Bile Acids as Modulators of Gut Microbiota Linking Dietary Habits and Inflammatory Bowel Disease: A Potentially Dangerous Liaison

Abstract

Devkota S, Wang Y, Musch MW, et al. Dietary-fat-induced taurocholic acid promotes pathobiont expansion and colitis in IL10−/− mice. Nature 2012;487:104–108. Epidemiologic, clinical, and experimental data have uncovered an important association between inflammatory bowel diseases (IBD) and several environmental factors such as diet, antibiotic usage, or microbial exposure during lifetime. These environmental factors might exert their main influences via modulation of the gut's microbiota and it is currently believed that IBD reflects a microbiota-driven disease in a genetically susceptible host. Overall, there is now increasing support that IBD's microbiota is disturbed (commonly described as dysbiosis), certain “pathobionts” exist, and IBD's microbiota is characterized by low diversity of species and a high density of mucosal surface colonization (Nat Rev Gastroenterol Hepatol 2012;9:599–608). Dietary factors are powerful modulators of the gut microbiota and there is accumulating evidence that dietary intake may affect development of IBD (Curr Opin Gastroenterol 2012;28:314–320). As such, high intake of total fat, polyunsaturated fatty acids, ω-6 fatty acids, and meats were identified as risk factors for developing both Crohn's disease and ulcerative colitis, whereas high intake of fruits and dietary fibers may reduce Crohn's disease incidence and high vegetable intake lowers risk for ulcerative colitis. Therefore, dietary habits could at least partially explain the rising incidence of IBD in many regions of the world, although the underlying mechanisms how diet alters microbiota are still poorly understood (Gastroenterology 2012;142:46–54). In addition to directly fueling microbiota with nutrients, recent studies have demonstrated how certain dietary components such as cruciferous vegetables interact with intestinal aryl hydrocarbon xenobiotic receptors and thereby regulate intestinal immune processes and the microbiota (N Engl J Med 2012;366:181–183). Apart from their detergent properties in lipid digestion, bile acids (BAs) have more recently been recognized to have additional hormonal actions that control a range of metabolic and immune functions throughout the body via dedicated BA receptors such as the nuclear BA receptor/farnesoid X receptor (FXR) and a G-protein–coupled receptor TGR5 (Gastroenterology 2011;140:1120–1125). In the gut, BA-activated FXR maintains epithelial barrier integrity by induction of multiple genes involved in intestinal mucosal defense against inflammation and microbes that, together with direct antibacterial detergents actions, help to control the gut microbiota. Conversely, alterations and genetic variants of FXR and TGR5, respectively, have been linked with IBD and pharmacologic activation of FXR (and possibly TGR5) improves experimental (dextran sulphate sodium-induced) colitis in mice. Finally, bacterial overgrowth owing to reduced intestinal BAs is a long known complication of cholestasis and end-stage liver disease (Gastroenterology 2011;140:1120–1125). Importantly, the gut microbiota is also able to alter BA metabolism and signaling through deconjugation and dehydroxylation resulting in formation of secondary, more hydrophobic BAs with altered ligand properties for BA receptors. As such, BA dysmetabolism through impaired microbiota enzymatic activity in IBD-related dysbiosis may affect luminal BA composition, thereby affecting the pro-/anti-inflammatory potential of BAs (Gut 2012, Epub Sept 19). In this article, Devkota et al (Nature 2012;487:104–108) provide novel, mechanistic insights into changes in BA metabolism as potential links between dietary fat contained in many Western-type diets and IBD. In a series of elegant experiments conducted in specific pathogen-free wild-type and genetically IBD-susceptible interleukin-10 knockout (IL10−/−) mice, the authors demonstrated that 3 weeks of saturated milk-derived fat (closely mimicking fats found in processed and confectionary foods), but not polyunsaturated (safflower oil) fat or low-fat diet, induced alterations in BA metabolism with increased formation of taurin-conjugated BAs (eg, taurocholic acid). These changes in bile composition increased the availability of organic sulphur, which in turn promoted the in vitro and in vivo expansion of a low-abundance, sulphite-reducing pathobiont and member of Deltaproteobacteria, Bilophila wadsworthia, resulting in increased susceptibility and severity of colitis in genetic susceptible IL10−/− (but not wild-type) mice by inducing a T-helper type 1 immune response. Although milk fat and polyunsaturated fat had similar effects on a higher abundance of Bacteroidetes and lower abundance of Firmicutes, changes that have previously been typically associated with fat consumption, only milk fat resulted in significant bloom of B wadsworthia. Notably, milk fat also aggravated the incidence and severity of dextran sulphate sodium-induced colitis in specific pathogen-free wild-type mice, which again was associated with B wadsworthia. As further evidence, mono-association with B wadsworthia and induction of colitis in germ-free IL10−/− mice could only be established with consumption of milk fat diet (but not polyunsaturated or low-fat diet) or gavage with taurocholic acid (but not glycocholic acid or PBS buffer) when mice were placed on a low-fat diet. Collectively, these findings provide for the first time a mechanistic basis for the link between certain dietary fats contained in Western diets, dysbiosis and development of colitis in genetically susceptible mice by revealing an unexpected role for certain BAs as growth factor for a specific pathobiont. These novel insights may close the mechanistic gap to original observations dating back from as far as the 1950s, suggesting already a role for milk fat and milk products in the pathogenesis of Crohn's disease (Trans Am Clin Climatol Assoc 1969;80:116–124). This landmark study significantly advances our understanding how dietary components can alter gut microbiota and has already received much attention, with multiple commentaries focusing mainly on dietary and immunological aspects in IBD pathogenesis. In addition, this study profoundly expands our current concepts on the gut–liver axis and BA (patho)biology and may consequently blaze the trail for some paradigm changes. Thus far, bile has been considered an antibacterial, which was mainly attributed to direct and indirect antimicrobial actions of BAs; moreover, in nontoxic ranges BAs exert anti-inflammatory hormonal actions via FXR and TGR5 (Gastroenterology 2011;140:1120–1125). However, some intestinal bacteria (eg, B wadsworthia identified in the current study, but also others such as Escherichia coli and Listeria) as well as some protozoa (eg, Giardia) are BA tolerant and their growth may even be favored in the presence of BAs, which conversely can suppress other symbiotic commensals. Notably, some culture media used for microbiological differentiation contain BAs. Taking on the potentially positive side of these findings, one may speculate that BA could also modulate gut microbiota as prebiotics. Importantly, this study reveals that BAs may act as double-edged sword in intestinal inflammation and unravels the mechanisms how certain gut microbes can use dietary-induced changes of bile composition to their own advantage, thereby mechanistically linking diet and intestinal inflammation. Milk fat-induced enrichment in taurine-conjugated BAs is a rich source of organic sulphur, favoring the growth of sulphite-reducing bacteria such as B wadsworthia expressing the sulphite reductase A gene allowing them to generate H2S. Apart from cytotoxic BA effects at high concentrations, bacterial BA-derived side products such as H2S or formation of more hydrophobic secondary BAs can act as gut mucosal barrier breakers, initiating intestinal inflammation. These aspects of BA pathobiology may also need to be considered when using (high-dose) BAs or their derivatives (eg, novel BA-based FXR and TGR5 ligands) for a range of inflammatory and metabolic liver/gastrointestinal disorders, including IBD. Notably, taurine supplementation in patients with cystic fibrosis increased taurine conjugation of BAs and improved steatorrhea without producing relevant side effects (Pediatr Res 1985;19:578–582). The current study is also breaking new ground by demonstrating for the first time a potential link between enteral luminal BAs and IBD pathogenesis. The association between IBD and liver diseases such as primary sclerosing cholangitis (PSC) has always been viewed as a classic example of the gut–liver axis with development of several attractive concepts such as translocation of bacterial products or primed T cell from the inflamed gut to the liver (Liver Int 2012;32:352–369). About 80% of PSC patients have IBD, mostly colitis with peculiar features (eg, right-sided predominance, mild to moderate clinical course) also referred to as PSC–IBD. In light of the findings of the current study, it is intriguing to turn around the concept of the gut–liver axis and develop the provocative alternative hypothesis that liver diseases (eg, PSC) may result in changes of BA composition favoring dysbiosis with outgrowth of specific pathobionts associated with PSC–IBD. This study also raises the question of whether (dietary-induced) changes in other biliary constituents, such as phospholipids and cholesterol, could also contribute to dysbiosis associated with IBD. This article calls for a revival of nearly forgotten or neglected techniques, such as bile sampling/analysis (achieved endoscopically or by string test), or fecal BA analysis. Changes of BA metabolism may not only be induced by exogenous factors such as diet, but also occur with progression of liver diseases—irrespective of their etiology—toward cirrhosis (Arch Intern Med 1999;159:2647–2658). So far, absence/reduced presence of bile have been mainly linked with bacterial translocation in end-stage liver disease. Future studies are needed that consider both qualitative and quantitative changes in biliary BA composition, which could favor outgrowth and translocation of certain pathobionts in disorders of the gut–liver axis. Apart from PSC this may be particularly relevant for nonalcoholic fatty liver disease, where diet-induced changes of hepatic fat content and gut microbiota could represent a double/multiple hit in the progression toward nonalcoholic steatohepatitis (Hepatology 2010;52:1836–1846). The hormonal actions of BAs are critically involved in fine-tuning hepatic glucose and lipid metabolism through activation of FXR and TGR5 (Gastroenterology 2011;140:1120–1125). The current study adds further complexity to the interplay between BAs and lipid metabolism by demonstrating that, conversely, dietary fat is able to modulate BA metabolism. This makes sense from the perspective of fat emulsification, because taurine conjugation of BAs allows a more efficient formation of micelles; however, the underlying molecular mechanisms how dietary fat changes BA metabolism remains elusive. Such potential mechanisms could include modification of the relevant steps in BA metabolism through lipid/fatty acid-regulated nuclear hormone receptors such as the peroxisome proliferator-activated receptor-α (J Lipid Res 2004;45:1051–1060). Inflammatory processes such as sepsis are also associated with profound changes in BA metabolism with massive repression of BA conjugation enzymes (PLoS Med 2012;9:e1001338.). It may be tempting to interpret such changes as self-protective mechanism against outgrowth of taurine-conjugate dependent pathobionts compromising the gut barrier function, changes that could complement the antimicrobial actions of accumulating serum BAs in sepsis. Notably, the effects on B wadsworthia expansion and colitis in the current study were only seen with taurine-conjugated, but not glycine-conjugated, BAs the latter being by far the predominating BA-conjugates in humans. Therefore, it remains unclear whether the rather subtle changes in the ratio between glycin– and taurine–BA conjugates observed under dietary lipid changes in humans (J Clin Invest 1965;44:1754–1765) are sufficient to alter the human gut microbiota. Notably, a similar study design in rats has not observed changes in sulphur metabolizing bacteria (but other genus such as Clostridia; Gastroenterology 2011;141:1773–1781), emphasizing the importance of species differences. It is, however, encouraging that B wadsworthia was also found in the (healthy) human gut. In summary, the present study highlights the key role of BAs in shaping the gut microbiota under the influence of certain (high-fat) diets and opens new horizons for BA-targeted therapeutic strategies and dietary interventions in gut inflammation. These important findings could have a major impact beyond IBD and associated liver diseases, and may also be relevant for a range of metabolic and neoplastic disorders that have been linked with dysbiosis and alterations of BA metabolism within the gut–liver axis, such as obesity, nonalcoholic fatty liver disease, and colorectal cancer. As the most relevant next step, future studies will have to explore potential changes in human biliary/fecal BA composition and the gut microbiota in well-characterized IBD patients as well as controls under evaluation of their current dietary habits. This should help to determine whether these astonishing experimental findings can be translated into clinical practice with potential therapeutic implications, including dietary interventional trials in this field.