LSCC tissues and corresponding adjacent normal margin (ANM) tissues were from patients (n = 30) who underwent partial or total laryngectomy in the Department of Head and Neck Surgery of Harbin Medical University Cancer Hospital from August 2016 to May 2017. Patients included in this study had neither radiotherapy/chemotherapy history before operation nor clinical treatment before being diagnosed. The tissues were preserved in liquid nitrogen within 5 min after excision and maintained at −80°C for later use. The ethics committee of Harbin Medical University Cancer Hospital approved the experiments in this study, and written consent was acquired from each patient. This study was conducted in obedience to the Declaration of Helsinki, and a follow-up study was performed for 3 years.

An LSCC cell line TU-177 and a Hela derivative Hep-2 (from American Type Culture Collection, ATCC) and normal human oral keratinocytes (NHOKs, from Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences) were cultured in RPMI1640 (72400120, GIBCO, NY, United States) containing 10% fetal bovine serum (FBS, 16140071, GIBCO, NY, United States) at 37°C in a 5% CO₂ incubator.

pcDNA3.1 plasmid with LNC-PINT overexpression (pcDNA3.1-LINC-PINT), ZEB1 overexpression (pcDNA3.1-ZEB1) and EZH2 overexpression (pcDNA3.1-EZH2), short-hairpin RNA of LINC-PINT (sh-LINC-PINT) and EZH2 (sh-EZH2) and corresponding negative controls (pcDNA3.1-NC and sh-NC) were bought from Shanghai Genechem Co., Ltd. (Shanghai, China). Cells were cultured in 60 mm culture dishes for 24 h at a density of 3.0 × 10⁵ cells/dish. Afterward, the cells were cultured with 3 μg of the aforementioned plasmid or shRNA, Lipofectamine 2,000 reagent (11668019, Invitrogen, CA, United States), Opti-MEM I Reduced Serum Medium (31985062, GIBCO, NY, United States) and 8 ng/ml polybrene (TR-1003, Sigma-Aldrich, St. Louis, MO, United States) for 48 h.

The cells were accordingly assigned into Control group, pcDNA3.1-LINC-PINT group, sh-LINC-PINT group, pcDNA3.1-EZH2 group, sh-EZH2 group, pcDNA3.1-ZEB1 group and pcDNA3.1-LINC-PINT + pcDNA3.1-ZEB1 group.

Proliferation rate of cells were measured by a CCK-8 kit (Beyotime, Shanghai, China). Transfected cells (3,000 cells/well) were pipetted into a 96-well plate, and fresh culture medium was replaced every day. Then, the proliferation rate of the cells were detected by the CCK-8 kit every 24 h for 5 consecutive days. In short, 10 μL of CCK-8 solution was added into each well and cultured with the cells for 2 h at 37°C, after which the absorbance was measured at 450 nm using the microplate reader SpectraMax M5 (Molecular Devices).

Tissues or cells were lyzed in 1 ml of Trizol reagent (Thermo Fisher Scientific, MA, United States) and total RNAs were isolated following the manufacturer’s directions. Reverse transcription was conducted to obtain cDNA with the assistance of M-MLV reverse transcriptase and arbitrary primers. Reaction conditions for qRT-PCR were configured in line with the instructions of Premix Ex Taq™II kit (Takara, Dalian, China), and the ABI7500 quantitative PCR instrument (Applied Biosystems, Shanghai, China) was employed for RT-PCR analysis. GAPDH was used for normalization, and relative mRNA levels of genes were analyzed by 2^−ΔΔCt method. ΔΔCt = [Ct _{(target gene)}–Ct _{(internal reference)}]_{experimental group}–[Ct _{(target gene)}–Ct _{(internal reference)}]_{control group}. Names and sequences of primers used in this study were listed (Table 1).

Tissues and cells were lyzed to obtain proteins using lysis buffer, and the protein concentration was then quantified by a BCA kit (23,227, Thermo Fisher, United States). Subsequently, the proteins were diluted by 5 × loading buffer and separated in 12% separating gel using electrophoresis for 90 min. After being blocked by 1 × PBS containing 5% (w/v) skim milk powder at room temperature for 1 h, the proteins were incubated with primary antibodies, including anti-EZH2 antibody (1:500, ab186006, Cambridge, United Kingdom), anti-ZEB1 antibody (1:500, ab203829, Cambridge, United Kingdom), anti-AKT antibody (1:500, ab38449, abcam, Cambridge, United Kingdom), anti-p-AKT antibody (1:500, ab8805, abcam, Cambridge, United Kingdom), anti-mTOR antibody (1:500, ab32028, abcam, Cambridge, United Kingdom), anti-p-mTOR antibody (1:500, ab109268, abcam, Cambridge, United Kingdom), anti-RPS6KB1 antibody (S6K1, 1:500, ab32529, abcam, Cambridge, United Kingdom), and anti-p-RPS6KB1 antibody (p-S6K1, 1:500, ab59208, abcam, Cambridge, United Kingdom) at 4°C overnight. Thereafter, the proteins were rinsed and incubated with the secondary antibody (1:500, ab150077, abcam, Cambridge, United Kingdom) at room temperature for 1 h before being photographed on BioSpectrum Imaging System (UVP, United States).

A Cytoplasmic and Nuclear RNA Purification Kit (cat no.21000; Norgen Biotek, Thorold, ON, United States) was used for the purification of nuclear and cytoplasmic RNA as previously described [24]. Purification was based on spin column chromatography. In brief, after the culture medium was removed, cells were washed with PBS once and lyzed by 200 μL of ice-cold Lysis Buffer J on ice for 5 min. The lysate was then pipetted into a microcentrifuge tube and centrifuged at 14,000 rpm and 4°C for 10 min. The supernatant containing cytoplasmic RNA was transferred into another RNase-free microcentrifuge tube, and the pellet containing nuclear RNA was retained. The tube containing cytoplasmic RNA or nuclear RNA was added 200 μL or 400 μL of Buffer SK, respectively, and then the cytoplasmic or nuclear RNAs were mixed with Buffer SK by vortexing for 10 s. The mixture was then mixed with 200 μL of 98% ethanol by vortexing for 10 s. Thereafter, the mixture was applied onto a spin column and centrifuged at 1,000 rpm for 1 min, after which the flowthrough was discarded. Wash Solution A (400 μL) was added to the column for 1 min of centrifugation and then the flowthrough was discarded, which was repeated twice to wash column. Afterward, the column was centrifuged for 2 min and then dried. The column was placed into an Elution tube and 50 μL of Elution Buffer E was added to the column. The tube was centrifuged at 2,000 rpm for 2 min, followed by 1 min of centrifugation at 14,000 rpm. Finally, the purified RNA was collected and quantified by qRT-PCR.

Cells (1 × 10⁵) from the experimental and control groups were pipetted into 12-well plates and a scratch was made by a 10 μL pipette tip vertically when the cell confluence reached 100%. Exfoliated cells and cell debris were washed off by DPBS (14190250, GIBCO, NY, United States) three times, and then the adherent cells were cultured in fresh DMEM supplemented with 2% FBS. Then, the cells were observed by an Olympus inverted microscope at 0 and 24 h under the same field to assess the changes of the scratch. Migration rate = (gap between the scratch at 0 h–gap between the scratch at 24 h)/gap between the scratch at 0 h. The results were from three independent experiments.

Invasiveness and migratory capacity of Hep2 and TU-177 cells were measured by using a Transwell (Corning, NY, United States). After transfection, cells (3 × 104) in the experimental and control groups were suspended in culture medium containing 1% FBS and seeded onto an apical chamber. Then, 0.8 ml of culture medium with 10% FBS was added onto a basolateral chamber. The insert used for Transwell invasion assay was encased with matrigel at 37°C for 2 h. Twenty four hours later, cells on the surface of the apical membrane were wiped off, and the cells passing the insert were maintained in 4% paraformaldehyde for 30 min. After being fixed, the cells were dyed with 10% Giemsa and rinsed with PBS three times and observed using an inverted microscope. The number of invasive cells was counted at the magnification of × 200 from 5 random fields and then averaged. The number of cells invaded and migrated in the control group was defined as 1, and normalized fold changes of invasive and migratory cells in the experimental groups were accordingly calculated. The results were acquired from three independent experiments.

A RIP kit was purchased from Guangzhou Saicheng Biotechnology Co., Ltd. (KT102–01, Guangdong, China). Approximate 4 × 10⁷ cells were homogenized with lysis buffer (1 ml) and rotated in a 4°C refrigerator for 1–2 h. Afterward, the cells were centrifuged at 4°C and 14,000 rpm for 10 min, and 10 μL of the supernatant was collected and termed Input group. A total volume of 100 μL of the supernatant was incubated with bead-antibody (5 μg, ab191250, ab75974, ab108252, abcam, Cambridge, and United Kingdom) complexes at 4°C overnight, and then the supernatant was removed on a magnetic rack. Subsequently, 1 ml of RIP buffer was added to wash the complexes and then the supernatant was discarded. Thereafter, each group (including Input group) was incubated with 117 μL of RIP buffer, 15 μL of 10% SDS, and 18 μL of proteinase K at 65°C for 45 min before being centrifuged at 3,000 rpm for 5 min. The supernatant from IP group were added with 250 μL of RIP wash buffer and 400 μL of the homogenate of phenol, chloroform and isoamylol (125:24:1) and mixed by vortexing. After being maintained at room temperature for 5 min, the mixture was centrifuged at 12,000 rpm and room temperature for 15 min. Upper aqueous phase (400 μL) was collected and then mixed with 1.2 ml of absolute ethanol and 2 μL of glycogen for precipitation overnight at −20°C. After precipitation, the mixture was centrifuged at 12,000 rpm and 4°C for 30 min. The supernatant was discarded before the beads were rinsed with 75% ethanol. The beads were centrifuged again, and the supernatant was removed. Finally, the beads were dissolved in DEPC, and RNAs were analyzed by qRT-PCR.

Specific FISH was conducted using LINC-PINT probes. The 5′CY3 labeled LINC-PINT probes were synthesized by Shanghai Genechem Co., Ltd. (Shanghai, China). All operations were made according to the instruction of the FISH kit (F03401, Genechem, Shanghai, China). Briefly, cells on the slides were fixed with ice-cold 4% paraformaldehyde and degraded at 70°C. The cells were incubated with 30 μg/ml LINC-PINT probes and then counterstained with DAPI after the residual probes were washed off. Finally, the cells were visualized and photographed under a fluorescent microscope.

Cells were lyzed to extract proteins, and then the isolated proteins were incubated with anti-EZH2 antibody (ab186006, abcam, Cambridge, United Kingdom) and 20 μL of protein-A/G-agarose beads at 4°C overnight. Following incubation, the precipitation was washed with lysis buffer four times and suspended in 5 × SDS-PAGE loading buffer. The suspension was boiled for 5 min before Western blot analysis.

Statistical analysis was carried out by using SPSS 18.0 (IBM Corp., Armonk, NY, United States) and GraphPad Prism 6.0 (GraphPad Software Inc.). Results in this study were from three independent experiments unless otherwise stated and the data were presented as average ± standard deviation. T-test measured the differences between two groups, and comparisons among multiple groups were assessed using One-way analysis of variance. P < 0.05 had statistical significance.

The clinical data of 30 patients enrolled in this study are presented in Table 2. To disclose the clinical significance of LINC-PINT in LSCC, we measured the relative mRNA level of LINC-PINT in LSCC tissues (n = 30) and in the corresponding ANM tissues (n = 30). The result of qRT-PCR analysis suggested that the mRNA level of LINC-PINT was dramatically downregulated in the LSCC tissues compared with the ANM tissues (Figure 1A, p < 0.01), which was consistent with the previous study [10]. To validate the association between LINC-PINT expression and LSCC malignancies, the clinicopathologic features of patients, including tumor node metastasis (TNM) stage, differentiation ability and lymph mode metastasis, were analyzed. With the median LINC-PINT expression as the cutoff value, patients were stratified into the high LINC-PINT group (n = 15) and low LINC-PINT group (n = 15). As presented in Figures 1B–D, the patients with lower LINC-PINT expression had advanced TNM stage and poor ability of differentiation (P < 0.01). The incidence of lymph node metastasis was much higher in patients with lower LINC-PINT expression when compared with those with higher LINC-PINT expression (P < 0.01). While there was no significant association between the expression of LINC-PINT and clinical variables such as age and sex (Figures 1E,F, P > 0.05). Moreover, the prognostic analysis based on the follow up data indicated that the lower expression of LINC-PINT predicted poor prognosis of LSCC patients (Figure 1G, P < 0.01). All these manifested that downregulated LINC-PINT expression was associated with LSCC progression.

To determine the functional role of LINC-PINT in cancer development, a Hela derivative Hep-2 and an LSCC cell line TU-177 were selected. qRT-PCR analysis of LINC-PINT expression exhibited a great decrease in Hep-2 and TU-177 cells compared to that in NHOK cells (Figure 2A, P < 0.01), which was consistent with the detection in LSCC tissues. We next overexpressed or inhibited LINC-PINT in Hep-2 and TU-177 cells. qRT-PCR suggested satisfactory transfection efficiencies of the pcDNA3.1-LINC-PINT and sh-LINC-PINT, as evidenced by increased level of LINC-PINT in the pcDNA3.1-LINC-PINT group and declined expression of LINC-PINT in the sh-LINC-PINT group (Figure 2B, P < 0.01). Then, we evaluated the viability of Hep-2 and TU-177 cells by CCK-8 assay. The results indicated that LINC-PINT overexpression suppressed the viability of Hep-2 and TU-177 cells, and LINC-PINT inhibition enhanced the cell viability (Figure 2C, P < 0.01). We also detected the influence of LINC-PINT on the migration and invasion of Hep-2 and TU-177 cells through performing Scratch and Transwell assays. Cells transfected with sh-LINC-PINT seemed to be more invasive than those in the Control group and the migration ability were also increased in the sh-LINC-PINT group; overexpression of LINC-PINT appeared to suppress the migration ability and invasiveness of Hep-2 and TU-177 cells (Figures 2D,E; Supplementary Figures S1A,B, P < 0.01). Overall, these data supported that overexpression of LINC-PINT suppressed cancer development in vitro.

To further illuminate the underlying mechanism by which LINC-PINT regulates cancer development, we investigated the target genes of LINC-PINT. Initially, the content of LINC-PINT in the cytoplasm and nuclei of Hep2 and TU-177 cells was analyzed by qRT-PCR analysis. The results revealed that LINC-PINT mainly expressed in the nuclei of Hep-2 and TU-177 (Figure 3A, P < 0.01), which was further identified by the results of RNA-FISH in Hep2 cells (Figure 3B). To investigate the role of EZH2 in LINC-PINT regulating cancer development, we measured the impact of LINC-PINT on EZH2 expression. qRT-PCR and Western blotting found that EZH2 was lowly expressed in the pcDNA3.1-LINC-PINT group and highly expressed in the sh-LINC-PINT group, compared to the Control group (Figures 3C,D; Supplementary Figure S2, P < 0.05). In addition, anti-EZH2 antibody was incubated with the total RNAs from Hep-2 and TU-177 cells to conduct RIP assay, and then qRT-PCR analysis was applied to quantify LINC-PINT expression. The results denoted that incubation of anti-EZH2 antibody increased the expression of LINC-PINT in Hep-2 and TU-177 cells compared to the cells incubated with anti-IgG antibody (Figures 3E,F, P < 0.01), which confirmed that LINC-PINT specifically bound EZH2.

Next, we screened the downstream genes of EZH2 to explore which genes were regulated by EZH2 in the LSCC cell line TU-177 and Hep-2 carcinoma cells. Hep-2 and TU-177 cells were cultured with pcDNA3.1-EZH2 or sh-EZH2, and then the expression levels of CDKN1C, ACE2, Robo4, KLF2, P4HA1, DUXAP8, DKK-1, and ZEB1 were quantified by qRT-PCR and Western blotting. In cells transfected with pcDNA3.1-EZH2, the expression levels of CDKN1C, ACE2, and ZEB1 were increased, while the levels of DUXAP8, DKK-1, and KLF2 were downregulated; Hep2 and TU-177 cells from the sh-EZH2 group presented massive decreases in the levels of CDKN1C, ACE2, and ZEB1 as well as increases in the levels of DUXAP8, DKK-1, and KLF2 (Figures 3G,H, P < 0.01); no significant changes were noticed in the expression levels of Robo4 and P4HA1 either in the pcDNA3.1-EZH2 or in the sh-EZH2 groups. CDKN1C, ACE2, and ZEB1 were selected for Co-IP assay to address the possible interaction with EZH2. Anti-EZH2 antibody was incubated with the proteins extracted from Hep-2 and TU-177 cells, and then we quantified the expression levels of CDKN1C, ACE2, and ZEB1 in the protein-antibody complexes. As exhibited in Figure 3I, the relative levels of CDKN1C and ACE2 were hardly detected while ZEB1 was abundant in the protein-antibody complexes. Taken together, LINC-PINT suppressed EZH2 expression, and the enhancer of EZH2 bound the promoter of ZEB1 to regulate ZEB1 expression.

The expression of ZEB1 in LSCC tissues was measured to disclose the effect of ZEB1 on LSCC progression. Compared with the ANM tissues, the mRNA level of ZEB1 was upregulated in the LSCC tissues (Figure 4A, P < 0.01). The patients were divided into high ZEB1 group (n = 15) and low ZEB1 group (n = 15), with the median ZEB1 expression as the cutoff value. The prognostic analysis manifested that higher expression of ZEB1 implied dismal prognosis of LSCC patients (Figure 4B, P < 0.01).

Based on the aforementioned findings, we next revealed whether LINC-PINT could inhibit cancer cell propagation, invasion and migration via ZEB1. Hep-2 and TU-177 cells were, respectively, introduced with pcDNA-3.1-LINC-PINT and sh-LINC-PINT, and then the expression of ZEB1 was assessed by qRT-PCR and Western blot analyses. Gain- or loss-of function experiment manifested that the mRNA and protein expression levels of ZEB1 were downregulated in the pcDNA3.1-LINC-PINT group and upregulated in the sh-LINC-PINT group (Figures 4C,D; Supplementary Figure S3A, P < 0.01), suggesting that LINC-PINT expression was inversely associated with ZEB1 in Hep2 and TU-177 cells.

Then, Hep-2 and TU-177 cells were transfected with pcDNA3.1-LINC-PINT, pcDNA3.1-ZEB1 or co-transfected with pcDNA3.1-LINC-PINT and pcDNA3.1-ZEB1, and cell proliferation, migration and invasion abilities were detected. As expected, the propagation, and invasion and migration rates of Hep2 and TU-177 cells were enhanced in the pcDNA3.1-ZEB1 group (vs. the Control group), and the cell viability and invasion and migration abilities were also increased in the pcDNA3.1-LINC-PINT + pcDNA3.1-ZEB1 group compared to the pcDNA3.1-LINC-PINT group (Figures 4E–G; Supplementary Figures S3B,C, P < 0.01). These data supported that the upregulation of ZEB1 could counteract, in part, the inhibitory action of LINC-PINT on the malignant properties of Hep2 and TU-177 cells.

To elucidate the mechanism regarding LINC-PINT regulating cancer progression, we determined the expression levels of AKT, mTOR, RPS6KB1 (S6K1) and their phosphorylated proteins in Hep-2 and TU-177 cells transfected with pcDNA3.1-LINC-PINT and sh-LINC-PINT. As shown in Figure 5A; Supplementary Figures S4A,B, the phosphorylated levels of AKT, mTOR and S6K1 were downregulated in the pcDNA3.1-LINC-PINT group, and significant increases in the phosphorylated levels of AKT, mTOR, and S6K1 were found in the sh-LINC-PINT group, compared to the Control group (P < 0.01).