Fas-associated Factor-1 Inhibits Nuclear Factor-κB (NF-κB) Activity by Interfering with Nuclear Translocation of the RelA (p65) Subunit of NF-κB*

Fas-associated factor-1 (FAF1) is a Fas-binding pro-apoptotic protein that is a component of the death-inducing signaling complex in Fas-mediated apoptosis. Here, we show that FAF1 is involved in negative regulation of NF-κB activation. Overexpression of FAF1 decreased the basal level of NF-κB activity in 293 cells. NF-κB activation induced by tumor necrosis factor (TNF)-α, interleukin-1β, and lipopolysaccharide was also inhibited by FAF1 overexpression. Moreover, FAF1 suppressed NF-κB activation induced by transducers of diverse NF-κB-activating signals such as TNF receptor-associated factor-2 and -6, MEKK1, and IκB kinase-β as well as NF-κB p65, one of the end point molecules in the NF-κB activation pathway, suggesting that NF-κB p65 might be a target molecule upon which FAF1 acts. Subsequent study disclosed that FAF1 physically interacts with NF-κB p65 and that the binding domain of FAF1 is the death effector domain (DED)-interacting domain (amino acids 181–381), where DEDs of the Fas-associated death domain protein and caspase-8 interact. The NF-κB activity-modulating potential of FAF1 was also mapped to the DED-interacting domain. Finally, overexpression of FAF1 prevented translocation of NF-κB p65 into the nucleus and decreased its DNA-binding activity upon TNFα treatment. This study presents a novel function of FAF1, in addition to the previously known function as a component of the Fas death-inducing signaling complex, i.e. NF-κB activity suppressor by cytoplasmic retention of NF-κB p65 via physical interaction.

FAF1 is able to enhance Fas-induced apoptosis. It has also been demonstrated that FAF1 can initiate apoptosis in the absence of any extrinsic death signals when overexpressed (2). Recently, FAF1 has been demonstrated to be a component of the death-inducing signaling complex (DISC) in Fas-mediated apoptosis (3).
Although FAF1 has demonstrated its apoptotic potential, it does not contain typical death motifs such as the death domain, death effector domain (DED), and caspase recruitment domain. Instead, FAF1 has sequence motifs that are present in the proteins of the ubiquitination pathway (4). FAF1 has two ubiquitin-homologous domains in its N terminus and a domain (the UX domain) homologous to proteins involved in the ubiquitination pathway. The three-dimensional structure of the UX domain of FAF1 as determined by NMR analysis demonstrates its structural similarity to ubiquitin (5). Despite the amino acid sequence and structural similarities to ubiquitin, functions related to the ubiquitination pathway have not yet been assigned to FAF1.
Members of the tumor necrosis factor (TNF) receptor (TNFR) family such as Fas, TNFR1, and TNFR2 play critical roles in various cellular responses, including cell survival and apoptosis (6,7). The initiation of signal transduction by these receptors requires adapter proteins to trigger cytoplasmic signal transduction cascades. The adapter proteins are not always specific to certain types of receptors. For example, the adapter molecule FADD is recruited not only to Fas, but also to TNFR1 and death receptor-3 (8). TRAF6 can act as an adapter for the interleukin-1 (IL-1) receptor as well as for TNFR2 (9,10). Different signaling cascades initiate depending on the combination of adapter molecules recruited to the receptors. In TNFR1 signaling, the recruitment of the TRADD⅐FADD complex leads to the activation of the caspase cascade, resulting in apoptosis, whereas that of the TRADD⅐TRAF2 complex leads to the activation of NF-B, resulting in survival (11,12).
FLICE-associated huge protein (FLASH) was identified by yeast two-hybrid assay using the duplicated DEDs of caspase-8 as probes, and its DED-recruiting domain interacts with the DED of FADD (13). FLASH is recruited to the Fas DISC, which is composed of Fas, FADD, and caspase-8. Overexpression of FLASH could potentiate Fas-mediated apoptosis after stimulation, and FLASH has a role in TNF-mediated NF-B activation via a TRAF2/NIK/IKK-dependent pathway (14).
FAF1 is similar to FLASH and FADD. Like FLASH, FAF1 binds to the DEDs of FADD and caspase-8 (3). FAF1 demonstrates similarity to FADD because both bind to Fas and * This work was supported by Ministry of Science & Technology and Korea Science and Engineering Foundation through the Research Center for Biomedicinal Resources of PaiChai University. 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.
NF-B is activated by various stimuli, including TNF, IL-1, and lipopolysaccharide (LPS) (18). NF-B regulates the expression of genes involved in the immune response, inflammatory process, and apoptosis (19,20). NF-B signals subsequently activate two subunit kinases, IKK␣ and IKK␤ (21)(22)(23). Phosphorylation of IB proteins leads to their ubiquitination and subsequent degradation by the 26 S proteasome. After degradation of IB, free NF-B translocates from the cytoplasm to the nucleus and activates NF-B-dependent genes (24).
In this study, we investigated whether FAF1 is involved in NF-B activity modulation. We demonstrate that FAF1 can inhibit NF-B activation by divergent signals, including TNF␣, IL-1␤, and LPS. This suppression of NF-B activity is contributed by the protein interaction between FAF1 and NF-B p65, leading to the inhibition of NF-B activation by preventing the translocation of NF-B p65 to the nucleus.
Luciferase Reporter Gene Assays-To determine whether FAF1 modulates NF-B activation by various signals, 293 cells were transfected with 1 g of test plasmid, 0.1 g of Ig-B-luc reporter plasmid, and 0.1 g of CMV-␤-galactosidase plasmid. To demonstrate that the action of FAF1 on NF-B is specific, 0.1 g of AP-1-luc plasmid, 0.5 g of pGL2-cIAP2Ϫ247-luc plasmid, and 0.5 g of pGL2-cIAP2Ϫ247(mB1,3)-luc reporter plasmid were tested as controls. The total DNA concentration in each transfection was kept constant by adjusting with the empty vector. Cells were left untreated or were stimulated at 36 h post-transfection with TNF␣ (20 ng/ml), IL-1␤ (20 ng/ml), or LPS (100 ng/ml) for 6 h. The cell lysates were assayed for luciferase activities according to the manufacturer's protocol (Promega, Madison, WI). Transfection efficiencies were normalized by measuring ␤-galactosidase activities.
Immunoprecipitation and Antibodies-At 48 h post-transfection, cells were washed with phosphate-buffered saline, lysed in lysis buffer (50 mM Tris-Cl (pH 8.0), 150 mM NaCl, 1 mM EDTA, 1% Nonidet P-40, and 0.4 mM phenylmethylsulfonyl fluoride), and then sonicated using a Branson Model 450 sonifier. The cell lysates was centrifuged at 14,000 ϫ g for 10 min, and the protein concentrations of the cell lysates were determined using a Bio-Rad protein assay kit. 100 g of supernatant was incubated for 4 h with 1 g of anti-FLAG monoclonal antibody (mAb) M2 (Sigma). The samples were further incubated with 30 l of protein A/G Plus-agarose (Santa Cruz Biotechnology, Santa Cruz, CA) for 2 h at 4°C. The beads were washed five times with lysis buffer and separated by SDS-PAGE. Immunocomplexes were then transferred to nitrocellulose membranes and analyzed by Western blotting using the appropriate primary antibodies and secondary antibodies coupled to peroxidase. Proteins were visualized by enhanced chemiluminescence (Amersham Biosciences). The antibodies used were as follows: anti-FLAG mAb M2, anti-␤-tubulin mAb, and anti-␤-actin mAb (Sigma) and rabbit anti-p65 polyclonal antibody (pAb) H10 and polyclonal anti-goat and anti-mouse IgG (Santa Cruz Biotechnology). Mouse anti-FAF1 mAb was provided by Dr. Jong-Seok Lim (Korea Research Institute of Bioscience and Biotechnology, Daejeon, Korea).
For detection of endogenous binding between FAF1 and NF-B p65, 293, NIH3T3, and NIH3T3-p65 Ϫ/Ϫ cells were treated with or without TNF␣ for 30 min. Cells were lysed in lysis buffer, and 100 g of supernatant was incubated with anti-FAF1 mAb and mouse IgG 1 , respectively, for 4 h at 4°C. Immunocomplexes were collected by incubation with protein A/G Plus-agarose for 2 h at 4°C and then analyzed by Western blotting as described above.
GST Pull-down Assay-293 cells were transfected with 1 g of pcDNA3.1-NF-B-p65 by the calcium phosphate precipitation method. At 48 h post-transfection, cells were harvested in lysis buffer as described above. GST-FAF1 deletion mutant fusion proteins were expressed in Escherichia coli BL21(DE3) cells and purified with glutathione-Sepharose 4B beads (Peptron, Daejeon) as specified by the manufacturer. For binding assay, 100 g of cell lysate was incubated with 1 g of either GST or GST-FAF1 deletion mutant fusion protein for 4 h at 4°C in lysis buffer. Samples were washed four times with lysis buffer, and the bound proteins were separated by SDS-PAGE and detected by Western blotting with rabbit anti-NF-B p65 pAb H10.
Electrophoretic Mobility Shift Assays-293 cells were transfected with pFLAG-FAF1 and harvested, and nuclear extracts were then prepared as described previously (27). Electrophoretic mobility shift assay was performed as described (26). Briefly, synthetic oligonucleotides were end-labeled with T4 polynucleotide kinase and [␥-32 P]ATP and incubated with nuclear proteins (4 g) at room temperature for 30 min in binding buffer (10 mM Tris-HCl (pH 7.5), 50 mM KCl, 1 mM dithiothreitol, 1 mM EDTA (pH 7.5), 5% glycerol, 2 g of poly(dI-dC), and 5% Nonidet P-40) for 20 min at room temperature, and the protein⅐DNA complexes were loaded onto a 5% native polyacrylamide gel and detected by autoradiography. To measure NF-B activation, an end-labeled, double-stranded, NF-B-specific oligonucleotide probe containing the two NF-B tandem repeats derived from the human immunodeficiency virus long terminal repeat (5Ј-ATCAGGGACTTTCCGCT-GGGGACTTTCCG-3Ј) was used (27). For control experiments, the region from bp Ϫ66 to Ϫ48 upstream of the surfactant protein A2 gene transcription initiation site, which contains the GT box (5Ј-TCT-CAGGGGTGGGGAAGAA-3Ј), was used (28).
Cell Fractionation-Nuclear and cytosolic fractions were isolated by differential centrifugation as follows. To prepare the nucleus, 2 ϫ 10 6 cells were lysed in lysis buffer for 15 min at 4°C to swell the cells. The cells were passed 10 -12 times through a 21-gauge needle, and cell lysates were centrifuged at 1000 ϫ g for 10 min at 4°C. The pellet was solubilized with 200 l of lysis buffer and used as the nucleus. The supernatant was centrifuged at 100,000 ϫ g for 1 h to separate the cytoplasmic fraction from the membrane fraction. The supernatant was immediately subjected to Western blot analysis.

RESULTS
Overexpression of FAF1 Inhibits NF-B Activation-To test a possible involvement of FAF1 in the NF-B activation pathway, we first examined whether overexpression of FAF1 can modulate NF-B activation using a transient transfection system. As shown in Fig. 1A, overexpression of FAF1 inhibited TNF␣-induced NF-B activity as well as its basal transcriptional activity in 293 cells. Direct targeting of FAF1 expression using antisense FAF1 increased the basal level of NF-B activation as well as NF-B activation induced by TNF␣. Inhibition of NF-B by FAF1 was decreased in a dose-dependent manner (Fig. 1B). To test the specificity of FAF1 for NF-B, we examined whether FAF1 can modulate the activity of other TNF␣induced transcription complexes such as AP-1. The experiment using the AP-1-luc reporter plasmid showed that FAF1 did not affect AP-1 activity in the presence or absence of TNF␣ (Fig.  1C, left). As an additional validation regarding the specificity of FAF1 for NF-B, we experimented using the luciferase reporter containing the NF-B-binding sites of the cIAP2 promoter (pGL2-cIAP2Ϫ247-luc) and the luciferase reporter containing mutant functional NF-B-binding elements (pGL2-cIAP2Ϫ 247(mB1,3)-luc) (25). The cIAP2 gene is transcriptionally reg-ulated by TNF␣ and other stimuli activating NF-B. As shown in Fig. 1C (middle and right), FAF1 down-regulated the transcription of pGL2-cIAP2Ϫ247-luc, but did not change that of pGL2-cIAP2Ϫ247(mB1,3)-luc. These results verify the specificity of FAF1 for NF-B. Next, we questioned whether the down-regulation of NF-B activity is influenced by caspase activation. To address this question, NF-B activation by overexpression of FAF1 was measured in the presence of a pancaspase inhibitor (benzyloxycarbonyl-VAD-fluoromethyl ketone) and in cells in which FAF1 and CrmA, a poxvirus-derived specific inhibitor of caspase-8, were cotransfected (29). As shown in Fig. 1D, our data confirm that the down-regulation of NF-B activity by FAF1 is independent of caspase activities.
FAF1 Suppresses NF-B Activation via Multiple Pathways-We questioned whether FAF1 can exert its negative regulatory effect on diverse NF-B-activating signals. When overexpressed, FAF1 inhibited TNF␣-, IL-1␤-, and LPS-mediated NF-B activation ( Fig. 2A). To further determine how FAF1 modulates NF-B activity, we tested whether FAF1 inhibits NF-B activation mediated via various signal transducers known to be NF-B activators downstream of the TNFR superfamily and IL-1 receptor. As shown in Fig. 2B, when overexpressed, FAF1 inhibited TRAF2-, TRAF6-, and MEKK1mediated NF-B activation, suggesting that FAF1 might function at a point where divergent signals merge in the NF-B activation pathways. Certainly, FAF1 inhibited NF-B activation by overexpression of IKK␤, a component of the IKK complex where divergent NF-B-activating signals converge, suggesting that FAF1 might function in the IKK complex and/or its downstream components in the NF-B activation pathway. Indeed, FAF1 inhibited NF-B activation by overexpression of NF-B p65, one of the end point molecules in the NF-B activation pathway. This suggests that NF-B p65 might be the target molecule upon which FAF1 acts.
FAF1 Interacts with NF-B p65-Based on the fact that FAF1 inhibits NF-B activation by overexpression of NF-B p65, we investigated whether FAF1 can physically interact with NF-B p65. 293 cells were transfected with pcDNA3.1-NF-B-p65 with or without pFLAG-FAF1. At 48 h post-transfection, cell lysates were immunoprecipitated using anti-FLAG mAb. Fig. 3A illustrates that FAF1 can interact with NF-B p65 directly in 293 cells. To investigate whether FAF1 and NF-B p65 are endogenously associated, we performed coimmunoprecipitation experiments using total cell lysates. As  1. FAF1 inhibits NF-B activation. A, FAF1 inhibits basal and TNF␣-stimulated NF-B activation. 293 cells were cotransfected with 0.1 g of Ig-B-luc reporter plasmid and 0.1 g of CMV-␤-galactosidase plasmid together with 1 g of pFLAG empty vector, pFLAG-FAF1, or pFLAG-antisense FAF1 as indicated. At 24 h post-transfection, cells were treated with TNF␣ (20 ng/ml) for 6 h or left untreated. Luciferase activity was then measured as described under "Experimen-shown in Fig. 3B, endogenous NF-B p65 was co-immunoprecipitated with FAF1 from 293 cell lysates induced by TNF␣ for 30 min. However, anti-FAF1 antibody did not co-immunoprecipitate NF-B p65 from the lysates of untreated 293 cells. The same experiment was performed with NIH3T3 cells, and equivalent results were obtained. However, no immunoprecipitation of FAF1 was detected in NIH3T3-p65 Ϫ/Ϫ cells. These results suggest that the TNF␣ signal might have facilitated FAF1 and NF-B p65 interaction in the cell.
The DED-interacting Domain (DEDID) of FAF1 Interacts with NF-B p65-To determine the FAF1 region responsible for the interaction with NF-B p65, GST pull-down assay was performed using GST-FAF1 fusion proteins. The extracts of 293 cells transfected with NF-B p65 were incubated with either GST or GST-FAF1 deletion mutant proteins (Fig. 4A). GST-FAF1 DEDID (amino acids 181-381) and GST-FAF1 bound to NF-B p65 (Fig. 4B), implying that the DEDID is sufficient for binding. It is worth mentioning that the DEDID is the region where the DEDs of FADD and caspase-8 also bind. To determine whether the binding affects NF-B activity, pFLAG-FAF1 and pFLAG-FAF1-DEDID were transfected, and NF-B activity was measured. When overexpressed, the FAF1 DEDID and FAF1 demonstrated NF-B activity-inhibiting potential (Fig. 4C).
FAF1 Inhibits Nuclear Translocation and Decreases the DNA-binding Activity of NF-B-To determine how FAF1/ NF-B p65 interaction modulates NF-B activity, we examined the influence of FAF1 on NF-B activity in gel shift experiments (Fig. 5A). An inducible protein⅐DNA complex was observed in the nuclear extracts upon TNF␣ treatment for 20 min. As expected, overexpression of FAF1 decreased significantly the DNA-binding activity, whereas FAF1 had no effect on GT box binding to DNA with or without TNF␣ treatment (Fig. 5B). We further confirmed the effects of FAF1 on the translocation of NF-B p65 into the nucleus of 293 cells with or without TNF␣ stimulation by Western blot analysis (Fig. 5C). TNF␣ stimulation increased the protein levels of NF-B p65 in the nucleus. However, FAF1 overexpression decreased significantly the nuclear translocation of NF-B p65, suggesting that FAF1 interferes with NF-B activity by preventing the nuclear translocation of NF-B p65 from the cytoplasm. DISCUSSION Our previous study showed that FAF1 functions as a component of the Fas DISC in Jurkat cells (3). In this study, we carried out experiments to examine whether FAF1 proteins can modulate the signal transduction pathway leading to NF-B activation, which is required for anti-apoptotic signals. Our study discloses a new function of FAF1. FAF1 negatively regulates NF-B activation by various extracellular stimuli. This is caused by the cytoplasmic retention of NF-B p65 by FAF1 via physical interaction.
The modulation of NF-B activation by death-inducing molecules has been reported. FADD, caspase-8, and FLASH were reported not only to induce cell death, but also to activate NF-B (14, 17, 30 -32). However, pro-apoptotic molecules such as ASK1 and prostate apoptosis response 4 negatively regulate NF-B activation (33)(34)(35). Because a pro-apoptotic FAF1 protein inhibits NF-B activation as do the ASK-1 and PAR-4 proteins, FAF1 can be added to the list of NF-B-down-regulating proteins with apoptotic potential.
There are other proteins that interact with NF-B p65. Wu et al. (36) reported that the promyelocytic leukemia protein PML, a pro-apoptotic protein, represses NF-B transactivation by interacting with NF-B p65 in the nucleus. In addition, histone deacetylase-1 and p202 also negatively regulate the transcriptional activity of NF-B by interacting with NF-B p65 (37,38). Those are nuclear proteins. In this respect, FAF1 presents a new regulatory mechanism for inhibition of NF-B activation, i.e. preventing nuclear translocation of NF-B p65 from the cytoplasm.
The interaction between FAF1 and NF-B p65 has been confirmed in two different cell types, i.e. 293 and NIH3T3 cells. The interaction occurred in TNF␣-treated cells, but barely in untreated cells, showing that FAF1/NF-B p65 interaction is tightly regulated and also stressing the physiological importance of this interaction. It is well known that the TNF␣ signal degrades IB and frees NF-B. Thus, it is conceivable that IB inhibits FAF1/NF-B p65 binding and that degradation of IB by the TNF␣ signal would free NF-B to bind to FAF1.
The DED is a pro-apoptotic signaling domain that mediates NF-B activation (17,30). DED-containing proteins such as caspase-8, caspase-10, and Casper can activate NF-B. One regulatory mechanism of the DED in the NF-B pathway occurs via the physical interactions of DED with NF-B pathway components. Caspase-8 interacts with multiple proteins in the NF-B pathway, such as receptor-interacting protein, NIK, IKK␣, and IKK␤ (30), and activates NF-B. The FAF1 DEDID interacts with the DEDs of FADD and caspase-8 (3). Binding of FAF1 to caspase-8 might prevent caspase-8 from binding to the NF-B pathway components. If FAF1/caspase-8 binding inhib-its the interactions of caspase-8 with NF-B pathway components, FAF1/caspase-8 binding would also reduce the positive influence of caspase-8 on NF-B activation. Thus, in addition to FAF1/NF-B p65 binding, FAF1/caspase-8 binding might have additional negative effects on NF-B activation.
FAF1 down-regulates NF-B activation and promotes apoptosis as a component of the Fas DISC. Even though both activities will drive cells in a pro-apoptotic direction, the relationship between these two activities and whether they converge in the same apoptotic effector molecule are not clear. However, these two activities seem to occur independently based on the fact that down-regulation of NF-B occurred in the presence of caspase inhibitors. This study shows that FAF1 is involved in dual signaling mechanisms, one via the Fas DISC and the other via down-regulation of NF-B. Thus, the combined action of the FAF1 protein may enhance the efficacy of receptor-mediated cell death.