Development of a novel class of peroxisome proliferator-activated receptor (PPAR) gamma ligands as an anticancer agent with a unique binding mode based on a non-thiazolidinedione scaffold

Article abstract of DOI:10.1016/j.bmc.2019.115122

We previously identified dibenzooxepine derivative 1 as a potent PPARγ ligand with a unique binding mode owing to its non-thiazolidinedione scaffold. However, while 1 showed remarkably potent MKN-45 gastric cancer cell aggregation activity, an indicator o

Full text of DOI:10.1016/j.bmc.2019.115122

Bioorganic&MedicinalChemistry27(2019)115122
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Development of a novel class of peroxisome proliferator-activated receptor
(PPAR) gamma ligands as an anticancer agent with a unique binding mode
based on a non-thiazolidinedione scaffold
⁎
Keisuke Yamamotoa,b, , Tomohiro Tamuraa, Rina Nakamuraa, Shintaro Hosoea,
Masahiro Matsubaraa, Keiko Nagataa, Hiroshi Kodairaa, Takeshi Uemoria, Yuichi Takahashia,
⁎⁎
Michihiko Suzukia, Jun-ichi Saitoa, Kimihisa Uenoa, Satoshi Shutob,c,
a Fuji Research Park, R&D Division, Kyowa Kirin, 1188, Shimotogari, Nagaizumi-cho, Sunto-gun, Shizuoka, Japan
b Faculty of Pharmaceutical Sciences, Hokkaido University, Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812, Japan
c Center for Research and Education on Drug Discovery, Hokkaido University, Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812, Japan
A B S T R A C T
We previously identified dibenzooxepine derivative 1 as a potent PPARγ ligand with a unique binding mode owing to its non-thiazolidinedione scaffold. However,
while 1 showed remarkably potent MKN-45 gastric cancer cell aggregation activity, an indicator of cancer differentiation-inducing activity induced by PPARγ
activation, we recognized that 1 was metabolically unstable. In the present study, we identified a metabolically soft spot, and successfully discovered 3-fluoro
dibenzooxepine derivative 9 with better metabolic stability. Further optimization provided imidazo[1,2-a]pyridine derivative 17, which showed potent MKN-45
gastric cancer cell aggregation activity and excellent PK profiles compared with 9. Compound 17 exerted a growth inhibitory effect on AsPC-1/AG1 pancreatic tumor
in mice. Furthermore, the decrease in the hematocrit (an indicator of localized edema, a serious adverse effect of PPARγ ligands) was tolerable even with oral
administration at 200 mg/kg in healthy mice.
1. Introduction
Considering the differentiation of pre-adipocytes based on PPARγ
activation, compounds that activate PPARγ might lead to the differ-
entiation induction of cancer cells. Indeed, the TZD derivative efatu-
tazone (CS-7017), a potent PPARγ full agonist, has been reported as a
cancer differentiation-inducing agent,8,20 inducing the differentiation
of several cancer, including anaplastic thyroid carcinoma,20 non-small
cell lung cancer,21 and pancreatic cancer.8 However, in the clinical
studies of efatutazone, some adverse effects, such as localized edema,
were observed, and these adverse effects are likely due to the structural
properties of the TZD moiety.20,22 Similar adverse effects, largely
characterized by fluid retention, have also been reported in both pre-
clinical and clinical studies of other PPARγ full agonists. Thus, avoiding
of fluid retention should be attempted in the development of PPARγ
ligands as antitumor drugs.
Peroxisome proliferator-activated receptor gamma (PPARγ) is a
member of the nuclear receptor superfamily and regulates the expres-
sion of genes involved in peripheral insulin sensitivity, adipogenesis
and glucose homeostasis.1–3 PPARγ heterodimerizes with retinoid X
receptor α (RXRα), a nuclear receptor, and binds to the DNA response
element (PPAR response element [PPRE]) of each target gene to induce
their transactivation activities.4
A number of selective PPARγ agonists have been developed, and the
structures of well-known PPARγ agonists, containing the thiazolidine-
dione (TZD) structure, e.g., pioglitazone and rosiglitazone, are shown in
Fig. 1.5–13 PPARγ agonists have been reported to be beneficial for
treating type 2 diabetes mellitus. PPARγ plays a key role in adipogen-
esis,14 particularly the differentiation from pre-adipocytes to adipo-
cytes, by regulating the expression of adipose-related genes, such as
adipocyte fatty acid-binding protein 2 (aP2),15 adipose differentiation-
related protein (Adfp) and Adiponectin.16 For instance, PPARγ agonists
strongly promote the differentiation of fibroblast-like cells such as 3T3-
L1 cells to adipocytes.17–19
We previously reported dibenzooxepine derivative 1,23 which has a
non-TZD structure. Compound 1, which was derived from high
throughput screening hit compound dibenzoazepine derivative
2
(Fig. 2), showed a unique binding mode and potent differentiation-in-
ducing activity of cancer cells. Efficient lead optimization of hit com-
pound 2 was likely to be troublesome due to the asymmetric center
⁎ Correspondence to: K. Yamamoto, Small Molecule Drug Research Laboratories, Fuji Research Park, Shimotogari, Nagaizumi-cho, Sunto-gun, Shizuoka 411-8731,
Japan.
⁎⁎ Correspondence to: S. Shuto, Faculty of Pharmaceutical Sciences, Hokkaido University, Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812, Japan.
E-mail addresses: keisuke.yamamoto.fn@kyowakirin.com (K. Yamamoto), shu@pharm.hokudai.ac.jp (S. Shuto).
https://doi.org/10.1016/j.bmc.2019.115122
Received 26 June 2019; Received in revised form 10 September 2019; Accepted 14 September 2019
Availableonline16September2019
0968-0896/©2019ElsevierLtd.Allrightsreserved.

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
Fig. 1. The representative PPARγ ligands.
Fig. 2. Identification of lead compound 1 from HTS hit compound 2.
adjacent to the nitrogen atom on the azepine ring. To eliminate the
chiral center, its bioisosteric scaffolds were designed based on the
conformational search of model substrates (R)-2′ and (S)-2′ (Fig. 3).23
As a result, we found that olefinic E/Z isomers 4 and 5 effectively mimic
the chiral structures of (R)-2′ and (S)-2′, respectively, and discovered
dihydrodibenzocycloheptene derivative 3, which has potent PPARγ
reporter activity comparable to that of 2. Further optimization of the
core structure and the benzimidazole moiety of 2 led to 1, which has a
novel binding mode to PPARγ ligand binding domain (LBD) and potent
MKN-45 aggregation activity, an indicator of cancer differentiation-
inducing activity.
structural optimization of 1 resulting in the discovery of a novel class of
anticancer PPARγ ligands, with reduced adverse effects of severe fluid
retention.
2. Result and discussion
2.1. Identification of 3-fluoro dibenzooxepine derivative 9
Fig. 4 shows the very short half-life of 1 when administered orally to
mice. To reduce lipophilicity of tricyclic ring and improve metabolic
stability, dihydrobenzo[6,7]oxepino[3,4-b]pyridine derivatives 6 and
7, in which either of the benzenes in the core structure was replaced
with pyridine in order to reduce the electron density of the aromatic
ring and thereby prevent oxidative metabolism, were designed and
synthesized (Fig. 5). The oxygen atom of 7 in the central seven-mem-
bered ring was shifted to the position adjacent to that of 1 for the
However, we recognized that the clinical development of 1 as an
anticancer agent was difficult due to its very poor metabolic stability.
Following the structural optimization of 1, we successfully developed a
promising drug candidate with a good PK profile that showed exerted
efficacy in a mouse model of pancreatic cancer. We herein report the
2

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
down-syn ((R)-2’) /down-Z (4)
down-anti ((S)-2’) /down-E (5)
Fig. 3. Overlay of the most stable conformations of the chiral model compounds (R)-2′ and (S)-2′ on their non-chiral olefinic bioisosters 4 and 5, respectively. Global
energy minimum conformations of these compounds were calculated by the MacroModel. The hydrogens, nitrogens, and 2-methyl group are colored white, blue and
magenta, respectively.
the right benzene moiety of dibenzooxepine was likely to be the main
metabolized site.
Based on these results, we attempted to introduce a fluorine atom at
the 1-, 2-, 3- or 4-position on the right benzene ring to block the oxi-
dative metabolism, likely by CYPs. All of the derivatives were suc-
cessfully synthesized except for the 1-fluoro derivative, the preparation
of which failed due to regioselectivity in the construction reaction of
the dibenzooxepine ring. All of these three fluoro-substituted com-
pounds exerted potent activities both in the reporter gene assay and
aggregation assay using MKN-45 cancer cells (Table 1). In the human
liver microsome metabolism evaluation, only 9, the 3-fluoro substituted
derivative, was remarkably stable (3.2 L/h/kg) compared to the other
derivatives 8 and 10 (> 24.9 and 22.6 L/h/kg, respectively). These
results suggested that the 3-position of dibenzooxepine was the main
metabolic site. This sort of medicinal chemistry approach to preventing
oxidative metabolism by introducing a fluorine on benzene ring has
been previously reported.24
Next, we investigated the in vivo mouse pharmacokinetic profile of 9
(Fig. 6). The 3-fluoro derivative 9 had a much longer half-life and better
AUC than lead compound 1 in mice. Thus, we identified the di-
benzooxepine derivative 9 with good pharmacokinetic profiles by the
introduction of a fluorine atom at the 3-position.
However, further structure optimization was needed to achieve a
better pharmacokinetic profile and reduce the pharmacologically ef-
fective dose. The acidic oxadiazolone moiety of 9 seemed to be essential
Fig. 4. Plasma concentrations of 1 after its oral administration in BALB/c mice
for binding to PPARγ-LBD and exerting PPARγ agonistic activity as
(30 mg/kg, n = 2).
demonstrated in other PPARγ ligands. The tricyclic core structure was
also deemed necessary in order to mimic the bioactive mode of lead
synthetic convenience.
compound 2, as we previously reported. We therefore conducted fur-
ther optimization of the benzimidazole moiety of 9 by referencing the
crystal structure of our compound with PPARγ-LBD.
The synthetic compounds were evaluated in vitro via an HEK293
PPARγ reporter gene assay and cell aggregation assay to assess the
differentiation-inducing activity against gastric cancer cell line MKN-
45, as in our previous study.23 In addition, compounds were treated
with human liver microsome to investigate the metabolic stability.
In the reporter gene assay, the left-pyridine derivative 6 showed an
EC50 value 86 nM higher than that of 1, and the activity of the right-
pyridine derivative 7 was significantly reduced (EC50 = 2194 nM).
However, the human liver microsome stability of 7 was significantly
better than that of 6 (2.06 L/h/kg vs. > 24.9 L/h/kg), suggesting that
The efficacies and EC50 values of compounds 1 and 8–10 in human
PPARγ/GAL4 transfected HEK293EBNA cells at 24 h after drug treat-
ment. The efficacy of each compound was calculated as the percentage
of the maximum activation obtained with pioglitazone at 1000 nM.
The aggregation of MKN-45 cells was evaluated using an IN Cell
Analyzer 1000 (GE Healthcare) after treatment of compounds 1 and
8–10 for 5 days. The aggregation efficacies were shown as the values
relative to the maximum efficacy of 1. EC50 values were determined
3

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
Fig. 5. Improvement of the human metabolic stability of lead compound 1 by modification of the tricyclic core-structure.
Table 1
Effects on PPARγ reporter, MKN-45 aggregation activities and human hepatic
intrinsic clearance of fluoro dibenzooxepine derivatives.
Compd. R1
X
Reporter gene assay
MKN-45 Aggregation
assay
hCLint
(L/h/kg)
EC50(nM) Efficacy(%) EC50(nM)
Efficacy
(%)
1
Me
H
H
2.4
9.5
0.6
1.5
3.3
7.0
100
102
104
101
> 24.9
> 24.9
3.2
8
2-F 12
3-F 7.4
4-F 4.4
6.9
9
H
11.7
8.7
3.6 1.0
10
H
1.6
3.8
22.6
EC50 values were determined using the XLFit.
The hCLint was calculated from the elimination rate constant of each compound
in human liver microsomes (n = 2).
Fig. 6. Plasma concentrations of 1 and 3-fluoro-dibenzooxepine derivative 9 in
BALB/c mice orally administered (30 mg/kg, n = 2).
using the XLFit.
carbon for the hydrophilic nitrogen atom in benzene of benzimidazole
(i.e. 11 and 12) resulted in decreased PPARγ reporter activities
(EC50 = 32 and 24 nM, respectively) compared to lead compound 1
(Table 2). We then attempted to convert the benzimidazole moiety to
other hydrophobic structures. As expected, thieno[3,4-d]imidazole de-
rivative 13 and imidazo[1,2-a]pyridine derivatives 14 showed very
potent aggregation activity (EC50 = 3.0 and 5.3 nM, respectively)
comparable to 9. Although the mouse AUC of 13 was very low, the AUC
2.2. Discovery of imidazo[1,2-a]pyridine derivative 17 and the binding
mode of 17
According to the X-ray crystal structure of 1 complex with PPARγ
LBD,23 the benzimidazole moiety of 1 was surrounded by hydrophobic
residues of the canonical site, consisting of the residues of Helix 3 (H3)
and Helix 5 (H5).25 This suggested that a bulky and/or polar substituent
on the benzimidazole ring seemed to be intolerant. The substitution of
4

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
Table 2
Effects on PPARγ reporter, MKN-45 aggregation activities and pharmacokinetic para-
meters of dibenzooxepine derivatives modified at the 8-substituted heterocycles.
EC50 values were determined using the XLFit.
N.T.; not tested.
5

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
of 14 was more than twice that of 9. The discovery of 14, which showed
the desirable aggregation activity and pharmacokinetic profile, en-
couraged us to conduct further optimization of the imidazo[1,2-a]pyr-
idine moiety.
The optimization studies showed that compound 17 exhibited the
most promising aggregation activity and PK profile. The binding mode
of 17 to PPARγ LBD was next investigated. Fig. 7A shows the X-ray
analyzed binding mode of 17, in which the acidic proton of ox-
adiazolone forms a hydrogen bond with the phenolic oxygen of Tyr473
and this interaction is essential for PPARγ activation.26 The imidazo
[1,2-a]pyridine moiety filled the canonical site and the right benzene
moiety of the tricyclic core structure was accommodated in the small
lipophilic pocket, surrounded by H3 and helix 11 (H11). Fig. 7B shows
the superimposition of the complex structure of 17 and that of 1.23 Even
though the placement of atoms at the side aromatic ring differed, the
structure of 17 overlapped well with that of 1.
We next designed and synthesized imidazo[1,2-a]pyridine deriva-
tives 15–21 (Table 2). Chlorine substituted derivatives 15, 16 and 17
exerted quite potent MKN-45 aggregation activity, with EC50 values of
2.0, 1.4 and 2.5 nM, respectively. The 7-chloro imidazo[1,2-a]pyridine
derivative
16
showed
an
excellent
mouse
PK
profile
(AUC = 10,200 ng·h/mL) compared to 14 (AUC = 3750 ng·h/mL), but
its metabolic stability was slightly lower than that of 17. Although 7-
fluoro imidazo[1,2-a]pyridine derivative 18 also had potent aggrega-
tion activity (EC50 = 1.7 nM) and good AUC (AUC = 5510 ng·h/mL) in
mice, its human microsome stability was not sufficient (hCLint = 3.9 L/
h/kg). Although the introduction of a methoxyethyl group at the 2-
position on imidazo[1,2-a]pyridine led to potent aggregation EC50 va-
lues (19: 1.3 nM and 20: 4.5 nM), the mouse PK profiles of these
compounds were not sufficient (AUC = 2450 and 1086 ng·h/mL, re-
spectively). In contrast, 8,9-dihydrofuro[3,2-c]imidazo[1,2-a]pyridine
derivative 21 showed an excellent mouse AUC (8760 ng·h/mL); how-
ever, the aggregation activity (EC50 = 19 nM) was slightly weaker than
that of 14 (EC50 = 5.3 nM).
Therefore, we selected 17, which had an excellent aggregation ac-
tivity and PK profile and a unique binding mode, as the candidate for
further in vivo evaluations.
The efficacies and EC50 values of compounds 9 and 11–21 in human
PPARγ/GAL4 transfected HEK293EBNA cells at 24 h after drug treat-
ment. The efficacy of each compound was calculated as the percentage
of the maximum activation obtained with pioglitazone at 1000 nM.
The aggregation of MKN-45 cells was evaluated using an IN Cell
Analyzer 1000 (GE Healthcare) after treatment of compounds 9 and
11–21 for 5 days. The aggregation efficacies were shown as the values
(A)
O
H
N
O
Me
N
O
N
F
N
Cl
17
(B)
O
H
N
O
Me
Me
N
O
N
N
Me
1
O
H
N
O
Me
N
O
N
F
N
Cl
17
Fig. 7. The X-ray crystal structures of ligands complexed with PPARγ(PDB: 6K0T). (A) The binding mode of 17 (ball and stick; green) to the PPARγ LBD. The proton
of the oxadiazolone ring of 17 interacted with the phenolic oxygen of Tyr473. 2ME: 2-mercaptoethanol, which was used in the protein purification buffer and found
in the ligand binding pocket as a covalent adduct to Cys285. (B) The superimposition of the complex structures of 17 (green) and 1 (gray, PDB: 6AD9).
6

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
relative to the maximum efficacy of 1.
The AUC was calculated from plasma drug concentrations in BALB/
c mice treated orally at 10 mg/kg (n = 2). The hCLint was calculated
from the elimination rate constant of each compound in human liver
microsomes (n = 2).
2.3. Effect of 17 on an AsPC-1 xenograft mouse model
We evaluated the in vivo antitumor activity of 17 using an AsPC-1
pancreatic cancer xenograft mouse model. Because MKN-45 is known as
a poorly differentiated cancer cell line, its engraftment was not ob-
served in SCID mice. The differentiation-inducing activity against AsPC-
1 cells by PPARγ activation was previously reported in preclinical
studies of efatutazone.8 Compared to the vehicle control, the oral ad-
ministration of 17 clearly reduced the tumor volume (Fig. 8). After
10 days, the tumor growth was significantly reduced at > 25 mg/kg
(oral administration), the point at which the antitumor efficacy in this
model was saturated. Therefore, the minimum effective dose was likely
to be between 12.5 and 25 mg/kg in this model.
Next, we measured the mRNA expression of angiopoietin-like 4
(angptl4),27 a PPARγ-regulated downstream gene, and keratin 20
(krt20),28 a cell-differentiation marker, of AsPC-1/AG1 tumor cells in
the xenografts. Compound 17 was orally administered at 100 mg/kg
twice a day for 14 days, and at 4 h after the last dose, the quantitative
polymerase chain reaction (qPCR) results of the 17-treated tumor
showed that this therapy markedly increased the expression of angptl4
and krt20 compared to vehicle control. These data suggested that the
tumor differentiation-inducing activity of the 17 in vivo model was due
to its PPARγ activation potency.
$: p < .05 between the vehicle-treated group and the 17 50 mg/kg-
treated group. *: p < .05 between the vehicle-treated group and the
17,100 mg/kg-treated group.
(B) The mRNA expression of angptl4 and krt20 in the xeno-
transplanted tumor. Compound 17 was orally administered at 12.5, 25,
50 and 100 mg/kg twice a daily for 14 days. The gene expression was
determined by quantitative PCR using an ABI PCR system.
2.4. Effect on fluid retention in CD-1 mice
In preclinical and clinical studies, fluid retention due to the ad-
ministration of PPARγ ligands has been reported.29,30 This adverse ef-
fect is likely to be caused by the TZD structure.20,22 We therefore in-
vestigated whether or not dibenzooxepine derivative 17, which has a
non-TZD structure, caused fluid retention.
To assess fluid retention, the hematocrit values of healthy CD-1
mice treated by 17 were measured (Fig. 9). At 50 mg/kg of oral ad-
ministration of 17, which was determined to be the effective dose in
AsPC-1 xenograft mouse experiments, the hematocrit value was slightly
decreased (17, 42.9%; vehicle, 44.1%). At 200 mg/kg of oral adminis-
tration, which was 8 times higher than the lower effective dose, the
decrease in the hematocrit value was still only slight (42.0%). There
were no significant differences between the vehicle and 17-treated
groups (50 and 200 mg/kg). Compared to the fluid retention induced by
pioglitazone administration in mice,29 that of 17 was considered to be
quite moderate.
Thus, severe fluid retention was successfully avoided by lead opti-
mization from the novel scaffold of dibenzooxepine without a TZD
moiety.
Fig. 8. Antitumor effect of 17 on AsPC-1/AG1 xenotransplanted mice. (A)
Tumor growth inhibition after the administration of 17 in the xenotransplanted
2.5. Chemistry
mice. Each plot represents the mean
SE of the relative tumor volume
(n = 5). #: p < .05 between the vehicle-treated group and the 17 25 mg/kg-
treated group. All of the p values were calculated by the Steel test. The fold-
inductions are shown as the values relative to baseline.
Compounds 8–13 were synthesized from 8-(bromomethyl)-dibenzo
[b,e]oxepin intermediates 22–24 (supporting information) as
7

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
Synthesis of imidazo[1,2-a]pyridine derivatives 14–18 were re-
ferred to the synthesis of Cu-catalyzed three component coupling re-
action, which was reported as one-pot synthesis of alipidem as shown in
Scheme 2.31 The reaction in the presence of Cu(OTf)2, 2-aminopyridine,
aldehyde and 8-ethynyl-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)
propanenitrile 31 (supporting information), produced 32–36. Key in-
termediate 31 was the mixture of E/Z isomers, and accordingly this
coupling reaction provided an E/Z mixture, which was separated by
silica column chromatography to afford 32–36 as (E)-isomers. Target
oxadiazolones 14–18 were prepared by the cyclization reaction de-
scribed above.
The introduction of an imidazo[1,2-a]pyridine ring using (E)-2-(3-
fluoro-8-formyldibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile 37
(supporting information) is shown in Scheme 3. A coupling reaction
with Grignard reagents and iodoaryl intermediates gave hydroxybenzyl
compounds, which were reduced with silyl reduction reagents to gen-
erate 38–40. The synthesis of oxadiazolone using 38–40 provided
target compounds 19–21.
3. Conclusion
Based on the results of the metabolite analysis of 1 and modification
of both benzene rings of the core structure, we designed fluorine-sub-
stituted derivatives. 3-fluoro dibenzooxepine derivative 9 was meta-
bolically stable, and accordingly exerted better PK than 1 in mice. The
further optimization of the benzimidazole moiety led to imidazo[1,2-a]
pyridine 17, which not only showed better mouse PK than 1, but also
maintained potent MKN-45 aggregation activity, probably derived from
the unique binding mode to PPARγ. Furthermore, compound 17 exerted
tumor growth inhibition in AsPC-1/AG1 pancreatic cancer xenografted
mice. Although TZD derivatives have been concerned severe fluid re-
tention, the hematocrit value of 17 was only slightly decreased, even
when administered at a much higher dose than its effective dose. This
result suggested that treatment with 17 would not induce the severe
Fig. 9. Hematocrit values in CD-1 mice treated orally with 17 twice a daily at
50 and 200 mg/kg. p = 0.671 between the vehicle-treated group and the 17
50 mg/kg-treated group. p = 0.176 between the vehicle-treated group and the
17,200 mg/kg-treated group.
summarized in Scheme 1. Treatment of imidazole compounds with
K2CO3 at room temperature, followed by addition of 22–24, afforded
25–30. The addition of hydroxylamine to the cyano group of 25–30,
followed by a ring-closure reaction using ethyl chloroformate and tert-
BuOK afforded the target oxadiazolone compounds 8–13.
Scheme 1. Synthesis of benzimidazole analoguesa
8

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
Scheme 2. Introduction of the substituents to benz [1,2-a]imidazole ringa
fluid retention.
4.2. Chemistry
In conclusion, we successfully developed potential anticancer agent
17 with promising properties. Further research using this compound for
clinical studies will be expected to offer novel options for cancer
therapy.
4.2.1. (E)-3-(1-(2-fluoro-8-((2-propyl-1H-benzo[d]imidazol-1-yl)methyl)
dibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-oxadiazol-5(4H)-one (8)
To a stirred solution of 25 (63 mg, 0.13 mmol) in ethanol (2 mL) was
added NH2OH solution 50 wt% in water (0.42 mL, 6.4 mmol), and the
solution was refluxed overnight. After the consumption of the starting
material, the reaction mixture was poured into water and extracted
twice with CHCl3. The combined organic layers were washed with
brine, dried over sodium sulfate, and filtered. The organic layer was
concentrated to dryness and the obtained residue was dissolved in
CH2Cl2 (2 mL). To a stirred solution was added triethylamine (26 μL,
0.19 mmol) and ethyl chloroformate (18 μL, 0.19 mmol) the solution
was stirred for 1 h at room temperature. The reaction mixture was
poured into saturated aqueous sodium bicarbonate solution and ex-
tracted twice with ethyl acetate. The combined organic layers were
washed with brine, dried over sodium sulfate, and filtered. The organic
layer was concentrated to give the residue. To a solution of the residue
in toluene (2 mL) was added tert-BuOK (28 mg, 0.25 mmol) and the
solution was stirred for 30 min at room temperature. The reaction
mixture was poured into 5% aqueous citric acid solution and extracted
twice with CHCl3. The combined organic layers were washed with
brine, dried over sodium sulfate, and filtered. The organic layer was
concentrated to give the residue. The obtained residue was then pur-
ified by flash column chromatography on silica gel (100:0 to 95:5
4. Experimental section
4.1. General methods
All reagents and solvents were procured from commercial sources
and used as received. Thin layer chromatography (TLC) was carried out
using Merck GmbH Precoated silica gel 60 F254. Chromatography on
silica gel was carried out using prepacked silica gel cartridges (Yamazen
Hi-Flash Column Silicagel or Wako Presep® Silicagel). Microwave re-
action was conducted using Biotage Initiator 2.5.
Chemical shifts in 1H NMR spectra were reported in δ values (ppm)
relative to tetramethylsilane. HPLC analyses were performed following
conditions: Waters Xbrige® C18 column (3.5 μm, 4.6 mm × 50 mm),
30 °C column temperature, 1.0 mL/min flow rate, photodiode array
detection (254 nm), and linear mobile phase gradient of 20%–90% B
over 5 min, holding for 3.5 min at 20% B (mobile phase A, 0.05% tri-
fluoroacetic acid in water; mobile phase B, acetonitrile), by which the
purities of final compounds were confirmed as > 95%. Mass spectra
were recorded on a Waters 2695 (ESI-MS).
9

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
Scheme 3. Synthesis of the side chain structure from 8-formyldihydrodibenzooxepinea
CHCl3/methanol) to give 8 (16 mg, 25%) as an amorphous. 1H NMR
(300 MHz, CDCl3): δ 0.96 (t, J = 7.4 Hz, 3H), 1.73–1.91 (m, 2H), 2.31
(s, 3H), 2.72–2.82 (m, 2H), 4.59 (d, J = 12.8 Hz, 1H), 5.28–5.42 (m,
3H), 6.71–6.97 (m, 4H), 6.97–7.06 (m, 1H), 7.08–7.25 (m, 4H),
7.58–7.73 (m, 1H). The proton of oxadiazolone was not observed. LC/
MS (ESI, [M + H]+, m/z) 497. HPLC: purity 96%, RT 3.82 min.
solution and extracted twice with CHCl3. The combined organic layers
were washed with brine, dried over sodium sulfate, and filtered. The
organic layer was concentrated to give the residue. The obtained re-
sidue was then purified by flash column chromatography on silica gel
(100:0 to 95:5 CHCl3/methanol) to give 9 (68 mg, 54%) as an amor-
phous. 1H NMR (300 MHz, CDCl3): δ 0.98 (t, J = 7.4 Hz, 3H), 1.74–1.94
(m, 2H), 2.28 (s, 3H), 2.79 (t, J = 7.4 Hz, 2H), 4.65 (d, J = 12.0 Hz,
1H), 5.34 (s, 2H), 5.45 (d, J = 12.0 Hz, 1H), 6.48–6.56 (m, 1H),
6.60–6.70 (m, 1H), 6.95–7.30 (m, 7H), 7.67–7.74 (m, 1H). The proton
of oxadiazolone was not observed. LC/MS (ESI, [M + H]+, m/z) 497.
HPLC: purity 99%, RT 3.90 min.
4.2.2. (E)-3-(1-(3-fluoro-8-((2-propyl-1H-benzo[d]imidazol-1-yl)methyl)
dibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-oxadiazol-5(4H)-one (9)
To a stirred solution of 26 (110 mg, 0.251 mmol) in ethanol
(1.3 mL) was added NH2OH solution 50 wt% in water (0.769 mL,
12.3 mmol), and the solution was refluxed overnight. After the con-
sumption of the starting material, the reaction mixture was poured into
water and extracted twice with CHCl3. The combined organic layers
were washed with brine, dried over sodium sulfate, and filtered. The
organic layer was concentrated to dryness and the obtained residue was
dissolved in CH2Cl2 (1.3 mL). To a stirred solution was added triethy-
lamine (70 μL, 0.50 mmol) and ethyl chloroformate (48 μL, 0.50 mmol),
and the solution was stirred for 1 h at room temperature. The reaction
mixture was poured into saturated aqueous sodium bicarbonate solu-
tion and extracted twice with ethyl acetate. The combined organic
layers were washed with brine, dried over sodium sulfate, and filtered.
The organic layer was concentrated to give the residue. To a solution of
the residue in toluene (1 mL)/THF (1 mL) was added tert-BuOK (85 mg,
0.75 mmol) and the solution was stirred for 30 min at room tempera-
ture. The reaction mixture was poured into 5% aqueous citric acid
4.2.3. (E)-3-(1-(4-fluoro-8-((2-propyl-1H-benzo[d]imidazol-1-yl)methyl)
dibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-oxadiazol-5(4H)-one
(10)
To a stirred solution of 27 (160 mg, 0.366 mmol) in ethanol
(1.8 mL) was added NH2OH solution 50 wt% in water (1.12 mL,
18.3 mmol), and the solution was refluxed overnight. After the con-
sumption of the starting material, the reaction mixture was poured into
water and extracted twice with CHCl3. The combined organic layers
were washed with brine, dried over sodium sulfate, and filtered. The
organic layer was concentrated to dryness and the obtained residue was
dissolved in CH2Cl2 (1.8 mL). To a stirred solution was added triethy-
lamine (102 μL, 0.732 mmol) and ethyl chloroformate (70 μL,
0.73 mmol), and the solution was stirred for 1 h at room temperature.
The reaction mixture was poured into saturated aqueous sodium
10

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
bicarbonate solution and extracted twice with ethyl acetate. The com-
bined organic layers were washed with brine, dried over sodium sulfate,
and filtered. The organic layer was concentrated to give the residue. To
a solution of the residue in toluene (1 mL)/THF (1 mL) was added tert-
BuOK (123 mg, 1.09 mmol) and the solution was stirred for 30 min at
room temperature. The reaction mixture was poured into 5% aqueous
citric acid solution and extracted twice with CHCl3. The combined or-
ganic layers were washed with brine, dried over sodium sulfate, and
filtered. The organic layer was concentrated to give the residue. The
obtained residue was then purified by flash column chromatography on
silica gel (100:0 to 92:8 CHCl3/methanol) to give 10 (80 mg, 44%) as
an amorphous. 1H NMR (300 MHz, DMSO‑d6): δ 0.88 (t, J = 7.4 Hz,
3H), 1.61–1.83 (m, 2H), 2.18 (s, 3H), 2.77 (t, J = 7.6 Hz, 2H), 5.06 (d,
J = 12.4 Hz, 1H), 5.44–5.61 (m, 3H), 6.86–6.98 (m, 1H), 6.98–7.25 (m,
6H), 7.24–7.32 (m, 1H), 7.42–7.52 (m, 1H), 7.53–7.66 (m, 1H). LC/MS
(ESI, [M + H]+, m/z) 497. The proton of oxadiazolone was not ob-
served. HPLC: purity 99%, RT 3.78 min.
layer was concentrated to give the residue. To a solution of the residue
in toluene (1 mL)/THF (1 mL) was added tert-BuOK (43 mg, 0.39 mmol)
and the solution was stirred for 30 min at room temperature. The re-
action mixture was poured into 5% aqueous citric acid solution and
extracted twice with CHCl3. The combined organic layers were washed
with brine, dried over sodium sulfate, and filtered. The organic layer
was concentrated to give the residue. The obtained residue was then
purified by recrystallized from isopropyl alcohol to give 12 (37 mg,
37%) as an amorphous. 1H NMR (300 MHz, CDCl3): δ 1.02–1.17 (m,
2H), 1.24–1.39 (m, 2H), 1.83–1.97 (m, 1H), 2.29 (s, 3H), 4.70 (d,
J = 12.6 Hz, 1H), 5.46–5.59 (m, 3H), 6.46–6.60 (m, 1H), 6.60–6.73 (m,
1H), 6.96–7.25 (m, 5H), 7.99–8.15 (m, 1H), 8.57 (br s, 1H). LC/MS
(ESI, [M + H]+, m/z) 530. HPLC: purity 99%, RT 4.57 min.
4.2.6. (E)-3-(1-(8-((2-cyclopropyl-1H-thieno[3,4-d]imidazol-1-yl)
methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-
oxadiazol-5(4H)-one (13)
To a stirred solution of 30 (160 mg, 0.360 mmol) in ethanol
(1.8 mL) was added NH2OH solution 50 wt% in water (1.11 mL,
18.1 mmol), and the solution was refluxed overnight. After the con-
sumption of the starting material, the reaction mixture was poured into
water and extracted twice with CHCl3. The combined organic layers
were washed with brine, dried over sodium sulfate, and filtered. The
organic layer was concentrated to dryness and the obtained residue was
dissolved in CH2Cl2 (1.8 mL). To a stirred solution was added triethy-
lamine (101 μL, 0.725 mmol) and ethyl chloroformate (69 μL,
0.73 mmol), and the solution was stirred for 1 h at room temperature.
The reaction mixture was poured into saturated aqueous sodium bi-
carbonate solution and extracted twice with ethyl acetate. The com-
bined organic layers were washed with brine, dried over sodium sulfate,
and filtered. The organic layer was concentrated to give the residue. To
a solution of the residue in toluene (1 mL)/THF (1 mL) was added tert-
BuOK (81 mg, 0.73 mmol) and the solution was stirred for 30 min at
room temperature. The reaction mixture was poured into 5% aqueous
citric acid solution and extracted twice with CHCl3. The combined or-
ganic layers were washed with brine, dried over sodium sulfate, and
filtered. The organic layer was concentrated to give the residue. The
obtained residue was then purified by recrystallized from isopropyl
alcohol to give 13 (65 mg, 36%) as an amorphous. 1H NMR (300 MHz,
DMSO‑d6): δ 0.90–1.19 (m, 4H), 2.04–2.20 (m, 1H), 2.17 (s, 3H), 4.96
(d, J = 12.5 Hz, 1H), 5.34 (s, 2H), 5.52 (d, J = 12.5 Hz, 1H), 6.58–6.72
(m, 2H), 6.74–6.86 (m, 1H), 6.95 (d, J = 2.6 Hz, 1H), 7.09 (d,
J = 7.9 Hz, 1H), 7.19–7.29 (m, 2H), 7.39–7.44 (m, 1H), 12.16 (br s,
1H). LC/MS (ESI, [M + H]+, m/z) 501. HPLC: purity 98%, RT 3.77 min.
4.2.4. (E)-3-(1-(8-((7-chloro-2-cyclopropyl-3H-imidazo[4,5-b]pyridin-3-
yl)methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-
oxadiazol-5(4H)-one (11)
To a stirred solution of 28 (175 mg, 0.372 mmol) in ethanol
(1.9 mL) was added NH2OH solution 50 wt% in water (1.14 mL,
18.6 mmol), and the solution was refluxed overnight. After the con-
sumption of the starting material, the reaction mixture was poured into
water and extracted twice with CHCl3. The combined organic layers
were washed with brine, dried over sodium sulfate, and filtered. The
organic layer was concentrated to dryness and the obtained residue was
dissolved in CH2Cl2 (1.9 mL). To a stirred solution was added triethy-
lamine (104 μL, 0.744 mmol) and ethyl chloroformate (71 μL,
0.74 mmol), and the solution was stirred for 1 h at room temperature.
The reaction mixture was poured into saturated aqueous sodium bi-
carbonate solution and extracted twice with ethyl acetate. The com-
bined organic layers were washed with brine, dried over sodium sulfate,
and filtered. The organic layer was concentrated to give the residue. To
a solution of the residue in toluene (1 mL)/THF (1 mL) was added tert-
BuOK (125 mg, 1.12 mmol) and the solution was stirred for 30 min at
room temperature. The reaction mixture was poured into 5% aqueous
citric acid solution and extracted twice with CHCl3. The combined or-
ganic layers were washed with brine, dried over sodium sulfate, and
filtered. The organic layer was concentrated to give the residue. The
obtained residue was then purified by flash column chromatography on
silica gel (100:0 to 90:10 CHCl3/methanol) to give 11 (82 mg, 42%) as
an amorphous. 1H NMR (300 MHz, CDCl3): δ 1.05–1.17 (m, 2H),
1.24–1.33 (m, 2H), 1.85–1.97 (m, 1H), 2.28 (s, 3H), 4.78 (d,
J = 12.8 Hz, 1H), 5.45–5.68 (m, 3H), 6.47–6.58 (m, 1H), 6.58–6.71 (m,
1H), 6.98–7.29 (m, 5H), 8.14 (d, J = 5.4 Hz, 1H). The proton of ox-
adiazolone was not observed. LC/MS (ESI, [M + H]+, m/z) 530. HPLC:
purity 98%, RT 5.13 min.
4.2.7. (E)-3-(1-(8-((2-cyclopropylimidazo[1,2-a]pyridin-3-yl)methyl)-3-
fluorodibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-oxadiazol-5(4H)-
one (14)
To a stirred solution of 32 (201 mg, 0.462 mmol) in ethanol (3 mL)
was added NH2OH solution 50 wt% in water (1.5 mL, 23 mmol), and
the solution was refluxed overnight. After the consumption of the
starting material, the reaction mixture was poured into water and ex-
tracted twice with CHCl3. The combined organic layers were washed
with brine, dried over sodium sulfate, and filtered. The organic layer
was concentrated to dryness and the obtained residue was dissolved in
CH2Cl2 (2 mL). To a stirred solution was added triethylamine (129 μL,
0.923 mmol) and ethyl chloroformate (89 μL, 0.92 mmol), and the so-
lution was stirred for 1 h at room temperature. The reaction mixture
was poured into saturated aqueous sodium bicarbonate solution and
extracted twice with ethyl acetate. The combined organic layers were
washed with brine, dried over sodium sulfate, and filtered. The organic
layer was concentrated to give the residue. To a solution of the residue
in toluene (2 mL) was added tert-BuOK (104 mg, 0.923 mmol) and the
solution was stirred for 30 min at room temperature. The reaction
mixture was poured into 5% aqueous citric acid solution and extracted
twice with CHCl3. The combined organic layers were washed with
4.2.5. (E)-3-(1-(8-((4-chloro-2-cyclopropyl-1H-imidazo[4,5-c]pyridin-1-
yl)methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-
oxadiazol-5(4H)-one (12)
To a stirred solution of 29 (91 mg, 0.19 mmol) in ethanol (2.0 mL)
was added NH2OH solution 50 wt% in water (0.64 mL, 9.7 mmol), and
the solution was refluxed overnight. After the consumption of the
starting material, the reaction mixture was poured into water and ex-
tracted twice with CHCl3. The combined organic layers were washed
with brine, dried over sodium sulfate, and filtered. The organic layer
was concentrated to dryness and the obtained residue was dissolved in
CH2Cl2 (2.0 mL). To a stirred solution was added triethylamine (54 μL,
0.39 mmol) and ethyl chloroformate (37 μL, 0.39 mmol), and the solu-
tion was stirred for 1 h at room temperature. The reaction mixture was
poured into saturated aqueous sodium bicarbonate solution and ex-
tracted twice with ethyl acetate. The combined organic layers were
washed with brine, dried over sodium sulfate, and filtered. The organic
11

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
brine, dried over sodium sulfate, and filtered. The organic layer was
concentrated to give the residue. The obtained residue was then pur-
ified by flash column chromatography on silica gel (100:0 to 95:5
CHCl3/methanol) to give 14 (62 mg, 27%) as an amorphous. 1H NMR
(300 MHz, CDCl3):δ 0.93–1.07 (m, 2H), 1.07–1.21 (m, 2H), 1.95–2.09
(m, 1H), 2.29 (s, 3H), 4.29–4.47 (m, 2H), 4.74 (d, J = 12.8 Hz, 1H),
5.50 (d, J = 12.8 Hz, 1H), 6.47–6.58 (m, 1H), 6.59–6.75 (m, 2H),
7.01–7.25 (m, 5H), 7.44–7.54 (m, 1H), 7.54–7.63 (m, 1H). The proton
of oxadiazolone was not observed. LC/MS (ESI, [M + H]+, m/z) 495.
HPLC: purity 95%, RT 3.78 min.
(d, J = 17.1 Hz, 1H), 4.44 (d, J = 17.1 Hz, 1H), 4.90 (d, J = 12.7 Hz,
1H), 5.49 (d, J = 12.7 Hz, 1H), 6.65 (dd, J = 10.8, 2.0 Hz, 1H),
6.74–6.84 (m, 1H), 6.84 (dd, J = 7.8, 2.0 Hz, 1H), 7.02 (d, J = 7.8 Hz,
1H), 7.16 (d, J = 7.8 Hz, 1H), 7.19–7.20 (m, 1H), 7.26 (s, 1H), 7.57 (d,
J = 2.0 Hz, 1H), 8.09 (d, J = 7.8 Hz, 1H).
The proton of oxadiazolone was not observed. LC/MS (ESI,
[M + H]+, m/z) 529. HPLC: purity 95%, RT 3.88 min.
4.2.10. (E)-3-(1-(8-((8-chloro-2-cyclopropylimidazo[1,2-a]pyridin-3-yl)
methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-
oxadiazol-5(4H)-one (17)
4.2.8. (E)-3-(1-(8-((6-chloro-2-cyclopropylimidazo[1,2-a]pyridin-3-yl)
methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-
oxadiazol-5(4H)-one (15)
To a stirred solution of 35 (100 mg, 0.213 mmol) in ethanol (2 mL)
was added NH2OH solution 50 wt% in water (0.42 mL, 6.4 mmol), and
the solution was refluxed for 1 day. After the consumption of the
starting material, the reaction mixture was poured into water and ex-
tracted twice with CHCl3. The organic layer was concentrated to dry-
ness and the obtained residue was dissolved in CH2Cl2 (2 mL). To a
stirred solution was added triethylamine (59 μL, 0.43 mmol) and ethyl
chloroformate (41 μL, 0.43 mmol), and the solution was stirred for 1 h
at room temperature. The reaction mixture was poured into saturated
aqueous sodium bicarbonate solution and extracted twice with ethyl
acetate. The combined organic layers were washed with brine, dried
over sodium sulfate, and filtered. The organic layer was concentrated to
give the residue. To a solution of the residue in toluene (1 mL)/THF
(1 mL) was added tert-BuOK (48 mg, 0.43 mmol) and the solution was
stirred for 30 min at room temperature. The reaction mixture was
poured into 5% aqueous citric acid solution and extracted twice with
CHCl3. The combined organic layers were washed with brine, dried
over sodium sulfate, and filtered. The organic layer was concentrated to
give the residue. The obtained residue was then purified by flash
column chromatography on silica gel (100:0 to 90:10 CHCl3/methanol)
to give 17 (30 mg, 27%) as an amorphous. 1H NMR (300 MHz, CDCl3):δ
0.86–1.16 (m, 4H), 1.92–2.10 (m, 1H), 2.27 (s, 3H), 4.27–4.42 (m, 2H),
4.68 (d, J = 12.6 Hz, 1H), 5.48 (d, J = 12.6 Hz, 1H), 6.43–6.57 (m,
1H), 6.57–6.68 (m, 2H), 7.02–7.20 (m, 5H), 7.50–7.58 (m, 1H).The
proton of oxadiazolone was not observed. LC/MS (ESI, [M + H]+, m/z)
529. HPLC: purity 97%, RT 3.93 min.
To a stirred solution of 33 (88 mg, 0.19 mmol) in ethanol (2 mL) was
added NH2OH solution 50 wt% in water (33 μL, 5.6 mmol), and the
solution was refluxed for 1 day. After the consumption of the starting
material, the reaction mixture was poured into water and extracted
twice with CHCl3. The organic layer was concentrated to dryness and
the obtained residue was dissolved in CH2Cl2 (2 mL). To a stirred so-
lution was added triethylamine (52 μL, 0.38 mmol) and ethyl chlor-
oformate (36 μL, 0.38 mmol), and the solution was stirred for 1 h at
room temperature. The reaction mixture was poured into saturated
aqueous sodium bicarbonate solution and extracted twice with ethyl
acetate. The combined organic layers were washed with brine, dried
over sodium sulfate, and filtered. The organic layer was concentrated to
give the residue. To a solution of the residue in toluene (2 mL) was
added tert-BuOK (42 mg, 0.38 mmol) and the solution was stirred for
30 min at room temperature. The reaction mixture was poured into 5%
aqueous citric acid solution and extracted twice with CHCl3. The
combined organic layers were washed with brine, dried over sodium
sulfate, and filtered. The organic layer was concentrated to give the
residue. The obtained residue was then purified by flash column
chromatography on silica gel (100:0 to 95:5 CHCl3/methanol) to give
15 (37 mg, 36%) as an amorphous. 1H NMR (300 MHz, DMSO‑d6):δ
0.95–1.20 (m, 4H), 1.94–2.07 (m, 1H), 2.28 (s, 3H), 4.34 (s, 2H), 4.80
(d, J = 12.7 Hz, 1H), 5.54 (d, J = 12.7 Hz, 1H), 6.50–6.73 (m, 2H),
7.00–7.30 (m, 5H), 7.45 (d, J = 9.8 Hz, 1H), 7.67 (s, 1H). The proton of
oxadiazolone was not observed. LC/MS (ESI, [M + H]+, m/z) 529.
HPLC: purity 96%, RT 3.97 min.
4.3. (E)-3-(1-(8-((2-cyclopropyl-7-fluoroimidazo[1,2-a]pyridin-3-yl)
methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-
oxadiazol-5(4H)-one (18)
4.2.9. (E)-3-(1-(8-((7-chloro-2-cyclopropylimidazo[1,2-a]pyridin-3-yl)
methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-
oxadiazol-5(4H)-one (16)
To a stirred solution of 34 (83 mg, 0.18 mmol) in ethanol (2 mL) was
added NH2OH solution 50 wt% in water (35 μL, 5.3 mmol), and the
solution was refluxed for 1 day. After the consumption of the starting
material, the reaction mixture was poured into water and extracted
twice with CHCl3. The organic layer was concentrated to dryness and
the obtained residue was dissolved in CH2Cl2 (2 mL). To a stirred so-
lution was added triethylamine (49 μL, 0.35 mmol) and ethyl chlor-
oformate (34 μL, 0.35 mmol), and the solution was stirred for 1 h at
room temperature. The reaction mixture was poured into saturated
aqueous sodium bicarbonate solution and extracted twice with ethyl
acetate. The combined organic layers were washed with brine, dried
over sodium sulfate, and filtered. The organic layer was concentrated to
give the residue. To a solution of the residue in toluene (1 mL)/THF
(1 mL) was added tert-BuOK (40 mg, 0.35 mmol) and the solution was
stirred for 30 min at room temperature. The reaction mixture was
poured into 5% aqueous citric acid solution and extracted twice with
CHCl3. The combined organic layers were washed with brine, dried
over sodium sulfate, and filtered. The organic layer was concentrated to
give the residue. The obtained residue was then purified by flash
column chromatography on silica gel (100:0 to 95:5 CHCl3/methanol)
to give 16 (42 mg, 46%) as an amorphous. 1H NMR (300 MHz,
DMSO‑d6):δ 0.82–0.99 (m, 4H), 2.09–2.20 (m, 1H), 2.15 (s, 3H), 4.38
To a stirred solution of 36 (3.9 g, 8.9 mmol) in DMSO (43 mL) was
added NH2OH solution 50 wt% in water (26.0 mL, 430 mmol), and the
solution was refluxed for 1 day. After the consumption of the starting
material, the reaction mixture was poured into water and the resulting
suspension was filtered. The obtained residue was dissolved in CH2Cl2
(43 mL). To
a stirred solution was added triethylamine (2.3 mL,
17 mmol) and ethyl chloroformate (1.6 mL, 17 mmol), and the solution
was stirred for 1 h at room temperature. The reaction mixture was
poured into saturated aqueous sodium bicarbonate solution and ex-
tracted twice with ethyl acetate. The combined organic layers were
washed with brine, dried over sodium sulfate, and filtered. The organic
layer was concentrated to give the residue. To a solution of the residue
in toluene (15 mL)/THF (15 mL) was added tert-BuOK (1.9 g, 17 mmol)
and the solution was stirred for 30 min at room temperature. The re-
action mixture was poured into 5% aqueous citric acid solution and the
suspension was filtered. The obtained residue was recrystallized in
ethanol (20 mL) to give 18 (1.7 g, 37%) as an amorphous. 1H NMR
(300 MHz, CDCl3):δ 0.91–1.18 (m, 4H), 1.89–2.08 (m, 1H), 2.29 (s,
3H), 4.26–4.50 (m, 2H), 4.78 (d, J = 12.8 Hz, 1H), 5.53 (d,
J = 12.8 Hz, 1H), 6.39–6.77 (m, 3H), 7.00–7.30 (m, 5H), 7.44–7.59 (m,
1H). The proton of oxadiazolone was not observed. LC/MS (ESI,
[M + H]+, m/z) 513. HPLC: purity 98%, RT 3.85 min.
12

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
4.3.1. (E)-3-(1-(8-((7-chloro-2-(methoxymethyl)imidazo[1,2-a]pyridin-3-
yl)methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-
oxadiazol-5(4H)-one (19)
4.3.3. (E)-3-(1-(8-((2-cyclopropyl-8,9-dihydrofuro[3,2-c]imidazo[1,2-a]
pyridin-3-yl)methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-
1,2,4-oxadiazol-5(4H)-one (21)
To a stirred solution of 39 (72 mg, 0.15 mmol) in ethanol (10 mL)
was added NH2OH solution 50%wt in water (0.467 mL, 7.63 mmol),
and the solution was refluxed for 20 h. After the consumption of
starting material, the reaction mixture was poured into water and ex-
tracted twice with CHCl3. The combined organic layers were washed
with brine, dried over sodium sulfate, and filtered. The organic layer
was concentrated to dryness and the obtained residue was dissolved in
CH2Cl2 (2 mL). To a stirred solution was added Et3N (43 μL, 0.31 mmol)
and ethyl chloroformate (29 μL, 0.31 mmol), and the solution was
stirred for 1 h at room temperature. The reaction mixture was poured
into water and extracted twice with CHCl3. The combined organic
layers were washed with brine, dried over sodium sulfate, and filtered.
The organic layer was concentrated to give the residue. To a solution of
the residue in toluene (1 mL)/THF (1 mL) was added tert-BuOK (34 mg,
0.31 mmol) and the solution was stirred for 5 min at room temperature.
The reaction mixture was poured into 5% aqueous citric acid solution
and extracted twice with CHCl3. The combined layers were washed
with brine, dried over sodium sulfate, and filtered. The organic layer
was concentrated to give the residue. The obtained residue was then
purified by reverse phase chromatography (80:20 to 10:90 0.05% tri-
fluoroacetic acid in water/acetonitrile) to give 19 (8.2 mg, 18%) as an
amorphous. 1H NMR (300 MHz, DMSO‑d6):δ 2.15 (s, 3H), 3.27 (s, 3H),
4.39 (s, 2H), 4.56 (s, 2H), 4.87 (d, J = 12.5 Hz, 1H), 5.48 (d,
J = 12.5 Hz, 1H), 6.65 (dd, J = 2.6, 10.8 Hz, 1H), 6.79 (d t, J = 2.7,
8.3 Hz, 1H), 6.81–6.96 (m, 1H), 7.18 (d, J = 7.9 Hz, 1H), 7.22 (d,
J = 6.9 Hz, 1H), 7.25 (d, J = 6.9 Hz, 1H), 7.27 (s, 1H), 7.70 (s, 1H),
8.16 (d, J = 7.6 Hz, 1H), 12.08 (br s, 1H). LC/MS (ESI, [M + H]+, m/z)
533. HPLC: purity 98%, RT 3.77 min.
To a stirred solution of 41 (377 mg, 0.789 mmol) in DMSO (4 mL)
was added NH2OH solution 50%wt in water (1.45 mL, 23.7 mmol), and
the solution was refluxed for 6 h. After the consumption of starting
material, the reaction mixture was poured into water and extracted
twice with CHCl3. The combined organic layers were washed with
brine, dried over sodium sulfate, and filtered. The organic layer was
concentrated to dryness and the obtained residue was dissolved in
CH2Cl2 (4 mL). To
a stirred solution was added Et3N (218 μL,
1.56 mmol) and ethyl chloroformate (150 μL, 1.56 mmol), and the so-
lution was stirred for 1 h at room temperature. The reaction mixture
was poured into water and extracted twice with CHCl3. The combined
organic layers were washed with brine, dried over sodium sulfate, and
filtered. The organic layer was concentrated to give the residue. To a
solution of the residue in toluene (1.7 mL)/THF (1.7 mL) was added tert-
BuOK (154 mg, 137 mmol) and the solution was stirred for 40 min at
room temperature. The reaction mixture was poured into 5% aqueous
citric acid solution and extracted twice with CHCl3. The combined
layers were washed with brine, dried over sodium sulfate, and filtered.
The organic layer was concentrated to give the residue. The obtained
residue was then purified by reverse phase chromatography (80:20 to
10:90 0.05% trifluoroacetic acid in water/acetonitrile) to give 21
(90 mg, 21%) as an amorphous. 1H NMR (300 MHz, CDCl3):δ 0.93–1.02
(m, 2H), 1.03–1.10 (m, 2H), 1.93–2.07 (m, 1H), 2.27 (s, 3H), 3.39 (t,
J = 9.2 Hz, 2H), 4.25–4.39 (m, 2H), 4.60–4.72 (m, 3H), 5.46 (d,
J = 12.5 Hz, 1H), 6.46 (d, J = 7.3 Hz, 1H), 6.52 (dd, J = 10.3, 2.6 Hz,
1H), 6.59–6.68 (m, 1H), 7.04–7.21 (m, 4H), 7.39 (d, J = 7.3 Hz, 1H).
The proton of oxadiazolone was not observed. LC/MS (ESI, [M + H]+
,
m/z) 537. HPLC: purity 99%, RT 4.13 min.
4.3.4. (E)-2-(2-fluoro-8-((2-propyl-1H-benzo[d]imidazol-1-yl)methyl)
dibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (25)
4.3.2. (E)-3-(1-(8-((8-chloro-2-(methoxymethyl)imidazo[1,2-a]pyridin-3-
yl)methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)ethyl)-1,2,4-
oxadiazol-5(4H)-one (20)
To a stirred solution of 22 (65 mg, 0.18 mmol) and 2-propylbenzo
[d]imidazole (32 mg, 0.20 mmol) in DMF (1.5 mL) was added po-
tassium carbonate (125 mg, 0.908 mmol), and the solution was stirred
for 4 h at 60 °C. After the consumption of the starting material, the
reaction mixture was poured into water and extracted twice with ethyl
acetate. The combined organic layers were washed with brine, dried
over sodium sulfate, and filtered. The organic layer was concentrated to
give the residue. The obtained residue was then purified by flash
column chromatography on silica gel (100:0 to 50:50 hexane/ethyl
acetate) to give 25 (63 mg, 79%) as an amorphous. 1H NMR (300 MHz,
CDCl3): δ 0.98 (t, J = 7.4 Hz, 3H), 1.71–1.93 (m, 2H), 2.23 (s, 3H),
2.74–2.91 (m, 2H), 4.73 (d, J = 12.4 Hz, 1H), 5.23–5.38 (m, 3H),
6.72–6.85 (m, 2H), 6.90–7.18 (m, 7H), 7.39–7.47 (m, 1H). LC/MS (ESI,
[M + H]+, m/z) 438.
To a stirred solution of 40 (100 mg, 0.211 mmol) in ethanol
(1.0 mL) was added NH2OH solution 50%wt in water (0.388 mL,
6.33 mmol), and the solution was refluxed overnight. After the con-
sumption of starting material, the reaction mixture was poured into
water and extracted twice with CHCl3. The combined organic layers
were washed with brine, dried over sodium sulfate, and filtered. The
organic layer was concentrated to dryness and the obtained residue was
dissolved in 1,4-dioxane (1.0 mL). To a stirred solution was added DBU
(48 μL, 0.32 mmol) and 1,1′‑carbonyldiimidazole (38 mg, 0.23 mmol)
and the solution was stirred for 1 h at 110 °C. The reaction mixture was
poured into 5% aqueous citric acid solution and extracted twice with
CHCl3. The combined organic layers were washed with brine, dried
over sodium sulfate, and filtered. The organic layer was concentrated to
give the residue. To a solution of the residue in toluene (1 mL)/THF
(1 mL) was added tert-BuOK (34 mg, 0.31 mmol) and the solution was
stirred for 5 min at room temperature. The reaction mixture was poured
into 5% aqueous citric acid solution and extracted twice with CHCl3.
The combined layers were washed with brine, dried over sodium sul-
fate, and filtered. The organic layer was concentrated to give the re-
sidue. The obtained residue was then purified by reverse phase chro-
matography (80:20 to 20:80 0.05% trifluoroacetic acid in water/
acetonitrile) to give 20 (47 mg, 41%) as an amorphous. 1H NMR
(300 MHz, DMSO‑d6):δ 2.52 (s, 3H), 3.28 (s, 3H), 4.38–4.45 (m, 2H),
4.62 (s, 2H), 4.88 (d, J = 12.4 Hz, 1H), 5.48 (d, J = 12.4 Hz, 1H),
6.59–6.68 (m, 1H), 6.75–6.91 (m, 2H), 6.97–7.07 (m, 1H), 7.16–7.27
(m, 2H), 7.27–7.33 (m, 1H), 7.40–7.48 (m, 1H), 8.08–8.17 (m, 1H),
12.11 (br s, 1H). LC/MS (ESI, [M + H]+, m/z) 533. HPLC: purity 98%,
RT 3.77 min.
4.3.5. (E)-2-(3-fluoro-8-((2-propyl-1H-benzo[d]imidazol-1-yl)methyl)
dibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (26)
To a stirred solution of 23 (117 mg, 0.330 mmol) and 2-propylbenzo
[d]imidazole (50 mg, 0.31 mmol) in DMF (1.8 mL) was added po-
tassium carbonate (216 mg, 1.56 mmol), and the solution was stirred
over night at room temparature. After the consumption of the starting
material, the reaction mixture was poured into water and extracted
twice with ethyl acetate. The combined organic layers were washed
with brine, dried over sodium sulfate, and filtered. The organic layer
was concentrated to give the residue. The obtained residue was then
purified by flash column chromatography on silica gel (100:0 to 50:50
hexane/ethyl acetate) to give 26 (115 mg, 84%) as an amorphous. 1H
NMR (300 MHz, CDCl3): δ 1.02 (t, J = 7.4 Hz, 3H), 1.77–1.97 (m, 2H),
2.23 (s, 3H), 2.80 (t, J = 7.4 Hz, 2H), 4.73 (d, J = 12.6 Hz, 1H),
5.32–5.48 (m, 3H), 6.51–6.58 (m, 1H), 6.59–6.69 (m, 1H), 6.94–7.07
13

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
(m, 3H), 7.11–7.29 (m, 3H), 7.43 (d, J = 7.8 Hz, 1H), 7.78 (d,
J = 7.5 Hz, 1H). LC/MS (ESI, [M + H]+, m/z) 438.
and washed with water twice. The obtained residue was then purified
by flash column chromatography on silica gel (100:0 to 50:50 hexane/
ethyl acetate) to obtain 30 (166 mg, 95%) as an amorphous. 1H NMR
(300 MHz, CDCl3): δ 1.00–1.09 (m, 2H), 1.18–1.26 (m, 2H), 1.73–1.84
(m, 1H), 2.25 (s, 3H), 4.80 (d, J = 12.8 Hz, 1H), 5.19–5.35 (m, 2H),
5.43 (d, J = 12.8 Hz, 1H), 6.29 (d, J = 2.6 Hz, 1H), 6.57 (dd, J = 9.9,
2.6 Hz, 1H), 6.61–6.69 (m, 1H), 6.94 (d, J = 2.6 Hz, 1H), 7.04 (d,
J = 8.8, 6.6 Hz, 1H), 7.16–7.19 (m, 1H), 7.27–7.32 (m, 1H), 7.47 (d,
J = 8.1 Hz, 1H). LC/MS (ESI, [M + H]+, m/z) 442.
4.3.6. (E)-2-(4-fluoro-8-((2-propyl-1H-benzo[d]imidazol-1-yl)methyl)
dibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (27)
To a stirred solution of 24 (141 mg, 0.393 mmol) and 2-propylbenzo
[d]imidazole (60 mg, 0.37 mmol) in DMF (2.2 mL) was added po-
tassium carbonate (258 mg, 1.87 mmol), and the solution was stirred
for 4 h at 60 °C. After the consumption of the starting material, the
reaction mixture was poured into water and extracted twice with ethyl
acetate. The combined organic layers were washed with brine, dried
over sodium sulfate, and filtered. The organic layer was concentrated to
give the residue. The obtained residue was then purified by flash
column chromatography on silica gel (100:0 to 50:50 hexane/ethyl
acetate) to give 27 (164 mg, 100%) as an amorphous. 1H NMR
(300 MHz, CDCl3): δ 0.99 (t, J = 7.4 Hz, 3H), 1.82–1.95 (m, 2H), 2.24
(s, 3H), 2.80 (t, J = 7.4 Hz, 2H), 4.91 (d, J = 12.4 Hz, 1H), 5.30–5.40
(m, 2H), 5.46 (d, J = 12.4 Hz, 1H), 6.79–6.89 (m, 2H), 6.98–7.31 (m,
6H), 7.40–7.50 (m, 1H), 7.72–7.82 (m, 1H). LC/MS (ESI, [M + H]+, m/
z) 438.
4.3.10. (E)-2-(8-((2-cyclopropylimidazo[1,2-a]pyridin-3-yl)methyl)-3-
fluorodibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (32)
To a stirred solution of 2-aminopyridine (136 mg, 1.44 mmol) in
toluene (5 mL) was added copper (I) chloride (10 mg, 0.10 mmol),
copper (II) trifluoromethanesulfonate (36 mg, 0.10 mmol), cyclopro-
panecarbaldehyde (0.32 mL, 4.4 mmol) and 2-(8-ethynyl-3-fluor-
odibenzo[b,e]oxepin-11(6H)-yliden)propanenitrile
31
(500 mg,
1.73 mmol), and the solution was stirred for 5 h at 120 °C. The reaction
mixture was filtered with celite pad. The filtrate was poured into water
and extracted twice with ethyl acetate. The combined organic layer was
washed with brine, dried over sodium sulfate, and filtered. The organic
layer was concentrated to give the residue. The obtained residue was
then purified by flash column chromatography on silica gel (100:0 to
34:66 hexane/ethyl acetate) to obtain 32 (205 mg, 32%) as an amor-
phous. 1H NMR (300 MHz, CDCl3): 0.92–1.08 (m, 2H), 1.08–1.18 (m,
2H), 1.91–2.09 (m, 1H), 2.23 (s, 3H), 4.24–4.51 (m, 2H), 4.76 (d,
J = 12.8 Hz, 1H), 5.41 (d, J = 12.8 Hz, 1H), 6.50–6.70 (m, 3H),
6.94–7.15 (m, 3H), 7.21–7.30 (m, 1H), 7.35–7.46 (m, 1H), 7.49–7.58
(m, 1H), 7.58–7.66 (m, 1H). LC/MS (ESI, [M + H]+, m/z) 436.
4.3.7. (E)-2-(8-((7-chloro-2-cyclopropyl-3H-imidazo[4,5-b]pyridin-3-yl)
methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (28)
To a stirred solution of 23 (250 mg, 0.698 mmol) and 7-chloro-2-
cyclopropylimidazo[4,5-b]pyridine (135 mg, 0.698 mmol) in DMF
(3.5 mL) was added potassium carbonate (480 mg, 3.49 mmol), and the
solution was stirred overnight at room temparature. After the con-
sumption of the starting material, the reaction mixture was poured into
water and extracted twice with ethyl acetate. The combined organic
layers were washed with brine, dried over sodium sulfate, and filtered.
The organic layer was concentrated to give the residue. The obtained
residue was then purified by flash column chromatography on silica gel
(100:0 to 50:50 hexane/ethyl acetate) to give 28 (177 mg, 54%) as an
4.3.11. (E)-2-(8-((6-chloro-2-cyclopropylimidazo[1,2-a]pyridin-3-yl)
methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (33)
To
a
stirred solution of 5-chloropyridin-2-amine (178 mg,
amorphous. 1H NMR (300 MHz, CDCl3):
δ
1.02–1.16 (m, 2H),
1.38 mmol) in toluene (7 mL) was added copper (I) chloride (27 mg,
0.28 mmol), copper (II) trifluoromethanesulfonate (100 mg,
1.25–1.38 (m, 2H), 1.83–1.97 (m, 1H), 2.26 (s, 3H), 4.76 (d,
J = 12.8 Hz, 1H), 5.41 (d, J = 12.8 Hz, 1H), 5.59–5.63 (m, 2H),
6.51–6.59 (m, 1H), 6.59–6.70 (m, 1H), 6.97–7.08 (m, 1H), 7.13–7.35
0.277 mmol), cyclopropanecarbaldehyde (0.156 mL, 2.07 mmol) and 2-
(8-ethynyl-3-fluorodibenzo[b,e]oxepin-11(6H)-yliden)propanenitrile
31 (400 mg, 1.38 mmol), and the solution was irradiated for 4 h with
the reaction temperature controlled around 120 °C. The reaction mix-
ture was filtered with a celite pad. The filtrate was poured into satu-
rated ammonium chloride solution and extracted twice with ethyl
acetate. The combined organic layer was washed with brine, dried over
sodium sulfate, and filtered. The organic layer was concentrated to give
the residue. The obtained residue was then purified by flash column
chromatography on silica gel (90:10 to 40:60 hexane/ethyl acetate) to
obtain 33 (88 mg, 14%) as an amorphous. 1H NMR (300 MHz, CDCl3): δ
0.99–1.04 (m, 2H), 1.09–1.15 (m, 2H), 1.95–2.02 (m, 1H), 2.24 (s, 3H),
4.30–4.35 (m, 2H), 4.78 (d, J = 12.7 Hz, 1H), 5.43 (d, J = 12.7 Hz,
1H), 6.54–6.58 (m, 1H), 6.62–6.67 (m, 1H), 7.01–7.07 (m, 2H), 7.11
(br s, 1H), 7.23 (d, J = 7.8 Hz, 1H), 7.42 (d, J = 7.8 Hz, 1H), 7.46 (d,
J = 9.8 Hz, 1H), 7.67 (d, J = 2.0 Hz, 1H). LC/MS (ESI, [M + H]+, m/z)
470.
(m, 3H), 7.38–7.48 (m, 1H), 8.11–8.24 (m, 1H).LC/MS (ESI, [M + H]+
,
m/z) 471.
4.3.8. (E)-2-(8-((4-chloro-2-cyclopropyl-1H-imidazo[4,5-c]pyridin-1-yl)
methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (29)
To a stirred solution of 23 (120 mg, 0.335 mmol) and 4-chloro-2-
cyclopropylimidazo[4,5-c]pyridine (66 mg, 0.34 mmol) in DMF
(2.0 mL) was added potassium carbonate (231 mg, 1.68 mmol), and the
solution was stirred overnight at room temparature. After the con-
sumption of the starting material, the reaction mixture was poured into
water and extracted twice with ethyl acetate. The combined organic
layers were washed with brine, dried over sodium sulfate, and filtered.
The organic layer was concentrated to give the residue. The obtained
residue was then purified by flash column chromatography on silica gel
(100:0 to 44:66 hexane/ethyl acetate) to give 29 (98 mg, 62%) as an
amorphous. 1H NMR (300 MHz, CDCl3):
δ 1.04–1.17 (m, 2H),
1.27–1.38 (m, 2H), 1.83–1.96 (m, 1H), 2.25 (s, 3H), 4.71–4.81 (m, 1H),
5.32–5.53 (m, 3H), 6.53–6.73 (m, 2H), 6.95–7.11 (m, 3H), 7.15–7.22
(m, 1H), 7.42–7.53 (m, 1H), 8.09–8.15 (m, 1H). LC/MS (ESI,
[M + H]+, m/z) 471.
4.3.12. (E)-2-(8-((7-chloro-2-cyclopropylimidazo[1,2-a]pyridin-3-yl)
methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (34)
To
a
stirred solution of 4-chloropyridin-2-amine (156 mg,
1.21 mmol) in toluene (6 mL) was added copper (I) chloride (24 mg,
0.24 mmol), copper (II) trifluoromethanesulfonate (88 mg, 0.24 mmol),
cyclopropanecarbaldehyde (0.136 mL, 1.82 mmol) and 2-(8-ethynyl-3-
fluorodibenzo[b,e]oxepin-11(6H)-yliden)propanenitrile 31 (350 mg,
1.21 mmol), and the solution was stirred overnight at 120 °C. The re-
action mixture was filtered with celite pad. The filtrate was poured into
saturated ammonium chloride solution and extracted twice with ethyl
acetate. The combined organic layer was washed with brine, dried over
sodium sulfate, and filtered. The organic layer was concentrated to give
the residue. The obtained residue was then purified by flash column
4.3.9. (E)-2-(8-((2-cyclopropyl-1H-thieno[3,4-d]imidazol-1-yl)methyl)-3-
fluorodibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (30)
To a stirred solution of 2-cyclopropyl-1H-thieno[3,4-d]imidazole
(65 mg, 0.40 mmol) in DMF (2.3 mL) was added potassium carbonate
(274 mg, 1.98 mmol) and (E)-2-(8-(bromomethyl)-3-fluorodibenzo[b,e]
oxepin-11(6H)-yliden)propanenitrile 23 (145 mg, 0.404 mmol), and the
solution was stirred overnight at room temperature. The reaction
mixture was poured into water and the resulting suspension was filtered
14

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
chromatography on silica gel (90:10 to 40:60 hexane/ethyl acetate) to
obtain 34 (73 mg, 13%) as an amorphous. 1H NMR (300 MHz, CDCl3):δ
0.98–1.05 (m, 2H), 1.08–1.15 (m, 2H), 1.94–2.03 (m, 1H), 2.23 (s, 3H),
4.30 (d, J = 17.2 Hz, 1H), 4.38 (d, J = 17.2 Hz, 1H), 4.76 (d,
J = 12.8 Hz, 1H), 5.41 (d, J = 12.8 Hz, 1H), 6.53–6.59 (m, 1H),
6.61–6.68 (m, 2H),6.99–7.06 (m, 1H), 7.08–7.10 (m, 1H), 7.20–7.24
(m, 1H), 7.41 (d, J = 7.7 Hz, 1H), 7.50–7.54 (m, 2H).
mixture was added the obtained residue (400 mg) and the mixture was
stirred overnight at room temperature. The reaction mixture was
poured into saturated aqueous ammonium chloride solution/saturated
aqueous sodium thiosulfate solution and extracted twice with ethyl
acetate. The combined organic layers were washed with brine, dried
over sodium sulfate, and filtered. The organic layer was concentrated to
give the residue. The obtained residue was then purified by flash
column chromatography on silica gel (100:0 to 70:30 CHCl3/methanol)
to give 38 (72 mg, 19%) as an amorphous. 1H NMR (300 MHz, CDCl3):δ
2.23 (s, 3H), 3.47 (s, 3H), 4.35–4.37 (m, 1H), 4.63–4.76 (m, 4H),
5.38–5.42 (m, 1H), 6.53–6.71 (m, 3H), 7.00–7.10 (m, 2H), 7.20–7.24
(m, 2H), 7.39–7.42 (m, 1H), 7.58–7.60 (m, 1H). LC/MS (ESI,
[M + H]+, m/z) 474.
LC/MS (ESI, [M + H]+, m/z) 470.
4.3.13. (E)-2-(8-((8-chloro-2-cyclopropylimidazo[1,2-a]pyridin-3-yl)
methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (35)
To
a
stirred solution of 3-chloropyridin-2-amine (101 mg,
0.786 mmol) in toluene (4 mL) was added copper (I) chloride (5.4 mg,
0.055 mmol), copper (II) trifluoromethanesulfonate (20 mg,
0.055 mmol), cyclopropanecarbaldehyde (88 μL, 1.2 mmol) amd 2-(8-
ethynyl-3-fluorodibenzo[b,e]oxepin-11(6H)-yliden)propanenitrile 31
(250 mg, 0.864 mmol), and the solution was stirred for 30 h at 120 °C.
The reaction mixture was filtered with celite pad. The filtrate was
poured into saturated ammonium chloride solution and extracted twice
with ethyl acetate. The combined organic layer was washed with brine,
dried over sodium sulfate, and filtered. The organic layer was con-
centrated to give the residue. The obtained residue was then purified by
flash column chromatography on silica gel (100:0 to 40:60 hexane/
ethyl acetate) to obtain 35 (61 mg, 15%) as an amorphous. 1H NMR
(300 MHz, CDCl3):δ 0.92–1.09 (m, 2H), 1.09–1.28 (m, 2H), 1.93–2.10
(m, 1H), 2.23 (s, 3H), 4.24–4.52 (m, 2H), 4.75 (d, J = 12.6 Hz, 1H),
5.40 (d, J = 12.6 Hz, 1H), 6.45–6.73 (m, 3H), 6.95–7.31 (m, 4H), 7.40
4.3.16. (E)-2-(8-((8-chloro-2-(methoxymethyl)imidazo[1,2-a]pyridin-3-
yl)methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (39)
To a stirred solution of 8-chloro-3-iodo-2-(methoxymethyl)imidazo
[1,2-a]pyridine (500 mg, 1.55 mmol) in THF (2.0 mL) was added iso-
propylmagnesium chloride (2 mol/L solution in THF, 0.775 mL,
1.55 mmol) and the solution was stirred for 1 h at −50 °C. To the re-
action mixture was added 37 (227 mg, 0.775 mmol) and stirred for 1 h
at −20 °C. After the consumption of the starting material, the reaction
mixture was poured into saturated aqueous ammonium chloride solu-
tion and extracted twice with CHCl3. The combined organic layers were
washed with brine, dried over sodium sulfate, and filtered. The organic
layer was concentrated to give the residue. The obtained residue was
then purified by flash column chromatography on silica gel (100:0 to
70:30 CHCl3/methanol) to give the residue (180 mg). To a stirred so-
lution of chlorotrimethylsilane (0.471 mL, 3.71 mmol) in hexane
(0.4 mL)/acetonitrile (0.4 mL) was added sodium iodide (556 mg,
3.71 mmol) and the mixture was stirred overnight at room temperature.
To the reaction mixture was added the residue (180 mg) and the mix-
ture was stirred overnight at room temperature. The reaction mixture
was poured into saturated aqueous ammonium chloride solution/satu-
rated aqueous sodium thiosulfate solution and extracted twice with
ethyl acetate. The combined organic layers were washed with brine,
dried over sodium sulfate, and filtered. The organic layer was con-
centrated to give the residue. The obtained residue was then purified by
flash column chromatography on silica gel (100:0 to 70:30 CHCl3/
methanol) to give 39 (77 mg, 44%) as an amorphous. 1H NMR
(300 MHz, CDCl3):δ 2.23 (s, 3H), 3.47 (s, 3H), 4.33–4.46 (m, 2H), 4.76
(s, 3H), 5.39 (d, J = 12.7 Hz, 1H), 6.53–6.57 (m, 1H), 6.62–6.68 (m,
2H), 7.00–7.04 (m, 1H), 7.10–7.11 (m, 1H), 7.23–7.30 (m, 2H),
7.39–7.41 (m, 1H), 7.62–7.64 (m, 1H). LC/MS (ESI, [M + H]+, m/z)
474.
(d, J = 7.8 Hz, 1H), 7.55 (d, J = 7.8 Hz, 1H). LC/MS (ESI, [M + H]+
m/z) 470.
,
4.3.14. (E)-2-(8-((2-cyclopropyl-7-fluoroimidazo[1,2-a]pyridin-3-yl)
methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (36)
To
a
stirred solution of 4-fluoropyridin-2-amine (141 mg,
1.23 mmol) in toluene (5 mL) was added copper (I) chloride (11 mg,
0.11 mmol), copper (II) trifluoromethanesulfonate (41 mg, 1.1 mmol),
cyclopropanecarbaldehyde (13 μL, 1.7 mmol) and 2-(8-ethynyl-3-
fluorodibenzo[b,e]oxepin-11(6H)-yliden)propanenitrile 31 (330 mg,
0.864 mmol), and the solution was stirred for 8 h at 120 °C. The reaction
mixture was filtered with celite pad. The filtrate was poured into water
and extracted twice with ethyl acetate. The combined organic layer was
washed with brine, dried over sodium sulfate, and filtered. The organic
layer was concentrated to give the residue. The obtained residue was
then purified by flash column chromatography on silica gel (100:0 to
50:50 hexane/ethyl acetate) to obtain 36 (25 mg, 4.3%) as an amor-
phous. 1H NMR (300 MHz, CDCl3):δ 0.90–1.18 (m, 4H), 1.92–2.07 (m,
1H), 2.23 (s, 3H), 4.21–4.44 (m, 2H), 4.75 (d, J = 12.8 Hz, 1H), 5.43
(d, J = 12.8 Hz, 1H), 6.43–6.74 (m, 3H), 6.99–7.31 (m, 4H), 7.33–7.62
(m, 2H). LC/MS (ESI, [M + H]+, m/z) 454.
4.3.17. (E)-2-(8-((2-cyclopropyl-8,9-dihydrofuro[3,2-c]imidazo[1,2-a]
pyridin-3-yl)methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)
propanenitrile (40)
4.3.15. (E)-2-(8-((7-chloro-2-(methoxymethyl)imidazo[1,2-a]pyridin-3-
yl)methyl)-3-fluorodibenzo[b,e]oxepin-11(6H)-ylidene)propanenitrile (38)
To a stirred solution of 7-chloro-3-iodo-2-(methoxymethyl)imidazo
[1,2-a]pyridine (700 mg, 2.17 mmol) in THF (4.5 mL) was added iso-
propylmagnesium chloride (2 mol/L solution in THF, 1.1 mL, 2.0 mmol)
and the solution was stirred for 1 h at −50 °C. To the reaction mixture
was added 37 (318 mg, 1.09 mmol) and stirred for 1 h at −20 °C. After
the consumption of the starting material, the reaction mixture was
poured into saturated aqueous ammonium chloride solution and ex-
tracted twice with CHCl3. The combined organic layers were washed
with brine, dried over sodium sulfate, and filtered. The organic layer
was concentrated to give the residue. The obtained residue was then
purified by flash column chromatography on silica gel (100:0 to 70:30
CHCl3/methanol) to give the residue (400 mg). To a stirred solution of
chlorotrimethylsilane (1.04 mL, 8.16 mmol) in hexane (0.8 mL)/acet-
onitrile (0.4 mL) was added sodium iodide (1.2 g, 8.2 mmol) and the
mixture was stirred overnight at room temperature. To the reaction
To a stirred solution of 2-cyclopropyl-3-iodo-8,9-dihydrofuro[3,2-c]
imidazo[1,2-a]pyridine (435 mg, 1.33 mmol) in THF (4.5 mL) was
added isopropylmagnesium chloride lithium chloride complex
(1.3 mol/L solution in THF, 1.15 mL, 1.50 mmol) and the solution was
stirred for 1 h at −40 °C. To the reaction mixture was added 37
(244 mg, 0.843 mmol) and stirred for 1 h at −20 °C. After the con-
sumption of the starting material, the reaction mixture was poured into
saturated aqueous ammonium chloride solution and extracted twice
with CHCl3. The combined organic layers were washed with brine,
dried over sodium sulfate, and filtered. The organic layer was con-
centrated to give the residue. The obtained residue was then re-
crystallized from isopropyl alcohol (20 mL) to afford a white solid
(362 mg). To a stirred solution of the resulting residue (362 mg,
0.729 mmol) in TFA (3.6 mL) was added triethylsilane (0.583 mL,
3.65 mmol) and the mixture was stirred for 1 h at 60 °C. The reaction
mixture was poured into saturated aqueous sodium bicarbonate solu-
tion and extracted twice with CHCl3. The combined organic layers were
15

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
washed with brine, dried over sodium sulfate, and filtered. The organic
layer was concentrated to give the residue. The obtained residue was
then purified by flash column chromatography on silica gel (100:0 to
90:10 CHCl3/methanol) to give 40 (386 mg, 87%) as an amorphous.1H
NMR (300 MHz, CDCl3): δ 0.91–1.01 (m, 2H), 1.03–1.11 (m, 2H),
1.95–2.03 (m, 1H), 2.23 (s, 3H), 3.48 (t, J = 8.8 Hz, 2H), 4.25–4.38 (m,
2H), 4.69 (t, J = 8.8 Hz, 2H), 4.75 (d, J = 12.7 Hz, 1H), 5.40 (d,
J = 12.7 Hz, 1H), 6.41 (d, J = 6.8 Hz, 1H), 6.55 (dd, J = 10.2, 2.4 Hz,
1H), 6.60–6.67 (m, 1H), 6.99–7.05 (m, 1H), 7.07–7.11 (m, 1H),
7.21–7.25 (m, 1H), 7.37–7.43 (m, 2H).
water. Male severe combined immunodeficient (SCID) mice were pur-
chased from CLEA Japan, Inc., and used at eight weeks of age. All an-
imal experiments were performed in compliance with protocols ap-
proved by the Institutional Animal Care and Use Committee of Kyowa
Hakko Kirin.
4.4.4. Pharmacokinetics studies
Each compound was orally administered to mice. Blood samples in
mice were collected from the tail vein at appropriate time-points and
centrifuged to obtain plasma. Diluted plasma samples were extracted
using protein precipitation with acetonitrile containing an internal
standard. The collected supernatants were subjected to LC-MS/MS
analysis using an atmospheric pressure chemical ionization interface in
positive ion mode. The compound concentrations in the plasma samples
were calculated using calibration standards curve in control plasma.
The AUC was obtained from the plasma concentration-time profiles
using noncompartmental analysis.
LC/MS (ESI, [M + H]+, m/z) 537.
4.4. Biological methods
4.4.1. Chimeric GAL4-PPARγ transactivation reporter assay
Test compounds were screened for agonist activity on PPARγ-GAL4
chimeric receptors in transiently transfected HEK293EBNA cells. The
cells were maintained in Dulbecco's Modified Eagle Medium (DMEM;
Thermo Fisher Scientific) containing 10% fetal bovine serum (Thermo
Fisher Scientific), 100 U/mL penicillin, and 100 μg/ml streptomycin
(Thermo Fisher Scientific) and incubated at 37 °C in 5% CO2. To
transfect the reporter construct into HEK293EBNA cells, cells were
seeded at 1 × 105 cells/ml in a tissue culture dish (Iwaki, Chiba,
Japan). After 24 h incubation, transfections were performed with
Superfect transfection reagent (QIAGEN) according to the instructions
of the manufacturer. In brief, pM-human PPARγ/GAL4 expression
vector and pZAC19-Luc vector were premixed and transfected into the
cells followed by 2.5 h incubation. After a further 24 h incubation with
growth medium, the transfected cells were seeded into 96-well assay
plates, and test compounds were added (1–3000 nM, n = 3 per con-
centration). The test compounds were initially dissolved in DMSO, and
then diluted in DMEM without any supplement. Steady-Glo luciferase
assay reagent (Promega) was used as a substrate, and the luciferase
activity was measured using the Microplate Scintillation and lumines-
cence counter TopCount NXT (Packard, Groningen, Netherlands). The
luciferase activity was normalized to that of pioglitazone at 1000 nM.
The maximum activation (efficacy) of pioglitazone was set at 100%.
The efficacy of each compound was calculated as the percentage of the
maximum activation obtained with pioglitazone. EC50 values were de-
termined by the concentration that was 50% of its maximum activity
using the XLFit.
4.4.5. Metabolic stability assay in human liver microsomes
An incubation mixture was prepared consisting of human liver mi-
crosomes, each compound and a potassium phosphate buffer containing
MgCl2 (pH 7.4). The reactions were initiated with the addition of
NADPH and held for 30 min at 37 °C, and then terminated by adding
ice-cold acetonitrile containing internal standards. The supernatant was
subjected to LC-MS/MS analysis. The in vitro disappearance rate con-
stant for each compound was determined by the slope of the linear
regression from log percentage remaining versus incubation time re-
lationships, and the intrinsic clearance was calculated from the elim-
ination rate constant, the microsomes concentration and liver micro-
somes amount per body weight.
4.4.6. Xenograft study
AsPC-1/AG1 cells (3 × 107) suspended in 0.1 mL PBS were in-
oculated subcutaneously into SCID mice. 17 or control (0.5% MC400)
administration into tumor-bearing SCID mice was started 7 days after
the tumor inoculation. Twenty-five mice were divided into 5 groups of
5 mice per group for 17 (12.5, 25, 50 or 100 mg/kg) or control, such
that the average tumor volumes were equivalent among the groups. The
tumor volume was monitored twice weekly and calculated by the fol-
lowing formula: tumor volume (mm3) = 0.5 × (major dia-
meter) × (minor diameter)2. Mice were treated with oral administra-
tion of 17 (12.5, 25, 50 or 100 mg/kg) or control twice a day for
14 days. The relative tumor volume in each mouse was represented as
V/V0, where V0 was the volume on Day0. The difference in the V/V0
between the control group and each treatment group was assessed by
the Steel test. Statistical analyses were performed using the SAS soft-
ware program (Release 9.1.3, SAS institute Inc.). In our study, P < .05
was considered significant.
4.4.2. MKN45 cell aggregation assay
MKN45 cells were aggregated when the cells were incubated with
PPARγ agonists. The cells were maintained in RPMI1640 medium
(Thermo Fisher Scientific) containing 10% fetal bovine serum (Thermo
Fisher Scientific), 100 U/mL penicillin, and 100 μg/ml streptomycin
(Thermo Fisher Scientific) and incubated at 37 °C in 5% CO2. Cells were
seeded at 2500 cells/well in 96-well assay plates and incubated with
test compounds, (1–1000 nM, n = 3 per concentration), for 5 days. Cell
images were captured using an IN Cell Analyzer 1000 (GE healthcare).
Nuclei were stained with Hoechst 33342 (SIGMA), and the area of
nuclei was calculated. An area exceeding 1015.58 μm2 was defined as
an aggregated cell cluster. The ratio of the area of the aggregated cell
clusters to the total cell area was calculated and set as the formation
rate of aggregation. The cell aggregation-inducing activity was nor-
malized to that of 1, with the maximum activity of 1 set as 100%. The
maximum activity of each compound was then calculated as the per-
centage of the maximum activity of 1. EC50 values were determined as
the concentration achieving 50% of the maximum activity using the
XLFit.
4.4.7. RT-qPCR
Tumor tissues were collected from AsPC-1/AG1 tumor-bearing mice
4 h after the last dose of bid ×14 oral administration. Total RNA was
extracted using RNeasy mini Kit (Catalog no. 74104; Qiagen). RNA was
converted to cDNA using SuperScript VILO cDNA synthesis kit (Cat
No.11754250; Invitrogen) according to the manufacturer's protocol.
qPCR was performed using Taqman Gene Expression Master Mix
(Catalog no. 4352042; Applied Biosysems) with cDNA and Taqman
probes (#KRT20 Hs000300643_m1 and ANGPTL4 Hs01191127_m1;
Applied Biosystems) on 7500 Fast Real-Time PCR system (Applied
Biosystems).
4.4.8. Hematocrit test
4.4.3. Animals
Male Crlj:CD1(ICR) mice (n = 10 per group) 7 to 8 weeks old at the
start of dosing were orally administered derivative 17 twice a day
(10 mL/kg/time) at doses of 0 (vehicle: 0.5%[w/v] methyl cellulose
400), 50 and 200 mg/kg for 4 days. On the day after the dosing period,
Male BALB/c mice were purchased from Charles River Ltd. (Tokyo,
Japan). They were maintained on a constant 12-h light/dark cycle in a
temperature- and humidity-controlled room and were given food and
16

K. Yamamoto, et al.
Bioorganic&MedicinalChemistry27(2019)115122
blood samples were collected in hematocrit tubes and the hematocrit
values were determined by the micro-hematocrit method.
3. Kim H-I, Ahn Y-H. Diabetes. 2004;53:S60–S65.
4. Lehrke M, Lazar MA. Cell. 2005;123:993–999.
5. Parulkar AA, Pendergrass ML, Granda-Ayala R, Lee TR, Fonseca VA. Ann Intern Med.
2001;134:61–71.
Declaration of competing interest
6. Lehmann JM, Moore LB, Smith-Oliver TA, Wilkison WO, Willson TM, Kliewer SA. J
Biol Chem. 1995;270:12953–12956.
7. Edvardsson U, Bergström M, Alexandersson M, Bamberg K, Ljung B, Dahllöf B. J Lipid
Res. 1999;40:1177–1184.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence this work.
8. Shimazaki N, Togashi N, Hanai M, et al. Eur J Cancer. 2008;44:1734–1743.
9. Copland JA, Marlow LA, Kurakata S, et al. Oncogene. 2006;25:2304–2317.
10. Motani A, Wang Z, Weiszmann J, et al. J Mol Biol. 2009;386:1301–1311.
11. Gregoire FM, Zhang F, Clarke HJ, et al. Mol Endocrinol. 2009;23:975–988.
12. Khandekar MJ, Banks AS, Laznik-Bogoslavski D, et al. Proc Natl Acad Sci U S A.
2018;115:561–566.
Acknowledgment
13. Hughes TS, Giri PK, de Vera IM, et al. Nat Commun. 2014;5:3571.
14. Barak Y, Nelson MC, Ong ES, et al. Mol Cell. 1999;4:585–595.
15. Medina-Gomez G, Gray S, Vidal-Puig A. Public Health Nutr. 2007;10:1132–1137.
16. Iwaki M, Matsuda M, Maeda N, et al. Diabetes. 2003;52:1655–1663.
17. Tamori Y, Masugi J, Nishino N, Kasuga M. Diabetes. 2002;51:2045–2055.
18. Kliewer SA, Lenhard JM, Willson TM, Patel I, Morris DC, Lehmann JM. Cell.
1995;83:813–819.
We thank Hiromi Nagai for supporting the synthesis of the tricyclic
compounds. We thank Satomi Nagata and Yoko Nakajima for the eva-
luation of the AsPC-1 xenograft mice. We thank Kenta Nakajima for
supporting our validation of the in vitro assay data. We thank Takahiro
Kabeya for his assistance with the preparation of the correlation chart
between reporter gene assay and aggregation assay. We thank Hideyuki
Onodera, Seiji Aratake and Kuniyuki Kishikawa for supporting editing
of this manuscript. We thank Daisuke Nakashima for screening the
compounds for PAINS.
19. Hamza MS, Pott S, Vega VB, et al. PLoS One. 2009;4:e4907.
20. Smallridge RC, Copland JA, Brose MS, et al. J Clin Endocrinol Metab.
2013;98:2392–2400.
21. Serizawa M, Murakami H, Watanabe M, Takahashi T, Yamamoto N, Koh Y. Cancer
Sci. 2014;105:683–689.
22. Choi JH, Banks AS, Kamenecka TM, et al. Nature. 2011;477:477–481.
23. Yamamoto K, Tamura T, Henmi K, et al. J Med Chem. 2018;61:10067–10083.
24. Shah P, Westwell AD. J Enzyme Inhib Med Chem. 2007;22:527–540.
25. Bae H, Jang JY, Choi S-S, et al. Chem Sci. 2016;7:5523–5529.
26. Einstein M, Akiyama TE, Castriota GA, et al. Mol Pharmacol. 2008;73:62–74.
27. Xu A, Lam MC, Chan KW, et al. Proc Natl Acad Sci U S A. 2005;102:6086–6091.
28. Chan CWN, Wong NA, Liu Y, et al. Proc Natl Acad Sci U S A. 2009;106:1936–1941.
29. Chang C-S, Tsai P-J, Sung J-M, et al. Am J Pathol. 2014;184:442–453 (The hematocrit
value reduced approximately 45% to 41% in pioglitazone-treated mice at 40 mg/kg).
30. Niemeyer NV, Janney LM. Pharmacotherapy. 2002;22:924–929.
31. Chernyak N, Gevorgyan V. Angew Chem Int Ed Engl. 2010;49:2743–2746.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmc.2019.115122.
References
1. Tontonoz P, Hu E, Spiegelman BM. Cell. 1994;79:1147–1156.
2. Willson TM, Lambert MH, Kliewer SA. Annu Rev Biochem. 2001;70:341–367.
17