, and Human Pulmonary Artery Endothelial Cell (HPAEC) lines were purchased from Lonza (Basel, Switzerland) and grown in EBM-2 supplemented with EGM-2 MV BulletKit. Human Cardiac Microvascular Endothelial Cell (HCMEC), Human Hepatic Sinusoidal Endothelial Cell (HHSEC), HRGEC, and HMBEC lines were purchased from ScienCell (Berlin, Germany) and grown in an endothelial cell medium containing FBS and endothelial cell growth supplement. The Human Saphenous Vein Endothelial Cell (HSaVEC) line was purchased from PromoCell (Heidelberg, Germany) and grown in an endothelial cell growth medium-2. Human vascular cells were maintained under conditions of 5% CO2 and 37 °C, below 80% confluency, and used in experiments at passage 7 or 8.
Endothelial cell 3D culture model
All eight human endothelial cell lines described above were cultured in collagen gels with 3D embedding and evaluated for their ability to form 3D luminal structures. Detailed characterization was performed on the three selected cell types (HMVEC-Lung, HMBEC, and HRGEC) based on receptor expression profiles. Cellmatrix Type I-A (Nitta Gelatin, Osaka, Japan) was diluted to 0.15% with dilute hydrochloric acid (Fujifilm Wako Pure Chemical Corporation, Osaka, Japan) at pH 3.0 and reconstituted according to the manufacturer’s protocol. The reconstituted collagen gel was mixed with cultured vascular endothelial cells at a concentration of 5 × 105 cells/150 µl. The collagen cell mixture solution was cultured at 37 °C in a CO2 incubator for 30 min in a glass-bottomed dish (Matsunami D141400, Osaka, Japan). After the gel hardened, 10 ng/ml of VEGF-A (PeproTech, Rocky Hill, NJ, USA) containing the culture medium was added. Following incubation in a humidified incubator at 37 °C and 5% CO2 for 2 days, CD31 staining was performed to identify luminal structures. Vascular endothelial cells cultured under 3D conditions were pretreated with 10 µM of LPA for 30 min, infected with SARS-CoV-2 (Wuhan strain and Delta strain) at a multiplicity of infection of 0.1, and incubated with 10 µM of LPA. Two days after infection, the 3D tubes were fixed with paraformaldehyde and stained with Alexa Fluor 488 anti-human CD31 antibody (BioLegend, San Diego, CA, USA). Images were obtained using a TCS SP8 confocal microscope (Leica Microsystems, Wetzlar, Germany). The 3D tube volume was analyzed using Volocity imaging software (Perkin Elmer, Waltham, MA, USA). The weakest 10% of Alexa Fluor 488 fluorescence signals were excluded from the analysis as gel autofluorescence signal noise. Furthermore, objects smaller than 500 µm3 were excluded from the volumetric analysis as endothelial cells that were not 3D tube structures.
Animals
Four-week-old male Syrian hamsters were purchased from Japan SLC (Shizuoka, Japan). The animals were housed in environmentally controlled rooms at the animal experimentation facility of Osaka University under standard conditions (22 ± 2 °C, 12 h light/dark cycle, standard laboratory diet, and water ad libitum). Animals were maintained in individually ventilated cages (ISOrat900 N, Tecniplast, Buguggiate, Italy) with appropriate bedding, enrichment, and air filtration. All experiments were performed according to the guidelines of the Osaka University Committee for Animal and Recombinant DNA Experiments (approval number 4062).
Animal infection experiments
Syrian hamsters were randomly assigned to experimental groups (n = 4 animals per group: mock infection control, SARS-CoV-2 infection, SARS-CoV-2 infection + LPA 10 mg/kg, and SARS-CoV-2 infection + LPA 30 mg/kg) following established protocols from our previous work20,21. Syrian hamsters were anesthetized with isoflurane and challenged with 1.0 × 106 PFU (in 60 µL) SARS-CoV-2 via intranasal routes. Infection was confirmed by the development of consistent respiratory symptoms and characteristic pulmonary pathology in SARS-CoV-2-inoculated animals, which were absent in mock-infected controls. To evaluate the therapeutic effects of LPA on vascular injury caused by SARS-CoV-2 infection, LPA (10 or 30 mg/kg, intraperitoneally) was administered daily from days 2 to 4 after infection. The selection of doses (10 and 30 mg/kg) was based on our previous studies with LPA in other mouse models18 and the treatment window (days 2–4) was chosen to reflect a clinically relevant scenario where treatment would begin after symptom onset. Five days post-infection, the hamsters were euthanized by isoflurane overdose followed by cervical dislocation, and the lungs were harvested for the experiments. All animal experiments with SARS-CoV-2 were performed at the Animal Biosafety Level 3 (ABSL3) facilities of the Research Institute for Microbial Diseases at Osaka University. Animal Experimentation, and the study protocol was approved by the Institutional Committee of Laboratory Animal Experimentation of the Research Institute for Microbial Diseases, Osaka University (approval number R02-08-0). All methods were performed in accordance with the relevant guidelines and regulations. All animal experiments were conducted in strict accordance with the guidelines for the care and use of laboratory animals established by the Ministry of Education, Culture, Sports, Science and Technology of Japan, and followed the ARRIVE guidelines for reporting animal research.
RNA isolation and expression analysis
Vascular endothelial cells cultured in planar or 3D cultures were collected, and RNA was isolated using the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. Total RNA was extracted from the lung tissue using TRIzol reagent. RNA was transcribed into cDNA using the PrimeScript RT Reagent Kit (TaKaRa, Shiga, Japan) according to the manufacturer’s instructions. Quantitative real-time PCR was performed on a LightCycler 96 (Roche, Rotkreuz, Switzerland) using the TB Green™ Premix Ex Taq™ Kit (TaKaRa) and was run in triplicate. The results were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) levels using the comparative threshold cycle method. Specific primers used in this experiment were as follows: LPA4: 5’-CCTAGTCCTCAGTGGCGGTA-3’ (sense) and 5’-CTTCAAAGCAGGTGGTGGTT-3’ (anti-sense); GAPDH: 5’-GAAGGTGAAGGTCGGAGTC-3’ (sense) and 5’-GAAGATGGTGATGGGATTTC-3’ (anti-sense); ACE2: 5’-GGATTCCATGAAGCTGTTGG-3’ (sense) and 5’-TCGTGAGTGCTTGTTTGAGC-3’ (anti-sense); and TMPRSS2: 5’-TGGAGCCGGATACCAAGTAG-3’ (sense) and 5’-GTTGGGCAGACACACTGGTT-3’ (anti-sense). Three independent samples were collected from each group, and mRNA was extracted from endothelial cells using RNeasy Plus Micro Kits (Qiagen) according to the manufacturer’s protocol. RNA libraries were prepared using the TruSeq Sample Prep v2 kit and sequenced on a HiSeq 2500 (Illumina, San Diego, CA, USA) in the 75-base single-end mode. CASAVA 1.8.2 software (Illumina) was used for base calling. Sequenced reads were mapped to the mouse reference genome sequence (mm9) using TopHat v2.1.0. Fragments per kilobase of exons per million mapped fragments values were determined using Cuffnorm v.2.2.1.
Tissue staining
Lung tissue samples harvested from hamsters were fixed in 4% paraformaldehyde, and cryo-tissue sections were prepared. Hematoxylin and eosin staining was performed to evaluate immune cell infiltration and pathological changes in lung tissues, with 20 random high-power fields (400× total magnification) scored for each condition. Scores were assigned to each parameter, including neutrophil infiltration in the alveolar and interstitial spaces, hyaline membrane, proteinaceous debris filling the airspaces, and alveolar septal thickening36. Histopathological evaluation was performed by a pulmonologist who assessed the extent of inflammatory lesions and vascular damage in a blinded manner using a standardized scoring system. The blood vessels of the lung sections were stained with Lycopersicon esculentum Lectin, DyLight 488 conjugate (Vector Laboratories, Burlingame, CA, USA), and TO-PRO-3 stain (ThermoFisher, Waltham, MA, USA). Tissue preparation and staining were performed as previously described37.
DEG analysis
The RNA-seq data in Fig. 3 were analyzed using iDEP (ver2.01) (http://bioinformatics.sdstate.edu/idep/)38,39. DEGs were identified using DESeq40 with a maximum FDR p-value of < 0.05 and a fold change of at least 1.5 between groups. We performed k-means clustering40 on the 2,000 most variable genes using the Gene Ontology biological process41 and KEGG pathway databases42. We analyzed the RNA-seq data (Fig. 5) obtained under the three experimental conditions to identify genes with differential expression patterns. The conditions were as follows: (1) Samples collected 5 days after SARS-CoV-2 infection (SARS-CoV-2); (2) samples collected 5 days after SARS-CoV-2 infection with subsequent LPA stimulation on days 2, 3, and 4 post-infection (SARS-CoV-2 + LPA); and (3) samples collected 5 days after mock SARS-CoV-2 infection served as controls (mock). Four samples were analyzed for each condition. We then extracted genes exhibiting expression differences under the following conditions: (1) Genes with increased expression post-infection compared to non-infection (SARS-CoV-2) were selected if the log2 ratio of the average gene expression in SARS-CoV-2-infected samples to that in non-infected samples was ≥ log2(1.5), and the average expression in SARS-CoV-2-infected samples was at least 1.0. (2) Genes with decreased expression under LPA stimulation compared to SARS-CoV-2 infection (SARS-CoV-2 + LPA) were selected if the log2 ratio of the average gene expression in SARS-CoV-2 + LPA samples to that in SARS-CoV-2-infected samples was ≤ log2(0.75), and the average expression in SARS-CoV-2-infected samples was at least 1.0. Using these criteria, we identified specific gene sets that exhibited altered expression in response to SARS-CoV-2 infection and subsequent LPA stimulation.
Gene enrichment analysis
To explore the functional characteristics of the genes altered by SARS-CoV-2 infection, we performed gene enrichment analysis on two groups: those with increased expression following SARS-CoV-2 infection and those exhibiting decreased expression upon LPA stimulation. Fisher’s exact test was used to identify functionally enriched gene groups using Gene Ontology sets from MsigDB (c5.all.v7.2. symbols.gmt). This approach delineates the functional pathways and biological processes involved in these gene sets.
Statistical analysis
All data are presented as mean ± SD. No data were excluded from the analyses. For the in vivo studies, age-matched and weight-matched hamsters were randomized into experimental or control groups. Statistical analysis was performed using the statcel4 software package (OMS, Saitama, Japan) with analysis of variance for all data, followed by the Tukey–Kramer multiple comparisons test. Two-sided Student’s t-tests were used to compare two groups. P < 0.05 was considered to indicate statistical significance.
