Patients and clinical data collection

We reviewed the medical records of 87 patients who were admitted to Shanghai General Hospital between 2016 and 2022. Based on their medical history, we stratified the enrolled patients into three groups: (a) patients with lung cancer and thrombosis, (b) patients with lung cancer only, and (c) patients with thrombosis but no lung cancer. In all cases, the diagnosis of lung cancer was confirmed through pathology. For cases of lung cancer combined with thrombosis, the diagnosis was only confirmed in individuals whose thrombosis was diagnosed after their lung cancer diagnosis. We retrospectively collected clinical characteristics and plasma biomarkers related to coagulation and tumors from all patients for analysis of similarities and differences. The variables were collected within one month of a lung cancer diagnosis or thrombosis diagnosis. This study was approved by the Ethics Review Board of Shanghai General Hospital (NO.2022KY011, Shanghai, China).

Bioinformatic analysis

We downloaded the single-cell transcriptome dataset, GSE131907, from the GEO database (https://www.ncbi.nlm.nih.gov/geo/), which included 58 sequences from 44 patients. Next, we selected lung adenocarcinoma (LUAD) tissues for further analysis. After quality control, a total of 42,570 cells were used for subsequent analysis. The data were analyzed with the "Seurat" package to perform Principal Component Analysis. Uniform Manifold Approximation and Projection was used to visualize cell types and clusters [19]. With the Find Clusters function in the Seurat package, we performed integrated clustering of expression values based on shared-nearest-neighbor graph clustering using the Louisville community detection-based method. We evaluated the robustness and further clustering analyses by using the clustering method at different resolutions with different representative markers. In total, we annotated 7 types of cells according to their representative genes, including MCs, T cells, myeloid cells, epithelial cells, B lymphocytes cells, fibroblasts and endothelial cells. The specific annotated marker genes can be found in Supplementary Table 1.

We downloaded the mRNA expression profiles and associated clinical information for 527 LUAD samples and 59 non-LUAD samples from The Cancer Genome Atlas database (TCGA, http://cancergenome.nih.gov/). After normalizing the RNA-seq data to transcripts per million and applying a log2 transform, we removed samples with incomplete clinical information for joint analysis of RNA-sequencing and clinical data.

We used the xCell algorithm to estimate the proportion of MCs in the samples from the TCGA data. Then, we sorted all LUAD samples by the proportion of MCs and determined the median value (0.00694578). Subsequently, we divided the LUAD samples into two groups based on their MC proportion: the high MC proportion group and the low MC proportion group. Next, we used CIBERSORT, EPIC, MCP-counter, quanTIseq, TIMER, and xCell algorithms to calculate the infiltration level of immune cells in the TME of LUAD [20,21,22,23,24,25]. We also used the Wilcoxon test to assess the association between the infiltration level of immune cells and MC proportion, and selected the immune cells with p < 0.0001 to visualize them as a heatmap. Additionally, we collected the lists of thrombosis-related signature genes, neutrophil marker genes, and NET signature genes from the reference literature (Supplementary Table 2) [26, 27]. We also obtained the list of 129 immunomodulators (Supplementary Table 3), including chemokines, interleukins, interferons, receptors, and other cytokines, from the study of Charoentong et al. [28]. Subsequently, we analyzed the relationship between the proportion of MCs and the expression levels of these genes and visualized them using the "ggplot2" R package.

Cell culture

Bone marrow-derived mast cells (BMMCs) were produced by culturing cells extruded from the femurs of 4–6 weeks C57BL/6 mice. Cells were cultured on Roswell Park Memorial Institute 1640 medium (Gibco, USA) containing 10% fetal bovine serum (FBS) (BI, Israel), 10 ng/ml recombinant interleukin-3 (rIL-3) (PeproTech, USA), 5 × 10-5 M β-mercaptoethanol (Sigma, USA), 2 mM L-glutamine (Gibco, USA), 10 mM HEPES (Beyotime, China), 10 × nonessential amino acids (Beyotime, China) and 1% penicillin–streptomycin (Beyotime, China). The medium was replaced per week and all cell cultures were grown at 37 °C in a humidified atmosphere with 5% CO2. BMMCs with 99% purity were harvested after four weeks of culture and identified by toluidine blue staining, FcεR1, and CD117 expression.

Human LUAD cell line A549 was purchased from Coweldgen Scientific Co., Ltd (Shanghai, China). The cells were cultured in Ham's F-12 K (Kaighn's) medium (BasalMedia, China) with 10% FBS (BI, Israel) and 1% penicillin–streptomycin (Beyotime, China) in a humidified 5% CO2 atmosphere at 37 °C.

Isolation of exosomes

Exosomes were isolated from A549 conditioned media. After the cells reached 90% confluence, the media was centrifuged at 300 g for 10 min, then at 10,000 g for 30 min at 4 °C, and filtered through a 0.22 μm membrane (Merck Millipore, Ireland) to remove cells, debris, and large vesicles. The exosomes were then isolated by ultracentrifugation at 120,000 g for 70 min at 4 °C, washed with phosphate buffer solution (PBS), and centrifuged again at 120,000 g for 70 min at 4 °C. The resulting pellets were resuspended in PBS and stored at -80 °C until use.

TME of exosomes

We prepared A549-derived exosome samples for transmission electron microscopy (TEM) analysis, as previously described [17]. Briefly, the exosome pellets were loaded onto thin bar carbon-coated copper grids with a 200 mesh and fixed with 2.5% glutaraldehyde. The samples were then washed, contrasted with 2% uranyl acetate, and embedded in a mixture of 0.4% uranyl acetate. Finally, examination was performed using an LEO 912AB Omega electron microscope (Carl Zeiss NTS, Oberkochen, Germany).

NTA of exosomes

The size distribution and concentration of exosomes diluted in PBS were analyzed by nanoparticle tracking analysis (NTA) using a ZetaVIEW® instrument (Particle Metrix, Germany). The equipment used a 405 nm excitation laser and was pre-calibrated for concentration using a 100 nm polystyrene latex reference standard (Applied Microspheres, Netherlands). The following parameters were used for the detection of exosomes (sensitivity: 85, shutter: 70, minimum brightness: 20, minimum size: 10, maximum size: 200) and the videos were taken at 30 frames per second. The data were analyzed using ZetaView software (Particle Metrix, Germany).

Confocal microscopy

Isolated exosomes were labeled with PKH67 (Sigma-Aldrich, USA) and then centrifuged again at 120,000 g for 70 min to remove excess dye. The labeled exosomes were co-incubated with BMMCs for 24 h. BMMCs were collected and washed twice with PBS, fixed with 4% formaldehyde for 15 min, and then washed twice with PBS. After labeling the nuclei with DAPI staining solution (Absin, China), the cells were photographed using spinning disc confocol microscopy and Leica SP8 confocal microscopy (Leica, Germany).

Degranulation assay of β-hexosaminidase release rate

The assay of BMMCs β-hexosaminidase release rate was performed using a previously described method [17], in which BMMCs (5 × 105 cells/ml, 0.5 ml) were prepared for culture in a 24-well plate with three replicate wells per group. The cells were washed twice and resuspended in 500 μl Tyrode's solution. Human SCF (50 ng/ml, PeproTech, USA), different concentrations of exosomes (50μg/well or 100μg/well), and 1μl DNP-IgE (1 mg/ml, Sigma, USA) were added to the different experimental groups for further analysis. The cells were incubated for 24 h, washed with Tyrode's solution, and the IgE group sensitized with DNP-HA (500 ng/well, Sigma, USA) for 45 min. All groups were incubated in an ice bath for 10 min to end the reaction. After incubation, the cell supernatant from each well was collected and centrifuged at 450 g for 5 min. Then, 50 μL of the supernatant was transferred to a separate well of a 96-well plate, followed by the addition of 50 μL of p-nitrophenyl-N-acetyl-beta-D-glucosaminide substrate (Aladdin, China). The plate was incubated at 37 °C for 60 min, and then 200 μL of stop buffer (400 mmol/L glycine, pH 10.4) was added to each well. The absorbance of each well was measured at 405 nm.

After discarding the supernatant, 100 μl of Triton X-100 (1%, Sigma, USA) was added to each well to fully lyse the cells at 37 °C for 30 min, after which the lysate was centrifuged at 10,000 × g for 30 min. Subsequently, the absorbance of the supernatant of the lysate was measured for each sample. The release rate (%) of β-hexosaminidase was calculated by dividing the absorbance of the supernatant by the absorbance of the cell lysate supernatant.

Cytokine array

We performed a cytokine antibody array using a mouse XL cytokine array kit (Cat. ARY028, R&D Systems, USA) following the instructions of the manufacturer. In brief, after treatment with 50 μg A549-derived exosomes or an equal volume of PBS for 24 h, the supernatant of BMMCs was collected, and the particulates were removed by centrifugation. Diluted supernatants (1:3) and the membranes in the kit were incubated on a rocking platform shaker at 4 °C overnight. After washing, the detection antibody cocktail was added to each membrane for one hour, followed by washing and addition of streptavidin–horseradish peroxidase for 30 min. After another round of washing and addition of Chemi Reagents Mix, immunoblot images were captured using a Tanon system, and the intensity of each spot was analyzed using Image J software.

Western blotting

Western blotting was performed as previously described. Briefly, exosomes and cell lysates were prepared using RIPA Lysis Buffer (Beyotime, China) supplemented with protease and phosphatase inhibitors (Bimake, USA). Proteins were separated by SDS-PAGE and transferred onto PVDF membranes. The membranes were blocked for 30 min in a blocking solution and incubated overnight at 4 °C with various primary antibodies (anti-TSG101, anti-Calnexin, anti-GAPDH, Abcam, USA; anti-CD81, Santa Cruz Biotech, Germany). After incubation with secondary antibodies for 1 h at room temperature, chemiluminescence detection was performed using the Omni-ECL™ Pico Light Chemiluminescence Kit (Epizyme Biotech, China).

Statistics

The statistical analysis was performed using SPSS 25.00, R (version 4.1.0), various R packages, and GraphPad Prism 8.0.2. The categorical variables were tested by the chi-square test. For continuous parameters, the t-test is used if the distribution is normal, and the Mann–Whitney U test if it is not normal. For all analyses, we defined statistical significance as p < 0.05.