Introduction

Bile acids are amphipathic steroid molecules synthesized in the liver from cholesterol. Bile acids were originally thought to simply aid in the digestion and absorption of dietary lipid in the small intestine [1]; however, work over the past several decades demonstrates that bile acid metabolism and signaling are key contributors to metabolic regulation and thus are promising therapeutic targets for metabolic diseases [2]. This review will summarize our current understanding of the role of bile acid metabolism and signaling in glucose regulation and will identify the knowledge gaps that need to be addressed to enable successful targeting of bile acid signaling and metabolism for diabetes treatment.

Changes in circulating bile acid profile have been implicated in the pathogenesis of insulin resistance and type 2 diabetes (T2D) [3]. Different bile acid subtypes exhibit varying degrees of hydrophobicity which is determined by factors such as state of ionization and by the number, position, and orientation of hydroxyl groups [4]. The relative amounts of hydrophobic versus hydrophilic bile acids determine the overall hydrophobicity of the bile acid pool [4]. T2D is associated with an increase in the hydrophobicity of the circulating bile acid pool in humans [3]. Consistent with this, hydrophilic bile acid subtypes, such as tauroursodeoxycholic acid (TUDCA), have been shown to protect against inflammation and improve insulin sensitivity in rodent models and patients with T2D [5,6,7]. In contrast, hydrophobic bile acid subtypes, such as deoxycholic acid (DCA), have been shown to promote inflammation and endoplasmic reticulum stress that are associated with impaired glucose regulation [8,9,10]. These data suggest that altering bile acid profiles may be an effective approach for the treatment of T2D.

Sterol 12-α-hydroxylase (CYP8B1) is a bile acid synthetic enzyme expressed primarily in hepatocytes [1]. CYP8B1 is required for the synthesis of 12-α-hydroxylated bile acids and thereby determines systemic bile acid profiles [1]. Genetic ablation of Cyp8b1 decreases bile acid profile hydrophobicity and protects against several metabolic diseases, including obesity and T2D, in mice [11,12,13,14,15]. Furthermore, genetic ablation of Cyp8b1 in ApoE−/− mice reduces development of atherosclerotic plaque [14]. Moreover, diabetic cholesterol-fed Cyp8b1−/− mice are protected against hypercholesterolemia and cholelithiasis [15]. Finally, Cyp8b1 null mice are resistant to fatty liver development [13]. These findings suggest that Cyp8b1 may be an effective target for the treatment of T2D and many of its associated comorbidities.

Several different bile acid receptors have been implicated in metabolic regulation, including the nuclear receptor farnesoid X receptor (FXR) and the transmembrane G protein-coupled receptor, TGR5. Activation of these receptors regulates bile acid, cholesterol, lipid, and glucose metabolism and plays a role in various pathological processes, including inflammation, fibrosis, and carcinogenesis. While the mechanisms by which these receptors regulate metabolic homeostasis are incompletely understood, the impact of FXR and TGR5 is likely mediated, at least in part, via bile acid metabolism [16•, 17•, 18].

Key Steps in Bile Acid Metabolism

Primary bile acids are synthesized in the liver and then converted into secondary bile acids through interactions with the gut microbiota. The biosynthesis of primary bile acids in the liver involves a series of enzymatic reactions in which the cholesterol ring is modified; the side chain is shortened and conjugated [19] with either glycine or taurine [20]. Bile acids are stored in the gallbladder and secreted into the gastrointestinal tract in response to feeding. Primary bile acids are deconjugated and dehydroxylated in the distal intestinal lumen by gut microbes to generate the secondary bile acids, deoxycholic acid (DCA) and lithocholic acid (LCA) [21]. The most abundant primary bile acids in humans are cholic acid (CA) and chenodeoxycholic acid (CDCA). In rodents, the majority of CDCA is converted to muricholic acid (MCA) [22]. While a primary bile acid subtype in bears [23], UDCA is derived from CDCA in humans and rodents through gut microbial modifications [24, 25]. Detailed descriptions of the enzymes and intermediates involved in bile acid metabolism can be found in several excellent review articles [1, 19]. We provide a summary of this complex process below.

Bile acid synthesis can occur through two pathways: the classic (neutral) and alternative (acidic) pathways. The classic pathway, which occurs in the liver, accounts for the majority of bile acid synthesis [26]. Cholesterol 7-α-hydroxylase (CYP7A1) is the rate-limiting enzyme in this pathway, while CYP8B1 determines bile acid profile [27]. In the alternative pathway, cholesterol is oxidized by sterol-27-hydroxylase (CYP27A1), the rate-limiting enzyme in this pathway, followed by 7-α-hydroxylation of the oxysterol intermediates by oxysterol 7-α-hydroxylase (CYP7B1) [28, 29]. CYP27A1 and CYP7B1 are expressed in the liver and in various extrahepatic sites including vascular endothelium and macrophages [30,31,32,33,34]. Bile acid synthesis is tightly controlled to ensure the maintenance of a healthy total bile acid pool, with approximately 95% of bile acids being recycled. FXR is considered a primary regulator of this homeostatic process; however, research increasingly supports a role for TGR5 as well.

Regulation of Glucose Homeostasis and Hepatic Bile Acid Metabolism by FXR

FXR is a key regulator of hepatic bile acid metabolism [35] that plays a role in a various disease processes, including inflammatory bowel disease, colorectal cancer, and T2D [36,37,38]. FXR, which is highly expressed in the liver and gastrointestinal tract, can be activated by both free and conjugated bile acids, with CDCA having the highest affinity. The order of potency of bile acid subtypes for FXR is CDCA>LCA=DCA>CA [39, 40]. However, hydrophilic bile acids, such as UDCA and MCA, cannot activate FXR [41], and tauro-conjugated β- and α-MCA may actually serve as naturally occurring FXR antagonists [42,43,44]. More recent work in humans treated with UDCA for 3 weeks prior to bariatric surgery suggested that UDCA might also exert antagonistic effects on FXR [45].

FXR is a key regulator of glucose and lipid metabolism [36, 46,47,48,49,50,51]. FXR improves glycemic control by stimulating insulin secretion from pancreatic β-cells [52], enhancing adipocyte insulin sensitivity [46] and inhibiting hepatic gluconeogenesis (Fig. 1) [53]. However, genetic ablation of FXR and FXR antagonism improve glucose regulation and liver lipid deposition [54, 55]. Furthermore, while Trabelsi et al. report that FXR inhibits secretion of the incretin hormone, glucagon-like peptide-1 (GLP-1), from intestinal L cells [56], Pathak et al. report that L cell FXR signaling increases GLP-1 secretion [17•]. Together, these conflicting reports on the role of FXR in glucose regulation suggest that further work is needed to fully understand the glucoregulatory function of FXR.

Fig. 1
figure 1

Glucoregulatory effects of FXR and TGR5. FXR-mediated glucoregulatory effects in various tissues are illustrated with purple arrows and TGR5-mediated pathways are illustrated with green arrows. Farnesoid X receptor (FXR), transmembrane G-coupled protein receptor (TGR5), glucagon-like peptide-1 (GLP-1)

Full size image

Bile acid-mediated activation of FXR suppresses bile acid synthesis in a homeostatic feedback loop. Specifically, hepatocyte FXR activation upregulates small heterodimer partner (SHP), which inhibits the transcription factors hepatic nuclear factor 4α (HNF4α) and liver receptor homolog-1 (LRH-1), thus reducing their binding to the bile acid response element in the Cyp7a1 and Cyp8b1 gene promoters and inhibiting transcription of these genes [57,58,59,60,61,62,63,64,65,66]. Activation of FXR in the intestine induces fibroblast growth factor 15 (FGF15 or human orthologue FGF19) secretion, ultimately activating hepatocyte FGF receptor 4 (FGFR4) and mitogen-activated protein kinase (MAPK)/extracellular receptor kinase 1/2 (ERK1/2) [67,68,69,70]. Activation of FXR downregulates Cyp8b1 expression by upregulating SHP and FGF15/19 levels [71, 72]. However, work in tissue-specific FXR knockout mice demonstrates that treatment with an FXR agonist in mice with loss of intestinal FXR, but intact hepatocyte FXR, results in decreased Cyp8b1 expression, but not Cyp7a1 expression [69, 73•]. In contrast, treatment with an FXR agonist in mice with loss of hepatic FXR, but intact intestinal FXR, results in downregulation of both Cyp8b1 and Cyp7a1 expressions [69, 73•]. These data suggest that hepatic FXR signaling is important for suppressing Cyp8b1 expression, while intestinal FXR signaling is critical for downregulating both Cyp7a1 and Cyp8b1 expressions [69, 73•].

While FXR signaling through SHP and FGF15/19 is a critical contributor to FXR-mediated regulation of bile acid metabolism, several recent studies reveal new FXR-mediated pathways (Fig. 2, dashed lines), suggesting that there is still much to learn about FXR regulation of bile acid metabolism. For example, MAFG has been identified as an FXR target gene that transcriptionally represses Cyp8b1 expression in mice [74]. While FXR is thought to primarily exert its effects on Cyp7a1 and Cyp8b1 expressions through transcriptional pathways, recent studies have identified important FXR-dependent post-transcriptional regulators. In particular, Tarling et al. reported that FXR activation upregulates the expression of the RNA-binding protein, ZFP36L1, to rapidly degrade Cyp7a1 mRNA [75•]. ZFP36L1-dependent regulation of bile acid metabolism may also contribute to obesity, as hepatocyte-specific Zfg36l1−/− mice are resistant to diet-induced obesity [75•]. Moreover, microRNAs (miRs), such as miR-144 [76] and miR-122a [77], may be downstream mediators of FXR action. For example, Li et al. found that miR-33a is induced in a mouse model of elevated hepatic bile acid synthesis (mice overexpressing Cyp7a1) to act as a homeostatic regulator inhibiting Cyp7a1 and Cyp8b1 mRNA expressions in response to elevated hepatic bile acid synthesis [78]. Moreover, FXR activation with GW4064 induces expression of miR-122a [77], a liver-specific microRNA. MiR-122a overexpression decreases Cyp7a1, but not Cyp8b1, mRNA levels [77]. The identification of microRNAs that regulate bile acid metabolism provides a potentially viable approach to target bile acid metabolism for the treatment of T2D, as microRNAs have already been successfully applied to the treatment of liver disease in humans [79, 80]. Overall, understanding the post-transcriptional pathways involved in regulation of bile acid metabolism will be critical in rationale design of bile acid-based pharmaceutical strategies for treating metabolic disease [81,82,83].

Fig. 2
figure 2

Established and emerging pathways involved in the regulation of Cyp7a1 and Cyp8b1 expressions by TGR5 and FXR. Established pathways are illustrated with solid arrows, and newly identified regulators of Cyp7a1 and Cyp8b1 expression are illustrated with dashed arrows. Regulators of Cyp7a1 expression are illustrated with blue arrows and regulators of Cyp8b1 expression are illustrated with red arrows. Farnesoid X receptor (FXR), transmembrane G-coupled protein receptor (TGR5), Cholesterol 7-α hydroxylase (Cyp7a1), sterol 12-alpha-hydroxylase (Cyp8b1), sterol 27 hydroxylase (Cyp27a1), oxysterol 7-α hydroxylase (Cyp7b1), cholic acid (CA), chenodeoxycholic acid (CDCA), deoxycholic acid (DCA), lithocholic acid (LCA), fibroblast growth factor 19 (FGF19), FGF receptor 4 (FGFR4), small heterodimer partner (SHP)

Full size image

FXR shows promise as a target for the treatment of metabolic and inflammatory disorders including T2D, primary biliary cirrhosis, nonalcoholic fatty liver disease (NAFLD), and nonalcoholic steatohepatitis [84,85,86,87,88]. In fact, an FXR agonist, obeticholic acid (Ocaliva©), has been approved by the Food and Drug Administration (FDA) to treat the rare liver disease, primary biliary cholangitis. However, FXR activation suppresses both Cyp7a1 and Cyp8b1 expressions in tandem, which can have detrimental effects on lipid metabolism as Cyp7a1 is the rate-limiting enzyme in bile acid synthesis. Specifically, inhibition of Cyp7a1 expression decreases the conversion of cholesterol to bile acids, resulting in lipid dysregulation [89, 90]. Therefore, FXR signaling has been reported to produce off-target effects, such as dyslipidemia [54, 91, 92]. For example, mice with targeted deletion of Cyp7a1 develop hypercholesterolemia and have an increased risk of developing atherosclerosis [89]. A marked decrease in bile acid synthesis and decreased Cyp7a1 expression are also associated with dyslipidemia and cholelithiasis [93, 94]. Thus, an improved understanding of the selective regulation of Cyp8b1 expression may lead to the development of bile acid-based therapeutics with fewer adverse side effects.

Regulation of Glucose Homeostasis and Hepatic Bile Acid Metabolism by TGR5

TGR5 is a transmembrane G protein-coupled receptor [95] that is ubiquitously expressed throughout the body, including endocrine glands, adipocytes, muscle, liver, and the gastrointestinal tract [96,97,98]. In the liver, TGR5 is present on liver sinusoidal endothelial cells [99], cholangiocytes [100], biliary epithelial cells [101], and Kupffer cells [102]. It has been speculated that TGR5 is not expressed on hepatocytes; however, Yang et al. reported that TGR5 is expressed in a human hepatocellular carcinoma cell line [103], and recent work reveals that TGR5 is present in canine hepatocytes [104].

The binding affinity of various bile acids to TGR5 differs as compared to FXR [95, 96]. Unlike FXR, hydrophobic bile acids have the highest affinity for TGR5, with the following rank order of potency: LCA>DCA>CDC>CA>UDCA. TGR5 activation by bile acids and synthetic agonists results in activation of the adenylyl cyclase, which in turn activates protein kinase A (PKA) signaling pathways [96]. Moreover, cell-specific signaling pathways are also activated [99, 102, 105].

Increased TGR5 signaling improves glucose regulation through several tissue-specific effects [106, 107]. TGR5 signaling in gastrointestinal enteroendocrine L cells enhances secretion of GLP-1 [107]. Recent work reveals that L cell TGR5 is located on the basolateral L cell membrane [108, 109]. TGR5 signaling in immune cells decreases inflammatory cytokine secretion and TGR5 signaling on adipocytes increases energy expenditure by promoting the beiging of white adipose tissue in mice [110,111,112]. Therefore, TGR5 may exert its metabolic effects, in part, by decreasing inflammation and promoting mitochondrial biogenesis [96, 113, 114]. However, TGR5 activation leads to unwanted side effects, including pruritus [115, 116], cholesterol gallstone formation [117], and cholestasis [118, 119]. Together, these data suggest that identifying and targeting downstream effectors of TGR5 may be a more effective approach with less risk for adverse side effects than directly targeting TGR5.

Several studies point to an important role for TGR5 in bile acid metabolism and the maintenance of a healthy bile acid profile [17•, 120]. Indeed, Pean et al. reported that genetic ablation of TGR5 leads to increased hydrophobicity of the biliary, circulating, and hepatic bile acid pools and revealed a protective effect of TGR5 during liver regeneration in mice [18]. Donepudi et al. report similar shifts in the gallbladder bile acid profile in Tgr5−/− compared to Tgr5+/+ mice in free-fed and fasted conditions [120], in parallel with decreased Cyp7b1 and Cyp27a1 mRNA levels. Thus, TGR5 can upregulate the alternative pathway of bile acid synthesis, which likely contributes to TGR5-dependent decreases in bile acid profile hydrophobicity [120]. Pathak et al. reported that TGR5 is a downstream target of FXR that is required for the effect of L cell FXR signaling to promote GLP-1 secretion [17•]. Specifically, this study identified an FXR response element on the human TGR5 gene promoter, which is highly conserved in the mouse Tgr5 gene [17•]. Thus, interactions between TGR5 and FXR may have implications for hepatic bile acid metabolism; however, further work is needed to comprehensively define the underlying molecular mechanisms.

Role of Bile Acids in Glucoregulatory Improvement After Bariatric Surgery

Bariatric surgery, including Roux-en-Y gastric bypass (RYGB) and vertical sleeve gastrectomy (VSG), is the most effective long-term treatment for obesity and results in high rates of T2D remission, even prior to weight loss [121, 122]. One well-established mechanism by which bariatric surgery improves glucose regulation is by increasing bile acid signaling [16•, 123, 124]. A reoccurring finding between vastly different bariatric procedures is increased circulating bile acid concentrations, in both humans and rodent models [16•, 124,125,126,127,128,129,130]. Therefore, bariatric surgery provides a useful model system with which to understand bile acid metabolism and how it may be targeted for T2D treatment.

The mechanisms by which bariatric surgery increases circulating bile acid concentrations remain incompletely defined, but appear to be dependent on anatomic re-arrangement of the gastrointestinal tract; Roux-en-Y gastric bypass, but not laparoscopic adjustable gastric banding, increases circulating bile acid concentrations in humans [131]. Furthermore, diverting bile to the distal gut increases circulating bile acid concentrations and improves glucose regulation, suggesting an important role for the distal gastrointestinal tract [132, 133]. Finally, studies in rodents have suggested that increased bile acid reabsorption may contribute to increased circulating bile acid concentrations after bariatric surgery. One study reports an increase in ileal apical sodium–bile acid transporter (ASBT) expression after VSG in mice [134]. Other studies report gut hypertrophy after various types of bariatric surgery in rodents, which likely enhances gut absorptive capacity [134,135,136]. However, further work is needed to fully define the mechanisms responsible for increased circulating bile acid concentrations after bariatric surgery.

FXR and TGR5 Contribute to Improved Glucose Regulation After Bariatric Surgery

Several studies in whole body TGR5 and FXR knockout mouse models demonstrate that increased bile acid receptor signaling contributes to the metabolic benefits of bariatric surgery [16•, 123, 124]. Using a whole body FXR knockout mouse model, Ryan et al. reported that FXR contributes to body weight loss and improved glucose regulation after VSG [124]. More recently, TGR5 has been shown to contribute to improved glucose regulation after VSG in mice [16•, 123]. TGR5-dependent improvement in glycemic control after VSG was associated with a TGR5-dependent decrease in the hydrophobicity of the circulating bile acid pool. This beneficial bile acid profile shift was associated with a TGR5-dependent reduction in hepatic CYP8B1 protein expression, with no effect on hepatic CYP7A1 expression [16•]. Ding et al. confirmed that TGR5 contributes to the glucoregulatory benefits of VSG, in part due to increased GLP-1 secretion after VSG [123]. By contrast, McGavigan et al. did not observe a TGR5-dependent increase in GLP-1 secretion after VSG. These discrepancies could be a result of differences in surgical model, strain, and/or the timing of experiments. Notably, all of the published work assessing the role of TGR5 and FXR in the benefits of bariatric surgery has been performed in whole body knockout mouse models which are prone to development of compensatory pathways. Therefore, further work is needed in inducible and tissue-specific knockout models to gain a deeper understanding of the mechanisms by which TGR5 and FXR contribute to improved glucose regulation after bariatric surgery.

In contrast, the role of TGR5 in Roux-en-Y gastric bypass-mediated improvements in glucose regulation is unclear, with one study reporting that TGR5 does not contribute to weight loss or improved glucose regulation after Roux-en-Y gastric bypass in mice [137]. This suggests that VSG and Roux-en-Y gastric bypass activate different signaling pathways to exert their beneficial metabolic effects. However, Zhai et al. observed increased ileal TGR5 expression after Roux-en-Y gastric bypass in mice and present data to suggest that TGR5 contributes to elevated GLP-1 production after Roux-en-Y gastric bypass, although this was not directly assessed in vivo [138].

Overall, bariatric surgery increases bile acid concentrations to increase TGR5 and FXR signaling and improve glucose regulation. However, the exact molecular mechanisms and tissue site(s) of action by which enhanced TGR5 and FXR signaling improve glucose regulation after bariatric surgery are unknown.

Conclusion

Targeting bile acid profile by inhibiting Cyp8b1 expression is a promising therapeutic modality for T2D. The role of FXR in the regulation of bile acid synthesis and Cyp8b1 expression is well documented; however, recent work suggests that TGR5 may also regulate Cyp8b1 expression. Furthermore, recent studies reveal nontraditional pathways in the regulation of bile acid metabolism, including key post-transcriptional mechanisms. Overall, our understanding of the regulation of bile acid metabolism by FXR and TGR5 is rapidly evolving; however, there remain significant gaps in our understanding of these processes. In particular, it will be important to define the pathway(s) by which Cyp8b1 expression is selectively regulated to enable pharmaceutical targeting of this promising therapeutic target.