The ever-expanding role of cytokine receptor DR3 in T cells

Abstract

Death Receptor 3 (DR3) is a cytokine receptor of the Tumor Necrosis Factor receptor superfamily that plays a multifaceted role in both innate and adaptive immunity. Based on the death domain motif in its cytosolic tail, DR3 had been proposed and functionally affirmed as a trigger of apoptosis. Further studies, however, also revealed roles of DR3 in other cellular pathways, including inflammation, survival, and proliferation. DR3 is expressed in various cell types, including T cells, B cells, innate lymphocytes, myeloid cells, fibroblasts, and even outside the immune system. Because DR3 is mainly expressed on T cells, DR3-mediated immune perturbations leading to autoimmunity and other diseases were mostly attributed to DR3 activation of T cells. However, which T cell subset and what T effector functions are controlled by DR3 to drive these processes remain incompletely understood. DR3 engagement was previously found to alter CD4 T helper subset differentiation, expand the Foxp3⁺ Treg cell pool, and maintain intraepithelial γδ T cells in the gut. Recent studies further unveiled a previously unacknowledged aspect of DR3 in regulating innate-like invariant NKT (iNKT) cell activation, expanding the scope of DR3-mediated immunity in T lineage cells. Importantly, in the context of iNKT cells, DR3 ligation exerted costimulatory effects in agonistic TCR signaling, unveiling a new regulatory framework in T cell activation and proliferation. The current review is aimed at summarizing such recent findings on the role of DR3 on conventional T cells and innate-like T cells and discussing them in the context of immunopathogenesis.

Introduction

The cytokine receptor DR3, also known as TRAMP, LARD, WSL-1, and TNFRSF25, is a member of the tumor necrosis factor receptor superfamily (TNFRSF), which comprises a group of 29 receptors in humans and 32 receptors in mice that bind to cytokines of the tumor necrosis factor (TNF) family. Within the TNFRSF, some receptors contain an intracellular motif known as the “Death Domain (DD)” that can trigger programmed cell death upon signaling. Such DD-containing TNFRSF members are colloquially known as “death receptors”, and they include the prototypic TNFR1 and Fas/CD95 as well as DR3, among others [1]. The DD motif is important in the signaling of death receptors because the homotypic trimerization of the intracellular domains mediates the recruitment and binding of critical downstream adaptor proteins [2]. Specifically, ligand-induced activation of the DD recruits the TNFR1-associated death domain (TRADD) adaptor protein, which plays a key role in TNF and DR3-mediated apoptosis by recruiting the FAS-associated death domain (FADD) [3]. Notably, some death receptors play roles other than inducing apoptosis, which include cell proliferation, activation, and effector cell differentiation. In this regard, DR3 was found to exert multiple effects in T cell biology, not only by inducing cell death but also by skewing T effector cell differentiation, expanding and augmenting Foxp3⁺ Treg immunity, and most recently, controlling the activation of innate-like T cells [4]. The current review aims to revisit our current knowledge of the roles and requirements of DR3 in T cells and to discuss them in the frame of newer findings that place DR3 as a key regulator of both conventional and innate-like T cells.

The cytokine receptor DR3 and its ligand TL1A

DR3 is a type 1 membrane protein that is expressed on various immune cells, including plasma cells, innate lymphoid cells (ILCs), and several subsets of T cells, including Foxp3⁺ Treg cells, and invariant Natural Killer T (iNKT) cells [5]. Outside of the immune system, DR3 is prominently expressed on lung fibroblasts and bronchial epithelial cells [6], as well as on intestinal epithelial cells [7], and endothelial cells [8], where it was found to be associated with tissue injury and controlling tumorigenesis. While highly interesting, the non-immune function of DR3, however, is not the topic of this review, which is mostly focused on its role in T lineage cells.

Encoded in the Tnfrsf25 gene and expressed as a 417 amino acid 47 kDa transmembrane protein, the DR3 receptor shows the highest sequence homology to TNFR1 among TNFRSF receptors but it is distinct in its expression and function [9]. Akin to all other members of the TNFRSF, DR3 is preassembled as a stable trimer on the cell membrane and binds its ligand, TNF-like cytokine 1A (TL1A), which is also a trimeric protein [10].

The cytokine TL1A, also known as TNF superfamily member 15 (TNFSF15) or vascular endothelial growth inhibitor (VEGI), is produced by endothelial cells and antigen-presenting cells (APCs) that include dendritic cells [11], macrophages [12], and monocytes [13], among others. While TL1A is constitutively expressed on endothelial cells, its expression can be further upregulated by activation of the Toll-like receptor (TLR) or Fcγ receptors (FcγRs) in APCs [13, 14]. TL1A is mainly produced as a type 2 transmembrane protein that self-assembles into stable trimers through interactions between TNF homology domains [15]. In addition to its membrane-bound form, however, TL1A is also found as a soluble protein, which can be generated either by alternative mRNA splicing or by matrix metalloproteinase-mediated proteolytic cleavage of the extracellular domain [10, 16, 17]. Whether soluble TL1A molecules that are generated by proteolysis differ in their function from soluble TL1A that is produced by alternative splicing is an important issue that still needs to be resolved. Soluble TL1A produced by proteolysis requires the expression and activation of the TNF-α converting enzyme (TACE), a membrane metalloproteinase [17]. In agreement, the inhibition of TACE by small molecules such as TAPI-1, strongly diminished the number of soluble TL1A in the supernatant of cells overexpressing TL1A [17]. How TACE expression is controlled to determine the abundance of soluble TL1A molecules that are shed from the cell surface, and whether such shedding is controlled either in conjunction with or independently of TL1A alternative splicing remain to be resolved.

Regarding TL1A mRNA splice isoforms, two alternative splice products, namely VEGI-251 and VEGI-192, have been proposed to be expressed in humans [18]. While they differ in their abundance, it is unclear whether they would also differ in their function [18]. Along these lines, whether soluble TL1A molecules that are generated by post-transcriptional mechanisms, i.e., alternative splicing, are functionally different from soluble TL1A produced by post-translational processes, i.e., proteolytic cleavage, also remains to be tested. Moreover, it would be critical to discern whether soluble TL1A has disparate effects from membrane-bound TL1A, which is quite likely based on earlier studies on soluble versus membrane-bound TNFα proteins. Reportedly, TNFα binds to both TNFR1 and TNFR2, but it only activates TNFR1 efficiently when in its soluble form [19]. Considering the structural similarities between DR3 and TNFR1, it is plausible that, akin to TNFR1, DR3 could also have different affinity towards membrane-bound versus soluble TL1A. In agreement, while the membrane form of TL1A promoted expression of inflammatory cytokines in the lung, soluble TL1A was unable to do so [20]. Such limitation was independent of the potency of soluble TL1A because it triggered both apoptosis and NF-κB activation, as was also the case for membrane-bound TL1A [21]. Thus, despite employing the same receptor and presumably triggering the same downstream signaling pathway, membrane-bound and soluble TL1A are not identical in their effector function. These results suggest that further studies are required to gain full insights into the mechanism of DR3 signaling by TL1A.

Signaling pathways of DR3

Two major pathways operate downstream of DR3 that promote either proapoptotic or proliferative and proinflammatory activities in T cells. The distinct outcome of DR3 signaling is determined through the combinatorial use of different signaling circuits, which include the activation of caspase 8, NF-κB, the MAP kinase (MAPK), and the PI3K pathway (Figure 1). Under steady-state conditions, the downstream effects of DR3 predominantly emanate from the proinflammatory cascade. In brief, proinflammatory DR3 signals get triggered by activation of the DD motif in the DR3 cytoplasmic tail, which then recruits the 34 kDa adaptor protein TNFR-associated death domain (TRADD) that also contains a DD and binds DR3 through the DD-DD interaction [22]. TRADD has no intrinsic kinase activity but is critical for DR3 signaling because DR3 downstream signaling is severely attenuated in TRADD-deficient T cells [23]. The major function of TRADD in DR3 is to provide a platform for the recruitment of downstream pathway molecules, such as TNF receptor-associated factor 2 (TRAF2), cellular inhibitor of apoptosis proteins 1 and 2 (cIAP1/2) and receptor-interacting protein kinase 1 (RIPK1) [24] Figure 1). Polyubiquitination of RIPK1 by cIAP1/2 activates MAPK, NK-κB, and PI3K signaling pathways [24], and the induction of gene expression downstream of these three pathways eventually leads to a proinflammatory response to DR3 signaling.

Figure 1: DR3 signaling can induce both proinflammatory and proapoptotic pathways.

TL1A is expressed either as a transmembrane protein or in a soluble form. The soluble form of TL1A can be generated by alternative splicing or the proteolytic cleavage of the membrane-bound form by matrix metalloproteinases, such as ADAM17. Two major pathways operate downstream of DR3 signaling that promote proapoptotic or proinflammatory activities through the activation of caspase 8, NF-κB, MAPK, and the PI3K pathway. Under steady-state conditions, the downstream effects of DR3 are predominantly proinflammatory. The proinflammatory DR3 signals get triggered by activation of the DD motif in the DR3 cytoplasmic tail, which recruits TRADD that allows the recruitment of downstream pathway molecules, such as TRAF2, cIAP1/2 and RIP1K. Polyubiquitination of RIPK1 by cIAP1/2 further activates MAPKs, NK-κB, and PI3K signaling pathways, and the induction of gene expression downstream of these three pathways eventually leads to a proinflammatory reaction to the DR3 signaling. The alternative pathway that induces programmed cell death is transduced through the caspase 8 cascade. When inducing apoptosis, the adaptor protein TRADD binds to the FADD and RIPK3, which then activates the caspase 8 pathway, leading to apoptotic cell death.

TRADD: TNFR-associated death domain; TL1A: Tumor Necrosis Factor-like cytokine 1A; ADAM17: A disintegrin and metalloprotease 17; DR3: Death Receptor 3; NF-κB: nuclear factor kappa light chain enhancer of activated B cells; MAPK: Mitogen activated protein kinase; PI3K: Phosphoinositide 3-kinase; DD: death domain; TRAF2: TNF receptor-associated factor 2; cIAP1/2: cellular inhibitor of apoptosis proteins 1 and 2; FADD: Fas-associated death domain; RIPK3: Receptor-interacting protein kinase 3.

The alternative DR3 signaling pathway that induces programmed cell death is transduced through the caspase 8 cascade. In this case, the adaptor protein TRADD binds to the Fas-associated death domain (FADD) and RIPK3, which then activate the caspase 8 pathway, leading to apoptotic cell death [25] (Figure 1). Activation of the proapoptotic pathway by DR3, however, is not the norm. As recently documented, soluble TL1A binding to DR3 preferentially induces the proinflammatory pathway and not apoptosis [26], and such was the case because NF-κB activation induced cIAP2, diverting the main downstream pathway towards inflammation [9]. In a simple yet elegant experiment, the role of NF-κB was reaffirmed as a central factor in DR3-mediated apoptosis when a NF-κB-specific inhibitor, but not a MAPK-specific inhibitor, diverted DR3 signaling into cell death of human erythroleukemic TF-1 cells [24]. Mechanistically, it turned out that NF-κB-specific inhibitors reduced the production of c-IAP2, and with interruption of c-IAP2 synthesis by RNA interference, the apoptotic cascade was induced upon TL1A-mediated ligation of DR3 [24]. Altogether, the current notion of DR3 signaling posits that the default signaling effects would run through proinflammatory rather than apoptotic pathways (Figure 1). However, it remains unclear what cellular factors or context would control the outcome of DR3 signaling into apoptosis versus activation. It is also uncertain if DR3 signaling would alternate into two disparate pathways depending on the cell type or whether it represents sequential events that can happen in the same cells. Finally, whether there would be a divergence in DR3 signaling effects based on the lineage and function of the cells that express DR3 will require further studies.

DR3 expression in T lineage cells

Unlike the ubiquitously expressed TNFR1, DR3 expression is more restricted in its tissue distribution and mostly found on T lineage cells among lymphocytes [27]. A DR3 requirement in T cell development was tested using DR3 germline knockout mice, but no discernible defects in lymphoid organ development, including the thymus, LN, spleen and Peyer’s patches, or lymphoid cell generation were observed [28]. Overall thymocyte development, including the differentiation from immature CD4⁻CD8⁻ double negative (DN) into CD4⁺CD8⁺ double positive (DP) cells and their positive selection, was also unaffected in DR3-deficient mice [28]. To examine whether DR3 would be essential for thymic negative selection, the HY TCR transgenic mouse model was used, in which the HY TCR recognizes a male-specific antigen in context of H-2^b, resulting in the negative selection of clonotypic TCR specificities in male mice [29]. Interestingly, DR3-deficient male mice contained substantially increased numbers and frequencies of HY TCR expressing mature CD8 thymocytes, especially in mice 10 weeks and older, suggesting that DR3-deficiency would impair negative selection [28]. These results suggested that DR3 is not required for the development and positive selection of T cells, but that it might be involved in the elimination of self-reactive thymocytes.

While immature DN and DP thymocytes lack DR3, DR3 expression is modestly induced on both mature CD4 and CD8 single-positive (SP) thymocytes [4]. In peripheral T cells, DR3 expression persists, but it is more pronounced on CD4 compared to CD8 T cells [10, 30]. Among peripheral CD4 T cells, Foxp3⁺ Treg cells display a greater abundance of DR3 compared to Foxp3-negative conventional CD4 T cells, and such is also the case for Foxp3⁺ Treg cells in the thymus [4]. Nonetheless, DR3 expression can be further induced on conventional CD4 T cells by CD3 ligation [31], and such was also the case for CD8 T cells, whereby TCR activation of naïve CD8 T cells increased their DR3 levels up to 4-fold [10]. These results are in agreement with earlier observations that DR3 mRNA and protein expression in naïve and resting T cells are low [32], but that activated T cells express copious amounts of DR3 [33].

Because of the distinct DR3 expression in naïve versus activated T cells, deciphering the physiological role of DR3 in T cells has not been easy. When stimulating human T cells with recombinant TL1A proteins, TL1A alone was insufficient to induce proliferation in resting T cells [32]. However, the pretreatment with TL1A significantly augmented anti-CD3/CD28 induced T cell proliferation in response to IL-2 (Figure 2). Interestingly, the cytokine production of such TL1A costimulated T cells was also substantially enhanced [32], suggesting that the TL1A/DR3 axis could be involved in driving the effector function of activated T cells. In this regard, it was questioned whether DR3 signaling could also play a role in the polarization of activated CD4 T cells. Naïve CD4 T cells can differentiate into different effector T cell subsets with distinct functions [34], but both in vitro and in vivo studies showed that DR3 signaling was not required for the polarization of naïve CD4 T cells into CD4 T helper 1 (Th1), Th2, Th9 or Th17 cells [35, 36]. On the other hand, the TL1A/DR3 axis was found to promote Th1 and Th17 cell differentiation in T-T and DC-T cell interaction-dependent manners, because TL1A-deficiency impaired the differentiation of naïve CD4 T cells into Th1/Th17 cells [37]. Such notion of a positive regulatory effect of DR3 in helper T cell differentiation was further supported by the observation that DR3 was required to establish brain immunopathology in experimental autoimmune encephalomyelitis, an autoimmune disease model dependent on Th17 and Th1 cell subsets. Specifically, in the absence of DR3, the number of T cells accumulating in the target organ as well as the production of effector cytokines, were found to be significantly reduced [35]. Collectively, while there is no clear consensus on the role of DR3 on effector T cell differentiation, it is evident that DR3 can promote the activation and cytokine production of T cells in the context of TCR stimulation [35, 38, 39].

Figure 2: DR3 plays a multifaceted role in adaptive immunity.

DR3 is a cytokine receptor of the TNF receptor superfamily that plays a multifaceted role in adaptive immunity. DR3 is expressed on the cell surface of various immune cells, including Treg cells, helper T cells, MAIT cells, and iNKT cells. Upon DR3 signaling, Treg cells expand and these expanded Treg cells are immunosuppressive ex vivo. Foxp3-negative CD4 T cells also express DR3, albeit at a lower level than that of Treg cells. Pretreatment with TL1A can augment T cell proliferation in response to anti-CD3/anti-CD28 stimulation. Innate-like T cells, such as MAIT and iNKT cells, also get activated in response to DR3 signaling. In vitro DR3 stimulation of MAIT cells increases their TNF-α production while iNKT cells induce the expression of activation markers. In the presence of TCR signaling, DR3 exerts costimulatory effects, which expands and activates iNKT cells, leading to robust cytokine secretion.

DR3: death receptor 3; TNF: tumor necrosis factor; Treg: regulatory T cell; MAIT cell: Mucosal-associated invariant T cell; iNKT cells: invariant Natural Killer T cells; Foxp3: forkhead box P3; TL1A: Tumor Necrosis Factor–like cytokine 1A.

DR3 in innate-like T cells

Because DR3 was frequently found to be expressed on end-differentiated effector T cells, it was only logical that DR3 expression should be also assessed on innate-like T cells, which correspond to a heterogenous population of unconventional T cells that are innately equipped with effector function. Innate-like T cells comprise a series of preactivated effector T cells that include thymic innate CD8 T cells, some γδ T cells, mucosal-associated invariant T (MAIT) cells, and invariant Natural Killer T (iNKT) cells, among others [4]. Akin to conventional T cells, iNKT cells are generated in the thymus upon TCR-mediated positive selection, but their TCR repertoire is extremely limited. In fact, all iNKT cells express an invariant Vα14-Jα18 TCRα chain, which restricts their antigen specificities to glycolipids that are presented in the context of the non-classical MHC-I molecule, CD1d. While they all express the same TCRα chain, iNKT cells are diverse in their phenotype and function. As such, iNKT cells can be divided into three major subsets, i.e., NKT1, NKT2 and NKT17 cells, based on their transcription factor and cytokine expression profiles [40]. Analogous to the CD4 T helper subsets, NKT1 cells express the transcription factor T-bet and produce IFNγ, whereas NKT2 cells express large amounts of the transcription factor PLZF and produce IL-4 [40, 41]. Along these lines, IL-17-producing iNKT cells are referred to as NKT17 cells and they uniquely express the nuclear factor RORγt [40, 41]. Importantly, iNKT cells are generated in the thymus as fully differentiated effector T cells. However, whether and when they would express DR3 has not been determined yet. In this regard, DR3 expression on iNKT cells was already documented in 2008 [11], but whether DR3 is an acquired versus an innate trait of iNKT cells had not been examined until recently.

Interestingly, DR3 was found to be ubiquitously expressed on all iNKT cells residing in peripheral tissues, including the spleen, LN, and lung [4]. Such broad expression markedly differed from the subset-specific DR3 expression on thymic iNKT cells, where DR3 was selectively found on NKT17 but not on other subsets of iNKT cells [4]. The molecular basis of distinct DR3 expression on thymic versus peripheral iNKT cells remains unclear. But since all peripheral iNKT cells are derived from thymic iNKT cells, it is likely that DR3 expression on peripheral iNKT cells is an acquired trait while DR3 on thymic NKT17 cells is developmentally programmed. Along these lines, NKT17 cells uniquely express the transcription factor RORγt, and the forced expression of RORγt was sufficient to induce ectopic DR3 expression on thymic non-NKT17 iNKT cells, such as NKT1 and NKT2 cells, as well as CD8 thymocytes. Mechanistically, RORγt ChIP-qPCR analysis showed that RORγt directly binds to the Tnfrsf25 gene, encoding DR3, presumably to directly drive its expression [4]. Thus, at least in thymic iNKT cells, DR3 expression appears to be positively regulated by RORγt. For peripheral iNKT cells, however, DR3 expression can be induced independently of RORγt because NKT1 and NKT2 cells do not express RORγt but are still abundant for DR3 in peripheral tissues [4]. How the DR3 regulatory mechanisms differ between thymic and peripheral iNKT cells is not known and remains to be clarified.

Understanding the role of DR3 in iNKT cells is another aspect that remains incompletely understood, but that has recently gained momentum by the identification of DR3 as a costimulatory molecule in iNKT cell activation [4]. The full activation of conventional T cells requires the TCR engagement with the co-ligation of costimulatory molecules, such as CD28. However, whether TCR activation of iNKT cells also requires costimulation, and whether it is CD28 that would provide such costimulatory signals are not clearly understood. In this regard, the agonistic glycolipid α-galactosylceramide (α-GalCer) dramatically increased the activation of thymic iNKT cells when employed in the presence of agonistic anti-DR3 monoclonal antibodies (clone 4C12) [11]. Consistent with the finding that DR3 is primarily expressed on thymic NKT17 cells, α-GalCer stimulation with agonistic anti-DR3 antibodies skewed the immune response toward NKT17 cells, promoting IL-17A production in the thymus [4]. The skewing of the iNKT subset response toward RORγt-positive NKT17 cells mirrored the Th17-polarizing effects of TL1A in CD4 T helper cells [37]. Thus, the co-ligation of DR3 strongly augments agonistic TCR signaling, and consistent with an effect as a costimulatory molecule, the activation of DR3 alone and in the absence of TCR stimulation failed to induce T cell activation. Because activation markers downstream of TCR signaling, such as CD69 and CD25, and Nur77 gene reporter expression were highly induced upon DR3 co-ligation in α-GalCer stimulated thymic iNKT cells [4], it is likely that DR3 would exert costimulatory functions through augmenting TCR signaling. Specifically, the costimulatory effect of DR3 was nullified in the presence of MAPK inhibitors, suggesting that DR3 targets the MAPK pathway for costimulation [4].

Consistent with DR3 being expressed on all subsets of peripheral iNKT cells, the coadministration of anti-DR3 and α-GalCer into experimental mice also triggered the activation of all iNKT cells, resulting in a systemic immune response [42] (Figure 2). Four days after α-GalCer/anti-DR3 stimulation, all peripheral iNKT cell acquired a highly activated phenotype that manifested in their dramatic expansion and accumulation in various tissues and organs, accompanied with splenomegaly but also thymic atrophy. Notably, the swift and dramatic loss of thymocyte numbers by α-GalCer/anti-DR3 but not by α-GalCer injection alone, suggested a costimulatory role for DR3 in TCR signaling and iNKT cell activation [42]. Collectively, these results expand the scope of DR3’s role in T cell immunity from transducing TL1A-triggered cytokine receptor signaling into a regulatory role of TCR signaling.

Because DR3 signaling intersects with the TCR signaling pathway, it would be important to examine and understand the role of DR3 in other T cell populations as well. In this regard, innate-like T cells other than iNKT cells were also found to express DR3, rendering them responsive to TL1A. MAIT cells correspond to such a population of DR3-expressing innate-like T cells, which comprises a group of unconventional T lymphocytes that express a semi-invariant TCR and recognize riboflavin-derived metabolites in the context of the MHC-I-like molecule, MR-1 [43]. MAIT cells are generated in the thymus from DP thymocyte precursors. However, unlike iNKT cells, MAIT cells only differentiate into IFNγ-producing MAIT1 and IL-17-producing MAIT17 cells, without the ability to become IL-4 producers [43]. Whether DR3 would play a regulatory role in MAIT cell immunity has been unclear, but it was proposed based on the observation that human MAIT cells upregulated DR3 expression upon activation by inflammatory cytokines. Moreover, the in vitro stimulation of human MAIT cells with recombinant TL1A drastically increased their TNFα secretion in a TCR signaling-independent manner [44]. Additionally, MAIT cell activation by TCR engagement through ligand-MR1 stimulation alone turned out to be insufficient to trigger their full activation [45], and it was the addition of cytokines, such as IL-12 and IL-18, that was necessary to achieve full MAIT cell activation [46]. Similarly, the DR3 stimulation by TL1A further augmented MAIT cell immune response, indicating that DR3 costimulation of MAIT cells could be an important mechanism to boost host-defense abilities, specifically in tissues abundant in TL1A, like the gut [47] (Figure 2). Collectively, these studies highlight a positive regulatory role of DR3 in MAIT cells, suggesting that the TL1A/DR3 axis could promote the activation and cytokine secretion of MAIT cells, both in conjunction with and independently of TCR signaling.

DR3 expression has been also reported for some γδ T cells, which correspond to another innate-like T cell population [48]. γδ T cells represent a small subset of unconventional T lymphocytes that preferentially homes to peripheral tissues and barrier sites [49]. Interestingly, γδ T cell numbers were significantly reduced among intraepithelial lymphocytes (IELs) in the small intestine when TL1A is absent, suggesting that DR3 signaling could contribute to their recruitment or homeostasis [48]. Moreover, the γδ T cells in the small intestinal IELs of TL1A-deficient mice showed reduced levels of the activating receptor NKG2D, suggesting lower activation potential of the intestinal γδ T cells in the absence of DR3 signaling [48]. These findings demonstrated that DR3 signaling can impact the activation of γδ T cells as well, further expanding the pool of TL1A/DR3-controlled innate-like T cells.

In recent years, the discovery of ILCs, which are derived from common lymphoid progenitors but lack T and B lineage markers [50], expanded the population of cells expressing DR3 [51]. ILCs constitute only a small proportion of lymphocytes but their innate phenotype allows them to secrete copious amounts of cytokines [52]. Like iNKT cells, there are three subtypes of ILCs based on their transcription factors and cytokine profiles. Group 1 ILCs are T-bet-positive and produce IFNγ akin to NKT1 cells. Group 2 ILCs (ILC2) express RORα and GATA-3 and they produce IL-13 and IL-5. Group 3 ILCs (ILC3) comprise a relatively heterogenous group composed of RORγt-positive cells secreting IL-17 and IL-22 [52]. Interestingly, the level of DR3 expression differed among these subsets, with ILC2 and ILC3 displaying the highest levels. In accordance, DR3 activation with TL1A costimulated ILC2 and ILC3 and augmented the secretion of IL-13 and IL-22, respectively [51]. Specifically in a pathological setting, DR3 was required for the expansion and function of ILC2 in both T cell-dependent and independent models of allergic diseases [53]. These data expanded the role of DR3 beyond T cells while suggesting an activating role for DR3 in ILCs.

DR3 in Foxp3⁺ Treg cells

Another major population of DR3-expressing T cells are the immunosuppressive Foxp3⁺ regulatory T (Treg) cells [54]. The cytokine receptor expression and cytokine requirement of Foxp3⁺ CD4 T cells are distinct from those of conventional CD4 T cells, as Treg cells express uniquely high levels of the IL-2 receptor α- and β-chain, and they mostly depend on IL-2, but not IL-7, for their differentiation and homeostasis [55, 56]. Treg cells also differ from conventional CD4 T cells as they express large amounts of the cytokine receptor DR3. Foxp3⁺ Treg cells naturally acquire DR3 expression along their thymic generation, but peripheral Treg cells that are induced from mature CD4 T cells in peripheral tissues also express large amounts of DR3 [57].Thus, DR3 expression is likely intrinsic to the molecular signature of Foxp3⁺ Treg cells among CD4 T cells [57].

Because DR3-deficient mice were unaffected in their generation and maintenance of Foxp3⁺ Treg cells, it was not immediately clear whether DR3 would play a role in Treg cell biology [58]. However, a single injection of agonistic anti-DR3 antibodies vigorously and selectively expanded Foxp3⁺ Treg cells in vivo, demonstrating that DR3 positively regulates Treg cell activation [54] (Figure 2). Importantly, the DR3-induced proliferation strictly depended on TCR engagement and IL-2 signaling, suggesting that DR3 plays a TCR costimulatory role rather than a TCR-independent role in Treg cell activation. Moreover, such DR3-expanded Treg cells were protective against allergic lung inflammation, and they remained suppressive ex vivo, cementing the role of DR3 as a costimulatory factor in Treg cells [54]. In marked contrast to such in vivo effects of DR3, however, it was also shown that both recombinant TL1A and agonistic anti-DR3 antibodies could directly inhibit the suppressive activities of Treg cells in vitro [59]. The molecular basis of these contradicting results remains unresolved. However, these observations indicate that the effects of the DR3-TL1A axis on Treg cells can be highly dependent on the experimental conditions.

To characterize the functional capacity of DR3 stimulated Treg cells, their suppressive abilities were assessed in a graft versus host disease (GvHD) model [60]. The adoptive transfer of C57BL/6 donor T cells (H-2^b) into haplotype mismatched irradiated BALB/c host mice (H-2^d) results in acute GvHD, which can be ameliorated by donor Treg cells [61]. Strikingly, a single injection of anti-DR3 agonistic antibodies was sufficient to expand the Treg cell pool in donor mice, such that adoptive transfer of CD4 donor T cells of DR3-treated mice improved the survival of the host mice compared to control donor T cells [60]. Detailed immunophenotyping of DR3 signaled Treg cells pointed the increased suppression not only to increased numbers of Treg cells, but also to the increased expression of effector molecules, such as LAG3, PD-1, LAP, and TIGIT [62]. Notably, both the transcription factor Helios and the costimulatory molecule ICOS were highly induced in DR3-treated Treg cells. Considering that Helios expression is associated with increased activation [63], and that ICOS expression marks increased effector functions of Tregs cells [64], the combined expression of these two molecules further supports the notion that DR3 treatment promotes Foxp3⁺ Treg cell activation. Collectively, these data indicate that DR3 is a functional molecule in Treg cells, modulating their TCR and cytokine receptor responsiveness to increase proliferation and activation [62].

DR3 in immunopathogenesis

Consistent with its broad expression on T lineage cells, DR3 has been associated with multiple immune diseases. A genome-wide case-control study, for example, showed that certain polymorphisms in the Tnfsf15 gene, encoding TL1A, were associated with Crohn’s disease (CD) in a Japanese cohort and with the onset of inflammatory bowel disease (IBD) in two European cohorts [65]. Such studies suggested that genetic variants in molecules of the DR3-TL1A signaling axis could contribute to increased susceptibility to T cell mediated autoimmune diseases.

A possible link between DR3 and gut inflammation was further demonstrated in animal models where DR3 expression was found to be associated with the pathogenesis of chronic ileitis [33]. While DR3 expression and TL1A production were significantly upregulated in the intestinal mucosa upon inflammation [33], DR3 expression was also substantially induced in activated lymphocytes through an alternative mRNA splice mechanism that upregulates membrane DR3 expression [33]. Moreover, DR3 signaling selectively induced the proliferation of effector memory phenotype but not naïve T cells, suggesting that DR3 could exacerbate inflammation, placing it as a key player in gut inflammation [33]. This notion was also confirmed in the dextran sodium sulfate (DSS)-induced chronic colitis model [66], where the TL1A treatment of CD4 T cells from gut-associated lymphoid tissues of DSS colitis mice showed enhanced in vitro cytokine production by Th1 and Th17 cells. Conversely, in vivo studies using blocking anti-TL1A antibodies attenuated the disease severity of established colitis by downregulating the activation of both Th1 and Th17 cells [66]. In agreement with its proinflammatory role in the gut, the forced expression of TL1A in either the myeloid or lymphoid cell compartments by TL1A-transgenesis induced mild ileitis in all the animals [67–69]. Most of these inflammatory changes were observed in the terminal ileum, similar to how IBD presents in patients [67–69]. Considering these multiple gain- and loss-of-function models of the DR3-TL1A axis, it is evident that aberrant DR3 signaling in lymphocytes can promote intestinal inflammation and pathogenesis.

Another autoimmune condition afflicted by dysregulated DR3 and/or TL1A expression is rheumatoid arthritis (RA). Soluble TL1A was found in the serum and synovial fluid of patients with RA, and TL1A levels positively correlated with disease severity [70]. The decline in TL1A levels after TNF blockade in RA patients, on the other hand, indicated that TL1A levels may be an effective biomarker for TNF activity [71]. Because immunohistochemistry revealed TL1A-positive cells being present in synovial tissues, these results also suggested the TL1A would directly affect the target tissue [70]. A more direct role of TL1A was shown by collagen-induced arthritis (CIA), a classical animal model of RA [72]. As exogenous TL1A administration into CIA-induced mice exacerbated the disease burden and promoted the generation of anti-collagen autoantibodies, these results suggested that DR3 signaling is pathogenic in the context of RA. Such relationship between DR3 and arthritic symptoms was corroborated by the finding that DR3 knockout mice showed resistance to the development of bone pathology in antigen-induced arthritis (AIA) [73]. Osteoclasts are involved in bone resorption that is often seen in later stages of RA, and detailed analysis of bone erosion in DR3-deficient animals showed reduction in osteoclast numbers while TL1A stimulation conversely promoted the generation of osteoclasts in vitro [73]. Collectively, these findings suggest that DR3 and TL1A are key elements in the cytokine network contributing to autoimmune pathology in rheumatoid arthritis.

TL1A is constitutively expressed by epithelial cells so that a potential link of the TL1A/DR3-axis with the pathogenesis of psoriasis has been suspected [74]. In this regard, TL1A expression was found to be highly upregulated in patients with psoriatic lesions, particularly in infiltrating inflammatory cells, keratinocytes, and vascular cells [75, 76]. Also, a role for Th17 immunity in psoriasis is well documented [77], whereby DR3 signaling is known to promote IL-17 production [78]. Considering these findings, it is reasonable to assume that DR3 could trigger psoriasis through promoting Th17 immunity, which was tested in the imiquimod (IMQ)-induced psoriasis model [79]. IMQ-treated mice developed skin lesions resembling that of psoriasis patients, and TL1A expression level was significantly higher in the skin of IMQ-treated mice compared to that of the control group. Moreover, exogenous TL1A exacerbated the psoriasiform phenotype whereas administration of anti-TL1A blocking antibodies alleviated the disease burden [79]. The pathogenic impact of TL1A on the disease status was correlated with enhanced T cell, neutrophil, and dendritic cell infiltration as well as increased mRNA levels of IFNγ and IL-17 in the skin of psoriatic mice [79]. Collectively, the TL1A-DR3 axis plays a pathogenic role also in psoriasis.

Further expanding the immunoregulatory effects of DR3, a recent study of acute respiratory distress syndrome (ARDS) showed substantial reduction in DR3 expression in the alveolar epithelium of septic-ARDS patients which correlated with their disease severity [80]. To further investigate this observation in an experimental setting, a mouse model using lipopolysaccharide (LPS)-induced ARDS was used. In this model, germline knockout and alveolar epithelium-specific conditional knockout of TL1A exacerbated alveolar inflammation in LPS-induced ARDS [80]. Unexpectedly, however, DR3 signaling turned out to be protective in ARDS development. On a similar note, a study on the role of DR3/TL1A axis on IBD using DSS-induced colitis showed that DR3 knockout mice developed more severe colon inflammation and had impaired intestinal epithelial cell regeneration, compared to their littermate controls [7]. This finding is particularly surprising because previous reports conversely showed pathogenic effects of DR3 in various animal models of IBD, including in the DSS-induced colitis model. However, it is also important to consider that the DSS-induced colitis model has been controversial for its less-than-ideal replication of the IBD setting [81]. Nonetheless, these two studies indicate a protective, rather than pathogenic role for the TL1A/DR3 axis, at least in the context of pulmonary and large intestinal inflammation, which is contrary to autoimmune conditions previously discussed.

Perspective

Over the past decade, remarkable advances in understanding DR3 biology in healthy and pathological settings have been made. Various animal models have provided valuable information about the importance of DR3 in disease models of asthma, colitis, arthritis, and psoriasis, among others. Moreover, newly discovered genetic links between DR3 signaling and human diseases provided key insights into the broad spectrum of functions mediated by the DR3/TL1A pathway. Clinical phenotypes associated with overexpression of TL1A and/or DR3 in genetically engineered animals have paved the way to identify diseases that might benefit from treatments targeting the DR3/TL1A axis. In this regard, IBD has been the primary disease target for assessing the efficacy of blocking TL1A/DR3 signaling in humans [82]. To date, three anti-TL1A drug candidates are being investigated in clinical trials: RVT-3101(originally PF-06480605) [82, 83], MK-7240 (originally PRA023) [84], and TEV-48574. Two phase 2 clinical trials have shown efficacy for ulcerative colitis (UC) (NCT04090411, NCT04996797), and there are several phase 2 (NCT05013905, NCT05499130) and phase 3 (NCT06052059) trials on-going for UC and CD patients. Depending on the success of these clinical trials, it is fair to expect that the range of the disease targets for these drugs can be broadened in the future, which is further supported by DR3 signaling being implicated in many other inflammatory conditions. Thus, the next decade will certainly bring exciting answers about the efficacy of these drugs into the clinic. Further research might also bring new revelations of this fascinating signaling pathway since many aspects regarding the differences in Treg and non-Treg cell activation in response to DR3 stimulation remain elusive. Finding answers to these questions might lead to the development of DR3-mediated Treg cell expansion and/or innate-like T cell costimulation based therapeutic strategies.

Highlights.

• Both proapoptotic and proinflammatory effects are attributed to DR3 activation.

• DR3 signaling in Foxp3⁺ Treg cells drives their expansion and effector functions.

• DR3 is expressed on iNKT cells and promotes their proliferation and activation.

• In conjunction with agonistic TCR signaling, DR3 acts as a costimulatory molecule.

• The role of DR3 in T cell immunity is intensely debated and requires further research.

Data availability:

No data were used for the research described in the article.