Roles of TRAF3 in T cells: many surprises

Abstract

Tumor necrosis factor receptor (TNFR)-associated factor 3 (TRAF3) is broadly involved in different receptor-mediated signaling pathways. Considerable progress was made recently in understanding the role of TRAF3 in T cell biology. Here we review these new findings about how TRAF3 participates in T cell development and function. The different roles of TRAF3 in distinct immune cells are also compared. That TRAF3 is required for T cell effector functions, and invariant Natural Killer T cell function and development, was unexpected. Another surprising finding is that TRAF3 normally restrains regulatory T cell development. It is now clear that TRAF3 regulates signaling to T cells not only through costimulatory members of the TNFR superfamily, but also through the T cell receptor complex, and cytokine receptors. The diverse roles it plays support the multifaceted nature of this molecule. How TRAF3 mediates integration of different signaling cascades is an important topic for future study.

Keywords:

• adaptor molecule
• IL-2 receptor
• invariant Natural Killer T cell
• regulatory T cell
• signaling pathway
• T cell
• T cell receptor
• TRAF3

Abbreviations

TNFR	=	Tumor necrosis factor receptor
TRAF3	=	TNFR-associated factor 3
LMP1	=	latent membrane protein-1
TLR	=	Toll-like receptor
NLR	=	nucleotide binding-oligomerization domain (NOD)-like receptor
RLR	=	retinoic acid-inducible gene (RIG)-I-like receptor
DC	=	dendritic cell
MΦ	=	macrophage
iNKT cell	=	invariant Natural Killer T cell
Treg cell	=	regulatory T cell
TCR	=	T cell receptor
MAPK	=	mitogen-activated protein kinase
NIK	=	NF-κB inducing kinase
IKK	=	IκB kinase
IL-2	=	interleukin-2
LMC	=	litter mate control
T-TRAF3−/−	=	CD4CreTRAF3flox/flox
SLAM	=	signaling lymphocyte activation molecule
IBD	=	inflammatory bowel disease
TCPTP	=	T cell protein tyrosine phosphatase
Tcm cell	=	central memory T cell
Tem cell	=	effector memory T cell
Jak1	=	Janus kinase 1
SOCS1	=	Suppressor of cytokine signaling 1
TFR	=	follicular Treg cell
TFH	=	follicular helper T cell
ICOS	=	inducible co-stimulator

Introduction

Tumor necrosis factor receptor (TNFR)-associated factor 3 (TRAF3) was initially discovered as a novel protein associating with the intracellular cytoplasmic domains of CD40 and its viral mimic, the Epstein-Barr virus latent membrane protein 1 (LMP1).^1-4 TRAF family proteins are defined as possessing a TRAF homology domain at the C-terminus, with the exception of the latest family member, TRAF7. Most TRAFs also have N-terminal Zn ring and Zn finger domains.^5-9 These structural similarities support the notion that they might be involved in similar signaling pathways. In fact, most TRAFs do bind to the cytoplasmic domains of one or more members of the TNFR superfamily, and participate in the signaling pathways mediated by these receptors. However, the nature of their regulatory functions is quite diverse, including both positive and negative regulation for distinct receptors and different cell types.^9,10 Accumulating data show that TRAFs also participate in signaling by receptors that do not belong to the TNFR family. These include several families of innate immune receptors, the Toll-like receptor (TLR) family, nucleotide binding-oligomerization domain (NOD)-like receptors (NLRs) and the retinoic acid-inducible gene (RIG)-I-like receptor (RLR) family.^11-13 TRAF molecules also play important roles in regulating signaling by cytokine receptors, such as the IL-17R in myeloid cells,¹⁴ and the IL-2R in T lymphocytes.¹⁵ The large repertoire of receptors that utilize and depend on TRAFs for signal transduction explains why these proteins regulate a plethora of biological functions. TRAF3 is one of the most multi-functional TRAF molecules.^8,16-19

TRAF3 forms a trimer in solution, stabilized by the N terminal α-helical segments. The entire structure resembles a mushroom, and is closely similar to the structure of TRAF2.^20-22 The C-terminal portion contains several ‘hot-spots’ for protein-protein interactions.^20,23-26 The TRAF domain mediates such interactions, including TRAF homo- and hetero-oligomerization, as well as interactions with receptors and other signaling molecules.²¹ Recent studies show that the TRAF N-terminus also plays critical roles in protein-protein interactions and other functions, including serving as an E3 ubiquitin ligase.^15,27,28 Thus, TRAF3 functions in many different ways to regulate signal transduction and the structural properties for different functions require more precise interpretation.

For a considerable time following its identification, the roles of TRAF3 in immune cell functions were not well-studied compared to TRAF2 and TRAF6. A major underlying reason for this neglect of TRAF3 is likely the manner in which TRAF functions were initially investigated, by exogenous overexpression in non-lymphoid cell lines such as 293T and HeLa. Immune functions could not be studied in such cell types, so functional readouts were typically activation of NF-κB or c-Jun kinase (JNK) reporter genes. Unlike TRAFs 2 and 6, TRAF3 doesn't activate these reporters, so may have been considered uninteresting.^29-32 In addition, whole body depletion of TRAF3 in mice results in early lethality, preventing use of this model to delineate the roles of TRAF3 in different mature cell types.³³

The first indication that TRAF3 is a much more interesting molecule than assumed followed the development and analysis of B lymphocyte cell lines that were made completely and specifically TRAF3-deficient, using an approach of gene targeting by homologous recombination, adapted for somatic cells. These studies revealed the unexpected finding that TRAF3 performs distinct and contrasting roles for CD40 versus its dysregulated viral mimic, LMP1.³⁴ Work with these cell lines also supported earlier findings that TRAF3 inhibits TRAF2-mediated effector functions in B cells.^35,36

This indication that TRAF3 can provide different roles to distinct receptors in the same cell prompted generation of the first strain of TRAF3^flox/flox mice, to provide a means of exploring whether TRAF3 plays not only receptor-specific, but also cell type-specific roles.³⁷ Studies of this and a second strain of such mice revealed that TRAF3 deficiency in B cells results in abnormally prolonged B cell homeostatic survival, as well as hyperresponsiveness to TLR stimulation.^37-39 This enhanced survival was initially attributed entirely to lack of TRAF3-facilitated degradation of the mitogen-activated protein kinase (MAPK) NF-κB inducing kinase (NIK), with subsequent constitutive nuclear localization of proteins of the NF-κB2 pathway, p52 and RelB.^37,38 However, subsequent use of the TRAF3^flox/flox mice to produce strains lacking TRAF3 in T lymphocytes, dendritic cells (DC), or macrophages (MΦ) and neutrophils demonstrate that all TRAF3 deficient (TRAF3^−/−) cell types have constitutive NF-κB2 activation, but only in B cells is this associated with enhanced survival.^39-42 Additionally, it was recently shown that the relationship between TRAF3, NIK, NF-κB2 and B cell survival is more complex than previously appreciated.²⁸ TRAF3 deficiency in DC impairs type 1 interferon, TNF-α and IL-10 production upon TLR stimulation, but results in enhanced production of IL-12.³⁹ A recent study shows that MΦ lacking TRAF3 also produce less type 1 interferon, more IL-6 and IL-12, but unaltered TNF-α in response to TLR ligands.⁴¹ Therefore, TRAF3 plays different roles in different cell types, and even in distinct myeloid cell types, its functions are overlapping, but not identical.

T cells express members of the TNFR superfamily, IL-1R and also some TLRs, indicating that TRAF3 is likely important for T cell function, in ways additional to suppressing the NF-κB2 signaling pathway, particularly as TRAF3^−/− T cells show no enhancement in survival. However, the roles played by TRAF3 in T lymphocytes were only minimally examined until the past several years, when it was revealed that there are multiple and very striking roles for TRAF3 in T cell biology. In this review, we will summarize the recent research progress toward understanding these roles of TRAF3.

Role of TRAF3 in T Cell Development

TRAF3 differentially regulates differentiation of specific T cell subsets

As stated above, TRAF3 depletion in all cell types examined to date leads to constitutive activation of the NF-κB2 signaling pathway. The early lethality of Traf3 germinal deletion can be rescued by simultaneous depletion of molecules in the NF-κB2 signaling pathway.^33,43 The NF-κB2 signaling pathway is important for cell survival, so it is considered as one of the reasons for the survival advantage of TRAF3^−/− B cells.^37,38 However, in CD4^CreTRAF3^flox/flox (T-TRAF3^−/−) mice, the proportion of CD4⁺ and CD8⁺ conventional T cells is not affected by the absence of TRAF3. The thymic size of T-TRAF3^−/− mice is comparable to that of TRAF3^flox/flox littermate control (LMC) mice, and the frequencies and numbers of thymocyte populations are normal.⁴⁰ Thus, depletion of TRAF3 from double positive (DP) thymocytes does not affect CD4⁺ and CD8⁺ conventional T cell lineage commitment or survival in the thymus. In addition, the proportions and absolute numbers of B cells and T cells are also normal in the spleen and lymph nodes in T-TRAF3^−/− mice compared to LMC.⁴⁰ These results demonstrate that deletion of Traf3 from thymocytes at the DP stage does not substantially affect conventional CD4⁺ and CD8⁺ T cell development and homeostasis.

However, further study of T cell subsets shows marked differences. T-TRAF3^−/− mice have a greater number of CD4⁺CD44^hi effector/memory T cells than LMC mice. In contrast, CD8⁺CD44^hiCD62L^hi central memory (Tcm) cells are markedly reduced in T-TRAF3^−/− mice in comparison to LMC mice, although CD8⁺CD44^hiCD62L^low effector memory T (Tem) cells and naïve T cells (CD8⁺CD44^lowCD62L^hi) do not show significant differences in number.⁴⁴ Furthermore, T-TRAF3^−/− mice exhibit increased frequency and numbers of CD4⁺CD25⁺Foxp3⁺regulatory T (Treg) cells,^15,40 but reduced invariant natural killer T (iNKT) cells in all lymphoid organs.⁴⁵ Together, these results indicate that although TRAF3 does not affect the total number of T cells, it plays different roles in regulating the proportions of distinct T cell subsets.

TRAF3 is required for iNKT cell development

The subset iNKT cells play critical roles in anti-tumor immunity, as well as being implicated in the pathogenesis of autoimmune and inflammatory diseases. Although the total number of T cells is not affected by the absence of TRAF3, iNKT cells are profoundly reduced in T-TRAF3^−/− mice,⁴⁵ indicating an important role of TRAF3 in iNKT cell development or survival. The development of iNKT cells is a complex process. Thymic iNKT cells can be divided into 4 developmental stages according to surface marker expression. Stage 0 and stage 1 iNKT cell development requires TCR signaling as well as signals delivered by signaling lymphocyte activation molecule (SLAM). Stages 2 and 3 of iNKT cell development require IL-15 signaling, which is also essential for mature iNKT cell homeostasis. Although all 4 stages can be found in thymus, the majority of stage 2 iNKT cells migrate to the periphery and acquire NK cell lineage markers.^46-48 Notably, during the transition from stages 1 to 2, the transcription factor T-bet is upregulated through TCR signaling.⁴⁸ T-bet further mediates IL-2/15Rβ chain (CD122) expression,⁴⁹ which is essential for activating IL-15 signaling during the later stages of development, and for mature iNKT cell proliferation and survival.

There are ∼10-fold fewer iNKT cells in the spleen, liver and thymus of T-TRAF3^−/− mice than in LMC. Our finding that the burst of proliferation of iNKT cells from stage 1 to stages 2 and 3 is defective in the absence of TRAF3 indicates that IL-15 signaling is affected. Indeed, IL-15-induced proliferation of TRAF3^−/− iNKT cells is diminished and IL-15 signaling is impaired. Expression of CD122 is reduced in stages 2 and 3 TRAF3^−/− iNKT cells compared to those of LMC. Furthermore, impaired TCR signaling in stage 1 iNKT cells does not efficiently upregulate T-bet, which is required for mediating CD122 expression.⁴⁵ Thus, the role played by TRAF3 in TCR signaling in stage 1 iNKT cells is instrumental for the transition to IL-15 signaling.

The findings that only later developmental stages of iNKT cells are impaired, but not stages 0 and 1, indicate that TCR signaling required for iNKT cell thymic selection at early developmental stages is unaltered. This result is consistent with our recent finding that T cell thymic selection is not affected in the absence of TRAF3.^15,40 In contrast, TCR signaling in peripheral TRAF3^−/− T cells is markedly diminished.⁴⁰ These results suggest that TRAF3 plays minor roles in TCR signaling during T cell development at early stages, but starts to exert its effect in mature T cells. The impaired TCR signaling in stage 1 iNKT cells is consistent with the mature status of stage 1 iNKT cells, implied by downregulation of CD24.⁴⁵ Taken together, these findings indicate that TRAF3 plays a critical role in the transition from TCR to IL-15 signaling pathways during iNKT cell development.

TRAF3 restrains Treg cell development in the thymus

Treg cells are an important immune cell population that keeps self-reactive cells in check and prevents over-activation during immune responses. Reduced Treg cell number or function may lead to autoimmune diseases, whereas increased Treg cell number or function can compromise normal immune responses and increase the risk of infection and cancers. Therefore, the number and function of Treg cells must be tightly controlled.

T-TRAF3^−/− mice exhibit doubled to tripled Treg cell numbers in all lymphoid organs compared to LMC mice. Moreover, this increase in Treg cells is not due to a survival or proliferation advantage, but because more Treg cells develop from the thymus in T-TRAF3^−/− mice. Consistent with this finding, TRAF3 deletion from mature Treg cells in Foxp3^CreTRAF3^flox/flox mice only slightly increases Treg cell number.⁵⁰ Interestingly, Treg precursor cells in T-TRAF3^−/− mice are unaltered, which further demonstrates that TRAF3 deficiency minimally affects early stage T cell development. The normal numbers of Treg precursor cells, but increased numbers of mature Treg cells suggest that the differentiation from precursors to Treg cells is more efficient in the absence of TRAF3. Indeed, more TRAF3^−/− than TRAF3^+/+ Treg precursor cells cultured in vitro in the presence of IL-2 become mature Treg cells.¹⁵

This enhanced IL-2 responsiveness of TRAF3^−/− Treg precursors was subsequently demonstrated to be a consequence of amplified IL-2 signaling, but not due to altered expression of IL-2Rs. In addition, thymic Treg cells emerge as early as at d1 after birth in T-TRAF3^−/− mice, a time when no Treg cells are found in LMC mice. This early appearance of Treg cells in the thymus is consistent with findings obtained with a mouse with transgenic expression of the IL-2R signaling molecule, the transcription factor STAT5.⁵¹ Thus, enhanced IL-2 signaling accelerates Treg cell development. By restraining IL-2 signaling, TRAF3 plays a critical role in controlling the transition from precursor cells to mature Treg cells. The molecular mechanism by which TRAF3 regulates IL-2R signaling is discussed below.

A recent study shows that the TNFR superfamily members OX40, GITR and TNFR2 also play important roles in the process of transition from Treg precursor to Treg cells.⁵² Because TRAF3 binds to the cytoplasmic domains of these receptors, it's possible that these TNFR superfamily members also contribute to the increased Treg cell numbers in the T-TRAF3^−/− mice. This possibility is a focus of ongoing investigations in our lab.

TRAF3 is essential for the homeostasis of central memory CD8⁺ T cells

The finding that CD8⁺ Tcm but not Tem cells and naïve T cells are remarkably reduced in T-TRAF3^−/− mice indicates that TRAF3 plays a unique role in the generation, proliferation or maintenance of CD8⁺ Tcm cells.⁴⁴ As induction of CD8⁺ Tcm cells is a goal of vaccine design and cancer immunotherapy, understanding how TRAF3 deficiency induces this striking phenotype may inform novel immunotherapies for infections and tumors. Our data show that TRAF3^−/− CD8⁺ Tcm cells undergo apoptosis faster and do not proliferate well to IL-15 stimulation, compared to those in LMC mice. Furthermore, IL-15 signaling in TRAF3^−/− CD8⁺ Tcm cells is defective.⁴⁴ Therefore, hypo-responsiveness of these cells to IL-15 is one reason underlying the reduction of this population in T-TRAF3^−/− mice. Our recent preliminary data show that the kinase NIK also plays critical roles for this specific T cell subset, as depletion of NIK from T-TRAF3^−/− mice restored normal CD8⁺ Tcm cell numbers (ZY & GAB, unpublished data). It was reported that the NF-κB2 pathway is required for efficient effector/memory T cell development.⁵³ However, constitutive activation of this pathway in T-TRAF3^−/− mice might also impair Tcm cell development or maintenance. Thus, multiple factors may contribute to the reduced CD8⁺ Tcm cell number in T-TRAF3^−/− mice. The underlying mechanism awaits further investigation.

TRAF3 in T Cell Functions

TRAF3 is required for conventional T cell functions

The role of TRAF3 in T cell biology was first studied by using TRAF3^−/− liver stem cells from the early-lethal TRAF3 knockout mouse to reconstitute the immune systems of irradiated wild type mice. The finding that T-cell-dependent (TD) but not T-cell-independent (TI) IgG production is impaired in TRAF3^−/− chimeric mice suggested that TRAF3 deficiency may impair T cell function, and defective antigen-specific recall responses further supported this notion.³³ However, as we now know that TRAF3 deficiency in other immune cells including B cells, DC and MΦ also impacts their function,^39,41 it could not be concluded that impaired T cell function was a cell intrinsic effect accounting for these early findings. In addition, whether T cell development is also affected in the absence of TRAF3 was not explored, which might also contribute to the phenotype of the chimeric recipients of TRAF3^−/− liver stem cells.

Targeted deletion of Traf3 specifically from T cells greatly improved our understanding of the role of TRAF3 in T cell function. Studies with this mouse model confirm that TRAF3 is required for TD antibody responses, particularly for the production of IgG1. In addition, TRAF3 deficiency in T cells compromises the ability of mice to resist L. monocytogenes infection. Antigen specific IFN-γ– or TNF-α–producing T cells are reduced in T-TRAF3^−/− mice, with greater reduction observed in CD4⁺ T cells than in CD8⁺ T cells.⁴⁰ These data demonstrate that TRAF3 is required for optimal T-cell function in vivo, although the increased Treg cells and reduced iNKT cells in these mice may also contribute to their reduced responses to infection and immunization.^15,45 To test whether TRAF3 deficiency affects the intrinsic function of conventional CD4⁺ and CD8⁺ T cells, purified Treg cell-depleted TRAF3-deficient and LMC T cells were stimulated with anti-CD3 and anti-CD28 Abs in vitro. Surprisingly, T cell proliferation and cytokine production were markedly reduced in the TRAF3^−/− T cells, indicating for the first time that TRAF3 is intrinsically involved in TCR+CD28-mediated signaling.³ As the role of TRAF3 in signaling by TRAF3-binding TNFR superfamily members expressed by T cells is incompletely understood, altered signaling by these receptors might also play roles in the compromised in vivo effector functions seen in T-TRAF3^−/− mice. However, upregulation of costimulatory TRAF3-binding receptors (OX40, 4–1BB, CD30, CD27) requires robust TCR signaling, so elucidation of the specific role that these receptors play in the T-TRAF3^−/− phenotype must await development of new model systems.

TRAF3 deficiency in B cells does not affect B cell proliferation.³⁷ However, TRAF3^−/− CD4⁺ and CD8⁺ T cells exhibit reduced proliferation upon anti-CD3+CD28 Ab stimulation, as fewer cells enter the cell cycle (S/G2/M phase).⁴⁰ In addition, in sharp contrast to the marked resistance to apoptosis in B cells lacking TRAF3, TRAF3 deficiency compromises T cell survival.^37,40 Although the underlying mechanism remains unclear, it further enhances the concept that TRAF3 plays different roles in distinct cell types.

TRAF3 is required for cytokine production by iNKT cells

Although iNKT cell numbers are remarkably reduced in the absence of TRAF3, a small number of iNKT cells can still mature. Analysis of these cells reveals that their surface receptors are either slightly decreased or unchanged. However, the percentage of iNKT cells producing IFN-γ and IL-4 upon α-Galactoceramide stimulation is drastically reduced in TRAF3^−/− compared with LMC iNKT cells.⁴⁵ This result is comparable with the phenotype of TRAF3^−/− conventional CD4⁺ and CD8⁺ T cells, which show defective cytokine production in response to TCR stimulation.⁴⁰

Role of TRAF3 in Treg cell function

The important role of TRAF3 in Treg cell development raises the question of whether it also plays a role in Treg cell function. This question was addressed by the complementary approaches of an in vivo induced inflammatory bowel disease (IBD) model and an in vitro suppressive assay. Results do not show a significant difference between LMC and TRAF3^−/− Treg cell function.¹⁵ In a more detailed study, it was revealed that suppression of Th1, but not Th17 cell function by TRAF3^−/− Treg cells is partially impaired. This study used the same IBD model as the earlier study, but the latter report does not address differences in disease parameters, such as mouse weight loss and histology. This second study also did not find differences in function of TRAF3^−/− Treg cells in the in vitro suppressive function assay.⁵⁰ However, both Foxp3^CreTRAF3^flox/flox mice and T-TRAF3^−/− mice exhibit more CD4⁺ CD44^Hi effector/memory T cells,^44,50 supporting the concept that TRAF3-deficient Treg cells might have defects in suppression of the Th1 response. Interestingly, although normal numbers of follicular Treg (T_FR) cells are found in Foxp3^CreTRAF3^flox/flox mice, antigen-stimulated production of T_FR cells is reduced. Furthermore, TRAF3 mediated signaling is essential for maintaining high levels of the inducible costimulator (ICOS) in T_FR cells, which is required for T_FR cell generation and inhibition of follicular helper T cells (T_FH) cells. Accordingly, Foxp3^CreTRAF3^flox/flox mice exhibit increased T_FH cells, germinal center formation and high affinity IgG Abs upon antigen stimulation.⁵⁰ Taken together, TRAF3 is partially required for general Treg cell function and specifically important for T_FR cell generation and function. How TRAF3^−/− Treg cells suppress distinct T cell subsets and other immune cells is an interesting topic for future exploration.

TRAF3 in T Cell Signaling

Requirement for TRAF3 in signaling by the TCR complex

The findings that in vitro stimulation of T cells with anti-CD3 and anti-CD28 Abs resulted in a decrease in cytokine production in TRAF3^−/− T cells indicate that TCR signaling is impaired in the absence of TRAF3. Indeed, a decrease in activation of key signaling events, including phosphorylation of Zap70, LAT, PLC-γ, and ERK upon anti-CD3 and anti-CD28 Ab stimulation was identified in TRAF3^−/− T cells.⁴⁰ In addition, Ca²⁺ influx, another output of TCR induction, was diminished in CD3+CD28 stimulated TRAF3^−/− T cells (ZY and GAB, unpublished data). Consistently, phosphorylation of ERK and p38 is also reduced in splenic TRAF3^−/− iNKT cells upon TCR stimulation.⁴⁵ These results suggest that TRAF3 is required for TCR signaling in conventional T cells as well as mature iNKT cells. To establish whether TRAF3 directly associates with the TCR complex, an immunoprecipitation of the TCR was performed. In CD3+CD28 stimulated T cells, TRAF3 associates with the TCR, though this interaction is not observed upon CD3 stimulation alone, so CD28 co-stimulation is required for optimal assocation.⁴⁰

In previous studies, the most proximal signaling event altered in the absence of TRAF3 was Zap70 phosphorylation. Newly acquired data identified a reduction in the activation of even more proximal signaling events. TRAF3^−/− deficient T cells stimulated with anti-CD3 and anti-CD28 Abs have reduced activation of the Src-family kinases, Fyn and Lck (ZY and GAB, unpublished). Further, TRAF3 interacted with the Src family kinase inhibitor Csk upon CD3+CD28 stimulation (AW and GAB, unpublished data). The molecular mechanisms of how TRAF3 regulates Csk localization or functions are currently under investigation. Compared with TRAF3, TRAF6 promotes TCR signaling by activating the IκB kinase (IKK) and IL-2 production in T cells.⁵⁴ A more recent study shows that TRAF6 regulates CD3+CD28-mediated induction of the transcription factor NFAT through the K63-lined ubiquitination of LAT.⁵⁵ This suggests further roles for TRAF proteins in TCR signaling.

TRAF3 restrains IL-2R signaling in certain T cell subsets

The binding of IL-2 to the IL-2R induces the activity of Janus kinase (Jak)1 and Jak3. These kinases are associated with IL-2R chains CD122 and CD132 (the common γ-chain), respectively. The Jaks phosphorylate each other and also IL-2Rβ. Three downstream signaling cascades are activated, including Jak/STAT5, PI3K/AKT and Ras/MAPK pathways. In contrast to STAT5, which can be directly phosphorylated by Jak3, additional intermediate molecules, such as the kinases Shc, Syk and Lck, are required for activation of the PI3K/Akt and Ras/MAPK pathways.^56-59 Once IL-2 signaling is activated, negative regulatory mechanisms will be initiated. These mechanisms include those mediated by Suppressor of cytokine signaling (SOCS)1 and SOCS3, which provide negative feedback to IL-2R signaling by associating with Jak1 and inhibiting its kinase activity.^61,62 Another negative feedback mechanism is provided by the phosphatase SHP-1, which dephosphorylates Jak1, and T cell protein tyrosine phosphatase (TCPTP), which can directly interact with Jak1 and Jak3 and dephosphorylate them.^63,64 The positive and negative regulatory mechanisms counterbalance each other and delicately control the strength of IL-2 signaling.

Our findings that TRAF3^−/− T cells exhibit elevated phosphorylation of Jak1 and Jak3 indicate that TRAF3 acts proximally to the IL-2R complex. Indeed, TRAF3 is found associated with Jak1 and Jak3 after IL-2 stimulation of conventional T cells. Further experiments show that TRAF3 is essential for TCPTP recruitment to the IL-2R complex. Moreover, TRAF3 interacts with TCPTP when both are expressed in 293T cells, and this interaction requires N-terminal Ring and Zn finger domains of TRAF3, but not its C-terminal TRAF domain.¹⁵ Therefore, TRAF3 is an important adaptor molecule linking TCPTP to the IL-2R complex, thereby restraining IL-2R signaling. The increased Treg cell numbers in TCPTP deficient mice also support these findings.⁶⁵

Interestingly, T cell TRAF3 deficiency affects phosphorylation of STAT5, but not ERK and AKT,¹⁵ indicating different regulatory mechanisms for the latter 2 signaling pathways. Compared to direct phosphorylation of STAT5 by Jak3, activation of PI3K/Akt and Ras/Erk requires additional intermediate molecules,^57–59 which could be one reason that TRAF3 plays less important roles in these signaling pathways. In addition to TRAF3, TRAF6 is also involved in IL-2 signaling. TRAF6 can compete with Jak1 for binding to the IL-2Rβ chain and negatively regulate signaling by the Jak1/Erk pathway; the PI3K/Akt and Jak/STAT5 pathways were not explored in this study.⁶⁶ Thus, it seems that TRAF6 and TRAF3 have different roles in IL-2R signaling, as they do in other receptor mediated signaling pathways in immune cells.

The ability of TRAF3 to regulate IL-2R signaling has important implications for more than Treg cell development, as IL-2 is an important cytokine in many aspects of T cell biology, and the signaling induced by IL-2 can be different in distinct T cell subsets. For example, although IL-2 activates S6 kinase in both CD8⁺ and CD4⁺ T cells, much higher activation is found in the former.⁵¹ In addition, IL-2 fails to activate the PI3K/Akt pathway in Treg cells due to high expression of the phosphatase PTEN.⁶⁷ IL-2 signaling is especially important for maintaining the homeostasis of mature Treg cells. However, our study shows that IL-2 signaling in TRAF3-deficient mature Treg cells is not changed, in striking contrast to that in Treg precursor cells and conventional T cells.¹⁵ One reason for this difference might be that Foxp3 expression in mature Treg cells mediates a unique gene-expression profile, which may contribute to the regulation of IL-2 signaling independently of TRAF3.⁶⁸ In addition, the constitutive activation of IL-2 signaling in these cells due to high expression of CD25 may activate a negative feedback loop to alter the regulatory mechanisms of IL-2 signaling. Consistent with the finding that IL-2 signaling and homeostasis of mature Treg cells show little change in T-TRAF3^−/− mice, Foxp3^CreTRAF3^flox/flox mice only exhibit slightly increased Treg cells.⁵⁰ Additionally, we find low expression of TRAF3 in mature Treg cells compared to conventional T cells, consistent with its minimal role in this population.¹⁵ Whether the low expression level of TRAF3 is due to degradation mediated by TNF superfamily members highly expressed in mature Treg cells awaits further investigation. These results emphasize the complexity of regulation of the IL-2 signaling pathway, which despite much investigation, remains incompletely understood.

TRAF3 in TNFR signaling pathways in T cells

TNFR-superfamily members 4–1BB (CD137), OX40 (CD134), GITR, CD27, CD40, HVEM and TNFR2, are essential co-stimulatory receptors on T cells.^69-72 TNFR regulation of the NF-κB signaling pathway occurs through the recruitment of TRAF adaptor proteins to the cytoplasmic tail of the receptor. TRAF3 interacts with the cytoplasmic domains of all the T cell expressing TNFR-SF members and subsequently regulates the NF-κB signaling pathway when overexpressed in 293T cells.^29,73 Furthermore, these co-expression studies established that the binding capacity of TRAF3 with individual TNFR-SF members is receptor dependent, with HVEM having the strongest interaction and CD40 having the weakest. Co-expression of TRAF3 with individual TNFR-SF members decreased the level of nuclear p52, a substrate of NF-κB, suggesting an inhibitory role of TRAF3 in TNFR signaling.⁷³ However, these studies must be explored using endogenous levels of relevant proteins, in T cells. The inhibitory role of TRAF3 in TNFR signaling is unique, as TRAF1, 2, 5 and 6 promote TNFR-mediated NF-κB signaling.^29,74,75 Our recent study shows that CD40 mediated signaling inhibits CD8⁺ T cell function in the unique environment of obese visceral adipose tissue.⁷⁶ CD40 signaling mediated TRAF3 degradation might weaken TCR signaling in these CD8⁺ T cells (ZY and GAB, unpublished data).

Due to the reduction in TCR signaling, TCR inducible expression of TNFR-SF members is weak in the TRAF3-deficient T cell mouse model, hindering the use of this mouse model for biochemical analysis (ZY and GAB, unpublished data). Thus, exploring the role of TRAF3 in T cell specific TNFR signaling pathways remains an area of interest.

NF-κB2 signaling in T cells

The NF-κB2 signaling pathway is essential for the development and maintenance of secondary lymphoid tissues.^77,78 While research has suggested that alterations in T cell-mediated NF-κB2 signaling leads to defects in CD4 T cell development and T cell mediated immune responses, additional evidence has shown the defects are in fact a result of alterations in DC-mediated NF-κB2 signaling.^79,80 Furthermore, thymic stroma requires NF-κB2 for lymphotoxin β receptor-mediated transcriptional regulation of Aire, a key regulator of central immune tolerance.^81,82 Thus, alterations in the NF-κB2 pathway leading to T cell mediated autoimmune diseases are not driven in a T cell intrinsic manner.

The NF-κB2 pathway is regulated by the levels of readily available NIK. NIK activity is tightly controlled by TRAF3-mediated recruitment of TRAF2 to the NIK-cIAP complex where TRAF2 induces the ubiquitination and degradation of NIK.^83,84 In T cells, stimulation of the 4–1BB receptor leads to TRAF3 degradation and NIK-induced activation of the NF-κB2 pathway.^83,84 Enhanced activation of NF-κB2 occurs in T cells deficient in TRAF3, but does not contribute to enhanced T cell survival, as seen in B cells.^37,38,40 Recently, an alternative isoform for TRAF3, traf3DE8, was identified to have a distinct role in NF-κB2 signaling. Unlike full length TRAF3, traf3DE8 activates rather than inhibits NF-κB2 signaling. Traf3DE8 disrupts the interaction of TRAF2 and NIK, preventing degradation of the latter.⁸⁵ This study confirms the diverse roles of TRAF3 in NF-κB2 activation and the need to further explore all roles of TRAF3 in T cell biology.

The reduction in TCR signaling seen in TRAF3-deficient T cells occurs prior to the induction of the NF-κB2 pathway, thus alterations in NF-κB2 activation cannot account for the effects of TRAF3 on TCR+CD28-mediated signaling and effector functions. However, it is possible that constitutive activation of NF-κB2 in TRAF3^−/− T cells could play a role downstream of other receptors that bind TRAF3, such as those discussed above.

Conclusions and Future Directions

Although it is known that TRAF3 can be involved in receptor mediated signals other than TNFR superfamily members, recent findings are still surprising, that TRAF3 modulates Treg cell development by restraining IL-2 signaling and conventional T cell function, and iNKT cell development and function by participating in TCR signaling. Many questions arise from these findings, such as whether TRAF3 regulates IL-2 signaling in other cell types, whether TRAF3 affects other cytokines, e.g. IL-7, IL-15 mediated signaling, whether other TRAFs are also involved in these signals, and how TRAF3 plays its roles when different signals are activated in T cells. Completely understanding these molecular mechanisms will allow us to better control T cell development and function and manipulate these parameters for therapeutic benefit.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Funding

This work was supported by the US National Institutes of Health R01AI28847 to G.A.B. and the US Department of Veterans Affairs (GAB). ZY received support from NIH T32AI007260. This work was based on work supported by the Office of Research and Development of the Veterans Health Administration of the US Department of Veterans Affairs.