Immortalized normal esophageal epithelial cell line NE1 was obtained from Professor George Tsao's laboratory (Department of Anatomy, The University of Hong Kong) in 2006. Chinese ESCC cell lines [HKESC1(HK), EC109 and EC9706] and six Japanese ESCC cell lines [KYSE30(K30), KYSE140(K140), KYSE180(K180), KYSE410(K410), KYSE510(K510), and KYSE520(K520)] were kindly provided by Professor Srivastava (Department of Pathology, The University of Hong Kong). The human ESCC cell lines HK, EC18, EC109, EC9706, K30, K140, K180, K410, K510 and K520 were cultured in DMEM (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% fetal bovine serum (Invitrogen; Thermo Fisher Scientific, Inc.). The oesophageal epithelial cell line (NE1) was cultured in Keratinocyte-SFM & EpiLife (1:1; Invitrogen; Thermo Fisher Scientific, Inc.), supplemented with bovine pituitary extract (BPE) (35 ug/ml; Invitrogen; Thermo Fisher Scientific, Inc.). All cell lines were cultured at 37°C in a 5% CO₂ incubator.

A total of 60 matched fresh ESCC samples and normal oesophageal epithelium specimens were obtained via surgical resection at Linzhou People's Hospital (Henan, China) between January 2015 and June 2015 for RNA extraction. Additionally, an aggregate of 266 formalin-fixed paraffin-embedded (fixed in 4% paraformaldehyde at room temperature for 10 h, in 4-mm thick sections) ESCC tissues and the comparing normal oesophageal epithelia were obtained from Linzhou Cancer Hospital (Henan, China) between January 2002 and February 2005 for the tumor tissue microarray (TMA). All patients enrolled in this investigation did not receive preoperative treatment. The clinical attributes of all patients are presented in Table I. The accompanying end focuses (time to the date of death were assessed). The present study was approved by the Institutional Ethics Review Board of the First Associated Hospital (Zhengzhou University), and written informed consent form was obtained from each patient.

Briefly, a TMA segment was de-paraffinized (dimethylbenzene I, II, IIII, each for 5 min) and rehydrated (alcohol gradient: 100, 95, 75 and 50%, each for 5 min), and the endogenous peroxidase activity was blocked using 3% hydrogen peroxide for a period of 15 min. For antigen retrieval, the TMA slide was microwave-treated (100°C) in 10 mM citrate buffer (pH 6.0, Na₃C₆H₅O₇·2H₂O: 1.2054 g, C₆H₈O₇·H₂O: 0.189 g, dissolving in 1 liter of ddH₂O) for 10 min. Then Bovine serum albumin (5%, Invitrogen; Thermo Fisher Scientific, Inc.) was applied to block non-specific binding at temperature for 1 h. The primary antibody used was rabbit anti-GRHL1 (cat. no. NBP1-81321, Novus Biologicals, LLC, Littleton, CO, USA), which was diluted to a ratio of 2 ug/ml and incubated with the slides for overnight at 4°C. The nucleus was counterstained with the use of Meyer's haematoxylin (at room temperature for 20 sec). Representative photomicrographs were captured with a Vectra (PerkinElmer, Shanghai, China). Slides were observed under light microscopy at ×40 magnification. GRHL1 immunoreactivity was calculated via the scores for the level of GRHL1-positive cells 1: ≥5%, <25%; 2: ≥25, <50; 3: ≥50, <75%; and 4: ≥75% and the force of GRHL1-positive staining (negative, 0; weak, 1; moderate, 2; or strong, 3). Enlightening outcomes were observed in 266 sets of ESCC cases. However, invalid cases included lost cases as well as Non-specific staining.

Total RNA was separated from the cell lines with the use of TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.). The extricated RNA (2 µg) was reverse transcribed with the use of the Takara PrimeScript™RT reagent kit with gDNA Eraser FastQuant RT kit (Takara Bio, Inc., Otsu, Japan), and the cDNA productions were stored at −20°C until subsequent use. RT-qPCR was conducted on the Roche LightCycler 480 sequence detection system (Roche Diagnostics, Basel, Switzerland) in 10 µl responses containing 0.5 µl reverse transcription product, 5 µl SYBR^® Green SuperMix (Roche Diagnostics), 0.25 µl of each PCR forward primer and reverse primer, and 4 µl ddH₂O. The reactions were pre-incubated at 95°C for 10 min, followed by 45 cycles of denaturation at 95°C for 15 sec and expansion at 60°C for 1 min, at that point the temperature was inclined from 60°C to 95°C (via automatic program processing) for the purpose of obtaining a melting curve. The accompanying PCR primers were: GRHL1, forward, 5′-CAAACGGCCAGTGTTGGTTC-3′, and reverse, 5′-TGCTCATCATCGCTTTGGTCG-3′; 18s, forward, 5′-GTAACCCGTTGAACCCCATT-3′, and reverse, 5′-CCATCCAATCGGTAGTAGCG-3′. The gene expression level was expressed as 2-∆∆Cq, where ∆Cq=Cq (gene)-Cq (18s) (8). The expression level in control cells was considered as 1. The experiments were conducted in triplicate. 18s was used as the standardization control. Agarose gel electrophoresis was conducted for the analysis of the mRNA expression level of GRHL1. The accompanying PCR primers were the same as RT-qPCR. PCR amplifications were conducted with the use of the GoTaq Green Master mix kit (Promega Corporation, Madison, WI, USA) and PCR cycling conditions adhered to the following: 95°C for 5 min, followed by 35 cycles of 95°C for 30 sec, 60°C for 30 sec and 72°C for 30 sec, with extension at 72°C for 7 min (S1000 Thermal Cycler; Bio-Rad Laboratories, Inc., Hercules, CA, USA; http://www.bio-rad.com). The PCR items were examined by 1.5% agarose gel electrophoresis.

A lentiviral construct containing GRHL1 (GeneCopoeia, Inc., Rockville, MD, USA) was packaged with the use of a Lenti-Pac™ HIV Expression Packaging kit (GeneCopoeia, Inc.) in 293 cells. GRHL1-treatment lentivirus was employed to steadily transfect ESCC cells (EC109 and HKESC1), in order to construct the GRHL1-overexpressing cells. Empty vector-transfected cells were established as the controls. Short hairpin (sh)RNA against GRHL1 (GeneCopoeia, Inc.) was transfected into KYSE510 cells with the use of a Lenti-Pac HIV Expression Packaging kit in 293 cells, according to the manufacturer's protocol. The cells transfected with the mixed inhibitor (NC; Shanghai GenePharma Co., Ltd., Shanghai, China) were employed as the negative controls.

Protein was extracted from ESCC cells by ice-cold lysis buffer supplemented with protease inhibitor cocktail (Pierce; Thermo Fisher Scientific, Inc.) and quantified using the BCA Protein Assay kit (Beyotime Institute of Biotechnology, Haimen, China). Protein samples (30 µg) were separated on a 15% SDS-PAGE gel and then transferred onto PVDF membrane (Millipore, Billerica, MA, USA). Following blocking with 5% skimmed milk in Tris-buffered saline with 0.1% Tween-20 (TBST) at room temperature for 2 h, the membranes were incubated overnight at 4°C with rabbit anti-human GRHL1 monoclonal primary antibody (1:250, cat. no. NBP1-81321, Novus Biologicals, LLC, Littleton, CO, USA). Subsequently, The membrane was rinsed with TBST and incubated with anti-rabbit IgG (1:2,000; cat. no. CST-14708, Cell Signaling Technology, Inc., Danvers, MA, USA) antibody conjugated to horseradish peroxidase at room temperature for 2 h. Blots were visualized with enhanced chemiluminescence (GE Healthcare, Chicago, IL, USA). β-tubulin primary antibody (1:1,000; cat. no. CST-2146, Cell Signaling Technology, Inc.) was used as a loading control.

In order to investigate cell growth, 1×103 cells (GRHL1, Vec-EC109; GRHL1, Vec-HK; shGRHL1, shCtr-K510) were plated in 96-well plates and the cell development rate was examined using a Cell Counting Kit-8 kit (Dojindo Molecular Technologies, Inc., Kumamoto, Japan), according to the manufacturer's protocol. For the foci formation assay, 1×103 cells were plated in 6-well plates, and following two weeks of culture (37°C, 5% CO₂), colonies comprised of >50 cells stained with 1% crystal violet were counted.

Data are presented as the mean ± standard error of the mean and SPSS standard V.22.0 (IBM Corp., Armonk, NY, USA) was employed for statistical analysis. The unpaired student's t test or two-way ANOVA (post-hoc test, mean ± SEM) were used to investigate differences among two groups or a multi-group comparison. Survival analysis was carried out with the use of the Kaplan-Meier and log-rank tests and the associations between GRHL1 expression and clinicopathological factors were calculated by Pearson's χ2 test or Fisher's exact test. Univariate and multivariate Cox proportional hazard regression model were used to calculate the survival hazard risk. P<0.05 was considered to indicate a statistically significant difference.

In an attempt to identify the associations between the GRHL1 protein expression and the clinical factors of patients with ESCC, the present study analysed the protein expression of GRHL1 in an arrangement of 266 paraffin-embedded ESCC TMA using immunohistochemistry. Delegate images of GRHL1 immunohistochemical staining in clinical ordinary oesophageal epithelial tissues and ESCC tissues are presented in Fig. 2A-H. Positive staining for GRHL1 was primarily observed in the nucleus of typical oesophageal epithelial cells and ESCC cells. Using a staining index of 5 as the cut-off point in the 266 patients with ESCC inspection, 164 (61.7%) patients were categorised as exhibiting low GRHL1 expression, and 102 (38.3%) patients exhibited a high expression of GRHL1. As presented in Table I, low GRHL1 expression was significantly associated with the tumor invasion (P=0.008), clinical stage (P=0.004) and patient mortality (P<0.001). No significant associations were observed between GRHL1 expression and other clinical factors, including age, sex, differentiation or lymph node (LN) metastasis (P>0.05).

The association between GRHL1 expression and OS in ESCC was examined using Kaplan-Meier analysis as well as the log-rank test. As evident from Fig. 3, a low GRHL1 expression was associated with a significantly reduced OS [hazard ratio (HR), 2.07 95% confidence interval (CI), 1.491–2.881; P<0.001; Fig. 3A]. Additionally, stage III–IV was associated with a reduced OS (HR, 2.345; 95% CI, 1.727–3.186; P <0.001; Fig. 3B). In addition, the early stage (stage I–II, n=178) and the advanced stage subgroup (stage III–IV, n=88) of patients with a low GRHL1 expression had a significantly reduced OS (stage I–II: HR, 1.840; 95% CI, 1.229–2.756; P=0.0031; and stage III–IV: HR, 2.032; 95% CI, 1.241–3.325; P=0.0048) in comparison with those who had a high GRHL1 expression (Fig. 3C and D). These outcomes indicated that the low expression of GRHL1 is associated with a reduced prognosis in ESCC. Univariate and multivariate Cox regression analyses were performed in order to investigate the impact of various clinical factors, including GRHL1 expression, sex, age, differentiation, tumor invasion, LN metastasis and clinical stage. As evident from Table II, the univariate analyses demonstrated that low GRHL1 expression was associated with notably reduced OS (P<0.001), compared with high GRHL1 expression. Furthermore, in multivariate analysis, poor differentiation was also associated with a significantly reduced OS (P=0.049), compared with patients with moderate and well differentiation. In univariate analysis, T3 (Fibre layer invasion) was associated with reduced OS (P<0.001), compared with T1 (mucosae invasion). LN metastasis was associated with significantly reduced OS (P<0.001), compared with patients without LN metastasis. The advanced stage of disease was also associated with significantly reduced OS (P<0.001), compared with early stage. The results also revealed that sex and age had no significant effect on OS (P>0.05), with multivariate analysis confirming that the low expression levels of GRHL1 and differentiation degrees were independent prognostic factors associated with the OS in ESCC (P<0.05).

In order to investigate the tumor-suppressive ability of GRHL1, GRHL1 was transfected into the EC109 and HKESC1 ESCC cell lines (termed GRHL1-EC109 and GRHL1-HK cells, respectively). EC109 and HKESC1 cells transfected with an empty vector (Vec-EC109 and Vec-HK, respectively) were used as controls. Expression of the GRHL1 gene and protein in these transfectants were confirmed using western blotting (Fig. 4A). KYSE510 cells were transfected using shGRHL1 (K510-SH2 and K510-SH3) or shCtr (K510-ctr). Expression of the depleted GRHL1 gene and protein in these transfectants were analysed by western blotting (Fig. 4B). In comparison with the control cells, the in vitro experiments revealed that the high expression of GRHL1 could successfully suppress tumorigenic capacity in its transfected cells, which was evidenced by a significant decrease in foci formation frequency (P<0.01, ANOVA; Fig. 4C-E), together with the inhibition of cell development rate (P<0.001, ANOVA; Fig. 4F-H).