Discovery of 3,5-Diphenyl-4-methyl-1,3-oxazolidin-2-ones as Novel, Potent, and Orally Available Δ-5 Desaturase (D5D) Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.7b01210

The discovery and optimization of Δ-5 desaturase (D5D) inhibitors are described. Investigation of the 1,3-oxazolidin-2-one scaffold was inspired by a pharmacophore model constructed from the common features of several hit compounds, resulting in the identification of 3,5-diphenyl-1,3-oxazolidin-2-one 5h as a novel lead showing potent in vitro activity. Subsequent optimization focused on the modification of two metabolic sites, which provided (4S,5S)-5i, a derivative with improved metabolic stability. Moreover, adding a substituent into the upper phenyl moiety further enhanced the intrinsic activity, which led to the discovery of 5-[(4S,5S)-5-(4fluorophenyl)-4-methyl-2-oxo-1,3-oxazolidin-3-yl]benzene-1,3-dicarbonitrile (4S,5S)-5n, endowed with excellent D5D binding affinity, cellular activity, and high oral bioavailability in a mouse. It exhibited robust in vivo hepatic arachidonic acid/dihomo-γ-linolenic acid ratio reduction (a target engagement marker) in an atherosclerosis mouse model. Finally, an asymmetric synthetic procedure for this compound was established.

Full text of DOI:10.1021/acs.jmedchem.7b01210

Subscriber access provided by LAURENTIAN UNIV
Article
Discovery of 3,5-Diphenyl-4-methyl-1,3-oxazolidin-2-ones as Novel,
Potent and Orally Available Delta-5 desaturase (D5D) Inhibitors
Jun Fujimoto, Rei Okamoto, Naoyoshi Noguchi, Ryoma Hara, Shinichi Masada,
Tetsuji Kawamoto, Hiroki Nagase, Yumiko Okano Tamura, Mitsuaki Imanishi, Shuichi
Takagahara, Kazuki Kubo, Kimio Tohyama, Koichi Iida, Tomohiro Andou, Ikuo
Miyahisa, Junji Matsui, Ryouta Hayashi, Tsuyoshi Maekawa, and Nobuyuki Matsunaga
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01210 • Publication Date (Web): 12 Oct 2017
Downloaded from http://pubs.acs.org on October 13, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of 3,5ꢀDiphenylꢀ4ꢀmethylꢀ1,3ꢀoxazolidinꢀ2ꢀones as Novel, Potent and Orally
Available Deltaꢀ5 desaturase (D5D) Inhibitors
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
‡
§
§
‡
Jun Fujimoto,* Rei Okamoto, Naoyoshi Noguchi, Ryoma Hara, Shinichi Masada, Tetsuji Kawamoto,
‡
ǁ
Hiroki Nagase, Yumiko Okano Tamura, Mitsuaki Imanishi, Shuichi Takagahara, Kazuki Kubo, Kimio
‡
‡
‡
Tohyama, Koichi Iida, Tomohiro Andou, Ikuo Miyahisa, Junji Matsui, Ryouta Hayashi, Tsuyoshi
§
ǁ
Maekawa and Nobuyuki Matsunaga
Pharmaceutical Research Division, Takeda Pharmaceutical Company Limited, 26ꢀ1, MuraokaꢀHigashi
2
ꢀchome, Fujisawa, Kanagawa, 251ꢀ8555, Japan
ABSTRACT
The discovery and optimization of deltaꢀ5 desaturase (D5D) inhibitors are described. Investigation of
the 1,3ꢀoxazolidinꢀ2ꢀone scaffold was inspired by a pharmacophore model constructed from the common
features of several hit compounds, resulting in the identification of 3,5ꢀdiphenylꢀ1,3ꢀoxazolidinꢀ2ꢀone 5h
as a novel lead showing potent in vitro activity. Subsequent optimization focused on the modification of
two metabolic sites, which provided (4S,5S)ꢀ5i, a derivative with improved metabolic stability. Moreover,
adding a substituent into the upper phenyl moiety further enhanced the intrinsic activity, which led to the
discovery of 5ꢀ[(4S,5S)ꢀ5ꢀ(4fluorophenyl)ꢀ4ꢀmethylꢀ2ꢀoxoꢀ1,3ꢀoxazolidinꢀ3ꢀyl]benzeneꢀ1,3ꢀdicarbonitrile
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(4S,5S)ꢀ5n, endowed with excellent D5D binding affinity, cellular activity and high oral bioavailability in
a mouse. It exhibited robust in vivo hepatic arachidonic acid/dihomoꢀγꢀlinolenic acid ratio reduction (a
target engagement marker) in an atherosclerosis mouse model. Finally, an asymmetric synthetic procedure
for this compound was established.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
ACS Paragon Plus Environment

Page 3 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Deltaꢀ5 desaturase (D5D) is a transꢀmembrane protein mainly expressed in the liver, and associated with
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
−3
polyꢀunsaturated fatty acid (PUFA) biosynthesis (Figure 1). In this pathway, an essential ωꢀ6 fatty acid,
linoleic acid, is desaturated at the 6ꢀposition by Deltaꢀ6 desaturase (D6D) to afford γꢀlinolenic acid, which
is converted by elongase to dihomoꢀγꢀlinolenic acid (DGLA). D5D is involved in the production of
arachidonic acid (AA) from DGLA by desaturation of the 5ꢀpostiion of DGLA. DGLA is known as a
precursor of antiꢀinflammatory eicosanoids, such as prostaglandin E1 (PGE1) and 15ꢀhydroxytrienoic
4
acid. In contrast, AA is a precursor of inflammatory unsaturated fatty acids, such as prostaglandins (PGs),
leukotrienes (LTs) and thromboxanes (TXs). Therefore, D5D inhibitors are expected to drive
antiꢀinflammatory properties through reduction of AA biosynthesis and simultaneous accumulation of
DGLA. This concept is supported by a report regarding D5Dꢀdeficient mice that displayed increases in
5
PGE1 and concurrent decreases in PGE2 levels compared to wildꢀtype mice.
So far, a few groups disclosed D5D inhibitors in literatures (Figure 2), but none have yet entered into
6
clinical trials. In 1998, Obukowicz et al. reported 4ꢀ(3ꢀmethylbutoxy)aniline (CPꢀ24879) with dual
7
D5D/D6D inhibition and 2ꢀaminoꢀNꢀ(3ꢀchlorophenyl)benzamide (CPꢀ74006) with selective D5D
inhibition, confirmed in rat liver microsomes and mastocytoma ABMCꢀ7 cells. Starting from the structure
of the latter, Lexicon Pharmaceuticals employed a cyclization approach, leading to the identification of
8
the indazole compound shown in Figure 2 that displayed in vivo D5D inhibition in mouse liver. Takeda
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
has also been interested in this target for its potential in antiꢀinflammatory drug development, and has
9
−11
disclosed fused and monocyclic pyrimidinones.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Herein, we describe our efforts to discover and optimize a novel series of D5D inhibitors. Starting
1
2
points were identified through a highꢀthroughput screening (HTS) campaign utilizing a binding assay,
which revealed a number of hit compounds. Among these, compounds 1−4 were highlighted due to their
possessing moderate to good potencies and low molecular weights (Figure 3). Moreover, we noticed their
structural similarities, which are characterized by aromatic rings at each end of the molecule and a spacer
containing a carbonyl group present in the form of an amide or urea linkage. Based on these features, we
13
constructed a pharmacophore model using Molecular Operating Environment (MOE) software as
illustrated in Figures 4a and 4b, and hypothesized that a novel scaffold for D5D inhibitors could be
discovered by the design of compounds matching this model. With this aim, two general structures,
1,3ꢀoxazolidinꢀ2ꢀones (5ꢀPh and 5ꢀBn in Figure 4c), were designed as compounds possessing two
aromatic rings linked by a central 1,3ꢀoxazolidinꢀ2ꢀone core. As shown in Figure 4d, both compounds fit
well with the 3ꢀpoint pharmacophore model, which encouraged us to embark on synthetic activity around
1,3ꢀoxazolidinꢀ2ꢀone derivatives. In this article, our medicinal chemistry efforts on the discovery of a
novel lead scaffold and further optimization to generate potent and orally available D5D inhibitors are
discussed.
4
ACS Paragon Plus Environment

Page 5 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ꢀ
CHEMISTRY
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The preparation of 3ꢀbenzylꢀ1,3ꢀoxazolidinꢀ2ꢀone 5a was carried out following the procedure shown in
Scheme 1. The oxazolidinꢀ2ꢀone compound 7 was formed by the reaction of 2ꢀaminoꢀ1ꢀphenylethanol (6)
and 1,1'ꢀcarbonyldiimidazole (CDI) in the presence of N,Nꢀdimethylꢀ4ꢀaminopyridine (DMAP). After
3
methanesulfonylation of 8 using methanesulfonyl chloride (MsCl) in the presence of triethyl amine (NEt ),
the mesylated intermediate was reacted with 7 to afford the target compound 5a. Scheme 2 illustrates the
synthesis of 5b−f. The Weinreb amide 10 was prepared from carboxylic acid 9, which was then treated
with 4ꢀfluorophenylmagnesium bromide to yield ketone 11. Compound 11 underwent reduction with
4
sodium borohydride (NaBH ) followed by oxazolidinꢀ2ꢀone ring formation by treatment with sodium
hydride (NaH) to afford the racemate 13. The 3ꢀbenzylꢀ1,3ꢀoxazolidinꢀ2ꢀone compound 5b was obtained
by the benzylation of 13 with 2,4ꢀdichlorobenzyl chloride. Intermediate 13 was also used for the
preparation of 3,5ꢀdiphenylꢀꢀ1,3ꢀoxazolidinꢀ2ꢀones 5c−f by a BuchwaldꢀHartwig coupling reaction. The
enantiomers (S)ꢀ5f and (R)ꢀ5f were obtained by chiral high performance liquid chromatography (HPLC)
separation of the racemic 5f. Compounds 5g and 5h were prepared as shown in Scheme 3. The ring
opening reaction of oxirane 14 with the appropriate anilines gave aminoethanols 15a and 15b, which
were converted to the corresponding target compounds 5g and 5h by 1,3ꢀoxazolidinꢀ2ꢀone formation
using CDI and DMAP.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The 1,3ꢀdiphenylꢀ4ꢀmethylꢀ1,3ꢀoxazolidinꢀ2ꢀone analogues 5i and 5j were obtained in 6 steps from the
commercially available starting material 16 (Scheme 4). Compound 16 was converted to phenyl ketone
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
18 via Weinreb amide formation followed by Grignard reaction. Subsequent reduction with NaBH and
removal of the tertꢀbutoxy carbonyl (Boc) protecting group under acidic conditions afforded the key
2ꢀaminoethanol intermediate 19 as a cis/trans mixture. According to the procedure shown in Scheme 2,
we first prepared the oxazolidinꢀ2ꢀone intermediate from 19 and attempted a BuchwaldꢀHartwig reaction,
however, the coupling reaction did not proceed, most likely due to steric hindrance of the 4ꢀmethyl group
of the oxazolidinꢀ2ꢀone. Thus, a Cu (I)ꢀcoupling reaction using 19 in 2ꢀpropanol was investigated
14
according to a reported procedure, and this approach provided the desired coupling product 20.
Formation of the 1,3ꢀoxazolizinꢀ2ꢀone ring from 20 afforded the product as a mixture of diastereomers,
which were separated into the each isomer 21 (trans) and 22 (cis) by silica gel column chromatography.
Cyano
derivatives
5i
and
5j
were
prepared
using
zinc
cyanide
and
3 4
tetrakis(triphenylphosphine)palladium(0) (Pd(PPh ) ) from the precursors 21 and 22, respectively. The
trans derivative 5i was subjected to chiral HPLC separation to yield enantiomers (4S,5S)ꢀ5i and
(
4R,5R)ꢀ5i.
Scheme 5 shows the synthetic procedure for 4ꢀgemꢀdimethylꢀ1,3ꢀoxazolidinꢀ2ꢀone derivative 5k. In a
manner analogous to Scheme 4, the 2ꢀaminoethanol intermediate 26 was obtained from carboxylic acid
23 in 3 steps. The following Cu(I)ꢀcoupling reaction proceeded even with the sterically hindered
6
ACS Paragon Plus Environment

Page 7 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
gemꢀdimethyl substrate 26 to provide the coupling product 27, which was converted into target compound
5k via oxazolidinꢀ2ꢀone ring formation and installment of a nitrile group using zinc cyanide and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3 4
Pd(PPh ) .
Preparation of 5l−n was conducted using key intermediate 19 as the starting compound (Scheme 6).
The Cu(I)ꢀmediated coupling of diastereomixture 19 and the appropriate aryl iodides followed by
1,3ꢀoxazolidinꢀ2ꢀone formation gave 30a and 30b as cis/trans mixtures, respectively. The bromo atom in
30a was converted to a cyano group by Pd(0)ꢀcatalyzed cyanation and following silicaꢀgel column
chromatography separation provided the trans 5l. In the case where chloro derivative 30b was used, the
cyanation reaction produced a mixture of both monoꢀ and diꢀcyano compounds, and their trans isomers
5
m and 5n were isolated by silica gel column chromatography. The resulting dicyano compound 5n was
further separated by chiral HPLC to yield (4S,5S)ꢀ5n. In addition, introduction of a cyano group from 30b
and chiral HPLC purification of the resulting diastereomeric mixture afforded the enantiomer
(4S,5S)ꢀ5m.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
Scheme 1. Synthesis of 3ꢀbenzylꢀ1,3ꢀoxazolidinꢀ2ꢀone 5a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
3
Reagents and conditions: (a) CDI, DMAP, THF, 80 °C, 46%; (b) (i) MsCl, NEt , THF, 0 °C; (ii) 7,
K
2
CO
3
, THF, 80 °C, then NaOH, THF, 0 °C, 2% in 2 steps.
a
Scheme 2. Synthesis of 5b−f
8
ACS Paragon Plus Environment

Page 9 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
3
Reagents and conditions: (a) N,Oꢀdimethylhydroxylamine hydrochloride, EDCI, HOBt, NEt , DMF,
0 °C to room temperature, 82%; (b) 4ꢀfluorophenylmagnesium bromide, THF, 0 °C to room temperature,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
87%; (c) NaBH , MeOH, 0 °C to room temperature, 98%; (d) NaH, THF, 0 °C to room temperature, 68%;
(
(
e) 2,4ꢀdichlorobenzyl chloride, NaH, tetrabutylammonium iodide, THF, 0 °C to room temperature, 95%;
f) ArBr or ArI, Pd
2
(dba)
3
, Xantphos, NaOtBu, toluene, 120 °C, microwave irradiation, 12−47%; (g)
chiral HPLC separation.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
Scheme 3. Synthesis of 5g and 5h
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
2
Reagents and conditions: (a) ArNH , 2ꢀpropanol, 90 °C; (b) CDI, DBU, DMF, 80 °C, 3−4% in 2 steps.
a
Scheme 4. Synthesis of 5i and 5j
a
Reagents and conditions: (a) N,Oꢀdimethylhydroxylamine hydrochloride, EDCI, HOBt, NEt , DMF,
3
room temperature, 81%; (b) 4ꢀfluorophenylmagnesium bromide, THF, 0 °C to room temperature, 76%;
(c) NaBH THF, MeOH, 0 °C, then HCl/EtOAc, room temperature, 38%; (d) 1ꢀbromoꢀ3ꢀiodobenzene,
4
,
10
ACS Paragon Plus Environment

Page 11 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CuI, NaOH, 2ꢀpropanol, 90 °C, 72%; (e) CDI, DBU, THF, 80 °C, 37% (21) and 27% (22); (f) zinc
cyanide, Pd(PPh
3
)
4
, DMF, 100 °C, 67% (5i) and 30% (5j); (g) chiral HPLC separation.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
Scheme 5. Synthesis of 5k
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
3
Reagents and conditions: (a) N,Oꢀdimethylhydroxylamine hydrochloride, EDCI, HOBt, NEt , DMF,
room temperature, 57%; (b) 4ꢀfluorophenylmagnesium bromide, THF, 0 °C to room temperature; (c)
NaBH4, THF, MeOH, 0 °C, then HCl/EtOAc, 0 °C to room temperature, 28% in 2 steps; (d)
1
ꢀbromoꢀ3ꢀiodobenzene, CuI, NaOH, 2ꢀpropanol, 90 °C; (e) CDI, DBU, THF, room temperature to 80 °C,
5
6% in 2 steps; (f) zinc cyanide, Pd(PPh
3
)
4
, DMF, 100 °C, 76%.
12
ACS Paragon Plus Environment

Page 13 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Scheme 6. Synthesis of 5l−n
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) ArI, CuI, NaOH, 2ꢀpropanol, room temperature to 100 °C; (b) CDI, DBU,
3 4
THF, 80 °C, 32% in 2 steps (30a) and 28% in 2 steps (30b) (c) (i) zinc cyanide, Pd(PPh ) , DMF, room
3 4
temperature, 100 °C (ii) chiral HPLC separation, 16% in 2 steps; (d) for 5l: 30a, zinc cyanide, Pd(PPh ) ,
DMF, room temperature to 100 °C, 49%; for 5m and 5n: 30b, zinc cyanide, Pd(PPh
temperature to 100 °C, 72% (5m) and 13% (5n); (e) chiral HPLC separation.
3 4
) , DMF, room
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ꢀ
RESULTS AND DISCUSSION
The compounds synthesized were evaluated for D5D binding activity in rat liver microsomes using
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
12
[ H]Tꢀ3364366 as a ligand of D5D. D5D inhibitory activities in HepG2 cells were also evaluated by
14
14
measuring the production of [ C]AA from [ C]DGLA. The in vitro activities are indicated as IC50
values.
The biological activities of representative screening hits and an initial set of compounds 5a, 5b, 5g and
5h are illustrated in Table 1. Screening hits 1 and 2 showed moderate to good D5D binding activity
sufficient for us to view them as reasonable starting points for drug discovery (IC50 39 and 360 nM,
respectively), and further enhancement of potency was sought from a novel scaffold based on a
1,3ꢀoxazolidinꢀ2ꢀone framework (5ꢀPh and 5ꢀBn in Figure 4c). In order to quickly assess the merit of our
design concept, 3ꢀbenzylꢀ1,3ꢀoxazolidinꢀ2ꢀone derivative 5a was first designed by installment of the
2,4ꢀdichloroꢀ5ꢀhydroxybenzyl moiety present in hit 1 into the 5ꢀBn scaffold. As expected, 5a displayed
3ꢀfold more potent D5D binding activity as compared to 1. Due to potential toxicity concerns the phenolic
1
5,16
group
of 5a was removed in the design of 5b, which showed somewhat moderated binding affinity
while cellular activity was enhanced, possibly due to improved cellular permeability. Encouraged by
these promising results, our attention was then directed to another 1,3ꢀoxazolidinꢀ2ꢀone series, the 5ꢀPh
derivatives, which also fit well with our pharmacophore model (Figure 4d). We took notice of the
2,5ꢀdimethoxy anilino moiety of hit compound 2, and introduced this fragment into the 5ꢀPh scaffold,
14
ACS Paragon Plus Environment

Page 15 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
resulting in the synthesis of 3ꢀphenylꢀ1,3ꢀoxazolidinꢀ2ꢀone derivative 5g. As we had hoped, both the
binding and cellular activities of 5g were high (IC50 13 and 18 nM, respectively), and further modification
showed that removal of the 2ꢀmethoxy substituent (5h) resulted in even more robust D5D inhibitory
activity, with singleꢀdigit nM IC50 values in both the binding and cellular assays. In addition to the
substantial enhancement in potency, 5g and 5h have an advantage in terms of lower lipophilicity,
resulting in higher ligandꢀlipophilic efficiency (LLE) values than the 5ꢀBn analogues (5a and 5b) and
other hit compounds. Therefore, 3,5ꢀdiphenylꢀ1,3ꢀoxazolidinꢀ2ꢀone 5h was selected as a lead compound
and proceeded to the next stage of study, in which we sought to further increase biological potency and
provide suitable pharmacokinetic (PK) parameters for use in oral administration.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The novel 3,5ꢀdiphenylꢀ1,3ꢀoxazolidinꢀ2ꢀone 5h showed potent activity, however, it also demonstrated
low metabolic stability in human and mouse liver microsomes (201 and 234 mL/min/mg, respectively) –
an important issue in our quest to discover orally active agents. In order to address this matter in a rational
way, metabolite analyses of 5h in human and mouse hepatocytes were conducted, which estimated phenol
Mꢀ1, Oꢀsulfate conjugate Mꢀ2 and Oꢀglucuronate conjugates (Mꢀ3 and Mꢀ4) as the apparent metabolites
of 5h (Figure 5). The analysis was also conducted in vivo, where five metabolites except Mꢀ3 (Figure 5)
were detected in liver after oral treatment of 5h as dietary admixture to mice. Both cases suggested that
the methoxy group and the 4ꢀmethylene of the 1,3ꢀoxazolinꢀ2ꢀone ring on the test compound
predominantly suffered from phase I oxidative metabolism and subsequent phase II conjugation reaction.
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Therefore, modification of these sites was explored with the goal of improving metabolic stability, which
is summarized in Table 2. First, removal of the metabolically labile methoxy group (5c) was examined,
which resulted in significantly reduced activity and showed the importance of retaining a substituent at
this position for substantial potency. Hence replacement of the methoxy with less metabolically
susceptible substituent, such as fluoro or chloro atom, was next conducted to provide compounds 5d and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5e, which resulted in partially recovered potency but insufficient improvement in microsomal stability.
Aiming further improvement of metabolic stability, we chosed a substituent that was not only less
metabolically labile but also polar, which resulted in the cyano derivative 5f showing the excellent
metabolic stability in human microsomes with moderate binding affinity. Comparing the stereoisomers of
5f revealed that the Sꢀenantiomer ((S)ꢀ5f) was the eutomer, showing relatively high D5D binding affinity
with an IC50 value of 49 nM, in clear contrast to the significantly less active Rꢀanalogue ((R)ꢀ5f). This
suggested that the configuration of the 5ꢀphenyl moiety might be crucial for gaining effective interaction
with the enzyme. Encouraged by the improved metabolic stability of 5f, we attempted to block another
metabolically labile position, the 4ꢀmethylene of the 1,3ꢀoxazolidinꢀ2ꢀone core. Initially, the
4ꢀgemꢀdimethyl derivative 5k was prepared for studying substituent effects on metabolism, which showed
expectedly high metabolic stability in both species supporting the validity of our drug design strategy.
Unfortunately, D5D binding affinity of 5k was more than 5ꢀtimes weaker than the corresponding
nonꢀmethylated compound 5f. We postulated that the dropꢀoff in the binding affinity was due to steric
16
ACS Paragon Plus Environment

Page 17 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
hindrance at the 4,4ꢀdimethylmethylene group, which could substantially alter the conformational rigidity
of the 5ꢀphenyl moiety that was deemed important for D5D binding. In order to test this assumption, the
energetic effects of altering the dihedral angle between the central 1,3ꢀoxazolidinꢀ2ꢀone and the 5ꢀphenyl
group were examined computationally using MOE. Two simplified compounds (a) and (b) shown in
Figure 6 were built as (S)ꢀenantiomers, and the variation in potential energy vs. dihedral angle was
calculated for each compound. The resulting calculations for compound (a), which has no substituents at
the 4ꢀposition, showed an energy minimum for a dihedral angle between 90 and 135° (potential energy ≤
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.5 kcal/mol). On the other hand, compound (b), which is a model of the gemꢀdimethyl analogue,
displayed an energy minimum with a narrower range of dihedral angles (95−115°). We concluded that
this difference in conformational rigidity is likely to result in a greater energy gap between the stable
conformer in solution and the bound, bioactive conformation for gemꢀdimethyl compound 5k, leading to
a decrease in potency. Although conclusive information regarding the bioactive conformation of our D5D
inhibitor was not available due to the difficulty in obtaining crystal structures of D5D to date, we
hypothesized that a substituent could be installed into the 4ꢀposition in a manner that would give rise to a
conformation similar to model (a), thus retaining potent D5D binding affinity. Hence 4ꢀmonomethyl
compounds (c) and (d) were examined by dihedral angle calculations for the models of trans and cis
derivatives, respectively. As a result, the trans model (c) showed a similar dihedral angle plot as (a),
while cis compound (d) provided results similar to the lessꢀactive gemꢀdimethyl compound (b). This
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
encouraged us to test the trans diastereomer with the expectation of retaining high binding affinity for
D5D. The trans derivative 5i was first prepared as a racemate, which showed more potent binding activity
(IC50 13 nM) than 5f. Moreover, the active enantiomer of 5i ((4S,5S)ꢀ5i) displayed higher potencies in
both binding and cellular assays, reaching IC50 values of less than 10 nM, which are more than 6ꢀfold
more potent than the corresponding derivative (S)ꢀ5f. The high potency of (4S,5S)ꢀ5i is possibly
explained by our prediction that the (4S)ꢀmethyl group could minimize the conformational strain/energy
gap between the lowꢀenergy and bioactive conformations (as discussed in Figure 6) and by additional
lipophilic interaction between this methyl group and the protein. In contrast, the binding affinity of cis
derivative 5j decreased relative to 5f, which is consistent with the calculations for model compounds (b)
and (c) (Figure 6).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In terms of metabolic stability, (4S,5S)ꢀ5i displayed excellent microsomal stability (0 mL/min/mg in
human and 24 mL/min/mg in mouse), which we reasoned is due to steric blockage of this position from
metabolic enzymes.
The high metabolic stability of compound (4S,5S)ꢀ5i was reflected in its excellent bioavailability, as
found in a mouse cassette dosing study (F = 87.2%). Encouraged by the promising PK profile, in vivo
D5D inhibition of (4S,5S)ꢀ5i was evaluated. The test compound was administered orally to C57BL/6J
mice for 4 days, and AA and DGLA content of liver samples were measured. The decrease rate of
AA/DGLA values relative to that of the control group as 100% was used as an index for assessing in vivo
18
ACS Paragon Plus Environment

Page 19 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
D5D inhibition of the test compound. In the course of previous studies of other chemotypes, we have
learned that the maximum level of in vivo D5D inhibition is projected to about 70% of the in vivo D5D
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
7
inhibition index after 4 days of treatment. Hence, we set our goal to achieve the same level of in vivo
efficacy with our novel 3,5ꢀdiphenylꢀ1,3ꢀoxazolidinꢀ2ꢀone derivatives. However, when (4S,5S)ꢀ5i was
administered orally at a dose of 3 mg/kg twice a day for 4 days, the in vivo D5D inhibition index was
considerably lower than our goal (42%, Table 3). In order to demonstrate higher in vivo effects, we
concluded that further enhancement of in vitro D5D inhibitory activity would be needed, leading us to
examine the effect of additional substituents on potency. After initial attempts to introduce a substituent
into the benzonitrile moiety, adding a substituent at the 5ꢀposition was found to be the most effective in
enhancing potency, as summarized in Table 3. For mitigating the potential for oxidative metabolism, the
introduction of a small and electronꢀwithdrawing substituent (fluoro atom) was first attempted (5l),
resulting in more than 2ꢀfold enhancement for both binding and cellular potencies compared to 5i in Table
2. As expected, compound 5l displayed higher in vivo D5D inhibition (63%) than (4S,5S)ꢀ5i when dosed
at 3 mg/kg twice a day for 4 days. Replacement of the fluoro with a chloro atom was then examined, and
the enantiomer (4S,5S)ꢀ5m exhibited robust in vitro activities with subꢀnM IC50 values (< 1.0 nM for
binding and 0.52 nM in cellular assay). For evaluating the in vivo effects of this compound, onceꢀdaily
dosing was tested, which demonstrated a maximum 71% in vivo D5D inhibition index. Although this
chloro derivative showed moderate CYP2C9 inhibition at 10 ꢀM (44.2%), exchanging the chloro atom
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
for a polar nitrile group ((4S,5S)ꢀ5n) was attempted, which showed excellent in vitro activities (IC50
=
1.5 nM in binding and cell assays) and a high in vivo D5D inhibition index (74%, 3 mg/kg, po) with
reduced CYP2C9 inhibition (28.5% at 10 ꢀM). As expected, enhancement of in vitro activities of the
compounds discussed in Table 3 resulted in increase of in vivo D5D inhibition effects. In addition, we
have found in our research of previous chemotypes that a durable plasma drug concentration would be
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
7
expected to positively impact on in vivo D5D inhibitory action. The plasma meanꢀresidenceꢀtime
(
MRT) values of 5l (3.8 h), (4S,5S)ꢀ5m (4.2 h) and (4S,5S)ꢀ5n (4.0 h) in a mouse cassette dosing study
were longer than for (4S,5S)ꢀ5i (3.4 h), which might result in more durable drug exposure to the target
organ (liver) as well as plasma, and may be attributable to their robust in vivo D5D inhibitory effects.
Among the compounds as discussed above, the dicyano derivative (4S,5S)ꢀ5n was identified as a
promising compound with high in vitro potency and robust in vivo D5D inhibitory effects. For further
biological evaluation, this compound was administered onceꢀdaily for 2 weeks to Western dietꢀfed
apolipoprotein E knockout (apoE KO) mice as an atherosclerosis model. As a result, 61%, 67% and 66%
in vivo D5D index reduction was achieved at 0.3, 1 and 3 mg/kg dosing, respectively (Figure 7a). Under
both single and repeated administrations, plasma drug concentrations increased in a doseꢀdependent
manner, and accumulation occurred in the repeatꢀdose regimen (Figure 7b).
Further profiling of (4S,5S)ꢀ5n showed no significant drugꢀdrug interaction risk with respect to
inhibition of CYP isoforms (3A4, 2C8, 2C9 and 2D6) and CYP3A induction. Moderate human
20
ACS Paragon Plus Environment

Page 21 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
etherꢀaꢀgoꢀgo related gene (hERG) inhibitory activity (30.7% inhibition at 10 ꢀM in a manual patch
clamp assay) was found to be a concern in terms of the potential for QT prolongation. In order to assess
the hERG inhibitory effect in vivo, an electrocardiogram (ECG) study using dogs was conducted under
anesthesia, which showed essentially no adverse cardiovascular effects at a dose of 5 mg/kg by
intravenous infusion. The safety margin was calculated from the maximum drug concentration in an ECG
study (7.15 ꢀg/mL at 5 mg/kg dose) and that in the apoE KO mouse study (0.28 ꢀg/mL at 1 mg/kg dose
in the repeated treatment), in which the compound showed the maximum level of in vivo D5D inhibition
index (67%, Figures 7a and 7b). As a result, the safety margin was calculated to be more than 25ꢀfold,
which supported our conclusion that the risk level of this compound for QT prolongation would be low.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The compound thus identified, (4S,5S)ꢀ5n, demonstrated a promising profile as a novel and orally
available D5D inhibitor, which encouraged us to develop an asymmetric synthetic procedure for this
compound. The first procedure we established is shown in Scheme 7. Starting from
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1ꢀ(4ꢀfluorophenyl)propanꢀ1ꢀone (31), bromination at the αꢀposition of the carbonyl group yielded 32,
which underwent a substitution reaction by dibenzylamine to afford 33. Reduction of the carbonyl group
of 33 by LAH predominantly produced (1RS,2RS)ꢀcompound 34 in high yield, probably due to the steric
hindrance of the dibenzyl group. Removal of the dibenzyl protecting group was carried out by
hydrogenation and following Bocꢀprotection of the amine moiety enabled application of a chiral HPLC
separation, which afforded (1S,2S)ꢀenantiomer 36 in more than 99% enantiomeric excess (ee). The Boc
group of 36 was removed under acidic conditions and the key chiral 2ꢀaminoethanol intermediate was
obtained as an HCl salt (37ꢀHCl). Using 37ꢀHCl, Cu(I)ꢀcatalyzed coupling with
1,3ꢀdibromoꢀ5ꢀiodobenzene was conducted to provide 38, which underwent 1,3ꢀoxazolidinꢀ2ꢀone
formation with CDI and DBU under conventional conditions (39). Conversion of the both bromo atoms
of 39 into nitrile groups by Pd(0)ꢀmediated coupling with zinc cyanide led to the target compound
(
4S,5S)ꢀ5n, which showed greater than 99% ee after recrystallization from EtOAc. The absolute
configuration of the chiral center of the obtained product was confirmed by singleꢀcrystal Xꢀray
diffraction analysis (Figure 8).
22
ACS Paragon Plus Environment

Page 23 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Finally, an asymmetric synthetic procedure for key chiral intermediate 37 was established that avoided
chiral HPLC separation (Scheme 8). Using commercially available BocꢀLꢀalanine (40) as a starting
material, Weinreb amide formation was conducted with N,Oꢀdimethylhydroxylamine hydrochloride to
provide 41, and subsequent reaction with 4ꢀfluorophenylmagnesium bromide afforded ketone 42.
Reduction of the carbonyl group to a hydroxyl group was conducted under MeerweinꢀPonndorfꢀVerley
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
conditions using Al(OiPr) and excess 2ꢀpropanol, which proceeds via a sixꢀmembered transition state by
1
8
coordination of the Lewis acidic aluminium and carbonyl group and afforded (1R,2S)ꢀisomer 43 with
greater than 99% diastereomeric excess (de) after recrystallization from EtOAc/hexane. For the
preparation of key intermediate (1S,2S)ꢀ2ꢀaminoꢀ1ꢀ(4ꢀfluorophenyl)propanꢀ1ꢀol (37), inversion at the
hydroxyl group of 43 by Mitsunobu reaction with 4ꢀnitrobenzoic acid was first attempted, which provided
the corresponding 4ꢀnitrobenzoate. The ester was then hydrolyzed, however, the analytical data of the
compound thus obtained was identical with starting material 43, not the desired (1S,2S)ꢀproduct. This
1
9
could be due to intramolecular aziridine formation and subsequent ringꢀopening caused by
4ꢀnitrobenzoic acid under the Mitsunobu conditions, which would give the (1R,2S)ꢀcompound with
retention of stereochemistry. Thus, as an alternative method, treatment of 43 with methanesulfonyl
chloride was attempted, which resulted in the formation of oxazolizinꢀ2ꢀone 44 in a high yield by an
2
0
N
intramolecular S 2 reaction leading to inversion of stereochemistry. Attempts at hydrolysis of 44 with
aqueous lithium hydroxide (LiOH) proved to be successful and to afford the key 2ꢀaminoethanol
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
intermediate 37 with the desired (1S,2S)ꢀconfiguration in > 99% ee, which can be led to compound
(4S,5S)ꢀ5n in a manner similar to that described in Scheme 7.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ
CONCLUSION
In this article, we have presented the discovery of novel D5D inhibitors through HTS, lead generation and
optimization studies. The common structural features of the initial hits inspired us to design a novel
scaffold, which resulted in the identification of 3,5ꢀdiphenylꢀ1,3ꢀoxazolidinꢀ2ꢀones as a new lead series
with high potency and LLE values. Subsequent lead optimization efforts first focused on the improvement
of microsomal stability, and metabolite identification studies of one derivative guided us toward
replacement of a metabolically labile substituent (a methoxy group) and steric blockage of another
metabolically susceptible site (the 4ꢀposition of the oxazolidinꢀ2ꢀone ring). In an attempt to further
enhance in vitro activity with the aim of achieving higher in vivo effects, the installment of a substituent
into the benzonitrile portion was explored, which contributed to potency enhancement and prolonged
drug concentration in plasma upon oral treatment. Among the derivatives synthesized, compound
(
4S,5S)ꢀ5n exhibited potent in vitro activity with IC50 value of 1.5 nM, and a 67% in vivo D5D inhibition
index at a dose of 1 mg/kg, po in apoE KO mice after 2ꢀweek treatment. These results led us to conclude
that these 3,5ꢀdiphenylꢀ1,3ꢀoxazolidinꢀ2ꢀone derivatives represent a novel and promising series for further
studies of D5D inhibition.
24
ACS Paragon Plus Environment

Page 25 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Scheme 7. Asymmetric synthesis of (4S,5S)ꢀ5n
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
2 2 2 3
Reagents and conditions: (a) Br , Et O, 0 °C to room temperature, 95%; (b) dibenzylamine, Na CO ,
2 2
acetonitrile, 60 °C, 89%; (c) LAH, Et O, room temperature, 81%; (d) H , Pd/C, MeOH, room temperature,
9
2 3
3%; (e) Boc O, NEt , THF, 0 °C to room temperature, 91%; (f) chiral HPLC separation; (g) HCl/EtOAc,
0
°C to room temperature; (h) 1,3ꢀdibromoꢀ5ꢀiodobenzene, CuI, NaOH, 2ꢀpropanol, room temperature to
90 °C, 49%; (i) CDI, DBU, THF, 100 °C, 89%; (j) zinc cyanide, Pd(PPh
3 4
) , DMF, room temperature to
100 °C, 43%.
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
Scheme 8. Asymmetric synthesis of 37
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: (a) N,Oꢀdimethylhydroxylamine hydrochloride, EDCI, HOBt, NEt , DMF,
3
room temperature, 88%; (b) 4ꢀfluorophenylmagnesium bromide, THF, room temperature, quant.; (c)
3 3
Al(OiPr) , 2ꢀpropanol, toluene, 50 °C, then recrystallization, 67%; (d) MsCl, NEt , THF, 50 °C, 87%; (e)
LiOH, water, 90 °C, 69%.
26
ACS Paragon Plus Environment

Page 27 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ꢀ
EXPERIMENTAL SECTION
1
General. H NMR spectra were obtained at 300 MHz on a BRUKERꢀ300 or a Varian VNMRSꢀ400 or
400 MHz on a Bruker Avance 400 MHz. Chemical shifts are given as δ values (ppm) using
tetramethylsilane as an internal standard. The following abbreviations as used to indicate multiplicities: s
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
=
singlet, d = doublet, t = triplet, q = quartet, m = multiplet, br s = broad singlet. Coupling constants (J
values) are given in hertz (Hz). Reactions were followed by TLC on Silica gel 60 F254 precoated TLC
plates (E. Merck) or NH TLC plates (Fuji Silysia Chemical Ltd.). Chromatographic separations were
carried out on PurifꢀPack silica columns (Si or NH, Shoko Scientific) or Universal silica columns (Si or
NH, Yamazen Corp.) using the indicated eluents. Preparative HPLC was performed on a Waters 2525
separations module (Lꢀcolumn2 ODS (20 × 150 mm ID), CERI, Japan). Element analyses were carried
+
out by Takeda Analytical Laboratories or Sumika Chemical Analysis Service, Ltd. LCꢀMS (ESI )
analyses were performed on a Waters ZQꢀ2000, a Agilentꢀ1200, a Shimazu LCꢀ20AD or a Shimadzu
LCMS 2010 with acetonitrile/water mobile phase. Microwave reactions were performed using a Biotage
Initiator microwave reactor (Biotage Japan Ltd.). Purity data were collected by a HPLC with Corona
CAD (Charged Aerosol Detector), Nano Quantity Analyte Detector (NQAD), or photo diode array
detector. The column used was a Capcell Pak C18AQ (50 mm × 3.0 mm ID, Shiseido, Japan) or
Lꢀcolumn 2 ODS (30 mm × 2.0 mm ID, CERI, Japan) with a temperature of 50 °C and a flow rate of 0.5
mL/min. Mobile phases A and B under a neutral condition were a mixture of 50 mM ammonium acetate,
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 79
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
H
2
O, and CH
3
3
CN (1:8:1, v/v/v) and a mixture of 50 mM ammonium acetate and CH CN (1:9, v/v),
respectively. The ratio of mobile phase B was increased linearly from 5% to 95% over 3 min, 95% over
the next 1 min. Mobile phases A and B under acidic conditions were a mixture of 0.2% formic acid in 10
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
mM ammonium formate and 0.2% formic acid in CH CN, respectively. The ratio of mobile phase B was
increased linearly from 14% to 86% over 3 min, 86% over the next 1 min. The purities of all test
compounds were assessed by analytical HPLC (> 95%).
3ꢀ(2,4ꢀDichloroꢀ5ꢀhydroxybenzyl)ꢀ5ꢀphenylꢀ1,3ꢀoxazolidinꢀ2ꢀone (5a). To
a
solution of
2,4ꢀdichloroꢀ5ꢀ(hydroxymethyl)phenol (8, 5.0 g, 25.9 mmol) in THF (200 mL) and NEt
3
(10.8 mL, 77.7
mmol) was added MsCl (4.0 mL, 51.8 mmol) at 0 °C. The mixture was stirred at 0 °C for 1 h. The
4
mixture was quenched with sat. aq. NH Cl at 0 °C and extracted with EtOAc (×2). The organic layer was
4
separated, washed with brine, dried over MgSO and concentrated under vacuum. This product (6.5 g)
2 3
and K CO (7.7 g, 55.5 mmol) were added to a solution of 7 (3.0 g, 18.5 mmol) in THF (160 mL) at room
4
temperature. The mixture was stirred at 80 °C overnight. The mixture was quenched with sat. aq. NH Cl
at 0 °C and extracted with EtOAc (×2). The organic layer was separated, washed with brine, dried over
4
MgSO and concentrated under vacuum. The residue was purified by column chromatography (silica gel,
eluted with 20%−70% EtOAc/hexane). The obtained product (5.7 g) was dissolved with THF (30 mL).
To the mixture was added 1 N aq. NaOH (41.0 mL, 41.0 mmol) at 0 °C. The mixture was stirred at 0 °C
for 30 min. The mixture was neutralized with 1 N aq. HCl at 0 °C and extracted with EtOAc. The organic
28
ACS Paragon Plus Environment

Page 29 of 79
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
layer was separated, washed with water and brine, dried over MgSO and concentrated under vacuum.
The residue was purified by column chromatography (silica gel, eluted with 20%−70% EtOAc/hexane) to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
yield 5a (0.20 g, 2%) as a white solid. H NMR (300 MHz, DMSOꢀd ):
δ 3.32−3.39 (m, 1H), 3.90 (t, J =
6
8
1
3
.9 Hz, 1H), 4.27−4.54 (m, 2H), 5.59 (t, J = 8.1 Hz, 1H), 7.02 (s, 1H), 7.31−7.45 (m, 5H), 7.50 (s, 1H),
2 3
0.58 (br s, 1H). Anal. Calcd for C16H13Cl NO : C, 56.82; H, 3.87; N, 4.14. Found: C, 56.50; H, 4.09; N,
+
.89. MS (ESI) m/z 338.0 [M + H] . HPLC purity 100%.
3
ꢀ(2,4ꢀDichlorobenzyl)ꢀ5ꢀ(4ꢀfluorophenyl)ꢀ1,3ꢀoxazolidinꢀ2ꢀone (5b). To a solution of 13 (320 mg,
1
.8 mmol) in THF (25 mL) was added NaH (60% oil dispersion, 92 mg, 2.3 mmol) at 0 °C. After being
stirred at room temperature for 5 min, 2,4ꢀdichlorobenzyl chloride (0.37 mL, 2.7 mmol) and
tetrabutylammonium iodide (65 mg, 0.18 mmol) were added to the reaction mixture. The mixture was
4
stirred at room temperature overnight. The mixture was quenched with sat. aq. NH Cl at 0 °C and
extracted with EtOAc (×3). The organic layer was separated, washed with 1 N aq. HCl and brine, dried
over MgSO
4
and concentrated under vacuum. The residue was purified by column chromatography (silica
1
gel, eluted with 5%−50% EtOAc/hexane) to yield 5b (570 mg, 95%) as a colorless oil. H NMR (300
MHz, CDCl ):
δ 3.34 (dd, J = 8.7, 7.5 Hz, 1H), 3.69−3.93 (m, 1H), 4.34−4.74 (m, 2H), 5.47 (t, J = 8.1 Hz,
3
1
2 2
H), 6.89−7.16 (m, 2H), 7.18−7.50 (m, 5H). Anal. Calcd for C16H12Cl FNO : C, 56.49; H, 3.56; N, 4.12.
+
Found: C, 56.44; H, 3.69; N, 4.13. MS (ESI) m/z 340.0 [M + H] . HPLC purity 99.2%.
5
ꢀ(4ꢀFluorophenyl)ꢀ3ꢀphenylꢀ1,3ꢀoxazolidinꢀ2ꢀone (5c). To a solution of 13 (91 mg, 0.50 mmol),
29
ACS Paragon Plus Environment