LncRNA MEG3 inhibits the progression of prostate cancer by modulating miR‐9‐5p/QKI‐5 axis

Abstract This study was designed to detecting the influences of lncRNA MEG3 in prostate cancer. Aberrant lncRNAs expression profiles of prostate cancer were screened by microarray analysis. The qRT‐PCR and Western blot were employed to investigating the expression levels of lncRNA MEG3, miR‐9‐5p and QKI‐5. The luciferase reporter assay was utilized to testifying the interactions relationship among these molecules. Applying CCK‐8 assay, wound healing assay, transwell assay and flow cytometry in turn, the cell proliferation, migration and invasion abilities as well as apoptosis were measured respectively. LncRNA MEG3 was a down‐regulated lncRNA in prostate cancer tissues and cells and could inhibit the expression of miR‐9‐5p, whereas miR‐9‐5p down‐regulated QKI‐5 expression. Overexpressed MEG3 and QKI‐5 could decrease the abilities of proliferation, migration and invasion in prostate cancer cells effectively and increased the apoptosis rate. On the contrary, miR‐9‐5p mimics presented an opposite tendency in prostate cancer cells. Furthermore, MEG3 inhibited tumour growth and up‐regulated expression of QKI‐5 in vivo. LncRNA MEG3 was a down‐regulated lncRNA in prostate cancer and impacted the abilities of cell proliferation, migration and invasion, and cell apoptosis rate, this regulation relied on regulating miR‐9‐5p and its targeting gene QKI‐5.

MicroRNAs (miRNAs), belongs to a class of small non-coding regulatory RNA molecules, also involved in many biological processes, such as cell development, differentiation, proliferation and apoptosis. 3,8 LncRNAs and miRNAs are thought to interact with one another to create an additional layer of regulational complexity. 9 Ke et al revealed that silencing the lncRNA HOTAIR could inhibit malignant biological developments of human glioma cells through modulation of miR-326. 10 In addition, long non-coding RNA TUG1 acted as a tumour promoting factor, could contribute to tumorigenesis of human osteosarcoma depending on regulating miR-9-5p/POU2F1 expression. 8 Furthermore, it has also been reported that MEG3 could bind to miR-9 and regulated the expression of E-cadherin and FOXO1. 11 Therefore, we hypothesized that MEG3 might function as ceRNA for miR-9-5p in prostate cancer.
The Quaking (QKI) protein belongs to the signal transduction and activation of RNA (STAR) protein family and is a key post-transcriptional regulator, 12 which involved in several cancer regulatory mechanisms, including gastric cancer, myocardial ischaemia and oligodendrocyte. 13 Recent findings indicated that genomic depletion of QKI-5 increases proliferation and dedifferentiation of cancer cells, which indicated that QKI-5 is a tumour suppressor for many cancer types. 14 In our study, a series of experiments carried out that QKI-5 involved in regulating the development of prostate cancer, which was consistent with the previous reports. 15 In this study, relied on detecting the down-regulated lncRNA MEG3 in prostate cancer, the interactions among MEG3, miR-9-5p and QKI-5 were investigated and the influences on prostate cancer were explored subsequently. Hence, lncRNA MEG3 regulated miR-9-5p/QKI-5 axis in prostate cancer.

| Clinical specimens
In this study, 85 pairs of prostate cancer tissues and adjacent tissues were collected from patients who underwent surgical resection at the Second Affiliated Hospital of Nanchang University. All patients were confirmed by experienced pathologists and none was received preoperative radiotherapy or chemotherapy. This study was approved by the Ethics Committee of the Second Affiliated Hospital of Nanchang University. All the tissues were stored in liquid nitrogen at À80°C for further experiments. Manassas, VA, USA). Lipofectamine TM 3000 (Invitrogen) was diluted with 250 lL DMEM medium. The cell transfection was followed by manufacturer's protocols. The carrier used in the experiment was pcDNA3.1. Cells were generally assigned to different groups as follows: (1) mock group and pcDNA3.1-MEG3 group; (2) negative control (NC) group, miR-9-5p mimics group, miR-9-5p inhibitor group and miR-9-5p mimics+ pcDNA3.1-MEG3 group; (3) NC group, miR-9-5p mimics group, pcDNA3.1-MEG3 group and miR-9-5p mimics+ pcDNA3.1-MEG3 group.
We identified the differentially expressed lncRNAs with a discriminating parameter of adjusted P value <.05 and fold change >2.

| QRT-PCR
Total RNA was isolated and reverse transcribed into cDNA according to M-MLV Reverse Transcriptase (TaKaRa, Tokyo, Japan). And then real-time PCR assay was proceed by SYBR Prime Script PLUS RT-RNA PCR Kit (TaKaRa). The reaction condition of PCR is 95°C for 30 seconds, 62°C for 40 seconds and repeat the process for 40 times in sequence. The relative expression levels were counted by 2 ÀDDCt method and StepOne software was employed to analysing the data. The internal reference we used was GAPDH, listed in Table 1 together with the primer sequence.

| Cell counting kit-8 (CCK-8) assay
The DU 145 cells were seeded in 96-well plates for evaluating proliferation ability by CCK-8 assay (Dojindo, Kumamoto, Japan). Cells were then cultured for 0, 1, 2, 3 or 4 days before addition 10 lL of CCK-8 (5 mg/mL) to the culture medium in each well. The absorbance at 450 nm and measured it by Thermo-max microplate reader (Thermo Fisher Scientific, Waltham, MA, USA).

| Transwell assay
The 24-well transwell chambers (Costar, Corning, Switzerland) with Matrigel-coated membranes were used for invasion assay. DU 145 cells were seeded on the RPMI 1640 medium and the invading cells in the lower chamber were stained by 0.1% crystal violet.
The cells were observed under the phase contrast microscope (9200) (Nikon).

| Luciferase reporter assay
The fragment of MEG3 containing the predicted miR-9-5p binding site was amplified by PCR. And then was cloned into a Dual-luciferase miRNA Target Expression Vector (Promega, Madison, WI, USA) for forming MEG3-wild-type (MEG3-Wt). The same approach was used to forming MEG3-mutated-type (MEG3-Mut).

| RNA immunoprecipitation (RIP) assay
RNA immunoprecipitation assay was performed by Imprint RNA immunoprecipitation kit (Sigma-Aldrich, St. Louis, MO, USA) referring to the recommended protocols of manufacturer. Firstly, IgG-induced chondrocytes were collected and resuspended in RIP lysis bufer (Solarbio), subsequently centrifuged at 12 000 g for 5 minutes. Ten, cell lysate was incubated with anti-Argonaute2 (anti-Ago2) or antiIgG (negative control) overnight at 4°C, followed by the addition of Protein A magnetic beads to get the immunoprecipitation complex. Total RNA was isolated using GenElute TM Total RNA Purifcation Kit (Sigma-Aldrich). Lastly, the relative enrichment of MEG3 and miR-9-5p were determined by RT-qPCR analysis.

| Western blot
DU 145 cells were lysed in RIPA buffer. Proteins were separated on 10% SDS-PAGE and electro-transferred to nitrocellulose membranes.
And then were incubated with anti-QKI-5 (0.5 mg/mL, Boston Biochem, Cambridge, MA, USA) and anti-GADPH (Boston Biochem) at 4°C overnight, and then incubated with a HRP-labelled secondary antibody IgG (Boston Biochem) at room temperature for 3 hours.

| Xenograft mice model in vivo
The

| Statistical analysis
Statistical analyses were performed with GraphPad Prism 6.0 software and data were expressed as mean AE SD. Statistical comparisons were made by one-way analysis of variance (ANOVA) and P < .05 instructed a statistically significant difference.
The differentially expressed lncRNAs were screened and showed in the volcano plot by microarray analysis ( Figure 1A). The top 10 up-and down-regulated lncRNAs were chosen to draw the heat map ( Figure 1B) and MEG3 ranked second among these lncRNAs.
The results of qRT-PCR revealed that the level of lncRNA MEG3 was significantly decreased in prostate cancer tissues ( Figure 1C).
Meanwhile, a lower level MEG3 was obtained in the 4 prostate WU ET AL. 3.3 | MEG3 served as a sponge of miR-9-5p and inhibited prostate cancer The targeting relationship between MEG3 and miR-9-5p was showed in Figure 3A. And then the luciferase reporter assay indicated that MEG3 served as a sponge of miR-9-5p. The results of luciferase reporter gene assay validated that cotransfected with MEG3-wt and miR-9-5p obtained a weakened luciferase activity compared with the miR-NC group and the MEG3-mut group ( Figure 3B). Furthermore, RIP assay was performed with Ago2 antibody to confirm the potentially endogenous interaction between MEG3 and miR-9-5p. The results revealed that MEG3 and miR-9-5p were largely captured by anti-Ago2 compared with the negative control in IgG-induced chondrocytes (Figure 3C). The results of qRT-PCR verified that miR-9-5p expression significantly increased in miR-9-5p mimics group. Besides, miR-9-5p was hardly expressed in the miR-9-5p inhibitor group. The miR-9-5p expression level of pcDNA3.1-MEG3 & miR-9-5p mimics group was same as NC group ( Figure 3D). CCK-8 assay showed that the miR-9-5p mimics group exhibited higher cell proliferative capacity than other 4 groups. The cell proliferative capacity in pcDNA3.1-MEG3 group was reduced, which was the same as miR-9-5p inhibitor group. Cell proliferative capacity of pcDNA3.1-MEG3 & miR-9-5p mimics group was same as NC group ( Figure 3E). The results of FCM assay revealed that cell apoptosis rate of pcDNA3.1-MEG3 group and miR-9-5p inhibitor group were significantly higher than other 3 groups, whereas miR-9-5p mimics group was the lowest and pcDNA3.1-MEG3 & miR-9-5p mimics group was same with NC group (Figure 3F,H). After transwell assay, cell invasion ability in pcDNA3.1-MEG3 group and miR-9-5p inhibitor group was weakened, whereas the ability in the miR-9-5p mimics group was strengthened. The pcDNA3.1-MEG3 & miR-9-5p mimics group was same with NC group (Figure 3G,I). Concluded from the above experiments, MEG3 served as a sponge of miR-9-5p to inhibit prostate cancer.
3.4 | MEG3 served as a sponge of miR-9-5p and regulated QKI-5 MiR-9-5p was found to have targeting relations with QKI-5 by miR-Base predictor. And then, the targeting relationship between miR-9-5p and QKI-5 was also showed in Figure 4A. The luciferase intensity of DU 145 cells indicated that cotransfected with QKI-5 wild-type and miR-9-5p was significantly reduced, whereas QKI-5 mutant type was hardly changed based on luciferase reporter assay ( Figure 4B).
The results of qRT-PCR verified that QKI-5 expression level of pcDNA3.1-QKI-5 group was significantly promoted than other 3 groups, whereas miR-9-5p mimics group was significantly reduced.
The results of FCM assay revealed that cell apoptosis rate of miR-9-  Figure 4G,H). QKI-5 expression level of miR-9-5p mimics group was significantly decreased, whereas that of pcDNA3.1-MEG3 group was increased detected by Western blotting ( Figure 4I). As a result, MEG3 served as a sponge of miR-9-5p to regulate QKI-5.

| MEG3 inhibited prostate cancer in vivo
To verify the oncogenic role of MEG3 in prostate tumorigenesis, a xenograft mouse model was constructed. Overexpression MEG3 decreased tumour size after 3 weeks results compared with NC group ( Figure 5A-C). According to Western blot results, the MEG3 overexpression group exhibited higher QKI-5 expression than that in NC group After 5 weeks ( Figure 5D). Moreover, immunohistochemical analysis revealed that MEG3 overexpression increased expression level of QKI-5 ( Figure 5E). Collectively, these results indicated that MEG3 inhibits prostate cancer in vivo.
LncRNAs involved in many cancer-related research mechanisms. 2 For example it is proved that SNHG1 is up-regulated in lung cancer

cells lines and in vitro experiments and down-regulation of SNHG1
suppresses lung cancer cell proliferation. 10  What's more, it's also been reported that up-regulation of MEG3 in prostate cancer cell lines induces cell apoptosis and G0/G1 phase arrest. 16 The results of our study focused on detecting the effect of lncRNA MEG3 in PCa, we revealed that MEG3 was a down-regulated lncRNA in PCa and it played a role of anti-cancer factor.
Mounting evidence suggested that lncRNAs may exert functions by targeting miRNAs. 17 Bian et al confirmed that UCA1 worked as a ceRNA in colorectal cancer, regulated the expression of miR-204-5p and induced the resistance to 5-FU in CRC. 18 He et al 9 found another lncRNA that regulated cardiac hypertrophy by targeting miRNA-489, named as CHRF. Moreover, a recent study revealed that lncRNA SNHG1 could promote cell progression by antagonizing miR-199a-3p in prostate cancer. 10 Also, it has been found that MEG3 exerts its function through negatively regulating miR-9 expression by directly interaction in vascular endothelial cells. 19 In present study, we also found that MEG3 functions as a ceRNA for miR-9-5p in prostate cancer, which indicated that a mutual antagonism between MEG3 and miR-9-5p on impacting prostate cancer progression. This is useful for understanding the molecular mechanisms associated with prostate cancer. downstream targeting molecule Bcl-2. 22 We revealed a result of MEG3 functioned as a ceRNA for miR-9-5p to regulate QKI-5, which was consistent with the previous reports. A report proved that MEG3 could regulate expression of E-cadherin and FOXO1 by competitively binding miR-9 in oesophageal squamous cell carcinoma. 11 Our findings revealed that overexpression of MEG3 and QKI-5 inhibited the progression of PCa, whereas overexpression of miR-9-F I G U R E 3 MEG3 served as a sponge of miR-9-5p and inhibited prostate cancer. A, The targeting relationship between MEG3 and miR-9-5p. B, The luciferase intensity of DU 145 cells transfected with MEG3 wild-type and miR-9-5p was significantly reduced. **P < .01, compared with miR-NC group. C, The interaction between MEG3 and miR-9-5p was detected by RNA immunoprecipitation (RIP) with Ago2 antibody. D, The expression of miR-9-5p in pcDNA3.1-MEG3 group was reduced, whereas miR-9-5p mimics group was increased. The miR-9-5p expression level in the miR-9-5p inhibitor group was the lowest. **P < .01, ***P < .001, compared with NC group, # P < .05, ## P < .01, compared with pcDNA3.1-MEG3 group, && P < .01, &&& P < .001, compared with miR-9-5p mimics group. E, CCK-8 assay showed that the miR-9-5p mimics group exhibited higher cell proliferative capacity. Cell proliferative capacity of pcDNA3.1-MEG3 group and miR-9-5p inhibitor group was decreased. *P < .05, **P < .01, compared with NC group, # P < .05, ## P < .01, compared with pcDNA3.1-MEG3 group, & P < .01, compared with miR-9-5p mimics group. F, H, FCM assay revealed that cell apoptosis rate of pcDNA3.1-MEG3 group and miR-9-5p inhibitor group were increased, whereas miR-9-5p mimics group was decreased. *P < .05, **P < .001, compared with NC group, # P < .05, ## P < .01, compared with pcDNA3.1-MEG3 group, & P < .05, && P < .01, compared with miR-9-5p mimics group. G, I, Cell invasion ability of pcDNA3.1-MEG3 group and miR-9-5p inhibitor group were decreased, whereas miR-9-5p mimics group was promoted by transwell assay. *P < .05, compared with NC group, # P < .05, ## P < .01, compared with pcDNA3.1-MEG3 group, & P < .05, && P < .01, compared with miR-9-5p mimics group. Scale bar = 20 lm 5p presented an inverse effects. However, there are also shortcomings in our research, for example the number of samples we used for this study was not large enough. In this study, we focused on QKI-5, which is just a member of QKI alternatively spliced mRNAs. Therefore, the further research on functions and mechanisms underlying of other QKI genes was needed to be carried out. In addition, there was not only one targeted miRNA have potential targeting relationship with MEG3 or QKI-5. However, we just focused on one targeting miRNA in this study.

| CONCLUSION
In conclusion, our study showed that MEG3 was a down-regulated lncRNA in human prostate cancer, and it could regulate miR-9-5p/ QKI-5 expression. We also demonstrated that overexpression of MEG3 could effectively inhibit the development of PCa through targeting miR-9-5p/QKI-5 axis in prostate cancer. Our findings provided novel evidences that lncRNAs acted as "microRNA sponges" in prostate cancer.