No description available.

Glioblastoma (GBM) is an aggressive brain cancer with devastating outcomes and few effective treatments. Although immunotherapy has shown promise in treating a variety of cancers, it is still unclear if and how it can be effectively used in GBM. Elucidating this will require a better understanding of the mechanistic role of immune cells and their interactions in the GBM tumor microenvironment. This thesis utilizes recent technological developments in cancer genomics and imaging to study the mechanisms underlying immunotherapy and the tumor microenvironment. First, we will provide background on our current understanding of GBM, its immune microenvironment, as well as modern sequencing and imaging methods. Second, we will present a longitudinal study of GBM patients before and after treatment with PD-1 immunotherapy. Only a small fraction of GBM patients respond to this type of therapy, so we perform genomic, transcriptomic, and spatial analyses to compare the molecular features of these rare responders versus non-responders. We show that clinical response to PD-1 immunotherapy in GBM is associated with specific molecular alterations and immune infiltration profiles that reflect the tumor’s clonal evolution during treatment. The most common infiltrating immune cells in GBM are macrophages, which are implicated in a wide variety of pro-tumor and anti-tumor roles. We then focus on this specific immune population by analyzing single-cell expression data from GBM tumors. We identify a novel macrophage subpopulation characterized by expression of the scavenger receptor MARCO, which drives tumor progression in GBM and is altered over the course of PD-1 immunotherapy. Next, we demonstrate that the methods we have developed for GBM are applicable to understanding the tumor microenvironments of other cancers as well. We analyze a cohort of melanoma cases to show that transcriptomic and imaging features can be combined to create a biomarker that stratifies patients into different risk groups. Finally, while most of the image analysis described so far has utilized histopathology, we include two appendices where we demonstrate new ways to process and analyze Magnetic Resonance Imaging (MRI) in GBM.

T-cell lymphomas (TCLs) are a highly aggressive and heterogeneous group of non-Hodgkin lymphomas derived from post-thymic mature T- and NK-cells broadly classified peripheral (PTCL) or cutaneous (CTCL), indicating either a nodal or skin-homing disease, respectively. One of the main clinical challenges contributing to dismal outcomes for TCL patients is both the lack of curative treatment strategies and the high rate of relapse for currently approved therapies, underscoring the need for identification of novel targeted therapies for the treatment of TCL. While combination therapeutic strategies have been proposed and show great promise in pre-clinical and clinical trials for PTCL and CTCL, none are yet approved. Additional contributing factors toward the difficulty in studying TCL and the high rate of therapeutic failure is the highly heterogeneous genetic and molecular mechanisms driving TCL as well as the poorly understood role of non-tumor microenvironment cells in the pathogenesis of TCL. Indeed, while several studies have suggested that tumor associated macrophages play both a significant functional role in supporting tumor maintenance and are therapeutically targetable, less is known about potential tumor supporting roles of other cell microenvironment populations. Here, I used an unbiased and high-throughput approach to discover novel drug combinations in CTCL and to characterize at the single-cell level relevant molecular mechanisms driving T-cell lymphomagenesis. First, I demonstrate that the combination of romidepsin, a selective class I HDAC inhibitor, with afatinib, an inhibitor of the epidermal growth factor receptor (EGFR) family, produces strongly synergistic antitumor effects in CTCL models, both in vitro and in vivo, using mechanisms of action that involve down-regulation of the JAK-STAT signaling pathway. This result suggests a potential therapeutic role for the combination of HDAC inhibitors with afatinib in the treatment of CTCL that had not been previously recognized. Second, we developed single-nuclei analysis on a cohort of 30 TCL (PTCL-NOS, AITL, epstein barr virus positive PTCL) and 6 normal patient samples to identify and deconvolute genomic and functional mechanisms contributing to T-cell lymphomagenesis. Here, I implemented a comprehensive bioinformatics pipeline for the analysis of sparse single-nuclei transcriptomic data and characterized heterogeneous molecular mechanisms driving T-cell lymphomagenesis, such as the upregulation of the PI3K-AKT-mTOR and WNT signaling pathways in a subset of 5/9 PTCL-NOS and 7/19 AITL samples, respectively. Additionally, I identified the enrichment of both the macrophage compartment in PTCL-NOS and AITL, and the specific enrichment of CD8+ T cells in AITL. These results suggest a correlation between patient-specific characteristics, such as mutational status, and possibly therapeutically targetable molecular mechanisms driving neoplastic cell growth that warrants further investigation.

Historically, the role of immune cells in the nervous system was predominantly examined throughthe lens of disease. In recent years, studies have shown that the complex, orchestrated events of immune activity throughout embryonic and postnatal critical periods are crucial for proper nervous system development. While previous studies have suggested limited immune heterogeneity in the adult brain, the diverse roles of the hippocampus in cognition and pathological development would suggest variation of immune cells in this region. Specifically, the hippocampus is known to be a site of adult neurogenesis. However, fundamental traits of immune cells in this region have not been well characterized. In chapter one, I present a summary of literature that discusses what was previously known of immune regulation of adult neurogenesis during health and disease. In chapter two, I compare different reporter lines and marker genes to evaluate responses in various cell types in the neurogenic niche and in other regions of the brain in the context of injury and pharmacological modulation. I discuss preliminary evidence suggesting microglial depletion may result in phenotypic changes in astrocytes throughout the hippocampus. In chapter three, I provide evidence of heterogeneity in myeloid-lineage cells in the hippocampus. I leveraged the highthroughput nature of cell suspension based single cell RNA-sequencing to collect transcriptomes of over 20,000 myeloid lineage cells from murine hippocampi. Using a series of bioinformatic techniques, I was able to computationally dissect different populations within this system and found spatial mapping of one distinct subset specifically localized to the neurogenic niche of the hippocampus. The transcriptomic signature of these cells alongside immunoreactivity to candidate genes, and morphological properties of this population resemble those of reactive microglia associated with the restriction of neurodegenerative diseases. In chapter four, I discuss how the immune landscape of the hippocampus responds to perturbation using a model of Focused Ultrasound mediated Blood-Brain Barrier opening. Subtypes of myeloid lineage cells change in composition and in transcriptomic response. We find distinct, temporally defined transcriptional responses in microglial and macrophage populations, indicating discrete roles for microglia and macrophages in immune activity during the transition from acute to chronic response. Together, these findings point towards diverse properties of microglia in the adult hippocampus.