miR-138-5p inhibits the metastasis of prostate cancer by targeting FOXC1

Background: This study aimed to uncover the regulatory effect of miR-138-5p on the metastasis of PCa cells, and further explore the potential regulatory mechanisms via regulating FOXC1. Methods: 60 pairs tumor specimens from PCa patients were collected to determine the expression level of miR-138-5p by qRT-PCR. Subsequently, over-expression of miR-138-5p were established to explore the proliferation and metastasis of miR-138-5p in PCa cell lines was analyzed by CCK-8, Tranwell assay and Wounding healing assay, respectively. Bioinformatics analysis and luciferase reporter gene assay were performed to search for the target genes of miR-138-5p, and FOXC1 was selected. Finally, the biological role of miR-138-5p and FOXC1 in the progression of PCa was clarified by a series of rescue experiments. Results: The results of qRT-PCR revealed that miR-138-5p was lowly expressed in PCa tissues and cell lines. Besdies, these PCa patients with low-miR-138-5p had a higher Gleason score, lymph node metastasis, bone metastasis and poor prognosis of PCa, compared with the patients with high-miR-138-5p. Over-expression of miR-138-5p inhibited the viability, migratory and invasive capacities of PC-3 and DU-145 cells. Bioinformatics analysis and luciferase reporter gene assay suggested that FOXC1 was predicted to be the target of miR-138-5p. Moreover, FOXC1 level was negatively correlated to that of miR-138-5p in PCa tissues. Importantly, FOXC1 could reverse miR-138-5p mimic induced-inhibition of PCa malignant progression. Conclusions: Downregulated miR-138-5p was closely associated with Gleason score, distant metastasis and poor prognosis of PCa patients. In addition, miR-138-5p alleviated the malignant progression of PCa by targeting and downregulating FOXC1. this study aimed to elaborate the possible roles of miR-138-5p and FOXC1 in the progression of PCa, as well as the association with the clinical characteristics and prognosis of PCa patients, so as to bring a new idea for clinical treatment of PCa. 6 V6.01 was used for data analyses. Data were expressed as mean ± standard deviation. Intergroup differences were analyzed by the t-test. Kaplan-Meier curves were introduced for survival analysis. Chi-square test was performed to evaluate the correlation between miR-138-5p level with pathological indexes of PCa patients. P < 0.05 was considered as statistically significant.

Although PSA test greatly improves the early-stage diagnostic rate of PCa, its benefit in decreasing the mortality of PCa remains controversial [9,10]. Like other malignancies, the malignant progress of PCa is a multi-step and multi-stage process, including inactivation of tumor suppressor genes and/or activation of proto-oncogenes [11,12]. Currently, target therapy based on tumor-related miRNAs presents a promising application, and the results of these studies showed that miRNA had a good application prospect in the diagnosis, treatment, prognosis and other aspects of cancer, to provide new ideas for the pathogenesis of PCa [13,14].
MiRNAs are small, endogenous non-coding RNAs that negatively regulate expressions of proteincoding genes at translational level [15,16]. MiRNAs exert the biological function by degrading or inhibiting translation of mRNAs [16]. It is reported that miRNAs are extensively involved in affecting cellular behaviors and disease progression [17,18]. Accumulating evidences have demonstrated the effects of abnormally expressed miRNAs on the occurrence and progression of tumors [17,19]. These certain miRNAs might be utilized for developing anti-tumor drugs or biological hallmarks [20]. About 30% human genome could be regulated by miRNAs, and most of human miRNAs (52%) locate on tumor-associated genomic regions or gene fragile sites [21,22]. MiR-138-5p is a newly discovered cancer-related miRNA, which has been confirmed to be down-regulated in pancreatic cancer, colorectal cancer and other malignant tumors [21,22]. However, its expression and function of miR-138-5p in PCa are rarely reported.
Bioinformatics analysis has been widely applied for analyzing genome to further uncover the gene expression pattern. By analyzing miR-138-5p profiling microarray, Forkhead box C1 (FOXC1) was selected to be the target genes of miR-138-5p. FOXC1 is an essential member of the forkhead box transcription factors and has been highlighted as an important transcriptional regulator of crucial proteins associated with a wide variety of carcinomas [25,26]. Based on the above series of researches, this study aimed to elaborate the possible roles of miR-138-5p and FOXC1 in the progression of PCa, as well as the association with the clinical characteristics and prognosis of PCa patients, so as to bring a new idea for clinical treatment of PCa.

Wound healing assay
After 48 h of transfection, the cells were inoculated in 6-well plates and grown to 90% confluence.
After the creation of an artificial wound in cell monolayer, medium with 1% FBS was replaced. After 24 hours, the wound closure was captured by a microscope.
GAPDH and U6 were used as internal references. Each sample was performed in triplicate, and

Western Blotting
The transfected cells were lysed using PRO-PREPTM lysis buffer, shaken on ice for 30 minutes, and centrifuged at 14,000 × g for 15 minutes at 4 °C. Total protein concentration was calculated by the PCaA Protein Assay Kit (Pierce, Rockford, Il, USA). Rabbit anti-human monoclonal antibodies against FOXC1 were purchased from Santa Cruz, USA; horseradish peroxidase-labeled goat anti-rabbit secondary antibody was purchased from Genscript. GAPDH was used as the internal reference control. Protein samples were separated by SDS-PAGE, transferred to PVDF membrane, and blocked with 5% skim milk powder for 1 h at room temperature. Primary antibodies were added for incubation overnight at 4 °C shaker. In the next day, the membrane was rinsed 3 times with TBST and incubated with second antibody for 1 h at room temperature. After that, the protein samples on the membrane were finally semi-quantitatively analyzed by alpha SP image analysis software.
Dual-luciferase reporter assay 3'-UTR of wild-type (WT) human FOXC1 gene, which contains a putative miR-138-5p binding DNA sequence, was amplified by PCR and inserted into a p-miR-reporter (Ambion, USA) to create a firefly FOXC1-WT luciferase vector. The putative miR-138-5p binding sequence on FOXC1 3'-UTR was then mutated to void miR-505 binding. The mutant (MUT) 3'-UTR was also inserted into p-miR-reporter to create a firefly FOXC1-MUT luciferase vector. Human HEK293T cells were transduced with NC mimic or miR-138-5p mimic, then cross-transfected with FOXC1-WT or FOXC1-MUT for 48 h. After that, relative luciferase activities were measured using a Dual-Luciferase Reporter assay (Promega, USA) according to the manufacturer's protocol.

In vivo xenograft vectors
The Animal Ethics and Use Committee approved the tumor-forming experiment in nude mice. 8-weekold male nude mice were purchased from the animal center and randomly divided into two groups (5 in each group). The PC-3 cells with miR-138-5p mimic were injected subcutaneously into the axilla of mice. Tumor size was monitored every 5 days; Then, after 6 weeks, the mice were sacrificed. The tumor volumes were calculated using the following formula: tumor volume = (width 2 x length)/2.

Statistically analysis
GraphPad Prism 6 V6.01 was used for data analyses. Data were expressed as mean ± standard deviation. Intergroup differences were analyzed by the t-test. Kaplan-Meier curves were introduced for survival analysis. Chi-square test was performed to evaluate the correlation between miR-138-5p level with pathological indexes of PCa patients. P < 0.05 was considered as statistically significant.
Results miR-138-5p was down-regulated in PCa tissues and cell lines associated with PCa progression. We first focused insight into the expression level of miRNAs form TCGA database, and miR-138-5p with significant statistical difference was finally selected (Fig. 1A). In order to investigate the role of miR-138-5p in PCa development, qRT-PCR was performed to evaluate the expression of miR-138-5p in PCa tissues and cells. As showed in Fig. 1B, miR-138-5p was downregulated in PCa tissues, compared with paracancerous normal tissues. Similarly, miR-138-5p was also down-regulated in PCa cell lines than that of Prostate epithelial cell line (RWPE-1) (Fig. 1E). miR-138-5p expression was correlated with clinicopathologic characteristics and overall survival in PCa patients The clinical and pathological characteristics and follow-up data of enrolled PCa patients were collected for further analyses. According to the median level of miR-138-5p, PCa patients were assigned into two groups as high-miR-138-5p level group and low-miR-138-5p level group. As shown in Table 1, the relationships between the expression of miR-138-5p and Age, Tumor size, Gleason score, Lymph node metastasis and Bone metastasis of PCa patients were analyzed. The results found that these PCa patients with low-miR-138-5p expression had a higher Gleason score, Lymph node metastasis and Bone metastasis of PCa, compared with the patients with high-miR-138-5p expression (Fig. 1C). In addition, Kaplan-Meier methods revealed the poor prognosis in PCa patients of low-miR-138-5p level group than that of high-miR-138-5p level group (Fig. 1D). Table 1 Association of miR-138-5p expression with clinicopathologic characteristics of prostate cancer. miR-138-5p alleviated the proliferation and metastasis of PCa cell lines To explore the biological function of miR-138-5p in PCa cell lines was analyzed by CCK-8 cell proliferation assay, Tranwell assay and Wounding healing assay, respectively. miR-138-5p overexpression vectors were successfully constructed in the PC-3 and DU-145 cell lines, respectively ( Fig. 2A). It was found by the CCK-8 assay that the cell proliferation ability of miR-138-5p mimic was remarkably decreased in PCa cell lines, compared with NC mimic (Fig. 2B). Transwell assay revealed that the metastasis ability of PCa cells was significantly decreased in miR-138-5p mimic, compared with NC mimic (Fig. 2C). In addition, Wound healing assay showed that the overexpression of miR-138-5p could hinder the invasion and crawling ability of PCa cell lines (Fig. 2D). These results suggested that miR-138-5p could inhibited cell proliferation and metastasis in PCa.
In addition, the results of qRT-PCR also revealed the same trend (Fig. 3D). To further uncover the biological role of FOXC1 in PCa, we constructed pcDNA3.1-NC and pcDNA3.1-FOXC1. qRT-PCR found that miR-138-5p level was found to be significantly down-regulated in PCa cell lines transfected with pcDNA-FOXC1, compared to that transfected with pcDNA-NC (Fig. 3E). Luciferase reporter assay verified that miR-138-5p could indeed combine with FOXC1 through specific sequences (Fig. 3F &   Fig. 3G). Additionally, a significant negative correlation was identified to detect the expression levels of miR-138-5p and FOXC1 in PCa tissues (Fig. 3H).

miR-138-5p negatively regulated FOXC1 to inhibit the malignant progression of PCa
To further explore the specific regulatory mechanisms in which miR-138-5p exactly regulated FOXC1 to inhibit malignant progression of PCa. Firstly, the overexpressed endogenous FOXC1 was established with a FOXC1 overexpressing plasmid, and an empty overexpressing plasmid NC, to transfect PCa cell lines with miR-138-5p mimic. qRT-PCR demonstrated that PCa cells transfected with pcDNA3.1-FOXC1 had significantly lower miR-138-5p expression levels than cells transfected with pcDNA3.1-NC (Fig. 4A). In addition, Western Blotting demonstrated that PCa cells transfected with pcDNA3.1-FOXC1 had significantly higher FOXC1 expression levels than cells transfected with pcDNA3.1-NC (Fig. 4B). Subsequently, overexpression of FOXC1 was demonstrated to be able to counteract the effects of miR-138-5p mimic on the metastasis of PCa cells by Tranwell assay and Wound healing assay (Fig. 4C&Figure 4D). Therefore, these results revealed that miR-138-5p could inhibit the malignant progression of PCa through modulating FOXC1.
Over-expression of miR-138-5p suppressed the PCa in vivo tumorigenicity In an in vivo tumorigenicity assay, NC mimic or miR-138-5p mimic transduced PC-3 cells were subcutaneously inoculated into the abdominal compartments of athymic nu/nu mice for 6 weeks. The volumes of PC-3 xenografts were calculated weekly. It showed that, in vivo tumor growth was significantly suppressed by miR-138-5p over-expression (P < 0.05; Fig. 5A&Figure 5B). Subsequently, we validated the reduction of weight in tumor-forming tissues of nude mice injected with miR-138-5p mimic (P < 0.05; Fig. 5C). The results of qRT-PCR revealed that the overexpression of miR-138-5p in the tumor tissues of nude mice could decrease miR-138-5p expression level (Fig. 5D). In addition, compared with NC mimic, FOXC1 expression dramatically decreased in the tumor tissues of nude mice with miR-138-5p mimic by Western Blotting (Fig. 5E). Immunohistochemistry showed that the level of miR-138-5p mimic-transduced PC-3 xenografts significantly decreased than NC mimictransduced xenografts (Fig. 5F).

Discussion
The etiology of PCa is complex, and no clear conclusion about the pathogenesis is found [1,7,9].
Nowadays, the understanding of the pathogenesis and biological behavior of PCa still has great limitations [8,9]. The development of PCa is a multi-factors process, influenced by a variety of biomolecules and regulated by signaling pathways [11,12]. With the progress of the Human Genome Project, researches on molecular level have been extensively conducted to detect the differential expression profile of tumor genes, which is of great significance for exploring the molecular mechanism of the development of PCa, and finding the molecular biomarkers in the early diagnosis and prognosis of PCa [13,14].
MiRNAs not only participate in normal physical biological processes, but also regulate tumor progression at transcriptional and post-transcriptional levels [15,16]. It is estimated that over 30% of human genes and cellular processes are regulated or controlled by miRNAs [17,20]. In the past, oncogenes or tumor-suppressor miRNAs were used to develop as drug targets for tumor treatment [15,18,20]. Nowadays, tumor-related miRNAs have been well concerned, exerting more crucial application in early-stage diagnosis, target therapy and effective prognosis of tumors [13,14,19]. As a member of tumor-associated miRNAs family, miR-138-5p is located on chromosome Xq38.13 [24].
Previous researches showed that miR-138-5p could inhibit the malignant progress of several human tumors [23,24]. In this study, miR-138-5p was down-regulated in PCa tissues and cell lines. Besides, these PCa patients with low-miR-138-5p expression had a higher Gleason score, Lymph node metastasis, Bone metastasis and poor prognosis of PCa, compared with the patients with high-miR-138-5p expression. Thus, the above results suggested that miR-138-5p might act as anti-tumor effect in the progression of PCa. In order to further investigate the biological function of miR-138-5p in PCa cell lines, CCK-8, Tranwell assay and Wounding healing assay were used to introduce that miR-138-5p mimic could inhibit the proliferation and metastasis of PCa cell lines. The above results provided a theoretical basis for revealing the mechanism of PCa development. Of course, the specific molecular mechanism of signal transduction in PCa need to further study.
Regulatory mechanism of miRNAs depends on the expressions and functions of their target genes [15,20]. A miRNA degrades target mRNA or suppress its translation by base pairing with 3'UTR of the mRNA [15,16]. The degree of base pairing decides the degradative or translation inhibitory effect of a miRNA, that is, complete base pairing leads to mRNA degradation; otherwise, translation inhibition is achieved [27,28]. MiRNAs only account for only 1% of the whole human genome, but they are able to regulate more than 30% protein-encoding genes [29]. Multiple miRNAs could precisely regulate a single target gene [16,22]. Bioinformatics analysis and luciferase reporter gene assay showed that FOXC1 was the target gene of miR-138-5p through predicting in the miRDB, TargetScan and StarBase.
FOXC1 level was up-regulated in PCa tissue and cell lines. Based on these findings, we hypothesized that miR-138-5p might act as a ceRNA in the progression of PCa and found that miR-138-5p negatively regulated FOXC1 expression in PCa tissue. In addition, qRT-PCR showed that compared with NC mimic, miR-138-5p mimic could decreased the level of FOXC1. In order to explore the associations between miR-138-5p and FOXC1 in the development of PCa, the overexpression of FOXC1 was found to reverse the metastasis ability of miR-138-5p mimic on PCa cells, thus promoting the malignant progression of PCa.

Conclusions
In summary, the down-regulated miR-138-5p was closely associated with Gleason score, Distant    Over-expression of miR-138-5p inhibited tumorigenic ability in nude mice. (A,B). Tumor volume growth curves were calculated for different nude mice after injection of NC mimic and miR-138-5p mimic, respectively; (C). Tumor weight growth curves were calculated after