Targeting ABCB6 induces ferroptosis in osteosarcoma cells via HIF1A/GPX4 pathway Authors: 梓立 林, 浩 罗, 湘尧 李, 伟 罗 Published: 2026 ABSTRACT ObjectiveOsteosarcoma (OS) is a malignant bone tumor that predominantly affects adolescents. Due to its early metastatic tendency and chemoresistance, long-term survival rates in patients have not shown significant improvement over time, highlighting the urgent need for novel molecular targets and therapeutic strategies. Ferroptosis, a form of programmed cell death driven by iron-dependent lipid peroxidation, offers a promising avenue for overcoming drug-resistant tumors. ATP-binding cassette subfamily B member 6 (ABCB6) belongs to the ATP-binding cassette (ABC) transporter superfamily and is mainly localised to the outer mitochondrial membrane, where it participates in haem synthesis and intracellular iron transport. Recent studies have indicated that ABCB6 is aberrantly expressed in various malignancies and regulates tumor progression, yet its role and mechanism in OS remain unexplored. The present study aims to systematically characterise the expression pattern, clinical prognostic significance, and biological functions of ABCB6 in OS, and to dissect the molecular mechanism by which ABCB6 governs ferroptosis in OS cells, thereby providing a theoretical foundation for ABCB6-targeted therapy.MethodsHuman OS cell lines (HOS and 143B) and human bone marrow mesenchymal stem cells (BMSCs) were used as research subjects. Stable ABCB6-knockdown OS cell lines were established via lentivirus-mediated RNA interference, and rescue experiments were performed using lentiviral overexpression of hypoxia-inducible factor 1-alpha (HIF1A). Cell proliferation was assessed by cell counting kit-8 (CCK-8) and colony formation assays; migration was evaluated by Transwell and scratch wound-healing assays; apoptosis was detected by flow cytometry. Ferroptosis-related phenotypic assays were employed to determine the impact of ABCB6 expression on ferroptosis. Transcriptome sequencing was conducted to analyse differentially expressed genes and enriched pathways; chromatin immunoprecipitation (ChIP) and dual-luciferase reporter assays were applied to verify the direct binding of HIF1A to the promoter of glutathione peroxidase 4 (GPX4) and its transcriptional regulation. Iron chelator and proteasome inhibitor interventions were used to validate the regulation of HIF1A protein stability by Fe2+. In vivo, ABCB6-knockdown and control 143B cells were injected subcutaneously into nude mice to establish a xenograft tumor model, and tumor growth was monitored.ResultsBoth messenger RNA (mRNA) and protein levels of ABCB6 were significantly higher in OS cell lines than in BMSCs (both P<0.05), and analysis of the Therapeutically Applicable Research to Generate Effective Treatments (TARGET)-OS cohort revealed that patients with high ABCB6 expression had markedly worse overall survival than those with low expression [hazard ratio (HR)=2.94, 95% confidence interval (CI) 1.36 to 6.36, P<0.05]. ABCB6 knockdown significantly suppressed OS cell proliferation and colony formation, reduced migratory capacity, and increased the apoptosis rate, accompanied by downregulation of N-cadherin and upregulation of E-cadherin and cleaved caspase-3 (all P<0.05). ABCB6 silencing markedly elevated the proportion of Liperfluo-positive cells and dihydroethidium (DHE) oxidation levels, increased intracellular Fe2+ concentration, decreased the glutathione (GSH)/oxidized glutathione (GSSG) ratio, reduced mitochondrial membrane potential, and induced typical ferroptotic mitochondrial morphological changes (shrinkage and cristae reduction) as observed by transmission electron microscope (all P<0.05); these oxidative alterations were reversed by the ferroptosis inhibitors. ABCB6 knockdown downregulated GPX4 at both protein and mRNA levels, whereas solute carrier family 7 member 11 (SLC7A11) expression remained unchanged (P>0.05). Transcriptome sequencing and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis showed significant enrichment of the HIF-1 signalling pathway in the control group (P<0.001); ABCB6 knockdown decreased HIF1A protein expression, and treatment with an iron chelator or a proteasome inhibitor reversed this degradation through Fe2+ chelation or inhibition of the ubiquitin-proteasome pathway, respectively (all P<0.05). Chromatin immunoprecipitation followed by quantitative polymerase chain reaction (ChIP-qPCR) and dual-luciferase reporter assays confirmed that HIF1A directly binds to the GPX4 promoter and activates its transcription. Overexpression of HIF1A in ABCB6-knockdown cells restored GPX4 expression and significantly rescued the proliferation inhibition, migration impairment, and apoptosis increase caused by ABCB6 silencing, while also restoring mitochondrial membrane potential (all P<0.05). In the xenograft model, tumors from the ABCB6-knockdown group exhibited significantly reduced volume and weight compared with the control group (all P<0.05); immunohistochemistry confirmed downregulation of HIF1A and GPX4, reduced Ki-67 and N-cadherin positivity, and increased cleaved caspase-3 and E-cadherin positivity in the knockdown group (all P<0.05). ConclusionThis study systematically uncovers, for the first time, the oncogenic role and prognostic value of ABCB6 in OS, and delineates the complete “ABCB6/Fe2+/HIF1A/GPX4” axis that governs ferroptosis in OS cells. These findings not only provide a novel biomarker for molecular subtyping and prognostic evaluation of OS, but also establish a theoretical basis for therapeutic strategies that target the ABCB6/HIF1A/GPX4 axis to induce ferroptosis. Future investigations may explore the upstream regulators of ABCB6 and its potential utility in reversing chemoresistance in OS.Graphical Abstract Targeting ABCB6 induces ferroptosis in OS cells. ABCB6: ATP-binding cassette subfamily B member 6; OS: Osteosarcoma; HIF1A: Hypoxia-inducible factor 1-alpha; GPX4: Glutathione peroxidase 4; ROS: Reactive oxygen species. Osteosarcoma (OS) is the most common primary bone malignancy, highly aggressive and metastatic, peaking in adolescents and children[1-2]. The global annual incidence is 2-3 per million[2]. Standard therapy, including surgical resection plus neoadjuvant/adjuvant chemotherapy, has improved localized disease 5-year survival to 60%-70%, but metastatic/recurrent OS has a poor 5-year survival of 20%-30%[1-2]. Genetic exploration is crucial for improving OS treatment[3-4]; however, the tumor's genetic complexity and instability hinder mechanistic studies and therapy development, highlighting the need to define oncogene and tumor suppressor roles in OS pathogenesis[3-4]. The human ATP-binding cassette (ABC) transporter superfamily comprises 48 genes in 7 subfamilies, many aberrantly expressed in cancers and linked to progression, drug resistance, and prognosis[5]. Recent studies[6-9] have indicated that ABC subfamily B member 6 (ABCB6) is associated with poor clinical outcomes and drug resistance in various cancers. Moreover, ABCB6 mediates ferroptosis, and modulating its expression can induce ferroptosis to suppress tumor progression[10-11]. However, the role and mechanism of ABCB6 in OS remain unexplored, warranting further investigation. Ferroptosis is iron-dependent lipid peroxide-driven cell death with tumor-suppressive functions[12-14]. Hypoxia-inducible factor 1-alpha (HIF1A) promotes ferroptosis resistance in solid tumors and its silencing sensitizes tumors to ferroptosis inducers, suggesting that targeting HIF1A can modulate ferroptosis[15-16]. HIF1A undergoes proteasomal degradation regulated by prolyl hydroxylases (PHDs) and the von Hippel-Lindau protein (pVHL), with O2 and Fe2+ serving as essential cofactors[17-19]. Notably, ABCB6 maintains iron homeostasis, which is key for both HIF1A protein degradation and ferroptosis[20-22]. Herein, we identified ABCB6 as an oncogene in OS via functional assays after ABCB6 silencing. Subsequent experiments revealed that ferroptosis underlies the anti-OS effects of ABCB6 knockdown. Mechanistically, ABCB6 silencing elevated intracellular Fe2+, leading to HIF1Aprotein degradation and ferroptosis induction. Increased HIF1Aprotein degradation further reduced glutathione peroxidase 4 (GPX4) transcription, enhancing ferroptosis in OS cells. Together, these findings show that ABCB6 acts as an oncogene and that targeting ABCB6 induces ferroptosis to suppress OS progression, offering new insights for OS research and therapy. Materials and methods Ethics statements The animal experiments in this study acquired approval from the Animal Care and Ethics Committee of Xiangya Hospital, Central South University (Approval number: 2023040746). Cells, reagents, and transfection Human bone marrow mesenchymal stem cells (BMSCs) were isolated from patients undergoing hip replacement at the Department of Orthopedics, Xiangya Hospital with informed consent. The human OS cell lines, HOS and 143B cells, and 293T cells were purchased from Procell Company (WuHan, China). Under the circumstance of 5% CO2 and 37 ℃, cells were cultured in Dulbecco’s modified Eagle medium (DMEM; Biological Industries, Israel) added with 1% penicillin/streptomycin (Gibco, USA) and 10% fetal bovine serum (Gibco, USA). To minimize potential off‑target effects, 2 independent short hairpin RNA (shRNA) sequences targeting ABCB6 (SH1 and SH2) and their respective scrambled control lentiviruses were purchased from General Biosystems (Nanjing, China) and used for gene silencing. Lentiviruses for HIF1A overexpression and its empty‑vector control were sourced from Genechem (Shanghai, China). HIF1A and GPX4 plasmids for dual-luciferase assays were also synthesized by General Biosystems (Nanjing, China). ABCB6-silenced and HIF1A-overexpressed OS cells were established via lentiviral infection and selected with puromycin (1.2 mg/mL) for 4 weeks. The following antibodies were used: anti-ABCB6 (121438, zenbio), anti-HIF1A (ab308433, ABclonal), anti-solute carrier family 7 member 11 (SLC7A11; 26864-1-AP, proteintech), anti-E-cadherin (ab40772, ABclonal), anti-N-cadherin (ab76011, ABclonal), anti-Ki-67 (ab15580, ABclonal), anti-GPX4 (30388-1-AP, proteintech), anti-cleaved caspase-3 (341034, zenbio), and anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH; 10494-1-AP, proteintech). Ferrostatin-1 (Fer-1; a lipid peroxidation scavenger), deferoxamine (DFO; an iron chelator), and MG-132 (a proteasome inhibitor) were purchased from MedChemExpress (MCE, USA) with the respective catalog numbers 347174-05-4, 138-14-7, and 133407-82-6. Cancer Genome Atlas Program data acquisition and processing The RNA sequencing and the corresponding clinical traits of 85 OS patients were downloaded from Therapeutically Applicable Research to Generate Effective Treatments (TARGET), acquiring from The Cancer Genome Atlas Program (TCGA) website (https://www.cancer.gov/ccg/research/genome-sequencing/tcga). The R package maxstat (Maximally Selected Rank Statistics with several P-value approximations, version 0.7-25.0) was utilized to compute the optimal cut‑off value for ABCB6. The minimum group sample size was set to >25% and the maximum group sample size to <75% of the total sample. The resulting optimal cut‑off value was 0.26. Based on this cut‑off, patients were divided into a high-expressed group and a low-expressed group. Subsequently, we analyzed the prognostic difference between the 2 groups using the survfit function from the R package survival and assessed the statistical significance of the difference using the log‑rank test. The data acquisition and utilization were in accordance in the policy of TCGA website. Immunohistochemistry staining of tissues For immunohistochemistry (IHC), tumor tissues were undergone fixation, permeabilization, dehydration, embedding, and section. Subsequently, tissue slides were incubated with anti-ABCB6, anti-Ki-67, anti-E-cadherin, anti-N-cadherin, anti-caspase-3, anti-HIF1A, anti-GPX4 primary and the corresponding secondary antibodies, and the details of IHC assays were performed as described in our previous study[23]. RNA extraction and real-time polymerase chain reaction The total RNA was extracted from human BMSCs and OS cells using the AG RNAex Pro RNA extraction kit (AG, Changsha, China) and then utilized for the synthesis of complementary DNA (cDNA) using the Reverse Transcription kit (AG, Changsha, China). The TBGreen Premix Pro Taq HS real-time polymerase chain reaction (qPCR) kit (AG, Changsha, China) was utilized for qPCR analysis. The ΔΔCq method was utilized to calculate the relative RNA expression of each sample and the results were expressed as 2-ΔΔCq. GAPDH was used as the internal control gene. The primers for qPCR are shown in the Table 1. Table 1Primers for qPCRGenesForward primer (5' to 3')Reverse primer (5' to 3') ABCB6 GGCTCTGGGGACTCTGGCACGTAGGGAGAGATGCGA HIF1A AAGCACAGTTACAGTATTCCAGCAGTCATCAGTGGTGGCAGTGGTAG GPX4 AGAGATCAAAGAGTTCGCCGCTCTTCATCCACTTCCACAGCG GAPDH GATTCCACCCATGGCAAATTCCTGGAAGATGGTGATGGGATTqPCR: Real-time polymerase chain reaction; ABCB6:ATP-binding cassette subfamily B member 6; HIF1A: Hypoxia-inducible factor 1-alpha; GPX4: Glutathione peroxidase 4; GAPDH: Glyceraldehyde-3-phosphate dehydrogenase. Primers for qPCR qPCR: Real-time polymerase chain reaction; ABCB6:ATP-binding cassette subfamily B member 6; HIF1A: Hypoxia-inducible factor 1-alpha; GPX4: Glutathione peroxidase 4; GAPDH: Glyceraldehyde-3-phosphate dehydrogenase. Western blotting Proteins were extracted with a mammalian total protein extraction kit (DE101-01, TransGen), resolved by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred to polyvinylidene fluoride (PVDF) membranes. After blocking with 5% non-fat milk for 1 hour, membranes were incubated with primary antibodies overnight at 4 ℃, followed by secondary antibodies for 1 hour, and detected using an enhanced chemiluminescence (ECL) kit (BL520A, Biosharp, China). Of note, for mechanistic exploration of the signaling axis, cells were treated with 100 μmol/L DFO for 24 hours to chelate intracellular Fe2+, and cells were exposed to 10 μmol/L MG‑132 for 6 hours to evaluate the potential proteasomal degradation of HIF1A. Cell proliferation and colony formation Proliferation was evaluated using a cell counting kit (CCK)-8 (Beyotime, China). About 3 000 cells/well were seeded in 96-well plates and processed according the manufacturer's instructions. For colony formation, 500 cells/well were seeded in 6-well plates and cultured for 2 weeks as previously described[23]. Transwell migration and scratch wound-healing assay Transwell migration assays were performed by seeding 2×104 cells in 200 µL serum-free medium into the upper chamber (24-well insert, Corning) with 800 µL complete medium below. For scratch assays, 3×105 cells were seeded in 6-well plates, scratched with a 1 mL tip at 90% confluence. Detailed procedures followed our previous study[23]. Flow cytometry In term of cell apoptosis assay, cells were digested using ethylene diamine tetraacetic acid (EDTA)-free trypsin, washed twice with pre-cooled phosphate buffered saline (PBS), and stained with the apoptosis detection kit (#FXP023, 4A BIOTECH, China). Subsequently, the stained samples were incubated in the dark 10 min at room temperature and analyzed by flow cytometry within 1 hour. For liperfluo and reactive oxygen species (ROS) analyses, OS cells were seeded in the 6-well plates and stained with liperfluo (Dojindo, Kumamoto, Japan) and dihydroethidium (DHE; US Everbright, Suzhou) probe in accordance to the manufacturer’s protocol. Subsequently, the stained cells were digested with trypsin, washed twice with pre-cooled PBS and detected by flow cytometry. In the ferroptosis rescue experiments, cells were co‑incubated with either 1 μmol/L Fer‑1 or 100 μmol/L DFO for 24 hours, in order to assess the reversal of ABCB6‑driven cell death and lipid peroxidation. Detection of intracellular Fe2+ levels Cells were cultured in the 24-well plate and then stained in 1 μmol/L FerroOrange (Dojindo, Kumamoto,Japan) diluted with Hank’s Balanced Salt Solution for exactly 30 min under the circumstance of 5% CO2 and 37 ℃. After staining, cells were imaged with a fluorescence microscope (Leica, Weztlar, German). Additionally, DFO (100 μmol/L, 24 hours) was employed to reverse the elevation of intracellular Fe2+ caused by ABCB6 knockdown. Detection of glutathione and oxidized glutathione Cells were seeded in the 6-well plate, made into the lysates, and centrifuged at 10 000 ×g at 4 ℃ for 10 min. Subsequently, the supernatant was collected, and the total glutathione (GSH) and oxidized glutathione (GSSG) levels were determined using GSH and GSSG Assay Kit (Beyotime, Jiangsu, China). The operation and measurement of this assay were in accordance with the manufacturer's instructions. Mitochondrial membrane potential detection Since ferroptotic cells typically display decreased mitochondrial membrane potential (MMP), we employed the 5,5’,6,6’-tetrachloro-1,1’,3,3’-tetraethyl-imidacarbocyanine iodide (JC-1) probe to detect MMP changes after ABCB6 knockdown. MMP was conducted using JC-1 MMP Assay Kit (Solarbio, Beijing, China). Cells were seeded in the 24-well plate. After cultured overnight, cells were stained with 1×JC-1 solution for 20 min in the dark and then imaged with a fluorescent microscope (Leica, Weztlar, German). JC-1 images were analyzed with ImageJ (Fiji). For each field, the mean fluorescence intensities of green (JC-1 monomer) and red (JC-1 aggregate) were measured after manually outlining regions of interest (ROI). The red/green ratio was calculated to reflect mitochondrial membrane potential. Transmission electron microscope Cells in 10 cm dishes were fixed with 2.5% glutaraldehyde for 5 min, scraped, and centrifuged. The pellet was fixed again in the dark for 30 min, post-fixed with 1% osmium tetroxide, dehydrated in ethanol and acetone, infiltrated with epoxy resin overnight, embedded, and polymerized at 60 ℃ for 48 hours. Sections were cut with an ultra-microtome (Leica EM UC7), stained with uranyl acetate and lead citrate, and observed by transmission electron microscope (TEM; Talos L120C G2) at 120 keV. Transcriptome sequencing Total RNA was extracted from ABCB6-silenced and control OS cells using TRIzol reagent (AG, Changsha, China) according to the manufacturer’s protocol. Transcriptome sequencing and analysis were performed by OE Biotech (Shanghai, China). Differential expression analysis (DEA) was conducted using DESeq2, with significantly differentially expressed genes (DEGs) defined as Q value<0.05 and fold change >2 or <0.5. Chromatin immunoprecipitation and dual-luciferase assays Chromatin immunoprecipitation (ChIP) assays were performed using a ChIP kit (Bersinbio, Guangzhou) in 143B and HOS cells according the manufacturer’s protocol. Immunoprecipitation used an anti-HIF1A antibody, with normal immunoglobulin (Ig) G as negative control and input chromatin as normalization control. Purified DNA was analyzed by qPCR, and data were presented as percentage of input. The primers are listed in Table 2. Table 2Primers for ChIP-qPCRTarget regionForward primer (5' to 3')Reverse primer (5' to 3') GPX4 promoter regionAAAGGTAAATAAGAGTTGGGGACACAAGAAGGCGGGGAAGGGChIP: Chromatin immunoprecipitation. Primers for ChIP-qPCR ChIP: Chromatin immunoprecipitation. The dual-luciferase assay was performed using Luc-PairTM Duo Luciferase HS Assay Kit (GeneCopoeia, Guangzhou, China). In brief, 293T cells were seeded in the 12-well plates and infected with the constructed luciferase reporter plasmid. After infected 48 hours, the cells were lysised and used for fluorescence detection in accordance with the manufacturer's instructions. Xenograft mouse models and in-situ tumor formation For xenograft mouse models, 4- to 6-week-old female nude mice were randomly divided into the ABCB6-knockdown and the negative control groups (n=5), and the corresponding 143B OS cells were injected in these mice. In the subsequent feeding days, the tumor volumes of mice were measured every 5 days, and tumors were harvested after 25 days, with their weights measured. Statistical analysis The quantitative data were expressed as the mean±standard deviation and tested by unpaired student's t-test. GraphPad Prism 8 were utilized for data processing, statistical analysis, and presentation of all the experimental results. The indicated P value was considered statistically significant. Results ABCB6expression was associated with OS patients' prognoses Compared with human BMSCs, the messenger RNA (mRNA) levels of ABCB6 were elevated in OS cells, particularly in HOS and 143B cells (both P<0.05, Figure 1A). Consistently, western blotting analysis confirmed that ABCB6 protein expression was also higher in OS cells than in BMSCs (P<0.05, Figure 1B). Subsequently, the TARGET-OS cohort was used to investigate the clinical significance of ABCB6 upregulation in OS and the Kaplan-Meier curve revealed that high ABCB6 mRNA expression was negatively associated with the prognosis of OS patients [hazard ratio (HR)=2.94, 95% confidence interval (CI) 1.36 to 6.36, P<0.05; Figure 1C]. To further assess the functional role of ABCB6 in OS progression, we knocked down ABCB6 expression in 143B and HOS cells (both P<0.05, Figure 1D). Transwell migration and scratch wound-healing assays demonstrated that ABCB6 knockdown significantly impaired the migratory capacity of OS cells (both P<0.05, Figure 1E and 1F). Meanwhile, flow cytometry analysis of cell apoptosis revealed that ABCB6 silencing markedly increased the apoptosis rate of OS cells (P<0.05, Figure 2A). Furthermore, we examined the effect of ABCB6 on OS cell proliferation. The results showed that reduced ABCB6 expression suppressed cell proliferation and colony formation, ultimately inhibiting OS growth (both P<0.05, Figure 2B and 2C). Finally, western blotting of apoptosis- and migration-related markers indicated that ABCB6 knockdown markedly decreased N-cadherin expression while increasing both E-cadherin and cleaved caspase-3 levels, suggesting that ABCB6 silencing promotes apoptosis and inhibits migration in OS cells (all P<0.05, Figure 2D). Taken together, these findings indicate that ABCB6 functions as an oncogene in OS, and its elevated expression correlates with a worse prognosis in OS patients. ABCB6 functioned as an oncogene in OS, and its knockdown suppressed OS cell migration A: The mRNA expression of ABCB6 was upregulated in OS cells. B: Compared with BMSCs, the protein expression of ABCB6 was upregulated in OS cells, 143B and HOS cells. C: TARGET-OS cohort showed that ABCB6 mRNA expression was negatively associated with the prognosis of OS patients. D: Validation of the knockdown efficiency of ABCB6 in OS cells. E: Transwell assay. F: Scratch wound-healing assay. The data represent the mean±standard deviation. *P<0.05, **P<0.01, and ***P<0.001 indicate a significant difference between the indicated groups. mRNA: Messenger RNA; ABCB6: ATP-binding cassette subfamily B member 6; OS: Osteosarcoma; BMSC: Bone marrow mesenchymal stem cell; TARGET: Therapeutically Applicable Research to Generate Effective Treatments; GAPDH: Glyceraldehyde-3-phosphate dehydrogenase; HR: Hazard ratio; CI: Confidence interval; NC: Negative control; OD: Optical density. Knockdown of ABCB6 promoted apoptosis and inhibited proliferation in OS cells A: Cell apoptosis assay. B: CCK‑8 assay, †P<0.05, ††P<0.01, and †††P<0.001 vs the NC group. C: Colony formation assay. D: Western blotting analysis revealed that ABCB6 knockdown significantly reduced N-cadherin expression while concurrently upregulating E-cadherin and cleaved caspase-3 levels. The data represent the mean±standard deviation. *P<0.05, **P<0.01, and ***P<0.001 indicate a significant difference between the indicated groups. CCK-8: Cell counting kit-8; PI: Propidium iodide. Targeting ABCB6 induced OS cells GPX4-associated ferroptosis First, the liperfluo assay, a specific indicator of ferroptosis, revealed that ABCB6 silencing significantly increased the proportion of liperfluo-positive cells, indicating the occurrence of ferroptosis (Figure 3A). Meanwhile, the DHE probe, used to assess cellular oxidation levels, showed that ABCB6-knockdown OS cells exhibited higher oxidation levels compared with controls (Figure 3B). The detection results of intracellular Fe2+ concentration demonstrated that ABCB6 silencing increased the intracellular Fe2+ concentration (Figure 3C). In contrast to the elevated oxidation level, the level of reducing substances, such as GSH/GSSG, decreased following ABCB6 knockdown (P<0.05, Figure 3D). The results of the JC-1 experiment showed that ABCB6 silencing reduced the MMP of OS cells, suggesting mitochondrial damage (P<0.05, Figure 3E). The results of the TEM examination further confirmed revealed damaged mitochondrial structures in ABCB6-knockdown OS cells (Figure 4A and 4B). Moreover, treatment with ferroptosis inhibitors reversed the elevated oxidation levels induced by ABCB6 silencing, further supporting the involvement of ferroptosis (Figure 4C and 4D). Western blotting analysis revealed that GPX4 expression was significantly reduced following ABCB6 knockdown (P<0.05), whereas SLC7A11 expression showed no significant change (P>0.05, Figure 4E). Additionally, qPCR analysis demonstrated that the mRNA levels of GPX4 were also downregulated upon ABCB6 silencing, indicating that ABCB6 regulates GPX4 expression at the transcriptional level (all P<0.05, Figure 4F and 4G). Taken together, these findings suggest that the anti-OS effect of ABCB6 knockdown likely stems from its induction of GPX4-associated ferroptosis in OS cells. Knockdown of ABCB6 induced oxidative stress, elevated Fe2+ levels, and reduced mitochondrial membrane potential in OS cells A: Liperfluo assays. B: ROS detection. C: Intracellular Fe2+ concentration detection. D: GSH and GSSG detection. E: Mitochondrion membrane potential. The data represent the mean±standard deviation. *P<0.05, **P<0.01, and ***P<0.001 indicate a significant difference between the indicated groups. GSH: Glutathione; GSSG: Glutathione disulfide; ROS: Reactive oxygen species; JC-1: 5,5',6,6'-Tetrachloro-1,1',3,3'-tetraethylbenzimidazolylcarbocyanine iodide. Ferroptosis inhibitor reversed ABCB6 knockdown‑induced oxidative stress in OS cells, and GPX4 served as a potential mediator of ABCB6 action A and B: Transmission electron microscope revealed that ABCB6 knockdown led to mitochondrial structural damage in 143B cells (A) and HOS cells (B). C and D: Ferroptosis inhibitors reversed the lipid peroxidation (C) and ROS (D) induced by ABCB6 knockdown. E: Expressional exploration of ferroptosis related molecules. F and G: qPCR of ABCB6 and GPX4 in 143B cells (F) and HOS cells (G). The data represent the mean±standard deviation. *P<0.05, **P<0.01, and ***P<0.001 indicate a significant difference between the indicated groups. SLC7A11: Solute carrier family 7 member 11; GPX4: Glutathione peroxidase 4; qPCR: Real-time polymerase chain reaction. ABCB6 silencing induced OS ferroptosis via Fe2+/HIF1A/GPX4 signaling pathway To further elucidate the mechanism by which ABCB6 regulates GPX4 expression, we performed transcriptome sequencing (Figure 5A). Gene Ontology (GO) analysis revealed that DEGs were enriched in transcriptional pathways, including positive regulation of transcription by RNA polymerase II and DNA binding, suggesting that ABCB6 may affect transcriptional activity in OS cells (Figure 5B). Moreover, the HIF-1 signaling pathway was significantly enriched in the negative control group (P<0.001, Figure 6A), and subsequent western blotting confirmed that ABCB6 knockdown decreased HIF1A protein levels (both P<0.05, Figure 6B). Fe2+ probe assays confirmed that DFO treatment reversed the ABCB6 knockdown-induced elevation of intracellular Fe2+ (Figure 6C), and western blotting showed that DFO also reversed the degradation of HIF1A (both P<0.05, Figure 6D). Subsequently, we employed a ubiquitination inhibitor to block HIF1A degradation. Western blotting analysis revealed that the inhibitor attenuated the degradation of HIF1A protein (both P<0.05, Figure 6E), indicating that ABCB6 likely promotes HIF1A protein degradation through elevated intracellular Fe2+ concentration. Transcriptome sequencing revealed the transcriptional landscape of OS cells following ABCB6 knockdown A: Volcano plot shows the differentially expressed genes between the ABCB6 knockdown group and the control group. B: Gene ontology analysis. HIF1A: Hypoxia-inducible factor 1-alpha; GSEA: Gene Set Enrichment Analysis. ABCB6 knockdown inhibited HIF1A expression A: GSEA analysis. B: Expressional exploration of HIF1A in ABCB6 knockdown cells. C: DFO could reduce the elevated intracellular Fe2+ concentration resulted from ABCB6 knockdown. D: DFO could reverse the degradation of HIF1A resulted from ABCB6 knockdown. E: Proteasome inhibitor could reverse the degradation of HIF1A resulted from ABCB6 knockdown. The data represent the mean±standard deviation. *P<0.05, **P<0.01, and ***P<0.001 indicate a significant difference between the indicated groups. FC: Fold change. ChIP-qPCR assays results showed significant enrichment of the GPX4 promoter region in anti-HIF1A immunoprecipitates compared with the IgG control (Figure 7A). Dual luciferase reporter assays further confirmed that HIF1A promotes GPX4 transcription (Figure 7B). Consistently, western blotting showed that ABCB6 knockdown reduced HIF1A expression and subsequently suppressed GPX4 expression (all P<0.05, Figure 7C). To validate whether the ABCB6 knockdown-induced downregulation of GPX4 is mediated by HIF1A, we overexpressed HIF1A in ABCB6-silenced cells, the results showed that HIF1A overexpression reversed the GPX4 downregulation caused by ABCB6 knockdown (all P<0.05, Figure 7D and 7E). ABCB6 silencing induced OS ferroptosis via Fe2+/HIF1A/GPX4 signaling pathway A: ChIP-qPCR of HIF1A enrichment at the GPX4 promoter. B: Dual-luciferase assay. C: Expressional exploration of HIF1A and GPX4 in ABCB6 knockdown cells. D: qPCR indicated that HIF1A overexpression reversed the decreased GPX4 in ABCB6 knockdown 143B cells and HOS cells. E: Western blotting indicated that HIF1A overexpression reversed the decreased GPX4 in ABCB6 knockdown 143B cells and HOS cells. The data represent the mean±standard deviation. *P<0.05, **P<0.01, and ***P<0.001 indicate a significant difference between the indicated groups. ChIP: Chromatin immunoprecipitation; Ig: Immunoglobulin; OE: Over expression. CCK-8 assays demonstrated that HIF1A overexpression rescued the proliferative ability suppressed by ABCB6 knockdown (Figure 8A and 8B). Transwell assays showed that HIF1A overexpression enhanced the migration of ABCB6-silenced OS cells (Figure 8C and 8D). Furthermore, HIF1A overexpression significantly inhibited apoptosis in ABCB6-knockdown OS cells (Figure 8E). In addition, MMP was increased upon HIF1A overexpression, indicating that HIF1A alleviates mitochondrial damage induced by ABCB6 silencing (Figure 8F). HIF1A overexpression reversed the biological functional inhibition in ABCB6 knockdown cells A and B: CCK-8 assays demonstrated that HIF1A overexpression reversed the proliferative inhibition in ABCB6 knockdown 143B cells (A) and HOS cells (B). C and D: Transwell assays exhibited that HIF1A overexpression reversed the weaken migrative inhibition in ABCB6 knockdown 143B cells (C) and HOS cells (D). E: Flow cytometry displayed that HIF1A overexpression reduced the apoptosis rate of 143B cells and HOS cells caused by ABCB6 knockdown. F: HIF1A overexpression reversed the reduced MMP in ABCB6 knockdown cells. The data represent the mean±standard deviation. *P<0.05, **P<0.01, and ***P<0.001 indicate a significant difference between the indicated groups. Taken together, our study demonstrates that ABCB6 silencing induces GPX4-related ferroptosis in OS cells via the Fe2+/HIF1A signaling pathway, thereby inhibiting OS progression. ABCB6 knockdown suppressed OS growth in vivo Xenograft tumor formation assay results showed that tumors derived from ABCB6-silenced stable cells exhibited significantly reduced volume and weight compared with the negative control group (both P<0.05, Figure 9A-9C). IHC staining for ABCB6 confirmed successful knockdown of ABCB6 in the shABCB6 groups (Figure 9D). Meanwhile, IHC analysis of tumor sections revealed that, compared with the negative control group, the expression levels of HIF1A and GPX4 were downregulated in the ABCB6 knockdown groups (both P<0.05, Figure 9D-9F). Subsequently, we found that in the ABCB6 knockdown groups, intratumoral cleaved caspase-3 levels were upregulated, whereas the expression of the proliferation marker Ki-67 was inhibited, compared with the negative control group (both P<0.05, Figure 9G-9I). Furthermore, IHC results revealed that E-cadherin expression was upregulated upon ABCB6 knockdown, whereas N-cadherin expression showed the opposite trend (both P<0.05, Figure 9J-9L). Collectively, our in vivo findings corroborated the in vitro results, demonstrating that ABCB6 silencing suppresses OS progression by inducing GPX4-related ferroptosis in OS cells via the Fe2+/HIF1A signaling pathway. Validation of anti-OS effects of ABCB6 in vivo A: Ex vivo tumors showed that ABCB6 knockdown inhibited the in vivo growth of OS. B: Tumor growth curves showed that tumor growth in the ABCB6-knockdown group was significantly inhibited compared with the control group. C: Tumor weight measurements revealed a significant difference between the ABCB6-knockdown group and the control group. D-F: IHC staining against HIF1A (D and E) and GPX4 (D and F). G-I: IHC staining against cleaved caspase-3 (G and H) and Ki-67 (G and I). J-L: IHC staining against E-cadherin (J and K) and N-cadherin (J and L). The data represent the mean±standard deviation. *P<0.05, **P<0.01, and ***P<0.001 indicate a significant difference between the indicated groups. Discussion The therapeutic bottleneck in OS necessitates exploring its molecular mechanisms to develop novel strategies[2-3, 24]. Since 2012, ferroptosis has been a promising approach for refractory cancers[12, 25-26]. Cancer cells resistant to conventional therapies often remain ferroptosis-susceptible, offering a complementary OS strategy[25, 27-28]. Inducing ferroptosis in OS may overcome resistance and immune non-responsiveness[29-31], thus representing a potential avenue to break the therapeutic dilemma in OS. In this study, we identified ABCB6 as an oncogene in OS that is negatively associated with patient prognosis. ABCB6 expression was upregulated in OS cell lines compared with hBMSCs, and Kaplan-Meier analysis of the TARGET-OS cohort revealed distinct prognoses between patients with differential ABCB6 expression. Previous studies have implicated ABCB6 in tumorigenesis, progression, and chemoresistance in various cancers[6, 32]; however, its role and mechanism in OS remain unexplored. Based on the differential expression and prognostic significance, we selected 2 OS cell lines with the highest ABCB6 upregulation for further investigation. Functional assays demonstrated that ABCB6 knockdown suppressed OS cell proliferation and migration while promoting apoptosis. We note that ABCB6 knockdown simultaneously induces both ferroptosis and apoptosis in OS cells; however, the temporal relationship and molecular crosstalk between these 2 processes remain unclear. Future studies using time-sequential inhibition experiments, e.g., blocking key nodes of apoptosis or ferroptosis, are required to determine whether they occur sequentially or independently in parallel. Moreover, a xenograft tumor formation assay in nude mice confirmed the in vivo tumor-suppressive effect of ABCB6 knockdown. Therefore, targeting ABCB6 exhibits therapeutic potential for OS, and elucidating its molecular mechanism will facilitate clinical translation and treatment optimization. Further mechanistic exploration revealed that the anti-OS effect of ABCB6 knockdown may be attributed to its induction of ferroptosis. ABCB6 silencing led to elevated oxidation levels, lipid peroxidation, and mitochondrial damage in OS cells; these effects were reversed by ferroptosis inhibitors, confirming that ABCB6 silencing triggers ferroptosis. Previous studies have shown that ABCB6 plays a role in regulating ferroptosis[10-11]. For instance, Yin, et al[11] reported that targeting ABCB6 suppresses the PI3K/AKT pathway to induce ferroptosis in multiple myeloma cells. In our study, both mRNA and protein levels of the ferroptosis-related gene GPX4 were significantly downregulated following ABCB6 knockdown. Transcriptomic analysis revealed enrichment of DEGs in transcription-related pathways, suggesting that ABCB6 knockdown may affect GPX4 transcription. Notably, the HIF-1 signaling pathway was enriched in the negative control group. Given that HIF1A is a transcription factor for ferroptosis-related genes[33-35], we hypothesized that HIF1A mediates the effect of ABCB6 knockdown on GPX4 expression. ChIP and dual-luciferase assays confirmed that HIF1A binds to the GPX4 promoter and promotes its transcription, which was further validated by western blotting. Hence, ABCB6 may induce OS cell ferroptosis via the HIF1A/GPX4 signaling pathway. We found that ABCB6 knockdown increased intracellular Fe2+ and downregulated the HIF-1 signaling pathway. Given that Fe2+ promotes HIF1A protein degradation[17-18, 36] and ABCB6 regulates iron homeostasis[20-23, 25, 37], we hypothesized that elevated Fe2+ leads to HIF1A protein degradation. Treatment with the iron chelator DFO or the proteasome inhibitor MG-132 reversed this degradation, confirming the hypothesis. Furthermore, HIF1A overexpression in ABCB6-silenced OS cells rescued proliferation and migration while inhibiting apoptosis, indicating that the anti-OS effect of ABCB6 knockdown partly depends on HIF1A protein degradation. Consistent with reports that HIF1A promotes ferroptosis resistance[15-16, 33, 38], ABCB6 knockdown reduced HIF1A expression and MMP, both of which were restored by HIF1A overexpression. Notably, GPX4 expression consistently paralleled HIF1A under all conditions, and ChIP and dual-luciferase assays confirmed that HIF1A binds to and transcriptionally activates GPX4. Collectively, ABCB6 silencing elevates intracellular Fe2+, which promotes HIF1A protein degradation, thereby suppressing GPX4 transcription and ultimately triggering ferroptosis in OS cells. In summary, this study identifies ABCB6 as an oncogene in OS, and targeting ABCB6 exerts potent anti-OS effects by inducing ferroptosis. Mechanistically, ABCB6 knockdown increases intracellular Fe2+ concentration, facilitating HIF1A protein degradation, which in turn suppresses GPX4 transcription and expression, ultimately triggering ferroptosis in OS cells.