Nod1 Participates in the Innate Immune Response to Pseudomonas aeruginosa*

The mammalian innate immune system recognizes pathogen-associated molecular patterns through pathogen recognition receptors. Nod1 has been described recently as a cytosolic receptor that detects specifically diaminopimelate-containing muropeptides from Gram-negative bacteria peptidoglycan. In the present study we investigated the potential role of Nod1 in the innate immune response against the opportunistic pathogen Pseudomonas aeruginosa. We demonstrate that Nod1 detects the P. aeruginosa peptidoglycan leading to NF-κB activation and that this activity is diminished in epithelial cells expressing a dominant-negative Nod1 construct or in mouse embryonic fibroblasts from Nod1 knock-out mice infected with P. aeruginosa. Finally, we demonstrate that the cytokine secretion kinetics and bacterial killing are altered in Nod1-deficient cells infected with P. aeruginosa in the early stages of infection.

The mammalian innate immune system recognizes pathogen-associated molecular patterns through pathogen recognition receptors. Nod1 has been described recently as a cytosolic receptor that detects specifically diaminopimelate-containing muropeptides from Gram-negative bacteria peptidoglycan. In the present study we investigated the potential role of Nod1 in the innate immune response against the opportunistic pathogen Pseudomonas aeruginosa. We demonstrate that Nod1 detects the P. aeruginosa peptidoglycan leading to NF-B activation and that this activity is diminished in epithelial cells expressing a dominant-negative Nod1 construct or in mouse embryonic fibroblasts from Nod1 knock-out mice infected with P. aeruginosa. Finally, we demonstrate that the cytokine secretion kinetics and bacterial killing are altered in Nod1deficient cells infected with P. aeruginosa in the early stages of infection.
Pseudomonas aeruginosa is a typical opportunistic pathogen that can cause a variety of systemic infections, particularly in immunocompromised patients such as those with cancer, AIDS, cystic fibrosis, and burns (1,2). The virulence of P. aeruginosa is multifactorial including several cell-associated and secreted proteins such as elastase A (3), phospholipase C (4), and those translocated through the type III secretion system (5).
The mammalian innate immune system recognizes pathogen-associated molecular patterns (PAMPs) 8 through pathogen recognition receptors (PRRs). PAMPs are conserved structures present in all or large groups of microorganisms but are absent in mammalian cells. PRRs are located either at the cell membrane, where they respond to extracellular PAMPs, or in the cytosol of host cells, where they respond to PAMPs that are present in the cytosol (6). Human Toll-like receptors interact mainly with extracellular PAMPs; for example, TLR5 interacts with bacterial flagellin (7), and TLR4 interacts with bacterial lipopolysaccharide (LPS) (8). Signaling through TLRs leads to the activation of NF-B and, subsequently, of NF-B target genes.
Recent studies have highlighted the importance of Nod1 and Nod2 in innate immunity. It is thought that they act as cytoplasmatic sensors of bacteria and bacterial products (5). Nod1 structurally contains a CARD domain, a centrally located nucleotide-binding oligomerization domain, and multiple C-terminal leucine-rich repeats. Nod1 engages in a homophilic CARD-CARD interaction with Rip2, a CARD-containing protein kinase that in turn interacts with IKK␥ and results in NF-B activation (9). Nod1 and Nod2 have been recognize muropeptides from bacterial peptidoglycan (PG), although they require distinct molecular motifs within PG to achieve this sensing (10).
Although traditionally considered as an extracellular pathogen, P. aeruginosa has been shown to be internalized by and to survive within different types of mammalian cells, in particular epithelial respiratory cells from nonjunctional epithelia (1). P. aeruginosa infections are usually associated with a marked host inflammatory response, and previous studies have shown that bacterial products such as LPS, flagellin, and CpG DNA from P. aeruginosa are strong stimuli for TLR4 (11), TLR5 (12), and TLR9 (13), respectively.
In this study we investigated the ability of P. aeruginosa to induce NF-B activation through Nod1. Our results demonstrate that in addition to data from literature that show the role of TLRs in the detection of products from P. aeruginosa, signaling through Nod1 constitutes a potential important back-up system against P. aeruginosa.

EXPERIMENTAL PROCEDURES
Cell Culture-The human embryonic kidney epithelial cell line HEK293T were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum and 2 mM L-glutamine. Cells were maintained in 95% air, 5% CO 2 at 37°C.
Isolation of Mouse Embryonic Fibroblasts (MEFs)-Wild-type and Nod1 knock-out mice were used to isolate MEFs as follows. The torsos of 13.5-day-old embryos were washed in cold phosphate-buffered saline (PBS), pH 7.2. After removal of the heart and liver, the embryos were transferred to a Petri dish containing fresh PBS, cut into small pieces with a sterile blade, and resuspended in trypsin-EDTA in Hanks' balanced salt solution (Invitrogen). The suspensions were incubated twice for 20 min at 37°C under magnetic stirring and then filtered and centrifuged at 1500 rpm for 5 min. The pellet was resuspended in 20 ml of complete medium (DMEM; 10% fetal calf serum, 2 mM L-glutamine, 50 units/ml penicillin) and cultured to 90% confluence. Each culture was expanded to five new dishes and frozen in liquid nitrogen. Before use, the cells were thawed in complete medium.
Bacteria and Other Reagents-The following P. aeruginosa strains were used in this study. PAO-1, PAK, and a PAK mutant lacking flagellin (PAK⌬fliC) (the last two strains were kindly provided by Dr. Steve Lory, Harvard Medical School, Boston, MA). tri-DAP was obtained by treatment of the corresponding MurNAc peptide with partially purified Escherichia coli N-acetylmuramoyl-L-alanine amidase according to the method of van Heijenoort et al. (14). Muramyl dipeptide (15) was obtained from Calbiochem (San Diego, CA). PG from P. aeruginosa PAO-1 was prepared as described previously (16). PG samples were lyophilized in a SpeedVac to estimate the amount of PG and determine the yield per colony-forming unit. Peptidoglycan samples were resuspended in pyrogene-free ultrapure water (Biochrom AG, Berlin, Germany). Amino acid and amino-sugar compositions were determined with a Hitachi L8800 analyzer (ScienceTec, Les Ulis, France) after hydrolysis of samples in 6 M HCl at 95°C for 16 h. Recombinant TNF-␣ was from Pharmingen (San Diego, CA).
Infection Protocol-Bacteria were grown in Trypticase soy broth at 37°C overnight under agitation, harvested by centrifugation, and resuspended in DMEM culture medium to an OD 660 of 0.1 (ϳ10 8 colonyforming units/ml). HEK293T cells or MEFs grown to 80% confluence in 24-well plates in DMEM supplemented with 10% fetal calf serum were infected with bacteria at a multiplicity of infection of 100 or stimulated with flagellin-enriched supernatants from P. aeruginosa cultures unless otherwise stated. After incubation for 1 h at 37°C, extracellular bacteria were removed by washing and by incubation for different periods with DMEM containing 300 g/ml gentamicin to kill the remaining extracellular bacteria. Flagellin-enriched supernatants were prepared as follows. Bacterial cells from 96-h-old cultures were harvested by centrifugation, and the supernatants were filter-sterilized with 0.22-m membranes and boiled at 100°C for 30 min to destroy proteins other than flagellin.
Plasmids, Transfection, and Luciferase Assays-Studies on the synergistic activation of NF-B by PGs were carried out as described previously (17). Studies on NF-B activation following P. aeruginosa infection were carried out as described (18). Briefly, cells were transfected with 75 ng of the reporter plasmid Ig-luc alone or in combination with 3 ng of Nod1-HA or a dominant-negative form of Nod1 (DN-Nod1-HA/⌬CARD4-HA). The pcDNA3.1 vector was used to balance the transfected DNA concentration. The PG preparations were used at 1 g/ml, and tri-DAP or muramyl dipeptide (MurNAc-L-Ala-D-isoGln) (10 pmol) were used as the positive and negative control for Nod1, respectively.
Western Blots-HEK293T cells were transfected as described above. Briefly, cells were harvested from 6-well plates to detect Nod1-HA expression. Cells were lysed and kept on ice until separation in SDS-8% PAGE. After separation, proteins were transferred to nitrocellulose membranes and probed with an anti-HA (Upstate Biotechnology). After three washes, membranes were incubated with an anti-rabbit IgG conjugated to horseradish-peroxidase. Proteins were visualized by Western Lightening ECL labeling (PerkinElmer Life Sciences). Blots were also probed with a control antibody against human ␤-actin.
Cytokine Measurements-Murine cytokines (TNF-␣, MIP-2, and KC) released into the medium were measured using BD Pharmingen opt enzyme immunoassay kits or R&D Systems' Duoset enzyme-linked immunosorbent assay kits.
Intracellular Bacterial Survival-MEFs obtained from wild-type and Nod1 knock-out mice were infected with P. aeruginosa for 1 h, washed, and incubated with gentamicin-containing culture medium. After different periods of incubation, the cells were washed with sterile PBS and lysed with 0.1% Triton X-100 in PBS. Viable counts of cell-associated bacteria were determined by serial plating of cell lysates.
Cell Viability Assays-The cell membrane integrity was analyzed by flow cytometry after propidium iodide labeling as described previously (2). Briefly, control non-infected or infected cells were detached from the microplate wells with a 0.05% trypsin, 0.02% EDTA solution and pooled with spontaneously detached cells present in the cell culture supernatants. After rinsing, cells were resuspended in PBS (pH 7.2, containing 1% bovine serum albumin) and incubated with propidium iodide at 5 g/ml for 5 min at room temperature. Samples were kept on ice and analyzed with a FACScalibur flow cytometer (BD Biosciences) equipped with a standard argon ion laser.
Statistics-Data were analyzed with the Student's t test and Mann-Whitney test, as appropriate. Differences in data values were considered significant at a p value of less than 0.05.

P. aeruginosa Induces NF-B Activation in Epithelial
Cells-We first investigated whether P. aeruginosa was able to induce NF-B activation in HEK293T cells transiently transfected with a Ig-luc reporter plasmid. Cells were exposed to different concentrations of a laboratory strain (PAO-1) for 1 h and treated for an additional 3-h period with gentamicin-containing culture medium to eliminate extracellular bacteria. The increase in luciferase activity of infected cells was shown to be up to 5.2 Ϯ 0.5-fold higher than the increase in non-infected control HEK293T cells. Additionally, the increase in luciferase activity was shown to be dependent on the number of bacteria added to the cells (Fig. 1A).
Flagella Do Not Account for NF-B Activation of P. aeruginosa-infected Cells-HEK293T cells, although not expressing TLR2 or TLR4, do express TLR5 (20), which recognize bacterial flagellin. Because in our assay HEK293T cells were exposed to a P. aeruginosa flagellated strain for 1 h before the gentamicin treatment, and flagellin monomers may be released from the bacterial cell, we could not exclude the participation of TLR5 in the response to P. aeruginosa infection detected in our assays. To determine whether flagella accounted for the NF-B activation, we compared the levels of luciferase activity following infection with the parental laboratory strain PAK and the nonflagellated mutant PAK⌬flic. As shown in Fig. 1B there was no significant difference in the luciferase activity of cells infected with PAK or PAK⌬flic. Gentamicin protection assays with PAK and PAK⌬fliC showed no difference in the invasive ability between these two strains (data not shown). Interestingly, the flagellin-enriched supernatant from the PAK strain was still capable of inducing high levels of NF-B activation, whereas the supernatant of the PAK⌬flic strain had no activity, demonstrating that the flagellin monomers of P. aeruginosa are a potent proinflammatory stimulus (Fig. 1B). The results suggest that flagellin monomers are a better stimulus for TLR5 than the assembled molecule, as described previously (21).
Nod1 Participates in the Innate Immune Response to P. aeruginosa Infection-The observation that PAK⌬flic strain could still induce NF-B activation, apparently in a TLR-independent manner, led us to hypothesize that the recently described cytosolic receptor Nod1 could be involved in the innate immune recognition of P. aeruginosa. It has P. aeruginosa Detection by Nod1 NOVEMBER 4, 2005 • VOLUME 280 • NUMBER 44 been demonstrated that Nod1 detects DAP-containing muropeptides from the PG of Gram-negative bacteria. To determine whether Nod1 is involved in P. aeruginosa detection, we first tested whether the highly purified PG of P. aeruginosa is recognized by Nod1. To this end, we co-transfected HEK293T cells with a Nod1 expression vector and the purified PG from P. aeruginosa. The PG induced a 7.0-fold increase in the luciferase activity when compared with the nonstimulated control ( Fig. 2A). In contrast, the transient transfection of a dominant-negative form of Nod1 (DN-Nod1) expression vector inhibited the activation of the co-transfected Ig-luc reporter gene in response to P. aeruginosa PG by ϳ3-fold (Fig. 2C). After the observation that P. aeruginosa PG was detected by Nod1, we questioned whether Nod1 would participate in vitro in the induction of NF-B during P. aeruginosa infection. To this end HEK293T cells transfected with increasing amounts of DN-Nod1 were infected with PAK and PAK ⌬fliC strains. The results showed that luciferase activity was inhibited in response to infection PAK and PAK⌬fliC strains in a dose-dependent manner (Fig. 3A). Accordingly, MEFs from Nod1-deficient mice infected with PAO-1 strains showed significant lower activation of NF-B (Fig. 3B).
Nod1 Deficiency Leads to Slower Cytokine Production Kinetics-We next investigated the consequences of Nod1 signaling in the production of proinflammatory cytokines and/or chemokines in response to P. aeruginosa infection. To this end MEFs from wild-type and Nod1 knock-out mice were infected with PAO-1 strain for different periods of time, and the supernatants were assayed for the presence of TNF-␣, MIP-2, and KC. P. aeruginosa infection did not lead to TNF-␣ or MIP-2 secretion either in the knock-out cells or in the control wild-type cells (data not shown). In contrast, it induced the secretion of high amounts of KC. At 3 h post-infection the wild-type cells secreted significantly more KC than the knock-out cells. However, at the end of 20 h postinfection both wild-type and knock-out cells secreted similar amounts of KC (Fig. 4A).

Nod1 Knock-out Cells Show Delayed Bacterial
Killing-Data from the literature have suggested that Nod2, expressed in intestinal epithelial cells, may serve as an antibacterial factor. The fact that Nod1 and Nod2 are structurally closely related and that both activate NF-B through the same adapter protein (22) led us to investigate the survival of internalized P. aeruginosa in wild-type and Nod1-deficient cells after different periods of time. As shown in Fig. 4B, it was again in the early periods post-infection (2 h) that we observed a significant difference between wild-type and knock-out cells. The number of intracellular P. aeruginosa in Nod1 knock-out cells is significantly higher than in wild-type cells. The difference in bacterial survival decreases with time, and after 10 h both wild-type and knock-out have eliminated intracellular P. aeruginosa. Flow cytometry analysis did not show differences in membrane permeability between wild-type and Nod1-deficient cells at 2 h post-infection, indicating that the bacterial killing was not due to the presence of gentamicin in the cells (data not shown).

DISCUSSION
The innate immune system recognizes and responds to an array of bacterial products, such as LPS, lipoproteins, and flagellin, through TLRs (7,8,16). These products are highly conserved among both pathogenic and commensal microorganisms. The recognition of PG by PRRs has been shown to be a important feature in the innate immunity of flies and mammals (23). In mammals TLR2 was the first PRR implicated in PG recognition. However, it has been recently demonstrated that TLR2  does not detect PG but likely detects lipoteicoic acid or lipoprotein contaminants in commercial PG preparations, which makes Nod proteins the only known receptors for PG in mammals (8). Nod1 was described as a cytosolic receptor that detects specifically DAP-containing muropeptides from Gram-negative bacterial peptidoglycan. Indeed it has been demonstrated that it plays a major role in the immune response against intracellular bacteria (18). More recently, it has been shown that Nod1 may participate in the detection of Helicobacter pylori, an extracellular pathogen that is able to translocate its PG into the host cells (24). In the epithelium, Nod1 plays an important role in the activation of NF-B following infection with Gram-negative bacteria that do not activate NF-B through TLR signaling pathways (9).
Induction of the host epithelial cell proinflammatory gene program is essential for rapidly activating the host mucosal inflammatory response and is important for bacterial clearance and host survival following infection (25). Several studies indicate that epithelial cells have at least two independent physiologically relevant signaling pathways for activating NF-B and epithelial cell proinflammatory genes in response to bacterial infection. One pathway involves TLRs. For example, pathogenic bacteria that have flagella and adhere to or invade epithelial cells can activate NF-B and NF-B target genes by signaling through the TLR5 receptor (7). Signaling through Nod1 provides epithelial cells with an independent pathway for rapidly activating NF-B and the subsequent mucosal inflammatory response, which can favor host survival following infection with bacteria that bypass TLR recognition (26). Importantly, both pathways converge at IB kinase, making IB kinase a key molecular target for manipulating mucosal inflammatory responses that are initiated by host epithelium.
We demonstrated the participation of Nod1, a cytosolic receptor, in the detection of a bacterium that is not considered to be an invasive pathogen. This has been demonstrated using an epithelial cell line that does not express TLRs (except for TLR5) but expresses Nod1 or a dominant-negative form of Nod1. In this case, Nod1 might represent a backup, in addition to the TLR signaling, in detection of intracellular bacteria or those capable of translocating PG into the cells. In addition, even the strain lacking flagella was able to induce NF-B activation via Nod1, once more strengthening the importance of this back-up system against bacteria that may bypass TLR recognition.
P. aeruginosa-induced production of inflammatory mediators is initiated by the interaction between the bacteria and the host cell receptors (27). In the context of an infection, bacteria likely activate the immune response through several PRRs simultaneously. A major effort has been made to identify the molecules responsible for the initiation of P. aeruginosa-induced inflammation. Several cell surface molecules have been implicated in the direct interaction with P. aeruginosa, including complement receptor 3 (28), gangliotetraosylceramide (29), and syndecan-1 (30). More recently, independent studies have demonstrated that LPS, flagellin, and CpG DNA from P. aeruginosa initiate inflammatory responses through TLR4, TLR5, and TLR9, respectively (12,13,31). Thus, even though some PRRs seem to play a more relevant role than others, it is likely that neither of these receptors functions individually as the only component responsible for P. aeruginosa-induced immune response in vivo. Furthermore, interaction between some of these receptors potentiates their activity. For example, it has been demon-strated recently that Nod1 and TLR4 act synergistically in the induction of proinflammatory cytokines (32,33).
However, despite all that, we still observed differences in the response to P. aeruginosa in MEFs from normal mice and mice deficient in Nod1, all of which express TLRs and other PRRs. This might be because of the elimination of the extracellular bacteria after 1 h of infection in the gentamicin exclusion assays, allowing the observation of a differential response to the intracellular bacteria. Should that be the case it demonstrates again the importance of an intracellular detection system that provides the cells with additional mechanisms to fight the bacteria that manage to reach the cytosol evading extracellular immune responses. Indeed, wild-type MEFs were able to eliminate intracellular bacteria more efficiently than Nod1 knock-out MEFs after 2 h of infection in gentamicin-containing medium. The mechanisms involved in Nod1dependent intracellular bacterial killing were not investigated in this study. Again, by analogy with TLRs, one can speculate that production of nitric oxide may be involved, but only future investigation will confirm or not such speculations.
Two major features of P. aeruginosa infection are the recruitment of neutrophils and the production of various cytokines and chemokines in the local tissue. KC is a CXC chemokine that recruits and activates neutrophils and is up-regulated during infection in epithelial cells. This chemokine is among the earliest new proteins produced at sites of inflammation, consistent with its role in regulating the early infiltration of neutrophils (34). In the present study, Nod-1-deficient cells secreted significantly less KC in the first hours of infection, which might lead to a delay in providing signals that are essential for the onset of inflammatory response by infected epithelial cells. A number of prior studies have observed differential expression of this chemokine, both in terms of cell type and temporal pattern (35,36). In accordance with our results, in several models of bacterial infection in mice, maximum KC expression occurs in the first hours, whereas MIP-2 (another CXC chemokine that we have tested and that was not detected in periods of infection of up to 24 h) peaks on day 2. Complex temporal patterns of chemokine expression in vivo have been reported previously in a number of settings; however, the regulation of such expression has not been explored in detail. In murine models of acute Gram-negative bacterial and fungal pneumonia, these chemokines appear crucial to the innate host defense. Specifically, in a murine model of Pseudomonas pneumonia, KC levels are associated with the presence of neutrophils in the lung (19). Furthermore, studies have demonstrated that site-specific transgenic expression of KC resulted in enhanced clearance of bacteria after Pseudomonas challenge (36). It will be interesting to look at whether this delayed secretion in Nod1 knock-out mice leads to decreased neutrophil recruitment and exacerbation of disease.
In summary, we conclude that Nod1 expressed by epithelial cells takes part in the activation of NF-B and the up-regulated production of an important epithelial cell chemoattractant in response to P. aeruginosa. Murine models of infection will help to further elucidate the role of Nod1 in P. aeruginosa infection.