Introduction

Lung cancer is one of major severe cancers with high incidence and mortality which has been listed as one leading cause of cancer-related deaths worldwide during the past decades [1, 2]. The development of lung cancer is reported to be associated with smoking, exposure to carcinogens, persistent pulmonary inflammation, and also susceptibility due to genetic mutations [3,4,5], however, the molecular pathogenesis underlying lung cancer development is still far from being fully understood. Pathologically, lung cancers are mainly classified into two subtypes including the small-cell lung carcinoma (SCLC) and non-small-cell lung carcinoma (NSCLC) [3]. NSCLC is traditionally termed due to its histological and cytological features significantly different from SCLC, and epidemiology reports show that NSCLC accounts for more than 80% lung cancer cases worldwide [3, 6]. Many chemotherapy drugs such as Cisplatin (diamminedichloroplatinum, DDP) and Vinorelbine are widely applied for NSCLC treatments, however, their efficacy are greatly impaired by the frequent drug resistances which usually lead to poor responses and high recurrence rates [7, 8]. The disclosure of molecular mechanisms underlying chemotherapy resistance are critical for surmounting chemotherapy resistance and developing novel drugs.

The resistance of NSCLC cells to chemotherapies are reported to be a complex cellular processes associated with numerous molecular events [7, 8]. Previous investigations reveal that acquired resistance of NSCLC cells to Cisplatin treatment is caused by the alteration of several DNA repair pathways such as the nucleotide excision repair (NER) pathway and homologous recombination (HR) pathway, which effectively modulate the sensitivity of NSCLC cells to Cisplatin [9]. Moreover, the resistance of NSCLC cell line A549 to Cisplatin treatment has been closely linked with suppressed G2/M cell cycle arrest and apoptosis, caused by altered expression of signaling proteins such as P21, mouse double minute 2 homolog and xeroderma pigmentosum complementation group C [10]. In addition, elevations of the Janus kinase 2 (JAK2) and signal transducers and activators of transcription 3 (STAT3) also contribute to the acquired resistance of A549 to Cisplatin treatment, and inhibition of the JAK/STAT pathway activation is shown to sensitize A549 cells to Cisplatin treatment [11]. The Cisplatin resistance of NSCLC also involves altered expression of many other signaling components such as PI3K/Akt pathway and TGF-β receptor 2 pathway [11, 12]. However, the molecular mechanisms driving the abnormal expressional alterations of these key signaling pathway genes associated with Cisplatin resistance in NSCLC cells remains poorly understood.

Circular RNAs (circRNAs) are a large group of non-coding RNA molecules which are newly characterized to modulate functional gene expression through sponging miRNA and interacting with RNA binding protein [13, 14]. Recent reports have established circRNAs as key regulators of various biological processes and pathogenic conditions such as NSCLC development [15, 16]. For instance, the circRNA hsa_circ_0007385 expression is reported to be significantly elevated in both NSCLC tissue and cell lines, which can act as an novel oncogene during NSCLC development and progression potentially through targeting the miR-181 expression [17]. More importantly, circRNAs are also significantly involved in the acquirement of multiple resistances of cancer cells to chemotherapies [18,19,20]. Has_circ_0006528 may play a role in breast cancer chemoresistance [18]. Hsa_circ_0000199 originating from the AKT3 gene enhances the Cisplatin resistance of gastric cancer cells by suppressing miR-198 and promoting PIK3R1 expression [21]. In addition, the high expression of circRNA PVT1 is reported to promote the resistances of osteosarcoma cells to Doxorubicin and Cisplatin treatments by regulating the expression of ATP-binding cassette subfamily B (ABCB1) gene [20]. Nevertheless, the expression of circRNAs which might contribute to the Cisplatin resistance of NSCLC cells are still largely unexplored.

In this study, differentially expressed circRNA profiles between NSCLC cell line A549 and its Cisplatin-resistant counterpart A549 DDP were characterized in large scale by deep sequencing. Moreover, the potential roles of hsa_circ_0096157 which was derived from the well-known long noncoding RNA-metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) was further investigated. MALAT1 has been previously established as one essential regulator of Cisplatin resistance in lung cancer cells by modulating cell proliferation, apoptosis and epithelial to mesenchymal transition [21, 22]. The elucidation of circRNA profile alteration and MALAT1-derived hsa_circ_0096157 in this study would provide novel insights into the potential functions of circRNAs in chemotherapy resistance of lung cancers.

Material and methods

Cell culture and reagents

The non-small-cell lung carcinoma A549 (#CC0202) and Cisplatin-resistant A549/DDP cells were purchased from the CELLCOOK company (Guangzhou, China) and cultured in the F12K medium (#N3520; Sigma-Aldrich) supplemented with 2.5 g/L NaHCO3 and 10% fetal bovine serum (FBS; Thermo Fisher Scientific) at 37 °C in a humified culture cabinet with supply of 5% CO2. A549/DDP cells were cultured in Cisplatin-free F12K medium under normal conditions for one week upon cell recovery, followed by culture in F12K medium containing 1 μg/ml DDP (Sigma) until cell confluency reached 90–95%. After cell passage, A549/DDP cells were first cultured in Cisplatin-free F12K medium until adherent to culture bottle and then maintained in F12K medium containing 2 μg/ml DDP.

CCK-8 and cisplatin IC50 assays

The half-maximal inhibitory concentration (IC50) of Cisplatin in A549 and A549/DDP cells were determined by measuring the cell viabilities using CCK-8 kit after treatments with 0, 2, 4, 8, 16, 32 and 64 μg/ml DDP for 24 h. The viabilities of A549 and A549/DDP cells were evaluated using the CCK-8 kit (#C0038; Beyotime, Beijing, China) as instructed by the manufacturer. Briefly, NSCLC cells were seeded in 96-well plates and cultured in F12K medium containing DDP at designated concentrations for 24 h, followed by incubation with the CCK-8 reagent for another 24 h at 37 °C. Cell viabilities were finally determined by measuring the absorbance at 450 nm (OD450) on a multiplex plate reader.

CircRNA and mRNA profiling and bioinformatics

To investigate the circRNA profile differences between A549 and A549/DDP cells, the total RNA samples were separately extracted from cultured A549 and A549/DDP cells using TRIzol reagent (Thermo Fishier Scientific; USA) according to the manufacturer’s instructions. CircRNAs profiles were sequenced and quantitated as previously introduced with minor modifications [23]. Briefly, RNA samples were first analyzed by the agarose gel electrophoresis and Agilent 2100 Bioanalyzer for determination of RNA integrity and size distribution, followed by removal of rRNA components using the Qiagen RiboMinus Eukayote Kit following the manufacturer’s instructions. The resultant RNA samples were then used for construction of RNA library using the NEBNext® Ultra™ II RNA Library Prep Kit (#E7770S; New England Biolabs, Beverly, USA) as instructed by the manufacturer, which were finally sequenced with a Hiseq 2000 system (Illumina, USA).

Clean reads were then subjected to alignment with the reference genome database via the Bowtie2 software [24], and read junctions were selected using the back-splice algorithm. CircRNAs were predicted and annotated with the de novo circRNA identification software CIRI [25]. Relative expressional levels of circRNAs between A549 and A549/DDP cells were assessed by the Mapped backsplicing junction reads per million mapped reads (RPM) through normalization to numbers of total reads. Significantly differentially expressed circRNAs were defined by an FDR (false discovery rate) of ≤ 0.05 and a log2Ratio of ≥ 1. The differentially expressed mRNAs between A549 and A549/DDP cells were also analyzed by deep sequencing as previously described [26]. The R software package (Version 1.0.8) was used to establish the hierarchical clustering, scatter plot and volcano plot of differentially expressed circRNAs and mRNAs. Gene ontology (GO) categorization of differentially expressed circRNAs were performed using the DAVID (Database for Annotation, Visualization and Integrated Discovery) website. Signaling pathways with significant enrichment of differentially expressed genes were analyzed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) method (https://www.genome.jp/kegg/). The interaction network between circRNAs and mRNAs were established with the cytoscape software (V. 3.2.1).

RT-PCR and quantitative RT-PCR

The circRNAs were confirmed in this study by RT-PCR method using opposite-directed and divergent primers, and differential expressions of circRNAs were validated by quantitative RT-PCR method. Briefly, total RNA samples were extracted from cultured NSCLC cells as introduced above. The cDNA library was established using the Prime Script RT Master Mix kit (#RR036A; Takara) following the manufacturer’s instructions. To validate the circular morphology of circRNAs, the divergent and corresponding opposite-directed primer pairs were designed for detection of circRNA and linear genomic DNA, respectively. GAPDH was also detected as the control. For analysis of circRNA expressional levels, the quantitative RT-PCR (qRT-PCR) was performed using the 2 × EasyTaq PCR SuperMix kit (#AS111-03; Transgen Biotech, Beijing, China) according to the manufacturer’s instructions. Primers used for PCR in this study were listed in Table 1.

Table 1 Primers used for RT-PCR and qRT-PCR assays
Full size table

Cell transfection

For knockdown of hsa_circ_0096157 expression in A549/DDP cells, siRNA sense (5′-GAGGGGTGAGGTGGGAGATTA-3′) and antisense (5′-TAATCTCCCACCTCACC.

CCTC-3′) were synthesized by the GenePharma company (Shanghai, China). Also, the DNA sequences coding the hsa_circ_0096157 were amplified and ligated with the LV003 plasmids for overexpressing hsa_circ_0096157 in A549 cells. NSCLC cells seeded and cultured 6-well plates for 24 h were then transfected with above siRNA sequences or recombinant circRNA-overexpressing plasmids using the Lipofectamine ™ 2000 transfection system (Invitrogen, USA) following the manufacturer’s instructions. 48 h after transfection, the expression of hsa_circ_0096157 in NSCLC cells were validated by quantitative RT-PCR method.

Cell proliferation

The proliferation rates of NSCLC cells were analyzed by the MTS method using the CellTiter 96® AQueous Non-Radioactive Cell Proliferation Assay (MTS) (#G5421) provided by the Promega Biotech Co., Ltd. (Beijing) following the manufacturer’s instructions. Briefly, NSCLC cells after specified transfections were then seeded in a 96-well plate (100 μl/well). Cells were incubated with 20 μl combined MTS/PMS solution for 2 h at 37 °C. Cell proliferation was finally evaluated by measuring the absorbance at 492 nm (OD492) using a plate reader.

Cell migration and invasion

The Transwell systems (Corning) were used to detect the migration and invasion capacities of NSCLC cells. Briefly, NSCLC cells cultured in serum-free F12K medium were seeded in the upper chamber of Transwell system, of which the lower chamber were filled with F12K medium containing 10% FBS. After culture under normal conditions for 24 h, NSCLC cell migrated into the lower chambers were then stained with crystal violet and observed for assessment of NSCLC cell migration capacity. For analysis of cell invasion, NSCLC cells cultured in serum-free medium were loaded in upper chamber of Transwell system with Matrigel-loaded inner sides. Cells invaded into the lower chambers 24 h later were counted for evaluation of cell invasion capacity. Image J software was used for semi-quantitative analysis of the cell number of migration and invasion.

Cell cycle progression

The progression of cell cycle in NSCLC cells after transfection were determined by flow cytometry combined with Propidium Iodide Flow Cytometry Kit (#ab139418; Abcam). Briefly, NSCLC cells were washed 5 min with PBS for three times, harvested by centrifuge at 1000×g for 10 min and then fixed with 70% ethanol for 24 h at 4 °C. After incubation with 200 µl 1X Propidium Iodide (PI) + RNase Staining Solution at 37 °C in darkness for 30 min, the cell cycle progression of NSCLC cells was finally quantitated by flow cytometry.

Cell apoptosis

The apoptosis of NSCLC cells was quantitatively analyzed by flow cytometry (#V13242; Thermo Fishier Scientific) using the Dead Cell Apoptosis Kit with Annexin V-FITC and PI following the manufacturer’s instructions. Briefly, NSCLC cells were washed with PBS, resuspended in cells in 1X Annexin-binding buffer and then incubated with 5 μl Annexin V-FITC and 1 μl 100 μg/ml PI working solution for 15 min at room temperature. The percentages of apoptotic NSCLC cells were finally measured by a flow cytometer. While, TUNEL assay also was performed for cell death detection following the manufacturer's instructions [27]. Apoptosis stained sections were examined by a fluorescence microscope (Carl Zeiss, Tokyo, Japan), and images of 5 random and non-overlapping fields were examined at an objective magnification of × 200 for analysis.

Western blot

Total proteins were extracted from NSCLC cells using the Western and IP Cell Lysis Buffer (#P0013J; Beyotime, Beijing, China) as instructed by the manufacturer. Approximate 30 μg protein were boiled at 100 °C for 5 min, separated by SDS-PAGE and transferred onto PVDF membrane (Millipore, USA). After being blocked with 5% lipid-free milk for 2 h at room temperature, the PVDF membranes were then incubated with primary anti-P21 (#ab218311; dilution 1:1000; Abcam), anti-CDK4 (#12,790; dilution 1:1000; CST), anti-Cyclin D1 (#55,506; dilution 1:1000; CST), anti-Bcl-2 (#ab32124; dilution 1:1000; Abcam), anti-E-cadherin (#3195; dilution 1:1000; CST), anti-N-cadherin (#13,116; dilution 1:1000; CST), anti-Vimentin (#6260; dilution 1:1000; Santa), and anti-GAPDH (#ab181602; dilution 1:1000; Abcam) at room temperature for 2 h, then the membranes were incubated with the appropriate horseradish peroxidase-conjugated secondary antibody (Santa Cruz Biotechnology) for 1 h at room temperature. The immune blots were finally developed with enhanced ECL substrates (Thermo Fishier Scientific). GAPDH was detected as the internal standard.

Statistical analysis

Quantitative data from at least three biological replicates were presented as mean ± standard deviation and analyzed by the SPSS 20.0 software. Significant differences were evaluated by the Student’s t test and ANOVA as appropriate. The significant differences were defined by a P value of < 0.05.

Results

Large-scale identification of circRNAs in A549 and A549/DDP cells

To investigate the circRNAs associated with Cisplatin resistance in NSCLC cells, we first confirmed the resistance of A549/DDP cells by measuring the IC50 of Cisplatin in both cultured A549 and A549/DDP cells. We found through CCK-8 assay that the IC50 of Cisplatin in A549/DDP cells was approximately 16 μg/ml, which was greatly higher than the IC50 of about 4 μg/ml of Cisplatin in A549 cells, showing the A549/DDP cells used in our study possessed high resistance to Cisplatin treatment (Fig. 1a). Subsequently, the circRNA profiles in both the A549 and A549/DDP cells were characterized by next-generation RNA sequencing (Fig. 1b–e). We observed that most circRNAs were identified with 1 to 100 backspliced reads (Fig. 1b). The lengths of identified circRNAs ranged between 1 to 3000 bp, and the majority of circRNAs possessed a length of about 400 bp (Fig. 1c). In addition, our deep sequencing analysis showed that circRNAs identified in NSCLC cells were distributed in almost every human chromosome, and the chromosomes #1 and #2 encoded the largest numbers of circRNAs identified in both the A549 and A549/DDP cells (Fig. 1d). Totally 7137 and 3700 circRNAs were identified in A549 and A549/DDP cells, respectively, and 2286 circRNAs were identified in both the A549 and A549/DDP cells which accounted for 26.7% of all 8551 circRNAs (Fig. 1e). The significantly different circRNA profiles between A549 and A549/DDP cells indicated the critical roles of circRNAs in development of Cisplatin resistance in NSCLC cells.

Fig. 1
figure 1

Large-scale circRNA profiles in A549 and A549/DDP cells by deep sequencing. a The IC50 values of A549 and A549/DDP cells determined by CCK-8 method. A549 and A549/DDP cells were separately treated with 0, 2, 4, 8, 16 and 32 μg/ml Cisplatin for 24 h, followed by cell viability analysis via CCK-8 assay. b The numbers of backspliced reads of circRNAs identified in A549 and A549/DDP cells by next-generation RNA sequencing. c The length distributions of circRNAs identified in A549 and A549/DDP cells via RNA sequencing. d Genomic distribution of circRNAs identified between A549 and A549/DDP cells. The circRNAs identified in A549 and A549/DDP cells were indicated in yellow and green bars, respectively. e A Venn diagram showing the total numbers of circRNAs identified in A549 and A549/DDP cells. DDP: Cisplatin; IC50: half-maximal inhibitory concentration; **P < 0.01

Full size image

Differentially expressed circRNAs between A549 and A549/DDP cells

For more insights into the involvement of circRNAs in NSCLC Cisplatin resistance, the expressional alteration of circRNAs between A549 and A549/DDP cells were also quantitatively analyzed during deep sequencing. Using FDR ≤ 0.05 and log2Ratio ≥ 1 as the threshold, we found that totally 776 circRNAs showed significantly differential expression between the A549 and A549/DDP cells, including 689 up-regulated and 87 down-regulated circRNAs in A549/DDP cells compared with A549 cells (Fig. 2a–c; Supplemental Table 1). The hierarchical clustering (Fig. 2a), scatter plot (Fig. 2b), and volcano plot (Fig. 2c) of these differentially expressed circRNAs clearly demonstrated the significantly distinct circRNA profiles in A549/DDP cells in comparison to A549 cells. Moreover, the expression levels of six identified circRNAs in this study, which were encoded by genes highly expressed in lung cancer or reportedly associated with lung cancer resistance to Cisplatin, were further validated by quantitative RT-PCR method (Fig. 2d). We found that the expression of hsa_circ_0096157 and hsa_circ_0030411 were significantly elevated in A549/DDP cells compared with the A549 cells, while the expression of hsa_circ_0005941 was then suppressed in A549/DDP cells (Fig. 2d). Significantly differential expression of circRNAs in A549/DDP cells compared with A549 cells further supported the roles of circRNAs in resistance of NSCLC cells to Cisplatin treatment. Specifically, the expression of hsa_circ_0096157 showed a 2.5-fold increase in A549/DDP cells (Fig. 2d). The circular morphology of hsa_circ_0096157 was also confirmed by agarose gel electrophoresis using divergent and opposite-directed primer pairs and hsa_circ_0096157 cleavage sites was also verified by Sanger sequencing (Fig. 2e). Therefore, hsa_circ_0096157 was selected for further functional investigations in following assays.

Fig. 2
figure 2

Significantly differential circRNA profiles between A549 and A549/DDP cells by deep sequencing. a The hierarchical clustering of differentially expressed circRNAs between the A549 and A549/DDP cells. Totally 689 up-regulated and 87 down-regulated circRNAs in A549/DDP cells in comparison with A549 cells were used for hierarchical clustering. b, c The scatter plot (b) and volcano plot (c) showing the expression of 689 up-regulated and 87 down-regulated circRNAs in A549/DDP cells. d The expressional levels of six representative circRNAs between the A549 and A549/DDP cells. Quantitative RT-PCR was performed to analyze the expression of circRNAs in A549 and A549/DDP cells. GAPDH was used as the internal standard. e Circular morphology for hsa_circ_0096157 in A549 cells was validated by agarose gel electrophoresis using divergent and opposite-directed primer pairs. Confirmation of splice junction underlying hsa_circ_0096157 by Sanger sequencing. DDP Cisplatin; MALAT1 metastasis-associated lung adenocarcinoma transcript 1, M marker, cDNA circular DNA, gDNA genomic DNA, ns not significant; ***P < 0.001

Full size image

Differential gene expression profiles between A549 and A549/DDP cells

To explore the potential biological processes and signaling pathways possibly regulated by circRNA during Cisplatin resistance development, we further analyzed the gene expression profile alterations between A549 and A549/DDP cells also by next-generation sequencing. Totally, we found the mRNA levels of 859 genes were significantly up-regulated and 1038 genes were down-regulated in A549/DDP cells compared with the A549 cells, using FDR ≤ 0.001 and log2Ratio ≥ 1 as the threshold (Fig. 3a–c; Supplemental Table 2). Through GO categorization, we found that these genes differentially expressed in A549 and A549/DDP cells were significantly enriched in various biological processes, such as cellular response to stimulus, cell communication, cell differentiation and proliferation, programmed cell death, cell adhesion, cell migration, and cell cycle (Fig. 3d).

Fig. 3
figure 3

Differential gene expressional profiles in A549 and A549/DDP cells. a A hierarchical clustering of differentially expressed mRNAs between A549 and A549/DDP cells. Totally 859 up-regulated and 1028 down-regulated mRNAs in A549/DDP cells were characterized by deep sequencing compared with A549 cells. b, c A scatter plot (b) and volcano plot (c) representing the up-regulation of 689 mRNAs and down-regulation of 1038 mRNAs in A549/DDP cells. d Categorization of differentially expressed mRNAs in A549/DDP cells by GO biological processes. e The KEGG pathways with significant enrichments of differentially expressed mRNAs in A549/DDP cells. DDP Cisplatin, GO gene ontology, KEGG Kyoto encyclopedia of genes and genomes

Full size image

Moreover, our KEGG pathway analysis showed that these significantly differentially expressed genes between A549 and A549/DDP cells were significantly enriched in many signaling pathways reportedly involved in Cisplatin resistance, including the PI3K/Akt, MAPK, Focal adhesion, TNF, Ras and the Wnt signaling pathways (Fig. 3e). In addition, our bioinformatics analysis revealed an extensive interaction networks between differentially expressed circRNAs and functional proteins associated with cell cycle progression and apoptosis (Supplemental Fig. 1). The large number of genes differentially expressed and its potential interaction with key functional proteins in A549/DDP cells further indicated that the acquirement of Cisplatin resistance in NSCLC cells involved alterations of various biological processes and signaling pathways, which might be modulated by above-characterized differentially expressed circRNAs.

Hsa_circ_0096157 knockdown repressed proliferation and cell cycle progression and promoted apoptosis of A549/DDP cells under Cisplatin treatment and non-treatment

To study the potential functions of hsa_circ_0096157 in Cisplatin resistance of NSCLC cells, we knocked down the expression of hsa_circ_0096157 in A549/DDP cells by transfection with specific siRNAs (Fig. 4a). Quantitative RT-PCR results showed that the expression of has_circ_0096157 was significantly downregulated in siRNA group compared to siNC group (Fig. 4a). Compared with non-Cisplatin treatment, Cisplatin treatment suppressed significantly the proliferation rates of A549/DDP cells (Fig. 4b). Interfering with the expression of has_circ_0096157 inhibited significantly the proliferation rates of A549/DDP cells under Cisplatin treatment and non-treatment (Fig. 4b). TUNEL and flow cytometry assays results showed that, compared with non-Cisplatin treatment, Cisplatin treatment promoted significantly the apoptosis of A549/DDP cells (Fig. 4c, d). Interfering with the expression of has_circ_0096157 promoted significantly the apoptosis of A549/DDP cells under Cisplatin treatment and non-treatment (Fig. 4c, d). On the contrary, compared with non-Cisplatin treatment, the flow cytometry results showed that Cisplatin treatment significantly blocked the cell cycle in G1 phase (Fig. 4e). Interfering with the expression of hsa_circ_0096157 significantly blocked the cell cycle in S phase and reduced G2 phase under Cisplatin treatment and non-treatment (Fig. 4e). These results clearly showed that hsa_circ_0096157 knockdown could effectively impair the proliferation and cell cycle progression of A549/DDP cells, however, promote apoptosis of A549/DDP cells under Cisplatin treatment.

Fig. 4
figure 4

Regulation of A549/DDP cell proliferation, apoptosis, and cycle progression by hsa_circ_0096157 knockdown under Cisplatin treatment and non-treatment. a Significant decrease of hsa_circ_0096157 expression in A549/DDP cells by transfection with siRNAs. Hsa_circ_0096157 expression was assessed by quantitative RT-PCR method. b The proliferation rates of A549/DDP cells with or without hsa_circ_0096157 knockdown under Cisplatin treatment and non-treatment. A549/DDP cells were treated with 3 μg/ml DDP for 24 h, and the proliferation was analyzed by MTS method. c, d The apoptosis of A549/DDP cells with or without hsa_circ_0096157 knockdown under Cisplatin treatment and non-treatment via TUNEL assay (c) and flow cytometry (d). Percentages of apoptotic A549/DDP cells were determined by Image J software and flow cytometry. e The cell cycle progression of A549/DDP cells with or without hsa_circ_0096157 knockdown under Cisplatin treatment and non-treatment. The percentages of A549/DDP cells at G1, S and G2 phases were quantitated by flow cytometry after treatment with 3 μg/ml DDP for 24 h. DDP Cisplatin, NC negative control, ( −) no Cisplatin treatment, ( +) treatment with 3 μg/ml Cisplatin for 24 h; *P < 0.05; **P < 0.01; ***P < 0.001; bar indicates 100 μm

Full size image

Hsa_circ_0096157 promoted proliferation and cell cycle progression and suppressed apoptosis of A549 cells under Cisplatin treatment and non-treatments

For further validation of hsa_circ_0096157 in Cisplatin resistance in NSCLC, the hsa_circ_0096157 was then overexpressed in A549 cells by transfection with the LV003-hsa_circ_0096157 recombinant plasmids (Fig. 5a). Quantitative RT-PCR results showed that the expression of has_circ_0096157 was significantly upregulated in has_circ_0096157 group compared to LV003 group (Fig. 5a). Compared with non-Cisplatin treatment, Cisplatin treatment suppressed significantly the proliferation rates of A549 cells (Fig. 5b). Overexpression of has_circ_0096157 promoted significantly the proliferation rates of A549 cells under Cisplatin treatment and non-treatment (Fig. 5b). TUNEL and flow cytometry assays results showed that, compared with non-Cisplatin treatment, Cisplatin treatment promoted significantly the apoptosis of A549 cells (Fig. 5c, d). Overexpression of has_circ_0096157 inhibited significantly the apoptosis of A549 cells under Cisplatin treatment and non-treatment (Fig. 5c, d). Moreover, compared with non-Cisplatin treatment, the flow cytometry results showed that Cisplatin treatment significantly blocked the cell cycle in G1 phase (Fig. 5e). Overexpression of hsa_circ_0096157 significantly promoted the conversion from S phase to G2 phase under Cisplatin treatment and non-treatment (Fig. 5e). Together, we proved here by these assays that hsa_circ_0096157 overexpression promoted the proliferation and cell cycle progression, however suppressed the apoptosis of A549 cells. More importantly, hsa_circ_0096157 overexpression inhibited the suppression of A549 cell proliferation and cell cycle progression as well as the inducement of A549 cell apoptosis caused by Cisplatin treatment.

Fig. 5
figure 5

Modulation of A549 cell proliferation, apoptosis, and cycle progression by hsa_circ_0096157 overexpression under Cisplatin treatment and non-treatment. a The increase of hsa_circ_0096157 expression in A549 cells by transfection with overexpressing plasmids. The hsa_circ_0096157 expression was measured by quantitative RT-PCR assay. b The changes of proliferation rates in A549 cells overexpressing hsa_circ_0096157 under Cisplatin treatment and non-treatment. After being treated with with 3 μg/ml DDP for 24 h, the proliferation of A549 cells was determined by the MTS method. c, d The numbers of apoptotic A549 cells with hsa_circ_0096157 overexpression under Cisplatin treatment and non-treatment via TUNEL assay (c) and flow cytometry (d). The number of apoptotic A549 cells were determined by Image J software and flow cytometry. e The alterations of A549 cell cycle progression induced by hsa_circ_0096157 overexpression under Cisplatin treatment and non-treatment. Flow cytometry was performed to analyze the percentages of A549 cells at the G1, S and G2 stages. DDP Cisplatin, ( −) no Cisplatin treatment, ( +) treatment with 3 μg/ml Cisplatin for 24 h; *P < 0.05; **P < 0.01; bar indicates 100 μm

Full size image

Hsa_circ_0096157 knockdown repressed migration and invasion in A549/DDP cells and promoted the migration and invasion in A549 cells

Based on the above effects of circRNA on cell proliferation, cell cycle progression, and apoptosis, we further analyzed the effects of hsa_circ_0096157 knockdown in A549/DDP cells and hsa_circ_0096157 overexpression in A549 cells on cell migration and invasion. The results showed that hsa_circ_0096157 knockdown resulted into significantly suppressed migration and invasion capacities in A549/DDP cells treated with Cisplatin or no Cisplatin (Fig. 6a, b). While hsa_circ_0096157 overexpression resulted into significantly promoted migration and invasion capacities in A549 cells treated with Cisplatin or no Cisplatin (Fig. 6c, d). Together, the above results indicated that hsa_circ_0096157 knockdown and overexpression played an important role in the migration and invasion of A549/DDP and A549 cells respectively.

Fig. 6
figure 6

Migration and invasion analysis of A549/DPP cell with hsa_circ_0096157 knockdown and A549 cell with hsa_circ_0096157 overexpression under Cisplatin treatment and non-treatment. a Regulation of A549/DPP cell migration and invasion by hsa_circ_0096157 knockdown. Migration and invasion of A549/DDP cells were analyzed using the Transwell system. b Semiquantitative analysis of the number of migration and invasion cells in graph A. c Regulation of A549 cell migration and invasion by hsa_circ_0096157 overexpression. Migration and invasion of A549/DDP cells were analyzed using the Transwell system. d Semiquantitative analysis of the number of migration and invasion cells in graph C. DDP Cisplatin, NC negative control; ( −) no Cisplatin treatment, ( +) treatment with 3 μg/ml Cisplatin for 24 h; *P < 0.05; **P < 0.01; bars indicate 100 μm

Full size image

Regulation of cell cycle signaling pathway and epithelial-mesenchymal transition (EMT) by hsa_circ_0096157 in A549 and A549/DDP cells

To investigate the molecular events underlying the regulation of Cisplatin resistance by hsa_circ_0096157 in NSCLC, the expression levels of several key components of the cell cycle signaling pathway were quantitatively analyzed in A549 cells overexpressing hsa_circ_0096157 and A549/DDP cells with hsa_circ_0096157 knockdown. We observed that the abundance of P21 and E-cadherin proteins in A549 cells were significantly down-regulated by hsa_circ_0096157 overexpression, however P21 and E-cadherin protein levels in A549/DDP cells were greatly elevated by hsa_circ_0096157 knockdown (Fig. 7a and b). Contrarily, the abundance of CDK4 protein in hsa_circ_0096157-overexpressing A549 cells was significantly higher than the A549 cells, but its level in A549/DDP cells with hsa_circ_0096157 knockdown was markedly lower than the A549/DDP cells (Fig. 7a and b). Furthermore, the protein levels of Cyclin D1, Bcl-2, N-cadherin, and Vimentin showed similar alterations as CDK4 in A549 and A549/DDP cells induced by hsa_circ_0096157 overexpression and knockdown, respectively (Fig. 7a and b), indicating that cell cycle progression might be inhibited by increased P21 protein level to inhibit the expression of CDK4-Cyclin D1 complex and decreased Bcl-2 protein level; and migration and invasion were suppressed by the increased E-cadherin and decreased N-cadherin and Vimentin expression levels through the hsa_circ_0096157 knockdown in A549/DDP cells. These molecular evidences further verified the regulation of cell cycle signaling pathway and EMT by hsa_circ_0096157 expression alteration, which mediated the acquirement of Cisplatin resistance in NSCLC cells.

Fig. 7
figure 7

Changes of the cell cycle signaling pathway and EMT marker in A549 and A549/DDP cells by hsa_circ_0096157. a Abundances of key protein components of the cell cycle signaling pathway in A549 cells with hsa_circ_0096157 overexpression and A549/DDP cells with hsa_circ_0096157 knockdown. Protein levels of P21, CDK4, Cyclin D1, Bcl-2, E-cadherin, N-cadherin, and Vimentin in A549 and A549/DDP were measured by western blotting. GAPDH was used as the internal standard. b Quantitation of P21, CDK4, Cyclin D1, Bcl-2, E-cadherin, N-cadherin, and Vimentin protein levels in A549 and A549/DDP cells. DDP Cisplatin, NC negative control, OE overexpression; *P < 0.05; **P < 0.01

Full size image

Discussion

The resistance of chemotherapy drugs such as Cisplatin to the treatment of NSCLC has become a huge challenge for cancerous treatment [7, 8]. The full elucidation of the molecular mechanisms underlying acquirement of Cisplatin treatment has been regarded as one essential basis for developing novel treatments for patients resistant to Cisplatin treatment. To provide novel insights into the epigenetic regulation of Cisplatin resistance in NSCLC, we performed a large-scale identification of circRNA profiles in A549 cells and A549/DDP cells with strong resistance to Cisplatin treatment via next-generation deep sequencing method. Totally 7137 and 3700 circRNAs were sequenced in A549 and A549/DDP cells, respectively, among which 776 circRNAs were significantly differentially expressed in A549/DDP cells compared with A549 cells. In consistence with the alteration of circRNA profiles, we also revealed a number of differentially expressed genes in A549/DDP cells associated with various biological processes and signaling pathways linked with cancer development and chemotherapy resistance. Furthermore, we also showed that the proliferation, migration, invasion, cell cycle progression, and apoptosis of A549 and A549/DDP cells were greatly altered by hsa_circ_0096157 knockdown or overexpression under Cisplatin treatment. These assays revealed great changes of circRNA profiles associated with Cisplatin resistance in NSCLC, and proved the hsa_circ_0096157 expression elevation as a driving actor for Cisplatin resistance in NSCLC cells.

Noncoding RNA molecules such as miRNAs and long non-coding RNAs (lncRNAs) have recently been established as essential parts of cellular epigenetic networks, which play critical roles in chemotherapy resistance in NSCLC [26, 28, 29]. For instance, the miR106a expression is greatly upregulated in DDP-resistant A549 cell line, which is reported to confer Cisplatin resistance in A549 cells by modulating the expression of ABCA1, a member of the adenosine triphosphatase binding cassette family [29]. It has been demonstrated by accumulating evidences that circRNAs serve as sponges of miRNAs in various biological and pathogenic processes, resulting into the substantial expressional alterations of key functional genes [30, 31]. However, the circRNAs which may be critically involved in Cisplatin resistance of NSCLC cells remains largely unexplored. The rapid development of next-generation RNA sequencing and combined computational database and analysis tools greatly facilitate identification and annotation of circRNA profiles under distinct contexts [32]. In the present study, we identified a great number of differentially expressed circRNAs in A549/DDP cells and these circRNAs were mainly distributed on chromosomes #1 and #2, indicating the significances of circRNA expression alterations on chromosomes #1 and #2 in Cisplatin resistance. Of note, these 689 upregulated circRNAs in A549/DDP cells greatly overnumbered those 87 down-regulated circRNAs, indicating circRNA expression elevation might serve as a common mechanism promoting NSCLC resistance to Cisplatin treatment. Combined with the great alterations of gene expression profiles revealed in this study, further functional investigation of differentially expression circRNAs and their regulatory networks would put new perspectives on epigenetic regulation of Cisplatin resistance acquirement.

Among the large set of up-regulated circRNAs in A549/DDP cells, we validated by quantitative RT-PCR that hsa_circ_0096157, hsa_circ_0030411, and hsa_circ_0005941 expression levels exhibited significant increase in A549/DDP cells. Previous reports showed hsa_circ_0096157 is formed by the ligation of long noncoding RNA MALAT1, which is implicated in cell proliferation and apoptosis regulation underlying the resistance of lung cancer to Cisplatin [21, 22]. Alpha-ketoglutarate dependent dioxygenase (FTO) knockdown also inhibited the colony formation ability of lung cancer cells [32], and it can be interfered that has_circ_0005941 derived from FTO gene may play an important role in A549/DDP resistance to cisplatin. To explore its biological functions during Cisplatin resistance in NSCLC, we silenced the highest expression level of hsa_circ_0096157 in A549/DDP cells, which resulted into greatly suppressed proliferation, migration, invasion, and cell cycle progression in A549/DDP cells, even under Cisplatin treatment. Previous reports already showed that Cisplatin-resistant NSCLC cells are featured by enhanced proliferation, migration, invasion, and cell cycle progression [33,34,35]. Also, the abrogation of proliferation and apoptosis alterations in Cisplatin-resistant NSCLC cells has been regarded as a promising strategy to overcoming the resistance of cancer cells to Cisplatin treatment [36, 37]. The effective modulations of A549/DDP cell proliferation and apoptosis as well as other cellular tumorigenesis-related processes by hsa_circ_0096157 knockdown hopefully suggested that targeting hsa_circ_0096157 expression might be utilized as a new method to increase Cisplatin sensitivity and overcome NSCLC resistance to Cisplatin. In consistence, the overexpression of hsa_circ_0096157 in A547 cells in this study produced completely opposite alterations of these cellular processes in A547 cells, which further validated the potential roles of hsa_circ_0096157 in development of Cisplatin resistance in NSCLC cells. These observations also disclosed new mechanisms mediating the tumorigenesis-regulating roles of MALAT1 and other long non-coding RNAs.

It has been well-documented that activation of cell cycle progression substantially promotes the development of Cisplatin in various human cancers [10, 37]. In NSCLC cells, the therapeutic effects of Cisplatin treatment has been partially attributed to its capacity of inducing G2/M cell cycle arrest in cancerous cells, while the development of Cisplatin resistance is mediated by the abrogation of G2/M cell cycle arrest during Cisplatin treatment [10]. In this study, we convincingly demonstrated that the cell cycle progression of A549/DDP cells could be significantly inhibited by hsa_circ_0096157 knockdown under Cisplatin treatment. On the contrary, hsa_circ_0096157 overexpression mitigated the Cisplatin-induced A549 cell cycle arrest. These results further highlighted the significance of cell cycle progression alteration during development of Cisplatin resistance in NSCLC cells. In consistence, we detected consistent alteration of four key signaling proteins regulating cell cycle progression in A549 and A549/DDP cells with altered hsa_circ_0096157 expression. The P21 protein is a well-known tumor suppressor protein which induces cell cycle arrest to inhibit tumorigenesis [38]. Also, the CDK4/Cyclin D1 axis and Bcl-2 protein act as essential checkpoint mechanism and regulatory signaling pathways driving cell cycle progression [39]. We also investigated the expression of EMT biomarkers such as E-cadherin, N-cadherin, and Vimentin, which were involved in the regulation of molecular mechanisms of tumor migration and invasion. Research reveals that overexpression of lncRNA H19 reduces the level of E-cadherin and increases the expression levels of Vimentin and N-cadherin to promoting the A549 cells invasion and metastasis [40]. While, our results also showed that E-cadherin protein level was greatly elevated and Vimentin and N-cadherin protein levels were downregulated in A549/DDP cells by hsa_circ_0096157 knockdown. The corresponding changes of these proteins in NSCLC cells by alteration of hsa_circ_0096157 further indicated the mediating roles of cell cycle regulation and migration in Cisplatin resistance of NSCLC, which deserve further investigations.

In summary, we characterized a large set of significantly differentially expressed circRNA profiles between A549 and A549/DDP cells through next-generation deep sequencing. Moreover, we disclosed for the first time that the elevated expression of hsa_circ_0096157 contributed to Cisplatin resistance by promoting proliferation, migration, invasion, and cell cycle progression; and suppressing apoptosis of NSCLC cells. These results provided novel insights into the noncoding RNAs-mediated Cisplatin resistances and a basis for developing novel treatments for Cisplatin-resistant lung cancer patients.