Parental A549 and PC9 human LUAD cells were purchased from Wuhan Servicebio Technology Co., Ltd. (Wuhan, China), whereas cisplatin-resistant A549/DDP cells were purchased from Wuhan Pricella Biotechnology Co., Ltd. (Wuhan, China). All cell lines were cryopreserved and stored in the Central Laboratory of Xuzhou Central Hospital. According to the suppliers’ product information, these cell lines were authenticated by short tandem repeat profiling and were negative for mycoplasma contamination. A549 and A549/DDP cells were cultured in high-glucose DMEM supplemented with 10% premium fetal bovine serum and 1% penicillin-streptomycin, whereas PC9 cells were cultured in RPMI-1640 medium containing the same supplements. All cells were maintained at 37 °C with 5% CO₂. Cells between passages 5 and 15 after thawing were used for all experiments. According to the supplier, the cisplatin-resistant A549/DDP subline was established from parental A549 cells through stepwise selection with increasing concentrations of cisplatin. Following thawing, the cells were allowed to recover in cisplatin-free medium and were subsequently exposed to progressively increasing concentrations of cisplatin (0.4, 0.8, 1.2, 1.6, and 2.0 μg/mL) to restore selective pressure and maintain the resistant phenotype. Once stable growth in the presence of 2.0 μg/mL cisplatin was achieved, the cells were routinely maintained in medium containing the same concentration of cisplatin. Before the experiments, A549/DDP cells were cultured in cisplatin-free medium for 48 h to minimize the potential acute effects of the maintenance drug.

Cells (5 × 10³/well) were seeded into 96-well plates. Under light-protected conditions, Cell Counting Kit-8 (CCK-8) reagent (10 μL; Dojindo, CK04) was added to each well, and the cells were then incubated for 1 h at 37 °C. Optical density (OD) was measured with a multifunctional microplate reader at 450 nm. The half-maximal inhibitory concentration (IC₅₀) was calculated from the OD values, and the resistance index (RI) was calculated as follows: RI = IC₅₀ of drug-resistant cells/IC₅₀ of parental cells.

A cell mitochondria isolation kit (Beyotime, C3601) was used for mitochondrial isolation. Briefly, 1 mL of mitochondrial isolation reagent was added to 2 × 10⁷ cells, followed by incubation on ice for 15 min. Subsequently, the cell suspension was transferred to a glass homogenizer and manually homogenized about 30 times. Purified mitochondria were collected through density gradient centrifugation. RNA-easy isolation reagent (Vazyme, R701) was used to extract RNA from the mitochondrial pellet through phase separation, precipitation, and washing steps. The extracted RNA was thoroughly mixed with DEPC-treated water before storage at -80 °C.

RNA was reverse transcribed into cDNA using SweScript All-in-One RT SuperMix for qPCR (ServiceBio, G3337). Using cDNA as the template, qPCR was performed with 2× Universal Blue SYBR Green qPCR Master Mix (ServiceBio, G3326). The reaction conditions were as follows: 95 °C for 30 s, 95 °C for 15 s, and 60 °C for 30 s, repeated for 40 cycles. The primer sequences were designed and synthesized by Wuhan Servicebio Technology Co., Ltd., and the specific sequences are listed in Table 1.

Probes targeting LINC01133 were synthesized by Wuhan Servicebio Technology Co., Ltd., and its target sequence is listed in Table 2.

After fixation with 4% paraformaldehyde and permeabilization with 0.1% Triton X-100, the cell slides were pre-hybridized for 1 h at 37 °C and hybridized with a 500 nM LINC01133 probe overnight at 37 °C in the dark. The slides were then sequentially washed, hybridized, and subjected to signal amplification with 500 nM branch probes (40 °C, 45 min) and 1:200 diluted 488-labeled signal probes (42 °C, 3 h). Then, 3% BSA was added to block the slides for 30 min, followed by overnight incubation with MitoTracker Orange CMTMRos (1.5 μM; Thermo Fisher, M7510) at 4 °C. Finally, DAPI was added for nuclear counterstaining. After the slides were mounted, images were acquired using a fluorescence microscope.

Lentiviral vectors for LINC01133 overexpression and RNA interference (RNAi) were prepared by Shanghai GeneChem Co., Ltd. The LINC01133-targeting siRNA sequences are listed in Table 3.

Cells (1 × 10⁵/well) were seeded into 6-well plates and transfected the following day. After 48 h, the cells were selected using 2 μg/mL puromycin to establish stable LINC01133-overexpressing A549 cells (A549-OE-LINC01133) and stable LINC01133-knockdown A549/DDP cells (A549/DDP-sh-LINC01133). In addition, negative control groups infected with empty vector virus (A549-Vector, A549/DDP-NC) and uninfected blank control groups (A549-Control, A549/DDP-Control) were established. The same LINC01133-overexpressing lentiviral vector and corresponding empty vector were used to establish stable PC9-OE-LINC01133 and PC9-Vector cells, respectively, followed by puromycin selection.

Complete medium (600 μL) containing the drug was added to the lower chamber of the 24-well plate, while the cell suspension (100 μL, 1 × 106 cells/mL) was added to the upper Transwell chamber. In the invasion assay, after dilution, Matrigel (Becton, 354230) was precoated onto the bottom of the upper chamber, followed by gelation at 37 °C for 3 h and hydration with serum-free medium for 30 min. After incubation, fixation, staining, and washing, the nonpenetrated cells in the upper chamber were gently removed using a cotton swab, and the samples were observed and recorded under a microscope.

After the cell samples were rinsed with phosphate buffered saline (PBS), an Annexin V-FITC/PI apoptosis detection kit (KeyGEN, K231205) was used for staining. The cell suspension was then filtered through a 200-mesh sieve before analysis using a flow cytometer.

Following fixation with 2.5% glutaraldehyde and 1% osmic acid, the cells were dehydrated with a gradient ethanol series, incubated with a mixture of 100% acetone and epoxy resin embedding medium (1:1 ratio) at ambient temperature for 2 h, transferred to pure embedding medium, and polymerized overnight at 70 °C. After ultrathin sections were prepared, they were double-stained with uranyl acetate and lead citrate. Finally, they were visualized using transmission electron microscopy (TEM).

Cells from 6-well plates and shredded tissue samples (homogenized with a glass homogenizer) were lysed with RIPA Lysis Buffer (Beyotime, P0013B) on ice. After centrifugation and collection of the supernatant, total cellular and tissue proteins were obtained. For mitochondrial protein extraction, the isolated mitochondrial pellet was lysed on ice for 20 min with 150 μL of mitochondrial lysis buffer (including 1 mM PMSF), followed by centrifugation and collection of the supernatant to obtain mitochondrial proteins.

Protein samples were analyzed using a BCA protein assay kit (Beyotime, P0012). Protein aliquots (30 μg) were separated by 12% SDS-PAGE (Vazyme, E304) and transferred to PVDF membranes using wet transfer. After blocking with 5% defatted milk in TBST for 2 h at ambient temperature, the membranes were incubated with diluted primary antibodies overnight at 4 °C. The primary antibodies used were as follows: anti-Caspase 1 (Proteintech, 81482-1, 1:500), anti-Cleaved Caspase 1 (Affinity, AF4005, 1:500), anti-GSDMD (Proteintech, 80918-1, 1:1,000), anti-GSDMD-NT (Abcam, ab215203, 1:1,000), anti-TOMM20 (Servicebio, GB111481, 1:1,000), anti-COX IV (Servicebio, GB11250, 1:1,000), and anti-GAPDH (Servicebio, GB15004, 1:3,000). The following day, the membranes were incubated with an HRP-labeled goat anti-rabbit IgG secondary antibody (Servicebio, GB23303, 1:5,000) for 1 h. Protein bands were visualized using an enhanced chemiluminescence (ECL) kit (NCM, P10100) on a Bio-Rad ChemiDoc imaging system. The gray values of the protein bands were analyzed using ImageJ software. The relative expression level of the target protein was expressed as the ratio of the gray value of the target protein band to that of the internal reference protein band. For each biological replicate, protein samples were independently prepared from a separate cell culture experiment, and n = 3 represents three independent biological replicates.

Human IL-18 (MULTI SCIENCES, EK118/2) and IL-1β (Thermo Fisher, 88–7261) uncoated enzyme-linked immunosorbent assay (ELISA) kits were used to measure IL-18 and IL-1β levels in the cell supernatants. Briefly, cytokine standards were serially diluted to generate standard curves for IL-18 and IL-1β. Absorbance was measured at 450 nm using a microplate reader. The absolute concentrations of IL-18 and IL-1β in each sample were calculated based on the corresponding standard curves and expressed as ng/mL or pg/mL.

Using a lactate dehydrogenase (LDH) Cytotoxicity Assay Kit (Beyotime, C0016), the LDH concentration was determined according to the specified protocols. A microplate reader was used to measure the absorbance at 490 nm.

Cells were seeded in black 96-well plates with clear bottoms. To assess mitochondrial and intracellular Ca²⁺ levels, cells were incubated with working solutions of Rhod-2 AM (Beyotime, S1062M) and Fluo-4 AM (Beyotime, S1060) in the dark. After probe loading and hydrolysis, fluorescence intensities were measured using a multifunctional microplate reader (mitochondrial Ca²⁺: Ex/Em = 549/578 nm; intracellular Ca²⁺: Ex/Em = 488/515 nm). For visualization, the cells were incubated with a combined staining working solution containing Rhod-2 AM (4 μM), MitoTracker Green (1 μM), and Hoechst 33342 (1 μM) at 37 °C in the dark for 20 min. Changes in mitochondrial Ca²⁺ were observed using a confocal microscope after the slides were mounted.

A mitochondrial membrane potential assay kit with JC-1 (Solarbio, M8650) was used to measure changes in the mitochondrial membrane potential. The fluorescence of JC-1 monomers and aggregates was measured at Ex/Em = 490/530 and 525/590 nm, respectively, using a microplate reader.

To assess mitochondrial ROS levels, 100 μL of MitoSOX™ Red mitochondrial superoxide indicator (Thermo Fisher, M36007) was added to the wells and incubated at 37 °C for 30 min in the dark. After the cells were rinsed twice with PBS, fluorescence intensities were measured using a multifunctional microplate reader (Ex/Em = 396/610 nm).

Mitochondrial adenosine triphospate (ATP) content was determined after mitochondrial lysis and centrifugation using an Enhanced ATP Assay Kit (Beyotime, S0027). ATP content was measured using a luminometer on a fluorescence microplate reader.

Total cellular DNA was extracted using a Universal Genomic DNA Purification Mini Spin Kit (Beyotime, D0063) in accordance with the specified protocols. PCR was performed to determine the relative mtDNA copy number, which was calculated as the ratio of the mitochondrial gene MT-CO1 to the nuclear gene B2M. The specific sequences are listed in Table 4.

All animal experiments were approved by the Laboratory Animal Management and Use Committee of the Experimental Animal Center of Wuhan Servicebio Technology Co., Ltd. (Approval No. Saiwei’er Animal [Fu] No. 2025207). The 5-week-old female BALB/c nude mice used in the present study were maintained in a specific pathogen-free (SPF) environment. Following acclimatization, mice were randomly assigned to six groups using a random-number method (n = 6). Investigators responsible for drug administration were aware of the group allocation. To minimize assessment bias, tumor measurements, tumor weighing, and quantitative histological analyses were performed under coded, blinded conditions by investigators who were unaware of the group assignments. Group identities were disclosed only after completion of data analysis. Cells in the logarithmic growth phase were harvested and resuspended at a density of 1 × 10⁷ cells/mL. Each mouse was subcutaneously inoculated in the right dorsal flank with 0.1 mL of the cell suspension. One week after inoculation, following confirmation of tumor formation, the mice received intraperitoneal injections of cisplatin (DDP; 5 mg/kg), disulfiram (DSF; 75 mg/kg), or lipopolysaccharide (LPS; 50 μg/kg), as appropriate for the assigned group, once every 3 days for 4 consecutive weeks. During the treatment period, tumor length (a) and width (b) were measured every 3-4 days using digital calipers, and tumor volume was calculated according to the formula V = a × b²/2. Body weight was recorded concurrently using an electronic balance. Following the final treatment, the mice were euthanized by cervical dislocation, and the tumors were excised and weighed. A portion of each tumor was immediately snap-frozen in liquid nitrogen and stored at -80 °C for quantitative reverse transcription-polymerase chain reaction (qRT-PCR) analysis. The remaining tissues were fixed, embedded in paraffin, sectioned, and subjected to immunohistochemical (IHC) and terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) staining.

Experimental results were obtained from three or more independent replicates and are represented as the mean ± standard deviation. Normally distributed data with equal variance were compared between groups using Student’s t test and among groups using one-way analysis of variance (ANOVA) followed by the LSD post hoc test. Non-parametric data were analyzed using the Mann-Whitney U test. The log-rank test was applied to analyze survival, while multifactorial designs were examined using multifactorial ANOVA. All statistical analyses and visualizations were performed using GraphPad Prism 9.5 software, with P < 0.05 indicating statistical significance.

The results of the CCK-8 assay revealed that relative to A549 cells, A549/DDP cells had markedly increased viability when treated with the same concentrations of cisplatin, verifying that these cells had a stable drug-resistant phenotype [Figure 1A and B]. Based on our group’s established method for mitochondrial RNA isolation and sequencing, combined with the results of a qRT-PCR assay, the LINC01133 mRNA level markedly increased within the mitochondria of A549/DDP cells compared with that in A549 cells [Figure 1C-F]. Subcellular localization using fluorescence in situ hybridization (FISH) revealed that the fluorescence signal for LINC01133 was markedly enhanced in A549/DDP cells and was accompanied by a high degree of colocalization with mitochondrial markers [Figure 1G]. Taken together, these results indicate that LINC01133 is upregulated in cisplatin-resistant A549/DDP cells and exhibits a mitochondria-enriched distribution pattern, suggesting that it may participate in mitochondria-related mechanisms underlying cisplatin resistance in LUAD.

To determine the clinical significance of LINC01133 in LUAD, the transcriptomic and clinical information of the TCGA-LUAD cohort was obtained from the UCSC Xena database. To minimize the potential influence of non-tumor biological factors, tissues with survival times of less than 90 days were excluded, and 473 LUAD tumor tissue samples were ultimately included for subsequent analysis. LINC01133 expression differed significantly between LUAD and normal lung tissues, with substantial expression heterogeneity and a distinct high-expression subset observed in LUAD tissues [Figure 2A]. Furthermore, Kaplan-Meier survival analysis revealed that LUAD patients with high LINC01133 expression, defined according to the optimal cutoff value of LINC01133 expression, exhibited significantly decreased survival rates compared with those with low LINC01133 expression [Figure 2B]. Consequently, LINC01133 expression is closely associated with unfavorable prognostic outcomes in LUAD patients, suggesting its potential as a biomarker for prognosis prediction and as a therapeutic intervention target.

To investigate the impact of LINC01133 on LUAD cisplatin resistance, this study successfully constructed stable LINC01133-overexpressing A549 cell lines and stable LINC01133-knockdown A549/DDP cell lines. After confirming altered expression by qRT-PCR, sh-LINC01133-3, which was the most effectively silenced gene, was selected for subsequent experiments [Figure 3A]. According to the results of the CCK-8 assays, LINC01133 overexpression significantly increased cisplatin resistance in A549 cells, whereas LINC01133 knockdown partially reversed the cisplatin-resistant phenotype and significantly increased cisplatin sensitivity in A549/DDP cells [Figure 3B]. Further cellular functional assays confirmed that overexpression of LINC01133 enhanced A549 cell proliferation, invasion, and migration, whereas LINC01133 knockdown effectively suppressed these malignant phenotypes of A549/DDP cells [Figure 3C and D]. Based on the above findings, LINC01133 not only regulates the malignant phenotype of LUAD cells but is also a key factor in maintaining their resistance to cisplatin.

To elucidate the cytotoxic mechanism of cisplatin and the regulatory role of LINC01133 in this process, we systematically analyzed cell death phenotypes and related signaling pathways. TEM revealed that following cisplatin treatment at the IC₅₀ for 48 h, both A549 and A549/DDP cells displayed typical features of pyroptosis, including cell swelling, organelle damage, and pyroptotic body formation on the cell membrane surface [Figure 4A]. Annexin V-FITC/PI double staining showed a significant increase in the late apoptotic/necrotic cell proportion after cisplatin treatment, indicating increased late-stage cell death and loss of plasma membrane integrity rather than specifically demonstrating pyroptosis [Figure 4B]. At the molecular level, Western blotting showed that cisplatin effectively activated the Caspase-1/GSDMD axis, as evidenced by the significant upregulation of GSDMD-NT and cleaved Caspase-1 [Figure 4C]. Further experiments confirmed that overexpression of LINC01133 inhibited cisplatin-induced Caspase-1/GSDMD pathway activation in A549 cells, reducing LDH release and secreted IL-1β and IL-18 levels. Conversely, knockdown of LINC01133 restored its activity in A549/DDP cells and upregulated these inflammatory markers [Figure 4D-F]. These findings suggest that the anti-tumor effect of cisplatin occurs partly through the activation of the Caspase-1/GSDMD-mediated pyroptosis pathway, whereas LINC01133 attenuates cisplatin cytotoxicity by inhibiting this pathway, thereby promoting drug resistance.

To exclude potential A549 cell-specific effects, PC9 LUAD cells with a distinct genetic background were selected to validate the regulatory effects of LINC01133 on cisplatin sensitivity and the pyroptosis pathway. Cisplatin dose–response analysis showed that PC9 cells exhibited the lowest IC₅₀, followed by A549 cells, whereas A549/DDP cells displayed the highest IC₅₀. These findings indicate that, under the present experimental conditions, PC9 cells can serve as a relatively cisplatin-sensitive LUAD cell model [Figure 5A]. qRT-PCR analysis showed that the basal expression level of LINC01133 was markedly higher in A549/DDP cells than in A549 and PC9 cells, whereas no significant difference was observed between A549 and PC9 cells [Figure 5B]. PC9-Vector and PC9-OE-LINC01133 cells were subsequently established by lentiviral transduction, and successful LINC01133 overexpression was confirmed by qRT-PCR [Figure 5C]. CCK-8 assays showed that LINC01133 overexpression significantly increased the cisplatin IC₅₀ of PC9 cells, indicating reduced cisplatin sensitivity [Figure 5D]. Western blotting demonstrated that cisplatin treatment markedly increased the expression levels of cleaved Caspase-1 and GSDMD-NT in PC9 cells, indicating activation of the Caspase-1/GSDMD pathway [Figure 5E]. In contrast, LINC01133 overexpression reduced the levels of these proteins, suggesting suppression of this pathway [Figure 5F]. A cross-model summary of the independently obtained IC₅₀ values showed that LINC01133 overexpression increased the cisplatin IC₅₀ in both A549 and PC9 cells, whereas LINC01133 knockdown decreased the IC₅₀ in A549/DDP cells [Figure 5G], indicating a consistent direction of regulation of cisplatin sensitivity. Western blotting indicated that cisplatin generally promoted activation of the Caspase-1/GSDMD pathway, although the magnitude of the response differed among the three cell models [Figure 5H]. Analysis of the normalized data showed that cisplatin treatment consistently increased the levels of cleaved Caspase-1 and GSDMD-NT. LINC01133 overexpression reduced the levels of both proteins in A549 and PC9 cells, whereas LINC01133 knockdown exerted the opposite effect in A549/DDP cells [Figure 5I]. Collectively, these findings demonstrate that LINC01133 exerts consistent regulatory effects on cisplatin sensitivity and the Caspase-1/GSDMD pathway across different LUAD cellular backgrounds.

To investigate the downstream mechanism through which LINC01133 regulates pyroptosis and drug resistance, we examined changes in mitochondrial structure and function. Staining with Fluo-4 AM and Rhod-2 AM fluorescent probes revealed that LINC01133 knockdown markedly increased the calcium ion concentration in the cytoplasm and mitochondrial matrix, indicating “calcium overload” [Figure 6A and B]. Fluorescence colocalization imaging revealed that in cells with evident pyroptosis, the mitochondrial calcium signal was enhanced and highly colocalized with the mitochondrial marker [Figure 6C]. TEM revealed that calcium overload was accompanied by severe damage to mitochondrial ultrastructure, including mitochondrial swelling, blurred boundaries, disrupted cristae, and vacuolization [Figure 6D]. Western blotting revealed that both inner and outer mitochondrial membrane markers (COX IV and TOMM20, respectively) were downregulated in the calcium overload group but upregulated in the LINC01133 overexpression group [Figure 6E]. Additionally, functional assays confirmed that interference with LINC01133 expression lowered mitochondrial membrane potential, increased mitochondrial ROS levels, reduced ATP content, and increased mtDNA copy number, whereas LINC01133 overexpression reversed these dysfunctions [Figure 6F-I]. These results suggest that LINC01133, by inhibiting the pyroptosis pathway, alleviates GSDMD-NT-mediated calcium influx^[22] and maintains mitochondrial calcium homeostasis, thereby protecting mitochondrial structural integrity and normal function and ultimately promoting cisplatin resistance in LUAD.

To elucidate the key effect of GSDMD on LUAD cell resistance to cisplatin, rescue experiments were conducted in A549/DDP cells. First, concentrations were determined by assessing cell viability after treatment with the GSDMD pathway activator LPS and inhibitor DSF, which were found to be 300 ng/mL and 10 μM, respectively [Figure 7A]. CCK-8 assays revealed that both silencing LINC01133 and LPS treatment improved cisplatin sensitivity in A549/DDP cells, as indicated by a decreased RI. Conversely, DSF intervention had the opposite effect, partially restoring the resistant phenotype. The results of the combination treatment experiments revealed that DSF partially attenuated the cisplatin-sensitizing effect induced by LINC01133 silencing, while LPS further strengthened this effect [Figure 7B]. The combined effects of DDP with LPS or DSF were quantitatively analyzed using the Chou-Talalay model. The results showed that the combination of DDP and LPS produced an additive effect, with a tendency toward synergism at higher effect levels, whereas the combination of DDP and DSF exhibited a stable antagonistic effect [Figure 7C]. These findings further support, at a quantitative level, the conclusion that activation of the GSDMD pathway enhances cisplatin cytotoxicity, whereas its inhibition reduces cisplatin sensitivity. At the molecular level, the Western blot results were consistent with the observed phenotypic changes: silencing LINC01133 and LPS treatment promoted Caspase-1 activation and enhanced GSDMD cleavage, as evidenced by increased cleaved Caspase-1 and GSDMD-NT levels, whereas DSF intervention inhibited the activation and cleavage of these pyroptosis-related proteins [Figure 7D]. ELISA results further supported the above regulatory pattern [Figure 7E]. Further analysis revealed that silencing LINC01133 and activating GSDMD increased the levels of calcium ions in the cytoplasm and mitochondria (calcium overload), whereas DSF treatment significantly reversed these effects [Figure 7F]. In summary, LINC01133 may regulate calcium homeostasis by modulating GSDMD-mediated pyroptosis, thereby contributing to the maintenance of cisplatin resistance in LUAD cells.

To validate the above mechanism in vivo, a BALB/c nude mouse xenograft tumor model was constructed in six experimental groups: Lv-A549-Vector, Lv-A549-OE-LINC01133, Lv-A549-OE-LINC01133 + LPS, Lv-A549/DDP-NC, Lv-A549/DDP-sh-LINC01133, and Lv-A549/DDP-sh-LINC01133 + DSF. The tumor tissues were excised immediately after the experiment [Figure 8A and B]. qRT-PCR results showed that LINC01133 expression in tumor tissues in each group was as expected, confirming successful construction of the animal model [Figure 8C]. Compared with those in the Lv-Vector group, tumor growth was significantly accelerated, and the final tumor weight was increased in the Lv-OE group; the combined use of LPS restored the tumor-suppressive effect of cisplatin. Conversely, relative to those in the Lv-NC group, tumor growth was inhibited, and the final tumor weight was reduced in the Lv-sh group; moreover, DSF administration attenuated the tumor-suppressive effect of cisplatin [Figure 8D and E]. The immunohistochemistry results indicated that, compared with the Lv-vector group, the Lv-OE group exhibited increased expression of Ki-67 and N-cadherin, whereas E-cadherin expression was decreased. In contrast, compared with the Lv-NC group, the Lv-sh group showed the opposite expression pattern for these markers [Figure 8F]. LINC01133 overexpression reduced TUNEL positivity and the expression of these pyroptosis-related proteins, and these effects were reversed by LPS; conversely, LINC01133 knockdown produced the opposite changes, which were attenuated by DSF [Figure 8G and H]. In summary, the results of the present study confirmed in an animal model that LINC01133 promotes malignant phenotypes including LUAD cell proliferation, invasion, and migration. The inhibition of the Caspase-1/GSDMD pathway-induced pyroptosis weakens the tumor-suppressive effect of cisplatin, thereby promoting cisplatin resistance in LUAD.