Article Text

Download PDFPDF

NOD2 mutations and Crohn’s disease: are Paneth cells and their antimicrobial peptides the link?
  1. M C Grimm1,
  2. P Pavli2
  1. 1Department of Medicine, St George Clinical School, University of NSW, Sydney, Australia
  2. 2Gastroenterology Unit, Canberra Hospital, Canberra, Australia
  1. Correspondence to:
    Dr P Pavli
    Gastroenterology Unit, Canberra Hospital, Woden ACT 2606, Canberra, Australia;

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

There is evidence that CARD15/NOD2 variants in Crohn’s disease are associated with diminished mucosal α-defensin production, resulting in impaired innate immunity


Paneth cells (reviewed by Cunliffe and colleagues1,2) are granulated epithelial cells found at the base of small intestinal crypts. They are most numerous in the terminal ileum and are not found in the normal colon. They secrete a variety of antimicrobial peptides and proteins, including α-defensins, lysozyme, and secretory phospholipase A2 (sPLA2), in response to bacterial products and cholinergic agents. There is a link between intestinal inflammation and Paneth cell number. In mouse models of enteric infection, Paneth cell hyperplasia occurs in response to signals from activated T cells, in mouse models of ileocolitis there is Paneth cell metaplasia, and in inflammatory bowel disease, Paneth cell metaplasia is also seen in the colon. Parasite infestations also induce an increase in the number of Paneth cells in the murine small intestine.

Defensins (reviewed by Ganz3) are a family of small (29–45 amino acid) cationic antimicrobial peptides that are divided into two main classes, the α- and β-defensins, on the basis of their disulphide bond pairing pattern. They are encoded by single genes as prepropeptides, which are then processed to mature active forms. The mammalian defensin genes probably evolved from a single ancestral gene but because of selection pressure and rapid evolution, comparisons between species is difficult.4

Six α-defensins have been identified in humans: four are neutrophil peptides and two human defensins (HD-5 and HD-6) are expressed in intestinal Paneth cells. HD-5 is expressed in Paneth cells of the duodenum, jejunum, and ileum, but is not found in the normal stomach or colon. In cases of gastritis, Crohn’s colitis, and ulcerative colitis, HD-5 is expressed in metaplastic Paneth cells.

In vitro, α-defensins have broad spectrum antimicrobial activity and are active against a variety of bacteria, fungi, parasites, and viruses. Neutrophil defensins kill bacteria engulfed in phagolysosomes by formation of microbial selective, membrane spanning pores which lead to cell lysis.

In contrast with enteric α-defensins, human β-defensin 1 (HBD1) and other members of the β-defensin family appear to be expressed by most epithelial cells of the small and large intestine. HBD-1 is expressed constitutively while HBD-2 is induced by stimuli that activate the transcription factor nuclear factor κB (NFκB).5


Hugot and colleagues6 and Ogura and colleagues7 independently identified mutations in CARD15/NOD2 that confer susceptibility to Crohn’s disease. The three main disease associated alleles (Arg702Trp (C2104T), Gly908Arg (G2722C) and 1007fs (3020insC)) and other variants cosegregate with disease in up to 30–50% of patients, depending on the population studied.

The protein product of this gene is thought to be a cytoplasmic receptor for muramyl dipeptide, a breakdown product of the peptidoglycan component of all bacterial cell walls.8,9 Lysozyme, including that derived from Paneth cells, is involved in this process by breaking the α-1,4 glycosidic linkages in the polysaccharide components of peptidoglycans. CARD15/NOD2 is thought to act through the NFκB pathway to regulate production of tumour necrosis factor (TNF)-α and other proinflammatory cytokines.10

Hisamatsu and colleagues11 showed that CARD15/NOD2 mRNA was expressed and CARD15/NOD2 protein was detectable in intestinal epithelial cells (IECs). Using stable transfectants, they showed that CARD15/NOD2 was a potent factor for recognition of intracellular bacteria in IECs and that this function was impaired in 3020insC variant CARD15/NOD2 expressing IECs.

Definitive studies of the distribution of CARD15/NOD2 in vivo showed expression in Paneth cells of the normal duodenum, jejunum, and terminal ileum (the greatest staining intensity was noted in terminal ileal crypts). In inflammatory bowel disease, CARD15/NOD2 expression in Paneth cells was increased and was also evident in metaplastic Paneth cells in the colon.12

Genotype-phenotype studies show that the known CARD15/NOD2 variants are more likely to have terminal ileal involvement.13–15

These observations raised the possibility that there was a link between CARD15/NOD2 variants and defective α-defensin release by Paneth cells, resulting in impaired innate immunity. Paneth cells detect and respond to bacteria, and the CARD15/NOD2 protein may mediate their interaction with bacterial products. Inherited mutations in CARD15/NOD2 could interfere with normal Paneth cell function resulting in a disordered immune response.

A paper by Wehkamp and colleagues16 in this issue of Gut provides supportive evidence that CARD15/NOD2 variants in Crohn’s disease are associated with diminished mucosal α-defensin production (see page 1658). Wehkamp et al examined expression of HD-5, HD-6, lysozyme, sPLA2, TNF, and IL-8 in 45 Crohn’s disease patients (24 with CARD15/NOD2 variants and 21 without) and 12 controls. They found that expression of HD-5 and HD-6 specifically was diminished in affected ileum, particularly in those bearing CARD15/NOD2 variants. Other proteins and cytokines were unaffected. In the colon, HD-5 and HD-6 expression was increased during inflammation in wild-type but not CARD15/NOD2 variant bearing patients. There are some unexplained results in this paper: why were TNF-α levels not increased in inflamed ileum? Why were lysozyme levels in unaffected wild-type ileum not the same as in controls? How do they explain the increase in α-defensins in inflamed colon while levels were decreased in the inflamed small bowel?


The predominant site of disease localisation in the CARD15/NOD2 variant bearing patients is the terminal ileum which is also the site of maximal Paneth cell density—guilt by association. The density of Peyer’s patches is also greatest in the terminal ileum. These organised collections of lymphoid cells play a role in the selective uptake of luminal antigens. They are also the site of the earliest lesion in Crohn’s disease—the aphthous ulcer, an area of focal ulceration over a Peyer’s patch (in the small intestine) or lymphoid follicle (in the colon). Perhaps Paneth cell α-defensins present in high levels in the terminal ileum inactivate luminal bacteria taken up by Peyer’s patches or colonic lymphoid follicles so the immune system need only deal with inactivated bacteria rather than viable microorganisms.

In addition, α-defensins (and their β-family counterparts) play a linking role between the innate immune system, characterised by rapid antimicrobial responses, and adaptive immunity instigated by B and T cells. The α-defensins are selectively chemotactic for resting naïve CD4+ T cells and immature dendritic cells although at concentrations several logs lower than their antimicrobial concentrations.17 Impaired release of defensins from Paneth cells might thereby augment this chemotactic potential while simultaneously reducing their antipathogen effector function. Similarly, α-defensins act as immunological adjuvants, heightening Th1 and especially Th2 immune responses to non-specific antigens18 and to tumours19; a reduction in mucosal defensin concentrations might allow skewing of the immune response to the Th1 profile characteristic of Crohn’s disease.

Most importantly, CARD15/NOD2 and α-defensins potentially are linked in the Paneth cell as recognition and effector molecules, respectively, for pathogen associated molecular pattern (PAMP) motifs: their colocalisation in these cells is difficult to ignore. Some of these PAMPs, including lipopolysaccharide, lipoteichoic acid, dsRNA and, significantly, muramyl dipeptide, have been shown in an elegant study to induce rapid release of defensins from murine ileal crypt Paneth cells.20 Indeed, this defensin release occurred within 15 minutes and therefore did not reflect an effect on gene expression. While Wehkamp et al did show an association between defective CARD15/NOD2 and α-defensin gene expression, a much more profound effect might yet be detected in secretion of stored α-defensins. An analogous situation may have been observed, unknowingly, in intestinal epithelial cell lines which are able to produce β-defensins: when transfected with wild-type CARD15/NOD2, these cells resist infection with Salmonella typhimurium; if transfected with mutated CARD15/NOD2, they are more easily infected.11

While many of the specific details linking genetic defects with reduced antimicrobial resistance in Crohn’s disease are still to be identified, there is strengthening circumstantial evidence to suggest that CARD15/NOD2 variant mediated deficiency of Paneth cell derived α-defensins is critical in the development of some forms of Crohn’s disease.


Direct proof of the functional significance of these peptides in vivo is scarce. How important is a specific defensin when there are other antimicrobial molecules with similar properties? How important are Paneth cells? In one transgenic mouse strain, ablating the entire Paneth cell population had no observable effect on host-microbial interactions.21 Moreover, mice survive even though they lack granulocyte defensins completely.

However, there is evidence that defensin deficiency does result in impaired host resistance: knockout mice that cannot process the prepropeptides and secrete inactive cryptdin precursors are more susceptible to specific infections than wild-type littermates. Transgenic mice that express (human) HD-5 (as well as the normal crytpdins) in Paneth cells are resistant to oral infections compared with controls, indicating that HD-5 is an effective luminal antimicrobial.22

Nevertheless, these and other studies of defensin deficiencies in experimental systems generally demonstrate susceptibility to specific infections rather than non-specific inflammation, in much the same way as selective IgA or other immunodeficiencies result in (generally) identifiable infections in humans. It is possible, as implied by Lala and colleagues,12 that in Paneth cells there is a Crohn’s disease related dysregulation of TNF gene expression (and presumably protein production or secretion) but there is much to be done to prove this hypothesis. If defensin deficiencies result in increased susceptibility to infections (either specific or non-specific) we have to be able to explain the “effectiveness” of immunosuppressive medication in controlling symptoms of CD—or at least not worsening the disease.

Much is made of the predominant terminal ileal involvement found in CARD15/NOD2 variant bearing patients. However, many of these patients have colonic involvement, generally in association with ileal disease, but sometimes exclusively. In the later stages of disease, there is Paneth cell metaplasia in the colon, but this is not present in the early phase. How does a defect of terminal ileal Paneth cell derived α-defensins result in left sided colonic involvement? There are plausible explanations other than Paneth cell density. For example, the terminal ileum is also the site of greatest density and size of Peyer’s patches; it may be the ability of the associated M cells to take up luminal antigens and deliver them to the underlying lymphoid structures that initiates Crohn’s disease. This same process may also induce aphthous ulceration over lymphoid follicles in the colon.

It may indeed be that there is a CARD15/NOD2 variant mediated deficiency of Paneth cell derived α-defensins but there are still numerous steps that need to be taken before there is conclusive proof. For example, there are no NFκB binding sites in the promoter regions of the Paneth cell α-defensins so the means by which CARD15/NOD2 variants affect α-defensin protein production and secretion need to be defined. Obtaining the evidence will be very difficult in the absence of appropriate experimental systems to study signal transduction in Paneth cells.

According to Sherlock Holmes “…when you have eliminated the impossible, whatever remains… must be the truth.” (The Sign of Four. Sir Arthur Conan Doyle). There is evidence implicating the Paneth cell but it is not beyond reasonable doubt. Also, there is another possibility: the newly arrived monocyte is known to express high levels of CARD15/NOD212 and to possess a plethora of antimicrobial and other effector functions. It remains a prime suspect and, like the Paneth cell, could also be regarded as guilty until proven innocent.

There is evidence that CARD15/NOD2 variants in Crohn’s disease are associated with diminished mucosal α-defensin production, resulting in impaired innate immunity


Linked Articles