FIN56-induced ferroptosis suppresses the development of endometriosis by augmenting mitochondrial ROS and lipid peroxidation via the ACACA/ARID5A/NOX4 axis

Endometriosis (EMs) is one of the most common chronic gynecologic diseases and affects ~15% of females of reproductive age worldwide. [ 1 – 3 ] Though it is known to be dependent on sex steroid hormones, the pathophysiology of EMs is not well understood. In addition, current nonsurgical treatment options for EMs have clear side effects, and the rate of symptom recurrence is high. [ 4 ] Thus, novel and effective therapies for the etiology of EMs are urgently needed to improve the treatment of EMs. Ferroptosis, a recently identified mechanism of programmed cell death mediated by iron and lipid peroxidation accumulation, is clearly distinct from other forms of programmed cell death, such as apoptosis and autophagy. [ 5 ] Dysregulation of ferroptosis is associated with various pathological conditions and diseases, such as cancer and neurodegenerative diseases. [ 6 , 7 ] To date, three classes of ferroptosis inducers (FINs) have been identified: Class 1 FINs (such as erastin), class 2 FINs (such as RAS-selective lethal small molecule 3 [RSL3] and ML162), and class 3 FINs (FIN56). [ 8 ] Despite extensive investigations on the role of class 1 and class 2 FINs in tumorigenesis and progression, class 3 FIN (FIN56), a novel FIN, has rarely been studied. A primary defect in EMs is abnormal eutopic endometrium characterized by resistance to ferroptosis, which allows cells to spread via retrograde menstruation to survive, implant, and establish endometriotic lesions within the abdominal cavity. [ 9 ] However, the molecular mechanisms underlying the survival of endometriotic lesions under iron overload and ferroptosis remain unclear. Lipogenic enzyme acetyl-CoA carboxylase 1 (ACC1) reportedly plays a crucial role in cancer invasion and metastasis. [ 10 – 12 ] ACC1 deficiency was indispensable for the epithelial-mesenchymal transition and metastasis induction in breast tumors. [ 11 ] Knockdown of ACACA compensates for AMP-activated protein kinase (AMPK) activation and facilitates anchorage-independent growth and solid tumor formation in vivo . [ 10 ] Accumulating evidence indicates a close link between ACC1-fatty acid synthase (FAS)-mediated fatty acid biogenesis and ferroptosis. [ 5 , 13 ] ACC1, encoded by ACACA , catalyzes the carboxylation of acetyl-CoA to generate malonyl-CoA, which is required for the synthesis of some polyunsaturated fatty acids (PUFAs) and therefore for ferroptosis. [ 14 ] Previous studies have shown that inactivation of ACC1 drives resistance to ferroptosis induced by a series of FINs, such as erastin, RSL3, and FIN56. [ 15 – 18 ] Although Shimada et al [ 14 ] have suggested that 5-(tetradecyloxy)-2-furoic acid (TOFA), an allosteric inhibitor of ACC1, is a potent suppressor of FIN56 and suppresses lipid reactive oxygen species (ROS) generation after FIN56 treatment, the detailed mechanism linking FIN56 to ACC1 is still unclear. Considering the tumor-like characteristics of EMs, we hypothesized that ACC1 may play a vital role in its initiation and progression. However, the precise characteristics and mechanisms of ACC1 and ferroptosis regulation in the ectopic lesion microenvironment remain to be clarified. Nicotinamide adenine dinucleotide phosphate (NADPH) oxidase (NOX)-mediated mitochondrial damage has been found to be involved in lipid peroxidation and iron homeostasis imbalance in various pathological conditions and diseases, such as in neurodegenerative diseases, human beta-cells, and vascular endothelial inflammation, which contributes to ferroptosis-dependent cytotoxicity. [ 6 , 19 , 20 ] In EMs, the ROS-buffering capacity of ectopic endometrial stromal cells (ecESCs) is exhausted to protect against oxidative damage; thus, these cells tend to be more sensitive to oxidants. [ 21 ] However, much remains unknown about the intracellular ROS determinants of ferroptosis in EMs. Moreover, the mechanisms by which NADPH oxidase (NOX) regulates ferroptosis through oxidative stress via the impairment of mitochondrial metabolism in EMs remain unclear. In EMs, ecESCs have been shown to be sensitive to erastin, which may be used as a new therapeutic strategy for EMs. [ 22 – 24 ] However, the role of other FINs, such as FIN56, in EMs remains poorly understood. Thus, the role of ferroptosis and the specific mechanism of ferroptosis regulation in EMs require further elucidation. On the basis of these previous findings, we aimed to evaluate whether the novel ferroptosis inducer FIN56 could potentially treat EMs by triggering ferroptosis in ectopic lesions and whether ACC1 and NOX4 participate in this process. We also aimed to identify the downstream pathways activated after ferroptosis induction in EMs.

This work was supported by grants from the National Natural Science Foundation of China (No. 81901455) and the Beijing Natural Science Foundation (No. 7194325).

All tissue samples were obtained with informed patient consent, and all experiments followed the ethical principles outlined by the 1975 Helsinki Declaration and its later amendments or comparable ethical standards. The experimental protocol was approved by the Institutional Review Board of Peking University (No. 2020 [518]). All procedures involving animals were approved by the Animal Care and Use Committee of Peking University First Hospital (No. 2022105) and performed in accordance with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (Eighth Edition). Ectopic endometrial tissues (EC) from the cyst walls of ovarian endometriomas and eutopic endometrial tissues (EU) were obtained from 20 women with EMs immediately after surgery, resulting in 20 self-controlled pairs. Detailed information can be obtained from Supplementary Table 1, http://links.lww.com/CM9/C574 . For the non-EMs group, eight normal eutopic endometrial tissues (NC) were collected from patients who underwent combined laparoscopy and hysteroscopy for tubal infertility and simple ovarian cysts. All patients (aged 23–42 years) had regular menstrual cycles and were free of hormonal medications for at least 3 months before surgery. The diagnoses for all the samples were histologically confirmed. Patients who were pregnant or had diabetes mellitus, hypertension, autoimmune disorders, or cancer were excluded from this study. These samples were collected during the proliferative phase and the secretory phase using both patient-reported last menstrual period date and histological criteria. All tissue samples were transferred to the laboratory and further processed. Portions of the samples were stored in Gibco Hank’s solution (24020117, Thermo Fisher Scientific, Waltham, MA, USA) for cell isolation under sterile conditions. The remaining samples were fixed in 4% paraformaldehyde and embedded in paraffin for histological analysis. Human ecESCs from ovarian endometriomas and eutopic endometrial stromal cells (euESCs) from eutopic endometrium were isolated from 20 tissue samples of women with EMs using the protocol previously described by Ryan et al [ 25 ] with minor modifications. Briefly, endometrial tissues were washed with Gibco phosphate-buffered saline (PBS) (10010023, Thermo Fisher Scientific) and then cut into pieces. Then, the pieces were digested with collagenase type I (1 mg/mL; C0130, Sigma–Aldrich, St. Louis, MO, USA) and deoxyribonuclease I (0.1 mg/mL; DN25, Sigma–Aldrich). Dissociated tissues were sequentially filtered through 70- and 20-μm nylon meshes to remove epithelial cells. Stromal cells were then harvested and cultured to confluence in Gibco DMEM/F12 (11330032, Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (F8687, Sigma–Aldrich). The purity of the stromal cells was >98%, as confirmed by immunofluorescence staining for vimentin, a specific marker of stromal cells. [ 26 ] The cells reached confluence after ~3 days, and cells at passages 3–4 were used for experiments. Detailed reagents, cell viability measurements, lipid peroxidation assay, [ 27 ] transmission electron microscopy (TEM) analysis, RNA extraction and reverse transcription quantitative polymerase chain reaction (RT–qPCR), Western blotting analysis, cell transfection and small-interfering RNA knockdown, Cell transfection and plasmid overexpression, next-generation RNA sequencing (RNA-seq), mitochondrial membrane potential assay, mitochondrial ROS (mtROS) production analysis, ATP assay, measurement of intracellular iron levels, immunofluorescence analysis, and chromatin immunoprecipitation (ChIP) assay can be found in the Supplementary materials, http://links.lww.com/CM9/C574 . Six- to eight-week-old C57BL/6N female mice, 16–18 g in body weight, were purchased from Vital River Laboratory (Beijing Vital River Laboratory Animal Technology Co., China). The mice were acclimated for at least 1 week prior to surgery. Endometriotic lesions were surgically induced by autotransplantation of a homologous uterine horn onto the peritoneal wall. The mice were anesthetized via isoflurane laparotomy, and the left uterine horn was exposed and excised. [ 28 , 29 ] This piece of tissue was placed in phosphate‑buffered saline (PBS). Subsequently, it was cut into two square pieces of 4 × 4 mm 2 . The fragments were sutured on each right and left peritoneal surface area, with the endometrial layer fronting a large vessel. Penicillin was administered (40,000 U/kg) for 3 days after the surgery. In addition, all mice received a subcutaneous injection of 0.1 mg/kg body weight 17β-estradiol (Sigma) every 2 days to stimulate and maintain the growth of the endometrial implants. Twenty mice with experimentally induced EMs were randomly divided into four groups of five mice each: (1) The PBS group (as a control); (2) the FIN56 group; (3) the FIN56 plus TOFA group; and (4) the TOFA group. Two weeks after the induction of EMs, the mice in the experimental groups were intraperitoneally injected with 2.5 mg/kg FIN56 (Sigma) and/or 2.5 mg/kg selective ACC1 inhibitor TOFA (Sigma) every 3 days for 2 weeks. Mice were injected with an equal dose of PBS and/or TOFA as the control group. After 2 weeks of drug delivery, the animals were sacrificed in a CO 2 chamber, and the endometriotic lesions were removed from the peritoneum of the mice. All lesions were individually measured, and lesion volumes were calculated with the following formula: Volume = 1/2 smallest diameter 2 × largest diameter. Half of the lesions were harvested for RNA examination (immediately frozen in liquid nitrogen and stored at −80°C), and the other half were kept in 4% paraformaldehyde solution for immunohistochemistry studies. Tissues were fixed with 4% paraformaldehyde overnight, gradually dehydrated, embedded in paraffin, cut into transverse sections, and then stained with hematoxylin and eosin (HE), Ki-67, and 4-hydroxynonenal (4-HNE). After HE staining, all lesions were evaluated to confirm EMs. Cell proliferation was quantified using Ki-67 (A20018, ABclonal) as a marker of active cell division. Cell lipid peroxidation was measured using 4-HNE (ab48506. Abcam). After being stored at 4°C for 2 h, the whole blood of mice was centrifuged to separate the serum and the plasma. For in vivo serum estradiol analysis, mouse serum estradiol (E2) levels were measured by an estradiol ELISA kit (MM-0566M1, Jiangsu Meimian Industrial Co., Yancheng, China) according to the manufacturer’s instructions. All experiments were performed at least three times. Continuous variables are presented as the mean with standard deviation (SD). Categorical data are presented as numbers and percentages. Comparisons of two groups were performed using two-tailed Student’s t -tests or the nonparametric Mann–Whitney U test. Comparisons of more than two groups were performed using one-way analysis of variance (ANOVA). To compare qualitative variables, Chi-square and Fisher’s exact tests were used as indicated. Statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) computer software version 22.0 (IBM Corp, NY, USA). Figures were constructed using Prism 8 (GraphPad Software, Inc., CA, USA). Each experiment was conducted independently, with at least three replicates. P values <0.05 were considered statistically significant.

Since ferroptosis is activated by different FINs, we isolated primary stromal cells from endometriotic cysts and then treated the cells with varying concentrations of FINs for different durations. As shown in Figure 1 A, FIN56 induced cell death in ecESCs, and this effect was abolished by the ferroptosis inhibitors ferrostatin-1 (Fer-1) and liproxstatin-1 (Liprox). The CCK-8 results revealed that FIN56 reduced the viability of ecESCs and that both Fer-1 and Liprox significantly reversed the FIN56-induced decrease in ecESC viability ( P <0.001, P <0.001, P <0.001, respectively [Figure 1 B and Supplementary Figure 1A, http://links.lww.com/CM9/C574 ]). In addition, other FINs, such as erastin and RSL3, also reduced the viability of ecESCs, and these effects were reversed by Fer-1 and Liprox ( P <0.001, P <0.001, P <0.001, respectively [Figure 1 C]; P <0.001, P <0.001, P <0.001, respectively [Figure 1 D and Supplementary Figure 1B, 1C, http://links.lww.com/CM9/C574 ]). To further understand how FIN56 induces cell death in EMs, we treated ecESCs with FIN56 in the absence or presence of different programmed cell death inhibitors. We confirmed that FIN56-induced cell death in ecESCs was abolished by the ferroptosis inhibitor Fer-1 but not by the apoptosis inhibitor Z-VAD-FMK (zVAD) or the necroptosis inhibitor necrostatin-1 (Nec-1; P <0.001, P <0.001, respectively [Figure 1 E and Supplementary Figure 1D, http://links.lww.com/CM9/C574 ]). Additionally, since lipid peroxidation is a hallmark of ferroptosis, we further investigated the effects of FIN56 on lipid peroxidation changes in ecESCs. As shown in Figure 1 F, the level of malondialdehyde (MDA), which is a lipid peroxidation product, increased in the FIN56-treated group compared with the dimethyl sulfoxide (DMSO)-treated group, and these effects were reversed by Fer-1 or Liprox ( P <0.001, P = 0.003, P = 0.002, respectively). Similarly, treatment of cultured ecESCs with FIN56 for 18 h significantly increased the lipid ROS level, as assessed by C11-BODIPY fluorescent dye. However, treatment with Fer-1 or Liprox decreased the accumulation of lipid ROS compared with that in the FIN56 treatment group ( P = 0.007, P = 0.004, P = 0.008, respectively [Figure 1 G]). Collectively, these data suggest that ecESCs are sensitive to FIN56 and that FIN56 can inhibit ecESC proliferation by inducing ferroptosis in vitro . FIN56 inhibits ecESC proliferation by inducing lipid peroxidation and ferroptotic cell death. (A) Bright-field images of ecESCs cultured with 10 μmol/L FIN56 in the presence of the indicated ferroptosis inhibitors (Fer-1, 5 μmol/L; Liprox, 400 nmol/L) for 18 h (scale bar = 100 μm). (B–D) EcESCs were treated with or without 10 μmol/L FIN56 ( vs . FIN56), 20 μmol/L erastin ( vs . erastin), and 0.06 μmol/L RSL3 ( vs . RSL3) in the presence of the indicated ferroptosis inhibitors (Fer-1, 5 μmol/L; Liprox, 400 nmol/L) in 96-well plates for 18 h ( n = 3). (E) EcESCs were treated with or without 10 μmol/L FIN56 in the presence of the indicated bioactive inhibitors (Fer-1, 5 μmol/L; zVAD, 10 μmol/L; Nec-1, 5 μmol/L) for 18 h ( vs . FIN56, n = 3). Cell viability was measured with a CCK-8 assay. (F) Intracellular MDA levels in ecESCs after incubation with or without 10 μmol/L FIN56 in the presence of the indicated bioactive inhibitors (Fer-1, 5 μmol/L; Liprox, 400 nmol/L) for 18 h ( vs . FIN56, n = 3). (G) Lipid peroxidation in ecESCs was assessed by flow cytometry with C11-BODIPY after treatment with or without 10 μmol/L FIN56 in the presence of the indicated bioactive inhibitors (Fer-1, 5 μmol/L; Liprox, 400 nmol/L) for 18 h, and the relative MFI of ROS is presented ( vs . FIN56, n = 3). DMSO: Dimethyl sulfoxide; EcESCs: Ectopic endometrial stromal cells; Fer-1: Ferrostatin-1; FIN56: Ferroptosis inducer 56; FITC: Fluorescein isothiocyanate; Liprox: Liproxstatin-1; MDA: Malondialdehyde; MFI: Mean fluorescence intensity; Nec-1: Necrostatin-1; ROS: Reactive oxygen species; RSL3: RAS-selective lethal 3; zVAD: zVAD-FMK. * P <0.001; † P <0.01; ns: Non-significant. Increased fatty acid biogenesis contributes to the induction of ferroptosis. [ 30 ] However, the role of ACC1 in ferroptosis and EMs has not yet been elucidated. To explore the role of ACC1 in EMs, we performed immunohistochemical (IHC) assays on both EU and EC tissues. The expression of ACC1, encoded by ACACA , in EC tissues was significantly lower than that in EU tissues [Supplementary Figure 1E, http://links.lww.com/CM9/C574 ]. Similarly, at the mRNA level, ACACA expression in the EC group was lower than that in the EU group, while no obvious difference in ACACA expression was observed between the EU and NC groups or between the proliferative phase and the secretory phase [Supplementary Figure 1F–1H, http://links.lww.com/CM9/C574 ]). At the cellular level, both the Western blotting and RT–qPCR results indicated a significant decrease in the expression level of ACC1 in ecESCs compared with that in euESCs ( P = 0.005, P = 0.001, respectively; Supplementary Figure 1I, 1J, http://links.lww.com/CM9/C574 ). To further explore the role of ACC1 in regulating ferroptosis in EMs, we induced ferroptosis in ecESCs using FIN56 and inhibited the function of ACC1 by TOFA. The addition of TOFA abrogated the decrease in cell viability induced by FIN56, thereby preventing ecESC death from ferroptosis [Figure 2 A, 2 B; P <0.001]. Moreover, the MDA concentration and lipid ROS production in the ecESCs were reduced by TOFA [Figure 2 C and 2 D; P = 0.016, P = 0.021, respectively], indicating that inhibition of ACC1 reduced the level of lipid peroxidation. Furthermore, TEM was used to observe the morphological features of the ecESC mitochondria. Notably, incubation of ecESCs with FIN56 resulted in mitochondrial structural alterations, such as a shorter appearance and increased membrane density, whereas TOFA partly protected the mitochondrial structure from ferroptosis-induced damage [Figure 2 E]. ACC1 inhibition suppresses lipid peroxidation and ferroptosis induced by FIN56 in ecESCs. (A) Bright-field images of ecESCs cultured with 10 μmol/L FIN56 for 18 h, with or without 0.5 μmol/L ACC1 inhibitor (TOFA) pretreatment for 1 h (scale bar = 100 μm). (B) CCK-8 assay after ecESCs were treated with or without 10 μmol/L FIN56 and 0.5 μmol/L TOFA for 18 h ( vs . FIN56, n = 3). (C) Lipid peroxidation in ecESCs was assessed by flow cytometry with C11-BODIPY after pretreatment with or without 0.5 μmol/L TOFA and 10 μmol/L FIN56 for 18 h ( vs . FIN56, n = 3). (D) Intracellular MDA levels in ecESCs after pretreatment with or without 0.5 μmol/L TOFA and 10 μmol/L FIN56 for 18 h ( vs . FIN56, n = 3). (E) TEM analysis of the mitochondrial ultrastructure in ecESCs pretreated with or without 0.5 μmol/L TOFA and 10 μmol/L FIN56 for 24 h. The structures indicated by the arrows are mitochondria (upper scale bar = 2 μm, lower scale bar = 1 μm). ACC1: Acetyl-CoA carboxylase 1; CCK-8: Cell counting kit-8; DMSO: Dimethyl sulfoxide; ecESCs: Ectopic endometrial stromal cells; FIN56: Ferroptosis inducer 56; FITC: Fluorescein isothiocyanate; MDA: Malondialdehyde; MFI: Mean fluorescence intensity; ROS: Reactive oxygen species; TEM: Transmission electron microscopy; TOFA: 5-(Tetradecyloxy)-2-furoic Acid. * P <0.05; † P <0.01; ‡ P <0.001; ns: Non-significant. To explore the possible ferroptosis-related genes responsible for ACC1-mediated FIN56-induced ferroptosis in EMs, we conducted next-generation RNA-seq analysis to identify differentially expressed genes (DEGs) during ACACA knockdown in FIN56-induced ecESCs, which resulted in the identification of 752 DEGs [Figure 3 A]. Then, Venn diagram analysis was performed to identify ferroptosis-related genes (listed in the FerroDb V2 database) [ 31 ] that were essential for ACACA knockdown in FIN56-induced ecESCs compared with the negative control siRNA-transfected group (RNA-seq). From this comparison, we identified nine candidate genes (seven drivers and two suppressors), which are listed in Figure 3 B. Furthermore, we used RT–qPCR and Western blotting to validate the differential expression of these genes in ACC1-mediated FIN56-induced ferroptosis in ecESCs. The results revealed that while FIN56 increased NOX4 expression in ecESCs, ACC1 inhibition by TOFA [Figure 3 C, 3 D; P = 0.002, P = 0.017, respectively] or knockdown by ACACA siRNA [Figure 3 E, 3 F; P <0.001, P = 0.021, respectively] significantly decreased NOX4 expression at both the mRNA and protein levels. Conversely, overexpression of ACACA by plasmid transfection increased NOX4 mRNA and protein expression in FIN56-induced ecESCs [Figure 3 G, 3 H; P = 0.001, P = 0.021, respectively]. The other eight candidate genes did not show a consistent pattern. Thus, NOX4 was selected as the most promising downstream target of ACC1-mediated ferroptosis. In addition, NOX4 inhibition by GKT137831 (GKT) conferred ferroptosis resistance induced by FIN56 in ecESCs, as measured by cell viability, lipid ROS production, and MDA concentration [Figure 4 A–C; P <0.001, P <0.001, P = 0.005, respectively]. As expected, the addition of GKT to ecESCs rescued the damages in mitochondrial structure that were altered by incubation with FIN56 [Figure 4 D]. Similarly, the results of the CCK-8, flow cytometry, and MDA assays indicated that NOX4 knockdown by siRNA conferred ferroptosis resistance induced by FIN56 in ecESCs [Figure 4 E–G; P = 0.002, P = 0.046, P = 0.034, respectively]. In addition, ACACA overexpression increased ferroptosis sensitivity induced by FIN56, but this increase in sensitivity was abolished by the inhibition of NOX4 (GKT, [Figure 4 H, 4 I]; P <0.001, P <0.001, respectively). Taken together, the above results indicate that ACC1 regulates FIN56-induced ferroptosis sensitivity through NOX4 in ecESCs. ACC1 regulates NOX4 expression in FIN56-induced ecESCs. (A) DEGs identified by RNA-seq analysis of ecESCs treated with 10 μmol/L FIN56 for 18 h after treatment with either siNC or si ACACA ( n = 3). (B) Venn diagram analysis showing the intersection of the DEGs associated with ACACA knockdown (data originating from our RNA-seq analysis) and ferroptosis-related genes (data originating from the FerroDb V2 database). (C) RT–qPCR and (D) Western blotting were used to determine the relative levels of NOX4 mRNA following different treatments: DMSO, FIN56 (10 μmol/L), FIN56 (10 μmol/L) plus TOFA (0.5 μmol/L), and TOFA (0.5 μmol/L) in ecESCs ( vs. FIN56, n = 3). (E) SiRNA-mediated knockdown of the ACACA reduced the mRNA level of NOX4 in ecESCs treated with 10 μmol/L FIN56 for 24 h ( vs. siNC plus FIN56). (F) Representative Western blots showing the protein levels of NOX4 and ACC1 in ecESCs transfected with either siNC (100 nmol/L) or si ACACA (100 nmol/L) for 24 h and then treated with or without 10 μmol/L FIN56 for 24 h ( vs. siNC plus FIN56, n = 3). (G) Overexpression of ACACA by plasmid transfection increased the mRNA level of NOX4 in ecESCs treated with 10 μmol/L FIN56 for 24 h ( vs. Vector plus FIN56; n = 3). (H) Representative Western blots showing the protein levels of NOX4 and ACC1 in ecESCs transfected with either the vector or the ACACA plasmid for 24 h and then treated with or without 10 μmol/L FIN56 for 24 h ( vs. Vector plus FIN56; n = 3). ACC1: Acetyl-CoA carboxylase 1; ACACA: Acetyl-CoA carboxylase alpha; ACO1: Aconitase 1; CYGB: Cytoglobin; DEGs: Differentially expressed genes; DMSO: Dimethyl sulfoxide; DUOX1: Dual oxidase 1; EcESCs: Ectopic endometrial stromal cells; FC: Fold change; FBXW7: F-box and WD repeat domain containing 7; FIN56: Ferroptosis inducer 56; FRGs: ferroptosis-related genes; MT1G: Metallothionein 1G; NC: Negative control; NOX4: Nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 4; OE: Overexpression; RNA-seq: RNA sequencing; TFR2: Transferrin receptor 2; TOFA: 5-(Tetradecyloxy)-2-furoic Acid; TXN: Thioredoxin; USP11: Ubiquitin specific peptidase 11. * P <0.05; † P <0.01; ‡ P <0.001; ns: Non-significant. ACC1 regulates FIN56-induced lipid peroxidation and ferroptosis sensitivity through NOX4 in ecESCs. (A) CCK-8 assay after ecESCs were treated with or without FIN56 (15 μmol/L) and NOX4 inhibitor GKT (20 μmol/L) for 18 h ( vs . FIN56; n = 3). (B) Lipid peroxidation and (C) intracellular MDA levels in ecESCs after pretreatment with or without 20 μmol/L GKT and 10 μmol/L FIN56 for 18 h ( vs . FIN56; n = 3). (D) TEM analysis of the mitochondrial ultrastructure in ecESCs pretreated with or without 20 μmol/L GKT and 10 μmol/L FIN56 for 18 h. The structures indicated by the arrows are mitochondria. (upper scale bar = 1 μm, lower scale bar = 500 nm). (E) CCK-8 assay after ecESCs were treated with or without FIN56 (15 μmol/L) in the presence of either siNC or si NOX4 for 18 h ( vs . siNC plus FIN56; n = 3). (F) Lipid peroxidation and (G) intracellular MDA levels in ecESCs after transfection with either siNC or si NOX4 and 10 μmol/L FIN56 for 18 h ( vs . siNC plus FIN56; n = 3). (H) EcESCs were transfected with either the vector or the ACACA plasmid for 24 h, and then a CCK-8 assay was used to measure cell viability after the ecESCs were treated with FIN56 (10 μmol/L) in the presence of either DMSO or GKT for 18 h ( vs . ACACA OE plus FIN56; n = 3). (I) Lipid peroxidation in ecESCs was assessed by flow cytometry with C11-BODIPY after transfection with either the vector or the ACACA plasmid and incubation with 10 μmol/L FIN56 in the presence of either DMSO or GKT for 18 h ( vs . ACACA OE plus FIN56; n = 3). ACC1: Acetyl-CoA carboxylase 1; ACACA: Acetyl-CoA carboxylase alpha; CCK-8: Cell Counting Kit-8; DMSO: Dimethyl sulfoxide; EcESCs: Ectopic endometrial stromal cells; FITC: Fluorescein isothiocyanate; FIN56: Ferroptosis inducer 56; GKT: GKT137831; MDA: Malondialdehyde; MFI: Mean fluorescence intensity; NOX4: Nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 4; OE: Overexpression; ROS: Reactive oxygen species; TEM: Transmission electron microscopy. * P <0.05; † P <0.01; ‡ P <0.001; ns: Non-significant. To further investigate the transcriptional mechanism responsible for the FIN56/ ACACA -mediated activation of NOX4 in EMs, we conducted Venn diagram analysis to determine the transcription factors (TFs) that were differentially expressed after ACACA knockdown and repressed or increased after FIN56 treatment of ecESCs [Supplementary Figure 1K, http://links.lww.com/CM9/C574 ]. According to our RNA-seq results, the expression of ARID5A was increased by the addition of FIN56 compared with the siNC+DMSO group, and it was downregulated after the transfection of si ACACA compared with that in the siNC group in FIN56-induced ecESCs. Therefore, we investigated whether ARID5A participates in FIN56/ ACACA -mediated transcriptional activation of NOX4 in EMs. After the addition of FIN56 to ecESCs, the mRNA levels of ARID5A increased; however, the upregulation of ARID5A expression was reversed by ACACA knockdown [Figure 5 A; P = 0.003]. In contrast, the overexpression of ACACA further promoted ARID5A expression in FIN56-induced ecESCs [Figure 5 B; P = 0.031]. This finding indicates that ACACA -mediated FIN56 induces ARID5A mRNA expression. To test whether ARID5A, a transcription factor, regulates NOX4 expression in FIN56-induced ecESCs, we conducted immunofluorescence analysis of FIN56-induced ecESCs. The results revealed that ARID5A downregulation reduced the expression of NOX4 [Supplementary Figure 1L, http://links.lww.com/CM9/C574 ]. Furthermore, the Western blotting results revealed that the expression of NOX4 decreased in response to ARID5A downregulation [Figure 5 C; P = 0.002]. To further test whether ARID5A participates in ACC1-mediated NOX4 expression, we conducted a reverse experiment in which ACACA was overexpressed and ARID5A was knocked down in ecESCs. The RT–qPCR and Western blotting results demonstrated that ARID5A knockdown effectively reversed the increase in NOX4 expression induced by ACACA overexpression [Figure 5 D, 5 E; P <0.001, P = 0.012, respectively]. In addition, the immunofluorescence results revealed that the inhibition of ACC1 by TOFA [Figure 5 F] or the knockdown of ACACA by siRNA transfection [Figure 5 G] downregulated the expression of ARID5A and NOX4 in FIN56-induced ecESCs compared with the control. Next, the JASPAR database was used to predict possible ARID5A-binding sites in the promoter region of NOX4 . We then performed ChIP analysis using antibodies against ARID5A with primers amplifying three regions of the NOX4 promoter between −931 and −138 bp upstream from the transcriptional start site. ChIP analysis revealed that FIN56 increased the binding of ARID5A to the NOX4 promoter, and this increase was reversed by ACACA knockdown [Figure 5 H; P = 0.015, P = 0.018, P = 0.020, respectively]. Overall, these results show that ARID5A serves as a positive transcription factor that promotes FIN56/ACC1-mediated NOX4 upregulation. ACC1 regulates NOX4 through ARID5A-mediated transcription. (A) SiRNA-mediated knockdown of the ACACA gene reduced the mRNA level of ARID5A in ecESCs treated with 10 μmol/L FIN56 ( vs. siNC plus FIN56, n = 3). (B) The overexpression of ACACA by plasmid transfection increased the level of ARID5A mRNA in ecESCs treated with 10 μmol/L FIN56 ( vs. Vector plus FIN56; n = 3). (C) Representative immunoblot analysis of the protein levels of NOX4 in control (siNC) and ARID5A-downregulated (si ARID5A ) human ecESCs ( vs. siNC plus FIN56; n = 3). (D) RT–qPCR ( n = 3) and (E) Western blotting ( n = 3) were used to determine the relative levels of NOX4 mRNA and protein after different treatments in 10 μmol/L FIN56-treated ecESCs: Vector plus siNC, ACACA OE plus siNC, Vector plus si ARID5A and ACACA OE plus si ARID5A ( vs. ACACA OE plus FIN56). (F) Representative fluorescence images of ARID5A and NOX4 in ecESCs cultured with 10 μmol/L FIN56 for 18 h with or without 0.5 μmol/L TOFA pretreatment for 1 h. DAPI was used as a nuclear marker (blue). The ARID5A protein was labeled with red fluorescence, and the NOX4 protein was labeled with green fluorescence. Scale bar, 25 μm. (G) Representative fluorescence images of ARID5A and NOX4 in ecESCs transfected with either siNC (100 nmol/L) or si ACACA (100 nmol/L) for 24 h, followed by treatment with or without 10 μmol/L FIN56 for 18 h. DAPI was used as a nuclear marker (blue). The ARID5A protein was labeled with red fluorescence, and the NOX4 protein was labeled with green fluorescence. Scale bar, 25 μm ( vs . siNC plus FIN56; n = 3). (H) NOX4 is a direct target of ARID5A. Using ecESC lysates subjected to different treatments (siNC plus DMSO, si ACACA plus DMSO, siNC plus FIN56, and si ACACA plus FIN56), ChIP–qPCR with an ARID5A antibody confirmed that ARID5A binds to the NOX4 promoter at three different binding sites ( vs . siNC plus FIN56, n = 3). ACC1: Acetyl-CoA carboxylase 1; ACC1: Acetyl-CoA carboxylase 1; ACACA: Acetyl-CoA carboxylase alpha; ARID5A: AT-rich interaction domain 5A; ChIP: Chromatin immunoprecipitation; DAPI: 4′,6-diamidino-2-phenylindole; DMSO: Dimethyl sulfoxide; EcESCs: Ectopic endometrial stromal cells; FIN56: Ferroptosis inducer 56; IgG: Immunoglobulin G; NC: Negative control; NOX4: Nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 4; OE: Overexpression; RT–qPCR: Reverse transcription quantitative polymerase chain reaction; TOFA: 5-(Tetradecyloxy)-2-furoic Acid. * P <0.05; † P <0.01; ‡ P <0.001; ns: Non-significant. The above results indicate that FIN56/ACC1 can stimulate NOX4 expression via the activation of ARID5A. Thus, we next investigated how FIN56/ACC1 modulates the expression of ARID5A and NOX4, thereby affecting ferroptosis sensitivity. P38 mitogen-activated protein kinase (p38 MAPK) has been reported to regulate ARID5A phosphorylation and NOX4-mediated ROS production, [ 32 , 33 ] but whether p38 MAPK participates in the regulation of FIN56/ACC1-induced ferroptosis has not yet been clarified. First, we investigated the effects of FIN56 on the p38 MAPK signaling pathway. Intriguingly, our findings revealed that FIN56 treatment significantly increased p38 MAPK phosphorylation (Figure 6 A; P = 0.007), whereas inhibition of ACC1 by TOFA suppressed p38 MAPK phosphorylation (Figure 6 A; P = 0.010). Similar results were obtained following ACACA siRNA‑mediated knockdown (Figure 6 B; P = 0.013 and P = 0.013, respectively). Furthermore, FIN56 treatment markedly elevated p38 MAPK phosphorylation (Figure 6 C; P = 0.015), and ACACA overexpression further enhanced this effect (Figure 6 C; P = 0.017). In addition, a p38-specific inhibitor, SB202190 (SB), abrogated the increase in ARID5A and NOX4 expression associated with ACACA overexpression at both the mRNA and protein levels [Figure 6 D, P <0.001, P <0.001, respectively; Figure 6 E, P = 0.004, P = 0.002, respectively]. Furthermore, the results of experiments measuring cell viability and lipid peroxidation consistently demonstrated that p38 MAPK inhibition can reverse the ACACA overexpression-induced increase in ferroptosis sensitivity [Figure 6 F, 6 G; P <0.001, P = 0.002, respectively]. In summary, our findings suggest that FIN56/ACC1 promotes ARID5A expression by modulating the p38 MAPK signaling pathway, thereby upregulating NOX4 expression. FIN56/ACC1 promotes the generation of lipid ROS and modulates the expression of ARID5A and NOX4 through the p38 MAPK signaling pathway. (A) Western blotting was used to detect the phosphorylated and total protein levels of p38 MAPK in ecESCs after pretreatment with or without 0.5 μmol/L TOFA and incubation with 10 μmol/L FIN56 for 24 h ( vs. FIN56, n = 3). (B) The phosphorylated and total protein levels of p38 MAPK in ecESCs after transfection with either siNC (100 nmol/L) or si ACACA (100 nmol/L) for 24 h, followed by treatment with or without 10 μmol/L FIN56 for 24 h ( vs. siNC plus FIN56, n = 3). (C) The phosphorylated and total protein levels of p38 MAPK in ecESCs after transfection with either the Vector or the ACACA plasmid for 24 h, followed by treatment with or without 10 μmol/LFIN56 for 24 h ( vs. Vector plus FIN56; n = 3). (D) RT–qPCR ( n = 3) and (E) Western blotting ( n = 3) were used to determine the relative levels of ARID5A and NOX4 mRNA and protein in ecESCs after transfection with either the Vector or the ACACA plasmid for 24 h, followed by treatment with or without the p38 MAPK inhibitor SB202190 and 10 μmol/L FIN56 ( vs. ACACA OE plus FIN56). (F) Cell viability ( n = 3) and (G) lipid peroxidation ( n = 3) in ecESCs were assessed by CCK-8 assay and flow cytometry with C11-BODIPY after treatment with FIN56 (10 μmol/L): Vector plus DMSO, ACACA OE plus DMSO, Vector plus SB202190, and ACACA OE plus SB202190 ( vs. ACACA OE plus FIN56). ACC1: Acetyl-CoA carboxylase 1; ACACA: Acetyl-CoA carboxylase alpha; ARID5A: AT-rich interaction domain 5A; CCK-8: Cell Counting Kit-8; DMSO: Dimethyl sulfoxide; EcESCs: Ectopic endometrial stromal cells; FIN56: Ferroptosis inducer 56; FITC: Fluorescein isothiocyanate; MFI: Mean fluorescence intensity; NC: Negative control; NOX4: Nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 4; OE: Overexpression; P38 MAPK: P38 mitogen-activated protein kinase; P-P38 MAPK: Phosphorylated P38 MAPK; RT–qPCR: Reverse transcription quantitative polymerase chain reaction; ROS: Reactive oxygen species; SB: SB202190; TOFA: 5-(Tetradecyloxy)-2-furoic Acid. * P <0.05; † P <0.01; ‡ P <0.001; ns : Non-significant. NOX4 has been proven to be a key regulator of lipid peroxidation during ferroptosis. [ 34 , 35 ] To examine whether NOX4 promotes lipid peroxidation via mitochondrial dysfunction in FIN56-induced ferroptosis, we analyzed the effects of NOX4 on mitochondrial functions, such as mtROS production, the MMP, and ATP production. In ecESCs, FIN56 treatment reduced ATP production, as assessed by a bioluminescence assay; decreased the MMP, as estimated by JC-1 fluorescence; and increased mtROS production, as estimated by mitoSOX fluorescence. Furthermore, these effects were abolished by the inhibition of ACC1, the p38 MAPK inhibitor SB, or the NOX4 inhibitor GKT [Figure 7 A–C]. Similarly, ACACA and NOX4 downregulation by siRNA transfection significantly increased the MMP [Figure 7 D; P = 0.003, P = 0.002, respectively], decreased mitoROS production [Figure 7 E; P = 0.002, P = 0.001, respectively] and increased ATP production [Figure 7 F; P = 0.042, P <0.001, respectively] in FIN56-treated ecESCs compared with those in control cells. Moreover, inhibition of NOX4 abolished the effects of ACACA overexpression on the MMP [Figure 7 G; P = 0.044], mtROS production [Figure 7 H; P = 0.006], and ATP levels [Figure 7 I; P = 0.008] in FIN56-induced ecESCs. These results suggest that the activation of ACC1 and NOX4 promotes ferroptosis by impairing mitochondrial function in FIN56-treated human ecESCs. Finally, we measured the intracellular iron level in ecESCs, which is crucial for ferroptosis. Our results demonstrated that the FIN56-induced accumulation of intracellular iron was abolished by ACC1 and NOX4 inhibition [Figure 7 J; P = 0.003, P = 0.015, respectively], accompanied by reduced lipid peroxidation in ecESCs [Figure 7 K; P = 0.001, P <0.001, respectively]. Taken together, our findings suggest that ACC1 and NOX4 increase ferroptosis-dependent cytotoxicity by impairing mitochondrial function and inducing mtROS in ecESCs. ACC1 and NOX4 increase ferroptosis-dependent cytotoxicity by impairing mitochondrial function and inducing mtROS in ecESCs. (A) The MMP levels of ecESCs treated with 10 μmol/L FIN56 in the presence of the indicated bioactive inhibitors (TOFA, 0.5 μmol/L; SB, 5 μmol/L; GKT, 20 μmol/L) for 18 h were measured via JC-1 staining ( vs . FIN56; scale bar, 200 μm). (B) MtROS levels in ecESCs treated with 10 μmol/L FIN56 in the presence of the indicated bioactive inhibitors (TOFA, 0.5 μmol/L; SB, 5 μmol/L; GKT, 20 μmol/L) for 18 h determined by mitoSOX staining ( vs . FIN56; scale bar, 200 μm, n = 3). (C) The levels of ATP production after ecESCs were treated with 10 μmol/L FIN56 in the presence of the indicated bioactive inhibitors (TOFA, 0.5 μmol/L; SB, 5 μmol/L; GKT, 20 μmol/L) for 18 h ( vs . FIN56, n = 3). (D) The MMP, (E) mtROS levels and (F) ATP production in ecESCs treated with 10 μmol/L FIN56 for 18 h after transfection with siNC, si ACACA or si NOX4 for 24 h ( vs . siNC plus FIN56; scale bar, 200 μm, n = 3). (G) The MMP, (H) mtROS levels, and (I) ATP production in ecESCs treated with or without NOX4 inhibitor GKT and 10 μmol/L FIN56 for 18 h after transfection of either the Vector or the ACACA plasmid for 24 h ( vs . ACACA OE plus FIN56; scale bar, 200 μm, n = 3). (J) The intracellular iron (cyto-Fe 2+ ) level in ecESCs treated with 10 μmol/L FIN56 in the presence of the indicated bioactive inhibitors (TOFA, 0.5 μmol/L; GKT, 20 μmol/L) for 18 h was determined by FerroOrange staining ( vs . FIN56; scale bar, 200 μm, n = 3). (K) Lipid peroxidation in ecESCs was assessed by flow cytometry with C11-BODIPY after treatment with FIN56 (10 μmol/L) and 0.5 μmol/L TOFA, or 20 μmol/L GKT ( vs. FIN56, n = 3). ACC1: Acetyl-CoA carboxylase 1; ACACA: Acetyl-CoA carboxylase alpha; ATP: Adenosine triphosphate; DMSO: Dimethyl sulfoxide; EcESCs: Ectopic endometrial stromal cells; FIN56: Ferroptosis inducer 56; FITC: Fluorescein isothiocyanate; GKT: GKT137831; MFI: Mean fluorescence intensity; MMP: Mitochondrial membrane potential; MtROS: Mitochondrial reactive oxygen species; NC: Negative control; NOX4: Nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 4; OE: Overexpression; SB: SB202190; TOFA: 5-(Tetradecyloxy)-2-furoic Acid. * P <0.05; † P <0.01; ‡ P <0.001; ns: Non-significant. This study has shown that FIN56 induces ferroptosis and that ACC1 inhibition in ecESCs confers resistance to ferroptosis in vitro ; thus, we hypothesized that FIN56 can promote ferroptosis and reduce the volume of ectopic lesions in vivo . To test the therapeutic efficacy of FIN56 in ectopic lesions and investigate the role of ACC1 in ferroptosis in vivo , we established a mouse model of EMs as previously described [Figure 8 A]. We subsequently divided the mice into four groups: PBS (control), FIN56, TOFA, and FIN56 plus TOFA, as shown in Figure 8 B. The results of HE staining revealed the formation of cystic endometriotic lesions by epithelial and stromal cells [Figure 8 C]. Notably, the application of FIN56 slowed the development of EMs, whereas FIN56 plus TOFA increased the ectopic lesion volume compared with that in the FIN56 treatment group [Figure 8 D; P = 0.002, P = 0.015, respectively]. In addition, the ectopic lesion weights of the FIN56-treated group were lower than those of the control group, whereas FIN56 plus TOFA reversed the reduction in lesion weight caused by FIN56 treatment [Figure 8 E; P = 0.015, P = 0.009, respectively]. None of these treatments affected the body weights of the mice [Figure 8 F]. The results of the IHC assays revealed that 4-hydroxynonenal (4-HNE, a marker of lipid peroxidation) levels were greater in the FIN56 treatment group than in the control group [Figure 8 G]. IHC assays also revealed a decrease in the expression of Ki-67 in the FIN56 treatment group compared with that in the control group, while the combination of FIN56 plus TOFA had the opposite effect. Similar trends were observed via RT–qPCR analysis, as the expression of cyclin D1 Ccnd1 decreased in the FIN56 treatment group compared with the control group [Figure 8 H, I; P = 0.004]. Previous studies have demonstrated that the growth of endometriotic tissue is estrogen dependent, [ 36 ] and the ELISA results revealed that the serum estrogen levels were lower in the FIN56 treatment group than in the control group [Figure 8 J; P = 0.002]. In summary, we propose a molecular model [Figure 8 K] in which FIN56 induces ferroptosis by affecting the function of ACC1, which in turn promotes activation of the p38 MAPK-ARID5A-NOX4 axis and oxidative stress-induced lipid peroxidation via impairment of mitochondrial function in ecESCs. This study provides evidence for the use of the novel ferroptosis inducer FIN56 in the treatment of EMs. FIN56 inhibits the growth of ectopic lesions in vivo . (A) Schematic diagram showing the establishment of the mouse EMs model. (B) Schematic diagram showing the therapeutic process. (C) Representative HE staining of glandular and stromal structures of EMs ectopic lesions in the four groups (scale bar = 20 μm). (D) A representative image and comparison of the volume of ectopic EMs lesions in the four groups ( n = 10 from 5 mice each group) on day 29 are shown. (E) Comparison of the weights of the EMs ectopic lesions in the four groups ( n = 10 from 5 mice each group) on day 29. (F) Line chart of the weights of the mice in the four groups ( n = 5 each group) at different time points, which showed no significant changes. (G) Representative IHC images of the four groups of EMs ectopic lesions stained with 4-HNE (a marker of lipid peroxidation) (upper scale bar = 50 μm, lower scale bar = 20 μm; n = 5). (H) RT–qPCR was used to determine the relative levels of Ccnd1 mRNA in the four groups of EMs ectopic lesions ( n = 5; vs . FIN56). (I) Representative IHC images showing Ki-67 (a marker of proliferation) staining in ectopic lesions from the four groups of EMs patients (upper scale bar = 50 μm, lower scale bar = 20 μm; n = 5). (J) ELISA was used to determine the serum estradiol levels in the four groups of EMs ectopic lesions ( n = 5; vs . FIN56). (K) A schematic diagram summarizing the findings: FIN56 induces ferroptosis by affecting the function of ACC1, which, in turn, promotes the activation of the p38 MAPK-ARID5A-NOX4 axis and oxidative stress-induced lipid peroxidation via impairment of mitochondrial function in ecESCs. ACC1: Acetyl-CoA carboxylase 1; ARID5A: AT-rich interaction domain 5A; Ccnd1: Cyclin D1; 4-HNE: 4-Hydroxynonenal; ELISA: Enzyme-linked immunosorbent assay; EMs: Endometriosis; FIN56: Ferroptosis inducer 56; HE staining: Hematoxylin-eosin staining; IHC: Immunohistochemical; NOX4: Nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 4; P38 MAPK: P38 mitogen-activated protein kinase; PBS: phosphate-buffered saline; RT–qPCR: Reverse transcription quantitative polymerase chain reaction; ROS: Reactive oxygen species; TOFA: 5-(Tetradecyloxy)-2-furoic Acid. * P <0.05; † P <0.01; ns : Non-significant.

Although EMs have been demonstrated to be associated with the dysregulation of iron homeostasis, [ 26 ] the role of ferroptosis in the ectopic endometrium has not yet been clearly examined. Accumulating evidence indicates that ferroptosis is a critical tumor suppression mechanism. Several tumor-suppressor and oncogenic signaling pathways have been shown to promote or suppress ferroptosis, respectively. [ 37 , 38 ] The central part of the ferroptosis process is the intracellular iron-dependent accumulation of lipid peroxides, which exceeds the buffering capability of ferroptosis defense systems, subsequently leading to membrane rupture and ultimately resulting in ferroptotic cell death. [ 5 , 38 ] Thus, the induction of ferroptosis is a new approach for cancer therapy. [ 39 ] However, the roles of ferroptosis in EMs have not been fully clarified. In patients with EMs, periodic bleeding and the accumulation of old blood establish an iron-overloaded environment, which increases the tolerance of ecESCs to high iron concentrations. [ 40 , 41 ] Ferroptosis is not conducive to cell survival; thus, modulating the sensitivity of eESCs to ferroptosis through downstream pathways may be a therapeutic direction in the future. [ 42 ] It was recently reported that macrophage-derived interleukin-33 upregulated SLC7A11 through the MAPK14/ATF3 pathway, resulting in protection against ferroptosis in ecESCs and ultimately contributing to ecESC survival and migration. [ 24 ] Thus, ferroptosis is correlated with ecESC growth and proliferation, and the induction of ferroptosis by different types of FINs may be an effective therapeutic strategy for the noninvasive treatment of EMs. [ 22 ] Here, we provide evidence that triggering ferroptosis in ectopic lesions can slow the progression of EMs, which is consistent with the findings of previous studies. [ 22 , 23 , 43 , 44 ] Despite extensive investigations of the role of class 1 and class 2 FINs in tumorigenesis and progression, class 3 FIN (FIN56), a novel FIN, has rarely been studied. Through pharmacologically accessible cell death profiling analysis, Shimada et al [ 14 ] first discovered that FIN56 promoted the degradation of glutathione peroxidase 4 (GPX4) and activated squalene synthase, an enzyme involved in isoprenoid biosynthesis, which further induced ferroptosis. FIN56 has been shown to trigger ferroptosis via a pathway that is not fully understood. Recent studies have shown that FIN56-triggered ferroptosis mechanistically depends on the autophagic machinery in bladder cancer cells. [ 35 ] Moreover, FIN56 triggers ferroptosis and inflammation in hippocampal neurons after subarachnoid hemorrhage by activating the NFE2L2 / TXNRD1 axis. [ 45 ] In the present study, we demonstrate that FIN56 induces ferroptosis in ecESCs by activating the ACC1/p38 MAPK/ARID5A/NOX4 axis, which further promotes mtROS accumulation and lipid peroxidation. ACC1 has emerged as an attractive therapeutic target for cancer therapy. [ 46 ] Inhibiting ACC1 has anti-cancer effects in several malignancies, including hepatocellular carcinoma. [ 47 ] However, other preclinical models have demonstrated tumor-promoting effects of ACC inhibition or impede response to other therapeutics through acetyl-CoA rerouting. For example, ACC1 inhibition increases acetylation of SMAD family member 2 (SMAD2) to promote epithelial-mesenchymal transition and metastasis in breast cancer. [ 11 ] Interestingly, inhibition of ACC1 has been reported to be involved in ferroptosis resistance. [ 14 , 15 , 18 ] AMPK-dependent phosphorylation of ACC promotes cell death resistance by increasing NADPH and reducing lipid peroxidation. Treatment with TOFA, an allosteric inhibitor of ACC1, potently inhibits lipid peroxidation and ferroptosis induced by erastin and RSL3 in mouse embryonic fibroblasts (MEFs). [ 16 ] In addition, compared with wild-type MEFs, ACC1 knockout MEFs are remarkably resistant to ferroptosis and accumulate fewer lipid hydroperoxides and less NADPH oxidation. Knockout of ACACA protects cells from erastin-induced mitochondrial damage and GPX inhibition. [ 17 ] However, the detailed mechanism linking FIN56 to ACC1 is still unclear. To the best of our knowledge, the roles of ACC1 and ferroptosis regulation in EMs have not yet been studied. Although EMs is considered as a benign lesion, it shares similar characteristics, including proliferation and metastasis, with cancers. [ 48 ] On the basis of these previous findings, we investigated the biological role of ACC1 in the pathogenesis of EMs and its possible mechanisms in inducing ferroptosis resistance. The present study revealed that the expression of ACACA was decreased in the ectopic lesions and the ecESCs of patients with EMs. Although ACC1 alone did not affect the viability of ecESCs, ACC1 inhibition or ACACA deficiency significantly reduced the cellular iron concentration, inhibited mitochondrial oxidative stress and lipid peroxidation, and rescued ecESCs from iron-induced ferroptosis, whereas the overexpression of ACACA facilitated FIN56-induced ecESC ferroptosis. Therefore, we hypothesized that low expression of ACACA in the ectopic lesions protected ecESCs from ferroptosis-dependent cytotoxicity, thereby enhancing the tolerance of ecESCs to ferroptosis. Lipid ROS can accumulate through the generation of superoxide and hydrogen peroxide through the upregulation of nicotinamide adenine dinucleotide phosphate oxidase (NOX). [ 49 ] NOX are a family of enzymes that produce ROS. [ 50 ] These enzymes are multisubunit membrane-bound enzymes that catalyze the reduction of oxygen to superoxide by using NADPH as an electron donor and oxygen as an electron acceptor. [ 50 ] Although NOX-derived ROS are essential for normal cell survival and angiogenesis, excessive ROS levels lead to cell death by damaging cellular components, including proteins, lipid bilayers, and chromosomes. [ 51 ] A growing body of evidence supports that the overproduction of ROS by abnormal NOX activation in the brain and cardiovascular system contributes to neurodegeneration and thrombosis, thus NOXs inhibitors show therapeutic effects in cardiovascular diseases and brain disorders. [ 52 , 53 ] Using RNA-seq analysis, we identified NOX4 as a potential downstream target of FIN56-induced ferroptosis mediated by ACACA . To date, the essential role of NOX4 as a ferroptosis promoter has been proven in many diseases. [ 19 , 20 , 27 ] In renal cell carcinoma, inhibition of NOX4 reduces erastin-induced cell death and lipid ROS through the WWTR1 - EMP1 axis. [ 27 ] In addition, an increase in NOX4 promotes excessive mtROS production in human astrocytes, accelerating ferroptosis through the accumulation of 4-HNE and morphological cytotoxicity. [ 6 ] For EMs, there is increasing evidence of a role for ROS and impaired mitochondrial function in terms of EMs development and embryo development impairment. [ 54 ] Mitochondrial superoxide dismutase (SOD2), an antioxidant enzyme, is highly expressed in the ectopic endometrium and promotes cell proliferation and migration in ovarian EMs by maintaining functional mitochondria. [ 21 ] However, much remains unknown about NOX4 and mtROS in EMs. Our study revealed that ecESCs could not eliminate lipid ROS produced by FIN56 induction. Additionally, NOX4 increased FIN56-induced ferroptosis-dependent cytotoxicity by impairing mitochondrial function and oxidative stress-induced lipid peroxidation in ecESCs, which was reversed by ACC1 inhibition. This may be partly attributed to the exhausted ROS-buffering capacity after heavily utilizing adaptation mechanisms in ecESCs; thus, ecESCs may be more sensitive to oxidative damage. [ 21 ] The ARID family of DNA-binding proteins is a group of TFs that have been implicated in various human diseases, including cancers and intellectual disability. The ARID family has been shown to be important in cell development, differentiation, proliferation, and tissue-specific gene expression. The associations and functions of some ARID family members, such as ARID1A and ARID1B, with human cancers have already been established and are well understood. [ 55 , 56 ] Recent advances in EMs research have revealed the important function of somatic ARID1A mutations in the upregulation of proangiogenic and prolymphangiogenic factors, reinforcing the epidemiological associations between EMs and EMs-associated ovarian cancers. [ 57 , 58 ] However, the functions of the ARID family in EMs or ferroptosis have not been fully characterized. As a transcription factor, ARID5A mediates both transcriptional and posttranscriptional regulatory effects that are implicated in immune regulation and cellular homeostasis. [ 59 ] Several publications have demonstrated that inhibiting Arid5a enhances several processes, such as invasion and metastasis, immune evasion, and the M1-like tumor-associated macrophage (TAM)-to-M2-like TAM transition. [ 60 , 61 ] Low expression of ARID5A is correlated with poor prognosis in lung and breast cancer patients. [ 62 , 63 ] In addition, Arid5a controls adipogenesis and obesity in mice to maintain metabolic homeostasis. [ 64 ] To date, the role of ARID5A in EMs has not been addressed. As predicted through RNA-seq analysis and the use of the bioinformatic website JASPAR, there may be correlations between ACACA, ARID5A and NOX4 . Knockdown of ACACA in ecESCs resulted in decreased ARID5A and NOX4 expression. Furthermore, this binding relationship between ARID5A and NOX4 was confirmed via a ChIP assay, which indicated that NOX4 was a direct target of ARID5A. Thus, the present study has revealed a crucial role of ARID5A in ferroptosis through its transcriptional regulation of the ferroptosis-related gene NOX4 . Another intriguing finding of our study is the restrained activation of the p38 MAPK pathway in ecESCs treated with FIN56. Previous research has highlighted the role of p38 MAPK in activating NOX4 in erastin-induced ferroptosis. [ 27 ] We also found that FIN56 treatment had a similar promoting effect on ARID5A and NOX4 expression via p38 MAPK activation. Treatment with the p38 MAPK inhibitor SB restrained the functions of ARID5A and NOX4 in ACACA -overexpressing ecESCs. By measuring the levels of Fe 2+ , lipid peroxidation and MDA, we confirmed the protective effects of p38 MAPK inhibitor against ferroptosis triggered by FIN56. Notably, there are several limitations to our study. Most of the samples were collected from EMs patients in the middle and advanced stages (stages III and IV) according to the rAFS. Further studies including patients in early stages (stages I and II) are necessary to acquire a more comprehensive understanding of disease maintenance and development. In addition, although the levels of iron and lipid peroxide within low-density lipoprotein (LDL) particles are greater in the peritoneal fluid of women with EMs than in those without, [ 65 , 66 ] ferroptosis has been demonstrated to play a dual role in both the eutopic and ectopic endometria of patients with EMs. [ 40 ] On the one hand, FINs can promote ferroptosis to inhibit the proliferation of ecESCs, which may lead to potential drugs for the treatment of EMs. [ 23 , 30 , 67 ] On the other hand, ferroptotic ESCs can release inflammatory cytokines and activate downstream regulatory pathways to promote angiogenesis in surrounding tissues. [ 68 ] Thus, further investigation is still warranted to fully elucidate the role of ferroptosis in EMs development. Overall, we propose that FIN56 can increase lipid peroxidation, which further triggers ferroptosis in ecESCs in vitro and in ectopic lesions in vivo . Dysregulated ACC1 protects ecESCs from FIN56-induced ferroptosis by downregulating NOX4 expression through the p38 MAPK/ ARID5A axis. The synergistic effect of FIN56 induction and ACACA overexpression on ferroptosis may be a new therapeutic strategy for EMs. Our study provides new evidence for understanding the pathogenesis of EMs and developing novel treatment strategies.

None.

The authors thank Professor Dingfang Bu for providing generous advice regarding this work.

The RNA sequencing datasets presented in this study can be found in the NCBI. The accession number is PRJNA1105642. Other data used and/or analyzed during the current study are available from the corresponding author upon reasonable request.