Identification of fused-ring alkanoic acids with improved pharmacokinetic profiles that Act as G protein-coupled receptor 40/free fatty acid receptor 1 agonists

Article abstract of DOI:10.1021/jm2012968

The G protein-coupled receptor 40 (GPR40)/free fatty acid receptor 1 (FFA1) has emerged as an attractive target for a novel insulin secretagogue with glucose dependency. We previously identified phenylpropanoic acid derivative 1 (3-{4-[(2′,6′-dimethylbiphenyl-3-yl)methoxy]-2-fluorophenyl} propanoic acid) as a potent and orally available GPR40/FFA1 agonist; however, 1 exhibited high clearance and low oral bioavailability, which was likely due to its susceptibility to β-oxidation at the phenylpropanoic acid moiety. To identify long-acting compounds, we attempted to block the metabolically labile sites at the phenylpropanoic acid moiety by introducing a fused-ring structure. Various fused-ring alkanoic acids with potent GPR40/FFA1 activities and good PK profiles were produced. Further optimizations of the lipophilic portion and the acidic moiety led to the discovery of dihydrobenzofuran derivative 53 ((6-{[4′-(2-ethoxyethoxy)-2′,6′-dimethylbiphenyl-3-yl]methoxy} -2,3-dihydro-1-benzofuran-3-yl)acetic acid), which acted as a GPR40/FFA1 agonist with in vivo efficacy during an oral glucose tolerance test (OGTT) in rats with impaired glucose tolerance.

Full text of DOI:10.1021/jm2012968

Article
pubs.acs.org/jmc
Identification of Fused-Ring Alkanoic Acids with Improved
Pharmacokinetic Profiles that Act as G Protein-Coupled Receptor 40/
Free Fatty Acid Receptor 1 Agonists
Nobuyuki Negoro,*,† Shinobu Sasaki,† Masahiro Ito,† Shuji Kitamura,† Yoshiyuki Tsujihata,† Ryo Ito,†
Masami Suzuki,† Koji Takeuchi,† Nobuhiro Suzuki,‡ Junichi Miyazaki,‡ Takashi Santou,†
Tomoyuki Odani,† Naoyuki Kanzaki,† Miyuki Funami,‡ Toshimasa Tanaka,† Tsuneo Yasuma,‡
and Yu Momose†
†Pharmaceutical Research Division, Takeda Pharmaceutical Company Limited, 26-1 Muraoka-Higashi 2-chome, Fujisawa, Kanagawa
251-8555, Japan
‡Pharmaceutical Research Division, Takeda Pharmaceutical Company Limited, 17-85 Jusohonmachi 2-chome, Yodogawa-ku, Osaka
532-8666, Japan
S
* Supporting Information
ABSTRACT: The G protein-coupled receptor 40 (GPR40)/
free fatty acid receptor 1 (FFA1) has emerged as an attractive
target for a novel insulin secretagogue with glucose depend-
ency. We previously identified phenylpropanoic acid derivative
1 (3-{4-[(2′,6′-dimethylbiphenyl-3-yl)methoxy]-2-
fluorophenyl}propanoic acid) as a potent and orally available
GPR40/FFA1 agonist; however, 1 exhibited high clearance
and low oral bioavailability, which was likely due to its
susceptibility to β-oxidation at the phenylpropanoic acid moiety. To identify long-acting compounds, we attempted to block the
metabolically labile sites at the phenylpropanoic acid moiety by introducing a fused-ring structure. Various fused-ring alkanoic
acids with potent GPR40/FFA1 activities and good PK profiles were produced. Further optimizations of the lipophilic portion
and the acidic moiety led to the discovery of dihydrobenzofuran derivative 53 ((6-{[4′-(2-ethoxyethoxy)-2′,6′-dimethylbiphenyl-
3-yl]methoxy}-2,3-dihydro-1-benzofuran-3-yl)acetic acid), which acted as a GPR40/FFA1 agonist with in vivo efficacy during an
oral glucose tolerance test (OGTT) in rats with impaired glucose tolerance.
leads to metabolic syndrome, a series of diseases that include
type 2 diabetes, obesity, and hyperlipidemia.9
INTRODUCTION
■
The G protein-coupled receptor 40 (GPR40)/free fatty acid
receptor 1 (FFA1), which is preferentially expressed in
pancreatic β-cells, was identified as a receptor for medium- to
long-chain free fatty acids (FFAs) such as oleic acid and linoleic
acid.1−3 The activation promotes the phospholipase C-
mediated hydrolysis of phosphatidylinositol 4,5-bisphosphate
into diacylglycerol and inositol triphosphate, activating protein
kinase C and mobilizing intracellular Ca2+ from the
endoplasmic reticulum. The elevated intracellular Ca2+
concentration resulting from the activation of GPR40/FFA1
by FFAs may amplify glucose-stimulated insulin secretion
(GSIS).4,5 Because the downregulation of GPR40/FFA1 by
small interfering RNA suppressed the augmentation of FFA-
induced insulin secretion in mouse insulinoma MIN6 cells,1
GPR40/FFA1 is thought to be involved in acute insulin
secretion by FFAs.
Insulin secretagogues such as sulfonylureas and meglitinides
are widely used as one of the first-line drugs in type 2 diabetes
therapy.10 Although sulfonylureas have strong insulinotropic
potencies, they promote insulin secretion, even with normal
plasma glucose levels.11 Thus, these agents can cause
hypoglycemia and apoptosis of pancreatic β-cells with long-
term therapy (secondary failure).12,13 To eliminate these safety
concerns, plasma concentrations of these agents must be
carefully controlled. Meglitinides are short-acting insulinotropic
agents, which do not have as great a potential for hypoglycemia
compared to sulfonylureas. Meglitinides must be administered
immediately prior to each meal, which makes their use
inconvenient for patients, and their effects are limited to the
improvement of postprandial hyperglycemia. Because of these
factors, agents that promote GSIS have the potential to replace
sulfonylureas as a first-line therapy in patients with type 2
diabetes; for example, dipeptidyl peptidase-4 (DPP-4) inhib-
Although FFAs acutely potentiate insulin secretion in
response to plasma glucose levels,6 chronically elevated FFAs
in plasma exhibit pleiotropic effects, including impaired β-cell
function and cell death.7,8 Additionally, an excess amount of
FFAs is an important component in insulin resistance, which
Received: September 29, 2011
Published: January 16, 2012
© 2012 American Chemical Society
1538
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
Figure 1. Representative GPR40/FFA1 agonists.
Figure 2. Confirmed structures of metabolites in plasma after oral administration of compound 1 in rats.
Figure 3. Design of fused-ring alkanoic acids.
itors or peptide-stable analogues of glucagon-like peptide-1
(GLP-1)14 have got much attention as a new-generation of
antidiabetic drugs.
pancreatic β-cells indicated exclusive efficacy on islets and the
reduced side effects derived from GPR40/FFA1 in other
tissues.
In the development of a novel insulinotropic agent, GPR40/
FFA1 is attractive because of a number of reasons. First,
GPR40/FFA1 agonists can directly increase Ca2+ concen-
trations in pancreatic β-cells, enabling a robust insulinotropic
effect on both postprandial and fasting hyperglycemia. Second,
because GPR40/FFA1-mediated insulinotropic actions are
glucose-dependent, there may be a low risk of hypoglycemia;
thus, we aimed to design a drug with a long duration in plasma.
Furthermore, the dominant expression of GPR40/FFA1 in
Figure 1 summerizes a variety of synthetic GPR40/FFA1
agonists such as phenylpropanoic acid and thiazolidinedione
derivatives reported by several groups14−21 and also represents
our clinical candidate TAK-875 under phase 3 development.22
We independently identified a diverse series of GPR40/FFA1
agonists as type 1 using an endogenous ligand-based design
(Figure 2).23 The compound exhibited potent agonist activity
and a significant insulinotropic effect in diabetic animal models.
In this compound series, introduction of substituents at the
1539
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
a
Scheme 1
a
Reagents and conditions: (a) AlCl3, toluene, reflux; (b) benzyl bromide, K2CO3, acetone, reflux; 91−94% (2 steps); (c) triethyl phosphonoacetate,
NaH, toluene, reflux; (d) H2 (balloon pressure), 10% Pd/C, EtOH or MeOH, rt, 54−89% (2 steps), 76−100% for 7e,f; (e) AlCl3, 1-dodecanethiol,
toluene, rt, 98%; (f) ethyl 4-chloroacetoacetate, H2SO4, rt, 84%; (g) 1 M NaOH aq, reflux, 83%; (h) H2SO4, MeOH, reflux, 70%; (i) benzyl chloride,
KF, CH3CN, reflux; 43%; (j) ethyl 4-bromobutyrate, Cs2CO3, DMF, 80 °C, 41%; (k) 2,6-dimethylphenylboronic acid, Pd(PPh3)4, 1 M Na2CO3 aq,
EtOH, toluene, reflux, 97%; (l) NaBH4, DME, THF, 0 °C, 83%; (m) 7a−f, ADDP, P(n-Bu)3, toluene, rt, 23−96%; (n) 2 M NaOH aq, MeOH or
EtOH, THF, rt, 53−80%.
ortho-position of the phenylpropanoic acid moiety might
improve susceptibility to β-oxidation and the resulting ortho-
fluoro derivative 1 showed moderate bioavailability. However,
the β-oxidation was not suppressed completely and the β-
oxidation metabolites of 1 such as cinnamic acid 2 and benzoic
acid 3 were actually observed in rat plasma (Figure 2). The PK
profile of 1 was not yet sufficient, especially regarding the
duration in plasma. Therefore, to develop a drug administered
once daily for the convenience for patients, we thought that it
would be needed to further improve the duration in plasma.
With a view to improving metabolic stability of 1 against β-
oxidation again, we developed a design strategy to block the α-
or β-position of the propanoic acid moiety by constructing a
fused-ring structure between the ortho-position of the phenyl
ring and either site of the propanoic acid (α- or β-position)
(Figure 3).
Herein, we describe the identification of fused-ring alkanoic
acids as novel GPR40/FFA1 agonists with low clearance and
high plasma exposure in rats consistent with the blockade of β-
oxidation. Additionally, we evaluated the species differences
between humans and rats in the binding affinity using a
GPR40/FFA1 homology model and determined the in vivo
efficacy by conducting an oral glucose tolerance test (OGTT)
in female Wistar fatty rats.
CHEMISTRY
■
The syntheses of the fused-ring alkanoic acids 14−19 are
detailed in Scheme 1.22 The preparation of the 5-
hydroxyindane intermediate 7a began with the conversion of
5-methoxy-1-indanone (5a) to 5-benzyloxy-1-indanone (6a).
Horner−Wadsworth−Emmons olefination of 6a, which was
followed by catalytic hydrogenation, afforded 7a. The 6-
membered analogue 7b was synthesized from 6-methoxy-1-
tetralone (5b) without exchanging the protecting group as
shown in the synthesis of 7a. Incorporation of the acetic acid
moiety, followed by deprotection of the methoxy group using
Node’s method,24 produced 7b. Similar conditions to 7a
provided the seven-membered analogue 7c from 2-methoxy-
6,7,8,9-tetrahydro-5H-benzo[7]annulen-5-one (5c), easily pre-
pared from 3-methoxybenzaldehyde in a four-step sequence.
The synthesis of the benzofuran intermediate 7d began with
resorcinol (8), followed by Pechmann reaction25 to afford 7-
hydroxy-4-chloromethylcoumarin (9). Treatment of 9 with an
aqueous base triggered the ring-opening and recyclization to
form the benzofuran ring,26 followed by esterification to yield
the desired product 7d. Hydrogenation of 7d resulted in the
dihydrobenzofuran intermediate 7e. The synthesis of 8-hydoxy-
tetrahydrobenzoxepine 7f commenced with the monobenzyla-
tion of 2,4-dihydroxybenzaldehyde (10), which was followed by
alkylation with ethyl 4-bromobutyrate and simultaneous
cyclization to yield 11. The catalytic hydrogenation of the
1540
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
a
Scheme 2
a
Reagents and conditions: (a) bromine, AcOH, rt, 88%; (b) cyanoacetic acid, NH4OAc, pyridine, toluene, reflux, 80%; (c) NaBH4, MeOH, satd
NaHCO3 aq, rt, 98%; (d) DMA, 180 °C, 88%; (e) NaNH2, NH3 aq, −33 °C, 48%; (f) AlCl3, 1-dodecyl methyl sulfide, toluene, 0 °C, 79%; (g) 13,
ADDP, P(n-Bu)3, toluene or THF, rt, 62−90%; (h) KOH, EtOH, H2O, rt to reflux, 82−99%; (i) LiAlH4, THF, rt, 88%; (j) p-TsCl, pyridine, rt, 89%;
(k) NaCN, DMSO, rt, 90%; (l) silica gel, NH3 aq, rt, 67%; (m) methyl propiolate, toluene, reflux; (n) heat, DMF, reflux, 34% (2 steps); (o) POCl3,
120 °C, 89%; (p) NBS, AIBN, CCl4, reflux, 63%; (q) diethyl malonate, NaH, toluene, rt, 74%; (r) NaH, toluene, reflux, 99%; (s) H3PO4, 185 °C,
85%; (t) 13, MsCl, Et3N, THF then 30, K2CO3, DMF, 70 °C; (u) NaBH4, MeOH, THF, 0 °C, 6% (2 steps); (v) SOCl2, pyridine, toluene, rt; (w)
diethyl malonate, NaH, THF, rt, 53% (2 steps); (x) 2 M NaOH aq, EtOH, THF, 0 °C then toluene, reflux, 38%; (y) ethyl bromoacetate, NaH, THF,
DMF, 4 °C to rt, 83%; (z) KOH aq, EtOH, THF, rt, 76%.
olefin furnished debenzylated product 7f. The left-hand
biphenylylmethanol 13 was prepared by Suzuki coupling and
subsequent reduction with sodium borohydride. Condensation
of 13 with phenols 7a−f by Mitsunobu reaction, followed by
hydrolysis, afforded the desired carboxylic acids 14−19.
finally the following thermal decarboxylation afforded the target
compound 32. The synthesis of the indole derivative 35 was
accomplished via attachment of the lipophilic region, insertion
of the acetate moiety, and basic hydrolysis.
Next, we fixed (2,3-dihydro-1-benzofuran-3-yl)acetic acid as
the acidic region and modified the lipophilic region (Scheme
3). The alcohol intermediates 42a−g were prepared by Suzuki
coupling between the appropriate bromide and boronic acid,
followed by the reduction of the ester group or the formyl
group. The synthesis of the 4′-alkoxybiphenyl intermediates
42h−i were obtained via alkylation of the 4′-hydroxybiphenyl
intermediate 44. Condensation of the obtained alcohols 42a−i
with the dihydrobenzofuran intermediate 7e, followed by basic
hydrolysis, afforded the desired carboxylic acids 45−53.
Next, we again conducted the modification of the acid region
with fixing the lipophilic side-chain as the {4′-(2-ethoxyethoxy)-
2′,6′-dimethylbiphenyl-3-yl}methyl group (Scheme 4). The
synthesis of the indan-2-carboxylate intermediate 56a or the
tetrahydronaphthalene-2-carboxylate intermediate 56c began
with the insertion of an ethoxycarbonyl group into the 6-
methoxy-1-indanone (54) or 6-methoxy-1-tetralone (57),
respectively. Reductive decarbonylation with TFA/triethylsi-
lane, followed by demethylation, afforded the desired
compounds 56a and 56c. The synthesis of the isobenzofuran
intermediate 56e was completed in a five-step sequence.
Sandmeyer hydroxylation of 5-aminophthalide (59) and
benzylation of the consequent hydroxy group gave the
Scheme 2 describes the synthesis of other fused-ring
analogues (25, 32, and 35). Construction of the benzocyclo-
butene backbone was achieved using a known method.27 The
propanenitrile intermediate 22, prepared from 3-methoxyben-
zaldehyde (20) in a four-step sequence, was treated with
sodium amide in liquid ammonia, followed by demethylation
with dodecyl methyl sulfide28 to afford 23. Introduction of the
biphenyl region into 23 and subsequently a one-carbon
homologation sequence provided the desired benzocyclobute-
neacetic acid 25. The synthesis of dihydrocyclopenta[b]-
pyridine derivative 32 was accomplished through multistep
reactions. The pyridone intermediate 27 was prepared in a
three-step sequence using a reported procedure,29,30 and it was
then treated with phosphorus oxychloride to elaborate
chloropyridine 28. After the benzylic bromination of 28, the
insertion of a malonate unit and the subsequent cyclization with
a base afforded the ketoester 29, which was decarboxylated by
heating to give hydroxypyridine 30. Next, introduction of a
biphenyl region to 30 and the subsequent reduction of the
carbonyl group produced the alcohol 31. Subsequent
chlorination of 31 with thionyl chloride, alkylation of the
resulting chloride with malonate anion, basic hydrolysis, and
1541
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
a
Scheme 3
a
Reagents and conditions: (a) ArBr, Pd(PPh3)4, Cs2CO3 or Na2CO3, EtOH, toluene, H2O, 70 °C, 65−94%; (b) ArB(OH)2, Pd(PPh3)4, Cs2CO3,
EtOH, toluene, 70−80 °C, 76%−quant; (c) NaBH4, EtOH or DME−THF or MeOH−THF, 0 °C, 70−99%; (d) LiAlH4, THF, 0 °C to rt, 95−96%;
(e) 2,6-dimethylphenylboronic acid, Pd2(dba)3, SPhos, K3PO4, toluene, H2O, 100 °C, quant; (f) R3-X, K2CO3, (KI), DMF, 70 °C, 89−92%; (g) 42,
ADDP, P(n-Bu)3, toluene, rt, 50−93%; (h) 2 M NaOH aq, MeOH, THF, rt, 55−92%.
benzyloxy compound 60. Treatment with an organolithium
reagent that was generated from tert-butyl acetate and the
subsequent deoxygenation with TFA/triethylsilane, followed by
debenzylation, produced the desired intermediate 56e. The
target carboxylic acids 61−65 were obtained under the similar
conditions described above. The synthesis of the oxyindolyl-
acetic acid 68 was initiated with 5-methoxyisatin (66). After
attachment of the acetate group, the obtained compound was
subjected to decarbonylation under hydrogenation conditions
in acidic solvent to afford the hydrolyzed oxyindole derivative,
which was reesterified and subsequently demethylated to result
in the phenol intermediate 67. Condensation of the lipophilic
portion, followed by hydrolysis under acidic condition,
produced the desired compound 68. The 6-aminodihydroben-
zofuran derivative 73 was synthesized from 3-aminophenol
(69) utilizing the 2-nitrobenzenesulfonyl (Ns) group.31 After
introduction of a Ns group, a similar reaction sequence that was
used for the synthesis of 7d afforded the benzofuran 71, which
was alkylated using Mitsunobu reaction, followed by
denosylation with mercaptoacetic acid to give 72. Finally, the
resultant benzofuran 72 was reduced under catalytic hydro-
genation conditions, followed by ester hydrolysis to give the
desired compound 73 as a hydrochloride salt.
RESULTS AND DISCUSSION
■
Agonist activities of the synthesized compounds were measured
with a fluorometric imaging plate reader (FLIPR) assay in
Chinese hamster ovary (CHO) cells expressing human
GPR40/FFA1 in the presence of 0.1% bovine serum albumin
(BSA). Binding affinities for human and rat receptors were also
measured using the membrane fraction of stably expressing
GPR40/FFA1 in the presence of 0.2% BSA.22
First, we examined whether various fused-ring alkanoic acid
derivatives, which were designed to suppress β-oxidation,
retained potent GPR40/FFA1 activity (Table 1). Nonaromatic
four- to six-membered ring fused-benzene derivatives, such as
14, 15, and 25, showed agonist activities comparable to the
phenylpropanoic acid 1, whereas the seven-membered ring
acetic acid 16 displayed decreased agonist activity. Interestingly,
moving the attachment site from the β-position to the α-
position of the propanoic acid moiety, as with 2,3,4,5-
tetrahydro-1-benzoxepine-4-carboxylic acid 19, recovered the
agonist activity. Thus, proper placement of the carboxylic group
and the phenyl ring may be important for the GPR40/FFA1
agonist activity, regardless of the size of the fused-ring or the
construction of a fused-structure at either the α- or β-position
1542
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
a
Scheme 4
a
Reagents and conditions: (a) diethyl carbonate, NaH, toluene, rt to 120 °C, 51−70%; (b) triethylsilane, TFA, rt; (c) AlCl3, 1-octanethiol, CH2Cl2, 0
°C, 32−67% (2 steps), 81% for 67-step 3; (d) H2SO4, sodium nitrite, H2O, 0 °C to reflux; (e) benzyl bromide, K2CO3, DMF, 60 °C, 21% (2 steps);
(f) tert-butyl acetate, lithium diisopropylamide, THF, −78 °C, (g) triethylsilane, TFA, CH2Cl2, rt, 16% (2 steps); (h) H2, Pd/C, EtOH, rt, 86%; (i)
42i, ADDP, (n-Bu)3P, toluene, rt, 22−85%; (j) 1 or 2 M NaOH aq, MeOH or EtOH, THF, rt, 57−91%; (k) TFA, toluene, rt, 72%; (l) ethyl
bromoacetate, NaH, DMF, 0 °C to rt, 78%; (m) H2 (balloon pressure), 10% Pd/C, 70% perchloric acid, AcOH, 50 °C, then SOCl2, EtOH, rt, 36%;
(n) 60% HClO4, AcOH, 50 °C, 18%; (o) 2-nitrobenzenesulfonyl chloride, pyridine, rt, 77%; (p) ethyl 4-chloroacetoacetate, H2SO4, rt, 51%; (q) 1 M
NaOH aq, rt, quant; (r) SOCl2, MeOH, rt, 62%; (s) 42i, DEAD, PPh3, toluene, rt; (t) mercaptoacetic acid, LiOH·H2O, DMF, rt 76% (2 steps); (u)
H2 (balloon pressure), 10% Pd/C, EtOH, rt, 66%; (v) 2 M NaOH aq, EtOH, THF, rt, then 4 M HCl/AcOEt, Et2O, rt, 85%.
of the propanoic acid moiety. Replacement of the methylene
linker at the 3-position of the dihydroindene ring (14) with an
ether linker (18) retained the agonist activity, suggesting that
the oxygen atom on the fused-ring had little impact on the
agonist activity. Aromatic fused-ring compounds with planar
structures, such as benzofuran (17) and indole (35), exhibited
reduced agonist activities. Nonaromatic ring fused-benzene
derivatives bearing an sp3 carbon at the β-position of the
propanoic acid moiety appeared to be in a more favorable
position of the carboxylic acid relative to compounds bearing an
sp2 carbon. Finally, introduction of a nitrogen atom into the
benzene ring, as with the fused pyridine derivative 32,
attenuated the agonist activity compared to the parent
compound 14. Generally, a pyridine ring is known as a π−
electron-deficient heterocycle, so we speculated that the erosion
of potency on 32 might be attributed to the decreased
capability of a π−electron interaction with the GPR40/FFA1
receptor because the π−π interaction between the ligand and
the receptor may be important for exerting GPR40/FFA1
activity, as we reported previously.22,23
Next, we screened the GPR40/FFA1 binding affinities in
order to identify species differences between humans and rats;
this information can be used to examine the in vivo
insulinotropic effects and glucose-lowering effects during an
OGTT using female Wistar fatty rats. Generally, human
binding affinities were correlated with human agonist activities
as shown in Table 1. However, for the rat receptor, the smaller
ring was preferable compared to the larger ring (five-membered
ring 14, 18 > six-membered ring 15 > seven-membered ring
19), except for the four-membered ring 25.
1543
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
Table 1. In Vitro Activities of Fused-Ring Alkanoic Acids
a
EC50 values are averages of n = 3 in the presence of 0.1% BSA. EC50 values and 95% confidence intervals of each compound were obtained with
b
c
Prism 5 software (GraphPad). All values are averages of n = 2 in the presence of 0.2% BSA. Log D values were determined at pH 7.4 according to a
reported method.33 Not determined (101% increase of control at 10 μM, 2% increase of control at 1 μM).
d
Thus, we identified various nonaromatic ring fused-benzene
derivatives with potent GPR40/FFA1 agonist activities and
human and rat binding affinities. Among these, the five-
membered fused-ring compounds (14 and 18) showed potent
activities and an acceptable profile regarding species differences.
Meanwhile, higher lipophilicity of the compounds might be
correlated with undesirable absorption, distribution, metabo-
lism, excretion, and toxicology (ADME-Tox) profiles.32
Accordingly, we selected the (2,3-dihydro-1-benzofuran-3-
yl)acetic acid derivative 18 as a template for further
investigation based on its low lipophilicity (see Log D value).
Having identified a favorable scaffold that was expected to
show a long-lasting PK profile, we then shifted our focus on
evaluating the lipophilic portion of the structure (Table 2). In a
previous report,23 we demonstrated that the orthogonal
conformation of the biphenyl moiety was important in
enhancing agonist activity and reducing the influence of
binding to serum albumin. With these observations in mind,
1544
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
Table 2. In Vitro Activities of (2,3-Dihydro-1-benzofuran-3-yl)acetic Acids
a
EC50 values are averages of n = 3 in the presence of 0.1% BSA. EC50 values and 95% confidence intervals of each compound were obtained with
b
c
Prism 5 software (GraphPad). All values are averages of n = 2 in the presence of 0.2% BSA. The Log D values were determined at pH 7.4
according to a reported method.33
we first explored bicyclic aromatic rings as an R1 substituent
that would be likely to produce a twisted conformation. The
most sterically hindered 2-methyl-1-naphthyl analogue 45
showed comparable agonist activity and higher binding
affinities in humans and rats relative to the parent compound
18. The bioisosteric replacement of the naphthyl group into the
3-benzothienyl group (46) showed comparable agonist activity
but reduced binding affinities that were likely due to the
absence of an ortho-methyl group, and the less hindered 5-
benzothienyl analogue 47 exhibited lower agonist activity and
binding affinities. Next, we evaluated the R2 group, the
substituent at the 6-position of the central phenyl ring. The
6-methoxy analogue 48 retained activity comparable to the
parent compound 18, and the more sterically bulky 6-benzyloxy
analogue 49 displayed significantly improved activity. These
results demonstrated the expected results that the steric
1545
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
Table 3. In Vitro Activities of Fused-Ring Alkanoic Acids
a
EC50 values are averages of n = 3 in the presence of 0.1% BSA. EC50 values and 95% confidence intervals of each compound were obtained with
b
c
Prism 5 software (GraphPad). All values are averages of n = 2 or 3 in the presence of 0.2% BSA. The Log D values were determined at pH 7.4
according to the reported method.33 Not determined (59% increase of control at 10 μM, 2% increase of control at 1 μM).
d
bulkiness of the biaryl moiety was important for enhancing the
activity of this series. Next, incorporation of a substituent at the
4′-position of the biphenyl group was investigated. Displace-
ment of one methyl group on the 2′,6′-dimethyl analogue 18 to
the 4′-position (50) retained agonist activity and binding
affinities in humans and rats. Furthermore, the 2′,4′,6′-trimethyl
analogue 51 also retained potency for the human receptor and
showed more potent binding affinities for both human and rat
receptors. These results indicated that the introduction of a
substituent at the 4′-position was found to be favorable for
human and rat binding affinities. Both the 4′-benzyloxy
analogue 52 and the 4′-(2-ethoxyethoxy) analogue 53 showed
potent agonist activities, suggesting the potential of the
incorporation of various substituents, especially a certain level
of a polar functionality for modulating drug-like parameters, at
the 4′-position of the biphenyl ring.
ring type and the tether on the carboxylic acid moiety, we again
explored the acidic portion with the aim of optimally projecting
the carboxylic acid moiety toward the acid-binding residue
(Table 3). We selected the 4′-(2-ethoxyethoxy)-2′,6′-dimethyl-
biphenylyl group on 53 with a relatively low Log D value (3.83)
and good PK profile (vide infra) as the lipophilic portion and
first screened indane and tetrahydronaphthalene derivatives
(61−64). Unfortunately, all analogues displayed decreased
potency, particularly with respect to rat receptor binding
affinity. On the other hand, the isobenzofuran analogue 65
retained agonist activity; furthermore, 65 exhibited higher
affinity to rat receptors than to human receptors. The oxyindole
derivative 68, which was intended to increase the polarity of the
fused-ring moiety (see low Log D value), showed decreased
potency. This result implied that the carbonyl group of 68
might interrupt the proper interaction with the receptor.
Replacement of the linker oxygen of 53 to nitrogen (73)
minimally affected the agonist activity but decreased the Log D
value (ca. 0.7). Subsequently, we found that (2,3-dihydro-1-
As described previously, the dihydrobenzofuran derivatives
possessed high affinity for the human receptor but reduced
affinity for the rat receptor. To examine the influence of the
1546
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
benzofuran-3-yl)acetic acid was the most effective acidic moiety
in this structural series and discovered that the nitrogen linker
also appeared to be optimal for potency.
Next, we considered the species differences between human
and rat GPR40/FFA1 using homology modeling based on the
crystal structure of bovine rhodopsin. The method for
modeling and ligand docking was described in a previous
paper.22 GPR40/FFA1 is well-conserved between humans and
rats; rat GPR40/FFA1 has a 75% identity to human GPR40/
FFA1 at the DNA level and an 82% identity to human GPR40/
FFA1 at the protein level.2 Among the amino acid residues that
interact with the ligand in the binding pocket, the only amino
acid residue that varies between human and rat might be
Leu186 (TM5), which is replaced with Phe in rat GPR40/
FFA1. We constructed a rat GPR40/FFA1 model by replacing
Leu186 with Phe. Then, we docked the dihydrobenzofuran
derivative 18 into this model and superimposed the six-
membered analogue 15 and the seven-membered analogue 19
(Figure 4A). In this model, the fused-rings of these compounds
were located adjacent to Phe186, indicating that the larger ring
caused steric repulsion with the phenyl group of Phe186. Thus,
the species differences in the potency between human and rat
might result from a decreased affinity due to the substitution of
Leu186 in the human GPR40/FFA1 receptor with the sterically
bulky Phe in the rat GPR40/FFA1 receptor.
Next, we compared the binding mode of the representative
compound 53 with 2-ethoxyethoxy group at the 4′-position of
the biphenyl ring and the dihydrobenzofuran lead 18 in the
human GPR40/FFA1 receptor (Figure 4B). Compound 53
overlaps well with 18, and the proposed binding mode of 53
shows several interactions of the {6-[(2′,6′-dimethylbiphenyl-3-
yl)methoxy]-2,3-dihydro-1-benzofuran-3-yl}acetic acid moiety.
The carboxylic acid group interacts with Arg183 (TM5) and
Arg258 (TM7), and the benzyl moiety forms a π−π interaction
with Trp72 (E-I loop). In addition, 2-ethoxyethoxy group that
is characteristic of 53 extends to the space between TM1 and
TM7 and might interact with Tyr12 (TM1) and Lys259
(TM7). This information also afforded the opportunity to
investigate the effect of the substituent at the 4′-position of the
biphenyl ring.
Table 4 shows our data regarding the rat PK profiles for the
selected fused-ring alkanoic acids. We initially investigated the
relationship between the fused-ring structure and PK profiles
(indane 14, tetrahydronaphthalene 15, dihydrobenzofuran 18,
and tetrahydrobenzoxepine 19). Each compound showed
superior PK profiles, in particular, a lower clearance and a
higher plasma exposure (see AUCpo,0−8 h) than phenyl-
propanoic acid 1 as we had expected. Among these, cyclic
ether derivatives (18 and 19) exhibited better PK profiles
compared to cycloalkane analogues (14 and 15). This
improvement was likely due to the blocking of the
metabolically labile benzylic position on the fused-ring (Figure
5, site A). The most potent 6-benzyloxy analogue 49 showed
high clearance, which resulted in decreased plasma exposure
(Figure 5, site B), but its bioavailability was still good (F =
52.4%). The rapid clearance of 49 from plasma was probably
due to its high tissue distribution consistent with its high
lipophilicity (Log D value: 4.53). Introduction of the methyl
group into the 4′-position of the distal phenyl group (51)
decreased the plasma exposure (AUCpo,0−8 h = 698.7 ng·h/mL).
In contrast, 4′-alkoxy analogues (52 and 53), masked at the
benzylic position of the distal phenyl ring (Figure 5, site C)
with an oxygen atom, showed a good PK profile. In particular,
Figure 4. (A) Docking model of GPR40/FFA1 in complex with 15
(yellow), 18 (red), and 19 (blue). (B) Overlay of 18 (red) and 53
(white) in complex with human GPR40/FFA1. (C) Two-dimensional
diagram showing the interaction of 53 with human GPR40/FFA1
constructed by MOE.
the 4′-ethoxyethoxy analogue (53) exhibited the lowest
clearance (211 mL/h/kg) and the highest plasma exposure
(AUCpo,0−8 h = 2837.4 ng·h/mL) among this series. Analogues
1547
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
a
alkanoic acid moiety, indicating that constructing a fused-ring
structure would be an efficient strategy for identifying long-
acting compounds.
Table 4. PK Profiles for Fused-Ring Alkanoic Acids
CLtotal
Cmax
(ng/mL)
Tmax
(h)
AUCpo,0−8 h
(ng·h/mL)
F
(%)
compd
(mL/h/kg)
On the basis of its favorable in vitro activity and PK profile,
compound 53 was selected and assessed for its ability to
improve glucose tolerance during an OGTT in female Wistar
fatty rats, which were used as a model developing obesity and
obesity-related features such as impaired glucose tolerance,
hyperinsulinemia, and hyperlipidemia.34 To evaluate both the
intrinsic and durable efficacy of 53, the compound was orally
administered 1 and 4 h prior to the glucose challenge (1 g/kg)
(these tests were referred as 1H-OGTT and 4H-OGTT,
respectively). Single oral dosing of 53 (1, 10 mg/kg) reduced
plasma glucose levels (Figure 6A) and augmented insulin
secretion (Figure 6C) during 1H-OGTT. The minimum
effective dose of the area under the curve of plasma glucose
14
15
18
19
49
51
52
53
65
73
1
464
782
296
293
1708
625
414
211
573
348
900
285.1
220.7
449.7
606.6
84.9
1.67
1.33
2.67
1.33
1.00
2.67
2.00
2.67
1.33
1.67
1.50
1701.3
925.7
47.8
56.9
69.4
68.6
52.4
43.3
69.4
59.5
49.2
70.2
21.5
2357.3
2801.4
308.9
162.8
316.9
465.4
193.1
439.0
86.0
698.7
1687.8
2837.4
871.2
2230.6
249.0
a
Rat cassette dosing at 0.1 mg/kg, iv and 1 mg/kg, po. All values are
averages of 3 rats. F indicates bioavailability.
(AUC0−120 min) (Figure 6B) and plasma insulin (AUCpre−120 min
)
(Figure 6D) was 10 mg/kg. Furthermore, 53 significantly
improved glucose tolerance and increased plasma insulin levels
during 4H-OGTT at a similar dose as 1H-OGTT (10 mg/kg)
(Figure 7A−D). Thus, the dihydrobenzofuran derivative
showed durable efficacy due to its desirable PK profile, so
this structural series was considered to have a superior profile
compared to the phenylpropanoic acid series with high plasma
clearance. Additionally, because dihydrobenzofuran derivatives
possess higher affinities to the human receptor than to the rat
receptor, they are expected to show more potent efficacy in
humans. These results encouraged us to focus on a
dihydrobenzofuran derivative as a promising lead compound
Figure 5. Presumed metabolic sites of fused-ring alkanoic acids.
with the amino linker (73) also showed high exposures and
good bioavailabilities that were comparable to the oxygen
analogue 53. Overall, the fused-ring alkanoic acids possessed
good PK profiles such as low clearance and high plasma
exposure, presumably by blocking β-oxidation of the phenyl-
Figure 6. Effects of 53 during an 1H-OGTT in female Wistar fatty rats. (A) and (C) show time-dependent changes of plasma glucose (PG) and
plasma insulin levels after single oral doses of 53 (1, 10 mg/kg) followed by 1 g/kg oral glucose challenge, respectively. Data in (B) and (D)
represent AUC0−120 min of PG levels and incremental AUC0−120 min of plasma insulin levels shown in (A) and (C), respectively. Values are mean SD
(n = 6). #: P ≤ 0.025 compared with control by one-tailed Williams’ test.
1548
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
Figure 7. Effects of 53 during an 4H-OGTT in female Wistar fatty rats. (A) and (C) show time-dependent changes of plasma glucose (PG) and
plasma insulin levels after single oral doses of 53 (10 mg/kg) followed by 1 g/kg oral glucose challenge, respectively. Data in (B) and (D) represent
AUC0−120 min of PG levels and incremental AUC0−120 min of plasma insulin levels shown in (A) and (C), respectively. Values are mean SD (n = 5−
6). +, P ≤ 0.05; ++, P ≤ 0.01 compared with control by Aspin−Welch test.
to identify new chemical entities with potent, durable, and safe
profiles.
EXPERIMENTAL SECTION
■
Chemistry. Reagents and solvents were obtained from commercial
sources and used without further purification. Reaction progress was
determined by thin layer chromatography (TLC) analysis on Merck
Kieselgel 60 F254 plates or Fuji Silysia NH plates. Chromatographic
purification was carried out on silica gel columns (Merck Kieselgel 60,
70−230 mesh or 230−400 mesh, Merck or Chromatorex NH-DM
1020, 100−200 mesh) or on Purif-Pack (SI: 60 μM or NH: 60 μM,
Fuji Silysia Chemical, Ltd.). Melting points were determined on a
CONCLUSION
■
We identified a series of fused-ring alkanoic acids as potent and
orally available GPR40/FFA1 agonists. Cyclization of the
phenylpropanoic acid moiety improved the PK profile,
especially with respect to clearance and plasma exposure.
This effect is consistent with the suppression of metabolic
pathways for β-oxidation. Our bioisosteric replacement of the
propanoic acid moiety with the fused-ring alkanoic acid moiety
provided an efficient strategy for retaining the activity against
the target molecule and improving the PK profile. Among this
series, (2,3-dihydro-1-benzofuran-3-yl)acetic acid was identified
as the best acidic moiety regarding human agonist activity and
rat binding affinity, and optimization of its biaryl moiety led to
the discovery of 4′-(ethoxyethoxy)-2′,6′-dimethylbiphenyl de-
rivative 53 with the lowest clearance and the highest plasma
exposure. Compound 53 exhibited a plasma glucose-lowering
effect and insulinotropic effect at an oral dose of 10 mg/kg in
both 1H-OGTT and 4H-OGTT in female Wistar fatty rats.
While 53 presented the concept of a long-acting GPR40
agonist, the effective dose (10 mg/kg) was relatively high in
1H-/4H-OGTTs in female Wistar fatty rats. In addition, the
lipophilicity of 53 was still high (Log D value: 3.83). Further
optimization to improve potency, physicochemical properties,
and ADME profiles suitable for a clinical candidate, resulting in
the discovery of TAK-875, will be reported in a forthcoming
publication.
̈
BUCHI B-545 melting point apparatus and were uncorrected. Proton
nuclear magnetic resonance (1H NMR) spectra were recorded on
Bruker Ultra Shield-300 (300 MHz) instruments. Chemical shifts are
given in parts per million (ppm) with tetramethylsilane as an internal
standard. Abbreviations are used as follows: s = singlet, d = doublet, t
= triplet, q = quartet, m = multiplet, dd = doublet of doublets, dt =
doublet of triplets, td = triplet of doublets, ddd = doublet of doublets
of doublets, br = broad. Coupling constants (J values) are given in
hertz (Hz). Low-resolution mass spectra (MS) were determined on a
Waters liquid chromatography−mass spectrometer system (MS),
using a CAPCELL PAK UG-120 ODS (Shiseido Co., Ltd.) column
(2.0 mm i.d. × 50 mm) with aqueous CH3CN (10−95%) containing
0.05% trifluoroacetic acid (TFA) and an HP-1100 (Agilent
Technologies) apparatus for monitoring at 220 nm. All MS
experiments were performed using electrospray ionization (ESI) in
positive ion mode. Analytical HPLC was performed on a Shimadzu
LC-VP instrument, equipped with CAPCELL PAK C18 UG120 S-3
μm, 2.0 mm × 50 mm column with a 4 min linear gradient from 90/10
to 5/95 and subsequently with a 1.5 min isocratic elution 5/95 A/B,
where A = H2O−0.1% TFA, B = CH3CN−0.1% TFA, at a flow rate of
0.5 mL/min, with UV detection at 220 nm, at column temperature of
25 °C, or performed on a Waters Quattro micro API (Agilent
HP1100, Gilson215) instrument, equipped with CAPCELL PAK C18
UG120 S-3 μm, 1.5 mm × 35 mm column, by gradient elution: 0.00
min (A/B = 100/0), 2.00 min (A/B = 0/100), 3.00 (A/B = 0/100),
1549
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
3.01 (A/B = 100/0), 3.30 (A/B = 100/0), where A = 2% CH3CN/
H2O with 5 mM NH4OAc, B = 95% CH3CN/H2O with 5 mM
NH4OAc, at a flow rate of 0.5 mL/min, with UV detection at 220 nm,
at column temperature of 40 °C. A part of compounds were assessed
by the following method. The analytical HPLC system consisted of
Prominence UFLC (Shimadzu Corporation, Japan), L-column2 ODS
column (30 mm × 2.1 mm I.D., 2 μm) (Chemicals Evaluation and
Research Institute, Japan) at column temperature of 50 °C and nano
quantity analyte detector, QT-500 (Quant technologies LLC, MN,
USA). The mobile phase A and B are a mixture of distillated water, 50
mmol/L ammonium acetate aqueous solution, and acetonitrile
(8:1:1,v/v/v) and a mixture of acetonitrile and 50 mmol/L ammonium
acetate aqueous solution (9:1,v/v), respectively. The flow rate
maintained 0.5 mL/min. The mixture ratio of mobile phase A and B
changed from 95/5 to 5/95 with a 2 min linear gradient and
subsequently with a 1 min isocratic elution 5/95. Elemental analyses
were carried out by Takeda Analytical Laboratories Limited and were
within 0.4% of the theoretical values unless otherwise noted. The
purity of compounds was assessed by elemental analysis or analytical
HPLC (>95%). Experimental procedures for compounds 14−19
including their intermediates were disclosed in the previous literature
(Supporting Information).21
4′-Hydroxy-2′,6′-dimethylbiphenyl-3-carbaldehyde (44). 4-
Bromo-3,5-dimethylphenol (43) 10.3 g, 51.0 mmol) and (3-
formylphenyl)boronic acid (7.67 g, 51.2 mmol) were dissolved in a
mixture of 1 M sodium carbonate aq (150 mL), EtOH (50 mL), and
toluene (150 mL). After argon substitution, tetrakis-
(triphenylphosphine)palladium(0) (2.95 g, 2.55 mmol) was added.
The reaction mixture was stirred at 80 °C under argon atmosphere for
24 h. The reaction mixture was cooled, and water was added. The
mixture was diluted with AcOEt, and the insoluble material was
filtered off through Celite. The organic layer of the filtrate was washed
with brine, dried over anhydrous magnesium sulfate, and concentrated.
The residue was purified by silica gel column chromatography
(AcOEt−hexane = 10:90−40:60) to give a solid, which was triturated
with hexane−AcOEt to give 44 (9.53 g, 83%) as pale-yellow crystals.
1H NMR (CDCl3) δ 1.97 (s, 6H), 4.69 (s, 1H), 6.62 (s, 2H), 7.42 (dt,
and tributylphosphine (0.448 mL, 1.80 mmol) in toluene (20 mL) was
added portionwise 1,1′-(azodicarbonyl)dipiperidine (0.454 g, 1.80
mmol) at 0 °C, and the mixture was stirred at room temperature under
nitrogen atmosphere for 15 h. Hexane (10 mL) was added, and the
insoluble material was removed by filtration. The filtrate was
concentrated, and the residue was purified by silica gel column
chromatography (AcOEt−hexane =5:95−30:70) to give methyl (6-
{[4′-(2-ethoxyethoxy)-2′,6′-dimethylbiphenyl-3-yl]methoxy}-2,3-dihy-
1
dro-1-benzofuran-3-yl)acetate (0.481 g, 89%) as a colorless oil. H
NMR (CDCl3) δ 1.25 (t, J = 7.1 Hz, 3H), 1.98 (s, 6H), 2.55 (dd, J =
16.5, 9.2 Hz, 1H), 2.75 (dd, J = 16.5, 5.4 Hz, 1H), 3.62 (q, J = 7.1 Hz,
2H), 3.71 (s, 3H), 3.75−3.85 (m, 3H), 4.14 (t, J = 5.1 Hz, 2H), 4.26
(dd, J = 9.2, 6.0 Hz, 1H), 4.75 (t, J = 9.2 Hz, 1H), 5.05 (s, 2H), 6.45−
6.50 (m, 2H), 6.68 (s, 2H), 7.01 (d, J = 8.1 Hz, 1H), 7.08 (dt, J = 7.0,
1.6 Hz, 1H), 7.16 (s, 1H), 7.35−7.44 (m, 2H). MS m/z 491 (M +
H)+. Step 2: To a solution of methyl (6-{[4′-(2-ethoxyethoxy)-2′,6′-
dimethylbiphenyl-3-yl]methoxy}-2,3-dihydro-1-benzofuran-3-yl)-
acetate (1.96 g, 4.00 mmol) in MeOH (20 mL) and THF (20 mL)
was added 2 M NaOH aqueous solution (6 mL), and the mixture was
stirred at room temperature for 3 days. The mixture was diluted with
water, acidified with 10% citric acid aqueous solution, and extracted
with AcOEt. The organic layer was washed with brine, dried over
MgSO4, and concentrated. The residue was purified by silica gel
column chromatography (AcOEt−hexane = 20:80−80:20) and
crystallized from heptane−AcOEt to give 53 (1.12 g, 59%) as
1
colorless prisms; mp 72 °C. H NMR (CDCl3) δ 1.25 (t, J = 7.1 Hz,
3H), 1.98 (s, 6H), 2.61 (dd, J = 16.8, 9.2 Hz, 1H), 2.80 (dd, J = 16.8,
5.4 Hz, 1H), 3.62 (q, J = 7.1 Hz, 2H), 3.75−3.85 (m, 3H), 4.14 (t, J =
5.0 Hz, 2H), 4.28 (dd, J = 9.2, 6.0 Hz, 1H), 4.75 (t, J = 9.2 Hz, 1H),
5.05 (s, 2H), 6.45−6.51 (m, 2H), 6.68 (s, 2H), 7.03−7.10 (m, 2H),
7.16 (s, 1H), 7.34−7.44 (m, 2H). MS m/z 477 (M + H)+. HPLC
purity (220 nm) 100%. Anal. Calcd for C29H32O6: C, 73.09; H, 6.77.
Found: C, 73.06; H, 6.73.
Ca Influx Activity of CHO Cells Expressing Human GPR40/
FFA1 (FLIPR Assay). CHO dhfr cells stably expressing human
GPR40/FFA1 (accession no. NM_005303) were plated and incubated
overnight in 5% CO2 at 37 °C. Then, cells were incubated in loading
buffer (recording medium containing 2.5 μg/mL fluorescent calcium
indicator Fluo 4-AM (Molecular Devices), 2.5 mmol/L probenecid
(Dojindo), and 0.1% fatty acid-free BSA (Sigma)) for 60 min at 37 °C.
Various concentrations of test compounds or γ-linolenic acid (Sigma)
were added into the cells and increase of the intracellular Ca2+
concentration after addition were monitored by FLIPR Tetra system
(Molecular Devices) for 90 s. The agonistic activities of test
compounds and γ-linolenic acid on human GPR40/FFA1 were
expressed as [(A − B)/(C − B)] × 100 (increase of the intracellular
Ca2+ concentration (A) in test compounds-treated cells, (B) in vehicle-
treated cells, and (C) in 10 μM γ-linolenic acid-treated cells). EC50
values and 95% confidence intervals of each compound were obtained
with Prism 5 software (GraphPad).
Preparation of CHO Membrane for GPR40/FFA1 Receptor
Binding Assay. Cell lines stably expressing human GPR40/FFA1 and
rat GPR40/FFA1 were used for the experiments. Each cell was
cultured in Minimum Essential Medium Alpha (MEM-Alpha,
Invitrogen) supplemented with 10% dialyzed Fetal-Bovine-Serum
(dialyzed FBS, Thermo Trace Ltd.), 100 unit/mL penicillin, and 100
unit/mL streptomycin in 5% CO2/95% air atmosphere at 37 °C. Cells
were harvested at confluence in Dulbecco’s Phosphate-Buffered-Saline
(D-PBS, Invitrogen) containing 1 mM EDTA and centrifuged. Cells
were homogenized in ice-cold membrane preparation buffer (50 mM
Tris-HCl (pH 7.5), 5 mM EDTA, 0.5 mM PMSF (Wako), 20 μg/mL
leupeptin, 0.1 μg/mL pepstatin A, 100 μg/mL Phosphoramidon,
Peptide Institute, Inc.) and centrifuged (700g, 10 min, 4 °C). The
supernatant was filtered through 40 μm Cell Strainer (BD Falcon) and
ultracentrifuged (100000g, 1 h, 4 °C) with Optima L-100 XP
Ultracentrifuge (Beckman Coulter). The precipitation was suspended
in the same buffer, and the protein concentration was determined with
the BCA Protein assay reagent (Pierce) following membrane
solubilization with 0.1% SDS and 0.1 M NaOH aqueous solution.
J = 7.7, 1.4 Hz, 1H), 7.59 (t, J = 7.6 Hz, 1H), 7.66 (t, J = 1.7 Hz, 1H),
7.86 (dt, J = 7.6, 1.5 Hz, 1H), 10.05 (s, 1H). MS m/z 227 (M + H)+.
[4′-(2-Ethoxyethoxy)-2′,6′-dimethylbiphenyl-3-yl]methanol
(42i). Step 1: A mixture of 44 (8.52 g, 37.7 mmol), 2-chloroethyl ethyl
ether (6.15 g, 56.6 mmol), K2CO3 (6.25 g, 45.2 mmol), and KI (1.25
g, 7.54 mmol) in DMF (40 mL) was stirred under nitrogen
atmosphere at 80 °C for 18 h. After evaporation of the solvent, the
residue was partitioned between water and AcOEt. The extract was
washed with brine, dried over anhydrous MgSO4, and concentrated.
The residue was purified by silica gel column chromatography
(AcOEt−hexane =5:95−25:75) to give 4′-(2-ethoxyethoxy)-2′,6′-
1
dimethylbiphenyl-3-carbaldehyde (10.0 g, 89%) as a colorless oil. H
NMR (CDCl3) δ 1.26 (t, J = 7.0 Hz, 3H), 1.99 (s, 6H), 3.62 (q, J = 7.0
Hz, 2H), 3.81 (t, J = 4.9 Hz, 2H), 4.15 (t, J = 4.9 Hz, 2H), 6.71 (s,
2H), 7.42 (dt, J = 7.5, 1.5 Hz, 1H), 7.58 (t, J = 7.5 Hz, 1H), 7.66 (t, J =
1.5 Hz, 1H), 7.86 (dt, J = 7.5, 1.5 Hz, 1H), 10.05 (s, 1H). MS m/z 299
(M + H)+. Step 2: To a solution of the obtained oil (4.21 g, 14.1
mmol) in MeOH (15 mL) and THF (30 mL) was added portionwise
NaBH4 (0.533 g, 14.1 mmol) at 0 °C, and the mixture was stirred
under nitrogen atmosphere at 0 °C for 2 h. The mixture was quenched
with water and 1 M HCl aqueous solution and extracted with AcOEt.
The extract was washed with brine, dried over anhydrous Na2SO4, and
concentrated. The residue was purified by silica gel column
chromatography (AcOEt−hexane = 20:80−60:40) to give a solid.
Recrystallization from hexane−AcOEt gave 42i (3.37 g, 79%) as
colorless crystals. 1H NMR (CDCl3) δ 1.25 (t, J = 7.1 Hz, 3H), 1.66 (t,
J = 5.9 Hz, 1H), 2.00 (s, 6H), 3.62 (q, J = 7.1 Hz, 2H), 3.80 (t, J = 5.1
Hz, 2H), 4.14 (t, J = 5.1 Hz, 2H), 4.73 (d, J = 5.9 Hz, 2H), 6.69 (s,
2H), 7.06 (d, J = 7.3 Hz, 1H), 7.12 (s, 1H), 7.33 (d, J = 7.3 Hz, 1H),
7.40 (t, J = 7.3 Hz, 1H). MS m/z 301 (M + H)+.
(6-{[4′-(2-Ethoxyethoxy)-2′,6′-dimethylbiphenyl-3-yl]-
methoxy}-2,3-dihydro-1-benzofuran-3-yl)acetic Acid (53). Step
1: To a mixture of 7e (0.250 g, 1.20 mmol), 42h (0.330 g, 1.10 mmol),
1550
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
The membrane suspension was stored at −80 °C until receptor
binding assay.
ASSOCIATED CONTENT
* Supporting Information
■
S
GPR40/FFA1 Receptor Binding Assay. The frozen cell
membranes were resuspended in ice-cold assay buffer (25 mmol/L
Tris-HCl (pH7.5), 5 mmol/L EDTA, 0.5 mmol/L PMSF, 20 μg/mL
leupeptin, 0.1 μg/mL pepstatin A, 0.05% CHAPS (Wako), and 0.2%
fatty-acid-free BSA (Sigma)) and used for receptor binding assay. To
determine the Kd values of 3-[4-({2′,6′-dimethyl-6-[(4-[3H])-
phenylmethoxy]biphenyl-3-yl}methoxy)phenyl] propanoic acid
(Amersham Biosciences) for human and rat GPR40/FFA1, binding
assays were performed in the presence of various concentrations of the
labeled ligand. After incubation at room temperature for 90 min, the
membranes were harvested GF/C filter plates (MILLIPORE) and
washed with ice-cold 50 mmol/L Tris-HCl (pH7.5) using FilterMate
Harvester (PerkinElmer). The membrane-associated radioactivities
were counted using TopCount liquid scintillation counter (Perki-
nElmer). Nonspecific binding was defined as binding in the presence
of 10 μmol/L of the unlabeled ligand. To determine the binding
affinities of test compounds to human and rat GPR40/FFA1, binding
assays were performed in the presence of both various concentrations
of test compounds and 2 nmol/L or 6 nmol/L of the labeled ligand.
The 50% inhibitory concentrations (IC50 values) of test compounds
for the labeled ligand were calculated using nonlinear regression
analysis in GraphPad Prism 3.0 (GraphPad Software). Ki values were
converted as Ki = IC50/{1 + (the concentration of the labeled ligand)/
Kd}.
Molecular Modeling. A homology model of GPR40/FFA1 was
constructed using the crystal structure of bovine rhodopsin (PDB code
1GZM),35 which was obtained from the RCSB Protein Data Bank, as a
structural template. An alignment of the amino acid sequences
between human/rat GPR40/FFA1 and rhodopsin was created using
ClustalX (version 2.0.11)36 and manually revised. Procedures of
homology modeling were performed in MOE (version 2008.10).37
The CL2 loop on the extra cellular domain was excluded except
Cys170 forming disulfide bond due to the difficulty of estimation. In
the previous step, compound 18 was docked into the obtained
receptor model using the program GOLD (version 4.1).38 Then, the
resultant docking modes with receptor models, replacing compound
18 with 15, 19, or 53, were subjected energy minimization with MOE
after connecting each residual substituent. In the energy minimization
process, the MMFF94s force field was used and the dielectric constant
was set to 2r, where r is the distance between two interacting atoms.
Pharmacokinetic Analysis in Rat Cassette Dosing. Test
compounds were administered as a cassette dosing to nonfasted rats.
After oral and intravenous administration, blood samples were
collected. The blood samples were centrifuged to obtain the plasma
fraction. The plasma samples were deproteinized with acetonitrile
containing an internal standard. After centrifugation, the supernatant
was diluted and centrifuged again. The compound concentrations in
the supernatant were measured by LC/MS/MS.
Synthetic procedures and spectroscopic data of final com-
pounds and intermediates not described in the manuscript.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Phone: +81-466-32-1279. Fax: +81-466-29-4536. E-mail:
Negoro_Nobuyuki@takeda.co.jp.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Dr. Nozomu Sakai, Satoshi Mikami, Nobuyuki
Amano, and Junichi Sakamoto for helpful discussions and Dr.
Tsuyoshi Maekawa for proofreading this manuscript.
ABBREVIATIONS USED
■
GPR40/FFA1, G protein-coupled receptor 40/free fatty acid
receptor 1; PK, pharmacokinetic; OGTT, oral glucose tolerance
test; GSIS, glucose-stimulated insulin secretion; FFA, free fatty
acid; DPP-4, dipeptidyl peptidase-4; GLP-1, glucagon-like
peptide-1; rt, room temperature; aq, aqueous; FLIPR,
fluorometric imaging plate reader; CHO, Chinese hamster
ovary; BSA, bovine serum albumin; ADME-Tox, absorption,
distribution, metabolism, excretion, and toxicology; CL, plasma
clearance; Cmax, maximum plasma concentration; AUC, area
under curve; Tmax, time of maximum plasma concentration;
AUC, area under curve; PG, plasma glucose
REFERENCES
■
(1) 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.; Shinohara, T.; Hinuma, S.; Fujisawa, Y.;
Fujino, M. Free fatty acids regulate insulin secretion from pancreatic β
cells through GPR40. Nature 2003, 422, 173−176.
(2) 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. The orphan G protein-coupled receptor
GPR40 is activated by medium and long chain fatty acids. J. Biol. Chem.
2003, 278, 11303−11311.
(3) Kotarsky, K.; Nilsson, N. E.; Flodgren, E.; Owman, C.; Olde, B. A
human cell surface receptor activated by free fatty acids and
thiazolidinedione drugs. Biochem. Biophys. Res. Commun. 2003, 301,
406−410.
(4) Shapiro, H.; Shachar, S.; Sekler, I.; Hershfinkel, M.; Walker, M.
D. Role of GPR40 in fatty acid action on the β cell line INS-1E.
Biochem. Biophys. Res. Commun. 2005, 335, 97−104.
Oral Glucose Tolerance Test (OGTT). The care and use of the
animals and the experimental protocols used in this research were
approved by the Experimental Animal Care and Use Committee of
Takeda Pharmaceutical Company Limited. Female Wistar fatty (WF)
rats were obtained from Takeda Rabics, Ltd. (Hikari, Japan). They
were fed a commercial diet CE-2 (Clea Japan Co.) and tap water ad
libitum. Female WF rats (17−19 weeks old) were fasted overnight and
orally given vehicle (0.5% methylcellulose) or compounds. All animals
were received an oral glucose load (1 g/kg) 1 or 4 h after drug
administration. Blood samples were collected from tail vein before
drug administration (pre), and just before glucose load (time 0), and
10, 30, 60, and 120 min after glucose load. Plasma glucose and plasma
insulin levels were measured by Autoanalyzer 7080 (Hitachi, Japan)
and radioimmunoassay (Millipore, USA), respectively. In the dose-
dependent study, statistical significances versus vehicle control were
assessed by one-tailed Williams test. Differences between two groups
were analyzed by Aspin−Welch test.
(5) Fujiwara, K.; Maekawa, F.; Yada, T. Oleic acid interacts with
GPR40 to induce Ca2+ signaling in rat islet β-cells: mediation by PLC
and L-type Ca2+ channel and link to insulin release. Am. J. Physiol.
Endocrinol. Metab. 2005, 289, E670−E677.
(6) Gravena, C.; Mathias, P. C.; Ashcroft, S. J. H. Acute effects of
fatty acids on insulin secretion from rat and human islets of
Langerhans. J. Endocrinol. 2002, 173, 73−80.
(7) Poitout, V.; Robertson, R. P. Minireview: Secondary β-cell failure
in type 2 diabetesa convergence of glucotoxicity and lipotoxicity.
Endocrinology 2002, 143, 339−342.
(8) Robertson, R. P.; Zhang, H. J.; Pyzdrowski, K. L.; Walseth, T. F.
Preservation of insulin mRNA levels and insulin secretion in HIT cells
1551
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552

Journal of Medicinal Chemistry
Article
by avoidance of chronic exposure to high glucose concentrations. J.
Clin. Invest. 1992, 90, 320−325.
(24) Node, M.; Kumar, K.; Nishide, K.; Ohsugi, S.; Miyamoto, T.
Odorless substitutes for foul-smelling thiols: syntheses and applica-
tions. Tetrahedron Lett. 2001, 42, 9207−9210.
(25) Conconi, M. T.; Guiotto, A.; Manzini, P.; Chilin, A.; Bassani, V.;
Parnigotto, P. P. 4-Hydroxymethyltetrahydro- and 4-hydroxymethyl-
benzopsoralen: synthesis and biological activity. Farmaco 1995, 50,
125−130.
(26) Fall, Y.; Santana, L.; Teijeira, M.; Uriarte, E. A convenient
synthesis of benzofuran-3-acetic acids. Heterocycles 1995, 41, 647−650.
(27) Kametani, T.; Nemoto, H.; Ishikawa, H.; Shiroyama, K.;
Fukumoto, K. A formal regio- and stereoselective total synthesis of
estrone. A convenient synthesis of D-homoestrone. J. Am. Chem. Soc.
1976, 98, 3378−890.
(28) Nishide, K.; Node, M. Development of odorless sulfur reagents
and their applications. J. Syn. Org. Chem. Jpn. 2004, 62, 895−910.
(29) Gao, Y.; Zhang, Q.; Xu, J. A convenient and effective method for
synthesizing β-amino-α,β-unsaturated esters and ketones. Synth.
Commun. 2004, 34, 909−916.
(30) Anghelide, N.; Draghici, C.; Raileanu, D. New dienamino esters
and their cyclization to α-pyridones of nicotinic acid. Tetrahedron
1974, 30, 623−632.
(31) Fukuyama, T.; Jow, C.-K.; Cheung, M. 2- and 4-Nitro-
benzenesulfonamides: exceptionally versatile means for preparation of
secondary amines and protection of amines. Tetrahedron Lett. 1995,
36, 6373−6374.
(32) Waring, M. J. Lipophilicity in drug discovery. Expert Opin. Drug
Discovery 2010, 5, 235−248.
(33) Yamagami, C.; Ogura, T.; Takao, N. Hydrophobicity parameters
determined by reverse-phase liquid chromatography I. Relationship
between capacity factors and octanol−water partition coefficients for
monosubstituted pyrazines and the related pyridines. J. Chromatogr.
1990, 514, 123−136.
(34) Ikeda, H.; Shino, A.; Matsuo, T.; Iwatsuka, H.; Suzuoki, Z. A
new genetically obese-hyperglycemic rat (Wistar fatty). Diabetes 1981,
30, 1045−1050.
(35) Li, J.; Edwards, P. C.; Burghammer, M.; Villa, C.; Schertler, G. F.
X. Structure of bovine rhodopsin in a trigonal crystal form. J. Mol. Biol.
2004, 343, 1409−1438.
(36) Larkin, M. A.; Blackshields, G.; Brown, N. P.; Chenna, R.;
McGettigan, P. A.; McWilliam, H.; Valentin, F.; Wallace, I. M.; Wilm,
A.; Lopez, R.; Thompson, J. D.; Gibson, T. J.; Higgins, D. G. Clustal
W and Clustal X version 2.0. Bioinformatics 2007, 23, 2947−2498.
(37) Chemical Computing Group Inc.; Chemical Computing Group
Inc., 1010 Sherbrooke Street W, Suite 910 Montreal, Quebec, Canada
H3A 2R7; www.chemcomp.com.
(38) CCDC, Business & Administration; CCDC, Business &
Administration, Cambridge Crystallographic Data Centre, 12 Union
Road, Cambridge CB2 1EZ, U.K.; http://www.ccdc.cam.ac.uk.
(9) Boden, G. Obesity, insulin resistance and free fatty acids. Curr.
Opin. Endocrinol. Diabetes Obes. 2011, 18, 139−143.
(10) Rendell, M. The role of sulphonylureas in the management of
type 2 diabetes mellitus. Drugs 2004, 64, 1339−1358.
(11) Doyle, M. E.; Egan, J. M. Pharmacological agents that directly
modulate insulin secretion. Pharmacol. Rev. 2003, 55, 105−131.
(12) Maedler, K.; Carr, R. D.; Bosco, D.; Zuellig, R. A.; Berney, T.;
Donath, M. Y. Sulfonylurea induced β-cell apoptosis in cultured
human islets. J. Clin. Endocrinol. Metab. 2005, 90, 501−506.
(13) Del Guerra, S.; Marselli, L.; Lupi, R.; Boggi, U.; Mosca, F.;
Benzi, L.; Del Prato, S.; Marchetti, P. Effects of prolonged in vitro
exposure to sulphonylureas on the function and survival of human
islets. J. Diabetes Complications 2005, 19, 60−64.
(14) For recent reviews, see: Bharate, S. B.; Nemmani, K. VS.;
Vishwakarma, R. A. Progress in the discovery and development of
small-molecule modulators of G-protein-coupled receptor 40
(GPR40/FFA1/FFAR1): an emerging target for type 2 diabetes.
Expert Opin. Ther. Pat. 2009, 19, 237−264.
(15) 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. Synthesis and activity of small molecule GPR40
agonists. Bioorg. Med. Chem. Lett. 2006, 16, 1840−1845.
(16) Zhou, C. Y.; Tang, C.; Chang, E.; Ge, M.; Lin, S. N.; Cline, E.;
Tan, C. P.; Feng, Y.; Zhou, Y. P.; Eiermann, G. J.; Petrov, A.; Salituro,
G.; Meinke, P.; Mosley, R.; Akiyama, T. E.; Einstein, M.; Kumar, S.;
Berger, J.; Howard, A. D.; Thornberry, N.; Mills, S. G.; Yang, L. H.
Discovery of 5-aryloxy-2,4-thiazolidinediones as potent GPR40
agonists. Bioorg. Med. Chem. Lett. 2010, 20, 1298−1301.
(17) 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. Synthesis and biological evaluation of 3-aryl-3-(4-
phenoxy)-propionic acid as a novel series of G protein-coupled
receptor 40 agonists. J. Med. Chem. 2007, 50, 2807−2817.
(18) Morrison, H.; Jona, J.; Walker, S. D.; Woo, J. C. S.; Li, L.; Fang,
J. Development of a suitable salt form for a GPR40 receptor agonist.
Org. Process Res. Dev. 2011, 15, 104−111.
(19) Christiansen, E.; Due-Hansen, M. E.; Urban, C.; Merten, N.;
Pfleiderer, M.; Karlsen, K. K.; Rasmussen, S. S.; Steensgaard, M.;
Hamacher, A.; Schmidt, J.; Drewke, C.; Petersen, R. K.; Kristiansen,
K.; Ullrich, S.; Kostenis, E.; Kassack, M. U.; Ulven, T. Structure−
activity study of dihydrocinnamic acids and discovery of the potent
FFA1 (GPR40) agonist TUG-469. ACS Med. Chem. Lett. 2010, 1,
345−349.
(20) 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. Discovery of potent and
selective agonists for the free fatty acid receptor 1 (FFA(1)/GPR40), a
potential target for the treatment of type II diabetes. J. Med. Chem.
2008, 51, 7061−7064.
(21) Christiansen, E.; Urban, C.; Grundmann, M.; Due-Hansen, M.
E.; Hagesaether, E.; Schmidt, J.; Pardo, L.; Ullrich, S.; Kostenis, E.;
Kassack, M.; Ulven, T. Identification of a Potent and Selective Free
Fatty Acid Receptor 1 (FFA1/GPR40) Agonist with Favorable
Physicochemical and in Vitro ADME Properties. J. Med. Chem.
2011, 54, 6691−6703.
(22) Negoro, N.; Sasaki, S.; Mikami, S.; Ito, M.; Suzuki, M.;
Tsujihata, Y.; Ito, R.; Harada, A.; Takeuchi, K.; Suzuki, N.; Miyazaki, J.;
Santou, T.; Odani, T.; Kanzaki, N.; Funami, M.; Tanaka, T.; Kogame,
A.; Matsunaga, S.; Yasuma, T.; Momose, Y. Discovery of TAK-875: a
potent, selective, and orally bioavailable GPR40 agonist. ACS Med.
Chem. Lett. 2010, 1, 290−294.
(23) Sasaki, S.; Kitamura, S.; Negoro, N.; Suzuki, M.; Tsujihata, Y.;
Suzuki, N.; Santou, T.; Kanzaki, N.; Harada, M.; Tanaka, Y.;
Kobayashi, M.; Tada, N.; Funami, M.; Tanaka, T.; Yamamoto, Y.;
Fukatsu, K.; Yasuma, T.; Momose, Y. Design, synthesis, and biological
activity of potent and orally available G protein-coupled receptor 40
agonists. J. Med. Chem. 2011, 54, 1365−1378.
1552
dx.doi.org/10.1021/jm2012968 | J. Med. Chem. 2012, 55, 1538−1552