microRNA-212 inhibits nasopharyngeal carcinoma progression through targeting TPD52 and inhibiting EMT and Wnt/β-catenin pathway

Downregulation of miR-212 in nasopharyngeal carcinoma patient tissues and cell lines

The level of miR-212 was explored in NPC cell lines and patient tissues by qRT-PCR. The relative level of miR-212 normalized to U6 in NPC tissues was decreased compared with normal tissues (p<0.01, Figure 1A). Similarly, the level of miR-212 was also observed lower in C666-1 and 6-10B NPC cell lines relative to the normal NP69 cells (p<0.01, Figure 1B). Moreover, the tumor/node/metastasis (TNM) stage (p = 0.017) and distant metastasis (p = 0.022, Table 1) in NPC patients were associated with downregulation of miR-212, and NPC patients expressing lower levels of miR-212 had a decreased overall survival (p = 0.0436, Figure 1C). We also analyzed miR-212 expression in NPC and normal tissues in public available study cohorts in TCGA database and found low miR-212 in NPC tissues (p = 0.0036, Figure 1D). These results thus indicated that NPC is associated with lower level of miR-212.

Figure 1. Downregulation of miR-212 in NPC tissues and cell lines. (A) Low miR-212 expression was observed in NPC tissues. N=33 in each group. (B) Relative MiR-212 expression levels in C666-1, 6-10B and NP69 cell lines. N=4 in each group. (C) Low miR-212 expression was associated with worse NPC prognosis. (D) Low miR-212 expression in NPC was found in public available study cohorts in TCGA database (N=102). *p < 0.05, ** p <0.01.

Upregulation of miR-212 in NPC cells hindered cell migration, proliferation and invasion

To pinpoint the effects of miR-212 on NPC pathogenesis, miR-212 mimics or inhibitors were transfected into C666-1 cells. MiR-212 expression was observed to be regulated by miR-212 mimics or inhibitor, indicating that transfection was effective (p<0.01, Figure 2A). Functionally, miR-212 overexpression inhibited proliferation of C666-1 cells. In contrast, miR-212 downregulation promoted cell proliferation in NPC cells (p<0.05 or 0.01, Figure 2B). Moreover, miR-212 mimics inhibited migration and invasion of C666-1 cell while the miR-212 inhibitor promoted cell migration and invasion as shown by transwell assay (p<0.01, Figure 2C, 2D). Similarly, miR-212 overexpression inhibited cell migration whereas downregulation of miR-212 promoted C666-1 cell migration in the wound healing assay (p<0.01, Figure 2E). Thus, these data suggested that miR-212 possesses an inhibitory capacity in the context of NPC tumorigenesis.

Figure 2. MiR-212 overexpression inhibited migration and proliferation of NPC. NC represented C666-1 cells with the negative control sequence. (A) Effect of miR-212 mimics or inhibitor on miR-212 expression in C666-1 cells. N=3. (B) miR-212 mimics or inhibitor influenced cell proliferation in C666-1 cells as detected by MTT assay. N=3. (C, D) miR-212 mimics or inhibitor regulated the migration and invasion of C666-1 cells as detected by Transwell assay. Representative images were shown (Magnification, 40×). Bar graphs indicate mean±SD of 3 independent experiments. (E) Wound healing assay of C666-1 cells transfected with NC, miR-mimics and miR-inhibitor. *p <0.05, **p <0.01.

MiR-212 inhibited EMT and inactivated the Wnt/β-catenin pathway in NPC cells

To gain a deeper understanding of the regulatory functions of miR-212, we sought to determine how miR-212 regulates EMT and the Wnt/β-catenin pathway in NPC. Overexpression of miR-212 in C666-1 and 6-10B cells reduced protein levels of EMT indicators, including Vimentin and N-cadherin but increased the amount of E-cadherin protein (Figure 3A). Unsurprisingly, knockdown of miR-212 led to increased Vimentin and N-cadherin expression while reducing the expression of E-cadherin (Figure 3A).

Figure 3. MiR-212 inhibited EMT and inactivated the Wnt/β-catenin pathway in NPC cells. (A) The protein levels of Vimentin, N-cadherin and E-cadherin were regulated by miR-212 mimics or inhibitor in C666-1 cells. NC means C666-1 cells with empty sequence. n=3-4. Data were quantified and normalized to the intensity of β-actin (left). (B) The protein levels of Wnt/β-catenin and downstream Cyclin D1 were analyzed by western blot. Both total β-catenin and nuclear localized β-catenin were shown. Lamin-B1 was used as the loading control for nuclear proteins. The western blots (right) were the representative images of 3-4 experiments. Data were quantified and normalized to the intensity of β-actin (left). *p <0.05, **p <0.01.

Because the Wnt/β-catenin pathway is engaged in the pathogenesis of cancers, we proposed that the Wnt/β-catenin pathway may also influence NPC progression. In Figure 3B, the expression of Wnt/β-catenin and downstream Cyclin D1 were studied by western blot. We showed both total β-catenin and nuclear localized β-catenin because the translocation of β-catenin from the cell membrane or the cytoplasm to the nuclear is an indicator of the canonical Wnt signaling pathway activation. We also analyzed the protein level of Cyclin D1 since it is a downstream target gene of the Wnt/β-catenin signaling pathway. High levels of Cyclin D1 protein promotes cell proliferation. Our data indicated downregulation of miR-212 activated Wnt/β-catenin pathway which may promote the proliferation of NPC. We found overexpression of miR-212 inhibited the expression of Wnt1, total β-catenin, nuclear β-catenin and the downstream protein Cyclin D1; while downregulation of miR-212 promoted the expression of Wnt1, β-catenin and Cyclin D1. Overall, these data indicated that miR-212 suppress the progression of NPC by impeding both EMT and the Wnt/β-catenin pathway.

MiR-212 targeting TPD52

To further investigate the mechanism by which miR-212 inhibits the formation of NPC, we searched for additional targets of miR-212 using the TargetScan database. It was predicted that TPD52 containing the binding sites of miR-212 (Figure 4A). To verify this prediction, we designed a luciferase reporter assay to test the interaction of miR-212 in the 3’ UTR of TPD52. Overexpression of miR-212 reduced WT-TPD52 luciferase activity, but had no impact on Mut-TPD52 luciferase activity (p<0.01, Figure 4B). Furthermore, miR-212 expression was negatively associated with TPD52 expression in NPC tissues (p<0.0001, R²=0.7522; Figure 4C). To further verify this antagonistic relationship, the impact of miR-212 on TPD52 expression was measured in C666-1 and 6-10B cells following overexpression or knockdown of miR-212 (Figure 4D). We found that TPD52 was downregulated by miR-212 mimics and upregulated by miR-212 inhibitor (p<0.01). Overall, these data demonstrated that miR-212 directly bound to TPD52 mRNA and subsequently caused a decrease in TPD52 expression in NPC.

Figure 4. TPD52 was a direct target of miR-212. NC represented C666-1 cells with empty sequence. (A) Predicted binding site between TPD52 and miR-212. (B) Luciferase reporter assay with WT-TPD52 and Mut-TPD52. N=3. (C) miR-212 was negatively correlated with TPD52 in NPC tissues. N=33. (D) miR-212 mimics or inhibitor regulated TPD52 expression. The western blots were the representative images of 3 experiments. Data were quantified and normalized to the intensity of β-actin. **p <0.01.

MiR-212 inhibited NPC progression via targeting TPD52

Through qPCR, TPD52 expression was found elevated in NPC tissues (p<0.01, Figure 5A). The protein levels of TPD52 were also increased in NPC tissues from patients (p<0.01, Figure 5B). Moreover, upregulation of TPD52 correlated with the poor prognosis in NPC patients (p=0.0357, Figure 5C). High TPD52 expression in NPC was found in publicly available study cohorts in TCGA database (p<0.01, Figure 5D). We also detected the enhanced expression of TPD52 in C666-1 and 6-10B cell lines (p<0.05, Figure 5E). These results indicate that TPD52 may be involved in NPC progression. Next, miR-212 mimics suppressed the expression of TPD52, while miR-212 inhibitor enhanced the expression of TPD52 in C666-1 cells, further confirmed their interaction in NPC (p<0.05, Figure 5F).

Figure 5. Upregulation of TPD52 in NPC patient tissue and cell lines. (A) TPD52 expression was analyzed in NPC tissues via qRT-PCR. N=33. (B) TPD52 protein levels were analyzed by western blot and normalized to β-actin (N=10). (C) Low TPD52 expression predicted good prognosis in NPC patients. (D) High TPD52 expression in NPC was found in public available study cohorts in TCGA database (N=102). (E) High TPD52 protein levels in NPC cell lines (N=3). (F) TPD52 expression was measured in C666-1 cells transfected with miR-212 mimics and inhibitor (N=3). *p <0.05, **p <0.01.

Functionally, the previously observed inhibition of cell proliferation induced by miR-212 mimics was impaired by overexpression of TPD52 (Figure 6A). Similar results were observed for the transwell assay of cell migration (Figure 6B) and invasion (Figure 6C), as well as the wound healing assay in C666-1 cells (Figure 6D). siRNA knockdown of TPD52 inhibited EMT and Wnt/β-catenin pathway, as shown by western blot of marker proteins in C666-1 cells (Figure 6E, 6F). On the contrary, overexpression of TPD52 stimulated the expression of EMT and Wnt/β-catenin proteins. Overall, these results suggested that miR-212 suppressed cell viability and EMT in NPC cells through targeting TPD52.

Figure 6. MiR-212 inhibited NPC through targeting TPD52. NC represented cells with empty sequence. (A) The effect of TPD52 on cell proliferation. N=3. (B, C) The effect of TPD52 on cell migration and invasion as shown by Transwell assay. Representative images were shown (Magnification, 40×). Bar graphs indicate mean±SD of 3 independent experiments. (D) The effect of TPD52 on cell migration as shown by wound healing assay. (E, F) Representative western blot analysis and quantified bar-graphs of EMT marker proteins and Wnt/β-catenin pathway proteins under up- or downregulation of TPD52. *p <0.05, **p <0.01.

MiR-212 inhibited tumor growth in vivo

In order to validate the effects of miR-212 on TPD52 expression and NPC tumor growth in vivo, the transplanted tumor model with miR-212 up- and downregulation were established in nude mice. Tumor volumes were compared among different groups. We found that the volume of C666-1 tumors with overexpression of miR-212 were considerably lower than those of the control (Figure 7A), and tumor volumes with miR-212 inhibition were much higher than the control, indicating miR-212 overexpression inhibited tumor growth. In order to confirm whether miR-212 regulates the TPD52 and Wnt/β-catenin pathway in vivo, western blot was performed using tumor tissues from different groups (Figure 7B). The result showed that TPD52 and Wnt/β-catenin pathway were downregulated under miR-212 overexpression in tumors in vivo. Immunohistochemical staining of tumor tissues indicated inhibition of miR-212 enhanced proliferation and invasion of tumor cells, as revealed by Ki-67 and CD31 staining, respectively (Figure 7C). The results suggested that miR-212 downregulated TPD52 and Wnt/β-catenin pathway and inhibited tumor growth in vivo.

Figure 7. MiR-212 inhibited tumor growth in vivo. (A) Tumor volume growth curve. n=6. (B) The expression of TPD52 and Wnt/β-catenin were detected by western blot. n=6. (C) The proliferation and angiogenesis of tumor tissue section as detected by Ki-67 and CD31 immunostaining, respectively. Representative images were shown. * p < 0.05; ** p < 0.01.

Clinical tissues

All experimental tissues were sampled from 33 patients with NPC at Heze Municipal Hospital and Linyi People’s Hospital. NPC patients who provided consent forms did not receive other treatments except for surgery. The study protocols and consent forms were approved by The Ethics Review Board of Anqiu People's Hospital (approval number 0015009). The surgically removed NPC or normal tissue was collected from the operation rooms, cut into small pieces (2×2×2 mm) and stored in -80° C before testing. Tissue sample was homogenized in Bullet Blender tissue homogenizer before RNA and protein isolation. For survival assay, we divided the 33 patients into two groups: low miR-212 expression group (n=15, >medium value) and high miR-212 expression group (n=18,

Cell lines

Normal nasopharyngeal epithelial cells NP69 (BNCC339267), NPC cell lines 6-10B (BNCC342441) and C666-1 (BNCC337872) were obtained from BeNa Culture Collection (BNCC, Beijing, China). The cells were maintained at 37° C in a humidified incubator with 5% CO₂. RPMI-1640 medium containing 10% fetal bovine serum (FBS) was used for cell culture.

Transfection

miR-212 inhibitors, miR-212 mimics and the negative control miR sequence (NC), and TPD52 overexpression plasmid were purchased from ACON (Hangzhou, China). The miRNA negative control RNA sequence was based on common miRNA structure for use as negative experimental control in human and had been analyzed by BLAST against all human sequences and miRNA sequences in the current miRBase Database. siRNA against TPD52 was purchased from Santa Cruz (sc-45341). Transfections were performed in C666-1 cells using Lipofectamine 2000 (ThermoFisher).

Animal study

In vivo tumorigenesis assay was approved by the Animal Protection and Ethics Committee of the Anqiu People's Hospital. BALB/c nude mice (8-10 weeks) were from Beijing Weitonglihua company (Beijing, China). For Xenograft, control C666-1 cells and cells with miR-212 overexpression or inhibition (5 × 10⁶) were suspended in normal saline (200 μl) and administrated subcutaneously. Tumor volumes were measured by using the formula: Volume (mm3) = (L x W x H) / 2, where L represents tumor length, W as tumor width, and H as tumor height. All mice in this study were euthanized by cervical traction 20 days following tumor cell inoculation.

qRT-PCR

RNA sample was extracted from C666-1 cells using Trizol reagent (Takara Bio, Dalian, China). cDNA was made via reverse transcription using the miScriptII RT kit (Qiagen, Shanghai, China). The cDNA solution, SYBR Green PCR Master Mix, and primers (Beijing Unique Biotechnology, Beijing, China) were mixed prior to pipetting into a 96-well plate. qRT-PCR was conducted using a 7300 System (Applied Biosystems, USA) and the relative expression of miR-212 and TPD52 were assessed using the 2^-ΔΔCt method with U6 and GAPDH as controls, respectively. The primers were as follows: miR-212-F, CGCTAACAGTCTCCAGTC; miR-212-R, GTGCAGGGTCCGAGG T. U6-F, CGCAAGGATGACACGCAAATTC; U6-R, AACGCTTCACGAATTTGCGT; TPD52-F, AACAGAACATTGCCAAAGGGT G; TPD52-R, TGACTGAGCCAACAGACGAAA; GAPDH-F, AGCCACATCGCTCAGACAC; GAPDH-R, GCCCAATACGACCAAATCC.

Western blot

Cells were lysed in RIPA buffer (Beyotime, China) to isolate total proteins. The NE-PER nuclear extraction kit was used for nuclear proteins extraction according to the kit manual (ThermoFisher, 78833). Protein concentration was measured by BCA method and 15 μg was loaded and separated in a 10% SDS-PAGE. After electrophoresis, protein bands were transferred to PVDF membranes (Millipore, USA) and blocked with 5% non-fat milk in TBST. The membrane was probed with primary antibodies (4° C overnight) and secondary antibodies (RT 1 hour). Finally, the target proteins were visualized using the enhanced chemiluminescence (ECL) protein detection kit (Millipore, USA) and analyzed by Image-Pro Plus software version 6.0 (Media Cybernetics, USA). Primary antibodies of E-cadherin (1:1000, ab1416), TPD52 (1:1000, ab244523), N-cadherin (1:500, ab76011), Wnt1 (1:1000, ab15251), Vimentin (1:1000, ab92547), β-catenin (1:1000, ab32572), Cyclin D1 (1:500, ab16663), β-actin (1:3000, ab8226) and Lamin-B1 (1:1000, ab16048) were from abcam.

Immunohistochemistry

Paraffin sections of mice tissues were dewaxed in xylene and a series of descending ethanol solutions. 0.5% Triton X-100 was used for penetration and 50% goat serum was used for blocking. Sections were then incubated with Ki-67 or CD31 primary antibody overnight and developed with the DAB kit (Vector Laboratories, USA). Hematoxylin was used for counterstain and samples were dehydrated and mounted. The images were analyzed by Image-Pro Plus.

Methyl thiazolyl tetrazolium (MTT) assay

MTT cell proliferation and cytotoxicity assay kit (Solelybio #C0009, China) was used to compare the cell growth rate strictly according to the kit instruction. The absorbance at 490 nm was measured using a microplate reader (Molecular Devices).

Transwell

For invasion assay, the above chamber surface of the Transwell (8-μm pore size membranes) was pre-coated with Matrigel (BD Biosciences, USA). C666-1 cells (1×10⁴ cells) were seeded in the above chamber, and the bottom one was filled with medium containing 20% FBS. After a 24-h incubation, the invaded cells were stained with 0.1% crystal violet and enumerated under a light microscope at 400× magnification. The Matrigel was not used in the migration assay, and other steps were the same as the invasion assay.

Wound healing

Cells were plated in 6-well plates until they formed a confluent monolayer. Cells were then scratched using a 100 μL pipette tip. The scratch wounds were captured using microscopy immediately after scratching and 24-48 h later.

Dual luciferase assay

The wildtype 3’-UTR of TPD52 containing the predicted miR-212 binding site (WT-TPD52) was mutated to disrupt the binding site through Q5 site-directed mutagenesis kit (NEB). The generated mutant 3’-UTR (Mut-TPD52) or WT-TPD52 was inserted into the pmirGLO dual-luciferase vector (Qcbio Science and Technologies Co., Ltd, Shanghai, China). MiR-212 mimics and the above vector were co-transfected into C666-1 cells. Dual-Glo Luciferase Assay System (Qcbio S & T, Shanghai, China) was used to measure luciferase activity 48 hours after transfection.

Statistical analysis

SPSS 13.0 (SPSS Inc., USA) and Graphpad Prism 6 (GraphPad, USA) were used for data analysis. Chi-squared Test, one-way analysis of variance followed by TUKEY post-hoc test were performed for multiple groups comparison. Student's t-test was performed between two groups. Survival curves were compared by Kaplan-Meier analysis with log-rank test. p<0.05 was considered as statistical significance.