TRIzol reagent and Lipofectamine 2000 were purchased from Invitrogen (Carlsbad, CA, USA). The First Strand cDNA Synthesis kit and SYBR Premix Taq were purchased from Takara (Dalian, Liaoning, China). MTT [3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide], bromodeoxyuridine (BrdU), anti-BrdU antibody and Smad signaling pathway inhibitor LND-193189 were purchased from Sigma (St. Louis, MO, USA). LND-193189 (100 nM) was used for cell treatment. DAPI, the BCA protein assay kit and the ECL Plus kit were obtained from Beyotime Institute of Biotechnology (Beijing, China). BD BioCoat Matrigel invasion chambers were obtained from BD Biosciences (San Jose, CA, USA). Primary antibodies against human COL11A1 and cleaved caspase-3 were obtained from Abcam (Cambridge, MA, USA). Anti-p-Smad2, anti-Smad2 and anti-GAPDH were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). The horseradish peroxidase-conjugated anti-rabbit and anti-mouse IgG secondary antibodies were purchased from Santa Cruz Biotechnology. Biotinylated- and Cy3-conjugated anti-rabbit secondary antibodies were from Boster (Wuhan, Hubei, China).

A total of 65 samples from NSCLC patients ranging in age from 20–75 years were surgically obtained at the Department of Thoracic Surgery at Fudan University Shanghai Cancer Center. NSCLC was diagnosed according to surgical and pathological findings, which were based on the guidelines described by the 6th edition of AJCC/UICC (Table I). After explaining the purpose of the study, written informed consent was obtained from each patient. Karnofsky performance scale (KPS) score for all patients was ≥60 points. Patients were excluded if they had received any prior treatments involving chemotherapy, radiotherapy or immunotherapy, or presented with other malignant diseases. Recurrent tumor tissues were obtain from 5 NSCLC patients who had received at least three cycles of standard chemotherapy consisting of intravenous (i.v.) cisplatin after surgery. This study was approved by the Institutional Ethics Review Boards of Fudan University Shanghai Cancer Center.

The human NSCLC cell lines (A549, H23, H520 and H1975) were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA) and cultured in RPMI-1640 medium (Gibco, Rockville, MD, USA) supplemented with 10% fetal calf serum (FBS), 100 U/ml penicillin G, 100 mg/ml streptomycin sulfate and 2 mmol/l glutamine (all from Gibco) at 37°C in a humidified incubator under an atmosphere of 5% CO₂ in air.

Human COL11A1 cDNA was amplified from H23 cells by PCR and constructed into the pcDNA3.1 vector using NheI and NotI restriction sites (forward primer sequences, ATAAGA ATGCGGCCGCATGGAGCCGTGGTCCTCTAGGT; and reverse primer sequences, ATAAGAATGCGGCCGCTTAGC CAAGAAAACAAACAGGACCAACTTC. COL11A1 siRNA and control siRNA were purchased from Santa Cruz Biotechnology. Transfection of the vector or siRNA to cells was performed using Lipofectamine 2000 according to the manufacturer's protocol.

Total RNA was extracted with TRIzol reagent from the NSCLC samples and cell lines according to the manufacturer's instructions. Total RNA (5 µg) was used for reverse transcription using the First Strand cDNA Synthesis kit. The expression of mRNA was examined by qRT-PCR with SYBR Premix Taq and Applied Biosystems 7500 Sequence Detection system. The relative expression levels of mRNA were normalized to GAPDH expression, and the amplification results for qRT-PCR were calculated using the 2^−ΔΔCt method. PCR reaction was performed using COL11A1 primers: 5′-TGGTGATCAGAATCAGAAGTTCG-3′, (forward) and 5′-AGGAGAGTTGAGAATTGGGAATC-3′ (reverse); and GAPDH, 5′-GTGGACATCCGCAAAGAC-3′ (forward) and 5′-AAAGGGTGTAACGCAACTA-3′ (reverse).

The paraffin-embedded tissue samples from postoperative patients were sectioned into 5-µm slices and embedded with paraffin after fixation with 10% formaldehyde. The samples were then deparaffinized in xylene, rehydrated and quenched the endogenous peroxidase activity by 3% hydrogen peroxide. Non-specific binding was blocked with 1% bovine serum albumin. The sections were then incubated overnight at 4°C with the COL11A1 primary antibody, and subsequently with the horseradish peroxidase-conjugated secondary antibody. The peroxidase was then developed by 3-amino-9-ethylcarbazole (AEC).

The MTT assay was used to assess cell proliferation and viability. In brief, the transfected NSCLC cells were seeded at a density of 2.5×10³ cells/well in a 100 µl volume of medium in 96-well plates and allowed to attach overnight. MTT (20 µl at 5 mg/ml) was added to the medium after treatment. After incubation at 37°C for 4 h, the supernatant was aspirated, and 100 µl DMSO was added to each well to dissolved the precipitates. Absorbance was then measured at 570 nm using a 96-well microplate reader. The survival ratio was calculated using the following equation: Survival ratio (%) = OD_treated/OD_control × 100. For cell counting, NSCLC cells were plated onto 35-mm dishes at 2×10³ cells/dish. Cells were trypsinized and counted with a hemocytometer on day 0, 1, 2, 3, 4, 5 and 6.

The BrdU assay was performed to analyze cell proliferation according to the manufacturer's recommendations. NSCLC cells were seeded on a 4-well chamber slide and cultured in growth medium overnight. BrdU-labeled medium was then changed and incubated at 37°C for 3 h. Cell medium was removed and the cells were fixed with ethanol for 20 min at −20°C. The cells were permeabilized with 0.1% Triton X-100 in phosphate-buffered saline (PBS) and blocked with 3% FBS in PBS solution. After several washes with PBS, the cells were incubated with the anti-BrdU reagent for 30 min at 37°C. The nuclei were then counterstained with DAPI. Images of each treatment were captured with an Olympus inverted microscope.

The wound-healing assay was used to evaluate the cancer cell motility capacity. In brief, the transfected NSCLC cells were seeded in 6-well plates and cultured overnight. When the culture reached nearly 90% confluency, the cell layer was wounded by dragging a 1,000-µl sterile pipette tip through the monolayer and washed with PBS twice to remove cellular debris. Cells were then cultured for up to 48 h with growth medium. Photographic images of the plates were acquired under a microscope, and the relative surface traveled by the leading edge was assessed.

Cell invasion activity was assessed using 24-well BD BioCoat Matrigel Invasion chambers (8-µm pores, growth factor reduced) according to the manufacturer's instructions. Transfected NSCLC cells (5×10⁴ cells/ml) were loaded into the top chamber with serum-free media. In the lower chamber, 450 µl of growth medium with 10% FBS was added. After 24 h of cultivation, the medium was removed and the Transwell chamber was gently wiped with a cotton swab. The migrated cells were fixed in 4% paraformaldehyde, stained with crystal violet solution and counted under a microscope.

NSCLC cells were washed with PBS and incubated in RIPA buffer (150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris pH 8.0, 5.0 mM EDTA pH 8.0, 0.5 mM dithiothreitol and 1 mM phenylmethylsulfonyl fluoride) containing protease inhibitor on ice for 30 min, and centrifuged at 12,000 × g for 15 min at 4°C. The protein concentration was determined using the BCA protein assay kit, and equal amounts of total protein were separated by 10% SDS-PAGE, and the protein was then transferred from the gel to a nitrocellulose membrane. The nitrocellulose membrane was then blocked with 5% BSA for 1 h. The membrane was then probed with a specific primary antibody and gently shacked at 4°C overnight. The next day, the membrane was washed three times with TBST buffer, and incubated with the secondary antibodies for 1 h at room temperature. Images were acquired using the ECL Plus kit.

Data were analyzed using SPSS software 19.0 for Windows (SPSS, Chicago, IL, USA). Results are represented as the mean value of at least three different experiments and expressed as the mean ± SD and analyzed by t-test or one-way ANOVA. P<0.05 was considered to indicate a statistically significant result.

COL11A1 has been reported to be upregulated in NSCLC and to be correlated with lymph node metastasis and poor prognosis (22). However, its function in NSCLC is still unknown. To investigate the role of COL11A1 in NSCLC development and recurrence, we first examined COL11A1 expression in 65 NSCLC tissues. When compared with the adjacent normal tissues, the expression of COL11A1 was found to be overexpressed in the NSCLC samples (Fig. 1A). We then compared COL11A1 expression in 36 NSCLC samples with lymph node metastasis with 29 patients without lymph node metastasis. As shown in Fig. 1B, COL11A1 was significantly upregulated in NSCLC tissues with lymph node metastasis. From five primary and paired recurrent NSCLC tissues, COL1A1 was found to be significantly increased in all recurrent samples (Fig. 1C). We next performed immunohistochemistry for COL11A1 in the primary and recurrent NSCLC tissues. COL11A1 staining of the recurrent cancer tissues was obviously stronger than that of the primary NSCLC (Fig. 1D).

To evaluate the functional effect of COL11A1, we examined COL11A1 expression in four NSCLC cell lines (A549, H23, H520 and H1975). Among these four cell lines, H23 cells with endogenous low COL11A1 expression and H520 cells with high COL11A1 expression were selected for in vitro studies (Fig. 2A). Western blot assays for COL11A1 expression were used to analyze the transfection efficiency in the H520 cells transfected with COL11A1 siRNA and H23 cells transfected with the COL11A1 plasmid (Fig. 2B). These cells were then subjected to MTT, cell counting and BrdU cell proliferation assays. Knockdown of COL11A1 in the H520 cells significantly inhibited cell proliferation. Conversely, ectopic overexpression of COL11A1 in the H23 cells markedly promoted cell proliferation (Fig. 2C and D). To determine whether COL11A1 regulates cell migration and invasion, we performed wound-healing and Transwell assays in the COL11A1-silenced H520 cells and the COL11A-overexpressing H23 cells. As shown in Fig. 3, depletion of COL11A1 reduced NSCLC cell migration and invasion ability, while an increase in COL11A1 expression enhanced cell migration and invasion.

To determine the active mechanism of COL11A1 that may contribute to the drug-resistant phenotype, we treated NSCLC H520 and H23 cells with different concentrations of cisplatin (0, 2, 4 or 8 µmol/l), and analyzed COL11A1 mRNA levels by qRT-PCR. As shown in Fig. 4A, a gradual increase in COL11A1 mRNA levels was observed in both the H520 and H23 cells in response to the cisplatin treatment. Consistently, western blot analysis further confirmed the cisplatin-induced COL11A1 expression (Fig. 4B). Moreover, we treated the H520 and H23 cells with 8 µmol/l cisplatin and tested the COL11A1 mRNA levels during a 24-h time period. COL11A1 mRNA reached a peak at 12 h and was increased 2-fold when compared to its level at 0 h (Fig. 4C). To assess whether COL11A1 mediates cell sensitivity to cisplatin, we performed MTT cell viability assays following treatment of cisplatin with different concentrations (0, 2, 4, 8, 16, 32 or 64 µmol/l). Cell viability was decreased by cisplatin in a dose-dependent manner in both the H520 and H23 cells. Knockdown of COL11A1 shifted the survival curve of H520 cells downwards, while overexpression of COL11A1 increased H23 cell survival (Fig. 5A, IC₅₀ of 30.72 µmol/l in control siRNA vs. 12.62 µmol/l in COL11A1 siRNA-transfected H520 cells; IC₅₀ of 29.04 µmol/l in vector control vs. 47.33 µmol/l in COL11A1-transfected H520 cells). We further analyzed the cell viability after treatment with 8 µmol/l cisplatin at different time-points (0, 12, 24, 36, 48, 60 and 72 h). COL11A1 silencing resulted in a significant reduction in cell viability. Conversely, COL11A1 overexpression enhanced cell viability (Fig. 5B).

COL11A1 has been reported to regulate the Smad signaling pathway (19). To address the intrinsic mechanisms by which COL11A1 regulates NSCLC cell tumorigenesis, we analyzed the effect of COL11A1 expression on Smad pathway activity. Downregulation of COL11A1 decreased p-Smad2, and upregulation of COL11A1 increased p-Smad2 (Fig. 6A). We next performed MTT, wound-healing and Transwell assays using the COL11A1-overexpressing H23 cells in the presence or absence of LND-193189, a Smad signaling inhibitor. As shown in Fig. 6B–D, overexpression of COL11A1 promoted NSCLC cell proliferation, migration and invasion, which was abolished by inhibition of Smad signaling. Moreover, LND-193189 was found to abrogate cisplatin resistance induced by COL11A1 (Fig. 6E).