Death Effector Domain-Only Polypeptides of Caspase-8 and -10 Specifically Inhibit Death Receptor-Induced Cell Death

Biochemical and Biophysical Research Communications 291, 484 – 493 (2002) doi:10.1006/bbrc.2002.6482, available online at http://www.idealibrary.com on

Death Effector Domain-Only Polypeptides of Caspase-8 and -10 Specifically Inhibit Death Receptor-Induced Cell Death Yoshiaki Shikama,* Lisong Shen,* ,1 Motokuni Yonetani,* Jun Miyauchi,† Toshiyuki Miyashita,* ,2 and Masao Yamada* *Department of Genetics, National Children’s Medical Research Center, Tokyo, Japan; and †Division of Pathology, Department of Clinical Laboratory, National Children’s Hospital, 3-35-31 Taishido, Setagayaku, Tokyo 154-8509, Japan

Received January 18, 2002

Caspase-8 and -10 are thought to be involved in a signaling pathway leading to death receptor-mediated apoptosis. The prodomains of these caspases are known to form fibrous structures in the perinuclear region when overexpressed, though the meaning of the structures remains unclear. In a previous study we showed that the overexpressed caspase-8 or -10 prodomain (PDCasp8 or PDCasp10) did not induce cell death, and we hypothesized that these prodomains interfere with the receptor-mediated cell death signaling pathway. Indeed, in 293, HeLa and Jurkat cells, cell death mediated by agonistic anti-Fas antibody, TRAIL or overexpression of full-length caspase-8 was significantly inhibited by overexpression of PDCasp8 or PDCasp10 which colocalized with the Golgi complex and with overexpressed FADD. However, when about 20 amino acid residues were deleted from either terminus of the caspase-10 prodomain (amino acid residue 1 to 219), the ability to inhibit Fas-mediated cell death was lost. Interestingly, these deletion mutants also lost the ability to make fibrous structures and to bind FADD, suggesting that FADD binding is important for their function, and that PDCasp8 and PDCasp10 act as dominant-negative inhibitors. © 2002 Elsevier Science (USA)

Abbreviations used: DED, death effector domain; DEF, death effector filament; DISC, death inducing signaling complex; DOX, doxycycline; FADD, Fas-associated death domain-containing protein; GFP, Green fluorescent protein; GM130, Golgi matrix protein of 130 kDa; IAP, inhibitor of apoptosis; PCR, polymerase chain reaction; PDCasp8, prodomain of caspase-8; PDCasp10, prodomain of caspase10; PS, phosphatidylserine; STS, staurosporine; TNF, tumor necrosis factor; TRAF, TNF receptor-associated factor; TRAIL, TNF-related apoptosis-inducing ligand; zVAD-fmk, carbobenzoxy-VAD-fluoromethyl ketone. 1 Present address: Laboratory Diagnostic Center, Shanghai Children’s Medical Center, Shanghai Second Medical University, Shanghai 200127, China. 2 To whom correspondence and reprint requests should be addressed. Fax: ⫹81-3-3414-3208. E-mail: [email protected]. 0006-291X/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.

Key Words: apoptosis; caspase; Fas/Apo-1; death effector domain; death effector filament.

Apoptosis is a type of cell death regulated by various genes that are remarkably conserved during evolution. It plays an important role in the development and the homeostasis of multicellular organisms. Therefore, defects in its regulation contribute to many major diseases including cancer. Apoptotic signals are mainly transduced through two relatively distinct pathways (1); the so-called “intrinsic” pathway mediated by mitochondria from which proapoptotic molecules such as cytochrome c are released into the cytosol, and an “extrinsic” pathway triggered by members of the TNF receptor superfamily such as Fas (Apo-1/CD95) or TNF-␣ receptor. Upon stimulation by the Fas ligand, for example, an adaptor molecule, FADD, is recruited to the Fas receptor through interaction between the death domains of both molecules. FADD then recruits caspase-8 (MACH/FLICE/Mch5) (2– 4) and caspase-10 (FLICE2/Mch4) (4, 5), members of apoptosis-inducing cysteine proteases (caspases), through homophilic interaction between the DEDs (6). Thus, a complex called DISC is formed (7) and the aggregation of caspase-8 or -10 in DISC triggers a cascade of caspase activation leading to the cleavage of cytosolic, cytoskeletal and nuclear proteins and chromosomal DNA. The biological importance of caspase-8 and -10 is underscored by the analysis of CASP8-deficient mice and the mutations of the CASP10 gene observed in patients with autoimmune lymphoproliferative syndrome (8, 9). In addition, caspase-8 acts as a tumor suppressor since the gene is silenced or mutated in some neuroblastomas and head and neck carcinomas (10 –12). Caspases are synthesized as inactive zymogens (procaspases) comprising an N-terminal prodomain and a catalytic protease domain that is further processed to a

484

Vol. 291, No. 3, 2002

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 1. Subcellular localization of prodomain-only polypeptides of caspase-8 and -10. HeLa cells were transiently transfected with EGFP-PDCasp8 (H–K, P–S) or EGFP-PDCasp10 (A–G, L–O). In D–G and H–K, cells were cotransfected with EGFP-PDCasp10 or -PDCasp8, respectively, along with HA-FADD. At 48 h after transfection, cells were stained with anti-␣-tubulin (B), anti-HA (E, I), or anti-GM130 (M, Q) antibody. Fluorescent as well as transmitted images (F, J, N, R) were taken using a confocal laser microscope. Merged images are shown in C, G, K, O, and S.

large and a small subunit upon exposure to apoptotic stimuli. Caspase-8 and -10 have a long N-terminal prodomain, each of which has two copies of the DED

that mediate protein-protein interactions as described above. In addition to the isoforms that contain the entire set of these components, prodomain-only iso-

485

Vol. 291, No. 3, 2002

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

forms of caspase-8 and -10 that lack protease activity have been reported (2, 13). Previously, we and others have reported that DED-only peptides, when overexpressed in cultured cell lines, form perinuclear filamentous structures called DEFs, although the biological significance of these isoforms remains controversial (14, 15). In this study, we describe the colocalization of DEFs with the Golgi complex. Moreover, our results reveal that DED-only polypeptides specifically inhibit death receptor-mediated cell death, suggesting a dominant negative function in extrinsic apoptosis pathway. MATERIALS AND METHODS Plasmids. The DNA fragments encoding various deleted forms of human PDCasp10 were amplified by PCR with a plasmid encoding a full-length PDCasp10, and then subcloned into pEGFP-C2 (Clontech) to generate fusion proteins with GFP. The other plasmids encoding EGFP-caspase fusion proteins were prepared as described previously (15). To generate tetracycline-regulated expression plasmids, pTet-PDC8 and pTet-PDC10, cDNAs encoding the entire coding regions of EGFP-PDCasp8 and EGFP-PDCasp10 obtained by PCR were subcloned into pTet-Splice (GIBCO-BRL). cDNA covering the coding region of mouse FADD was obtained by reverse transcription-PCR with primers designed from the reported sequence. It was then subcloned into pcDNA-3XHA (kindly provided by Dr. John C. Reed) to generate HA-FADD fusion protein. The authenticity of all constructs was confirmed by DNA sequencing. Cell cultures and transfections. The human embryonic kidney cell line 293 and human cervical cancer cell line HeLa were maintained in DMEM supplemented with 10% fetal calf serum, 50 U/ml of penicillin and 0.1 mg/ml of streptomycin at 37°C in a humidified atmosphere of 5% CO 2. Cells were transfected with the plasmids described above using Effectene reagent (Qiagen) according to the manufacturer’s protocol. A pan-caspase inhibitor, zVAD-fmk (Peptide Institute, Osaka, Japan), was added in some experiments to a final concentration of 50 ␮M. Jurkat Tet-On cell line which expresses the reverse tetracycline-controlled transactivator (16) was obtained from Clontech and maintained in RPMI 1640 medium supplemented with 10% fetal calf serum, 100 ␮g/ml of G418, 50 U/ml of penicillin and 0.1 mg/ml of streptomycin. Jurkat Tet-On cells were transfected with either pTet-PDC8 or pTet-PDC10 together with pPGK-Hyg using Effectene reagent (Qiagen) according to the manufacturer’s protocol. Transfected cells were allowed to recover for 48 h before the addition of 360 ␮g/ml of hygromycin and the selection of resistant clones by limiting dilution in 96-well microtiter plates. The resistant clones were subjected to immunoblotting to evaluate the expression of introduced genes. Antibodies. The following antibodies were used for immunoblotting and immunofluorescence: rabbit anti-GFP polyclonal antibody (MBL, Nagoya, Japan), rat anti-HA monoclonal antibody (clone 12CA5) (Roche), mouse anti-GM130 monoclonal antibody (clone 35) (BD Biosciences), mouse anti-␣-tubulin monoclonal antibody (clone DM 1A) (Sigma), mouse anti-caspase-8 antibody (clone 5F7) (MBL), and rabbit anti-caspase-3 polyclonal antibody (kindly provided by Dr. John C. Reed). Immunostaining and confocal microscopy. At 24 h posttransfection, HeLa cells growing on Lab-Tek II chamber slides (Nalge Nunc International) were fixed with 4% paraformaldehyde at 4°C for 1 h, permeabilized with 0.1% Triton X-100 for 1 h, and then blocked for 1 h with preblock solution (5% skim milk and 1% bovine serum

FIG. 2. Comparison of cell loss mediated by caspase-8 or -10 and their prodomains. 293 cells were cotransfected with 0.9 ␮g each of the indicated plasmids in combination with 0.3 ␮g of luciferase plasmid. A plasmid encoding GFP non-fusion protein (GFP) was used to adjust the total amount of DNA. Cells were harvested 48 h after transfection and luciferase activity was measured. In this assay, luciferase activity represents the relative number of living cells remaining in the culture after the transfection. Data are presented as the mean ⫾ SD obtained from three independent experiments done in triplicate.

albumin in 10 mM Tris–HCl (pH 8.0), 150 mM NaCl and 0.1% Tween 20) at room temperature. The slides were next incubated for 1 h with the antibody against ␣-tubulin, GM130 or HA, and then with tetramethylrhodamine isothiocyanate-labeled anti-mouse immunoglobulin (DAKO). A fluorescent image was obtained using a confocal microscope (Fluoview FV300, Olympus, Tokyo, Japan). Western blotting and coimmunoprecipitation assay. Immunoprecipitations and Western blot analyses were performed as described previously (15, 17). Briefly, 200 ␮g of cell lysate was immunoadsorbed with the antibody against GFP for 6 h at 4°C and then with protein A/G agarose (Santa Cruz Biotechnology) for 16 h at 4°C. The immunoprecipitated pellets were washed four times with lysis buffer, and finally resuspended in SDS sample buffer. Bound proteins were eluted by boiling before SDS–polyacrylamide gel electrophoresis. The resolved proteins were transferred to nitrocellulose membranes, incubated with the antibodies described above, and visualized by using enhanced chemiluminescence (Amersham). Luciferase assay. 293 or HeLa cells growing on 6-well culture plates were cotransfected using Effectene reagent with indicated plasmids and pGV-C2 (Wako Chemicals, Osaka, Japan) which encodes luciferase. The amounts of the plasmids used are indicated in the figure legends. For the assay of Fas-induced cell death, 100 ng/ml of agonistic anti-Fas antibody (clone CH11) (MBL, Nagoya, Japan) together with 5 ␮g/ml of cycloheximide was added to each well 24 h after transfection. After another 24 h, the cells were washed once with PBS to remove dead and floating cells. They were then harvested and subjected to a luciferase assay as described previously (15). To quantify TRAIL-induced cell death, various concentrations of recombinant human TRAIL (R&D Systems) together with 1 ␮g/ml of monoclonal anti-6⫻ histidine cross-linking antibody (R&D Systems) were added at 24 h posttransfection. After 5 h of incubation, the cells were harvested and subjected to the luciferase assay. Flow cytometry. Jurkat clones were harvested after the culture with various combinations of DOX, anti-Fas, STS and recombinant TRAIL. Cells were washed twice with cold PBS and stained with phycoerythrin-conjugated annexin V (Pharmingen) according to the manufacturer’s protocol. Flow cytometric analysis was performed using a FACSort (Becton–Dickinson).

486

Vol. 291, No. 3, 2002

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 3. Interaction of PDCasp10 deletion mutants with FADD and inhibition of Fas-mediated cell death. (A) Expression plasmids encoding various deletion mutants are depicted. Numbers refer to amino acid positions. The regions homologous to the DEDs are indicated by shaded boxes. Boundaries of the DEDs are based on Ref. 4. The percentage of fiber-positive cells among GFP-positive cells derived from three separate experiments was determined 24 h after the transfection of HeLa cells and is indicated at the right. At least 200 cells were counted under a fluorescent microscope for each transfection. (B) 293 cells were cotransfected with various combinations of the plasmids as indicated at the top in the presence of zVAD-fmk. Cell lysates were obtained 48 h after the transfection. The coimmunoprecipitated FADD was detected by immunoblotting with anti-HA antibody (top). Total lysates were blotted with anti-HA (middle) or anti-GFP antibody (bottom). (C) 293 cells were transfected with 1.5 ␮g of the indicated constructs in combination with 0.3 ␮g of luciferase plasmid. At 24 h after the transfection, cells were treated with 100 ng/ml of anti-Fas antibody (CH11) and 5␮g/ml of cycloheximide. Cells were harvested 24 h after the treatment and luciferase activity was measured. (D) 293 cells were transfected and treated as in C. At various time points after the treatment with anti-Fas antibody, cells were harvested and luciferase activity was measured. Data are presented as the mean ⫾ SD obtained from three independent experiments. *P ⬍ 0.05, **P ⬍ 0.01, vs GFP.

RESULTS DEFs colocalize with FADD and the Golgi complex. To date, DEFs formed by PDCasp8 and PDCasp10 have not been reported to colocalize with any known proteins or cytoskeletal elements, including mitochondria, actin, tubulin and vimentin (Refs. 13, 14, and Figs. 1A–1C). Since an adaptor protein FADD binds to caspase-8 and -10 through its DED domain, we investigated if FADD is recruited to DEFs. In contrast to a previous report (14), FADD itself did not form filamentous structures at least not in our experimental conditions using HeLa cells but was distributed evenly both in the nucleus and in the cytoplasm (Fig. 1E, arrow). However, when cotransfected with PDCasp8 or PDCasp10, FADD was significantly recruited to DEFs, although a diffuse intracellular distribution was still observed (Figs. 1D–1K). A juxtanuclear pattern of prodomain distribution is reminiscent of the Golgi distribution. Indeed, PDCasp8 and PDCasp10 colocalized

with GM130, a cis-Golgi network marker (18), in HeLa cells (Figs. 1L–1S). An identical pattern of colocalization was observed in several cell types, including 293, SHSY5Y (human neuroblastoma cells) and HepG2 (human hepatoma cells) (data not shown). Caspase-8-mediated cell death is inhibited by overexpression of PDCasp8 or PDCasp10. We previously demonstrated that, in contrast to wild-type caspase-8 and -10, PDCasp8 and PDCasp10 did not induce cell death when transiently overexpressed (15). To further explore the function of these prodomain-only isoforms, we next addressed the question of whether PDCasp8 or PDCasp10 can inhibit cell death mediated by overexpression of the wild-type in a dominant negative fashion. As shown in Fig. 2, a marked cell loss was observed when 293 cells were transfected with wild-type caspase-8 (Casp8) or -10 (Casp10), whereas neither the protease-dead mutants (C/S) nor prodomain-only polypeptides (PDCasp8, PDCasp10) demonstrated a

487

Vol. 291, No. 3, 2002

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 4. Induced expression of the prodomains of caspase-8 and -10 in Jurkat cells. (A) Jurkat clones were cultured in the presence (lanes 2, 4, 6, 8) or absence (lanes 1, 3, 5, 7) of 150 ng/ml of DOX. Cell lysates were obtained after 2 days of culture and subjected to Western blotting with anti-GFP antibody. B–E, Jurkat clones, PD8-2 (B and C) and PD10-2 (D and E), were cultured in the presence of DOX for 2 days and subjected to confocal microscopy. The expression of GFP-prodomain caspase-8 (B) or -10 (D) and the merged images of GFP-fusions and transmitted micrographs (C and E) are shown. F, Jurkat clones were seeded at a density of 2.5 ⫻ 10 5/ml and grown in the presence or absence of DOX (150 ng/ml). Cell densities were measured at various time points.

Vol. 291, No. 3, 2002

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

decrease in luciferase activity. When an equal amount of plasmid encoding PDCasp8 or PDCasp10 was cotransfected with wild-type caspase-8, the cell deathinducing activity of caspase-8 was markedly inhibited. Unexpectedly, that of caspase-10 was not inhibited by the expression of the prodomain of caspase-8 or -10. Even when cells were transfected with an increased amount of the PDCasp8 or PDCasp10 construct, cell death-inducing activity of caspase-10 was not inhibited (data not shown), suggesting that there is a distinct pathway to apoptosis triggered by overexpression of caspase-10. Both modules of DED are required for formation of DEF, interaction with FADD, and inhibition of Fasmediated cell death. The prodomain of caspase-10 is composed of 219 amino acids comprising two copies of DED. We next prepared a series of deletion mutants as shown in Fig. 3A, each of which was fused with EGFP at its N-terminus. The expression of proper-sized GFP fusion proteins was confirmed by immunoblotting (Fig. 3B, bottom). First, the percentage of fiber-positive cells among transfected cells was determined under a fluorescent microscope (Fig. 3A). When 20 or less amino acid residues were deleted from either terminus of PDCasp10, the mutants retained the ability to make fibrous structures. In contrast, mutants with 30 or more amino acid deletions at either terminus did not make fibrous structures. Next we examined the ability of these mutants to interact with FADD by immunoprecipitation. In addition to full-length PDCasp8 and 10, only the mutants that could make fibrous structures, PDCasp10(20 –219) and PDCasp10(1–202), interacted with FADD (Fig. 3B, top). A primary role of caspase-8 and -10 is to transduce death receptor-mediated cell death signals. Therefore, we examined whether PDCasp8, PDCasp10 and deletion mutants thereof were able to inhibit Fas-mediated cell death (Fig. 3C). First of all, as reported (2, 5), protease-dead mutants of these caspases (Casp8(C/S), Casp10(C/S)) inhibited Fas-mediated cell death, demonstrating the specificity of the experiments. As anticipated, PDCasp8 and PDCasp10 also significantly inhibited cell death, which was evident after 16 h of anti-Fas treatment (Fig. 3D). In addition, as in a previous experiment, only fiber-forming mutants interfered with the cell death induced by agonistic anti-Fas antibody. Taken together, both of the DED copies in PDCasp10 were required for the formation of DEF, interaction with FADD, and inhibition of caspase-8mediated cell death. Inducible expression system of DED peptides in Jurkat cells. Since Fas signaling plays a critical role in lymphocyte homeostasis, we next employed the human T-cell line Jurkat known to be sensitive to anti-Fas treatment. Since we initially assumed that the continuous expression of the prodomain would be toxic to

cells, we utilized a tetracycline-controlled expression system. Jurkat Tet-On cells were stably transfected with the expression plasmid that encodes GFP-fused PDCasp8 or PDCasp10. Among more than 40 clones obtained, we selected two clones for further study that express comparative levels of PDCasp-8 or PDCasp-10 upon treatment with DOX, a tetracycline derivative, with little leakage of expression in the absence of DOX (PD8-2 and PD10-2, Fig. 4A, lanes 3, 4, 7, 8). Two more clones that did not express introduced genes were used as negative controls (PD8-1 and PD10-1, Fig. 4A, lanes 1, 2, 5, 6). As in 293 or HeLa cells, these prodomains also formed DEFs in Jurkat cells (Figs. 4B– 4E, arrows). However, they were not as remarkable as those in 293 or HeLa cells and observed in only a fraction of fluorescence-positive cells. In most of the cells, fluorescence signals were homogeneously distributed throughout the cell (Figs. 4D and 4E, arrowheads). As shown in Fig. 4F, the expression of these prodomains had no effect on the proliferation of Jurkat clones. As discussed later, not all PD8-2 and PD10-2 cells expressed the introduced gene after the addition of DOX. However, fluorescence-positive populations neither increased nor decreased during a long-term culture of these clones for at least 20 days, suggesting that the expression of DED prodomains does not result in a growth advantage or disadvantage (data not shown). DED peptides specifically inhibit Fas-mediated apoptosis in Jurkat cells. Using the established Jurkat clones described above, we next explored if prodomainonly polypeptides of caspase-8 or -10 also inhibit Fasmediated cell death in Jurkat cells. As shown in Fig. 5A, not all cells expressed the introduced gene upon treatment with DOX, which is frequently observed in other inducible expression systems. Therefore, in this experiment, we gated GFP-positive and GFP-negative populations, and analyzed exposure of PS as an early event in apoptosis by using annexin V, a PS-binding protein. As shown in Fig. 5A, the treatment with the anti-Fas antibody increased the number of annexin V-positive cells in the GFP-negative population (from 11 to 31%), but not in the GFP-positive population (from 8 to 10%). This indicates that the expression of the caspase-8 prodomain protects Jurkat cells from Fas-mediated apoptosis. The expression of PDCasp8 induced by DOX had no effect on the externalization of PS (compare the top two rows in Fig. 5A), again arguing against a proapoptotic function of PDCasp8. In contrast, the treatment with STS, a protein kinase inhibitor, increased annexin V positivity both in GFPpositive and -negative populations, suggesting that the antiapoptotic effect of these prodomains is specific to the Fas-mediated pathway. Similar results were obtained in PD10-2 clone that inducibly expressed PDCasp10 (Fig. 5B). In contrast to 293 cells (Fig. 3D), Jurkat cells eventually underwent apoptosis after 24 h

489

Vol. 291, No. 3, 2002

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 5. Flow cytometric analysis of Jurkat clones. (A) Clone PD8-2 was cultured in the presence or absence of 150 ng/ml of DOX for 24 h. Agonistic anti-Fas antibody, CH11, (␣-Fas, 150 ng/ml) or staurosporine (STS, 500 nM) was added to some of the cultures 6 or 16 h prior to the analysis, respectively. Cells were then harvested, stained with annexin V-PE, and subjected to flow cytometry. GFP-positive and -negative populations were gated as shown in the left panels and histograms of PE intensity were analyzed for each population (middle and right). The percentages of annexin V-positive cells are also indicated. (B) The percentages of annexin V-positive cells are summarized. The results obtained from the same experiment using 3 other clones are also included. GFP positive populations were virtually absent in clones PD8-1 and PD10-1. Data are presented as the mean ⫾ SD obtained from three independent experiments. *P ⬍ 0.05, **P ⬍ 0.01. (C) Time course of cell death assessed by annexin V positivity after the treatment of PD8-2 and PD10-2 cells with anti-Fas as described in A. (D) Activation of caspases in PD10-2. Clone PD10-2 was cultured in the presence or absence of 150 ng/ml of DOX for 24 h. Agonistic anti-Fas antibody (␣-Fas, 150 ng/ml) was added to some of the cultures 6 h prior to the analysis. Cell lysates were then obtained and subjected to SDS–PAGE followed by immunoblotting using antibodies against caspase-8 i) and caspase-3 (bottom).

of anti-Fas treatment regardless of the expression of PDCasp8 or PDCasp10 (Fig. 5C). This is probably due to the difference in cell type, experimental system, and the levels of expression. Caspase cascade upon anti-Fas treatment is diminished by DED peptides. Upon the stimulation of Jurkat cells with Fas ligand or anti-Fas agonistic antibodies, caspase-8 is first proteolytically activated in DISC leading to the activation of downstream caspases such as caspase-3. We therefore examined whether the expression of DED peptides has any effect on the activation of caspases. When PDCasp8 was not expressed, the amount of procaspase-8 or -3 decreased accompanied by the production of cleaved and activated caspase-8 and -3 after the addition of CH11 in culture (Fig. 5D, lanes 1 and 2). However, when the expression of PDCasp8 was induced by DOX, the proteolytic activation of both caspases was markedly inhibited (Fig.

5D, lanes 3 and 4). This implies that the inhibition of Fas-mediated apoptosis by the prodomain-only peptide of caspase-8 is an upstream event of caspase activation. PDCasp-8 and PDCasp-10 also inhibit TRAILinduced apoptosis. Recently, another member of the TNF family, TRAIL or Apo-2L, was identified (19, 20). TRAIL interacts with its receptor, death receptor 4 (DR4) or TRAIL-R1 and DR5 (or TRAIL-R2) and initiates apoptosis mainly in transformed and malignant cells (21). To examine whether PDC8 and PDC10 can also inhibit TRAIL-induced apoptosis, we treated HeLa cells, which are more sensitive to TRAIL than 293 cells (data not shown), and Jurkat clones with human recombinant TRAIL and analyzed the extent of cell death by luciferase assay and flow cytometry, respectively, as described above. As shown in Fig. 6, PDCasp8 and PDCasp10 also protected HeLa and Jurkat cells from TRAIL-induced cell death.

490

Vol. 291, No. 3, 2002

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

FIG. 6. Inhibition of TRAIL-mediated cell death by PDCasp8 and PDCasp10. (A) HeLa cells were transfected with 1.5 ␮g of the indicated constructs with 0.3 ␮g of luciferase plasmid. A plasmid encoding GFP non-fusion protein (GFP) was used as a negative control. At 24 h after the transfection, cells were treated with indicated concentrations of human recombinant TRAIL and 1 ␮g/ml of monoclonal anti-6⫻ histidine cross-linking antibody. Cells were harvested 5 h after the treatment and luciferase activity was measured. Data are presented as the mean ⫾ SD obtained from three independent experiments performed in triplicate. (B) Jurkat clone PD10-2 was cultured in the presence or absence of 150 ng/ml of DOX for 24 h. Human recombinant TRAIL (20 ng/ml) along with anti-6⫻ histidine antibody (1 ␮g/ml) was added to some of the cultures 6 h prior to the analysis. Cells were then harvested, stained with annexin V-PE, and subjected to flow cytometry. Data are presented as in Fig. 5A.

491

Vol. 291, No. 3, 2002

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

DISCUSSION We previously demonstrated that PDCasp8 and PDCasp10 form filamentous structures termed as DEFs (15). In contrast to the previous results of others in which DEFs were proapoptotic (13, 14, 22), DEFs did not induce apoptosis in our experimental conditions when transiently overexpressed. Cells with DEDs had an intact nucleus and mitochondrial membrane potential as assessed by Hoechst 33342 and Mitotracker Red staining, respectively (15). It is likely, at least in part, that these conflicting results are due to the difference in cell lines and constructs used, and expression levels of transfected genes. Moreover, in this paper, we, for the first time, described that both PDCasp8 and PDCasp10 were localized to the Golgi complex and selectively inhibited anti-Fas- and TRAIL-mediated apoptosis in 293, HeLa and Jurkat cells. These contradictory results remind us of the function of FLIP (Casper/ I-FLICE/CASH/FLAME-1/MRIT/CLARP/usurpin), independently identified by several groups, which has two copies of the DED like PDCasp8 and PDCasp10 and is deficient in protease activity. When overexpressed, Casper is proapoptotic in some cell lines and experimental conditions, but not in others (23–28). In the latter cases, Casper protects cells from Fas- or TNFR-mediated apoptosis. For example, FLIP L, a longer isoform of FLIP, caused cell death only when high concentrations of the expression vector were transfected (25). In other reports, FLIP L inhibited TNFR-mediated cell death in HeLa cells. In 293-T cells, however, it resulted in marked cytotoxicity (28). At least four models can be proposed for the mechanism by which PDCasp8 and PDCasp10 inhibit apoptosis mediated by death receptors. Since prodomain-only DED polypeptides bind to FADD, they may be recruited to DISC upon stimulation with anti-Fas or TRAIL and then compete with full-length caspase-8 or -10 for the binding to FADD. The second possibility is that these polypeptides recruit FADD to DEFs leading to the interference in the translocation of FADD to DICS upon stimulation with anti-Fas. Third, caspase-8 and its homologs including caspase-10 activate NF-␬B through their DEDs (Y. Shikama, T. Miyashita, unpublished data and ref. 29, 30). NF-␬B suppresses apoptosis, at least in part, through the induction of genes for TRAF1, TRAF-2, c-IAP1 and c-IAP2 (31). According to this scenario, DED-only polypeptides that lack protease activity may switch function from proapoptotic to antiapoptotic. To our knowledge, the only caspase that is localized at least in part in the Golgi complex is caspase-2 (32). The biological significance of this localization, however, is unclear and contradictory results on its localization have also been reported (15, 33). Recently, death receptors such as Fas and TNFR1 were reported to be localized mainly in the Golgi complex (34, 35). Since

death receptors in the Golgi complex are not responsive to their ligands and have to translocate to the plasma membrane to trigger death signals, the fourth possibility, that PDCasp8 or PDCasp10 exerts its antiapoptotic function by sequestering Fas or TNFR1 in the Golgi complex through adaptor molecules such as FADD, is also plausible. All models fit well with our findings that the inhibition of apoptosis is an upstream event of caspase activation. In any case, since the ability of PDCasp8 or PDCasp10 to bind FADD well correlated with antiapoptotic function, the interaction with FADD seems to play an important role in inhibiting death receptor-mediated cell death. Both caspase-8 and -10 have isoforms that lack protease domains, and the expression of the caspase-10 isoforms is reported to be highly variable among normal tissues, as well as among various tumor cell lines (2, 13). Because our study showed that the expression of prodomain-only polypeptides of caspase-8 and -10 affected the sensitivity of death receptor-mediated apoptotic signals, the mRNA expression profiles of caspase-8 and -10 isoforms may have an effect on the regulation of cell death, and their imbalance may lead to tumorigenesis or immunological disorders. ACKNOWLEDGMENTS We are grateful to Dr. John C. Reed for providing pcDNA-3XHA and anti-caspase-3 antibody. We thank Yuko Ohtsuka and Atsuko Asaka for excellent technical assistance. We also thank Kayoko Saito for secretarial assistance. This study was supported in part by grants for Brain Research and for Genome Research from the Ministry of Health, Labour and Welfare, a Grant-in-Aid for Scientific Research and a Grant for Organized Research Combination System from the Ministry of Education, Culture, Sports, Science and Technology, Japan.

REFERENCES

492

1. Roy, S., and Nicholson, D. W. (2000). J. Exp. Med. 192, F21-F25. 2. Boldin, M. P., Goncharov, T. M., Goltsev, Y. V., and Wallach, D. (1996) Cell 85, 803– 815. 3. Muzio, M., Chinnaiyan, A. M., Kischkel, F. C., O’Rourke, K., Shevchenko, A., Ni, J., Scaffidi, C., Bretz, J. D., Zhang, M., Gentz, R., Mann, M., Krammer, P. H., Peter, M. E., and Dixit, V. M. (1996) Cell 85, 817– 827. 4. Fernandes-Alnemri, T., Armstrong, R. C., Krebs, J., Srinivasula, S. M., Wang, L., Bullrich, F., Fritz, L. C., Trapani, J. A., Tomaselli, K. J., Litwack, G., and Alnemri, E. S. (1996) Proc. Natl. Acad. Sci. USA 93, 7464 –7469. 5. Vincenz, C., and Dixit, V. M. (1997) J. Biol. Chem. 272, 6578 – 6583. 6. Nagata, S. (1997) Cell 88, 355–365. 7. Kischkel, F. C., Hellbardt, S., Behrmann, I., Germer, M., Pawlita, M., Krammer, P. H., and Peter, M. E. (1995) EMBO J. 14, 5579 –5588. 8. Varfolomeev, E. E., Schuchmann, M., Luria, V., Chiannilkulchai, N., Beckmann, J. S., Mett, I. L., Rebrikov, D., Brodianski, V. M., Kemper, O. C., Kollet, O., Lapidot, T., Soffer, D., Sobe, T., Avraham, K. B., Goncharov, T., Holtman, H., Lonai, P., and Wallach, D. (1998) Immunity 9, 267–276.

Vol. 291, No. 3, 2002

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

9. Wang, J., Zheng, L., Lobito, A., Chan, F. K., Dale, J., Sneller, M., Yao, X., Puck, J. M., Straus, S. E., and Lenardo, M. J. (1999) Cell 98, 47–58. 10. Teitz, T., Wei, T., Valentine, M. B., Vanin, E. F., Grenet, J., Valentine, V. A., Behm, F. G., Look, A. T., Lahti, J. M., and Kidd, V. J. (2000) Nat. Med. 6, 529 –535. 11. Mandruzzato, S., Brasseur, F., Andry, G., Boon, T., and van der Bruggen, P. (1997) J. Exp. Med. 186, 785–793. 12. Hopkins-Donaldson, S., Bodmer, J. L., Bourloud, K. B., Brognara, C. B., Tschopp, J., and Gross, N. (2000) Cancer Res. 60, 4315– 4319. 13. Ng, P. W., Porter, A. G., and Janicke, R. U. (1999) J. Biol. Chem. 274, 10301–10308. 14. Siegel, R. M., Martin, D. A., Zheng, L., Ng, S. Y., Bertin, J., Cohen, J., and Lenardo, M. J. (1998). J. Cell Biol. 141, 1243– 1253. 15. Shikama, Y., U, M., Miyashita, T., and Yamada, M. (2001). Exp. Cell Res. 264, 315–325. 16. Gossen, M., Freundlieb, S., Bender, G., Muller, G., Hillen, W., and Bujard, H. (1995) Science 268, 1766 –1769. 17. U, M., Miyashita, T., Ohtsuka, Y., Okamura-Oho, Y., Shikama, Y., and Yamada, M. (2001) Cell Death Differ. 8, 377–386. 18. Nakamura, N., Rabouille, C., Watson, R., Nilsson, T., Hui, N., Slusarewicz, P., Kreis, T. E., and Warren, G. (1995) J. Cell Biol. 131, 1715–1726. 19. Pitti, R. M., Marsters, S. A., Ruppert, S., Donahue, C. J., Moore, A., and Ashkenazi, A. (1996) J. Biol. Chem. 271, 12687–12690. 20. Wiley, S. R., Schooley, K., Smolak, P. J., Din, W. S., Huang, C. P., Nicholl, J. K., Sutherland, G. R., Smith, T. D., Rauch, C., and Smith, C. A. (1995) Immunity 3, 673– 682. 21. Pan, G., O’Rourke, K., Chinnaiyan, A. M., Gentz, R., Ebner, R., Ni, J., and Dixit, V. M. (1997) Science 276, 111–113. 22. Tsukumo, S. I., and Yonehara, S. (1999) Genes Cells 4, 541–549.

23. Han, D. K., Chaudhary, P. M., Wright, M. E., Friedman, C., Trask, B. J., Riedel, R. T., Baskin, D. G., Schwartz, S. M., and Hood, L. (1997) Proc. Natl. Acad. Sci. USA 94, 11333–11338. 24. Hu, S., Vincenz, C., Ni, J., Gentz, R., and Dixit, V. M. (1997) J. Biol. Chem. 272, 17255–17257. 25. Irmler, M., Thome, M., Hahne, M., Schneider, P., Hofmann, K., Steiner, V., Bodmer, J. L., Schroter, M., Burns, K., Mattmann, C., Rimoldi, D., French, L. E., and Tschopp, J. (1997) Nature 388, 190 –195. 26. Shu, H. B., Halpin, D. R., and Goeddel, D. V. (1997) Immunity 6, 751–763. 27. Srinivasula, S. M., Ahmad, M., Ottilie, S., Bullrich, F., Banks, S., Wang, Y., Fernandes-Alnemri, T., Croce, C. M., Litwack, G., Tomaselli, K. J., Armstrong, R. C., and Alnemri, E. S. (1997) J. Biol. Chem. 272, 18542–18545. 28. Goltsev, Y. V., Kovalenko, A. V., Arnold, E., Varfolomeev, E. E., Brodianskii, V. M., and Wallach, D. (1997) J. Biol. Chem. 272, 19641–19644. 29. Chaudhary, P. M., Eby, M. T., Jasmin, A., Kumar, A., Liu, L., and Hood, L. (2000) Oncogene 19, 4451– 4460. 30. Hu, W. H., Johnson, H., and Shu, H. B. (2000) J. Biol. Chem. 275, 10838 –10844. 31. Wang, C. Y., Mayo, M. W., Korneluk, R. G., Goeddel, D. V., and Baldwin, A. S. J. (1998) Science 281, 1680 –1683. 32. Mancini, M., Machamer, C. E., Roy, S., Nicholson, D. W., Thornberry, N. A., Casciola-Rosen, L. A., and Rosen, A. (2000) J. Cell Biol. 149, 603– 612. 33. Colussi, P. A., Harvey, N. L., and Kumar, S. (1998) J. Biol. Chem. 273, 24535–24542. 34. Bennett, M., Macdonald, K., Chan, S. W., Luzio, J. P., Simari, R., and Weissberg, P. (1998) Science 282, 290 –293. 35. Jones, S. J., Ledgerwood, E. C., Prins, J. B., Galbraith, J., Johnson, D. R., Pober, J. S., and Bradley, J. R. (1999) J. Immunol. 162, 1042–1048.

493