Modelling Glioblastoma Angiogenesis in Drosophila larvae

Abstract

Glioblastoma multiforme (GBM) is the most common malignant type of cancer of the Central Nervous System (CNS) and is characterized as highly angiogenic. Poor prognosis combined with short lifespan after diagnosis (⁓18 months) and resistance to the most invasive treatments highlight the need to understand the disease pathology at the molecular level. Angiogenesis plays an important role on tumor cell survival and malignancy, but the precise mechanisms that control the interactions between glial cells and blood vessels remain unknown. Drosophila melanogaster has emerged as an excellent model for GBM studies because glia-specific overexpression of EGFR/Ras and PI3K-Akt, the major pathways upregulated in human GBMs, causes low- and high-grade gliomas. Our goal is to use Drosophila larvae to understand the mechanisms governing GBM tracheogenesis, the functional and molecular equivalent of angiogenesis. In this study we used immunofluorescence and confocal microscopy to characterize glia-tracheae interactions in established Drosophila GBM models. Specifically, we used the Gal4 – UAS system to overexpress the RasV12 oncogene, the constitutively active catalytic subunit of PI3K, Dp110CAAX, alone or in combination, in glial cells of Drosophila larvae. We found that the mitotic index in Dp110CAAX overexpressing brains was moderately increased, whereas brains overexpressing RasV12 alone or in combination with Dp110CAAX exhibited strikingly increased mitosis compared to control brains. The levels of brain mitosis correlated with increased central brain size. In addition, brains overexpressing RasV12 alone had significantly increased number of apoptotic cells compared to controls, Dp110CAAX, and Dp110CAAX; RasV12 brains. To understand glia-trachea interactions, we needed to develop a model to visualize the trachea in GBM-bearing larvae. Since we were using the Gal4 – UAS system to overexpress the tumor-causing genes, we needed a different method to highlight the brain tracheal cells. We tested available trachea-specific QF-QUAS and LexA-LexAop drivers, but they did not allow brain trachea visualization. Therefore, we reverted to using the QF -QUAS system to overexpress RasV12 in glial cells and the Gal4-UAS system to visualize the trachea, a combination that proved successful. The mitotic activity of this model agreed with our results from the brains overexpressing oncogenes using the Gal4-UAS system. Finally, our 3D imaging suggests increased tracheogenesis in brains overexpressing RasV12.