The cancer tissues and adjacent noncancerous lung tissues of 50 NSCLC patients were acquired from the pathological specimen repository of The First Affiliated Hospital of Nanchang University, China. Time to metastasis was calculated based on the date of initial treatment to the date of investigator-assessed radiographic organ or node metastasis. The data were censored at the last follow-up or when patients died without metastasis.

The human bronchial epithelial cell lines (BEAS-2B) and human NSCLC cell lines A549 and SK-MES-1 were obtained from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). Both cell lines were grown in high-glucose Dulbecco’s modified Eagle’s medium (DMEM, BI, Israel) supplemented with 10% foetal bovine serum (FBS, Gibco, Grand Island, USA) and 1% penicillin and streptomycin solution (Solarbio, China) at 37°C in a 5% CO₂ humidified atmosphere.

The Ethics Committee of the Medical Innovation Center of the First Affiliated Hospital of Nanchang University approved the animal experimental protocol. Eight-week-old female SD rats, and four-week-old female BALB/c nude mice were purchased from Hunan SJA Laboratory Animal Co., Ltd. The mice were fed under specific pathogen-free (SPF) conditions in accordance with the regulations of the institution.

Neutrophils were isolated from the peripheral blood of healthy donors and SD rats with the peripheral blood neutrophil extraction kit (Solarbio, China). Isolated primary neutrophils were maintained in RPMI 1640 medium (BI, Israel) supplemented with 10% FBS. Giemsa staining and trypan blue viability assays were utilized to determine neutrophil purity (> 98%) and vitality (> 95%).

Neutrophils were plated and allowed to adhere in 6-well plates for 1 h before treatment with 100 nM phorbol-12-myristate-13-acetate (PMA, Sigma, USA) for 4 h at 37°C in 5% CO₂; this treatment allowed NETsosis to occur. Then, based on the protocol recommended by previous studies (36), NETs were harvested following a multistep centrifugation protocol. For quantification, equivalent numbers of neutrophils (1×10⁷/well) grown in 6-well plates were stimulated with PMA (100 nM) for 4 h to generate NETs. Then, the supernatants were slowly and gently extracted and washed twice to remove impurities without disrupting the NETs. The supernatants containing NETs were collected and centrifuged to purify the NETs, which were finally stored at −80°C for subsequent experiments.

Paraffin-embedded lung tissues from NSCLC patients were cut into 5-μm-thick sections for Immunofluorescence assays. The paraffin sections were deparaffinized and rehydrated. The sections were heated in Tris/EDTA buffer (pH 9.0) for antigen retrieval. Isolated neutrophils were seeded on coverslips in 24-well plates. Then, both cells and thin sections of lung tissues were fixed, permeabilized, blocked and stained with primary antibodies against citrullinated histone H3 (cit-H3) (1:250, Abcam, ab5103, UK), myeloperoxidase (MPO) (1:50, Abcam, ab90810, UK) or Ly6g (1:100, Abcam, ab25377, UK) overnight at 4 °C with shaking, followed by incubation with secondary antibodies conjugated to Alexa Fluor 488 (1:200, Elabscience, China) and Alexa Fluor 594 (1:200, Elabscience, China) for 1 h at 37°C. The nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI, Boster Biological Technology, China) for 5 min. Neutrophil-produced NETs were imaged by fluorescence microscopy (Zeiss) or confocal laser scanning microscopy (Leica) to assess their components. The colocalization of NETs with cit-H3, MPO, Ly6g or DNA was observed. The quantification of NETs was analyzed using Image J software.

Twelve hours after NETs stimulation, cells were treated with TRIzol (Invitrogen, Carlsbad, USA) and then sent to OE Biotech Co., Ltd. (Shanghai, China) for transcriptome RNA microarray analysis. A |log2 (fold change)| ≥ 2 and P < 0.01 were used as the thresholds to determine whether lncRNA expression was upregulated or downregulated. Volcano plots and heatmaps showing differential lncRNAs were generated with R soft and related Bioconductor packages. The top 20 upregulated signalling pathways were identified by Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses.

MIR503HG-overexpressing and empty vector A549 cells were stimulated with NETs the day before the mouse metastasis model was established. These two kinds of A549 cells (2 × 10⁶/mouse) were injected into BALB/c nude mice through the tail vein. After 8 weeks, the mice were euthanized, and the lungs were collected and subjected to H&E and histological staining. Metastasis burden, number of metastasized tumours, volume, and maximum size were evaluated.

Paraffin-embedded lung tissues from the above mouse model were cut into 5-μm-thick sections for immunohistochemical (IHC) analysis. Briefly, the paraffin sections were deparaffinized and rehydrated. The sections were heated in Tris/EDTA buffer (pH 9.0) for antigen retrieval and incubated in 3% hydrogen peroxide for 10 min at room temperature to block endogenous peroxidase activity. For IHC staining, the primary antibodies anti-N-cadherin (1:50, Abcam, ab76011, UK), anti-E-cadherin (1:200, Proteintech, 20874-1-AP, China), and anti-vimentin (1:200, Abcam, Ab92547, UK) were incubated with the tissue samples overnight at 4°C. Images of the IHC-stained slides were visualized and analysed at 100× and 400× magnification utilizing an ordinary optics microscope (Zeiss).

A549 and SK-MES-1 cells were seeded on coverslips in 24-well plates overnight. After preprocessing, the cells were fixed with 4% paraformaldehyde (Solarbio, Beijing, China) for 20 min and permeabilized with 0.5% Triton X-100 (Solarbio, China) for 15 min at room temperature. After blocking with 5% goat serum for 1 h at room temperature, the cells were incubated with primary antibodies against NLRP3 (1:50, Proteintech, 19771-1-AP, China), Caspase1 (1:50, Proteintech, 22915-1-AP, China), and p50 (1:200, Cell Signaling Technologies, 13586, USA) overnight at 4°C with shaking. The cells were incubated with Alexa Fluor 488-conjugated secondary antibodies (1:200, Elabescience, China) for 1 h at 37 °C. The nuclei were stained with DAPI (Boster Biological Technology, China) for 5 min. The samples were visualized by fluorescence microscopy (Zeiss) or confocal laser scanning microscopy (Olympus).

ROS in A549 and SK-MES-1 cells were evaluated by using the ROS assay kit (Elabscience, China). Briefly, NSCLC cells after preprocessing were incubated with 10 µM 2,7-dichlorodihydrofluorescein diacetate (DCFH-DA) protected from light at 37°C for 30 min, and then followed by washing with PBS to remove excess fluorescence probe. The active cell nucleus was stained using Hoechst 33342. Fluorescence microscopy (Zeiss) was applied to test the fluorescence intensities in NSCLC cells.

The human IL-1β and IL-18 ELISA kit (Boster Biological Technology, China) was performed to assess the level of IL-1β and IL-18 in NSCLC cells supernatant according to the manufacturer’s instructions. Brief, the culture medium was collected after the removal of cell debris by centrifugation at a speed of 1000×g for 10 min. The level of IL-1β and IL-18 in supernatants were analyzed by ELISA kit. Absorbance at 450 nm was detected with a microplate reader (ThermoScientific).

NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Fisher Scientific, 78833, USA) were used to determine the MIR503HG location in A549 and SK-MES-1 cells. According to the instructions of the kit, briefly, dry cells were collected in 1.5-mL microcentrifuge tubes (1-10×10⁶ cells/tube) and washed with PBS, and ice-cold mixture that contained Cytoplasmic Extraction Reagent (CER I), CER II and Nuclear Extraction Reagent (NER) (200:11:100 µL) was added. The tubes were vigorously vortexed at the highest setting and centrifuged at 16,000 ×g for 5 min; then, the supernatant was transferred to a new tube. The above steps were repeated once more, and the extracts were obtained. All the steps were performed on ice, and then, the samples were purified using the FastPure^Ⓡ Cell/Tissue RNA Isolation Kit V2 (Vazyme, Nanjing, China) following the manufacturer’s protocol to isolate the nuclear and cytoplasmic RNA. GAPDH (predominantly in the cytoplasm) and U6 (enriched in the nucleus) were used as controls.

A549 and SK-MES-1 cells were seeded in dishes at an appropriate density and cultured in a 37°C incubator overnight. The cells were transfected with plasmids and siRNAs with Hieff Trans liposomal transfection reagent (Yeasen, Shanghai, China) according to the manufacturer’s instructions. The cells were collected 24-48 h after transfection for Western blotting or qRT-PCR analysis.

The NLRP3 expression plasmid was purchased from Vigene Biosciences (Shandong, China). The p50-pcDNA3.1 plasmids were generated by inserting the full-length sequence of p50 cDNA, which was amplified by PCR, into the pcDNA3.1 vector, and then, the constructed p50-pcDNA3.1 plasmid was sent to Genewiz Biotechnology Co. Ltd. (Suzhou, China) for DNA sequencing. siRNAs targeting p50, MIR503HG and the negative controls were designed and synthesized by Gemar Pharmaceutical Technology Co. Ltd. (Shanghai, China). The sequences are as follows:

p50-siRNA-1: 5′-GCUAUAAUCCUGGACUCUUTT-3′;

p50-siRNA-2: 5′-GCAAUCAUCCACCUUCAUUTT-3′.

MIR503HG-siRNA-1: 5′-CCUCUCCCACCAUUUCUUUTT-3′;

MIR503HG-siRNA-2: 5′-GACAAGAACUAAAGUGGAATT-3′;

Total RNA was harvested using the FastPure^Ⓡ Cell/Tissue Total RNA Isolation Kit V2 (Vazyme, Nanjing, China) following the manufacturer’s protocol. Real-time quantitative polymerase chain reaction (RT-qPCR) analyses were conducted utilizing a SYBR Green Kit (TransGen Biotech, Beijing, China). The 20-μL qRT-PCR mixture included 10 μL of 2× PCR master mix, 1 μL of 10 μM primers, 7 μL of RNase-free water, and 2 μL of the reverse-transcription template. The qRT-PCR amplification conditions were as follows: 95°C for 10 min, followed by 40 cycles at 95°C for 5 s and 60°C for 30 s. Human GAPDH was utilized as an internal control. The sequences of the PCR primers are provided in Supplementary Table 1.

Equal amounts of proteins were used for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The proteins were transferred to a polyvinylidene fluoride membrane (PVDF, Merck Millipore, Darmstadt, Germany). The membranes were incubated with 5% buttermilk for 1 h at 25°C and then incubated with primary antibodies at 4°C overnight with gentle shaking. The primary antibodies included anti-N-cadherin (ab76011), anti-Vimentin (ab92547), anti-cit-H3 (ab5103), (1:1000, Abcam, UK), anti-NLRP3 (19771-1-AP), anti-E-cadherin (20874-1-AP), anti-Caspase1 (22915-1-AP), anti-GAPDH (60004-1-Ig) (1:1000, 1:5000, 1:1000, and 1:20000 Proteintech, China), anti-p50 (1:1000, Cell Signaling Technologies, 13586, USA), anti-p-p50 (sc-271908), anti-p65 (sc-8008), anti-p-p65 (sc-136548) (all at 1:500, Santa Cruz Biotechnology, USA), and anti-Histone H3 (1:1000, ImmunoWay Biotechnology, YT2163, USA). The next day, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies specific for rabbit or mouse primary antibodies (Proteintech, China) at a dilution of 1:2000 at 25°C for 1 h. The protein bands were detected utilizing ECL Detection Reagents (Proteintech, China).

For the cell invasion assay, the upper chamber of the Transwell 24-well plates (8 μm pores, Corning, NY, USA) was coated with 5% Matrigel (Corning, NY, USA). A549 and SK-MES-1 cells (6×10⁴) suspended in 200 µl of serum-free medium were plated in the upper chamber after treatment. The lower chambers were filled with 600 μL of culture medium supplemented with 10% FBS. The NSCLC cells were permitted to invade through the Matrigel for 48 h, and the cells on the bottom side of the membrane were fixed using 4% paraformaldehyde, the cells on the upper surface were removed with cotton swabs, and then, the invaded cells were with crystal violet. The numbers of invaded cells were counted in five random fields under an optical microscope with ImageJ software.

For wound healing assays, treated cells were incubated in 6-well culture plates and grown to approximately 90% confluence. Scratches were manually established in the cell monolayers with a P200 pipette tip, which guaranteed that all wound widths were consistent. PBS was used to remove cell debris, and then, the scratched cells were cultured in medium supplemented with 1% serum that eliminated the effects of cell proliferation on migration. The wounds in three random fields were photographed at 0 h and 36 h, and the wound widths were measured with ImageJ software.

The data are shown as the mean ± standard deviation (SD). Statistical analysis was performed using Student’s t test or one-way analysis of variance (ANOVA) to compare the means between two groups or multiple groups, respectively, followed by the nonparametric Wilcoxon rank sum test. The risk of NSCLC patient metastasis was determined with the Kaplan-Meier method. All the analyses were performed using GraphPad Prism 8.0.2 software (San Diego, CA, USA). P < 0.05 was considered to be statistically significant.

To investigate whether released NETs are differentially expressed between healthy donors (HD) and NSCLC, we first isolated neutrophils from the blood of HD or NSCLC and observed the cell nuclear morphology by Giemsa staining to estimate neutrophil purity. Under the microscope, we observed the bulk multilobular nucleus of neutrophils (Figure 1A). Trypan blue dye exclusion assays revealed that the neutrophil viability was > 97% (Figure 1B). We observed that neutrophils isolated from NSCLC patients spontaneously generated NETs, whereas neutrophils from HD produced fewer NETs by immunofluorescence. The capacity of NETs formation from NSCLC patients was higher than HD (Figure 1C). We further detected the formation of NETs in NSCLC clinical tissue specimens by immunofluorescence staining. We found that NETs were largely accumulated in the NSCLC tumor tissues, while fewer NETs were formed in normal lung tissues (Figure 1D). Activated neutrophils exhibited NETs formation under specific conditions. Freshly isolated neutrophils were evenly seeded in 6-well plates at 1×10⁷/well and stimulated with 100 nM PMA for 4 h. After 4 h of PMA stimulation, light cloud-like substances appeared at the bottom of the 6-well plates. To further confirm whether these structures were NETs, several well-recognized NETs markers, including MPO and cit-H3, were measured by immunofluorescent staining. These markers were more highly expressed after PMA stimulation, and the structures were almost destroyed by treatment with DNase I (Figure 1E). In addition, we observed the colocalization of these protein markers with extracellular spider mesh-like DNA around neutrophils.

According to a previous study, the effect of NETs on tumours is not limited to their ability to physically capture cells (37). To explore the potential impact of NETs on the metastasis behaviour of NSCLC, NSCLC cells (A549 and SK-MES-1) were cocultured with PMA-induced NETs, DNase I, or NETs and DNase I together. The invasion assay results showed that NETs significantly promoted the number of invaded NSCLC cells, and these effects were abrogated by DNase I treatment, as the DNA structures of the NETs were digested by DNase I (Figure 1F). Moreover, this phenomenon was also observed in wound healing assays; isolated NETs promoted the migration of A549 and SK-MES-1 cells (Figure 1G). To investigate the possibility that NETs might induce a switch from epithelial to mesenchymal properties, the expression of EMT marker proteins was evaluated. As shown in Figure 1H, NETs promoted the EMT process of NSCLC cells, which was characterized by decreased expression of epithelial markers, such as E-cadherin, and increased expression of mesenchymal markers, such as N-cadherin and vimentin. This effect was suppressed by the digestion of NETs with DNase I (Figure 1H). These data suggest that PMA-induced neutrophil NETs formation contributes to the metastatic potential of NSCLC cells.

To identify candidate lncRNAs that participate in the pro-metastasis effect of NETs, a transcriptome RNA microarray analysis of A549 cells treated with NETs was performed. We identified 99 differentially expressed lncRNAs (fold change ≥ 4 and P < 0.01), of which 59 were upregulated and 40 were downregulated in the NETs-treated group compared to the untreated group (Figure 2A). According to bioinformatics analysis, the expression of lncRNA MIR503HG was downregulated in the NETs-stimulated A549 cell group (Figure 2B). We investigated the expression of MIR503HG in NETs-stimulated NSCLC cells (A549 and SK-MES-1 cells) by qRT-PCR. Consistently, the expression level of MIR503HG was dramatically lower (P < 0.01) in both A549 and SK-MES-1 cells treated with NETs for 12 h than in the control cells (Figure 2C), which was consistent with the data generated from our microarray analysis.

To further examine the clinicopathological and prognostic implications of MIR503HG in patients with NSCLC, we next assessed tumour tissue data of NSCLC patients in The Cancer Genome Atlas (TCGA) and demonstrated decreased expression of MIR503HG in both lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) tissues compared to the corresponding adjacent normal lung tissues (Figure 2D). Then, we analysed the expression of MIR503HG in 50 pairs of NSCLC tissues and corresponding normal tissues from our cohort by qRT-PCR. Our results showed that the expression of MIR503HG was significantly downregulated (P < 0.01) in NSCLC tissues compared to the corresponding adjacent normal tissues (Figure 2E). The Kaplan–Meier analysis revealed that NSCLC patients with a lower level of MIR503HG had a higher (P = 0.045) risk of metastasis (Figure 2F). These data indicate that MIR503HG was expressed at low levels in NSCLC tissues or stimulated NETs and might act as a tumour suppressor.

To explore the underlying mechanism by which lncRNA MIR503HG functions in NSCLC, we subsequently explored the subcellular distribution of MIR503HG by subcellular fractionation and qRT-PCR analyses. The results of qRT-PCR, using GAPDH and U6 as controls, showed that MIR503HG was primarily located in the nucleus of A549 and SK-MES-1 cells (Figure 2G).

To fully understand the effect of MIR503HG on NETs metastasis promotion, we further detected the migration, invasion and EMT of NSCLC cells overexpressing MIR503HG. MIR503HG was overexpressed in A549 and SK-MES-1 cells via a stable retroviral expression system, and its overexpression was validated by qRT-PCR (Figure 3A). We also examined the expression of MIR503HG in normal human bronchial epithelial cell lines (BEAS-2B), NSCLC cells and cells overexpressed. We found that MIR503HG is generally overexpressed in BEAS-2B compared to NSCLC cell lines (Figure S1A). Wound healing assays showed that MIR503HG overexpression significantly impaired the NETs-induced migration abilities of NSCLC cells (Figure 3B). Similarly, Transwell assays revealed that MIR503HG overexpression inhibited the invasion abilities of cells treated with NETs (Figure 3C). Furthermore, the results showed that the expression of the EMT-related epithelial marker E-cadherin was increased but that of the mesenchymal markers N-cadherin and Vimentin was decreased in MIR503HG-overexpressing cells treated with NETs (Figure 3D). However, we did not observe significant changes for those markers in siRNA MIR503HG knockdown BEAS-2B cells (Figures S1B, C).

To explore the potential inhibitory effect of MIR503HG on NETs-induced NSCLC cell metastasis in vivo, MIR503HG-overexpressing NSCLC cells and the corresponding control NSCLC cells were treated with NETs one day before intravenous injection. Then, NETs-treated tumour cells were injected into mice via the tail vein, and the mice were sacrificed after 8 weeks (Figure 3E). The number of nodules on the surfaces of the lungs was counted (Figure 3F), and the presence of metastatic nodules inside the lung and the tumour size were also confirmed by H&E staining (Figure 3G). MIR503HG significantly reduced the NETs-induced lung metastasis of cancer cells. We quantified three features of tumours, namely, the number of nodules, tumour volume, and maximum size, to estimate the burden of lung metastasis. The results revealed that all these measures were significantly decreased compared with those in the control mice (Figure 3H). Furthermore, the results of IHC staining of tumour tissues suggested that the rate of E-cadherin positivity was increased and N-cadherin and vimentin expression was decreased in the groups with stable MIR503HG expression (Figure 3I). All these results suggest that MIR503HG could reverse the facilitation of NSCLC metastasis by NETs both in vitro and in vivo.

NETs formation could induce the release of large amounts of cytokines, and inflammasomes are considered key mediators of this process in tumour cells. Nevertheless, inflammasomes have also been shown to suppress the antitumour immune response and play a vital role in tumour cell migration and invasion (38). KEGG pathway analyses of the differentially expressed genes suggested that NETs triggered responses associated with inflammation, including the NOD-like receptor signalling pathway and NF-kappa B signalling pathway (Figure 4A). qRT-PCR was performed to validate the RNA microarray data (Figure 4B; Figure S2A). The expression of a set of NLRP3 inflammasome-associated genes, including NLRP3, Caspase1, IL-1β, and IL-18, was upregulated in a time-dependent manner. NETs stimulation for 12 h induced the most significant increase in the expression levels of NLRP3, Caspase1, and IL-18 compared to the other groups, while IL-1β reached its highest expression levels after NETs stimulation for 24 h (P < 0.05). We further validated the protein expression levels of NLRP3 and Caspase1 by Western blotting, and the results were consistent with the qRT-PCR results (Figure 4C). Immunofluorescence staining also supported the idea that NETs treatment could increase NLRP3 and Caspase1 expression in NSCLC cells (Figure S2B). Compared with the control group, the results of ELISA showed that the secretion level of IL-1β and IL-18 in the NSCLC cells supernatant of the NETs-treated group increased obviously (Figure S2C). Then, we used DCFH-DA to examine the effect of NETs on ROS levels in the intracellular. Intracellular ROS levels in both A549 and SK-MES-1 were significantly elevated after 12 h of NETs stimulation (Figure 4D). We next examined the role of NLRP3 inflammasomes by treating cells with the NLRP3 receptor-specific small molecule inhibitor MCC950 before NETs treatment. The results showed that MCC950 treatment before NETs treatment significantly decreased the mRNA expression levels of NLRP3 inflammasome-related genes (Figure 4E) and decreased the protein expression levels of NLRP3 and Caspase1 (Figure 4F). As expected, MCC950 reversed the protein expression levels of EMT-related markers after NETs treatment (Figure 4G). In the Transwell and wound healing assays, the number of invasive and migratory NETs-treated NSCLC cells was significantly decreased by treatment with MCC950 compared with the NSCLC cells treated with NETs alone (Figures 4H, I). Taken together, these results suggested that NLRP3 inflammasomes contribute to NETs-mediated metastasis promotion in NSCLC.

Next, we investigated whether MIR503HG inhibited the NETs-triggered metastasis of NSCLC by affecting the NLRP3 inflammasome. We first examined the protein and mRNA expression levels of NLRP3 inflammasome components in MIR503HG-overexpressing A549 and SK-MES-1 cells treated with NETs. Our Western blotting results revealed that the expression of NLRP3 and Caspase1 was significantly decreased (P < 0.05) when MIR503HG was overexpressed in NETs-stimulated NSCLC cells (Figure 5A). Examination of NLRP3 and Caspase1 mRNA expression by qRT-PCR also confirmed these results(Figure 5B). The mRNA expression of lncRNA MIR503HG was negatively associated with NLRP3 expression in NETs-treated NSCLC cells (Figure 5C). A549 and SK-MES-1 cells were transfected with NLRP3-pENTER or empty vector, and the levels of NLRP3 were validated by Western blotting (Figure S3A). Then, we transfected NLRP3-pENTER into NSCLC cells stably overexpressing MIR503HG and treated these cells with NETs. NLRP3-pENTER transfection abolished the downregulation of the N-cadherin and vimentin protein levels and the upregulation of the E-cadherin level in the two cell lines overexpressing MIR503HG (Figure 5D). Transwell and wound healing assays also indicated that the overexpression of NLRP3 restored the NSCLC cell invasion and migration abilities that had been inhibited by MIR503HG overexpression (Figures 5E, F). These results indicated that MIR503HG might indirectly or directly regulate the expression of NLRP3 inflammasome components, and the NLRP3 inflammasome mediated the effect of MIR503HG on the inhibition of NETs-triggered NSCLC metastasis.

Activation of the NLRP3 inflammasome is associated with the NF-κB signalling pathway. The NF-κB pathway is one of the most important signalling pathways, and this pathway upregulates NLRP3 and pro-IL-1 protein expression (39). In addition, KEGG analysis revealed that the NF-κB signalling pathway ranked in the top 20 pathways in our NETs-stimulated A459 microarray data (Figure 4A). To further investigate whether NF-κB was involved in promoting the NETs-induced expression of NLRP3 inflammasome components, we first examined whether the NF-κB signalling pathway was activated in A549 and SK-MES-1 cells after NETs treatment. Our results showed that NETs treatment increased the phosphorylation of NF-κB (p-p50 and p-p65), and this effect was entirely abolished by DNase I treatment (Figure 6A). Furthermore, confocal microscopy showed that the translocation of p50 into the nucleus was increased by NETs and inhibited by DNase I (Figure 6B). This result indicated that NETs activated the NF-κB signalling pathway in NSCLC cells. To further verify the role of NF-κB in the NETs-induced expression of NLRP3 inflammasome-related proteins, we designed p50-siRNA to knockdown p50 expression in A549 and SK-MES-1 cells and validated the levels of p50 by Western blotting and qRT-PCR (Figures S3B, C). P50-siRNA significantly suppressed the expression of NLRP3 and Caspase1 in NETs-stimulated NSCLC cells, as shown by immunofluorescence (Figure 6C and Figure S3D). Similarly, Western blotting analysis showed that the expression of NLRP3 inflammasome-related proteins in NETs-stimulated NSCLC cells was decreased when the cells were transfected with p50-siRNA compared to the negative control (Figure 6D). We further assessed the effect of NETs on the levels of EMT-related proteins and the metastatic ability of NSCLC cells after p50 expression was knocked down. Silencing p50 expression with p50-siRNA reversed the NETs-induced expression of EMT-related proteins (Figure 6E). Furthermore, the invasion and wound healing assay results showed that the number of invaded and migrated NSCLC cells after NETs treatment was decreased by the downregulation of p50 expression (Figures 6F, G). These results confirmed that NETs triggered the expression of NLRP3 inflammasome components and enhanced the metastatic ability of NSCLC cells via activation of the NF-κB signalling pathway.

The above results indicate that downregulation of MIR503HG expression could activate the NF-κB signalling pathway and induce the expression of NLRP3 inflammasome components, which in turn promotes tumour metastasis in NETs-treated NSCLC cells. We further explored the relationship between MIR503HG and the NF-κB/NLRP3 inflammasome via regulation of NF-κB expression. We treated MIR503HG-overexpressing A549 and SK-MES-1 cells with NETs and then measured the total and phosphorylated p50 and p65 protein levels by Western blotting. The results indicated that overexpression of MIR503HG decreased the NETs-induced phosphorylation of the NF-κB subunits p50 and p65 in the two NSCLC cell lines (Figure 7A). Then, these two NSCLC cell lines were transfected with p50-pcDNA3.1 or empty vector, and the levels of p50 were validated by qRT-PCR and Western blotting (Figures 7B, C). We further transfected p50-pcDNA3.1 into NSCLC cells stably overexpressing MIR503HG after NETs treatment. Upregulation of p50 expression abolished the downregulated expression of NLRP3 inflammasome components in NSCLC cells stably overexpressing MIR503HG and treated with NETs (Figure 7D). Transwell and wound healing assays also indicated that overexpression of p50 restored the NSCLC cell invasion and migration abilities that had been inhibited by MIR503HG overexpression (Figures 7E, F). These data revealed that MIR503HG suppresses the NETs-induced expression of NLRP3 inflammasome components and promotion of metastasis by inhibiting NF-κB activation.