Down-regulation of microRNA-144 in air pollution-related lung cancer

Air pollution has been classified as a group 1 carcinogen in humans, but the underlying tumourigenic mechanisms remain unclear. In Xuanwei city of Yunnan Province, the lung cancer incidence is among the highest in China, owing to severe air pollution generated by the combustion of smoky coal, providing a unique opportunity to dissect lung carcinogenesis. To identify abnormal miRNAs critical for air pollution-related tumourigenesis, we performed microRNA microarray analysis in 6 Xuanwei non-small cell lung cancers (NSCLCs) and 4 NSCLCs from control regions where smoky coal was not used. We found 13 down-regulated and 2 up-regulated miRNAs in Xuanwei NSCLCs. Among them, miR-144 was one of the most significantly down-regulated miRNAs. The expanded experiments showed that miR-144 was down-regulated in 45/51 (88.2%) Xuanwei NSCLCs and 34/54 (63%) control region NSCLCs (p = 0.016). MiR-144 interacted with the oncogene Zeb1 at 2 sites in its 3′ untranslated region, and a decrease in miR-144 resulted in increased Zeb1 expression and an epithelial mesenchymal transition phenotype. Ectopic expression of miR-144 suppressed NSCLCs in vitro and in vivo by targeting Zeb1. These results indicate that down-regulation of miR-144 is critical for air pollution-related lung cancer, and the miR-144-Zeb1 signalling pathway could represent a potential therapeutic target.

Scientific RepoRts | 5:14331 | DOi: 10.1038/srep14331 highly polluted regions (HPR) lends a unique opportunity to dissect carcinogenesis that is specifically related to air pollution. We utilized this opportunity by systematically analyzing the abnormalities in the cancer genomes, genome-wide DNA methylation, non-coding RNAs (miRNAs and lncRNAs), and inflammatory factors, in patients from this region. Alterations found in HPR lung cancers were tested in patients from control regions (CR) where smoky coal was not used to compare the difference in the carcinogenic mechanism between HPR and CR lung cancer. Here, we report our results in the assessment of the expression of miRNAs in HPR non-small cell lung cancers (NSCLCs).

Identification of aberrant miRNAs in Xuanwei NSCLCs. Tumour tissues and adjacent normal
lung tissues were obtained with informed consent from 105 patients (51 from HPR and 54 from CR) with previously untreated lung adenocarcinoma (AD) or squamous cell carcinoma (SCC) (Table 1). Firstly, we performed miRNA microarray analysis for ten NSCLCs (6 from HPR and 4 from CR; Fig. 1A) and compared the expression of the miRNAs in tumour samples to that in their paired normal lung tissues to identify abnormal miRNA expression patterns. In total, we found 23 down-regulated and 7 up-regulated miRNAs in the 10 patients ( Fig. 1A and Table 2). MiR-3195, miR-3656 and miR-144-3p (hereafter, miR-144) were the 3 most significantly down-regulated miRNAs ( Fig. 1 and Table 2). We compared miRNA expression profiles between HPR and CR patients and found 21 miRNAs that were differentially expressed in HPR patients compared to CR cases. These included 13 down-regulated and 2 up-regulated miRNAs that were seen only in HPR and 5 down-regulated and 1 up-regulated miRNAs that were detected only in CR patients. Three miRNAs were down-regulated in NSCLCs from both regions (Fig. 1B). To verify the results of the miRNA microarrays, we tested the expression of miR-3195, miR-3656, miR-144, miR-1915, and miR-451a in additional NSCLCs from HPR and CR. We found that the expression of miR-3195 in 18 NSCLC tumor samples did not differ from their paired normal lung tissues (Fig. 1C). The expression of miR-3656, miR-1915, and miR-451a was tested in 44 to 55 NSCLCs, and the results showed that these miRNAs were down-regulated in tumour samples compared to paired normal lung tissues (Fig. 1C). Notably, the expression of miR-144 was decreased by at least 4-fold in 68/105 (64.8%) NSCLCs compared to matched adjacent normal lung samples (Fig. 1D, E).
Down-regulation of miR-144 is frequently seen in NSCLCs from HPR. We found that the expression of miR-144 was much lower in tumour samples than paired adjacent normal lung tissues in 45/51 (88.2%) and 34/54 (63%) NSCLCs from HPR and CR (Fig. 1F), respectively. The down-regulation of miR-144 in tumor samples was more frequently seen in HPR than CR NSCLCs (p = 0.016), indicating the association between air pollution and low miR-144 expression (Table 1).
To test whether air pollutants could down-regulate miR-144, several PAH compounds including benzo(a)pyrene (BaP), benzo(a)pyrene diol epoxide (BPDE), dibenzo[a, h]anthracene (DBA), and benzo[g, h, i]perylene (BzP) as well as the tobacco-specific carcinogen nicotine-derived nitrosamine ketone (NNK) were used to treat normal human bronchial epithelial 16HBE cells 11   Effects of miR-144 on NSCLC cell migrations. We analysed the expression of miR-144 in human lung cancer cell lines by quantitative RT-PCR and found that the expression of this miRNA was low in the lung cancer cell lines 95D, H1975, EPLC, and A549, whereas 16HBE cells had a relatively high expression level of miR-144 ( Fig. 2A). To determine whether miR-144 was a putative tumour suppressor, an miR-144 mimic was transfected into A549 and 95D cells by using Lipofectamine or into A549-luciferase cells by using lentivirus-mediated cell transfection (Fig. 2B), and cell migration assays were performed. By transwell and wound healing assays, we found that miR-144 dramatically suppressed the migration abilities of the cells (Fig. 2C, D). In contrast, suppression of miR-144 by using an miRNA inhibitor enhanced cell migration in both of the lung cancer cell lines (Fig. 2E, F).
The expression of Zeb1 in HPR NSCLCs. We tested the expression of Zeb1 in the 105 NSCLCs, and found that Zeb1 expression in tumour samples was much higher than in paired normal lung tissues (Fig. 3H). The expression of Zeb2 was also elevated in the patients' tumour samples (Fig. S2), which was consistent with previous reports 16 . Using Pearson correlation analysis, we evaluated the potential correlation between miR-144 and Zeb1 expression and found that miR-144 was inversely associated with Zeb1 expression (Fig. 3I).

Zeb1 inhibition recapitulates the tumour suppressing effect of miR-144.
To evaluate whether down-regulation of Zeb1 was involved in the miR-144-induced suppression of cell migration, A549 and 95D cells were transfected with a Zeb1-specific siRNA (siZeb1) that caused down-regulation of Zeb1 at the mRNA (Fig. 4A) and protein (Fig. 4B) levels. Silencing of Zeb1 led to up-regulation of E-Cadherin and down-regulation of Vimentin (Fig. 4B). Furthermore, we found that silencing of Zeb1 significantly suppressed the cell migration of A549 and 95D cells in the transwell (Fig. 4C) and wound-healing (Fig. 4D) assays. Whereas transfection of miR-144 suppressed the migration of cancer cells, co-transfection of Zeb1 attenuated this effect (Fig. 4E). Consistent with this, transfection of miR-144 into the cells caused down-regulation of Zeb1 and Vimentin and up-regulation of E-Cadherin, whereas ectopic expression of Zeb1 partially reversed these effects (Fig. 4F).

MiR-144 inhibits cancer progression in vivo.
To test the in vivo tumour suppressing effect of miR-144, either the miR-control or miR-144 was stably transfected into A549-luciferase cells, which were then injected into the tail veins of SCID Beige mice. We showed that, compared to mice injected with miR-control-transfected cells, mice injected with miR-144-A549-luciferase cells showed a three-fold reduction in lung tumour volume detected by the IVIS Spectrum Imaging System (Fig. 4G). The body weights of the mice harboring miR-control-A549-luciferase cells decreased rapidly, whereas the body weights of mice bearing miR-144-A549-luciferase cells decreased gradually (Fig. 4H), reflecting the tumour burden and progression. The immunohistochemistry (Fig. 4I) and Western blot (Fig. 4J) assays showed that both Zeb1 and Vimentin were down-regulated and E-Cadherin was up-regulated in tumours harvested from mice inoculated with the miR-144-A549-luciferase cells.

Discussion
MiRNAs are small non-coding RNAs that post-transcriptionally mediate the expression of their target genes by perfect or imperfect binding to the 3′ -UTR 17 , the 5′ -UTR 18 , or the open reading frame of the target mRNAs 19 , thereby causing mRNA degradation or inhibition or enhancement of mRNA translation. It has been estimated that miRNAs regulate more than 60% of human genes 20 . MiRNAs participate in diverse biological processes, including cell proliferation, apoptosis, migration and invasion 21 . Some miRNAs have been shown to be involved in the tumourigenesis of lung cancer, and miRNA expression profiles are diagnostic and prognostic markers of lung cancer 22 . For instance, microRNA-135b promotes lung cancer metastasis by targeting tumour suppressor LZTS1 and multiple key components in the Hippo pathway such as LATS2, β -TrCP and NDR2 23 . MicroRNA-193a-3p and − 5p function as tumour suppressors and inhibit the metastasis of lung cancer by down-regulating the ERBB4/PIK3R3/ mTOR/S6K2 signalling pathway 24 . MiR-21 25 , miR-17-92 26 and miR-196a 27 are up-regulated and function as oncogenes, whereas let-7 28 , miR-194 29 and miR-126 30 suppress carcinogenesis. However, abnormalities in miRNAs in air pollution-related lung cancer have not been investigated.
PAHs, the main carcinogens in HPR 9 , are ubiquitous environmental pollutants that are generated primarily through incomplete combustion of carbon-containing materials. In cellular or animal models, PAHs down-regulate miR-34c 31 , miR-21, miR-221, miR-222, and miR-429 32 , and up-regulate miR-34a 33,34 , miR-181a, miR-181b, and miR-181d 35 . Deng et al. 36 performed a study in healthy male coke oven workers to identify miRNAs associated with PAH exposure, and found that urinary 4-hydroxyphenanthrene and/or plasma BPDE-Alb adducts were associated with lower miR-24-3p, miR-27a-3p, miR-142-5p, and miR-28-5p expression. Urinary 1-hydroxynaphthalene, 2-hydroxynaphthalene, 2-hydroxyphenanthrene, and the sum of monohydroxy-PAHs were associated with higher miR-150-5p expression. Here, we performed miRNA microarrays to screen for abnormal miRNAs in air pollution-related lung cancer, using samples from a unique HPR and found that miR-144 and its family member miR-451 37 were significantly down-regulated in the patients. We expanded this observation and reported that miR-144 was down-regulated in 45/51 (88.2%) HPR NSCLCs, whereas in CR patients miR-144 was suppressed in 34/54 (63%) tumour samples (p = 0.016). Our results provided the first evidence of association between a miRNA and air pollution-related lung cancer. However, treatment of 16HBE cells with the carcinogens BaP, BPDE, BzP, DBA and NNK for up to 15 days did not result in the down-regulation of miR-144, suggesting that other carcinogens may be responsible for the down-regulation of this miRNA; therefore, investigations are warranted to uncover the carcinogen(s).
MiR-144 can modulate TRAIL-induced apoptosis by targeting caspase-3 38 . Down-regulation of miR-144 is associated with colorectal cancer progression via activation of the mTOR signalling pathway 39 . MiR-144 inhibits NSCLC cell growth and induces apoptosis by down-regulating ZFX 40 , and re-establishing miR-144 in gastric cancer restores the chemosensitivity 41 . We showed that Zeb1, a transcriptional repressor 42 that inhibits E-cadherin and promotes EMT and metastasis 12,43 , is a target of miR-144. MiR-144 interacts with Zeb1 via two sites in Zeb1's 3′ -UTR, and mutations in these sites abrogate miR-144's repression of Zeb1 functions (Fig. 3), a result in agreement with previous reports 44, 45 . In lung cancer, the expression of Zeb1 is up-regulated 46 , whereas knockdown of Zeb1 results in dramatic growth inhibition 47 . MiR-144 could also bind Zeb2 at positions 1003-1023 in its 3′ -UTR (Fig. 3A), but in NSCLC cells overexpression or knockdown of miR-144 had no effect on the expression of Zeb2 at the mRNA and protein levels (Fig. 3B, F, G). In NSCLCs, miR-200b and miR-200c, which are able to bind Zeb2 at positions 1015-1036 of its 3′ -UTR, were up-regulated 48,49 , and may therefore antagonize the effects of miR-144 on Zeb2. Hence, our results indicated that miR-144 is an important tumour suppressor that is inactivated during malignant transformation, and the miR-144-Zeb1 signal pathway could represent a rational therapeutic target.

Methods
Patients and tissue samples. Use of the samples was approved by the Institutional Review Board of the Institute of Zoology, Chinese Academy of Sciences, and the local research ethics committees of all participating hospitals. The methods were performed in accordance with the approved guidelines. Tumour tissues and adjacent normal lung tissues were obtained with informed consent from 105 patients with previously untreated lung adenocarcinoma (AD) or squamous cell carcinoma (SCC) at local hospitals. The diagnosis of lung cancer was confirmed by at least 3 pathologists, and the HPR patients enrolled met the following criteria: (1) patients were residents of Xuanwei where the smoky coal was used; (2) patients resided in their communities and never stayed in other regions for a long period of time (6 months or more); (3) patients had previously untreated primary lung cancer; and (4) patients' tissue samples were taken at the time of surgery and quickly frozen in liquid nitrogen. The tumour samples contained a tumour cellularity of greater than 60% and the matched control samples had no tumour content. The clinical and pathological data for these patients are shown in Table 1 and Fig. 1A.
Cell culture. The NSCLC cell lines A549 and 95D, human bronchial epithelial cell line 16HBE, and a human embryonic kidney cell line HEK293T were cultured in DMEM (Hyclone, Logan, UT, USA) medium supplemented with 10% foetal bovine serum (Hyclone). The cells were treated with PAHs (BaP, BPDE, DBA, and BzP) or the tobacco specific carcinogen NNK. Cell viability was estimated by trypan blue dye exclusion analysis, and cell proliferation was measured by the MTT assay 50 .
MiRNA microarray and quantitative RT-PCR. Total RNA was isolated using TRIzol (Invitrogen, Frederick, MD, USA) and the miRNeasy mini kit (Qiagen, Hilden, Germany), quantitated by using the NanoDrop 1000 (Thermo Scientific, Wilmington, DE, USA), labelled by using the miRCURY ™ Hy3 ™ /Hy5 ™ Power labelling kit (Exiqon, Vedbaek, Denmark) and hybridized onto the miRCURY ™ LNA Array (v.16.0; Exiqon). Following the washing steps, the slides were scanned using the Axon GenePix 4000B microarray scanner (Axon Instruments, Foster City, CA, USA), and images were imported into the GenePix Pro 6.0 software (Axon Instruments) for grid alignment and data extraction. All expressed data were normalized using the Median normalization method 51  Wound-healing assay and in vitro migration assay. For the wound-healing assay, the cells (4 × 10 5 /wells) were seeded into six-well plates and transfected the next day with a miR-144 mimic or a miR-control at a final concentration of 100 nM. Twenty-four hours after transfection, wounds were created in the cell monolayer using a p200 micropipette tip. The healing process was followed for the next 48 hours. For the transwell assay, transwell inserts (6.5 mm diameter and 8 μ m pore size; Corning Inc., Corning, NY, USA) were rehydrated by adding serum-free medium for at least 1 hour. The cells transfected with the miR-144 mimic in serum-free medium were seeded (2 × 10 4 cells) into the inserts (the upper chamber). Complete medium was used as a chemoattractant in the bottom chamber. After 24 hours of migration, the cells in the upper surface of the insert membrane were removed by wiping with a cotton swab, and cells in the lower surface were fixed with methanol, stained with crystal violet, and counted by microscopy.

Lentivirus-mediated cell transfection and transduction.
Pre-miR-144 DNA sequences were amplified from human genomic DNA, subcloned into the EcoR I and BamH I sites downstream of the CMV promoter in the pCDH vector (kindly provided by Dr. Wanzhu Jin at the Institute of Zoology, Chinese Academy of Sciences), and verified by DNA sequencing. The primers used for the genomic PCR amplification of miR-144 were as follows: 5′ -GCGCGAATTCGAGATCTTAACAGACCCTAGCTC -3′ (forward primer) and 5′ -GCGCGGATCCGTGCCCTGGCAGTCAGTAGG-3′ (reverse primer). Infectious virus particles were harvested 48 hours after co-transfection of pCDH-miR-144 or pCDH with the lentivirus packing vector (psPAX2 and pMD2G) into HEK293FT cells. A549 cells were infected with lentiviruses in medium containing polybrene (8 mg/ml). One week after infection, the cells were sorted by flow cytometry.
Scientific RepoRts | 5:14331 | DOi: 10.1038/srep14331 Western Blotting. Cells were lysed in RIPA buffer supplemented with a protease inhibitors cocktail (Sigma, St. Louis, MO, USA). Proteins (20 μ g) were subjected to 10-15% SDS-PAGE, electrophoresed and transferred on to a nitrocellulose membrane. After blocking with 5% non-fat milk in Tris-buffered saline, the membrane was washed and incubated with the indicated primary and secondary antibodies and detected using the Luminescent Image Analyser LSA 4000 (GE, Fairfield, CO, USA). Animal studies. The animal studies were approved by the Institutional Review Board of the Institute of Zoology, Chinese Academy of Sciences. The methods were performed in accordance with the approved guidelines. Six-week-old SCID Beige mice were maintained under specific pathogen-free (SPF) conditions. A549-luciferase cells (1 × 10 6 ) stably expressing miR-144 or the miR-control were injected into the lateral tail veins of the mice (6 per group). After 30 days, the tumours were monitored with the IVIS Spectrum Imaging System (Caliper Life Sciences; Hopkinton, MA, USA). For the immunohistochemistry (IHC) assay, sections were fixed in formalin and embedded in paraffin, incubated with primary antibodies overnight, and then incubated with anti-rabbit IgG secondary antibodies. Detection was conducted using 3, 3′ -diaminobenzidine (DAB, Zhongshan Golden Bridge Biotechnology Co., Ltd, Beijing, China) and haematoxylin.
Statistical analysis. Experimental data were presented as the mean ± SD of three independent experiments. The differences between groups were estimated using an independent two-tailed Student's t-test (normal distribution data) or Wilcoxon rank sum test (non-normal distribution data), and the association between miR-144 expression and the Zeb1 level was analysed by Pearson correlation analysis. P values less than 0.05 were considered statistically significant in all cases.