Regulation of bile acid metabolism: New insights from inside


  • Potential conflict of interest: Nothing to report.

Sayin SI, Wahlstrom A, Felin J, Jantti S, Marschall HU, Bamberg K, et al. Gut microbiota regulates bile acid metabolism by reducing the levels of tauro-beta-muricholic acid, a naturally occurring FXR antagonist. Cell Metab 2013;17:225-235. (Reprinted with permission).


Bile acids are synthesized from cholesterol in the liver and further metabolized by the gut microbiota into secondary bile acids. Bile acid synthesis is under negative feedback control through activation of the nuclear receptor farnesoid X receptor (FXR) in the ileum and liver. Here we profiled the bile acid composition throughout the enterohepatic system in germfree (GF) and conventionally raised (CONV-R) mice. We confirmed a dramatic reduction in muricholic acid, but not cholic acid, levels in CONV-R mice. Rederivation of Fxr-deficient mice as GF demonstrated that the gut microbiota regulated expression of fibroblast growth factor 15 in the ileum and cholesterol 7a-hydroxylase (CYP7A1) in the liver by FXR-dependent mechanisms. Importantly, we identified tauroconjugated beta- and alpha-muricholic acids as FXR antagonists. These studies suggest that the gut microbiota not only regulates secondary bile acid metabolism but also inhibits bile acid synthesis in the liver by alleviating FXR inhibition in the ileum.


In recent years we have witnessed a tremendous increase in research on the role of gut microbiota (GM) in many aspects of physiology and pathophysiology of vertebrates.[1] The relevance of this topic is reflected in large-scale projects, such as the Human Microbiome Project in North America ( and the MetaHIT project in Europe (, that are searching for connections between GM and multiple conditions spanning from cardiovascular or metabolic diseases such as obesity and diabetes mellitus to behavioral disorders. Studies in both mice and humans are helping to disclose the effects of GM on host physiology through modulation of the metabolism of dietary or endobiotic compounds present in the intestinal lumen.

With regard to liver diseases, GM had also gained renewed attention with major focus in alcoholic and nonalcoholic fatty liver disease as well as cirrhosis.[2, 3] Now, Sayin et al.[4] add to the field providing new data on how GM influences the bile acid (BA) pool size and composition throughout the enterohepatic system in mice. These may be very relevant findings, since BAs are now considered key endobiotic molecules that, as recently disclosed, perform multiple and crucial physiological functions. In fact, BAs seem to be much more than simple detergents that facilitate dietary fat digestion and absorption. Recent evidence supports a regulatory role of BAs in several metabolic pathways related to lipid and sugar handling[5] and show that extrahepatic actions in tissues such as brown adipose tissue or skeletal muscle may influence whole-body metabolism.[6]

Regulation of BA homeostasis is an essential component of liver physiology. Advances in bile research have shown that BA metabolism is governed by complex transcriptional networks within the enterohepatic circulation (EHC) that regulate both BA synthesis and transport in the liver and intestine. BA synthesis involves a complex multistep process that requires the actions of more than a dozen enzymes, most of them belonging to the cytochrome P450s (CYPs) superfamily, that are subjected to a fine and redundant regulation.[7] BA synthesis begins with the formation of 7α-hydroxycholesterol by cholesterol 7α-hydroxylase (CYP7A1), the rate-limiting enzyme of the so-called “classic” pathway, and is followed by several enzymatic steps such as sterol 12α-hydroxylation by sterol 12α-hydroxylase (CYP8B1) that directs BA synthesis to cholic acid (CA). The “alternative” pathway leads to the formation of chenodeoxycholic acid (CDCA) and under normal conditions is a minor pathway. Major advances have been made in recent years regarding the regulation of BA synthesis, particularly in the regulation of CYP7A1 by nuclear receptors such as farnesoid X receptor (FXR), an intracellular BA sensor, and also by ileal-derived molecules such as fibroblast growth factor 15 (FGF15).[7]

GM forms secondary BAs (such as deoxycholic [DCA] and lithocholic acid [LCA]) through a series of reactions including deconjugation, oxidation, and epimerization, thus expanding the chemical diversity of BA pool.[8] Previous work in germfree (GF) rodents have shown that GM, in addition to modulating BA pool composition, also influences BA pool size, with GF animals exhibiting a larger BA pool than conventionally raised (CONV-R) counterparts.[9] The underlying molecular mechanisms to these differences remained unknown. Sayin et al.[4] now provide elegant mechanistic data on how GM influences the BA pool size and composition throughout the EHC. To gain insights into their research question, a comprehensive assessment of BA metabolism including BA pool size determination, profiling of BA composition, and measurement of the expression of hepatic and intestinal genes involved in BA synthesis, metabolism, and transport was carried out in both GF and CONV-R mice. The authors found that colonization of the intestine by GM is associated with a marked (−70%) reduction in the BA pool size in CONV-R mice with respect to GF animals. The underlying mechanisms of this change involve modulation of BA metabolism at several levels (Fig. 1). First, CONV-R mice exhibit a decreased hepatic BA synthesis, which is associated with a reduced expression and activity of Cyp7a1. This is likely related to the inhibitory action of the ileal entero-hormone Fgf15 on Cyp7a1, since the expression of Fgf15 is up-regulated in the distal ileum of CONV-R. Second, a decreased BA reabsorption in the distal ileum and an increased fecal BA excretion also contributed to the reduction of BA pool size in the CONV-R mice group. This finding is explained by a reduced expression of the ileal BA transporter, Asbt (Slc10a2), in this experimental group. Lastly, the authors show that GM strongly influences BA pool composition by specifically decreasing the proportion of tauro-beta-muricholic acid (TβMCA), resulting in a reduced βMCA/CA ratio in CONV-R mice. Of note, livers from the latter experimental group had a 70% decrease in the content of this BA compared with GF counterparts, thus explaining the lower BA pool size in these animals. The authors mechanistically explain their findings showing that antibiotic treatment promoted a marked suppression of Fgf15 expression in the ileum and a corresponding increase of Cyp7a1 expression in the liver, confirming that GM influences BA metabolism through the Fgf15-mediated negative feedback of Cyp7a1. Moreover, they demonstrated that this phenomenon is FXR-dependent using Fxr knockout mice rederived as GF. However, one inconsistency remained. This is that in spite of the fact that CONV-R and GF mice have comparable levels of the FXR agonist CA, they showed different expression of FXR target genes. The authors solve this conflict by showing that muricholic acid derivatives (TαMCA and TβMCA) act as FXR antagonists in both in vivo and in vitro experiments involving ileal stimulation with taurocholic acid. Thus, decreased ileal levels of MCA derivatives in CONV-R mice actually result in increased Fxr activation in the enterocyte due to the alleviation of βMCA-mediated FXR-antagonism.

Figure 1.

Gut microbiota regulation of BA metabolism. A reduction in MCA derivatives content in the enterocyte activates the nuclear receptor Fxr, presumably due to alleviation of MCA-induced receptor antagonism. Fxr activation leads to increased production of Fgf15, which in turn down-regulates Cyp7a1 in the liver and contributes to a reduction in the BA pool size. Down-regulation of the apical bile salt transporter (Asbt), likely due to FXR activation, promotes an increase in fecal BA excretion which further contributes to the BA pool shrinkage. Changes in BA pool size and composition may have metabolic implications.

The results of Sayin et al.[4] are significant since they demonstrate that GM influences BA homeostasis beyond the simple microbial metabolism, also acting as a direct regulator of CYP7A1 through the FGF15 pathway. Additionally, the regulation of the BA pool components by GM and the GM-induced shrinkage of BA pool may also have metabolic implications. Of note, recent findings by Watanabe et al.[10] suggest that a reduced BA pool size may translate into reduced energy expenditure in brown adipose tissue, insulin resistance, and accumulation of triglycerides in the liver under high-fat-diet feeding conditions. Thus, it may be hypothesized that a larger BA pool present in GF mice may contribute to the reported resistance of these mice from diet-induced obesity.[1] In addition, metabolic implications could also result from deactivation of other nuclear receptors having BAs as physiological ligands such as vitamin D receptor, the pregnane-X-receptor, and the constitutive androstane receptor that play a role in a myriad of metabolic pathways.

Another interesting finding is that βMCA is an FXR antagonist. Assuming that FXR antagonism is not beneficial to the hepatocyte, this could represent a harmful side of hydrophilic BA therapy. In this line, FXR-antagonistic effects of very high ursodeoxycholic acid levels has been reported in humans.[11] On these grounds, one could speculate that FXR antagonism may be related to the increased risk of adverse outcomes in patients with primary sclerosing cholangitis treated with high doses of ursodeoxycholic acid.

Finally, it should be kept in mind that these findings cannot be directly extrapolated to humans, since mice and men exhibit multiple differences in biliary physiology.[12, 13] For example, with regard to bile acid pool composition, MCA and its derivatives are exclusively present in rodents, while in humans the BA pool is predominantly constituted by CA. Also, some enzymatic pathways such as rehydroxylation of secondary BAs like deoxycholic acid, which results in its conversion to CA, are not present in human beings. These differences determine marked variations in BA pool hydrophyllicity between both species and may have importance in their response in pathological settings such as cholestasis and increased levels of DCA, since the latter is a powerful activator of a myriad specific cell signaling pathways and receptors (i.e., EGFR, protein kinase C, β-catenin) with potential effects on cells of the EHC.

Finally, although GM in mouse and humans is similar at the division level, the majority of mouse gut species is unique and can account for differences in their influence in BA homeostasis. Thus, further research in human beings is needed to confirm the current findings.

  • Pablo Quintero, Ph.D.

  • Marco Arrese, M.D.

  • Departamento de Gastroenterología

  • Pontificia Universidad Católica de Chile Santiago, Chile