Crohn disease (CD) usually develops in the young adult, with the majority of cases diagnosed when the patient is between 15 and 35 years of age. However, CD can affect people at any age and approximately 10% of cases are in patients younger than 18 years of age at presentation. CD predominantly affects white patients, with a prevalence rate of 150/100,000. Inflammatory bowel disease (IBD) is largely a disease of the industrialized world, especially the United States and Europe, and is more common in urban areas and northern climates. No clear-cut Mendelian pattern of inheritance in CD has been established, but CD has a known genetic component, with 25% of patients with CD having a family member with some form of IBD (1).
Signs and symptoms of CD include frequent (bloody) diarrhea, abdominal pain, fatigue, loss of appetite, weight loss, fever, stomatitis, and perianal fistula or fissures. A proportion of patients present with extraintestinal manifestations such as arthritis, erythema nodosum, or pyoderma gangrenosum. In pediatric cases of CD, growth failure is observed in 75% of patients (1).
Clear evidence exists for the activation of the immune response in CD. The lamina propria is infiltrated with lymphocytes, macrophages, and other cells of the immune system. In any immune response a specific antigen serves as a trigger for the response and as a target for the effector arm of the response. During the past 35 years, an intensive search has been conducted for the antigens that trigger the immune response in CD. Immune activation in CD is largely confined to the gastrointestinal tract; therefore, the search for the antigenic trigger has focused on the intestinal lumen. Most of the foreign antigens in the intestinal lumen are of microbial or dietary origin (2).
Three major hypotheses on the antigenic trigger in CD have been postulated. One hypothesis is that the antigenic triggers are microbial pathogens that have not yet been identified because of fastidious culture requirements. According to this hypothesis, the immune response in CD is an appropriate but ineffective response to these pathogens. Various viruses and bacteria have been proposed as candidate organisms, but little evidence has been found to support any of these organisms as having a causative role in CD. The second hypothesis is that the antigenic trigger in CD is some common dietary antigen or usually nonpathogenic microbial agent against which the patient mounts an abnormal immune response. In healthy people a finely tuned, low-grade chronic inflammation is present in the intestinal lamina propria. Presumably, this chronic inflammation is a product of chronic exposure of the lamina propria to luminal antigens. Failure to suppress this inflammatory response could result in the uncontrolled immune activation seen in CD. As a result of failure of normal suppressor mechanisms, immune activation in CD may be an inappropriate vigorous and prolonged response to some normal luminal antigens. The third hypothesis for CD is that the antigenic trigger is one expressed on the patient's own cells, particularly intestinal epithelial cells. This is an autoimmune theory for the pathogenesis of CD. In this theory, the patient mounts an appropriate immune response against some luminal antigen, either dietary or microbial; however, because of similarities between proteins on epithelial cells and the luminal antigen, the patient's immune system also attacks the epithelial cells (3).
Crohn disease is a consequence of a disturbance in the normal immunological unresponsiveness of the mucosal immune system to components of the mucosal microflora. The hyperresponsiveness to these components that ensues gives rise to the T helper type 1 (Th1) cell–mediated inflammation that underlies all forms of the disease. Activated Th1 cells secrete cytokines, such as tumor necrosis factor (TNF)-α, interleukin (IL)-2, and interferon (IFN)-γ. IFN-γ, in turn, activates macrophages, causing them to secrete excess proinflammatory cytokines (eg, IL-1β, TNF-α). Activated macrophages contribute to epithelial injury by secreting TNF-α and reactive oxygen species and by recruiting neutrophils, which also produce free radicals. Neutrophils are also recruited by IL-8 secreted by epithelial cells following activation or injury by bacteria. Neutrophils release reactive oxygen species, and oxygen species injure the epithelial cells. Together, the macrophages and neutrophils produce prostaglandin E2 and leukotriene B4 that contribute to the vasodilation and enhanced vascular permeability characteristic of CD (4). CD bears the immunological stigmata of an exaggerated CD4+ Th1 response. Thus, intestinal CD4+ T cells isolated from patients with CD produce large amounts of the Th1 signature cytokine IFN-γ. Mucosal macrophages from patients with CD also produce large amounts of the Th1-inducing cytokines IL-12, IL-18, and TNF-α. Th1 cell resistance to apoptosis and increased cell cycling in CD inflammation appear to be sustained by these cytokines (5). Blocking the pathways that confer resistance of Th1 cells to apoptotic stimuli and using drugs that enhance mucosal T cell death, such as the immunosuppressive agent azathioprine or the antibody to TNF-α infliximab, are effective in downmodulating intestinal inflammation (6).
Identifying the particular antigen(s) that drive the Th1 inflammatory response in the face of the large amount of potential antigens in the gut has proven difficult. Nevertheless, the likelihood is that bacterial antigens are involved, because stimulation of mucosal CD4 cells from patients with CD with extracts of their own commensal flora can induce IFN-γ production (7). Clinical observations also support a role for antigens derived from the commensal flora. Thus, for example, the antibiotic metronidazole is of therapeutic benefit in CD of the distal colon, eliminating the commensal flora and resulting in decreased inflammation (8).
The cellular and molecular mechanisms of interaction between intestinal mucosal cells and the resident luminal bacteria in healthy individuals and patients with CD is not yet fully understood, but is an area of active investigation. Recently, mutations in NOD2/CARD15, encoding an intracellular bacteria-sensing protein expressed mainly by macrophages and dendritic cells, have been associated with CD.
Different theories exist for nuclear factor–κB (NF-κB) activation in NOD2 variants. Mutations in NOD2 cause a loss of function of NOD2, resulting in a decreased suppression of the Toll-like receptor-2 (TLR2)–driven Th1 response via NF-κB. Conversely, it has been hypothesized that mutated NOD2 enhances (ie, causes gain of function) the sensitivity of macrophages to the nucleotide-binding oligomerization domain protein (NOD) ligand muramyl dipeptide (MDP), potentiating NF-κB activity. This review focuses on the functional consequences of NOD2/CARD15 mutations in the pathogenesis of CD.
PATTERN RECOGNITION MOLECULES
The host mucosa is exposed to vast numbers of metabolically active microbial cells and cell wall components, such as lipopolysaccharide (LPS), a component of the outer membrane of Gram-negative bacteria, and peptidoglycan (PGN), a complex amino sugar present in Gram-negative and Gram-positive bacteria. A unique feature of host–microbial interactions in the intestine is the lack of proinflammatory responses in the mucosa exposed to the resident luminal microflora while retaining the capability to respond to luminal pathogenic bacteria via the recruitment of acute inflammatory cells from the systemic circulation. The gut epithelium itself can also directly sense commensal bacteria and pathogens; integral to this are the mammalian pattern recognition receptors (PRRs), which recognize conserved structures of bacteria and viruses and generally activate proinflammatory pathways alerting the host to infection. Two different classes of PRRs are involved (9,10). The TLRs are usually associated with cell membranes and have an external leucine-rich repeat (LRR) recognition domain and an intracellular IL-1 receptor (IL-1R)–like signaling domain that activates intracellular signaling pathways (11). To date, 11 members of the TLR family have been identified in mammals. The subcellular localization of different TLRs correlates to some extent with the molecular patterns of their ligands. TLR1, TLR2, and TLR4 are located on the cell surface and are recruited to phagosomes after activation by their respective ligands. By contrast, TLR3, TLR7, and TLR9, all of which are involved in the recognition of nucleic acid–like structures, are not expressed on the cell surface. For example, TLR9 has recently been shown to be expressed in the endoplasmic reticulum (11). After ligand binding TLRs/IL-1Rs dimerize and undergo the conformational change required for the recruitment of downstream signaling molecules (10). These include the adaptor molecule myeloid differentiation primary response protein 88 (MyD88), which activates the serine/threonine kinases of the IL-1R–associated kinase family (IRAK) and TNF receptor–associated factor 6 (TRAF6), subsequently leading to the degradation of the inhibitor of NF-κB activity protein IκB. This results in the activation of NF-κB and its translocation to the nucleus (12). In addition to TLRs, microbial products can be recognized by members of the NOD family.
NUCLEOTIDE-BINDING OLIGOMERIZATION DOMAIN
The NODs are cytosolic proteins that contain a nucleotide-binding oligomerization domain. These proteins include key regulators of apoptosis and pathogen resistance in mammals and plants. A large number of NODs contain leucine-rich repeats (LRRs), hence referred to as NOD-LRR proteins. Genetic variation in several NOD-LRR proteins, including human Nod2, Cryopyrin, and MHC class II transactivator (CIITA), as well as mouse neuronal apoptosis inhibitor protein (NAIP) 5, is associated with inflammatory disease or increased susceptibility to microbial infections. Nod1, Nod2, Cryopyrin, and Ipaf have been implicated in protective immune responses against pathogens. Together with Toll-like receptors, Nod1 and Nod2 appear to play important roles in innate and acquired immunity as sensors of bacterial components. Specifically, Nod1 and Nod2 participate in the signaling events triggered by host recognition of specific motifs in bacterial peptidoglycan and, upon activation, induce the production of proinflammatory mediators. NAIP5 is involved in host resistance to Legionella pneumophila through cell autonomous mechanisms, whereas CIITA plays a critical role in antigen presentation and development of antigen-specific T lymphocytes. Thus, NOD-LRR proteins appear to be involved in a diverse array of processes required for host immune reactions against pathogens.
Family of NOD-LRR Proteins
The NOD-LRR proteins are also referred to as the CATERPILLER (CARD, transcription enhancer, R [purine]–binding, pyrin, lots of leucine repeats) family (13). NOD-LRR proteins are expressed primarily in immune cells, although the expression of certain proteins, such as Nod1, is ubiquitous. The majority of animal and plant NOD-LRR proteins are composed of 3 distinct functional domains: an aminoterminal effector domain involved in signaling, a centrally located regulatory NOD-domain, and carboxyl-terminal LRRs that serve as a ligand-recognition domain (Fig. 1).
The effector domains of mammalian NOD-LRR proteins are structurally variable, linking these proteins to multiple signaling pathways and biological functions. These effector domains are involved in homophilic and heterophilic interactions with downstream signaling partners. The diversity of the effector domains allows NOD-LRR proteins to interact with a wide array of binding partners and to activate multiple signaling pathways. Effector domains involved in homophilic association include the caspase-recruitment domain (CARD) (14) and the pyrin domain (PYD, also called DAPIN and PAAD) (15). Both the PYD and the CARD belong to the death domain-fold family characterized by 6 α-helices that are tightly packed and include the death domain (DD) and death effector domain (DED) (16). Only 3 human NOD-LRR proteins possess amino-terminal CARDs, whereas NOD-LRR proteins possessing a PYD are by far the most numerous and include 14 proteins designated as NACHT, leucine-rich repeat, and pyrin domain-containing family proteins (NALPs).
Several NOD-LRR proteins contain amino-terminal sequences that are not involved in homophilic protein interactions, including NAIPs and CIITA (17,18). Instead of a CARD or PYD, the amino termini of NAIPs are composed of amino-terminal baculovirus-inhibitor-of-apoptosis repeats (BIRs) (19). CIITA is a transcriptional coactivator involved in the regulation of major histocompatibility complex class (MHC) genes, especially class II (MHC-II) (17,19). CIITA contains an amino-terminal transcriptional activation domain that is essential for MHC gene transactivation through its interaction with multiple nuclear factors, including CBP/p300, RFX5, NF-Y, and CREB (19) (Fig. 2).
Detection of Microbial Products by NOD-LRR Proteins
Whereas initial studies identified lipopolysaccharide (LPS) as a NOD2 ligand (20), it is now well established that the NOD1 and NOD2 ligands are the peptidoglycan (PGN)–derived peptides γ-D-glutamyl-meso-diaminopimelic acid (iE-DAP) (21,22) and MDP, respectively (23,24). Because PGN from Gram-positive and Gram-negative bacteria contains MDP, NOD2 functions as a general sensor of most, if not all, bacteria. By contrast, because PGNs from Gram-positive bacteria do not contain iE-DAP (except for PGNs derived from specific Gram-positive bacteria such as Listeria and Bacillus species and from many Gram-positive bacteria in the soil) (25), NOD1 mainly senses products from Gram-negative bacteria. Confirmation of these specificities has come from the finding that macrophages isolated from NOD1- or NOD2-deficient mice are completely unresponsive to their respective ligands (21,26). The NOD protein ligands need to reach the LRR domains of the respective NOD protein for activation of this protein to be initiated. However, information about how this is accomplished is sparse at present, especially in the case of antigen-presenting cells (APCs). One possibility that relates to phagocytic cells such as macrophages or dendritic cells (DCs) is that these cells generate the peptide ligands by ingesting whole bacteria and then digesting them in phagolysosomes (27,28). In epithelial cells a slightly different process may occur in that the apical peptide transporter PEPT1 seems to play a role in the delivery of MDP. This is indicated by the finding that MDP that is taken up by PEPT1 into colonic epithelial cells subsequently mediates the activation of NF-κB (29). In addition, it has recently been shown that Helicobacter pylori can “inject” PGN into cells through a type IV secretion system, which is encoded by a pathogenicity island (30). This discovery indicates that PGN can enter cells by various mechanisms that involve bacteria–host interactions. After small peptides derived from PGN have been released into the cytosol, they are thought to interact with NOD1 or NOD2 through the LRR domains of these molecules. However, it should be noted that, as is the case for activation of most TLRs by their respective ligands (31), there is as yet no direct evidence for the binding of the NOD1 and NOD2 ligands to these domains. The postulated interaction is then proposed to initiate the activation of NOD1 and NOD2 through the induction of a complex conformational change (32,33). Our understanding of this change comes in part from studies of activation of apoptotic protease-activating factor 1 (APAF1), an NLR-family (NACHT-LRR family domain present in NAIP, CIITA, HET-E, and TP1-LRR families) member that is involved in caspase activation and apoptosis (34,35). Activation of APAF1 is initiated by the interaction of its WD40 domain with its ligand (cytochrome c), as well as by the binding of dATP or ATP to an ATP-binding cassette (ABC) or oligomerization cassette in the NOD. The molecule then undergoes self-oligomerization, which enables it to bind its downstream effector molecule, caspase-9, through a CARD–CARD interaction. The large molecular complex that is formed in this way, which is known as the apoptosome, then facilitates activation of the bound caspase-9, possibly by bringing caspase molecules into juxtaposition (34,35). That this activation model applies to NLRs in general (and to NOD1 and NOD2 in particular) is indicated by the presence of structural similarities between NOD proteins and APAF1: the N-terminal region of the central NOD in NLRs contains an ABC and an oligomerization module. At least in the case of NOD2, the introduction of mutations into the ABC region abolishes NOD2 signaling (33). In addition, it has been shown that both NOD1 and NOD2 undergo self-oligomerization following the binding of PGN-derived ligand (32,33). In one model of NOD-protein activation based on the APAF1–caspase-9 pathway, Inohara et al (25) proposed that oligomerization of NOD1 or NOD2 also allows binding to a downstream effector molecule through a CARD–CARD interaction, in this case involving the Rip-like interacting caspase-like apoptosis-regulatory protein kinase (RICK; also known as RIP2 or CARDIAK), and this, in turn, leads to crossactivation of RICK. However, further work is necessary to establish this possibility.
Signal Transduction Pathways of NOD-LRR
One of the main outcomes of NOD1 and NOD2 activation by their respective ligands is the activation of NF-κB. Whereas such activation was clearly evident in iE-DAP- or MDP-stimulated epithelial cells that had been transfected with constructs encoding wild-type NOD1 or NOD2, it was reduced in cells that were transfected with constructs encoding a mutated NOD2 that had alterations in the LRR domain (23).
Consistent with this, after stimulation with MDP, translocation of NF-κB subunits to the nucleus is observed in human and mouse APCs that have intact NOD2 but not in APCs that are deficient in NOD2 or have a mutation in NOD2 (23,34). The activation of NF-κB by NOD1 and NOD2 occurs exclusively through the downstream effector molecule RICK. This is demonstrated by the finding that transfection of RICK-deficient fibroblasts with constructs encoding NOD1 or NOD2 results in severely defective NF-κB activation (35). However, it should be noted that RICK-deficient macrophages also have reduced cytokine responses following stimulation with LPS, lipoteichoic acid, and PGN, indicating that TLR2 and TLR4 may also use RICK as a downstream effector molecule, although the existence of a TLR4–RICK pathway is controversial (35). RICK is a CARD-containing serine/threonine kinase that physically associates with the CARD(s) of NOD1 and NOD2 through CARD–CARD interactions (36,37). As shown recently by Abbott et al (38), following its activation by NOD2, RICK mediates K63-linked polyubiquitylation of inhibitor of NF-κB (IB)-kinase-γ (IKK-γ; also known as NEMO; the key member of the IKK complex) at a unique ubiquitylation site (the lysine residue at position 285). As shown previously, K63-linked polyubiquitylation is associated with activation of the NF-κB pathway (39), and in the case of the RICK–IKK-γ interaction, it is indeed followed by phosphorylation of IKK-γ and downstream activation of NF-κB, leading to the translocation of transcriptional components of NF-κB to the nucleus. So, in activating the IKK complex, RICK activates an E3 ubiquitin ligase that promotes K63-linked polyubiquitylation or inhibits an enzyme (eg, cylindromatosis protein) that de-ubiquitylates proteins that are modified with K63-linked polyubiquitin, so RICK does not require its own kinase activity for this function. Recently, it has been shown that activated NOD2 (but not NOD1) also interacts with the intracellular molecule GRIM19 (gene associated with retinoid-IFN–induced mortality 19) and that such an interaction may be required for optimal NF-κB activation. However, neither the structural basis of this interaction nor the mechanism of its relation to NF-κB activation is known (40). Another outcome of NOD1 and NOD2 activation is the activation of the mitogen-activated protein kinase (MAPK) pathway. Thus, stimulation of wild-type macrophages, but not NOD2-deficient macrophages, with MDP leads to activation of p38MAPK and extracellular signal–regulated kinase (ERK) (41,42). In addition, activation of NOD1 by its ligand leads to the activation of JUN N-terminal kinase (JNK) (43). The mechanism of such NF-κB-independent signaling is unknown at present. Finally, it has been shown in transfection and immunoprecipitation studies that NOD2 binds procaspase-1, and when cells are transfected with constructs encoding NOD2 and procaspase-1, NOD2 induces IL-1β secretion (44). Because caspase-1 is required for processing pro-IL-1β (45) into mature IL-1β, NOD2 may bind procaspase-1 through a CARD–CARD interaction in the same way that it binds RICK and, in doing so, convert the procaspase into a caspase. However, whether NOD2 has this function under physiological conditions remains to be seen.
GENETIC NOD2 POLYMORPHISMS
A role for genetic factors in CD was first suggested by epidemiological studies showing familial aggregation of disease and by twin studies that reported greater concordance for disease in monozygotic twins compared with dizygotic twins (46). During the past 8 years, this evidence has been supplemented by molecular data from genome-wide linkage studies of multiple affected IBD families. These studies have been remarkably successful in identifying a number of susceptibility loci, with convincing replication shown for at least 7 loci (IBD1–7; Fig. 1). Some loci have been shown to be specific to ulcerative colitis (eg, IBD2) or CD (IBD1), whereas others confer common susceptibility on IBD. Collectively, these genome scans reaffirm the concept that IBDs are complex genetic disorders with several predisposing genes (47).
In 2001, 3 independent groups reported the identification of the first CD susceptibility gene, NOD2 (renamed CARD15 by the international nomenclature committee), on chromosome 16q12 (IBD1) (47). This major breakthrough firmly established a role for genetics in determining susceptibility to CD and has provided proof of principle that model-free linkage analyses may be used successfully to identify disease susceptibility gene loci. Recent studies have highlighted a number of associations between genotype and phenotype. These studies suggest that genetics also may influence the clinical manifestations of CD including disease location, behavior, natural history and response, and side affects of drug therapy (46). These discoveries may allow more accurate prediction of disease, permitting the implementation of highly specific therapy tailored to an individual's genotype.
NOD2/CARD15 has gained recent prominence through its association with increased susceptibility to CD. Thirty nonconservative mutations, also called variants or polymorphisms, associated with CD have been identified within the NOD2/CARD15 gene, but only 3 are common. Lesage et al (48) showed that the 3 common mutations—Arg702Trp, Gly908Arg, and Leu1007fsinsC—account for 82% of the mutated alleles. The relative risk to develop the disease when carrying 1 mutation is 2 to 3, but increases dramatically to 20 to 40 in case of 2 mutations (49). Approximately 40% to 50% of patients carry at least 1 mutation in the NOD2/CARD15 gene but heterogeneity has been reported, and in at least 3 populations (Japanese, Korean, and black), the gene is not implicated in CD (50–54). The prevalence of CD in western Europe is 1 to 2/1000, so it is possible to deduce from these relative risks that probability of developing the disease is 4% to 8% in the group with 2 mutations. However, the penetrance is modest: <10% of all people carrying 2 CARD15 risk alleles will develop CD, which means that other genes and environmental stimuli are needed for disease expression (55). Structure of the CARD15 gene and location of the CD-associated variants are shown in Figure 3.
The mechanism by which CARD15 mutations cause susceptibility to CD is poorly understood: 3 main views are being considered at present (Fig. 4).
Because signaling via mutated NOD2 proteins leads to defective activation of the transcription factor NF-κB, one proposal is that mutations cause deficient NF-κB–dependent Th1 responses and increased susceptibility to infection. Two recent studies used genetically altered mouse models to address this hypothesis and have come to surprisingly different conclusions.
In the first study, Meada et al (56) introduced into mice a mutation of CARD15 homologous to the major mutation in human CD (Leu007fsinsCys), which resulted in a truncated protein lacking the last 33 amino acids. This model theoretically mimics the genetic defect in CD. They examined the effect of the NOD2 mutation on NF-κB activation in bone marrow–derived macrophages after stimulation with various TRL ligands or TRL2 ligand (PGN) plus MDP. No differences between wild-type and mutated mice were seen after stimulation, except in the case of stimulation with MDP alone, where the authors found that macrophages of these mutant mice had enhanced sensitivity to MDP relative to WT counterparts with increased production of IL-1β, a target of the proinflammatory mediator NF-κB. These data, together with the finding that knock-in mice were more susceptible to dextran sulfate sodium–induced colitis and showed increased amounts of IL-1β, IL-6, and cyclooxygenase-2 protein in colons relative to wild-type counterparts, indicated that the frameshift mutation associated with CD is a gain of function that results in disease associated with IL-1β (and perhaps IL-6) production. Thus, the authors argue that individuals with NOD2 mutations may have an enhanced responsiveness to bacterial PGN, resulting in high levels of production of proinflammatory cytokines by intestinal macrophages.
This model cannot explain in vitro data showing that epithelial cells transfected with CARD15 that contains CD-associated mutations have defective NF-κB activation in response to stimulation with MDP. More important, it also cannot explain why peripheral blood mononuclear cells (PBMCs) isolated from patients with CD who have a frameshift mutation in CARD15 show a marked defect in IL-1β production rather than increased IL-1β production. Finally, these knock-in mice did not have any abnormality in the production of Th1 cytokines, which is an almost universal finding in CD (57).
To investigate this gain of function hypothesis in human, Zelinkova et al (58) used monocyte-derived DCs from patients with CD who carried double-dose NOD2 mutations and wild-type controls. Mature DCs were stimulated with MDP and the production of the proinflammatory cytokines TNF-α and IL-12 was measured. Mature DCs from NOD2 mutants showed significantly increased IL-12p70 and TNF-α production upon stimulation with MDP compared with wild-type controls. These first observations in humans support the hypothesis of NOD2 variants associates with CD acting as a gain-of-function allele supporting the regulatory role of NOD2 as the basic cause of disease.
Conversely, Kobayashi et al (42) proposed a loss-of-function mutation in NOD2 in vivo that affects mainly epithelial cells rather than macrophages. They generated NOD2−/− mice using a targeting construct to replace the NOD, which is essential for the activation of the protein. The mutant animals displayed no symptoms of intestinal inflammation when observed for as long as 6 months, and there was no significantly enhanced susceptibility to colitis in the dextran sulfate sodium model. Their mutant mice showed a lack of responsiveness to MDP in several assays. The mice also developed a more severe infection with Listeria monocytogenes when given orally versus systematically, indicating a loss of control over intestinal infection, but not an overall suppressed ability to defend against the pathogen in NOD2-deficient mice. To investigate potential genes that may be induced by NOD2 during intestinal infection, they isolated RNA samples from wild-type and NOD2−/− terminal ileum Paneth cells before and after Listeria infection and screened them by microarray analysis. The most significant difference was in the expression of a subgroup of cryptdins (analogous to α-defensins in human). Cryptidins are antimicrobial peptides that are produced in intestinal Paneth cells of mice, and their antimicrobial activity is important in suppressing infection with pathogenic bacteria. Their results indicate that NOD2 is essential in the detection of bacterial MDP and the regulation of cryptdin (α-defensins in human) expression in Paneth cells. Paneth cells are located at the (terminal) ileum, the side where patients with CD who have NOD2 gene mutations are mostly affected. These data support the concept that NOD2 mutations produce a kind of immunodeficiency state that predispose humans to a type of bacteria.
A third resolution emerged from a study by Watanabe et al (34) in which a second mouse strain was also made completely deficient in NOD2 (target deletion of exon 1); these mice had a reduced response to MDP, yet had enhanced responses to PGN inducing elevated levels of IL-12. To explain this finding, the authors invoked an additional signal delivered through TLR2, which was supported by studies with a second TRL2 ligand and purified MDP. Thus, activation of normal NOD2 inhibited signals codelivered through TLR2. A loss of function mutation of NOD2 together with TLR2 signals delivered by other bacterial products will result in enhanced cytokine responses by macrophages (or DCs) to commensal bacteria and result in inflammation. The data from Watanabe and colleagues (34) emphasize the importance of innate immunity in driving a chronic inflammatory disease. Targeting TLR2 signaling may therefore be a useful approach in the treatment of individuals with CD who carry the NOD2 mutation. Another outcome of this study is the possibility of using NOD2 ligands as anti-inflammatory agents. This may explain the anti-inflammatory effects of certain Gram-positive bacteria used in the treatment of IBD.
In addition to the results in mice, Wehkamp et al (59) suggest that the expression of α-defensins is diminished in humans with CD, particularly those who have NOD2 gene mutations. They compared mucosal levels of the human α-defensins—epithelial human defensin 5 (HD5) and epithelial human defensin 6 (HD6)—in patients with CD with respect to NOD2 genotypes and in healthy controls. They found diminished HD5 and HD6 expression in the ileum of patients carrying a NOD2 mutation compared with controls or those with nonileal CD. In addition to the NOD2−/− mouse model, they suggest that the defensin deficiency secondary to NOD2 mutations in humans could lead to a breakdown of the mucosal barrier with a secondary inflammation, leading to the development of CD.
Recent reports have shown synergy between NOD2 activation and several TLR ligands in cellular responses. van Heel et al (60) demonstrated, using primary human cells of differing NOD2 genotypes, that NOD2 stimulation normally synergistically enhances TRL9 responses (TNF-α and IL-8 secretion) and that synergy is lost in CD-associated NOD2 homozygotes, with implications for TRL-mediated intestinal homeostasis and inflammation. PBMCs were stimulated with CpG DNA (TRL9 ligand) and MDP. MDP stimulation of PBMCs from normal individuals results in a 2- to 3-fold enhancement of CpG DNA stimulation of PBMC production of TNF-α and IL-8, whereas such enhancement is not seen in PBMCs from patients with CD bearing NOD2 mutations. Concomitant studies of IL-12 secretion were not reported in this study or in a previous study of normal individuals and patients with CD with NOD2 mutations (61), perhaps because secretion of IL-12 by PBMCs is low and thus difficult to assess. It is not known whether the loss of the enhancing effect of NOD2 is counterbalanced by loss of an inhibitory effect and whether the same or similar findings would be obtained if the authors had studied intestinal cells that differ considerably from peripheral cells in response to various stimuli. In any case, based on these findings, van Heel and colleagues propose that synergistic cytokine response between TLR9 and NOD2 may be beneficial in maintaining intestinal homeostasis and that the lack of such synergism is a cause of CD (60).
SUMMARY AND CONCLUSIONS
A tightly regulated response allows the immune system to coexist with the large amount of antigenic material present in the gastrointestinal tract in the form of commensal microorganisms and food antigens while retaining the ability to respond to pathogens. However, in genetically susceptible individuals, alterations in responses to the resident luminal bacteria may lead to the development of CD, a complex multifactorial disease whose pathogenesis is still not well understood. Errors in interpretation or regulation of immune perception and responsiveness disrupt mucosal homoeostasis and predispose the individual to uncontrolled or pathological inflammation. Tissue damage in most patients with CD can be accounted for by the downstream effects of activated Th1 cells. Th1 cell differentiation takes place when T cells interact with APCs that produce proinflammatory cytokines in response to exposure to bacteria. In CD the type 1 activation is exaggerated and results in the secretion of excess proinflammatory cytokines, such as IL-1β and TNF-α. At the same time, anti-inflammatory cytokines such as IL-10 and TGF-β, which are responsible for downregulating the Th1 response, are not produced to counteract this misbalance in CD (61).
Recent studies have begun to define the mechanisms through which crucial protein recognition receptors may regulate intestinal innate immunity. Cooperative and competitive interactions may occur between different bacterial and nonbacterial ligands via TLRs and NODs leading to differential proinflammatory and anti-inflammatory immune responses in different cell types. TLRs and NOD2 are significantly involved in host defense and tissue repair responses, thus crucially maintaining mucosal homeostasis. TLR signaling protects intestinal epithelial barrier and maintains tolerance, whereas NOD2 signaling exerts antimicrobial activity and prevents bacterial invasion. Thus, both receptors collectively exhibit distinct features that ensure commensal as well as mucosal homeostasis. Imbalance of the complex interactions between commensal microorganisms and PRRs may result in tissue injury and subsequent inflammation of the intestinal mucosa. Aberrant TLR and/or NOD signaling may stimulate diverse inflammatory responses, leading to the chronic intestinal inflammation seen in CD.
The discovery of an association of NOD2 variants with CD in humans has led to intensive research in this field for the last few years. Unraveling the signaling transduction cascade of NOD2 has proven difficult, and the link to the development of CD has been demonstrated by epidemiological and linkage studies, but the exact cellular mechanism responsible for the CD phenotype in NOD2 variants needs further analysis. Wild-type NOD2 signaling by its specific ligand MDP leads to NF-κB activation and the subsequent production of proinflammatory cytokines. However, a loss-of-function mutation of NOD2 variants would lead to deficient NF-κB activation and subsequent decrease of Th1 responses. Kobayashi et al (42) and Wehkamp et al (34) showed that normal NOD2 function is responsible for the production of antibacterial peptides (ie, defensins) in Paneth cells. In NOD2 variants, deficient defensin production leads to diminished bacterial clearance at the epithelial surface and enhanced activation of macrophages or DCs within the intestine. Together these effects would heighten the cellular immune responses mediated by high levels of IL-12 and TNF-α. This theory is particularly interesting because Paneth cells are concentrated most highly in the terminal ileum, which is the most common site of inflammation in CD. Furthermore, NOD2 mutations have been consistently associated with ileal involvement in CD. Conversely, Meada et al (56) and Zelinkova et al (57) found that macrophages and dendritic cells, respectively, with NOD2 mutations, had enhanced responsiveness to MDP, releasing excess proinflammatory cytokines compared with wild-type cells. This surprising finding of a gain of function is at odds with previously generated data. Despite some reservations, it remains possible that more complex interactions between mutated NOD2 and downstream signaling complexes result in enhanced NF-κB activation in macrophages and DCs in response to MDP. Finally, Watanabe et al (34) propose that intact NOD2 signaling prevents the development of CD by controlling PGN-mediated Th1 responses via TLR2. A loss-of-function mutation of NOD2 together with TLR2 signals delivered by other bacterial products will result in enhanced cytokine responses by macrophages or DCs to commensal bacteria and result in inflammation.
Inflammation not only results from an upregulation of proinflammatory proteins by APCs or effector T cells but also is balanced by regulatory cells that secrete IL-10 or TGF-β, which are cytokines known to inhibit T cell proliferation. Netea et al (62) demonstrated that NOD2 mutations in mononuclear cells produce less IL-10 in response to various bacterial ligands compared with wild-type cells. The IL-10/TNF-α ratio showed only half of the patients bearing the WT allele, favoring mucosal inflammation. Finally, linkage studies show an association between the TLR4 polymorphism Asp299Gly and the development of CD. However, other research groups were not able to significantly reproduce these findings.
Further studies of the physiological and pathophysiological mechanisms within this network of possible cell–cell, ligand–ligand, and PRR–PRR signaling interactions that may favor or prevent CD could lead to promising, novel approaches that may differentially exploit the TLR/NOD pathways and lead to novel therapeutic strategies. It is likely that differential therapeutic strategies will need to include agonists as well as antagonists of PRRs, taking into account differences of PRR pathophysiology at different stages of disease as well as phenotypic and genotypic heterogeneity between distinct subgroups of patients with CD. Prophylactic application of selective TLR/NOD2 ligands could enhance desired commensal-mediated tissue-protective processes to prevent disease. When an acute inflammatory episode breaks out, some of the untoward effects of intestinal inflammation could be stopped by blocking uncontrolled signal transduction by specific TLR/NOD2 inhibitors, thus dampening the tissue-destructive effects. One key element to this disease-modifying approach may be to stop rather than entirely eliminate the dysregulated innate responses in CD. In this context, careful assessments of adverse effects will be critical when modulating such fundamental host defense pathways of innate immunity. Given the rapid and exciting advancements of research in this field over the last few years, it is reasonable to presume that more immunological evidence and concrete directions for the potential value of these PRRs as therapeutic targets in CD will emerge in the near future.
1. Head K, Jurenka J. Inflammatory bowel disease part II: Crohn's disease pathophysiology and conventional and alternative treatment options. Altern Med Rev 2004; 9:360–401.
2. Philpott DJ, Viala J. Towards an understanding of the role of NOD2/CARD15 in the pathogenesis of Crohn's disease. Clin Gastroenterol 2004; 18:555–568.
3. Yamada T, Alpers DH, Laine L, et al
. Textbook of Gastroenterology, Vol 1. 3rd ed Philadelphia: Lippincott Williams & Wilkins; 1999.
4. Ohkusa T, Nomura T, Sato N. The role of bacterial infection in the pathogenesis of inflammatory bowel disease. Intern Med 2004; 43:534–539.
5. MacDonald TT, Monteleone G. IL-12 and Th1 immune responses in human Peyer's pathes. Trends Immunol 2001; 22:244–247.
6. Dhillon S, Loftus EV. Medical therapy of Crohn's disease. Curr Treat Options Gastroenterol 2005; 8:19–30.
7. Duchmann R, Kaiser I, Hermann E, et al
. Tolerance exists towards resident intestinal flora but is broken in active inflammatory bowel disease (IBD). Clin Exp Immunol 1995; 102:448–455.
8. Greenberg GR. Antibiotics should be used as first-line therapy for Crohn's disease. Inflamm Bowel Dis 2004; 10:318–320.
9. Akira S, Takeda K. Toll-like receptor signaling. Nat Immunol 2004; 4:499–511.
10. Russell RK, Wilson DC, Satsangi J. Unraveling the complex genetics of inflammatory bowel disease. Arch Dis Child 2004; 89:598–603.
11. Cario E. Bacterial interactions with cells of the intestinal mucosa: toll-like receptors and NOD2. Gut 2005; 54:1182–1193.
12. Dunne A, O'Neill LA. The interleukin-1 receptor/Toll-like receptor superfamily: signal transduction during inflammation and host defense. Sci STKE 2003; 171:re3.
13. Harton JA, Linhoff MW, Zhang J, et al
. Cutting edge: CATERPILLER: a large family of mammalian genes containing CARD, pyrin, nucleotide-binding, and leucine-rich repeat domains. J Immunol 2002; 169:4088–4093.
14. Hofmann K, Bucher P, Tschopp J. The CARD domain: a new apoptotic signalling motif. Trends Biochem Sci 1997; 22:155–156.
15. Bertin J, DiStefano PS. The PYRIN domain: a novel motif found in apoptosis and inflammation proteins. Cell Death Differ 2000; 7:1273–1274.
16. Liepinsh E, Barbals R, Dahl E, et al
. The death-domain fold of the ASC PYRIN domain, presenting a basis for PYRIN/PYRIN recognition. J Mol Biol 2003; 332:1155–1163.
17. Steimle V, Otten LA, Zufferey M, et al
. Complementation cloning of an MHC class II transactivator mutated in hereditary MHC class II deficiency (or bare lymphocyte syndrome). Cell 1993; 75:135–146.
18. Roy N, Mahadevan MS, McLean M, et al
. The gene for neuronal apoptosis inhibitory protein is partially deleted in individuals with spinal muscular atrophy. Cell 1995; 80:167–178.
19. Reith W, Mach B. The bare lymphocyte syndrome and the regulation of MHC expression. Annu Rev Immunol 2001; 19:331–373.
20. Ogura Y, Bonen DK, Inohara N, et al
. A frameshift mutation in NOD2 associated with susceptibility to Crohn's disease. Nature 2001; 411:603–606.
21. Chamaillard M, Hashimoto M, Horie Y, et al
. An essential role for NOD1 in host recognition of bacterial peptidoglycan containing diaminopimelic acid. Nat Immunol 2003; 4:702–707.
22. Girardin S, Boneca I, Carneiro LAM, et al
. Nod1 detects a unique muropeptide from Gram-negative bacterial peptidoglycan. Science 2003; 300:1584–1587.
23. Inohara N, Ogura Y, Fontalba A, et al
. Host recognition of bacterial muramyl dipeptide mediated through NOD2. Implications for Crohn's disease. J Biol Chem 2003; 278:5509–5512.
24. Girardin SE, Boneca I, Viala J, et al
. Nod2 is a general sensor of peptidoglycan through muramyl dipeptide (MDP) detection. J Biol Chem 2003; 278:8869–8872.
25. Inohara N, Chamaillard M, McDonald C. NOD-LRR proteins: role in host–microbial interactions and inflammatory disease. Annu Rev Biochem 2004; 74:355–383.
26. Pauleau AL, Murray PJ. Role of Nod2 in the response of macrophages to Toll-like receptor agonists. Mol Cell Biol 2003; 23:7531–7539.
27. Gupta DK, Theisen N, von Figura K, et al
. Comparison of biosynthesis and subcellular distribution of lysozyme and lysosomal enzymes in U937 monocytes. Biochim Biophys Acta 1985; 847:217–222.
28. Araki Y, Nakatani T, Makino R, et al
. Isolation of glucosaminyl-α (1–4)-muramic acid and phosphoric acid ester of this disaccharide from acid hydrolysates of peptidoglycan of Bacillus cereus
AHU 1356 cell walls. Biochem Biophys Res Commun 1971; 42:684–690.
29. Vavricka SR, Musch MW, Chang JE, et al
. hPepT1 transports muramyl dipeptide, activating NF-κB and stimulating IL-8 secretion in human colonic Caco2/bbe cells. Gastroenterology 2004; 127:1401–1409.
30. Viala J, Chaput C, Boneca IG, et al
. Nod1 responds to peptidoglycan delivered by the Helicobacter pylori
cag pathogenicity island. Nat Immunol 2004; 5:1166–1174.
31. Bell JK, Mullen GE, Leifer CA, et al
. Leucine-rich repeats and pathogen recognition in Toll-like receptors. Trends Immunol 2003; 24:528–533.
32. Inohara N, Koseki T, Lin J, et al
. An induced proximity model for NF-κB activation in the Nod1/RICK and RIP signalling pathways. J Biol Chem 2000; 275:27823–27831.
33. Tanabe T, Chamaillard M, Ogura Y, et al
. Regulatory regions and critical residues of NOD2 involved in muramyl dipeptide recognition. EMBO J 2004; 23:1587–1597.
34. Watanabe T, Kitani A, Murray PJ, et al
. W. NOD2 is a negative regulator of Toll-like receptor 2- mediated T helper type 1 responses. Nat Immunol 2004; 5:800–808.
35. Kobayashi K, Inohara K, Hernandez LD, et al
. RICK/Rip2/CARDIAK mediates signalling for receptors of the innate and adaptive immune systems. Nature 2002; 416:194–199.
36. Ogura Y, Inohara N, Benito A, et al
. Nod2, a Nod1/Apaf-1 family member that is restricted to monocytes and activates NF-kappaB. J Biol Chem 2001; 276:4812–4818.
37. Inohara N, Koseki T, del Peso L, et al
. Nod1, an Apaf-1-like activator of caspase-9 and nuclear factor-kappaB. J Biol Chem 1999; 274:14560–14567.
38. Abbott DW, Wilkins A, Asara JM, et al
. The Crohn's disease protein, NOD2, requires RIP2 in order to induce ubiquitinylation of a novel site on NEMO. Curr Biol 2004; 14:2217–2227.
39. Zhou H, Wertz I, O'Rourke K, et al
. Bcl10 activates the NF-κB pathway through ubiquitination of NEMO. Nature 2004; 427:167–171.
40. Barnich N, Hisamatsu T, Aguirre JE, et al
. GRIM-19 interacts with nucleotide oligomerization domain 2 and serves as downstream effector of anti-bacterial function in intestinal epithelial cells. J Biol Chem 2005; 280:19021–19026.
41. Pauleau AL, Murray PJ. Role of Nod2 in the response of macrophages to Toll-like receptor agonists. Mol Cell Biol 2003; 23:7531–7539.
42. Kobayashi KS, Chamaillard M, Ogura Y, et al
. Nod2-dependent regulation of innate and adaptive immunity in the intestinal tract. Science 2005; 307:731–734.
43. Girardin SE, Tournebize R, Mavris M, et al
. CARD4/Nod1 mediates NF-κB and JNK activation by invasive Shigella flexneri
. EMBO Rep 2001; 2:736–742.
44. Damiano JS, Oliveira V, Welsh K, et al
. Heterotypic interactions among NACHT domains: implications for regulation of innate immune responses. Biochem J 2004; 381:213–219.
45. Martinon F, Tschopp J. Inflammatory caspases: linking an intracellular innate immune system to autoinflammatory diseases. Cell 2004; 117:561–574.
46. Hugot JP, Chamaillard M, Zouali H, et al
. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn's disease. Nature 2001; 411:599–603.
47. Ahmad T, Tamboli CP, Jewel D, et al
. Clinical relevance of advances in genetics and pharmacogenetics of IBD. Gastroenterology 2004; 126:1533–1549.
48. Lesage S, Zouali H, Cezard JP, et al
. NOD2/CARD15 mutational analysis and genotype- phenotype correlation in 612 patients with inflammatory bowel disease. Am J Hum Genet 2002; 70:845–857.
49. Economou M, Trikalinos TA, Loizou KT, et al
. Differential effects of NOD2 variants on Crohn's disease risk and phenotype in diverse populations: a metaanalysis. Am J Gastroenterol 2001; 99:2393–2404.
50. Sugimura M, Kinouchi Y, Takahashi S, et al
. NOD2/CARD15 mutational analysis in Japanese patients with Crohn's disease. Clin Genet 2003; 63:160–162.
51. Inoue N, Tamura K, Kinoudhi Y, et al
. Lack of common NOD2 variants in Japanese patients with Crohn's disease. Gastroenterology 2003; 123:86–91.
52. Leong RWL, Armuzzi A, Ahmad T, et al
. NOD2/CARD15 gene polymorphism and Crohn's disease in the Chinese population. Aliment Pharmacol Ther 2003; 17:1465–1470.
53. Bonen DK, Niclolae DL, Moran T, et al
. Racial differences in NOD2 variation: characterization of NOD2 in African-Americans with Crohn's disease. Gastroenterology 2002; 122(Suppl):A-29.
54. Croucher PJ, Mascheretti S, Hampe J, et al
. Haplotype structure and association to Crohn's disease of CARD15 mutations in two ethnically divergent population. Eur J Hum Genet 2003; 11:6–16.
55. Vermeire S. NOD2/CARD15: relevance in clinical practice. Best Pract Res Clin Gastroenterol 2004; 18:569–575.
56. Maeda S, Hsu L, Liu H, et al
. Nod 2 mutation in Crohn's disease potentiates NF-κB activity and IL-1β processing. Science 2005; 307:734–738.
57. Strober W, Murray PJ, Kitani A, et al
. Signalling pathways and molecular interactions of NOD1 and NOD2. Nat Rev Immunol 2006;(6):9–20.
58. Zelinkova Z, de Kort F, Pronk I, et al
. Functional consequences of NOD2 deficiency in Crohn's disease patients peripheral blood monocytes derived dendritic cells. Gastroenterology 2005; 128:A510.
59. Wehkamp J, Harder J, Weichenthal M, et al
. NOD2 (CARD15) mutations in Crohn's disease are associated with diminished mucosal0020-defensin expression. Gut 2004; 53:1658–1664.
60. van Heel DA, Ghosh S, Hunt KA, et al
. Synergy between TLR9 and NOD2 innate immune responses is lost in genetic Crohn's disease. Gut 2005; 54:1553–1557.
61. Netea MG, Ferwerda G, de Jong DJ, et al
. Nucleotide-binding oligomerization domain-2 modulates specific TLR pathways for the induction of cytokine release. J Immunol 2005; 174:6518–6523.
62. Netea MG, Kullberg BJ, de Jong DJ, et al
. NOD2 mediates anti-inflammatory signals induced by TLR2 ligands: implications for Crohn's disease. Eur J Imuunol 2004; 34:2052–2059.