Using the force to find the peptides you’re looking for

T cells are an essential cell type of our adaptive immune system, helping to both detect and eliminate pathogens that may have infected our bodies. This essential function is mediated by the T cell antigen receptor complex (TCR) expressed at the plasma membrane, which can interrogate the intracellular state of host cells by scanning for pathogen-derived peptides presented on the surface of host cells. For us to remain healthy, T cells must be able to respond extremely robustly to ensure a resolution of the diseased state, even when there are only a few TCR ligands expressed on the infected cells. In PNAS, Feng et al. (1) show that exertion of force on the TCR during ligand binding is required for the extraordinary potency toward ligand binding observed in vivo.

The ligand for the TCR is a pathogen-derived peptide bound within the MHC protein (pMHC). In contrast to receptors that bind soluble ligands, the pMHC ligand is a membrane protein expressed at the surface of antigen-presenting cells. This inherently leads to tension, or a pulling force, being generated between the two cells upon TCR/pMHC engagement. This is compounded by the relatively small height of the receptor complex (∼13 nm) compared with the apposition of two cells (∼100 nm) normally restricted by the glycocalyx surrounding cells, driving the two membranes away from their natural resting state.

The forces typically found for biologically relevant processes are in the low pico-Newton (pN; 10⁻¹² N) range. Experimentally manipulating proteins with forces of this magnitude can be achieved using a variety of biophysical methods (2, 3). In the current study, Feng et al. (1) used an optical trap (4), also known as tweezers, to apply a force on the TCR complex in a direct and quantitative manner. This was achieved by trapping a polystyrene bead, functionalized with the pMHC ligand for a TCR of known specificity, in the center of a tightly focused laser beam (Fig. 1A), which provides a sufficient spatial gradient of light intensity to overcome the bead movements driven by thermal Brownian motion. This bead can then be brought in close enough proximity to the T cell surface to precipitate TCR/pMHC interactions to drive cellular activation (Fig. 1B). This work is a continuation of the long-standing collaboration between the Lang and Reinherz groups, who have used this approach previously to measure the force and directionality requirements of TCR triggering (5–7). While measuring the forces created during TCR triggering is useful in its own right, the power of the trapping approach is manifest when combined with a measured cellular output of the force-induced input. In this case, the increased concentration of intracellular Ca²⁺ ions observed on T cell activation was concomitantly measured using a fluorescent indicator, a technique that Lang and coworkers (7, 8) have developed previously. A crucial part of this new work is the rigorous quantification the authors have used to accurately measure the pMHC density on the beads used to activate the T cells, through calibrated flow cytometry and single-molecule total internal reflection fluorescence imaging.

Fig. 1.

Force exertion on the TCR drives T cell activation. (A) Side-view representation of the experimental setup. The infrared laser beam is sufficiently focused to create an optical trap, which can immobilize a polystyrene bead functionalized with pMHC ligands and brought into contact with a T cell through precise motions of a piezo translation stage. Arrows show approximate force vectors in the normal and shear modes. (B) Schematic of initial TCR engagement by pMHC bound to a bead (streptavidin omitted). The spatial force gradient of the optical trap is represented by concentric dashed circles, and the actin cytoskeleton underlying the plasma membrane that maintains membrane stiffness is shown. (C) Rapid movement of the stage (ΔX) exerts (shear) force on the TCR, inducing receptor triggering and the recruitment of adaptor proteins, which link the TCR to myosin motor proteins that “step” along actin. (D) T cell activation, measured by increased intracellular Ca²⁺ concentration, is observed after the bead has returned to the trap center.

The first part of the current study shows that in the absence of an applied force, T cells only respond when presented with beads bearing a high density of pMHC ligands, a concentration that is unlikely to be achievable physiologically. However, by exerting a force on the bound TCR, it is now possible to observe T cell activation with only a few pMHC molecules per bead. To generate this force on the receptor, the trapped bead first engages the T cell under low force, and the cell is then rapidly displaced (<150 ms) using a piezo translation stage (Fig. 1C). As the optical trap can be approximated by a simple spring attached to the trapped bead with a stiffness previously determined (∼0.1 pN/nm in this study), forces up to 30 pN can be applied in a defined orientation by displacing the cell a known distance (∼0–300 nm).

This important result extends the conclusions from previous work by these groups (6, 7), and the quantitative approach provides strong evidence that the application of force on the TCR is essential for T cell activation at physiological pMHC densities. Davis and coworkers (9, 10) have previously shown that T cell activation can be driven by potentially even a single pMHC ligand when presented on a bone fide antigen-presenting cell. Given this latter result, this does indeed imply that the “bridging” between the two cells caused by TCR/pMHC engagement should be able to exert force at an equivalent level to that applied by the optical trap. Feng et al. (1) then go on to demonstrate that applying shear force (in the plane of the plasma membrane) is generally more efficient than when exerted in the normal direction (perpendicular to membrane), principally using beads with an average of 29 pMHC.

It should be noted that the two modes of pulling on the TCR present quite different insults to the plasma membrane and underlying cytoskeleton, making it harder to directly compare their effects. In shear mode, the bound TCR is essentially “dragged” back through the plasma membrane as the bead returns to the trap center, potentially without distorting the membrane structure. Conversely, applying a load in the normal direction must create a significant membrane protrusion as the bead returns toward the trap center, which would inherently alter membrane bending and the underlying cytoskeletal structure. These deformations may also drive segregation of large phosphatases such as CD45 that could enhance signaling through the kinetic-segregation model (11, 12). Indeed, although Feng et al. (1) suggest that a shear force is the most efficient way to activate the T cells (in terms of number of binding events required), it was interesting that the most robust receptor triggering using one to two ligands was always observed when pulling in the normal direction, where every cell measured was now found to respond. Following their rationale, a single bound receptor loaded with all of the exerted force might be as effective as pulling in the normal direction; perhaps this might be the most likely scenario in situ when pMHC density is low and independent force-generating regions are formed at the T cell surface.

The remainder of the study focuses on the shear mode of force-induced activation. Feng et al. (1) find that the rate at which the bead returns to the trap center after applying load correlates with cell activation, with triggered cells associated with faster relaxation times. Very intriguingly, there was clear evidence of abrupt steps in these temporal traces of bead displacement. The authors showed these steps were lost when the actin cytoskeleton was disrupted with Cytochalasin D but were unaffected when microtubules were disrupted with Nocodazole. They also found that Blebbistatin, a drug that inhibits myosin II, had a similar effect to actin disruption. This suggests that these motor proteins could be responsible for the active transport observed (Fig. 1D), although as the authors also point out, an alternative explanation is the loss of stabilizing cross-links within the actin cytoskeleton that are also dependent on myosin II proteins (13). It would be interesting to know if the observed stepping behavior remained even with a TCR complex devoid of intracellular sequences so that it could no longer directly interact with motor proteins, something that should be possible in the current system because all of the components of the TCR complex were transduced into the TCR-negative cell line used.

The stepped nature of the relaxation traces leads to the final result presented in the study. Feng et al. (1) used an experiment with a second, nontrapped bead to provide evidence that even unbound TCRs can be co-opted into an immune synapse, which they speculate is also driven by motor-driven coupling in the absence of direct pMHC binding. They use this data to argue against the serial triggering hypothesis (14), which posits that a single pMHC can sequentially trigger multiple TCRs. This is indeed a very intriguing and provocative experiment, and more work will be required to square their conclusion with the evidence supporting serial triggering (15), including why high-affinity pMHC interactions can sometimes paradoxically drive decreased downstream cell activation (16).

In summary, the Lang and Reinherz groups have demonstrated that a single TCR is likely to have a force of ∼10 pN exerted on it when it is engaged by a cognate pMHC ligand (5–7). Feng et al. (1) present one of the first quantitative investigations into the relationship between ligand density and force generation, and should be applauded for their rigorous approach to tackling this problem. Through an alternative strategy, using short DNA duplex sequences with defined rupture forces, Salaita and coworkers (17) have also measured the force dependence of TCR/pMHC engagement, finding forces in the range of 12–19 pN for efficient T cell activation, agreeing well with the present result. It is worth pointing out that in all of the studies so far addressing this problem, the pMHC is bound to a solid support, either a bead or glass surface. Clearly, the native pMHC-presenting cell will not have equivalent properties, and far less attention is given to how force generation on the “other side” is applied. Perhaps TCR engagement with the underlying cytoskeleton may relatively stiffen the T cell surface, something that pMHC may not be able to do in its own membrane. How this force application affects peptide discrimination by the TCR also remains unresolved.