Betulinic Acid Inhibits Endometriosis Through Suppression of Estrogen Receptor β Signaling Pathway

Endometriosis is an inflammatory gynecological disorder characterized by endometrial tissue growth located outside of the uterine cavity in addition to chronic pelvic pain and infertility. In this study, we aim to develop a potential therapeutic treatment based on the pathogenesis and mechanism of Endometriosis. Our preliminary data showed that the expression of estrogen receptor β (ERβ) was significantly increased, while ERα was significantly decreased, in endometriotic cells compared to normal endometrial cells. Further investigation showed that betulinic acid (BA) treatment suppressed ERβ expression through epigenetic modification on the ERβ promoter, while had no effect on ERα expression. In addition, BA treatment suppresses ERβ target genes, including superoxide dismutase 2 (SOD2), nuclear respiratory factor-1 (NRF1), cyclooxygenase 2 (COX2), and matrix metalloproteinase-1 (MMP1), subsequently increasing oxidative stress, triggering mitochondrial dysfunction, decreasing elevated proinflammatory cytokines, and eventually suppressing endometriotic cell proliferation, mimicking the effect of ERβ knockdown. On the other hand, gain of ERβ by lentivirus infection in normal endometrial cells resulted in increased cell proliferation and proinflammatory cytokine release, while BA treatment diminished this effect through ERβ suppression with subsequent oxidative stress and apoptosis. Our results indicate that ERβ may be a major driving force for the development of endometriosis, while BA inhibits Endometriosis through specific suppression of the ERβ signaling pathway. This study provides a novel therapeutic strategy for endometriosis treatment through BA-mediated ERβ suppression.


INTRODUCTION
Endometriosis (EMS) is a common inflammatory gynecological disease characterized by the presence of endometrial-like lesions located outside the uterus, causing chronic pelvic pain and infertility (1), and affects approximately 5-10% of women during their reproductive ages (2). Diagnosis for EMS is mostly dependent on surgical visualization. Retrograde menstruation is considered to be an important mechanism, while other factors, including endocrine and metabolic pathway, altered immunity, inflammatory response, and metaplasia of the coelom, may also be involved (2). Current treatment for EMS primarily focuses on alleviating symptoms, such as through pain relief, surgical removal, and hormone suppression (3), while development of novel targeted treatment for EMS based on a deeper understanding of its pathological mechanism is still an urgent requirement.
Endometriosis is estrogen dependent, and the expression of estrogen receptor (ER)b in EMS is significantly higher than those in endometrial tissues. High levels of ERb suppress ERa, resulting in very high ratios of ERb/ERa (4,5). Recently, it has been reported that ERb plays an important role in the pathogenesis of EMS by modulating apoptosis and inflammasome (6). Additionally, ERb modulates mitochondrial function through nuclear respiratory factor-1 (NRF1) (7) and regulates the basal expression of superoxide dismutase 2 (SOD2) to minimize oxidative stress (8)(9)(10). In addition, ERb modulates the expression of cyclooxygenase-2 (COX2) (1,4,11,12) for inflammation (13) and regulates the expression of matrix metalloproteinases 1 (MMP1) for tissue degradation, cell invasion, and growth (14)(15)(16). It has been reported that ERb plays an important role in EMS pathogenesis by modulating its target genes (6), including NRF1, SOD2, COX2, and MMP1, and specific ERb suppression, such as ERb antagonist (17), has been used for treatment of endometriosis, but with many of limitations and side effects. In this case, development of a nontoxic and specific ERb suppression agent for treatment of endometriosis is still quite necessary.
Betulinic acid (BA) is identified as a pentacyclic triterpene that can be isolated from many natural plants and has been characterized to possess several biological properties, including its ability to inhibit tumor growth (18,19) through apoptosis (20) and suppress leukemia (21,22) through modulation of mitochondrial function and oxidative stress (23) while also having an antiviral effect in inhibiting the HIV (18,24), EBV (25), and HBV viruses (23). Recently, it has been reported that BA suppresses the estrogen signaling pathway either directly or indirectly (26)(27)(28), while the detailed mechanism remains largely unknown (29). Here, we hypothesize that BA inhibits endometriosis through ERb suppression.
In an effort to study the potential mechanism and effect of BA on endometriosis, we evaluated the related gene expression in different cells, finding that ERb expression was significantly increased in endometriotic cells compared to endometrial cells and BA treatment significantly inhibited the expression of ERb and its target genes. Further investigation showed that BA suppresses ERb expression through epigenetic changes on the ERb promoter. Also, BA treatment significantly increased oxidative stress, induced mitochondrial dysfunction, normalized elevated levels of proinflammatory cytokines, and subsequently suppressed cell proliferation and growth in endometriotic cells, mimicking the effect of ERb knockdown. We conclude that BA inhibits endometriosis by suppression of the ERb signaling pathway, providing a novel strategy for potential treatment of endometriosis.

MATERIALS AND METHODS
A detailed description can be found in Supplementary Information (see Data S1), and the primers used in this study are shown in Table 1.

Human Cell Lines
Human Endometrial Epithelial Cells HEEC (#ABC-TC4601) and Immortalized Human Endometriotic Epithelial Cell Line 12Z (#ABI-TC278D) were obtained from ACCEGEN Biotechnology. The human primary endometrial epithelial cells (EM) and primary endometriotic epithelial cells (EMT) were a kind gift from Dr. Haimou Zhang (from Hubei University) (30). In some of the experiments, the cells for HEEC, EM, and EMT were conditionally immortalized using a hTERT lentivirus vector with an extended life span to achieve higher transfection efficiency and experimental stability (10,31). The cells were maintained in DMEM/F12 medium supplemented with 10% FBS, 1% penicillin and streptomycin, and 1% sodium pyruvate at 37°C with 5% CO2.

Construction of Human Reporter Plasmid
In order to construct ERb reporter plasmids, the gene promoters (2kb upstream of the transcription start site plus first exon) were amplified from Ensembl ID: ESR2-201 ENST00000267525.10 by PCR from human genomic DNA and subcloned into the pGL3-basic vector (#E1751, Promega) using restriction sites of Mlu I and Hind III with the following primers: ERb forward: 5'-gcgcacgcgt-att tca aga cga gcc tgg cca -3' (Mlu I) and ERb reverse: 5'gtac-aagctt-ctg ttt aca ggt aag gtg tgt -3' (Hind III). To map promoter activity, the related deletion promoter constructs were generated by PCR methods and subcloned into the pGL3-basic vector (23).

Generation of Human ERb Expression Lentivirus
The human cDNA for ERb was obtained from Open Biosystems. The cDNA for human ERb was subcloned into the pLVX-Puro vector (from Clontech) with the restriction sites of Xho1 and Xba1 using the below primers: ERb forward primer: 5'-gtac -ctcgag-atg gat ata aaa aac tca cca -3' (Xho1) and ERb reverse primer: 5'-gtactctaga-tca ctg ctc cat cgt tgc ttc -3' (Xba1). The ERb or empty control (CTL) was expressed through Lenti-X ™ Lentiviral Expression Systems (from Clontech) per manufacturers' instructions (32).

Preparation of ERb Knockdown (shERb)
The shRNA lentivirus plasmids for human ERb (sc-35325-SH) or non-target control (sc-108060) were purchased from Santa Cruz Biotechnology. The related lentiviruses for ERb and empty control (CTL) were expressed through Lenti-X ™ Lentiviral Expression Systems (from Clontech) per manufacturers' instructions. The purified and condensed lentiviruses were used for in vitro gene knockdown. The knockdown efficiency was confirmed by more than 65% of mRNA reduction compared to the control group in cells using real time PCR (see Table 1).

Immunostaining
The treated cells were transferred to cover slips, and the cells were fixed in 4% paraformaldehyde for 20 min before being incubated with 0.3% Triton X-100 in PBS for 15 min. After blocking with 5% goat serum in PBS at room temperature for 30 min, cells were incubated with anti-mouse antibody for either 8-oxo-dG (# 4354-MC-050, from Novus Biologicals) or Ki-67 (# sc-101861, from Santa Cruz Biotechnology) for 12 h at 4°C and subsequently with secondary antibody Alexa Fluor 488. The cover slips were then mounted by antifade Mountant with DAPI (staining nuclei, in blue). The photographs were taken using a Confocal Laser Microscope (Leica, 20× lens) and quantitated by Image J. software (34).
Statistical Analysis. The data was given as mean ± SEM, and all the experiments were performed at least in quadruplicate unless otherwise indicated. The unpaired Student's t-tests or one-way analysis of variance (ANOVA) followed by the Turkey −Kramer test was used to determine statistical significance of different groups by SPSS 22 software, and a P value of <0.05 was considered significant (35,36).

Expression of ERb and Its Target Genes Is Significantly Increased, While Expression of ERa Decreased, in Endometriotic Cells Compared to Endometrial Cells
Different cells, including Human Endometrial Epithelial Cells (HEEC), immortalized Human Endometriotic Epithelial Cell Line 12Z, human primary endometrial epithelial cells (EM), and primary endometriotic epithelial cells (EMT), were used for gene expression of ERb, ERa and its target genes, including SOD2, NRF1, MMP1, and COX2. The results showed that 12Z and EMT cells have ERb mRNA levels that are 17.2 and 19.6 times higher of than HEEC cells, respectively; On the other hand, ERa mRNA levels were deceased to 45 and 21% in endometriotic 12Z and EMT cells, respectively, compared to its related endometrial HEEC and EM cells. In addition, SOD2 mRNA levels in 12Z and EMT cells increased to 286 and 216%, respectively; NRF1 mRNA levels increased to 244 and 198%, respectively; MMP1 mRNA levels increased to 226 and 187%, respectively; and COX2 mRNA levels increased to 314 and 211%, respectively, compared to HEEC cells. Furthermore, we found that there was no difference in mRNA levels of ERb, NRF1, and MMP1 in EM cells, but levels were 1.45 and 1.43 times higher, respectively, then in HEEC cells (see Figure 1A). In addition, we measured the protein levels for those genes and found that ERb levels in 12Z and EMT cells were 3.11 and 3.42 times higher than that of HEEC cells, respectively, and an expression pattern similar to that of the mRNA was observed for genes of SOD2, NRF1, MMP1, and COX2 (see Figures 1B, C, and S1). We conclude that expression of ERb and its target genes is significantly increased, while expression of ERa decreased, in endometriotic cells compared to endometrial cells.

Betulinic Acid Suppresses the ERb Signaling Pathway by Modulation of Epigenetic Changes on the ERb Promoter and Subsequent Expression
We first evaluated the potential effect of betulinic acid (BA) on the gene expression of ERb and its target genes. The conditional immortalized HEEC or 12Z cells were either treated with 20 µM betulinic acid (BA) for 24 h or infected by a lentivirus of the empty control (CTL), ERb expression (↑ERb), or ERb knockdown (shERb) groups, before being harvested for gene expression analysis. The results showed that HEEC/↑ERb and 12Z/CTL treatments result in ERb mRNA levels that are 13.0 and 13.6 times higher than that of HEEC cells. Additionally, HEEC/↑ERb and 12Z/CTL treatments have increased SOD2 mRNA levels to 215 and 264%, respectively; increased NRF1 mRNA levels to 226 and 187%, respectively; increased MMP1 mRNA levels to 234 and 208%, respectively; and increased COX2 mRNA levels to 295 and 229%, respectively, compared to the HEEC/CTL group. This indicates that a gain of ERb by lentivirus infection in endometrial cells (HEEC/↑ERb) increases expression of ERb and its target genes, mimicking the effect of endometriotic cells, such as the 12Z/CTL group. On the other hand, ER knockdown in 12Z cells (12Z/shERb) decreased the expression of ERb and its target genes to levels that were similar to that of the HEEC/CTL group, and interestingly, BA treatment (12Z/BA) mimicked the effect of shERb and normalized endometriosis (12Z cells)-mediated increased expression of ERb and its target genes (see Figure 2A). We also measured the protein levels for those genes and found that HEEC/↑ERb and 12Z/CTL treatments resulted in ERb levels that were 3.41 and 2.68 times higher than that of HEEC cells, respectively, and an expression pattern similar to that of the mRNA was observed for the SOD2, NRF1, MMP1, and COX2 genes (see Figures 2B, C, and S2). These results indicate that BA can suppress gene expression of ERb and its target genes. We then measured the potential effect of BA on ERb/ ERa expression using dose-response curve in endometriotic 12Z cells, which showed that 20 µM of BA achieved the maximum inhibition effect on ERb mRNA expression after a 24-h treatment, while BA treatment showed no effect on ERa expression (see Figure  2D). We also investigated the potential effect in endometrial HEEC cells, the results showed that 20 µM of BA had no significant effect on the expression of ERa and ERb, while when the dose of BA reached to 30 mM, it increased ERa, but decreased ERb expression very slightly (see Figure S3), indicating that BA treatment can specifically suppress ERb expression in endometriotic cells. Afterwards, we investigated the potential molecular mechanism for betulinic acid-mediated ERb suppression. In order to identify the betulinic acid responsive element on the ERb promoter, a series of progressive 5'-promoter deletion constructs for the ERb promoter were generated by PCR methods, including -2000, -1600, -1200, -800, -400, -300, -200, -100, and -0 deletion constructs (numbered according to Ensembl gene ID: ESR2-201 ENST00000267525.10, and the transcription start site was marked as 0). All of these constructs were transfected into 12Z cells for the analysis of ERb reporter activity in the presence of either control (CTL) or 20 µM betulinic acid for 24 h, and the betulinic acid-induced relative reporter activities (% control) were calculated. Our results showed that BAinduced relative reporter activities were around 60% compared to the control group for the deletion constructs of -2000, -1600, -1200, -800, -400, and -300, while the reporter activities reached around 100% for the deletion constructs of -200, -100, and -0, indicating that BAinduced reporter activity suppression was restored in the deletion constructs of -200, -100, and -0, and the BA-responsive transcriptional element is located in the range of -300~0 on the ERb promoter (see Figure 2E). We also evaluated the potential epigenetic changes in the range of -300~0 on the ERb promoter by ChIP analysis. The results showed that there was no significant difference in any of the different treatments on either histone H3 (K9, K14, K18, K23, K27) or H4 (K5, K8, K12, K16) acetylation (see Figure  S4A). Furthermore, we found that betulinic acid had no effect on histone H4 methylation, including H4K20me1, H4K20me3, and H4K3me1 (see Figure S4B). Finally, we evaluated the potential effect of different treatments on H3 methylation. The results  showed that there was no effect on the methylation of H3K9me3 and H3K27me2, while 12Z/CTL treatment decreased H3K9me2 and H3K27me3 to 43 and 51%, respectively, compared to the HEEC/ CTL group; ERb knockdown (12Z/shERb) showed no effect, while BA treatment (12Z/BA) completely reversed 12Z-mediated decreased H3 methylation on the ERb promoter (see Figure 2F).
Our results indicate that betulinic acid suppresses the ERb signaling pathway by modulation of epigenetic changes on the ERb promoter and subsequent expression.

Betulinic Acid Mimics ERb Knockdown-Induced Oxidative Stress in Endometriotic Cells
We evaluated the effect of ERb knockdown and betulinic acidinduced oxidative stress in endometriotic cells. The conditional immortalized HEEC or 12Z cells were either treated by 20 µM betulinic acid (BA) for 24 h or infected by either an empty control (CTL) or ERb knockdown (shERb) lentivirus, and the cells were then harvested for analysis. Our results showed that ERb knockdown (12Z/shERb) and BA treatment (12Z/BA) increased ROS formation to 352 and 319%, respectively (see Figure 3A), and increased 3nitrotyrosine formation to 223 and 178%, respectively (see Figure  3B), compared to the HEEC/CTL group; while the oxidative stress showed no significant difference between HEEC and 12Z cells. We then investigated the effect of stimulus TNFa on oxidative stress in those cells (6). The results showed that 12Z cells had significant higher levels of ROS formation (see Figure S5A), 3-nitrotyrosin formation (see Figure S5B), and 8-OHdG formation (see Figure S5C) in 12Z cells compared to HEEC cells in the presence of TNFa. We next measured the effect of betulinic acid on DNA damage. The results showed that treatments of 12Z/shERb and 12Z/BA increased 8-OHdG formation to 247 and 216%, respectively (see Figure 3C) and increased gH2AX formation to 178 and 231%, respectively (see Figures 3D, E, and S6). Next, we measured SOD2 activity, and the results showed that 12Z/CTL group increased SOD2 activity to 149%, while treatments of 12Z/shERb and 12Z/BA decreased SOD2 activity to 64 and 74%, respectively, compared to the HEEC/CTL group (see Figure 3F). Finally, we evaluated the 8-oxo-dG formation using immunostaining quantitation. The results showed that treatments of 12Z/shERb and 12Z/BA increased 8-oxo-dG formation to 241 and 256%, respectively, compared to the HEEC/CTL group, while the 12Z/CTL group showed no difference (see Figures 3G, H). Our results indicate that betulinic acid mimics ERb knockdown-induced oxidative stress in endometriotic cells.

Betulinic Acid Mimics ERb Knockdown-Induced Mitochondrial Dysfunction and Apoptosis in Endometriotic Cells
We evaluated the effect of ERb knockdown and betulinic acidinduced mitochondrial dysfunction in endometriotic cells. The results showed that mitochondrial DNA copies (see Figure 4A) and intracellular ATP levels (see Figure 4B) in the 12Z/CTL group were significantly increased to 236 and 169%, respectively, compared to the HEEC/CTL group. ERb knockdown (12Z/  shERb) completely normalized this effect, while betulinic acid treatment (12Z/BA) partly normalized this effect. We measured the caspase-3 activity in different treatments. The results showed that 12Z/CTL had no effect, while 12Z/shERb group increased caspase-3 activity to 191% compared to the HEEC/CTL group, and 12Z/BA mimicked the effect of 12Z/shERb (see Figure 4C). Additionally, mitochondrial membrane potential (DYm) of the 12Z/CTL group significantly increased to 131% compared to the HEEC/CTL group, and treatments of 12Z/shERb and 12Z/BA completely normalized this effect (see Figure 4D). Finally, we evaluated the apoptosis rate. The results showed that 12Z/CTL had no effect, while apoptosis rate in the 12Z/shERb group increased to 638% compared to the HEEC/CTL group, and 12Z/BA mimicked the effect of 12Z/shERb (see Figures 4E, F ). In addition, the caspase-3 activity and apoptosis rate showed no significant difference between HEEC and 12Z cells. We then investigated the effect of stimulus TNFa in those cells (6). The results showed that 12Z cells had significant lower levels of caspase-3 activity (see Figure S7A) and apoptosis rate (see Figure S7B) in 12Z cells compared to HEEC cells in the presence of TNFa. The results suggest that betulinic acid mimics ERb knockdown-induced mitochondrial dysfunction and apoptosis in endometriotic cells.

ERb Knockdown Reduces Endometriosis-Mediated Elevated Proinflammatory Cytokine Secretion, and Betulinic Acid Treatment Mimics This Effect
We evaluated the potential effect of ERb knockdown and betulinic acid-mediated proinflammatory cytokine release in endometriotic cells. We first evaluated the mRNA levels of proinflammatory cytokines, and the results showed that 12Z/ CTL group significantly increased mRNA levels of IL1b, IL6, TNFa, and PGE2 to 245, 215, 189, and 231%, respectively, compared to the HEEC/CTL group, while 12Z/shERb treatment completely normalized this effect, and 12Z/BA treatment completely normalized the mRNA levels of IL6 and TNFa, but partly normalized mRNA levels of IL1b and PGE2 (see Figure 5A). Furthermore, we measured cytokine release from cell culture supernatants for IL1b (see Figure 5B), IL6 (see Figure 5C), TNFa (see Figure 5D), and PGE2 (see Figure 5E), and an expression pattern similar to that of the mRNA was observed. Our results indicate that ERb knockdown reduces endometriosis-mediated elevated proinflammatory cytokine secretion, and betulinic acid treatment mimics this effect.

ERb Knockdown Reduces Endometriosis-Mediated Cell Proliferation, and Betulinic Acid Treatment Mimics This Effect
We evaluated the potential effect of ERb knockdown and betulinic acid-mediated cell proliferation in endometriotic cells. First, cell proliferation was measured by thymidine incorporation. The results showed that 12Z/CTL group significantly increased thymidine incorporation to 290% compared to the HEEC/CTL group, and treatments of 12Z/shERb and 12Z/BA completely normalized this effect (see Figure 6A). We then evaluated cell migration (see Figure  6B) and invasion (see Figure 6C). The results showed that the 12Z/ CTL group significantly increased cell migration and cell invasion to 241 and 257%, respectively, compared to the HEEC/CTL group, and treatments of 12Z/shERb and 12Z/BA partly normalized the effect on cell migration, but completely normalized the effect on cell invasion. We also evaluated cell colony formation (see Figure 6D)  Figures 6E, F). The results showed that cell colony formation and the ratio of Ki-67 positive cells in the12Z/CTL group significantly increased to 336 and 342%, respectively, compared to the HEEC/CTL group, and 12Z/shERb treatment completely normalized this effect, while 12Z/BA treatment partly normalized this effect. Our results indicate that ERb knockdown reduces endometriosis-mediated cell proliferation, and betulinic acid treatment mimics this effect.

Gain of ERb in Endometrial Cells Promotes Cell Proliferation, While BA Treatment Diminishes This Effect
We investigated the potential effect of ERb expression and BA treatment in endometrial cells. The conditional immortalized HEEC cells were infected by either empty control (CTL) or ERb expression (↑ERb) lentivirus, or ERb expression (↑ERb) lentivirus plus 20 µM betulinic acid (BA) (↑ERb/BA) for 24 h, then the cells were harvested for analysis. We first evaluated the ERb/ERa mRNA level, and the results showed that ERb expression (↑ERb) increased ERb mRNA level to 317% compared to control (CTL) group, while BA treatment (↑ERb/BA) significantly reduced ERb level to 44% compared to ↑ERb group; on the other hand, ERb expression (↑ERb) treatment decreased ERa mRNA to 57%, and BA treatment (↑ERb/BA) completely reversed this effect ((see Figure 7A). We next measured the oxidative stress. We found that ↑ERb treatment slightly increased ROS formation (see Figure 7B) and 3nitrotyrosine formation (see Figure 7C), while ↑ERb/BA  treatment further significantly potentiated the oxidative stress compared to CTL group. We next measured the mitochondrial function. The results showed that ↑ERb treatment slightly decreased apoptosis rate (see Figure 7D) and increased mitochondria membrane potential (see Figure 7E), while ↑ERb/BA treatment significantly further increased the apoptosis rate, and restored mitochondria membrane potential to normal level compared to CTL group. We also measured the proinflammatory cytokine release. The results showed that ↑ERb treatment significantly increased cytokine releases, including IL1b (see Figure 7F), IL6 (see Figure 7G), TNFa (see Figure 7H), and PGE2 (see Figure 7I), while ↑ERb/BA treatment completely or partly restored cytokine release to normal levels compared to CTL group. We finally evaluated the cell proliferation. The results showed that ↑ERb treatment significantly increased thymidine incorporation (see Figure 7J) and ratio of Ki-67 positive cells (see Figure 7K), while ↑ERb/BA treatment completely or partly restored this effect to normal levels compared to CTL group. Our results indicate that gain of ERb in endometrial cells promotes cell proliferation and proinflammatory cytokine release, while BA treatment diminishes this effect through increased oxidative stress and apoptosis.

DISCUSSION
In this study, we demonstrated that gene expression of ERb and its target genes are significantly increased in endometriotic cells compared to endometrial cells, and BA treatment suppresses ERb expression by epigenetic changes on the ERb promoter in endometriotic cells. In addition, BA treatment increases oxidative stress, induces mitochondrial dysfunction, and decreases proinflammatory cytokine release, subsequently inhibiting endometriosis.

Effect and Role of ERb and ERa in Endometriosis
Our results showed that ERb expression was significantly increased in endometriotic cells as well as ERb target genes, including SOD2, NRF1, COX2, and MMP1, subsequently contributing to endometriosis (EMS) development (4,5), while ERa expression was significantly decreased, and the ERb:ERa ratio in endometriotic 12Z cells is increased to 38:1 compared to endometrial HEEC cells as calculated from data in Figure 1A, which is consistent with previous report (4,5). In addition, increased SOD2 expression results in decreased reactive oxygen species (ROS) formation and minimized oxidative stress and increases the invasiveness of cell growth (37), while increased NRF1 expression triggers mitochondrial replication through Tfam (transcription factor A, mitochondrial) (7, 38), providing stronger respiration and metabolic energy for cell proliferation (39). On the other hand, increased COX2 expression involves with inflammation and COX2-derived prostaglandin E2 (PGE2) biosynthesis contributes to EMS-related pain and infertility (1,13), while increased MMP1 expression is involved with tissue degradation, menstrual bleeding, and invasion of seeded  endometriotic explants (14,15). In addition, gain of ERb in normal endometrial cells significantly increased cell proliferation, proinflammatory cytokine release together with decreased apoptosis (6). We conclude that increased ERb expression may be the major driving force for EMS development, and ERb is the potential therapeutic target for clinical treatment of endometriosis (40).

Potential Effect of Betulinic Acid on ERb Expression
Our results showed that betulinic acid (BA) treatment significantly suppresses the expression of ERb and its target genes through epigenetic changes on the ERb promoter, subsequently suppressing cellular proliferation and growth. The BA-mediated epigenetic modifications may be transferred into daughter cells during cell proliferation, making the BAmediated ERb suppression effect stable and long-lasting. In addition, it has been reported that BA is widely available from common natural sources with relative non-toxicity, making BA a potential novel candidate for drug development for rescuing the epigenetic changes (29).

Potential Effect of Betulinic Acid on Endometriosis Growth
Our results showed that betulinic acid can significantly inhibit the endometriotic cell growth through specific suppression of ERb and its target genes in endometriotic cells (28). Betulinic acid treatment significantly increases oxidative stress and DNA damage, inhibits the proinflammatory cytokines release (41), resulting in apoptosis and suppressed cell proliferation in endometriotic cells. Furthermore, the betulinic acid has little effect on the expression of ERa, and it significantly suppresses ERb expression in endometriotic cells, but has no effect in normal endometrial cells, indicating that betulinic acid can specifically inhibit ERb expression in endometriosis. Our data showed that betulinic acid may be a potential novel candidate for clinical treatment of endometriosis.

CONCLUSIONS
The endometriotic cells have high expression of ERb and its target genes, and betulinic acid can specifically suppress ERb signaling pathway by epigenetic modification on the ERb promoter, subsequently suppressing endometriosis development. We conclude that betulinic acid inhibits endometriosis through suppression of ERb signaling pathway.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.