MiR-876-5p modulates head and neck squamous cell carcinoma metastasis and invasion by targeting vimentin

Cell culture, plasmids, and transfection

The CAL27 and HEK293T (293T) cell lines were obtained from the American Type Culture Collection (ATCC). Human HNSCC WSU-HN4 and WSU-HN6 (hereafter simplified to HN4 and HN6) cell lines were obtained from Dr. Silvio Gutkind of the NIH (Bethesda, MD) [24, 25]. CAL27 cells (authenticated at July 2017) and HEK293T cells (authenticated at March 2018) were tested by short tandem repeat testing done in the Fragment Analysis Facility, Tian Lin Biology and Technology Company (Shanghai, China). The two cell lines utilized in the experiment (WSU-HN4, WSH-HN6) showed distinct genotypes not matching with any in the databases. All the cell lines were recovered and maintained in Dulbecco’s modified Eagle’s medium containing 10% FBS at 37 °C in an atmosphere of 5% CO₂. The hsa-miR-876-5p mimics, the hsa-miR-876-5p inhibitors and corresponding negative controls were purchased from GenePharma (Shanghai, China). The RNA sequences for the primers mentioned above were as follows: miR-876-5p mimics, sense 5′-UGGAUUUCUUUGUGAAUCACCA-3′ and antisense 5′-GUGAUUCACAAAGAAAUCCAUU-3′; mimic control, sense 5′-UUCUCCGAACGUGUCACGUTT-3′ and antisense 5′-ACGUGACACGUUCGGAGAATT-3′; miR-876-5p inhibitors 5′-UGGUGAUUCACAAAGAAAUCCA-3′; and inhibitor control 5′-CAGUACUUUUGUGUAGUACAA-3′. The vimentin overexpression and negative control plasmids were obtained from GeneCopoeia (EX-vimentin-M02 and EX-EGFP-M02, respectively). The CAL27 and HN6 cells were transfected with special vimentin overexpression and negative control plasmids for 8 h using Lipofectamine 2000 (Invitrogen, USA) following the manufacturer’s instructions. The efficacy of transfection was tested by qRT-PCR and western blotting. For the manual alteration of miR-876-5p expression, cells (5 × 10⁵ per well in 6-well plates) were cultured to 80% confluence in complete growth medium; then, the medium was replaced with serum-free medium with miRNA mimics or inhibitors using Lipofectamine 2000 for 4–6 h according to the manufacturer’s instructions.

Patients and sample collection

Real-time reverse transcription (RT)-PCR

Total RNA was extracted using a TRIzol reagent (Invitrogen) according to the manufacturer’s protocol. For qRT-PCR detection of mature miR-876-5p expression, we purchased the Bulge-Loop™ miRNA qRT-PCR Primer Set and Control Primer Set (RiboBio, Guangzhou, China). RNA (2 μg) was converted into cDNA using the RevertAid First Strand cDNA Synthesis Kit (Thermo). QRT-PCR was accomplished using the FastStart Universal SYBR Green Master Mix (Rox) (Roche) in the ABI PRISM^® 7300 real-time PCR system (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s instructions. GADPH and U6 were used as endogenous controls. We used dissociation curves to monitor non-specific amplification. The relative expression level was computed using the 2^−ΔΔCt method. The sequences for sense and antisense primers are as follows: vimentin 5′-TGA GTA CCG GAG ACA GGT GCA G-3′ (sense) and 5′-TAGCAG CTT CAA CGG CAA AGT TC-3′ (antisense) and GAPDH 5′-GAA GGT GAA GGT CGG AGT C-3′ (sense) and 5′-GAG ATG GTG ATG GGA TTT C-3′ (antisense). For miRNA quantification, Bulge-loop™ miRNA qRT-PCR Primer Sets (one RT primer and a pair of qPCR primers for each set) specific for U6 and miR-876-5p were designed by RiboBio (Guangzhou, China).

Western blotting

Total protein was extracted from cells and lysed for 30 min using lysis buffer (Beyotime Shanghai, China). All proteins were resolved using sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) with 10% polyacrylamide gels and then transferred to polyvinylidene difluoride (PVDF) membranes (Millipore, MA), which were blocked with 5% BSA in phosphate-buffered saline (PBS) containing Tween 20 (PBS-T) for 2 h at room temperature. The blots were then probed with primary antibodies specific for vimentin (1:1000; Proteintech, 60330-1-Ig, China) or beta-actin (1:1000; Bioworld, I102, China) overnight at 4 °C, washed twice with TBST, and incubated with horseradish peroxidase-conjugated (HRP) secondary antibodies (Zhongshan Golden Bridge Bio, China) for 1 h at room temperature. Finally, the protein bands were detected using Immobilon Western Chemiluminescent HRP substrate (Millipore) and visualized using the ImageQuantLAS 4000 mini imaging system (General Electrics).

Invasion assays

Cell invasion ability was analyzed using Transwell filters (8 mm pore size; Millipore). Transwell inserts with 8-mm pores were coated with Matrigel (Matrigel:DMEM = 1:9; 50 µL per well; BD Bioscience, Franklin Lakes, NJ, USA). The cells (1 × 10⁵) were plated in 200 µL of serum-free medium in the upper chamber, while 500 µL of medium containing 10% FBS was used as the chemoattractant and placed into the lower chamber. After incubating the cells for 24 or 48 h at 37 °C, the non-invading cells remaining on the upper side of the filter were gently removed with cotton swabs. The invading cells on the lower membrane were fixed with 4% paraformaldehyde (PFA) for 30 min and stained with crystal violet for 5 min.

Scratch assays

Cells were cultured to 90% confluence in 6-well plates and then scratched in the central area with a sterile 10-µL pipette tip. Floating cells and debris were carefully removed with PBS, and the culture medium was replaced with a serum-free medium. Wounded cell migration was observed under a microscope, and images of the same wound area were captured over time.

Cell counting kit-8 (CCK-8) experiments

Cells were seeded in 96-well microplates at a density of 2 × 10³ cells per well. Cells were incubated in new medium containing 10% CCK-8 reaction solution (Selleckchem, Houston). After incubation for 2 h, the absorbance was measured on a spectrophotometer microplate reader (Multiskan MK3, Thermo) at a wavelength of 450 nm according to the manufacturer’s instructions. Three independent experiments were performed.

Immunofluorescence staining

Briefly, HN6 and CAL27 cells were grown on cover slips for 24 h, and the cells were fixed in 4% PFA and permeabilized in 1% Triton. After incubating overnight with primary antibody against vimentin (1:100, Proteintech, 60330-1-Ig, China), the cells were incubated with FITC-conjugated ATF4 Rabbit Polyclonal antibody (1:500, Proteintech, FITC-10835, China) and counterstained with DAPI (Beyotime Shanghai, C1002, China). Cells were subsequently viewed by fluorescence microscopy (ZEISS, Germany).

Immunohistochemistry

Tumor specimens were fixed in 10% neutral-buffered formalin for 24 h, followed by standard tissue processing and embedding. Sections were cut at 4 μm and dried overnight at 37 °C onto microscope slides. The tissue sections were stained with primary antibodies against vimentin (1:2000, Proteintech, 60330-1-Ig, China) and Ki67 (1:4000, Proteintech, 27309-1-AP, China) overnight following secondary antibody incubation for 30 min. All the sections were counterstained using hematoxylin and were dehydrated, cleared, and mounted before being examined using a microscope (DM4000B, Leica, Germany).

Vector construction and dual-luciferase reporter assays

For luciferase assays, the potential miR-876-5p binding site in the vimentin 3′-UTR was predicted by TargetScan (http://www.targetscan.org) to be at position 47–56 from the vimentin stop codon site. All vimentin 3′-UTR sequences with the wild-type or mutant seed region were synthesized and cloned into the Pezx-FR02 vector (Genecopoeia, USA) downstream from the luciferase stop codon; the new vectors were designated vimentin-WT or vimentin-MUT, respectively. 293T cells were cotransfected with 50 nM miR-876-5p mimics or negative control and 1 µg of vimentin-WT or vimentin-MUT using Lipofectamine 2000 (Invitrogen, USA). Cells were harvested at 48 h after transfection, and luciferase activities were analyzed by the Luc-Pair Duo-Luciferase Assay Kit 2.0 (Genecopoeia, USA).

Study in vivo

All animal studies were approved and supervised by the Animal Care and Use Committee of Nanjing Medical University, and all procedures were performed in accordance with the institutional animal welfare guidelines of Nanjing Medical University. All mice were maintained in specific pathogen-free (SPF) animal facilities with a 12 h day/12 h night cycle. Male BALB/c-nu mice (5 weeks old) were obtained from the Animal Research Center of Nanjing Medical University. To evaluate the effect of miR-876-5p on tumorigenesis and metastasis, 1 × 10⁶ CAL27 cells (0.2 mL) were suspended in Matrigel (BD) and were injected under the mice dorsa. When the tumors reached 3.0–5.0 mm in diameter (6 weeks old), the mice were randomly assigned into two groups (n = 7 each). Agomir-876-5p or negative control (NC) (RiboBio, Guangzhou, China) was directly injected into the implanted tumor at the dose of 5 nmol (in 50 µL of PBS) per mouse every 3 days for six times. Tumor diameter (mm) was measured every 2 days for 3 weeks, and tumor volume (V) was monitored by measuring the length (L) and width (W) with Vernier calipers and calculated with the formula V = (L × W²)/2. The tumor specimens were fixed in 4% PFA solution for 12–24 h and paraffin embedded for further analysis. For hepatic dissemination assays, HN6 cells (1 × 10⁶/mL) transfected with agomir-876-5p or NC (RiboBio, Guangzhou, China) were suspended in 200 µL of PBS for each mouse (3 per group). The cells were injected into nude mice through the lateral tail vein. After 3 weeks, the mice were killed, and the liver tissue was dissected, fixed with 4% PFA solution and embedded in paraffin. All sections were stained in hematoxylin and eosin (H and E) to analyze the tumor cells, and the number of micrometastases on the liver surface was calculated.

Statistical analysis

SPSS18.0 software was used for statistical analysis. All data were presented as mean ± standard deviation (SD) of three independent experiments and evaluated using the Student’s t-test. One-way analysis of variance (ANOVA) was used for comparison, and P < 0.05 was considered statistically significant (*P < 0.05, **P < 0.01, and ***P < 0.001).

The expression of miR-876-5p and vimentin in HNSCC cell lines and HNSCC tissues

To evaluate the impact of miR-876-5p on HNSCC metastasis and progression, we performed quantitative PCR analysis to compare the miR-876-5p expression in metastatic HNSCC and non-metastatic HNSCC. In 20 cases of non-metastatic HNSCC tissues and HNSCC tissues with lymph node metastasis (Table 1), our results revealed that miR-876-5p expression was significantly lower in HNSCC tissues with lymph node metastasis than that in HNSCC without metastatic primary tumors (Fig. 1a). We then searched for potential miR-876-5p targets using TargetScan7.1 (http://www.targetscan.org) and found that the 3′-UTR of vimentin contains a potential miR-876-5p binding site (Fig. 1b). It has been reported that vimentin is usually expressed in normal mesenchymal cells and is critical for malignant cell invasiveness [26]. In HNSCC samples, we found that the vimentin mRNA expression levels were markedly upregulated in metastatic HNSCC tissues compared with non-metastatic HNSCC tissues, implying the complementary relationship between miR-876-5p and vimentin (Fig. 1c). Next, we examined the relative levels of vimentin and miR-876-5p in HN4, HN6 and CAL27 cells. High vimentin expression levels were detected in HN6 cells, but much lower vimentin levels were detected in HN4 and CAL27 cells (Fig. 1d). However, high miR-876-5p transcript levels were verified in HN4 and CAL27 cells, while much lower miR-876-5p levels were found in HN6 cells (Fig. 1e). Therefore, miR-876-5p expression levels may be negatively associated with vimentin levels in these HNSCC cells. Then, we performed transwell assays to evaluate the invasive ability of HNSCC cells. The results revealed that HN6 had the greatest invasiveness, HN4 had the worst, and CAL27 was between these two (Fig. 1f). Thus, we speculated that as a mesenchymal marker, vimentin may play a crucial role in HNSCC cell invasiveness and miR-876-5p may target vimentin to regulate HNSCC cell invasion ability.

MiR-876-5p significantly inhibits cell migration and invasion in HNSCC cell lines

To evaluate the effect of miR-876-5p on invasion and migration in HNSCC, we transfected HN6 and CAL27 cells with miR-876-5p mimics or inhibitors. Transwell and scratch assays were utilized to verify the alteration of cell invasion and migration. Following transfection, increased miR-876-5p expression in HNSCC cells was confirmed by quantitative RT-PCR (Fig. 2a). In the miR-876-5p mimics group, miR-876-5p was overexpressed, while the invasion and migration activity of HNSCC was decreased markedly (Fig. 2b, c). On the other hand, compared with negative control cells, cells with miR-876-5p knocked down presented improved cell invasion and migration (Fig. 2b, c). Our data indicated that miR-876-5p inhibits the metastatic potential of HNSCC cells.

Moreover, to determine the alteration of cell proliferation, CCK-8 assays and flow cytometry were performed in HNSCC cells. Compared with the negative control, the miR-876-5p mimics or inhibitors did not substantially affect HNSCC cell proliferation (Additional file 1: Figure S1A, B).

Vimentin is a direct target of miR-876-5p in HNSCC

To further examine the potential negative regulatory effect of miR-876-5p on vimentin, we transiently transfected HN6 and CAL27 cells with the miR-876-5p mimics or negative control. Vimentin mRNA expression in miR-876-5p mimic-transfected HNSCC cells was only 27% and 21% of that in cells transfected with negative control, respectively (Fig. 3a). Next, vimentin protein expression was quantified by western blot analysis after miR-876-5p transfection. The vimentin expression level was decreased in the miR-876-5p mimics group. In contrast, after cells were transfected with miR-876-5p inhibitors, vimentin protein expression was upregulated (Fig. 3b). Moreover, we investigated vimentin expression in HNSCC cells by immunofluorescence; As expected, vimentin expression was deregulated in the miR-876-5p mimics group and was upregulated after transfection with the miR-876-5p inhibitors (Fig. 3c). Therefore, our data suggested that miR-876-5p may limit the metastatic potential of HNSCC cells by inhibiting vimentin.

To investigate whether miR-876-5p binds the 3′-UTR of target mRNA, we generated a luciferase reporter wild-type vector that contained the vimentin 3′-UTR with the putative miR-876-5p binding sites. Correspondingly, we constructed a mutant reporter vector that contained the vimentin 3′-UTR with a mutation at the putative miR-876-5p binding site. Co-transfection experiments revealed that miR-876-5p downregulated the luciferase activity of the wild-type vector by 50%, but this decrease was not present when the mutant vector was transfected (Fig. 3d). Taken together, these results indicate that vimentin is a direct, downstream target of miR-876-5p.

Vimentin overexpression promotes HNSCC cell migration and invasion

HN6 and CAL27 cells were transfected with special vimentin-overexpressing plasmids using Lipofectamine 2000. After transfection, the mRNA and protein expression level were obviously increased according to the qRT-PCR and western blot analysis (Fig. 4a, b). By using scratch and transwell assays, we evaluated whether vimentin upregulated the migratory and invasive ability of HNSCC cells. As expected, vimentin overexpression significantly induced HNSCC cell invasion abilities (Fig. 4c). According to the results in Fig. 4d, HN6 and CAL27 cell migratory abilities were increased significantly in the vimentin-overexpressing group. These results revealed that vimentin significantly regulates HNSCC cell invasion and migration.

Vimentin is a critical mediator of miR-876-5p in HNSCC cells

To further investigate whether miR-876-5p targets vimentin to regulate HNSCC cell invasion and migration, we divided into four groups: miR-876-5p-NC co-transfection with vimentin-NC, miR-876-5p mimics co-transfection with vimentin overexpression, miR-876-5p-NC together with vimentin overexpression, and miR-876-5p mimics together with vimentin-NC. As shown in Fig. 5a, transfection of vimentin plus miR-876-5p mimics significantly attenuated vimentin expression. Transwell assays revealed that cell invasiveness was higher in the over expressed vimentin group and lower in the miR-876 mimics group than that in the control group. No significant difference in the number of invasive cells per field was observed between the miR-876-5p mimics with the overexpression vimentin group and control group (Fig. 5b, c). Similarly, in scratch assays, vimentin increased HNSCC cell migratory ability, whereas miR-876-5p mimics blocked this change (Fig. 5d, e). Hence, our findings demonstrated that miR-876-5p regulates HNSCC cell invasion and migration via targeting vimentin.

MiR-876-5p inhibits vimentin expression and suppresses the distal metastasis of HNSCC cells in vivo

Since miR-876-5p substantially suppressed HNSCC cell migration and invasion in vitro, we further investigated the biological role of miR-876-5p in vivo. After incubating with agomir-876-5p, 2 × 10⁶ CAL27 cells were injected into the right dorsal subcutaneous tissue of nude mice. Two weeks after inoculation, agomir-876-5p was injected into the tumor tissues once every 3 days (5 nmol/mice). After 3 weeks of treatment, all mice were sacrificed, and the tumors extracted from the mice were presented in Fig. 6a, b. We found that the volume of the tumors formed by miR-876-5p-overexpressing CAL27 cells was not significantly different from that of the tumors formed by NC-expressing cells. IHC staining showed that the expression of the metastasis-related protein vimentin was upregulated, but proliferation-related protein Ki-67 [27] was not decreased in the tissue samples from the miR-876-5p-transfected nude mice. This data demonstrated that miR-876-5p has no significant effect on tumor proliferation in vivo (Fig. 6c).

To further determine the effect of miR-876-5p overexpression on HNSCC cell metastasis in vivo, we performed the distal dissemination assays using a nude mouse model. BALB/c nude mice were injected with agomir-876-5p or NC transfected cells via the tail vein (group 2, n = 3). At 24 days post injection, all mice were sacrificed. For the distal dissemination assay, the liver tissues were removed and subjected to histological examination. As expected, miR-876-5p substantially reduced the number of metastatic HNSCC cell clumps on the surface of the liver (Fig. 6d). These data suggest that miR-876-5p suppresses HNSCC cell metastasis in vivo.