Discovery of N-arylpyrroles as agonists of GPR120 for the treatment of type II diabetes

Article abstract of DOI:10.1016/j.bmcl.2018.02.013

The discovery of a novel series of N-arylpyrroles as agonists of GPR120 (FFAR4) is discussed. One lead compound is a potent GPR120 agonist, has good selectivity for related receptor GPR40 (FFAR1), has acceptable PK properties, and is active in 2 models of Type 2 Diabetes in mice.

Full text of DOI:10.1016/j.bmcl.2018.02.013

Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of N-arylpyrroles as agonists of GPR120 for the treatment of
type II diabetes
⇑
Michael P. Winters , Zhihua Sui, Mark Wall, Yuanping Wang, Joseph Gunnet, James Leonard, Hong Hua,
Wen Yan, Arthur Suckow, Austin Bell, Wilmelenne Clapper, Celia Jenkinson, Peter Haug,
Tatiana Koudriakova, Norman Huebert, William V. Murray
Cardiovascular and Metabolism Research, Janssen Research and Development LLC, 1400 McKean Rd., Spring House, PA 19477, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The discovery of a novel series of N-arylpyrroles as agonists of GPR120 (FFAR4) is discussed. One lead
compound is a potent GPR120 agonist, has good selectivity for related receptor GPR40 (FFAR1), has
acceptable PK properties, and is active in 2 models of Type 2 Diabetes in mice.
Ó 2018 Published by Elsevier Ltd.
Received 11 January 2018
Revised 5 February 2018
Accepted 7 February 2018
Available online xxxx
Keywords:
GPR120
FFAR4
N-arylpyrrole
Type 2 diabetes
GPR40
oGTT
DIO mouse
The prevalence of type 2 diabetes is increasing and has become
a public health crisis in much of the world.1 The disorder is charac-
terized by hyperglycemia, insulin-secreting b-cell dysfunction, and
insulin resistance.1 Furthermore, many of the co-morbidities asso-
ciated with diabetes, such as cardiovascular disease, kidney dis-
ease, neuropathy, and retinopathy, are mediated by inflammatory
processes that are common in diabetics.2
GPR120 (Free Fatty Acid Receptor 4) is a 7-transmembrane
GPCR that is activated by omega-3 fatty acids and is mainly
expressed in the intestine, adipose, and macrophages.3 The activa-
tion of GPR120 has been implicated in a number of processes such
as release of gastrointestinal peptides including glucagon-like pep-
tide-1 (GLP-1), adipogenesis, lipogenesis, and glucose tolerance.3
Additionally, GPR120 is postulated to mediate anti-inflammatory
and insulin-sensitizing effects in the adipose and macrophages.3
These benefits suggest that GPR120 agonists have the potential
to be an effective treatment for type 2 diabetes.
cle would give an increase in potency over 1 and 2, or improved
ADME or pharmacokinetic properties compared to the existing
leads. In this vein, we decided to explore N-arylpyrroles 3 as
GPR120 agonists.5
The initial SAR from the isothiazole series indicated that elec-
tron-withdrawing substituents on the pyrrole ring at the 3-posi-
tion, in particular CF3, were favored for GPR120 agonist activity.4
In addition, concerns about the instability of the electron-rich pyr-
role ring compel the incorporation of electron-withdrawing groups
as well.6 Synthesis of the 3-trifluoromethylpyrrole analogs starts
with the bromination of N-phenylsulfonylpyrrole 4 at the 3-posi-
tion followed by installation of the 2-carbomethoxy to give 5
(Scheme 1).5,7 Treatment with methyl 2,2-difluoro-2-(fluorosul-
fonyl)acetate in the presence of copper iodide gives an inseparable
mixture of 5 and 6.8 Reduction of the bromine of 5 with hydrogen
and Pd/C enables the isolation of 6 from the undesired reduced
product. Finally, deprotection of the phenylsulfonyl with TBAF
gives the desired methyl 3-(trifluoromethyl)-1H-pyrrole-2-car-
boxylate 7. Treatment with an arylboronic acid in the presence of
copper acetate and pyridine gives the N-arylpyrrole 8.9 Reduction
of the ester with LAH followed by Mitsunobu reaction gives 10.
Hydrolysis of the ester gives the desired acid 11, while reduction
of the ester with LAH gives the alcohol 12.
Thiophene 1 and isothiazole 2 were existing leads from our
GPR120 agonist program (Fig. 1).4 While continuing investigations
of those series, we were probing various substitutions of the cen-
tral 5-membered heterocyclic ring to see if changing the heterocy-
⇑
Corresponding author.
E-mail address: mwinter4@its.jnj.com (M.P. Winters).
https://doi.org/10.1016/j.bmcl.2018.02.013
0960-894X/Ó 2018 Published by Elsevier Ltd.
Please cite this article in press as: Winters M.P., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.02.013

2
M.P. Winters et al. / Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
Fig. 1. Heterocyclic GPR120 Agonists.
Scheme 2. Reagents and Conditions: (a) ICH2CO2Et, NaOAc; (b) EtOCHCHCOCF3,
DCM, 40 °C; (c) DBU, EtOAc, 120 °C, sealed tube.
Because the trifluoromethylation step from the synthetic
hGPR120 potency and was air stable. This established the CF3 as
an optimal functional group at the 3-position of the pyrrole ring
and was used in further analogs.
scheme above was not amenable to scale-up, an improved synthe-
sis of intermediate 16 was developed (Scheme 2). Condensation of
4-chloroaniline 13 with ethyl 2-iodoacetate gives 14. Addition to
4-ethoxy-1,1,1-trifluorobut-3-en-2-one gives 15, which is cyclized
with DBU in moderate yield to give the desired pyrrole 16.10 This
new route enabled the multi-gram synthesis of lead compounds
for in vivo studies.
We first made a selection of compounds with substitution at the
3-position of the pyrrole to explore the SAR of this position
(Table 1). At this early stage, we only made compounds with the
2,3-dimethyl- or 2,3-difluoro- substitution on the phenylpropionic
acid moiety because these groups were known to afford good
potency in other series. The unsubstituted pyrrole 17 was active
in the calcium mobilization assay, but discolored over time indicat-
ing air oxidation of this compound. Installing bromo 18 and chloro
19 at this position gave a similar level of potency, with the added
benefit of air stability. The cyano group 20 gave a large loss in
potency while the CF3 group 21 and 22 gave a 2-fold boost in
Next, we explored the SAR around the aryl ring attached to the
pyrrole nitrogen in the presence of the 3-CF3 and the 2,6-difluo-
rophenylpropionic acid sidechain (Table 2). The EC50 value of the
unsubstituted phenyl 23 was about 100 nM. The 4-methyl 24
was 2-fold more potent, however ethyl 25 and bromo 26 did not
show improvement over the phenyl. 4-chloro 27 was a slight
improvement, but most other substitutions on the ring resulted
in a loss of potency in the calcium assay. Addition of a pyridyl
nitrogen at the 3-position also gave a loss in activity compared
to the phenyl.
Scheme 1. Reagents and Conditions: (a) (i) Br2, HOAc; (ii) LDA, MeOCOCl; (b) (i) FSO2CF2CO2CH3, CuI, NMP, 80 °C; (ii) H2, 10% Pd/C, EtOH; (c) TBAF, THF; (d) R1B(OH)2, Cu
(OAc)2, pyridine, DCM; (e) LAH, THF, 0 °C; (f) R2-phenol, ADDP, Bu3P, toluene, 60 °C; (g) LiOH, THF, MeOH, water; (h) LAH, THF, 0 °C.
Please cite this article in press as: Winters M.P., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.02.013

M.P. Winters et al. / Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
3
Table 1
SAR of the 3-position of the pyrrole.
Cpd. No.
R1
R2
R3
R4
hGPR120 Ca2+ EC50 (nM)a
17
18
19
20
21
22
H
Br
Cl
CN
CF3
CF3
4-Br
4-Cl
4-Cl
4-Cl
4-Cl
4-Cl
Me
F
Me
F
Me
F
Me
F
Me
F
Me
F
164
161
173
1370
88
91
a
hGPR120 Ca2+ assay in HEK-293 cells (n > 2, average values, SEM < 25%).
Table 2
SAR of the N-aryl/heteroaryl group.
Cpd. No.
R1
X
hGPR120 Ca2+ EC50 (nM)a
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
H
4-Me
4-Et
4-Br
4-Cl
4-F
4-OMe
3-F
3-F-4-Cl
3-F-4-Me
2-F-4-Cl
2-F-4-Me
2,4-di-F
3,4-di-F
4-Me
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
N
101
50
117
112
80
134
184
448
368
262
179
118
156
175
289
877
197
Lastly, we interrogated the SAR of the phenylpropionic acid
sidechain (Table 3). The EC50 value of the 2,6-difluoropropionic
acid 27 was 80 nM, while reduction to the propanol sidechain 40
gave a boost in activity to 43 nM. Conversely, when the substitu-
tion on the phenyl was modified to 2,3-dimethyl 21 or 2,3-difluro
22, reduction to the propanol 41 and 42 led to a decline in potency
over the parent acid. The combination of 2,3-difluoro substitution
with the introduction of an ethyl group at R6 43 gave an EC50 of
40 nM, the most potent compound in the series. Methyl substitu-
tion was introduced on the carbon
a- to the acid 44 to try to
4-Et
4-MeO
N
N
improve metabolic stability, but disappointingly gave a 3-fold loss
in potency. When R6 is F 28, the 2,6-diflurophenylpropionic acid
derivative had an EC50 of 134 nM. Changing the 2,6-di-F to 2-
methyl 46 or 2,3-di-methyl 47 gave a 2- to 3-fold loss in activity.
Substitution at the 3-position of the central aryl ring with CF3 48
or Cl 49 afforded analogs with poor potency, while the methyl ester
50 resulted in a complete loss of hGPR120 activity.
hGPR120 Ca2+ assay in HEK-293 cells (n > 2, average values, SEM < 25%).
a
Table 3
SAR of the phenylpropionic acid sidechain.
Cpd. No.
R1
R2
R3
R4
R5
R6
hGPR120 Ca2+
EC50 (nM)a
27
40
21
41
22
42
43
44
45
28
46
47
48
49
50
F
F
F
F
Me
Me
F
F
F
F
F
H
H
Me
Me
F
F
F
H
H
H
H
Me
CF3
Cl
H
H
H
H
H
H
H
H
Me
Me
H
H
H
H
H
H
COOH
CH2OH
COOH
CH2OH
COOH
CH2OH
COOH
COOH
COOH
COOH
COOH
COOH
COOH
COOH
COOMe
Cl
Cl
Cl
Cl
Cl
Cl
Et
Cl
F
F
F
F
F
80
43
88
533
91
H
H
H
H
H
F
162
40
Early ADME data was used to try to prioritize the more potent
compounds for pharmacokinetic studies and ultimately in vivo glu-
cose-lowering experiments. Three of the more potent compounds
(24, 27 and 43) from the primary screen were profiled for their
cytochrome P450 inhibition (Cyp), their microsomal stability (Intr.
Cl t1/2) and their plasma protein binding (PPB) (Table 4). All 3 com-
pounds had low cytochrome P450 inhibition with slight activity
against 2C9 and 2C19, and all were highly plasma protein bound
in rat, mouse, and human. The one distinguishing factor for
290
547
134
280
346
834
923
>5000
F
F
F
Me
H
H
H
F
H
Me
H
H
F
F
Cl
hGPR120 Ca2+ assay in HEK-293 cells (n > 2, average values, SEM < 25%).
a
Table 4
Cyp inhibition, microsomal stability and PPB for selected compounds.
Cpd.
No.
hGPR120 Ca2+
EC50 (nM)
Cyps 3A4, 2D6,
Cyp 2C9
M)
Cyp 2C19 Intr. CL human
Intr. CL mouse
t1/2 (min)
Intr. CL rat
t1/2 (min)
PPB human
(% bound)
PPB mouse
(% bound)
PPB rat
(% bound)
1A2 (
lM)
(
l
(
lM)
t1/2 (min)
24
27
43
50
80
40
>10
>10
>10
4.2
7
7
>10
10
9
7
51
15
85
155
7
<4
132
<4
99.92
99.92
99.98
99.59
99.63
99.86
99.85
99.83
99.92
Table 5
PK of 27 in mouse, rat and dog.
Species
IV dose (mg/kg)
IV t1/2 (h)
IV CL (mL/min/kg)
IV Vss (L/kg)
PO dose (mg/kg)
PO t1/2 (h)
PO Cmax (ng/mL)
PO AUClast (h*ng/mL)
F (%)
Mouse
Rat
Dog
1.15
2
1
4.14
1.64
4.72
4.9
3.2
2.45
1.14
0.42
0.52
14.6
10
5
3.45
2.87
4.37
18,667
11,632
8923
81,714
60,877
47,897
166
116
145
Please cite this article in press as: Winters M.P., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.02.013

4
M.P. Winters et al. / Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
compound 27 was its much lower intrinsic clearance in human,
mouse and rat liver microsomes versus compounds 24 and 43,
most likely due to the oxidation of the 4-alkyl substituents.
Metabolite ID studies of 27 showed that the major metabolites
were cleavage of the central ether linkage and b -oxidation of the
parent propionic acid sidechain, while the acyl glucuronide of the
parent was detected in minor amounts. Due to its suitable meta-
bolic profile, compound 27 was chosen for further studies.
The in vitro activity of 27 was assessed in multiple species and
cell lines to determine if it was suitable for in vivo experiments. The
human GPR120 Ca2+ EC50 in the primary high-expressing HEK-293
transfected cell line was 80 nM, while the potencies for rat
and mouse were 713 nM and 193 nM respectively. The human
b -arrestin EC50 in a CHO-K1 cell line was 69 nM, while the human
Ca2+ EC50 in HT-29 cells, a low-expressing endogenous cell line,
was 137 nM. Together, this data indicated that 27 should be suffi-
ciently potent to see in vivo effects in mice, but perhaps not in rats
due to its high rEC50
.
We were also interested in the selectivity of 27 for GPR120 over
the related GPR40 (FFAR1) receptor.11 In humans, the selectivity
was found to be about 42-fold, with an hGPR40 EC50 of 3340 nM,
while the mouse GPR40 EC50 was found to be 3520 nM, 18-fold
selectivity. Although 27 maintains some potency for GPR40, we
believed that this low level of activity would not compromise
our efforts to observe clear effects of GPR120 agonism in vivo. To
evaluate selectivity in a more general sense, 27 was submitted
for testing against 50 GPCR’s, ion channels, and transporters. All
were below 50% inhibition except for 5HT2b (62% inhibition),
5HT5a (58%), norepinephrine transporter (51%) and Cl- channel
(66%) at 10 lM concentration.
(a)
A U C -3 0 to 9 0 D e lta
A U C -3 0 to 9 0 D e lta
(a)
2 0 0 0 0
+ 8 %
1 5 0 0 0
1 0 1 1 3
4 9 2 0 *
7 3 1 9
7 4 2 9
-9 %
1 5 0 0 0
1 0 0 0 0
5 0 0 0
0
-8 3 %
-5 4 %
-8 8 %
* *
1 0 0 0 0
5 0 0 0
0
* * *
-5 0 0 0
-1 0 0 0 0
)
k
k
k
p
C
p
p k
m
p k
m
e
l
e
l
7
2
7
2
M
m
m p
c
i
c
i
3
0
.
m
0
1
%
.1
d
p
d
p
h
5
0
0 .
1
3 .
.
e
a x
0
7
7
7
2
v
v e h
(
m
S
2
2
2 7
C m
T
C
e
l
d
d
d
p
d
c
K O
p
p
p
W
i
h
K O
m
W T
C m
C m
C m
C
V e
T re a tm e n t G ro u p
(b)
T re a tm e n t G ro u p
(b)
G lu co se -3 0 to 9 0 D e lta G lu c o se C u rv e
G lu co se -3 0 to 9 0 D e lta G lu c o se C u rv e
5 0 0
4 0 0
3 0 0
2 0 0
1 0 0
0
3 0 0
2 5 0
2 0 0
1 5 0
1 0 0
5 0
0
-5 0
-2 0
0
2 0
4 0
6 0
8 0
-3 0
0
3 0
6 0
9 0
T im e (m in )
T im e (m in s )
Fig. 2. Acute oGTT of 27 in DIO mice (*** = P < 0.001, ** = P < 0.01 1 Way ANOVA
Fig. 3. Acute ipGTT in wild-type and GPR120 knockout mice (* = P < 0.05 1 Way
w/Tukey post test).
ANOVA w/Tukey post test).
Please cite this article in press as: Winters M.P., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.02.013

M.P. Winters et al. / Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
5
Table 6
27 in a 2-week DIO mouse model of type 2 diabetes.
Group
Number Glucose baseline: Glucose final: Glucose vehicle-subtracted Insulin baseline:
Insulin final:
Insulin vehicle-subtracted
of mice
mean mg/dL (SD) mean mg/dL
(SD)
LS mean Change (95% CI)
Mean pg/mL (SD) mean pg/mL (SD) LS mean change
(95% CI)
Vehicle
27 10 mg/kg QD
27 10 mg/kg BID 10
27 30 mg/kg QD
11
11
187 (24)
202 (35)
203 (35)
201 (36)
184 (30)
163 (13)
170 (27)
151 (16)
–
16,959 (4943)
17,395 (5930)
16,966 (7312)
17,067 (6144)
7304 (2967)
7756 (3357)
6225 (2744)
3856 (1133)
–
–23.7 (À48.5; 1.1)
+457 (À2498; 3412)
À16.4 (À41.9; 9.1)
À1079 (À4106; 1948)
À35.6a
À3447a
10
(À61.0; À10.2)
(À6474; À420)
À36.6a
À5969a
Rosi 3 mg/kg
11
188 (29)
148 (12)
15,956 (6091)
1348 (25 0)
(À61.1; À12.0)
(À8928; À3011)
a
Significantly different from vehicle.
The pharmacokinetics of 27 was obtained in mouse, rat and dog
Acknowledgements
to determine if it had a suitable profile to investigate the in vivo
effects of the GPR120 mechanism (Table 5). 27 had an adequate
half-life, low clearance in all 3 species, and high blood levels and
bioavailability. The low volume of distribution (Vss) indicates that
27 mainly resides in the plasma compartment, most likely due to
its high plasma protein binding. Although not ideal, this profile
was thought to be sufficient for in vivo studies in rodents.
An anesthetized guinea pig study was performed to gauge the
cardiovascular safety of 27. There was no effect on the PQ, QRS,
QT or QTcB intervals, heart rate, body temperature or ECG
morphology, however 27 did significantly decrease mean
arterial blood pressure by 10% relative to vehicle at 22.4 mg/kg
(Cmax 105,000 ng/mL), the highest concentration tested.12
Next, compound 27 was evaluated in an oral glucose tolerance
test (oGTT) in diet-induced obese (DIO) mice to assess the conse-
quences of GPR120 agonism in an acute model of type 2 diabetes
(Fig. 2). 27 was dosed orally at doses from 0.1 to 3.0 mg/kg 30
min (time-30) prior to the oral glucose administration (time 0),
and glucose levels were measured from 30 min pre-glucose to
90 min post-glucose. 27 lowered glucose in a dose-dependent
manner, significantly lowering glucose at 1 and 3 mg/kg. The 3
mg/kg dose lowered glucose to an extent similar to saxagliptin, a
DPP-4 inhibitor positive control, at 1 mg/kg.13
Thanks to the Discovery Sciences ADME and Analytical teams,
and the Cardiovascular Center of Excellence in Beerse for their con-
tributions to this work.
A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at https://doi.org/10.1016/j.bmcl.2018.02.013.
References
1. (a) Hu F. Diabetes Care. 2011;34:1249;
(b) Beltran-Sanchez H, Harhay MM, McElligott S. J Am Coll Cardiol. 2013;62:697.
2. (a) Wellen KE, Hotamisligil GS. J Clin Invest. 2005;115:1111;
(b) Navarro JF, Mora C. Nephrol Dial Transplant. 2005;20:2601;
(c) Williams MD, Nadler JL. Curr Diabetes Rep. 2007;7:242.
3. (a) Zhang D, Leung PS. Drug Des Dev Ther. 2014;1013:8;
(b) Cornall LM, Mathai ML, Hryciw DH, McAinch AJ. Drug Disc Today.
2014;19:670;
(c) Oh DY, Walenta E, Akiyama TE, et al. Nat Med. 2014;20:942;
(d) Li A, Li Y, Du L. Fut Med Chem. 2015;7:1457;
(e) Moniri NH. Biochem Pharmacol. 2016;110–111:1;
(f) Li Z, Qiu Q, Geng X, Yang J, Huang W, Qian H. Exp Opin Invest Drugs.
2016;25:871;
(g) Sparks SM, Aquino C, Banker P, et al. Bioorg Med Chem Lett. 2017;27:
1278;
To confirm that the glucose-lowering effect of compound 27
was indeed on target, an intra-peritoneal GTT was performed at
10 mg/kg in GPR120 knockout mice and in wild-type mice
(Fig. 3).4a While 27 lowered glucose to a significant extent in the
wild-type mice, it did not show any appreciable difference in
acute glucose-lowering in the GPR120 KO mice compared to
vehicle thus confirming that the glucose-lowering effect is due to
agonism of GPR120, and not an off-target effect from GPR40, for
example.
(h) McCoull W, Bailey A, Barton P, et al. J Med Chem. 2017;60:3187;
(i) Adams GL, Velasquez F, Jayne C, et al. ACS Med Chem Lett. 2017;8:96;
(j) Azevedo CMG, Watterson KR, Wargent ET, et al. J Med Chem. 2016;59:8868;
(k) Cox JM, Chu HD, Chelliah MV, et al. ACS Med Chem Lett. 2017;8:49.
4. (a) Zhang X, Cai C, Sui Z, et al. ACS Med Chem Lett. 2017;8:947;
(b) Zhang X, Cai C, Winters M, et al. Bioorg Med Chem Lett.. 2017;27:3272(c) Z.
Sui C. Cai X. Zhang, US Pat Appl, 20140275179 A1 20140918.
5. Z. Sui, M.P. Winters, All synthetic protocols for each of the compounds
described in this paper, and all biological assay protocols not described here,
are included in the following patent, US Pat. WO 2015134038 A1 20150911 or
in ref. 4 (a).
Next, compound 27 was assessed in a 2-week DIO mouse model
of type 2 diabetes to determine its effects on metabolic parameters
after repeat dosing (Table 6).14 Compound 27 was delivered at 10
mg/kg QD, 10 mg/kg BID, and 30 mg/kg QD along with a positive
control, rosiglitazone at 3 mg/kg QD, for 15 days. Both rosiglita-
zone and the 30 mg/kg dose of 27 significantly reduced both
fasting glucose and insulin compared to vehicle. 27 at 30 mg/kg
also significantly lowered body weight 4.6% relative to vehicle in
this study. These effects on both fasting glucose and insulin at
30 mg/kg QD were confirmed in a second 15-day DIO assay
conducted in the same manner, and give credence to the insulin-
sensitizing effects of GPR120 agonism.15
In summary, we have discovered a new class of heterocyclic
GPR120 agonists with good in vitro potency and acceptable phar-
macokinetic properties that show positive effects in both acute
and chronic rodent models of type 2 diabetes. We hope to report
on further development of this series in the future.
6. Synthesis of methyl 3-chloro-1H-pyrrole-2-carboxylate was conducted
following the procedure in WO 2008154271 and then employed in an
analogous fashion as in Scheme 1. The 3-Br compound was synthesized from
intermediate
5 following Scheme 1. The 3-CN compound was made by
cyanation of the 3-Br product of the Mitsunobu reaction with Zn(CN)2/Pd
(DPPF)2.
7. Ohta T, Fukuda T, Ishibashi F, Iwao M. J Org Chem. 2009;74:8143.
8. Chen Q, Wu S. J Chem Soc Chem Commun. 1989;705–706.
9. Chan DMT, Monaco KL, Wang R, Winters MP. Tet Lett. 1998;39:2933.
10. Bartnik R, Bensadat A, Cal D, et al. Bull Soc Chim Fr. 1997;134:725.
11. (a) Sekiguchi H, Kasubuchi M, Hasegawa S, Pelisch N, Kimura L, Ichimura A.
Curr Diabetes Rev. 2015;11:107;
(b) Defossa E, Wagner M. Bioorg Med Chem Lett. 2014;24:2991;
(c) Sameer M. Curr Drug Targets. 2016;17:1292;
(d) Ishimura A, Hasegawa S, Kasubuchi M, Kimura I. Front Pharmacol. 2014;5:1.
12. This finding is most likely due to the high dose of 27 and not the GPR120
mechanism because other series did not show this effect. No further
cardiovascular safety studies were performed with 27.
13. The plasma exposure from the 3 mg/kg dose was predicted to be 3974 ng/mL
from the mouse PK study which corresponds to 8.64
lM 27. Factoring in the
high plasma protein binding in the mouse (99.63%) gives an effective
concentration of 27 of 32 nM, lower than the 193 nM EC50 in mouse, but
Please cite this article in press as: Winters M.P., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.02.013

6
M.P. Winters et al. / Bioorganic & Medicinal Chemistry Letters xxx (2018) xxx–xxx
not so low that it would be inconceivable to see an effect. See Smith DA, Kerns
EH. Nature Rev. Drug Disc.. 2010;9:929. for a discussion on the limitations of
using plasma protein binding in drug discovery.
On Day 15, animals were fasted for 6 hours prior bleeding for determination of
glucose and insulin levels. Statistical significance of changes in glucose and
insulin from baseline were determined by ANCOVA. It is noted that the modest
decrease in insulin levels and the slight decline in body weight in the vehicle
group on day 15 relative to baseline may have been due to stress from the
twice-daily handling of the animals.
14. Chronic study methods: Diet-induced obese mice from an internal colony of
C57Bl/6 mice that were at least 21 weeks of age and had been on a high-fat diet
for at least 15 weeks were sorted based on plasma glucose levels and body
weight. The mice weighing 52–54g at study start, were sham-dosed for 5 days
and then administered the indicated dose in 0.5% methylcellulose for 15 days.
15. In a duplicate 15-day DIO mouse study, 30 mg/kg 27 lowered fasting glucose
28.5% compared to vehicle and lowered fasting insulin 4759 pg/mL.
Please cite this article in press as: Winters M.P., et al. Bioorg. Med. Chem. Lett. (2018), https://doi.org/10.1016/j.bmcl.2018.02.013