Cell lines
Murine LLC cell line (#CRL-1642) was purchased from the American Type Culture Collection. B16-F10 cell line was a gift from J. Tao of the Department of Dermatology, Wuhan Union Hospital, Tongji Medical College, Huazhong University of Science and Technology. Subsequently, LLC cells were transfected with the pcDNA3-OVA plasmid (Genecreale) to obtain LLC-OVA cells. Mouse T cell hybridoma cells B3Z (#BFN608006444) were purchased from Bluefbio Biology Technology Development. Murine lung epithelial cells (MLE, #iCell-m036), murine colon adenocarcinoma cell line (MC38, #iCell-m032) and human bronchial epithelioid cells (BEAS-2B, #iCell-h023) were purchased from the Cellverse. Murine embryonic fibroblast cell line (NIH/3T3, #CL-0171) was purchased from the Pricella. These cell lines were cultured in DMEM (Gibco) supplemented with 10% FBS (Newzerum) at 37 °C in 5% CO2–95% air.
Mice
C57BL/6 mice were purchased from Hubei Beiente Biotechnology. They were kept in specific pathogen-free facility in the Wuhan Laboratory Animal Center of Tongji Medical College. All animal-related procedures have been performed according to the Declaration of Helsinki and gained permission from the Animal Care and Use Committee of Tongji Medical College ((2023) IACUC number 3958).
Preparation and isolation of MW-TMPs
We established an MW-assisted method to collect TMPs from tumour cells. After 24 h of incubation under standard conditions, the cell culture medium was replaced with a serum-free medium. The Petri dishes were subsequently placed into a domestic MW oven (setting condition 700 W, Midea) and subjected to MW heating for 20 s, a condition we designated as H20. To compare the products from different MW conditions, several settings were applied to LLC cells: L10 (175 W, 10 s), L20 (175 W, 20 s), L30 (175 W, 30 s), H10 (700 W, 10 s), H20 (700 W, 20 s) and H30 (700 W, 30 s). After 48 h of culture, the cell-conditioned medium from each MW treatment was collected for isolation of MW-TMPs. TMPs were extracted according to our previous study via a serial centrifugation strategy9. In brief, the medium was centrifuged at 200 × g for 10 min and 2,000 × g for 30 min at 4 °C to remove cells and debris, respectively. Second, pellet was obtained after 18,000 × g for 60 min at 4 °C and then resuspended in PBS. Finally, MW-TMPs were washed once and then stored using PBS. For drug-loaded MW-TMPs, the incubation method was used to encapsulate MW-TMPs with small molecules, such as MTX. After MW treatment, 200 μM MTX was added into the culture medium and the extraction of MW-TMPs-MTX followed the aforementioned procedures.
TMT quantification proteomics
UV-TMPs were prepared and isolated based on our former protocols10. Abundant proteins of UV-TMPs (named as U1, U2 and U3) and MW-TMPs (named as M1, M2, M3 and M4) as Mus_musculus samples were identified via single mass spectrometry run using the TMT quantitative method. Totally, 747,975 spectrums were generated, 19,595 peptides and 3,967 proteins were detected with 1% false discovery rate, and the further protein annotation with the help of multiple bioinformatic databases, including KEGG and Gene Ontology, was carried out. For differential protein analysis, an automated software, named IQuant, was applied and proteins with 1.5-fold change and P < 0.05 were regarded as differential expression12.
High-performance liquid chromatography of MW-TMPs-MTX
MW-stimulated LLC cells were incubated with 50 μM, 100 μM, 200 μM and 300 μM MTX for 24 h. MW-TMPs-MTX was collected and resuspended in ddH2O. Samples were sonicated for 10 min and passed through a 0.2 μm water film. As for the cumulative MTX release analysis by MW-TMPs, prepared MW-TMPs-MTX were incubated in the environment of pH = 7.4 and 37 °C and the stirring of 200 rpm at intervals of 4 h, 6 h, 12 h and 24 h. Then high-performance liquid chromatography was utilized to detect the drug content in MW-TMPs. Mobile phase (phase A, 0.025 mol l−1 potassium dihydrogen phosphate buffer (pH 5.5); phase B, acetonitrile) with the liquid samples (MTX or MW-TMPs-MTX) went through the chromatographic column (Athena-C18, temperature, 30 °C; flow rate, 1.0 ml min−1) under 303 nm wavelength in the model of isocratic elution. According to the chromatograms of standards and samples, the concentration of MTX loaded in MW-TMPs-MTX was analysed. Drug loading efficiency was evaluated with the ratio of MTX package quality and TMP quality39.
Detection of the ICD biomarkers
LLC cells were treated with either 20 μg ml−1 MW-TMPs, UV-TMPs or PBS, respectively. In addition, some LLC cells were pretreated with the HMGB1 inhibitor (glycyrrhizic acid, #HY-N0184) before MW-TMPs stimulation. The expression of canonical ICD markers, including HMGB1, ATP and calreticulin (CRT), was detected 24 h post-stimulation. HMGB1 expression levels were analysed by the WB assay (Supplementary Table 2). HMGB1 concentrations in the cell supernatant were measured using an ELISA kit (Bioswamp, catalogue number MU30043) following the manufacturer’s guidelines. CRT expression was evaluated by FCM and visualized through immunofluorescence (Supplementary Table 4). ATP content in the LLC supernatant was quantified using an ATP assay kit (Beyotime, catalogue number S0026).
MW-TMPs fluorescence labelling and uptake in vitro and in vivo
MW-TMPs were labelled with lipophilic membrane tracers DiI (Beyotime, #C1036), DiO (Beyotime, #C1038) or DiR (MedChemExpress, #HY-D1048) referring to the manufacturer’s instruments. After 10 min incubation in darkness at room temperature, MW-TMPs were washed using PBS twice to remove free dye solutions.
Fluorescence-labelled MW-TMPs were added to tumour cells. At various incubation intervals, cell medium was removed and PBS was used to wash cells twice. Cell nuclei were stained with DAPI and then the internalization of MW-TMPs was assessed by a fluorescence digital scanner (3DHISTECH, Pannoramic SCAN) or FCM (BD LSRFortessa X-20).
Female subcutaneous tumour-bearing C57BL/6 mice were intraperitoneally injected with DiR-labelled MW-TMPs. After post-administration of 1 h, 12 h and 24 h, the presence of MW-TMPs in tumour sites and freshly removed vital organs were visualized under bioluminescence imaging (Bruker MS FX Pro Imaging System). Excitation and emission wavelengths were set at 600 nm and 570 nm, respectively.
Immune cell extraction and induction
Murine femurs from C57BL/6 mice (6–8 weeks) were dissociated under sterile conditions, followed by immediate centrifugation at 12,000 rpm for 1 s to isolate bone marrow cells. The cells were treated in 1 ml red blood cell lysis buffer for 3 min and then neutralized with RPMI-1640 medium supplemented with 10% FBS. The cell suspension was then centrifuged at 1,800 rpm, 5 min at 4 °C. For induction of BMDCs, primary bone marrow cells were cultured in RPMI-1640 medium supplemented with 10% FBS, 20 ng ml−1 interleukin-4 (IL-4) (Peprotech, #214-12) and GM-CSF (Peprotech, #315-03) for 1 week. In addition, BMDMs were induced by culturing bone marrow cells in RPMI-1640 medium supplemented with 10% FBS and 20 ng ml−1 M-CSF (Peprotech, #315-02) for 1 week.
Lymphocytes were extracted from the spleens of 6–8-week-old C57BL/6 mice using the mouse CD3+ T cell isolation kit (Vazyme, catalogue number CS101) according to the manufacturer’s instructions. The isolated spleen lymphocytes were incubated in RPMI-1640 medium supplemented with 10% FBS (Gibco), 20 ng ml−1 IL-2 (Peprotech), HEPES solution (100×, Procell), β-mercaptoethanol (1,000×, Procell) and sodium pyruvate solution (100×, Procell).
To assess the immune-stimulating capacity of tumour cells treated with MW-TMPs, LLC tumour cells were treated with LLC-derived MW-TMPs for 24 h. These resulting LLC cell pellets were then co-cultured with BMDCs for another 24 h. Subsequently, these pretreated BMDCs were co-cultured with either B3Z cell line or isolated CD3+ T cell for 48 h, allowing for the evaluation of the proportion of mature DCs, CD4+ T, CD8+ T cells, as well as the secretion of IFNγ (#1210002, Dakewe) and IL-2 (#1210203, Dakewe).
Subcutaneous and orthotopic tumour-bearing model and in vivo treatment
LLC cells (5 × 105 cells suspended in 100 μl DMEM solution) were subcutaneously inoculated into the unilateral posterior flanks of C57BL/6 mice (6 weeks). After approximately 1 week of observation, they were randomized into several groups (n = 6). The tumour volume and body weight were recorded every other day until the end of the intervention. For lung orthotopic tumour models, 6 × 105 LLC-LUC cells were injected into the exposed lung tissue of mice. After 3 days, the orthotopic tumour models were randomly classified into three groups and give specific intervention. The growth of orthotopic tumour was monitored by using the IVIS Spectrum imaging system to detect the spontaneous luminescence formed after d-luciferase binds to the substrate. The treatment (100 μl PBS, 3 μg g−1 MPs, 5 μg g−1 anti-PD-L1 antibody) was administrated intraperitoneally once every 2 days. Mice were sacrificed after the fifth intervention. Subcutaneous or naked tumour volume was estimated as follows: volumeSubcutaneous tumour = length × width2 × 0.5; volumeNaked tumour = length × width × height × 0.5. To estimate the antitumour efficiency, the formula about tumour growth inhibition rate (TGI %) was used: TGI (%) = (1-mean volumetreatment/mean volumecontrol) × 100%. According to the ethics committee’s regulations, the maximal tumour burden permitted is 2,000 mm3. We confirm that the maximal tumour burden was not exceeded during the study.
scRNA sequencing
When the intervention was completed, fresh subcutaneous tumours were stored in magnetic-activated cell sorting tissue storage solution (#130100008) and cell suspensions were prepared by tissue homogenate. The single-cell suspension was added to the single-cell sorting honeycomb plate and excessive magnetic beads with barcodes were put into the plate. During the phase of cell lysis, the RNA was labelled as bounding to the barcode. Subsequently, reverse transcription of RNA and cDNA synthesis were carried out to obtain cDNA library for subsequent sequencing. Paired-end sequencing was conducted in the NovaSeq 6000 sequencing platform to get the transcriptome results. BD Rhapsody Analysis pipeline was applied to identify the UMI sequence and the cell tag sequence alignment to the reference genome (GRCm39). After removing low-quality cells, the normalization methods ‘LogNormalize’ and ‘ScaleData’ were used to standardize and scale each gene and then linear dimensional reduction was performed according to the PCA score. Uniform manifold approximation and projection (UMAP) techniques were chosen to visualize the cell clusters (the visualized PCA dimension = 38). The raw scRNA sequencing datasets have been deposited in the NCBI database (GSE289959).
Pseudotime trajectory analysis and calculation of signature score
According to differential expression genes across cells, monocle package (version 2.26.0) was used for analysing the trajectory processing of DCs, CD8+ T cells and TAMs to infer their dynamic differentiation and functions. To quantify the function of immune cells, including DCs, CD8+ T cells and NK cells, the ‘AllModuleScore’ function (Seurat version 4.4.0) was used to calculate the enrichment scores according to signature gene sets (Supplementary Table 8).
Cell-cell interaction and pathway enrichment analysis
We used the classic ‘CellChat’ package (version 1.6.1) to predict and analyse the strength of interactions between different cell types, mainly including tumour cells and various immune cells, as well as DC subpopulations and T cells, based on cell gene expression. After 1,000 pairing tests, ligand–receptor pairs with significant enrichment and P-value significance were extracted (Supplementary Table 7). On the basis of the differential expression genes among treated and untreated groups by performing ‘FindAllMarkers’, KEGG and hallmark pathway analysis were investigated to estimate the differential activities of cells comparing groups.
Distribution of tumour-infiltrating immunocytes
The detection of immune cells in tumour tissues using FCM was carried out as previously described10. In brief, single-cell suspension was collected through grind, digestion, filtration and lysis of red blood cells. Cells may need stimulation by PMA (0.1 μg ml−1, Solarbio, P6741-1 mg) and ionomycin (1 μg ml−1, Aladdin, 1139530-1 mg) before evaluation of concentrations of Gzmb, IFNγ and IL-4. Suspension was blocked by anti-mouse CD16/32 antibody. Next, multiple fluorescent antibodies (Supplementary Table 3) were stained in darkness for 30 min at 4 °C. Finally, FCM instrument (Beckman Coulter, DxFLEX) was used to analyse the infiltration of immune cells in the TME.
Tissue in situ immunofluorescence
Tissues of tumours, spleens and lymph nodes were acquired at the end of intervention and washed with PBS before they were fixed and embedded in paraffin. Then the sections underwent dewaxing, hydration, antigen retrieval and quenching of endogenous peroxidases. The tissues were blocked with 2% bovine serum albumin (BSA) and then incubated with fluorescence-coupled antibodies (Supplementary Table 4) overnight at 4 °C. Next, nuclei in tissues were stained with DAPI. Finally, the sections were scanned using a fluorescence microscope scanner.
Acquisition and treatment of human MPE samples
Human MPE samples were obtained from lung cancer patients undergoing thoracentesis to alleviate fluid accumulation at Wuhan Union Hospital, Tongji Medical College, Huazhong University of Science and Technology. The study protocol was approved by the ethical committee of Tongji Medical College, Huazhong University of Science and Technology and all MPE samples were collected with patients’ informed consent ((2019) IEC (S1041)).
Approximately 500 ml of fresh MPE samples per patient were subjected to centrifugation at 1,000 × g for 15 min, resulting in separation into supernatants and cell pellets. A portion of the cell pellets was immediately labelled with FCM antibodies to characterize the cellular composition within the pleural effusion. A list of the anti-human antibodies used is summarized in Supplementary Table 3. All antibodies for human FCM were used at a 1:100 dilution. Another portion of the cell pellets was seeded into agarose gel-coated 96-microwell plates for the subsequent formation of 3D multicellular spheroids. After 5 days, we used 3D multicellular spheroids that formed in the wells to represent a human solid TME for further immunomodulation experiments with various TMP types. The formed 3D multicellular spheroids were treated with various TMPs for 6 days, and spheroid sizes were recorded. At the end point of observation, we immobilized 3D multicellular spheroids with 4% paraformaldehyde and stained them with DAPI for the subsequent imaging. In addition, cytotoxicity assays were conducted using Live/Dead detection kits on the multicellular spheroids to evaluate their viability at the end of the experimental observation.
The remaining cells underwent treatment with red blood cell lysis buffer to remove erythrocytes, followed by washing twice with PBS. Tumour cells were then isolated by magnetic-activated cell sorting using magnetic beads conjugated with human Epcam microbeads (Miltenyi Biotec) following the manufacturer’s instructions. The sorted tumour cells were subjected to either UV or MW stimulation for the preparation of UV-TMPs and MW-TMPs. Simultaneously, a smaller fraction of the whole-cell pallets was seeded into different 12-well plates for analyses of cell uptake of MW-TMPs and immune activation detection.
To assess cellular uptake capacity of TMPs by various cell types within the MPE, Dil-labelled MPs were incubated with MPE pallets in the 12-well plates. After a 24 h incubation period, FCM was used to quantify the phagocytosis levels of Dil-MW-TMPs by individual cell types. Alternatively, some cell pellets were fixed for further immunofluorescence.
To further explore the impact of UV-TMPs and MW-TMPs on immune cells, they were introduced to additional 12-well plates. Upon 24 h of incubation, FCM was utilized to evaluate cell proportion and functional changes in the MPEs from each patient. Concurrently, the cell supernatant was collected to measure ATP concentrations using an ELISA kit. Another portion of the cells was lysed and denatured to facilitate the detection of CRT expression via WB.
Zebrafish PDX models
To establish the PDX models in zebrafish, lung cancer tumour tissues obtained from the patient during the surgical operation were processed into single-cell suspensions using mechanical and enzymatic digestion. These cells were subsequently labelled with Vybrant CM-DiI (Thermo Fisher Scientific). In parallel, peripheral blood samples from the patients were collected to isolate peripheral blood mononuclear cells (PBMCs), which were activated in vitro and stained with DiO fluorescent dye (Thermo Fisher Scientific). The labelled tumour cells and PBMCs were mixed in a 1:1 ratio and microinjected into the yolk sacs of 2 days post-fertilization (dpf) wild-type AB zebrafish from Hunter Biotech, at a density of 300 cells per fish. Then the zebrafish were all reared in fish-rearing water at 35 °C. Regarding the water quality, 200 mg of instant sea salt was added to each 1 l of reverse osmosis water, with the electrical conductivity ranging from 450 μS cm−1 to 550 μS cm−1; the pH value was between 6.5 and 8.5; and the hardness was 50–100 mg l−1 CaCO3. The license number for the utilization of laboratory animals is SYXK (Zhejiang) 2022-0004.
At 3 dpf, zebrafish showing consistent tumour cell expression were selected under the microscope and randomly divided into 3 groups for culture in 6-well plates with a 3 ml well volume, with 10 zebrafish per group. A549-MW-TMPs or A549-UV-TMPs were administered into the yolk sacs of zebrafish models, alongside a model control group. Two days after the intervention, fluorescence intensity was assessed by photographing 10 zebrafish per group under a fluorescence microscope. The images were processed using NIS-Elements D 3.20 advanced image-processing software to evaluate the antitumour growth efficacy of TMPs.
Statistical analysis
The experiments in the study were repeated independently at least three times. Statistical analysis was performed using the software GraphPad 8.0. All data were shown as mean values ± s.d. Significant comparison between two groups was determined through unpaired two-tailed Student’s t-test. For more than two groups, one-way analysis of variance (ANOVA) with further multiple comparisons using post-Turkey’s multiple comparison test was carried out. P < 0.05 indicated statistical significance.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.