The Yins of T Cell Activation

See allHide authors and affiliations

Science's STKE  04 Jan 2005:
Vol. 2005, Issue 265, pp. re1
DOI: 10.1126/stke.2652005re1


Transcription factors activated in response to T cell receptor (TCR) signaling include nuclear factor of activated T cells (NFAT) family, which is highly phosphorylated and thereby maintained in the cytoplasm of resting T cells, the nuclear factor NF-κB, which is kept in the cytoplasm of resting cells through its association with the inhibitor protein IκB, and activating protein–1 (AP-1), which is only transcribed after TCR stimulation. Negative regulators of TCR signaling can be divided into two groups: Class 1 regulators help maintain the quiescent state of unstimulated T cells, whereas class 2 regulators are themselves transcriptionally induced in response to TCR signaling and serve to limit and terminate the activating signal. Class 1 regulators include the autoinhibitory domain of the phosphatase calcineurin; IκB and its transcriptional activators Foxj1 and Foxo3a; and various transcriptional coregulators that inhibit interleukin-2 (IL-2) production. Class 2 regulators include the calcipressins, which, like NFATp and NFAT4 are feedback inhibitors of calcineurin-NFAT signaling, IκB, and the mitogen-activated protein kinase (MAPK) phosphatases, which inhibit MAPK signaling and thus the nuclear localization of AP-1 components.


A mature T cell spends most of its life span circulating in the periphery in a dormant state, much like that of a computer set in standby mode. Only when it encounters an antigen presented by an antigen-presenting cell will the dormant T cell become activated to initiate an adaptive immune response against the source of the antigen, often an invading virus or a foreign tissue or organ transplanted into the host. The cell surface signal is generated when the TCR binds to a major histocompatability complex (MHC)–antigen complex at the same time that a myriad of coreceptors, including CD28, bind to their respective ligands on the antigen-presenting cell. Subsequently, the signal emanating from the TCR and its coreceptors on the T cell surface is transmitted through the cytoplasm into the nucleus to reprogram gene expression, causing both up- and down-regulation of gene transcription. This signal transduction process leads to the production of various cytokines, which, in turn, help to drive the proliferation of T cells, as well as the activation of other components of the adaptive immune system. Given the potential destructive power of an active immune system to the host (as well as to its appropriate targets), the extent to which T cells are activated is under stringent control. As a result, T cell activation is accompanied by attenuation of the activation signal, which eventually leads to termination of the signal. Although much progress has been made in the past few decades toward an understanding of the molecular network involved in the TCR-mediated signal transduction responsible for T cell activation, the negative regulators that are required to keep a T cell in its dormant state and to terminate the T cell activation signal have only recently begun to emerge. In this brief review, some of the negative regulators of T cells uncovered in the past few years, especially those that are associated with known positive regulatory proteins or signaling modules, are highlighted.

The negative regulators of T cell activation reported to date can be classified into two categories: those that are involved in the maintenance of the quiescent state of dormant T cells before TCR signaling (which I will call class 1 ) and those whose transcription is induced as a consequence of TCR signaling and are involved in the feedback inhibition and termination of TCR signaling (which I will call class 2). One obvious difference between these two classes of negative regulators is that TCR signaling has to overcome the first class of negative regulators to activate gene transcription. How TCR signaling removes pre-existing negative regulators constitutes an important mechanistic question, which will be addressed for all of the class 1 negative regulators covered in this review.

One of the hallmarks of TCR signaling is the production of interleukin-2 (IL-2), a cytokine that is largely responsible for subsequent T cell proliferation and clonal expansion through an autocrine signaling loop activated by IL-2 binding to the high-affinity IL-2 receptor (13). TCR-mediated IL-2 transcription has been used as a model system to map out the signaling network linking the TCR at the plasma membrane with the promoters of genes encoding various cytokines in the cell nucleus. All of the negative regulators covered in this review help to maintain, directly or indirectly, the "off" state of the IL-2 promoter (class 1 ) or to attenuate and terminate the signal that activates it (class 2).

Class 1 Negative Regulators of TCR Signaling in Resting T Cells

In resting T cells, the transcription of IL-2 and other activation-associated cytokines is suppressed by a number of regulators (Fig. 1). Before turning to these negative regulators, it is first helpful to take a look at the transcription factors that are activated as a result of TCR signaling and that positively regulate IL-2 transcription to see how they are kept inactive in quiescent T cells.

Fig. 1.

Class 1 negative regulators of T cell activation. Class 1 negative regulators, which include the calcineurin autoinhibitory domain, IκB, the transcription factors Foxo3a and Foxj1, and the transcriptional coregulators Cabin1 and Tob, help maintain the quiescent state of unstimulated T cells. Negative regulators are blue; transcription factors that are either neutral or capable of serving as either stimulatory or inhibitory regulators are green.

At least three different families of transcription factors are activated by TCR signaling and subsequently bind cooperatively to the IL-2 enhancer region to cause transcriptional activation of the IL-2 gene. They are the nuclear factor of activated T cells (NFAT) family, the activating protein (AP)-1 family, and the nuclear factor NF-κB family (4). In resting T cells, both NFAT and NF-κB are kept in the cytosol in an inactive form (4). NFAT exists as a highly phosphorylated protein with its nuclear localization signal masked. NF-κB is held in the cytosol through its tight association with a family of inhibitor proteins, the IκBs. AP-1 is absent in resting T lymphocytes and is transcriptionally induced only by TCR signaling. For each of these transcription factors, there thus seems to be a negative control mechanism to ensure that they are prevented from entering into the nuclear compartment in resting T cells. Among the class 1 negative regulators to be covered in this review are the autoinhibitory domain of calcineurin, which is responsible for keeping calcineurin in an inactive state, thus holding NFAT in the cytosol by preventing dephosphorylation of NFAT by calcineurin, and the forkhead transcription factors Foxj1 and Foxo3a, which activate the expression of members of the IκB family to prevent nuclear translocation of NF-κB. In addition, the myocyte enhancer factor 2 family of transcription factors (MEF2)-Cabin1-class II histone deacetylase (HDAC) complex and the Tob-Smad complex, two complexes that directly bind to the IL-2 promoter and silence it through chromatin remodeling, will also be discussed.

The autoinhibitory domain in calcineurin serves as a negative inhibitor of the calcium-calcineurin-NFAT signaling module

The calcium- and calmodulin-dependent protein phosphatase calcineurin plays an essential role in transmitting the calcium signal that occurs after TCR stimulation from the cytosol into the nucleus by way of its major substrate, the phosphorylated form of NFAT (5, 6). Calcineurin exists as a heterodimer consisting of a catalytic subunit and a regulatory subunit. To ensure that calcineurin is not active in resting T cells in the absence of an appropriate calcium signal, an autoinhibitory domain that serves as a pseudosubstrate is present at the C terminus of the catalytic subunit of calcineurin. This autoinhibitory domain binds to the active site of the phosphatase domain of calcineurin and, as such, can be viewed as an intramolecular inhibitor of calcineurin and, thus, of the entire calcium-calcineurin-NFAT signaling module (7).

The calcineurin "on" switch is built into the calcineurin catalytic subunit in the form of a calmodulin-binding domain that lies in close proximity to the autoinhibitory domain (7). Thus, after TCR signaling and the accompanying influx of calcium into the cytosolic compartment, activated calmodulin binds to the calmodulin-binding domain and removes the autoinhibitory domain from the active site of calcineurin to allow it to act on such substrates as NFAT.

Foxo3a and Foxj1 as negative regulators of NF-κB

Expression of the NF-κB inhibitor IκB appears to be under dynamic control in resting T cells. In other words, IκB seems to undergo steady, albeit slow, turnover even without an activating signal and is constantly replenished through the synthesis of new mRNA. Quiescent T cells actively synthesize IκB to ensure that NF-κB is not activated. The transcription factors that activate IκB expression have been recently identified as two members of the forkhead family, Foxo3a and Foxj1 (8, 9).

The forkhead transcription factors consist of a large family of proteins, many of which have been shown to regulate T cell homeostasis. To date, two members of the forkhead family, Foxj1 and Foxo3a, have been unambiguously shown to be actively involved in the negative regulation of resting T cells. Foxj1 and Foxo3a seem to have overlapping, but not completely redundant, functions in T cells, as will be further elaborated below.

The initial hint that Foxj1 may play a role in maintaining the quiescent state of T cells came from the observation that this protein is down-regulated in lymphoid cells derived from lupus-prone mice (8), in which there is an abnormal T cell activation. In wild-type mice, foxj1 is highly expressed in primary naïve T (and B) cells and is down-regulated when both the TCR and CD28 pathways are stimulated. Knockout of foxj1 led to early death in utero, preventing characterization of its function in T cells. Chimeric animals with a foxj1−/− lymphoid system in a recombinase-activating gene knockout (RAG−/−) background were therefore generated and found to exhibit multiorgan systemic inflammation, elevated production of TH1 cytokines, and hyperproliferation of T cells. This is attributed, in large part, to dysregulation of the NF-κB signaling pathway. One of the major isoforms of IκB, IκBβ, is absent in foxj1-deficient T cells, which accounts for the hyperactivation of NF-κB in resting foxj1−/− T cells and the hypersensitivity of those cells to TCR stimulation.

The role of Foxo3a in the negative regulation of T cell activation was not apparent until the generation and characterization of knockout mice. Data from work with cell lines suggested that the primary function of Foxo3a was to control apoptosis, because overexpression of Foxo3a led to induction of apoptosis (10). Unexpectedly, Foxo3a-deficient T cells have no death phenotype: They are not resistant to apoptosis (9). Instead, they show hyperproliferation and substantial elevation of both TH1 cytokines (which promote inflammatory responses) and TH2 cytokines (which promote B cell activation) after TCR signaling (9). Further analysis revealed that Foxo3a-deficient T cells have a high basal level of NF-κB activity, owing to the lack of expression of two IκB isoforms, IκBβ, as in foxj1−/− T cells, and IκBε, indicating that Foxo3a, like Foxj1, is required for the production of IκB in resting T cells. This is consistent with the temporal expression patterns of both Foxj1 and Foxo3a, which are expressed in resting T cells, much as are many housekeeping genes. Comparison of the phenotypes of foxj1−/− and foxo3a−/− T cells suggests that Foxj1 plays a secondary role relative to Foxo3a in controlling IκB expression and TCR-dependent cytokine production, because the foxo3a−/− phenotype is much more severe and encompasses all of the foxj1−/− phenotype. It is interesting that Foxj1 expression is also turned off in Foxo3a-deficient cells, which suggests that Foxo3a lies upstream of Foxj1 in a common signaling pathway.

NF-κB is subject to inhibition by multiple members of the IκB family, including IκBα, β, and ε. It is noteworthy that in foxj1−/− and foxo3a−/− animals, only IκBβ or IκBβ and ε are absent, respectively, with no change in the expression level of the remaining IκB family members (8, 9). These results clearly indicate the lack of functional redundancy among the three IκB proteins in the control of basal level of NF-κB activity. The presence of IκBα did not seem to affect the up-regulation of NF-κB in both foxj1−/− and foxo3a−/− strains. That Foxo3a and Foxj1 keep NF-κB inactive to maintain long-term T cell quiescence by activating the expression of IκBβ and ε instead of IκBα is also reminiscent of a recent computational model that predicts that IκBα is responsible for negative feedback inhibition of NF-κB for fast turnoff, whereas IκBβ and ε are involved in regulation of NF-κB during long-term responses (11).

After TCR signaling, Foxj1 and Foxo3a are inactivated by distinct mechanisms. Foxj1 expression is dramatically down-regulated at the mRNA level by T cell activation as judged by real time reverse-transcription polymerase chain reaction (RT-PCR) (8). Although the precise molecular mechanism of Foxj1 down-regulation remains unclear, it may occur through inhibition of Foxj1 transcription, or accelerated degradation of its mRNA, or a combination of both. In contrast, inactivation of Foxo3a can occur through at least two posttranslational mechanisms, nuclear export mediated by Akt (also known as protein kinase B, PKB) and 14-3-3 (12, 13), and Akt-dependent, proteosome-mediated degradation (14). Although evidence for these alternative pathways of Foxo3a inactivation has been demonstrated in cell types other than T lymphocytes, it would not be surprising if one or both mechanisms operate in TCR-mediated Foxo3a inactivation to pave the way for NF-κB activation during TCR signaling.

Cabin1 and class II HDACs repress the IL-2 and likely the AP-1 promoters through MEF2-dependent chromatin remodeling

Cabin1, also known as cain, was initially identified as a calcineurin-binding protein that blocks calcineurin activity (15, 16). Cabin1 also interacts with the transcription factor MEF2 (17). MEF2 comprises four isoforms encoded by four different genes in the human genome, MEF2A to MEF2D. Although MEF2 was discovered in muscle cells and was shown to play a key role in regulating muscle-specific gene expression, it is now known to be ubiquitously expressed. In T lymphocytes, MEF2D is the most abundant isoform. Among many of the target genes of MEF2 is c-jun, a member of the AP-1 family (18). MEF2 is required for the transcription of c-jun and is likely to mediate AP-1 production during TCR signaling (Fig. 2).

Fig. 2.

The MEF2-Cabin1-mSin3-HDAC and MEF2-HDAC(II) calcium signaling module is involved in both repression and activation of the IL-2 promoter. These modules help keep IL-2 transcriptionally silent in the resting T cell. After TCR signaling, calcium-calmodulin displaces MEF2 from the Cabin1-mSin3-HDAC module (or from HDACII), which allows MEF2 to interact with the transcriptional coactivator p300. Color codes are the same as in Fig. 1, except that positive regulators are red. Ag, antigen.

In resting T cells and other cell types, MEF2 is repressed by various members of a superfamily of transcriptional repressors that includes Cabin1 and class II HDACs. [HDACs are divided into class I and II, on the basis of their homology to the prototypical yeast enzymes yRPD3 and yHDA1, respectively (19)]. Although Cabin1 does not have intrinsic HDAC activity, it is associated with mSin3A, which recruits class I HDACs to form a multiprotein chromatin-remodeling complex (20). Overexpression of Cabin1ΔCNBD, a truncation mutant that lacks the calcineurin-binding domain but retains the MEF2-binding domain, causes substantial inhibition of TCR-mediated IL-2 reporter activation. The strongest evidence for a role of the MEF2-Cabin1 signaling module in TCR signaling comes from a partial Cabin1 knockout animal in which the C-terminal MEF2-binding domain, along with the calcineurin-binding domain of Cabin1, is genetically deleted (21). These so-called Cabin1ΔC mice exhibit increased production of mRNA for IL-2 and other cytokines when stimulated with TCR agonists, such as antibody against CD3 or phorbol 12-myristate 13-acetate plus ionomycin. Further analysis of Cabin1ΔC T cells reveals no obvious defects in calcineurin-NFAT signaling, ruling out the possibility that the hyperactivation phenotype is a consequence of dysregulation of calcineurin activity. This raises the possibility that the loss of Cabin1 binding to MEF2 decreases the threshold for activation of MEF2, which may be directly involved in the transcriptional regulation of IL-2 and other cytokines.

Indeed, careful examination of the IL-2 promoter reveals a potential, albeit imperfect, MEF2-binding site in close proximity to the TATA box (22). Electrophoretic mobility shift assay with oligonucleotides derived from the IL-2 promoter shows that this site is capable of binding to MEF2. Chromatin immunoprecipitation–PCR assay demonstrated that MEF2 is indeed associated with the endogenous IL-2 promoter, and mutation of the putative MEF2-binding site, with the use of transiently transfected plasmid, abolishes the binding of MEF2 to the IL-2 promoter. These data strongly suggest that MEF2 directly binds to the IL-2 promoter and is likely involved in recruiting HDACs either indirectly through Cabin1 (for class I) or directly (for class II HDACs) to repress the expression of IL-2 in resting T cells. As such, the MEF2-Cabin1-mSin3-HDAC1or 2 or MEF2-HDAC (II) complexes represent another class of negative regulatory protein complexes that keep the IL-2 promoter silent. Given the established association of MEF2 with the c-jun promoter, it can be inferred that the same repressive complexes also help to silence the c-jun promoter in resting T cells.

The phenotype of Cabin1ΔC mice, although clear-cut, is not as dramatic as that of Foxo3a- or Foxj1-deficient animals, because it does not cause T cell activation in the absence of stimulation. This is likely because of the functional redundancy of the class II HDACs. It remains to be determined which class II HDACs are present in quiescent T cells and at what levels relative to Cabin1. That Cabin1ΔC animals give rise to hypersensitive T cells suggests that Cabin1 is a major MEF2 repressor that cannot, in resting T cells, be functionally replaced in its entirety by the class II HDACs.

After TCR signaling, the MEF2-mediated repression of IL-2 and c-jun expression by Cabin1 and class II HDACs is relieved by a unique calcium- and calmodulin-dependent process. Thus, both Cabin1 and class II HDACs contain calmodulin-binding domains (16, 23, 24). Note that the calmodulin-binding domains in Cabin1 and class II HDACs overlap with their MEF2-binding domains and, thereby, make the binding of calmodulin and MEF2 to either Cabin1 or class II HDACs mutually exclusive. There thus exists a direct calcium signaling pathway from the cytosol to the nucleus by way of calmodulin, which, when activated, binds to Cabin1 and class II HDACs and competitively removes them from MEF2 and so makes room for such coactivators as p300 (25). In addition to the derepression by calmodulin directly, the interaction of class II HDACs with MEF2 in muscle cells is also susceptible to calmodulin-dependent kinase regulation. Calcium/calmodulin-dependent protein kinase (CaMK) or a downstream kinase can phosphorylate class II HDACs, to create a docking site for 14-3-3 proteins. After binding to 14-3-3 proteins, class II HDACs are exported from the nucleus into the cytosol, separating them from MEF2. More recent evidence shows that Cabin1 also undergoes CaMK-dependent phosphorylation and subsequent 14-3-3–dependent nuclear export during T cell activation (26), much as do class II HDACs (26). It seems that more than one pathway is operative to synergistically relieve MEF2 from the repression by Cabin1 and class II HDACs during TCR signaling.

MEF2-Cabin1 (or MEF2-HDAC II) represents a novel calcium-dependent signaling module independent of the well-established calcineurin-NFAT signaling module. Why should there be two different calcium-signaling modules to regulate the same gene for IL-2? Unlike NFAT, which is sequestered in the cytosol in resting T cells, MEF2 is constitutively bound to the IL-2 promoter, which makes it possible for MEF2 to recruit Cabin1 and class II HDACs to repress the promoter. When calcium signaling begins, the calcium-dependent replacement of Cabin1 and class II HDACs by p300 histone acetyltransferase allows rapid remodeling of the IL-2 promoter in preparation for transcriptional initiation. Thus, the primary role of MEF2 is likely to remodel chromatin, whereas the major function for NFAT and other inducible transcription factors may be to recruit the basal transcription machinery to the IL-2 promoter region, even though these inducible factors also interact with p300. In that sense, the calcineurin-NFAT and the MEF2-Cabin1-p300 calcium signaling modules play nonredundant roles in the transcriptional activation of the IL-2 gene and T cell activation. As such, neither is dispensable for TCR signaling to the IL-2 promoter, as has been amply demonstrated by the potent inhibition of TCR signaling by the calcineurin inhibitors CsA and FK506, and by short interfering RNA (siRNA) knockdown of MEF2 (22).

Tob-Smads as negative regulators of IL-2 transcription

Tob is a member of a novel protein family that shares a common function, suppression of cell proliferation or of cell cycle progression (27). Unlike many other cell types, in which Tob expression is rapidly induced after stimulation with growth factors (2830), Tob is constitutively expressed in quiescent or anergic T cells (31). Tob expression is almost eliminated within 3 hours of stimulation with a combination of antibodies against CD3 and CD28, although not after stimulation with either antibody alone. It is interesting that Tob expression also stops after activation of protein kinase C as a result of treatment with phorbol ester. The expression of Tob in quiescent T cells leads to the down-regulation of the expression of a number of genes essential for cell cycle progression, including genes for cyclin E, cyclin A, and Cdk2, and the up-regulation of p27kip1, a negative regulator of the cell cycle. In addition to its role in the cell cycle, Tob also represses IL-2 transcription.

Tob does not itself bind DNA, which suggests that, as does Cabin1, it may serve as a transcriptional coregulator. Indeed, Tob alters the DNA binding and transactivation activity of Smad family of proteins. In osteoblasts, Tob negatively regulates the bone morphogenetic protein signaling pathway by interacting with Smad1, 5, and 8, blocking their transactivating activity (32). In human Jurkat T cells, Tob associates primarily with Smad2 and Smad4 in the nucleus (31). Although Tob enhances Smad-dependent transcription of a reporter gene, consistent with the enhanced DNA binding activity of Smad4, Tob-Smad protein complexes repress activation of IL-2 transcription. Transcriptional inhibition of IL-2 by Tob and Smad is not mediated through interference with any of the three major inducible transcription factors, NFAT, AP-1, or NF-κB. Instead, the Tob-Smad complex seems to exert its effect by binding to a –105 bp negative regulatory element on the IL-2 promoter. More recent evidence suggests that Tob is likely to repress gene expression through recruitment of HDAC1 (33). It is thus likely that, similar to Cabin1, after binding to the IL-2 promoter, the Tob-Smad2 or 4 complexes recruit HDAC to silence the IL-2 promoter by means of chromatin remodeling.

Paradoxically, ectopic expression of Tob stimulates a Smad-dependent luciferase reporter gene, while inhibiting an IL-2 reporter gene (31). It thus appears that the effect of Tob on the transactivation activity of Smad proteins depends on the target promoter sequence. This is similar to the dual effects of NFAT, which can either inhibit or activate IL-2 transcription, as is discussed below.

Tob expression is high in resting or anergic T cells. What drives Tob expression? The lung Kruppel-like transcription factor (LKLF) is one candidate (34). Like Tob, LKLF is a bona fide negative regulator of T cell activation. LKLF-deficient T cells show hyperproliferation and enhanced expression of activation markers on the T cell surface (35). It remains to be seen whether the hypothesis that, in addition to regulating its known target genes, LKLF is an upstream transcriptional regulator of the expression of Tob is valid.

Class 2 Negative Regulators of TCR Signaling After T Cell Activation

Unlike the class 1 negative regulators of T cell activation, the class 2 inhibitors are either absent or are kept inactive by the class 1 inhibitors in quiescent T cells. A hallmark of class 2 negative regulators is that they become active only after TCR signaling (Fig. 3).

Fig. 3.

(A) Oversimplified depiction of TCR signaling pathways leading to the activation of NFAT, AP-1, and NF-κB, which are covered in this review. (B) Class 2 negative regulators of T cell activation. Class 2 negative regulators, such as calcipressin, NFATp, NFAT4, IκB, and the MKPs, are transcriptionally induced after TCR signaling and are involved in its feedback inhibition and termination. Color codes are the same as in Fig. 2.

Calcipressins are feedback inhibitors of calcineurin

The calcineurin-NFAT signaling pathway is subject to feedback inhibition at the level of both calcineurin and NFAT. Although Cabin1 can associate with and inhibit the phosphatase activity of calcineurin in a TCR-dependent manner (15), knockout of the calcineurin-binding domain of Cabin1 does not cause an obvious defect in calcineurin activity in Cabin1ΔC knockout T cells (21). This may be because of either the existence of a functionally redundant calcineurin inhibitor or because Cabin1 has an alternative role in calcineurin function (besides direct inhibition). Another class of calcineurin-binding proteins, known as MCIPs (modulatory calcineurin-interacting proteins), RCNs (regulators of calcineurin), CBP1 (calcineurin binding protein 1), or calcipressins (referred to hereafter as calcipressins, because this name was adopted in the paper dealing with their T cell functions), has been shown to block the calcineurin signaling pathway as feedback inhibitors.

Calcipressins (Csps) are a family of structurally related proteins that are highly conserved from yeasts to humans (36, 37). Many insights into the function of calcipressins have come from studies on their role in the regulation of calcineurin signaling pathway in yeast and fungi (38, 39), as well as in mammalian cells (40, 41). Knockout of the calcipressin genes in both yeast and mouse leads to dysregulation of the calcineurin pathway. In mouse, for example, Csp-1 (MCIP1) deficiency leads to cardiac hypertrophy, a phenotype consistent with that of calcineurin transgenic mice, which overexpress a constitutively active form of calcineurin and show cardiac hypertrophy (42).

Calcipressin-1 deficiency causes dysregulation of TCR signaling (43). In comparison with their wild-type littermates, calcipressin-1–deficient animals exhibit a decrease in the threshold for transcriptional activation of IL-2 and genes encoding other cytokine in response to TCR signaling. In particular, the gene that encodes Fas ligand, which is responsible for calcineurin-dependent cell death during secondary stimulation, is expressed during primary T cell activation, which leads to premature T cell death and suggests that calcipressin-1 deletion leads to hyperactivation of calcineurin. Consistent with calcipressin’s function as an endogenous calcineurin inhibitor, calcineurin activity from calcipressin-1–deficient T cells is higher than that in wild-type T cells, which explains the lower transcriptional activation threshold for IL-2, Fas ligand, and other cytokines. Because there are three functionally redundant isoforms of calcipressin, more than one member of the family is probably expressed in T cells; it is possible that a more pronounced phenotype of T cell activation would be seen in the presence of double or triple knockouts of calcipressins.

NFATp and NFAT4 negatively regulate T cell activation

The NFATs were identified as a family of transcription factors that convert a calcium signal generated in the cytosol into a transcriptional output in the nucleus, and they have been conventionally viewed as positive regulators of transcription of targeted genes (5, 6). Three members of the NFAT family of transcription factors, NFATc (also known as NFATc1 or NFAT2), NFATp (also known as NFATc2 or NFAT1), and NFAT4 (also known as NFATc3), are expressed at appreciable levels in lymphoid cells. The existence of multiple members of the NFAT family raises the question as to whether they are distinct and functionally redundant. Given that NFAT is known to transmit the calcium-calcineurin signal from the cytosol into the nucleus to activate the transcription of IL-2 and other cytokine genes, one would have expected that NFAT-deficient T cells should have had, to some degree, defects in TCR-stimulated expression of these cytokines. Unexpectedly, NFATc knockout T cells have no defect in TCR-mediated IL-2 production, although they do exhibit defects in the production of TH2 cytokines such as IL-4 (44, 45). Furthermore, knockout of NFATp leads to enhanced TCR signaling and hyperproliferation of T cells (46, 47). Similar hyperactivation of peripheral T cells is seen in NFAT4 knockout T cells (48). Thus, both NFATp and NFAT4 appear to play a negative role during TCR signaling and T cell activation. Nevertheless, double knockout of both NFATc and NFATp leads to decreased production of IL-2 and other markers of T cell activation (49).

Members of the NFAT family are capable of activating gene transcription; that has been unambiguously established, because mutation of the NFAT-binding sites in the minimal IL-2 promoter abrogates its transcriptional activation in response to TCR agonists (50). How could the same proteins serve as negative and positive regulators of transcription? Part of the answer may come from structural analysis of various NFAT-DNA complexes and may lie in the DNA sequence-dependence of transcriptional regulation by NFAT (51). NFAT is capable of binding to NF-κB binding sites of the HIV 5′ long terminal repeat (LTR) (52). A comparison of the complexes of NFATp dimer-DNA and NFATp-Fos-Jun-DNA reveals that the change in DNA sequence from the NFAT-AP1 composite site in the IL-2 promoter to the NF-κB binding site of the HIV LTR alters the conformation of NFATp, particularly the C-terminal half of its Rel homology DNA binding domain, which allows NFATp to either form an asymmetric dimer with itself or to interact with AP-1 protein bound to an adjacent site (51). It would not be surprising if such conformational changes after NFAT binding to different DNA elements could lead to its association with either transcriptional coactivators or corepressors. These allosteric effects of DNA on the conformation of transcription factors have been seen in several classes of transcription regulators including the steroid hormone receptor (53), the POU domain factor Pit-1 (54), and NF-κB (55). In fact, it has been shown that the calcineurin-NFAT signaling pathway negatively regulates the expression of cyclin-dependent kinase (Cdk)4, and this repression is attributed to recruitment of HDACs to the Cdk4 promoter by activated NFATc (56). This may also account for the ability of NFAT to induce anergy in the absence of AP-1 when TCR is activated without costimulation signals (57).

It is noteworthy that of the three NFAT family members that are expressed in T cells, only transcription of NFATc is induced by TCR signaling (5). The steady increase in NFATc concentration after TCR signaling will alter the balance among different members of the NFAT family, which may also have functional consequences on the activity of the NFAT family as a whole to act as either positive or negative transcriptional regulator of IL-2 and other cytokines after TCR activation.

IκB as a feedback inhibitor of NF-κB signaling pathway

Although IκB expression appears to be dictated by the forkhead transcription factors Foxj1 and Foxo3a in quiescent T cells, these transcription factors become irrelevant for IκB expression after T cell activation, because they are either degraded or exported into the cytosol (preventing their contact with the IκB promoter). However, a highly conserved mechanism exists for feedback inhibition of NF-κB through the NF-κB–dependent activation of IκB transcription, which replenishes cells with newly synthesized IκB proteins (58). There are a number of other alternative mechanisms by which the NF-κB signaling pathway is terminated, many, if not all, of which may apply to TCR-mediated NF-κΒ activation (58).

MAP kinase phosphatases as activation-induced inhibitors of the MAPK pathway during T cell signaling

Of the three transcription factors induced by TCR signaling that are covered in this review, AP-1 is the only one that is directly regulated by MAPKs. Among the two prominent AP-1 components, c-Fos is known to be regulated by extracellular signal-regulated kinases (ERKs) and c-Jun is regulated by c-Jun N-terminal kinase (JNK) (59, 60). After TCR signaling, newly synthesized c-Fos and c-Jun have to undergo phosphorylation by their respective MAPKs to enter the nucleus, where they heterodimerize and then bind to the IL-2 promoter in complex with NFAT. Thus, MAPK pathways play important roles in AP-1 activation and TCR-mediated IL-2 transcription.

The MAPK pathways are subject to feedback inhibition by MAPK phosphatases (MKPs), because the latter are transcriptionally induced after activation of the TCR-stimulated MAPK pathway. Two of the MKPs, MKP5 and MKP6, have been shown to negatively regulate T cell activation by inhibiting the activity of their respective MAPK targets.

MKP6 was identified as a protein that interacts with the cytoplasmic tail of the CD28 coreceptor. MKP6 is induced when CD28 costimulates primary human T cells (61). Retrovirus-driven ectopic expression of a dominant-negative form of MKP6 increases the expression of IL-2 in response to TCR and CD28 costimulation. An increase in IL-2 production is also obtained by using a mutant CD28 coreceptor, which can no longer interact with MKP6. Taken together, these observations suggest that MKP6 is involved in negative regulation of MAPK pathways in response to CD28 costimulation.

MKP5 was initially characterized as a phosphatase that acts preferentially on JNK and p38 (62). Although MKP5 is most abundant in liver and skeletal muscle, the presence of a DNA marker of its expression suggests that it is also expressed in the spleen among other tissues. Recently, a Mkp5 knockout animal was generated, which exhibited up-regulation of JNK activity and the consequent dysregulation of both innate and adaptive immune responses (63). The phenotype of Mkp5 knockout animals revealed the complexity of the role of this phosphatase in T cell activation. Mkp5-deficient CD4+ T cells showed a decrease in their capacity to proliferate in response to stimulation by antibodies to both CD3 and CD28, suggesting that MKP5 is required for T cell proliferation. In contrast, after differentiation into TH1 and TH2 cells, production of the corresponding cytokines, interferon-γ (IFN-γ) and IL-4, is dramatically enhanced in Mkp5 knockouts compared with wild-type control cells. In an in vivo model of the T cell response to lymphocytic choriomeningitis virus (LCMV) infection, Mkp5-deficient T cells show no defects in the primary response to the virus as determined by viral clearance. However, after rechallenge with LCMV, the Mkp5 knockout T cells produce greater amounts of different cytokines, including tumor necrosis factor-α, IFN-γ, IL-4, and IL-2, which suggests that MKP5 may be involved in controlling the secondary immune response to infection, possibly by controlling the threshold of activation, particularly in memory T cells.


Much of the work on TCR signaling and T cell activation in the past has been focused on the identification and characterization of positive regulators. Only more recently has attention has been paid to the negative regulators of T cell activation found from the cell surface (64) to the nucleus. The life cycles of T cells are dictated by their encounters with an appropriate antigen and are characterized by a relatively short-lived activation state, followed by differentiation or apoptosis. Negative regulators are essential not only to maintain the quiescent state of T cells but also to attenuate and terminate an immune response after the elimination of the inciting pathogens. A more deliberate and systematic approach is likely to unravel even more negative regulators of T cell activation and signaling, which may enrich our understanding of T cell biology while providing promising new strategies and targets for development of immune system-based therapies for a plethora of human diseases.


  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.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
View Abstract

Stay Connected to Science Signaling

Navigate This Article