Identification of Differentially Expressed microRNAs in Metastatic Ovarian Cancer

Background: The molecular mechanisms of ovarian cancer (OC) remain unclear. We sought to comprehensively identify miRNAs that are aberrantly expressed in OC. Methods: Differentially expressed miRNAs were screened from six pairs of primary and metastatic OC tissues; their possible functions were then analyzed and target genes were predicted by bioinformatics. Then gene expression profiling results were established by reverse transcription quantitative polymerase chain reaction and western blot. Target binding between miR-7-5p and TGFβ2 was validated by dual-luciferase reporter assay. Results: Fifteen miRNAs and 10 target mRNAs were differentially expressed in primary and metastatic OC tissues. ITGB3, TGFβ2 and TNC correlated to miRNA function in metastatic OC. Compared with primary OC, RNA levels of hsa-miR-141-3p, hsa-miR-7-5p and hsa-miR-187-5p in metastatic OC tissues were potently decreased ( p < 0.05). However, a statistically prominent difference in hsa-miR-584-5p level between the two groups ( p > 0.05) was not observed. Comparing to primary OC, TGFβ2 and TNC were markedly increased ( p < 0.05). Luciferase activity was remarkably decreased after co-transfection of a wild-type TGFβ2 3’-UTR plasmid and miR-7-5p compared with a control plasmid, but no remarkable change after co-transfection of mutant TGFβ2 3’-UTR and miR-7-5p was demonstrated. Conclusions: Fifteen miRNAs and 10 mRNAs were differentially expressed in metastatic OC tissues compared with primary OC tissues, which suggested that they may participate in invasive and metastatic processes. Hsa-miR-141-3p, hsa-miR-187-5p and hsa-miR-7-5p expression was prominently lower in metastatic OC than in primary OC, while TGFβ2 and TNC expression was markedly higher in metastatic OC tissues. Hsa-miR-7-5p may bind to the TGFβ2 3’-UTR to inhibit its expression. support the detection of miRNA expression but fail to allow the parallel detection of miRNA and target genes; as a result, it is impossible to comprehensively understand microRNA–target gene regulatory networks. In our study, we analyzed the miRNA expression profiles of primary OC tissues and their respective metastases from six patients using a miRStar™ Human Cancer Focus miRNA and Target mRNA PCR Arraychip. This chip contained 184 cancer-related miRNAs and 178 target mRNAs. All 184 miRNAs were carefully screened and identified from the latest literature, including miRNAs relating to such common cancers as gastric, liver, lung, breast, colorectal, and prostate. Numerous studies provided support for 178 target mRNAs, and their expression correlated to the functional activity of miRNAs in tumors. With an innovative experimental design, the miRStar™ Human Cancer Focus miRNA and Target mRNA PCR Array chip allows the synchronous detection of miRNA and target mRNA. It can be used not only to rapidly detect expression changes of cancer-related miRNAs in test samples, but can also be used to evaluate the activity of these miRNAs against target mRNAs, thus facilitating the screening of core functional miRNAs in diseases and the rapid comprehension of their regulatory mechanisms and functions. versus OC while and were markedly up-regulated. Hsa-miR-7-5p and TGFβ2 were correlated in terms of mRNA and protein expression. Hsa-miR-7-5p directly bound the TGFβ2 3'-UTR to inhibit its expression and thus affect the invasive and metastatic processes of OC. Hsa-miR-7-5p may prove to be a potential strategy for targeted metastatic OC therapies. However, several limitations existed in the present study. Firstly, the present study had a small sample size. Thus, it is necessary to conduct further studies with a larger number of samples to validate our results. Furthermore, future studies will be required to investigate the underlying molecular mechanisms between hsa-miR-7-5p and metastatic OC.


Background
Ovarian cancer (OC) is the deadliest gynecological cancer due to the absence of symptoms at the early stage [1,2]. More concretely, due to the non-specificity of the symptoms of OC, a diagnosis is always made at advanced stages of cancer that have poor 5-year overall survival rate which is about 30-40% [3]. Moreover, the widespread and distant metastasis is observed in 59% OC patients and closely correlated with poor prognosis of ovarian cancer [4]. Therefore, it is pivotal to study invasive and metastatic mechanisms of OC. The incomplete elucidation of the molecular pathways underlying OC development poses a great challenge to improving clinical outcomes and having a greater depth of understanding underlying the mechanisms of cancer metastasis that would contribute toward combating this disease.
MicroRNAs (miRNA), small RNA molecules (21-23 nucleotides [nt]), function as potent modulators of gene expression through mRNA translation blockade or RNA interference [5], and have become a hotspot of tumor research over the recent years. The dysregulation of miRNAs may act as novel oncogenes or anti-oncogenes in ovarian cancer by regulating target genes [6]. For instance, miR-7 was found to depress EGFR/ERK pathway to decrease OC cell invasion [7]. Stated thus, a hypothesis can be drawn that miRNAs participate in OC metastasis via their target genes. Therefore, differentially expressed miRNAs and mRNAs were screened from primary and metastatic OC tissues using chip technique. We found 15 miRNAs and 10 mRNAs that were differentially expressed and that participate in the invasive and metastatic processes of OC. The chip results were validated by reverse transcription quantitative polymerase chain reaction (RT-qPCR), western blot and dual-luciferase reporter assay.

Study participants
All 31 pairs of metastatic and primary OC tissues were collected from patients who had received surgery at Shengjing Hospital of China Medical University from 2014-2016. Of these, six pairs were used for chip experiments and 25 pairs were employed for RT-qPCR validation. Tissue samples resected during operations were immediately placed into 1.5-mL centrifuge tubes with RNase-removal high-pressure treatment, rapidly placed in liquid nitrogen for quick freezing, and then transferred to a -80ºC cryogenic freezer for preservation. All patients were diagnosed with ovarian serous cystadenocarcinoma by postoperative paraffin pathology, and their clinical staging was made according to the International Federation of Gynecology and Obstetrics (FIGO) staging system for ovarian cancer (2009). None of the patients had received radiotherapy, chemotherapy or other special treatments prior to surgery. Patients who provided tissue specimens for chip experiments were aged 48-64 years (mean: 57.5 years), and those providing samples for RT-qPCR validation were aged 38-80 years (median: 57 years). Data and tissues were harvested upon receiving of the informed consent of patients and approval by the ethics committee of Shengjing Hospital of China Medical University.

RNA extraction from tissues
A Trizol Kit (Invitrogen Inc., Carlsbad, CA, USA) was applied for isolation of total RNA from metastatic and primary OC tissues. RNA purity was estimated using a NanoDrop® ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), and assessed by the ratio of optical density (OD) 260 nm/280 nm. A ratio between 1.8 and 2.0 signified the sample to be highly pure.

RT-qPCR
A total of 25 pairs of metastatic and primary OC tissues frozen at -80 °C were collected. Extraction of total RNA from tissues and cells was conducted using TRIzol reagent (Invitrogen). A mixture (14.5 µL) of 2.0 µg RNA, 1 µL Oligo, 1.6 µL deoxy-ribonucleoside triphosphate Mix, and H 2 O were incubated at 65 °C in water for 5 min and then put on ice for 2 min. After centrifugation, the mixture was mixed with reverse transcription reaction liquid, and placed at 37 °C for 1 min. The mRNA was reversely transcribed into cDNA, and the sample then temporarily placed on ice. All cDNA template samples were prepared in a PCR amplification system that included 5 µL of 2 × Taq MasterMix (Arraystar), 0.  (Table 1) were designed by Sangon Biotech (Shanghai, China). The relative expression level of TGFβ2 and TNC or hsa-miR-141-3p, hsa-miR-187-5p, hsa-miR-7-5p and hsa-miR-584-5p was standardized to β-actin or U6 expression and was calculated using the 2 −ΔΔCt method. Table 1 Primer sequences for RT-qPCR
After plasmid transfection, a. 293T cells in the logarithmic growth phase were prepared into a suspension that was then cultured in a 24-well plate so as to achieve 85% confluence in an incubator (BB15; Thermo Fisher Scientific) with saturated humidity and 5% CO 2 at 37 °C. The medium was changed every 24 h and subcultures were undertaken every 72 h. After removal of medium, cells were trypsinized for 3 min and then the reaction was ended by adding DMEM containing 10% FBS. Afterward, cells were made into a single cell suspension using pipette tips.

Statistical analysis
All data were summarized as mean ± standard deviation and analyzed by t test using SPSS 23.0 statistical software (IBM Corp., Armonk, New York, USA). A value of p < 0.05 was considered appreciable.
For molecular functions, differentially expressed genes mostly took part in RNA polymerase II distal enhancer sequence-specific DNA binding transcription factor activity involved in positive regulation of transcription (p < 0.001, Enrichment scroe > 3; Table 8). The results of KEGG enrichment analysis are presented in Fig.s 3-4. Pronounced differences between 10 pathways were found, of which ITGB3, TGFβ2 and TNC were linked to the pathway-MicroRNAs in Cancer.  Screening of miRNA-mRNA with potential target relationship Through miRNA microarray analysis, miRNA-mRNA with a potential target relationship was screened and their binding sites analyzed on a bioinformatics prediction website (Fig. 5).
Hsa-miR-7-5p and TGFβ2 selected as target miRNAs The expression of four miRNAs and TGFβ2 and TNC mRNA expression were determined by RT-qPCR.
As documented in Fig. 6, hsa-miR-7-5p, hsa-miR-141-3p and miR-187-5p expression was markedly decreased in metastatic compared with primary OC tissue (p < 0.05). As displayed in Fig. 7, TGFβ2 and TNC mRNA expression was enhanced in metastatic OC compared with primary OC tissue (p < 0.05). Results coincided with the results of microarray analysis. TGFβ2 was chosen as our target gene.
The protein level of TGFβ2 is elevated in metastatic OC tissue TGFβ2 protein level was evaluated by western blot. TGFβ2 protein level was elevated in metastatic OC tissue versus that in primary OC tissue (p < 0.05; Fig. 8). The enhanced TGFβ2 may contribute to metastasis in OC progression.

Discussion
MicroRNAs play a pivotal part in multiple cellular functions such as proliferation, apoptosis, differentiation and development. However, aberrant miRNAs occurs in carcinogenesis and metastasis during cancer progression [8]. As is well known, miRNAs can post-transcriptionally mediate a number of genes through a combination of specific sequences in target mRNA molecules [17]. MicroRNAs were differently expressed in OC [18]. Furthermore, miR-187 has been found to exert a dual role in OC by regulating the disabled homolog-2 gene [19]. Gao et al. also found that miR-141 acted as a potential diagnostic and prognostic biomarker for OC [20], which was concurred with our results. Zhu et al. demonstrated that miR-7-5p suppresses cell migration and invasion by targeting SOX18 in pancreatic ductal adenocarcinoma [21]. Another study also claimed that miR-7-5p under-expression was associated with recurrence in glioblastoma patients, and that its overexpression decreased glioblastoma cell stemness [22]. Furthermore, it was also elucidated that ectopic expression of miR-7 functioned as an anti-oncogene in OC by repressing cell invasion and proliferation [7].
MiR-7-5p has been found to target and regulate the genes as SATB1 and PARP1 in some cancer diseases as we have mentioned before [23,24]. Furthermore, in this study, we found that TGFβ2 was targeted by hsa-miR-7-5p. The results of miRNA-mRNA conjoint analysis with our chip showed that hsa-miR-7-5p and TGFβ2 were differentially expressed in metastatic and primary OC tissues at a striking level, and may have a targeted regulatory relationship. To confirm this, we further validated the expression of four miRNAs (including hsa-miR-7-5p), TGFβ2 and TNC in 25 pairs of metastatic and primary OC tissues, a larger sample size. We found that TGFβ2 was negatively targeted by miR-7-5p using a dual luciferase reporter assay. MiR-7-5p may have several target genes similar to miR-137; such target genes are critical oncogenic factors, including TGFβ2, that could further regulate brain tumorigenesis [25]. In addition, we found that OC metastases expressed higher levels of TGFβ2 and TNC compared with primary OC tissues. This suggested that TGFβ2 was closely associated with OC metastasis. In studying potential mechanisms, we identified TGFβ2 as an mRNA target using bioinformatics analysis, as well as functional and binding assays. Taken together, our findings suggested that hsa-miR-7-5p and TGFβ2 were inversely correlated with regard to mRNA and protein expression; they demonstrated a targeted regulatory relationship that could influence the invasive and metastatic processes of OC. The TGFβ pathway takes part in many cellular processes, including cell proliferation, differentiation, extracellular matrix accumulation, tissue repair, immune and inflammatory responses. TGFβ2 is an isoform of TGFβ in mammals [26] and is abnormally expressed in various cancers such as human melanoma and hepatocellular carcinoma [27,28]. In addition, many researchers have also explored the role of TGFβ in OC. Cao and his colleagues reported that TGFβinduced transglutaminase gave rise to EMT and a cancer stem cell phenotype that consequently enhanced ovarian tumor metastasis [29]. TGFβ induced EMT and a more invasive phenotype in epithelial OC cells in collaboration with the EGF pathway, indicating TGFβ may be a promising target candidate for the treatment of metastatic OC in future [30]. TNC is an extracellular matrix glycoprotein that shows forced expression in cell proliferation and migration, and in EMT during organogenesis [31]. Moreover, a recent study highlighted that serum TNC levels were much higher in patients with epithelial OC than in normal controls. Such high serum TNC levels were related to poorer overall survival, which was consistent with our findings [32].

Conclusion
In conclusion, our study suggested that several differentially expressed miRNAs and mRNAs exist as putative target genes of these miRNAs in primary and metastatic OC tissues, which may mediate invasion and metastasis by OC. As shown by PCR validation with a larger sample size, three miRNAs (hsa-miR-141-3p, hsa-miR-187-5p, and hsa-miR-7-5p) in metastatic OC tissues were obviously downregulated versus primary OC tissues, while TGFβ2 and TNC were markedly up-regulated. Hsa-miR-7-5p and TGFβ2 were negatively correlated in terms of mRNA and protein expression. Hsa-miR-7-5p directly bound the TGFβ2 3'-UTR to inhibit its expression and thus affect the invasive and metastatic processes of OC. Hsa-miR-7-5p may prove to be a potential strategy for targeted metastatic OC       The expression of hsa-miR-7-5p, miR-187-5p, and miR-141-3p is decreased in metastatic OC tissue as measured by RT-qPCR. *p < 0.05 vs. primary OC tissues; hsa, Homo sapiens; miR, microRNA; RT-qPCR, reverse transcription quantitative polymerase chain reaction. The mRNA expression of TGFβ2 and TNC are elevated in metastatic OC tissue as measured by RT-qPCR. Panels A and B, The mRNA expression of TGFβ2 and TNC in metastatic OC tissue increased when compared with primary OC tissue as shown in bar graphs. Panels C and D, The mRNA expression of TGFβ2 and TNC in metastatic OC tissue increased when compared with primary OC tissue as shown in scatter plots; *p < 0.05 vs. primary OC tissue; TGFβ2, transforming growth factor beta 2; TNC, tenascin-C; RT-qPCR, reverse transcription quantitative polymerase chain reaction. in metastatic OC tissue was increased when compared with primary OC tissue as shown by bar graphs. *p < 0.05 vs. metastasis OC tissue; TGFβ2, transforming growth factor beta 2; GAPDH, glyceraldehyde phosphate dehydrogenase.