A single-cell multi-omic approach to the analysis of T cell differentiation

This thesis aims to investigate T cell differentiation through the bioinformatic analysis of single-cell multi-omic data. T cells are an important part of the adaptive immune system, involved in the immune response to infections and cancer. Single-cell technologies have advanced to the point where multiple modes of data, such as gene and protein expression, can be assayed on the same cells. Greater understanding of T cell differentiation pathways at the single-cell level can help in the design of immunotherapies to treat cancer and autoimmune disease. The thesis begins by presenting a multi-omic workflow that combines scRNA-seq and T cell receptor (TCR) sequence extraction. The principles developed for this workflow were applied to investigate T cell differentiation in two scenarios. The first scenario was an application of single-cell multi-omics to the study of CD8+ T cell peripheral tolerance mechanisms in a mouse model. This work demonstrated that tolerance is a distinct differentiation program to functional effector responses, and T cells progressively commit to the tolerised state over the first 60hrs post exposure to triggering antigen. A gene signature for the tolerised state was identified, containing genes uniquely upregulated in tolerised cells. Quiescent and Proliferating clusters were found in tolerised cells, indicating that a proportion of cells exit cell cycle within each division. The second scenario was an investigation of the differentiation of CD4+ CAR T cells in vivo, and the evolution of a lymphoma derived from these cells. Three cell types, proliferating, cytotoxic and resting, were observed within the malignant CAR T-cells, and these types were also observed within non-malignant CAR T and endogenous CD4+ T cells. The lymphoma was characterised by expression of the NF-κB transcription factor in all three cell types, while each cell type had differing expression levels for several other known oncogenes. This thesis has contributed to the understanding of T cell differentiation in tolerance and CAR T therapy, and has helped meet the challenge of increasingly large and complex single- cell datasets through the development of bioinformatic workflows to integrate samples from multiple patients and sequencing technologies, and integrate gene, protein, TCR sequence, cell division count and somatic mutation data at the single-cell level