|Home | About | Journals | Submit | Contact Us | Français|
Regulatory T (Treg) cells are crucial enforcers of immune homeostasis. Their characteristic suppressive function largely arises from an equally unique pattern of gene expression. A complex network of factors and processes contribute to this ‘signature’ Treg gene expression landscape. Many of these alter the level and activity of the Treg-defining transcription factor Foxp3. Since stable expression of Foxp3 is important for the ability of Treg cells to successfully prevent excessive or inappropriate immune activation, uncovering the mechanisms regulating Foxp3 level is required for the understanding and therapeutic exploitation of Tregs. While transcriptional regulation of the Foxp3 gene has been studied in depth, additional regulatory layers exist controlling the expression and activity of this key transcription factor. These include less-defined mechanisms active at the posttranslational level. These pathways are just beginning to be elucidated. Here we summarize emerging evidence for distinct, posttranslationally active, ubiquitin-dependent pathways capable of controlling the activation and expression of Foxp3 and the function of Tregs. These pathways offer untapped opportunities for therapeutic fine-tuning of Tregs and their all-important restraint of the immune system.
The immune system is capable of annihilating both invading pathogens and malignant threats from within. However, containment of this onslaught is necessary to spare the host from injury resulting from overzealous or misdirected immune activation. Regulatory T cells (Tregs) play a critical role in enforcing immune tolerance by moderating immune response intensity and by preventing self-directed responses (1, 2). In the cancer settling, however, elevated Treg frequencies contribute to the state of immune tolerance that is counterproductive to the mounting of an effective anti-tumor response (3).
Several T-cell subsets display suppressive function either as their recognized principal function or at certain points during their participation in an immune response (4–7). Perhaps the most important and certainly the most intensely studied Treg cells are those marked by characteristically high, constitutive expression of CD25 and the transcription factor Forkhead box protein 3 (Foxp3). While transient expression of Foxp3 can be seen in non-Tregs upon activation (8–10), the continuous expression of the transcription factor is key for enforcing the immune-suppressing phenotype of Tregs. Reflecting this, mutation of Foxp3 in mice and humans results in dramatic and widespread loss of immune homeostasis (11, 12).
Foxp3+ Tregs cells deploy a number of mechanisms to counteract immune activation including the production of anti-inflammatory cytokines [e.g. interleukin-10 (IL-10), transforming growth factor-β (TGFβ), and IL-35], the expression of coinhibitory molecules [e.g. cytotoxic T-lymphocyte antigen-4 (CTLA-4) and lymphocyte activation gene 3 (LAG3)], the modulation of antigen-presenting cell (APC) activity, the scavenging of growth factors, and the disruption of effector cell metabolism (13). Execution of these suppressive functions depends upon the expression of a repertoire of immune-suppressing factors and the silencing of T-effector genes in Tregs. The establishment and maintenance of this Treg signature gene expression pattern is anchored by Foxp3 and a cadre of co-regulatory factors that interact with the ‘master regulator’ (14). Understanding the mechanisms underlying Foxp3 expression and Treg function may reveal critical regulatory junctures that could serve as targets for therapeutic interventions capable of modulating the immune response during cancer, chronic infection, or autoimmune disease.
Transcriptional activation at the Foxp3 locus has been the focus of intense investigation. The molecular events responsible for Foxp3 transcription have been particularly well studied during the generation of Tregs in the thymus. This organ is a critical site of T-cell development and central tolerance. Thymic mechanisms of positive selection ensure that developing T cells are capable of recognizing antigen. Cells with strongly self-reactive T-cell receptors (TCRs), however, fall victim to negative selection and are culled.
High affinity TCR-antigen-MHC interactions also initiate the development of thymically derived (t)Tregs. TCR engagement on maturing T cells triggers the activation of protein kinase Cθ (PKCθ), which then phosphorylates the scaffold protein CARMA1 [caspase-recruitment domain (CARD) membrane-associated guanylate kinase (MAGUK) protein 1] leading to recruitment of Bcl10 (B-cell lymphoma 10) and MALT1 (mucosal associated lymphoid tissue 1), yielding the CBM complex. This complex serves as a molecular platform that facilitates the activation and nuclear translocation of NFκB family members by driving the phosphorylation and subsequent polyubiquitination and degradation of the inhibitor IκB [inhibitor of nuclear factor κB (NF-κB)] subunit (2, 15), perhaps one of the most appreciated examples of ubiquitin-mediated regulation in the immune system. Mice genetically deficient in the key players of this pathway [PKCθ, BCL10, CARMA1, TAK1 (TGFβ-activated protein kinase-1), IKK2, and c-Rel] display reduced tTreg output (16).
Signaling cascades initiated at the TCR, combine with those triggered by the B7/CD28 costimulatory axis, which, are also required for proper thymic development and peripheral maintenance of Tregs (17–22), culminating in the recruitment of several key transcription factors to the Foxp3 promoter. Activator protein-1 (AP-1), nuclear factor of activated T cells (NFAT), NFκB family members (particularly c-Rel), the Runx-CBFβ complex, and Foxp3 itself have been shown to bind and activate transcription of the Foxp3 gene (2, 23–28). The Foxo proteins, Foxo1 and Foxo3a, also bind the promoter of Foxp3 (and other Treg-associated genes, such as Ctla4) and drive Foxp3 expression (29).
Both the initiation and maintenance of Foxp3 transcription are highly dependent on key conserved noncoding sequences (CNS). The NFκB family member c-Rel, drives development of Tregs in the thymus (24–26) by serving as a ‘pioneer factor’ responsible for turning on transcription at the Foxp3 gene by binding one of these sequences, the conserved non-coding sequence 3 (CNS3), which is critical for Foxp3 induction in the thymus and periphery (24–26, 28, 30, 31). Continued expression of Foxp3 after egress from the thymus depends on another regulatory site known as CNS2. This region, rich in CpG residues, has also been referred to as the Treg-specific demethylated region (TSDR).
As suggested by its name, the CpG elements within the TSDR of the Foxp3 locus are extensively hypomethylated in ex vivo isolated Tregs displaying relatively stable expression of Foxp3 under a variety of conditions. It has been reported that the demethylated state of this region in stable tTregs is chemically initiated in thymic Treg precursors after TCR activation (32). Demethylated elements in the CNS2 TSDR as well as the Foxp3 promoter itself and other critical regulatory loci serve as preferential binding sites for several transcription factors including c-Rel, Creb, ATF, Runx-Cbfb, Ets (25, 27, 33, 34), and Foxp3 itself (28). Interestingly, while initiation of Foxp3 expression in developing tTregs requires TCR activation and is critical for long-term commitment to the Treg phenotype, distinct Foxp3-independent outcomes of TCR signaling also appear to contribute to the epigenetic signature of Tregs that underlies their function (35). Together, it seems, upregulated Foxp3 and the Treg-specific epigenetic signature enforce continued transcriptional commitment to the Treg gene expression profile long after thymic egress and circulation in the periphery.
Cytokine signaling pathways also play an important part in the regulation of Foxp3 expression at the level of transcription. Signaling through the IL-2/IL-2R axis has been a known requisite for functional Tregs. IL-2R knockout mice experience autoimmunity and are deficient in their Treg pools (36, 37). Along with IL-2, other cytokines relying on the common γ-chain (γc) receptor are highly important for tTreg generation as well as the induction and maintenance of Foxp3+ Tregs in the periphery. Reflecting this, knockout of these cytokines or the common γc severely and broadly reduces Foxp3+ T cells (2). Furthermore, signal transducer and activator of transcription 5 (STAT5), the crucial downstream mediator of the signaling triggered by these cytokines, is capable of binding the promoter and CNS2 enhancer of the Foxp3 gene (38–41). Indeed, mice lacking either STAT5 or JAK3 have substantially reduced Treg populations (42) and constitutive activation of the STAT5 pathway can rescue Treg levels even in the absence of IL-2. Furthermore, high levels of surface CD25, the high affinity IL-2 receptor, has been linked to enhanced stability of Foxp3 expression and Treg suppressive function (43–45).
TGFβ signaling also plays a role in tTreg generation. It has been reported to be crucial, along with IL-2 signaling, for the generation of tTregs (46). This may be attributable to Treg-survival promoting effects (47) or distinct mechanisms. Still other studies using TGF-β RII-deficient mice suggest that TGFβ is not required for tTreg generation (48). While the role of TGFβ signaling in the transcriptional upregulation of Foxp3 in developing tTregs is debated, the singular importance of this cytokine for extrathymically derived Tregs is clear.
Even though the majority of Foxp3+ Tregs arise in the thymus, a significant and functionally important Treg population is induced in peripheral tissues from naive CD4+ T-cell precursors. In these peripherally induced Tregs (pTregs or iTregs when induced ex vivo), the circumstances of Foxp3 upregulation are distinct from those active in the thymus.
Cytokine signals from TGFβ are required for pTreg and iTreg differentiation (48). Smad3 activation downstream of TGF-β/TGFβ receptor engagement, while not required for tTreg generation, is necessary for pTreg generation (49). Furthermore, the binding of active Smad3 to the CNS1 enhancer of the Foxp3 gene is a key event in the induction of the transcription factor during pTreg but not tTreg differentiation (28). Reflecting this, CNS1-deficent mice, having a defective pTreg compartment, fail to properly enforce immune tolerance at barrier sites such as the gut but are spared from aggressive autoimmunity by their essentially normal tTregs (50).
The cytokine IL-2 also is important for induction of Foxp3 in iTregs (48). Additionally supplementation with this cytokine, as well as exogenous supplies of TGFβ can stabilize Foxp3 expression in iTregs that can lack the redoubtable epigenetic character (i.e. the hypomethylated TSDR) of their relatively more stable tTregs counterparts. Notably, Chen et al. (51) showed that IL-2/anti-IL-2 complexes promote TSDR demethylation in TGFβ-induced Tregs, stabilizing Foxp3 expression and suppressive function in vitro and in vivo while IL-2 blockade leads to loss of Foxp3 expression and Treg function. Accordingly, IL-2 treatment of lymphopenic mice was found to prevent the loss of Foxp3 in adoptively transferred Tregs (52).
Activation of TCR and costimulation signaling cascades are important for extrathymic Treg differentiation, as is the case for tTregs. However, conditions of strong or overzealous activation of these pathways are counterproductive in the generation pTregs (53, 54). Particularly, during the skewing of naive T cells towards an iTregs fate, activation of the phosphoinositide 3-kinase (PI3K)/mammalian target of rapamycin (mTOR)/AKT pathway downstream of TCR-, CD28/B7-, and cytokine-initiated signaling can interfere with the upregulation of Foxp3 (53, 54). Conversely, mTOR inhibition or genetic knockout in naive T cells preferentially results in Foxp3+ T-cell induction over other T-helper lineages (55). Furthermore, mTOR inhibition can stability of Foxp3 expression (56, 57) and in established Tregs it also bolsters their suppressive function and prevents aberrant expression of effector T genes (58).
The negative effect of this pathway on Foxp3 expression occurs through the antagonism of Foxo proteins (Foxo1 and Foxo3a). These known activators of Foxp3 and other Treg-associated genes help enforce a suppressive phenotype and are negatively regulated by PI3K/mTOR/AKT signaling through posttranslational modification. Specifically, phosphorylation of Foxo proteins by AKT prevents their nuclear translocation, leaving them sequestered in the cytosol, and negating their participation in Treg gene expression (59, 60).
With the mechanisms driving transcription of the Foxp3 gene in Tregs being many and complex, it is not surprising that some Foxp3+ Tregs display remarkably stable expression of the transcription factor and staunch fidelity to a suppressive phenotype. Nevertheless, Tregs, under certain (typically inflammatory) conditions, have been reported to lose Foxp3 expression and their characteristic function. As our grasp of Foxp3 regulation becomes more complete and our view of Treg behavior grows more nuanced, these incongruous views on Treg stability may be reconciled. For instance, heterogeneity within Foxp3+ Tregs (based on CD25 expression level) appears to underlie some differences in phenotypic stability (61). Incompletely understood factors and pathways regulating Foxp3 beyond the level of gene transcription might also bring about changes in Treg function. Appreciation for these will likely clarify our concept of Treg functional stability and the potentially dynamic behavior of these cells during immune responses. As discussed later, a deeper understanding of these mechanisms may also shed light on how Treg suppressive function can be responsive to environmental stimuli.
In addition to the transcriptional and epigenetic mechanisms controlling Foxp3 expression and Treg function, other pathways exist that apparently hold considerable sway over these important elements of immune control. The transcript-level action of microRNAs (miRNAs), for instance, is key in shaping the T-cell immune response and its regulation as well. These short, non-coding RNAs arise through the sequential processing of primary transcripts mediated by two class III RNases known as Drosha and Dicer. miRNAs suppress the expression of target gene products through the RNA-induced silencing complex resulting in the degradation of transcript or inhibition of its translation (62).
The importance of miRNAs for Treg lineage stability is clearly demonstrated by mice with Treg specific deletion of Dicer or Drosha. Deficiencies in either enzyme critical for miRNA processing causes autoimmune pathology reminiscent of that seen with Foxp3 mutation (63). Dicer−/− Tregs are highly unstable and ineffective suppressors taking on production of effector cytokines and showing perturbed expression of Treg-associated gene products including neuropillin, glucocorticoid-induced tumor necrosis factor receptor (GITR), and CTLA-4 as well as loss of Foxp3 expression (64) and homeostasis (65).
Tregs display unique microRNA profiles and among these characteristic microRNAs is miR-155, which is key for Treg fitness under adverse conditions. By targeting the STAT signaling pathway inhibitor known as SOCS1, miR-155 enhances STAT5 activation, increasing responsiveness to the Treg-sustaining cytokine IL-2 (66). miR-10a is also highly expressed by Tregs (67, 68), and its expression prevents in vivo Foxp3 loss. miR-10a was also shown, along with miR-182, to be important in establishing cytokine-influenced heterogeneity among Tregs (69).
While these and other miRNAs have been identified as important positive regulators for Treg suppressive function, it is possible other miRNAs can negatively regulate the function of Treg. For instance, a specific member of the miR-17-92 cluster appears to oppose Treg function. This highly conserved, prototypical example of a polycistronic miRNA gene encodes six miRNAs (miR-17, miR-18a, miR-19a, miR-20a, miR-19b-1, and miR-92-1) (70). Over-expression of this miRNA cluster in hematopoietic cells results in autoimmunity (71) and negatively impact Treg induction (72). We have also found that miR-17 levels are suppressed in Tregs but are elevated by the Treg-destabilizing cytokine IL-6. Furthermore, miR-17-mediated downregulation of the important Foxp3 interacting coregulatory factor Eos was found to result in a destabilized Treg phenotype marked by reduced suppressive function in vitro and in vivo (our unpublished results). In human Tregs, miR142-3p and miR-31 are downregulated to preserve expression of their targets, which include Foxp3 (73, 74). A role for miR21 and miR210 in the downregulation of Foxp3 was also recently shown (75). In all, these findings illustrate that mechanisms outside of transcription can significantly affect the expression of Foxp3 and the function of Tregs.
Mechanisms active at the protein-level are also important in Treg biology. As discussed in the following sections, posttranslational modifications of Foxp3 have major consequences for the stability and function of the transcription factor. The processes of acetylation, phosphorylation, and ubiquitination are important in this newly appreciated layer of Foxp3 regulation and Treg functional modulation.
Lysine acetyl transferases or histone acetyl transferases (HATs) acetylate lysine residues in diverse target proteins including Foxp3. Acetylation of Foxp3 has been shown to stabilize its cellular protein pool. Furthermore, this modification enhances the DNA binding ability and therefore the activity of Foxp3 as a transcriptional regulator (76, 77). In line with this, lysine/histone deacetylatases (HDACs) negatively affect Foxp3 expression and Treg function, while HDAC inhibitors promote Foxp3 expression in Tregs and enhance their suppressive performance (76, 78–81).
Phosphorylation of Foxp3 has also been observed to have consequences for Treg function. Modification of serine residue 418 (Ser418) has been reported to promote Foxp3-mediated gene repression in Tregs under inflammatory conditions. Conversely, phosphatase activity induced by the proinflammatory and Treg-antagonizing cytokine TNFα counteracts the functional enhancement afforded by this modification (82). Interestingly cyclin-dependent kinase 2 (CDK2), which is activated downstream of the TCR, recruits cyclin E to phosphorylate sites in the N-terminal ‘repressor’ domain of Foxp3 (83), a region shown to be important for recruitment of coregulatory factors and binding partners such as Eos, IRF4, HIF-1, and c-Jun (84–86) (members of the so called ‘Foxp3 interactome’) (87). In line with observations that CDK2-deficient Tregs are more suppressive than their wildtype counterparts, mutation of the serine and threonine residues targeted for phosphorylation by cdk2 leads to increased Foxp3 half-life and function (88)
More recently, Li et al. (89) reported that in human Tregs, the kinase PIM1 interacts with Foxp3 (at the C-terminal domain) and phosphorylates Ser422, which interferes with the transcription factor’s ability to bind DNA. Additionally, they found that the Foxp3-enhancing modification at Ser418 prevents the inhibitory phosphorylation at Ser422 (89). Taken together, these studies suggest that Foxp3 function can be positively or negatively modulated by site-specific, possibly cross-regulating, phosphorylation-mediated pathways (82, 83, 89).
Of the more than 200 types of posttranslational modification, ubiquitination along with phosphorylation may be the most prevalent and are certainly the most intensely studied regulatory modifications. Ubiquitination refers to the covalent attachment of the small (8 kDa, 76 amino acid) ubiquitin protein to specific lysine residues of target proteins. This process occurs in three sequential, highly regulated steps.
The initial step is carried out by a ubiquitin-activating enzyme (termed ‘E1’) that forms an ATP-dependent, thioester bond between a cysteine residue in its active site and the C-terminal carboxyl group of a ubiquitin molecule (90). Next, the ubiquitin molecule is transferred from the E1 to a cysteine residue in the active site of one of ~40 ubiquitin-conjugating enzymes (E2) via a trans-acylation reaction. In the third step of the process, any of the multitudinous E3 ubiquitin ligases facilitates a reversible isopeptide bond between the target protein’s lysine residue and the C-terminal glycine of the ubiquitin molecule (91, 92).
Ubiquitin modifications can consist of monomers or polyubiquitin chains. In the case of the latter, the di-ubiquitin linkage type – determined by the lysine residue point of attachment – determines the type and consequences of the modification. The seven types of linkages (one for each lysine residue in the ubiquitin protein) are known as K6, K11, K27, K29, K33, K48, and K63. Additionally ubiquitin subunits can be joined at the amino-terminal methionine residue resulting in an eighth linkage type termed M1 or linear linkage (93). Processes of target protein degradation, activation, and cellular localization can depend upon the specific ubiquitin linkage type.
K48-type polyubiquitination is well known for precipitating the degradation of modified targets via the 26S proteasome (90). For over two decades, ubiquitination-driven proteolysis has been recognized as a means to remove cellular pools of damaged or misfolded proteins (92). It is also well established that K63-type modifications are important for the activity of various cell signaling pathways including those triggered by DNA damage and cytokines (91). Linkage type and substrate specificities are afforded by the considerable diversity among the expansive E3 ligase superfamily.
Appropriately, many diverse cellular processes depend upon ubiquitin-mediated events. Among these processes are those governing the immune response from its activation to its restraint. Indeed the importance of ubiquitination-mediated events for activation of the signaling cascades triggered by TLR ligands and inflammatory cytokines are well known and have been reviewed in detail elsewhere (91). In the following sections, we discuss recent discoveries that bring to light several ubiquitin-dependent pathways that modulate Treg generation, function, and stability, some of which directly regulate Foxp3 expression and activity.
There are many examples of ubiquitin-mediated processes that influence the biology of Tregs. As mentioned above, the thymic generation of Foxp3+ Tregs is very much influenced by ubiquitin-mediated regulation of NFκB signaling. TCR/costimulation triggered activation of PKCθ in Treg-precursors leads to assemblage and activation of the CBM complex. This complex serves as a platform for recruitment of the kinase IKK, which phosphorylates IκB, a modification that leads to its polyubiquitination and proteasomal degradation, freeing NFκB family members, such as c-Rel, from the inactivating complex allowing its nuclear translocation (15) and participation in the transcriptional activation of the Foxp3 gene. The role of ubiquitin-mediated events in Treg biology does not stop here, though.
Casitas-B-lineage lymphoma protein-b (Cbl-b) is another E3 ligase highly relevant to Tregs. This RING-type E3 ligase is known to play a role in setting the threshold for T-cell activation (94). By nonproteolytic tagging the regulatory p85 subunit of PI3K, Cbl-b prevents the association of the kinase with CD28 and the ζ chain of the TCR and activation of downstream signaling. This moderates T-cell activation and promotes T-cell anergy and immunological tolerance. Its importance is evidenced by the hyperproliferative phenotype seen in T cells from mice lacking Cbl-b and the proclivity for autoimmune disease seen in these mice (94, 95).
When TCR activation is coupled with CD28 costimulation, Cbl-b itself becomes ubiquitinated and thus marked for proteasomal degradation. In this way, costimulatory signaling can itself dampen effects of Cbl-b, working to lower the threshold for T-cell activation through this crosstalk. In contrast, coinhibitory signals from CTLA-4 engagement stabilize Cbl-b levels, restraining TCR signaling (95, 96).
This E3 ligase has been implicated in multiple processes relevant to Tregs. By restraining PI3K signaling, Cbl-b also counters Akt/mTOR activation, which negatively effects Foxp3 expression. Cbl-b is particularly key for iTregs as the process of Foxp3 induction is defective in the absence of this ligase. Foxp3 upregulation can be rescued in Cbl-b−/− mice, however, by inhibition of PI3K (60), clearly demonstrating its role as a negative regulator of this pathway. Cbl-b has also been reported to be critical for signaling triggered by TGFβ, making its action even more important for iTregs, that depend heavily on this cytokine to drive their differentiation. As discussed later, this ligase has also been reported to directly affect Foxp3 in tTregs as well (97).
Another ligase important in Tregs is named for the ‘Itchy’ phenotype displayed by mice lacking it. ITCH, a HECT-domain type E3 ligase, also promotes the generation and maintenance of Tregs. Like those of Cbl-b−/− mice, ITCH-deficient T cells are hyperproliferative. ITCH−/− mice are prone to excessive, Th2-dominated inflammation, and their T cells resist suppression by Tregs and the anti-inflammatory cytokine TGFβ (98).
ITCH is also involved in TGFβ signaling which is of central importance in the conversion of naive CD4+ T cells into extrathymic Tregs. Specifically, ITCH has been linked to the phosphorylation and activity of Smad2 (99) and the monoubiquitination of a transcription factor known as TGFβ-inducible early gene 1 product (TIEG1). The latter modification promotes TIEG1’s nuclear translocation, which allows it to bind and transactivate the Foxp3 promoter. Reflecting this, naive T cells from both ITCH and TIEG1-deficient mice induce Foxp3 poorly under iTreg skewing conditions compared to wildtype cells. In these mice, the function of iTregs is also adversely affected as they fail to suppress airway inflammation in vivo (100). ITCH also stabilizes Foxp3 expression during the Treg skewing process. Deficiency in either ITCH or Nedd4 family–interacting protein 1 (Ndfip1), a TGFβ-upregulated adapter molecule necessary for ITCH-mediated polyubiquitination, impairs in vivo and in vitro differentiation of Tregs, a defect linked to ITCH’s importance as a specific suppressor of Th2-associated gene expression in Tregs (99).
In addition to its role in Treg induction, ITCH also seems to be important for the function of established Tregs. Mice with Foxp3-restricted ITCH deficiency show signs of multi-organ autoimmunity. Despite being as frequent as their wildtype counterparts and as functional in vitro, ITCH-deficient Tregs cannot control Th2-type inflammation, particularly in the lung and gut. (101). This defect was tied to the acquisition of Th2-like attributes such as heightened STAT6 activation and expression of GATA-3 and IL-4 (101).
The ubiquitinating enzyme known as ‘gene related to anergy in lymphocytes’ (GRAIL) is a RING finger E3 ligase important for the induction of anergy in T cells and immune tolerance. In the absence of GRAIL, T cells are hyperproliferative and are less sensitive to suppression than their wildtype counterparts. Interestingly, while GRAIL is also upregulated by Tregs (102), populations of these regulators in GRAIL−/− mice appear normal as does the process of in vitro Foxp3 induction in TGFβ-treated naive CD4+ T cells. However, Tregs from these mice are less suppressive than their wildtype counterparts, and despite expressing wildtype levels of Foxp3 mRNA, these cells express genes typical of Th17 cells. This uncharacteristic effector T-cell gene expression was attributed to the abnormally high levels of IL-21 made in the absence of GRAIL-mediated checks on NFAT signaling (103). Therefore, it seems that in addition to an important role in the induction of anergy, GRAIL preserves Treg function by restraining inappropriate gene expression capable of undermining their suppressive phenotype.
Another ubiquitin-associated facilitator of Treg function is known as Ubc13. While actually a K63-type E2 ubiquitin conjugating enzyme known to be involved in TNFα signaling and NFκB activation (104), this factor plays an important role in preventing inappropriate T-effector genes in Tregs (105). Mice specifically lacking Ubc13 in Foxp3+ cells experience widespread inflammation linked to abnormally proliferative, functionally defective Tregs that were also found to express the proinflammatory cytokines IFNγ and IL-17 (105). The Treg defect seen in the absence of Ubc13 appears, in part, to be the result of decreased levels of IL-10 and SOCS1 (an important regulator of IFNγ and IL-17 production) (106). In Ubc13−/− Tregs, constitutive activation of IKK2, a Ubc13 target in the NFκB pathway (which is disrupted in Ubc13-deficient T cells) could rescue SOCS1 expression. Furthermore, administration of a SOCS1 mimetic peptide could partially restore the Treg phenotype in vitro and in vivo (105). In this way, Ubc13 contributes to the focusing of Treg gene expression.
The examples described above illustrate how sensitive Tregs and their part in maintaining immunologic homeostasis can be to ubiquitination-mediated regulation. While these involve the cytokine- and receptor-initiated pathways central to Tregs, an element most central to the acquisition and maintenance of the Treg phenotype, the Treg master regulator Foxp3, also answers to ubiquitin-mediated sway. Recently uncovered pathways leading to the direct regulation of Foxp3 protein by ubiquitination are discussed in the following section.
Hypothesizing that Foxp3 expression might be regulated at the protein level, several groups have begun to explore the potential ubiquitin-dependent pathways directly controlling the stability and activity of this key transcription factor. We and our colleagues found that, in Tregs, the cellular pool of Foxp3 protein is subject to what is likely constant turnover. This process was found to be facilitated by K48-type polyubiquitination followed by degradation via the proteasome (107, 108). In keeping with the notion that Treg function can be disrupted by inflammatory cues, the cellular half-life of Foxp3 can be markedly shortened and the bulk Foxp3 pool is reduced by exposure to a range of inflammatory stresses in vitro including LPS, proinflammatory cytokine, and heat shock (108). Encouraged by these findings, we set out to determine the molecular events involved in this ubiquitin-dependent downregulation.
Mass-spectrometry analysis of the proteins that interact with Foxp3 identified the heat shock 70kDa protein (Hsp70) to be among the binding partners of this critical Treg regulator. Hsp70 is a well-known chaperone molecule that collaborates with a stress-activated, U-box domain type E3 ubiquitin ligase known as C-terminal of Hsp-70-interacting protein (CHIP) or Stub1 (109). Since Stub1 is recruited to mediate the degradation of numerous target proteins that include the transcription factors Runx2 and HIF-1α (110, 111), we suspected it to be the driving force behind degradative Foxp3 polyubiquitination. Interestingly, Stub1, while not highly expressed in Tregs at baseline, can be upregulated in mouse Tregs under the same conditions capable of downmodulating Foxp3 in vitro. This, as well as an observed physical interaction between Foxp3 with Stub1 further suggested the involvement of this E3 ligase in Foxp3 protein loss (108).
Evidence that Stub1 is involved in the protein-level downregulation of Foxp3 came from retroviral overexpression of Stub1 in Tregs, which undermined levels of Foxp3 and the suppressive function of these cells in vitro as well and in vivo. Compromised expression of several Treg-associated genes and the inappropriate upregulation of effector T-cell cytokines were detected upon forced Stub1 upregulation. Conversely, when Stub1 induction in cell lines and stressed Tregs was hampered by siRNA knockdown, Foxp3 expression was stabilized. Likewise, silencing Stub1 also bolstered Treg function in vitro and in vivo (108). In all, these studies identify an unappreciated pathway for the regulation of Treg function by environmental cues working at the protein level.
A number of unanswered questions exist concerning Stub1 and ubiquitin-mediated Foxp3 downregulation. For instance, to what degree do the levels of Stub1 induced under physiologic conditions disrupt Treg function? Experimental overexpression approaches do not address this. Our findings suggest that high expression of Stub1 should precipitate dramatic loss of Treg function. It is possible, however, that low or intermediate Stub1 activation (potentially in response to proportionally moderate inflammation severities) only incrementally reduces the transcriptionally active pool of Foxp3. This, in turn, begs the question of what consequences for Treg function arise from such an intermediate or temporary Foxp3 loss? It is conceivable that a partial or brief deactivation of Treg provides a respite from Treg-mediated suppression following Foxp3 downregulation may allow the effective initiation of an effector response. The likely more resilient mechanisms promoting Foxp3 transcription, which may remain active in Tregs independent of a reduced Foxp3 protein pool (or partially so), could ensure that even though Treg function may be paused, a hardwired blueprint of suppressor phenotype persists, providing for a ready pathway back to the restoration of suppressive function and immune homeostasis. Indeed, the intersection of protein and transcription-level regulatory mechanisms in this and other aspects of Treg biology are ripe for speculation and future investigation.
Widespread Treg deactivation in response to inflammatory cues is certainly not evolutionarily tenable. These cells must be able to function in inflamed tissues to correct ongoing immune responses. Controlled, reversible release of T-effector genes (involved in proliferation, trafficking, survival, etc.) from Foxp3-mediated repression, however, may ensure the transition from ‘peacetime’, resting Tregs to activated or ‘effector’ Tregs. This subset of Foxp3 Tregs capable of expressing the transcription factors and chemokine receptors more traditionally viewed as effector T-helper cells are thought to be more adept at reaching inflammation sites, proliferating, and controlling inflammation in peripheral tissues (112).
In the face of danger signals, a fraction of the Tregs that fail to upregulate Stub1 (or are for some other reason insensitive to Foxp3 protein loss) may come to represent a greater proportion of the total Treg complement. Such a selection process would eventually yield a more stable and likely more suppressive Treg pool. Indeed, the observation that a small population of Tregs retains Foxp3 in the face of forced Stub1 expression (108) is compatible with this notion. Alternatively, it is possible that additional pathways for the induction of new suppressor cells or their expansion could be at play to prevent a complete loss of immune control. Whether or not potential specialization, pausing, or selective mechanisms can be pushed too far by correspondingly extreme inflammation or activating signals to cause long-lasting Treg dysfunction, like that seen in autoimmune disease, remains to be explored.
While the findings mentioned above reveal that Stub1 is an important regulator of Foxp3 and the function of established Tregs, there is reason to suspect that it may play different roles in other aspects of Treg biology. Recently Stub1 was found to mediate the activating, nondegradative ubiquitination of CARMA1, a participant in NFκB signaling (88, 89, 113), which is important for initiating Foxp3 transcription in developing tTregs (114). This finding suggests that Stub1 may unexpectedly contribute to Foxp3 induction in developing tTregs. However, since overzealous activation of this pathway, perhaps by robust TCR/CD28 stimulation, can interfere with extrathymic Treg induction (115) and the stability of Foxp3 expression by mature Foxp3+ Tregs, a critical role for Stub1 in NFκB signaling is compatible with the notion of it being a Foxp3-destabilizing force in the periphery. Recently, a study speaking to this issue reported a role for the E3 ligases Stub1, Cbl-b in the regulation of Foxp3 protein in thymic Tregs.
In addition to IL-2 signaling, costimulation through CD28 is necessary for Treg generation in the thymus and is thought to be important for their persistence in the periphery as well. For example, CD28−/− mice have reduced Treg populations in the thymus and spleen. Interestingly, these mice also display enhanced Cbl-b levels. Zhao et al. (97) recently found that simultaneous knock out of Cbl-b in CD28-deficient mice can partially rescue Treg generation in the thymus. These authors would go on to report that the enhanced expression of Foxp3 in these double knockouts reflect a ubiquitin/proteasome-dependent downregulation of Foxp3 by Cbl-b. The ligase was found to interact with Foxp3 and, in concert with Stub1, facilitate the polyubiquitination and degradation of Foxp3 in tTregs upon TCR activation (97). Supporting this notion, shRNA knockout of either Stub1 or Cbl-b could stabilize Foxp3 protein levels and reduce the extent of Foxp3 polybuiquitination. Moreover, in vivo administration of the proteasome inhibitor PS-341 (bortezomib) to CD28−/− mice corrected their defect in tTreg generation in spite of elevated Cbl-b expression, confirming the involvement of this degradative pathway in Foxp3 regulation in the thymus. Instead of a predominantly positive role for Stub1 in the CARMA1-mediated priming of Foxp3 transcription in tTregs, these results seem to suggest that its role as a protein-level antagonist is active and in concert with Cbl-b, may oppose the pro-Treg effects of costimulation in the thymus. Also noteworthy is the contrasting roles played by Cbl-b in peripheral and thymic Foxp3 induction: positive in the former, and apparently negative in the latter.
We have also found ubiquitin-involved mechanisms of posttranslational Foxp3 regulation to be at work during T-cell differentiation. Naive CD4+ T cells are capable of acquiring remarkably specialized effector functions in response to unique cytokine signals. The notoriously proinflammatory Th17 lineage and the immune-suppressing iTregs, while functionally opposite, share common elements in their differentiation pathways. TGFβ, for instance is required for both Th17 and iTreg generation (116). High concentrations of TGFβ among other factors can sustain Foxp3 expression and commitment to the iTreg fate. STAT3-activating cytokines, such as IL-6, in contrast promote RORγt and IL-17 upregulation (116, 117). Since Foxp3 expression has been noted in the early stages of both lineages, and Foxp3 is known to actively suppress RORγt driven expression of Th17-associated genes (118), timely removal of Foxp3 protein from cells poised at the crossroads of these divergent fates is necessary for optimal commitment to Th17 differentiation.
We found that the oxygen sensor and mediator of the cellular response to hypoxia known as hypoxia-inducible factor -1 (HIF-1) plays a key role in this process. The ability of HIF-1 to act as a transcriptional activator is controlled by a ubiquitin-dependent process. In the presence of oxygen, critical proline residues on HIF-1’s α subunit (HIF-1α) are hydroxylated by the prolyl hydroxylases or PHDs. These modified residues serve as a docking site for the Von Hippel-Lindau protein (VHL), which recruits the Elongin-C-Elongin-B-Cullin-2-E3- ubiquitin ligase complex that polyubiquitinates HIF-1α. Addition of this K48-type polyubiquitin chain marks HIF-1α for degradation via the 26S proteasome. When oxygen is scarce, unhydroxylated HIF-1α accumulates, associates with its constitutively expressed β subunit, and mobilizes the cellular response to hypoxia and a range of stimuli through the activation of numerous metabolic and immunologically relevant genes (119, 120).
During Th17 differentiation, HIF-1 is induced in naive CD4+ T cells, even in the presence of oxygen in a STAT3-dependent manner. Suggesting that HIF-1 is important for Th17 differentiation, HIF-1α-deficient naive CD4+ T cells (from CD4cre+/HIF-1αflox/flox mice) display stunted upregulation of Th17 genes, including those encoding RORγt and IL-17 compared to wildtype cells. Further investigation would reveal that HIF-1 promotes expression of several Th17-linked genes through the activation of RORγt expression and function (121).
In the absence of HIF-1, T cells show reciprocal upregulation of Foxp3 protein but not transcript (121), suggesting a HIF-1 dependent mechanism for rapid downregulation of Foxp3 at the crossroads of Th17 and iTreg differentiation. Similar results were also reported by Shi et al. (122). In our study, the mechanism behind this effect was found to depend upon HIF-1’s ability to physically interact with Foxp3 as well as its own ubiquitin-mediated degradation pathway. Mutant HIF-1 molecules rendered insensitive to oxygen-dependent modification and degradation failed to bring about Foxp3 downregulation. Knocking down components of the HIF-1 degradation machinery and chemical inhibition of the proteasome also stabilized Foxp3 levels. Furthermore we found that Foxp3 itself can be ubiquitinated in iTregs. These findings suggest a model of ‘piggy-back’ co-modification/co-degradation as a means of HIF-1-mediated Foxp3 protein loss (121). The molecular events involved in such a co-degradation scenario remain to be elucidated. The identity of the E3 ligase actually responsible for Foxp3 modification, for instance, is not definitively known.
The influence of this HIF-1 mediated, ubiquitin/proteasome-dependent pathway on the stability of the Foxp3+ protein pool of established Tregs was not directly addressed. Intriguingly, HIF-1 has been reported to positively affect transcription of the Foxp3 gene with potential consequences for the stability of Treg function under inflammatory conditions (123). It is possible that the transcriptional and protein-level pathways controlling Foxp3 expression may be poised in opposition and uncharacterized factors may determine the winner of such a ‘tug-of-war’. While HIF-1 deficient T cells, in our hands, upregulate Foxp3 message on pace with their wildtype counterparts during in vitro T-cell differentiation (121), under different conditions (i.e. in vivo inflammation, extreme hypoxia, etc.), HIF-1 driven transcription at Treg-critical loci such as Ctla4 and Foxp3 might be more necessary for optimal Treg stability.
Speaking to the role of HIF-1 in Treg stability, a recent report by Hsiao et al. (124) showed that elevated HIF-1 levels actually have negative effects on Foxp3 expression and the stability of Treg function. In this study, knocking out a promoter of HIF-1 degradation, known as Deltex (DTX1), in a Treg-restricted manner did not remarkably change the baseline Foxp3 levels, expression of Treg-associated molecules and co-regulatory factors, or the in vitro suppressive function of Tregs. DTX1-deficient Tregs, however, were shown to be less effective suppressors than their wildtype counterparts in vivo (using models of airway inflammation and colitis). This deficiency was attributed to unstable expression of Foxp3 seen along side the elevated HIF-1 levels in the absence of DTX1. Importantly, the authors showed that simultaneous HIF-1- and DTX1-deficiency in Tregs restored their in vivo Foxp3 expression level and largely corrected the inability of DTX1-deficient Tregs to prevent colitis (124). Interestingly, DTX1 did not appear to drive the direct ubiquitination of HIF-1. Instead it was found to bind HIF-1 promoting hydroxylation of proline residues. The authors also showed that DTX expression could prevent hypoxia-induced reduction of the Foxp3 protein pool in T cells (124). These results do not appear at first glance to support a model for Foxp3-HIF-1 co-degradation via the proteasome as, in the absence of DTX1, Foxp3 protein loss was more pronounced along side reduced HIF-1 degradation. With many of the details of DTX1’s Foxp3 stabilizing effects still to be dissected, it is possible that intricacies of DTX and PHD-mediated proline hydroxylation have unique and presently unknown consequences for the Foxp3-HIF-1 interaction or polyubiquitin-tagging of Foxp3. It is also possible that some HIF-1 degradation is still occurring even though DTX1-deficiency greatly enhanced the size of the HIF-1 protein pool, providing for some co-degradtion to deplete Foxp3. This in combination with the likely increased expression of HIF-1 activated genes could contribute to an instable Treg phenotype. Regardless, this study’s findings at least suggest that high levels of HIF-1 protein can lead to unstable Foxp3 expression by committed Tregs – an important step in understanding the potentially dynamic role of HIF-1 in Treg phenotypic stability.
The duration of hypoxia appears to be crucial in determining the overall impact of HIF-1 in T cells. Prolonged hypoxia halts HIF-1 degradation due to the stalled action of the PHDs, VHL, and the proteasome, and in theory the co-degradation of Foxp3 as well. This would likely lead to accumulation of both proteins and HIF-1’s transcriptional role may become more prominent. Unpublished results from our group are in line with this. We find that transgenic mice constitutively expressing a hydroxylation resistant HIF-1α molecule (impervious to polyubiquitin-driven degradation) do display modestly enhanced Foxp3 levels under certain conditions. Curiously, this is seen along side uncharacteristic, effector T cell gene expression (our unpublished results). This may be due to transcriptional direction of effector T cell genes by HIF-1 or the metabolic consequences of HIF-1 stabilization. Indeed, while the metabolic programming accomplished by HIF-1 in effector T cell differentiation may be wholly incompatible with the preferred lipid oxidation dominated metabolism of Tregs (125).
In contrast, during intermittent or transient hypoxia, or upon the normoxic induction of HIF-1 by cytokines, active oxygen-dependent turnover of suspected HIF-1/Foxp3 complexes may shift the relative importance to the posttranslational role in downregulating Foxp3. While several studies employing cycling hypoxic and normoxic culture conditions have found that Th17 differentiation can be enhanced under these conditions in a HIF-1-mediated fashion (126, 127), the effects of these conditions on levels of Foxp3 protein deserve further investigation.
Discovery that a hypoxia-induced miRNA known as miR210 actually targets the HIF-1 transcript adds another layer of complexity to the study of HIF-1 and hypoxia duration and the stability of Tregs. In this study, hypoxic priming (24 hours of hypoxic culture) resulted in a surge in HIF-1 level and, in agreement with a previous study (126), an enhancement of Th17 skewing upon reoxygenation of differentiating T cells. Longer hypoxic culture periods, however, saw less of build up of HIF-1 coincident with increased miR210 upregulation. Knocking out miR210 resulted in even higher levels of HIF-1 and HIF-1-dependent Th17 differentiation following hypoxic priming compared to those seen in wildtype T cells (127). While the existence of this message-level feedback mechanism complicates the study of hypoxia duration on the posttranslational actions of HIF-1, they do provide a nice example of how prolonged versus transient HIF-1 can have significantly different outcomes.
Ubiquitination of target proteins does not only lead to degradation. Indeed, nonproteolytic ubiquitination can interfere with protein-protein interactions (128) or may even enhance the activity or direct the cellular trafficking of key signaling molecules in leukocytes (91). While the mechanisms of K48-type Foxp3 modification described above function to dwindle cellular pools of the transcription factor, recent work indicates that Foxp3 levels and activity can be positively affected by a unique pathway of ubiquitination.
The E3 ligase known as TNF receptor-associated factor 6 (TRAF6) is an important signaling adapter molecule in the cascades triggered at the IL-1R, TLRs, and other TNFR superfamily members (e.g. CD40, TRANCE) (129). TRAF6 is unique among the TRAF family member since it participates in both CD40/TRANCE and TLR signaling pathways through directly interacting with receptor molecules or adapter factors like MyD88, respectively.
CD40 and TLR signaling is important for the activation or ‘licensing’ of antigen presenting cells including DCs to produce inflammatory cytokines and become effective primers of T cells (129). Reflecting its importance in this process, TRAF6 deficiency in DCs impedes activation induced MHCII upregulation and production of IL-6 and IL-12 (130).
Despite this role in the priming of a T-cell-mediated immune response, TRAF6 deficiency in mice results multi-organ autoimmunity (131, 132). Additionally, TRAF6 deficiency has been shown to result in enhanced Th17 commitment (133). The role played by TRAF6 in setting the balance between immune activation and restraint, and its role in Tregs is therefore the subject of much investigation.
Previously, Shimo et al. (134) found that TRAF6 deficient mice harbor reduced Tregs with reduced Foxp3 expression levels, primarily in their thymi compared to wildtype mice. Interestingly, in this and another study, loss of TRAF6 did not hinder pTreg levels or TGF-β-driven iTreg differentiation (133). While these findings suggest that TRAF6 is uniquely important for the TCR/NFκB- mediated induction of Foxp3 transcription and generation of tTregs, an additional role for this E3 ligase specifically in maintaining the suppressive phenotype of Tregs in the periphery was recently suggested (135).
Muto et al. (135) recently found that mice lacking TRAF6 specifically in Tregs (Foxp3cre+/TRAF6flox/flox) are prone to dermatitis and multi-organ inflammation. T cells from these mice were also shown to display elevated activation markers at baseline suggestive of insufficient immune control. Curiously, they found that Tregs were actually present at higher frequencies in the conditional knockout mice. However, the authors found that Foxp3 expression by these TRAF6-deficient Tregs was resoundingly unstable in vivo, and despite being suppressive in vitro, these Tregs fail to control naive T-cell-induced colitis in lymphopenic mice. Additionally, TRAF6 deficiency led to considerable upregulation of effector T-cell-associated transcription factors and cytokines by Tregs (e.g. GATA3, Tbet, IL-4, IFN) (135), reinforcing the importance of TRAF6 in maintaining the characteristic gene expression of Tregs. While these findings clearly implicate TRAF6 as a Treg-stabilizing factor, the mechanism by which it promotes continued Foxp3 expression remained uncertain. Our group has uncovered an unappreciated, ubiquitin-mediated pathway relevant to TRAF6 and Tregs.
Previously, we found that Foxp3 was subject to K48-linked polyubiquitination resulting in degradation of the transcription factor (108, 121). Further investigation has revealed that Foxp3 is also subject to K63-type ubiquitin tagging detectable even the sites of K48-linked polyubiquitin chain attachment are mutated. Screening of multiple E3 ligases for the ability to interact and modify Foxp3 revealed that TRAF6 was in fact capable of facilitating this additional mode of ubiquitination.
Since K63-type ubiquitination can dictate the intracellular distribution and signaling activity of numerous target proteins (128), we hypothesized that modification of Foxp3 by TRAF6 promotes the nuclear localization and gene regulating activity of the key transcriptional factor. Indeed, immunostaining of cell lines and primary T cells reveal that upon genetic TRAF6 deficiency, Foxp3 occurs diffusely throughout the cell, while TRAF6 competent cells, as expected, display primarily nuclear Foxp3 staining. Given this aberrant localization of non-K63 modified Foxp3, TRAF6 deficiency markedly disrupts Foxp3’s capacity to suppress target gene transcription in reporter assays (authors’ unpublished results).
Demonstrating the importance of TRAF6 in Tregs, TRAF6f/f Foxp3Cre+ mice display heightened leukocyte proliferation at baseline and fail to support the growth of implanted B16 melanoma cells, unlike their wildtype counterparts. In keeping with the notion that TRAF6 is necessary for Treg-mediated immune restraint, the severely stunted tumor progression in TRAF6f/f Foxp3Cre+ mice coincided with a markedly enhanced anti-tumor response evidenced by heightened production of proinflammatory cytokines by leukocytes and reduced Treg frequencies infiltrating the tumor and lymphoid tissues (authors’ unpublished results). These findings largely support those of Muto et al. (135) and confirm the importance of TRAF6 as a stabilizer of the Treg phenotype across diverse immunological scenarios. Importantly, we reveal that this phenomenon involves an unanticipated role for TRAF6 as a director of Foxp3 cellular distribution that is dependent on a unique, ubiquitination-mediated event.
We found that TLR/MyD88 signaling was important for the upregulation of Stub1 in Tregs and the downregulation of Foxp3 protein (108). Since TRAF6 is a known participant in the signaling cascades set off by the same inflammation-associated stimuli that appear to negatively impact Foxp3 through degradative polyubiquitination, it is possible that the context of these signaling pathways may be critical in determining their ultimate consequences for Tregs. Indeed since TRAF6 sits at the convergence point of multiple signaling pathways, integration of distinct signals, possibly depending upon unique adapter molecule activity, may afford stimuli-specific nuance to TLR and TRAF6 activation (136). It is also important to consider the potential MyD88 and TRAF6-independent aspects of certain TLR cascades (TLR3, TLR4) as well as the recently uncovered cross-talk with STAT pathways (137). It is also possible that TRAF6’s pro-Treg effects stem from a unique aspect of non-TLR-triggered signaling pathways involving the molecule such as those arising from CD40-CD40L interaction (138).
While these results present TRAF6 as an enticing target for anti-cancer immunotherapy aimed at breaking tolerance, the well-known involvement of this ligase in NFκB signaling pathways is cause for reservation. Should TRAF6 inhibition hinder effector T-cell activation (downstream of the TCR, TLR, or inflammatory cytokines, for example) while undermining the functional stability of Tregs, the counterproductive effects may prove an ineffective means of enhancing anti-tumor immunity.
Encouragingly, TRAF6 has been reported to be dispensable for antigen-initiated activation of T cells in the peripheral tissues and for NFκB activation in these cells (131, 139). In agreement, we find that TRAF6 loss by the entire T-cell compartment (achieved CD4+cre/TRAF6flox/flox mice), largely phenocopies Foxp3-cre-restricted TRAF6 deficiency in terms of impaired support of implanted tumors and Treg function, and enhancement of inflammatory cytokine production as well (authors’ unpublished results). Muto et al. (135) have also observed enhanced, rather than inhibited leukocyte activation and inflammation upon whole TRAF6 knockout (135). In light of this, it stands to reason that Treg-mediated immune restraint and not immune activating mechanisms may be uniquely dependent on TRAF6 – making it likely that chemical inhibition of TRAF6 will be a highly effective strategy for unleashing the anti-tumor immune response.
It is clear that several E3 ligases strongly influence the generation and function of Tregs either through the regulatory modification of signaling molecules important for these cells or by directly regulating the Foxp3 protein pool. The process of protein ubiquitination is, however, a highly reversible one. Removal of ubiquitin tags from modified proteins is catalyzed by substrate-specific proteases known as deubiquitinases (DUBs) (140). Given the importance of E3 ligases for Treg relevant processes and Foxp3 itself, it should not surprise that enzymes capable of undoing their action should be equally important in the biology of Tregs and immune control.
One such DUB is known as cylindromatosis tumor suppressor (CYLD). NFκB, TLR, and TNF signaling pathways are regulated by this DUB as it targets the intermediates TRAF2 and NEMO with significant consequences for Tregs. CYLD also negatively regulates signaling and TGFβ responsiveness, suppressing extrathymic Treg induction and the pTreg pool. CYLD knockout mice display enhanced peripheral Tregs and T cells from these mice are more apt to upregulate Foxp3 ex vivo (141, 142). By antagonizing NFκB signaling, CYLD appears to inhibit tTreg generation as CYLD deficiency or mutations impairing target binding are accompanied by elevated Treg frequencies (141, 143). Despite yielding increased Treg presence, however, CYLD dysfunction appears to result in less functional Tregs (143), possibly reflecting the distinct consequences NFκB overactivation can have for Treg development and maintenance.
The Th2 factor GATA3 is upregulated in Tregs upon TCR activation, and it contributes to Treg phenotype stability and Foxp3 expression (144, 145). Recently, another DUB known as USP21 was found to deubiquitinate GATA3 stabilizing its expression in human Tregs. Interestingly, Foxp3 directly activates expression of USP21, and siRNA knockdown of USP21 downregulates both GATA3 and Foxp3 protein. It is likely that by stabilizing GATA3 levels, USP21 indirectly supports Foxp3 expression and Treg function (146).
Since direct ubiquitination of Foxp3 can drive its degradation, it stands to reason that counteracting this process should preserve levels of this important regulatory hub of the Treg phenotype. Indeed, van Loosdregt and colleagues (107) recently discovered a specific DUB capable of reversing Foxp3 polyubiquitination, preserving both its expression and Treg function. Based on past observations of Foxp3 polyubiquitination and its negative association with Foxp3 stability (76, 121, 147), the authors set out to identify the DUB responsible for modulating Tregs. Confirming the importance of deubiquitination for stable Treg function, treatment of human Tregs with a pan-DUB inhibitor reduces their effectiveness in an in vitro suppression assay. Furthermore, mass spectroscopy analysis identified USP7, a DUB also reported as a stabilizer of NFκB family members (148), to be not highly expressed by Tregs and capable of interacting with Foxp3 and mediating its deubiquitination. Linking the specific activity of this DUB to stable Treg function, the authors of this study showed that knocking down USP7 diminishes both Foxp3 levels and Treg suppression in vitro, while its overexpression stabilizes and augments the Foxp3 protein pool and suppressive function (107). Strikingly, the ability of murine Tregs to suppress colitis in vivo was found to be significantly reduced upon pre-treatment with either the pan-DUB inhibitor or shRNA against USP7 prior to lymphopenic recipients of colitogenic naive CD4+ T cells (107). These findings support a role for the cellular ubiquitination/protein-degradation machinery in regulating Foxp3 in Tregs. They also suggest that DUB targeting may be an effective strategy to break Treg-enforced tolerance to fight cancer.
Recent work from our group and our collaborators suggest the existence of an additional DUB working to directly stabilize Foxp3 protein and Treg function. The DUB known as USP44, which has previously been studied in the context of chromosomal segregation and the development of T-cell leukemias (149, 150) is highly expressed by Tregs. Knockout of USP44 renders mutant Tregs less stable and less effective than their wildtype counterparts in vivo (our unpublished results). These results suggest that multiple DUBs may contribute to the deubiquitination-dependent preservation of the Foxp3 protein pool. This may reflect a need for redundancy, a conceivable notion given the importance of stable Foxp3 expression for proper Treg function, or it may suggest a condition- or Treg subset-specific reliance on individual DUBs. These possibilities are under investigation.
These studies and others that describe ubiquitin-dependent Foxp3 downregulation demonstrate the dynamic nature of Foxp3 protein expression and make clear the importance of protein-level Treg modulation. They also reveal a means of overcoming inflammation-induced disruption of the Treg phenotype. These findings suggest an exciting avenue for novel immunotherapies. Specifically, the targeting of Foxp3-preserving DUBs may be an effective means to break tolerance and muster an effective anti-tumor immune response in cancer patients.
There are numerous examples of ubiquitin-sensitive pathways and factors that are highly relevant to the multifaceted biology of Tregs. Some influence signaling pathways critical for Foxp3 induction (e.g. TGFβ and NFκB signaling), while others involve the direct ubiquitination of the ‘master’ regulator Foxp3. In this review, we summarize findings showing that distinct Foxp3-ubiquitinating pathways exist with sometimes starkly contrasting effects on the induction, functional potency, and stability of the Foxp3-protein pool in Tregs (Fig. 1). These findings suggest that specific targeting of pathways responsible for the addition or removal of specific ubiquitin modifications on this key Treg transcription factor will have broad immunotherapeutic relevance. Ubiquitin-mediated Foxp3 regulation, only recently brought to light, awaits further exploration. A number of pressing issues and possibilities exist at the writing of this review.
A pressing question asks how these pathways of ubiquitin-dependent Foxp3 regulation (posttranslational mechanisms in general) fit into our evolving concepts of Treg functional stability, heterogeneity with the Foxp3+ Treg population, and the potentially dynamic behavior of Tregs during immune responses.
It is now appreciated that considerable heterogeneity exists within Foxp3+ Tregs. tTregs and pTregs differ markedly in the factors governing initiation and maintenance of Foxp3 transcription. Also, the contributions of these subsets to general and regional immune homeostasis also appear to differ (50). Furthermore, among Tregs, the stability of Foxp3 expression and suppressive function has been shown to vary and appear positively associated with the relative expression of CD25 (43, 61).
It is possible that heterogeneity also exists among Foxp3+ Tregs in terms of their susceptibility to posttranslational modes of Foxp3 regulation. With the revelation that unique pathways of ubiquitination-driven Foxp3 degradation exist, it is tempting to speculate that differential expression or activity of key molecular participants in these pathways could spell the difference between a Treg likely to lose Foxp3 protein upon contacting inflammatory cues and a completely stable one. In this way, the balance between Foxp3-stabilizing (TRAF6/USP7-mediated) and Foxp3-degrading (Stub1-mediated) processes may contribute to the functional heterogeneity of Foxp3+ Tregs.
In keeping with the notion that Foxp3+ Tregs are a considerably diverse population of suppressor cells, both human and mouse Tregs appear to be able to take on some of the attributes of the specialize T-helper lineages while interestingly maintaining suppressive function (151, 152). It has been suggested that to control highly specialized T-helper responses, Tregs must take on shades of these lineages. STAT3 and Tbet expression, for instance, has been reported to be critical for control of Th17- and Th1-dominated responses, respectively (112, 153). Also in the generation of activated or ‘effector’ Tregs from ‘central’ or resting Tregs, Tregs dramatically change their localization and role in promoting immune homeostasis upon exposure to inflammation-associated stimuli (154). It is possible that posttranslation mechanisms, including ubiquitin-mediated degradation of Foxp3, may provide temporary deviation from the normal Treg gene expression pattern. Reducing Foxp3’s enforcement of Treg-associated gene expression patterns by downsizing levels of the transcription factor may very well allow for some limited expression of T-effector genes (Tbet, CXCR3 etc.), the elements of suppressor cell specialization. Future investigations will likely determine what role, if any, is played by these mechanisms in the apparently varied and dynamic behavior of Tregs.
Another evolving concept in Treg biology is the responsiveness of these cells to environmental cues. While full blown plasticity of the Treg lineage remains controversial (155), there are numerous examples of both up- and down-modulation of Treg function in response to external inputs (156). Since a recurring theme in the growing literature concerning posttranslational Foxp3 regulation is the direction of these processes by external stimuli, it is possible that they are central in the modulation of Treg function by the microenvironment. Inflammatory cues including proinflammatory cytokines (IL-6, TNFα) have been reported to drive dephosphorylation and degradative polyubiquitination of Foxp3 (82, 108) while counteracting the Foxp3-stabilizing/enhancing acetylation modifications (77). Certain ‘danger signals’ including microbial products and stresses (LPS. heat shock, hypoxia) also appear to engage pathways leading to posttranslational downmodulation of Foxp3 expression and Treg function (107, 108, 124).
It was recently found that Treg transcriptional character could be galvanized by exposure to intense inflammation. In a study by Arvey et al. (157), a wholesale loss of immune homeostasis was induced by the temporary removal of Tregs in mice. The Tregs that repopulated these mice showed increased suppressive potential as well as strict adherence to Treg-associated patterns of gene expression (157). This study suggests that unspecified inflammatory elements from the microenvironment may positively impact the Treg pool. When taken together with the revelation of inflammation-triggered, posttranslational pathways to Foxp3 downregulation, it also suggests that despite being functional disabled, Tregs as a population are likely to regain their suppressive function and control over the immune system. As discussed earlier, the highly reversible and timely action of protein-level Foxp3 downregulation likely to be operating alongside those governing transcription at the Foxp3 gene, may provide a mechanism for pausing the suppressive action of Tregs while maintaining much of the transcriptional landscape of these cells. Hypothetically, this temporary reduction in the active Foxp3 protein pool could allow for (i) greater amplitudes of immune activation needed to combat invading threats; (ii)) less-restrained expansion of Tregs, allowing them to keep pace with effector cells; or (iii) some diversification of their gene expression profile to enhance Treg fitness or trafficking potential (Fig. 2). While the immune homeostasis is possible in this model upon removal of the external signals leading to Foxp3 protein loss, it is also possible that extreme insult may irrevocably compromise elements of the Treg suppressive repertoire resulting in immune pathology. A better grasp of the side-by-side action of transcriptional and posttranslational paths of Foxp3 regulation will no doubt shed light on this subject in the near future.
In the cancer setting, Tregs tend to be enriched, either locally in developing tumors or systemically at advanced stages, and the immune-suppressing nature of these cells hampers the anti-tumor response permitting tumor growth (158). Sabotaging Tregs or blocking their function has been explored as an immunotherapeutic measure to enhance the anti-tumor immune response and increase the effectiveness of anti-cancer therapies including tumor-vaccines. An obvious drawback of prolonged or extensive Treg depletion or broad inhibition of immunoregulatory checkpoints is the potential for a loss of immune control and immune pathologies.
Fortunately, complete or continued Treg ablation may not be necessary for a timely and effective mobilization of immune forces against the tumor. In mice, transient Treg depletion is adequate to suppress tumor burdens and enhance the effect of interventions including radiation therapy and an anti-cancer vaccine (159, 160). Importantly, a single infusion of therapeutic anti-CD25 antibody treatment has also shown promise as a means to reduce Foxp3+ Treg frequencies and improve the anti-tumor response in human patients (161). Additionally, localized Treg sabotage may also be enough to efficiently unleash the anti-tumor response (162). Therefore, strategies aimed at brief, temporary, or localized disruption of Tregs or their function may be effective and safe additions to the cancer-fighting immunotherapeutic arsenal.
Ubiquitination is by nature reversible and may, as a pathway of Treg modulation, afford a considerable degree of responsiveness to external stimuli. It is possible, as discussed above, that ubiquitin-dependent downregulation of Foxp3 protein in Tregs results in temporary disruption of their suppressive function. If so, therapeutically activating the factors responsible for this degradative pathway, or targeting those leading to Foxp3-stabilizing posttranslational modifications may provide a sought after avenue for rapid yet transient Treg down-modulation and robust anti-tumor immunity. On the other hand, preserving Foxp3 expression and function through stabilizing ubiquitin-dependent means may be useful in restoring immune homeostasis in patients with inflammatory or autoimmune disease. Further characterization of the molecular players involved in these ubiquitin-driven pathways of Foxp3 degradation and activation may reveal multiple potential targets for therapies aimed at either bolstering or interrupting the tolerance-promoting potential of Treg cells.
Funding support comes from grants from the Melanoma Research Alliance, the National Institutes of Health (RO1AI099300 and RO1AI089830), “Kelly’s Dream” Foundation, the Janey Fund, and the Seraph Foundation, and gifts from Bill and Betty Topecer and Dorothy Needle. FP is a Stewart Trust Scholar, JB is supported by a Crohn’s and Colitis Foundation of America Research Fellowship.
The authors declare no financial conflicts of interest.