LAUSR

Glioblastoma multiforme (GBM) represent the most devastating form of high-grade glioma (HGG) affecting adults and children. Despite a multi-therapeutic approach consisting in surgery, radio- and chemo-therapy the prognosis remains poor. Several models such as xenograft, animal models and recently organoids, have been developed in order to investigate the physiopathology of GBM. Although several mouse models (where either gain- or loss-of function of genes/pathway found altered in patients induce tumour formation) have been generated many aspects of how the tumour is formed, evolves, infiltrates and recurs after treatments remain unclear. Another challenge is the creation of proper GBM models showing intratumor heterogeneity found in the patient tumour but missing in the animal model generated so far. To overcome the lack of that human tumour characteristic, we decided to generate new model of GBM using a gene screen-based approach in mice. Firstly, we analysed the genes found amplified or mutated in GBM patients; secondly, we over-expressed the candidate genes (combination of multiple genes found to be mutated or amplified in specific GBM patients) in the subventricular zone (SVZ) of P2 mouse brain. Thirdly, we analysed the formation of tumour after 2 months. We over-expressed roughly 50 combinations in newborn mice and found that only 3 successfully led to the formation of lesions positive for proliferation and brain tumour marker (i.e. GFAP). We are currently characterising the tumours by DNA methylation analysis and RNA-seq (for further confirmation of the cancer subtype and intratumor heterogeneity) or investigating the cell of origin of the tumour. Moreover, we are testing the tumorigenicity of specific in GBM-amplified/mutated gene combinations in hiPSC-derived cerebal organoids. Our data suggest that a gene screen-based approach can be used for quickly and easily assaying the tumorigenicity of genes found amplified/mutated in GBM patients as well as the biology behind such complex process.