Primary cultures of normal esophageal epithelial cells (NEECs) were established in our laboratory. NEECs were established from fresh biopsies of adjacent non-cancerous esophageal tissue, according to a previous report (17). The NEECs and ESCCs were grown at 37°C in 5% CO₂ with keratinocyte serum-free medium, with 40 μg/ml bovine pituitary extract, 1.0 ng/ml epidermal growth factor, 100 U/ml penicillin and 100 μg/ml streptomycin.

The esophageal cancer cell lines, including NECC, TE1, TE13, Eca109, EC9706, KYSE140 and KYSE30, were obtained from the Cell Bank of the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). All cell lines were grown in RPMI-1640 medium (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum, 100 μg/μl streptomycin and 100 μg/μl penicillin in a humidified incubator containing 5% CO₂ at 37°C.

The present study was conducted on a total of 25 samples from ESCC patients with nodal involvement, which were histopathologically and clinically diagnosed at the First Affiliated Hospital of Zhengzhou University in 2012. For the use of these clinical materials for research purposes, prior patient consent and approval from the Institutional Research Ethics Committee were obtained.

Total RNAs were extracted from cells with TRIzol reagent (Invitrogen). For the detection of KRAS mRNA, cDNA was synthesized from 1 mg of total RNA by means of a reverse reaction kit according to the manufacturer’s instructions (Promega, Madison, WI, USA). Human GAPDH was amplified in parallel as an internal control. For miR-27a, reverse transcription and qRT-PCR reactions were performed by means of a qSYBR-Green-containing PCR kit (Shanghai Genepharma Co., Ltd., Shanghai, China), and U6 snRNA was used as an endogenous control for miRNA detection. Expression of each gene was quantified by measuring cycle threshold (Ct) values and normalized using the ΔCt method [2^{Ct(reference) − Ct(target)}] relative to U6 snRNA or GAPDH.

luc-UTR vectors were constructed by cloning the predicted miR-27a target region or its mutant control into the NheI and SalI sites of the pmirGLO luciferase vector (Promega) using the PCR generated fragments. The oligonucleotide pairs contained the Kpn1 internal site for clone confirmation: sense-wt: 5′-CTAGCTAGGTACCTTGAACT AGCAATGCCTGTGAAAG-3′ and antisense-wt: 5′-TCG ACTTTCACAGGCATTGCTAGTTCAAGGTACCTAG-3′; sense-mut: 5′-CTAGCTAGGTACCTTATTATAGCAATG CACACAGAAG-3′ and antisense-mut: 5′-TCGACTTCTGT GTGCATTGCTATAATAAGGTACCTAG-3′. Bold indicates NheI and SalI sites; underlining indicates the Kpn1 site; italics indicates the mutated sites.

Synthesized RNA duplexes of scramble miRNA, miR-27a, and their inhibitors anti-scramble and anti-miR-27a were obtained from Baoxin Bio-Technology Co., Ltd. (Zhengzhou, China). To construct a vector expressing miR-27a, the precursor sequence of miR-27a (MI0000085) was synthesized, annealed and then inserted into the BamHI-HindIII fragment of the pGCsi/U6 vector (GeneChem, Shanghai, China). A construct including the non-specific miRNA was used as a negative control. The miR-27a knockdown lentivirus was purchased from Baoxin Bio-Technology.

The KRAS-expressing vector was constructed by cloning full-length KRAS cDNA into the eukaryotic expression vector pcDNA3.1(+) (Invitrogen). The empty pcDNA3.1(+) vector was used as a negative control.

TE-1 cells were infected with the miR-27a lentivirus or the control lentivirus expressing a scrambled miRNA. All cells were selected with 500 mg/l G418 to generate two stable monoclonal cell lines (a stable cell line expressing miR-27a, TE-1-miR-27a and a control stable cell line, TE-1-scramble).

To establish stable miR-27a knockdown cell lines, TE-1 cells were transduced with the miR-27a knockdown lentivirus or the control lentivirus and selected with 5 mg/l puromycin.

For miRNA and pcDNA3.1-KRAS combination experiments, TE-1-miR-27a and TE-1-scramble cells were transfected with pcDNA3.1-KRAS or empty vector using Lipofectamine 2000 (Invitrogen).

TE-1-miR-27a cells were transfected with pmirGLO-KRAS-wt, pmirGLO-KRAS-mut or pmirGLO-ctrl using Lipofectamine 2000 (Invitrogen). Luciferase activity was measured 24 h after transfection using the Dual-Glo luciferase assay system (Promega). The Renilla luciferase activity served as internal control.

Cells were seeded onto 96-well plates at a density of 5×10⁴ cells/well in 100 μl medium. All cells were maintained in a humidified 37°C incubator with 5% CO₂. 20 μl 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) solution (5 g/l in phosphate-buffered saline) was added to each well of the microplate, and the absorbance at 570 nm was measured by a microplate reader. After a 4-h incubation, the number of viable cells was measured a point, was set 5 re-wells.

TE-1 cells (5×10⁶/well) were seeded to 90% confluence in a 6-well plate for overnight culture. The following day a scratch was made through the center of each well using a 200-μl pipette tip, creating an open ‘scratch’ or ‘wound’ that was clear of cells. The dislodged cells were removed by three washes with complete culture media, and cells were incubated under standard conditions. Migration into the open area was documented at 72 h post-scratching.

Western blot analyses were performed as previously described (18). Antibodies against KRAS (Sigma-Aldrich), ERK (Santa Cruz Biotechnology), phosphorylated ERK (Santa Cruz Biotechnology), and c-Fos (Santa Cruz Biotechnology) were obtained from Cell Signaling Technology. The anti-β-actin antibody was from Santa Cruz Biotechnology.

Five-week-old male nude athymic BALB/c nu/nu mice were used for examining tumorigenicity. To evaluate the role of miR-27a in tumor formation, TE-1-miR-27a cells, TE-1-scramble control, TE-1-anti-miR-27a cells, or TE-1-anti-scramble control were propagated and inoculated subcutaneously into the dorsal flanks of nude mice (2×10⁶ cells in a 0.2-ml volume). Tumor size was measured every 5 days. After 30 days, the mice were sacrificed, necropsies were performed and the tumors were weighed. Tumor volumes were determined according to the following formula: A × B²/2, where A is the largest diameter and B is the diameter perpendicular to A. The experiments were performed using five mice per group, and all animal procedures were performed in accordance with institutional guidelines.

Statistical evaluation of data was performed using SPSS 13 analysis software (SPSS, Chicago, IL, USA). Independent-samples t-test, or paired-samples t-test were used to evaluate statistical significance. Spearman’s correlation tests were used to evaluate the pair-wise expression correlation between miR-27a and KRAS. Data are shown as means ± SEM. The significant level was set at P<0.05.

Published datasets archived in the publicly available Gene Expression Omnibus (GEO) repository were used for re-analysis using GEO2R (NCBI online gene expression tool) (19). Datasets published in GEO reference series GSE13937 (20) were used for microRNA expression analysis of hsa-miR-27a in ESCC compared with adjacent normal esophageal tissues. These datasets used OSU-CCC Human and Mouse MicroRNA Microarray version 3.0 array platform. Normalized signal intensity value was analyzed using the paired-samples t-test analysis. The expression of hsa-miR-27a was decreased in tumors when compared with that in adjacent normal tissues from ESCC patients with nodal involvement (n=20; P=0.041). However, in ESCC patients without nodal involvement, this difference was not significant (n=24; P=0.785) (Fig. 1A).

We used quantitative real-time PCR (qRT-PCR) to measure mature miR-27a expression levels in ESCC tissues of patients with nodal involvement and cell lines. miR-27a was significantly downregulated in ESCC tissues when compared with the paired adjacent normal tissues (paired-samples t-test, n=25; P=0.025) (Fig. 1B). In six esophageal carcinoma cell lines, the expression of miR-27a was decreased, particularly in TE-1 cells (independent-samples t-test; P=0.003) (Fig. 1C).

To determine the role of miR-27a in ESCCs, we performed a bioinformatic search (TargetScan and PicTar) to identify putative mRNA 3′-UTR targets for the mature miR-27a. We found that miR-27a had a seed region that matched the 3′-UTR of human KRAS (nucleotides 343–349; NM_033360) (Fig. 2A). To verify that KRAS is a direct target of miR-27a, the KRAS 3′-UTR containing the miR-27a binding site was cloned into the pmirGLO control vector downstream of the luciferase ORF. This reporter construct was used to transfect 293 cells. Co-transfection of miR-27a with the wt KRAS 3′-UTR construct in 293 cells resulted in a significant inhibition of luciferase activity compared with the negative control (Fig. 2B). Mutagenesis of the miR-27a binding site within the KRAS 3′-UTR abolished the ability of miR-27a to regulate the luciferase expression (Fig. 3B). In addition, overexpression of miR-27a in TE-1 cells strongly reduced the endogenous protein and mRNA levels of KRAS when compared with these levels in the control (Fig. 2C and D). In clinical ESCC specimens, there was an inverse correlation between the mRNA/protein levels of KRAS and the level of miR-27a expression (Fig. 2E). These data suggest that miR-27a acts as a tumor suppressor by targeting KRAS in ESCC.

KRAS activation can trigger several important signaling pathways, such as the Ras/Raf/MEK/ERK pathways, most of which regulate cell proliferation, survival and invasion. Therefore, we investigated the possibility that miR-27a regulates those pathways by targeting KRAS. Upregulation of miR-27a through transfection of miR-27a in TE-1 cells suppressed the levels of phosphorylated ERK and its downstream effector c-Fos. We also observed that knockdown of miR-27a, through transfection of anti-miR-27a, in TE-1 cells increased the levels of ERK (Fig. 3A). Results of the western blot analysis demonstrated that miR-27a is a negative regulator of the ERK pathway.

Subsequently, rescue experiments were performed by overexpressing the KRAS vector (without an endogenous 3′-UTR) in miR-27a-treated cells. TE-1 cells were first transfected with miR-27a and then with the KRAS-encoding vector 48 h later. The miR-27a-induced downregulation of KRAS was rescued upon the introduction of KRAS, and the phosphorylation level of ERK was altered in a similar manner (Fig. 3B). These observations suggest that miR-27a inhibits the ERK pathway by targeting KRAS.

Overexpression of miR-27a markedly attenuated cell proliferation, and knockdown of miR-27a promoted cell proliferation in TE-1 cells (Fig. 4A). Forced expression of KRAS (without an endogenous 3′-UTR) rescued the cell growth inhibition of miR-27a partially, suggesting that miR-27a regulates cell growth through targeting KRAS (Fig. 4A). Scratch assays were used to measure cell migration or to observe the healing of scratches in cancer cell monolayers. Overexpression of miR-27a significantly inhibited the ability of TE-1 cells to heal scratch assays. However, knockdown of miR-27a significantly promoted this behavior (Fig. 4B). In nude mouse xenograft models, TE-1 cells with miR-27a overexpression developed smaller tumors (both in tumor volume and weight) compared with TE-1 cells with normal miR-27a (scramble). Accordingly, miR-27a knockdown in TE-1 cells promoted tumor formation (both in tumor volume and weight) in the nude mouse xenograft experiments (Fig. 4C).