Development of thymocytes involves two distinct outcomes resulting from superficially similar events. Recognition by thymocytes of major histocompatibility complex (MHC) proteins plus peptides leads to their rescue from apoptosis (positive selection), and recognition of antigenic peptide induces cell death (negative selection). Antigen analogues, and sometimes low concentrations of antigenic peptide, induce positive selection; such analogues are often antagonists of mature T-cell clones. Various models seek to explain how recognition of different peptide/MHC complexes leads to such different outcomes: quantitative models relate response to the affinity, avidity or kinetics of T-cell-antigen receptor (TCR) binding, whereas qualitative models require conformational or spatial changes in the TCR or associated molecules to modulate signal transduction. We have used surface plasmon resonance to measure the kinetics of TCR interactions with positively and negatively selecting ligands to distinguish between these models, and find that affinity correlates to the outcome of selection. A 'window' of affinity resulting in positive selection extends over a 1-log range starting threefold below the affinity for negative selection.

No abstract

The widely accepted kinetic proofreading theory proposes that rapid TCR dissociation from a peptide/MHC ligand allows for stimulation of early but not late T cell activation events, explaining why low-affinity TCR ligands are poor agonists. We identified a low-affinity TCR ligand which stimulated late T cell responses but, contrary to predictions from kinetic proofreading, inefficiently induced early activation events. Furthermore, responses induced by this ligand were kinetically delayed compared to its high-affinity counterpart. Using peptide/MHC tetramers, we showed that activation characteristics could be dissociated from TCR occupancy by the peptide/MHC ligands. Our data argue that T cell responses are triggered by a cumulative signal which is reached at different time points for different TCR ligands.

The kinetics of interaction between TCR and MHC-peptide show a general relationship between affinity and the biological response, but the reported kinetic differences between antigenic and antagonistic peptides are very small. Here, we show a remarkable difference in the kinetics of TCR interactions with strong agonist ligands at 37 degrees C compared to 25 degrees C. This difference is not seen with antagonist/positive selecting ligands. The interaction at 37 degrees C shows biphasic binding kinetics best described by a model of TCR dimerization. The altered kinetics greatly increase the stability of complexes with agonist ligands, accounting for the large differences in biological response compared to other ligands. Thus, there may be an allosteric, as well as a kinetic, component to the discrimination between agonists and antagonists.