Non‐coding RNA MFI2‐AS1 promotes colorectal cancer cell proliferation, migration and invasion through miR‐574‐5p/MYCBP axis

Abstract Objective Long non‐coding RNAs (lncRNAs) and microRNAs (miRNAs) play essential roles in the tumour progression. LncRNAs mostly act as competing endogenous RNAs (ceRNAs) by sponging miRNAs. This study aimed to study the association of a novel lncRNA MFI2‐AS1 with miR‐574‐5p/MYCBP axis in the development of colorectal cancer (CRC). Methods Ninety‐four CRC tissues and paired adjacent non‐tumour tissues were included in our study. The relative expression level of MFI2‐AS1 was detected, and its relationship with clinico‐pathological factors was analysed. Then, the CRC cells lines (LoVo and RKO) were transfected with MFI2‐AS1 siRNA, miR‐574‐5p mimics and inhibitors. Cell proliferation, migration, invasion, cell cycle distribution and DNA damage in response to different transfection conditions were examined. Dual‐luciferase reporter assay was performed to identify the target interactions between MFI2‐AS1 and miR‐574‐5p, miR‐574‐5p and MYCBP. Results LncRNA MFI2‐AS1 and MYCBP were up‐regulated in CRC tissues when compared with adjacent non‐tumour tissues. The expression levels of MFI2‐AS1 were significantly associated with tumour histological grade, lymph and distant metastasis, TNM stage and vascular invasion. Both MFI2‐AS1 siRNA and miR‐574‐5p mimics inhibited proliferation, migration and invasion in LoVo and RKO cells. The transfection of miR‐574‐5p inhibitor showed MFI2‐AS1 siRNA‐induced changes in CRC cells. Dual‐luciferase reporter assay revealed target interactions between MFI2‐AS1 and miR‐574‐5p, miR‐574‐5p and MYCBP. Conclusions These findings suggested that lncRNA MFI2‐AS1 and MYCBP have promoting effects in CRC tissues. LncRNA MFI2‐AS1 promoted CRC cell proliferation, migration and invasion through activating MYCBP and by sponging miR‐574‐5p.


| INTRODUC TI ON
Increasing evidences show the crucial roles of genetic and epigenetic dysregulation in the generation and development of tumours. [1][2][3] Recently, the multifunctional roles of non-coding RNAs, including microRNA (miRNA) and long non-coding RNA (lncRNA), have been identified, which are deeply interested and extensively studied by the researchers.
Long non-coding RNAs are important and play diverse roles in regulating genetic transcription, 4 modulating embryonic and neoplastic differentiation, 5,6 pathogenesis of multiple diseases 7,8 and drug resistance in tumour cells. [9][10][11] LncRNA by itself cannot regulate these biological processes. The most widely recognized theory is that the lncRNA acts as a competing endogenous RNA (ceRNA) by decoying miRNAs to regulate the expression of miR-NA's target genes and change the level of transcription and translation products. In addition, some lncRNAs have similar structure to mRNAs, including a polyA tail and promoter region. The latter empowers the lncRNAs to manage the transcription of its downstream genes directly.
Theoretically, in ceRNA, lncRNA acts as a miRNA sponge or trapper to modulate miRNA-mRNA axis-mediated biological processes.
In the field of cancer, there are countless studies that show the regulation of lncRNAs in cancer pathogenesis, development, metastasis and prognosis. For instance, the widely studied lncRNAs, including MALAT1, H19 and HOTAIR, have been reported to promote tumour aggressiveness. [12][13][14][15] The expression of lncRNA HOTAIR in epithelial cancer cells increased breast cancer invasiveness and metastasis via enforcing the expression of polycomb repressive complex 2 14 ; and MALAT1 expression promotes the aggressiveness of renal cell carcinoma (RCC). 12 These data suggested the important roles of lncRNAs in tumorigenesis and development.
Colorectal cancer (CRC) is a common cause of cancer-related deaths worldwide with high morbidity, especially in the older population. 16,17 The 5-year survival rate of patients with CRC varied in stages, and at stage IV or metastatic stage is approximately 10%. 18 There are several studies that discussed the association of lncRNA dysregulation with the occurrence and progression of CRC. [19][20][21][22][23] These studies demonstrated the complex mechanism of CRC pathogenesis by lncRNAs.
According to a recent study, the lncRNA MFI2-AS1 (chr3:196729777-196731615) has been identified and was up-regulated in pancreatic cancer cells and sporadic localized clear-cell RCC (ccRCC). 24,25 Flippot et al showed that the RNA MFI2-AS1 expression was strongly associated with the recurrence and poor disease-free survival (DFS) of patients with sporadic localized clear-cell RCC. 25 Using bioinformatics analysis, we predicted that MFI2-AS1 sponged miR-574-5p, a candidate oncogene that play multifunctional roles in cancer metastasis, including thyroid carcinoma 26,27 and CRC. 28 However, no study suggested the association of lncRNA MFI2-AS1/miR-574-5p axis with the tumorigenesis or progression of CRC. Hence, in order to investigate this, we detected the expression of lncRNA MFI2-AS1 in CRC tissues and in vitro experiments were performed to assess the association of potent lncRNA MFI2-AS1/ miR-574-5p axis with CRC development.

| Clinical specimens
A total of 94 CRC tissues and paired adjacent non-tumour tissues were collected from patients undergoing surgical resection from January 2013 to June 2014 at Xiangya Hospital, Central South University.
None of them had received radiotherapy or chemotherapy prior to surgical resection. All isolated samples were snap frozen and then stored at −80°C before RNA extraction. Our experimental protocol was approved by the Ethics Committee of Central South University.
Written informed consent was obtained from all participants.

| Data mining
The data of the expression of MFI2-AS1 in CRC were acquired from the Gene Expression Profiling Interactive Analysis (GEPIA) (http:// gepia.cancer-pku.cn/), a bioinformatics database based on TCGA and GTEx, facilitating various online analysis including differential expression analysis, box plotting, patient survival analysis and so on.
The relevance of DFS and overall survival (OS) rates with the expression of MFI2-AS1 was also analysed by the GEPIA online database. 29

| Fluorescence in situ hybridization
The in situ expression of lncRNA MFI2-AS1 was examined using RNA fluorescence in situ hybridization (FISH) staining as previously described. The fluorescent RNA FISH probe (dye on the 3′ end, coupled with Alexa Fluor 594) was purchased from Genechem Co. Ltd. (China).
The cells were then fixed on slides using 4% paraformaldehyde and incubated with probes (50 nmol/L) at 37°C overnight. Cells were then counterstained with propidium iodide (PI). Digital fluorescent photographs were captured by using a fluorescent microscope (Olympus).
All experiments were performed in triplicates. All cells were incubated in RPMI 1640 medium at 37°C in 5% CO 2 for 72 hours.

| Total RNA isolation and quantitative realtime PCR
Trizol reagent (TaKaRa) was used to extract total RNA from tissue samples and cultured cells. The first cDNA strand was synthesized according to the methods provided by Bestar qPCR RT kit (DBI Bioscience) and amplified using specific primers (Invitrogen) listed in Table 1. Quantitative real-time PCR (qRT-PCR) analysis was performed to determine the relative expression level of lncRNA and miRNA according to the following conditions: 94°C for 2 minutes, followed with 40 cycles of 94°C for 20 seconds, 58°C for 20 seconds, 72°C for 20 seconds and finally extended at 72°C for 4 minutes.
Amplification was implemented using a Bestar ® Sybr Green qPCR master mix kit (DBI Bioscience) on Agilent Stratagene Mx3000P RT-PCR machine (Agilent Technologies). The relative expression level of detected genes was determined using the 2 −∆∆Ct methods.
GAPDH and U6 gene were used as internal reference genes for lncRNA and miRNA, respectively.

| In vitro transwell assay
Cell invasion ability was detected using 24-well invasion chambers (Coring) coated with Matrigel (BD Biosciences). The upper chamber was supplemented with serum-free RPMI 1640 medium and 1 × 10 5 CRC cells, and the lower chamber was filled with full RPMI 1640 medium supplemented with 10% FBS (Hyclone). The chambers were then maintained at 37°C in 5% CO 2 for 48 hours. Subsequently, the cells that were adhered to the undersurface of the filter membranes were removed, and the invaded cells were then fixed and stained with crystal violet for cell counting. An olympus microscope (CX41) was used to capture the digital photographs at five arbitrarily selected (non-overlapped) fields. Then the average number of the invaded cells was counted.

| Wound healing assay
Cell migration was detected using in vitro wound healing assay.

Cells were placed into 24-well plates and incubated for 24 hours
for the formation of monolayer on the bottom plate. After that, a straight line was scratched onto the monolayer using a 200 μL micropipette tip. After 48 hours, the wound width was analysed using Axio Observer, the microscope (Carl Zeiss AG). All experiments were performed in triplicates.

| Flow cytometric analysis
Cell cycle distribution was detected using flow cytometry. Cells were transfected under different conditions for 48 hours and then were harvested and fixed. For cell cycle distribution, cells were incubated with PI for 20 minutes. Then, the cell cycle distribution was analysed using the BD FACS Calibur flow cytometry (BD Biosciences).

| Hoechst fluorescent staining
Hoechst fluorescent staining was performed to detect DNA damage and cell apoptosis. In brief, the cells were plated in 96-well plate TA B L E 1 The PCR primers used in this study

Gene name
Primers

| Western blot analysis
The fold change of MYCBP protein in CRC tissues and trans-
Beforehand, plasmids were constructed into pMIR by inserting ei-

| Statistical analyses
All statistical analyses were performed using GraphPad Prism 6.0 software. Data were expressed as mean ± standard deviation. Differences in the demographic characteristics between groups were analysed using unpaired t test or chi-square test. P < 0.05 was considered to be statistically significant.

| MFI2-AS1 is up-regulated in CRC tissues
The results of the box plots revealed that MFI2-AS1 expression was significantly higher in CRC tissues by analysing the data form GEPIA ( Figure 1A). The survival curves of CRC patients showed that the expression level of MFI2-AS1 was significantly associated with DFS rate (P < 0.05; Figure 1B) and OS rate (P < 0.05; Figure 1C) by GEPIA. This cell line, where its expression was down-regulated (P < 0.05, Figure 1E).
Moreover, we found that the expression of MFI2-AS1 was related with several clinico-pathological factors, and high MFI2-AS1 was significantly correlated with tumour histological grade, lymph involvement, distant metastasis, TNM stage and vascular invasion (P < 0.05 for all, Table 2). There was no significant association found between MFI2-AS1 expression and age, gender, T stage, pre-operative serum CEA and CA 19-9 levels, and the presence of perineural invasion (P > 0.05, Table 2).

| Inhibition of MFI2-AS1 impedes CRC cell proliferation and metastasis
Using FISH technique, we detected the expression of lncRNA MFI2-AS1 in the cytoplasm of CRC cells ( Figure 1F). In order to investigate whether the MFI2-AS1 expression was associated with CRC development and metastasis, the CRC cell lines (LoVo and RKO) were transfected with siRNA target lncRNA MFI2-AS1 (Figure 2A). The results showed that the inhibition of MFI2-AS1 expression dramatically suppressed the cell viability (P < 0.01, Figure 2B), wound healing speed (P < 0.05, Figure 2C) and invasion of LoVo and RKO cells (P < 0.05, Figure 2D) compared with blank control. Further, flow cytometry analysis showed that the inhibition of lncRNA MFI2-AS1 expression increased the percentage of cells at G1 stage and reduced the cells at S stage ( Figure 3A and Figure S1). Hoechst 33258 staining showed that siRNA transfection increased Hoechst 33258-positive cells in LoVo and RKO cells, respectively ( Figure 3B). These data suggested that the MFI2-AS1 inhibition suppressed CRC cell proliferation and invasion via arresting the cell cycle at G1 phage.

| MiR-574-5p is a negative target of MFI2-AS1
There are accumulating evidences showing that lncRNA regulates the biological processes by sponging miRNAs. In this study, a series of miRNAs were predicted be potential in sponging by MFS1-AS1.
We predicted that the miR-574-5p was a target of lncRNA MFI2-AS1 by DIANA tools ( Figure 4A) since the fold change in the expression of miR-574-5p was extraordinary higher than other miRNAs ( Figure   S2). Dual-luciferase reporter assay showed that the administration of miR-574-5p mimic significantly reduced the relative luciferase intensity in 293T cells transfected with wild-type 3′-UTR sequence containing putative miRNA binding sites (P < 0.05, Figure 4B). In addition, the MFI2-AS1 inhibition by siRNA significantly up-regulated Note: Low, fold change lower than 0.5. High, fold change larger than 0.5 (cut-off = 2.71).

| MiR-574-5p suppresses CRC cell proliferation and metastasis by targeting MYCBP
and PTPRU were all the target genes of miR-574-5p by TargetScan Human. MYCBP expression was changed more significantly after miR-574-5p depletion, and so we predicted MYCBP as a target of miR-574-5p ( Figure 6C, 122 ~ 128 nt). Dual-luciferase reporter assay showed that the administration of miR-574-5p mimics significantly reduced the relative luciferase intensity (P < 0.01, Figure 6D). Moreover, we found that the expression of MYCBP protein was reduced by miR-574-5p mimics in LoVo and RKO cells ( Figure 6E). These data suggested that miR-574-5p expression might suppress CRC cell proliferation and invasion via MYCBP expression inhibition.
F I G U R E 2 MFI2-AS1 depletion impedes CRC cell proliferation and metastasis. A, the relative expression of MFI2-AS1 in cells transfected with MFI2-AS1 siRNA and corresponding negative control. B, cell viability analysis by CCK-8 assay. C, wound healing assay. Cells are transfected with siRNA target MFI2-AS1 for 48 h. D, cell invasion by transwell migration assay. Magnification, ×400. * and ** note P < 0.05 and P < 0.01 vs NC, respectively. NC, negative control

| MFI2-AS1 regulates CRC cell proliferation and metastasis and MYCBP expression by sponging miR-574-5p
According to the dual-luciferase reporter assay results, we speculated that lncRNA might function as a ceRNA for MYCBP in regulating CRC cell proliferation. We determined the up-regulated expression of MYCBP protein in CRC tumour tissues by comparing with adjacent tissues ( Figure 7A). The MFI2-AS1 siRNA-induced expression of miR-574-5p in CRC cells was dramatically suppressed by administration of miR-574-5p inhibitor (P < 0.01, Figure 7B). In contrast, the decreased MYCBP protein by MFI2-AS1 siRNA was up-regulated by miR-574-5p inhibitor ( Figure 7C). These results suggested that the expression of MYCBP was competitively regulated by miR-574-5p and lncRNA F I G U R E 3 Cell cycle distribution and apoptosis analysis. A, cell cycle analysis was performed using flow cytometry. B, Hoechst staining for cells in response to siRNA. Cells were transfected with siRNA target MFI2-AS1 for 48 h. * notes P < 0.05 vs NC. NC, negative control F I G U R E 4 MFI2-AS1 targets to miR-574-5p. A, the predicated bind sites of miR-574-5p to MFI2-AS1. B, dualluciferase reporter assay. C, the relative expression level of miRNA in CRC cells transfected with MFI2-AS1 siRNA. *, ** and *** note P < 0.05, P < 0.01 and P < 0.001 vs NC, respectively. MUT, mutant_MFI-AS1, NC, negative control, WT, wild type MFI2-AS1. Further analysis showed that MFI2-AS1 siRNA inhibited cell viability (P < 0.01, Figure 7D), migration (P < 0.01, Figure 7E,F) and invasion (P < 0.01, Figure 7G). The increased percentage of cells at G1 phase (P < 0.01, Figure 8A and Figure S4) and Hoechst 33258-positive cells ( Figure 8B) were significantly rescued by the administration of miR-574-5p inhibitor. These data revealed that the MFI2-AS1 siRNA inhibited CRC cell proliferation and metastasis and MYCBP inhibition could be rescued by miR-574-5p inhibitor, suggesting that MFI2-AS1 promoted CRC metastasis via sponging miR-574-5p. LncRNA MFI2-AS1 is a novel lncRNA identified from drug-resistant pancreatic cancer cells using next-generation RNA sequencing. 24 Flippot et al showed that the up-regulation of MFI2-AS1 was associated with poor survival of patients with sporadic localized ccRCC, and patients with undetectable MFI2-AS1 had favourable outcomes. 25 We identified that the expression of MFI2-AS1 was significantly up-regulated in CRC tumour tissues when compared with adjacent  33 However, MiR-574-5p is also identified as an oncogene in several cancers, including thyroid carcinoma 26,27 and CRC. 28  Our present study showed that the miR-574-5p mimic-induced proliferation inhibition, migration and invasion of CRC cells were associated with MYCBP. We found that both MYCBP and MFI2-AS1 were up-regulated in CRC tissues when compared with adjacent non-tumour tissues. These data suggested the association of MFI2-AS1/ miR-574-5p/MYCBP axis with CRC pathogenesis and development. and miR-574-5p as well as between miR-574-5p and MYCBP was detected using dual-luciferase reporter assay. Taken together, our study suggested the facilitative role of lncRNA MFI2-AS1 in CRC through MYCBP and by sponging miR-574-5p.

ACK N OWLED G EM ENTS
This study was supported by the Nature Scientific Foundation of

CO N FLI C T O F I NTE R E S T
Authors declared no conflict of interest.

R E FE R E N C E S
F I G U R E 8 MiR-574-5p inhibition rescues MFI2-AS1 siRNA-influenced cell cycle distribution, apoptosis and DNA damage. A, Cell cycle analysis was performed using flow cytometry. Cells were transfected with miR-574-5p mimics for 48 h. B, Hoechst staining for cells in response to miR-574-5p mimics and MFI-AS1 siRNA. Cells were transfected with different conditions for 48 h. * and ** note P < 0.05 and P < 0.01 vs. Blank, respectively. # and ## note P < 0.05 and P < 0.01 vs siRNA, respectively. NC, negative control