Oncogenic Role of miR-200c-3p in High-Grade Serous Ovarian Cancer Progression via Targeting the 3′-Untranslated Region of DLC1

High-grade serous ovarian cancer (HGSC) is the most common ovarian cancer with highly metastatic properties. A small non-coding RNA, microRNA (miRNA) was discovered to be a major regulator in many types of cancers through binding at the 3′-untranslated region (3′UTR), leading to degradation of the mRNA. In this study, we sought to investigate the underlying mechanisms involved in the dysregulation of miR-200c-3p in HGSC progression and metastasis. We identified the upregulation of miR-200c-3p expression in different stages of HGSC clinical samples and the downregulation of the tumor suppressor gene, Deleted in Liver Cancer 1 (DLC1), expression. Over expression of miR-200c-3p in HGSC cell lines downregulated DLC1 but upregulated the epithelial marker, E-cadherin (CDH1). Based on in silico analysis, two putative binding sites were found within the 3′UTR of DLC1, and we confirmed the direct binding of miR-200c-3p to the target binding motif at position 1488–1495 bp of 3′UTR of DLC1 by luciferase reporter assay in a SKOV3 cell line co-transfected with vectors and miR-200c-3p mimic. These data showed that miR-200c-3p regulated the progression of HGSC by regulating DLC1 expression post-transcription and can be considered as a promising target for therapeutic purposes.


Introduction
Ovarian cancer is the fifth most common cancer in women, and is the leading cause of death for patients with gynecological cancer. High levels of genetic and molecular heterogeneity are reported for epithelial ovarian cancer (EOC), and these epithelial cells can be transformed into different histological subtypes, either resembling the fallopian tube layer (serous, 52%), endometrium (endometrioid, 10%), endocervix glands (mucinous, 6%), or other vaginal cells (6%); the rest are rare histotypes [1,2].
Serous ovarian cancer (SOC) is classified into two categories, namely low-grade serous carcinoma (LGSC), which is a prototype for type I ovarian cancer and comprises less than five percent of EOC cases, and high-grade serous carcinoma (HGSC), which is a prototype for type II ovarian cancer. LGSC patients are often diagnosed at an early stage (stage I or II) with low mortality rates and tumor growth. Hence, the prognosis for this category is better, compared to HGSC [3][4][5]. About 70% to 80% of the total EOC subtypes are HGSC and are frequently associated with TP53 mutation [6]. Patients are often diagnosed at

Ethics Statement
The institutional ethics committee approved this study involving human tissue collection (reference no: UKM PPI/111/8/JEP-2017-268). All patients involved provided signed informed consent.

Human Tissues and Cell Lines
Twenty-nine specimens of early (stage I and II) and late (stage III and IV) stages of HGSC taken from primary ovarian tumor, and 21 unmatched normal ovarian surface tissues obtained from the Department of Obstetrics and Gynecology, Faculty of Medicine, Universiti Kebangsaan Malaysia Medical Centre (UKMMC), Malaysia, were snap frozen and stored at −80 • C. Written consent for tissue donation was obtained from every patient before surgery. The unmatched normal tissues were obtained from patients with a benign gynecological disorder who underwent total abdominal hysterectomy with bilateral salpingo-oophorectomy. Patients who underwent chemotherapy or those with secondary tumors were excluded from this study. All the cancer samples were then histologically verified by a pathologist using hematoxylin and eosin (H&E) staining to contain more than 80% malignant cells, while no cancer or inflammatory cells were detected in normal tissues.

RNA isolation and quantitative Real Time-PCR (qRT-PCR)
Total RNA containing miRNA was isolated by using a miRNeasy kit (Qiagen, Germantown, CA, USA) according to the manufacturer's protocol. The purity of the total RNA was assessed using a NanoDrop spectrophotometer (Thermo Scientific, Waltham, MA, USA) with criteria of purity met by A260/280 ratios of 1.8-2.1. Total RNA was then converted into cDNA and micDNA by using a RT2 First Strand kit and a miScript II RT kit (Qiagen, Germantown, CA, USA), respectively. Quantitative RT-PCR (qRT-PCR) was performed using an RT2 SYBR green PCR kit (Qiagen, Germantown, CA, USA) for mRNA and a miScript SYBR green PCR kit (Qiagen) for miRNA, using a CFX96 Real-Time System (BioRad, Hercules, CA, USA). We purchased primers for miR-200c-3p (catalog no. MS00003752) and the reference gene, RNU6 (catalog no. MS00033740), and primers for mRNA DLC1 (catalog no. PPH00438B), CDH1 (catalog no. PPH00135F), and the reference gene PPIA (catalog no. PPH01319G) from Qiagen. The relative expression was calculated using the 2-∆∆CT method.

Transient miR-200c-3p transfection
Each HGSC cell line, CAOV3 and SKOV3, was cultured in 6-well plates with 5 × 105 cells/well with antibiotic-free culture medium until reaching 70% confluence. Cells were then transiently transfected with 150 nM (with transfection efficiency 90%, Supplementary Figure S1) miRCURY LNA™ miR-200c-3p mimic or inhibitors, and their respective negative control (Exiqon, Vedbaek, Denmark), together with Lipofectamine ® RNAimax (Invitrogen) and Opti-MEM ® I without serum and antibiotic (Invitrogen, Carlsbad, CA, USA) for 48 h. Cells were then collected and we immediately continued with RNA extraction and protein extraction. The DLC1 mRNA and protein expression analyses were made using qRT-PCR and Western blot, respectively. All treatments were tested in biological triplicate.

Protein Extraction and Western Blot Assay
Each cell line was lysed with RIPA buffer with 0.1 M dithiothreitol, protease, and phosphatase inhibitors (Thermo Scientific, Waltham, MA, USA). The total protein was quantified using a BCA protein assay (Thermo Scientific, Waltham, MA, USA). The protein (25 µg) from each sample was loaded into 10% SDS-PAGE gel, separated by electrophoresis, and transferred onto a PVDF membrane (Biorad, Hercules, CA, USA). The membrane was blocked with 5% skimmed milk and incubated with primary antibody against DLC1 (Santa Cruz, CA, USA) and β-actin (Abcam, Cambridge, MA, USA) at 1:2000 overnight at 4 • C. Incubation with secondary antibody HRP (Abcam, Cambridge, MA, USA) at 1:5000 was performed on the next day for an hour at room temperature and visualized by chemiluminescence (ECL, Abcam, USA) with a gel doc Amersham Imager 6000 (GE Life Sciences, Buckinghamshire, UK). Band intensity on the PVDF membranes was analyzed with Image J software and normalized to β-actin level.

Prediction of miRNA target sites in the 3 UTR of DLC1
There are many methods and web-based services recently developed for the integrated analysis of miRNA and mRNA expression data. MicroRNA.org [16] and TargetScan [17] were used to find the target prediction of miR-200c-3p on the 3 UTR of DLC1.

Cloning of the 3 UTR of DLC1
The 3 UTR of DLC1 (nucleotides 4842-7412 of NM_182643.3) containing two putative miR-200c-3p binding sites were amplified from human genomic DNA. Then, reverse complement of 3 UTR of DLC1, and two additional wild-type (WT) sequences containing one of the binding sites were synthesized by PCR with the primers ( Table 1). All forward and reverse primers contained SbfI and XbaI restriction enzyme sequence (underlined), respectively, at the 5 end to produce flanking restriction sites at the end of the PCR products to facilitate cloning into the pmirGLO dual-luciferase miRNA target expression vector (Promega, Madison, WI, USA). All the amplicons and vectors were digested with SbfI and XbaI, ligated, and transformed into 10-beta competent E. coli (New England Biolabs, Ipswich, MA, USA) for cloning purposes.  Another four constructs with an isolated sequence containing two small regions of 3 UTR, namely, WT_A (46bp) and WT_B (53bp), and two mutant-type constructs with substitution mutated on the first three nucleotides of the seed sequence, i.e., MT_A (46bp) and MT_B (53 bp) were constructed ( Table 2). All four vectors were constructed and validated by Gen Script, USA.

Luciferase Reporter Assay
SKOV3 cells (4 × 105 cells/well) were cultured in 96-well plate and co-transfected with three pmol miR-200c-3p mimic or negative control, and 100 ng of either wild-type or mutanttype pmirGLO plasmid for 48 h using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). Cells were then assayed with a Dual-Luciferase Reporter Assay System (Promega, Madison, WI, USA) according to the manufacturers protocols. All treatments were tested in biological triplicate.

Statistical Analysis
All data were expressed as mean ± standard error of mean (SEM). Statistical analyses were performed using GraphPad Prism 5.0 (GraphPad Software, Inc., La Jolla, CA, USA). Student's t-test analyzed the statistical significance between groups. Pearson's correlation analysis determined the correlation between miR-200c-3p vs. DLC1.A difference with p < 0.05 was considered to be statistically significant.

Clinical and Pathological Characteristics of Samples
We collected 29 primary ovarian tumor tissues from HGSC patients (17 and 12 samples were at the early and late stage, respectively) who met the inclusion criteria and 21 unmatched normal ovarian surface epithelium tissues from patients with benign gynecological diseases who underwent total abdominal hysterectomy with bilateral salpingooophorectomy, as the control. All the tissues obtained were histologically validated using hematoxylin and eosin (H&E) staining by a pathologist to confirm that the normal ovary tissue samples were free from cancer or inflammatory cells and that the cancer tissues contained more than 80% malignant cells. The inclusion criteria included patients with primary epithelial ovarian cancer, and who had not undergone chemotherapy. Table 3 shows a summary of the collected clinical samples.

Expression Levels of miR-200c-3p and DLC1 in Clinical Specimens.
The relative expression level of miR-200c-3p in HGSC tissues was found to be significantly upregulated, by 19.13-fold compared to normal tissues (p < 0.0001) (Figure 1a). The trend of the relative expression level of this miRNA was found to be significantly upregulated in early stage (I and II) by 14.89-fold, and followed by the late stages (III and IV) by 25.14-fold as compared to the normal (p < 0.0001). This showed the expression of miR-200c-3p increased with the progression of the disease (Figure 1b).
On the other hand, the relative expression level of DLC1 in cancer tissues was found to be significantly downregulated, by 0.4-fold compared to the normal tissues (p < 0.0001) ( Figure 1c). Similar results were found in different stages of HGSC, where the relative expression level was downregulated significantly in both early, by 0.37-fold (p < 0.001), and late stage, by 0.35-fold (p < 0.01), compared to normal (Figure 1d). expression level was downregulated significantly in both early, by 0.37-fold (p < 0.001), and late stage, by 0.35-fold (p < 0.01), compared to normal (Figure 1d).
The Pearson correlation coefficient (r) between the relative expression levels of miR-200c-3p and DLC1 was found to have a moderate correlation, with the value of r = −0.5217 and p = 0.002 (Figure 1e). This result showed that there was a negative correlation between miR-200c-3p and DLC1 expression, which suggested DLC1 was a target of miR-200c-3p in the HGSC samples.  The Pearson correlation coefficient (r) between the relative expression levels of miR-200c-3p and DLC1 was found to have a moderate correlation, with the value of r = −0.5217 and p = 0.002 (Figure 1e). This result showed that there was a negative correlation between miR-200c-3p and DLC1 expression, which suggested DLC1 was a target of miR-200c-3p in the HGSC samples.

DLC1 Is a Direct Target of miR-200c-3p in an In Vitro Model of HGSC
The bioinformatic prediction tools, microRNA.org (http://www.microrna.org, accessed on 20 May 2021) and TargetScan (http://www.targetscan.org, accessed on 20 May 2021) were accessed on 6 February 2019 have suggested that there are two putative binding sites for miR-200c-3p in the 3 UTR of DLC1 (Supplementary Figure S2), which are an 8-mer seed CAGTATTA (1488-1495 bp from the start of 3 UTR) and a 7-mer-m8 seed CAGTATT (2519-2526 bp from the start of 3 UTR) of the transcript variant 1, as shown in Figure 2a. To investigate the relationship between miR-200c-3p and DLC1 expression through miRNA:mRNA binding at the 3 UTR, a firefly luciferase reporter vector was constructed by inserting a wild-type (WT) 3 UTR of DLC1 containing two binding sites in the 3 end of the luciferase gene and the sequence was confirmed by Sanger sequencing (Supplementary Figure S3). The data indicated that overexpression of miR-200c-3p remarkably reduced the luciferase activity of DLC1 3 UTR by 80% (p < 0.0001) (Figure 2b). correlation was shown between miR-200c-3p and DLC1 in both early and late stage of high grade serous carcinoma samples (e). Significant values are indicated as follows: ** p < 0.01 and *** p < 0.001.

DLC1 is a Direct Target of miR-200c-3p in an In Vitro Model of HGSC
The bioinformatic prediction tools, microRNA.org (http://www.microrna.org, accessed on 20 May 2021) and TargetScan (http://www.targetscan.org, accessed on 20 May 2021) were accessed on 6 February 2019 have suggested that there are two putative binding sites for miR-200c-3p in the 3'UTR of DLC1 (Supplementary Figure S2), which are an 8-mer seed CAGTATTA (1488-1495 bp from the start of 3'UTR) and a 7-mer-m8 seed CAGTATT (2519-2526 bp from the start of 3' UTR) of the transcript variant 1, as shown in Figure 2a. To investigate the relationship between miR-200c-3p and DLC1 expression through miRNA:mRNA binding at the 3'UTR, a firefly luciferase reporter vector was constructed by inserting a wild-type (WT) 3'UTR of DLC1 containing two binding sites in the 3' end of the luciferase gene and the sequence was confirmed by Sanger sequencing (Supplementary Figure 3). The data indicated that overexpression of miR-200c-3p remarkably reduced the luciferase activity of DLC1 3'UTR by 80% (p < 0.0001) (Figure 2b).  However, in order to identify the specific binding area of miR-200c-3p on DLC1 3 UTR, two firefly luciferase reporter vectors containing site A or site B, namely WT-A or WT-B, respectively, were constructed. Figure 3a shows that overexpression of miR-200c-3p significantly reduced the luciferase activity, by 44%, in WT-A (p < 0.05), but not in WT-B. This showed that the binding area of miR-200c-3p could be within the 1672 bp of WT-A.
tivity in the presence of miR-200c-3p mimic (p < 0.001). To be more specific, the inhibition of luciferase activity was mainly observed by the first putative binding site, which is site A, since the deletion could result in the loss of over 90% of luciferase inhibitory activity. These results confirmed that miR-200c-3p repressed the DLC1 gene through its binding in the 3'UTR at site A.
To confirm the specific target binding site of miR-200c-3p on 3′UTR, a small region containing 46 bp for site A, and 53 bp for B was isolated into different vectors, namely, WT-A (46bp) and WT-B (53bp), respectively. Additionally, two vectors containing three substitution mutations on the first three nucleotides (5′-3′) of seed sequences were constructed. With the presence of miR-200c-3p, luciferase activities were significantly reduced by 25% and 20% in both WT-A (46bp) and WT-B (53bp), respectively (p < 0.05). However, the inhibitory effects were attenuated in both mutant-type containing vectors (Figure 3f). All the constructed vectors are shown in Supplementary Figure 4.

Regulation of DLC1 by miR-200c-3p in HGSC Cell Lines
To obtain additional evidence on the role of miR-200c-3p towards DLC1, we investigated further, by transiently transfected miR-200c-3p mimic or inhibitor in HGSC cell lines, which were CAOV3 (primary tumor) and SKOV3 (isolated from ascites), together with their respective negative controls. Figure 4a shows that the relative expression of DLC1 mRNA was significantly downregulated (p < 0.001) in both CAOV3 and SKOV3 cell lines transfected with miR-200c-3p mimic, by 0.06-fold and 0.72-fold, respectively. On the other hand, the relative expression of DLC1 mRNA was shown to be significantly upreg- We further investigated whether these two binding sites were repressed by miR-200c-3p; three additional luciferase vectors were constructed containing the mutant DLC1 3 UTR sequence with deletions for both binding sites (MT-AB) and either of the two binding sites (MT-A or MT-B). As shown in Figure 3b, MT-A and MT-AB attenuated the inhibitory effects on the luciferase activity in the presence of miR-200c-3p mimic. However, the luciferase activity was significantly inhibited for MT-B construct of the luciferase activity in the presence of miR-200c-3p mimic (p < 0.001). To be more specific, the inhibition of luciferase activity was mainly observed by the first putative binding site, which is site A, since the deletion could result in the loss of over 90% of luciferase inhibitory activity. These results confirmed that miR-200c-3p repressed the DLC1 gene through its binding in the 3 UTR at site A.
To confirm the specific target binding site of miR-200c-3p on 3 UTR, a small region containing 46 bp for site A, and 53 bp for B was isolated into different vectors, namely, WT-A (46bp) and WT-B (53bp), respectively. Additionally, two vectors containing three substitution mutations on the first three nucleotides (5 -3 ) of seed sequences were constructed. With the presence of miR-200c-3p, luciferase activities were significantly reduced by 25% and 20% in both WT-A (46bp) and WT-B (53bp), respectively (p < 0.05). However, the inhibitory effects were attenuated in both mutant-type containing vectors (Figure 3f). All the constructed vectors are shown in Supplementary Figure S4.

Regulation of DLC1 by miR-200c-3p in HGSC Cell Lines
To obtain additional evidence on the role of miR-200c-3p towards DLC1, we investigated further, by transiently transfected miR-200c-3p mimic or inhibitor in HGSC cell lines, which were CAOV3 (primary tumor) and SKOV3 (isolated from ascites), together with their respective negative controls. Figure 4a shows that the relative expression of DLC1 mRNA was significantly downregulated (p < 0.001) in both CAOV3 and SKOV3 cell lines transfected with miR-200c-3p mimic, by 0.06-fold and 0.72-fold, respectively. On the other hand, the relative expression of DLC1 mRNA was shown to be significantly upregulated in both CAOV3 cells, by 1.69-fold (p < 0.01), and SKOV3 cells, by 6.76-fold (p < 0.0001), after transfection with miR-200c-3p inhibitor (Figure 4b). However, there was no significant difference in DLC1 protein level in both cell lines transfected with miR-200c-3p mimic or inhibitor compared to the control (Figure 4c). These results gave a reliable indicator of the role of miR-200c-3p in regulating DLC1 at the mRNA level in HGSC progression.

Regulation of miR-200c-3p in Epithelial-Mesenchymal Transition through Modulation of DLC1 Expression
To explore the association between miR-200c-3p regulation on DLC1 and its effect on the progression and metastasis of HGSC, CAOV3 and SKOV3 cell lines were transiently transfected with miR-200c-3p mimic or inhibitor. The relative expression levels of EMT/MET markers, E-cadherin (CDH1), in addition to DLC1 were assessed in comparison to their respective negative controls. Figure 5a shows that in the presence of miR-200c-3p mimic, the relative expression of CDH1 was significantly upregulated in CAOV3 cells, by 1.27-fold (p < 0.05), and in SKOV3 cells, by 2.32-fold (p < 0.01), in comparison to the controls. However, in Figure 5b, when the cells were transfected with miR-200c-3p inhibitors, there was a significant downregulation in the expression of CDH1 in both CAOV3 and SKOV3 cells, by 0.72-fold and 0.44-fold, respectively (p < 0.01). These results showed that miR-200c-3p could be a regulator for EMT/MET through the modulation of DLC1 expression, which plays an essential role in HGSC progression and metastasis.

Regulation of miR-200c-3p in Epithelial-Mesenchymal Transition through Modulation of DLC1 Expression
To explore the association between miR-200c-3p regulation on DLC1 and its effect on the progression and metastasis of HGSC, CAOV3 and SKOV3 cell lines were transiently transfected with miR-200c-3p mimic or inhibitor. The relative expression levels of EMT/MET markers, E-cadherin (CDH1), in addition to DLC1 were assessed in comparison to their respective negative controls. Figure 5a shows that in the presence of miR-200c-3p mimic, the relative expression of CDH1 was significantly upregulated in CAOV3 cells, by 1.27-fold (p < 0.05), and in SKOV3 cells, by 2.32-fold (p < 0.01), in comparison to the controls. However, in Figure 5b, when the cells were transfected with miR-200c-3p inhibitors, there was a significant downregulation in the expression of CDH1 in both CAOV3 and SKOV3 cells, by 0.72-fold and 0.44-fold, respectively (p < 0.01). These results showed that miR-200c-3p could be a regulator for EMT/MET through the modulation of DLC1 expression, which plays an essential role in HGSC progression and metastasis.

Discussion
Increasing evidence has shown that aberrant changes in miRNAs, especially the miR-200 family, play critical roles in the progression and carcinogenesis of ovarian cancer [18,19]. Previously, we shortlisted upregulated miRNAs in SOC samples using miRNA

Discussion
Increasing evidence has shown that aberrant changes in miRNAs, especially the miR-200 family, play critical roles in the progression and carcinogenesis of ovarian cancer [18,19]. Previously, we shortlisted upregulated miRNAs in SOC samples using miRNA microarray analysis, where miR-200c-3p was the most upregulated in SOC compared to normal ovarian tissues. Even though both low-grade and high-grade SOC were used in the study, more than 90% of the samples were high-grade [14]. Thus, in the present study, we further evaluated the expression of miR-200c-3p in different stages of HGSC compared to normal tissues, and identified the binding region at 3 UTR of the targeted gene.
In this study, we clearly proved that the expression pattern of miR-200c-3p in HGSC tissues at different stages was found to parallel the aggressiveness of the cancer cells, where the expression was the highest in the late stages. These results were in concordance with a previous study that showed miR-200c-3p was significantly associated with the advanced stages of HGSC [18], lymph node metastasis, and poor clinical outcome [20]. Simultaneously, there was a downregulation of the tumor suppressor gene, DLC1, a computational prediction gene identified from the bioinformatic analysis.
DLC1 (also known as ARHGAP7 and STARD12 in humans) is a gene that encodes the RhoGAP protein critically involved in cytoskeleton regulation, angiogenesis, and cell migration by inactivating Rho GTPase [21,22]. It is a tumor suppressor gene that could be dysregulated by genetic and epigenetic mechanisms in many types of cancers, such as lung, breast, gastric, liver, and ovarian cancer [23][24][25][26][27]. From a meta-analysis study, it was shown that DLC1 expression was downregulated in the advanced stages compared to the earlystage cancer in 19 eligible studies, including epithelial ovarian cancer (EOC), but those studies did not mention the histological subtypes of EOC [28]. Hence, our findings showed that DLC1 was the most downregulated in the late stage of HGSC and might be associated with high expression of miR-200c-3p through its direct binding at 3 UTR. Furthermore, our findings confirmed that there was a negative correlation between miR-200c-3p and DLC1 expression in HGSC tissues.
From the bioinformatic analysis, it was found that there are two putative miR-200c-3p binding sites within the 3 UTR of DLC1. For the first time, in the present study, we verified and confirmed miR-200c-3p direct binding on 3 UTR of DLC1 in luciferase reporter experiments in SKOV3 cell line. Based on these results, the main binding site was found to be at the first site of 3 UTR, a conserved target site with an 8-mer seed sequence, 5 -CAGTATTA-3 . This was then subsequently verified by mutating the target site, which retracted the binding actions by miR-200c-3p. This study was in an agreement with a previous study, where in an UTR fragment containing two sites, the repression at the 8-mer site was the most favorable. However, the repression at both sites in one UTR fragment was greater than the expected site individually [29]. This could be due to synergistic effects, for which the exact mechanism remains unknown [29,30].
To overcome the 'off-target' effects which occur through the same identification based on seed recognition as endogenous miRNA targeting, gain and loss of miRNA function tests are required [29]. Herein, we investigated the expression of miR-200c-3p in HGSC cell lines, and the results confirmed that miR-200c-3p downregulated DLC1, and that when miR-200c-3p expression was inhibited, DLC1 expression increased. However, this in vitro study demonstrated that the downregulation of DLC1 in the presence of over expression of miR-200c-3p only took place at an mRNA level. Hence, the regulation of miR-200c-3p at 3 UTR of DLC1 would cause mRNA degradation instead of translation repression. Nonetheless, this is still unclear, due to the complexity of the underlying mechanisms involved.
EMT and MET, an epithelial to mesenchymal transition and vice versa, are the transdifferentiations between epithelial cells and motile mesenchymal cells, which is a critical process in metastasis. Multiple established markers served as core markers in the assessment of EMT/MET, such as up-or downregulation of CDH1, a hallmark of EMT that disrupts adherens junctions, and which is customarily accompanied by a reduction in the expression of CDH2, VIM, and others [31,32]. A previous study demonstrated that over expression of DLC1 could inhibit EMT by increasing the expression of epithelial markers [33]. According to another report, the inhibitory effects of CDH1 could be attenuated by knocking down DLC1 epigenetically [34]. However, the present study displayed contradictory results, which showed a negative regulation between DLC1 and CDH1 when miR-200c-3p was overexpressed.
In this study, even though the high expression of miR-200c-3p was found to cause downregulation of a tumor suppressor gene DLC1, at the same time it also led to upregulation of CDH1 expression in both primary and metastatic tumor cells. Based on this data, we suggested that over expression of miR-200c-3p could reduce DLC1 mRNA and hence, inhibited EMT and promoted MET in HGSC, as observed by the upregulation of the epithelial marker, E-cadherin. Our previous study confirmed that over expression of miR-200c stimulated cell proliferation and colony formation but attenuated cell migration and invasion [14]. Moreover, MET could also be regulated by miR-200c through a double feedback loop with the two E-cadherin transcriptional regulators, ZEB1 and ZEB2, and influencing TGF-β and EGFR signaling cascades [35,36]. Eventually, this caused cells to re-epithelialize to form metastatic foci at a secondary site.
The use of the SKOV3 cell line might be a possible limitation of our study. The SKOV3 cells was mostly referred as 'serous' due to some uncertainties about putative TP53 mutation. However, the genetic analysis showed that SKOV3 not only had a mutation on TP53, but also in ARID1A, where it was most likely to be endometrioid or clear cell carcinoma [37,38]. Therefore, a better-defined system model with valid clinical phenotypes needs to be considered in future studies.
In the future, miR-200c-3p could have a great potential as a diagnostic marker for early detection, as well as a prognostic marker to complement the current cancer staging system in ovarian cancer [39,40]. Furthermore, for therapeutic purposes, with the target sequence known, it could be proposed to form an miRNA sponge containing binding sites of the target gene to compete with the 'authentic target' [41]. The use of miRNA as a target in medical therapies demonstrates its role as an alternative in combating debilitating cancers, such as ovarian cancer

Conclusions
In conclusion, upregulated miR-200c-3p expression in both early and late stage of HGSC could downregulate DLC1, as a tumor suppressor. At the same time, miR-200c-3p could regulate EMT/MET in HGSC cells. Therefore, we suggest that a combination of miR-200c-3p and the molecular target sites of DLC1 could be a potential regulator in HGSC for therapeutic purposes and to reverse cancer progression in this deadly disease.