bioRxiv2023
The T cell receptor (TCR) is a key component of the adaptive immune system, recognizing foreign antigens and triggering an immune response. Competing models exist to explain the high sensitivity and selectivity of the TCR in discriminating self from non-self antigens, particularly models using kinetic proofreading (KP), kinetic segregation (KS), and combinations of the two. In this paper, we consider the role and importance of KS in TCR activation, using two models: classic KP (cKP), without KS, where antigen-TCR binding is required for activation, and a combination of KP and KS (KS-KP), where only residence within a close contact is required for activation. Building on previous work, our computational model is the first to permit a head-to-head comparison of these models in silico. While we find that both models can be used to explain the probability of TCR activation across much of the parameter space, we find biologically important regions in the parameter space where significant differences in performance can be expected. Furthermore, we show that the available experimental evidence may favour the KS-KP model over cKP. Our results may be used to motivate and guide future experiments to determine highly accurate computational models for the TCR. Author summaryThe T cell receptor (TCR) is a master of reliable sensing: it detects faint signals (rare ligands derived from foreign proteins) over high noise (abundant ligands derived from the bodys own proteins) to set T cells on a course to exterminate pathogens and tumours, a process that is central to our immune response. Despite decades of studying TCR signalling, we still do not know how the TCR can be so exceptionally sensitive and accurate. It is widely believed that kinetic proofreading (KP), in which the TCR binds to an antigen and triggers a series of phosphorylation steps prior to activation, plays an important role. However, recent results suggest that kinetic segregation (KS), in which binding is not required, is also important. These models are mutually exclusive, and yet both appear to explain various aspects of T cell activation. Our work directly addresses this puzzle. We develop a computational modeling framework which can simulate TCR activation by both KP-based and KS-based models, making it possible to compare them in silico for the first time. Using this framework, we find conditions under which the two models provide different responses, and we show that the limited experimental evidence to date is consistent with KS, which should motivate further investigation.