Genomic and proteomic analysis of normal and diseased tissues have yielded an abundance of molecular information for diagnostic and potential therapeutic targets. Changing the target of analysis from poorly accessible cells within tissues to easily accessible vascular endothelium has theoretical advantages in tissue-specific targeting. In this study, we sought to map a large-scale proteome of microvascular endothelium in human non-small cell lung cancer (NSCLC) and normal lung tissues, and identify lung cancer-related endothelial cell (EC)-selective proteins. Endothelial cells were isolated within NSCLC tissues and adjacent-normal lung tissue of lung cancer patients by using CD31-immunomagnetic beads. The complex proteins from the ECs were separated by one-dimensional gel electrophoresis, and the proteins in each gel band were digested by trypsin. Peptides were separated by online reverse-phase liquid-chromatography and analyzed by electrospray ionization (ESI) ion trap tandem mass spectrometry. Approximately 600-1000 proteins were identified in each individual sample. Five patient cases of paired individual data, extracted from the protein identification data sets of both normal- and cancer-derived ECs, were analyzed by subtractive proteomics. An average of 300 proteins was specifically identified from each lung cancer-derived EC isolate, compared to normal lung-derived ECs. With the use of several comparative analyses, we identified among those 300 proteins, 16 common candidate proteins that were detected in at least 3 of 5 cases specific to lung cancer-derived ECs. Proteins selectively identified in cancer-derived ECs, including coatomer protein complex, subunit gamma (COPG), and peroxiredoxin 4 (PRDX4), were validated by Western blot analysis. In an additional experiment in which 16 cancer samples were analyzed by immunohistochemistry, PRDX4, thymopoietin (TMPO), and COPG were confirmed to be abundantly expressed in lung cancer-derived ECs and in cancerous lung cells. Further ongoing analysis of these 16 candidate proteins will determine their potential applicability to NSCLC-specific diagnosis and therapeutics.

These results suggest that miR-149 and miR-196a may be involved in the pathogenesis of NSCLC, and that rs2292832 and rs11614913 can be used as prognostic markers for patients with surgically resected early-stage NSCLC.

Telomere shortening leads to genomic instability that drives oncogenesis through the activation of telomerase and the generation of other mutations necessary for tumor progression. This study was conducted to determine the impact of telomere shortening on the survival of patients with early stage non-small cell lung cancer (NSCLC). Relative telomere length in tumor tissues was measured by quantitative polymerase chain reaction in 164 patients with surgically resected NSCLC. The association between telomere length and overall survival (OS) and disease-free survival (DFS) was analyzed. When the patients were categorized into quartiles based on telomere length, those patients with the 1st quartile (shortest) of telomere length had a significantly worse OS and DFS compared to patients with the 2nd to the 4th quartiles of telomere length (adjusted hazard ratio for OS = 2.67, 95% confidence interval = 1.50-4.75, P = 0.001; and adjusted hazard ratio for DFS = 1.92, 95% confidence interval = 1.17-3.14, P = 0.01). An association between telomere length and survival outcome was more pronounced in squamous cell carcinomas than adenocarcinomas (P-value of test for homogeneity for OS and DFS = 0.05 and 0.02, respectively). Telomere length of tumor tissues is an independent prognostic factor in patients with surgically resected early stage NSCLC.