T cell factor, the effector transcription factor of the WNT signaling pathway, was so named because of the primary observation that it is indispensable for T cell development in the thymus. Since this discovery, the role of this signaling pathway has been extensively studied in T cell development, hematopoiesis, and stem cells; however, its functional role in mature T cells has remained relatively underinvestigated. Over the last few years, various studies have demonstrated that T cell factor can directly influence T cell function and the differentiation of Th1, Th2, Th17, regulatory T cell, follicular helper CD4+ T cell subsets, and CD8+ memory T cells. In this paper, we discuss the molecular mechanisms underlying these observations and place them in the general context of immune responses. Furthermore, we explore the implications and limitations of these findings for WNT manipulation as a therapeutic approach for treating immune-related diseases.
The WNT signaling pathway (named after a combination of the drosophila Wingless gene and mouse Int gene) is highly conserved between species, where it has been shown to play an essential role in controlling a variety of developmental processes, including asymmetric cell division, stem cell pluripotency, and cell fate specification (1–3). In humans, 19 different WNT family members, which are lipid-modified, secreted glycoproteins, have been identified. Because they are insoluble and therefore hard to purify, it remains difficult to perform in vitro functional assays. Fifteen WNT receptors and coreceptors have been described; therefore, the number of possible WNT/WNT receptor interactions are enormous, adding to the challenges of functional analyses. Uncontrolled WNT signaling has been associated with a variety of diseases, including gastrointestinal malignancy, osteoporosis, and neurodegeneration (1, 4). In addition to the importance of WNT signaling for stem cells, it has also proved indispensable for the development of T lymphocytes in the thymus. Although the role for WNT proteins regulating mature T cell function has remained a relatively understudied area, evidence for the importance of these proteins has been increasing over the last few years. These recent advances will be discussed in this review.
Although alternative WNT signaling pathways have been described, we will only focus on the canonical WNT pathway (Fig. 1). The transcriptional activity of T cell factor (TCF) is induced by direct association with β-catenin (3, 5). The expression of β-catenin levels is carefully modulated by a destruction complex consisting of adenomatous polyposis coli, axis inhibition protein (AXIN), glycogen synthase kinase 3β (GSK3β), and casein kinase 1 (CK1). Extracellular WNT directly modulates β-catenin degradation by the destruction complex. In the absence of WNT, β-catenin is phosphorylated by CK1 and GSK3β, thereby generating a binding site for the ubiquitin ligase β-transducin repeat-containing protein (βTRCP), resulting in polyubiquitination and subsequent degradation of β-catenin by the proteasome. Interaction of WNT with one of 10 known Frizzled receptors in humans, together with its coreceptor receptor-related protein 5 (LRP5) or LRP6, leads to disruption of the destruction complex and inactivation of the kinases. β-catenin is therefore not degraded, and cytoplasmic β-catenin levels rapidly increase followed by nuclear translocation. β-catenin competes with TCF repressor proteins such as Groucho and thereby initiate TCF-mediated transcription (6). The TCF family consists of four members: TCF1 (encoded by the TCF7 gene), TCF3, TCF4, and LEF1, belonging to the high-mobility group (HMG) transcription factors, which associate to specific DNA sequences, which comprise a 5′-AAGATCAAAGG-3′ optimal binding sequence (7, 8). Although the four TCF family members are highly homologous, small structural differences result in distinct affinities when associating with DNA or cofactors (9). In addition, distinct isoforms have been described for each TCF family member as a result of alternative exons, different splice acceptor sites, and alternative promoter usage (10, 11). These isoforms can be divided into two groups: the long isoforms (bigger than 42 kDa), which contain a β-catenin binding site at their N terminus, and the shorter isoforms (25–40 kDa), which lack this binding site and can therefore act as transcriptional repressors (10).
TCF was so named because of its essential role in T cell development and thymocyte proliferation (5). TCF1-deficient mice have a partial block at the double-negative and immature single-positive stages of T cell differentiation in the thymus (5). This phenotype can be restored by full-length TCF1 but not by a TCF1 isoform that lacks the β-catenin binding site, demonstrating β-catenin dependence (12). Although T cell differentiation in LEF1-deficient mice is unaffected, mice lacking both TCF1 and LEF1 display a total block of thymocyte development, indicating functional redundancy between these highly related transcription factors (13). The observation that TCF1 appears more important for T cell development compared with LEF1 is also reflected in their expression levels in the thymus, where TCF1 is expressed at around a 10-fold higher level (14), whereas deletion of TCF3 or TCF4 does not influence T cell development in mice. A similar block in T cell differentiation has been observed using ectopic expression of the cell-autonomous inhibitor of β-catenin and TCF (ICAT) (15) and in mice deficient in WNT1, WNT3a, or WNT4 (16, 17). At the later CD4+CD8+ double-positive stage during thymic development, TCF1 and LEF1 have also been demonstrated to influence T cell differentiation by controlling CD4+ lineage choice. These transcription factors can directly induce the expression of Zbtb7b (encoding Thpok), a gene crucial for promoting CD4+ T cell fate, by binding its conserved enhancer element (18). Similarly, TCF1 and LEF1 control CD8+ fate by repressing CD4+ lineage–associated genes such as CD4, RORC, and FOXP3 via demethylation of their histones due to intrinsic HDAC activity of TCF1 and LEF1 (19). TCF1 was also demonstrated to actively coordinate opening of chromatin rather than only binding to open chromatin, as TCF-1 expression in fibroblasts was enough to create de novo chromatin accessibility sites associated with T cell restricted genes (20). Together, these studies demonstrate that TCF1 and LEF1 can directly control both CD4+ and CD8+ lineage identity.
Although a role for WNT proteins in regulation of T cell development has been extensively demonstrated, surprisingly, the effect of WNT signaling in mature/peripheral T cell biology has remained relatively understudied (21, 22). However, over the last few years, a role for TCF in regulating T cell function has started to gain traction. In this paper, we will review this role of WNT signaling in peripheral T cell populations.
CD8+ T cells
The importance of WNT signaling for peripheral T cells has been best studied in the generation of CD8+ memory T cells, which mediate secondary immune responses against invading pathogens. Mouse CD8+ T cells activated using adenomatous polyposis coli in vitro in the presence of TWS119 or SB216763 (both GSK-3β inhibitors and thus WNT mimetics) show arrested CD8+ development into effector T cell populations. This results in increased numbers of multipotent and self-renewing CD44lowCD62LhighSCA-1highCD122highBCL-2high memory T cells (23, 24). However, neither stimulation with WNT3a nor the use of T cells genetically modified to either activate or inhibit β-catenin was able to modulate altered CD44lowCD62Lhigh expression. These observations may suggest that results with GSK-3β inhibitors may not actually be β-catenin/TCF specific but rather are mediated through a parallel GSK-3β–dependent pathway (Fig. 2).
Memory T cell responses have been predominantly investigated using TCF1-deficient mice. Tcf7−/− (coding for TCF1) CD8+ T cells demonstrated a mild (∼50%) reduction in cellular expansion during primary immune response in both an OVA and a lymphocytic choriomeningitis virus model (25, 26). Upon rechallenge, the secondary expansion was severely impaired in TCF1-deficient CD8+ T cells. Analysis of CD8+ T cell memory phenotypes indicated that Tcf7−/− cells did not make the transition from effector memory to central memory based upon expression of CCR7, CD62L, or KLRG1 markers (25, 26). Ectopic expression of p45 TCF1, a larger variant containing the β-catenin binding site, but not p33 TCF1, a variant lacking the β-catenin binding site, could restore memory in TCF1-deficient cells, demonstrating that association with β-catenin is necessary (26). Correspondingly, secondary expansion was also impaired in mice in which the hematopoietic compartment was deficient for both β-catenin and γ-catenin (26). Consistently, increased TCF1 transcriptional activity, achieved by a combined p45 TCF1 and stable β-catenin expressing double transgene, resulted in significantly better secondary expansion and increased bacterial clearance. Although these experiments using knockout mice provide valuable information, it remains to be determined whether changes in the concentrations of WNT, as a ligand of the pathway, have relevant effects for the memory CD8+ T cells compartment in vivo.
Similar to observations in the thymus, TCF1, and LEF1 appear to have redundant roles in CD8+ memory cell development (27). TCF1 can directly increase the expression of Eomes, which in turn enhance the expression of IL-2Rβ, which is critical for the maintenance of CD8+ central memory cells (25). It has been reported that β-catenin stability can be increased upon TCR stimulation (28). TCR stimulation during the primary infection could, in this manner, activate TCF1 transcriptional activity and subsequently promote the generation of CD8 memory T cells. In addition, it was demonstrated that CD8+ effector cells with high WNT signaling yielded a better memory response compared with effector cells with low WNT signaling (29). Correspondingly, after cessation, virus-specific CD8+ T cells with high TCF1 expression were less exhausted, less differentiated, and demonstrated increased survival and increased recall proliferation compared with TCF1low cells (30, 31). In addition, these TCF1-expressing T cells were essential for CD8+ T cell expansion as a result of PD-1 blockade during chronic infection (32, 33).
One report investigated the role of TCF1 on effector CD8+ T cell differentiation from naive CD8+ T cells. Using Wnt reporter mice (Axin2-LacZ), Tiemessen et al. (34) demonstrate that Wnt signaling activity is lowest in CD8+ effector cells compared with the naive or memory population, which corresponds with the very low expression of TCF1 in this cell subset. Upon in vitro stimulation with anti-CD3/CD28, the percentage of IFN-γ–producing cells was significantly increased in naive CD8+ T cells from Tcf7−/− mice compared with wild-type. In a CMV-specific Tcf7−/−mouse model, CD8+ T cells expanded more rapidly, which correlated with increased expression of TNF-α, IFN-γ, and effector cell marker KLRG1. Reconstitution of different TCF1 isoforms (containing or lacking the β-catenin binding site) all reduced IFN-γ expression, indicating that the differentiation of CD8+ effector T cells can be WNT/β-catenin independent (34). In what form TCF1 functions as a transcription factor in a β-catenin–independent manner and whether TCF1 levels alone regulate the differentiation toward effector or memory remains to be determined. In addition, it would be informative to gain further understanding of the (extracellular) signals that regulate TCF1 expression in naive CD8+ T cells and thereby skew differentiation toward effector cells rather than memory cells.
CD4+ Th1 cells
The CD4+ Th1 cell lineage, which can be characterized by expression of the transcription factor T-bet, is critical for coordinating immune responses against bacteria and protozoa. Th1 cells can produce proinflammatory cytokines, including IFN-γ, TNF-α, and IL-2, which promote CTL proliferation, macrophage activation, and NO. The percentage of IFN-γ–producing T cells is increased in TCF1-deficient mice, and TCF1 mRNA is downregulated in naive CD4+ cells differentiated into Th1 cells compared with undifferentiated cells (Fig. 2) (35). Furthermore, both IFN-γ and T-bet expression is reduced in T cells cocultured with colorectal cancer cells. These tumor cells produce large amounts of WNT ligands, thereby increasing active β-catenin levels in nearby T cells. This effect was reverted by WNT antagonist DKK1, indicating that WNT signaling indeed impairs Th1 T cells differentiation (36). In TCF1-deficient cells, ectopic expression of a TCF1 mutant that is unable to associate with β-catenin–decreased IFN-γ expression, suggesting that the TCF1 function is not solely dependent on β-catenin (35). These observations were strengthened by a study that used mice deficient for specifically the β-catenin–binding p45 large isoform of Tcf7. After viral infection, CD4+ T cells from these mice demonstrated comparable expression of IFN-γ and T-bet. However, the formation of memory Th1 cells was impaired in these mice (37). It is uncommon for TCF1 to be transcriptionally active independently of β-catenin. In addition, there are no confirmed TCF binding sites in the Ifng promoter. Therefore, it is unlikely that the effect of TCF1 on IFN-γ is mediated by direct transcriptional repression. More likely, TCF1 is part of a larger transcriptional complex that together impairs T-bet (38). TCF1 expression is higher in T-bet knockout mice but lowered by re-expression of T-bet in these cells. In addition, T-bet was found to directly associate with the Tcf7 promoter, indicating direct transcriptional repression by T-bet, which is not uncommon for this transcription factor (Fig. 3) (38). The microRNA miR-24 has also been demonstrated to regulate TCF-1. Transgenic mice overexpressing miR-24 in a T cell–specific manner demonstrated increased numbers of IFN-γ–producing Th1 T cells, which was, in part, the consequence of reduced TCF-1 expression (39). Interestingly, miR-24 expression is increased in patients with rheumatoid arthritis and ulcerative colitis, both Th1-associated diseases, suggesting that the miR-24–TCF1 axis might contribute to these disease mechanisms (40, 41). Together, these data suggest that TCF1 fulfills an important role in limiting the capacity of T-bet to induce IFN-γ expression. How TCF1 transcriptional activity in these cells is modulated and under what circumstances this happens remains unclear and requires further research.
CD4+ Th2 cells
The Th2 lymphocyte subset is crucial for effective immune protection against large extracellular pathogens such as parasites. These cells, hallmarked by the expression of GATA-3, secrete proinflammatory cytokines, including IL-4, IL-5, IL-9, IL-13, and IL-25, and one of their major functions is to help drive B cell responses by promoting their proliferation and Ab production.
WNT signaling can modulate the early stages of Th2 development (Fig. 2). TCF1/β-catenin induces expression of GATA-3 (35, 42). Both GATA-3 expression and the expression of its transcriptional target IL-4 is reduced in TCF1-deficient cells, in ICAT transgenic T cells, upon incubation with DKK1 (dickkopf-1, inhibitor of the interaction of LRP5/6 with WNT), or in cells in which β-catenin was knocked down, demonstrating that this pathway is β-catenin–dependent (Fig. 3). Correspondingly, forced expression of β-catenin induced the expression of GATA-3 via direct association of TCF1 with the GATA-3 promoter (35). Furthermore, in an allergic asthma mouse model, IL-4 production was severely impaired in bronchoalveolar lavage fluid of TCF1-deficient mice, indicating that TCF1 deletion impairs Th2 responses in vivo (35).
β-catenin has been demonstrated to associate with special AT-binding protein-1 (SATB1) at the GATA-3 promoter (42). Ectopic expression of SATB1 in T cells increased GATA-3 expression, whereas SATB1 knockdown resulted in lower GATA-3 expression. In addition, stimulation with WNT3a also increased the expression of SATB1 transcriptional targets (42).
Human PBMC stimulated with IL-4, especially in combination with anti-CD3/28, demonstrate reduced expression of both TCF1 and LEF1 mRNA and protein levels (43). This effect of IL-4 is mediated by STAT6 directly associating with the TCF1 locus region (9); however, it is uncommon for STAT6 to be a transcriptional repressor. Maier et al. (9) demonstrated that predominantly, the short isoforms of TCF1 that use alternate promoters compared with the long isoforms are downregulated by IL-4 stimulation. The complex signaling pathways by which IL-4 can specifically decrease the shorter TCF1 isoform therefore still remain to be fully determined. In contrast, it has been reported that LEF1 knockdown in Jurkat T cells results in increased IL-4 mRNA levels (43). In addition, Hossain et al. (44) demonstrated that ectopic expression of LEF1 in differentiated Th2 cells repressed expression of IL-4, IL-5, and IL-13. This was mediated by a direct interaction between LEF1 and GATA-3, which abrogated GATA-3 DNA binding activity. These results are in contradiction with the dogma that WNT signaling favors Th2 differentiation and function, which might be the result of the differences between TCF1 and LEF1. Although both transcription factors are very similar, small differences in DNA affinity or the ability to associate with cofactors could result in distinct functional activities. Another potential explanation could be timing, as most experiments that demonstrate a positive effect of TCF1 on GATA-3 activity were performed during or directly after Th2 differentiation, whereas the two studies that demonstrate a negative effect of LEF1 on IL-4 transcription were performed in a Jurkat T cell line or fully differentiated Th2 cells, in which the expression of Wnt signaling pathway components or cofactors might be different. Additional research is required to fully understand the relevance of WNT signaling for Th2 immune responses. Analysis of the expression of WNTs or TCF1 isoforms in patients with Th2-associated diseases might reveal more insight regarding the role of WNT signaling for immune homeostasis in vivo.
CD4+ Th17 cells
Th17 cells have been associated with many autoimmune diseases, including multiple sclerosis, rheumatoid arthritis, and psoriasis (45). These cells are characterized by the expression of the master transcription factor RAR-related orphan receptor-γt (RORγt) and fulfill a powerful role in microbial protection at epithelial barriers by the secretion of IL-17A, IL-17F, and IL-22. TCF1-deficient mice demonstrate increased percentages of Th17 cells in the thymus and periphery and increased susceptibility for experimental autoimmune encephalomyelitis, a Th17-mediated disease (Fig. 2) (46, 47). TCF1 was demonstrated to associate with the IL-17 promoter, indicating that TCF1 may directly inhibit IL-17 transcription, although it remains unknown whether this is β-catenin dependent (Fig. 3) (46, 47). In addition, Th17 cell differentiation in vitro is significantly improved in TCF1-deficient cells and in the presence of secreted frizzled-related protein 1 (sFRP1), DKK-1, or WNT inhibitory factor 1 (WIF-1), which all inhibit WNT (Fig. 1) (46, 47). Correspondingly, Th17 differentiation was impaired when the T cells were cultured in the presence of WNT3a, WNT5a, or the GSK-3β inhibitor TWS119 (48). Additionally, Lee et al. (48) demonstrated that differentiating naive CD4+ T cells into Th17 cells in the presence of sFRP1 increased TGF-β receptor I, II, and III mRNA expression and SMAD-2 and SMAD-3 phosphorylation, which are crucial mediators in the TGF-β signaling pathway. These results may explain the impaired Th17 differentiation in the presence of WNT, as TGF-β signaling is important for Th17 differentiation. Together, these data suggest that TCF1 functions as a direct transcriptional repressor of IL-17 in the presence of β-catenin.
In contrast, Keerthivasan et al. (49) demonstrated that TCF/β-catenin can induce differentiation toward Th17 cells. A CD4CreCtnnbex3 mouse was used in which β-catenin cannot be degraded. Both the IL-17–positive T cell population and IL-17 serum concentrations in these animals was significantly increased compared to controls. Correspondingly, RORγt expression was increased in CD4+ cells from the gut and spleen of CD4CreCtnnbex3 mice, which was most likely the result of TCF1-mediated transcription, as chromatin immunoprecipitation–sequencing data showed TCF1 to associate with the RORγt locus precisely at loci where H3KAc peaks were enhanced by β-catenin activation. These data are strengthened by a study demonstrating that RORγt expression is decreased in thymocytes of TCF1-deficient mice and increased in β-catenin transgenic mice (although not significant) (50). RORγt promoter activity was increased by coexpression of TCF1 and β-catenin in a luciferase reporter assay but was reduced when TCF1 binding sites in the RORγt promoter were mutated (50).
The different strategies used in gene targeting and discrepancies between the usage of splenocytes or thymocytes could be the cause of the contrasting results observed for the effect of WNT signaling on Th17 differentiation. Therefore, additional research is needed to assess under precisely which circumstances TCF1 can regulate Th17 differentiation.
Follicular helper T cells
Follicular helper T cells, also known as Tfh cells, are mostly located in the B cell follicles of secondary organs such as the spleen and lymph nodes because of their B cell follicle homing receptor CXCR5. Tfh cells provide help for germinal center B cells to differentiate into high-affinity Ab-producing plasma cells and memory B cells through the production of IL-4 and IL-21 and the expression of CD40L. Tfh cells are dependent on their transcriptional repressor BCL-6 for their differentiation from naive T cells, as Tfh differentiation is absent in Bcl6−/− T cells (51–53). Tfh differentiation was reduced in mice with a T cell compartment that was deficient for TCF1 or LEF1 but predominantly in the double knockouts (Fig. 2) (54–56). Similarly, germinal center formation and antiviral responses were impaired in these mice, as was the expression of the Tfh signature genes such as Il4, IL6ra, IL6st, Ascl2, Bcl6, and Icos. These differences were not the result of indirect effects during T cell development in the thymus, as comparable results were obtained in a model where Tcf7 is deleted after thymic development (Tcf7fl/flxhCD2-Cre, hCD2-cKO), in a model where TCF1 was knocked down in the periphery using an RNAi approach, and in a model where Tcf7 deletion was induced by tamoxifen administration immediately before lymphocytic choriomeningitis virus infection (Tcf7fl/flxErt2-Cre) (54–56). TCF-1 was demonstrated to directly repress the promoter activity of BLIMP-1 (Prdm1) and CD25 (Il2ra), which both promote Th1 differentiation and bind the Bcl6 promoter and increase its expression. Therefore, TCF1 can regulate Tfh differentiation through multiple pathways: by inhibiting Th1 differentiation by promoting factors such as BLIMP-1 that also inhibit Tfh differentiation and by directly inducing BCL-6 expression (Fig. 3). Additional research is required to investigate whether this knowledge can be translated to the clinic to increase antiviral responses. Furthermore, better characterization of the main regulators of both TCF1 expression and activity would increase our understanding of the mechanisms/situations that determine whether a T cell will differentiate toward a Th1 or Tfh cell.
Regulatory T cells
Regulatory T (Treg) cells are of critical importance for maintaining immune homeostasis and tolerance through a variety of incompletely understood cellular mechanisms. The transcription factor FOXP3 is crucial for both Treg cell function and development (57, 58). Treg cells can differentiate in the thymus (natural Treg cells; nTreg cells) or differentiate from naive CD4+ T cells in the periphery (induced Treg cells). Tcf1 deficiency increased the differentiation of nTreg cells in the thymus compared with WT mice. Correspondingly, Tcf7 heterogeneity, resulting in reduced Tcf1 expression, also increased nTreg cell differentiation, although less pronounced compared with the knockouts (59). These observations correlate with the fact that Foxp3 can repress Tcf1 expression, resulting in decreased levels of both Foxp3 mRNA and protein in Treg cells compared with conventional CD4+ T cells (59).
Ectopic expression of stabilized β-catenin in murine Treg cells significantly increased survival as a result of upregulation of Bcl-XL and downregulation of Bax, important mediators of apoptosis (60). Increased Treg cell survival also resulted in lower disease scores in an inflammatory bowel disease mouse model. These observations were considered to be the result of increased Treg cell survival because suppression mediated by stable β-catenin–transduced Treg cells was similar to control Treg cells. c-myc expression is normally induced by TCF as a result of β-catenin activation, and surprisingly, c-myc expression was lower in the Treg cells that were transduced with stable β-catenin (61–63). These results suggest that in this experimental setting, the effect of stable β-catenin may not actually be mediated through canonical WNT signaling and TCF activation.
WNT signaling has been shown to abrogate Treg cell–mediated suppression (Fig. 2) (64). Activation of WNT signaling using three independent GSK-3β inhibitors or addition of WNT3a-impaired Treg cell–mediated suppression in vitro. Correspondingly, impairment of WNT signaling using a WNT production inhibitor (IWP-2), a WNT antagonist (FZ8CRD), or TCF1-deficient cells improved Treg cell function. Furthermore, in contrast to wild-type Treg cells, APCmin/+ Treg cells, which have increased levels of β-catenin, could not resolve inflammation in a colitis mouse model. Moreover, treatment of animals with a GSK-3β inhibitor as a WNT mimetic increased the onset of arthritis in an arthritis mouse model (64). Similarly, Foxp3CreCtnnbex3 mice in which β-catenin levels are specifically increased in the Treg cell compartment showed increased inflammation in both the small intestine and colon (49). In addition, Foxp3 expression and Treg cell–mediated suppression was decreased in CD4CreCtnnbex3 mice (49). Interestingly, WNT3a transcript expression was found to be increased in PBMC from inflamed joints of juvenile idiopathic arthritis patients (64). Although Treg cell–mediated suppression in these patients is normally impaired compared to controls, inhibiting WNT production using IWP-2 restored suppression by Treg cells from juvenile idiopathic arthritis patients to control levels in vitro. Analyses of the molecular mechanism behind these observations revealed that TCF1 and FOXP3 can both bind close to the transcriptional start site of more than 900 genes in Treg cells (64). These data suggest a model in which without WNT, FOXP3 is the dominant regulator of these genes because TCF1 is not activated by β-catenin (Fig. 3). Upon infection, extensive Treg cell–mediated suppression would hamper an efficient immune response. Mononuclear cells at the site of inflammation could modulate Treg cell suppressive capacity via WNT production. In autoimmunity, inappropriate WNT production might result in abrogated Treg cell function and uncontrolled inflammation.
Taken together, these studies demonstrate that TCF1 is not only important for regulating T cell development in the thymus but also functions as an important mediator for immunity in mature T cells. TCF activity has been shown to play crucial roles in CD8+ T cell memory acquisition as well as in orchestrating appropriate Th subset responses. Because WNT signaling fulfills such a central role for immune responses, it seems a promising target to manipulate the immune system for therapeutic purposes. For example, the incidence of psoriasis and autoimmune thyroid disease is increased in patients with manic depression that are treated with lithium salts (65–68). The GSK3β inhibitor LiCl was demonstrated to impair Treg cell–mediated suppression and could therefore attribute to the onset of autoimmunity through this pathway (64).
The observation that TCF plays a role in multiple aspects of T cell function suggests that additional molecular mechanisms must drive specificity in distinct T cell subsets under specific conditions. Generation and use of TCF-1 knockout mice for specific T cell subsets would help gain further insight in these control mechanisms. Similarly, inducible knockout or inducible overexpression experiments could help to better understand the complex mechanisms regulating TCF function. The overall outcome of TCF manipulation may be somewhat unpredictable; therefore, additional studies designed to evaluate the overall outcome of manipulating WNT signaling in vivo are warranted. In addition, another possible challenge to overcome is cell type specificity. Increased β-catenin expression is well known to drive colon cancer through enhanced TCF transcriptional activity (69). Increasing WNT signaling to inhibit Treg cell–mediated suppression and promote anticancer immunity, for example, has the caveat that it might promote malignancies as well. Varying levels of WNT signaling was reported to be crucial for the regulation of hematopoietic stem cells, myeloid precursors, and T lymphoid precursors during hematopoiesis (21). This so-called goldilocks phenomenon proposes that only the optimal amount of TCF activity can result in a desired outcome.
Although TCF is generally regarded as a transcriptional activator in the presence of β-catenin, various studies indicate that TCF can repress transcription in different T cell subsets. Although it is uncommon, these studies did demonstrate that β-catenin can indeed mediate transcriptional repression by LEF/TCF through incompletely understood pathways (reviewed in Ref. 70). Because TCF was found to be a transcriptional repressor in different T cell subsets, these alternative pathways that mediate TCF transcriptional repression might be used more often in T cells compared with other cell types. Furthermore, studies have demonstrated a β-catenin–independent activity of TCF. It is important to better understand this form of TCF functionality because these processes would not be sensitive to upstream manipulation of the WNT pathway, such as GSK3β inhibition. It remains unclear how WNT signaling in T cells is regulated by its environment and which cells contribute. Recently, a novel mouse model was developed that allows WNT3 visualization in vivo (71). This epitope-tagged Wnt3 knock-in mouse can now be used to evaluate which cells can produce WNT3 and under which immunological circumstances. Comparable mouse models specific for other WNT proteins are required to provide a better understanding of the complete picture. It was demonstrated that vascular endothelial cells can produce WNT1, 2B, 4, 5A, and 8B, dendritic cells can express WNT5a, and macrophages, WNT7b, but the exact signals that trigger WNT expression remain to be determined (72–74). WNT3A production was found to be increased in CD4+T cells upon TCR stimulation, but whether also other WNTs or WNT-associated proteins are affected by TCR stimulation is still unknown (64). Additional research is also required to define which WNT-Frizzled receptor combinations are relevant for T cells and how inflammatory signals can modulate WNT signaling in vivo.
Although further research is necessary, it is clear that WNT signaling is important for the function of mature T cells. Manipulating this signaling cascade may therefore provide a powerful therapeutic approach to control immune responses in the future.
The authors have no financial conflicts of interest.