PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Nature. Author manuscript; available in PMC 2012 February 6.
Published in final edited form as:
PMCID: PMC3273423
NIHMSID: NIHMS238135

TCR–peptide–MHC interactions in situ show accelerated kinetics and increased affinity

Abstract

The recognition of foreign antigens by T lymphocytes is essential to most adaptive immune responses. It is driven by specific T-cell antigen receptors (TCRs) binding to antigenic peptide–major histocompatibility complex (pMHC) molecules on other cells1. If productive, these interactions promote the formation of an immunological synapse2,3. Here we show that synaptic TCR–pMHC binding dynamics differ significantly from TCR–pMHC binding in solution. We used single-molecule microscopy and fluorescence resonance energy transfer (FRET) between fluorescently tagged TCRs and their cognate pMHC ligands to measure the kinetics of TCR–pMHC binding in situ. When compared with solution measurements, the dissociation of this complex was increased significantly (4–12-fold). Disruption of actin polymers reversed this effect, indicating that cytoskeletal dynamics destabilize this interaction directly or indirectly. Nevertheless, TCR affinity for pMHC was significantly elevated as the result of a large (about 100-fold) increase in the association rate, a likely consequence of complementary molecular orientation and clustering. In helper T cells, the CD4 molecule has been proposed to bind cooperatively with the TCR to the same pMHC complex. However, CD4 blockade had no effect on the synaptic TCR affinity, nor did it destabilize TCR–pMHC complexes, indicating that the TCR binds pMHC independently of CD4.

Surface plasmon resonance (SPR) and microcalorimetry experiments with engineered soluble proteins generally show weak TCR–pMHC binding, with affinities (Kd values) in the range 1–100 µM, t1/2 values on the order of seconds, and association rates usually ranging from 1,000 to 10,000 M−1 s−1 (refs 46). Nevertheless, T cells are highly specific and sensitive for antigen, able to detect even a single antigenic pMHC complex among structurally similar yet non-stimulatory pMHCs7,8. However, although this is useful for comparative purposes, these measurements do not account for the many constraints and special features of the synaptic environment and might not reflect what occurs in situ. In particular, the restricted intercellular volume should greatly accelerate the association rate and enhance serial engagement911. Favourable molecular alignment of TCR and MHC, as well as any molecular pre-clustering, could also drastically affect binding1214. Although the volume effect can be approximated, the influence of the others is essentially unknown. We have therefore devised a method to measure synaptic TCR binding to pMHC directly, using single-molecule microscopy and FRET between a donor fluorophore on the TCR and an acceptor fluorophore on the peptide bound to a MHC molecule.

We used T-cell blasts from two different TCR transgenic T-cell lines, 2B4 and 5c.c7, which are specific for the same moth cytochrome c peptide (MCC 88–103) bound to the class II MHC molecule IEk. To permit the use of highly sensitive and rapid total internal reflection (TIRF) microscopy, we used a modified planar lipid bilayer system as a surrogate antigen-presenting cell surface, presenting IEk complexes, ICAM-1 adhesion molecules and B7-1 co-stimulatory polypeptides (see Methods).

Structural analysis of an Fab fragment derived from the monoclonal anti-TCR-β antibody H57 complexed with the Cβ region of a TCR indicated that portions of its binding site might be close enough to the carboxy terminus of a peptide bound to a MHC (about 41 Å)15,16 to allow FRET in situ to measure TCR binding. We therefore constructed a single-chain variable fragment (scFv) of this antibody with mutations that could be labelled at three different sites (J1, J2 and J3). Their location and proximity to the C terminus of a peptide bound to the IEk molecule are shown (Fig. 1a).

Figure 1
FRET as a sensor for TCR–ligand interactions

On addition of the Cy3-labelled J1 reagent to 5c.c7 transgenic T cells on bilayers containing IEk/MCC(C)-Cy5, we observed the rapid onset of a FRET signal, which was corrected for donor bleed-through and acceptor cross-excitation (see Methods). Consistent with the FRET results was the observation that this signal disappeared immediately after ablation of the FRET acceptor, together with an increase in donor intensity (Fig. 1b). Reversing the labels produced comparable results (Fig. 1b and Supplementary Fig. 1a). The average synaptic FRET yield (as measured by donor recovery after acceptor bleaching; see Methods) was 10–12% for the agonists IEk/MCC(C)-Cy5 and IEk/K5(C)-Cy5 (Fig. 1c). FRET was highly dependent on the ligand: no synaptic FRET was observed between the J1-Cy3-labelled TCR and the null ligands IEk2M(C)-Cy5 or IEk/MCC null(C)-Cy5, even in the presence of unlabelled agonists (IEk/MCC) to ensure T-cell activation (Fig. 1b, c). The FRET strength correlated with both the density of labelled IEk/peptide(C)-Cy5 and the ligand potency (Supplementary Fig. 1b). No FRET was detectable when T cells were labelled with a Cy3-conjugated anti-CD4 Fab (GK1.5), further validating signal specificity (Fig. 1c).

To assess potential bystander FRET between TCR–pMHC complex bound J1-Cy3 and neighbouring pMHCs, we tested the J1, J2 and J3 probes in conjunction with the agonist ligand IEk/K5 labelled at either the C or N terminus of the peptide. These combinations gave rise to six inter-dye distances ranging from about 41 Å to about 100 Å. Average synaptic FRET depended strictly on the distance between the dye pair, as expected, given that yields of FRET decline inversely with the sixth power of the distance (Fig. 1d). Hence, this assay is specific for individual TCR–ligand interactions, and bystander effects are negligible.

We also used single-molecule FRET (smFRET) microscopy to derive dissociation rates. After correcting for donor bleed-through and acceptor cross-excitation, we identified smFRET events as those that both co-localized with single acceptor molecules and also appeared and disappeared in a single step—a hallmark of single-molecule behaviour (Fig. 2a and Supplementary Figs 2–7). Furthermore, the appearance of these events was specific to label and ligand: other label combinations, including J3-Cy3 or IEk/MCC-null(C)-Cy5, did not produce such events (data not shown). Reversing the label proportion by employing IEk/MCC(C)-Cy5 in high abundance and J1-Cy3 in low abundance also led to smFRET events (Supplementary Figs 8–10). By monitoring the duration of these FRET signals during synapse formation and correcting for photobleaching (see Methods), we determined the dissociation rates for both the 5c.c7 TCR and the 2B4 TCR with IEk/MCC at 37 °C and for 5c.c7 with the weak agonist T102S (Fig. 2b, Table 1 and Supplementary Fig. 1c). These data followed single-exponential decay kinetics, which is consistent with a simple dissociation of the TCR–pMHC complex. Unexpectedly, the t1/2 values for these interactions were much shorter than those measured in vitro by SPR, especially for the 5c.c7 TCR (Fig. 2c). Across a temperature range we found that synaptic t1/2 values were consistently 3–12-fold shorter than those measured in vitro, especially at higher temperatures. Recently, a continual ‘flow’ of actin in synapse-forming T cells from the periphery to near the centre of the synapse has been shown17. To determine whether this could be the cause of the shorter t1/2 values, we employed the actin-depolymerizing drugs cytochalasin D and latrunculin A. These treatments greatly prolonged synaptic TCR–pMHC interactions, indicating that actin dynamics destabilizes TCR–ligand interactions either directly or indirectly (Fig. 2d). Under these conditions the t1/2 is almost identical to that measured by SPR (4.23 s or 4.78 s in situ, versus 3.5 s in vitro) with the synaptic t1/2 being slightly greater; this is expected because diffusion is more limited in that environment.

Figure 2
Synaptic off-rates measured through smFRET
Table 1
Summary of measured binding constants

To estimate the affinity, we calculated the concentrations of the reactants and their products by first determining the local density (ρ) of total TCRs, pMHCs and TCR–pMHC complexes in a representative synapse. This was done on the basis of the average intensities of single-molecule fluorescence (Supplementary Fig. 21). To ensure a TCR labelling efficiency of at least 95%, we analysed synapses for 2.5 min or less at 24 °C (where the t1/2 for scFv is about 50 min). With this image information (Fig. 3) we calculated the average local effective two-dimensional Ka (2D Ka) (Ka = 1/Kd) and, using the koff determined above, the kon (kon = Kakoff; Fig. 3a; see Methods). Hotspots of TCR binding are apparent, indicating that both local cellular parameters and the intrinsic chemistry of the interactions influence binding behaviour.

Figure 3
Quantitative image analysis yields effective synaptic 2D Kd, 2D Ka and 2D kon

We also measured the affinities within individual TCR microclusters. To conduct our studies at higher temperatures we quantified the TCR occupancy of TCRs in individual TCR microclusters through their individual FRET yield (as measured by acceptor bleaching), which is independent of the degree of TCR labelling. This and a knowledge of the pMHC density in the bilayer allowed us to calculate the effective 2D Kd values (see Methods). The 2D Kd values for individual TCR microclusters varied significantly (about 250-fold), most probably because of locally active cellular parameters (Fig. 3b). We give the median in situ Kd values for comparison in Table 1.

To derive the effective synaptic three-dimensional Kd (3D Kd) of the interactions taking place within a TCR microcluster for comparison with the corresponding value measured by SPR, we converted the area for which binding was observed into the volume of the intermembrane space. To calculate the concentration we divided the density by the distance of 13.4 nm between the opposed membranes. This width is also the approximate maximum distance at which a TCR and pMHC could interact if fully extended and perpendicular to the plane of their respective membranes18. It is also consistent with electron microscopic analyses of synapses19. For the interaction between 5c.c7 TCR and IEk/MCC at 37 °C (at a density of 30 IEk/MCC(C)-Cy5 µm−2), this resulted in a median Kd of 4.8 µM, which is about 8.3-fold lower than the value determined in vitro, despite the about 12-fold accelerated decay rate that we observed in situ. Assuming steady-state conditions, we calculated a median synaptic kon of 1.38 × 106 M−1 s−1, an almost 100-fold increase compared with the SPR value at 37 °C (Fig. 3c and Table 1). We also determined the dependence of the TCR affinity on the density of ligands presented on the bilayer at 37 °C. Although constant at low and moderate pMHC densities, the median synaptic affinity decreased at ligand densities approaching that of the TCR, suggesting that some TCRs are refractory to binding (Fig. 3d and Table 1).

Synaptic binding properties also correlated with signalling potency: TCR binding was strongest for the agonist IEK/MCC, followed by the weak agonist IEK/T102S and the antagonists IEK/T102G and IEK/K99R (Fig. 3e). IEK/T102S showed a single-exponential decay curve (Supplementary Fig. 1c) with a t1/2 2.4-fold lower (0.7 s) than that of IEk/MCC. This difference is less than sevenfold, as seen with the equivalent SPR measurements (Table 1) and probably reflects the impaired signalling activity of this ligand (see below).

Of special interest with regard to helper T-cell recognition is the role of the CD4 co-receptor, which binds to a non-polymorphic region within MHC class II with a barely measurable affinity (250 µM or more)20, yet has a main role in T-cell sensitivity7. A widely held view is that, to initiate signalling, simultaneous binding of TCR and CD4 to the same pMHC ligand directly recruits the TCR-proximal kinase p56lck to the TCR through the association of p56lck with CD4 (refs 21, 22). Other models suggest that CD4 is associated with the TCR but does not bind the same MHC class II as its associated TCR7,22. To distinguish between these two models, we first compared the average synaptic TCR–pMHC FRET yield in the absence and presence of CD4 blockade, because the FRET yield is related to the average synaptic TCR–pMHC affinity (Supplementary Fig. 12). CD4 blockade had little or no effect on the synaptic affinity between TCR and pMHC, both early and later in synapse formation. smFRET analysis revealed that CD4 blockade slightly stabilized interactions (Fig. 4a). An explanation could be that CD4 blockade attenuates T-cell signalling (Fig. 4b, c). This could dampen cellular motility, which destabilizes TCR–pMHC interactions. Thus we repeated CD4 blocking in the presence of pp2 (a p56lck inhibitor) or cytochalasin D, which completely abrogated the CD4-mediated difference in the binding off-rate (Fig. 4d, e). Thus, the increased TCR–pMHC stability is due to signalling effects. Taken together, these results show that TCRs bind to ligands independently of CD4 and are not, as previously proposed, joined to CD4 molecules as a physical unit. TCR and CD4 could nevertheless still bind the same agonist pMHC simultaneously21, because independent and simultaneous binding events are not mutually exclusive, whether in solution or within the synapse. Because CD4 engagement boosts T-cell activation and sensitivity throughout synapse formation (Fig. 4b, c, and Supplementary Fig. 13), the function of CD4 seems likely to be that of signal reinforcement through its intracellular association with p56lck once a TCR has engaged a viable ligand21.

Figure 4
Effect of CD4 on TCR signalling and ligand binding

Our data offer new insights into both how TCRs discriminate between different peptide-MHC ligands and also the unique challenges of finding and deriving the maximal signalling potential from even very rare agonists. In view of our data, the phosphorylation of TCR-associated CD3 signalling chains—that is, the first step in response to receptor engagement—is certain to result from frequent yet short-lived binding events. This means that, for the 5c.c7 TCR with its t1/2 of about 100 ms, there could be roughly 3,000 separate receptor–ligand interactions in 1 min if ten ligands were engaged simultaneously in a synapse in situ, as is required for maximal calcium activation and stable synapse formation7. If the ligands were fully occupied and each interaction gave rise to a single signalling event, there would be a significant rate of TCR signal transduction, equivalent to triggering about 10% of the TCRs in the cell.

How could rare pMHCs be kept occupied by TCRs, which are relatively sparse and have such unstable interactions? The explanation may lie in ‘protein islands’, distinct 10–200-nm membrane micro-domains highly enriched for particular membrane proteins12,13. We find that clusters of TCRs are expressed on a subset of these islands, as also suggested by others14. Such TCR islands could efficiently ‘scan’ the surface of an adjacent cell and provide a high local density of TCRs surrounding a ligand to keep it continually engaged to generate the maximum possible signal. This is also consistent with cholesterol-depletion experiments in which TCR binding was no longer punctate: binding was decreased fivefold at higher and moderate pMHC densities and was no longer detectable at low densities (Supplementary Fig. 14). Receptor and ligand aggregation might also be facilitated by weak lateral interactions, substrate topography, membrane rigidities and line tension23,24.

Differences in kinetic binding parameters could be substantially enhanced in a synaptic environment. Pre-oriented binding partners, molecular clustering and active T-cell scanning mechanisms would raise the frequency of otherwise rare molecular encounters. In contrast, interactions with a low ΔΔG might be even less stable as a result of cellular forces acting against them. In particular, we found that the 5c.c7-TCR of ‘average’ affinity was more severely destabilized (about 12-fold) than a higher-affinity 2B4-TCR (fourfold). This suggests that this unexpected activity in the synapse could help select for better-quality ligands and against weaker ones.

We have used a new imaging strategy to measure the binding of the TCR to its peptide–MHC ligand. Within an immunological synapse this shows that this very specialized but ubiquitous environment enhances the ability of otherwise very weak and rare ligands to drive important biological effects. With antibody pairs, for example, this FRET strategy could be used to characterize other molecular interactions as well. It seems likely that at least some of the binding characteristics found here will be shared with other cell-surface receptors on other cells, because molecules of this type generally have affinities and dissociation rates in the same range as TCRs10,11,18.

METHODS SUMMARY

We take advantage of FRET to measure in situ the synaptic interactions between cell-bound TCRs and their cognate ligands, pMHC, embedded in a planar lipid bilayer. We engineered a site-specifically labelled TCR-reactive scFv to tag cell-surface-located TCRs. When the TCR is bound to its ligand, its associated scFv brings its label (FRET donor) close enough to the FRET acceptor dye attached to the MHC-embedded peptide to give rise to a FRET signal. We recorded smFRET to measure the t1/2 values of synaptic TCR–pMHC interactions. We combined smFRET and bulk FRET measurements to determine synaptic Kd values and kon values. We applied antibody-mediated blockade of CD4 to assess its contribution to TCR–pMHC binding.

Full Methods and any associated references are available in the online version of the paper at www.nature.com/nature.

Supplementary Material

01

Acknowledgements

We thank F. E. Tynan, A. K. Chakraborty and S. R. Quake for helpful discussions, and P. P. J. Ebert for help with the flow cytometry experiments. This work was supported by grants from the National Institutes of Health (RO1 AI52211) and the Howard Hughes Medical Institute to M.M.D. J.B.H. is partly supported by the Immunology Frontier Research Center (iFREC) consortium, M.A.M. received a postdoctoral fellowship from the Wilhelm Macke Stiftung, and E.W.N. from the American Cancer Society’s Steven Stanley and Edward Albert Bielfelt Fund. L.O.K. is supported by a National Science Foundation predoctoral fellowship. M.A., M.B. and G.J.S. received grants from the Austrian Science Fund (project Y250-B10) and the GEN-AU project of the Austrian Ministry for Science and Research.

Footnotes

Supplementary Information is linked to the online version of the paper at www.nature.com/nature.

Author Contributions J.B.H., G.J.S. and M.M.D. conceived the project. Reagents and experimental systems were designed and tested by J.B.H. unless indicated otherwise. M.A., M.A.M. and G.J.S. provided important expertise in single-molecule microscopy. M.A. conducted all single dye tracing experiments and all smFRET experiments performed in Linz. J.B.H. conducted all bulk FRET experiments and all smFRET experiments at Stanford. J.B.H. and M.A.M. laid the experimental foundation in the initial phase of the project. M.A.M. provided important expertise in setting up the microscope system at Stanford. B.F.L. provided the expression system for ICAM-1 and B7-1. E.W.N. conducted all SPR measurements. M.B. conducted all TOCCSL experiments. L.O.K. and B.F.L. contributed important ideas. J.B.H. and M.M.D. wrote the manuscript.

Reprints and permissions information is available at www.nature.com/reprints.

The authors declare no competing financial interests.

References

1. Davis MM, et al. T cells as a self-referential, sensory organ. Annu. Rev. Immunol. 2007;25:681–695. [PubMed]
2. Grakoui A, et al. The immunological synapse: a molecular machine controlling cell activation. Science. 1999;285:221–227. [PubMed]
3. Monks CR, Freiberg BA, Kupfer H, Sciaky N, Kupfer A. Three-dimensional segregation of supramolecular activation clusters in T cells. Nature. 1998;395:82–86. [PubMed]
4. Corr M, et al. T cell receptor-MHC class I peptide interactions: affinity, kinetics, and specificity. Science. 1994;265:946–949. [PubMed]
5. Matsui K, Boniface JJ, Steffner P, Reay PA, Davis MM. Kinetics of T-cell receptor binding to peptide/I-Ek complexes: correlation of the dissociation rate with T-cell responsiveness. Proc. Natl Acad. Sci. USA. 1994;91:12862–12866. [PubMed]
6. Krogsgaard M, et al. Evidence that structural rearrangements and/or flexibility during TCR binding can contribute to T cell activation. Mol. Cell. 2003;12:1367–1378. [PubMed]
7. Irvine DJ, Purbhoo MA, Krogsgaard M, Davis MM. Direct observation of ligand recognition by T cells. Nature. 2002;419:845–849. [PubMed]
8. Sykulev Y, Joo M, Vturina I, Tsomides TJ, Eisen HN. Evidence that a single peptide-MHC complex on a target cell can elicit a cytolytic T cell response. Immunity. 1996;4:565–571. [PubMed]
9. Bell GI. Models for the specific adhesion of cells to cells. Science. 1978;200:618–627. [PubMed]
10. Dustin ML, Ferguson LM, Chan PY, Springer TA, Golan DE. Visualization of CD2 interaction with LFA-3 and determination of the two-dimensional dissociation constant for adhesion receptors in a contact area. J. Cell Biol. 1996;132:465–474. [PMC free article] [PubMed]
11. Zhu DM, Dustin ML, Cairo CW, Golan DE. Analysis of two-dimensional dissociation constant of laterally mobile cell adhesion molecules. Biophys. J. 2007;92:1022–1034. [PubMed]
12. Lillemeier BF, et al. TCR and LAT occur in separate membrane domains and concatenate during activation. Nature Immunol. 2010;11:90–96. [PMC free article] [PubMed]
13. Lillemeier BF, Pfeiffer JR, Surviladze Z, Wilson BS, Davis MM. Plasma membrane-associated proteins are clustered into islands attached to the cytoskeleton. Proc. Natl Acad. Sci. USA. 2006;103:18992–18997. [PubMed]
14. Schamel WW, et al. Coexistence of multivalent and monovalent TCRs explains high sensitivity and wide range of response. J. Exp. Med. 2005;202:493–503. [PMC free article] [PubMed]
15. Garcia KC, et al. Structural basis of plasticity in T cell receptor recognition of self peptide-MHC antigen. Science. 1998;279:1166–1172. [PubMed]
16. Wang J, et al. Atomic structure of an αβ T cell receptor (TCR) heterodimer in complex with an anti-TCR Fab fragment derived from a mitogenic antibody. EMBO J. 1998;17:10–26. [PubMed]
17. Kaizuka Y, Douglass AD, Varma R, Dustin ML, Vale RD. Mechanisms for segregating T cell receptor and adhesion molecules during immunological synapse formation in Jurkat T cells. Proc. Natl Acad. Sci. USA. 2007;104:20296–20301. [PubMed]
18. van der Merwe PA, McNamee PN, Davies EA, Barclay AN, Davis SJ. Topology of the CD2–CD48 cell-adhesion molecule complex: implications for antigen recognition by T cells. Curr. Biol. 1995;5:74–84. [PubMed]
19. Choudhuri K, Wiseman D, Brown MH, Gould K, van der Merwe PA. T-cell receptor triggering is critically dependent on the dimensions of its peptide-MHC ligand. Nature. 2005;436:578–582. [PubMed]
20. Xiong Y, Kern P, Chang H, Reinherz E. T cell receptor binding to a pMHCII ligand is kinetically distinct from and independent of CD4. J. Biol. Chem. 2001;276:5659–5667. [PubMed]
21. Li QJ, et al. CD4 enhances T cell sensitivity to antigen by coordinating Lck accumulation at the immunological synapse. Nature Immunol. 2004;5:791–799. [PubMed]
22. Krogsgaard M, et al. Agonist/endogenous peptide-MHC heterodimers drive T cell activation and sensitivity. Nature. 2005;434:238–243. [PubMed]
23. Fricke GM, Thomas JL. Receptor aggregation by intermembrane interactions: a Monte Carlo study. Biophys. Chem. 2006;119:205–211. [PubMed]
24. Weikl TR, Lipowsky R. Pattern formation during T-cell adhesion. Biophys. J. 2004;87:3665–3678. [PubMed]