T-Switch: A specificity-based engineering platform for the development of T cell therapeutics

Abstract

Many promising targets for adoptive T cell therapy (ACT) are overexpressed self-antigens. However, T cells bearing T cell receptors (TCRs) with high potency and affinity for a self-antigen are eliminated during thymic selection. The affinity of the remaining low-avidity TCRs can be improved through conventional mutagenesis approaches, but isolating these TCRs is labor intensive, and their subsequent affinity enhancement can generate toxic cross-reactivities. To bypass T cell tolerance and to obtain potent and safe T cell therapeutics, we developed T-Switch, an in vitro TCR engineering platform for the creation, modification, and comprehensive profiling of TCRs that can target tumor-associated self-proteins for ACT. In this approach, we bypass T cell tolerance by first raising T cells from the natural repertoire that recognize a related ‘foreign’ peptide that differs by one amino acid from a self-antigen. Then, we modulate the fine specificity of the TCR by directed evolution of the peptide binding region to switch its specificity to the selfantigen of interest. T-Switch emphasizes the capacity to engineer TCR specificity while avoiding the pitfalls of affinity enhancement by selecting against a closely related epitope and thereby making stepwise changes to fine tune TCR potency and specificity. The first part of this dissertation describes proof-of-concept experiments where we created comprehensive libraries of different cytomegalovirus (CMV) specific-TCRs that, after in vitro selection, switched specificities to closely related antigens. The engineered TCRs showed robust T cell activation after ligand recognition and are of equal or higher efficiency than the parental receptor. Crucially, comprehensive assessment of their off-target cross-reactivity across the viral proteome showed no additional promiscuity or off-target specificities of the engineered TCRs similar to the parental TCR. This is key because it stands in stark contrast to previously engineered TCRs that suffered from increased cross-reactivity. The second part of the dissertation describes the application of T-Switch to the engineering of TCRs to a tumor-associated antigen, Tyrosine Hydroxylase (TH), and validating their translational potential through multiple in vitro and in vivo assays of safety and efficacy. Importantly, we detected no off-target recognition by the engineered TCR variants as measured against the human proteome. Thus, T-Switch represents a valuable technology for the creation of collections of highly sensitive synthetic TCRs for T cell-based immunotherapies.