Using novel models of glioma for cancer discovery science
Brennan, Paul Martin
MetadataShow full item record
The prognosis for patients diagnosed with glioma has changed little over the past two decades. Many therapies that appeared promising in preclinical studies have been unsuccessful in the clinic. In an attempt to address this problem I developed a method for the efficient derivation of glioma primary cultures from fresh human brain tumours. These cultures are enriched for putative cancer stem-like cells that are thought to be responsible for glioma initiation, therapy resistance and recurrence. This mechanism of tumour development is a departure from the traditional multistep model of cancer. It is hoped that preclinical models incorporating glioma stem-like cells will more effectively recapitulate the biology of human disease and so better predict the likely clinical efficacy of inhibitor compounds tested in vitro and in the preclinical setting. In contrast to the majority of the existing literature, I identified two distinct tumourderived glioma stem-like cell phenotypes in my primary cultures that I have called ‘branched’ and ‘flat.’ The branched cells had similarities to the radial glia-like cells previously described in glioma stem-like cultures. In contrast, the flat cells had mesenchymal-like features. I discuss the implications of these observations for understanding glioma cell biology. I describe the development of high content phenotypic assays that incorporate these putative glioma stem-like cells. I screened inhibitor compounds of the PI3 kinase pathway, which is important in glioma cell behaviour, and identified that PIK75, a drug that targets the p110α catalytic subunit of PI3 kinase, inhibited growth of all the primary cells tested. I examined PIK75 activity in some detail. In vivo models of glioma are used to validate the findings of in vitro compound screening, so I describe my attempt to develop a novel genetically engineered mouse model designed to initiate glioma formation from the glioma stem-like cell. Surprisingly, these mice actually developed malignant peripheral nerve sheath tumours and that gave me a novel insight into the pathogenesis of this rare disease. This also informed future work on my long-term goal of generating a genetic model of glioma that recapitulates human disease.