Discovery and optimization of a novel series of liver X receptor-α agonists

Article abstract of DOI:10.1016/j.bmcl.2005.12.015

A novel series of hexafluorocarbinols were discovered as potent activators of the liver X receptor-α using a fluorescence polarization assay. Structure-activity relationship study led to the identification of compounds that are more potent agonists than t

Full text of DOI:10.1016/j.bmcl.2005.12.015

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1638–1642
Discovery and optimization of a novel series of liver X
receptor-a agonists
Leping Li,a,* Jiwen Liu,a Liusheng Zhu,a Serena Cutler,a Hirohiko Hasegawa,b
Bei Shana and Julio C. Medinaa
aAmgen Inc., 1120 Veterans Boulevard, South San Francisco, CA 94080, USA
bDainippon Sumitomo Pharma Co., Ltd, 1-98, Kasugadenaka 3-Chome, Konohana-ku, Osaka 554-0022, Japan
Received 14 November 2005; revised 4 December 2005; accepted 6 December 2005
Available online 28 December 2005
Abstract—A novel series of hexafluorocarbinols were discovered as potent activators of the liver X receptor-a using a fluorescence
polarization assay. Structure–activity relationship study led to the identification of compounds that are more potent agonists than
the endogenous ligand, 24(S), 25-epoxycholesterol, with similar efficacy. Several compounds, including T0901317, were shown to
have desirable pharmacokinetic profiles suitable for in vivo studies.
Ó 2005 Elsevier Ltd. All rights reserved.
Liver X receptors (LXRa and LXRb) belong to a super-
family of nuclear hormone receptors that function as
transcription factors.1 LXRa is expressed at high level
in liver, adipose tissue, and macrophages, whereas LXRb
is expressed ubiquitously. The discovery of oxidized
derivatives of cholesterol as the endogenous ligands
and the subsequent gene targeting studies in mice provid-
ed strong evidence that LXRa plays an important role in
cholesterol metabolism.2a,b LXRs form heterodimers
with the retinoid X receptors (RXRs) and regulate the
expression, directly or indirectly, of a number of genes
involved in cholesterol and fatty acid metabolism,
including cholesterol transporters and cholesterol
metabolizing enzymes.3,4 Potent modulators of LXRs,
LXRa in particular, with adequate pharmacokinetic
properties in animals are not only of particular value
as research tools for the study of their pharmacological
roles but also provide an opportunity for developing
treatments for a number of pathophysiological states
including dyslipidemia, atherosclerosis, and diabetes.5
ILRKLLQE LXRa binding motif was labeled with the
fluorescent light emitting rhodamine moiety. The FP as-
say measures a compound’s ability to enhance the bind-
ing of the co-activator to LXRa. A unique compound 1
(Fig. 1, N-methyl-N-[4-(2,2,2-trifluoro-1-hydroxy-1-tri-
fluoromethyl-ethyl)-phenyl]-benzenesulfonamide) was
found to be a potent and efficacious activator of LXRa.
Using tritium-labeled compound 1 in a ligand competi-
tion assay, we also demonstrated that the cold com-
pound and 24(S),25-epoxycholesterol competed
effectively for the binding of radiolabeled ligand and
suggested that endogenous ligands bind at the same site
of the LXRa receptor.6 We carried out an extensive
structure–activity relationship (SAR) study to fully
explore the potential of this lead series and to identify
compounds with adequate pharmacokinetic properties
for continued investigation in in vivo models. One of
these compounds, 2 (T0901317), has been the prototyp-
ical LXRa agonist used extensively in a number of
important studies by us and others.6,7a–c In this paper,
In search of nonsteroidal LXR activators, we screened
our compound collection using a fluorescence polariza-
tion (FP) assay. In this homogeneous biochemical assay,
a fragment of the co-activator (SRC1) containing the
F3C
F3C
OH
OH
O
O
O
O
CF3
CF3
S
S
N
N
CH3
Keywords: Hexafluorocarbinol; Liver X receptor-a; LXRa agonist;
T0901317; Cholesterol transporter.
CF3
2 (T0901317)
1
*
Corresponding author. Tel.: +1 650 244 2522; fax: +1 650 244
2015; e-mail: lepingl@amgen.com
Figure 1. LXRa agonists.
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.12.015

L. Li et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1638–1642
1639
we provide a detailed account of the identification of
T0901317 and other potent LXRa activators.8
O
O
b
a
OH
O
OCH3
F3C
13
Cl
Cl
S
O O
S
O O
The 1,1,1,3,3,3-hexafluoro-2-hydroxy-propan-2-yl group
present in the structure was most conveniently introduced
by a Friedel–Crafts type of reaction between an
aniline and hexafluoroacetone hydrate in the presence
of an acid, as shown in Scheme 1.9 The R group on aniline
4 can be introduced prior to the hexafluoro-2-hydroxy-
propan-2-yl moiety or it can be installed by alkylation
of the sulfonamide intermediate 5, which allows for
easy structural diversification (to 6a–h). The ring-fused
analogs (8a–d) were synthesized following similar
procedures by starting with indoline, 1,2,3,4-tetrahydro-
quinoline, 3,4-dihydro-2H-benzo[1,4]oxazine, and 2,3,4,
5-tetrahydro-1H-benzo[b]azepine, respectively. In addi-
tion to various sulfonamide analogs, several amides
and carbamates (represented by 7a and 7b) were prepared.
11
10
OH
CF3
CH3
OCH3
c
CH3
N
N
Ph
S
O O
Ph
S
O O
12
Scheme 2. Reagents and conditions: (a) TMSCHN2, MeOH, rt, 5 h,
88%; (b) N-methylaniline, 2,6-lutidine, acetone, 60 °C, 24 h, >80%; (c)
CF3-TMS, TBAF, THF, rt, 24 h, 50%.
F3C
F3C
F3C
OH
OH
OH
a
O2N
b
H2N
CF3
CF3
CF3
An alternative route was required in order to install the
1,1,1,3,3,3-hexafluoro-2-hydroxy-propan-2-yl group to
an electronically deficient phenyl ring, such as the re-
verse sulfonamide compound represented by 13. The
presence of an electron-withdrawing group on the phen-
yl ring retards the regiospecific introduction of the hexa-
fluorocarbinol unit. We developed a convenient method,
in which two trifluoromethyl units from trifluoromethyl-
trimethylsilane (TMS-CF3) were added to a methyl ester
in the presence of tetra-n-butylammonium fluoride, to
complete the transformation (Scheme 2).
14
15
16
18
F3C
CH3
F3C
H
N
OH
CF3
OH
c
d
Ph
Ph
N
S
S
CF3
O O
O O
17
Scheme 3. Reagents and conditions: (a) HNO3 (90%), H2SO4 (concd),
0 °C, 30 min, rt, 2 h, 90%; (b) H2, Pd/C, EtOH, rt, 6 h, 100%; (c)
PhSO2Cl, 2,6-lutidine, acetone, 60 °C, 24 h, 80%; (d) CH3I, NaH
(1 equiv), DMF, rt, 2 h, 60%.
To prepare compounds in which 1,1,1,3,3,3-hexafluoro-
2-hydroxy-propan-2-yl and the sulfonamido substituent
are in a 1,3-relationship on the central phenyl ring, we
employed a nitration step (from 14 to 15) to achieve
the desired regiospecificity, taking advantage of the
meta-directing ability of the hexafluorocarbinol group
(Scheme 3). The nitro compound was reduced to aniline
16, which, in turn, was converted to the N-methyl sul-
fonamide 18 using a sequence similar to that described
in Scheme 1.
transfected with an expression plasmid for human
LXRa and a luciferase reporter plasmid containing an
LXR response element. The cells were treated with vary-
ing concentrations of test compounds. The EC50 values
and maximal transactivation at LXRa are reported
here. However, it is worth mentioning that these ana-
logs, while they exhibit high specificity over all other
nuclear receptors tested, are generally not LXR subtype
specific.6a
The hexafluorocarbinol moiety was found to be impor-
tant for LXRa activation, as data in Table 1 indicated.
Replacing one of the trifluoromethyl groups with a
methyl group gave a compound that was fully effica-
cious but with much reduced potency.10 Blocking of
the hydroxyl group also significantly lowered potency.10
The presence of both trifuoromethyl groups is believed
as not only important to maintain the proper orienta-
tion of the hydroxyl group but also necessary to ensure
the strong acidity of the OH group. We postulated that
the relatively high acidity (pKa was determined to be 8.5)
of the hydroxyl group might be important for interac-
tion with basic residues on the LXRa receptor providing
the anchoring interaction. This hypothesis was support-
ed by a recent X-ray crystallographic study reported by
Stefansson et al.11
The lead optimization was primarily guided by a cell-
based assay for assessing a compound’s ability to acti-
vate the LXRa receptor. HEK293 cells were transiently
F3C
F3C
OH
OH
b
CF3
CF3
a
O
O
S
HN
R
Ar
N
if R = H
HN
R
H
5
3
4
b
c
d
F3C
F3C
OH
OH
O
CF3
CF3
O
O
S
L
N
Ar
N
R
R
6a-h
7a-b
Unlike the head hexafluorocarbinol group, the benzene-
sulfonamide region can tolerate a wide range of struc-
tural modifications. As shown in Table 2, a variety of
substituted benzenesulfonamides were potent agonists
Scheme 1. Reagents and conditions: (a) hexafluoroacetone trihydrate,
p-TsOH, 130 °C, 14 h, 80%; (b) ArSO2Cl, 2,6-lutidine, acetone, 60 °C,
24 h, 90%; (c) R-I, NaH (1 equiv), DMF, rt, 2 h, 60%; (d) Bz-Cl, or
i-BuOCOCl and Hunig’s base, 80%.

1640
L. Li et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1638–1642
Table 1. LXRa agonist potency and intrinsic efficacy of compounds
varying the carbinol function
Table 2. Impact of N-substituents on LXRa agonist potency and
intrinsic efficacy
R
F3C
OH
O
O
S
CF3
N
CH3
Y
Compound
LXRa12
Y
Compound
LXRa12
Max activation
R
EC50 (lM) Max activation
EC50 (lM)
NC
O
O
S
24(S),25-Epoxycholesterol
0.5
5.4
5.3
0
6a
6b
0.7
6.2
6.2
N
CH3
F3C
O
S
O
OH
1
0.2
NC
0.02
N
CF3
H
CH3
OH
1a
>5
O
O
S
CF3
N
6c
6d
6e
6f
2
4.4
4.9
5.2
6.3
H3C
CH3
OH
NC
1b
5.0
1
5.1
4.5
CF3
F3C
O
O
O
O
S
S
H2N
0.08
0.03
0.6
N
OCH3
CF3
1c
CH3
O
O
Ph
S
N
CH3
of LXRa with high efficacy. In general, substitutions at
the ortho- and para-positions to the hexafluorocarbinol
group tended to lower potency, while meta-substitution
enhanced potency, as exemplified by entries 6a, 6b, and
6c. Other meta-substituents were introduced in order to
modulate the physicochemical properties of the mole-
cules. A sulfonamide was not a prerequisite for activity,
since benzamide (7a) and isobutoxycarbamate (7b) were
shown to maintain LXRa activity. We also found that
the LXRa agonist potency and efficacy were well re-
tained when the sulfonamide and amide linkages were
reversed, indicating that the ability to maintain the sub-
stituents in the correct special orientation is essential.
O
O
S
S
S
N
CH3
O
O
N
S
6g
0.04
0.1
0.1
0.1
0.2
6.1
3.9
5.1
5.3
4.0
CH3
CH3
O
7a
N
CH3
O
H3C
7b
Table 3 summarized the results from a broad survey of
substituents and substitution patterns around the cen-
tral phenyl ring. A variety of N-alkyl substituents were
well tolerated. Fused ring systems that resulted from
connecting the N-alkyl group to the adjacent phenyl
group, as in 8a through 8d, maintained LXRa activity.
Presumably the ring structures, while restricting rotation
freedom, were able to preserve the active conformation.
However, the central phenyl ring is sensitive to structur-
al changes. A substituent ortho to the hexafluorocarbi-
nol head group was detrimental to activity, as seen in
example 9b. A 1,4-spacial arrangement was preferred
for the hexaflurocarbinol and the alkylamino substituent
(1 vs 18). Insertion of a methylene unit between the two
groups caused a lowering in potency, as seen for 19.
N
O
CH3
CH3
CH3
N
13
S
O O
CH3
N
13a
O
at 5 mg/kg. The exposure results are summarized in
Table 4.13
While the highest AUC value was observed for com-
pound 6d, the N-trifluoroethyl analog T0901317 ap-
peared to be advantageous when considering the
exposure and the potency. T0901317 produced excellent
exposures following oral dosing at 5 mg/kg in mice
(AUC = 3421 lg h/L and t1/2 = 3 h, n = 3) and in dogs
(AUC = 7301 lg h/L and t1/2 = 5.8 h, n = 2).13
Selected potent and efficacious LXRa agonists were
characterized for their pharmacokinetic properties in
animals in order to identify compounds that are suitable
for acute and chronic dosing for further pharmacologi-
cal investigations in vivo. The compounds were dosed
to male Sprague–Dawley rats (n = 3) by oral gavage

L. Li et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1638–1642
1641
Table 3. Impact of central ring substituents on LXRa agonist potency
and intrinsic efficacy
Table 4. Exposure after oral dosing at 5 mg/kg in rats
Compound
AUC
(lg h/L)a
Average concentration
(fold over EC50)
F3C
Ar
OH
O
O
T0901317
6g
6d
5920
3204
51
18
16
4
S
CF3
N
14,876
1899
1037
R
6k
11a
1
Compound
LXRa12
Max activation
Ar
a AUC, area under the curve.
N
EC50 (lM)
R
the 1,4-relationship to the head hexafluorocarbinol
group on the central phenyl ring. Our SAR investiga-
tions have led to the identification of a number of highly
potent and efficacious LXRa agonists. Several of them,
including T0901317, were shown to have adequate phar-
macokinetic profiles suitable for in vivo pharmacologi-
cal investigations.
5a
4
4.3
6.1
N
H
T0901317
0.01
N
CF3
Acknowledgments
N
6h
0.008
4.8
The authors thank Drs. Juan Jaen and Jin-Long Chen
for their guidance and Dr. Matthew Wright for conduct-
ing pharmacokinetic studies.
8a
8b
8c
0.05
0.03
0.1
3.8
3.9
4.5
N
N
References and notes
1. Willy, P. J.; Umesono, K.; Ong, E. S.; Evans, R. M.;
Heyman, R. A.; Mangelsdorf, D. J. Gene Dev. 1995, 9,
1033.
2. (a) Repa, J. J.; Mangelsdorf, D. J. Nat. Med. 2002, 8,
1243; (b) Janowski, B. A.; Willy, P. J.; Devi, T. R.; Falck,
J. R.; Mangelsdorf, D. J. Nature 1996, 383, 728; (c)
Lehmann, J. M.; Kliewer, S. A.; Moore, L. B.; Smith-
Oliver, T. A.; Oliver, B. B.; Su, J. L.; Sundseth, S. S.;
Blanchard, D. E.; Spencer, T. A.; Willson, T. M. J. Biol.
Chem. 1997, 272, 3137.
N
N
O
8d
9a
0.01
0.2
4.2
5.2
3. Janowski, B. A.; Grogan, M. J.; Jones, S. A.; Wisely, G.
B.; Liewer, S. A.; Corey, E. J.; Mangelsdorf, D. J. Proc.
Natl. Acad. Sci. USA 1999, 96, 266.
N
4. Peet, D. J.; Turley, S. D.; Ma, W.; Janowski, B. A.;
Lobaccaro, J.-M. A.; Hammer, R. E.; Mangelsdorf, D. J.
Cell 1998, 93, 693.
CH3 Cl
5. For a recent review on the therapeutic utilities of LXR
modulators, see: Collins, J. L. Curr. Opin. Drug Discov.
Dev. 2004, 24, 1755.
9b
>5
0
N
Cl
CH3
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) Schultz, J. R.; Tu, H.; Luk, A.; Repa, J. J.;
Medina, J. C.; Li, L.; Schendner, S.; Wang, S.; Thoolen,
M.; Mangelsdorf, D. J.; Lustig, K. D.; Shan, B. Gene Dev.
2000, 14, 2831.
CH3
N
18
19
1
3
4
CH3
N
3.7
7. For selected studies with T0901317, see: (a) Cao, G.;
Liang, Y.; Broderick, C. L.; Oldanm, B. A., et al. J. Biol.
Chem. 2003, 278, 1131; (b) Efanov, A. M.; Sewing, S.;
Bokvist, K.; Gromada, J. Diabetes 2004, 53(Suppl. 3), S75;
(c) Koldamova, R. P.; Lefterov, I. M.; Staufenbiel, M.;
Wolfe, D.; Huang, S.; Glorioso, J. C.; Walter, M.; Roth,
M. G.; Lazo, J. S. J. Biol. Chem. 2005, 280, 4079.
8. A preliminary disclosure of this work was made earlier:
Medina, J. C.; Li, L.; Cutler, S.; Hasegawa, H.; Liu, J.;
Zhu, L.; Schultz, J. R.; Shan, B. Book of Abstracts 221st
National Meeting of the American Chemical Society San
From our systematic structure–activity relationship
investigation, we found that the hexafluorocarbinol moi-
ety is essential for activation of the LXRa receptor.
Maintaining the strong acidity and preserving the prop-
er orientation of the hydroxyl group are key factors. A
variety of sulfonamido, acetamido, and alkoxyacetami-
do substituents are allowed and are preferred to be in

1642
L. Li et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1638–1642
Diego, CA April 1–5, 2001; American Chemical Society:
Washington DC, 2001; MEDI-180.
9. Farah, B. S.; Gilbert, E. E.; Sibilia, J. P. J. Org. Chem.
1965, 30, 998.
10. Note: Compound 1a was obtained from the reduction of
the corresponding trifluoroacetophenone with sodium
borohydride in methanol, while compound 1b was pro-
duced by the addition of a methyl Grignard reagent to the
same trifluoroacetophenone.
12. HEK293 transient transfection cell assay using a wild-type
LXRa expression plasmid and a luciferase reporter gene
containing two copies of an LXR response element. EC50:
effective concentration to elicit 50% of the maximum
response. Values are means of three experiments; standard
derivation is 25%. Maximal activation value represents
the maximal amount of luciferase activity produced by
a compound relative to the activation produced by a
vehicle control.
11. 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.
13. For the pharmacokinetic studies, compounds were
formulated as suspension in saline containing 1%
Tween 80 and 1% methylcellulose, and were given by
oral gavage.