DEFINITION OF BIOLOGICAL RESPONSES THROUGH THE ANALYSIS OF GENE EXPRESSION PROFILES

The aim of this PhD project was the development of a pipeline for the analysis of expression data and a set of of different strategies to extract biological informations from micrarray experiments. The computational pipeline for processing raw microarray data (images) to define gene expression levels, to provide experiment quality assessment and significativity statistical tests, was implemented in R, using mostly Bioconductor packages. The first fase had as purpose the determination of the gene function combining experiments of silecing with the gene expression analysis. Caspase-2 is a member of a cystein-protease family that carry out important roles in the apoptosis and in the inflammation. Altough it is highly conserved from the evolutionary point of view, in the literature several contradictory results are found. Being expressed at high level during the neurological development and with a strong involvement in the apoptotic processes in the adult central nervous system, we decided to proceed with the silecing of the gene that codifies for this enzyme using glioblastoma cells, a very aggressive cerebral tumor. The comparative analysis of expression profiles of silenced cells respect to the control ones, highlighted the relation between CASP2 and genes involved in the cholesterol metabolism. Previuos studies have suggested for this enzime a role in the control of intracellular level of this metabolite. Therefor, we decided to use data stored in public databases in order to to extend the investigation, including all the other caspases and all the genes in same way connected to cholesterol. After we had obtained the data related to several different experiments, we went ahead with the computation of the correlation between expression levels and, then, based of these values, with the clustring analysis in order to see which among the caspases has the same corralational profile. After that, the analysis was expanded to normal brain and liver tissues, in order to know whether the situation observed in the patological condition is unique or if it can be overlayed to that present in normal tissues. In the second phase, I performed an analysis of expression data with a completely different purpose. The aim of this project was the definition of the signaling pathways and of the resistence mechanisms induced by the treatment of cancer cells obtained from patients affected by cronic lymphocytic leukemia and treated with a new category of ubiquitin proteasome system (UPS) inhibitors. Through the comparison of trascriptional profiles before and after the treatment, many genes connected with the drug action at cellular level, whose expression was altered by the UPS inhibitor, were identified. Furthermore, considering the difference in terms of responsiveness of the analized patients, we could determine some genes responsible of the different efficacy of the farmacological treatment.