Bile Acids, Microbiota, and Cystic Fibrosis: Channeling Intestinal FXR Signals

Enterohepatic bile acid (BA) homeostasis is coordinated through an intricate series of networks that facilitate exquisite feedback of hepatic BA synthesis and secretion. These networks in turn promote systemic signaling through intestinal enterokines and also directly by BA liganding nuclear hormone receptors, including farnesoid X receptor (FXR) and G protein–coupled receptor-5 (TGR5). The key steps in intestinal BA homeostasis involve active ileal uptake of conjugated BA species via the apical sodium-dependent BA transporter (ASBT), following which intracellular BAs then ligand and activate FXR, which in turn activates transcription of the ileal enterokine fibroblast growth factor (FGF) 19 (FGF15 in mice). After secretion, FGF15/19 enters the portal vein and signals via the hepatic receptor complex FGFR4/βKLOTHO, resulting in transcriptional repression of CYP7A1 and reducing BA synthesis.

Patients with cystic fibrosis exhibit widespread abnormalities including thickened, viscous mucus that promotes intestinal obstruction and dysbiosis as well as BA malabsorption.¹ In this issue of Cellular and Molecular Gastroenterology and Hepatology, Bijvelds et al² have examined some of the underlying mechanisms of intestinal BA homeostasis by using mice with germline Cftr deletion (CF mice). They demonstrate impaired Fxr signaling in CF mice as evidenced by decreased ileal FGF15 and Shp mRNA. Also, although the total BA pool size was unchanged in CF mice, luminal primary BA composition shifted with increased cholic acid and decreased β-muricholic acid without changing abundance of the secondary BA deoxycholic acid, suggesting increased primary BA synthesis rather than altered microbial BA dehydroxylation. They also confirmed dysbiosis in CF mice, demonstrating increased relative abundance of Firmicutes and decreased Bacteroidetes phyla. They further show the functional impact of these taxonomic shifts, with increased ileal expression of mRNAs encoding mucin glycosylation (beta-1,4-galactasyltransferase, fucalpha1-2 fucosyltransferase) as well as mRNAs encoding inflammatory markers, acute phase proteins (Saa1, Saa3), and Toll-like receptor signaling. They then treated CF mice with ciprofloxacin/metronidazole in drinking water and showed that many of those phenotypes were reversed, including reduced expression of inflammatory markers and, of relevance to BA homeostasis, increased ileal FGF15 and Shp mRNA, as well as decreased hepatic Cyp7a1. Those findings suggest that antibiotic treatment restores important elements of BA homeostatic regulation in CF mice. The authors also demonstrated an organ-intrinsic component to some of these signaling events because both ileal explants treated with lipopolysaccharide and ileal organoids treated with a mixture of interferon-γ, tumor necrosis factor-α, and interleukin 1β demonstrated an impaired response to a synthetic FXR ligand (GW4064). Taken together, the findings suggest that altered microbial taxa in CF mice modulate ileal Fxr signaling and contribute to the impairment of systemic BA homeostasis, which is further suggested by the blunted secretion of FGF15 after cholic acid feeding of CF mice. The mechanisms that link these cytokine signals to relative Fxr resistance is unknown but may relate to previously demonstrated nuclear factor κβ binding motifs in several FXR-target genes, which revealed a plausible mechanism for transcriptional inhibition of Fxr expression.³

How do these findings fit into what is known about intestinal FXR signaling, and what questions do these findings raise? First, the findings demonstrate that intestinal BA uptake is impaired without a change in ASBT localization or expression, suggesting that the apical BA transport machinery is intact. This is important because one might predict that the altered mucus glycocalyx might physically impair BA diffusion and lead to reduced uptake. Indeed, this prediction was recently examined in studies where CF mice offered water supplemented with polyethylene glycol (PEG 4000) in drinking water exhibited reduced BA malabsorption and restored Fxr signaling along with a shift in fecal BA species.⁴ The critical question of whether there were accompanying shifts in microbial taxa in those studies was not addressed, and the role of altered mucus composition after antibiotic treatment still remains to be understood. It may be that altering the viscosity or other components of the physical mucous barrier exerts a primary role on microbial colonization, and so it would be of interest to compare microbial taxa from the adherent mucus layer in the distal intestine with that from luminal content, as was performed here.² It is also worth noting that microbial BA metabolism plays an important role in the metabolic outcomes of FXR agonism. For example, intestine-restricted Fxr activation (with fexaramine) alters microbial taxa and increases lithocholic acid production, which then activates TGR5, whereas antibiotic treatment abolished the beneficial metabolic effects of fexaramine-induced Fxr activation in both wild-type and leptin receptor deficient db/db mice.⁵ These findings introduce another level of complexity in predicting the net systemic impact of changes in intestinal Fxr activation.

The findings raise broad implications for both understanding BA homeostasis and the clinical care of CF patients who are routinely exposed to courses of antibiotics. A further translational component to the current studies stems from recent work showing humans with mutations in CFTR exhibit reduced serum levels of FGF19 and elevated serum levels of 7α-hydroxy-4-cholesten-3-one (C4),⁶ a surrogate for 7α-hydroxylase activity. Importantly, those parameters of abnormal BA production were partially reversed in CF patients treated with ivacaftor,⁶ suggesting that even partial correction of the underlying channel defect may help restore BA homeostasis.