2-Aryl-N-acyl indole derivatives as liver X receptor (LXR) agonists

Article abstract of DOI:10.1016/j.bmcl.2007.06.017

Structure-activity relationship studies on a series of Boc-indole derivatives as LXR agonists are described. Compound 1 was identified as an LXR agonist through structure-based virtual screening followed by high-throughput gene profiling. Replacement of t

Full text of DOI:10.1016/j.bmcl.2007.06.017

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4442–4446
2-Aryl-N-acyl indole derivatives as liver X receptor (LXR) agonists
Sunil Kher,a Kirk Lake,a Ila Sircar,a Madhavi Pannala,a Farid Bakir,a James Zapf,b
Kui Xu,c Shao-Hui Zhang,c Juping Liu,c Lisa Morera,c Naoki Sakurai,c
Rick Jackc and Jie-Fei Chenga,*
aDepartment of Chemistry, Tanabe Research Laboratories USA, Inc., 4540 Towne Centre Court, San Diego, CA 92121, USA
bDepartment of Computational Discovery, Tanabe Research Laboratories USA, Inc., 4540 Towne Centre Court,
San Diego, CA 92121, USA
cDepartment of Biology, Tanabe Research Laboratories USA, Inc., 4540 Towne Centre Court, San Diego, CA 92121, USA
Received 30 March 2007; revised 31 May 2007; accepted 4 June 2007
Available online 10 June 2007
Abstract—Structure–activity relationship studies on a series of Boc-indole derivatives as LXR agonists are described. Compound 1
was identified as an LXR agonist through structure-based virtual screening followed by high-throughput gene profiling. Replace-
ment of the indan linker portion in 1 with an open-chain linker resulted in compounds with similar or improved in vitro potency
and cellular functional activity. The Boc group at the N-1 position of the indole moiety can be replaced with a benzoyl group. The
SAR studies led to the identification of compound 8, a potent LXRb agonist with an EC50 of 12 nM in the cofactor recruitment
assay.
Ó 2007 Elsevier Ltd. All rights reserved.
Liver X receptors (LXRa and LXRb, or NR1H3 and
NR1H2)) belong to the type 2 family of the nuclear hor-
mone receptor superfamily that function as transcrip-
tion factors.1 LXRa is expressed at a high level in
liver, adipose tissue, and macrophages, whereas LXRb
is ubiquitously expressed. LXRs form heterodimers with
retinoid X receptors (RXR) and regulate the expression
of a number of genes involved in cholesterol homeosta-
sis2 and fatty acid metabolism.3 LXR was also recently
reported as a glucose sensor involved in liver carbohy-
drate metabolism.4 Upon agonist binding, the DNA
binding domain (DBD) of LXR interacts with LXR
response elements on target genes to initiate the tran-
scription process. One LXR target gene is the ATP-bind-
ing cassette transporter ABCA1, which is involved in
reverse cholesterol transport (RCT) from macrophages
to high-density lipoproteins (HDL) in the plasma.5
Increasing RCT through LXR agonism is a potential
therapeutic approach for a number of pathophysiologi-
cal states including dyslipidemia and atherosclerosis.6
Since cholesterol accumulation in pancreatic b-cells is re-
ported to result in b-cell dysfunction in type 2 diabetes,7
increasing expression of b-cell ABCA1 through LXR
agonism would reduce the cholesterol accumulation, im-
prove insulin secretion and glucose homeostasis in b-cell,
and would be a viable approach for diabetes. A number
of small molecule LXR agonists such as GW3965 (I)8a,b
and T0901317 (II)8c have been described (Fig. 1). Some
LXR agonists were shown to raise plasma HDL levels
in mice and showed anti-diabetic activity in rodent mod-
els of type 2 diabetes.9,10 However, most LXR agonists
also induce the expression of sterol response element
binding protein-1c (SREBP1c) gene, which controls the
entire fatty acid biosynthetic pathway and increases lipo-
genesis and hepatic steatosis.
In order to identify selective LXR modulators that
would have the potential to increase reverse cholesterol
transport with minimum or absence of lipogenic activ-
ity, we first conducted a structure-based virtual screen-
ing using LXRb homology model based on
crystallography. Compounds (ꢀ1500) with high docking
score were selected and profiled on twelve LXR regu-
lated genes including LXRa gene using high-throughput
genomic technology (HTG, Inc., Tucson, AZ).11 An in-
dan derivative 1 (Fig. 1) was found to up-regulate
ABCA1 gene expressions to the extent similar to
T0901317 or GW3965 (Table 3), while showing less reg-
ulation on SREBP-1c genes in THP-1 differentiated
Keywords: Liver X receptor; LXR agonist; Indole; ABCA1; SREBP-1c.
*
Corresponding author. Tel.: +1 858 622 7029; fax: +1 858 558
9383; e-mail: jcheng@trlusa.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.06.017

S. Kher et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4442–4446
4443
CF3
OH
CF3
O
O
S
O
O
NH
S
HOOC
N
O
N
N
O
CF3
T0901317 (II)
O
Cl
GW3965 (I)
1
CF3
Figure 1. LXR modulators.
macrophages. Compound 1 was subsequently confirmed
to be an LXR agonist in both a cofactor recruitment as-
say and a transactivation reporter assay (Table 3). Initial
SAR indicated that an acyl indole moiety is essential for
activity.12 We herein describe the chemistry efforts on
the indan linker modifications to explore the struc-
ture–activity relationship.
rank-ordered by EC50 value for LXRb (Tables 1 and
2). As observed with indan linker compounds,12 these
compounds tended to show slight selectivity for LXRb
over LXRa in the HTRF assay (Table 3). SAR com-
pounds with good potency against LXRb and/or with
good selectivity were further profiled in a transactivation
reporter14 assay and cellular functional assays15,16 (Table
3). Early SAR indicated that the indan moiety in 1 serves
only as a linker to bring the Boc-indole and the sulfon-
amide functionalities together in a certain shape. We
envisioned that a ring open system would be a viable sur-
rogate for the indan ring. Indeed, phenethyl amine linker
at either C-3 or C-4 position retained the potency with
the C-4 substitution being slightly more potent than
the corresponding C-3 substituted analog (11 vs 10) (Ta-
ble 1). Compounds with shorter (CH2) and longer
[(CH2)3] spacers were tolerated (12, 13), whereas aniline
derivative (14) suffered a 7-fold loss in potency. The 2-
fold increase in potency seen with 2-thienyl sulfonamide
15 (EC50 = 58 nM) followed a similar SAR trend ob-
served in the aminoindan linker series.12 Substitution
on the thiophene ring (16) reduced the potency.
Indole sulfonamide compounds with an open-chain lin-
ker replacing the indan ring were prepared according to
Scheme 1. The requisite substituted bromobenzene sul-
fonamides (3) were obtained from commercially avail-
able amines 5 via sulfonylation. Suzuki coupling of
intermediates 3 with N-Boc-2-indole boronic acid 2 gave
rise to the desired products 4. Amide derivatives 6 were
prepared in a similar fashion.
Suzuki reaction of N-Boc-2-indole boronic acid 2 with
para-substituted bromobenzene provided compounds 7
with functionality other than sulfonamides or benz-
amides. Removal of the Boc group at the N-1 position
of the indole moiety followed by benzoylation resulted
in benzamide derivatives 8 and 9.
As shown in Table 2, benzamide derivatives appear to
be less potent than the corresponding sulfonamides if
the spacers are methylene or ethylene (11, 12 vs 17,
The compounds prepared above were tested in LXRa
and b HTRF cofactor recruitment assays13 and were
NHCOAr
NH2
n
n
c, b
N
O
Br
5
O
6
a
NHSO2Ar
n
B(OH)2
b
NHSO2Ar
+
n
Br
N
N
O
O
O
O
3
4
2
d
R
e
R
8: R = CH2CH2CN
9: R = CH2CH2NHCOPh
N
N
O
O
O
7
Scheme 1. Reagents and conditions: (a) ArSO2Cl, DIEA, CH2Cl2, rt; (b) 2Pd(PPh3)4, Cs2CO3, toluene/ethanol/water (10:1:1); (c) ArCOCl, DIEA,
CH2Cl2, rt; (d) p-R-PhBr, Pd(PPh3)4, Cs2CO3, toluene/ethanol/water (10:1:1); (e) TFA, CH2Cl2, rt then PhCOCl, DIEA, CH2Cl2, rt.

4444
S. Kher et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4442–4446
Table 1. Open-chain analogs of compound 1
plays a key role in the binding to the LBD. It is worth
noting that amide bond orientation in the linker portion
did not affect the activity (18 vs 30).
n
NHSO2R
4
N
Boc
Previously we found that the replacement of the Boc
group with even closely related carbamate, amide or
urea moieties led to a complete or near-complete loss
of activity in the indan series.12 However, in the current
open-chain linker series, the Boc moiety can be replaced
by a benzoyl group. For instance, compound 9 with an
N-benzoyl group showed similar activity as its Boc ana-
log (30). Compound 8, which possesses a cyano group
on the terminal and a benzoyl group on the indole nitro-
gen, demonstrated a 10-fold increase in potency over the
corresponding Boc derivative 22.
Compound Attachment
position
n
R
EC50 (lM)a
LXRb
1
10
11
12
13
14
15
16
—
3
—
2
—
0.19
0.42
0.12
0.22
0.35
1.32
0.058
Ph
Ph
Ph
Ph
Ph
4
2
4
1
4
3
4
0
4
2
2-Thienyl
4
2
5-(2-Pyridinyl)- 0.23
2-thienyl
A few selected compounds with varying activity and
selectivity profiles in cofactor recruitment assays were
further tested in a reporter transactivation assay for
LXRb as well as other cell-based functional assays. In
general, compounds with good potency in the cofactor
recruitment assay maintained good potency in the cellu-
lar reporter assay and the cholesterol efflux assay with
some exceptions (e.g., 25). Most of compounds in the
series are partial agonists as compared to GW3965 in
the LXRb reporter transactivation assay. Compound
8, which showed similar potency in the cofactor recruit-
ment assay but with a slightly better selectivity profile
(LXRb over LXRa) than GW3965 and T0901317, dem-
onstrated almost identical gene expression profiles as
GW3965 for ABCA1 and SREBP1c. LXRb reporter
transactivation assay appears to correlate with ABCA1
gene regulation as well as cholesterol efflux in THP-1
cells (Table 3). Compound 8 induced more significant
lipogenesis than GW3965, indicating that the LXRa
activity may be a more important contributing factor
to the lipogenesis than the LXRb activity. However,
due to the lack of LXRa reporter transactivation data,
no clear correlation can be drawn among LXRa potency
and SREBP-1c regulation or lipogenesis.
a HTRF assay results are mean values of duplicate samples in a single
experiment.
Table 2. Linear linker derivatives as LXR agonists
R2
N
R1
Compound R1
R2
EC50 (lM)a
LXRb
17
18
19
20
21
22
23
24
25
26
27
28
29
30
8
Boc
–CH2NHCOPh
–(CH2)2NHCOPh
–(CH2)3NHCOPh
–CH2NH2
1.8
Boc
Boc
Boc
Boc
Boc
Boc
Boc
Boc
Boc
Boc
Boc
Boc
Boc
1.75
0.36
5.6
–CH2CN
–(CH2)2CN
0.3
0.13
0.38
0.42
0.22
0.057
4.9
–(CH2)3OH
–O(CH2)2OH
–(CH2)2CO2H
–(CH2)2CONH2
–(trans-CH2@CH)CONH2
–(CH2)2CONHMe
0.49
–(CH2)2CONHCH2CH(CH3)2 1.51
–(CH2)2CONHPh
1.78
0.012
1.78
COPh –(CH2)2CN
COPh –(CH2)2NHCOPh
Docking studies suggested that compound 1 fits nicely in
the LXRb ligand binding domain (Fig. 2). The indole
aromatic ring system is close to Trp457 in the AF-2 (he-
lix 12) and forms a p–p interaction while the Boc car-
9
a HTRF assay results are mean values of duplicate samples in a single
experiment.
bonyl group forms
a hydrogen bond with the
conserved His435 in helix 10/11, thereby locking it into
an agonistic conformation.17 Compound 8 overlaps well
with 1 and fits nicely in the active site. The benzoyl
group at the indole nitrogen occupies the hydrophobic
region in a similar fashion as the Boc group. The 2-phe-
nyl group in 8 or the indan ring in 1 seems to be close to
the Phe 271, while the terminal cyano group is located in
proximity to Leu314 via a hydrogen bonding interac-
tion. The sulfonamide group may have an interaction
with Arg319 (not shown).
18). However, when the spacer is extended to a propyl-
ene moiety, both sulfonamides and benzamide are
equally potent (19 vs 13). Free amino group-containing
compound 20 is about 10-fold less potent than the cor-
responding cyano derivative (21). Increase in the linker
length in the cyano derivative improved the activity
slightly (22 vs 21). A hydroxyl group (23, 24) or a car-
boxylic acid group (25) on the terminal seems to be well
tolerated. A dramatic improvement was observed with
the primary amide (26), which showed a 4-fold improve-
ment in the activity over the corresponding carboxylic
acid. Substitutions on the amide nitrogen resulted in a
decrease in the potency (28, 29, and 30 vs 26). Introduc-
tion of a,b-unsaturation (27) led to more than 80-fold
loss of the activity (EC50 = 5.9 lM), indicating that the
hydrogen-bond interaction on the terminal portion
In summary, we have described a series of potent LXR
agonists based on the initial hit compound 1 identified
by using a combination of virtual screening and high-
throughput gene profiling. The rigid indan linker in
the initial hit compound 1 could be replaced with an
open-chain linker and the potency was retained or

S. Kher et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4442–4446
Table 3. Profile of selected LXR agonists
4445
Compound
HTRF
a,b
EC50
HTRF
a,b
EC50
Reporter
a
EC50
ABCA1
inductiond
(THP-1)
SREBP-1c
inductiond
(HepG2)
Efflux
a
EC50
Maximum
effluxd
Lipogenesis
inductiond
(HepG2)
LXRa
LXRb
LXRb
(THP-1)
1
22
1.5
0.206
0.155
0.225
0.057
0.012
0.035
0.015
0.065 (43%)c
0.576 (45%)c
4.58 (76%)c
0.285 (70%)c
0.042 (39%)c
0.015 (100%)c
0.011 (84%)c
8.4
6.0
3.1
2.5
0.680
2.47
7.0
1.6
2.2
1.4
1.9
1.7
1.8
1.6
3.1
2.5
1.2
2.7
4.0
2.7
3.6
1.19
1.15
1.11
0.22
0.367
0.04
25
26
5.5
5.7
Nde
4.0
2.4
8
10.3
10.7
8.4
5.3
5.2
0.051
0.010
0.033
GW3965
T0901317
6.7
a Values shown are in micromolar.
b All compounds including T0901317 are full agonists compared with GW3965 in both LXRa and LXRb HTRF assays.
c Values shown are in percent of maximum induction by GW3965.
d Values shown are in ratio of compound (10 lM) versus control (vehicle only).
e Not determined.
4. Mitro, N.; Mak, P. A.; Vargas, L.; Godio, C.; Hampton, E.;
Molteni, V.; Kreusch, A.; Saez, E. Nature 2007, 445, 219.
5. Tontonoz, P.; Mangelsdorf, D. J. Mol. Endocrinol. 2003,
17, 985.
6. (a) Attie, A. D.; Kastelein, J. P.; Hayden, M. R. J. Lipid
Res. 2001, 42, 1717; (b) Repa, J. J.; Mangelsdorf, D. J.
Nat. Med. 2002, 8, 1243.
7. Brunham, L. R.; Kruit, J. K.; Pape, T. D.; Timmins, J. M.;
Reuwer, A. Q.; Vasanji, Z.; Marsh, B. J.; Rodrigues, B.;
Johnson, J. D.; Parks, J. S.; Verchere, C. B.; Hayden, M.
R. Nat. Med. 2007, 13, 340.
8. (a) Collins, J. L.; Flvush, 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.; Wilson,
T. M. J. Med. Chem. 2002, 45, 1963; (b) Jaye, M. C.;
Krawiec, J. A.; Campobasso, N.; Smallwood, A.; Qiu, C.;
Lu, Q.; Kerrigan, J. J.; De Los Frailes Alvaro, M.;
Laffitte, B.; Liu, W.; Marino, J. P., Jr.; Meyer, C. R.;
Nichols, J. A.; Parks, D. J.; Perez, P.; Sarov-Blat, L.;
Seepersaud, S. D.; Steplewski, K. M.; Thompson, S. K.;
Wang, P.; Watson, M. A.; Webb, C. L.; Haigh, D.;
Caravella, J. A.; Macphee, C. H.; Wilson, T. M.; Collins,
J. L. J. Med. Chem. 2005, 48, 5419; (c) Li, L.; Liu, J.; Zhu,
L.; Cutler, S.; Hasegawa, H.; Shan, B.; Medina, J. C.
Bioorg. Med. Chem. Lett. 2006, 16, 1638.
9. Cao, G.; Liang, Y.; Broderick, C. L.; Oldham, B. A.;
Beyer, T. P.; Schmid, R. J.; Zhang, Y.; Stayrook, K. R.;
Suen, C.; Otto, K. A.; Miller, A. R.; Dai, J.; Foxworthy,
P.; Gao, H.; Ryan, T. P.; Jiang, X. C.; Burris, T. P.;
Eacho, P. I.; Etgen, G. J. J. Biol. Chem. 2003, 278, 1131.
10. Michael, L. F.; Schkeryantz, J. M.; Burris, T. P. Mini-Rev.
Med. Chem. 2005, 5, 729.
Figure 2. An overlay of compounds 1 and 8 docked into the active site
of ligand-binding domain of LXRb based on X-ray cocrystal structure
of GW3965.
improved. The sulfonamide functionality, which may
act as a hydrogen bonding donor in the initial hit, could
be replaced by hydrogen bonding acceptors such as a cy-
ano group. These compounds tend to be moderately
selective for LXRb in HTRF cofactor recruitment assay.
Replacement of N-Boc moiety in the indole portion with
a benzoyl group not only retained or improved the po-
tency, but also improved the metabolic stability (not
shown). Compound 8 was identified as the most potent
and stable compound with an EC50 of 12 nM. Although
they are active in cell-based functional assays such as
cholesterol efflux in THP-1 macrophages, they also up-
regulate the undesirable SREBP1-c gene expression
and cause lipogenesis in the HepG2 cells.
11. High-throughput genomics screening was performed using
the ArrayPlate technology in a contract service provided
by HTG, Inc. The ArrayPlate technology is a quantitative
nuclease protection assay for multiplex gene profiling
(http://www.htgenomics.com/).
12. Bakir, F.; Kher, S.; Pannala, M.; Wilson, N.; Nguyen, T.;
Sircar, I.; Takedomi, K.; Fukushima, C.; Zapf, J.; Xu, K.;
Zhang, S.-H.; Liu, J.; Morera, M.; Schneider, L.; Sakurai,
S.; Jack, R.; Cheng, J.-F. Bioorg. Med. Chem. Lett. 2007,
17, 3473.
13. The HTRF cofactor peptide recruitment assay was mod-
ified from a previous report. Briefly, polyhistidine-tagged
human LXRa (2 nM) or LXRb (1 nM) ligand-binding
domain (Roche Diagnostics, Indianapolis, IN) was mixed
with the test compound, 20 nM biotin-SRC1 peptide
(Synpep, Dublin, CA), 5 nM streptavidin-allophycocya-
nin, and europium-labeled anti-polyhistidine antibody (1
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. 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. U.S.A. 1999, 96, 206.
3. 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.

4446
S. Kher et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4442–4446
and 0.5 nM for a and b, respectively) in 50 mM Tris, pH
acLDL and 200 nM PMA. Cells were then incubated
with vehicle or compound for 6 h in serum-free media
containing 2 mg/ml BSA. Subsequently, 5 lg/ml of APO-
AI was added in fresh, serum-free medium along with
vehicle or compound for an additional 18-h incubation.
Cholesterol efflux was monitored by quantifying the
radioactivity in the cell supernatant and the data are
presented as fold-induction versus control (vehicle only).
Results shown are mean values of triplicate samples in a
single experiment.
7.5, 50 mM KCl, 1 mM EDTA, 0.1% BSA, and 1 mM
DTT. The final volume of the mixture was 40 ll in a 384-
well assay plate. The mixture was incubated at room
temperature for 1 h with shaking. Time-resolved fluores-
cence was measured at 615 and 665 nm on an LJL Analyst
plate reader. The ratio of 665/615 was used to calculate
EC50 values of test compounds.
14. COS-7 cells were transfected for 24 h with plasmids of the
human LXRb receptor (OriGene, Rockville, MD), the
LXR reporter +-GACCAGCAGTAACCTTGACCAGC
AGTAACCTTGACCAGCAGTAACCT (prepared in
house), and the RXRa receptor (OriGene, Rockville,
MD). Subsequently, cells were treated with vehicle or
compound for 16 h. LXR reporter activation was moni-
tored by quantifying the luciferase activity in the cell
lysate. EC50 values were calculated from mean values of
quadruplicate samples in a single experiment.
15. Compound induction of cholesterol efflux was measured
as described previously with small modifications. Briefly,
THP-1 cells were differentiated into macrophages in 48-
well tissue culture plates by 30-h treatment with 200 nM
PMA. The cells were then labeled with 0.6 lCi of 1,2-3H
(N)-cholesterol for 18 h in the presence of 16 lg/ml
16. To measure compound induction of lipogenesis, HepG2
cells in a 48-well tissue culture plate were pre-treated with
vehicle or compound for 24 h. [14C]-glycerol (1 lCi) was
added and the cells were cultured for another 48 h.
Cellular triglycerides were extracted, separated by TLC,
and quantified on a Storm 820 phosphorimager (GE
Healthcare, Giles, UK). Lipogenesis was presented as
fold-induction versus control (vehicle only). Results
shown are mean values of triplicate samples in a single
experiment.
17. 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.;
Wilson, T. M. J. Biol. Chem. 2003, 278, 27138.