Role for Erbin in Bacterial Activation of Nod2

Identification of proteins that interact with NOD2 by Y2H screening.Yeast two-hybrid (Y2H) screening and data analysis were performed by Hybrigenics, S.A., Paris, France.

Bait cloning. NOD2 was PCR amplified and cloned into a Y2H vector optimized by Hybrigenics. The bait construct was checked by sequencing the entire insert and was subsequently transformed into the Saccharomyces cerevisiae strain L40 GAL4 (14).

Y2H screening.A human activated leukocyte and mastocyte random-primed cDNA library, transformed into the Y187 yeast strain and containing 10 million independent fragments, was used for mating. A high mating efficiency was obtained by using a specific mating method (29a, 29b, 29c). The screen was first performed on a small scale to adapt the selective pressure to the intrinsic property of the bait. Neither toxicity nor autoactivation of the bait was observed. The full-scale screen was then performed under conditions ensuring a minimum of 50 million interactions tested in order to cover five times the primary complexity of the yeast-transformed cDNA library (37). Seventy-one million interactions were actually tested with Nod2. After selection on medium lacking leucine, tryptophan, and histidine, 39 positive clones were picked and analyzed, and the corresponding prey fragments were amplified by PCR and sequenced at their 5′ and 3′ junctions. Sequences were then filtered and divided into contigs as described previously (12) and compared to the GenBank database using BLASTN (3). A predicted biological score (PBS) was used to assess the reliability of each interaction, as described previously (12). Briefly, the PBS relies on two different levels of analysis, as follows. Firstly, a local score takes into account the redundancy and independency of prey fragments as well as the distributions of reading frames and stop codons in overlapping fragments. Secondly, a global score takes into account the interactions found in all the screens performed at Hybrigenics using the same library. The PBSs have been shown to positively correlate with the biological significance of interactions (37, 41).

Construction of Nod2- and Erbin-encoding plasmids.Plasmids encoding amino-terminally FLAG-tagged Nod2, deletion constructs, and the indicated domains as well as the expression plasmids for the Nod2 mutant proteins were cloned by PCR. Nod2 mRNA derived from a Nod2 cDNA expression plasmid (kindly provided by G. Nunez) was used as the template, and the PCR products were cloned using BamHI and XhoI restriction sites into the pCMV-Tag2B vector (Stratagene). The FLAG-Nod1 expression plasmid was constructed following the same strategy, using a Nod1 cDNA expression vector as the template. Myc-tagged expression plasmids for Erbin and deletion mutants thereof were a kind gift from L. Mei and P. Dai (20, 21). All constructs used were subjected to full-length sequencing (MilleGen, France).

Generation of monoclonal antibodies against Nod2.An internal peptide of Nod2 (₁₂₉HPARDLQSHRPAIVRRL₁₄₅) was synthesized and coupled to keyhole limpet hemocyanin or ovalbumin (PSL, Heidelberg, Germany). Rats were immunized with 50 μg peptide-keyhole limpet hemocyanin using CPG 2006 and incomplete Freund adjuvant as an adjuvant. Supernatants were tested in a differential enzyme-linked immunosorbent assay and analyzed by Western blotting using extracts from HEK 293T cells and transiently transfected HEK 293T cells expressing FLAG-Nod2. Anti-Nod2 7E11 and 4A11 of the rat immunoglobulin G2a (IgG2a) subclass were used in this study.

Cell culture, indirect immunofluorescence microscopy, and bacterial infection.HEK 293T, HT29, and HeLa cells were obtained from the ATCC. Cells were cultivated at 37°C with 5% CO₂ in Dulbecco's modified Eagle's medium (Gibco-BRL) supplemented with 10% heat-inactivated fetal calf serum (Gibco-BRL) and penicillin-streptomycin (100 IU/ml and 100 mg/ml, respectively; Gibco-BRL). For indirect immunofluorescence microscopy, cells were seeded on HCl-treated coverslips, fixed in 3% paraformaldehyde in phosphate-buffered saline, and permeabilized with 0.5% Triton X-100 for 5 min. Cells were incubated in 3% bovine serum albumin in phosphate-buffered saline. Staining was done by subsequent incubation of primary and secondary antibodies in 3% bovine serum albumin. Primary antibodies were rat anti-Nod2 7E11 (1:100; this study), mouse anti-FLAG M2 (1:100) (F3165; Sigma-Aldrich), rabbit anti-Myc (1:500) (A-14; Santa Cruz Biotechnology), rabbit anti-Erbin serum (1:500) (20), and rabbit anti-S. flexneri 5a lipopeptide (1:500; kindly provided by J. Mounier and A. Phalipon, Institut Pasteur). Primary antibodies were detected with Texas Red-conjugated goat anti-mouse IgG (1:500; Sigma-Aldrich), Alexa 488-conjugated goat anti-rabbit IgG (1:500; Invitrogen Molecular Probes), and fluorescein isothiocyanate-conjugated goat anti-rat IgG (1:500; Jackson Research) secondary antibodies. DNA was stained with Hoechst 33342 (0.1 μg/ml; Invitrogen Molecular Probes).

Bacterial infection of HeLa and HEK 293T cells was performed using the S. flexneri strain M90T afaE as described previously (36). M90T afaE is a wild-type (WT) invasive strain of S. flexneri serotype 5a harboring the plasmid pIL22, which encodes the afimbrial adhesin from uropathogenic Escherichia coli (9). Briefly, bacteria were added to the cells, which were transferred to serum- and antibiotic-free medium and incubated for 10 min at room temperature prior to transfer to 37°C (time zero).

Immunoprecipitation and immunoblotting.For immunoprecipitation, HEK 293T cells were transiently transfected, using FuGene6 (Roche) according to the manufacturer's conditions, with the indicated plasmids (1 μg plasmid per 6-cm dish) and incubated for 48 h. Cells were lysed in NP-40 buffer (150 mM NaCl, 1% NP-40, 50 mM Tris-HCl, pH 7.5) containing phosphatase inhibitors (20 mM β-glycerophosphate, 5 mM NaF, 100 μM Na₃VO₄, and Complete protease inhibitor cocktail [Roche]). Lysates were cleared for 20 min at 14,000 × g at 4°C. Immunoprecipitation was subsequently carried out for 4 h at 4°C by adding anti-FLAG beads (M2 gel; Sigma-Aldrich) or 9E10 anti-Myc antibody (sc-40; Santa Cruz Biotechnology) and protein G-Sepharose matrix (Amersham Pharmacia) to the cell extracts. The beads were precipitated by centrifugation steps and washed five times in NP-40 buffer before sodium dodecyl sulfate loading buffer was added. Typically, about 10 to 20 times more precipitate than input was loaded into the gel. Proteins were separated by Laemmli sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred by semidry Western transfer to a nitrocellulose membrane (Bio-Rad). Proteins were detected by incubation of the membrane subsequently with primary and secondary antibodies and by a final incubation with SuperSignal West Femto maximum sensitivity substrate (Pierce). Primary antibodies were mouse anti-FLAG M2 (1:1000) (F3165; Sigma-Aldrich), rabbit anti-Myc (1:1,000) (A-14; Santa Cruz Biotechnology), rabbit anti-Erbin serum (1:1,000) (20), rat anti-Nod2 4A11 and 7E11 (1:100; this study), mouse anti-β-tubulin (1:4,000) (T-7816; Sigma-Aldrich), and rabbit anti-IκBα (1:1,000) (C-21; Santa Cruz Biotechnology). Secondary antibodies were horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (1:4,000) (170-6616; Bio-Rad), HRP-conjugated goat anti-rabbit IgG (1:4,000) (170-6515; Bio-Rad), and HRP-conjugated goat anti-rat IgG (1:4,000; Jackson Research Laboratory).

Luciferase reporter assays.Activation of NF-κB was measured using a luciferase reporter assay as described previously (36). Briefly, HEK 293T cells were seeded in 24-well plates and transfected using Fugene6 (Roche), 0.25 ng Nod2 or Nod2 K305R, 75 ng Igk-luciferase NF-κB reporter (31), and 25 ng β-galactosidase expression plasmid per well. Cells were directly stimulated with 200 nM MDP or 0.01 μg/ml tumor necrosis factor (TNF), as indicated. After 16 h of incubation, the cells were lysed and luciferase activity was measured. The mean and standard error of the mean (SEM) were calculated from triplets, and luciferase activity was normalized as a ratio to β-galactosidase activity. For short hairpin RNA (shRNA) experiments, cells were transfected with the indicated amount of an Erbin-specific shRNA plasmid (pBLOCK-it; Invitrogen), a kind gift of P. Dai and L. Mei (10), by using FuGene 72 h prior to transfection with the reporter and stimulation. Western analysis of the Erbin protein amounts after shRNA treatment was conducted with the lysates used for the luciferase activity test. For infection, S. flexneri M90T afaE was added to the transfected cells, and the medium was replaced with Dulbecco's modified Eagle's medium containing 50 μg/ml gentamicin 15 min after incubation at 37°C.

Identification of Erbin.In order to identify new interaction partners of the human Nod2 protein, a yeast two-hybrid screen was applied. We found that full-length Nod2 was functional in a yeast two-hybrid system by using a directed interaction test with the known binding partner RIP2 (data not shown). Since Nod2 is expressed in leukocytes (13, 18, 34), a yeast two-hybrid screen using full-length Nod2 as bait was therefore performed with a human leukocyte/mastocyte library. Interestingly, one of the interacting clones obtained encoded the CARD of RIP2, which is known to interact with Nod2 (34), thereby validating our approach. The clones with the highest probability scores for interaction among the candidates were two independent clones spanning the LRR domain of the human LAP family member Erbin (6, 7) (Fig. 1). Moreover, the expression profile reported for Erbin was closely related to the expression pattern of Nod2, with both proteins being highly expressed in CD14⁺ monocytes and CD33⁺ myeloid cells in both humans and mice (39).

• Open in new tab
• Download powerpoint

FIG. 1.

Identification of Erbin in a yeast two-hybrid screen. Schematic representations of Erbin and Nod2 are shown. LRR domains are represented by bars, dark rounded boxes represent PDZ domains, and light rounded boxes represent CARDs. The NACHT domain is depicted as a black box. Numbers give amino acid positions. The two partial clones of Erbin identified in the screen are shown (middle).

The physical interaction of Nod2 and Erbin was subsequently confirmed in human cells by coimmunoprecipitation from transiently transfected HEK 293T cells. Endogenous Erbin protein was found to coprecipitate with FLAG-Nod2 (Fig. 2A). Moreover, ectopically expressed Myc-Erbin, but not the matrix alone, coprecipitated with FLAG-Nod2 (Fig. 2B). The interaction between Nod2 and Erbin was shown to be specific, as no binding of Erbin was observed with the Nod2-related protein Nod1 (Fig. 2A), nor did Nod2 bind the LAP protein Scribble, a close relative of Erbin (7; data not shown). Both a P-loop mutant of Nod2 (K305R) and a Crohn's disease-associated frameshift mutant of Nod2 (L1007fs) are known to be compromised in sensing of MDP (15, 22, 33, 34, 40). Whereas Erbin bound to WT Nod2, no significant interaction was observed with Nod2 K305R or Nod2 L1007fs (Fig. 2C).

• Open in new tab
• Download powerpoint

FIG. 2.

Nod2 and Erbin physically interact in human cells. Western blot analysis results of coimmunoprecipitations from HEK 293T cells are shown. Signals obtained in immunoprecipitates (IP) and total lysates (input), using the indicated antibodies, are shown. The matrix alone was used in control precipitations (CTRL). Asterisks indicate the specific proteins. (A) FLAG-Nod2 and FLAG-Nod1 were ectopically expressed in HEK 293T cells, and lysates were precipitated using anti-FLAG antibody. Coprecipitated endogenous Erbin was detected using anti-Erbin-specific serum. (B) FLAG-Nod2 and Myc-Erbin were coexpressed, and lysates were immunoprecipitated using anti-Myc antibody. Coprecipitated FLAG-Nod2 was detected using anti-FLAG antibody. (C) Interaction of mutants of Nod2 with Erbin. FLAG-tagged Nod2 and the K305R and L1007fs mutants thereof were coexpressed with Myc-Erbin, and lysates were subsequently immunoprecipitated using anti-FLAG antibody. Coprecipitated Myc-Erbin was detected using anti-Myc antibody. (D) Mapping of Nod2 interaction domain in Erbin. Myc-tagged Erbin and fragments thereof spanning the indicated amino acid residues were coexpressed together with FLAG-Nod2 in HEK 293T cells. Lysates were immunoprecipitated using anti-FLAG antibody, and precipitated proteins were detected using anti-Myc or anti-FLAG antibody. Data for short and long exposures of the signal are shown. (E) Mapping of Erbin interaction domain in Nod2. FLAG-tagged Nod2 and fragments spanning the indicated amino acid residues were coexpressed together with Myc-Erbin in HEK 293T cells. Lysates were immunoprecipitated using anti-FLAG antibody, and precipitated proteins were detected using anti-Myc or anti-FLAG antibody. Schematic drawings of the deletion constructs used are depicted in the lower panels.

Taken together, these findings show that Erbin forms a complex with Nod2 in human cells and that the affinity for Nod2 seems to depend on the MDP-sensing functionality of the Nod2 protein.

Mapping of interaction domains.To establish the protein domains responsible for the interaction between the two proteins, coimmunoprecipitation experiments with Erbin and Nod2 deletion constructs were conducted with lysates of transiently transfected HEK 293T cells. FLAG-Nod2 not only bound to the full-length Myc-Erbin protein but also showed binding at nearly the same affinity to a deletion protein of Erbin lacking the PDZ domain (amino acids [aa] 1 to 1279) (Fig. 2D, lanes 1 and 3). On the other hand, no interaction was observed with an amino-terminal deletion protein of Erbin containing the PDZ domain alone (aa 965 to 1371) (Fig. 2D, lane 6). However, a mutant of Erbin lacking the LRR domain (aa 391 to 1371) still showed very weak binding, and also the LRR domain (aa 1 to 391) and the central domain of Erbin (aa 391 to 965) itself bound weakly to Nod2 (Fig. 2D, lanes 2, 4 and 5).

Using the same approach with Nod2, the expected binding of full-length Erbin to Nod2 was observed, whereas no binding was observed to Nod1 or the matrix alone (Fig. 2E). Furthermore, a mutant lacking the LRR domain of Nod2 (aa 1 to 600) bound Myc-Erbin, although to a lesser extent (Fig. 2E). Importantly, no interaction was observed with a mutant lacking the two CARDs of Nod2 (aa 250 to 1040) or with the NACHT domain (aa 250 to 600) or CARD (aa 1 to 250) alone (Fig. 2E).

In conclusion, these results support the hypothesis that the domain responsible for mediating the interaction with Nod2 lies within the carboxyl terminus of the LRR domain of Erbin. On the side of Nod2, the interaction is mediated by the CARDs of Nod2; however, these domains seemed not to be sufficient to mediate the interaction.

Localization in human cells.To explore the site of complex formation between Nod2 and Erbin within the cell, the subcellular localization of the two proteins in human cells was examined. HeLa cells were transfected transiently with FLAG-tagged human Nod2 and Myc-Erbin, and the proteins were visualized using the indicated antibodies. Consistent with recent reports describing the localization of Erbin in human cells, the Erbin antibody gave a strong signal at the cell cortex and a weak signal in the cytosol (6, 29, 30) (Fig. 3A). As reported recently, we also observed Nod2 at the cell membrane and in the cytoplasm (4, 30) (Fig. 3A). Depending on the expression level of ectopically expressed Nod2, the protein was found in the cytosol and was more or less prominent at the cell membrane and in membrane ruffles, where the merged signals of Nod2 and Erbin revealed partial colocalization (Fig. 3A). Notably, about half of the transfected cells exhibited nuclear localization of the ectopically expressed Nod2 protein (Fig. 3A and B).

• Open in new tab
• Download powerpoint

FIG. 3.

Subcellular localization of Nod2 and Erbin. (A) Indirect immunofluorescence micrographs of HeLa cells transfected with a FLAG-Nod2 construct. Fixed cells were stained with Erbin-specific serum and anti-FLAG antibody. Signals obtained with the two antibodies (left and middle panels) are presented. Merged images, with staining of DNA, are shown in the right panels. The FLAG-specific signal is shown in green, and the Erbin-specific signal is shown in red. (B) Indirect immunofluorescence micrographs of HeLa cells cotransfected with a FLAG-Nod2 construct and an Erbin shRNA-expressing plasmid. Fixed cells were stained with Erbin-specific serum and anti-FLAG antibody. Signals obtained with the two antibodies (left and middle panels) are presented. A merged image, with staining of DNA, is shown in the right panel. The FLAG-specific signal is shown in green, and the Erbin-specific signal is shown in red. The arrows indicate a cell with highly reduced Erbin protein levels. (C) Characterization of Nod2-specific monoclonal antibodies. (Left panel) Western analysis of total cell lysates from HEK 293T cells transiently transfected with FLAG-Nod2, probed using the indicated antibodies. Running of a protein standard is shown to the right. (Right panel) Western analysis of total cell lysates from HT29 cells showing endogenous Nod2 and from HEK 293T cells transiently transfected with FLAG-Nod2 as a positive control, probed with 4A11 antibody. (D) Indirect immunofluorescence micrographs of HT29 cells. Fixed cells were stained with the Erbin-specific serum and the 7E11 antibody. Signals obtained with the two antibodies (left and middle panels) are presented. A merged image, with staining of DNA, is shown in the right panel. The Nod2-specific signal is shown in green, and the Erbin-specific signal is shown in red. Bars, 10 μm.

Erbin was initially reported to play a role in targeting the ErbB2 receptor to the membrane (6), and Erbin associates with basolateral membranes though an LRR-domain-dependent mechanism (29). Recently, a domain within the LRR of Nod2 was also reported to be essential for the membrane association of Nod2 (4). The deletion of the LRR domain in Nod2 interrupted the interaction with Erbin (Fig. 2C and E), making it tempting to speculate that Erbin might target Nod2 to the membrane. To investigate a possible involvement of Erbin in localizing Nod2 at the membrane, endogenous Erbin was silenced by shRNA. HeLa cells transiently cotransfected with FLAG-Nod2 and the Erbin shRNA plasmid showed a clear reduction in the Erbin-specific immunofluorescence signal in transfected cells (Fig. 3B, arrows). However, the membrane localization pattern of ectopically expressed Nod2 in these cells was indistinguishable from that in normal HeLa cells (Fig. 3B, arrows). Thus, Erbin seems not to be essential for the targeting of Nod2 to the membrane.

To determine the localization of endogenous Nod2, Nod2-specific monoclonal antibodies were generated. Two clones, 4A11 and 7E11, specifically recognized the ectopically expressed FLAG-Nod2 protein expressed in HEK 293T cells but gave no signal in lysates from untransfected HEK 293T cells (Fig. 3C; data not shown). Whereas 4A11 gave the strongest signal in Western blot analysis, 7E11 gave the best signal by immunofluorescence and recognized endogenous Nod2 protein in lysates from HT29 cells, which were shown to express higher levels of endogenous Nod2 (4) (Fig. 3C). Using these antibodies and a polyclonal serum specific for Erbin (20), colocalization of Nod2 and Erbin at the cell membrane in undifferentiated cells from the human colon cancer cell line HT29 was observed (Fig. 3D). Notably, the Nod2 antibody also gave signals in the nuclei of HT29, SW480, HeLa, and HEK cells (Fig. 3D; data not shown).

Taken together, these data provide evidence that a partial pool of Nod2 is localized in an Erbin-independent manner at the cell surface, where it partially colocalizes with Erbin.

Influence of Erbin on Nod2-mediated NF-κB activation.Nod2 triggers the NF-κB signaling pathway via activation of RIP2 (26, 34). This activation can be studied specifically in HEK cells transiently transfected with Nod2 and a NF-κB reporter construct by using the Nod2-specific elicitor MDP (15, 17, 40). Since HEK 293T cells do express detectable amounts of Erbin (Fig. 2A and 4B), Nod2-mediated NF-κB activation was assessed in these cells, and the influence of Erbin on this signaling pathway was tested by shRNA-mediated Erbin knockdown. As shown in Fig. 4B, the Erbin shRNA significantly reduced the amount of total Erbin protein detected by Western analysis. Quantification of NF-κB reporter activity in these cells showed that a knockdown of Erbin correlated with higher reporter activity when cells were treated with MDP; however, no correlation between the Erbin protein level and basal Nod2-driven NF-κB reporter activity was observed (Fig. 4A). In contrast, ectopic expression of Erbin resulted in reduced Nod2/MDP-mediated NF-κB reporter activity (Fig. 4C). This influence on Nod2/MDP-mediated NF-κB activation was specific, as TNF-mediated NF-κB reporter activity was unaltered when increasing amounts of Erbin were expressed (Fig. 4C).

• Open in new tab
• Download powerpoint

FIG. 4.

Influence of Erbin expression on Nod2-mediated NF-κB activation. (A) HEK 293T cells seeded in 24-well plates were transiently transfected with 0, 100, or 250 ng of a shRNA plasmid targeting Erbin mRNA. At 72 h posttransfection, luciferase activity was measured 16 h after transfecting the cells with the reporter and Nod2 plasmids. Cells were either stimulated with 200 nM MDP or left untreated (CTRL). Percentages of luciferase activity in the lysates relative to that in cells without Erbin shRNA treatment are presented. Error bars show SEM. (B) Western blot analysis of the total cell lysate used for panel A (plus MDP). Erbin was detected with anti-Erbin serum (upper lanes). The loading control was β-tubulin (lower lanes). (C) HEK 293T cells seeded in 24-well plates were transiently transfected with the indicated amounts of Myc-Erbin expression plasmid, together with plasmids encoding an NF-κB-driven luciferase reporter and Nod2. Sixteen hours after stimulation with 200 nM MDP (left) or 0.01 μg/ml TNF (right), luciferase activity was measured. Percentages of luciferase activity in the lysates relative to that in cells transfected with empty vector are presented. Error bars show SEM.

In conclusion, these results support an involvement of Erbin as a negative regulator of the MDP-mediated activation of NF-κB by Nod2.

Nod2/Erbin interaction during bacterial infection.Nod2 is proposed to be the major sensor for intracellularly located, bacterially derived MDP (15, 28). Accordingly, we found that NF-κB activation triggered by invasion of HEK 293T cells with the gram-negative bacterium S. flexneri was enhanced when Nod2, but not the Nod2 K305R mutant, was ectopically expressed (Fig. 5A). Since Erbin was shown to modulate NF-κB activation after elicitation of Nod2, the subcellular localization of Erbin and Nod2 was analyzed during biological activation of Nod2, using the S. flexneri infection model. For this purpose, HeLa cells transiently transfected with FLAG-Nod2 were infected with the invasive S. flexneri strain M90T expressing the E. coli AfaE adhesin to mediate a higher infection efficiency. During bacterial invasion, in all infected cells, the Nod2 signal was predominant at the entry foci of the bacteria (Fig. 5B and C). At the same time, the Erbin signal was also enhanced at the site of bacterial entry, where it showed colocalization with Nod2 (Fig. 5C). S. flexneri induces membrane ruffling at the site of entry, leading to a higher surface area at the entry site (32), which underscores the observation that both Nod2 and Erbin are localized at the cell membrane.

• Open in new tab
• Download powerpoint

FIG. 5.

Analysis of Nod2 and Erbin during bacterial infection. (A) HEK 293T cells seeded in 24-well plates were transiently transfected with 0.25 ng Nod2, Nod2 K305R, or empty vector. At 16 h posttransfection, luciferase activity was measured 4 h after incubation of the cells with S. flexneri M90T afaE. Luciferase activities normalized to the β-galactosidase activity in the lysates are presented. Error bars show SEM. (B) Indirect immunofluorescence micrographs of S. flexneri-infected HeLa cells transiently transfected with FLAG-Nod2. Cells were incubated with S. flexneri M90T afaE for 30 min, fixed, and stained with anti-FLAG and anti-lipopolysaccharide antibodies. Signals obtained with anti-FLAG antibody are shown (green in the merged figure). Bacteria are shown in red in the merged figure, together with staining of DNA (blue). (C) Indirect immunofluorescence micrographs of S. flexneri-infected HeLa cells transiently transfected with FLAG-Nod2 and Myc-Erbin. Cells were incubated with S. flexneri M90T afaE for 30 min, fixed, and stained with anti-FLAG and anti-Myc antibodies. Signals obtained with the anti-FLAG and anti-Myc antibodies are shown. A merged image of the Myc-Erbin and FLAG-Nod2 signals, together with staining of DNA, is shown in the right panel. The FLAG-specific signal (Nod2) is shown in green, and the Erbin-specific signal is shown in red. Bar, 10 μm. (D) Western analysis of coimmunoprecipitations conducted with S. flexneri-infected and noninfected HEK 293T cells. Cells were transiently transfected with FLAG-Nod2 or cotransfected with FLAG-Nod2 and Myc-Erbin, respectively, and infected with S. flexneri M90T afaE for the indicated times. Subsequently, cells were lysed, and cell extracts were immunoprecipitated using anti-FLAG antibody. The obtained signals for precipitated FLAG-Nod2 and coprecipitated endogenous (right panels) or ectopically expressed (left panels) Erbin are shown. Infection was monitored by detection of IκBα in the total cell extract (lower lanes).

The coincidence of the two proteins being present at the bacterial entry focus makes it tempting to speculate that this is of biological relevance in sensing bacteria through Nod2. To examine this possibility in greater detail, we asked whether colocalization at the site of bacterial entry could reflect changes in the stoichiometry of the Nod2/Erbin complex. To this end, HEK 293T cells were transiently transfected with FLAG-Nod2 and subsequently infected with S. flexneri M90T afaE. At the indicated times after infection, cells were lysed, and equal amounts were subjected to immunoprecipitation using anti-FLAG antibody. Coprecipitated endogenous Erbin protein was detected by Western analysis. Monitoring the amount of IκBα in the total lysates showed a degradation of IκBα 60 min after infection, as expected for successful bacterial invasion using this infection model (36). Importantly, a change in the amounts of coprecipitated endogenous and ectopically expressed Erbin was observed, whereas virtually equal amounts of Nod2 were precipitated at each time point of the experiment, and the Erbin signal in the total lysate did not change over time (Fig. 5D and data not shown). The amount of coprecipitated Erbin protein was lower than that in uninfected cells at early times of infection (10 to 20 min) and increased over time to reach a peak at around 30 to 40 min postinfection (Fig. 5D). To rule out the possibility that this might reflect small changes in Erbin expression after infection that are not detectable by Western analysis, the same experiment was conducted with coexpression of ectopic FLAG-Nod2 and Myc-Erbin. Again, a clear change over time in the amount of coprecipitated Myc-Erbin was observed, supporting the finding of the previous experiment (Fig. 5D).

These results show conclusively that S. flexneri triggers Nod2-mediated NF-κB activation, and they support a role of Erbin in bacterially triggered Nod2 activation. Indeed, Nod2 and Erbin are present at the entry foci during S. flexneri infection of human cells, and the stoichiometry of the Nod2/Erbin complex is altered during infection.