A total of 65 NSCLC tumor tissues and their adjacent normal lung tissues were collected between January 2016 and January 2017 at The Second Xiangya Hospital (Changsha, China) by surgical resection for RT-qPCR analysis. The patients included 41 males and 24 females with a mean age of 57.6 years (range, 18–74 years). A total of 73 paraffin-embedded NSCLC and paired adjacent normal lung tissues were provided by the Department of Pathology of The Second Xiangya Hospital and used for immunohistochemistry. Tissues were obtained from 43 male and 30 female patients with a mean age of 58.2 years (range, 19–75 years). The inclusion criteria were as follows: Definitive diagnosis of NSCLC, aged between 18 and 75 years, stage I to stage III tumor, and Tumor Node Metastasis classification (6) data were available. The pathological data, including tumor type, tumor differentiation degree, tumor size, lymph node metastasis and clinical stage, were also collected. The protocol of this study was reviewed by the Ethics Committee of Human Study of the Second Xiangya Hospital (Changsha, China). This approved study was performed in accordance with the ethical standards of the Declaration of Helsinki (as revised in Brazil 2013). Written informed consent was obtained from the subjects for use of their tissue.

The total RNA was extracted from tissues using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA), according to the manufacturer's protocol. cDNA was synthesized using RevertAid^™ First Strand cDNA Synthesis kit (Fermentas; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. RT-qPCR was performed using SYBR^® Green Master mix for Real-Time RT-PCR kit (Takara Bio, Inc., Otsu, Japan), according to the manufacturer's protocol. The BMP-6 gene was amplified using forward primer, 5′-CCCTTCATGGTGGCTTTCTT-3′ and reverse primer, 5′-GAGCGATTACGACTCTGGTTCTGTTGTC-3′. The BMP-6 was amplified at the following thermocycling conditions: 94°C, 1 min, followed by 28 cycles of 94°C, 30 sec; 59°C, 40 sec; and 72°C, 2 min. β-actin was amplified using forward primer, 5′-GCACCACACCTTCTACAATGAG-3′ and reverse primer, 5′-GATAGCACAGCCTGGATAGCA-3′, as an internal control. Each sample was amplified 3 times and the raw data were averaged for mean. Relative gene expression was quantified using the 2^−ΔΔCq in log-10 scale method (7). The normalized log10 value of BMP-6 expression was used for statistical analysis.

The tissues were fixed in 10% formalin solution for 48 h at room temperature and then paraffin-embedded. Tissues were sectioned at 5 µm. Immunohistochemical staining was performed using a Two-step Immunohistochemical Staining kit (OriGene Technologies, Inc., Beijing, China) according to the manufacturer's protocols. After dewaxing, rehydration in descending alcohol series (100–70%), antigen repair at 98°C for 20 min in sodium citrate buffer (10 mM Sodium citrate, 0.05% Tween 20, pH 6.0), and blocking in 10% donkey serum (Sigma-Aldrich; Merck KGaA, Darmstadt, German) in tris-buffer for 2 h at room temperature, sections were incubated with rabbit anti-human BMP-6 monoclonal antibody (dilution, 1:50; cat. no. ab101056; Abcam, Cambridge, UK) overnight at 4°C. Following rinsing with PBS three times for 30 sec each, the sections were incubated with horseradish peroxidase-conjugated goat-anti rabbit secondary antibody (dilution, 1:1,000 dilution; cat. no. ab6721; Abcam) for 2 h at room temperature. After the slices were subject to 3,3′-diaminobenzidine treatment for 5 min at room temperature, hematoxylin staining for 30 sec at room temperature and dehydration, the slices were then sealed and observed under a light microscope. The normal prostate tissue slices provided with the kit were used as a positive control and PBS instead of the primary antibody was used as the negative control. The staining was scored by two experienced pathologists in the Department of Pathology of The Second Xiangya Hospital. The brown granules indicated positive staining of BMP-6. The whole slice was observed with a magnification of ×40 to determine the tumor infiltration edge. A total of 10 randomly selected fields of view were selected under a high magnification (×400) and 100 tumor cells were counted for each field. The intensity of staining was scored as 1 for negative, 2 for positive and 3 for strong positive. The percentage of positively stained cells (the score was 2 or 3) was also calculated. A tissue was defined as negative staining (−) when the tissue was scored 1 or <10% cells were scored 2 or 3, weakly positive staining (+) when 10–30% cells were scored 2 or 3, positive staining (++) when 31–50% cells were scored 2 or 3 and strongly positive staining (+++) when >50% cells were scored 2 or 3.

The Kaplan-Meier plotter online database (http://kmplot.com/analysis/index.php?p=service&cancer=lung#) was computationally used to analyze the association between BMP-6 mRNA expression and overall survival of patients with NSCLC. The Kaplan-Meier survival curve, hazard ratio (HR), and log-rank P-value were obtained. The high expression group was defined when the BMP-6 mRNA expression in the patient was higher or equal to the median mRNA expression of all NSCLC samples. The low expression group was the patients with BMP-6 mRNA expression lower than the median.

The human lung cancer cell lines including H125, A549, H23, H460, H520, H1299, PC9 and Human Bronchial Epithelial Cell (HBE) were purchased from the Shanghai Institute for Biological Sciences (http://english.sibs.cas.cn/rs/fs/). Cells were cultured in Dulbecco's modified Eagle medium or RPMI-1640 medium with 10% fetal calf serum. (Thermo Fisher Scientific Inc.) in 5% CO₂ at 37°C.

H460, H1299, A549, and H520 cells (1×10⁴) were seeded in 96-well plates overnight and treated without (control), or with 5 or 50 ng/ml active BMP-6 recombinant protein (GeneTex, Inc., Irvine, CA, USA) for 48 h. Cell viability was determined using Cell Counting Kit-8 (CCK-8; Dojindo Molecular Technologies, Inc., Rockville, MD, USA), according to the manufacturer's protocol. The plates were read at microplate reader at 450 nm and the optical density value was normalized with the control.

The supernatant of BMP-6 in the culture of all tested cells was measured using a Human BMP-6 ELISA kit (cat. no. ab99984; Abcam), according to the manufacturer's protocol.

The data were analysed using SPSS 18.00 (SPSS, Inc., Chicago, IL, USA) and were presented as the mean ± standard error of the mean. The Mann-Whitney U and Kruskal-Wallis test were used to analyze the data for two groups and multiple groups, respectively. The Nemenyi test was used for a post-hoc test for multiple comparisons. While the Pearson χ² test was used to analyze the data of immunohistochemistry. P<0.05 was considered to indicate a statistically significant difference.

The expression of BMP-6 mRNA in NSCLC and adjacent tissues was detected by RT-qPCR. The level of BMP-6 mRNA in NSCLC tissues was significantly reduced, compared with adjacent normal lung tissues (P<0.05; Fig. 1). There were no significant differences in BMP-6 mRNA expression between NSCLC patients with different tumor types, tumor differentiations, tumor size, clinical stage, and with or without lymph node metastasis (P>0.05; Fig. 2).

Immunohistochemical staining of BMP-6 expression in 73 NSCLC and adjacent normal lung tissues revealed that BMP-6 was positively expressed in a number of lung cancer tissues and negatively expressed in the majority of tumor tissues (Fig. 3). The positive rate of BMP-6 expression in 73 lung cancer tissues was 26.03% (19/73). In normal lung tissue, BMP-6 expression was positively expressed in the majority of the samples. The positive rate of BMP-6 expression in 73 adjacent normal lung tissues was 89.04% (65/73). The expression of BMP-6 was significantly reduced in NSCLC tissues, compared with adjacent lung tissues (χ²=59.32; P<0.001; Table I).

The association of BMP-6 mRNA expression with the overall survival was analyzed using the Kaplan-Meier plotter online database. The database contained a total of 1,926 cases of NSCLC with available data for BMP-6 mRNA expression and overall survival rate. The online analysis using the Kaplan-Meier plotter database calculated a HR value of 0.83, and log-rank P-value of 0.0068, indicating that BMP-6 mRNA level can function as a predictive factor, and that reduced BMP-6 mRNA expression (lower than the median) is associated with a poor prognosis of patients. The Kaplan-Meier survival curve was presented in Fig. 4.

The expression of BMP-6 mRNA in cells (Fig. 5A) and the concentration of BMP-6 protein in the supernatant of cultured cells (Fig. 5B) were measured. The mRNA and protein expression of BMP-6 were significantly decreased in lung cancer cells, compared with that in HBE cells. To further validate the role of BMP6 in lung cancer cells, the H460, H1299, A549 and H520 cells were treated with active BMP6 recombinant protein for 48 h, where it was revealed that 5 or 50 ng/ml BMP-6 protein significantly inhibited cell proliferation in H460 (Fig. 5C), H1299 (Fig. 5D), A549 (Fig. 5E) and H520 (Fig. 5F) cells, compared with the control group. These data indicate that BMP-6 is a suppressor in human lung cancer cells.