3,4,5-Trisubstituted isoxazoles as novel PPARδ agonists. Part 2

Article abstract of DOI:10.1016/j.bmcl.2006.08.052

A series of PPARδ-selective agonists was investigated and optimized for a favorable in vivo pharmacokinetic profile. Isoxazole LCI765 (17d) was found to be a potent and selective PPARδ agonist with good in vivo PK properties in mouse (C_max = 5.1 μM, t_1/2 = 3.1 h). LCI765 regulated expression of genes involved in energy homeostasis in relevant tissues when dosed orally in C57BL6 mice. A co-crystal structure of compound LCI765 and the LBD of PPARδ is discussed.

Full text of DOI:10.1016/j.bmcl.2006.08.052

Bioorganic & Medicinal Chemistry Letters 16 (2006) 5488–5492
3,4,5-Trisubstituted isoxazoles as novel PPARd agonists. Part 2
Robert Epple,a,* Mihai Azimioara,a Ross Russo,a Yongping Xie,a Xing Wang,a
Christopher Cow,a John Wityak,a Don Karanewsky,a Badry Bursulaya,b
Andreas Kreusch,b Tove Tuntland,c Andrea Gerken,c Maya Iskandar,c Enrique Saez,c
H. Martin Seidelc and Shin-Shay Tianc
aDepartment of Medicinal Chemistry, The Genomics Institute of the Novartis Research Foundation,
10675 John Jay Hopkins Drive, San Diego, CA 92121, USA
bDepartment of Structural Biology, The Genomics Institute of the Novartis Research Foundation,
10675 John Jay Hopkins Drive, San Diego, CA 92121, USA
cDepartment of Lead Discovery, The Genomics Institute of the Novartis Research Foundation,
10675 John Jay Hopkins Drive, San Diego, CA 92121, USA
Received 10 July 2006; revised 8 August 2006; accepted 9 August 2006
Available online 22 August 2006
Abstract—A series of PPARd-selective agonists was investigated and optimized for a favorable in vivo pharmacokinetic profile.
Isoxazole LCI765 (17d) was found to be a potent and selective PPARd agonist with good in vivo PK properties in mouse
(Cmax = 5.1 lM, t1/2 = 3.1 h). LCI765 regulated expression of genes involved in energy homeostasis in relevant tissues when dosed
orally in C57BL6 mice. A co-crystal structure of compound LCI765 and the LBD of PPARd is discussed.
Ó 2006 Elsevier Ltd. All rights reserved.
The peroxisome proliferator-activated receptors are
ligand-activated transcription factors belonging to the
nuclear hormone superfamily.1,2 In recent years, the
three distinct subtypes (PPARa, PPARc, and PPARd)
have received great attention in both academic and
pharmaceutical research due to their involvement in glu-
cose and lipid homeostasis.3–7
an improved PPARd selectivity profile in order to distin-
guish the PPARd-driven effects from those derived from
activation of either PPARc or/and PPARa subtypes.
We have recently described the identification of an isox-
azole scaffold HTS hit and its optimization to the lead
compound 1 (Fig. 1) with exceptional selectivity for
PPARd.11 Unfortunately, all analogs in the series up
to this point had shown rather poor bioavailability in
mouse. Further optimization was necessary to solve
the issue. Since neither replacement of the amide func-
tionality in position 4 nor any change of substituent in
position 5 led to an improvement in bioavailability, we
focused our efforts on changing the core heterocycle
and the head group substituent bearing the carboxylate
in position 3.
In light of the success of the hypolipidemic fibrates and
the TZD class of insulin sensitizers acting through acti-
vation of PPARa1 and PPARc,8 respectively, pharma-
ceutical companies have focused efforts mostly on
developing more potent and selective agonists on these
two PPAR subtypes. More recent reports have raised
the awareness for the third subtype, PPARd, suggesting
a likewise important role in energy homeostasis.9,10 The
synthetic PPARd ligands used in these studies show
significant cross-reactivity to other PPAR subtypes,
which underscores the need for tool compounds with
Switching the nitrogen and oxygen of the isoxazole gave
an ‘inverted’ isoxazole 2 with similar activities, but also
similarly poor in vivo properties when compared to
compound 1. Changes from isoxazoles to any other het-
erocycles such as pyrazoles 3 and 4, imidazoles 5, and
triazoles 6 were not tolerated and resulted in inactive
analogs (Fig. 1).12
Keywords: Nuclear hormone receptor; PPAR agonist; PPARd; Energy
homeostasis; Metabolic xyndrome; Isoxazoles.
*
Corresponding author. Tel.: +1 858 812 1720; fax: +1 858 812
1848; e-mail: repple@gnf.org
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.08.052

R. Epple et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5488–5492
5489
1O
investigated changes in the head group phenyl substit-
uents and/or their substitution patterns. Starting from
the appropriate hydroxy-phenylacetic acids, hydroxy-
phenylpropionic acids or hydroxy-phenoxyacetic acids
12 the carboxylate functionality was esterified, then
the hydroxyl was triflated and converted to the cya-
nides 13 using zinc (II) cyanide (Scheme 2). Alterna-
tively, when aromatic halides were available as
starting materials instead of phenols, the triflation step
could be omitted. Raney-nickel alloy reduction in for-
mic acid, oxime formation, and chlorination gave the
reactive chloroxime intermediates 14. The cyclization
partners 16 were synthesized by microwave-assisted
amide formation from the corresponding b-ketoesters
15 with amine 10, followed by deprotonation using
potassium bis(trimethylsilyl)amide. The cyclization of
intermediates 14 and 16 proceeded regioselectively un-
der mild conditions, and saponification with lithium
hydroxide gave the free carboxylates 17 in overall fair
to good yields.
Cl 2 N
N
5
O
4
3
NH
Cl
HO2C
O
O
Cl
2
1
EC50 (PPARδ)
0.08 μM (66%)
EC50 (PPARδ)
0.10μM (62%)
N
N
N
N
N
N
N
N
N
3
4
5
6
inactive
Figure 1. Changing the isoxazole core heterocycle led to loss of activity
in all compounds except compound 2.
Table 1 shows a list of selected analogs and their in vitro
activities in PPAR transactivation assays. All analogs
presented were metabolically stable after incubation
with mouse, rat, and human liver microsomes, and be-
haved similarly in a standard kinetic solubility assay.
Based on an acceptable in vitro activity profile (EC50-
(PPARd) < 100 nM), selected compounds were dosed
in C57BL6 mice for an assessment of their in vivo phar-
macokinetic properties. The mice were dosed po at
20 mpk in a 0.5% CMC suspension. The maximal con-
centration (Cmax) and half-life (t1/2) were calculated
from a 24 h time course study.
Next, we directed our attention toward position 3 of the
isoxazole. Most PPAR agonists connect the functional
head group to the lipophilic tail group via a flexible link-
er. In our case, inserting a flexible linker between the
head group and isoxazole core was accomplished by
reacting 3-methyl-5-phenyl-isoxazole-4-carbonyl chlo-
ride 7 with tert-butanol (Scheme 1). The methyl group
was brominated using NBS to give isoxazole 8, followed
by ether formation with the appropriate phenol or thi-
ophenol 9. Removal of the tert-butyl protecting group,
activation of the acid with thionyl chloride, and amide
formation with the amine 10 gave an intermediate ester,
which was saponified to the final products 11. Insertion
of a flexible linker obliterated activity in all cases, which
is exemplified in Table 1 by compound 11a.
All analogs disclosed in the first publication11 were rap-
idly cleared when dosed iv, regardless of the nature of
the substituents in positions 4 and 5 of the isoxazoles.
This was also reflected in their low Cmax (<0.3 lM)
and short half-lives (<0.5 h) when dosed po, exemplified
by compound 1 in Table 1. Notably, these compounds
Since a direct single bond between head group phenyl
and isoxazole is required for PPAR activation, we
MeO2C
Y
XH
O
O
O
N
N
N
a), b)
R
HO2C
Y
Br
X
Cl
c)
X = O, S
OtBu
OtBu
O
O
O
R
Y = CH2, (CH2)2, OCH2
R = H, Me, Cl
8
7
9
Cl
Cl
d)-g)
H2N
O
10
O
N
HO2C Y
X
NH
Cl
O
R
O
Cl
11
Scheme 1. Reagents and conditions: (a) t-BuOH, pyridine, DCM, rt, 12 h (70%); (b) NBS, CCl4, rt, 12 h (43%); (c) Cs2CO3, CH3CN, rt, 2 h (>90%);
(d) TFA/DCM 1:5, rt, 7 h (>90%); (e) SOCl2, benzene, 80 °C, 2 h (80–90%); (f) compound 10, K2CO3, DCM, rt, 12 h (>80%); (g) 1 N LiOH, THF, rt,
12 h (95%).

5490
R. Epple et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5488–5492
Table 1. In vitro transactivation and in vivo pharmacokinetic properties of selected examples
R2
Cl
Cl
O
N
H
N
O
R1
O
Compound
R1
R2
Cell-based transactivationa
In vivo PK (mouse)
20 mpk po
huPP ARa
EC50 (lM)
huPP ARc
EC50 (lM)
huPP ARd
Cmax (lM)
t1/2 (h)
EC50 (lM)a/% eff
Cl
CO2H
1
>10
>10
0.08/66%
<0.3
<0.5
Cl
CO2H
11a
17a
17b
17c
17d
17e
17f
17g
17h
17i
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
O
CO2H
0.72/42%
>10
CO2H
O
CO2H
0.20/60%
0.07/83%
0.97/38%
0.56/70%
1.07/52%
0.09/65%
0.05/76%
0.06/83%
<0.3
5.1
<0.5
3.1
CO2H
CO2H
CO2H
MeO
CO2H
O
<0.3
1.9
<0.5
1.5
CO2H
CO2H
CO2H
Cl
17j
5.6
1.1
a Concentration of test compound that produced 50% of the maximal reporter activity (100% control: GW501516); all data within 25%.
shared a para substituted 3-chloro-phenylacetic acid in
position 3 of the isoxazole (R1).
A breakthrough was achieved when the para substitution
pattern of the phenylacetic acid (compound 17a) was
changed to meta (compound 17d, LCI765). Compound
LCI765 not only displayed good in vitro activity and
selectivity for the PPARd subtype, but also showed a tre-
Removing the chloride (compound 17a) led to an
approximately 10-fold reduction in activity. Switching
from phenylacetic acids to benzoic acid analogs obliter-
ated activity, regardless of the carboxylate being in
ortho, meta or para (compound 17b) position. On the
other hand, extending the carboxylate substituent to
phenyl propionic acids or phenoxyacetic acids (exempli-
fied by compound 17c) retained activity, but did not lead
to an improvement of the in vivo PK profile.
mendous improvement in its in vivo PK profile (Cmax
=
5.1 lM, t1/2 = 3.1 h). LCI765 was tested in vivo in an acute
dosing study in C57BL6 mice.13 Similar to the PPARd-se-
lective ligand GW501516,14–16 LCI765 regulated expres-
sion of genes involved in glucose and lipid metabolism
(PDK4 and ApoAIV), and energy uncoupling (UCP3)
in skeletal muscle and intestinal tissues (Fig. 2).

R. Epple et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5488–5492
5491
O
O
R2 = Alkyl, Aryl
R2
OMe
15
Cl
Cl
g), h)
H2N
O
10
OH
Cl
N
HO2C
Y
MeO2C
Y
X
MeO2C
Y
OK
O
Cl
CN
a)-c)
d)-f)
O
R2
N
H
R1
R1
R1
Cl
12
X = OH, Cl, Br
13
14
16
i), j)
Y = CH2, (CH2)2, OCH2
R1 = H, Me, Cl, OMe
O
N
R2
NH
HO2C
Y
Cl
R1
O
O
Cl
17
Scheme 2. Reagents and conditions: (a) MeOH, H2SO4, reflux, 8 h, quant.; (b) Tf2O, NEt3, DCM, 0 °C, 1 h (>95%); (c) Zn(CN)2, Pd(PPh3)4, DMF,
80 °C, 30 h (60–80%); (d) Raney-alloy, H2CO2, 110 °C, 10 h (50–60%); (e) H2NOHÆHCl, Na2CO3, H2O, rt, 2 h, (>90%); (f) NCS, DMF, HCl(g), rt,
30 min; (g) compound 10, toluene, MW (160 °C, 10 min) (40–60%); (h) KN(TMS)2, Et2O, rt, 1 h; (i) CH3CN, rt, 3 h (60–80%); (j) 1 N LiOH, THF,
rt, 12 h (>95%).
Gene Regulation
5
4
3
2
1
0
GW 501516(10mpk)
LCI765(10mpk)
**
**
**
**
**
*
PDK4
(SKM)
UCP3
(SKM)
ApoAIV
(Jejunum)
Figure 2. Acute dosing-induced gene assay of LCI765. Tissues were
harvested from C57BL/6 mice dosed for days with vehicle,
3
GW501516 or LCI765. The expression of various genes was analyzed
by real-time PCR analysis. Fold induction is reported relative to
vehicle control. *p value < 0.03, **p value < 0.01.
Extension of the carboxylate substituent (compound
17e) or additional substituents (cf. compound 17f) did
not lead to improved in vitro activities. Changes of the
amide substituent in position 4 of the isoxazole core
were not well tolerated, following the trend of narrow
SAR described in the earlier communication.11 In gener-
al, alkyl substituents in position 5 (R2) (e.g., compound
17g) had deleterious effects on PPARd in vitro activity.
Small heterocycles as in compound 17h were tolerated,
but lacked favorable PK. The only other active and bio-
available compounds in addition to the plain phenyl
substituted analog LCI765 displayed an ortho substitut-
ed phenyl ring as in compounds 17i and 17j.
Figure 3. Superposition of co-crystal structures of the PPARd LBD
with LCI765 (blue) and GW2433 (yellow). The image was generated
using Maestro GUI (Schrodinger, Inc.).
deviation of backbone atoms between residues in the
LBD (29 residues) of our crystal structure and the equiv-
alent residues in the co-crystal structure of PPARd/
18
˚
GW2433 (pdb code 1GWX ) is ꢀ0.7 A, indicating a
very similar overall conformation. The orientation of
head group and tailgroup is in accordance with the
docking results of the original HTS hit used in the de-
sign of the isoxazole series.11 However, the model would
fail to predict the tolerance of large substituents in posi-
tion 5 of the isoxazole. The most prominent conforma-
tional change from the PPARd/GW2433 structure used
LCI765 was co-crystallized with the ligand binding do-
main (LBD) of PPARd (Fig. 3).17 The root mean square

5492
R. Epple et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5488–5492
6. Berger, J.; Moller, D. E. Annu. Rev. Med. 2002, 53, 409.
7. Willson, T. M.; Brown, P. J.; Sternbach, D. D.; Henke, B.
R. J. Med. Chem. 2000, 43, 527.
8. Willson, T. M.; Cobb, J. E.; Cowan, D. J.; Wiethe, R. W.;
Correa, I. D.; Prakash, S. R.; Beck, K. D.; Moore, L. B.;
Kliewer, S. A.; Lehmann, J. M. J. Med. Chem. 1996, 39,
665.
9. Leibowitz, M. D.; Fievet, C.; Hennuyer, N.; Peinado-
Onsurbe, J.; Duez, H.; Berger, J.; Cullinan, C. A.;
Sparrow, C. P.; Baffic, J.; Berger, G. D.; Santini, C.;
Marquis, R. W.; Tolman, R. L.; Smith, R. G.; Moller, D.
E.; Auwerx, J. FEBS Lett. 2000, 473, 333.
in the docking study to the co-crystal structure with
LCI765 is observed for the dihedral angle chi-1 of
Leu330, whose value is different by ꢀ100°.19 This move-
ment is necessary in order to enlarge a cavity lined by
residues Phe327, Leu330, Val334, Leu339, Ile364, and
Lys367, so that it can accommodate the phenyl ring in
position 5 of the isoxazole. The corresponding cavity
does not exist in PPARa and PPARc where both iso-
forms have the bigger Met residues in place of Val334
and Ile364, respectively. This might explain the observed
PPARd selectivity of 3,4,5-trisubstituted isoxazoles.
10. Oliver, W. R., Jr.; Shenk, J. L.; Snaith, M. R.; Russell, C.
S.; Plunket, K. D.; Bodkin, N. L.; Lewis, M. C.; Winegar,
D. A.; Sznaidman, M. L.; Lambert, M. H.; Xu, H. E.;
Sternbach, D. D.; Kliewer, S. A.; Hansen, B. C.; Willson,
T. M. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 5306.
11. Epple, R.; Russo, R.; Azimioara, M.; Cow, C.; Xie, Y.;
Wang, X.; Wityak, J.; Karanewsky, D.; Gerken, A.;
Iskandar, M.; Saez, E.; Seidel, H. M.; Tian, S.-S. Bioorg.
Med. Chem. Lett. 2006, 16, 4376.
12. The synthetic procedures of all analogs described in this
communication except compounds 3–6 and 11a can be
found in the following patent application: Epple, R.;
Russo, R.; Azimioara, M.; Xie, Y. WO 2005113519.
Synthetic procedures of compounds 3–6 and 11a can be
found in the supplementary material.
13. C57BL/6 male mice (age 8–9 weeks, Charles River
Laboratories) were orally gavaged once daily with the
compounds at 10 mg/kg or with vehicle (0.5% CMC: 2%
Tween 80) for 3 days. On the 4th day and 4 h after the last
dose was administered, mice were euthanized. Samples of
skeletal muscle (quadriceps), adipose tissue, jejunum, and
ileum were collected, rapidly frozen in liquid nitrogen, and
stored at À80 °C. Total RNA was isolated from the tissues
and SYBR Green quantitative real-time-PCR was per-
formed on an ABI PRISM 7900HT Sequence Detection
system (Applied biosystems). For each sample, the quan-
tity of the target gene and the endogenous reference
GAPDH was determined to obtain a normalized target
value. Data were analyzed using SDS 2.0 software
(Applied biosystems).
In summary, PPARd-selective isoxazoles 17d, 17i, and
17j exhibiting greatly improved mouse in vivo pharma-
cokinetic properties were identified, while the in vitro
activity and selectivity of the three isoxazoles was main-
tained or improved compared to compound 1. This was
achieved by optimization of the head group in position 3
of the isoxazole core of compound 1. Compound
LCI765 (17d) regulates expression of genes involved in
energy homeostasis in relevant tissues when dosed orally
in C57BL6 mice. Additionally, a co-crystal structure of
compound LCI765 with the PPARd ligand binding do-
main revealed the formation and occupancy of a new
hydrophobic cavity. This cavity may be formed exclu-
sively in the PPARd isoform due to the smaller amino
acid side chains lining the pocket as compared to the
other isoforms. Compound LCI765 represents the first
example to occupy this side pocket and underlines the
flexibility of the PPAR ligand binding domain to accom-
modate various diet-derived lipids and their metabolites.
LCI765 should be a useful tool compound to elucidate
the pharmacological consequences of selective PPARd
activation.
Acknowledgments
We gratefully acknowledge Drs. T. R. Vedananda,
Sandra Teixeira, and Peter G. Schultz for their helpful dis-
cussions and support. We also thank ActiveSight, San
Diego, for their support in our co-crystallization efforts.
14. Dressel, U.; Allen, T. L.; Pippal, J. B.; Rohde, P. R.;
Lau, P.; Muscat, G. E. O. Mol. Endocrinol. 2003, 17,
2477.
15. Wang, Y.-X.; Lee, C.-H.; Tiep, S.; Yu, R. T.; Ham, J.;
Kang, H.; Evans, R. M. Cell 2003, 113, 159.
16. Tanaka, T.; Yamamoto, J.; Iwasaki, S.; Asaba, H.; Ikeda,
Y.; Watanabe, Y.; Uchiyama, Y.; Sumi, K.; Iguchi, H.;
Ito, S.; Doi, T.; Hamakubo, T.; Naito, M.; Auwerx, J.;
Yanagisawa, M.; Kodama, T.; Sakai, J. Proc. Natl. Acad.
Sci. U.S.A. 2003, 100, 15924.
17. Atomic coordinates of the co-crystal structure have been
deposited with the PDB and are accessible under the code
2J14.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2006.08.052
18. Xu, H. E.; Lambert, M. H.; Montana, V. G.; Parks, D. J.;
Blanchard, S. G.; Brown, P. J.; Sternbach, D. D.;
Lehmann, J. M.; Wisely, G. B.; Willson, T. M.; Kliewer,
S. A.; Milburn, M. V. Mol. Cell 1999, 3, 397.
19. The amino acid labels in Figure 3 are given in consistency
with published crystal structures, but do not correspond to
actual PPARd sequence numbering. The actual numbers
for residues Phe327, Leu330, Val334, Leu339, Ile364, and
Lys367 are Phe291, Leu294, Val298, Leu303, Ile328, and
Lys331, respectively.
References and notes
1. Issemann, I.; Green, S. Nature 1990, 347, 645.
2. Mangelsdorf, D. J.; Evans, R. M. Cell 1995, 83, 841.
3. Cheng, P. T. W.; Mukherjee, R. Mini-Rev. Med. Chem.
2005, 5, 741.
4. Staels, B.; Fruchart, J.-C. Diabetes 2005, 54, 2460.
5. Evans, R. M.; Barish, G. D.; Wang, Y.-X. Nat. Med. 2004,
10, 355.