Lower growth arrest‐specific 5 level in endometrium is related to endometriosis via promoting cell proliferation and angiogenesis

We explored the roles of GAS5 in the development of endometriosis, by measuring the levels of GAS5 by qRT‐PCR in 23 endometrial samples from controls without endometriosis, and paired eutopic and ectopic endometrial samples from 33 patients with endometriosis. The normal GAS5 expression range was defined according to the GAS5 level in the control group. As shown in Figure  1(A) , the GAS5 levels were lower in both eutopic and ectopic endometrium, while ectopic tissues have the lowest GAS5 expression. When focusing on eutopic tissues from endometriosis patients, we noticed that 10 samples have the GAS5 level in the normal range (GAS5 level in the control group (1.00 ± 0.15), and 7 eutopic samples have lower or similar GAS5 expression than the ectopic tissues. So, the total 33 patients were divided into 2 subgroups according the GAS5 level in eutopic tissues. Subgroup‐1 are patients with normal GAS5 expression in eutopic tissues and lower GAS5 in ectopic tissues ( n  = 10); the subgroup‐2 consisted of patients with lower GAS5 level in both eutopic and ectopic tissues ( n  = 23). Growth arrest‐specific 5 (GAS5) level is lower in patients with endometriosis and the primary endometrial stromal cells (ESCs) exhibit increased viability, dysregulated cell cycle, and deceased apoptosis. Total RNA and protein were extracted from tissue samples from patients with endometriosis and controls without endometriosis. The levels of GAS5 were examined by qRT‐PCR (A and B). The protein levels of p27, cleaved caspase‐3, and cleaved PARP1 were examined by immunoblotting, and then quantified by ImageJ (C). Primary ESCs were separate from endometrium, and then subjected to flow cytometry for cell cycle analysis (D), MTT assay for viability detection (E), and flow cytometry for apoptosis analysis (F). Results were analyzed by one‐way ANNOVA and a p  < 0.05 was considered significant. * p  < 0.05, ** p  < 0.01, *** p  < 0.001 GAS5 has been found to inhibit cell proliferation by upregulating p27 in cancer cells. 21 , 22 To understand whether lower GAS5 level in endometrium relates to p27 dysregulation, the protein level of p27 was examined by immunoblotting. As shown in Figure  1(C) and Figure  S1 (A), the expression of p27 was significant lower in ectopic tissues from all patients. Subgroup‐2 patients also have a lower p27 level in the eutopic tissues. Meanwhile, the apoptosis markers cleaved caspase‐3 and cleaved PARP1 were also lower in ectopic endometrium from all patients when compared with control (Figure  1C ; Figure  S1 A) suggesting that lower GAS5 level may be related to dysregulated cell cycle, proliferation and apoptosis in endometrial cells. Subsequently, the primary ESCs were separated from 5 eutopic endometriums from 5 controls without endometriosis participants, both eutopic and ectopic endometrium from 5 subgroup‐1 patients and 5 subgroup‐2 patients. Flow cytometry and MTT assay were employed to examine cell cycle and viability. As shown in Figure  1(D) and Figure  S1 (B), the ESCs from ectopic endometriums of all patients have reduced distribution in G0/G1 phase and increased distribution in S and G2/M phase. Meanwhile, ESCs from ectopic endometrium have increased cell viability (Figure  1(E) ) and decreased apoptotic cell number (Figure  1(F) ). These results indicated that GAS5 downregulation in ESCs may be related to promoted cell proliferation and cell cycle progression by downregulating p27. To confirm the hypothesis above, the GAS5 levels in ESCs from controls without endometriosis were knocked down by siRNAs, and then the cells were subjected to immunoblotting, flow cytometry and MTT assay. As shown in Figure  2(A) , the GAS5 levels were reduced to 22.1% and 19.4% in siRNAs treated ESCs, and the p27 protein levels were reduced significantly (Figure  2B ). Meanwhile, the GAS5 knockdown ESCs have increased cell viability (Figure  2C ), reduced distribution in G0/G1 phase and increased distribution in S and G2/M phases (Figure  2D ). Reversibly, in GAS5 overexpressed ESCs (Figure  2E ), the cells have increased p27 level (Figure  2F ), and upregulated viability (Figure  2G ). Meanwhile, the GAS5 overexpressed ESCs have increased distribution in G0/G1 phase and decreased distribution in S and G2/M phases (Figure  2H ). Growth arrest‐specific 5 (GAS5) controls endometrial stromal cells (ESCs) cell cycle by regulating p27 expression. GAS5 was knocked down by siRNAs for 48 h, and then the cells were subjected to qRT‐PCR (A), immunoblotting (B), MTT assay (C), and flow cytometry analysis (D). Results were analyzed by one‐way ANNOVA and a p  < 0.05 was considered significant. * p  < 0.05, ** p  < 0.01. ESCs were transfected by GAS5 overexpression vector for 48 h, and then the cells were subjected to qRT‐PCR (E), immunoblotting (F), MTT assay (G), and flow cytometry analysis (H). Results were analyzed by paired t‐test and a p  < 0.05 was considered significant. * p  < 0.05, ** p  < 0.01 It is reported that endometriotic lesions usually have a dense vascularization which is a result of dysregulated angiogenesis processes. 23 , 24 Recently, a growing body of evidences indicated that GAS5 also played a role in the processes of angiogenesis. 25 , 26 To explore the function of GAS5 in the endothelial cells during the pathophysiological processes of endometriosis, the primary endothelial cells were separated from 5 control endometrium, eutopic and ectopic from 5 endometrium of subgroup‐1 patients and 5 endometrium of subgroup‐2 patients. Reduced GAS5 levels were also observed in endometrial cells from the eutopic tissues of all patients (Figure  3A ; Figure  S2 A). Meanwhile, the results of in vitro capillary formation assays indicated that endothelial cells from ectopic endothelium formed more microvascular (Figure  3B ; Figure  S2 B). To explore the underlying mechanisms, the level of vascular endothelial growth factor A, which is a promoter for angiogenesis, and the levels of TIMP3 and TMSB10, which are angiogenesis inhibitors, were examined by immunoblotting. As shown in Figure  3(C) and Figure  S2 (C), the protein level of TIMP3 was significantly reduced in endothelial cells from eutopic and ectopic endothelium, which may contribute to the upregulated angiogenesis. It is reported that TIMP3 is confirmed to be a direct target of miR‐181b. 27 , 28 Since miR‐181 family members have the same seed sequences through which they may interact with the same target sequences, and meanwhile, GAS5 was identified to inhibit miR‐181c through direct interaction in pancreatic cancer cells. 29 So, we hypothesized that TIMP3 may be also a direct target of miR‐181c, and the overexpression of GAS5 may contribute to the downregulation of TIMP3 through repressing miR‐181c in endothelial cells. Primary endothelial cells from patients with endometriosis have reduced growth arrest‐specific 5 (GAS5) level and increased capillary formation capacity. Primary endothelial cells were separate from endometrium tissues, and then subjected to qRT‐PCR (A and D), capillary formation assay (B) and immunoblotting (C). Results were analyzed by one‐way ANNOVA and a p  < 0.05 was considered significant. * p  < 0.05, ** p  < 0.01 To confirm the hypothesis above, the levels of miR‐181b and miR‐181c in the primary endothelial cells were examined by qRT‐PCR, and upregulated miR‐181c was observed in the endothelial cells from patients with endometriosis (Figure  3D ; Figure  S2 D). Meanwhile, the interaction between TIMP3 and miR‐181c was predicted by using TargetScan ( http://www.targetscan.org/vert_70/ ) (Figure  S3 A). A segment of 356 bp TIMP3 3′UTR containing the predicted miR‐181c target sites were cloned into pmirGLO vector to generate the reporter vector. The vector with five nucleotides mutation in the predicted miR‐181c binding site was constructed as the mutant reporter vectors (Figure  S3 A). HEK293T cells were transfected with the TIMP3 reporter vector with or without miR‐181c for 48 h, and then the cells were subjected to dual luciferase assay. As shown in Figure  S3 B, the luciferase activity was significantly reduced in miR‐181c mimic treated cells and upregulated in miR‐181c inhibitor treated cells. When five nucleotides in the predicted miR‐181c binding region were mutated, the luciferase activity was not repressed by miR‐181c. These results indicated that miR‐181c repressed the luciferase expression through targeting the 3′UTR of TIMP3. Subsequently, primary endothelial cells were transfected with miR‐181c mimic or inhibitor for 48 h, and then subjected to immunoblotting. As shown in Figure  S3 (C), the endogenous TIMP3 protein level was repressed by miR‐181c mimic and upregulated by miR‐181c inhibitor. These results indicated that TIMP3 is a direct target of miR‐181c in endothelial cells. To understand the biological function of GAS5 in the endothelial cells, GAS5 was knocked down by siRNAs, and followed by qRT‐PCR to examine the expression of miR‐181c and immunoblotting to detect the protein level of TIMP3. As shown in Figure  4(A) , the level of GAS5 was reduced to 12.4% and 16.5% by the two siRNAs separately. Meanwhile, the miR‐181c level was increased up to 6.22‐ and 5.39‐fold, and the TIMP3 protein levels were repressed by GAS5 siRNAs. When GAS5 was overexpressed in the endothelial cells, the miR‐181c was significantly reduced and the TIMP3 protein level was upregulated (Figure  4B ). These results indicated that GAS5 negatively regulated miR‐181c expression in the endothelial cells and finally controlled the TIMP3 expression. Growth arrest‐specific 5 (GAS5) inhibits the capillary formation through regulating miR‐181c‐TIMP3 axis. The primary endothelial cells were transfected with GAS5 siRNAs for 48 h, and then subjected to qRT‐PCR and immunoblotting (A). The primary endothelial cells were transfected with GAS5 overexpression vector for 48 h, and then the cells were subjected to qRT‐PCR and immunoblotting (B). The primary endothelial cells were transfected with GAS5 overexpression vector, with or without miR‐181c mimic or siTIMP3 for 48 h, and then subjected to capillary formation assay (C) and immunoblotting (D). Results were analyzed by one‐way ANNOVA or paired t ‐test, and a p  < 0.05 was considered significant. * p  < 0.05, ** p  < 0.01 Finally, to confirm that GAS5 regulate endothelial cells angiogenesis through controlling miR‐181c‐TIMP3 axis, the primary endothelial cells were transfected with GAS5 expression vector, with or without miR‐181c mimic or siTIMP3. Forty‐eight hours after transfection, the cells were subjected to in vitro capillary formation assay and immunoblotting. As shown in Figure  4(C) , GAS5 overexpression endothelial cells formed very few capillaries, which were partially restored by miR‐181c or siTIMP3. Meanwhile, TIMP3 expression was significantly upregulated by GAS5, which was totally repressed by miR‐181c or siTIMP3 (Figure  4D ). These results indicated that GAS5 inhibits angiogenesis via modulating miR‐181c‐TIMP3 axis.

A total of 33 Chinese‐Han women with endometriosis in Shijiazhuang Obstetrics and Gynecology Hospital were included in this study (mean age: 29.2 years, range: 24–33 years) (2015–2019). Based on endometrial histology and last menstrual period, they were in the mid‐late proliferative ( N  = 15) and early mid secretory ( N  = 18) phase of the menstrual cycle. All women underwent laparoscopic surgical examination of the abdominal cavity and complete excision of endometriotic tissue. None of the patients had received preoperative hormonal therapy. The presence of endometriosis was suspected either by clinical or ultrasonography examination and confirmed by surgical findings and postoperative pathological examination. Laparoscopic examination of the abdominal cavity excluded the presence of any other pelvic pathology that could potentially confound the data. The menstrual phase was identified according to the day of the reproductive cycle and histological analysis of the endometrium. Samples were collected between April 2015 and December 2016. Twenty‐three asymptomatic women without endometriosis who underwent surgery for laparoscopic tubal sterilization were collected as the control group (mean age: 29.6 years, range: 26–33 years) (June 2015–March 2017). Absence of the disease was confirmed after surgical examination of the abdominal cavity. All patients provided written informed consent prior to participating in this study. This study was approved by The Fourth Hospital of Shijiazhuang ethics committee. All the medical accords of the patients were obtained from Medical records management office of Shijiazhuang Obstetrics and Gynecology Hospital. Each sample was divided and used for cell isolation, total RNA and protein extraction. Endometrial stromal cells (ESCs) from women without endometriosis were cultured as described. 19 Briefly, the minced eutopic endometrium was digested with collagenase type α (0.1%; Sigma‐Aldrich, St. Louis, MO, USA) for 30 min at 37°C and then filtrated through a 200 μmol/L wire sieve to remove debris. Following gentle centrifugation, the supernatant was discarded, and the cells were resuspended in 1:1 formula of Dulbecco's modified Eagle's medium (DMEM) /F‐12 (Hyclone, Logan, UT, USA). ESCs were separated from epithelial cells by passage over a 400 μmol/L wire sieve. The filtered suspension was layered over Ficoll and centrifuged at 2000 rpm for 20 min to further remove leukocytes and erythrocytes. The middle layer of cells was collected and washed with D‐Hanks buffer. The ESCs were placed in a culture flask and allowed to adhere for 20 min. The adherent stromal cells were cultured as a monolayer in flasks with DMEM/F‐12 containing 10% fetal calf serum (FCS; Hyclone), 20 mmol/L HEPES, 100 IU/mL penicillin, and 100 μg/mL streptomycin. For microvascular endothelial cell isolation, endometrial specimens were minced in PBS and digested with type VII high‐purity collagenase (Sigma‐Aldrich). The digest was filtered through a 70‐μm cell strainer to remove undigested tissue fragments. The filtrate was treated with 1 mg/mL of biotinylated UEA‐1 lectin (Sigma‐Aldrich), which binds selectively to microvascular endothelial cells in human endometrial sections. Biotinylated UEA‐1 labeled cells were separated from non‐labeled cells by panning on activated surface/AIS MicroCELLector flasks (Applied Immune Sciences, Menlo Park, CA, USA) coated with streptavidin (Sigma‐Aldrich). The cells were released and cultured in DMEM F‐12 medium, containing endothelial cell growth factor supplemented with 15% stripped fetal calf serum. Total RNA was extracted from tissues and cells using Trizol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instruction. RNA concentration and purity were determined using a model ND‐1000 spectrophotometer (Nanodrop Technologies, Wilmington, DE, USA). Only samples with absorbance ratios 260 nm/280 nm of ~2.0, and 260 nm/230 nm of 1.9–2.2 were considered for inclusion in the study. The level of GAS5 was quantified by qRT‐PCR using SYBR Green Real‐Time PCR Master Mix (Thermal Fisher Scientific), with β‐actin level as the loading control. The miRNAs expression was determined by qRT‐PCR using commercial TaqMan miRNA probes and primers. The level of U6 snRNA was used as loading control. Each sample in each group was measured in triplicate and the experiment was repeated at least three times. The method of −ΔΔct was used to determine the relative level of target genes. The primers' sequences are: ACTB‐F: 5′‐CACCATTGGCAATGAGC‐3′, ACTB‐R: 5′‐AGGTCTTTGCGGATGTC‐3′; GAS5‐F: 5′‐TGAGGAACTTCGGAGAAGA‐3′, GAS5‐R: 5′‐GACACAACTGTCCATAAG‐3′. Protein extracts were boiled in sodium dodecyl sulfate/β‐mercaptoethanol sample buffer, and 30 μg samples were loaded into each lane of 10% polyacrylamide gels. The proteins were separated by electrophoresis, and the proteins in the gels were blotted onto a polyvinylidene fluoride membrane (Amersham Pharmacia Biotech, St. Albans, Herts, UK) by electrophoretic transfer. The membrane was incubated with one of the primary antibodies for 2 h at 37°C. The specific protein‐antibody complex was detected by using horseradish peroxidase conjugated secondary antibodies. Detection by the chemiluminescence reaction was carried using ECL kit (Pierce, Appleton, WI, USA). The β‐actin signal was used as a loading control. Antibody information: rabbit anti‐p27 polyclonal antibody (Cell Signaling Technology, Inc., cat. No. 3686S), rabbit anti‐cleaved caspase‐3 monoclonal antibody (Cell Signaling Technology, Inc., cat. No. 9664S), rabbit anti‐cleaved PARP1 monoclonal antibody (Cell Signaling Technology, Inc., cat. No. 5625S), and mouse anti‐β‐actin monoclonal antibody (Cell Signaling Technology, Inc., cat. No. 3700S). Cell proliferation was estimated by the 3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide (MTT) assay. Cells were seeded in wells of 96‐well plates at low density (2 × 10 3 ) in DMEM‐F12 medium and allowed to attach overnight. The cells were then transfected with siRNA or GAS5 expressing vector for 48 h. Twenty microliters of MTT (5 mg/mL) (Sigma‐Aldrich) were added to each well, and the cells were incubated for a further 4 h. The absorbance was recorded at 570 nm with a 96‐well plate reader after addition of dimethyl sulfoxide. The cells were transfected with siRNA or GAS5 expressing vector for 48 h. The cells were collected and incubated with anti‐annexin V fluorescein isothiocyanate antibody and then stained with propidium iodide. The apoptotic cells were determined by flow cytometry after analysis using FlowJo software (v10.4.1) (Tree Star, Inc., Ashland, OR, USA). The capillary formation assay was processed following the reported protocol with minor revisions. 20 Briefly, growth factor‐reduced Matrigel (BD Biosciences) was seeded into 96‐well plate and solidified at 37°C for 1 h. The primary endothelial cells were then seeded on top of the Matrigel at 40,000 cells/well. Images of each well were taken at 10× magnification using an invert microscope after 8 h of incubation at 37°C. Each condition was assessed in triplicate. Tube formation was quantified using ImageJ, which measured the total tube length, and the total number of tubes, loops, and branching points. Data were analyzed using SPSS Statistical Package version 16 (SPSS Inc., Chicago, IL, USA). Two‐tailed Student's t ‐test was used to calculate statistical significance between the two comparator groups. Values with p  < 0.05 was considered statistically significant.

Although retrograde menstruation might induce endometriosis, but it does not result in endometriosis in all women, suggesting that some other factors, such as cytokines, chemokines, inflammatory cells, and dysregulated gene expression, are required for the implantation and growth of ectopic endometrium. In the patients with endometriosis, ectopic endometrium has been found to exhibit multiple molecular abnormalities, including cell apoptosis, viability, and cell cycle, 30 , 31 , 32 , 33 that contribute to the dissemination, implantation, and proliferation of endometriotic cells at ectopic sites. Lower level of p27 was observed in the endometrium of women with endometriosis, but the reasons for this remains unclear. 33 Here, we examined the expression of p27 in the eutopic and ectopic endometrium and found p27 levels were lower in patients with endometrium, which is consistent with previous findings. 33 We also identified lower GAS5 level in the endometrium of patients with endometriosis. We confirmed that lower GAS5 level in ESCs contributes to cell viability and cell cycle dysregulation through desuppressing p27 expression. Our findings revealed the mechanism of p27 downregulation in the eutopic endometrium. However, we found that p27 level was not only reduced in eutopic tissues. In this study, subgroup‐2 patients have reduced p27, c‐caspase 3, and c‐PARP1 in both eutopic and ectopic tissues. This phenomenon may be a result of other factors such as genetic background or different endometriosis subtypes which needs to be further investigated. A sufficient blood supply is also important for the growth of endometriotic lesions, and dysregulated angiogenesis is necessary for dense vascularization in endometriotic lesions. 23 , 24 We found reduced TIMP3 levels in the primary endothelial cells from the patients with endometriosis and controls without endometriosis, which was consistent with the results reported by Konstantinos et al. 34 TIMP3 has antitumor properties, including inhibition of neovascularization. 35 The present study showed that GAS5 repressed miR‐181c expression in endothelial cells, and that TIMP3 is a direct target of miR‐181c. We identified a regulatory correlation between GAS5 and TIMP3 in endothelial cells, and partially unveiled the anti‐angiogenesis function of GAS5. This study has the following limitations. We cannot yet explain why GAS5 level was lower in patients with endometriosis. Although we unveiled the function of GAS5 in ESCs and endothelial cells, the expression of GAS5 in other types of cells in the endometrium remains obscure and our findings require validation in vivo. In conclusion, lower GAS5 level might relate to endometriosis by regulating cell proliferation, apoptosis, the cell cycle, and angiogenesis.

Endometriosis, defined as the retrograde passage of endometrial tissue outside the uterus, is a common and multifactorial gynecological disorder that affects 10% to 15% of reproductive age women and 20% to 50% of infertile women. 1 It is associated with dysmenorrhea, dyspareunia, chronic pelvic pain, and irregular uterine bleeding; it also tends to recur and associated with ovarian cancer. 2 Laparoscopic surgery is the most prevalent method used to treat ovarian endometriosis, but recurrent rate remains at 40%. 3 Therefore, understanding the underlying mechanisms of the initiation and pathophysiological mechanisms of endometriosis is important for new treatment methods development. Long noncoding RNAs (lncRNAs) are a group of more than 200 nt, nonprotein coding RNAs, some of which are dysregulated in many gynecological states including endometriosis. 4 , 5 , 6 , 7 Growth arrest‐specific 5 (GAS5) is a 725 bp lncRNA which accumulates in growth‐arrested cells. 8 GAS5 is downregulated in several types of cancers and is associated with uncontrolled cell apoptosis. 9 , 10 , 11 Recently, emerging evidences indicate that GAS5 is downregulated in endometrial cancers, but the role in endometriosis is unclear. 12 , 13 MicroRNAs (miRNAs) are negative regulators that repress gene expression by directly binding to the 3′‐untranslated region (UTR) of mRNA. miRNAs play important roles in maintaining normal human body physiologic conditions. Abnormal miRNA expressions are associated with human diseases including psychiatric disorders and malignant cancers. 14 , 15 , 16 Recent miRNA profiles of endometriosis has been established in affected tissues and eutopic endometria, and in endometrium tissues from control patients without endometriosis. 17 , 18 However, the roles of miRNAs in endometriosis pathogenesis remains obscure, and more functional studies are needed to unveil the complex roles of dysregulated miRNAs. Our assessment of GAS5 in 16 normal, 20 eutopic, and 17 ectopic endometrial tissue revealed that GAS5 regulates endometrial cells growth and endothelial cells angiogenesis.

All authors declare no conflict of interest.

Fig. S1 GAS5 is relative low in patients with endometriosis and the primary ESCs exhibit increased viability, dysregulated cell cycle and deceased apoptosis. Total RNA and protein were extracted from tissue samples from patients with endometriosis and controls without endometriosis. The protein levels of p27, cleaved caspase‐3 and cleaved PARP1 were examined by immunoblotting, and then quantified by Image J (A). Primary ESCs were separate from endometrium, and then subjected to flow cytometry for cell cycle analysis (B), MTT assay for viability detection (C), and flow cytometry for apoptosis analysis (D). Results were analyzed by One‐way ANNOVA and a p < 0.05 was considered significant. *p < 0.05, **p < 0.01,***p < 0.001. Fig. S2 Primary endothelial cells from patients with endometriosis have reduced GAS5 level and increased capillary formation capacity. Primary endothelial cells were separate from endometrium tissues, and then subjected to qRT‐PCR (A and D), capillary formation assay (B) and immunoblotting (C). Results were analyzed by One‐way ANNOVA and a p < 0.05 was considered significant. *p < 0.05, **p < 0.01. Fig. S3 miR‐181c inhibits TIMP3 expression via targeting 3'UTR. (A) predicted miR‐181c target sites in TIMP3 3'UTR and mutant vector generation. (B) Wildtype or mutant TIMP3 reporter vector were transfected with miR‐181c mimic or inhibitor into HEK293T cells were 48 hours, and then the cells were subjected to dual luciferase assay. Results were analyzed by paired t‐test and a p < 0.05 was considered significant. *p < 0.05, **p < 0.01. (C) primary endothelial cells were transfected with miR‐181c mimic or inhibitor for 48 hours, and then subjected to immunoblotting.