Discovery and structure-guided optimization of tert-butyl 6-(phenoxymethyl)-3-(trifluoromethyl)benzoates as liver X receptor agonists

Article abstract of DOI:10.1016/j.bmcl.2015.07.047

Abstract To obtain potent liver X receptor (LXR) agonists, a structure-activity relationship study was performed on a series of tert-butyl benzoate analogs. As the crystal structure analysis suggested applicable interactions between the LXR ligand-binding

Full text of DOI:10.1016/j.bmcl.2015.07.047

Bioorganic & Medicinal Chemistry Letters 25 (2015) 3914–3920
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery and structure-guided optimization of tert-butyl
6-(phenoxymethyl)-3-(trifluoromethyl)benzoates as liver
X receptor agonists
b
b
b
b
Yumi Matsui a, , Takahiro Yamaguchi , Takanori Yamazaki , Masayuki Yoshida , Masami Arai ,
⇑
Naoki Terasaka b, Shoko Honzumi b, Kenji Wakabayashi a, Shinko Hayashi b, Daisuke Nakai b,
Hiroyuki Hanzawa a, Kazuhiko Tamaki b,
⇑
a Daiichi Sankyo RD Novare Co., Ltd, 1-16-13, Kita-Kasai, Edogawa-ku, Tokyo 134-8630, Japan
b R&D Division, Daiichi Sankyo Co., Ltd, 1-2-58 Hiromachi, Shinagawa-ku, Tokyo 140-8710, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 31 March 2015
Revised 7 July 2015
Accepted 16 July 2015
Available online 23 July 2015
To obtain potent liver X receptor (LXR) agonists, a structure–activity relationship study was performed on
a series of tert-butyl benzoate analogs. As the crystal structure analysis suggested applicable interactions
between the LXR ligand-binding domain and the ligands, two key functional groups were introduced. The
introduction of the hydroxyl group on the C6-position of the benzoate part enhanced the agonistic activ-
ity in a cell-based assay, and the carboxyl group in terminal improved the pharmacokinetic profile in
mice, respectively. The obtained compound 32b increased blood ABCA1 mRNA expression without
plasma TG elevation in both mice and cynomolgus monkeys.
Keywords:
Liver X receptors
LXR agonists
Ó 2015 Elsevier Ltd. All rights reserved.
Atherosclerosis
tert-Butyl benzoate
Structure-based drug design
The liver X receptor
a
and b (LXR
a
and b) are members of the
is mainly expressed
biosynthesis, including fatty acid synthase (FAS) and stearoyl CoA
desaturase-1 (SCD-1).9 For therapeutic use, the agonists which
nuclear hormone receptor superfamily.1 LXR
a
in the liver, adipose tissue, and macrophage, whereas LXRb is
expressed ubiquitously. LXRs regulate lipid metabolism and
reverse cholesterol transport by controlling the gene expression
of ApoE2 and ATP-binding cassette transporters A1 and G1
(ABCA1 and ABCG1),3 which play an important role in cholesterol
efflux from adipocyte and macrophage.4 Various research shows
have a wide therapeutic window between the induction of
ABCA1 and TG elevation are desirable.
A mammalian two-hybrid assay using LXRb and SRC-1 was
developed for a high-throughput screening (HTS), which identified
a novel N-Boc-indole analog 3 as a ligand to LXRs (Fig. 2).
Separately from the chemical modification of 3 with the N-Boc-
indole core intact, early chemical modification of the indole ring
in 3 led to 4, a potent binder to LXRs (Fig. 2). On the other hand,
4 showed only moderate agonist efficacy in a cell-based assay
(Table 1). In addition, the plasma exposure level of 4 after oral
administration was fairly low, and 4 did not sufficiently show a
desired in vivo biological activity in mice.
In this letter, we describe the chemical modification of 4 to
enhance the transcriptional activation in a cell-based assay, and
to change the physicochemical property with the intent to improve
the in vivo pharmacokinetic (PK) behavior, using a structure-
guided drug design.
liver
X receptors (LXRs) as potential therapeutic targets for
atherosclerosis by enhancing the cholesterol efflux promotion
from macrophages.5
The synthetic LXR agonists, such as T0901317 (1) and GW3965
(2) (Fig. 1, Table 1), have been described as increasing the ABCA1
expression and reducing atherosclerotic plaque in a LDLRꢀ/ꢀ
mouse model.6,7 However, these compounds also increase the
plasma and the liver triglyceride (TG) levels by induction of the
sterol regulatory element-binding protein-1c (SREBP-1c).8 SREBP-
1c activates the transcription of genes involved in fatty acid
In an effort to find a key to enhance the efficacy in the cell-
⇑
Corresponding authors. Tel.: +81 3 5696 8264; fax: +81 3 5696 4768 (Y.M.);
tel.: +81 3 3492 3131; fax: +81 3 5436 8563 (K.T.).
based assay, the X-ray crystal structure of a human LXR
a ligand-
E-mail addresses: matsui.yumi.gk@rdn.daiichisankyo.co.jp (Y. Matsui),
tamaki.kazuhiko.s2@daiichisankyo.co.jp (K. Tamaki).
binding domain (LBD) in complex with 4 and a peptide derived
http://dx.doi.org/10.1016/j.bmcl.2015.07.047
0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.

Y. Matsui et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3914–3920
3915
F3C
O
OH
CF3
CO2H
O
O
N
S
N
Cl
CF3
CF3
GW-3965 (2)
T0901317 (1)
Figure 1. Chemical structures of LXR agonists.
F3C
O
O
O
N
O
N
H
O
O
O
O
N
3
4
Ki (µM) LXRα:n.a.
LXRβ: 0.47
Ki (µM) LXRα:0.87
LXRβ:0.18
Figure 2. HTS hit compound and its derivative.
Table 1
Profiles of 1, 2, 4 and the corresponding C6-hydroxy analog 12
F3C
O
O
R6
O
O
N
a
b
FP assay
Ki ( M)
CTF assay
l
EC50
(l
M) (% efficacy)
Compound
R6
LXR
a
LXRb
LXR
a
LXRb
Figure 3. Crystal structure of LXR
structure of LXR
homodimer in the crystal. LXR
4 is shown as a space-filling representation. AF2 region (residues 437–447) of
LXR -LBD are colored in yellow, other parts of LXR -LBD in gray, and SRC-1 peptide
in green. Carbon atoms of compound 4 in cyan, oxygen in red, nitrogen in blue, and
fluoride in light blue. (B) Compound 4 in the ligand binding site of LXR -LBD.
Compound 4, and the side chains of His421 and Trp443 are shown as a stick
representation. Other residues included in binding are shown as line
a
-LBD in complex with 4. (A) The overall
a
-LBD in complex with 4 and SRC-1 peptide. LXR
a-LBD exists as a
1
2
4
12
0.23
0.21
0.87
0.27
0.22
0.16
0.18
0.17
0.03 (100)
0.54 (40)
0.31 (27)
0.28 (89)
0.13 (100)
0.10 (72)
0.21 (28)
0.26 (76)
a
-LBD and SRC-1 peptide are depicted as ribbons, and
H
OH
a
a
a
FP assay: 1 constantly shows 0.20–0.30
CTF assay: efficacy was shown as a relative value based on 1 (100%).
l
M of the Ki value toward both LXRs.
a
b
4
a
representation. The carbon atoms of His421 and Trp443 are colored in magenta
and yellow, respectively. The color scheme of 4 is the same as used in A.
from steroid receptor coactivator-1 (SRC-1 peptide) was deter-
mined.10 LXR
a-LBD adopts the three-layered a-helical sandwich
structure seen in a typical nuclear receptor LBD fold11 (Fig. 3A).
The C-terminal AF-2 helix is packed against the ligand binding
pocket, providing the binding site for SRC-1 peptide. The ligand-
binding pocket in which 4 binds is a large cavity, that is, sur-
rounded by Helix-3, Helix-5, Helix-6, Helix-7, Helix-11, and
mean square deviation of 0.9 Å over the backbone atoms of 221
residues. Thus, the interactions between 4 and LXRa-LBD would
be maintained in LXRb-LBD with 4.
Although the published crystal structures of LXRs with agonists
have indicated the importance of hydrogen bonding or electro-
static interaction between the agonist and His435 in LXRb (corre-
Helix-12. Intermolecular interaction between 4 and the LXRa-
a)
on Helix10/11,12 our crystal
LBD are dominated by hydrophobic contacts. As Figure 3B shows,
the trifluoromethyl benzyl moiety of 4 binds deep in the pocket,
forming van der Waals contacts with the residues Phe257,
Thr258, Leu331, His421, Gln424, Leu428, Leu435, and Trp443.
The tert-butyl acetate moiety occupies the hydrophobic part which
consists of Ile295, Met298, Leu299, Thr302, Phe315, Phe326,
Phe335, and Ile339. The right part benzene ring and the N-methy-
lacetamide substituent contact Leu260, Ser264, Met298, Glu301,
Thr302, Arg305, and Phe315, facing toward the open space sur-
sponding to His421 in LXR
structure has revealed that 4 has only van der Waals interaction
with His421 in LXR -LBD. For example, Williams et al. reported
a
that 1 formed a hydrogen bond with and locked His435 of LXRb
in the position to make an edge to face interaction with Trp457
of LXRb on the AF2 helix, stabilizing LXRb-LBD in the active confor-
mation (His-Trp switch).12 Compound 2 also formed the electro-
static interaction by the trifluoromethyl group with the Ne2
atom of His435 in LXRb-LBD with 3.3 Å distance (PDB ID:
1PQ6).13 Unlike these reported agonists, 4 does not have a hydro-
rounded by Helix3, Helix5, and an antiparallel b-sheet. LXRa-LBD
(residues from 204 to 447) shares 78% sequence identity with
LXRb-LBD (residues from 218 to 461), and these residues of the
gen bond or an electrostatic interaction to His421 of LXRa-LBD,
but contacts with the Cd2 atom of His421 by the carbonyl oxygen
in the tert-butyl benzoate moiety with 3.9 Å distance and by the
fluoride atom on the trifluoromethyl group with 4.0 Å distance.
Presuming that the van der Waals interactions of 4 with His421
LXR
the overall structure of LXR
LBD, as superimposing of the 4 bound LXR
holesterol bound LXRb-LBD (PDB ID: 1P8D)12 exhibits a small root
a
ligand-binding pocket are conserved in LXRb. Additionally,
-LBD is very similar to that of LXRb-
-LBD on the epoxyc-
a
a
of LXRa-LBD were not sufficient to lock His421 in the proper

3916
Y. Matsui et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3914–3920
F C
F C
3
F C
3
3
HO
Br
CO tBu
a, b
c
O
O
O
COCl
2
CO tBu
N
2
N
5
4
7
6
F C
3
F C
3
F C
3
f, g
d, e
h
OH
MOMO
MOMO
MeO OMe
10
HO
CHO
MeO
OMe
9
8
F C
F C
3
3
i, j, k
O
O
MOMO
O
HO
O
CO tBu
2
MeO
OMe
11
N
N
12
Scheme 1. Reagents and conditions: (a) t-BuOK, THF, 0 °C (81%); (b) NBS, AIBN,
PhH, reflux (53%); (c) 7, Cs2CO3, DMF (97%); (d) (MeO)3CH, CSA, MeOH, 50 °C; (e)
MOMCl, i-Pr2NEt, CH2Cl2, 0 °C–rt (93%, 2 steps); (f) n-BuLi, TMEDA, Et2O, ꢀ30 °C–rt
then DMF at ꢀ30 °C; (g) NaBH4, THF–MeOH, 0 °C–rt (52%, 2 steps); (h) 7, TMAD, n-
Bu3P, THF (51%); (i) 4 N HCl–THF, 50 °C; (j) NaClO2, NaH2PO4, 2-methyl-2-butene,
H2O–t-BuOH; (k) Me2NCH(OtBu)2, toluene, reflux (40%, 3 steps). TMEDA = N,N,N0,N0-
tetramethylethylenediamine; TMAD = N,N,N0,N0-tetramethylazodicarboxamide.
Figure 4. Overlay of 2 in LXRb-LBD (PDB ID: 1PQ6) on 4 in LXR
and 4 are shown as stick representations, with the carbon atoms colored in orange
and cyan, respectively. LXR -LBD is shown as a surface representation with the
color scheme of carbon in gray, nitrogen in blue, and oxygen in red.
a-LBD. Compounds 2
a
F C
3
HO
position for the His-Trp switch, we tried to modify the compound
to form a more robust interaction with His421. Based on the crystal
structure, the compound which had the hydroxyl group on the C6
position of the tert-butyl benzoate part was predicted to achieve a
O
n
a, b
O
O
n
OMe
CO tBu
14a-14c
2
OH
13a-13c
hydrogen bond with the N
To test our hypothesis, the C6-hydroxy derivative, 12, was
designed. The synthetic routes of and 12 are shown in
Scheme 1. Tert-butyl esterification and bromination of acid chlo-
ride 5 provided the key intermediate, 6. By benzylation of phenol
714 with 6, 4 was prepared. On the other hand, C6-hydroxyl
group-installed 12 was synthesized from benzaldehyde 8.15
ortho-Lithiation of benzaldehyde dimethyl acetal 9, obtained by
the successive protection of phenol and aldehyde of 8, was fol-
lowed by a reaction with DMF and reduction by NaBH4 to give
e2 atom of His421.
F C
3
HO
4
HO
O
O
a, e, f
c, d
O
N
O
NH
CO tBu
CO H
2
2
2
N
15
16
17
Scheme 2. Reagents and conditions: (a) 6, K2CO3, DMF, rt–0 °C (70–80%); (b) 3 N aq
NaOH, THF–MeOH (29–59%); (c) allyl chloroformate, Et3N, CH2Cl2, 0 °C–rt; (d)
K2CO3, MeOH, rt (74%, 2 steps); (e) Pd(PPh3)4, pyrrolidine, 1,4,-dioxane-H2O, rt
(85%); (f) succinic anhydride, Et3N, CH2Cl2, 0 °C–rt (68%).
another key intermediate, 10.
A Mitsunobu reaction using
TMAD16 of 10 and 7 led to 11. Deprotection of compound 11 under
an acidic condition followed by oxidation and tert-butyl esterifica-
tion provided 12.
carboxyl methyl group of 2 binds. Thereby, we supposed that the
carboxyl group could be fused into 4 by replacing or extending
the N-methylacetamide group with the tethered carboxyl group.
Based on this idea, carboxylic acid analogs were designed and syn-
thesized (Scheme 2).
The LXR binding and the efficacy of transcriptional activation of
4 and 12 were evaluated in the fluorescence polarization (FP)
assay17 and the cell transfection (CTF) assay18
,
respectively
Compounds 14a–c were prepared from phenols (13a–c)19 with
the methoxycarbonyl group and 6 in 2 steps (benzylation and
hydrolysis). Compound 17 was synthesized from commercially
(Table 1). Though the effect of the C6-hydroxyl group installation
on Ki in the FP assay and EC50 in the CTF assay was small, 12
achieved a three-fold improvement of the efficacy in the CTF assay,
as expected.
In parallel to the work to achieve the interaction with His421,
we focused on improving the physicochemical properties of our
LXR ligand. Because the obtained tert-butyl benzoate analogs
tended to show poor solubilities (data not shown), we assumed
that the low solubility was one of the reasons for low oral absorba-
bility of 4. Hence we tried to introduce the hydrophilic group to
tert-butyl benzoate analogs.
A clue for the modification to improve solubility was also
obtained from the crystal structures. While searching for a suitable
hydrophilic group to introduce, we focused our attention on 2,
which had a terminal carboxyl group in the structure. In the crystal
structure of LXRb-LBD with 2,13 the carboxyl group of 2 interacts
available 15. Phenol 15 was converted to 16 in
2 steps.
Benzylation of 16 with 6, followed by deprotection of the N-alloc
group and amidation with succinic anhydride gave 17.
First, compounds which directly tethered the carboxyl group by
several lengths of the linear methylene chain instead of N-methy-
lacetamide moiety were tested. As seen from the data in Table 2,
the binding affinities of 14a and 14c to LXRs were drastically
decreased. Nevertheless, 14b retained a three-fold decreased bind-
ing affinity for LXRb compared to 4. Compound 17 which contained
the extended carboxyl methyl group at the terminal part of 4 also
retained moderate binding affinities for LXRs. We suspected that
the decreased binding affinities of the carboxylic acid analogs were
due to improper interactions of the carboxyl group with the LXRs.
Because there was a wide space between the ligand binding pocket
and the loop where the carboxyl group was intended to interact,
flexible linkers may not have held the carboxyl group at an appro-
priate location to interact with Leu316 on the loop. Therefore, the
benzene ring was selected as a rigid linker to fix the terminal car-
boxyl group, and biphenyl analogs were designed for the following
chemical modification.
with Arg319 (Arg305 in LXR
LXR ) on the loop between bsheet-1 and bsheet-2. Overlay of the
structure of LXRb-LBD with 2 on that of LXR -LBD with 4 showed
a) on Helix5 and Leu330 (Leu316 in
a
a
that the N-methylacetamide group of 4 occupies a position similar
to that of the phenyl ring of the phenyl acetic acid moiety of 2
(Fig. 4). However, 4 does not occupy the space where the terminal

Y. Matsui et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3914–3920
3917
Table 2
Table 3
Binding affinities of the carboxyl group introduced compounds
Binding affinities of the carboxyl group introduced biphenyl compounds
F3C
F3C
O
O
O
O
R1
O
O
R2
FP assay
Ki (lM)
FP assay
Ki ( M)
l
Compound
R2
Compound
R1
LXRa
LXRb
LXR
a
LXRb
25a
25b
25c
25d
m-COOH
p-COOH
m-CH2COOH
p-CH2COOH
6.3
3.4
14a
14b
14c
17
(CH2)3COOH
(CH2)4COOH
(CH2)5COOH
N(CH3)CO(CH2)2COOH
N.D.
N.D.
8.6
4.9
0.63
1.2
n.a.a
5.1
n.a.
0.36
0.26
1.0
4.0
0.66
a
n.a.: >10 lM.
N.D.: not determined since a reasonable sigmoid curve could not be obtained.
product of 28 and 29. A Mitsunobu reaction of 30a with 10 using
ADDP,20 followed by acid hydrolysis led to 31a. Synthesis of 32a
was carried out by the succeeding O-allylation of 31a, oxidation
to carboxylic acid, tert-butyl esterification, deprotection of the allyl
group, and hydrolysis of methyl ester. In the same manner, 32b
was synthesized from 30b and 10 via aldehyde 31b.
The synthesis of biphenyl derivatives (25a–d) are shown in
Scheme 3. 4-Iodo benzene analog 19, derived from 18 and 6, was
converted to boronate ester 20. Compounds 25a–d were prepared
by Suzuki coupling with the appropriate combination of the left
hand (19, 20) and the right hand (21–24) parts. The following
hydrolysis of methyl ester was required for 25b and 25d. Toward
the synthesis of the C6-hydroxyl group-installed analog, 30a was
prepared by the Suzuki coupling of 26 and 27. Compound 30b
was also obtained from deprotection of the Suzuki coupling
Although the binding affinities of phenyl benzoic acids (25a and
25b) were not attractive, biphenylacetic acids (25c and 25d) main-
tained similar potency values to 4 in the LXRb binding. (Table 3).
Based on the structure–activity relationship (SAR) of evaluated
compounds, biphenylacetic acid derivatives having the hydroxyl
group at the C6-position of the tert-butyl benzoate part were
designed, expecting to show both improved efficacy of transcrip-
tional activation and oral absorbability. As listed in Table 4, para-
biphenylacetic acid 32b showed higher potencies and efficacies
than the corresponding meta-analog 32a in the CTF assay. It was
also noteworthy that 32b showed a potent cholesterol efflux activ-
ity.21 Hence 32b was selected for further in vivo studies.
Compound 32b exhibited a much higher Cmax and plasma concen-
tration at 6 h after oral administration at a dose of 10 mg/kg than
the lead compounds 4 and 12 in mice (Table 5).
F C
3
HO
I
19
R1 =
6
O
b
66%
O
a
94%
I
B
CO tBu
2
20 R1=
R1
18
21
O
B(OH)
CO H
2
23
F C
2
3
19
20
20
c
21
c
O
29%
5.1%
24
B(OH)
2
CO H
CO tBu
2
22
c, d
2
19
22
c, d
CO Me
n
2
25a-25d
32% 2 steps
43% 2 steps
Br
To verify the ligand design, the crystal structure of LXRa-LBD in
CO H
2
23
complex with 32b and SRC-1 peptide was determined.22 As we
have intended, the hydroxyl group of 32b forms a hydrogen bond
to His421 with the length of 3.4 Å (Fig. 5A). Although the location
HO
Br
Br
OMe
Br
+
O
B
O
30a
O
e
67%
CO Me
2
24
26
27
of the carboxyl group of compound 32b is slightly shifted in LXRa-
LBD compared with that of 2 in LXRb-LBD, one of the carboxyl
group oxygens has a hydrogen bond to the main-chain of Leu316
(Fig. 5B), which is comparable to the one seen in 2 with LXRb-LBD.
MeO
O
+
30b
OMe
B(OH)
2
f, g
29
28
75% 2 steps
F C
3
HO
O
10
HO
30a
Table 4
CO Me
2
CHO
OMe
h, i
Profiles of compounds which have both the hydroxyl group and the carboxyl group
31a
45% 2 steps
O
30b
10
h-j
F3C
O
j-m, d
33% 5 steps
80% 3 steps
HO
F C
3
F C
3
k-m, d
O
O
O
O
R2
O
HO
48% 4 steps
CHO
CO H
CO tBu
O
2
2
31b
OMe
32a, 32b
CTF assay EC50
(% efficacy)
(lM)
a
Cholesterol efflux assay
Scheme 3. Reagents and conditions: (a) K2CO3, DMF, rt; (b) bis(pinacolato)diboron,
PdCl2(dppf), KOAc, DMSO–CH2Cl2, 80 °C; (c) Pd(PPh3)4, K2CO3, DMA–H2O, 90 °C; (d)
3 N aq NaOH, THF–MeOH; (e) Pd(PPh3)4, K2CO3, DMA–H2O, 110 °C; (f) Pd(PPh3)4,
1 M aq Na2CO3, toluene–EtOAc, reflux; (g) 1 N BCl3 in CH2Cl2, n-Bu4NI, CH2Cl2,
ꢀ78 °C–rt; (h) ADDP, n-Bu3P, THF, rt; (i) p-TsOH, acetone, rt; (j) allylbromide, K2CO3,
DMF, 50 °C; (k) NaClO2, NaH2PO4, 2-methyl-2-butene, H2O-t-BuOH, rt; (l)
Me2NCH(OtBu)2 toluene, reflux; (m) Pd(PPh3)4, pyrrolidine, DMF, rt ADDP = 1,10-
(azodicarbonyl)dipiperidine.
Compound R2
LXR
m-CH2COOH 1.3 (39)
p-CH2COOH 0.51 (79) 0.16(51) 69(1), 57(0.1)
a
LXRb
% efficacy (dose/lM)
32a
32b
0.53 (27) 58(1), 3(0.1)
a
Cholesterol efflux assay: % efficacy was shown at the concentration indicated
with parenthesis ( M). The relative value is based on the activity that 1 shows
efficacy (100%) at 1 M concentration.
l
l

3918
Y. Matsui et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3914–3920
Table 5
Plasma concentration of the compounds
Plasma concentrationa
Compound
Cmax
(
lg/ml)
At 6 h (lg/ml)
4
12
32b
1.7
1.0
8.9
1.0
0.67
5.3
a
The compounds were orally administered to mice at a dose of 10 mg/kg.
Figure 6. Effect of 32b and 1 in C57Bl/6 mice. Animals were dosed daily by oral
gavage (n = 5/group) for 7 days before measurements. ⁄: p<0.05 as compared with
the vehicle group by Dunnett’s test.
Figure 7. Effect of 32b in cynomolgus monkeys. Animals were dosed daily by oral
gavage (n = 6/group) for 14 days. Results are represented as mean + SE. (A) Line plot
of blood ABCA1 mRNA levels on Day 1 and on Day 14. (B) Triglyceride level.
Figure 5. Compound 32b in the ligand binding site of LXR
shown as a stick representation with the carbon atoms in green. The color scheme
for LXR -LBD is the same as that used in Figure 3B. Hydrogen bonds are shown as
red broken lines. (A) the C6-hydroxyl group is hydrogen bonding to His421. (B) 2 in
LXRb-LBD is overlaid on 32b in LXR -LBD. Compound 2 is shown in stick with
a-LBD. Compound 32b is
a
a
carbon atoms in orange while LXRb-LBD is not shown. The carboxyl group of 32b
slightly shifted from that of 2.
proliferator-activated receptor gamma coactivator 1
and activating signal cointegrator 2 (ASC2), were identified as the
potential determinants. PGC1 has been reported as a potent tran-
a (PGC1a)
a
scriptional coactivator for LXRa, and potentiates its transcriptional
The effects of 32b and 1 on blood ABCA1, liver FAS and liver
SCD-1 mRNA levels, and on plasma TG levels were investigated
in C57Bl/6 mice after repeated oral administration at 1 mg/kg for
7 days (Fig. 6). Both 32b and 1 significantly elevated the blood
ABCA1 mRNA levels. On the contrary, 32b did not affect the liver
FAS mRNA levels, liver SCD-1 mRNA levels, nor plasma TG levels
while 1 elevated them. To further investigate the pharmacological
property, 32b (0, 0.1, 0.3, 1, 3 or 10 mg/kg) was orally administered
to cynomolgus monkeys for 14 days. Treatment with 32b signifi-
cantly elevated blood ABCA1 mRNA expression in a dose depen-
dent manner on both Day 1 and Day 14 (Fig. 7A). Importantly,
32b did not affect plasma TG after repeated dosing as well as single
dosing (Fig. 7B).
activity upon the SREBP-1c gene promoter.24 ASC2 has also been
identified as a physiologically important transcriptional coactiva-
tor of LXR
of 32b for LXR
than that of 1, while both 32b and 1 induced the recruitment of
PGC1 or ASC2 to LXRb to almost the same extent (Fig. 8). This
a
in regard to lipid metabolism in the liver.25 The effect
a
on PGC1a or ASC2 recruitment was less efficacious
a
characteristic cofactor recruitment property of 32b might be one
of the reasons why 32b has the favorable pharmacological prop-
erty of ABCA1 expression without FA expression, SCD1 expression,
nor TG elevation.
In summary, a SAR study on a series of tert-butyl benzoate ana-
logs was performed to obtain a potent LXR agonist. The 4 bound
LXRa-LBD structure, along with the 2 bound LXRb-LBD structure,
To elucidate why 32b had the wide window between ABCA1
led us to introduce the hydroxyl group at the C6-position of the
tert-butyl benzoate core to enhance the efficacy in the cell-based
reporter assay, and to introduce the carboxyl group tethered by
and TG elevation,
a fluorescence resonance energy transfer
(FRET) assay was carried out using LXRs and synthetic peptides
from transcriptional cofactors.23 Two cofactors, peroxisome-

Y. Matsui et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3914–3920
3919
Collins, J. L.; Tangirala, R. K.; Tontonoz, P. J. Clin. Invest. 2007, 117, 2337; (e)
Gabbi, C.; Warner, M.; Gustafsson, J.-A. Mol. Endocrinol. 2009, 23, 129.
6. (a) Repa, J. J.; Turley, S. D.; Lobaccaro, J.-M. A.; Medina, J.; Li, L.; Lustig, K.; Shan,
B.; Heyman, R. A.; Dietschy, J. M.; Mangelsdorf, D. J. Science 2000, 289, 1524; (b)
Terasaka, N.; Hiroshima, A.; Koieyama, T.; Ubukata, N.; Morikawa, Y.; Nakai, D.;
Inaba, T. FEBS Lett. 2003, 536, 6.
7. (a) Collins, J. L.; Fivush, A. M.; Watson, M. A.; Galardi, C. M.; Lewis, M. C.; Moore,
L. B.; Parks, D. J.; Wilson, J. G.; Tippin, T. K.; Binz, J. G.; Plunket, K. D.; Morgan, D.
G.; Beaudet, E. J.; Whitney, K. D.; Kliewer, S. A.; Willson, T. M. J. Med. Chem.
1963, 2002, 45; (b) Joseph, S. B.; McKilligin, E.; Pei, L.; Watson, M. A.; Collins, A.
R.; Laffitte, B. A.; Chen, M.; Noh, G.; Goodman, J.; Hagger, G. N.; Tran, J.; Tippin,
T. K.; Wang, X.; Lusis, A. J.; Hsueh, W. A.; Law, R. E.; Collins, J. L.; Willson, T. M.;
Tontonoz, P. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 7604.
8. Schultz, J. R.; Tu, H.; Luk, A.; Repa, J. J.; Medina, J. C.; Li, L.; Schwendner, S.;
Wang, S.; Thoolen, M.; Mangelsdorf, D. J.; Lustig, K. D.; Shan, B. Genes Dev. 2000,
14, 2831.
9. Shimomura, I.; Shimano, H.; Korn, B. S.; Bashmakov, Y.; Horton, J. D. J. Biol.
Chem. 1998, 273, 35299.
10. Crystals were obtained by mixing equal volumes of the precipitant solution (8–
18% (w/v) polyethylene glycol 2000 monomethyl ether, 0.18 M ammonium
sulfate, and 0.1 M Tris–HCl (pH 7.5)) and the mixture of LXRa-LBD, the
compound, and SRC-1 peptide (20 mM Tris–HCl (pH 8.0), 50 mM sodium
chloride, 0.1 mM EDTA, and 1 mM DL-dithiothreitol). The atomic coordinates
for the crystal structure of LXRa-LBD in complex with 4 and SRC-1 peptide
were deposited in the Protein Data Bank, with the PDB accession code, 5AVI.
11. Weatherman, R. V.; Fletterick, R. J.; Scanlan, T. S. Annu. Rev. Biochem. 1999, 68,
559.
12. Williams, S.; Bledsoe, R. K.; Collins, J. L.; Boggs, S.; Lambert, M. H.; Miller, A. B.;
Moore, J.; McKee, D. D.; Moore, L.; Nichols, J.; Parks, D.; Watson, M.; Wisely, B.;
Willson, T. M. J. Biol. Chem. 2003, 278, 27138.
Figure 8. Effects of 32b and 1 on cofactor recruitment to LXRs. Effects of 32b and 1
on PGC1a or ASC2 recruitment to LXR were examined by a fluorescence resonance
energy transfer assay. ⁄p<0.05, as compared with the vehicle control group and 1
group or 32b group and 1 group using Student’s t-test. The data are represented as
13. Färnegårdh, M.; Bonn, T.; Sun, S.; Ljunggren, J.; Ahola, H.; Wilhelmsson, A.;
Gustafsson, J. A.; Carlquist, M. J. Biol. Chem. 2003, 278, 38821.
mean + SE (n = 4). (A) PGC1
ASC2 to LXRb.
a to LXRa. (B) ASC2 to LXRa. (C) PGC1a to LXRb. (D)
14. Fernando, C. R.; Calder, I. C.; Ham, K. N. J. Med. Chem. 1980, 23, 1153.
15. Daly, A. M.; Gilheany, D. G. Tetrahedron: Asymmetry 2003, 14, 127.
16. Tsunoda, T.; Otsuka, J.; Yamamiya, Y.; Itô, S. Chem. Lett. 1994, 23, 539.
17. Fluorescence polarization (FP) assay: assay buffer was composed of 10 mM
HEPES (pH 7.0), 0.1 mM EDTA 3Na, 10 mM DTT, 20% (v/v) Ultrapure BGG 5 mg/
ml (Invitrogen Co.), 0.5 nM fluorescein isothiocyanate (FITC)-labeled T0901317
biphenyl at one terminal to improve the oral absorbability. The
obtained compound, 32b, exhibited a potent LXR agonistic activity
in a cell-based assay, showed good exposure of post-oral adminis-
tration, and induced blood ABCA1 mRNA expression without
affecting the plasma TG levels in both mice and cynomolgus
monkeys.
and
1
lM
His-LXR
a
/FLAG-RXR
a
or
1 lM His-LXRb/FLAG-RXRa. FP assay
employed 38
l
l of buffer with 2
ll of test compound in DMSO, which was
placed at room temperature overnight before measurement.
18. Cell transfection (CTF) assay: The African green monkey kidney-derived cell
line and CV-1 cells were cultured in DMEM supplemented with 10% charcoal/
dextran-treated FBS (HyClone), 1% penicillin-streptomycin, 1% Gluta Max-1
(Invitrogen Co.), and 1 mM sodium pyruvate (Invitrogen Co.) at 37 °C. On the
day before transfection, the CV-1 cells were seeded in 96-well assay plates at
Acknowledgments
2.0 ꢁ 104 cells/well and placed for 20 h. Expression plasmid pcDNA-hLXR
a or
We thank Yasuko Kono for her evaluation of the FP assay, Ayano
Ogawa-Hiroshima for her evaluation of the cholesterol efflux
assay, Tomoko Takayama for the protein production, and Rie
Nakagomi-Hagihara for the pharmacokinetic study. The HTS screen
pcDNA-hLXRb and luciferase reporter plasmid pTAL-LXRE including LXR
response element (ref. Willy P.J. et al.: Genes Dev. 1995, 9, 1033.) were
transfected with Lipofectamine Reagent (Invitrogen Co.). After 3 h, 70 ll of the
DMEM medium was added. Then, DMSO solutions of the test compounds or
DMSO alone (for the control) were added. The cells were further incubated for
about 20 h .The luciferase assay was performed using Luciferase Assay Reagent
(Promega Co.). The luminescence intensity in each well was measured using a
luminescence plate reader (Analyst HT, Molecular Devices). The luminescence
intensity was represented as counts per second (CPS). The luminescence
intensity data were analyzed by the Sigmoid Emax model.
was carried out at X-Ceptor (now Exelixis). The initial LXRa-LBD
crystallization protocol was developed at Syrrx (now Takeda
California) and further optimized at Sankyo (now Daiichi Sankyo
RD Novare).
19. (a) Hazeldine, S. T.; Polin, L.; Kushner, J.; White, K.; Corbett, T. H.; Horwitz, J. P.
Bioorg. Med. Chem. 2005, 13, 3910; (b) Yi, C.-S.; Martinelli, L. C.; Blanton, C.;
DeWitt, Jr. J. Org. Chem. 1978, 43, 405; (c) Chu, L.; Hutchins, J. E.; Weber, A. E.;
Lo, J.-L.; Yang, Y.-T.; Cheng, K.; Smith, R. G.; Fisher, M. H.; Wyvratt, M. J.; Goulet,
M. T. Bioorg. Med. Chem. Lett. 2001, 11, 509.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2015.07.
047. These data include MOL files and InChiKeys of the most
important compounds described in this article.
20. Tsunoda, T.; Yamamiya, Y.; Itô, S. Tetrahedron Lett. 1993, 34, 1639.
21. Cholesterol efflux assay: RAW264.7 cells were seeded in a 96-well plate and
cultured in DMEM supplemented with 1% FBS and 1% penicillin-streptomycin,
including 0.2
washing the cells with 0.2% albumin bovine serum (BSA, Sigma–Aldrich Co.)/
PBS, 100
l/well of apoAI(ꢀ) medium or apoAI(+) medium (DMEM including
10 g/ml apoAI (Biogenesis)) was added. The test compounds (final
concentrations: 0.01, 0.1, and 1 M) and DMSO alone were added and the
cells were further incubated for about 24 h. Then, 75 l of the medium in each
well was transferred to a Luma Plate (PerkinElmer, Inc.) and dried in the draft
chamber. The cells were washed with PBS and 250 l of MICROSCINT 20
l
Ci/ml [4-14C]-cholesterol (PerkinElmer, Inc.) for 2 days. After
l
References and notes
l
l
1. (a) Willy, P. J.; Umesono, K.; Ong, E. S.; Evans, R. M.; Heyman, R. A.;
Mangelsdorf, D. J. Genes Dev. 1995, 1033, 9; (b) Jaye, M. Curr. Opin. Investig.
Drugs. 2003, 1053, 4.
2. Laffitte, B. A.; Repa, J. J.; Joseph, S. B.; Wilpitz, D. C.; Kast, H. R.; Mangelsdorf, D.
J.; Tontonoz, P. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 507.
3. (a) Venkateswaran, A.; Laffitte, B. A.; Joseph, S. B.; Mak, P. A.; Wilpitz, D. C.;
Edwards, P. A.; Tontonoz, P. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 12097; (b)
Wang, N.; Lan, D.; Chen, W.; Matsuura, F.; Tall, A. R. Proc. Natl. Acad. Sci. U.S.A.
2004, 101, 9774; (c) Nakamura, K.; Kennedy, M. A.; Baldán, Á.; Bojanic, D. D.;
Lyons, K.; Edwards, P. A. J. Biol. Chem. 2004, 279, 45980; (d) Baldán, Á.; Bojanic,
D. D.; Edwars, P. A. J. Lipid Res. 2009, 50, S80.
l
l
(PerkinElmer, Inc.) was added, and then incubated at room temperature
overnight. Cellular (assay plate) and medium (Luma Plate) radioactivities
(cpm) were measured by TOPCOUNT (B991201, Packard). ApoAI-dependent
cholesterol efflux (%) was calculated according to the following formula:
Cholesterol efflux (%) = counts in medium (cpm)/[counts in medium
(cpm) + counts in cells (cpm)] * 100 ApoAI-dependent cholesterol efflux
(%) = cholesterol efflux in the case of apoAI(+) medium (%) ꢀ mean
cholesterol efflux in the case of apoAI(ꢀ) medium (%).
4. Joseph, S. B.; Tontonoz, P. Curr. Opin. Pharmacol. 2003, 3, 192.
22. The atomic coordinates for the crystal structure of LXRa-LBD in complex with
5. For reviews on the therapeutic utilities of LXR modulators, see (a) Collins, J. L.
Curr. Opin. Drug Discov. Dev. 2004, 7, 692; (b) Bradley, M. N.; Tontonoz, P. Drug
Discov. Today: Therapeutic Strategies 2005, 2, 97; (c) Geyeregger, R.; Zeyda, M.;
Stulnig, T. M. Cell. Mol. Life Sci. 2006, 63, 524; (d) Bradley, M. N.; Hong, C.; Chen,
M.; Joseph, S. B.; Wilpitz, D. C.; Wang, X.; Lusis, A. J.; Collins, A.; Hseuh, W. A.;
32b and SRC-1 peptide were deposited in the Protein Data Bank, with the PDB
accession code, 5AVL.
23. Recruitment of the PGC1
a
peptide (residues 130–154) and ASC2 peptide
& Flag-RXR
(residues 1481–1510) as transcriptional cofactors to His-LXR
a
a

3920
protein and His-LXRb & Flag-RXR
Y. Matsui et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3914–3920
a
protein were investigated with a FRET
shaker. After one night of incubation at 4 °C, the emission of fluorescence
excited by a laser at 320 nm was measured at wavelengths of 620 nm (B-count,
cpm) and 665 nm (A-count, cpm) using a fluorescence plate reader (WALLAC
EnVision, PerkinElmer, Inc.).
assay, based on the non-radiative energy transfer of the long-lived emission
from the europium cryptate donor t the cross-linked allophycocyanin (XL665)
acceptor. Two microliters of 1 mM compound or DMSO alone (for the DMSO
control) was pipetted into the wells of a 384 well plate (n = 4). Then, 8
His-LXR & Flag-RXR mixture solution or the His-LXRb & Flag-RXR
solution was added to each well and the contents of the wells were mixed
using a plate shaker. After 60 min of incubation at room temperature, 10 l of
the PGC1 mixture solution or the ASC2 mixture solution was added to the
individual wells and the contents of the wells were mixed using the plate
l
l of the
24. (a) Oberkofler, H.; Schraml, E.; Krempler, F.; Patsch, W. Biochem. J. 2003, 371,
89; (b) Oberkofler, H.; Schraml, E.; Krempler, F.; Patsch, W. Biochem. J. 2004,
381, 357.
25. Kim, S. W.; Park, K.; Kwak, E.; Choi, E.; Lee, S.; Ham, J.; Kang, H.; Kim, J. M.;
Hwang, S. Y.; Kong, Y. Y.; Lee, K.; Lee, J. W. Mol. Cell. Biol. 2003, 23, 3583.
a
a
a mixture
l
a