Medulloblastoma (MB) is a common and frequently lethal tumor of childhood. While a number of genetically engineered mouse models for MB have been generated, the genetic events that govern MB in children remain uncertain, in part, due to a lack of physiologically relevant humanized mouse models. Since MB is thought to be derived from neural stem cells (NSC) and granule neuron precursor cells (GNPC), this would suggest that a representative model of human MB requires the orthotopic transplantation of transformed NSC and/or GNPC in mice. The oncogene MYCN is a strong candidate to initiate transformation of these cells, as MYCN drives proliferation in cerebellar development and has been shown to promote MB formation in preneoplastic cells. Consistent with this observation, we have found that a stabilized mutant of MYCN (T58A) can drive MB formation in mouse cerebellar stem cells. In this fellowship application, we propose to generate humanized mouse models of MB by first manipulating human induced pluripotent stem cells (iPSC) to generate iPSC predisposed to form MB (via loss of expression for candidate susceptibility genes including PTCH), then subsequently forcing differentiation along NSC and GNPC lineages, followed by transducing MYCN (WT and T58A) and transplanting the resulting cells orthotopically into immunocompromised mice. Predisposing mutations will be introduced in iPSC by the recently described approach of Transcription Activator-Like Effector Nucleases (TALEN) to efficiently edit the genome and mimic a true loss of expression for the endogenous candidate genes. This approach of transforming normal human iPSC into tumor cells would recapitulate what can occur during MB initiation in humans. The models produced by this proposal could be modified to examine passenger events and to gain an understanding of early events that produce this tumor type. Successful completion of these studies will generate accurate and highly predictive humanized mouse models to understand the biology and genetics of human MB. These models can then be utilized to evaluate treatment algorithms for children with high-risk, MB.
Our project aims to produce a model of human medulloblastoma in a mouse that closely resembles the progression from a normal human stem cell to a medulloblastoma cancer cell. Since this model will be relevant for this human disease, it will be a valuable tool to determine the efficacy of various forms of treatment for patients with medulloblastoma and gain accurate and predictive results. The standard of care for medulloblastoma patients will likely be influenced based on therapeutic experiments utilizing this model.