The human esophageal squamous cell carcinoma cell lines (EC1, EC9706, EC109, TE1, TE13, Kyse140 and Kyse450) were grown in Dulbecco's modified Eagle's medium (HyClone, Logan, UT, USA) supplemented with 10% FBS [Biological Industries (BI), Inc., Cromwell, CT, USA] at 37°C with 5% CO₂. Human immortalized normal esophageal epithelial cell line (Het-1A) was a kind gift from Professor R. Liu (Southeast China University) and was cultured in Bronchial Epithelial Cell Medium (BEGM; BulleKit).

Human esophageal squamous cell carcinoma tissue array was purchased from Xi'an Alenabio Biotech Co. Ltd. (Xi'an, China). IHC staining was carried out with specific CREB antibody (Abcam Trading Co. Ltd., Shanghai, China). Briefly, the tissue array sections (5 µm) were dehydrated and peroxidase blocked. Primary antibodies were added and incubated at 4°C overnight, followed by staining with a Histostain-SAP kit (ZSGB-BIO, Beijing, China). The slides were counterstained with hematoxylin. The stained slides were observed by microscopy, and images were acquired. Based on staining intensity, samples were classified into five groups from the lowest density (−) to the highest (++++) as previously described (17,18).

Esophageal cancer and adjacent esophageal tissues were collected from esophageal squamous cell carcinoma patients undergoing resection at the Linzhou Cancer Hospital (Linzhou, Henan, China) from July 2012 to September 2014. Histological diagnosis and tumor-node-metastasis (TNM) stages of cancers were determined in accordance with the American Joint Committee on Cancer (AJCC) manual criteria for esophageal cancer.

Knockdown of CREB was carried out using siRNA oligonucleotides. The sequences of the siRNA are as follows: siCREB-1, CCAAGUUGUUGUUCAAGCU; siCREB-2, GAGAGAGGUCCGUCUAAUG; siControl, UUCUCCGAACGUGUCACGU.

Kyse450 and EC1 cells were transfected with siControl or siCREB, and then proteins were collected for western blot analysis, using antibodies against CREB (Abcam Trading Co. Ltd.), P27, WEE1, cleaved caspase-3, cleaved poly(ADP)-ribose polymerase (PARP) (Cell Signaling, Boston, MA, USA). CDC2, CDK2, cyclin B, cyclin A1, cyclin D, cyclin E (Abgent Biotech Co., Ltd., Suzhou, China). Tubulin (Shanghai Likun Trade Co. Ltd., Shanghai, China) was used as a control. Densitometric analysis for the quantification relative to tubulin was performed using the ImageJ software.

The cell viability was determined by Cell Counting Kit-8 (CCK-8; Beyotime Biotech Co., Haimen, China). EC1 and Kyse450 cells were transfected with siControl or siCREB for 24 h, and then seeded into 96-well plates (2.5×10³ cells/well) for 72 or 96 h. Subsequently, CCK-8 solution was added to each well and incubated at 37°C for 2 h. The absorbance at 450 nm was measured.

For the clonogenic assay, cells were transfected with siControl or siCREB, seeded into 6-well plates with 500 cells/well in triplicate, and then incubated for 12 days. The colonies were fixed by 4% paraformaldehyde, stained using crystal violet solution, and then the colonies with >50 cells were counted.

EC1 and Kyse450 cells were transfected with siControl or siCREB for 72 h, and then cells were harvested, fixed in 70% ethanol at −20°C, stained with propidium iodide (PI; 50 µg/ml) containing RNase A (30 µg/ml) (both from Sigma, St. Louis, MO, USA) at 37°C for 30 min, and analyzed for cell cycle profile by flow cytometry (FACScan; Becton-Dickinson, Franklin Lakes, NJ, USA). Data were analyzed using ModFit LT software (Verity Software House, Inc., Topsham, ME, USA). The expression of cell cycle-related proteins was detected using indicated antibodies.

Cells were transfected with siControl or siCREB for 96 h. Apoptosis was determined using the Annexin V-FITC/PI apoptosis kit (BioVision, Inc., Mountain View, CA, USA) according to the manufacturer's instructions. The activities of caspase-3 were measured using the CaspGLOW assay kit (BioVision, Inc.) according to the manufacturer's instructions. Cell proteins after transfection were collected, and then cleaved PARP and caspase-3 were detected.

The invasion assay was carried out in Matrigel (Becton-Dickinson)-coated Transwell inserts with a pore size of 8-µm, as previously described (17). Briefly, the inserts were pre-coated with Matrigel. EC1 cells (3×10⁴) transfected with siRNA for 48 h were seeded in serum-free medium in the upper chamber, whereas medium with 10% FBS was added to the lower well. After incubating for 24 h, the cells in the upper chamber were carefully removed with a cotton swab. The inserts were fixed in methanol, stained with 0.4% crystal violet, observed under microscopy, and images were acquired. Then, the dye was eluted by 33% acetic acid and detected at OD 570 nm.

For wound-healing assay, cells were seeded on 6-well plates and the confluent monolayer was scratched by a plastic pipette, then cells were washed three times with PBS. Images were captured at 0 or 36 h after wounding.

The stable EC1 cell line with lentivirus targeting CREB was established as previously described (19,20). BALB/c nude female mice were subcutaneously injected with 5×10⁶ EC1 cells stably expressing lenti-shCREB or lenti-shControl, respectively. Tumor growth was observed and tumor area was recorded twice a week with a FluorVivo Model-300 imaging system (INDEC BioSystems, Santa Clara, CA, USA) (19). At the time of sacrifice, tumor tissues were harvested, photographed and weighed. Animal experiments were performed in accordance with animal protocols approved by the Institutional Animal Care and Use Committee of Zhengzhou University.

The statistical significance of differences between groups was assessed using GraphPad Prism 5 software (GraphPad Software Inc., La Jolla, CA, USA). The t-test was used for the comparison of parameters between groups. For all tests, two levels of significance (*P<0.05; **P<0.01) were applied.

To investigate whether CREB served as anti-esophageal cancer target, we firstly examined the expression levels of CREB by immunochemistry (IHC) staining of the tissue array derived from human ESCC. Results showed that CREB was overexpressed in ESCC tissues (Fig. 1A), which was confirmed by western blot assay using the esophageal cancer and adjacent esophageal tissues from esophageal squamous cell carcinoma patients (Fig. 1B). Furthermore, the overexpressed CREB was positively correlated with lymph node metastasis and TNM stage of ESCC patients (Table I). However, CREB was expressed higher in ESCC cell lines than human immortalized normal esophageal epithelial cell line (Het-1A) (Fig. 1C). These results implied that CREB may be an attractive anti-ESCC target.

According to the above results, we next examined the effect of knockdown CREB on cell growth. Results showed that the expression of CREB was effectively downregulated using specific siRNA (Fig. 2A). Silencing of CREB inhibited cell growth by cell proliferation (Fig. 2B) and colony formation assay (Fig. 2C).

To elucidate the growth suppression mechanism by CREB silencing, the cell cycle profile was examined after knockdown of CREB. As shown in Fig. 3, knockdown of CREB induced S cell cycle arrest (Fig. 3A) and downregulated the expression of cyclin A1 and D (Fig. 3B).

We next examined whether apoptosis was also responsible for the growth inhibition effect of CREB silencing. Results showed that knockdown of CREB-induced apoptosis, as evident by increased Annexin V-positive cells (Fig. 4A), improved caspase-3 activity (Fig. 4B) and enhanced expression level of cleaved caspase-3 and cleaved PARP (Fig. 4C).

Statistical analysis results showed that the overexpressed CREB was correlated with lymph node metastasis (Table I). So the expression of CREB was downregulated by siRNA and the effect on cell invasion and metastasis was examined. Results showed that silencing CREB inhibited ESCC cell invasion and metastasis by Transwell (Fig. 5A) and wound healing assay (Fig. 5B).

To further investigated the growth suppressive effect of CREB knockdown in vivo, BALB/c nude female mice were subcutaneously injected with 5×10⁶ EC1 cells stably expressing lenti-shCREB (marked as shCREB) or lenti-shControl (marked as shControl), respectively. Tumor growth was observed and tumor area was recorded twice a week. Results showed that CREB silencing suppressed tumor growth over time while control tumors grew rapidly, as revealed by real-time images of tumors (Fig. 6A), tumor growth curve (Fig. 6B; P<0.05), tumor size (Fig. 6C) and tumor weight analysis (Fig. 6D; P<0.05).