Diffuse intrinsic pontine glioma (DIPG) is the most lethal childhood cancer and causes death in virtually all children within 1-2 years of diagnosis. Because DIPG tumor tissues are not available from surgery, autopsy provides the only and the most precious opportunity to collect tumor tissues. The objective of our proposal is therefore to determine if autopsied DIPG tissues can be used to develop orthotopic (intra-brain stem) xenograft mouse models that will look like and behave like human DIPGs, so that they can be used to understand tumor biology and to test new therapies. Our central hypothesis is that some tumor cells in the autopsied DIPG tumors can survive in the postmortem period of severe hypoxia and starvation and form new tumors if these surviving tumor cells are placed back in mouse brain stems, which share maximum similarities to the nature habitat of human DIPGs. A series of tests will be performed to accomplish the following specific aims: Aim 1: To develop a panel of orthotopic xenograft models for pediatric DIPG through direct engraftment of autopsy tumor cells into the brain stems of immunodeficient mice. Aim 2: To confirm that the resultant intra-brain stem xenograft tumors replicate the histopathological and invasive growth of the original patient tumors. Aim 3: To identify new therapeutic targets in the xenograft tumors through comprehensive whole genome gene expression profiling, genotyping and sequencing of the xenograft tumors. Completion of our proposed study will not only provide a set of critically needed and clinically relevant animal models for DIPG and identify series of novel therapeutic targets, but also provide a new concept that can potentially revolutionize the use of autopsied tumor tissues for biological and preclinical studies of late/lethal stage of human cancers by demonstrating that the viable and tumorigenic tumor cells can be harvested from postmortem tissues.