Modulation of Tumor Necrosis Factor and Interleukin-1-dependent NF-κB Activity by mPLK/IRAK*

The innate immune response is an important defense against pathogenic agents. A component of this response is the NF-κB-dependent activation of genes encoding inflammatory cytokines such as interleukin-8 (IL-8) and cell adhesion molecules like E-selectin. Members of the serine/threonine innate immune kinase family of proteins have been proposed to mediate the innate immune response. One serine/threonine innate immune kinase family member, themouse Pelle-likekinase/human interleukin-1receptor-associated kinase (mPLK/IRAK), has been proposed to play an obligate role in promoting IL-1-mediated inflammation. However, it is currently unknown whether mPLK/IRAK catalytic activity is required for IL-1-dependent NF-κB activation. The present study demonstrates that mPLK/IRAK catalytic activity is not required for IL-1-mediated activation of an NF-κB-dependent signal. Intriguingly, catalytically inactive mPLK/IRAK inhibits type 1 tumor necrosis factor (TNF) receptor-dependent NF-κB activation. The pathway through which mPLK/IRAK mediates this TNF response is TRADD- and TRAF2-independent. Our data suggest that in addition to its role in IL-1 signaling, mPLK/IRAK is a component of a novel signal transduction pathway through which TNF R1 activates NF-κB-dependent gene expression.

The innate immune response is rapidly activated upon exposure to environmental stimuli and discriminates between self and nonself. The innate immune response may also be involved in determining whether an acquired immune response is required following pathogenic invasion (1). Studies in plants, insects, and mammals have revealed that serine/threonine innate immune kinase family members are components of the innate immune response (2). Serine/threonine innate immune kinase mediate early developmental decisions and, in adult tissue, mediate transactivation of genes whose products are involved in host defense. A Drosophila serine/threonine innate immune kinase family member, Pelle, is a maternal-effect gene that is also required for protection against fungal infections (2,3). Plant serine/threonine innate immune kinase family members include Pto and Pti-1, which mediate disease resistance in the tomato (4). Two human serine/threonine innate immune kinase family members, interleukin-1 receptor associated kinases IRAK and IRAK-2 have been linked to signaling through IL-1 1 receptor family members (5)(6)(7).
IRAK was identified by co-purification with the IL-1RI and was subsequently shown to bind the IL-1RAcP directly (5, 6, 9 -11). IRAK-2, like IRAK, is in a complex with IL-1RI and IL-1RAcP and is a component of a transduction pathway downstream of MyD88 (6). IRAK or IRAK-2 overexpression activates the NF-B-dependent E-selectin gene promoter (6,11). In response to IL-1 stimulation, the IRAK protein becomes phosphorylated (5,19). Whether IRAK phosphorylation is an autophosphorylation event or the result of phosphorylation by an independent kinase is not known. Nor is it known whether IRAK phosphorylation is an activation event and/or is required for IL-1-targeted degradation of IRAK by the proteosome (19). Two types of IRAK and IRAK-2 mutants have been described: mutants that consist of the amino terminus of IRAK or IRAK-2, and for IRAK, a mutant in which a lysine in the putative ATP binding site has been changed to serine (IRAKK239S). Overexpression of the amino terminus of IRAK or IRAK-2 inhibits IL-1-dependent activation of NF-B activity, thus implicating IRAK and IRAK-2 in IL-1 signaling (6,7,11,20). The effect of IRAKK239S overexpression on IL-1-dependent NF-B activation has not been reported.
We identified a mouse homologue of Pelle, a Drosophila serine/threonine innate immune kinase family member (21). Based upon sequence identity (5) and chromosomal location (22), mPLK is the mouse homologue of human IRAK. The mPLK protein contains intrinsic protein kinase activity (21). Although IRAK and IRAK-2 share sequence similarity, IRAK-2 lacks key residues thought to be critical for protein kinase activity. 2 This led us to question whether mPLK/IRAK protein kinase catalytic activity is essential in NF-B activation and IL-1 signaling. Dominant-negative alleles of serine/threonine protein kinases have been generated either by mutating the lysine in protein kinase subdomain II or the aspartic acid in protein kinase subdomain VII (23). In mPLK/IRAK, these sites correspond to amino acid residues Lys-239 and Asp-358, respectively. Because mutation of the corresponding lysine residue in other kinases does not inevitably result in loss of catalytic activity (24), we generated a catalytically inactive kinase by mutating the aspartic acid in protein kinase subdomain VII of mPLK/IRAK. We show here that introduction of a D358N mutation in subdomain VII of mPLK/IRAK abrogates the ability of mPLK/IRAK to induce NF-B activity. Furthermore, mPLK/IRAKD358N functions as a dominant-negative allele because it inhibits wild-type mPLK/IRAK activity. We have used this mutation to confirm a role for mPLK/IRAK in IL-1 signaling and to identify a role for mPLK/IRAK in the TNF RI signaling pathway.

EXPERIMENTAL PROCEDURES
Plasmid Construction and Plasmids-The IL-8-CAT and IL-8-LUC reporter constructs contained bp Ϫ1451-ϩ44 of the IL-8 gene (25) immediately upstream of either the bacterial chloramphenicol acetyltransferase gene (pCAT-Basic, Promega, Madison, WI) or the firefly luciferase cDNA (pGL3, Promega). The indicated mPLK cDNAs were subcloned into a mammalian expression vector that placed them under the control of the cytomegalovirus immediate-early gene promoter (pCMV). The mPLK construct contained the wild-type mPLK cDNA (amino acids 1-711; Ref. 21). The ⌬mPLK cDNA encoded amino acids 33-711. The cimPLK mutant contained a point mutation at amino acid 358 (D358N) and, in in vitro kinase assays, lacked catalytic activity. 2 The NIK plasmid construct contained the wild-type NIK cDNA, and ciNIK encoded NIKKK429 -430 amino acids (13). TRAF2 encoded the wild-type TRAF2 cDNA, ⌬TRAF2 encoded amino acids 87-501 (26), TRADD encoded the wild-type TRADD cDNA, and ⌬TRADD encoded amino acids 102-312 (27). The full-length mPLK cDNA was tagged with a myc epitope at the carboxyl terminus (Invitrogen, San Diego, CA); full-length NIK cDNA contained a carboxyl-terminal FLAG epitope. Epitope-tagged mPLK and NIK constructs were determined to be biologically active in preliminary studies). 2 Cell Culture and Transfections-Human embryonic kidney epithelial cells (293 cell line) were maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, penicillin/streptomycin, and glutamine. The C3H10T1/2 mouse embryo fibroblast cell line was maintained in Basal's modification of Eagle's media supplemented with glutamine, penicillin/streptomycin, and 10% fetal bovine serum. For transient transfection assays, C3H10T1/2 mouse embryo fibroblasts were plated at a density of 2 ϫ 10 5 cells/60-mm dish, and Ca 3 (PO 4 ) 2 precipitates were used to introduce plasmid constructs into cells as described previously (28). To monitor transfection efficiency, precipitates also included a reporter construct containing the luciferase gene under control of the herpes simplex virus thymidine kinase promoter (Tk-LUC; ATCC, Manassas, VA) or the ␤-galactosidase gene under control of the Rous sarcoma virus LTR (RSV-␤GAL; Ref. 29). Cultures were harvested 24 or 48 h after transfection. Individual assays were normalized for luciferase activity when IL-8-CAT reporter construct activity was examined or were normalized for ␤-galactosidase activity when IL-8-LUC reporter construct activity was examined. Luciferase (Promega, Madison, WI) and ␤-galactosidase (TROPIX, Inc., Bedford, MA) activities were assayed according to the manufacturer's specifications, and data are presented as a ratio between the IL-8-CAT and Tk-LUC or IL-8-LUC and RSV-␤GAL. Data are from two to three independent experiments performed in duplicate or triplicate with similar qualitative results.
Antibodies, Immunoprecipitation, and Western Blotting-Monoclonal mouse anti-c-myc antibody was purchased from Roche Molecular Biochemicals. The anti-FLAG M2 monoclonal antibody was purchased from Eastman Kodak Co. Chromatographically purified mouse IgG was purchased from Zymed Laboratories Inc. Laboratories (South San Francisco). Bacterially expressed HIS/T7-tagged mPLK was used as an immunogen to raise polyclonal antisera in rabbits (HRP Inc., Denver, PA). Crude antisera were affinity-purified using denatured mPLK protein immobilized on Immobilon-P membrane (Millipore).
The C3H10T1/2 mouse embryo fibroblast cell line was used in the mPLK-TNF RI co-immunoprecipitation assays. Approximately 5 ϫ 10 5 cells were seeded into 100-mm tissue culture dishes 24 h before treatment with recombinant mouse TNF␣ (100 units/ml) for 5 or 15 min. Cell monolayers were harvested, and immunocomplexing assays were performed as described below. For the mPLK-NIK co-immunoprecipitation assays, the human embryonic kidney epithelial cells (293 cell line; ATCC) were used. Approximately 2 ϫ 10 6 cells were seeded into 60-mm tissue culture dishes and grown in 5% CO 2 at 37°C. Plasmid constructs encoding myc-tagged mPLK and FLAG-tagged NIK were transfected into cells 16 h later by calcium phosphate precipitation (28). 48 h later, cell monolayers were harvested.
Cell monolayers were rinsed with phosphate-buffered saline and lysed in 1 ml of immunoprecipitation (IP) lysis buffer (10 mM HEPES, pH 7.4, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, and complete protease inhibitors). Cell debris was removed by centrifugation, and cell lysates were incubated with the indicated immunoprecipitation antibody. After a 16-h incubation at 4°C, protein A-Sepharose (10% (v/v) slurry) was added to antibody-containing cell lysates, and reactions were subject to an additional 2-h incubation at 4°C. Immunocomplexes, collected by centrifugation, were washed two times in IP wash buffer (same as lysis buffer, except [Triton X-100] was 0.1%). Washed material was resuspended in Laemmli buffer, denatured, and subjected to SDSpolyacrylamide gel electrophoresis in 8% reducing polyacrylamide gels. Separated proteins were transferred to Immobilon-P (Millipore) for Western analysis.

RESULTS
mPLK/IRAK Activates NF-B-Treatment of a variety of cell types with inflammatory cytokines, like IL-1␣ and TNF␣, results in transactivation of genes involved in mediating immune and inflammatory responses, including the IL-8 and E-selectin genes (25,30). Overexpression of human IRAK induces transactivation of the NF-B-dependent E-selectin gene promoter (11). Therefore we first confirmed that overexpression of mPLK, the mouse IRAK homologue, transactivates an NF-B-dependent gene promoter in mouse cells. Transient transfection of mouse embryo fibroblasts with a mammalian expression vector containing the mPLK cDNA led to approximately a 5-fold increase in the activity of the NF-B-dependent IL-8 gene promoter (Fig. 1A). This level of mPLK-mediated induction of NF-B activity is comparable with that detected when IRAK is overexpressed in human cells (11). These data are consistent with the proposal that mPLK, like its human homologue, IRAK, lies in a signaling pathway upstream of NF-B. Substitution of an asparagine for an aspartic acid residue (D358N) in the Mg 2ϩ -ATP binding site of mPLK created catalytically inactive mPLK protein (cimPLK). 2 In contrast to results obtained with mPLK, overexpression of cimPLK did not result in transactivation of the IL-8 gene promoter (Fig. 1A). A catalytically active mPLK mutant lacking the first 33 amino acids corresponding to helix 1 of the putative mPLK/IRAK death domain (⌬mPLK; Ref. 21) also did not transactivate the IL-8 gene promoter (Fig. 1A). Thus, mPLK/IRAK catalytic activity and an intact amino-terminal death domain are required for mPLK to induce transactivation of the IL-8 gene promoter.
To determine whether cimPLK functions in a dominant-negative manner, mouse embryo fibroblasts were co-transfected with mPLK and cimPLK. In a dose-dependent manner, cim-PLK inhibited the ability of mPLK to induce IL-8 promoter activity (Fig. 1B). Similar results were obtained with ⌬mPLK. 2 The IL-8 gene promoter, like the E-selectin promoter, contains NF-B and AP1 cis-acting elements (25,31). Thus, the observations described above do not define which of these cisacting elements is activated by mPLK. To verify that the mPLK effect is mediated through NF-B, the effect of mPLK overexpression on NF-B-dependent promoters was examined. Like the IL-8 gene promoter, overexpression of mPLK increased E-selectin promoter activity (4-fold; Fig. 1C) and stimulated the activity of a reporter construct under the control of tandem NF-B sites (see Fig. 3C; Ref. 32). Because the E-selectin and IL-8 promoters also contain binding sites for AP1 family members, we examined whether mPLK affected the activity of an IL-11 promoter construct known to be AP1-dependent (33). Although TNF␣ treatment led to a 4-fold increase in IL-11 promoter activity, overexpression of mPLK had no effect on IL-11 promoter activity (Fig. 1D). These data demonstrate that mPLK/IRAK lies in a signaling pathway upstream of NF-B.
The NF-B-inducing kinase, NIK, is a component of the type I IL-1 receptor (IL-1RI)-signaling cascade leading to NF-B activation (13). We therefore determined whether induction of the IL-8 gene promoter by mPLK required NIK. Transfection of mouse embryo fibroblasts with NIK led to an approximate 18-fold increase in IL-8 gene promoter activity (Fig. 2). Cotransfection with cimPLK or ⌬mPLK had no effect upon NIKmediated activation of the IL-8 gene promoter (Fig. 2). Malinin et al. (13) describe a dominant-negative allele of NIK, ciNIK (K429A,K430A), capable of blocking IL-1 or TNF␣-induced transactivation of the E-selectin gene promoter. However this construct has no effect on p65-induced E-selectin gene promoter activity. We examined whether ciNIK overexpression would affect the ability of mPLK to induce IL-8 gene promoter activity. Results of these assays suggested that mPLK/IRAKmediated transactivation of the IL-8 gene promoter requires NIK (Fig. 2).
IL-1 Signaling Does Not Require mPLK/IRAK Activity-NIK and mPLK/IRAK are thought to mediate activation of NF-B through the IL-1RI (11,13). IRAK-2, which lacks conserved residues in key protein kinase subdomains, can also modulate IL-1-induced NF-B activation (6). Therefore we determined whether mPLK/IRAK catalytic activity is necessary for IL-1-dependent activation of the IL-8 gene promoter. Overexpression of wild-type mPLK did not block IL-1 induction of IL-8 gene promoter activity (Fig. 3A). Interestingly, neither cimPLK nor ⌬mPLK decreased IL-1 induction of IL-8 promoter activity (Fig. 3A). We therefore examined whether the link between IL-1 and mPLK/IRAK activity in mouse embryo fibroblasts was similar to that described for mPLK/IRAK activity in human cells. Overexpression of amino-terminal IRAK (amino acids 1-208 or 1-215) or IRAK-2 (amino acids 1-96) mutants inhibit IL-1-dependent NF-B activation in human cells (6,7,11). Therefore we prepared an mPLK mutant that contained only amino-terminal residues 1-156. Overexpression of the mPLK mutant containing only amino-terminal residues 1-156 in mouse embryo fibroblasts blocked IL-1-dependent NF-B activation. 2 This result is consistent with that described for amino-terminal IRAK mutants expressed in human cells (6,7,11) and suggests that mPLK activity in mouse cells functions in a manner similar to that proposed for IRAK in human cells. However, when mouse embryo fibroblasts transfected with dominant-negative mPLK/IRAK were treated with IL-1␣ for varying times, higher levels of IL-8 promoter activity were detected as compared with IL-1-treated fibroblasts transfected with vector alone (Fig. 3B). To confirm that the enhanced response in IL-1-treated cultures overexpressing cimPLK was Thus, it appears that NF-B-dependent signaling through the IL-1RI can be blocked by overexpressing amino-terminal mPLK/IRAK residues or enhanced by overexpressing catalytically inactive mPLK/IRAK.
TNF RI Signaling Is mPLK/IRAK-dependent-NIK also mediates activation of NF-B through the TNF RI, which contains a death domain that associates with other death domain-containing proteins during signaling (13,34,35). The presence of a death domain motif in mPLK/IRAK (21) suggested that mPLK/IRAK may be a component of the TNF␣ as well as the IL-1-signaling pathways. Treatment of mouse embryo fibroblasts with TNF␣ results in approximately a 30-fold increase in IL-8 gene promoter activity (Fig. 3C). When mouse embryo fibroblasts transfected with mPLK are treated with TNF␣, no additional increase in IL-8 gene promoter activity is observed (Fig. 3C). However, when mouse embryo fibroblasts are transfected with cimPLK or ⌬mPLK, TNF␣-induced transactivation of the IL-8 gene promoter is decreased (Fig. 3C). The cimPLKand ⌬mPLK-mediated decrease in TNF␣-induced activation of the IL-8 gene promoter was dose-dependent; however, overex-pression of either cimPLK or ⌬mPLK did not completely abrogate a TNF-dependent signal. 2 Analysis of cells derived from TNF RI and TNF RII nulligenic animals revealed that TNFinduced NF-B activity is mediated solely through TNF RI (36). Consistent with this observation, cimPLK decreased IL-8 promoter activity in mouse fibroblasts treated with human TNF␣, 2 which on mouse cells signals exclusively through TNF RI (37). Taken together, these data demonstrate that mPLK/IRAK is a component of a TNF RI signaling pathway and provide further evidence for mPLK/IRAK as a component of the IL-1-signaling pathway.
IL-1-mediated activation of NF-B-dependent promoters is inhibited by a mutated version of TRAF6 (11,12). We found that dominant-inhibitory TRAF6 (⌬TRAF6289 -522) inhibits IL-1-dependent activation of the IL-8 gene promoter 10-fold but did not significantly inhibit TNF-or mPLK-dependent activation (1.4-fold decrease under either condition). 2 These results support those of Cao and co-workers (11,12), whereby similar amounts of the dominant-inhibitory TRAF6 reduced an IL-1dependent signal 10-fold and reduced TNF-and the IRAK-dependent signals 1.5-fold.
TNF-induced NF-B activity is mediated in part by TNF RI-associated death domain (TRADD) and two TRADD-recruited proteins: TNF receptor associated factor-2 (TRAF2) and the TNF receptor-interacting protein (RIP; Refs. 38 -41). To determine whether mPLK/IRAK is required for TRADD/ TRAF2/RIP-mediated activation of the IL-8 gene promoter, the effects of cimPLK and ⌬mPLK on TRADD and TRAF2-induced IL-8 gene promoter activity were measured. Neither cimPLK nor ⌬mPLK blocked the ability of TRADD (Fig. 4B) or TRAF2 (Fig. 4A) to induce activation of the IL-8 gene promoter.
We next determined if TRADD or TRAF2 were required for mPLK/IRAK-mediated activation of NF-B. In contrast to mPLK, which does not transactivate AP1-dependent promoters (Fig. 1D), TRAF2 activates AP1 as well as NF-B-dependent promoters (42). The IL-8 gene promoter, like the E-selectin gene promoter, contains AP1 and NF-B sites. Therefore, the requirement for TRADD and/or TRAF2 in mPLK/IRAK signaling was evaluated with a reporter construct that only contains NF-B sites ([NF-B] 3 -LUC; Ref. 32). Neither ⌬TRADD nor ⌬TRAF2, which decrease NF-B activation through TNF RI and TNF receptor family members (34,43), blocked mPLKmediated activation of the NF-B-dependent reporter (Fig. 4C). Transfection of mPLK with wild type or mutant TRAF2 (or TRADD) had an additive effect on induction of NF-B-dependent (Fig. 4C) as well as IL-8 gene promoter activity (Fig. 4, A  and B). A similar additive effect was detected when wild type or dominant-negative mPLK constructs were co-transfected with wild-type TRADD (or TRAF2; Fig. 4, A-C). These data indicate that mPLK/IRAK lies in a TRADD/TRAF2-independent signaling pathway.
TNF RI signaling in part is thought to be mediated by the recruitment of death domain-containing proteins, which bind to the carboxyl-terminal death domain in TNF RI (13,34,35). Because mPLK contains a death domain and overexpression of cimPLK decreases IL-8 promoter activity in mouse fibroblasts treated with human TNF␣, 2 which signals exclusively through TNF RI (37), we next determined whether the mPLK protein associates with TNF RI. TNF RI anti-sera was used to immunoprecipitate proteins from cell lysates prepared from mouse embryo fibroblasts that had been treated with TNF␣ for 5 or 15 min. SDS-polyacrylamide gel electrophoresis followed by Western blot analysis of the immunocomplexes revealed the presence of mPLK protein in the immunocomplexes generated with the TNF RI antisera in unstimulated cells. Furthermore the amount of mPLK protein associated with TNF RI appeared to increase in response to TNF treatment (Fig. 5). When mouse embryo fibroblasts were transfected with myc-tagged cimPLK, immunocomplexes generated with TNF RI anti-sera contained cimPLK (data not presented). Finally, similar studies were performed in the human 293 cell line, and myc-tagged mPLK was found in TNF RI-containing complexes (data not presented). Because cimPLK did not affect NIK activity (Fig. 2), these results suggest the ability of cimPLK to block a TNF signal is most likely mediated at the level of the TNF RI receptor.
mPLK/IRAK and TRAF family members function upstream of NIK (13,34). The WKI motif found in TRAF proteins is required for binding to NIK (34). Because mPLK/IRAK contains a similar motif (WHL; Refs. 5 and 21), we postulated that the mPLK/IRAK and NIK proteins may complex. To test this hypothesis, human embryonic kidney epithelial cells (293 cell line) were co-transfected with plasmids encoding myc-tagged mPLK and FLAG-tagged NIK. Myc or FLAG antisera were used to immunoprecipitate proteins from cell lysates. Immunoprecipitated proteins were subjected to SDS-polyacrylamide gel electrophoresis, followed by Western blot analysis. mPLK and NIK were detected in immunoprecipitates generated with either the FLAG or the myc antisera (Fig. 6, A and B). However, neither mPLK nor NIK were in immunocomplexes prepared with an unrelated mouse IgG (Fig. 6, A and B, lanes 1 and 2,  respectively). These data indicate that mPLK/IRAK and NIK can complex in cells. DISCUSSION mPLK/IRAK has been linked to signaling through IL-1 receptor family members (5-7) and has been shown to have protein kinase activity in vitro (21). However, the importance of mPLK/IRAK protein kinase activity in signaling has not been addressed. Frequently, mutations within the ATP binding site of protein kinases not only encode nonfunctional protein kinases but also interfere with the function of the wild-type protein (23). Thus, this mutation can be described as a dominant negative allele of mPLK/IRAK (44) and will be useful for further dissection of mPLK/IRAK function. We report here that an mPLK/IRAK mutant lacking catalytic activity (D358N) was unable to induce the activity of NF-B-dependent promoters. Furthermore, the D358N mPLK/IRAK mutant decreased the ability of wild-type mPLK/IRAK to activate an NF-B-dependent promoter in a dose-dependent fashion.
Interestingly, overexpression of cimPLK does not inhibit the ability of IL-1 to induce the activity of an NF-B-dependent promoter. In fact, overexpression of cimPLK enhances an IL-1-dependent signal. This result suggests that mPLK/IRAK catalytic activity is not required for its role in the IL-1-signaling pathway. IRAK-2 has been described as a relative of IRAK that can also enhance an IL-1 signal (6). IRAK-2 is quite similar to IRAK; however, it lacks key residues in several of the highly conserved protein kinase subdomains and is unlikely to be catalytically active. Thus cimPLK may enhance an IL-1 signal by mimicking IRAK-2. In this context, independent of catalytic activity, IRAK2 and/or cimPLK may subserve a scaffolding function and facilitate formation of signaling complexes. Alternatively, in response to IL-1, phosphatidylinositol 3-kinase activity is also increased (45), which, independent of IRAK/ mPLK, may effect changes in NF-B-dependent signaling. FIG. 5. Endogenous mPLK and TNF RI co-precipitate. Cell lysates were prepared from control, nontreated mouse embryo fibroblasts or mouse embryo fibroblasts treated with recombinant mouse TNF␣ (100 units/ml) for 5 or 15 min. Cellular proteins immunocomplexed with control mouse IgG or with TNF RI antisera were separated by SDSpolyacrylamide gel electrophoresis, Western blots were prepared and probed with mPLK or TNF RI antisera (see "Experimental Procedures" for details).
Enhancement of the IL-1-dependent signal in the presence of cimPLK/IRAK suggests mPLK/IRAK catalytic activity may negatively regulate IL-1-initiated signaling. mPLK/IRAK mutants that inhibit or enhance an IL-1-dependent signal may have therapeutic utility. Clearly, identification of targets that can be used to block IL-1-dependent signaling is important for down-regulating inflammatory responses. As important, however, may be the identification of targets that enhance/activate an inflammatory response in an otherwise immunocompromised host.
TRAF6 has been linked to IL-1 and mPLK/IRAK signaling (11,12). Although overexpression of a dominant-inhibitory TRAF6 mutant decreases IL-1 signaling, this mutant has a weaker inhibitory effect on IRAK/mPLK. In fact, mutant TRAF6 interferes with IRAK/mPLK and TNF similarly. Our own observations have confirmed these findings. Thus the inhibitory effect of the TRAF6 mutant is much more robust in the context of the IL-1-signaling pathway than in the mPLK/IRAK or TNF RI signaling pathways. Interestingly, TRAF6 was recently shown to complex with and affect signaling through the low affinity nerve growth factor receptor (46), another member of the TNF receptor superfamily (47).
In addition to a protein kinase catalytic domain, mPLK/ IRAK also contains an amino-terminal domain that resembles the death domain of proteins linked to TNF RI signaling (35). Indeed, endogenous mPLK and TNF RI proteins can be found complexed. Moreover, overexpression of cimPLK decreases the ability of TNF to induce the activity of NF-B-dependent promoters. Thus, in contrast to the IL-1-signaling pathway, the catalytic activity of mPLK/IRAK is critical for TNF signaling and suggests that mPLK/IRAK substrates are likely to be components of the TNF signaling pathway.
Overexpression of catalytically inactive NIK, the protein kinase that is believed to phosphorylate IB kinases (14 -18), blocks the activity of mPLK/IRAK. However, overexpression of wild-type NIK in the presence of catalytic inactive mPLK results in activation of NF-B-dependent promoters. These data suggest that mPLK/IRAK is upstream of NIK and suggests a TNF signaling pathway in which mPLK/IRAK is important for transmitting a signal from TNF RI to NIK. In support of this hypothesis, mPLK/IRAK protein can complex with NIK. mPLK/IRAK signaling is independent of the TNF-signaling molecules TRAF2 and TRADD, indicating that mPLK/IRAK represents a previously undescribed TNF RI signaling pathway. The latter observation is consistent with the analysis of TRAF2 nulligenic animals, which revealed that TNF R1 can mediate NF-B activation in a TRAF2-independent manner (43,48). These results suggest a model whereby TNF binding to TNF RI leads to activation of mPLK/IRAK protein kinase activity and the subsequent phosphorylation of mPLK/IRAK substrates, leading to the activation of NF-B.
Our data places mPLK/IRAK in the TNF RI signaling pathway and confirms a role for mPLK/IRAK in the IL-1-signaling pathway. These data also suggest that the regulation of mPLK/ IRAK activity may be more complicated than previously appreciated. In the IL-1 pathway, mPLK/IRAK catalytic activity is not required and, thus, is similar to the role of RIP in TNF signaling. Although cells lacking RIP are defective in TNF signaling, the defect can be reversed by expression of wild-type or catalytically inactive RIP (40,49). How RIP or mPLK/IRAK may transduce these effects is unclear. Our data suggests that TNF and IL-1 signaling may be coordinated at the level of mPLK/IRAK. In response to IL-1, the mPLK/IRAK protein is targeted for degradation (19). Interestingly, CD30 and TNF RII potentiate TNF RI-induced apoptosis by inducing the degradation of TRAF2 (50). It thus seems possible that mediation of cross-talk between the IL-1 and TNF RI signaling pathways may occur through the targeted degradation of mPLK/IRAK.