TRIM26 as a dual regulator of ferroptosis and chemoresistance in gastric cancer through HSF1 ubiquitination and exosomal miR-24-3p signaling.

Introduction
Gastric cancer (GC) is the fourth most prevalent malignancy worldwide and remains one of the leading causes of cancer-related mortality, accounting for approximately 1.1 million new cases and over 700,000 deaths annually [[1], [2], [3]]. Despite advances in surgical interventions, targeted therapies, and immunotherapy, the prognosis for GC patients remains dismal, with a five-year survival rate lingering around 20 % [4]. This poor outcome is largely attributable to late-stage diagnosis, limited therapeutic options, high recurrence rates, and widespread metastasis. Additionally, both intrinsic and acquired resistance to chemotherapy, especially platinum-based regimens such as cisplatin combined with taxanes or fluoropyrimidines, pose major obstacles to effective treatment, with nearly 80 % of patients eventually experiencing relapse due to drug resistance and tumor progression [[5], [6], [7], [8], [9], [10]]. These clinical challenges underscore the urgent need to identify novel molecular regulators that drive GC pathogenesis and therapeutic resistance.
In recent years, the ubiquitin-proteasome system (UPS) has emerged as a central node in cancer biology, governing essential cellular processes such as proliferation, apoptosis, DNA repair, and stress responses [11,12]. Among its key players, the tripartite motif-containing (TRIM) family of E3 ubiquitin ligases has gained increasing attention for its diverse roles in oncogenic signaling and tumor suppression. TRIM proteins orchestrate selective protein ubiquitination and degradation, thereby modulating pathways involved in cell cycle regulation, migration, invasion, immune evasion, and chemoresistance [[13], [14], [15]]. Notably, TRIM26 has been identified as a tumor suppressor in various cancers, including non-small cell lung cancer, hepatocellular carcinoma, and glioblastoma, where it regulates survival and metastatic behavior by targeting key oncogenic substrates [[16], [17], [18], [19], [20]]. However, the role of TRIM26 in GC remains largely unexplored, and its contribution to chemoresistance and tumor–stromal interactions is poorly understood.
Ferroptosis, a regulated, iron-dependent form of non-apoptotic cell death characterized by lipid peroxidation and reactive oxygen species (ROS) accumulation, has recently emerged as a crucial mechanism in overcoming chemotherapy resistance [21,22]. Unlike apoptosis, which many tumor cells evade through genetic mutations, ferroptosis offers an alternative avenue for eradicating therapy-resistant cancer cells [[23], [24], [25]]. Inducing ferroptosis has therefore garnered considerable interest as a therapeutic strategy in GC. Intriguingly, emerging data suggest that TRIM26 may influence ferroptotic signaling, although its mechanistic role in this context remains elusive. Given that ferroptosis is tightly regulated by oxidative stress, glutathione metabolism, and intracellular iron levels, deciphering how TRIM26 interfaces with this pathway could provide important insights into its tumor-suppressive functions and therapeutic potential in GC.
In this study, we investigate the multifaceted role of TRIM26 in gastric cancer, with a particular focus on its regulation of ferroptosis, influence on chemoresistance, and interaction with the tumor microenvironment. Our findings reveal that TRIM26 promotes ferroptosis by facilitating the ubiquitin-mediated proteasomal degradation of heat shock factor 1 (HSF1), a known regulator of antioxidant responses. This action leads to reduced intracellular glutathione (GSH), elevated ROS, and increased malondialdehyde (MDA) levels, hallmarks of ferroptosis. Moreover, we demonstrate that TRIM26 loss promotes malignant progression and resistance to cisplatin. We also uncover a previously unrecognized stromal–tumor interaction mediated by cancer-associated fibroblasts (CAFs), which secrete exosomes enriched with miR-24–3p This microRNA directly targets the 3′ untranslated region (3′ UTR) of TRIM26, leading to its post-transcriptional suppression. This stromal signaling axis significantly contributes to TRIM26 downregulation in the tumor microenvironment and facilitates the development of chemoresistance. These findings provide new evidence for the functional impact of CAF-derived exosomal miRNAs in modulating tumor suppressor pathways in GC [[26], [27], [28], [29], [30], [31], [32], [33]].
Furthermore, by employing clinically relevant cisplatin-resistant GC cell models (MGC803/DDP and AGS/DDP), we demonstrate that prolonged cisplatin exposure leads to marked suppression of TRIM26 expression, which correlates with a 5.6-fold increase in IC50 values and enhanced metastatic capacity. TRIM26 silencing promotes epithelial–mesenchymal transition (EMT) and invasiveness, whereas TRIM26 restoration re-sensitizes resistant cells to cisplatin, suggesting a critical role for TRIM26 in regulating drug sensitivity through EMT-dependent and independent mechanisms.
Together, our findings position TRIM26 as a pivotal regulator of ferroptosis and chemoresistance in GC, driven by both cell-intrinsic mechanisms and microenvironmental cues. By elucidating the interplay between TRIM26, ferroptotic signaling, stromal-derived miRNAs, and chemotherapy response, this study highlights TRIM26 as a compelling therapeutic target and prognostic marker in gastric cancer.

Materials and methods

Data acquisition and clinical samples
Gene expression profiles were obtained from The Cancer Genome Atlas Stomach Adenocarcinoma (TCGA-STAD) cohort (https://portal.gdc.cancer.gov/) and the Gene Expression Omnibus (GEO) datasets GSE54129 and GSE146996 (https://www.ncbi.nlm.nih.gov/geo/). Raw expression data were normalized using the log2 transformation in R (version 4.2.1) following standard preprocessing steps. Homogeneity of variance and normality were assessed for each dataset. Differential expression analyses were conducted using the Mann–Whitney U test for the GEO datasets, and paired t-tests were employed for the TCGA-STAD cohort.
Additionally, 124 paired gastric cancer (GC) tissue samples and adjacent normal gastric tissues were collected from patients undergoing surgical resection for primary GC between November 2020 and December 2021. All patients provided written informed consent, and the study adhered to the principles outlined in the Declaration of Helsinki. Ethical approval was granted by an international and institutional bioethics committee. The tissue samples were immediately snap-frozen in liquid nitrogen and stored at –80 °C until use. Detailed clinicopathological data were recorded for all patients, including age, sex, tumor-node-metastasis (TNM) staging, histopathological classification, and prior treatment history. Inclusion criteria required histologically confirmed primary GC with no history of chemotherapy or radiotherapy. Patients with concurrent malignancies, autoimmune disorders, or severe systemic conditions were excluded to avoid confounding variables. Comprehensive clinical characteristics are summarized in Table 1.

Cell lines, culture conditions, and induction of chemoresistance
Human gastric cancer cell lines (MGC803, AGS, HGC27, MKN45) and the normal gastric epithelial cell line GES-1 were obtained from the American Type Culture Collection (ATCC). All cell lines were cultured in DMEM supplemented with 10 % fetal bovine serum (FBS), 2 mM l-glutamine, 1 mM sodium pyruvate, and 1 % penicillin-streptomycin (Gibco, Thermo Fisher Scientific, USA). Cultures were maintained at 37 °C in a humidified incubator with 5 % CO2. Medium was refreshed every 2–3 days, and cells were routinely passaged at 70–80 % confluence. For gene modulation experiments, transfection was performed using Lipofectamine 3000 (Thermo Fisher Scientific, USA) following the manufacturer’s protocol. Cells were transfected with 120 nM of miR-24–3p mimics, miR-24–3p inhibitors, or respective negative controls (Guangzhou RiboBio, China) for 8 h, followed by incubation in complete medium for 24–48 h prior to analysis. To explore the effect of HSF1 inhibition on TRIM26-mediated ferroptosis and chemoresistance, cells were treated with KRIBB11 (Sigma-Aldrich, Cat. SML0534), a specific inhibitor of heat shock factor 1 (HSF1) transcriptional activity. KRIBB11 was dissolved in DMSO and added to cell culture media at a final concentration of 10 μM, 24 h after seeding. Control groups were treated with 0.1 % DMSO to account for vehicle effects. Cells were harvested at 24- and 48-hours post-treatment for RNA extraction, protein analysis, ROS quantification, and ferroptosis-related assays. This treatment allowed us to pharmacologically mimic TRIM26-mediated degradation of HSF1 and validate its downstream impact on ferroptotic signaling pathways.
To generate cisplatin-resistant models, MGC803 and AGS cells were gradually exposed to increasing concentrations of cisplatin (Sigma-Aldrich, USA) over a three-month period, as described previously [34]. The resulting resistant cell lines, designated MGC803/DDP and AGS/DDP, were maintained in medium containing 1 μg/mL cisplatin and cultured in drug-free medium for 48 h before use in experiments to avoid acute drug effects.

Transfection reporter assay
Potential targets of miR-24–3p were predicted using a combination of bioinformatic algorithms, including PITA (https://genie.weizmann.ac.il/), RNA22 (https://cm.jefferson.edu/rna22/), PicTar (https://pictar.mdc-berlin.de/), TargetScan (https://www.targetscan.org/), and miRanda (https://www.mirdb.org/). To experimentally validate the predicted interaction between miR-24–3p and the 3′-untranslated region (3′-UTR) of TRIM26, a dual-luciferase reporter assay was conducted. The wild-type (WT) TRIM26 3′-UTR sequence, containing the putative miR-24–3p binding site, was cloned into the pMIR-reporter vector using HindIII and SpeI restriction sites. A corresponding mutant (MUT) construct, with specific point mutations introduced into the miR-24–3p seed sequence region, was generated by site-directed mutagenesis. The target sequence was excised with restriction endonucleases and ligated into the pMIR-reporter plasmid using T4 DNA ligase. Both WT and MUT constructs were sequence-verified and co-transfected with miR-24–3p mimics into gastric cancer (GC) cells using Lipofectamine 3000. After 60 h of transfection, cells were lysed, and luciferase activity was quantified using a dual-luciferase assay kit (ShengGong Biological Technology, Shanghai, China). Luciferase activity was normalized to Renilla luciferase to control for transfection efficiency.

In situ hybridization
To localize and visualize miR-24–3p expression, fluorescence in situ hybridization (FISH) was performed using 3′- and 5′-digoxigenin-labeled locked nucleic acid (LNA) probes complementary to miR-24–3p (Sigma, USA). Although the primary focus of the study was gastric cancer, spinal cord samples were used for methodological optimization. Tissue sections were initially pre-treated to remove any residual paraffin and fixed appropriately. Following this, the probe hybridization step was performed on a FISH system with denaturation at 80 °C for 15 min, followed by overnight hybridization at 42 °C. The next day, tissues were sequentially washed at 25 °C and 75 °C for 5 min each to remove nonspecifically bound probes. After drying, DAPI (4′,6-diamidino-2-phenylindole) was applied dropwise to counterstain the nuclei. Slides were examined under a fluorescence microscope (Olympus, Japan), and images were captured for analysis. The specificity and localization of miR-24–3p signals were interpreted by comparing probe-treated sections to negative controls lacking probes.

TRIM26 overexpression and knockdown
To elucidate the functional significance of TRIM26 in gastric cancer, both overexpression and knockdown strategies were employed. For TRIM26 overexpression, MGC803 cells were transfected with a pcDNA3.1-TRIM26 expression plasmid (Addgene-85,133) using Lipofectamine 3000 (Invitrogen, Thermo Fisher Scientific), following the manufacturer’s recommended protocol. Cells transfected with empty pcDNA3.1 served as vector controls. For gene silencing, ON-TARGETplus SMARTpool siRNAs targeting TRIM26 (Dharmacon, Horizon Discovery) were transfected into MGC803/DDP, AGS/DDP, MGC803, and AGS cells using DharmaFECT 2 reagent in serum-free medium. To investigate the downstream regulatory relationship between TRIM26 and HSF1, parallel knockdown experiments were conducted using siRNAs targeting HSF1 (Invitrogen, Thermo Fisher Scientific). Scrambled siRNA sequences were used as negative controls in all knockdown experiments. Transfected cells were harvested 48 h post-transfection. Knockdown and overexpression efficiencies were validated by quantitative real-time PCR (qRT-PCR) and Western blotting, confirming modulation of TRIM26 expression at both mRNA and protein levels relative to control conditions.

CCK-8 assay
Cell viability was evaluated using the Cell Counting Kit-8 (CCK-8) assay (Biyuntian Biotechnology Co., Ltd., China), following the manufacturer’s protocol. Gastric cancer (GC) cells were seeded into 96-well plates at a density of 1 × 105 cells per well and incubated overnight to allow for adherence. After 24 h, 10 μL of CCK-8 reagent was added to each well, and the plates were incubated at 37 °C for 4 h. Absorbance was measured at 450 nm using a microplate reader (Bio-Rad, Hercules, CA, USA). All experiments were performed in triplicate, and data were presented as mean ± standard deviation (SD) from three independent experiments.

In vitro ubiquitination assay for TRIM26-Mediated ubiquitination of HSF1
To determine whether TRIM26 directly mediates the ubiquitination of HSF1, an in vitro ubiquitination assay was performed using purified recombinant proteins. His-tagged HSF1 and FLAG-tagged TRIM26 proteins were purchased from Abcam and Sino Biological, respectively. The assay was carried out using a commercial ubiquitination kit (Boston Biochem), with slight modifications to accommodate TRIM26 as the E3 ubiquitin ligase. The reaction mixture (30 µL total volume) consisted of 100 ng E1 ubiquitin-activating enzyme, 250 ng E2 conjugating enzyme (UbcH5c), 500 ng ubiquitin (Sigma-Aldrich), 200 ng recombinant HSF1, and 200 ng TRIM26, all incubated in ubiquitination buffer (50 mM Tris–HCl, pH 7.5; 2 mM ATP; 5 mM MgCl₂; 0.6 mM DTT) at 37 °C for 2 h. Negative control reactions lacking TRIM26 or E1 enzyme were included to assess reaction specificity. Following incubation, reactions were terminated by boiling in SDS sample buffer, and protein products were resolved by SDS-PAGE, followed by immunoblotting. Anti-HSF1 antibody (Cell Signaling Technology, #12,972) was used to detect native and ubiquitinated HSF1. Anti-ubiquitin antibody (Santa Cruz Biotechnology, sc-8017) was employed to confirm polyubiquitination.

Colony formation assay
The effect of TRIM26 and HSF1 on the proliferative capacity of gastric cancer cells was assessed using a colony formation assay. Experimental and control groups were established according to the study design. Cells were seeded at 500 cells per well into six-well plates and cultured under standard conditions for 12 days, with media refreshed every three days. After incubation, cell colonies were fixed with 4 % paraformaldehyde, stained with 0.1 % crystal violet, and photographed. The number and size of colonies were quantified to evaluate the impact of TRIM26 overexpression or knockdown and HSF1 modulation on long-term cell survival and clonogenicity.

CAFs and exosome isolation and transmission electron microscopy (TEM)
Cancer-associated fibroblasts (CAFs) and normal fibroblasts (NFs) were isolated from gastric cancer (GC) tumor and adjacent non-tumorous tissues, respectively, using the markers α-SMA, FAP, and FSP1 for differential identification. After culturing for 48 h, conditioned media were collected for exosome extraction. Exosomes were isolated via sequential differential centrifugation. Briefly, supernatants were centrifuged at 300 × g, 2000 × g, and 10,000 × g for 10 min each to remove cells, debris, and larger vesicles. The clarified supernatant was then ultracentrifuged at 100,000 × g for 70 min to pellet exosomes. The exosomal pellet was resuspended in PBS and recentrifuged under the same ultracentrifugation conditions for further purification. Exosome identity was verified by Western blotting for canonical exosomal markers, including CD63, Tsg101, and ALIX. The morphology of exosomes was visualized using transmission electron microscopy (TEM). For TEM analysis, purified exosomes were stained with 2 % phosphotungstic acid for 5 min and observed under a transmission electron microscope to confirm vesicular structure and size.

MTT assay
All cell lines were seeded in a 96-well plate at a density of 5 × 104 cells per well and treated with cisplatin at concentrations of 1, 2, 5, 10, 20, 40, 60, 80, 100, 200, 400, and 600 µM, or left untreated, for 48 h. The MTT assay was performed as previously described [35]. Cell viability was expressed as the percentage of cytotoxicity, calculated using the following formula:
The IC25 and IC50 values for each cell line were determined by plotting a dose-response curve using GraphPad Prism 9 for Windows. The curve was fitted using a variable slope model, and IC50 was calculated according to the formula:

BrdU cell proliferation assay
Cells were seeded in 96-well plates at a density of 1 × 105 cells per well and treated with cisplatin at concentrations corresponding to previously determined IC25 or IC50 values, or left untreated. After 48 h of incubation at 37 °C in a 5 % CO2 atmosphere, the medium was replaced, and a BrdU proliferation assay (Sigma-Aldrich, Merck) was performed according to the manufacturer’s instructions. Absorbance was measured at dual wavelengths of 450 nm and 595 nm using a Labsystems Multiskan RC spectrophotometric plate reader (Thermo Fisher Scientific). The proliferation rate was expressed as the percentage of BrdU-positive cells, indicating DNA synthesis and active cell proliferation.

Wound healing assay
To evaluate cell migratory capacity, GC cells were seeded into 24-well plates at a density of 5 × 105 cells per well and cultured until a confluent monolayer was achieved. A mechanical wound (scratch) was made through the center of each well using a sterile pipette tip. Detached cells were removed by gently washing the wells twice with PBS containing 1 % FBS. Fresh medium supplemented with 1 % FBS was added to minimize proliferation and focus on migration. Wound closure was monitored in real time using the SPARK multimode reader (Tecan, Switzerland), equipped with a phase-contrast microscope and CO2 incubation chamber. Time-lapse imaging was conducted over a 24-hour period (12 intervals at 2-hour intervals). Wound areas were quantified using ImageJ software, and cell migration was determined by measuring the reduction in wound width at 24 h relative to the initial gap at 0 h.

Transwell invasion assay
Cell invasive ability was evaluated using Corning® HTS Transwell®−24-well permeable supports (Corning, USA). Serum-free medium was added to the upper chamber, while medium supplemented with 10 % fetal bovine serum (FBS) was placed in the lower chamber to act as a chemoattractant. Transfected and serum-starved cells were seeded into the upper chamber and incubated at 37 °C in a 5 % CO2 for 24 h. Following incubation, non-invading cells on the upper side of the membrane were carefully removed. The membranes were fixed and stained with 0.1 % crystal violet (Kaigen, China) for 15 min. Cells that had invaded to the underside of the membrane were visualized under a light microscope and counted in five randomly selected fields per well to assess invasion efficiency.

Invasion through extracellular matrix (ECM) membranes
To further evaluate the invasive potential of gastric cancer cells, we performed an ECM-based invasion assay using the QCM™ 24-Well Cell Invasion Assay Kit (Sigma-Aldrich, Merck) following the manufacturer’s protocol. Briefly, cells were serum-starved for 18 h in serum-free medium, washed twice with PBS, and detached using the provided Harvesting Buffer. After centrifugation at 300 × g for 5 min, cells were resuspended in Quenching Medium (serum-free DMEM containing 5 % BSA). Cells were seeded at a density of 5 × 105 cells per insert into the upper chamber of the ECM-coated invasion inserts. The lower chamber contained either serum-free medium (control) or medium supplemented with 10 % FBS as a chemoattractant. Plates were incubated at 37 °C in a 5 % CO₂ for 48 h. Following incubation, the inserts were transferred into clean wells containing Cell Detachment Buffer and incubated for 30 min. The detached cells were subsequently mixed with Lysis/Dye buffer for 15 min at room temperature. The resulting fluorescent signal was transferred to a 96-well plate and measured using a fluorescence plate reader (Fluoroskan Ascent FL, Thermo Fisher Scientific) with an excitation/emission filter set of 480/520 nm. Results were expressed as relative fluorescence units (RFU) to indicate invasion efficiency.

Ferroptosis, oxidative stress markers, and GPX4 activity assay
Total glutathione (GSH) levels were quantified using a commercial detection kit (Beyotime, S0052) at 412 nm. Lipid peroxidation was assessed by measuring malondialdehyde (MDA) levels (Beyotime, S0131S) at 532 nm. Reactive oxygen species (ROS) production was evaluated using the DCFH-DA probe (Beyotime, S0033M) and quantified using a fluorescence microplate reader. GPX4 activity was analyzed using a commercial cellular glutathione peroxidase assay kit (GPx Activity Kit; ab102530), according to the manufacturer’s instructions. Finally, the absorbance of 340 nm was measured by a microplate reader (Thermo Fisher Scientific, USA).

qRT-PCR analysis
Total RNA was extracted using TRIzol reagent (Invitrogen) and reverse transcribed into cDNA using PrimeScript™ RT Master Mix (Takara, Japan). qRT-PCR was performed using SYBR® Green (Thermo Fisher Scientific) following the manufacturer's instructions. GAPDH was used as an internal control, and relative gene expression was calculated using the 2-∆CT method [36].

Western blot analysis
Protein lysates were prepared using RIPA buffer, and total protein concentration was determined using the BCA assay. Protein samples were resolved on 10–12 % SDS-PAGE gels and transferred onto PVDF membranes. Membranes were blocked with 5 % non-fat milk containing 0.05 % Tween-20 and incubated overnight at 4 °C with primary antibodies. After washing, membranes were incubated with secondary antibodies for 2 h at room temperature. Protein bands were visualized using enhanced chemiluminescence (ECL) reagents and quantified using ImageJ software.

Flow cytometry
Sample preparation and cell staining have been performed as described before [24]. 1 × 105 cell/100 µl of PBS with 2 mM EDTA and 0.5 % BSA were stained with mouse monoclonal antibodies anti-E-cadherin-PE or anti-N-cadherin-Alexa Fluor 488 or anti-Vimentin-PE (Beckton Dickinson, USA) or the appropriate mouse isotype IgG1κ control (Beckton Dickinson and R&D) and analyzed using a FACS BD LSR II flow cytometer (Beckton Dickinson) equipped with BD FACS Diva Software. The results were analyzed with FlowJo v10.8.1 software and are presented as the median fluorescence intensity (MFI), which reflects the surface expression of the target proteins.

Immunofluorescence staining and confocal microscopy
Cells were seeded (5 or 7 × 104 cells/well) on permanox plastic 8-well chamber slides (Nunc™ Lab-Tek™ Chamber Slide, Thermo Fisher Scientific) and incubated in growth medium for 24 h to adhere. For immunofluorescence staining, cells, before or 48 h after transfection with siRNA, were fixed by a 15-min incubation with 4 % paraformaldehyde, washed four times with PBS (Thermo Fisher Scientific) and incubated for 1 hour with an open/block buffer (0,2 % Triton X-100 / 5 % BSA / PBS). TRIM26 was stained by over-night incubation of cells in staining buffer (0,2 % Triton X-100 / 3 % BSA/PBS) with anti-TRIM26 mouse antibody in humidity chamber at 4 °C. After triple wash with PBS cells were stained for 1 hour with Alexa Fluor-488-conjugated, secondary, anti-mouse (AB_2,534,106) antibody, in the dark at room temperature. Cell membranes and nuclei were visualized by a 10 min incubation in the dark with 2 μM Alexa-Fluor 594-conjugated wheat germ agglutinin (WGA) and 5μg/ml Hoechst 33,342 (in 3 % BSA in PBS) respectively. Stained cells were visualized using a confocal microscope (Nikon d-Eclipse C1) and analyzed with EZ-C1 version 3.6 software (Nikon, Japan).

Statistical analysis
Statistical evaluation of obtained results has been performed using data from either four or five different experiments, depending on the chosen statistical test. RT-PCR data were calculated with t-test with the use of DataAssist 3.01 software. The rest of results were firstly analyzed with Shapiro-Wilk test to assess the normality of distribution and according to these results, the nonparametric Wilcoxon’s singed rank test or Mann-Whitney U test or student’s t-test were applied with the use of Graphpad 9.5.1 for Windows. Survival rates of GC patients were calculated by Log-rank analysis and Kaplan-Meier, and prognostic factors were assessed using COX regression model. Statistical significance was defined as p ≤ 0.05.

Results

TRIM26 is downregulated in gastric cancer tissues and cells
To explore the potential involvement of TRIM26 in gastric cancer (GC), we analyzed gene expression profiles from three independent datasets: TCGA-STAD, GSE54129, and GSE146996. Among the intersecting differentially expressed genes, TRIM26 emerged as a uniquely and consistently downregulated gene in GC tissues compared to adjacent normal tissues across all datasets (Fig. 1A). To further assess its clinical relevance, we examined the association between TRIM26 expression and various clinicopathological features in a cohort of 124 GC patients. As summarized in Table 1, low TRIM26 expression was significantly associated with advanced TNM stage, lymph node metastasis, and increased depth of tumor invasion. However, no significant correlation was observed with tumor location, tumor size, histological differentiation, patient age, or gender. Kaplan-Meier survival analysis demonstrated that GC patients with low TRIM26 expression exhibited significantly poorer overall survival compared to those with high expression levels (Fig. 1B). This prognostic implication was further validated using data from the Kaplan-Meier Plotter online tool (Fig. 1C). Moreover, univariate Cox regression analysis confirmed that TRIM26 expression functions as an independent prognostic factor for gastric cancer outcomes (Fig. 1D). To validate these findings at the molecular level, we measured TRIM26 expression in GC tissues. Both mRNA and protein levels of TRIM26 were significantly downregulated in GC tissues relative to matched normal gastric tissues (Fig. 1E–F). Similarly, in vitro analyses revealed markedly lower TRIM26 expression in various GC cell lines compared to the normal human gastric epithelial cell line GES-1 at both the mRNA and protein levels (Fig. 1G–H). Collectively, these data suggest that TRIM26 downregulation is a common and clinically relevant feature in GC, implicating it as a potential tumor suppressor gene.

TRIM26 suppresses GC cell proliferation, migration, and invasion
To elucidate the functional role of TRIM26 in gastric cancer, we conducted gain- and loss-of-function experiments in MGC-803 GC cells by transfecting them with TRIM26-targeting siRNA or overexpression plasmids. The transfection efficiency was confirmed at the mRNA and protein levels (Fig. 2A). Functional assays revealed that TRIM26 knockdown significantly increased cell viability (Fig. 2B), enhanced migratory capacity (Fig. 2C), and promoted invasion (Fig. 2D) in MGC-803 cells. Conversely, TRIM26 overexpression markedly suppressed these malignant phenotypes, as demonstrated by reduced cell viability, migration, and invasion (Figs. 2B–D). The results of the colony formation assay further supported the tumor-suppressive role of TRIM26. Cells overexpressing TRIM26 formed fewer and smaller colonies compared to control cells, indicating impaired clonogenic potential (Fig. 2E). At the molecular level, TRIM26 modulation influenced the expression of epithelial–mesenchymal transition (EMT)-related markers. Specifically, E-cadherin expression was downregulated in TRIM26-deficient cells, while N-cadherin and Vimentin levels were upregulated, consistent with a mesenchymal and invasive phenotype. Conversely, TRIM26 overexpression restored E-cadherin expression and suppressed N-cadherin and Vimentin levels (Fig. 2F). Together, these findings provide compelling evidence that TRIM26 functions as a tumor suppressor by inhibiting GC cell proliferation, migration, invasion, and EMT, thereby contributing to the attenuation of gastric cancer progression.

TRIM26 regulates ferroptosis via HSF1 ubiquitination and degradation
Given the increasing evidence linking ferroptosis to tumor suppression, we investigated whether TRIM26 influences ferroptosis in gastric cancer (GC) cells. Treatment with erastin, a known ferroptosis inducer, significantly reduced GC cell viability. However, silencing TRIM26 using siRNA effectively rescued the viability of erastin-treated cells, suggesting a suppressive effect of TRIM26 knockdown on ferroptosis (Fig. 3A). Further biochemical analyses revealed that TRIM26 knockdown attenuated oxidative stress, as evidenced by reduced ROS and malondialdehyde (MDA) levels and elevated glutathione (GSH) in response to erastin (Fig. 3B–D). At the molecular level, TRIM26 silencing upregulated the ferroptosis-inhibitory protein GPX4, while downregulating COX-2 and ACSL4, two markers associated with pro-ferroptotic activity (Fig. 3E). These results collectively support the conclusion that TRIM26 promotes ferroptosis, and its depletion suppresses this process in GC cells. To elucidate the molecular mechanism by which TRIM26 regulates ferroptosis, we examined its interaction with Heat Shock Factor 1 (HSF1), a known inhibitor of ferroptosis. Surprisingly, quantitative RT-PCR revealed that TRIM26 manipulation had no effect on HSF1 mRNA levels, regardless of whether TRIM26 was overexpressed or silenced (Fig. 3F). This suggested a post-transcriptional regulatory mechanism. Given that TRIM26 functions as an E3 ubiquitin ligase, we hypothesized that it may regulate HSF1 through ubiquitin-mediated protein degradation. To test this, we conducted an in vitro ubiquitination assay using purified recombinant proteins. The results confirmed that TRIM26 directly catalyzes the polyubiquitination of HSF1, as polyubiquitinated HSF1 species were only observed in reactions containing all necessary components: E1, E2, ubiquitin, and TRIM26 (Fig. 3G). Omission of either TRIM26 or E1 abolished this effect, demonstrating specificity. Immunoblotting with anti-HSF1 and anti-ubiquitin antibodies confirmed the presence of ubiquitinated HSF1 species. Consistent with these findings, TRIM26 overexpression significantly increased HSF1 ubiquitination levels in cells (Fig. 3H), while silencing TRIM26 led to elevated HSF1 protein levels, and conversely, TRIM26 overexpression reduced HSF1 protein levels (Fig. 3I). These results support a model in which TRIM26 functions as a negative regulator of HSF1 via ubiquitination-mediated degradation. To confirm the functional impact of TRIM26 and HSF1 signaling on ferroptosis, we again measured ferroptosis-related indicators. TRIM26 knockdown resulted in decreased ROS and MDA, and increased GSH levels, consistent with ferroptosis suppression (Fig. 3J–L). Additionally, GPX4 enzymatic activity was significantly elevated in TRIM26-deficient cells (Fig. 3M), alongside upregulated GPX4 protein expression and downregulation of COX-2 and ACSL4 (Fig. 3N). Importantly, silencing HSF1 reversed the anti-ferroptotic effects of TRIM26 knockdown, including normalization of ROS, MDA, and GSH levels, and restoration of ferroptosis-related protein expression (Fig. 3J–N). Taken together, these results demonstrate that TRIM26 promotes ferroptosis in GC cells through the degradation of HSF1, thereby acting as a key regulator in the ferroptotic signaling pathway.

TRIM26 modulates GC progression by regulating HSF1-mediated ferroptosis
To further explore whether the tumor-suppressive effects of TRIM26 are mediated through HSF1-dependent ferroptosis, we assessed the phenotypic consequences of TRIM26 knockdown in gastric cancer cells. As shown in Fig. 4A–C, TRIM26 silencing significantly promoted cell proliferation, migration, and invasion, suggesting enhanced tumorigenic potential. However, co-silencing of HSF1 effectively rescued these malignant phenotypes, reversing the aggressive behavior induced by TRIM26 knockdown. At the molecular level, TRIM26 knockdown led to downregulation of E-cadherin, a key epithelial marker, and upregulation of N-cadherin and Vimentin, markers associated with epithelial–mesenchymal transition (EMT), further supporting the role of TRIM26 in suppressing tumor progression (Fig. 4D). To directly assess the functional relevance of HSF1 in TRIM26-mediated ferroptosis, we utilized KRIBB11, a selective small-molecule inhibitor of HSF1 transcriptional activity. Treatment of GC cells with KRIBB11 resulted in significant downregulation of HSF1 protein levels and its transcriptional targets (Fig. 4E). Notably, KRIBB11 phenocopied the effects of TRIM26 overexpression by increasing ferroptotic stress, as demonstrated by decreased GSH, elevated ROS, and increased MDA levels (Fig. 4F). These findings collectively demonstrate that TRIM26 exerts its tumor-suppressive functions by promoting HSF1 degradation, which in turn enhances ferroptosis. Thus, the TRIM26–HSF1 axis represents a critical regulatory pathway in GC progression, offering new insights into the mechanisms underlying tumor suppression via ferroptosis and providing a potential target for therapeutic intervention.

miR-24–3p targets TRIM26 and negatively regulates its expression
Quantitative RT-PCR analysis revealed that among the four candidate miRNAs assessed, miR-24–3p was significantly upregulated in gastric cancer (GC) tissues compared to adjacent normal tissues (Fig. 5A). This differential expression was further validated using fluorescence in situ hybridization (FISH), which confirmed increased miR-24–3p levels in GC tissues (Fig. 5B). These findings suggest a potential pathological role for miR-24–3p in GC. To identify upstream regulators of TRIM26, TargetScan was employed, which predicted miR-24–3p as a potential post-transcriptional regulator (Fig. 5C). To validate this interaction, a luciferase reporter assay was conducted. Co-transfection of miR-24–3p mimics with the wild-type TRIM26 3′UTR reporter led to a significant reduction in luciferase activity, whereas no effect was observed in the mutant TRIM26 reporter group (Fig. 5D), confirming the direct binding of miR-24–3p to the TRIM26 3′UTR. Further qRT-PCR experiments demonstrated that miR-24–3p mimics markedly increased miR-24–3p expression, while miR-24–3p inhibitors effectively reduced its levels (Fig. 5E). Consistently, GC cells exhibited elevated miR-24–3p expression compared to the normal gastric epithelial cell line GES-1 (Fig. 5F). Importantly, transfection with miR-24–3p mimics led to a significant downregulation of TRIM26, while inhibition of miR-24–3p resulted in its upregulation (Fig. 5G). These findings collectively demonstrate that miR-24–3p directly targets and suppresses TRIM26 expression, highlighting its regulatory role in GC pathogenesis.

CAF-derived exosomal miR-24-3p mediates TRIM26 downregulation in GC
Cancer-associated fibroblasts (CAFs) were phenotypically distinct from normal fibroblasts (NFs), exhibiting markedly higher expression of α-SMA, FAP, and FSP-1, confirming their activated state (Fig. 6A, B). Exosomes isolated from CAFs showed an enrichment of canonical exosomal markers, including CD63, TSG101, and Alix, compared to NF-derived exosomes (NFs-Ex) (Fig. 6C). Transmission electron microscopy (TEM) confirmed the characteristic morphology of exosomes from both sources (Fig. 6D). Interestingly, miR-24–3p expression was significantly higher in CAFs and their secreted exosomes (CAFs-Ex) compared to NFs and their exosomes (Fig. 6E), implicating a role for CAF-derived exosomal miR-24–3p in the tumor microenvironment. To assess the functional consequences of exosomal transfer, MGC-803 GC cells were treated with CAFs-Ex after miR-24–3p inhibition. While the inhibitor effectively suppressed miR-24–3p and restored TRIM26 expression, this effect was reversed upon CAFs-Ex treatment, which led to increased miR-24–3p expression and decreased TRIM26 expression in recipient GC cells (Fig. 6F, G). These results strongly suggest that CAF-derived exosomal miR-24–3p contributes to the downregulation of TRIM26 in GC cells, establishing a novel axis of stromal-tumor communication that may promote tumor progression.

TRIM26-mediated functional changes in GC cells during cisplatin resistance acquisition
To investigate the relationship between TRIM26 expression and cisplatin resistance, we established cisplatin-resistant gastric cancer cell lines, MGC803/DDP and AGS/DDP. These resistant cells displayed significantly elevated IC50 values (MGC803/DDP: 17.78 µM vs. 3.16 µM in parental MGC803; AGS/DDP: 19.95 µM vs. 6.31 µM in parental AGS), demonstrating acquired chemoresistance (Fig. 7A). Dose-response curves for IC50 determination are shown in Fig. 7B Consistently, both MGC803/DDP and AGS/DDP cells maintained higher proliferation rates under cisplatin treatment compared to their parental counterparts (Fig. 7C).
Cisplatin-resistant GC cells exhibited features characteristic of epithelial–mesenchymal transition (EMT), including downregulation of E-cadherin and upregulation of N-cadherin and Vimentin (Fig. 7D). Functional assays further revealed enhanced migratory and invasive capacities in resistant cells, as evidenced by wound healing and transwell invasion assays (Fig. 7E, F). Notably, TRIM26 expression progressively declined in MGC803/DDP and AGS/DDP cells (Fig. 7G). Immunohistochemical analysis supported these findings, with staining intensity levels that corresponded well to TRIM26 protein expression detected by Western blotting (Fig. 7H). These results suggest that loss of TRIM26 is associated with EMT-like features and enhanced chemoresistance, indicating its potential role in regulating cisplatin sensitivity in GC.

TRIM26 silencing exacerbates chemoresistance in GC cells
To investigate the functional role of TRIM26 in cisplatin resistance, we evaluated the effects of TRIM26 knockdown in the cisplatin-resistant GC cell lines MGC803/DDP and AGS/DDP. Silencing TRIM26 significantly increased the IC50 values for cisplatin in both cell lines, indicating a further enhancement of chemoresistance (Fig. 8A). Corresponding dose–response curves are depicted in Fig. 8B In line with these findings, cell proliferation assays revealed that, when treated with cisplatin at their respective IC50 concentrations, TRIM26-silenced cells exhibited significantly higher proliferative capacity than their non-transfected counterparts (Fig. 8C), further supporting the role of TRIM26 in modulating drug sensitivity. Western blot analysis demonstrated that TRIM26 knockdown exacerbated the epithelial–mesenchymal transition (EMT) phenotype in resistant cells, as evidenced by a further reduction of E-cadherin and increased expression of N-cadherin and Vimentin (p < 0.01) (Fig. 8D). To assess the metastatic potential of TRIM26-silenced cells, we performed wound healing and Matrigel-coated transwell invasion assays. The results revealed that TRIM26 knockdown significantly enhanced both migratory and invasive capabilities of MGC803/DDP and AGS/DDP cells compared to non-silenced controls (Fig. 8E, F).
Collectively, these findings demonstrate that loss of TRIM26 function intensifies cisplatin resistance and promotes aggressive cellular behavior in GC, underscoring its critical role in regulating chemoresensitivity and metastatic potential in drug-resistant gastric cancer cells.

Discussion
Gastric cancer (GC) remains a leading cause of cancer-related mortality globally, primarily due to late-stage diagnosis, the emergence of chemoresistance, and aggressive metastatic behavior. Despite substantial progress in therapeutic strategies, overcoming drug resistance remains a major clinical obstacle [37,38]. In this study, we identify TRIM26, a member of the tripartite motif (TRIM) family of E3 ubiquitin ligases, as a critical tumor suppressor that promotes ferroptosis and inhibits epithelial–mesenchymal transition (EMT) in GC. Our findings reveal that TRIM26 depletion stabilizes heat shock factor 1 (HSF1), thereby enabling ferroptosis resistance and metastatic progression. Importantly, we uncover a novel tumor–stroma interaction in which cancer-associated fibroblasts (CAFs) secrete exosomal miR-24–3p to post-transcriptionally silence TRIM26 in tumor cells. This miR-24–3p/TRIM26/HSF1 axis drives chemoresistance and disease progression in gastric cancer. Survival analysis using three independent cohorts (TCGA-STAD, GSE54129, and GSE146996) demonstrated that low TRIM26 expression is significantly associated with poor overall survival, whereas high levels of HSF1 and miR-24–3p correlate with worse clinical outcomes. These results support the clinical relevance of our mechanistic findings and underscore the prognostic value of the TRIM26–HSF1–ferroptosis axis in GC. Ferroptosis has emerged as a distinct form of regulated cell death with profound implications for overcoming therapeutic resistance in various malignancies, including GC [39,40]. Ferroptosis is a distinct form of programmed cell death characterized by iron-dependent lipid peroxidation and reactive oxygen species (ROS) accumulation, both of which are tightly regulated by cellular antioxidant systems [21,22]. Tumors often develop mechanisms to suppress ferroptosis, facilitating survival under therapeutic pressure [41]. Our results establish TRIM26 as a potent ferroptosis inducer in GC through its ability to ubiquitinate and destabilize HSF1, a key transcriptional regulator of redox homeostasis. To validate the functional relevance of HSF1, we employed KRIBB11, a selective small-molecule inhibitor of HSF1 transcriptional activity. KRIBB11 restored ferroptotic sensitivity in TRIM26-deficient GC cells, confirming that HSF1 mediates resistance to ferroptosis in this context. These findings are consistent with prior studies linking HSF1 to antioxidant defense and drug resistance [42,43]. Mechanistically, TRIM26 knockdown led to HSF1 stabilization, which in turn elevated glutathione (GSH) levels, suppressed lipid peroxidation (as indicated by decreased malondialdehyde (MDA)), and reduced ROS production, collectively shielding cells from ferroptotic death. These observations build on earlier work suggesting that pharmacological inhibition of HSF1 sensitizes cancer cells to ferroptosis-inducing agents like erastin [44]. However, our study identifies TRIM26 as the upstream E3 ubiquitin ligase responsible for HSF1 degradation, highlighting a previously unrecognized regulatory mechanism of ferroptosis in GC. Restoration of ferroptosis sensitivity through HSF1 inhibition in TRIM26-silenced cells further reinforces the functional importance of this pathway and supports its potential as a therapeutic target. Beyond intrinsic cellular mechanisms, the tumor microenvironment (TME), and particularly CAFs, plays a vital role in promoting GC progression and resistance to therapy [[45], [46], [47], [48]]. Here, we demonstrate that CAF-derived exosomes deliver miR-24–3p to GC cells, where it binds to the 3′ untranslated region (3′UTR) of TRIM26 mRNA and suppresses its expression. This aligns with growing evidence that stromal exosomal miRNAs actively modulate tumor behavior. For instance, CAF-derived miR-522 has been shown to inhibit ALOX15-mediated lipid peroxidation and promote chemoresistance [37], while fibroblast-derived miR-21 facilitates EMT in breast cancer by regulating transcriptional programs [49]. Our findings establish miR-24–3p as a stromal-derived oncogenic miRNA (oncomiR) that impairs TRIM26-dependent ferroptosis, thereby enhancing tumor survival and chemoresistance. Notably, clinical analysis of GC patient samples revealed that miR-24–3p expression is significantly upregulated in tumor tissues and exhibits an inverse correlation with TRIM26 levels. This pattern is consistent with observations in other malignancies, such as breast cancer, where miR-24–3p has been shown to suppress apoptotic pathways [50]. Functionally, inhibition of miR-24–3p restored TRIM26 expression in GC cells and re-sensitized them to cisplatin treatment, underscoring the translational promise of targeting stromal–tumor communication via exosomal miRNAs.
Chemoresistance continues to pose a formidable obstacle in GC management, with cisplatin resistance frequently coinciding with heightened metastatic potential. In our study, prolonged exposure to cisplatin resulted in the resistant MGC803/DDP and AGS/DDP cells, which exhibited marked TRIM26 downregulation, EMT activation, and increased invasive capacity. Loss of TRIM26 correlated with upregulation of mesenchymal markers such as N-cadherin and Vimentin, coupled with a reduction in epithelial marker E-cadherin, an expression profile characteristic of EMT. These findings reinforce the notion that TRIM26 functions as an EMT suppressor. Notably, silencing TRIM26 in cisplatin-resistant MGC803/DDP and AGS/DDP cells further exacerbated chemoresistance, leading to an increase in the half-maximal inhibitory concentration (IC50) of cisplatin, while TRIM26 restoration reversed these phenotypic changes. These data align with clinical evidence indicating that TRIM26 downregulation is frequently observed in advanced-stage hepatocellular carcinoma cancer (HCC) [[51], [52], [53]].
The dual functionality of TRIM26, promoting ferroptosis and suppressing EMT, suggests that it acts as a molecular gatekeeper of tumor plasticity, integrating survival and death pathways in a context-specific manner. This may explain its divergent roles across cancer types; for instance, TRIM26 has been reported to facilitate proliferation in osteosarcoma via stabilization of oncogenic signaling pathways [18]. Such tissue-specific effects underscore the need for precision approaches in therapeutic targeting of TRIM26.
From a translational standpoint, our findings offer two major therapeutic avenues. First, restoring TRIM26 expression or targeting miR-24–3p could help overcome GC chemoresistance. Development of TRIM26-activating compounds or an inhibitor of miR-24–3p oligonucleotides, ideally delivered via tumor-targeting nanoparticles, may enhance treatment efficacy. Second, inhibition of HSF1 or disruption of CAF-derived exosome signaling represents a complementary strategy. Notably, HSP90 inhibitors, which destabilize HSF1 and its client proteins, may synergize with ferroptosis inducers to enhance antitumor responses [54]. However, due to the pleiotropic roles of TRIM26 in immune modulation, as observed in nasopharyngeal carcinoma where it affects immune cell function [55,56]. This raises concerns regarding potential off-target effects associated with systemic TRIM26 modulation. Therefore, future research should focus on the development of tissue-specific drug delivery platforms, such as exosome-mimetic nanoparticles, to selectively target tumor cells while minimizing toxicity.

Conclusion
In this study, we identified TRIM26 as a key tumor suppressor in gastric cancer (GC), regulating ferroptosis and chemoresistance through the ubiquitin-mediated degradation of HSF1. TRIM26 downregulation stabilizes HSF1, suppressing lipid peroxidation and ROS accumulation, thereby promoting tumor survival. Additionally, cancer-associated fibroblasts (CAFs) facilitate TRIM26 suppression via exosomal miR-24–3p transfer, contributing to chemoresistance. In cisplatin-resistant GC models, TRIM26 loss enhances EMT and metastatic potential, while its restoration reverses these effects. These findings highlight TRIM26 as a potential therapeutic target, suggesting that TRIM26-based therapies or ferroptosis inducers combined with chemotherapy could improve treatment outcomes in advanced GC. However, further investigations should explore TRIM26-targeted therapies in preclinical models and investigate combinatorial strategies integrating ferroptosis inducers with conventional chemotherapies to improve outcomes in advanced GC.

Funding
Not applicable.

Data availability
All data are available and the correspondent can be contacted if requested.

Ethics approval
All experimental procedures were reviewed and approved by the institutional animal care and use committee and adhered to the international guidelines for ethical conduct and complied with the ARRIVE guidelines and relevant national and institutional regulations (UD-2020-2). Animal handling and experimentation followed the protocols of the Animal Management and Welfare Committee of the University of Helsinki, Finland.

Consent to participate
Written informed consent was obtained from individual or guardian participants.

CRediT authorship contribution statement
Nouf S. Al-Abbas: Writing – original draft, Validation, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization.

Declaration of competing interest
The author declares no competing financial interests.