A549, H1299, HepG2, MCF-7, HeLa and Cos-7 cell lines were used in the present study. The A549 and H1299 cell lines were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA) (nos. CCL-185 and CRL-5803, respectively). These cells were seeded in RPMI-1640 medium and fetal bovine serum (FBS) to a final concentration of 10%. HepG2, MCF-7, HeLa and Cos-7 cells used in this study were purchased from ATCC (nos. HB-8065, HTB-22, CCL-2 and CRL-1651, respectively). These cells were seeded in Dulbecco's modified Eagle's medium-high glucose (DMEM-HG; Hyclone; GE Healthcare Life Sciences, Logan, UT, USA) supplemented with 10% FBS) (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA). All of the aforementioned cell lines were cultured at 37°C in a humidified air with 5% CO₂.

Human Nrf2 (Gene ID: 4780) was inserted into expression vector (pcDNA3.1-). ShNrf2 and its corresponding control were kindly provided by Professor Jian Dong (North Carolina State University, Raleigh, NC, USA). miR-144-3p mimic, miR-144-3p inhibitor or the appropriate negative controls (NC) of miRNA mimic (mimic NC) and miRNA inhibitor (inhibitor NC) were obtained from Guanzhou RiboBio Co., Ltd. (Guanzhou, China). For transfection experiments, the cells were cultured in growth medium without antibiotics at 60% confluence for 2 days, and then transfected with transfection reagent (FuGENE^® HD; Roche Diagnostics, Basel, Switzerland) according to the manufacturer's instructions. After incubation for 6 h, the medium was removed and replaced with normal culture medium for 48 h.

Cell growth was estimated by a modified MTT assay. As a measurement of cell growth, the cells were seeded onto a 96-well dish and grown in medium containing 10% FBS. After the cells were treated daily with cisplatin (1, 2, 4, 8, 16, 32 and 64 µM) for 24 h, the MTT reagent (2.5 mg/ml) was added and the optical density (570 nm) was measured by ELISA reader.

Total RNA, including miRNA, was extracted by using a e.Z.N.A miRNA kit (Omega Bio-Tek, Inc., Norcross, GA, USA) according to the manufacturer's protocol, and the sample was reverse-transcribed using M-MLV reverse transcriptase (Promega Corporation, Madison, WI, USA). Real-time PCR was performed using Applied Biosystems Step One™ Real-Time PCR System. Fast SYBR^® Green Master Mix was obtained from Applied Biosystems; Thermo Fisher Scientific, Inc. Data presented as the relative expression levels of Nrf2, HO-1, NQO-1, Bcl-2 and caspase-3 were normalized by GAPDH. The primers for the PCR analysis are listed in Table I. Amplification of U6 small nuclear RNA served as an endogenous control used to normalize miR-144-3p expression data. Thermocycling conditions were as follows: 95°C for 5 min followed by 40 cycles at 95°C for 10 sec and 60°C for 30 sec, then a melting curve analysis from 60 to 95°C every 0.2°C for 1.5 min was obtained. Each sample was analyzed in triplicate, and quantified using the 2^−ΔΔCq method (18).

For the western blot analysis, protein samples were extracted from the cells with Protein Extraction Reagent (Pierce; Thermo Fisher Scientific, Inc.). The concentrations of the proteins were determined using the BCA Quantification kit (Beyotime Institute of Biotechnology, Beijing, China) for subsequent sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The proteins (20 µg) were separated by SDS PAGE (10%) and transferred onto a PVDF membrane. The membrane was blocked using 5% non-fat milk at 25°C for 1 h, and then incubated with primary antibodies overnight at 4°C. The antibodies used were as follows: Anti-human GAPDH antibody (dilution 1:2,000; cat. no. 97166; Cell Signaling Technology, Inc., Danvers, MA, USA), anti-human Nrf2 antibody (dilution 1:1,000; cat. no. sc-81342; Santa Cruz Biotechnology, Inc., Dallas, TX, USA), anti-human HO-1 antibody (dilution 1:1,000; cat. no. 5853; Cell Signaling Technology, Inc.), anti-human NQO-1 antibody (dilution 1:1,000; cat. no. ab28947; Abcam, Cambridge, UK), anti-human Bcl-2 antibody (dilution 1:1,000; cat. no. sc-509) and anti-human caspase-3 antibody (dilution 1:1,000; cat. no. sc-271759) (both from Santa Cruz Biotechnology, Inc.). Then, the membrane was incubated with IRDyeTM-800 conjugated anti-mouse or anti-rabbit secondary antibodies (dilution 1:5,000; cat. nos. 115-005-146 and 115-005-144; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) for 1 h at room temperature. The protein signals were visualized with the Odyssey Infrared Imaging System (LI-COR Biosciences, Lincoln, NE, USA). GAPDH expression was used as an internal control.

The luciferase constructs were as follows: i) Nrf2-3′-UTR-WT: The Nrf2 3'-UTR region was fused to the pmirGLO reporter vector; ii) Nrf2-3′-UTR-MUT: The same as the Nrf2-3′-UTR-WT, except that the miR-144-3p binding site was mutated; iii) miR-144-3p-luc: The miR-144-3p promoter region (about 2000 bp) was fused to the pGL3-Basic reporter vector; iv) Mut-miR-144-3p-luc: The same as miR-144-3p-luc, except that the Nrf2 binding site was mutated. The cells (2×10⁵/well) were plated in 24-well plates. To ascertain the effect of miR-144-3p on Nrf2, Cos-7 cells were co-transfected with miR-144-3p mimic or mimic NC in combination with Nrf2-3′-UTR-WT or Nrf2-3′-UTR-MUT. In order to explore the mechanism of Nrf2 acting on miR-144-3p, Cos-7 cells were co-transfected with Nrf2 or control (pcDNA3.1-) in combination with miR-144-3p-luc, or mut-miR-144-3p-luc. Cells were harvested 48 h after transfection and luciferase activity was assessed using the Dual Luciferase Assay System (Promega Corp.). The results were expressed as a fold induction relative to the cells transfected with the control after normalization to Renilla activity. In the dual luciferase assay results, all columns represented the mean result of three independent experiments and the error bars represented the standard deviation.

We used a commercial Chromatin Immunoprecipitation (ChIP) Assay kit (Merck KGaA, Darmstadt, Germany) by following the manufacturer's instructions. After the treatment, each test group was incubated with 1% formaldehyde to cross-link DNA-protein complexes. After washing with ice-cold PBS three times, the cells were lysed in SDS lysis buffer. Then lysates were sonicated to shear DNA to ~200-1,000 bp fragments. We then used an anti-Nrf2 antibody to immunoprecipitate the cross-linked protein at 4°C overnight. IgG acted as the negative control. The DNA was used as a template for PCR and utilized the Nrf2 binding site. The PCR products were separated on 1% agarose gel.

High throughput sequencing data and prognostic data were derived from the TCGA database (GSE56036) and LinkedOmics database (ID-3650). The software of TargetScan was used to predict gene targets. Data were expressed as the mean ± SE, accompanied by the number of experiments performed independently, and analyzed by t-tests. Differences at P<0.05 were considered to be statistically significant.

There was a significant difference of miR-144-3p expression in tumor tissues and normal tissues in patients with platinum insensitivity drugs by analyzing high-throughput sequencing data from the TCGA database (GSE56036). The results revealed that miR-144-3p expression was significantly decreased in lung cancer tissues compared with adjacent non-tumor tissues (Fig. 1A-C). The emergence of drug resistance often indicates a poor prognosis. To assess the clinical significance of miR-144-3p overexpression in lung cancer, we evaluated the association between miR-144-3p levels and patient clinicopathological characteristics based on the TCGA database (GSE56036) and LinkedOmics database (ID-3650). Kaplan-Meier survival analysis revealed that patients with higher miR-144-3p levels had longer overall survival and progression-free survival times than those who had lower levels of miR-144-3p (Fig. 1D). These findings indicated that the expression of miR-144-3p was decreased in lung cancer tissues and associated with poor prognosis. Thus, we performed experimental research on whether miR-144-3p was related to cisplatin resistance.

To further investigate whether miR-144-3p was involved in drug resistance in lung cancer, we transfected miR-144-3p mimic or miR-144-3p inhibitor to lung cancer cells (Fig. 2A and B) and then exposed the cells to cisplatin with rising levels of concentration for 24 h. The viability of lung cancer cells that were transfected with miR-144-3p were more sensitive to cisplatin in comparison with the mimic NC group (Fig. 2C and D). However, the miR-144-3p inhibitor did the opposite, the resistance of cells was significantly enhanced to cisplatin (Fig. 2C and D). These results indicated that miR-144-3p inhibited cisplatin resistance in lung cancer cells. Thus, targeting miR-144-3p could reverse the cisplatin resistant behavior in lung tumor cells.

In order to further explore how miR-144-3p regulated drug resistance, we used TargetScan to predict potential miR-144-3p target genes. Nrf2 was identified during the scanning process. Nrf2, as a nuclear transcription factor, which binds to the ARE sequences and activates the molecules downstream, was considered to render cancer cells resistant to drugs including cisplatin. We next investigated whether miR-144-3p could regulate the Nrf2 pathway during the cisplatin resistance process in lung cancer cells.

Within 24 h of cisplatin treatment, we examined the mRNA and protein levels of Nrf2 in lung cancer cells after transfection with the miR-144-3p mimic or miR-144-3p inhibitor. Data revealed that overexpressed miR-144-3p effectively suppressed the mRNA and protein levels of Nrf2 in lung cancer cells (Fig. 3A and B). However, the mRNA and protein levels of Nrf2 were upregulated in the miR-144-3p inhibitor-transfected lung cancer cells (Fig. 3C and D). Concurrently, the expression of Nrf2 downstream target genes, which are involved in drug resistance, were examined. As revealed in Fig. 3E-H, the expression of HO-1, NQO1 and Bcl-2 was downregulated significantly in the miR-144-3p mimic-transfected group compared to the miR-144-3p mimic-NC group. However, the miR-144-3p inhibitor-transfected group exhibited the opposite results. Then, we investigated the expression of caspase-3 by Real-time PCR and western blot assays. The results confirmed that the expression of caspase-3 and miR-144-3p exhibited a positive association trend (Fig. 3E-H). These results revealed that miR-144-3p could inhibit the expression of Nrf2 against the drug resistance of lung cancer cells.

To further determine that the effect of miR-144-3p in cisplatin resistance was achieved by regulating Nrf2, we generated two stable lung cancer cell lines with Nrf2 knocked down (shNrf2 group). Then miR-144-3p was overexpressed or inhibited in the shNrf2 stable lung cancer cells. We used Real-time PCR to assess the mRNA levels of HO-1, NQO1, Bcl-2 and caspase-3. Concurrently, western blotting was also used to detect the protein levels of HO-1, NQO1, Bcl-2 and caspase-3. The results confirmed that, when Nrf2 was knocked down, miR-144-3p lost the ability to regulate the resistance in cisplatin-treated lung cancer cells (Fig. 4A and B). All of the aforementioned results indicated that miR-144-3p required Nrf2 to promote cisplatin sensitivity in lung cancer cells.

As reported in research, miRNAs are known to negatively modulate the expression of targeted genes by completely or partially binding with the 3′-untranslated regions (3′-UTR). Thus, we constructed two plasmids of luciferase reporter gene vectors named Nrf2-3′-UTR-WT and Nrf2-3′-UTR-MUT respectively. We then examined the effect of miR-144-3p on the Nrf2 3′-UTR by luciferase assay. The results confirmed that miR-144-3p could bind to the 3′-UTR region of Nrf2 directly, reducing the mRNA level of Nrf2 (Fig. 5A and B). The experimental results revealed that miR-144-3p mediated Nrf2 expression by targeting the 3′-UTR in lung cancer cells.

Previous research confirmed that miR-144-3p could inhibit the cisplatin resistance of lung cancer cells via Nrf2, and we revealed the possible molecular mechanisms between miR-144-3p and Nrf2 in this process. Notably, we found that Nrf2 could reverse regulate the expression of miR-144-3p. There was one potential ARE box on the miR-144-3p promoter region. As revealed in Fig. 6A and B, the results of Real-time PCR indicated that the mRNA level of miR-144-3p was positively associated with the mRNA level of Nrf2, whether Nrf2 was overexpressed or knocked down. Nrf2 regulated target genes by binding to the ARE box existing in the promoter region. There was one potential ARE box on the miR-144-3p promoter region. The results of the luciferase assay indicated that the transcriptional activity of miR-144-3p promoter could be upregulated by Nrf2. However, the miR-144-3p transcriptional activity was not affected when the ARE box was mutated (Fig. 6C). These results may indicate that Nrf2 affected the miR-144-3p transcriptional activity by binding to the ARE box. To further assess the mechanism of this regulation, we used chromatin immunoprecipitation (ChIP) to investigate in lung cancer cells. Data revealed that Nrf2 could be combined with the ARE box, consistent with results of the luciferase assay (Fig. 6D).

In order to clarify whether miR-144-3p played a crucial role to cisplatin resistance in other tumor cells, we exposed HepG2, HeLa, MCF-7 to cisplatin with rising levels of concentration for 24 h after transfection miR-144-3p mimic or miR-144-3p inhibitor (Fig. 7A-C). The viability of these tumor cells that were transfected with miR-144-3p were more sensitive to cisplatin in comparison with the mimic NC group. Conversely, the resistance of cells after transfection the miR-144-3p inhibitor was significantly enhanced to cisplatin. These results indicated that miR-144-3p could also regulate cisplatin resistance in other tumor cells.