Stromal cell-derived itaconate promotes endometriosis via macrophage NRF2 and lysosomal pH modulation

Abstract

Endometriosis (EM) is driven by immune dysregulation and macrophage dysfunction, yet the underlying mechanisms remain unclear. Here, metabolomic profiling revealed excessive itaconate accumulation in EM lesions, primarily due to elevated cis-aconitate decarboxylase 1 (ACOD1) expression in ectopic stromal cells (ESCs). ESC-derived itaconate was internalized by peritoneal macrophages, where it suppressed pro-inflammatory activity and phagocytosis, thereby facilitating ESC survival and dissemination. Mechanistically, itaconate exerted dual regulatory effects on macrophages: it activated NRF2 signaling to repress the transcription of pro-inflammatory genes, and it enhanced lysosomal acidification, thereby reducing lysosomal calcium release, which in turn inhibited p38–MAPK activation and further attenuated pro-inflammatory gene expression. In vivo, ACOD1 inhibition restored macrophage function and reduced lesion burden, while exogenous 4-octyl itaconate aggravated disease progression. These findings define a novel “ESC–ACOD1–itaconate–macrophage” axis that mediates immunosuppression in EM and identify ACOD1 as potential therapeutic targets.

Graphical abstract

ESC-derived itaconate remodels macrophage function and drives immune microenvironment dysregulation in endometriosisEctopic endometrial stromal cells (ESCs) in endometriosis upregulate ACOD1, leading to increased itaconate synthesis and secretion. This ESC-derived itaconate is taken up by peritoneal macrophages, where it elevates lysosomal acidity, inhibits NOX2 and MCOLN2-mediated calcium signaling, and activates NRF2-driven antioxidant responses. These coordinated changes suppress macrophage pro-inflammatory and clearance functions, resulting in immune microenvironment dysregulation and facilitating lesion persistence. Our findings uncover a novel ESC–ACOD1–itaconate–macrophage metabolic–immune axis, providing new insight into endometriosis pathogenesis and suggesting promising targets for precision therapy.

1. Introduction

Endometriosis (EM) is a prevalent chronic inflammatory disorder affecting approximately 10% of women of reproductive age worldwide. It is defined by the ectopic growth of functional endometrial tissue beyond the uterine cavity, resulting in significant clinical manifestations such as chronic pelvic pain and infertility [1,2]. endometrial stromal cells (ESCs) constitute the predominant cellular component within EM lesions and serve as a critical driver of disease progression due to their enhanced survival capacity and invasive potential [3]. Although hormone therapy and surgical excision remain the primary treatment modalities, drug-related adverse effects and high recurrence rates underscore the urgent need for novel and precision-based therapeutic strategies.

Dysregulation of the immune microenvironment is an important feature of the pathogenesis of EM, especially the dysfunction of macrophages in eliminating ectopic ESCs [[4], [5], [6], [7]]. As a core component of the innate immune system, peritoneal macrophages play a key role in recognizing and eliminating ectopic endometrial tissues [8,9]. However, in EM patients, their pro-inflammatory function is markedly suppressed, with significantly reduced phagocytic and cytotoxic capacities [10], leading to immune escape of ectopic ESCs and disease progression [11,12]. Recent studies have shown that metabolic reprogramming plays an important role in regulating the immune function of macrophages [[13], [14], [15]], suggesting that metabolic products may play a key role in the immune suppression mechanism of EM.

Itaconate is a critical immunomodulatory metabolite synthesized by cis-aconitate decarboxylase 1 (ACOD1), which is encoded by immune response gene 1 (IRG1) [16,17], and is predominantly found in activated macrophages [18]. Accumulating evidence demonstrates that itaconate exerts potent anti-inflammatory effects through modulation of key signaling pathways such as NRF2, SDH, and TFEB [[18], [19], [20]], thereby suppressing inflammatory responses in various pathological conditions including osteoarthritis and atherosclerosis [[21], [22], [23], [24], [25]]. Furthermore, itaconate derived from tumor-associated macrophages has been implicated in promoting tumor immune evasion via its immunosuppressive properties [[26], [27], [28], [29]]. Despite these advances, the functional role of itaconate in EM and its underlying regulatory mechanisms remain unclear. Notably, optimal lysosomal acidity is essential for macrophage pro-inflammatory function, and alterations in lysosomal pH can profoundly affect macrophage activity [[30], [31], [32]], suggesting that itaconate may modulate macrophage-mediated immune responses by regulating lysosomal pH.

This study demonstrates for the first time that ectopic ESCs in EM abnormally highly express ACOD1 and secrete itaconate, significantly inhibiting the pro-inflammatory and phagocytic functions of peritoneal macrophages. Our mechanistic study demonstrates that itaconic acid not only activates the NRF2 signaling pathway but also inhibits NOX2 to enhance lysosomal acidification, thereby synergistically suppressing macrophage-mediated immune clearance of ectopic ESCs. These findings define the "ESC–ACOD1–itaconate–macrophage" metabolic–immune regulatory axis, highlighting the pivotal role of ESC–macrophage crosstalk in remodeling the endometriotic immune microenvironment. Importantly, this work provides novel targets and a theoretical framework for EM immunotherapy, suggesting that modulation of the ACOD1–itaconate pathway may offer safer and more effective treatment strategies for EM patients.

2. Materials and methods

2.1. Sex as a biological variable

As endometriosis is relevant only to females, only female mice and humans were involved in the present studies.

2.2. Ethics statement

All animal experiments were conducted in accordance with the Guide for the Care and Use of Laboratory Animals and were approved by the Ethics Committee of the Medical Integration and Practice Center of Shandong University (SDULCLL2022-2-17; approval date: November 4, 2022). Human sample collection and use were approved by the same committee (SDULCLL2022-1-21; approval date: November 4, 2022), and written informed consent was obtained from all participants.

2.3. Human sample collection

Peritoneal fluid, eutopic endometrial tissue, and ectopic endometrial tissue were obtained from patients with endometriosis (mean age: 35.2 years; range: 22-50 years; experimental group). The ectopic endometrial tissue was collected from ovarian endometriotic lesions. Control samples (mean age: 32.5 years; range: 22-50 years) were collected from women without endometriosis undergoing surgery for tubal adhesions or ovarian cysts. All participants were recruited from Shandong Provincial Hospital, Shandong University. The diagnosis of endometriosis was confirmed intraoperatively and by histopathology. None of the participants had received steroid hormone therapy, had pelvic inflammatory disease, or related complications within 3–6 months prior to sampling.

2.4. Isolation and culture of primary human cells

Peritoneal fluid was collected intraoperatively, diluted with 100 mL sterile saline, and reaspirated. Cells were collected by centrifugation (4 °C, 400 g, 10 min), red blood cells lysed (R1010, Solarbio; Beijing, China), and the remaining cells either analyzed by flow cytometry or cultured in complete DMEM medium (10569044, Gibco; Carlsbad, USA) supplemented with 10% fetal bovine serum (FBS; 35-081-CV, Corning; NY, USA) and 1% penicillin-streptomycin (PS; 100 × , P1400, Solarbio) at 37 °C, 5% CO₂.

ESCs were isolated from normal, eutopic and ectopic endometrial tissues. Tissues were minced, digested with 0.25% collagenase II/IV at 37 °C for up to 60 min, filtered (70 μm and 40 μm), centrifuged (400 g, 5 min), and cultured in complete DMEM medium at 37 °C, 5% CO₂. Cell purity was confirmed by immunofluorescence staining for vimentin and cytokeratin.

2.5. mouse model of endometriosis

C57BL/6 female mice (6–8 weeks old) were purchased from Jiangsu Gempharmatech Biotechnology Co., Ltd. (Jiangsu, China) and maintained in a specific pathogen-free (SPF) facility at the Experimental Animal Center of Shandong University (Jinan, China). Estrous cycles were synchronized with subcutaneous estradiol benzoate (E2) (3 μg/mouse) (MCE; Nanjing, China). Uteri from donor mice were minced (<1 mm³), pooled, and suspended in PBS, then intraperitoneally injected into recipient mice; after injection, tissue fragments primarily adhere and establish lesions around the uterine adnexa, forming adhesions and endometriotic foci [[33], [34], [35]]. Recipients received estradiol benzoate on days 0, 5, and 10 to support lesion establishment. Mice were euthanized 16 days post-implantation for analysis. Clodronate liposomes and control liposomes (PBS) (Yeasen) were injected intraperitoneally every three days to remove peritoneal macrophages. IRG1-IN-1 (MedChemExpress, HY-148335) was prepared as stocks in DMSO (final DMSO ≤0.1%) with matched vehicle controls; dosing followed the study schedule, and vehicle-only controls were handled identically.

2.6. Isolation of human peripheral blood mononuclear cells (PBMCs)

Peripheral blood was obtained from healthy female volunteers aged 20-40 years, with informed consent. Blood samples were collected in anticoagulant tubes, and human peripheral blood mononuclear cells were isolated using Human Peripheral Blood Lymphocyte Isolation Solution (Prepared by Ficoll) (P8900, Solarbio). CD14⁺ mononuclear cells were purified using the Human CD14 Selection Kit (17858, Stem cell). The purified cells were plated in complete DMEM medium supplemented with 20 ng/mL recombinant human M-CSF (MCE) and cultured at 37 °C in a humidified atmosphere of 5% CO₂ for 5 days to generate macrophages.

2.7. Isolation and culture of primary mouse cells

Peritoneal cells were collected by injecting 5 mL cold PBS into the peritoneal cavity, and aspirating the fluid. After centrifugation (4 °C, 400 g, 10 min), cells were either analyzed by flow cytometry or cultured in complete DMEM medium at 37 °C, 5% CO₂.

Lesions and uteri were minced (<1 mm³), digested with 0.25% collagenase II/IV at 37 °C for <60 min, filtered (70 μm/40 μm), centrifuged (1000 rpm, 5 min), and processed as above.

2.8. Generation of bone marrow-derived macrophages (BMDMs)

Femurs and tibias from euthanized 7-week-old female C57BL/6 mice were flushed with PBS. Bone marrow cells were filtered (40 μm), centrifuged (400 g, 5 min), and red blood cells lysed. Cells were cultured in DMEM with 20 ng/mL macrophage colony-stimulating factor (M-CSF) for 5 days to induce BMDMs.

2.9. Metabolomics analysis

Endometrial and lesion tissues (∼5 mm diameter) from patients and mice were snap-frozen in liquid nitrogen and stored at −80 °C. Peritoneal cells were centrifuged (4 °C, 400 g, 10 min). The cell pellet was resuspended in complete DMEM medium and plated in culture dishes for 1 h at 37 °C with 5% CO₂ to allow macrophages to adhere. The adherent cells were collected, washed three times with PBS, gently scraped, pelleted by centrifugation, and immediately snap-frozen in liquid nitrogen for storage at −80 °C.

For analysis of ESC-derived metabolites, isolated primary ESCs were counted and seeded at a density of 2 × 10⁶ cells per well. Cells were cultured in serum-free DMEM for 48 h. The culture supernatant was then collected, centrifuged to remove cell debris, and the clarified supernatant was aliquoted and snap-frozen in liquid nitrogen for storage at −80 °C.

All samples were analyzed by Hangzhou Lianchuang Biotechnology for targeted and untargeted metabolomics.

2.10. Protein extraction and western blot

Tissues and cells were lysed in radioimmunoprecipitation assay lysis buffer (RIPA, P0013B, Beyotime; Suzhou, China) with phenylmethylsulfonyl fluoride (PMSF) on ice (15 min), centrifuged (12000 rpm, 20 min, 4 °C), and denatured. Proteins were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to PVDF membranes (Millipore, Billerica). After blocking in 5% milk in TBST, membranes were incubated overnight at 4 °C with primary antibodies: Beta Actin Monoclonal antibody(1:20000, 66009-1-Ig; Proteintech), Beta Tubulin Recombinant antibody (1:5000, 80713-1-RR; Proteintech), Anti-IRG1 antibody (1:1000; ab222411; abcam; Cambridge, UK), Nrf2 monoclonal antibody (1:2000; A21176; abclonal; Wuhan, China), iNOS Polyclonal antibody (1:500; 22226-1-AP; Proteintech), NOX2 Polyclonal antibody (1:3000; 19013-1-AP; Proteintech), p-p38 MAPK Polyclonal antibody (1:2000; 28796-1-AP; Proteintech), and p38 MAPK Polyclonal antibody (1:4000; 14064-1-AP; Proteintech). After washing, membranes were incubated with HRP-conjugated secondary antibodies, and signals were detected using CHAMPCHEMI 910. Band intensity was quantified with ImageJ and normalized to Actin or Tubulin.

2.11. Immunofluorescence

Tissue specimens were fixed in 4% paraformaldehyde (P1110; Solarbio), paraffin-embedded, and sectioned at 4 μm thickness. After deparaffinization and rehydration, antigen retrieval was performed in 1 × Tris-EDTA buffer (C1038; Solarbio) using microwave heating. Slides were blocked with BSA (ST023, Beyotime) for 30 min at room temperature, then incubated overnight at 4 °C with primary antibodies: anti-IRG1 (1:300, PA5-102893, Invitrogen; Carlsbad, CA, USA) and anti-vimentin (1:600, 60330-1-Ig, Proteintech). After washing with PBS, sections were incubated with corresponding fluorescent secondary antibodies for 30 min at ∼26 °C. Nuclear staining was performed with DAPI (ab104139, abcam). Images were acquired using a fluorescence microscope.

2.12. RNA extraction and quantitative PCR

Total RNA was extracted from cultured cells using TRIzol reagent (Invitrogen), and reverse-transcribed into cDNA with Evo M-MLV RT premix (Accurate Biotechnology) following the manufacturer’s instructions. Quantitative PCR was performed using Quant Studio 1 (Thermo Fisher Scientific; Waltham, MA, USA). Primer sequences are listed in Supplemental Table. ACTIN served as the internal control. Relative gene expression was calculated using the 2^∧−ΔΔCt method.

2.13. Flow cytometry

Collected cells were washed twice with ice-cold PBS and centrifuged (350 g, 5 min). For surface markers such as CD45 (anti-mouse CD45; 157213, BioLegend; San Diego, CA, USA) and F4/80 (anti-mouse F4/80; 157305, BioLegend), cells were incubated with diluted primary antibodies at 4 °C in the dark for 30 min. For intracellular markers such as iNOS (anti-mouse iNOS; Invitrogen), cells were permeabilized with BD Perm/Wash buffer (BD Biosciences; San Jose, CA, USA) for 15 min, then incubated with primary antibodies at 4 °C in the dark for 30 min. After staining, cells were washed twice, centrifuged, and resuspended in permeabilization buffer for flow cytometric analysis.

2.14. Cell culture and transfection

Human siIRG1,mouse siNrf2 or scramble siRNAs (50 nM, Sangon) were transfected into cells by Lipofectamine RNAiMAX transfection reagent (Life Technologies) for 8 h. Then cell media was replaced and continued culture for 24-72 h.

2.15. Cell migration assay

Mouse peritoneal lavage fluid was centrifuged (350 g, 10 min), and the cell pellet was resuspended in complete DMEM and plated in dishes. Cells were incubated at 37 °C, 5% CO₂ for 2 h to allow macrophage adhesion. Adherent macrophages were gently collected with a cell scraper and seeded into 24-well plates at 3 × 10⁵ cells per well.

BMDMs were similarly collected and seeded at 3 × 10⁵ cells per well. Cells were randomly assigned to two groups: LPS group (100 ng/mL LPS; Sigma-Aldrich, St. Louis, MO, USA, with 0.1‰ DMSO added after 2 h) and LPS + 4-OI group (100 ng/mL LPS, followed by 250 μM 4-OI [MCE] and 0.1‰ DMSO after 2 h). After 12 h of stimulation, the medium was replaced with fresh complete medium. Macrophages derived from peripheral blood of humans are seeded into 24-well plates at 3 × 10⁵ cells per well.

Transwell inserts (8 μm pore size; Thermo Fisher Scientific) were placed in each well. Mouse primary ESCs (≤2 passages) and human primary ESCs (with siRNA treatment), resuspended in serum-free medium (1 × 10⁵ cells per chamber), were added to the upper compartment. After 24 h, migrated cells on the lower surface were fixed and stained with crystal violet for quantification.

2.16. Macrophage phagocytosis assay

Mouse ESCs (mESCs) and human primary ESCs (with siRNA treatment) were labeled with PKH67 (Sigma-Aldrich), a membrane dye detectable via the FITC channel. PKH67-labeled ESCs were co-cultured with mouse peritoneal macrophages, LPS/LPS + 4-OI-treated BMDMs (as described above) and macrophages derived from peripheral blood of humans at a 1:4 ratio (ESCs: macrophages). After 8 h, cells were collected, centrifuged at 4 °C (350 g, 5 min), and washed twice with PBS. Macrophages were stained with anti-F4/80-APC antibody (17-4801-82, APC, eBioscience; San Diego, CA, USA). Flow cytometry was performed using a Beckman Coulter instrument, and data were analyzed with CytExpert (v2.4.0.28, Beckman Coulter; Brea, CA, USA). The percentage of PKH67⁺ cells among F4/80⁺ macrophages was quantified as a measure of phagocytosis.

2.17. Co-culture experiments

Mouse ESCs or human ectopic/eutopic ESCs (3 × 10⁵ cells per chamber) were seeded into 0.4 μm pore-size transwell inserts (140620, Thermo Fisher Scientific) and placed in 6-well plates. Mouse BMDMs (2 × 10⁶ cells/well) or plate-adherent macrophages purified from the ascites of EM patients and non-EM controls (1.5 × 10⁶ cells/well) were plated in the lower chamber. After 16 h of co-culture, transwell inserts were removed and RNA was extracted from BMDMs or patient-derived macrophages for subsequent analysis.

2.18. Cell viability assay (CCK-8)

Cell viability was assessed using the Cell Counting Kit-8 (CCK-8; C0038, Beyotime). Briefly, BMDMs were seeded into 96 well plates at 2 × 10⁴ cells per well in 100 μL complete medium and allowed to adhere overnight. The indicated treatments (e.g., LPS, 4-OI, DPI, CQ, or ML-SA1) were applied for the specified durations. CCK-8 reagent was then added at 10 μL per well (final 1:10, v/v), and plates were incubated for 1 h at 37 °C protected from light. Absorbance at 450 nm was recorded on a microplate reader with medium plus CCK-8 (no cells) as the blank. Cell viability was calculated as (A450sample − A450blank)/(A450control − A450blank) × 100%. Each condition was measured in technical triplicate and repeated in at least three independent experiments.

2.19. Masson’s trichrome staining

Lesion tissues from 4-OI treated and control EM model mice were rapidly fixed in 4% paraformaldehyde (Solarbio), paraffin-embedded, and sectioned at 4 μm thickness. After deparaffinization and rehydration, sections were stained with the Modified Masson's Trichrome Stain Kit (G1346; Solarbio) according to the manufacturer's instructions. Bright-field images were captured at 4 × , 10 × , and 20 × magnifications. Quantitative analysis was performed using ImageJ software (version 1.5; Wayne Rasband).

2.20. Preparation of LNPs

SM-102 (HY-134541, MCE), DSPC (HY–W040193, MCE), cholesterol (C8667-1G, Sigma), and DMG-PEG 2000 (HY-112764, MCE) were dissolved in ethanol at a ratio of 50:10:38.5:1.5. siRNA (siIrg1, NC, Sangon) was dissolved in citrate buffer (10 mM). The lipids were rapidly injected into the mRNA solution and pipetted vigorously 80–100 times, followed by vortexing for 10 s. After adding an equal volume of PBS, the mixture was dialyzed for 1 h. The resulting LNPs were administered to mice via intraperitoneal injection.

2.21. Lysosomal acidity measurement

Macrophages were gently collected by scraping and resuspended at 1 × 10⁶ cells/mL. For lysosomal pH detection, cells were stained with LysoSensor Yellow/Blue DND-160 (1:1000, L7545, Thermo Fisher Scientific) in pre-warmed complete medium at 37 °C for 30 min, then incubated for an additional 3 min at 37 °C. After washing with ice-cold PBS, cells were transferred to black 96-well plates (1 × 10⁶ cells/200 μL per well). Fluorescence was measured at Ex 360 nm/Em 440 nm and Ex 360 nm/Em 550 nm on a Synergy H1 microplate reader (BioTek) following the addition of 10 μM valinomycin and 10 μM nigericin. A standard curve for lysosomal pH was generated using the intracellular pH calibration buffer kit (P35379, Thermo Fisher Scientific).

For LysoSensor Green staining (17535, Thermo Fisher Scientific), BMDMs were incubated with the probe at 37 °C for 30 min, collected, washed with PBS, and analyzed for mean fluorescence intensity by flow cytometry.

2.22. Intracellular calcium measurement

Collected macrophages (patient-derived peritoneal macrophages or BMDMs) were incubated with 5 μM Fluo-4 AM (F14201, Thermo Fisher Scientific) in PBS at 37 °C. Real-time calcium levels were assessed by measuring Fluo-4 AM fluorescence intensity upon ionomycin stimulation using flow cytometry. For quantification of intracellular calcium content, BMDMs were analyzed using a colorimetric calcium assay kit (S1063S, Beyotime) according to the manufacturer’s instructions.

2.23. NOX2 activity assay

Collected BMDMs were washed three times with ice-cold PBS, resuspended in ice-cold PBS, and lysed by ultrasonic disruption. NOX2 enzyme activity was measured using a commercial assay kit (Nanjing Jiancheng Bioengineering Institute) according to the manufacturer’s instructions. Protein concentrations were determined in parallel using a BCA assay kit (AR1189, Boster).

2.24. Statistical analysis

Data were analyzed and visualized using GraphPad Prism 10.2.3 (GraphPad, USA) and are presented as mean ± standard deviation. Normality and homogeneity of variance were assessed prior to statistical testing. Differences between two groups were evaluated using an unpaired two-tailed Student’s t-test (P < 0.05). For comparisons among multiple groups, one-way ANOVA with Tukey’s post hoc test was used for normally distributed data, while the Kruskal–wallis test was applied for non-parametric data (P < 0.05).

2.25. Data availability

All study data are included in the article, supplemental information, and the Supporting Data Values file.

3. Result

3.1. ACOD1-driven itaconate production in human ectopic endometrial stromal cells

To assess metabolic alterations in endometriotic lesions, we first performed untargeted metabolomic profiling to compare the metabolite spectra of lesion tissues from patients with EM and healthy controls. The results revealed significant differences in the metabolic profiles between the EM and control groups (Supplemental Fig. 1). Pathway enrichment analysis further indicated that pathways such as tricarboxylic acid (TCA) cycle, and amino acid biosynthesis were significantly enriched in EM lesions (Fig. 1A). Notably, multiple organic acids-including itaconic acid and citric acid-were elevated to varying degrees in endometriotic lesions (Fig. 1B), with itaconate exhibiting a particularly pronounced increase (Fig. 1C). To investigate the source of this accumulation, we next examined the expression of aconitate decarboxylase 1 (ACOD1), the key enzyme responsible for itaconate synthesis. ACOD1 protein levels were markedly higher in endometriotic lesions compared with both eutopic endometrium from EM patients and normal endometrium from healthy controls (Fig. 1D and E). Immunofluorescence co-localization analysis revealed that ACOD1 upregulation was predominantly localized to ectopic ESCs (Fig. 1F). Consistently, isolated ectopic ESCs displayed significantly elevated ACOD1 mRNA and protein levels relative to eutopic and normal ESCs (Fig. 1G–I), and secreted markedly higher amounts of itaconate into culture supernatants (Fig. 1J). In contrast, ACOD1 expression in peritoneal macrophages from EM patients remained unchanged (Fig. 1K). These findings identify ectopic ESCs as the principal cellular source of itaconate in EM lesions, linking metabolic reprogramming in EM.

Fig. 1.

Aberrant activation of the itaconate metabolic pathway in endometriotic lesions and stromal cells from Patients (A) Quantitative enrichment analysis of altered metabolic pathways among ectopic lesions, eutopic endometrium, and normal endometrium, visualized as a bubble plot. (B) Hierarchical clustering of differential metabolites in the three tissue types. Heatmap values: row-wise z-scores (no log transform). Colors indicate relative levels; scale centered at 0 (≈[−1,1]). (C) Quantification of itaconate in ectopic lesions (n = 10), eutopic endometrium (n = 5), and normal endometrium (n = 5). (D, E) Western blot analysis (D) and quantification (E) of ACOD1 protein in lesion tissues from endometriosis (EM) and non-EM patients (n = 12/group). (F) Representative immunofluorescence staining for DAPI (nuclei, blue), Vimentin (red), and ACOD1 (green) in normal, eutopic, and ectopic endometrial tissues; scale bars, 150 μm. (G, H) Western blot analysis (G) and quantification (H) of ACOD1 in stromal cells derived from normal (nor-ESC), eutopic (eu-ESC), and ectopic (ec-ESC) endometrium (n = 3/group). (I) qRT-PCR analysis of ACOD1 mRNA in nor-ESC, eu-ESC, and ec-ESC (n = 3/group). (J) Itaconate levels in cell supernatants of nor-ESC (n = 3), eu-ESC (n = 3), and ec-ESC (n = 6).(K) Western blot analysis of ACOD1 expression in peritoneal macrophages from EM and non-EM patients; LPS-stimulated macrophages from non-EM patients serve as positive control. (Data are presented as mean ± SEM. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001, ns: not significant.) EM, endometriosis; NC, non-EM control; LPS, lipopolysaccharide; Eutopic: endometrial tissue collected from the uterine cavity of the same patient (in situ endometrium); Ectopic: endometriosis lesions collected from intraperitoneal sites.

3.2. Mouse ectopic endometrial stromal cells exhibit ACOD1-Dependent itaconate production

To further investigate the role of itaconate in EM, we established mouse model of EM (Fig. 2A and B). Untargeted metabolomic profiling of murine lesions revealed a metabolic landscape closely resembling that of human EM lesions. Pathway enrichment analysis further highlighted significant upregulation of pathways involved in carbon metabolism and amino acid biosynthesis in EM mouse lesions (Fig. 2C). Analysis of Krebs cycle intermediates showed elevated levels of itaconate and citrate, with itaconate exhibiting the most pronounced increase (Fig. 2D and E), supporting a pivotal role for dysregulated ESC metabolism in EM pathology. At the protein level, ACOD1 was markedly upregulated in EM mouse lesions compared with eutopic and normal endometrium (Fig. 2F and G). Consistently, ectopic ESCs isolated from mice exhibited significantly elevated ACOD1 expression at both protein and mRNA levels relative to eutopic and normal ESCs (Fig. 2H–J), closely mirroring the patterns observed in human EM samples. These findings confirm that dysregulated ACOD1-itaconate metabolism in ESCs is conserved across species, reinforcing a pivotal role for ESC metabolic reprogramming in shaping the EM immune microenvironment.

Fig. 2.

Activation of the itaconate metabolic pathway in a mouse model of endometriosis (A) Schematic diagram of the mouse endometriosis model. (B) Representative intra-abdominal images. Top: normal uterus from control mice. Middle and bottom: recipient mice with endometriotic lesions (red circles); lower panels show magnified views. (C) Bubble plot of quantitative enrichment analysis showing altered metabolic pathways among ectopic lesions, eutopic endometrium, and normal endometrium in mice. (D) Heat map of selected metabolites across normal, eutopic, and ectopic tissues from the mouse endometriosis model. Row-wise z-scored intensities (no log transform); colors reflect relative abundance per metabolite, centered at 0 (≈[−1, 1]). (E) Quantification of itaconate in ectopic lesions (n = 10), eutopic endometrium (n = 5), and normal endometrium (n = 5). (F, G) Western blot analysis (F) and quantification (G) of ACOD1 protein in lesion tissues from endometriosis (EM) and control mice (n = 3/group). (H, I) Western blot analysis (H) and quantification (I) of ACOD1 in stromal cells from normal (nor-ESC), eutopic (eu-ESC), and ectopic (ec-ESC) endometrium (n = 3/group). (J) qRT-PCR analysis of Acod1 mRNA in nor-ESC, eu-ESC, and ec-ESC (n = 3/group); LPS-stimulated macrophages from control mice serve as positive control. (Data are presented as mean ± SEM. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns: not significant.) E2, estradiol benzoate; EM, endometriosis; NC, non-EM control; LPS, lipopolysaccharide.

3.3. The ACOD1-Itaconate axis promotes lesion progression and serves as a therapeutic target in endometriosis-

To investigate the functional role of the ACOD1/itaconate pathway in EM, we employed both pharmacological, genetic, and gain-of-function approaches. Pharmacologically, IRG1-IN-1, a previously reported ACOD1 inhibitor that suppresses ACOD1 activity in vivo [36], was administered via intraperitoneal injection (Fig. 3A). This treatment markedly reduced lesion size in EM mice (Fig. 3B–D). Complementarily, lipid nanoparticle-mediated delivery of siRNAs targeting Irg1 (siIrg1-LNPs) similarly reduced lesion burden and robustly downregulated ACOD1 expression within lesions (Fig. 3E–I), confirming the pathogenic contribution of ACOD1. Conversely, administration of 4-octyl itaconate (4-OI), a cell-permeable itaconate derivative that mimics endogenous itaconate, significantly increased lesion weight, stiffness, and fibrosis (Fig. 3J–N). Together, these complementary interventions demonstrate that the ACOD1/itaconate pathway drives lesion growth and tissue remodeling in EM, highlighting it as a critical mechanistic and therapeutic axis.

Fig. 3.

Itaconate Pathway Intervention Alters Lesion Burden and Fibrosis in Endometriosis Mice (A)Schematic of IRG1-IN-1 (0.5 mg/kg) or PBS administration in the mouse endometriosis (EM) model. (B) Representative images of endometriotic lesions in PBS- and IRG1-IN-1-treated mice. (C) Gross morphology of excised lesions from each group. (D) Quantification of lesion weight (n = 6/group). (E) Schematic of the experimental design for si-Irg1-LNP administration in EM mice. (F, G) Representative images (F) and gross morphology (G) of lesions in si-ctrl-LNP and si-Irg1-LNP groups. (H) Quantification of lesion weight (n = 6/group). (I) Western blot analysis of ACOD1 expression in lesions from si-ctrl-LNP and si-Irg1-LNP groups.(J) Schematic of the experimental design for 4-OI intervention in EM mice. (K) Representative images of endometriotic lesions after treatment with PBS or 4-OI (red arrows). (L) Gross morphology of lesions from each group. (M) Quantification of lesion weight (n = 6/group). (N) Masson's trichrome staining of lesions with quantification of fibrotic area (n = 6/group). (O) Measurement of itaconate content in peritoneal macrophages from EM and non-EM mice (n = 8/group). (P) Western blot analysis of ACOD1 in peritoneal macrophages from EM and control mice E2, estradiol benzoate; EM, endometriosis; LNP, lipid nanoparticle; LPS, lipopolysaccharide; 4-OI, 4-octyl itaconate; PBS, phosphate-buffered saline; si-Irg1, small interfering RNA targeting Irg1.

3.4. ESC-derived itaconate reprograms macrophage function to promote EM progression

To determine whether ESC-derived itaconate modulates immune cells within the EM microenvironment, we focused on peritoneal macrophages, central effectors in EM pathogenesis. Macrophages from EM mice exhibited markedly elevated intracellular itaconate (Fig. 3O), despite unchanged ACOD1 expression (Fig. 3P), indicating uptake of ESC-secreted itaconate rather than endogenous synthesis. This establishes a paracrine metabolic circuit linking ESC metabolic reprogramming to macrophage modulation. Flow cytometry revealed that macrophages in the peritoneal cavity of EM mice exhibited a reduced proportion of iNOS⁺ macrophages (Supplemental Fig. 2). To extend these findings to the human context, peritoneal macrophages from EM patients displayed significantly lower pro-inflammatory gene expression compared to healthy controls (Fig. 4A). In co-culture assays, only ectopic ESCs led to a similar downregulation of pro-inflammatory gene expression as observed in the EM group, but not normal ESCs (Fig. 4A), demonstrating that ectopic ESCs actively inhibit pro-inflammatory macrophage activation. Additionally, silencing IRG1 in ectopic ESCs significantly reduced their ability to inhibit macrophage phagocytosis and diminished the macrophage-mediated promotion of ESC migration in co-culture (Fig. 4B–D), directly implicating the ACOD1/itaconate pathway. In vivo, genetic knockdown of Acod1 via siIrg1-LNP restored pro-inflammatory macrophage polarization (Fig. 4E and F). Pharmacological inhibition of ACOD1 with IRG1-IN-1 similarly restored this phenotype (Fig. 4G and H) and reduced lesion itaconate (Fig. 4I). In contrast, 4-OI decreased the number of iNOS⁺ macrophages and suppressed inflammatory mediator production (Fig. 4J). Functionally, 4-OI promoted macrophage-driven ESC migration and impaired macrophage phagocytosis (Fig. 4K–M). Supporting in vitro experiments using mouse BMDMs and mESCs further confirmed that 4-OI suppresses macrophage activation and function (Supplemental Fig. 3A–G), but has no direct effect on ESC migration when macrophages are absent (Supplemental Fig. 3H). Ectopic ESC-conditioned media (ESC-CM) suppressed macrophage activation to a level comparable to 250 μM 4-OI (Supplemental Fig. 3I); cell viability under these conditions is shown (Supplemental Fig. 3J). Strikingly, depletion of peritoneal macrophages with clodronate liposomes completely abolished the pro-endometriotic effects of 4-OI (Fig. 4N–Q; Supplemental Fig. 3K–M), establishing macrophages as indispensable mediators of itaconate-induced disease exacerbation. Together, these findings demonstrate that ESC-derived itaconate reprograms peritoneal macrophages into an immunosuppressive state, thereby driving lesion persistence and progression, and highlight macrophages as indispensable mediators of ACOD1/itaconate-driven EM pathology.

Fig. 4.

Itaconate-mediated modulation of macrophage activity alters endometriotic lesion progression (A) mRNA expression of IL1B, IL6, TNFA, and iNOS in peritoneal macrophages from endometriosis (EM) and non-EM patients, as well as non-EM macrophages co-cultured with either nor-ESC or ectopic ec-ESC for 12 h (n = 3/group).(B) Migration of si-Irg1 or NC-treated ectopic ESCs was assessed after co-culture with PBMCs for 48 h. Quantification of migrated cells in five random fields per group are shown (n = 3/group). (C-D) PKH67-labeled human ectopic ESCs pretreated with si-Irg1 or NC were co-cultured with PBMCs for 8 h. Phagocytosis of ESCs by macrophages was analyzed by flow cytometry with representative gating (C), PKH67 signal and quantification of PKH67-positive macrophages (D) (n = 6/group).(E,G) Flow cytometry analysis and quantification of iNOS⁺ peritoneal macrophages from mouse model in Fig. 3E (E), Fig. 3A (G). (n = 6/group). (F,H) mRNA levels of Il1b, Il6, Nos2, and Tnf in peritoneal macrophages from mouse model in Fig. 3E (E), Fig. 3A (G) (n = 6/group).(I)Quantification of itaconate in endometriotic lesion tissue by LC–MS from mice treated with IRG1-IN-1 or vehicle. (n = 3/group).(J) Flow cytometry analysis and quantification of iNOS⁺ peritoneal macrophages from mouse model in Fig. 3J (n = 6/group).(K,L) Migration of mESCs induced by peritoneal macrophages from PBS- or 4-OI-treated mice was assessed by transwell assay; representative images and quantification of migrated cells are shown (n = 5/group). (M) Flow cytometry analysis and quantification of phagocytosis of PKH67-labeled mESCs by peritoneal macrophages (n = 6/group).(N) Schematic of the experimental design for clodronate liposome-mediated macrophage depletion and 4-OI intervention in EM mice. (O, P) Representative images (O) and gross morphology (P) of endometriotic lesions in control, clodronate, PBS, and 4-OI groups. (Q) Quantification of lesion weight in mice treated with clodronate liposomes or control liposomes, with or without 4-OI (n = 6/group). (Data are presented as mean ± SEM. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.) E2, estradiol benzoate; EM, endometriosis; LNP, lipid nanoparticle; LPS, lipopolysaccharide; 4-OI, 4-octyl itaconate; Clod, clodronate liposome; PBS, phosphate-buffered saline; si-Irg1, small interfering RNA targeting Irg1.

3.5. Itaconate Suppresses inflammation through NRF2 activation

Previous studies have suggested that activation of the nuclear factor erythroid 2-related factor 2 (NRF2) pathway is a key mechanism underlying the anti-inflammatory activity of itaconate [[37], [38], [39]]. To determine whether NRF2 mediates the immunomodulatory effects of itaconate in EM, we first examined its role in vitro. In lipopolysaccharide (LPS)-stimulated bone marrow-derived macrophages (BMDMs), treatment with 4-octyl itaconate (4-OI) markedly increased NRF2 protein expression, as detected by Western blotting (Fig. 5A and B) and upregulated classical NRF2 target genes (Supplemental Fig. 4A). Pharmacological inhibition of NRF2 with ML385 abrogated this upregulation (Fig. 5C and D) and partially reversed the suppressive effects of 4-OI on macrophage activation. Specifically, ML385 restored the pro-inflammatory phenotype (Fig. 5E) and increased production of pro-inflammatory cytokines (Fig. 5F). Consistently, NRF2 knockdown in BMDMs attenuated the ability of 4-OI to suppress pro-inflammatory activation (Fig. 5G; Supplemental Fig. 4B), indicating that Nrf2 activation is required for the anti-inflammatory actions of 4-OI in macrophages.

Fig. 5.

Inhibition of NRF2 Signaling Reverses the Anti-inflammatory and Anti-endometriotic Effects of Itaconate in Macrophages and a Mouse Model of Endometriosis (A,B) Western blot analysis (A) and quantification (B) of NRF2 protein levels in bone marrow-derived macrophages (BMDMs) pretreated with LPS (100 ng/mL), 4-octyl itaconate (4-OI, 250 μM), or both for 12 h (n = 3/group). (C,D) Western blot analysis (C) and quantification (D) of NRF2 in BMDMs treated with LPS, 4-OI, and the NRF2 inhibitor ML385 (2.5 μM) for 12 h (n = 3/group). (E) Flow cytometry analysis and quantification of iNOS⁺ BMDMs after indicated treatments (n = 3/group). (F) mRNA expression of pro-inflammatory genes (Il1b, Il6, Nos2, Tnf) in BMDMs under different conditions (n = 3/group). (G) Flow cytometry analysis and quantification of iNOS⁺ BMDMs following NRF2 knockdown (siNrf2) with or without 4-OI, compared to negative control (NC) (n = 3/group).(H) Schematic of the experimental design for ML385 administration in a mouse model of endometriosis. (I) Representative images of endometriotic lesions in PBS- and ML385-treated mice (lesions marked by red circles). (J) Gross morphology of lesions in both groups(n = 6/group). (K) Quantification of lesion weight (n = 6/group). (L) Flow cytometry analysis and quantification of iNOS⁺ peritoneal macrophages from EM mice treated with PBS or ML385 (n = 6/group). (M) mRNA expression of Il1b, Il6, Nos2, and Tnf in peritoneal macrophages from each group (n = 6/group). (Data are presented as mean ± SEM. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.) E2, estradiol benzoate; EM, endometriosis; LPS, lipopolysaccharide; 4-OI, 4-octyl itaconate; ML385, NRF2 inhibitor.

To further validate these findings in vivo, we administered intraperitoneal injections of ML385 in the EM mouse model (Fig. 5H). Compared with control mice, ML385 treatment significantly decreased lesion burden, as reflected by lower lesion weight (Fig. 5I–K). Moreover, peritoneal macrophages isolated from ML385-treated mice displayed enhanced pro-inflammatory activity, evidenced by an increased proportion of pro-inflammatory macrophages (Fig. 5L) and elevated expression of pro-inflammatory gene (Fig. 5M). Together, these results demonstrate that NRF2 signaling plays a critical role in mediating the anti-inflammatory effects of 4-OI both in vitro and in vivo. Notably, within the same assay window, itaconate showed no comparable induction of NRF2 protein or target genes (Supplemental Fig. 4C–D), consistent with differences in delivery and electrophilicity relative to 4-OI, which we explicitly position as a delivery-facilitating experimental tool rather than a biochemical surrogate of endogenous itaconate. Although NRF2 blockade reduced the effects of 4-OI, a residual response remained. This could reflect incomplete NRF2 blockade and/or engagement of other pathways; accordingly, we next investigated mechanisms that may account for the remaining activity.

3.6. Itaconate regulates lysosomal function to suppress macrophage pro-inflammatory activity

Lysosomes have emerged as central hubs linking cellular clearance to immunometabolic signaling. In macrophages, lysosomal function is tightly regulated by both biogenesis and acidity, and alterations in lysosomal pH profoundly impact macrophage phenotype [32]. Anti-inflammatory macrophages typically exhibit more acidic lysosomes compared with their pro-inflammatory counterparts [32], and experimental modulation of lysosomal acidity has been shown to reshape macrophage function [40,41]. We first examined lysosomal properties in macrophages from EM patients and observed that peritoneal macrophages displayed significantly higher lysosomal acidity compared with controls (Fig. 6A). To assess whether itaconate modulates lysosomal function, LPS-stimulated PBMCs and BMDMs were treated with 4-OI and analyzed using LysoSensor Green. 4-OI markedly enhanced lysosomal acidity (Fig. 6B; Supplemental Fig. 5A–B), which was further confirmed by direct pH measurements showing a reduction in lysosomal pH from 5.69 to 5.07 in pro-inflammatory macrophages (Fig. 6C; Supplemental Fig. 5C). After extended stimulation, itaconate also enhanced lysosomal acidity, consistent with the effect seen with 4-OI (Supplemental Fig. 5D–E). Notably, the lysosomal alkalinizing agent chloroquine (CQ) abolished the acidifying effect of 4-OI (Fig. 6D; Supplemental Fig. 5F) and reversed its inhibitory impact on macrophage pro-inflammatory phenotype (Fig. 6E and F). These findings indicate that lysosomal acidification contributes to the anti-inflammatory actions of itaconate.

Fig. 6.

Itaconate Regulates Lysosomal Function, Calcium Signaling, and p38 MAPK Pathway in Macrophages (A) Lysosomal acidification in ascites-derived macrophages from non-endometriosis (NC) and endometriosis (EM) patients assessed by LysoSensor fluorescence intensity (n = 6/group). (B) Lysosomal acidification in PBMC-derived macrophages after LPS,IL-4 and 4-OI treatment (n = 3/group). (C) Lysosomal pH value of BMDMs after LPS and 4-OI treatment (n = 3/group). (D, E) Effect of chloroquine (CQ) on LysoSensor intensity (D) and proportion of iNOS⁺ BMDMs (E) after LPS and 4-OI treatment (n = 3/group). (F) mRNA levels of Il1b, Il6, Nos2, and Tnf in BMDMs under indicated treatments (n = 3/group). (G) Intracellular Ca²⁺ dynamics in BMDMs after treatments, measured by Fluo-4. (H) Intracellular Ca²⁺ dynamics in ascites-derived macrophages from NC and EM patients. (I) Quantification of intracellular calcium in peritoneal macrophages from NC and EM patients (n = 3/group). (J) Quantification of intracellular calcium in BMDMs (n = 5/group). (K) mRNA expression of lysosomal calcium channel genes (MCOLN1, MCOLN2, TPC1, TPC2) in ascites-derived macrophages from NC and EM patients (n = 3/group). (L) lysosomal calcium channel genes mRNA in PBMC-derived macrophages co-cultured with normal ESCs (Nor-ESC), eutopic ESCs (Eu-ESC), or ectopic ESCs (Ec-ESC) from patients (n = 5/group). (M − N) Fluo-4-based Ca²⁺ dynamics in BMDMs treated with LPS, 4-OI, ML-SA1 (M), or CQ (N). (O) Flow cytometry analysis and quantification of iNOS⁺ BMDMs after indicated treatments (n = 3/group). (P) mRNA levels of pro-inflammatory genes in BMDMs under different treatments (n = 3/group). (Q) Western blot and quantification of p-p38 and total p38 in BMDMs with LPS ± 4-OI (n = 3/group). (R–S) Western blot and quantification of iNOS, p-p38, and total p38 in BMDMs treated with LPS, 4-OI, and CQ (R), or ML-SA1 (S) (n = 3/group).

(Data are presented as mean ± SEM. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.)

EM, endometriosis; NC, non-EM control; LPS, lipopolysaccharide; 4-OI, 4-octyl itaconate; CQ, chloroquine; ML-SA1, MCOLN channel agonist; PBMC, peripheral blood mononuclear cell; ESC, endometrial stromal cell.

Given that lysosomal pH regulates Ca²⁺ release [[42], [43], [44]], we next investigated whether the acidifying effect of 4-OI impacts lysosomal Ca²⁺ signaling. We found that reduced lysosomal pH in 4-OI-treated macrophages was accompanied by decreased intracellular Ca²⁺ levels (Fig. 6G). Consistently, peritoneal macrophages from EM patients displayed diminished intracellular Ca²⁺ compared with controls (Fig. 6H), a result validated in BMDMs and peritoneal macrophages using calcium quantification assays (Fig. 6I and J). Mechanistically, analysis of lysosomal Ca²⁺ release channels revealed that mucolipin-2 (Mcoln2) was downregulated in peritoneal macrophages from EM patients (Fig. 6K), and 4-OI treatment similarly suppressed Mcoln2 expression in PBMCs, peritoneal macrophages from mice and BMDMs (Fig. 6L; Supplemental Fig. 5G–H). Importantly, the decline in intracellular Ca²⁺ induced by 4-OI was prevented by either CQ or the MCOLN2 agonist ML-SA1 (Fig. 6M and N; Supplemental Fig. 5I). Functionally, ML-SA1 treatment abolished the inhibitory effects of 4-OI on macrophage activation, restoring the macrophage pro-inflammatory phenotype (Fig. 6O and P). At the signaling level, 4-OI suppressed activation of p38 MAPK, a critical regulator of macrophage pro-inflammatory responses [45,46] that is modulated by intracellular Ca²⁺ [47,48], in LPS-stimulated BMDMs (Fig. 6Q). This suppression was reversed by CQ or ML-SA1 treatment (Fig. 6R and S). Together, these findings reveal a mechanism by which itaconate promotes lysosomal acidification, downregulates Mcoln2 expression, and reduces lysosomal Ca²⁺ release, thereby inhibiting p38 MAPK signaling and restraining pro-inflammatory macrophage phenotype.

3.7. Itaconate suppresses NOX2/ROS to enhance lysosomal acidification and inhibit macrophage inflammation

The maintenance of acidic lysosomal pH depends primarily on proton pumping mediated by the vacuolar-type H⁺-ATPase (V-ATPase) -mediated proton pumping and counterbalanced by superoxide production from NADPH oxidase 2 (NOX2) [49,50]. While V-ATPase expression was unaffected, 4-OI markedly downregulated gp91^phox mRNA (the main subunit of NOX2) in PBMCs, BMDMs, and peritoneal macrophages from EM patients and mice (Fig. 7A and B; Supplemental Fig. 6A–B). NOX2 protein remained unchanged at 12 h but decreased at 48 h following 4-OI stimulation (Fig. 7C and D; Supplemental Fig. 6C–D), while enzyme activity was already reduced at 12 h(Fig. 7E), suggesting early post-translational regulation. Similar effects were observed with itaconate stimulation (Supplemental Fig. 6E–F). Flow cytometric analysis using CellROX Green demonstrated that 4-OI decreased ROS production in pro-inflammatory BMDMs (Fig. 7F and G), similar to the NADPH oxidase inhibitor diphenyleneiodonium (DPI) (Supplemental Fig. 6G–H). DPI also decreased lysosomal pH (Supplemental Fig. 6I), suppressed the pro-inflammatory phenotype (Fig. 7H and I; Supplemental Fig. 6J), and decreased intracellular Ca²⁺ (Fig. 7J), mirroring 4-OI effects. These findings were further validated in vivo. Intraperitoneal administration of 4-OI or DPI to EM mice (Fig. 7K) significantly increased lesion weight compared with controls, with no significant difference between the two treatments (Fig. 7L–N). Similarly, the proportion of pro-inflammatory macrophage subsets (Fig. 7O and P; Supplemental Fig. 6K) and pro-inflammatory phenotype (Fig. 7Q) were comparably decreased by both 4-OI and DPI. Collectively, these results indicate that 4-OI inhibits NOX2-derived ROS, thereby enhancing lysosomal acidification and suppressing pro-inflammatory macrophage phenotype.

Fig. 7.

Itaconate Suppresses NOX2-Derived ROS to Regulate Macrophage Function and Lesion Progression in Endometriosis (A-B) qPCR analysis of NOX2 mRNA in peritoneal macrophages (PMs) from human (A) and mouse (B) NC and EM groups (n = 3/group). (C-D) Western blot and quantification of NOX2 protein in PBMC-derived macrophages treated with LPS or LPS + 4-OI (n = 5/group). (E) NOX2 enzyme activity in BMDMs (n = 4/group). (F, G) Flow cytometry and quantification of ROS production in BMDMs after LPS or LPS + 4-OI (n = 3/group). (H, I) Flow cytometry and quantification of iNOS⁺ BMDMs after LPS, LPS + 4-OI, or LPS + 4-OI + DPI treatment (n = 4/group). (J) Intracellular Ca²⁺ dynamics in BMDMs measured by Fluo-4 after LPS, 4-OI, or DPI treatment. (K) Schematic of 4-OI and DPI intervention in the mouse endometriosis model. (L, M) Representative images (L) and gross morphology (M) of endometriotic lesions after PBS, 4-OI, or DPI treatment. (N) Quantification of lesion weight (n = 6/group). (O, P) Flow cytometry and quantification of iNOS⁺ peritoneal macrophages (percentage and MFI) in peritoneal lavage (n = 6/group). (Q) mRNA levels of Il1b, Il6, Nos2, and Tnf in peritoneal macrophages after treatments (n = 6/group). (Data are presented as mean ± SEM. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001, ns: not significant.)

E2, estradiol benzoate; EM, endometriosis; NC, non-EM control; LPS, lipopolysaccharide; 4-OI, 4-octyl itaconate; DPI, NOX2 inhibitor.

4. Discussion

Immune microenvironment dysregulation is central to the pathogenesis of endometriosis (EM), with macrophages playing a pivotal regulatory role [51]. Although reported macrophage proportions in EM vary across studies, there is consistent evidence that peritoneal macrophages exhibit a restrained pro-inflammatory and clearance phenotype, which compromises the elimination of ectopic endometrial stromal cells (ESCs). [52,53]. However, the molecular and metabolic mechanisms underlying this immunosuppressive state have remained incompletely defined. In this study, we identify a previously unrecognized metabolic–immune regulatory circuit in EM, which we term the “ESC-ACOD1-itaconate–macrophage” axis. Through integrated metabolomic analyses and functional assays, we demonstrate that ectopic ESCs, rather than immune cells, exhibit elevated ACOD1 expression and represent the predominant source of itaconate within EM lesions. Importantly, ACOD1 expression was not increased in macrophages, challenging the prevailing assumption that itaconate accumulation in inflammatory settings is primarily macrophage-derived and instead positioning ESCs as active metabolic regulators of the immune microenvironment.

Untargeted metabolomic profiling revealed that, aside from the shared elevation of ACOD1/itaconate, human EM lesions and the murine implantation model display limited pathway-level concordance. This observation is consistent with known constraints of the murine EM model and underscores its contextual nature [34,54]. Accordingly, we do not interpret the mouse metabolome as a surrogate for the full human metabolic network. Rather, we restrict mechanistic conclusions to the conserved itaconate-centered signaling modules observed in both systems-most notably the NRF2 and NOX2-related pathways-while avoiding extrapolation of broader pathway differences to human disease. Functionally, ESC-derived itaconate is taken up by peritoneal macrophages, where it suppresses pro-inflammatory cytokine production and impairs phagocytic clearance capacity, thereby promoting ESC persistence and dissemination. Both exogenous itaconate and its derivative 4-octyl itaconate (4-OI) recapitulated these effects in vitro and in vivo, inducing an immunosuppressive and reparative macrophage phenotype. Conversely, inhibition of ACOD1 restored macrophage inflammatory and phagocytic functions and reduced lesion burden, whereas exogenous 4-OI exacerbated disease progression. Together, these findings establish ESC-derived itaconate as a central driver of macrophage-mediated immune dysregulation in EM.

Mechanistically, we delineate how itaconate regulates macrophage function through interconnected redox, lysosomal, and signaling pathways. Previous studies have largely emphasized itaconate-mediated inhibition of succinate dehydrogenase, activation of NRF2, and suppression of mitochondrial ROS and NLRP3 inflammasome activity. [55]. In parallel, succinate accumulation and SUCNR1 signaling have been implicated in EM-associated macrophage activation [56], consistent with our metabolomic observations. Within this metabolic context, we show that 4-OI robustly activates NRF2 signaling in macrophages, inducing antioxidant gene expression and dampening inflammatory responses. Importantly, NRF2 inhibition or knockdown attenuated-but did not abolish-these effects, supporting NRF2 involvement without assigning it exclusive causality.

Beyond NRF2, our data reveal a previously underappreciated role of the NOX2–lysosome-Ca²⁺ axis in itaconate-mediated immune regulation. Itaconate rapidly inhibits NOX2 activity, preceding detectable changes in protein expression, and molecular interaction analyses suggest direct interference with NOX2 complex assembly via the p47^phox subunit (Supplemental Fig. 6L–M). Reduced NOX2-derived ROS enhances lysosomal acidity, a critical determinant of lysosomal function that has received less attention than TFEB-mediated biogenesis in prior studies [23]. Lysosomal acidification, in turn, suppresses MCOLN2-mediated Ca²⁺ release, leading to reduced activation of the stress-responsive p38 MAPK pathway. Pharmacological reopening of lysosomal Ca²⁺ channels with ML-SA1 restored Ca²⁺ mobilization and p38 phosphorylation, supporting a NOX2/lysosome/Ca²⁺/p38 signaling route. Importantly, we interpret these pathways as functionally coupled rather than independent. NRF2 and p38 are known to engage in bidirectional crosstalk [57,58], and in our system, partial NRF2 inhibition leaves residual anti-inflammatory activity that may reflect incomplete blockade rather than a distinct NRF2-independent mechanism. Accordingly, we frame NRF2 activation and NOX2/Ca²⁺/p38 modulation as interacting arms within a unified signaling network, rather than as separate or parallel pathways. To interrogate these mechanisms experimentally, we employed 4-OI as a delivery-enhanced tool compound, while explicitly acknowledging its chemical and kinetic differences from itaconate [59]. 4-OI rapidly activated NRF2, whereas itaconate required prolonged exposure and optimized buffering conditions to elicit comparable directional effects on lysosomal acidity and NOX2-dependent ROS. These differences likely reflect delivery and electrophilicity rather than divergent biology. Importantly, LC-MS analysis of patient-derived ESC-conditioned media revealed cumulative itaconate concentrations sufficient to exert functional effects on macrophages, supporting the physiological relevance of our findings.

Regarding ACOD1 regulation in ESCs, canonical inducers such as LPS and IFN-γ did not upregulate ACOD1 in normal ESCs, indicating that alternative regulatory mechanisms operate in ectopic lesions. EM exhibits tumor-like metabolic reprogramming [60], potentially driven by hypoxia, inflammatory cytokines [61], iron overload, and oxidative stress-factors [62] that may converge on ACOD1-itaconate metabolism through NRF2-and NOX2-related pathways. Elucidating these upstream regulatory networks represents an important direction for future investigation. From a translational perspective, ACOD1 inhibition emerges as a promising therapeutic strategy. While small-molecule ACOD1 inhibitors remain in early developmental stages, genetic targeting approaches have shown efficacy in reducing lesion burden. Building on this concept, we demonstrate that LNP-mediated delivery of siIrg1 restores macrophage inflammatory and clearance functions and suppresses EM progression in vivo, offering a potentially more precise and safer intervention. Further development of ESC-or lesion-targeted delivery systems may enhance therapeutic specificity and clinical feasibility.

In conclusion, we establish that ESC-derived itaconate orchestrates an integrated network involving NRF2 activation, NOX2/ROS suppression, lysosomal acidification, and Ca²⁺-dependent p38 signaling to restrain macrophage pro-inflammatory and clearance functions. This metabolic-immune axis drives immune microenvironment remodeling and disease progression in EM, highlighting ACOD1-itaconate signaling as a compelling target for precision therapy. More broadly, our findings provide a conceptual framework for understanding metabolic-immune interactions in chronic inflammatory diseases.

CRediT authorship contribution statement

Zhaoyang Zhong: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. Shuang Wang: Conceptualization, Formal analysis, Investigation, Methodology, Software, Supervision. Qianhui Ren: Methodology, Resources. Xue Jiao: Software, Supervision. Le Xu: Investigation. Xiaoyu Dong: Data curation. Na Li: Investigation. Hongwei Guan: Methodology. Ran Chu: Investigation, Methodology, Software, Supervision. Ming Yuan: Conceptualization, Data curation. Jincheng Liu: Conceptualization, Funding acquisition. Yanbo Du: Conceptualization. Keke Wei: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration. Lei Yan: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources. Guoyun Wang: Conceptualization.

Declaration of competing interest

The authors have declared that no conflict of interest exists.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.