Down expression of lnc-BMP1-1 decreases that of Caveolin-1 is associated with the lung cancer susceptibility and cigarette smoking history

Lnc-BMP1-1 is down-regulated in lung cancer patients and is associated with cigarette smoke history

The clinical characteristics of the 276 patients with LC included in this study has been published in our previous studies conducted with the same subjects [39], another 17 patients were newly enrolled, the clinical characteristic of the study subjects are shown in Supplementary Table 2. As shown in Figure 1A, lnc-BMP1-1 is down-regulated in lung cancer tissues, when comparing to their adjacent normal tissues (P=8.514×10^-7). Therefore, we considered that lnc-BMP1-1 is associated with lung cancer development and might exert tumor suppressor function.

Figure 1. The decrease of lnc-BMP1-1, Cav-1 expression in lung cancer population and the expression of Inc-BMP1-1 in lung cancer cell lines. (A) The expression of lnc-BMP1-1 was decreased in lung cancer patients; (B) The expression of Cav-1 was decreased in lung cancer patients; (C) The relationship between lnc-BMP1-11 and Cav-1 expression; (D) The expression of lnc-BMP1-1 in types of lung cancer cells. RNA expression was analyzed using RT-PCR, the RNA levels were normalized against β-actin mRNA. Each bar represents the means ± SD of three independent experiments.

The clinical data of these subjects are shown in Table 1. The smoking history is associated with the expression of lnc-BMP1-1 (P=0.027). Patients with lower expression level of lnc-BMP1-1 are more likely to be cigarette smokers or the ever ones, suggesting a possible association of lnc-BMP1-1 and cigarette smoking-related genes. Lnc-BMP1-1 is also associated with distant metastasis, though it isn’t a protection factor (P=0.002).

Effect of lnc-BMP1-1 on cell proliferation and migration

According to the relationship between SFTPC and adenocarcinoma, with qrt-PCR detection (Figure 1D), we selected A549, GLC82 and PC9 adenocarcinoma cell types to further examine the lnc-BMP1-1 function. Through observation with fluorescence microscope, we found that the transfection efficiency of lnc-BMP1-1 in these three types of cells are above 80%; compared to that of the control cells, the fold change of lnc-BMP1-1 expression in ‘A549-BMP’, ‘GLC82-BMP’, ‘PC9-BMP’ cells are about 700, 3000, 4000 times, respectively (Supplementary Figure 1).

Through detecting the cell viability, we found that the proliferation of A549-BMP cells was inhibited by the over-expressed lnc-BMP1-1 when compared to A549-NC cells (Figure 2A), the proliferation of GLC82-BMP cells had also been inhibited when compared to GLC82-NC cells (Figure 2C); however, the inhibition on the proliferation of PC9-BMP cells is not obvious (Figure 2B).

Figure 2. The cell growth or migration of lung cancer cells with lnc-BMP1-1 over-expression vs. NC cells. (A) The cell proliferation of A549-BMP vs. A549-NC, (B) PC9-BMP vs. PC9-NC, (C) GLC82-BMP vs. GLC82-NC, respectively; (D) Comparison of the wound widths (40x) of lung cancer cells with Inc-BMP1-1 over-expression vs. NC cells after 72 hrs, the relative bar graph are A549-BMP vs. A549-NC(E), PC9-BMP vs. PC9-NC(F), GLC82-BMP vs. GLC82-NC(G); the migration capacity of both A549-BMP and GLC82BMP cells were reduced; however, there's no significant difference of PC9-BMP vs. PC9-NC cells; with transwell migration experiments (40 x), the cell migration capacity of A549-BMP vs. A549-NC was proved to be decreased again (H), the relative bar graph was A549-BMP vs. A549-NC(I), PC9-BMP vs. PC9-NC(J), GLC82-BMP vs. GLC82NC (K), respectively. Data are represented as means ± SD. *P < 0.05, ns P> 0.05

With Cell Wound Scratch (Figure 2D) and Transwell (migration) assay (Figure 2H), we found that over-expression of lnc-BMP1-1 could inhibit the migration of A549-BMP cells when compared to A549-NC cells. But no significant difference was observed in GLC82-BMP and PC9-BMP cells, compared to their GLC82-NC and PC9-NC cells, respectively.

The Bcl-2 protein level is decreased by over-expressed lnc-BMP1-1

Bcl-2 protein level is a most frequently used indicator for apoptosis resistance of cancer cells [40–42]. Compared to that of the A549-NC cells, the Bcl-2 protein level in A549-BMP cells is decreased by over-expressed lnc-BMP1-1 (Figure 3A), indicating a decline of anti-apoptosis capacity of the A549-BMP cells, which also means that the over-expressed lnc-BMP1-1 is helpful in improving the malignant phenotype of A549 cells, a result in accordance to the migration experiment results above. The protein level of Bcl-2 increased slightly in PC9-BMP cells, but no change was observed in the above migration experiment. We have thus concluded that lnc-BMP1-1 has a different regulation mechanism in PC9 cells than in A549 and GLC-82 cells. Therefore, the PC9 cells are ruled out from our further experiments.

Figure 3. The RNA or protein expression changes of lung cancer cells with lnc-BMPl-1 over-expression vs. NC cells, the CSE vs. DMSO treated 16HBE cells, and 5-AzaC or TSA treatment. (A) The protein expression of Bcl-2, Cav-1 and DNMT3a in different lung cancer cell lines with lnc-BMP1-1 over-expression vs. NC cells; the protein expression were detected using western blotting analyses, protein expression was normalized against GAPDH or α-tubulin protein; the value above each band indicates the fold change of protein expression relative to their control; (B) The cell viability of CSE (vs. DMSO) treated 16HBE cells at 5^th passage; (C) The cell viability of CSE (vs. DMSO) treated 16HBE cells at 10^th passage; (D) The expression of lnc-BMP1-1, Cav-1 were reduced in CSE treated 16HBE cells both in 5^th and 10^th passage, and made greater decline in 10^th than in 5^th passage (the same gene was compared in 5^th and 10^th generation of CSE treated 16HBE cells, #P<0.05, ##P<0.01). The mRNA expression of Cav-1 in A549-BMP vs. A549-NC (E), GLC82-BMP vs. GLC82-NC(F), respectively; In A549 cells treated with TSA, the expression of lnc-BMP1-1 (G) and Cav-7 (H) were increased, while DNMT3a (I) was decreased; The protein expression of HDAC2 (J) was decreased in A549-BMP cells. Data are represented as means ± SD. *P< 0.05, **P<0.01.

Cav-1 is probably the target gene of lnc-BMP1-1

To clarify how lnc-BMP1-1 functions in lung cancer development, we first tried to find the downstream genes that are regulated by it. Based on the systematical analysis, three genes are taken into consideration, they are ‘ATP-binding cassette, sub-family A (ABC1), member 3 (ABCA3)’, ‘Caveolin-1 (Cav-1)’, and ‘natriuretic peptide receptor A (NPR1)’. Combining bio-information and literature analysis, NPR1 and Cav-1 were suggested as target genes regulated by lnc-BMP1-1. The expression of NPR1 and Cav-1 were detected in tissue samples by qrt-PCR.

Only the expression of Cav-1 is significantly decreased in lung cancer population, and positively correlated with that of lnc-BMP1-1 (R=0.211, P < 0.001) (Figure 1B and 1C). In 293 pairs of lung cancer tissues and their adjacent normal tissues (Table 2), Cav-1 is also found to be associated with smoking (P = 0.011).

The expression of lnc-BMP1-1 and Cav-1 in 16HBE cells treated with cigarette smoke extract (CSE)

The expression of lnc-BMP1-1 in 16HBE cells treated with cigarette smoke extract (CSE) was shown, including the 5^th and 10^th generations of 16HBE cells. As to Figure 3B and 3C, be it DMSO or CSE treatment, the cell viability alteration was not obvious. In comparison to that in the control cells (DMSO treated only), the lnc-BMP1-1 expression in 5^th and 10^th generation cells is decreased, indicating the sensitivity of lnc-BMP1-1 to cigarette smoke. Notably, the expression of lnc-BMP1-1 expression in 10^th generation is undetectable. Meanwhile, the Cav-1 expression is also decreased along with the decreased of lnc-BMP1-1 expression. Both the expression of lnc-BMP1-1 and Cav-1 showed greater decline in the 10^th than in the 5^th generation, these results further reinforce our viewpoint that the expression of lnc-BMP1-1 and Cav-1 in human bronchial epithelial cells might be decreased because of the exposure to cigarette smoking (Figure 3D).

The mRNA and protein level of Cav-1 are increased by over-expressed lnc-BMP1-1

When compared A549-BMP to A549-NC cells, both the mRNA and protein level of Cav-1 are increased as a result of the over-expression of lnc-BMP1-1 (Figure 3E and 3A). Meanwhile, no alteration of Cav-1 was observed in GLC-82 cell (Figure 3F and 3A), both on the mRNA and the protein level. It is possible that Cav-1 is not regulated by lnc-BMP1-1 in the GLC-82 cell line. Therefore, we select the A549 cell line to further explore the transcription impact that lnc-BMP1-1 has on Cav-1.

The tumor growth in nude mice of A549 cells is inhibited by the over-expression of lnc-BMP1-1

21 days after injection, the tumor in nude mice injected with A549-NC cells (A549-NC nude mice) had grown larger in size than that in nude mice injected with A549-BMP cells (A549-BMP nude mice). At day 28 (Figure 4A), the average weight of tumors in the A549-BMP nude mice group is 0.35g ± 0.17g, while it is 0.48g ± 0.55g in the A549-NC nude mice group; the average size of tumors in the A549-BMP nude mice group is 0.40 ± 0.28 cm³, while it is 1.07 ± 1.15 cm³ in theA549-NC nude mice group (Figure 4B). The represented photo of H & E staining of the tumor tissues of nude mice is shown in Figure 4C. The above data show the inhibition of tumor growth by the over-expression of lnc-BMP1-1 in subcutaneous tumorigenesis of nude mice.

Figure 4. The tumor growth in nude mice and the results of anti-cancer drug trial in A549-BMP vs. A549-NC cells. (A) Tumor of A549-BMP vs A549-NC cells; (B) Tumor volume and tumor weight; (C) The HE staining of nude mice tumor tissues; (D) The drug sensitivity to doxorubicin hydrochloride was enhanced in A549-BMP vs A549-NC cells. Data are represented as means ± SD. *P< 0.05, **P<0.01.

Explore the mechanism of lnc-BMP1-1 in regulating Cav-1

To understand in which epigenetic regulatory manner lnc-BMP1-1 exerts its function and what are the relevant molecules, A549 cells were treated with 5-AzaC and TSA, respectively. As a result the expression of lnc-BMP1-1 (Figure 3G), Cav-1 mRNA (Figure 3H) was increased, while that of DNMT3a was decreased (Figure 3I), but not that of DNMT3b and DNMT1 (supplementary Figure 2A and 2B). As shown in Figure 3G–3I, we found that histo-acetylation modification might be the dominant mechanism through which lnc-BMP1-1 exerts function. In response to the TSA treatment, the expression of lnc-BMP1-1 and Cav-1 are increased significantly. We’ve carried out the detection of the methylation status of the promoter of Cav-1, and no difference was found between that in A549-BMP and in A549-NC cells (supplementary Figure 2C), even though there is obvious change in DNMT3a protein expression (Figure 3A). The above results again proved that histo-acetylation regulation pathway is more dominant than the DNA methylation regulation.

As to Figure3J, the HDAC2 protein is decreased in A549-BMP cells, which might be responsible for the transcription promotion of Cav-1 by lnc-BMP1-1. We have previously found that TSA, an HDACIs, could increase the expression of both lnc-BMP1-1 and Cav-1 (Figure 3G and 3H), indicating a possible interaction of lnc-BMP1-1 and HDAC2. To some degree, we might conclude that the inhibited transcription of Cav-1 resulted from decreased histo-acetylation level caused by decrease of lnc-BMP1-1.

The enhancement of the drug sensitivity of Doxorubicin hydrochloride (DOX) by over-expressed lnc-BMP1-1 in A549 cells

As to Figure 4D, at the concentration of 50nM of Doxorubicin hydrochloride (DOX), the cell viability was 37% of A549-BMP cells when compared to 50% of A549-NC cells. Over-expressed lnc-BMP1-1 markedly enhanced the drug sensitivity of Dox to A549 cells. However, there’s no improvement in sensitivity to Cisplatin of A549 cells resulting from the over-expression of lnc-BMP1-1(Supplementary Figure 2D).

Study subjects

All study subjects selected in this study were Han Chinese originated from Southern or Eastern China. A total of 293 samples of cancer tissues and paired normal tissues were collected in this study. Among the 293 samples, 199 pairs of samples were collected between 2008 and 2015 from the Cancer Hospital Affiliated with Guangzhou Medical University, and1^st Affiliated Hospital and Cancer Hospital Affiliated with Kunming University, and 94 pairs of samples were collected between 2007 and 2016 from the 1^st Affiliated Hospital of Soochow University. The patients in the study have no genetic connections with one another. The study was approved by the Research Ethics Committee of Guangzhou Medical University (No. GMU201481473040) and followed clinical research guidelines. Written informed consents were obtained from all patients who participated in this study.

qRT-PCR

Total cellular RNA was isolated with TRIzol (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. To detect the RNA expression of related genes, we first synthesised the strand cDNA using the RevertAidTM First Strand cDNA Synthesis Kit (Thermo Scientific, Carlsbad, CA, USA). Next, qRT-PCR was performed using the FastStart Universal SYBR Green Master kit (Roche, Mannheim, Germany) on an ABI7900HT PCR instrument with specific primers (shown in the Supplementary Table 1). All primers were synthesised by Invitrogen Ltd. and listed in Supplementary Table 1. β-actin was used as an internal control. The delta-delta CT method was used to quantify gene expression using the 7900 System SDS Software according to the recommended protocol.

Cell culture

The human lung cancer cell lines (A549, GCL82, and PC9) and a normal bronchial epithelial cell line (16HBE) were purchased from the Cell Bank of Type Culture Collection of the Chinese Academy of Science, Shanghai Institute of Cell Biology. The A549, GCL82, and PC9 cell lines were cultured in RPMI-1640 medium (Gibco, life technologies, California, USA) with 10% fetal bovine serum (FBS) and penicillin (100UI/ mL)/streptomycin (100mg/mL) in standard conditions with 5% carbon dioxide at 37.

Cigarette smoke extract (CSE) preparation and acute exposure

Commercial normal cigarettes (Double Happiness, Guangzhou Tobacco, Guangzhou, China) yield 15mg of tar and 1.4mg of nicotine under a standard smoking regimen. Cigarette burning-produced puffs were collected using an atmosphere sampler with a glass accessory containing 100% Dimethyl sulfoxide (DMSO, AR), then purified water was added into the compound, forming a 2% diluent, termed as cigarette smoke extract (CSE). 16HBE cells were treated with 10mL CSE for 24h and made a rest for another 24h; the cells grown every 3 days are considered one generation. The 5^th and 10^th generation cells were employed for the next detection.

Cell transfection

To understand the function of lnc-BMP1-1 in lung cancer, the human lung cancer adenocarcinoma cell lines A549, GLC82 and PC9 were transfected with over-expressed lnc-BMP1-1 lentivirus vectors. The cDNA sequence of lnc-BMP1-1 was obtained from Lncipedia database (version4.0), the coding fragment of lnc-BMP1-1 was synthesised by PCR and then inserted into the pLVX-shRNA1 vector (Clontech Corporation, Tokyo, Japan) using the BamH I and EcoR I restriction sites.

The over-expression of lnc-BMP1-1 in lung adenocarcinoma cell line A549, GLC82 and PC9 were denoted as ‘A549-BMP’, ‘GLC82-BMP’, ‘PC9-BMP’ and their corresponding empty vector control cells were denoted as ‘A549-NC’, ‘GLC82-NC’, ‘PC9-NC’ in the following text.

Cell proliferation

Cell growth was measured by cell counting with the CCK-8 cell proliferation assay kit (Dojindo, Tokyo, Japan). In a 96-well plate, 1000 cells in 100 μL serum-free RPMI-1640 medium were plated per well. The plates were incubated at 37°C for 24 h, and 10 μL CCK-8 was added to each well. The cells were cultured for an additional 4 h at 37°C. With a reference wavelength of ‘650 nm’, the absorbance wavelength of the sample was 450 nm.

Cell migration (Wound scratch and Trans-well migration)

Wound scratch method of the lentivirus-vectors-transfected A549, GLC82 and PC-9 cell lines were carried out to understand the contribution of lnc-BMP1-1 to the cell migration capacity. Transfected cells were seeded on 24-well plates and then grown to confluence. The confluent monolayers were scratched with a pipette tip and maintained under certain conditions according to the protocol. Plates were washed once with fresh medium to remove non-adherent cells and then photographed. The percentage of open spaces covered by migrated cells was calculated with ImageJ software.

Trans-well migration assay of the lentivirus-vectors-transfected A549, GLC82 and PC-9 lung adenocarcinoma cell lines was conducted in an 8μm Transwell chamber (Costar, Corning Incorporated, NY). Cells grew to near confluence in a 5×5 cm culture vessel and then were placed in serum-free RPMI-1640 medium (Gibco) for 24 h. Cells were trypsinized and re-suspended in serum-free RPMI-1640 medium with 0.1% BSA. We seeded 2×10⁴ cells in the chamber, which was put in a 24-well plate containing RPMI-1640 with 20% FBS (Gibco). After 48 h of culture, non-migrating cells in the top section of the chambers were removed by swabbing. Migrated cells were counted manually in 10 random fields and their numbers were averaged.

Xenograft tumor formation assay in nude mice

With A549-BMP and A549-NC cells, nude mice tumor transplantation experiments were employed to study the influence of lnc-BMP1-1 on tumor growth inhibition in vivo. Five-week-old BALB/c nude mice were provided by Guangdong Medical Animal experiment Centre (Foshan, China). Transformed and control cells were washed twice and suspended with serum-free RPMI-1640 medium. Then, 5×10⁶ cells were resuspended in culture medium and injected subcutaneously into the neck pads of nude mice. The mice were housed in a pathogen-free environment and monitored every 2 days for tumor formation. All mice were sacrificed after 28 days, and the tumors were removed, weighed and fixed in 10% buffered formalin for pathological examination. This study was approved by the Ethics Committee of Guangzhou Medical University (NO. GZYDW201503512).

5-AzaC and TSA treatment

DNA methylation and histo-acetylation modification are two important mechanisms in epigenetic regulatory. To clarify through which epigenetical regulation mechanism lnc-BMP1-1 functions on Cav-1, A549 cells were treated with 5-AzaC (5 μmol/L) or TSA (100 nmol/L) for 24 h, respectively. Then the expression of lnc-BMP1-1, Cav-1, and DNMTs were compared with that of untreated cells.

5-AzaC (Sigma) was dissolved in DMSO and stored at -20°C. TSA (Sigma) was dissolved in absolute ethanol at a concentration of 4 mM, shielded from light and stored at -20°C. Exponentially grown cells were treated with 5-AzaC (5 μmol/L) or TSA (100 nmol/L) for 24 h.

Methyl-specific PCR (MSP)

MSP was carried out to analyze the DNA methylation level of Cav-1 promoter. Methylation-specific PCR (MSP) was performed as previously described [37]. Briefly, genomic DNA was extracted as recommended by the EZNA-DNA kit (Omega). Next, 1 μg of purified DNA was subjected to bisulfite modification using a CpGenome DNA Modification Kit (Chemicon International). The bisulfite-treated DNA was stored at -80 until further used, MSP performed was performed using bisulfite-treated DNA, methylation primers and non-methylation primers. The primer sets used in this study were as follows: 5′-TTATTTCGAAGCGTTTGGGAG-3′ and 5′-AACACTCGTTTACATCTAATCG-3′ for the methylated reaction, and 5′-TTATTTTGAAGTGTTTGGGAG-3′ and 5′-AACACTCATTTACATCTAATCA-3′ for the unmethylated reaction [55].

Western blotting

Total protein was extracted with cell lysis buffer (Cell Signaling Technology, Beverly, MA, USA) according to the manufacturer’s instructions, and the protein concentration was determined using the Bicinchoninic Acid (BCA) assay kit (Kangwei technology, Beijing, China). Western blotting analyses were performed as previously described [38]. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH), α-tubulin primary antibodies were purchased from Cell Signaling Technology (Beverly, MA, USA); Cav-1, DNMT3a and Bcl-2 primary antibodies were purchased from Abcam (Cambridge, MA, USA). The horseradish peroxidase (HRP)-labelled goat anti-mouse and goat anti-rabbit IgG secondary antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA).

Hematoxylin and eosin staining of tumor tissues of nude mice

The tumor or adjacent normal tissues of nude mice were stained with Harris’ hematoxylin solution for 6 h at a temperature of 60~70 °C and were then rinsed in tap water until the water was colorless. Then 10% acetic acid and 85% ethanol in water were used to differentiate the tissue 2 times for 2 h and 10 h, respectively, and the tissues were rinsed with tap water. In the bluing step, we soaked the tissue in saturated lithium carbonate solution for 12 h and then rinsed it with tap water. Finally, staining was performed with eosin Y ethanol solution for 48 h.

Anti-cancer drug sensitivity

To understand whether lnc-BMP1-1 could enhance the sensitivity of A549 cells to anti-cancer drug, including Cisplatin (DDP, 200~600μg/mL) and Doxorubicin hydrochloride (DOX, 10~50 nM). An observation was made every 24h and cell viability was checked using CCK8 kit at the end point of drug treatment.

Statistical analysis

All tests were finished with SPSS19.0. Paired t-test was used to analyze the different expression levels of genes in lung cancer tissues and adjacent non-cancerous tissues. Linear regression models were used to analyze the correlation between lnc-BMP1-1 and Cav-1. Chi-square test and Logistic regression model were used to analyze the expression of lnc-BMP1-1 and clinical characteristics of lung cancer patients. Independent t-test was used to analyze the cell migration capability. Repetitive measurement deviation analysis was used to analyze the difference in cell proliferation. All statistical tests were two sides, P < 0.05 was considered statistically significant.