PROJECT 1: PROJECT ABSTRACT Glioblastoma (GBM) is among the most lethal and heterogeneous cancers. However, investigations of intratumor heterogeneity, evolution, and therapeutic target selection in GBM rely on a tiny fraction of the whole tumor. Precision therapies are therefore destined to fail if the target is heterogeneous. Conversely, tumor wide targets likely arise in the earliest clonal expansion and should be present in most or all tumor cells. Similarly, gene expression and epigenetic regulatory programs that are heterogeneous in a tumor may be driven by the microenvironment or sub-clonal genetic alterations, whereas programs that are active in all tumor regions could potentially represent the tumor’s cell lineage of origin. We hypothesize that an imaging-based tumor wide approach will elucidate the genetic alterations of the earliest clonal expansion for each patient and the cell lineage from which the tumor arises. The approach could also allow us to redefine tumor heterogeneity, evolution, and target selection. To evaluate our hypothesis we will build an intratumor sample cohort from 25 newly diagnosed and 25 recurrent GBM, including patient-matched pairs. Each sample will have a 3D spatial coordinate for mapping back into the corresponding 3-dimensional (3D) tumor representation we will create (Aim 1). A ShinyApp will be designed for easy exploration of the data in the original context within the tumor. In Aim 2, we will define the evolutionarily earliest tumor clone for each patient, including its genetic alterations and the cell lineage from which the tumor arises. This will be accomplished by applying Hi-C epigenome analysis and RNA sequencing, and other methods, to the samples from Aim 1. The assays are highly complementary and provide comprehensive analyses of genome alterations, 3D chromatin contacts and the gene expression programs they regulate. From preliminary profiles of 103 samples from 10 GBM patients we discovered tumor wide genetic alterations and diverse, large scale structural variants from Hi-C data that have a common feature – they produce gene fusions that are expressed tumor wide. We will distinguish heterogeneous expression and gene regulatory programs from those that are tumor wide and may represent the cell lineage of origin, as highlighted in our preliminary data. We will determine whether the tumor wide genetic alterations and expression programs are retained over time. Heterogenous alterations and programs will be explored for relationships with advanced imaging parameters. In Aim 3 we will determine the transforming capacity and therapeutic value of tumor wide genetic alterations such as gene fusions that create immunogenic neoantigens. By delineating the earliest tumor clone we will lay a foundation for human GBM cell models that represent main root GBM cells of individual patients. Focusing on tumor wide targets will enable development of therapeutic strategies that avoid a major source of treatment fail...