miR-27a-3p targeting RXRα promotes colorectal cancer progression by activating Wnt/β-catenin pathway

This study aimed to elucidate how miR-27a-3p modulates the Wnt/β-catenin signaling pathway to promote colorectal cancer (CRC) progression. Our results showed that the expression of miR-27a-3p was up-regulated in CRC and closely associated with histological differentiation, clinical stage, distant metastasis and CRC patients’ survival. miR-27a-3p mimic suppressed apoptosis and promoted proliferation, migration, invasion of CRC cells in vitro and in vivo. Whereas miR-27a-3p inhibitor promoted apoptosis and suppressed proliferation, migration, invasion of CRC cells in vitro and in vivo. Furthermore, RXRα was the target gene of miR-27a-3p in CRC. miR-27a-3p expression negatively correlated with RXRα expression in CRC tissues. The underlining mechanism study showed that miR-27a-3p/RXRα/Wnt/β-catenin signaling pathway is involved in CRC progression. In conclusion, our findings first demonstrate that miR-27a-3p is a prognostic and/or potential therapeutic biomarker for CRC patients and RXRα as miR-27a-3p targeting gene plays an important role in activation of the Wnt/β-catenin pathway during CRC progression.


miR-27a-3p expression correlates with CRC patient survival
To determine the significance of miR-27a-3p in CRC, we first performed quantitative real-time PCR on fresh human CRC specimens and their matched adjacent nontumor colorectal tissues (ANT). miR-27a-3p expression was higher in 10 (67%) of 15 CRC samples compared with their respective ANT ( Figure 1A). High miR-27a-3p expression was found in 3 of 6 CRC cell lines including HCT116, HT29, and SW1116 cell lines compared with human colonic epithelial cell line NCM460 ( Figure 1B). To further investigate the correlation of miR-27a-3p expression with clinicopathological features of CRC, we detected miR-27a-3p expression in 100 pairs of paraffin-embedded human CRC samples and respective non-tumor colorectal tissues by real-time PCR analysis. The result showed miR-27a-3p expression significantly increased in CRC compared with ANT. Notably, miR-27a-3p expression was associated with the pathological differentiation, Dukes staging, and distance metastasis of CRC (Table 1). The 3-year overall survival rate in CRC patients with miR-27a-3p high expression was 75.3%. However, the 3-year overall survival rate in CRC patients with miR-27a-3p low expression was 87.9%. There was a significant difference (p<0.01). The 5-year overall survival rate (71.1%) was significantly lower in CRC patients with miR-27a-3p high expression than 80.5% of the 5-year overall survival rate in CRC patients with miR-27a-3p low expression (p<0.05). Kaplan-Meier survival analysis showed that CRC patient survival was less in high miR-27a-3p expression group compared with low miR-27a-3p expression group (p=0.047) ( Figure 1C). Univariate Cox regression analysis showed that histological differentiation, Dukes staging, N classification, M classification, and miR-27a-3p expression were significantly associated with the prognosis of CRC patients. Multivariable Cox regression analysis showed that Dukes staging and miR-27a-3p expression were independent prognostic factors for CRC patients (Table 2-

MiR-27a-3p promotes CRC growth, migration and invasion and suppresses apoptosis in vitro and in vivo
We then examined the biological function of miR-27a-3p in CRC cells. As shown in Figure 2A-2D, cell proliferation was suppressed significantly in HCT116 cells transfected with miR-27a-3p inhibitor and enhanced in SW480 cells transfected with miR-27a-3p mimic compared with the control group, respectively. In addition, miR-27a-3p inhibitor and mimic significantly suppressed and promoted clone formation in HCT116 and SW480 cells compared with the control group at 400 and 600 different cell densities, respectively. Flow cytometry analysis showed that elevated apoptosis and cell cycle G1-S phase arrest were found in HCT116 transfected with miR-27a-3p inhibitor compared with the control group, respectively. Decreased apoptosis and cell cycle G1-S phase arrest were found in SW480 transfected with miR-27a-3p mimic compared with the control group, respectively ( Figure 2E-2H). miR-27a-3p mimic significantly increased SW480 cell migration by scratch wound healing assay and cell migration assay, respectively. Similarly, cellular invasion assay showed that miR-27a-3p mimic significantly increased SW480 cell invasion ( Figure  3A-3C). miR-27a-3p inhibitor significantly suppressed HCT116 cell migration and invasion compared with the control group, respectively (Figure 3D-3E).   To further investigate the role of miR-27a-3p on CRC growth and metastasis in vivo, we performed xenograft tumor assays using HCT116 and SW480 cells. As shown in Figure 4A-4C, xenograft tumor growth, weight, and volume were significantly inhibited in HCT116 treated with miR-27a-3p antagomir compared with the control group, respectively. However, xenograft tumor growth, weight, and volume significantly increased in SW480 treated with miR-27a-3p agomir compared with the control group, respectively ( Figure 4D-4F). The effect of miR-27a-3p on CRC metastasis was performed in metastatic animal model using HT29 cells. The results showed that pulmonary metastatic CRC was found in 2 of 6 mice treated with miR-27a-3p agomir but no metastatic CRC tumor in the lung (0/6 mice) was found in the control group ( Figure 4G). miR-27a-3p expression was significantly higher in pulmonary metastatic CRC tumor in treatment group than that in lung of control group by real-time PCR analysis (p<0.01, Figure 4H). RXRα is the target gene of miR-27a-3p and indispensable to miR-27a-3p-mediated oncogenic role in CRC To clarify the molecular mechanism of miR-27a-3p in CRC progress, we predicted that miR-27a-3p UGACACU could completely bind ACTGTGA of RXRα mRNA 3'UTR using TargetScan, Diana Tools, PicTar, and miRanda software. First, we analyzed miR-27a-3p and RXRα expression in 100 samples of CRC tissues. As shown in Table 4, High RXRα expression was found in 33% (12/36) of miR-27a-3p low expression group, low RXRα expression was found in 86% (55/64) of miR-27a-3p high expression group. A significantly negative correlation between miR-27a-3p and RXRα expression was found in CRC tissues (r=-0.227, p=0.023). Further study showed that RXRα mRNA and protein expression was increased in HCT116 cells transfected with miR-27a-3p inhibitor compared with the control group, respectively. RXRα mRNA and protein expression was suppressed in SW480 cells transfected with miR-27a-3p mimic compared with the control group, respectively ( Figure 5A-5D). To further investigate whether the predicted binding site of miR-27a-3p to the 3'UTR of RXRα was specific, we cloned the 3'UTR of RXRα (RXRα-WT) and its mutant variant (RXRα-Mut) into the downstream of luciferase reporter gene. Luciferase reporter assay showed that miR-27a-3p mimic-WT significantly suppressed luciferase activity of RXRα-WT 3'-UTR in SW480 and HCT116 compared with the control group, respectively. The suppressed luciferase activity of RXRα-WT 3'-UTR was reversed by miR-27a-3p mimic-Mut in HCT116 and SW480 cells, respectively. However, miR-27a-3p mimic-WT or miR-27a-3p mimic-Mut did not affect the luciferase activity of RXRα-Mut 3'-UTR in SW480 and HCT116 compared with the control group, respectively ( Figure 5E-5G). We further performed RNA Binding    Figure 5H-5I).
Immunoblotting showed that the ubiquitination of β-catenin was dramatically augmented by RXRα overexpression in HCT116 cells treated with MG132 compared with the control group ( Figure 8A). RXRα could be regulated by several protein kinases. To determine whether GSK3β phosphorylates RXRα and induces activation of Wnt/β-catenin pathway. Myc-RXRα and Flag-GSK3β plasmids were cotransfected in HCT116 cells. The anti-Myc antibody recognized the migrating upward band of RXRα, which stands for the phosphorylated RXRα. Phosphorylation inhibitor-calf intestinal alkaline phosphatase (CIP) impaired RXRα phosphorylation by GSK3β. Further study showed that LiCl, 6-bromoindirubin-30-oxime (BIO), and SB415286, which are inhibitors of GSK3β, suppressed RXRα phosphorylation by GSK3β in HEK293T and HCT116, respectively. Coimmunoprecipitation assay showed that there was a direct interaction between RXRα and GSK3β through GSK3β serine sites in HEK293 cells. However, transfection of mutant-Flag-GSK3β-K85R impaired the effect of RXRα binding to GSK3β serine sites. Moreover, GSK3β kinaseactivated mutant plasmid-GSK3β-S9A induced RXRα phosphorylation in HEK293T cells ( Figure 8B-8F).
Next, we examined potential p-Ser sites of the RXRα protein using the Phosphor motif Finder program and identified some hypothetical p-Ser sites, which mainly located within N-domain 100 amino acids, a residue in RXRα predicted to be a target of Ser and other kinases. To determine whether GSK3β phosphorylates RXRα through one or more specific p-Ser sites of RXRα, we constructed various RXRα deletion mutants and cotransfected these various DNAs and Flag-GSK3β into human HEK293T cells. Our result showed that deletion mutants of the N20 (ΔN20), N40 (ΔN40), and N60 (ΔN60) induced RXRα phosphorylation, respectively. However, deletion mutants of the N80 (ΔN80) and N100 (ΔN100) had no effect on Western blot showed that deletion mutants of the N20 (ΔN20), N40 (ΔN40), and N60 (ΔN60) induced RXRα phosphorylation by GSK3β, respectively. However, deletion mutants of the N80 (ΔN80) and N100 (ΔN100) had no effect on GSK3β-induced RXRα phosphorylation in HEK293T cells.
GSK3β-induced RXRα phosphorylation ( Figure 8G-8H). Three p-Ser candidate sites at Ser49, Ser66 and Ser78 were identified within the targeted domain of RXRα. To ascertain that Ser49, Ser66, and Ser78 are major p-Ser sites of RXRα induced by GSK3β, Ser49, Ser66 and Ser78 were mutated to Ala (A) in the full-length RXRα protein, respectively. GSK3β-induced p-Ser was markedly reduced for RXRα S49A , RXRα S66A and RXRα S78A in HEK293 cells compared with the stimulated p-Ser of RXRα WT by western blot analysis. The physical interaction of endogenous RXRα and GSK3β was found in HCT116 by co-immunoprecipitation assay. Since casein kinase 1α (CK1α) mediates GSK3β-induced p-Ser of protein, our result showed that CK1α inhibitor increased RXRα expression and suppressed β-catenin expression in HCT116 and in a dose-dependent manner ( Figure 9A-9C).

DISCUSSION
CRC is one of the frequently seen malignancies in the world. Kara et al. have analyzed a total of 54 CRC and normal colon tissue samples of 42 healthy controls using a high-throughput real-time PCR method and found that miR-27a-3p was significantly deregulated in CRC [21]. Large-scale microRNA expression profiling was performed in serum samples from 427 CRC patients and 276 healthy donors using Illumina small RNA sequencing and a diagnostic four-microRNA signature consisting of miR-23a-3p, miR-27a-3p, miR-142-5p and miR-376c-3p was established [11]. miR-27a-3p was an independent predictive factor for recurrence in clear cell renal cell carcinoma [15]. However, expression of miR-27a-3p was consistently down-regulated in hepatocellular carcinoma cell lines and tissue samples and suppressed tumor metastasis [13]. Expression of microRNA-27a was increased in CRC stem cells [17]. The miR-27acalreticulin axis affects drug-induced immunogenic cell death in human colorectal cancer cells [18]. Our data showed that expression of miR-27a-3p was up-regulated in CRC cell lines and tissues and closely associated with histological differentiation, clinical stage, distant metastasis and CRC patients' survival. miR-27a-3p expression was an independent prognostic factor for CRC patients. However, Bao et al. have reported microRNA-27a is a tumor suppressor in colorectal carcinogenesis and progression by targeting SGPP1 and Smad2 [19]. Our result was different from Bao's report, the reason includes: 1) we studied miR-27a-3p, Bao et al. studied microRNA-27a, whether microRNA-27a-3p and/or microRNA-27a-5p was studied is not clear. 2) we performed miR-27a-3p expression in large number of CRC with 100 samples, Bao et al. only used 41 CRC samples to detect microRNA-27a expression. This issue needs further study in larger samples of CRC.
As for the role of miR-27a-3p in malignant tumors, overexpression of miR-27a-3p promoted gastric cancer cell proliferation in vitro as well as tumor growth in vivo [12]. High expression of miR-27a enhances proliferation and invasion of CRC cells [22]. Consistent with other reports, our data showed that miR-27a-3p mimic suppressed apoptosis and promoted proliferation, migration, invasion of CRC cells in vitro and in vivo, whereas miR-27a-3p inhibitor suppressed CRC progression. The results suggest that miR-27a-3p plays an important role in promoting carcinogenesis of CRC.
In conclusion, our findings first demonstrate that miR-27a-3p is a prognostic and/or potential therapeutic biomarker for CRC patients and RXRα as miR-27a-3p targeting gene plays the critical role in activation of Wnt/ β-catenin pathway during CRC progression.

Clinical samples and patient information
100 pairs of paraffin-embedded archived CRC and adjacent non-tumor colorectal mucosal tissues (ANT) were collected from our Department between January 2000 and December 2013. Fifteen pairs of fresh CRC and ANT were collected in 2013. No patients had received chemotherapy and/or radiotherapy before operation. The histopathology of the disease was determined by two pathologists according to the criteria of the World Health Organization. Clinical staging was done according to Dukes staging. For the research purposes of these clinical materials, prior patient's consents and approval from the Institutional Research Ethics Committee were obtained. Detailed clinical information about these patients, including age, gender, clinical stage, histological differentiation, T classification, N classification, and distant metastasis status, is summarized in Table 1. Follow-up information was available for all patients.

Cell lines and small interfering RNA (siRNA) sequences
The human CRC cell lines SW480 and SW620 were maintained in Leibovitz's L-15 Medium (Invitrogen, Carlsbad, CA). HCT116 was grown in McCoy's 5A Medium (Invitrogen). LOVO and SW1116 were cultured in RPMI-1640 medium (Invitrogen). HT29 was maintained in Dulbecco's modified Eagle's medium (Invitrogen). The human colonic epithelial cell line NCM460 was cultured in RPMI-1640 medium. All lines were purchased from Cell bank of Chinese Academy of Science (Shanghai, China) and authenticated by their karyotypes, and detailed gene expression in 2015. All medium were supplemented with 10% (v/v) fetal bovine serum (Invitrogen), 1×antibiotic/antimycotic (100units/mL streptomycin, 100units/mL penicillin, and 0.25 mg/ mL amphotericin B). All cell lines were cultured in humidified incubator at 37°C with 5% CO 2 .

Analysis of cell cycle and apoptosis
5 × 10 5 cells were seeded in 6-well plates and incubated overnight until 50-60% confluence. The cells were transfected with 100nM miR-27a-3p mimic or inhibitor or RXRα-siRNA and harvested at 48h, washed in cold PBS, fixed with 80% ethanol for 8 h at 4°C, then stained with propidium iodide buffer (50mg/ml propidium iodide, 0.1% sodium citrate and 0.1% Triton X-100) for 3h at 4°C. Non-specific control miRNA mimic or inhibitor or scrambled-siRNA was used as the control group, respectively. 2× 10 4 cells were analysed for cell cycle and apoptosis using a Becton Dickinson FACScan (Becton Dickinson Immunocytometry Systems, San Jose, CA). The percentage of cells in each phase of the cell cycle and apoptotic cells was quantified using Cell Quest software. This experiment was performed in triplicate.

Colony formation assay
After 48h transfection with miR-27a-3p mimic or inhibitor or RXRα-siRNA, the cells at 4× 10 2 and 6× 10 2 per well were plated in 6-well plates and grown for 2 weeks, respectively. Then, the cells were washed twice with phosphate buffer saline (PBS), fixed with 4% paraformaldehyde and stained with 0.5% crystal violet for 15 min. The number of colonies in 10 random view fields was counted under a microscope and the average number of colonies was achieved. The experiment was triplicated independently.

Scratch wound migration assay
2× 10 5 SW480 cells were seeded in 6-well plates and grown to 60% confluency in complete medium. The cells were transfected with 100nM miR-27a-3p mimic. Non-specific control miRNA mimic was used as the control group. After 24h transfection, vertical scratches were then made using a 100μl plastic filter tip to create a 'wound' of approximately 100μm in a diameter. To eliminate dislodged cells, culture medium was removed and wells were washed with PBS. The average distance of migration cells was determined under an inverted microscope at 0, 12, 24 and 48h. The experiment was performed in triplicate.

Transwell migration and invasion assays
Migration and invasion assays were carried out in Transwell chambers containing polycarbonate filters (8μm pore size; Corning Incorporated, Life Sciences, NY, USA). After transfected with 100nM miR-27a-3p mimic or inhibitor or RXRα-siRNA for 48h, HCT116 cells and SW480 cells (migration/2×10 4 cells; invasion/2×10 5 cells) in a 500μl volume of serumfree medium were placed in the upper chambers and incubated at 37°C with 5% CO 2 for 24 hours, while a 200μl volume of medium containing 15%FBS was added to the lower chamber as chemoattractant. Cells were allowed to invade through the matrigel (BD Biosciences) or migrate for 24 hours at 37°C with 5% CO 2 . Following invasion or migration, cells were fixed with 4% formaldehyde and stained with 1% crystal violet. Cells on the upper surface of the filters were removed by wiping with a cotton swab. Cells counts were the mean number of cells per fields of view. Three independent experiments were performed and the data were presented as mean ± standard deviation (SD).
The quantity of miR-27a-3p in each CRC tissues relative to its paired ANT was calculated using the equation [RQ = 2 -ΔΔCT , ΔΔCT= (CTmiRNA-CTU6)T -(CTmiRNA-CTU6)N]. The expression level of miR-27a-3p was classified into low expression and high expression group compared with RQ ratio =RQ(CRC)/RQ(ANT). The geometric mean of housekeeping gene GAPDH was used to normalize the variability at mRNA expression levels. All experiments were performed in triplicate.

Immunohistochemistry staining
The working concentrations of primary antibody for the detection of RXRα (Santa Cruz Biotechnology, Santa Cruz, CA) was 1:200. Immunohistochemical staining was independently assessed by two observers who had no knowledge of the clinicopathologic data. The degree of RXRα immunostaining was scored as our previously reported [10].

Cloning of RXRα 3'UTR and RNA binding protein immuno-precipitation (RIP) assay
A RXRα 3'UTR was amplified from CRC cell line-HCT116 cDNAs and cloned into the XhoI/NotI www.impactjournals.com/oncotarget site of a psiCHECK-2 vector (Promega, Madison, WI). For mutagenesis of the miR-27a-3p binding site, a QuickChange site-directed Mutagenesis Kit (Agilent Technologies, Palo Alto, CA) was used according to the manufacturer's instructions. RIP experiments were performed using a Magna RIP™ RNA-Binding Protein Immunoprecipitation Kit (Millipore) according to the manufacturer's instructions. Briefly, cells were harvested and lysed in RIP Lysis Buffer. RNAs were immunoprecipitated with an antibody against RBP and protein A/G magnetic beads. The magnetic bead bound complexes were immobilized with a magnet and unbound materials were washed off. Then, RNAs were extracted and analyzed by quantitative real time-PCR. The specific primers for RXRα were followed: forward: 5′-CAG CTG CAT TCT CCC ATG AG-3′; reverse: 5′-GTT CAT AGG TGA GCT GAG CTG-3′. Antibody for RIP assays of AGO2 was purchased from Cell Signaling Technology (2897S).

Co-Immunoprecipitation and immunoblotting analysis
Cells lysates were incubated with 5ug antibody on a rotator overnight at 4°C. The protein-antibodyprotein A/G agarose complexes were prepared by adding protein A/G agarose beads (Invitrogen) for an hour at 4°C. After extensive washing with RIPA lysis buffer, the immnoprecipitated complexes were resuspended in reducing sample buffer and boiled for 10 minutes. After centrifugation to pellet the agarose beads, supernatants were subjected to SDS-PAGE and immunoblotting.

Luciferase reporter assays
To examine the effect of RXRα on β-catenin/TCF/ LEF signaling, cells were plated in 96-well plate at a density of 1×10 4 cells/well and incubated overnight in FBSsupplemented medium. Cells were transiently transfected with RXRα expression plasmid using Lipofectamine 2000. After 6 hours, cells were transfected with β-catenin-TCF luciferase reporter construct-TOPFlash (Upstate Biotechnology, Lake Placid, NY). Luciferase activity was measured with the Dual Luciferase reporter assay system (Promega Corporation, Madison, WI).

Xenograft tumor model
Female BALB/c-nude mice (4-5 weeks old and weighing 15-18g) were housed under pathogen-free conditions. SW480 and HCT116 cells were trypsinized, washed twice with serum-free medium and reconstituted in serum-free medium DMEM, mixed 1:1 with Matrigel (Becton-Dickinson) and then inoculated subcutaneously into the right flank of each nude mouse. A local miR-27a-3p antagomir or agomir treatment was initiated when the tumor was palpable at a volume of approximately 20 mm 3. The mice were randomly assigned into treatment and negative control groups (n=6 mice/group) and given intratumour injection with 2nM miR-27a-3p antagomir or agomir or non-specific control miRNA dissolved in 30μl PBS every 3 days. We modified antagomir or agomir with 2-O-methyl and conjugated cholesterol to the ends of the antagomir or agomir, which can retain full potency of the antagomir or agomir, confers substantial nuclease resistance, improves bio-distribution and facilitates entry into cells. The treatment time was 12 days. Tumor size was measured every 3 days, using a digital caliper, and the tumor volume was calculated according to the formula: tumor volume (mm 3 ) = length×width 2 ×0.5. At the end of the experiment, all mice were sacrificed and the total weights, tumor weights, and the tumor volumes were recorded. All the experiments were performed following the Guide for the Care and Use of Laboratory Animals (National Institutes of Health publication).
As for CRC metastatic animal model, after transient transfection for 24 hours, HT29 cells were trypsinized, washed twice with serum-free medium DMEM and reconstituted in serum-free medium DMEM and then introduced through tail-vein injection (10 6 cells/mouse,6 mice/group). Subsequently, miR-27a-3p agomir or negative control (10nmol per mouse) was injected into the lateral vein in the tails of the nude BALB/c mice once every 6 days. Mice were sacrificed at 39 days (after 6 times injection) and lung, liver, and kidney tissues were fixed in 4% neutral formaldehyde and performed with haematoxylin and eosin staining. The CRC metastatic tumors were calculated by grossly and histological observation.

Statistical analyses
Groups from cell culture and in vivo experiments were compared using an unpaired, two-tailed Student's tests and results were presented as mean ± SD. For CCK8 assay, comparison was done by univariate variance analysis (two-way ANOVA). Statistical analyses were performed using SPSS 16.0 statistical software. p<0.05 was considered to be statistically significant.