|Home | About | Journals | Submit | Contact Us | Français|
The inability to coordinate the signaling pathways that lead to proper cytokine responses characterizes the pathogenesis of inflammatory diseases such as Crohn's Disease. The Crohn's Disease susceptibility protein, NOD2, helps coordinate cytokine responses upon intracellular exposure to bacteria, and this signal coordination by NOD2 is accomplished, in part, through K63-linked polyubiquitin chains that create binding surfaces for the scaffolding of signaling complexes.
In this work, we show that the NOD2 signaling partner, RIP2, is directly K63 polyubiquitinated by ITCH, an E3 ubiquitin ligase which when lost genetically, causes widespread inflammatory disease at mucosal surfaces. We show that ITCH is responsible for RIP2 polyubiquitination in response to infection with listeria monocytogenes. We further show that NOD2 can bind polyubiquitinated RIP2, and while ITCH E3 ligase activity is required for optimal NOD2:RIP2-induced p38 and JNK activation, ITCH inhibits NOD2:RIP2-induced NFκB activation. This effect can be seen independently at the whole genome level by microarray analysis of MDP-treated Itch−/− primary macrophages.
These findings suggest that ITCH helps regulate NOD2-dependent signal transduction pathways and as such, may be involved in the pathogenesis of NOD2-mediated inflammatory disease.
The Crohn's Disease susceptibility protein, NOD2, is activated by intracellular exposure to both gram-positive and gram-negative bacteria (1-3). Upon activation, it synergizes with Toll-like Receptors (TLRs)1 to help generate a cytokine response that is carefully measured to precisely deliver the correct cytokines, the correct amount of those cytokines and the correct duration of release of those cytokines (1-3). Loss of function, Crohn's Disease-associated NOD2 alleles lose this ability and cause genetic Crohn's Disease (1-3), while gain of function NOD2 alleles cause the granulomatous disease Blau Syndrome and a subset of sporadic Early Onset Sarcoidosis (EOS) cases (4, 5). The perturbation of NOD2 signaling in diverse inflammatory diseases suggests that the coordination of inflammatory signaling pathways by NOD2 is critical and highly regulated.
Upon activation, NOD2 forms a complex with RIP2 (6-8). RIP2 is an obligate component for NOD2 signaling as RIP2-knockout mice cannot generate NFκB, p38 or JNK activation in response to the NOD2 agonist, muramyl dipeptide (MDP, a breakdown product of bacterial peptidoglycan) (9-11). Upon activation, the NOD2:RIP2 complex causes Lysine-63 (K63) polyubiquitination of Lysine-285 on NEMO (7, 8), a component of the NFκB signaling pathway that is necessary for NFκB activation (12). Although evolving, current NOD2:RIP2 signaling models imply that this polyubiquitination of NEMO is responsible for binding to the TAK1 complex, allowing TAK1 to phosphorylate and activate IKK, thereby generating an NFκB response (8, 13-16). Recently, work from independent labs has shown that RIP2 is also K63 polyubiquitinated in response to the in vitro agonist, MDP and in response to intracellular bacterial infection (14, 15). This polyubiquitination occurs on Lysine-209 (K209) of RIP2 and is essential for NOD2:RIP2-induced NFκB activation (15). Although one study presented data suggesting that TRAF6 might be mediating RIP2 polyubiquitnation (14), RIP2 polyubiquitination in TRAF6-deficient MEFs was similar to that seen in WT MEFs (15), and although NOD2 activation can cause autoubiquitination and activation of TRAF6 (8), TRAF6 inhibition by either dominant negative constructs or by siRNA has no effect on NFκB activation by the NODs or RIP2 (8, 10). Given these disparate findings, the E3 ligase(s) mediating RIP2 polyubiquitination is/are unknown, and the role that these potential E3 ligases play in affecting not just NFκB activation, but also p38 and JNK activation downstream of NOD2 activation is unknown.
Ubiquitination plays a key role not only in the NOD2:RIP2 signaling pathway, but also in other innate immune and inflammatory pathways. The TLRs, the IL-1 receptor and the TNF receptor all required lysine-63 (K63) polyubiquitination of signaling proteins for optimal cytokine release (16). While the more common Lysine-48 (K48) polyubiquitin linkage targets a protein to the proteosome for degradation, K63 polyubiquitin linkages are thought to serve a scaffolding function for signaling proteins such that signaling pathways can be activated and cytokine release achieved (16).
In this work, we find that ITCH (AIP4), a HECT-domain containing E3 ubiquitin ligase (17-21), directly ubiquitinates RIP2 to allow differential NOD2:RIP2 signaling through p38, JNK and NFκB. Itch was first identified as the gene underlying the itchy mouse phenotype (17, 18). This mouse has a variety of inflammatory and autoimmune phenotypes (17-20). Mucosal surfaces of these mice show varying degrees of inflammation and, on a C57BL/6J background, the itchy mice die of pulmonary pneumonitis (17-20). When itchy mice are crossed into a Rag1-deficient background, autoimmune-mediated death no longer occurs (21). However, mild mucosal inflammation was present in the itchy animals lacking Rag1, indicating a putative role for this E3 ligase that is independent of the adaptive immune system and is instead important in the innate immune response (21). By linking ITCH to the NOD2:RIP2 signaling pathway, this manuscript provides a potential mechanism for ITCH's role in the innate immune system.
To identify E3 ligases that can polyubiquitinate RIP2, a panel of E3 ligases known to be involved in innate immunity and inflammation were screened for their ability to induce RIP2 polyubiquitination. TRAF2 (an E3 ligase activated by TNF (22)), TRAF3 (an E3 ligase activated by innate immune signals (23, 24)), TRAF4 (an E3 ligase that downregulates Toll-like receptor (TLR) signaling (25)), TRAF6 (an E3 ligase known to be activated by TLR and NOD signals (14)) and ITCH (a ubiquitously expressed E3 ligase, which mediates inflammatory signaling and T cell activation (18-21)) were transfected into 293 cells with HA-tagged ubiquitin and NTAP (Streptavidin binding peptide)-tagged RIP2. All of the E3 ligases transfected were capable of autoubiquitination (Supplemental Figure 1B) suggesting that they were active in this system. After transfection, lysates were generated, and RIP2 was purified under stringent washing conditions (1 M NaCl, 1% SDS) through the use of Streptavidin-agarose beads. Western blotting showed that ITCH expression caused strong RIP2 polyubiquitination and that none of the TRAFs tested significantly increased RIP2 polyubiquitination (Figure 1A – ubiquitin blot in Supplemental Figure 1A, note: all ubiquitin expression blots are shown in the Supplemental data section of the manuscript (Supplemental Figures 1, 3, 4, 6)). To show that the RIP2 polyubiquitination was due to ITCH's E3 ligase activity, the catalytic Cysteine in the HECT domain of ITCH was mutated to an Alanine (C830A ITCH). When this catalytically inactive ITCH was expressed with RIP2, RIP2 polyubiquitination was lost (Figure 1B) and another HECT domain containing E3 ligase (E6AP) did not cause such increased RIP2 ubiquitination (Supplemental Figure 2), suggesting that the catalytic activity of ITCH was necessary for RIP2 polyubiquitination and that this process was specific to ITCH. To then determine if ITCH was the E3 ubiquitin ligase causing K209 ubiquitination, K209 on RIP2 was conservatively mutated to an arginine (K209R). Despite this mutation, ITCH-mediated polyubiquitination of RIP2 was unchanged when compared to wild-type RIP2 (Figure 1C). Further experiments showed only a small role for TRAF6 in ITCH-induced RIP2 ubiquitination. ITCH did not appreciably increase TRAF6 ubiquitination (Figure 2A). NOD2 co-expression with TRAF6 did not increase RIP2 polyubiquitination (Figure 2B). Inhibition of TRAF6 expression had a minimal effect on ITCH-induced RIP2 ubiquitination (Figure 2C) while inhibition of the E2 mediating TRAF6 activity (Ubc13) did not affect ITCH-induced RIP2 ubiquitination (Figure 2D). The fact that ITCH causes ubiquitination of a site distinct from K209 and that the TRAF6/Ubc13 axis plays a minimal role suggests that the ITCH-mediated of RIP2 ubiquitination site(s) may be causing distinct signaling effects within the cell.
ITCH-induced RIP2 polyubiquitation could be direct or through an intermediary E3 ligase. Initial mapping experiments showed that the CARD domain of RIP2 was required ITCH-induced ubiquitination (Supplemental Figure 3A). We have found that truncation of the CARD domain is essential for recombinant RIP2 expression in recombinant systems, so we utilized 293 cell-expressed, purified full length NTAP-RIP2 for in vitro ubiquitination assays. For these assays, ITCH and the catalytically inactive (C830A ITCH) ITCH mutant were expressed as GST-fusion proteins in bacteria (Supplemental Figure 3B). To determine the optimal E2 for the reaction, ITCH was incubated with E1, ATP, NTAP-RIP2 and a panel of E2s in standard in vitro ubiquitination assays. In the presence of UbcH5, UbcH6, and UbcH7, ITCH could directly ubiquitinate RIP2 (Supplemental Figure 3C). After obtaining this information, an in vitro ubiquitination assays was performed to show that there was not a co-purifying bacterial E3 ligase in the recombinant ITCH preparation. No RIP2 polyubiquitination was identified in the presence of C830A ITCH (Figure 3A), indicating that ITCH could directly ubiqutinate RIP2.
The lysine linkage on the ubiquitin itself influences the fate of the polyubiquitinated protein. ITCH has previously been shown to help synthesize K29-linked (26), K48-linked (19) and K63-linked (27) polyubiquitin chains. RIP2's polyubuiquitin chain linkage was determined in three ways. First, in vitro ubiquitination assays were performed using recombinant, purified ubiquitin molecules with specific lysine mutations. In these mutant ubiquitins, one lysine was mutated to an arginine while the other lysines were intact. For instance, K6R ubiquitin contains a lysine mutated to an arginine at position 6 while all the other lysines on that mutant are intact. After in vitro ubiquitination reactions were performed, RIP2 was purified in high stringency wash conditions, and western blotting was performed. While ubiquitin mutants containing lysines mutated to arginines at positions 6, 11, 29 and 48 could all form polyubiquitin chains on RIP2 which were indistinguishable from wt ubiquitin, K63R ubiquitin could not form these chains (Figure 3B), indicating that K63-linked polyubiquitin chains are formed on RIP2 by ITCH in vitro. As a second independent test to determine the ubiquitin linkage, a ubiquitin construct which contains lysine only at position 48 (K48--only) or a ubiquitin construct which contains lysine only at position 63 (K63-only) were transfected into 293 cells with NTAP-RIP2 and ITCH. After purifying RIP2 from the lysate, Western blotting showed that only the K63-only ubiquitin allowed ITCH-mediated RIP2 polyubiquitination (Figure 3C), indicating again that ITCH induced K63-linked polyubiquitin chain formation on RIP2. As a last independent test of the ubiquitin linkage, we utilized a deubiquitinase, A20, that recognizes specific proteins with K63-polyubiquitinated chains and deubiquitinates those proteins (28, 29). As A20 downregulates the NOD2:RIP2 signaling pathway (29, 30) and as A20 has been shown to deubiquitinate K63-linkages on RIP2 (29), we reasoned that if RIP2 was K63 polyubiquinated by ITCH, then A20 should recognize the K63-linked chains and remove them. To this end, A20 was expressed with RIP2 and ITCH. A20 expression caused a significant decrease in the amount of ITCH-induced polyubiquitinated RIP2 (Figure 3D), further suggesting that ITCH caused K63-linked polyubiquitination of RIP2. Thus, three separate experimental methods suggest that ITCH polyubiquitinates RIP2 through K63-specific linkages.
Because ITCH-induced K63-linked polyubiquitination of RIP2 may affect NOD2:RIP2-induced signaling pathways, we wanted to determine if RIP2 polyubiquitination affected NOD2 binding and if NOD2 could bind to polyubiquitinated RIP2. For this reason, Omni-tagged NOD2 was expressed with RIP2 with either ITCH or catalytically inactive C830A ITCH. NOD2 was immunoprecipitated. Western blotting showed that NOD2 could bind to RIP2 in the presence of ITCH or C830A ITCH; however, when ITCH was present, there was not only a significant shift in the mobility of total RIP2 (Figure 4A, second panel from bottom), but also a significant shift in the mobility of the RIP2 that was bound to NOD2 (Figure 4A, top panel). This shift in the RIP2 bound to NOD2 was only present in cells expressing ITCH and not in cells expressing catalytically inactive C830A ITCH (Figure 4A, top panel). To further show that NOD2 can bind to polyubiquitinated RIP2, RIP2 was again expressed with HA-tagged ubiquitin, Omni-tagged NOD2 and a limited amount of ITCH. NOD2 was initially purified via the 6XHis component of the Omni-tag via Nickel beads and then eluted in 200 mM Imidazole. This NOD2 isolate was then split into 2 fractions. The first fraction was subjected to Streptavidin-agarose to isolate the fraction of RIP2 bound to NOD2. The second fraction was subjected to anti-HA immunoprecipitation to isolate the ubiquitinated proteins present in the NOD2 isolate. Within the NOD2 isolate, ubiquitinated RIP2 was present as shown by the ability of the HA antibody to immunoprecipitate a significant amount of RIP2 (Figure 4B, second panel from top, lane 4) and by the ability of the streptavidin-agarose to precipitate a significant amount of ubiquitinated RIP2 (Figure 4B, upper panel, lane 3). These findings suggest that NOD2 can bind to polyubiquitinated RIP2.
We then sought to determine the effect of ITCH on NOD2:RIP2 signaling pathways. NOD2:RIP2 complex activation causes not only NFκB activation, but also activation of a number of signaling pathways including JNK and p38 (11, 14, 31). To determine ITCH's role in NOD2:RIP2 signaling, 4 individual siRNAs were utilized to inhibit ITCH's expression. Inhibition of ITCH expression by these siRNAs has limited effect on activation of JNK, IKKβ or p38 by overexpression of downstream MAP3Ks (Supplemental Figure 5A-D). Each siRNA targeting ITCH (and a control siRNA) was transfected with RIP2 and either FLAG-JNK (Figure 5A), HA-IKKβ (Figure 5B) or FLAG-p38 (Figure 5C). Immunoprecipitations were then performed to isolate JNK, IKKβ or p38. Western blotting showed that in these cells in which the individual siRNAs were transfected, ITCH expression was significantly reduced (bottom panel, Figures 5A, 5B and 5C). Loss of ITCH expression caused significant decreases in both JNK (Figure 5A) and p38 (Figure 5C) activity. In contrast, loss of ITCH expression enhanced RIP2-induced IKKβ activation (Figure 5B). To further independently validate this finding, catalytically inactive C830A ITCH was used. In these conditions, C830A ITCH can act in a dominant negative manner (19, Supplemental Figure 6A) and has limited effect on activation of JNK, IKKβ or p38 by overexpression of downstream MAP3Ks (Supplemental Figure 6B-D). Dominant negative C830A ITCH did, however, cause loss of RIP2-induced activation of both JNK and p38 (Supplemental Figures 6B, 6C). Consistent with published results on the role of ITCH in negatively regulating TNF-induced NFκB signaling (21), ITCH expression caused a strong decrease in RIP2-induced IKKb activation while expression of C830A ITCH not only reversed this inhibition, but also enhanced RIP2-induced IKKβ activation (Supplemental Figure 6B).
Because these experiments utilized overexpression and in vitro ubiquitination assays, we attempted to determine if loss of ITCH expression could affect RIP2 polyubiquitination in more endogenous systems. RIP2 is an essential component of cellular NFκB and MAP kinase signaling pathways initiated by cellular infection by the gram-positive intracellular pathogen, Listeria monocytogenes (11, 31-33). Listeria monocytogenes enters the body by infecting the gastrointestinal epithelium. For this reason, we utilized the gastrointestinal epithelial cell line, HT-29, which expresses NOD2 (34) to determine if loss of ITCH expression affected RIP2 polyubiquitination in response to intracellular Listeria infection. After transfection with the ITCH targeting siRNAs, HT-29 cells were infected with Listeria monocytogenes at an MOI of 3:1. Endogenous RIP2 was then immunoprecipitated. Under these conditions, ITCH expression was significantly inhibited and RIP2 polyubiquitination was attenuated in cells in which ITCH expression was inhibited (Figure 6A). To determine the signaling effects of listerial infection of HT-29s in which ITCH expression was inhibited, HT-29 cells were infected with listeria under the conditions described above. 45 minutes after washing extensively with PBS and then adding Gentamycin to serum-free media to kill any remaining extracellular bacteria, lysates were generated, and western blotting was performed. Activity of JNK and p38 was attenuated in cells in which ITCH expression was inhibited. In contrast, NFκB activity (as shown by total IκBα and phospho-IκBα) was increased relative to the control siRNA-transfected cells (Figure 6B). Because listeria monocytogenes contains numerous PAMPs that will activate additional innate immune signaling pathways, a partial response is not surprising. In all though, the inhibition of RIP2 ubiquitination (Figure 6A), coupled with the signaling responses seen in Figure 6B, suggest that ITCH affects NOD2:RIP2 signaling.
To study the MDP-induced signaling alterations due to loss of ITCH expression in a genetic system, age (8 week old) and sex-matched C57BL/6J Wild-Type (WT) and itchy mice were sacrificed, and their bone marrow was harvested to obtain bone-marrow-derived macrophages. After differentiation, these macrophages were treated with MDP for 60 minutes. RIP2 was then immunoprecipitated under stringent washing conditions. Western blotting showed that RIP2 was ubiquitinated in these cells upon MDP treatment, but that this ubiquitination was very strongly attenuated in the itchy macrophages (Figure 7A). Signaling experiments were then performed. Macrophages from itchy mice had significantly lower levels of MDP-induced JNK and p38 activity and significantly higher levels of NFκB activity (as shown by increased degradation of IκBα and increased levels of phospho-IκBα (Figure 7B)). The itchy cells did have increased basal phosphorylation of IκBκ, indicating the possibility of increased basal NFκB activation. In all, these signaling experiments closely mirror the results found in both the 293 overexpression system and in the HT-29 infection model.
To next independently determine if ITCH was in the NOD2:RIP2 pathway, WT and Itch−/− macrophages were either left untreated or were treated with 10 μg/mL for 16 hours. Total RNA was extracted and microarray analysis was performed. After subtracting out baseline gene expression differences between the WT and Itch−/− macrophages, whole genome clustering analysis showed that WT macrophages could be distinguished from Itch−/− macrophages upon MDP treatment (Figure 7C). This experiment was performed in duplicate, and both results are shown. The data was then analyzed for the inducible expression of known NFκB target genes. Itch−/− macrophages had significantly higher expression of NFκB target genes including BCL2, CCL3, IL-1, TNF, ICAM and TNFAIP2 (Figure 7D). These microarray analyses show that Itch−/− cells have a different genetic response to MDP stimulation and that a number of NFκB target genes are upregulated. These results, when coupled with the biochemistry/molecular biology of Figures Figures11--5,5, the siRNA results of Figure 6 and the signaling differences between WT and Itch−/− macrophages in Figure 7A and B strongly suggest that ITCH-mediated RIP2 ubiquitination has functional consequences.
The work presented in this manuscript suggests that ITCH directly ubiquitinates RIP2 to specify NOD2 signaling responses. RIP2 has previously been shown by three independent groups to be K63-polyubiquitinated (14, 15, 35). One group mapped the site of polyubiquitination on RIP2 to lysine-209 (K209) (15). This K209 ubiquitination site is required for NOD2:RIP2-induced NFkB activation (15), however, while this group found the ubiquitination to be independent of TRAF6 (14), another group presented data suggesting that TRAF6 was the E3 ligase (134. For this reason, it is unclear if K209 ubiquitination is mediated through the TRAF proteins and the E2 ligase Ubc13 (14, 15). Our results, like those of the Nunezand Inohara labs, suggest TRAF6-independent ubiquitination machinery for NOD2:RIP2 signaling. Unlike the TRAF proteins, ITCH induced RIP2 ubiquitination does not require Ubc13 (Figure 2D). Inhibition of TRAF6 expression showed only a small effect on ITCH-induced RIP2 ubiquitination (Figure 2C), and prior work from our lab (8) and Gabriel Nunez's lab (36) showed that while TRAF6 could be activated by NOD2, neither dominant negative TRAF6 (36) nor siRNA inhibition of TRAF6 (8) had any bearing on NOD:RIP2-induced NFκB activation. Thus, these results, coupled with the findings in this manuscript, suggest that ITCH-induced RIP2 ubiquitination is not dependent on TRAF6. Additionally, a report published while this manuscript was in review also suggests that RIP2 ubiquitination is independent of TRAF6 (35). The Saleh group found that the cIAP proteins could both K48 and K63 ubiquitinate RIP2 and that depending on the cell type, this ubiquitination was required for p38, JNK and/or NFκB signaling (35). While our own unpublished findings indicate that cIAP-1 causes predominantly K48-linked polyubiquitination of RIP2, given the presence of at least 1000 E3 ubiquitin ligases in the genome, it is not surprising that multiple E3 ubiquitin ligases can act on a protein to exert signaling effects. Further work will be required to determine the interplay between the cIAPs and ITCH in NOD signaling pathways.
ITCH, itself, plays a role in the immune system. While it is ubiquitously expressed, mice lacking Itch develop uncontrolled inflammatory responses at mucosal surfaces (17, 18, 21). The lethal aspect of the phenotype is dependent on the adaptive immune response as itchy mice crossed onto a Rag1-deficient background show no death due to pulmonary pneumonitis. However, the inflammatory phenotype partially persists in these mice (21), suggesting that ITCH also has a role in regulating the innate immune response. The work described here work suggests that ITCH may do this by modulating signaling downstream of NOD2:RIP2 activation. Synthesizing these data with previously published reports (6-8, 13-15, 30, 36), we hypothesize that upon cellular infection/MDP exposure, RIP2 dissociates from the protein kinase, MEKK4 to bind to NOD2. Upon binding NOD2, a subset of RIP2 becomes ubiquitinated on K209 to stimulate NEMO ubiquitination and subsequent NFκB activation. Separately, a second subset of RIP2 is polyubiquitinated by ITCH to activate JNK and p38 signaling. We hypothesize that ITCH competes with the K209 E3 ligase to dictate whether JNK and p38 or whether NFκB is the predominant downstream pathway following NOD2:RIP2 activation. It will be important to determine the ITCH-induced polyubiquitination site on RIP2 and to determine if differential polyubiquitination between K209 and the ITCH-mediated site dictate signal specificity downstream of NOD2. It will also be important to determine if the phenotype of the itchy mice is influenced by the NOD2 signaling pathway and if loss of components of the NOD2 signaling pathway can complement the inflammatory phenotype seen in the itchy mice. Lastly, a key question remains. Where in the cell is NOD2:RIP2 signaling taking place? NOD2 can localize to the membrane upon exposure to MDP (37). MDP uptake by macrophages appears to be due to an endocytic process involving pannexin-1, and MDP is ultimately delivered to and possibly escapes from the phagolysosome (38). Because ITCH has been published to localize to endocytic compartments (19, 26), we suspect that, upon MDP uptake, the NOD2:RIP2 physiologic interaction with ITCH localizes to endocytic compartments and that ITCH helps regulate the initial signaling events after MDP uptake. In this light, an alternative explanation for the data presented in the manuscript is that ITCH ubiquitination causes preferential degradation of RIP2 to decrease NFκB activation. Despite these questions and need for future studies, this work suggests an important regulatory role for ITCH in NOD2:RIP2 signaling pathways and suggests that ITCH may have a role in the pathophysiology of NOD2-driven inflammatory diseases.
HEK-293 and HT-29 (ATCC) cells were grown in DMEM containing 5% FBS. Transfections were performed by calcium phosphate precipitation as previously described (7). Immunoprecipitations were performed in 50 mM Tris (pH 7.5), 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, 1 mM EGTA, 1 mM β-glycerophosphate, 1 mM PMSF, 1 mM NaVO4, 10 nM Calyculin A in the presence of protease inhibitor cocktail (Sigma). After adding Protein G sepharose beads, IPs were washed at least 5 times before Western blotting. For IP-ubiquitination assays, cells were lysed in high-salt modified Cell Signaling Lysis buffer containing 1 M NaCl and from 0.25%-1% SDS as indicated in the text. Western blotting was performed as previously described (7).
Myc (9E10), RIP2, actin and Omni antibodies were obtained from Santa Cruz Technology. P4D1 (ubiquitin) Myc (rabbit), phospho IKKα/β, phospho-IκBα, IKKβ, IκBα, NEMO, phospho-JNK, JNK, phospho-p38 and total p38 antibodies were obtained from Cell Signaling Technology. NTAP-RIP2 was generated by subcloning the Omni-RIP2 construct (EcorI-XhoI) into the NTAP vector (Stratagene). Omni-RIP2, HA-ubiquitin, Myc-NEMO K399R, HA-RIP2, Omni-NOD2, FLAG-TAK1, FLAG-p38, HA-MEKK4, FLAG-JNK, ITCH, C830A ITCH and HA-IKKβ were used as previously described (7, 8, 30, 36). The K209R RIP2 construct was generated by Quickchange site-directed mutagenesis (Stratagene).
ITCH and C830A ITCH were subcloned into pGEX-4T vectors and purified using standard methodology. For the in vitro ubiquitination assays, E1, the indicated E2s, wt ubiquitin and the indicated ubiquitin mutants were all obtained from Boston Biochem. Reaction mixtures consisted of a buffer containing 25 mM HEPES (pH 7.4), 100 mM NaCl, 1 mM DTT, 10 μM MG132, 1 μM ubiquitin aldehyde, 4 mM ATP and 50 ng/mL E1, 1 μg/mL ubiquitin, 200 ng/mL E2, 150 ng/mL ITCH and NTAP-RIP2 purified from transfected HEK 293 cells. Reactions were allowed to proceed at 37°C for 1 hour. The reactions were then diluted in 900 μL of Cell Signaling lysis buffer containing 1 M NaCl and 0.5% SDS, and RIP2 was then immunopurified from the in vitro reaction.
Four separate siRNAs were purchased from Qiagen. The sequences of these were as follows: siRNA1: CACGGGCGAGUUUACUAUGUA, siRNA2: CAAGAGCUAUGAGCAACUGAA, siRNA3: AUGGGUAGCCUCACCAUGAAA siRNA4: UGCCGCCGACAAAUACAAAUA. HT-29 cells were transfected using Dharmafect 4 (Fisher) according to the manufacturer's protocol.
Listeria monocytogenes were grown overnight in BHI media at 37°C. The next morning, the listeria were diluted 1:10 and allowed to grow for another 45 minutes such that they were in exponential growth phase. OD600's were performed to quantitate the numbers of listeria and the exponential growth rate of that listerial culture. Listeria was added to HT-29 cells at a MOI of 3:1. 45 minutes after the addition of listeria to the HT-29 cells, fresh media containing Gentamycin Sulfate (50 μg/mL final concentration) was added to kill extracellular bacteria. The infection was then allowed to proceed for another hour before lysates were generated, protein concentrations were standardized and RIP2 was immunoprecipitated. High stringency washing and Western blotting and signaling assays on lysates were performed as described above.
Total RNA was isolated from WT and Itch−/− macrophages via a Qiagen kit. Integrity was analyzed by agarose gel electrophoresis, and RNA quantities were standardized. Microarray analysis on the RNA was performed by the Cleveland Clinic Research Institute Core Facility using Illumina Mouse-6 bead chips. Raw Illumina microarray data were quintile normalized and significant genes (P < 0.05, adjusted for false discovery rate) were identified using the lumi package in R (39). Hierarchical clustering and heatmap analysis of significant genes were performed using Cluster 3.0 and Java TreeView (40).
This work was supported by R01GM86550-01 (DWA), R21AI076886-01 (DWA), the Burroughs Wellcome CABS 10061206.01 (DWA) and R01HD056369 (PCS). We thank XiaoXia Li (CCF, Cleveland, OH) for supplying the FLAG-tagged TRAF2, 3 and 4 constructs.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 Abbreviations used: TLR: Toll-like receptor, EOS: Early onset sarcoidosis, NFκB: nuclear factor kappa B, , MDP: muramyl dipeptide, TLR: Toll-like receptor, IP: immunoprecipitation, BMDM: bone marrow-derived macrophages, IKK: I Kappa Kinase, GI: gastrointestinal