ILPIP, a novel anti-apoptotic protein that enhances XIAP-mediated activation of JNK1 and protection against apoptosis.

We have previously described a new aspect of the Inhibitor of Apoptosis (IAP) family of proteins anti-apoptotic activity that involves the TAK1/JNK1 signal transduction pathway (1,2). Our findings suggest the existence of a novel mechanism that regulates the anti-apoptotic activity of IAPs that is separate from caspase inhibition but instead involves TAK1-mediated activation of JNK1. In a search for proteins involved in the XIAP/TAK1/JNK1 signaling pathway we isolated by yeast two-hybrid screening a novel X chromosome-linked IAP (XIAP)-interacting protein that we called ILPIP (hILP-Interacting Protein). Whereas ILPIP moderately activates JNK family members when expressed alone, it strongly enhances XIAP-mediated activation of JNK1, JNK2, and JNK3. The expression of a catalytically inactive mutant of TAK1 blocked XIAP/ILPIP synergistic activation of JNK1 thereby implicating TAK1 in this signaling pathway. ILPIP moderately protects against interleukin-1beta converting enzyme- or Fas-induced apoptosis and significantly potentiates the anti-apoptotic activity of XIAP. In vivo co-precipitation experiments show that both ILPIP and XIAP interact with TAK1 and tumor necrosis factor receptor-associated factor 6. Finally, expression of ILPIP did not affect the ability of XIAP to inhibit caspase activation, further supporting the idea that XIAP protection against apoptosis is achieved by two separate mechanisms: one requiring JNK1 activation and a second involving caspase inhibition.

From the ‡Department of Immunology, The Scripps Research Institute, La Jolla, California 92037, §Rigel Inc., South San Francisco, California 94080, and ¶The Genomics Institute of the Novartis Research Foundation, San Diego, California 92121 We have previously described a new aspect of the Inhibitor of Apoptosis (IAP) family of proteins anti-apoptotic activity that involves the TAK1/JNK1 signal transduction pathway (1, 2). Our findings suggest the existence of a novel mechanism that regulates the antiapoptotic activity of IAPs that is separate from caspase inhibition but instead involves TAK1-mediated activation of JNK1. In a search for proteins involved in the XIAP/TAK1/JNK1 signaling pathway we isolated by yeast two-hybrid screening a novel X chromosomelinked IAP (XIAP)-interacting protein that we called ILPIP (hILP-Interacting Protein). Whereas ILPIP moderately activates JNK family members when expressed alone, it strongly enhances XIAP-mediated activation of JNK1, JNK2, and JNK3. The expression of a catalytically inactive mutant of TAK1 blocked XIAP/ILPIP synergistic activation of JNK1 thereby implicating TAK1 in this signaling pathway. ILPIP moderately protects against interleukin-1␤ converting enzyme-or Fas-induced apoptosis and significantly potentiates the anti-apoptotic activity of XIAP. In vivo co-precipitation experiments show that both ILPIP and XIAP interact with TAK1 and tumor necrosis factor receptor-associated factor 6. Finally, expression of ILPIP did not affect the ability of XIAP to inhibit caspase activation, further supporting the idea that XIAP protection against apoptosis is achieved by two separate mechanisms: one requiring JNK1 activation and a second involving caspase inhibition.
Apoptosis is an active process in which an individual cell responding to internal and/or external cues commits suicide. Apoptosis is involved in many diverse homeostatic processes in multicellular organisms, both during development and in the mature organism, and is increasingly being recognized as a biological process of critical importance, not only in normal physiology but also in the pathogenesis of diseases such as cancer and neurodegenerative diseases (3)(4)(5). Apoptosis is genetically determined and controlled through the expression of an increasing number of genes, many of which are conserved in nematodes, mammals, and viruses (6).
Caspases, a family of cysteine proteases, are the most extensively studied activators of apoptosis. Among the anti-apoptotic gene products is the IAP 1 (Inhibitor of Apoptosis) family of proteins. Initially discovered in baculovirus, where they were shown to be involved in suppressing the host cell death response to viral infection (7,8), IAP homologues were then isolated in Drosophila, Caenorhabditis elegans, yeast, and mammalian organisms. To date, seven members of the IAP family have been identified in mammalian cells, with XIAP being the most extensively studied member of the family (9 -16). IAPs have been shown to protect against a wide spectrum of apoptotic triggers. The diversity of stimuli against which the IAPs suppress apoptosis is greater than that observed for any other family of apoptotic inhibitors. At least one suggested mechanism of IAP apoptotic suppression appears to be through direct caspase inhibition. Several of the human IAP family proteins have been reported to directly bind and inhibit specific members of the caspase family (17).
XIAP has been shown to participate in the BMP signaling pathway by binding with both the BMP receptor and the adaptor molecule TAB1, which is a co-activator of TAK1, thus linking the BMP receptors to TAB1-TAK1 and therefore participating in the bone morphogenic protein-signaling pathway involved in mesoderm induction and patterning in early Xenopus embryos (18). XIAP also stimulates NF-B via the TAK1 signaling pathway (19). Consistent with these findings, we have recently described an alternative mechanism for the IAP anti-apoptotic protection, which is distinct from caspase inhibition and involves activation of the MAPK JNK1 through the TAB1⅐TAK1 complex (1,2,20).
XIAP involvement in signal transduction pathways is still poorly characterized. Few proteins have been shown to interact with XIAP. A recently described XIAP-interacting protein is SMAC/DIABLO, which promotes caspase activation by binding and inhibiting the IAPs (21,22). XAF1 has also been reported to bind to XIAP and inhibit its anti-apoptotic effect apparently by triggering the redistribution of XIAP from the cytosol to the nucleus (11). However, none of these interactions has been correlated with the XIAP-mediated activation of JNK1. This led us to search for new XIAP-interacting proteins and to investigate their role in the XIAP/TAK1/JNK1 signaling cascade.

Yeast Two-hybrid Screening and Isolation of ILPIP
A cDNA fragment encoding human XIAP was amplified by PCR and inserted into the XhoI site of pBD-GAL4 Cam (Stratagene, LA Jolla, CA) to generate pBD-GAL4 Cam/XIAP. After verification of the sequence, Saccharomyces cerevisiae CG1945 cells (CLONTECH, Palo Alto, CA) were sequentially transformed with pBD-GAL4 Cam/XIAP and human fetal brain library (Matchmaker, CLONTECH, 10 8 cfu/ml) in pACT using a lithium acetate transformation protocol. Selection was done by growth on synthetic (SD) medium (yeast nitrogen base/dextrose/essential amino acids) lacking histidine, uracil, and tryptophan (CLONTECH). Twenty clones exhibiting activation of the lacZ reporter gene were identified among 3 ϫ 10 6 transformants by the ␤-galactosidase assay. A few clones showing a strong reproducible interaction with XIAP were chosen. Plasmids were isolated from positive yeast colonies by a glass bead phenolchloroform extraction protocol (CLONTECH). Escherichia coli DH5␣ bacteria cells were transformed with the plasmids and bacteria containing the pACT vector were selected on ampicillin-resistant plates. The pACT plasmids were isolated from E. coli and restriction-mapped (XhoI), and the sequence of the insert was determined. The partial cDNA two-hybrid clone was used to design a probe to screen a human liver Uni-ZAP cDNA lambda library (Stratagene) and 5Ј-rapid amplification of cDNA ends techniques. Phage plaques were isolated and screened to verify the cDNA size by PCR using the T3 and T7 primers. The phagemid pBluescript vector carrying the cDNA of nine individual clones was isolated by in vivo excision from the Uni-ZAP vector according to the manufacturer's instructions (Stratagene). The isolated cDNAs were sequenced, and sequences were analyzed using the GCG Sequence Analysis software package (Madison, WI). The full-length ILPIP gene encoding an ϳ2.4-kb cDNA was subcloned by BamHI-XhoI restriction sites into pcDNA3 vector encoding an HA tag at the N terminus. The sequence contains an initiator methionine, a stop codon, and a poly-adenylation tail and provides evidence for a novel gene, which we have named ILPIP. Northern blot was performed using the partial ILPIP cDNA isolated from the two-hybrid system following standard procedures.

Transfection and Cell Culture
Human embryonic kidney cells (293T) were grown at 37°C in 5% CO 2 in Dulbecco's modified Eagle's medium with 10% fetal bovine serum, 2 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin. For transfection, each well of a six-well plate was seeded with 7 ϫ 10 5 cells. Cells were transfected 18 h later using LipofectAMINE Plus reagent (Invitrogen) for 3 h and incubated for 18 h before lysis. MCF7-Fas cells were grown in RPMI 1640 containing 10% fetal bovine serum, 200 g/ml G418, and 100 g/ml hygromycin and grown at 37°C in 5% CO 2 . For transfection each well of a six-well plate was seeded with 2.5 ϫ 10 5 cells and 24 h after plating were transfected using LipofectAMINE Plus reagent (Invitrogen). 24 h after transfection, cells were treated with anti-Fas antibody (150 ng/ml). After 16 h, cells were fixed and stained as described below.

Cell Lysis and Kinase Assay
Cell lysis was performed for 30 min at 4°C with lysis buffer (25 mM Hepes (pH 7.6), 1% Triton X-100, 137 mM NaCl, 3 mM ␤-glycerophos-phate, 3 mM EDTA, 0.1 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride). Expression of MAPK proteins was quantified by densitometry after Western blot analysis, and equivalent amounts were immunoprecipitated at 4°C for 2 h. The immunoprecipitates were washed twice with lysis buffer and twice with kinase buffer (see below) before performing the kinase assay. HA-and Myc-tagged proteins were immunoprecipitated using 20 l of agarose-Protein A (Pierce) preincubated with anti-HA antibody or anti-Myc antibody (5 g, Roche Molecular Biochemicals and Upstate Biotechnology, respectively) and FLAGtagged proteins with 20 l of agarose conjugated with the M2 anti-FLAG monoclonal antibody (Sigma Chemical Co.).
In vitro kinase assays was performed as previously described (1) with the difference that a 30-min incubation time was used for the detection of JNK2 and JNK3 kinase activity. UV activation of JNK1 was carried out in 293T cells as previously reported (1).

Detection of Apoptotic Cells
␤-Galactosidase Staining-Cells were transfected with the indicated plasmids together with a ␤-galactosidase expressing vector and stained with X-gal (reagent for ␤-galactosidase expression) to allow visualization of transfected cells and morphology observation. Quantification of apoptotic cells was determined at the microscope by counting over five fields for each sample. Apoptotic cells appear to be smaller, rounder, and show condensed and misshapen nuclei compared with viable cells, which are flat, well spread, and with easily discernible nuclei. Protein expression for all the transfected constructs was assessed by Western blot on duplicate lysates of original transfections used for the apoptosis assays.
Annexin V-PE/FACS Analysis-Cell were C. O. transfected with the indicated plasmids together with green fluorescent protein vector (CLONTECH Laboratories) to allow quantitation of transfection efficiency. Annexin V-PE staining was performed as suggested by the manufacturer (BD PharMingen). Briefly, adherent cells were detached from the plates and centrifuged for 5 min at 1000 rpm. After removing the supernatant, cells were washed with Annexin V binding buffer and centrifuged again, and the supernatant was decanted by inversion of the tube. Cells were resuspended, and Annexin V-PE conjugate (5 l) was added to each sample, incubated for 10 min in the dark, and then analyzed by FACS within 1 h.
Death by apoptosis was quantified both by X-gal staining of cells and Annexin V-PE-FACS analysis for each experiments. The results obtained using the two different techniques were comparable, and therefore only the x-gal data are shown.

Co-immunoprecipitations and Immunoblot Assays
Cells were washed extensively and lysed in 200 l of lysis buffer containing 50 mM Hepes, 100 mM NaCl, 2 mM EDTA, 10% glycerol, 1% Nonidet P-40, 14 mM pepstatin A, 100 mM leupeptin, 3 mM benzamidine, 1 mM phenylmethylsulfonyl fluoride, 1 mM sodium pyrophosphate, 10 mM sodium orthovanadate, 100 units/ml aprotinin, 100 mM sodium fluoride. After incubation for 30 min on ice, cell lysates were centrifuged (14,000 rpm, 10 min, 4°C), and the supernatants were recovered. Cell lysates were pre-cleared three times for 20 min at 4°C with 20 l of protein A-Sepharose beads and mixed with specified antibodies for 3 h at 4°C under constant agitation. Immune complexes were allowed to bind to 20 l of protein A-Sepharose beads overnight, beads were washed three times with lysis buffer, and the washed beads were resuspended in 30 l of Laemmli buffer and boiled for 10 min. Immunoprecipitates were separated on 12% SDS-PAGE and transferred to nitrocellulose membranes. Filters were blocked with 5% nonfat milk in blocking buffer (TBS, 50 mM Tris-HCl, pH 7.5; 150 mM NaCl; 0.1% Tween 20), and incubated with the specified antibody for 2 h and with peroxidase-conjugated secondary antibody for 1 h at ambient temperature. Specific bands were revealed using the ECL Plus system (Amersham Biosciences).

In Vitro Binding Assays
In vitro translation of TAK1 was performed using standard procedures (Promega). XIAP-GST protein was expressed from a pGEX vector (Amersham Biosciences) and purified as suggested by the manufacturer and JNK1-HIS protein was purchased from Santa Cruz Biotechnology. Glutathione-or nickel-nitrilotriacetic acid-conjugated beads (from Sigma and Qiagen, respectively) were used to precipitate XIAP-GST and JNK1-HIS. Binding assays were performed in lysis buffer, and TAK1 interaction with XIAP or JNK1 was detected by Western blot using an anti-TAK1 antibody (Santa Cruz Biotechnology).
FIG. 1. ILPIP predicted sequences and homologies. A, nucleotide and amino acid sequence of ILPIP. ILPIP cDNA was cloned as described in the text. Starting methionines for ILPIP␣ and ILPIP␤ are highlighted in boldface at amino acid numbers 1 and 139, respectively. The ILPIP␣ and ILPIP␤ sequences have been submitted to GenBank TM under accession number AY093697. B, predicted amino acid homology of ILPIP␣ with known proteins. By protein sequence homology, ILPIP is related to serologically defined breast cancer antigen NY-BR-96. Both ILPIP and

Caspase Activation in Cytosolic Extracts
Cytosolic extracts from transfected 293T (100-mm dishes) were prepared essentially as described previously (24), with several modifications (25). Briefly, cells were washed once with ice-cold buffer A and pelleted by centrifugation. Packed cell pellets were suspended in 1-2 volumes of buffer A, incubated on ice for 20 min, and then disrupted by 15-30 passages through a 26-gauge needle. Cell extracts were clarified by centrifugation at 16,000 ϫ g for 10 min, and the resulting supernatants were used for "cell-free" assays. For initiating caspase activation, 10 M horse heart cytochrome c (Sigma) together with 1 mM dATP was added, and the assays were incubated at 30°C for 10 min. 1 l (10 g of total protein) was measured for caspase activity by monitoring the release of AFC DEVD-containing synthetic peptides using continuously reading instruments as described previously (26). Fluorogenic 7-amino-4-trifluoromethyl coumarin (AFC) caspase substrate (Ac-DEVD-AFC) was purchased from Sigma.

RESULTS
XIAP Bait Identifies ILPIP in a Two-hybrid Screening-As a first step in the characterization of the XIAP/JNK1/TAK1/ TAB1 interactions, we used a XIAP-Gal4 DNA binding domain fusion construct as bait in the two-hybrid system to screen a human fetal brain cDNA-Gal4 activation domain function library (Matchmaker, CLONTECH). Among the selected primary transformants, positive yeast colonies were independently identified and isolated. Plasmids were isolated from positive yeast cells, and the sequence of the insert was determined. This process lead to the identification of a 700-bp cDNA sequence. This partial cDNA two-hybrid clone was used to design a probe to screen a human liver Uni-ZAP cDNA lambda library (Stratagene). The library screen yielded the parent 2.4-kb cDNA containing an initiator methionine, a stop codon, and a poly-adenylation tail that encoded for a 418-amino acid protein that we have named ILPIP␣ (Fig. 1A). The predicted ILPIP amino acid sequence revealed that ILPIP is identical to ALS2CR2, a novel gene identified within the ALS2 (juvenile amyotrophic lateral sclerosis) critical region (27). At the protein level, the strongest homologies were found with serologically defined breast cancer antigen NY-BR-96 (ϳ20% identity, 46% homology) and with SPAKs (STE20/SPS1-related proline alanine-rich kinases), a member of the Ste20/SPS1 family of kinases and other Ste20-like kinases (Fig. 1B, up to ϳ20% overall amino acid identity, ϳ45% similarity). ILPIP has a putative protein kinase domain (amino acids 58 -369, Fig. 1C) with a hypothetical ATP-binding site located at the N terminus of the protein and several serine, threonine, or tyrosine active and phosphorylation sites toward the C terminus (Fig. 1C). In the same screening, a shorter isoform encoding a protein of 280 amino acids was isolated (Fig. 1A). The shorter isoform was named ILPIP␤ and originates from a methionine at amino acid 139 of the ILPIP␣ due to an insertion  vector. In vitro kinase assay was performed using ATF-2 as substrate. Kinase activity was quantitated by PhosphorImager analysis and is expressed as -fold induction relative to the basal level of phosphorylation of each JNK. UV was used as a positive control for each JNK activation (data not shown for JNK2 and JNK3). Western blots showing expression levels of JNK1, XIAP, ILPIP␣, or ILPIP␤ are shown for each experiment. Expression of ILPIP␣ or ILPIP␤ corresponds to proteins of 52 or 35 kDa, respectively. Asterisks indicate the presence of an unspecific band that appears when the anti-HA antibody is used.
NY-BR-96 exhibit significant homology to SPAK and other Ste20/SPS1-related kinases. The alignment was generated by the DNASTAR megalign program using the ClustalW method. C, schematic representation of the ILPIP␣ domains. The predicted amino acid sequence of ILPIP␣ encodes for a protein 418 amino acids long. ILPIP␣ has a "putative" protein kinase domain (amino acids 58 -369). The ATP-binding site is located at the N terminus of the protein. Several putative serine-, threonine-, or tyrosine-active and phosphorylation sites are shown.
in the 5Ј region of the gene (data not shown), which introduces two stop codons before the initial ILPIP␣ methionine. The existence of the ILPIP␣ and ␤ isoforms was confirmed by RT-PCR on RNA extracted from several different cell lines. Two PCR products were obtained, and sequence analysis matched the original sequences for ILPIP␣ and ␤. Northern blot analysis revealed that ILPIP is expressed in normal tissues as a single 2.4-kb transcript with higher levels of expression in muscle, liver, and heart (Fig. 2). ILPIP was also expressed as a 2.4-kb band in human embryonic kidney cells (293T) (Fig. 2).
ILPIP Moderately Activates JNK Family Members and Strongly Enhances the XIAP-mediated Activation of JNK1, JNK2, and JNK3-We have previously shown that XIAP-me-diated activation of JNK1 is necessary for its anti-apoptotic (1, 2). Thus we addressed the question whether expression of ILPIP alone or ILPIP and XIAP together would have an effect on MAPK activation. 293T cells were co-transfected with vectors encoding for JNK1, JNK2, or JNK3 in the absence or presence of increasing concentrations of plasmids encoding for ILPIP or XIAP, alone or in combination. Expression of MAPK proteins was quantified by densitometry after Western blot analysis and equivalent amounts were immunoprecipitated at 4°C for 2 h. Kinase activity was measured using ATF-2 as substrate. As previously reported XIAP expression increased JNK1 activity. Expression of ILPIP␣ or ILPIP␤ also resulted in a slight increase in JNK activity. However, co-expression of ILPIP␣ with XIAP, and to a lesser degree ILPIP␤, resulted in a marked synergistic activation of JNK1 that was comparable to UV activation (Fig. 3A). Cooperative activation of JNK2 and JNK3 was also observed, although to a lesser extent (Fig. 3B). Expression of ILPIP␣ or ILPIP␣ plus XIAP had no effect on p38 or ERK2 activation (data not shown). Therefore ILPIP acts synergistically with XIAP in specifically activating JNK family members.
ILPIP Moderately Protects against ICE-or Fas-induced Apoptosis and Significantly Potentiates the Anti-apoptotic Activity of XIAP-XIAP is known to protect against a variety of apoptotic stimuli. Because XIAP and ILPIP synergistically activated JNK1 and XIAP-mediated activation of JNK1 is important for the protection against apoptosis, we investigated whether ILPIP and XIAP also cooperatively protect against apoptosis. 293T cells were transfected with a plasmid encoding for ICE-␤-galactosidase in the presence of increasing amounts of XIAP or ILPIP DNA, alone or in combination. A green fluorescent protein-expressing vector was used as negative control. Each apoptotic assay was quantified both with X-gal staining of cells and Annexin V-PE-FACS analysis. The results obtained using the two different techniques were comparable, and therefore only the data from representative experiments performed with X-gal are shown. Expression of ILPIP was able to partially inhibit ICE-induced apoptosis (Fig. 4A), although this effect was not as pronounced as that of XIAP. However, ILPIP remarkably enhanced XIAP protection against ICE-induced apoptosis (up to 90% of viable cells), suggesting that ILPIP and XIAP cooperatively protect against ICE-induced apoptosis. Similar results were obtained in MCF7/Fas cells FIG. 6. XIAP and ILPIP interact with TAK1 and TRAF6. A, in vivo interaction of XIAP or ILPIP␣ with TAK1. Vectors encoding XIAP-FLAG or ILPIP␣-FLAG were co-transfected with wt TAK1-HA or AKT-HA in 293T cells. Cell extracts were immunoprecipitated using anti-FLAG antibody-conjugated beads. Co-precipitated TAK1 or AKT was detected by Western blot analysis with an anti-HA antibody. Cell extracts were directly subjected to immunoblot analysis to check for protein expression. B, interaction of XIAP or ILPIP␣ with TRAF6. Vectors encoding XIAP-FLAG or ILPIP␣-FLAG were co-transfected with wt TRAF6-Myc or deletion mutant TRAF6-Myc (TRAF6⌬-Myc) in 293T cells. Cell extracts were immunoprecipitated using anti-Myc antibody. Co-precipitated XIAP or ILPIP␣ was detected by Western blot analysis with an anti-FLAG antibody. Vectors encoding for XIAP-Myc and ILPIP␣-FLAG were also transfected in 293T cells to check for co-precipitation between XIAP and ILPIP␣. Cell extracts were immunoprecipitated using anti-Myc antibody, and co-precipitated ILPIP␣ was detected by Western blot analysis with an anti-FLAG antibody. C, interaction of XIAP or ILPIP␣ with TAB1 and TAB2. Vectors encoding XIAP-HA or ILPIP␣-HA were co-transfected with TAB1 or TAB2 in 293T cells. Cell extracts were subjected to immunoprecipitation with anti-TAB1 or anti-TAB2 antibody and co-precipitated XIAP or ILPIP␣ were detected by Western blot analysis using anti-HA antibody. Cell extracts were subjected to immunoblot analysis to check protein expression. D, in vitro interaction of XIAP with ILPIP␣. GST-XIAP or GST recombinant proteins were incubated with glutathione-conjugated beads, and in vitro translated ILPIP␣ protein was added. Co-precipitation of ILPIP␣ was detected by Western blot using an anti-HA antibody. Input proteins were detected by Western blot using anti-GST antibodies. when apoptosis was induced by treating the cells for 18 h with anti-Fas antibody (Fig. 4B). ILPIP␣ partially protected against Fas-induced apoptosis and significantly potentiated the protective effect of XIAP.
XIAP and ILPIP␣ Synergistic Activation of JNK1 Involves TAK1-We have previously reported that XIAP-mediated activation of JNK1 involves the MAP3 kinase TAK1. To investigate if the synergistic effect of ILPIP on XIAP-mediated JNK1 activation was also dependent on TAK1, 293T stably transfected cells expressing either LacZ control vector or catalytically inactive TAK1 (TAK1 (KW)) were transfected with vectors encoding for XIAP or ILPIP␣, alone or in combination, and JNK1 activation was determined. 293T cells stably expressing a catalytically inactive mutant of ASK1 (ASK1 (KM)) were also used as a control of specificity. XIAP-, ILPIP␣-, or XIAP/ILPIP␣mediated JNK1 activation was inhibited in the presence of TAK1 (KW) (Fig. 5B), whereas LacZ or ASK1 (KM) had no effect (Fig. 5, A and C). These data suggest that activation of JNK1 by ILPIP␣ or XIAP and the synergistic activation of JNK by ILPIP and XIAP are dependent upon TAK1.
XIAP and ILPIP Interact with TAK1 and TRAF6 -Because TAK1 appears to transmit the signal between the XIAP, ILPIP␣, and JNK1, we investigated the possibility that XIAP and ILPIP␣ would physically interact with TAK1. In an in vivo binding assay, vectors encoding XIAP-FLAG or ILPIP␣-FLAG were co-transfected with wt TAK1-HA or AKT-HA as a control of specificity in 293T cells. Cell extracts were immunoprecipitated using anti-FLAG antibody and analyzed by Western blot with an anti-HA antibody (Fig. 6A). Cell extracts were also directly subjected to immunoblot analysis to check for protein expression. TAK1 was found to co-precipitate with XIAP and ILPIP␣, suggesting that an interaction exists between these proteins. Results were confirmed by reverse co-precipitation using anti-HA antibody (data not shown).
It has been previously reported that TAK1, TRAF6, TAB1, and TAB2 associate in a complex (28 -30). Therefore we addressed the question whether XIAP and ILPIP␣ would also bind to TRAF6, TAB1, and TAB2. Vectors encoding TRAF6-Myc or a deletion mutant of TRAF6 (TRAF6⌬-Myc) were cotransfected with ILPIP␣-FLAG or XIAP-FLAG in 293T cells. Cell extracts were immunoprecipitated using anti-Myc antibody and analyzed by Western blot with an anti-FLAG antibody (Fig. 6B). TRAF6 was found to co-precipitate with XIAP and ILPIP␣ therefore suggesting that an interaction exists between these proteins. Interestingly, TRAF6⌬ also co-precipitate with XIAP and ILPIP␣ indicating that the interactions occur through the TRAF domain of TRAF6. Cell extracts were also directly subjected to immunoblot analysis to confirm expression of the respective proteins. Results were confirmed by reverse co-precipitation using anti-FLAG antibody (data not shown). As a control of specificity, TRAF2 was also assayed for coprecipitation with ILPIP and found to be negative (data not shown).
To determine if the adaptor molecules TAB1 and TAB2 were also part of this complex, vectors encoding TAB1 or TAB2 were co-transfected with XIAP-HA ILPIP␣-HA in 293T cells. Cell extracts were immunoprecipitated using anti-TAB1 or anti-TAB2 antibody and analyzed by Western blot with an anti-HA antibody. Co-precipitation of XIAP with TAB1 was confirmed as previously published (18). Interestingly, binding between XIAP and TAB2 was also detected. Surprisingly, ILPIP␣ did not interact either with TAB1 or TAB2 suggesting that ILPIP␣ interaction with TAB1 and TAB2 is achieved through XIAP and TAK1 (Fig. 6C).
Because ILPIP␣ was cloned as an XIAP-interacting protein in a yeast two-hybrid screening, we also addressed the question whether XIAP and ILPIP␣ were able to co-precipitate in an in vivo binding assay. Vectors encoding ILPIP␣-FLAG and XIAP-Myc were co-transfected in 293T cells, and cell extracts were immunoprecipitated using anti-Myc antibody and analyzed by Western blot with an anti-FLAG antibody. Surprisingly, no interaction was detected (Fig. 6B). Similar results were obtained using HA tagged-XIAP thus excluding the possibility that the lack of interaction observed could have been due to the low expression of XIAP-Myc. This result may be explained by the fact that the interaction between XIAP and ILPIP␣ is transient and unstable and thus undetectable with this method.
To determine if ILPIP could directly interact with XIAP, we used in vitro translated ILPIP␣ protein and recombinant GST-XIAP in in vitro binding assays. In these studies, GST protein was also used as negative control. XIAP-GST or GST proteins were incubated with glutathione-conjugated beads. ILPIP␣ was added to the reactions, and bound proteins were separated on SDS gels and analyzed by Western blots using an anti-HA antibody. ILPIP␣ was found to associate with XIAP but not with GST (Fig. 6D). Thus, these data are consistent with those observed in our yeast two-hybrid studies, which suggested that XIAP and ILPIP interact directly. The totality of these results supports the idea that XIAP, ILPIP␣, TAK1, TRAF6, TAB1, and TAB2 are likely to co-exist in a complex.
ILPIP␣ Does Not Affect XIAP Inhibition of Caspase Activation-XIAP has been reported to be a strong inhibitor of caspase activity (25,31). However, we have previously shown that XIAP protection against apoptosis requires JNK1 and TAK1 activation and does not affect the ability of XIAP to inhibit caspase activity. Because ILPIP␣ dramatically enhances the XIAP-mediated activation of JNKs, passes through TAK1, and significantly potentiates the anti-apoptotic activity of XIAP against ICE-or Fas-induced apoptosis, we investigated whether expression of ILPIP␣ would influence the ability of XIAP to inhibit caspase activity.
293T cells were transfected with control vector or vectors encoding XIAP or ILPIP␣, alone or in combination, and the effect on caspase activity was detected by measuring cleavage FIG. 7. ILPIP␣ does not affect XIAP inhibition of caspase activation. Effect of ILPIP␣ on XIAP inhibition of caspase activity. 293T cells were seeded in 100-mm dishes and transfected with plasmids encoding for XIAP or ILPIP␣ alone (3 g) or in combination. Empty vector was also transfected alone as a control (6 g). Cell extracts were prepared, and cytochrome c was added to induce proteolytic processing of pro-caspase-3. Caspase activity was measured by monitoring the release of AFC-DEVD-containing synthetic peptides. The last two columns represent the results obtained using cell extracts expressing XIAP alone or with ILPIP␣, diluted 10-fold. of short fluorogenic peptides (24,25) in a cell-free system where exogenously added cytochrome c induces proteolytic activation of caspase-9 and subsequently caspase-3 in cytosolic extracts. As expected, expression of XIAP strongly suppressed cytochrome c-induced caspase activation, whereas expression of ILPIP␣ alone had no effect. Co-expression of ILPIP␣ with XIAP did not appear to inhibit nor enhance the ability of XIAP to block caspase activation (Fig. 7). To rule out the possibility that the effect of ILPIP␣ on XIAP inhibition of caspases activity was too subtle to be detected when caspase activity is near completely inhibited by XIAP, we diluted the XIAP and XIAP/ ILPIP containing extracts 10-fold with the control extracts. Under these conditions cytochrome c-mediated activation of caspases is significantly increased above background, however, ILPIP␣ did not affect the inhibition mediated by XIAP. Combined, these data suggest that ILPIP␣ cooperatively enhances XIAP protection against apoptosis by a mechanism that is independent of inhibition of caspase activity. DISCUSSION XIAP has been shown to protect against a wide spectrum of apoptotic triggers. A suggested mechanism of IAP apoptotic suppression appears to be through direct caspase inhibition, in fact several of the human IAP family proteins have been reported to directly bind and inhibit specific members of the caspase family. Our findings suggest the existence of an alternative mechanism regulating the anti-apoptotic activity of XIAP that is separate from caspase inhibition and involves the TAB1/TAB2/TAK1/JNK1 signaling cascade. In an attempt to characterize the pathway connecting XIAP to JNK1 we performed a yeast two-hybrid screening and isolated ILPIP, a novel XIAP-interacting protein. The characterization of ILPIP functional properties highlighted some interesting features. First, ILPIP expression slightly activates JNK and strongly enhances JNK1 activation when co-expressed with XIAP. Activation of MAPKs has been reported to regulate the activities of many transcription factors and regulatory molecules and is required for the regulation of inflammatory responses, cell proliferation, and apoptosis (32). In particular, the involvement of the JNK family in apoptotic cell death has been most actively studied. JNK activation is observed in apoptosis induced by a variety of stimuli in different cell types. However, the consequence of its activation has been contradictory, resulting in protection from apoptosis in some cases and induction of apoptosis in others (32)(33)(34)(35)(36). Despite these apparent contradictions, there is a growing consensus that correlation between activation of JNK and protection or induction of apoptosis is stimuliand/or cell type-dependent (37)(38)(39)(40). With this in mind, we investigated whether XIAP/ILPIP synergistic activation of JNK1 was correlated with the ability of XIAP to protect against apoptosis. Interestingly, ILPIP moderately protects against ICE-or Fas-induced apoptosis and significantly potentiates the anti-apoptotic activity of XIAP therefore further supporting the idea that XIAP-mediated activation of JNK1 promotes cell survival.
XIAP and ILPIP synergistic activation of JNK1 involves TAK1 as demonstrated by inhibition of JNK1 activity using a catalytically inactive form of TAK1. It has been previously reported that XIAP participates in the BMP signaling pathway by binding with the BMP receptor, the adaptor molecule TAB1 (2,18), and with TAK1 (2). The findings, that ILPIP also co-precipitates with TAK1, that both XIAP and ILPIP bind to TRAF6, and that XIAP also binds to TAB2, further support the idea that these molecules behave in a functional complex.
Surprisingly, we were unable to detect association between XIAP and ILPIP in cells, suggesting that such an interaction may be transient. That could be explained by the possibility of ILPIP being a kinase, as predicted by the homology with serine/ threonine kinases and by the presence of a putative kinase domain. This possibility is currently under investigation in our laboratory. A direct interaction between XIAP and ILPIP␣ was demonstrated in an in vitro binding assay thus supporting the original interaction showed by the yeast two-hybrid system.
Finally, expression of ILPIP did not affect XIAP inhibition of caspase activation further supporting the idea that XIAP protection against apoptosis is achieved by two separate mechanisms: one requiring JNK1 activation and a second involving caspase inhibition. The interaction between XIAP, ILPIP, and TRAF6 may also suggest that XIAP might be involved in the IL-1 inflammatory response, which TRAF6 has been previously shown to regulate (41).
XIAP has been reported to interact with a few other proteins all of which act as negative regulators. Among these is SMAC/ DIABLO, which is a nuclear-encoded, mitochondrially localized protein that is released in response to apoptotic stimuli and acts as a negative regulator of XIAP anti-apoptotic function (21,22). Similarly Omi/HtrA2, a serine protease localized in the mitochondria, inhibits the protective effect of XIAP by binding to its BIR-3 domain (42). A third negative regulator of XIAP is XAF1, which is thought to exert its effect through sequestering XIAP in the nucleus (43). Importantly, we show here that ILPIP is the first protein able to potentiate the anti-apoptotic effect of XIAP instead of antagonizing it.
Taken together our results describe a novel XIAP-interacting protein that acts as a co-factor enhancing XIAP-mediated activation of JNK1 and the caspase-independent protection of XIAP against apoptosis.