Alan Williams, who was loathe to consider conformational change-based mechanisms “because they [are invoked to] explain everything”, was the first to emphasize that the paucity of T-cell receptor (TCR) ligands on antigen-presenting cells (APCs) probably meant that the TCR would not be triggered by dimer formation because the binding kinetics would be too fast to allow adjacent TCRs to be coengaged by ligands.
He argued that TCRs are more likely to be triggered at the level of individual complexes, proposing the idea to private audiences that the key event is the recruitment of the coreceptors CD4 and CD8 to ligand-engaged complexes. However other work showed that TCR triggering enhances the binding of CD4 and CD8 to MHC class II and class I molecules, respectively, indicating that coreceptor recruitment to the complex itself is dependent on prior triggering of the TCR, rather than vice versa. A very notable observation of Hua Xu and Dan Littman revealed the importance of the Src versus the kinase domain of Lck, emphasizing the central role of complex stabilization in TCR triggering.
Somewhat earlier Tim Springer observed that, owing to their large size, glycocalyx components including receptor-type protein tyrosine phosphatases such as CD45 will likely be excluded from sites of TCR/ligand engagement. The important corollary of this idea was that the size-dependent exclusion of the tyrosine phosphatase from the vicinity of tyrosine kinase-dependent signalling receptors would probably increase the half-lives of phosphorylated species in that region. However, this did not solve the problem of how the TCR becomes phosphorylated. An unexpected and, at the time, overlooked finding by Gary Koretzky and others was that the TCR is apparently constitutively phosphorylated, as shown by treatment of resting T cells with the tyrosine phosphatase inhibitor pervanadate. The amount of phosphorylation apparent in these experiments was so great it was not immediately obvious how ‘ligand-induced’ kinase activity could have any additional effect, raising the possibility that the size-dependent removal of phosphatases alone from regions of TCR complex formation could be responsible for increased phosphorylation of the TCR. The most obvious candidate phosphatase was CD45 since it comprises >90% of the membrane-associated tyrosine phosphatase activity in T cells and because it seemed large enough to be passively excluded from regions of close contact between a T cell and APC. On the other hand, the abundance of CD45, occupying as much as 10% of the cell surface, coupled with its high enzymatic activity and relatively broad specificity, was expected to suppress TCR triggering in resting cells.
A final issue concerned how a receptor that failed to “see” cognate antigen differed from one that did so during contact with APCs. In other words, how would the presence of ligand influence the phosphorylation state of the TCR? We speculated that the phosphatase-deficient regions of the cellular contact occupied by the TCR might initially be so small that phosphorylated but unproductively engaged complexes diffuse from the region of close contact rapidly enough to be dephosphorylated before they can initiate inappropriate “downstream” signals. These considerations lead to the proposal (1) (Fig. 1), with Anton van der Merwe, of the “kinetic-segregation” (KS) model of TCR triggering, as it is now known. The model postulates that low levels of TCR phosphorylation are maintained by an equilibrium between kinases and CD45, until it is disturbed locally in favour of the kinases when CD45 is sterically excluded from hypothetical structures called “close-contact zones” where TCRs and other small proteins such as CD2 engage their ligands, resulting in net receptor phosphorylation.
For an animation of the KS model, please see Movie 1.
Fig. 1: Dimensions of CD2 and other T-cell surface molecules: implications for T-cell triggering
(a) Scale models are shown of some of the key molecules present at the cell surface during the interaction of T cells with APCs, based on crystallography or electron microscopy data. (b) On resting T cells, the TCR/CD3 complex and other molecules involved in antigen recognition are expected to be randomly distributed among much larger molecules such as CD45. (c) It is proposed that binding of CD2 to its ligand(s) leads to close membrane approximation within multiple contact zones between the T cell and APC. In addition to optimizing the interaction between the TCR and peptide-MHC, membrane approximation may exclude CD45 from the contact zone, thereby extending the half-life of any tyrosine-phosphorylated signalling intermediate in the adjacent cytoplasm. This model postulates that TCR triggering requires stable complexes (e.g. complex II), which remain in the contact zone long enough for completion of all the requisite steps in the signal transduction pathway. For clarity, only CD2, CD4/Lck, CD45 and the TCR/CD3 complex have been depicted in (b) and (c). This model of TCR triggering is described in more detail in (1). Abbreviations: APC, antigen-presenting cell; GPI, glycosylphosphatidylinositol; IgSF, immunoglobulin superfamily; ITAM, immunoreceptor tyrosine-based activation motif; LFA-1, leukocyte function-associated molecule 1; MHC, major histocompatibility complex; TCR, T-cell receptor.
- Davis SJ, van der Merwe PA. (1996) The structure and ligand interactions of CD2: implications for T-cell function. Immunol Today. 17, 177-87.
KS Model Papers
Oliveira MI, Gonçalves CM, Pinto M, Fabre S, Santos AM, Lee SF, Castro MA, Nunes RJ, Barbosa RR, Parnes JR, Yu C, Davis SJ, Moreira A, Bismuth G, Carmo AM. (2012), Eur J Immunol. 42, 195-205Details
A new pathway of CD5 glycoprotein-mediated T cell inhibition dependent on inhibitory phosphorylation of Fyn kinase
Bamberger M, Santos AM, Gonçalves CM, Oliveira MI, James JR, Moreira A, Lozano F, Davis SJ, Carmo AM. (2011), J Biol Chem. 286, 30324-36Details
Davis SJ, van der Merwe PA. (2003), Trends Immunol. 24, 624-6Details
van Der Merwe PA, Davis SJ. (2002), Science. 295, 1479-80Details
The immunological synapse: required for T cell receptor signalling or directing T cell effector function?
Davis SJ, van der Merwe PA. (2001), Curr Biol. 11, R289-91Details
Cytoskeletal polarization and redistribution of cell-surface molecules during T cell antigen recognition
van der Merwe PA, Davis SJ, Shaw AS, Dustin ML. (2000), Semin Immunol. 12, 5-21Details