"Glioblastoma multiforme (GBM) is the most malignant type of brain tumor in adults with a median survival of one year. The hallmarks of GBM are invasion and neo-angiogenesis. Recently, gene expression studies have established that overexpression of a “mesenchymal” gene signature co-segregate with the poor-prognosis group of glioma patients. We have revealed that two transcription factors, Stat3 and C/EBPβ behave as master regulators of the mesenchymal gene signature. Overexpression of these two transcription factors reprograms neural stem cells along the aberrant mesenchymal lineage, whereas gene silencing in glioma cancer initiating cells (CSC) reduce tumor growth and aggressiveness. Recently, the Cancer Genome Atlas Network has provided information on genomic abnormalities that have allowed the expansion of previous glioblastoma classification studies by associating known subtypes with specific genomic alterations. In particular, EGFR amplification and NF1 and PTEN loss have been linked with the mesenchymal tumor subtype. In this proposal, we would like to identify the genetic events that cause the aberrant expression of Stat3 and C/EBPβ and mesenchymal transformation of gliomas. We have hypothesized that this might occurs by (1) gene copy number variation, (2) mutation in the gene promoter, (3) gene mutation of an upstream transcriptional regulator or, alternatively (4) aberrant DNA methylation of the gene promoter. Will analyze Stat3 and C/EBPβ gene promoter as well as the gene status in GBM samples by performing copy number variation studies and promoter sequencing. The status of CpG methylation at the promoter will be evaluated by pyrosequencing. Finally, we will test the effect of EGFR amplification, as well as of NF1 and PTEN deletion. We will mimic gene deletion or amplification by silencing or overexpressing these genes in glioma CSC and evaluating the effects on cell invasion and tumor formation, both in vitro and in vivo."