Targeting the NR3C1‐ACSL4 Axis Triggers Ferroptosis to Overcome Radioresistance in Prostate Cancer Authors: Jing Yang, Puyuan Chen, Yawen Zheng, Feiyue Shi, Min Wang, Luwei Xu, Xiaowei Wei, Xiumei Wang Published: 2026 ABSTRACT ABSTRACTRadioresistance in prostate cancer demands innovative sensitization strategies. We identified the glucocorticoid receptor nuclear receptor subfamily 3 group C member 1 (NR3C1) as a key negative regulator of radiosensitivity linked to poor prognosis. NR3C1 transcriptionally upregulates the lipid‐metabolizing enzyme, acyl‐CoA synthetase long‐chain family member 4 (ACSL4), thereby enhancing cell proliferation, migration, and radioresistance. High ACSL4 expression sensitizes cells to ferroptosis inducers that amplify lipid peroxidation and restore radiosensitivity. Using the clinically applicable ferroptosis inducer, dihydroartemisinin (DHA), we found that DHA synergizes with ACSL4 to trigger ferroptosis, sensitizing radioresistant cells and xenografts to radiation. This effect was characterized by elevated lipid peroxidation and was reversed by the ferroptosis inhibitor deferoxamine. Collectively, our study revealed that the NR3C1‐ACSL4 axis regulates lipid peroxidation and promotes radioresistance. Targeting this axis with DHA and using ACSL4 as a biomarker represents a promising preclinical strategy to overcome radioresistance in prostate cancer, pending further clinical validation. Abbreviations acyl‐CoA synthetase long‐chain family member 4 annexin A1 annexin A2 androgen receptor amphiregulin dihydroartemisinin glucocorticoid receptor hydroxysteroid 17‐beta dehydrogenase 4 nuclear receptor subfamily 0 group B member 1 nuclear receptor subfamily 3 group C member 1 pre‐B‐cell leukemia homeobox 1 prostate cancer polyunsaturated fatty acids SMAD family member 3 signal transducer and activator of transcription 5B Introduction Prostate cancer (PCa) is the leading cause of cancer‐related mortality in men [1]. Radiotherapy is a key treatment for both localized and advanced PCa [2, 3]. However, the development of radioresistance is a major clinical obstacle, leading to disease recurrence and progression [4, 5]. The mechanisms driving radioresistance in PCa include enhanced antioxidant defense and DNA repair capacity through pathways such as the Nrf2/Keap1 signaling axis, Bcl‐2 mutations, and alterations in speckle‐type poxvirus and zinc finger protein; regulating hypoxia, androgen receptor variants, and ferroptosis regulators; and enhancing genomic DNA stability through polypyrimidine tract binding protein 1‐mediated alternative splicing of DNA methyltransferase 3b [6, 7]. However, these mechanisms are highly heterogeneous and interconnected, and clinical translation remains slow. Therefore, elucidating the molecular drivers of radioresistance and developing effective strategies to overcome it is a critical unmet need in PCa management. The glucocorticoid receptor (GR), encoded by the nuclear receptor subfamily 3 group C member 1 (NR3C1) gene, plays a complex role in cancer biology [8, 9, 10, 11]. In PCa, GR signaling has been implicated in disease progression and resistance to androgen receptor‐targeted therapies [12]. Clinical observations suggest that GR activation may be associated with adverse outcomes [13]; however, attempts to broadly block GR signaling have not yielded significant prognostic improvements [14], likely due to its extensive and pleiotropic downstream network. This highlights the necessity of identifying precise, therapeutically actionable downstream effectors of the GR that govern specific phenotypes, such as radioresistance. Ferroptosis, iron‐dependent cell death driven by lipid peroxidation, has emerged as a therapeutic vulnerability in cancer [15, 16]. Acyl‐CoA synthetase long‐chain family member 4 (ACSL4) is a key enzyme that enriches cellular membranes with polyunsaturated fatty acids (PUFAs), thereby supplying the substrate necessary for lipid peroxidation and determining cellular sensitivity to ferroptosis inducers [17, 18]. Intriguingly, while ACSL4 promotes ferroptosis, its function in certain cancers has also been linked to aggressive tumor phenotypes and therapy resistance [19, 20], suggesting a context‐dependent role that warrants investigation in the setting of radiotherapy. In this study, we provided evidence that the NR3C1‐ACSL4 axis contributes to radioresistance in PCa. ACSL4‐high tumors displayed specific vulnerability to ferroptosis inducers, which facilitates radiosensitization. Dihydroartemisinin (DHA) exploited this vulnerability by inducing ferroptosis and synergizing with radiotherapy. These findings identify a potentially targetable GR‐ACSL4‐ferroptosis pathway and support further investigation of DHA as a biomarker‐informed radiosensitizing strategy. Methods and Materials Cell Lines and Cell Culture Conditions The human prostate cancer cell lines LNCaP, 22Rv1, DU145, and PC3 and the normal prostate‐derived cell lines WPMY‐1 and RWPE‐1 were obtained from Shanghai Zhong Qiao Xin Zhou Biotechnology Co Ltd. LNCaP, DU145, PC3, and RWPE‐1 cells were cultured in RPMI‐1640 (#KGL1503‐500, KeyGen Biotech, Nanjing, China) supplemented with 10% fetal bovine serum (FBS; #A5256701, Gibco, Brazil). 22Rv1 and WPMY‐1 cells were cultured in high‐glucose DMEM (#KGL1206‐500, KeyGen Biotech) supplemented with 10% FBS. All cells were maintained at 37°C under a humidified atmosphere with 5% CO2. Cell Transfection To achieve gene silencing, specific small interfering RNAs (siRNAs) were transfected with KeygenMAX 3000 Transfection Reagent (#KGA9705, KeyGEN, China). Target‐specific and non‐targeting control siRNAs were obtained from Tsingke Biotech Co. Ltd. (Beijing, China). Stable NR3C1 knockdown was performed using lentiviruses carrying an NR3C1‐specific shRNA (#P55188, MIAOLING Biology, Wuhan, China). Stable ACSL4 overexpression in LNCaP cells was achieved using lentiviruses carrying pLV3‐CMV‐ACSL4‐3 × FLAG‐CopGFP (#P94237, MIAOLING Biology). The siRNA sequences used were as follows: siNR0B1: GCUGAGAUUCAUCAAUGCCAA; siAXNA1: CCCUAUUUCAUGAUAUCAGAA; siAXNA2: CGGGAUGCUUUGAACAUUGAA; siNR3C1: GUGUCACUGUUGGAGGUUAUU; siAREG: CCUGGCUAUAUUGUCGAUGAU. Compounds PRGL493 (#HY‐139180, MedChemExpress, New Jersey, USA), erastin (#HY‐15763, MedChemExpress), RSL3 (#HY‐100218A, MedChemExpress), deferoxamine (#HY‐B1625, MedChemExpress), and dihydroartemisinin (#D140839, Aladdin Scientific, Shanghai, China) were used as described in the figure legends. Immunoblotting Cells were lysed in cell lysis buffer for western and IP (#P0013, Beyotime, Shanghai, China) supplemented with protease and phosphatase inhibitors (#KGB5101‐100, KeyGen Biotech). Total protein extraction was quantified using a BCA Protein Assay Kit (#P0010, Beyotime) and then denatured by heating at 95°C for 10 min in SDS‐PAGE loading buffer (#WB2001, NCM biotech, Suzhou, China). Approximately 20 μg of protein per sample was separated on a 10% gradient SDS PAGE (#PG112, Epizyme Biotech, Shanghai, China) and subsequently transferred to polyvinylidene fluoride membranes (#IPVH00010, Merck‐Millipore, USA). The membranes were blocked with 5% nonfat milk and probed overnight (12 h) at 4°C with the following antibodies: anti‐NR3C1 (#24050‐1‐AP, Proteintech Group, Wuhan, China), anti‐annexin A1 (ANXA1) (#YP‐Ab‐03701, UpingBio, Hangzhou, China), anti‐annexin A2 (ANXA2) (#YP‐Ab‐03702, UpingBio), anti‐amphiregulin (AREG) (#YP‐Ab‐16,058, UpingBio), antinuclear receptor subfamily 0 group B member 1 (NR0B1) (#YP‐Ab‐06843, UpingBio), anti‐β‐actin (#66009‐1‐Ig, Proteintech Group), anti‐ACSL4 (#22401‐1‐AP, Proteintech Group), anti‐signal transducer and activator of transcription 5B (STAT5B) (#51072–2‐AP, Proteintech Group), anti‐pre‐B‐cell leukemia homeobox 1 (PBX1) (#YP‐Ab‐09660, UpingBio), anti‐SMAD family member 3 (SMAD3) (#YP‐Ab‐02017, UpingBio), and anti‐glutathione peroxidase 4 (GPX4) (#67763‐1‐Ig, Proteintech Group). Subsequently, the membranes were incubated with secondary antibodies (#SA00001‐1 and #SA00001‐2, Proteintech Group) for 60 min at room temperature and visualized by chemiluminescence. Quantitative Real‐Time PCR RNA was isolated from cells with Total RNA Extraction Reagent (#R401‐01, Vazyme Biomed, Nanjing, China). Quantitative PCR analysis was conducted in triplicate with the HiScript II One Step quantitative real‐time PCR (qRT‐PCR) Probe Kit (#Q222‐01, Vazyme Biomed) on a LightCycler 96 instrument (Roche Life Science, Basel, Switzerland). Gene expression levels were calculated by the 2−ΔΔCt method after normalizing Ct values to actin beta (ACTB). All primers were commercially synthesized (Tsingke Biotech Co. Ltd.). The sequences of primers were as follows: ACTB forward primer: 5′‐CCGCGAGAAGATGACCCAGA‐3′; ACTB reverse primer: 5′‐GGATAGCACAGCCTGGATAGCAA‐3′; NR3C1 forward primer: 5′‐TTCATCGAACTCTGCACCCC‐3′; NR3C1 reverse primer: 5′‐TTTGCCCATTTCACTGCTGC‐3′; ANXA1 forward primer: 5′‐GCCCCTGGATGAAACACTGA‐3′; ANXA1 reverse primer: 5′‐TCCTCAGATCGGTCACCCTT‐3′; ANXA2 forward primer: 5′‐GCTCGGGATCTCTATGACGC‐3′; ANXA2 reverse primer: 5′‐CACTGCGGGAGACCATGATT‐3′; AREG forward primer: 5′‐TGTCGCTCTTGATACTCGGC‐3′; AREG reverse primer: 5′‐ATGGTTCACGCTTCCCAGAG‐3′; NR0B1 forward primer: 5′‐TGCTCTTTAACCCGGACGTG‐3′; NR0B1 reverse primer: 5′‐GCGTCATCCTGGTGTGTTCA‐3′; ACSL4 forward primer: 5′‐AGGTGGTTTTGCACTGGTGA‐3′; ACSL4 reverse primer: 5′‐TGGTTCCTCAGCTCCTTCCT‐3′; PBX1 forward primer: 5′‐GGGCAGGTTCAGACAACTCA‐3′; PBX1 reverse primer: 5′‐GGCCTGCTCGTATTTCTCCA‐3′; SMAD3 forward primer: 5′‐CGCATGAGCTTCGTCAAAGG‐3′; SMAD3 reverse primer: 5′‐TCTGGGTGAGGACCTTGTCA‐3′; STAT5B forward primer: 5′‐CAGACCAAGTTTGCAGCCAC‐3′; STAT5B reverse primer: 5′‐ATTGCGGGTGTTCTCGTTCT‐3′. Colony Formation Assay The cells were plated in 6‐well plates at a density of 200 cells per well. After cell attachment, the indicated treatments (DMSO, Erastin, RSL3, or dihydroartemisinin) were added 2 h before irradiation (IR). After irradiation, the cells were cultured for 12 days. Colonies were fixed with 4% paraformaldehyde for 15 min and stained with 1.5% crystal violet for 20 min. The number of clones was obtained by a combination of spot counting using Image J software and manual counting. Subsequently, the surviving fraction was calculated as the number of colonies formed divided by the number of cells plated and normalized to the plating efficiency of the untreated control group. Calculation of the Bliss Synergy Ratio and the Sensitization Enhancement Ratio Bliss Synergy Ratio (Inhibition Fraction Version) The inhibition fraction was defined as ƒ = 1 − S, where S is the surviving fraction. For each combination of drug (D) and radiation (R), the expected inhibition under independence was calculated as follows: ƒbliss = ƒD + ƒR − ƒD × ƒR, where ƒD and ƒR were the mean inhibition fractions from the respective single‐agent controls (same concentration or dose). The bliss synergy ratio (BSR) was then calculated as follows: BSR = ƒobs/ƒbliss. BSR > 1, = 1, and < 1 indicate synergy, additivity, and antagonism, respectively. Sensitization Enhancement Ratio A linear regression of ln S against radiation dose D (Gy) was performed using all single‐irradiation data: ln S = α − βD. For each combination well with observed survival S comb and actual radiation dose D comb, the dose required to achieve the same survival with radiation alone was back‐calculated: Dsingle = (α − ln S comb)/β. The sensitization enhancement ratio (SER) was then calculated as follows: SER = D single/D comb. An SER > 1 indicated radiosensitization. Irradiation Conditions Irradiation (IR) was performed using 6 MV X‐rays generated by a medical linear accelerator (Varian Clinac iX, Varian Medical Systems, California, USA). Cells were irradiated at prescribed doses ranging from 2 to 8 Gy, with a source‐to‐surface distance (SSD) of 100 cm and a monitor‐unit rate of 600 MU/min. Cell Viability Assay Cells were plated in 96‐well plates at a density of 5000 cells per well during the logarithmic growth phase. Following a 48‐h incubation in complete medium supplemented with or without treatments (DMSO, PRGL493, erastin, RSL3, deferoxamine, or dihydroartemisinin), the cells were then incubated with a cell counting kit‐8 reagent (#K1018, APEXBIO, USA) at 37°C for 2 h. Absorbance was measured at 450 nm using a microplate reader, and the cell viability was calculated and normalized to that of the control group. Transwell Assay For the cell migration assay, PCa cells were seeded into transwell inserts at a density of 1 × 104 cells per well and cultured until they reached approximately 90% confluency. The medium in the upper chamber was replaced with a serum‐free medium, whereas the lower chamber was filled with complete medium containing 20% FBS. After 48 h of incubation, cells were fixed with 4% paraformaldehyde and stained with 1.5% crystal violet for 20 min, followed by five washes with phosphate‐buffered saline (PBS; #KGL2206‐500, KeyGEN). Non‐migrating cells on the upper surface of the membrane were gently removed using a cotton swab. The inserts were observed and photographed using a microscope (#IX51, Olympus, Japan). Measurement of Lipid Peroxidation PCa cells (4 × 104 cells/well) were plated in a 20‐mm confocal dish and cultured overnight to achieve adhesion. Following treatment, cells were washed with PBS, and incubated in serum‐free medium supplemented with 2.5 μM of BODIPY581/591 C11 (#D3861, Invitrogen, Massachusetts, USA). The cell suspension was incubated for 30 min at 37°C in a 5% CO2 atmosphere. Subsequently, the cells were washed thrice with PBS and incubated with Hoechst 33342 for 10 min for nuclear staining. After the final wash, images were acquired using a fluorescence microscope (#IX51, Olympus) at 100× magnification to visualize lipid peroxidation via green fluorescence. Chromatin Immunoprecipitation‐qPCR A total of 3.0 × 106 PCa cells were plated in a 15‐cm dish and incubated for 48 h prior to processing. Chromatin was crosslinked by treating cells with 1% formaldehyde for 15 min at 25°C, and the reaction was subsequently stopped with 125 mM glycine. The cells were then washed with ice‐cold PBS containing protease inhibitors and harvested. Chromatin was sheared to an average size of 200–500 bp by sonication (20 cycles of 5‐s pulses interspersed with 10‐s intervals at 50 W output) in ChIP sonication buffer. The fragment size distribution was confirmed by 1.5% agarose gel electrophoresis. For immunoprecipitation, 50 μg of sheared chromatin was incubated with an anti‐NR3C1 (#24050‐1‐AP, Proteintech Group) overnight at 4°C, followed by incubation with rProtein A/G MagPoly Beads (#BK0045‐01, ACE Biotechnology) for 2 h. Bead‐bound complexes were subjected to a series of stringent washes (high‐salt, low‐salt, and TE buffer). Following reversal of crosslinks at 65°C for 8 h with proteinase K and RNase, the immunoprecipitated DNA was purified. The enrichment of specific genomic regions was quantified by qPCR. Primers targeting sequences within the ACSL4 promoter region (spanning 3 kb upstream of the transcription start site) and a negative primer located distally from this region were designed as follows: ACSL4 promoter forward primer 1: 5′‐GGGAAATATTTTGCTCCTGGGC‐3′; ACSL4 promoter reverse primer 1: 5′‐GCTGTCTTCATCACCTACCACA‐3′; ACSL4 promoter forward primer 2: 5′‐CCTCTTCAACTTGGGTGCAGT‐3′; ACSL4 promoter reverse primer 2: 5′‐ATTGTCGGTCTCTTAGCCCAG‐3′; Negative forward primer: 5′‐CCCGTGAAGTTTAACACGCA‐3′; Negative reverse primer: 5′‐AGGCTGCTCTTGGCATTAGT‐3′. Luciferase Reporter Assay The firefly reporter plasmids were constructed using the pGL3 vector. The wild‐type and mutant ACSL4 promoter sequences containing the putative binding site were cloned into pGL3‐Basic, generating wild‐type pGL3‐Basic‐ACSL4 promoter and mutant pGL3‐Basic‐ACSL4 promoter, respectively (#PPL01602‐2c, #PPL01602‐2d, Public Protein/Plasmid Library, Nanjing, China). For each transfection, the firefly reporter plasmid (wild‐type or mutant) was co‐transfected with Renilla reporter plasmid (pRL‐TK) and an NR3C1 overexpression plasmid. Untransfected cells served as background controls, while pGL3‐Basic and empty pcDNA served as empty vector controls. The cells were harvested 24 h post‐transfection. Firefly and Renilla luciferase activities were measured using the Dual Luciferase Reporter Gene Assay Kit (#11402ES60, Yeasen, Shanghai, China) on a Tecan M200 microplate reader, and relative luciferase ratios were calculated according to the manufacturer's instructions. Ferrous Ion Assay To detect intracellular iron ions, PCa cells were seeded in a 15‐mm confocal dish at a density of 3 × 104 cells per well and incubated overnight. The cells were subsequently treated with DHA (1 μM) for 24 h. After treatment, the cells were washed with PBS and incubated with 5 μM Fe2+ indicator (FeRhoNox‐1, #MX4558, Maokangbio, Shanghai, China) for 1 h. The cells were then washed thrice with PBS and incubated with Hoechst 33342 for 10 min. After washing five times with PBS, the cells were imaged using a fluorescence microscope (#IX51, Olympus) at 40× magnification. To quantify ferrous ions, PCa cells were seeded in a 60‐mm culture dish at a density of 6 × 105 and incubated for 24 h. The cells were subsequently harvested, and iron ions were quantified using a ferrous ion assay kit (#S1068S, Beyotime). Briefly, the cells were lysed in lysis buffer in an ice‐water bath, and a ferrous ion probe was added to the extracted supernatant to produce a colored compound. The absorbance of each colored compound was measured at 593 nm. Standard solutions of ferrous ions at different concentration gradients were used to plot the standard curve. The ferrous ion concentration in the sample was calculated according to the standard curve and total protein concentration of the sample. Reduced Glutathione Assay PCa cells were seeded in a 100‐mm culture dish at a density of 4 × 106 cells per dish and cultured overnight. Prior to irradiation, the cells were pretreated with 1 μM DHA for 2 h and then continuously incubated for an additional 22 h after IR. The cells were harvested and washed with PBS. The glutathione (GSH) levels were measured using a GSH and GSSG Assay Kit (#S0053, Beyotime). Transmission Electron Microscopy PCa cells were seeded in a 100‐mm culture dish at a density of 4 × 106 cells per dish and incubated with DHA for 24 h. The cells were harvested and washed with PBS. To observe mitochondrial structure, the cells were fixed in a solution containing 2.5% glutaraldehyde in 0.1 M PBS (pH 7.4). After fixation, samples were washed thrice with 0.1 M PBS for 15 min each. Post‐fixation was performed with 1% osmium tetroxide for 2 h at room temperature, followed by another three washes with 0.1 M PBS for 15 min each. Dehydration was performed using a graded ethanol series (50%, 75%, 80%, 95%, 100% I, and 100% II) followed by propylene oxide (100% I and 100% II) at each step for 15 min. The samples were infiltrated with mixtures of propylene oxide and SPI 812 resin (1:1 for 2 h at room temperature and then 1:2 overnight at room temperature), followed by two changes of pure resin for 2 h each. The specimens were embedded in fresh resin and polymerized in an oven at 37°C and then 60°C for 48–72 h. Ultrathin sections (60–80 nm) were cut using a Leica UC7 ultramicrotome with a Daitome diamond knife (45°) and collected on copper grids. Sections were stained with 2% uranyl acetate for 30 min in the dark, rinsed with deionized water, and stained with lead citrate for 15 min. After rinsing and drying, the sections were observed under a JEM 1400 transmission electron microscope, and images were acquired for analysis. In Vivo Validation Animal experiments were conducted following a protocol approved by the Institutional Animal Care and Use Committee (IACUC) of the Research Committee of Nanjing Medical University and Nanjing First Hospital (No. IACUC‐2112040, No. DWSY‐26014783). To observe the radiation sensitivity of PCa with different ACSL4 expression levels, 5‐week‐old male athymic BALB/c nude mice (n = 5; GemPharmatech Co. Ltd., Nanjing, China) were subcutaneously injected with 6 × 106 LNCaP cells or ACSL4‐overexpressing LNCaP cells suspended in a 1:1 mixture of cell suspension and high‐concentration Matrigel from NEST Biotechnology (GelNEST, #211352, Wuxi, China), with a final injection volume of 100 μL per mouse. When the tumor volume reached 120 mm3, the mice were irradiated with three fractions of 2 Gy localized irradiation every 4 days. The tumor volume was measured every 2 days, and a tumor growth curve was drawn. To validate the radiosensitizing function of dihydroartemisinin, animals were randomly assigned to the experimental groups by a researcher using a simple randomization method (manual drawing of the subjects from their housing cages). Five‐week‐old male athymic BALB/c nude male mice (n = 5; GemPharmatech Co. Ltd.) were injected subcutaneously with 3 × 106 PCa cells suspended in a 1:1 mixture of cell suspension and high‐concentration Matrigel from NEST Biotechnology (GelNEST, #211352), with a final injection volume of 100 μL per mouse. When the tumor volume reached 100 mm3, the mice were irradiated with four fractions of 2 Gy localized irradiation every 2 days, with or without oral gavage administration of DHA (25 mg/kg body weight [21]) administered 2 h before each irradiation. Tumor volume and body weight were measured every 2 days from Days 1 to 9, and every day from Days 10 to 16. The tumors were excised 4 h after the last irradiation for iron ion and malondialdehyde (MDA) analyses. Bioinformatics Database The Gene Expression Omnibus (GEO) dataset GSE134499 was used to identify differentially expressed genes between radiotherapy‐resistant and radiotherapy‐sensitive prostate cancer cells. The top 50 significantly upregulated and downregulated genes were selected for Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses, respectively. Correlations between the expression levels of candidate genes (NR3C1, ANXA1, ANXA2, AREG, and NR0B1) and biochemical recurrence (BCR)–free survival were assessed using the GSE70770 dataset from GEO. Next, the GSE116918 dataset from GEO was analyzed to validate the correlation between ACSL4 expression and BCR time following radical radiotherapy. Additionally, an interaction network was constructed using the STRING database based on the top 50 genes identified from the Cancer Genome Atlas (TCGA) dataset that exhibited the strongest correlation with NR3C1 expression in prostate cancer. Finally, the JASPAR database was used to predict the potential transcription binding sites for NR3C1 within the promoter region of ACSL4. Molecular Docking The chemical structure of dihydroartemisinin was obtained from the PubChem database. The three‐dimensional structure of the target protein, NR3C1, was obtained from the Protein Data Bank and the AlphaFold Protein Structure Database. Protein‐ligand docking analysis was subsequently performed using the CB‐Dock2 web server [22, 23]. Statistical Analyses All experiments were performed with at least three replicates, and data are presented as the mean ± SD. Statistical significance was determined using an unpaired Student's t‐test or one‐way ANOVA, followed by Dunnett's or Bonferroni's multiple comparison tests using GraphPad Prism (San Diego, CA, USA), with a p‐value < 0.05 considered significant. Results NR3C1 Overexpression Impairs Radiosensitivity in Prostate Cancer To identify the key pathways influencing radiosensitivity in PCa, we analyzed the GEO dataset GSE134499 to compare the gene expression profiles of radiosensitive and radioresistant cell lines. Enrichment analysis of the top 50 upregulated and downregulated genes revealed a significant enrichment of glucocorticoid receptor‐related signaling pathways in radioresistant cells (Figure 1a). Among the molecules involved in glucocorticoid signaling, NR3C1, ANXA1, ANXA2, AREG, and NR0B1 were significantly upregulated in the radioresistant cells (Figure 1b). We examined the protein levels of these candidates in four PCa cell lines with varying malignancies (LNCaP, 22RV1, DU145, and PC3) and assessed their radiosensitivity using CCK‐8 assays. NR3C1 protein expression increased progressively in LNCaP, 22RV1, PC3, and DU145 cells (Figure 1c), whereas their radiosensitivity decreased correspondingly (Figure 1d). Analysis of the clinical GEO dataset GSE70770 showed that high NR3C1 expression was significantly associated with shorter BCR‐free survival (Figure S1a–e). Among the knockdown of NR3C1, ANXA1, ANXA2, AREG, and NR0B1 in DU145 and PC3 cells, only siNR3C1 most potently enhanced radiation‐induced cell death in both cell lines, underscoring the critical role of NR3C1 in modulating the radiosensitivity of PCa cells (Figure S1f–i). NR3C1 silencing increased the radiosensitivity of DU145 cells (Figure 1e). These findings suggest an inverse correlation between NR3C1 expression and radiosensitivity in prostate cancer. NR3C1 Expression Is Inversely Correlated with Radiosensitivity in PCa. (a) Pathway enrichment analysis of the top 50 differentially expressed genes between radioresistant and radiosensitive PCa cell lines from the GEO dataset GSE134499. (b) Relative mRNA expression levels of the significantly differentially expressed genes NR3C1, ANXA1, ANXA2, AREG, and NR0B1 which are involved in glucocorticoid receptor signaling regulation, in radioresistant versus radiosensitive PCa cell lines from the GEO dataset GSE134499. (c) Western blot analysis of NR3C1, ANXA1, ANXA2, AREG, and NR0B1 protein levels in PCa cell lines LNCaP, 22Rv1, DU145, and PC3. (d) The radiosensitivity of PCa cell lines LNCaP, 22Rv1, DU145, and PC3 was compared using a colony formation assay. (e) Changes in radiosensitivity following NR3C1 knockdown in DU145 cells were assessed by a colony formation assay. ***p < 0.001. NR3C1 Positively Regulates ACSL4 Expression Previous clinical studies have indicated that blocking NR3C1 does not significantly improve the prognosis of patients with prostate cancer [14], potentially because glucocorticoid receptor signaling regulates an extensive downstream network, making precise functional blockade challenging. To address this, we aimed to pinpoint specific NR3C1 downstream molecules critical for radiosensitivity. Analysis of the top 50 genes coexpressed with NR3C1 in prostate cancer identified SMAD3, STAT5B, PBX1, and ACSL4 as potential direct targets (Figure S2a). The initial screening of mRNA levels in LNCaP, 22RV1, PC3, and DU145 cells revealed a positive correlation between NR3C1, ACSL4, and SMAD3 (Figure 2a). We hypothesized that NR3C1 directly regulates one or more of these candidates. Using gain‐ and loss‐of‐function models (NR3C1‐overexpressing 22RV1 and NR3C1‐knockdown DU145 cells, respectively) (Figure S2b,c), we found that ACSL4 protein levels consistently changed with NR3C1 manipulation, indicating positive regulation (Figure 2b,c). As NR3C1 is a transcription factor [12], we bioinformatically predicted its binding sites in the ACSL4 promoter (Figure 2d) and validated its direct binding using ChIP‐qPCR assays, confirming NR3C1‐mediated transcriptional regulation of ACSL4 (Figure 2e). To further validate the direct transcriptional regulation of ACSL4 by NR3C1, we constructed luciferase reporter plasmids carrying either the wild‐type or a mutant version of the putative NR3C1‐binding site within the ACSL4 promoter (Figure 2f). Relative luciferase activity was significantly higher for the wild‐type construct (hACSL4‐WT) than for the mutant construct (hACSL4‐MUT) (Figure 2g), indicating that NR3C1 binds to the ACSL4 promoter and directly regulates its transcription. NR3C1 transcriptionally regulates ACSL4 in PCa. (a) qRT‐PCR was performed to measure the relative mRNA expression levels of NR3C1, ACSL4, PBX1, SMAD3, and STAT5B in PCa cell lines LNCaP, 22Rv1, DU145, and PC3. (b) Following NR3C1 overexpression in 22Rv1 cells, western blot analysis was used to detect the protein levels of ACSL4, PBX1, SMAD3, and STAT5B. (c) Following NR3C1 knockdown in DU145 cells, western blot analysis was used to detect the protein levels of ACSL4, PBX1, SMAD3, and STAT5B. (d) Potential binding sites for the NR3C1 transcription factor within the ACSL4 gene promoter region were predicted using the JASPAR database. (e) Chromatin immunoprecipitation (ChIP) assay was conducted to validate the binding of NR3C1 to the ACSL4 promoter region. The DNA fragments immunoprecipitated by NR3C1 in the ChIP assay were quantified using qPCR. (f) Construction of wild‐type and mutant ACSL4 promoter luciferase reporter plasmids. (g) Relative firefly luciferase activity of the wild‐type (hACSL4‐WT) and mutant (hACSL4‐MUT) ACSL4 promoter constructs. *p < 0.05, **p < 0.01, ***p < 0.001; ns, no significant difference. ACSL4 Contributes to Malignancy and Radioresistance in PCa To investigate the function of ACSL4 in PCa, we first treated PCa cells with the ACSL4‐specific inhibitor, PRGL493. Cell growth assessment using the CCK‐8 assay revealed that ACSL4 inhibition significantly suppressed PCa cell proliferation (Figure 3a). Transwell migration assays showed that ACSL4 inhibition markedly reduced the migratory capacity of cells (Figure 3b). To further evaluate the effect of ACSL4 on radiosensitivity, we generated ACSL4‐knockdown DU145 cells (Figure S3a) and ACSL4‐overexpressing 22RV1 (Figure S3b). Colony formation assays showed that ACSL4 silencing significantly enhanced the cellular radiosensitivity (Figure 3c, Figure S3c). Conversely, ACSL4 overexpression in 22RV1 cells promoted cell growth, migration, and radioresistance (Figure 3d–f, Figure S3d). ACSL4 promotes cell proliferation, migration, and radioresistance in PCa. (a) Cell proliferation was assessed using a CCK‐8 assay to determine the effect of the ACSL4‐specific inhibitor PRGL493 on the growth of DU145 and PC3 cells. (b) A transwell assay was used to examine the impact of PRGL493 on the migratory ability of DU145 and PC3 cells. (c) A colony formation assay was employed to assess radiosensitivity of DU145 cells following ACSL4 knockdown. (d) A CCK‐8 assay was performed to analyze the proliferation of 22Rv1 cells after ACSL4 overexpression. (e) A transwell assay was conducted to evaluate the migration of 22Rv1 cells following ACSL4 overexpression. (f) A colony formation assay was used to determine the radiosensitivity of 22Rv1 cells after ACSL4 overexpression. (g) Tumor growth curves of LNCaP xenografts with or without ACSL4 overexpression in nude mice following radiotherapy. (h) Time to biochemical recurrence (BCR) after radical radiotherapy in patients with high versus low ACSL4 expression from the GEO dataset GSE116819. To investigate whether ACSL4 correlates with radiosensitivity in AR‐positive cell lines and in vivo, we stably overexpressed ACSL4 in LNCaP cells using lentiviral infection (Figure S3e). LNCaP cells transfected with an empty vector were used as controls. The cells were injected subcutaneously into nude mice to establish xenograft tumors and assess their radiosensitivity. Tumor growth curves showed that the radiosensitivity of tumors in the ACSL4 overexpression group was significantly reduced compared with the control group (Figure 3g), suggesting that ACSL4 expression levels correlate with radiosensitivity in both AR‐positive PCa cells and in vivo. Furthermore, in the GEO dataset GSE116918, patients with high ACSL4 expression (relative expression higher than 8.5) had a significantly shorter time to biochemical recurrence after radical radiotherapy than those with low ACSL4 expression (Figure 3h). These findings indicate that ACSL4 expression is clinically associated with the efficacy of radiotherapy in patients with prostate cancer. Ferroptosis Inducers Restore Radiosensitivity in ACSL4‐High PCa As a classic ferroptosis‐promoting gene, high ACSL4 expression increases the abundance of PUFA in cell membranes, which supplies the necessary substrate for ferroptosis and consequently confers heightened sensitivity to its inducers [18]. Across the four PCa cell lines examined, ACSL4 expression increased progressively in the following order: LNCaP, 22Rv1, PC3, and DU145 (Figure S4a). Therefore, we assessed the sensitivity of these cell lines to ferroptosis inducers. The results revealed that PC3 and DU145 cells, which expressed ACSL4 at very high levels, exhibited significantly greater sensitivity to ferroptosis inducers (erastin, RSL3) than 22Rv1 and LNCaP cells with lower ACSL4 expression (Figure 4a,b). The overexpression of ACSL4 in 22Rv1 cells further enhanced their sensitivity to ferroptosis inducers (Figure 4c). Clonogenic survival assays demonstrated that ferroptosis inducers significantly increased the radiosensitivity of 22Rv1, PC3, and DU145 cells, with more pronounced effects in PC3 and DU145 cells (Figure S4b, Figure 4d,e). Staining with a lipid peroxidation probe showed that the combined treatment with a ferroptosis inducer and irradiation markedly elevated lipid peroxidation levels. Furthermore, ferroptosis induced lipid peroxidation was higher in PC3 and DU145 cells than that in 22Rv1 cells (Figure S4c, Figure 4f,g). The radiosensitizing effect of the ferroptosis inducer was substantially reversed by the specific ACSL4 inhibitor, PRGL493 (Figure S4d). Taken together, these findings indicate a positive correlation between ACSL4 expression levels and both sensitivity to ferroptosis inducers and their efficacy in radiosensitizing PCa cells. Ferroptosis inducers enhance radiosensitivity in PCa cells with high ACSL4 expression. (a, b) The CCK‐8 assay was performed to compare the sensitivity of LNCaP, 22Rv1, DU145, and PC3 cells to the ferroptosis inducers Erastin and RSL3. (c) A CCK‐8 assay was used to measure sensitivity to Erastin and RSL3 following ACSL4 overexpression in 22Rv1 cells. (d, e) A colony formation assay was employed to evaluate the radiosensitivity of DU145 and PC3 cells treated with a combination of the ferroptosis inducers (Erastin or RSL3) and irradiation. (f, g) Lipid peroxidation levels in DU145 and PC3 cells treated with Erastin or RSL3 in combination with irradiation were measured using the BODIPY 581/591 C11 probe. *p < 0.05, **p < 0.01, ***p < 0.001; ns, no significant difference. Dihydroartemisinin Synergizes With ACSL4 to Induce Ferroptosis and Radiosensitize PCa Drugs, such as sorafenib and artemisinin, induce ferroptosis [24, 25]. Specifically, active derivatives of artemisinin can trigger ferroptosis by targeting HSD17B4, a protein essential for lipid metabolism, and directly induce lipid peroxidation [25]. We performed molecular docking of dihydroartemisinin with the HSD17B4 protein, and the results revealed a potential binding energy of −9.0 Kcal/mol between dihydroartemisinin and HSD17B4 (Figure S5a), suggesting that dihydroartemisinin may also target HSD17B4, leading to lipid peroxidation. Therefore, we used dihydroartemisinin to explore its synergistic radiosensitizing effects. CCK8 assay results demonstrated that radiotherapy‐resistant PCa cells with high ACSL4 expression were sensitive to dihydroartemisinin (Figure 5a). Clonogenic assays further showed that dihydroartemisinin significantly enhanced the radiosensitivity of radiotherapy‐resistant PCa cells (Figure 5b,c, Figure S5b,c). Additionally, the combination of dihydroartemisinin and radiotherapy markedly increased lipid peroxidation levels in PCa cells (Figure 5d, Figure S5d). Furthermore, artemisinin significantly upregulated ferrous ion levels (Figure 5e) and reduced GSH content (Figure 5f), GPX4 expression (Figure S5e), and disrupted mitochondrial structure (Figure 5g) in PCa cells, thereby potently inducing ferroptosis. However, the ferroptosis inhibitor deferoxamine reversed the radiosensitizing effect of dihydroartemisinin (Figure 5h). These findings indicated that dihydroartemisinin cooperatively induces ferroptosis and sensitizes radiotherapy in PCa cells with high ACSL4 expression. This conclusion was validated in a nude mouse model (Figure 6a–e). Furthermore, the DHA concentration used in this study exhibited low toxicity to the normal prostate‐derived cell lines WPMY‐1 and RWPE‐1 (Figure 6f) and no significant effect on nude mouse body weight (Figure 6g), indicating that the DHA dosage regimen employed here is minimally toxic to normal cells and tissues, both in vitro and in vivo. A diagram of the mechanism discussed in this study is presented in Figure 7. Dihydroartemisinin enhances radiosensitivity in PCa by inducing ferroptosis. (a) A CCK‐8 assay was performed to assess the sensitivity of PC3 and DU145 cells to dihydroartemisinin (DHA). (b) A colony formation assay was used to evaluate the radiosensitizing effect of DHA (1 μM) in combination with irradiation (IR) on DU145 cells. (c) Sensitization enhancement ratio (SER) and bliss synergy ratio (BSR) values derived from the DHA and IR dose matrix for radiosensitization assessment. (d) Lipid peroxidation levels in DU145 cells treated with DHA (1 μM) in combination with IR were measured using the BODIPY 581/591 C11 probe. (e) Ferrous ion (Fe2+) staining using a fluorescent probe and quantitative analysis. (f) Measurement of glutathione (GSH) levels. (g) Mitochondrial morphology observed by transmission electron microscopy. Yellow arrows indicate mitochondria. (h) A CCK‐8 assay was conducted to investigate the effect of deferoxamine (DFO) on the radiosensitizing efficacy of DHA. *p < 0.05, **p < 0.01, ***p < 0.001; ns, no significant difference. Dihydroartemisinin enhances radiosensitivity in a nude mouse xenograft model. (a) Effect of DHA and irradiation (IR) on tumor size in nude mouse xenografts, with a photograph of resected tumors. (b) Effect of DHA and IR on tumor growth in nude mouse xenografts, presented as tumor growth curves. (c) Tumor weights. (d) Tumor iron content measured using a tissue iron assay kit. (e) Tumor lipid peroxidation levels measured using an MDA assay kit. (f) IC50 determination of DHA in normal prostate‐derived cell lines WPMY‐1 and RWPE‐1. (g) Body weight changes in nude mice in DHA‐treated and untreated groups. *p < 0.05, **p < 0.01, ***p < 0.001. Targeting the NR3C1‐ACSL4 Axis attenuates malignant progression and radioresistance in PCa. In PCa, NR3C1 drives tumor proliferation, migration, and radioresistance by transcriptionally upregulating ACSL4. Treatment with either the ACSL4‐specific inhibitor PRGL493 or the ferroptosis inducer DHA effectively reverses this radioresistance. Discussion NR3C1 is associated with a poor prognosis in PCa [13]. Previous studies indicate that it can partially compensate for androgen receptor (AR) function, representing a key adaptive mechanism contributing to endocrine therapy resistance [10, 26]. In addition to promoting endocrine therapy resistance in PCa, NR3C1 has also been found to be involved in malignant progression, therapy resistance, and immunotherapy evasion in other cancers [9, 27, 28, 29]. Consequently, NR3C1 may have multifaceted and complex functions in tumor biology. However, the role of NR3C1 in the radiosensitivity of PCa remains unclear. Here, we show that high NR3C1 expression is strongly correlated with reduced radiosensitivity in PCa. However, clinical trials targeting NR3C1 have not shown consistent patient benefits [14]. Therefore, there is a need to identify therapeutically actionable downstream molecules of NR3C1. ACSL4, a central enzyme in lipid metabolism, drives the accumulation of PUFAs when upregulated [20]. Recent evidence indicates that ACSL4‐mediated PUFA biosynthesis provides essential substrates for ferroptosis, a regulated form of cell death [18]. Consequently, tumors exhibiting high ACSL4 expression may be particularly vulnerable to ferroptosis inducers that enhance therapeutic response [30]. Interestingly, ACSL4 also serves as a key regulator of fatty acid metabolism and is capable of remodeling the phospholipid composition of cell membranes and modulating steroidogenesis [31]. In this study, we demonstrated that the glucocorticoid receptor, NR3C1, transcriptionally regulates ACSL4 in aggressive PCa, thereby contributing to radioresistance. Importantly, the ferroptosis inducers Erastin and RSL3 substantially restored radiosensitivity in these resistant models. However, neither Erastin nor RSL3 is currently approved for clinical use, highlighting the need to identify clinically applicable ferroptosis‐promoting agents for translation into therapeutic strategies. Numerous studies have established that ferroptosis induction can sensitize cancer cells to both radiotherapy and chemotherapy [16]. Recent studies have uncovered several novel ferroptosis‐based therapeutic strategies, including multifunctional Fe‐EGCG@RSL3 nanomedicine that synergistically induces ferroptosis, biomimetic β‐lapachone and hemin nanoassemblies for ferroptosis‐related tumor nanomedicine, and a dual‐channel electron expressway that triggers a piezocatalytic ferroptosis‐mitochondrial catastrophe cascade [32, 33, 34]. Additionally, considerable research has focused on identifying clinically approved drugs with ferroptosis‐inducing properties as viable candidates for therapeutic sensitization. Examples include sorafenib, sulfasalazine, and artemisinin derivatives, which promote ferroptosis via distinct mechanisms [35]. Combining such compounds with standard therapies offers a promising strategy for exploiting the metabolic vulnerability of treatment‐resistant tumors [36]. This is particularly relevant for highly aggressive malignancies that often exhibit enriched PUFA metabolism or upregulated lipid peroxidation pathways [37]. Notably, active artemisinin derivatives have been shown to induce ferroptosis by targeting HSD17B4, a protein essential for lipid metabolism, and directly stimulating lipid peroxidation [25]. In line with this evidence, our study demonstrates that DHA similarly targets HSD17B4 and acts synergistically with ACSL4 to exacerbate lipid peroxidation in aggressive PCa cells, thereby triggering ferroptosis and enhancing radiosensitivity. In conclusion, our study defined the NR3C1‐ACSL4 axis as a dual regulator that promotes radioresistance while concurrently enhancing ferroptosis sensitivity in PCa cells. This duality raises the possibility of a biomarker‐guided strategy, in which high ACSL4 expression could guide the addition of ferroptosis inducers such as DHA to radiotherapy, although this approach requires further preclinical and clinical validation. Author Contributions Jing Yang: methodology, data curation, investigation, visualization, formal analysis, writing – original draft, writing – review and editing. Puyuan Chen: methodology, investigation, formal analysis, visualization. Yawen Zheng: methodology, investigation, formal analysis. Feiyue Shi: resources. Min Wang: resources. Luwei Xu: supervision, resources, project administration. Xiaowei Wei: writing – original draft, resources, supervision, project administration, funding acquisition. Xiumei Wang: funding acquisition, writing – original draft, writing – review and editing, methodology, investigation, formal analysis, data curation, visualization. Funding This work was funded by the National Natural Science Foundation of China (No. 82473435) and the Natural Science Foundation of Jiangsu Province (No. BK20231127) to Xiaowei Wei along with the Nanjing Science and Technology Development Foundation (No. YKK24120) to Xiumei Wang. Ethics Statement The Institutional Animal Care and Use Committee of Nanjing Medical University and Nanjing First Hospital approved all animal procedures (No. IACUC‐2112040 and No. DWSY‐26014783), and experiments were carried out following the approved protocol. Consent The authors have nothing to report. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Figure S1: NR3C1 is Associated with biochemical recurrence and radiosensitivity in PCa. (a–e) Correlation analysis between the expression levels of NR3C1, ANXA1, ANXA2, AREG, and NR0B1 and time to biochemical recurrence (BCR) in the GEO clinical dataset GSE70770; (f, g) qRT‐PCR validation of siRNA‐mediated knockdown efficiency for NR3C1, ANXA1, ANXA2, AREG, and NR0B1 in DU145 and PC3 cells; (h, i) A CCK‐8 assay was used to evaluate the radiosensitivity of DU145 and PC3 cells following siRNA‐mediated knockdown of NR3C1, ANXA1, ANXA2, AREG, and NR0B1. *p < 0.05, **p < 0.01, ***p < 0.001, and ns, no significant difference. Figure S2: Identification of NR3C1‐Related Molecules in PCa. (a) A protein–protein interaction network of the top 50 genes correlated with NR3C1 expression in prostate cancer clinical samples from the TCGA database was constructed using the STRING database. (b) Western blot analysis was performed to detect NR3C1 protein levels following NR3C1 overexpression in 22Rv1 cells. (c) Western blot analysis was performed to detect NR3C1 protein levels following NR3C1 knockdown in DU145 cells. Figure S3: Validation of ACSL4 Knockdown and Overexpression in PCa. (a) Western blot analysis was performed to detect ACSL4 protein levels following ACSL4 knockdown in DU145 cells. (b) Western blot analysis was performed to detect ACSL4 protein levels following ACSL4 overexpression in 22Rv1 cells. (c) A colony formation assay was used to assess the radiosensitivity of DU145 cells following ACSL4 knockdown. (d) A colony formation assay was used to assess the radiosensitivity of 22Rv1 cells following ACSL4 overexpression. (e) LNCaP cells stably overexpressing ACSL4 were generated via lentiviral transduction, as indicated by green fluorescence. Figure S4: ACSL4 Expression Modulates the Radiosensitizing Efficacy of Ferroptosis Inducers in PCa. (a) Western blot analysis was performed to compare the basal ACSL4 protein levels in LNCaP, 22Rv1, DU145, and PC3 cells. (b) A colony formation assay was used to assess the radiosensitivity of 22Rv1 cells treated with the ferroptosis inducers Erastin or RSL3 in combination with irradiation. (c) Lipid peroxidation levels in 22Rv1 cells treated with the ferroptosis inducers Erastin or RSL3 in combination with irradiation were detected using the BODIPY 581/591 C11 probe. (d) A CCK‐8 assay was performed to analyze changes in the radiosensitizing effects of ferroptosis inducers following ACSL4 inhibition with PRGL493 in PC3 cells. *p < 0.05, **p < 0.01, ***p < 0.001, and ns, no significant difference. Figure S5: Dihydroartemisinin Targets HSD17B4 to Promote Ferroptosis in PCa. (a) Schematic representation of the predicted binding mode and binding energy of DHA with HSD17B4 based on molecular docking. (b) A colony formation assay was used to assess the radiosensitivity of DU145 cells after treatment with DHA (1 μM). (c) A colony formation assay was used to assess the radiosensitizing effect of DHA across the DHA and irradiation (IR) dose matrix in DU145 cells. (d) Lipid peroxidation levels in PC3 cells treated with DHA (1 μM) in combination with IR were measured using the BODIPY 581/591 C11 probe. (e) Western blot analysis was performed to compare GPX4 protein levels in PCa cells.