Discovery of an oxybenzylglycine based peroxisome proliferator activated receptor α selective agonist 2-((3-((2-(4-chlorophenyl)-5-methyloxazol-4- yl)methoxy)benzyl)(methoxycarbonyl)amino)acetic acid (BMS-687453)

Article abstract of DOI:10.1021/jm9016812

An 1,3-oxybenzylglycine based compound 2 (BMS-687453) was discovered to be a potent and selective peroxisome proliferator activated receptor (PPAR) α agonist, with an EC₅₀ of 10 nM for human PPARα and -410-fold selectivity vs human PPAR- in PPA

Full text of DOI:10.1021/jm9016812

2
854 J. Med. Chem. 2010, 53, 2854–2864
DOI: 10.1021/jm9016812
Discovery of an Oxybenzylglycine Based Peroxisome Proliferator Activated Receptor r Selective
Agonist 2-((3-((2-(4-Chlorophenyl)-5-methyloxazol-4-yl)methoxy)benzyl)(methoxycarbonyl)-
amino)acetic Acid (BMS-687453)
,
†
†
†
†
†
†
Jun Li,* Lawrence J. Kennedy, Yan Shi, Shiwei Tao, Xiang-Yang Ye, Stephanie Y. Chen, Ying Wang,
†
†
†
†
†
†
†
†
†
‡
Andr ꢀe s S. Hern ꢀa ndez, Wei Wang, Pratik V. Devasthale, Sean Chen, Zhi Lai, Hao Zhang, Shung Wu, Rebecca A. Smirk,
Scott A. Bolton, Denis E. Ryono, Huiping Zhang, Ngiap-Kie Lim, Bang-Chi Chen, Kenneth T. Locke,
†
†
^
^
^
§
§
Kevin M. O’Malley, Litao Zhang, Rai Ajit Srivastava, Bowman Miao, Daniel S. Meyers, Hossain Monshizadegan,
§
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
Debra Search, Denise Grimm, Rongan Zhang, Thomas Harrity, Lori K. Kunselman, Michael Cap, Pathanjali Kadiyala,
#
Vinayak Hosagrahara, Lisa Zhang, Carrie Xu, Yi-Xin Li, Jodi K. Muckelbauer, Chiehying Chang, Yongmi An,
#
¥
ꢀ
‡
Stanley R. Krystek, Michael A. Blanar, Robert Zahler, Ranjan Mukherjee, Peter T. W. Cheng, and Joseph A. Tino
†
‡
†
†
†
‡
§
Metabolic Diseases Chemistry, Metabolic Diseases/Atherosclerosis Biology, Lead Evaluation, Preclinical Candidate Optimization,
^
#
Department of Chemical Synthesis, Macromolecular Crystallography, Gene Expression and Protein Biochemistry, and CADD,
¥
ꢀ
Bristol-Myers Squibb, Building 13, P.O. Box 5400, Princeton, New Jersey 08543-5400
Received November 13, 2009
An 1,3-oxybenzylglycine based compound 2 (BMS-687453) was discovered to be a potent and selective
peroxisome proliferator activated receptor (PPAR) R agonist, with an EC50 of 10 nM for human
PPARR and ∼410-fold selectivity vs human PPARγ in PPAR-GAL4 transactivation assays. Similar
potencies and selectivity were also observed in the full length receptor co-transfection assays.
Compound 2 has negligible cross-reactivity against a panel of human nuclear hormone receptors
including PPARδ. Compound 2 demonstrated an excellent pharmacological and safety profile in
preclinical studies and thus was chosen as a development candidate for the treatment of atherosclerosis
and dyslipidemia. The X-ray cocrystal structures of the early lead compound 12 and compound 2 in
complex with PPARR ligand binding domain (LBD) were determined. The role of the crystal structure
of compound 12 with PPARR in the development of the SAR that ultimately resulted in the discovery of
compound 2 is discussed.
Introduction
levels have also been correlated with increased cardiovascular
3
risk. Hence, a combination of intensive LDLc lowering along
with HDLc elevation as well as reduction of triglyceride levels
may greatly benefit the treatment of atherosclerosis in cornary
disease patients.
Atherosclerosis/cardiovascular disease is the leading cause
of death for adults in developed countries. Among the pre-
dominant risk factors for atherosclerosis are high levels of
low-density lipoprotein-cholesterol (LDLc ) and triglycerides
and low levels of high-density lipoprotein-cholesterol
a
The peroxisome proliferator activated receptor R (PPARR)
is a member of the intracellular nuclear hormone receptor
superfamily of transcription factors. Upon binding of ligand
agonists, there is a conformational change that leads to the
modulation of a number of PPARR responsive genes. These
genes in turn have pleiotropic effects on plasma lipoprotein
levels, atherosclerosis, insulin sensitization, and inflamma-
(HDLc). Although considerable progress has been made in
the discovery of therapeutics that lower LDLc (especially
statins), atherosclerosis still remains a leading cause of mor-
tality in the developed countries. Several clinical studies have
indicated the limitations of the strategy of reducing athero-
sclerotic cardiovascular disease by lowering LDLc alone
4
1
tion. The endogenous ligands for PPARR are believed to
be fatty acids, and synthetic ligands include the fibrate class
without treatment of other lipid risk factors. Thus, there
has been a gradual realization that additional risk factors
beyond LDLc need to be targeted in order to address the very
substantial residual risk of cardiovascular disease. HDLc
levels have been found to be inversely correlated with the risk
of coronary artery diseases (CAD). Several clinical trials have
shown a marked decrease in the incidence of CAD
of hypolipidemic drugs (fenofibrate, gemfibrozil, and
4
a,14
bezafibrate) currently in clinical use.
decades, the fibrates have been broadly utilized for the treat-
For the past several
5
ment of dyslipidemia and hypertriglyceridemia. This class of
drugs is also used as combination therapy for diabetics, and
6
their new applications are continuously being explored.
2
with increased plasma HDLc. Additionally, high triglyceride
Recently, a delayed-release formulation of fenofibrate for
use along with diet has been shown to help in lowering
triglycerides and LDLc, as well as raising HDLc in dyslipi-
demic patients. Clinical trials with this new formulation of
fenofibrate have demonstratedthat whenusedincombination
with the most commonly prescribed statins, it has helped
*
To whom correspondence should be addressed. Telephone: 609-818-
7
123. Fax: 609-818-6810. E-mail: jun.li@bms.com.
a
Abbreviations: PPAR, peroxisome proliferator activated receptor;
LBD, ligand binding domain; LDLc, low-density lipoprotein-cholesterol;
HDLc, high-density lipoprotein-cholesterol.
r
pubs.acs.org/jmc
Published on Web 03/10/2010
2010 American Chemical Society

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7 2855
Figure 1. Small molecule PPARR agonists.
a
Scheme 1
a
Reagents and conditions: (a) Et
THF, 99%; (d) K
3
N, MeOH, room temp; (b) NaBH
4
3
, 0 ꢀC to room temp, 95% for two steps; (c) methyl chloroformate, aq NaHCO ,
2
CO
3
, MeCN, 70 ꢀC, 71%; (e) LiOH, THF, room temp, 93%.
patients manage all three key lipids better than the corre-
7
sponding therapies alone. Despite the general concern for the
with minimized side effects. Herein we report the oxybenzyl-
glycine based compound 2 (Figure 1) as a potent, highly
selective PPARR agonist with an excellent preclinical safety
profile.
side effects and carcinogenetic potential of PPAR class drugs,
in vitro and in vivo potential antitumor properties of feno-
fibrate and other PPARR agonists through direct and indirect
antiangiogenic effects, as well as anti-inflammatory activity,
Chemistry
8
have also been reported recently. Those findings may further
The synthesis of compound 2 is described in Scheme 1.
Reductive amination of 3-hydroxybenzaldehyde 3 with gly-
cine methyl ester hydrochloride 4 afforded the secondary
amine 5 as a colorless solid in 95% yield. Condensation of
amine 5 with methyl chloroformate gave the methyl carba-
mate 6 in 99% yield as a light-yellow oil. Compound 6 was
reacted with the chloromethyloxazole 7 in the presence of base
at 80 ꢀC to give the methyl ester 8 as a colorless solid in 71%
yieldafter column chromatography. Hydrolysisofester 8 with
aqueous lithium hydroxide gave compound 2 as a colorless
solid in 93% yield. The compounds 9-32 were synthesized in
analogous fashion as described above.
stimulate the study of the potential clinical benefits of feno-
fibrate or other PPARRagonists in cancer treatments, along or
in combination with other therapies, or as a potential tumor-
preventative agent, in addition to its antiatherosclerosis.
In spite of their use in clinical settings, the fibrate drugs are
very weak affinity ligands for PPARR, which results in the
relatively high doses (e.g., 200 mg of fenofibrate) needed to
achieve clinical efficacy, and unwanted side effects may occur
at these high doses. These dose-limiting side effects of the
fibrates may be limiting their broader clinical usage, thus
preventing maximal efficacy of these drugs in protecting
against cardiovascular disease. In the past several years,
several potent PPARR selective agonists have been progressed
9
Results and Discussion
into various phases of clinical development. However, the
development of most of these potent and selective PPARR
agonist clinical candidates has been suspended and none of
them have reached the market because of various reasons
including (primarily) safety concerns. A potent and effica-
cious PPARR agonist with an excellent safety profile may
provide an opportunity for the treatment of atherosclerosis
and dyslipidemia as well as further lowering the risk of CAD
During the course of our concurrent work on PPARR/γ
dual agonists, we made the following observations in our
1
0
transactivation assay (see Table 1). Compound 9 (Figure 2),
which has a two-carbon ethoxy linker between the 1,
4-oxybenzylglycine and phenyloxazole, is a relatively selective
PPARγ agonist, whereas the corresponding analogue 10,
which has a one-carbon linker, is almost equipotent at

2
856 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7
Li et al.
a
Table 1. In Vitro Transactivation EC50 and Binding IC50 Data of Early Leads
compd
R-EC50 (nM)
γ-EC50 (nM)
γ/R EC50 ratio
R-IC50 (nM)
γ-IC50 (nM)
γ/R IC50 ratio
9
289.3
39.1
16.5
8.8
75.4
41.5
0.26
1.06
8.54
941.8
811.2
610.1
347.0
141.4
247.8
162.0
0.15
0.30
0.27
8.04
10
11
12
141.2
1321
150.0
2789.0
a
Compounds were tested for agonist activity on hPPAR-GAL4 HEK transactivation assay. Full PPARR intrinsic activity (relative to fenofibric acid)
was observed for all tested compounds.
Figure 2. Early lead evolution.
PPARR and PPARγ. On the other hand, the closely related
also observed between the oxazole nitrogen, water, and Thr
279. The remaining interactions between ligand and protein
are hydrophobic in nature. In particular, the tolyl carbamate
moiety was observed to bind into a hydrophobic pocket
defined by Ile 272, Phe 273, Leu 347, Phe 351, Ile 354, and
Met 355; no hydrogen bonds or ionic interactions with the
surrounding residues existed. We believed that SAR investi-
gations couldbeconducted within thisportionof the molecule
to further optimize compound binding affinity or reduce
potential liabilities. Additionally, the function of the chloro
substituent of the phenyl group was also unclear in this
binding mode in the PPARR crystal structure, although it
significantly increased the PPARR selectivity in our in vitro
assay. For example, compound 18 (see Table 3), a close
analogue of 12 with a hydrogen rather than Cl at the para-
position, lost almost 6-fold in PPARR potency.
1
,3-oxybenzylglycine analogue 11 was an 8-fold selective
PPARR agonist vs PPARγ. With this in mind, a hybrid
compound 12, which incorporated a one-carbon linker into
the 1,3-oxybenzylglycine framework of 11, was prepared. As
anticipated, this compound was a highly potent PPARR
agonist with >200-fold PPARR selectivity vs PPARγ. The
binding affinity also gave a similar selectivity trend for these
lead compounds. Although compound 12 fulfills our criteria
for in vitro PPARR potency and selectivity vs PPARγ, it has
significant issues that preclude it from further advancement,
including ion channel activity (93% inhibition at 30 μM in
hERG and 58% inhibition at 10 μM in sodium channel patch
clamp assays, respectively) and CYP-450 inhibitory activity
(
e.g., IC = 1.2 μM for the CYP 2C-9 isozyme). Additionally,
50
compound 12 only showed weak to moderate effects in
standard in vivo efficacy models, e.g., failing to lower LDLc
levels in high fat fed hamsters at doses up to 10 (mg/kg)/day in
a 21-day study.
With lead compound 12 in hand, we initiated a systematic
SAR study that targeted each of its separate putative phar-
macophores. The transactivation assay (EC ) was employed
5
0
To attempt to understand the selectivity of 12, an X-ray
cocrystal structure of 12 with the PPARR ligand binding
domain (LBD) was determined to 2.1 A resolution (Figure 3).
as our primary assay for SAR studies because of its better
correlation with the in vivo efficacy in our hands. Our first
approach focused on the left-hand portions of the molecule.
To test the indirect hydrogen-bond interaction between the
oxazole nitrogen with the PPARR binding pocket as shown in
Figure 3, three close analogues of 12 were prepared (Table 2).
Interestingly, the differences in binding affinities between
these compounds were smaller than we expected. However,
the results from the functional assay did indicate that the
regiochemical orientation of the original oxazole of 12 is very
important/optimal. For example, compound 13 (where the
11
˚
Indeed, the 1, 3 oxybenzylglycine central core was found to fit
well within the PPARR binding pocket. As expected, the well-
recognized hydrogen-bonding network was observed between
the carboxylic acid of 12 and neighboring residues (His 440,
Tyr 464, Tyr 314, and Ser 280), which is believed to be
critical for the functional activity of PPARR ligands. In
addition to these standard carboxylate interactions with the
binding pocket, an interesting indirect hydrogen bond was

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7 2857
attenuated the PPARR but maintained the PPARγ transacti-
vation activity. Interestingly, a bulky aliphatic substituent
such as tert-butyl significantly improved the PPARγ potency
while maintaining PPARR potency. This modification re-
sulted in a very potent and well balanced PPARR/γ agonist
2
0 (EC50 ratio γ/R = 2). On the other hand, a 4-phenyl
substituent only slightly improved PPARγ potency but con-
siderably reduced PPARR functional potency, also resulting
in a balanced PPARR/γ agonist 21 (EC50 ratio γ/R = 1.2) but
with relatively weak overall functional potency. Unexpect-
edly, a saturated, polar six-membered analogue with N-linked
morpholine (22) slightly improved functional potency and
selectivity compared to the aromatic phenyl-substituted ana-
logue 21. These results prompted us to further examine the
effect of varying the electronegativity of the substituent at the
4
(
-phenyl position. A polar and electron withdrawing group
CN) at this position provided compound 23 with good
PPARR potency with 93-fold PPARR selectivity vs PPARγ
in the transactivation assay. Interestingly, an electron-donating
substituent such as methoxy gave a compound 24 with similar
functional potency and selectivity to 23, although their
binding affinities are significantly different. This result indi-
cated that the electronic effect of the substituents on the
phenyl ring had minimal consequences on functional potency
or selectivity but may play a more influential role in binding
affinity. Steric effects may play a more dominant role in
determining PPARγ functional activity in this portion of the
molecule. For example, two comparably sized substituents
that have different electronic properties, isopropyl (26) and
trifluoromethyl (27) at the para position of the phenyl,
afforded analogues with very similar functional potencies
and selectivities. On the other hand, analogues with a small
methyl group (25) and a bulky tert-butyl group (20) have very
different PPARγ functional potencies, thus resulting in a
significant disparity in selectivity, although they have similar
PPARR potency. The SAR trend of the binding affinity is not
clear for this portion of the molecule. Overall, the SAR study
of this part of the molecule has indicated that the 4-Cl-phenyl
moiety was optimal, providing the highest level of PPARR
potency and selectivity. Additionally, none of the phenyl-
substituted analogues with acceptable potency or selectivity
showed any significant improvement in comparison with the
lead compound 12 with respect to CYP 2C9 inhibition or
electrophysiological hERG activity (Table 4).
Figure 3. The X-ray crystal structure of PPARR LBD with com-
pound 12 is illustrated at 2.1 A resolution. Compound 12 is shown as
thick sticks and also as a 2D representation. Protein side chains
˚
within 3.9 A of the compound are shown as thin sticks. Residues of
the hydrophobic pocket surrounding the tolyl group are shown with
˚
orange carbons. 2F
o
c
- F electron density is shown as light-blue
mesh contoured at 1σ around the compound. Water molecules are
shown as red spheres. Hydrogen bonds are shown as black dashed
lines. PDB deposition number for PPARR and compound 12 is
3
KDU.
oxazole 4- and 5-substituents are reversed relative to 12) was
significantly less active than 12 in the PPARR functional assay
and had significantly reduced PPARR selectivity vs PPARγ.
Inaddition, compounds 14 and 15, wherea larger thiazole and
“
1
reversed” thiazole respectively replace the original oxazole of
2, exhibited similarly attenuated PPARR binding affinity.
Focusing on eliminating both the cardiovascular liabilities
and the CYP inhibitory activities of the lead molecule 12, we
then turned our attention to modify its other pharmaco-
phores. As indicated in the analysis of the X-ray structure of
12, although the tolyl carbamate group of the “right-hand”
portion of the molecule fits well into a hydrophobic pocket
defined by residues Ile 272, Phe 273, Leu 347, Phe 351, Ile 354,
and Met 355, we hypothesized that alternative/additional
functionalities could be explored and may be tolerated in this
region. To test this hypothesis, we first examined the cyclo-
hexyl carbamate analogue 28, which is the aliphatic counter-
part of compound 12. As anticipated, compound 28 possessed
similar PPARR potency and excellent selectivity. Systemati-
callyreducing the cycloalkyl ringsize fromsix tothree resulted
in compounds 29-31, all of which provided excellent PPARR
potency and good to excellent selectivity. We then extended
this portion of the SAR study to noncyclic alkyl carbamates.
This effort revealed that small, noncyclic alkyl carbamates
provided even better PPARR selectivity. For example,
compound 32, an n-propyl carbamate, achieved >200-fold
Unlike their corresponding oxazole analogues, the two thia-
zole analogues 14 and 15 also showed similar potency in the
functional PPARR assay. All these results suggested that the
indirect hydrogen-bond interaction in this portion of the
molecule might not be critical for optimal binding affinity.
However, our SAR study at this portion of the molecule
indicated thatthe oxazolemoietywith the substitution pattern
of 12 is optimal and plays a critical role in maintaining the
potent PPARR functional activity.
Our next approach focused on the “left hand” portion of
the molecule, the phenyloxazole moiety, with the initial focus
being the effects of the phenyl ring substituents on PPARR/γ
activity. The results are shown in Table 3. Moving the
p-chloro group of 12 to the ortho and meta positions (16,
1
7) reduced both PPARR potency (EC ) and selectivity,
5
0
suggesting that the para-position may be optimal for further
SAR studies to optimize the phenyl substituent(s). A small
substituent such as fluorine (19) at the para-position slightly

2
858 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7
Li et al.
a
Table 2. In Vitro Activity of Alternative Heterocyclic Analogues
a
Compounds were tested for agonist activity in the hPPAR-GAL4 HEK transactivation assays. Full PPARR intrinsic activity (relative to fenofibric
acid) was observed for all tested compounds. n = 1-3.
a
Table 3. In Vitro Activity of Substituted Phenyloxazole Analogues
compd
R
R-EC50 (nM)
γ-EC50 (nM)
γ/R EC50 ratio
R-IC50 (nM)
γ-IC50 (nM)
γ/R IC50 ratio
12
16
17
18
19
20
21
22
23
24
25
26
27
4-Cl
2-Cl
3-Cl
4-H
4-F
8.8
1321
1100
520
150
16
3.9
15
49
2
347
420
422
NA
362
1324
2885
1725
473
375
372
526
569
2789
1039
384
8.0
2.5
68.0
134
48.6
20.5
10.0
340
108
37.0
15.8
9.1
0.9
705
NA
NA
6.3
1007
20.1
398
2300
483
4-C(CH
4-Ph
3
)
3
0.36
0.69
0.69
15.3
2.3
1.2
3.9
93
71
75
32
59
1988
1194
7257
873
4-morpholine
4-CN
4-MeO
417
3445
1124
684
4-Me
4-CH(Me)
4-CF
1361
497
3.6
2
8.7
12.0
284
705
0.94
4.8
3
2737
a
Compounds were tested for agonist activity in the hPPAR-GAL4 HEK transactivation assays. Full PPARR intrinsic activity (relative to fenofibric
acid) was observed for all tested compounds. n = 1-3.
R/γ selectivity. Following up on this SAR trend, further
reducing the carbamate size led to the discovery of lead
compound 2, a methyl carbamate with excellent PPARR
potency and 410-fold PPARR/γ selectivity.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7 2859
Table 4. CYP 2C-9 Isozyme Inhibition and hERG Inhibition IC50
parameter
12
13
14
15
16
17
19
23
21
53.0
24
25
27
CYP 2C9 IC50 (μM)
hERG flux IC50 (μM)
1.2
29.3
3.4
20.0
3.7
>80
7.1
10.0
3.7
>80
0.5
7.3
8.6
38
2.8
8.6
3.5
25.9
4.4
26
<15% of standard) against a panel of human nuclear hor-
mone receptors, including PPARδ LXR and RXR.
A human ApoA1 transgenic mouse model was employed to
evaluate the impact of compound 2 on serum HDLc and
triglyceride levels. PPARR agonists such as fibrates lower
triglycerides and raise HDLc levels in humans. One of the
pathways by which this is accomplished is by increasing
ApoA1, the main protein component of the HDL particle.
However, in normal mice, the murine promoter causes a
reduction (rather than an increase, as in humans) in apoA1
1
8a
and thereby reduces HDL levels.
Therefore, transgenic
mice overexpressing human ApoA1 under the control of the
natural human (rather than the murine) ApoA1 promoter are
widely used to evaluate the HDLc elevating properties of
1
8b
PPARR agonists. Compound 2 and fenofibrate were thus
evaluated in the human ApoA1 transgenic mouse model in a
1
0-day dose response study. These mice still express mouse
PPARR, and compound 2 in the GAL4-mouse PPARR assay
has an EC50 value of 426 nM, which is 43-fold less than its
activity at human PPARR. Therefore, the use of high doses of
compound 2 would be expected to be needed to demonstrate
efficacy. As shown in Table 6, serum ApoA1 protein levels
after 10 days of treatment of compound 2 were increased in a
dose-dependent manner, with the maximal effect likely seen at
the 50-100 (mg/kg)/day dose. These data corroborate the
HDLc elevations observed at these doses (Table 6). As a
reference, fenofibrate at the 100 (mg/kg)/day dose raised
HDLc by 62% after 10 days. By comparison, compound 2,
after 10 days of treatment, also showed dose-dependent
increases in serum HDLc and reductions in triglycerides.
The maximal effect (plateau) was achieved at the 50 (mg/
kg)/day dose where the elevation of HDLc was 135% and the
triglyceride lowering was 78%. These data clearly demon-
strated that compound 2 can robustly elevate HDLc and
lower triglycerides in the human ApoA1 transgenic mouse
model.
˚
Figure 4. X-ray crystal structures of PPARR LBD to 2.7 A resolu-
tion confirming the binding mode of compound 2. Protein side
˚
chains within 4 A of compound are shown as thin sticks. Residues
surrounding the carbon atom of the methyl carbamate are shown
with orange carbons. All other carbons are shown in purple.
Compound is shown as thick sticks and also as a 2D representation.
c
electron density is shown as light-blue mesh contoured at
σ around the compound. Hydrogen bonds are shown as black
2
1
F
o
- F
dashed lines. PDB deposition number for PPARR and compound 2
is 3KDT.
The X-ray crystal structure of compound 2 bound to
1
PPARR was determined to 2.7 A resolution (Figure 4).
1
˚
The structure confirmed the anticipated binding mode of
compound 2 and revealed very similar binding conformations
between compounds 12 and 2 to PPARR. As observed
for the X-ray structure of compound 12, the same hydrogen
bonding network is observed between the carboxylic acid and
residues Tyr 464, His 440, Tyr 314, and Ser 280. Interestingly,
the ether oxygen of the carbamate of compound 2 flips
in orientation relative to compound 12, leaving the methyl
group extended toward Phe 273, His 440, and Val 444, and
the oxygen extended toward the now empty hydrophobic
pocket.
The high fat fed hamster was another animal model used to
further evaluate the in vivo efficacy of compound 2. Hamsters
are very responsive to a high fat diet; the fasting triglyceride
levels and plasma total cholesterol levels (mainly LDLc) are
elevated following the high fat feeding, but HDLc levels
remain relatively unchanged. The advantage of the hamster
model is that the level of hepatic cholesterol synthesis in
hamsters is similar to that in humans, and the serum
lipid profile of high fat diet-fed hamsters resembles that of
dyslipidemic humans with pronounced LDL and VLDL
The potency and selectivity of compound 2 were further
confirmed by testing it in co-transfection assays in HepG2
cells using full length human PPARR and PPARγ. In this
assay, compound 2 showed excellent PPARR potency
1
9
cholesterol peaks. The hamster has been shown to be a
useful preclinical model of human lipoprotein metabolism
2
0
and atherosclerosis. In our study, the hamster model was
primarily used for evaluation of the compound’s efficacy to
lower triglyceride and LDLc levels.
(
(
EC50 = 47 nM) with ∼50-fold selectivity vs PPARγ
EC50 = 2400 nM). This result correlated well with the data
observed from the primary GAL-4 HEK transactivation
based screening assays as shown in Table 5. Compound 2
was found to be less potent in rodent PPARR functional
assays, with a moderate EC50 of 426 nM for mouse and 488
nM for hamster but remains a full PPARR agonist in both
Evaluation of the efficacy of compound 2 was conducted in
a dose response study ranging from 1 to 10 (mg/kg)/day for 3
weeks. Table 7 shows the fasting plasma lipid parameters at
the conclusion of the study. Compound 2 lowered plasma
triglycerides by 60% at the 3 (mg/kg)/day dose and 91% at the
10 (mg/kg)/day dose, suggesting that the effect of compound2
on triglycerides lowering reaches a plateau at the 10 (mg/kg/)
1
2
species. Importantly, compound 2 showed negligible activ-
ity (EC50 for transactivation of >25 μM and efficacies of

2
860 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7
Li et al.
a
Table 5. In Vitro Activity of Aliphatic Carbamate Analogues
compd
X
R-EC50 (nM)
γ-EC50 (nM)
γ/R EC50 ratio
R-IC50 (nM)
γ-IC50 (nM)
γ/R IC50 ratio
2
2
3
3
3
2
8
9
0
1
2
cyclohexyl-
cyclopentyl-
cyclobutyl-
cyclopropylmethyl-
n-propyl-
7.5
9.2
12.5
6.3
6.0
10
843
1482
905
112
161
72
513
408
228
351
336
260
4498
9444
8.8
23.1
5293
5336
23.2
15.2
826
131
209
410
1257
4100
10310
>15000
30.7
Me-
>57
a
Compounds were tested for agonist activity on hPPAR-GAL4 HEK transactivation assay. Full PPARR intrinsic activity (relative to fenofibric acid)
was observed for all tested compounds. n = 1-3.
a
Table 6. Effect of Compound 2 and Fenofibrate on Plasma Parameters in Human ApoA1 Transgenic Mice
serum hu-ApoA1 mg/dL ( SEM
(% change)
TG mg/dL ( SEM
(% change)
HDLc mg/dL ( SEM
(% change)
treatment
vehicle
693 ( 51
189 ( 32.9
158 ( 18.5
fenofibrate, 100 (mg/kg)/day
compound 2, 10 (mg/kg)/day
compound 2, 50 (mg/kg)/day
compound 2, 100 (mg/kg)/day
a
1957 ( 138 (182%*)
1163 ( 61 (67.88%*)
2219 ( 176 (220%*)
2511 ( 158 (262%*)
64.1 ( 3.0 (-66%*)
95.8 ( 3.7 (-49.4%*)
72.6 ( 7.1 (-61.7%*)
40.8 ( 3.2 (-78.5%*)
255 ( 21.3 (61.7%*)
250 ( 29.6 (58.1%*)
372 ( 20.5 (135.4%*)
364 ( 27.1 (130.1%*)
p < 0.05 compared to vehicle-treated group (n = 10). ApoA-I transgenic mice (n = 10) were treated for 10 days with compound dosed by oral
gavage. Blood was drawn after a 4 h fast on day 10 after the final dose to measure plasma lipids. Human ApoA1 protein in the serum was measured by a
using apolipoprotein A1 kit (Polymedco). Also see ref 12.
a
Table 7. Effect of Compound 2 on Plasma Parameters in High Fat-Fed Hamsters
fenofibrate
(100 mg/kg)
compound 2
(1 mg/kg)
compound 2
(3 mg/kg)
compound 2
(10 mg/kg)
entry
vehicle
TG mg/dL ( SEM (% change)
LDLc mg/dL ( SEM (% change)
a
644 ( 139
326 ( 53
132 ( 20 (-79%)
93 ( 12 (-71%)
784 ( 195 (þ21%)
196 ( 21 (-40%)
258 ( 51 (-60%)
97 ( 10 (-70%)
58 ( 5 (-91%)
42 ( 4 (-87%)
p < 0.05 versus vehicle control. Compound 2 lower serum triglycerides and LDLc in fat fed hamsters. Male Syrian golden hamsters on a high fat diet
were dosed daily by oral gavage for 21 days. Blood samples were drawn for serum lipid measurements after an 18 h fast and 24 h after the last
dose. Fenofibrate (Feno) at 100 mpk was used as a positive control. The compounds and doses are indicated. Data represent the mean ( SEM (n = 8).
Also see ref 12.
a
Table 8. Pharmacokinetic Profile of 2
species
mouse
dose route
dose (mg/ kg)
T
max (h)
C
max (μM)
AUC (μM h)
CLPl ((mL/min)/kg)
10.8
V
ss (L/kg)
T
1/2 (h)
F (%)
3
b
iv
po
6
19.3
33.9
1.3
3.0
b
12
5
0.25
23.8
88
91
58
75
c
ia
rat
48 ( 9
4.3 ( 0.9
9.5 ( 0.7
8.9 ( 4.1
0.7 ( 0.1
1.8 ( 0.8
3.5 ( 1.7
3.2 ( 0.2
7.9 ( 4.1
11.9 ( 4.0
po
iv
10
1
0.4 ( 0.1
0.9 ( 0.1
0.8 ( 0.3
43 ( 14
1.0 ( 0.6
1.9 ( 0.7
88 ( 10
3.9 ( 0.3
4.5 ( 2.0
5.8 ( 3.6
8.5 ( 4.9
c
dog
po
iv
2
monkey
1
po
2
a
b
For each experimental study, n g 3. AUC0-8h reported here, not AUCINF
.
Intra-arterial administration.
day dose. Similar effects were also observed on LDLc with
7
respectively. Overall, compound 2 significantly lowers plasma
triglycerides and LDLc in the chronic dyslipidemic hamster
model.
As shown in Table 8, compound 2 has an excellent phar-
macokinetic profile across all tested animal species. The oral
absorption was rapid, with Tmax ranging between 0.25 and
oral dose), 43 μM in rat (10 mg/kg oral dose), 1 μM in dog
(2 mg/kg oral dose), and 1.9 μM/mL in monkeys (2 mg/kg oral
dose). Compound 2 exhibited low plasma clearance in the
mouse, rat, and monkey and moderate plasma clearance in
the dog, and the volume of distribution ranged from 0.7 L/kg
(rat) to 3.5 L/kg (cynomolgus monkey), which was comparable
to the total body water in the rat and greater than total body
water in the mouse, dog, and monkey. The half-life of com-
pound 2 ranged from 3 h in mouse to 12 h in cynomolgus
monkeys. Compound 2 also possessed excellent absolute oral
0% and 87% reductions at the 3 and 10 (mg/kg)/day doses,
0.9 h in mouse, rat, dog, and cynomulgus monkey. The
corresponding Cmax values were 23.8 μM in mouse (12 mg/kg

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7 2861
bioavailability ranging from 58% (dog) to 91% (rat). Addi-
tionally, compound 2 has excellent pharmaceutical properties,
with a crystalline aqueous solubility being 280 μg/mL at pH 6.5,
increasing to >4 mg/mL at pH 7.9.
mixtures of solvent C (10% MeOH/90% H
solvent D (90% MeOH/10% H O/0.1% TFA) or of solvent E
10% CH CN/90% H O/0.1% TFA) and solvent F (90%
CH CN /10% H O/0.1% TFA). All other reagents and solvents
were obtained from commercial sources and were used without
further purification.
2
O/0.1% TFA) and
2
(
3
2
3
2
Compound 2 was not a significant inhibitor of human
CYP1A2, CYP2C8, CYP2C9, CYP2C19, CYP2D6, or
CYP3A4(IC50 > 40μM). In addition, noinduction of human
PXR was observed up to 50 μM. No glutathione conjugates
were detected when compound 2 at 30 μM was incubated
with mouse, rat, or human liver microsomes fortified with
glutathione(GHS, 5 mM). This result suggests thatthe overall
formation of oxidative reactive metabolites may be low in
humans in vivo. No in vitro liabilities were noted in extensive
screens for receptor/enzyme binding inhibition, human hepato-
cyte toxicity, bacterial mutagenicity, or CHO cell clasto-
genicity. In comparison to lead compound 12, compound 2
has a significantly improved cardiac ion channel liability
profile. Compound 2 showed minimal activity against both
hERG (1.8% at 10 μM and 4.0% at 30 μM in a patch-clamp
assay) and sodium channels (at 10 μM drug, 6.9% inhibition
at 1 Hz and 9.5% at 4 Hz). Additionally, no drug-related
changes in cardiovascular parameters [e.g., hemodynamic and
electrocardiographic (ECG) effects] were observed at up to
Methyl 2-(3-Hydroxybenzylamino)acetate (5). To a solution
of glycine methyl ester hydrochloride 4 (84.86 g, 0.67 mol) in
MeOH (900 mL) was added Et N (68.29 g, 0.675 mol). After
3
1
5 min, a solution of 3-hydroxybenzaldehyde (75 g, 0.614 mol)
in MeOH (500 mL) was added. After being stirred for 1 h,
(5.7 g,
the reaction mixture was cooled to 0 ꢀC, and NaBH
150 mmol) was then added portionwise over 20 min. After the
mixture was stirred for 1 h, volatiles were removed in vacuo at
4
4
5-50 ꢀC from the reaction mass. The resulting residue was
partitioned between EtOAc (500 mL) and water (500 mL), with
the aqueous layer being washed again with EtOAc(200 mL). The
combined organic extracts were washed with brine, dried
(
product 5 as a pale-yellow solid (113.9 g, 95%), which was used
2 4
Na SO ), then concentrated in vacuo to afford the desired
1
in the next step without further purification. H NMR (CDCl ,
3
400 MHz) δ 7.19 (t, J = 7.9 Hz, 1H), 6.87 (m, 1H), 6.81 (m, 1H),
6.72 (m, 1H), 3.76 (s, 2H), 3.74 (s, 2H), 3.43 (s, 3H); H NMR
1
(DMSO-d , 400 MHz) δ 9.26 (s, 1H), 7.07 (t, J = 8 Hz, 1H), 6.72
6
(
(
(
s, 1H), 6.69 (d, J = 8 Hz, 1H), 6.61 (d, J = 1 Hz, 1H), 3.76
1
s, 2H), 3.61 (s, 3H), 3.28 (s, 2H), 2.35 (s, 1H); C NMR
3
2
0 mg/kg compound 2 in a single-dose monkey telemetry
DMSO-d
6
, 100 MHz) δ 172.6, 157.4, 141.7, 129.9, 118.6,
study. Ames testing also showed that compound 2 was
not mutagenic to the tester strains TA 98 and TA 100 at up
to 5000 μg per plate.
þ
1
14.8, 113.7, 52.1, 51.3, 49.3. LCMS [M þ H] : 196.1.
Methyl 2-((3-Hydroxybenzyl)(methoxycarbonyl)amino)-
acetate (6). To a stirred 0 ꢀC solution of compound 5 (65.0 g,
33 mmol) in THF (325 mL) and saturated aqueous NaHCO
260 mL) was added dropwise methyl chloroformate (25.7 mL,
33 mmol) over 20 min under nitrogen. The mixture was stirred at
0 ꢀC for 45 min and extracted with EtOAc (2 ꢀ 260 mL). The
organic extracts were dried over anhydrous Na SO , filtered, and
In summary, compound 2 is a potent, orally active PPARR
selectiveagonist thatis highly efficacious in elevatingHDLc in
human ApoA1 transgenic mice and lowering LDLc in dyslipi-
demic hamsters in chronic studies. Compound 2 also robustly
lowers plasma triglycerides in both animal models. The like-
lihood of drug-drug interaction of compound 2 should be
minimal because of its negligible inhibition in all tested CYP
isozymes or induction in the human PXR transactivation
assay. No cardiovascular pharmacology safety issues were
identified with in vitro screens as well as in preclinical
animal models. On the basis of its excellent pharmacokinetic
and pharmcodynamic properties, as well as superior in
vitro liability profile, compound 2 was selected as a develop-
ment candidate for further evaluation for the treatment of
atherosclerosis.
3
(
3
3
2
4
concentrated in vacuo to give crude product 6 (83.8 g, 99.4%) as a
yellow oil. The material was used directly in the next step without
1
further purification. H NMR (CDCl , 400 MHz) δ 7.19 (m, 1H),
3
6
3
1
1
5
.78 (m, 3H), 6.14 (br s, 1H), 4.56 (s, 1H), 4.52 (s, 1H), 3.98 (s, 1H),
.90 (s, 1H), 3.83 (s, 1.5H), 3.77 (s, 1.5H), 3.74 (s, 1.5H), 3.72 (s,
.5H). C NMR (CDCl , 100 MHz) δ170.2, 170.1, 157.4, 157.2,
3
13
56.6, 138.3, 138.2, 129.9, 129.8, 120.1, 119.4, 115.0, 114.8, 114.3,
3.3, 52.2, 51.4, 51.0, 47.8, 47.3. LCMS [M þ H] : 254.2.
þ
Methyl 2-((3-((2-(4-Chlorophenyl)-5-methyloxazol-4-yl)methoxy)-
benzyl)(methoxycarbonyl)amino)acetate (8). To a solution of com-
pound 6 (83.8g, 331 mmol) inMeCN (700 mL) was addedoxazole
chloride 7 (80.1 g, 331 mmol) and anhydrous K
93 mmol). The mixture was heated at 70 ꢀC for 21 h under
nitrogen, cooled to 5 ꢀC, poured into saturated aqueous NH Cl
(1000 mL), and extracted withEtOAc (400 mL). The organiclayer
was dried over anhydrous Na SO , filtered, and concentrated in
2 3
CO (137 g,
9
Experimental Section
1
General Chemistry Methods. H (400 MHz) and C (100 MHz)
4
13
NMR spectra were recorded on a JEOL GSX400 spectrometer
using Me Si as an internal standard unless otherwise noted.
2
4
4
vacuo to yield the product (155.8 g) as a yellow syrup, which was
purified by flash column chromatography (2.5 kg silica gel, elution
with 20-50% EtOAc/heptane) to give the desired product 8
LC-MS spectra were obtained on a Shimadzu HPLC and
Micromass Platform using electrospray ionization. HRMS spectra
were obtained on a Micromass LCT in lockspray with electro-
spray ionization. Analytical HPLC analyses were performed on a
Shimadzu instrument using one of the following reverse phase
methods, with UV detection set at 220 nm: (method A) Phenom-
enex S5 ODS 4.6 mm ꢀ 50 mm column, gradient elution 0-100%
1
(108.1 g, 70.7%) as a colorless solid. Mp 83.4 ꢀC. H NMR
3
(CDCl , 400 MHz) δ 7.97 (m, 2H), 7.42 (d, 2H, J = 8.4 Hz), 7.28
(m, 1H), 6.89 (m, 3H), 4.98 (s, 2H), 4.59 (s, 1H), 4.53 (s, 1H), 3.97
(s, 1H), 3.88 (s, 1H), 3.80 (s, 1.5H), 3.72 (s, 1.5H), 3.73 (s, 1.5H),
13
3
3.72 (s, 1.5H), 2.45 (s, 3H); C NMR (CDCl , 100 MHz) δ170.5,
B/A over 4 min (solvent A = 10% MeOH/H
2
O containing 0.1%
170.4, 159.5, 159.3, 159.2, 157.5, 157.3, 147.8, 138.9, 136.6, 136.6,
132.5, 130.2, 130.1, 129.4, 127.8, 126.3, 126.3, 121.4, 120.7, 114.9,
114.6, 114.3, 114.1, 62.5, 53.6, 52.5, 51.7, 51.4, 48.1, 47.6, 10.9;
H PO , solvent B = 90% MeOH/H O containing 0.1% H PO ),
4
3
4
2
3
flow rate 4 mL/min; (method B) Zorbax S5 SB-C18 4.6 mm ꢀ
5 mm column, gradient elution 0-100% B/A over 8 min (solvent
A = 10% MeOH/H O containing 0.1% H PO , solvent
þ
7
HRMS m/e 459.1323 (M ). Anal. Calcd for C23
2 6
H23ClN O : C,
2
3
4
60.20; H, 5.05; N, 6.10. Found: C, 60.41; H, 5.08; N, 6.00.
B = 90% MeOH/H O containing 0.1% H PO ), flow rate
2
2-((3-((2-(4-Chlorophenyl)-5-methyloxazol-4-yl)methoxy)benzyl)-
(methoxycarbonyl)amino)acetic Acid (2). To a stirred solution of
methyl ester 8 (108 g, 235 mmol) in THF (732 mL) and water
(366 mL) was added LiOH H O (24.6 g, 585.9 mmol). The
mixture was stirred at room temperature under nitrogen for 2 h
and diluted with EtOAc (200 mL). The solution was brought to
3
4
2.5 mL/min. The purity of all final compounds is g95%, deter-
mined by analytic HPLC method A and confirmed by analytic
HPLC method B.
Preparative HPLC was carried out on an automated Shimadzu
system using YMC ODS C18 5 μm preparative columns with
3
2

2
862 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7
Li et al.
2
2
pH 2 by the addition of aqueous 1 N HCl. The organic layer was
separated, and the aqueous layer was extracted with EtOAc
beamline. The images were processed and scaled using HKL2000
(Supporting Information). Refinement and model building were
carried out using the program MIFit (Supporting Information).
23
(
2 ꢀ 250 mL). The combined organic extracts were washed with
11
water (2 ꢀ 150 mL), dried over anhydrous Na
2
SO , and con-
4
The structures have been deposited into the PDB.
centrated in vacuo to give crude 2 (97 g, 93%). HPLC analysis
showed that the purity of this batch was 98.5%.
In Vitro Assays. A homogeneous, fluorescent polarization
PPARR and PPARγ binding assay was used as the primary
screen for determining the PPARR and PPARγ binding affinity
of compounds. The human functional activity of PPARR and
PPARγ agonists was determined by using the GAL4-LBD
For recrystallization, crude compound 2 (210 g, material
combined from several batches) was dissolved in hot EtOAc
1
3
(
6
1200 mL) at 78 ꢀC, then was cooled to room temperature over
0 min, then further cooled to 5 ꢀC. The slurry was stirred at 5 ꢀC
1
5,17
assays as previously described.
The in vitro hamster, rat,
for 40 min and filtered. The filter cake was washed with cold
EtOAc (2 ꢀ 100 mL). The colorless solid was dried under
vacuum at 55 ꢀC for 8 h until a constant weight was obtained.
The weight of the solid was 191 g (91% recovered yield). HPLC
and mouse PPARR functional activities were tested in the
chimeric GAL4/PPARR assay format described for human
1
5,17
PPARR as above.
calculated using XLfit 4 parameter fit and floating all para-
The data are reported as an EC value
50
1
16
analysis showed that the purity of this batch was >99%. H
meters. Full length human PPARR and PPARγ co-transfec-
tion assays in HepG2 cells were employed for further testing the
leading compounds as reported by Mukherjee et al.
In Vivo Assays. Human apoA1 Transgenic Mice Lipid Stu-
dies. Male 6-8 week old human apoA1 transgenic mice were
randomly assigned into different treatment groups and weighed
and dosed by oral gavage (5 mL/kg body weight) once a day in
the morning with vehicle alone or with compound and allowed
free access to food and water. The study duration was 10 days.
After dosing on day 10, mice were fasted for 4 h and sacrificed by
NMR (DMSO-d
6
, 500 MHz, 65 ꢀC) δ 12.47 (s, 1H), 7.93 (d, J =
.8 Hz, 2H), 7.56 (d, J = 8.4 Hz, 2H), 7.25 (t, J = 8.1 Hz, 1H),
.91-6.98 (m, 2H), 6.86 (d, J = 7.5 Hz, 1H), 4.98 (s, 2H), 4.44 (s,
12,17
8
6
2
d
1
1
1
3
H), 3.85 (s, 2H), 3.62 (s, 3H), 2.43 (s, 3H). C NMR (DMSO-
6
, 126 MHz, 19.8 ꢀC) δ 170.7,170.6, 158.2, 157.9, 156.4, 147.8,
39.2, 135.0, 132.1, 129.2, 127.3, 125.6, 120.1, 119.7, 114.0,
13.4, 61.1, 52.6, 51.0, 50.6, 48.4, 47.9, 39.5, 9.9. HRMS-
þ
(M þ H) = 445.1173 (Δ = 1.4 ppm). Anal. Calcd for C22
ClN O : C, 59.39; H, 4.75; N, 6.29; Cl, 7.97. Found: C, 59.40; H,
21
H -
2
6
4
.74; N, 6.22; Cl, 8.03.
Crystallography. Protein Expression and Purification.
CO asphyxiation, and blood samples were collected in serum-
2
separating tubes via cardiac puncture for lipid measurements.
Livers were dissected out, weighed, and quickly frozen in liquid
nitrogen for future RNA analysis. Human apoA1 concentration
in serum was measured using the apolipoprotein A1 kit
(Polymedco).
Hamsters Lipid Studies. Male Syrian golden hamsters were
acclimated to 12 h light/dark reverse light cycle for 7 days with
high fat diet, then dosed daily by oral gavage for 21 days while
on the same diet. At the end of the experiment, blood samples
were drawn retro-orbitally after an 18 h fast and 24 h after the
last dose for the determination of serum lipid levels. Livers were
dissected out for mRNA analysis.
Statistical Analysis. Mean values for body weight and food
consumption, values obtained from clinical laboratory tests,
and organ weights of treated groups were compared to those of
the control group using Dunnett’s test. Statistical comparisons
across dose groups were performed using Tukey all pair com-
parison. A p value of <0.05 was considered as significant
changes. Data are expressed as the mean ( SEM.
PPARR LBD (E196-Y468) protein was expressed and purified
2
1
as described by Cronet et al. with the following modifications.
Freshly transformed E. coli (BL21-DE3, Novagen) were grown
at 37 ꢀC to an OD600nm of 0.6 in M9 minimal media supple-
mented with casamino acids (Difco), trace minerals, and 30 μg/
mL kanamycin. Cultures were chilled on ice for 30 min. Then
overexpression was induced by addition of 0.4 mM isopropyl-β-
D-thiogalactopyranoside, and the cultures were incubated for
1
8 h at 20 ꢀC. Harvested cells were disrupted using a high-
pressure homogenizer (Rannie). Supernatant loaded onto Ni-
NTA was eluted with a 0-500 mM imidazole gradient. The
histidine tag was removed using 10 units of human R-thrombin
(
Enzyme Research Labs) per 1.0 mg of purified protein for 2 h.
At the completion of the cleavage reaction, thrombin was
removed using a benzamidine resin. After removal of the
histidine tag, PPARR was further polished on Q-Sepharose
HP column (Amersham-Pharmacia Biotech) and protein elu-
tion was carried out with 10 column volumes of 20-500 mM
NaCl gradient. The final samples were flash-frozen at 1.0 mg/
mL concentration and stored at -80 ꢀC.
Acknowledgment. The authors thank Dr. Joel C. Barrish
for proofreading the manuscript. The Discovery Toxicology
department at Bristol-Myers Squibb is acknowledged for the
preclinical safety evaluation of compounds.
Protein Crystallization. PPARR protein at 1.0 mg/mL in
2
0 mM Tris-HCl, pH 8.0, 0.15 M sodium chloride, 1 mM TCEP,
and 10% glycerol was used for the crystallization trials with
compounds 2 and 12. Deoxy Big CHAP (DBC) was added to a
final concentration of 0.7 mM before the addition of 5-fold
molar excess compound to the protein solution. The complex
was incubated at 4 ꢀC overnight, concentrated to 8.0 mg/mL,
and then diluted to 2.0 mg/mL with water just prior to crystal-
lization. Crystal trials were performed at room temperature
using the hanging-drop vapor-diffusion method. The hanging
drop contained 1 μL of protein solution and 1 μL of reservoir
solution (26-27% PEG 4000, 200 mM ammonium acetate,
Supporting Information Available: Protocols and methods of
in vitro and in vivo assays, analytical and spectroscopic data for
compounds 12-32, and X-ray crystallographic data of com-
pounds 12 and 2. This material is available free of charge via the
Internet at http://pubs.acs.org.
References
1
2
0 mM magnesium acetate, and 20 mM Tris-HCl, pH 7.0, or
00 mM MES, pH 6.5). Rod-shaped crystals of PPARR com-
(1) (a) Cannon, C. P.; Braunwald, E.; McCabe, C. H.; Rader, D. J.;
Rouleau, J. L.; Belder, R.; Joyal, S. V.; Hill, K. A.; Pfeffer, M. A.;
Skene, A. M. Intensive Versus Moderate Lipid Lowering with
Statins after Acute Coronary Syndromes. N. Engl. J. Med. 2004,
plexed with compound were obtained within 4-6 days.
Protein Structure Determination and Refinement. Crystals
were transferred briefly into a solution of 25% (v/v) glycerol,
3
50, 1495–1504. (b) LaRosa, J. C.; Grundy, S. M.; Waters, D. D.; Shear,
C.; Barter., P.; Fruchart, J. C.; Gotto, A. M.; Greten, H.; Kastelein, J. J.;
Shepherd, J.; Wenger, N. K. Intensive Lipid Lowering with Atorvastatin
in Patients with Stable Coronary Disease. N. Engl. J. Med. 2005, 352,
2
6% PEG 4000, 200 mM ammonium acetate, 10 mM
magnesium acetate, and 20 mM Tris-HCl, pH 7.0, and flash-
cooled in preparation for cryodata collection. Data of PPARR
with compound 12 were collected at the Advanced Photon
Source, Argonne National Laboratory, on the IMCA-CAT
beamline ID-17. Data of PPARR with compound 2 were
collected at Brookhaven National Laboratory on the SLS X25
1
425–1435. (c) Linsel-Nitschke, P.; Tall, A. R. HDL as a Target in the
Treatment of Atherosclerotic Cardiovascular Disease. Nat. Rev. Drug
Discovery 2005, 4, 193–205.
(
2) (a) Frick, M. H.; Elo, O.; Haapa, K.; Heinonen, O. P.; Heinsalmi,
P.; Helo, P.; Huttunen, J. K.; Kaitaniemi, P.; Koskinen, P.;
Manninen, V.; et al. Helsinki Heart Study: Primary-Prevention

Article
Trial with Gemfibrozil in Middle-Aged Men with Dyslipidemia.
Safety of Treatment, Changes in Risk Factors, and Incidence of
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7 2863
Wang, X.; Thompson, R. C.; Dominianni, S. J.; Kauffman, R. F.;
Singh, J.; Bean, J. S.; Bensch, W. R.; Barr, R. W.; Osborne, J.;
Montrose-Rafizadeh, C.; Zink, R. W.; Yumibe, N. P.; Huang, N.;
Luffer-Atlas, D.; Rungta, D.; Maise, D. E.; Mantlo, N. B. Design and
Synthesis of a Potent and Selective Triazolone-Based Peroxisome
Proliferator-Activated Receptor Agonist. J. Med. Chem. 2003, 46,
5121–5124. (f) Nomura, M.; Tanase, T.; Ide, T.; Tsunoda, M.; Suzuki,
M.; Uchiki, H.; Murakami, K.; Miyachi, H. Design, Synthesis, and
Evaluation of Substituted Phenylpropanoic Acid Derivatives as Human
Peroxisome Proliferator Activated Receptor Activators. Discovery of
Potent and Human Peroxiome Proliferator Activated Rceptor
Subtype-Selective Activators. J. Med. Chem. 2003, 46, 3581–3599.
(g) Shi, G. Q.; Dropinski, J. F.; Zhang, Y.; Santini, C.; Sahoo, S. P.;
Berger, J. P.; MacNaul, K. L.; Zhou, G.; Agrawal, A.; Alvaro, R.; Cai,
T.-Q.; Hernandez, M.; Wright, S. D.; Moller, D. E.; Heck, J. V.; Meinke,
P. T. Novel 2,3-Dihydrobenzofuran-2-carboxylic Acids: Highly Potent
and Subtype-Selective PPARR Agonists with Potent Hypolipidemic
Activity. J. Med. Chem. 2005, 48, 5589–5599. (h) Yamazaki, Y.; Abe,
K.; Toma, T.; Nishikawa, M.; Ozawa, H.; Okuda, A.; Araki, T.; Oda, S.;
Inoue, K.; Shibuya, K.; Staels, B.; Fruchart, J. C. Bioorg. Med. Chem.
Lett. 2007, 17, 4689–4693.
Coronary Heart Disease. N. Engl. J. Med. 1987, 317, 1237–1245.
(
b) Rubins, H. B; Robins, S. J.; Collins, D.; Fye, C. L.; Anderson, J. W.;
Elam, M. B.; Faas, F. H.; Linares, E.; Schaefer, E. J.; Schectman, G.;
Wilt, T. J.; Wittes, J. Gemfibrozil for the Secondary Prevention of
Coronary Heart Disease in Men with Low Levels of High-Density
Lipoprotein Cholesterol. Veterans Affairs High-Density Lipoprotein
Cholesterol Intervention Trial Study Group. N. Engl. J. Med. 1999,
341, 410–418.
(
3) (a) Goldbourt, U.; Yaari, S.; Medalie, J. H. Isolated Low
HDL Cholesterol as a Risk Factor for Coronary Heart Disease
Mortality. A 21-Year Follow-Up of 8000 Men. Arterioscler.,
Thromb., Vasc. Biol. 1997, 17, 107–113. (b) Genest, J. J.; McNamara,
J. R.; Salem, D. N.; Schaefer, E. J. Prevalence of Risk Factors in
Men with Premature Coronary Artery Disease. Am. J. Cardiol. 1991,
67, 1185–1189.
(4) (a) Cheng, T. W. P.; Mukherjee, R. PPARs as Targets for Meta-
bolic and Cardiovascular Dieases. Mini-Rev. Med. Chem. 2005, 5,
7
41–753. (b) Willson, T. M.; Brown, P. J.; Sternbach, D. D.; Henke, B.
R. The PPARs: From Orphan Receptor to Drug Discovery. J. Med.
Chem. 2000, 43, 527–550. (c) Sher, T.; Yi, H. F.; McBride, O. W.;
Gonzalez, F. J. cDNA Cloning, Chromosomal Mapping, and Functional
Characterization of the Human Peroxisome Proliferator Activated
Receptor. Biochemistry 1993, 32, 5598–5604. (d) Mukherjee, R.;
Jow, L.; Noonan, D.; McDonnell, D. P. Human and Rat Peroxisome
Proliferator Activated Receptors (PPARs) Demonstrate Similar Tissue
Distribution but Different Responsiveness to PPAR Activators.
J. Steroid. Biochem. Mol. Biol. 1994, 51, 157–166.
(10) Devasthale, P. V.; Chen, S.; Jeon, Y.; Qu, F.; Shao, C.; Wang, W.;
Zhang, H.; Cap, M.; Farrelly, D.; Golla, R.; Grover, G.; Harrity,
T.; Ma, Z.; Moore, L.; Ren, J.; Seethala, R.; Cheng, L.; Sleph, P.;
Sun, W.; Tieman, A.; Wetterau, J. R.; Doweyko, A.; Ghandreasena,
G.; Chang, S. Y.; Humphreys, W. G.; Sasseville, V. G.; Biller, S. A.;
Ryono, D. E.; Selan, F.; Hariharan, N.; Cheng, R. T. W.
Design and Synthesis of N-[(4-Methoxyphenoxy)carbonyl]-N-
[[4-[2-(5-methyl-2-phenyl-4-oxazolyl)ethoxy]phenyl]methyl]glycine
[Muraglitazar/BMS-298585], a Novel Peroxisome Proliferator-
Activated Receptor R/γ Dual Agonist with Efficacious Glucose
and Lipid-Lowering Activities. J. Med. Chem. 2005, 48, 2248–
2250.
(11) PDB deposition number for PPARR and compound 2 is 3KDT.
PDB deposition number for PPARR and compound 12 is 3KDU.
(12) Mukherjee, R.;Locke, K. T.;Miao, B.;Meyers, D.;Monshizadegan,
H.; Zhang, R.; Search, D.; Grimm, D.; Flynn, M.; O’Malley, K. M.;
Zhang, L.; Li, J.; Shi, Y.; Kennedy, L. J.; Blanar, M.; Cheng, P. T.;
Tino, J. A.; Srivastava, R. A. Novel Peroxisome Proliferator-
Activated Receptor R Agonists Lower Low-density Lipoprotein
and Triglycerides, Raise High-density Lipoprotein, and Synergisti-
cally Increase Cholesterol Excretion with a Liver X Receptor
Agonist. J. Pharmacol. Exp. Ther. 2008, 327, 716–726.
(13) Seethala, R.; Golla, R.; Ma, Z.; Zhang, H.; O’Malley, K.; Lippy, J.;
Cheng, L.; Mookhtiar, K.; Farrelly, D.; Zhang, L.; Hariharan, N.;
Cheng., P. T. A Rapid, Homogeneous, Fluorescence Polarization
Binding Assay for Peroxisome Proliferator-Activated Receptors
Alpha and Gamma Using a Fluorescein-Tagged Dual PPAR
Alpha/Gamma Activator. Anal. Biochem. 2007, 363, 263–274.
(14) Kliewer, S. A.; Sundseth, S. S.; Jones, S. A.; Brown, P. J.; Wisely,
G. B.; Koble, C. S.; Devchand, P.; Wahli, W.; Willson, T. M.;
Lenhard, J. M.; Lehmann, J. M. Fatty Acids and Eicosanoids
Regulate Gene Expression through Direct Interactions with Per-
oxisome Proliferator-Activated Receptors Alpha and Gamma.
Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 4318–4323.
(
5) (a) Staels, B.; Maes, M.; Zambon, A. Fibrates and Future PPARR
Agonists in the Treatment of Cardiovascular Diesease. Nat. Clin.
Pract. Cardiovasc. Med. 2008, 5, 542–553. (b) Keating, G. M.;
Ormrod, D. Micronised Fenofibrate. Drugs 2002, 62, 1909–1944.
6) (a) Athyros, V. G.; Papageorgiou, A. A.; Athyrou, V. V.; Demi-
triadis, D. S.; Kontopoulos, A. G. Atorvastatin and Micronized
Fenofibrate along and in Combination in Type 2 Diabetes with
Combined Hyperlipidemia. Diabetes Care 2002, 25, 1198–1202. (b)
Grundy, S. M.; Vega, G. L.; Yuan, Z.; Battisti, W. P; Brady, W. E;
Palmisano, J. Effectivenness and Tolerability of Simvastatin Plus
Fenofibrate for Combined Hyperlipidemia (SAFARI trial). Am.
J. Cardiol. 2005, 95, 2088–2093.
(
(
(
7) Trilipix is trademark of Abbott Laboratories, IL.
8) (a) Grabacka., M.; Plonka., P. M.; Urbanska, K.; Reiss, K.
Peroxisome Proliferator-Activated ReceptorR Activation De-
creases Metastatic Potential of Melanoma Cells in Vitro via
Down-Regulation of Akt. Clin. Cancer Res. 2006, 12, 3028–3036.
(
b) Pozzi, A.; Ibanez, M. R.; Gatica, A. E.; Yang, S.; Wei, S.; Mei, S.;
Falck, J. R.; Capdevila, J. H. Peroxisomal Proliferator-Activated
Receptor-R-Dependent Inhibition of Endothelial Cell Proliferation
and Tumorigenesis. J. Biol. Chem. 2007, 282, 17685–17695. (c)
Yokoyama, Y.; Xin, B.; Shigeto, T.; Umemoto, M.; Kasai-Sakamoto,
A.; Futagami, M.; Tsuchida, S.; Al-Mulla, F.; Mizunuma, H. Clofibric
Acid, a Peroxisome Proliferator-Activated Receptor R Ligand, Inhibits
Growth of Human Ovarian Cancer. Mol. Cancer Ther. 2007, 6, 1379–
1386. (d) Panigrahy, D.; Kaipainen, A.; Huang, S.; Butterfield,
C. E.; Barnes, C. M.; Fannon, M.; Laforme, A. M.; Chaponis, D. M.;
Folkman, J.; Kieran, M. W. PPAR.R Agonist Fenofibrate Suppresses
Tumor Growth through Direct and Indirect Angiogenesis Inhibition.
Proc. Natl. Acad. Sci. U.S.A. 2008, 105 (3), 985–990.
(15) Mukherjee, R. PPARs: Versatile Targets for Future Therapy for
Obesity, Diabetes and Cardiovascular Diseases. Drug News Per-
spect. 2002, 15, 261–267.
(16) Osumi, T.; Wen, J. K; Hashimoto, T. Two Cis-Acting Regulatory
Sequences in the Peroxisome Proliferator-Pesponsive Enhancer
Region of Rat Acyl-CoA Oxidase Gene. Biochem. Biophys. Res.
Commun. 1991, 175, 866–871.
(17) Mukherjee, R.; Sun, S; Santomenna, L.; Miao, B.; Walton, H.;
Liao, B.; Locke, K.; Zhang, J. H.; Nguyen, S. H.; Zhang, L. T.;
Murphy, K.; Ross, H. O.; Xia, M. X.; Teleha, C.; Chen, S. Y.;
Selling, B.; Wynn, R.; Burn, T.; Young, P. R. Ligand and Coacti-
vator Recruitment Preferences of Peroxisome Proliferator
Activated Receptor Alpha. J. Steroid Biochem. Mol. Biol. 2002,
81, 217–225.
(
9) (a) Sierra, M. L.; Beneton, V.; Boullay, A.-B.; Boyer, T.; Brewster,
A. G.; Donche, F.; Forest, M.-C.; Fouchet, M.-H.; Gellibert, F. J.;
Grillot, D. A.; Lambert, M. H.; Laroze, A.; Le Grumelec, C.;
Linget, J. M.; Montana, V. G.; Nguyen, V.-L.; Nicodeme, E.; Patel,
V.; Penfornis, A.; Pineau, O.; Pohin, D.; Potvain, F.; Poulain, G.;
Ruault, C. B.; Saunders, M.; Toum, J.; Xu, H. E.; Xu, R. X.;
Pianetti, P. M. Substituted 2-[(4-Aminomethyl)phenoxy]-2-
methylpropionic Acid PPARR Agonists. 1. Discovery of a Novel
Series of Potent HDLc Raising Agents. J. Med. Chem. 2007, 50,
685–695. (b) Brown, P. J.; Winegar, D. A.; Plunket, K. D.; Moore, L. B.;
Lewis, M. C.; Wilson, J. G.; Sundseth, S. S.; Koble, C. S.; Wu, Z.;
Chapman, J. M.; Lehmann, J. M.; Kliewer, S. A.; Willson, T. M. A
Ureido-thioisobutyric Acid (GW9578) Is a Subtype-Selective PPARRR
Agonist with Potent Lipid-Lowering Activity. J. Med. Chem. 1999, 42,
(18) (a) Berthou, L.; Duverger, N.; Emmanuel., F.; Langou €e t, S.;
Auwerx, J.; Guillouzo, A.; Fruchart, J. C.; Rubin, E.; Den ꢁe fle,
P.; Staels, B.; Branellec, D. Opposite Regulation of Human versus
Mouse Apolipoprotein A-I by Fibrates in Human Apolipoprotein
A-I Transgenic Mice. J. Clin. Invest. 1996, 97, 2408–2416. (b) Rubin,
E. M.; Ishida, B. Y.; Clift, S. M.; Krauss., R. M. Expression of Human
Apolipoprotein A-I in Transgenic Mice Results in Reduced Plasma
Levels of Murine Apolipoprotein A-I and the Appearance of Two
New High Density Lipoprotein Size Subclasses. Proc. Natl. Acad.
Sci. U.S.A. 1991, 88, 434–438.
3
785–3788. (c) Brown, P. J.; Stuart, L. W.; Hurley, K. P.; Lewis, M. C.;
Winegar, D. A.; Wilson, J. G.; Wilkison, W. O.; Ittoop, O. R.; Willson,
T. M. Identification of a Subtype Selective Human PPARR Agonist
through Parallel-Array Synthesis. Bioorg. Med. Chem. Lett. 2001, 11,
1
225–1227. (d) Miyachi, H.; Nomura, M.; Tanase, T.; Suzuki, M.;
Murakami, K.; Awano, K. Enantio-Dependent Binding and Transacti-
vation of Optically Active Phenylpropanoic Acid Derivatives at Human
Peroxisome Proliferator-Activated Receptor Alpha. Bioorg. Med.
Chem. Lett. 2002, 12, 333–335. (e) Xu, Y.; Mayhugh, D.; Saeed, A.;
(19) Wang, P. R.; Guo, Q.; Ippolito, M.; Wu, M.; Milot, D.; Ventre, J.;
Doebber, T.; Wright, S. D.; Chao, Y. S. High Fat Fed Hamster, a
Unique Animal Model for Treatment of Diabetic Dyslipidemia

2
864 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7
Li et al.
with Peroxisome Proliferator Activated Receptor Alpha Selective
Agonists. Eur. J. Pharmacol. 2001, 427, 285–293.
20) (a) Spady, D. K.; Dietschy, J. M. Sterol Synthesis in Vivo in 18
Tissues of the Squirrel Monkey, Guinea Pig, Rabbit, Hamster, and
Rat. J. Lipid Res. 1988, 24, 303–315. (b) Nistor, A.; Bulla, A.; Filip,
D. A.; Radu, A. The Hyperlipidemic Hamster as a Model of Experi-
mental Atherosclerosis. Atherosclerosis 1987, 68, 159–173.
PPARR and -γ Ligand Binding Domain in Complex with AZ
242; Ligand Selectivity and Agonist Activation in the PPAR
Family. Structure 2001, 9, 699–706.
(
(22) Otwinowski, Z.; Minor, W. Processing of X-Ray Diffraction Data
Collected in Oscillation Mode. In Methods in Enzymology, Macro-
molecular Crystallography, Part A; Carter, C. W., Jr., Sweet, R. M.,
Eds.; Academic Press: New York, 1997; Vol. 276, pp 307-326.
(23) MIFit Software. Consortium Version for Windows, version 7.1.rc1;
Rigaku Americas Corporation: The Woodlands, TX, 2003-2007.
(
21) Cronet, P.; Petersen, J.; Folmer, R.; Blomberg, N.; Sj o€ blom, K.;
Karlsson, U.; Lindstedt, E.-L.; Bamberg, K. Structure of the