Evaluation of anti-diabetic effect and gall bladder function with 2-thio-5-thiomethyl substituted imidazoles as TGR5 receptor agonists

Article abstract of DOI:10.1016/j.bmcl.2017.02.069

A novel series of 2-thio-5-thiomethyl substituted imidazoles was discovered to be potent TGR5 agonists that possessed glucose-lowering effects while inhibiting gall bladder emptying in mice.

Full text of DOI:10.1016/j.bmcl.2017.02.069

Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Evaluation of anti-diabetic effect and gall bladder function with 2-thio-
5-thiomethyl substituted imidazoles as TGR5 receptor agonists
⇑
Xuqing Zhang , Zhihua Sui, Jack Kauffman, Cuifen Hou, Cailin Chen, Fuyong Du, Thomas Kirchner, Yin Liang,
Dana Johnson, William V. Murray, Keith Demarest
Cardiovascular and Metabolic Research, Janssen Research & Development, LLC, Welsh & McKean Roads, Box 776, Spring House, PA 19477, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A novel series of 2-thio-5-thiomethyl substituted imidazoles was discovered to be potent TGR5 agonists
that possessed glucose-lowering effects while inhibiting gall bladder emptying in mice.
Ó 2017 Elsevier Ltd. All rights reserved.
Received 15 December 2016
Revised 23 February 2017
Accepted 26 February 2017
Available online xxxx
Keywords:
TGR5
Imidazole
OGTT
Gall bladder
Takeda G-protein-coupled receptor 5 (TGR5), also known as
GPR 131, GPBAR1, or M-BAR is G protein-coupled receptor acti-
vated by bile acids (BAs).1 It is broadly expressed in human tissues,
such as the GI tract, gall bladder, spleen, lung, brown adipose tissue
and placenta.2–4 Bile acids induce glucagon like peptide-1 (GLP-1)
secretion from primary intestinal enteroendocrine cells by signal
transduction through Gs protein-mediated cyclic adenosine
monophosphate (cAMP) accumulation via TGR5. The process could
regulate homeostasis of blood glucose through a variety of mecha-
nism including promoting glucose-induced insulin secretion, sup-
pressing glucagon release, delaying gastric emptying, promoting
satiety, and increasing glucose disposal in the peripheral tis-
sues.5–8 Moreover, activation of TGR5 in brown adipose tissue
has been proposed to increase energy expenditure through the
induction of type 2 iodothyronine deiodinase (D2). Therefore,
TGR5 activation provides a promising strategy for treatment of
type 2 diabetes mellitus and associated metabolic disorders. In
particular, a potent and selective TGR5 agonist may be beneficial
for the treatment of type 2 diabetes and obesity.
through activation of the TGR5 receptor in the epithelium of gall
bladder by administration of either steroidal or non-steroidal
TGR5 agonists. To alleviate or eliminate the strong side effect on
the gall bladder, increasing efforts have been made to develop
gut-restricted TGR5 agonists specifically targeting the GI tract
while avoiding systemic exposure.22,23 Therefore, GI restricted
TGR5 agonists are a potential anti-diabetes mellitus strategy to tar-
get the intestine locally to minimize this unwanted side effect.
However, it is still highly challenging to pursue adequate anti-dia-
betic effect by solely stimulating the production of GLP-1 secretion
from the gut without all other benefits on increasing energy home-
ostasis, enhancing insulin sensitivity and improving glucose toler-
ance through systemic exposure approach.
Herein, we report our effort to develop novel and selective TGR5
agonists for the treatment of metabolic diseases.24,25 The goal of
the program was to identify an orally efficacious TGR5 agonist
from this series to enable evaluation on both anti-diabetic effect
and gall bladder function in disease-relevant preclinical models.
Examination of non-steroidal TGR5 agonist chemotypes in the lit-
erature9–15 led us to postulate that a common pharmacophore 1
may participate in ligand-receptor binding (Fig. 1). This prompted
us to synthesize a few small chemical libraries based on template 1
by modifying the core heteroaryl ring as well as the linkers A–E for
preliminary screening. A hit scaffold 2 was discovered to display
moderate TGR5 agonistic activity in murine enteroendocrine cells
(STC-1 cells) expressing TGR5. To further explore 2, we then
conducted detail SAR studies to improve TGR5 activity in STC-1
cells.
There have been several reports in the literature of both selec-
tive steroidal TGR5 agonists and synthetic, non-steroidal TGR5
agonists.9–23 While both chemo-types exhibited potent TGR5 ago-
nist activity in cell-based functional assays and were orally effica-
cious in lowering glucose levels in rodents, some adverse side
effects, of which mostly in gall bladder, have been observed
⇑
Corresponding author.
E-mail address: xzhang5@its.jnj.com (X. Zhang).
http://dx.doi.org/10.1016/j.bmcl.2017.02.069
0960-894X/Ó 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Zhang X., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.069

2
X. Zhang et al. / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
Z
activity with EC50 returning to sub
14. Unfortunately, introduction of a slightly bigger acetyl substitu-
tion on the basic aniline abolished TGR5 activity as evidenced by
15 with EC50 >50 lM in both STC-1 and NCI-716 cells. We then
shifted our effort on the right-side of the structure. In contrast to
the low tolerance for modification at the left-side linker, 4-substi-
tution on the right-side phenyl ring was tolerated for modification
as illustrated by 4-CN substituted phenyl analogue 16 with EC50 of
290 nM in STC-1 cells. Furthermore, incorporation of more polar
carboxylic acid (17) or tetrazole group (19) still maintained low
lM range in STC-1 cells for
N
B
D
H
E
A
A
C
S
N
X
Library Synthesis
1
Y
A, B, C, D: O, NH, S or CH2
E: direct bond, CH2
2
Hit Scaffold
:
5 or 6 membered heteroaryl
H
Fig. 1. Pharmacophore-based design of novel TGR5 agonists.
l
M of EC50 in STC-1 cells as compared to 16, providing an efficient
To explore SAR on this scaffold, an efficient synthesis was devel-
path to improve compound solubility with maintaining moderate
in vitro TGR5 agonistic potency. Unfortunately, further increasing
polarity by oxidation of the 2-S group on the imidazole ring into
sulfoxide resulted in 6–7-fold of drop in TGR5 agonistic potency
oped and utilized (Scheme 1). Aniline 3 was coupled with ethyl 2-
bromoacetate 4 under weakly basic conditions followed by acyla-
tion of the aniline adduct to give ethyl N-acetyl-N-(4-fluo-
rophenyl)glycinate 5 in 68% yield over two steps. Intermediate 5
was then treated with potassium ethoxide and ethyl formate to
afford the adduct 6 in 65% yield, this underwent acid catalyzed
cyclization by the treatment with KSCN to yield imidazole 7 in
71% yield. Alkylations of thiol 7 with various benzyl halides or
mesylates in the presence of potassium carbonate followed by
DIBAL reduction gave the corresponding imidazolylmethanols 9.
Compounds 9 were coupled with 3,4-di-methoxy-phenol, 3,4-di-
methoxy-thiophenol or 3,4-dimethoxy-aniline via the correspond-
ing mesylates to afford the target compounds 10.
We began our SAR exploration on the left side of the structure
(Table 1). Consistent with the SAR of a structurally similar series
discovered in our group earlier,24 modification on the left side aro-
matic ring identified 3,4-di-methoxy as optimal for TGR5 agonistic
activity (data not shown). Polar groups (A) on the linker were
poorly tolerated while lipophilic S containing linker displayed
potent TGR5 agonistic activity. As evidenced by 11 with EC50 of
in STC-1 cells (EC50 of 44.9 lM in STC-cells for 18), suggesting
polarity was not tolerated on the central imidazole ring. Given
the moderate TGR5 potency achieved by the right-side phenyl ring
substituted with a polar group such as carboxylic acid or tetrazole
at the 4-position, other polar amide and sulfonamide analogues
20–26 were explored. To our delight, compared with the carboxylic
acid analogue 17, the corresponding amides with terminal sulfonic
acid (20), terminal quaternary ammonium salt (21) or D-arginine
group displayed 2–15-fold EC50 boost in STC-1 cells. While signif-
icantly less potent in STC-1 cells, 21 was interesting since its qua-
ternary ammonium salt structure could limit its intestinal
permeability. Such a GI restricted agent for targeting the intestine
locally might minimize unwanted side effect on the gall bladder.
We then introduced a more polar sulfonamide moiety into the side
chain on the phenyl ring. As expected, 23 bearing a small sulfon-
amide group maintained good TGR5 agonistic potency with EC50
of 590 nM in STC-1 cells. However, combination of polar sulfon-
amide and quaternary ammonium salt group into the side chain
failed to maintain good TGR5 agonistic potency in both STC-1
and NCI H716 cells. As exemplified by 24 and 25, substitution of
one quaternary ammonium salt group dramatically reduced
160 nM in STC-1 cells and EC50 of 5.68 lM in NCI H716 cells,
increasing polarity by replacing the S with a O or a NH group
resulted in dramatic loss or abolishment on TGR5 agonistic
potency (EC50 of 5.68 lM for 12 and EC50 of >50 lM for 13 in
STC-1 cells). For the in vitro potency evaluation, both 11 and 12
showed a preference for mouse TGR5 in STC-1 cells over human
TGR5 in NCI H76 cells. One potential reason is due to low amino
acid sequence identity of the two receptors.3 However, expression
level of TGR5 in different cell lines would also play a critical role on
the potency evaluation. It is possible that rather lower expression
of human TGR5 attributed to the lower human TGR5 potency
observed in NCI-716 cells. Masking of the polar and basic free
amino group of 13 with a formyl group revived TGR5 agonistic
TGR5 agonistic potency (24, EC50 of 9.36
lM; 25, EC50 of
11.75 M) in STC-1 cells. Furthermore, a bis quaternary ammo-
l
nium salt group on the side chain abolished activity (26). To eval-
uate both anti-diabetic effects and gall bladder function, 19 was
selected as the tool compound for in vivo studies to help us decide
whether a TGR5 agonist devoid of undesired gall bladder effects
could be achieved.
Compound 19 was evaluated for its in vivo glucose lowering
activity with an oral glucose tolerance test (OGTT) in C57 BL/6 mice
N
F
F
OK
SH
O
3)
F
4)
1) 2)
N
O
O
O
O
N
N
Br
NH2
O
O
O
O
O
3
4
5
F
6
7
F
F
F
N
R2
R2
N
N
5)
S
R2
7)
S
O
6)
S
N
R1
HO
R1
N
R1
A
N
O
O
O
F
F
F
8
9
10
Scheme 1. Reagents and conditions: 1) AcONa, EtOH, 2 h, 80 °C (80%); 2) AcCl, TEA, 3 h, 0–30 °C (85%); 3) EtOK, HCO2Et, toluene, overnight, 30 °C (65%); 4) KSCN, conc. HCl,
2 h, 90 °C (71%); 5) K2CO3, acetone, 1 h, 30 °C (75–92%); 6) DIBAL, 2 h, 0–30 °C (45–70%); 7) SOCl2, overnight, 0–25 °C then Cs2CO3, 3,4-di-MeO-Ph-AH (A = O, S or NH), 2 h,
25 °C (50–75%).
Please cite this article in press as: Zhang X., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.069

X. Zhang et al. / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
3
Table 1
SAR of R1, R2 and A of 2-benzylthio substituted imidazoles 10.
F
R2
N
S
R1
A
N
O
O
F
10
ID
R1
R2
A
STC-1
NCI H716
(human)b
(mouse)a
EC50
(l
M)
EC50 (lM)
11
12
13
14
15
16
17
18
19
Cl
Cl
Cl
Cl
Cl
F
F
F
F
H
H
H
H
S
O
NH
NCHO
NAc
S
S
SO
S
0.16
5.68
>50
0.99
>50
0.29
6.08
44.9
2.06
5.68
17.8
>50
21.75
>50
4.94
>50
>50
H
CN
CO2H
CO2H
42.62
N
N
NH
N
20
F
S
0.38
8.26
HO3S HN
O
21
22
F
F
S
S
2.3
>50
HN
NH
N
O
0.76
36.8
H
N
N
H
H2N
O
O
OH
23
24
25
F
F
F
S
S
S
0.59
9.36
16.63
29.55
20.75
O
S
N
N
N
O
O
S
N
O
H
11.75
>50
O
S
N
N
O
HO
26
F
S
>50
O
S
Me3N
Me3N
N
O
a
STC-1: murine enteroendocrine cell lines expressing TGR5.
NCI-H716: human enteroendocrine cell lines expressing TGR5.
b
(Fig. 2). The area under the curve AUC0–120 min (absolute AUC0–120
min) for glucose levels versus time were calculated after oral
administrations of 19 at 30 mg/kg and 50 mg/kg in 20% 2-hydrox-
ypropyl-b-cyclodextrin. It was found to cause 9.8% and 12.0%
reduction in blood glucose absolute AUC0–120 min at 30 mg/kg and
50 mg/kg respectively compared with the vehicle control group
(Fig. 2). A pharmacokinetic profile in vivo revealed the absorption
First, 19 was evaluated for its ability to inhibit on the egg yolk-
induced reduction of gall bladder size in CD-1 mice (Fig. 3).27 The
mice were dosed orally at 5 ml/kg 20% HPBCD or a suspension of
19 in the same vehicle at 50 mg/kg. One hour later, saline or 30%
egg yolk suspension was orally administrated at 20 ml/kg. Gall
bladder emptying was assessed by comparing gall bladder weights
from control and egg yolk administered mice. It was clear that con-
traction and emptying of the gall bladder was induced by egg yolk
compared with the vehicle/saline group. Pleasingly, 19 at 50 mg/kg
did not block the egg yolk-induced gall bladder emptying effect at
15 min. It exhibited relatively low systemic exposure in CD-1 mice
of 19. The plasma drug exposure was measured as 1.18 lM
45 min post oral dose at 50 mg/kg. While moderately absorbed,26
19 displayed an acute glucose-lowering effect and improved glu-
cose tolerance in C57 BL/6 mice induced by a relatively high drug
dose. In the context of reported TGR5 agonist driven gall bladder
toxicity via systemic administration, peripherally biased 19 was
selected for further evaluation of its gall bladder filling effect in
multiple rodent models.
plasma, with 0.02 lM at 1.25 h, which might explain that it did not
significantly inhibit the gall bladder emptying effect.
Next, we sought to further explore the effect of TGR5 agonist 19
on egg yolk-induced gall bladder emptying in fasted C57 BL/6 mice
Please cite this article in press as: Zhang X., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.069

4
X. Zhang et al. / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
Fig. 4. Effect of 19 on gall bladder function in fasted C57 BL/6 mice, 50 mg/kg po vs.
30 mg/kg sc: On-way ANOVA Analysis, **P < 0.01 vs Vehicle control; ***P < 0.001 vs
Vehicle control.
Fig. 2. Compound 19 reduced AUC of blood glucose in C57 BL/6 mice (OGTT): One-
way ANOVA Analysis (*P < 0.05 vs Vehicle).
weighed 15 min later. When given 50 mg/kg PO, 19 had no signif-
icant effect on the gall bladder weight. However, 19 dramatically
inhibited the gall bladder emptying when given 30 mg/kg subcuta-
neously compared to the vehicle control. The plasma exposure
levels were also measured 45 min post dosing of 19. It exhibited
15.8
pared with 1.03
l
M of systemic exposure via subcutaneous injection com-
M via oral administration, suggesting that a gut
l
restricted TGR5 agonist could maintain its therapeutic benefit
while reducing the potential for toxic effects on gall bladder
emptying.
Given its glucose-lowering effect without a significant gall blad-
der filling effect via oral administration in the acute models using
C57 BL/6 mice, 19 was evaluated further in a 2 week chronic study
using db/db mice. After twice a day oral dose (bid) of 19 (50 mg/kg)
to db/db mice for one week, 19 decreased the fed blood glucose
level prior to a fasted OGTT study on day 7 without a significant
effect on the body weight. It also displayed significant glucose-
lowering effect and improved glucose homeostasis in an OGTT
study on day 7 (Fig. 5). However, two of eight mice in this study
expired on day 11. Upon the necropsy after the study, it was found
that 19 impaired the gall bladder function in the db/db mice. The
gall bladder weight increased dramatically compared with that of
the control group; moreover, the gall bladders turned dark red con-
taining bile which was not visibly clear, indicating the severe gall
bladder filling effect. This data convincingly demonstrated that
19 as a TGR5 agonist inhibited the gall bladder contraction and
impaired the gall bladder emptying upon chronic oral administra-
tion. While compound 19 had low systemic exposure, it was still
Fig. 3. Effects of 19 on gall bladder weight in CD-1 mice at 50 mg/kg PO: One-way
ANOVA-Dunnett’s Analysis ***P < 0.001 vs Vehicle/Saline group.
(Fig. 4). To investigate whether systemic exposure could affect the
gall bladder emptying effect, we decided to run head-to-head com-
parison of gall bladder function between oral administration (PO)
and subcutaneous injection (SC) of 19. Fasted C57 mice received
19 at an oral dose of 50 mg/kg or subcutaneous injection of
30 mg/kg followed by oral administration of egg yolk 30 min later.
Then the animals were sacrificed and the gall bladders were
Please cite this article in press as: Zhang X., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.069

X. Zhang et al. / Bioorganic & Medicinal Chemistry Letters xxx (2017) xxx–xxx
5
It had no significant effect on egg yolk-induced gall bladder emp-
tying over 15 min in CD-1 mice. Assessed by another acute study,
19 exhibited no significant effect on the gall bladder emptying at
oral administration of 50 mg/kg but significantly inhibited the gall
bladder emptying when given subcutaneously at 30 mg/kg. In a
2 week chronic study using db/db mice, 19 significantly improved
glucose homeostasis in an OGTT study on day 7 but severely
blocked the gallbladder emptying.
Acknowledgments
The authors thank ADME/PK and Lead Generation Biology
teams at Janssen R&D for their technical assistance. The authors
gratefully acknowledge Dr. James Lanter for scientific discussion
and revising the paper.
A. Supplementary data
Day 7 - OGTT Absolute AUC
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2017.02.
069.
80000
60000
40000
20000
0
*
References
1. Tiwari A, Maiti P. Drug Disc. Today. 2009;14:523–530.
2. Maruyama T, Miyamoto Y, Nakamura T, et al. Biochem Biophys Res Commun.
2002;298:714–719.
3. Kawamata Y, Fujii R, Hosoya M, et al. J Biol Chem. 2003;278:9435–9440.
4. Thomas C, Pellicciari R, Pruzanski M, Auwerx J, Schoonjans K. Nat. Rev. Drug
Disc.. 2008;7:678–693.
5. Watanabe M, Houten SM, Mataki C, et al. Nature. 2006;439:484–489.
6. Prawitt J, Caron S, Staels B. Curr. Diab. Rep.. 2011;11:160–166.
7. Pols TW, Noriega LG, Nomura M, Auwerx J, Schoonjans K.
2011;54:1263–1272.
8. Chen X, Lou G, Meng Z, Huang W. Exp Diab Res. 2011. 853501-853501.
9. Evans KA, Budzik BW, Ross SA, et al. J. J. Med. Chem.. 2009;52:7962–7965.
10. Herbert MR, Siegel DL, Staszewski L, et al. Bioorg Med Chem Lett.
2010;20:5718–5721.
11. Budzik BW, Evans KA, Wisnoski DD, et al. Bioorg Med Chem Lett.
2010;20:1363–1367.
J Hepatol.
e
9
1
cl
i
d
h
p
e
C
V
12. Charton J, Deprez B, Leroux F, Staels B, Muhr-Tailleux A, Hennuyer N, Lestavel S,
Lassalle M, Dubanchet B. WO 2015189330 A1, 2015.
13. Bissantz C, Dehmlow H, Martin R. E, Obst S. U, Richter H, Ullmer C.
US20100105906A1, 2010.
14. Maruyama M. WO2010016552A1, 2010.
15. Bollu V, Boren BC Dalgard JE, Flatt BT, Haq N, Hudson S, Mohan R, Morrissey M,
Pratt B, Wang T-L. WO 2010093845A1 2010.
One Way ANOVA - Dunnett's t Test
Fig. 5. Effects of 19 on glycemic control and gall bladder function in a 2 week study
in db/db mice (50 mg/kg PO bid): One-way ANOVA analysis *P < 0.05 vs Vehicle
group.
16. Phillips DP, Gao W, Yang Y, et al. J Med Chem. 2014;57:3263–3282.
17. Zou Q, Duan H, Ning M, et al. Eur J Med Chem. 2014;82:1–15.
18. Duan H, Ning M, Chen X, et al. J Med Chem. 2012;55:10475–10489.
19. Piotrowski DW, Futatsugi K, Warmus JS, et al. ACS Med. Chem. Lett..
2013;4:63–68.
20. Futatsugi K, Bahnck KB, Brenner MB, et al. Med. Chem. Comm.. 2013;4:205–210.
21. Agarwal S, Patil A, Aware U, et al. ACS Med. Chem. Lett.. 2016;7:51–55.
22. Duan H, Ning M, Zou Q, et al. J Med Chem. 2015;58:3315–3328.
23. Cao H, Chen Z-X, Wang K, et al. Sci Rep. 2016;6:28676–28683.
24. Zhang X, Wall M, Sui Z. WO2015160772A1, 2015.
sufficient to cause gall bladder filling at this dose/exposure proba-
bly due to the drug accumulation in the gall. Overall, the anti-dia-
betic benefit for a low absorbed TGR5 agonist likely could not be
separated from concurrent pharmacology in the gallbladder.
In summary, we designed and synthesized a series of 2-thio-5-
thiomethyl substituted imidazoles as TGR5 receptor agonists cul-
minating in the identification of compound 19 as a suitable tool
to assess in vivo. This molecule displayed moderate in vitro activity
in both murine TGR5 expressing STC-1 cells and human TGR5
expressing NCI H716 cells. It was chosen as the tool compound
for evaluation on both anti-diabetic effect and gall bladder func-
tion. Compound 19 with low oral exposure significantly reduced
blood glucose excursion during an OGTT study in C57 BL/6 mice.
25. Zhang X, Sui, Z, Kauffman J, Du F, Kirchner T, Hou C, Liang Y, Johnson D, Murray
WV, Demarest K. Abstracts of Papers, 252nd ACS National Meeting
Exposition, Philadelphia, PA, United States, 2016, August 21–25, MEDI-384.
&
26. A full PK study was conducted on compound 19 in C57 BL/6 mice and the date
was reported in the supporting information.
27. Valsecchi B, Toson CJ. Pharm. Meth.. 1982;7:193–195.
Please cite this article in press as: Zhang X., et al. Bioorg. Med. Chem. Lett. (2017), http://dx.doi.org/10.1016/j.bmcl.2017.02.069