1-Methyl-tryptophan attenuates regulatory T cells differentiation due to the inhibition of estrogen-IDO1-MRC2 axis in endometriosis

The percentage of T reg cells in peritoneal fluid is higher in patients with endometriosis than in healthy women. 12 As shown in Figure 1a and b , the proportion of peritoneal T reg cells in EMS stage III/IV patients was significantly higher than that in EMS stage I/II patients and healthy women. TGF- β 1 expression in peritoneal T reg cells also showed a similar trend ( Figure 1c and d ); however, IL-10 expression in T reg cells did not (data not shown). These data demonstrated that the percentage of peritoneal T reg cells increases with exacerbation of endometriosis. To verify whether IDO1 influences the differentiation of T reg cells in the ectopic lesion, we established co-culture systems to mimic the microenvironment of the peritoneal cavity ( Supplementary Figure 1 ) and detected T reg cell differentiation and expression of functional molecules in T reg cells when treated with 1-MT or not. The highest percentage of T reg was found in the naive T cell-macrophage-ESC co-culture system, compared with that in two other co-culture systems ( Figure 2a and b ). Therefore, we used this co-culture system for further analysis. After adding 1-MT, a specific inhibitor of IDO1, to the co-culture system, the differentiation of T reg cells decreased significantly, especially that of IL-10 + T reg cells ( Figure 2c and d ), whereas, TGF- β 1 + T reg cell remained unchanged (data not shown). Moreover, 1-MT-pretreated ESCs-educated T reg performed less suppressive function, as the divided percentage of CD4 + CD25 − (T eff ) cells was higher than that of E+T reg group, while ESCs-educated T reg cells owned a more suppressive function compared with non-pretreated-T reg ( Figure 2e and f ). These results indicate that IDO1 is involved in the differentiation and suppressive function of T reg cells in endometriosis. To investigate whether IDO1 participates in the differentiation of T reg cells in endometriosis in vivo , we established an endometriosis-disease mouse model ( Supplementary Figure 2 ). The results showed that 1-MT distinctly inhibited the percentage of T reg cells in peritoneal fluid of mice ( Figures 2g and h ), especially IL-10 and CD 73 expression in peritoneal T reg cells ( Figures 2i and k ), which were consistent with the in vitro results. Patients with endometriosis show high local estrogen levels. 7 Additionally, IDO1 expression in ectopic ESCs is higher than that in normal ESCs, 6 leading us to consider that estrogen may regulate the expression of IDO1 in the ectopic lesion. We found that IDO1 expression in estrogen-conditioned ESCs and estrogen-conditioned macrophages were obviously higher than that in the control groups ( Figures 3c–f ). Besides, the effect of ESCs on up-regulating the expression of IDO1 in macrophages was more significant than that with estrogen alone ( Figures 3e and f ), which indicates a crosstalk between ESCs and macrophages that related to IDO1 expression. To identify which sub-unit estrogen receptor (ER) that allows estrogen to promote the expression of IDO1 in ESCs, we blocked ER α , ER β , and ER respectively. We found that IDO1 expression in ESCs was down-regulated regardless of whether a single or both sub-unit ERs were blocked. This suggests that both sub-units of ER are involved in this activity, especially ER β ( Figures 3g and h ). Although the percentage of T reg cells in ectopic lesions of the estrogen receptor inhibitor (ER i ) group showed little changes in vivo (data not shown), the percentage of TGF- β 1 + , IL-10 + , CD73 + , and CTLA-4 + T reg cells decreased ( Supplementary Figure 3 ). Considering that endometriosis is an estrogen-dependent disease, and that estrogen enhances Foxp3 expression and T reg cell function, 13 we explored whether estrogen regulates the differentiation of T reg cells in endometriosis. After adding estrogen to the co-culture system, the differentiation of T reg cells notably increased ( Figures 4a and b ), which indicates that local high estrogen level participates in inducing the differentiation of T reg cells in the peritoneal fluid of patients with endometriosis. As mentioned above, estrogen promotes the expression of IDO1 in the ectopic lesion, and IDO1 influences the differentiation of T reg cells, 5 raising the possibility that estrogen may promote the differentiation of T reg cells via up-regulation of IDO1 expression. Compared with that in the estrogen group, the differentiation of T reg cells in estrogen plus 1-MT group was lower ( Figures 4a and b ), which manifested that 1-MT inhibits T reg cell differentiation induced by estrogen. As outlined above, IDO1 influences the differentiation of T reg cells, especially IL-10 + T reg cells. To identify factors that possibly connect IDO1 and IL-10, and are involved in interactions of IL10, IDO1, Foxp3, and TGF- β , we performed a search of the KEGG database as part of the bioinformatics analysis. We found that molecules up-stream of IL-10 include C-type lectin domain family 4 member M (CLEC4M), mannose receptor C, type 1 (MRC1), mannose receptor C, type 2 (MRC2), STAT6, phospholipase A2 receptor1 (PLA2R1), platelet-activating factor receptor (PTAFR), and CD209 ( Figure 5a ). After treating ESCs with estrogen, 1-MT, or estrogen plus 1-MT, only MRC2 expression increased in the estrogen-treated group compared with that in the control group ( Figure 5b ), which was similar to our previous result where estrogen up-regulated IDO1 expression in ESCs ( Figures 3c and d ). Thus, the results demonstrate that 1-MT also promotes the expression of MRC2 in ESCs to a greater extent than estrogen ( Figures 5b and c ), which means that MRC2 is downstream to IDO1 and estrogen. MRC2 is a constitutively recycling endocytic receptor belonging to the mannose receptor family. 19 The expression of MRC2 in ectopic ESCs was significantly lower than that in normal ESCs ( Figure 5d ). Combined with evidence that the expression of IDO1 in ectopic ESCs is higher than that in normal ESCs, 6 and 1-MT up-regulates the expression of MRC2, we reasoned that high levels of IDO1 might lead to low expression of MRC2 in ectopic ESCs. Besides, after silencing MRC2 in ESCs, the expression of IDO1 increased ( Figures 5e and g ), which indicates a negative feedback between MRC2 and IDO1. Therefore, the level of Ki-67 in MRC2-silenced ESCs increased compared with that in the vector group ( Figures 5e and f ), which is consistent with the notion that ectopic ESCs have a stronger ability for proliferation. 6 Similarly, the expression of Ki-67 and IDO1 also increased in ectopic lesions after MRC2 shRNA was intraperitoneal injected to the peritoneal cavity of endometriosis mice model in vivo ( Figures 5h and i ). According to the findings above, MRC2 is downstream to IDO1, and IDO1 is involved in the differentiation of T reg in ectopic lesion, hinting the possibility that MRC2 may participate in the activity that IDO1 regulates the differentiation of T reg in endometriosis. When MRC2-silenced ESCs were co-cultured with naive CD4 + T cells and monocytes-derived macrophages, the percentage of CD4 low T reg and CD4 high T reg cells were higher in the MRC2-silenced group than that in the vector group, especially CD4 high T reg cells ( Figures 6a and b ). Moreover, CD4 high T reg cells from the MRC2-silenced group showed a more immunosuppressive phenotype, with higher expression of TGF- β 1, IL-10, CD39 and CTLA-4, than that in CD4 high T reg cells in the vector group and CD4 low T reg cells in MRC2-silenced group ( Figures 6c and e ). Furthermore, MRC2-silenced ESCs-educated T reg performed stronger suppressive function, as divided percentage of T eff cells was lower than that of vector group, which indicated that MRC2 is involved in T reg suppressive function ( Figures 6f and g ). In vivo , at two weeks after administrating MRC2-shRNA via the peritoneal cavity of the endometriosis-disease mouse model, the percentage of peritoneal T reg cells and T reg cells in ectopic lesions were significantly higher than that in the vector group ( Figures 6h, i, l and m ), and mainly comprised TGF- β 1 + and CTLA-4 + T reg cells in the peritoneal fluid ( Figures 6j and k ). These results demonstrate that MRC2 is not only responsible for the differentiation of T reg cells, mainly that of CD4 high T reg cells; but also promotes T reg immunosuppressive function in endometriosis, which may exacerbate the development of endometriosis. To determine whether 1-MT improves the condition of endometriosis in vivo , we measured the total number and weight of ectopic lesions and found that both parameters decreased notably in the 1-MT administered group compared with that in the control group, as similar as ERi administered group ( Figures 7a, b and d ). In the MRC2-shRNA administered group, the total number of ectopic lesions did not significantly increased ( Figure 7c ), but total weight of ectopic lesions was obviously higher than that in the vector group ( Figure 7e ), which indicates that MRC2 is responsible for ectopic lesion growth while not for the spread of the implant. Moreover, 1-MT notably restrained the proliferation of ectopic lesions in vivo , in terms of lower expression of Ki-67 compared with that in the control group, as seen with the ER i group ( Figures 7f and g ). These results suggest that 1-MT reverses the condition of endometriosis in vivo .

The human ethics committee of Gynecology and Obstetrics Hospital, Shanghai Medical School, Fudan University, People's Republic of China approved this study. Written informed consent was obtained from each patient enrolled. Endometriotic tissues were acquired from premenopausal patients who had undergone laparoscopic surgery for the treatment of ovary endometriosis cysts ( n =55; aged 22–46 years) diagnosed by pathological examination. Control endometrial tissues were obtained from premenopausal patients who had undergone panhysterectomy surgery for multiple myoma ( n =15; aged 46–52 years). Peritoneal fluid was aspirated from the cul de sac at the beginning of the laparoscopic procedure under general anesthesia, obtained from patients with ovary endometriosis ( n =16; aged 25–46 years), or patients who had laparoscopic surgery for uterine fibroid without evidence of endometriosis ( n =12; aged 29–41 years) at the Obstetrics and Gynecology Hospital of Fudan University, between August 2014 and September 2016. Samples of peritoneal fluid contaminated by blood were excluded from the study. None of women had received hormonal therapy within 6 months prior to tissue collection. The stage of endometriosis was diagnosed according to the revised American Society for Reproductive Medicine staging (1997). Peripheral blood samples (5–15 ml) were obtained from healthy volunteers ( n =35; aged 25–38 years old). All samples were collected under sterile conditions. All collected peritoneal fluid samples were centrifuged at 1600 rpm, at 4 °C for 9 min, then supernatant was discarded, and cells were fixed with 4% par formaldehyde (Sheng Gong, Shanghai, China) for 35 min, at 4 °C in the dark. Then cells were washed twice with phosphate-buffered saline (PBS; Hyclone, Logan, UT, USA). After centrifugation and removal of the supernatant, the fixed cells were resuspended in Foxp3 Perm Buffer (10 × , Biolegend, San Diego, USA) according to the manufacturer protocol. Finally, these cells were labeled with flow cytometry antibodies according to the manufacturer protocol and detected by flow cytometry. Endometrial stromal cells (ESCs) were purified as previously described. 37 Endometriotic tissues were cut into<1mm-thick sections and digested with Dulbecco's Modified Eagle Medium (DMEM)/F12 (Hyclone, Logan, UT, USA) containing collagenase type IV (0.1% Sigma, San Francisco, CA, USA) with constant agitation for 40 min at 37 °C. The resulting suspension was then filtered through 100- to 70- μ m nylon strainers (Becton Dickinson, Franklin Lakes, NJ, USA). After the filtrate was centrifuged at 1600 r.p.m. for 9 min at 4 °C, the supernatant was discarded. Finally, ESCs were resuspended in DMEM/F-12 containing 10% fetal bovine serum (FBS; Hyclone) in the presence of 100 U/ml penicillin and 100 mg/ml streptomycin, and placed in culture flasks at 37 °C under 5% CO 2 . The culture medium was replaced with fresh medium every 3 days. Peripheral blood mononuclear cells (PBMCs) were isolated by Lymphoprep (Stemcell Technologies Inc., Vancouver, BC, Canada) density gradient centrifugation. CD14 + cells were obtained by positively selection using CD14 + cells micro-magnetic beads according to the manufacturer instructions (Miltenyi Biotec, Bergisch Gladbach, Germany). Naïve CD4 + T cells were obtained by negatively selection from the remaining immune cells according to the manufacturer instructions (Miltenyi Biotec). The purity of CD14 + cells was confirmed by flow cytometry, using PE/Cy7 anti-human CD14 monoclonal antibody (mAb) (Biolegend, San Diego, USA, clone: HCD14), was found to be>95%. Ectopic ESCs were cultured in 24-well plates (Corning, Steuben County, NY, USA) at a density of 1 × 10 5 cells/well. The co-culture systems were established by incubating 2 × 10 5 monocytes with ESCs or alone, adding of estrogen (10 −8 M; Sigma), 1-MT (0.05 mM; Sigma), or estrogen (10 −8 M; Sigma) plus 1-MT (0.05 mM; Sigma). Meanwhile, naive CD4 + T cells were cultured in 24-well plates that coated with monoclonal anti-CD3 (5  μ g/ml; Biolegend) and monoclonal anti-CD28 (1  μ g/ml; Biolegend), in the presence of recombinant human IL-2 (50 ng/ml; Biolegend). The monocytes-derived macrophages and 2 × 10 5 naive CD4 + T cells were collected to co-culture with ESCs in 1 ml medium/well for 5 days ( Supplementary Figure 1 ). Peripheral blood mononuclear cells (PBMCs) were isolated by Lymphoprep (Stemcell Technologies Inc.) density gradient centrifugation. CD4 + CD25 + T cells were obtained by positively selection using CD4 + CD25 + T cells (Regulatory T cells) micro-magnetic beads according to the manufacturer instructions (Miltenyi Biotec). CD4 + CD25 − cells were obtained by negatively selection according to the manufacturer instructions (Miltenyi Biotec). CD4 + CD25 + T cells were cultured in 96-well plates that coated with monoclonal anti-human CD3 (5  μ g/ml; Biolegend, clone: OKT3) and monoclonal anti-human CD28 (1  μ g/ml; Biolegend, clone: CD28.2), in the presence of recombinant human IL-2 (50 ng/ml; Biolegend), while CD4 + CD25 − cells were cultured in 24-well plates that coated with monoclonal anti-CD3 (5  μ g/ml; Biolegend) and monoclonal anti-CD28 (1  μ g/ml; Biolegend). T reg cells were collected to culture with non-treated ESCs, 1-MT-pretreated ESCs, vector-pretreated ESCs, and MRC2-silenced ESCs or not for 48 h after T reg cells proliferation for two weeks. After that, T reg cells from these groups respectively cultured with paired CFSE-labeled CD4 + CD25 − T cells (T eff ) for 48 h and then detected by flow cytometry. Fluorochrome-conjugated antibodies the following human antigens were used for flow cytometry analysis: Alexa Fluor 488 anti-human Foxp3 mAb (Biolegend, clone: 206D), PE anti-human LAP (TGF- β 1) mAb (Biolegend, clone: TW4-6H10), PE/Cy7 anti-human CD4 mAb (Biolegend, clone: RPA-T4), APC anti-human IL-10 mAb (Biolegend, clone: JES3-9D7), PerCP/Cy5.5 anti-human CD73 (Ecto-5'-nucleotidase) mAb (Biolegend, clone: AD2), Brilliant Violet 421 TM anti-human CD152 (CTLA-4) mAb (BD Biosciences, clone: BNI3), Brilliant Violet 510 TM anti-human CD39 mAb (Biolegend, clone: A1), PE/Cy7 anti-human CD14 mAb (Biolegend, clone:HCD14), APC/Cy7 anti-human CD45 (Biolegend, clone: HI30), Alexa Fluor 488 anti-human Vimentin mAb (BD Biosciences, clone: RV202), PE anti-human Ki-67 mAb (Biolegend, clone: Ki-67), Anti-human Indoleamine2,3-dioxygenase/IDO APC-conjugated mAb (R&D System, Inc., Minneapolis, USA, clone: 700838), CFSE Cell Division Tracker Kit (Biolegend). Additionally, fluorochrome-conjugated antibodies against the following mouse antigens were used for flow cytometry analysis: anti-mouse CD4 fluorescein isothiocyanate (FITC) mAb (eBioscience, clone: GK1.5), FITC anti-mouse Vimentin mAb (BD Biosciences), anti-mouse/rat Foxp3 PE mAb (eBioscience, clone: FJK-16 s), anti-mouse/rat Ki-67 PE mAb (eBioscience, clone: P46013 ), PE/Cy7 anti-mouse IL-10 mAb (Biolegend, clone: JES5-16E3), PerCP/Cy5.5 anti-mouse CD73 mAb (Biolegend, clone: TY/11.8), APC anti-mouse LAP(TGF- β 1) mAb (Biolegend, clone: TW7-16B4), Brilliant Violet 421 TM anti-mouse CD152 mAb (Biolegend, clone: UC10-4B9), and anti-mouse IDO eFluor® 660 (eBioscience, clone: mIDO-48). Surface staining was performed at 4 °C for 30 min. Intracellular staining was performed using Foxp3 Fix/Perm Buffer Set (4 × , Biolegend) according to the manufacturer protocol. The resulting data were analyzed using the LYSYS II software program (Becton Dickson, Franklin Lakes, NJ, USA). 2 × 10 5 monocytes were co-cultured with ectopic ESCs on 24-well plates (Corning) or alone, adding estrogen (10 −8 M; Sigma) or not. After 48 h, ESCs and monocytes-derived macrophages from different groups were collected, and then detected the expression of IDO1 by flow cytometry. Ectopic ESCs were cultured in 24-well plates (Corning) at a density of 1 × 10 5 cells/well. Pretreated ESCs with estrogen receptor α inhibitor (ER αi , 10 −6 M; Sigma), estrogen receptor β inhibitor (ER βi , 10 −6  M; Sigma), estrogen receptors inhibitor (ER i , 10 −6 M; Sigma) or not. After 24 h, inhibitors were washed off by PBS (Ji-Nuo, Hangzhou, China), and then added estrogen (10 −8 M; Sigma) to these groups except control group. After 48 h, the expression of IDO1 in ESCs from different groups was detected by flow cytometry. Total RNA from ESCs treated with estrogen (10 −8  M; Sigma), 1-MT (0.05 mM; Sigma), estrogen (10 −8  M; Sigma) along with 1-MT (0.05 mM; Sigma), or knockdown of MRC2 with shRNA for 48 h was extracted using the Trizol reagent (Life Technologies, Carlsbad, CA, USA), according to the manufacturer instructions. Total RNA (2  μ g) was reverse transcribed into first-stand cDNA (TaKaRa Bio Inc., Japan) following the manufacturer protocol, which was then used as a template for polymerase chain reaction (PCR) amplification. Real-time PCR was performed using ABI PRISM TM 7900 Sequence Detector (Applied Biosystems, Warrington, UK). The primers sequences used are listed in Table 1 . PCRs was carried out for 40 cycles using the following conditions: denaturizing at 95 °C for 30 s, annealing at 95 °C for 5 s, and elongation at 60 °C for 34 s. The expression levels of the samples were expressed as arbitrary units defined by the 2 −ΔΔCT method. All measurements were performed in triplicate. The specificity of the product was assessed by melting curve analysis. Protein expression level of MRC2 and β -actin were measured by western blot as previously described. 38 The primary antibodies used were anti-MRC2 mAb (1:1000; Abcam Cambridge, MA, USA), and β -actin rabbit mAb (1:1000; Cell Signaling Technology, Inc., Danvers, MA, USA). Horseradish peroxidase-conjugated goat anti-rabbit IgG (1:2000; Arigobio, Taiwan, ROC) was used as the secondary antibody. Quantitative analysis of the relative density of the bands obtained by western blot was performed using Image J (W.S. Rasband, National Institutes of Health, Bethesda, MD, USA). Ectopic ESCs were cultured on 24-well plates with DMEM/F-12 (Hyclone) plus 10% FBS (Hyclone) in the presence of 100 U/ml penicillin and 100 mg/ml streptomycin, at 37 °C under 5% CO 2 . Lipofectamine 3000 (Invitrogen; Life Technologies, Carlsbad, CA, USA), OPTI-MEM (Gibco BRL, Gaithersburg, MD, USA), and the pGPU6/GFP/Neo-MRC2 short hairpin RNA (shRNA) (GenePharma, Shanghai, China) were used to transfected ESCs according to the manufacturer protocol. We used seven-week-old female C57B/L6 mice (Slaccas Animal Laboratory, Shanghai, China) to establish an endometriosis-disease model as previously described. 39 , 40 Each donor mouse was administered 200 μ l estrogen (10 −8 M; Sigma) via the peritoneal cavity. One week later, donor mice were sacrificed and the uterine horn were collected and minced. Every two recipient mice were then intraperitoneally injected with minced uterine horn tissue from one donor mouse equally. All procedures were performed under aseptic conditions. One week after uterine tissue injection, mice were randomly divided into five groups and each received an intraperitoneal injection of 200 μ l PBS (Ji-Nuo), estrogen receptors blockers (10 −6  M; Sigma), 1-MT (0.05 mM; Sigma), vector (GenePharma, Shanghai, China), or mouse MRC2 shRNA (GenePharma, Shanghai, China) every week. Vector and mouse MRC2 shRNA were mixed with Lipofectamine 3000 (Invitrogen) and OPTI-MEM (Gibco BRL), according to the manufacturer protocol. Two weeks later, we measured the total number and weight of ectopic lesions, the expression of Ki-67 and IDO1 within ectopic lesions, the percentage and the expression of functional molecules of T reg in total ectopic lesions and peritoneal fluid of mice from groups administered PBS, estrogen receptor blockers, 1-MT, vector and MRC2 shRNA ( Supplementary Figure 2 ). All studies were set up to include three wells per condition, and each experiment was independently repeated more than three times. Data collected from each independent experiment were analyzed using the Graphpad Prism (Graphpad software Inc., La Jolla, CA, USA) statistical package. Paired t -test of variance was performed when appropriate. Differences were considered statistically significant at ±S.D.<0.05.

Although endometriosis is an estrogen-dependent disease, it is not enough to determine the pathogenesis of endometriosis from the endocrine perspective alone. Multiple factors participate in the development and maintenance of endometriosis, including immunological dysfunction, genetic susceptibility, psychological factors, and environment factors. Among these, the immunological aspect of endometriosis has recently been extensively studied. Up to 90% of women of reproductive age exhibit retrograde menstruation, whereas only 6–10% of these women develop endometriosis. 20 This indicates that a microenvironment of immune tolerance is formed within the ectopic lesion in patients with endometriosis, where the function of menstrual debris clearance by the local immune responses is ineffectively. 21 Several types of immune cells are involved in the formation of the local immune tolerance environment in endometriosis, such as T reg cells, macrophages, and natural-killer (NK) cells. Disturbance of T reg cells, which are responsible for self-tolerance, maintenance of immune homeostasis, and immunosuppressive functions, 22 aggravates the condition of endometriosis. In this study, we have demonstrated not only that the percentage of T reg cells in the peritoneal fluid of endometriosis patients higher than that in healthy women, which is consistent with results of previous published studies, 23 but also that the augmentation of T reg cell percentage, especially that of TGF- β 1 + T reg cells, occurs in parallel with endometriosis exacerbation. This suggests that the detection of peritoneal T reg cells may be used as an indicator to assess the severity of endometriosis. However, the origin of excessive T reg cells in the peritoneal fluid of patients with endometriosis is yet unclear. Three assumptions may elucidate this clinic phenomenon. One is homing of peripheral T reg cells to the peritoneal fluid as the percentage of T reg cells decreases in the peripheral blood of EMS patients. 23 The second is that local naive T cells differentiate into T reg cells under the influence of various factors involved; and the third is self-proliferation of local T reg cells. In fact, all of these assumptions may be involved in the increase of peritoneal T reg cell percentage in endometriosis. IDO1 is essential for the generation and function of T reg cells. 14 , 24 , 25 Initially, IDO1 was thought as the first line of host defense against infectious pathogens, as it causes tryptophan shortage, which in turn restricts mammalian cell growth, 26 especially that of T cells. 27 Subsequent studies revealed that IDO1 has multiple immunological functions, such as suppressing T-cell responses, regulating functions of T reg cells, 14 and promoting immune tolerance. 15 Evidence of higher IDO1 expression in ectopic ESCs than in normal ESCs 6 and high percentage of peritoneal T reg cells in endometriosis 23 demonstrates that IDO1 may play an important role in inducing the differentiation of T reg cells in the ectopic lesion, which is a probable reason for excessive T reg cells in the peritoneal fluid of patients with endometriosis. In this study, we established a naive T cell-macrophage-ESC co-culture system to mimic the local microenvironment of the peritoneal cavity. The results of our study show that IDO1 participates in the differentiation of T reg cells in the ectopic lesion, evidenced by inhibition of T reg cell differentiation, especially that of IL-10 + T reg cells, by 1-MT. In addition, 1-MT-pretreated ESCs-educated T reg cells suppressed the proliferation of T eff cells less effectively compared with non-treated ESCs-educated T reg cells, which indicates that 1-MT is involved in T reg cell suppressive function. Moreover, estrogen promoted the expression of IDO1 in both ESCs and macrophages, which, on a side-note, explains that ectopic ESCs have high expression of IDO1 in endometriosis. Therein, the expression of IDO1 in ESCs-educated macrophages was notably higher than that in both estrogen-treated and untreated monocytes-derived macrophages. This demonstrates a crosstalk between ESCs and macrophages, involving IDO1. Besides, estrogen was found to induce T reg cell differentiation in the ectopic lesion and IDO1 was involved in this process, wherein, 1-MT down-regulated T reg cell differentiation induced by estrogen. These results indicate that estrogen regulates the expression of IDO1 in the ectopic lesion to induce the T reg cell differentiation. Based on the finding that 1-MT mainly influences the differentiation of IL-10 + T reg cells; we performed a search of the KEGG database to identify possible molecules that connect IDO1 and IL-10. The results show that MRC2, which is an upstream molecule of IL-10, is essential for the differentiation and function of T reg cells. Most researches on MRC2 to date have focused on its role in the development of cancer, as it can promote breast tumor growth, 17 co-operate with the matrix metalloproteinase to remodel of extracellular matrix that attenuate renal fibrosis, 28 and predict prognosis of hepatocellular carcinoma 18 and prostate cancer. 29 Besides, MRC2 is also closely related to collagen turnover. 30 In our study, we found that both estrogen and 1-MT promoted the expression of MRC2 in ESCs, which indicates that MRC2 is downstream to estrogen and IDO1. Therein, the regulatory ability of 1-MT was stronger than that of estrogen. Combined with the low expression of MRC2 by ectopic ESCs and MRC2 expression promoted by estrogen, we reasoned that other factors might be responsible for decreased MRC2 expression by ESCs. As shown in Figures 5b and c , we observed that 1-MT promoted the expression of MRC2 by ESCs to a greater extent than estrogen, whereas, decreased IDO1 levels caused an increase in MRC2 expression. Therefore, it is a possibility that increased IDO1 levels down-regulate the expression of MRC2 by ESCs. In addition, negative feedback was noted between IDO1 and MRC2, evidenced by increase in IDO1 expression when MRC2 in ESCs was silenced. Meanwhile, MRC2 negatively regulated proliferation of ESCs as the expression of Ki-67 increased in si-MRC2 ESCs compared with vector group. To explore the possibility that MRC2 influences the differentiation of T reg cells, we co-cultured MRC2-silenced ESC with macrophages and naive CD4 + T cells. Compared with that in the vector group, the percentage of T reg cells increased in the MRC2-silenced group, especially CD4 high T reg cells, which demonstrate that MRC2 plays a key role in the differentiation of T reg cells in endometriosis. Therefore, MRC2-silenced ESCs-educated T reg cells acquired stronger suppressive function than vector-pretreated ESCs-educated T reg cells, which indicates that MRC2 negatively regulates T reg cell suppressive function. To our knowledge, this is the first report demonstrating the immunological functions of MRC2 as an important regulator of T reg cell differentiation and function, and showing that estrogen and IDO1 are up-stream to MRC2. These results illustrate that estrogen-IDO1-MRC2 axis is involved in the differentiation of T reg cells in endometriosis. Consistent with the in vitro results above, 1-MT dramatically reversed the condition of endometriosis in vivo , irrespective of total number or weight of ectopic lesions, expression of Ki-67 in total ectopic lesions, or the percentage of peritoneal T reg cells. These data suggest that locally applied 1-MT may clinically relieve sufferings of patients with endometriosis. Besides, MRC2 significantly induced the differentiation of T reg cells, increased weight of total ectopic lesions, and promoted the expression of Ki-67 in ectopic lesions in vivo , which is consistent with the in vitro results, suggesting that MRC2 is involved in the growth of ectopic lesions and T reg differentiation in endometriosis ( Supplementary Figure 4 ). Although endometriosis is a benign gynecological disease, its biological activities are similar to that of cancer, including metastasis, plantation, angiogenesis, immune tolerance, and recurrence. 31 , 32 In the present study, we focused on IDO1, MRC2, and T reg cells, which have been reported in previous cancer studies, 17 , 33 – 35 to identify the endocrine-immune mechanism for the high percentage of peritoneal T reg cells in patients with endometriosis. The results revealed that the percentage of T reg cells in the peritoneal fluid increases as endometriosis progresses. Therein, we found that the lower expression of MRC2 in ectopic ESCs significantly promoted the differentiation and function of T reg cells. This is a novel function of MRC2, identified in this study. Future work would involve identifying probable mechanisms by which MRC2 influences the differentiation of T reg cells, including the transcription factors involved in this process and interactions of MRC2 and T reg functional molecules. Moreover, the possibility that MRC2 influences the biological activities of ESCs in endometriosis remains to be explored. Common treatment options for endometriosis, including progestogens, ovulation induction, GnRH analog and surgery, however, typically do not provide long-term relief. 36 Pharmacotherapy for endometriosis affects the physiological functions of estrogen and this limits its long-term use. In addition, high recurrence rate renders endometriosis clinically intractable. Therefore, to treat endometriosis, it is essential to identify possible targets that do not influence the physiological functions of estrogen. The in vivo results from this study showed that 1-MT, a specific inhibitor of IDO1, can notably improve the condition of endometriosis, as similar as ER inhibitor, which provide evidences that blocking IDO1 in ectopic lesions may be a novel treatment option for endometriosis, not only by virtue of its effects on biological activities of ESCs, but also on local immune-tolerance environment effected by inhibition of the differentiation and functions of T reg cells. To date, 1-MT is undergoing clinical phase II and I trials for application in cancer immunotherapy. 15 Intrauterine devices containing 1-MT can be considered for the treatment of endometriosis. These data indicate that endometriosis is more than a disease involving endocrine disorders, and also involves immunological factors. Moreover, it should be taken into consideration that assessment of the local immunological status of patients with endometriosis is essential for the treatment choice to be made.