Overall, 92 PC samples and their corresponding adjacent tissues were selected from specimens collected from patients diagnosed at Anhui Provincial Hospital (Hefei, China) between June 2008 and June 2012. Detailed pathological and clinical data (including age, sex, tumour location, nerve invasion, degree of differentiation, histological type, depth of invasion, lymph node metastasis and TNM stage) were obtained from the medical records of each patient. The samples were from 56 male and 36 female patients aged 38–77 years (median, 54 years). Samples were included in the present study based on the 7th edition of the Union for International Cancer Control TNM staging system (14). Patients who had received radiotherapy or chemotherapy prior to surgery were not included. The specimens were fixed in 4% formalin at 37°C for 2 h and embedded in paraffin for pathological analysis and confirmation of the diagnosis. The clinical follow-up data of the patients were obtained from the PC database of Anhui Provincial Hospital. The present study was approved by the Human Research Ethics Committee of the Anhui Provincial Hospital and each patient provided informed consent.

Immunohistochemistry for GRHL2 was performed on each tissue sample. The tissue samples were dissected into 4-µm sections on silanised glass slides. Protein expression was detected with 2-step immunohistochemistry. In brief, deparaffinised and hydrated sections were treated with 0.3% hydrogen peroxide in methanol for 15 min at room temperature in order to block endogenous peroxidase activity, and washed in PBS (3 times for 3 min each), whereby antigen retrieval was conducted in citrate buffer (cat. no. P0081; Beyotime Institute of Biotechnology, Haimen, China; pH 6.0) for 10 min at 100°C. Following 3 more PBS washes (3 min each), the sections were stained with a GRHL2 monoclonal antibody (cat. no. ab86611, 1:200, Abcam, Cambridge, UK) for 2 h at 37°C, and washed again with PBS (3 times for 3 min each). Subsequently, the sections were incubated with horseradish peroxidase (HRP) universal IgG antibody polymer (cat. no. sc69786, 1:1,000, Santa Cruz Biotechnology, Inc., Dallas, TX, USA) for 15 min at 37°C, followed by 3 PBS washes (3 min each). Each section was treated with 50 µl diaminobenzadine working solution (DAB Horseradish Peroxidase Color Development kit; cat. no. P0202; Beyotime Institute of Biotechnology) at room temperature for 3–10 min, followed by a wash in PBS. All sections were counterstained with haematoxylin for 1–2 min at room temperature for the purpose of enabling the morphology of the tissue to be observed using a light microscope.

The expression of GRHL2 in the tumour samples was semi-quantitatively assessed using ImageJ 1.8.0 software (National Institutes of Health, Bethesda, MD, USA). The proportion of positive cells was graded as follows: 0, <1%; 1, 1–30%; 2, 30–70%; and 3, >70%. Intensity of staining was stated as none (score, 0), weak (score, 1), moderate (score, 2) or strong (score, 3). The immunoreactivity was calculated according to the intensity of staining and the percentage of positive cells. The 2 scores were multiplied with each other and the eventual immunostaining score was determined: A final score of 0, 1, 2 or 3 was considered as low expression, and a score of 4, 6 or 9 was classified as high expression. All specimens were evaluated separately by 2 pathologists who were blinded to the clinical data, and discrepancies were resolved by consensus.

PANC-1, BxPC-3 and HPDE6-C7 were purchased from the Cell Bank of Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). Cell lines were grown in Dulbecco's modified Eagle' medium (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc.) at 37°C in a humidified atmosphere containing 5% CO₂.

The cells were lysed in RIPA cell lysis buffer (cat. no. P0013B; Beyotime Institute of Biotechnology) with 1 mM phenylmethylsulfonyl fluoride at 4°C for 30 min, with vortexing every 10 min, followed by centrifugation at 13,800 × g for 10 min at 4°C. BCA kit (cat. no. P0009; Beyotime Institute of Biotechnology) was used to quantify the protein concentration. The supernatant protein solutions were diluted to 4 mg/ml and kept at −80°C until further use. A total of 50 µg of denatured protein for each sample was separated on a 10% SDS-PAGE gel and transferred onto a polyvinylidene fluoride membrane. The membranes were blocked with 5% non-fat milk in TBS/Tween-20 for 2 h at room temperature and blotted with primary antibodies of rabbit monoclonal anti-GRHL2 (cat. no. ab86611; 1:200; Abcam) and mouse monoclonal anti-β-actin (cat. no. TA811000; 1:400; OriGene Technologies, Inc., Beijing, China) overnight at 4°C. Following washing with TBS/Tween-20, the membranes were incubated with anti-rabbit or anti-mouse secondary antibodies, conjugated with HRP (catalog nos. A0208 and A0216; 1:1,000, Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) for 1 h at room temperature. Following 3 washes, the membrane was visualised with an enhanced chemiluminescence system (cat. no. P0018AM; Beyotime Institute of Biotechnology). The housekeeping control in this experiment was β-actin.

All statistical analyses were performed using the statistical package SPSS 17.0 (SPSS Inc., Chicago, IL, USA). The χ² test was used to analyse the immunohistochemistry results. Survival analysis was performed using Kaplan-Meier survival curves with the log-rank test. A multivariate Cox regression analysis was used to identify significant independent prognostic factors. All P-values were 2-sided and P<0.05 was considered to indicate statistically significant differences.

For the purpose of evaluating the levels of GRHL2 protein in PC tissue samples, immunohistochemistry was performed on 92 PC specimens and their matched adjacent healthy pancreatic tissues. Representative images of high and low expression of GRHL2 in PC and adjacent normal tissue samples are presented in Fig. 1. The GRHL2 staining was primarily observed in the cytoplasm of the tumour cells. High expression of GRHL2 was observed in the majority of the PC tissues (50/92 samples) and in a number of the non-cancerous pancreatic tissues (34/92 samples). The number of PC tissue samples with high GRHL2 expression was significantly higher than the number of non-cancerous tissue samples of the same phenotype (P=0.018; Table I).

The results of the western blotting were in agreement with those of the immunohistochemistry. As demonstrated in Fig. 2, the amount of GRHL2 protein in the PC cell lines (PANC-1 and BxPC-3) was higher compared with that from the normal pancreatic cell line (HPDE6-C7).

In order to evaluate the biological significance of GRHL2 in PC, the associations between PC tissue GRHL2 levels and clinicopathological parameters, including age, sex, tumour size, tumour location, nerve invasion, lymphatic metastasis, pathological grade, depth of invasion and TNM stage, were analyzed. As demonstrated in Table II, the increased GRHL2 expression was associated with higher tumour differentiation (P=0.018) and N classification (P=0.024). By contrast, no significant association was identified between the expression of GRHL2 and any of the other parameters tested. These findings imply that GRHL2 may serve an important role in the progression of PC.

A Kaplan-Meier survival curve was plotted in order to compare the OS time and the GRHL2 expression data (Fig. 3). Patients whose samples revealed high expression of GRHL2 presented with a worse prognosis than those with low GRHL2 expression (P<0.001; Fig. 3). Multivariate survival analysis further revealed the high GRHL2 levels in the tumour tissue to be an independent prognostic marker for poor OS time (P=0.001; Table III).