c-E10 Is a Caspase-recruiting Domain-containing Protein That Interacts with Components of Death Receptors Signaling Pathway and Activates Nuclear Factor-κB*

Members of the tumor necrosis factor receptor superfamily induce apoptosis via interaction with FADD and regulate cell growth and differentiation through TRADD and TRAFs molecules. While screening for molecules involved in the regulation of death receptor signaling, we identified a novel protein, c-E10. c-E10 contains an amino-terminal caspase-recruiting domain (CARD) and shares a sequence homologous with E10, a viral CARD-containing protein that binds to c-E10. In transfection experiments c-E10 oligomerizes, binds to the cytoplasmic portion of TRAIL receptor 1 (DR4) and coprecipitates with TRADD. Expression of c-E10 under the control of a doxycycline-dependent transcriptional transactivator results in NF-κB activation, which is inhibited by dominant negative forms of TRAF2 and NIK kinase. Thus, our results suggest that c-E10 is an adapter protein that activates NF-κB through a molecular pathway involved in death receptor signaling.

Members of the TNF 1 receptor superfamily, which comprises TNFR-1 and -2, Fas/APO-1, DR3, and the TRAIL receptors DR4 and DR5, play a significant role in regulating many biological functions, including cell activation and proliferation, cytokine production, and apoptosis (1,2). Two distinct signaling cascades, leading either to cellular apoptosis or activation of the transcription factors NF-B, are elicited by these receptors, and both rely on the association of proteins containing conserved domains, including the death domain (3), the death effector domain, and the CARD motif.
The signal transduction machinery that couples a subset of these receptors to initiation of the cell death cascade is well understood. In the case of Fas, for example, FasL binds to and induces Fas receptor clustering, which results in FADD recruitment to the receptor complex (4,5). FADD, in turn, binds to caspase-8, thereby connecting death receptors to the caspase proteases (2,6,7). Present in the cell as proenzymes, these cysteine proteases are rapidly activated by proteolytic cleavage following death receptor stimulation and, once activated, constitute the effector component of the apoptotic machinery (8). Under particular circumstances, however, exposure to TNF results in the activation of the transcription factors NF-B and AP1 through the recruitment of the IKK complex and the JNKs pathway, which mediate the induction of immunoregulatory genes and exert positive effects on cell survival and proliferation (9 -13).
TRADD is a TNFR-1-associated signal transducer involved in activating both pathways (14). This adapter protein, which is recruited to TNFR-1 upon TNF binding, interacts with FADD to initiate the death cascade and engages several proteins such as TRAF1, TRAF2, and RIP that regulate the activation of noncytotoxic pathways (13,(15)(16)(17)(18). Thus, TRADD is capable of transducing signals that either trigger or suppress cell death. Regulation of molecular interactions in close proximity to the death receptor complex is therefore essential in determining whether a cell dies or survives following TNF stimulation. In this study, we present a novel protein, c-E10, that coprecipitates with TRADD and activates NF-B through a TRAF 2/NIKdependent pathway.

MATERIALS AND METHODS
Two-hybrid Screening-The two-hybrid screening was conducted using the Matchmaker system (CLONTECH) according to the manufacturer's instructions. Briefly, yeast strain HF7c, expressing a GAL4-DR4 fusion protein, was transformed with a human peripheral blood leukocyte cDNA library cloned into the pGAD10 vector (CLONTECH). 2 ϫ 10 6 clones were analyzed. Transformed yeast were selected on SD/agar plates lacking leucine, tryptophan, and histidine for 5 days at 30°C. Selected colonies were blotted onto filter paper, permeabilized in nitrogen liquid, and placed on another filter soaked in Z buffer (60 mM Na 2 HPO 4 , 40 mM NaH 2 PO 4 , 10 mM KCl, 1 mM MgSO 4 , 37.5 mM ␤-mercaptoethanol) containing 1 mM 5-bromo-4-chloro-3-indolyl-␤-D-galactoside. Colonies that developed color were re-streaked on selective plates to allow plasmid segregation and tested again for ␤-galactosidase ac-* The Basel Institute for Immunology was founded and is supported by F. Hoffmann-La Roche and Co. Ltd., CH-4005 Basel, Switzerland. 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.
¶ To whom correspondence should be addressed.

TABLE I
Interaction of c-E10 with DR4 and TNFR-1 in the yeast two-hybrid Yeast HF7c was transformed with clone T6 fused to GAL4-activating domain together with the indicated cDNAs fused to GAL4 DNA-binding domain. The interactions were examined by a filter assay for ␤-galactosidase activity; assays were done for five to ten independent transformants. Yeast colonies were scored as positive when a bright color developed in 2-5 h; a negative was scored when color failed to develop within 12 h. Quantitative liquid ␤-galactosidase assay was performed on individual colonies using o-nitrophenyl ␤-D-galactopyranoside as substrate. Data represent mean Ϯ S.D., n ϭ 3-5. tivity. The liquid ␤-galactosidase assay was performed according to the manufacturer's instructions using chlorophenol red-␤-D-galactopyranoside as a substrate. cDNA Cloning, Northern and Western Blot Analysis and Antibodies-The cDNA insert contained in clone T6 was used to screen a human spleen cDNA library (Stratagene). Nucleotide sequence of a phage clone, as well as the cDNA insert of clone T6, was determined on both strands using a Sequenase kit (Amersham Pharmacia Biotech). The cDNA insert contained in clone T6 was also used to probe a human multiple tissue Northern blot (CLONTECH) according to the manufacturer's protocol.
The polypeptides RTSSRKRAGKLLDYLQE and FPDGATNN-LSRSNSDESNFSEK were used to generate rabbit polyclonal antisera. The whole rabbit serum was affinity purified on CNBr-activated Sepharose beads (Amersham Pharmacia Biotech) coupled with the antigenic peptide. Proteic lysates from human tissues were obtained from CLONTECH.
Recombinant histidine-tagged proteins were made in Escherichia Human adult tissue Northern blots were probed with the coding fragment of c-E10 cDNA. pbl, peripheral blood leukocytes; small int., small intestine. Signal is clearly detected in the lane containing mRNA from ovary following a longer exposure. B, predicted amino acid sequence of c-E10. The polypeptide encoded by clone T6 is underlined. C, sequence alignment of CARD sequences present in c-E10 and v-E10. Double dot indicates identical residues, single dot represents conservative changes. Alignments were done with the Align program. D, immunoblot analysis of c-E10 protein. 10 g of total lysate from the indicated tissue resolved by SDS-PAGE and blotted on nitrocellulose membrane. The immunoblots were hybridized with affinity purified antisera raised against two polypeptides contained in c-E10 and described under "Materials and Methods." coli BL21 strain using the pET expression system (Novagen) and purified with nickel-nitrilotriacetic acid-agarose beads (Qiagen).
Cell Culture and Transfection-293 cells were cultured in Dulbecco's modified Eagle's medium, 10% fetal calf serum and were transfected by calcium phosphate method. HeLa Tet-On (CLONTECH) clones expressing c-E10 and T6 were established and maintained according to the manufacturer's instructions. Tet-On clones and HeLa cells were transfected with LipofectAMINE (Life Technologies, Inc.). Doxycycline and cycloheximide were obtained from Sigma.
In Vitro Transcription/Translation-The cDNA encoding for v-E10 was transcribed and translated in vitro in the presence of L-[4,5-3 H]leucine with the TNT-coupled lysate system (Promega) following the manufacturer's instructions.
In Vitro Binding and Coprecipitation-293 cells were transfected for 7 h in 6-well plates with 5 g of indicated plasmid DNA by calcium phosphate precipitation. 16 -24 h after transfection, cells from each well were lysed in 200 l of lysis buffer (150 mM NaCl, 50 mM Tris, 1 mM EGTA, 0.5% Nonidet P-40, and a mixture of protease inhibitors). For histidine-tagged protein binding, 50 l of 293 lysate were mixed with 20 l of the indicated agarose-bound histidine-tagged protein in 1 ml of modified E1A buffer (50 mM Hepes, pH 7.6, 250 mM NaCl, 10% glycerol, 0.1% Nonidet P-40). Samples were incubated for 1-2 h at 4°C, washed 8 times by pulse centrifugation with 1 ml of modified E1A buffer, and resuspended in 40 l of sample buffer. 5 l of the reaction were loaded for SDS-polyacrylamide gel electrophoresis and Western blot analysis.
Cell Death and Luciferase Assay-HeLa cells were transfected with 2 g of the indicated cDNAs together with 0.2 g of CMV ␤-galactosidase (CLONTECH). 24 h after transfection, ␤-galactosidase activity was visualized by fixing the cells in 0.2% glutaraldehyde for 10 min followed by staining in phosphate-buffered saline containing 20 mM each K 3 Fe(CN) 6 and K 4 Fe(CN) 6 -H 2 O and 1 mg/ml 5-bromo-4-chloro-3-indolyl-␤-D-galactoside for 1-3 h at 37°C. The number of live blue cells and blue cells with apoptotic morphology were counted in at least four fields. Cell viability of HeLa Tet-On clones was determined using CellTiter 96 (Promega). To assess NF-B activation, HeLa Tet-On clones and HeLa cells were transfected with 2 g of the indicated cDNAs together with pNF-B-luc (CLONTECH) in 12-well plates. Tet-inducible clones were treated with 1 mM doxycycline for 24 h or left repressed. Cells were then lysed, and luciferase activity was determined with luciferase assay system (Promega). A plasmid expressing ␤-galactosidase was added to the transfection mixture for normalization of the efficiency of transfection.

RESULTS
To identify novel proteins involved in death receptor signaling, a yeast two-hybrid screen was performed with the intracellular portion of DR4 fused to the DNA binding domain of GAL4. A plasmid library of fusions between the GAL4 transcription activation domain and cDNAs from peripheral blood leukocytes was screened for interaction in the yeast reporter strain HF7c. One library clone (T6), containing an open reading frame coding for a 119-amino acid polypeptide, was identified that specifically interacted with DR4 and with TNFR-1, but not with related family members (Table I).
The tissue distribution of clone T6 was examined by Northern blot analysis, and a single ϳ3-kilobase transcript was detected in all tissues analyzed (Fig. 1A). The full-length cDNA, encoding for a 233-amino acidic residue protein, was isolated from a spleen cDNA library (Fig. 1B). A FASTA search showed that this protein shares significant homology with E10, a   FIG. 2. Interaction of c-E10 with TRADD. A, 293 cells were transfected with full-length or truncated versions of c-E10 provided with an hemagglutinin epitope together with the indicated cDNA expressed in pCMV-FLAG. After transfection, samples were immunoprecipitated with an ␣-hemagglutinin (HA) antibody and analyzed for coprecipitating proteins by Western blot with ␣-FLAG. B, v-E10 binds to c-E10. Recombinant histidine-tagged v-E10 was purified with nickel-nitrilotriacetic acid-agarose beads and mixed with lysates from 293 cells transfected with plasmids expressing c-E10 or control proteins. After washing, agarose beads were boiled in SDS-sample buffer, separated by SDS-PAGE, and transferred on nitrocellulose membrane subsequently probed with ␣-FLAG monoclonal antibody. A fraction of the reaction mixtures was stained with Coomassie to visualize recombinant v-E10 protein. C, 35 S-labeled v-E10 was rotated with agarose-immobilized c-E10 in E1A buffer. Beads were washed, boiled in SDS-sample buffer, resolved by SDS-PAGE, and autoradiographed. Agarose-immobilized TRAIL served as control for a specific binding. A fraction of the reaction mixtures was stained with Coomassie and aligned to visualize histidine-tagged proteins. equine herpesvirus 2 protein of unknown function (19), therefore we named it cellular-E10 (c-E10) and refer to the viral protein as v-E10. v-E10 contains an amino-terminal CARD motif (20), which is conserved in c-E10 (Fig. 1C). The CARD motif was originally described as a conserved sequence present in some caspases and in several proteins involved in regulation of apoptosis (20). The recently determined solution structure of the CARD domain (21) revealed that it is arranged in a topology similar to the death domain (22) and, in fact, it contains sequence motifs suggestive of the death domain (20). The homology between c-E10 and RAIDD, another CARD-containing protein (23), is also significant.
Two rabbit antisera were raised against polypeptides contained in the c-E10 amino acidic sequence. Immunoblot analysis performed on cell lysates from several human tissues indicates that most of endogenous c-E10 migrates with an apparent molecular mass of 48 kDa, and a fraction of the protein migrates with a molecular mass of 38 kDa (Fig. 1D). Both signals of 38 and 48 kDa were specific because the immunoreactivity was abolished by competition with the corresponding immunogenic peptide (data not shown).
Next, c-E10 interactions were analyzed by cotransfection experiments. As shown in Fig. 2A, whereas the association of c-E10 with DR4 is barely detectable in mammalian cells, fulllength c-E10, but not the truncated proteins encoded by T6 and c-E10 1-127, clearly coprecititates with TRADD. In addition, c-E10 oligomerizes via an omophilic association of its CARD region. Conversely, TNFR-1 and proteins involved in TNFR-1 signaling, including TRAF-1, -2, -6, and RIP, did not associate with c-E10 in vivo. The discrepancy of this biochemical data with the two-hybrid result (Table I) may be because of the different sensitivity of the experimental systems. We also examined the interaction of several tagged proteins with immobilized v-E10. Full-length c-E10, but not clone T6, binds to v-E10, suggesting that the CARD regions are required for this interaction (Fig. 2B). Likewise, 35 S-labeled v-E10 binds to c-E10 but not to clone T6 (Fig. 2C).
Given the involvement of CARD-containing molecules in the induction of apoptosis in mammalian cells, we first tested whether overexpression of c-E10 is by itself sufficient in triggering the cell death pathway. To this end, HeLa cells were transiently transfected with c-E10 and scored for apoptotic morphology. As shown in Fig. 3A, overexpression of c-E10 only weakly induces cell death, as compared with overexpression of  , n ϭ 4). B, expression of c-E10 under the control of a doxycycline-dependent transcriptional transactivator. cDNAs encoding c-E10 and T6 were tagged with a FLAG epitope and transfected in HeLa Tet-On cells. Isolated clones were left untreated or cultured in medium containing 1 mM doxycycline. After 24 h, 2 g of total cell lysate were separated by SDS-PAGE, and expression of the transfected cDNAs was analyzed by Western blot using ␣-FLAG monoclonal antibody. C, expression of c-E10 does not interfere with TNF-induced cell death. Cells were cultured in 96-well plates (1 ϫ 10 4 cells/well) in regular medium or medium containing 1 mM doxycycline for 24 h. TNF was added at the indicated final concentration together with cycloheximide (0.8 g/ml) to induce apoptosis. 16 h later, cell viability was measured by a colorimetric method. FADD or RIP, which both result in massive cell death. To better characterize the function of c-E10 in the cell death cascade, we established HeLa cell clones expressing either c-E10 or T6 cDNA under the control of a doxycycline-dependent transcriptional transactivator (24). Clones showing the highest induction with the lowest background expression were chosen for experimental analysis (Fig. 3B). In these HeLa clones, the induction of either c-E10 or T6 expression does not result in cell death ( Fig. 3C and data not shown). Also, as shown in Fig. 3C, expression of neither c-E10 nor T6 polypeptide has a marked effect on apoptosis induced by TNF; similar results were obtained when cell death was evoked by exposure to Fas or TRAIL (data not shown). These results suggest that the efficacy of c-E10 in inducing apoptosis might depend on the level of expression of the protein.
To examine a possible role for c-E10 in the activation of the transcription factor NF-B, the inducible clones were transfected with a NF-B-dependent luciferase reporter gene, and the cells were left repressed or treated with doxycycline to induce c-E10 expression. As shown in Fig. 4, induction of c-E10 expression results in NF-B activation. In this experimental system, the polypeptide encoded by clone T6 only partially triggers that response. Likewise, activation of NF-B was readily observed in 293 and Hela cells transiently transfected with c-E10 (Fig. 4B). Expression of c-E10 1-127, which contains the CARD, resulted in a similar or even higher activating response. This result, in light of the nonassociation of c-E10 1-127 with TRADD ( Fig. 2A), suggests that c-E10 acts downstream TRADD and that the CARD domain is itself able to recruit downstream effectors leading to NF-B activation. Thus, TRADD/c-E10 interaction explains only in part the activation of NF-B mediated by c-E10.
Because NF-B activation by death receptors is in part mediated by the TRAF/NIK pathway, we tested whether dominant negative forms of these proteins inhibited c-E10-dependent NF-B activation (Fig. 4). As expected, a dominant negative form of IB efficiently prevents NF-B activation when coexpressed together with c-E10. Dominant negative versions of c-E10 Activates NF-B TRAF2, but not TRAF6, had the same inhibitory effect. Furthermore, NF-B activity by c-E10 is strongly inhibited by NIK DN and only partially by ⌬MEKK K-M, the dominant negative form of the other immediate upstream activatory kinase of the IKK complex (25). Therefore, we concluded that activation of NF-B induced by c-E10 is TRAF2/NIK-dependent. DISCUSSION The capacity to activate the transcription factor NF-B is shared by a number of receptors related to TNFR-1, such as Wsl/DR3/Apo-3/TRAMP/LARD (26 -28), the two TRAIL receptors DR4 and DR5 (29,30), and Fas/Apo1 (31,32). The pleiotropic activities of these receptors are modulated by a regulated recruitment of a number of cytoplasmic signaling proteins to the receptor complex. Activation of NF-B is mediated by association of TRADD with the serine/threonine kinase RIP and TRAF-2 (15,16). Transfection experiments have in fact shown that overexpression of TRAF-2 is in itself sufficient to induce NF-B activation (33), and TNFR-1-induced NF-B activation is reduced and delayed in TRAF-2-deficient mice (34).
In this study, we present a novel CARD-containing protein that, upon overexpression, induces activation of NF-B via a TRAF-2/NIK-dependent pathway. TRADD-TRAF-2 interactions are also involved in the activation of JNK (9,17,35); overexpression of c-E10, however, does not result in phosphorylation of JNK (data not shown).
Other CARD-containing proteins have been shown to be involved in the activation of NF-B: RIP2/CARDIAK/RICK, a CARD-containing protein kinase related to RIP, also signals NF-B activation (36,38,39). However, whereas RIP2 interacts with specific members of the TRAF family, c-E10 does not directly recruit TRAFs protein, and its association with TRADD suggests that c-E10 functions upstream of TRAF-2. Despite the fact that a number of proteins containing the CARD motif operate in apoptotic pathways by modulating caspase activity (36,37), our results indicate that c-E10 only weakly triggers the cell death cascade and, although cloned for this ability to interact with DR4, it does not interfere with FAS, DR4, and TNFR1 induced apoptosis. Thus, the identification of c-E10 provides additional evidence that CARDcontaining proteins not only modulate apoptosis by regulating caspase activity but may also be critical mediators of NF-B activation response.
While this work was in preparation, Willis et al. (40) reported that c-E10, which they named bcl10, is often mutated in MALT lymphomas and in a number of other tumor types, suggesting that this protein may be involved in the pathogenesis of several human malignancies. Particularly, they found that the most recurrent mutations in c-E10 were nucleotidic insertions or deletions distal to the CARD, which resulted in truncated c-E10 protein lacking the carboxyl-terminal domain but still effective in activating NF-B. In our experiments, the carboxylterminal domain of c-E10 is responsible for the binding to DR4 in yeast. The mechanism involving apoptosis induced by DR4 and its ligand TRAIL has recently raised a great deal of attention, because whereas normal cells are resistant, tumoral cells are sensitive to TRAIL-induced cell death (41)(42)(43)(44). Therefore, these molecules are extremely interesting for their potential application in cancer therapy and, in fact, repeated treatments with TRAIL effectively suppressed tumor growth in vivo without detectable damage for normal tissues (45). Thus, the participation of c-E10 in signaling pathways shared by death receptors offers an attractive perspective in explaining how mutations in c-E10 result in cellular transformation.