Glycochenodeoxycholic acid

The relationship between postprandial bile acid concentration, GLP-1, PYY and ghrelin

Summary

Background Gut hormones peptide YY (PYY) and glucagon-like peptide-1 (GLP-1) play an integral role in appetite control and energy homeostasis. Entero-endocrine L-cells can be stimulated by nutrients and or bile acids to co-secrete PYY and GLP-1. The aim of this study was to determine the response of bile acids, PYY, GLP-1 and ghrelin after a test meal.

Design Twelve subjects with a BMI of 22Æ8 (0Æ52) kg/m2 [mean (SEM)] received a 400 kcal test meal after which blood samples were taken every 30 min from 0 to 180 min. PYY, GLP-1 and ghre- lin were measured by radioimmunoassays. Fractionated bile acids were measured by HPLC-MSMS.

Results PYY positively correlated with glycochenodeoxycholic acid (GCDCA) (rs = 0Æ23, P = 0Æ03) and taurochenodeoxycholic acid (TCDCA) (rs = 0Æ26, P = 0Æ02). GLP-1 positively correlated with GCDCA (rs = 0Æ22, P = 0Æ047) and glycodeoxycholic acid (GDCA) (rs = 0Æ3, P = 0Æ005). Ghrelin negatively correlated with GDCA (rs = –0Æ45, P £ 0Æ0001), TCDCA (rs = –0Æ23, P = 0Æ034) and taurodeoxycholic acid (TDCA) (rs = –0Æ44, P £ 0Æ0001).

Conclusion PYY and GLP-1 responses correlated with chenode- oxycholic acid (CDCA) counterparts, whereas ghrelin negatively correlated with deoxycholic acid (DCA) counterparts. Specific bile acids may thus differentially affect entero-endocrine cells.

Introduction

Bile acids have several endocrine effects and are involved in three main signalling pathways; mitogen-activated protein kinase path- ways,1,2 via the G-protein coupled receptor TGR53,4 and via nuclear hormone receptors such as farnesoid X receptor a (FXRa).5–7 Administration of bile acids to mice increases energy expenditure in brown adipose tissue via induction of the cyclic- AMP-dependent thyroid hormone-activating enzyme type 2 iodothyronine deiodinase (D2).8 In vitro treatment of brown adipocytes and human skeletal myocytes with bile acids increases D2 activity and oxygen consumption.8 In addition, bile acid stim- ulation of FXRa activates a signalling pathway culminating in increased energy expenditure and decreased adiposity.9–11 In humans, circulating bile acid concentrations correlate with mea- sures of insulin sensitivity12, and administration of bile acids has been found to prevent high fat diet-induced obesity and develop- ment of insulin resistance in mice.13 Modulation of bile acid homeostasis using bile acid sequestrants such as cholestyramine and colesevelam has been shown to improve glycaemic control in patients with type 2 diabetes mellitus.14–16 The mechanisms by which bile acids improve glycaemic control are largely thought to be FXRa-independent and are instead mediated by binding to TGR5, leading to cAMP generation and activation of the intracel- lular type 2 thyroid hormone deiodinase.17 Bile acids also act via the phosphoinositide 3-kinase (PI3 kinase)/protein kinase B (AKT) pathway, directly promoting insulin signalling and glyco- gen synthase activation, thus aiding insulin-dependent control of glucose metabolism in the liver.18 In addition, bile acid activation of TGR5 has been found to stimulate glucagon-like peptide-1 (GLP-1) production in vitro, which promotes insulin secretion and thus improves glucose metabolism.

Bile acids constitute a large family of molecules composed of a steroid structure with four rings and a five or eight carbon side chain terminating in a carboxylic acid. Each bile acid has a specific number and orientation of hydroxyl groups (–OH). The principal primary bile acids in humans are cholic acid (CA) and chenodeoxy- cholic acid (CDCA). Ursodeoxycholic acid (UDCA) is the 7b epi- mer of CDCA. Bile acids can be conjugated to either glycine or taurine, forming bile salts. Glycochenodeoxycholic acid (GCDCA) and taurochenodeoxycholic acid (TCDCA) are the glycine and tau- rine conjugates of CDCA, respectively, and glycocholic acid (GCA) and taurocholic acid (TCA) are the glycine and taurine conjugates of CA, respectively. Bile acids can also be conjugated in phase II metabolic reactions such as N-acetylglucosaminidation at the C7 position with a beta-hydroxyl group.19 Bacteria in the intestinal tract can convert primary bile acids into secondary bile acids by removing a hydroxyl group, for example CDCA becomes lithochol- ic acid (LCA) and CA becomes deoxycholic acid (DCA) (Fig. 1).

Gut hormones peptide YY (PYY) and GLP-1 play an integral role in appetite control and energy homeostasis. PYY and GLP-1 are secreted by L-cells of the distal small intestine in response to nutri- ents and bile acids.20–22 Exaggerated responses of PYY and GLP-1 may contribute to the weight loss and improvement in glycaemic control following gastric bypass surgery for the treatment of obesity. The mechanisms for increased concentrations of PYY and GLP-1 after gastric bypass are not entirely understood. The ‘hind- gut hypothesis’, described by Mason,23 proposes that altered anatomy following bariatric surgery accelerates the delivery of nutrient-dense chyme to the distal intestine, which stimulates increased gut hormones secretion. In addition, plasma bile acid concentrations are doubled in individuals after gastric bypass surgery compared to obese individuals who have not had bariatric surgery24, and it is possible that co-delivery of these bile acids within the chyme enhances the stimulation of gut hormone release postgastric bypass.

A role of bile acids in appetite suppression was proposed as early as 1968.25 Several studies indicate that bile acids DCA and TCA combined with fatty acids are more significant than fatty acids alone in stimulating PYY release in the ileum.21,26–29 Bile acid infu- sions of TCDCA and DCA have been shown to increase plasma PYY in rabbits,27,30 and DCA infusion in the human colon stimu- lates PYY release in a dose-dependent manner.21 Importantly, the concentrations of bile acids used in these studies are consistent with the physiological concentrations of bile acids found in the intestinal lumen postprandially,31 supporting a physiological role of bile acids in the stimulation of PYY release. However, not all bile acids have been found to stimulate PYY release with hyodeoxycholic acid and UDCA having no effect on PYY release.27,32 The reason why PYY would be stimulated by specific bile acids is not clear. The postprandial peak concentrations of GLP-1 and total bile acid concentrations inversely correlate with fasting and postprandial glu- cose after gastric bypass.24

Ghrelin is secreted from X/A-like cells of the fundus of the stom- ach33,34 and may play a role in regulating premeal hunger, meal ini- tiation and long-term energy balance.35,36 Circulating ghrelin concentrations increase during fasting and decrease after eating37 while anticipation of meals may contribute to its secretion.38 The mechanisms by which ghrelin promotes hunger and increased energy intake are not entirely clear. There is some evidence to sug- gest that ghrelin may exert its effects by acting on the central ner- vous system.39 Obese subjects have lower fasting plasma ghrelin concentrations than lean controls,40 although appetite is com- monly increased, suggesting that higher body mass is associated with increased gastric responsiveness to ghrelin.41 A relationship between ghrelin and bile acids has not been described.The aim of this study was to investigate the associations between the responses of bile acids, PYY, GLP-1 and ghrelin after a standard 400 kcal test meal.

Methods

All studies were performed according to the principles of the Decla- ration of Helsinki. The Local Research and Ethics Committee at King’s College Hospital approved the study (05/Q0703/4). Written informed consent was obtained.

Subjects and sample collection

We studied 12 healthy male subjects (age 21Æ2 (0Æ46) years, BMI 22Æ8 (0Æ52) kg/m2 [mean (SEM)]. Subjects fasted overnight from 23:00 before ingesting a 400 kcal test meal (46Æ6 g carbohydrate, 28Æ5 g fat and 10Æ4 g protein) at 09:00. EDTA blood samples were taken every 30 min from 0 to 180 min.

Fig. 1 The structure of bile acids. X unconjugated – OH, X tauro-conjugated – NHCH2CH2CH2SO3H, X glyco-conjugated – NHCH2COOH; CA, cholic acid; CDCA, chenodeoxycholic acid; DCA, deoxycholic acid; LCA, lithocholic acid; UDCA, ursodeoxycholic acid.

Bile acid and hormone measurement

Fractionated bile acids were measured using HPLC-MSMS as previ- ously described by Tagliacozzi et al.42 Bile acids analysed were CDCA, DCA, cholic acid (CA), UDCA, LCA and their respective gly- cine (G-) and taurine (T-) conjugates. The method was linear between 0Æ1 and 10 lmol/l for all bile acids and their conjugates with intra-assay coefficient of variations (CV) ranging from 3Æ6% to 8Æ0% and inter-assay CVs ranging from 1Æ5% to 6Æ8%. The lower limit of quantitation was 0Æ1 lmol/l. LCA and UDCA and their conjugates and CDCA were below the detection limit. Glucose was measured using the glucose oxidase method on the Advia 2400 analyser (Sie- mens Healthcare Diagnostics, Frimley, UK). Insulin was measured using a two-site sandwich chemiluminescent immunoassay on the Advia Centaur analyser (Siemens Healthcare). The intra-assay CV was 3Æ3%, and inter-assay CV was 4Æ8%. PYY was measured by an established in-house radioimmunoassay as previously described by Adrian et al.43 The assay measured the biologically active compo- nents, both the full length (PYY1–36) and the fragment (PYY3–36). The detection limit was 5 pmol/l, and the intra- and inter-assay CVs were 5Æ8% and 9Æ8%, respectively. GLP-1 was measured by an estab- lished in-house radioimmunoassay as previously described by Krey- mann et al.44 The detection limit was 7Æ5 pmol/l, and the intra-assay CV was 6Æ1%. Ghrelin was measured by the commercially available Phoenix Pharmaceutical assay kit as previously described by Patter- son et al.45 The assay measured total human ghrelin (acylated and des-acylated), and the intra-assay CV was 8Æ8%.

Statistical analysis

Results are reported as mean (SEM) and correlation co-efficient (rs). Statistical analysis was performed using Analyse-It for Excel
Table 1. Baseline and peak values [mean (SEM)] of glycine- and taurine- conjugated bile acids in plasma following a 400 kcal test meal
(Analyse-It, Leeds, UK). Data were normally distributed, and a Spearman correlation was used. Significance is shown by a P < 0Æ05. Results None of the unconjugated bile acids changed significantly through- out the study period, but all glycine- and taurine-conjugated bile acids and total bile acids increased significantly (Table 1). The pre- dominant bile acid measured was GCDCA, followed by GCA and glycodeoxycholic acid (GDCA). All other bile acids had peak con- centrations of <1Æ00 lmol/l. Table 2 and Fig. 2 show the changes in plasma hormones and glucose before and after the standardized 400 kcal meal. Correlations were sought between individual types of bile acids, PYY, GLP-1 and ghrelin. Bile acids that significantly correlated with PYY were GCDCA (rs = 0Æ23, P = 0Æ03), total glycine-conju- gated bile acids (rs = 0Æ23, P = 0Æ03), TCDCA (rs = 0Æ26, P = 0Æ02). Discussion The CDCA group of bile acids GCDCA and TCDCA correlated with PYY. GCDCA also correlated with GLP-1. GCDCA and TCDCA share a common chemical structure with –OH hydroxyl- ation at C7 and a single hydrogen ion at C12 (Fig. 1). The positive correlation found between GCDCA and PYY and GLP-1 and TCDCA and PYY suggests that the specific chemical structure of bile acids may play a role in stimulating L-cells. UDCA and its con- jugates and CDCA also share this chemical structure, but concen- trations were below the detection limit of our assay, and therefore, we are unable to comment whether a similar correlation is possible with PYY and GLP-1. The DCA group of bile acids GDCA, TDCA and TCDCA corre- lated negatively with ghrelin. DCA counterparts are hydroxylated at both C7 and C12. Bile acids may contribute through an endocrine mechanism to the inhibition of ghrelin secretion from P/D1 cells of the fundus of the stomach postprandially. Ghrelin and cholecysto- kinin (CCK) are negatively correlated.46 CCK is a hormone released from the I-cells of the small intestine in response to lipid ingestion which stimulates the release of bile from gallbladder and also acts as a satiety signal. It is possible that increased CCK levels postprandial- ly stimulate increased bile acid secretion which may in turn inhibit ghrelin synthesis through neural or endocrine mechanisms. Limitations of our study included the homogenous nature of our volunteers, and therefore, the data can not be extrapolated to women or older people. The detection limit of the assay was insuffi- cient for determination of a correlation with UDCA, its conjugates and CDCA. The number of subjects was too small to draw firm conclusions, but the data suggested that the total bile acid and GCDCA response may have been biphasic in 9 of 12 subjects (Fig. 4). It is possible that with 30 min sampling, the peaks and troughs may have been missed in all subjects. The biphasic response of bile acids in some of our subjects may also have accounted for the low rs values of our correlations. Conclusion This study suggests that postprandial conjugated bile acids correlate with appetite regulating gut peptides PYY, glucagon-like pep- tide-1 and ghrelin. The postprandial responses of the chenodeoxycholic acid counterparts primarily correlated with pep- tide YY and glucagon-like peptide-1 responses, whereas the post- prandial deoxycholic acid counterparts correlated with changes in ghrelin. Specific bile acids may play a part in entero-endocrine cell stimulation or inhibition. Targeted experiments, specifically with bile acids hydroxylated at C7, may now be considered to determine the therapeutic value of these bile acids in reducing weight through stimulation of peptide YY and glucagon-like peptide-1 and inhibition of ghrelin. Fig. 4 Line graphs demonstrating the response of (a) glycochenodeoxy- cholic acid (GCDCA) and (b) total bile acids following a 400 kcal test meal in nine subjects. Expressed as mean (SEM).