Research ArticleImmunology

Act1, a U-box E3 Ubiquitin Ligase for IL-17 Signaling

See allHide authors and affiliations

Science Signaling  13 Oct 2009:
Vol. 2, Issue 92, pp. ra63
DOI: 10.1126/scisignal.2000382

This article has a correction. Please see:

Abstract

Interleukin-17 (IL-17), a proinflammatory cytokine mainly produced by cells of the T helper 17 (TH17) lineage, is required for host defense against bacterial and fungal infections and plays a critical role in the pathogenesis of inflammatory and autoimmune diseases. Act1 is an essential adaptor molecule in IL-17–mediated signaling and is recruited to the IL-17 receptor (IL-17R) upon IL-17 stimulation through an interaction between its SEFIR domain and that of the IL-17R. Here, we report that Act1 is a U-box E3 ubiquitin ligase and that its activity is essential for IL-17–mediated signaling pathways. Through the use of the Ubc13-Uev1A E2 complex, Act1 mediated the lysine-63–linked ubiquitination of tumor necrosis factor receptor–associated factor 6 (TRAF6), a component of IL-17–mediated signaling. Deletion and point mutations of the Act1 U-box abolished Act1-mediated ubiquitination of TRAF6 and impaired the ability of Act1 to restore IL-17–dependent signaling and expression of target genes in Act1−/− mouse embryonic fibroblasts. We also showed that the lysine-124 residue of TRAF6 was critical for efficient Act1-mediated ubiquitination of TRAF6 and for the ability of TRAF6 to mediate IL-17–induced activation of nuclear factor κB. Thus, we propose that Act1 mediates IL-17–induced signaling pathways through its E3 ubiquitin ligase activity and that TRAF6 is a critical substrate of Act1, which indicates the importance of protein ubiquitination in the IL-17–dependent inflammatory response.

Introduction

The discovery of an inflammatory T helper 17 (TH17) cell population distinct from the classical TH1 and TH2 lineages has challenged the paradigm of T cell biology and provided new understanding about T cell–mediated immunity. Interleukin-17 (IL-17), a key proinflammatory cytokine mainly produced by TH17 cells, is required for host defense against extracellular microorganisms and contributes to the development and pathogenesis of inflammatory and autoimmune diseases (14). IL-17 is increased in abundance in many inflammatory conditions, such as multiple sclerosis, rheumatoid arthritis, lung airway infections, and psoriasis. IL-17–deficient mice display a reduced severity of experimental autoimmune encephalomyelitis and collagen-induced arthritis compared to that of wild-type mice, indicating the essential role of IL-17 in these inflammatory conditions. The main function of IL-17 is to coordinate local tissue inflammation by increasing the concentrations of inflammatory and neutrophil-mobilizing cytokines and chemokines in various cells, including fibroblasts, endothelial cells, epithelial cells, and astrocytes. Although studies have begun to unravel the complexity of IL-17–dependent signaling events (58), the precise molecular mechanisms involved in IL-17–mediated pathways remain unclear. Identification of intermediate signaling components and an understanding of their signaling mechanisms are crucial for the development of therapeutic strategies to block this major proinflammatory pathway.

IL-17 signals through a heteromeric receptor complex composed of IL-17 receptor A (IL-17RA) and IL-17RC, members of the IL-17R family (9, 10). Both IL-17RA and IL-17RC belong to a SEFIR protein family, which is defined by the presence of a conserved cytoplasmic SEFIR domain, a protein domain named for SEFs [similar expression to fibroblast growth factor (FGF) genes] and IL-17R. The SEFIR domain is responsible for homotypical interactions between proteins (11). Act1 is an essential component in IL-17 signaling and is required for IL-17–dependent immune responses (7, 8). Act1 contains two tumor necrosis factor (TNF) receptor–associated factor (TRAF)–binding sites (TBs), a helix-loop-helix domain at its N terminus, and a coiled-coil domain at its C terminus (12). Act1 has a SEFIR domain located within its coiled-coil region and, therefore, is also a member of the SEFIR protein family. Upon the binding of IL-17 to its receptor, Act1 is recruited to IL-17R through a SEFIR-dependent interaction, which is followed by the recruitment of transforming growth factor β–activated kinase 1 (TAK1) and the E3 ubiquitin ligase TRAF6, which mediate downstream signaling events (7, 13).

Protein ubiquitination is an important posttranslational modification that is required for many cellular functions (14, 15). Protein ubiquitination sequentially involves three types of enzyme: a ubiquitin-activating enzyme (E1), which activates ubiquitin in an adenosine triphosphate (ATP)–dependent process; a ubiquitin-conjugating enzyme (E2), which accepts activated ubiquitin from E1 to form an E2-ubiquitin thioester; and a ubiquitin protein ligase (E3), which binds to the E2 enzyme and to the substrate and mediates the formation of an isopeptide bond between the carboxyl terminus of ubiquitin and an ɛ-amino group of a lysine residue on the target protein. The E3 ubiquitin ligases, together with E2, determine the specificity of their substrates to mediate diverse biological functions. Three families of E3 ubiquitin ligases have been described: HECT (homology to E6AP C terminus), RING (really interesting new gene), and U-box. Different lysine residues of ubiquitin can be used to link to other ubiquitin molecules to form distinct polyubiquitin chains. Lys48 (K48)-linked polyubiquitination chains usually mark the substrate for proteasomal degradation. In contrast, Lys63 (K63)-linked polyubiquitination is involved in nonproteolytic functions, such as mediating protein-protein interactions and cell signaling.

Here, we found that Act1 contains a U-box–like region and is a member of the U-box type 3 ubiquitin ligase family. We showed that Act1 functioned as a previously unidentified U-box type E3 ubiquitin ligase that mediated K63-dependent ubiquitination of TRAF6, which was critical for IL-17–mediated signaling. Thus, Act1 represents the first example of how a U-box type E3 ligase may regulate immune responses through its effect on IL-17 signaling.

Results

Act1 is a U-box type E3 ubiquitin ligase

Although Act1 is an essential adaptor molecule in IL-17 signaling, the precise mechanisms by which it mediates IL-17 signaling remain to be defined. Thus, we searched for the effector domain(s) of Act1 important for IL-17–dependent signaling pathways. A Hidden Markov Model search of the SMART database (http://smart.embl-heidelberg.de) with the sequence of human Act1 identified a region between amino acid residues 273 and 338 as homologous to the U-box domain class of ubiquitin ligases. This portion of Act1 was aligned with selected U-box sequences (Fig. 1A). A number of residues that are part of the hydrophobic core in known U-box structures are conserved in Act1. Based on this alignment, we used the Swiss-Model interface (http://swissmodel.expasy.org) (16) to thread the Act1 sequence onto the template U-box structure from the C terminus of heat shock protein of 70 kD (Hsp70)–interacting protein (CHIP) (17, 18). As expected from the sequence conservation, several key hydrophobic residues (Leu303 and Leu324, Fig. 1B) constituted part of the hydrophobic core in this model. In addition, the model predicted that several surface-exposed residues involved in interactions between U-boxes as defined by CHIP and E2 enzymes, including Pro318, Val319, His282, and Lys321, are conserved in Act1 (Fig. 1B). These residues are putatively located in two long loops (loop 1 and loop 2), which define much of the interaction surface between U-boxes and E2 enzymes.

Fig. 1

Act1 is a U-box type E3 ligase. (A) Alignment of the U-box sequences in Act1 (TRAF3IP2, NP_671733) and other U-box–containing proteins. The conserved amino acid residues are boxed. (B) The proposed structure of the U-box of Act1 superimposed on that of CHIP. (C) In vitro polyubiquitination assays with GST-Act1 as the E3 ubiquitin ligase in combination with different E2 enzymes. (D) In vitro polyubiquitination assays with different controls. (E) GST-Act1 and Ubc13-Uev1A were used in polyubiquitination assays in combination with various ubiquitin mutant proteins. (F) All purified GST fusion Act1 proteins (10 μg) used in this study were run on SDS gels and stained with Coomassie Blue. Different domains of Act1 (G), U-box deleted Act1 (D-Ubox) (H), or the U-box region and U-box mutants (I) were used as E3 enzymes in in vitro polyubiquitination assays. The polyubiquitination reactions were analyzed by Western blotting with an antibody specific for ubiquitin. All experiments were repeated at least three times and representative blots are presented.

To explore the potential role of Act1 as an E3 ubiquitin ligase, in vitro ubiquitination assays were performed with different E2 enzymes. Act1 used the Ubc13-Uev1A E2 complex to specifically catalyze polyubiquitination in this in vitro assay, which suggested that Act1 had E3 ubiquitin ligase activity (Fig. 1C). The specificity of this polyubiquitination assay was confirmed by different control reactions (Fig. 1D). We noticed that Act1-mediated polyubiquitination was as efficient as that of TRAF6, a known E3 ligase. Although K48-mediated polyubiquitination targets a protein for proteasomal degradation, polyubiquitination chains linked through K63 of ubiquitin mediate protein-protein interactions and cell signaling (14). Therefore, it was important to determine whether Act1-mediated polyubiquitination was linked through Lys48 or Lys63 of ubiquitin. Mutants of ubiquitin, including K48 (in which all Lys residues except Lys48 were mutated to Arg), K63 (in which all Lys residues except Lys63 were mutated to Arg), K48R (in which Lys48 was mutated to Arg), and K63R (in which Lys63 was mutated to Arg), were examined for their ability to form polyubiquitin chains in vitro with purified Act1 as the E3 ligase and Ubc13-Uev1A as the E2 enzyme. Act1 was able to mediate polyubiquitination of the ubiquitin mutants K63 and K48R but not of K48 and K63R (Fig. 1E), which suggested that Act1 used Ubc13-Uev1A as an E2 enzyme and mediated polyubiquitination of target proteins through Lys63 of ubiquitin. This further implied that Act1 functions to enhance protein-protein interactions and cell signaling.

To test whether the predicted U-box of Act1 was responsible for its E3 ubiquitin ligase activity, different deletion mutants of Act1 were generated. Purified glutathione-S-transferase (GST)–tagged wild-type and mutant Act1 proteins were analyzed by SDS–polyacrylamide gel electrophoresis (SDS-PAGE) (Fig. 1F). Whereas neither the N terminus (residues 1 to 200) nor the C terminus (residues 400 to 574) of Act1 showed E3 ubiquitin ligase activity, the middle region (residues 200 to 400) showed E3 activity in the in vitro polyubiquitination assay (Fig. 1G). On the other hand, deletion of the U-box (D-Ubox, residues 250 to 350 deleted) abolished the E3 activity of Act1 (Fig. 1H). Together, these results suggested that the predicted U-box was responsible for Act1-mediated E3 activity.

Consistent with the deletion analysis, the purified recombinant protein of the U-box–like region (residues 250 to 350) indeed contained E3 activity (Fig. 1I). Structural modeling of the U-box of Act1 with that of CHIP predicted several key hydrophobic residues (including Leu303 and Leu324) as part of the hydrophobic core in the U-box–like region of Act1 (Fig. 1B). Whereas mutation of Leu303 to Arg completely abolished the E3 activity of the U-box–like domain, mutation of Leu324 to Arg greatly reduced the E3 activity. In addition, the model predicted that several surface-exposed residues involved in interactions between the U-box region as defined by CHIP and E2 enzymes (17, 18) were conserved in Act1 (Fig. 1B), including Pro318 and Val319. Whereas mutation of Pro318 to Gly reduced the E3 activity of the U-box region of Act1, mutation of Val319 to Arg completely abolished its activity (Fig. 1H). These structural and functional data support the notion that the E3 activity of Act1 is likely conferred by its U-box–like structure.

The U-box domain of Act1 is essential for IL-17–mediated signaling and gene expression

The expression of several known IL-17 target genes and their proteins, including the chemokine KC, granulocyte-macrophage colony-stimulating factor (GM-CSF), and IL-6 were analyzed by real-time reverse transcription–polymerase chain reaction (RT-PCR) assays and by enzyme-linked immunosorbent assay (ELISA) of Act1−/− mouse embryonic fibroblasts (MEFs) reconstituted with wild-type Act1, the D-Ubox mutant of Act1, or the point mutant L303G after treatment with tumor necrosis factor–α (TNF-α), IL-17, or both (Fig. 2A). Gene expression induced by IL-17 alone or in synergy with TNF-α was greatly reduced in Act1−/− MEFs reconstituted with the D-Ubox mutant of Act1 compared with that of Act1−/− MEFs reconstituted with wild-type Act1 (Fig. 2, B and C). The L303G mutant Act1, which had no E3 activity (Fig. 1I), also showed an impaired ability to restore IL-17–mediated gene expression in Act1−/− MEFs (Fig. 2, B and C). Specific IL-17–mediated signaling events were also examined in Act1−/− MEFs reconstituted with wild-type Act1 or its D-Ubox mutant. Although wild-type Act1 was able to restore the IL-17–induced phosphorylation and degradation of inhibitor of nuclear factor κB (NF-κB) α (IκBα) and the phosphorylation of c-Jun N-terminal kinase (JNK) and extracellular signal–regulated kinase (ERK) in Act1−/− MEFs, the D-Ubox mutant showed much reduced activity (Fig. 2C). These data show that the U-box domain of Act1 is required for IL-17–mediated signaling, implicating U-box–mediated E3 ubiquitin ligase activity in Act1-mediated IL-17 signaling.

Fig. 2

The U-box domain is essential for IL-17–mediated signaling and the expression of target genes. Act1−/− MEFs were reconstituted with empty vector or with FLAG-tagged mouse wild-type (WT) Act1, the D-Ubox Act1 mutant, or the L303 Act1 mutant by retroviral infection. Forty-eight hours after infection, cells were treated with TNF-α (10 ng/ml), IL-17 (50 ng/ml), or both for 3 hours. (A) Western blotting analysis of whole-cell extracts for the detection of mouse Act1 (WT) or its D-Ubox or L303G mutants. (B) The abundances of KC, GM-CSF, and IL-6 mRNAs were measured by real-time RT-PCR. Results were calculated as arbitrary units after normalization to the abundance of glyceraldehyde-3-phosphate dehydrogenase(GAPDH) mRNA. The experiment was repeated five times and the data are shown as the mean arbitrary units ± SEM. *P < 0.05. (C) The abundance of KC, GM-CSF, and IL-6 proteins in the cell supernatant were measured in triplicate by ELISA. The experiment was repeated five times and the data are shown as the mean concentration ± SEM. *P < 0.05. (D) Act1−/− MEFs reconstituted with either empty vector, FLAG-tagged mouse wild-type Act1, or its D-Ubox mutant treated with IL-17 and analyzed by Western blotting with the indicated antibodies. Experiments were repeated at least three times, and representative blots are presented.

Act1 ubiquitinates TRAF6 through Lys63 in a U-box–dependent manner

We then attempted to identify the substrate(s) of the U-box E3 ubiquitin ligase activity of Act1. Previous findings directed our effort toward exploring TRAF6 as a potential substrate. First, TRAF6 is an important component in IL-17 signaling (13). IL-17–induced activation of NF-κB (as determined by measurement of the phosphorylation of IκBα and of NF-κB–dependent luciferase reporter activity) and phosphorylation of JNK were abolished or greatly reduced in Traf6−/− MEFs compared to those in wild-type MEFs, although IL-17 still induced the phosphorylation of ERK in the absence of TRAF6 (Fig. 3, A and B). Second, IL-17 induced interactions between IL-17R and certain modified forms of TRAF6 (as detected through apparent mobility shifts on Western blots), as well as with Act1 in HeLa cells overexpressing the IL-17R (7) (Fig. 3C). Immunoprecipitation of TRAF6 from cell lysates and analysis by Western blotting with an antibody against ubiquitin showed that IL-17 induced the ubiquitination of TRAF6 (Fig. 3D). Furthermore, we also found that IL-17 induced the ubiquitination of TRAF6 in wild-type MEFs, whereas deficiency in Act1 completely abolished this process (Fig. 3E).

Fig. 3

Act1 ubiquitinates TRAF6 through Lys63 in a U-box–dependent manner. (A) WT and Traf6−/− MEFs were treated with IL-17, and whole-cell lysates were analyzed by Western blotting with the indicated antibodies. (B) WT and Act1−/− MEFs were treated with IL-17 and analyzed by NF-κB luciferase assay. The experiment was repeated five times, and the data are shown as the mean luciferase activity ± SEM. *P < 0.05. HeLa cells transfected with a plasmid encoding FLAG-tagged IL-17R were treated with IL-17 and samples were immunoprecipitated (IP) with an antibody against FLAG (M2) (C) or an antibody against TRAF6 (D), followed by Western blotting analysis with the indicated antibodies. (E) Cell lysates of WT or Act1−/− MEFs treated with IL-17 were used in immunoprecipitation assays with an antibody against TRAF6 and analyzed by Western blotting with the indicated antibodies. (F) GST-fusion proteins of Act1, the U-box of Act1, or the D-Ubox mutant of Act1 were used as E3 ubiquitin ligases in ubiquitination reactions with GST-TRAF6-dR as the substrate, followed by Western blotting analysis of samples with an antibody against TRAF6. (G) In vitro ubiquitination assays with GST-TRAF6-dR as the substrate and different ubiquitin mutant proteins. (H) HEK 293 cells were cotransfected with plasmids encoding FLAG-tagged TRAF6 and HA-tagged ubiquitin (Ub) or myc-tagged Act1. Cell lysates were immunoprecipitated with an antibody against FLAG. (I) Cell lysates of Act1−/− MEFs reconstituted with either control vector or plasmids encoding FLAG-tagged WT Act1 or its D-Ubox mutant were treated with IL-17 and samples were immunoprecipitated with an antibody against TRAF6 and analyzed by Western blotting with the indicated antibodies. All experiments were repeated at least three times, and representative blots are presented. WCE, whole-cell extract.

TRAF6 is itself a RING-type E3 ubiquitin ligase, and by interacting with the Ubc13-Uev1A E2 complex, it catalyzes the formation of Lys63-linked polyubiquitination chains (19, 20). TRAF6 is also ubiquitinated upon stimulation of cells with IL-1R and Toll-like receptor (TLR) ligands (19, 20), a modification that is crucial for TRAF6 to mediate the recruitment and ubiquitination of downstream signaling molecules [including components of the TAK1 complex and IκB kinase (IKK) γ] that lead to the activation of NF-κB and JNK. The current view holds that TRAF6 can act as an E3 to ubiquitinate itself (“auto-ubiquitination”); however, a study has shown that TRAF6 undergoes poor auto-ubiquitination in the presence of Uev1A (21). The in vitro polyubiquitination assay (with the Ubc13-Uev1A E2 complex) showed that Uev1A favored the polyubiquitin chain synthesis reaction rather than substrate modification on TRAF6 itself. Thus, we hypothesized that there should exist another E3 upstream of TRAF6 that would facilitate its ubiquitination.

To test whether Act1 functioned as an E3 ligase to ubiquitinate TRAF6, we first examined whether Act1 could directly ubiquitinate TRAF6 in an in vitro polyubiquitination assay. Because TRAF6 functions as E3 ubiquitin ligase, we used a bacterially expressed mutant form of TRAF6 from which the RING domain was deleted (TRAF6-dR) as the substrate. The ubiquitination of TRAF6 was catalyzed by either wild-type Act1 or its U-box region, whereas deletion of the U-box domain of Act1 (the D-Ubox mutant) abolished the ability of Act1 to ubiquitinate TRAF6 (Fig. 3F). Furthermore, in vitro polyubiquitination assays with different ubiquitin mutants (K48, K63, K48R, and K63R) showed that Act1 mediated the Lys63-linked ubiquitination of TRAF6 (Fig. 3G). Overexpression of wild-type Act1 led to increased ubiquitination of TRAF6 compared to that in control cells, whereas the D-Ubox mutant of Act1 failed to induce ubiquitination (Fig. 3H), suggesting that Act1 can ubiquitinate TRAF6 in cells and that the U-box domain of Act1 is indeed critical for this process.

To examine the requirement of the U-box domain for IL-17–mediated ubiquitination of TRAF6, Act1−/− MEFs were reconstituted with vector alone, wild-type Act1, or its D-Ubox mutant and treated with IL-17 at different times. IL-17–induced ubiquitination of TRAF6 was completely abolished when the U-box was deleted, suggesting that the U-box domain of Act1 is crucial for ubiquitination of TRAF6 (Fig. 3I). We found that Act1 contained two putative TRAF-binding sites, including TB1 (residues 38 to 42) and TB2 (residues 333 to 337). Whereas mutation of TB1 or TB2 had minimal effect on the interaction of Act1 with TRAF6, mutation of both TB1 and TB2 (TB1+2 mutant) led to a substantial reduction in the interaction between Act1 and TRAF6 (fig. S1A). IL-17–induced activation of NF-κB and phosphorylation of JNK were greatly reduced in Act1−/− MEFs reconstituted with the TB1+2 mutant compared with Act1−/− MEFs reconstituted with wild-type Act1 (fig. S1, B and C). Moreover, IL-17–mediated ubiquitination of TRAF6 was also greatly impaired in Act1−/− MEFs reconstituted with the TB1+2 mutant (fig. S1D). These data collectively indicate that Act1 can mediate Lys63-linked ubiquitination of TRAF6 in cells.

Act1 promotes Lys63-linked polyubiquitination of TRAF6 at Lys124

Lys124 of TRAF6 is the primary ubiquitin acceptor site for the polyubiquitination of TRAF6 (22); mutation of Lys124 of TRAF6 impairs the activation of NF-κB in response to IL-1 or RANKL (receptor activator of NF-κB ligand). To address whether Lys124 of TRAF6 was also important for IL-17–mediated activation of NF-κB, the phosphorylation and degradation of IκB were examined in IL-17–treated Traf6−/− MEFs reconstituted with wild-type TRAF6, TRAF6-K124R, or TRAF6-C70A (Fig. 4A). The C70A mutant of TRAF6 is a RING domain point mutant that abolishes the E3 ligase activity of TRAF6 (22) (Fig. 4D). The phosphorylation and degradation of IκB were considerably impaired in Traf6−/− MEFs reconstituted with either the TRAF6-K124R or the TRAF6-C70A mutants, which suggested that Lys124 and the ligase activity of TRAF6 were critical for IL-17–mediated activation of NF-κB. The impaired NF-κB activity associated with the TRAF6-K124R and TRAF6-C70A mutants was also confirmed by NF-κB luciferase reporter assays (Fig. 4B). Furthermore, bacterially expressed GST fusion proteins of wild-type TRAF6 and its mutants (K124R and C70A) were used as substrates for in vitro polyubiquitination assays (Fig. 4, C and D). Act1 indeed promoted the ubiquitination of TRAF6 and TRAF6-C70A but not that of TRAF6-K124R (Fig. 4C), suggesting that Act1 polyubiquitinates TRAF6 primarily on Lys124.

Fig. 4

Act1 promotes Lys63-linked polyubiquitination of TRAF6 at Lys124. (A) Traf6−/− MEFs reconstituted with control vector or plasmids encoding WT TRAF6 or its mutants (K124R or C70A) were treated with IL-17 and analyzed by Western blotting with the indicated antibodies. (B) Traf6−/− MEFs reconstituted with either control vector or plasmids encoding WT TRAF6 or its mutants (K124R or C70A) were analyzed by NF-κB luciferase assay. The experiment was repeated five times. The data are shown as the mean luciferase activity ± SEM. *P < 0.05. GST-fusion proteins of WT TRAF6 or of its mutants (K124R or C70A) were used as substrates in ubiquitination assays, followed by Western blotting analysis with antibodies against TRAF6 (C) or ubiquitin (D). Experiments were repeated at least three times, and representative blots are presented.

Discussion

Here, we have shown that Act1, a receptor-proximal component of the IL-17 signaling pathway, is a U-box–type E3 ubiquitin ligase. The U-box–mediated E3 ligase activity of Act1 was crucial for IL-17–induced downstream signaling events, including the activation of NF-κB, JNK, and ERK, and for the expression of genes encoding proinflammatory factors, including KC, GM-CSF, and IL-6, thus providing a molecular mechanism for Act1-mediated IL-17R signaling. Specifically, Act1 mediated Lys63-linked ubiquitination of TRAF6, an important component of the IL-17R signaling pathway. We showed that Act1-mediated ubiquitination of TRAF6 occurred on Lys124 of TRAF6, and that this was required for IL-17–induced activation of NF-κB.

The U-box of Act1 (amino acid residues 273 to 338) was identified through a Hidden Markov Model search of the SMART database, followed by sequence alignment with the U-boxes of known structure from CHIP [PDB code 2C2V (17, 18), Prp (23), and AtPUB (24)]. The alignment was used to generate a model structure for the Act1 U-box with the Swiss-Model interface (16). The overall reasonableness of the model geometry and conformation were verified by a number of parameters, including pseudo-energies calculated by QMEAN (−1204 kJ/mol) (25) and DFIRE (−3250 kJ/mol) (26). Measurements of local structure quality (ANOLEA and GROMOS energies) supported these numbers, especially in the core regions of the fold. The mean S score of the model (calculated in ProQRes) (27) was 0.65, which is considered reasonable given the relatively low sequence homologies among the template structures and Act1. The overall sequence analysis and structural modeling provided sufficient confidence in the identity of the U-box domain of Act1. Our studies showed that this predicted U-box region was indeed able to confer an E3 ubiquitin ligase activity to Act1 in vitro. Although the removal of the U-box abolished the E3 ligase activity of Act1, the U-box alone was sufficient to mediate polyubiquitination. Point mutations of the key residues in the U-box (predicted by sequence alignment and structural modeling with known U-box–containing E3 structures from CHIP) abolished its E3 activity, confirming the specificity of the Act1 U-box domain. Together, our results suggest that Act1 contains a U-box with potent E3 ubiquitin ligase activity.

Act1 is an essential component in the IL-17R signaling pathway and is required for IL-17–dependent immune responses (7, 8). IL-17 signals through a heteromeric receptor complex composed of IL-17RA and IL-17RC. Act1 is recruited to the IL-17R through its SEFIR domain, and we have now identified the U-box as a second functional domain of Act1. Whereas the SEFIR domain is an interaction domain, the U-box domain of Act1 is an effector domain required for Act1 to transduce signals from the IL-17R. Inactivation of the E3 ligase activity of Act1 diminished the ability of Act1 to mediate IL-17–induced activation of NF-κB, JNK, and ERK. Specifically, we identified TRAF6 as one critical ubiquitination substrate of Act1. U-box–mediated ubiquitination of TRAF6 by Act1 was required for the IL-17–dependent activation of NF-κB.

TRAF6 is a central signaling molecule that is used by various receptors, including the IL-1R, TLRs, and CD40 (19, 20, 2830). TRAF6 plays an important role in the activation of TAK1, which leads to the activation of IKK and NF-κB. Upon stimulation of TLRs or IL-1R, TRAF6 undergoes K63-linked ubiquitination. The TAK1-binding proteins, TAB2 and TAB3, bind preferentially to polyubiquitin chains on TRAF6 through a highly conserved zinc finger domain (31). TRAF6 is a RING-domain containing E3 ubiquitin ligase, and it mediates the ubiquitination and subsequent activation of TAK1 and the IKK complex (32, 33). In this study, we found that the RING domain-deleted mutant TRAF6, TRAF60 C70A, failed to restore IL-17–induced phosphorylation and degradation of IκB and NF-κB–dependent luciferase activity in Traf6−/− MEFs. Our results show that the E3 ubiquitin ligase activity of TRAF6 is also critical for IL-17–mediated activation of NF-κB.

The critical question has been whether TRAF6 undergoes self-ubiquitination or another E3 ligase is required for its ubiquitination (22). Our results showed that TRAF6 was not efficiently ubiquitinated by itself when incubated with the Ubc13-Uev1A E2 complex in vitro. This observation is consistent with a study that showed that TRAF6 is unable to ubiquitinate itself in the presence of Uev1A (21). The addition of Act1 to the reaction promoted the modification of TRAF6, and the U-box of Act1 was required for Act1-mediated ubiquitination of TRAF6 in vitro and in cells. Furthermore, we mapped Lys124 of TRAF6 as the primary acceptor site for Act1-mediated ubiquitination and found that Lys124 of TRAF6 was required for IL-17R–mediated signaling. IL-17–dependent ubiquitination of TRAF6 and the activation of NF-κB were greatly impaired in Traf6−/− MEFs reconstituted with the TRAF6 K124R mutant protein. Based on these findings, we propose the following model for IL-17–induced, Act1- and TRAF6-mediated activation of NF-κB. Act1 is recruited to IL-17R upon its stimulation with IL-17, which is followed by the recruitment of TRAF6, which is ubiquitinated by Act1. Polyubiquitinated TRAF6 then recruits the complex containing TAK1 and TAB2 or TAB3 and mediates the ubiquitination and activation of the TAK1 and IKK complexes, which results in the activation of NF-κB.

Although deficiency in TRAF6 abolished IL-17–induced activation of NF-κB and JNK in MEFs, IL-17–mediated phosphorylation of ERK was intact; however, the U-box region of Act1 was required for IL-17–induced activation of NF-κB, JNK, and ERK. Furthermore, whereas Act1 is required for IL-17–mediated stabilization of messenger RNA (mRNA) (34), TRAF6 is dispensable for this posttranscriptional regulation. These results suggest that the E3 ubiquitin ligase activity of Act1 is probably required for more IL-17–mediated signaling events than those mediated by TRAF6, which implies the presence of additional Act1 substrates in the IL-17 signaling pathway.

In summary, our study shows that Act1 is a previously unidentified U-box–type E3 ubiquitin ligase whose activity is critical for IL-17–mediated, TRAF6-dependent activation of NF-κB. Furthermore, our study showed that TRAF6 could only poorly ubiquitinate itself and that Act1 functioned as an E3 ligase to efficiently promote the polyubiquitination of TRAF6. Modified TRAF6 then recruits and ubiquitinates downstream signaling molecules (components in the TAK1 and IKK complex), leading to the activation of NF-κB. Thus, Act1-mediated ubiquitination of TRAF6 represents a new regulatory mechanism in IL-17 signaling pathway.

Materials and Methods

Reagents and cell culture

The following recombinant mutant forms of ubiquitin proteins were obtained from Boston Biochem: K48 (in which all Lys residues except Lys48 were mutated to Arg), K63 (in which all Lys residues except Lys63 were mutated to Arg), K48R (in which Lys48 was mutated to Arg), and K63R (in which Lys63 was mutated to Arg). Ubiquitin and the antibodies against FLAG (M2) were obtained from Sigma. Antibodies against TRAF6, ubiquitin, HA, myc, actin, and IκB were obtained from Santa Cruz Biotechnology. Antibodies against phospho-IκB, phospho-JNK, and phospho-ERK were from Cell Signaling. HEK 293 cells and MEFs were cultured as previously described (7).

Plasmids

FLAG-tagged TRAF6 and myc-tagged Act1 were cloned into the pCMV vector for transient expression. Human Act1, truncated Act1 constructs including amino acid residues 1 to 200, 200 to 400, 400 to 574, the D-Ubox (in which residues 250 to 350 were deleted), the U-box (residues 250 to 350), and point mutants in the U-box of Act1 including L303R, P308G, V319R, and L324R, as well as wild-type TRAF6 and its mutants [TRAF6-K124R, TRAF6-C70A, and TRAF6-dR (residues 1 to 108 deleted)] were cloned into pGEX-KG for the production and purification of GST-fusion proteins. Mouse Act1, its D-Ubox, and the TB1, TB2, and TB1+2 mutants of Act1, as well as wild-type TRAF6 and its point mutants (TRAF6-K124R and TRAF6-C70A) were cloned into pMSCV-IRES-GFP for retroviral infection.

In vitro ubiquitination assay

The polyubiquitination assays were performed in a 10-μl reaction volume in buffer [20 mM tris-HCl (pH 7.5), 2 mM ATP, 5 mM MgCl2] at 37°C for 1 hour, together with the following components provided accordingly: 100 ng of E1 enzyme, 100 ng of E2 enzyme, 100 ng of individual E3 enzymes, and 5 μg of ubiquitin or of mutant ubiquitin proteins. To assay Act1-mediated polyubiquitination of TRAF6, in addition to E1, E2, and ubiquitin, 100 ng of GST-TRAF6-dR or of its mutants (K124R or C70A) were added as substrates, together with or without wild-type Act1 or Act1 mutants in the same reaction buffer.

Purification of GST-fusion proteins

The GST-fusion proteins were purified by affinity chromatography through Glutathione Sepharose 4 Fast flow beads (Amersham Biosciences) according to the manufacturer’s protocol.

Immunoprecipitations and luciferase assays

Cells were lysed in buffer A [0.5% Triton X-100, 20 mM Hepes (pH 7.4), 150 mM NaCl, 12.5 mM β-glycerophosphate, 1.5 mM MgCl2, 10 mM NaF, 2 mM dithiothreitol, 1 mM sodium orthovanadate, 2 mM EGTA, 20 μM aprotinin, 1 mM phenylmethylsulfonyl fluoride] or, to prevent possible protein-protein interactions, in buffer B (buffer A with 0.1% SDS and 0.5% deoxycholate). Cell extracts were incubated with 1 μg of the appropriate antibodies overnight at 4°C with 20 μl of protein A Sepharose beads. After incubation, beads were washed four times with lysis buffer, resolved by SDS-PAGE, and analyzed by Western blotting. NF-κB luciferase reporter assays were performed as previously described (7).

Retroviral infections and real-time RT-PCR assays

Retroviral infections and complementary DNA (cDNA) synthesis were performed as described previously (7). Real-time RT-PCR assays were performed with the SYBER Green PCR Master Mix kit (Applied Biosystems). The primers used were as follows: KC, 5′-TAGGGTGAGGACATGTGTGG-3′ (forward) and 5′-AAATGTCCAAGGGAAGCGT-3′ (reverse); IL-6, 5′-GGACCAAGACCATCCAATTC-3′ (forward) and 5′-ACCACAGTGAGGAATGTCCA-3′ (reverse); GM-CSF, 5′-GGCCTTGGAAGCATGTAGAGG-3′ (forward) and 5′-GGAGAACTCGTTAGAGACGACTT-3′ (reverse).

Statistical analyses

The statistical analyses were performed by one-way analysis of variance, followed by multiple pairwise comparisons of the treatments, or by the Student’s t test, where appropriate.

Acknowledgments

This research was supported by NIH Research Project grant RO1 AI065470 (X.L.) and National Multiple Sclerosis Society grant RG 4065-A-1 (X.L.). The work was also partly supported by an International Collaboration grant (30928025) from China and by a Senior Investigator Award from the American Asthma Foundation (X.L.).

Supplementary Materials

www.sciencesignaling.org/cgi/content/full/2/92/ra63/DC1

Fig. S1. Mutation of the TRAF-binding sites of Act1 impairs IL-17–mediated activation of NF-κB.

References and Notes

View Abstract

Navigate This Article