Hsa_circ_0063526 promotes the development of endometriosis by sponging miRNA-141-5p

Background: Endometriosis can cause various social burdens. Abnormal circular RNAs level lead to changes of related gene expression, thereby mediating the occurrence and development of a series of diseases. The research of circRNA in endometriosis is still in its infancy. This study will explore the role of hsa_circ_0063526 with microRNA-141-5p in the development of endometriosis. Methods: The expression of genes was detected by RT-qPCR. Transwell, wound-healing, EdU assay were performed on endometriosis patient’s End1 / E6E7 cell. PCR and immunohistochemistry were used to detect the expression of candidate regulatory genes in ectopic lesions in endometriosis mice model. Results: therapy reduce endometriosis development. The pleiotropic nature of miR-141-5p and si-hsa_circ_0063526 therapy suggests that multiple complementary mechanisms play a role in endometriosis. These effects suggest that endometriosis may be treated more comprehensively without the systemic side effects of current drugs. But further dose-response and safety studies are needed before they can be used in humans.


Introduction
Endometriosis is a common estrogen-dependent chronic disease, affecting about 10% of childbearingaged women, and 20-50% of the infertile patients were combined with endometriosis, the disease causes infertility, and further malignant transformation (1). The etiology and pathogenesis of endometriosis are not clear. Sampson rst proposed the theory of menstrual blood re ux, the endometrial glandular epithelium and mesenchymal cells ow back into the fallopian tube and the abdominal cavity with menstrual blood, implanted in the basin of abdominal organs such as the ovaries or pelvic peritoneum.
The cells continue to grow there, so form the endometriosis (2). Although endometriosis is a benign disease with benign histomorphology, its clinical behaviors show similar characteristics of malignant tumors' invasion, metastasis, and implantation, so it is also called "benign cancer". Lang proposed the "Eutopic endometrium determinism" of endometriosis, that is, the adhesion, invasion, and growth of endometrium tissue in "ectopic" are determined by the property of the endometrium itself, and the menstrual blood re ux is only a potential auxiliary pathway (2). Recently, the pathogenesis of endometriosis increasingly attributes to a variety of genes and factors closely related to genetics (3).
Treatment for endometriosis has many side effects, including stopping the menstrual cycle during medication, etc (4)(5)(6)(7). New diagnostic indicators and non-hormonal therapy for endometriosis are urgently needed.
MicroRNAs are highly conserved and considered important post-transcriptional regulators (8). At present, relevant studies on the interaction between miRNA and endometriosis are gradually attracting attention.
In our research team, microRNA solexa sequencing and RT-qPCR veri cation were used to test the serum samples of endometriosis patients in the early stage (13)(14)(15)(16), the result showed the expression level of miR-141 was lower than control (17). Studies have demonstrated the down-regulation of the mir-200 family (Include miR-141-5p) in endometriosis patients' plasma (17).
Several studies have explored the use of miR-141-5p to treat a range of diseases. For example, Kim et al previously synthesized miR-141-5p hydrogel that shown effective results in liver cancer mice (18). Rekker et al reported that in the serum of endometriosis patients, the expression of circulating microRNA-200s family including miR-141-5p was lower than control (11). Liang et al. used miR-200c to suppress endometriosis in cells and a mice model (19). Members of our research team also reported that the reduction of miR-141-5p in endometriosis tissue promotes the development of endometriosis mainly by regulating EMT(epithelial-mesenchymal transformation) (20).
CircRNAs exist in eukaryotic cells as covalently closed rings, without 5 'or 3' polarity or polyadenylate, and are more conserved and stable than linear RNAs. It used thought to be a rare type of RNA, just "waste products" from the process of RNA cutting. More and more evidence showed that circRNAs are involved in the proliferation, invasion, and metastasis of various diseases (21)(22)(23)(24). CircRNAs regulate gene expression levels mainly by binding to proteins or sponging microRNAs (25). These molecular interactions will open up new prospects for the diagnosis and treatment of endometriosis. However, the pattern and potential role of circRNAs in the tissue of patients with endometriosis have not been elucidated (20,24,26).
In our previous study, we did microarray circRNAs expression detection and RT-qPCR veri cation in endometriosis versus control, we found in the endometriosis group, hsa_circ_0063526 (circ-RanGAP1) expression level is higher, bioinformatics analysis revealed that circ-RanGAP1 and miRNA-141-5p have complementary binding sites (27). Furthermore, we hypothesized that circRNAs may involve in the development of endometriosis. We aim to explore the correlation between hsa_circ_0063526 with miR-141-5p and endometriosis.

Clinical specimens
Tissue samples were collected from child-bearing-aged women who underwent surgery from February 2019 to January 2020 in the department of gynecology of The Second Xiangya Hospital. Fresh proliferation stage endometriosis cyst samples and endometrial tissue were snap-frozen and stored in a liquid nitrogen container. The study was approved by the ethics committee of The Second Xiangya Hospital. Before the samples were taken, informed consent was assigned by each patient. The clinicopathological data of endometriosis patients, including age, histological subtype, clinical-stage, cyst size was retrieved (Supplementary Table 1-1).
Experimental group: The ectopic endometriosis group of EMs patients, a total of 31 patients, all of whom were diagnosed with ovarian endometriosis by laparoscopy plus histopathology. The endometriosis lesions were collected intraoperatively as the ectopic endometrium group (American Fertility Society stage III-IV).
Control group: 10 patients with secondary infertility caused by fallopian tube obstruction factors con rmed by the combined operation of uterus and abdomen. The pathological diagnosis is proliferation stage endometrium. The sample size was calculated by calculator on powerandsamplesize.com.

1.4 Primers used for RT-qPCR detection
All primers were veri ed by PubMed BLAST and previous articles. The primers used in the experiment are shown in Table 1.

Bioinformatics analysis of hsa_circ_0063526 complementary miRNAs
Miranda, TargetScan, RNA22 v2, RNAhybrid software was used to predict the potential target miRNAs and the target genes of the miRNAs involved. The miRNA complementing hsa_circ_0063526 was predicted, and further research and veri cation were conducted.

Dual-luciferase reporter assays
The recombinant plasmid of double luciferase reporter hsa_circ_0063526-WT and hsa_circ_0063526-MUT was designed and synthesized according to the complementary pairing sequence of miR-141-5p and hsa_circ_0063526-MUT. 293T cells were seeded with 1.0×10 4 cells per well. The miRNA mimics or non-target Control vector and target gene 3 'UTR double reporter vector or mutant vector were diluted in 5µL Opti-MEM medium, and the transfection reagent was diluted with 0.25µL Lipo6000 in 5µL Opti-MEm medium. Before plasmids and Mimics were added to the cells, a 90µL culture medium was added to each well. Mimics transfection concentration was 50nM and plasmid concentration was 50ng/ well. Each group was set with 3 replicates. After 48h of transfection, the medium was extracted and added with PBS. Luciferase reagent was added and shaken for 10min. The medium was transferred to LUMITRAC™ 200 96 well white cell culture plate for uorescence determination. Finally, 30 µL Stop reagent was added to each well for spectrophotometric determination.

Wound healing assays
To detect the migration of cells, we conducted wound healing assays in vitro. Sterile 100µL Tips were used to draw on the con uent monolayer cells in the cell culture Wells. Washed with PBS and put the 6 well plates back into the CO 2 at 37℃ for 48 hours. The cells were observed under an inverted microscope. The width of the scratch was measured and recorded using ImageJ software.

Induction of endometriosis in mice
Endometrium was obtained from 18 childbearing-aged women who underwent hysterectomy for uterine myoma at The Second Xiangya Hospital. According to Noyes et al. criteria (1975), the menstrual cycle was histologically con rmed as the proliferative phase. The patients were not received hormone therapy for at least 6 months. All experimental protocols were approved by The Ethics Committee under the declaration of Helsinki, and all women signed informed consent. The endometrium is obtained by an endometrial specimen collector (J-ES-090500; Cook, USA) and cut into 2 mm diameter fragments and incubated in DMEM, PEN/STREP (Changsha, China) culture medium at 4℃ before implantation. Female nude mice aged 5 to 6 weeks were purchased and reared in the Department of Laboratory Animals in Central South University. They maintained ve animals in each cage during a 12-hour light and 12-hour dark cycle (8 a.m. to 8 p.m.) in a barrier system.
The modi ed endometriosis model previously used in our laboratory and wildly by previous scholars was used to induce endometriosis in 30 mice. According to this protocol, endometrial tissue fragments of patients of the same size(2 mm) were sutured to the peritoneal surface of each mice. Mice were anesthetized by sodium pentobarbital and were laparotomies through a midline incision. Two pieces of endometrial tissue fragments were sutured by 5 − 0 polyglactin suture on the surface of the left and right abdominal wall respectively. Then, the peritoneum and skin were closed.

Evaluation of lesions
After microRNA-141-5p-agomir and si-hsa_circ_0063526-agomir treatment for 2 weeks, the animals were cervical dislocated, and the endometriosis lesions were collected from the abdominal cavity of mice. The volume of lesions was calculated by the formula of minimum diameter² * maximum diameter/2. Left abdominal lesions were preserved in RNA stabilized solution (Qiagen, Germany) for extraction of mRNA, RT-qPCR was used to detect gene expression, and the right abdominal lesions were preserved in 4% polyformaldehyde solution for immunohistochemical study. Endometriosis was observed under a light microscope after H & E staining by a third researcher who was concealed from the group allocation. Image J (NIH, USA) was used for the calculation of the lesion area.

Immunohistochemistry
We used 4% paraformaldehyde to x lesions and para n to embed lesions. Tissue was cut into slices about 5 microns and xed on glass slides, then boiled in sodium citrate (pH = 6) solution in high pressure for 15 minutes to repair the antigen. 10% goat serum was used for antigen blocking. Slides were incubated at 4℃ overnight with anti-E-cadherin (1:1500; Proteintech, USA), anti-N-cadherin (1:1500; Proteintech, USA), anti-Vimentin (1:1500; Proteintech, USA) antibodies to determine protein expression. And dyed with DAB (Well-Biology, Changsha, China). Tissue sections were restained with hematoxylin (Well-Biology, Changsha, China). Stained section images were taken with OLYMPUS BX63 (OLYMPUS, Japan) and analyzed by Image-pro Plus 7.0.

Statistical analysis
All statistical analysis was performed in GraphPad Prism 7.0 software (Lahora, USA). All in vitro experiments were carried out three times. Data are mainly expressed as mean ± standard deviation, mean ± standard error (%), or count. T-test or Wilcoxon rank-sum test were used to compare the two groups. Kruskal-Wallis test or one-way ANOVA was used for comparison of three or more groups. Whether there is a correlation between hsa_circ_0063526 and the expression level of miR-141-5p was detected by Pearson correlation analysis. P < 0.05 was the standard for a statistically signi cant difference.

Role of the funding source
This study was funded by The National Natural Science Foundation of China (81671437,81771558) and Shenzhen People's Hospital. The sponsors helped with design, interpretation of data, and decide submission.

Comparison of relative expression of hsa_circ_0063526 between endometriosis patients and the control group
RT-qPCR was used to analyze the expression level of hsa_circ_0063526 in the endometriosis group versus the control group. The results were shown in Fig. 1. Compared with the control group, the hsa_circ_0063526 level was higher in the endometriosis group. We found that the difference of means ± SEM was 0.6856 ± 0.1735 (P < 0.01).

Bioinformatics analysis and Luciferase assay about hsa_circ_0063526 and miRNA-141-5p
We used bioinformatics analysis (starBase, www.starBase.sysu.edu.cn, RNAhybrid software) to predict the downstream target miRNA of hsa_circ_0063526 ( Figure.2a). We performed a double luciferase reporter assay on the binding site (Fig. 2b&c). The double luciferase assay showed that the cotransfection of hsa_circ_0063526 wild-type and miR-141-5p mimic signi cantly reduced luciferase activity compared with the control group. Compared with the co-transfection of hsa_circ_0063526 mutant and miR-141-5p mimic, the reporter gene expression was rescued, and it indicated that miR-141-5p mimic could bind to hsa_circ_0063526.

correlation between miR-141-5p and hsa_circ_0063526 expression
RT-qPCR was used to determine the relative expression of miR-141-5p between endometriosis patients and the control group (Fig. 3a). We found the difference of means ± SEM was 0.5994 ± 0.08425 (P < 0.01). Pearson correlation analysis of the expression levels of hsa_circ_0063526 and miR-141-5p in the lesion tissue of endometriosis showed a negative correlation (r= -0.4267, p = 0.02, Fig. 3b). Pearson correlation analysis further indicated the interaction between hsa_circ_0063526 and miR-141-5p.
Due to the overexpression of hsa_circ_0063526, we constructed different siRNAs to knockdown hsa_circ_0063526. To determine the knockdown e ciency, three uorescent small interfering RNAs were designed and synthesized. Fluorescent images showed successful transfection of siRNA into End1/E6E7 cells. Results showed that interference sequence 3 most signi cantly reduced the expression of hsa_circ_0063526 ( g.s1).
After transfection with si-hsa_circ_0063526, the expression level of hsa_circ_0063526 in End1/E6E7 cells in the si-hsa_circ_0063526 group was signi cantly decreased (P < 0.05, Fig. 3c), and the expression level of miR-141-5p was signi cantly increased (P < 0.05, Fig. 3d), compared with that in the Blank group and the si-NC group. The results showed that hsa_circ_0063526 could negatively regulate the expression level of miR-141-5p in End1/E6E7 cells.

Down-regulation of hsa_circ_0063526 can inhibit the proliferation, invasion, and migration of endometriosis cells
Migration, invasion, and proliferation of ectopic endometrial cells are essential pathological processes for the development of endometriosis. Therefore, 72 hours after transfection of si-hsa_circ_0063526, this study further explored the changes of cell migration and invasion and cell proliferation ability after downregulation of hsa_circ_0063526. First of all, the ability to invasion by Transwell experiments, after End1 / E6E7 cell transfected with hsa_circ_0063526 siRNA, the cells entering the lower chamber was relatively lesser compared to the control group, cell invasion ability decreased. Difference between means ± SEM = 70.75 ± 10.99 (P < 0.05, Fig. 4). The ability of cell migration was tested by wound healing experiment, after End1 / E6E7 cell transfected hsa_circ_0063526 siRNA, the width of the cell scratch was wider than that of the control group. The ability of cell migration was decreased. Difference between means ± SEM=-333.3 ± 72.12 (P < 0.05, Fig. 5). Further, the effect of down-regulation of hsa_circ_0063526 on cell proliferation was detected by using EdU cell proliferation assay. 72 h after siRNA transfection, compared with the si-NC control group, the EdU test results showed that the proportion of End1/E6E7 cells in the mitotic stage was lesser after the down-regulation of hsa_circ_0063526, and the proliferation capacity of the cells was decreased after siRNA transfection. Difference between means ± SEM = 32.38 ± 6.639 (P < 0.05, Fig. 6). PCR and ELISA results showed that the expression of E-cadherin mRNA, an important epithelial marker of EMT, was up-regulated (P < 0.05, Fig. 7).
3.5 Down-regulation of miR-141-5p can rescue the proliferation, invasion, and migration of endometriosis inhibited by hsa_circ_0063526.
The above experimental results showed that knockdown hsa_circ_0063526 inhibited the development of endometriosis. Inhibition of microRNA-141-5p can rescue the inhibitory effect brought by knockdown of hsa_circ_0063526, that is to say, hsa_circ_0063526 promotes the development of endometriosis through sponging (inhibition) of microRNA-141-5p.

Comparison of miR-141-5p treatment and pathological changes
No adverse reactions were observed in mice treated with miR-141-5p. At the end of miRNA-141-5p treatment, mice were sacri ced by cervical dislocation and endometriosis lesions were collected. First, we evaluated the volume of the lesions between the miR-141-5p treatment group and the control group. All lesions were cystic. Lesions in miR-141-5p treatment groups were relatively small. (Figure8 a,b,c) Besides, the pathological difference of the endometrial part is further compared by the histological area.
The histological area of all lesions was observed under the microscope to exclude muscle and destructed parts. The histological area of endometriosis in abdominal lesions of the miR-141-5p treatment group was signi cantly smaller than the two control groups (p < 0.05, Figure 8 d-g).

Expression difference of endometriosis-related genes
The differences in gene expression related to endometriosis were detected by RT-qPCR. The expression of some genes known to promote the development of endometriosis decreased in the miR-141-5p group. Expression of N-cadherin, Vimentin, K-ras, MAPK-14, ER-α, ER-β, ZEB-1 was decreased in the miR-141-5p treatment group (P 0.05 Figure 9). The quantitative decrease in gene expression is 6.5-fold for K-Ras, 1.81-fold for MAPK14, 13.9 fold for ER-α, 97.2 fold for ER-β, 4.63 fold for N-Cadherin, 2.37 fold for Vimentin, 3.98-fold for ZEB-1 in 141 treated group compared to saline group. Expression levels of miR-141, E-Cadherin, ZO-1(Zona Occludens 1) were increased in the miR-141-5p treatment group (P 0.05, Figure 9). Expression levels of Notch, IL-6 was not statistically signi cant between treatment and control groups (P 0.05, Figure 9).
Immunohistochemistry results showed that the protein expression level of the miR-141-5p treatment group was signi cantly changed. The intensity of N-cadherin and vimentin in stromal cells decreased in the miR-141-5p treatment group( Figure 10).

Comparison of si-hsa_circ_0063526 treatment and pathological changes
No adverse reactions were observed in mice treated with si-hsa_circ_0063526 agomir. At the end of si-hsa_circ_0063526 agomir treatment, mice were sacri ced by cervical dislocation and endometriosis lesions were collected. We evaluated the volume of the lesions between the si-hsa_circ_0063526 agomir treatment group and the control group. All lesions were cystic. Lesions in si-hsa_circ_0063526 agomir treatment groups were relatively small. (Figure11a) Besides, the pathological difference of the endometrial part is further compared by the histological area.
The histological area of all lesions was observed under the microscope to exclude muscle and destructed parts. The histological area of endometriosis in abdominal lesions of the miR-141-5p treatment group was signi cantly smaller than the two control groups (p < 0.05, Figure 11b-e).
3.9 Endometriosis-related gene expression differences RT-qPCR was used to detect the gene expression differences associated with endometriosis. Compared with the control group, some genes known to promote the development of endometriosis were decreased in the si-hsa_circ_0063526 agomir group. MAPK-14, K-ras, N-cadherin, Vimentin, ER-α, ER-β were decreased in the si-hsa_circ_0063526 agomir group (P < 0.05 Figure 12). The decrease folds of quantitative gene expression were 5.4 times of K-ras, 3.4 times of MAPK14, 7.5 times of ER-α, 5.2 times of ER-β, and 2.1 times of N-cadherin. The expression levels of E-cadherin and Zona occludens 1 (Zona Occludens 1) in the si-hsa_circ_0063526 agomir group were increased (P<0.05, Figure 12).
Immunohistochemical results showed that E-cadherin staining intensity was signi cantly increased in the si-hsa_circ_0063526 treatment group ( Figure 13). The results are consistent with the overall research results, indicating that our research results have good stability and reliability.

Discussion
Endometriosis is a common estrogen-dependent chronic disease that affects about 10% women of childbearing age. Its histomorphology is benign, but its clinical behavior shows similar characteristics of malignant tumor' invasion, metastasis, and implantation (3).
The study of circRNA in endometriosis is still in its infancy, and its pattern and potential role in endometriosis tissues have not been elucidated. (20,28,29). Most of the studies on circRNA are preliminary, and more in-depth studies are needed to provide a su cient basis for circRNA to be a diagnostic marker or therapeutic target for endometriosis (21,24,25,30).
We veri ed hsa_circ_0063526 was increased in endometriosis lesions, and the invasion, migration, and proliferation of End1/E6E7 cells in endometriosis patients were decreased after knockdown of hsa_circ_0063526. At the same time, bioinformatics has shown that hsa_circ_0063526 may bind miR-141-5p as a target, the dual-luciferase reporter assay further suggesting miR-141-5p can be combined with hsa_circ_0063526 and in patients' endometriosis lesions the expression of both has a negative correlation, our preliminary studies suggest that miR-141-5p regulates the development of endometriosis, It suggests that hsa_circ_0063526 may play a regulatory role in the occurrence and development of endometriosis through miR-141-5p.
Endometriosis is also known as "benign cancer". It is characterized by aggressive metastatic growth and a high risk of recurrence, similar to a malignant tumor (31,32). So, they have something in common in terms of pathogenesis. EMT endows cells with the ability to invade and metastasize, including reducing apoptosis, suppressing the immune response, and obtaining stem cell characteristics. It not only plays a key role in embryonic development but is also involved in tissue healing, organ brosis, and the development of cancer (33). Simpson's theory of menstrual blood re ux suggests that EMT is one of the pathogenesis of endometriosis. Small mesothelial cells after EMT no longer provide a lamellar protective barrier between the basal layer and the lacunae. Because there is no protective barrier, endometrial cells tend to adhere to the peritoneal matrix, resulting in endometriosis. The expression of epithelial markers in endometriosis was down-regulated and the expression of mesenchymal markers was up-regulated (34). In addition to changes in cell markers, there is also evidence that EMT plays a key role in endometriosis. For example, cells die as soon as they leave ECM(extracellular matrix) or stick to the wrong place, a phenomenon known as anoikis, which is one of the challenges to Simpson's re ux hypothesis. EMT can resist apoptosis and assist the metastasis of ectopic lesions (35). Second, the endometrium is formed by the transformation of the stromal cells during the development of the urogenital system of the embryo.
Due to the retention of some stromal origin imprints, endometrial epithelial cells tended to revert to the original stromal state through EMT (36,37). We found that the EMT activity in the si-hsa_circ_0063526 group was decreased, and the expression level of E-cadherin, the epithelial marker, was increased.
The Ras/MAPK signaling pathway is important in embryo development, differentiation, proliferation, cell death, etc. K-ras is upregulated in endometriosis. K-ras activation is a classical method to establish a mouse model of spontaneous endometriosis (38,39). Here, we found that the expression level of K-ras was lower in the miR-141-5p and si-hsa_circ_0063526 group.
Compared with linear RNA, circRNAs have a more stable structure and can resist the degradation of multiple RNA enzymes. Therefore, circRNAs could be a diagnostic and therapeutic target for endometriosis (40,41). Currently, no studies have been reported on the role of hsa_circ_0063526 in endometriosis, but other circRNA studies have shown that circRNA has potential value in prognosis prediction or early diagnosis of endometriosis. For the treatment of endometriosis, several possible therapeutic targets may be extracted from this regulatory pathway.
Endometriosis is considered to be an estrogen-dependent disease, ER-α and β have important roles in endometriosis. A lot of studies also used microRNA to treat estrogen-dependent disease in breast cancer.
Let-7 miRNAs were down-regulated in breast cancer-initiating cells. In these cells, the recovery of miR-let-7 resulted in a decrease of proliferation in NOD/SCID mice in vitro, and decrease tumorigenesis and metastasis (42). ER-α β expression was signi cantly inhibited with miR-141-5p and si-hsa_circ_0063526 suggesting that si-hsa_circ_0063526/miR-141-5p pathway blocks estrogen stimulation in the treatment. These data reveal that miR-141-5p and si-hsa_circ_0063526 therapy may speci cally block the sex hormone pathway in patients with endometriosis without systemic side effects of estrogen de ciency.
In this study, miR-141-5p and si-hsa_circ_0063526 were used for local therapy. After systematic administration, oligonucleotides are di cult to reach the desired tissue because they are degraded by the liver and removed from the blood. (43). Therefore, many vectors have been used to improve the stability of oligonucleotides. However, these drugs are di cult to achieve targeted drug delivery (18). We believe that miR-141-5p and si-hsa_circ_0063526 have the best therapeutic effect as an intraperitoneal injection (44).
However, since the animal model could not completely simulate the patient's condition, the model could not re ect the role of in ammation, angiogenesis, and other activities in endometriosis, and the side effects of the therapy could not be observed as thoroughly as the patient. Before being used in humans, dose-response and safety studies should be conducted.
In summary, miR-141-5p and si-hsa_circ_0063526 therapy reduce endometriosis development. The pleiotropic nature of miR-141-5p and si-hsa_circ_0063526 therapy suggests that multiple complementary mechanisms play a role in endometriosis. These effects suggest that endometriosis may be treated more comprehensively without the systemic side effects of current drugs. But further dose-response and safety studies are needed before they can be used in humans. writing the initial draft of the paper. Lipin Li contributed to re ning the ideas, carrying out additional analyses. Xiaoling Fang nalizes the revision of this paper. Xinyue Zhang has veri ed the underlying data.

Declaration of interests
No con ict of interest exists.