Scientific summary

Overview

Our activities are directed at understanding the cell biology of the leukocyte cell surface, and integrating these findings into coherent and systematic models of T-cell recognition and activation. The scope of our work extends from the molecular to the cellular scale, drawing on X-ray crystallography, interaction studies, single molecule fluorescence-detection methods, and global gene expression and bioinformatic analyses. We approach the problem of understanding cell surface function from a largely structural point of view because this places the most useful constraints on possible explanations of biological function. Wherever we can we also consider the immunological implications of our observations.

We were initially interested in the problem of how cell-cell contacts can be specific yet weak enough to allow cells to engage reversibly. To get at this question we needed to determine the structures of interacting adhesion molecules. Such proteins tend to be heavily glycosylated, which generally inhibits their crystallization. We therefore had to establish, in the early 1990’s, new methods for dealing with the “glycosylation problem” [1] (see [2] for the current implementation of these methods). This subsequently allowed us to crystallize a number of cell surface molecules. Our initial crystals yielded the first overall structure of a cell adhesion molecule, CD2 [3], which suggested that charged residues, since they are clustered in the ligand binding site, would have an unusually prominent role in adhesive interactions involving CD2. Subsequent mutational work supported the idea that electrostatic contacts are ideally suited to weak, specific recognition because electrostatic complementarity is required to compensate for the removal, upon binding, of water interacting with charged residues in the binding interface [4]. We argued that this represents a new type of protein interaction, i.e. a “magnetic swipe-card”-like form of recognition distinct from the classical “lock-and-key” paradigm. When we determined the structure of the CD2 ligand, LFA-3, we were able to verify our new concept by predicting the overall structure of the CD2/LFA-3 complex, based on the principle of maximizing electrostatic complementarity only [5].

We have now moved on from the analysis of adhesion molecules to cell surface signaling proteins. The CD28/B7 family, i.e. CD28, CTLA-4, B7-1 and B7-2, are involved in both activating and inhibitory “costimulatory” signaling in T cells. These molecules determine, to a considerable extent, the outcome of immune responses and are therefore excellent potential targets for immunotherapy. Our structure of soluble B7-1 [6] revealed, very unexpectedly, that it is likely to dimerize at the cell surface, a result that we have since confirmed for cell surface-expressed B7-1 using bioluminescence resonance energy transfer (BRET) [7]. Because B7-1 binds another dimer, the inhibitory protein CTLA-4, we predicted that these two molecules would form extremely stable 1D “zipper-like” arrays at the T-cell surface. Such arrays were subsequently seen in crystals of the complex of B7-1 and CTLA-4 [8]. This is important because it implies that B7-1 may have evolved specifically to enhance strong inhibitory signaling by CTLA-4, a suggestion supported by detailed studies of the binding properties of these molecules [9]. This idea is also strongly supported by collaborative work with Patric Nilsson’s group (Skövde), with whom we have developed in silico methods to model the synaptic accumulation of costimulatory complexes based on rigorous biophysical and expression data [10]. The structure of the B7-1/CTLA-4 complex also showed that contact between these molecules, although still weak, is very different from that for CD2/LFA-3. This implies that, for cell surface molecules, it is probably more important that binding is weak and specific than how this is achieved. We have now reviewed our work on adhesive and costimulatory protein interactions in Annual Reviews [11] and Nature Immunology [12].

It seems to us that one of the outstanding questions in T-cell surface biology concerns the mechanism of receptor “triggering” by monovalent receptors, such as the T-cell receptor (TCR) and CD28, that interact with generally monomeric ligands and are reliant on extrinsic protein kinases. For various reasons, early explanations based on receptor aggregation or dimerization, or ligand-induced conformational changes, appear to us to be inadequate. In 1996, we proposed, with Anton van der Merwe (Oxford), a somewhat counter-intuitive mechanism for TCR triggering that is based on kinetic principles and on the physical segregation of key signaling molecules [13]. Our concept, which we call the kinetic-segregation (KS) model, seems to explain most existing data relating to TCR triggering and circumvents the problems inherent in the early models. We propose that large molecules such as the phosphatase, CD45, are segregated from smaller signaling proteins, such as the TCR, according to size and that specific ligand contacts hold the TCR in the “pro-signaling”, phosphatase-depleted microenvironment of the contact zone, ensuring that the stochastic phosphorylation of the TCR leads to downstream signaling. An important point is that the KS model requires that the TCR only has to engage peptide-MHC molecules and that otherwise it is a largely passive molecule. This fits very well with what we know about the structure of the TCR. There is now also very good experimental support, mostly from the van der Merwe group, for the notion that the size of CD45 is a critical factor in TCR triggering.

An indication that the TCR is not a special case, and that the triggering of other molecules might be based on KS-like principles, came from our structural analyses of CD28 complexed with mitogenic (or “superagonistic”) and non-mitogenic antibodies [14]. The only substantive difference between mitogenic and non-mitogenic antibody/CD28 complexes is that the latter extend ~100Å further from the cell surface. Noting that antibodies need to be immobilized to be active, we predicted that the differences in the sizes of the antibody complexes affects the degree of access of small and large signaling molecules, such as phosphatases and kinases, to the receptor. Our findings support a unified concept wherein cell surface receptors are triggered by local changes in the ratio of kinase versus phosphatase activity [15]. We have now reviewed the development of the KS model and considered the extent to which it explains key features of TCR triggering [15]. (Please see the /ks_model and /antibody pages for animations of TCR and superagonistic antibody-induced triggering).

Whilst concepts such as the KS model can be valuable for guiding experiments, their real usefulness would, in principle, be restricted for as long as the full set of components of such systems remain unknown. For this reason we undertook a systematic analysis of the set of transcripts encoding molecules expressed at the T-cell surface, based on serial analysis of gene expression. This method, also known as SAGE, generates very short (14-21 bp) tags specific for individual transcripts that can be cloned and sequenced extremely efficiently, allowing systematic “open” sampling of cellular transcriptomes. Somewhat unexpectedly, in answer to the question, “How well do we know the T-cell surface?”, our analysis of ~60,000 SAGE tags [16] indicated that we pretty much fully understand the T-cell specific composition of the resting T-cell surface. In particular, all the components of the T-cell triggering machinery appear to be known.

This finding places new emphasis on the analysis of protein function. An important new focus of our work, therefore, involves studying the organization and behaviour of the known triggering proteins at the single-molecule level, using ultra-sensitive fluorescence microscopic methods. With these approaches we hope to solve the mystery of receptor triggering. With David Klenerman (Cambridge), we have shown that the TCR is comprised of single ?? heterodimers using two-colour coincidence detection [17]. This implies that, in the very first instance at least, triggering relies on the passive association of individual, monovalent TCR complexes with MHC molecules rather than the reorganization of existing bi- or multi-valent complexes. Using BRET, we have shown that most T-cell surface proteins are monomeric and tested the widely held idea that G protein-coupled receptors (GPCRs) are invariably oligomeric, which we thought was an unnecessary complication [7]. To do this, it was necessary to establish a new analytical framework for generating and interpreting BRET data. Our proposal that proteins such as the
?2-adrenergic receptor are monomeric has not been generally well received by the GPCR community.

Finally, we have extended our SAGE analyses to include 1,000,000 tags from a resting and activated CD4+ T-cell clone, giving us >99% coverage of the transcriptome of the T cell. It is our intention to eventually provide the entire transcriptome to the research community in the form of a heuristic, fully annotated web-based resource, the iT-cell, which should assist the systems-level analysis of T-cell function by others.

Key papers

  1. Davis SJ, Puklavec MJ, Ashford DA, Harlos K, Jones EY, Stuart DI and Williams AF (1993) Expression of soluble recombinant glycoproteins with predefined glycosylation: application to the crystallization of the T-cell glycoprotein CD2. Prot Eng 6, 229-232.
  2. Chang VT, Crispin M, Aricescu AR, Harvey DJ, Nettleship JE, Fennelly JA, Yu C, Boles KS, Evans EJ, Stuart DI, Dwek RA, Jones EY, Owens RJ, Davis SJ (2007) Glycoprotein structural genomics: solving the glycosylation problem. Structure. 15, 267-73.
  3. Jones EY, Davis SJ, Williams AF, Harlos K and Stuart DI (1992) Crystal structure at 2.8Å resolution of a soluble form of the cell adhesion molecule CD2. Nature 360, 232-239.
  4. Davis SJ, Davies EA, Tucknott MG, Jones EY, van der Merwe PA (1998) The role of charged residues mediating low affinity protein-protein recognition at the cell surface by CD2. Proc Natl Acad Sci USA 95, 5490-5494.
  5. Ikemizu S, Sparks LM, van der Merwe PA, Harlos K, Stuart DI, Jones EY, Davis SJ (1999) Crystal structure of the CD2-binding domain of CD58 (lymphocyte function-associated antigen 3) at 1.8-A resolution. Proc Natl Acad Sci USA 96, 4289-4294.
  6. Ikemizu S; Gilbert RJC; Fennelly JA; Collins AV; Harlos K; Jones EY; Stuart DI, Davis SJ (2000) Structure and dimerisation of a soluble form of B7-1(CD80). Immunity 12, 51-60.
  7. James JR, Oliveira MI, Carmo AM, Iaboni A, Davis SJ (2006) A rigorous experimental framework for detecting protein oligomerization using bioluminescence resonance energy transfer. Nat Methods. 3, 1001-6.
  8. Stamper CC, Zhang Y, Tobin JF, Erbe DV, Ikemizu S, Davis SJ, Stahl ML, Seehra J, Somers WS, Mosyak L. (2001) Crystal structure of the B7-1/CTLA-4 complex that inhibits human immune responses. Nature 410, 608-611.
  9. Collins AV, Brodie DW, Gilbert RJC, Iaboni A, Stuart DI, van der Merwe PA, Davis SJ (2002) Interaction properties of costimulatory molecules revisited. Immunity 17, 201-210.
  10. Jansson A, Barnes E, Klenerman P, Harlen M, Sorensen P, Davis SJ, Nilsson P (2005) A theoretical framework for quantitative analysis of the molecular basis of costimulation. J Immunol. 175, 1575-85.
  11. van der Merwe PA, Davis SJ (2003) Molecular interactions mediating T cell antigen recognition. Annu Rev Immunol. 21, 659-84.
  12. Davis SJ, Ikemizu S, Evans EJ, Fugger L, Bakker TR, van der Merwe PA (2003) The nature of molecular recognition by T cells. Nature Immunol 4, 217-224.
  13. Davis SJ and van der Merwe PA (1996) The structure and ligand interactions of CD2: implications for T-cell function. Immunol Today 17, 177-187.
  14. Evans EJ, Esnouf RM, Manso-Sancho R, Gilbert RJ, James JR, Yu C, Fennelly JA, Vowles C, Hanke T, Walse B, Hunig T, Sorensen P, Stuart DI, Davis SJ (2005) Crystal structure of a soluble CD28-Fab complex. Nat. Immunol. 6, 271-279.
  15. Davis SJ, van der Merwe PA. (2006) The kinetic-segregation model: TCR triggering and beyond. Nat Immunol. 7, 803-9.
  16. Evans EJ, Hene L, Sparks LM, Dong T, Retiere C, Fennelly JA, Manso-Sancho R, Powell J, Braud VM, Rowland-Jones SL, McMichael AJ, Davis SJ (2003) The T cell surface – how well do we know it? Immunity 19, 213-223.
  17. James JR, White SS, Clarke RW, Johnasen AM, Dunne PD, Sleep DL, Fitzgerald WJ, Davis SJ, Klenerman D (2007) Single molecule-level analysis of the subunit composition of the T-cell receptor on live T cells. Proc Natl Acad Sci USA 104, 17662-17667.

home | research | papers | simon davis | ed evans | co-workers | contacts | search