Glycodeoxycholic Acid Inhibits Primary Bile Acid Synthesis With Minor Effects on Glucose and Lipid Homeostasis in Humans

Abstract

Background

Bile acids play vital roles in control of lipid, glucose, and energy metabolism by activating Takeda G protein-coupled receptor 5 and Farnesoid X receptor, the latter promoting production of the endocrine-acting fibroblast growth factor 19 (FGF19). Short-term administration of single bile acids has been reported to enhance plasma levels of GLP-1 and to enhance energy expenditure. However, prolonged bile acid supplementation (eg, of chenodeoxycholic acid for gallstone dissolution) has been reported to have adverse effects.

Study design

In this proof-of-concept study, we assessed the safety and metabolic effects of oral glycine-conjugated deoxycholic acid (GDCA) administration at 10 mg/kg/day using regular and slow-release capsules (mimicking physiological bile acid release) over 30 days in 2 groups of each 10 healthy lean men, respectively.

Main findings

GDCA increased postprandial total bile acid and FGF19 concentrations while suppressing those of the primary bile acids chenodeoxycholic acid and cholic acid. Plasma levels of 7α-hydroxy-4-cholesten-3-one were reduced, indicating repressed hepatic bile acid synthesis. There were minimal effects on indices of lipid, glucose, and energy metabolism. No serious adverse events were reported during GDCA administration in either capsule types, although 50% of participants showed mild increases in plasma levels of liver transaminases and 80% (regular capsules) and 50% (slow-release capsules) of participants experienced gastrointestinal adverse events.

Conclusion

GDCA administration leads to elevated FGF19 levels and effectively inhibits primary bile acid synthesis, supporting therapy compliance and its effectiveness. However, effects on lipid, glucose, and energy metabolism were minimal, indicating that expanding the pool of this relatively hydrophobic bile acid does not impact energy metabolism in healthy subjects.

Bile acids act as fat emulsifiers in the small intestine and are the major end-products of cholesterol catabolism (1). Bile acids also function as (postprandial) signalling molecules involved in control of lipid, glucose, and energy metabolism and . therefore, bile acid signalling pathways may be interesting targets for treatment of cardiometabolic disorders such as hypercholesterolemia and type 2 diabetes mellitus (2, 3).

The primary bile acids cholic acid (CA) and chenodeoxycholic acid (CDCA) are synthesized from cholesterol in the liver, subsequently conjugated to glycine (G) or taurine (T), and a major part is stored in the gallbladder until meal-induced intestinal release (4). In the gut, the primary bile acids can be deconjugated by the microbiome and converted to the secondary bile acids deoxycholic acid (DCA), lithocholic acid, and ursodeoxycholic acid (5). Conjugated bile acids are taken up by ileal active transport and dehydroxylated and deconjugated bile acids via passive absorption in uncharged form throughout the intestinal tract (6). The enterohepatic cycle is completed when bile acids return via the portal vein to the liver where they are resecreted into bile as part of this cycle. Only a minor share (∼5% per cycle) is lost via fecal excretion, representing the major route for cholesterol removal. Approximately 5% of absorbed bile acids escape hepatic uptake and spill over into the systemic circulation, with a characteristic postprandial rise in plasma bile acid concentrations (4, 7).

Postprandial signalling by bile acids is considered to be mainly mediated via activation of the bile acid receptors Takeda G protein-coupled receptor 5 (TGR5) and Farnesoid X receptor (FXR) (1, 2). TGR5 activation results in the intestinal release of the insulinotropic glucagon-like peptide-1 (GLP-1) in rodents (8). FXR activation inhibits bile acid synthesis via intestinal activation of fibroblast growth factor 19 (FGF19) expression and subsequent activation of the hepatic receptor for FGF19, or directly via effects of liver FXR. Apart from inhibiting bile acid synthesis, animal studies indicate that FGF19 increases energy expenditure and hepatic lipid oxidation, improves insulin sensitivity, and may reduce lipogenesis (9-12). Furthermore, the presence of bile acid receptors in brown and white fat, adrenals, and brain suggests physiological roles of (postprandial) bile acid signalling at these sites (13, 14).

Agonizing TGR5 and FXR has beneficial metabolic effects in preclinical models (8, 10). In recent years, a number of specific synthetic FXR and TGR5 agonists have been developed with varying results. For example, the FXR agonist, Obeticholic acid, a semisynthetic bile acid that improves liver fibrosis and key components in nonalcoholic steatohepatitis, whereas Compound SB-756050, a proven TGR5 agonist, had no significant effects on glucose metabolism (15, 16). Individual bile acid species have different binding affinities for TGR5 and FXR (2, 3). The secondary bile acid DCA has agonistic properties for both TGR5 and FXR (2, 3, 17). Interestingly, DCA administration lowered plasma cholesterol in humans, which has been suggested to represent a FXR-mediated effect (18-23). In an earlier study, we showed that a single dose of orally administered glycine conjugated DCA (GDCA) increased postprandial GLP-1 excursions in healthy male subjects, along with lowered postprandial glucose excursion implicating bile acid-induced TGR5 activation in the small intestine (24) and hence suggests potential for the treatment of insulin resistance and type 2 diabetes mellitus.

Safety issues and adverse side effects (eg, elevated liver transaminases and gastrointestinal complaints) occur frequently after prolonged bile acid supplementation as seen in treatment for gallstones (25). When side effects occur, dose reduction of oral bile acids is warranted. A possible manner to prevent these adverse effects is to administer bile acids in a conjugated form because conjugation to glycine or taurine lowers pKa and hence passive reabsorption and reduces bile acid cytotoxicity (1, 4). In this proof-of-concept study, we investigated the safety and metabolic effects of oral GDCA administration in regular and slow-release capsules (ie, mimic physiological bile acid release) for 30 days in 2 groups of each 10 healthy lean men, respectively.

Methods and Materials

Participants

Twenty healthy lean men finished the study (Fig. 1). Inclusion criteria were body mass index (BMI) between 18.5 and 25.0 kg/m² or BMI between 25.0 and 30 kg/m² and waist circumference between 79 and 94 cm (ie, lean waist circumference), aged 18 years or older at the time of signing the informed consent, and a Homeostatic Model Assessment of Insulin Resistance (HOMA) index ≤2.0 (calculated as fasting glucose [mmol/L] × fasting insulin [pmol/L]/135). Exclusion criteria were use of any medication that potentially influences bile acid metabolism, cholecystectomy, gastrointestinal diseases, consumption of three or more units of alcohol per day, tobacco use, drugs abuse, participation in an intervention study within 3 months before this study, and weight decrease or increase above 10% in the previous 3 months. Oral and written informed consent was obtained from all participants before the start of the study and was in agreement with the principles of the Declaration of Helsinki (2013). The study was approved by the Medical Ethics Committee of the Academic Medical Center, Amsterdam, The Netherlands (METC number 2017_133). The study was registered at the Netherlands Trial Register (https://trialsearch.who.int; NL6526).

Study Design

The aim of this study was to assess the metabolic effect and feasibility of 10 mg/kg/day GDCA administration (bile salt: sodium glycodeoxycholate) in regular hard gelatin capsules (Capsugel, Lonza USA, New Jersey, USA) (hereafter GDCA-R) and slow-release capsules (DR capsules, Capsugel, Lonza USA, New Jersey, USA) (hereafter GDCA-S) for 30 days once daily in the morning. The dosage of the GDCA supplementation was based on our previous study (24) and the literature on bile acid treatment. A synopsis of these studies can be found as Supplementary Table S1 (26). Twenty participants were assigned to 2 groups: 10 participants received 10 mg/kg/day GDCA in regular capsules for 30 days and the other 10 participants received 10 mg/kg/day GDCA in slow-release capsules for 30 days. We investigated the metabolic effects after 15 and 31 days (±2 days range in test days) using a mixed meal test design.

Study Safety

An independent data safety monitoring board defined stop criteria before the enrollment of the first participant and reviewed all (serious) adverse events (AEs) that occurred over the course of this study.

For the assignment to GDCA-R or GDCA-S, we followed a predefined scheme including 3 steps to monitor safety. First, 5 participants were assigned to GDCA-R. When no serious adverse events (SAEs) occurred, 5 participants were assigned to GDCA-S. Finally, after SAE evaluation, the final 10 participants were assigned to GDCA-R or GDCA-S in a fixed order.

We monitored liver tests (amino alanine transferase [ALT], aspartate amino transferase [AST], gamma glutamyl transpeptidase [GGT], alkaline phosphatase, and total bilirubin at 7, 15, and 31 days after the start of GDCA administration.

A step-down procedure was implemented in case of elevations of liver function tests. If liver parameters increased between 2 and 4 times of the upper reference limit, the GDCA dose was to be reduced to 5 mg/kg/day. Moreover, monitoring of liver tests was intensified to 2 times per week during treatment. If liver tests were increased to 4 times the upper reference limit, the GDCA administration was immediately discontinued and study participation stopped. The step-down procedure was designed after careful discussion with the data safety monitoring board. Four to 6 weeks after the last study day, the investigators called the participants to check whether (S)AEs occurred after GDCA discontinuation.

Mixed Meal Test Protocol

After an overnight fast, participants reported at 08:00 Am at the Experimental and Clinical Research Unit of the Amsterdam University Medical Centres, location Academic Medical Center, Amsterdam, The Netherlands. Three days before the study days, participants were asked to refrain from alcohol use and strenuous physical exercise (defined as >1 hour per day). Body composition was measured with whole-body air displacement plethysmography (Bodpod, Cosmed, Rome, Italy). Fasting resting energy expenditure (REE) and the respiratory quotient (RQ) were assessed with indirect calorimetry (Quark, Cosmed) using a ventilated hood system (27). For venous blood withdrawal, a cannula was placed into an antecubital vein. Blood samples were taken just before meal ingestion at time point 0 and 15, 30, 45, 60, 90, 120, 150, 180, 240, and 300 minutes after ingestion of the standardized liquid mixed meal (Nutricia Compact, Nutricia, Zoetermeer, The Netherlands) containing all the macronutrients (16% protein, 35% fat, and 49% carbohydrates). The amount of administered energy was 25% of fasting REE (measured at day 1). Indirect calorimetry was repeated postprandially at time point 120 and 240. Beta-cell insulin sensitivity was evaluated with HOMA of B-cell function (HOMA-B) (28). Participants collected 24-hour stool samples for the analyses of fecal bile acid and cholesterol concentrations. Plasma and feces were stored at −80 °C until analyses.

Laboratory Analyses

Plasma total bile acid concentrations were determined with an enzymatic method (Diazyme Laboratories) and individual plasma and fecal bile acids were measured by liquid chromatography mass spectrometry as described (29). FGF19 (RRID:AB_355750) and 7α-hydroxy-4-cholesten-3-one (C4, liquid chromatography mass spectrometry) were determined with in-house developed assays (30). Total cholesterol, high-density lipoprotein cholesterol (HDL-C), and triglycerides (TGs) were measured with commercial assays (Diasys and WAKO) on the Selectra autoanalyzer (Sopachem, Ochten, The Netherlands). Low-density lipoprotein cholesterol (LDL-C) concentrations were calculated with the Friedewald formula (31). Apolipoprotein (Apo) A1 (RRID:AB_3096880), ApoB (RRID:AB_3096879), and ApoE (RRID:AB_3096881) were measured with a commercial nephelometric assay on the Selectra autoanalyzer. Bedside plasma glucose concentrations were measured with the glucose oxidation method (EKF Diagnostics, Barleben/Magdeburg, Germany). Insulin concentrations were measured with the Atellica IM (RRID:AB_2941780). Total GLP-1 (intact plus the primary metabolite) was measured by specific radioimmunoassay as previously described (RRID:AB_2892202) (32). The 24-hour stool samples, used for the analysis of fecal bile acid and cholesterol concentrations (33), were homogenized using a bag mixer (P. bagmixer, Interscience “Ermergo” jumbomix, service number 2401079) and diluted with distilled water (Milli-Q) in a 1:1 ratio. Fecal cholesterol concentrations were quantified as the sum of coprostanol, epicoprostanol, and cholesterol.

Statistical Analyses

Data are presented as mean ± SEM. The distribution of the data was assessed with the Shapiro-Wilk test. Differences in body composition or fasting parameters between baseline and day 15 or 31 were analyzed with 1-way repeated measures ANOVA, with a Bonferroni correction for adjusted P values for multiple comparisons if the data were normally distributed. Otherwise the Friedman test was used with the Dunn test as correction for adjusted P values of multiple comparisons. To assess differences in postprandial responses (including multiple time points) between the baseline and day 15 or 31, we used mixed-effects models with multiple comparisons (Bonferroni correction applied). The mixed meal tests were set as fixed effect, whereas the different time points and participant numbers were set as random effects. We excluded the postprandial data from day 31 of 1 participant in the GDCA-R group because he did not consume the complete liquid mixed meal as a result of dyspepsia. SPSS Statistics 26 (IBM, Armonk, New York, United States) and GraphPad Prism (GraphPad Software Inc., La Jolla, California, United States) were used for statistical analyses and graph design respectively. A P value ≤.05 was considered as statistically significant.

Results

Clinical Characteristics

Twenty healthy lean men completed the study (Fig. 1, Table 1). Twenty-six participants were recruited for the study. Five participants did not meet inclusion criteria. One participant was excluded because of a rise in ALT >4 times the upper limit. In the GDCA-R group, 8 participants completed 10 mg/kg/day for 30 days, whereas 2 participants completed the study with the dose reduction to 5 mg/kg/day for 30 days. In the GDCA-S group, 9 participants completed 10 mg/kg/day for 30 days, whereas 1 participant completed the study with the dose reduction to 5 mg/kg/day for 30 days.

Participant weights remained stable during treatment in both groups, and GDCA administration did not alter fat or fat-free mass distribution (Supplementary Fig. S1) (26).

GDCA Administration Inhibits Bile Acid Synthesis in the Fasting State

GDCA-R did not alter fasting total bile acid concentrations (Table 2). Regarding bile acid composition, GDCA-R increased fasting GDCA concentrations, whereas those of the primary bile acids CA (unconjugated and glycine-conjugated forms) and CDCA (unconjugated form) decreased (Supplementary Table S2) (26). Fasting FGF19 concentrations were transiently increased at day 15 but were indistinguishable from baseline, at day 31 (Table 2). Fasting C4, a marker for bile acid synthesis (via the predominant classical pathway), was decreased at day 15 and remained decreased until day 31 (Table 2).

In contrast to GDCA-R, GDCA-S administration increased fasting total bile acid concentrations at day 31 (Table 2). Concerning individual bile acids, GDCA-S increased all forms of DCA in the fasting state (Supplementary Table S2) (26). There were only minor changes in concentrations of the other individual bile acids (Supplementary Table S2) (26). Comparable to GDCA-R, fasting FGF19 concentrations were elevated after GDCA-S only at day 15, and this was accompanied by decreased fasting C4 concentrations (Table 2), which partially normalized by day 31.

GDCA Increases Plasma FGF19 Levels and Inhibits Postprandial Primary Bile Acid Synthesis

In the postprandial state, GDCA-R led to augmented plasma total bile acid responses on both day 15 and day 31, as depicted in Fig. 2A. GDCA concentrations 3 hours after the mixed meal test (ie, T180) at day 31 increased, whereas DCA and TDCA levels remained unchanged. The primary bile acids CA and GCDCA showed slightly decreased concentrations at day 15, whereas those of other bile acids remained stable (Supplementary Table S3) (26).

GDCA-R administration increased postprandial FGF19 excursions assessed at both days 15 and 31 (Fig. 2B). The increase in FGF19 was accompanied by decreased C4 concentrations (Fig. 2C).

GDCA-S increased postprandial total bile acid responses after days 15 and 31 (Fig. 2D), similar to GDCA-R. The postprandial total concentration of glycine-conjugated bile acids increased after 15 days, whereas those of total unconjugated and taurine-conjugated bile acids remained unchanged (Supplementary Table S3) (26). Concerning individual bile acids, GDCA-S increased all forms of DCA in the postprandial state at days 15 and 31 (Supplementary Table S3) (26). The concentrations of glycine-conjugated forms of the primary bile acids CA and CDCA decreased, and GDCA-S also lowered TCA (Supplementary Table S3) (26). Postprandial FGF19 responses were increased at both days 15 and 31 (Fig. 2E), whereas postprandial C4 concentrations decreased on those days (Fig. 2F).

GDCA Administration Results in Only Minor Changes in Fecal Bile Acid and Total Cholesterol Concentrations

GDCA-R administration did not significantly impact on the total concentrations of fecal bile acids or cholesterol, as shown in Table 2. There was an increase in fecal GDCA concentrations at day 31 (Supplementary Table S4) (26). Furthermore, GDCA-R administration resulted in minor changes in individual fecal bile acid profiles, including decreased levels of GCDCA and GCA (Supplementary Table S4) (26).

Similar to GDCA-R, GDCA-S administration had no substantial effect on the total excretion of fecal bile acids. However, there was a transient increase in cholesterol concentrations noted at day 15 (Table 2). Additionally, we observed increased fecal excretion of GDCA, DCA, and TDCA (Supplementary Table S4) (26). Notably, the excretion of unconjugated and glycine-conjugated primary bile acids, CDCA and CA, decreased (Supplementary Table S4) (26).

GDCA Administration Prompts Minor Alterations in Plasma Lipids Composition in the Fasting and Postprandial State

GDCA-R administration decreased fasting plasma total cholesterol, LDL-C, and ApoB concentrations (Table 2). Fasting HDL-C, ApoA1, ApoE, and TG concentrations remained unchanged (Table 2). GDCA-R decreased postprandial plasma responses of total cholesterol, LDL-C, and ApoB at both days 15 and 31 (Supplementary Fig. S2B-D) (26). Postprandial TGs, HDL-C, ApoA1, and TG concentrations were unchanged (Supplementary Fig. S2A and E, F) (26).

In contrast to GDCA-R, GDCA-S only decreased plasma fasting ApoE concentrations at day 31, whereas cholesterol, lipoproteins, and TGs remained unchanged (Table 2). In the postprandial state, plasma concentrations of total cholesterol, LDL-C, and ApoB were decreased at day 31, whereas HDL-C and ApoA1 were transiently decreased at day 15 (Supplementary Fig. S3C-D) (26), and HDL-C was decreased at days 15 and 31 (Supplementary Fig. S3E-F) (26). TGs did not change (Supplementary Fig. S3A) (26).

GDCA-S Administration Increases Fasting and Postprandial GLP-1 Concentrations

GDCA-R increased fasting plasma glucose concentrations at days 15 and 31, whereas GDCA-R did not alter fasting insulin concentrations (Table 2). Consequently, GDCA-R increased Homeostatic Model Assessment of Insulin Resistance at day 15 but not at day 31. HOMA-B was not affected (Table 2). Postprandial glucose and insulin excursions were unchanged (Fig. 3A-B). Fasting and postprandial GLP-1 concentrations were unaffected by GDCA (Fig. 3C).

GDCA-S administration had no effects on fasting or postprandial glucose or insulin concentrations (Table 2, Fig. 3D-E). However, GDCA-S administration increased fasting and postprandial GLP-1 concentrations (Table 1, Fig. 3F).

GDCA Administration has Transient Effects on Resting Energy Expenditure in the Postprandial State

GDCA-R significantly increased fasting and postprandial REE at day 15, but these effects were not sustained after 31 days (Table 2, Supplementary Fig. S4A) (26). We did not observe any effects on fasting or postprandial RQs, fat oxidation, or carbohydrate oxidation (Table 2, Supplementary Fig. S4B-D) (26).

In contrast, GDCA-S decreased fasting and postprandial REE at day 15, but this effect was not present at day 31 (Table 2, Supplementary Fig. S5A) (26). No significant effects were noted on fasting or postprandial RQs or substrate oxidation (Table 2, Supplementary Fig. S5B-D) (26).

GDCA Administration Caused No Serious Adverse Events

Table 3 provides an overview of the reported AEs with GDCA-R (GDCA in regular capsules) and GDCA-S (GDCA in slow-release capsules) administration.

No SAEs occurred during GDCA-R administration. One participant was excluded after 3 weeks because of elevated ALT (>4 times the upper limit) and was replaced. Two participants required a dose reduction to 5 mg/kg of GDCA-R because of AEs: 1 had persistent diarrhea for 3 weeks and the other had elevated liver transaminases. Both AEs resolved after dose reduction. The most frequently reported AEs with GDCA-R were transient gastrointestinal complaints and elevated liver transaminases. Five participants observed changes in stool patterns, whereas 2 had persistent diarrhea. Five participants had reversible ALT elevations and 4 had increased AST. During the course of GDCA-R ingestion, ALT significantly increased after 15 and 31 days, whereas AST showed a tendency to increase (Supplementary Fig. S6) (26). Total bilirubin, AP, and GGT concentrations did not exceed upper reference limits during the study period (data not shown).

In the GDCA-S group, no SAEs occurred. One participant required a dose reduction because of elevated ALT concentrations after 2 weeks of administration. Only 2 participants experienced transient gastrointestinal complaints (Table 2). Five participants had elevated ALT and 3 had elevated AST (Table 2). During the GDCA-S course, ALT increased, whereas AST showed a tendency to increase (Supplementary Fig. S6) (26). Total bilirubin, AP, and GGT were not affected (data not shown).

All AEs resolved after discontinuation of GDCA-R and GDCA-S.

Discussion

This is the first proof-of-concept study to demonstrate the feasibility of daily GDCA administration in both regular and slow-release capsules for 30 days in healthy lean men. Despite anticipated effects on plasma FGF19 levels and primary bile acid synthesis, only small effects were observed on clinically relevant metabolic parameters reflecting lipid and glucose metabolism and energy expenditure.

A strength of this study is the ability to confirm therapy compliance by the increased FGF19 and concomitantly reduced C4 levels. C4 is a bile acid synthesis intermediate and reflects accurately the rate of hepatic bile acid production (34). FXR activation inhibits bile acid synthesis either directly in the liver via the SHP-LRH1 pathway or indirectly via intestinal induction of FGF19 expression (9). Our study demonstrates that GDCA administration in both regular and slow-release capsules inhibited primary bile acid synthesis (ie, reduces plasma C4 concentrations) in the fasting and postprandial states. GDCA administration in regular capsules decreased the contributions of the primary bile acids CA and CDCA, which are synthesized from cholesterol in the liver to total plasma bile acids (1, 4). However, GDCA administration in the slow-release capsules had less effect on plasma concentrations of these primary bile acids. There was no difference in concentration of C4, suggesting a similar lack of synthesis inhibition. It might be possible that the GDCA-S administration, presumably followed by delayed bacterial dehydroxylation and more distal absorption, caused this difference, although differential effects on gut microbial conversion of bile acids are not excluded (35).

The administration of GDCA in regular capsules decreased plasma total cholesterol concentrations. This reduction was predominantly attributed to a decrease in LDL-C levels (∼19% reduction). This may be explained by the fact that FXR inhibits the hepatic expression of ApoB as suggested by previous research (36). Indeed, ApoB levels were found to be reduced following GDCA administration, indicating an effect of the activation of FXR by bile acids on lipoprotein production by the liver or the small intestine.

In contrast to the decreased LDL-C concentrations in our study, previous projects involving the use of FXR agonists such as obeticholic acid (a synthetic FXR agonist) and NGM282 (an FGF19 analogue) demonstrated an increase in LDL-C concentrations in humans as a result of inhibited bile acid synthesis (ie, reduced conversion of cholesterol to bile acids and diminished removal from the liver, followed by homeostatic down-regulation of LDL receptor activity) (37, 38). However, bile acid synthesis was also inhibited in our study as evidenced by low C4 concentrations.

The question arises why GDCA administration failed to influence lipid, glucose, and energy homeostasis in our study. A previous study with CDCA (15 mg/kg/day for 2 days) led to increased energy expenditure in women (39). This suggests that bile acid administration has the potential to modulate energy metabolism. The absence of a response to GDCA administration might point a gender specific effect. Moreover, it could be argued that this may be due to an insufficient dose, but, intestinal FXR activation as monitored by the increase of FGF19 was clearly observed (9). The lack of additional effect on lipid and glycemic control suggests that under “normal” control conditions hydrophobic bile acids do not significantly control energy metabolism in humans. Studies in preclinical models also show no or limited effects in healthy animals but do so in obese or insulin-resistant states (40).

GDCA treatment with a dose of 10 mg/kg/day in regular and slow-release capsules for 30 days proved to be safe, as evidenced by the absence of SAE during administration or following discontinuation. However, it is noteworthy that some participants exhibited increased liver transaminase levels and experienced gastrointestinal complaints. In contrast to our hypothesis, the glycine conjugated form of DCA did not mitigate these AEs and, in fact, induced gastrointestinal complaints similar to administration of unconjugated DCA, as previously reported in studies (18, 22, 23, 41). The elevated liver transaminases observed in our study (50% in each group) align with findings of Einarsson et al. (41), who observed similar elevations (50%) in AST and ALT after administration of unconjugated DCA in the same dose. One plausible explanation is that the GDCA dose of 10 mg/kg/day may have been too high for certain individuals, as evidenced by the normalization of both increased liver transaminases and stool patterns following a dose reduction to 5 mg/kg/day in these cases (N = 3). Another explanation is that a substantial portion of ingested GDCA may undergo immediate deconjugation in the proximal intestinal tract (4). Consequently, this process could lead to relatively elevated DCA concentrations within the enterohepatic cycle, resulting in a bile acid pool with higher hydrophobicity, as reported in prior research (4, 17). Interestingly, by now, we had treated a patient with a primary bile acid synthesis disorder for more than 4 years successfully with GDCA (unpublished data that are currently prepared for submission), which resulted in near normalization of plasma bile acid levels.

Our proof-of-concept study has some limitations. First, a limited number of healthy male participants was included, which may be too small to detect prominent metabolic effects, and it is unclear whether our findings can be translated to women (42). We performed this study in healthy volunteers to understand the (patho)physiological response to prolonged bile acid supplementation in subjects with an intact enterohepatic cycle. Therefore, we can only speculate on the effects of GDCA in patients with the metabolic syndrome or type 2 diabetes mellitus. However, because of the elevated liver enzymes we observed in combination with no discernible effect on bile acid levels, we doubt if we should investigate the effects in patients with the metabolic syndrome or type 2 diabetes mellitus. These patients often have metabolic-associated liver disease, which puts them at risk for liver injury (43). Additionally, this study lacked the statistical power required to make a comparative analyses of the effects between the regular and slow-release capsules. We did not include a placebo arm, which is another limitation of this study. Nevertheless, this paper represents the first study of extended dosing of glycine-conjugated DCA, and consequently, because of safety reasons, our ability to include participants was restrained. This study proves that it is difficult to modulate bile acid levels and ensuing biological effects in humans with an intact enterohepatic cycle. Finally, we did not measure the newly identified bile acid conjugations (ie, phenylalanocholic acid, tyrosocholic acid, and leucocholic acid) in plasma or feces, which were recently published by Quinn et al (44). Nevertheless, taking all the data together, our results indicate that 2 to 4 weeks administration of a hydrophobic bile acid does not influence human physiology significantly.

Conclusions

Our proof-of-concept study shows that GDCA administration effectively leads to elevated FGF19 levels and inhibits primary bile acid synthesis, indicating both therapy compliance and its effectiveness. However, the effects on lipid, glucose, and energy metabolism were minimal, indicating that under healthy conditions hydrophobic bile acids do not significantly control energy metabolism in humans. Apparently, this may be related to the fact that it is difficult to modulate peripheral bile acid levels and obtain ensuing biological effects in humans with an intact enterohepatic cycle.

GDCA administration, regardless of regular of slow-release capsules, was found to be feasible. However, it was associated with elevated liver transaminases in 50% of the participants and a significant part of the participants reported gastrointestinal discomfort. Given these side effects, it is crucial to recognize the limitations and potential challenges with this approach to modulate bile acid metabolism. It is essential to consider alternative treatments or modifications to GDCA administration that can mitigate these side effects. This may involve refining doses, delivery methods, or identifying patient subgroups that may be better suited to this treatment.

Funding

E.C.E.M., S.M., and M.R.S. are funded by ZorgOnderzoek Nederland (ZON) en het gebied Medische Wetenschappen (MW) van NWO ZonMW/Diabetes fonds [grant number 95105011]. M.N. is supported by a personal Nederlandse Organisatie voor Wetenschappelijk Onderzoek VICI grant 2020 [09150182010020]. J.J.H. is funded by Novo Nordisk Foundation Center for Basic Metabolic Research [grant number NNF 18CC0034900].

Author Contributions

E.C.E.M. designed the study, performed clinical experiments, did statistical analysis, and wrote and edited the manuscript. S.M. performed clinical experiments and edited the manuscript. U.A. and S.W.O.D. executed laboratory analysis and reviewed the manuscript. J.A.R. designed the study and reviewed the manuscript. J.J.H. and B.H. executed laboratory analyses and reviewed the manuscript. F.K. executed laboratory analyses and reviewed the manuscript. M.N. reviewed the manuscript. F.G.S. executed laboratory analyses and reviewed the manuscript. A.K. executed laboratory analyses and reviewed the manuscript. E.M.K. and M.R.S. designed the study, reviewed and edited the manuscript. M.R.S. is guarantor of the article.

Disclosures

J.J.H. is a member of advisory boards for NovoNordisk. M.N. is a scientific advisory board member of Caelus Pharmaceuticals and Advanced Microbiome Interventions, the Netherlands. E.C.E.M., S.M., U.A., S.W.O.D., J.A.R., B.H., F.K., F.G.S., A.K.G., E.M.K. and M.R.S. have no conflicts of interest.

Data Availability

Restrictions apply to the availability of some or all data generated or analyzed during this study to preserve patient confidentiality or because they were used under license. The corresponding author will on request detail the restrictions and any conditions under which access to some data may be provided.

Clinical Trial Information

The study was registered at the International Clinical Trial Registry Platform (https://trialsearch.who.int; NL6526).

References

Abbreviations

 
• AE

  adverse event

 
• ALT

  amino alanine transferase

 
• AST

  aspartate aminotransferase

 
• Apo

  apolipoprotein

 
• BMI

  body mass index

 
• C4

  7α-hydroxy-4-cholesten-3-one

 
• CA

  cholic acid

 
• CDCA

  chenodeoxycholic acid

 
• DCA

  deoxycholic acid

 
• FGF19

  fibroblast growth factor 19

 
• FXR

  Farnesoid X receptor

 
• G

  glycine

 
• GDCA

  glycodeoxycholic acid

 
• GDCA-R

  glycodeoxycholic acid in regular capsules

 
• GDCA-S

  glycodeoxycholic acid in slow-release capsules

 
• GGT

  gamma glutamyl transpeptidase

 
• GLP-1

  glucagon-like peptide 1

 
• HDL-C

  high-density lipoprotein cholesterol

 
• HOMA-B

  Homeostatic Model Assessment of B-cell function

 
• LDL-C

  low-density lipoprotein cholesterol

 
• REE

  resting energy expenditure

 
• RQ

  respiratory quotient

 
• SAE

  serious adverse event

 
• TGR5

  Takeda G protein-coupled receptor 5

 
• TG

  triglyceride