The human NLCLC cell lines, A549, NCI-H460, NCI-H1299 and NCI-H358, were purchased from the American Type Culture Collection (Manassas, VA, USA), and the HCC827, NCI-H292, and NCI-H838 cells were purchased from the Chinese Academy of Sciences (Shanghai, China). All cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; HyClone; GE Healthcare Life Sciences, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, Carlsbad, CA, USA) and maintained in a humidified incubator at 37°C with 5% CO₂.

Serum, tumor tissues and corresponding adjacent normal tissues for reverse transcription-quantitative polymerase chain reaction (RT-qPCR) analysis were randomly collected from 78 patients with NSCLC who underwent curative resection between January, 2008 and December, 2010 at Jiangxi Chest Hospital (Nanchang, China) and The First Affiliated Hospital of Nanchang University (Nanchang, China). In addition, 22 healthy serum controls were randomly collected from Jiangxi Chest Hospital and The First Affiliated Hospital of Nanchang University. Retrospective data were also obtained from 46 patients with advanced recurrent NSCLC receiving gemcitabine and cisplatin (GP) who had undergone NSCLC resection 2-60 months prior to the GP therapy; patient demographics and overall survival (OS) were recorded. Written informed consent was obtained from all patients and healthy donors, and the study was approved by the institutional Ethics Review Committee of Jiangxi Chest Hospital and The First Affiliated Hospital of Nanchang University.

The immunohistochemistry procedure for IGF1R was performed according to a previously described protocol (15). The percentage of IGF1R-positive cells was scored in five groups: 0 (0%), 1 (1 to ≤25%), 2 (25 to ≤50%), 3 (50 to ≤75%) and 4 (>75%). Groups 0, 1, and 2 groups were defined as low expression, whereas groups 3 and 4 were defined as high expression.

Western blot analysis was performed as described in a previous study (16). Cell lysates were collected and centrifuged for 15 min at 13,200 × g, 4°C. The supernatant was transferred to a clean tube and proteins concentrations were then quantified using the BCA kit (Pierce, Rockford, IL, USA). Proteins were then separated on 10% SDS-PAGE gels and transferred onto nitrocellulose membranes. The membranes were then blocked with 5% skim milk for 2 h at room temperature and incubated overnight at 4°C with primary antibodies. Primary antibodies against IGF1R (ab39675; Abcam, Cambridge, UK), AKT (ab179463; Abcam), p-AKT (ab176657; Abcam) and tubulin (ab6160; Abcam) were used. Immune complexes were then detected by incubating the nitrocellulose membranes with HRP-conjugated goat anti-mouse/rabbit antibody (ab6789/ab6721; Abcam) for 2 h at room temperature, followed by exposure of the membrane to enhanced chemiluminescence reagents (Pierce).

Total RNA was isolated using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) from serum, cell lines, frozen tumor tissues and corresponding adjacent normal tissues according to the manufacturer’s protocol. To detect the levels of lncRNA and mRNA, 1 µg sample of total RNA was used as a template for single strand cDNA synthesis performed using random primers and Primescript reverse transcriptase kit (Takara Bio, Inc., Otsu, Japan) according to the manufacturer’s instructions. GAPDH served as the internal control. To detect the levels of miR-223, we used a MicroRNA Assays kit (Applied Biosystems, Foster City, CA, USA). miR-223 data was normalized to endogenous U6 small RNA. The 2^-∆∆Cq method was used for calculating the relative expression levels as reference (17). The primer sequence: EGFR-AS1, forward, 5′-TCTGT CAGCT CC TTG CACCT C-3′ and reverse, 5′-TGCTC AGTGT GGTCT GATGT C-3′; GAPDH, forward, 5′-TGTTC GTCAT GGGTG TGAAC-3′ and reverse, 5′-ATGGC ATGGA CTGTG GTCAT-3′; miR-223, forward, 5′-AGCTG GTGTT GTGAA TCAGG CCG-3′ and reverse, 5′-TGGTG TCGTG GAGTC G-3′; U6 forward, 5′-CTCGC TTCGG CAGCA CA-3′ and reverse, 5′-AACGC TTCAC GAATT TGCGT-3′. RT-qPCR analysis was performed using a SYBR-Green PCR kit (Takara Bio, Inc.) according to the manufacturer’s protocol. RT-qPCR was performed as follows: pre-denaturation at 95°C for 5 min, followed by 40 cycles of denaturation at 95°C for 10 sec and elongation at 60°C for 30 sec.

To assess cell proliferation, the cells were seeded into a 96-well plate at a density of 500 cells per well. At various time-points (24, 48, 72, 96 and 120 h), CCK-8 solution (Dojindo Molecular Technologies, Inc., Tokyo, Japan) was added to each well, and each plate was incubated for 2 h at 37°C with 5% CO₂. The optical density was measured at 450 nm. For the cisplatin sensitivity assay, the cells were seeded in 96-well plates at a density of 2,000 cells per well. After 24 h, the cells were cultured in complete medium containing the indicated concentrations (0, 2, 4, 6, 8 and 10 µM) of cisplatin or gemcitabine (0, 10, 20, 30, 40 and 50 µM) (both from Qilu Pharmaceutical Co., Ltd., Jinan, China) for 72 h at 37°C with 5% CO₂, and the sensitivity of the cells to cisplatin was measured using a CCK-8 assay.

To construct the reporter vector, EGFR-AS1 [wild-type (wt) or miR-223 target mutant (mu), respectively] amplified and fused to a into the luciferase reporter pGL3 Dual-Luciferase miRNA Target Expression Vector (Invitrogen; Thermo Fisher Scientific, Inc.). Mutant reporter vectors were performed as instructions using the QuickMutation kit (Beyotime, Shanghai, China). The NSCLC NCI-H358 cells were seeded into 24-well plates and transfected with the different reporter plasmid including pGL3-wtEGFR-AS1 or pGL3-muEGFR-AS1, together with miR-223 mimics or negative control using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.). Forty-eight hours after the transfection, luciferase activity was then measured using a Dual Luciferase Reporter Assay system (Promega Corp., Madison, WI, USA) according to the manufacturer’s protocol.

The full-length of EGFR-AS1 sequence was transfected into Trans-OETM plasmid (GenePharma, Shanghai, China) and the empty Trans-OE™ vector was used as a Mock. The specific short-hairpin RNA (shRNA) against human EGFR-AS1 was cloned into pENTRTM/U6 plasmid (GenePharma). A non-targeting shRNA (shNC; GenePharma) was used as a negative control. Cells (5×10⁴ cells/well) were seeded into 6-well plates and were transfected with above plasmid using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer’s instructions. The target sequences for EGFR-AS1 shRNAs was 5′-GACGC AUGAA UGCGA UCUU-3′.

The prediction programs (TargetScan and miRGen) were used to identify potential binding sites with miRNA in the EGFR-AS1 and the target association was found between the EGFR-AS1 and miR-223 among the results.

Experimental data in the present study are presented as the mean ± standard deviation. Statistical analyses were performed using the SPSS software package (version 16.0; SPSS Inc., Chicago, IL, USA) with Student’s t-test or one-way analysis of variance. Multiple groups were compared by the one-way ANOVA and post hoc Dunnett’s test, and data among three groups were compared by the Kruskal-Wallis test followed by a Mann-Whitney U post hoc test with Bonferroni’s correction. A χ² test was used to analyze the association between the expression of EGFR-AS1 and clinicopathological parameters. Survival curves were analyzed using the Kaplan-Meier method and assessed using a log-rank test. The Cox proportional hazard regression model was applied to identify independent prognostic factors. The association between the expression of EGFR-AS1 and mRNA expression of IGF1R in NSCLC tissues was analyzed using Pearson’s correlation. P<0.05 was considered to indicate a statistically significant difference.

The level of EGFR-AS1 was detected in 78 paired NSCLC tissues and adjacent normal tissues using RT-qPCR analysis. As shown in Fig. 1A, the expression of EGFR-AS1 was significantly increased in 61.5% (48/78) of the tumor tissues compared with the adjacent normal tissues. Furthermore, the expression level of EGFR-AS1 was positively associated with the clinical stage in patients with NSCLC (Fig. 1B). To further investigate the role of EGFR-AS1 in NSCLC, the association between EGFR-AS1 and the clinicopathologic characteristics of patients with NSCLC was analyzed. As shown in Table I, the results indicated that the expression of EGFR-AS1 was positively associated with tumor size (P=0.004) and clinical stage (P=0.016). For further analysis, the patients were dichotomized into EGFR-AS1^High (n=48) or EGFR-AS1^Low (n=30) groups. Statistically, there was a positive association between the expression level of EGFR-AS1 and the cumulative recurrence rate (P=0.009; Fig. 1C) and, to a lesser extent, the OS rate (P=0.012; Fig. 1D). Multivariate analysis revealed that the expression of EGFR-AS1 in tumors was an independent prognostic predictor for cumulative recurrence and OS in patients with NSCLC (Tables II and III).

Subsequently, the present study analyzed retrospective data from 46 patients with advanced recurrent NSCLC receiving gemcitabine and cisplatin (GP) who had undergone NSCLC resection 2-60 months prior to the GP therapy; patient demographics were recorded (Table IV). The expression levels of EGFR-AS1 were then measured using RT-qPCR analysis, and Kaplan-Meier survival analysis revealed that the OS for the EGFR-AS1high group was lower than that for the EGFR-AS1^Low group (Fig. 2A and B); therefore, it was hypothesized that high expression levels of EGFR-AS1 lead to NSCLC GP resistance. To examine whether EGFR-AS1 was present in patient plasma, RNA was extracted from the plasma of 46 patients with NSCLC. As shown in Fig. 2C, EGFR-AS1 was detectable in the plasma and positively correlated with that in NSCLC tissues. Compared with healthy donors, the plasma levels of EGFR-AS1 were increased in patients with NSCLC, particularly in GP-resistant patient plasma (Fig. 2D). Together, these results suggested that plasma EGFR-AS1 may be a promising non-invasive biomarker for predicting the response to GP therapy in patients with NSCLC.

To further determine the biological roles of EGFR-AS1 in NSCLC, RT-qPCR analysis was performed to detect the relative expression levels of EGFR-AS1 in different NSCLC cell lines. The A549 and HCC827 cells expressed relatively high levels of EGFR-AS1, whereas the NCI-H1299 and NCI-H358 cells expressed relatively low levels of EGFR-AS1 (Fig. 3A). NSCLC cell lines with stably enforced or reduced expression of EGFR-AS1 were established (Fig. 3B and C). A cell proliferation assay revealed that the enforced expression of EGFR-AS1 significantly increased cell proliferation, whereas the reduced expression of EGFR-AS1 inhibited cell proliferation (Fig. 4A-D). Furthermore, the enforced expression of EGFR-AS1 significantly promoted NSCLC cell resistance to gemcitabine and cisplatin, whereas reduced EGFR-AS1 increased the sensitivity of NSCLC cells to gemcitabine and cisplatin (Fig. 4E-H).

The bioinformatics analysis revealed that the EGFR-AS1 transcript contains a potential binding site for miR-223 (Fig. 5A), which is downregulated in several types of cancer, suggesting that EGFR-AS1 may act as a ceRNA in NSCLC cells. To confirm whether EGFR-AS1 is a ceRNA of miR-223 in NSCLC, luciferase reporter assays were performed following the transfection of NSCLC NCI-H358 cells containing a synthetic miR-223 mimic with the wild-type or mutated miR-223 target sequence of EGFR-AS1. The results showed that miR-223 reduced the luciferase activity in cells trans-fected with wild-type EGFR-AS1 but had no effect on the luciferase activity in cells transfected with mutant EGFR-AS1 (Fig. 5B). In previous studies, IGF1R was verified as a target gene of miR-223 in NSCLC cells (18,19). In the present study, EGFR-AS1 elevated the mRNA and protein expression of IGF1R, leading to the upregulation of the activity of AKT (Fig. 5C and D). Furthermore, the mRNA expression of IGF1R in human NSCLC tissues was positively correlated with the level of EGFR-AS1 (Fig. 5E). The expression of miR-223 in human NSCLC tissues was negatively correlated with the level of IGF1R (Fig. 5F).

To further determine the biological roles of EGFR-AS1 in NSCLC, NCI-H1299 and NCI-H358 cell lines with stable forced expression of the miR-223 target sequence EGFR-AS1 mutant were established. A cell proliferation assay revealed that forced miR-223 target sequence mutant EGFR-AS1 expression neither increased cell proliferation nor promoted NSCLC cell resistance to cisplatin and gemcitabine (Fig. 6A-D). Additionally, mutant EGFR-AS1 did not elevate the mRNA or protein expression of IGF1R in NSCLC cells (Fig. 6E and F). Subsequently, the expression of IGF1R was investigated in NSCLC tissues. It was found that the expression of IGF1R was significantly increased in 66.7% (52/78) of tumor tissues compared with adjacent normal tissues (Fig. 6G).

The patients were dichotomized into a miR-223^High (n=31) or a miR-223^Low (n=47) group. The results showed that there was a significant negative association between the expression of miR-223 and cumulative recurrence rate (P=0.019; Fig. 7A) and, to a greater extent, the OS rate (P=0.006; Fig. 7B). In addition, the patients were dichotomized into a IGF1R^High (n=56) or IGF1R^Low (n=22) group. Statistically, there was a significant positive association between the expression of IGF1R and the cumulative recurrence rate (P=0.039; Fig. 7C) and, to a lesser extent, the OS rate (P=0.042; Fig. 7D).