Novel selective small molecule agonists for peroxisome proliferator-activated receptor δ (PPARδ) - Synthesis and biological activity

Article abstract of DOI:10.1016/S0960-894X(03)00207-5

We report the synthesis and biological activity of a new series of small molecule agonists of the human Peroxisome Proliferator-Activated Receptor δ (PPARδ). Several hits were identified from our original libraries containing lipophilic carboxylic acids. Optimization of these hits by structure-guided design led to 7k (GW501516) and 7l (GW0742), which shows an EC₅₀ of 1.1 nM against PPARδ with 1000-fold selectivity over the other human subtypes.

Full text of DOI:10.1016/S0960-894X(03)00207-5

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1517–1521
Novel Selective Small Molecule Agonists for Peroxisome
Proliferator-Activated Receptor ꢀ (PPARꢀ)—Synthesis and
Biological Activity
Marcos L. Sznaidman,a,y Curt D. Haffner,a Patrick R. Maloney,a Adam Fivush,a
Esther Chao,a Donna Goreham,a Michael L. Sierra,b Christelle LeGrumelec,b H. Eric Xu,a
Valerie G. Montana,a Millard H. Lambert,a Timothy M. Willson,a William R. Oliver, Jr.a
and Daniel D. Sternbacha,*
aGlaxoSmithKline, Five Moore Drive, Research Triangle Park, NC 27709-3398, USA
´
GlaxoSmithKline, Centre de Recherches, 25 Avenue du Quebec, 91951 Les Ulis Cedex, France
b
Received 7 January 2003; revised 14 February 2003; accepted 27 February 2003
Abstract—We report the synthesis and biological activity of a new series of small molecule agonists of the human Peroxisome
Proliferator-Activated Receptor d (PPARd). Several hits were identified from our original libraries containing lipophilic carboxylic
acids. Optimization of these hits by structure-guided design led to 7k (GW501516) and 7l (GW0742), which shows an EC50 of 1.1
nM against PPARd with 1000-fold selectivity over the other human subtypes.
# 2003 Elsevier Science Ltd. All rights reserved.
The Peroxisome Proliferator-Activated Receptors
(PPARs) comprise a family of ligand-activated tran-
scription factors belonging to the nuclear receptor gene
superfamily.1 Three mammalian PPAR subtypes, com-
monly known as PPARa, PPARg and PPARd were
originally cloned as orphan receptors.2ꢀ4 Each subtype
contains a signature DNA binding domain (DBD) and
a ligand-binding domain (LBD), both of which are
necessary for ligand activated gene transcription. The
PPARs function through heterodimers with the 9-cis-
retinoic acid receptor (RXR) to regulate lipid and car-
bohydrate metabolism. The role of PPARa in the reg-
ulation of hepatic lipid metabolism was uncovered when
it was shown to be the target receptor for the fibrate
class of antihyperlipidemic drug molecules.2 Likewise,
the role of PPARg in the regulation of insulin sensitivity
was discovered when it was shown to function as the
cellular receptor of the glitazone class of antidiabetic
drugs.5 In contrast, PPARd does not appear to be a
receptor for any known classes of drugs.6 As such, the
biological role of this subtype has remained elusive due,
in part, to its broad tissue expression and the lack of
good chemical tools with which to study its pharmacol-
ogy. Thus, part of the challenge in determining the
function of PPARd has been the identification of potent
and selective ligands for use as chemical tools.7 It is
important to note that all of the ligands published to
date either have low affinity for PPARd or lack selec-
tivity over the other PPARs.8ꢀ12
To elucidate the function of PPARd we set out to iden-
tify potent and selective ligands for this subtype using
combinatorial chemistry and structure-based drug
design.
Chemistry
The general synthesis of compounds 7a–l is depicted in
Table 1. First, the corresponding phenoxy or thiophe-
noxy methyl ester (5a–l) was coupled to the appropriate
chloromethyl (4a–c, Scheme 1) or hydroxymethyl (3a–b,
Scheme 1) derivative by using either method A or B,
respectively. Method A is the Williamson synthesis of
an ether (or thioether) bond by displacement of the
*Corresponding author. Tel.: +1-919-483-1424; fax: +1-919-315-
5668; e-mail: dds3687@gsk.com
yCurrent address: Norak Biosciences, PO Box 14769, Research Trian-
gle Park, NC 27709-3398, USA
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00207-5

1518
M. L. Sznaidman et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1517–1521
Table 1. General synthetic routea,b
condensation of a-bromoacetic ester 10 with hydroxy-
acetophenone 11 afforded ester 12. Baeyer–Villiger oxi-
dation of the ketone to an ester afforded 13, which after
removal of the acetate group afforded the phenoxy ester
5h. Thiophenoxy esters 5i and 5j were synthesized as
follows: first, either phenoxyacid 5g or 5h was con-
densed with dimethylthiocarbamoyl chloride to afford
the protected phenoxy esters 14 or 15, respectively.
Rearrangement at high temperatures afforded the pro-
tected thiophenoxy esters 16 or 17, and final deprotec-
tion with sodium methoxide afforded the final
thiophenoxy esters 5i (not isolated, used in situ in the
next step) and 5j, respectively.
Compd
X
R1 R2
Y
Z
Method Overall yield
(%)
7a
7b
7c
7d
7e
7f
7g
7h
7i
7j
CH2
CH2CH2
CH=CH Me
CH2CH2 Me
CH2CH2
—
CH2O
CH2O
CH2O
CH2O
CH2O
CH2O
H
H
H
H
H
H
H
H
H
H
H
H
H
F
O
O
O
O
O
O
O
O
S
S
S
S
A
B
38
17
25
38
51
50
14
63
10
57
52
50
A
A
B
S
H
H
O
O
O
O
O
O
S
A
B
H
SAR Studies
Me
H
A
A
A
A
A
Table 2 shows the EC50 for compounds 7a–l in the
transient transactivation assay against the three human
PPAR subtypes: a, g and d; and binding IC50 against
the human PPARd receptor. Since the PPARs were
known to bind lipophilic carboxylic acids13 we decided
to screen existing libraries with these structural proper-
ties.14 As a result, several PPARd agonists were dis-
covered, compounds 7b and 7c being the most potent
with single digit nM activity in the PPARd cell based
reporter assay. The selectivity versus PPARa and
PPARg in this assay was 35- and 80-fold respectively for
7b and greater than 100-fold on both receptors for 7c.
While the binding data was in good agreement with the
cell based reporter assay for 7b, the cinnamic acid deri-
vative (7c) showed an anomalous low binding affinity
for PPARd (IC50=1.0 mM, Table 2). The cocrystal
structure of 7c with the human PPARd ligand-binding
domain provided a possible explanation for the diver-
gent assay results. Although 7c was synthesized as the
pure E-cinnamic acid, the X-ray structure (Fig. 1)
showed that the double bind had isomerized to the Z-
configuration during the crystallization process. Since
7b is a propionic acid derivative it could readily assume
the s-cis geometry needed for high affinity binding to
PPARd while 7c could not. However, within the time
course of the cell-based assay it is possible that 7c could
isomerize to the Z-isomer, which was observed in the
Me
Me
Me
S
S
S
7k GW501516
7l GW0742
S
aMethod A: Cs2CO3, CH3CN, 16 h, rt.
bMethod B: Ph3P, DEAD, THF, 48 h, rt.
reactive chlorine of 4a–c with a phenoxide (or thio-
phenoxide) salt and method B is the Mitsunobu con-
densation of an alcohol (3a–b) with an acidic phenol or
thiophenol. The intermediate esters (6a–l) were then
hydrolyzed to the corresponding carboxylic acids 7a–l.
Synthesis of intermediates 3a–b and 4a–c is depicted in
Scheme 1. The ethyl a-chloro-b-keto ester was con-
densed with the thioamide 1a to form the substituted
thiazole 2a. Reduction with LiAlH4, afforded the alco-
hol 3a, which upon treatment with methansulfonyl
chloride afforded the chloromethyl thiazole 4a. In a
similar fashion the oxazoles 3b and 4b, and thiazoles 3c
and 4c were synthesized. Scheme 2 shows the synthesis
of the non-commercially available phenoxy (5c, 5d and
5h) and thiophenoxy (5i and 5j) esters. Compound 5c
was obtained by Heck coupling of methyl acrylate (8)
and bromophenol 9. Hydrogenation of the double bond
of 5c afforded compound 5d in quantitative yield.
Phenoxy ester 5h was obtained in three steps. First,
Scheme 1. Synthesis of intermediates 3a, 3b, 3c, 4a, 4b and 4c.

M. L. Sznaidman et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1517–1521
1519
Scheme 2. Synthesis of phenoxy esters 5c, 5d and 5h, and thiophenoxy esters 5i and 5j.
Table 2. Activity of compounds 7a–l in cell based transactivation assaya and binding assayb against human PPAR receptors
Transactivation EC50 (mM)a
Binding IC50 (mM)b
hPPARd
Compd
hPPARa
hPPARg
HPPARd
7a
7b
7c
7d
7e
7f
7g
7h
7i
7j
6.9ꢁ1.7
3.0ꢁ1.5
3.2ꢁ7.3
1.7ꢁ0.17
0.28ꢁ0.089
Inactive
7.9ꢁ1.3
7.04ꢁ3.6
2.7ꢁ3.3
4.2ꢁ2.2
3.3ꢁ1.9
Inactive
9.8ꢁ8.0
Inactive
6.4ꢁ3.9
1.61ꢁ2.6
0.85ꢁ0.29
2.0ꢁ1.3
1.9ꢁ0.39
0.09ꢁ0.02
0.02ꢁ0.006
0.03ꢁ0.03
0.06ꢁ0.02
5.9ꢁ5.5
1.0ꢁ0.07
0.03ꢁ0.002
1.0ꢁ0.03
0.06c
0.09ꢁ0.003
5.5ꢁ0.28
Inactive
3.2ꢁ1.7
0.47ꢁ0.18
0.03ꢁ0.002
0.42ꢁ0.18
0.01ꢁ0.001
0.001ꢁ0001
0.001ꢁ0002
0.34ꢁ0.008
0.03ꢁ0.001
0.09ꢁ0.002
0.003ꢁ0.0001
0.001ꢁ0.0001
0.001ꢁ0.00001
Inactive
2.5ꢁ1.6
1.1ꢁ0.15
1.1ꢁ0.109
7k GW501516
7l GW0742
aEC50, the concentration of test compound required to induce 50% of the maximum alkaline phosphatase activity (SEM).
bKi, the concentration of test compound required to achieve an apparent Ki value according to the equation Ki=IC50/(1+[L]/Kd), where IC50 is the
concentration of test compound required to inhibit 50% of the specific binding of the radioligand, [L] is the concentration of the radioligand used,
and Kd is the dissociation constant for the radioligand at the receptor.
cAverage of two measurements.
crystal structure. Further confirmation of this hypoth-
esis resulted from the reduction of the double bond in 7c
to yield propionic acid 7d, which had comparable
activity to 7c in the cell-based assay and importantly,
showed 37-fold higher affinity than 7c for PPARd in the
binding assay.
effect on PPARd binding activity however the PPARd
selectivity versus PPARa in the cellular assay dropped
from 80-fold to 5-fold. Pharmacokinetic measurements
with 7e in mice revealed that it had a very short half-life
(<10 min) due to rapid and extensive metabolism. The
inactive benzoic acid 7f was identified (by mass spec) as
the primary metabolite, presumably resulting from
b-oxidative degradation of the propionic acid side
chain. Thus, although compounds 7b, 7c, 7d and 7e
Changing the heterocycle in the propionic acid 7b
(thiazole) to an oxazole (cf. 7e in Table 2) had little

1520
M. L. Sznaidman et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1517–1521
EC50=8.8, >10, 0.03, PPARa, PPARg, and PPARd,
respectively).
When dosed to insulin-resistant middle-aged rhesus mon-
keys, 7k causes a dramatic dose-dependent rise in serum
HDLc while lowering the levels of LDLc, fasting trigly-
cerides, and fasting insulin.16 At the cellular level, com-
pound 7k increased expression of the reverse cholesterol
transporter ATP-binding cassette A1 (ABCA1) and
induced apolipoprotein A1-specific cholesterol efflux.
Conclusions
We were able to identify the first truly selective PPARd
agonists, GW501516 (7k) and GW0742 (7l), starting
with high-throughput screening, then using combina-
torial chemistry to develop small focused libraries, and
finally using structure guided lead optimization. These
compounds will be valuable tools for determining the
biological activity of the PPARd receptor and as poten-
tial therapeutic agents for the treatment of diseases
associated with raised serum triglycerides and low levels
of HDLc.
Figure 1. X-ray data showing the binding interactions of cinnamic
acid 7c to the hPPARd receptor.
were potent PPARd agonists in vitro, they were unsui-
table for use as chemical tools in vivo.
The phenyl acetic acid derivative 7a is unable to
undergo b-oxidation but its weak PPARd activity
(EC50=1.77 mM) and the fact that crystallographic
studies13 indicated that a two-atom spacer between the
carboxylic acid and the aryl ring would be preferred led
us to explore other alternatives. Replacement of the
propionic side chain of 7e with an isosteric oxyacetic
acid side chain in 7g would also serve to block b-oxida-
tion. Unfortunately, 7g was significantly less potent on
PPARd (0.47 mM) and showed no selectivity over
PPARg.
Acknowledgements
We gratefully acknowledge Kim Adkison, Kelli Plun-
ket, Raymond Merrihew and Lisa Leesnitzer, for their
help in pharmacokinetics, transient transfection and
binding determinations.
Early modeling studies,15 later confirmed by crystal-
lographic studies,13 showed that there is a lipophilic
pocket in the PPARd ligand-binding domain that could
accommodate small substituents at the ortho position of
the aromatic ring. The enhancement of PPARd activity
by methyl substitution at the ortho position can be seen
by comparing 7d with 7b (3-fold enhancement) and 7h
with 7g (15-fold enhancement). Comparable effects were
observed in the binding assay. Notably, the selectivity
versus PPARa and PPARg remained high.
References and Notes
1. Willson, T. M.; Brown, P. J.; Sternbach, D. D.; Henke,
B. R. J. Med. Chem. 2000, 43, 527.
2. Issemann, I.; Green, S. Nature 1990, 347, 645.
3. Kliewer, S. A.; Forman, B. M.; Blumberg, B.; Ong, E. S.;
Borgmeyer, U.; Mangelsdorf, D. J.; Umesono, K.; Evans,
R. M. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 7355.
4. Schmidt, A.; Endo, N.; Rutledge, S. J.; Vogel, R.; Shinar,
D.; Rodan, G. A. Mol. Endocrinol. 1992, 6, 1634.
5. Lehmann, J. M.; Moore, L. B.; Smith-Oliver, T. A.; Wilk-
ison, W. O.; Willson, T. M.; Kliewer, S. A. J. Biol. Chem.
1995, 270, 12953.
6. Willson, T. M.; Wahli, W. Curr. Opin. Chem. Biol. 1997, 1,
235.
7. Kliewer, S. A.; Lehmann, J. M.; Willson, T. M. Science
1999, 284, 757.
8. Brown, P. J.; Smith-Oliver, T. A.; Charifson, P. S.; Tom-
kinson, N. C. O.; Fivush, A. M.; Sternbach, D. D.; Wade,
L. E.; Orband-Miller, L.; Parks, D. J.; Blanchard, S. G.;
Kliewer, S. A.; Lehmann, J. M.; Willson, T. M. Chem. Biol.
1997, 4, 909.
9. Berger, J.; Leibowitz, M. D.; Doebber, T. W.; Elbrecht, A.;
Zhang, B.; Zhou, G.; Biswas, C.; Cullinan, C. A.; Hayes, N. S.;
Li, Y.; Tanen, M.; Ventre, J.; Wu, M. S.; Berger, G. D.;
Mosley, R.; Marquis, R.; Santini, C.; Sahoo, S. P.; Tolman,
R. L.; Smith, R. G.; Moller, D. E. J. Biol. Chem. 1999, 274,
6718.
Replacement of the oxygen closest to the oxazole ring
with sulfur, led to compound 7i which showed similar
cellular PPARd activity (PPARd EC50=0.42 mM) and
better selectivity against PPARg and PPARa when
compared to 7g (PPARd EC50=0.47 mM). Combining
both of these modifications led to compound 7j, which
showed
a dramatic increase in PPARd activity
(EC50=0.006 mM) and selectivity (ꢂ400-fold against
PPARa and 250-fold against PPARg). Replacing oxa-
zole with thiazole (compound 7k) showed further
improvement in PPARd activity (EC50=0.001 mM) as
well as selectivity (1100 against PPARa, and 800 against
PPARg). Compound 7k also showed excellent selectiv-
ity on the murine receptors (EC50=2.5, 1.0, 0.02 mM for
PPARa, PPARg, and PPARd, respectively) and against
other nuclear or non-nuclear receptors.16 Finally com-
pound 7l demonstrated that addition of a fluorine ortho
to the CF3 group, maintained potency and selectivity on
the human and murine receptors (murine PPAR
10. Johnson, T. E.; Holloway, M. K.; Vogel, R.; Rutledge,
S. J.; Perkins, J. J.; Rodan, G. A.; Schmidt, A. J. Steroid Bio-
chem. Mol. Biol. 1997, 63, 1.

M. L. Sznaidman et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1517–1521
1521
11. Brooks, C. D. W.; Summers, J. B. J. Med. Chem. 1996, 39,
2629.
12. Vosper, H.; Patel, L.; Graham, T. L.; Khoudoli, G. A.;
Hill, A.; Macphee, C. H.; Pinto, I.; Smith, S. A.; Suckling,
K. E.; Wolf, C. R.; Palmer, C. N. A. J. Biol. Chem. 2001, 276,
44258.
13. 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.
14. Maloney, P. R.; Parks, D. J.; Haffner, C. D.; Fivush,
A. M.; Chandra, G.; Plunket, K. D.; Creech, K. L.; Moore,
L. B.; Wilson, J. G.; Lewis, M. C.; Jones, S. A.; Willson, T. M.
J. Med. Chem. 2000, 43, 2971.
15. Lambert, M. H., Unpublished data.
16. 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.; Stern-
bach, D. D.; Kliewer, S. A.; Hansen, B. C.; Willson, T. M.
Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 5306.