Current smoking-specific gene expression signature in bronchial epithelium is enhanced in squamous cell lung cancer

Cigarette smoking is the main risk factor for the development of squamous cell lung carcinoma (SCC). However, the smoking-related molecular changes in SCC have not been studied. We wanted to identify genes in both histologically normal bronchial epithelium and SCC samples that are differentially expressed between current and ex-smokers. In addition, to analyze the levels of the smoking-related genes identified in normal bronchial epithelium with the levels in SCC. Gene expression profiles were generated using Agilent whole human genome microarrays in 28 laser microdissected normal bronchus epithelial samples and in 35 laser microdissected SCC samples of current and ex-smokers. Levels of 246 genes, mainly related to oxidative stress response, were significantly different between normal bronchial epithelium of current and ex-smokers. No significant differences were associated with the smoking status in SCC samples. Analysis of the 246 smoking specific genes in SSC from current and ex-smokers also revealed no differences. As a next step, we compared the levels of the smoking-specific gene signature between normal bronchial epithelium from ex- and current smokers to all 34 SCC samples. Twenty-two percent of the upregulated genes are further upregulated in SCC as compared to current smokers. Expression of the downregulated genes was even further downregulated for 79% of the genes in SCC as compared to bronchus epithelium of current smokers. The downregulated genes included several tumour suppressor genes. This study shows that genes upregulated in normal bronchial epithelium of current smokers are expressed at similar levels in SCC samples, while levels of downregulated genes were significantly further reduced in SCC. This indicates that these downregulated genes play a role in SCC oncogenesis. PATIENTS AND TISSUE SPECIMENS For each patient, the number of packyears and current smoking status was assessed. Individuals were divided in ex-smokers, who quitted smoking at least two years before surgery, or current smokers, who continued smoking until surgery. We collected cartilaginous bronchial tissues (diameter >2mm) from 28 patients during thoracotomy for a lung tumour (NSCLC) (n=21) or during lung transplantation (n=7). Tumour samples were obtained from 35 patients with primary SCC. For two cases, the bronchus and tumour sample were derived from the same patient. We excluded patients using inhaled, intravenous or oral steroids before surgery. Patients who were treated with chemotherapy and patients with alpha-1-antitrypsin deficiency were also excluded. All tissue samples were snap frozen and stored at -80ºC. The study protocol was consistent with national ethical and professional guidelines (“Code of Good Conduct; Dutch Federation of Biomedical Scientific Societies”). LASER DISSECTION MICROSCOPY Histologically normal bronchial epithelial cells (consisting of basal, ciliated and goblet cells) were microdissected from central bronchus sections and vital tumour cells from SCC sections. An area of approximately 20 - 25 x106 μm2 was collected from 8 μm frozen sections using P.A.L.M. Microlaser Technology system (P.A.L.M., Bernried, Germany) in lysis buffer (Macherey-Nagel, Düren, Germany). RNA ISOLATION Total RNA was isolated and purified from the microdissected cells with a Nucleospin RNA II kit (Macherey-Nagel). The quantity of DNA-free total RNA was measured using a Nanodrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE). RNA quality was assessed by the presence and ratio of 18S and 28S rRNA bands combined with a low baseline, monitored with RNA 6000 PicoChip (Agilent, Palo Alto, CA) on the 2100 bioanalyzer (Agilent). MICROARRAY APPROACH Agilent Whole Human Genome Oligo Microarrays, design number 012391 (GPL1708) (Agilent, Palo Alto, CA) were used. Briefly, 150 ng of total RNA was amplified for one round using MessageAmp II (Ambion, Austin, USA). mRNA amplification, ULS labeling, Cy-dye coupling, hybridisation and data extraction were performed at ServiceXS (Leiden, the Netherlands) according to manufacturer’s instructions. A dye-swap was included and samples were hybridised using a randomized approach (scheme available as supplementary table in manuscript), based on the non-competitive conditions of a two-colour array. T22 failed quality test