miR-373-3p inhibits EMT via regulation of TGFβR2 in choriocarcinoma

Background: Previous studies have indicated that early metastasis is a major cause of mortality in patients with choriocarcinoma. However, what determines whether early metastasis of choriocarcinoma has occurred is unknown. The emerging role of miRNA in regulating cancer development and progression has been recognized. MiR-373 has been shown to play pivotal roles in tumorigenesis and metastasis. However, whether miR-373 functions to promote choriocarcinoma metastasis is not clear. The purpose of this study is to determine the function of miR-373-3p in the progression of this cancer. Methods: In this study, we first compared epithelial-to-mesenchymal transition (EMT)-related markers, which were inversely correlated with miR-373-3p expression, in trophoblast and choriocarcinoma cell lines. Using PCR and Western blot, upregulation of miR-373-3p was observed to inhibit EMT progression. Similarly, gain- and loss-of-function studies revealed that ectopic miR-373-3p overexpression inhibited the metastasis of choriocarcinoma cells. Results: Our results revealed that miR-373-3p acted as an EMT inhibitor in JEG-3 and JAR cells; this was due to its mediation of the TGF-β signalling pathway, which was responsible for EMT. Bioinformatic analysis demonstrated that miR-373-3p interacted with the 3' untranslated region of TGFβR2 mRNA, and then Western blot and dualluciferase reporter gene assays verified this interaction. Conclusions: Our findings suggest that miR-373-3p upregulation partly accounts for TGFβR2 downregulation and leads to a restraint of EMT and metastasis. MiR-373-3p may therefore serve as a valuable potential target in the treatment of choriocarcinoma.

extracellular signals can initiate an EMT programme, and members of the transforming growth factor-β (TGFβ) family of cytokines are the primary and best characterized inducers of EMT [11,12]. TGF-β signalling pathway is activated through heteromeric complex of TGF-βRI and TGF-βRII. TGF-β binds to TβRII and induces phosphorylation and activation of TβRI. After interacting with TβRI, phosphorylated Smad2/3 dissociate to form a heterotrimeric complex with Smad4; and then the complex translocates to the nucleus, where it regulates gene transcription [13]. Thus, TGF-βRII plays a crucial role in the TGF-β signalling pathway.
In this study, we compared EMT-related markers, which were inversely correlated with miR-373-3p expression, in trophoblast and choriocarcinoma cell lines. An increase or decrease in miR-373-3p can release or reduce repression of its direct target TGFβRII, respectively. TGFβRII is also part of the TGFβ signalling pathway, which is the molecular mechanism that mediates EMT. Our data showed that miR-373-3p played an inhibitor role in choriocarcinoma, which was due to its specific mediation of the TGFβRII signalling pathway and inhibition of EMT progression. (1:1000 dilution) and β-actin (1:500 dilution). The membranes were washed and incubated with horseradish peroxidase-conjugated secondary antibodies (goat anti-mouse or antirabbit; 1:1000 dilution). Immunoreactive bands were detected using enhanced chemiluminescence (ECL) substrate (Pierce, Rockford, IL, USA) and imaged using the Image Quant LAS4000 system (GE Company, Pittsburgh, PA, USA).

In vitro migration, invasion, and wound scratch assays
The cells were collected and resuspended in serum-free DMEM at a concentration of 4×10 5 cells/ml, as determined by cell counts. Then, the cell suspension solution was seeded into the upper chambers (200μl/well), and the bottom chambers were filled with DMEM containing 10% FBS (1ml/well). We used CORNING (USA) Transwell chambers (Lot 3422, 6.5mm Diameter Inserts; 8-μm pore size; polycarbonate membrane) to quantify cell migration. For the Transwell invasion assay, CORNING (USA) Transwell chambers (Lot 354480, polycarbonate membrane, Matrigel-coated) were used. After they were cultured at 37°C for 24h, the cells that had not penetrated the polycarbonate membrane were wiped off with a cotton swab. The membrane was removed, fixed in 4% paraformaldehyde and stained with crystal violet solution. Cell migration and invasion were determined by counting five random high-power fields (Olympus Corp, Tokyo, Japan).
For the wound scratch assay, cells were plated and grown until confluence after which the cells were scratched using sterile tips. Cellular migration (toward the scratched area) was assessed after 24 h.

Luciferase assay
In the luciferase reporter vector, the wild-type or mutant 3'-UTR of human TGFβR2 was cloned into the downstream of the renilla luciferase gene in pGL3 vectors (GenePharma, Shanghai, China). Cells were seeded onto 24-well plates (1x10 5 cells per well) the day before transfection. Cells were then transfected with the firefly luciferase reporter plasmid including either the wild-type or mutant 3'-UTR of TGFβR2 (50 ng per well) and the pRL-TK Renilla luciferase reporter (10 ng per well). The cells were then transfected with miR-373-3p mimics or negative control (20 µM) using jetPRIME® in vitro DNA & siRNA transfection reagent (Poly plus).After 24 h of transfection, cells were lysed with 1× reporter lysis buffer, and firefly and renilla luciferase activities were measured using the Dual-Luciferase® Reporter Kit (Promega) according to the manufacturer's instructions. Firefly luciferase activity was standardized to renilla activity as a control.

Statistical analysis
All experiments were repeated at least three times. Data are presented as the mean± standard deviation (SD). Statistical analysis was performed using one-way analysis of variance (ANOVA) and Student's t test. A P-value <0.05 was considered statistically significant and is indicated by asterisks in the figures. Experiments were performed using cells at a logarithmic growth phase.

Paraffin-embedded human choriocarcinoma tissue
We included ten formalin-fixed paraffin-embedded (FFPE) choriocarcinoma samples from the pathological archives of the Affiliated Hospital of Chengde Medical College, Chengde Third Hospital and the Second Affiliated Hospital of Hebei Medical University. Ten women with normal early pregnancy and a single gestational sac were randomly selected, and the gestational villus were obtained at Chengde Women's and Children's Hospital.

miRNA microarray analysis
Total RNA in formalin-fixed paraffin-embedded (FFPE) sections was extracted using a Recover All™ Total Nucleic Acid Isolation Kit (Ambion-1975). MiRNA expression analysis was conducted on total RNA samples containing small RNAs using the human miRNA microarray, Release19.0, 8x60K (Agilent Technologies). For control and test RNAs, the synthesis of target miRNA probes and hybridization was performed using a miRNA Labeling Reagent and Hybridization Kit (Agilent Technology) according to the manufacturer's instructions. The hybridized microarrays were washed according to the manufacturer's washing protocol (Agilent Technology). Images of hybridized arrays were obtained using a DNA microarray scanner (Agilent Technology) and were quantified with Feature Extraction Software (Agilent Technology). Data normalization and the selection of genes with a fold change (FC) > 2 were performed using GeneSpring GX software (Agilent Technology).

Quantitative real-time PCR analysis (qRT-PCR)
Total RNA was isolated using the RNAgents® Total RNA Isolation System (Promega) with DNase I (Invitrogen) treatment. Then, 2 µg RNA, oligo (dt) 20 primers and the M-MLV First-Strand Synthesis System kit was used to synthesize cDNA. Real-time PCR was performed using the SYBR® Green I on GoTaq® qPCR Detection System (Promega) according to the manufacturer's instructions. The results were analysed using the comparative threshold cycle method with GAPDH as an internal control. The results were normalized to GAPDH levels using the formula ΔCt (Cycle threshold) = Ct of target gene -Ct of GAPDH. The mRNA level of the control group was used as the baseline; therefore, ΔΔCt was calculated using the formula ΔΔCt = ΔCt of target gene -ΔCt of the baseline.
The fold change in the mRNA level was calculated as fold = 2 -ΔΔCt . Primers used in this study were synthesized by Sangon Biotech Technology (Shanghai, China) (see Table 1). FFPE samples were cut into 1-3 mm cores. Total RNA was isolated using the Qiagen RNeasy FFPE protocol. For analysis of miRNA expression, miRNA-specific reverse transcription was performed with a GenePharma MicroRNA Reverse Transcription Kit according to the manufacturer's instructions. Quantitative RT-PCR was performed using GenePharma PCR Master Mix; U6 small nuclear RNA was used as an internal control. All reactions were performed in triplicate.

Western blot analysis
Total proteins were extracted from each group using RIPA buffer (Thermo Scientific, Rockford, IL, USA) and were quantified using a BCA kit (Thermo Scientific). The proteins were separated by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidenedifluoride (PVDF) membranes. After blocking with 5% bovine serum albumin, the blots were probed with the appropriate primary antibodies (1:1000 dilution) and β-actin (1:500 dilution). The membranes were washed and incubated with horseradish peroxidase-conjugated secondary antibodies (goat anti-mouse or antirabbit; 1:1000 dilution). Immunoreactive bands were detected using enhanced chemiluminescence (ECL) substrate (Pierce, Rockford, IL, USA) and imaged using the Image Quant LAS4000 system (GE Company, Pittsburgh, PA, USA).

In vitro migration, invasion, and wound scratch assays
The cells were collected and resuspended in serum-free DMEM at a concentration of 4×10 5 cells/ml, as determined by cell counts. Then, the cell suspension solution was seeded into the upper chambers (200μl/well), and the bottom chambers were filled with DMEM containing 10% FBS (1ml/well). We used CORNING (USA) Transwell chambers (Lot 3422, 6.5mm Diameter Inserts; 8-μm pore size; polycarbonate membrane) to quantify cell migration. For the Transwell invasion assay, CORNING (USA) Transwell chambers (Lot 354480, polycarbonate membrane, Matrigel-coated) were used. After they were cultured at 37°C for 24h, the cells that had not penetrated the polycarbonate membrane were wiped off with a cotton swab. The membrane was removed, fixed in 4% paraformaldehyde and stained with crystal violet solution. Cell migration and invasion were determined by counting five random high-power fields (Olympus Corp, Tokyo, Japan).
For the wound scratch assay, cells were plated and grown until confluence after which the cells were scratched using sterile tips. Cellular migration (toward the scratched area) was assessed after 24 h.

Luciferase assay
In the luciferase reporter vector, the wild-type or mutant 3'-UTR of human TGFβR2 was cloned into the downstream of the renilla luciferase gene in pGL3 vectors (GenePharma, Shanghai, China). Cells were seeded onto 24-well plates (1x10 5 cells per well) the day before transfection. Cells were then transfected with the firefly luciferase reporter plasmid including either the wild-type or mutant 3'-UTR of TGFβR2 (50 ng per well) and the pRL-TK Renilla luciferase reporter (10 ng per well). The cells were then transfected with miR-373-3p mimics or negative control (20 µM) using jetPRIME® in vitro DNA & siRNA transfection reagent (Poly plus).After 24 h of transfection, cells were lysed with 1× reporter lysis buffer, and firefly and renilla luciferase activities were measured using the Dual-Luciferase® Reporter Kit (Promega) according to the manufacturer's instructions. Firefly luciferase activity was standardized to renilla activity as a control.

Statistical analysis
All experiments were repeated at least three times. Data are presented as the mean± standard deviation (SD). Statistical analysis was performed using one-way analysis of variance (ANOVA) and Student's t test. A P-value <0.05 was considered statistically significant and is indicated by asterisks in the figures.

mesenchymal-epithelial transition occurs in choriocarcinoma cells but not in trophoblasts
We found an interesting phenomenon that mesenchymal-epithelial transition (MET) These results suggest that trophoblast cells exhibit stronger mesenchymal characteristics.
Conversely, choriocarcinoma cells exhibit relatively more epithelial characteristics.

Trophoblasts have strong invasion and migration potential
To further detect if the EMT phenotype is the main reason through which cells obtain the capacity to migrate and invade, a wound healing assay was performed to examine whether choriocarcinoma and trophoblast cells have different abilities to migrate into a scratched area. The scratch wound assay simultaneously indicated that the choriocarcinoma cell lines JEG-3 and JAR showed a weaker ability to migrate compared with the trophoblast line, as demonstrated by quantification and representative images shown in Fig.2A. This indicates that trophoblasts exhibit a stronger migration ability. A Transwell chamber assay was then performed to analyse choriocarcinoma cell migration and invasion. Compared with trophoblast cells, the numbers of migrated and invasive cells were less in choriocarcinoma JAR and JEG-3 cells (Fig.2B).

MiR-373-3p expression is upregulated in choriocarcinoma tissue and cell lines and is inversely correlated with EMT marker expression
Three choriocarcinoma tissue samples and three gestational villus tissues were included in a microarray analysis, and the fold change method was employed to conduct a differential analysis of the miRNA expression levels in these two groups. Most members of the miR-520/373 family demonstrated expression changes that attracted our attention. Compared with gestational villus tissues, the average expression level of miR-373-3p was significantly increased (Fig.3A)

MiR-373-3p inhibits the invasion and migration ability of JEG-3 and JAR cells
In vitro gain-of-function analyses were performed to explore the effects of miR-373-3p function by transfecting miR-373-3p mimics and a miR-373-3p inhibitor into JAR and JEG-3 cells. Transwell migration experiments were then performed to explore the effects of miR-373-3p on migration ability. It was shown that miR-373-3p mimics can significantly inhibit the migration ability of JEG-3 cells. On the contrary, a miR-373-3p inhibitor can promote the migration ability of JEG-3 and JAR cells (Fig.4A-4B). Consistent with this hypothesis, Matrigel-coated Transwell assays showed that miR-373-3p overexpression can significantly inhibit the invasion ability of these two cell lines. However, low miR-373-3p expression can enhance the invasion ability of JEG-3 and JAR cells (Fig.4C-4D). Collectively, our data suggest that miR-373-3p may inhibit the invasion and migration ability by influencing EMT in JEG-3 and JAR cells.

TGF-β signalling induces EMT in choriocarcinoma
Since TGF-β signalling is one of the most important pathways that functions to maintain an epithelial cell phenotype and proper cell-cell junctions, we sought to understand whether TGF-β affected the EMT process. TGF-β treatment decreased the mRNA levels of epithelial markers and increased the levels of mesenchymal markers in a dose-dependent (Fig.5A) and time-dependent manner (Fig.5B) in JEG-3 cells. The same results can be seen in JAR cells (Fig.5C-D). To confirm these findings, the protein levels of epithelial markers and mesenchymal markers were evaluated. TGF-β treatment decreased the protein levels of epithelial markers and upregulated the protein levels of mesenchymal markers in JEG-3

MiR-373-3p targets TGFβR2, which influences EMT in choriocarcinoma
A bioinformatic analysis using 3 miRNA-target gene databases (targetscan.org, miranda.org and mirdb.org) showed that TGFβR2 mRNA is a presumed target of miR-373-3p. The potential target region for miR-373-3p in the TGFβR2 mRNA 3'UTR are shown in Fig.7A. To verify that miR-373-3p can regulate TGFβR2 mRNA, we investigated TGFβR2 protein expression in JEG-3 cells transfected with a miR-373-3p mimic and a miR-373-3p inhibitor. The expression of TGFβR2 protein was downregulated in successful transfection of the miR-373-3p mimic group (Fig.7B). Dual-luciferase reporter assays were implemented to validate whether regulation exists. Dual-luciferase reporter vectors containing either the mutant or wild-type 3'UTR of TGFβR2 mRNA were then constructed and cotransfected into JEG-3 (Fig.7C) and JAR (Fig.7D) cells together with the miR-373-3p mimic or non-specific control (NC) miRNA. The results showed that cells transfected with pmirGLO-wt-TGFβR2 group was significantly responsive to miR-373-3p overexpression and that these cells also exhibited a weaker reaction to the NC because both cell lines express miR-373. However, no significant differences were observed in the relative luciferase activity between the miR-373-3p mimic group and the NC group of cells cotransfected with pmirGLO-mut-TGFβR2 (Fig.7C-7D). The results imply that miR-373-3p acts directly on the 3'-UTR of TGFβR2 mRNA and that it negatively regulates TGFβR2 expression.
We further explored the functional significance of miR-373-3p and found that it activated TGFβ signalling, which increased EMT in choriocarcinoma cells. Recent studies and our previous data [14,15] showed that TGFβ signalling can activate EMT in choriocarcinoma [16,17]. Given the active roles of TGFβ in regulating EMT, we then questioned whether TGFβR2 is a potential regulator involved in this process. JEG-3 and JAR cells were pretreated for 12h with TGFβ before transfection with the miR-373-3p mimic for 6h. The

Discussion
EMT is a highly conserved process by which cells lose epithelial characteristics and acquire a mesenchymal, migratory phenotype. EMT is involved in the process of tissue and organ development at the embryonic stage. In recent years, most scholars focus on the EMT process as it relates to tumour progression and metastasis. Most tumours may undergo EMT, which is a hall mark of aggressive cancer; they may also acquire invasive and metastatic potential, which leads to a poor patient prognosis. Different from other tumours, trophoblast cells and their tumours are special. Several studies have reported that trophoblast cells and malignant cells share many similar phenotypic properties including proliferation, migration and invasion of neighbouring tissues. The cellular mechanisms of EMT in trophoblasts are poorly understood, but attempts have been made to use trophoblast cells to understand the EMT process from a cancer perspective because of the high invasive capacity of trophoblasts [18]. In this study, we present an exciting discovery that trophoblast cells have a stronger mesenchymal phenotype and greater ability to invade and migrate, which is different from the phenotype of most other cancers. Indeed, many reports have indicated that trophoblast cells undergo EMT, which allows them to migrate and infiltrate into the maternal decidua and vessels to ensure normal gestation [19,20]. Few studies on epithelial-to-mesenchymal transition in choriocarcinoma cells have been published. One study even suggests that gestational trophoblastic neoplasms do not display EMT features [21].
Based on our findings, it appears that trophoblast cells exhibit strong expression of mesenchymal markers and that choriocarcinoma cells exhibit strong expression of epithelial markers. However, in this study, we describe EMT as an immotile, polarized process through which epithelial cells undergo a number of biochemical changes to attain mesenchymal cell characteristics. Actually, EMT is a dynamic and somewhat unstable procedure, and thus future research should use dynamic and changeable methods to further understand the EMT process in the occurrence and transformation of trophoblast cell malignancy. EMT is closely related to invasion, and shallow invasion is a characteristic feature of foetal growth restriction, while abnormal deep invasion is associated with placenta accrete, and uncontrolled invasion is associated with choriocarcinoma [22, 23].
Hence, clarifying the potential regulatory mechanism of EMT in trophoblast cells and tumour development may provide novel diagnostic and therapeutic perspectives of malignancy.
Recent work suggests that miRNAs are important molecular regulators that can modulate EMT [24]. Hence, clarifying the potential regulatory mechanism of miRNAs in choriocarcinoma development and the invasive characteristics of trophoblast cells may provide novel diagnostic and therapeutic perspectives for trophoblastic malignancies. In this study, q-PCR validation revealed that miR-373-3p expression was enhanced in choriocarcinoma cells compared with trophoblasts and was inversely correlated with the occurrence of EMT. MiR-373 has been shown to serve as a tumour-suppressing miRNA in gastric cancer because it downregulates vimentin [25]. One study suggested that miR-373-3p acts as an oncogenic miRNA, which when upregulated, targets CD44 to promote invasion and metastasis in breast cancer [26]. However, the role of miR-373-3p remains controversial because the expression of miR-373-3p is abnormal in various types of tumours, which influences proliferation, invasion and metastasis.
TGFβ may induce EMT through multiple distinct signalling mechanisms, including regulation of tight junction formation by certain cytoplasmic proteins and phosphorylation by ligand-activated receptors of SMAD transcription factors. TGFβRII can directly phosphorylate both SMAD2 and SMAD3 proteins [27][28][29]. Our studies indicate a direct connection between TGFβ1 and EMT in choriocarcinoma. According to our observations, TGFβ1 recognizes and activates the receptor TGFβR2, which subsequently activates its downstream EMT mediators.
Given the roles of miR-373-3p in regulating EMT and inhibiting invasion and metastasis activity, we hypothesized that miR-373-3p may also regulate TGFβ signalling, which regulates EMT in choriocarcinoma cells. Our results demonstrated that TGFβR2, an apical regulator of TGFβ signalling, as a direct functional effector of miR-373-3p, which is in agreement with many studies showing that TGFβR2 is downregulated in various cancers.
Apart from this, increased expression of miR-373-3p in choriocarcinoma cells was significantly associated with downregulated expression of TGFβR2 and a less aggressive mesenchymal phenotype.
In conclusion, miR-373-3p was upregulated in choriocarcinoma and miR-373-3p overexpression inhibited the EMT process and invasion/migration in vitro. These data suggest that miR-373-3p overexpression may have potential in the development of a new therapy for choriocarcinoma. Finally, our results provide a first step in understanding the pathologenesis of these rare gestational neoplasms and reveal novel molecular mechanisms. However, trophoblasts are unique since they undergo migration and invasion to establish their roles in the placenta. During the process in which trophoblasts transform into choriocarcinoma cells, intrinsic and extrinsic mechanisms precisely control this transition. In particular, EMT is a dynamic equilibrium course, and more research is  -GATAGAGAACGCATTGCCACATA (Sense); -ACCTTCCATGACAGACCCCTTAA (Antisense) CDH2 -TGTTTGTCCTTACTGTTGCTGC (Sense); -TTCTTCTTGGCGAATGATCTTAG (Antisense) Vimentin -GACCTGCTCAATGTTAAGATGGC (Sense); -CAGAGGGAGTGAATCCAGATTAGTT (Antisense) ZEB1 -GCAGTCTGGGTGTAATCGTAAAT (Sense); -CTTGCAGTTTGGGCATTCATAT (Antisense) has-miR373 mimics (5' to 3') GAAGUGCUUCGAUUUUGGGGUGU ACCCCAAAAUCGAAGCACUUCUU has-miR373-3p inhibitor (5' to 3') ACACCCCAAAAUCGAAGCACUUC