In the paper by Wapenaar et al (see page 10.1136/gut.2007.133132) the authors have taken a fascinating approach to identifying shared mechanisms involved in the genesis of either coeliac disease or inflammatory bowel disease (IBD).1 They argue that in these two prototypical inflammatory diseases of the intestinal mucosa there exists reasonable evidence for a defect in barrier function that appears to be required before the development of disease. Furthermore, as both diseases have strong genetic components they speculated that these syndromes might share common genetic defects in the control of intestinal barrier function. They used a genetic association analysis approach and through this identified two adapter protein coding genes that were associated with coeliac disease in patients from both Great Britain and The Netherlands. They went on to demonstrate that one of these genes was also associated with ulcerative colitis in a Dutch patient cohort.

These observations are important not only for the conclusions reached in the paper but also in the broader context. Until recently, it was believed that IBD, such as Crohn’s disease, represented dysregulation of the adaptive immune system. Over the past decade, however, there has been increasing recognition of the importance of both epithelial barrier function and innate immunity in the genesis of intestinal inflammation. In the broadest sense these two factors could be argued to be different aspects of the same basic system. Within the gastrointestinal tract there is significant exposure to foreign compounds that can drive systemic inflammation through a variety of mechanisms. The gut has a tremendous number of defence mechanisms that have evolved to manage this ever-changing threat (fig 1). In general terms these include the ability to manage commensal flora in preference to pathogenic organisms, the secretion of toxic molecules such as defensins, the scavenging and binding of luminal organisms by specifically formulated mucins, the presence of regulated tight and adherens junctions between epithelial cells that regulate the passage of potentially pro-inflammatory molecules and the presence of both intra and extracellular pattern recognition molecules that can regulate immunological responsiveness to environmental stimuli. Finally, the adaptive immune system, which sits on top of this large defensive system, can fine tune the responses to a wide variety of environmental agents. It is an amazingly complex system that in most of us functions extremely well!

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Intestinal defence. This schematic broadly outlines, in a simplified manner, the three levels of defence that make up what we refer to as intestinal barrier function. Level A is external to the epithelial cells and consists of host bacteria, secreted mucin and antibacterial products such as defensins and immunoglobulins. Level B represents the epithelial cells. These have rigid and relatively impermeable brush border membranes. The route between cells is sealed by tight and adherens junctions. Integral to these cells is the ability to react in a pre-programmed manner to foreign compounds using pattern recognition receptors found either on the surface of the cells (Toll receptors) or internally such as NOD2/CARD15. Finally, level C represents the subepithelial layer. Here there are important interactions that take place primarily under the control of the adaptive immune system and in response to pathogenic compounds that make it through the other lines of defence.

Given the complexity of this defensive system, however, it is not at all surprising that defects in many of these important systems could ultimately lead to inflammatory disease. Furthermore, as the mucosal immune system is “educated” primarily in the gut and these cells subsequently migrate elsewhere, it is perhaps not surprising that defects in these systems may lead to inflammatory disease that can be expressed at sites distant to the intestine. This is becoming increasingly apparent in human disease and in animal models of disease.

It is beyond the scope of this commentary to review each aspect of mucosal defence exhaustively and there have been excellent reviews recently.2 I would, however, like to discuss one aspect of gut barrier function; that being abnormal epithelial permeability and disease. The genetic abnormalities described in this paper would appear to fit most closely with this system.

Abnormal permeability refers to a measurable increase in flux of small water-soluble compounds across the paracellular pathway of the small intestine. The rate of movement across this pathway is regulated primarily by the functional state of the tight and adherens junction. These, in turn, are controlled by a complex array of intracellular proteins within the enterocyte as well as the protein composition of the junctions themselves. Increased permeability can be observed as a result of action by inflammatory cytokines (such as tumour necrosis factor α, IL17 or IFN-γ), bacterial interactions with the enterocyte, migration of inflammatory cells across the epithelium, nutrient transporter activation, noxious environmental agents or it may exist de novo, without apparent cause.3^–6 In the latter case this may be secondary to an alteration in the protein composition of the junctions or presumably their regulatory systems. In the paper by Wapenaar et al,1 in this issue of Gut, it is a gene association in these regulatory proteins that appears to be shared.

It has long been appreciated that individuals who are at increased risk of the development of IBD demonstrate evidence of increased gastrointestinal permeability.7 8 Furthermore, similar evidence is present in animal models of both coeliac and IBD. Perhaps most important is the observation that increased permeability precedes the development of disease, suggesting that it is an important pathophysiological event. In IBD such evidence is available from human and animal studies, whereas in coeliac disease this has been demonstrated in animal models.

Our current understanding of coeliac disease is that a cereal protein, highly resistant to human digestive processes secondary to enrichment with glutamine and proline, “leaks” across the intestinal epithelium and in the susceptible host initiates an inflammatory cascade that we call “coeliac disease”.9 10 Of the several peptide fragments known to induce disease the most potent is a 33 amino acid oligomer. This is not a small peptide and it begs the question as to how this peptide can cross a relatively tight epithelium? In this regard there are intriguing data from an animal model of coeliac disease. Inbred Irish setter dogs have been described to have a gluten-sensitive enteropathy that is similar to human coeliac disease. As we know that these animals will ultimately develop the disease, it affords us the opportunity to ask whether there is an abnormality present in the intestinal epithelium that precedes exposure to gluten. Intuitively, it makes sense that these animals should have abnormal intestinal permeability so that these peptides can cross the epithelium and initiate disease. This has been tested by evaluating small intestinal permeability in animals that had never been exposed to gluten. Susceptible pups were weaned from maternal milk directly onto a gluten-free diet and when tested had clear evidence of increased permeability, despite never having been exposed to gluten.11 These data argue strongly that abnormalities in intestinal permeability may exist in the absence of disease and may, at the very least, predate the appearance of inflammatory disease.

An important point to recognise is that to a large extent we experience our environment through the small intestine. This includes antigens derived both from food and the resident flora of the intestine. If any of the above-noted defensive mechanisms breaks down inflammatory disease may develop. Although this inflammation may occur within the wall of the intestine, such as in coeliac or Crohn’s disease, it is important to recognise that inflammation need not be restricted to the gut. There is now evidence to suggest that abnormal small intestinal permeability, and consequent dysregulation of an inflammatory response to a luminal agent, can result in disease distant to the intestine. Perhaps the best studied example is type 1 diabetes.

The BB rat develops type 1 diabetes. Although this is an immune-mediated destruction of islet cells intriguing evidence has been presented that in this animal model the inflammatory process involves the gut. Feeding these animals a hydrolysed diet, low in intact protein antigens, can largely prevent the development of disease.12 Again this begs the question as to whether these protein mediators need to gain access to the mucosal immune system and if so does this require abnormal intestinal permeability. In the BB rat there are now data to answer these questions. First, there is clearly increased small intestinal permeability that precedes the development of insulitis.13 Second, it has now been suggested that the increased permeability is secondary to the abnormally increased secretion of a luminal hormone, zonulin, which increases epithelial permeability. Antagonising the effects of this hormone with a peptide that competes with zonulin for the receptor both normalises the increased permeability and also prevents the appearance of diabetes.14 This is strong evidence that not only is abnormal permeability present before disease but that it is also required for the development of disease. These data, observed in an animal model, take on renewed interest in the light of recent data demonstrating that humans, studied at the onset of diabetes, also have increased intestinal permeability.15

These findings reinforce the importance of our intestinal defence system. They point out to us that there exists a pathological mechanism for the generation of disease that is perhaps under-appreciated. Our exposure to the antigenic universe largely occurs through the small intestine and we are only beginning to appreciate the complexity of these interactions. Abnormal permeability seems to be clearly involved in the genesis of IBD, coeliac disease and diabetes.16 There are also, however, new findings that suggest a similar role in other poorly understood diseases such as irritable bowel syndrome and perhaps atopic eczema or asthma.17^–22

A paradigm to consider is that some inflammatory diseases may require a minimum of three conditions to be present for the development of disease. The first requirement is an abnormality of the immune system to interact inappropriately with an environmental antigen to develop disease. The second is clearly the presence of the antigen or inciting agent—no antigen, no disease. Finally, the third requirement is that the antigen must reach the immune system and that may require abnormal intestinal permeability or some other breach in the intestinal barrier. The data that we have at hand support this broad concept and point us perhaps to new ways of treating inflammatory disease.

The paper by Wapenaar et al1 is important from this perspective. The authors took advantage of the recognised association between two inflammatory diseases and did so in the context of this new paradigm. They identified a shared genetic predisposition for these diseases that involves proteins important in the regulation of epithelial junctions. To some extent this is not surprising. The clinical implications are, however, much more provocative. These data suggest that an important step in the genesis of these diseases involves inappropriate interactions between luminal antigens and the mucosal immune system. As such, we would expect similarly increased susceptibility to these diseases in patients or animal models with abnormalities in mucin production, defensin secretion, pattern recognition molecules, epithelial permeability or any other important part of intestinal barrier function. What is important, however, is that this provides a novel therapeutic target for the management of disease. Rather than trying to suppress inflammation downstream in the affected tissue would it not make more sense to target the initiators of inflammation before the disease develops? If the identified defect involves increased epithelial permeability, as is the case in the diseases discussed in this paper, could this be a useful therapeutic target? There are already targets identified for therapy such as the zonulin pathway and the identification of others may lead to a variety of potentially synergistic therapies.23

Finally, I believe that it is worth considering one additional and important point. From a clinical perspective we already know that many other diseases are associated with IBD or coeliac sprue. There is a diverse array of systemic inflammatory diseases that we routinely observe clinically in these settings. The causal relationships between these diseases have, however, remained obscure. Perhaps we need to revisit these disease associations in the context of the paradigm presented above and look at techniques similar to those used so well by Wapenaar and colleagues1 in this issue of Gut.