LncRNA LINC00261 overexpression suppresses the growth and metastasis of lung cancer via regulating miR-1269a/FOXO1 axis

LC and adjacent normal lung tissue samples were obtained from 78 LC patients who received surgical resection between July 2015 and July 2017 in the Henan Provincial Chest Hospital, and all the samples were stored at − 80 °C. This study was approved by the Ethics Committee of Henan Provincial Chest Hospital. The clinical features of all the participants, including gender, age, TNM stage, lymph node metastasis, histological type, distant metastasis and tumor size, were collected from the hospital. The relationships between LINC00261 expression and clinic features in LC patients were shown in Table 1. The relatively high or low expression level of LINC00261 was defined according to the median level of the expression.

The normal human bronchial epithelial cells (BEAS-2B, CBP60577), LC lines (A549 (CBP60084), NCI-H1299 (CBP60053), NCI-H23 (CBP60132), SPC-A1 (CBP60153)) were purchased from Cobioer Co., Ltd. (China). The LC cell line L78 was obtained from Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences. All the cells were cultured in RPMI-1640 medium (61870044, ThermoFisher, USA) containing 10% FBS (16140071, ThermoFisher, USA) at 37 °C with 5% CO₂.

Chloroform and isopropanol methods were used to isolate total RNAs from the tissues and cells. NanoDrop 2000 (ND-2000-GL, Thermo Scientific, USA) was used to quantify the RNAs. To determine the levels of LINC00261 and FOXO1, reverse-transcription and qRT-PCR were performed using the PrimeScript™ II 1st Strand cDNA Synthesis Kit (6210B, Takara, Japan), SYBR^® Green PCR Master Mix (4312704, ABI, USA) and Bio-Rad CFX 96 Touch Real-Time PCR Detection System (1855196, Bio-Rad, China). GAPDH served as a reference gene. The loop RT primer sequence was 5′-GTCGTATCCAGTGCGTGTCGTGGAGTCGGCAATTGCACTGGATA-CGACCCAGTAGC-3′, and used for detecting the expression of miR-1269a. U6 snRNA served as an internal reference gene. Parameters for qRT-PCR were as follows: at 95 °C for 5 min, 40 cycles at 95 °C for 15 s, at 60 °C for 30 s, and at 70 °C for 10 s. The relative expression was calculated by 2^−ΔΔCt method. All primers for qRT-PCR were shown in Table 2.

For cell transfections, 100 pmol miR-1269a mimic (miR10005923-1-5, Ribobio, China) was added into Opti-MEM medium (31985062, Thermofisher, USA) containing Lipofectamine 2000 (11668019, Thermofisher, USA) and mixed for 20 min at room temperature. Next, the mixture was added into a 6-well cell culture plate to culture the cells (2 × 10⁵ cells/well) at 37 °C with 5% CO₂ for 8 h. Then, the medium was replaced by RPMI-1640 containing 10% FBS. After transfection for 24 h, the cells were used for later detection.

Full-length cDNAs of LINC00261 and FOXO1 (Tsingke Co., Ltd.) were inserted into pCDH-CMV vector (CD513B-1, System Biosciences, USA) and then infected into 293T cells (CBP60439, Cobioer, China) to produce a lentivirus, which was used to infect A549 and SPC-A1 cells (2 × 10⁵ cells/well) in the 6-well plate. After 72 h, the cells were collected to determine the efficiencies of LINC00261 and FOXO1 overexpression. Cells were selected using 2 µg/mL puromycin starting on day 4 after the virus infection. Following assays were carried out 2 weeks after the infection.

After cell incubation, the cells (3000 cells/well) were seeded into a 96-well plate. 10 µL CCK-8 70-CCK801 (MultiSciences, China) was added into each well for 4 h at 37 °C. Then the absorbance value at 490 nm was detected by the SpectraMax plus 384 Microplate Reader (PLUS 384, Molecular Devices, USA). The medium containing only 10 μL CCK-8 reagent served as the blank control, and each group were set up with 9 parallel wells.

The cells (1 × 10² cells/well) were seeded into a 6-well plate to be cultured for 7 d, and the medium was changed every 3 d. After the culture, the cells were fixed by paraformaldehyde at room temperature for 30 s, stained by 0.1% crystal violet solution (548-62-9, Aladdin, China) at room temperature for 30 min, and washed by PBS twice. Next, the cells were observed under a microscope (TS100, Nikon, Japan). Colony formation number and relative colony formation were calculated from 10 randomly selected fields under a microscope.

The cells (2 × 10⁵ cells/well) were grown into a 6-well plate, cultured for 24 h, and then treated by 500 µL Trypsin (25300054, Thermofisher, USA) at 37 °C for 1 min. The cells were collected by centrifuging at 500×g, 4 °C for 5 min.

To determine cell apoptosis, the cells were resuspended in 300 μL Annexin-V kit (70-AP101-100-AVF, MultiSciences, China) binding buffer, and then mixed with 5 μL Annexin-V-FITC solution at RT for 15 min. The nuclei were stained by 5 µL PI for 5 min, and then another 200 μL binding buffer was added into the cells. For cell cycle arrest analysis, the cells were fixed in 70% ethanol at 4 °C overnight and stained by PI/RNase A mixture at 37 °C for 30 min in the dark. Finally, the cell cycle was analyzed by flow cytometer (342973, BD Biosciences, USA). FACSCalibur flow cytometer (342973, BD Biosciences, USA), and BD FACSCanto™ system software v2.4 (646602, BD Biosciences, USA) were used to detect the cell apoptosis and cell cycle.

A wound was created by using a 10 μL pipette tip after the cells were grown to 100% confluence. The cells were then starved and cultured for 48 h. The scratch area was measured from 3 fields in each group under a microscope (TS100, Nikon, Japan) at 0 h and 48 h.

The cells (2 × 10³ cells/well) were seeded in serum-free media contained the upper chamber of transwell plate (3428, Corning, USA) pre-coated with 40 μL Matrigel, while 600 μL normal medium was added into the lower chamber. After incubation for 24 h, the cells invaded from the upper chamber into the lower were fixed by methanol and stained by 0.1% crystal violet (548-62-9, Aladdin, China) at room temperature for 30 min. Then, 10 fields under a microscope were randomly selected to count invaded cells. The experiments were performed in triplicate.

Total proteins were isolated from the cells on ice using the lysis buffer (89901, Thermo Scientific, USA), and the protein concentration was measured using the BCA Protein Assay kit (23225, ThermoFisher, USA). The proteins were separated by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred onto a PVDF membrane. The membrane was then blocked by 5% non-fat milk at RT for 10 min, and exposed to the following primary antibodies: Anti-Ki-67 (359 kD, 1:5000, ab92742, Abcam, UK), Anti-PCNA (29kD, 1:1000, ab18197, Abcam, UK), Anti-MMP-3 (54kD, 1:1000, ab53015, Abcam, UK), MMP-9 (95kD, 1:1000, ab73734, Abcam, UK), Anti-GAPDH (36 kD, 1:5000, ab8245, Abcam, UK) at 4 °C for 12 h. Anti-mouse HRP secondary antibody (1:2000, ab6728, Abcam, UK) was incubated with the membrane at RT for 2 h for detection of GAPDH. Anti-rabbit HRP secondary antibody (1:2000, ab97051, Abcam, UK) was used for detection of other proteins. Then, the membrane was treated by chemiluminescence solution (WBKLS0100, Millipore, USA) at room temperature in the dark for 1 min. The signals were finally developed by using SignalFire™ ECL reagent (6883, CST, USA).

StarBase (http://​starbase.​sysu.​edu.​cn/​) was used to predict the interaction between LINC00261 and miRNA, and the interaction between miR-1269a and its target mRNA was predicted by Targetscan7.2 (http://​www.​targetscan.​org/​vert_​72/​). Potential binding site of LINC00261 (LINC00261-WT, 5′-CAAGAGGCAATGGTCCCAGTCCAA-3′), mutation sequence (LINC00261-MUT, 5′-CATCATCGATGCCTGCGTCAGGTA-3′), potential binding site of FOXO1 3′UTR (FOXO1-WT, 5′-AGGAACTGAGAGAAGCAGTCCAA-3′) and the mutation sequence (FOXO1-MUT, 5′-AGGAACTGAGAGAAGGTCAGGTA-3′) were inserted into dual-luciferase reporter vector (pmirGLO, E1330, Promega, USA). Those reporter plasmids were then transfected into the cells with miR-1269a mimic. Cells transfected with reporter plasmid only served as controls. After culture for 24 h at 37 °C, the luciferase activity of cells was determined by dual-luciferase assay system (E1910, Promega, USA) and Microplate Luminometer (11300010, Berthold, Germany). The firefly luciferase activity was normalized to that of renilla luciferase activity.

We purchased 12 8-week-old male BALB/C nude mice (weighting 20 g–25 g) from Vitalriver Corporation (China). The mice were kept in the Typical SPF Laboratory Animal room at 23 °C and provided with free access to sterilized food and water. The cells (1 × 10⁶) stably infected by pCDH-CMV empty vector (NC) or pCDH-CMV-LINC00261 were resuspended in PBS/matrigel mixture on ice and then subcutaneously injected into the nude mice (with 6 mice in each group). The tumor volumes were measured at day 5, 10, 15, 20, 25 and 30. Finally, all the mice were sacrificed and the tumors were weighted. Animal experiments were approved by the Institutional Animal Care and Use Committee of Henan Provincial Chest Hospital.

FOXO1 primary antibody (1:500; ab70382; Abcam, UK) was added to the cells at 4 °C overnight. Horseradish peroxidase-labeled anti-rabbit IgG H&L secondary antibody (1:2000; ab205718; Abcam, UK) was then added to the cells and incubated at 37 °C for 30 min. DAB developer (Biyuntian Biotechnology Co., Ltd.) was used to develop the brown-yellow positive color. The tumor sections were counterstained by Hematoxylin for 3 min, differentiated with 1% hydrochloric acid alcohol, made transparent by xylene, and fixed with gum seal. The sections were then observed under a light microscope (Olympus, 100× and 200× Magnification) and photographed.

The statistics were analyzed using Graphpad Prism 5.02 software (La Jolla, CA, USA). The data are shown as mean ± standard deviation. Correlations of LINC00261 expression and clinic pathological characters were analyzed by Pearson’s Chi square test. Kaplan–Meier analysis was performed to plot the overall survival curves. The differences among groups were analyzed by one-way ANOVA, followed by Tukey’s post hoc test. Gene expression differences in LC and adjacent normal lung tissues were analyzed by paired t-test. Spearman’s correlation analysis was performed to evaluate the expression correlations among LINC00261, miR-1269a and FOXO1. P < 0.05 was defined as a significant difference.

In this study, we measured the level of LINC00261 in LC tissues and adjacent normal lung tissues. RT-PCR result showed that LINC00261 expression was down-regulated in LC tissues (P < 0.001, Fig. 1a) and in A549, H1299, H23, SPC-A1 and L78 cells, especially in A549 and SPC-A1 cells. Thus, A549 and SPC-A1 cells were used in further study (P < 0.001, Fig. 1c). In addition, the Kaplan–Meier analysis revealed that low expression LINC00261 was associated with a shorter overall survival time (P < 0.05, Fig. 1b). Pearson’s Chi square test indicated that the expression of LINC00261 was correlated with TNM stage (P = 0.039), lymph node metastasis (P = 0.020) and distant metastasis (P = 0.039) in LC patients (Table 1).

To investigate the role of LINC00261 in LC, LINC00261 was constitutively overexpressed in A549 and SPC-A1 cells. QRT-PCR results showed that the expression of LINC00261 in A549 and SPC-A1 cells were increased by LINC00261 overexpression 72 h after the infection (P < 0.001, Fig. 2a, b). Two weeks after the infection, corresponding assays were performed to determine effects of LINC00261 overexpression on cell viability, colony formation, cell apoptosis and cell cycle arrest. CCK-8 assay showed that overexpressed LINC00261 suppressed the viabilities of A549 and SPC-A1 cells (P < 0.05, Fig. 2c, d), and inhibited colony formation of A549 and SPC-A1 cells (P < 0.001, Fig. 2e–g). Ki-67 and PCNA (Proliferating Cell Nuclear Antigen) were cell proliferation-associated markers, and the effects of LINC00261 overexpression on Ki-67 and PCNA expressions were explored by Western blot, the results indicated that LINC00261 overexpression significantly reduced the levels of Ki-67 and PCNA in A549 and SPC-A1 cells (P < 0.001, Fig. 2h–k). Furthermore, flow cytometry demonstrated that LINC00261 overexpression promoted apoptosis (P < 0.001, Fig. 3a–c) and cell cycle arrest in G1 phase (P < 0.001, Fig. 3d–f) of A549 and SPC-A1 cells. Additionally, the migration and invasion of A549 and SPC-A1 cells upon constitutive LINC00261 overexpression were determined, and the results showed that LINC00261 overexpression suppressed the migration and invasion of LC cells (All P < 0.001, Fig. 4a–f). The expressions of metastasis-associated markers (MMP-3 and MMP-9) were also measured by Western blot, and we found that LINC00261 overexpression inhibited the expressions of MMP-3 and MMP-9 in LC cells (All P < 0.001, Fig. 4g–j).

Furthermore, the function of LINC00261 in LC in vivo was explored. The xenograft was established in the mice by injecting A549 or SPC-A1 cells (constitutively overexpressing LINC00261 or empty vector control) via the ventral side into BALB/c nude mice. The tumors in the LINC00261 overexpression group were significantly smaller and lighter than those in NC group (P < 0.001, Fig. 5a–e). IHC showed that the expression levels of FOXO1 were increased in LINC00261 overexpression group (Fig. 5f–i), while RT-qPCR showed that the expression of miR-1269a was significantly decreased in LINC00261 overexpression group (Fig. 5j–k).

StarBase predicted that LINC00261 might function as a sponge for miR-1269a (Fig. 6a). Moreover, dual-luciferase reporter assays demonstrated that miR-1269a mimic decreased luciferase activities of A549 and SPC-A1 cells transfected with LINC00261-WT luciferase reporter plasmid, while the luciferase activities of A549 and SPC-A1 cells transfected with LINC00261-WT luciferase did not change (Fig. 6b–c). Furthermore, Targetscan7.2 predicted a potential binding site of miR-1269a at position 65–72 of FOXO1 3′UTR (Fig. 6d), and dual-luciferase reporter assays confirmed that miR-1269a mimic suppressed the luciferase activity of FOXO1-WT reporter plasmid, but it did not change the luciferase activity of FOXO1-MUT reporter plasmid (Fig. 6e–f).

QRT-PCR and spearman correlation analysis were further performed to determine the expressions of LINC00261, miR-1269a and FOXO1 in LC. The results showed that LINC00261 overexpression reduced the expression of miR-1269a but increased that of FOXO1 in LC cells (Fig. 7a, b). The expression level of miR-1269a was up-regulated in LC tissues, but that of FOXO1 was down-regulated in LC tissues (Fig. 7c, d). The spearman correlation analysis demonstrated that the expression of LINC00261 was negatively correlated with miR-1269a (r = − 0.595, P < 0.001), similarly, the expression of FOXO1 was negatively correlated with miR-1269a (r = − 0.558, P < 0.001), and the expression of LINC00261 was negatively correlated with FOXO1 (r = − 0.442, P < 0.001).

To confirm the regulative relation between miR-1269a and FOXO1 in LC, A549 and SPC-A1 cells were co-transfected with FOXO1 and miR-1269a, and the results were compared with those transfected with FOXO1, miR-1269a or empty vector. The expression of FOXO1 was increased by FOXO1 overexpression but reduced by miR-1269a mimic, and co-transfection with FOXO1 and miR-1269a could reverse these effects (P < 0.001, Fig. 8a, b). Furthermore, the viability, colony formation, apoptosis, cell cycle, migration and invasion of A549 and SPC-A1 cells were determined. The results indicated that miR-1269a promoted cell growth, migration and invasion and suppressed apoptosis and cell cycle arrest, but FOXO1 overexpression reversed effects on LC cells. More importantly, co-incubation of FOXO1 overexpression plasmid and miR-1269a mimic could neutralize the effects of FOXO1 and miR-1269a (Figs. 8, 9, 10).