Design, synthesis and Structure-activity relationship studies of new thiazole-based free fatty acid receptor 1 agonists for the treatment of type 2 diabetes

Article abstract of DOI:10.1016/j.ejmech.2016.02.040

The free fatty acid receptor 1 (FFA1/GPR40) has attracted interest as a novel target for the treatment of type 2 diabetes. Several series of FFA1 agonists including TAK-875, the most advanced compound terminated in phase III studies due to concerns about liver toxicity, have been hampered by relatively high molecular weight and lipophilicity. Aiming to develop potent FFA1 agonists with low risk of liver toxicity by decreasing the lipophilicity, the middle phenyl of TAK-875 was replaced by 11 polar five-membered heteroaromatics. Subsequently, systematic exploration of SAR and application of molecular modeling, leads to the identification of compound 44, which was an excellent FFA1 agonist with robustly hypoglycemic effect both in normal and type 2 diabetic mice, low risks of hypoglycemia and liver toxicity even at the twice molar dose of TAK-875. Meanwhile, two important findings were noted. First, the methyl group in our thiazole series occupied a small hydrophobic subpocket which had no interactions with TAK-875. Furthermore, the agonistic activity revealed a good correlation with the dihedral angle between thiazole core and the terminal benzene ring. These results promote the understanding of ligand-binding pocket and might help to design more promising FFA1 agonists.

Full text of DOI:10.1016/j.ejmech.2016.02.040

Accepted Manuscript
Design, Synthesis and Structure−–activity Relationship Studies of New Thiazole-
Based Free Fatty Acid Receptor 1 Agonists for the Treatment of Type 2 Diabetes
Zheng Li, Qianqian Qiu, Xue Xu, Xuekun Wang, Lei Jiao, Xin Su, Miaobo Pan,
Wenlong Huang, Hai Qian
PII:
S0223-5234(16)30120-9
10.1016/j.ejmech.2016.02.040
EJMECH 8391
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 23 August 2015
Revised Date: 14 February 2016
Accepted Date: 15 February 2016
Please cite this article as: Z. Li, Q. Qiu, X. Xu, X. Wang, L. Jiao, X. Su, M. Pan, W. Huang, H. Qian,
Design, Synthesis and Structure−–activity Relationship Studies of New Thiazole-Based Free Fatty Acid
Receptor 1 Agonists for the Treatment of Type 2 Diabetes, European Journal of Medicinal Chemistry
(2016), doi: 10.1016/j.ejmech.2016.02.040.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Design, Synthesis and Structure−Activity Relationship Studies of New Thiazole-Based Free
Fatty Acid Receptor 1 Agonists for the Treatment of Type 2 Diabetes
Zheng Lia, Qianqian Qiua, Xue Xub, Xuekun Wanga, Lei Jiaoa, Xin Sua, Miaobo Pana, Wenlong
Huanga, c*, Hai Qiana, c*
aCenter of Drug Discovery, State Key Laboratory of Natural Medicines, China Pharmaceutical
University, 24 Tongjiaxiang, Nanjing 210009, PR China.
bKey Laboratory of Drug Quality Control and Pharmacovigilance, China Pharmaceutical
University, 24 Tongjiaxiang, Nanjing 210009, PR China.
cJiangsu Key Laboratory of Drug Discovery for Metabolic Disease, China Pharmaceutical
University, 24 Tongjiaxiang, Nanjing 210009, PR China
Graphical Abstract
11 polar heteroaromatics were designed and synthesized to improve the lipophilicity and liver
toxicity of TAK-875.
Co-corresponding authors: Wenlong Huang and Hai Qian, Centre of Drug Discovery, State Key Laboratory of
Natural Medicines, China Pharmaceutical University, 24 Tongjiaxiang, Nanjing 210009, China.
Tel: +86-25-83271302; Fax: +86-25-83271302.
E-mail: ydhuangwenlong@126.com (W.l. Huang), qianhai24@163.com (H. Qian).

ACCEPTED MANUSCRIPT
Design, Synthesis and Structure−Activity Relationship Studies of New Thiazole-Based Free
Fatty Acid Receptor 1 Agonists for the Treatment of Type 2 Diabetes
Zheng Lia, Qianqian Qiua, Xue Xub, Xuekun Wanga, Lei Jiaoa, Xin Sua, Miaobo Pana, Wenlong
Huanga, c*, Hai Qiana, c*
aCenter of Drug Discovery, State Key Laboratory of Natural Medicines, China Pharmaceutical
University, 24 Tongjiaxiang, Nanjing 210009, PR China.
bKey Laboratory of Drug Quality Control and Pharmacovigilance, China Pharmaceutical
University, 24 Tongjiaxiang, Nanjing 210009, PR China.
cJiangsu Key Laboratory of Drug Discovery for Metabolic Disease, China Pharmaceutical
University, 24 Tongjiaxiang, Nanjing 210009, PR China
Abstract
The free fatty acid receptor 1 (FFA1/GPR40) has attracted interest as a novel target for the
treatment of type 2 diabetes. Several series of FFA1 agonists including TAK-875, the most
advanced compound terminated in phase III studies due to concerns about liver toxicity, have been
hampered by relatively high molecular weight and lipophilicity. Aiming to develop potent FFA1
agonists with low risk of liver toxicity by decreasing the lipophilicity, the middle phenyl of
TAK-875 was replaced by 11 polar five-membered heteroaromatics. Subsequently, systematic
exploration of SAR and application of molecular modeling, leads to the identification of
compound 44, which was an excellent FFA1 agonist with robustly hypoglycemic effect both in
Co-corresponding authors: Wenlong Huang and Hai Qian, Centre of Drug Discovery, State Key Laboratory of
Natural Medicines, China Pharmaceutical University, 24 Tongjiaxiang, Nanjing 210009, China.
Tel: +86-25-83271302; Fax: +86-25-83271302.
E-mail: ydhuangwenlong@126.com (W.l. Huang), qianhai24@163.com (H. Qian).
1

ACCEPTED MANUSCRIPT
normal and type 2 diabetic mice, low risks of hypoglycemia and liver toxicity even at the twice
molar dose of TAK-875. Meanwhile, two important findings were noted. First, the methyl group
in our thiazole series occupied a small hydrophobic subpocket which had no interactions with
TAK-875. Furthermore, the agonistic activity revealed a good correlation with the dihedral angle
between thiazole core and the terminal benzene ring. These results promote the understanding of
ligand-binding pocket and might help to design more promising FFA1 agonists.
Keywords: FFA1 agonist, GPR40, Liver toxicity, Type 2 diabetes, Heteroaromatics, Dihedral
angle.
1. Introduction
Type 2 diabetes mellitus (T2DM) is a global epidemic characterized by impaired glucose
homeostasis due to insulin deficiency and tissue resistance to insulin-stimulated glucose uptake
and utilization.[1, 2] Despite an inspiring array of pharmacotherapies for T2DM, most of current
hypoglycemic agents are associated with undesirable side effects such as body weight gain, risk of
hypoglycemia, and gastric symptoms.[3-6] Hence, a novel drug with increased safety and
durability in controlling blood glucose levels is still an urgent need for T2DM.[7, 8]
The G protein-coupled receptor 40 (GPR40)/free fatty acid receptor 1 (FFA1), the prominent
ones of novel antidiabetic targets in the last decade, play a key role in amplifying
glucose-stimulated insulin secretion (GSIS) on pancreatic β-cells without the associated risk of
hypoglycemia.[9-12] The mechanism of action of FFA1 is mainly couples with the G protein
α-subunit of the Gq family, which activates phospholipase C, leading to the production of inositol
triphosphate and release of intracellular Ca2+ from the endoplasmic reticulum, which are
associated with enhanced insulin secretion in a glucose concentration-dependent manner.[13, 14]
Recently, a variety of synthetic FFA1 agonists that contain acidic moieties have been reported
(Figure 1),[15-23] of which the compounds TAK-875, AMG-837 and LY2881835 have reached to
clinical trials. However, the known FFA1 ligands reported in the literature to date have been often
hampered by relatively high molecular weight and lipophilicity even beyond the scope of
“Rule-of-Five” (red mark in Figure 1), which is associated with poor water-solubility, metabolic
instability, toxicity, high promiscuity, and correlates with a higher risk of attrition in clinical
2

ACCEPTED MANUSCRIPT
trials.[24-28] Therefore, researches have suggested that clogP values should not exceed 4−5 in the
optimization process.[24, 29] At present, a lot of classic strategies have been extensively adopted
to decrease the lipophilicity such as introduction of hydrophilic group and scaffold hopping with
polar skeleton.[25, 30-32] With this perspective in mind, our research was focused on decreasing
the lipophilicity of the most advanced compound TAK-875, which was terminated in phase III
studies due to concerns about liver toxicity in December 2013.[33-35] Our previous study also
identified a series of phenoxyacetic acid derivatives as a more hydrophilic acidic moiety compared
with the dihydrobenzofuran of TAK-875.[36] In this study, we envisioned that replacement of the
middle benzene ring of TAK-875 with a polar five-membered heteroaromatics was a worthy
approach for the decrease of clogP value, to potentially reduce the risk of liver toxicity. Hence, 11
heteroaromatics with the combination of different heteroatom number and location were designed
and synthesized to explore the nature of ligand-binding pocket and select an optimal
heteroaromatics for further SAR studies (Figure 2). Among them, compound 26, with robust
FFA1 agonistic activity (77.3 nM), was selected to replace the terminal benzene ring with a series
of heterocycles in order to expand the SAR studies. Then, various substituents in the terminal
benzene ring of 26 were also introduced to optimize the FFA1 agonistic activity. Based on the
molecular modeling study and analysis of the SAR, we discovered that the methyl group of 26, the
most important fragment in our thiazole series, efficiently occupied a small hydrophobic
subpocket which had no interactions with TAK-875. Furthermore, the agonistic activity revealed a
good correlation with the dihedral angle between thiazole core and the terminal benzene ring. Of
all the synthesized compounds, compound 44 with excellent FFA1 agonistic activity (48.7 nM)
and antihyperglycemic effect, revealed a low risks of hypoglycemia and liver toxicity, was
considered to be a promising drug candidate worthy of further investigation.
Put Figure 1 and Figure 2 here
2. Results and Discussion
2.1. Chemistry
The synthetic routes of intermediates 2a-g are summarized in Scheme 1. The commercially
3

ACCEPTED MANUSCRIPT
starting material benzoylhydrazide 1 was treated with chloroacetyl chloride to afford the desired
product 2, which was further converted to intermediate 2a or 2b by dehydrating with POCl3 in
acetonitrile at reflux for 4 hrs or cyclizing with Lawesson’s reagent, respectively.[37, 38] Similar
conditions to 2a or 2b provided the intermediate 2c or 2d from the common intermediate 4,
formed from commercially available 2-Aminoacetophenone 3 treated with chloroacetyl chloride.
Cyclization of the starting material 2-Bromoacetophenone with ethyl thiooxamate generated the
desired ester 1a, which was followed by reduction using NaBH4 and methanol,[39] and further
chloridized to yield the desired intermediate 2e. The synthesis of the intermediate 1b was started
from the intermolecular cyclization between phenylhydrazine and ethyl 2, 4-dioxopentanoate.[40]
While the intermediate 1c was prepared by ring-closure reaction using hydroxylamine
hydrochloride and ethyl 2,4-dioxo-4-phenylbutanoate 10, which was easily obtained in
oxaloylation of acetophenone by Claisen condensation with diethyl oxalate in the presence of
NaOEt.[41, 42] The obtained ester 1b or 1c was subsequently converted to compounds 2f or 2g
according to the method for the synthesis of 2e.
The chlorinated intermediates 2h-i and 14a-u were synthesized starting from various
benzamide 11a-u as shown in Scheme 2. The ethyl 2-chloroacetoacetate was condensed with
benzamide 11a or various thiobenzamide 12a-u to form the oxazole 1d or substituted thiazole
13a-u.[43] The thiazole 1e was obtained from the intermolecular cyclization between ethyl
bromopyruvate and thiobenzamide, which was afforded from benzamide by treatment with
Lawesson’s reagent in high yield.[44] Next, the obtained ester 1d-e or 13a-u was reduced with
NaBH4 in the presence of methanol, followed by treating with thionyl chloride catalyzed by DMF,
generated the chlorinated intermediates 2h-i and 14a-u.
The synthesis of the target compounds 20 to 50 is depicted in Scheme 3. The
dihydrobenzofuran intermediate 15 was synthesized via published procedures.[15] Condensation
of the obtained chlorinated intermediates 2a-i or 14a-u with 15 by using Williamson ether
synthesis, followed by basic hydrolysis, afforded the desired carboxylic acids 20 to 27 and 29 to
50. The synthesis of triazole core in 28 was accomplished by classical click chemistry method.[45]
The starting material 15 was treated with 3-bromoprop-1-yne in acetone at the presence of
potassium carbonate to give 16 in high yield. Azidobenzene 18 was obtained by treating aniline
with NaNO2 followed by NaN3 in 50% acetic acid. The intermediates 16 reacted smoothly with
4

ACCEPTED MANUSCRIPT
azidobenzene 18 at ambient temperature in the presence of catalytic amount of copper sulfate and
sodium ascorbate in 80 % methanol, followed by basic hydrolysis, giving the target compound 28.
Put Scheme 1, 2 and 3 here
2.2. Biological activities and discussion
2.2.1. FFA1 agonistic activity and SAR study
In order to explore an optimal polar heteroaromatic to replace the middle phenyl of TAK-875, the
agonistic activities of the synthesized 11 heteroaromatic compounds with relative lower molecular
weight and clogP values were investigated by a fluorometric imaging plate reader (FLIPR) assay
in Chinese hamster ovary (CHO) cells expressing human FFA1 (Table 1). The oxdiazole core (20)
with lowest clogP value (clogP = 1.705) was initially designed to decrease the lipophilicity, but the
agonistic activities of 20 far more deviated the desired effect (6.31% agonistic activity at 300 nM).
An approximately 2-fold increase in potency over the compound 20 was obtained only by
replacing the oxygen atom of 20 with sulphur atom in compound 21. We speculated that the
improvement of the activity was associated with the larger hydrophobic effect and electronic
distribution of sulphur atom than oxygen atom.[30] A similar improvement in potency was also
observed by converting compound 22 to compound 23, which was intended to increase the
hydrophobic interaction with Gly139 and Val84 by removing the left-hand nitrogen of 21, and this
design strategy was directly inspired by the crystal structure of FFA1 bound to TAK-875.[46]
However, position exchange of sulphur and nitrogen atom in 23 to afford 24 resulted in a
significant drop of activity, attributed to the strict electrical interaction in the binding pocket of
FFA1. On the basis of the above analysis, we replaced the right-hand nitrogen with methyl in 20 to
further explore the electrical and hydrophobic effect with ligand-binding pocket, and the obtained
compound 25 showed a 3-fold increase in potency compared with the parent compound 20.
Consistent with before-mentioned findings, replacement of the oxygen atom in 25 with sulphur
atom gave compound 26, a superior agonist with a 67.41% agonistic activity at 300 nM, showed
the great potential for further optimization. Interestingly, switched the sites of sulphur and
nitrogen atom and removed the methyl in 26 to afford 27 appeared to diminish the in vitro
5

ACCEPTED MANUSCRIPT
agonistic activity. Meanwhile, other combinations of heteroatoms such as triazole (28), parazole
(29) and isoxazole (30) were designed to decrease the lipophilicity, but none of them showed the
desired potency, further confirming that the importance of electrical and hydrophobic effect in
ligand.
Put Table 1 here
To better understand the robustly agonistic activity of 26 and the SAR for FFA1 agonists, a
molecular modeling study based on the recently reported crystal structure of FFA1 (PDB
accession code: 4PHU) was performed by using the Molecular Operating Environment (MOE)
(Figure 3). As shown in Figure 3A, the interaction mode of compound 26 was nearly perfect
overlap with TAK-875, and the carboxylate moiety of them showed the same hydrogen bonding
interactions with Tyr91, Arg183 and Arg2258 in FFA1. Interestingly, the methyl group of 26 occupied
a small hydrophobic subpocket (Figure 3) which had no interactions with TAK-875. It was
suggested that this hydrophobic interaction was crucial for the robustly agonistic activity of 26
rather than other five-membered heteroaromatic compounds evaluated above.
Put Figure 3 here
Based on these results above, we therefore selected this scaffold of 26 as our starting point
for further modification. Our optimized efforts were directed to replace the terminal benzene ring
with a series of heterocycles in order to expand the SAR studies (Table 2). Gratifyingly, the
bioisosteres 3-thiophene derivative 32, with lower lipophilicity, exhibited a slight improvement on
potency in comparison with the parent 26. However, the agonistic activity of 31 was significantly
lower than 32 only by moving the sulphur atom from the 3-position to the 2-position of the
thiophene core. Concerns about potential metabolic issues associated with the thiophene ring led
us to introduce electron-withdrawing chlorine in thiophene ring to afford 33 while appeared to
diminish the agonistic activity.[47] The bulkier heterocycles such as benzofuran and naphthalene
have been reported to improve the agonistic activity of TAK-875 series,[17] however, a similar
6

ACCEPTED MANUSCRIPT
optimization in 34 and 35 exhibited a markedly reduced agonistic activity in our thiazole system.
One possible rational explanation is that the electron donors including oxygen and nitrogen atoms
are available for interaction with low-lying σ* orbitals of the C−S bond, and the noncovalent
sulfur interaction could restrict conformational flexibility of a molecule just as the intramolecular
hydrogen-bonding interaction.[30, 48, 49] As shown in Figure 4, the thiophene ring (31) and the
benzofuran (34) tend to be coplanar with the middle thiazole core, which might hindered the
hydrophobic interaction highlighted in Figure 3 owing to deviating the favorable conformation.
Put Table 2 and Figure 4 here
In order to select a suitable lead compound for further modification, the most potent
compounds 26 and 32 were used to evaluate the oral glucose tolerance test (OGTT) in normal ICR
mice (Figure 6A). Although the in vitro activity of 32 had more potential than 26, the
hypoglycemic effect of 32 was significantly lower than 26. We speculated that the weaker in vivo
effect of 32 may be attributable to a faster metabolism for thiophene ring of 32.[47, 50]
These results prompted a comprehensive evaluation on various substituents in the terminal
benzene ring of 26, an orally bioavailable FFA1 agonist. As shown in Table 3, for the
ortho-substituted compounds, the agonistic activity of 26 (2-H, 1.20 Å) ꢀ 42 (2-F, 1.47 Å) ꢀ
39 (2-Cl, 1.75 Å) ꢀ 36 (2-Me, 1.80 Å) ꢀ 45 (2-CF3, 2.20 Å) suggested that the steric effect
(Van der Waals radius) in the 2-position might influence agonistic activity of FFA1.[51] This
phenomenon also appeared in the naphthalene ring (35), seen as cyclic analogue of 2-methyl
benzene (36). These results further demonstrated the SAR of the present series was different from
the TAK-875 series which obtained the best activity by introducing substituents in the 2-position
of terminal benzene ring.[16-18, 52] Given the dihedral angle between thiazole core and the
terminal benzene ring with substituents in the 4-position ꢁ 3-position ꢁ 2-position (Table 3),
which revealed a good correlation with FFA1 agonistic activity: the agonistic activity of the
para-substituted compounds (38, 41, 44 and 47) ꢀ meta-substituted compounds (37, 40, 43 and
46) ꢀ‐ortho-substituted compounds (36, 39, 42 and 45), respectively. Therefore, the abnormal
SAR could be attributed to the larger space restriction induced by the larger dihedral angle, which
7

ACCEPTED MANUSCRIPT
changed the favorable conformation especially blocked the hydrophobic interaction of methyl
group on thiazole. With this beneficial experience for avoid substitution at ortho-position, the
3,5-disubstituted (49) and 3,4,5-trisubstituted methoxy group (50) were introduced to increase the
solvation effects with the water molecules which are exposed outside the receptor, but the
agonistic activity of them was inferior to 48 (4-methoxy group), implying that the polysubstituted
methoxy groups may introduce an unfavorable steric interaction with the surface of FFA1. After
evaluation of SAR in all the previously described areas, a clear SAR picture and the key
pharmacophore of this chemical scaffold were developed and summarized in Figure 5.
Put Table 3 and Figure 5 here
2.2.2. Oral Glucose Tolerance Test evaluated in the normal mice
Based on these results above, the optimized compounds 38, 41, 43 and 44 were selected to
evaluate the in vivo hypoglycemic effects in normal ICR mice by OGTT. The time-dependent
changes of blood glucose and the area under the curve (AUC0-2h) of the plasma glucose levels are
shown in Figure 6. All of the selected compounds showed good hypoglycemic activity in
accordance with in vitro agonistic activity. Among them, the hypoglycemic effect of compound 44
was in close proximity to the positive control TAK-875, the most advanced compound once in
phase III studies.
Put Figure 6 here
2.2.3. Dose-response relationship of 44 explored in normal ICR mice
On the basis of the in vivo hypoglycemic effects, the most potent compound 44 (10, 30 and 60
mg/kg) was further investigated for the dose-response relationship of lowering blood glucose level
in normal mice. As shown in Figure 7A, compound 44 demonstrated a dose-proportional decrease
in blood glucose levels with comparable efficacy at 30 mg/kg compared to TAK-875 at 20 mg/kg
during an OGTT. Furthermore, the plasma glucose curve of 44 tended to flat after 60 min at the
dose of 60 mg/kg. This result might be at least in part rationalized by the low risk of
8

ACCEPTED MANUSCRIPT
hypoglycemia showed in 44.
Put Figure 7 here
2.2.4. Effects of 44 on the risk of hypoglycemia
Obtaining a positive result in pharmacological study, the risk of hypoglycemia was subsequently
assessed in fasting normal ICR mice by oral administrating a high dose of 44 in comparison with
glibenclamide to further confirm the above speculation. As shown in Figure 7B, glibenclamide
(15 mg/kg) lowered plasma sugar levels far below normal fasting levels. In contrast, compound 44,
even at the high dose of 60 mg/kg, only slightly reduced fasting glucose levels in ICR mice, and
the change of plasma glucose levels was much smaller than that caused by administration of
glibenclamide. Thus, our results indicated that compound 44 revealed a low risk of hypoglycemia,
a serious adverse effect to sulfonylureas, which are widely used as one of the first-line drugs for
the treatment of T2DM.
2.2.5. Hypoglycemic effects of 44 explored in type 2 diabetic C57BL/6 mice
To further assess antihyperglycemic effects of 44 in the diabetic state, STZ-induced type 2 diabetic
C57BL/6 mice were used to evaluate the OGTT of 44, the most potent agonist among our
synthetic compounds. As shown in Figure 8, the compound 44 was significantly improved the
hyperglycemia state of diabetic mice as TAK-875. The results demonstrated that compound 44,
with a low risk of hypoglycemia, has a great potential for improving the diabetic state.
Put Figure 8 here
2.2.6. Effects of 44 on the risk of liver toxicity
For evaluating the risk of liver toxicity, a 30 days chronic study was subsequently carried out in
normal ICR mice. In this study, vehicle, TAK-875 (20 mg/kg, 38 mmol/kg) and 44 (30 mg/kg, 75
mmol/kg) was orally administered to normal mice once daily for 30 days. At the end of study,
several parameters indicative of liver function including alanine aminotransferase (ALT), aspartate
9

ACCEPTED MANUSCRIPT
transaminase (AST) and total bilirubin (TBIL) in serum were measured. As shown in Table 4, the
ALT and AST, the main biomarkers of liver injury, was significantly increased in TAK-875 group
compared with vehicle group, the result demonstrated a relatively high risk of liver toxicity
observed in TAK-875. Gratifyingly, the compound 44 group, even at the twice molar dose of
TAK-875, showed comparable values of ALT and AST similar to that of vehicle group. These
results suggested that compound 44, with lower molecular weight and lipophilicity, successfully
reduced the risk of liver toxicity compared to TAK-875.
Furthermore, in this chronic 44 treatment study, no particular side effects were observed at
the compound 44 group, indicating the possibility that minimized off-target pharmacology and
metabolic toxicity.
Put Table 4 here
3. Conclusion
With the aim of developing potent FFA1 agonists with reduced lipophilicity to decrease the risk of
liver toxicity, we have identified a new series of thiazole FFA1 agonists by comprehensive
evaluating 11 heteroaromatics with the combination of different heteroatom number and location.
Subsequently, the optimal lead compound 26 was selected to replace the terminal benzene ring
with a series of heterocycles, and various substituents in the terminal phenyl of 26 were also
introduced to optimize the agonistic activity. Among them, the preferred compound 44 exhibits
potent agonistic activity on FFA1 and produces a robustly hypoglycemic effect both in normal and
type 2 diabetic mice. Besides, compound 44, even at the high dose of 60 mg/kg, revealed a low
risk of hypoglycemia. In further studies, a 30 days chronic 44 treatment study was carried out in
normal mice, even at the twice molar dose of TAK-875, revealed a lower risk of liver toxicity
compared with TAK-875, and no particular side effects were observed at compound 44 treated
group. All of these results demonstrated that compound 44 was meaningful for further
investigation, and the information obtained from the SAR studies in our thiazole series might help
to design more active FFA1 agonists that are structurally related.
4. Experimental section
10

ACCEPTED MANUSCRIPT
4.1. General chemistry
All starting materials, reagents and solvents were obtained from commercial sources and used
without further purification unless otherwise indicated. Purifications by column chromatography
were carried out over silica gel (200-300 mesh) and monitored by thin layer chromatography
performed on GF/UV 254 plates and were visualized using UV light at 254 and 365 nm. Melting
points were taken on a RY-1 melting-point apparatus and were uncorrected. NMR spectra were
recorded on a Bruker ACF-300Q instrument (300 MHz for 1H NMR and 75 MHz for 13C NMR
spectra), chemical shifts are expressed as values relative to tetramethylsilane as internal standard,
and coupling constants (J values) were given in hertz (Hz). LC/MS spectra were recorded on a
Waters liquid chromatography-mass spectrometer system (ESI). Elemental analyses were
performed by the Heraeus CHN-O-Rapid analyzer. TAK-875 was synthesized as previously
reported.[15]
1
The physical characteristics, H NMR, 13C NMR, MS and elemental analysis data for all
intermediates and target molecules, were reported in the supporting information.
4.2. Molecular modeling
Docking simulations were performed using MOE (version 2008.10, The Chemical Computing
Group, Montreal, Canada). The crystal structure of FFA1 (PDB ID: 4PHU) was downloaded from
the Protein Data Bank (PDB). Prior to ligand docking, the structure was prepared with Protonate
3D and a Gaussian Contact surface was draw around the binding site. Subsequently, the active site
was isolated and the backbone was removed. The ligand poses was filtered using Pharmacophore
Query Editor. The compound structures were placed in the site with Pharmacophore method and
then ranked with the London dG scoring function. For the energy minimization in the pocket,
MOE Forcefield Refinement was used and ranked with the London dG scoring function.
4.3. Biological methods
4.3.1. Ca2+ influx activity of CHO cells expressing human FFA1 (FLIPR Assay)
CHO cells stably expressing human FFA1 (accession no. NM_005303) were seeded into 96-well
plates at a density of 15K cells/well and incubated 16 h in 5% CO2 at 37 °C. After, the culture
medium in the wells was removed and washed with 100 ꢀL of Hank’s Balanced Salt Solution.
11

ACCEPTED MANUSCRIPT
Then, cells were incubated in loading buffer (containing 2.5 ꢀg/mL fluorescent calcium indicator
Fluo 4-AM, 2.5 mmol/L probenecid and 0.1% fatty acid-free BSA) for 1 h at 37 °C. Various
concentrations of test compounds or γ-linolenic acid (Sigma) were added into the cells and the
intracellular calcium flux signals were monitored by FLIPR Tetra system (Molecular Devices).
The agonistic activities of test compounds on human FFA1 were expressed as [(A−B)/(C−B)]×100
(increase of the intracellular Ca2+ concentration (A) in the test compounds-treated cells and (B) in
vehicle-treated cells, and (C) in 10 ꢀM γ-linolenic acid-treated cells). EC50 value of selected
compound was obtained with Prism 5 software (GraphPad).
4.3.2. Animals and Statistical Analysis of the Data
Male ICR mice (18-22 g) and male C57BL/6 mice (18-22 g) were purchased from Comparative
Medicine Centre of Yangzhou University (Jiangsu, China), acclimatized for 1 week before the
experiments. The animal room was maintained under a constant 12 h light/black cycle with
temperature at 23 ± 2 °C and relative humidity 50 ± 10% throughout the experimental period.
Mice were allowed ad libitum access to standard pellets and water unless otherwise stated, and the
vehicle used for drug administration was 0.5% sodium salt of Caboxy Methyl Cellulose aqueous
solution for all animal studies. All animal experiments were performed in compliance with the
relevant laws and institutional guidelines, and our experiments have been approved by the
institutional committee of China Pharmaceutical University.
Statistical analyses were performed using specific software (GraphPad InStat version 5.00,
GraphPad software, San Diego, CA, USA). Unpaired comparisons were analyzed using the
two-tailed Student’s t-test, unless otherwise stated.
4.3.3. Oral Glucose Tolerance Test in male ICR mice
Normal ICR mice 10 weeks old were fasted overnight (12 h), weighted, bled via the tail tip, and
randomized into 8 groups (n = 6). Mice were administrated orally with a single doses of vehicle,
TAK-875 (10 mL·kg-1; 20 mg·kg-1; 38 mmol·kg-1), or selected compounds (10 mL·kg-1; 38
mmol·kg-1) and subsequently dosed orally with 30% glucose aqueous solution (3 g·kg-1) after half
an hour. Blood samples were collected immediately before drug administration (-30 min), before
glucose challenge (0 min), and at 15, 30, 45, 60 and 120 min post-dose. The blood glucose was
12

ACCEPTED MANUSCRIPT
measured by blood glucose test strips (SanNuo ChangSha, China).
4.3.4. Dose-response relationship of 44 explored in male ICR mice
To investigate dose-response relationship of 44, normal ICR mice 10 weeks old were fasted
overnight (12 h), weighted, bled via the tail tip, and randomized into 5 groups (n = 6). Mice were
administrated orally with a single doses of vehicle, TAK-875 (10 mL·kg-1; 20 mg·kg-1), or
compound 44 (10 mg·kg-1, 30 mg·kg-1, 60 mg·kg-1) and subsequently dosed orally with 30%
glucose aqueous solution (3 g·kg-1) after half an hour. Blood samples were collected immediately
before drug administration (-30 min), before glucose challenge (0 min), and at 15, 30, 45, 60 and
120 min post-dose. The blood glucose was measured by blood glucose test strips (SanNuo
ChangSha, China).
4.3.5. Effects of 44 on the risk of hypoglycemia
10 weeks old male normal ICR mice were fasted overnight and randomized into 3 groups (n = 6).
Compound 44 (60 mg·kg-1), glibenclamide (15 mg·kg-1), or vehicle was orally administered, and
blood was collected from tail vein immediately before administration (0 h) and at 30, 60, 90, 120
and 180 min after administration and measure blood glucose as described above.
4.3.6. Hypoglycemic effects of 44 explored in type 2 diabetic mice
Male C57BL/6 mice after 1 week adaptation were fed with high-fat diet (45% calories from fat,
from Mediscience Ltd., Yangzhou, China) ad libitum for 4 weeks to induce insulin resistance and
then injected intraperitoneally (i.p.) with low dose of STZ (10 mL·kg-1; 80 mg·kg-1). The mice
were fed with high-fat-diet for another 4 weeks, and the development of diabetes was confirmed
by measuring blood glucose levels. The mice with fasting blood glucose level 11.1 mmol/L or
higher were considered to be diabetic and were used in the experiment as type 2 diabetic mice
model.[53, 54]
Type 2 diabetic C57BL/6 mice were fasted overnight (12 h), weighted, bled via the tail tip,
and randomized into 3 groups (n = 6), another group of normal fasting C57BL/6 mice was added
as negative control. Mice were administrated orally with a single doses of vehicle, TAK-875 (10
mL·kg-1; 20 mg·kg-1), or compound 44 (10 mL·kg-1; 30 mg·kg-1) and subsequently dosed orally
13

ACCEPTED MANUSCRIPT
with 20% glucose aqueous solution (2 g·kg-1) after half an hour. Blood samples were collected
immediately before drug administration (-30 min), before glucose challenge (0 min), and at 15, 30,
45, 60 and 120 min post-dose. The blood glucose was measured by blood glucose test strips
(SanNuo ChangSha, China).
4.3.7. Risk of liver toxicity of 44 evaluated in normal ICR mice
Normal ICR mice were dosed daily with the vehicle, TAK-875 (10 mL·kg-1; 20 mg·kg-1) or
compound 44 (10 mL·kg-1; 30 mg·kg-1) by gavage administration for 30 days. The body weights
were measured every 5 days and the dosage was adjusted according to the most recent body
weight. All animals were observed daily and any abnormal state were recorded. At the end of
treatment, mice were fasted overnight (12 h), blood samples were drawn from orbit and
centrifuged at 3500 rpm/min for 10 min to separate serum. ALT, AST and TBIL in serum were
measured using automatic biochemical analyzer (Beckman Coulter, AU5811, Tokyo, Japan).
Acknowledgements
This study was supported by grants from the National Natural Science Foundation of China
(Grants 81172932 and 81273376), the Natural Science Foundation of Jiangsu Province (Grant
BK2012356), and the Project Program of State Key Laboratory of Natural Medicines, China
Pharmaceutical University (Grant JKGZ201103).
14

ACCEPTED MANUSCRIPT
References
[1] R.A. DeFronzo, From the triumvirate to the ominous octet: a new paradigm for the treatment of
type 2 diabetes mellitus, Diabetes, 58 (2009) 773-795.
[2] G. Danaei, M.M. Finucane, Y. Lu, G.M. Singh, M.J. Cowan, C.J. Paciorek, J.K. Lin, F. Farzadfar,
Y.-H. Khang, G.A. Stevens, M. Rao, M.K. Ali, L.M. Riley, C.A. Robinson, M. Ezzati, C. Global
Burden Metab Risk Factors, National, regional, and global trends in fasting plasma glucose and
diabetes prevalence since 1980: systematic analysis of health examination surveys and epidemiological
studies with 370 country-years and 2.7 million participants, Lancet, 378 (2011) 31-40.
[3] O.J. Phung, J.M. Scholle, M. Talwar, C.I. Coleman, Effect of noninsulin antidiabetic drugs added to
metformin therapy on glycemic control, weight gain, and hypoglycemia in type 2 diabetes, JAMA, 303
(2010) 1410-1418.
[4] M.A. Avery, C.S. Mizuno, A.G. Chittiboyina, T.W. Kurtz, H.A. Pershadsingh, Type 2 diabetes and
oral antihyperglycemic drugs, Curr Med Chem, 15 (2008) 61-74.
[5] S.A. Stein, E.M. Lamos, S.N. Davis, A review of the efficacy and safety of oral antidiabetic drugs,
Expert Opin Drug Saf, 12 (2013) 153-175.
[6] H.E. Lebovitz, Type 2 diabetes mellitus-current therapies and the emergence of surgical options,
Nat. Rev. Endocrinol., 7 (2011) 408-419.
[7] A.A. Tahrani, C.J. Bailey, S. Del Prato, A.H. Barnett, Management of type 2 diabetes: new and
future developments in treatment, Lancet, 378 (2011) 182-197.
[8] X. Xu, G. Wang, T. Zhou, L. Chen, J. Chen, X. Shen, Novel approaches to drug discovery for the
treatment of type 2 diabetes, Expert Opin Drug Dis, 9 (2014) 1047-1058.
[9] Y. Itoh, Y. Kawamata, M. Harada, M. Kobayashi, R. Fujii, S. Fukusumi, K. Ogi, M. Hosoya, Y.
Tanaka, H. Uejima, H. Tanaka, M. Maruyama, R. Satoh, S. Okubo, H. Kizawa, H. Komatsu, F.
Matsumura, Y. Noguchi, T. Shinobara, S. Hinuma, Y. Fujisawa, M. Fujino, Free fatty acids regulate
insulin secretion from pancreatic β cells through GPR40, Nature, 422 (2003) 173-176.
[10] C.P. Briscoe, M. Tadayyon, J.L. Andrews, W.G. Benson, J.K. Chambers, M.M. Eilert, C. Ellis,
N.A. Elshourbagy, A.S. Goetz, D.T. Minnick, P.R. Murdock, H.R. Sauls, U. Shabon, L.D. Spinage, J.C.
Strum, P.G. Szekeres, K.B. Tan, J.M. Way, D.M. Ignar, S. Wilson, A.I. Muir, The orphan G
protein-coupled receptor GPR40 is activated by medium and long chain fatty acids, J Biol Chem, 278
(2003) 11303-11311.
[11] L.A. Stoddart, N.J. Smith, G. Milligan, International Union of Pharmacology. LXXI. Free fatty
acid receptors FFA1,-2, and-3: pharmacology and pathophysiological functions, Pharmacol. Rev., 60
(2008) 405-417.
[12] P. Wellendorph, L.D. Johansen, H. Bräuner-Osborne, Molecular pharmacology of promiscuous
seven transmembrane receptors sensing organic nutrients, Mol. Pharmacol., 76 (2009) 453-465.
[13] H. Shapiro, S. Shachar, I. Sekler, M. Hershfinkel, M.D. Walker, Role of GPR40 in fatty acid action
on the β cell line INS-1E, Biochem. Biophys. Res. Commun., 335 (2005) 97-104.
[14] K. Fujiwara, F. Maekawa, T. Yada, 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., 289 (2005) 670-677.
[15] N. Negoro, S. Sasaki, S. Mikami, M. Ito, M. Suzuki, Y. Tsujihata, R. Ito, A. Harada, K. Takeuchi,
N. Suzuki, J. Miyazaki, T. Santou, T. Odani, N. Kanzaki, M. Funami, T. Tanaka, A. Kogame, S.
Matsunaga, T. Yasuma, Y. Momose, Discovery of TAK-875: a potent, selective, and orally bioavailable
GPR40 agonist, ACS Med. Chem. Lett., 1 (2010) 290-294.
15

ACCEPTED MANUSCRIPT
[16] S. Mikami, S. Kitamura, N. Negoro, S. Sasaki, M. Suzuki, Y. Tsujihata, T. Miyazaki, R. Ito, N.
Suzuki, J. Miyazaki, T. Santou, N. Kanzaki, M. Funami, T. Tanaka, T. Yasuma, Y. Momose, Discovery
of phenylpropanoic acid derivatives containing polar functionalities as potent and orally bioavailable G
protein-coupled receptor 40 agonists for the treatment of type 2 diabetes, J. Med. Chem., 55 (2012)
3756-3776.
[17] N. Negoro, S. Sasaki, M. Ito, S. Kitamura, Y. Tsujihata, R. Ito, M. Suzuki, K. Takeuchi, N. Suzuki,
J. Miyazaki, T. Santou, T. Odani, N. Kanzaki, M. Funami, T. Tanaka, T. Yasuma, Y. Momose,
Identification of fused-ring alkanoic acids with improved pharmacokinetic profiles that act as G
protein-coupled receptor 40/free fatty acid receptor 1 agonists, J. Med. Chem., 55 (2012) 1538-1552.
[18] N. Negoro, S. Sasaki, S. Mikami, M. Ito, Y. Tsujihata, R. Ito, M. Suzuki, K. Takeuchi, N. Suzuki, J.
Miyazaki, T. Santou, T. Odani, N. Kanzaki, M. Funami, A. Morohashi, M. Nonaka, S. Matsunaga, T.
Yasuma, Y. Momose, Optimization of (2, 3-dihydro-1-benzofuran-3-yl) acetic acids: discovery of a
non-free fatty acid-like, highly bioavailable G protein-coupled receptor 40/free fatty acid receptor 1
agonist as a glucose-dependent insulinotropic agent, J. Med. Chem., 55 (2012) 3960-3974.
[19] S.C. McKeown, D.F. Corbett, A.S. Goetz, T.R. Littleton, E. Bigham, C.P. Briscoe, A.J. Peat, S.P.
Watson, D.M. Hickey, Solid phase synthesis and SAR of small molecule agonists for the GPR40
receptor, Bioorg. Med. Chem. Lett., 17 (2007) 1584-1589.
[20] J.B. Houze, L. Zhu, Y. Sun, M. Akerman, W. Qiu, A.J. Zhang, R. Sharma, M. Schmitt, Y. Wang, J.
Liu, J. Liu, J.C. Medina, J.D. Reagan, J. Luo, G. Tonn, J. Zhang, J.Y.-L. Lu, M. Chen, E. Lopez, K.
Nguyen, L. Yang, L. Tang, H. Tian, S.J. Shuttleworth, D.C.H. Lin, AMG 837: a potent, orally
bioavailable GPR40 agonist, Bioorg. Med. Chem. Lett., 22 (2012) 1267-1270.
[21] S.P. Brown, P.J. Dransfield, M. Vimolratana, X. Jiao, L. Zhu, V. Pattaropong, Y. Sun, J. Liu, J. Luo,
J. Zhang, S. Wong, R. Zhuang, Q. Guo, F. Li, J.C. Medina, G. Swaminath, D.C.H. Lin, J.B. Houze,
Discovery of AM-1638: a potent and orally bioavailable GPR40/FFA1 full agonist, ACS Med. Chem.
Lett., 3 (2012) 726-730.
[22] Y. Wang, J. Liu, P.J. Dransfield, L. Zhu, Z. Wang, X. Du, X. Jiao, Y. Su, A.-r. Li, S.P. Brown, A.
Kasparian, M. Vimolratana, M. Yu, V. Pattaropong, J.B. Houze, G. Swaminath, T. Thanhvien, N.
Khanh, Q. Guo, J. Zhang, R. Zhuang, F. Li, L. Miao, M.D. Bartberger, T.L. Correll, D. Chow, S. Wong,
J. Luo, D.C.H. Lin, J.C. Medina, Discovery and optimization of potent GPR40 full agonists containing
tricyclic spirocycles, ACS Med. Chem. Lett., 4 (2013) 551-555.
[23] E. Defossa, M. Wagner, Recent developments in the discovery of FFA1 receptor agonists as novel
oral treatment for type 2 diabetes mellitus, Bioorg. Med. Chem. Lett., 24 (2014) 2991-3000.
[24] C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J. Feeney, Experimental and computational
approaches to estimate solubility and permeability in drug discovery and development settings, Adv.
Drug Deliv. Rev., 64 (2012) 4-17.
[25] E. Christiansen, C. Urban, M. Grundmann, M.E. Due-Hansen, E. Hagesaether, J. Schmidt, L.
Pardo, S. Ullrich, E. Kostenis, M. Kassack, T. Ulven, 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., 54 (2011) 6691-6703.
[26] M.J. Waring, Lipophilicity in drug discovery, Expert Opin Drug Dis, 5 (2010) 235-248.
[27] G.M. Keserü, G.M. Makara, The influence of lead discovery strategies on the properties of drug
candidates, Nat Rev. Drug Discov, 8 (2009) 203-212.
[28] P.D. Leeson, R.J. Young, Molecular Property Design: Does Everyone Get It?, ACS Med. Chem.
Lett., 6 (2015) 722-725.
16

ACCEPTED MANUSCRIPT
[29] M.P. Gleeson, Generation of a set of simple, interpretable ADMET rules of thumb, J. Med. Chem.,
51 (2008) 817-834.
[30] N.A. Meanwell, Synopsis of Some Recent Tactical Application of Bioisosteres in Drug Design, J
Med Chem 54 (2011) 2529-2591.
[31] N.A. Meanwell, Improving drug candidates by design: a focus on physicochemical properties as a
means of improving compound disposition and safety, Chem. Res. Toxicol., 24 (2011) 1420-1456.
[32] E. Christiansen, M.E. Due-Hansen, C. Urban, M. Grundmann, R. Schröder, B.D. Hudson, G.
Milligan, M.A. Cawthorne, E. Kostenis, M.U. Kassack, T. Ulven, Free fatty acid receptor 1
(FFA1/GPR40) agonists: mesylpropoxy appendage lowers lipophilicity and improves ADME
properties, J. Med. Chem., 55 (2012) 6624-6628.
[33] C.F. Burant, P. Viswanathan, J. Marcinak, C. Cao, M. Vakilynejad, B. Xie, E. Leifke, TAK-875
versus placebo or glimepiride in type 2 diabetes mellitus: a phase 2, randomised, double-blind,
placebo-controlled trial, Lancet, 379 (2012) 1403-1411.
[34] K. Kaku, T. Araki, R. Yoshinaka, Randomized, double-blind, dose-ranging study of TAK-875, a
novel GPR40 agonist, in Japanese patients with inadequately controlled type 2 diabetes, Diabetes.
Care., 36 (2013) 245-250.
[35] M.S. Hedrington, S.N. Davis, Discontinued in 2013: diabetic drugs, Expert. Opin. Inv. Drug., 23
(2014) 1703-1711.
[36] X. Wang, T. Zhao, B. Yang, Z. Li, J. Cui, Y. Dai, Q. Qiu, H. Qiang, W. Huang, H. Qian, Synthesis
and biological evaluation of phenoxyacetic acid derivatives as novel free fatty acid receptor 1 agonists,
Bioorgan Med Chem, 23 (2015) 132-140.
[37] J. Balsells, L. DiMichele, J. Liu, M. Kubryk, K. Hansen, J.D. Armstrong, Synthesis of [1, 2, 4]
triazolo [4, 3-α] piperazines via highly reactive chloromethyloxadiazoles, Org. Lett., 7 (2005)
1039-1042.
[38] P. Kotian, S.W. Mascarella, P. Abraham, A.H. Lewin, J.W. Boja, M.J. Kuhar, F.I. Carroll, Synthesis,
Ligand Binding, and Quantitative Structure-Activity Relationship Study of 3β-(4'-Substituted
phenyl)-2β-heterocyclic Tropanes: Evidence for an Electrostatic Interaction at the 2β-Position, J. Med.
Chem., 39 (1996) 2753-2763.
[39] N. Boechat, J.C.S. da Costa, J. de Souza Mendonça, P.S.M. de Oliveira, M. Vinıcius Nora De
́
Souza, A simple reduction of methyl aromatic esters to alcohols using sodium borohydride–methanol
system, Tetrahedron lett., 45 (2004) 6021-6022.
[40] A. Schmidt, T. Habeck, M.K. Kindermann, M. Nieger, New pyrazolium-carboxylates as structural
analogues of the pseudo-cross-conjugated betainic alkaloid Nigellicine, J. Org. Chem., 68 (2003)
5977-5982.
[41] C.R. Hauser, F.W. Swamer, J.T. Adams, The Acylation of Ketones to Form βꢂDiketones or βꢂ
Keto Aldehydes, Org. React., 8 (1954) 59-196.
[42] E.E. Royals, The use of sodium methoxide in the Claisen reaction, J. Am. Chem. Soc., 67 (1945)
1508-1509.
[43] K. Nicolaou, M. Zak, B.S. Safina, A.A. Estrada, S.H. Lee, M. Nevalainen, Total synthesis of
thiostrepton. Assembly of key building blocks and completion of the synthesis, J. Am. Chem. Soc., 127
(2005) 11176-11183.
[44] J.Y. Hwang, R.R. Attia, F. Zhu, L. Yang, A. Lemoff, C. Jeffries, M.C. Connelly, R.K. Guy,
Synthesis and evaluation of sulfonylnitrophenylthiazoles (SNPTs) as thyroid hormone receptor–
coactivator interaction inhibitors, J. Med. Chem., 55 (2012) 2301-2310.
17

ACCEPTED MANUSCRIPT
[45] H.C. Kolb, M. Finn, K.B. Sharpless, Click chemistry: diverse chemical function from a few good
reactions, Angew. Chem. Int. Edit., 40 (2001) 2004-2021.
[46] A. Srivastava, J. Yano, Y. Hirozane, G. Kefala, F. Gruswitz, G. Snell, W. Lane, A. Ivetac, K.
Aertgeerts, J. Nguyen, A. Jennings, K. Okada, High-resolution structure of the human GPR40 receptor
bound to allosteric agonist TAK-875, Nature, 513 (2014) 124–127.
[47] D.J. St. Jean Jr, C. Fotsch, Mitigating heterocycle metabolism in drug discovery, J. Med. Chem.,
55 (2012) 6002-6020.
[48] M. Iwaoka, S. Takemoto, S. Tomoda, Statistical and Theoretical Investigations on the
Directionality of Nonbonded S···O Interactions. Implications for Molecular Design and Protein
Engineering, J. Am. Chem. Soc., 124 (2002) 10613-10620.
[49] B.R. Beno, K.-S. Yeung, M.D. Bartberger, L.D. Pennington, N.A. Meanwell, A survey of the role
of noncovalent sulfur interactions in drug design, J. Med. Chem., 58 (2015) 4383–4438.
[50] D.K. Dalvie, A.S. Kalgutkar, S.C. Khojasteh-Bakht, R.S. Obach, J.P. O'Donnell,
Biotransformation reactions of five-membered aromatic heterocyclic rings, Chem Res Toxicol, 15
(2002) 269-299.
[51] A. Bondi, van der Waals volumes and radii, J. Phys. Chem., 68 (1964) 441-451.
[52] S. Sasaki, S. Kitamura, N. Negoro, M. Suzuki, Y. Tsujihata, N. Suzuki, T. Santou, N. Kanzaki, M.
Harada, Y. Tanaka, M. Kobayashi, N. Tada, M. Funami, T. Tanaka, Y. Yamamoto, K. Fukatsu, T.
Yasuma, Y. Momose, Design, synthesis, and biological activity of potent and orally available G
protein-coupled receptor 40 agonists, J. Med. Chem., 54 (2011) 1365-1378.
[53] M.S. Winzell, B. Ahrén, The high-fat diet-fed mouse a model for studying mechanisms and
treatment of impaired glucose tolerance and type 2 diabetes, Diabetes, 53 (2004) 215-219.
[54] J.A. Koehler, L.L. Baggio, B.J. Lamont, S. Ali, D.J. Drucker, Glucagon-like peptide-1 receptor
activation modulates pancreatitis-associated gene expression but does not modify the susceptibility to
experimental pancreatitis in mice, Diabetes, 58 (2009) 2148-2161.
18

ACCEPTED MANUSCRIPT
Table 1: In vitro activities and select physicochemical properties of 11 heteroaromatics 20 to 30
compd
TAK-875
20
Het
Act%(300nM)a
76.01
Act%(100nM)b
65.32
Mw
clogPc
4.697
1.705
524.63
352.35
6.31
1.54
13.65
10.15
19.76
0.23
3.36
2.34
5.26
0.16
7.15
54.81
0.15
0.76
0.98
0.84
368.41
351.36
367.42
367.42
365.39
381.45
367.42
351.36
364.40
351.36
2.507
2.932
3.597
3.597
3.403
4.298
3.597
2.702
3.73
21
22
23
24
25
26
27
28
29
30
21.17
67.41
5.67
0.89
1.23
1.67
3.232
a Agonist activities mean values at a screening concentration of 300 nM were obtained from three
independent experiments.
b Agonist activities mean values at a screening concentration of 100 nM were obtained from three
independent experiments.
c clogP values were estimated with ChemDraw Ultra, version 12.0.
19

ACCEPTED MANUSCRIPT
Table 2: In vitro activities and physicochemical properties of designed compounds 31 to 35
compd
TAK-875
26
Het
Act%(100nM)a
65.32
hFFA1 EC50 (nM)b
Mw
clogPc
4.697
4.298
29.6
77.3
524.63
381.45
54.81
35.15
56.47
ND
387.47
387.47
4.189
3.979
31
32
68.5
8.34
14.78
8.53
ND
ND
ND
421.91
421.47
431.51
4.923
5.068
5.472
33
34
35
ND = Not determined.
a Agonist activities mean values at a screening concentration of 100 nM were obtained from three
independent experiments.
b EC50 values for FFA1 activities represent the mean of three determinations.
c clogP values were estimated with ChemDraw Ultra, version 12.0.
20

ACCEPTED MANUSCRIPT
Table 3: In vitro activities and dihedral angle of designed compounds 36 to 50
Act% (100nM)a
65.32
54.81
10.68
47.16
54.85
12.36
42.97
53.35
25.76
56.90
61.63
7.76
hFFA1 EC50 (nM)b
29.6
compd
TAK-875
26
R
Calcd dihedral angle θ (deg)c
89.6
4.9
H
2-Me
77.3
ND
50.4
12.6
3.2
36
3-Me
118.5
73.5
37
4-Me
38
2-Cl
ND
46.7
13.3
7.8
39
3-Cl
137.5
85.6
40
4-Cl
41
2-F
ND
36.5
10.6
7.4
42
3-F
64.8
43
4-F
48.7
44
2-CF3
3-CF3
4-CF3
4-OMe
3,5-diOMe
3,4,5-triOMe
ND
58.6
15.3
4.8
45
35.68
50.35
46.78
17.53
7.47
ND
46
96.7
47
123.4
ND
4.3
48
6.3
49
ND
16.6
50
ND = Not determined.
a Agonist activities mean values at a screening concentration of 100 nM were obtained from three
independent experiments.
b EC50 values for FFA1 activities represent the mean of three determinations.
c Dihedral angle between thiazole core and the terminal benzene ring was calculated by Dihedral
Plot in MOE, version 2008.10.
21

ACCEPTED MANUSCRIPT
Table 4: Effects of 44 on the values of AST, ALT and TBIL
Groups
Vehicle
ALT (IU/L)
35.0±4.3
AST (IU/L)
133.1±13.1
181.1±15.3#
141.8±14.8*
TBIL (IU/L)
4.1±0.4
TAK-875 (20 mg/kg)
44 (30 mg/kg)
48.3±4.7#
37.5±4.1*
4.0±0.8
4.0±0.5
#
Results are expressed as mean ± SD for six mice in each group. pꢃ0.05 compared to vehicle
group by Student’s t-test. *pꢃ0.05 compared to TAK-875 group by Student’s t-test.
22

ACCEPTED MANUSCRIPT
Figure 1: Selected examples of synthetic GPR40 agonists. clogP values are calculated with
ChemDraw Ultra 12.0 using the “clogP” option.
23

ACCEPTED MANUSCRIPT
Figure 2: Our strategy to decrease the lipophilicity of TAK-875 by replacing the middle benzene
ring with 11 heteroaromatics.
24

ACCEPTED MANUSCRIPT
Figure 3: The interaction mode of compound 26 bound to FFA1. Key residues are labeled in red,
and hydrogen bonding interactions are represented by yellow dashed lines. (A) Overlay of
TAK-875 (green) and 26 (blue) bound to FFA1. (B) Compound 26 bound to FFA1.
25

ACCEPTED MANUSCRIPT
Figure 4: Presumed conformational preference based on the interaction between the lone pair (O
and N) and C−S σ* orbitals (noncovalent N···S interaction and O···S interaction).
26

ACCEPTED MANUSCRIPT
Figure 5: SARs and the key pharmacophore summary.
27

ACCEPTED MANUSCRIPT
Figure 6: Effects of compounds on blood glucose levels during an OGTT in normal ICR mice. A,
B and C represent time dependent changes of plasma glucose levels. D shows the AUC0-2h of
blood glucose levels. Values are mean ± SEM (n = 6). *pꢃ0.05, **pꢃ0.01 compared to vehicle
normal mice by Student’s t-test.
28