Tissues are heterogeneous communities of cells that work together to achieve a higher-order function. Large-scale single cell RNA-sequencing (scRNA-seq) offers an unprecedented opportunity to systematically map the transcriptional programs underlying this diversity. However, extracting biological signal from noisy, high-dimensional scRNA-seq data requires carefully designed, statistically robust methodology that makes appropriate assumptions both for the data and for the biological question of interest. This thesis explores computational approaches to finding biological signal in scRNA-seq datasets. Chapter 2 focuses on preprocessing and cell-centric approaches to downstream analysis that have become a mainstay of analytical pipelines for scRNA-seq, and includes dissections of lineage diversity in high grade glioma and in the largest neural stem cell niche in the adult mouse brain. Notably, the former study suggests that heterogeneity in high grade glioma arises from at least two distinct biological processes: aberrant neural development and mesenchymal transformation. Chapter 3 presents a flexible approach for de novo discovery of gene expression programs without an a priori structure across cells, revealing subtle properties of a spatially sampled high grade glioma that would not have been apparent with previous approaches. Chapter 4 leverages our prior work and a unique tissue resource to build a unified reference map of human T cell functional states across tissues and ages. We discover and validate a novel pan-T cell activation marker and a previously undescribed kinetic intermediate in CD4+ T cell activation. Finally, ongoing work defines key programs of gene expression in tissue-associated T cells in infants and adults and predicts their candidate regulators.

Mice with human immune systems, generated by transplanting human CD34+ cells into immunodeficient mice, are essential tools for studying phenomena unique to the human immune system or poorly reproduced in existing mouse models. Human immune tolerance induction, function and autoimmunity have been poorly modeled in conventional murine models, which often have poor predictive value for preclinical development. Models that allow the study of human immune cells with the reproducibility and flexibility of small animal models are required. In our lab, humanized mouse models have been used to study preclinical protocols for human xenotolerance induction and to better understand the immunological underpinnings of human autoimmunity. These are each areas of critical unmet medical need. Xenotolerance-inducing protocols may be necessary to allow long-term survival of a transplanted pig organ in a human patient, and, with more than 113,000 Americans currently waiting for a life-saving organ, the need to expand the pool available for transplantation is urgent. Additionally, clinical options for patients with autoimmune diseases are limited. Currently, most patients with autoimmunity are only diagnosed after significant immune damage of target organs. Predicting who will develop autoimmunity – and who will not – before damage occurs would be very useful but is currently very difficult. Small animal models that can better help us understand how human autoimmunity develops could help us develop protocols for early detection and even prevention. We have developed a personalized immune model to study the development of an individual patient’s immune system in a transplanted mice to better understand immune abnormalities that underlie autoimmunity. We have used existing humanized mouse models to answer important questions related to human xenotolerance induction and autoimmunity, but in the studies described here we have worked to extend our capacity to use these models to study human T cell development and peripheral function. We would like to be able to study both the initial selection of T cell receptors (TCRs) in the thymus based on their ability to recognize antigen in the context of presenting MHC without reacting unduly to self-antigen, as well as in the periphery, where T cells interact with peripheral antigen-presenting cells (APCs) to maintain homeostasis and respond to antigen. First, we have incorporated TCR transgenesis into our humanized mouse models to allow greater precision in studying thymic selection in our humanized mice. Developing a system for this would allow us to study in greater detail mechanisms of human xenotolerance induction, including confirming that a swine thymus can support positive selection of T cells with human-restricted TCRs to allow a future xenotransplantation patient to maintain immune competence, while also robustly tolerizing human T cells expressing pig-reactive TCRs. We will also expand this system to study the thymic selection of human T cells with autoreactive TCRs to better understand mechanisms of central tolerance and understand how they fail in autoimmunity. Finally, while processes of thymic selection are critical for human T cell development and function, peripheral interactions also have a large impact on human T cell function and homeostasis and may contribute to the development of autoimmunity. For these interactions to occur appropriately requires robust engraftment and reconstitution of APCs, especially of myeloid and B cell lineages, in transplanted immunodeficient mice. APC reconstitution tends to be suboptimal in humanized mice and is even more so in mice transplanted with patient-derived CD34+ cells. Better characterization of human APC populations and their progenitors could allow us to develop approaches to improve long-term human APC reconstitution in patient-derived humanized mice, allowing us to more fully model patient peripheral T cell function.

The immune system plays a critical role in the prevention and eradication of cancerous lesions. Indeed, cancer immunology is a rapidly growing area of study that has already generated several FDA-approved treatments and cellular therapies to address mechanisms by which various cancers evade immune clearance. Recent studies look to expand the clinical use of these current therapies and to identify novel targets for future treatments in order to meet the unmet medical needs of the cancer patient population. Recently, multiple correlative studies have identified Carabin (Tbc1d10c) as a potential biomarker for cancer prognosis, including head and neck squamous cell carcinoma (HNSCC), breast cancer, and melanoma. Two mechanistic studies have shown that Carabin acts as a negative feedback inhibitor of canonical TCR and BCR signaling during lymphocyte activation. One group demonstrated that Carabin inhibits CD4+ T-cell activation by binding to and inhibiting the actions of Ras and calcineurin, leading to decreased NFAT and AP-1 transcriptional activity. The second group corroborated the impact of Carabin signaling on the Ras/MAPK pathway in B cells and implicated a role for Carabin in autoimmune diseases in both mice and humans. Collectively, these studies suggest Carabin’s potential role in chronic inflammation and as a pro-tumorigenic target in human cancers. The data presented in Chapters 2-3 demonstrate an immunosuppressive role for Carabin in tumorigenesis. Using three murine tumor models, we identified a novel cancer phenotype in immune competent germline Carabin-ablated (Carabin-/-) mice: these mice showed a twofold decrease in tumor growth and an increase in tumor-free survival compared to wild-type (Carabin+/+) mice. Further assessment identified Carabin expression localized to cells of the immune lineage within the tumor microenvironment (TME), and tumor immunophenotyping showed a twofold increase in the percent of Carabin-/- total and activated CD8+ T cells infiltrating the tumors. Carabin-/- CD8+ T cells displayed an increase in TCR activation and tumor cell killing with no impact on proliferation or migration, indicating that the identified tumor outcome phenotype is due to a suppressive action on the CD8+ TCR activation pathway. Adoptive transfer of tumor antigen-restricted CD8+ T cells into immune-deficient Rag2-/- mice led to reduction of tumor growth in mice receiving Carabin-/- CD8+ T cells. Thus, the data in Chapters 2-3 demonstrate that Carabin deficiency confers tumor resistance via increased CD8+ T-cell anti-tumor activity. This anti-tumor activity is due to an increase in basal NF-κB activity specifically within CD8+ T cells. NF-κB perturbation is the result of a twofold increase in MEKK3 (Map3k3) protein and its downstream phosphorylation of the IKK complex to activate canonical NF-κB. MEKK3 knockdown by siRNA rescued the Carabin-/- in vitro molecular and cellular phenotype without impacting Carabin+/+ CD8+ T cells, and therefore supports the assertion that Carabin signaling is mediated by downstream MEKK3 activity. The NF-κB pathway is critical for T-cell activation and effector function. NF-κB perturbation was selective to CD8+ T cells and not found in CD4+ T cells. Thus, Carabin may be a novel target to mediate NF-κB signaling specifically in CD8+ T cells to improve their effector function within the TME without simultaneously impacting NF-κB in neoplastic or immunosuppressive cells. This is the first study to identify a causative link between Carabin and solid tumor malignancies, to demonstrate a unique mechanism for Carabin in the CD8+ T-cell response to tumorigenesis, and to suggest Carabin as a novel CD8+ T-cell-specific NF-κB inhibitor.

Indiana University-Purdue University Indianapolis (IUPUI) / Regulatory T (Treg) cells represent an important layer of immune-regulation indispensible for curtailing exuberant inflammatory responses and maintaining self-tolerance. Treg cells have translational potential for autoimmunity, inflammation, transplantation and cancer. Therefore, delineating the molecular underpinnings underlying the development, suppressor function and stability of Tregs is particularly warranted. The transcriptional repressor Bcl6 is a critical arbiter of helper T cell fate, promoting the follicular helper (Tfh) lineage while repressing Th1, Th2 and Th17 differentiation. Bcl6-deficient mice develop a spontaneous and severe Th2-type inflammatory disease including myocarditis and pulmonary vasculitis, suggesting a potential role for Bcl6 in Treg cell function. Bcl6-deficient Treg cells are competent in controlling Th1 responses, but fail to control Th2 inflammation in an airway allergen model. Importantly, mice with Bcl6 deleted specifically in the Treg lineage develop severe myocarditis, thus highlighting a critical role for Bcl6 in Treg-mediated control of Th2 inflammation. Bcl6-deficient Tregs display an intrinsic increase in Th2 genes and microRNA-21 (miR-21) expression. MiR-21 is a novel Bcl6 gene target in T cells and ectopic expression of miR-21 directs Th2 differentiation in non-polarized T cells. MiR-21 is up-regulated in mouse models of airway inflammation and also in human patients with eosinophilic esophagitis and asthma. Thus, miR-21 is a clinically relevant biomarker for Th2-type pathologies. Our results define a key function for Bcl6 in repressing Gata3 function and miR-21 expression in Tregs, and provide greater understanding of the control of Th2 inflammatory responses by Treg cells.