Courtesy of Michael Dustin
We can do things that haven't been done before, I think, ever in cell biology," exclaims Mark Davis of Stanford University. His 3-D, fluorescence video-microscopy system allows him to count the number of antigen receptors being stimulated on a given T cell, and to follow that cell through time. Davis asks of the single cell, "What do you need in the way of signals to get synapses? And what is the sensitivity of a T cell to antigen?"
The synapse Davis refers to is the immunological synapse, where T cells receive their marching orders; understanding it is vital to appreciating how an immune response is set in motion. "We like to think of the immunological synapse as the brain," says cell biologist Abraham (Avi) Kupfer. "It's like a control center of the activation process."
SIMPLICITY ITSELF In its simplest incarnation, the immunological synapse (IS) consists of two pairs of molecules. Michael Dustin and colleagues at Washington University put freely diffusible MHC-peptide (major histocompatability complex class II plus bound peptide) with ICAM-1 (intracellular adhesion molecule-1) into an artificial lipid bilayer. By adding T cells, the researchers induced T-cell antigen receptor (TCR) and lymphocyte function-associated antigen-1 (LFA-1) to form "kind of a bull's-eye-like pattern," he recalls.
The presence of those molecules is "sufficient to form this pattern," says Dustin, now an associate professor at New York University. But he wants to know what would happen in vivo: "There are probably a lot more interactions going on." In fact, a Web site from Washington University's Synapse Group (now outdated) lists more than 20 molecules that are involved in the IS.1 Many are membrane-bound proteins, such as TCR and MHC, on T cells and antigen presenting cells (APCs). Others are cytoplasmic proteins. Some are found in vesicles, or associated with the cytoskeleton.
GIVE IT A SMAC Kupfer dubbed the bull's-eye structure a "supramolecular activation cluster," or SMAC. The center is termed the cSMAC; the surrounding ring is the pSMAC.2 They "are not static structures, but highly dynamic," says Kupfer, of the National Jewish Medical and Research Center. "Their composition changes in a highly orchestrated, four-dimensional manner--namely time and space." One of the most dramatic changes that happens within the first half-hour of contact is a "very dynamic inversion, with movement of the TCRs to the center, with the formation of this ring of adhesion molecules," Dustin explains.
"We were surprised to find that the TCR started out in the outside ring," says Andrey Shaw, one of Dustin's collaborators, noting that Kupfer's original work, which used a mixture of T cells and APCs, spun together and then fixed, had the TCR at the IS's center. The easiest way to explain the inversion, Shaw says, is that the TCR first engages the MHC-peptide on the immature synapse's periphery and is then drawn into the center. Davis, Shaw, Dustin, and Washington University's Paul Allen proposed a model in which an initial LFA-1/ICAM-1 interaction slows the T cell long enough for the TCR to sample the MHC-peptide. If it recognizes its cognate antigen, the TCR would signal the T cell to stop, and would then move toward the synaptic center.3
Each T cell has a single TCR specificity, which it uses to recognize a specific antigen. Once the TCR is engaged, the T cell must determine the best course of action. That process requires complicated sets of signals (conveying, for example, information such as the nature, abundance, and distribution of an offending bacteria) to be transmitted back and forth, creating what Davis calls a dialogue between those cells. "It actually helps the cells to decide, early on ... whether they should mount a productive response, or abort, or even commit suicide by apoptosis," Kupfer says, emphasizing that not every encounter results in a productive response.
Courtesy of Darrell Irvine
SIGNAL FIRST, SYNAPSE LATER The Dustin, Allen, and Shaw groups have shown that signaling through the TCR actually precedes IS formation.4 By the time Kupfer's mature synapse appears, signaling through the TCR seems to have all but disappeared. "Central clustering of the TCR generated by the immunological synapse formation does not function to initiate or sustain TCR signaling," the paper emphasizes. "These observations challenge current ideas about the role of immunological synapses in T-cell activation."
One possibility, which they do not advocate, "is that the thing maybe acts like a gasket," Dustin says. "It provides a molecular basis for this idea that ... materials are secreted or released by the T cell into the junction." (This concept reiterates the original analogy to the neural synapse. In the late 1980s, Kupfer and others showed that a T cell would reorient its secretory machinery toward its target, and secrete soluble vesicles into the junction between it and an APC, such as a B cell. But, Kupfer notes, "Directed secretion actually takes place several hours after the initiation of the contact, because the B cells and T cells need to first produce the cytokines.")
The paper speculates that the IS is involved in TCR downregulation and endocytosis, perhaps providing a platform for attenuating the TCR signaling by means of other receptors. Shaw puts it this way: "If the synapse is not required for [TCR] signaling, and is not required for activation, then actually the synapse is required for deactivation, and turning off signaling, and getting TCRs out of there."
However, that work, Shaw says, is open to misinterpretation. "You have to be careful, because the systems are highly optimized. Some of the conclusions aren't necessarily relevant in vivo," he notes. "My gut feeling," he continues, "is that in a real physiological system, when we're looking at real antigens that are at very limiting concentrations in the body, the synapse will play a very important role in potentiating signaling." Shaw recently submitted a paper claiming that TCR signaling does occur in the cSMAC, but that the receptors are internalized and degraded too quickly to be noticed.
Kupfer notes that there are multiple stages and checkpoints in T-cell activation that are critical for its decision-making. This, he explains, "provides a mechanism for a really high sensitivity to antigen, but also allows for high selectivity or specificity." He recently proposed "that new TCR-induced signals generated after the formation of SMACs would be required to trigger the productive activation of T cells by APCs."5 As the response progresses, "new components replace older molecular components," he says.
Courtesy of Michael Dustin
GOING HOLLYWOOD Meanwhile, Davis and his Stanford colleagues have been making 3-D movies of the activation process.6 Davis laments that the field had been subjected to a "tyranny of biochemistry" that tends to draw conclusions based on the average of what the population is doing, prompting Davis to ask his questions about individual interactions.
They labeled peptides and used fluorescent microscopy to detect the exact number of MHC-peptides one T cell encounters. By tracking the cell's calcium flux (one of the prime responses to signaling) and the movement of proteins involved in the IS, they found that a single MHC-peptide was enough for the T cell to trigger a calcium response and to stop. It took about 10 peptides to make a synapse.
The minimal signaling required to activate the T cell is not clear, Davis says. "You probably don't need a complete synapse, but you probably need pieces of the synapse. And what those pieces are exactly is not clear. And what the synapse is ultimately doing in the structure we know of is not clear."
Others are questioning the very notion of an IS. Mattias Gunzer and Stephan Grabbe, University of Munster, Germany, postulated that T cells are constantly attaching to, crawling on, and detaching from APCs, with the duration of contact averaging 6-12 minutes when the cells are observed interacting on a collagen matrix. They write: "These dynamic and short-lived encounters favor sequential contacts with the same or other [dendritic cells] and trigger calcium influx, upregulation of activation markers, T-blast formation, and proliferation."7
Dustin, Allen, and Shaw counter that although collagen may make T cells behave this way, the cells are unlikely to encounter collagen when interacting with APC in the T-cell area of the lymph node and are unlikely to behave that way at that primary site of naïve T-cell activation. They write: "By contrast, T cells in the dermis and other solid tissues would always be exposed to abundant collagen, and thus would interact in a serial encounter mode that might be ideal for T-cell effector functions."8
Josh P. Roberts (firstname.lastname@example.org) is a freelance writer in Minneapolis, Minn.
1. "Immunological synapse fact book," Washington University, St. Louis; available online at http://pathbox.wustl.edu/~synapse/
2. C.R. Monks et al., "Three-dimensional segregation of supramolecular activation clusters in T cells," Nature, 395:82-6, 1998.
3. A. Grakoui et al., "The immunological synapse: A molecular machine controlling T cell activation," Science, 285:221-7, 1999.
4. K.H. Lee et al., "T cell receptor signaling precedes immunological synapse formation," Science, 295:1539-42, 2002.
5. B.A. Freiberg et al., "Staging and resetting T cell activation in SMACs," Nat Immun, 3:911-7, 2002.
6. D.J. Irvine et al., "Direct observation of ligand recognition by T cells," Nature, 419:845-9, 2002.
7. M. Gunzer et al., "Antigen presentation in extracellular matrix: Interactions of T cells with dendritic cells are dynamic, short lived, and sequential," Immunity, 13:323-32, 2002.
8. M.L. Dustin et al. "Environmental control of immunological synapse formation and duration," Trends Immunol, 22:192-4, 2001.