Synthesis and SAR of potent LXR agonists containing an indole pharmacophore

Article abstract of DOI:10.1016/j.bmcl.2009.01.004

A novel series of 1H-indol-1-yl tertiary amine LXR agonists has been designed. Compounds from this series were potent agonists with good rat pharmacokinetic parameters. In addition, the crystal structure of an LXR agonist bound to LXRα will be disclosed.

Full text of DOI:10.1016/j.bmcl.2009.01.004

Bioorganic & Medicinal Chemistry Letters 19 (2009) 1097–1100
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and SAR of potent LXR agonists containing an indole pharmacophore
a
d
d
e
David G. Washburn a, , Tram H. Hoang , Nino Campobasso , Angela Smallwood , Derek J. Parks ,
Christine L. Webb b, Kelly A. Frank c, Melanie Nord c, Chaya Duraiswami d, Christopher Evans c,
Michael Jaye b, Scott K. Thompson a
*
a Department of Chemistry, Molecular Discovery Research, GlaxoSmithKline Pharmaceuticals, 709 Swedeland Road, King of Prussia, PA 19406, USA
b Department of Biology, Molecular Discovery Research, GlaxoSmithKline Pharmaceuticals, 709 Swedeland Road, King of Prussia, PA 19406, USA
c Department of Drug Metabolism and Pharmacokinetics, Molecular Discovery Research, GlaxoSmithKline Pharmaceuticals, 709 Swedeland Road, King of Prussia, PA 19406, USA
d Metabolic Pathways Centre for Excellence in Drug Discovery and Computational and Structural Chemistry, Molecular Discovery Research, GlaxoSmithKline Pharmaceuticals,
709 Swedeland Road, King of Prussia, PA 19406, USA
e Molecular Discovery Research GlaxoSmithKline, Research Triangle Park, NC 27709, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A novel series of 1H-indol-1-yl tertiary amine LXR agonists has been designed. Compounds from this ser-
ies were potent agonists with good rat pharmacokinetic parameters. In addition, the crystal structure of
an LXR agonist bound to LXRa will be disclosed.
Received 24 July 2008
Revised 29 December 2008
Accepted 6 January 2009
Available online 9 January 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
LXR
Liver X receptors
Nuclear hormone receptors
Pulmonary inflammation
Diabetes
Atherosclerosis
Liver X receptors (LXR
a
and LXRb) are members of the type 2
cholesterol efflux assays (EC50 = 50 nM); however, it was less
potent in the LXR
FRET assay (EC50 = 175 nM).4–6
family of the nuclear hormone receptors and function as transcrip-
tion factors that mediate cholesterol and lipid metabolism.1 LXRb
is ubiquitously expressed in a variety of different tissues, while
a
As part of an effort to identify LXR agonists with improved
potency, a structure-based design strategy centering around 2
was initiated. Analysis of the X-ray crystal structure of compound
2 and other LXR agonists revealed an unoccupied hydrophobic
pocket adjacent to the phenyl acetic acid region of 2.6,7 Toward this
end, we set out to explore suitable replacements for the phenyl
acetic acid moiety of 2.
Synthesis of the analogs of 2 is outlined in Scheme 1. Reductive
amination of benzaldehyde 4 with 2,2-diphenylethanamine 3
yielded the secondary amine 5. Subsequent treatment of 5 with
base and 3-bromo-1-propanol or 1,3-dibromopropane yielded the
desired alcohol 6 or bromide 7, respectively. Intermediates 6 and
7 served as important building blocks for exploration of replace-
ments for the phenyl acetic acid group. Target heteroaromatic
ethers 8–15 were then prepared by either Mitsunobu coupling of
6 or nucleophilic displacement of alkyl bromide 7.
LXRa is primarily expressed in the liver, adipose tissue, and macro-
phage.1 These ligand-activated transcription factors form heterodi-
mers with retinoid X receptors (RXR) and regulate the expression
of a number of genes involved in reverse cholesterol transport,
fatty acid and glucose metabolism.2 In addition, LXRs inhibit lipo-
polysaccharide-induced macrophage expression of inflammatory
mediators such as inducible nitric oxide synthase, cyclooxygenase
2 and interleukin-6 and inhibit inflammation in vivo.3 Conse-
quently, LXR agonists have been studied as a potential therapy
for pulmonary inflammation, diabetes, and atherosclerosis.2
Recently, a number of structurally distinct LXR agonists have
been reported. These are highlighted by 24(S),25-epoxycholesterol
(1) and the synthetic ligand GW3965A (2) (Chart 1).4–6 Compound
2, identified from array synthesis, was described as a potent LXR ab
agonist. This compound showed good potency and full agonist
Analogs 8–15 all displayed potencies in the LXRab FRET assay
activity in the LXRb FRET (EC50 = 25 nM) and mouse macrophage
comparable to the lead 2 (Table 1); however, only 15 displayed full
agonism (>80% compared to the reference agonist 2). In addition,
compound 15 showed improved potency in the mouse macro-
phage cholesterol efflux assay compared to 2. Due to these results,
we focused our attention on modifying compound 15.
* Corresponding author. Tel.: +1 610 270 4941.
E-mail address: dave.g.washburn@gsk.com (D.G. Washburn).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.01.004

1098
D. G. Washburn et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1097–1100
O
H
O
HO
O
N
HO
1
2
Cl
CF3
Chart 1. Structures of 24(S),25-epoxycholesterol (1) and GW3965A (2).
Table 1
LXRa
b FRET assay data for compounds 2 and 8–18a
O
a
HN
H
+
Cl
RO
N
H2N
Cl
CF3
CF3
b
3
4
5
Cl
CF3
Compound
R
LXR
nM (% eff.)
a
EC50
,
LXRb EC50
nM (% eff.)
,
MM-Efflux
EC50, nM
c
RO
N
X
N
O
N
6: X = OH
7: X = Br
2
175 (100%)
25 (100%)
50
8-11
Cl
Cl
HO
HN
CF3
CF3
Scheme 1. Reagents and conditions: (a) NaB(OAc)3H, (2,2-diphenylethyl)amine,
AcOH, CH2Cl2,; (b) 3-bromo-1-propanol, NaI, K2CO3, CH3CN or 1,3 dibromopropane,
K2CO3, CH3CN; (c) ROH, DIAD, THF or ROH, K2CO3, CH3CN.
8
9
23 (25%)
63 (65%)
15 (30%)
26 (68%)
ND
ND
O
A docked pose of 15 overlaid with the crystal structure of 2
bound to LXRb suggested that while the indole was predicted to
better fill the hydrophobic pockets, key hydrogen bonding to
Leu330 and Arg319 were likely not be exploited as observed with
2 (Fig. 1).7,8
With this in mind, efforts to extend the reach of the indole
nitrogen of 15 were initiated. Reaction of 15 with base and electro-
philes produced compounds 16–18 (Table 1).9 While these com-
pounds were all equipotent to 15 in the LXRb FRET assay, they
O
O N
H
N
N
10
11
110 (17%)
>3000
60 (44%)
ND
ND
H
N
N
all showed improved LXR
17 displayed only partial LXR
a
potency. Furthermore, while 16 and
agonist activity, 18 demonstrated
510 (25%)
N
a
full agonism against both isoforms. In addition, 18 exhibited
excellent cellular activity in a mouse macrophage cholesterol
efflux assay (Scheme 2).
The crystal structures determined for 2 (bound to LXRb) and 18
N
12
13
790 (25%)
630 (19%)
560 (33%)
288 (36%)
ND
ND
(bound to LXRa) show clear distinctions between the two com-
pounds (Fig. 2).7,10 The indole ring of compound 18 but not the
phenyl ring of 2 appears to fill the leucine rich hydrophobic pocket.
There is also a hydrogen binding interaction (2.8 Å) between one of
the carboxylate oxygens of 18 and the NH backbone of Leu314
N
N
N
H
N
(LXRa). In addition, there is a stronger salt bridge interaction be-
tween Arg303 (LXRa) and one of the carboxylate oxygens of com-
O
pound 18 with the distances being 3.1 Å and 3.2 Å while the
corresponding interactions for compound 2 with Arg319 (LXRb)
are 3.6 and 3.8 Å, respectively.
O
Having prepared a compound with target in vitro characteris-
tics, we next evaluated its rat pharmacokinetic properties (Table
2). In general, compound 18 displayed attractive PK properties
including low clearance, small Vdss, moderate half-life, and good
oral bioavailability.
H
N
14
226 (36%)
23 (78%)
ND

D. G. Washburn et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1097–1100
1099
Table 1 (continued)
Compound
R
LXR
nM (% eff.)
a
EC50
,
LXRb EC50
nM (% eff.)
,
MM-Efflux
EC50, nM
15
60 (90%)
15 (104%)
5.0
ND
ND
HN
16
17
39 (57%)
20 (50%)
12 (86%)
7 (72%)
CH3N
CH3SO2N
O
HO
18
14 (132%)
4 (105%)
15
N
Figure 2. Superposition of crystal structures of 18 (green) bound to the LXR
a
-LBD &
2 (cyan) bound to the LXRb-LBD. Enzyme backbone and key backbone residues are
EC50 = concentration of compound that leads to half-maximal activity.
shown in the color of the ligand to which they correspond.
% efficacy (eff. = efficacy) normalized to 2. ND = no data available.
Table 2
In vivo rat pharmacokinetic data for compound 18a,b
Parameter
Compound
18a
Dose (iv, p.o., mg/kg)
CLp (mL/min/kg)
Vdss (L/kg)
1.1, 2.1
14.6 (1.7)
1.15 (0.29)
1.35 (.5)
t
, i.v. (h)
½
Oral Cmax (ng/mL)
, p.o. (h)
1808 (134)
2.9 (1.2)
t
½
Oral%F
45.4 (5.8)
a
Values are means of three experiments, standard deviation is given in
parentheses.
b
Male Sprague–Dawley rats (<1 year old) were used.
showed much improved potency and efficacy in the LXR
a FRET,
and potencies in the LXRb FRET and in mouse macrophage choles-
terol efflux assays were also improved. In addition, this indole-
based LXR agonist described has properties suitable for in vivo
testing, the results of which are to be described in a further
publication.
References and notes
Figure 1. Overlay of the crystal structure of 2 (cyan) bound to the LBD of LXRb and
15 (green).
1. (a) Jaye, M. Curr. Opin. Investig. Drugs 2003, 4, 1053; (b) Collins, J. L. Curr. Opin.
Drug Discov. Dev. 2004, 7, 692; (c) Lund, E. G.; Menke, J. G.; Sparrow, C. P.
Arterioscl. Throm. Vas. 2003, 237, 1169.
2. (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)
Levin, N.; Bischoff, E. D.; Daige, C. L.; Thomas, D.; Vu, C. T.; Heyman, R. A.;
Tangirala, R. K.; Schulman, I. G. Arterioscl. Throm. Vasc. 2005, 25, 135; (c) Laffitte,
B. A.; Chao, L. C.; Li, J.; Walczak, R.; Hummasti, S.; Joseph, S. B.; Castrillo, A.;
Wilpitz, D. C.; Mangelsdorf, D. J.; Collins, J. L.; Saez, E.; Tontonoz, P. Proc. Natl.
Acad. Sci. U.S.A. 2003, 100, 5419; (d) Chisholm, J. W.; Hong, J.; Mills, S. A.; Lawn,
R. M. J. Lipid Res. 2003, 44, 2039; (e) Smoak, K.; Madenspacher, J.; Jeyaseelan, S.;
Williams, B.; Dixon, D.; Poch, K. R.; Nick, J. A.; Worthen, G. S.; Fessler, M. B. J.
Immunol. 2008, 180, 3305.
a
O
N
R2
N
O
N
HN
16-18
15
Cl
Cl
CF3
CF3
3. Joseph, S. B.; Castrillo, A.; Laffitte, B. A.; Mangelsdorf, D. J.; Tontonoz, P. Nat.
Med. 2003, 9, 213.
4. (a) Lehmann, J. M.; Kliewer, S. A.; Moore, L. B.; Smith-Oliver, T. A.; Blanchard, D.
E.; Spencer, T. A.; Willson, T. M. J. Biol. Chem. 1997, 272, 3137; (b) Meinke, P. T.;
Wood, H. B.; Szewczyk, J. W. Annu. Rep. Med. Chem. 2006, 41, 99.
5. 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.
2002, 45, 1963.
Scheme 2. (a) NaH, electrophile, DMF.
In summary, replacement of the phenyl acetic acid head group
of the lead compound 2 with 1H-indol-1-yl acetic acid group led to
a novel orally bioavailable LXR agonist 18. Compared to 2, 18

1100
D. G. Washburn et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1097–1100
6. (a) 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; (b) 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.
Gene Dev. 2000, 14, 2831. Role of LXRs in control of lipogenesis.
dibromopropane (2.60 mL, 25.6 mmol) and CH3CN (25.6 mL). After the
reaction mixture was refluxed overnight, it was poured into saturated NH4Cl
and extracted with EtOAc (3Â). The organic layers were combined, dried over
Na2SO4, filtered, and concentrated. Purification via silica gel chromatography
yielded the product (1.10 g, 84%). 1H NMR (CDCl3): d 7.57 (d, J = 7.6 Hz, 1H),
7.33–7.14 (m, 12H), 4.18 (t, J = 7.6 Hz, 1H), 3.79 (s, 2H), 3.24 (t, J = 6.4 Hz, 2H),
3.15 (d, J = 7.6 Hz, 2H), 2.72 (t, J = 6.4 Hz, 2H), 1.95 (m, 2H) MS(ES) m/e 510.2
[M+H]+. MS(ES) m/e 510.2 [M+H]+. Step 3: To a sealed tube (3-bromo-propyl)-
7. (a) Faernegardh, M.; Bonn, T.; Sun, S.; Ljunggren, J.; Ahola, H.; Wilhelmsson, A.;
Gustafsson, J.-A.; Carlquist, M. J. Biol. Chem. 2003, 278, 38821; (b) Svensson, S.;
Oestberg, T.; Jacobsson, M.; Norstroem, C.; Stefansson, K.; Hallen, D.; Johansson,
I. C.; Zachrisson, K.; Ogg, D.; Jendeberg, L. EMBO J. 2003, 22, 4625; (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) Hoerer, S.; Schmid, A.; Heckel, A.;
Budzinski, R.-M.; Nar, H. J. Mol. Biol. 2003, 334, 853.
8. Compound 15 was docked using the program, Flo+, version 0203 as shown in
McMartin, C.; Bohacek, R. S. J. Comput.-Aided Mol. Des. 1997, 11, 333–334. The
protein coordinates containing polar hydrogen were converted to Macromodel
format using Flo+ tools. All residues within a 20 Å sphere centered on a residue
identified visually as central in the binding site were selected, and the rest of
the protein atoms were removed. The residues lining the binding site pocket
(approximately 10 Å from same residue near the center of the active site) were
selected to allow movement during minimization steps. The remaining
residues were held rigid during all docking and minimization calculations.
The mcdock algorithm, which relies on a Monte Carlo perturbation/fast search/
energy minimization algorithm was used for this study. Two thousand steps of
perturbation were performed and the twenty-five top-ranked poses were
retained. Visual inspection of the interactions made by the ligand within the
active site and the relative strain energies of the protein and ligand in each
pose was used to determine the best docked pose. The figures were made using
the Pymol software. DeLano, W.L. The PyMOL Molecular Graphics System
(2002) DeLano Scientific, San Carlos, CA, USA on the World Wide Web; http://
www.pymol.org.
(2-chloro-3-trifluoromethyl-benzyl)-2,2-diphenyl-ethyl-amine
(250 mg,
0.49 mmol), 4-hydroxyindole (79 mg, 0.50 mmol), K2CO3 (388 mg,
2.50 mmol), and CH3CN (3 mL) were added and heated at 90 °C overnight.
The mixture was filtered, concentrated, and purified via silica gel
chromatography to yield the product (270 mg, 98%). 1H NMR (CDCl3): d 8.14
(s, 1H), 7.45 (d, J = 7.6 Hz, 1H), 7.30–7.05 (m, 14H), 6.87 (m, 1H), 6.61 (m, 1H),
6.44 (d, J = 7.6 Hz, 1H), 4.22 (m, 1H), 3.99 (t, J = 6 Hz, 2H), 3.86 (s, 2H), 3.21 (d,
J = 7.6 Hz, 2H), 2.87 (t, J = 6 Hz, 2H), 2.01 (m, 2H). MS(ES) m/e 563.0 [M+H]+.
Step 4: To a solution of NaH (60% in oil) (6.79 g, 172.2 mmol) and DMF
(100 mL),
a mixture of (2-chloro-3-trifluoromethyl-benzyl)-(2,2-diphenyl-
ethyl)-[3-(1H-indol-4-yloxy)-2-methyl-propyl]-amine (9.70 g, 17.2 mmol)
and DMF (100 mL) was added. After the mixture stirred 0.25 h, BrCH2COOH
(12.0 g, 84.9 mmol) was added and the mixture was stirred overnight. The
reaction mixture was then poured into 1 N HCl and extracted with EtOAc. The
organic layers were combined, dried over Na2SO4, filtered, and concentrated.
Purification via preparative HPLC yielded the product which was dissolved in
Et2O, acidified with 1 N HCl in Et2O, and concentrated to afford the desired
product as the HCl salt (7.16 g, 67%). mp 197–199 °C; 1H NMR (CDCl3): d 7.97
(b, s, 1H), 7.60 (d, J = 8 Hz, 1H), 7.26–7.20 (m, 11H), 6.99 (m, 2H), 6.85 (d,
J = 2.4 Hz, 1H), 6.58 (dd, J = 8.8 Hz, 2 Hz, 1H), 6.42 (d, J = 3.2 Hz, 1H), 4.78 (s,
2H), 4.49 (m, 1H), 4.12 (s, 2H), 3.65 (m, 2H), 3.56 (m, 2H), 2.99 (m, 2H), 1.90 (m,
2H). MS(ES) m/e 621.2 [M+H]+.
10. Protein was expressed, purified and crystallized as in Ref. 6. The X-ray
diffraction data were collected at sector 17ID at the Advanced Photon Source,
Argonne National Laboratory. The X-ray diffraction images were processed
with HKL2000 (Otwinowski, Z. and Minor, W. Methods Enzymol. 1997, 276,
307–326). The protein crystallized in the space group C2 with cell dimensions
a = 122.2 Å, b = 90.0 Å, c = 101.6 Å, and b = 111.9 degr and with two
heterodimers per asymmetric unit. The Rmerge and completeness of the data
to 2.06 Å resolution were 8.2% and 93%, respectively. The structure was solved
by molecular replacement using PHASER (McCoy, A. J., Grosse-Kunstleve, R. W.;
Adams, P. D.; Winn, M. D.; Storoni, L.C.; Read, R. J. J. Appl. Cryst. 2007, 40, 658–
674.) and using starting coordinates from PDB id = 2ACL. The model was built
and refined with COOT (Emsley, P. & Cowtan, K. Acta. Cryst. D. 2004, 60 2126–
2132) and PHENIX.REFINE (Afonine, P. V.; Grosse-Kunstleve, R. W. & Adams; P.
D. 2005. CCP4 Newsl. 42, contribution 8). The model was refined to an R factor
of 20% and R free of 25%. Protein coordinates were deposited: RCSB ID code
rcsb050408 and PDB ID code 3FC6.
9. Example preparation: Compound 18. Step 1: To a solution of 2,2-diphenethyl-
amine (2.0 g, 10.0 mmole) and 2-chloro-3-trifluoromethylbenzaldehyde
(2.33 g, 11.0 mmole) in CH2Cl2 (20 mL) was added NaB(OAc)3
H (2.36 g,
11.0 mmole) and AcOH (2.0 mL). After the reaction mixture stirred overnight,
solvent was removed and the residue was washed with saturated NaHCO3, and
extracted with EtOAc (3Â). The organic extracts were dried over Na2SO4,
filtered, and concentrated. Purification via silica gel column chromatography
(silica gel 60, EM Science) afforded the product (3.0 g, 76% yield). 1H NMR
(CDCl3): d 7.62 (d, J = 8 Hz, 1H), 7.58 (d, J = 7.2 Hz, 1H), 7.35–7.22 (m, 11H), 4.26
(t, J = 7.6 Hz, 1H), 4.00 (s, 2H), 3.28 (d, J = 7.6 Hz, 2H). MS(ES) m/e 390.0 [M+H]+.
Step 2: K2CO3 (0.53 g, 3.85 mmol) was added to a solution of [2-chloro-3-
trifluoromethyl-benzyl](2,2-diphenylethyl)amine)
(1.00 g,
2.56 mmol),