NR4A1 is Involved in Fibrogenesis in Ovarian Endometriosis

Endometriosis, which is characterized by the implantation and growth of endometrial glands and stroma on extra-uterine sites, is a common cause of pelvic pain and subfertility in women of reproductive age [1]. Severe symptoms of endometriosis affect 6%–10% of the female population worldwide [2]. Among them, 50%–60% have pelvic pain and up to 50% have infertility [2]. Histologically, ectopic endometrial glands and stroma are encompassed by dense fibrous tissue [3, 4]. Fibrogenesis is an important process in endometriosis and exerts a significant negative impact on patients with the condition. Previous studies demonstrated that excess fibrosis around endometriotic lesions leads to scarring, pain, and impaired tissue function [5, 6]. Furthermore, the ovarian cortex might be involved and consequently the follicle reserve can be reduced by severe fibrosis [7, 8].

Transforming growth factor-beta (TGF-β) is a multifunctional growth factor that plays important roles in the regulation of cell proliferation, angiogenesis, differentiation, and the immune response [9]. It is also believed to be a key molecule in the fibrotic process [10]. Overexpression of TGF-β has been found in a number of fibrocontractive diseases, such as skin scarring [11], pulmonary fibrosis [12], and peritoneal adhesion [13]. On the other hand, blockage of TGF-β suppresses experimentally induced fibrosis in the skin, lungs, and liver [14, 15]. Previous studies have shown that, compared with their healthy counterparts, women with endometriosis have higher levels of TGF-β in serum and peritoneal fluid [16, 17]. This elevated level of TGF-β plays a crucial role in fibrosis in endometriosis [18].

NR4A1 (also known as TR3, Nur77m, or NGF-IB) belongs to the steroid/thyroid hormone receptor superfamily [19]. It is implicated in a wide array of cellular events, such as glucose and lipid metabolism [20, 21], apoptosis [22], and vascular homeostasis [23]. As an immediate early gene, it is subject to alteration by multiple stimuli, including growth factors, inflammatory factors, and stress, in a variety of cells and organs [24]. NR4A1 has been shown to be a strong and important activator of the TGF-β/Smad signaling pathway [25], and it also recruits a repressor complex containing SP1, SIN3A, CoREST, LSD1, and HDAC1 to the TGF-β target gene, thereby forming a negative feedback loop [26]. However, the role of NR4A1 in the fibrosis of endometriosis has not yet been studied.

In this study, we examined the fibrosis markers alpha-smooth muscle actin (ɑ-SMA), type 1 collagen 1 (COL1A1), connective tissue growth factor (CTGF), and fibronectin (FN) and found that inhibition of NR4A1 could promote TGF-β1-induced overexpression of fibrotic markers in human ectopic endometrial stromal cells (EESCs) and normal endometrial stromal cells (NESCs). Our study also indicated that Cytosporone B (Csn-B), a selective agonist that activates NR4A1 [27], could effectively decrease the TGF-β1-induced elevated expression of fibrosis markers in vitro. Animal experiments showed that a lack of NR4A1 promoted the progression of fibrosis and that Csn-B effectively suppressed fibrogenesis in mouse endometriotic tissue. We are led to conclude that NR4A1 may be a therapeutic target for the treatment of fibrosis in endometriosis.

Tissue samples were obtained from patients aged 20–41 years who underwent laparoscopy for ovarian endometriosis and hysteroscopy for primary infertility. None of them had received hormonal treatments in the prior 6 months, and all of them had regular menstrual cycles (26–32 days). Endometriotic tissues were taken from 23 patients (age: median, 31.0 years; range, 20–41 years) who had histopathologically proven endometriosis. In addition, proliferative-phase endometrial tissues were collected from 15 women without endometriosis (age: median, 29.5 years; range, 24–36 years) during the hysteroscopy by using an endometrial curette. Each sample was divided into three portions, with one portion fixed in 4% paraformaldehyde and embedded in paraffin, one frozen at –80°C for western blotting, and one stored in Dulbecco’s modified Eagle’s medium (DMEM)/F-12 (Hyclone, Logan, UT) for cell isolation. The research protocol was approved by the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China. Prior to tissue collection, written informed consent was obtained from each participant.

After being washed 3–5 times with phosphate-buffered saline (PBS; Hyclone,) to remove traces of blood, tissues collected from both women with endometriosis and healthy women were manually minced into 1–2 mm³ fragments and incubated in DMEM/F-12 with 0.2% type II collagenase (Sigma-Aldrich, St. Louis, MO) for 60 min at 37°C. Stromal cells were separated from epithelial cells by filtration through a 40-µm nylon strainer (BD Falcon, San Jose, CA), as described previously [28]. The isolated cells were plated into Primaria flasks (Corning, Corning, NY) containing DMEM/F-12 [consisting of 20% fetal bovine serum (ScienCell, San Diego, CA), 0.1 mg/mL streptomycin, 100 U/mL penicillin, and 0.25 µg/mL amphotericin B (Sigma-Aldrich)] and incubated at 37°C in 95% air plus 5% CO₂. Red blood cells were removed by PBS 24 h later. The cells reached confluence in about 3 days, and cells of 2–3 passages were used for experiments. To determine the purity and homogeneity of EESCs and NESCs, immunofluorescent staining was performed using antibodies against vimentin (GB13225, Servicebio, Wuhan, China), CD10 (23898-1-AP, Proteintech, Wuhan, China), cytokeratin 7 (GB11066, Servicebio), and CD45 (GB11192, Servicebio). The purity of stromal cells was > 95% as assessed by positive staining for vimentin and CD10 and negative staining for cytokeratin 7 and CD45.

EESCs and NESCs were seeded in culture media 24 h before transfection into 60-mm dishes (4 × 10⁵ cells per well) for western blotting, 6-well plates (2 × 10⁵ cells per well) for quantitative real-time RT-PCR, or 24-well plates (2 × 10⁴ cells per well) for immunofluorescence. siRNA transfection was carried out in serum-free DMEM/F-12 by using NEOFECT^TM DNA transfection reagent (NEOFECT, Beijing, China) and 20 nM siRNA. The siRNA sequences were as follows: S1, GGACAGAGCA GCUGCCCAAdTdT; S2, CAGUCCAGCC AUGCUCCUCdTdT. Control siRNA or validated human NR4A1 siRNAs (GENE, Shanghai, China) were mixed into the culture media and incubated for 24 h for quantitative real-time RT-PCR or for 48 h for western blotting and immunofluorescence. Mock-transfected cells were used as a negative control. They were treated with 5 ng/mL TGF-β1 (PeproTech, Rocky Hill, NJ [29]) 6 h after siRNA transfection.

After being incubated with Csn-B (1 μmol; Santa Cruz Biotechnology, Santa Cruz, CA [26]), siRNA, and TGF-β1 for 24 h, total RNA from cells was extracted with Invitrogen TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. RNA quantification and purification were carried out at a 260/280 ratio on a NanoDrop 2000 spectrophotometer (NanoDrop Technologies, Wilmington, DE). A 260/280 ratio ranging from 1.8 to 2.0 was regarded as satisfactory purity. Complementary DNA was synthesized from total RNA using a reverse transcription PrimeScript RT Reagent Kit (Takara Biomedical Technology, Dalian, China).

The gene expression levels of ɑ-SMA, COL1A1, CTGF, and FN were determined by quantitative realtime PCR using the SYBR Premix Ex Taq kit (Takara Biomedical Technology). The PCR was performed in duplicate and run on a StepOnePlus Real-Time PCR System (Applied Biosystems, Foster City, CA). As an internal housekeeping control gene, GAPDH served to normalize the amount of cDNA within each sample. The primer sequences are listed in Table 1. Experiments were performed three times.

After being incubated with Csn-B, TGF-β1, and MK2206 (100 µm; Selleck Chemicals, Houston, TX [30]), proteins from cells were extracted, electrophoresed on sodium dodecyl sulfate-polyacrylamide gels, and transferred to polyvinylidene difluoride membranes (PVDF; Millipore Corp., Bedford, MA). After blockage with 5% nonfat milk for 1 h in Tris-buffered saline containing 0.1% Tween-20 (TBST), the membranes were incubated at 4°C overnight with primary antibodies against NR4A1 (1: 400 dilution; no. 3960, Cell Signaling Technology), phosphorylated-NR4A1 (P-NR4A1; 1: 200 dilution; no. sc-16992, Santa Cruz Biotechnology), AKT (1: 1000 dilution; no. 4691, Cell Signaling Technology), and phosphorylated AKT (P-AKT; 1: 1000 dilution; no. 4060, Cell Signaling Technology). The membranes were washed with TBST for 30 min and then immunoblotted with HRP-conjugated secondary antibodies (Proteintech) at room temperature for 1 h. Blots were developed by enhanced chemiluminescence (ECL; Beyotime, Beijing, China). Autoradiograms were densitometrically determined by using the Quantity One software package (Bio-Rad, Hercules, CA). GAPDH was used as a loading control and was visualized using monoclonal antibodies against GAPDH (1: 1000 dilution; no. 60004-1-Ig, ProteinTech). Western blots were analyzed using ImageJ software (version 1.41). Experiments were performed three times.

Formalin-fixed, paraffin-embedded human and mouse tissue sections were deparaffinized in xylene and rehydrated. The tissue slides were incubated with fresh prewarmed antigen retrieval buffer, and the buffer was heated to boiling. The tissue slides were then equilibrated to room temperature for 30 min. After being added to 3% H₂O₂ and rinsed with PBS, the slides were incubated with antibodies against NR4A1 (1: 50 dilution; no. PAB11925, Bio-Swamp) or P-NR4A1 (1: 100 dilution; no. sc-16992, Santa Cruz Biotechnology) at 4°C overnight. HRP-conjugated antibodies (ZSGB-BIO, Beijing, China) were used as secondary antibodies. After three washes with PBS, the tissue slides were incubated with the secondary antibody for 30 min at room temperature. Finally, the slides were color-reacted with DAB and counterstained with hematoxylin. To quantify immunostained cells, we used a computerized image analysis system consisting of an automatic digital slide scanner (Pannoramic Midi II; 3DHistech, Budapeste, Hungary) and an image-viewing software (Pannoramic Viewer; 3DHistech). The immunostained score (percentage of immunostained surface × mean staining intensity) was computed [31]. In each sample, 10 non-overlapping fields were analyzed.

For immunofluorescence, 4% PFA-fixed, 0.25% Triton X-100-permeabilized EESCs and NESCs were incubated with antibodies against NR4A1 (1: 100 dilution; no. sc-5569, Santa Cruz Biotechnology) and ɑ-SMA (1: 100 dilution; no. ab7817, Abcam) at 4°C overnight after treatment with TGF-β1, Csn-B, or NR4A1 siRNA. Fluorescent antibodies (1: 200 dilution; no. 150105 and 150075, Abcam, Cambridge, UK) were used as secondary antibodies. Cell nuclei were stained by using DAPI (ZSGB-BIO). In addition, F-actin was dyed with FITC-Phalloidin (Sigma-Aldrich). Stained cells were observed under a Nikon Eclipse 80i microscope (Nikon, Louisville, KY).

NR4A1^-/- mice (B6; 129S2-Nr4a1^tm1Jmi/J) were purchased from Jackson Laboratory (Bar Harbor, ME). Female NR4A1^-/- and littermate wild-type controls (WT; 7–8 weeks old, 23–25 g) were used in this study. Female nude mice (7–8 weeks old, 20–23 g) were procured from Vital River Laboratory Animal Technology Co. Ltd. (Beijing, China). The mice were maintained in autoclaved cages at 26°C with regular 12-h light–dark cycles and were fed sterile food and water. All procedures for the mouse experiments were reviewed and approved by the Animal Care Committee of Tongji Medical College, Wuhan, China.

Endometriotic lesions were experimentally induced by auto-transplantation of uterine horn tissue into the peritoneal cavity as previously described [32]. Briefly, animals were anesthetized by intraperitoneal injection of pentobarbital sodium (40 mg/kg; Solarbio, Beijing, China). A mid-ventral incision was made to expose the uterus. The horn was securely ligated at the utero-cervical junction and at the uterotubal junction, and then the part of the uterus between the two ligations was excised. The uterine horn, after being opened longitudinally, was placed in PBS and cut into four pieces measuring approximately 3–4 mm². Four tissue blocks of equal size from one uterine horn were transplanted onto the abdominal wall and anchored with 8-0 suture. The skin and peritoneum were closed with 5-0 suture. 17-β-estradiol-3-benzoata (30 µg/kg; Aladdin, Shanghai, China) was intramuscularly injected at the time of surgery and every 3 days thereafter [33] for 21 days. The mice were sacrificed on day 21 for collection of endometriotic implants.

A nude mouse endometriosis model was established as previously described [34]. Endometrial tissue from 8 patients without endometriosis was collected, placed in sterile PBS, and then minced. Next, 0.5 mL of 10 endometrial fragments at the proliferation phase (1–2 mm³) was injected subcutaneously into nude mice by 18-gauge needles. The mice were randomly divided into two groups after the injection. The first group of 30 mice was used for examining the changes in NR4A1 and P-NR4A1 during the development of endometriosis. 17-β-estradiol-3-benzoata was injected at the time of the endometrial tissue implantation and at 3-day intervals. The mice were sacrificed for tissue collection on days 3, 5, 7, 14, and 21. The second group of 30 mice was randomly further divided into two subgroups. In the first subgroup, the NR4A1 agonist Csn-B (Sigma-Aldrich), dissolved in dimethyl sulfoxide (DMSO; Sigma-Aldrich), was injected into the abdominal cavity at a dosage of 13 mg/kg twice weekly [27]. Animals in the second subgroup were injected with the same solvent without Csn-B. Treatment with Csn-B or DMSO started on the day of implantation. 17-β-estradiol-3-benzoata was injected at the time of the endometrial tissue implantation and every 3 days thereafter for 21 days. The mice were sacrificed for tissue collection on day 21.

Paraffin-embedded endometriotic tissue sections were stained with Masson trichrome, Sirius red, and hematoxylin as previously described [35]. Sirius red and Masson were quantified according to the staining intensity [36, 37]. The staining was scored from 0 to 4 (0 indicated no staining, 4 indicated maximal staining). Ten non-overlapping fields were analyzed.

GraphPad Prism 5 (GraphPad Software Inc., San Diego, CA) was used for statistical analysis. Comparisons between different groups were made using the Student's t-test and Kruskal-Wallis test. Statistical significance was set at P< 0.05.

Western blotting was performed to determine NR4A1 and P-NR4A1 expression levels in 15 normal endometrial and 23 ovarian endometriotic tissue samples. No significant differences were observed in the total NR4A1 level between endometrial and endometriotic tissue (Fig. 1A). Furthermore, treatment with antibodies conjugated to its phosphorylated form (P-NR4A1) showed that NR4A1 phosphorylation was augmented in endometriotic tissue compared with endometrial tissue (Fig. 1A). Immunohistochemical analysis of 15 normal endometrial and 23 ovarian endometriotic tissue samples showed that endometriotic stromal cells had higher expression of P-NR4A1 than endometrial stromal cells (Fig. 1B). Then, we cultured NESCs extracted from normal endometrial tissue and EESCs from endometriotic tissue to determine whether the expression of NR4A1 and P-NR4A1 in stromal cells was altered by TGF-β1 stimulation. By exposing NESCs and EESCs to TGF-β1 for 96 h as a persistent stimulation, the level of NR4A1 protein from NESCs began to rise at 12 h and then decreased to control levels within 72 h; the level of NR4A1 protein from EESCs began to increase at 6 h and then decreased, but had still not reached the control level by 96 h (Fig. 1C). Meanwhile, under identical experimental conditions, the concentration of P-NR4A1 in both NESCs and EESCs rose steadily over time (Fig. 1C). Immunofluorescent staining showed that the expression of NR4A1 in NESCs was upregulated at 24 h compared with 0 h and had decreased to control levels by 72 h (Fig. 1D). Taken together, these data showed that long-term stimulation with TGF-β1 promoted the phosphorylation of NR4A1 in NESCs and EESCs.

To understand the role of reduced NR4A1 function in the pathogenesis of fibrosis in endometriosis, NESCs and EESCs were transfected with NR4A1 siRNA. We studied two NR4A1 siRNAs. Experiments in NESCs indicated that NR4A1 siRNA of S1 lowered NR4A1 protein expression by 30% (P< 0.01) and that NR4A1 siRNA of S2 reduced the level of NR4A1 protein by 40% (P< 0.01) (Fig. 2A). Both S1- and S2-transfected NESCs had lower expression of NR4A1 mRNA compared with mock-transfected cells (Fig. 2B). The immunofluorescent staining showed that NR4A1 siRNA S2 transfection led to a higher expression of ɑ-SMA and increased the formation of stress fibers in TGF-β1-treated NESCs (Fig. 2C). After incubation with TGF-β1, the mRNA expression of ɑ-SMA, FN, COL1A1, and CTGF was increased in S2-transfected NESCs and EESCs (Fig. 2D). Collectively, these data indicated that the decreased NR4A1 promoted the expression of fibrosis markers in TGF-β1-treated NESCs and EESCs.

In order to further explore whether the activation of NR4A1 could decrease fibrosis in vitro, we incubated stromal cells with TGF-β1 and Csn-B. Western blotting results showed that Csn-B significantly decreased the TGF-β1-induced phosphorylation of NR4A1 and marginally increased the expression of NR4A1 in NESCs (Fig. 3A). Immunofluorescence staining revealed that treatment with Csn-B significantly reduced the expression of ɑ-SMA and the formation of stress fibers induced by TGF-β1 in NESCs (Fig. 3B). In addition, treatment with Csn-B significantly decreased the TGF-β1-dependent increase in the expression of ɑ-SMA, FN, COL1A1, and CTGF mRNA in both endometrial and endometriotic stromal cells (Fig. 3C).

To investigate whether AKT participates in the TGF-β1-induced phosphorylation of NR4A1, we incubated NESCs with TGF-β1 and MK2206 (an inhibitor of AKT). MK2206 significantly decreased the level of phosphorylated AKT and, at the same time, reduced the phosphorylation of NR4A1 (Fig. 4A). These data showed that AKT is involved in the TGF-β1-induced phosphorylation of NR4A1 in NESCs.

The expression of NR4A1 and P-NR4A1 during the development of endometriosis was determined in 30 nude mice. Ten human endometrial fragments were subcutaneously injected into each nude mouse. On day 21, 3–5 endometriotic lesions were found under the skin in each mouse. Five nude mice were evaluated in each group. Immunohistochemical staining using NR4A1 antibody indicated that the level of NR4A1 was significantly higher on day 5 and 7 than on day 0 and 3. On the other hand, the expression of NR4A1 on day 14 and 21 was decreased compared with day 5 and 7, but the level remained higher than that on day 0 (Fig. 5A). Furthermore, immunohistochemical staining using P-NR4A1 antibody indicated that the level of P-NR4A1 started to increase on day 5 and rose steadily during the development of endometriotic lesions (Fig. 5A).

To investigate the loss of NR4A1 function in endometriosis-associated fibrosis in vivo, endometriosis was surgically induced in NR4A1^-/- and littermate WT control mice by auto-transplantation of uterine tissue (Fig. 6A). All of the mice developed endometriosis with endometriotic tissue containing stromal and glandular cells. Western blotting and immunohistochemical staining using NR4A1 antibody showed that NR4A1^-/- ectopic lesions had weak NR4A1 expression (Fig. 6B and C). NR4A1^-/- ectopic lesions were strongly stained by Sirius red and Masson trichrome staining (Fig. 6C). The results suggested that loss of NR4A1 promoted fibrogenesis in mice with endometriosis.

A mouse endometriosis model was established by xenograft (Fig. 7A). Animals were monitored on a daily basis and body weight was recorded. No significant difference was found in growth rate between Csn-B-treated and untreated mice. The numbers and sizes of the endometriotic lesions were also not significantly different between the two groups on day 21. However, western blotting and immunohistochemistry indicated that Csn-B-treated mice had a higher level of total NR4A1 and lower level of P-NR4A1 in endometriotic lesions (Fig. 7B and C), and the endometriotic tissues in Csn-B-treated mice were weakly stained by Sirius red and Masson trichrome staining (Fig. 7C). Collectively, these data indicated that Csn-B effectively inhibited the fibrogenesis in nude mice with endometriosis.

This study demonstrated that persistent stimulation with TGF-β1 phosphorylated NR4A1 in endometrial stromal cells through an AKT-dependent mechanism. NR4A1 deficiency promoted fibrosis in endometriosis and Csn-B antagonized this effect both in vitro and in vivo. These results may provide a new target for the treatment of endometriosis.

Tissues can be damaged after multiple acute or chronic stimulations, such as autoimmune reactions, infections, and mechanical injuries [38]. The repair process involves a regenerative phase and a fibrotic phase [38]. Hence, fibrosis is often regarded as an uncontrolled wound-healing response [39] and is characterized by the presence of abundant fibrous connective tissue [40]. TGF-β is believed to be a key mediator in this process [41]. Although the pathogenesis of endometriosis is still unclear, one of the most widely accepted hypotheses is the implantation theory [1]. In this theory, endometriosis originates from retrograde menstruation of endometrial tissue, which passes through the fallopian tubes into the peritoneal cavity and implants in the peritoneal cavity or on the abdominal organs.

The endometrial fragments proliferate and cause inflammation and adhesions [42]. Here, we see fibrosis in endometriosis as an uncontrollable healing response induced by TGF-β1. Dysfunction of stromal cells after injury may regulate the fibrotic process [43]. Our study showed that stimulation with TGF-β1 promoted the expression of fibrotic markers in NESCs and EESCs.

NR4A1 is a transcription factor expressed in various and tissues that serves a range of biological functions [44]. NR4A1 bi-directionally regulates gene expression by binding to the promoter of its downstream target gene [44, 45]. NR4A1 has been implicated in hypercholesterolemia [46], bladder cancer growth [47], and other events. This study showed that inhibition of NR4A1 promoted the expression of fibrotic markers in vitro and that loss of NR4A1 exacerbated fibrosis in a mouse endometriosis model. These findings suggest that NR4A1 participates in the fibrogenesis of endometriosis. As a member of nuclear orphan receptors, NR4A1 has been shown to act independently of ligands [48]. Hence, the activity of NR4A1 could be regulated at both a protein level and a post-translational modification level, such as via phosphorylation [49]. In 1990, Fahrner found that NR4A1 could be phosphorylated [50], and subsequent studies confirmed this finding and further indicated that the activity of NR4A1 was decreased by phosphorylation [48, 51]. Jiang et al [52]. showed that patients with uterine adenomyosis had a lower level of NR4A1 expression in endometrial tissues compared with fertile controls. However, this study showed that expression of NR4A1 was not significantly different between endometriotic tissue from women with ovarian endometriosis and endometrial tissue from healthy women. Furthermore, the P-NR4A1 level was increased in endometriotic tissue. This discrepancy might be because the samples were from different sources.

Because a previous study found that the ratio of stromal cells to epithelial cells in endometriotic tissue was different from that in the endometrium [53], it is difficult to determine whether the different expression of NR4A1 and P-NR4A1 protein in the two kinds of tissues was caused by stromal cells. However, our immunohistochemical results indicated that stromal cells may contribute more to this difference. NR4A1 expression was higher on day 21 than on day 0 in the nude mouse endometriosis model, possibly because of the short time of endometriotic lesion development. AKT has been reported to phosphorylate NR4A1 in gastric cancer cells [54]. Our study showed that long-term stimulation with TGF-β1 phosphorylated NR4A1 and that MK2206 could partly counteract this process. These findings indicated that AKT is involved in the post-translational regulation of NR4A1 in endometrial stromal cells.

Csn-B is isolated from an endophytic fungus named Dothiorella sp. Strain HTF3 [55]. It specifically binds to the ligand-binding domain of NR4A1 and then stimulates the transactivational activity of NR4A1 [27]. Moreover, Csn-B does not have any such effects in NR4A1-null mice or cells, suggesting that Csn-B is an agonist of NR4A1 [27]. Recent studies found that Csn-B reduced hypercholesterolemia-induced inflammation [46], strongly blocked the secretion of Salmonella pathogenicity-associated proteins [55], and inhibited the growth of androgen-dependent bladder cancer cells [47]. In the present study, Csn-B effectively decreased the expression of fibrosis markers in vitro and attenuated fibrogenesis in a nude mouse model of endometriosis. Therefore, Csn-B may be a drug candidate for the treatment of endometriosis. Because the endometriotic lesions of nude mice developed under the skin instead of in the ovary, the mouse model used in this study has severe limitations compared with human ovarian endometriosis. Considering that Csn-B is a natural product and may be difficult to extract on a large scale, analogs of Csn-B might be useful in future clinical applications.

Our study showed that persistent stimulation of endometriotic tissues with TGF-β1 decreased the activity of NR4A1 through AKT-dependent phosphorylation. The devitalized NR4A1 lost its function as a depressor of TGF-β1 signaling and promoted fibrogenesis in ovarian endometriosis. Csn-B treatment significantly reduced the expression of fibrotic markers in vitro and inhibited fibrogenesis in mouse endometriosis. This study suggests that NR4A1 is involved in the fibrosis of ovarian endometriosis and that Csn-B, an agonist of NR4A1, is a potential candidate for the treatment of endometriosis.

Y.G. and G.J. were principally responsible for the research design and performed analyses. X.Z. and Z.Y. were involved in the study design, experimentation, result analysis, and manuscript drafting. Y.S., J.H., M.Y., and Y.L. conducted the experiments. All authors approved the final version of the paper.

This work was supported by grants from the National Natural Science Foundation of China (No. 81472406 and No. 81272816).

The authors declare no conflict of interest.

X. Zeng and Z. Yue contributed equally to this work.