CTLA-4: acting at the synapse.

Successful immune cell control requires a delicate balance of positive and negative regulatory signals. Costimulatory pathways involving molecules such as CD28, inducible costimulator (ICOS), 4-1BB, and CD40L are essential coactivators of proliferation, cytokine production, and cell migration. To balance these signals, cell surface molecules like Fas, tumor necrosis factorreceptor (TNFR), and programmed death-1 (PD-1) decrease T cell responses. Most prominent among these T cell regulators is cytotoxic T lymphocyte antigen-4 (CTLA-4), a homolog of CD28 that can interact with the CD28 ligands CD80 and CD86 (1). CTLA-4 inhibits T cell responses in a T cell receptor(TCR)dependent manner. For example, T cells treated with soluble CTLA-4 monoclonal antibody (mAb) have enhanced T cell proliferation and cytokine production, whereas surfaceimmobilized CTLA-4 mAb inhibits interleukin-2 (IL-2) production and cell cycle progression both in vitro and in vivo (2, 3). In addition, CTLA-4–deficient mice exhibit a CD4+ T cell lymphoproliferative disorder that leads to death of the animals a few weeks after birth (4, 5). Several models have been proposed to explain the molecular basis for CTLA-4 inhibition (Figure 1A). The first model proposes that CTLA-4 successfully competes against CD28 for CD80 or CD86 binding because CTLA-4 binds to them with much stronger avidity than does the CD28 costimulatory molecule (Ligand Competition Model) (6). Support for this model comes from studies showing that mutant CTLA-4 lacking a functional intracytoplasmic domain can suppress T cell function both in vitro (7), and in vivo (8). A second model suggests that CTLA-4 inhibits important downstream signaling pathways involving the extracellular signalregulated kinase (ERK) and nuclear factor-kappa B (NFB) pathways (9, 10), subsequent to TCR-CD3 association and CD28 receptor engagement (Distal Blockade Model). In this model, CTLA4-mediated inhibition depends on, but is downstream of, the TCR signal. Finally, a third model suggests that CTLA-4 associates with the immunological synapse and attenuates proximal signal initiated by the TCR (Proximal Blockade Model) (11, 12). The immunological synapse is formed at the interface between the T cell and antigen-presenting cell (APC) membranes and is the hot spot for T cell activation. Upon T-cell–APC engagement, critical components of the TCR signal transduction machinery relocalize to this region initiating a biochemical cascade that leads to full T cell activation. In many instances, the interface forms highly organized scaffolds that create an immunological synapse (13, 14). Several early studies suggested that upon Tcell–APC engagement, CTLA-4 is exported to the cell surface at the site of cell–cell interaction, where CTLA-4 can regulate TCRproximal signals (15, 16). These studies have been confirmed and extended by the work of Egen and Allison who demonstrate that CTLA-4 is at first localized in the uropod of activated T cells, where the microtubule organizing center (MTOC) is located (11). When T cells are restimulated with weak-agonist peptide-bearing APCs, CTLA-4–containing vesicles in the T cells rapidly relocalize just beneath the APC–T cell contact site. By comparison, stimulation with a strong agonist peptide induces translocation of CTLA-4 to the cell surface and association with the immunological synapse. At least two mechanisms regulate the cell surface expression of CTLA-4: tyrosine phosphorylation-dependent clathrin-mediated internalization (17–21), and the active release of the molecule from the intracellular compartment to the cell surface (22, 23). CTLA-4 has a tyrosine-based FVYVKM motif within its cytoplasmic tail that strongly interacts—in the dephosphorylated state—with the clathrin adaptor molecules AP1 and AP2. This interaction results in rapid internalization and trafficking of CTLA4 to intracellular compartments (16–21). Many kinases can phosphorylate this tyrosine-based motif in vitro, and therefore prevent CTLA-4 from associating with clathrin and becoming internalized. The TCR-associated Src family kinases, Lck and Fyn, are the most likely tyrosine kinases involved, considering their localization within the immunological synapse (24, 25). This hypothesis is also consistent with observations showing that active translocation of CTLA-4 depends on TCR signal and is thus ZAP70– and Lck-dependent (22), because Lck activation causes surface stabilization of CTLA-4 by direct phosphorylation of a regulatory tyrosine within the CTLA-4 tail (24, 25). Thus, the TCR signal might influence CTLA-4 expression by both translocation and stabilization, although whether this active transport depends on lysosome secretion or other mechanisms remains unclear (23). CTLA-4 has been observed at the immunological synapse; however, there has been no direct biochemical basis to explain its inhibitory activity. CTLA-4 can directly associate with the TCR chain of the TCR–CD3 complex, and subsequent dephosphorylation of the chain has been observed (12). These observations provided the basis for a model suggesting that the negative regulatory activity of this cell surface protein depends on the association of a tyrosine phosphatase because CTLA-4 lacks known intrinsic enzymatic activity. In this regard, two protein tyrosine phosphatases—Src homology 2 (SH2) domain–containing tyrosine phosphatase-1 (SHP-1) and SHP-2—can associate with the cytoplasmic tail of CTLA-4 (19, 26, 27); however, the role of these enzymes in the function of CTLA-4 is controversial because SHP-2 can positively regulate cell activities (28), and CTLA-4, in some cellular contexts, negatively regulates signals in the absence of SHP-1 (29). PP2A, a protein serine-threonine phosphatase can associate with CTLA-4 (30). This protein phosphatase is known to negatively regulate the Ras-ERK pathway in some systems (31). This finding also fits with the localization of CTLA-4 to the immunological synapse, because Ras is activated within the Viewpoint