J Biol Chem, Vol. 275, Issue 13, 9882-9889, March 31, 2000

Regulation of the NF-B Activation Pathway by Isolated Domains of FIP3/IKK, a Component of the IB- Kinase Complex^*

Jianjiang Ye, Xueping Xie, Leonid Tarassishin, and Marshall S. Horwitz^§

From the Albert Einstein College of Medicine, Department of Microbiology and Immunology, Bronx, New York 10461

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

FIP3, isolated as a type 2 adenovirus E3-14.7-kDa interacting protein, is an essential component of the multimeric IB- kinase (IKK) complex and has been shown to interact with various components (Fas receptor-interacting protein, NF-B-inducing kinase, IKK) of the NF-B activation pathway. FIP3 has also been shown to repress basal and tumor necrosis factor (TNF) -induced NF-B activity as well as to induce cell death when overexpressed. The adenovirus E3-14.7-kDa protein (E3-14.7K) is an inhibitor of TNF-induced cell death. In the current study, we generated deletion mutants to map the domains of FIP3, which are responsible for its various functions. The NF-B inhibitory activity and the E3-14.7K binding domains were mapped at the carboxyl half of the FIP3 protein. We also found that the carboxyl-terminal half of FIP3 blocked TNF-induced IB- phosphorylation and subsequent degradation, which suggests that the stabilization of the cytoplasmic inhibitor of NF-B underlies the FIP3 inhibition of NF-B activity. The amino-terminal 119 amino acids were responsible for the FIP3-IKK and FIP3-IKK interaction, and the middle of the protein (amino acids 201-300) appeared to be both the FIP3 self-association domain as well as the FIP3-Fas receptor-interacting protein interaction domain. Thus, FIP3 might serve as a scaffold protein to organize the various components of the IB- kinase complex. Whereas the full-length protein is required for efficient cell death, the amino-terminal 200 amino acids are sufficient to cause rounding and detachment of the cells from the monolayer.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

NF-B¹ is a transcription factor important for modulations of inflammatory responses, cell proliferation, and apoptosis (1-4). The most common form of NF-B is the heterodimer consisting of RelA(p65) and p50 subunits (5, 6). In most nonstimulated cell types, NF-B is retained in the cytoplasmic compartment through its association with a group of inhibitory proteins known as IBs (7-9). Three isoforms of IB have been described, namely, IB-, -, and - (4, 7, 10, 11). A variety of stimuli could lead to the phosphorylation of IBs at two serine residues located at the amino terminus of the protein (Ser-32 and Ser-36 for IB-), including the pro-inflammatory cytokines tumor necrosis factor  (TNF) and interleukin 1, bacterial lipopolysaccharide, protein kinase inhibitors and viral products (12-17). Upon phosphorylation, IBs are multiubiquitinated and targeted to the 26 S proteasome for degradation (18-21). This releases NF-B to translocate into the nucleus where it activates transcription upon binding to its target DNA (22).

A large multicomponent kinase complex responsible for phosphorylation of IB- has been identified (23), and two kinases, IKK and IKK (also known as IKK1 and IKK2), have been cloned and characterized (24-28). IKK and IKK both phosphorylate IB- in vitro (24, 25). Knockout studies of IKK and IKK in mice have revealed their critical and noninterchangeable roles in regulation of NF-B activity (29-34). Recently, we and others using different approaches have discovered another component of the IKK complex, FIP3, also known as NF-B essential modulator (NEMO), IKK and IB kinase-associated protein 1 (35-38). We cloned FIP3, a 14.7K-interacting protein, from a human cDNA library, after it was identified in our yeast two-hybrid studies designed to look for cellular proteins that could interact with the adenovirus E3-14.7K protein (35). The E3-14.7-K protein is an inhibitor of TNF-induced apoptosis (39). FIP3 has been shown to be essential for the activation of the NF-B pathway, and mutations of FIP3 have been identified as the reason for the nonresponsiveness of two cell lines to NF-B-inducing stimuli (36). Antisense studies also showed that FIP3 plays an essential regulatory role in the IKK complex (37). Despite its importance in activation of the IB- kinase, it also plays an inhibitory role in the NF-B pathway (35). When expression of FIP3 is increased by transient transfection, it down-regulates both basal and TNF-induced NF-B activity (35). FIP3 has been shown to interact with IKK, RIP, and NF-B-inducing kinase and to form homotypic oligomers (35, 36); however, the structural basis for these interactions and the regions of FIP3 that are responsible for these interactions are not well characterized.

TNF treatment of cells also leads to activation of caspases through the interaction of TNFR1 and TRADD, FADD complex (40-45), or TNFR1, RIP and the adaptor protein RAIDD/CRADD (46-48), and thus results in activation of an apoptotic pathway. FIP3 is also an apoptosis-inducing protein when overexpressed (35), but the mechanism by which FIP3 is integrated into the TNF apoptotic pathway has not been elucidated.

In this manuscript we present the results of deletion studies of FIP3, which assign various FIP3 functions to different domains of the protein. We mapped the IKK- and IKK-interacting activity to the amino-terminal 119 amino acids. The FIP3 self-association and the RIP interaction domains were located in a 100-amino acid segment in the middle of the protein. We also assigned the 14.7K-interacting and NF-B down-regulation function of FIP3 within the carboxyl-terminal half of the protein. The primary cell rounding function that results in detachment of the cells from the monolayer was mapped to the amino-terminal 200 amino acids of the FIP3 protein, whereas efficient cell killing required the full-length FIP3. Furthermore, we demonstrated that FIP3 prevented TNF-induced IB- phosphorylation and degradation and located this activity to the same domain that is responsible for down-regulation of NF-B.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Lines and Reagents-- 293 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, penicillin (50 units/ml), and streptomycin (50 µg/ml). Mouse anti-T7 monoclonal antibody and HRP-conjugated anti-T7 antibody were purchased from Novagen. Mouse anti-FLAG and antipolyhistidine antibodies were from Sigma. Rabbit anti-IB- antibody, rabbit anti-IKK antibody, and mouse anti--tubulin antibody were purchased from Santa Cruz Biotechnology.

Plasmids-- pcDNA3-FLAG-14.7K, pcDNA3-T7-FIP3 full-length, and a mutant with deletion of the amino-terminal 179 amino acids were constructed as described (35). All other FIP3 mutants were made by a polymerase chain reaction-based method using Pfu polymerase. The mutagenesis scheme is shown in Fig. 1A, and mutant designations are described in the legend to Fig. 1A. A BamHI site was introduced into all 5'-primers, and an XhoI site was introduced into all 3'-primers to facilitate cloning. Primer sequences are available upon request. All mutants were cloned into the BamHI and XhoI sites of the pcDNA3-T7 vector in-frame with the 5'-T7 tag. The fidelity of the vector-insert junction was confirmed by sequencing. The constructs expressing FLAG-tagged IKK (pFLAG-CMV-IKK) and polyhistidine-tagged IKK (pHIS-CMV-IKK) were kindly provided by Jun Li (Boehringer Ingelheim). pFLAG-CMV-RIP, which expresses FLAG-tagged RIP protein, was generously provided by David Wallach (Weizmann Institute, Israel). The NF-B-dependent luciferase reporter construct (pIg-Luc) has been previously reported (35, 49). pGEX-IB--(1-53) and pGEX-IB- mutant-(1-53), which express the wild type and mutant form (Ser-32 and Ser-36 changed to Ala) of the amino-terminal 53-amino acid fragment of IB- as GST fusions were kindly provided by Sergey Trushin (Mayo Clinic, Rochester, MN). pGreen-Lantern-1, which expresses the green fluorescent protein under the control of the CMV promoter, was purchased from Life Technologies, Inc., and pCH110, which expresses -galactosidase under the control of the SV40 promoter, was from Amersham Pharmacia Biotech.

Transfection and Luciferase Reporter Assay-- 5 × 10⁵ 293 cells grown on 6-well plates were transfected with a total amount of 1.0 µg of DNA and 8 µl of LipofectAMINE (Life Technologies, Inc.) according to a manufacturer-provided protocol. Luciferase activity was measured with a luciferase assay kit (Roche Molecular Biochemicals) using Monolight 2010 of The Analytical Luminescence Laboratory. Relative luciferase activity was normalized using -galactosidase activity, which was expressed from the co-transfected pCH110.

Morphological Studies of Transfected Cells-- 293 cells were transfected with test plasmids and pGreen-Lantern-1, which served as a co-transfection marker facilitating the identification of transfected cells and monitoring transfection efficiency (50). Forty-eight hours after transfection cells were examined and photographed using a fluorescent microscope with a fluorescein isothiocyanate filter.

Quantification of Apoptosis-- The amount of apoptosis was measured by a quantitative enzyme-linked immunosorbent assay kit (Roche Molecular Biochemicals) using mouse antihistone and anti-DNA antibodies to detect mono- and oligonucleosomes. The fold increase of cell death in the experimental samples over control is taken as the apoptosis enrichment factor, which serves as an index of cell death.

Immunoprecipitations-- Twenty-four hours post-transfection, 293 cells were lysed with 1.5% Nonidet-P40, 0.25 M NaCl, 50 mM HEPES (pH 7.4), and 1x complete proteinase inhibitor mixture (Roche Molecular Biochemicals). Lysates were precleared with normal mouse IgG and zysorbin (Zymed Laboratories Inc.) and then immunoprecipitated with mouse monoclonal anti-FLAG antibody or mouse monoclonal anti-polyhistidine antibody and zysorbin. Immunoprecipitates were washed three times with lysis buffer, subjected to 12% SDS-polyacrylamide gel electrophoresis, and transferred to nitrocellulose membranes. The membranes were blotted with HRP-conjugated anti-T7 antibody or with nonconjugated anti-T7 antibody and HRP-conjugated rabbit anti-mouse IgG antibody. Western blots were developed with the enhanced chemiluminescence system (NEN Life Science Products).

In Vitro Kinase Assay-- 2 × 10⁶ 293 cells grown on 10-cm plates were transfected with 1.5 µg of FLAG-IKK and 3.0 µg of FIP3 by the LipofectAMINE method (Life Technologies, Inc.) according to the manufacturer's protocol. Twenty-four hours after transfection, cell were treated with TNF (20 ng/ml) for 15 min and then lysed. The cell lysates were immunoprecipitated with anti-FLAG monoclonal antibody and the immunoprecipitates were washed extensively with lysis buffer followed by kinase buffer (20 mM HEPES, pH 7.5; 20 mM -glycerophosphate; 10 mM MgCl₂; 100 µM Na₃VO₄; 2 mM dithiothreitol; and 1× complete protease inhibitor mixture (Roche Molecular Biochemicals)). The immunoprecipitates were then incubated for 20 min at 30 °C with 20 mM ATP, 2 µCi of [-³²P]ATP and 2 µg of GST-IB--(1-53) or GST-IB--(1-53) mutant with substitutions of serines at position 32 and 36 by alanines as substrates in the kinase buffer. The reaction mixtures were resolved by 10% SDS-polyacrylamide gel electrophoresis, and the gel was dried and developed by autoradiography. In parallel, the gel was subjected to Western blot analysis using anti-IKK antibody. The intensity of specific bands was quantified by densitometry.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

FIP3 Deletion Mutants Are Expressed and Stable in 293 Cells-- The 293 human embryonic kidney cell line was transfected with various FIP3 deletion mutants or vector alone as a negative control and FIP3 wild type as a positive control. One day post-transfection, the expression of FIP3 and various mutants was analyzed by Western blot analysis. As shown in Fig. 1B, all FIP3 mutants were expressed in 293 cells with comparable molar levels. The mobility of some of them was irregular, probably because of differential post-translational modifications.

View larger version (33K): [in this window] [in a new window]   Fig. 1.   A, scheme of FIP3 mutagenesis. FL, full-length FIP3; ND100, ND179, ND200 and ND300 are FIP3 mutants with deletions of amino-terminal 100, 179, 200, and 300 amino acids, respectively; CD100, CD200, CD240, and CD300 are FIP3 deletions of carboxyl-terminal 100, 200, 240, and 300 amino acid sequences, respectively; ZFD is the FIP3 mutant deleted of the carboxyl-terminal zinc-finger domain (amino acids 397-419); LZ1D, LZ2D, and LZ3D are FIP3 mutants deleted of the first leucine-zipper domain (amino acids 128-149), the second leucine-zipper domain (amino acids 260-281), and the third leucine-zipper domain (amino acids 322-343) respectively. B, expression of FIP3 mutants. 5 × 105 293 cells were transfected with 0.2 µg of pcDNA3-T7-FIP3 (full-length) and mutants. The pGreen-lantern-1 (0.2 µg) was included in every transfection in this experiment and all subsequent experiments to facilitate monitoring transfection efficiency. The total DNA amount in all transfections was brought up to 1.0 µg/well with control plasmid pcDNA3-T7. Twenty-four hours after transfection, cells were lysed with 1× SDS sample buffer, and the lysates were analyzed by Western blot using anti-T7 antibody. M, protein molecular weight marker. All DNAs were cloned into the pcDNA3-T7 vector. Lane 1, empty vector control; lane 2, FIP3 full-length; lane 3, ND100; lane 4, ND179; lane 5, ND200; lane 6, ND300; lane 7, CD100; lane 8, CD200; lane 9, CD240; lane 10, CD300; lane 11, ZFD; lane 12, LZ1D; lane 13, LZ2D; lane 14, LZ3D.

The Carboxyl-terminal Half of FIP3 Is Required for Its Interaction with Adenovirus 2 E3-14.7-kDa Protein-- FIP3 was initially cloned by its interaction with the adenovirus E3-14.7K protein in the yeast two-hybrid system (35). Co-immunoprecipitation studies were used to define the domains in FIP3, which mediate its interaction with the viral protein. Wild type FIP3 and amino-terminal deletion mutants ND100 and ND179 were all co-immunoprecipitated with the antibody against the FLAG-tag, which is on the 14.7K (Fig. 2A, lanes 2-4). ND200 was co-immunoprecipitated at a much lower efficiency (Fig. 2A, lane 5), whereas ND300 was not precipitated at all (lane 6). This suggested that amino acids 180-200 are required for full-scale FIP3-14.7K association and amino acids 201-300 are essential for the interaction to occur. The carboxyl-terminal deletion mutants did not co-precipitate with 14.7K (Fig. 2A, lanes 7-10), arguing the last 100 amino acids are necessary for the interaction. We further tested the FIP3 mutants with deletions in the three leucine-zipper domains and the carboxyl-terminal zinc-finger domain in this interaction assay. Deletions of the carboxyl-terminal zinc-finger domain (amino acids 397-419) or the second leucine-zipper domain (amino acids 260-281) compromised the FIP3-14.7K interaction (Fig. 2A, lanes 11 and 13), whereas the first and the third leucine-zipper domains are not important (Fig. 2A, lanes 12 and 14). FLAG-tagged 14.7K protein was expressed well in all transfections (Fig. 2B), and FIP3 mutants were all expressed at levels comparable to Fig. 1B. From these studies, we concluded that the carboxyl-terminal half of the FIP3 protein is involved in the interaction between FIP3 and 14.7K, but the third leucine-zipper domain located in this region is not important.

View larger version (43K): [in this window] [in a new window]   Fig. 2.   Carboxyl-terminal regions of FIP3 including the zinc-finger and leucine-zipper 2 are required for its association with the adenovirus protein E3-14.7-kDa. 5 × 105 293 cells were transfected with 0.2 µg of T7-tagged FIP3 or mutants and 0.6 µg of FLAG-tagged 14.7 kDa plus 0.2 µg of pGreen-lantern 1 as a co-transfection marker to monitor transfection efficiency. Twenty-four hours after transfection, cells were lysed with buffer containing Nonidet P-40. A, the lysates were immunoprecipitated with anti-FLAG antibody and blotted with anti-T7 antibody to test the interaction between various FIP3 mutants and E3-14.7K. M, protein molecular weight marker; H, heavy chain of immunoglobulin; L, light chain of immunoglobulin; Lanes 1-14 indicate the corresponding control, wild type, or FIP3 mutants in each lane as in Fig. 1B. B, cell lysates were also blotted with anti-FLAG antibody to check the expression of E3-14.7K protein.

The FIP3 Self-association and FIP3-RIP Interaction Domains Map to the Region between Amino Acid 201 and 300-- FIP3 is able to form dimers or trimers, and FIP3 also interacts with RIP (35-37). We used co-immunoprecipitation assays to map the region in FIP3 through which the FIP3-FIP3 association or FIP3-RIP interaction occurred. Wild type FIP3 associated with itself intermolecularly (Fig. 3A, lane 2) and interacted with RIP (Fig. 3C, lane 2) as expected. Amino-terminal deletions of 100, 179, or 200 amino acids did not affect the FIP3-FIP3 association (Fig. 3A, lanes 3-5), nor did they affect the FIP3-RIP interaction (Fig. 3C, lanes 3-5). Further deletion of 100 more amino acids abolished the FIP3-FIP3 and FIP3-RIP interaction (Fig. 3, A and C, lane 6). Carboxyl-terminal deletion of 100 amino acids did not change the FIP3-FIP3 or FIP3-RIP interaction either (Fig. 3, A and C, lane 7), whereas deletion of the carboxyl-terminal 200 amino acids or more abolished these interactions (Fig. 3, A and C, lanes 8-10). These data suggested that the 100-amino acid segment in the middle of the FIP3 protein is required for the FIP3-FIP3 or FIP3-RIP interaction. The FIP3 mutant with a deletion of the second leucine-zipper domain, which falls in this region (260-281), showed significantly weaker interaction with wild-type FIP3 (Fig. 3A, lane 13) but did not affect FIP3-RIP interaction (Fig. 3C, lane 13). Deletions of other leucine-zipper domains or the zinc-finger domain, which are outside of this region, did not affect the FIP3-FIP3 or FIP3-RIP interaction (Fig. 3, A and C, lanes 11, 12, and 14). The expression of FLAG-tagged wild type FIP3 and RIP are shown in Fig. 3, B and D, respectively, as controls. Also the expression of FIP3 mutants was monitored in these interaction studies by Western blot, and comparable expression levels were observed (data not shown). These co-immunoprecipitation analyses suggested that the region located between amino acids 201 and 300 is necessary for FIP3 self-association and for the FIP3-RIP interaction, and the second leucine-zipper domain, which covers the amino acids 260-281, is important but not absolutely required for FIP3 self-association.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   A domain of one hundred amino acids (201-300) in the middle of FIP3 protein is required for FIP3 self-association and FIP3-RIP interaction. 5 × 105 293 cells were transfected with 0.2 µg of T7-tagged FIP3 mutants together with 0.2 µg of FLAG-tagged wild type FIP3 (A and B) or with 0.2 µg of T7-tagged FIP3 mutants together with 0.1 µg of FLAG-tagged RIP and 0.1 µg of p35 (C and D). The baculovirus inhibitor of apoptosis p35 was used to block RIP-induced cell death. Total amount of DNA for each transfection was brought up to 1.0 µg with control plasmid. Twenty-four hours after transfection, cells were lysed; the lysates were immunoprecipitated with anti-FLAG antibody and blotted with anti-T7 antibody (A) or HRP-conjugated anti-T7 antibody (C) to test the association of various FIP3 mutants with wild type FIP3 or with RIP. Cell lysates were also blotted with anti-FLAG antibody to check the expression of FLAG-FIP3 (B) or FLAG-RIP (D). ns, nonspecific bands; other letter or number designations are the same as those in Fig. 2.

The Amino-terminal 119 Amino Acids of FIP3 Are Necessary and Sufficient for Its Interaction with IKK and IKK-- FIP3 is a key component of the IB- kinase complex; it interacts with IKK in vivo (36) and with IKK both in vivo and in vitro (36-38). We mapped the FIP3 interaction with each of the IKK proteins to study whether FIP3 interacts with IKK or IKK through the same or distinct regions. We used polyhistidine-tagged IKK and FLAG-tagged IKK in the following co-immunoprecipitation studies. Wild type FIP3 interacted with IKK (Fig. 4A, lane 2) and IKK (Fig. 4C, lane 2) as expected and is shown as positive controls. Amino-terminal deletions of 100 amino acids or more abolished the association between FIP3 and IKK or IKK (Fig. 4, A and C, lanes 3-6), indicating that the first 100 amino acids are required for FIP3 to interact with IKK or IKK. Carboxyl-terminal deletions of 100, 200, 240, or 300 amino acids did not affect the association of FIP3 and IKK or IKK significantly (Fig. 4, A and C, lanes 7-10). These data suggested that these regions are not important for FIP3 to interact with IKK or IKK and the amino-terminal 119 amino acids are necessary and sufficient for the FIP3-IKK or FIP3-IKK interaction to occur properly. As anticipated from these results, none of the deletions in the leucine-zipper domains or in the zinc-finger domain compromised the FIP3-IKK or FIP3-IKK interaction (Fig. 4, A and C, lanes 11-14). The expression of polyhistidine-tagged IKK and FLAG-tagged IKK is shown in Fig. 4, B and D, as controls.

View larger version (34K): [in this window] [in a new window]   Fig. 4.   The amino-terminal 119 amino acids of the FIP3 protein are required for both FIP3-IKK and FIP3-IKK interactions. 293 cells were similarly transfected with 0.2 µg of T7-tagged FIP3 mutants and 0.2 µg of polyhistidine-tagged IKK (A and B) or FLAG-tagged IKK (C and D). Twenty-four hours after transfection, cells were lysed, and the cell lysates were immunoprecipitated with antipolyhistidine antibody or anti-FLAG antibody. The immunoprecipitates were analyzed by Western blot with HRP-conjugated anti-T7 tag antibody (A) or nonconjugated anti-T7 antibody (C) to test the interaction of FIP3 mutants with IKK or IKK, respectively. Cell lysates were also immunoblotted with antipolyhistidine antibody (B) or anti-FLAG antibody (D) to check the expression of HIS-IKK or FLAG-IKK. Letter and number designations are as described in previous figures. Arrows highlight interacting protein bands in C, lanes 8 and 9.

The Carboxyl Half of FIP3 Protein Is Responsible for the Down-regulation of TNF-induced NF-B Activity-- FIP3 is an essential component of the NF-B activation pathway (36); however, when FIP3 is overexpressed in cells, it also causes down-regulation of both basal or TNF-induced NF-B activity (Ref. 35 and Fig. 5, A-C, upper panels, column 2). When the amino-terminal 100, 179, or 200 amino acids were deleted, FIP3 retained its activity in down-regulation of NF-B (Fig. 5, A-C, upper panels, columns 3-5), implying that the amino-terminal half of the protein is dispensable for its inhibitory effect on NF-B. In agreement with this, deletion of the first leucine-zipper domain (amino acids 128-149) did not affect the inhibitory function of FIP3 (Fig. 5, A--C, upper panels, column 12). When 300 amino acids were removed from the amino terminus of the protein, FIP3 lost most of its capacity to block NF-B activation (Fig. 5, A-C, upper panels, column 6), suggesting that the region between amino acids 201 and 300 is crucial for this activity. When FIP3 was deleted of 100 amino acids from the carboxyl terminus, it also lost a significant portion of its activity (Fig. 5, A-C, upper panels, column 7), suggesting that the carboxyl 100 amino acids are also important for FIP3 to be fully functional. As expected, deletion of 200, 240, or 300 amino acids from the carboxyl terminus all compromised the inhibitory role of FIP3 in the NF-B activation pathway (Fig. 5, A-C, upper panels, columns 8-10). Deletions of the FIP3 carboxyl-terminal zinc-finger domain (amino acids 397-416, column 11), the second or the third leucine-zipper domain (amino acids 260-281, column 12; 322-343, column 14) were less inhibitory than the amino-terminal leucine-zipper domain deletion (column 12) on NF-B activity. These suggested that these domains are all required for FIP3 to elicit its effect fully, but neither of them is the sole determinant of the FIP3 function. The inhibitory effect on NF-B activity is dose-dependent, as we observed more inhibition when higher amount of FIP3 or its mutants were used (Fig. 5, A-C, upper panels). The corresponding expression of FIP3 and its mutants was shown on the lower panels in Fig. 5, A-C, by Western blot. From the above analyses we concluded that the carboxyl half of the FIP3 protein (amino acids 201-419) was responsible for its function as an inhibitory component of the NF-B modulation pathway.

View larger version (34K): [in this window] [in a new window]   Fig. 5.   The carboxyl-terminal half of the FIP3 protein down-regulates NF-B activity. 5 × 105 293 cells were transfected with 20 (A), 100 (B), and 200 ng (C) of FIP3 or mutant plasmids together with 0.1 µg of pCH110 and 0.05 µg of pIg-Luc. TNF (10 ng/ml) was added at 18 h post-transfection. Cells were harvested 24 h post-transfection and analyzed for luciferase activity and -galactosidase activity. The normalized relative luciferase units are shown in the upper panels (A-C). The expression of the FIP3 and its mutants was examined by Western blot using anti-T7 antibody and is shown in the lower panel of A-C.

The Amino-terminal Domain of FIP3 Is Not Required to Block TNF-induced IB- Degradation-- FIP3 augmented the kinase activity of IKK when overexpressed and assayed in cytoplasmic extracts (Ref. 38 and Fig. 6C); however, the intracellular NF-B luciferase reporter activity was inhibited (Ref. 35 and Fig. 5, upper panels, column 2). Thus, there appears to be an uncoupling of IKK kinase activity and NF-B activation. TNF treatment induced rapid degradation of IB- within 30 min, and IB- was resynthesized within 1 h after treatment (Fig. 6A, lane 1). When the effect of FIP3 overexpression on TNF-induced IB- degradation was examined, we found that this process was partially blocked in the presence of intact FIP3 (Fig. 6A, lane 2) and even more effectively inhibited by the amino-terminal 100-amino acid deletion (ND100) of FIP3 (Fig. 6A, lane 3). That the protection was only partial in both of these examples might be explained by the fact that TNF stimulated all the cells, but FIP3 could only protect IB- in the fraction of the cells that were transfected. It was noted that after 10 min of TNF treatment all FIP3 mutants attenuated IB- degradation to various extents. However, the residual amounts of IB- were least when ND300 (lane 6) and CD300 (lane 10) were added. The different effects of these mutants were much more evident after 30 min of TNF treatment. Deletion of 179 and 200 amino acids from the amino terminus reduced the protective effect of FIP3 moderately at 30 min after the addition of TNF (Fig. 6A, lanes 4 and 5). However, further deletion of 100 amino acid from the amino terminus made FIP3 incapable of protecting IB- from TNF-induced degradation at 30 min (Fig. 6A, lane 6). These results suggest that the amino-terminal 200 amino acids of FIP3 are less important for the inhibition of IB- degradation but imply the importance of the region from 200-300 amino acids. When FIP3 was deleted of 100, 200, 240, or 300 amino acids from the carboxyl terminus, it was more compromised in its activity to block IB- degradation (Fig. 6A, lanes 7-10). Deletion of the first leucine-zipper (amino acids 128-149, lane 12) affected the activity of FIP3 similarly to ND100. Interestingly, FIP3 retained some activity in this assay even after deletion of the zinc-finger (lane 11) and the second leucine-zipper (lane 13), which are located within the carboxyl half of the FIP3 protein. These results suggested that all but the amino-terminal 200 amino acids of the FIP3 protein are necessary for inhibition of TNF-induced IB- degradation, but some small regions in the carboxyl part of the FIP3 protein are not absolutely required. FIP3 or its mutants did not affect the expression level of the housekeeping gene -tubulin, which is shown in Fig. 6B as a control. It appeared that the mobility of the protected IB- (Fig. 6A, 10 and 30 min) was not changed compared with IB- in nonstimulated controls (Fig. 6A, 0 min). This suggests that these IB- molecules were not ubiquitinated, as ubiquitination would result in considerable retardation of protein mobility. However, it was not clear whether they were phosphorylated, as this post-translational modification only results in slight mobility shifts. When these IB- molecules were examined with an antibody to phosphorylated IB- to see whether they were modified after TNF stimulation in the presence of FIP3, no phosphorylation was detected (data not shown). This suggested that overexpression of FIP3 intracellularly protected IB- from TNF-induced phosphorylation. In contrast, when assayed in vitro using GST-IB--(1-53) as substrate, the IKK kinase activity was increased 4-fold by FIP3 overexpression (Fig. 6C, I). The activity was specific for the serines at position 32 and 36, as a mutant of IB- with substitutions of these serine residues by alanines were not phosphorylated (Fig. 6C, II). After normalization of the kinase activity against IKK protein level (Fig. 6C, III), we still observed a significant increase (Fig. 6C, IV). Thus, we concluded that FIP3 induces the uncoupling of the activated IKK and its substrate IB- in vivo.

View larger version (34K): [in this window] [in a new window]   Fig. 6.   The amino-terminal 200 amino acids of the FIP3 protein are not required to prevent TNF-induced IB- degradation. A, 5 × 105 293 cells were transfected with 0.3 µg of FIP3 or mutant plasmids. Cells were lysed 24 h post-transfection. Before lysis, cells were either treated with TNF (10 ng/ml) for 10, 30, and 60 min or used as untreated controls. The lysates were then analyzed by Western blot using anti-IB- antibody. Lane numbers have the same designations as in previous figures. B, cell lysates were blotted with anti--tubulin antibody to monitor the expression of this housekeeping gene. C, 293 cells were transfected with FIP3 and IKK, and the cell lysates were subjected to in vitro kinase assay according to the protocol described under "Experimental Procedures" using GST-IB--(1-53) wild type (I) and mutant (II) as substrates. The expression of IKK was examined by Western blot using anti-IKK antibody and is shown in C, III. The numbers underneath C, I and C, III are the relative levels of kinase activity and IKK expression, respectively. The normalized kinase activity was shown in C, IV.

Full-length FIP3 Is Required for Cell Death, but the Cell Rounding Activity of FIP3 Could Be Mapped to the Amino-terminal Half of the Protein-- We showed previously that FIP3 when overexpressed caused apoptotic cell death (35) and proceeded to define the death-inducing domain of the FIP3 protein in the current study. 293 cells were either mock-transfected or transfected with FIP3 wild type or mutants together with pGreen-Lantern 1 as a co-transfection marker. Wild-type FIP3 induced a considerable amount of cell rounding by 24-48 h post-transfection. By 48 h we observed a significant decrease of GFP-expressing cells attached to the monolayer (Ref. 35 and Fig. 7I, B), whereas control plasmid-transfected cells were flat and polygonal with protruding processes characteristic of normal 293 cells (Fig. 7I, A). When the amino-terminal 200 amino acids were deleted, FIP3 lost its activity to induce cell rounding and to detach cells from the monolayer (Fig. 7I, C). Deletions of the carboxyl-terminal 200 amino acids did not abrogate the cell-rounding activity (Fig. 7I, D). These results suggested that the amino-terminal 200 amino acids are required for the FIP3 protein to induce cell rounding and detachment. We then quantified apoptosis caused by FIP3 and some of the mutants using an enzyme-linked immunosorbent assay-based method that measures the amount of free nucleosomes released during apoptosis. We found that the wild type FIP3 elicited more that 4-fold enrichment of free nucleosome production as compared with an empty vector control (Fig. 7II). However, efficient apoptosis required both amino and carboxyl termini of FIP3 protein, as deletions from either end rendered the FIP3 protein ineffective in causing cell death (Fig. 7II). Interestingly, the mutant CD200 that caused efficient cell rounding did not induce significant amount of apoptosis (Fig. 7II). This argues that the cell rounding and apoptosis are separate events, and detachment from the monolayer does not necessarily lead to cell death.

View larger version (62K): [in this window] [in a new window]   Fig. 7.   Full-length FIP3 protein is required for efficient cell killing, but the primary cell rounding activity of FIP3 is mapped at the amino-terminal half of the protein. I, 5 × 105 293 cells were transfected with 0.8 µg of pcDNA3-T7-FIP3 and mutants ND200 and CD200 together with 0.2 µg of pGreen-lantern-1 as a co-transfection marker, which helped discriminate between transfected and nontransfected cells. Forty-eight hours post-transfection, cells were examined under the fluorescent microscope. A, vector control; B, FIP3 wild type; C, ND200; D, CD200. II, cell death was quantified by enzyme-linked immunosorbent assay according to the method described under "Experimental Procedures."

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

FIP3 is an essential component of the IB kinase complex but does not itself have kinase activity as an isolated protein (36, 38). Because FIP3 interacts with multiple components of the NF-B activation pathway, e.g. RIP, NF-B-inducing kinase, IKK, and IKK (35, 36, 38), it appears that FIP3 might serve as a scaffolding protein to organize the formation of the multisubunit IKK complex. Some previous studies have resulted in interesting but conflicting observations on whether the amino or carboxyl domain of the FIP3 protein is involved in IKK complex binding (37, 38, 51). The current study defines the domains in FIP3 responsible for its association with IKK as well as other components of the NF-B signal transduction pathway. We found that the amino-terminal 119-amino acid region is responsible for the association of FIP3 with IKK and IKK of the IB kinase complex, in agreement with the observations of two other groups (38, 51). Although FIP3 and IKK did not interact in vitro (36, 38), FIP3 and IKK interact quite strongly in vivo when both are overexpressed after transient transfection (29, 36). This suggests that their interaction might be direct, because an indirect interaction would require very high level expression of an endogenous bridging molecule. IKK, the most likely bridging molecule thus far proposed, is dispensable for the FIP3-IKK interaction to occur as demonstrated by recent IKK knock-out studies (29, 33). Through the amino-terminal domain, FIP3 might interact with IKK and IKK homo- or heterodimers, depending on the abundance of each oligomer in the cells.

FIP3 itself forms dimers and trimers (37).² This interaction provides an additional possibility for FIP3 to interact with and organize multiple components of the pathway and also provides another level of regulation. We mapped the FIP3 self-association domain in the middle of the protein from amino acids 201 to 300. The fact that FIP3 forms homotypic trimers implies that there must be more than one FIP3-FIP3 interaction domain; thus, there might be two subdomains in the 201-300 region, which are both capable of mediating the FIP3-FIP3 association or there is an outside domain not identified by our studies. The same region (201-300) is also required for RIP binding; however, the second leucine-zipper domain that is located in this region is important for the FIP3-FIP3 interaction but not for the FIP3-RIP interaction. This suggests that different subdomains in this region might be utilized to mediate the FIP3-FIP3 and FIP3-RIP interactions. We are currently investigating whether or not there is mutual inhibition between FIP3 oligomerization and the FIP3-RIP interaction. Our preliminary data also suggest that there are two separate NF-B-inducing kinase-interacting domains in the FIP3 protein, one between amino acids 120 and 179 and the other between amino acids 201 and 300 (data not shown). These studies suggest that FIP3 might play a scaffolding role through its association with different components to organize and regulate the IB- kinase complex (Fig. 8).

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Schematic diagram showing the domains of FIP3 responsible for various functions. Shaded areas indicate regions important for a specified function. Subdomains that are important for a specified interaction are further darkened, and those that do not appear to be important for a particular function are left blank.

Stimulation of the TNF pathway leads to both apoptosis and activation of NF-B (42). Activation of the NF-B pathway is thought to protect against cell death (52). The underlying mechanism proposed is that NF-B activates the expression of cellular anti-apoptotic genes, e.g. IAPs, which inhibit the apoptotic branch of the TNF pathway (53, 54). Because FIP3 down-regulates NF-B and causes cell death, it might have been expected that cell death would be the result of the inhibition of the protective effect of NF-B activation. However, this does not appear to be the explanation, because the carboxyl-terminal half of the FIP3 protein, which effectively down-regulates the NF-B activity, does not cause cell death. Nonetheless, it is still possible that NF-B down-regulation complements FIP3 in achieving its maximum apoptosis-inducing activity. Consistent with this latter possibility, deletion of the carboxyl-terminal NF-B repression domain compromised the activity of FIP3 to induce cell death (Fig. 7II). Peptide inhibitors of the ICE-like caspase family (YVAD-CHO) or the CED-3 subfamily (DEVD-CHO) do not appear to block FIP3-induced cell death (data not shown), suggesting that these caspases might not be involved in the FIP3 cell death pathway. We are currently investigating whether FIP3 activates the caspase 2 pathway through its association with RIP, which has been shown to be involved in the activation of this caspase through the caspase recruitment domain-containing adaptor protein RAIDD/CRADD (46).

FIP3 overexpression was also shown to block TNF-induced IB- degradation (Fig. 6A) while simultaneously activating IKK as measured by an in vitro kinase assay (Ref. 38 and Fig. 6C). However, the IB- subpopulation is not phosphorylated, suggesting that the endogenous IB- is not available as a substrate for the activated IKK. In addition, the mobility of the stabilized IB- did not appear to change from nonstimulated controls, arguing that IB- was not ubiquitinated. It is still not clear how FIP3 on one hand acts as an essential component of the IKK complex and on the other hand inhibits NF-B activation. One possibility is that FIP3 is the entry point for feedback inhibition of the NF-B pathway. In cells, the NF-B activity needs to be tightly regulated, and this makes IB degradation and NF-B activation a transient event on most occasions. FIP3 might be modified after activation of the NF-B pathway, and the modified FIP3 might play a role opposite to its unmodified form. When overexpressed, a significant amount of FIP3 protein might mimic the function of the modified form of endogenous FIP3. The FIP3 mutant ND100 might mimic the modified form better and thus be more potent in stabilizing IB- (Fig. 6A, lane 3). TNF treatment does not affect the endogenous FIP3 protein level (data not shown); thus the post-stimulus modification might occur post-translationally, and we are currently testing this possibility.

It is interesting to note that the NF-B down-regulation activity of FIP3 is independent of its association with the kinase components of the IKK complex (Fig. 8). This implies that these two functions are discrete properties of FIP3 protein. FIP3 might only act as a scaffolding protein in the formation and activation of the IKK complex, whereas the NF-B down-regulation may require an additional function of FIP3.

The domain responsible for the binding of FIP3 to the adenovirus inhibitor of apoptosis E3-14.7K was also mapped at the carboxyl-terminal half, which is the region involved in NF-B modulation. This suggests that 14.7K might act on the NF-B pathway to protect cells from apoptosis induced by TNF, either by enhancing the positive role of FIP3 in activation of NF-B or by blocking the inhibitory function of FIP3. The net effect would be an induction of higher NF-B activity, which would overcome cell death.

	FOOTNOTES

^* This work was supported by the National Institutes of Health Grant RO1 CA72963 (to M. S. H. and J. Y.), National Institutes of Health Cancer Center Core Grant CA13330 (to M. S. H.), and the Forchheimer Foundation (to M. S. H.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The data in this paper will be submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Sue Golding Graduate Division of Medical Sciences, Albert Einstein College of Medicine, Yeshiva University.

^§ To whom correspondence should be addressed. Tel.: 718-430-2230; Fax: 718-430-8702; E-mail: horwitz@aecom.yu.edu.

² L. Tarassishin and M. S. Horwitz, unpublished observations.

	ABBREVIATIONS

The abbreviations used are: NF-B, nuclear factor B; IB, inhibitor of B; TNF, tumor necrosis factor ; IKK, IB kinase; E3-14.7-kDa, adenovirus early region 3-14.7-kDa protein; FIP3, E3-14.7-kDa-interacting protein; RIP, Fas receptor-interacting protein; HRP, horseradish peroxidase; GST, glutathione S-transferase.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES