The remarkable ability of bile acids to solubilize polar lipids during digestion has generally been considered the sole function of conjugated bile acids in the small intestine. Work during the past decade (see below) has suggested that luminal conjugated bile acids have a second function: to inhibit the growth of bacteria in the small intestine. In a recent issue of PNAS, Inagaki et al. (4) present strong evidence for a previously undescribed mechanism by which conjugated bile acids mediate their antimicrobial effects in the distal small intestine. They show here that conjugated bile acids regulate expression of host genes whose products promote innate defense against luminal bacteria.

The human small intestine is relatively devoid of microbes under normal conditions (10⁴ to 10⁵ colony-forming units/ml) and has a high conjugated bile acid concentration, averaging 10 mM during digestion. In liver cirrhosis in both humans (5) and animals (6), bile acid secretion is decreased and bacterial overgrowth occurs. In animals, bile duct ligation also leads to bacterial overgrowth in the small intestine (7, 8). These observations, plus studies showing that bile and unconjugated bile acids inhibit bacterial growth in vitro (9, 10), led to the hypothesis that the high concentration of conjugated bile acids in the small intestinal lumen is an important factor in the paucity of microbes in the proximal small intestine. Apparent proof for this hypothesis came when it was shown that the feeding of bile or conjugated bile acids in conditions of bile acid deficiency in the intestine abolished bacterial overgrowth and reduced bacterial translocation to intestinal lymph nodes (6, 8). Unfortunately, in vitro studies (11) suggest that the antimicrobial effect of conjugated bile acids is quite weak when compared with that of unconjugated bile acids and cast doubt on the validity of the effect being mediated solely by conjugated bile acids. A possible explanation for this paradox is that the antimicrobial effect of administered conjugated bile acids may be comediated by fatty acids (partly present as soaps) that are associated with the conjugated bile acids in mixed micelles in the proximal small intestine. Long-chain fatty acids have been known for many decades to have potent antimicrobial effects (12, 13). Nonetheless, irrespective of the mechanism, these experiments provided strong evidence for a second physiological function of conjugated bile acids in the proximal small intestine: to prevent bacterial overgrowth by their antimicrobial activity.

Inagaki et al. (4) present compelling evidence, based on studies in mice, that the antibacterial effect of conjugated bile acids in the distal small intestine is mediated by a cellular pathway involving the nuclear receptor farnesoid X receptor (FXR), an orphan receptor that is activated by conjugated bile acids. Activation of FXR by conjugated bile acids induced the expression of genes whose products prevent bacterial overgrowth and promote epithelial integrity. The authors first determined that intestinal FXR mRNA levels were three times higher in the ileal epithelium, where bile acids are absorbed, than in the epithelium of the proximal small intestine. To test which genes are regulated by FXR in the ileum, they administered GW4064, a potent FXR agonist developed by Glaxo Wellcome (14). Using microarray analysis of ileal RNA, several genes with potential functions in mucosal defense were found to be up-regulated by GW4064, including inducible nitric oxide synthetase (iNOS), whose enzymatic product, nitric oxide, is well known to possess direct antimicrobial activity.

The authors then performed bile duct ligation to determine whether such up-regulation was associated with suppression of bacterial overgrowth in vivo. As anticipated, bile duct ligation in WT mice caused an >10-fold increase in aerobic bacteria and a doubling of anaerobic bacteria in ileal and cecal contents. It also caused bacterial invasion of the intestinal mucosa and increased aerobic bacterial translocation to mesenteric lymph nodes. These notable effects of bile duct ligation were abolished by oral administration of GW4064 in WT mice but not in mice genetically deficient in FXR (developed by Frank Gonzalez and his colleagues at the National Institutes of Health; ref. 14). Thus, the bacterial overgrowth and bacterial translocation that had been attributed to a conjugated bile acid deficiency in older studies were corrected by GW4064 despite conjugated bile acid levels in the small intestine being negligible because of bile duct ligation. Based on these results, the authors propose that conjugated bile acids activate FXR and that such activation, in turn, induces the expression of gene products that promote antimicrobial defense and epithelial integrity in the distal small intestine.

Missing in the experimental design of this paper was an examination of the effects of conjugated bile acid administration. If bile acids affect microbial levels in the distal small intestine through only FXR-dependent mechanisms, one would expect that their administration, like that of GW4064, would have no effect in FXR-deficient mice. Practically, such experiments may be confounded by severe toxicity, and even death, because administration of conjugated bile acids to bile duct-ligated mice or FXR-deficient mice results in bile acid accumulation in the hepatocyte with consequent induction of apoptosis/necrosis (15). The paper also did not examine proximal intestinal content, so it is not known whether GW4064 influenced the presumed bacterial overgrowth in the duodenum and jejunum.

It must be further noted that although the identification of FXR-regulated genes by microarray analysis suggests their involvement in antibacterial intestinal defense, such data are merely correlative and not necessarily causal. Indeed, mice lacking iNOS are not known to have bacterial overgrowth in the small intestine (16), even though they are far more susceptible to infections by enteric bacterial pathogens (17). Thus, up-regulation of iNOS may not be solely responsible for the observed powerful effects of FXR activation. The particular microarray used in the work covered only ?50% of all murine genes, and only a single time point was examined, leaving open the possibility that other FXR target genes are expressed in the small intestine and responsible for the effects of GW4064. Finally, FXR may mediate antimicrobial functions independent of changes in gene expression as probed by microarray analysis.

Bile acid concentrations fall to low, submicellar levels in the terminal ileum because of active bile acid absorption (18), and fatty acids are unlikely to be present as fatty acid absorption occurs in the proximal small intestine. Reflux of bacteria-rich cecal content across the ileocecal valve results in continuous bacterial seeding of the terminal ileum. A defense mechanism that inhibits luminal bacterial growth based on the carrier-mediated entry of conjugated bile acids into the ileal enterocyte should be more robust than one based on the luminal concentration of bile acids or fatty acids, because these substances are present in very low concentrations in the ileal lumen. If this reasoning is correct, conjugated bile acids, probably together with fatty acids, inhibit bacterial growth in the proximal small intestine directly by their pharmacological properties and in the distal small intestine indirectly by their signaling properties. This concept is illustrated in Fig. 1.

Multiple factors contribute to the low bacterial content of the human small intestine (19). These factors include gastric acidity that kills ingested organisms, IgA that is secreted in bile and by the intestinal epithelium, antimicrobial peptides (e.g., defensins) that are secreted by Paneth and other epithelial cells, and propulsive intestinal motility. Long-chain fatty acids solubilized in conjugated bile acid-mixed micelles also may play a role in the proximal small intestine as discussed. Inagaki et al. (4) describe the secretion in the distal small intestine of as-yet-unidentified microbial agents evoked by conjugated bile acids activating FXR in the ileal enterocyte. These different mechanisms are likely to act synergistically in controlling microbial levels in the small intestine

In addition to their surfactant properties that render them powerful solubilizers and bacteriostatic agents, bile acids are now known to signal a variety of systems in the liver and intestine by interaction with multiple nuclear receptors (20). This seminal paper from the laboratory of Steven Kliewer describes an important signaling function for conjugated bile acids in the distal small intestine. It also confirms the power of genetic techniques and the availability of a nuclear receptor agonist to elucidate unexpected and previously undescribed physiological functions in vertebrates. Future studies should elucidate the relative roles of conjugated bile acids and/or fatty acids in promoting antibacterial defenses in the small intestine and define which gene products resulting from bile acid-induced FXR activation promote such defenses in distal intestinal content.