Intestinally-targeted TGR5 agonists equipped with quaternary ammonium have an improved hypoglycemic effect and reduced gallbladder filling effect

Abstract

TGR5 activation of enteroendocrine cells increases glucagon-like peptide 1 (GLP-1) release, which maintains glycemic homeostasis. However, TGR5 activation in the gallbladder and heart is associated with severe side effects. Therefore, intestinally-targeted TGR5 agonists were suggested as potential hypoglycemic agents with minimal side effects. However, until now no such compounds with robust glucose-lowering effects were reported, especially in diabetic animal models. Herein, we identify a TGR5 agonist, 26a, which was proven to be intestinally-targeted through pharmacokinetic studies. 26a was used as a tool drug to verify the intestinally-targeted strategy. 26a displayed a robust and long-lasting hypoglycemic effect in ob/ob mice (once a day dosing (QD) and 18-day treatment) owing to sustained stimulation of GLP-1 secretion, which suggested that robust hypoglycemic effect could be achieved with activation of TGR5 in intestine alone. However, the gallbladder filling effect of 26a was rather complicated. Although the gallbladder filling effect of 26a was decreased in mice after once a day dosing, this side effect was still not eliminated. To solve the problem above, several research strategies were raised for further optimization.

agonists equipped with quaternary ammonium have an improved

R

A

P

Hua Cao, Zhi-Xiang Chen,, KaiWang, Meng-Meng Ning, Qing-An Zou, Ying Feng, Yang-LiangYe, Ying Leng & Jian-Hua Shen

ob/ob

TGR5 (Takeda G-protein-coupled receptor 5), also known as GPBAR1, M-BAR, or GPCR19, was identied rst as a G protein-coupled receptor responsive to bile acids (BAs) in 20021,2. It shows: high expression in the gallbladder; moderate expression in the intestine, spleen and placenta; and low expression in the lung, brown adipose tissue (BAT), skeletal muscle, and brain35. TGR5 activation in enteroendocrine cells6 increases the release of GLP-1 which maintains homeostasis of blood glucose by promoting glucose-induced insulin secretion, suppressing glucagon release, delaying gastric emptying, promoting satiety, and increasing glucose disposal in the peripheral tissues7,8. In brown adipose tissue and skeletal muscle TGR5 mediates energy expenditure through a BATGR5 cAMPD2 signaling pathway9. Therefore, TGR5 activation provides a promising strategy for treatment of type 2 diabetes mellitus and associated metabolic disorders1012. Thus TGR5 has drawn considerable attention from both academia and industry1318.

However, TGR5 activation in other tissues can cause some side eects, of which those in the gallbladder and heart are the main concerns. Assays in mice have revealed that TGR5 activation in the epithelium of the gallbladder by administration of either bile acids derivatives (e. g. INT-777, 1 developed by Intercept Pharmaceuticals, Fig.1) or synthetic small molecule TGR5 agonist (e. g. 2 developed by our team, Fig.1) causes smooth-muscle relaxation, prevents bile secretion, and greatly increases gallbladder volume14,19,20. Several absorbed TGR5 agonists have been shown to change heart rate and blood pressure in dogs15,21,22. Therefore, it was suggested that localized activation of TGR5 within the intestinal tract while avoiding systemic exposure (i. e. intestinally-targeted) could be a promising anti-diabetes mellitus strategy with minimal side eects23,24. While no intestinally-targeted

State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica (SIMM), Chinese Academy of

SCIENTIFIC REPORTS

1

www.nature.com/scientificreports/

Figure 1. Structures of several TGR5 agonists and their EC50 (concentration for 50% of maximal eect) values on human TGR5 (hTGR5).

Figure 2. Design strategy of intestinal-targeted TGR5 agonists.

TGR5 agonists with robust activity were reported, especially in a diabetic model, there was still doubt about the validity of this strategy. The rst concern was whether robust hypoglycemic efficacy could be achieved by TGR5 activation in the intestine alone without additional eects in the brown adipose tissue or skeletal muscle. The second concern was whether the possible side eects in gallbladder and heart could be eliminated by low systemic drug concentration. Our research team once found a PEG8 compound (3, Fig.1) with low systemic exposure, and thus its gallbladder lling eect was reduced. It displayed a moderate hypoglycemic efficacy in normal mice (ICR (Institute of Cancer Research) mice)25; however no signicant eect in diabetic model ob/ob mice was observed.

It was apparent that the side eect in heart could be minimized successfully as the drug concentration was decreased in plasma decreases. Therefore, although the gallbladder lling eect is more challenging, it becomes the main focus of our study and key parameter in drug discovery of TGR5 agonist.

Quaternary ammonium is present widely in bile acid sequestrants (BASs) such as cholestyramine (4, Fig.2), colesevelam and colestilan26. BASs can bind to BAs in the intestine and act as cholesterol-lowering polymer drugs27. BASs are barely absorbed owing to their high molecular weight and positive charge28. Recent studies have revealed that BASs can improve glycemic control through induction of energy expenditure, enhance glucose utilization and indirect activation of TGR529,30. Quaternary ammonium plays an important part in binding to BAs (so as to improve glycemic control) and the non-absorbed prole. Besides, quaternary ammonium was considered

SCIENTIFIC REPORTS

2

www.nature.com/scientificreports/

A to B B to A

Efflux ratio

Papp (106cm/s) Papp (106cm/s)2 6.75 5.58 0.8 9a 0.55 0.43 0.8 26a 0.06 0.8 14.1

Table 1. Apparent Permeability (Papp) in Caco-2 Cells of 2, 9a, 26a. A to B indicates the experiment from apical to basolateral side, B to A indicates the experiment from basolateral to apical side.

Compd

Figure 3. OGTT of 26a (100mg/kg) and 2 (50mg/kg) in ICR mice. (A) Blood glucose concentration aer administration of 26a and 2; (B) blood glucose AUC0120 min aer administration of 26a and 2. Compound 26a, 2, and 0.25% CMC (carboxymethyl cellulose sodium, control) were administered (p.o.) to ICR mice (n=78)1.5h before oral glucose loading (4g/kg). Blood glucose levels were measured berefore and aer glucose loading. *p<0.05, **p<0.01 vs. control. Error bar indicateds SEM (standard error of the mean).

by Exelixis and Searle (now Pzer) in the development of low absorbed (intestinally targeted) drugs3133. Up to

now, only dimer TGR5 agonist 3 with large molecular weight was reported to be low-absorbed, and no other approaches of reducing gallbladder lling eect were reported. We rst introduced the quaternary ammonium of BASs to TGR5 agonist. We hypothesized that this approach would increase their hypoglycemic efficacy and greatly decrease systemic exposure to minimize the risk of gallbladder toxicity.

Herein we report the validation of the intestinally-targeted strategy with our newly found low-absorbed TGR5 agonist 26a.

Results

Medicinal chemistry work (including design, structure-activity relationships and biological assays in vitro) is available in Supplementary Information. The design strategy is illustrated in Fig.2. By replacing the pyridine ring of 2 (an absorbed TGR5 agonist, human TGR5 EC50 (concentration for 50% of maximal eect) = 1.5 nM, mouse TGR5 EC50 = 14 nM) with a thiophene ring, we obtained 9a with best activity in vitro (human TGR5 EC50 = 0.55 nM, mouse TGR5 EC50 = 2.8 nM). To our delight, 9a (Papp = 0.55 106cm/s, Table1) showed decreased Caco-2 cell permeability compared with 2 (Papp = 6.75 106cm/s). Quaternary ammonium was incorporated to 9a to yield a series of TGR5 agonists, among which 26a (human TGR5 EC50 = 4.1 nM,

mouse TGR5 EC50 = 0.71 nM) displayed the best activity in vitro and extremely low Caco-2 cell permeability (Papp=0.06106cm/s). 26a could be classied as a low-permeability agent.

26a met our primary design strategy: low permeability in cell membranes, and highly potent TGR5 activity. Whats more, 26a exhibited low cLogP=2.30 (vs. cLogP=5.5 for 2, calculated by ChemBioDraw Ultra 12.0) and high solubility in water (1.8mg/ml vs.441ng/ml for 2, determined by LC-MS/MS). In addition, activation of the farnesyl X receptor (FXR) was not observed up to 100M of 26a, thus suggested that 26a was a highly selective TGR5 agonist. To investigate whether 26a with a low-permeability prole could display a robust hypoglycemic eect and a reduced gallbladder lling eect, once a day oral dosing (QD) and long-term eect of 26a in animal models were determined. A pharmacokinetic prole in vivo was also determined to reveal the absorption and distribution of 26a.

The activity of 26a was investigated using an oral glucose tolerance test (OGTT) in ICR (Institute of Cancer Research) mice. Oral administration of 26a and 2 resulted in a similar and signicant decrease in blood glucose (Fig.3A,B); the area under the glucose levels vs. time curve (AUC) was decreased by 18% and 16% for 26a and 2, respectively, compared with that of the control group.

Next we evaluated the eect of 26a in gallbladder lling in ICR mice (Fig.4A,B). Aer once a day oral dose of 2 (50mg/kg) to ICR mice, gallbladder area and bile weight increased 143% and 108%, respectively, compared with the control group. Encouragingly, the gallbladder lling eect was decreased in the group treated with 26a

SCIENTIFIC REPORTS

3

www.nature.com/scientificreports/

Figure 4. (A) Relative gallbladder area and (B) relative bile weight aer once a day oral dose. Aer the OGTT experiment, mice were refed for 3h, the gallbladder was removed and the area measured using a vernier caliper. The relative gallbladder area was calculated from the length multiplied by the width of the gallbladder. Bile weight was measured using analytical balances. *p<0.05; **p<0.01; ***p<0.001. Error bar indicates SEM.

Figure 5. Eects of 26a on blood glucose levels in ob/ob mice. Compounds 2 (50mg/kg), 26a (100mg/kg) and 0.25% CMC (control) were administered (p.o.) to 2h-fasted ob/ob mice (male, n=8), and blood glucose measured before dosing or 2, 4, 6, 8, 10, and 24h aer dosing. Mice were refed at 6h aer dosing. *p<0.05, **p<0.01 vs. control. Error bar indicates SEM.

(100 mg/kg), as gallbladder area increased 83% compared with the control group, while bile weight increased approximately 58% compared with the control group, which was not a signicant dierence.

ob/ob Mice. To further assess efficacy, the hypoglycemic eect of 26a was tested at once a day oral dose (QD) to ob/ob mice (a genetic type 2 diabetes model with impairment of leptin production). 26a (100 mg/kg, Fig.5) displayed a more robust glucose-lowering eect compared with 2 (50 mg/kg), and this eect was maintained until at least 24h post-dosing.

A GLP-1 secretion assay of 26a in ob/ob mice was consistent with the robust and long-lasting hypoglycemic eect. Persistent stimulation of GLP-1 was found up to the 24 h of our experiment (Fig.6): GLP-1 stimulation using 26a was 2.8-, 3.9-, and 4.9- fold at 6, 12, and 24 h, respectively, over that observed in the control group. These data conrmed that the hypoglycemic eect of 26a was reliant mainly on the eect of GLP-1 stimulation from enteroendocrine cells. In addition, the GLP-1 stimulation eect of 26a was enhanced considerably by linagliptin (a dipeptidyl peptidase (DPP)-4 inhibitor that prevents GLP-1 degradation). Total eect of GLP-1 stimulation was 8.3-, 24-, and 35- fold at 6, 12, and 24h, respectively, over that in the control group. These results highlighted the considerable therapeutic potential of combining TGR5 agonist 26a with a DPP-4 inhibitor in treatment of type 2 diabetes.

The robust hypoglycemic eect of 26a in mice was consistent with our design purpose. Nevertheless, whether improvement in activity was owing to the BAs binding eect of quaternary ammonium was not known. In a subsequent 18-day treatment of 26a in ob/ob mice (Part 8 of the Supplementary Information), a signicant change in levels of total cholesterol was not observed, suggesting that 26a could not chelate BAs. A pharmacokinetic (PK) study of 26a was carried out to have a better understanding of its high efficacy. Aer once a day oral dose (QD) of 26a (100mg/kg) to ob/ob mice, plasma and intestinal tissue were collected and analyzed. 26a exhibited relatively low systemic exposure in ob/ob mice plasma, with Cmax (peak drug concentration) of 0.071g/mL at 2h (Fig.7A). Additionally, no accumulation was found as the drug concentration fell below 0.01g/mL aer 10 h. The low-absorbed prole of 26a could explain the reduced gallbladder lling eect, which was consistent with our

SCIENTIFIC REPORTS

4

www.nature.com/scientificreports/

Figure 6. Study of GLP-1 secretion by 26a in ob/ob mice. Four groups of ob/ob mice (female, n=89 for each time point per group) were administered 0.25% CMC (control), 26a (100mg/kg), linagliptin (3mg/kg) and 26a (100mg/kg) plus linagliptin (3mg/kg). Blood samples were collected aer 6, 12, and 24h. All the animals were fasted for 6hours before collecting blood samples at the indicated time points. At other times, animals werefree accessed to water and food. All time point shared the same control group collected aer 12h. **p<0.01; ***p<0.001 vs. control. Error bar indicates SEM. The number above the error bar displays the fold-change over that in the control group.

Figure 7. Concentration of 26a in plasma (A) and the intestinal tissue (duodenum, jejunum, ileum, colon, B) of ob/ob mice. Once a day oral dose of 26a (100mg/kg) was administered to 12h-fasted ob/ob mice (male, n=3). The blood and intestinal tissue samples were collected before dosing or 2, 4, 6, 10, 16, and 24h aer dosing. Density of intestinal tissue was taken as 1g/ml. Error bar indicates SEM.

design strategy. The drug level of 26a in intestinal tissue (Fig.7B) was relatively high with a Cmax (at 2h) of 136, 92, 331, and 27 g/ml for the tissue of duodenum, jejunum, ileum, colon, respectively. The level of 26a in intestinal tissue was greater than 0.8g/ml even at 24h. The high concentration of 26a in intestinal tissue could have accounted for the robust and long-lasting hypoglycemic eect and persistent GLP-1 stimulation in ob/ob mice. The intestinally-targeted prole of 26a was proven by the high concentration ratio between intestinal tissue and plasma. As mentioned above, TGR5 has moderate expression in the intestinal tract and low expression in BAT and skeletal muscle; therefore, we could assume that the robust hypoglycemic eect was owing mainly to TGR5 activation in enteroendocrine cells. That is, local activation of TGR5 in the intestinal tract can elicit a long-lasting eect on glucose levels.

Finally, the gallbladder lling eect of 26a was measured in ob/ob mice. Once a day oral dose of 26a (100mg/kg, Fig.8A,B) did not increase gallbladder area or bile weight signicantly. However, aer the 3-day administration, gallbladder area and bile weight were increased signicantly by 134% and 129%, respectively. These data suggested that treatment in ob/ob mice for 3 days is a more sensitive and suitable model to judge the gallbladder lling eect.

The gallbladder lling eect of ob/ob mice aer 3-day treatment was not in accordance with the low systemic exposure in the pharmacokinetic study. Hence, further experimentation to ascertain the distribution of 26a in the gallbladder was carried out. Aer the gallbladder lling assay, samples of plasma, bile, and gallbladder tissue were collected and evaluated for drug concentration. The concentration of 26a in plasma was relatively low (183ng/mL), but was higher in the gallbladder and bile (11230 and 78051851ng/ml, respectively). High exposure in bile suggested that 26a was excreted from serum to bile. We hypothesized that active transport had an important role in 26a accumulation in bile. However, the search for corresponding transporters involved in

SCIENTIFIC REPORTS

5

www.nature.com/scientificreports/

Figure 8. (A) Relative gallbladder area and (B) relative bile weight aer once a day oral dose or 3-day treatment. 26a (100mg/kg) or 0.25% CMC (control) were administered (p.o.) to overnight-fasted ob/ob mice (male, n=3 for control group, n=56 for group of 26a) for once or 3 days. 1h aer the nal dose, mice were re-fed for 3h and then the gallbladder was removed and the area measured using a vernier caliper. The relative gallbladder area was calculated from the length multiplied by the width of the gallbladder. Bile weight was measured using analytical balances. **p<0.01; ***p<0.001. Error bar indicates SEM.

Figure 9. Long-term study of 26a in ob/ob mice. Compound 26a (50mg/kg or 100mg/kg) and 0.25% CMC (control) were administered (p.o.) to ob/ob mice (male, n=10) for 18 days. (A) Non-fasting blood glucose and (B) fasting blood glucose in a long-term study of 26a in ob/ob mice. (C) HbA1c level at day 0 (before dosing)

and day 18 (aer the nal dose) in ob/ob mice. (D) Triglyceride level aer the 18-day treatment. *p<0.05, **p<0.01, ***p<0.001 vs. control. Error bar indicates SEM.

hepatic uptake and biliary excretion was not successful (data not shown). Another possible factor contributing to high exposure of 26a in bile was that the bile was highly concentrated in the gallbladder. Further studies were underway to have a better understanding of this phenomenon.

Long-term Study in ob/ob mice. Because of its robust hypoglycemic effect and relatively low gall-bladder filling effect, 26a was evaluated further in a long-term study in ob/ob mice for 18 days. Persistent glucose-lowering eect was observed upon 18-day administration of 26a. Non-fasting and fasting blood glucose were decreased dramatically upon 18-day treatment in ob/ob mice (Fig.9A,B), and the glucose-lowering eect was dose-dependent. Hemoglobin A1c (HbA1c) refers to glycated or glycosylated hemoglobin and can reect the long-term blood glucose level. 26a (100mg/kg) could decrease the HbA1c levels signicantly by 0.94% on day 18 (Fig.9C), whereas 26a (50mg/kg) could decrease the HbA1c levels by 0.48%, but the dierence was not signicant.

SCIENTIFIC REPORTS

6

www.nature.com/scientificreports/

The triglyceride level (Fig.9D) was decreased signicantly by 22% aer 18-day treatment of 26a (100 mg/kg). These data suggested the potential therapeutic eect of 26a in type 2 diabetes and metabolic diseases.

No signicant changes in levels of alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, or total bilirubin were observed (data not shown). Besides, body weight of mice was not aected in the 18-day treatment, suggesting no other apparent toxicity of 26a.

Conclusion

We designed and synthesized a series of intestinally-targeted TGR5 agonists, among which 26a displayed the best in vitro activity and extremely low Caco-2 cell permeability, thus it was chosen for validation of the intestinally-targeted strategy. The PK study conrmed its low absorption into systemic circulation and high concentration in the intestinal-tissue, which was in accordance with our design strategy. In the meantime, 26a (100 mg/kg) exhibited robust and long-lasting hypoglycemic eect in ICR and ob/ob mice. Furthermore, 26a exhibited a consistent hypoglycemic eect throughout the 18-day treatment of ob/ob mice. These data conrmed that a robust hypoglycemic eect could be achieved by intestinal activation of TGR5 independent of systemic exposure. Pleasingly, a decreased gallbladder lling eect was observed aer once a day oral dose of 26a (100mg/ kg) in ICR mice compared with that of absorbed TGR5 agonist 2. However, aer 3-day treatment with 26a (100mg/kg) in ob/ob mice, gallbladder area and bile weight were increased signicantly by 134% and 129% compared with that of the control group, respectively. Further pharmacokinetic study revealed that the drug concentration ratio between bile and plasma of 26a was relatively high, which suggested that low level of 26a in plasma was secreted to gallbladder. However, the search for corresponding transporters involved in hepatic uptake and biliary secretion yield no good results. The high correlation between drug concentration in bile rather than plasma with gallbladder lling eect suggests that gallbladder-nonabsorbed instead of systemic-nonabsorbed prole is the key to later optimization, and our further studies on the reduction of the secretion from plasma to gallbladder of intestinally-targeted agonists are underway. Furthermore, the cooperative eect in stimulating GLP-1 secretion of 26a with a DPP-4 inhibitor suggested that dose reduction of TGR5 agonist may be permitted when combined with DPP-4 inhibitors, thus the side eects might be minimized.

Taken together, our study revealed that TGR5 agonism in intestine alone could bring robust hypoglycemic activity. Although the gallbladder lling eect of 26a was not eliminated, it was reduced compared with that of systemic absorbed TGR5 agonist. This strategy could provide a foundation for further studies on non-absorbed TGR5 agonists or other intestinally-targeted agents.

Methods

Medicinal chemistry work and experimental procedure are available in Supplementary Information. Animal experiments were carried out according to the Guidelines for the Care and Use of Laboratory Animals and were approved by the Animal Care and Use Committee, Shanghai Institute of Materia Medica, Chinese Academy of Sciences.

1. Maruyama, T. et al. Identication of membrane-type receptor for bile acids (M-BAR). Biochem. Biophys. Res. Commun. 298, 714719 (2002).

2. Kawamata, Y. et al. A G protein-coupled receptor responsive to bile acids. J. Biol. Chem. 278, 94359440 (2003).3. Pols, T. W., Noriega, L. G., Nomura, M., Auwerx, J. & Schoonjans, K. The bile acid membrane receptor TGR5: a valuable metabolic target. Dig. Dis. 29, 3744 (2011).

4. Keitel, V. et al. The bile acid receptor TGR5 (Gpbar-1) acts as a neurosteroid receptor in brain. Glia 58, 17941805 (2010).5. Vassileva, G. et al. Targeted deletion of Gpbar1 protects mice from cholesterol gallstone formation. Biochem. J. 398, 423430 (2006).6. Katsuma, S., Hirasawa, A. & Tsujimoto, G. Bile acids promote glucagon-like peptide-1 secretion through TGR5 in a murine enteroendocrine cell line STC-1. Biochem. Biophys. Res. Commun. 329, 386390 (2005).

7. Donnelly, D. The structure and function of the glucagon-like peptide-1 receptor and its ligands. Br. J. Pharmacol. 166, 2741 (2012).8. Thomas, C. et al. TGR5-mediated bile acid sensing controls glucose homeostasis. Cell Metab. 10, 167177 (2009).9. Watanabe, M. et al. Bile acids induce energy expenditure by promoting intracellular thyroid hormone activation. Nature 439, 484489 (2006).

10. Chen, X., Lou, G., Meng, Z. & Huang, W. TGR5: a novel target for weight maintenance and glucose metabolism. Exp. Diabetes Res.

2011, 853501 (2011).

11. Pols, T. W., Noriega, L. G., Nomura, M., Auwerx, J. & Schoonjans, K. The bile acid membrane receptor TGR5 as an emerging target in metabolism and inammation. J. Hepatol. 54, 12631272 (2011).

12. Zhong, M. TGR5 as a therapeutic target for treating obesity. Curr. Top. Med. Chem. 10, 386396 (2010).13. Evans, K. A. et al. Discovery of 3-aryl-4-isoxazolecarboxamides as TGR5 receptor agonists. J. Med. Chem. 52, 79627965 (2009).14. Briere, D. A. et al. Novel Small Molecule Agonist of TGR5 Possesses Anti-Diabetic Eects but Causes Gallbladder Filling in Mice. PloS One 10, e0136873 (2015).

15. Phillips, D. P. et al. Discovery of triuoromethyl(pyrimidin-2-yl)azetidine-2-carboxamides as potent, orally bioavailable TGR5 (GPBAR1) agonists: structure-activity relationships, lead optimization and chronic in vivo efficacy. J. Med. Chem. 57, 32633282 (2014).

16. Martin, R. E. et al. 2-Phenoxy-nicotinamides are potent agonists at the bile acid receptor GPBAR1 (TGR5). ChemMedChem 8, 569576 (2013).

17. Agarwal, S. et al. Discovery of a potent and orally efficacious TGR5 receptor agonist. ACS Med. Chem. Lett. 7, 5155 (2016).18. Fryer, R. M. et al. G protein-coupled bile acid receptor 1 stimulation mediates arterial vasodilation through a K(Ca)1.1 (BK(Ca))-dependent mechanism. J. Pharmacol. Exp. Ther. 348, 421431 (2014).

19. Li, T. et al. The G protein-coupled bile acid receptor, TGR5, stimulates gallbladder lling. Mol. Endocrinol. 25, 10661071 (2011).20. Duan, H. et al. Design, synthesis, and antidiabetic activity of 4-phenoxynicotinamide and 4-phenoxypyrimidine-5-carboxamide derivatives as potent and orally efficacious TGR5 agonists. J. Med. Chem. 55, 1047510489 (2012).

21. Piotrowski, D. W. et al. Identication of tetrahydropyrido[4,3-d]pyrimidine Amides as a new class of orally bioavailable TGR5 agonists. ACS Med. Chem. Lett. 4, 6368 (2013).

22. Futatsugi, K. et al. Optimization of triazole-based TGR5 agonists towards orally available agents. MedChemComm 4, 205210 (2013).

SCIENTIFIC REPORTS

7

www.nature.com/scientificreports/

23. Xu, Y. Recent Progress on Bile Acid Receptor Modulators for Treatment of Metabolic Diseases. J. Med. Chem. doi: 10.1021/acs. jmedchem.5b00342 (2016).

24. Hodge, R. J. & Nunez, D. J. The Therapeutic Potential of TGR5 Agonists. Hope or Hype? Diabetes Obes. Metab. doi: 10.1111/ dom.12636 (2016).

25. Duan, H. et al. Discovery of intestinal targeted TGR5 agonists for the treatment of type 2 diabetes. J. Med. Chem. 58, 33153328 (2015).

26. Handelsman, Y. Role of bile acid sequestrants in the treatment of type 2 diabetes. Diabetes Care 34 Suppl 2, S244250 (2011).27. Tziomalos, K. Lipid-lowering agents in the management of nonalcoholic fatty liver disease. World J. Hepatol. 6, 738744 (2014).28. Hansen, M., Sonne, D. P. & Knop, F. K. Bile acid sequestrants: glucose-lowering mechanisms and efficacy in type 2 diabetes. Curr. Diab. Rep. 14, 482490 (2014).

29. Harach, T. et al. TGR5 potentiates GLP-1 secretion in response to anionic exchange resins. Sci. Rep. 2, 430436 (2012).30. Prawitt, J., Caron, S. & Staels, B. Glucose-lowering eects of intestinal bile acid sequestration through enhancement of splanchnic glucose utilization. Trends Endocrinol. Metab. 25, 235244 (2014).

31. Boliu, V. et al. Inventors; Exelixis, Inc., assignee. TGR5 agonists. Patent Cooperation Treaty (PCT) WO2011071565A1. 2011 Jun 16.32. Flatt, B. T. & Mohan, R. Inventors; Exelixis, Inc., assignee. TGR5 agonists: imidazole and triazole compounds containing a quaternary nitrogen. Patent Cooperation Treaty (PCT) WO2014100021A1. 2014 Jun 26.

33. Flatt, B. T. & Mohan, R. Inventors; Exelixis, Inc., assignee. TGR5 agonists having an imidazole or triazole core with subtituent having a quaternary nitrogen. Patent Cooperation Treaty (PCT) WO2014100025A1. 2014 Jun 26.

This work was financially supported by grant from National Nature Science Foundation of China (Grant 81473093), grant from State Key Laboratory of Drug Research (SIMM1403ZZ-03) and grant from Shanghai Institute of Materia Medica (CASIMM0120152030).

Author Contributions

H.C. designed and performed experiments, prepared the figures and wrote the manuscript. Z.-X.C., K.W., M.-M.N., Q.-A.Z., Y.F. and Y.-L.Y. performed experiments and reviewed the manuscript. J.-H.S. and Y.L. designed experiments and reviewed the manuscript.

Supplementary information accompanies this paper at http://www.nature.com/srep

Competing nancial interests: The authors declare no competing nancial interests.

How to cite this article: Cao, H. et al. Intestinally-targeted TGR5 agonists equipped with quaternary ammonium have an improved hypoglycemic eect and reduced gallbladder lling eect. Sci. Rep. 6, 28676; doi: 10.1038/srep28676 (2016).

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

SCIENTIFIC REPORTS

8