SCM-198 was kindly gifted by Dr. Zhu Yizhun’s laboratory. Female C57BL/6 mice (6–8 weeks old) were purchased from Shanghai JieSiJie Laboratory Animal Co. Ltd. (Shanghai, China). After 2 weeks of adaptation, the mice were randomly selected as the donors of the EMS model. Donor mice were intraperitoneally injected with 17β estradiol (E2) (#E2758, Sigma, St. Louis, MO, USA) (0.2 µg/g weight) thrice for a week as previously reported (36). Vaginal smears were used to select estrus mice as the recipients of EMS mouse models. As previously described (36), the uteri of donor mice were minced together, and then the tissue debris was intraperitoneally injected into recipient mice (the number of the donor uteri was equal to the number of the recipient mice). Since the ectopic lesions were well developed within 1 week after injection, a 7-day formulation of EMS ectopic lesions was administered in this study as previously described (37).

To investigate the effects of SCM-198 on the pathogenesis of EMS, recipient mice were randomly divided into three groups: the EMS group, the EMS+SCM-198 low-dose group (EMS+SCM-198 L, 7.5 mg/kg), and the EMS+SCM-198 high-dose group (EMS+SCM-198 H, 15 mg/kg). According to the corresponding dose (once daily for a week), a 200-µl aliquot of SCM-198 was intraperitoneally injected into each recipient mouse. The mice in the EMS group were given Phosphate buffer saline (PBS) at the same posology. One week later, all of the mice were sacrificed. The endometriotic tissue, uterus, and peritoneal fluid were collected for subsequent analyses.

Ectopic endometrial tissues of 46 women (aged 22–45 years) with ovarian EMS were obtained via laparoscopic surgery, and normal endometrial samples were collected from 10 healthy women (aged 23–46 years) by uterine curettage at the Obstetrics and Gynecology Hospital of Fudan University. The demographic and obstetrical characteristics of the enrolled participants are summarized in Table 1 . Each sample (at least 500 mg) was collected under sterile conditions. ESCs were isolated according to a previously described method (38–41). This method yields ESCs with more than 95% purity, as confirmed by using immunocytochemical staining of vimentin.

Briefly, the endometrial tissues were minced (2–3-mm pieces) and digested in Dulbecco's Modified Eagle's Medium (DMEM)/F-12 containing collagenase type IV (0.1%, Sigma, USA) for 30 min at 37°C. Then, the dispersed cells were filtered through a 400-mesh wire sieve to remove the undigested tissue pieces containing glandular epithelium. After gentle centrifugation, the supernatant was discarded, and the cells were resuspended in DMEM/F-12 containing 10% fetal bovine serum (Gemini, Calabasas, CA, USA), 100 IU/ml penicillin (Sigma, USA), 100 μg/ml streptomycin (Sigma, USA), and 1 μg/ml amphotericin B (Sangon, Shanghai, China) at 37°C in 5% CO₂. Each clinical sample was an independent source of ESCs. Freshly isolated ESCs were cultured overnight in a 25-cm² flask (Corning, USA) per sample. In the next day, those cells that did not adhere were washed away, and those that adhered were largely stromal cells (2–3 × 10⁶/flask), which could attain 85%–90% fusion. After trypsin digestion, ESCs were seeded into the six-well plate at a density of 3–5 × 10⁵/well for further experiments.

The immunohistochemical sections were kept at 60°C for 2 h. Xylene and gradient alcohol were used to dewax and rehydrate the sections. The sections were incubated with 3% hydrogen peroxide and 5% bovine serum albumin successively to block endogenous peroxidase. Tissue sections were incubated with anti-mouse estrogen receptor (ER)α (#ab32063, Abcam, Cambridge, UK) and progesterone receptor (PR) (#ab101688, Abcam, UK) overnight in a humid chamber at 4°C. The sections were washed thrice with PBS for 5 min each time and covered with peroxidase-conjugated goat anti-rabbit or mouse IgG (#GK500710, Gene Teck, San Francisco, CA, USA) for 30 min. Next, they reacted with 3,3-diaminobenzidine (DAB), and the nucleus was stained with hematoxylin. Finally, the slices were dehydrated in gradient alcohol and xylene and then mounted.

The total proteins of ESCs, mouse uterine tissue, and ectopic lesions were extracted by a radioimmunoprecipitation assay (RIPA) buffer (Beyotime, Shanghai, China) supplemented with protease and phosphatase inhibitors (Sigma, USA). The protein concentration was measured using a BCA (Bicinchoninic Acid) protein assay kit (Beyotime, China). After denaturation, equal amounts of protein were separated via Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and then wet-transferred to polyvinylidene difluoride membranes. Nonspecific binding sites were blocked by incubating the membranes with 5% skim milk in Tris-buffered saline with 0.1% Tween 20 (TBS-T) for 1 h. Next, the membranes were incubated overnight at 4°C with primary antibodies (1:1,000) against aromatase (#14528, CST, Boston, USA), ERα (#ab32063, Abcam, UK), PRB (#ab32085, Abcam, UK), LC3B (#3868, CST, USA), BECN1 (#ab207612, Abcam, UK), Bcl-2 (#2870, CST, USA), Bax (#12105, CST, USA), FN1 (#ab2413, Abcam, UK), vimentin (#5741, CST, USA), α-tubulin (#ab7291, Abcam, UK), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (#10112, Arigo, Taiwan, China). Subsequently, membranes were incubated with appropriate horseradish peroxidase (HRP) -conjugated anti-rabbit (#65351, Arigo, China) or anti-mouse (#65350, Arigo, China) IgG secondary antibodies for 1 h at room temperature. After washing with TBS-T thrice, the immunopositive bands on the blots were visualized on the enhanced chemiluminescence detection system (Merck Millipore, USA) using chemiluminescent HRP substrate (#WBKLS0100, Millipore, Boston, MA, USA).

The corrected expression value of genes or transcriptomes, the corrected value of the fold change, the P-value, and the false discovery rate (FDR) value were obtained by DESeq2. We considered transcripts as differentially expressed if the P-value was <0.05 and the fold change was either >1.2 or <0.83333. The GO (Gene Ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) databases were used for functional enrichment and pathway enrichment, respectively. Bubble charts and volcano plot were produced using the ggplot and cluster profiler packages of R version 4.0.4.

Total RNA was extracted using TRIzol reagent (Invitrogen, CA, USA) and then reverse-transcribed into first-strand complementary DNA (cDNA) (Takara, Kyoto, Japan) per the manufacturer’s instructions. The synthesized cDNA was amplified using the ABI PRISM 7900 Sequence Detection System (Applied Biosystems, CA, USA) with specific primers and SYBR Green (Takara, Japan). Triplicate samples were examined for each condition. A comparative threshold cycle value was normalized for each sample using the 2^-ΔΔ Ct method.

The supernatants of normal endometrial stromal cells (nESCs) and differently treated eESCs were harvested and assayed by ELISA per the manufacturer’s instructions (estrogen, #CSB-E07286h, CUSABIO, Shanghai, China; TNF-α, #BDEL-0049, Biodragon, Beijing, China) to detect the secretion levels of estrogen and TNF-α.

The aromatase overexpression (Aromatase^over) plasmid and negative control plasmid were purchased from Shanghai Genechem Co., Ltd. (Shanghai, China). Aromatase siRNA (siAromatase) and control siRNA were purchased from Shanghai Genepharma Co., Ltd. (Shanghai, China). The Aromatase^over plasmid and negative control plasmid (Ctrl) were transfected into eESCs by liposome transient transfection when the fusion degree reached approximately 70%–80% in a six-well plate. Transfected cells were incubated at 37°C for 24 h and then collected for further study. The transfection process of aromatase-silencing (siAromatase) was similar to that of the overexpressed aromatase transfection.

Prism 8 (GraphPad) was used for data analysis. Statistical significance was determined by either Student’s t-test for two-group analyses or the one-way ANOVA for multiple group comparisons. Continuous data were presented as the mean ± SD. The threshold for statistical significance was set at P < 0.05.

Firstly, we used mouse models to investigate whether SCM-198 could alleviate the development of EMS. Figure 1A illustrates the general process of establishing EMS mouse models. As shown in Figures 1B , C, SCM-198 significantly decreased the weights and sizes of mouse ectopic lesions. However, we found no significant difference in the number of ectopic lesions in EMS mice treated with or without SCM-198 ( Figure 1C ). Then, we separated whole ectopic lesions from mice and stained the sections with hematoxylin and eosin (H&E). Through microscopic observation, we found that the outer layers of ectopic lesions were coated with fibrotic tissue. We then measured the thickness of the surrounding fibrotic tissue under the microscope and found that SCM-198 reduced the wall thickness of EMS lesions ( Figure 1D ). In addition, Masson staining revealed that SCM-198 significantly reduced collagen accumulation in ectopic lesions ( Figure 1E ). Western blotting results revealed that SCM-198 inhibited the expression of antiapoptotic protein Bcl-2 and promoted the expression of proapoptotic protein Bax in ectopic lesions at both low and high doses ( Figure 1F ). In line with in vivo results, in vitro analyses revealed that SCM-198 inhibited Bcl-2 and promoted Bax expression in human eESCs ( Figure 1G ). Meanwhile, the levels of fibrosis-related molecules such as fibronectin 1 (FN1) and vimentin were also reduced in human eESCs after SCM-198 treatment ( Figure 1H ). These results suggest that SCM-198 is able to accelerate apoptosis and attenuate the growth and fibrosis of EMS both in vivo and in vitro.

To investigate the underlying mechanism of SCM-198 in restraining EMS, we performed RNA-seq in ectopic lesions of EMS mice that were either treated with SCM-198 or not. We observed a total of 1,616 differentially expressed genes, with 701 genes being upregulated and 915 genes being downregulated in SCM-198-treated ectopic lesions ( Figure 2A ). GO enrichment and KEGG pathway analyses revealed that SCM-198 reduced the levels of autophagy inhibitor molecules and inhibited the ER pathway in ectopic lesions ( Figures 2B, C ).

To confirm the results of the bioinformatics analysis, we first analyzed the expression of autophagy-related genes in ectopic lesions. The results presented in Figure 2D show that SCM-198 could extensively promote the mRNA expression of autophagy-related proteins such as Map1lc3b, Becn1, Ulk1, Atg3, Atg4b, Atg5, Atg7, Gabarap, Atg9a, and Atg10 in ectopic lesions. Furthermore, Western blotting results confirmed that SCM-198 could promote autophagy by increasing the ratio of LC3B-II/I and BECN1 expression ( Figure 2E ). Meanwhile, SCM-198 reversed the imbalance of ERα and PR in EMS ectopic lesions by upregulating PR and downregulating ERα expressions ( Figures 2F–H ). However, the expression of ERβ was not significantly decreased by SCM-198 ( Figure 2F ). These results indicate that SCM-198 could promote autophagy and reverse the imbalance of ERα/PR in EMS.

Then, we assessed the levels of estrogen, hormone receptors, and autophagy in eESCs from EMS patients. Higher production of estrogen ( Figure 3A ) and upregulated ERα ( Figure 3B ) were observed in eESCs. Compared with nESCs, LC3B-II/I and BECN1 were downregulated in eESCs, indicating a lower autophagy level in eESCs ( Figure 3C ). To explore the relationship between estrogen signaling and autophagy, eESCs were treated with E2. The results showed that E2 treatment dose-dependently increased ERα and inhibited autophagy by reducing LC3B-II/I and BECN1 ( Figure 3D ). Previous studies have demonstrated that ERα inhibited autophagy in eESCs (10). Thus, high local estrogen production led to an increase in ERα, which further inhibited autophagy. Progesterone resistance in the ectopic endometrium is mainly mediated by the decrease in PRB (the isoform of PR). Therefore, we focused on the effect of SCM-198 on the expression of PRB in eESCs and found that PRB was decreased in eESCs ( Figure 3E ). Progesterone increased PRB expression and promoted autophagy in a dose-dependent manner ( Figure 3F ). In addition, PR silencing downregulated the autophagy of eESCs by decreasing LC3B-II/I and BECN1, implying that the decrease in PR contributed to the hypo-autophagy state of eESCs ( Figure 3G ). Together, these results suggest that high local estrogen levels lead to increased ERα expression, and the ERα/PRB imbalance in ectopic lesions promotes hypo-autophagy in eESCs.

To investigate whether SCM-198 could promote autophagy by inhibiting estrogen signaling, we treated eESCs with SCM-198. The results showed that SCM-198 downregulated the estrogen level and ERα expression in a dose-dependent manner ( Figures 4A, B ) and enhanced autophagy by upregulating LC3B-II/I and BECN1 levels of eESCs. Meanwhile, no significant change in the expression of ERβ was detected ( Figure 4B ). Importantly, augmented ERα expression and the inhibited autophagy induced by E2 were reversed by SCM-198 in eESCs ( Figure 4C ). In addition, SCM-198 dose-dependently upregulated PRB expression ( Figure 4D ) and reversed the inhibitory autophagy mediated by PR silencing ( Figure 4E ). Furthermore, by using autophagy inhibitor 3-MA (3-Methyladenine), we proved that low autophagy levels were conducive to the antiapoptosis of ESCs ( Figure 4F ). Also, SCM-198 exerted proapoptotic effects on eESCs by promoting autophagy ( Figure 4F ).

Together, these results imply that SCM-198 can enhance autophagy by inhibiting the estrogen pathway and promoting PRB expression, which promotes the apoptosis of eESCs.

Disordered inflammation and endocrine factors promoted the growth of ectopic lesions in EMS. To study the association between inflammation and endocrine signals in EMS, we first detected the level of the pro-inflammation cytokine TNF-α and the expression of aromatase (a key enzyme of estrogen production) in eESCs. As shown in Figure 5A , the mRNA expression and concentration of TNF-α were significantly increased and the expression of aromatase was also upregulated in eESCs ( Figure 5B ). Next, eESCs were treated with TNF-α or R-7050, a tumor necrosis factor receptor (TNFR) antagonist. TNF-α significantly promoted estrogen signaling by increasing aromatase and ERα levels ( Figure 5C ) and elevating estrogen concentration ( Figure 5D ). Although TNF-α and R7050 had no significant effect on the expression of ERβ, TNF-α inhibited the expression of PRB, suggesting that it aggravated the endocrine disorder by inhibiting progesterone signaling ( Figure 5C ). While aromatase silencing decreased estrogen levels and ERα expression ( Figures 5E, F ), its overexpression had the opposite effect ( Figures 5G, H ). The results indicated that aromatase-estrogen signaling positively regulated ERα. To demonstrate whether TNF-α positively regulates estrogen-ERα signaling through aromatase, we silenced aromatase under the treatment of TNF-α. As shown in Figures 5I , J , aromatase silencing wiped out the promotive effect of TNF-α on estrogen production and ERα expression. These data indicate that TNF-α can upregulate the aromatase-estrogen-ERα pathway and reduce PRB expression. Inflammatory disorders can promote an imbalance of estrogen and progesterone signals and, thus, accelerate the development of EMS.

To investigate whether SCM-198 can promote autophagy by inhibiting the TNF-α-mediated imbalance of estrogen and progesterone signals, we treated eESCs with different concentrations of SCM-198. SCM-198 significantly decreased the concentration of TNF-α ( Figure 6A ) and the expression of aromatase ( Figure 6B ). Furthermore, SCM-198 ameliorated aromatase-estrogen-ERα signaling and augmented PRB levels by inhibiting TNF-α ( Figures 6B–D ). The antiapoptotic effect mediated by the TNF-α-estrogen/progesterone signaling-low autophagy axis could be abated by SCM-198 ( Figures 6C, D ). These results suggest that inflammation suppresses autophagy via estrogen and progesterone signaling, thereby inhibiting the apoptosis of eESCs. SCM-198 could restore the balance of estrogen and progesterone signaling by reducing TNF-α and, eventually, promote autophagy and accelerate the apoptosis of eESCs ( Figure 7 ).