TNAP, a Novel Repressor of NF-κB-inducing Kinase, Suppresses NF-κB Activation*

NF-κB-inducing kinase (NIK) has been implicated as an essential component of NF-κB activation. However, the regulatory mechanism of NIK signaling remains elusive. We have identified a novel NIK interacting protein, TNAP (for TRAFs and NIK-associated protein). In mammalian cells, TNAP physically interacts with NIK, TRAF2, and TRAF3 but not IKK1 or IKK2. TNAP specifically inhibits NF-κB activation induced by tumor necrosis factor (TNF)-α, TNF receptor 1, TRADD, RIP, TRAF2, and NIK but does not affect IKK1- and IKK2-mediated NF-κB activation. Knockdown of TNAP by lentiviral-mediated small interference RNA potentiates TNF-α-induced NF-κB activation. TNAP suppresses NIK kinase activity and subsequently reduces p100 processing, p65 phosphorylation, and IκBα degradation. These data suggest that TNAP is a repressor of NIK activity and regulates both the classical and alternative NF-κB signaling pathways.

The NF-B family of transcription factors is a key regulator of inflammation, immune responses, oncogenesis, apoptosis, and neuronal signaling (1)(2)(3)(4)(5). To date two principal pathways for NF-B activation have been characterized: a classical and an alternative pathway (2,6,7). The classical pathway involves most forms of NF-B, especially the p65 (RelA)/p50 dimer. It is triggered by tumor necrosis factor (TNF)-␣, 1 interleukin-1, lipopolysaccharide (LPS), and CD40 ligand (CD40L) and to a lesser extent by lymphotoxin-␤ (LT-␤) and B cell-activating factor. Activation of this pathway depends on the IB kinase (IKK) signalsome, which consists of at least two catalytic subunits (IKK1 or IKK-␣ and IKK2 or IKK-␤) and a regulatory subunit (IKK-␥ or NEMO). The IKK complex phosphorylates the inhibitor proteins of NF-B (IBs) to induce their ubiquiti-nation and degradation, resulting in NF-B translocation to the nucleus and transcriptional activation of specific target genes. This pathway is crucial for the activation of innate immunity and inflammation. The alternative pathway involves p100 processing and nuclear translocation of RelB/p52 dimers. It is activated by LT-␤, CD40L, LPS, and B cell-activating factor and depends on IKK1 homodimers (8 -12). This alternative pathway is crucial for secondary lymphoid organ development, maturation of B cells, and adaptive humoral immunity.
NF-B-inducing kinase (NIK) has been implicated in activating both the classical and alternative pathways. NIK was originally identified as a mitogen-activated protein kinase kinase kinase that interacts with TNF receptor-associated factor 2 (TRAF2) and activates NF-B when expressed in mammalian cells (13). NIK has been characterized as a downstream component of the TNF signaling pathway (14,15), which is activated directly or indirectly by cytoplasmic adaptor proteins like TNF receptor-associated death domain (TRADD) (16), receptor interacting protein (RIP) (17,18), or TRAF2 (16,19). A kinase inactive mutant of NIK can block NF-B activation by various stimuli, such as TNF-␣, interleukin-1, and LPS. In bone marrow, TNF-␣ and interleukin-1 direct pluripotent mesenchymal stem cell toward an osteoblast fate and suppress adipogenesis through the NIK signaling (20). NIK interacts with both IKK1 and IKK2 (14,21). For the heterodimer of IKK complex, NIK activates IKK1 and IKK2 in a directional manner (22). Genetic studies with NIK mutant mice (23)(24)(25) and knockout mice (26,27) have demonstrated that NIK plays an essential role in the signaling pathway of LT-␤ receptor but not TNF-␣ receptor, implying that NIK functions in a receptor-and cell type-specific manner. Further studies confirm that NIK is indispensable in the IKK1-mediated phosphorylation and processing of p100 and the activation of the alternative pathway of NF-B activation (10, 12, 28 -33). NIK also has been demonstrated to phosphorylate the p65 transactivation domain (34,35), to activate mitogen-activated protein kinase pathways (36,37), to be involved in signaling of epidermal growth factor receptor (38), to establish self-tolerance (39), and to promote neurite formation (40). These observations indicate that NIK exerts a wide range of biological functions. However, the mechanisms for NIK activation and regulation are not well understood. Using a yeast two-hybrid screen with NIK as bait, we have identified a novel protein, designated TNAP (for TRAFs and NIK-associated protein), which directly binds to NIK and inhibits NIK kinase activity. Studies via overexpression and knockdown of TNAP demonstrate that TNAP negatively regulates NF-B activation upstream of the IKK complex and regulates both the classical and alternative NF-B signaling pathways.

EXPERIMENTAL PROCEDURES
Yeast Two-hybrid Screening-The yeast two-hybrid system Matchmaker II (Clontech) was used. The bait, encoding the N-terminal domain (nucleotides 1-437) of NIK, was inserted in frame into the GAL4 DNA-binding domain in the yeast expression vector pBridge (Clontech). The yeast strain G1945 was transformed first with pBridge-NIK (1-437) and then with a human brain cDNA expression library fused to the GAL4 transactivator domain in expression vector pACT2 (Clontech).
Transfections and Generation of Stable Cell Lines-For coimmunoprecipitation and Western blot experiments, the cells were seeded in 10-cm dishes. For reporter gene assay and EMSA, the cells were seeded in 6-well dishes. For immunocytochemistry, the cells were seeded in 8-well chamber slides. The adherent cells (HEK293T) were transfected by the standard calcium phosphate method. Suspension cells (Jurkat) were transfected with FuGENE 6 (Roche Applied Science). After cotransfection of pRK-TNAP with pTK-Hyg selection vector (Clontech, Palo Alto, CA) into HEK293T cells, stable clones were selected with hygromycin (20 ng/ml).
Reporter Gene Assay-The Dual-Light chemiluminescent reporter gene assay for the combined detection of luciferase and ␤-galactosidase (Tropix) was performed following the manufacturer's protocols. Within the same experiment, each transfection was performed in triplicate, and when necessary, empty control plasmid was added to each sample to ensure each transfection received the same amount of total DNA. To normalize for transfection efficiency and protein concentration, a constant amount of RSV-␤-galactosidase plasmid was added to all transfections. The luciferase activity was normalized to ␤-galactosidase expression levels. The data are expressed as the averages Ϯ S.D. of one representative experiment.
GST-TNAP Fusion Protein Expression and Antibody Production-GST fusion proteins were expressed and prepared according to standard procedures using expression vector pGEX-4T-2 (Amersham Biosciences). The purified fusion protein was used as an immunogen to raise polyclonal antisera in rabbits (Proteintech Group Inc.). The final anti-serum was affinity-purified with a GST-TNAP and a GST column.
In Vitro Translation, GST Pull-down, Immunoprecipitation, and Western Blot-IKK1 and NIK were in vitro translated and labeled with [ 35 S]methionine (TNT kit; Promega). For pull-down experiments, 15 l of in vitro translation mixture was incubated overnight at 4°C with equal amounts of bacterially expressed GST fusion proteins coupled to 20 l of glutathione-Sepharose. The beads were washed six times with 20 mM Tris, pH 8, 1 mM EDTA, 50 mM NaCl, 0.5% Nonidet P-40. The bound proteins were eluted in SDS-PAGE sample buffer, separated on 10% SDS-PAGE, and detected by fluorography. Immunoprecipitation and Western blot were performed essentially as described (48). Antibodies against phospho-IKK1(Ser-176/180), phospho-p65(Ser-536) and p65 (Cell Signaling Technology), p52 (Upstate Biotechnology, Inc.), and phospho-IB␣ and IB␣ (Santa Cruz) were obtained from the indicated resources.
In Vitro and in Vivo Kinase Assays-FLAG or Myc-tagged fusion proteins TNAP, NIK, and IKK (K44M) overexpressed in HEK293T cells were purified through immunoprecipitation. An equal amount of IKK1(K44M) protein was used as substrate, and the in vitro kinase assay was performed as described (49). For the in vivo kinase assay, HEK293T cells were cotransfected with pRK-Myc-NIK and pRK-FLAG-IKK1(K44M) with or without pRK-Myc-TNAP. Twenty-four hours after transfection, the cells were washed once in phosphate-free Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% dialyzed, heat-inactivated, fetal bovine serum and starved for 2 h in the same medium. 32 P-Labeled orthophosphoric acid (0.5 mCi) was then added to the cells. After incubation for 4 h, the cells were harvested for immunoprecipitation with anti-FLAG antibody (Sigma) as described (48). Immunoprecipitated complexes were fractionated by SDS-PAGE, transferred to nitrocellulose membranes, and analyzed by autoradiography. The same membrane was immunoblotted with anti-Myc (Santa Cruz) and anti-FLAG antibodies to determine the expression levels.
Electrophoretic Mobility Shift Assays-The Cells were starved with media containing 0.2% fetal bovine serum for 20 -24 h before treatment for indicated times. The nuclear extracts and EMSA were performed as described (50,51).
RT-PCR and Northern Blot-Total RNA was isolated with TRIzol reagent (Invitrogen). Two micrograms of RNA pretreated with DNase I for 2 h at 37°C was used to synthesize cDNA using SuperScript TM II reverse transcriptase with random hexanucleotide as primers. A touchdown PCR was performed on the cDNA using TNAP (76 -140) specific primers: the antisense (21-mer), 5Ј-ATGCCTCTGCACCCCAGCCTG-3Ј and the sense (21-mer), 5Ј-AAATATGGCTGGGGCAGCAAT-3Ј.
The tissue-specific expression of TNAP was examined by hybridization of a human multiple tissue Northern blot containing 2 g of poly(A) ϩ RNA (OriGene). A PCR-produced probe of 192 bp was [ 32 P]dCTP-labeled with random priming. The hybridizations were conducted in the ExpressHyb hybridization solution (Clontech) at 65°C for 2 h according to the manufacturer's protocol. The same blots were then stripped and hybridized with the 32 P-labeled 380-bp cDNA fragment PstI-digested from pRK-FLAG-NIK as well as 300 bp of PCR product of human ␤-actin.
Immunohistochemistry-For immunohistochemical studies, free floating sections of spinal cord were stained as described (52). The human spinal cords were obtained from the Pathology Department at the University of Miami.
Quantitative Image Analysis-The images from EMSA, Western blot, and kinase assays were scanned and analyzed with the Quantity One software from Bio-Rad. The data are presented as adjusted volume times the total area.

RESULTS
Cloning and Characterization of TNAP-To identify proteins associated with the NF-B signaling pathway, we performed a yeast two-hybrid screen of an adult human brain cDNA library using NIK as bait. From 3 ϫ 10 6 yeast transformants, 40 clones were both histidine-and ␤-galactosidase-positive. Sequence analysis confirmed the presence of several known gene products that interact with NIK, such as TRAF3 (41). We also identified several novel clones. One of them, TNAP, is 660 bp in length with an open reading frame of 140 amino acids (Gen-Bank TM accession number: AF463496). TNAP coincides with nucleotides 87239 -86607 of human clone RP11-429E17 on chromosome 13q13.1-14.2. No known protein sequence or motif was found to resemble TNAP. The 140 amino acids share significant homology (40 -60%) with several un-named small protein products.
To determine the transcript size and tissue distribution of TNAP, Northern blot analysis of human adult tissues (Ori-Gene) was performed. As shown in Fig. 1A, a single transcript of 660 bp is expressed in all the tissues examined, with high expression in testis, brain, colon, liver, and lung (Fig. 1A). The tissue distribution of TNAP mRNA is consistent with NIK mRNA expression, although the expression level is different between TNAP and NIK in most tissues. The relative amount of mRNA was confirmed by Northern blot with human ␣-actin. RT-PCR of RNAs extracted from different cell lines demonstrates that TNAP mRNA is constitutively expressed in HL60 (human myeloid leukemia cell line), BV2 (mouse microglia cell line), and RN33B (rat neuronal cell line) ( Fig. 1B and data not shown).
To determine whether the TNAP gene encodes an expressed protein, we raised a rabbit polyclonal antibody against a GST-TNAP fusion protein. The specificity of this affinity-purified antibody was tested using double-labeled immunofluorescent staining and Western blot analysis of HEK293T cells overexpressing Myc-TNAP. As shown in Fig. 1C, anti-TNAP antibody specifically recognizes the purified Myc-TNAP fusion protein.
To determine the endogenous protein expression of TNAP, Western blot analysis of different cell lines was performed. As shown in Fig. 1D, the anti-TNAP antibody recognizes a band of ϳ16 kDa, the predicted size for TNAP in HL60 cells, which is consistent with the expression of TNAP mRNA demonstrated by RT-PCR (Fig. 1B). In the white matter of the spinal cord, TNAP-like immunoreactivity is detectable in glial cells (Fig. 1E).
Physical Interaction of TNAP with NIK-We cloned the open reading frame of TNAP into a mammalian expression vector to verify the interaction between TNAP and NIK in mammalian cells. Coimmunoprecipitation and Western blot analysis in transiently transfected HEK293T cells demonstrate that TNAP specifically interacts with NIK ( Fig. 2A). TNAP also interacts with TRAF2 and TRAF3 but does not interact with IKK1 or IKK2 (Fig. 2A). The direct interaction between TNAP and NIK, but not TNAP and IKK1, was confirmed by GST pull-down assay using bacterially expressed GST-TNAP fusion protein and in vitro translated NIK or IKK1 protein (Fig. 2B). To examine which region of TNAP interacts with NIK, we made several deletion mutants of TNAP and transiently expressed these mutants in HEK293T cells. The expression levels of TNAP and its mutants were confirmed by immunofluorescent cytochemistry with an anti-Myc antibody. As shown in Fig. 2C, the C-terminal fragment (amino acids 87-140) interacts with NIK, but the N-terminal fragment (amino acids 1-70) does not. To determine whether endogenous interactions between TNAP and NIK occur, coimmunoprecipitation studies in HL60 cells were performed. The affinity-purified anti-TNAP antibody works well for immunoprecipitation (Fig. 1C). We show that the anti-TNAP antibody coimmunoprecipitates endogenous NIK in HL60 whole cell lysates (Fig. 2D). Taken together, these studies demonstrate that TNAP interacts with NIK.
Suppression by TNAP of NF-B Activation-NIK has been demonstrated to be important in NF-B activation (1, 4, 13, 49, 53-56). To test whether TNAP modulates NF-B activation, we performed NF-B-dependent reporter gene assays. Transient expression of TNAP alone does not affect NF-B-dependent gene expression but significantly inhibits TNF-␣-induced NF-B activation in HEK293T and Jurkat cells (Fig. 3A). However, TNAP does not affect interferon-␥-induced interferon response factor-1-dependent gene expression in either HEK293T or Jurkat cells (Fig. 3A), suggesting that the inhibition by TNAP of NF-B-dependent gene expression is specific. Stable clones constitutively expressing TNAP also exhibit significant inhibition of TNF-␣-induced NF-B activation (Fig. 3A).
To confirm the physiological function of endogenous TNAP in regulating NF-B activation, we used lentiviral-mediated RNA interference (46,57,58) to confer stable knockdown (59, 60) of HEK293T cells were cotransfected with pRK-FLAG-NIK and various TNAP mutants as indicated. After 24 h, TNAP was immunoprecipitated with anti-Myc or mouse control IgG, and NIK was detected by Western analysis using anti-FLAG antibody. The expression level of NIK and TNAP mutants were verified by Western blot and immunofluorescent (IF) staining. D, interaction of endogenous TNAP and NIK. Whole lysate of HL60 cells was immunoprecipitated with affinity-purified anti-TNAP antibody or control rabbit Ig G followed by Western blot with anti-NIK antibody.
the TNAP expression. The efficiency and specificity of lentiviral-mediated short hairpin RNA (shRNA) for TNAP were examined by Western blot analysis of the overexpressed Myctagged TNAP fusion protein and RT-PCR of the endogenous TNAP mRNA (Fig. 3C). Two of three constructs for TNAP shRNA efficiently reduce TNAP expression and one less efficiently (Fig. 3C). To examine the biological relevance of TNAP shRNA, we performed NF-B-dependent reporter gene assay. In addition, TNAP knockdown increases TNF-␣-induced NF-B-dependent gene expression (Fig. 3C). Taken together, these data suggest that TNAP is an inhibitor of TNF-␣-induced NF-B signaling.
TNAP Functions Upstream of IKK-It is well known that TNF-␣ induces NF-B activation through the TNFR1-TRADD-RIP-TRAF2-NIK-IKK signaling pathway (1, 4, 49, 54 -56). To test which step of the NF-B signaling pathway is the target of TNAP, we coexpressed these proteins together with TNAP and determined whether TNAP affects NF-B activation using luciferase reporter gene assay. As shown in Fig. 4A, transient expression of TNAP reduces TNFR1, TRADD, RIP, TRAF2, and NIK-mediated NF-B activation but does not affect IKK1 and IKK2-mediated NF-B activation. In addition, EMSA confirmed that TNAP reduces TRAF2 and NIK-mediated NF-B-DNA binding activity but does not affect IKK1-mediated NF-B-DNA binding activity (Fig. 4A). These results are consistent with the data presented in Fig. 2A and suggest that TNAP exerts its NF-B inhibitory effect by interfering with a component of NF-B signaling pathway upstream of IKK. Inhibition by TNAP of NIK Kinase Activity-Because TNAP interacts with NIK and blocks NIK-mediated NF-B activation upstream of IKK complex, we hypothesized that TNAP affects the kinase activity of NIK. IKK1 is a preferential substrate for NIK kinase (14,21,22). Overexpression of IKK1 induces potent autophosphorylation (22). Therefore, we selected a kinase-inactive form of IKK1 (IKK1(K44M)) as a substrate for NIK. To verify the possible effect of TNAP on NIK kinase activity, we first performed an in vivo kinase assay using 32 P-labeled inorganic phosphate and autoradiography (Fig. 4B, left upper  panel) and then a Western blot analysis using a specific antibody against phosphorylated IKK1 (Fig. 4B, left lower panel). As previously described (13,14,61,62), transient expression of NIK results in constitutive activation of IKK1 in HEK293T cells (Fig. 4B). Coexpression of TNAP markedly reduces the phosphorylation of IKK1 induced by expression of NIK (Fig.  4B). Stimulation with TNF-␣ increases NIK-mediated phosphorylation of IKK1. Coexpression of TNAP blocks TNF-␣induced NIK kinase activity to phosphorylate IKK1 (Fig. 4B,  left lower panel). To determine whether TNAP directly affects NIK kinase activity, we performed an in vitro kinase assay using the immunoprecipitated complex. As shown in Fig. 4B  (right upper panel), TNAP reduces TNF-␣-induced NIK kinase activity. However, TNAP alone induces a marginal increase in the constitutive NIK activity, which is inconsistent with the in vivo results. This inconsistency may reflect the different status of in vivo and in vitro conditions. To further determine whether endogenous TNAP inhibits the NIK-mediated phosphorylation of IKK1, we overexpressed NIK and IKK1(KM) in a stable knockdown cell line expressing lentiviral-mediated TNAP shRNA. Western blot analysis was performed using anti-phospho-IKK1 antibody to measure IKK1 transphosphorylation levels. As shown in Fig. 4B (right lower panel), NIK-mediated IKK1 phosphorylation is higher in the TNAP knockdown cells (Fig. 2C) than that in the control cells infected with lentivirus carrying empty shRNA vector. Taken together, these data indicate that TNAP suppresses NIK kinase activity.
Suppression by TNAP of p100 Processing-NIK is essential for p100 processing (10, 12, 28 -33). Because TNAP inhibits NIK kinase activity, we hypothesize that TNAP may affect NIK-mediated p100 processing. Consistent with previous reports, CD40L, LT-␣/␤, and LPS induce the processing of p100 inhibits NIK kinase activity. Wild-type or stable knockdown HEK293T cells were cotransfected with indicated vectors. For in vivo kinase assay, the cells were labeled with 32 P-labeled inorganic phosphate, and then immunoprecipitation and autoradiography was performed (left upper panel), or cell lysates were directly subjected to Western blot with anti-phospho-IKK1 antibody (left and right lower panels). For in vitro kinase assay, the cell lysates were subjected to immunoprecipitation followed by in vitro kinase reaction and autoradiography (right upper panel). The amount of immunoprecipitated IKK1 was determined by Western blot (WB) to the same blot. into p52 in HEK293T cells (10, 12, 28 -30, 63). Transient expression of TNAP inhibits the inducible processing of p100 (Fig.  5A). Overexpression of NIK induces robust processing of p100 to p52 (10,12,28,30,63). Expression of TNAP alone does not affect p100 processing but blocks NIK-mediated p100 processing (Fig. 5A, right upper panel). To examine whether endogenous TNAP regulates inducible processing of p100, we treated TNAP stable knockdown cells with CD40L, LT-␣/␤, and LPS. As shown in Fig. 5A (left lower panel), lentiviral-mediated TNAP shRNA increases p100 processing induced by CD40L, LT-␣/␤, and LPS. TNAP knockdown also enhances NIK-mediated p100 processing (Fig. 5A, right lower panel). These data suggest that TNAP inhibits p100 processing and targets NIK activity.
Inhibition by TNAP of p65 Phosphorylation-Phosphoryla-tion of p65 has been shown to regulate NF-B transactivation induced by cytokines, LPS, and LT-␣/␤ (34, 64 -68). Both IKK1 and IKK2 have been implicated in this process (34,35,66,67,69,70). Recent studies demonstrate that NIK-IKK1 cascade plays a crucial role in p65 phosphorylation (35,68). To address whether TNAP regulates p65 phosphorylation, we performed Western blot analysis of HEK293T cells that either overexpress or have a stable knockdown of TNAP. Overexpression of NIK induces p65 phosphorylation on Ser-536, which is reduced by transient overexpression of TNAP (Fig. 5B, left panel). TNF-␣ stimulation induces rapid phosphorylation of p65, consistent with several reports (65)(66)(67). Knockdown of TNAP increases TNF-␣-induced phosphorylation of p65 (Fig. 5B, right panel). These data suggest that TNAP is an inhibitory regulator for TNF-␣-induced NIK-mediated phosphorylation of p65. Suppression by TNAP of TNF-␣-induced Phosphorylation and Degradation of IBa-IB␣ phosphorylation and degradation is critical for NF-B activation by cytokines (71,72). To test whether TNAP affects TNF-␣-induced IB␣ degradation, we performed Western blot analysis in transient and stable TNAP cell lines. TNAP reduces TNF-␣-induced IB␣ degradation when transiently expressed in HEK293T cells (data not shown). Stable expression of TNAP stabilizes IB␣ at 5 and 15 min after TNF-␣ stimulation, and delays IB␣ degradation induced by TNF-␣ in all of the stable cell lines tested (Fig. 5C and data not shown). Stable knockdown of endogenous TNAP by lentiviral-mediated shRNA increases TNF-␣-induced Ser-32/36 phosphorylation of IB␣ (Fig. 5C, right panel). DISCUSSION In most cells, two separate but overlapping pathways regulate NF-B activation: the classical and alternative pathways. Both pathways rely on the IKK kinases for activation. However, each pathway has distinct activation requirements as well. For example, IKK2 and IKK-␥ are essential for the classical pathway, but IKK1 and NIK are indispensable for the alternative pathway (2, 6, 7, 10, 12, 28 -33). In this study we describe the cloning and initial characterization of a protein, TNAP, which regulates activation of both the classical and alternative pathways. We demonstrate that TNAP blocks NF-B activation upstream of the IKK complex. TNAP physically associates with NIK and attenuates NIK kinase activity. Subsequently, TNAP blocks p100 processing, p65 phosphorylation, and IB␣ degradation.
TNAP was cloned in a yeast two-hybrid screen of a human brain cDNA library using NIK as bait. The endogenous interaction between TNAP and NIK in mammalian cells was confirmed by coimmunoprecipitation studies. We also determined that TNAP interacts with TRAF2 and TRAF3 but not IKK1 or IKK2. Further studies using overexpression and small interference RNA knockdown of TNAP showed that TNAP negatively regulates NIK kinase activity. These observations suggest that TNAP suppresses NF-B activation by interfering with NIK-and/or TRAF-mediated signaling pathways.
Numerous in vitro studies have shown that NIK is an important intermediate in ligand-induced NF-B activation (13,14,61,62). However, genetic studies using a naturally occurring mutant, aly mice (23)(24)(25), and NIK gene knockout (26,27) mice demonstrate that NIK is not essential for TNF-␣-induced NF-B activation but is required for lymphotoxin-induced NF-B activation. It remains a possibility that NIK regulates NF-B activation in a cell type-and stimulus-specific manner (26,27,73). Recent studies have determined that NIK is indispensable for p100 processing, mediated by IKK1 (10, 12, 28 -33). However, the mechanism of NIK activation or inactivation has not been well elucidated. Our studies show that TNAP negatively regulates NIK signaling. TNAP is also an important addition to the family of negative regulators for NF-B activation, such as A20 (74), TANK (75), TTRAP (76), TRIP (77), ZIN (78), and CYLD (79 -81). Although our data demonstrate that TNAP inhibits NIK kinase activity and p100 processing, we have not elucidated the mechanisms. TNAP does not have any known phosphatase domains, nor does it share any sequence similarity with other domains that may explain its inhibitory effects on NF-B activation. Future studies will explore the mechanisms through which TNAP inhibits NIK kinase activity.
We provide evidence that TNAP retards TNF-␣-induced and NIK-mediated IB␣ degradation, suggesting that it also regulates the classical activation pathway. Because genetic studies strongly suggest that NIK is not an essential intermediate in TNF-␣-induced NF-B activation (26,27), TNAP may inhibit another intermediate in this classical pathway. Previous stud-ies have shown that TRAF2 is critically involved in NF-B activation induced by TNF-␣ (16,82) and CD40L (83)(84)(85)(86). We demonstrate that TNAP associates with TRAF2 and that it significantly inhibits TRAF2-mediated NF-B activation and CD40L signaling. Therefore, it is likely that TNAP-mediated inhibition of IB␣ proteolysis is due in part to its interaction with TRAF2 and perhaps TRAF3 (87,88). In addition to regulating TNF-␣ and TRAF2-dependent NF-B activation, TNAP may regulate additional signaling pathways activated by TNF-␣ and TRAF2 such as c-Jun N-terminal kinase and p38 (89). Future studies are planned to determine whether TNAP inhibits TNF-␣ and CD40L signaling through a TRAF2-dependent mechanism.
TNAP is a relatively small protein (16 kDa) with no known regulatory domains that could help explain its inhibitory mechanisms. TNAP shares a high degree of homology with several small proteins whose functions are unknown at this time. Therefore, TNAP may represent a novel family of small signaling molecules that function to regulate NF-B and other important signaling intermediates. Because TNAP regulates both the classical and alternative NF-B activation pathways, it may have potential for the development of therapies targeting diseases related to inappropriate NF-B activation such as cancer and inflammation (2).