A Novel Single Amino Acid Deletion Caspase-8 Mutant in Cancer Cells That Lost Proapoptotic Activity*

Caspase-8 is an important initiation caspase that activates the caspase cascade during death receptor-mediated apoptosis. We here report a novel caspase-8 mutant with a naturally occurring deletion of leucine 62 (ΔLeu62casp-8). ΔLeu62casp-8 has a shorter half-life than its wild-type counterpart. Unlike wild-type caspase-8, ΔLeu62casp-8 failed to interact with wild-type caspase-8 or with the adaptor protein FADD. ΔLeu62casp-8 lost its proapoptotic activity in mammalian cells. The leucine 62 therefore is critical for caspase-8 function, and the mutation may be one of the mechanisms through which some types of cancer cells escape from programmed cell death.

Recent studies have made substantial progress in delineating the signal transduction pathways that couple the initiation caspases to downstream cellular effects. Increasing evidence suggests that defects in these pathways may cause disease (1). Cancer cells may become resistant to death receptor-induced apoptosis because of the expression of decoy receptors, inactivation of death receptors, or loss of their distal signaling molecules. For example, homozygous deletion of the death receptor DR4 gene may lead to TRAIL resistance (17). Mutation in Fas was found in T-lineage acute leukemia (18,19). Caspase-8 gene silencing due to DNA methylation was found in childhood neuroblastomas (20). Activation of survival signaling pathways, such as up-regulation of NF-B, FLIP protein, or members of the IAP family, could also confer on tumor cells resistance to death receptor-mediated apoptosis (21). In this article, we report a novel, naturally occurring caspase-8-defective mutant with a single amino acid deletion of leucine 62 (⌬Leu62casp-8) that we discovered in A431 human vulvar squamous carcinoma cells. Functional analysis indicates that ⌬Leu62casp-8 lost its proapoptotic activity when overexpressed in mammalian cells, suggesting that the leucine 62 is critical for caspase-8 function and that the deletion of the amino acid may play a role in tumorigenesis.

EXPERIMENTAL PROCEDURES
Transfection of Cells with Expression Vectors-Cell transfection was performed with the FuGENE TM -6 transfection kit (Roche Diagnostics Corp., Indianapolis, IN) according to the manufacturer's instructions.
Immunoprecipitation and Immunoblotting-Cells were scraped off from culture dishes with a rubber scraper and lysed in a Nonidet P-40 lysis buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.5% Nonidet P-40, 50 mM NaF, 1 mM Na 3 VO 4 , 1 mM phenylmethylsulfonyl fluoride, 25 g/ml leupeptin, 25 g/ml aprotinin). Protein concentrations were measured by using the Coomassie Plus protein assay reagent (Pierce Chemical Co.). Equal amounts of cell extracts were used for immunoprecipitation and immunoblotting analysis as we previously described (22,23).
RNA Extraction, cDNA Synthesis, RT-PCR, and cDNA Subcloning-Total RNA from the cell lines was extracted using a modified chloroform/phenol procedure (TRIZOL TM , Invitrogen, Carlsbad, CA). First strand of cDNA was generated using reverse transcriptase (RTase) (Roche) following the manufacturer's protocol, and subsequently amplified by PCR using the Expand TM High Fidelity PCR System (Roche) and the following primer sets: caspase-8 forward primer, 5Ј-CGGGATCCG-CCACCATGGACTTCAGCAGAAATC-3Ј and the reverse primer, 5Ј-T-CCCCCGGGCACCATCAATCAGAAGGG-3Ј. The semiquantitative PCR control transcript GAPDH was amplified by the specific primers provided by CLONTECH Laboratories, Inc. (Palo Alto, CA). Amplified fragments were confirmed by DNA sequencing and subcloned into the pcDNA3.1-His C version plasmid (Invitrogen).
Northern Blot Analysis-Total RNA was prepared as described above. 10 g of total RNA was loaded into each well, electrophoretically * This work was supported by a research award from the Bristol-Myers Squibb Company and by the Cancer Center Core Grant CA16672 from the National Cancer Institute. 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.
Protein Stability Analysis-MCF7 cells were transiently transfected with pcDNA3.1-Casp8 or pcDNA3.1-⌬Leu62Casp8 for 15 h. The cells were then metabolically labeled with 100 Ci/ml [ 35 S]methionine (PerkinElmer Life Sciences) for 30 min, followed by a 2-h chase period. Cell lysates were harvested at each of the indicated time points for immunoprecipitation with anti-HisG monoclonal antibody (Invitrogen). The immunoprecipitates were separated by SDS-PAGE and followed by autoradiography and quantification by a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).
Protein in Vitro Translation and Binding Assay and GST Pulldown Assay-The pcDNA3.1-Casp8, pcDNA3.1-⌬Leu62Casp8, and pcDNA3.1-FADD were in vitro transcribed and translated in the presence of L-[ 35 S]methionine using the TNT kit from Promega Corp. (Madison, WI). 5 l of the in vitro translated products were incubated with 5 g of GST, GST-DED, or GST-⌬DED fusion proteins, respectively, in 500 l of GST binding buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol, and 0.1% Tween 20) at 4°C for 2 h, followed by 1 h additional incubation with glutathione-Sepharose beads. After washing the beads three times with 500 l of buffer, the bound proteins were eluted with glutathione, analyzed by SDS-PAGE, and stained with Coomassie Blue. For the GST pull-down assay, 200 g of total cell lysates were incubated with 5 g of GST, GST-DED, or GST-⌬DED fusion proteins in 500 l of GST binding buffer at 4°C overnight. The GST and GST fusion proteins were collected by glutathione-Sepharose beads. After extensively washing with 500 l of buffer, the recovered proteins were resolved by SDS-PAGE and followed by immunoblotting analysis with specific antibodies against caspase-8 or FADD.
In Vitro Processing and Activation of Caspase-8 by Granzyme B-In vitro translated, 35 S-labeled caspase-8 and ⌬Leu62Casp8 proteins were incubated with or without recombinant granzyme B (GraB) in a reaction buffer (50 mM HEPES, pH 7.4, 0.1% CHAPS, 0.1 M NaCl, 10% sucrose, and 10 mM dithiothreitol) at 37°C for 4 h, and the reaction products were analyzed by SDS-PAGE and visualized with autoradiography. To examine the activities of the products following GraB processing, a specific GraB inhibitor (Enzyme System Products, Livermore, CA) was added for 15 min followed by incubating with the putative substrate PARP at 37°C for another 2 h. The reaction mixture were resolved by SDS-PAGE and followed by immunoblotting analysis with specific antibody against PARP (CHUL Research Center, Laval University, Quebec).

RESULTS AND DISCUSSION
In contrast to many other cell lines, the A431 human vulvar squamous carcinoma cells are insensitive to TRAIL-mediated cytotoxic effect. Fig. 1A shows the result of treating a panel of six tumor cell lines with TRAIL. Exposure of the cell lines to TRAIL induced apoptosis, measured by an apoptosis ELISA, except in A431 cells. TRAIL is known to induce apoptosis in a caspase-8-dependent manner (25,26). Western blot analysis indicated that A431 cells had an extremely low level of caspase-8 protein (Fig. 1B). The low expression of caspase-8 in A431 cells could not be attributed to caspase-8 gene methylation, which was found in neuroblastomas (20), because treatment of A431 cells with the DNA demethylation agent 5-aza-2Ј-deoxycytidine failed to activate the expression (data not shown). Northern blot analysis indicated that A431 cells expressed a considerable level of caspase-8 mRNA, compared with the other five cell lines (Fig. 1C). This result further suggested that the low expression level of caspase-8 observed in the A431 cells could not be attributed to changes at the transcription level. We further examined the effects on these cells of two additional caspase-8-dependent apoptosis inducers (the anti-Fas agonistic antibody CH-11 and TNF␣), and two caspase-8-independent chemotherapeutic agents (paclitaxel (Bristol-Myers Squibb Company, Princeton, NJ) and cisplatin). A431 cells expressed detectable levels of Fas and TNFR1 (data not shown); however, these cells were resistant to CH-11 or TNF␣, but not to paclitaxel or cisplatin (Fig. 1D), indicating the presence of a defective caspase-8-initiated apoptotic pathway in A431 cells.
To explore the underlying mechanism of low expression and low activity level of caspase-8 in A431 cells, we used a semi- quantitative RT-PCR to amplify the coding sequence of caspase-8 from A431 cells. H460 non-small cell lung carcinoma and MDA468 breast cancer cells served as controls ( Fig. 2A). Consistent with the Northern blot data, the PCR products from all three cell lines each displayed a single band with similar size and intensity. However, sequencing of the PCR products indicated that there was an in-frame deletion of three consecutive base pairs (coding for leucine 62) found in A431 cells, compared with the published sequence of caspase-8 in Gen-Bank TM (Fig. 2B). Caspase-8 is composed of two DEDs at the amino terminus and an ICE-like activity domain at the carboxyl terminus (6,7). The missing leucine 62 is located in the first DED of caspase-8. This novel caspase-8 mutant is hereafter termed ⌬Leu62casp-8. When transiently transfected into MCF7 breast cancer cells, ⌬Leu62casp-8 was expressed and even showed a slightly higher expression level than wild-type caspase-8 in our well controlled and repeated experiments (Fig.  2C). However, a pulse-chase experiment with [ 35 S]methionine metabolically labeled wild-type caspase-8 and ⌬Leu62casp-8 indicated that, regardless of its relatively high protein level, ⌬Leu62casp-8 had a faster degradation rate, with a half-life of Ͻ30 min compared with Ͼ1 h for wild-type caspase-8 (Fig. 2D), which may contribute to the low level of caspase-8 (⌬Leu62casp-8) in A431 cells.
Caspase-8 is proteolytically activated by oligomerization following its recruitment to the FADD molecule upon death receptor activation (4 -7). Overexpression of caspase-8 alone was sufficient to induce caspase-8 oligomerization and autoactivation, triggering the caspase-8-initiated cell death pathway (7). We therefore hypothesized that the relative lower level of transfected wild-type caspase-8 could be due to an autoactivation-related autocleavage, which might not be the case for transfected ⌬Leu62casp-8. To address this hypothesis, we examined the capability of ⌬Leu62casp-8 to undergo oligomerization and interact with FADD. GST-fused wild-type caspase-8 DED domain (GST-DED) and ⌬Leu62casp-8 DED domain (GST-⌬DED) (Fig. 3A) were used, respectively, for in vitro binding assays with in vitro translated FADD, wild-type caspase-8, or ⌬Leu62casp-8 (Fig. 3B). An interaction was found between FADD and the wild-type caspase-8 DED but not between FADD and the ⌬Leu62casp-8 DED (Fig. 3C). Furthermore, wild-type caspase-8 interacted with the caspase-8 DED but not with the ⌬Leu62casp-8 DED, and the ⌬Leu62casp-8 DED failed to interact with either wild-type caspase-8 or itself (Fig. 3D). To further confirm these in vitro results, the proteinprotein interaction assays were repeated with the caspase-8 or FADD expressed by cells. Interactions of the wild-type caspase-8 DED with caspase-8 were found in the lysates of MCF7 and Jurkat cells but not in the lysate of A431 cells; no interaction of ⌬Leu62casp-8 DED with caspase-8 was found in the lysate of any cell line (Fig. 3E, top). Interactions of wildtype caspase-8 DED with FADD were found in all three cell lines, but ⌬Leu62casp-8 DED and FADD did not interact (Fig.  3E, bottom). These data strongly demonstrate that ⌬Leu62casp-8 was unable to interact with itself, wild-type caspase-8, or the adaptor protein FADD.
To determine whether ⌬Leu62casp-8 retained its enzymatic activity, in vitro translated wild-type caspase-8 and ⌬Leu62casp-8 were incubated with recombinant granzyme B, a serine protease that activates caspase-8 by proteolytic processing (7). Both wild-type caspase-8 and ⌬Leu62casp-8 were efficiently processed by granzyme B (Fig. 3F). The granzyme Bprocessed wild-type caspase-8 and ⌬Leu62casp-8 showed equal activities in cleaving PARP (Fig. 3G). To block any potential direct effect of granzyme B on PARP, in this assay, a selective inhibitor of granzyme B was used before PARP was incubated with granzyme B-processed caspase-8 or ⌬Leu62casp-8. The results indicate that although ⌬Leu62casp-8 was defective in protein-protein interaction, it could still be processed by granzyme B and was enzymatically active once processed.
Caspase-8 can trigger apoptosis when overexpressed, presumably through proximity-induced autoactivation (6,7,27). To further investigate whether ⌬Leu62casp-8 remained proapoptotic, wild-type caspase-8 and ⌬Leu62casp-8 were tran- GST-bound proteins were resolved by SDS-PAGE, followed by autoradiography. E, GST pull-down assay. The GST fusion proteins in A were incubated with the lysates of A431, MCF7, and Jurkat cells as indicated. GST-bound proteins, along with aliquots of the cell lysates, were resolved by SDS-PAGE, followed by Western blot analyses with antibodies directed against caspase-8 or FADD. F, processing of wildtype caspase-8 and ⌬Leu62casp-8 by granzyme B (GraB). In vitro translated, 35 Slabeled caspase-8 and ⌬Leu62casp-8 proteins were incubated with or without 1 g of recombinant GraB at 37°C for 4 h, and the reaction products were resolved by SDS-PAGE, followed by autoradiography. G, cleavage of PARP by GraB-processed wild-type caspase-8 and ⌬Leu62casp-8. The GraB-processed wild-type caspase-8 and ⌬Leu62casp-8 protein were mixed with 30 M of GraB-specific inhibitor (Enzyme System Products) for 15 min prior to a 2-h incubation with 200 ng of recombinant PARP protein (Alexis Corp., San Diego, CA). The reaction products were resolved by SDS-PAGE, followed by Western blot analysis with specific antibody against PARP.
siently expressed in A431 and in MCF7 cells (Fig. 4A). An apoptosis-specific ELISA demonstrated induction of apoptosis in both A431 and MCF7 cells following transfection with wildtype caspase-8 but not with control vector or ⌬Leu62casp-8 (Fig. 4B). In addition, typical apoptotic changes in morphology were seen in wild-type caspase-8-transfected MCF7 cells but not in the ⌬Leu62casp-8-transfected cells (Fig. 4C). Similar changes were also observed in A431 cells (data not shown). We therefore conclude that ⌬Leu62casp-8 lost its proapoptotic activity in vivo. Regardless of its high expression level, the failure of ⌬Leu62casp-8 to mediate apoptosis in A431 cells indicates that the loss of its function was the critical factor contributing to the resistance of A431 cells to treatment with TRAIL, TNF␣ or the anti-Fas agonistic antibody shown in Fig. 1. The results also indicate that the down stream partners of the caspase-8initiated pathway appear to be intact since expression of the wild-type caspase-8 in A431 cells can trigger apoptosis in the cells.
In addition to the DNA methylation found in neuroblastomas (20), two mutation sites in caspase-8 have recently been reported (28,29). A mutation that modifies the stop codon of caspase-8 and adds an Alu repeat to the coding region was found in a head and neck cancer cell line (28). Its ability to trigger apoptosis was reduced relative to that of wild-type caspase-8 but was not totally abolished. Another missense mutation (alanine to valine) at the caspase-8 gene codon 96 was found in a neuroblastoma cell line lacking caspase-8 expression, but no functional studies were explored (29). We showed FIG. 4. Loss of proapoptotic activity of ⌬Leu62casp-8. A and B, induction of apoptosis by wild-type caspase-8 but not by ⌬Leu62casp-8. A431 and MCF7 cells were individually transiently transfected with pcDNA3.1 control vector, or the vectors containing wild-type caspase-8, or ⌬Leu62casp-8 using the FuGENE TM -6 reagent for 20 h, followed by cell lysis for the detection of wild-type caspase-8, ⌬caspase-8 and ␤-actin expressions by Western blot analysis (A), and for quantification of apoptosis by ELISA (B). C, induction of nuclear condensation by wild-type caspase-8 but not by ⌬Leu62casp-8. MCF7 cells were transiently transfected with vectors described in A and B, the cells were then stained with 1 g/ml Hoechst 33342 in phosphate-buffered saline for 5 min at 37°C and then observed and photographed under a fluorescent microscope (arrow, condensed nuclei).
here that deletion of the leucine 62 in caspase-8 (⌬Leu62casp-8) dramatically altered the proapoptotic function of caspase-8. Our findings are significant. From the mechanistic point of view, the results indicate that leucine 62 located in the first DED of caspase-8 plays a critical role for caspase-8 oligomerization and for its interaction with FADD for triggering the caspase cascade. From the cancer biology point of view, our observation, along with several other types of reported caspase-8 genetic changes in cancer cells, including gene deletion, methylation, and point mutation, suggests a possible mechanism by which some cancer cells may escape this genetically programmed, caspase-8-initiated cell death. Restoration of the caspase-8-mediated cell death pathway may be therefore a rational strategy for cancer therapeutics.