Neural Stem Cells in Brain Tumor Development

Malignant astrocytomas are highly invasive and incurable brain tumors. Mouse models that genetically resemble the human disease are valuable tools in understanding the pathogenesis of these malignancies. We previously reported mouse models based on conditional inactivation of the human astrocytoma-relevant tumor suppressors Nf1, p53 and Pten. Through somatic loss of heterozygosity, these mice develop varying grades of astrocytic malignancy with 100% penetrance. Studies on our tumor suppressor mouse models indicated a central role for neural stem cells and stem cell-like cancer cells in malignant astrocytoma formation. Using stereotactic viral cre-mediated approach, we demonstrate that targeting of tumor suppressor inactivating mutations in the subventricular zone (SVZ) where neural stem and progenitor cells reside is both necessary and sufficient to induce astrocytoma formation. We also show evidence of spontaneous differentiation and infiltration of these cancer-initiating cells in situ during tumor development. These studies have so far shown that neural stem cells or its progeny can give rise to astrocytomas. Neural stem cells, which have unlimited self-renewal potential, produce transit amplifying cells, or progenitor cells, which undergo limited mitoses before differentiating into more mature cell types. By genetically targeting transit amplifying cells using the Ascl1-creERT2 transgenic mouse, we show that tumor suppressor inactivation in the progenitor compartment alone induces malignant astrocytoma formation. Defects in proliferation, differentiation, and migration are likewise found several months prior to advanced disease. This establishes both neural stem and progenitor cells as cells of origin of malignant astrocytomas in our tumor suppressor mouse models. In another study, we isolated and characterized a population of stem cell-like cancer cells from murine astrocytomas that are enriched for tumor cells compared to primary tumor tissue, exhibit aberrant stem cell properties, and are tumorigenic in vivo. We demonstrate resistance to a known chemotherapeutic agent and the migratory capacity of these cells. We also investigated the mechanisms involved in astrocytoma progression and maintenance by gene expression analysis. Genomic profiling of tumor-derived neurosphere-forming cells from conditional astrocytoma mouse models show prominent dysregulation of genes involved in neurodevelopmental processes and transcriptional regulation, particularly the hox transcription factors, in high-grade astrocytomas.
Taken together, we have demonstrated that neural stem and progenitor cells are the origins of malignant astrocytoma in tumor suppressor mouse models. We have established a system by which molecular mechanisms of tumor development can be further investigated and performed genomic profiling of tumor-derived neurosphere-forming cells, suggesting a possible role for homeobox transcription factors in malignant astrocytoma formation. These mouse models thus represent powerful tools in understanding various aspects of cancer development that otherwise cannot be explored in humans. Further studies will provide a better understanding of the biology of these tumors and will hopefully pave the way for more effective therapeutic approaches for these devastating diseases.