NOD1 and NOD2: Signaling, Host Defense, and Inflammatory Disease

Summary

The nucleotide-binding oligomerization domain (NOD) proteins, NOD1 and NOD2, the founding members of the intracellular NOD-like receptor family, sense conserved motifs in bacterial peptidoglycan and induce pro-inflammatory and anti-microbial responses. Here we discuss recent developments about the mechanisms by which NOD1 and NOD2 are activated by bacterial ligands, the regulation of their signaling pathways, and their role in host defense and inflammatory disease. Several routes for the entry of peptidoglycan ligands to the host cytosol to trigger activation of NOD1 and NOD2 have been elucidated. Furthermore, genetic screens and biochemical analyses have revealed mechanisms that regulate NOD1 and NOD2 signaling. Finally, recent studies suggest several mechanisms to account for the link between NOD2 mutations and susceptibility to Crohn’s disease. Further understanding of NOD1 and NOD2 should provide new insight into the pathogenesis of disease and the development of new strategies to treat inflammatory and infectious disorders.

Introduction

The innate immune system carries out important functions including the recognition and clearance of infectious organisms through the detection of microorganisms by germline encoded pathogen recognition receptors (PRRs). The nucleotide-binding oligomerization domain (NOD) proteins, NOD1 and NOD2, represent two well characterized PRRs of the NOD-like receptor (NLR) family that sense conserved fragments found in the cell wall of many types of bacteria and activate intracellular signaling pathways that drive pro-inflammatory and anti-microbial responses. In contrast to PRRs such as Toll-like receptors (TLRs) that recognize microbial ligands at the cell surface or within endosomes, NOD1 and NOD2 sense bacterial products in the host cytosol providing another level of microbial surveillance that is often associated with pathogen invasion. Here, we focus on recent discoveries of the molecular mechanisms that regulate NOD1 and NOD2 activation and signaling, the disruption of these pathways in inflammatory diseases, and efforts to develop small molecules capable of modulating these pathways for therapeutic applications. Specifically, we highlight recent knowledge about the mechanism of recognition and activation of NOD receptors, regulation of NOD1 and NOD2 signaling pathways, and new evidence linking NOD2 to the development of Crohn’s disease (CD).

NOD1 and NOD2: ligand recognition and activation

Sequence homology searches identified NOD1 as the first NLR member (Bertin et al., 1999; Inohara et al., 1999). NOD1 encodes an intracellular multi-domain scaffolding protein consisting of a caspase activation and recruitment domain (CARD) domain, a nucleotide-binding oligomerization domain (NOD), and multiple leucine rich repeats (LRRs). This led to the discovery of NOD2, a closely related protein with an additional CARD domain (Ogura et al., 2001b). Early work showed that NOD1 and NOD2 mediate activation of the nuclear factor-kappa B (NF-κB) family of transcriptional regulators in response to distinct peptidoglycan (PGN) fragments (Inohara et al., 2001). Subsequent studies demonstrated that NOD1 can be activated by γ-d-glutamyl-meso-diaminopimelic acid (iE-DAP), a motif present in many Gram^−ve bacteria and certain Gram^+ve bacteria (Chamaillard et al., 2003a; Girardin et al., 2003a; Hasegawa et al., 2006). Additional work revealed that muramyl dipeptide (MDP), a PGN motif widely distributed among both Gram^+ve and Gram^−ve bacteria, is sufficient to trigger NOD2 activity (Girardin et al., 2003b; Inohara et al., 2003). Ultimately NOD1 and NOD2 signaling contributes to host defense via the production of pro-inflammatory cytokines and anti-microbial molecules (Kobayashi et al., 2005; Masumoto et al., 2006).

The mechanisms by which bacterial PGN enters cells and activate NOD1 and NOD2 receptors remain poorly understood but multiple routes of entry have been reported (Figure 1). Two recent studies have revealed a role for endosomes in the activation of NOD1 and NOD2 signaling (Irving et al., 2014; Nakamura et al., 2014). In one study, two peptide transporters, SLC15A3 and SLC15A4, were shown to transport MDP across endosome membranes of phagosomes that have internalized bacteria via phagocytosis and that NOD1 and NOD2 localize to these membranes (Nakamura et al., 2014). This study is consistent with previous reports showing decreased NOD1-mediated cytokine production in Slc15A4^−/− mice (Sasawatari et al., 2011) and reduced NF-κB activation after NOD1 stimulation upon SLC15A4 down-regulation (Lee et al., 2009). Notably, SLC15A3 is present near a susceptibility loci associated with CD by GWAS (Jostins et al., 2012). While the subcellular localization and expression of NOD1 and NOD2 vary depending on the cell type, the source and route of entry of the ligand also influence NOD1 and NOD2 complex formation. Indeed, outer membrane vesicles (OMVs) from a variety of bacteria (Irving et al., 2014; Thay et al., 2014) activate NOD1 and NOD2 signaling revealing an important role for vesicle internalization in the delivery of ligands to cytosolic NOD1 and NOD2 receptors. Given that soluble NOD1 and NOD2 ligands are sufficient to stimulate host cells there are clearly multiple routes for these molecules to enter the cell.

Figure 1. Potential mechanisms for bacterial recognition by NOD1 and NOD2.

NOD1 and NOD2 sense intracellular PGN fragments from bacteria. Host cells can internalize PGN by multiple routes such as phagocytosis of bacteria and subsequent bacterial degradation (1), uptake of PGN fragments from bacteria derived extracellular outer membrane vesicles (2), transport across host membranes via channels, pore-forming molecules or bacterial secretion systems (3), endocytosis (4), or from neighboring cells (5). Once inside the cell, NOD1 and NOD2 activation typically involves their re-localization to various intracellular locations such as the plasma and endosomal membranes via different adaptor molecules that are differentially expressed in host cells. For example, the activation of NOD2 by intracellular pathogens induces the formation of an autophagosome, which is mediated by ATG16L1.

While the role of NOD1 and NOD2 in sensing bacterial PGN fragments has been reported for over a decade, it was not clear whether these receptors respond directly to bacterially-derived molecules or indirectly through some other intermediate. Recent experimental evidence supports a direct interaction between bacterial l-Ala-γ-d-Glu-mesoDAP (Tri-DAP) and NOD1 using surface plasmon resonance (SPR) and atomic force microscopy (Laroui et al., 2011) and between MDP and NOD2 by SPR (Grimes et al., 2012) and biotin-labeled MDP pull-downs (Mo et al., 2012). While these studies agree on a direct interaction between receptor and ligand, the SPR data suggest a role for the LRRs of both NOD1 and NOD2 in mediating the interaction while the pull-down experiments implicate the NOD region. Notably, SPR analysis between NOD2 and MDP revealed a relatively high affinity interaction (K_d ~ 50 nM), while the NOD1-Tri-DAP interaction was significantly weaker (K_d ~ 30 µM), which may reflect differences in the ligand used or in assay conditions. A direct interaction between labeled PGN fragments and NOD1 is also supported by fluorescence energy transfer (Irving et al., 2014). Together, these studies support a direct interaction between NOD1 and NOD2 and their bacterial ligands but they do not rule out a role for accessory molecules that may facilitate ligand recognition similar to the function of MD-2 as a TTLR4 co-receptor for bacterial LPS (Park et al., 2009). Further studies to define the molecular details of these interactions will require structural data.

Existing structural analyses of the NLR family member NLRC4 and Apaf-1, an NLR-related protein that regulates apoptosis, have provided insight into potential mechanisms of NOD1 and NOD2 activation (Hu et al., 2013; Park et al., 2009). NOD1 and NOD2 share a NOD module composed of nucleotide-binding domain (NBD), winged-helix (WH), and helix domains (HD1 and HD2). The ADP-mediated interaction between the NBD and WH is important for stabilizing the closed conformation while the LRRs occlude these domains rendering NLR proteins in a monomeric state (Hu et al., 2013). Upon ligand binding to the LRRs, the HD2 mediates conformational changes of the NBD-WH-HD1 domains allowing ADP/ATP exchange, self-oligomerization, and downstream signaling (Lechtenberg et al., 2014; Riedl et al., 2005). Consistently, HD2 mutations in NOD2 result in either gain- or loss-of-function (Hu et al., 2013; Tanabe et al., 2004). Furthermore, similar gain- and loss-of-function NOD2 mutations are associated with early onset sarcoidosis (EOS)/Blau syndrome and CD, respectively (Caso et al., 2014; Franchi et al., 2009). Complementation between NOD2 mutations in HD2 and LRRs further supports this mechanism of NOD2 activation (Tanabe et al., 2004). Collectively, available evidence suggests that NOD1 and NOD2 reside in an auto-inhibited monomeric state in the cytosol and upon ligand recognition, they undergo conformational changes that promote their activation. However, further studies are needed including structural data to understand the regulation of NOD1 and NOD2 activation.

NOD1 and NOD2 signaling and regulation

Once the conformation of NOD1 and NOD2 is open, the proteins self-oligomerize and recruit receptor-interacting serine/threonine-protein kinase 2 (RIPK2) through homotypic CARD-CARD interactions, resulting in close proximity of RIPK2- IκB kinase (IKK) complexes (Inohara et al., 1999). The RIPK2 Ser-Thr kinase then mediates the recruitment and activation of the TAK1 Ser-Thr kinase, which is a prerequisite for activation of the IKK complex and MAPK pathway (Figure 2). IKK-mediated phosphorylation of the NF-κB inhibitor IκBα leads to its polyubiquitination (pUb) and subsequent degradation through the proteasome allowing NF-κB to translocate to the nucleus and influence the expression of downstream target genes. The post-translation modification of proteins with Ubiquitin (Ub) affects many steps in the NF-κB pathway. For example, NOD1 and NOD2 signaling is regulated via Lys63-linked pUb of RIPK2 on Lys209, which is necessary for the recruitment of the TAK1 complex (Hasegawa et al., 2008; Ogura et al., 2001b). Multiple groups have identified several different E3 ligases capable of binding and catalyzing Lys48-linked RIPK2 pUb including cellular inhibitor of apoptosis 1 (cIAP1) and cIAP2 (Bertrand et al., 2009), ITCH (Tao et al., 2009), and Pellino3 (Yang et al., 2013), suggesting a role for these proteins in regulating NOD1 and NOD2 signaling. Notably, like NOD2, mutations in X-linked inhibitor of apoptosis protein (XIAP), an E3 ligase that binds and polyubiquitinates RIPK2, are associated with CD (Zeissig et al., 2014) and impaired NOD1 and NOD2 signaling (Damgaard et al., 2013). Additional E3 ligases that influence NOD1 and NOD2 signaling have been identified including TNF receptor associated factor 2 (TRAF2) and TRAF5 (Hasegawa et al., 2008), TRAF4 (Marinis et al., 2011), and RNF34 (Zhang et al., 2014). The linear ubiquitin chain assembly complex (LUBAC) complex, which consists of the E3 ligase RNF31 (also known as HOIP) and directs the linear pUb of NF-kappa-B essential modulator (NEMO), also plays a positive role in NOD2-induced NF-κB signaling (Damgaard et al., 2012).

Figure 2. NOD2 signaling pathways for gene activation.

NOD2 interacts directly with intracellular bacterial PGN fragments containing the MDP motif. Ligand recognition relieves intra-molecular autoinhibitory interactions leading to NOD oligomerization. Recruitment of the downstream RIPK2 Ser/Thr kinase occurs through CARD-CARD domain interactions. Subsequent activation of the NF-κB and MAPK pathways results in the transcriptional up-regulation of pro-inflammatory and host defense genes. Multiple steps in the pathway are regulated either positively or negatively by post-translational modifications such as phosphorylation and pUb events. Multiple regulatory genes act to influence various steps in the pathway often in a cell-type dependent manner.

The removal of Ub by proteases can further fine-tune NOD1 and NOD2 signaling. For instance, A20 was the first Ub protease identified to negatively regulate NOD2 signaling (Hitotsumatsu et al., 2008). More recently, the deubiquitinase OTULIN was shown to dampen NOD2 signaling (Fiil et al., 2013) while knockdown of the Ub-specific protease USP8 and several of it interacting partners also negatively regulate NOD2-induced interleukin-8 (CXCL8) production (Warner et al., 2014). Other roles for Ub in the regulation of NOD1 and NOD2 signaling have been revealed. For example, NOD2 protein is destabilized by Lys48-linked pUb and subsequent degradation via the E3 ligase TRIM27 (Lee et al., 2012; Zurek et al., 2012). This may partially account for the phenomenon of MDP tolerance where MDP pretreatment significantly decreases subsequent NOD2 activation upon restimulation with MDP in both human (Hedl et al., 2007) and mouse (Kim et al., 2008a) cells. Similarly, direct pUb of NOD1 has been reported (Zhang et al., 2014). Structural details of the interaction between the CARD domain of NOD1 and Ub have also been resolved using NMR (Ver Heul et al., 2013) and X-ray crystallography (Ver Heul et al., 2014). Interestingly, mutations in the CARD domain of NOD2 analogous to the Ub interacting interface of NOD1 result in loss of Ub binding and increased NOD2-induced CXCL8 secretion.

NOD1 and NOD2 function in pathogen recognition and immunity

Consistent with a role for NOD1 and NOD2 in host responses against bacterial infection, Nod1^−/− and Nod2^−/− mice show enhanced susceptibility to several pathogens (Franchi et al., 2009; Philpott et al., 2014). Because bacteria can be sensed by multiple PRRs, it is not surprising that in most cases, deficiency of NOD1 and/or NOD2 have only modest effects on pathogen clearance in vivo (Philpott et al., 2014). Consistently, NOD1 and NOD2 play redundant roles with TLRs in the detection of bacteria and production of pro-inflammatory molecules, as both signaling pathways lead to NF-κB and MAPK activation (Park et al., 2007; Tada et al., 2005). A cooperative role for TLR4 and NOD1 was also revealed in the mobilization of mature hematopoietic stem cells to the spleen upon Escherichia coli infection (Burberry et al., 2014). Furthermore, the activity of NOD1 and NOD2 becomes important when TLR signaling is absent or reduced in vivo (Kim et al., 2008b). In line with these observations, TLR ligand-insensitive epithelial cells still respond to NOD1 ligands and NOD1-stimulatory bacteria (Kim et al., 2004). Additionally, NOD1 and NOD2 signaling is enhanced by pretreatment with TLR ligands, such as LPS or viral infection through a type I IFN-dependent mechanism both in vitro and in vivo (Kim et al., 2014).

NOD1 is widely expressed by a variety of cell types such as epithelial cells, stromal cells, and endothelial cells (Inohara et al., 1999; Park et al., 2007). Stimulation of intestinal epithelial cells with NOD1-activating molecules induces the production of chemokines and the recruitment of acute inflammatory cells in vivo (Masumoto et al., 2006). NOD1 contributes to the activation of immune responses in epithelial cells infected with several Gram^−ve bacteria including Shigella flexneri (Girardin et al., 2003a; Kim et al., 2010) and Helicobacter pylori (Allison et al., 2009). In both infection models, either intracellular delivery of S. flexneri, LPS-containing NOD1 ligand or injection of H pylori PGN fragments via a type IV secretion system, promotes NOD1-dependent activation of NF-κB in epithelial cells (Allison et al., 2009; Girardin et al., 2001). H. pylori can also induce the synthesis of type I interferon via NOD1 (Watanabe et al., 2010), and consequently, Nod1^−/− mice are more susceptible to H. pylori infection demonstrating the importance of NOD1 in pathogen recognition in vivo (Viala et al., 2004).

NOD1 is also involved in the sensing of enteroinvasive E. coli (Kim et al., 2004), Pseudomonas aeruginosa (Travassos et al., 2005), and Campylobacter jejuni (Zilbauer et al., 2007), Pasteurellaceae NI1060 (Jiao et al., 2013) and Gram^+ve Clostridium difficile (Hasegawa et al., 2011). Consistently, Nod1^−/− mice are more susceptible to C. difficile infection, which was associated with impaired C. difficile clearance, increased commensal translocation, and defective recruitment of neutrophils to the infected site (Hasegawa et al., 2011). Similarly, NOD1 is important for neutrophil recruitment in response to accumulation of the pathobiont NI1060 at damaged gingival sites, which results in mouse periodontitis (Jiao et al., 2013). Furthermore, Nod1^−/− mice show increased susceptibility to lung infection with Legionella pneumophila which was associated with reduced neutrophil recruitment to the lungs (Frutuoso et al., 2010).

In contrast to NOD1, the expression of NOD2 is limited to certain cell types such as hematopoietic cells (Ogura et al., 2003). In the intestine, NOD2 is expressed in Paneth cells and stem cells (Barnich et al., 2005; Nigro et al., 2014), where it senses many types of PGN which vary in their level of NOD2-stimulatory activity across bacterial species (Hasegawa et al., 2006), and activates NF-κB and MAPK signaling in the context of S. pneumoniae (Opitz et al., 2004) and E. coli (Theivanthiran et al., 2012) infections. NOD2 is also involved in sensing intracellular L. monocytogenes (Kobayashi et al., 2005), Salmonella typhimurium (Hisamatsu et al., 2003), S. flexneri (Kufer et al., 2006), and Mycobacterium tuberculosis (Ferwerda et al., 2005). Several studies have reported a role for NOD2 in host defense in vivo. For instance, S. pneumoniae recognition by NOD2 induced the production of CC-chemokine ligand 2 (CCL2), leading to the recruitment of inflammatory macrophages that are necessary for bacteria clearance in the lung (Davis et al., 2011). Similarly, NOD2 promotes the clearance of Citrobacter rodentium by the induction of CCL2 and Th1 immune responses in the intestine (Kim et al., 2011a) and of Staphylococcus aureus in the skin (Hruz et al., 2009).

Given the role of both NOD1 and NOD2 in the innate immune response against bacterial pathogens, it is not surprising that pathogens have evolved mechanisms to evade detection and clearance by the host. Indeed, L. monocytogenes can escape NOD1-mediated detection by modifying its PGN via N-deacetylation (Boneca et al., 2007). A similar evasion mechanism, through the modification of cell wall PGN, has also been described for H. pylori to avoid NOD1-mediated detection (Chaput et al., 2006).

In addition to the role of NOD proteins in innate immune responses to bacterial infections, there is mounting evidence that NOD1 and NOD2 signaling influences adaptive immune responses. PGN derivatives have been identified as adjuvants of antigen-specific IgG production (Ellouz et al., 1974). As expected, the adjuvant activity of MDP, including antigen-specific IgG responses, was mediated by NOD2 in vivo (Girardin et al., 2003b; Inohara et al., 2003; Kobayashi et al., 2005). NOD2 also regulates Th17 cell responses (Brain et al., 2013; van Beelen et al., 2007). In mice, stimulation with either NOD1 or NOD2 alone leads to predominantly Th2-dependent adaptive immune responses, whereas co-stimulation with TLR agonists promotes the priming of Th1, Th2, as well as Th17 cell immune responses (Fritz et al., 2007; Magalhaes et al., 2008). Furthermore, NOD1 and NOD2 stimulation induces OX40 ligand, which is important for Th2-oriented acquired immunity (Duan et al., 2010; Magalhaes et al., 2011). Both NOD1 and NOD2 contribute to IL-6-dependent induction of mucosal Th17 responses during early stages of intestinal infection with C. rodentium and S. typhimurium (Geddes et al., 2011). Notably, radiation-resistant non-hematopoietic cells play a role in triggering NOD1-mediated Th2 immune responses (Fritz et al., 2007), although the mechanisms involved remains unknown. Together these studies provide evidence for a role of NOD1 and NOD2 signaling in regulating adaptive immune responses.

NOD2 is required for the “training” of monocytes induced by tuberculosis vaccination Indeed, monocytes can be functionally reprogrammed to exhibit enhanced and lasting protective functions not only against tuberculosis, but also against secondary non-mycobacterial challenges, through a NOD2-dependent mechanism (Kleinnijenhuis et al., 2012). Overall, this work suggests a role for NOD2 in modulating the adaptive features of innate immunity.

NOD2 function and CD

Polymorphisms in NOD2 are the strongest known genetic risk factors in the development of CD (Hugot et al., 2001; Ogura et al., 2001a). Three common NOD2 variants (R702W, G908R, and L1007insC) and multiple minor variants in the C-terminal LRR region and the HD2 are linked to the development of CD (Chamaillard et al., 2003b). Individuals with just one NOD2 variant are only mildly increased in their risk of developing CD while those homozygous or compound heterozygous for NOD2 variants exhibit a 20- to 40-fold increased risk (Hugot et al., 2007). However, the majority of individuals homozygous for NOD2 mutations do not develop CD, indicating that environmental factors, altered immune regulation, and/or the gut microbiota are critical for disease development. Consistently, no spontaneous intestinal inflammation occurs in Nod2^−/− mice or knock-in mice homozygous for the CD-associated L1007insC NOD2 variant, housed in SPF facilities (Kim et al., 2011b; Kobayashi et al., 2005; Pauleau and Murray, 2003).

The mechanisms by which NOD2 mutations contribute to CD pathogenesis remain unclear, but several hypotheses have been proposed (Figure 3). The first suggests that variants of NOD2 lead to impaired NOD2 activation in response to MDP stimulation, resulting in a “loss-of-function” phenotype (Inohara et al., 2003). Consistently, induction of inflammatory cytokines after MDP stimulation is impaired in monocytes homozygous for CD-associated NOD2 mutations (Inohara et al., 2003; van Heel et al., 2005). Overall, this hypothesis suggests that NOD2 mutations impair bacterial recognition and clearance, which, in turn, lead to aberrant inflammation through NOD2-independent pathways. Because monocytes and neutrophils express NOD2, defects in bacterial recognition could involve phagocytic cells recruited to intestinal sites (Ogura et al., 2001b). Furthermore, some intestinal bacteria rely on NOD2 recognition for activation of innate immune responses, which is impaired in macrophages harboring CD-associated NOD2 mutations (Kim et al., 2011b). Paneth cells located in the crypts of the small intestine that secrete anti-bacterial proteins and peptides such as α-defensins express NOD2 (Lala et al., 2003; Ogura et al., 2003). Notably CD-associated NOD2 mutations confer susceptibility to the development of ileal, but not colonic lesions, corresponding to the location of Paneth cells (Gasche and Grundtner, 2005). Furthermore, altered expression of α-defensins was observed in patients with ileal CD (Wehkamp et al., 2005). However, whether the reduced expression of α-defensins is a primary pathogenic event or an epiphenomenon of inflammation reflecting Paneth cells loss due to mucosal damage remains controversial (Simms et al., 2008). In mice, NOD2 regulates the expression of a subgroup of Paneth cell-expressed α-defensins (Kobayashi et al., 2005) and the accumulation of both commensal and pathogenic bacteria in the terminal ileum (Petnicki-Ocwieja et al., 2009). However, other studies have shown no difference in the expression of anti-microbial molecules between Nod2^−/− and WT mice (Robertson et al., 2013; Shanahan et al., 2014). So, it is still controversial whether CD-related inflammation is triggered by impaired anti-microbial activity within the intestinal crypts and abnormal accumulation of pathogenic bacteria and/or pathobionts.

Figure 3. NOD2 and Crohn’s disease.

NOD2 plays a crucial role in regulating intestinal homeostasis. By sensing microbiota-derived PGN fragments, NOD2 activates NF-κB, which, in turn, leads to the production of antimicrobial peptides (AMPs) in Paneth cells (PC) that provide a barrier between the microorganisms and the epithelial layer. Activation of NOD2 in dendritic cells (DCs) leads to production of the interleukin-23 (IL-23) thus promoting an early mucosal T helper 17 (Th17) cell response that enhances barrier protection by inducing IL-22 and regenerating islet-derived protein IIIγ (REGIIIγ). NOD2 activation in stromal cells also promotes CC-ligand 2 (CCL2)-mediated recruitment of inflammatory monocytes (Mo) to the intestine. Interaction between NOD2 and ATG16L1 promotes autophagosome formation in intestinal epithelial cells (IEC) and intraepithelial bacterial clearance. Crohn’s disease-associated NOD2 variants perturb many aspects of immune homeostasis including reduced MDP sensing in both macrophages (Mac) and DCs, impaired anti-microbial responses in Paneth cells, and altered autophagy leading to defective barrier function and/or bacterial clearance. These alterations along with the development of dysbiosis may lead to enhanced mucosal adherence and translocation of bacteria. MDP, muramyl-dipeptide, GC, globet cells, IELs, intraepithelial lymphocytes, N, neutrophils, Th1, T helper 1.

Initial work showed that Nod2^−/− mice have an altered composition of the gut microbiota with a net increase in the abundance of Bacteroidetes and Firmicutes phyla in the feces and terminal ileum compared to WT mice (Petnicki-Ocwieja et al., 2009; Rehman et al., 2011). However, other studies did not identify significant differences in the composition of the gut microbial community in Nod2^−/− mice (Robertson et al., 2013; Shanahan et al., 2014), raising the possibility that the imbalance in bacterial communities found in Nod2^−/− mice may be due to maternal transmission of the microbiota and/or other environmental factors independent of genotype. However, the controversy of whether NOD2 deficiency promotes dysbiosis requires further study given that small intestinal abnormalities have been reported in Nod2^−/− mice including goblet cell alterations, which are associated with an increased number of interferon-γ-producing intraepithelial lymphocytes (Ramanan et al., 2014). Notably, these abnormalities correlate with the selective expansion of the commensal Bacteroides vulgatus, which can exacerbate intestinal inflammation in Nod2^−/− mice treated with the NSAID piroxicam. However, the biological relevance of these intestinal alterations observed in Nod2^−/− mice as well as the relative abundance of B. vulgatus in patients with CD need to be further investigated.

Another hypothesis that links NOD2 mutations to the development of CD involves the recently recognized role of NOD2 in the autophagy pathway. NOD2 interacts with and recruits the CD-associated autophagy protein ATG16L1 to the plasma membrane at bacterial entry sites (Travassos et al., 2010). Furthermore, NOD2 stimulation promotes autophagosome formation in host cells which is associated with either an increase in killing of S. typhimurium or antigen presentation (Cooney et al., 2010; Homer et al., 2010). Notably, CD-associated NOD2 variants are defective in ATG16L1 recruitment and exhibit impaired bacteria-induced autophagy in a cell-type specific manner (Homer et al., 2010). Furthermore, patients with ileal CD are impaired in the secretion of Paneth cell-derived α-defensins, a defect that may arise from mutations in NOD2 or autophagy genes (Cadwell et al., 2008; Wehkamp et al., 2004; Wehkamp et al., 2005). Loss of function of Atg16L1 was also associated with Paneth cell abnormalities in mice, although pathology associated with Paneth cell function relied on prior exposure to mouse norovirus (Cadwell et al., 2008; Cadwell et al., 2010). However, no defects in autophagy have been reported in Nod2^−/− mice in vivo. Therefore, further work is needed to ascertain the functional interplay between CD-associated NOD2 and ATG16L1.

Finally, another hypothesis suggests that NOD2 negatively regulates TLR signaling and that NOD2 mutations result in deregulated TLR signaling and enhanced inflammation. Consistently, TLR-induced IL-12 production was increased in macrophages and DCs deficient in Nod2 (Watanabe et al., 2006). This model proposes that NOD2 acts as a brake to dampen immune responses and that NOD2 mutations lead to a deregulated TLR-mediated Th1 response in the intestine. However, the bulk of studies failed to reveal an increase in the production of cytokines, such as IL-12, in macrophages or human DCs deficient in NOD2 after TLR stimulation (Kim et al., 2008b; Kramer et al., 2006; Park et al., 2007; Salucci et al., 2008). Furthermore, NOD2 positively regulates the production of IL-10 in response to microbial stimulation and CD-associated NOD2 variants inhibit MDP-induced IL-10 transcription in human monocytes (Noguchi et al., 2009; Wagener et al., 2014). The latter likely reflects, in part, the overall impairment of immune responses to MDP in cells homozygous for CD-associated NOD2 mutations.

Early evidence showed that macrophages isolated from knock-in mice expressing the CD-associated L1007insC NOD2 variant exhibit increased NF-κB activation and IL-1β secretion after MDP stimulation (Maeda et al., 2005). However, the interpretation of the latter studies is difficult because subsequent examination of this mouse model showed a duplication of the 3′ end of the WT Nod2 locus, which was targeted by the mutation (Maeda et al., 2005). Furthermore, an independently generated knockin L1007insC NOD2 mouse model revealed that the mutation acts as a loss-of-function (Kim et al., 2011b) consistent with human studies.

Potential for the therapeutic modulation of and NOD1 and NOD2 signaling

Several groups are attempting to develop small molecules capable of modulating NOD1 and NOD2 signaling in order to influence pathogen clearance and inflammation (Geddes et al., 2009; Jakopin, 2014; Maisonneuve et al., 2014). For instance, molecules capable of enhancing or blocking NOD2 signaling would have application in diseases such as CD and EOS which have been associated with NOD2 loss- and gain-of-function mutations, respectively. The development of small molecules capable of modulating NOD1 and NOD2 signaling may also have potential applications in other infectious and inflammatory disease such asthma, arthritis, leprosy, graft versus host disease, and periodontitis which have each been associated with deregulated NOD1 and/or NOD2 signaling (Correa et al., 2012; Jiao et al., 2014; Philpott et al., 2014) but await further functional validation. One of the first successful approaches to modulate NOD2 signaling using small molecules came from the development of kinase inhibitors capable of blocking RIPK2 (Jun et al., 2013; Tigno-Aranjuez et al., 2010; Tigno-Aranjuez et al., 2014) while others identified anti-inflammatory activities for diterpene-based molecules via their ability to specifically block NOD2 signaling (Bielig et al., 2010). More recently, libraries of small molecules were screened identifying lead compounds capable of influencing NOD1 (Khan et al., 2011; Rickard et al., 2013) and NOD2 signaling (Correa et al., 2011; Rickard et al., 2013). In most cases, the molecular mechanisms by which these small molecules act to regulate NOD1 and NOD2 signaling remain poorly understood. It is possible that these small molecules act directly affecting NOD protein stability, ligand recognition, and/or ATP binding and hydrolysis activity. Alternatively, they may target any one of the numerous regulatory proteins in the NOD1 and NOD2 pathway revealed by recent siRNA screens (Lipinski et al., 2012; Warner et al., 2014).

While most of the small molecule screens identified NOD1 and NOD2 inhibitors, some investigators have attempted to identify NOD1 and NOD2 immunostimulatory molecules with potential applications as vaccine adjuvants or anti-microbial agents. Early efforts in this area involved the chemical modification of core PGN ligands to enhance NOD1 and NOD2 activity. Structure-activity relationship experiments revealed chemical modifications that either attenuate or enhance the stimulatory activity of the NOD1 ligand iE-DAP (Agnihotri et al., 2011; Jakopin et al., 2013) or MDP (Jakopin et al., 2012; Rubino et al., 2013). These studies suggest that new ligands can be developed to enhance NOD1 and NOD2 immunity.

In addition to the role of NOD1 and NOD2 in regulating gut homeostasis and protection against colitis, there is evidence that NOD1 and NOD2 regulate colitis-associated cancer. For example, mice lacking NOD1 (Chen et al., 2008), RIPK2, or NOD2 (Couturier-Maillard et al., 2013) develop more tumors in colitis-associated colon cancer models. There is also data suggesting a role for NOD1 and NOD2 in tumors outside of the intestine (da Silva Correia et al., 2006). Together these studies support the modulation of NOD1 and NOD2 signaling as a therapeutic strategy in the treatment of pro-inflammatory diseases and cancers; however, additional studies will be required to determine the negative and positive effects of targeting the NOD1 and NOD2 pathway in order to balance the inhibition of detrimental inflammation against immune suppression and susceptibility to infection.

Concluding Remarks

Compelling evidence indicate that NOD1 and NOD2 are involved in the recognition of highly conserved PGN molecules that enter the host cytosol through phagocytosis, endocytosis, endosomal membrane transporters, bacterial secretion systems, pore-forming toxins or outer membrane vesicles. NOD1 and NOD2 directly recognize PGN molecules leading to their activation and induction of pro-inflammatory and anti-microbial molecules. Biochemical analyses and genetic screens have revealed complex mechanisms that regulate NOD1 and NOD2 signaling. NOD1 and NOD2 play a role in the clearance of invading pathogens and act redundantly and cooperatively with TLRs in the detection of bacteria and production of pro-inflammatory molecules. Experimental evidence suggests several mechanisms by which NOD2 mutations regulate the susceptibility to CD. Although our knowledge about NOD1 and NOD2 has increased, several key questions remain including the precise mechanism of NOD1 and NOD2 activation by PGN molecules and how NOD2 acts in the intestine to regulate the susceptibility to CD. Undoubtedly, elucidating these questions will provide new insights into the mechanisms of host defense and the pathogenesis of inflammatory disease.