Human colorectal carcinoma cell lines (HCT116, CCL-247™; Caco2, HTB-37™; - SW480, CCL-228™; HT29, HTB-38™; DLD1, CCL-221™; and COLO205, CCL-222™) were obtained from the American Type Culture Collection (Manassas, VA, USA). The DKO1 (CVCL-9798) cell line was purchased from the ExPASy (Cellosaurus, Lausanne, Switzerland). All cell lines, excluding HCT116 cells, were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, Thermo Fisher Scientific, Inc.) and 1% penicillin/streptomycin, whereas the HCT116 cells were cultured in McCoy’s 5a Medium (WelGENE, Inc., Daegu, Korea) supplemented with 10% FBS and 1% penicillin/streptomycin at 37°C in a humidified atmosphere with 5% CO₂.

FOXA1 cDNA was subcloned into the pcDNA6-myc vector (Invitrogen, Thermo Fisher Scientific, Inc.). The FOXA1 construction was verified by sequencing. FOXA1 small interfering (si)RNA (GGAGGAGAGAUAAG UUAUA-dTdT) and scrambled siRNA (AllStars Negative Control siRNA, cat. no. 1027281) were purchased from Bioneer (Daejeon, Korea) and Qiagen GmbH (Hilden, Germany), respectively. To transfect siRNA, the SW480 and HCT116 cells were seeded into 6-well plates at a density of 3×10⁵ cells/well at 37°C and were 40-60% confluent at the time of transfection. For FOXA1 knockdown, 20 µM of FOXA1 siRNA (FS) was transfected with 5 µl Lipofectamine™ RNAiMAX reagent (Invitrogen; Thermo Fisher Scientific, Inc.). To overexpress FOXA1, 0.5 µg of FOXO1-pcDNA6-myc vector (FV) was transfected with 5 µl Lipofectamine™ 2000 reagent (Invitrogen, Thermo Fisher Scientific, Inc.). Scrambled siRNA (SS) and empty-pcDNA6-myc vector (EV) were used as a negative control, respectively. Following incubation for 24 h at 37°C, identification of the expression of FOXA1 was performed by western blotting and the transfected cells were used in the following experiments.

The transfected HCT116 and DLD1 cells were seeded into a 96-well plate at a density of 1×10⁴ cells/well and incubated for 24 h at 37°C. Cell viability was determined using a water-soluble tetrazolium salt (DoGen, Daeillab, Seoul, Korea). Following application, the absorbance at 450 nm was measured on a microplate reader (Infinite M200; Tecan, Mannedorf, Switzerland). Each experiment was performed in triplicate wells and was repeated at least three times.

The DLD1 and HCT116 cells at a density of 5×10⁵/well were seeded into a 6-well plate and incubated for 24 h at 37°C prior to FOXA1 transfection. The transfected cells were collected and resuspended in 100 µl of binding buffer (BD Biosciences, San Diego, CA, USA) for 20 min at 4°C. The cells were incubated with 7-amino-actinomycin D (BD Biosciences) and Annexin V-APC (BD Biosciences) for 20 min at room temperature. To analyze the number of apoptotic cells, a FACSCalibur flow cytometer (BD Biosciences) and WinMDI version 2.9 (The Scripps Research Institute, San Diego, CA, USA) were used.

The DLD1 and HCT116 cells at a density of 5×10⁵/well were seeded into a 6-well plate and transfected with FS and FV for 24 h at 37°C. The transfected cells were fixed in ice-cold 70% ethanol for 1 h to determine the cell cycle distribution and then washed twice with PBS. The cells were incubated in 100 µl of 10 µg/ml ribonuclease A (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) for 20 min at 37°C. Following this, 100 µl of 50 µg/ml propidium iodide (Sigma-Aldrich; Merck KGaA) was added for 30 min at room temperature in the dark. Cell cycle analysis was performed using a FACSCalibur flow cytometer (BD Biosciences) and WinMDI version 2.9 (The Scripps Research Institute).

The transfected cells were lysed in RIPA buffer containing Halt™ Protease inhibitor and Halt™ Phosphatase inhibitor cocktail (Thermo Fisher Scientific, Inc.) for 30 min in an ice bath. The protein concentration of the lysate from transfected cells was measured using the BCA™ protein assay (Thermo Fisher Scientific, Inc.). The proteins (20 µg per lane) were separated using 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto a polyvinylidene fluoride (PVDF) membrane (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The transferred membranes was blocked with 5% bovine serum albumin (BSA) for 1 h at room temperature and then incubated overnight at 4°C with primary antibodies at 1:1,000 dilution. Antibodies against the following proteins were used: FOXA1 (cat. no. ab170933, Abcam, Cambridge, UK), cleaved poly (ADP-ribose) polymerase (PARP, cat. no. 5625), p53 upregulated modulator of apoptosis (PUMA; cat. no. 4976), Bax (cat. no. 2772), BH3 interacting domain death agonist (Bid; cat. no. 2002), Bcl-xL (cat. no. 2764), cyclin D1 (cat. no. 2926), p21 (cat. no. 2947), cyclin-dependent kinase 2 (CDK2, cat. no. 2546), CDK4 (cat. no. 2906), myeloid cell leukemia-1 (Mcl-1; cat. no. 5453), phos pho-signal transducer and activator of tran scription-3 (phospho-STAT3, cat. no. 9145), STAT3 (cat. no. 9139) from Cell Signaling Technology, Inc. (Danvers, MA, USA), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH, cat. no. FL-335) from Santa Cruz Biotechnology, Inc.. The membranes were washed four times with Tris-buffered saline-0.1% Tween-20 (TBS-T) and were incubated with a horseradish peroxidase-conjugated secondary antibody (anti-rabbit, cat. no. 7074, anti-mouse, cat. no. 7076, Cell Signaling, Technology, Inc.) at 1:2,000 dilution for 1 h at room temperature. Following washing with TBS-T, the protein bands were developed using an Enhanced Chemiluminescent reagent (Amersham, GE Healthcare Life Sciences) and were analyzed on the luminescence image analyzer LAS-4000 (Fujifilm, Tokyo, Japan). The bands of the immunoblot were quantified using Multi-Gauge software (ver. 3.0, Fujifilm).

The transfected cells were plated in the upper well of Transwell filter chambers (8.0-µm pore size; Costar, Cambridge, MA, USA) coated with 1% gelatin. Fibronectin (10 µg/ml, Sigma-Aldrich; Merck KGaA) was added as a chemoattractant to 0.2% BSA (Sigma-Aldrich; Merck KGaA) containing medium in the lower chamber. Following overnight incubation at 37°C, the cells invaded to the lower surface of the upper chamber were fixed with 70% ethanol for 1 min at room temperature and stained with Hemacolor^® Rapid staining solution (Merck KGaA) following the manufacturer’s protocol. Under a light microscope (magnification, ×200, Olympus BX51, Olympus, Tokyo, Japan), the stained upper chambers were placed on the corner squares of a hemocytometer and images were captured. Subsequently, the stained cells were counted in five selected fields (each 0.25 mm², four smaller squares). The results are expressed as the mean ± SE of the number of cells/field of three individual experiments.

Formalin-fixed, paraffin-embedded tissue blocks were obtained from 403 patients who underwent surgery for colorectal cancer at Chonnam National University Hwasun Hospital (Jeonnam, Korea) between July 2004 and June 2006. The tissue blocks were selected by viewing original pathologic slides and selecting blocks that showed the junction between normal colorectal epithelium and tumor tissue. Patient characteristics, including sex, age at the time of surgery, histologic grade and stage, were obtained by examining medical records and contacting pathologists and physicians when necessary. No patient had received anticancer therapy prior to surgery. The tumors were staged in accordance with the American Joint Committee on Cancer staging system (24). Overall survival was measured from the time of surgery until follow-up on December 31, 2013. The present study was approved by the Institutional Review Board of the Chonnam National University Hwasun Hospital (Jeonnam, Korea; CNUHH-2017-163). This was a retrospective study and written informed consent was obtained from each participant prior to tissue acquisition at the time of hospitalization. All participants provided written consent for their information to be stored in the hospital database and used for research.

The paraffin-embedded tissue samples (4-µm-thick tissue sections) were deparaffinized using xylene and rehydrated with ethanol at graded concentrations (100-60%). The samples were then boiled in citrate buffer (pH 6.0, Dako, Agilent Technologies, Inc., Carpentaria, CA, USA) for 10 min in a pressure boiler for antigen retrieval. Endogenous peroxidase activity was blocked using peroxidase-blocking solution (Dako, Agilent Technologies, Inc.) for 10 min at room temperature, and non-specific reactivity was blocked with Dako^® Protein Block Serum-Free solution (Dako, Agilent Technologies, Inc.) for 30 min at room temperature. The samples were incubated with primary anti-FOXA1 (1:200 dilution; cat. no. ab170933, Abcam) antibodies for 1 h at room temperature. Bound antibody was visualized with the DakoReal™ Envision HRP/DAB detection system (Dako, Agilent Technologies, Inc.). Nuclear counterstaining was performed with Mayer’s hematoxylin solution (Sigma-Aldrich, Merck KGaA). The stained tissues were observed and images were captured under a light microscope (Olympus BX51). Immunohistochemical staining was assessed by two independent pathologists blinded to the patient clinical outcome data. FOXA1 staining intensity was scored as follows: 0, no staining; 1, weak; 2, moderate; 3, strong. The percentage of the stained area was scored as 0 for the absence of positive staining in tumor cells, 1 for positive staining in <10% of the tumor cells, 2 for positive staining in 10-50% of the tumor cells, and 3 for positive staining in >50% of the tumor cells. The final score index was obtained by multiplying the intensity and area percentage scores. The final scores for the 403 tumor samples ranged between 0.0 and 12.0 with a mean of 6.0. Samples with a total score of ≥6 were designated as positive for FOXA1 expression, whereas those with a total score of <6 were designated as negative for FOXA1 expression.

Tumor cell apoptosis was determined using the terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) system (Promega, Madison, MA, USA) according to the manufacturer’s instructions. Briefly, the tissues were deparaffinized, rehydrated through a graded alcohol series (100-60%) and incubated in permeabilization solution. Labeling was performed by adding the terminal deoxynucleotide transferase enzyme reaction mix to tissue sections mounted on slides. Following washing with TBS-T, the slides were incubated with the enzyme substrate 3,3-diaminobenzidine (DAB) for color development, which was used to localize the labeled cells. The apoptotic index (AI) was calculated as the number of TUNEL-positive cells per 1,000 tumor cell nuclei. Proliferation of the tumor cells was visualized by immunohistochemical staining with Ki-67 (1:500 dilution, cat. no. ab833, Abcam) antibody. The samples were incubated with primary anti-ki-67 diluent for 1 h at room temperature. Bound antibody was visualized with the DakoReal™ Envision HRP/DAB detection system (Dako, Agilent Technologies, Inc.). Nuclear counterstaining was performed with Mayer’s hematoxylin solution (Sigma-Aldrich, Merck KGaA). The stained tissues were observed and images were captured under a light microscope (Olympus BX51). Nuclei stained with Ki-67 antibody were considered positive. The Ki-67 labeling index (KI) was defined as the number of Ki-67-positive nuclei per 1,000 tumor cell nuclei.

Statistical analyses were conducted using the Statistical Package for Social Sciences software version 20.0 (IBM Corp., Armonk, NY, USA). A χ² test was used to analyze the association between the expression of FOXA1 and clinicopathological parameters. The survival rates were calculated using the Kaplan-Meier method, and the statistical significance of differences was examined using the log-rank test. The Student’s t-test was used for the analysis of association between the expression of FOXA1 and the apoptosis and proliferation of cells in human colorectal cancer. Experimental differences between the FOXA1 knockdown or overexpression group and control group were tested with the Student’s t-test. Each experiment was repeated at least three times. P<0.05 was considered to indicate a statistically significant difference.

To investigate the protein expression of FOXA1 in colorectal cancer cells, various human colorectal cancer cell lines, including HCT116, DLD1, DKO1, SW480, Caco2, COLO205 and HT-29, were subjected to western blot analysis. Among these cells, the protein expression of FOXA1 was the highest in DLD1 cells and the lowest in HCT116 cells (Fig. 1A). FS or FV were used to modulate the endogenous protein expression of FOXA1 in DLD1 and HCT116 cells, respectively. The protein expression of FOXA1 was specifically decreased at the protein level by transfection of FS in DLD1 cells and was increased by the transfection of FV in HCT116 cells (Fig. 1B).

The number of invading FS-transfected DLD1 cells was significantly decreased, compared with those in the SS-transfected cells (P<0.05). By contrast, the FV-transfected HCT116 cells exhibited a significantly increased number of invading cells compared with the EV-transfected cells (P<0.05) (Fig. 2).

To determine the potential effects of FOXA1 on cell proliferation, cells were subjected to a cell proliferation assay 2 days after transfection with FS or FV. Proliferating cells, as determined by absorbance, were decreased significantly in the FS-transfected cells, compared with the SS-transfected cells in DLD1 cells (P<0.05). By contrast, the FV-transfected HCT116 cells exhibited significantly increased proliferation, compared with the EV-transfected cells (P<0.05) (Fig. 3).

Flow cytometric analyses were performed to evaluate the impact of FOXA1 on apoptosis and cell cycle distribution. The apoptotic rate of cells transfected with FS was significantly increased, compared with those transfected with SS in DLD1 cells (8.65±4.9, vs. 12.69±5.0%; P<0.01). In addition, the apoptotic rate was decreased in HCT116 cells following the overexpression of FOXA1 (15.14±2.3, vs. 13.89±0.2%; Fig. 4A). To determine the activation of caspases during the knockdown and overexpression of FOXA1, caspase-specific activities were detected. The expression of cleaved PARP was increased in DLD1 cells following the knockdown of FOXA1, and was decreased in HCT116 cells following the overexpression of FOXA1 (Fig. 4B). Whether the impact of FOXA1 on apoptosis is associated with the modulation of apoptosis regulatory proteins was further examined. As shown in Fig. 4B, FOXA1 knockdown led to an increase in the pro-apoptotic protein, PUMA. By contrast, the overexpression of FOXA1 led to a decrease in PUMA and an increase in anti-apoptotic proteins Bid and Mcl-1. The effects of FOXA1 on cell cycle distribution and the regulators involved in cell cycle distribution were evaluated in human colorectal cancer cells. The overexpression of FOXA1 inhibited cell cycle arrest in the subG1 phase of HCT116 cells, and its knockdown induced cell cycle arrest in DLD1 cells (Fig. 5A). As shown in Fig. 5B, positive regulators of the cell cycle, including cyclin D1, CDK2 and CDK4, exhibited significantly decreased protein levels, whereas the negative regulator of CDKI-p21 was significantly increased by FOXA1 knockdown in DLD1 cells. The protein levels of cyclin D1 and CDK2 were significantly increased and the protein level of p21 was significantly decreased by the overexpression of FOXA1 in HCT116 cells (Fig. 5B).

To examine whether FOXA1 activates oncogenic signaling pathways in human colorectal cancer cells, the present study determined the phosphorylation level of STAT3 signaling protein using western blotting. It was found that the phosphorylation of STAT3 was decreased by FOXA1 knockdown in DLD1 cells. By contrast, the phosphorylation of STAT3 was increased by the overexpression of FOXA1 in HCT116 cells (Fig. 6).

To examine the prognostic role of FOXA1 in the progression of human colorectal cancer, the present study investigated the expression of FOXA1 immunohistochemically in formalin-fixed, paraffin-embedded tissue sections obtained from 403 patients with colorectal cancer, and results were associated with the clinicopathological data of these patients. FOXA1 protein staining was either absent or weak in the normal colorectal mucosa (Fig. 7A and B). The immunohistochemical staining of FOXA1 protein was predominantly identified in the nucleus of colorectal cancer cells and was not detected in the tumor stroma (Fig. 7C and D). The survival rates of patients with colorectal cancer and the association between the expression of FOXA1 and clinicopathological parameters in these patients were analyzed. It was observed that the expression of FOXA1 was significantly associated with differentiation, cancer stage, lymphovascular invasion, depth of invasion and lymph node metastasis (P=0.006, 0.001, 0.034, 0.009 and 0.002, respectively; Table I). In addition, the overall survival rates of patients with FOXA1-positive tumors were significantly lower than those of patients with FOXA1-negative tumors (P<0.001; Fig. 8).

All tumor samples were subjected to a TUNEL assay and immunohistochemical staining of Ki-67 to determine apoptosis and cell proliferation in the tumor cells. The AI of the 403 tumor samples ranged between 0.9 and 19.9, with a mean AI of 8.7±6.3. No significant association was observed between the expression of FOXA1 and AI (P=0.152). The KI of the 403 tumor samples ranged between 32.4 and 97.3, with a mean KI of 62.4±18.5. The mean KI value of FOXA1-positive tumors was 74.8±14.9, which was significantly lower than that of FOXA1-negative tumors (P<0.001) (Table II).