miR-27a regulates cisplatin resistance and metastasis by targeting RKIP in human lung adenocarcinoma cells

To better understand the biological mechanisms of chemoresistance in lung adenocarcinoma cells and search for the reversion opportunities, we made use of a cisplatin-sensitive and derived resistant lung cancer cell line pair (A549 and A549/ CDDP). Compared with parental A549 cells, A549/CDDP cells displayed spindle shape and separated from one other (Figure 1A), MTT assay was then performed, and the IC50 of A549/CDDP cells increased 15.7 fold (Figure 1B). Moreover, western blotting demonstrated that the protein expression of E-cadhein decreased, while that of vimentin dramatically increased in A549/CDDP cells (Figure 1C). Additionally, invasion assay demonstrated that the invasion significantly increased in A549/CDDP as compared with A549 cells (Figure 1D).

Based on the miRNA microarray data, 13 miRNAs were found to be differentially expressed (>2-fold change) in A549/CDDP cells compared with A549 cells (Additional file 1: Table S1), among which miR-27a was the most up-regulated one (5.6-fold change). The result was validated via real-time quantitative RT-PCR (Figure 1E).

To investigate the association of miR-27a expression with lung adenocarcinoma chemoresistance against cisplatin, A549 and A549/CDDP cells were transfected with miR-27a mimics and miR-27a inhibitors respectively. As shown in Figure 2A, miR-27a mimics increased vimentin, but suppressed E-cadherin expression in A549 cells. Invasion assay demonstrated that miR-27a mimics synergistically enhanced the invasion of A549 cells (Figure 2B). On the other hand, silencing miR-27a expression in A549/CDDP cells using anti-sense oligonucleotides reduced vimentin and increased the expression of E-cadherin (Figure 2A). Furthermore, miR-27a inhibitors synergistically suppressed the invasion of A549/CDDP cells (Figure 2B). Therefore, upregulation of miR-27a may play an important role in chemotherapy-induced EMT of lung adenocarcinoma cells. Moreover, MTT assay revealed that A549 cells transfected with miR-27a mimics showed greatly decreased sensitivity to cisplatin as indicated by substantially increased IC50 values (Figure 2C). In contrast, suppression of the miR-27a level in A549/CDDP cells resulted in an enhanced sensitivity to cisplatin (Figure 2D). Similar results were obtained in H1395 and H1299 cells (Additional file 2: Figure S1).

To investigate the effect of miR-27a expression on chemosensitivity of lung adenocarcinoma in vivo, A549 cells stably expressing miR-27a by lentivirus were subcutaneously inoculated into nude mice. When the average tumor size reached ≈ 50 mm³, cisplatin was administered via intraperitoneal injection at a dose of 5 mg/kg, 1 dose every other day, with 3 doses in total. As shown in Figure 3A, the number of lung metastasis nodules was dramatically increased in miR-27a overexpression group when compared with control. Next, we engineered A549/CDDP cells to stably inhibit miR-27a with a lentivirus-mediated antagomir. The results showed that suppression of miR-27a decreased the number of lung metastases (Figure 3B). These results suggested that miR-27a could regulate the response to cisplatin in vivo.

By employing open access softwares (TargetScan and PicTarget), RKIP was chosen as a preferred candidate target gene of miR-27a because of the putative binding site within its 3′UTR (Figure 4A) and lower RKIP protein expression in A549/CDDP cells (Figure 4B). Western blot showed that overexpression of miR-27a in A549 cells significantly repressed RKIP protein expression compared to cells transfected with negative control (Figure 4C). Relatively, downregulation of miR-27a by inhibitors in A549/CDDP cells led to a moderate increase of RKIP protein level (Figure 4C). To verity whether RKIP is the direct downstream target of miR-27a, a fragment of RKIP 3′UTR containing the putative miR-27a binding site was cloned into a luciferase reporter vector. Luciferase reporter assays showed that up-regulation of miR-27a significantly decreased the relative luciferase activity of RKIP 3′UTR in A549 cells, but had no effect on the mutant of RKIP 3′UTR (Figure 4D). Taken together, these results suggest that miR-27a down-regulates RKIP expression by directly targeting its 3′UTR.

To further examine whether RKIP is involved in miR-27a-induced chemoresistance, we performed loss-of-function and gain-of-function analyses. Firstly, A549 cells were transfected with si-RKIP or negative control. Western blotting analysis confirmed that the expression of RKIP was suppressed (Figure 5A). As expected, RKIP knockdown significantly increased vimentin, reduced E-cadherin and decreased sensitivity to cisplatin in A549 cells (Figure 5A). Subsequently, we employed an expression construct that encodes the entire RKIP coding sequence but lacks the 3′UTR. Ectopic expression of RKIP partially rescued miR-27a-mediated EMT and cisplatin resistance in miR-27a-overexpressing cells (Figure 5B). Collectively, these data suggest that miR-27a regulate chemoresistance of lung adenocarcinoma cells at least in part by targeting RKIP.

To better understand the association between miR-27a and RKIP expression, a total of 30 clinical tumor tissue samples were collected from patients with advanced lung adenocarcinoma and divided into^“sensitive^” and ^“insensitive^”groups according to the patient’s response to cisplatin-based chemotherapy. As shown in Figure 6A, miR-27a was significantly up-regulated in the^“insensitive^”group tissues (n = 17) compared with that in the ^“sensitive^”group ones (n = 13). On the contrary, RKIP mRNA expression level was significantly down-regulated in the^“insensitive^” group tissues (Figure 6B). The inverse correlation between miR-27a and RKIP mRNA expression was verified by linear regression analysis (r = −0.691, P < 0.01) (Figure 6C). We then analyzed the association of miR-27a expression with survival of patients. As shown in Figure 6D, Patients with high miR-27a expression showed significantly shorter overall survival than those with low miR-27a expression (P < 0.01).

Human lung adenocarcinoma A549 cells were purchased from the American Type Culture Collection and cultured in RPMI 1640 medium containing 10% fetal bovine serum (FBS) with 100 μg/ml penicillin /streptomycin at 37°C with 5% CO₂. The cisplatin-resistant A549 cell line (A549/CDDP) was established and preserved in a 40 μmol/L final concentration of cisplatin in our laboratory.

miR-27a mimics and negative control mimics (NC), miR-27a inhibitors (anti-miR-27a) and negative control inhibitors (anti-NC) and RKIP siRNAs were synthesized by GenePharma Company (Shanghai, China). Transfection was performed with Lipofectamine 2000 (Invitrogen, CA, USA) according to the manufacturer’s protocol. Total RNA and protein were prepared 48 h after transfection and were used for qRT-PCR or Western blot analysis.

A total of 30 lung adenocarcinoma tissues were collected from patients with advanced lung adenocarcinoma who received chemotherapy at Yinzhou People’s Hospital (Ningbo, China) between January 2009 and March 2010. Informed consent was obtained from all subjects and this study was approved by the Clinical Research Ethics Committee of Yinzhou People’s Hospital. Patients met all of the following criteria: primary lung adenocarcinoma; histological diagnosis of lung adenocarcinoma with at least 1 measurable lesion; clinical stage IIIB-IV; first-line chemotherapy either with cisplatin 100 mg/m² and docetaxel 75 mg/m² or cisplatin 100 mg/m² and gemcitabine 1000 mg/m² administered every 3 weeks for a maximum of 5 cycles. Samples were divided into “sensitive” (complete response or partial response) and “insensitive” (stable disease or progressive disease) groups according to the patient’s responses assessed via medical image analysis and detection of serum tumor markers after 4 or 5 cycles of cisplatin-based chemotherapy.

Total RNA was extracted from the cultured cells and the lung adenocarcinoma tissue specimens using Trizol Reagent (Invitrogen, CA, USA) according to the manufacturer’s protocol. The expression level of mature miR-27a was measured by TaqMan miRNA assays (Applied Biosystems, CA, USA) according to the provided protocol, U6 snRNA levels were used for normalization. RKIP expression was measured by SYBR green qPCR assay (Takara, Dalian, China) and GAPDH was used as an endogenous control.

Protein extracts were prepared by a modified RIPA buffer with 0.5% sodium dodecyl sulfate (SDS) in the presence of proteinase inhibitor cocktail (Complete mini, Roche, Indianapolis, IN, USA). Polyacrylamide gel electrophoresis, tank-based transfer to Immobilon Hybond-C membranes (Amersham Biosciences) and immunodetection were performed with standard techniques. Antibodies to RKIP (catalog no. sc-28837), E-cadherin (catalog no. sc-8426), vimentin (catalog no. sc-32322) and β-actin (catalog no. sc-1616) were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Signals were visualized with SuperSignal® West Pico chemoluminescent substrate (Pierce, Rockford, III, USA) by exposure to films.

Cells were seeded into 96-well plates (2 × 10³ cells/well) directly or 24 hours after transfection and allowed to attach overnight. Freshly prepared cisplatin was then added with different final concentrations. Forty-eight hours later, cell viability was assessed via 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) assay as described previously [35].

2 × 10⁵ cells were added into the upper chamber of the insert precoated with Matrigel (ECM gel, Sigma-Aldrich, St. Louis, MO). Cells were plated in medium without serum, and medium containing 10% fetal bovine serum in the lower chamber served as chemoattractant. After several hours of incubation, the cells that did not invade through the pores were carefully wiped out with cotton wool, and the filters were fixed by treatment with 95% ethanol for 30 minutes and stained with 0.2% crystal violet solution for 30 minutes. Invasive cells adhering to the undersurface of the filter were counted (five fields/chamber; 0.24 mm²/field) using an inverted microscope, and each experiment was repeated three times.

Wild-type 3′untranslated region (3′UTR) of RKIP containing predicted miR-27a target sites were amplified by PCR from A549 cell genomic DNA. Primers used: Forward: GAT CTG CAG GGG TTA GCT TGG GGA CCT GAA C; Reverse: GAT CAT ATG AGA GTG ACA TAC TGA TGC CTA C. Mutant 3′UTRs were generated by overlap-extension PCR method. Both wild-type and mutant 3′UTR fragments were subcloned into the pGL3-control vector (Promega, Madison, WI) immediately downstream of the stop codon of the luciferase gene. DNA fragment coding RKIP protein was amplified by PCR from A549 cell cDNA, and cloned into pCMV-Myc expression vector (Clonetech, Mountain View, CA). Primers used: Forward: GCT GAA TTC ATG CCG GTG GAC CTC AGC AAG T; Reverse: CTG CTC GAG CTA CTT CCC AGA CAG CTG CTC G. For luciferase assay, the reporter plasmid was cotransfected with a control Renilla luciferase vector into A549 cells in the presence of either miR-27a or NC. After 48 h, cells were harvested, and the luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega, Madison, WI, USA).

Five-week-old female BALB/c nude mice were purchased from the Animal Center of Zhejiang University (Hangzhou, China). For in vivo chemosensitivity and metastasis assays, A549 cells ( infected with either the miR-27a-overexpressing lentivirus or the mock lentivirus) and A549/CDDP cells (infected with either the miR-27a-knockdown lentivirus-mediated antagomir or the antagomir-NC) were subcutaneously inoculated into nude mice (six per group, 1 × 10⁶ cells for each mouse). Tumor growth was examined every other day, and tumor volumes were calculated using the equation V = A × B²/2 (mm³), where A is the largest diameter and B is the perpendicular diameter. When the average tumor size reached ≈ 50 mm³, cisplatin was administered via intraperitoneal injection at a dose of 5 mg/kg, 1 dose every other day, with 3 doses in total. After 2 weeks, all mice were sacrificed. Transplanted tumors were excised, and tumor tissues were used to perform hematoxylin & eosin (H&E) staining. All research involving animal complied with protocols approved by the Zhejiang medical experimental animal care commission.

Statistical analyses were performed using SPSS 16.0 software (SPSS Inc.). All data from 3 independent experiments were expressed as mean ± SD. Differences were assessed by two-tailed Student’s t test. P < 0.05 was considered statistically significant.