Preview

PDF, English Download (2MB) | Terms of use

Abstract

Environmental exposure such as tobacco smoke is the principal cause of most lung cancer cases worldwide. However, only a small proportion of heavy smokers develop lung cancer which suggests that other factors such as genetic and/or epigenetic interindividual variations may be responsible for individual disease susceptibility. The overall aim of this study was to determine germline copy number variations (CNVs) associated with early-onset lung cancer risk and to further investigate the genetic and epigenetic interplay of microRNAs (miRNAs) and genes located in two candidate CNVs on 8q24.3 and 11p15.5 in lung cancer. A genome wide association study (GWA) had been performed using the Illumina Infinium platform Human Hap550 BeadChip on 492 early-onset lung cancer cases and 487 population based controls. Two computational CNV detection algorithms, QuantiSNP and PennCNV, were applied to this existing data set to identify CNVs and the overlapping CNVs between the two algorithms were further analyzed for association with the disease. Ten CNVs were significantly associated with early-onset lung cancer. Two CNVs were selected for strength of association and for containing miRNAs and genes likely to be relevant for lung cancer. To assess their functional relevance in non-small cell lung carcinoma (NSCLC), qPCR based expression analysis and quantitative methylation analysis using the MassCLEAVETM assay of genes and miRNAs in these regions were performed on NCSLC and matched normal lung tissue. The expression analysis showed that miR-661 on 8q24.3 was significantly upregulated in lung tumor compared to normal. The putative miR-661 promoter was hypomethylated in tumor tissue and revealed a significant negative correlation with expression in tumor. Additionally, the loss of methylation at these sites was significantly associated with worse outcome independent from stage, histology and gender. The most significant changes in the gain CNV region on 11p15.5 were seen for miR-210 and Plakophilin 3 (PKP3) which both were significantly upregulated in NSCLC. Promoter hypomethylation at the transcription start site of PKP3 was inversely correlated with expression in NSCLC, suggesting that methylation regulates the PKP3 expression. For further functional analysis of the two miRNAs, predicted targets were identified in silico and 3´UTR luciferase reporter assays for the predicted targets and expression analysis after ectopic overexpression in A549, H1299 and H1703 lung cancer cell lines were carried out to determine whether a direct link between the miRNAs and the targets could be shown. The results showed that mitogen associated protein 3 kinase 3 (MAP3K3) and Cadherin1 (CDH1) are direct targets of miR-661, suggesting that miR-661 has oncogenic properties in lung cancer. Furthermore, miR-210 was shown to target the tumor suppressor gene Runt related transcription factor 3 (RUNX3), a transcription factor known to be involved in lung development and to be a crucial regulator of cell proliferation. The results from this study suggest that CNV analysis of GWAs data for lung cancer risk can point to functionally important regions in the genome that are deregulated in NSCLC and may contribute to lung tumorigenesis. Further investigation of the relevance of these CNVs to early-onset lung cancer risk is needed to confirm our suggested finding of two risk markers. Furthermore, additional analyses on the functional role of miR-661 in lung cancer are desirable to elucidate to what extent this miRNA contributes to tumorigenesis. Taken together, this study provides evidence that interplay between genetic variations and epigenetic deregulation plays a pivotal role in NSCLC pathogenesis.