The Adaptor Protein TRADD is Essential for TL1A/DR3 Signaling

Abstract

TRADD is a key effector protein of Tumor Necrosis Factor Receptor-1 (TNFR1) signaling. However, the role of TRADD in other death receptor signaling pathways, including Death Receptor-3 (DR3), has not been completely characterized. Previous studies utilizing overexpression systems suggested that TRADD is recruited to the DR3 complex in response to the DR3 ligand, TL1A, indicating a possible role in DR3 signaling. Using T cells from TRADD knockout mice, we here demonstrate that the response of both CD4+ and CD8+ T cells to TL1A is dependent upon the presence of TRADD. TRADD knockout T cells therefore lack the appropriate proliferative response to TL1A. Moreover, in the absence of TRADD, both the stimulation of MAP kinase signaling and activation of NF-κB in response to TL1A are dramatically reduced. Unsurprisingly, TRADD is required for recruitment of RIP1 and TRAF2 to the DR3 signaling complex and for the ubiquitination of RIP1. Thus, our findings definitively establish an essential role of TRADD in DR3 signaling.

Introduction

Death Receptor 3 (DR3, TNFRSF25, WSL-1, LARD, TRAMP, Apo3) is a member of the TNF-receptor superfamily and a member of death receptor subfamily that has seven additional members, including TNFR1, FAS, DR4, DR5, DR6, NGFR and EDAR (1). All the receptors of this subfamily contain a death domain (DD) as a part of their intracellular domain, and of these seven, DR3 has the highest homology to TNFR1 (2–5). However, in contrast to the ubiquitous expression of TNFR1, DR3 expression is restricted to lymphocyte-enriched tissues such as thymus, spleen, small intestine and PBL (2–6), and has been shown to be especially upregulated in activated T cells (5, 6). The ligand to DR3 receptor is a member of the TNF superfamily and is called TL1A (TNF-like ligand 1A or TNFSF15). Originally thought to be predominantly expressed by endothelial cells (7), later studies have shown that TL1A can be produced by variety of other cells including monocytes, macrophages, plasma cells, dendritic cells and T cells (8–10).

TNFR1 signaling has been very well characterized. The TNFα-dependent activation of TNFR1 leads to formation of a signaling complex that contains several important molecules, including TRADD, RIP and TRAF2 (11–13). These molecules are responsible for mediating the activation of MAP kinases and NF-κB. In some cases, a secondary complex (complex II) dissociates from the receptor and associates with FADD (14), and this complex is responsible for caspase activation and apoptosis. This is in contrast to other death receptors, such as Fas, DR4 and DR5, which are capable of binding directly to FADD in the primary receptor complex (1). The generation of RIP1 and TRAF2 knock-out mice did much to unveil the important functions of these molecules in MAP kinase and NF-κB activation downstream of TNFR1 (15, 16). However, TRADD has been always considered to be the first adaptor molecule that is recruited to TNFR1 (17) and TRADD-deficient mice generated recently by our lab (18) and others (19, 20) made it possible to definitively establish the role of TRADD in TNFR1 signaling that occurs mainly through recruitment of these molecules.

A majority of studies on the mechanisms of DR3 signaling have been conducted previously using overexpression systems. Similar to TNFR1, DR3 induces both NF-κB activation and apoptosis when ectopically overexpressed in cell lines (2–5). Of the adaptor proteins tested in overexpression studies, TRADD was the only protein that had a strong interaction directly with DR3 (2, 4), while the association of RIP and TRAF2 with DR3 was very week and greatly enhanced in the presence of transfected TRADD (4). In these systems FADD was also shown to be recruited to DR3 upon coexpression of TRADD, and association of FADD directly with DR3 was observed only at very high expression levels (3). In addition, DR3-mediated apoptosis was dependent on the presence of FADD and caspase-8 in embryonic mouse fibroblasts (15, 21). These studies do suggest that TRADD is the primary adaptor recruited to DR3 upon TL1A stimulation. However, a detailed examination of the physiological role of TRADD in DR3 signaling pathways has not yet been undertaken.

We therefore sought to examine DR3 signaling events in cells derived from TRADD knockout mice. Here we show that in the absence of TRADD, TL1A cannot induce proper JNK and NF-κB activation. This is largely due to the inability of DR3 to effect the recruitment of the critical adaptor proteins TRAF2 and RIP1 in TRADD-deficient cells, thus leading to a nonfunctional DR3 signaling complex. Moreover, TL1A does not promote T cell proliferation in TRADD-deficient cells. We also confirm that TL1A-induced DR3 signaling does not result in T cell apoptosis.

Materials and Methods

Reagents

Anti-phospho-JNK antibody was purchased from Biosource. Anti-TRADD and TRAF2 antibodies were purchased from Santa Cruz Biotechnology. Anti-JNK, RIP, Fas, CD28, CD3ε, anti-CD4-PE and anti-CD8-APC antibodies were from Pharmingen. Anti-phospho-IκBα, anti-phospho-p38, anti-p38, anti-phospho ERK and anti-ERK were purchased from Cell Signaling. Recombinant murine TL1A, TNFα, IL-2 and anti-DR3 antibody were from R&D Systems. PMA was from Alexis Biochemicals. SMAC mimetic (SM-164) was a kind gift from Dr. Shaomeng Wang.

Animals

TRADD knockout mice were described previously (18). 3–4 week old TRADD knockout and wild type littermates were used to isolate T cells.

T cell purification

Purification of CD4⁺ and CD8⁺ lymph node T cells or pan T cells was performed using MACS CD4⁺ and CD8⁺ T cell or pan T cell isolation kits (Miltenyi Biotec), respectively according to the manufacturer’s instructions. Briefly lymph node cell populations were depleted of non-T cells by incubation with biotin-conjugated antibody cocktail and anti-biotin magnetic micro-beads followed by separation on MACS columns. Purity of isolated T cells was more than 97%.

Treatment of cells

Before any treatment T cells were pre-activated with plate-bound anti-CD3 (1 μg/ml) antibodies for overnight.

In vitro cell proliferation assay

T cells were washed with PBS twice, resuspended at 10⁶/ml in PBS, warmed in 37°C water bath for 10 min, mixed with 1 μl of 5mM carboxy-fluorescein diacetate, succinimidyl ester stock (Vybrant CFDA SE cell tracer kit, Molecular Probes Invitrogen) and incubated in 37°C water bath for 10 min. Reaction was stopped with ice cold complete RPMI 1640 media. Cells were then plated in 24-well plates at 0.5×10⁶/well, harvested after 72 hours, stained with anti-CD4 or CD8 Abs and analyzed on FACS Calibur (Becton Dickinson) using FlowJo software.

Western blot analysis and co-immunoprecipitation

After treatments as described in the figure legends, cells were collected and lysed in M2 buffer (20mM Tris at pH 7, 0.5 % NP-40, 250mM NaCl, 3mM EDTA, 3mM EGTA, 2mM DTT, 0.5mM PMSF, 20mM β glycerol phosphate, 1mM sodium vanadate, 1μg/ml leupeptin). Cell lysates were fractionated by SDS-polyacrylamide gel and Western blotted. The proteins were visualized by enhanced chemiluminescence (ECL), according to the manufacturer’s instruction (Amersham). For immunoprecipitation assays, cells were treated with TL1A as indicated in the legend and then collected in lysis M2 buffer. The lysates were mixed and precipitated with anti-DR3 antibodies (R&D) and protein G-Agarose beads by incubation at 4 °C for overnight. The beads were washed five times with lysis buffer, and the bound proteins were resolved in 4–20 % SDS-polyacrylamide gels and detected by Western blot analysis.

Electromobility shift assay (EMSA)

Nuclear extracts were isolated using the Biovision Nuclear/Cytosol fractionation kit following the manufactures guidelines. Gel shifts were performed with the Promega Gel shift assay system using 10 μg of nuclear extract and following the recommendations outlined in the manufacturer’s protocol. All consensus oligos were also purchased from Promega.

Cytotoxicity assay

TL1A- and Fas-induced cell death was determined using tetrazolium dye colorimetric test (MTS test). The MTS absorbance was then read using a plate reader at 490 nm.

Results

TL1A can induce several events in T cells, including the activation of NF-κB and MAP kinases (7, 22). In addition, TL1A also enhances proliferation and cytokine production of T cells stimulated through the T cell receptor (23). We therefore sought to determine whether TRADD was important for these events. CD4⁺ and CD8⁺ T cells were purified from lymph nodes of wild-type and TRADD^−/− mice and labeled with CFSE. Since TL1A increases T cell proliferation better when the CD28 costimulatory signal is absent (23), we pre-activated T cells with anti-CD3 only and treated with TL1A or IL-2 for 72 hours. In contrast to wild-type T cells, which proliferated very well in response to TL1A (as seen by the dilution peaks on the CFSE histogram), CD4⁺ TRADD-deficient T cells showed very minimal proliferation (Fig.1a). Similar differences were seen in the CD8⁺ T cells (Fig. 1b). However, IL-2-induced proliferation was minimally affected by the absence of TRADD in either CD4⁺ or CD8⁺ T cells, suggesting that the defect in the proliferative response was specific for TL1A treatment.

Figure 1. TRADD-deficient T cells do not proliferate in response to TL1A treatment.

CFSE labeled wild type and knockout T cells were treated with TL1A (50 μg/ml; top panels) or IL-2 (100U/ml; bottom panels) for 72 hours, then stained with anti-CD4 or anti-CD8 antibodies and analyzed by flow cytometry. Histograms are gated on CD4⁺ (a) and CD8⁺ cells (b).

Next we sought to determine whether the lack of proliferation was due to a deficiency in the downstream signaling pathways, and whether TRADD has a role in DR3-mediated MAP kinase signaling and NF-κB activation. We examined phosphorylation of JNK and IκBα in CD3-activated T cells after TL1A treatment. As shown on Fig. 2a and 2b, wild-type CD4⁺ and CD8⁺ T cells demonstrated a potent activation of these two pathways. In contrast, TL1A induced only a very weak phosphorylation of JNK and minimal phosphorylation of IκBα in TRADD-deficient T cells. Similar results were obtained if T cells were pre-activated with CD3 in the presence of CD28 or if they were not pre-activated at all (data not shown). This was true despite similar expression levels of DR3 in wild-type and knockout cells (data not shown). The defect in signaling was specific for TL1A since the PMA-dependent activation of JNK was normal in both wild type and TRADD knockout cells (Fig. 2c). Phosphorylation of p38 and ERK was not observed in response to TL1A stimulation, whereas both kinases were strongly activated when T cells were treated with PMA (Supplemental Fig.1).

Figure 2. TRADD is required for MAP kinase and NFκB activation.

a–c, Western blot analysis of lysates of wild type and TRADD knockout CD4⁺ (a), CD8⁺ (b) T cells and pan-T cells (c) treated with TL1A (50 μg/ml) (a,b) for the indicated times or PMA (100nM) (c). d, Nuclear extracts of wild type and knockout T cells treated with TNFα (30 ng/ml) and TL1A (50 μg/ml) for indicated times were analyzed by EMSA.

We also assessed NF-κB DNA-binding activity in wild-type T cells versus TRADD-deficient T cells using EMSA. TNFα treatment was used as a positive control (lane 1). Strong NF-κB DNA-binding activity was observed in wild-type T cells 30 minutes after TL1A treatment and was decreased by 4 hours (Fig. 2d). However we observed little NF-κB binding to DNA in TRADD-deficient T cells in the presence or absence of both stimuli and we observed little, if any, increase at 6 h (Supplemental Fig.2) or 8h (not shown). Thus our data suggest that TRADD is necessary for potent DR3-mediated signaling in response to TL1A.

Based on the data obtained earlier from the overexpression experiments (2, 4) it is believed that, similar to TNFR1, DR3 interacts with TRADD and then TRADD recruits RIP and TRAF proteins to the signaling complex. To observe the formation of the physiological DR3 signaling complex we performed immunoprecipitation experiments with DR3-specific antibodies. In agreement with the previous results, TRADD, RIP and TRAF2 were recruited to DR3 after 5 minutes of TL1A treatment in wild-type T cells (Fig.3). Similar to TNFR1-associated RIP, DR3-associated RIP was presumed to be ubiquitinated based on the pattern seen in western blotting with an anti-RIP antibody. In contrast, in TRADD-deficient T cells we observed neither RIP1 nor TRAF2 recruitment to DR3, indicating that TRADD is required for the recruitment of these two critical signaling molecules to the DR3 signaling complex in T cells.

Figure 3. TRADD is required for formation of the functional DR3 signaling complex.

Cell extracts from wild type and TRADD knockout T cells treated with TL1A (50 μg/ml) for 5 min were immunoprecipitated with anti-DR3 (ip:DR3) and analyzed by western blotting. Input, 2% of extract before immunoprecipitation. Ub-, ubiquitinated.

It has been shown previously that overexpression of DR3 induces apoptosis in cell lines (2–5) and the human leukemia cell line TF-1, which has high endogenous levels of DR3, undergoes apoptosis when treated with TL1A in the presence of cycloheximide (7, 22). However, consistent with a previous study (7), we confirmed that primary T cells from lymph nodes were resistant to cell death under these conditions (data not shown). We also tested whether thymocytes can be sensitized to cell death. T cells from wild type and TRADD knockout mice were treated with cycloheximide and TL1A or anti-Fas antibodies for 16 hours. Apoptosis induced by Fas crosslinking was the same in wild type and TRADD-deficient thymocytes (Fig. 4a, Supplemental Fig.3a). No cell death was observed after TL1A treatment in the presence or absence of cycloheximide. Additionally, crosslinking of the flag-tagged TL1A with the flag-specific M2 antibodies did not result in apoptosis of the T cells (data not shown). Furthermore, when the SMAC mimetic SM-164 was used instead of cycloheximide it again sensitized thymocytes to Fas- but did not result in TL1A-induced apoptosis (Fig. 4b).

Figure 4. TL1A does not induce apoptosis in primary T cells.

MTS assay of the viability of wild type and TRADD knockout T cells treated for 24 hours with cycloheximide (C) alone (10μg/ml), cycloheximide and TL1A (50 μg/ml) or cycloheximide and anti-FAS (10ng/ml) (a) and SMAC mimetic (SM) alone (10nM), SMAC mimetic and TL1A (50 μg/ml) or SMAC mimetic and anti-FAS (10ng/ml) (b). Error bars shown represent SEM.

Discussion

Due to the lack of knockout mice the involvement of adapter protein TRADD in TNFR1 signaling had been a controversial issue for a long time. Recently our lab and others have generated TRADD knockout mice and have demonstrated that TRADD is indispensable for MAP kinase and NF-κB activation as well as apoptotic and necrotic cell death in TNFR1 signaling (18–20). Interestingly, TRADD turned out to be not solely a TNFR1 adaptor molecule, but also an adaptor in several other signaling pathways. In the absence of TRADD, TRIF-dependent signaling events downstream of TLR3 and TLR4 are attenuated, suggesting that TRADD functions outside of death receptor signaling (18–20). TRADD has also been proposed to serve as a negative regulator of IFN-γ signaling through the formation of a complex with STAT1α (24). The antiviral pathways mediated through RIG-like helicases is also dependent on TRADD, and TRADD is a necessary adaptor to mediate the downstreamCardif-dependent immune responses, such as the activation of the NF-κB and IRF3 transcription factors and the production of IFN-β (25). A role for TRADD in other death receptor signaling pathways has also been suggested. The p75^NTR receptor was shown to interact with TRADD in breast cancer cells after stimulation with nerve growth factor (26). Most other studies in relationship to other death receptors have relied on overexpression studies. For example, both DR4 and DR5 can interact with TRADD when overexpressed in cell lines (27). Likewise, the implication of the involvement of TRADD in the DR3 signaling pathway has also largely been based on overexpression data.

We took advantage of our recent generation of TRADD knockout mice and demonstrated that under more physiological conditions, TRADD plays an important role in DR3 signaling. Importantly, based on our immunoprecipitation experiments, TRADD is required for the recruitment of both RIP and TRAF2 to the DR3 complex in T cells, suggesting that TRADD is the primary adapter molecule involved in the DR3 signaling pathway, just as it is in TNFR1 signaling. This is consistent with the closer homology of the DR3 death domain to the TNFR1 receptor (40% identical in mouse) than other FADD binding death receptors are to TNFR1, such as the TRAIL receptor DR4/5 (28% identity), and Fas (13% identity).

Due to loss of complex formation, the TL1A-induced phosphorylation of JNK and IκBα were dramatically reduced in TRADD deficient T cells and no NF-κB DNA-binding activity was observed. A very small amount of residual phosphorylation of JNK and IκBα may be due to weak interaction of RIP and TRAF2 with DR3 (4). The downregulation of signaling appears to have a physiologically relevant consequence in TRADD-deficient T-cells, since we observed that the TL1A-induced proliferation of both CD4⁺ and CD8⁺ T cells was abrogated. Our data is consistent with recent data that also showed that TRADD deficient CD4⁺ T cells failed to proliferate in response to TL1A treatment (20).

DR3 was given its name Death Receptor-3 because of the death domain present in the C-terminal intracellular domain of the protein. While receptors with death domains such as Fas and DR4/5, are clearly associated with cell death in physiological contexts, in previous studies TL1A has only been shown to induce apoptosis in overexpression systems and in immortalized cell lines that endogenously express DR3, while primary T cells, the likely physiological TL1A targets, did not undergo cell death when treated with TL1A (7). Our findings are consistent with this study as we were unable to induce cell death in primary T cells in response to TL1A. Interestingly, the presence of cycloheximide, which usually sensitizes most TNFR1-expressing cells to TNFα-induced cell death, did not sensitize T- cells to cell death. Likewise, SMAC mimetic, while sensitizing the primary T-cells to FasL, did not sensitize the cells to TL1A. While we did not examine this possibility, these data may suggest that DR3 is not capable of inducing the formation of a secondary “complex II”-type complex (14) where FADD is recruited and then activates caspase-8. Thus, in contrast to TNFR1 signaling that can activate both survival and death pathways depending on the cellular context, the TL1A-DR3 pathway may be solely a pro-inflammatory and pro-survival pathway, rather than also acting as a pro-apoptotic pathway.

DR3 has been implicated in many inflammatory and autoimmune diseases including graft-versus-host disease (7), allergic lung inflammation (28), inflammatory bowel disease (8, 9), rheumatoid arthritis, and experimental autoimmune encephalomyelitis (23, 29). In all the cases TL1A plays an activating role, augmenting T cell responses and worsening the course of diseases, consistent with the fact that stimulation of DR3 does not lead to apoptosis in primary T cells. Since TNFα also plays an important role in inflammation in many of these diseases, and TRADD is involved in both DR3 and TNFR1 signaling pathways, approaches targeting TRADD might be taken into consideration when designing new therapeutic strategies for treatment of inflammatory T cell-mediated diseases.