Costimulation: Building an Immunological Synapse

See allHide authors and affiliations

Science  29 Jan 1999:
Vol. 283, Issue 5402, pp. 649-650
DOI: 10.1126/science.283.5402.649

A central event in the development of immunity is the activation of the T cell. At the center of this process is the T cell receptor (TCR), which triggers activation by a specific interaction with antigen [usually a foreign peptide bound to, or “presented by,” the organism's own major histocompatibility complex (MHC) molecule on the surface of an antigen-presenting cell]. Because of the small size of the TCR, its low affinity toward antigen, and the limited numbers of antigens on the antigen-presenting cell, an elaborate adhesion complex must be formed to allow the TCR to contact, sample, and then be activated by the rare antigenic ligands (1). This specialized contact area has been termed the immunological synapse (2, 3). Reports by Wülfing and Davis in a recent issue (4) and Viola et al. (5) on page 680 of this issue focus attention on how the immune synapse is built and unify what had been thought of as two distinct signals needed for efficient T cell activation.

Efficient T cell activation requires engagement of at least two types of T cell surface receptors. This phenomenon has been interpreted in terms of a “two-signal model,” which proposes that T cell activation requires one signal from the TCR and a second signal from a “costimulator” molecule. Although many molecules have been implicated as costimulators, CD28 has become the archetype for costimulatory molecules. Engagement of CD28 either by its ligand on the antigen-presenting cell [B7 (CD80)] or by antibody can strongly enhance TCR signaling responses. Although current models suggest that CD28 functions as a specific activator of the Jun kinase JNK or the nuclear transcription factor NF-κB, CD28 engagement by itself is not sufficient. Activation of either JNK or NF-κB always requires coengagement of the TCR. This has led to a counterproposal that costimulation might function to amplify the signals transduced by the TCR (1).

In a recent issue of Science, Wülfing and Davis (4) demonstrated a novel mechanism for costimulation in formation of the immune synapse. They demonstrate that costimulation initiates active directional transport of protein and lipid domains to the area of cell-cell contact. This transport process requires myosin and correlates with enhanced, as well as sustained, signaling—a hallmark of costimulation. In these experiments, directed transport could be stimulated by either CD28 or LFA-1 engagement, but occurred most efficiently when both were engaged together. Wülfing and Davis propose that costimulation works by activating an actin-myosin-driven transport process that delivers receptors and signaling complexes to the contact area. In this study, however, the transport process appears to be indiscriminate, and the key cargo was not clearly identified.

In this issue of Science, Viola et al. (5) show that the cargo for the actin-based transport mechanism is the 70-nm-diameter lipid rafts also referred to as caveolae or detergent-insoluble glycolipid domains (6). The rafts are initially distributed evenly on the T cell surface and remain so after engagement of TCRs by beads coated with antibody to the TCR. Remarkably, engagement of CD28 together with the TCR recruits essentially all the rafts to the contact area. This correlates with an increased lifetime for tyrosine phosphate, which may occur through phosphatase exclusion, and increased consumption of the Lck kinase, indicative of greater tyrosine kinase activation. It has been suggested recently that engaged TCRs migrate into rafts (7). Viola et al. now demonstrate that it is the rafts that migrate to engaged TCRs and CD28.

These studies suggest that costimulation modulates the signaling environment around the engaged TCRs. Rafts are rich in kinases and adapter molecules that are required for T cell activation (8). In addition, the rafts' topological features are also compatible with their promoting sustained TCR engagement. Because glycolipids and small glycophosphatidylinositol-anchored molecules such as CD59, DAF, alkaline phosphatase, and Thy-1 are concentrated in rafts, these domains may represent regions of reduced steric hindrance where interaction of the short TCR and MHC would be favored. In addition, cholesterol in lipid rafts may increase membrane rigidity and enhance the affinity of membrane protein interactions.

The costimulation-initiated transport mechanism appears capable of transporting anything linked to actin. Because both positive and negative regulators of T cell signaling may be associated with the actin cytoskeleton (9), how does the process achieve selectivity? One type of selectivity is demonstrated in the extreme by the movies of Wülfing and Davis (10): size selectivity. Large beads are transported to the edge of the synapse, but are excluded because they are too big to enter. On a molecular scale, integrins, the group of adhesion molecules that includes LFA-1, can generate effective occlusive barriers that exclude large molecules from contact areas (11). We and others have proposed that molecules such as CD2 that interact with ligands to generate very small gaps (<15 nm) between apposed membrane are also involved in large-molecule exclusion (1, 12). If the actin-based transport process can convey molecules to the center of the immunological synapse, then these barriers could be conceived of as molecular filters allowing only small molecules to enter the contact area, while excluding molecules with larger ectodomains (see the figure).

Topological anatomy of the immune synapse.

(Left) The pattern of LFA-1 and CD2 engagement in an activated T cell contact with an artificial membrane containing CD58 (blue) and ICAM-1 (red) (3). Arrowhead indicates the area of engagement of respective ligands. Arrow indicates exclusion of 20-nm-diameter ICAM-1 from the 15-nm contact formed by interaction of CD2 and CD58. (Right) The position of sequential molecular filters (40 and 15 nm) that are encountered by complexes such as rafts as they are transported toward the center of the immune synapse. The TCR clusters in the central 15-nm region (13).

The conventional view of T cell signaling was that each type of receptor generates its own distinct signal or has its own “voice.” This collection of independent voices from the surface was then harmonized (integrated) in the nucleus to regulate transcription. The new concept that is emerging suggests that the immune synapse functions to tune, adjust, and amplify a single voice, the signal transduced by the TCR. TCR signaling is intimately associated with contact formation because extended cell contact is required to maintain TCR engagement. This new concept is supported by recent studies of molecular organization of components in the immunological synapse and by the demonstration that specific transport mechanisms organize the contact area. Although we do not yet know whether CD28 and LFA-1 produce specific biochemical signals to initiate this transport process, a new paradigm for immunological costimulation is emerging that is built around the central role of contact formation in T cell activation.


View Abstract

Stay Connected to Science

Navigate This Article