Gene 534 (2014) 66–71
Contents lists available at ScienceDirect
Gene journal homepage: www.elsevier.com/locate/gene
A common polymorphism in pre-microRNA-146a is associated with lung cancer risk in a Korean population Hyo-Sung Jeon a,b,1, Yong Hoon Lee c,1, Shin Yup Lee b,c,1, Ji-Ae Jang a, Yi-Young Choi a, Seung Soo Yoo b,c, Won Kee Lee d, Jin Eun Choi a,b, Ji Woong Son e, Young Mo Kang c, Jae Yong Park a,b,c,⁎ a
Department of Biochemistry and Cell Biology, School of Medicine, Kyungpook National University, Dong In 2Ga 101, Daegu 700-422, South Korea Lung Cancer Center, Kyungpook National University Medical Center, 807, Hoguk-ro, Buk-gu, Daegu 702-210, South Korea Department of Internal Medicine, School of Medicine, Kyungpook National University, Dong In 2Ga 101, Daegu 700-422, South Korea d Statistics Center, School of Medicine, Kyungpook National University, Dong In 2Ga 101, Daegu 700-422, South Korea e Department of Internal Medicine, Konyang University Hospital, Gasuwon-dong, Seo-gu, Daejeon 302-718, South Korea b c
a r t i c l e
i n f o
Article history: Accepted 7 October 2013 Available online 19 October 2013 Keywords: Lung cancer Risk factor miR-146a MicroRNA
a b s t r a c t Introduction: MicroRNAs (miRs) play important roles in the development and progression of human cancers. MiR-146a down-regulates epidermal growth factor receptor and the nuclear factor-κB regulatory kinase interleukin-1 receptor-associated kinase 1 genes that play important roles in lung carcinogenesis. This study was conducted to evaluate the association between rs2910164CNG, a functional polymorphism in the pre-miR146a, and lung cancer risk. Material and methods: The rs2910164CNG genotypes were determined in 1094 patients with lung cancer and 1100 healthy controls who were frequency matched for age and gender. Results: The rs2910164 CG or GG genotype was associated with a significantly decreased risk for lung cancer compared to that of the CC genotype (adjusted odds ratio = 0.80, 95% confidence interval = 0.66–0.96, P = 0.02). When subjects were stratified according to smoking exposure (never, light and heavy smokers), the effect of the rs2910164CNG genotype on lung cancer risk was significant only in never smokers (adjusted odds ratio = 0.66, 95% confidence interval = 0.45–0.96, P = 0.03, under a dominant model for the C allele) and decreased as smoking exposure level increased (Ptrend b 0.001). In line with this result, the level of miR-146a expression in the tumor tissues was significantly higher in the GG genotype than in the CC or CG genotype only in never-smokers (P = 0.02). Conclusions: These findings suggest that the rs2910164CNG in pre-miR-146a may contribute to genetic susceptibility to lung cancer, and that miR-146a might be involved in lung cancer development. © 2013 Elsevier B.V. All rights reserved.
1. Introduction MicroRNAs (miRs) are endogenous small (~22 nucleotides) noncoding RNAs that down-regulate gene expression by complimentary
Abbreviations: miRs, microRNAs; SNPs, single nucleotide polymorphisms; EGFR, epidermal growth factor receptor; BRCA1, breast cancer 1, early onset; NF-κB, nuclear factor-κB; IRAK1, interleukin-1 receptor-associated kinase 1; RT-PCR, reverse transcriptionpolymerase chain reaction; HWE, Hardy–Weinberg equilibrium; aORs, adjusted odds ratios; CIs, confidence intervals; SCCs, squamous cell carcinomas; ACs, adenocarcinomas; SCLC, small cell lung cancers; PH, P-values of the homogeneity test. ⁎ Corresponding author at: Lung Cancer Center, Kyungpook National University Medical Center, 807, Hoguk-ro, Buk-gu, Daegu 702-210, South Korea. Tel.: +82 53 200 2631; fax: +82 53 200 2027. E-mail addresses: [email protected] (H.-S. Jeon), [email protected] (Y.H. Lee), [email protected] (S.Y. Lee), [email protected] (J.-A. Jang), [email protected] (Y.-Y. Choi), [email protected] (S.S. Yoo), [email protected] (W.K. Lee), [email protected] (J.E. Choi), [email protected] (J.W. Son), [email protected] (Y.M. Kang), [email protected] (J.Y. Park). 1 These authors contributed equally to this paper. 0378-1119/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gene.2013.10.014
binding to the 3′-untranslated region of target messenger RNAs, thereby repressing translation or decreasing mRNA stability (Valencia-Sanchéz et al., 2006). More than 1000 miRs have been identified in the human genome, each of which can regulate multiple genes (Griffiths-Jones et al., 2006). Increasing evidence indicates that miRs play important roles in the development and progression of human cancers, largely by targeting genes that are key regulators of cell proliferation and survival, DNA repair, and the immune response (Esquela-Kerscher and Slack, 2006). MiRs represent two opposing roles, by behaving as tumor promoters or tumor suppressors depending on the tissue type and the presence of specific targets (Fabbri et al., 2007). In addition, miR profiling studies have demonstrated that an abnormal miR expression pattern is different based on tumors originating from various tissues (Wouters et al., 2011). Single nucleotide polymorphisms (SNPs) in miR sequences could affect miR processing and/or target selection, and thereby contribute to cancer development and progression (Hu et al., 2008; Wu et al., 2008). miR-146a down-regulates several cancer-related genes, including epidermal growth factor receptor (EGFR), breast cancer 1, early onset
H.-S. Jeon et al. / Gene 534 (2014) 66–71
(BRCA1), and the nuclear factor (NF)-κB regulatory kinase interleukin-1 receptor-associated kinase 1 (IRAK1) (Garcia et al., 2011; Y. Li et al., 2010). In addition, the rs2910164CNG polymorphism in pre-miR-146a, which results in a change from a C:U pair to a mismatched G:U pair in its stem region, affects the expression of mature miR-146a and miR146a target genes (Jazdzewski et al., 2008, 2009; Shen et al., 2008). Several studies have reported that this SNP affects susceptibility to various human cancers (Guo et al., 2010; Jazdzewski et al., 2008; Permuth-Wey et al., 2011; Shen et al., 2008; Xu et al., 2008, 2010). However, the results of these previous studies are heterogeneous across different cell types of cancer: the 2910164C allele has been associated with early-onset familial breast and ovarian cancer (Shen et al., 2008) and an increased risk of adult glioma (Permuth-Wey et al., 2011), whereas the 2910164C allele has been associated with a decreased risk of esophageal squamous cell carcinoma, hepatocellular carcinoma and prostate cancer (Guo et al., 2010; Xu et al., 2008, 2010). Because the over-expression of EGFR and activation of NF-κB are common in lung cancer, we carried out a case–control study to evaluate the effect of the pre-miR-146a rs2910164GNC polymorphism on lung cancer risk.
67
2.3. Quantitative reverse transcription-polymerase chain reaction (RT-PCR) Quantitative RT-PCR was performed to determine the expression of miR-146a according to the miR-146a rs2910164CNG genotypes. RNAs from NSCLC tissues and paired non-malignant lung tissues (n = 69; genotype distribution: 14 CC, 43 CG, and 12 GG in all cases; 12 CC, 28 CG, and 7 GG in smokers; 2 CC, 15 CG, and 5 GG in never-smokers) were isolated using Trizol (Invitrogen, Carlsbad, USA) according to the manufacturer's instructions. Expression levels of miR-146a were determined using TaqMan MicroRNA Assays (Applied Biosystems, CA, USA). The 2−ΔΔCT (where CT is threshold cycle) method (Livak and Schmittgen, 2001) was used to calculate relative expression level of miR-146a, using small nuclear RNA U6 as an endogenous control to normalize the expression of mature miRs. All real-time PCR was performed in triplicate using a LightCycler 480 (Roche Applied Science, Mannheim, Germany) according to the manufacturer's protocol.
2.4. Statistical analysis 2. Materials and methods 2.1. Study population This case–control study consisted of 1094 patients with lung cancer and 1100 healthy controls. The subjects were recruited from an ongoing lung cancer molecular epidemiological study, as described previously (Bae et al., 2012; Lee et al., 2010). In brief, eligible cases included all patients newly diagnosed with primary lung cancer between January 2000 and December 2003 at Kyungpook National University Hospital, Daegu, Korea. There were no age, gender, histological, or stage restrictions, but patients with a prior history of cancer were excluded from this study. Control subjects were randomly selected from a pool of healthy volunteers who visited the general health check-up center at Kyungpook National University Hospital during the same period. The control subjects were frequency matched (1:1) to the cases based on gender and age (±5 years). All case and control subjects were ethnic Koreans residing in Daegu or the surrounding regions. This study was approved by the Institutional Review Board of Kyungpook National University Hospital, Daegu, Korea, and written informed consent was obtained from all participants. Genomic DNA samples of the cases and healthy controls were provided by the National Biobank of Korea, which is supported by the Ministry of Health, Welfare, and Family Affairs. Never smokers were defined as subjects who had smoked less than 100 cigarettes during their lifetime. A former smoker was defined as one who had stopped smoking at least 1year before either a diagnosis of lung cancer (cases) or the date the informed consent form was signed (controls). The cumulative cigarette dose (pack-years) was calculated using the formula: pack-years = packs per day × years smoked.
2.2. Genotyping Genomic DNA was extracted from peripheral blood lymphocytes by proteinase K digestion and phenol/chloroform extraction. The genotypes were determined by a polymerase chain reaction–restriction fragment length polymorphism assay. All genotyping analyses were blinded with respect to the case/control status to ensure quality control. Two researchers independently examined the gel images and performed a repeat assay if they did not reach a consensus on the tested genotype. Approximately 10% of the samples were randomly selected to be genotyped again by a different investigator, and the results were 100% concordant. Selected polymerase chain reaction-amplified DNA samples (n = 10 for each genotype) were examined by DNA sequencing, and the results were also 100% concordant.
Differences in demographic characteristics, selected variables, and the frequencies of the genotypes between the cases and controls were compared using Student's t-test for continuous variables or the χ2 test for categorical variables. Deviations of the genotype frequencies in the controls from those expected under the Hardy–Weinberg equilibrium (HWE) were assessed by a goodness-of-fit χ2 test, as implemented through SAS Genetics. Unconditional logistic regression analysis was used to calculate adjusted odds ratios (aORs) and 95% confidence intervals (CIs) with adjustment for possible confounders (age and pack-years of smoking as continuous variables; and gender as a nominal variable). In addition to the overall association analysis, we performed a stratified analysis by age (median age), gender, smoking status, and tumor histology to further explore the association between the genotypes and lung cancer risk in each stratum. A homogeneity test was performed to compare the difference between genotype-related ORs of the different groups. All statistical analyses were performed using Statistical Analysis System software version 9.1.3 (SAS Institute, Cary, NC, USA).
Table 1 Characteristics of the study population. Variables Age (years) Mean ± SD Sex Male Female Smoking status Current Former Never Pack-yearsc Mean ± SD Stage I II III IV Histological types Squamous cell ca. Adenoca. Large cell ca. Small cell ca. a b c
Cases (n = 1094)
Controls (n = 1100)
60.7 ± 9.3
60.6 ± 9.3
0.70a
837 (76.5)b 257 (23.5)
840 (76.4) 260 (23.6)
0.94b
618 (56.5) 247 (22.6) 229 (20.9)
394 (35.8) 336 (30.6) 370 (33.6)
b0.0001b
39.6 ± 20.1
30.4 ± 16.6
b0.0001a
239 (21.8) 63 (5.8) 372 (34.0) 420 (38.4) 461 (42.1) 466 (42.6) 28 (2.6) 139 (12.7)
t-test. χ2 test. In current and former smokers.
P
68
H.-S. Jeon et al. / Gene 534 (2014) 66–71
3. Results The demographics of the cases and controls are shown in Table 1. No significant differences were observed between the cases and controls for mean age or gender distribution, suggesting adequate matching based on these two variables. The case group had a higher prevalence of current smokers than that in the control group (P b 0.0001), and the number of pack-years in smokers was significantly higher in cases compared to that in controls (39.6 ± 20.1 vs. 30.4 ± 16.6; P b 0.0001). These differences were controlled in the multivariate analyses. Of the 1094 lung cancer cases, 461 (42.1%) were squamous cell carcinomas (SCCs), 466 (42.63%) were adenocarcinomas (ACs), 28 (2.6%) were large cell carcinomas, and 139 (12.7%) were small cell lung cancers (SCLC). Distribution of the rs2910164CNG genotypes among controls was in HWE. Distribution of the rs2910164CNG genotypes was significantly different between the cases and controls (CC, CG, and GG genotypes; 33.7%, 45.8% and 20.5%, respectively vs. 28.5%, 49.3% and 22.2%, respectively; P = 0.03) and the frequency of the polymorphic G allele was significantly lower in cases than that in controls (0.43 vs. 0.47, P = 0.02; Table 2). Individuals with the CG or GG genotype were at a significantly decreased risk of lung cancer when compared to those with the CC genotype (aOR = 0.80, 95% CI = 0.66–0.96, P = 0.02). The association between the rs2910164CNG genotypes and lung cancer risk was further examined after stratification according to age (≤median age vs. N median age), gender, smoking exposure level, and histological type of lung cancer (Table 3). The effect of the CG or GG genotype on lung cancer risk was similar in younger and older individuals, as well as in males and females [P-values of the homogeneity test (PH) = 0.62 and 0.48, respectively). However, when stratified according to smoking exposure level (never, light [≤median pack-years] and heavy [N median pack-years] smokers), the protective effect of the CG or GG genotype on lung cancer risk was significant only in never smokers and decreased as smoking exposure level increased (aOR for never smokers = 0.66, 95% CI = 0.45–0.96, P = 0.03: aOR for light smokers = 0.79, 95% CI = 0.59–1.04, P = 0.10; and aOR for heavy smokers = 0.91, 95% CI = 0.64–1.30, P = 0.62; Ptrend b 0.001). When the lung cancer cases were categorized by tumor histology, the CG or GG genotypes were associated with a significantly decreased risk of AC and SCLC compared to that of the CC genotype (aOR for AC = 0.74, 95% CI = 0.58–0.93, P = 0.01; and aOR for SCLC = 0.68, 95% CI = 0.460.99, P = 0.04, respectively), whereas no significant association was observed between the genotypes and SCC risk. To identify the functional effect of miR-146a rs2910164CNG, we evaluated the relationship between the rs2910164CNG genotypes and expression level of mature miR-146a in tumor tissues and paired nonmalignant lung tissues (Fig. 1). The mature miR-146a expression level Table 2 Pre-miR-146a rs2910164CNG genotypes of cases and controls, and their association with lung cancer risk. Genotype Cases, n (%)
Controls n (%)
Pa
Pb Adjusted ORb (95% CI)
Minor allele frequency Case Control Pa
CC CG
368 (33.7) 312 (28.5) 0.03 0.43 500 (45.8) 540 (49.3)
GG
223 (20.5) 244 (22.2)
Ptrend CG + GG vs. CC
0.47
0.02 1.00 0.80 (0.66– 0.98) 0.79 (0.62– 1.00) 0.80 (0.66– 0.96)
0.03 0.05 0.03 0.02
a Two-sided χ2 test for either genotype distributions or allele frequencies between the cases and controls. b Odds ratios (95% confidence intervals) and their corresponding P-values were calculated by unconditional logistic analysis, adjusted for age, gender and pack-years of smoking.
was significantly lower in tumor tissues than in non-malignant lung tissues (P = 0.001). There was no significant difference in relative expression levels of miR-146a by the rs2910164CNG genotypes in either tumor tissues or non-malignant lung tissues (P = 0.34 and P = 0.79, respectively). When the tumor tissues and the paired non-malignant tissues were separated by smoking history, there was no significant association of rs2910164 genotypes with the relative expression level of miR-146a in smokers. However, the relative expression level of miR146a in the tumor tissues of never-smokers was significantly higher in the GG genotype compared with CC or CG genotypes (P = 0.02). 4. Discussion We investigated the influence of pre-miR-146a rs2910164CNG on lung cancer risk by conducting a hospital-based case–control study. This SNP was associated with a significantly decreased risk of lung cancer under a dominant model for the minor G allele. The protective effects were significantly related to smoking exposure level. In addition, the level of miR-146a expression by genotypes in the tumor tissues was modified by smoking exposure. These findings suggest that miR-146a may be involved in lung carcinogenesis and that the rs2910164CNG could be used as a marker for genetic susceptibility to lung cancer. Several studies have investigated the functional effect of rs2910164CNG on the expression of mature miR-146a and target genes and its association with cancer risk; however, results have been inconsistent. Shen et al. (2008) reported that the rs2910164C allele decreases BRCA1 expression by causing an increase in mature miR146a expression and increases the risk of early onset familial breast and ovarian cancers. Hung et al. (2012) also reported that the rs2910164 CC genotype increases miR-146a expression in oral SCCs and is associated with more advanced disease. These studies (Hung et al., 2012; Shen et al., 2008) suggest that miR-146a plays an oncogenic role in tumorigenesis and that the rs2910164C allele confers a higher expression level of mature miR-146a and stronger oncogenic capacity than the rs2910164G allele. However, in contrast to these studies (Hung et al., 2012; Shen et al., 2008), Xu et al. (2010) found that the rs2910164 CC genotype decreases miR-146a expression and is associated with reduced prostate cancer risk, suggesting that miR146a acts as an oncogene in the development of prostate cancer, but that the rs2910164 CC genotype confers a protective effect against development of the disease. In contrast, several studies have shown that miR-146a functions as a tumor suppressor (Bhaumik et al., 2008; Y. Li et al., 2010; Lin et al., 2008); re-expression of miR-146a reduces invasion and metastasis of pancreatic, breast, and prostate cancers by down regulating EGFR and two key adaptor proteins, IRAK-1 and TNF receptor-associated factor 6 in the interleukin-1 and Toll-like receptor signaling pathway, which positively regulate NF-κB activity (Bhaumik et al., 2008; Y. Li et al., 2010; Lin et al., 2008; Taganov et al., 2006). Comparable with these studies, the rs2910164C allele reduces expression of mature miR-146a, resulting in an increased risk for papillary thyroid carcinoma (Jazdzewski et al., 2008). However, Yue et al. (2011) reported that the rs2910164 GG genotype reduces miR146a expression and increases cervical cancer risk, suggesting that although miR-146a acts as a tumor suppressor, the rs2910164C is protective against the development of cervical cancer. In addition, in a recent meta-analysis using 18 studies, Wang et al. reported that there was no significant association of rs2910164 with the overall cancer risk (Wang et al., 2012b). Although it is hard to decipher the different results across studies, several genetic and environmental factors relevant to this SNP, such as different molecular pathogenesis in different cancers and different genetic backgrounds, might be significant factors in the discrepancies. In human cancers, miR levels are modulated by genetic (deletion, amplification, or translocation) and epigenetic alterations (Calin et al., 2004; Davalos and Esteller, 2010); thus miR expression profiling is extensively different according to where the tumors originated. In addition, one miR can regulate
H.-S. Jeon et al. / Gene 534 (2014) 66–71
69
Table 3 Association between pre-miR-146a rs2910164CNG genotypes and lung cancer risk according to age, gender, smoking status, and tumor histology. Variables
Age (years) ≤62 N62 Gender Male Female Smoking status Never ≤39 pys N39 pys Ptrend Histology SCC AC LCC SCLC
Cases
Pa
Controls
CC
CG + GG
CC
CG + GG
202 (33.2) 166 (34.4)
407 (66.8) 316 (65.6)
185 (28.3) 127 (28.7)
468 (71.7) 316 (71.3)
0.06 0.06
Adjusted OR (95% CI)c
PHb
CC
CG + GG
P
1.00 1.00
0.83 (0.65–1.10)d 0.75 (0.56–1.00)d
0.13 0.05
0.62
e
0.48
278 (33.3) 90 (35.0)
556 (66.7) 167 (65.0)
239 (28.5) 73 (28.3)
599 (71.5) 185 (71.7)
0.03 0.10
1.00 1.00
0.84 (0.68–1.03) 0.72 (0.49–1.05)e
0.10 0.09
83 (36.3) 137 (34.9) 148 (31.5)
146 (63.8) 255 (65.1) 322 (68.5)
95 (25.8) 153 (29.8) 64 (29.8)
273 (74.2) 360 (70.2) 151 (70.2)
0.01 0.10 0.65
1.00 1.00 1.00
0.66 (0.45–0.96)f 0.79 (0.59–1.04)f 0.91 (0.64–1.30)f
0.03 0.10 0.62 b0.001
141 (30.7) 165 (35.5) 10 (35.7) 52 (37.4)
318 (69.3) 300 (64.5) 18 (64.3) 87 (62.6)
312 (28.5) 312 (28.5) 312 (28.5) 312 (28.5)
784 (71.5) 784 (71.5) 784 (71.5) 784 (71.5)
0.37 0.01 0.40 0.03
1.00 1.00 1.00 1.00
0.94 (0.73–1.21)g 0.74 (0.58–0.93)g 0.74 (0.34–1.64)g 0.68 (0.46–0.99)g
0.94 0.01 0.46 0.04
0.44 0.22
0.17 0.57 0.16
Abbreviations: pys, pack-years of smoking; AC, adenocarcinoma; LCC, large cell carcinoma; SCC, squamous cell carcinoma; SCLC, small cell lung carcinoma. a Two-sided chi-square test for genotype distribution between the cases and controls. b P-values of test for homogeneity. c Odds ratios (ORs), 95% confidence intervals (CIs) and their corresponding P-values were calculated by unconditional logistic regression analysis. d Adjusted for gender and smoking status. e Adjusted for age and smoking status. f Adjusted for age and gender. g Adjusted for age, gender and smoking status.
multiple genes and one gene can be regulated by multiple miRs (Berezikov et al., 2005; Friedman et al., 2009). Therefore, miR-146a might play different roles in different cancer cell and tissue types, and the effect of rs2910164CNG on tumorigenesis might be different in various cancers depending on miR-146a-regulating genes and its target genes in each cell and tissue type of cancer. Accordingly, meta-analyses investigating the association of polymorphisms in miR-146a and cancer risk, most of which included various types of cancers, are not likely to show a significant association between the SNP and all types of cancers as a whole. In addition, most of the recent meta-analyses included a single association study on lung cancer risk because it was the only one published on the association of rs2910164 with lung cancer risk, where Tian et al. reported the same direction of association as our result, although not significant (Tian et al., 2009). Therefore, additional studies investigating the association between rs2910164 and lung cancer risk and meta-analyses of those studies are needed to confirm the association of miR-146a with lung cancer risk. To clarify the effect of pre-miR-146a rs2910164CNG on cancer risk, further well-designed studies with a large sample size are necessary. In addition, because the frequency of the rs2910164G allele among the healthy controls is significantly different across different ethnicities, ranging from 0.26 to 0.80 (Wang et al., 2012a), the association of the SNP with cancer risk should be evaluated in diverse ethnic populations. In the present study, we evaluated the effect of this SNP on lung cancer risk in a large population. Moreover, the Korean population is genetically homogenous, reducing the risk of confounding due to population stratification. Additionally, another major strength of the present study was the inclusion criteria for the controls. Emerging evidence indicates that miR-146a regulates innate and adaptive immune and inflammatory responses and is deregulated in several diseases, including autoimmune disorders such as rheumatoid arthritis and osteoarthritis (L. Li et al., 2010, 2012; Perry et al., 2008; Rusca and Monticelli, 2011). We included only healthy subjects in the control group, who were free of disease on a health check-up. Therefore, the effect of nonmalignant disease on the genotype distribution in our control group can be excluded. We also controlled for selection bias related to hospital-based case–control studies. Given that most patients with lung cancer are treated at the University Hospital in Korea, the demographics and clinical characteristics of the patients with lung cancer in the current study were compatible with those of a nationwide
lung cancer survey (In et al., 2009). Furthermore, as all patients who were diagnosed with lung cancer at the University Hospital were included in this study, it is reasonable to assume that the case group represented the lung cancer cases in our community. In addition, because the age and gender distribution of non-participating controls were similar to those of the participating controls, self-selection bias was unlikely. Moreover, the controls were extensively evaluated for other SNPs (Bae et al., 2012; Lee et al., 2010), and the genotype distribution for each SNP met the conditions for HWE, which further supports the randomness of our control subjects. Taken together, these strengths increased the reliability of our finding of an association between the pre-miR-146a rs2910164CNG and lung cancer risk. An interesting finding of our study was that the rs2910164CNG genotypes had a more pronounced association with lung cancer risk in never smokers. This difference may be attributable to differences in the lung carcinogenesis pathways between ever and never smokers. Although tobacco smoking is a major cause of lung cancer, it has been estimated that 15% of lung cancers in men and 53% in women are never smokers, accounting for 25% of all lung cancers worldwide. In addition, lung cancer in never and ever smokers is a distinct disease that has considerable differences in etiology, genetic/molecular changes, and clinico-pathological features (Sun et al., 2007). Therefore, rs2910164CNG could confer a different impact on lung cancer risk in never and ever smokers. Because this study was designed to evaluate the effect of the SNP on overall lung cancer risk, the stratification analysis according to smoking exposure level may have had a type I error (due to multiple comparisons) and/or a type II error (due to the small number of subjects in the subgroups). Therefore, additional studies with larger sample sizes are required to confirm our findings. In the present study, we investigated the effect of the rs2910164CNG on miR-146a expression. The mature miR-146a expression level was significantly lower in tumor tissues than in non-malignant lung tissues, suggesting that miR-146a participates in lung carcinogenesis as a tumor suppressor. It should be noted that comparing the levels of a microRNA from tumors and normal tissues without microdissection could have some drawbacks due to possible differences in the cell composition of samples: normal tissues may contain a relatively larger proportion of stroma that hematopoietic cells abundant in the miR-146a may populate. There was no significant difference of mature miR-146a expression levels by the rs2910164CNG genotypes overall. However,
70
H.-S. Jeon et al. / Gene 534 (2014) 66–71
P = 0.79
P = 0.34
400
Total 300
Tumor
200 100 0
B
m 146 CG
GG
CC
P = 0.39
600
m 146 CG
GG
would affect EGFR pathway and implicate in the pathogenesis of NSCLC in never-smokers. In conclusion, we found a significant association between the premiR-146a rs2910164CNG SNP and lung cancer risk. The effect of this SNP on lung cancer risk was more pronounced in never smokers. In addition, this association was in line with the level of miR-146a expression by genotypes in the tumor tissues of never-smokers. These results suggest that this SNP contributes to genetic susceptibility to lung cancer, particularly in never smokers. Further studies on the biological function of this SNP are needed to understand the role of miR-146a in determining lung cancer risk. Conflict of interest statement
P = 0.93
All authors declare no conflict of interest.
P = 0.17
500
Acknowledgment 400
Smoker
m_Tumor
m_Normal
Mature miR-146a expression
Tumor
500
CC
300 200 100 0
This work was supported in part by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology (NRF-2010-0004700), and in part by grant no. NRF-2011-0013670. References
CC
C
CG
m ir146a
GG
CC
P = 0.48
600 500
CG
m ir146a
GG
P = 0.05
P = 0.33
P = 0.02
400
Neversmoker
n_m_Tumor
n_m_Normal
Mature miR-146a expression
Normal 600
Normal
Mature miR-146a expression
A
300 200 100 0 CC
CG
n_m ir146a
GG
CC
CG
n_m ir146a
GG
Fig. 1. Effect of the rs2910164CNG genotypes on miR-146a expression. Relative expression levels of mature miR-146a and the association with the rs2910164CNG genotypes (14 CC, 43 CG, and 12 GG in all cases; 12 CC, 28 CG, and 7 GG in smokers; 2 CC, 15 CG, and 5 GG in never-smokers) were examined in 69 tumors and paired non-malignant lung tissues. Relative expression levels of mature miR-146a by the rs2910164CNG genotypes in tumors and normal lung tissues among all cases (A), smokers (B), and never-smokers (C). The horizontal lines within the boxes represent the median values; the upper and lower boundaries of the boxes represent 75th and 25th percentiles, respectively; whiskers extend from the boxes to 1.5 times the interquartile range. P values, Student's t-test.
in line with the association of rs2910164 with lung cancer risk, the level of miR-146a expression in tumor tissues was significantly higher in GG genotype compared with CC or CG genotypes only in never-smokers after stratifying by smoking exposure. EGFR pathway plays a crucial role in many carcinogenic processes such as proliferation, angiogenesis, invasion, and metastasis, and resistance to apoptosis. EGFR is frequently overexpressed in NSCLC, and its mutation has been known as a frequent oncogenic driver mutation in NSCLC of never-smokers (Li et al., 2011). EGFR has been reported as the target of miR-146a: miR-146a down-regulated the expression of EGFR and downstream signaling, inhibited cell growth, and induced apoptosis in NSCLC cells (Chen et al., 2013). Taken together, it can be assumed that the miR-146a rs2910164C to G change may increase miR-146a expression, thereby down-regulate EGFR expression and downstream signaling, leading to cell growth inhibition and induction of apoptosis. Further studies are needed to investigate whether miR-146a rs2910164CNG
Bae, E.Y., et al., 2012. Replication of results of genome-wide association studies on lung cancer susceptibility loci in a Korean population. Respirology 17, 699–706. Berezikov, E., Guryev, V., van de Belt, J., Wienholds, E., Plasterk, R.H., Cuppen, E., 2005. Phylogenetic shadowing and computational identification of human microRNA genes. Cell 120, 21–24. Bhaumik, D., Scott, G.K., Schokrpur, S., Patil, C.K., Campisi, J., Benz, C.C., 2008. Expression of microRNA-146 suppresses NF-kappaB activity with reduction of metastatic potential in breast cancer cells. Oncogene 27, 5643–5647. Calin, G.A., et al., 2004. Human microRNA genes are frequently located at fragile sites and genomic regions involved in cancers. Proc. Natl. Acad. Sci. U. S. A. 101, 2999–3004. Chen, G., et al., 2013. miR-146a inhibits cell growth, cell migration and induces apoptosis in non-small cell lung cancer cells. PLoS One 8, e60317. Davalos, V., Esteller, M., 2010. MicroRNAs and cancer epigenetics: a macrorevolution. Curr. Opin. Oncol. 22, 35–45. Esquela-Kerscher, A., Slack, F.J., 2006. Oncomirs — microRNAs with a role in cancer. Nat. Rev. Cancer 6, 259–269. Fabbri, M., Ivan, M., Cimmino, A., Negrini, M., Calin, G.A., 2007. Regulatory mechanisms of microRNAs involvement in cancer. Expert. Opin. Biol. Ther. 7, 1009–1019. Friedman, R.C., Farh, K.K., Burge, C.B., Bartel, D.P., 2009. Most mammalian mRNAs are conserved targets of microRNAs. Genome Res. 19, 92–105. Garcia, A.I., et al., 2011. Down-regulation of BRCA1 expression by miR-146a and miR146b-5p in triple negative sporadic breast cancers. EMBO Mol. Med. 3, 279–290. Griffiths-Jones, S., Grocock, R.J., van Dongen, S., Bateman, A., Enright, A.J., 2006. miRBase: microRNA sequences, targets and gene nomenclature. Nucleic Acids Res. 34, D140–D144. Guo, H., et al., 2010. A functional variant in microRNA-146a is associated with risk of esophageal squamous cell carcinoma in Chinese Han. Fam. Cancer 9, 599–603. Hu, Z., et al., 2008. Genetic variants of miRNA sequences and non-small cell lung cancer survival. J. Clin. Invest. 118, 2600–2608. Hung, P.S., Chang, K.W., Kao, S.Y., Chu, T.H., Liu, C.J., Lin, S.C., 2012. Association between the rs2910164 polymorphism in pre-mir-146a and oral carcinoma progression. Oral Oncol. 48, 404–408. In, K.H., et al., 2009. Lung cancer patients who are asymptomatic at diagnosis show favorable prognosis: a Korean lung cancer registry study. Lung Cancer 64, 232–237. Jazdzewski, K., Murray, E.L., Franssila, K., Jarzab, B., Schoenberg, D.R., de la Chapelle, A., 2008. Common SNP in pre-miR-146a decreases mature miR expression and predisposes to papillary thyroid carcinoma. Proc. Natl. Acad. Sci. U. S. A. 105, 7269–7274. Jazdzewski, K., et al., 2009. Polymorphic mature microRNAs from passenger strand of premiR-146a contribute to thyroid cancer. Proc. Natl. Acad. Sci. U. S. A. 106, 1502–1505. Lee, S.Y., et al., 2010. Polymorphisms in the caspase genes and the risk of lung cancer. J. Thorac. Oncol. 5, 1152–1158. Li, L., Chen, X.P., Li, Y.J., 2010a. MicroRNA-146a and human disease. Scand. J. Immunol. 71, 227–231. Li, Y., et al., 2010b. miR-146a suppresses invasion of pancreatic cancer cells. Cancer Res. 70, 1486–1495. Li, C., et al., 2011. Spectrum of oncogenic driver mutations in lung adenocarcinomas from East Asian never smokers. PLoS One 6, e28204. Li, J., et al., 2012. miR-146a, an IL-1beta responsive miRNA, induces vascular endothelial growth factor and chondrocyte apoptosis by targeting Smad4. Arthritis Res. Ther. 14, R75. Lin, S.L., Chiang, A., Chang, D., Ying, S.Y., 2008. Loss of mir-146a function in hormonerefractory prostate cancer. RNA 14, 417–424.
H.-S. Jeon et al. / Gene 534 (2014) 66–71 Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods 25, 402–408. Permuth-Wey, J., et al., 2011. A functional polymorphism in the pre-miR-146a gene is associated with risk and prognosis in adult glioma. J. Neurooncol. 105, 639–646. Perry, M.M., Moschos, S.A., Williams, A.E., Shepherd, N.J., Larner-Svensson, H.M., Lindsay, M.A., 2008. Rapid changes in microRNA-146a expression negatively regulate the IL1beta-induced inflammatory response in human lung alveolar epithelial cells. J. Immunol. 180, 5689–5698. Rusca, N., Monticelli, S., 2011. MiR-146a in immunity and disease. Mol. Biol. Int. 2011, 437301. Shen, J., Ambrosone, C.B., DiCioccio, R.A., Odunsi, K., Lele, S.B., Zhao, H., 2008. A functional polymorphism in the miR-146a gene and age of familial breast/ovarian cancer diagnosis. Carcinogenesis 29, 1963–1966. Sun, S., Schiller, J.H., Gazdar, A.F., 2007. Lung cancer in never smokers-a different disease. Nat. Rev. Cancer 7, 778–790. Taganov, K.D., Boldin, M.P., Chang, K.J., Baltimore, D., 2006. NF-kappaBdependent induction of microRNA miR-146, an inhibitor targeted to signaling proteins of innate immune responses. Proc. Natl. Acad. Sci. U. S. A. 103, 12481–12486.
71
Tian, T., et al., 2009. A functional genetic variant in microRNA-196a2 is associated with increased susceptibility of lung cancer in Chinese. Cancer Epidemiol. Biomarkers Prev. 18, 1183–1187. Valencia-Sanchez, M.A., Liu, J., Hannon, G.J., Parker, R., 2006. Control of translation and mRNA degradation by miRNAs and siRNAs. Genes Dev. 20, 515–524. Wang, J., Bi, J., Liu, X., Li, K., Di, J., Wang, B., 2012a. Has-miR-146a polymorphism (rs2910164) and cancer risk: a meta-analysis of 19 case–control studies. Mol. Biol. Rep. 39, 4571–4579. Wang, J., et al., 2012b. The association of miR-146a rs2910164 and miR-196a2 rs11614913 polymorphisms with cancer risk: a meta-analysis of 32 studies. Mutagenesis 27, 779–788. Wouters, M.D., van Gent, D.C., Hoeijmakers, J.H., Pothof, J., 2011. MicroRNAs, the DNA damage response and cancer. Mutat. Res. 717, 54–66. Wu, M., et al., 2008. Genetic variations of microRNAs in human cancer and their effects on the expression of miRNAs. Carcinogenesis 29, 1710–1716. Xu, T., et al., 2008. A functional polymorphism in the miR-146a gene is associated with the risk for hepatocellular carcinoma. Carcinogenesis 29, 2126–2131. Xu, B., et al., 2010. A functional polymorphism in Pre-miR-146a gene is associated with prostate cancer risk and mature miR-146a expression in vivo. Prostate 70, 467–472. Yue, C., et al., 2011. Polymorphism of the pre-miR-146a is associated with risk of cervical cancer in a Chinese population. Gynecol. Oncol. 122, 33–37.
Contents lists available at ScienceDirect
Gene journal homepage: www.elsevier.com/locate/gene
A common polymorphism in pre-microRNA-146a is associated with lung cancer risk in a Korean population Hyo-Sung Jeon a,b,1, Yong Hoon Lee c,1, Shin Yup Lee b,c,1, Ji-Ae Jang a, Yi-Young Choi a, Seung Soo Yoo b,c, Won Kee Lee d, Jin Eun Choi a,b, Ji Woong Son e, Young Mo Kang c, Jae Yong Park a,b,c,⁎ a
Department of Biochemistry and Cell Biology, School of Medicine, Kyungpook National University, Dong In 2Ga 101, Daegu 700-422, South Korea Lung Cancer Center, Kyungpook National University Medical Center, 807, Hoguk-ro, Buk-gu, Daegu 702-210, South Korea Department of Internal Medicine, School of Medicine, Kyungpook National University, Dong In 2Ga 101, Daegu 700-422, South Korea d Statistics Center, School of Medicine, Kyungpook National University, Dong In 2Ga 101, Daegu 700-422, South Korea e Department of Internal Medicine, Konyang University Hospital, Gasuwon-dong, Seo-gu, Daejeon 302-718, South Korea b c
a r t i c l e
i n f o
Article history: Accepted 7 October 2013 Available online 19 October 2013 Keywords: Lung cancer Risk factor miR-146a MicroRNA
a b s t r a c t Introduction: MicroRNAs (miRs) play important roles in the development and progression of human cancers. MiR-146a down-regulates epidermal growth factor receptor and the nuclear factor-κB regulatory kinase interleukin-1 receptor-associated kinase 1 genes that play important roles in lung carcinogenesis. This study was conducted to evaluate the association between rs2910164CNG, a functional polymorphism in the pre-miR146a, and lung cancer risk. Material and methods: The rs2910164CNG genotypes were determined in 1094 patients with lung cancer and 1100 healthy controls who were frequency matched for age and gender. Results: The rs2910164 CG or GG genotype was associated with a significantly decreased risk for lung cancer compared to that of the CC genotype (adjusted odds ratio = 0.80, 95% confidence interval = 0.66–0.96, P = 0.02). When subjects were stratified according to smoking exposure (never, light and heavy smokers), the effect of the rs2910164CNG genotype on lung cancer risk was significant only in never smokers (adjusted odds ratio = 0.66, 95% confidence interval = 0.45–0.96, P = 0.03, under a dominant model for the C allele) and decreased as smoking exposure level increased (Ptrend b 0.001). In line with this result, the level of miR-146a expression in the tumor tissues was significantly higher in the GG genotype than in the CC or CG genotype only in never-smokers (P = 0.02). Conclusions: These findings suggest that the rs2910164CNG in pre-miR-146a may contribute to genetic susceptibility to lung cancer, and that miR-146a might be involved in lung cancer development. © 2013 Elsevier B.V. All rights reserved.
1. Introduction MicroRNAs (miRs) are endogenous small (~22 nucleotides) noncoding RNAs that down-regulate gene expression by complimentary
Abbreviations: miRs, microRNAs; SNPs, single nucleotide polymorphisms; EGFR, epidermal growth factor receptor; BRCA1, breast cancer 1, early onset; NF-κB, nuclear factor-κB; IRAK1, interleukin-1 receptor-associated kinase 1; RT-PCR, reverse transcriptionpolymerase chain reaction; HWE, Hardy–Weinberg equilibrium; aORs, adjusted odds ratios; CIs, confidence intervals; SCCs, squamous cell carcinomas; ACs, adenocarcinomas; SCLC, small cell lung cancers; PH, P-values of the homogeneity test. ⁎ Corresponding author at: Lung Cancer Center, Kyungpook National University Medical Center, 807, Hoguk-ro, Buk-gu, Daegu 702-210, South Korea. Tel.: +82 53 200 2631; fax: +82 53 200 2027. E-mail addresses: [email protected] (H.-S. Jeon), [email protected] (Y.H. Lee), [email protected] (S.Y. Lee), [email protected] (J.-A. Jang), [email protected] (Y.-Y. Choi), [email protected] (S.S. Yoo), [email protected] (W.K. Lee), [email protected] (J.E. Choi), [email protected] (J.W. Son), [email protected] (Y.M. Kang), [email protected] (J.Y. Park). 1 These authors contributed equally to this paper. 0378-1119/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gene.2013.10.014
binding to the 3′-untranslated region of target messenger RNAs, thereby repressing translation or decreasing mRNA stability (Valencia-Sanchéz et al., 2006). More than 1000 miRs have been identified in the human genome, each of which can regulate multiple genes (Griffiths-Jones et al., 2006). Increasing evidence indicates that miRs play important roles in the development and progression of human cancers, largely by targeting genes that are key regulators of cell proliferation and survival, DNA repair, and the immune response (Esquela-Kerscher and Slack, 2006). MiRs represent two opposing roles, by behaving as tumor promoters or tumor suppressors depending on the tissue type and the presence of specific targets (Fabbri et al., 2007). In addition, miR profiling studies have demonstrated that an abnormal miR expression pattern is different based on tumors originating from various tissues (Wouters et al., 2011). Single nucleotide polymorphisms (SNPs) in miR sequences could affect miR processing and/or target selection, and thereby contribute to cancer development and progression (Hu et al., 2008; Wu et al., 2008). miR-146a down-regulates several cancer-related genes, including epidermal growth factor receptor (EGFR), breast cancer 1, early onset
H.-S. Jeon et al. / Gene 534 (2014) 66–71
(BRCA1), and the nuclear factor (NF)-κB regulatory kinase interleukin-1 receptor-associated kinase 1 (IRAK1) (Garcia et al., 2011; Y. Li et al., 2010). In addition, the rs2910164CNG polymorphism in pre-miR-146a, which results in a change from a C:U pair to a mismatched G:U pair in its stem region, affects the expression of mature miR-146a and miR146a target genes (Jazdzewski et al., 2008, 2009; Shen et al., 2008). Several studies have reported that this SNP affects susceptibility to various human cancers (Guo et al., 2010; Jazdzewski et al., 2008; Permuth-Wey et al., 2011; Shen et al., 2008; Xu et al., 2008, 2010). However, the results of these previous studies are heterogeneous across different cell types of cancer: the 2910164C allele has been associated with early-onset familial breast and ovarian cancer (Shen et al., 2008) and an increased risk of adult glioma (Permuth-Wey et al., 2011), whereas the 2910164C allele has been associated with a decreased risk of esophageal squamous cell carcinoma, hepatocellular carcinoma and prostate cancer (Guo et al., 2010; Xu et al., 2008, 2010). Because the over-expression of EGFR and activation of NF-κB are common in lung cancer, we carried out a case–control study to evaluate the effect of the pre-miR-146a rs2910164GNC polymorphism on lung cancer risk.
67
2.3. Quantitative reverse transcription-polymerase chain reaction (RT-PCR) Quantitative RT-PCR was performed to determine the expression of miR-146a according to the miR-146a rs2910164CNG genotypes. RNAs from NSCLC tissues and paired non-malignant lung tissues (n = 69; genotype distribution: 14 CC, 43 CG, and 12 GG in all cases; 12 CC, 28 CG, and 7 GG in smokers; 2 CC, 15 CG, and 5 GG in never-smokers) were isolated using Trizol (Invitrogen, Carlsbad, USA) according to the manufacturer's instructions. Expression levels of miR-146a were determined using TaqMan MicroRNA Assays (Applied Biosystems, CA, USA). The 2−ΔΔCT (where CT is threshold cycle) method (Livak and Schmittgen, 2001) was used to calculate relative expression level of miR-146a, using small nuclear RNA U6 as an endogenous control to normalize the expression of mature miRs. All real-time PCR was performed in triplicate using a LightCycler 480 (Roche Applied Science, Mannheim, Germany) according to the manufacturer's protocol.
2.4. Statistical analysis 2. Materials and methods 2.1. Study population This case–control study consisted of 1094 patients with lung cancer and 1100 healthy controls. The subjects were recruited from an ongoing lung cancer molecular epidemiological study, as described previously (Bae et al., 2012; Lee et al., 2010). In brief, eligible cases included all patients newly diagnosed with primary lung cancer between January 2000 and December 2003 at Kyungpook National University Hospital, Daegu, Korea. There were no age, gender, histological, or stage restrictions, but patients with a prior history of cancer were excluded from this study. Control subjects were randomly selected from a pool of healthy volunteers who visited the general health check-up center at Kyungpook National University Hospital during the same period. The control subjects were frequency matched (1:1) to the cases based on gender and age (±5 years). All case and control subjects were ethnic Koreans residing in Daegu or the surrounding regions. This study was approved by the Institutional Review Board of Kyungpook National University Hospital, Daegu, Korea, and written informed consent was obtained from all participants. Genomic DNA samples of the cases and healthy controls were provided by the National Biobank of Korea, which is supported by the Ministry of Health, Welfare, and Family Affairs. Never smokers were defined as subjects who had smoked less than 100 cigarettes during their lifetime. A former smoker was defined as one who had stopped smoking at least 1year before either a diagnosis of lung cancer (cases) or the date the informed consent form was signed (controls). The cumulative cigarette dose (pack-years) was calculated using the formula: pack-years = packs per day × years smoked.
2.2. Genotyping Genomic DNA was extracted from peripheral blood lymphocytes by proteinase K digestion and phenol/chloroform extraction. The genotypes were determined by a polymerase chain reaction–restriction fragment length polymorphism assay. All genotyping analyses were blinded with respect to the case/control status to ensure quality control. Two researchers independently examined the gel images and performed a repeat assay if they did not reach a consensus on the tested genotype. Approximately 10% of the samples were randomly selected to be genotyped again by a different investigator, and the results were 100% concordant. Selected polymerase chain reaction-amplified DNA samples (n = 10 for each genotype) were examined by DNA sequencing, and the results were also 100% concordant.
Differences in demographic characteristics, selected variables, and the frequencies of the genotypes between the cases and controls were compared using Student's t-test for continuous variables or the χ2 test for categorical variables. Deviations of the genotype frequencies in the controls from those expected under the Hardy–Weinberg equilibrium (HWE) were assessed by a goodness-of-fit χ2 test, as implemented through SAS Genetics. Unconditional logistic regression analysis was used to calculate adjusted odds ratios (aORs) and 95% confidence intervals (CIs) with adjustment for possible confounders (age and pack-years of smoking as continuous variables; and gender as a nominal variable). In addition to the overall association analysis, we performed a stratified analysis by age (median age), gender, smoking status, and tumor histology to further explore the association between the genotypes and lung cancer risk in each stratum. A homogeneity test was performed to compare the difference between genotype-related ORs of the different groups. All statistical analyses were performed using Statistical Analysis System software version 9.1.3 (SAS Institute, Cary, NC, USA).
Table 1 Characteristics of the study population. Variables Age (years) Mean ± SD Sex Male Female Smoking status Current Former Never Pack-yearsc Mean ± SD Stage I II III IV Histological types Squamous cell ca. Adenoca. Large cell ca. Small cell ca. a b c
Cases (n = 1094)
Controls (n = 1100)
60.7 ± 9.3
60.6 ± 9.3
0.70a
837 (76.5)b 257 (23.5)
840 (76.4) 260 (23.6)
0.94b
618 (56.5) 247 (22.6) 229 (20.9)
394 (35.8) 336 (30.6) 370 (33.6)
b0.0001b
39.6 ± 20.1
30.4 ± 16.6
b0.0001a
239 (21.8) 63 (5.8) 372 (34.0) 420 (38.4) 461 (42.1) 466 (42.6) 28 (2.6) 139 (12.7)
t-test. χ2 test. In current and former smokers.
P
68
H.-S. Jeon et al. / Gene 534 (2014) 66–71
3. Results The demographics of the cases and controls are shown in Table 1. No significant differences were observed between the cases and controls for mean age or gender distribution, suggesting adequate matching based on these two variables. The case group had a higher prevalence of current smokers than that in the control group (P b 0.0001), and the number of pack-years in smokers was significantly higher in cases compared to that in controls (39.6 ± 20.1 vs. 30.4 ± 16.6; P b 0.0001). These differences were controlled in the multivariate analyses. Of the 1094 lung cancer cases, 461 (42.1%) were squamous cell carcinomas (SCCs), 466 (42.63%) were adenocarcinomas (ACs), 28 (2.6%) were large cell carcinomas, and 139 (12.7%) were small cell lung cancers (SCLC). Distribution of the rs2910164CNG genotypes among controls was in HWE. Distribution of the rs2910164CNG genotypes was significantly different between the cases and controls (CC, CG, and GG genotypes; 33.7%, 45.8% and 20.5%, respectively vs. 28.5%, 49.3% and 22.2%, respectively; P = 0.03) and the frequency of the polymorphic G allele was significantly lower in cases than that in controls (0.43 vs. 0.47, P = 0.02; Table 2). Individuals with the CG or GG genotype were at a significantly decreased risk of lung cancer when compared to those with the CC genotype (aOR = 0.80, 95% CI = 0.66–0.96, P = 0.02). The association between the rs2910164CNG genotypes and lung cancer risk was further examined after stratification according to age (≤median age vs. N median age), gender, smoking exposure level, and histological type of lung cancer (Table 3). The effect of the CG or GG genotype on lung cancer risk was similar in younger and older individuals, as well as in males and females [P-values of the homogeneity test (PH) = 0.62 and 0.48, respectively). However, when stratified according to smoking exposure level (never, light [≤median pack-years] and heavy [N median pack-years] smokers), the protective effect of the CG or GG genotype on lung cancer risk was significant only in never smokers and decreased as smoking exposure level increased (aOR for never smokers = 0.66, 95% CI = 0.45–0.96, P = 0.03: aOR for light smokers = 0.79, 95% CI = 0.59–1.04, P = 0.10; and aOR for heavy smokers = 0.91, 95% CI = 0.64–1.30, P = 0.62; Ptrend b 0.001). When the lung cancer cases were categorized by tumor histology, the CG or GG genotypes were associated with a significantly decreased risk of AC and SCLC compared to that of the CC genotype (aOR for AC = 0.74, 95% CI = 0.58–0.93, P = 0.01; and aOR for SCLC = 0.68, 95% CI = 0.460.99, P = 0.04, respectively), whereas no significant association was observed between the genotypes and SCC risk. To identify the functional effect of miR-146a rs2910164CNG, we evaluated the relationship between the rs2910164CNG genotypes and expression level of mature miR-146a in tumor tissues and paired nonmalignant lung tissues (Fig. 1). The mature miR-146a expression level Table 2 Pre-miR-146a rs2910164CNG genotypes of cases and controls, and their association with lung cancer risk. Genotype Cases, n (%)
Controls n (%)
Pa
Pb Adjusted ORb (95% CI)
Minor allele frequency Case Control Pa
CC CG
368 (33.7) 312 (28.5) 0.03 0.43 500 (45.8) 540 (49.3)
GG
223 (20.5) 244 (22.2)
Ptrend CG + GG vs. CC
0.47
0.02 1.00 0.80 (0.66– 0.98) 0.79 (0.62– 1.00) 0.80 (0.66– 0.96)
0.03 0.05 0.03 0.02
a Two-sided χ2 test for either genotype distributions or allele frequencies between the cases and controls. b Odds ratios (95% confidence intervals) and their corresponding P-values were calculated by unconditional logistic analysis, adjusted for age, gender and pack-years of smoking.
was significantly lower in tumor tissues than in non-malignant lung tissues (P = 0.001). There was no significant difference in relative expression levels of miR-146a by the rs2910164CNG genotypes in either tumor tissues or non-malignant lung tissues (P = 0.34 and P = 0.79, respectively). When the tumor tissues and the paired non-malignant tissues were separated by smoking history, there was no significant association of rs2910164 genotypes with the relative expression level of miR-146a in smokers. However, the relative expression level of miR146a in the tumor tissues of never-smokers was significantly higher in the GG genotype compared with CC or CG genotypes (P = 0.02). 4. Discussion We investigated the influence of pre-miR-146a rs2910164CNG on lung cancer risk by conducting a hospital-based case–control study. This SNP was associated with a significantly decreased risk of lung cancer under a dominant model for the minor G allele. The protective effects were significantly related to smoking exposure level. In addition, the level of miR-146a expression by genotypes in the tumor tissues was modified by smoking exposure. These findings suggest that miR-146a may be involved in lung carcinogenesis and that the rs2910164CNG could be used as a marker for genetic susceptibility to lung cancer. Several studies have investigated the functional effect of rs2910164CNG on the expression of mature miR-146a and target genes and its association with cancer risk; however, results have been inconsistent. Shen et al. (2008) reported that the rs2910164C allele decreases BRCA1 expression by causing an increase in mature miR146a expression and increases the risk of early onset familial breast and ovarian cancers. Hung et al. (2012) also reported that the rs2910164 CC genotype increases miR-146a expression in oral SCCs and is associated with more advanced disease. These studies (Hung et al., 2012; Shen et al., 2008) suggest that miR-146a plays an oncogenic role in tumorigenesis and that the rs2910164C allele confers a higher expression level of mature miR-146a and stronger oncogenic capacity than the rs2910164G allele. However, in contrast to these studies (Hung et al., 2012; Shen et al., 2008), Xu et al. (2010) found that the rs2910164 CC genotype decreases miR-146a expression and is associated with reduced prostate cancer risk, suggesting that miR146a acts as an oncogene in the development of prostate cancer, but that the rs2910164 CC genotype confers a protective effect against development of the disease. In contrast, several studies have shown that miR-146a functions as a tumor suppressor (Bhaumik et al., 2008; Y. Li et al., 2010; Lin et al., 2008); re-expression of miR-146a reduces invasion and metastasis of pancreatic, breast, and prostate cancers by down regulating EGFR and two key adaptor proteins, IRAK-1 and TNF receptor-associated factor 6 in the interleukin-1 and Toll-like receptor signaling pathway, which positively regulate NF-κB activity (Bhaumik et al., 2008; Y. Li et al., 2010; Lin et al., 2008; Taganov et al., 2006). Comparable with these studies, the rs2910164C allele reduces expression of mature miR-146a, resulting in an increased risk for papillary thyroid carcinoma (Jazdzewski et al., 2008). However, Yue et al. (2011) reported that the rs2910164 GG genotype reduces miR146a expression and increases cervical cancer risk, suggesting that although miR-146a acts as a tumor suppressor, the rs2910164C is protective against the development of cervical cancer. In addition, in a recent meta-analysis using 18 studies, Wang et al. reported that there was no significant association of rs2910164 with the overall cancer risk (Wang et al., 2012b). Although it is hard to decipher the different results across studies, several genetic and environmental factors relevant to this SNP, such as different molecular pathogenesis in different cancers and different genetic backgrounds, might be significant factors in the discrepancies. In human cancers, miR levels are modulated by genetic (deletion, amplification, or translocation) and epigenetic alterations (Calin et al., 2004; Davalos and Esteller, 2010); thus miR expression profiling is extensively different according to where the tumors originated. In addition, one miR can regulate
H.-S. Jeon et al. / Gene 534 (2014) 66–71
69
Table 3 Association between pre-miR-146a rs2910164CNG genotypes and lung cancer risk according to age, gender, smoking status, and tumor histology. Variables
Age (years) ≤62 N62 Gender Male Female Smoking status Never ≤39 pys N39 pys Ptrend Histology SCC AC LCC SCLC
Cases
Pa
Controls
CC
CG + GG
CC
CG + GG
202 (33.2) 166 (34.4)
407 (66.8) 316 (65.6)
185 (28.3) 127 (28.7)
468 (71.7) 316 (71.3)
0.06 0.06
Adjusted OR (95% CI)c
PHb
CC
CG + GG
P
1.00 1.00
0.83 (0.65–1.10)d 0.75 (0.56–1.00)d
0.13 0.05
0.62
e
0.48
278 (33.3) 90 (35.0)
556 (66.7) 167 (65.0)
239 (28.5) 73 (28.3)
599 (71.5) 185 (71.7)
0.03 0.10
1.00 1.00
0.84 (0.68–1.03) 0.72 (0.49–1.05)e
0.10 0.09
83 (36.3) 137 (34.9) 148 (31.5)
146 (63.8) 255 (65.1) 322 (68.5)
95 (25.8) 153 (29.8) 64 (29.8)
273 (74.2) 360 (70.2) 151 (70.2)
0.01 0.10 0.65
1.00 1.00 1.00
0.66 (0.45–0.96)f 0.79 (0.59–1.04)f 0.91 (0.64–1.30)f
0.03 0.10 0.62 b0.001
141 (30.7) 165 (35.5) 10 (35.7) 52 (37.4)
318 (69.3) 300 (64.5) 18 (64.3) 87 (62.6)
312 (28.5) 312 (28.5) 312 (28.5) 312 (28.5)
784 (71.5) 784 (71.5) 784 (71.5) 784 (71.5)
0.37 0.01 0.40 0.03
1.00 1.00 1.00 1.00
0.94 (0.73–1.21)g 0.74 (0.58–0.93)g 0.74 (0.34–1.64)g 0.68 (0.46–0.99)g
0.94 0.01 0.46 0.04
0.44 0.22
0.17 0.57 0.16
Abbreviations: pys, pack-years of smoking; AC, adenocarcinoma; LCC, large cell carcinoma; SCC, squamous cell carcinoma; SCLC, small cell lung carcinoma. a Two-sided chi-square test for genotype distribution between the cases and controls. b P-values of test for homogeneity. c Odds ratios (ORs), 95% confidence intervals (CIs) and their corresponding P-values were calculated by unconditional logistic regression analysis. d Adjusted for gender and smoking status. e Adjusted for age and smoking status. f Adjusted for age and gender. g Adjusted for age, gender and smoking status.
multiple genes and one gene can be regulated by multiple miRs (Berezikov et al., 2005; Friedman et al., 2009). Therefore, miR-146a might play different roles in different cancer cell and tissue types, and the effect of rs2910164CNG on tumorigenesis might be different in various cancers depending on miR-146a-regulating genes and its target genes in each cell and tissue type of cancer. Accordingly, meta-analyses investigating the association of polymorphisms in miR-146a and cancer risk, most of which included various types of cancers, are not likely to show a significant association between the SNP and all types of cancers as a whole. In addition, most of the recent meta-analyses included a single association study on lung cancer risk because it was the only one published on the association of rs2910164 with lung cancer risk, where Tian et al. reported the same direction of association as our result, although not significant (Tian et al., 2009). Therefore, additional studies investigating the association between rs2910164 and lung cancer risk and meta-analyses of those studies are needed to confirm the association of miR-146a with lung cancer risk. To clarify the effect of pre-miR-146a rs2910164CNG on cancer risk, further well-designed studies with a large sample size are necessary. In addition, because the frequency of the rs2910164G allele among the healthy controls is significantly different across different ethnicities, ranging from 0.26 to 0.80 (Wang et al., 2012a), the association of the SNP with cancer risk should be evaluated in diverse ethnic populations. In the present study, we evaluated the effect of this SNP on lung cancer risk in a large population. Moreover, the Korean population is genetically homogenous, reducing the risk of confounding due to population stratification. Additionally, another major strength of the present study was the inclusion criteria for the controls. Emerging evidence indicates that miR-146a regulates innate and adaptive immune and inflammatory responses and is deregulated in several diseases, including autoimmune disorders such as rheumatoid arthritis and osteoarthritis (L. Li et al., 2010, 2012; Perry et al., 2008; Rusca and Monticelli, 2011). We included only healthy subjects in the control group, who were free of disease on a health check-up. Therefore, the effect of nonmalignant disease on the genotype distribution in our control group can be excluded. We also controlled for selection bias related to hospital-based case–control studies. Given that most patients with lung cancer are treated at the University Hospital in Korea, the demographics and clinical characteristics of the patients with lung cancer in the current study were compatible with those of a nationwide
lung cancer survey (In et al., 2009). Furthermore, as all patients who were diagnosed with lung cancer at the University Hospital were included in this study, it is reasonable to assume that the case group represented the lung cancer cases in our community. In addition, because the age and gender distribution of non-participating controls were similar to those of the participating controls, self-selection bias was unlikely. Moreover, the controls were extensively evaluated for other SNPs (Bae et al., 2012; Lee et al., 2010), and the genotype distribution for each SNP met the conditions for HWE, which further supports the randomness of our control subjects. Taken together, these strengths increased the reliability of our finding of an association between the pre-miR-146a rs2910164CNG and lung cancer risk. An interesting finding of our study was that the rs2910164CNG genotypes had a more pronounced association with lung cancer risk in never smokers. This difference may be attributable to differences in the lung carcinogenesis pathways between ever and never smokers. Although tobacco smoking is a major cause of lung cancer, it has been estimated that 15% of lung cancers in men and 53% in women are never smokers, accounting for 25% of all lung cancers worldwide. In addition, lung cancer in never and ever smokers is a distinct disease that has considerable differences in etiology, genetic/molecular changes, and clinico-pathological features (Sun et al., 2007). Therefore, rs2910164CNG could confer a different impact on lung cancer risk in never and ever smokers. Because this study was designed to evaluate the effect of the SNP on overall lung cancer risk, the stratification analysis according to smoking exposure level may have had a type I error (due to multiple comparisons) and/or a type II error (due to the small number of subjects in the subgroups). Therefore, additional studies with larger sample sizes are required to confirm our findings. In the present study, we investigated the effect of the rs2910164CNG on miR-146a expression. The mature miR-146a expression level was significantly lower in tumor tissues than in non-malignant lung tissues, suggesting that miR-146a participates in lung carcinogenesis as a tumor suppressor. It should be noted that comparing the levels of a microRNA from tumors and normal tissues without microdissection could have some drawbacks due to possible differences in the cell composition of samples: normal tissues may contain a relatively larger proportion of stroma that hematopoietic cells abundant in the miR-146a may populate. There was no significant difference of mature miR-146a expression levels by the rs2910164CNG genotypes overall. However,
70
H.-S. Jeon et al. / Gene 534 (2014) 66–71
P = 0.79
P = 0.34
400
Total 300
Tumor
200 100 0
B
m 146 CG
GG
CC
P = 0.39
600
m 146 CG
GG
would affect EGFR pathway and implicate in the pathogenesis of NSCLC in never-smokers. In conclusion, we found a significant association between the premiR-146a rs2910164CNG SNP and lung cancer risk. The effect of this SNP on lung cancer risk was more pronounced in never smokers. In addition, this association was in line with the level of miR-146a expression by genotypes in the tumor tissues of never-smokers. These results suggest that this SNP contributes to genetic susceptibility to lung cancer, particularly in never smokers. Further studies on the biological function of this SNP are needed to understand the role of miR-146a in determining lung cancer risk. Conflict of interest statement
P = 0.93
All authors declare no conflict of interest.
P = 0.17
500
Acknowledgment 400
Smoker
m_Tumor
m_Normal
Mature miR-146a expression
Tumor
500
CC
300 200 100 0
This work was supported in part by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology (NRF-2010-0004700), and in part by grant no. NRF-2011-0013670. References
CC
C
CG
m ir146a
GG
CC
P = 0.48
600 500
CG
m ir146a
GG
P = 0.05
P = 0.33
P = 0.02
400
Neversmoker
n_m_Tumor
n_m_Normal
Mature miR-146a expression
Normal 600
Normal
Mature miR-146a expression
A
300 200 100 0 CC
CG
n_m ir146a
GG
CC
CG
n_m ir146a
GG
Fig. 1. Effect of the rs2910164CNG genotypes on miR-146a expression. Relative expression levels of mature miR-146a and the association with the rs2910164CNG genotypes (14 CC, 43 CG, and 12 GG in all cases; 12 CC, 28 CG, and 7 GG in smokers; 2 CC, 15 CG, and 5 GG in never-smokers) were examined in 69 tumors and paired non-malignant lung tissues. Relative expression levels of mature miR-146a by the rs2910164CNG genotypes in tumors and normal lung tissues among all cases (A), smokers (B), and never-smokers (C). The horizontal lines within the boxes represent the median values; the upper and lower boundaries of the boxes represent 75th and 25th percentiles, respectively; whiskers extend from the boxes to 1.5 times the interquartile range. P values, Student's t-test.
in line with the association of rs2910164 with lung cancer risk, the level of miR-146a expression in tumor tissues was significantly higher in GG genotype compared with CC or CG genotypes only in never-smokers after stratifying by smoking exposure. EGFR pathway plays a crucial role in many carcinogenic processes such as proliferation, angiogenesis, invasion, and metastasis, and resistance to apoptosis. EGFR is frequently overexpressed in NSCLC, and its mutation has been known as a frequent oncogenic driver mutation in NSCLC of never-smokers (Li et al., 2011). EGFR has been reported as the target of miR-146a: miR-146a down-regulated the expression of EGFR and downstream signaling, inhibited cell growth, and induced apoptosis in NSCLC cells (Chen et al., 2013). Taken together, it can be assumed that the miR-146a rs2910164C to G change may increase miR-146a expression, thereby down-regulate EGFR expression and downstream signaling, leading to cell growth inhibition and induction of apoptosis. Further studies are needed to investigate whether miR-146a rs2910164CNG
Bae, E.Y., et al., 2012. Replication of results of genome-wide association studies on lung cancer susceptibility loci in a Korean population. Respirology 17, 699–706. Berezikov, E., Guryev, V., van de Belt, J., Wienholds, E., Plasterk, R.H., Cuppen, E., 2005. Phylogenetic shadowing and computational identification of human microRNA genes. Cell 120, 21–24. Bhaumik, D., Scott, G.K., Schokrpur, S., Patil, C.K., Campisi, J., Benz, C.C., 2008. Expression of microRNA-146 suppresses NF-kappaB activity with reduction of metastatic potential in breast cancer cells. Oncogene 27, 5643–5647. Calin, G.A., et al., 2004. Human microRNA genes are frequently located at fragile sites and genomic regions involved in cancers. Proc. Natl. Acad. Sci. U. S. A. 101, 2999–3004. Chen, G., et al., 2013. miR-146a inhibits cell growth, cell migration and induces apoptosis in non-small cell lung cancer cells. PLoS One 8, e60317. Davalos, V., Esteller, M., 2010. MicroRNAs and cancer epigenetics: a macrorevolution. Curr. Opin. Oncol. 22, 35–45. Esquela-Kerscher, A., Slack, F.J., 2006. Oncomirs — microRNAs with a role in cancer. Nat. Rev. Cancer 6, 259–269. Fabbri, M., Ivan, M., Cimmino, A., Negrini, M., Calin, G.A., 2007. Regulatory mechanisms of microRNAs involvement in cancer. Expert. Opin. Biol. Ther. 7, 1009–1019. Friedman, R.C., Farh, K.K., Burge, C.B., Bartel, D.P., 2009. Most mammalian mRNAs are conserved targets of microRNAs. Genome Res. 19, 92–105. Garcia, A.I., et al., 2011. Down-regulation of BRCA1 expression by miR-146a and miR146b-5p in triple negative sporadic breast cancers. EMBO Mol. Med. 3, 279–290. Griffiths-Jones, S., Grocock, R.J., van Dongen, S., Bateman, A., Enright, A.J., 2006. miRBase: microRNA sequences, targets and gene nomenclature. Nucleic Acids Res. 34, D140–D144. Guo, H., et al., 2010. A functional variant in microRNA-146a is associated with risk of esophageal squamous cell carcinoma in Chinese Han. Fam. Cancer 9, 599–603. Hu, Z., et al., 2008. Genetic variants of miRNA sequences and non-small cell lung cancer survival. J. Clin. Invest. 118, 2600–2608. Hung, P.S., Chang, K.W., Kao, S.Y., Chu, T.H., Liu, C.J., Lin, S.C., 2012. Association between the rs2910164 polymorphism in pre-mir-146a and oral carcinoma progression. Oral Oncol. 48, 404–408. In, K.H., et al., 2009. Lung cancer patients who are asymptomatic at diagnosis show favorable prognosis: a Korean lung cancer registry study. Lung Cancer 64, 232–237. Jazdzewski, K., Murray, E.L., Franssila, K., Jarzab, B., Schoenberg, D.R., de la Chapelle, A., 2008. Common SNP in pre-miR-146a decreases mature miR expression and predisposes to papillary thyroid carcinoma. Proc. Natl. Acad. Sci. U. S. A. 105, 7269–7274. Jazdzewski, K., et al., 2009. Polymorphic mature microRNAs from passenger strand of premiR-146a contribute to thyroid cancer. Proc. Natl. Acad. Sci. U. S. A. 106, 1502–1505. Lee, S.Y., et al., 2010. Polymorphisms in the caspase genes and the risk of lung cancer. J. Thorac. Oncol. 5, 1152–1158. Li, L., Chen, X.P., Li, Y.J., 2010a. MicroRNA-146a and human disease. Scand. J. Immunol. 71, 227–231. Li, Y., et al., 2010b. miR-146a suppresses invasion of pancreatic cancer cells. Cancer Res. 70, 1486–1495. Li, C., et al., 2011. Spectrum of oncogenic driver mutations in lung adenocarcinomas from East Asian never smokers. PLoS One 6, e28204. Li, J., et al., 2012. miR-146a, an IL-1beta responsive miRNA, induces vascular endothelial growth factor and chondrocyte apoptosis by targeting Smad4. Arthritis Res. Ther. 14, R75. Lin, S.L., Chiang, A., Chang, D., Ying, S.Y., 2008. Loss of mir-146a function in hormonerefractory prostate cancer. RNA 14, 417–424.
H.-S. Jeon et al. / Gene 534 (2014) 66–71 Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods 25, 402–408. Permuth-Wey, J., et al., 2011. A functional polymorphism in the pre-miR-146a gene is associated with risk and prognosis in adult glioma. J. Neurooncol. 105, 639–646. Perry, M.M., Moschos, S.A., Williams, A.E., Shepherd, N.J., Larner-Svensson, H.M., Lindsay, M.A., 2008. Rapid changes in microRNA-146a expression negatively regulate the IL1beta-induced inflammatory response in human lung alveolar epithelial cells. J. Immunol. 180, 5689–5698. Rusca, N., Monticelli, S., 2011. MiR-146a in immunity and disease. Mol. Biol. Int. 2011, 437301. Shen, J., Ambrosone, C.B., DiCioccio, R.A., Odunsi, K., Lele, S.B., Zhao, H., 2008. A functional polymorphism in the miR-146a gene and age of familial breast/ovarian cancer diagnosis. Carcinogenesis 29, 1963–1966. Sun, S., Schiller, J.H., Gazdar, A.F., 2007. Lung cancer in never smokers-a different disease. Nat. Rev. Cancer 7, 778–790. Taganov, K.D., Boldin, M.P., Chang, K.J., Baltimore, D., 2006. NF-kappaBdependent induction of microRNA miR-146, an inhibitor targeted to signaling proteins of innate immune responses. Proc. Natl. Acad. Sci. U. S. A. 103, 12481–12486.
71
Tian, T., et al., 2009. A functional genetic variant in microRNA-196a2 is associated with increased susceptibility of lung cancer in Chinese. Cancer Epidemiol. Biomarkers Prev. 18, 1183–1187. Valencia-Sanchez, M.A., Liu, J., Hannon, G.J., Parker, R., 2006. Control of translation and mRNA degradation by miRNAs and siRNAs. Genes Dev. 20, 515–524. Wang, J., Bi, J., Liu, X., Li, K., Di, J., Wang, B., 2012a. Has-miR-146a polymorphism (rs2910164) and cancer risk: a meta-analysis of 19 case–control studies. Mol. Biol. Rep. 39, 4571–4579. Wang, J., et al., 2012b. The association of miR-146a rs2910164 and miR-196a2 rs11614913 polymorphisms with cancer risk: a meta-analysis of 32 studies. Mutagenesis 27, 779–788. Wouters, M.D., van Gent, D.C., Hoeijmakers, J.H., Pothof, J., 2011. MicroRNAs, the DNA damage response and cancer. Mutat. Res. 717, 54–66. Wu, M., et al., 2008. Genetic variations of microRNAs in human cancer and their effects on the expression of miRNAs. Carcinogenesis 29, 1710–1716. Xu, T., et al., 2008. A functional polymorphism in the miR-146a gene is associated with the risk for hepatocellular carcinoma. Carcinogenesis 29, 2126–2131. Xu, B., et al., 2010. A functional polymorphism in Pre-miR-146a gene is associated with prostate cancer risk and mature miR-146a expression in vivo. Prostate 70, 467–472. Yue, C., et al., 2011. Polymorphism of the pre-miR-146a is associated with risk of cervical cancer in a Chinese population. Gynecol. Oncol. 122, 33–37.
