Design, synthesis and pharmacology of 1,1-bistrifluoromethylcarbinol derivatives as liver X receptor β-selective agonists

Article abstract of DOI:10.1016/j.bmcl.2015.04.080

A novel series of 1,3-bistrifluoromethylcarbinol derivatives that act as liver X receptor (LXR) β-selective agonists was discovered. Structure-activity relationship studies led to the identification of molecule 62, which was more effective (E_max) and selective toward LXRβ than T0901317 and GW3965. Furthermore, 62 decreased LDL-C without elevating the plasma TG level and significantly suppressed the lipid-accumulation area in the aortic arch in a Bio F₁B hamster fed a diet high in fat and cholesterol. We demonstrated that our LXRβ agonist would be potentially useful as a hypolipidemic and anti-atherosclerotic agent. In this manuscript, we report the design, synthesis and pharmacology of 1,3-bistrifluoromethylcarbinol derivatives.

Full text of DOI:10.1016/j.bmcl.2015.04.080

Bioorganic & Medicinal Chemistry Letters 25 (2015) 2668–2674
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis and pharmacology of 1,1-bistrifluoromethylcarbinol
derivatives as liver X receptor b-selective agonists
Minoru Koura, Takayuki Matsuda, Ayumu Okuda, Yuichiro Watanabe, Yuki Yamaguchi, Sayaka Kurobuchi,
⇑
Yuuki Matsumoto, Kimiyuki Shibuya
Tokyo New Drug Research Laboratories, Pharmaceutical Division, Kowa Co., Ltd, 2-17-43, Noguchicho, Higashimurayama, Tokyo 189-0022, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 17 December 2014
Revised 3 April 2015
Accepted 23 April 2015
Available online 5 May 2015
A novel series of 1,3-bistrifluoromethylcarbinol derivatives that act as liver X receptor (LXR) b-selective
agonists was discovered. Structure–activity relationship studies led to the identification of molecule 62,
which was more effective (Emax) and selective toward LXRb than T0901317 and GW3965. Furthermore,
62 decreased LDL-C without elevating the plasma TG level and significantly suppressed the lipid-accumu-
lation area in the aortic arch in a Bio F1B hamster fed a diet high in fat and cholesterol. We demonstrated
that our LXRb agonist would be potentially useful as a hypolipidemic and anti-atherosclerotic agent. In
this manuscript, we report the design, synthesis and pharmacology of 1,3-bistrifluoromethylcarbinol
derivatives.
Keywords:
Liver X receptor (LXR) b-selective
LDL-C
1,1-Bistrifluoromethylcarbinol
Anti-atherosclerosis
Ó 2015 Elsevier Ltd. All rights reserved.
In our drug discovery program dedicated toward the develop-
ment of novel anti-atherosclerotic agents, we have focused our con-
tinuing efforts on LXR agonists that offer the possibility of reverse
cholesterol transport (RCT) from atherosclerotic lesions. LXRs are
ligand-activated transcription factors involved in cholesterol meta-
increased the HDL-C levels without a significant TG elevation and
resulted in a decrease in the lipid-accumulation areas in the aortic
arch in a Bio F1B hamster fed a diet high in fat and cholesterol.6
However, further developments and investigations on the
potency and selectivity toward LXRb and improvements in the
physical properties and pharmacokinetics are required. In this
manuscript, we report the modification of the structure of the head
moiety in our molecular design (head-to-tail) and identify 62 as the
optimized molecule. For this purpose, we focused on the His435-
Trp457 activation switch7 to enhance the interaction between the
ligand and the LXRb receptor. First, we speculated that the trifluo-
romethyl group may be a key factor in the head structure for LXR
activation, as shown in Figures 1 and 2. Second, we aimed to modify
the head structure of 36 and transformed its structure into the fol-
lowing four types of compounds (4, 5, 6 and 7)8,9: (1) the removal of
the carbonyl group at the 2-position of the chromene moiety and
the reduction of the double bond gave the chromane structure 4,
(2) the contraction of the six-membered ring in 4 and the incorpo-
ration of an additional trifluoromethyl group yielded the 1,3-dihy-
droisobenzofuran structure 5, (3) the cleavage of the C–O bond in
the five-membered ring of 5 gave the carbinol structure 6, and (4)
the incorporation of a n-propyl group on the benzene ring yielded
the bis-n-propyl phenyl structure 7 (Fig. 3).
bolism, glucose homeostasis, inflammation and lipogenesis.1 LXR
a
(known as NR1H3) is the dominant subtype in the liver, small intes-
tine, and macrophages, whereas LXRb (known as NR1H2) is dis-
tributed ubiquitously. LXR activation is known to regulate the
ATP-binding cassette transporter A1 (ABCA1), ABCG1 and
ABCG5/G8 expression and cholesterol metabolism to increase the
high-density lipoprotein cholesterol (HDL-C) levels and to serve
in cholesterol efflux and excretion.2
Several studies have reported that T0901317 and GW3965 exhi-
bit the potentially useful property of increasing the plasma HDL-C
and decreasing atherosclerotic lesions in mouse models of
atherosclerosis (Fig. 1).3,4 However, elevations in the plasma and
hepatic triglyceride (TG) levels were observed as significant side
effects.5 This lipogenesis originates from the expression of sterol
regulatory element-binding protein-1c (SREBP-1c), which is
induced by LXRa activation. To suppress this undesirable effect in
the LXR agonist, we made efforts to differentiate between these
two LXR subtypes and discovered that the 2-oxochromene deriva-
tive 3 can act as a LXRb-selective agonist. Derivative 3 also
The head of 4, which was the requisite 8-n-propyl-4-(trifluo-
romethyl)chroman-7-ol (11), was prepared from 7-hydroxy-
8-n-propyl-4-(trifluoromethyl)-2H-chromen-2-one (8) through a
three-step protocol, as depicted in Scheme 1. The reduction of 8
⇑
Corresponding author. Tel.: +81 42 391 6211; fax: +81 42 395 0312.
E-mail address: k-sibuya@kowa.co.jp (K. Shibuya).
http://dx.doi.org/10.1016/j.bmcl.2015.04.080
0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.

M. Koura et al. / Bioorg. Med. Chem. Lett. 25 (2015) 2668–2674
2669
propyl-3-hydroxybenzoate (15b). The reduction of 15a with
LiAlH4 afforded 3-hydroxymethyl-2-n-propylphenol (16), which
was subsequently halogenated using tetrabutylammonium tribro-
mide to give the bromobenzene 17. To distinguish the two hydroxyl
groups in 17, we first protected the primary alcohol with 3,4-dihy-
dro-2H-pyran (DHP) in the presence of p-TsOH and then protected
the phenol of 18 with chloromethyl methyl ether (MOMCl) in the
presence of NaH to produce the bis-protected bromobenzene 19.
The lithiation of 19 with n-BuLi was followed by reaction with
1,1,1,3,3,3-hexafluoroacetone to give the carbinol 20. After the
deprotection of THP, the cyclization of the resulting two hydroxyl
groups
CF3
Cl
N
CF3
F3C
HO
O
O
S
O
N
CO2H
F3C
GW3965 (2)
T0901317 (1)
LXRα EC50 (Emax): 0.5 μM (100%)
LXRβ EC50 (Emax): 0.4 μM (100%)
α
μ
LXR EC50 (Emax): 1.3 M (60%)
LXRβ EC50 (Emax): 0.4 μM (62%)
Figure 1. Structures of LXR agonists from Tularik and GlaxoSmithKline.
in 21 was performed through the Mitsunobu reaction10 to form the
1,3-dihydroisobenzofuran intermediate 22. The cleavage of the
MOM ether in 22 under acidic conditions furnished the desired
phenol 23.
CF3
O
The head of 6, the requisite 4-(1,1,1,3,3,3-hexafluoro-2-(hydro-
xy)propan-2-yl)-2-n-propylphenol (28), was prepared from methyl
4-hydroxybenzoate (24) through a nine-step procedure, as depicted
in Scheme 3. Methyl 3-n-propyl-4-hydroxybenzoate (25) was
obtained through the following steps: (i) the allylation of 24 with
allyl chloride, (ii) a thermal Claisen rearrangement of phenoxy allyl
ether, and (iii) the conversion of the allyl group into an n-propyl
group via hydrogenation. The benzylation of phenol 25 followed
by the hydrolysis of the methyl benzoate gave the benzoic acid
26. The acid chloride was obtained through the reaction of 26 with
thionyl chloride, which was then bis-trifluoromethylated using
CF3TMS in the presence of tetramethylammonium fluoride to pro-
vide the 1,1-bistrifluoromethylcarbinol 27. After the protection of
the carbinol with MOMCl, the benzyl ether was deprotected via
hydrogenolysis to yield 28.
The head of 7, the requisite, 4-(1,1,1,3,3,3-hexafluoro-2-(meth-
oxymethoxy)propan-2-yl)-2,6-di-n-propylphenol (32), was pre-
pared through a procedure similar to that described above.
Alkyl phenyl ketone 34 was obtained by a Friedel–Crafts acyla-
tion11 of 33 with an alkylacid anhydride. Imidazolidine-2,4-dione
35 was prepared via a Bucherer–Bergs reaction12 of ketone 34 using
NaCN and (NH4)2CO3, as depicted in Scheme 4.
O
N
NH
O
O
O
Tail
O
Head
3b
LXRα EC50 (Emax):
-
μM (2%)
LXRβ EC50 (Emax): 1.4 μM (31%)
Figure 2. Head-to-tail design of our discovered LXRb-selective agonist.
with LiAlH4 afforded the alcohol 9, which was subsequently bromi-
nated using a combination of triphenylphosphine and carbon tetra-
bromide to give the bromide 10. The cyclization of 10 under basic
conditions provided the desired chroman 11.
The head of 5, the requisite 4-n-propyl-1,1-bis-(trifluo-
romethyl)-1,3-dihydroisobenzofuran-5-ol (23), was prepared from
methyl 3-hydroxybenzoate (12) through 12 steps, as depicted in
Scheme 2. Methyl 3-allyloxybenzoate (13) was obtained by the
allylation of 12 with allyl chloride in the presence of a base. A ther-
mal Claisen rearrangement of 13 gave a mixture of methyl 2-allyl-
3-hydroxybenzoate (14a) and methyl 4-allyl-3-hydroxybenzoate
(14b), and this reaction was followed by hydrogenation with Pd/C
in MeOH under a hydrogen atmosphere to provide a mixture of
methyl 2-n-propyl-3-hydroxybenzoate (15a) and methyl 4-n-
A series of 1,1-bistrifluoromethylcarbinol derivatives including
37 and 39 was prepared as depicted in Scheme 5. The alkylation
CF3
CF3
O
CF3
F3C
O
O
O
O
O
O
Head
3
4
5
CF3
CF3
F3C
HO
F3C
HO
O
O
6
7
Figure 3. Molecular design to modify the head structure of 3.
CF3
O
CF3
CF3
CF3
a
b
c
HO HO
OH
Br HO
OH
O
OH
O
OH
8
9
10
11
Scheme 1. Reagents and conditions: (a) LiAlH4, THF, 0 °C, 1 h; (b) PPh3, CBr4, THF, rt, 10 min; (c) K2CO3, DMF, rt, overnight, 70% yield after the three steps.

2670
M. Koura et al. / Bioorg. Med. Chem. Lett. 25 (2015) 2668–2674
R2
a
c
b
MeO
MeO
MeO
MeO
OH
O
OH
O
O
O
O
1
R1
2
12
13
14a
: R = allyl, R = H
14b: R1 = H, R2 = allyl
R2
Br
f
d
e
HO
HO
OH
OH
OH
R1
15a: R1 = n-Pr, R2 = H
1
2
15b
16
19
17
CF3
: R = H, R = n-Pr
F3C
Br
Br
g
h
k
i
HO
THPO
THPO
THPO
OMOM
OMOM
OMOM
OH
18
20
CF3
CF3
CF3
F3C
F3C
F3C
O
j
HO
HO
O
OMOM
OH
21
22
23
Scheme 2. Reagents and conditions: (a) allyl chloride, K2CO3, DMF, 100 °C, overnight, 95%; (b) PhNMe2, 210 °C, 19 h, 80%; (c) H2, Pd/C, MeOH, rt, 2 h, 56%; (d) LiAlH4, THF, rt,
2.5 h, 91%; (e) n-Bu4NBr3, CH2Cl2/MeOH, rt, 1 h, 67%; (f) DHP, p-TsOH-H2O, CH2Cl2, rt, 2 h, 78%; (g) MOMCl, NaH, DMF, 0 °C to rt, overnight, 83%; (h) (i) n-BuLi, THF, À78 °C,
30 min and then À40 °C, 1.5 h; (ii) CF3COCF3, THF, À78 °C, 10 min then 0 °C, 3 h, 65%; (i) aq AcOH, THF, rt, 2 h, 90%; (j) DEAD, PPh3, CH2Cl2, rt, 2 h, 93%; (k) aq HCl, MeOH, rt,
2 h, 96%.
O
O
O
a, b, c
d, e
f, g
MeO
MeO
HO
OH
OH
OBn
24
27
25
26
CF3
CF3
F3C
MOMO
F3C
HO
h, i
OH
OBn
28
O
O
f, g
a, b, c
d, e
MeO
HO
25
OH
OBn
29
30
CF3
CF3
F3C
HO
F3C
MOMO
h, i
OBn
OH
31
32
Scheme 3. Reagents and conditions: (a) allyl chloride, K2CO3, DMF, 50 °C, overnight, 99%; (b) PhNMe2, 210 °C, 12 h, 60%; (c) H2, Pd/C, MeOH, rt, 2 h, 99%; (d) BnBr, K2CO3,
DMF, 80 °C, 2 h, 99%; (e) 2N NaOH aq, EtOH, 50 °C, 2 h, 98%; (f) SOCl2, reflux, 1 h; (g) CF3TMS, Me4NF, DME, À78 °C to rt, 67–72% after these two steps; (h) MOMCl, NaH, THF,
rt, overnight, 89–99%; (i) H2, Pd/C, MeOH, rt, 2 h, 85–99%.
of 28 and 32 with 1,4-dibromobutane and their subsequent reac-
tion with imidazolidine-2,4-dione 35 gave 36 and 38, respectively.
The MOM ethers of 36 and 38 were then cleaved under acidic con-
ditions to afford 37 and 39, respectively.
To investigate what type of head structure is favorable, we eval-
uated the activities of the four newly designed derivatives (4, 5, 6
and 7) in combination with two types of tail structure (a, b) through
GAL4-LXRab luciferase assays. We defined two types of tail

M. Koura et al. / Bioorg. Med. Chem. Lett. 25 (2015) 2668–2674
2671
O
Table 1
O
R1
LXR activities of the newly designed head structurea
a
b
R2
R1
R2
HN
R2
NH
R3
O
O
R4
33
34
35
A
R3 = H or n-propyl
N
N
R5
O
Scheme 4. Reagents and conditions: (a) (R1CO)2O, AlCl3, DMC, rt, 90%; (b) NaCN,
(NH4)2CO3, EtOH (aq), 70 °C, 85%.
a: R4 = R5 = Me
O
b: R4 = 4-methoxyphenyl, R5 = H
Head
Tail
Tail LXRa EC50 (%)c LXRb EC50 (%)c
b
b
Compound Head structure
CF3
structure for the following reasons: (1) 1-N-methyl-5,5-dimethyl-
imidazolidine-2,4-dione (tail a) was the basic structure used for
diversity and (2) 5-(4-methoxyphenyl)-5-methyl-imidazolidine-
2,4-dione (tail b) was the satisfactory structure in 3b.13 The results
were summarized and are shown in Table 1.
a
b
2.4 (7)
ia (2)
1.0 (51)
1.4 (31)
3
O
O
CF3
O
O
The chroman moiety 4a markedly decreased the potency (EC50
)
and efficacy (Emax) toward LXRb compared with 3a, whereas the
dihydroisobenzofuran moiety 5a and 5b increased the EC50 and
Emax values toward LXRb compared with 3a and 3b, respectively.
To our delight, the 1,1-bistrifluoromethylcarbinol moieties 6a, 7a
and 7b significantly enhanced the Emax values toward both LXRs.14
Therefore, we decided to subject the moieties 6 and 7 as promising
core structures in the head-to-tail design to further development.
Although ‘tail a’ was superior to ‘tail b’ in terms of the EC50 and
Emax values, further improvement in the selectivity (EC50
a
b
nd (3)
—
2.2 (21)
—
4
5
6
7
O
CF3
F3C
O
a
b
2.2 (82)
1.7 (81)
0.86 (95)
1.0 (104)
O
LXRa/EC50 LXRb) was limited. We then focused on the substitutable
CF3
F3C
HO
‘tail b’ and investigated which substituents at R1 and R2 were more
favorable on the imidazoline-2,4-dione moiety. To improve the EC50
and Emax values and the selectivity of the optimized combinations,
we evaluated a series of mono-n-propyl 1,1-bistrifluoromethyl-
carbinol (37) and bis-n-propyl 1,1-bistrifluoromethylcarbinol (39)
compounds composed of ‘tail b’ with a linker through GAL4-
a
b
3.2 (97)
2.8 (12)
1.9 (162)
1.8 (32)
O
CF3
F3C
HO
LXR
a/b luciferase assays. The results are summarized in Table 2.
a
b
0.71 (107)
1.6 (51)
0.56 (142)
1.4 (66)
O
In a series of mono-n-propyl type compounds 37 (6b, 40–60), in
the cases in which R1 was a methyl group and R2 was a variety of
substituents on the benzene in 5-phenyimidazolidine-2,4-dione
(40–56), the 3,4-methylendioxyphenyl group in 49 and 2,3-dihy-
drobenzo[b][1,4]dioxin-6-yl group in 50 showed greater EC50 values
for LXRb compared with 6b. The 4-phenyl group in 54 exhibited a
greater Emax value toward LXRb. In contrast, the 4-dimethy-
laminophenyl group in 51, the 4-chlorophenyl group in 52, the
3,4-dichlorophenyl group in 53, the 4-trifluoromethylphenyl group
in 55 and the 4-nitorophenyl group in 56 decreased the Emax values
for LXRb compared with 6b. In addition, if R1 was an ethyl or n-pro-
pyl group instead of a methyl group, the EC50 and Emax values
toward LXRb were increased (57, 58, 59, 60) compared with 6b. In
a series of bis-n-propyl type compounds 39 (7b, 61–69), the 4-
ethoxyphenyl group in 61 and the 4-isopropylphenyl group in 62
showed desirable Emax values for LXRb and selectivities (Emax for
nd = not determined.
ia = inactive at 10 M.
l
a
The GAL4-LXR luciferase assay was performed with a maximal dose of 30 lM.
The results are given as the means from two independent experiments.
The EC50 data are reported in
The Emax is defined as the percentage ratio of the maximum fold induction for
the test compound to the fold induction for T0901317 at 10
experiment.13
b
lM.
c
lM in the same
Based on the results presented in Table 2, we selected 49, 57 and
62 for further evaluation because their profiles showed lower Emax
values for LXRa and higher Emax values for LXRb. These compounds
outperformed our previously discovered LXRb agonist 3b in terms
LXRb/Emax for LXRa).
CF3
CF3
F3C
MOMO
F3C
HO
O
O
R1
R1
R2
R2
a, b
c
N
N
O
28
32
O
O
O
O
NH
NH
O
36
38
37
39
CF3
CF3
F3C
MOMO
F3C
HO
O
O
R1
R2
R1
R2
a, b
c
N
O
N
O
NH
NH
Scheme 5. Reagents and conditions: (a) 1,4-dibromobutane, K2CO3, DMF, rt, 71–94%; (b) 35, K2CO3, DMF, rt, 99%; (c) HCl/EtOAc, rt, 81–82%.

2672
M. Koura et al. / Bioorg. Med. Chem. Lett. 25 (2015) 2668–2674
Table 2
LXR activities of mono-n-propyl 37 and bis-n-propyl 39 in the 1,1-bistrifluoromethylcarbinol derivativesa
CF3
CF3
F3C
HO
F3C
HO
O
O
R2
R1
R2
R1
N
N
O
O
NH
NH
O
O
37 (6b, 40 - 60)
R2
39 (7b, 61-69)
Compound
R1
LXR
a
EC50 (%)c
LXRb EC50 (%)c
b
b
6b
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7b
61
62
63
64
65
66
67
68
69
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Et
4-OMe
H
4-Me
2.8 (12)
2.0 (42)
2.2 (37)
1.9 (8)
1.8 (32)
12 (15)
2.0 (23)
2.2 (64)
ia (0)
4-i-Pr
2-OMe
3-OMe
3,4-diOMe
4-OEt
4-O-i-Pr
4-OCF3
3,4-OCH2O
3,4-O(CH2)2O
4-N(Me)2
4-Cl
3,4-diCl
4-Ph
4-CF3
4-NO2
4-OMe
3,4-OCH2O
3,4-OCH2O
3,4-O(CH2)2O
4-OMe
4-OEt
4-O-i-Pr
4-OCF3
2.7 (7)
2.4 (15)
2.4 (13)
1.3 (89)
1.8 (67)
2.8 (33)
1.7 (54)
1.2 (71)
2.2 (8)
2.2 (70)
1.3 (65)
1.1 (53)
2.8 (51)
1.2 (125)
0.5 (118)
0.4 (171)
2.7 (27)
1.8 (29)
1.1 (43)
1.0 (174)
1.8 (25)
1.9 (39)
1.0 (150)
0.7 (172)
1.3 (72)
0.7 (172)
1.4 (66)
0.8 (148)
1.2 (146)
2.7 (61)
0.1 (101)
0.6 (70)
1.7 (25)
1.6 (58)
1.1 (81)
0.7 (118)
2.2 (23)
1.7 (4)
1.4 (108)
2.5 (10)
2.2 (51)
2.3 (48)
1.5 (72)
2.3 (42)
1.5 (88)
1.6 (51)
0.9 (43)
1.1 (26)
1.8 (62)
0.4 (89)
0.5 (94)
1.8 (21)
1.2 (31)
1.2 (85)
0.6 (118)
Et
n-Pr
Et
Me
Me
Me
Me
Me
Me
Me
Me
Et
3,4-OCH2O
3,4-O(CH2)2O
4-i-Pr
4-Ph
4-OMe
Et
3,4-OCH2O
ia = inactive at 10
lM.
a
The GAL4-LXR luciferase assay was performed with a maximal dose of 30 lM. The results are given as the means from two independent
experiments.
b
c
The EC50 data are reported in
lM.
The Emax is defined as the percentage ratio of the maximum fold induction for the test compound to the fold induction for T0901317 at 10
l
M
in the same experiment.13
of the dose-response curves of LXR
in Figure 4.
a
and LXRb activities, as shown
and a slight elevation in the plasma TG level. Additionally, 62 sig-
nificantly decreased the TC level (53%), whereas T0901317
increased it.
Based on these results, the suppression of lipid accumulation by
62 could be attributed to the marked reduction of the LDL-C level
under the conditions of a diet high in fat and cholesterol. In con-
trast, these findings suggested that the significant reduction of
HDL-C could not substantially serve the RCT from atherosclerotic
lesions. The obtained pharmacological effectiveness should not be
To discriminate between the pharmacokinetic properties of
these agonists, we conducted an in vitro test of liver clearance
(CL) in mice, hamsters and humans. The metabolic stability of the
1,1-bistrifluoromethylcarbinol derivative 62 was greatly improved
and superior to those of 3b, 49 and 57, as shown in Table 3.
We selected 62, which displayed a relatively weak activity for
LXRa and a higher activity for LXRb and metabolic resistance, and
then conducted an in vivo test of 62 in a Bio F1B hamster fed a diet
high in fat and cholesterol (Table 4). The Bio F1B hamster was more
responsive to diet-induced atherosclerosis than mice and an ade-
quate atherosclerosis model used for the evaluation of lipid-modu-
lating agents.15 T0901317, at a dose of 10 mg/kg, significantly
decreased the area of lipid accumulation in the aortic arch (40% rel-
ative to the control) but significantly elevated the plasma TG level
(251%) and TC level (124%) despite decreasing both the HDL-C
(76%) and LDL-C (75%) levels.16 The oral administration of 62 at a
dose of 100 mg/kg considerably decreased the area of lipid accumu-
lation, which was accompanied with a reduction in the LDL-C (45%)
identified as a direct anti-atherosclerotic effect but rather a
hypolipidemic one. Accordingly, we monitored the plasma drug
concentration of 62 and its major metabolite 70 after the oral
administration of 62 at
a dose of 100 mg/kg, as shown in
Figure 5.17 Compound 62 was oxidatively metabolized to afford
the carboxylic acid 70 and then disappeared. The Cmax of 62 was
406 ng/mL (0.6
lM) and did not reach the EC50 level (1.2 lM) asso-
ciated with LXRb agonistic action.
This result was insufficient to upregulate ABCA1 in the periph-
eral blood, resulting in a decrease in the HDL-C level. In contrast,
the long exposure time (6–8 h) to 62 in the intestine was sufficient

M. Koura et al. / Bioorg. Med. Chem. Lett. 25 (2015) 2668–2674
2673
CF3
F3C
HO
Metabolic
Oxidation
62
OH
O
O
70
Figure 5. Drug concentrations of 62 and metabolite 70.17
cholesterol. However, issues remain with 62, including the potency,
selectivity and PK profiles. An even more favorable modulator pro-
file than that of 62 is required for human therapeutic use. Further
development is underway.
Figure 4. Dose–response curves of T0901317, GW3965, 3b, 49, 57 and 62.
Acknowledgments
We thank Dr. S. Tanabe, Director of Tokyo New Drug Research
Laboratories, Kowa Co., Ltd, for his support.
Table 3
CL (
L/min/mg protein) of each animal obtained through an in vitro assay*
l
Supplementary data
Compound
Mouse
Hamster
Human
3b
49
57
62
397
73
65
500
351
295
49
393
31
73
Supplementary data (pharmacological experimental details,
pharmacokinetic experimental details, general preparation proce-
dure and analytical data of all of the synthetic compounds) associ-
ated with this article can be found, in the online version, at http://
dx.doi.org/10.1016/j.bmcl.2015.04.080.
44
42
*
The CL values of compounds 3b, 49, 57, 62 were assessed using hepatic
microsomes from each animal (mouse, hamster and human).
References and notes
1. (a) Nuclear Receptors Nomenclature Committee Cell 1999, 97, 161; (b) Willy, P.
J.; Umesono, K.; Ong, E. S.; Evans, R. M.; Heyman, R. A.; Mangelsdorf, D. J. Genes
Dev. 1995, 9, 1033.
Table 4
In vivo study with T0901317 and 6214
2. (a) Michael, J. Curr. Opin. Invest. Drugs. 2003, 4, 1053; (b) Calkin, A. C.; Tontonoz,
P. Nat. Rev. Mol. Cell. Bio. 2012, 13, 213; (c) Fievet, C.; Staels, B. Biochem.
Pharmacol. 2009, 77, 1316; (d) Zhang, Y.; Chan, J.; Cummins, C. Clin. Lipidol.
2009, 4, 29–40; (e) Tontonoz, P.; Mangelsdorf, D. J. Mol. Endocrinol. 2003, 17,
985; (f) Lund, E. G.; Menke, J. G.; Spalow, C. P. Arterioscler. Thromb. Vasc. Biol.
2003, 23, 1169; (g) Calkin, A. C.; Tontonoz, P. Arterioscler. Thromb. Vasc. Biol.
2010, 30, 1513.
Compound Dose
TC
(%)
HDL-
C (%)
LDL-
C (%)
Plasma
TG (%)
Area of lipid
accumulation (%)
(mg/
kg)
T0901317
10
124* 76*
75*
251*
40*
62
10
87
98
94
97
86
30
100
96
92
91
145*
112
84
3. Terasaka, N.; Hiroshima, A.; Koieyama, T.; Ubukata, N.; Morikawa, Y.; Nakai, D.;
Inaba, T. FEBS Lett. 2003, 536, 6.
53*
75*
45*
49*
4. 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.
*
p<0.05; statistical analysis was conducted using Dunnett’s test. The % value was
calculated relative to the control.
5. (a) 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; (b) Quinet, E. M.; Savio, D. A.; Halpern, A. R.; Chen, L.; Schuster, G. U.;
Gustafsson, J.-Å.; Basso, M. D.; Nambi, P. Mol. Pharmacol. 2006, 70, 1340.
6. (a) Matsuda, T.; Okuda, A.; Watanabe, Y.; Miura, T.; Ozawa, H.; Tosaka, A.;
Yamazaki, K.; Yamaguchi, Y.; Kurobuchi, S.; Koura, M.; Shibuya, K. Bioorg. Med.
Chem. Lett. 2015, 25, 1274; (b) Preparation of compound 3: Int. Patent Appl.
WO2009107387, 2009.
7. (a) Svensson, S.; Ostberg, T.; Jacobsson, M.; Norstrom, C.; Stefansson, K.; Hallen,
D.; Johansson, I. C.; Zachrisson, K.; Ogg, D.; Jendeberg, L. EMBO J. 2003, 22, 4625;
(b) Hoerer, S.; Schmid, A.; Heckel, A.; Budzinski, R. M.; Nar, H. J. Mol. Biol. 2003,
334, 853; (c) 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; (d) Farnegardh, M.;
to inhibit the absorption of cholesterol from the intestine by the
upregulation of ABCA1,18 and this compound plays a substantial
role in inhibiting cholesterol absorption. Therefore, we discovered
the potential utility of the hypolipidemic agent.
In conclusion, the 1,1-bistrifluoromethylcarbinol moiety (head)
played an important role as a potent and selective LXRb agonist.
The novel LXRb agonist 62 decreased TC without a significant TG
elevation and resulted in a decrease in the lipid-accumulation area
in the aortic arch in a Bio F1B hamster fed a diet high in fat and

2674
M. Koura et al. / Bioorg. Med. Chem. Lett. 25 (2015) 2668–2674
Bonn, T.; Sun, S.; Ljunggren, J.; Ahola, H.; Wilhelmsson, A.; Gustafsson, J. A.;
Carlquist, M. J. Biol. Chem. 2003, 278, 38821.
Kawamine, K.; Shibuya, K.; Sato, F.; Sawanobori, K.; Watanabe, T.; Miyazaki, A.
Atherosclerosis 2007, 191, 290.
8. Preparation of the type 5 compounds: Int. Patent Appl. WO2009122707, 2009.
9. Preparation of the type 6 and 7 compounds: Int. Patent Appl. WO2008065754,
2008.
16. Interestingly, Srivastava reported that T0901317 does not alter the LDL-C level
but results in a three-fold increase in the TG level and a 50% increase in the
HDL-C level under the fed condition. Our pharmacological experimental details
are described in the Supplementary data.
10. Mitsunobu, O. Synthesis 1981, 1.
11. (a) Friedel, P.; Crafts, J. M. Compt. Rend. 1877, 84, 1392; (b) Friedel, P.; Crafts, J.
M. Compt. Rend. 1877, 84, 1450.
12. (a) Bergs, H. Ger. Pat. 566,094, 1929; (b) Bucherer, H. T.; Fischbeck, H. T. J. Prakt.
Chem. 1934, 140, 69; (c) Bucherer, H. T.; Steiner, W. T. J. Prakt. Chem. 1934, 140,
291; (d) Ware, E. Chem. Rev. 1950, 46, 403.
17. A solution of 62 in PEG400 was orally administered to a hamster fed the CE-2
chow diet. Blood samples (heparin plasma) were collected from a forearm vein
0.5, 1, 2, 3, 4, 6, 8, 10 and 24 h after administration. The drug concentrations of
62 and 70 in the supernatant were measured by HPLC-MS/MS. Our
pharmacokinetic experimental details are described in the Supplementary data.
18. In an in vivo study with a Bio F1B hamster, we measured the gene expression of
ABCA1 in the intestine and peripheral blood as shown in the Table. ABCA1 in
the intestine was upregulated in a dose-dependent manner, but ABCG5 and
ABCG8 were not significantly altered. Increased intestinal ABCA1 expression
contributes to the decrease in cholesterol absorption. See the reference. Plat, J.;
Mensink, R. P. FASEB J. 2002, 16, 1248–1253. Similarly, ABCA1 in the peripheral
blood was not sufficiently upregulated to contribute to the increases in HDL-C
and RCT. These results led us to draw the previously mentioned conclusion.
13. CHOK-1 cells with the LXRa/GAL4 or LXRb/GAL4 hybrid and stably expressing
the GAL4-responsive reporter vector pG5luc were seeded at a density of 20,000
cells/well on a 96-well plate in HAM-F12 medium containing 10% immobilized
bovine fetal serum, 100 units/mL penicillin G, and 100
sulfate and incubated under a wet atmosphere with 5% CO2 at 37 °C. After 24 h,
medium with the test compound at various concentrations (0.01 M, 0.1 M,
M, 10 M, and 30 M) was added, and the cells were incubated for an
additional 24 h. The effect of the test compound on the activation of luciferase
transcription via LXR - or LXRb-LBD was measured using Bright-Glo (Promega)
lg/mL streptomycin
l
l
1
l
l
l
a
as a luciferase assay substrate and measuring the luminescence intensity with
an LB960luminometer (Berthold Technologies).
14. (a) The acidic hydroxyl oxygen atom of 1,1-bistrifluoromethyl-carbinol is
located in a favorable position for donating a proton to the N
e imidazole
nitrogen of His435 and results in strong hydrogen-bonding interaction.
a
Bennett, D. J.; Carswell, E. L.; Cooke, A. J.; Edwards, A. S.; Nimz, O. Curr. Med.
Chem. 2008, 15, 195; (b) The Amgen group reported that blocking the hydroxyl
group significantly reduced the potency in a series of T0901317 derivatives,
which indicates that the 1,1-bistrifluoro-methylcarbinol moiety plays an
Compound
Dose
(mg/kg)
ABCA1
intestine
(fold)
ABCA1
peripheral
blood
ABCG5
intestine
(fold)
ABCG8
intestine
(fold)
important role in LXR
a activation. These reports prompted us to design the
(fold)
disconnection of the C–O–C bond of 1,3-dihydroisobenzofuran. Li, L.; Liu, J.;
Zhu, L.; Cutler, S.; Hasegawa, H.; Shan, B.; Medina, J. C. Bioorg. Med. Chem. Lett.
2006, 16, 1638.
T0901317
62
10
10
30
100
3.4⁄
2.1⁄
4.0⁄
5.2⁄
1.3
1.0
0.8
1.0
1.5
1.2
1.5
1.3
1.4
1.3
1.4
1.3
15. Srivastava reported that the hyperlipidemic Bio F1B hamster has been proven to
be an adequate atherosclerosis model for the evaluation of lipid-modulating
agents. Srivastava, R. A. K. Atherosclerosis 2011, 214, 86–93. However, the above
experimental condition requires 21 weeks for the evaluation. To reduce the
experimental period to 10 weeks, we conducted our established protocol as
described in the following Letter: Ikenoy, M.; Yoshinaka, Y.; Kobayashi, H.;
⁄p<0.05; the statistical analysis was conducted using Dunnett’s test.
The fold value was calculated relative to the control.