S100A10 promotes tumorigenesis and metastasis in lung adenocarcinoma by regulating JUND/TNC axis-mediated EMT Authors: Zhixin Zheng, Anjie Peng, Ziyao Xi, Haibin Ou, Jiong Liu, Aoqing Ye, Jing Shao, Feng Li, Yu Liu Published: 2026 ABSTRACT Highlights•S100A10 is highly expressed in lung adenocarcinoma and correlates with poor prognosis.•S100A10 drives epithelial-mesenchymal transition by activating the JUND/TNC axis.•The small-molecule inhibitor [D-Leu-4]-OB3 is identified as a targeted ligand of S100A10.•[D-Leu-4]-OB3 effectively suppresses lung cancer growth and metastasis in vitro and in vivo. Introduction Non-small cell lung cancer (NSCLC) is the most common type of lung cancer and has a personalised treatment system guided by staging and molecular characteristics [1,2]. Early-stage patients primarily undergo surgical resection, while those with locally advanced cases rely on a combination of radiotherapy, chemotherapy and immunotherapy [3]. Furthermore, many patients have limited treatment options due to the absence of clear therapeutic targets or inadequate responses to immunotherapy [[4], [5], [6], [7]]. Consequently, there is an urgent need to investigate the key molecular mechanisms underlying the onset and progression of NSCLC, and to identify novel therapeutic targets to overcome the current treatment challenges in NSCLC. NSCLC is primarily categorised as either LUAD or lung squamous cell carcinoma (LUSC) [[8], [9], [10]]. These two types of cancer exhibit distinct molecular profiles and pathological characteristics [[11], [12], [13]]. The identification of driver mutations in LUAD, such as EGFR, ALK and ROS1, has in recent years revolutionised treatment paradigms through the use of targeted therapies [[14], [15], [16], [17]]. Meanwhile, immune checkpoint inhibitors have shown efficacy in a subset of patients [18,19]. Nevertheless, the therapeutic landscape for LUAD is limited by various challenges, including primary and acquired resistance to targeted agents, the ineffectiveness of immunotherapy in certain patient groups, and the absence of actionable targets in a significant proportion of cases [20]. Furthermore, tumour heterogeneity and the evolution of resistance mechanisms continue to hinder durable responses [21]. Therefore, despite significant progress, there is an urgent need to identify new molecular drivers and develop more effective treatment strategies to enhance long-term outcomes in LUAD [22]. S100A10 is a key member of the S100 calcium-binding protein superfamily, which has diverse and tissue-specific biological functions [23]. Under physiological conditions, it typically binds to adhesion protein A2 (ANXA2, also known as p36), forming a heterotetrameric complex (AIIt) which plays a central regulatory role in several key physiological processes [24,25]. Furthermore, S100A10 regulates various physiological activities, including exocytosis, endocytosis, intercellular adhesion, ion channel conductance and membrane-cytoskeleton interactions [26]. S100A10 plays a key regulatory role in the instability of plaques and the progression of disease during acute coronary syndrome, modulating the function of vascular endothelial cells and the release of inflammatory cytokines [27]. In Parkinson's disease, S100A10 contributes to pathological damage processes by influencing α-synuclein aggregation and dopaminergic neuron apoptosis, which correlates with disease severity [28]. In recent years, the role of S100A10 in oncology has received increasing attention, and it has been identified as a pro-cancer factor in various malignant tumours [23]. In breast cancer, the S100A10-ANXA2 heterotetramer maintains tumour stemness traits and is associated with chemoresistance development [26,28]. In colorectal cancer, S100A10 is highly expressed in tumour tissues, making it a potential biomarker for diagnosis and prognosis while promoting tumour progression by regulating the invasion and infiltration capabilities of tumour cells [29]. However, the role of S100A10 in lung cancer, particularly LUAD, has not been reported. EMT is a core biological process that regulates the malignant progression of LUAD [30,31]. It plays a pivotal role in tumour invasion, metastasis, therapeutic resistance and recurrence [32,33]. EMT is activated through multiple signaling pathways such as TGF-β/Smad and Wnt/β-catenin, and causes LUAD cells to lose their epithelial characteristics and acquire mesenchymal traits [30]. Cancer cells enhance their migration and invasion capabilities, thereby promoting local invasion and metastatic spread [34,35]. EMT also confers significant resistance to various anticancer therapies, including chemotherapy and radiotherapy, by enhancing cell survival, promoting drug efflux, activating DNA repair mechanisms and inducing a quiescent, slow-cycling state associated with cancer stem cell characteristics [33,36]. These mechanisms constitute a major cause of treatment failure and recurrence [30]. Therefore, elucidating the molecular mechanisms by which EMT regulates the malignant behaviour of LUAD cells and targeting EMT phenotype represent promising strategies for improving patient outcomes [9,37]. Despite increasing recognition of S100A10 in cancer research, its specific contribution and regulatory mechanisms within LUAD remain inadequately explored. While existing studies suggest associations or roles in other tumor types, systematic investigations into the direct effects of S100A10 on distinct biological processes relevant to LUAD are lacking. Moreover, although S100A10 shows promise as a therapeutic target based on its oncogenic functions and expression patterns across different tumors, effective small-molecule inhibitors have yet to be fully developed. Existing literature provides limited information regarding targeted inhibition of this protein in the context of LUAD. Virtual screening has emerged as a pivotal computational approach in modern drug discovery, enabling the efficient identification of potential lead compounds from vast molecular libraries [38]. By leveraging molecular docking and scoring functions, this method predicts the binding affinity and interaction modes between small molecules and target proteins, thereby prioritizing candidates for experimental validation [39]. Given its cost-effectiveness and high throughput, virtual screening is particularly valuable for targeting proteins with known structures but lacking commercially available inhibitors [40]. In this study, we employed structure-based virtual screening to identify a novel small-molecule inhibitor targeting S100A10, with the aim of exploring its therapeutic potential in LUAD. Our study aimed to systematically investigate the role of S100A10 in LUAD, a major histological subtype of NSCLC. We sought to clarify its regulatory effects on cell proliferation, migration, invasion and EMT. Additionally, we aimed to elucidate downstream molecular mechanisms involved and to screen for potential small-molecule inhibitors targeting S100A10. Our work may provide novel biomarkers for prognostic evaluation of LUAD, identify new therapeutic targets and candidate drugs for targeted treatment, and advance the development of effective LUAD therapeutics. Materials and methods Cell culture Bronchial epithelial cells and human LUAD cell lines A549(RRID:CVCL_A549) and PC9(RRID:CVCL_E4DE) were purchased from Zhifan Biological Co., Ltd. The cells were cultured in high-glucose DMEM medium supplemented with 10% high-quality unfrozen fetal bovine serum and 1% penicillin-streptomycin. Cell counts were performed using Countess™ cell counting plates to ensure uniform seeding density. The cells were cultured in a standard humidified incubator at 37 °C in air containing 5% CO₂. All human cell lines have been authenticated using STR profiling within the last three years. All experiments were performed with mycoplasma-free cells. Reagents The following commercial kits were used in this study according to the manufacturers' instructions: Plasmid Mini-Prep Kit (DP103), Endotoxin-Free Plasmid Maxi-Prep Kit (DP118), and Tissue/Cell Total RNA Extraction Kit (GDP661-E) were from TianGen (Beijing, China). The Nuclear and Cytoplasmic Protein Extraction Kit (P0028), BCA Protein Assay Kit (P0010), and SDS-PAGE Gel Preparation Kit (P00125) were purchased from Biyuntian (Shanghai, China). Kits from HYCEZMBIO (Wuhan, China) included the Dual-Luciferase Reporter Assay Kit (HYHGF001) and Cell Counting Kit-8 (HYCCK8). Additionally, the DAB Chromogen Kit (C1001) was obtained from Bioqiandu (Wuhan, China), and the Hematoxylin-Eosin (HE) Staining Kit (G1120) was from Servicebio (Wuhan, China). Lentiviral transduction and plasmid construction Specific primers (Table S1) were cloned into the lentiviral vector pLVX-IRES-GFP to construct the S100A10 overexpression plasmid (pLVX-S100A10-GFP). Two independent short hairpin RNAs (shRNAs) targeting S100A10 were cloned into the pLKO.1-puro vector (Table S1) and a non-targeting control shRNA (Table S1) were cloned into the pLKO.1-puro vector. Cells were plated and cultured overnight to reach 30% confluency for lentiviral infection. The cells and viruses were incubated for 8 h, the medium was replaced, and the incubation was continued for 24–48 h. Then, the lentivirus-infected cells were screened with 2 μg/mL puromycin. The cells were seeded into a 6-well plate at 60% confluence for transient expression of S100A10, and were transfected with the S100A10 plasmid using Lipofectamine 2000 according to the manufacturer’s protocol and collected or treated 24–48 h post-transfection. Western blot Total cellular and tissue proteins were performed using RIPA lysis buffer supplemented with protease and phosphatase inhibitors, following the manufacturer’s protocol. Protein concentrations were quantified using a BCA protein assay kit with bovine serum albumin as standard. Equal amounts of protein (30 μg per lane) were separated by electrophoresis on 4–20% MOPS precast gels at constant voltage (150 V). Subsequently, proteins were transferred to 0.45 μm PVDF membranes using transfer system. Membranes were blocked with 5% non-fat dry milk in for 1 h at room temperature, followed by incubation with primary antibodies (diluted in buffer according to manufacturer’s recommendations) overnight at 4 °C with gentle agitation. After three 10-min washes with TBST (Tris-buffered saline with 0.1% Tween-20), membranes were incubated with appropriate HRP-conjugated secondary antibodies (Table S2) for 1 h at room temperature. Protein bands were visualized using enhanced chemiluminescence (ECL) substrate and imaged with a chemiluminescence detection system. Band intensities were quantified using ImageJ software with β-actin as loading controls. CCK-8 assay Cells (5 × 10³ cells/well) were seeded into 96-well cell culture plates and cultured overnight. At 0, 24, 48, 72, and 96 h post-seeding, 10 μL of Cell Counting Kit-8 (CCK-8) reagent was added to each well, and the plates were incubated at 37 °C for 2 h. Absorbance at 450 nm was measured using a microplate reader. For drug treatment, cells were exposed to [D-Leu-4]-OB3 at a concentration of 0.625 μM, and cell viability was assessed at 48 h. Each experiment was performed in triplicate. Colony formation assay A total of 500 cells were seeded into six-well cell culture plates and cultured for 14 days, with the medium changed every three days. The colonies were fixed with 4% paraformaldehyde for 15 min, stained with a 1% aqueous solution of crystal violet, Counting clones after imaging to determine formation ratio. Colonies containing >50 cells were included in the analysis and the colony formation rate was calculated using the formula: (number of colonies/number of seeded cells) × 100%. Transwell invasion assay 24-well Transwell chambers with an 8-μm pore size (TCS003024, jetbiofil, Guangzhou,China) that had been pre-coated with a 1:8 dilution of Matrigel were used. Cells (5 × 10⁴ cells/well) in serum-free medium were added to the upper chamber and 600 μL of medium supplemented with 20% FBS was added to the lower chamber. After 24 h, the non-invading cells were removed with a cotton swab and the invading cells were fixed, stained with crystal violet and captured using microscopy. For the drug treatment, the cells were pre-incubated with [D-Leu-4]-OB3 for 24 h before seeding. Wound healing assay The stable cells were seeded and cultured in 6-well cell culture plates in a density of 8 × 105 cells/well. A straight wound was created using a sterile 200 μL pipette tip. The cells were then washed with PBS buffer and cultured in a medium supplemented with 1% FBS. Images of the wound were captured at 0, 24 and 48 h using a light microscope. Wound closure was measured to assess the migration rate of the cells in the wound field. Experimental animals Female BALB/c-nu mice (4–6 weeks, 18–22 g weight) were purchased from Wuhan Life Origin Technology Co., Ltd. and housed in a specific pathogen-free (SPF) animal facility in Wuhan University Laboratory Animal Center. The facility strictly adhered to SPF-grade laboratory animal management protocols, maintaining temperatures between 22 and 26 °C and a relative humidity of 40–60%. A 12-h light/12-h dark circadian rhythm lighting system was used to simulate natural physiological rhythms. The housing cages and bedding were sterilized using high-pressure steam. Feed and drinking water were sterile, standard products that were replaced regularly to ensure a clean, stable environment. All animal-related procedures throughout the experiment complied with ethical guidelines for laboratory animals and were approved by the Wuhan University Animal Ethics Committee (Approval No ZN2025035). Subcutaneous xenograft model A549 cells (2 × 10⁶ cells/mouse) suspended in PBS mixed with Matrigel (1:1) were subcutaneously injected into nude mice (n = 6 per group). Nude mice were randomly assigned to the following groups: Vector (empty vector), S100A10-OE (S100A10 overexpression), shNC (non-targeting shRNA), shS100A10 (S100A10 knockdown), S100A10-sh + S100A10-pcDNA(empty vector), and S100A10-sh + pcDNA3.1-TNC (TNC overexpression). For drug treatment, mice with tumors (approximately 100 mm³) were injected intraperitoneally with [D-Leu-4]-OB3 (2 μg/kg) or DMSO daily. Tumor volume was measured every 3 days using a Vernier caliper (V = length × width² × 0.5), and the final tumor weight was recorded. Tail vein metastasis model Stable A549 cells with the indicated genetic manipulations were harvested during the logarithmic growth phase, washed twice with PBS, and resuspended in sterile PBS. A total of 1 × 106 cells per mouse were injected into the lateral tail vein of female BALB/c-nu mice (4–6 weeks old, n = 6 per group). After injection, the mice were maintained under specific pathogen-free conditions and monitored regularly for general condition and survival throughout the experimental period. Six weeks after cell inoculation, the mice were euthanized under anesthesia. Major organs were carefully dissected, and liver tissues were collected, photographed, and subjected to subsequent histological and immunohistochemical analyses. Immunohistochemical (IHC) staining First, perform stepwise dehydration in alcohol. After antigen retrieval, wash the sections with PBS on a decolourising shaker. Then, block endogenous peroxidase activity with a 3% hydrogen peroxide solution and wash with PBS. Use an immunohistochemistry pen to draw a hydrophobic circle around the tissue. Add 3% bovine serum albumin to block at room temperature. Then, add the primary antibody and incubate overnight at 4 °C in a humid chamber. Wash away the unbound primary antibody with PBS, then add the secondary antibody, diluted to the specified concentration. Incubate at room temperature in the dark. After washing with PBS, add the DAB chromogenic reagent and observe the colour development process under a microscope. When the positive signals are clearly visible, rinse the sections in distilled water. Counterstain with haematoxylin, differentiate with a 1% hydrochloric acid and ethanol solution for approximately one second, then rinse under running water after staining with methylene blue. Finally, dehydrate in a gradient of alcohols, air-dry briefly and mount with a neutral binder. Under bright-field microscopy, haematoxylin-stained nuclei appear blue, while regions expressing the target antigen show brown or light brown staining. RNA sequencing (RNA-seq) and data analysis Total RNA was extracted from S100A10-sh and shNC A549 cells using the Tissue/Cell Total RNA Extraction Kit. The quality of the RNA was assessed using a NanoDrop 2000 and an Agilent 2100 Bioanalyzer. RNA-seq library construction and sequencing were performed using the DNBSEQ platform by Wuhan BGI Gene Technology Service Co., Ltd. Clean reads were mapped to the human genome (GRCh38) using Hisat2, and gene expression was normalized to FPKM. Differentially expressed genes (DEGs) were identified using DESeq2 (|log₂FC| > 1, adj. P < 0.05). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses, as well as Gene Set Enrichment Analysis (GSEA), were performed using clusterProfiler and GSEA software (v4.3.2), respectively. Heatmaps were generated using the pheatmap R package. Quantitative real-time PCR (qRT-PCR) Total RNA (1 μg) was reverse-transcribed into complementary DNA (cDNA) using the OloScript All-in-One RT EasyMix for qPCR. Quantitative PCR (qPCR) was performed using the 2×Q3 SYBR qPCR Master Mix on a StepOnePlus Real-Time PCR System. Relative gene expression was calculated using the 2⁻ΔΔCt method, with GAPDH as the reference gene. The primers used are listed in Table S3. PCR reactions were carried out in 0.2 mL PCR Single Tubesand 0.1 mL Flat-Cap 8-Tube Strips. Dual-Luciferase reporter assay The TNC promoter region was cloned into the pGL3-Basic vector. HEK-293T cells (2 × 10⁴ cells/well) were seeded in 96-well white-bottom, white-wall plates and co-transfected with TNC-promoter-pGL3, pcDNA3.1-JUND (JUND overexpression), and pRL-TK (Renilla control) using Lipomaster2000. Prepare the Firefly Luciferase Reaction Reagent by mixing the Luciferase Reaction Substrate (50×) with the Luciferase Reaction Buffer I at a volume ratio of 1:49. To prepare the Sea Urchin Luciferase Reaction Reagent (Luciferase Reaction Reagent II), mix Luciferase Reaction Substrate II (50×) with Luciferase Reaction Buffer II at the same ratio. To process samples, add the appropriate volume of Cell Lysis Buffer to the cells according to the format of the culture plate . Incubate at room temperature for 10 min, then scrape the cells into centrifuge tubes. Centrifuge at 12,000×g for 10 min at 2–8 °C, then collect the supernatant. For fluorescence activity detection: Transfer 10–20 μL of the supernatant to an opaque 96-well plate. Add 100 μL of room temperature Luciferase Reaction Reagent II and mix thoroughly. Detect firefly luciferase activity at a wavelength of 350–700 nm for one second using a chemiluminescence reader. Then, add 100 μL of room temperature Luciferase Reaction Reagent II, mix well and detect at 380–780 nm for 1 s to record sea squirt luciferase activity. The final relative activity index is calculated as the ratio of the firefly luciferase fluorescence reading to the sea squirt luciferase fluorescence reading. Virtual screening Virtual screening was performed using the Schrödinger Maestro 12.8 software programme to identify potential small-molecule ligands for the human S100A10 protein. Three-dimensional structure visualisation was conducted using PyMOL software. The crystal structure of human S100A10 (PDB ID: 1A4P) was obtained from the Protein Data Bank (PDB) for pre-processing of the protein. Hydrogen atoms were added to the protein using the Protein Preprocessing Wizard module. Energy minimisation was then performed using the OPLS2005 force field with a 0.30 Å root mean square deviation (RMSD) constraint to maintain structural integrity. A receptor grid box was then generated using the Receptor Grid Generation module and centred on the Site 1 binding site predicted for S100A10. The LigPrep module in Schrödinger software converted the 2D structures of the library compounds into optimised 3D structures. Molecular docking was then performed using the Virtual Screening Workflow module in conjunction with the Glide module, which calculates docking based on the geometric and energetic compatibility of the receptor and ligand. Molecular docking The crystal structure of the S100A10 receptor protein was downloaded from the Protein Data Bank (PDB) with the corresponding ID 1A4P. The two-dimensional structure of the ligand molecule [D-Leu-4]-OB3 was drawn and optimised using ChemDraw software. Based on an analysis of the S100A10 active site pocket using PyMOL or UCSF Chimera software, the centre coordinates of the docking grid box were fixed within this active site region for docking parameter configuration. The exhaustiveness parameter was adjusted to enhance the thoroughness of the conformation search. The docking command was then executed to initiate molecular docking calculations. Upon completion, AutoDock Vina automatically output the binding free energy (ΔG) for each conformation, thereby quantifying the binding affinity between S100A10 and [D-Leu-4]-OB3. Simultaneously, PyMOL (v2.5.2) was used to visualise the optimal binding conformation. The key interactions formed between the two molecules (e.g. hydrogen bonds, hydrophobic interactions and salt bridges) and the amino acid residues involved were analysed in detail. Microscale thermophoresis (MST) Recombinant human S100A10 protein with a His-tag was fluorescently labelled using NHS-fluorescein, in accordance with the manufacturer's recommended protocol, to ensure efficient, site-specific conjugation. For the binding affinity assay, a series of gradient dilutions of [D-Leu-4]-OB3 were prepared at concentrations ranging from 0.007 μM to 100 μM. Each dilution was then co-incubated with 50 nM of the fluorescently labelled S100A10 protein in MST buffer (50 mM Tris–HCl (pH 7.4), 150 mM NaCl and 0.05% Tween 20) to maintain the physiological environment and protein stability during the experiment. After thorough incubation to reach binding equilibrium, MST signals reflecting the interaction between [D-Leu-4]-OB3 and S100A10 were detected using a Monolith NT.115 instrument with optimised parameters. The obtained MST raw data were then imported into MO.Affinity Analysis software (NanoTemper Technologies, Germany) for processing and fitting. Finally, the dissociation constant (K_d), representing the binding affinity between the two molecules, was calculated. Thermal shift assay A thermal stability analysis was conducted on the recombinant S100A10-GFP fusion protein under experimental conditions, both with and without [D-Leu-4]-OB3. The samples then underwent a gradient heating programme using a temperature-controlled thermal cycler. The temperature was increased gradually from 37 °C to 69 °C at a rate of 1 °C every three minutes. After heating, all samples were centrifuged at 4 °C. The intact protein fraction was analysed by SDS-PAGE on a precast 12% polyacrylamide gel under denaturing conditions. ImageJ software was then used to analyse the intensity of the S100A10-GFP protein band in each lane quantitatively. The melting temperature of S100A10-GFP was determined by fitting normalised density measurements to an S-shaped dose-response curve using GraphPad Prism software. All experiments were repeated three times to ensure statistical reliability, and the data are presented as the mean ± standard deviation (SD). Statistical analysis All in vitro and in vivo experiments were performed independently at least three times. Student’s t-test was used to detect differences with GraphPad Prism 8.0 (GraphPad Software Inc., La Jolla, CA, USA) or SPSS software (SPSS Inc., Chicago, IL, USA). The quantitative results are presented as mean±SEM. A P < 0.05 was considered significant (* P < 0.05, ⁎⁎P < 0.01, ⁎⁎⁎P < 0.001, ns: not significantly different). Results S100A10 is highly expressed in LUAD and associated with poor prognosis, but not in LUSC In order to clarify the differential expression of S100A10 in non-small cell lung carcinoma (NSCLC), we downloaded TPM-formatted sequencing data from The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) projects that had undergone standardised processing via the UCSC XENA Toile workflow. The results revealed that S100A10 exhibited high expression in lung cancer (p < 0.05). Further analysis showed that S100A10 was significantly overexpressed primarily in LUAD, with no significant difference observed in LUSC (Fig. 1A). We then conducted a prognostic analysis of S100A10 in NSCLC. The results showed that LUAD patients with high S100A10 expression had poorer overall survival (OS) and progression-free survival (PFS), whereas S100A10 expression was unrelated to prognosis in LUSC patients (Fig. 1B). We also examined S100A10 expression levels in normal lung bronchial epithelial cells (BEAS-2B) and in several lung cancer cell lines (A549, PC9, etc.). Western blot (WB) results indicated that S100A10 expression was higher in A549, PC9,H1299 and H1975 cells than in BEAS-2B cells (p < 0.05). Additionally, immunofluorescence (IF) analysis of S100A10 localisation in lung adenocarcinoma cells revealed that S100A10 was primarily localised to the cell membrane and cytoplasm in A549 and PC9 cells (Fig. 1D). These findings suggest that S100A10 is significantly overexpressed in LUAD (a subtype of NSCLC) and is associated with poorer patient prognosis, though no such association was observed with LUSC.Fig. 1Expression pattern and prognostic significance of S100A10 in non-small cell lung cancer (NSCLC). (A) S100A10 expression levels in normal lung tissues and lung cancer tissues based on TCGA and GTEx datasets. Subgroup analysis shows significantly increased S100A10 expression in lung adenocarcinoma (LUAD), but not in lung squamous cell carcinoma (LUSC). Data are presented as log₂(TPM + 1). (B) Kaplan–Meier survival analyses of overall survival (OS) and progression-free survival (PFS) according to S100A10 expression levels in LUAD and LUSC cohorts. High S100A10 expression is associated with poorer OS and PFS in LUAD, but not in LUSC. Hazard ratios (HR) and P values are indicated in the plots. (C) Western blot analysis of S100A10 protein expression in human bronchial epithelial cells (BEAS-2B) and LUAD cell lines (A549, PC9, H1299, and H1975), with β-actin used as a loading control. Quantification of relative protein expression levels is shown on the right. (D) Immunofluorescence staining of S100A10 in A549 and PC9 cells. S100A10 (green) is primarily localized in the cytoplasm and cell membrane, and nuclei are counterstained with DAPI (blue). Scale bars: 50 μm. Data are presented as mean ± SEM from at least three independent experiments. Statistical significance was determined using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant).Fig 1 dummy alt text Expression pattern and prognostic significance of S100A10 in non-small cell lung cancer (NSCLC). (A) S100A10 expression levels in normal lung tissues and lung cancer tissues based on TCGA and GTEx datasets. Subgroup analysis shows significantly increased S100A10 expression in lung adenocarcinoma (LUAD), but not in lung squamous cell carcinoma (LUSC). Data are presented as log₂(TPM + 1). (B) Kaplan–Meier survival analyses of overall survival (OS) and progression-free survival (PFS) according to S100A10 expression levels in LUAD and LUSC cohorts. High S100A10 expression is associated with poorer OS and PFS in LUAD, but not in LUSC. Hazard ratios (HR) and P values are indicated in the plots. (C) Western blot analysis of S100A10 protein expression in human bronchial epithelial cells (BEAS-2B) and LUAD cell lines (A549, PC9, H1299, and H1975), with β-actin used as a loading control. Quantification of relative protein expression levels is shown on the right. (D) Immunofluorescence staining of S100A10 in A549 and PC9 cells. S100A10 (green) is primarily localized in the cytoplasm and cell membrane, and nuclei are counterstained with DAPI (blue). Scale bars: 50 μm. Data are presented as mean ± SEM from at least three independent experiments. Statistical significance was determined using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant). High S100A10 expression promotes cell proliferation of LUAD in vitro and in vivo To investigate the biological function of S100A10 in LUAD, we generated stable cell models with S100A10 knockdown and overexpression in A549 and PC9 cell. Our results revealed that, compared to shNC group, S100A10 protein expression was significantly reduced in the shS100A10 group, especially in S100A10-sh2 group (p<0.05), while the OE group exhibited high levels of the S100A10-GFP fusion protein than vector group (p<0.05). To evaluate the effect of S100A10 on cell growth, CCK-8 and colony formation assays were conducted to assess cellular proliferation activity. The CCK-8 assay results showed that, compared to the shNC group, the S100A10-sh2 group exhibited significantly reduced cell viability. In contrast, the OE group exhibited markedly increased proliferation activity compared to the vector group (Fig. 2B). The colony formation assay also revealed that the S100A10-sh2 group exhibited significantly reduced colony-forming capacity compared to the shNC group, while the OE group demonstrated markedly enhanced colony formation compared to the vector group. Collectively, these findings indicate that highly expressed S100A10 promotes cell proliferation of LUAD in vitro (Fig. 2C). Additionally, we validated the effect of S100A10 on tumor proliferation of LUAD in vivo through a xenograft model of nude mice. Consistent with the in vitro experimental results, the tumor volume and weight in the S100A10-sh2 group was significantly decreased compared to the shNC group. Conversely, the tumor size and tumor weight in the OE group were increased significantly compared to the vector group (Fig. 2D). Subsequently, we also evaluated Ki67 expression of tumor using IHC staining assay. Results showed that the ratio of Ki67-positive cells was significantly decreased in the S100A10-sh2 group, and increased in the OE group (Fig. 2E). Taken together, these results all indicate that high S100A10 expression promotes cell proliferation of LUAD both in vitro and in vivo.Fig. 2S100A10 promotes LUAD cell proliferation in vitro and in vivo. (A) Western blot analysis confirming S100A10 knockdown and overexpression efficiency in A549 and PC9 cells. Two independent shRNAs (sh1 and sh2) were used for S100A10 silencing, and S100A10-GFP was used for overexpression. β-actin was used as a loading control. Quantification of relative protein expression levels is shown below. (B) Cell proliferation assessed by CCK-8 assay in A549 and PC9 cells with S100A10 knockdown or overexpression. Cell viability was measured at the indicated time points (0–120 h) and expressed as OD450 values. (C) Colony formation assays in A549 and PC9 cells. Representative images show the number of colonies formed under different S100A10 expression conditions. (D) Subcutaneous xenograft tumor model in nude mice. Representative images of excised tumors are shown. Tumor weight and tumor volume growth curves are presented for S100A10 overexpression and knockdown groups. (E) Immunohistochemical (IHC) staining of Ki67 in tumor tissues derived from xenograft models. Representative images are shown to indicate proliferative activity. Scale bars: 50 μm. Data are presented as mean ± SEM from at least three independent experiments. Statistical significance was determined using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant).Fig 2 dummy alt text S100A10 promotes LUAD cell proliferation in vitro and in vivo. (A) Western blot analysis confirming S100A10 knockdown and overexpression efficiency in A549 and PC9 cells. Two independent shRNAs (sh1 and sh2) were used for S100A10 silencing, and S100A10-GFP was used for overexpression. β-actin was used as a loading control. Quantification of relative protein expression levels is shown below. (B) Cell proliferation assessed by CCK-8 assay in A549 and PC9 cells with S100A10 knockdown or overexpression. Cell viability was measured at the indicated time points (0–120 h) and expressed as OD450 values. (C) Colony formation assays in A549 and PC9 cells. Representative images show the number of colonies formed under different S100A10 expression conditions. (D) Subcutaneous xenograft tumor model in nude mice. Representative images of excised tumors are shown. Tumor weight and tumor volume growth curves are presented for S100A10 overexpression and knockdown groups. (E) Immunohistochemical (IHC) staining of Ki67 in tumor tissues derived from xenograft models. Representative images are shown to indicate proliferative activity. Scale bars: 50 μm. Data are presented as mean ± SEM from at least three independent experiments. Statistical significance was determined using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant). Increased S100A10 expression promotes cell metastasis of LUAD To explore the role of S100A10 on cell metastasis in LUAD, we examined the effects of S100A10 on cell migration and invasion both in vitro and in vivo. As illustrated in Fig. 3A, the transwell invasion assay results revealed that S100A10 overexpression significantly enhanced the invasive capacity of both A549 and PC9 cells. Conversely, S100A10 knockdown significantly reduced the number of invading A549 and PC9 cells (Fig. 3A). Furthermore, the scratch assay results indicated that S100A10 knockdown decreased the capacity of migrated A549 and PC9 cells, while S100A10 overexpression promoted cell migration (Fig. 3B). To further investigate the role of S100A10 in tumor dissemination in vivo, we established an experimental metastasis model by tail-vein injection of LUAD cells. At the experimental endpoint, metastatic lesions were detected in liver tissues of tumor-bearing mice. Quantitative analysis showed that S100A10 knockdown significantly reduced the number of metastatic nodules observed in the liver, whereas S100A10 overexpression increased the metastatic burden detected in liver tissues (Fig. 3C). In addition, we performed IHC staining assay to detect the metastasis associated markers in metastatic liver tissues. The IHC staining results revealed that S100A10 knockdown exhibited markedly increased E-cadherin expression and significantly decreased N-cadherin and vimentin expression in metastatic liver tissues, and S100A10 overexpression significantly decreased E-cadherin expression alongside increased N-cadherin and vimentin expression(Fig. 3D). Overall, these results suggest that S100A10 plays a role in promoting LUAD metastasis.Fig. 3S100A10 promotes LUAD cell migration and invasion in vitro and in vivo. (A) Transwell invasion assays in A549 and PC9 cells with S100A10 overexpression or knockdown. Representative images and quantification of invaded cells are shown. (B) Wound healing assays evaluating cell migratory capacity in A549 and PC9 cells at 0, 24, and 48 h. Representative images and quantitative analysis of migration rates (wound closure percentage) are presented. (C) Experimental metastasis model established by tail-vein injection of LUAD cells. Representative images of liver tissues with metastatic nodules are shown (indicated by arrows). (D) Immunohistochemical (IHC) staining of EMT-related markers in liver tissues, including E-cadherin (E-cad), N-cadherin (N-cad), and vimentin. Representative images are shown. Scale bars: 50 μm.Data are presented as mean ± SEM from at least three independent experiments. Statistical significance was determined using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant).Fig 3 dummy alt text S100A10 promotes LUAD cell migration and invasion in vitro and in vivo. (A) Transwell invasion assays in A549 and PC9 cells with S100A10 overexpression or knockdown. Representative images and quantification of invaded cells are shown. (B) Wound healing assays evaluating cell migratory capacity in A549 and PC9 cells at 0, 24, and 48 h. Representative images and quantitative analysis of migration rates (wound closure percentage) are presented. (C) Experimental metastasis model established by tail-vein injection of LUAD cells. Representative images of liver tissues with metastatic nodules are shown (indicated by arrows). (D) Immunohistochemical (IHC) staining of EMT-related markers in liver tissues, including E-cadherin (E-cad), N-cadherin (N-cad), and vimentin. Representative images are shown. Scale bars: 50 μm.Data are presented as mean ± SEM from at least three independent experiments. Statistical significance was determined using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant). The tumor metastasis effect of S100A10 in LUAD is associated with epithelial-mesenchymal transition (EMT) To investigate the regulatory mechanism of S100A10 in LUAD cells, RNA-sequencing was used to analyze transcriptome differences between the S100A10 knockdown group (S100A10-sh2) and the control group (shNC). The differential expression genes (DEGs) results were visualized via volcano plots. At the set thresholds of |log₂FC|>1 and a corrected P-value of P < 0.05, we screened a total of 5920 DEGs, including 3060 downregulated genes (blue dots) and 2860 upregulated genes (red dots) (Fig. 4A). Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed on DEGs to investigate the biological function and involved signaling pathway of S100A10 (Fig. 4B). These analysis results indicates that S100A10 regulates extracellular matrix remodelling and tumor invasion by modulating cytokine signaling network and pro-inflammatory pathways in the tumor microenvironment. To further validate the pro-metastatic mechanism of S100A10 in LUAD, gene set enrichment analysis (GSEA) was performed using the transcriptome data of RNA-seq. GSEA results showed that multiple gene sets were enriched and involved in the metastatic role of S100A10 in LUAD, including TNF-a signaling via NFKB, EMT, Apical junction, INF-a response, and INF-gamma response (p<0.05). We noticed that EMT (NES value: −1.828; FDR <0.05) (Fig. 4C) may play a critical role in S100A10-mediated LUAD cell metastasis. We also generated heatmaps to display all the differentially expressed genes in the EMT gene sets in RNA-seq (Fig. 4D). All above bioinformatics analysis results of RNA-seq indicated that EMT play a role in LUAD metastasis and correlated with S100A10 expression.Fig. 4S100A10 regulates gene expression and promotes epithelial–mesenchymal transition (EMT) in LUAD cells. (A) Volcano plot showing differentially expressed genes (DEGs) between S100A10 knockdown (S100A10-sh) and control (shNC) A549 cells. Genes with |log₂ fold change| > 1 and adjusted P < 0.05 are highlighted. (B) Gene Ontology (GO) enrichment analysis of DEGs, showing significantly enriched biological processes, cellular components, and molecular functions. (C) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of DEGs. Significantly enriched pathways are shown based on gene ratio and adjusted P values. (D) Gene set enrichment analysis (GSEA) of hallmark pathways, including epithelial–mesenchymal transition (EMT), TNFα signaling via NF-κB, apical junction, interferon alpha response, and interferon gamma response. (E) Heatmap of representative EMT-related genes derived from RNA-seq data, comparing S100A10 knockdown and control groups. (F) Quantitative real-time PCR (qRT-PCR) validation of selected candidate genes identified from RNA-seq analysis in S100A10 knockdown and overexpression models. (G) Venn diagram showing overlapping transcription factors predicted to regulate TNC expression based on hTFtarget, KnockTF, and TRRUST databases. (H) Western blot analysis of S100A10 and JUND expression in S100A10 knockdown cells, with β-actin used as a loading control. Quantification of protein levels is shown below. (I) Dual-luciferase reporter assay evaluating the transcriptional activity of the TNC promoter in the presence of JUND overexpression. (J) Nuclear and cytoplasmic fractionation assays showing the distribution of JUND and phosphorylated JUND (p-JUND) in S100A10 knockdown and control cells. β-tubulin and histone H3 were used as cytoplasmic and nuclear loading controls, respectively. Data are presented as mean ± SEM from at least three independent experiments. Statistical significance was determined using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant).Fig 4 dummy alt text S100A10 regulates gene expression and promotes epithelial–mesenchymal transition (EMT) in LUAD cells. (A) Volcano plot showing differentially expressed genes (DEGs) between S100A10 knockdown (S100A10-sh) and control (shNC) A549 cells. Genes with |log₂ fold change| > 1 and adjusted P < 0.05 are highlighted. (B) Gene Ontology (GO) enrichment analysis of DEGs, showing significantly enriched biological processes, cellular components, and molecular functions. (C) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of DEGs. Significantly enriched pathways are shown based on gene ratio and adjusted P values. (D) Gene set enrichment analysis (GSEA) of hallmark pathways, including epithelial–mesenchymal transition (EMT), TNFα signaling via NF-κB, apical junction, interferon alpha response, and interferon gamma response. (E) Heatmap of representative EMT-related genes derived from RNA-seq data, comparing S100A10 knockdown and control groups. (F) Quantitative real-time PCR (qRT-PCR) validation of selected candidate genes identified from RNA-seq analysis in S100A10 knockdown and overexpression models. (G) Venn diagram showing overlapping transcription factors predicted to regulate TNC expression based on hTFtarget, KnockTF, and TRRUST databases. (H) Western blot analysis of S100A10 and JUND expression in S100A10 knockdown cells, with β-actin used as a loading control. Quantification of protein levels is shown below. (I) Dual-luciferase reporter assay evaluating the transcriptional activity of the TNC promoter in the presence of JUND overexpression. (J) Nuclear and cytoplasmic fractionation assays showing the distribution of JUND and phosphorylated JUND (p-JUND) in S100A10 knockdown and control cells. β-tubulin and histone H3 were used as cytoplasmic and nuclear loading controls, respectively. Data are presented as mean ± SEM from at least three independent experiments. Statistical significance was determined using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant). S100A10 promotes TNC expression via upregulating transcription factor JUND To explore the specific signaling pathways through which S100A10 regulates EMT in LUAD cells, we employed quantitative reverse transcription PCR (qRT-PCR) to evaluate the expression changes of 11 candidate genes with differential expression in EMT process identified by GSEA. Compared with the shNC group, the expression of genes including TNC, MEST, CXCL8, TIMP3, CXCL1, PRSS2, and WNT5A were significantly lower in the S100A10-sh group, and SAT1, DKK1, SNTB1 were upregulated in S100A10-sh group. Conversely, TNC and SAT1 were significantly upregulated in the OE group, and MEST was downregulated in the OE group. The results showed that, of the 11 candidate genes, only TNC was downregulated in the S100A10 knockdown group and upregulated in the S100A10 overexpression group (Fig. 4F). To screen for the upstream transcription factors that regulate TNC expression, we integrated the hTFtarget, KnockTF and TRRUST databases for cross-validation. Three transcription factors were overlay in three databases, including JUND, SMAD3, and SMAD4. Then, further experimental results indicated that JUND was identified as the key transcription factor that regulated TNC transcription directly (Fig. 4G). As shown in Fig. 4H, western blot analysis results showed that JUND expression was significantly reduced in the S100A10-sh group compared to the shNC group. Dual luciferase assays revealed significantly enhanced TNC promoter transcription activity in the JUND group compared to the NC group (Fig. 4I). These findings provide preliminary evidence of the existence of the JUND-TNC transcriptional regulatory pathway in LUAD and offer key insights into the molecular mechanisms of the S100A10-JUND-TNC signaling axis in tumor invasion. In addition, nuclear fractionation analysis demonstrated that knockdown of S100A10 markedly reduced the nuclear accumulation of JUND and phosphorylated JUND compared with the shNC group(Fig. 4 J), suggesting that S100A10 facilitates JUND phosphorylation and nuclear translocation. S100A10 promotes LUAD cell migration and invasion through TNC-mediated EMT To further validate the molecular mechanism by which S100A10 regulates LUAD cell EMT, migration, and invasion through upregulation of TNC, we constructed TNC-overexpressing cells for rescue experiments. As shown in Fig. 5A and B, our results demonstrated that TNC overexpression promoted LUAD cell migration and invasion, and reversed the inhibitory effects of S100A10 knockdown on LUAD cell migration and invasion, confirming TNC as a key downstream effector of S100A10. We further examined the expression of key EMT-related markers in LUAD cells following S100A10 knockdown and TNC supplementation via Western blot analysis. The results showed that S100A10 knockdown increased E-cadherin expression and reduced N-cadherin and vimentin expression, while TNC overexpression effectively reversed these changes in EMT-related markers (Fig. 5C). Subsequently, we established a subcutaneous xenograft tumor model to further validate the mechanism by which the S100-TNC axis regulates EMT, migration and invasion in lung cancer in vivo. Our results showed that S100A10 knockdown significantly reduced tumor weight and volume, while TNC overexpression promoted tumor growth and markedly reversed the inhibitory effects of S100A10 knockdown in vivo (Fig. 5D). Immunohistochemical assay results showed that S100A10 knockdown led to increased E-cadherin expression and decreased N-cadherin and vimentin expression. TNC overexpression decreased E-cadherin expression while increasing N-cadherin and vimentin levels (Fig. 5E), thereby reversing the inhibitory effects caused by S100A10 knockdown. These results demonstrate that TNC is an indispensable downstream effector in S100A10-regulated EMT phenotypic conversion, and the S100A10-TNC signaling axis promotes malignant progression of LUAD by driving EMT.Fig. 5Validation of the S100A10-TNC axis in regulating LUAD cell EMT, migration, invasion, and tumor growth. (A) Transwell invasion assays in A549 and PC9 cells under different conditions: control (NC-S100A10 + NC-TNC), S100A10 knockdown (SH-S100A10 + NC-TNC), TNC overexpression (NC-S100A10 + OE-TNC), and rescue group (SH-S100A10 + OE-TNC). Representative images and quantification of invaded cells are shown. (B) Wound healing assays evaluating migratory capacity in A549 and PC9 cells at 0, 24, and 48 h under the indicated treatment conditions. Migration rates are quantified as wound closure percentage. (C) Western blot analysis of S100A10, TNC, and EMT-related markers (E-cadherin, N-cadherin, and vimentin) under the indicated conditions. β-actin was used as a loading control. Quantification of protein expression levels is shown. (D) Subcutaneous xenograft tumor model showing tumor growth under different treatment conditions. Representative images of tumors, tumor weight, and tumor volume growth curves are presented. (E) Immunohistochemical (IHC) staining of S100A10, TNC, and EMT markers (E-cadherin, N-cadherin, and vimentin) in tumor tissues derived from xenograft models. Representative images are shown. Scale bars: 50 μm. Data are presented as mean ± SEM from at least three independent experiments. Statistical significance was determined using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant).Fig 5 dummy alt text Validation of the S100A10-TNC axis in regulating LUAD cell EMT, migration, invasion, and tumor growth. (A) Transwell invasion assays in A549 and PC9 cells under different conditions: control (NC-S100A10 + NC-TNC), S100A10 knockdown (SH-S100A10 + NC-TNC), TNC overexpression (NC-S100A10 + OE-TNC), and rescue group (SH-S100A10 + OE-TNC). Representative images and quantification of invaded cells are shown. (B) Wound healing assays evaluating migratory capacity in A549 and PC9 cells at 0, 24, and 48 h under the indicated treatment conditions. Migration rates are quantified as wound closure percentage. (C) Western blot analysis of S100A10, TNC, and EMT-related markers (E-cadherin, N-cadherin, and vimentin) under the indicated conditions. β-actin was used as a loading control. Quantification of protein expression levels is shown. (D) Subcutaneous xenograft tumor model showing tumor growth under different treatment conditions. Representative images of tumors, tumor weight, and tumor volume growth curves are presented. (E) Immunohistochemical (IHC) staining of S100A10, TNC, and EMT markers (E-cadherin, N-cadherin, and vimentin) in tumor tissues derived from xenograft models. Representative images are shown. Scale bars: 50 μm. Data are presented as mean ± SEM from at least three independent experiments. Statistical significance was determined using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant). [D-Leu-4]-OB3 was identified as an inhibitor targeting S100A10, [D-Leu-4]-OB3 suppresses LUAD invasion, migration, and epithelial-mesenchymal transition by regulating the S100A10-TNC axis To identify S100A10-targeting drugs, we employed virtual screening techniques to screen small-molecule compounds targeting S100A10, and a small-molecule inhibitor [D-Leu-4]-OB3 was obtained. Then, computer-aided molecular docking was used to simulate and analyze the interaction between [D-Leu-4]-OB3 and S100A10 (Fig. 6A). To further validate the targeting interaction between [D-Leu-4]-OB3 and S100A10, Microacoustic titration (MST) and microcalorimetric titration experiments were employed. MST results showed that [D-Leu-4]-OB3 binds to S100A10 with moderate micromolar affinity (Fig. 6B). Microcalorimetric measurements indicate that the thermal stability of the S100A10 protein decreases gradually with increasing temperature. Meanwhile, S100A10 bound to OB3 degrades faster at elevated temperatures. (Fig. 6C). Both MST and microcalorimetry experiments confirmed that [D-Leu-4]-OB3 could directly bind to S100A10 and promote degradation of S100A10.Fig. 6[D-Leu-4]-OB3 targets S100A10 and inhibits LUAD cell migration and invasion. (A) Molecular docking analysis showing the predicted binding mode of [D-Leu-4]-OB3 with S100A10. Key interactions, including hydrogen bonds and hydrophobic contacts, are illustrated. (B) Microscale thermophoresis (MST) assay demonstrating the binding affinity between [D-Leu-4]-OB3 and recombinant S100A10 protein. Binding curves and normalized fluorescence (Fnorm) are shown. (C) Thermal shift assay (TSA) evaluating the thermal stability of S100A10 in the presence or absence of [D-Leu-4]-OB3. Representative immunoblots and quantification of relative protein abundance across increasing temperatures are presented. (D) Wound healing assays assessing cell migration in A549 and PC9 cells treated with [D-Leu-4]-OB3 at 0, 24, and 48 h. Migration rates are quantified as wound closure percentage. (E) Transwell invasion assays showing the invasive capacity of A549 and PC9 cells treated with [D-Leu-4]-OB3. Representative images and quantification of invaded cells are shown. (F) Western blot analysis of S100A10, TNC, and EMT-related markers (E-cadherin, N-cadherin, and vimentin) in A549 and PC9 cells treated with [D-Leu-4]-OB3. β-actin was used as a loading control. Quantification of protein expression levels is shown. (G) Subcutaneous xenograft tumor model evaluating the antitumor effects of [D-Leu-4]-OB3 in vivo. Representative images of mice and excised tumors are shown, along with tumor weight and tumor volume growth curves in A549 and PC9 models. Data are presented as mean ± SEM from at least three independent experiments. Statistical significance was determined using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant).Fig 6 dummy alt text [D-Leu-4]-OB3 targets S100A10 and inhibits LUAD cell migration and invasion. (A) Molecular docking analysis showing the predicted binding mode of [D-Leu-4]-OB3 with S100A10. Key interactions, including hydrogen bonds and hydrophobic contacts, are illustrated. (B) Microscale thermophoresis (MST) assay demonstrating the binding affinity between [D-Leu-4]-OB3 and recombinant S100A10 protein. Binding curves and normalized fluorescence (Fnorm) are shown. (C) Thermal shift assay (TSA) evaluating the thermal stability of S100A10 in the presence or absence of [D-Leu-4]-OB3. Representative immunoblots and quantification of relative protein abundance across increasing temperatures are presented. (D) Wound healing assays assessing cell migration in A549 and PC9 cells treated with [D-Leu-4]-OB3 at 0, 24, and 48 h. Migration rates are quantified as wound closure percentage. (E) Transwell invasion assays showing the invasive capacity of A549 and PC9 cells treated with [D-Leu-4]-OB3. Representative images and quantification of invaded cells are shown. (F) Western blot analysis of S100A10, TNC, and EMT-related markers (E-cadherin, N-cadherin, and vimentin) in A549 and PC9 cells treated with [D-Leu-4]-OB3. β-actin was used as a loading control. Quantification of protein expression levels is shown. (G) Subcutaneous xenograft tumor model evaluating the antitumor effects of [D-Leu-4]-OB3 in vivo. Representative images of mice and excised tumors are shown, along with tumor weight and tumor volume growth curves in A549 and PC9 models. Data are presented as mean ± SEM from at least three independent experiments. Statistical significance was determined using Student’s t-test (*P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant). Subsequently, we evaluated the effect of [D-Leu-4]-OB3 on the invasion and migration LUAD cells using cell scratch assays and Transwell experiments. Scratch assay results showed that [D-Leu-4]-OB3 effectively inhibited lung cancer cell migration (Fig. 6D). Transwell assays revealed that [D-Leu-4]-OB3 significantly reduced the number of invasive lung cancer cells (Fig. 6E). Western blot analysis also showed that the drug-treated groups significantly decreased the expression of S100A10, TNC, N-cadherin and vimentin, while increased E-cadherin expression markedly (Fig. 6F). These results suggest that [D-Leu-4]-OB3 exerts anti-tumor metastastic effects by inhibiting the S100A10-TNC axis mediated EMT in LUAD. Furthermore, we further established a subcutaneous xenograft tumor model to validate the inhibitory effect of [D-Leu-4]-OB3 on LUAD in vivo. As shown in Fig. 6 G, [D-Leu-4]-OB3 significantly reduced the weight and the volume of A549 and PC9 subcutaneous tumor by approximately 60% and 70%, respectively. Immunohistochemical analysis also revealed significantly reduced expression of S100A10, TNC and N-cadherin/vimentin in the [D-Leu-4]-OB3-treated groups compared to the vector group, while E-cadherin expression was increased. These results demonstrate that [D-Leu-4]-OB3 exerts anti-tumor effects by inhibiting S100A10-TNC axis mediated-EMT in LUAD both in vitro and in vivo, providing promising prospects for clinical application. Discussion Lung cancer is one of the most prevalent and lethal malignancies worldwide, and NSCLC accounts for over 80% of cases [41,42]. Recurrence and distant metastasis are the main factors that contribute to the poor prognosis of patients with NSCLC [9,43,44]. Although targeted therapy and immunotherapy have significantly improved survival rates for some patients, many still face therapeutic challenges due to the lack of effective targets or the development of resistance [3,18,45,46]. Therefore, identifying novel tumour-driving molecules and understanding the regulatory mechanisms involved are crucial for developing new therapeutic strategies. Given that the clinical association of S100A10 was significant in LUAD but not in LUSC, the present study mainly focuses on LUAD. As members of the calmodulin superfamily, S100 family proteins play a key role in regulating tumour cell proliferation, invasion and microenvironment remodelling [23,29]. One such protein, S100A10, has been shown to be highly expressed and associated with poor prognosis in various cancers, including colorectal and breast cancer [[47], [48], [49]]. However, its biological functions and regulatory mechanisms in LUAD remain poorly understood. In this study, we demonstrated through multiple experimental approaches that S100A10 promotes the EMT process by activating the JUND-TNC signalling axis, thereby facilitating the proliferation, invasion and metastasis of LUAD cells. We also identified and validated a small-molecule S100A10 inhibitor, [D-Leu-4]-OB3, which effectively inhibits the malignant progression of LUAD in vitro and in vivo by blocking the S100A10-mediated signalling pathway. This compound provided a promising new target for targeted therapy of LUAD. In order to investigate the role of S100A10 in LUAD, we conducted in vitro and in vivo experiments which revealed a close association between S100A10 and invasive migration in LUAD. These results implied that S100A10 could be an unfavourable prognostic indicator for LUAD, playing a pivotal oncogenic role in LUAD cell proliferation and tumour formation in vivo via EMT. The functional mechanism of S100A10 may involve regulating the cell cycle or inhibiting apoptosis. Further validation of these hypotheses is required in subsequent studies through cell cycle flow cytometry analysis and apoptosis assays. To elucidate the molecular mechanism by which S100A10 regulates EMT, we analyzed the transcriptomes of S100A10-knockdown and control A549 cells via RNA sequencing (RNA-seq). Bioinformatics analysis indicated that S100A10 may promote metastasis by activating EMT-related transcriptional programs. Subsequent quantitative polymerase chain reaction (qPCR) validation demonstrated that S100A10 expression levels were strongly correlated with TNC. As an extracellular matrix glycoprotein, TNC has been shown to be highly expressed in multiple cancers and to promote metastasis by activating the integrin signaling pathway or regulating EMT. Therefore, we inferred that TNC may be a key downstream effector molecule of S100A10. Furthermore, our experiments revealed that S100A10 promotes the transcription of its downstream target gene, TNC, by activating JUND phosphorylation. This drived EMT and enhanced the invasive and metastatic capabilities of LUAD cells. The discovery of the 'S100A10-JUND-TNC-EMT' regulatory axis provided a novel theoretical framework for understanding the molecular mechanisms of LUAD metastasis. Due to the pivotal oncogenic role of S100A10 in LUAD, the development of small-molecule inhibitors that target S100A10 is of significant clinical value. This study used computer-aided drug design to screen S100A10 inhibitors, which were then validated through subsequent experiments. Integrated structural biology and biophysical results confirm that [D-Leu-4]-OB3 specifically inhibits S100A10. This compound effectively suppressed the malignant progression of LUAD in both in vitro and in vivo models by specifically inhibiting S100A10 and blocking its downstream JUND-TNC-EMT signalling axis. The development of this inhibitor provided a highly promising new therapeutic candidate for targeted LUAD treatment. Its clinical potential can be further evaluated through pharmacokinetic and toxicological studies, as well as preclinical xenograft models such as PDX. Although this study shed light on the function and regulatory mechanisms of S100A10 in LUAD, there were several limitations. Firstly, the upstream signalling pathways through which S100A10 activates JUND phosphorylation were not thoroughly investigated. For example, it is unclear whether S100A10 activates the MAPK/ERK or PI3K/Akt pathways by binding to receptors such as ANXA2, thereby regulating JUND phosphorylation. Further exploration through experiments such as immunoprecipitation (Co-IP) and phosphoprotein arrays is warranted. Secondly, the antitumour activity of [D-Leu-4]-OB3 was validated only in cell-line-derived xenograft (CDX) models and not in patient-derived xenograft (PDX) models, which more accurately reflect clinical tumour heterogeneity. The clinical translational potential of this inhibitor should be evaluated using PDX models in future studies. It should be noted that, due to the incomplete collection of lung specimens and the lack of longitudinal in vivo imaging data in the present study, we were unable to quantitatively compare the metastatic burden between the lungs and liver. Although tail-vein injection models are commonly associated with frequent pulmonary colonization, our data primarily demonstrate the role of S100A10 in promoting the hepatic colonization of circulating tumor cells. Future studies incorporating orthotopic transplantation models and advanced in vivo imaging techniques will be required to comprehensively evaluate the impact of S100A10 on primary tumor growth in the lung and the spectrum of multi-organ metastasis. Future research could take the following directions: Firstly, the role of S100A10 in drug resistance in LUAD should be investigated further, particularly in terms of its potential contribution to resistance to epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) or chemotherapy drugs through the regulation of EMT or DNA damage repair mechanisms. If this hypothesis is confirmed, novel combination therapy strategies could be developed for drug-resistant LUAD patients. Secondly, exploring the interaction between S100A10 and the tumour microenvironment is another important area for research. RNA-seq analysis indicated that inflammation-related pathways were significantly downregulated following S100A10 knockdown. This suggests that S100A10 may influence tumour-associated macrophages (TAMs) or T cell infiltration by modulating cytokine secretion, such as that of CXCL8. This would remodel the tumour immune microenvironment. Future research could employ techniques such as single-cell sequencing or immunofluorescence staining to investigate the impact of S100A10 on the cellular composition of the tumour microenvironment. This could provide new insights into the integration of immunotherapy and targeted therapy. Conclusion This study investigated the role of S100A10 in LUAD. The results indicated that high S100A10 expression correlates with a poor prognosis. In vitro and in vivo functional assays revealed that S100A10 primarily regulates EMT via the S100A10-JUND-TNC regulatory axis. This promoted LUAD cell proliferation, colony formation, invasion and migration, and accelerated tumour growth and metastasis. Notably, the small-molecule inhibitor [D-Leu-4]-OB3 binds directly to S100A10 and promotes its degradation, thereby suppressing EMT, reducing LUAD cell proliferation, and inhibiting malignant progression of LUAD. This suggested that S100A10 is a key oncogenic driver, and that [D-Leu-4]-OB3 has potential as a targeted therapeutic agent for LUAD. Consent for publication Not applicable. Ethics approval and consent to participate All animal experiments were conducted in accordance with the guidelines for laboratory animal welfare and ethics. The study protocol was approved by the Institutional Animal Care and Use Committee (IACUC) of the Center for Animal Experiment at Wuhan University (Approval No ZN2025035). All methods were carried out in compliance with the relevant regulations and guidelines. Declaration of generative AI use Generative AI and AI-assisted technologies were NOT used in the preparation of this work. Funding This work was supported by the 10.13039/501100001809National Natural Science Foundation of China [grant number 82172841]; Zhongnan Hospital of 10.13039/501100007046Wuhan University [grant number ZNJC202229]; and Zhongnan Hospital of Wuhan University [grant number PTXM2025004]. CRediT authorship contribution statement Zhixin Zheng: Writing – review & editing, Writing – original draft, Funding acquisition, Formal analysis. Anjie Peng: Methodology, Investigation. Ziyao Xi: Validation, Supervision. Haibin Ou: Validation, Supervision. Jiong Liu: Project administration, Methodology. Aoqing Ye: Writing – original draft, Software. Jing Shao: Visualization, Investigation. Feng Li: Writing – review & editing, Validation. Yu Liu: Data curation, Conceptualization. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.