RARβ2 hypermethylation is associated with poor recurrence-free survival in never-smokers with adenocarcinoma of the lung
- Yujin Kim†1,
- DongHao Jin†1,
- Bo Bin Lee1,
- Eun Yoon Cho2,
- Joungho Han2,
- Young Mog Shim3 and
- Duk-Hwan Kim1, 4Email author
© Kim et al.; licensee BioMed Central. 2015
Received: 13 October 2014
Accepted: 27 February 2015
Published: 19 March 2015
This study was aimed at investigating if the effect of RARβ2 hypermethylation on recurrence-free survival (RFS) in non-small cell lung cancer (NSCLC) depends on one’s smoking status and specific interacting proteins.
We retrospectively analyzed the expressions of five proteins using immunohistochemistry in archival formalin-fixed and paraffin-embedded tissues from 578 NSCLC patients who had undergone surgical resection from 1994 through 2004. Promoter methylation of RARβ2 was assessed by bisulfite pyrosequencing. Recurrence was found in 268 (46%) of 578 NSCLCs with a median follow-up period of 4.8 years. Overexpression of β-catenin, c-MET, cyclin D1, and EGFR occurred in 55%, 72%, 51%, and 41% of the patients, respectively. E-cadherin expression was negative in 62% of the patients, and RARβ2 hypermethylation was found in 37%. The abnormal expression of c-MET (P = 0.002) and EGFR (P = 0.001) was found to be highly prevalent in never-smokers. RARβ2 hypermethylation was significantly associated with poor recurrence-free survival (RFS) in 128 never-smokers with adenocarcinoma (P = 0.01) For parsimonious model building, the five proteins were clustered into three groups (β-catenin and E-cadherin; c-MET; cyclin D1 and EGFR) by an unsupervised hierarchical clustering and were included in a multivariate analysis. Cox proportional hazard analysis showed that RARβ2 hypermethylation was significantly associated with poor RFS in 128 never-smokers with adenocarcinoma (adjusted hazard ratio [HR] = 2.19, 95% confidence interval [CI] = 1.28 to 3.47; P = 0.009), after adjusting for interacting proteins.
The present study suggests that RARβ2 hypermethylation may be an independent prognostic factor of RFS in never-smokers with adenocarcinoma of the lung.
KeywordsLung cancer RARβ2 Methylation Never-smoker Recurrence
Despite significant advances in diagnosis and treatment over the past two decades, lung cancer is the leading cause of cancer-related death in the world, and the overall five-year survival rate remains under 15% [1,2]. The poor prognosis of lung cancer is partially due to the high rate of recurrence after surgical resection. Given that some patients may benefit from postoperative adjuvant therapy shortly after resection, the discovery of biomarkers for the identification of patients who are at a high risk of recurrence has become increasingly important for novel targeted therapy and for improving recurrence prevention. In recent years, a number of genetic and epigenetic molecules have been discovered for the purpose, and novel therapeutic agents are being developed that target those molecules.
Retinoic acid (RA) induces growth inhibition and apoptosis mainly by regulating gene expression through its nuclear receptors, known as retinoic acid receptors (RARs) and retinoid X receptors (RXRs). The receptors bind to a retinoic acid response element (RARE), consisting of two 5′-AGGTCA-3′ direct repeats of the consensus half-site sequence 5′-AGGTCA-3′, in the promoter region of the target genes to initiate their transcription. Both RAR and RXR receptors have three subtypes (α, β, and γ) which have distinct functions. Of these receptors, the retinoic acid receptor beta (RARβ) is expressed primarily in epithelial cells and plays a central role in mediating the growth inhibition of different types of cancer cells by retinoic acid. The loss and reduction of RARβ expression have been reported in a large percentage of patients with lung cancer. The human RARβ gene has four isoforms (β1, β2, β3, and β4), and the β2 isoform is most abundant among the isoforms. In addition, several studies have indicated that RARβ2 is primarily responsible for limiting the growth of lung cancer cells [3-5]. Indeed, expression of RARβ2 in RARβ-negative lung cancer cells has been shown to restore retinoic acid-induced growth inhibition , and transgenic mice expressing antisense or other constructs that down-regulate RARβ2 developed lung cancer .
The aberrant methylation of CpG islands at a promoter region of tumor suppressor genes is an epigenetic change that induces the transcriptional silencing of a gene. The RARβ2 promoter contains a CpG island in the 5′-untranslated region that is frequently methylated in cancer cells. Abnormal hypermethylation of the RARβ2 CpG island has been found in approximately 40% of non-small cell lung cancers (NSCLCs) and identified as part of a mechanism by which RARβ2 expression is repressed [6-8]. In contrast to many previous reports about the tumor-suppressive effect of RARβ in lung cancer, Kurie et al.  reported a negative effect of RARβ expression on overall survival of stage I NSCLC. We also reported a protective effect of RARβ2 hypermethylation on the development of second primary lung cancers (SPLCs) in smokers with NSCLC . Based on these reports, we hypothesized that an effect of RARβ2 hypermethylation on the recurrence in NSCLC may differ according to exposure to tobacco smoke.
Cyclin D1, β-catenin, and E-cadherin are known to interact with RAR [10-18], and the altered expression of EGFR is known to occur at a high prevalence in never-smokers [19,20]. In an effort to investigate whether RARβ2 hypermethylation contributes to the development of recurrence in NSCLC according to statuses of smoking and interacting proteins, we analyzed the methylation status of RARβ2 and the expression statuses of cyclin D1, E-cadherin, β-catenin, EGFR, and c-MET in 578 NSCLC patients.
Correlation of expression levels among five proteins differs according to smoking status
Clinicopathological features ( N = 578)
60 ± 11
Tumor size (cm)a
4.1 ± 2.4
Pack-years of smokinga
32 ± 23
The effect of RARβ2 hypermethylation on RFS in never-smokers is different according to histology
Lung cancer histology is known to be associated with smoking status: adenocarcinoma develops more frequently in never-smokers, whereas other cell types are more common in ever-smokers. This suggests that the relationship between RFS and RARβ2 hypermethylation in never-smokers may be confounded by a histologic subtype. Therefore, the data was stratified by histologic subtypes, and RFS in never-smokers were further analyzed under this classification. For adenocarcinoma cases, five-year RFS rates were 27% and 46% in patients with RARβ2 hypermethylation and in those without, respectively, and this difference was statistically significant (P = 0.01; Figure 3C). However, RFS in cases with other histologic subtypes was not found to be significantly different between patients with and without RARβ2 hypermethylation (P = 0.28; Figure 3D).
No heterogeneity of association between RARβ2 hypermethylation and recurrence across the proteins
To adjust for effect modifiers or confounding factors in the relationship between RARβ2 hypermethylation and RFS in never-smokers with adenocarcinoma, we evaluated if the population odds ratios between RARβ2 hypermethylation and the risk of recurrence were uniform across the altered expression of individual proteins in never-smokers with adenocarcinoma. The Breslow-Day test for homogeneity of the odds ratio showed no evidence of heterogeneity in cases of RARβ2 hypermethylation across the altered expression of β-catenin (P = 0.18), c-MET (P = 0.48), cyclin D1 (P = 0.27), E-cadherin (P = 0.95), and EGFR (P = 0.68). The altered expression of individual proteins might not be an effect modifier in the relationship between RARβ2 hypermethylation and recurrence in our population. Therefore, we did not construct various contingency tables as if they had been drawn from distinct populations and computed a single summary measure rather than individual hazard ratio across the proteins for the 128 cases.
No association between RARβ2 hypermethylation and protein expression levels
Univariate Cox proportional hazards analysis of RFS in never-smokers with adenocarcinoma ( N = 128)
HR (95% CI)
0.98 (0.95 to 1.04)
1.25 (1.12 to 1.40)
0.64 (0.40 to 0.97)
1.32 (0.97 to 1.75)
1.72 (1.41 to 2.08)
1.71 (1.16 to 2.57)
0.82 (0.54 to 1.24)
0.77 (0.46 to 1.25)
1.09 (0.72 to 1.65)
1.14 (0.73 to 1.78)
1.10 (0.74 to 1.63)
Multivariate Cox proportional hazards analysis of RFS
Multivariate Cox proportional hazards analysis a of RFS according to RARβ2 hypermethylation, stratified by histology, in never-smokers
(N = 128)
1.28 to 3.47
B. Other cell types
(N = 23)
0.71 to 16.23
Several studies have suggested that retinoids may have harmful effects in the prevention of lung cancer. However, most studies about the tertiary chemopreventive effect of retinoids in surgically treated lung cancer have focused primarily on smokers, and relatively few studies have examined the chemopreventive effect of retinoids in never-smokers. Although tobacco smoking accounts for the majority of lung cancer as a presumed etiology, it is estimated that 15% of male and 53% of female lung cancer patients are lifelong never-smokers worldwide . More and more studies have suggested that lung cancers occurring in never-smokers may be a distinctive disease due to unique molecular profiles [22-25]. To understand the effect of RARβ2 hypermethylation on recurrence according to smoking status in NSCLC, we analyzed RARβ2 hypermethylation and expression levels of five proteins in 584 NSCLCs. The proteins were adjusted as confounding factors in the relationship of RARβ2 hypermethylation with recurrence.
Among the five proteins analyzed in this study, β-catenin and E-cadherin are known to interact with RAR [10-16]. β-catenin is a protein associated with the cytoplasmic tail of E-cadherin, and the level of β-catenin in the cytoplasm determines the activation of Wnt responsive genes. Nuclear β-catenin interacts with the TCF/LEF (T cell factor/lymphocyte enhancer factor) family of transcription factors and induces the transcription of genes involved in cell proliferation and/or inhibition of apoptosis. Thus, constitutive activation of β-catenin may be a critical step in tumorigenesis among divergent types of cancers. Retinoid signaling has been found to be a potent inhibitor for Wnt/β-catenin signaling in colon and breast cancer cells , neuronal cells , bronchial epithelial cells , and ES cells . E-cadherin is also known to interact with RARs. In breast cancer cells, RA up-regulates the function of the invasion-suppressor complex E-cadherin/catenin . Cadherin expression and function are necessary to mediate the effects of RA on adhesion and differentiation .
The interaction between retinoid signaling and cyclin D1 has also been reported in a variety of cancer cells. Retinoic acid triggers G1 cell cycle arrest through the degradation of cyclin D1 via the ubiquitin-proteasome pathway . Ectopic expression of cyclin D1 sensitizes ER-positive breast cancer cells to a retinoic acid-induced mitochondrial death pathway . The activity of the cyclin D1 reporter was reduced in response to retinoic acid in breast cancer cells independently of β-catenin/TCF signaling . However, in this study, the altered expression of cyclin D1 was not associated with RARβ2 hypermethylation (Figure 3A) or recurrence. These observations suggest that interaction between RARβ2 signaling and E-cadherin/β-catenin or cyclin D1 may be different in never-smokers. In addition, RARβ2 hypermethylation may affect RFS irrespective of an alteration of interacting proteins in never-smokers.
In this study, RARβ2 hypermethylation was not associated with expression levels of the five proteins (Figure 4A). We further analyzed the association between RARβ2 expression in 48 fresh-frozen tumor tissues and the expression of the other five proteins (data not shown). Twenty-six (54%) of 48 tumor tissues were found to lose RARβ2 expression. No association was found between RARβ2 expression and the expression of five proteins in the 48 samples. However, the association showed a different pattern in samples stratified by histology: RARβ2 expression showed a reciprocal relationship with EGFR expression in adenocarcinoma and with c-MET expression in squamous cell carcinoma, suggesting that RARβ2 may interact with EGFR or c-MET in a tissue-dependent manner. Further study in a larger sample is required to clearly understand the relationship between RARβ2 and EGFR (or c-MET) according to histological subtypes.
Mutations in the epidermal growth factor receptor (EGFR) gene are more common in lung cancers from never-smokers compared to ever-smokers in both men and women. EGFR overexpression in this study was found in 71 (47%) of 151 never-smokers and was more frequent in adenocarcinoma (56%) than in squamous cell carcinoma (34%); results which were consistent with previous reports [19,20]. The c-MET is a proto-oncogene that codes for a protein known as hepatocyte growth factor receptor (HGFR), which has tyrosine kinase activity. In this study, overexpression of c-MET was found at a higher prevalence in never-smokers compared to ever-smokers (Figure 1B; P = 0.002). However, the altered expression of β-catenin, cyclin D1, and E-cadherin occurred at a similar frequency in never-smokers and ever-smokers. Based on these observations, some of the molecular alterations in never-smokers with NSCLC may be similar to those typically seen in lung cancer associated with tobacco smoking, but separate and distinct molecular features seem to exist in never-smokers.
EGFR overexpression in this study was not associated with RFS in never-smokers with adenocarcinoma (Table 2). To understand if EGFR mutations can affect RFS, we sequenced the exons 18, 19, and 21 of EGFR in 48 fresh-frozen tumor tissues (data not shown). The deletion in exon 19 or L858R missense mutation in exon 21 were found in 17 (34%) of 48 tumor tissues. The effect of the EGFR mutation on RFS was tested by Kaplan-Meier survival curve. Patients with an EGFR mutation showed a trend toward poor RFS, though it was not statistically significant (P = 0.08). The lack of an effect of the EGFR mutation on RFS in this study may have resulted from the small sample size or from the lack of targeted chemotherapy of tyrosine kinase inhibitors. The effect of the EGFR mutation on RFS in never-smokers with adenocarcinoma warrants further study in a large prospective cohort.
Is it reasonable to recommend the intake of retinoids for the chemoprevention of recurrence in never-smokers who have undergone surgical resection for NSCLC? The response to retinoic acid may be different according to a patient’s exposure to tobacco smoke. A significant restoration of RARβ expression and reduction of metaplasia were found in former-smokers treated with 9-cis-retinoic acid . Treatment of 9-cis-RA in former-smokers modulated serum concentrations of insulin-like growth factor (IGF)-I, IGF binding protein (IGFBP)-3, and their molar ratio (IGF-I/IGFBP-3) . In contrast to our previous finding that the hypermethylation of RARβ2 gene in current-smokers had a beneficial effect on the development of SPLC in NSCLC (8), the present study suggests that RARβ2 hypermethylation may be associated with poor RFS in never-smokers with adenocarcinoma. Taken together, the combination of retinoic acid and the demethylating agent is recommended for the prevention of recurrence irrespective of alteration of the five proteins in never-smokers with adenocarcinoma. In addition, retinoic acid may be beneficial for the primary prevention of lung cancer in never-smokers with high risk factors, such as particular genetic mutations predisposing them to cancer development and for the secondary prevention of disease progression in never-smokers with premalignant lesions.
This study was severely limited by several factors. First, the number of cases with other cell types rather than adenocarcinoma was very small. For the application of retinoic acid in a clinical practice, the effect of RARβ2 hypermethylation on RFS in never-smokers with adenocarcinoma as well as other cell types should be further validated by prospective large-scale studies. Second, the molecular mechanisms underlying the finding that RARβ2 hypermethylation contributes to poor RFS in never-smokers with adenocarcinoma should be studied. Third, other proteins interacting with RAR that were not analyzed in this study also need to be analyzed to clearly understand the effect of retinoic acid on RFS in never-smokers.
RARβ2 hypermethylation is known to occur in approximately 40% of NSCLCs. The present study suggests that RARβ2 hypermethylation may be associated with poor RFS in never-smokers with adenocarcinoma.
A total of 578 formalin-fixed paraffin-embedded (FFPE) tissues were obtained from NSCLC patients undergoing surgical resection between May 1994 and September 2004 at the Samsung Medical Center in Seoul, Korea. The median duration of follow-up was 57 months. A written informed consent was obtained from the patients before surgery. This study was approved by the Samsung Medical Center (SMC) Institutional Review Board (IRB). Post-operative follow-up for the detection of local and distant recurrence was performed as previously described [2,8], and recurrence was evaluated from information obtained from our hospital’s electronic medical records and those from other hospitals, as of November 2013. All cases of cancer were classified based on the guidelines of the tumor-node-metastasis (TNM) staging system of the American Joint Committee on Cancer . Never-smoker was defined as a person who has never smoked or who has smoked fewer than 100 cigarettes in his or her lifetime.
Tissue microarrays (TMAs) of NSCLCs and immunohistochemical staining of the five proteins were conducted as previously described . The sections were incubated overnight at 4°C with the following primary antibodies: β-catenin (clone 17C2; Novocastra, Buckinghamshire, UK) at 1:100 dilution, c-MET (3D4, Zymed, South San Francisco, CA, USA) at 1:200 dilution, cyclin D1 (SP4, Lab Vision, Fremont, CA, USA) at 1:50 dilution, E-cadherin (4A2C7, Zymed, South San Francisco, CA) at 1:100 dilution, or EGFR (1005, Santa Cruz, CA, USA) at 1:100 dilution. Detection of immunoreactivity by each antibody was performed by EnvisionTM+ peroxidase (Dako, Carpinteria, CA, USA). The peroxidase activity for each antibody was visualized by applying diaminobenzidine chromogen containing 0.05% hydrogen peroxide for 3 to 8 min at room temperature. Nuclei were then counterstained with Mayer’s hematoxylin, and negative controls were included in each run by omitting the primary antibody.
Interpretation of immunohistochemical staining
The slides were read by two authors (EYC, JH), who were blinded to the clinical outcome of the patients, to reduce inter-observer variability. The correlation of scores between the two readers of immunohistochemical staining was calculated, and samples showing kappa values less than 0.20, poor inter-rater reliability, were removed from further analysis. Cyclin D1 was stained in the cytoplasm and nucleus, but only nuclear staining of cancer cells was considered positive. The immunoreactivity of EGFR and E-cadherin was found mainly in the membrane of neoplastic cells, and membrane staining was considered in scoring. The c-MET is localized in both the cell membrane and cytoplasm of cancer cells, and immunoreactivity in the cytoplasm and membrane was semi-quantitatively graded by considering the intensity of staining. Immunoreactivity for β-catenin was observed in the membrane, cytoplasm, and nucleus, and the intensity and proportion of positive staining cells only in the cytoplasm were evaluated for scoring.
The expression levels of five proteins were calculated by multiplying the intensity score (0, none; 1, weak; 2, moderate; 3, strong) and the proportion score of positive staining tumor cells (0, absent; 1, 0% to 10%; 2, 10% to 50%; 3, 50% to 80%; 4, >80%). Cutoff values of the composite scores for abnormal expression were determined by using an internal control consisting of 62 normal lung cores from 152 lung cancer blocks. To determine the cutoffs for abnormal expression of individual protein, we first analyzed the distribution of the composite scores for abnormal expression of β-catenin, cyclin D1, EGFR, E-cadherin, and c-MET proteins between the normal and tumor tissues. Based on the distribution, we set several cutoffs, compared false positives and false negatives at different cutoffs, and finally determined the cutoffs of the composite score that minimized false positives and false negatives. The β-catenin, c-MET, cyclin D1, and EGFR were considered to be overexpressed if the composite score was greater than or equal to 2. E-cadherin was defined as negative if the composite score of membrane staining was less than 2.
To microdissect tumor areas, we cut formalin-fixed, paraffin-embedded tissues into 10-μm-thick sections. Before DNA extraction, we performed hematoxylin-eosin staining to locate the tumor areas and microdissected areas containing at least 75% neoplastic tissue. Five hundred nanograms of genomic DNA was purified using the DNeasy Blood & Tissue Kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s protocol and then bisulfite-converted using the EZ DNA methylation kit (Zymo Research, Irvine, CA, USA). The methylation levels of CpG islands at the P2 promoter region of the RARβ2 gene were analyzed by pyrosequencing of bisulfite-treated DNA, as previously described by Lee et al. . Pyrosequencing was carried out on a PSQ 96MD system with the PyroMark PCR kit (Cat. No. 978703, Qiagen), and data were analyzed using the Q-CpG software (v.1.0.9, Pyrosequencing AB, Uppsala, Sweden).
Quantitative reverse transcription-PCR (qRT-PCR)
The expression of RARβ2 was measured using TaqMan assays following the manufacturer’s instructions on the ABI 7500 Fast Real-Time PCR system (Applied Biosystems, Waltham, MA, USA). One microgram of total RNA was reverse transcribed using the SuperScript VILO cDNA synthesis kit (Cat. No. 11754-050, Invitrogen, Carlsbad, CA, USA), and the resulting cDNA was used for real-time PCR. The GAPDH was included as an internal reference for RARβ2 expression. TaqMan primer sets (Hs00977141-mH for RARβ2 and Hs0275899_g1 for GAPDH) were purchased from Life Technologies. TaqMan probe sequences for RARβ2 and GAPDH were FAM-CCTGCCTGGACATCCTGATTCTTAG-MBG (110 bp) and FAM-GACTCATGACCACAGTCCATGCCAT-MBG (93 bp), respectively. RARβ2 expression values were normalized to the reference by using the comparative Ct method.
In the univariate analysis, the t-test (or Wilcoxon rank-sum test or Kruskal-Wallis test) was used for the analysis of continuous variables. Correlations among expression levels of the five proteins were compared using Spearman’s rank correlation coefficients. The prognostic significance of the five proteins and RARβ2 on recurrence-free survival (RFS) was assessed by Kaplan-Meier survival curves, and the significance of the difference in RFS between two groups was evaluated using the log-rank test. The hazard ratios of independent factors for RFS were estimated by multivariate Cox proportional hazard regression analysis. All statistical analyses were two sided with a 5% type I error rate. Expression patterns of the five proteins were recognized using an unsupervised hierarchical cluster analysis.
non-small cell lung cancer
methylation-specific polymerase chain reaction
epidermal growth factor receptor
The authors wish to thank Eunkyung Kim and Jin-Hee Lee for data collection and management and Hoon Suh for sample collection. This work was supported by grants from the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2014R1A1A2053325) and from the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare (HI14C1979), Republic of Korea.
- Jermal A, Siegel R, Ward E, Hao Y, Xu J, Murray T, et al. Cancer statistics, 2008. CA Cancer J Clin. 2008;58:71–96.View ArticleGoogle Scholar
- Kim JS, Kim JW, Han J, Shim YM, Park J, Kim DH. Cohypermethylation of p16 and FHIT promoters as a prognostic factor of recurrence in surgically resected stage I non-small cell lung cancer. Cancer Res. 2006;66:4049–54.View ArticlePubMedGoogle Scholar
- Moghal N, Neel BG. Evidence of impaired retinoic acid receptor-thyroid hormone receptor AF-2 cofactor activity in human lung cancer. Mol Cell Biol. 1995;15:3945–59.PubMed CentralPubMedGoogle Scholar
- Bérard J, Laboune F, Mukuna M, Massé S, Kothary R, Bradley WE. Lung tumors in mice expressing an antisense RARbeta2 transgene. FASEB J. 1996;10:1091–7.PubMedGoogle Scholar
- Toulouse A, Morin J, Dion PA, Houle B, Bradley WE. RARbeta2 specificity in mediating RA inhibition of growth of lung cancer-derived cells. Lung Cancer. 2000;28:127–37.View ArticlePubMedGoogle Scholar
- Virmani AK, Rathi A, Zöchbauer-Müller S, Sacchi N, Fukuyama Y, Bryant D, et al. Promoter methylation and silencing of the retinoic acid receptor-beta gene in lung carcinomas. J Natl Cancer Inst. 2000;92:1303–7.View ArticlePubMedGoogle Scholar
- Zöchbauer-Müller S, Fong KM, Virmani AK, Geradts J, Gazdar AF, Minna JD. Aberrant promoter methylation of multiple genes in non-small cell lung cancers. Cancer Res. 2001;61:249–55.PubMedGoogle Scholar
- Kim JS, Lee H, Kim H, Shim YM, Han J, Park J, et al. Promoter methylation of retinoic acid receptor beta 2 and the development of second primary lung cancers in non-small-cell lung cancer. J Clin Oncol. 2004;22:3443–50.View ArticlePubMedGoogle Scholar
- Kurie JM, Lotan R, Lee JJ, Lee JS, Morice RC, Liu DD, et al. Treatment of former smokers with 9-cis-retinoic acid reverses loss of retinoic acid receptor-beta expression in the bronchial epithelium: results from a randomized placebo-controlled trial. J Natl Cancer Inst. 2003;95:206–14.View ArticlePubMedGoogle Scholar
- Vermeulen SJ, Bruyneel EA, van Roy FM, Mareel MM, Bracke ME. Activation of the E-cadherin/catenin complex in human MCF-7 breast cancer cells by all-trans-retinoic acid. Br J Cancer. 1995;72:1447–53.View ArticlePubMed CentralPubMedGoogle Scholar
- Byers S, Pishvaian M, Crockett C, Peer C, Tozeren A, Sporn M, et al. Retinoids increase cell-cell adhesion strength, beta-catenin protein stability, and localization to the cell membrane in a breast cancer cell line: a role for serine kinase activity. Endocrinol. 1996;137:3265–73.Google Scholar
- Easwaran V, Pishvaian M, Salimuddin, Byers S. Cross-regulation of β-catenin-LEF/TCF and retinoid signaling pathways. Curr Biol. 1999;9:1415–8.View ArticlePubMedGoogle Scholar
- Shah S, Hecht A, Pestell RG, Byers SW. Trans-repression of β-catenin activity by nuclear receptors. J Biol Chem. 2003;278:48137–45.View ArticlePubMedGoogle Scholar
- Otero JJ, Fu W, Kan L, Cuadra AE, Kessler JA. Beta-catenin signaling is required for neural differentiation of embryonic stem cells. Dev. 2004;131:3545–57.View ArticleGoogle Scholar
- Li Y, Wu RL, Wang X, Chen WS. Effects of retinoic acid on the signal transduction of beta-catenin/Tcf of cultured bronchial epithelial cells. Zhonghua Yi Xue Za Zhi. 2004;84:1130–1.PubMedGoogle Scholar
- Liu T, Lee YN, Malbon CC, Wang HY. Activation of the beta-catenin/Lef-Tcf pathway is obligate for formation of primitive endoderm by mouse F9 totipotent teratocarcinoma cells in response to retinoic acid. J Biol Chem. 2002;277:30887–91.View ArticlePubMedGoogle Scholar
- Boyle JO, Langenfeld J, Lonardo F, Sekula D, Reczek P, Rusch V, et al. Cyclin D1 proteolysis: a retinoid chemoprevention signal in normal, immortalized, and transformed human bronchial epithelial cells. J Natl Cancer Inst. 1999;91:373–9.View ArticlePubMedGoogle Scholar
- Niu MY, Ménard M, Reed JC, Krajewski S, Pratt MA. Ectopic expression of cyclin D1 amplifies a retinoic acid-induced mitochondrial death pathway in breast cancer cells. Oncogene. 2001;20:3506–18.View ArticlePubMedGoogle Scholar
- Kosaka T, Yatabe Y, Endoh H, Kuwano H, Takahashi T, Mitsudomi T. Mutations of the epidermal growth factor receptor gene in lung cancer: biological and clinical implications. Cancer Res. 2004;64:8919–23.View ArticlePubMedGoogle Scholar
- Sonobe M, Manabe T, Wada H, Tanaka F. Mutations in the epidermal growth factor receptor gene are linked to smoking-independent, lung adenocarcinoma. Br J Cancer. 2005;93:355–63.View ArticlePubMed CentralPubMedGoogle Scholar
- Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin. 2005;55:74–108.View ArticlePubMedGoogle Scholar
- Subramanian J, Govindan R. Molecular genetics of lung cancer in people who have never smoked. Lancet Oncol. 2008;9:676–82.View ArticlePubMedGoogle Scholar
- Rudin CM, Avila-Tang E, Harris CC, Herman JG, Hirsch FR, Pao W, et al. Lung cancer in never smokers: molecular profiles and therapeutic implications. Clin Cancer Res. 2009;15:5646–61.View ArticlePubMed CentralPubMedGoogle Scholar
- Subramanian J, Govindan R. Lung cancer in ‘Never-smokers’: a unique entity. Oncol (Williston Park). 2010;24:29–35.Google Scholar
- Yano T, Miura N, Takenaka T, Haro A, Okazaki H, Ohba T, et al. Never-smoking nonsmall cell lung cancer as a separate entity: clinicopathologic features and survival. Cancer. 2008;113:1012–8.View ArticlePubMedGoogle Scholar
- Lee HY, Chang YS, Han JY, Liu DD, Lee JJ, Lotan R, et al. Effects of 9-cis-retinoic acid on the insulin-like growth factor axis in former smokers. J Clin Oncol. 2005;23:4439–49.View ArticlePubMedGoogle Scholar
- Edge SB, Byrd DR, Compton CC, Fritz AF, Greene FL, Trotti A. American Joint Committee on cancer. In: AJCC cancer staging manual. 7th ed. New York: Springer; 2010. p. 253–70.Google Scholar
- Ko E, Lee BB, Kim Y, Lee EJ, Cho EY, Han J, et al. Association of RASSF1A and p63 with poor recurrence-free survival in node-negative stage I-II non-small cell lung cancer. Clin Cancer Res. 2013;19:1204–12.View ArticlePubMedGoogle Scholar
- Lee ES, Issa JP, Roberts DB, Williams MD, Weber RS, Kies MS, et al. Quantitative promoter hypermethylation analysis of cancer-related genes in salivary gland carcinomas: comparison with methylation-specific PCR technique and clinical significance. Clin Cancer Res. 2008;14:2664–72.View ArticlePubMedGoogle Scholar
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.