Multi-omics analysis and modelling of T cell immunity and interferon signalling pathways for engineered T cell therapy

This thesis examines the nature of T cells through the lens of genomics. It delves into type I interferon (IFN)-mediated signalling in the tumour microenvironment (TME), focusing on the GMP-AMP synthase (cGAS)-stimulator of IFN genes (STING) pathway, and explores novel strategies to optimise T cell-based immunotherapy for triple-negative breast cancer (TNBC). In the first chapter, gene expression patterns of distinct T cell subtypes are studied to gain insight into the diverse T cell landscape. Cell type annotation methods are evaluated for their ability to identify closely-related T cell populations in single-cell RNA sequencing (scRNA-seq) data. In the following chapter, single-cell and spatial transcriptomics methods, mainly scRNA-seq, single-cell immune profiling, Slide-seq, in situ sequencing and Molecular Cartography enable the comprehensive analysis of immune infiltrates in tissue biopsies from patients with TNBC. The advantages and limitations of emerging spatial transcriptomics technologies are put into context. Tumour-infiltrating T cells were characterised at varying levels of resolution and cGAS was found to be most highly expressed in proliferating T cells. Tissue-wide spatial patterns revealed a strong relationship between programmed cell death-1/programmed cell death ligand-1 and cGAS but not STING. In addition, gene transcripts encoding the downstream protein kinase TANK-binding kinase 1 were found to spatially associate with hypoxia markers. Finally, T cells genetically engineered to express chimeric antigen receptors (CARs) are investigated using a systems biology approach. An agent-based model is developed to explore the effect of CAR tuning on on-target off-tumour toxicity and therapy efficacy. Gene regulatory networks are used to present readily testable hypotheses for combination strategies using immune checkpoint inhibitors, oncolytic viruses, and cGAS-STING-targeting treatments to potentiate anti-tumour responses. The work presented therein will feed into a novel computational framework to enable the modelling of patient-specific responses to T cell-based immunotherapy and synergistic combinations.