A total of 30 patients (age range, 47–75 years; mean age, 61.31±7.0 years; female, 6.67%; male, 93.33%) diagnosed with LSCC and cervical lymph node metastasis at the Second Affiliated Hospital of Xi'an Jiaotong University (Xi'an, China) between January 2015 and December 2016 were enrolled in the present study. All patients provided written informed consent, and the present study was approved by the Ethics Committee of the Second Affiliated Hospital of Xi'an Jiaotong University. Samples were collected from the patients who underwent resection (partial laryngectomy or total laryngectomy), apart from those whose pathology types were not squamous cell carcinoma. Patients who only underwent biopsy without resection were excluded from the present study. Patients who received radiation or chemotherapy before surgery were also excluded. Pairs of laryngeal carcinoma tissues, adjacent non-tumor tissues (normal tissue was obtained 1.0 cm from tumor edge) and metastatic cervical lymph nodes were either fixed in 10% formalin at room temperature for 12–24 h or frozen in liquid nitrogen at −196°C after surgical resection till use.

The human LSCC TU-686 (Procell Life Science & Technology Co., Ltd.), AMC-HN-8 (BeNa Culture Collection; Beijing Beina Chunglian Biotechnology Research Institute), TU-177 (Jennio Biotech Co., Ltd.) and 293T (Procell Life Science & Technology Co., Ltd.) cell lines were cultured in DMEM (Thermo Fisher Scientific, Inc.) supplemented with 10% FBS (HyClone; Cytiva) and 50 U/ml penicillin-streptomycin at 37°C in a humidified atmosphere of 5% CO₂. Furthermore, the aforementioned cell lines were treated with different concentrations (0, 1, 2.5, 5 and 10 µM) of 5-Aza-2′-deoxycytidine (5-Aza-dC; cat. no. S3196; Selleck Chemicals) for 24, 48 and 72 h at 37°C. Different concentrations of 5-Aza-dC were added to DMEM for treatments.

To generate lentiviruses that overexpress and target short hairpin RNA (shRNA) expression of HNF1A-AS1, a full-length cDNA of HNF1A-AS1 and oligonucleotides that encode HNF1A-AS1 shRNAs [shRNA-non-specific control (NC), shRNA-853, shRNA-1525 and shRNA-2048; Table SI] which were designed by Shanghai GenePharma Co., Ltd. and named according to the catalogue number were cloned into the pLVX-mCMV-ZsGreen-PGK-Puro vector or the pLVX-shRNA2-Puro vector (Biowit Technologies, Ltd.), respectively. All vectors were verified by Sanger sequencing (19). The lentiviral vectors were transfected into sub-confluent 293T cells together with packaging plasmid and envelope plasmid (Biowit Technologies, Ltd.) using Lipofectamine 2000 reagent (Invitrogen; Thermo Fisher Scientific, Inc.) to produce lentiviral particles. The lentiviruses in the medium were collected 48 h later and were concentrated by ultracentrifugation at 448 × g for 10 min at room temperature. The target cells (cultured in 6-cm plate) were infected by removing the old culture medium and replacing it with 0.5 ml diluted viral supernatant (MOI, 20). A total of 0.5 ml complete medium with Polybrene was added to each well. The plates were placed in a 37°C incubator with 5% CO₂. Fresh media were added at 24 h after infection. Following another incubation for 48 h, puromycin (1 µg/ml) selection of stably transduced cells was started. Then, GFP-positive cells were sorted by flow cytometry. Cells were suspended in PBS at 1×10⁷ cells/ml and filtered prior to sorting by FACSAria II (BD Biosciences). FACSDiva software (version 6.1.3; BD Biosciences) was installed on the FACSAria II to operate the instrument and analyze the measurements.

RT-qPCR was used to detect the expression levels of HNF1A-AS1 (Homo sapiens; Table SI) in the clinical specimens and cultured cell lines (cat. no. 9108/9109, RNAiso Plus, total RNA extraction reagent for RNA extraction; cat. no. RR036A, PrimeScript™ RT Master Mix, Perfect Real Time for reverse transcription; cat. no. RR820A, TB Green™ Premix Ex Taq™ II, Tli RNaseH Plus for RT-PCR; Takara Bio, Inc.). Reverse transcription was performed at 37°C for 15 min followed by 85°C for 5 sec and 4°C hold. The RT-qPCR thermocycling conditions were as follows: Initial denaturation at 95°C for 30 sec; 40 cycles of denaturation at 95°C for 5 sec, and annealing and elongation at 60°C for 30 sec; final extension 72°C for 15 min (20,21). β-actin was used as the reference gene (forward, 5′-AGCGAGCATCCCCCAAAGTT-3′; reverse, 5′-GGGCACGAAGGCTCATCATT-3′). The relative expression levels of HNF1A-AS1 in each group (fold change compared with the control group) was calculated using the following formula: RQ=2^−ΔΔCq (22). Each reaction was performed in triplicate.

The expression levels of E-cadherin (cat. no. ab1416; mouse; dilution, 1:500; Abcam), N-cadherin (cat. no. ab76057; rabbit; dilution, 1:500; Abcam), Vimentin (cat. no. ab92547; rabbit; dilution, 1:300; Abcam), Snail (cat. no. ab53519; goat; dilution, 1:100; Abcam) and Slug (Snail2; cat. no. 12129-1-AP; rabbit; dilution, 1:100; ProteinTech Group, Inc.) in clinical specimens and cultured cell lines were detected by IHC. Tissues were fixed in 10% neutral-buffered formalin at room temperature for 24 h and embedded in paraffin. Sections (4-µm thick) were deparaffinized in four changes of xylene for 5 min each. Sections were hydrated in 100, 90 and 80% ethanol for 5 min each. Subsequently, sections were rinsed in water. Antigens were retrieved with citrate buffer (cat. no. C9999; Sigma-Aldrich; Merck KGaA) and sections were microwaved for 15 min at 95°C. Sections were blocked with 5% BSA (cat. no. A3059; Sigma-Aldrich; Merck KGaA) in PBS containing 0.1% Tween-20 (cat. no. 85114; Thermo Fisher Scientific, Inc.) for 1 h at room temperature. Sections were incubated with the aforementioned primary antibodies at 4°C overnight. Sections were incubated with the secondary HRP antibody (dilution, 1:1,000; cat. nos. P0449, K4001 and K4003; Dako; Agilent Technologies, Inc.) for 1 h at room temperature. Images were captured under a light microscope (IX51; Olympus Corporation). Immunoreactivity was scored based on the number of positive cells and staining intensity, using Image-Pro Plus 6.0 software (Media Cybernetics, Inc.). Immunohistochemical scores (IHS) were determined by multiplying the quantity and staining intensity scores as follows: i) The quantity was rated on a scale of 0–4: 0, no staining; 1, <10% of cells stained; 2, 10–50% of cells stained; 3, 51–80% of cells stained; and 4, 81–100% of cells stained; and ii) staining intensity was rated on a scale of 0–3: 0, negative; 1, weak; 2, moderate; and 3, strong. Theoretically, the scores could range between 0 and 12. An IHS of 9–12 was considered strong immunoreactivity, 5–8 was considered moderate, 1–4 was considered weak, and 0 was considered negative (23). Immunofluorescence staining was performed as previously described (24).

Cultured cells were grown on cover slips or chamber slides. Culture media were removed and cells were rinsed with PBS. The cells were fixed with 4% formaldehyde for 15 min at room temperature. Sections were blocked with 5% BSA (cat. no. A3059; Sigma-Aldrich; Merck KGaA) in PBS containing 0.1% Triton™ X-100 (cat. no. 85111; Thermo Fisher Scientific, Inc.) for 1 h at room temperature. Cells were incubated with the primary antibodies at the aforementioned concentrations overnight at 4°C in the dark. Sections were incubated with the secondary Cy5-conjugated antibody (dilution, 1:500; cat. nos. BA1031 and BA1032; Boster Biological Technology) for 1 h at room temperature in the dark. Sections were mounted with Prolong Gold medium-DAPI (cat. no. P36931; Invitrogen; Thermo Fisher Scientific, Inc.). Images were captured under a fluorescence microscope (IX51; Olympus Corporation).

The proteins were extracted using a RIPA buffer kit (cat. no. P0013B; Beyotime Institute of Biotechnology). Protein concentration was determined using a BCA protein concentration determination kit (cat. no. P0010; Beyotime Institute of Biotechnology). Protein (20 µg/lane) was loaded onto an SDS-PAGE precast gel (8–16%; cat. no. P0058B; Beyotime Institute of Biotechnology). The proteins were transferred to a nitrocellulose membrane (cat. no. 88018; Thermo Fisher Scientific, Inc.). Membranes were blocked with 5% BSA (cat. no. A3059; Sigma-Aldrich; Merck KGaA) in TBS with 0.1% Tween-20 (cat. no. 85114; Thermo Fisher Scientific, Inc.) for 1 h at room temperature. The proteins were assessed by incubation with the specific primary antibodies, including E-cadherin, Snail1, Vimentin and N-cadherin (all diluted 1:1,000), as well as Slug (cat. no. 9585; rabbit; dilution, 1:800; Cell Signaling Technology, Inc.), CyclinD1 (cat. no. 55506; rabbit; dilution, 1:1,000; Cell Signaling Technology, Inc.), proliferating cell nuclear antigen (PCNA; cat. no. 13110; rabbit; dilution, 1:1,000; Cell Signaling Technology, Inc.) and GADPH (cat. no. 10494-1-AP; rabbit; dilution, 1:1,000; ProteinTech Group, Inc.) overnight at 4°C. Sections were incubated with the secondary HRP antibodies (dilution, 1:5,000; cat. no. BA1051, BA1060 and BA1054; Boster Biological Technology) for 1 h at room temperature. The membranes were visualized using a ChemiDoc MP system (Bio-Rad Laboratories, Inc.) with ECL substrate (EMD Millipore). Densitometric semi-quantification of the bands was performed using ImageJ software (v. 1.52n; National Institutes of Health).

Cell proliferation capacities were evaluated using a Cell Counting Kit-8 assay (Beyotime Institute of Biotechnology) according to the manufacturer's protocol. The results were plotted as mean ± SE of three separate experiments for each experimental condition.

The cell cycle analysis was performed using a Cell Cycle Staining kit [cat. no. CCS012; PI staining; Hangzhou Multi Sciences (Lianke) Biotech Co., Ltd.], and analyzed by flow cytometry (FACScan^®; BD Biosciences) with CellQuest software (version 5.2.1; BD Biosciences). The ratios of cells in the G₀/G₁, S and G₂/M phases were calculated and compared with those in the control groups.

The cell migration ability was measured by scratch assay analysis with 100% confluence after starving cells in media containing 0.1% FBS overnight. A 20-µl pipette tip was used to create the scratch. Then, the cells were cultured in the medium with a reduced concentration (1%) of serum to control for the influence of proliferation and avoid apoptosis for 0 (baseline), 24 and 48 h, and then the images of the same view were captured (light microscope; IX51; Olympus Corporation). ImageJ software was used to measure the wound area at different time points. The width of the wound divided by the width of the background field was used to calculate changes over time. The cell invasion ability was assessed using 6.5-mm Transwell chambers (Corning Inc.) Matrigel was thawed for 24 h in ice at 4°C in a refrigerator. A total of 100 µl Matrigel (0.4 mg/ml) was loaded into the upper chamber and incubated for 24 h at 4°C in a CO₂ incubator. After 24 h, 750 µl DMEM complete medium with 10% FBS was added into the lower chamber. In the upper chamber, 200 µl cell suspension (2×10⁵ cells/ml in only DMEM) was added onto the Matrigel-coated cell culture insert and incubated for 24 h. After 24 h, the medium in the upper chamber was removed, followed by washing with PBS. Cells were fixed with 4% formaldehyde for 30 min at room temperature and stained with 0.4% crystal violet for 10 min at room temperature. Non-invaded cells were removed with cotton swabs. Invasive cells were counted in four random microscopic fields (light microscope; IX51; Olympus Corporation). Each assay was performed in triplicate.

The UCSC genome browser public database (http://genome.ucsc.edu) and Methprimer 1.0 (http://www.urogene.org/methprimer/) tools were used to identify the CpG islands of HNF1A-AS1 (2 kb upstream of the transcriptional start site) (25). The prediction of CpG islands was performed using the following parameters: >100 bp window, percentage of C plus G >50%, and observed/expected CpG ratio >0.6. Since the number of patients with LSCC in The Cancer Genome Atlas (https://portal.gdc.cancer.gov/) database was not sufficient, 530 patients with head and neck squamous cell carcinoma included in the TCGA HNSC cohort were analyzed retrospectively (26).

Total DNA was extracted from cell lines and tissue specimens using the MiniBEST Genomic DNA Extraction kit (cat. no. 9765; Takara Bio, Inc.). The EZ DNA Methylation-Gold kit (Zymo Research Corp.) was used for converting bisulfate for subsequent methylation analysis. The primers used for MSP and BSP were complementary to the promoter regions of HNF1A-AS1 (Table SI). The bisulphate-modified DNA was used as template for amplification in MSP detection (cat. no. R110A; Takara Bio, Inc.; methylated: 98°C for 10 sec, 61.3°C for 5 sec and 72°C for 30 sec, 40 cycles; unmethylated: 98°C for 10 sec, 58.1°C for 5 sec and 72°C for 30 sec, 40 cycles), and the PCR products were subjected to 1.5% agarose (cat. no. AG006; Canvax Biotech) gel electrophoresis (100 V; 40 min). Tris-Acetate-EDTA buffer was used in the preparation and running of the gel, and 10 µl GelRed (cat. no. 41003-T; Biotium, Inc.) was added to 100 ml agarose gel. Subsequently, 6 µl sample (5 µl and 1 µl 6X DNA loading buffer; cat. no. R0611; Thermo Fisher Scientific, Inc.) per lane was loaded on an agarose gel. The Gel Doc EZ imaging system (Bio-Rad Laboratories, Inc.) came equipped with Image Lab™ software version 3.0, with auto image capture and auto analysis, and was used for semi-quantification. Additionally, the modified DNA was amplified, purified, subcloned into pMD19-T vector (Takara Bio, Inc.) and transformed into E. coli (DH5-alpha; Biowit Technologies, Ltd.), cultured in Luria broth medium (LB; cat. no. 10855001; Thermo Fisher Scientific, Inc.) with 50 µg/ml ampicillin at 37°C with 0.04% CO₂, according to the manufacturer's protocol. Finally, five isolated colonies that were grown on LB agar plates (Thermo Fisher Scientific, Inc.) containing ampicillin with X-gal (0.2 mg/ml)/IPTG (1 mM) were picked, and underwent sequencing and analysis using an ABI 3730 DNA Sequencer (Applied Biosystems; Thermo Fisher Scientific, Inc.) for BSP detection.

Briefly, 8-week-old male Bal/Bc nude mice (n=44; weight, 22.95±0.41 g) were supplied by the Institute of Zoology, Xi'an Jiaotong University Health Science Center (Xi'an, China). The animals were fed and raised in an ultraclean specific-pathogen-free laminar flow rack with a constant temperature (20–26°C), humidity (40–50%) and 12/12-h light/dark cycle. Food and fresh water were accessible at all times. All animal experiments were performed according to the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (27) and were authorized by the Medical Ethics Committee of Xi'an Jiaotong University. Animal experiments were performed in May 2018.

For in vivo tumorigenicity experiments, DMEM without FBS and Matrigel (cat. no. 356234; BD Biosciences) were mixed 1:1 as resuspension solution. The cells (TU-686/shRNA-2048, TU-177/HNF1A-AS1 and control cells;100 µl; 1×10⁶) were subcutaneously injected into the right flank of each mouse (five mice per group). The mice were monitored for weight, respiration, ability to ambulate, taking food, drinking, tumor size, ulceration, infection, and necrosis by a specialized technician at the animal facility. The tumor volume of each mouse was determined by measuring two of its dimensions and calculated using the following formula: Tumor volume=length × width²/2. The humane endpoints included a rapid weight loss of 15% within a few days and a tumor diameter >1.5 cm (subcutaneous xenografts) in any single dimension. Body weights of mice are shown in Fig. S1A and B. Tumor-bearing mice were euthanized using CO₂ (20% of the volume of the chamber per min). The time interval between injection and euthanasia was 43 days.

For in vivo tumor cervical lymph node metastasis, the present study used an orthotopic xenograft model of head and neck cancers as described previously (28,29). The indicated cells (TU-686/shRNA-2048, TU-177/HNF1A-AS1 and control cells; 30 µl; 2×10⁵) (28,29) were injected submucosally into the tongue of nude mice (6 mice per group) (30). The resuspension solution was the same as for flank injection. Tumor-bearing mice were euthanized by CO₂ (20% of the volume of the chamber per min). The time interval between injection and euthanasia was 21 days. Finally, submucosal tongue tumors and cervical lymph nodes were surgically excised, weighed and imaged. Pathological examinations of cervical lymph nodes were performed to confirm metastasis.

Statistical analysis was performed using SPSS v19.0 (IBM Corp.) and GraphPad Prism 5.0 (GraphPad Software, Inc.). All data are presented as the mean ± SD. In vitro experiments were performed in triplicate. An unpaired two-tailed Student's t-test, one-way ANOVA with Tukey's post hoc test and the χ² test were used to analyze the data. P<0.05 was considered to indicate a statistically significant difference.

The results of RT-qPCR demonstrated that the expression levels of HNF1A-AS1 were significantly downregulated in LSCC tissues and metastatic lymph node samples compared with in the adjacent non-tumor tissues (P<0.05; Fig. 1A). The results of HE staining are shown in Fig. 1B. It was demonstrated that the downregulation of HNF1A-AS1 may serve an important role in the development and progression of LSCC.

To further ascertain whether the downregulation of HNF1A-AS1 expression was caused by promoter hypermethylation in primary LSCC, the methylation frequency in the promoter region of HNF1A-AS1 in HNSCC tissues and metastatic cervical lymph nodes was examined and was significantly higher than that in the corresponding normal tissues according to the TCGA database (Fig. 1C).

Furthermore, the methylation status of HNF1A-AS1 in 30 paired tissues in patients was analyzed using BSP, CpG methylation of HNF1A-AS1 promoter was detected in all primary LSCC tissues (88.1%) and metastatic cervical lymph nodes (95.2%), and the results revealed significantly lower methylation in the matched normal adjacent tissues (54.7%; Fig. 1D and E). Similar to the BSP results, apparent methylation of HNF1A-AS1 in the promoter region was also detected to be higher in primary LSCC tissues and metastatic cervical lymph nodes than normal adjacent tissues by MSP (Fig. 1F). These results implied that hypermethylation of CpG islands in the promoter region was the major regulatory mechanism of HNF1A-AS1 downregulation in LSCC.

Compared with HNF1A-AS1 expression in AMC-HN-8 cells, its expression was high in the TU-686 cell line, but was decreased in the TU-177 cell line as demonstrated by RT-qPCR (P<0.01; Fig. S2A). Therefore, TU-686 and TU-177 cell lines were selected as research targets for further functional studies. To downregulate the expression levels of HNF1A-AS1 endogenously in laryngeal cells, small shRNA-853, shRNA-1525 and shRNA-2048 were separately transfected into TU-686 cells and the interference effects were confirmed by RT-qPCR analysis. At 48 h post-transfection, the expression levels of HNF1A-AS1 were significantly decreased by shRNA-1525 (50%; P<0.01) and shRNA-2048 (75%; P<0.0001) compared with scrambled shRNA-NC (Fig. S2B). Furthermore, HNF1A-AS1 expression in TU-177 cells was upregulated 10 times following lentiviral vector transfection (pLVX-HNF1A-AS1 vs. pLVX-control; Fig. S2C). Considering the remarkable interference and overexpression effects of HNF1A-AS1, three stable cell lines (TU-686/shRNA-1525, TU-686/shRNA-2048 and TU-177/HNF1A-AS1) were used for subsequent experiments.

When compared with the negative control group, knockdown of HNF1A-AS1 resulted in a significant increase in cell proliferation in TU-686 cells at 48 and 72 h. Conversely, overexpression of HNF1A-AS1 decreased the cell proliferation in TU-177 cells at 24, 48 and 72 h (Fig. 2A). Additionally, the downregulation of HNF1A-AS1 promoted TU-686 cell transition from G₀/G₁ to S phase (Fig. 2B and C), but HNF1A-AS1 overexpression arrested cell cycle at G₀/G₁ phase (Fig. 2B and C). Furthermore, knockdown of HNF1A-AS1 in TU-686 cells significantly promoted the expression of cell cycle-related protein cyclin D1 and PCNA; however, these were inhibited by overexpression of HNF1A-AS1 in TU-177 cells (Fig. 2D and E). Overall, overexpression of HNF1A-AS1 could inhibit the proliferation of LSCC cells and this was partially accompanied by G₀/G₁ arrest.

The cell migratory abilities were investigated by scratch assay analysis and the results revealed that HNF1A-AS1 downregulation significantly enhanced the migratory abilities of LSCC cells (Fig. 3A and B). In addition, a Transwell invasion assay demonstrated that decreased expression of HNF1A-AS1 following transfection with TU-686/shRNA-2048 significantly enhanced the invasive capacity of TU-686 cells (105.1±0.4 vs. 94.8±1.2; P<0.01; Fig. 3C and D), whereas HNF1A-AS1 overexpression inhibited the invasive abilities of TU-177 cells (44.9±0.7 vs. 63.4±0.3; P<0.001; Fig. 3C and D). The expression levels of EMT-related markers were assessed as EMT serves a key role in tumor invasion and metastasis. The results of western blotting and immunofluorescence assays demonstrated that knockdown of HNF1A-AS1 notably downregulated E-cadherin expression and upregulated Snail1, Slug, Vimentin and N-cadherin expression, and vice versa when HNF1A-AS1 was overexpressed (Fig. 4). These data suggested that HNF1A-AS1 contributed to the invasion and metastasis of laryngeal cancer cells partly by affecting the EMT process.

Xenograft tumor (subcutaneous) growth models revealed that tumor sizes were markedly increased following injection with TU-686/shRNA-2048 cells, while injection with TU-177/HNF1A-AS1 cells was associated with a reduction compared with the control group (Fig. 5A-C). The cervical lymph node metastasis model (submucosal injection in tongue) gave similar results to the xenograft subcutaneous model. Tumor weights were increased for injection with TU-686/shRNA-2048 cells, while they were decreased for injection with TU-177/HNF1A-AS1 cells (Fig. 5D-F). This indicated that HNF1A-AS1 could promote laryngeal tumor growth either by subcutaneous injection or by submucosal injection into the tongue of nude mice. Additionally, the nude mice injected with TU-686/shRNA-2048 cells submucosally into the tongue had more metastatic cervical lymph nodes compared with those injected with TU-686/sh-NC cells, while pathological examination confirmed the opposite effects for injection of TU-177/HNF1A-AS1 cells (Fig. 5G and H), demonstrating the inhibitory effect of HNF1A-AS1 on tumor lymph node metastasis. Furthermore, compared with that in the control group, E-cadherin expression was significantly downregulated, while Snail1, Slug, Vimentin and N-cadherin expression was significantly upregulated in tumors injected subcutaneously using TU-686/shRNA-2048 cells in nude mice according to western blotting and IHC. By contrast, upregulation of HNF1A-AS1 was associated with the opposite effects in vivo (Fig. 6). Collectively, these findings demonstrated that HNF1A-AS1 expression inhibited tumorigenic, as well as metastatic, abilities in a xenograft tumor model of nude mice via EMT.

To explore the inactivation mechanisms of HNF1A-AS1 in LSCC, the sequence of HNF1A-AS1 was analyzed and it was revealed that the promoter region of HNF1A-AS1 was rich in CpG dinucleotides. The MethPrimer program was used to predict the distribution of CpG islands of HNF1A-AS1, and one CpG island (114 bp; −792 to −905 bp) was predicted within the promoter region of HNF1A-AS1 (Fig. 7A).

To clarify whether downregulation of HNF1A-AS1 was regulated by methylation in the CpG islands, TU-686 and TU-177 cells were treated with DNA methyltransferase inhibitor 5-Aza-dC, and the results demonstrated that the expression levels of HNF1A-AS1 were significantly increased in TU-686 and TU-177 cells following treatment with 5-Aza-dC at a dose of 5 and 10 µM for 48 h, revealing the crucial role of abnormal methylation in the inactivation of HNF1A-AS1 in LSCC cell lines (Fig. 7B).

Subsequently, a 148 bp segment in this region (−796 to −943 bp) that includes 7 CpG sites and spans the core promoter was analyzed using BSP (Fig. 7C). The hypermethylated CpG sites of TU-686 and TU-177 cells were 100 and 85.7%. Following treatment with 5 and 10 µM 5-Aza-dC, the hypermethylated CpG sites of TU-686 cells were decreased to 74.3 and 45.7%, respectively, and the ones of TU-177 cells were decreased to 71.4 and 31.4%, respectively (Fig. 7C). Similar to the BSP results, apparent methylation of HNF1A-AS1 in the promoter region was also detected in TU-686 and TU-177 cells by MSP, in which the methylation status was partly reversed following 5-Aza-dC treatment. This revealed the importance of promoter methylation in the downregulation of HNF1A-AS1 expression (Fig. 7D).

The results demonstrated that the migration and invasion abilities of TU-686 and TU-177 cells were significantly decreased after demethylation of HNF1A-AS1 by 5-Aza-dC treatment (Fig. 8A-D). In addition, western blotting indicated that demethylation of the CpG island in the promoter of HNF1A-AS1 significantly upregulated E-cadherin expression and downregulated Snail1, Slug, Vimentin and N-cadherin expression (Fig. 8E and F). These data suggested that low HNF1A-AS1 expression which contributed to the invasion and metastasis of laryngeal cancer cells by affecting the EMT process may be the result of hypermethylation of HNF1A-AS1 in its promoter region.