A comprehensive search for microRNAs with expression profiles modulated by oncogenic KRAS: Potential involvement of miR-31 in lung carcinogenesis
- Koji Okudela
- Takeshisa Suzuki
- Shigeaki Umeda
- Yoko Tateishi
- Hideaki Mitsui
- Yohei Miyagi
- Kenichi Ohashi
- Published online on: July 18, 2014 https://doi.org/10.3892/or.2014.3339
- Pages: 1374-1384
Lung cancer is one of the most common causes of cancer-related mortality in the developed world (1,2). Recurrence has been reported in a large proportion of lung cancer patients in spite of successful resection of the primary tumor (1,2). Although some lung tumors are known to be sensitive to conventional chemotherapeutic agents or certain molecular targeting agents, many are not (3,4). Thus, further understanding of the molecular mechanisms involved in lung carcinogenesis is essential for developing novel therapeutic strategies.
Our previous studies, which involved a comprehensive search for the downstream targets of oncogenic KRAS, identified important molecules involved in lung carcinogenesis (3,5). Mutations in driver oncogenes, such as KRAS, EGFR, BRAF, and ALK, were found to mutually and exclusively occur, and each of these mutations has been shown to transmit the common essential oncogenic signal that promotes carcinogenesis (6–8). Therefore, the downstream targets of oncogenic KRAS participate not only in KRAS-mediated carcinogenesis, but also in other oncogene-mediated carcinogeneses (3,5). Thus, an investigation into the downstream targets of oncogenic KRAS is considered to be an available strategy that can identify common important molecular mechanisms involved in lung cancer.
Small non-protein coding RNAs that regulate messenger RNA levels, namely microRNAs (miRNAs), have been identified and subsequently implicated in the pathogenesis of various diseases. The present study investigated the expression profiles of miRNAs modulated by oncogenic KRAS in airway epithelial cells. We focused on miR-31 from the miRNAs that were differentially expressed, and evaluated its potential role in the development of lung cancer.
Materials and methods
Cell lines and culture
An immortalized human airway epithelial cell line [16HBE14o, simian virus 40 (SV40)-transformed human bronchial epithelial cells] described by Cozens et al (9) was kindly provided by D.C. Grunert (California Pacific Medical Center Research Institute). A sub-clone of 16HBE14o cells, described as NHBE-T in this study, was used. Human lung cancer cell lines (A549, H358, H2087, H23, EKVX, H226, H827, H1819, H441, H4006, HOP62, H1299 and H460) and a human embryonic kidney cell line (HEK293T) were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). The human lung cancer cell line LC2/ad was purchased from the Riken Cell Bank (Tsukuba, Japan). The human lung cancer cell lines, PC1, PC9 and HARA were from Immuno-Biological Laboratories Co. (Gunma, Japan). The human lung cancer cell lines, TKB1, TKB2, TKB4, TKB5, TKB6, TKB7, TKB8, TKB14 and TKB20 were obtained from Dr Hiroshi Kamma via Dr Takuya Yazawa (Kyorin University School of Medicine) (10). Primary small airway epithelial cells (SAEC) were purchased from Sanko Kagaku (Tokyo, Japan).
The pro-retrovirus vector bearing KRAS [wild-type (WLD), V12, H61, V12/S35, V12/G37 and V12/C40 mutants] (11) and PIK3CA (WLD, E545K mutant, and H1047R mutant) were previously described (12). Complementary DNA (cDNA) coding EGFR (NM_001964) and BRAF (NM_004333) was PCR-amplified using cDNA from NHBE-T cells as a template and inserted into the pro-retrovirus vector pQCXIP (BD Clontech, Palo Alto, CA, USA). Mutant EGFR (deletion: E745-A750) was PCR-amplified using cDNA from the PC9 lung cancer cell line. EGFR (L858R mutant) and BRAF (V600E mutant) were generated by the method of site-directed mutagenesis. DNA fragments including the miR-31 coding region (MIMAT0000089) flanking approximately a 100 base-pair margin in both directions was PCR-amplified and inserted into the pro-retrovirus vector pLHCX (BD Clontech). Vectors bearing a sense and antisense strand of cDNA were obtained. A pNsi-based miR-31 blocking vector was purchased from Takara Bio Inc. (Kyoto, Japan).
Retroviral-mediated gene transfer
The pro-retrovirus vectors bearing the desired constructs and the pLC10A1 retrovirus-packaging vector (Imgenex, San Diego, CA, USA) were cotransfected into HEK293T cells with Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA, USA). Forty-eight hours after the transfection, conditioned medium was recovered as a viral solution. The desired genes were introduced by incubating cells with the viral solution containing 10 μg/ml of polybren (Sigma, St. Louis, MO, USA). Cells stably expressing the desired genes were selected with 250 μg/ml of hygromycin B or 1,000 μg/ml of neomycin (both from Invitrogen). The pooled clones were used for analyses as follows.
miRNA expression profiling
miRNA expression profiles in empty vector (MOCK)-, KRAS/WLD-, KRAS/V12-, and KRAS/H61-vector transduced NHBE-T cells were comprehensively evaluated with human miRNA microarray (Human miRNA Microarray kit release 16.0, 8×60K, cat. no. G4870A; Agilent Technologies, Santa Clara, CA, USA). Total RNA was extracted from the cells immediately after completing the selection (5 days post-transduction) using the RNA easy kit (Qiagen, Hilden, Germany). Labeling and hybridization was performed according to the manufacturer’s recommendations (Agilent Technologies).
Colony formation assays
Cells (appropriate count of 1.0×104 to 5.0×104) were seeded onto a 10 cm culture dish (Iwaki, Tokyo, Japan), and grown for 10 days. Cells were fixed with methanol and Giemsa-stained, and colonies visible in scanned images were counted.
Growth curve assays
Cells (2.5×105) were seeded onto a 10 cm culture dish, and were grown in DMEM (Sigma) with 10 FBS (Sigma) to a semi-confluent state for 5–7 days. Cells were counted, and 2.5×105 cells were seeded again onto a 10 cm dish. Several passages were repeated in the same manner. The sum of population doublings (PDLs) at each point was calculated by the formula SPDLn = log2 (countn/2.5×105) + SPDLn−1.
Soft agar colony formation assays
Cells (1.25×104) were cultured and grown in 1 ml of DMEM-based 0.3% agar (agar noble; Becton-Dickinson, Sparks, MD, USA) containing 10% FBS in 3.5 cm culture dishes (Iwaki) for 4 weeks. The agars were fixed with a buffered 4% paraformaldehyde solution, and colonies visible in scanned images were counted.
Treatment with an inhibitor for MEK and PI3K
Following infection of the retrovirus vector, cells stably expressing KRAS were selected for 3 days, and were then harvested. Cells were treated with an inhibitor for MEK (PD98059, 50 μM) or PI3K (LY294002, 50 μM) [both from Cell Signaling Technology (CST), Danvers, MA, USA] and their combination for the last 24 h before the harvest.
Comprehensive search for the downstream target of miR-31
Gene expression in the empty vector (MOCK)-, miR-31 sense strand (SS)-, and miR-31 antisense (AS) strand-transduced NHBE-T cells was comprehensively evaluated with a human gene chip microarray (SurePrint G3 Human Gene Expression 8×60K v2 Microarray kit, cat. no. G4851B; Agilent Technologies). Total RNA was extracted from the cells immediately after completing the selection (5 days post-transduction) using the RNA easy kit (Qiagen). Labeling and hybridization were performed according to the manufacturer’s recommendations (Agilent Technologies). Transcripts whose signal values in the miR-31 SS-transduced cells were 5-fold higher or lower than that in both the empty vector-transduced cells and the miR-31 antisense strand-transduced cells were extracted according to the same flow chart as described in a previous study (13).
miRNA expression profiling modulated by oncogenic KRAS
A comprehensive evaluation of the expression of miRNA revealed that oncogenic KRAS-transduced cells had different expression profiles from empty vector- and WLD KRAS-transduced cells, as KRAS/V12- and KRAS/H61-transduced cells were classified into a distant branch from the others on a dendrogram described by hierarchical clustering analysis (Fig. 1). The miRNAs that had higher or lower levels in oncogenic KRAS (KRAS/V12 and KRAS/H61)-transduced cells than in both mock (MOCK)- and WLD KRAS (KRAS/WLD)-transduced cells are listed in Table I. Quantitative RT-PCR analysis confirmed that the transduction of oncogenic KRAS induced the largest change in let-7i and miR-31 (Fig. 2). Therefore, we focused on these two miRNAs.
Empty vector (MOCK)-, wild-type KRAS/WLD, oncogenic KRAS/V12, and oncogenic KRAS/H61 were transduced into NHBE-T cells. miRNA expression in the transfectants was comprehensively evaluated with a gene chip microarray. A hierarchical clustering analysis (Ward’s method) was performed and a dendrogram was described.
Level of miRNA whose expression is modulated by oncogenic KRAS. The copy number of the targeted miRNA and U6 snRNA was measured by quantitative RT-PCR. The level of miRNA normalized to that of U6 snRNA is presented. MOCK, empty vector-transduced NHBE-T; WLD, wild-type KRAS-transduced NHBE-T; V12, oncogenic mutant of KRAS/V12-transduced NHBE-T; H61, oncogenic mutant of KRAS/H61-transduced NHBE-T.
Let-7i and miR-31 expression in lung cancer cell lines
Let-7i and miR-31 expression levels were analyzed by quantitative RT-PCR analysis. Let-7i was expressed at various levels in all the cells examined. Let-7i levels were higher in cancer cell lines than in non-cancerous immortalized cell lines and primary airway epithelial cells (Fig. 3). In contrast, miR-31 expression levels were lower in cancer cell lines, and some cell lines [27.5%, 11/40, the loss of miR-31 expression was frequently detected in LCC cell lines (80.0%, 4/5)] completely lost its expression (Fig. 2). These results prompted us to further investigate the potential significance of the loss of miR-31 expression in lung carcinogenesis.
Level of let-7i and miR-31 in non-cancerous and cancer cell lines. The copy number of let-7i, miR-31 and U6 snRNA was measured by quantitative RT-PCR. The level of (A) let-7i or (B) miR-31 normalized to that of U6 snRNA is presented. AEC, airway epithelial cells (non-cancerous cells); ADC, adenocarcinoma cell lines; SQC, squamous cell carcinoma cell lines; LCC, large cell carcinoma cell lines; SCC, small cell carcinoma cell lines.
Effect of the restoration and knockdown of miR-31 on cell growth
The restoration of the miR-31 SS, but not the antisense strand, markedly suppressed the growth of cancer cell lines that lost miR-31 expression, as it decreased the formation of colonies and prolonged the doubling time (Fig. 4, TKB1 and H441 cells were examined. A representative result in TKB1 cells is shown). On the other hand, the knockdown of miR-31 expression in cell lines that retained its expression resulted in an enhancement in their growth activities in both ordinary media and soft agar (Fig. 5, HARA and A549 cells were examined. A representative result in HARA cells is shown).
Biological effect of the restoration of miR-31 in lung cancers losing its expression. The representative results from the examination of TKB1 cells are presented. Empty vector (MOCK), the sense strand of miR-31 (SS), and the antisense strand of miR-31 (AS) were transduced. Following a brief selection, the surviving cells were harvested and counted, and 2.0×104 were re-seeded onto a 10 cm dish. (A) After 14 days, the cells were methanol-fixed and Giemsa-stained. (B) The means and standard deviations (error bars) of colony counts from triplicate experiments are presented. (C) Cells selected were grown and passed several times. Cumulated population doublings are presented. Cells harvested immediately after the selection process were examined for the expression of miR-31 and U6 snRNA by quantitative RT-PCR. (D) The level of miR-31 normalized to that of U6 snRNA is presented.
Biological effect of the knockdown of miR-31 in lung cancer cells retaining its expression. The representative results from the examination of HARA cells are presented. The control vector bearing the scrambled random sequence (scl) and the inhibitory vector coding short hairpin RNA targeting for miR-31 (α-miR-31) were transduced. Following a selection for 10 days, the surviving cells were harvested and counted, and 2.0×104 were re-seeded onto a 10 cm dish. (A) After 14 days, the cells were methanol-fixed and Giemsa-stained. (B) The means and standard deviations (error bars) of colony counts from triplicate experiments are presented. The cells after the selection were grown and passed several times. (C) Cumulated population doublings are described. The selected cells (1.25×104) were cultured and grown in 1 ml of DMEM-based 0.3% agar containing 10% FBS in 3.5 cm culture dishes for 4 weeks. (D) The agars were fixed with a buffered 4% paraformaldehyde solution. (E) The means and standard deviations (error bars) of colony counts from triplicate experiments are presented. The cells harvested immediately after the selection process were examined for expression of miR-31 and U6 snRNA by quantitative RT-PCR. (F) The level of miR-31 normalized to that of U6 snRNA is presented.
Comprehensive search for potential genes modulated by miR-31
A comprehensive evaluation of gene expression profiles revealed that miR-31 SS-transduced cells had a different expression profile from empty vector (mock)- and miR-31 antisense strand (AS)-transduced cells, as miR-31SS-transduced cells were classified into the most distant branch on a dendrogram described from a hierarchical clustering analysis (Fig. 6). Transcripts with expression levels that were >5-fold higher in miR-31SS-transduced cells than in mock- and miR-31AS-transduced cells were extracted (Tables II and III).
mRNA expression in the empty vector (MOCK)-, the miR-31 sense strand (SS), and the miR-31 antisense strand (AS)-transduced NHBE-T cells was comprehensively evaluated with a gene chip microarray. A hierarchical clustering analysis (Ward’s Method) was performed and a dendrogram was described.
Involvement of different oncogenic pathways in the induction of miR-31
The expression of miR-31 was also induced by the transduction of oncogenic EGFR [deletion mutation (deletion: E746-A750), point mutation L858R], but not by oncogenic BRAF (V600E) or oncogenic PIK3CA (E545K, H1047R) (Fig. 7). The expression of miR-31 was consistently not induced by specific mutants in KRAS that activated MEK, RAL-GDS or PI3K (Fig. 7). Inhibitors for MEK (the BRAF-MEK-ERK pathway) or PI3Kinase (the PIK3CA-mediated pathway) interfered with the induction of miR-31 by oncogenic KRAS.
The modulation of miR-31 expression through the essential oncogenic pathway. (A) The level of miR-31 in NHBE-T cells transduced with different mutants of KRAS, EGFR, BRAF and PIK3CA was evaluated. The copy number of miR-31 and U6 snRNA was measured by quantitative RT-PCR. The level of miR-31 normalized to that of U6 snRNA is presented. PRNT, parental NHBE-T; MOCK, empty vector-transduced NHBE-T; WLD, wild-type KRAS-transduced NHBE-T; V12, oncogenic mutant of KRAS/V12-transduced NHBE-T; H61, oncogenic mutant of KRAS/H61-transduced NHBE-T; S35, a mutant of KRAS activating only the MEK pathway-transduced NHBE-T; G37, a mutant of KRAS activating only the Ral-GDS pathway-transduced NHBE-T; C40, a mutant of KRAS activating only the PI3K pathway-transduced NHBE-T. (B) The level of miR-31 in NHBE-T cells treated with the inhibitors for MEK (PD98059, 50 μM), PI3K (LY294002, 50 μM), and their combination (PD98/LY29, 50 μM each) for 24 h was evaluated. CTL, control; VIC, vehicle.
The findings of an initial study on breast cancer suggested that miR-31 could be a tumor suppressor that especially inhibited the invasive and metastatic spread of neoplastic cells (14). Previous studies have supported this initial observation and identified the potential molecular mechanisms responsible for the inhibition of invasion and metastasis by miR-31 (15). miR-31 was also shown to be downregulated in gastric cancer and was suggested to function as a tumor suppressor (16). In contrast, previous studies reported that miR-31 was upregulated in colorectal cancer and may also promote the invasive and metastatic spread of neoplastic cells (17,18). Thus, the potential role of miR-31 in carcinogenesis may differ depending on the type of cancer. The expression of miR-31 in lung cancer has generally been reported to be higher in tumor tissue than in the corresponding non-tumorous tissue (19–25). However, miR-31 levels have been shown to vary in primary lung tumors. Some tumors expressed miR-31 at lower levels than the corresponding non-tumorous tissue or did not express it at all. miR-31 expression was markedly reduced or completely lost in some lung cancer cell lines (Fig. 3). This result suggested that miR-31 may be a suppressor in lung carcinogenesis. The restoration of miR-31 in lung cancer cell lines that had lost its expression markedly attenuated their growth activities (Fig. 4). The knockdown of miR-31 expression in cell lines that retained its expression enhanced oncogenic phenotypes such as anchorage-independent growth activity (Fig. 5). These results supported miR-31 being a suppressor. However, a recent study demonstrated the oncogenic role of miR-31 in lung carcinogenesis using experiments with different cell lines from the ones used in the present study (26). miR-31 may play a pleiotropic role in the development of individual tumors. Further investigations are required to resolve this issue.
In the present study, miR-31 was upregulated not only by oncogenic KRAS, but also by oncogenic EGFR in vitro (Fig. 7). Neither oncogenic BRAF nor oncogenic PIK3CA induced its expression (Fig. 7). Similarly, the specific mutant of KRAS that activated the MEK-mediated pathway only and another specific mutant that activated the PI3K-mediated pathway only did not induce the expression of miR-31 (Fig. 7). On the other hand, the MEK inhibitor, but not PI3K inhibitor, attenuated the oncogenic KRAS-induced expression of miR-31 (Fig. 7). These results suggest that miR-31 could be a common player that regulates the KRAS/EGFR-mediated essential oncogenic pathway (Fig. 8). The MEK-mediated pathway may be necessary, but not sufficient for the induction of miR-31 expression (Fig. 8). The PI3K-mediated pathway may not be involved in its regulation (Fig. 8). A novel pathway may be inducing the expression of miR-31 in cooperation with the MEK-mediated pathway.
The putative role of miR-31 in carcinogenesis is described in a scheme. Physiologically, miR-31 may be induced by stimuli promoting cell growth in response to tissue damage and may control regenerative reactions (the left panel). In carcinogenesis, autonomous growth stimuli due to oncogenic mutations may also induce miR-31 that may interfere with unlimited growth. If the machinery to induce miR-31 is disrupted, autonomous stimuli for growth is augmented and further promote the progression of carcinogenesis (the right panel).
This comprehensive search for the potential target of miR-31 revealed that the downregulated transcripts included many molecules mediating the cytokine/chemokine signaling pathway (Tables II and III). TNFSF15, in particular, has been published on an online site as the predicted downstream target for miR-31 (http://mirdb.org/miRDB/index.html). TNFSF15, a member of the tumor necrosis factor (TNF) and TNF receptor superfamilies, regulates cell growth and apoptosis in an autocrine manner, and has been reported to suppress cancer cell growth by inhibiting angiogenesis (27). Thus, TNFSF15 could be a factor that participates in miR-31-induced growth modulations. Future studies that focus on the potential downstream targets of miR-31 may provide insight into the molecular mechanisms involved in lung cancer.
The present study comprehensively searched miRNAs, the expressions of which were regulated by oncogenic KRAS, and focused on miR-31 in order to investigate its potential involvement in lung carcinogenesis. The expression of miR-31 was markedly attenuated in lung cancer cell lines. The restoration of miR-31 in lung cancer cell lines that lost its expression attenuated their growth activities. The knockdown of miR-31 expression in cell lines that retained its expression enhanced oncogenic phenotypes. These results suggest that miR-31 may be a suppressor, the loss of which may promote lung carcinogenesis.
This study was supported by the Japanese Ministry of Education, Culture, Sports and Science (Tokyo Japan), and by a grant from the Yokohama Medical Facility (Yokohama, Japan). We especially thank Emi Honda and Misa Otara (Kanagawa Prefectural Cardiovascular and Respiratory Center Hospital, Yokohama, Japan) for their assistance.
Hoffman PC, Cohen EE, Masters GA, et al: Carboplatin plus vinorelbine with concomitant radiation therapy in advanced non-small cell lung cancer: a phase I study. Lung Cancer. 38:65–71. 2002. View Article : Google Scholar : PubMed/NCBI
Okudela K, Yazawa T, Ishii J, et al: Down-regulation of FXYD3 expression in human lung cancers: its mechanism and potential role in carcinogenesis. Am J Pathol. 175:2646–2656. 2009. View Article : Google Scholar : PubMed/NCBI
van Eijk R, Licht J, Schrumpf M, et al: Rapid KRAS, EGFR, BRAF and PIK3CA mutation analysis of fine needle aspirates from non-small-cell lung cancer using allele-specific qPCR. PLoS One. 6:e177912011.
Xu J, He J, Yang H, et al: Somatic mutation analysis of EGFR, KRAS, BRAF and PIK3CA in 861 patients with non-small cell lung cancer. Cancer Biomark. 10:63–69. 2012.PubMed/NCBI
Cozens AL, Yezzi MJ, Kunzelmann K, et al: CFTR expression and chloride secretion in polarized immortal human bronchial epithelial cells. Am J Respir Cell Mol Biol. 10:38–47. 1994. View Article : Google Scholar : PubMed/NCBI
Yazawa T, Kamma H, Fujiwara M, et al: Lack of class II transactivator causes severe deficiency of HLA-DR expression in small cell lung cancer. J Pathol. 187:191–199. 1999. View Article : Google Scholar
Okudela K, Yazawa T, Suzuki T, Sugimura H and Kitamura H: Role of 3′-phosphoinositides in oncogenic KRAS-induced modulation of shape and motility of airway epithelial cells. Pathol Int. 59:28–37. 2009.
Okudela K, Yazawa T, Woo T, et al: Down-regulation of DUSP6 expression in lung cancer: its mechanism and potential role in carcinogenesis. Am J Pathol. 175:867–881. 2009. View Article : Google Scholar : PubMed/NCBI
Cottonham CL, Kaneko S and Xu L: miR-21 and miR-31 converge on TIAM1 to regulate migration and invasion of colon carcinoma cells. J Biol Chem. 285:35293–35302. 2010. View Article : Google Scholar : PubMed/NCBI
Slaby O, Svoboda M, Fabian P, et al: Altered expression of miR-21, miR-31, miR-143 and miR-145 is related to clinicopathologic features of colorectal cancer. Oncology. 72:397–402. 2007. View Article : Google Scholar : PubMed/NCBI
Guan P, Yin Z, Li X, Wu W and Zhou B: Meta-analysis of human lung cancer microRNA expression profiling studies comparing cancer tissues with normal tissues. J Exp Clin Cancer Res. 31:542012. View Article : Google Scholar : PubMed/NCBI
Jang JS, Jeon HS, Sun Z, et al: Increased miR-708 expression in NSCLC and its association with poor survival in lung adenocarcinoma from never smokers. Clin Cancer Res. 18:3658–3667. 2012. View Article : Google Scholar : PubMed/NCBI
Gao W, Yu Y, Cao H, et al: Deregulated expression of miR-21, miR-143 and miR-181a in non small cell lung cancer is related to clinicopathologic characteristics or patient prognosis. Biomed Pharmacother. 64:399–408. 2010. View Article : Google Scholar : PubMed/NCBI
Võsa U, Vooder T, Kolde R, et al: Identification of miR-374a as a prognostic marker for survival in patients with early-stage nonsmall cell lung cancer. Genes Chromosomes Cancer. 50:812–822. 2011.PubMed/NCBI
Xing L, Todd NW, Yu L, Fang H and Jiang F: Early detection of squamous cell lung cancer in sputum by a panel of microRNA markers. Mod Pathol. 23:1157–1164. 2010. View Article : Google Scholar : PubMed/NCBI
Tan X, Qin W, Zhang L, et al: A 5-microRNA signature for lung squamous cell carcinoma diagnosis and hsa-miR-31 for prognosis. Clin Cancer Res. 17:6802–6811. 2011. View Article : Google Scholar : PubMed/NCBI
Meng W, Ye Z, Cui R, et al: MicroRNA-31 predicts the presence of lymph node metastases and survival in lung adenocarcinoma patients. Clin Cancer Res. 19:5423–5433. 2013. View Article : Google Scholar : PubMed/NCBI