C57BL/6J male mice (6–8 week old) were purchased from Guangdong Medical Laboratory Animal Center (Guangzhou, China). Mice were randomly divided into two groups. One group was exposed to CS generated from 9 filter-tipped cigarettes in a 342 L fume chamber (60 cm × 57 cm × 100 cm) each time, twice a day, 6 days per week. Each CS exposure was lasted for over 2 h per time with the interval between two CS exposures more than 4 h. To better demonstrate the effect of CS on the expression levels of Nqo1 and Nqo1-AS1 in lung tissues of mice, mice in the CS group were exposed to CS for 1 week, 1 month and 3 months. Moreover, chronic CS-induced COPD mouse model was constructed as we did before in order to elucidate the distribution of Nqo1-AS1 in lung tissues of mice (Zhang et al., 2018b). The Red Roses cigarettes (manufactured from Guangdong Cigarette Factory) emitting 13 mg tar and 1.3 mg nicotine per cigarette were used in this experiment. Mice in the control group were housed in a smoke-free environment. All experimental procedures were approved by the Animal Care and Use Committee of The First Affiliated Hospital of Guangzhou Medical University.

A total of seven patients with COPD and seven healthy individuals were recruited between March and May at 2018 in The First Affiliated Hospital of Guangzhou Medical University. Clinical data, including age, smoking information and lung function were collected. Blood samples were obtained with written informed consents from all participants. Whole blood samples were drawn and centrifuged at 1,000×g for 15 min. Plasma supernatants were collected and stored at − 80° C until analysis. PBMCs were isolated using lymphocyte separation medium according to the method described previously. This study was approved by the Ethics Committee of The First Affiliated Hospital of Guangzhou Medical University (Ethic Ref No.GZMC 2009-08-1336) and adhered to the Declaration of Helsinki as described previously.

To verify whether Nqo1-AS1 was able to encode protein, bioinformatics analysis and in vitro translation assay were performed as described previously (Liu et al., 2020). Briefly, the RNA sequences of Nqo1-AS1, Nqo1 and Hotair were put into the Coding Potential Calculator http://cpc.cbi.pku.edu.cn/ and Coding-Potential Assessment Tool http://lilab.research.bcm.edu/cpat/index.php. Next, the open reading frame (ORF) sequence of Nqo1-AS1, Nqo1 or Hotair was predicted using the ORF finder database (https://www.ncbi.nlm.nih.gov/orffinder), respectively. Then the predicted ORF sequence of Nqo1-AS1, Nqo1 or Hotair was synthesized and subcloned into the BsrGI and XhoI sites of pcDNA3.1-EGFP vector (Invitrogen). Next, the recombinant plasmid pcDNA3.1-EGFP- Nqo1-AS1 (pc-EGFP- Nqo1-AS1), pcDNA3.1-EGFP-Nqo1 (pc-EGFP-Nqo1) or pcDNA3.1-EGFP- Hotair (pc-EGFP- Hotair) was transfected into the mle-12 cells using Lipofectamine 3000 (Thermo) according to the manufacturer’s instructions, respectively. After transfection for 72 h, the nuclei were stained with DAPI (Beyotime), and the immunofluorescence of cells was observed using a fluorescence microscope. Pc-EGFP-Nqo1 was used as a positive control. Pc-EGFP- Hotair was used as a negative control.

Nqo1-AS1 expression was checked by in situ hybridization in lung tissues of mouse. The RNA ISH probe mixture of Nqo1-AS1, Gapdh or U6 RNA was synthesized and labeled with digoxigenin from Biosense Bioscience Co. Ltd. (Guangzhou, China). The probe sequences for RNA ISH were as follows: Nqo1-AS1 antisense probe: 5′- TAT​TTA​GGT​GTG​TAT​GCA​TAC​GTG​AGC​CAT​GGC​GCG​CCC​TGT​GGA-3′; Nqo1-AS1 sense probe: 5′- TCC​ACA​GGG​CGC​GCC​ATG​GCT​CAC​GTA​TGC​ATA​CAC​ACC​TAA​ATA -3′; Gapdh: 5′-TAA​GCA​GTT​GGT​GGT​GCA​GGA​TGC​ATT​GCT​GAC​AAT​CTT​GAG​TGA​GTT​GTC​ATA​TTT​CTC GTGGTTCACACCCATCA -3′. The Gapdh or U6 RNA probe was used as a positive control. The Nqo1-AS1 sense probe was used as a negative control. RNA ISH was performed as previously described (Mehta-Mujoo et al., 2019). Briefly, the mouse lung tissues were first fixed and embedded with paraffin. Embedded specimens were sectioned at 4 μm thickness. Then sample sections were incubated in graded alcohols, 3% hydrogen peroxide for 30 min and pre-hybridization solution for 2 h. After that, digoxigenin-labeled probes were added in the hybridization solution and incubated with the sections at 37°C overnight in the dark. Next, the sections were incubated with anti-DIG and horseradish peroxidase and observed. The staining scores were assessed based on both immunostaining intensity and the proportion of positive staining cells. The immunostaining intensity was scored 0–3 as follows: 0 (negative staining), 1 (weak staining), 2 (moderate staining) and 3 (dark staining).The proportion of positive staining cells was evaluated as follows: 0 (no positive cells), 1 (< 10%), 2 (10–50%) and 3 (> 50%). Expression of Nqo1-AS1 was evaluated by the final score that was multiplication of the immunostaining intensity and the proportion of positive staining cells. The final scores were divided into two levels: low expression (≤4) and high expression (> 4).

Nuclear and cytoplasmic RNA isolation in the mle-12 cells was performed using the Cytoplasmic and Nuclear RNA Purification Kit (Norgen, Belmont, CA, United States) according to the manufacturer’s instruction.

The mouse alveolar epithelium mle-12 cells were obtained from Shanghai fuxiang biotechnology co., LTD., (Shanghai, China).The mle-12 cells were cultured in DMEM/F12 supplemented with 2% fetal bovine serum (Gibco,United States), 100 IU/mL peniciliin and 100 μg/ml streptomycin, and maintained in a humidified atmosphere of 5% CO₂ at 37°C.

CSE was prepared according to the method described previously (Zhang et al., 2018b). Briefly, two cigarettes (Red Roses, China Tobacco Guangdong Zhongyan Industry CO. Ltd., tar, 13 mg; nicotine, 1.3 mg) without filter were burned and then the smoke was collected and finally bubbled through 10 ml serum-free DMEM medium with the use of a vacuum-pump. The resulting solution was filtered through a 0.22 μm filter to remove particles and bacteria and the pH was adjusted to 7.4. The obtained solution was represented 100% CSE and applied to mle-12 cells within 30 min of preparation.

The full-length (FL) Nqo1-AS1, Nqo1-overlapping region (OL) of Nqo1-AS1, Nqo1-non-overlapping region (NOL) of Nqo1-AS1, FL-Nqo1 mRNA were PCR amplified using the SuperScript® III First-Strand Synthesis System (Invitrogen) and subcloned into the ApaI and NotI , NheI and NotI, NheI and XbaI, Nhel and NotI sites of pcDNA3.1 vector (Invitrogen), named pc-Nqo1-AS1, pc-Nqo1-AS1- OL, pc-Nqo1-AS1-NOL or pc-Nqo1, respectively. The primers used were as follows: pc- Nqo1-AS1:5′-ATAAGAATGCGGCCGCGTTTCTTTGCTTTAGCC-3′ (forward), 5′-TTG​CGG​GCC​CGA​TAG​TTC​TGC​CAT​AAC​AAC-3’ (reverse); pc- Nqo1-AS1-OL: 5′-CTA​GCT​AGC​GAT​GTG​TGA​TGT​ATT​CAT​TTA​TTT​CG-3′ (forward), 5′-ATA​AGA​ATG​CGG​CCG​CGA​TAG​TTC​TGC​CAT​AAC-3′ (reverse); pc- Nqo1-AS1-NOL: 5′-GTT​TCT​TTG​CTT​TAG​CCT​GGC​T-3′ (forward), 5′-AGA​TGG​TGG​AGC​ATG​CCT​TTA​A-3′ (reverse); pc-Nqo1: 5′- CTA​GCT​AGC​AGG​CTC​AGC​TCT​TAC​TAG​CCT​AG-3′ (forward), 5′-ATA​AGA​ATG​CGG​CCG​CGA​TGT​GTG​ATG​TAT​TC-3′ (reverse). The 3′UTR of Nqo1 was PCR amplified using the SuperScript® III First-Strand Synthesis System (Invitrogen) and subcloned into the Xhol and XbaI sites of Dual-Luciferase reporter plasmid pmirGLO vector (Promega), named pmirGLO-Nqo1 3′UTR.The primers used were as follows: 5′-CCG​CTC​GAG​GGA​TTT​TTT​TCC​TAA​CAT​ATA​GTT​AGA​C-3′ (forward), 5′-GCT​CTA​GAG​ATG​TGT​GAT​GTA​TTC​ATT​TAT​TTC​G-3′ (reverse). The pcDNA3.1 empty vector was used as a control. A total of 5  ×  10⁵ mle-12 cells were seeded in 6-well plates, and cultured for 18–24 h, and 80–90% cells were used for transfection. Then transfection was performed on cells using Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions. After transfection for 24 h, cells were treated with 0 and 0.5% CSE for 24 h and then harvested for analysis.

siRNA specially targeting Nqo1-AS1 (Nqo1-AS1 siRNA) and scrambled negative control siRNA (siRNA CTL) were synthesized by GenePharma (Shanghai, China).The Nqo1-AS1 siRNA sequence was 5′-GCA​UGU​UGC​UGU​GUG​CCU​ATT-3′ and the siRNA CTL sequence was 5′-UUC​UCC​GAA​CGU​GUC​ACG​UTT-3′. A 20 μM siRNA solution was transfected into the mle-12 cells using HiPerFect Transfection (Qiagen) according to the manufacturer’s instructions. After 24 h, more than 95% of the mle-12 cells were still viable. Cells were then treated with 0 and 0.5% CSE for 24 h prior to being collected, and analyzed.

Intracellular ROS was measured using 2, 7-dichlorodihydrofluorescein diacetate (DCFH-DA, Beyotime, Shanghai, China) according to the manufacturer’s recommendation. Briefly, the mle-12 cells were seeded into 96-well plates and incubated with 10 mM DCFH-DA for 20 min at 37°C. Cells were washed with 1 × PBS and resuspended in DMEM. Then, DCF fluorescence was detected using fluorescence spectrophotometer (Thermo, MA, United States). The cells treated with Rosup (50 mg/ml) for 30 min were used as positive controls.

Concentrations of MDA in the mle-12 cells, the mouse lung tissues and serums from patients with COPD and healthy individuals were measured using MDA Assay kit (Beyotime, Shanghai, China) according to the manufacturer’s instructions. Briefly, the mle-12 cells and the homogenate of mouse lung tissues were lysed in lysis buffers for 30 min on ice, respectively. The lysates were then centrifuged at 10,000×g for 10 min at 4°C, and the supernatants were collected. The serums from patients with COPD and healthy individuals were collected as well. After being treated with thiobarbituric acid (TBA) working solution, the supernatants and the serums were heated at 100°C for 15 min and then cooled down and centrifuged at 1,000×g for 10 min, respectively. The absorbance was measured spectrophotometrically at 532 nm.

The levels of GSH and GSSG in the mle-12 cells, the mouse lung tissues and serums from patients with COPD and healthy controls were measured using GSH and GSSG Assay Kit (Beyotime, Shanghai, China) according to the protocols of manufacturer. Briefly, the mle-12 cells, the mouse lung tissues and the serums from patients with COPD and healthy individuals were treated with protein removal reagent M solution at 4°C for 10 min and then centrifuged at 10,000×g for 10 min, respectively. The supernatants were collected. Then the GSH test solution and 0.5 mg/ml NADPH solution were added to a 96-well plate containing a standard solution of GSH or the mentioned supernatants. The absorbance was measured spectrophotometrically at 412 nm. Then the content of reduced GSH or GSSG and the ration of reduced GSH/GSSG were calculated.

The total RNA was extracted from mouse lung tissues, cultured cells or PBMCs from patients with COPD and healthy individuals using Trizol reagent (Invitrogen) and reversely transcribed to cDNA using PrimeScript™ RT reagent Kit (TaKaRa, China). QRT-PCR expression analysis was performed on CFX96–C1000 system (Bio-Rad, CA) using SsoFast™ EvaGreen® supermix kit (Bio-Rad). Primers used for qRT-PCR were as follows: mouse Nqo1-AS1 non-overlapping region (Nqo1-AS1-NOL): 5′-TTG​GAA​TGC​TGA​GAC​CCT​GT-3′ (forward), 5′-GGA​GTG​AAA​ACA​CGT​GGC​TT-3′ (reverse); mouse Nqo1-AS1 overlapping region (Nqo1-AS1-OL): 5′-TCG​GGC​TAG​TCC​CAG​TTA​GA-3′ (forward), 5′-AAG​TTA​GTC​CCT​CGG​CCA​TT-3′ (reverse); mouse Nqo1 non-overlapping region (Nqo1-NOL): 5′-GGA​AGC​TGC​AGA​CCT​GGT​GA-3′ (forward), 5′-CCT​TTC​AGA​ATG​GCT​GGC​A-3′ (reverse); mouse Nqo1 overlapping region (Nqo1-OL): 5′-TCG​GGC​TAG​TCC​CAG​TTA​GA (forward), 5′-AAG​TTA​GTC​CCT​CGG​CCA​TT-3′ (reverse); mouse Gapdh: 5′-AGG​TCG​GTG​TGA​ACG​GAT​TTG-3′ (forward), 5′-GGG​GTC​GTT​GAT​GGC​AAC​A-3′ (reverse); Nqo1-AS1 human homologue: 5′-TAT​GGC​AGA​AGG​GAA​TTG​CT (forward), 5′-GCT​TTG​TAA​TTG​AAA​GCA​AGA​AA (reverse); human NQO1: 5′-GAA​GAG​CAC​TGA​TCG​TAC​TGG​C-3′ (forward), 5′-GGA​TAC​TGA​AAG​TTC​GCA​GGG-3′ (reverse); human GAPDH: 5′-ACA​ACT​TTG​GTA​TCG​TGG​AAG​G-3′ (forward), 5′-GCC​ATC​ACG​CCA​CAG​TTT​C-3′ (reverse).A primer sequence for mouse Nqo1-NOL has been previously described (Amara et al., 2012). Primer sequences of mouse Gapdh, human NQO1 and human GAPDH were retrieved from PrimerBank Database (http://pga.mgh.harvard.edu/primerbank/). The relative expression of each gene was normalized to Gapdh expression and calculated using the 2^−△△Ct method.

To detect whether Nqo1-AS1 was associated with Nqo1 mRNA, RNase protection assay was performed as previously described (Xia et al., 2021). Briefly, pc-Nqo1-AS1-OL, pc-Nqo1-AS1-NOL or pcDNA3.1 was cotransfected with pc-Nqo1 into mle-12 cells. After transfection for 48 h, the total RNA was extracted from the cells. RNA sample was digested by DNaseI (Invitrogen) and followed by RNAse A + T cocktail (AM2286, Thermo Fisher Scientific) treatment at 37°C for 30 min. Then, the RNA sample was extracted using RNeasy kits (QIAGEN) and was reversely transcribed to cDNA as described above. Nqo1-AS1 overlapping region of Nqo1 (Nqo1-OL), Nqo1-AS1 non-overlapping region of Nqo1 (Nqo1-NOL) and mouse Gapdh mRNA were amplified by PCR and analyzed by agarose gel electrophoresis. Gapdh PCR product was used as a control. Next, to detect whether Nqo1-AS1 increased Nqo1 mRNA stability, mle-12 cells were transfected with pc-Nqo1-AS1-OL, pc-Nqo1-AS1-NOL or pcDNA3.1 for 24 h, then further exposed to 10 μg/ml α-amanitin (MedChemExpress) for 0 h, 6 h, 12 h, 18 and 24 h. Finally, the cells were harvested and RNA was extracted and analyzed by qRT-PCR.

To detect whether Nqo1-AS1 increased the stability of Nqo1 mRNA by binding to its 3′UTR, luciferase reporter assay was performed. Briefly, pmirGLO-Nqo1-3′UTR was cotransfected with Nqo1-AS1 siRNA or siRNA CTL, pc- Nqo1-AS1-OL, pc- Nqo1-AS1-NOL or pcDNA3.1 into mle-12 cells. After transfection for 48 h, cells were harvested and were lysed with lysis buffer. Firefly and Renilla luciferase activities were measured using dual-luciferase reporter assay kit (Promega) according to the manufacturer’s instructions.

The mle-12 cells and the homogenate of mouse lung tissues were lysed in RIPA lysis buffer with PMSF for 30 min on ice. Total protein concentration was measured using BCA protein assay (Beyotime Biotechnology, China). Protein samples were separated by SDS-PAGE and then transferred to PVDF membranes. The PVDF membranes were blocked with 5% skim milk and then incubated with the primary antibodies NQO1 (1:20,000, Abcam Biotechnology, Cambridge, MA, United States) and β-actin (1:5,000, Abcam Biotechnology, Cambridge, MA, United States) overnight at 4°C. After being washed with TBST, the membranes were incubated with the secondary antibody at room temperature for 2 h. Protein bands were detected with ECL reagents (CoWin Biosciences, China) and then visualized using Tanon 5200 chemiluminescence imaging system (Tanon, Shanghai, China).Scanned images were quantified with Image-Pro 6 software.

Data were presented as mean ± standard error of the mean (SEM). Normality of the variables was evaluated using a Kolmogorov–Smirnov test. Analysis of parametric variables were performed using the Student t test or one-way ANOVA followed by Bonferroni correction for multiple comparisons, while the analysis of non-parametric variables were performed using the chi-square test. All statistical analyses were performed using SPSS version 13.0 software. The correlation between Nqo1-AS1 human homologues expression and the smoking amount of patients with COPD or healthy individuals was evaluated by Pearson’s correlation. A p-value less than 0.05 was regarded as statistically significant.

In previous study, we reported that lncRNA Fantom3_F830212L20 and Nqo1 were co-expressed lncRNA and protein-coding gene (Zhang et al., 2018b). To investigate the association between genome loci of Fantom3_F830212L20 and Nqo1, bioinformatics analysis was performed. The results showed that Fantom3_F830212L20 oriented in antisense direction with respect to Nqo1, which formed a “tail to tail” pairing pattern with 460 bp full complementarity between each other. We named this lncRNA as Nqo1 antisense transcript 1 (Nqo1-AS1) (Figure 1A).RNA ISH revealed that the majority of Nqo1-AS1 expression existed in alveolar epithelial cells of mouse with chronic CS exposure, whereas the positive staining was occasionally observed in mouse without CS exposure (Figure 1B). Moreover, subcellular fractionation assay shown that Nqo1-AS1 mainly located in the cytoplasm of mouse alveolar epithelium (Figure 1C). To verify whether Nqo1-AS1 had a coding potential, bioinformatics analysis and an in vitro translation assay were performed as described previously (Zhang et al., 2017). The Coding Potential Caculator computational algorithm predicted that Nqo1-AS1 had a very low protein coding potential, similar to Hotair, a well-known lncRNA, whereas Nqo1 was predicted to code for protein (Table 1).Then the recombinant plasmid pc-EGFP- Nqo1-AS1 with the predicted ORF sequence of Nqo1-AS1 was overexpressed in the mle-12 cells (Table 2). Pc-EGFP-Nqo1 was used as a positive control, and pc-EGFP- Hotair was used as a negative control. Immunofluorescence staining displayed that EGFP was hardly detected in cells transfected with pc-EGFP- Nqo1-AS1 or pc-EGFP- Hotair, whereas it was easily detectable in cells transfected with Pc-EGFP-Nqo1. These results suggest that Nqo1-AS1 mainly locates in the cytoplasm of mouse alveolar epithelium and has a very low protein coding potential.

The RNA sequence of Nqo1-AS1, HOTAIR or Nqo1 was put into the Coding Potential Calculator (CPC) algorithm version 2. CPC2 was available freely at http://cpc2.cbi.pku.edu.cn. Hotair was used as a negative control. Nqo1 was used as a positive control.

The predicted ORF sequence of Nqo1-AS1, Hotair or Nqo1 was obtained from the ORF finder database (https://www.ncbi.nlm.nih.gov/orffinder), respectively. Hotair was used as a negative control. Nqo1 was used as a positive control.

Nqo1-AS1 human homologue is both positively correlated with smoking amount and Nqo1 mRNA expression in PBMCs of patients with COPD or healthy controls

To assess whether Nqo1-AS1 human homologue expression was associated with smoking amounts and Nqo1 mRNA expression, the expression levels of Nqo1-AS1 human homologue and Nqo1 mRNA in PBMCs of patients with COPD or healthy controls were examined, and the correlation between these two gene expressions and smoking amounts of patients with COPD or healthy controls were analyzed. A total of seven patients with COPD and seven healthy controls were enrolled in this study. The general characteristics of study participants were summarized in Table 3. As compared to the control group, the GSH concentration and the GSH/GSSG ratio in serum of patients with COPD were lower, whereas concentrations of GSSG and MDA were higher (Figures 2A–D). The expression levels of Nqo1-AS1 human homologue and Nqo1 mRNA in PBMCs from patients with COPD were significantly upregualted than those of the control group (Figures 2E,F). Correlation analysis shown that Nqo1-AS1 human homologue expression was both positively associated with smoking amounts and Nqo1 mRNA expression (Figures 2G–I). These results indicate that Nqo1-AS1 human homologue is both positively associated with Nqo1 mRNA expression and smoking amounts of patients with COPD.

Given that Nqo1-AS1 human homologue is positively correlated with Nqo1 mRNA expression in PBMCs of patients with COPD, we speculated that Nqo1-AS1 might also be positively associated with Nqo1 mRNA expression in lung tissue of mice exposed to cigarette smoke. QRT-PCR revealed that both Nqo1-AS1 and Nqo1 mRNA expressions were upregulated in lung tissue of mice exposed to CS for 1 week, 1 month and 3 months in comparison with control animals. Correlation analysis shown that Nqo1-AS1 expression was positively associated with Nqo1 mRNA expression in lung tissues of mice (Figures 3A–C). These results suggest that the expression levels of Nqo1-AS1 and Nqo1 mRNA are elevated in lung tissue of mice exposed to CS, and there is a positive correlation between Nqo1-AS1 and Nqo1 mRNA expression.

Given that CS exposure was closely associated with the expression levels of Nqo1-AS1 (or its human homologue) and Nqo1 mRNA in lung tissues of mouse or PBMCs from patients with COPD, we speculated that CSE might have effects on the expression levels of Nqo1-AS1 and Nqo1 in vitro. The murine alveolar mle-12 cells were treated with CSE at different concentrations for varied time points, and the expression levels of Nqo1-AS1 and Nqo1 mRNA were detected. Compared to the control group, cells treated with 0% CSE, Nqo1-AS1 and Nqo1 mRNA expressions in cells exposed to CSE (0.3–1%) for 6 h, 12 and 24 h were significantly upregulated. Furthermore, Nqo1-AS1 and Nqo1 mRNA expressions were increased after being exposed to CSE in a concentration dependent manner (Figures 4A,B).Additionally, Nqo1-AS1 and Nqo1 mRNA expressions were significantly increased after being exposed to CSE for 6 h, yet both these two gene expressions were gradually decreased in response to the increasing CSE duration (Figures 4C,D). Line chart depicted the trends of Nqo1-AS1 and Nqo1 mRNA expressions in cells exposed to CSE, which elevated with the increase of CSE concentration whereas decreased with the increase of the CSE exposure duration (Figures 4E,F). These results indicate that Nqo1-AS1 and Nqo1 mRNA expressions are closely correlated with CSE concentration, and both of them were elevated mainly in the early stage of exposure of cells to CSE.

Compared to the control group, cells treated with 0% CSE, Nqo1 protein level in mle-12 cells exposed to CSE (0.3–1%) for 6 h, 12 and 24 h were significantly enhanced (Figure 5A). Similarly, compared to cells treated with CSE (0.3–1%) for 0 h, Nqo1 protein level in cells was significantly increased after being exposed to CSE (0.3–1%) for 24 h (Figure 5B). Line chart displayed the trends of Nqo1 protein level in cells exposed to CSE, which was enhanced with the increase of CSE concentration and exposure duration (Figure 5C). These results indicate that CSE enhances Nqo1 protein level of mle-12 in a dose-and time-dependent manner.

Accumulating evidences suggest that CS induced oxidative stress plays a critical role in the pathological mechanism of COPD (Kim et al., 2019). To elucidate the effect of Nqo1-AS1 against CSE-induced oxidative stress, mle-12 cells were transfected with Nqo1-AS1 siRNA or siRNA CTL prior to being treated with 0% CSE or 0.5% CSE for 24 h. Then oxidative stress parameters such as GSH, GSSG, MDA and ROS in cells were measured. The cells treated with Rosup (50 mg/ml) were used as positive controls. Compared to cells transfected with siRNA CTL and followed by 0% CSE treatment for 24 h, the GSH content and [GSH/GSSG] ratios in cells transfected with siRNA CTL and followed by 0.5% CSE treatment were significantly reduced, whereas the levels of GSSG, MDA and ROS were significantly enhanced. Knockdown of Nqo1-AS1 worsen the decrease of GSH content and [GSH/GSSG] ratios in cells due to CSE exposure, whereas increasing the levels of GSSG, MDA and ROS (Figures 6A–E). On the contrary, the overexpression of Nqo1-AS1 was able to rescue the decrease of GSH content and [GSH/GSSG] ratios due to CSE, and to alleviate the CSE-increased MDA and ROS in cells (Figures 6F–J). These results lend strong support to our hypothesis that Nqo1-AS1 has a protective effect on CSE-induced oxidative damage to mle-12 cells in vitro.

It has been demonstrated that Nqo1 functions as a crucial antioxidant enzyme and is able to bind to Serpina1 mRNA thereby having effect on COPD progression (Di Francesco et al., 2016). We then determined whether Nqo1-AS1 regulated the expressions of Nqo1 and Serpina1 expression. The Nqo1-AS1 siRNA was transfecting into mle-12 cells and the interference efficiency was detected (Figure 7A). QRT-PCR and western blotting shown that knockdown of Nqo1-AS1 expression significantly decreased Nqo1 at mRNA and protein levels. Moreover, down regulation of Nqo1-AS1 inhibited the CSE-induced upregulation of Nqo1 (Figures 7B,C). Interestingly, we also observed that silencing Nqo1-AS1 down-regulated the Serpina1 mRNA expression, and even aggravated the decrease of Serpina1 mRNA expression due to CSE (Figure 7D). To better evaluate the effect of Nqo1-AS1 on the CSE-induced Nqo1 and Serpina1 expressions, we further upregulated the Nqo1-AS1 expression through transfecting pc-Nqo1-AS1 plasmids into the mle-12 cells. The transfection efficiency of pc-Nqo1-AS1 was detected (Figure 7E). As expected, the overexpression of Nqo1-AS1 increased Nqo1 at mRNA and protein levels in cells with or without CSE treatment (Figures 7F,G). In addition, overexpressing Nqo1-AS1 not only enhanced Serpina1 mRNA expression but also rescued CSE-induced downregulation of Serpina1 mRNA (Figure 7H). These results suggest that Nqo1-AS1 is able to regulate CSE-induced Nqo1 and Serpina1 expressions.

Since bioinformatics analysis indicated that Nqo1-AS1 was able to form RNA-RNA hybrid with Nqo1 mRNA (Table 4), and Nqo1-AS1 and Nqo1 formed a “tail to tail” pairing pattern with 460 bp full complementarity between each other, we then determined whether Nqo1-AS1 was physically associated with Nqo1 mRNA. RNase protection assay shown that Nqo1-AS1-OL, but not Nqo1-AS1-NOL, protected the overlapping part of Nqo1 (Nqo1-OL) mRNA from RNase digestion by forming the RNA duplexes between Nqo1-AS1 and Nqo1 mRNA, whereas the non-overlapping part of Nqo1 (Nqo1-NOL) mRNA and Gapdh mRNA was totally digested (Figures 8A–C). Gapdh PCR product was used as a control. It was the RNA duplexes formation between the overlapping part of Nqo1-AS1 (Nqo1-AS1-OL) and the overlapping part of Nqo1 (Nqo1-OL) mRNA that protected both of them from RNase digestion (Figure 8D).To further determine whether Nqo1-AS1 regulated Nqo1 mRNA stability, we silenced the Nqo1-AS1 expression by transiently transfecting Nqo1-AS1 siRNA and followed by a-amanitin treatment to block new RNA synthesis in the mle-12 cells. QRT-PCR analysis revealed that knockdown of Nqo1-AS1 decreased Nqo1 mRNA stability (Figure 8E).On the contrary, overexpression of Nqo1-AS1 by transfecting pc- Nqo1-AS1 plasmids into the mle-12 cells increased the stability of Nqo1 mRNA (Figure 8F). As the overlapping region of Nqo1-AS1 antisense paired with Nqo1 3′UTR, we constructed a luciferase reporter plasmid that carried the Nqo1 3′UTR, named pmirGLO-Nqo1-3′UTR.A 594-bp fragment of the Nqo1-3′UTR, which contained the entire overlapping region with Nqo1-AS1, was inserted downstream of the luciferase reporter gene in the pmirGLO-Nqo1-3′UTR vector. We examined the effects of Nqo1-AS1 knockdown or overexpression on pmirGLO-Nqo1-3′UTR activity. The results shown that knockdown of Nqo1-AS1 reduced the luciferase activity of pmirGLO-Nqo1-3′UTR, whereas Overexpressing Nqo1-AS1-OL, but not Nqo1-AS1-NOL, increased the luciferase activity of pmirGLO-Nqo1-3′UTR (Figures 8G,H).Taken together, these results demonstrate that Nqo1-AS1 increases the Nqo1 mRNA stability and promotes the expression of Nqo1 through forming RNA-RNA hybrid with Nqo1 mRNA.

The RNA sequences of Nqo1-AS1 and Nqo1 were put into the lncRNATargets (http://www.cuilab.cn/lnctar) to analyze the potential interaction between each other.