Defining the stem cell origins of brain tumours containing non-neural cells

A striking feature of some central nervous system (CNS) tumours is the presence of cell types indicative of mixed neural/glial and mesodermal/mesenchymal (e.g. skeletal/smooth muscle, cartilage, bone) differentiation. These "neuromesodermal" (NM) tumours affect both young children and adults and they are highly aggressive being linked to a poor prognosis. For example, a considerable fraction of gliomas, the most common adult brain tumours, display mesenchymal features with elements of bone, cartilage and muscle differentiation and some of them contain a distinct muscular sarcomatous component (gliosarcomas). Defining the cell entity driving the formation of these tumours will be pivotal for the design of effective therapeutic interventions against these malignancies. The proposed PhD project aims to examine the hypothesis that NM tumours arise from stem cells which resemble multipotent embryonic precursors, namely neuromesodermal progenitors (NMPs) or neural crest (NC) precursors. During early embryonic development NMPs give rise to spinal cord neurons as well as bone, cartilage and skeletal/smooth muscle whereas NC cells generate peripheral neurons, melanocytes and mesenchymal cell types such as smooth muscle and cartilage. We speculate that NM tumourigenesis is the result of either a mutation-driven conversion of NMPs/NC progenitors into cancer-initiating cells during embryogenesis or a "reprogramming in situ" event where corruption of the genetic/epigenetic machinery within neural/glial cells results into their conversion into NMP/NC-like cells which in turn trigger tumorigenesis within a favorable niche.

Research plan
Current conventional methods for establishing brain tumour cell lines are designed to select for neural stem cell (NSC)-like cells reflecting the hypothesis that the cell-of-origin of all brain tumours is a transformed neural progenitor. Thus adherent glioma stem cell lines have been successfully established using the same conditions as those employed for the propagation of neural stem cells. However, this approach cannot be employed for the isolation of bipotent NM cells resembling NMPs/NC cells. We have recently defined the optimal culture conditions for the induction of NMPs and NC progenitors from human pluripotent stem cells (hPSCs) in vitro. Adult brain tumours will be dissociated and equal numbers of the resulting cells will be cultured in parallel in NMP, NC and neural stem cell (NSC) conditions (positive control). Successful attachment, proliferation and subsequent passaging in NMP or NC culture media would mark the presence of NM bipotent stem cells in the starting tumour. We have already obtained promising preliminary data showing that glioma-derived cells can proliferate in NMP induction conditions. The presence of bipotent cells in the resulting cell lines will be confirmed by monitoring the expression of NMP and NC markers using immunofluorescence/qPCR and in vitro differentiation assays. RNA-seq analysis will be carried out comparing the transcriptomes of NM tumour cell lines to hPSC-derived NMPs and NC cells as well as established human NSC lines to define candidate drivers of oncogenic transformation. Assessment of tumourigenic potential by xenografting/stereotactic injection into NOD-SCID mice will also be performed. If time permits we will also carry out small molecule screens aiming to identify compounds exerting cytotoxic/cytostatic effects on the NM tumour cell lines.