PARP-1 couples β-catenin/TCF4 signaling to epithelial-mesenchymal transition in endometriosis

Human studies were approved by the Institutional Review Board of Beijing Obstetrics and Gynecology Hospital, Capital Medical University, and written informed consent was obtained from all participants. All procedures complied with the Declaration of Helsinki. Animal experiments were approved by the institutional Animal Welfare and Ethics Committee and conducted in accordance with ARRIVE guidelines. Endometrial tissues were collected at Beijing Obstetrics and Gynecology Hospital (Capital Medical University) from women with ovarian endometriosis undergoing laparoscopic ovarian endometrioma cystectomy, from whom paired eutopic endometrium (EE) and ovarian endometriotic lesions (OEL) were obtained, and from women without endometriosis undergoing laparoscopic myomectomy with concomitant hysteroscopic endometrial biopsy, providing control endometrium (CE). All specimens were obtained in the proliferative phase. Overall clinical characteristics are summarized in Supplementary Tables S2 – S3 (total n  = 18 donors: 9 controls and 9 endometriosis patients, yielding 9 CE, 9 EE, and 9 OEL specimens). For each assay (IHC, Western blot, and RT–qPCR), a subset of samples from three independent donors per group (CE, EE, and OEL; n  = 3 each) was analyzed as biological replicates (selected based on tissue availability and assay requirements). Fresh samples were snap-frozen in liquid nitrogen for protein/RNA extraction or fixed in 10% neutral buffered formalin for immunohistochemistry. Total RNA was extracted for cDNA synthesis and quantitative PCR of PARP1 and EMT-related genes (E-cadherin, N-cadherin, vimentin, β-catenin, TCF4), normalized to GAPDH. Protein lysates were separated, transferred to membranes, and probed with primary antibodies: anti–PARP-1 (rabbit monoclonal, Abcam, ab191217; 1:1000 dilution), anti–E-cadherin (rabbit polyclonal, Proteintech, 20874-1-AP; 1:10000), anti–N-cadherin (rabbit polyclonal, Proteintech, 22018-1-AP; 1:5000), anti–vimentin (rabbit polyclonal, ABclonal, A19607; 1:10000), anti–β-catenin (rabbit polyclonal, BIOSS, bs-1165R; 1:2000), anti–TCF4 (rabbit monoclonal, Abcam, ab217668; 1:5000), and anti–β-actin (mouse monoclonal, Proteintech, 66009-1-Ig; 1:10000). HRP-conjugated secondary antibodies (1:50000) were applied, and signals detected by enhanced chemiluminescence. Section (4-µm) were deparaffinized, subjected to antigen retrieval in citrate buffer (pH 6.0), blocked with 3% H₂O₂, and incubated overnight with primary antibodies (1:100-1:2000). Detection employed biotinylated secondary antibodies, streptavidin–HRP, and DAB chromogen; slides were counterstained with hematoxylin. Negative controls omitted the primary antibody. Staining intensity and localization in epithelial and stromal compartments were scored by a blinded observer. 12Z human epithelial endometriotic cells (12Z cells; ZOMANBIO; an immortalized human endometriotic epithelial cell line) were maintained in DMEM/F12 supplemented with 10% FBS and 1% penicillin-streptomycin at 37 °C in 5% CO₂. PARP-1 knockdown was achieved by transfecting 50 nM PARP1-specific siRNA or control siRNA (Obio Technology) using Lipofectamine RNAiMAX (Invitrogen). For overexpression, cells were transfected with pcDNA3-PARP1 or empty vector using Lipofectamine 3000 (Thermo Fisher). Forty-eight hours post-transfection, PARP1 mRNA and protein levels were confirmed by qPCR and Western blot. To inhibit Wnt/β-catenin signaling, cells were treated with JW74 (5 µM; MedChemExpress) for 24 h, alone or combined with PARP-1 overexpression. Protein expression and transcript levels were assessed by Western blot using the antibodies described above, and quantitative PCR assays. Cell viability under each condition was measured by CCK-8 assay. Scratch wound-healing assays were performed on confluent 12Z cell monolayers scratched with a pipette tip; wound closure was imaged at 0 h and 48 h, and percent closure calculated. For Transwell assays, 1 × 10⁵ cells in serum-free medium were seeded into the upper chamber of 8-µm pore inserts, with 10% FBS in the lower chamber; after 24 h, migrated cells were fixed in paraformaldehyde, stained with crystal violet, and counted in five random fields. All assays were performed in triplicate. Nuclear extracts (500 µg) from 12Z cell were prepared, incubated with anti–PARP-1 antibody, and immune complexes captured with Protein A/G agarose beads. After extensive washing, bound proteins were eluted in SDS sample buffer, separated by electrophoresis, and probed by Western blot for β-catenin, TCF4, or PARP-1. Normal IgG IPs served as negative controls. All procedures were approved by the Animal Welfare and Ethics Committee of Beijing Obstetrics and Gynecology Hospital, Capital Medical University (approval BOGH21-2505-1) and conducted in the hospital’s SPF facility (license SYXK [Beijing] 2022-0010). Female BALB/c mice (SPF, 7–8 weeks, ~ 20 g) were procured from Beijing Viton Lihua Experimental Animal Technology Co., Ltd. (production license SCXK [Beijing] 2021-0006) and acclimated for 7 days. Anaesthesia was induced with sodium pentobarbital 45 mg/kg, intraperitoneal (i.p.); depth was verified by loss of pedal withdrawal and diminished corneal reflexes. Euthanasia was performed under deep anaesthesia by cervical dislocation using an institutional small-animal device in accordance with IACUC/AVMA guidelines and institutional SOPs. Endometriosis-like lesions were established by syngeneic intraperitoneal implantation of uterine endometrial tissue fragments from donor BALB/c females ( n  = 10). Donor uteri were minced into ~ 1 mm³ fragments, suspended in sterile PBS, and implanted into recipient BALB/c females via intraperitoneal injection under anesthesia. Recipients were randomly assigned to sham (laparotomy without implantation; n  = 3), vehicle ( n  = 5), or olaparib ( n  = 5) groups (total n  = 23). Two mice in the vehicle group died prior to treatment initiation (day 14 post-implantation) due to peri-procedural/anaesthesia-related complications and were excluded from efficacy analyses according to pre-specified criteria. Therefore, the final numbers analyzed were sham n  = 3, vehicle n  = 3, and olaparib n  = 5. Outcome assessors were blinded to group allocation. From day 14 to day 28 after implantation, mice in the treatment group received olaparib at 50 mg/kg intraperitoneally once daily 25 , while vehicle controls received an equal volume of vehicle on the same schedule. Olaparib (MedChemExpress, HY-10162) was prepared fresh daily. On day 28, ectopic lesions were identified at necropsy, enumerated, and measured using digital calipers (long axis, a; short axis, b). Lesion volume was calculated as V = a × b² × 0.52, and total lesion volume per mouse was defined as the sum of all measured lesions; animals without visible lesions were recorded as 0. As a complementary gross endpoint, the uterus mass with adherent ectopic lesions (without dissecting lesions from surrounding tissues) was recorded for each animal. Because lesions were frequently adherent and could not be cleanly separated from surrounding tissues, lesion-adherent uterine tissue (rather than “lesion-only” tissue) was collected for protein extraction and Western blot analysis of PARP-1, E-cadherin, N-cadherin, Vimentin, β-catenin, and TCF4. For the sham group, uterine tissue collected at day 28 served as the reference tissue for immunoblotting. Data are presented as mean ± SEM unless otherwise indicated. Statistical analyses were performed using SPSS (version 26.0). For comparisons among ≥ 3 groups, one-way ANOVA followed by Tukey’s post hoc test was applied when data met assumptions of normality and homogeneity of variance. When distributions were non-normal (e.g., skewed/zero-inflated lesion-burden data), nonparametric tests were used and data are presented as median (IQR); specifically, the primary in vivo efficacy comparison between the EMS+vehicle and EMS+olaparib groups was evaluated using a two-tailed exact Mann–Whitney U test. A two-sided p value < 0.05 was considered statistically significant.

Western blot and qPCR analyses showed significant variations in PARP-1 and EMT marker levels among normal endometrium (CE), eutopic endometrium from endometriosis patients (EE), and ovarian endometriotic lesions (OEL). PARP-1 expression was lower in CE and EE, and highest in OEL ( p  < 0.05) (Fig.  1 b-d). EMT markers mirrored this pattern: E-cadherin significantly decreased in OEL ( p  < 0.001), while N-cadherin and vimentin notably increased ( p  < 0.05) (Fig.  1 b-d). EE exhibited intermediate EMT alterations. Additionally, total β-catenin and TCF4, critical Wnt pathway components, were elevated in OEL compared to CE ( p  < 0.05) (Fig.  1 b-d). Immunohistochemistry confirmed these molecular trends, showing strong nuclear PARP-1 staining, reduced E-cadherin, and enhanced β-catenin and TCF4 staining mainly in OEL tissues, with moderate changes in EE (Fig.  1 a). Fig. 1 PARP-1 aligns with epithelial–mesenchymal transition (EMT)-marker remodeling in endometriosis tissues. a Representative immunohistochemistry for poly(ADP-ribose) polymerase-1 (PARP-1), N-cadherin, vimentin, E-cadherin, β-catenin, and T-cell factor 4 (TCF4) in normal endometrium (CE), eutopic endometrium from women with endometriosis (EE), and ovarian endometriotic lesions (OEL). Stronger nuclear/epithelial PARP-1 staining, reduced E-cadherin, and enhanced β-catenin/TCF4 are evident in OEL; EE shows intermediate changes. Scale bars, 50 μm. n  = 3 donors per group. b Immunoblots of the same markers in CE, EE, and OEL (GAPDH loading control). n  = 3 donors per group. c Densitometric quantification normalized to GAPDH and expressed relative to CE. d Relative mRNA expression (qPCR; fold-change vs. CE). n  = 3 donors per group. Bars show mean ± SEM. One-way ANOVA with Tukey’s post hoc test (two-sided): * p  < 0.05, ** p  < 0.01, *** p  < 0.001. IHC scoring was performed blinded to group. Uncropped blots are provided. PARP-1 aligns with epithelial–mesenchymal transition (EMT)-marker remodeling in endometriosis tissues. a Representative immunohistochemistry for poly(ADP-ribose) polymerase-1 (PARP-1), N-cadherin, vimentin, E-cadherin, β-catenin, and T-cell factor 4 (TCF4) in normal endometrium (CE), eutopic endometrium from women with endometriosis (EE), and ovarian endometriotic lesions (OEL). Stronger nuclear/epithelial PARP-1 staining, reduced E-cadherin, and enhanced β-catenin/TCF4 are evident in OEL; EE shows intermediate changes. Scale bars, 50 μm. n  = 3 donors per group. b Immunoblots of the same markers in CE, EE, and OEL (GAPDH loading control). n  = 3 donors per group. c Densitometric quantification normalized to GAPDH and expressed relative to CE. d Relative mRNA expression (qPCR; fold-change vs. CE). n  = 3 donors per group. Bars show mean ± SEM. One-way ANOVA with Tukey’s post hoc test (two-sided): * p  < 0.05, ** p  < 0.01, *** p  < 0.001. IHC scoring was performed blinded to group. Uncropped blots are provided. In 12Z cells, PARP-1 knockdown increased E-cadherin and reduced mesenchymal markers ( p  < 0.05), accompanying a more epithelial morphology (Fig.  2 a, c,d). Conversely, PARP-1 overexpression decreased E-cadherin and increased mesenchymal markers ( p  < 0.05) (Fig.  2 a, c,d); β-catenin redistribution consistent with pathway engagement was observed. The Wnt inhibitor JW74 partially attenuated PARP-1-overexpression effects, with partial restoration of epithelial markers and modest reductions of mesenchymal markers (Fig.  2 a, c,d). Functionally, PARP-1 knockdown reduced viability (CCK-8, p  < 0.001), wound closure ( p  < 0.001), and Transwell migration/invasion ( p  < 0.001), whereas overexpression showed the opposite pattern (viability p  < 0.001; migration/invasion p  < 0.01). JW74 mitigated these phenotypes (Fig.  3 a–d). Fig. 2 PARP-1 modulation associates with EMT features and couples to β-catenin/TCF4 in 12Z cells. a Immunoblot analysis of PARP-1 and EMT/Wnt markers in 12Z cells under the indicated conditions [control, empty vector (EV), PARP-1 overexpression (PARP-1-OE), PARP-1 knockdown (siRNA-PARP-1), and Wnt inhibition (JW74)]. GAPDH served as loading control. b Co-immunoprecipitation with anti-PARP-1 detecting association with β-catenin and TCF4; IgG and input controls shown. c Densitometric quantification of immunoblots normalized to GAPDH and expressed relative to EV. d Relative mRNA expression by qPCR (fold-change vs. EV). Data are mean ± SEM from ≥ 3 independent experiments. One-way ANOVA with Tukey’s post hoc test (two-sided): * p  < 0.05, ** p  < 0.01, *** p  < 0.001. Uncropped blots are provided. PARP-1 modulation associates with EMT features and couples to β-catenin/TCF4 in 12Z cells. a Immunoblot analysis of PARP-1 and EMT/Wnt markers in 12Z cells under the indicated conditions [control, empty vector (EV), PARP-1 overexpression (PARP-1-OE), PARP-1 knockdown (siRNA-PARP-1), and Wnt inhibition (JW74)]. GAPDH served as loading control. b Co-immunoprecipitation with anti-PARP-1 detecting association with β-catenin and TCF4; IgG and input controls shown. c Densitometric quantification of immunoblots normalized to GAPDH and expressed relative to EV. d Relative mRNA expression by qPCR (fold-change vs. EV). Data are mean ± SEM from ≥ 3 independent experiments. One-way ANOVA with Tukey’s post hoc test (two-sided): * p  < 0.05, ** p  < 0.01, *** p  < 0.001. Uncropped blots are provided. Fig. 3 PARP-1 influences motility and viability in 12Z cells. a Representative images of scratch-wound closure (0 h, 48 h) and Transwell migration (24 h) under the indicated conditions: control, empty vector (EV), PARP-1 overexpression (PARP-1-OE), PARP-1 knockdown (siRNA-PARP-1), and Wnt inhibition (JW74). Scale bars, 100 μm. b Quantification of wound-closure rate (% of initial gap). c Quantification of migrated cells per field (Transwell). d Cell viability (CCK-8; % of control). Bars show mean ± SEM from ≥ 3 independent experiments (technical duplicates unless stated). One-way ANOVA with Tukey’s post hoc test (two-sided): * p  < 0.05, ** p  < 0.01, *** p  < 0.001. PARP-1 influences motility and viability in 12Z cells. a Representative images of scratch-wound closure (0 h, 48 h) and Transwell migration (24 h) under the indicated conditions: control, empty vector (EV), PARP-1 overexpression (PARP-1-OE), PARP-1 knockdown (siRNA-PARP-1), and Wnt inhibition (JW74). Scale bars, 100 μm. b Quantification of wound-closure rate (% of initial gap). c Quantification of migrated cells per field (Transwell). d Cell viability (CCK-8; % of control). Bars show mean ± SEM from ≥ 3 independent experiments (technical duplicates unless stated). One-way ANOVA with Tukey’s post hoc test (two-sided): * p  < 0.05, ** p  < 0.01, *** p  < 0.001. Co-immunoprecipitation showed that PARP-1 co-precipitated with β-catenin and TCF4 in 12Z cells (Fig.  2 b), indicating an association within β-catenin/TCF4-containing immunocomplexes. This finding supports complex-level proximity, but does not by itself establish that PARP-1 catalytic activity or direct PARylation is required for β-catenin/TCF4-dependent transcription. In the mouse endometriosis model, treatment with olaparib (a PARP inhibitor) reduced ectopic lesion burden compared with vehicle. Lesion volume per mouse (undetectable lesions scored as 0) was decreased in the EMS+olaparib group (median [IQR] 0 [0–0.65] mm³; n  = 5) relative to the EMS+vehicle group (96.89 [12.28–297.27] mm³; n  = 3; Mann–Whitney U test, two-tailed exact p  = 0.0357) (Fig.  4 a–b). As a complementary gross endpoint, the mass of the uterus with adherent ectopic lesions was also reduced by olaparib (142.3 [132.4–145.6] mg vs. 324.8 [244.0–365.5] mg; p  = 0.0357) (Fig.  4 c). A complete set of macroscopic images (total N  = 11) and individual animal-level measurements are provided in Supplementary Figure S1 and Supplementary Table S1 . Lesion-adherent uterine tissues from olaparib-treated mice exhibited higher E-cadherin and lower N-cadherin, vimentin, β-catenin, and TCF4 compared with vehicle (Fig.  4 d–e), consistent with attenuation of EMT-aligned remodeling in vivo. Fig. 4 PARP inhibition reduces ectopic lesion burden and attenuates EMT/Wnt-aligned remodeling in a mouse endometriosis model. a Representative gross morphology of the uterus with adherent ectopic lesions in sham, EMS+vehicle, and EMS+olaparib groups. Dashed outlines and arrowheads indicate ectopic lesions. Scale bars, 5 mm. b Total ectopic lesion volume per mouse (sum of all lesions; undetectable lesions were scored as 0). c Uterus mass with adherent ectopic lesions (lesions were not dissected; therefore, this represents gross uterine mass with adherent lesions). In b-c, each dot represents one mouse; boxes indicate the interquartile range (IQR), center line indicates the median, and whiskers indicate min–max. Statistical comparison between EMS+vehicle ( n  = 3) and EMS+olaparib ( n  = 5) was performed using a two-tailed exact Mann–Whitney U test: lesion volume, p  = 0.0357; uterus mass with adherent ectopic lesions, p  = 0.0357. Sham mice ( n  = 3) are shown as a procedural reference. d Representative immunoblots of PARP-1 and indicated EMT/Wnt markers in sham uterine tissue (reference) and EMS lesion-adherent uterine tissue from vehicle- or olaparib-treated mice. e Densitometric quantification normalized to GAPDH ( n  = 3 biological replicates per group). P  < 0.05, P  < 0.01 (as indicated). PARP inhibition reduces ectopic lesion burden and attenuates EMT/Wnt-aligned remodeling in a mouse endometriosis model. a Representative gross morphology of the uterus with adherent ectopic lesions in sham, EMS+vehicle, and EMS+olaparib groups. Dashed outlines and arrowheads indicate ectopic lesions. Scale bars, 5 mm. b Total ectopic lesion volume per mouse (sum of all lesions; undetectable lesions were scored as 0). c Uterus mass with adherent ectopic lesions (lesions were not dissected; therefore, this represents gross uterine mass with adherent lesions). In b-c, each dot represents one mouse; boxes indicate the interquartile range (IQR), center line indicates the median, and whiskers indicate min–max. Statistical comparison between EMS+vehicle ( n  = 3) and EMS+olaparib ( n  = 5) was performed using a two-tailed exact Mann–Whitney U test: lesion volume, p  = 0.0357; uterus mass with adherent ectopic lesions, p  = 0.0357. Sham mice ( n  = 3) are shown as a procedural reference. d Representative immunoblots of PARP-1 and indicated EMT/Wnt markers in sham uterine tissue (reference) and EMS lesion-adherent uterine tissue from vehicle- or olaparib-treated mice. e Densitometric quantification normalized to GAPDH ( n  = 3 biological replicates per group). P  < 0.05, P  < 0.01 (as indicated). Across patient tissues, an endometriotic epithelial cell model, and an in vivo lesion model, PARP-1 abundance aligned with β-catenin/TCF4-linked epithelial–mesenchymal transition (EMT) features, and these features were modulated by genetic perturbation and attenuated by pharmacologic PARP inhibition.

Across patient tissues, an endometriotic epithelial cell model, and an in vivo lesion model, we observed concordant alignment of PARP-1 abundance with β-catenin/TCF4-linked epithelial–mesenchymal transition (EMT) features, and these features were pharmacologically attenuable. In ovarian endometriotic lesions, PARP-1 was increased relative to control endometrium and accompanied by reduced E-cadherin and increased N-cadherin/vimentin. In 12Z cells, PARP-1 gain and loss produced reciprocal remodeling of EMT markers and motility, and co-immunoprecipitation detected β-catenin and TCF4 in PARP-1 immunocomplexes, indicating association at the complex level. In mice, olaparib treatment was associated with reduced lesion burden and partial normalization of EMT markers. Taken together, these observations support a model in which PARP-1 functions as a druggable interface that couples the Wnt/β-catenin transcriptional apparatus to EMT-aligned phenotypes, rather than as a singular upstream driver; however, our data do not establish that PARP-1 catalytic activity or direct PARylation of pathway components is required for the PARP-1–β-catenin/TCF4 association or downstream EMT regulation. This interpretation is consistent with contemporary views of PARP-1 as a chromatin-associated regulator that intersects transcriptional programs beyond canonical roles in DNA damage responses 15 . The nuclear enrichment of PARP-1 we observed in lesions accords with its chromatin-proximal functions and echoes reports nominating PARP-1-together with EpCAM and folate receptor-α-as an intraoperative imaging biomarker for endometriotic tissue 14 . Co-variation of PARP-1 with β-catenin/TCF4 and mesenchymal markers is compatible with a partial EMT state, a framework increasingly invoked to explain lesion plasticity, fibrosis, and invasive behavior in endometriosis 16 . Within the inflammation-oxidative stress-fibrosis nexus, advanced oxidation protein products have been shown to activate EMT pathways, and non-human primate models document lesion progression accompanied by EMT and fibroblast-to-myofibroblast transition 17 , 18 . Given the reported roles of PARP-1 in NF-κB/TGF-β signaling and oxidative-stress responses, PARP-1 provides a plausible molecular conduit linking these axes to mesenchymal bias. Although implantation and decidualization were not directly interrogated here, disequilibrium of Wnt signaling and EMT is well recognized to perturb endometrial receptivity by disrupting epithelial-stromal crosstalk and suppressing factors such as IGFBP1 and prolactin 19 . In light of broader evidence implicating inflammation, endocrine disruption, and oxidative stress in reproductive dysfunction among women with endometriosis 20 , the PARP-1-β-catenin/TCF4-EMT interface described here may contribute to a less receptive endometrial milieu; this hypothesis warrants prospective evaluation using receptivity biomarkers and fertility endpoints. Therapeutically, our observation that olaparib reduced lesion volume and partially reversed EMT-aligned markers provides supportive pharmacologic evidence that PARP activity participates in these phenotypes 21 , nonetheless, this should be interpreted as supportive rather than definitive proof of catalytic dependence. Experience from oncology suggests rational combinations-e.g., PARP inhibition with Wnt-pathway modulators or immune-directed agents-may enhance efficacy 22 , 23 . Convergence with inflammatory pathways is also compelling: blockade of the IL-33–ST2 axis mitigates β-catenin-driven EMT in endometriosis models, suggesting that co-targeting PARP-1 and inflammatory signaling merits investigation 24 . Together, these data advocate for biomarker-guided, mechanism-anchored trials using EMT/Wnt readouts for stratification and pharmacodynamic assessment. This study has limitations. The in vivo experiment was a pilot with limited sample size due to attrition, and lesion volume was highly skewed/zero-inflated; therefore, effect estimates should be interpreted cautiously and warrant confirmation in larger cohorts. Our in vivo treatment window was short-term; therefore, we did not assess durability after prolonged exposure or drug withdrawal, potential adaptive escape/recurrence, or longer-term effects on normal tissues. In addition, we did not perform dose–response/minimal-effective-dose optimization or comprehensive safety profiling, which should be addressed in future longer-term studies. While we provide a complementary gross endpoint (uterus with adherent lesion mass), we did not perform a systematic toxicity assessment (hematology/biochemistry) or comprehensive reproductive-safety evaluation (estrous cycling, endometrial receptivity/decidualization markers, ovarian reserve). In vitro mechanistic experiments were performed in 12Z cells, an immortalized endometriotic epithelial cell line; thus, these findings may not fully capture the heterogeneity of primary endometriotic epithelial cells and warrant validation in primary cells or patient-derived models in future work. In human sampling, all specimens were obtained in the proliferative phase to reduce cycle-related heterogeneity, but this limits generalization to other cycle phases. Mechanistically, we did not directly interrogate PARP-1 catalytic mutants, measure PARylation of β-catenin/TCF components, employ β-catenin/TCF transcriptional reporters, or compare multiple PARP inhibitors with differing mechanisms; consequently, pathway engagement was inferred from complex association and marker shifts. In summary, our multi-layer evidence links PARP-1 to β-catenin/TCF4-associated EMT programs that are susceptible to pharmacological inhibition, nominating PARP-1 as a tractable mechanistic node with translational relevance in endometriosis. Prospective studies connecting PARP-targeted strategies to receptivity biomarkers and fertility outcomes-and embedding these strategies within Wnt- and inflammation-aware combinations-will be essential to define clinical impact.

Endometriosis affects up to 10% of women of reproductive age and is a major contributor to subfertility and reduced quality of life 1 , 2 . Although histologically benign, ectopic endometrial tissue displays traits of invasion, migration and implantation that can distort pelvic anatomy, sustain inflammation and impair endometrial function 3 , 4 . These clinical features, together with lesion recurrence and fibrosis, highlight the need to define tractable biological nodes that connect molecular programs to disease-relevant phenotypes. Epithelial–mesenchymal transition (EMT) has been proposed as a central program that enables epithelial cells to acquire mesenchymal features and motility, thereby facilitating lesion establishment and persistence at ectopic sites 5 , 6 . In eutopic endometrium from women with endometriosis, partial EMT has been linked to reduced epithelial integrity and impaired decidualization, processes that may diminish receptivity 7 , 8 . Thus, dysregulated EMT provides a unifying framework to interpret tissue plasticity in lesions and perturbed function in the uterus. The Wnt/β-catenin pathway, required for endometrial development and embryo–endometrium communication, is frequently dysregulated in endometriosis 9 , 10 . β-Catenin partnering with T-cell factor 4 (TCF4, also known as TCF7L2, a T-cell factor in the Wnt/β-catenin pathway) offers a transcriptional conduit through which upstream Wnt cues can align with EMT-related gene expression, suggesting this axis as a logical point for mechanistic inquiry. Poly(ADP-ribose) polymerase-1 (PARP-1) is best known for roles in DNA damage responses but also acts as a chromatin-associated regulator that interfaces with transcriptional complexe 11 . PARP-1 has been reported to augment EMT-related transcription through interactions with signaling mediators, and to modulate β-catenin/TCF4-dependent programs in Wnt-activated settings 12 , 13 . Increased PARP-1 expression has been described in endometriotic lesions 14 , yet whether PARP-1 coordinates β-catenin/TCF4 activity with EMT programs in endometriosis—and whether such features are attenuable by pharmacologic inhibition—remains unclear. Here, we test the hypothesis that PARP-1 couples β-catenin/TCF4 activity to EMT-aligned phenotypes in endometriosis. We integrate tissue profiling, cellular gain- and loss-of-function studies, complex association analyses, and an in vivo lesion model to determine alignment across contexts and to evaluate whether PARP inhibition can modulate these features.

Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2 Supplementary Material 3 Supplementary Material 3 Supplementary Material 4 Supplementary Material 4 Supplementary Material 5 Supplementary Material 5 Supplementary Material 6 Supplementary Material 6