If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The C-terminal Tails of Tumor Necrosis Factor-related Apoptosis-inducing Ligand (TRAIL) and Fas Receptors Have Opposing Functions in Fas-associated Death Domain (FADD) Recruitment and Can Regulate Agonist-specific Mechanisms of Receptor Activation*
Department of Cancer Biology and Comprehensive Cancer Center, Wake Forest University School of Medicine, Winston-Salem, North Carolina 27157Department of Pharmacology, University of Colorado Health Sciences Center at Fitzsimmons, Aurora, Colorado 80045-0508
* This work was supported by grants from the Susan G. Komen Breast Cancer Foundation (to L. R. T.), the United States Army Breast Cancer Research Program (to L. R. T. and A. T.), and National Institutes of Health Grant GM61694 (to J. C. R.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. § Current address: Dept. of Microbiology and Immunology, Vanderbilt University Medical Center, Nashville, TN 37232. ∥ Current address: National Institutes of Health Chemical Genomics Center, Bethesda, MD 20892.
Members of the tumor necrosis factor (TNF) superfamily of receptors such as Fas/CD95 and the TNF-related apoptosis-inducing ligand (TRAIL) receptors DR4 and DR5 induce apoptosis by recruiting adaptor molecules and caspases. The central adaptor molecule for these receptors is a death domain-containing protein, FADD, which binds to the activated receptor via death domain-death domain interactions. Here, we show that in addition to the death domain, the C-terminal tails of DR4 and DR5 positively regulate FADD binding, caspase activation and apoptosis. In contrast, the corresponding region in the Fas receptor has the opposite effect and inhibits binding to the receptor death domain. Replacement of wild-type or mutant DR5 molecules into DR5-deficient BJAB cells indicates that some agonistic antibodies display an absolute requirement for the C-terminal tail for FADD binding and signaling while other antibodies can function in the absence of this mechanism. These data demonstrate that regions outside the death domains of DR4 and DR5 have opposite effects to that of Fas in regulating FADD recruitment and show that different death receptor agonists can use distinct molecular mechanisms to activate signaling from the same receptor.
The abbreviations used are: TNF, tumor necrosis factor; FasL, Fas ligand; TRAIL, TNF-related apoptosis-inducing ligand; DD, death domain; DED, death effector domain; DISC, death inducing signaling complex; FADD, Fas-associated death domain; TRADD, TNF receptor-associated death domain; GFP, green fluorescent protein; CHX, cyclo-heximide; mAb, monoclonal antibody; FACS, fluorescence-activated cell sorting.
1The abbreviations used are: TNF, tumor necrosis factor; FasL, Fas ligand; TRAIL, TNF-related apoptosis-inducing ligand; DD, death domain; DED, death effector domain; DISC, death inducing signaling complex; FADD, Fas-associated death domain; TRADD, TNF receptor-associated death domain; GFP, green fluorescent protein; CHX, cyclo-heximide; mAb, monoclonal antibody; FACS, fluorescence-activated cell sorting.
superfamily of receptors induce a variety of cellular responses including apoptosis, cellular differentiation, and proliferation. A subfamily of these receptors contains a death domain (DD) that is essential for transducing the apoptotic signal. Fas (CD95) is the best characterized member of this family. Binding of Fas Ligand (FasL) to a preformed Fas trimer (
). These activated receptors signal the apoptotic response by recruiting FADD to the cytoplasmic DD of the receptor to form the death-inducing signaling complex (DISC). NMR studies of the Fas DD indicate that it consists of six anti-parallel α-helices followed by an unstructured tail that negatively regulates Fas activity (
). This leads to cleavage to a fully processed, active form of the caspase that can dissociate from the receptor complex. Once activated, these caspases can cleave and activate effector caspases such as caspase-3 to induce the characteristic phenotypes associated with apoptosis.
Signaling by TNF-related apoptosis inducing ligand (TRAIL) is less understood. Interest in TRAIL has arisen because of its reported ability to induce apoptosis in tumor cells without affecting normal cells (
), suggesting that it may be useful for treating cancer. There are two “signaling” receptors, DR4 and DR5, which, similar to Fas, contain DDs followed by a short C-terminal tail. Agonistic antibodies that specifically recognize these receptors provide another way to induce apoptosis (
Here we show that FADD can bind directly to DR4 and DR5. However, unlike binding of FADD to Fas in which the C-terminal tail of Fas inhibits FADD recruitment, the C-terminal tails of DR4 and DR5 are required for efficient FADD binding, caspase cleavage and apoptosis. To better understand the importance of the C-terminal tail we made DR5-deficient BJAB cells then introduced wild-type DR5 or a DR5 mutant that lacks the C-terminal tail. While most DR5 stimuli required the C-terminal tail for apoptotic signaling, one agonistic antibody was able to induce apoptosis in the absence of the C-terminal tail, indicating that the process of receptor activation is agonist-dependent. These data suggest a model for TRAIL signaling in which the C-terminal tail that is outside the death domain provides an additional regulatory role.
MATERIALS AND METHODS
Reagents—Antibodies and reagents were purchased from the following sources: caspase-8 and caspase-3 antibodies (Cell Signaling, Beverly, MA), Anti-GFP (Chemicon, Temecula, CA), anti-actin (Sigma), recombinant human TRAIL (R&D Systems, Minneapolis, MN). Agonistic human monoclonal antibodies against DR4 (HGS-ETR1) and DR5 (HGS-ETR2 and R2-A) were provided by Human Genome Sciences (Rockville, MD). mAb631 (agonistic DR5) was from R&D Systems. Human IgG and mouse IgG, as well as anti-mouse and anti-human IgG Fc polyclonal antibodies were from Sigma.
Plasmids—FLAG-tagged versions of DR4 and DR5 in pcDNA3.1 were used for cloning (
). Full-length DR4 and DR5 cDNAs were cloned into the EcoRI and XhoI sites of pcDNA3.1-Puro (+). L334F DR5 was made by site-directed mutagenesis of pcDNA-DR5. Amino acids 1–454 of DR4 and 1–400 of DR5 followed by a TAG stop codon were cloned into the EcoRI and BamHI sites of pcDNA3.1-Puro(-) to generate DR4ΔT (14 amino acids truncation) and DR5ΔT (12 amino acid truncation). pGB14 contains the Gal4 DNA binding domain in a YCp vector. DR4 Cyto (amino acids 272–469), DR4 CytoΔT (amino acids 272–454), DR5 Cyto (amino acids 209–412), DR5 CytoΔT (amino acids 209–400), Fas Cyto (amino acids 177–335) and Fas CytoΔT (amino acids 191–320) were made by PCR from the corresponding cDNAs and cloning the resulting products into pGB14, pBTM-116, pEGFP-C2, and pFLAG-C2. Full-length catalytically inactive caspase-8 was used to generate pGB14-Caspase8. pACT3 plasmids have been described previously (
). Full-length FADD was cloned into the EcoRI and XhoI sites of pFLAG-C2. A more complete description of plasmids, maps, and sequences are available upon request.
TRAIL Receptor Surface Expression—Samples were prepared using the manufacturer's protocol and stained with antibodies against DR5 (R&D systems) or DR4, DcR1, and DcR2 (Alexis, San Diego, CA).
Cell Lines—HeLa cells were maintained in Dulbecco's modified Eagle's medium + 10% fetal bovine serum. BJAB and Jurkat cells were maintained in RPMI 1640 + 10% FBS. BJAB and Jurkat stable cell lines were generated by electroporating pcDNA3.1-Puro constructs and selecting for stable transformants as previously described (
DR5-deficient BJAB Cells—DR5-deficient BJAB cells were made by mutagenizing BJAB cells with 2 μg/ml ICR191 (Acros, St. Louis, MO) for 3 h. After a week of recovery, cells were treated with 1 μg/ml CHX and 1 μg/ml cross-linked mAb631 for 3 days. Cells were then washed three times with phosphate-buffered saline and plated in growth media + 1 μg/ml mAb631. After several days, clonal populations were produced by limited dilution. Individual clones were tested for resistance to mAb631 and screened for DR5 surface expression.
Immunoprecipitation—1.5 × 106 HeLa cells were transfected with 2 μg of pEGFP constructs plus 2 μg of FLAG constructs using FuGENE 6 (Roche Applied Science). Cells were lysed in Triton X-100 lysis buffer (50 mm Tris, pH 8.0, 150 mm NaCl, 5 mm EDTA, 1% Triton X-100, 1× protease inhibitors), and the soluble fraction was incubated with 30 μl of M2-agarose (Sigma) for 4 h at 4 °C. The beads were washed four times in TBS, and precipitated GFP fusions were detected by immunoblotting.
Two-hybrid Assays—Strains Y190 (Mata his3 ade2 trp1 leu2 gal4 gal80 cyh2, LYS2::Gal1-HIS3 URA3::Gal1-LacZ) and DY6877 (Mata ade2 can1 his3 leu2 lys2 trp1 URA3::8xLexA-LacZ) were used for directed two-hybrid assays.
Cytotoxicity Assays—Jurkat or BJAB cells were seeded in 96-well plates at a density of 1.0 × 106 cells per ml in growth medium + 1.0 μg/ml cycloheximide. Ligands were serially diluted then added to each well. Agonistic DR4 (HGS-ETR1) and agonistic DR5 (mAb631 or HGS-ETR2) were cross-linked with an equal amount of anti-human or anti-mouse Fc before serial dilution. After 22 h, MTS reagent (Promega, Madison, WI) was added to each well and incubated for an additional 2 h. Percent survival was calculated relative to control wells containing no ligand or antibody.
Caspase Activation—Jurkat or BJAB cells were seeded at a density of 1.0 × 106 cells per ml in growth media + 1.0 μg/ml cycloheximide. Agonistic antibodies were cross-linked with an equal amount of anti-mouse or anti-human Fc and unless otherwise stated, used at the following concentrations: mouse or human IgG, mAb631, and HGS-ETR2 at 50 ng/ml; human IgG and HGS-ETR1 (αDR4) at 200 ng/ml; TRAIL was used at 50 ng/ml. Cells were incubated for 6 h, washed once with phosphate-buffered saline, then harvested in Triton X-100 lysis buffer.
DISC Immunoprecipitations—2 × 107 BJABDR5 DEF cells expressing the various forms of DR5 were treated with 1.0 μg/ml mAb631 or 0.5 μg/ml ETR2 cross-linked with the appropriate anti-Fc for 30 min. Precipitations were carried out using agarose-protein A/G beads essentially as described (
Direct Binding of FADD to DR4 and DR5 Requires the C-terminal Tail of Each Receptor—To test if FADD could bind directly to DR4 and DR5, the cytoplasmic domains of each receptor (Fig. 1A, DR4 Cyto and DR5 Cyto) were tested for their ability to interact with FADD in a directed two-hybrid assay. As it has been reported that the C-terminal tail of Fas inhibits FADD binding (
), FADD bound to the cytoplasmic region of Fas only when the C-terminal tail was removed. Surprisingly, the opposite was true for DR4 and DR5; FADD was able to bind only when the C-terminal tails of DR4 and DR5 were intact. This was of interest because removal of the C-terminal tail leaves the six α-helices of the death domains intact, suggesting that interaction of FADD with DR4 and DR5 is regulated by regions outside the receptor death domain.
To test whether the C-terminal tail of DR4 and DR5 is required for interaction with FADD in mammalian cells, we performed co-precipitation assays in HeLa cells using FLAG-tagged FADD and GFP-tagged DR constructs. FLAG complexes were precipitated with an anti-FLAG antibody and interaction of each DR construct was detected by immunoblotting for GFP. Both DR4 Cyto and DR5 Cyto bound to FADD while DR fusions lacking the C-terminal tail did not (Fig. 1C).
If the C-terminal tail of DR4 and DR5 is necessary for FADD binding, then overexpression of the cytoplasmic domains of these receptors should induce apoptosis while DR4 or DR5 lacking the C-terminal tails should not. To test this hypothesis, we transfected GFP-tagged DR5 constructs into HeLa cells. The DR5 cytoplasmic domain (DR5 Cyto) caused cells to round up and die while either GFP alone or the DR5 cytoplasmic domain without the C-terminal tail (DR5 CytoΔT) had no effect (Fig. 1D). Similar results were obtained with GFP-tagged versions of DR4 (data not shown). These data suggest that FADD binding and apoptosis through DR4 and DR5 require the C-terminal tail of the receptor. Surprisingly, this is the opposite of the Fas receptor where the C-terminal tail is inhibitory to FADD binding and receptor function (
The C-terminal 14 Amino Acids of DR4 Are Required for Apoptosis and Caspase Activation—To test whether the 14 C-terminal amino acids of DR4 are necessary for transducing the apoptotic signal by receptor activation, we made BJAB cells that stably express GFP and DR4ΔT (removal of amino acids 455–469 from full-length DR4 constituting the C-terminal tail). We hypothesized that DR4ΔT should act as a dominant-negative and inhibit DR4 induced apoptosis because it can interact with endogenous receptors and bind ligand but should inhibit the apoptotic signal as it is unable to bind FADD. We obtained several DR4ΔT clones and the expression level of DR4 for two representative clones is shown in Fig. 2A. Because BJAB cells express both DR4 and DR5, we selectively induced TRAIL receptor-mediated apoptosis using agonistic monoclonal antibodies that are specific for each receptor. All agonistic antibodies are IgG1 subtypes and were cross-linked with anti-IgG antibody before treatment. BJAB-GFP and BJAB-DR4ΔT cells were treated with increasing amounts of a DR4 agonistic antibody (αDR4) and cytotoxicity was measured. Clones expressing DR4ΔT, which has an intact DD but no C-terminal tail were completely resistant to αDR4 while GFP-expressing cells were sensitive (Fig. 2B). To test if this was caused by the stable expression of DR4ΔT and not some other defect in apoptosis, the same cells were treated with an agonistic antibody against DR5 (αDR5). As shown in Fig. 2C, the dose responses for αDR5-induced cytotoxicity in cells expressing GFP and DR4ΔT are nearly identical indicating apoptotic signaling downstream of the DR4 receptor is intact.
The immediate effect of ligand binding to a death receptor is the recruitment of FADD and procaspase-8. This leads to caspase-8 activation and proteolytic cleavage (
). If FADD is unable to bind to DR4ΔT, then cells expressing this construct should not process caspase-8 upon DR4 receptor activation. BJAB cells were treated with either nonspecific IgG or αDR4, and caspase-8 processing was examined by immunoblotting. Caspase-8 was processed from its proform into its p43 and p41 cleavage products as well as the p18 active subunit in BJAB cells expressing GFP whereas clones expressing DR4ΔT showed only very small amounts of the p43/p41 subunit (Fig. 2D). Caspase-3 processing was also examined as it is cleaved and activated by caspase-8. GFP-expressing cells showed high amounts of caspase-3 processing in response to αDR4 while DR4ΔT expressing clones showed very little. We conclude that the 14 C-terminal amino acids of DR4 that are outside the death domain are required for efficient FADD binding and activation of the apoptotic cascade.
The C-terminal 12 Amino Acids of DR5 Are Required for Apoptosis and Caspase Activation—To test whether the C-terminal tail is required for DR5 function, we made BJAB cells that stably express DR5ΔT (full-length receptor lacking amino acids 401–412 constituting the C-terminal tail). The expression level of DR5 for two representative clones as well as GFP control cells is shown in Fig. 3A. Clones expressing DR5ΔT showed increased resistance to αDR5 compared with GFP cells (Fig. 3B) as indicated by an approximate 50-fold shift in the dose response. Apoptotic signaling was otherwise intact because these cells were still sensitive to αDR4 (Fig. 3C). The same cells were treated with nonspecific IgG or αDR5 and caspase processing was examined by immunoblotting. As shown in Fig. 3D, both caspase-8 and caspase-3 were processed into their active subunits in BJAB-GFP cells but not in the clones expressing DR5ΔT when stimulated with αDR5. We therefore conclude that similar to DR4, the C-terminal tail of DR5 is required for cytotoxicity and caspase processing.
Identification of DR5-deficient BJAB Cells and Complementation with DR5 Transgene—Because experiments using dominant negative proteins rely on overexpression, we examined the effects of DR5ΔT in a DR5-deficient background. Because no DR5-deficient cell lines have been identified (
), we made DR5-deficient BJAB cells. BJAB cells were randomly mutagenized with ICR191 followed by selection for clones resistant to αDR5-induced death (Fig. 4A). Because resistance could be caused by loss of DR5 or some other component of the apoptotic pathway such as FADD or caspase-8, each clone was also tested for sensitivity to αDR4 and FasL. We reasoned that clones with no functional DR5 receptor should be sensitive to these two stimuli while those with defects in downstream signaling would be resistant. The subset of clones sensitive to DR4 and FasL were screened for loss of DR5 surface expression by FACS. A clone we designated BJABDR5 DEF did not express any detectable DR5, was resistant to mAb631 and sensitive to both αDR4 and FasL (Fig. 4, B–E).
If the defect in BJABDR5 DEF cells is loss of DR5 expression rather than loss of receptor transport as has been recently observed (
) or unrelated anti-apoptotic defects, then replacement of DR5 into these cells should restore DR5 signaling. We introduced DR5 into BJABDR5 DEF cells and identified a clone that stably expressed surface DR5 at levels comparable to wild-type BJAB cells as determined by FACS analysis (Fig. 4B). These cells showed similar cytotoxicity compared with wild-type BJAB cells when treated with αDR5, αDR4, and FasL (Fig. 4, C–E). Because replacement of DR5 into BJABDR5 DEF cells completely restores DR5 induced signaling, we conclude that the only DR5 signaling defect in BJABDR5 DEF cells is lack of DR5 expression.
The C-terminal Tail of DR5 Is Required for Signaling in Response to mAb631 but Not HGS-ETR2—To test whether DR5ΔT had activity in the absence of endogenous DR5 receptor, DR5ΔT was stably expressed in BJABDR5 DEF cells. As a control for a DR5 receptor that cannot bind FADD because of a death domain defect, DR5 (L334F) was also expressed in the DR5-deficient background. A mutation of leucine 334 to phenylalanine in the death domain of DR5 is analogous to the lpr mutation in Fas, which abolishes receptor function by preventing FADD binding (
). Neither DR5 (L334F) nor DR5ΔT are able to bind FADD as determined by co-immunoprecipitation experiments in mammalian cells (Fig. 5A). To test whether these mutations completely destroyed protein function, we tested each DR5 construct for interaction with the cytoplasmic domain of wild-type DR5. DR5 forms homomeric complexes and this association occurs through interaction between the DR5 cytoplasmic domain (
). Each DR5 construct co-precipitated with FLAG-DR5 indicating that the cytoplasmic domains of DR5 (L334F) and DR5ΔT are functionally intact (Fig. 5B).
We next made cell lines stably expressing DR5 (L334F) or DR5ΔT. FACS analysis was used to identify clones expressing surface levels of DR5 (L334F) or DR5ΔT similar to BJABDR5 DEF-DR5 (Fig. 5C). Cytotoxicity in BJABDR5 DEF cells or clones expressing exogenous DR5, DR5 (L334F) and DR5ΔT was measured after treatment with several DR5 agonistic antibodies. When treated with mAb631, the agonistic DR5 antibody used in previous experiments, BJABDR5 DEF cells expressing wild-type DR5 were sensitive whereas cells expressing DR5ΔT or DR5 (L334F) were resistant (Fig. 5D). A similar pattern of sensitivity was observed when the same cells were treated with a second DR5 agonistic antibody, R2-A (data not shown). These observations were consistent with our previous data in which removal of the C-terminal tail from DR5 prevented signaling through the receptor. A different result was obtained when we treated cells with a third DR5 agonistic antibody, HGS-ETR2. BJABDR5 DEF cells expressing exogenous DR5 or DR5ΔT were sensitive (Fig. 5E) although DR5ΔT cells required higher doses of HGS-ETR2. In contrast, cells expressing DR5 (L334F) were completely resistant to HGS-ETR2. These data suggest that the tail of DR5 is not absolutely required for signaling induced by HGS-ETR2 as it is with other DR5 agonistic antibodies. Since DR5 (L334F) cells are resistant to every DR5 agonist but DR5ΔT cells are differentially sensitive we conclude that the death domain is essential while the tail of DR5 mediates sensitivity of the receptor to different agonists.
Because HGS-ETR2 was able to induce apoptosis in BJABDR5 DEF cells expressing both DR5 and DR5ΔT, but mAb631 was able to induce apoptosis only in cells expressing DR5, we reasoned that these two antibodies might differentially recruit FADD to the truncated DR5 receptor. To test this hypothesis we performed DISC immunoprecipitation experiments with BJAB DR5 DEF cells or cells expressing DR5, DR5 (L334F), or DR5ΔT and immunoblotted for FADD. mAb631 recruited FADD to the DR5 receptor only in BJABDR5 DEF cells expressing wild-type DR5 (Fig. 5F). However, ETR2 was able to recruit FADD in cells expressing DR5 and cells expressing DR5ΔT. FADD was recruited to a lesser extent in cells expressing DR5ΔT compared with DR5 consistent with the dose response data in which DR5ΔT-expressing cells were sensitive to ETR2, but to a lesser extent than DR5-expressing cells.
Next, we assessed caspase processing in response to saturating doses of mAb631 (1 μg/ml) and HGS-ETR2 (500 ng/ml). mAb631 was able to induce caspase-8 and caspase-3 processing only in BJABDR5 DEF cells expressing DR5 (Fig. 5G). Consistent with the dose response data, treatment with HGS-ETR2 resulted in the processing of caspase-8 and caspase-3 in BJABDR5 DEF cells expressing DR5, and cells expressing DR5ΔT. These data suggest that the C-terminal tail of DR5 is necessary for FADD recruitment through most stimuli. Because other stimuli such as the HGS-ETR2 antibody can overcome this defect and recruit FADD to induce caspase activation, diverse agonists to the DR5 receptor might activate the receptor through different mechanisms.
TRAIL Signaling through DR5 Is Prevented by Removal of the C-terminal 12 Amino Acids—We next wished to assess whether TRAIL activation of DR5 requires the C-terminal tail similar to mAb631 or whether TRAIL can activate the receptor independent of the C-terminal tail like HGS-ETR2. To test any inhibitory effect associated with removal of the C-terminal tail in response TRAIL, we made Jurkat cells that stably express GFP or DR5ΔT. Jurkat cells express very low levels of DR4 (
) so almost all TRAIL signaling is through DR5. We identified several clones that expressed high levels of exogenous DR5ΔT all with similar responses to TRAIL; the expression level of DR5ΔT in a representative clone is shown in Fig. 6A. Jurkat cells expressing DR5ΔT showed decreased cytotoxicity compared with cells expressing GFP when treated with TRAIL (Fig. 6B). Apoptosis signaling was still intact in DR5ΔT Jurkat cells because all cell types were equally sensitive to death from FasL (Fig. 6C). When treated with TRAIL, Jurkat-GFP cells showed high levels of caspase-8 and caspase-3 processing, whereas cells expressing DR5ΔT showed very little (Fig. 6D). These data indicate that TRAIL signaling through DR5 requires the C-terminal tail similar to signaling with mAb631.
Although there is much interest in TRAIL signaling because of its reported ability to kill tumor cells through FADD-dependent apoptosis while not affecting normal cells, the mechanism by which FADD is recruited to DR4 and DR5 is unclear. While some reports implicate TRADD (
), we thought that removal of the corresponding regions of DR4 and DR5 would enhance interaction of FADD with these receptors. Interestingly, the opposite result was obtained and removal of the C-terminal tails (14 amino acids from DR4 and 12 amino acids from DR5) inhibits FADD binding. Although Fas and TRAIL receptors are similar in that FADD is recruited directly, this suggests that the requirements for FADD binding are different. Removal of the C-terminal tails from DR4 and DR5 leaves the DD structurally intact because DR4 and DR5 mutants that lack the C-terminal tail are still able to bind their full-length counterpart. Thus, unlike the situation with Fas, the DDs of TRAIL receptors are not sufficient for FADD binding. Fas is the only mammalian death receptor with a solved structure for the DD (
) observed that the region encompassing the death domain was well defined while the carboxyl terminal amino acids were disordered. Structural studies of the DR4 and DR5 intracellular domain will be important and may demonstrate that the corresponding C-terminal region of TRAIL receptors is more structured.
DR4 and DR5 can form homomeric or heteromeric complexes (
). Thus, one would expect that the tail-less DR4 receptor might partially block signaling through DR5 because this heteromeric complex should not bind FADD through the DR4ΔT cytoplasmic domain. However, we observed that signaling through DR5 that was activated by the DR5-specific antibody was unaltered in cells expressing DR4ΔT. The same was true for cells expressing DR5ΔT; responses to αDR4 were similar to those of GFP-expressing cells. This suggests that heteromeric complexes do not play an important role in apoptotic signaling through DR4 and DR5. In agreement with this view, Kischkel et al. (
). In this case the anti-apoptotic protein Bcl-xL could block death induced by Fas agonistic antibodies but not FasL suggesting that different agonists can signal through distinct mechanisms. Our data uncover a potential mechanism for such differential sensitivity in TRAIL signaling because different agonistic antibodies can display different requirements for regions in the intracellular domain of DR4 or DR5. Thus, binding of different agonists to the extracellular domain causes subtly different conformational changes in the intracellular domain that define the FADD binding surface of the receptor.
We propose a model for TRAIL signaling in which the C-terminal tail plays a regulatory role in the recruitment of FADD to DR5 (and DR4). In this model, DR5 exists as a preformed trimer held together by interactions between the cytoplasmic domains of each receptor. In the absence of bound ligand the FADD binding surface of the trimers is not accessible to FADD (Fig. 7A). Binding of TRAIL or an agonistic antibody to the receptor trimer causes a conformational change that is normally facilitated by the C-terminal tail. This alteration exposes the FADD binding surface leading to the recruitment of FADD and initiation of the apoptotic cascade. In the absence of the C-terminal tail, binding of some ligands such as TRAIL or mAb631 is not sufficient to trigger this conformational change (Fig. 7B), FADD cannot be recruited and there is no apoptotic signal. However, some agonists such as HGS-ETR2 are able to overcome this layer of regulation and recruit FADD even in the absence of the C-terminal tail (Fig. 7C). These data indicate that activation of TRAIL receptors has mechanistic differences compared with activation of the Fas receptor and raise the possibility that it will be feasible to activate subtly different signaling mechanisms from the same receptor using different agonists.
We thank Dr. Marcus Peter for providing BJAB cells and Minna Poukkula for help with DR4 and DR5 receptor staining.