The present study was approved by the Ethics Committee of Subei People's Hospital (Yangzhou, China) and written informed consent was obtained from all participants.

A total of 10 pairs of 4% formalin-fixed for 24 h at room temperature, paraffin-embedded (FFPE) NSCLC tissues and corresponding matched non-tumor lung tissues were collected (from 4 patients with squamous cell carcinoma and 6 patients with adenocarcinoma) between May 2012 and October 2012, at Subei People's Hospital. A further 56 FFPE NSCLC tissues (from 31 patients with squamous cell carcinoma and 25 patients with adenocarcinoma) were obtained between February 2008 and December 2009. All tissue samples were collected prior to treatment with chemoradiotherapy. Each sample was confirmed by histopathological evaluation using hematoxylin and eosin staining. Clinical data were recorded at the time of resection and patients were prospectively followed-up to ascertain vital status (the last follow-up date was April 30, 2012).

Human lung cancer A549 and H1299 cells and normal bronchial epithelial BEAS-2B cells were purchased from the Cell Resource Center, Shanghai Institute of Biochemistry, China. (http://www.sibcb.ac.cn/) BEAS-2B cells were derived by transforming human bronchial epithelial cells with an adenovirus 12-simian virus 40 construct, as previously described (20). The cell lines were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum (both from Wisent, Inc., St. Bruno, QC, Canada) at 37°C in a humidified air atmosphere containing 5% carbon dioxide.

A549 and H1299 cells were seeded onto 24-well plates on day 0, exposed to the DNA methylation inhibitor 5-aza-2′-deoxycytidine (5-aza-dC; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) at a final concentration of 5 µmol/l between day 1 and day 3 at 37°C in a humidified air atmosphere containing 5% CO₂ and were harvested under Trypsin-EDTA digestion harvested for RT-qPCR analysis of miR-137 expression on day 4.

Total RNA was extracted from cultured cells using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA). Total tissue RNA was extracted from FFPE tissue sections using the miRNeasy FFPE kit (Qiagen, Inc., Valencia, CA, USA) according to the manufacturer's protocol. Paraffin was removed from freshly cut FFPE tissue sections each up to 10-µm thick using deparaffinization solution using the miRNeasy FFPE kit (Qiagen, Inc.), and samples under went protease digestion at room temperature to release RNA from the sections, then short incubation (at 56°C for 15 min, then at 80°C for 15 min) to reverse formalin cross-linking of the released nucleic acids and DNase digestion to remove DNA. Total RNA (including miRNAs) was dissolved in 20 µl RNase-free water. RNA concentrations were measured using a NanoDrop-1000 (Thermo Fisher Scientific, Inc.) and RNA integrity was determined by 1.5% agarose gel electrophoresis.

RNA was reverse transcribed using the RevertAid First Strand cDNA kit (Thermo Fisher Scientific, Inc.) according to the manufacture's protocolin combination with a stem-loop primer for miRNA-137U6 small nuclear RNA was used as an internal control to normalize the expression levels of miRNA-137. The primer sequences are presented in Table I. Briefly, 1.0 µg total RNA was combined with 4.0 µl Ribo Lock RNase inhibitor, 2.0 µl dNTP mix (10 mM each) and 1.0 µl RevertAid M-MuLV reverse transcriptase in a total reaction volume of 20 µl, which was incubated on an ABI Prism 7900HT Fast Real-Time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.) at 25°C for 5 min, 42°C for 60 min and 70°C for 5 min.

RT-qPCR was performed using LightCycler^® 480 SYBR-Green I Master mix on a LightCycler^® 480 Real-Time PCR system (both Roche Diagnostics, Basel, Switzerland). Each 10 µl PCR mixture contained 1 µl 20-fold diluted reverse transcription product, 5 µl SYBR-Green Master mix, 2 µl RNase-free water, and 1 µl forward and reverse primers. The reactions were incubated at 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec and 65°C for 60 sec. Relative miRNA-137 expression was calculated using the 2^−ΔΔCq method (21), where ΔCq is the difference in threshold cycles (Cq) for the target and reference = Cq(miRNA-137)-Cq(U6). RT and PCR primers were synthesized by Shanghai Sheng Gong Biology Engineering Technology Service, Ltd. (Shanghai, China).

The miR-137 CpG is lands were identified using EMBOSS Software Version 6.3.1 (Institut Pasteur, Paris, France) and EMBOSS (http://gensoft.pasteur.fr/docs/EMBOSS/6.3.1/). Total tissue DNA was extracted from FFPE tissue scrolls using the Qiagen EpiTect Plus FFPE Bisulfite kit (Qiagen, Inc.) according to the manufacturer's protocol. FFPE tissue scrolls were deparaffinized, followed by proteinase digestion and de-cross-linking as aforementioned in RNA isolation paragraph. The DNA bisulfite reaction was then set up and performed using an ABI Prism 7900HT Fast Real-Time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.) using the Qiagen EpiTect Plus FFPE Bisulfite kit (Qiagen, Inc.) according to the manufacturer's protocol. Upon completion of the bisulfite conversion, modified DNA was purified and eluted, the DNA concentrations were measured using a NanoDrop-1000 (Thermo Fisher Scientific, Inc.) and the samples were stored at −20°C for further analysis.

The methylation status of the miR-137 promoter in the FFPE tissue sample was determined by MS-qPCR, as previously described (19). Modified DNA (10 ng) was subjected to PCR amplification on an ABI Prism 7900HT Fast Real-Time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.) at 90°C for 5 min, followed by 40 cycles of 95°C for 15 sec, 60°C for 30 sec and 72°C for 15 sec, then a 10 min final extension at 72°C. The PCR products were diluted 500-fold with water, and 1 µl aliquots were subjected to MS-qPCR using a LightCycler^® 480 SYBR-Green I Master (Roche Diagnostics) at 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec and 51°C for 60 sec, calculated with the formula 2^−ΔCq[Cq(methylated)-Cq (unmethylated)] (21), and expressed as a percentages. qPCR primer sequences are listed in Table I.

Data were analyzed using the SPSS 18.0 statistical software (SPSS, Inc., Chicago, IL, USA). Values presented as the mean ± standard deviation and were analyzed using one-way analysis of variance with post hoc analysis by least significant difference (LSD) test. The correlation between miR-137 expression and tissue methylation levels was evaluated using Pearson's correlation analysis. Descriptive statistics were used to compare the demographic and clinicopathological characteristics (sex, age, pathology, smoking, status, tumor size, histologic grade, T category, lymph node metastasis and clinical stage) (22) of the study population stratified by the miR-137 promoter methylation status, and categorical variables were compared using analysis of variance with post hoc analysis by LSD test. The univariate Kaplan-Meier method was used to estimate disease-free survival and overall survival rates, and survival differences were compared using the log-rank test. In analysis of disease-free and overall survival rates, patients who succumbed prior to recurrence were considered censored at the point of mortality. A multivariable Cox proportional hazards model was used to assess the prognostic value of the level of miR-137 promoter methylation for disease-free and overall survival rates, following adjustment for sex, age, histological grade, T category, age and smoking status. P<0.05 was considered to indicate a statistically significant difference.

The expression of miR-137 was determined by qRT-PCR (Fig. 1). miR-137 was markedly downregulated in lung cancer A549 and H1299 cells compared with that in normal lung bronchial epithelial BEAS-2B cells (Fig. 1A). In line with the cell line analyses, the expression of miR-137 was significantly downregulated in the 10 FFPE lung cancer tissue specimens compared with that in the paired adjacent normal lung tissues (P=0.037; Fig. 1B and C).

CpG plot EMBOSS Software version 6.3.1 software (http://gensoft.pasteur.fr/docs/EMBOSS/6.3.1/; Institut Pasteur, Paris, France) identified 2 CpG islands located close to the miR-137 gene (Fig. 1D) and promoter hypermethylation has previously been reported to be responsible for the repression of miR-137 (14). Therefore, miR-137 expression was analyzed in a lung cell line treated with the methyltransferase inhibitor 5-aza-CdR (5 µmol/l). Untreated control cells expressed lower levels of miR-137, whereas 5-aza-CdR-treated cells expressed higher levels of miR-137 (Fig. 1A). The methylation status of the CpG sites was consistent with the levels of miR-137 expressed in the lung cancer cell lines.

Subsequently, miR-137 expression and the promoter methylation status in the 56 FFPE lung cancer tissues were compared using RT-qPCR and MS-qPCR. A significant negative correlation between the expression level of miR-137 and miR-137 promoter methylation was observed in the lung cancer tissues (Pearson's correlation, r=−0.343, P=0.01; Fig. 2), suggesting that promoter methylation silences miR-137 in lung cancer.

To determine whether the miR-137 promoter methylation status is associated with lung cancer, the association between the miR-137 promoter methylation status and the clinicopathological characteristics of lung cancer were further analyzed. Low levels of miR-137 promoter methylation were significantly associated with smoking, positive lymph node metastasis and advanced clinical stage (P=0.027, P=0.004 and P=0.021, respectively) (Table II). There was no significant association between miR-137 promoter methylation and sex, age, pathology, tumor size, histological grade or T category (Table II).

Kaplan-Meier survival analysis and multivariate Cox proportional hazards analysis were employed to evaluate the association between miR-137 promoter methylation and prognosis in NSCLC. In the present study, 1 patient succumbed to pneumonia shortly after surgery and 8 patients were lost to follow-up. Patients were stratified according to the median relative miR-137 promoter methylation level (43.79%) in the tumor specimens from the remaining 47 tumor tissues; low miR-137 promoter methylation (n=24; ≤median) and high miR-137 promoter methylation group (n=23; >median). In univariate Kaplan-Meier analysis, patients with high levels of miR-137 promoter methylation exhibited significantly poorer disease-free survival rates (P=0.034; Fig. 3A), but no significant difference was observed in overall survival rates (P=0.136; Fig. 3B), compared with patients with low levels of miR-137 promoter methylation.

In the multivariable Cox proportional hazards analysis, adjusting for gender, age, histologic grade, smoking status and T stage, the results showed that patients with high levels of miR-137 promoter methylation had a higher risk of disease-free death (hazard ratio, 3.333; 95% CI, 1.108–10.028, P=0.032, Table III) compared with the patients with low levels of miR-137 promoter methylation; levels of miR-137 promoter methylation was not significantly associated with overall survival (P=0.366).