Optimization of a novel series of N-phenylindoline-5-sulfonamide-based acyl CoA:monoacylglycerol acyltransferase-2 inhibitors: Mitigation of CYP3A4 time-dependent inhibition and phototoxic liabilities

Article abstract of DOI:10.1016/j.bmc.2015.06.003

Acyl CoA:monoacylglycerol acyltransferase-2 (MGAT2) has emerged as a potential peripheral target for the treatment of obesity and metabolic disorders. We previously identified a novel series of N-phenylindoline-5-sulfonamide derivatives exemplified by 2 a

Full text of DOI:10.1016/j.bmc.2015.06.003

Accepted Manuscript
Optimization of a Novel Series of N-Phenylindoline-5-sulfonamide-based Acyl
CoA:Monoacylglycerol Acyltransferase-2 Inhibitors: Mitigation of CYP3A4
Time-dependent Inhibition and Phototoxic Liabilities
Kenjiro Sato, Hiroki Takahagi, Osamu Kubo, Kousuke Hidaka, Takeshi
Yoshikawa, Masahiro Kamaura, Masanori Nakakariya, Nobuyuki Amano,
Ryutaro Adachi, Toshiyuki Maki, Kazumi Take, Shiro Takekawa, Tomoyuki
Kitazaki, Tsuyoshi Maekawa
PII:
S0968-0896(15)00497-6
DOI:
http://dx.doi.org/10.1016/j.bmc.2015.06.003
BMC 12359
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
24 April 2015
1 June 2015
2 June 2015
Please cite this article as: Sato, K., Takahagi, H., Kubo, O., Hidaka, K., Yoshikawa, T., Kamaura, M., Nakakariya,
M., Amano, N., Adachi, R., Maki, T., Take, K., Takekawa, S., Kitazaki, T., Maekawa, T., Optimization of a Novel
Series of N-Phenylindoline-5-sulfonamide-based Acyl CoA:Monoacylglycerol Acyltransferase-2 Inhibitors:
Mitigation of CYP3A4 Time-dependent Inhibition and Phototoxic Liabilities, Bioorganic & Medicinal Chemistry
(
2015), doi: http://dx.doi.org/10.1016/j.bmc.2015.06.003
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Optimization of a Novel Series of N-Phenylindoline-
-sulfonamide-based Acyl CoA:Monoacylglycerol
5
Acyltransferase-2 Inhibitors: Mitigation of CYP3A4
Time-dependent Inhibition and Phototoxic
Liabilities
Kenjiro Sato,* Hiroki Takahagi, Osamu Kubo, Kousuke Hidaka, Takeshi Yoshikawa, Masahiro
Kamaura, Masanori Nakakariya, Nobuyuki Amano, Ryutaro Adachi, Toshiyuki Maki, Kazumi
Take, Shiro Takekawa, Tomoyuki Kitazaki, and Tsuyoshi Maekawa
Pharmaceutical Research Division, Takeda Pharmaceutical Co., Ltd., 26-1, Muraoka-Higashi 2-
Chome, Fujisawa, Kanagawa 251-8555, Japan
*
Corresponding author; Phone: +81 466 32 1072; Fax: +81 466 29 4475;
E-mail: kenjirou.satou@takeda.com
1

Abstract
Acyl CoA:monoacylglycerol acyltransferase-2 (MGAT2) has emerged as a potential peripheral
target for the treatment of obesity and metabolic disorders. We previously identified a novel
series of N-phenylindoline-5-sulfonamide derivatives exemplified by 2 as potent and orally
bioavailable MGAT2 inhibitors. Despite its attractive potency, further assessment revealed that
this compound exhibited time-dependent inhibition (TDI) of cytochrome P450 3A4 (CYP3A4).
To remove the undesirable CYP3A4 TDI activity, structural modification was focused on the 2,4-
difluoroaniline moiety on the basis of the assumption that this moiety would be involved in
mechanism-based inhibition of CYP3A4 via oxidative metabolism. This led to the finding that
the introduction of 4-chloro-2,6-difluoroaniline significantly improved CYP3A4 TDI risk.
Further optimization resulted in the discovery of N-(4-chloro-2,6-difluorophenyl)-1-{5-[1-
methyl-3-(trifluoromethyl)-1H-pyrazol-5-yl]pyrimidin-2-yl}-7-(2-oxopyrrolidin-1-yl)-2,3-
dihydro-1H-indole-5-sulfonamide (27c) with potent MGAT2 inhibitory activity (IC50 = 7.8 nM)
and excellent ADME-Tox profiles including metabolic stability, oral bioavailability, and
CYP3A4 TDI. In a mouse oral fat tolerance test, compound 27c effectively and dose-
dependently suppressed the elevation of plasma triacylglycerol levels after oral administration at
doses of 1 and 3 mg/kg. We also discuss mitigation of the phototoxic liability of biaryl
derivatives on the basis of the HOMO–LUMO gap hypothesis during the course of optimization
efforts.
Keywords
MGAT2 inhibitor, acyltransferase, triacylglycerol, obesity, metabolic disease, indoline, time-
dependent inhibition, CYP3A4, phototoxicity
2

Abbreviations
ACAT, acyl CoA:cholesterol acyltransferase; ADME, absorption, distribution, metabolism and
excretion; APCI, atomospheric pressure cheimcal ionization; CM, chylomicron; dba,
dibenzylideneacetone; CYP, cytochrome P450; DDI, drug–drug interaction; DGAT, acyl
CoA:diacylglycerol acyltransferase; DIEA, N,N-diisopropylethylamine; DMPK, drug
metabolism and pharmacokinetics; HOMO, highest occupied molecular orbital; HLM, human
liver microsome; LLE, ligand lipophilicity efficiency; LPL, lipoprotein lipase; LUMO, lowest
unoccupied molecular orbital; MGAT, acyl CoA:monoacylglycerol acyltransferase; MLM,
mouse liver microsome; MRT, mean residence time; OFTT, oral fat tolerance test; TEA,
triethylamine; TDI, time-dependent inhibition ; TG, triacylglycerol.
Graphical Abstract
3

1
. Introduction
Triacylglycerol (TG) constitutes a major class of neutral lipids and serves as an essential form
of energy storage in the human body. TG also plays an important role as a carrier of fatty acids to
peripheral tissues. However, excess storage of TG in adipose and nonadipose tissues causes
various pathological conditions such as obesity, insulin resistance, dyslipidemia, and hepatic
1
, 2
steatosis.
In addition, several studies have indicated that TG-rich lipoproteins are an
3
-5
independent risk factor for cardiovascular disease. Therefore, an appropriate modulation
strategy for TG metabolism is expected to restore abnormally elevated levels of TG and improve
TG-related cardiometabolic disorders.
One potential therapeutic approach is to inhibit the biosynthesis of TG. Acyl
CoA:monoacylglycerol acyltransferase-2 (MGAT2) is a member of the DGAT2 gene family of
enzymes and catalyzes the synthesis of diacylglycerol from monoacylglycerol and fatty acyl
6
, 7
CoA, which is a crucial step in TG synthesis by the MGAT pathway. This enzyme is highly
expressed in the intestine and is considered to play an important role in dietary fat absorption.
Recent research demonstrated that MGAT2 gene knockout mice were resistant to olive oil-
induced hypertriglyceridemia, high-fat induced obesity, and metabolic disorders such as insulin
8
-10
resistance, hypercholesterolemia, and fatty livers.
Interestingly, the MGAT2-deficient mice
displayed higher energy expenditure than the wild-type mice. These preclinical data highlight
that MGAT2 is critically involved in fat absorption as well as energy metabolism, suggesting that
the inhibition of MGAT2 could serve as a novel peripheral target for the treatment of obesity and
metabolic disorders. Furthermore, the facts that the MGAT2-deficient mice normally gained
weight on a low-fat diet and displayed no apparent abnormalities, including fatty stool, would
warrant the safety profile of MGAT2 inhibitors.
4

11, 12
Till date, several chemical classes of MGAT2 inhibitors have been identified;
however,
there have been no clinical studies using MGAT2 inhibitors. We previously reported the
discovery of a novel series of N-phenylindoline-5-sulfonamide derivatives as potent, selective,
1
3
and orally bioavailable MGAT2 inhibitors, as exemplified by 2 (Figure 1). Compound 2
exhibited potent MGAT2 inhibitory activity and suppressed the elevation of TG in a mouse oral
fat tolerance test (OFTT) after oral administration. These attractive profiles motivated us to
evaluate this compound in detail in terms of drug–drug interactions (DDIs) and toxicity profiles.
Unfortunately, compound 2 exhibited time-dependent inhibition (TDI) of cytochrome P450 3A4
(
CYP3A4). Because CYP3A4 is the most abundant CYP isoform and is responsible for the
oxidative metabolism of a wide variety of clinical drugs, TDI of this metabolizing enzyme may
1
4
preclude the drug development of compounds from the perspective of DDIs. Accordingly, our
subsequent optimization effort was directed toward addressing the CYP3A4 TDI issue to
produce a clinical candidate with safe profiles. In this paper, we report the design, synthesis, and
biological evaluation of N-phenylindoline-5-sulfonamide derivatives as novel MGAT2 inhibitors,
which led to the identification of compound 27c with potent MGAT2 inhibitory activity and
favorable ADME-Tox profiles as well as in vivo efficacy. We also describe an approach to
mitigate the resulting phototoxic liability in the course of our optimization campaign.
Figure 1. N-Phenylindoline-5-sulfonamide-based MGAT2 inhibitors.
5

2
. Chemistry
Scheme 1 details the synthesis of the N-(4-chloro-2,6-difluorophenyl)indoline-5-sulfonamide
1
3
derivatives 9a and 9b. As described in a previous report, the sulfonyl chloride 3 smoothly
reacted with 2,4-difluoroaniline or p-anisidine in N,N-dimethylacetamide without the addition of
a base to give the corresponding sulfonamide in good yields. However, the reaction of 3 with 4-
chloro-2,6-difluoroaniline resulted in a poor yield (10–20% yield) under the same condition,
probably because of the low basicity and nucleophilicity of the aniline. After condition screening,
we found that this sulfonamidation proceeded cleanly under refluxing in 2,2,2-trifluoroethanol as
a solvent, which afforded much better results than conditions using a base such as pyridine, N,N-
diisopropylethylamine, or sodium hydride. The slightly acidic nature of 2,2,2-trifluoroethanol
may be involved in the positive effect for this reaction because the use of ethanol or ethyl acetate
as a solvent for this type of reaction gave no or little desired product. For substrates with an acid-
labile substituent, the addition of molecular sieves 4A was effective in scavenging the generated
hydrogen chloride. After the reduction of the nitro group of 4 with zinc powder, the desired
aniline 5 was obtained in a good isolated yield. Acylation of 5 with 4-chlorobutyryl chloride
followed by intramolecular cyclization using sodium hydride furnished lactam 7. The
benzyloxycarbonyl (Cbz) group of 7 was removed by hydrolysis under a basic condition to give
8
, which was subjected to urea or carbamate formation at the indoline 1-position to provide the
final product 9a or 9b. The requisite N-(2,2,3,3,3-pentafluoropropyl)piperidine 11 was prepared
from 4-hydroxypiperidine and 2,2,3,3,3-pentafluoropropyl trifluoromethanesulfonate.
a
Scheme 1. Synthesis of indoline-5-sulfonamide derivatives 9a and 9b
6

a
Reagents and conditions: (a) 4-chloro-2,6-difluoroaniline, 2,2,2-trifluoroethanol, MS4A,
reflux; (b) Zn, AcOH, rt, 86% over 2 steps; (c) 4-chlorobutyryl chloride, DMA, 0 °C to rt; (d)
NaH, DMF, 0 °C to rt, 91% over 2 steps; (e) NaOH, H O, THF, MeOH, rt, 87%; (f) 4-
trifluoromethyl)aniline, triphosgene, pyridine, THF, 0 °C; then 8, TEA, THF, 0 °C to rt, 83%;
g) 11, triphosgene, TEA, THF, 0 °C to rt; then 8, THF, 0 °C to rt, 87%; (h) 2,2,3,3,3-
2
(
(
pentafluoropropyl trifluoromethanesulfonate, TEA, THF, 0 °C to rt, 88%.
To efficiently access 7-(2-oxopyrrolidin-1-yl)indoline-5-sulfonamide derivatives with various
anilines at the 5-position, we employed the method shown in Scheme 2, in which the lactam ring
was constructed at the 7-position prior to sulfonamide formation. The nitro group of 12 was
selectively reduced with sodium hydrosulfite to give aniline 13, which was converted to the
lactam 15 by a two-step sequence. Palladium-catalyzed coupling of 15 with benzyl mercaptan
followed by oxidative chlorination using N-chlorosuccinimide (NCS) in aqueous acetic acid gave
a common intermediate 17. After sulfonamidation of 17 with anilines, 18a–c were transformed to
the final products 20a–c in analogy with compound 9b in Scheme 1.
a
Scheme 2. Synthesis of indoline-5-sulfonamide derivatives 20a–c
7

a
Reagents and conditions: (a) Na
chloride, DMA, 0 °C to rt; (c) NaH, DMF, 0 °C to rt, 82% over 2 steps; (d) BnSH, Pd
Xantphos, DIEA, toluene, reflux, 96%; (e) NCS, AcOH, H O, rt, 81%; (f) 2-fluoro-4-
trifluoromethyl)aniline, 2,2,2-trifluoroethanol, 80 °C, 87%; (g) 2,4,6-trifluoroaniline, pyridine,
2
S
2
O
4
, EtOH, THF, H
2
O, 60 °C, 73%; (b) 4-chlorobutyryl
2
(dba)
3
,
2
(
rt, 70%; (h) H (1 atm), Pd on carbon, THF, MeOH, rt, 93%–quant.; (i) 11, triphosgene, TEA,
2
THF, 0 °C to rt, 59–79%.
Scheme 3 describes the synthetic steps of 1-phenylindoline derivatives. The introduction of a
phenyl group at the indoline 1-position of 8 was achieved by Buchwald–Hartwig amination
using RuPhos precatalyst, which was reported as an effective catalyst for C–N bond formation
1
5, 16
between aryl halides and secondary amines,
underwent a selective bromination upon treatment with one equivalent of N-bromosuccinimide
NBS) to give 22. Finally, Suzuki–Miyaura coupling reaction of 22 with an arylboronic acid
to afford 21 in a moderate yield. Compound 21
(
provided the desired biphenyl derivative 23.
a
Scheme 3. Synthesis of indoline-5-sulfonamide derivative 23
8

a
t
Reagents and conditions: (a) PhBr, RuPhos precatalyst, RuPhos, NaO Bu, DME, reflux, 65%;
b) NBS, DMF, 0 °C, 99%; (c) 4-(trifluoromethyl)phenylboronic acid, Pd(PPh , K CO , THF,
reflux, 63%.
(
3
)
4
2
3
1
-(Pyrimidin-2-yl)indoline derivatives 27a–f were synthesized using the methods depicted in
Scheme 4. Nucleophilic substitution reaction of 8 with 2-chloro-5-nitropyrimidine cleanly
proceeded under refluxing in THF to furnish 24 in a good yield. This type of S Ar reaction of 8
N
required an electron-withdrawing group on the pyrimidine ring, such as a nitro or alkoxycarbonyl
group, for high conversion to the desired compounds. The nitro group of 24 was reduced by
hydrogenation using Adams’ catalyst (platinum oxide) to give 25, which was subsequently
subjected to Sandmeyer reaction using n-pentyl nitrite and copper(II) bromide to provide 5-
bromopyrimidine 26. The final compounds 27a–c were obtained by palladium-catalyzed
coupling reactions of 26 with aryl boronic acids. In contrast, copper-catalyzed N-arylation of the
substituted pyrazoles with 26 provided 27d–f. The chemical structure of 27e was confirmed by
X-ray crystallographic analysis of a single crystal structure (Figure 2), and the structure of 27f
1
7
was determined by nuclear magnetic resonance (NMR) experiments.
a
Scheme 4. Synthesis of indoline-5-sulfonamide derivatives 27a–f
9

a
Reagents and conditions: (a) 2-chloro-5-nitropyrimidine, THF, reflux, 98%; (b) H
PtO , THF, EtOH, rt, 93%; (c) n-pentyl nitrite, CuBr , CH CN, 0 °C to 60 °C, 45%; (d)
ArB(OH) , Pd(PPh , Na CO , THF/H O or DME/H O, 110–130 °C (microwave), 50–87% (for
7a and 27b); (e) ArB(OH) , Pd(PPh , K CO , THF, reflux, 38% (for 27c); (f) substituted
pyrazole, CuI, 8-quinolinol, K CO , DMSO, 130°C, 11–42% (for 27d–f).
2
(1 atm),
2
2
3
2
3
)
4
2
3
2
2
2
2
3
)
4
2
3
2
3
Figure 2. ORTEP of 27e, thermal ellipsoids are drawn at 50% probability.
Scheme 5 details the synthesis of compounds with a cyclic urea at the indoline 7-position. In
1
0

the urea formation reaction of 5 with 2-chloroethyl isocyanate, the addition of a catalytic amount
of hydrogen chloride was found to accelerate the reaction rate and suppress undesired urea
formation on the sulfonamide nitrogen. The obtained 28 was subsequently subjected to intra-
molecular cyclization upon treatment with sodium hydride to give cyclic urea 29 in good isolated
yields over two steps. Compound 29 was elaborated to the final product 34 in analogy with 27c
in Scheme 4.
a
Scheme 5. Synthesis of indoline-5-sulfonamide derivative 34
a
Reagents and conditions: (a) 2-chloroethyl isocyanate, cat. HCl, THF, 70 °C; (b) NaH, DMF,
0
°C to rt, 77% over 2 steps; (c) NaOH, H
nitropyrimidine, THF, reflux, 92%; (e) H (1 atm), PtO
nitrite, CuBr , CH CN, rt, 39%; (g) 1-methyl-3-trifluoromethyl-1H-pyrazole-5-boronic acid
pinacol ester, Pd(PPh , K CO , THF, 100 °C (microwave), 46%.
2
O, THF, MeOH, reflux, 92%; (d) 2-chloro-5-
2
2
, THF, EtOH, rt, 97%; (f) tert-butyl
2
3
3
)
4
2
3
3
. Results and Discussion
The newly synthesized compounds were tested for MGAT2 inhibitory activity using
membrane fractions from the MGAT2 expression. To obtain in-depth profiles of N-
11

phenylindoline-5-sulfonamide derivatives for the development of novel MGAT2 inhibitors, we
conducted further evaluation of compounds 1 and 2 in terms of ADME-Tox profiles. We found
that these compounds displayed TDI of CYP3A4, discouraging their further development (Table
1
). TDI of CYP3A4 was evaluated by a single-point assay, in which test compounds were
preincubated with human liver microsomes in the presence of NADPH before measuring
CYP3A4 enzymatic activity toward its specific substrate. Preincubation of compound 1 or 2 with
microsomes led to a significant loss of the activity of CYP3A4 relative to that under a condition
without preincubation. Another series of indoline-5-sulfonamide-based MGAT2 inhibitors with a
non-aromatic ring connected by a carbamate linkage at the 1-position, represented by 20c, also
exhibited possible liability for CYP3A4 TDI. Accordingly, our subsequent optimization effort
was directed toward addressing the undesirable CYP3A4 TDI issue of this series of compounds.
Table 1. MGAT2 inhibitory activity and CYP3A4 TDI profiles
a
b
hMGAT2
IC50 (nM)
CYP3A4 TDI
c
Compound
LogD7.4
(% remaining)
1
.0
1
2
13
43
59
2.5
3.1
3.7
(
0.89–1.2)
3
.4
(
3.0–3.9)
4
.1
2
0c
(
3.4–4.9)
a
Inhibitory activity against human MGAT2. IC50 values are presented as means of duplicate
bexperiments with their 95% confidence intervals in parentheses.
CYP3A4 time dependent inhibition assay (n = 2). The remaining activity of CYP3A4 after
preincubation with a test compound was determined.
1
2

c
LogD value at pH 7.4.
We assumed that the 2,4-difluoroaniline moiety, a common structural motif for 1, 2, and 20c,
would be the cause of mechanism-based inactivation (MBI) of CYP3A4. Presumed mechanisms
of MBI of 2,4-difluoroaniline derivatives are depicted in Figure 3. It has been reported that 4-
1
8, 19
fluoroanilines are relatively easily subject to CYP-catalyzed defluorination,
and this
2
0, 21
metabolism has the potential to cause bioactivation.
We hypothesized that oxidative
metabolism of the benzene ring of 2,4-difluoroanilines could generate reactive metabolites such
as an epoxide or iminoquinone, which may be attacked by an amino acid residue of the CYP3A4
active site to form a covalent bond, leading to an irreversible inactivation of CYP3A4. According
to this hypothesis, structural modifications to the 2,4-difluoroaniline moiety could mitigate TDI
liability. Our compound design was based on the following concepts: 1) replacement of the
fluorine atom at the 4-position with a substituent more resistant to metabolism and 2)
introduction of an electron-withdrawing group onto the benzene ring to lower susceptibility to
oxidative metabolism.
Figure 3. Postulated mechanism of CYP3A4 inactivation
1
3

Table 2 summarizes results of CYP3A4 TDI of representative compounds together with their
MGAT2 inhibitory activities. Gratifyingly, the conversion of the 4-F group of 20c to a
trifluoromethyl group significantly improved CYP3A4 TDI risk (20a). Reducing the electron
density of the aniline moiety by introducing an additional fluorine atom also conferred a
beneficial effect on CYP3A4 TDI profile, and the 2,4,6-trifluoroaniline derivative 20b did not
inactivate CYP3A4 after preincubation. These results clearly indicated that the 2,4-
difluoroaniline moiety of 1, 2, or 20c was the cause of CYP3A4 TDI. Unfortunately, the
improvement of CYP3A4 TDI of compounds 20a and 20b was accompanied by a 10–20-fold
decline in potency in comparison with that of the corresponding 2,4-difluoroaniline derivative
2
0c. Further exploration of the aniline ring substitution led to the identification of 4-chloro-2,6-
difluoroaniline as a desirable aniline motif for this series of compounds that could eliminate
CYP3A4 TDI liability and minimize the decrease in potency as shown by compound 9b. In
2
2
terms of ligand lipophilicity efficiency, compound 9b displayed the highest LLE value among
2
0a, 20b, and 9b. The impact of 4-chloro-2,6-difluoroaniline on CYP3A4 TDI was also
confirmed by the comparison of 1-carbamoylindoline derivatives 9a and 2, although a
concomitant loss of potency was observed as in the case of the carbamates (9b vs 20c).
Table 2. MGAT2 inhibitory activity and CYP3A4 TDI profiles
a
b
1
2
hMGAT2
IC50 (nM)
CYP3A4 TDI
(% remaining)
c
d
Compound
R
R
LogD7.4
LLE
1
4

4
2
2
0a
0b
93
4.1
3.3
(
32–55)
7
5
2
95
82
89
3.4
3.8
3.3
3.7
3.9
4.2
(
(
(
58–96)
1
9
9
b
15–24)
3
0
9
a
26–34)
a
Inhibitory activity against human MGAT2. IC50 values are presented as means of duplicate
bexperiments with their 95% confidence intervals in parentheses.
CYP3A4 time dependent inhibition assay (n = 2). The remaining activity of CYP3A4 after
preincubation with a test compound was determined.
c
23
LogD value at pH 7.4.
d
LLE = pIC50 − LogD7.4
.
Given the desired aniline moiety with balanced profiles, our subsequent optimization effort
was directed to structural modification at the indoline 1-position with the aim of enhancing
MGAT2 inhibitory activity. We applied a strategy to rigidify the substituent at the indoline 1-
position to restrict conformation of the terminal hydrophobic motif, which is important for
1
3
interaction with MGAT2. Thereby, several biaryl-type substituents at the 1-position were
designed by the conversion of the urea linkage of 9a to an aryl ring (Figure 4).
Figure 4. Design stage from 1-carbamoylindoline derivatives to 1-arylindoline derivatives
The MGAT2 activities of the ensuing biaryl derivatives are presented in Table 3. The biphenyl
1
5

analog 23 exhibited dramatically enhanced MGAT2 inhibitory activity with an IC50 value of 0.25
nM. The LLE value of compound 23 was also considerably improved compared with that of the
urea linkage analog 9a, with a change from 4.2 to 5.1, suggesting that this structural
transformation achieved a larger increase in potency than simply expected from the change in
lipophilicity. Hence, it is supposed that the introduction of the biaryl alignment at the indoline 1-
position enabled the terminal trifluoromethylphenyl ring to adopt the desired orientation for
efficient interaction with the presumed hydrophobic site of MGAT2. Furthermore, the 5-
phenylpyrimidine derivative 27a, in which two nitrogen atoms were embedded at the
corresponding positions to the hetero atoms of the urea linkage of 9a, also displayed potent
MGAT2 inhibitory activity with a subnanomolar IC50 value. This modification substantially
lowered lipophilicity and thereby further improved the LLE value to 5.4. It was also confirmed
that compounds 23 and 27a displayed no CYP3A4 TDI liability (data not shown).
Table 3. MGAT2 inhibitory activity and phototoxicity profiles
b
Phototox
(% of control)
a
hMGAT2
IC50 (nM)
HOMO–LUMO
d
e
Compound
Ar
X
Y
c
LogD7.4
LLE
gap (eV)
UV−
UV+
2 ± 0
0
.25
2
3
CH
N
CH
CH
2
2
95 ± 5
7.90
8.18
4.5
3.8
5.1
5.4
(
(
0.22–0.29)
0
.69
2
7a
105 ± 8
54 ± 6
0.59–0.81)
1
6

0
.42
2
2
7b
7d
N
N
N
CH
CH
CH
2
2
2
98 ± 6
109 ± 7
105 ± 1
37 ± 16
84 ± 4
91 ± 5
8.01
8.31
8.31
4.1
3.1
3.1
5.3
5.1
5.0
(
0.37–0.47)
6
.2
(
5.5–7.1)
8
.2
2
7e
(
6.7–10)
2
.1
2
7f
N
CH
CH
2
2
112 ± 4
98 ± 7
8.32
2.9
5.8
(
1.9–2.4)
7
.8
2
7c
N
N
99 ± 8
96 ± 8
8.60
8.60
2.8
2.7
5.3
5.7
(
(
6.6–9.2)
3
.7
3
4
NH
111 ± 7
107 ± 6
3.2–4.3)
a
Inhibitory activity against human MGAT2. IC50 values are presented as means of duplicate
bexperiments with their 95% confidence intervals in parentheses.
Phototoxicity assay using BALB/c 3T3 cell in the presence of 6.25 µM of compounds.
cPercentages of control are represented as means ± standard deviation (n = 3).
The HOMO–LUMO energy gap (eV) of each simplified biarylaniline I calculated by AM1
method using MOPAC (23-I, 27a-I, 27b-I, 27d-I, 27e-I, 27f-I, 27c-I, and 34-I).
d
LogD value at pH 7.4.
LLE = pIC50 − LogD7.4
e
.
Despite their attractive potencies for MGAT2 inhibition, it was found that the biaryl
compounds 23 and 27a exhibited cytotoxicity against BALB/c 3T3 cells under UV irradiation
condition, whereas cytotoxic effect was not observed under a condition without UV exposure
(
Table 3), thereby indicating potential phototoxic liability. Phototoxicity is one of the drug-
induced adverse effects, and it is characterized by an exaggerated inflammatory response to
2
4
sunlight caused by a photoreactive chemical. Although whether a chemical causes
phototoxicity in vivo depends on the distribution of the compound to light-exposed tissues such
as the skin or the eyes as well as its photoreactivity, it is necessary to develop a drug candidate
that has essentially no potential photoreactivity in terms of patient safety.
1
7

We assumed that the biaryl motif at the indoline 1-position of 23 or 27a would be a
chromophore to induce phototoxicity because non-biaryl derivatives such as 9a did not exert a
UV-induced cytotoxic effect (data not shown). To remove the potential phototoxic liability, we
implemented a strategy of increasing energy gaps between the highest occupied molecular orbital
2
5,
(
HOMO) and the lowest unoccupied molecular orbital (LUMO), focusing on the biaryl motif.
2
6
27
A semiempirical molecular orbital method (AM1) using MOPAC was applied in Molecular
2
8
Operating Environment software (MOE) to calculate the HOMO, LUMO, and HOMO–LUMO
energy gaps of biarylanilines I, which are simplified versions of each compound (Table 3).
According to the calculation, the HOMO orbital of 23-I was localized on the phenyl ring A,
whereas the LUMO orbital was mainly found in the phenyl ring B (Figure 5). On the basis of the
localization of the HOMO and LUMO orbitals, we reasoned that reduction of the electron
density of ring A or enhancement of the electron density of ring B would lead to widening of the
HOMO–LUMO energy gap. Indeed, the conversion of phenyl ring A to an electron-deficient
pyrimidine ring increased the HOMO–LUMO gap, with a change from 7.90 (23-I) to 8.18 eV
(
27a-I). This transformation improved the phototoxicity risk to some extent, judging from the
BALB/c 3T3 cell assay, although compound 27a still exhibited phototoxic liability. In contrast,
the introduction of an electron-withdrawing fluorine atom on the terminal phenyl ring of 27a to
give 27b reduced the HOMO–LUMO gap and resulted in a similar level of UV-induced
cytotoxicity, as displayed by 27a. These results were in good agreement with our hypothesis
based on the relationship between phototoxic potential and the HOMO–LUMO energy gaps of
the biaryl moiety.
1
8

Figure 5. Molecular orbital of HOMO and LUMO of compound 23-I calculated by MOE
(
MOPAC, AM1).
We selected 1-(pyrimidin-3-yl)indoline derivatives for further optimization in terms of potent
inhibitory activity, lower lipophilicity, and improved phototoxicity liability. The AM1 calculation
revealed that the molecular orbital localization of HOMO and LUMO of 27a-I was similar to
that of 23-I, as shown in Figure 5. To further widen the HOMO–LUMO gap, we designed
compounds with an electron-rich 5-membered heterocycle at the 5-position on the pyrimidine
ring while maintaining a terminal trifluoromethyl group that could effectively interact with the
1
3
presumed hydrophobic site of MGAT2. As expected, compared with 27a, the pyrazole
derivatives 27d and 27e exhibited larger HOMO–LUMO gaps and greatly improved phototoxic
liability (Table 3). These compounds have desirable lipophilicity, with LogD values of
approximately 3, and they retained high LLE values of MGAT2 inhibitory activity exceeding 5.
Furthermore, the installation of a methyl group at the 5-position on the pyrazole ring of 27e
enhanced MGAT2 inhibitory activity without affecting the phototoxic potential (27f). It is worth
noting that compound 27f displayed lower lipophilicity than 27e (LogD = 2.9 (27f) and 3.1
(
27e)) despite the methyl addition. The calculated LogP value could not predict this change in
molecular properties (cLogP = 5.05 (27f) and 4.78 (27e)). This can be rationalized by the
1
9

disruption of an intramolecular hydrogen bond between the pyrazole nitrogen atom at the 2-
position and the hydrogen atom on the pyrimidine ring by the methyl addition of 27f. Because
the enhanced potency of 27f by the methyl addition for MGAT2 inhibition was not likely driven
by lipophilicity, it is assumed that the methyl group would have a direct productive interaction
with MGAT2.
Following the N-linked 1-arylpyrazoles 27d–f, a C-linked 5-arylpyrazole 27c, in which a
beneficial methyl group was introduced on the pyrazole ring, was also investigated. Compound
2
7c displayed potent MGAT2 inhibitory activity, with an IC50 value of 7.8 nM. In addition, this
compound had a widened HOMO–LUMO energy gap and exhibited clearly improved phototoxic
liability relative to 27a. To further enhance MGAT2 inhibitory activity without increasing
lipophilicity, the electron density on the carbonyl group at the indoline 7-position was increased
13
by replacement with a cyclic urea motif according to the knowledge we previously reported.
This approach was found to be effective in the biaryl series, and the consequent compound 34
exhibited improved potency and reduced lipophilicity, resulting in a higher LLE value. In
addition, no phototoxicity was observed for this compound. Taken together, compounds 27f, 27c,
and 34 demonstrated balanced profiles in terms of MGAT2 inhibitory activity, desirable
lipophilicity, and removed phototoxic liability, which made these compounds attractive
candidates for further evaluation.
Prior to the in vivo evaluation, the selected compounds 27f, 27c, and 34 were evaluated for
mouse MGAT2 inhibitory activity, metabolic stability, and pharmacokinetic profiles (Table 4).
All three compounds also exhibited potent inhibitory activity against mouse MGAT2, with IC50
values of approximately 1 nM and good metabolic stability toward both human and mouse
microsomes. The low microsomal clearances of 27f and 27c were reflected in low in vivo
2
0

clearances in mice (CLtotal = 471 and 205 mL/h/kg, respectively), and these two compounds
displayed favorable pharmacokinetic profiles with good oral bioavailability (F = 52% and 50%,
respectively) and prolonged MRTs (2.6 and 3.3 h, respectively) after oral administration at the
dose of 1 mg/kg. In sharp contrast to 27c, the corresponding cyclic urea analog 34 displayed
modest oral bioavailability and low plasma concentration despite its low microsomal clearance.
Given a comparable aqueous solubility of 34 relative to 27c (data not shown), this result was
likely attributed to the poor membrane permeability of 34 (PAMPA permeability = 166 (27c) and
4
0 (34) nm/s). A similar negative impact of the one-atom conversion to cyclic urea on
pharmacokinetic profiles was also previously observed in the series of 1-carbamoylindoline
1
3
derivatives. Consequently, a lactam ring serves as a desirable substituent at the indoline 7-
position with a balanced profile of potency, physicochemical profiles, and pharmacokinetic
parameters.
Table 4. MGAT2 inhibitory activity, metabolic stability, and pharmacokinetic profiles
b
c, d
CLint (µL/min/mg)
Pharmacokinetic profiles in mice
a
e
mMGAT2
IC50 (nM)
PAMPA
Compound
c
c
d
d
d
Vdss
CLtotal
AUC
MRT
(h)
F
(nm/sec)
HLM
ND
ND
4
MLM
27
(
mL/kg)
(mL/h/kg)
(ng·h/mL)
(%)
1
.1
2
7f
1019
471
1120
2.6
3.3
2.3
52
185
166
40
(
0.82–1.5)
2
.4
2
7c
ND
33
539
205
2436
72
50
16
(
2.1–2.7)
1
.9
3
4
1964
2368
(
1.6–2.3)
a
Inhibitory activity against mouse MGAT2. IC50 values are presented as means of duplicate
bexperiments with their 95% confidence intervals in parentheses.
Human/mouse liver microsomal clearance. ND: The decrease of the compound concentration
was not observed.
c
0
1
.1 mg/kg, iv.
mg/kg, po.
d
e
PAMPA permeability at pH 7.4.
2
1

Compound 27c was selected as a representative compound and evaluated for target selectivity.
This compound exhibited IC50 values of more than 10 µM against DGAT1, DGAT2, and ACAT1,
thereby demonstrating excellent MGAT2 selectivity over related acyltransferases. The potent
MGAT2 inhibitory activity, high selectivity over other acyltransferases, and favorable
pharmacokinetic profile of 27c motivated us to evaluate the in vivo efficacy by oral
administration.
We conducted OFTT using C57BL/6J mice to examine the compound’s effect on
hypertriglyceridemia following olive oil challenge. To measure the accumulation of plasma TG,
mice were pre-treated with a lipoprotein lipase (LPL) inhibitor, Pluronic F127 (Poloxamer 407),
2
9
which inhibits the hydrolysis of plasma TG by LPL. Compound 27c was orally administered 6
h before the oral olive oil challenge, and the plasma chylomicron TG concentration was
monitored for 4 h. As shown in Figure 6, the plasma TG excursion was effectively suppressed by
the administration of 27c. The inhibitory effect of 27c on TG elevation was dose-dependent and
significant (p < 0.025) at doses of 1 and 3 mg/kg. The inhibition degree of TG by 27c at the dose
of 3 mg/kg was comparable with that observed in MGAT2 gene knockout mice (unpublished
data), suggesting that this compound nearly completely inhibited MGAT2 in vivo at this dose.
These results promised that compound 27c would exhibit beneficial pharmacological effects for
the treatment of metabolic disorders.
2
2

Figure 6. Effect of 27c on plasma TG elevation during oral fat tolerance test in C57BL/6J mice.
7c (0.3, 1, 3 mg/kg) was orally administered 6 h prior to olive oil load. (A) Changes in plasma
2
chylomicron/TG (CM/TG) levels after olive oil load. (B) Area under the curve (AUC) of plasma
CM/TG levels during 4 h after oil load. Data are the mean ± SD (n = 5). #, p ≤ 0.025 versus
vehicle by William's test.
Compound 27c was further assessed in terms of the DDI potential. As shown in Table 5,
eliminated liability for CYP3A4 TDI of 27c was confirmed. In addition, this compound
exhibited little or no reversible inhibition against major CYP isoforms (1A2/2C8/2C9/2D6/3A4)
at the concentration of 10 µM. These results were suggestive of a safe DDI profile of 27c, and
they warrant further development of this compound.
Table 5. CYP3A4 TDI and reversible CYP inhibition profiles of 27c
a
CYP3A4 TDI
Reversible CYPs inhibition
Compound
(
% remaining)
(% inhibition at 10 µM)
2
3

1
A2
2C8
32
2C9
17
2D6
< 10
3A4
< 10
27c
107
< 10
a
CYP3A4 time dependent inhibition assay (n = 2). The remaining activity of CYP3A4 after
preincubation with a test compound was determined.
4
. Conclusion
MGAT2 has emerged as a potential peripheral target for the treatment of obesity and metabolic
disorders. To develop a novel class of MGAT2 inhibitors with desirable DMPK and toxicological
profiles, we continued our optimization campaign of N-phenylindoline-5-sulfonamide
derivatives for addressing the potential CYP3A4 TDI liability. Structural modification was
focused on the 2,4-difluoroaniline motif on the basis of the hypothesis of oxidative bioactivation
of this moiety. Replacement of the fluorine atom at the 4-position with a substituent more
resistant to metabolism was combined with the introduction of an additional fluorine atom on the
benzene ring to reduce electron density, resulting in the identification of 4-chloro-2,6-
difluoroaniline derivatives that lacked CYP3A4 TDI liability, with minimal loss of MGAT2
inhibition potency. Subsequently, biaryl substituents were introduced at the 1-position to enhance
MGAT2 inhibitory activity. The resulting phototoxic liability was successfully removed by an
appropriate ring alignment based on the HOMO–LUMO gap hypothesis, which led to the
discovery of 27c. Compound 27c exhibited potent and selective MGAT2 inhibitory activity,
desirable physicochemical properties, favorable oral bioavailability, and prolonged
pharmacokinetic profiles in mice. These profiles were reflected in the potent in vivo efficacy, and
2
7c thereby effectively suppressed plasma TG excursion after olive oil loading in OFTT. Taken
together, these results suggest that 27c could have a therapeutic potential for the treatment of
metabolic diseases, including obesity, as observed in MGAT2 gene knockout mice. Further
pharmacological evaluation of this compound will be disclosed in due course.
2
4

5
5
. Experimental section
.1. Chemistry
Melting points were determined in open capillary tubes on a Büchi melting point apparatus
1
13
B545 and are uncorrected. H NMR spectra and C NMR spectra were recorded on Bruker
AVANCE III (300 MHz) or Bruker Advance III plus (400 MHz) spectrometer. Chemical shifts
are given in parts per million (ppm) downfield from tetramethysilane (δ) as the internal standard
in deuterated solvent and coupling constants (J) are in Hertz (Hz). Data are reported as follows:
chemical shift, integration, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, quint =
quintet, dd = doublet of doublets, dt = doublet of triplets, dq = doublet of quartets, td = triplet of
doublets, tt = triplet of triplets, tq = triplet of quartets, qd = quartet of doublets, qt = quartet of
triplets, ddd = doublet of doublet of doublets, m = multiplet, bs = broad signal), and coupling
constants. All solvents and reagents were obtained from commercial suppliers and used without
further purification. Liquid-liquid extraction in work-up procedures was generally conducted
with 30–200 times volume of the indicated organic solvents to the reaction substrates (v/w).
Thin-layer chromatography (TLC) was performed on Merck silica gel plates 60F254. Silica gel
column chromatography was performed on Purif-Pack (SI or NH, SHOKO SCIENTIFIC). LC–
MS analysis was performed on a Shimadzu Liquid Chromatography–Mass Spectrometer System,
operating in APCI (+ or −) or ESI (+ or −) ionization mode. Analytes were eluted using a linear
gradient of 0.05% TFA containing water/acetonitrile or 5 mM ammonium acetate containing
water/acetonitrile mobile phase. The purities of compounds submitted for biological evaluation
were > 95% as determined by elemental analyses within ±0.4% of the calculated values. Yields
are not optimized.
2
5

5
.1.1. Benzyl 7-amino-5-[(4-chloro-2,6-difluorophenyl)sulfamoyl]-2,3-dihydro-1H-indole-1-
carboxylate (5)
1
3
Step A. A mixture of compound 3 (108 g, 271 mmol), 4-chloro-2,6-difluoroaniline (53.2 g,
3
25 mmol), molecular sieves 4A (pellet, 300 g, unactivated), and 2,2,2-trifluoroethanol (1084
mL) was stirred at reflux for 24 h. After cooling to room temperature, insoluble materials were
filtered off through a pad of celite. The filtrate was concentrated under reduced pressure and the
i
residue was diluted with AcOEt– Pr
2
O (1:1 v/v). Insoluble materials were filtered off and the
filtrate was concentrated under reduced pressure to give benzyl 5-[(4-chloro-2,6-
difluorophenyl)sulfamoyl]-7-nitro-2,3-dihydro-1H-indole-1-carboxylate as a brown amorphous
solid (4, 142 g). This product was used for the next step without further purification. MS
−
1
(
ESI/APCI) m/z 522.1 [M − H] . H NMR (300 MHz, DMSO-d ) δ 3.19–3.29 (2H, m), 4.27 (2H,
6
t, J = 8.4 Hz), 5.20 (2H, s), 7.26–7.53 (7H, m), 7.80–8.06 (2H, m), 10.35 (1H, s).
Step B. To a mixture of compound 4 (142 g) obtained above in acetic acid (1506 mL) was
added portionwise zinc powder (89.0 g, 1.36 mol) at room temperature. The mixture was stirred
with a mechanical stirrer at room temperature for 13 h under N atmosphere. THF was added to
2
the reaction mixture followed by stirring at room temperature for 1.5 h. Insoluble materials were
removed by filtration through a pad of celite and the filter cake was washed with THF. After the
filtrate was concentrated under reduced pressure, the residue was suspended in AcOEt and stirred
i
at 70 °C. Pr
2
O was added to the mixture at 70 °C followed by stirring for 10 min at the same
temperature. After cooling to room temperature, the precipitated solid was collected,
i
i
successively washed with AcOEt– Pr
2
O (1:2 v/v) and Pr
2
O, and dried under reduced pressure to
give the title compound as a pale purple solid (5, 116 g, 235 mmol, 86% over 2 steps). MS
+
1
(
ESI/APCI) m/z 494.2 [M + H] . H NMR (400 MHz, DMSO-d ) δ 2.99 (2H, t, J = 8.0 Hz), 4.09
6
2
6

(
2H, t, J = 8.1 Hz), 5.22 (2H, s), 5.67 (2H, br s), 6.85 (1H, s), 6.96 (1H, s), 7.31–7.57 (7H, m),
9
.57–10.03 (1H, m).
.1.2. Benzyl
5
5-[(4-chloro-2,6-difluorophenyl)sulfamoyl]-7-(2-oxopyrrolidin-1-yl)-2,3-
dihydro-1H-indole-1-carboxylate (7)
To a mixture of compound 5 (98.5 g, 199 mmol) in DMA (985 mL) was added 4-
chlorobutyryl chloride (33.7 g, 239 mmol) at 0 °C. The mixture was stirred at room temperature
overnight. The reaction mixture was quenched with water at 0 °C and extracted with AcOEt. The
organic layer was separated, washed with water and brine, dried over MgSO
under reduced pressure. The residue was dissolved in DMF (1 L) and thereto sodium hydride
60% oil dispersion, 17.5 g, 437 mmol) was added at 0 °C. The mixture was stirred at room
temperature for 2 h. The reaction was quenched by addition of saturated aqueous NH Cl
solution at 0 °C and the mixture was extracted with AcOEt. The organic layer was separated,
4
, and concentrated
(
4
washed with water and brine, dried over MgSO . After removal of the solvent, the residue was
4
i
crystallized from AcOEt– Pr
2
O to give the title compound as a light brown crystal (102 g, 182
+
1
mmol, 91% over 2 steps). MS (ESI/APCI) m/z 562.3 [M + H] . H NMR (400 MHz, DMSO-d
6
)
δ 1.93–2.06 (2H, m), 2.15–2.25 (2H, m), 3.12 (2H, t, J = 7.8 Hz), 3.68 (2H, t, J = 6.0 Hz), 4.13
13
(
(
(
2H, t, J = 7.8 Hz), 5.16 (2H, s), 7.30–7.44 (7H, m), 7.48 (2H, br s), 10.05 (1H, s). C NMR
101 MHz, DMSO-d ) δ 18.2, 27.6, 30.0, 48.9, 49.7, 66.9, 112.7 (t, J = 16.9 Hz), 113.1–113.4
m), 120.3, 124.3, 126.1, 127.5, 127.9, 128.3, 132.7 (t, J = 12.8 Hz), 135.4, 136.3, 137.0, 140.1,
6
1
51.6, 158.9 (dd, J = 253.5, 5.5 Hz), 171.9. Mp 197–199 °C.
.1.3. N-(4-Chloro-2,6-difluorophenyl)-7-(2-oxopyrrolidin-1-yl)-2,3-dihydro-1H-indole-5-
5
sulfonamide (8)
A mixture of compound 7 (97.5g, 173 mmol), 8 N aqueous NaOH solution (651 mL), MeOH
2
7

(
560 mL), and THF (560 mL) was stirred at room temperature for 3 days under N
2
atmosphere.
After cooling to 0 °C, the reaction mixture was neutralized with 1 M hydrochloric acid. The
mixture was extracted with AcOEt, washed with water and brine, dried over MgSO
4
, and
O to
i
concentrated under reduced pressure. The residue was crystallized from AcOEt–hexane– Pr
2
give the title compound as a light brown solid (64.8 g, 152 mmol, 87%). MS (ESI/APCI) m/z
+
1
4
3
28.1 [M + H] . H NMR (400 MHz, DMSO-d
6
) δ 2.02–2.15 (2H, m), 2.38 (2H, t, J = 7.8 Hz),
.03 (2H, t, J = 8.5 Hz), 3.51–3.68 (4H, m), 6.07 (1H, br s), 7.16 (1H, s), 7.25 (1H, s), 7.38 (2H,
1
3
d, J = 7.9 Hz), 9.61 (1H, br s). C NMR (101 MHz, DMSO-d
6
) δ 18.4, 28.4, 30.7, 46.2, 48.4,
13.0–113.3 (m), 118.9, 120.8, 124.2, 126.9, 131.2, 132.3 (t, J = 13.2 Hz), 151.0, 159.0 (dd, J =
53.8, 5.9 Hz), 173.5. Mp 146–149 °C.
.1.4. 5-[(4-Chloro-2,6-difluorophenyl)sulfamoyl]-7-(2-oxopyrrolidin-1-yl)-N-[4-
1
2
5
(
trifluoromethyl)phenyl]-2,3-dihydro-1H-indole-1-carboxamide (9a)
To a mixture of 4-(trifluoromethyl)aniline (113 mg, 0.701 mmol) and pyridine (0.056 mL,
0
.70 mmol) in THF (2 mL) was added triphosgene (72.8 mg, 0.245 mmol) at 0 °C. After stirring
at 0 °C for 40 min, the mixture was added to a mixture of compound 8 (150 mg, 0.351 mmol)
and triethylamine (0.147 mL, 1.05 mmol) in THF (2 mL) at 0 °C. The resulting mixture was
stirred at room temperature overnight. The reaction was quenched by addition of 0.02 M
hydrochloric acid and the mixture was extracted with AcOEt. The organic layer was successively
washed with aqueous NaHCO
3
solution and brine, dried over MgSO , and concentrated under
4
reduced pressure. The residue was purified by silica gel column chromatography (hexane/AcOEt
i
=
70/30 to 30/70) and crystallized from acetone– Pr
2
O to give the title compound as a white solid
+
1
(
179 mg, 0.292 mmol, 83%). MS (ESI/APCI) m/z 615.3 [M + H] . H NMR (300 MHz, DMSO-
d
6
) δ 1.92–2.06 (2H, m), 2.13–2.24 (2H, m), 3.18 (2H, t, J = 8.2 Hz), 3.76 (2H, t, J = 6.8 Hz),
2
8

4
1
.21 (2H, t, J = 8.3 Hz), 7.36–7.51 (4H, m), 7.60–7.69 (2H, m), 7.70–7.78 (2H, m), 9.56 (1H, s),
13
0.02 (1H, s). C NMR (101 MHz, DMSO-d ) δ 18.5, 28.0, 30.3, 48.6, 50.8, 112.8 (t, J = 16.9
6
Hz), 113.1–113.4 (m), 118.7, 120.2, 122.2 (q, J = 31.8 Hz), 124.1, 124.5 (q, J = 271.4 Hz), 125.8,
1
5
25.7–125.9 (m), 132.7 (t, J = 13.2 Hz), 134.4, 136.6, 141.2, 143.5, 151.4, 158.9 (dd, J = 253.8,
.9 Hz), 171.7. Mp 212–214 °C. Anal. Calcd for C26
H20ClF
5
N
4
O
4
S: C, 50.78; H, 3.28; N, 9.11.
Found: C, 50.91; H, 3.20; N, 8.95.
.1.5. 1-(2,2,3,3,3-Pentafluoropropyl)piperidin-4-yl
5
5-[(4-chloro-2,6-
difluorophenyl)sulfamoyl]-7-(2-oxopyrrolidin-1-yl)-2,3-dihydro-1H-indole-1-carboxylate
9b)
To a mixture of 1-(2,2,3,3,3-pentafluoropropyl)piperidin-4-ol (11, 98.0 mg, 0.420 mmol) and
(
triethylamine (0.059 mL, 0.42 mmol) in THF (2 mL) was added triphosgene (41.6 mg, 0.140
mmol) at 0 °C. The mixture was stirred at room temperature for 2 h. To the mixture was added
dropwise a solution of compound 8 (120 mg, 0.280 mmol) in THF (2 mL) at 0 °C. The resulting
mixture was stirred at room temperature overnight. The reaction was quenched by addition of
0
.02 M hydrochloric acid and the mixture was extracted with AcOEt. The organic layer was
washed with aqueous NaHCO
3
solution and brine, dried over MgSO , and concentrated under
4
i
reduced pressure. The residue was crystallized from AcOEt– Pr
2
O to give the title compound as a
+
1
white solid (167 mg, 0.243 mmol, 87%). MS (ESI/APCI) m/z 687.3 [M + H] . H NMR (300
MHz, DMSO-d ) δ 1.53–1.74 (2H, m), 1.77–1.94 (2H, m), 2.01–2.17 (2H, m), 2.21–2.38 (2H,
m), 2.53–2.66 (2H, m), 2.70–2.86 (2H, m), 3.03–3.29 (4H, m), 3.72 (2H, t, J = 6.9 Hz), 4.02–
6
1
3
4
.18 (2H, m), 4.64 (1H, dt, J = 7.6, 3.6 Hz), 7.30–7.57 (4H, m), 10.04 (1H, s). C NMR (101
MHz, DMSO-d ) δ 18.3, 27.5, 30.1, 30.2, 49.0, 49.6, 50.7, 55.1 (t, J = 22.4 Hz), 70.3, 112.7 (t, J
16.9 Hz), 113.1–113.4 (m), 112.0–123.4 (m), 120.3, 124.3, 125.9, 132.7 (t, J = 13.2 Hz), 135.3,
6
=
2
9