Discovery of 5-aryloxy-2,4-thiazolidinediones as potent GPR40 agonists

Article abstract of DOI:10.1016/j.bmcl.2009.10.052

Systematic structure-activity relationship (SAR) studies of a screening lead led to the discovery of a series of thiazolidinediones (TZDs) as potent GPR40 agonists. Among them, compound C demonstrated an acute mechanism-based glucose-lowering in an intraperitoneal glucose tolerance test (IPGTT) in lean mice, while no effects were observed in GPR40 knock-out mice.

Full text of DOI:10.1016/j.bmcl.2009.10.052

Bioorganic & Medicinal Chemistry Letters 20 (2010) 1298–1301
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of 5-aryloxy-2,4-thiazolidinediones as potent GPR40 agonists
*
Changyou Zhou , Cheng Tang, Eric Chang, Min Ge, Songnian Lin, Eric Cline, Carina P. Tan, Yue Feng,
Yun-Ping Zhou, George J. Eiermann, Aleksandr Petrov, Gino Salituro, Peter Meinke, Ralph Mosley,
Taro E. Akiyama, Monica Einstein, Sanjeev Kumar, Joel Berger, Andrew D. Howard, Nancy Thornberry,
*
Sander G. Mills, Lihu Yang
Merck Research Laboratories, Rahway, NJ 07065, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Systematic structure–activity relationship (SAR) studies of a screening lead led to the discovery of a series
of thiazolidinediones (TZDs) as potent GPR40 agonists. Among them, compound C demonstrated an acute
mechanism-based glucose-lowering in an intraperitoneal glucose tolerance test (IPGTT) in lean mice,
while no effects were observed in GPR40 knock-out mice.
Received 15 December 2008
Revised 9 October 2009
Accepted 12 October 2009
Available online 15 October 2009
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
GPR40
Agonist
5-Aryloxy-2,4-thiazolidinediones
Diabetic
Hypoglycemia
GDIS
GSIS
G-Protein-coupled receptor 40, GPR40, which is abundantly ex-
pressed in the pancreas, functions as a receptor for long-chain free
fatty acids (FFAs). Activation of GPR40 by several naturally occur-
ring medium to long-chain FFAs enhances insulin secretion
through the amplification of intracellular calcium signaling (via
IP3 generation) and the exocytosis of insulin secretory granules
(via PKC activation). The specificity of these responses is illustrated
by lack of glucose-dependent insulin secretion (GDIS) amplifica-
tion by FFAs following the reduction of GPR40 expression by siR-
NA.1 Therefore, there are several potential advantages of GPR40
as a drug target for the treatment of type 2 diabetes. First, since
GPR40-mediated insulin secretion is glucose dependent, there is
reduced risk of hypoglycemia. Second, the limited tissue distribu-
tion of GPR40 in islets suggests that there would be less chance
for side effects associated with GPR40 activity in other tissues.
Third, the identification of synthetic GPR40 agonists suggests that
small molecule drugs can potentially be developed for this target.
In addition, GPR40 agonists may potentially be used in combina-
tion with other anti-diabetes drugs that have distinct mechanisms
of action such as insulin sensitizers or inhibitors of hepatic glucose
production. It has been reported by GSK scientists that aryl cyclo-
propionic acids2 (compound A (Cpd A), Fig. 1) and aryl propionic
acids3 (compound B (Cpd B), Fig. 1) are potent GPR40 agonists. Sev-
eral examples carboxylic acid GPR40 agonists have also been re-
ported recently.4 Some thiazolidinedione (TZD) drugs that target
peroxisome proliferator-activated receptor (PPARs) are also identi-
fied as GPR40 agonists.5 In our recent efforts to identify novel
GPR40 agonists, we screened about 2000 in-house compounds
and identified a partial human GPR40 agonist (relative to fatty
acids) with a functional EC50 of 480 nM.6 Because of its structural
similarity to known TZDs that interact with PPARs, this compound
was tested for its binding ability to several PPAR isoforms, and it
was found to be inactive in binding assays at concentrations up
to 10 lM against human PPAR a-, b-, and c-isoforms. Herein we re-
port our structure–activity relationship (SAR) studies of this lead
which led to the identification of a potent and bioavailable
GPR40 agonist (compound C (Cpd C), Fig. 1).
As shown in Scheme 1, one general method of constructing target
compounds starts from key intermediate 3, which can be prepared
by coupling of a phenol or thiophenol (1) and a halogen-substituted
benzaldehyde (2) in the presence of a base. Alternatively, haloarene
(4) can be coupled to 4-hydroxyl/thiol benzaldehyde (5) to obtain
intermediate 3.
The initial route for the preparation of target compounds was
developed based on carbanion chemistry (Scheme 2). Dianion (7),
generated in situ by treatment of heterocycle (6) (e.g., 2,4-thiazo-
lidinediones(TZD), 1,2,4-oxadiazolidine-3,5-dinone (OZD), succini-
mide, and hydantoin) with more than 2 equiv of a strong base such
as LDA or NaHMDS in a mixture of THF/HMPA, can be reacted with
* Corresponding authors. Tel.: +1 732 594 4660; fax +1 732 594 3407.
E-mail address: lihu_yang@merck.com (L. Yang).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.10.052

C. Zhou et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1298–1301
1299
reducing agent such as lithium borohydride to give the desired
product (9) as a racemic mixture (Scheme 3).
O
OH
Most of the target compounds were prepared by a convergent
procedure as described for TZD below and shown in Scheme 4.
First, unsaturated TZD (12) was prepared starting from 4-methoxy-
benzaldehyde by the same procedure as depicted in the Scheme 3.
Further reduction with lithium borohydride gave saturated TZD
(13) which was deprotected with boron tribromide to yield hydro-
xyl TZD (14). Final convergent coupling with aryl halides was per-
formed using Cs2CO3 as a base and with or without copper iodide
as the catalyst.
S
GSK Compound A
F3C
N
H
hGPR40 pEC50 = 8
O
N
OH
O
GSK Compound B
hGPR40 pEC50 = 7
N
H
The SAR of lead compound was systematically investigated
according to four parts of the molecule which we named as head
(TZD), central (phenyl ring), linker, and tail (phenyl ring) as shown
in Figure 2.
An initial SAR study on the heterocyclic head part of the lead
molecule indicated that the original TZD ring was optimal, and
replacement of it with an OZD (X = O), a succinimide (X = CH2),
or other heterocycles significantly reduced the functional GPR40
activity (Table 1).
We then investigated the SAR of the linker region. The results
showed that para-phenyloxy linker provided the most active ago-
nists (9-7). Removal of the para-oxygen, switch of the para-linker
to a meta- or ortho-linker, or replacement with other linkers re-
sulted in markedly reduced intrinsic potency (Table 2).
O
O
S
Screening Hit
NH
F
hGPR40 EC50=480 nM
O
SAR Study in this paper
Cl
O
O
S
Compound C
NH
N
F3C
hGPR40 EC50= 10 nM
O
Figure 1. GPR40 agonists.
X
O
ArYH
+
Route 1
Ar
O
NH
CHO
a
O
1
R
Ar
S
CHO
a
R
2
3
Y
+
Ar
R
10
O
O
HY
a
CHO
R
ArX
+
X = S
S
b
3
Route 2
CHO
NH
4
O
R
O
5
O
Ar
S
6
NH
Scheme 1. Reagents and conditions: (a) Cs2CO3 (5 equiv), DMA, 120–150 °C, 1–
R
12 h. X = halide, Y = S or O.
9
O
Scheme 3. Reagents and conditions: (a) NaOAc (1.5 equiv) neat, 150 °C for 2 h; (b)
LiBH4 (1.2 equiv), THF–pyridine (4:1, v/v), reflux, 2 h.
O
OLi
N
X
PhY
X
PhY
NH
CH2Br
8
9
7
O
OLi
b
O
O
MeO
HO
MeO
S
a
a
S
NH
NH
+
O
CHO
R
O
O
12
X
R
6
NH
O
11
O
6
O
MeO
S
c
S
Scheme 2. Reagents and conditions: (a) LDA (2.5 equiv), THF–HMPA (1:1, v/v),
À78–0 °C for 30 min, then À78 °C; (b) 8 (1.2 equiv), HMPA, À78 to 0 °C for 2 h. X = S,
O, CH2, CH2CH2; Y = O, S.
b
NH
NH
R
O
R
O
14
13
O
a commercially available benzylbromide (8) to provide the target
compound (9) as a racemic mixture.
O
d
Ar
S
NH
The second route involves direct condensation of the substi-
tuted benzaldehyde (3), either commercially available or prepared
according to Scheme 1, with TZD (X = S) in the presence of sodium
acetate at elevated temperature. It was discovered that the con-
densation proceeds effectively by simply melting all the starting
materials in an open flask without use of any solvent. The resulting
unsaturated intermediate compound (10) was treated with a
4
ArX
R
O
9
Scheme 4. Reagents and conditions: (a) NaOAc (1.5 equiv) without solvent, 150 °C
for 2 h; (b) LiBH4 (1.2 equiv), THF–pyridine (4:1, v/v), reflux, 2 h; (c) BBr3
(1.2 equiv), CH2Cl2, 0 °C–rt, 1 h; (d) Cs2CO3 (5 equiv), DMA, 80–120 °C, from 1 to
12 h.

1300
C. Zhou et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1298–1301
Table 3
Linker
O
SAR on tail part of lead compound
O
O
O
Ar
S
S
NH
NH
F
O
O
Tail
*
Central
Entry
Ar
EC50 (nM)
Head
9-17
9-18
9-19
9-20
9-21
9-22
9-23
9-24
9-25
9-26
9-27
9-1
2-Me-Ph
3-Me-Ph
4-Me-Ph
3,5-diMe-Ph
2,3-diMe-Ph
2,4-diMe-Ph
2,5-diMe-Ph
3,4-diMe-Ph
2-Cl-Ph
124
203
547
96
Figure 2. Lead analysis.
334
121
238
727
1189
318
240
101
207
78
Table 1
SAR on the head part
Cl
O
O
X
3-Cl-Ph
4-Cl-Ph
NH
Cl
2,4-diCl-Ph
3,5-diCl-Ph
2-Cl,4-CF3-Ph
3,5-diMe,4-CN-Ph
1-Naphthyl
O
9-28
9-29
9-30
9-31
Entry
X
EC50 (nM)*
120
135
9-1
9-2
9-3
9-4
9-5
9-6
S
O
101
4017
6451
FLIPR functional assay (with 1% BSA).6
*
CH2
CH2CH2
OCH2
NMe
6% @40
lM
23% @40
lM
2% @40
l
M
Table 4
SAR on pyridine-2-yl tail
FLIPR functional assay (with 1% BSA).6
*
O
O
S
R
NH
Table 2
SAR on linker part
N
R
O
O
S
EC50*(nM)
Y
NH
Entry
N
O
9-32
9-33
9-34
9-35
9-36
9-37
9-38
9-39
9-40
9-41
9-42
9-43
9-44
9-45
3-CF3-Py
4-CF3-Py
5-CF3-Py
6-CF3-Py
3-Me-Py
4-Me-Py
5-Me-Py
6-Me-Py
2287
237
558
1635
1087
4282
4474
2620
1872
218
*
Entry
–Y–
EC50 (nM)
9-7
9-8
9-9
9-10
9-11
9-12
9-13
9-14
9-15
9-16
p-PhO–
m-PhO–
o-PhO–
p-PhS–
307
4700
11,270
5064
9283
2094
1082
6178
5300
2219
p-PhCH2–
5-Cl-Py
p-PhCH@CH–
p-PhCH2CH2–
p-PhCH2O–
p-PhCO–
4-CF3-6-Me-Py
4-Me-6-CF3-Py
3-Cl-5-CF3-Py
3,5-diCl-Py
Isoquinolonyl
688
69
90
144
p-Ph–
*
FLIPR functional assay (with 1% BSA).6
*
FLIPR functional assay (with 1% BSA).6
Therefore, we focused our SAR effort on the optimization of the
tail region of the lead. Although various aromatic substitutions re-
sulted in reasonable activity, 3,5- or 2,4-di-substitutions (such as
9-1, 9-20, and 9-29) are most potent (Table 3). In addition, rela-
tively good potencies were observed when C2 and C3 of the phenyl
ring were fused to form a naphthyl ring, as represented by com-
pound 9-31.
Encouraged by the above results, we went further to investigate
various heterocyclic tails. It was discovered that although incorpo-
ration of mono-substituted pyridine-2-yl tails did not produce sig-
nificant improvement compared to the corresponding phenyl
analogs, the employment of 3,5-disubstituted pyridine-2-yl tails
afforded a series of highly potent GPR40 agonists, especially, 3-
chloro-5-trifluoromethylpyridine-2-yl derivative 9-43, which was
the most potent compound in this series (Table 4). Good potency
was also observed when R1, R2 were connected together to form
an isoquinoline (compound 9-45).
The most exciting results were observed when we moved our
SAR study to the central phenyl ring. As shown in Table 5, introduc-
tion of substitution(s) on the central phenyl ring led to dramatic
changes in potency. 2-Methyl substitution increased potency ꢀ7-
fold to give the most potent compound 9-46 in this series. Slightly
bigger groups such as a chloro or a trifluoromethyl somewhat de-
creased potency (9-48, 9-49, and 9-50), while a bulkier group such
as methoxy suppressed GPR40 activity dramatically (9-51, 9-52).
When the central phenyl ring was replaced with a naphthyl moi-
ety, a complete loss of potency was observed (9-54 and 9-55).
Of the GPR40 agonists described herein, several compounds
were evaluated for their pharmacokinetic (PK) properties in
mouse. In general, these compounds possess good PK profiles.
For example, Cpd C (9-46) demonstrated good oral bioavailability
in mice at 2 mg/kg dose and excellent overall PK characteristics
(F ꢀ100%, AUCpo = 18
l
M h kg/mg, Cl = 1.9 mL/min/kg, T1/2 = 5.1 h,
Tmax = 3 h, Cmax = 1.4
l
M). Thus, Cpd C was selected to probe the

C. Zhou et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1298–1301
1301
Table 5
SAR on central phenyl ring
reduced blood glucose excursion in a dose-dependent manner with
doses ranging from 3 to 100 mg/kg. The maximum efficacy (73%
inhibition of AUCGLu) was achieved at ꢀ30 mg/kg with a corre-
Cl
O
5
O
4
sponding plasma concentration of 37 lM measured 2 h post-dose.
6
1
S
In this study, GLP-1 mimetic exendin-4 was included as a positive
control which completely inhibited glucose excursion at a concen-
tration of 0.0025 mg/kg. Cpd C was also found to significantly en-
hance glucose-dependent insulin secretion (GDIS) in isolated
islets (data no shown).6
To demonstrate that the observed Cpd C-induced glucose-low-
ering was GPR40-dependent, the effects of the ligand on blood glu-
cose excursion during an IPGTT were investigated again in a cohort
of GPR40-/- mice and littermate WT mice. The administration of
30 mg/kg Cpd C again resulted in a significant suppression of glu-
cose AUC during IPGTT in the WT mice (Fig. 3A). In contrast, the
same dose of the compound exerted no inhibition of blood glucose
excursion in the GPR40-/- mice (Fig. 3B). The above findings dem-
onstrate that robust glucose-lowering in normal WT mice by Cpd C
is mediated by GPR40.
NH
M
3
R
W
2
O
*
Entry
R
M
W
EC50 (nM)
9-46
9-47
9-48
9-49
9-50
9-51
9-52
9-53
9-54
9-55
2-Me
2-Me
2-Cl
2-Cl
N
CH
N
CF3
CF3
CF3
Cl
Cl
Cl
CF3
CF3
CF3
CF3
10
18
26
63
N
2-CF3
2-MeO
2-MeO
2,6-diMe
CH
CH
N
N
N
88
1407
13,850
5454
23,587
2,3-CH@CH–CH@CH–
2,3-CH@CH–CH@CH–
CH
46%@40 lM
*
FLIPR functional assay (with 1% BSA).6
In summary, a series of thiazolidinediones was discovered as
potent GPR40 agonists through systematic SAR studies of a screen-
ing lead. Among them, compound C possesses excellent pharmaco-
kinetic properties in mouse, and has demonstrated an acute
mechanism-based glucose-lowering in intraperitoneal glucose tol-
erance test in lean mice. Further SAR studies as well as optimiza-
tion of the overall profiles of these compounds will be reported
in due course.
References and notes
1. (a) Itoh, Y.; Kawamata, Y.; Harada, M.; Kobayashi, M.; Fujii, R.; Fukusumi, S.; Ogi,
K.; Hosoya, M.; Tanaka, Y.; Uejima, H.; Tanaka, H.; Maruyama, M.; Satoh, R.;
Okubo, S.; Kizawa, H.; Komatsu, H.; Matsumura, F.; Noguchi, Y.; Shinobara, T.;
Hinuma, S.; Fujisawa, Y.; Fujino, M. Nature 2003, 422, 173; (b) Briscoe, C. P.;
Tadayyon, M.; Andrews, J. L.; Benson, W. G.; Chambers, J. K.; Eilert, M. M.; Ellis,
C.; Elshourbagy, N. A.; Goetz, A. S.; Minnick, D. T.; Murdock, P. R.; Sauls, H. R.;
Shabon, U.; Spinage, L. D.; Strum, J. C.; Szekeres, P. G.; Tan, K. B.; Way, J. M.;
Ignar, D. M.; Wilson, S.; Muir, A. I. J. Biol. Chem. 2003, 278, 11303; for recent
reviews, see (c) Telvekar, V. N.; Kundaikar, H. S. Curr. Drug Targets 2008, 9, 899;
(d) Medina, J. C.; Houze, J. B. Ann. Rep. Med. Chem. 2008, 43, 75; (e) Bharate, S. B.;
Nemmani, K. V. S.; Vishwakarma, R. A. Exp. Opin. Ther. Pat. 2009, 19, 237.
2. Garrido, D. M.; Corbett, D. F.; Dwornik, K. A.; Goetz, A. S.; Littleton, T. R.;
McKeown, S. C.; Mills, W. Y.; Smalley, T. L.; Briscoe, C. P.; Peat, A. J. Bioorg. Med.
Chem. Lett. 2006, 16, 1840.
3. McKeown, S. C.; Corbett, D. F.; Goetz, A. S.; Littleton, T. R.; Bigham, E.; Briscoe, C.
P.; Peat, A. J.; Watson, S. P.; Hickey, D. M. B. Bioorg. Med. Chem. Lett. 2007, 17,
1584.
4. (a) Song, F. B.; Lu, S. F.; Gunnet, J.; Xu, J. Z.; Wines, P.; Proost, J.; Liang, Y.;
Baumann, C.; Lenhard, J.; Murray, W. V.; Demarest, K. T.; Kuo, G. H. J. Med. Chem.
2007, 50, 2807; (b) Tikhonova, I. G.; Sum, C. S.; Neumann, S.; Engel, S.; Raaka, B.
M.; Costanzi, S.; Gershengorn, M. C. J. Med. Chem. 2008, 51, 625; (c) Bharate, S. B.;
Rodge, A.; Joshi, R. K.; Kaur, J.; Srinivasan, S.; Kumar, S. S.; Kulkarni-Almeida, A.;
Balachandran, S.; Balakrishnan, A.; Vishwakarma, R. A. Bioorg. Med. Chem. Lett.
2008, 18, 6357; (d) Christiansen, E.; Urban, C.; Merten, N.; Liebscher, K.; Karlsen,
K. K.; Hamacher, A.; Spinrath, A.; Bond, A. D.; Drewke, C.; Ullrich, S.; Kassack, M.
U.; Kostenis, E.; Ulven, T. J. Med. Chem. 2008, 51, 7061; (e) Humphries, P. S.;
Benbow, J. W.; Bonin, P. D.; Boyer, D.; Doran, S. D.; Frisbie, R. K.; Piotrowski, D.
W.; Balan, G.; Bechle, B. M.; Conn, E. L.; Dirico, K. J.; Oliver, R. M.; Soeller, W. C.;
Southers, J. A.; Yang, X. J. Bioorg. Med. Chem. Lett. 2009, 19, 2400.
Figure 3. IPGTT results in WT and KO mice.
5. Kotarsky, K.; Nilsson, N. E.; Olde, B.; Owman, C. Pharmacol. Toxicol. 2003, 93, 249.
6. Fluorometric imaging plate reader (FLIPR) based intracellular calcium assay is
described herein: Tan, C. P.; Feng, Y.; Zhou, Y. P.; Eiermann, G. J.; Petrov, A.;
Zhou, C. Y.; Lin, S. N.; Salituro, G.; Meinke, P.; Mosley, R.; Akiyama, T. E.; Einstein,
M.; Kumar, S.; Berger, J. P.; Mills, S. G.; Thornberry, N. A.; Yang, L. H.; Howard, A.
D. Diabetes 2008, 57, 2211.
effects of these small molecule GPR40 agonists on glucose excur-
sion in normal lean mice. To this end, we were excited to find that
oral administration of Cpd C, one hour prior to dextrose challenge
in an intraperitoneal glucose tolerance test (IPGTT), significantly