Gut Permeability and Colitis

Article Outline

• Comment
• Copyright

Arieta MC, Madsen K, Doyle J, et al. (Department of Medicine, University of Alberta, Walter C. Mackenzie Health Science Centre, Edmonton, Alberta, Canada). Reducing small intestinal permeability attenuates colitis in the IL10 gene-deficient mouse. Gut 2009;58:41–48.

Gut epithelial permeability, the integrity of its physical barrier function, plays an important part in the development of inflammatory bowel disease and autoimmunity in certain animal models. Because increased gut permeability has intriguing associations with Crohn's disease, celiac disease, and HIV/AIDS, it follows that emerging information concerning the regulation of this physical barrier could point to new therapies for testing in human diseases. One example of this research is the discovery of zonulin regulation of gut epithelial permeability. A mammalian homolog of a Vibrio cholerae toxin, zonulin is secreted by lamina propria cells and activates apical receptors on gut epithelial cells leading to phosphorylation of tight junction proteins and increased permeability owing to enlargement of the paracellular space. Recently, a novel peptide inhibitor has been developed to block zonulin from binding to its receptor, with its effects limited to the small intestine where the receptor is located. Human trials with this drug (AT-1001, Alba Therapeutics) are underway in celiac disease, a setting where excess zonulin secretion occurs and increased gut permeability can precede clinical disease (Lancet 2000;355:1518–1519). This report in Gut suggests that inhibition of zonulin binding can reverse the preexisting small bowel permeability defect in a well-characterized mouse model of human Crohn's disease, the IL-10–deficient mouse, and furthermore can ameliorate features of the colitis, a site distant from the small bowel permeability defect (interestingly, the colon itself is unaffected directly by zonulin). These findings would further suggest that permeability defects in the small bowel contribute to the immune response at distant sites such as the colon (a paradigm already suspected in the BB rat model of autoimmune type 1 diabetes).

These investigators used the IL-10–deficient (IL10^−/−) mouse as a model of colitis; histologically detectable colonic inflammation develops spontaneously at around 10 weeks after conventional animal housing (ie, after exposure to environmental bacteria). These mice typically do not show histologic evidence of small intestinal enteritis, however. This increase in permeability also was not dependent on excessive inflammatory cytokine production (interferon-γ and tumor necrosis factor-α) in the small intestine. IL10^−/− mice and controls were treated with the zonulin antagonist AT-1001 (added to the drinking water) or placebo for 17 weeks. Using validated measures of small intestinal and colonic permeability, they showed that immediately after weaning, the IL10^−/− mice had increased small bowel permeability compared with wild-type control mice at all weekly timepoints measured. The small bowel permeability defect was prevented by treatment with the zonulin-antagonist AT-1001. This improvement in the epithelial layer leakiness was seen using both in vivo (lactulose/mannitol fractional excretion ratio) and in vitro (mannitol flux and transepithelial electrical potential difference using mucosa mounted in Ussing chambers) methods.

Importantly, there are also apparent effects of AT-1001 treatment on the course of the colitis, but the data from the studies of the colonic response raise some questions. First, we do not know whether there are baseline permeability defects in the colon similar to what exist in the small bowel of IL10^−/− mice right after weaning. This is important to know; a zonulin-independent colonic permeability defect could contribute to the spontaneous colitis despite effects on small bowel permeability. However, the increase in colonic permeability measured as increased mannitol flux and lowered electrical resistance in mucosal sections, at the 8-week time point just preceding the first signs of colonic damage (increased sucralose excretion at 10 weeks), is prevented by treatment with the zonulin antagonist. Furthermore, although treatment with the zonulin antagonist significantly prevents increases in colonic permeability (sucralose excretion) over time, it is interesting that the placebo and treatment groups have the same effect on bringing colonic permeability back to baseline by the end of the 17-week study period, suggesting that critical timepoints for zonulin effects on colonic responses may be early in the induction of the spontaneous colitis. Additionally, there may be some methodologic issues with the use of sucralose excretion in this model; at 17 weeks, the sucralose excretion scores are the same, whereas the histologic scores and inflammatory cytokine secretion are significantly different (and factors such as epithelial cell apoptosis and epithelial ulceration can increase sucralose excretion); it is not clear why the sucralose excretion is not higher in a more inflamed colon. Finally, although measures of neutrophil infiltration of colonic tissue were unchanged at the end of treatment between groups, the inflammatory cytokine secretion increases seen in placebo-treated animals were prevented by zonulin antagonist treatment, similar to what was seen in the small intestine.

Back to Article Outline

Comment 

This report is timely for 2 reasons. First, the role of increased intestinal permeability in disease pathophysiology is being expanded beyond its traditionally accepted role as being limited to gastrointestinal diseases. Increased permeability of the gut epithelial layer has long been recognized as a factor contributing to the severity of local intestinal lamina propria inflammation, like in Crohn's disease and celiac disease. However, increased gut permeability is also a factor contributing to the development of autoimmune type I diabetes in the BB rat model where it is a necessary but not sufficient component for disease initiation (Am J Physiol 1999;276:G951–G957). Interestingly, type I diabetic humans also have increased small bowel permeability and even increased zonulin serum levels (Diabetes 2006;55:1443–1449); increased small bowel permeability may be an early event in autoimmune diabetes, but it is not known if a permeability defect is actually a requirement for diabetes to develop, and if so whether it would be a constitutional or a transient defect. Regardless, the result of the permeability defect is a presumed increase in antigen exposure with consequent autoimmune inflammation establishing itself with increased frequency in genetically susceptible hosts. On the other hand, gut permeability defects can result in general immune activation, as seems to be the case in HIV/AIDS. Increased microbial translocation of the gut epithelium as reflected by increased plasma lipopolysaccharide levels is found in patients with chronic HIV and AIDS. This can be reduced by successful antiretroviral therapy. In fact, small bowel permeability (measured by mannitol flux and impedence spectroscopy on endoscopic biopsies) is significantly elevated in untreated versus treatment-suppressed HIV patients and seronegative controls (Gut 2009;58:220–227). In this case, enhanced gut permeability is likely related to epithelial barrier response to local gut infection and cytokine production, and the systemic effect of enhanced immune activation (with lymphocyte proliferation and retrovirus replication) is thought to play a role in HIV disease progression.

The second reason this report is timely is that AT-1001 is currently being evaluated for treatment of celiac disease in humans. With the introduction of this agent into clinical trials, there is the potential for testing this drug in a number of autoimmune and inflammatory states that may rely on increased gut permeability for initiation or persistence of the disease.

This report does raise an interesting question about the mechanism of increased gut permeability in inflammatory bowel disease and in this mouse model. Assuming that the IL10^−/− mice have baseline increases in small bowel permeability (but not colonic permeability) before developing colitis, and by correcting this defect the colitis can be prevented, how then does the small intestine control the colon immune response? Perhaps as in the case of the BB rat, a more permeable small bowel allows increased exposure to luminal antigen, and naïve T cells becoming antigen experienced in mesenteric lymph nodes of the small intestine then home to the lamina propria throughout the gut, including the colon where they are activated (notably, in the AT-1001–treated mice, interferon-γ secretion was significantly reduced in the small and large bowel). Alternatively, the small bowel permeability defect may lead to a general immune activation that can lower the threshold development of colitis on a susceptible genetic background.

This report aims to demonstrate that control of increased gut epithelial permeability can alter the natural history of colitis. This is most clearly seen in the data showing early precolitis improvement in small bowel and colonic permeability (before 10 weeks) and later suppression of inflammatory cytokine production in the colon of zonulin antagonist-treated mice. Of course this model of spontaneous colitis in IL10^−/− mice does not recapitulate the immune background of Crohn's disease patients, but it is difficult to not point out that patients with Crohn's disease and their unaffected first-degree relatives also display increased small bowel permeability and even a heightened sensitivity to nonsteroidal anti-inflammatory drug–induced increases in permeability. It would be interesting to know whether the permeability changes are associated more with the occurrence of Crohn's colitis compared with isolated small bowel disease. Having said this, it is not known whether this permeability defect in human Crohn's disease is zonulin dependent or not. Furthermore, once inflammation is active, it is well recognized that inflammatory cytokines can increase epithelial barrier permeability, leading to increased exposure of the lamina propria to luminal contents, suggesting that targeting cytokines as well as targeting the components of the epithelial barrier itself could both contribute to normalizing permeability defects.

With this in mind, maintenance of the epithelial barrier either directly by stabilizing the tight junction, or indirectly by blocking inflammatory cytokines, is a novel approach to treating IBD.