Main

Lung cancer is the leading cause of cancer deaths in both men and women worldwide, with over a million deaths annually.1 Histopathologically, lung cancer is classified as adenocarcinoma, squamous carcinoma, large cell carcinoma, small cell carcinoma, and other subtypes that are less frequently diagnosed.2 The first three types of lung cancer collectively accounts for 85% of lung cancers.

Tobacco smoke is the predominant etiologic risk factor for lung cancer.3 Carcinogens present in tobacco smoke or their intermediate metabolites might bind covalently to DNA at certain specific sites, forming bulky adducts and leading to gene mutations. However, a fraction of lung cancers, especially adenocarcinomas, arise in patients who have never smoked, indicating that host susceptibility is a factor in lung cancer carcinogenesis. It is still not fully understood whether the specific host susceptibility, alone or together with the environmental factors including tobacco smoking, has a unique or a synergetic role in the carcinogenesis of lung cancer. In lung cancer, heritable genetic changes can occur at chromosomal level with bulky loss, gain or translocation of chromosomes. At molecular level, the changes may be mutations in specific genes such as single-nucleotide polymorphism (SNP) or deletion. In addition to gene structural changes, reversible changes in gene expression that may be independent of changes in the primary DNA sequence can occur. These so-called epigenetic changes include DNA methylation, histone modifications, and abnormal expression of non-coding RNAs including microRNAs (miRNAs). These genetic and epigenetic alternations interact at all stages of cancer development, working together to promote cancer progression.4

In this review, we summarize the characteristic genetic and epigenetic changes observed in lung cancers and the implication of these changes in patients for early detection, survival prediction, and treatment responses.

Genetic changes

Chromosomal Aberration—Loss of Heterozygosity

Somatic alterations in cellular DNA are common in almost all human cancers, including lung cancers. By analysis of 371 lung adenocarcinomas using dense SNP arrays, Weir et al5 found that 26 of 39 autosomal chromosome arms showed consistent large-scale copy-number gain or loss and 31 focal alterations, including 24 amplifications and 7 homozygous deletions. Previous studies suggested that fractional allelic losses located primarily on chromosomes 1p, 2p, 2q, 3p, 6q, 7p, 9p, 12p, 16p, 17p, 17q, 19p, and 21q occur more frequently in lung carcinomas than in adjacent normal tissues,6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 indicating the existence of tumor-suppressor genes or potential candidates, such as HLJ1 at 1p31, FHIT at 3p14, RASSF1A, FUS1, LIMD1, SEMA3B, and SEMA3F at 3p21, p16 at 9p21, and p53 at 17p13.

Genetic changes similar to those found in lung cancers can be detected in the non-malignant bronchial epithelium of current and former smokers, and such changes may persist for many years after smoking cessation. Wistuba et al9 reported that biopsy specimens from smokers with either normal or preneoplastic lung tissue showed loss of heterozygosity (LOH) at multiple chromosomal sites, a phenomenon frequently observed in carcinoma in situ and invasive cancer. However, no genetic alterations were detected in non-smokers.9 Recently, Yendamuri et al18 reported increased deletions of the 3p22.1 and 10q22.3 regions in the bronchial epithelium in the lung harboring the tumor and the tumor itself compared with the contralateral normal bronchial epithelium in 122 squamous cell carcinoma/adenocarcinoma patients. These results indicated that the detection of genetic changes, such as LOH, might identify individuals at high risk for developing lung cancer.

Certain chromosomal changes are related to lung cancer's metastasis. Wrage et al19 reported that the loss of 4q, especially 4q12–q23, in primary lung cancer was significantly associated with bone marrow metastasis. In addition, the same loss was also found to be common in brain metastases from lung cancer patients. Therefore, a specific pattern of genomic changes, such as 4q deletion, might drive an early hematogenous dissemination of lung cancer cells.

The LOH or other chromosomal abnormalities observed in lung cancers could be associated with patient survival. Bepler et al20 observed an association between LOH at 11p15.5 and poor survival in 180 lung cancer patients with adenocarcinoma, squamous cell carcinoma, and large cell carcinoma. Marsit et al12 reported that LOH at 9p13 was a significant predictor of improved survival, whereas the homozygous deletion was associated with poor survival in 100 squamous cell carcinoma/adenocarcinoma patients undergoing surgical resection. Tseng et al21 linked LOH at 1p36.23 with smoking, squamous carcinoma, and late-stage disease. Furthermore, LOH at q37.3 and 6p21–p22 were significantly associated with poor prognosis in squamous cell carcinoma/adenocarcinoma patients, using both univariate and multivariate Cox regression analyses. These markers can potentially be used for early lung cancer detection and prognosis and for potentially identifying novel clones of new tumor-suppressor genes that might contribute to carcinogenesis of squamous cell carcinoma/adenocarcinoma when their functions were lost. These findings need to be examined further in larger studies to confirm the association of LOH with patient survival.

p53 Mutation

Mutations in the p53 tumor-suppressor gene are one of the most frequent changes identified in human tumor cells. The common mutations, usually occurring in the DNA-binding domains of p53 gene, lead to the formation of the mutant forms with altered amino-acid sequences that lack DNA-binding activity.

It has been reported that 40–60% of lung cancers are associated with mutations in the p53 gene,22, 23 and smoking is a primary factor inducing p53 gene mutation. Evidence indicated that these mutations were more commonly observed in tobacco-associated lung cancer (26–71%) than in lung cancer of never-smokers (10–47%).24, 25, 26, 27, 28, 29, 30 A significant dose–response relation between tobacco smoking and p53 gene mutations in lung cancer has been reported. In addition to the mutation frequency, the p53 mutational signatures, that is, the ratio of transitions, transversions and deletions, and the mutational spectrum, such as the distribution of mutations along the gene, are distinct in lung cancers between smokers and never-smokers. Tobacco-associated lung cancer is characterized by a high frequency of G to T transversion with a pronounced coding strand bias of 93%, whereas lung cancer in never-smokers shows a higher proportion of G to C transversions and G to A transitions at CpG dinucleotides,23 suggesting that G to T transversion is a molecular signature of mutagenesis by distinct exogenous factors such as tobacco smoking.

p53 protein expression probably is a prognostic factor for survival in lung cancer patients. In a meta-analysis of 74 eligible papers dealing with p53 assessment in lung cancer, the combined hazard ratios indicated that an abnormal p53 status had an unfavorable effect on survival at each clinical stage of squamous cell carcinoma/adenocarcinoma.31 Recently, in a total of 131 cases of primary lung adenocarcinoma, both univariate and multivariate analyses showed that overexpression of p53 protein was an independent prognostic factor in node-negative lung adenocarcinoma.32 In another 266 lung cancer patients, patients with stage I disease and p53 mutation had a hazard ratio of 1.79 (95% confidence interval, 1.04–3.10) for overall survival compared with patients with the wild-type p53 gene.33 However, the role of p53 as a prognostic factor for survival in lung cancer is still controversial. Kosaka et al34 found that p53 mutations were not an independent prognostic factor in their cohort of patients with lung adenocarcinoma. Similarly, Lim et al35 also reported that p53 mutations did not have a survival effect in 88 lung cancer patients with squamous cell carcinoma, adenocarcinoma, and large cell carcinoma. The controversy is probably due to the different methods or antibodies used to detect p53 gene mutation or protein expression, or the variation in stage classification of lung cancer patients. Therefore, further clinical studies are required to determine whether p53 gene mutation can be used as a predictive marker for lung cancer patients.

The role of p53 as a chemosensitivity predictive factor in lung cancer has also been studied. In a phase III intergroup trial that randomly assigned 482 patients with completely resected stage IB and II squamous cell carcinoma/adenocarcinoma either to receive four cycles of adjuvant cisplatin plus vinorelbine or to be observed without further intervention, Tsao et al36 found untreated p53-positive patients had significantly shorter overall survival than did patients with p53-negative tumors (hazard ratio 1.89; 95% confidence interval, 1.07–3.34; P=0.03). However, these p53-positive patients also showed a significantly greater survival benefit from adjuvant chemotherapy (hazard ratio 0.54; P=0.02) compared with patients with p53-negative tumors (hazard ratio 1.40; P=0.26; interaction P=0.02). The results suggested that p53 protein overexpression is a significant prognostic marker of shortened survival as well as a predictive marker for patients who might benefit from adjuvant chemotherapy after surgical resection. Similarly, a retrospective study of 55 patients with adenocarcinoma/squamous cell carcinoma also suggested that FHIT−/p53+ status might be a biological variable affecting the efficacy of carboplatin/gemcitabine treatment in patients.37

EGFR Mutation

The epidermal growth factor receptor (EGFR) family comprises c-erbB-1 (EGFR, HER-1), c-erbB-2 (HER-2/neu), c-erbB-3, and erbB-4, located in the cellular membrane. On binding of ligands such as transforming growth factor α and epidermal growth factor (EGF), the receptors form homodimers or heterodimers with other family members, resulting in autophosphorylation of key tyrosine residues in the receptor cytoplasmic domain and in further activation of downstream signaling events, including PI3K/Akt/mTOR, Ras-Raf-MEK-ERK, and JAKs-STATs, that trigger anti-apoptosis, cell proliferation, angiogenesis, tumor invasion, and metastasis.38

Higher frequencies of mutations in the EGFR tyrosine kinase domain are found in adenocarcinomas from Asian patients (25–50%) in comparison with those in North American and Western European patients (10%).39 The most common EGFR mutations observed in lung cancer are small in-frame deletions in exon 19 and point mutations that result in substitution of arginine for leucine at amino acid 858 (L858R) in exon 21, which account for 50% and 40% of total EGFR mutations in lung cancers with exclusion of small cell carcinoma, respectively. Both point mutations in exon 18 and in-frame insertions and point mutations in exon 20 account for 5% of EGFR mutations in lung cancer with exclusion of small cell carcinoma.40 These mutations induce oncogenic transformation in vitro and in vivo through constitutive activation of EGFR. It was reported that EGFR mutations, including those involving exons 18, 19, and 20 and L858R, can transform fibroblasts and lung epithelial cells in the absence of exogenous EGF with constitutive autophosphorylation of EGFR, Shc phosphorylation, or STAT pathway activation.41 Furthermore, transgenic mice with inducible expression of EGFR exon 19 deletion mutants or the L858R mutation in type II pneumocytes developed lung adenocarcinoma after sustained EGFR mutant expression, confirming their oncogenic potential.42

Transformation by most EGFR mutants led to dramatic tumor regression by small molecular tyrosine kinase inhibitors, such as erlotinib and gefitinib.41, 42, 43 This is consistent with the results from clinical trials, demonstrating an underlying association between mutations in the EGFR tyrosine kinase domain and tyrosine kinase inhibitors responsive lung carcinomas. A retrospective review of studies using tyrosine kinase inhibitors treatment found an average response rate of 77% (range, 30–100%) in mutation-positive cases, with most series reporting response rates >60%, compared with 10% in mutation-negative cases (range, 0–33%).39 Furthermore, with tyrosine kinase inhibitors treatment, patients with EGFR mutation-positive tumors showed improved survival, with a median survival of up to 30 months compared with patients without EGFR mutations.39 In addition to EGFR mutation, lung cancer patients with high EGFR gene copy numbers in the tumor, as detected by fluorescent in situ hybridization (FISH), also showed a higher response rate and better survival in patients treated with erlotinib.44 However, EGFR FISH-positive status predicted worse survival in untreated patients.44 This suggests that patients EGFR copy-number change may also be benefit from tyrosine kinase inhibitors treatment.

Resistance to tyrosine kinase inhibitors over time has been reported in patients who had advanced lung carcinomas with EGFR exon 19 deletions or L858R mutations and who initially responded to the EGFR tyrosine kinase inhibitors treatment. A major mechanism of resistance to tyrosine kinase inhibitors is secondary resistance mutations. A threonine-to-methionine substitution at position 790 (T790M) in exon 20 of the EGFR tyrosine kinase domain has been observed in primary45 and secondary46, 47 tyrosine kinase inhibitors-resistant lung carcinomas. T790M mutation confers tyrosine kinase inhibitors resistance by either activating wild-type EGFR or increasing the ATP affinity of the oncogenic L858R mutation.48 Other secondary resistance mutations including D761Y, L747S, and T854A have been reported but seem to be rare.49 Another possible mechanism of tyrosine kinase inhibitors secondary resistance is that some other pathway might be activated to bypass the EGFR pathway, such as HER3-dependent activation of PI3K caused by MET amplification,50 which was detected in about 20% of lung carcinomas from patients who acquired resistance to gefitinib or erlotinib.50, 51 Furthermore, in half of the lung carcinomas with MET amplification, T790M mutation was also found.50, 51 An in vitro study of a multi-kinase inhibitor (XL880) with potent activity against MET reported inhibited growth of the NCI-H820 lung cancer cell line, which harbors MET amplification in addition to a drug-sensitive EGFR mutation and the drug-resistant T790M change.51 These results suggest that a MET inhibitor might be potentially used to treat lung cancer patients with acquired resistance to gefitinib or erlotinib. In addition, many strategies to inhibit downstream EGFR signaling are also being evaluated as potential targets for lung cancer therapy. Of these, the Ras-Raf-MEK-MAPK, PI3K-Akt-mTOR, and phospholipase C-PKC pathways have been most intensively studied alone or in combination with EGFR-targeting agents for lung cancers.52 However, these drugs are still undergoing phase I or II clinical trial to determine their toxicity and effectiveness. In combination with more than one target inhibitors might overcome the drug resistance of the single-target therapy and improve patient survival.

EGFRvIII, an in-frame deletion of exons 2–7 from the EGFR extracellular domain commonly observed in gliomas, was initially reported to be present in 16% of squamous cell carcinoma.53 Tissue-specific expression of EGFRvIII in the murine lung may lead to the EGFRvIII-dependent development of adenocarcinomas.54 However, a later analysis showed that this mutation was present in only a small fraction (5%) of squamous carcinomas, and was not found in 123 cases of human lung adenocarcinoma,54 suggesting a limited role of EGFRvIII in lung cancer.

Another EGFR family member, HER2, was expressed at a higher level in about 40–60% primary lung cancers including squamous cell carcinomas, adenocarcinomas, and large cell carcinomas compared with normal lungs55, 56 and was correlated with poor clinical prognostic indicators such as advanced clinical stage.56 ErbB2 kinase domain mutations, mostly in the form of in-frame insertions in exon 20 and rare mis-sense substitutions, were found in 1–10% of lung adenocarcinomas57, 58, 59 and mutually exclusive with mutations in EGFR and K-ras in the same tumor.57 Like EGFR mutations, HER2 mutations have similar associations with female sex, non-smoking status, and Asian ethnic background in patients with adenocarcinoma.57 Furthermore, insertion mutations in exon 20 of ErbB2 also enhanced the tyrosine kinase activity of ErbB2 and resulted in resistance to EGFR tyrosine kinase inhibitor-targeted therapy.60

The overexpression and hyperactivation of another two ErbB family members ErbB3 and ErbB4 have been studied previously.61, 62, 63 However, mutations in the ErbB3 and ErbB4 kinase domain are rare in lung cancers,64, 65, 66 suggesting that they might be of limited value for molecular-targeted therapy.

K-ras Mutation

K-ras is one of the three human ras oncogenes (K-ras, H-ras, and N-ras) that regulate cellular proliferation by playing a role in the transduction of signals across cellular membranes. Mutations of the ras proteins contribute to the development of cancer. The mutations in K-ras occurs in 4–30% of squamous cell carcinomas/adenocarcinomas,30, 67, 68, 69, 70, 71 with a relatively higher mutation frequency of 12–57% in adenocarcinomas, compared with 2–9% in squamous and other types of lung cancers.30, 67, 69 Mutant K-ras was associated with poor overall survival of lung cancer patients.35

The percentage of K-ras mutations present in lung adenocarcinomas from smokers is markedly higher than reported for non-smokers with lung adenocarcinoma (10–43 vs 0–8%), unlike EGFR tyrosine kinase domain mutations primarily occurring in lung adenocarcinomas of non-smokers.25, 69, 72, 73, 74, 75, 76 Most K-ras mutations in lung adenocarcinomas from smokers are characteristic G to T transversions, a typical mutation type induced by benzo(a)pyrene diolepoxide in vitro.77 This mutation occurs at the first two bases of codon 12 (normally GGT, which codes for glycine), resulting in the mutant codons TGT (cysteine) or GTT (valine),67, 72, 73, 74 whereas never-smokers were significantly more likely than former or current smokers to have a transition mutation (G-->A).78 In addition, the K-ras codon 12 mutation pattern in lung adenocarcinomas from smokers is distinct from that in gastrointestinal malignancies,79 suggesting that K-ras codon 12 may be a specific target of the mutagenic activity of tobacco smoke.

The EGFR tyrosine kinase domain and K-ras mutations are mutually exclusive, in that no tumors examined had mutations in both genes. The exclusivity indicates that, in smokers, tobacco carcinogens might specifically induce ras signaling pathways through mutations in K-ras, whereas in never-smokers the unidentified carcinogens might selectively target the EGFR pathway through mutations in EGFR. Because of the exclusivity of EGFR and K-ras mutations, researchers have studied the possible role of K-ras mutation as an indicator of EGFR tyrosine kinase inhibitors’ sensitivity. In a study of advanced lung adenocarcinomas treated with EGFR tyrosine kinase inhibitors, K-ras mutation with or without increased EGFR copy number suggested disease progression.80

LKB1 Mutation

The genetic alterations of the LKB1 gene (also known as STK11), which has been implicated in the regulation of multiple biological processes and functions as a tumor-suppressor gene,81 occurs more commonly in lung adenocarcinomas and large cell carcinomas than in squamous cell carcinomas and small cell carcinomas.82, 83, 84, 85 In a recent study, Matsumoto et al86 found that LKB1 mutations were significantly more frequent in lung cancer cell lines with K-ras mutations than in those without. More importantly, LKB1 mutations were found in 8% of male smokers with lung cancer but in none of the 64 female smokers and/or non-smokers,86 indicating that lung cancer LKB1 genetic alterations possibly correlate with smoking in men.

ALK Rearrangement

A small inversion within chromosome 2p, resulting in the formation of a fusion gene comprising portions of the echinoderm microtubule-associated protein-like 4 (EML4) gene and the anaplastic lymphoma kinase (ALK) gene, has recently been identified in lung squamous cell carcinomas/adenocarcinomas.87 Intronic sequences downstream of exons 13, 20, and 6 of EML4 are fused to intron 19 of ALK to generate variants 1, 2, and 3 of EML4–ALK, respectively.87, 88 This fusion tyrosine kinase showed transforming potential both in vivo and in vitro. In lung carcinomas with ALK gene rearrangements, ALK protein was expressed and phosphorylated, leading to the activation of multiple signaling pathways that contribute to cell survival and transformation, such as Akt and Erk1/2. The ALK inhibitor TAE684 could completely abolish these phosphorylations and suppress cell growth.89

In lung cancer patients, the EML4–ALK fusion was detected in 2–7% of tumors, with a higher percentage in adenocarcinomas and a complete absence in carcinomas of other types.87, 90, 91, 92, 93, 94, 95 In addition, ALK rearrangement was associated with younger age and never-smoker status, similar to findings for EGFR mutation.91, 93, 94 However, EML4–ALK fusion was mutually exclusive with EGFR or K-ras mutation.87, 92, 94, 96 Therefore, ALK inhibitors may provide a means to control lung adenocarcinomas in patients with genomic ALK rearrangements, for whom effective treatments are rarely available.

Epigenetic changes

Methylation

Aberrant methylation of cytosine at the promoter regions of genes is one of the major mechanisms of the downregulation or upregulation of genes in lung cancers. An increasing number of genes have been intensively investigated for their methylation status in lung cancers, including p16, RASSF1A, APC, RARβ-2, CDH1, CDH13, DAPK, MGMT, ASC/TMS1, FHIT, hSRBC, TSLC1, DAL-1, and PTEN.97 These genes are involved in cancer cell-cycle regulation, proliferation, apoptosis, cell adhesion, mobility, and DNA repair. Aberrant DNA methylation provides another mechanism for the inactivation of tumor-suppressor genes along with the genetic mechanisms to promote lung cancer occurrence and progression.

Methylation of genes has been shown to be associated with the smoking history of patients with lung cancer. In lung adenocarcinomas/squamous cell carcinomas, the frequency of p16, MGMT, RASSF1, MTHFR, and FHIT promoter methylation was significantly higher among smokers than never-smokers.98, 99, 100, 101 On the other hand, methylation in certain genes, such as RASSF2, TNFRSF10C, BHLHB5, and BOLL,102, 103 was higher in lung cancers from never-smokers than those from smokers, suggesting smoking may target specific genes for methylation.

Given the aberrant gene methylation observed in lung cancers, researchers have studied the roles of methylation in lung cancers for early detection, risk assessment, disease progression, and prognosis. The RASSF1A, APC, ESR1, ABCB1, MT1G, and HOXC9 genes were more frequently methylated in stage I lung adenocarcinomas/squamous cell carcinomas than the non-cancerous lesions,104 whereas the prevalence of hDAB2IP, H-cadherin, DAL-1, and FBN2 methylation was associated significantly with advanced stage of lung cancer,105, 106, 107 indicating that these genes might be involved in different phases of lung cancer progression.

Gene methylation has also been studied for the roles of genes as prognostic markers. Hypermethylation of RASSF1A, PTEN, DAPK, p16, Wif-1, CXCL12, DLEC1, MLH1, H-cadherin, APC, RUNX3, H-cadherin, SPARC, and DAL-1 was significantly associated with poor prognosis in patients with surgically resected lung adenocarcinomas/squamous cell carcinomas.101, 106, 108, 109, 110, 111, 112, 113 In addition, methylation of 14-3-3 sigma in pretreatment serum DNA was found to be an independent prognostic factor for survival in patients with lung adenocarcinoma, squamous cell carcinoma, and large cell carcinoma who were treated with platinum-based chemotherapy.114

Previous researchers have reported that methylation, in addition to occurring in tumor tissues, can also be detected in blood samples or in exfoliated material of the aero-digestive tract epithelium from lung cancer patients, such as the methylation of RASSF1A,115, 116 p16, H-cadherin,117 MAGE A1, and MAGE B2.118 Moreover, it became evident that methylation of certain genes is also detectable in cancer-free smokers. Therefore, gene methylation might be used as a surrogate marker for screening in high-risk populations. Belinsky et al119 found that methylation of p16, PAX5-b, MGMT, DAPK, GATA5, and RASSF1A in sputum collected within 18 months of lung cancer diagnosis was associated with a >50% increased lung cancer risk. Furthermore, the concomitant methylation of three or more of the above six genes was associated with a 6.5-fold increased risk, with values of 64% for both sensitivity and specificity.

DNA methylation is mediated by DNA methyltransferases (DNMTs), which transfer a methyl group to the 5′-position of cytosine. Three members of the DNMT family (DNMT1, DNMT2, DNMT3a, 3b) have been cloned in mammals. DNMT1, DNMT3a, and DNMT3b proteins are highly expressed in lung carcinoma, particularly in smokers.120 Furthermore, their overexpression correlated with hypermethylation in the tumor-suppressor gene promoters.120, 121 This is consistent with the observation in vivo that the tobacco-specific carcinogen, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), attenuated DNMT1 degradation and enhanced DNMT1 nuclear accumulation and hypermethylation of the promoters of tumor-suppressor genes. In lung carcinomas, DNMT1 and DNMT3b overexpression was significantly associated with poor prognosis.120, 121, 122

Histone Modification

In addition to gene promoter methylation, histone modification is another epigenetic control of gene transcription. Recent findings have confirmed that histones are not merely simple ‘DNA-packaging’ proteins but rather dynamic regulators of gene activity that undergo many post-translational chemical modifications, including acetylation, methylation, phosphorylation, ubiquitination, and sumoylation. The status of acetylation and methylation of specific lysine residues contained within the tails of nucleosomal core histones is crucial in regulating chromatin structure and gene expression.123

Changes in global levels of several histone modifications are predictive of the clinical outcome of lung cancers. By immunohistochemistry and a recursive partitioning analysis, 138 patients with stage I and II lung adenocarcinomas/squamous cell carcinomas/large cell carcinomas were classified into seven distinct prognostic groups based on TNM stage, histology, and histone modifications: histone 3 lysine 4 dimethylation (H3K4diMe), histone 3 lysine 9 acetylation (H3K9Ac), and histone 2A lysine 5 acetylation (H2AK5Ac).124 In a recent study, Seligson et al125 found that lower cellular levels of histone H3 lysine 4 dimethylation (H3K4me2) and H3K18 acetylation (H3K18ac) predict significantly poorer survival probabilities for lung cancer patients.

Histone deacetylation is mediated by histone deacetylases (HDACs), which work synergically to alter the chromatin condensation status and repress transcription with DNMTs and a family of methylated DNA-binding proteins.126 In general, high HDAC activity is associated with condensed, transcriptionally inactive chromatin. In addition to this epigenetic function, it is now recognized that certain HDACs also exhibit important cytoplasmatic function by controlling the acetylation status and function of numerous cytoplasmic proteins and transcription factors.127 Several authors have examined HDAC expression in lung cancer specimens. Sasaki et al128 reported that the mRNA and protein expressions of HDAC1 correlated with the progression of lung carcinomas, although there was no difference in mRNA expression between tumor and adjacent non-tumor lung tissue. However, Bartling et al129 found that HDAC3 was upregulated in squamous lung cancers compared with non-tumorous lung tissues. In a group of 72 patients with lung adenocarcinoma/squamous cell carcinoma/large cell carcinoma, mRNA expression levels of HDAC class I (HDACs 1–3 and 8) and class II (HDACs 4–7, 9, and 10) genes in cancer tissues were measured using real-time RT–PCR. Reduced expression of each class II HDAC gene was significantly associated with poor prognosis with HDAC10 as the strongest predictor for poor prognosis.130 These results suggested that HDAC might be involved in lung cancer occurrence, progression, and prognosis and that inhibition of HDAC activity might be a possible target for lung cancer treatment. Previous studies found that HDAC inhibitor suppressed the PI3K/Akt131 and Src/Raf/MEK/ERK1/2132 signaling pathways, resulting in the downregulation of the anti-apoptotic proteins Bcl-2 and Bcl-xL, upregulation of the pro-apoptotic protein Bax, and the induction of time-dependent apoptosis in both adenocarcinoma133 and small cell carcinoma cells.133, 134 Coincident with inhibition of ERK1/2 and PI3K/AKT survival pathways, the HDAC inhibitor FK228 enhanced JNK and p38MAPK signaling,132 whereas an SIRT1 inhibitor, Sirtinol, impaired activation of Ras/MAPK pathways in response to EGF and insulin-like growth factor-I.135 Furthermore, another HDAC inhibitor, trichostatin A, suppressed the levels of COX-2 mRNA and protein expression, which were correlated with an inhibition in prostaglandin E2 synthesis in lung adenocarcinoma cells.133 However, HDAC inhibitors do not always function in inducing apoptosis of lung cancer cells despite their ability to effectively inhibit deacetylase activity. HDAC inhibitors could stimulate NF-κB, resulting in expression of NF-κB-dependent genes such as IL-8, Bcl-XL, and MMP-9, which cause failure to induce apoptosis in lung carcinoma cells.136 Therefore, members of the HDAC family have a potential role as lung cancer treatment targets.

MicroRNAs

miRNAs are a family of small RNA molecules (∼22 nt) that regulate specific gene expression post-transcriptionally. Abnormal expression of miRNAs are believed to be involved in the initiation and progression of human cancer.137 Such miRNAs as miR-126, miR-31, miR-519c, Let-7a, miR-133B, miR-15a, miR-16, and miR-183 have been found to regulate lung cancer cell proliferation, migration and invasion by targeting specific molecules, including Crk, EGFL7, VEGF, LATS2, PPP2R2A, HIF-1α, NIRF, MCL-1, BCL2L2, cyclins D1, D2 and E1, and Ezrin.138, 139, 140, 141, 142, 143, 144, 145, 146

Abnormally expressed miRNAs may have potential as markers in lung cancer diagnosis, treatment response, and prognosis. Has-miR-205, which suppresses epithelial-to-mesenchymal transition, has been identified as a highly specific marker for squamous carcinoma of the lung.147, 148 A well-studied miRNA in lung cancer is miR-21. In a group of 48 lung adenocarcinomas/squamous cell carcinomas, overexpression of mature miR-21 was found in 52% of cases compared with their corresponding non-cancerous tissues. During the follow-up period, mature miR-21 was upregulated in 16 (55%) of 29 patients who had relapse and in 15 (65%) of 23 patients who died. Mature miR-21 overexpression correlated with overall patient survival (P=0.027), suggesting that overexpression of mature miR-21 could be an independent negative prognostic factor for overall survival in patients with lung adenocarcinoma/squamous cell carcinoma patients.149 In a recent study, miR-21 expression in sputum specimens was significantly higher in patients with lung adenocarcinoma/squamous cell carcinoma (76.32±9.79) than in cancer-free individuals (62.24±3.82) (P<0.0001). Furthermore, detection of miR-21 expression had 70% sensitivity and 100% specificity for the diagnosis of lung cancer, compared with 48% sensitivity and 100% specificity for sputum cytology, suggesting that measurement of altered miRNA expression in sputum could be a useful non-invasive approach for the diagnosis of lung cancer.150 Studies also showed that lower let-7151 and miR-34152 expression predicted short survival or a high probability of relapse in patients with lung adenocarcinomas/squamous cell carcinomas/large cell carcinomas, whereas high miR-146b153 and miR155154 expression correlated with poor survival. Through analysis of miRNA expression in patients with lung adenocarcinoma/squamous cell carcinoma patients, a five-miRNA signature including miR-221, let-7a, miR-137, miR-372, and miR-182* was identified and validated as an independent predictor of cancer relapse and survival.155 In addition, expression levels of miR-486, miR-30d, miR-1, and miR-499 in serum could be used to predict survival for patients with lung adenocarcinoma/squamous cell carcinoma.156

Compared with gene expression-based predictive classifier, certain miRNA signatures were more informative in predicting survival. For example, miR-146b alone was found to have a predictive accuracy for prognosis in ∼78% of patients with lung squamous cell carcinoma,153 better than the overall predictive accuracy of 68% for a 50-gene signature.157 This is probably due to miRNAs being upstream regulators of gene expression with hundreds of downstream targets, suggesting that miRNAs may have more powerful prediction abilities than their target genes.

Conclusions

The development of lung cancer is a complex process, involving a series of genetic and epigenetic changes. Smoking directly induces the gene mutation, while methylation, HDACs, and miRNAs might indirectly affect carcinogenesis via modulating gene mutation and expression. A better understanding of how these factors participate in lung cancer development and progression would provide us with powerful tools for lung cancer prevention, diagnosis, and treatment by identifying practical molecular markers correlated with clinical parameters. Furthermore, understanding the molecular characteristics of lung cancers would aid in targeted therapy development, as in the example of EGFR inhibitors including monoclonal antibodies (eg, cetuximab) and small-molecule tyrosine kinase inhibitors (eg, erlotinib, gefitinib), which were developed and applied clinically based on our knowledge of EGFR mutation in lung cancer patients.158 To date, clinical utilization has been approved for very few biomarkers. Tremendous efforts are required to translate the findings generated from basic research to clinical application in lung cancers. Further validation of these potential biomarkers should involve careful selection of the markers, larger sample sizes with long-term follow-up, and multi-center studies. In addition, molecular testing for multiple biomarkers may generate higher sensitivity or specificity for clinical application than a single marker. Similarly, multi-target therapies that interfere with more than one pathway might be more effective than single-target agents in treating the deadly disease of lung cancer.