INTEGRATED ANALYSIS OF DNA COPY NUMBER AND GENE EXPRESSION DATA IN LUNG CANCER MODELS OF RESISTANCE TO TARGETED THERAPY

Motivation: Tyrosine Kinase inhibitors (TKIs) constitute the most promising frontier of the last decade in the field of cancer treatment. However, the development of resistance mechanisms often makes the tumour insensitive to TKI-targeted therapy. Identification of the factors responsible for the emergence of resistance remains a major clinical challenge for cancer treatment. Non Small Cell Lung Cancer (NSCLC) with mutations activating the Epidermal Growth Factor Receptor (EGFR) provides a suitable model for studying TKI resistance mechanisms. The majority of patients, after an initial positive response to EGFR-TKI drugs (e.g. Erlotinib), develops resistance to treatment for the onset of EGFR secondary mutations and/or the activation of alternative signaling pathways. The present work aims to identify genes and signaling pathways involved in the development of TKI-resistance. To this end, we developed cellular models of NSCLC and designed cellular, molecular and bioinformatics analyses to correlate genotypes to phenotypes. Methods: We generated six cell lines resistant to EGFR-TKI treatment by exposing NSCLC cell lines, harboring EGFR activating mutations in Exon 19, to Erlotinib (ERL) for 5 months. The selected resistant lines were then characterized at the cellular and molecular level to identify phenotypes associated with ERL-resistance. Briefly, immunofluorescence and confocal microscopy analysis as well as cell viability, wound healing and anchorage-independent cell growth assays were used to study cell morphology, cell growth, cell migration, and tumorigenic features. Sequence analysis, quantitative PCR and Western blot were used to characterize mutations, gene copy number variations (CNVs) and gene expression profiles of selected receptor tyrosine kinases (namely EGFR, KRAS, MET, AXL, AKT) that have been previously reported to be mutated or dysregulated in NSCLC. Next, CNVs and gene expression data were assayed for both parental and resistant derivative cell lines at genome-scale level by using microarray technology. We used R (cran.r-project.org/) and Bioconductor (www.bioconductor.org/) for data processing and analysis. The IGV genome browser (www.broadinstitute.org/igv/) and Gorilla web tool (cbl-gorilla.cs.technion.ac.il/) will be used, respectively, for interactive exploration and integrative analysis of the different data levels and for gene functional enrichment analysis. Results: Preliminary cellular and molecular characterization of the NSCLC cell lines highlighted several features associated to an ERL-resistant phenotype, including altered cell migration and cell adhesion, Epithelial-to-Mesenchimal-Transition (EMT) features and metabolic dysregulation. In-depth analysis is still on-going. Moreover, comparison of array data between resistant and sensitive cells array data identified several chromosomal regions that undergone amplification or deletion, as well as lists of thousands of genes aberrantly expressed in selected drug-resistant cell lines, and hundreds of genes aberrantly expressed in all resistant cell lines. Interestingly, preliminary functional enrichment analysis of genes, found consistently altered in all resistant cell lines, indicates their involvement in key biological processes. Further bioinformatics studies are in progress.