E3 Ligases in T Cell Anergy—Turning Immune Responses into Tolerance

See allHide authors and affiliations

Science's STKE  13 Jul 2004:
Vol. 2004, Issue 241, pp. pe29
DOI: 10.1126/stke.2412004pe29


Peripheral tolerance is an important strategy used by the immune system to prevent self-reactive lymphocytes from attacking host tissues. A variety of mechanisms contribute to peripheral tolerance, among them activation-induced cell death, suppression by regulatory T cells, and T cell anergy or unresponsiveness. Recent work has led to a better understanding of the cell-intrinsic program that establishes T cell anergy. A major insight is that during the induction phase of anergy, incomplete stimulation (T cell receptor stimulation without costimulation) leads via calcium influx to an altered gene expression program that includes up-regulation of several E3 ubiquitin ligases. When the anergic T cells contact antigen-presenting cells, intracellular signaling proteins are monoubiquitinated and targeted for lysosomal degradation, thus decreasing intracellular signaling and also resulting in decreased stability of the T cell–antigen-presenting cell contact. We propose a molecular program leading to T cell anergy and discuss other proteins that may play a role.

Multicellular organisms have evolved elaborate immune defenses to protect themselves against pathogens—organisms lower on the evolutionary scale that infect or parasitize the host. In the absence of infection, the immune cells responsible for recognizing pathogens must be held in check and blocked from attacking host tissues. In consequence, the immune system, more so than other organs, incorporates a multilayered system of checks and balances designed to prevent destruction of self while still supporting swift and decisive immune responses to foreign antigens. These include the elimination of self-reactive B and T cells during development, development of unresponsiveness ("anergy") in mature self-reactive cells that have escaped or survived negative selection, and suppression of overactive immune responses by specific populations of regulatory T cells (1, 2).

In this Perspective, we focus on T cell anergy, especially the transcriptional and signaling mechanisms involved. We describe the evidence that at least one form of T cell anergy involves a broad-based, calcium- and calcineurin-dependent program implemented at several levels (Fig. 1). The program begins with transcriptional up-regulation of a large number of anergy-associated genes (35) and ends with localized degradation of specific signaling proteins (6). We discuss one of the most striking features of the program, the involvement of ubiquitinating enzymes that participate in the pathway of endosomal-lysosomal trafficking and lysosomal protein degradation (4, 6). Finally, we consider how a greater understanding of the biochemical mechanisms of T cell anergy may lead to improved methods of either inducing or breaking T cell tolerance.

Fig. 1.

Model depicting the calcium- and calcineurin-dependent program of T cell anergy and its consequences for signal transduction. (A to C) A responsive T cell in which TCR engagement in the presence of costimulation produces a calcium signal (light red) and stimulates an immune response, represented by interleukin 2 (IL-2) production. (D to F) A T cell that has been rendered anergic by sustained calcium signaling (light red) as a result of TCR stimulation in the absence of costimulation. During the calcium-dependent phase of anergy induction (D), the E3 ligases Itch, Cbl-b, and Grail are transcriptionally up-regulated. In (B) and (E), the T cells engage in contact with antigen-presenting cells (APCs). T cell contacts are mediated by LFA-1 proteins (purple) and TCRs (orange). Gray vesicles represent endocytosis at the contact area and vesicle trafficking. In the anergic T cell (E), Cbl-b, Itch, and the related protein Nedd4 move to the membrane attracted by an unknown factor (red), as a result of which the TCR and associated signaling proteins are monoubiquitinated and rerouted into the pathway of lysosomal degradation. In (F), degradation of PLC-γ1 and PKCθ diminishes overall signal transduction through the TCR. Because PLC-γ1 and PKCθ signaling is required to maintain LFA-1 affinity, the immunological synapse disintegrates and the strength of T cell–APC contact is lost. This abbreviates the duration of T cell–APC contact and further attenuates the response.

Costimulation and T Cell Tolerance

What determines whether a T cell becomes anergic? In the simplest model, anergy is imposed by an encounter with antigen in the absence of input from costimulatory receptors, such as CD28 (7). The level of costimulation provided by dendritic cells and other antigen-presenting cells (APCs) depends on signaling through Toll receptors, which sense the presence of pathogens by recognizing conserved molecular structures associated with microorganisms (8). Lack of costimulation, which indicates to self-reactive T cells that there is no infectious or other "danger" to the organism, appears to induce T cell anergy. For instance, anergy can be induced in vivo by treatment with superantigens or with soluble high-dose antigens that stimulate the T cell receptor (TCR), but not costimulatory receptors (7). T cells can also be rendered anergic by exposure to tolerogenic dendritic cells, which typically express major histocompatibility complex (MHC) proteins but bear only low levels of costimulatory ligands (9, 10). In a mouse model of human schistosomiasis, the initial proinflammatory T helper cell TH1-type host response to Schistosoma mansoni egg antigens is curtailed by the development of T cell anergy; this correlates with loss of B7 costimulatory ligands from APCs while the expression of TCR ligands (class II MHC proteins) is maintained (11). A caveat is that B7 proteins are ligands not only for the costimulatory receptor CD28, which provides a positive signal for T cell activation, but also for a counterreceptor, CTLA4, which provides a late negative signal (12). As such, B7 proteins contribute not only to immunogenic signaling but also to the induction of tolerance (13). Furthermore, signaling through costimulatory receptors is required to maintain the function of regulatory T cells (8, 9). Because the role of costimulation in T cell tolerance is complex and not yet understood at a molecular level, we confine our discussion of T cell anergy to the simple dichotomy of TCR stimulation with or without costimulation.

NFAT and the Calcium-Induced Anergy Program

One of the negative regulatory loops that induces T cell anergy involves activation of the transcription factor NFAT, under conditions where its transcriptional partner AP-1 (comprising Fos and Jun) is not concomitantly activated (3, 7). In T cells stimulated through both the TCR and costimulatory receptors, NFAT cooperates in the nucleus with AP-1 (Fos-Jun) proteins to stimulate the expression of a large number of genes essential to the productive immune response (14). In contrast, anergy-inducing stimuli—TCR stimulation without costimulation or prolonged exposure to the Ca2+ ionophore ionomycin—evoke low, sustained elevations of intracellular Ca2+ concentration that support sustained activation of NFAT, but not of AP-1 or nuclear factor κB (NF-κB) (3, 7). Anergized T cells express a novel set of anergy-associated genes distinct from those that are characteristic of the productive immune response. The expression of these same anergy-associated genes is also increased in vivo in T cells from orally tolerized mice (3). Anergy induction is blocked by the calcineurin inhibitor cyclosporin A (CsA), which inhibits NFAT activation (3, 7); conversely, constitutively active NFAT can induce T cell anergy in the absence of AP-1 (3). Thus, NFAT is capable of regulating two contrasting aspects of T cell function: productive activation in the presence of AP-1, but also a nonoverlapping anergy program in the absence of AP-1.

The dual role of NFAT in T cell activation and anergy explains the finding that in transplant patients, interference with costimulatory pathways leads to tolerance induction, but paradoxically, the process of tolerance induction is blocked by CsA (15). It also explains the unexpected phenotype—increased responsiveness rather than immune deficiency—of mice lacking the predominant family member NFAT1 (16, 17). Presumably, other NFAT proteins can compensate for the activating effects of NFAT1, but not for its negative role.

E3 Ligases as Targets of Calcium-Induced Anergy

Among the targets of the calcium- and calcineurin-dependent pathway of anergy induction are three E3 ligases, Itch, Cbl-b, and Grail (4, 6). Itch is a HECT-domain E3 ligase that has catalytic function (that is, is capable of forming thioester bonds with ubiquitin); Cbl-b is a RING finger–containing adaptor E3 ligase; and Grail is a RING finger–containing transmembrane E3 ligase that localizes to endosomes and other vesicular structures in cells (4). All three are transcriptionally up-regulated during the induction phase of T cell anergy and are likely to participate functionally in the effector phase. Cbl-b and the related protein Cbl are clearly involved in the process of TCR down-regulation (18), and they also modulate the function of various other signaling proteins, including ZAP70, phosphoinositide 3-kinase (PI3K), and phospholipase C–γ1 (PLC-γ1) (18, 19). When anergic T cells are restimulated after anergy induction, Itch and the closely related HECT-domain E3 ligase Nedd4 become localized to detergent-insoluble membranes, where they target essential signaling proteins, including PLC-γ1 and protein kinase θ (PKCθ) for degradation, thereby diminishing calcium mobilization and subsequent TCR signaling (6) (Fig. 1). Because signaling is required to maintain T cell–APC contact (20), this delivers a double whammy, attenuating the responses of the anergic T cell not only because signaling proteins are degraded, but also because the overall duration of APC contact is markedly abbreviated (6). That is, through the action of E3 ligases, the anergy program elicits only a transient burst of signal transduction that is not sustained enough to trigger productive activation in the anergic T cells. Thus, a checkpoint is established in which the magnitude of the T cell response is determined by the duration of signaling from activated receptor complexes.

Consistent with roles for Itch and Cbl-b in T cell anergy, mice deficient in either protein are prone to autoimmune disease and their T cells are hyperactivated (19, 2123). Grail-deficient mice are not yet available, but retroviral overexpression of Grail blocks interleukin-2 (IL-2) and IL-4 production by T cells (4).

Anergy and the Endosomal-Lysosomal Pathway of Protein Degradation

We suggest that the anergy program involves internalization and monoubiquitination of the TCR and associated signaling proteins, followed by endosomal sorting of the monoubiquitinated proteins for lysosomal degradation (Fig. 1). This is based on our findings that PLC-γ1 degradation in anergic T cells is not inhibited by a proteasome inhibitor and that anergic T cells display monoubiquitinated PKCθ, which would not be recognized by the proteasome (6). A very similar program of monoubiquitination-dependent protein degradation is involved in termination of signaling through the epidermal growth factor (EGF) receptor (EGFR) (2426), a receptor tyrosine kinase. EGF binding to the EGFR stimulates receptor internalization, and EGF-dependent monoubiquitination of the EGFR by the cytoplasmic RING finger E3 ligase Cbl (a homolog of Cbl-b) triggers a sorting event on the surface of the early endosome. Sorting involves recognition of the monoubiquitinated receptors by proteins, such as Hrs and Tsg101, that contain ubiquitin-binding domains; transport of the monoubiquitinated receptors to invaginations that form on the membrane of the endosome; and eventual budding off of vesicles into the lumen of the endosome (now termed a multivesicular body). The monoubiquitinated proteins that are sorted into the internal vesicles are delivered for lysosomal degradation, whereas the unsorted (nonubiquitinated) proteins that remain on the endosomal membrane are recycled back to the plasma membrane.

Notably, the E3 ligases implicated in T cell anergy have counterparts involved in endosomal-lysosomal degradation in other systems. Rsp5, the yeast homolog of the mammalian HECT-domain E3 ligase Nedd4, monoubiquitinates several plasma membrane proteins, either by recognizing a PPXY motif or by binding to the adaptor molecule Bsd2 (27, 28), a homolog of N4WBP5. The expression of N4WBP5 mRNA and protein is induced in stimulated T cells in a CsA-sensitive manner (29). Similarly, the HECT-domain E3 ligase AIP4, the human homolog of murine Itch, monoubiquitinates ligand-activated CXCR4 (a chemokine receptor) and induces its sorting to the lysosomal degradation pathway through interactions with Hrs and Vps4 proteins (30). In yeast, the RING finger–containing E3 ligase Tul1, which contains several transmembrane domains, targets misfolded proteins for degradation and is essential for the formation of multivesicular bodies that originate in the Golgi (31). Tul1 does not have a homolog in metazoans; however, as a transmembrane RING finger E3 ligase that localizes to intracellular vesicles, it resembles the transmembrane E3 ligase Grail, the expression of which is increased at both mRNA and protein levels during anergy induction in T cells (4, 32).

Physiologic T cell activation most likely involves only a few TCR-peptide-MHC complexes that are nonetheless able to trigger a productive response (33). The study of protein regulation through monoubiquitination promises a new view of signal transduction in which signal intensity is modulated by the cellular localization, activation status, and half-life of individual proteins within specific signaling complexes. Targeting for lysosomal degradation of either the activated receptor itself or interacting adaptor proteins and signaling enzymes is an ideal mechanism for rendering T cells—whether naïve, mature, anergic, or chronically stimulated—less responsive or unresponsive to stimulation. One should not forget the emerging evidence, however, that endocytosis has functions in cellular activation that go beyond simple sequestration of receptors and signal transducers; endosomes may act as signaling units or even as unique messengers delivering instructions for epigenetic changes to the nucleus (34, 35).

Inferences from Hyperproliferative Phenotypes of Knockout Mice

Duality of signaling is a consistent feature of the immune system. In a very large number of cases, a single molecule or stimulus triggers opposing responses that must be integrated: one promoting signal propagation, the other antagonizing the response. The negative feedback mechanism can operate with diverse kinetics, attenuating the positive response as it develops, triggering an increase in molecules that shut off signal transduction at later times, or dampening the response to subsequent stimulation as observed in T cell anergy. Well-known examples of negative feedback loops that attenuate TCR signaling include increased abundance and exocytosis of CTLA4 (12); T cell receptor down-modulation (36, 37) through the actions of Cbl and Cbl-b (18); activation of the phospholipid phosphatase PTEN (38); metabolism of the second messenger diacylglycerol to phosphatidic acid by diacylglycerol kinase ζ (39); and degradation of signaling molecules, as discussed here. Not surprisingly, targeted deletion of these negative regulators results in hyperactivation of the immune response, with phenotypes ranging from lymphocyte hyperproliferation, splenomegaly, and lymphadenopathy to frank autoimmune disease. The pressure to maintain a balance between positive and negative signaling also explains the bias of the immune system to select for signals that are of intermediate strength. Signal strength has a major role in determining not only activation versus anergy induction, but also thymic selection (40), T helper cell differentiation (41), and lymphocyte activation and death (42).

Consistent with our hypothesis that lysosomal degradation plays a role in anergy, T cell hyperproliferation phenotypes have been observed after inactivation of certain proteins known or presumed to be involved in the monoubiquitination and lysosomal sorting pathway. For instance, CD2AP is an adaptor protein with homology to CIN85, which has an established role in endocytosis (43, 44). CD2AP is necessary for full segregation of the TCR away from LFA-1 during formation of the immunological synapse, and CD2AP-deficient T cells hyperproliferate in response to antigen stimulation (45). Haploinsufficiency of CD2AP is associated with a decreased number of multivesicular bodies and with glomerular disease in both humans and mice (46). In another example, the suppressors of T cell signaling Sts-1 and Sts-2 each contain an SH3 domain and a ubiquitin-associated (UBA) domain [which binds monoubiquitinated proteins (47)], and their combined inactivation results in T cell hyperproliferation and an autoimmune-prone phenotype similar to that observed in Cbl-b–deficient mice (48). Biochemically, T cells from Sts-1 and Sts-2–deficient mice show hyperactivation of Zap70 and its downstream targets, and they accumulate mono- and oligoubiquitinated Zap70 (48). Recently, Sts proteins were shown to bind Cbl proteins and inhibit Cbl-mediated down-regulation of EGFR as well as Cbl-mediated inhibition of Zap-70 (49, 50). It will be interesting to determine whether Sts and CD2AP proteins participate in the program of calcium-dependent T cell anergy and lysosomal degradation described here.


Analysis of T cell anergy suggests an important role for calcium signaling in the initial induction phase and, for E3 ligases, monoubiquitination and lysosomal degradation in the effector phase. How might these insights be translated into the clinic? On one hand, inhibiting the NFAT–AP-1 interaction might be expected to convert immunity into tolerance at the transcriptional level, a good scenario for treating autoimmune disease. On the other hand, blocking the E3 ligases involved in endosomal-lysosomal trafficking might be expected to block the development of tolerance—for instance, to cancer cells—and this approach could have an application in cancer immunotherapy.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
View Abstract

Stay Connected to Science Signaling

Navigate This Article