Comprehensive transcript and protein profiling reveals ERα rather than ERβ as the predominant estrogen receptor in human endometriotic lesions

Endometriosis (EMS) is a chronic estrogen -dependent inflammatory disease. Although several studies have suggested a key role for estrogen receptor ERβ in EMS lesion development, its detection has been challenged by the lackof specificity of many ERβ antibo dies. To clarify the status of sex steroid receptors in the endometrium and matched EMS lesions , we perform RNAScope and immunohistochemistry on a tissue microarray cohort, mapping the expression of estrogen receptors ERα and ERβ, as well as progesterone and androgen receptors (PR and AR). We find that ER α is the predominant estrogen receptor in epithelial and stromal compartments across lesion types, including ovarian endometriomas. By contrast, ER β expression remains low and is mainly restricted to endoth elial cells. ER α expression is reduced in stromal cells across lesion types relative to matched endometrium and in the epithelium of superficial peritoneal lesions. In deep endometriosis lesions, reduced stromal ER α expression is associated with a signific ant increase in epithelial PR and AR expression, suggesting compartment-specific perturbations of ERα/PR signaling and broader remodeling of steroid hormone responses. Together, these findings identify ER α as the dominant but dysregulated estrogen receptor in EMS lesions and support steroid receptor profiling as a framework for lesion stratification and improved endometriosis diagnosis. ARTICLE IN PRESS ARTICLE IN PRESS

Endometriosis (EMS) is a chronic inflammatory gynecologic disease, whose growth, progression and persistence are strongly influenced by estradiol -dependent. It affects nearly 10% of reproductive -aged women, and globally impacting 190 million women 1-5,6. It is characterized by the presence of endometrial tissue (glands and stroma) outside the uterus, including the ovaries and other pelvic structures. This disease can be classified into three subtypes depending on its localization: ovarian endometriomas (OMA), superficial peritoneal lesions (SPE), and deep endometriosis (DE ), also referred to as deep infiltrating endometriosis (D IE), defined clinically as lesions that infiltrate >5 mm under the peritoneal surface). Women with EMS experience severe pelvic pain leading to a major impairment in the quality of life, and EMS represents a major cause of infertility. The heterogeneity in the disease presentation makes it difficult to diagnose, resulting in a diagnostic delay of 6 to 10 years1. The human endometrium is a highly dynamic tissue with a remarkable regenerative capacity, undergoing cyclic phases of proliferation, differentiation, and shedding under the influence of ovarian steroid hormones, particularly estrogen and progesterone7-10. These ovarian hormones exert their actions by binding to th eir respective estrogen receptors , namely estrogen receptor alpha (ERα), estrogen receptor beta ( ERβ) and the progesterone receptor (PR). In addition to the classical nuclear estrogen receptor, the G protein -coupled estrogen receptor (GPER, also known as GP R30) has been proposed to mediate rapid non -genomic estrogen signaling. However, accumulating structural and functional evidence11 indicates that GPER does not function as a classical estrogen receptor, supporting the central role of ERα and ERβ in mediating estrogen signaling in the endometrium. Current medical treatments for endometriosis such as oral contraceptive pills, gonadotropin -releasing hormone (GnRH) agonists, hormone-releasing intrauterine devices, and subdermal implants act by suppressing ovarian steroid production and/or modulating ho rmonal signaling, thereby reducing estrogenic stimulation of endometriotic lesions. However, these therapies do not eradicate the disease and instead induce a temporary, hormonally suppressed state, with symptoms frequently recurring after treatment discontinuation Given the hormo ne-dependent nature and marked heterogeneity of endometriosis, accurately defining the steroid receptor profile of endometrio sis lesions is essential for improving our understanding of disease pathophysiology and for the developmen t of more effective therapeutic strategies. In the human endometrium of healthy women, ERα is strongly expressed in both epithelial and stromal cells during the proliferative phase and mediates the proliferative effects of estradiol⁶. Proteins encoded by human ESR2 splice variants (ER β2, ER β5) have also been detected in human endometrium 12,13 and endometrial cells14, although their functional significance remains incompletely understood. In contrast, the expression patterns of ER α and ERβ receptors in endometriosis are variable and remain a matter of debate 15. While increased expression of both ESR1 (encoding ER α protein) and ESR2 (encoding ER β protein) mRNAs have been reported in eutopic and ectopic tissues from women with endometriosis 16, numerous studies suggest a relative imbalance characterized by downregulation of ESR1 transcripts and upregulation of ESR2 transcripts in endometriotic lesions 17-20. Notably, Matsuzaki and colleagues 17 demonstrated that cyclic ovarian hormone variations differentially regulate E SR1 and E SR2 mRNA expression in endometriotic tissue and highlighted distinct expression patterns between ovarian and peritoneal lesions. These observations suggest that steroid receptor expression in endometriosis may be influenced not only by menstrual cycle -dependent hormonal f luctuations but also by lesion subtype (OMA, SPE, DE) and exposure to exogenous hormonal treatments21. A marked increase in ESR2 transcript ARTICLE IN PRESS ARTICLE IN PRESS and ERβ protein levels have been particularly reported in ovarian lesions, especially within the stromal compartment22-25, and elevated ER β levels have also been described in primary stromal cell cultur es derived from women with endometriosis24. Furthermore, higher ESR2 expression in OMA compared with SPE and DE lesions has been associated with differential responses to suppressive hormonal therapies²⁰. Notably, a study published in Cell by Han et al. (2015)26 concluded that ERβ plays a central and driving role in the pathogenesis of endometriosis, promoting lesion survival and inflammatory signaling. This work, which has been extensively cited, has substantially shaped the prevailing view that ER β predominates over ERα in endometriotic tissue and represents a key therapeutic target. Given the broad impact of this

on the field, a rigorous reassessment of ER β expression using validated and highly specific methodological approaches appears part icularly warranted. While functional studies have suggested a role for ERβ in lesion biology, accurate determination of its expression pattern in human tissues remains essential to correctly interpret these findings. In this context, rigorous validation of ERβ detection methods is essential. Accurate assessment of ERβ protein expression has been challenged by evidence showing that many antibodies against ERβ lack specificity, with the notable exception of the monoclonal antibody PPZ0506 27, which has demonstrated specificity but has been rarely used in previous ly published studies. Consequently, , many findings regarding ERβ protein levels may warrant reassessment using validated detection methods, and th e reported imbalance between ERα and ERβ remains uncertain. To unambiguously assess steroid receptor expression in EMS, we performed a comprehensive analysis of ERα and ERβ expression at both RNA and protein levels in the endometrium of 49 women with or without endometriosis and compared eutopic endometrium with endometriotic lesions. We also analyzed the expression of PR (progesterone receptor) and AR (androgen receptor), given their key roles in menstrual cycle regulation. Using tissue-microarrays (TMAs), composed of matched normal endometrium and EM S lesions, we performed multi-level testing of ERα, ERβ, PR, Ki-67 and AR expression in a cohort of patients operated for endometriosis in comparison to healthy women. Our analysis combines RNA fluorescent in situ hybridization (RNA -FISH) and immunohistochemistry (IHC)on adjacent tissue sections, enabling parallel assess ment of mRNA and protein expression. This approach enabled the examination of matched patient samples, including eutopic endometrium and various types of endometriotic lesions, while accounting for the menstrual cycle phase and lesion heterogeneity (OMA, SPE and DE). In contrast to previous stud ies, we detected a high prevalence of ERα and a strikingly low to undetectable expression of ERβ across all lesion types, including ovarian endometriomas with ER β detectable only in endothelial cells of the endometrium. While ERα expression varies across the menstrual cycle, decreasing in late secretory phase, analysis of matched tissues demonstrated that both SPE and DE lesions exhibit reduced stromal ERα expression relative to the corresponding eutopic endo metrium and allowed direct comparison of ER transcript and protein expression within the same lesions . Strikingly, DE lesions exhibit the most pronounced differences, characterized by upregulated expression of progesterone (PR) and androgen receptors (AR), and were associated with decreased cellular proliferation. These findings highligh t the predominant expression of ERα in both epithelial and stromal compartments of EM S lesions, while also suggesting a complex disruption of estrogenic homeostasis between these cell types, accompanied by significant alterations in hormonal paracrine signaling pathways. ARTICLE IN PRESS ARTICLE IN PRESS

Predominant ESR1 expression with limited ESR2 transcript detection in endometrium and endometriotic lesions To accurately characterize and compare ERα and ERβ expression in endometriotic lesions at different locations (SPE, DE and OMA), biopsies were collected from women with (EM S) or without endometriosis (Healthy women, HW). These samples were used to generate TMAs, enabling the simultaneous analys is and comparison of expression across multiple tissue specimens on a single slide (Fig. 1 A and Supplementary Fig. 1A-H). ESR1 and ESR2 mRNA expression w as first assessed using RNA in situ hybridization (RNA-FISH) with multiplex fluorescent RNAscope probes, enabling single cell detection of both ESR1 (in green) and ESR2 (in purple) mRNAs simultaneously (Fig. 1B-D). Normal ovarian tissue containing an antral follicle was also included in two spots, as a positive control for ESR2 expression (Fig. 1B). In antral follicles, granulosa cells become involved in estradiol production, which is associated with strong ER α expression, while ER β remains detectable 28. This control further validated the reliable detection of both ESR1 and ESR2 transcripts. H-score quantification of mRNAs signals was then performed across all tissue samples at single cell resolution, separating expression in epithelial and stromal c ompartments and taking the menstrual phase ( proliferative and secretory phases) into account , as determined by PAEP and CPM immunofluorescence staining (Supplementary Fig. 1I). The maximum H-score of 300 corresponds to 100% of cells exhibiting more than 10 spots per cell or 20% of cells containing clusters whereas an H-score of 5 indicates that only 5% of cells exhibit a single spot (see Materials and Methods). High levels of ESR1 mRNA expression (shown in green) were observed across all endometrial tissue samples including both control (HW) and endometriotic (EM S) endometrium, as well as SPE and DE lesions. ESR1 mRNAs was also present in ovarian lesions, with a H-score of 132 for epithelium and 36 for stromal comp artment (Fig. 1 C). The cell distribution of mRNA confirms that ESR1 mRNA is the predominant transcript for ERs, expressed in both epithelial and stromal cells of the endometrium of healthy (HW) or endometriotic (EM S) women with similar levels of expression. Specifically, ESR1 expression had a very high H-score (close to the maximum possible score of 300) with values of 273.8±5.4 in epithelial cells of endometrium from healthy women (HW) versus 268.6±5.4 in those of endometriotic women (EMS); and 215.6±9 in stromal cells from HW versus 195.5±14 in EMS. Similarly, ESR1 mRNA was highly expressed in endometriotic lesions, with H-scores of 236.4±20.5 in epithelial cells of SPE lesions and 236.7±20.7 in those of DE lesion during the proliferative phase (Fig 1E-F). Consistently, ESR1 expression was significantly reduced throughout the menstrual cycle, with a 30-35% decrease in the secretory phase compared with the proliferative phase in the endometrium. No s ignificant differences were observed between among endometrium, HW or EMS tissues, nor between EMS endometrium and endometriotic lesions. However, significant differences between the proliferative and secretory p hases were observed for all tissue types. In contrast, ESR2 mRNA expression (in purple) was barely detectable in both endometrial tissues and lesions (Fig. 1C-D), with only a few isolated spots observed in epithelial and stromal cells across all tissues examined. The specificity of the ESR2 probe , however, was validated by s trong signal in the positive control (ovarian tissues) where granulosa cells exhibited pronounced ESR2 expression (H -score = 111 in ovarian follicles, Fig. 1 B). As expected, ESR2 expr ession in the endometrium was low, with H -scores of 7.6 ± 3 in epithelial cells of healthy women and 4.8 ± 0.8 in those of women with endometriosis. Similarly, ARTICLE IN PRESS ARTICLE IN PRESS stromal cells showed low ESR2 expression, with H -scores of 11.1 ± 1 in HW and 5.9 ± 1.5 in EMS patients during the proliferative phase. U nexpectedly, ESR2 was notably absent in the ovarian lesions analyzed, showing negligible ESR2 expression (H -score = 0 -3, Fig. 1 C). Notably, the rare ESR2-positive signals detected in the endometrial stroma were freq uently distributed along aligned cells, particularly during the secretory phase, consistent with a morphology suggestive of endothelial cell s. This was further supported by ERG immunostaining (a nuclear endothelial marker ) on serial sections , which showed spatially corresponding labeling (Fig. 2G). To independently confirm the predominant expression of ESR1 and not ESR2 transcripts in endometrial tissues, we took advantage of two publicly available single-cell RNA-sequencings atlases of endometriotic tissues, that regroup all the major cell types found in the endometrial tissues (from Fonseca aet al.29, Fig. 1H and Garcia-Alonso et al. 30, Fig. 1 I). These scRNAseq data confirmed the predominant expression of ESR1 in both epithelial and stromal cells of all tissue types and almost no expression of ESR2. ESR1 was also modulated along the menstrual cycle, with a decrease of expression during the secretory phase compared to the proliferative phase (Fig. 1J). In agreement with our data, ESR2 transcript expression was present at low levels in endothelial cells, and was nearby undetectable in e pithelial and stromal compartments (Fig. 1K). Overall, these data demonstrate that ESR1 transcripts are highly expressed in endometriotic lesions and eutopic endometrium of EM S women whereas ESR2 transcripts remains lowly expressed, and detectable in endometrial endothelial cells. ERβ protein is expressed at low levels in endometriosis lesions, including ovarian lesions We then performed immunohistochemistry, using different validated antibodies against ER α and ERβ to evaluate expression of the steroid hor mone receptors at the protein level. To ensure consistency and protocol validation, the clinically approved ER α antibody (clone SP1, Roche Diagnostics) was used along with the two validated monoclonal PPZ0506 27, and CWKF12 31 antibodies for ERβ immunolabeling. First, we cross-validated the specificity of the anti-ERβ antibodies, using cell lines transfected or not with ERβ- expressing plasmids (see Materials and Methods). ERβ expression was detected in both MDA-MB-231 and MCF7 transfected cells by RT -qPCR and Western blot (Supplementary Fig. 2A-B) as well as by immunohistochemistry which showed strong immunolabeling, and thereby validated the accuracy of the antibodies used (Fig. 2A). In ovarian positive controls containing an antral follicle, in which ESR1 and ESR2 transcripts were expressed at similar levels (Fig. 2B, left panel), strong ERα labeling was observed (middle panel), whereas both anti -ERβ antibodies produce only faint immunostaining, mainly in granulosa cells, with comparable results for the two anti-ERβ antibodies (Fig. 2B, right panels). Analysis of ER β expression in endometrial tissues using the PPZ0506 antibody revealed only weak labeling in rare cells from both control and EMS women, as well as in all lesions tested, with less than 2.5 % of cells being ERβ-positive, independently of their location ( SPE, DE or OMA , Fig. 2 C-D) and of the menstrual phase. In the endometrium, both anti-ERβ antibodies showed similarly weak staining in a small subset of cells with a morphology suggestive of endothelial cells, consistent with the relatively higher ESR2 transcript expression detected in this cell population (Fig. 1 G). The presence of ERβ -positive cells remained very low, w ith H-scores of only 2.2 ± 1.4 in epithelial cells and 1.3 ± 0.9 in stromal cells of the endometrium from HW women during phase 2. No increased expression was observed in any endometriotic lesion, including OMA. These findings indicate that the percentage of ERβ-positive cells is ARTICLE IN PRESS ARTICLE IN PRESS low (between 2-3 %) and remains nearly undetectable in endometriotic tissues. Therefore, we focused on the expression of ERα for the remainder of our study. ERα and Ki -67 expression are unchanged in control versus EM S endometrium, whereas PR is decreased in the stroma of EMS women during the late secretory phase We then assessed the variation of ERα expression in endometrium of both control and EM S women together with the progesterone receptor (PR) along the menstrual cycle, distinguishing between the proliferative and secretory phases (Fig. 3). Importantly, Ki -67 as a marker of cell proliferation, was highly expressed in the epithelium of the endometrium in both control women and those with endometriosis, during the proliferative phase, compared to the secretory phase, confirming the accurate identification of the menstrual phase (Fig. 3A-B). ERα protein was highly expressed in endometrium of control women, both in epithelial and stromal cells (Fig 3 A-B). Similar ERα expression levels and distribution were observed with two independent antibody clones (SP1, recognizing all ER α splice variants, and 6F11, specific for the full -length isoform32) (Supplementary Fig. 3). A marked and significant reduction in ERα protein expression was observed during the late secretory phase in both epithelial and stromal cel ls of the endometrium, independently of the disease status in HW and EMS women. In epithelial cells, ERα decreased from 98.1± 0.6 to 41.7±14 in HW women (P<0.0001) and from94.4±2 to 46.3±10.9 (P<0.01) in EMS women (Fig. 3B, right panel). We also quantified PR protein expression, a key hormonal regulator of the female reproductive system. In epithelial cells, PR expression closely paralleled that of ERα expression, with a marked decrease during the late secretory phase in both control and EM S endometrium. In contrast, stromal PR expression remained stable across the menstrual cycle in HW women (82.3% versus 80.2% in the proliferative and late secretory phases, respectively) whereas it was significantly reduced in stromal cells from EMS women during the late secretory phase (49.1% versus 87.1 % in P and LS, respectively) , indicating a difference between healthy and EMS women. Overall, ERα expression in the endometrium did not differ significantly between HW and women with endometriosis, whereas PR expression was reduced in the stromal compartment of EMS women compared to HW controls. ERα and PR are broadly preserved but vary across endometriotic lesion types We then compared ER α and PR expression in endometriotic lesions ( SPE, DE and OMA) with that of eutopic endometrium within the same patients (Fig. 4 A). No significant differences were observed between the proliferative and secretory phase, likely because early and late secretory phases were no longer distinguished, except for the PR expressio n in eutopic endometrium and DE lesions. Overall, ERα and PR expression remained preserved across all tissue types in both epithelial and stromal cells (Fig. 4B). However, some heterogeneity was observed, particularly, a significant reduction in the percentage of ERα- positive cells in the stromal compartment, depending of lesion type. During the proliferative phase, 81.3% of stromal cells were ERα-positive in the endometrium , compared with 52% in SPE lesions, 74% in DE lesions and 43.8% in OMA lesions . PR expression was also significantly reduced during the proliferative phase in the epithelial cells of OMA lesions compared with endometrium, SPE and DE lesions. Moreover, the percentage of PR-positive cells in the stromal compartments differ ed significantly, according to the ARTICLE IN PRESS ARTICLE IN PRESS lesion types and menstrual phase. The proliferative marke r, Ki-67, also showed substantial variability across samples. We also assessed AR expression on these TMAs, as androgen signaling play also critical roles in the female reproductive tract. We could confirm that AR expression is significantly the highest in the stromal cells of the upper functional layer of epithelium of control women during the beginning of the menstrual cycle (proliferative phase and early secretory phase), as already observed by Gibson and his collaborators (2020)8 (Supplementary Fig. 4). AR expression remai ns high in the stromal cells of the lesions, and surprisingly, is now also detected in the epithelial cells of the DE lesions, although at a level below 10%. DE lesions show altered expression patterns of steroid receptors such as ERα, PR and AR, indicating a potential disruption in paracrine signaling between stromal and epithelial cells To evaluate potential variations in steroid receptor expression patterns between lesions and matched eutopic endometrium, we assessed the expression of ERα, PR, AR and Ki-67 markers in EM S lesions and their corresponding matched endometrium from the same patients (Supplementary Fig. 5). This paired comparison was performed to minimize potential variability related to menstrual cycle phase. Overall, we observed substantial inter-patient and intra -patient heterogeneity in the abundance of ER α- and PR- positive cells, particularly in the lesions. No consistent directional change (increase or decrease) was identified between matched eutopic endometrium and lesion samples. To a ddress the marked inter -patient variability in receptor expression, we performed paired comparisons between lesions and their matched eutopic endometrium. The percentage of ER α-, PR-, Ki- 67-, and AR-positive cells was therefore compared in matched lesional and eutopic tissues (left panels of Fig. 5A, C, E, G for epithelial cells and Fig. 5B, D, F, H for stromal cells). To further quantify these paired differences, we also calculated the change in the percentage of positive cells between each lesion and its matched eutopic endometrium (lesion – EUT). These values were then summarized by lesion subtype s (SPE, DE, OMA) to evaluate the average direction and magnitude of receptor expression changes relative to the matched endometrium (right panels of Fig. 5A –H). Positive values indicate higher expression in lesions relative to the matched eutopic endometrium, whereas negative values indicate lower expression. ERα expression was significantly decreased in both epithelial and stromal compartments of SPE lesions (P = 0.02 and P = 0.007, respectively) whereas epithelial ERα expression remained unchanged in DE lesions. In SPE lesions, ERα expression decreased by an average of 16.6% in epithelial cells and 21% in stromal compartments with reduction reaching up to 65% (Fig. 5A-B, left panels). Stromal ERα expression was also significantly reduced across all lesion types, with the strongest decrease observed in OMA lesions (average 42%) . In DE lesions, the stromal decrease in ERα was associated with a small but significant increase in epithelial PR expression compared to matched endometrium (P = 0.011, Fig. 5 C, left panel ) with an average increase of 10% and values reaching up to with a +78%. Although no significant overall change in stromal PR expression was observed (P=0.07), individual analysis showed that around 25% of DE lesions exhibited reductions of up to 40%. These changes were associated with a marked significant decrease in Ki-67 expression in DE lesions (P<0.0001, Fig. 5E middle panel) with an average reduction of 21% and a maximum decrease of up to 79%. In parallel, AR expression was significantly increased in epithelial cells of DE lesions (P=0.021, Fig. 5 G, left panel ) with an average increase of 3.9 % and values reaching up to 31%. Finally, OMA lesions showed a significant decrease in both ERα and PR expression in stromal cells. ARTICLE IN PRESS ARTICLE IN PRESS Overall, d espite the marked heterogeneity in steroid hormone receptor expression across lesions, these findings highlight differences in ERα expression across the disease . In the SPE lesions, a small but significant decre ase in E Rα expression was observed in both epithelial and stromal cells compared to matched eutopic endometrium. In contrast , DE lesions exhibit an altered ERα/PR expression, characterized by decreased stromal ERα and small but significant increase in epithelial PR, indicating that the hormonal signaling of these cells of endometrial origin is altered. In endometrium, ERα is highly expressed in epithelial and stromal cells thr oughout the estrogen - dominant phase and decline during the secretory phase, while PR is maintained in the endometrial stroma and decline in the epithelium 7. We then assessed the correlation between ERα expression and that of PR, Ki-67 and AR across all tissue types, distinguishing expression in epithelial and stromal cells, respectively (Table 1 ). Spearman correlation analysis demonstrated a positive correlation between ERα and PR expression in both epithelial and stromal c ompartments of the endometrium and SPE lesions, whereas this association was weaker in DE lesions. ERα and AR expression were also positively correlated, but only in stromal cells of the endometrium and SPE lesions, with no significant association in epithelial cells or DE lesions. In contrast, the correlation between ER α and Ki -67 was confined to epithelial cells of the endometrium and OMA lesions and was absent in all other cell types and tissues. These findings support the conclusion that ER α remains the predominant estrogen receptor in endometriotic lesions, while its coordinated expression with PR is preserved in endometrium and SPE lesions but diminished in DE lesions. To further investigate the altered ER α/PR relationship in endometriosis, we performed multiplex RNAscope for ESR1 and PR on the same tissue sections, to determine whether the differential correlations observed at the protein level were also recapitulated at the transcript level ( Supplementary Fig. 6A). A significant positive correlation between ESR1 and PR transcripts was observed in both epithelial and stromal cells of the endometrium, regardless of whether the w omen had endometriosis (Supplementary Fig. 6B). In contrast, this correlation was lost in all lesion types and in both cellular compartments. Together, these results support the presence of altered ESR1 –PR transcriptional coordination in endometriotic lesions, consistent with changes in steroid receptor signaling, and suggest impaired ERα/PR-mediated paracrine signaling between epithelial and stromal compartments in endometriosis.

Millions of women are affected by endometriosis and receive hormonal treatments to suppress estrogen signaling, thereby reducing the stimulation of endometrial tissue. This tissue is dynamic and highly sensitive to the effects of estrogen and progesterone because it expresses high -affinity hormone receptors. However, resistance to the proposed hormonal treatments occur, and a comprehens ive analysis of expression of these nuclear receptors ERα, ERβ, PR and A R is required to improve our understanding of the disease and develop more effective therapeutic strategies. The use of bulk tissue analysis techniques, such as RT-PCR and Western blotting to unravel receptor expression in endometriotic lesions, has important limitations, as these methods provide only an average measure of expression across the tissue and do not allow assessment of cell -specific localization or tissue distribution. Furthermore, the use of non -specific antibodies for ERβ induced confusion with regards to the presence of ERβ receptor in endometriosis27. Our study, combin ing RNAscope multiplex technology and immunohistochemistry on TMAs, demonstrates that estrogen receptor alpha (ERα) is the predominant estrogen receptor expressed at both ARTICLE IN PRESS ARTICLE IN PRESS the transcriptional and translational level in the endometrium as well as in all types of endometriotic lesions studied —SPE, DE and OMA. ERα expression was 10 - to 100 -fold higher than ER β, in line with previous reports17,33. Consistently, ESR2 expression was minimal, whereas ESR1 levels remained markedly higher in eutopic endometrium and endometriotic lesions. At the protein level, ERβ was detected only at very low levels in epithelial and stromal cells from both endometrial and endometriotic tissues, with mean expression below 2.5%. This near absence of ER β was reproducibly confirmed using two independently validated antibodies: PPZ0506, directed against amino acids 2 –8827 and CWK-F12, developed by B. Katzenellenbogen and independently validated by Nelson et al. (2017) 31, directed against amino-acids 256–505 . The absence of ERβ expression was also observed in ovarian lesions, despite the possibility of residual ERβ expression from normal granulosa cells . Th ese results contrast with previous studies reporting higher ESR2 mRNA expression in endometriomas than in superficial or deep infiltrating endometriosis lesions, or in normal uterine endometrium 18,23,24,34. Nevertheless, functional studies have shown that ERβ overexpression in a murine model of endometriosis was associated with enhanced disease progression26. Overall, the discrepancies between our findings and previous reports are likely due to methodological differences, including the use of highly specific probes for ESR2 transcripts and validated antibodies with confirmed ERβ specificity27,31. Moreover, we detected low levels of ESR2 mRNA expression in vascular cells of the endometrium from both control and EMS women, consistent with the presence of low ERβ protein levels. Although ERα is well established as a key mediator of the vasculoprotective effects of estrogens , its role appears particularly prominent in large arteries such as the aorta and carotid arteries , where it contributes to cardiovascular pro tection in females , by promoting vasodilation, angiogenesis, and anti -inflammatory responses35,36. In contrast, ERβ has been detected in certain microvascular beds, and has been previously been identified in primary human uterine microvascular endothelial cells 37,38. Andersson et al. 27 also described ERβ expression in endothelial cells from endometrial cancer. Together, these independent observations support the findings of our manuscript. Specifically, Tamura (2013)37 showed that estradiol increases COX -2 protein expression and PGE2 production in these cells through estrogen receptor signaling, most likely via ERβ, since ERα was not detected in these cells. Additionally, Greaves and her collaborators38 reported that ERβ activation differentially regulates endothelial cell function in the endometrium and myometrium , with in some cases opposing effects. In endometrial endothelial cells, ERβ activation was associated with reduced angiogenic activity. These findings suggest that ERβ may have a specific role in endometrial endothelial cells, particularly at the end of the secretory phase, contributing to the complex regulation of uterine vasculature across the menstrual cycle. Future directed studies are now warranted to determine how endothelial ERα and ERβ in these endothelial cells affect uterine physiology and endometriosis. Besides the predominant expression of ERα in all endometrial tissues, our findings also confirm the marked heterogeneity of ERα expression in endometriotic lesions, a feature previously reported in the literature21. Beyond the variability associated with lesion location and disease stage, a key contributing factor is the dynamic fluctuation in receptor expression across the menstrual cycle , particularly the pronounced decrease in ERα and PR levels in epithelium during the l ate secretory phase, which follows the rise in progesterone. Although this temporal variation has long been recognized 7,8, it has not always been adequately considered in previous analyses, which may have contributed to inconsistencies in the literature. ARTICLE IN PRESS ARTICLE IN PRESS Using tissue microarrays and matched samples, we direc tly compared endometriotic lesions and eutopic endometrium under identical conditions, despite di sease heterogeneity. W e observed reduced ERα expression in SPE lesions in both epithelial and stromal compartments, whereas in DE and OMA lesions, this decrease was predominantly stromal. In DE lesions, this stromal loss of ERα may contribute to the reduced epithelial proliferation observed, consistent with evidence that estrogen-driven epithelial proliferation is mediated by ERα-positive stromal cells39. These alterations in ERα expression also suggest that the estrogen-responsive stromal compartment is always affected across lesion types, which may in turn influence steroid receptor crosstalk and downstream hormonal responses. Accordingly, PR expression displayed lesion - and compartment -specific variations, with a modest increase in the epithelium of DE lesions and a decrease in the stroma of OMA lesions. Together with previous reports of variable PR expression 40,41, these findings highligh t the complex regulation of progesterone signaling in endometriosis. Notably, we found a loss of coordinated ERα and PR expression in ectopic lesions at both the transcriptional and translational levels, in contrast to the correlated expression pattern maintained in the endometrium ( Table 1 and Supplementary Fig. 6B). Because ERα /PR regulatory feedback normally operates across the menstrual cycle in endometrium 42, disruption of this relationship is consistent with altered steroid receptor signaling and supports the concept of progesterone resistance in endometriosis. Importantly, our results indicate that PR expression is not uniformly reduced, but instead varies by lesion type, menstrual phase and cellular compartment, highlighting the heterogeneity of progesterone signaling acro ss endometriotic lesions. This heterogeneity may have clinical relevance, as altered progesterone receptor expression has been associated with response to progestin therapy in endometriosis43. Together, these observations suggest that progesterone resistance in endometriosis may reflect not only changes in PR abundance and signaling , but also altered estrogen receptor ER α coordination and downstream signaling. Finally, while androgens play some role in epithelial and stromal cross talk 4, being normally expressed in stromal cells during the early secretory phase protecting these cells against stress or apoptosis44, the present work also reported an elevated AR expression in epithelial cells of DE lesions (Fig. 5G). Once more, this AR elevated expression in epithelial cells is the signs of strong dysregulated pathways indicating that these endometrial lesions, although endometrial -like in appearance, with epithelial and stromal cells, are not functionally the same. Further studies on the role of these AR expression in the regulation of endometrial function are needed to better understand their actions in the pathogenesis of endometriosis. Only a small subset of lesions exhibited very low steroid receptor expression (having less than 20% ERα and PR -positive cells ), confirming that ERα and PR remain key markers for both detection and stratification Notably, dysregulation of these receptors was more frequent in DE lesions compared with SPE lesions. The proportion of ERα- and PR-positive cells therefore represent a useful classification criteria. However, the marked heterogeneity of endometriotic lesions, as demonstrated by our findings and those of others, must be more fully acknowle dged, as it poses a challenge to establishing a straightforward classification or prognostic model based on steroid receptor expression. These findings are of particular interest, since they are important to challenge response to hormonal treatments. Spec ific agonists and antagonists for ERα and ERβ have been respectively generated45 while ERα also acts via complex nuclear and membrane actions on a tissue -specific manner. Some ligands, such as estetrol, a natural estrogen have a specific pharmacological activation profile46, that ARTICLE IN PRESS ARTICLE IN PRESS have shown exciting differences in endometriosis that wi ll need further evaluation for endometriosis treatment47. Notably, selective inhibition of ERβ activity was also found to significantly reduced ectopic lesion growth in a mouse model 26. However, while ERβ is not highly expressed in EMS lesions, this current work completely rules out the possibility to use these selective antagonists to block the development of lesions. In contrast, ERα is the only form of ER receptor expressed in EMS lesions, highlighting that it is the primary ER target to design therapeutic strategies. Overall, our findings indicate that ERα is the predominant ER expressed in epithelial and stromal cells of the endometrium and lesions, with limited evidence for opposing ERβ activity. This expression pattern may promote an environment that favors ERα -driven proliferative and inflammatory responses whereas ERβ might have a more restricted role in endometrial vasculature. These results challenge the notion of broad ER β expression in endometriosis and highlight the relevance of ERα and PR profiling for lesion stratification, diagnosis and development of hormone -based therapeutic strategies. The dysregulated receptor landscape observ ed in stromal and epithelial compartments further emphasizes the need for continued research to better understand estrogen signaling in endometriosis. ARTICLE IN PRESS ARTICLE IN PRESS

Patient cohorts The study involved biopsies from 15 healthy women and 34 women with endometriosis in different phases of the menstrual cycle: proliferative (n= 7 healthy and n= 17 endometriotic); early secretory (n= 3 healthy and n= 7 endometriotic; late secretory (n= 5 healthy and n= 10 endometriotic )). Cycle phase dating was defined by surgeon, pathologists and confirmed by immunofluorescence staining using antibodies against CPM (proliferative phase) and PAEP (end of secretory phase) 48. The endometriotic biopsies from different locations (classified as SPE, DE or OMA by the surgeon) were obtained from 34 endometriotic women who underwent hysterectomy surgery or laparoscopic surgery in the Gynecological Surgery Department of the Toulouse University Hospital. To compare with endometrium from healthy women, normal endometrium from 15 healthy women undergoing laparoscopic surgery for non-malignant gynecologic indications were used as controls. All samples were obtained from pre -menopausal women aged 22-49 years old (median age-35 years old), with regular menstrual cycles, and no hormonal contraceptives or no recent hormonal uterine device usage in the past 2 months (See Supplementary Table 1). Endometrial biopsy was obtained using the Pipelle de Cornier biopsy device. Written informed consent was received prior to participation. Patients’ samples were obtained after written informed consent prior to participation, in accordance with the Declaration of Helsinki and stored at the “CRB Cancer des Hôpitaux de Toulouse (BB -0033-00014)” collection. According to the French law, CRB Cancer collection has been declared to the Ministry of Higher Education and Research (DC-2008-463 and DC-2020-4074) and obtained a transfer agreement (AC-2013- 1955 and AC-2025-7615) after approval by the ethics committee (Comité de Protection des Personnes Sudouest et outre mer II, CPP). Clinical and biological annotations of the samples have been declared to the CNIL (Comité National Informatique et Libertés). Tissue Microarrays (TMAs) 3 different TMAs were generated from the endometrial biopsies or endometriotic lesions of women included in the cohort (described above), using an automated tissue microarrayer (Ex cilone, Elancourt, France). Representative endometrial tissue -rich areas, previously annotated by a certifi ed pathologist, were selected from the formalin -fixed paraffin -embedded (FFPE) patient tissue blocks. Each TMA consisted of cores with a diameter of 2 mm randomly distributed and included duplicate cores per lesion per woman when possible. Tissue sections (thickness: 4μm) were prepared from the TMAs for RNAs-cope In Situ Hybridization (TMA-1 and -2) or immunohistochemistry (all TMAs). Determination of menstrual phase by Immunofluorescence staining with CPM and PAEP antibodies The TMA slides were deparaffinized using Toluene (386001 Carlo Erba), and an acid unmasking procedure was performed (H -3300, Vector laboratories). Samples were permeabilized and blocked with blocking buffer (1% BSA (Sigma), 4% normal goat serum (ab7481, Abcam), 0.1% TritonX -100 (Sigma) in PBS) 1h at RT. Slides were incubated with primary antibodies CPM (HPA002657 -100UL, Sigma , lot 27155, RRID: AB_1078398) and PAEP (ab17247, lot 1022890 -1, Abcam; RRID: AB_2159754) overnight at 4°C. Then sections were incubated with the conjugated secondary antibodies Alexa Fluor 647 (Abcam) and the nuclei were counterstained with DAPI (Invitrogen, Life Technologies). The samples were then mounted with ARTICLE IN PRESS ARTICLE IN PRESS Mounting Medium (S3023, DAKO) and acquired using a LSM900 Fluorescence Microscope (Zeis s). The images were further processed using QuPath v0.5.0. Cell culture and transfection, and preparation of cytoblocks MDA-MB-231 (ATCC-HTB-26) and MCF7 (ATCC-HTB-22) cells were transiently transfected with either the pSG5puro-ERβ plasmid or the empty vector (gift of P. Balaguer, IRCM, Montpellier, France 49) as a control using the Lipofectamine 3000 (L3000-015, Invitrogen, Villebon-Sur-Yvette, France). Cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM/F12; P04 -41250, Dutscher , Ber nolsheim, France ) supplemented with 5% fetal calf serum (S181W-500; Biowest , Nuaillé, France ) and antibiotics (P0781, Sigma, Lezennes, France) at 37 °C under 5% CO2. Transfected MDA -MB-231 cells were formalin -fixed and paraffin -embedded using the Shandon™ Cytoblock™ Cell Block Preparation System, according to the manufacturer’s protocol. Gene expression analysis Cells were freeze -dried and then RNA was extracted using TRIzol (Invitrogen, Carlsbad, CA). 1000 ng of RNA was reverse transcribed (RT) at 25°C for 10 min and then at 37°C for 2 h using the High Capacity cDNA reverse transcriptase kit (Applied Biosystems). For gene expression, qPCR w as performed using SsoFast EvaGreen Supermix (Bio -Rad, Marnes -La coquette, France ) with primers efficiency validated using standard curves method (95% < efficiency < 105%). Gene expression was quantified using the comparative Ct (threshold cycle) method. HPRT gene was used as housekeeping gene to normalize the mRNA. The primer sequences were: ESR2: Forward: 5’ - GTC AGG CAT GCG AGT AAC AA - 3’; Reverse: 5’ - GGG AGC CCT CTT TGC TTT TA - 3’; HPRT: Forward: 5’ - TGC TTT CCT TGG TCA GGC AGT - 3’; Reverse: 5’ - CTT CGT GGG GTC CTT TTC ACC - 3’. Exploration of public scRNAseq datasets We used publicly available datasets from Fonseca et al 29 (Accession number GSE213216) and Garcia - Alonso et al. 30 (Accession numbers E-MTAB-10287). RDS files were download and processed using Seurat. Figures were generated using the FeaturePlot function in Seurat, or using ggplot2 package. Western blot analysis The cells were lysed using lysis buffer (150mM NaCl, 50mM Tris-HCl (pH 7,5), 1% NP40, 1mM EDTA, 5mM NaF, 1mM orthovanadate, 0,5mM DTT, proteinase inhibitors (Complete™ EDTA -free, Roche, Boulogne- Billancourt, France), 0,1% SDS)) Total proteins were separated on a 10% SDS/PAGE gel and transferred to nitrocellulose membranes using Trans Blot Turbo RTA Transfer Kit 0.2 µm Nitrocellulose. The ladder was the PageRuler Plus Prestained Protein ladder (10-180 kDa) from ThermoFisher Sientific , Illkirch, France . The following primary antibodies were used: anti-ERβ (clone PPZ0506, Cat#PP-PPZ0506-00, lot A -2, RRID:AB_2293861, R&D Systems , 1/1 000; anti-ERβ CWKF12 (DSHB depository, 1/200) and anti-GAPDH (clone 6C5,sc-32233, lot 71628, RRID:AB_627679, Santa Cruz , dilution 1/2000 ). Then, revelation was performed using HRP -conjugated secondary antibodies (Goat anti -mouse IgG (HRP), Cat# 91196, RRID:AB_2940774, Cell Signaling Technology ) and visualized by ECL detection according to the manufacturer’s instructions (Amersham Biosciences/GE He althcare), using ChemiDoc Imaging System (Bio-Rad). Bands were quantified using ImageLab. ARTICLE IN PRESS ARTICLE IN PRESS Multiplex Fluorescent RNAScope assay on BOND RX The RNAScope LS Multiplex Fluorescent assay (ref. 322800, 32327, ACD, BioTechne, MN, USA) was used according to manufacturer’s procedures (technical note: UM 322800/Rev B) and performed on the BOND RX research advanced automated slide stainer (Leica Biosystems, Nusslo ch, Germany). Formalin -fixed paraffin embedded (FFPE) TMA tissue sections of 4µm were prepared and air dried overnight. Quality of RNA of the tested samples was verified and validated using the RNAscope  2.5 LS 3-plex Positive Control Probe-Hs (ref. 320868, ACD, BioTechne, MN, USA). Non -specific staining was absent as confirmed by the use of RNAscope 2.5 LS Multiplex Negative Control Probe dapB (Bacillus subtilis strain, ref. 320878, ACD, BioTechne, MN, USA). Probes targeting the widely expressed Peptidylprolyl Isomerase B (PPIB) and against DNA-directed RNA polymerase II subunit RPB1 (PolR2A) were used as positive controls, to ensure about the RNA integrity. PPIB is a commonly used housekeeping gene for this purpose, while POLR2A serves as an alternative p ositive control, particularly in proliferating tissues. FFPE TMA tissue slides were heat - pretreated using ER2 pretreatment solution (pH8, Leica Biosystems, Nussloch, Germany) for 15 minutes at 95°C and subsequently incubated using RNAscope  2.5 LS Protease III (ref. 322800, ACD, BioTechne, MN, USA) during 15 minutes. The target probes used in this study were were as follows: RNAscope™ 2.5 LS Probe - Hs-ESR1-C2 (ref. 310308 -C2 targeting region 1251 -2376, ACD, BioTechne, MN, USA) and RNAscope™ 2.5 LS Probe- Hs-ESR2 ( ref. 470158-C1 targeting region 456 - 1995, BioTechne SAS, MN, USA) or the combination of RNAscope™ 2.5 LS Probe - Hs-ESR1 (ref. 310308 targeting region 1251 -2376, Bio- Techne SAS, MN, USA)and RNAscope™ 2.5 LS Probe - Hs-PGR-C2 (ref. 589758-C2 targeting region 1609 - 2579, BioTechne SAS, MN, USA) These target probes recognized all splice variants and were visualized using OPAL TM dyes (AKOYA Biosciences, Marlborough, USA) with excitation and emission wavelengths compatible with our whole slide im aging system were used: OPAL TM 570 and OPAL TM 650 for ESR2 and ESR1, or OPAL TM 520 and OPAL TM 570 for ESR1 and PR (1/1500 in 1x Plus Automation Amplification Diluent, AKOYA Biosciences, Marlborough, USA). The tissue slides were counterstained using RNAscope LS Multiplex DAPI (AKOYA Biosciences, Marlborough, USA) and mounted with Invitrogen TM ProLongTM Gold Antifade Mounting medium (Life Technologies, ThermoFisher Scientific, California, USA). Multispectral fluorescence imaging was performed using a n AxioScan Z1 (Carl Zeiss Microscopy, Oberkochen, Germany) whole-slide scanner with appropriate narrow band -pass excitation and emission filters and specific dichroic mirrors (Semrock Inc., Rochester, NY, USA), and with a multi -channel solid- state light engine (Colibri 7, Carl Zeiss Microscopy, Oberkochen, Germany) equipped with 7 LEDs covering the entire visible spectrum from UV to far -red (370 – 648 nm). Analog to digital image sampling was performed at 16 -bit (65 536 grey levels, 37 000:1 dynamic range) with a high -resolution scientific complementary metal oxide semiconductor (sCMOS sensor with 2 048 x 2 048 cells of size 6.5 x 6.5 µm each) Peltier -cooled monochrome camera (Orca Flash 4.0 V3, Hamamatsu Photonics K.K., Japan), to achieve a final scan resolution of 0.32 µm/pixel. RNAscope quantification Quantification of RNAscope  signals was performed at 20× magnification and the images were further processed and quantified using QuPath (v0.5.0) in a blinded manner. Quantification was conducted across ARTICLE IN PRESS ARTICLE IN PRESS multiple regions of each biopsy to ensure reliable assessment and to account for tissue heterogeneity. When duplicate biopsy spots were available, both spots were analyzed independently, and the mean of the two measurements was used for subsequent analysis. Epithelium and stroma were differentiated in the analysis. An H-score was used to determine the class of each cell and assign a value to each cell type for each TMA spot. Different class of cells were distinguished, de pending of the RNA spot number per cells, i.e. Class 0 (0 spot/cell), Class 1 (1 -5 spots/cell); Class 2 (6-10 spots/cell) and Class 3 (>10 spots/cell or >20% cluster in cells). The following formula was then used to calculate the average number of RNA/cell: (0 x % cells class 0) + (1 x % cells class 1) + (2 x % cells class 2) + (3 x % cells class 3). So, the maximum H-score is 300 if all cells are class 3. ERβ Immunohistochemistry Formalin-fixed paraffin embedded (FFPE) tissue sections (TMA and cell block) of 4µm were prepared and air dried overnight. ERβ Immunohistochemistry was automated on the AS48 automated stainer (Agilent Technologies, CA, USA). Dewaxing and antigen retrieval was performed using a PT Link pressure cooker (Agilent Technologies, CA, USA ) for 25 minutes at 97°C using En VisionTM FLEX target retrieval solution (pH6, ref. K800521, Agilent Technologies, CA, USA). Primary anti -ERβ antibodies ( clone PPZ0506, lots XF3616191 and 797560-25, RRID: AB_2717280, ThermoFisher Scientific, 1/200 and lot A-2, RRID:AB_2293861, R&D Systems , in Envision FLEX antibody diluent (Agilent technologies, CA, USA), and CWKF12 (DSHB depository, 1/200 in Envision FLEX antibody diluent (ref. K800621, Agilent technologies, CA, USA)) were incubated for 30 minutes at room temperature and visualized using the En VisionTM FLEX HRP detection system (ref. K800221 Agilent technologies, CA, USA). Slides were subsequently counterstained using haematoxylin/eosin (ref. K800821, Agilent technologies, CA, USA), dehydrated and mounted using xylene-based mounting in a Tissue Tek automated cover slipper (Sakura FineTek Europe, AV, The Netherlands). Stained slides were digitized with a Panoramic 250 Flash II digital microscope (3DHISTECH, Budapest, Hungary) equipped with a Zeiss Plan -Apochromat 20X NA 0.8 objective and a CIS VCC -FC60FR19CL 4 - megapixel CMOS sensor (unit cell size 5.5 x 5.5 µm) mounted on a 1.6X optical adaptor, to achieve a scan resolution of 0.24 μm/pixel in the final image (corresponds to 41.1X magnification at the hig hest optical resolution in traditional microscopy). ERα, PR, AR and Ki67 Immunohistochemistry Automated classical immunohistochemical (IHC) stain was performed using the Benchmark ULTRA (Roche, Ventana Medical Systems, Innovation Park Drive Tucson, Arizona 85755 USA) on FFPE tissue sections (4µm). After dewaxing, tissue slides were heat pre-treated using a CC1 (pH8) buffer (05424569001, Roche Diagnostics, IN, USA) at 98°C. The slides were blocked for endogenous peroxidase activity and incubated with primary anti-ERα (clone SP1, Cat# 06523838001, lot N11424, RRID: AB_2335977; Roche Diagnostics, IN, USA or clone 6F11, Cat# MA1-80216, lot OL1796492, RRID: AB_930763, Thermo Fisher Scientific), anti- PR (clone 1E2, Cat# 05278392001, lot N09748, RRID: AB_2335976, Roche Diagnostics, IN, USA) , anti-AR (clone SP107, Cat# 06523838001, lot V0005472, RRID: AB_10903299 Roche Diagnostics, IN, USA) and anti-ERG (Cat# 790-4576, lot N16218, RRID:AB_2861321, Roche Diagnostics) antibodies. The targets were then visualized sing the UltraView DAB Detection kit (05269806001, IN, USA). For Ki67 (clone 30 -9, Cat # 05278384001, RRID:AB_2631262, lot N60087, Roche Diagnostics, IN, USA) targets were visualized using ARTICLE IN PRESS ARTICLE IN PRESS the OptiView DAB detection kit ( 06396500001, Roche Diagnostics, IN, USA). The tissue slides were counterstained using hematoxylin II (05277965001, Roche Diagnostics, IN, USA) for 8 minutes followed by post-coloration using Bluing reagent for 4 minutes at room temperature (05266769001, Ro che Diagnostics, IN, USA). The slides were then dehydrated (ethanol and xylene) and mounted using xylene - based mounting (Sakura Tissue -Tek®, Sakura Finetek Europe, AV, The Netherlands). All antibodies used were validated by the pathology lab, following ISO15189 recommendations. Stained slides were digitized with a Panoramic 250 Flash II digital microscope (3DHISTECH, Budapest, Hungary) equipped with a Zeiss Plan -Apochromat 20X NA 0.8 objective and a CIS VCC -FC60FR19CL 4 - megapixel CMOS sensor (unit cell size 5.5 x 5.5 µm) mounted on a 1.6X optical adaptor, to achieve a scan resolution of 0.24 μm/pixel in the final image (corresponds to 41.1X magnification at the highest optical resolution in traditional microscopy). Statistics and Reproducibility Statistical analysis was performed with the PRISM software v10.1.2 (GraphPad Software, San Diego, CA, USA) (See Supplementary Data 1). Results are expressed as mean ± SEM. Statistical differences between endometrium and endometriotic lesions were assessed using Wilcoxon matched-pairs signed-rank test, two-way ANOVA with Tukey’s multiple comparison tests, as appropriate. Spearman’s correlation coefficients and simple linear regression were used for correlation analyses (*: P < 0.05; **: P < 0.01; ***: P < 0.001). Author Contributions: A.G., M.R., P.S performed the experiments, analyzed the data and contributed to the draft manuscript. AB and N.V.A performed immunostainings and RNAscope multiplex and helped with TMAs. K.B. helped with immunofluorescence stainings. LB conducted the bioinformatic re -analysis of the scRNA -seq data. E.C., A.W provided material and helped to edit the manuscript. F.L., and E.C. analyzed and critically reviewed the data, obtained funding and wrote the manuscript. P.L. helps with funding. F .L and E.C. designed, conceived and supervised the study. F.L. wrote the manuscript with input of LB, CF and JFA, and editing was performed by all the authors. Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Data availability: Supplementary Table 1 provides information on patient status, menstrual cycle phase, and the analyses in which each sample was included, namely RNAscope (Figure 1 and Supplementary Fig. 6 ) or immunohistochemistry (IHC; Figures 2–5). The raw data underlying the charts and graphs are provided in the corresponding Source Data file, as indicated in the relevant figure legends. Results of the statistical analyses are presented in Supplementary Data 1. All results associated with the datasets used to generate the figures are available in Supplementary Data2. Conflict-of -interest: ARTICLE IN PRESS ARTICLE IN PRESS The authors have declared that no conflict of interest exists. Fundings The work at Inserm U1297 was supported by the National Institute of Health and Medical Research (INSERM), University Paul Sabatier - Université de Toulouse; University Hospital Center of Toulouse, Région-Occitanie-Midi-Pyrénées-GRAINE-ENDOTREAT, Urosphere, the Endofrance Association and ANR- 23-CE17-0010-01 (EDISON) . A. Gargaros was funded by INSERM and Région -Occitanie Midi -Pyrénées CEBBOC. A. Buffeteau was funded by Société Française d'Endocrinologie (SFE). P. Singla was supported by ANR-23-CE17-0010-01 (EDISON) and K. Boriak obtained funding from ANR-PAUSE (Ukraine).

We thank P. Balaguer for providing the plasmid pSG5puro-ERβ and its empty vector. We acknowledge A. Lucas and C. Bernis from the We-Met Functional Biochemistry Facility (INSERM U1297, Toulouse, France), R. Florès-Florès from the imaging plateau of the TRI platform Genotoul (INSERM U1297, Toulouse, France), Dr. P. Brousset and Dr. Anne Gomez-Mascard from the Department of Pathological Anatomy and Cytology of IUCT, Toulouse for their help on Tissue collection, immunostainings and RNAscope ® technology. We gratefully acknowledge François-Xavier Frenois from the Imag’IN Platform of the University institute of Cancer (https://www.ibisa.net/plateformes/imag-in-368.html, Toulouse) for the slide scanning of TMAs, Sophie Péries from the CRB (centre de ressources biologiques, CHU Toulouse) for their help in anatomopathological analysis and generating the TMAs, Estelle Michau d and Lou -Na Redoute, from Inserm U1297, for their help with RNAscope quantification. We gratefully acknowledge all the surgeons in the Department of Gynecological Surgery at the Toulouse University Hospital Center for their contribution to this work. We would like to thank all women who participated in this study. ARTICLE IN PRESS ARTICLE IN PRESS

1 Horne, A. W. & Saunders, P. T. K. SnapShot: Endometriosis. Cell 179, 1677 -1677 e1671 (2019). https://doi.org/10.1016/j.cell.2019.11.033 2 Zondervan, K. T. et al. Endometriosis. Nature reviews. Disease primers 4, 9 (2018). https://doi.org/10.1038/s41572-018-0008-5 3 Zondervan, K. T., Becker, C. M. & Missmer, S. A. Endometriosis. N Engl J Med 382, 1244-1256 (2020). https://doi.org/10.1056/NEJMra1810764 4 Saunders, P. T. K. & Horne, A. W. Endometriosis: Etiology, pathobiology, and therapeutic prospects. Cell 184, 2807-2824 (2021). https://doi.org/10.1016/j.cell.2021.04.041 5 Chapron, C., Marcellin, L., Borghese, B. & Santulli, P. Rethinking mechanisms, diagnosis and management of endometriosis. Nature reviews. Endocrinology 15, 666 -682 (2019) https://doi.org/10.1038/s41574-019-0245-z 6 Blanco, L. P., Salmeri, N., Temkin, S. M., Shanmugam, V. K. & Stratton, P. Endometriosis and autoimmunity. Autoimmunity reviews 24, 103752 (2025). https://doi.org/10.1016/j.autrev.2025.103752 7 Critchley, H. O. D., Maybin, J. A., Armstrong, G. M. & Williams, A. R. W. Physiology of the Endometrium and Regulation of Menstruation. Physiol Rev 100, 1149 -1179 (2020). https://doi.org/10.1152/physrev.00031.2019 8 Gibson, D. A., Simitsidellis, I., Collins, F. & Saunders, P. T. K. Androgens, oestrogens and endometrium: a fine balance between perfection and pathology. J Endocrinol 246, R75-R93 (2020). https://doi.org/10.1530/JOE-20-0106 9 Wang, W. et al. Single-cell transcriptomic atlas of the human endometrium during the menstrual cycle. Nat Med 26, 1644-1653 (2020). https://doi.org/10.1038/s41591-020-1040-z 10 Mareckova, M. et al. An integrated single -cell reference atlas of the human endometrium. Nat Genet 56, 1925-1937 (2024). https://doi.org/10.1038/s41588-024-01873-w 11 Liu, H. et al. Structural and functional evidence that GPR30 is not a direct estrogen receptor. Cell Res 34, 530-533 (2024). https://doi.org/10.1038/s41422-024-00963-y 12 Critchley, H. O. et al. Wild-type estrogen receptor (ERbeta1) and the splice variant (ERbetacx/beta2) are both expressed within the human endometrium throughout the normal menstrual cycle. J Clin Endocrinol Metab 87, 5265-5273 (2002). https://doi.org/10.1210/jc.2002-020502 13 Juhasz-Boss, I. et al. Endometrial expression of estrogen receptor beta and its splice variants in patients with and without endometriosis. Arch Gynecol Obstet 284, 885 -891 (2011). https://doi.org/10.1007/s00404-010-1768-7 14 Collins, F. et al. The ERbeta5 splice variant increases oestrogen responsiveness of ERalphapos Ishikawa cells. Endocr Relat Cancer 27, 55-66 (2020). https://doi.org/10.1530/ERC-19-0291 15 Chantalat, E. et al. Estrogen Receptors and Endometriosis. Int J Mol Sci 21 (2020). https://doi.org/10.3390/ijms21082815 16 Pellegrini, C. et al. The expression of estrogen receptors as well as GREB1, c -MYC, and cyclin D1, estrogen-regulated genes implicated in proliferation, is increased in peritoneal endometriosis. Fertil Steril 98, 1200-1208 (2012). https://doi.org/10.1016/j.fertnstert.2012.06.056 17 Matsuzaki, S. et al. Expression of estrogen receptor alpha and beta in peritoneal and ovarian endometriosis. Fertil Steril 75, 1198-1205 (2001). https://doi.org/10.1016/s0015-0282(01)01783-6 18 Xue, Q. et al. Promoter methylation regulates estrogen receptor 2 in human endometrium and endometriosis. Biol Reprod 77, 681-687 (2007). https://doi.org/10.1095/biolreprod.107.061804 19 Dyson, M. T. et al. Genome-wide DNA methylation analysis predicts an epigenetic switch for GATA factor expression in endometriosis. PLoS genetics 10, e100 4158 (2014). https://doi.org/10.1371/journal.pgen.1004158 ARTICLE IN PRESS ARTICLE IN PRESS 20 Fujimoto, J., Hirose, R., Sakaguchi, H. & Tamaya, T. Expression of oestrogen receptor -alpha and - beta in ovarian endometriomata. Mol H um Reprod 5, 742 -747 (1999). https://doi.org/10.1093/molehr/5.8.742 21 Colgrave, E. M. et al. Extensive heterogeneity in the expression of steroid receptors in superficial peritoneal endometriotic lesi ons. Reprod Biomed Online 48, 103409 (2024). https://doi.org/10.1016/j.rbmo.2023.103409 22 Brandenberger, A. W. et al. Oestrogen receptor (ER) -alpha and ER -beta isoforms in normal endometrial and e ndometriosis-derived stromal cells. Mol Hum Reprod 5, 651 -655 (1999). https://doi.org/10.1093/molehr/5.7.651 23 Trukhacheva, E. et al. Estrogen receptor (ER) beta regulates ERalpha expression in stroma l cells derived from ovarian endometriosis. J Clin Endocrinol Metab 94, 615 -622 (2009). https://doi.org/10.1210/jc.2008-1466 24 Han, S. J. et al. A new isoform of steroid receptor coactivator-1 is crucial for pathogenic progression of endometriosis. Nat Med 18, 1102-1111 (2012). https://doi.org/10.1038/nm.2826 25 Monsivais, D. et al. ERbeta- and prostaglandin E2 -regulated pathways integrate cell proliferation via Ras-like and estrogen -regulated growth inhibitor in endometriosis. Mol Endocrinol 28, 1304 - 1315 (2014). https://doi.org/10.1210/me.2013-1421 26 Han, S. J. et al. Estrogen Receptor beta Modulates Apoptosis Complexes and the Inflammasome to Drive the Pathogenesis of Endometriosis. Cell 163, 960 -974 (2015). https://doi.org/10.1016/j.cell.2015.10.034 27 Andersson, S. et al. Insufficient antibody validation challenges oestrogen receptor beta research. Nat Commun 8, 15840 (2017). https://doi.org/10.1038/ncomms15840 28 Suzuki, T. et al. Immunohistochemical distribution of progesterone, androgen and oestrogen receptors in the human ovary during the menstrual cycle: relationship to expression of steroidogenic enzymes. Hum Reprod 9, 1589 -1595 (1994). https://doi.org/10.1093/oxfordjournals.humrep.a138757 29 Fonseca, M. A. S. et al. Single-cell transcriptomic analysis of endometriosis. Nat Genet 55, 255-267 (2023). https://doi.org/10.1038/s41588-022-01254-1 30 Garcia-Alonso, L. et al. Mapping the temporal and spatial dynamics of the human endometrium in vivo and in vitro. Nat Genet 53, 1698-1711 (2021). https://doi.org/10.1038/s41588-021-00972-2 31 Nelson, A. W. et al. Comprehensive assessment of estrogen receptor beta antibodies in cancer cell line models and tissue reveals critical limitations in reagent specificity. Mol Cell Endocrinol 440, 138- 150 (2017). https://doi.org/10.1016/j.mce.2016.11.016 32 Chantalat, E. et al. The AF-1-deficient estrogen receptor ERalpha46 isoform is frequently expressed in human breast tumors. Breast Cancer Res 18, 123 (2016). https://doi.org/10.1186/s13058-016- 0780-7 33 Fujimoto, R. et al. Androgen receptors, 5 alpha -reductase activity and androgen -dependent proliferation of vascular smooth muscle cells. J Steroid Biochem Mol Biol 50, 169-174 (1994). 34 Bulun, S. E. et al. Role of estrogen receptor -beta in endometriosis. Semin Reprod Med 30, 39-45 (2012). https://doi.org/10.1055/s-0031-1299596 35 Arnal, J. F. et al. Estrogen receptors and endothelium. Arterioscler Thromb Vasc Biol 30, 1506-1512 (2010). https://doi.org/10.1161/ATVBAHA.109.191221 36 Arnal, J. F. et al. Membrane and Nuclear Estrogen Receptor Alpha Actions: From Tissue Specificity to Medical Implications. Physiol Rev 97, 1045 -1087 (2017). https://doi.org/10.1152/physrev.00024.2016 37 Tamura, M., Deb, S., Sebastian, S., Okamura, K. & Bulun, S. E. Estrogen up-regulates cyclooxygenase- 2 via estrogen receptor in human uterine microvascular endothelial cells. Fertil Steril 81, 1351-1356 (2004). https://doi.org/10.1016/j.fertnstert.2003.09.076 ARTICLE IN PRESS ARTICLE IN PRESS 38 Greaves, E., Collins, F., Critchley, H. O. & Saunders, P. T. ERbeta -dependent effects on uterine endothelial cells are cell specific and mediated via Sp1. Hum Reprod 28, 2490 -2501 (2013). https://doi.org/10.1093/humrep/det235 39 Cooke, P. S. et al. Stromal estrogen receptors mediate mitogenic effects of estra diol on uterine epithelium. Proc Natl Acad Sci U S A 94, 6535 -6540 (1997). https://doi.org/10.1073/pnas.94.12.6535 40 Shao, R., Cao, S., Wang, X., Feng, Y. & Billig, H. The elusive and controversial r oles of estrogen and progesterone receptors in human endometriosis. Am J Transl Res 6, 104-113 (2014). 41 Marquardt, R. M., Kim, T. H., Shin, J. H. & Jeong, J. W. Progesterone and Estrogen Signaling in the Endometrium: What Goes Wrong in Endometriosis? Int J Mol Sci 20 (2019). https://doi.org/10.3390/ijms20153822 42 Patel, B. et al. Role of nuclear progesterone receptor isoforms in uterine pathophysiology. Hum Reprod Update 21, 155-173 (2015). https://doi.org/10.1093/humupd/dmu056 43 Flores, V. A., Vanhie, A., Dang, T. & Taylor, H. S. Progesterone Receptor Status Predicts Response to Progestin Therapy in Endometriosis. J Clin Endocrinol Metab 103, 4561 -4568 (2018). https://doi.org/10.1210/jc.2018-01227 44 Marshall, E. et al. In silico analysis identifies a novel role for androgens in the regulation of human endometrial apoptosis. J Clin Endoc rinol Metab 96, E1746 -1755 (2011). https://doi.org/10.1210/jc.2011-0272 45 Nilsson, S., Koehler, K. F. & Gustafsson, J. A. Development of subtype-selective oestrogen receptor- based therapeutics. Nat Rev Drug Discov 10, 778-792 (2011). https://doi.org/10.1038/nrd3551 46 Davezac, M. et al. The different natural estrogens promote endothelial healing through distinct cell targets. JCI insight 8 (2023). https://doi.org/10.1172/jci.insight.161284 47 Patino-Garcia, D. et al. Estetrol Increases Progesterone Genetic Response without Triggering Common Estrogenic Effects in Endometriotic Cell Lines and Primary C ultures. Biomedicines 11 (2023). https://doi.org/10.3390/biomedicines11041169 48 Zieba, A. et al. The Human Endometrium -Specific Proteome Defined by Transcriptomics and Antibody-Based Profiling. OMICS 19, 659-668 (2015). https://doi.org/10.1089/omi.2015.0115 49 Escande, A. et al. Evaluation of ligand selectivity using reporter cell lines stably expressing estrogen receptor alpha or beta. Biochem Pharmacol 71, 1459 -1469 (2006). https://doi.org/10.1016/j.bcp.2006.02.002 ARTICLE IN PRESS ARTICLE IN PRESS Table 1: Spearman’s correlation coefficients between ERα/PR, ERα/Ki-67 and ERα/AR expressions across tissue type, distinguishing epithelial and stromal compartments – ND: Not determined Color scale indicates the strength and direction of the correlation: dark blue represents strong positive correlations, light blue moderate p ositive correlations, grey weak correlations, and red negative correlations. ND: not determined. Tissu Epithelium / Stroma Correlation ERα / PR Correlation ERα / Ki-67 Correlation ERα / AR Endometrium (EUT) Epithelium 0.85 (P<0.0001) 0.8 (P<0.0001) 0.56 (P<0.001) Stroma 0.88 (P<0.0001) 0.47 (P<0.05) 0.85 (P<0.0001) Superficial Lesion (SUP) Epithelium 0.91 (P<0.0001) 0.6 ns 0.55 ns Stroma 0.72 (P<0.05) 0.3 ns 0.78 (p<0.05) Deep Lesion (DIE) Epithelium 0.43 (P<0.05) -0.04 0.3 ns Stroma 0.54 (P<0.01) 0.07 0.67 (P<0.001) Endometrioma (OMA) Epithelium 0.8 ns 0.8 ns ND Stroma 0.7 ns -0.051 0.5 ns ARTICLE IN PRESS ARTICLE IN PRESS Figure Legends: Figure 1. Predominant ESR1 expression with limited ESR2 transcript detection in endometrium and endometriotic lesions (A) Representative haematoxylin and eosin (H&E)-stained tissue microarray (TMA) spot prepared from a paraffin-embedded deep endometriotic lesion. (B-D) Fluorescent RNAscope hybridization for ESR2 (pink dots) and ESR1 (green dots) transcripts with DAPI (grey) In (B) ovarian follicle, (C) ovarian lesion and (D) endometrium from healthy women (HW) and women with endometriosis (EMS), together with matched SPE and DE lesions in the proliferative (upper panels) and secretory (lower panels) phases. Right panels show higher magnification of the boxed regions in the left panels. In B, the theca (T) and granulosa (G) cells are delineated. Scale bar, 50 µm for full images and 25 µm for insets. (E-F) Quantification of ESR1 and ESR2 mRNA transcripts using the histoscore (H-score) method in epithelial (upper panels) and stromal (lower panels) compartments. (E) Endometrium from HW (n = 6 in P and n = 5 in S) and EMS (n = 9 in P and n = 4 in S) women. (F) SPE (n = 3 in P and n = 4 in S) and DE (n = 6 in P and n = 5 in S) lesions. Data are presented as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 by two-way ANOVA. P, proliferative phase, S, secretory phase. (G) Representative fluorescent in situ hybridisation for ESR2 (yellow) with DAPI (white) in endometrium (upper panel), together with anti-ERG immunostaining on a serial section. Scale bar, 50 µm (H-I) UMAP visualisation of major cell types identified using single -cell RNA sequencing (left panels) by Fonseca et al. (2023) (H) and Garcia-Alonso et al. (2023) (I), showing ESR1 (middle panels) and ESR2 (right panels) mRNA expression in endometrium and lesions of women with endometriosis. (J-K) Mean ESR1 and ESR2 transcript expression in epithelial cells across the menstrual cycle (J) and across the major cell types analysed (K) using datasets from Fonseca et al. (2023). Figure 2. ERβ protein is expressed at low levels in endometriosis lesions, including ovarian lesions (A) Immunohistochemical staining of MDA -MB-231 cells transfected with ESR2 cDNA or control vector, used as a positive control, with two anti-ERβ antibodies (PPZ0506 and CWKF12). Scale bar, 50 µm. (B) Left panel: fluorescent RNAscope hybridization of an ovarian follicle used as a positive control, showing ESR2 (red dots) and ESR1 (green dots) transcripts with DAPI (blue). Middle and right panels: immunohistochemical staining of an ovarian follicle using anti-ERα antibody (SP1) and anti-ERβ antibodies ARTICLE IN PRESS ARTICLE IN PRESS (CWKF12, PPZ0506) as indicated. The lower panels shows higher magnification of the boxed region. Brown indicates positive immunohistochemical staining a nd blue indicates haematoxylin counterstain. Arrows indicate examples of cells with positive nuclear staining. The theca (T) and granulosa (G) cells are delineated. Scale bar, 50 µm for full images and 25 µm for insets. (C) Representative images of immuno histochemical staining of endometrium from women with endometriosis (EMS) or without endometriosis (HW), and of SPE, DE and OMA lesions, using anti -ERβ antibody (PPZ0506). Insets show higher magnification of the boxed regions. Scale bar, 50 µm. Arrows indicate examples of cells with positive nuclear staining. Scale bar, 50 µm for full images and 25 µm for insets. (D) Quantification of ERβ-positive cells (%) in epithelial and stromal compartments of endometrium from women with endometriosis (EMS; n = 10 in P, n = 4 in ES and n = 4 in LS) or without endometriosis (HW; n = 6 in P, n = 3 in ES and n = 6 in LS) across the menstrual cycle (left panels), and in epithelial and stromal compartments of SPE (n = 4 in P and n = 4 in S), DE (n = 13 in P and n = 9 in S) and OMA (n = 1 in P and n = 4 in S) lesions compared with endometrium from women with endometriosis (right panels). P, proliferative phase; ES, early secretory phase; LS, late secretory phase and S, secretory phase (combined ES + LS). Data are presented as mean ± SEM. Statistical differences between groups were assessed by two- way ANOVA. Figure 3. Hormone receptor expression and proliferation are largely unchanged in endometrium from women with or without endometriosis, except for stromal PR in the late secretory phase (A) Representative images of immunohistochemical staining of serial endometrial sections from women without endometriosis (left panels) or with endometriosis (right panels) using anti-ERα (SP1), anti-PR (1E2) and anti-Ki-67 (30-9) antibodies during the proliferative (P), early secretory (ES) and late secretory phase (LS). Scale bar, 50 µm. (B) Quantification of ERα, PR and Ki -67-positive cells (%) in epithelial (l eft panels ) and stromal ( right panels) compartments of endometrium from women without endometriosis (HW; n=15 including n=6 in P, n=3 in ES, n=6 in LS) or with endometriosis (EMS; n=21 or 22 including n=14 for ER α or Ki-67 or n=15 for PR in P, n=7 in ES, and n=8 in LS). Data are depicted by mean ± SEM. Statistical differences between groups were assessed by 2-way ANOVA with Tukey’s multiple comparison test. * P<0.05; ** P<0.01; ***P<0.001; **** P<0.0001 Figure 4. ERα remains the predominant estrogen receptor in endometriotic lesions (A) Representative images of immunohistochemical staining of matched eutopic endometrium (EUT) and lesions (SPE, DE, OMA) from the same patients with EMS during the proliferative (left panels) and secretory (right panels) phases, using anti-ERα (SP1), anti-PR (1E2) and anti-Ki-67 (30-9) antibodies. Scale bar, 50 µm. (B) Quantification of ERα, PR and Ki -67-positive cells (%) in epithelial (upper panels) and stromal (lower panels) compartments of eutopic endometrium (EUT), SPE, DE and OMA lesions from women with endometriosis across the proliferative (P) and secretory (S) phases. Data are presented as mean ± SEM. Statistical differences between groups were assessed by two -way ANOVA. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Sample sizes were as follows: EUT, n = 14 for ERα and Ki-67 and n = 15 for PR in P, and n = 14 for all markers in S; SPE, n = 5 for ERα and PR and n = 14 for Ki-67 in P, and n = 5 for all markers in S; DE, n = 5 for ERα and ARTICLE IN PRESS ARTICLE IN PRESS PR and n = 14 for Ki-67 in P, and n = 16 for ERα and Ki-67 and n = 17 for PR in S; OMA, n = 3 for all markers in P and n = 4 for all markers in S. Figure 5. Paired comparison of ERα, PR, Ki -67 and PR expression between endometriotic lesions (SPE, DE and OMA) and matched eutopic endometrium from women with endometriosis Lesions were compared with matched eutopic endometrium obtained from the same patient. (A-H) Left panels: comparison of ERα, PR, Ki-67 and AR-positive cells (%) in epithelial (A,C,E,G) and stromal (B,D,F,H) compartments between eutopic endometrium and matched SPE lesions (n = 8 for ER α and PR; n = 6 for Ki -67; n = 7 for AR), DE lesions (n = 34 for ER α; n = 35 for PR; n = 32 for Ki -67; n = 33 for AR) or OMA lesions (n = 6 for ER α, PR and Ki -67; n = 7 for AR). Data are shown as individual paired values presented as mean ± SEM. Right panels: difference in the percentage of ER α, PR, Ki -67 and AR -positive cells in epithelial and stromal compartments between lesions and matched eutopic endometrium from the same patient (lesion − EUT). Number of paired samples: SPE, n = 8; DE, n = 22; OMA, n = 4. Data are shown as individual paired values, with paired differences presented as mean ± SEM. Positive values indicate higher expression in lesions relative to matched eutopic endometrium. Statistical differences were assessed by two -sided Wilcoxon signed -rank test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. ARTICLE IN PRESS 0.2 0.0 0.1 0.3 P S P S P S P S 0 100 200 300 Hscore P S P S P S P S 0 100 200 300 Hscore Eutopic endometrium and lesions (EMS) AProliferativeSecretory Endometrioma Nucleus / ESR1 / ESR2 Nucleus / ESR1 / ESR2 Ovarian follicle Hscore: Granulosa cells ESR1: 139 &ESR2: 111 Hscore: Epithelium (ESR1: 132 & ESR2:0) Stroma (ESR1: 36 and ESR2: 3) B D E I C F Superficial lesion (SPE) Nucleus / ESR1 / ESR2 Deep lesion (DE) Nucleus / ESR1 / ESR2 Endometrium (EUT) Nucleus / ESR1 / ESR2 Endometrium Nucleus / ESR1 / ESR2 Heathly woman (HW) J K ESR1 ESR2 OMAEUT 0.000 0.005 0.010 0.015 0.0 0.1 0.2 0.3 0.4 0.5 mean_expression Menstrual.Cycle Proliferative Secretory ECT EUT ECT OMA H Endometriotic woman (EMS) Endometrium (HW vs EMS) SPEEUT DE G Epithelium ESR2ESR1 mean_expression ESR2ESR1 Stromal compartement ESR2ESR1 EMSHWn EpitheliumStromal compartement Phase : **** ; Statut : ns; Interaction : ns Phase : ns ; Statut : ns; Interaction : ns Phase : *** ; Statut : ns; Interaction : ns Phase : ns ; Statut : *; Interaction : ns P S P S 0 5 10 15 20 Hscore P S P S 0 5 10 15 20 Hscore P S P S P S P S P S P S 0 100 200 300 Hscore P S P S P S P S P S P S 0 100 200 300 Hscore Phase : **** ; Tissue : ns; Interaction : ns Phase : ns ; Tissue : ns; Interaction : ns Phase : ** ; Tissue : ns; Interaction : ns Phase : ns ; Tissue : ns; Interaction : ns P S P S P S 0 5 10 15 Hscore P S P S P S 0 5 10 15 Hscore T G Nucleus / ERG Nucleus / ESR2 50µm 50µm 25µm25µm 50µm 50µm 50µm 50µm 50µm 50µm 50µm 50µm 25µm 25µm 25µm 25µm 25µm 25µm 25µm 25µm ARTICLE IN PRESS ARTICLE IN PRESS ESR1 / ESR2 ERα ERβ (PPZ0506)ERβ (CWK-F12) T G T G T G T G T G T G P ES LS P ES LS 0 2 4 6 8 10 % ERβ+ cells P ES LS P ES LS 0 2 4 6 8 10 % ERβ+ cells P S P S P S P S 0 2 4 6 8 10% ERβ+ cells P S P S P S P S 0 2 4 6 8 10% ERβ+ cells BA Ovarian lesion (OMA) Deep lesion (DE) Endometrium Superficial lesion (SPE) Endometrium C Endometriotic woman (EMS)Heathly woman (HW) MDA-MB-231 ERβ+ MDA-MB-231 ERβ- ERβ (PPZ0506) ERβ (CWK-F12) D Epithelium Stromal compartment Stromal compartmentEpithelium Endometrium EUT SPE DE OMA Eutopic endometrium and endometriotic lesions HW EMS Phase : ns ; Statut : ns; Interaction : ns Phase : ns ; Statut : ns; Interaction : ns Phase : ns ; Statut : ns; Interaction : ns Phase : ns ; Statut : ns; Interaction : ns HW EMS EUT SPE DE OMA 25µm 25µm 25µm 25µm 25µm 25µm 25µm 25µm 25µm ARTICLE IN PRESS ARTICLE IN PRESS P ES LS P ES LS 0 25 50 75 100 % ERα+ cells P ES LS P ES LS 0 25 50 75 100 % PR+ cells P ES LS P ES LS 0 25 50 75 100 % Ki-67+ cells P ES LS P ES LS 0 25 50 75 100 % ERα+ cells P ES LS P ES LS 0 25 50 75 100 % PR+ cells ✱✱ ✱✱✱ ✱✱✱ P ES LS P ES LS 0 25 50 75 100 % Ki-67+ cells Proliferative (P) Early secretory (ES) Late secretory (LS) Healthy woman (HW) Endometriotic woman (EMS) A B Proliferative (P) Early secretory (ES) Late secretory (LS) ERα PR Ki-67 Stromal compartmentEpithelium ERα PR Ki-67 ERα PR Ki-67 HW EMS Phase : **** ; Statut : ns; Interaction : ns Phase : **** ; Statut : ns; Interaction : ns Phase : *** ; Statut : ns; Interaction : ns Phase : ** ; Statut : ns; Interaction : ns Interaction : * Phase : ns ; Statut : ns; Interaction : ns ARTICLE IN PRESS ARTICLE IN PRESS A Proliferative Secretory B ERα PR Ki-67 EUT OMA DE SPE ERα PR Ki-67 Stromal compartment Epithelium ERα PR Ki-67 EUT SPE DE OMA EUT SPE DE OMA P S P S P S P S 0 25 50 75 100 % ERα+ cells P S P S P S P S 0 25 50 75 100 % PR+ cells ✱✱✱ ✱✱ ✱✱✱ ✱✱✱ ✱✱ P S P S P S P S 0 25 50 75 100 % Ki-67+ cells P S P S P S P S 0 25 50 75 100 % ERα+ cells P S P S P S P S 0 25 50 75 100 % PR+ cells P S P S P S P S 0 25 50 100 % Ki-67+ cells Phase : ns ; Tissue : ns; Interaction : ns Interaction : * Phase : p=0.065 ; Tissue : ns; Interaction : ns Phase : ns ; Tissue : *; Interaction : ns Phase : * ; Tissue : * ; Interaction : ns Phase : ns ; Tissue: ns; Interaction : ns ARTICLE IN PRESS ARTICLE IN PRESS ARTICLE IN PRESS 0 20 40 60 80 100 % ERα+ cells ✱ p = 0.0625 -100 -50 0 50 % ERα+ cells lesions - EUT 0 20 40 60 80 100 % ERα+ cells ✱✱✱✱✱ -100 -50 0 50 100 % ERα+ cells lesions - EUT 0 20 40 60 80 100 % PR+ cells ✱ -50 -25 0 25 50 100 % PR+ cells lesions - EUT 0 20 40 60 80 100 % PR+ cells ✱ -50 -25 0 25 50 % PR+ cells lesions - EUT 0 20 40 60 80 100 % Ki-67+ cells ✱✱✱✱p=0.0625 -100 -50 -50 0 50 50 100 % Ki-67+ cells lesions - EUT 0 10 20 30 40 50 50 100 % Ki-67+ cells ✱✱ -40 -20 0 20 40 % Ki-67+ cells lesions - EUT 0 5 10 15 20 50100 % AR+ cells ✱ 0 5 10 10 30 50 % AR+ cells lesions - EUT 0 20 40 60 80 100 % AR+ cells p=0.0625 -50 -25 0 25 50100 % AR+ cells lesions - EUT Epithelium Stromal compartmentA C G E EUT SPE DE OMA SPE - EUT DE - EUT OMA - EUT EUT SPE DE OMA SPE - EUT DE - EUT OMA - EUT B F D H EUT SPE DE OMA SPE - EUT DE - EUT OMA - EUT EUT SPE DE OMA SPE - EUT DE - EUT OMA - EUT EUT SPE DE OMA SPE - EUT DE - EUT OMA - EUT EUT SPE DE OMA SPE - EUT DE - EUT OMA - EUT EUT SPE DE OMA SPE - EUT DE - EUT OMA - EUT EUT SPE DE OMA SPE - EUT DE - EUT OMA - EUT ARTICLE IN PRESS ARTICLE IN PRESS