Design, synthesis, and structural analysis of phenylpropanoic acid-type PPARγ-selective agonists: Discovery of reversed stereochemistry-activity relationship

Article abstract of DOI:10.1021/jm101233f

Peroxisome proliferator-activated receptor gamma (PPARγ) is a ligand-mediated transcription factor with roles in glucose, lipid, and lipoprotein homeostasis, and PPARγ ligands are expected have therapeutic potential in these as well as other areas. We rep

Full text of DOI:10.1021/jm101233f

J. Med. Chem. 2011, 54, 331–341 331
DOI: 10.1021/jm101233f
Design, Synthesis, and Structural Analysis of Phenylpropanoic Acid-Type PPARγ-Selective Agonists:
Discovery of Reversed Stereochemistry-Activity Relationship†
Masao Ohashi,‡ Takuji Oyama,§ Izumi Nakagome, Mayumi Satoh,^ Yoshino Nishio,‡ Hiromi Nobusada,‡
Shuichi Hirono, Kosuke Morikawa,§ Yuichi Hashimoto,# and Hiroyuki Miyachi*,‡
‡Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama University, 1-1-1, Tsushima-Naka, Kita-ku,
Okayama 700-8530, Japan, §Institute for Protein Research, Osaka University, 6-2-3, Furuedai, Suita, Osaka 565-0874, Japan,
School of Pharmaceutical Sciences, Kitasato University, 5-9-1 Shirokane, Minato-ku, Tokyo 108-8641, Japan,
^Tumor Therapy Project, The Tokyo Metropolitan Institute of Medical Science,
Tokyo Metropolitan Organization for Medical Research, 2-1-6 Kamikitazawa, Setagaya-ku, Tokyo 156-8506, Japan, and
#Institute of Molecular and Cellular Biosciences, The University of Tokyo, 1-1-1, Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan
Received September 21, 2010
Peroxisome proliferator-activated receptor gamma (PPARγ) is a ligand-mediated transcription factor
with roles in glucose, lipid, and lipoprotein homeostasis, and PPARγ ligands are expected have thera-
peutic potential in these as well as other areas. We report here the design, synthesis, crystallographic
analysis, and computational studies of R-benzylphenylpropanoic acid PPARγ agonists. Interestingly,
these compounds show a reversal of the stereochemistry-transactivation activity relationship observed
with other phenylpropanoic acid ligands.
1. Introduction
energy metabolism, such as TNF-R, leptin, and adiponectin
genes.5
The nuclear receptors (NRsa) form a superfamily of ligand-
dependent transcription factors that control diverse biological
functions, including reproduction, development, homeosta-
sis, and immune function. This superfamily includes receptors
for steroid hormones, thyroid hormones, retinoids, and vita-
min D as well as a large number of orphan receptors. On the
basis of the elucidated human genome sequence, 48 NRs are
speculated to exist in humans.1 Among them, much attention
has been focused on the peroxisome proliferator-activated
receptors (PPARs) over the past two decades, mainly because
of their roles in glucose, lipid, and lipoprotein homeostasis.
PPARs are activated by endogenous saturated and unsat-
urated fatty acids and their metabolites as well as synthetic
ligands.2 Three subtypes, PPARR, PPARδ, and PPARγ, have
been isolated to date, and each subtype appears to be differ-
entially expressed in a tissue-specific manner. The most exten-
sivelystudied subtypeis PPARγ, which is expressed inadipose
tissue, macrophages, vascular smooth muscle, and components
of the immune system.3 PPARγ is reported to be a principal
regulator of adipocyte differentiation,4 but recent molecular-
biological studies have indicated that its activation is also
linked to the expression of many important genes that affect
The range of therapeutic potential for PPAR agonists is
currently expanding well beyond lipid, lipoprotein, and glu-
cose homeostasis, and PPAR biology and pharmacology are
attracting enormous interest. For example, PPARγ activation
attenuates the expression of inducible nitric oxide (iNOS) and
cyclooxygenase-2 (COX-2) as well as the production of pro-
inflammatory cytokines.6 PPARγ was initially noted to be
highly expressed in adipose tissue, but later studies demon-
strated that PPARγ is also expressed widely in tumors orig-
inated from various organs. Ligand-mediated activation of
PPARγ inhibits cell proliferation and/or induces apoptosis or
terminal differentiation by upregulating the expression of
cyclin-dependent kinase (CDK) inhibitors, including P18,
P21, and P27.7 PPARγ also promotes cell cycle arrest by
inhibiting CDK activity in several tumor cell lines.8 Angio-
genesis, the formation of new blood vessels, is a critical step in
solid tumor growth,9 and PPARγ activation inhibits the expres-
sion of at least three important genes involved in angiogenic
processes, i.e., VEGF, VEGF receptor 1, and urokinase plas-
minogen activator (uPA).10 Therefore, PPARγ is considered
as a therapeutic target for certain human malignancies. Fur-
ther, on the basis of findings that the glitazone-class antidia-
betic agents are ligands of PPARγ,11 much research interest
has been focused on NRs as therapeutic targets for the treat-
ment of diabetes and dyslipidemia. These examples remind us
that PPARs are pleiotropic NRs, and the PPAR subtypes
appear to be potential, though somewhat overlapping, ther-
apeutic targets for the treatment of not only metabolic dis-
orders but also inflammation, cancer, neurodegeneration,
wounds, etc. The range of possible applications of PPAR
ligands has certainly not yet been fully explored.
†PDB ID codes: 3AN3, hPPARγ LBD-(S)-7; 3AN4, hPPARγ
LBD-(R)-7.
*To whom correspondence should be addressed. Phone: þ81-086-
251-7930. E-mail: miyachi@pharm.okayama-u.ac.jp
a Abbreviations: PPAR, peroxisome proliferator-activated receptor;
NR, nuclear receptor; iNOS, inducible nitric oxide; COX-2, cycloox-
ygenase-2; CDK, cyclin-dependent kinase; VEGF, vascular endothelial
growth factor; uPA, urokinase plasminogen activator; AF, activation
function; DBD, DNA-binding domain; RE, response element; LBD,
ligand-binding domain; THF, tetrahydrofurane; PDC, pyridinium di-
chromate; PDB, Protein Data Bank; JST, Japan Science and Technol-
ogy Corporation; JSPS, Japan Society for the Promotion of Science.
We have been engaged in structural development studies of
NR ligands (agonists andantagonists) for over 10years, based
r
2010 American Chemical Society
Published on Web 12/03/2010
pubs.acs.org/jmc

332 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Ohashi et al.
Figure 1. Representative PPARs agonists. 1-3: PPARγ agonists. 4: PPARR/δ dual agonist. 5: PPARR/δ dual agonist. 6: PPAR pan agonist.
Figure 2. Our design concept for transition from PPAR pan agonist to PPARγ selective agonist.
on our working hypothesis concerning the NR ligand super-
family,12 and we have already successfully designed and syn-
thesized a series of substituted phenylpropanoic acid human
PPARR-selective agonists,13 PPARR/δ dual agonist,14 PPARδ-
selective agonists,15 and PPAR R/δ/γ pan agonist.16 Here, we
focus on phenylpropanoic acid-type PPAR agonists selective
for the PPAR γ subtype.
In designing PPARγ-selective agonists, we started from our
previously reported PPAR R/δ/γpan agonist 6(Figure1), which
exhibited submicromolar EC50 values for all PPAR subtypes
(EC50s of the transactivation activity against hPPARR,hPPARδ,
and hPPARγ are 61, 120, and 43 nM, respectively) (Figure 1).
The appearance of the PPARγ activity of 6 was considered
likely to have been due to the introduction of the sterically
bulky adamantyl group at the hydrophobic tail part. There-
fore we turned our attention to the substituent at the R-position
of the carboxylic acid group. Considering the presence of the
“benzophenone pocket” of the PPARγ-LBD,17 we prepared
several compounds with bulky substituents and found that
the introduction of a benzyl group retained the PPARγ-
agonistic activity while considerably decreasing the PPARR-
and PPARδ-agonisticactivities (datanot shown) inthe case of
the racemic compound 7 (Figure 2).
using Evans’s asymmetric alkylation method18 as the key
step, as depicted in Scheme 1. 5-Formylsalicylic acid (8) was
benzyl-esterified and n-propoxylated to give compound 9. The
formyl group of 9 was reduced with NaBH4 (10) and bromi-
nation of the hydroxyl group afforded the bromomethyl deriv-
ative (11). (S)-N-3-Phenylpropionyl-4-benzyloxazolidinone
was treated with 11 according to Evans’s asymmetric alkyla-
tion protocol, followed by hydrogenolysis to afford benzoic
acid derivative 13. This was reduced with BH3-THF (14), and
then oxidation with PDC afforded the formyl derivative 14.
14 was amide-alkylated with adamantan-1-ylbenzamide in
the presence of trifluoroacetic acid and triethylsilane as the
reducing agent, followed by removal of the chiral auxiliary
to afford the desired (R)-configuration product (R)-7 with
98% ee.19 The antipodal (S) enantiomer was similarly pre-
pared, using (R)-N-3-phenylpropionyl-4-benzyloxazolidinone
as the reagent, with equivalent optical purity.
2.2. Biological Activities of (R)-7 and (S)-7. The PPARs trans-
activation activities of R-benzylphenylpropanoic acid (R)-7 and
(S)-7, together with the results for (S)-6 (PPAR pan agonist)
and rosiglitazone (PPARγ-selective agonist used clinically as an
antidiabetic agent), are summarized in Figure 3.
Neither of the enantiomers exhibited apparent transacti-
vation activity toward PPARR or PPARδ at the high con-
centration of 10 μM, while they both showed potent and
selective PPARγ transactivation activity. The EC50 values of
these compounds were similar or somewhat superior to those
of the structurally similar PPAR pan agonist (S)-6. It is of par-
ticular interest that (R)-7 is a more potent PPARγ-selective
agonist than the antipodal (S)-enantiomer (the EC50 values
of the transactivation activities of (R)-7 and (S)-7 for hPPARγ
were 3.60 and 22.0 nM, respectively, in our assay system).
In agreement with the above result, (R)-7 exhibited more
potent 3T3-L1 adipocyte differentiation activity than (S)-7.
In this paper, we present a series of optically active R-benzyl-
substituted phenylpropanoic acid-type PPARγ-selective ago-
nists. Interestingly, these compounds show a reversal of the
stereochemistry-transactivation activity relationships previ-
ously reported for other phenylpropanoic acid ligands. This
apparent discrepancy was resolved by means of X-ray crystal-
lographic analysis and computational chemistry based on the
obtained X-ray data.
2. Results and Discussion
2.1. Chemistry. Optically active R-benzylphenylpropanoic
acid PPARγ-selective agonists, (R)-7 and (S)-7, were prepared

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 333
Scheme 1. Synthetic Route to the Present Series of Compoundsa
a Reagents and conditions: (a) (1) BnBr, KHCO3, DMF, rt, overnight, quant, (2) nPrI, K2CO3, DMF, 60 ꢀC, overnight, quant; (b) NaBH4, EtOH,
rt 2 h, 91%; (c) PBr3, diethylether, 0 ꢀC, 1.5 h, 52%; (d) (R)-3-(3-phenylpropanoyl)-4-benzyloxazolidin-2-one, LiHMDS, dehydrated THF, -50 ꢀC to
0 ꢀC, 2 h, 64%; (e) H2, 10% Pd-C, AcOEt, rt, 3 h, 55%; (f) (1) BH3-THF, dehydrated THF, 0 ꢀC, overnight; (2) activated MnO2,CH2Cl2, rt, overnight,
65% (2 steps); (g) 4-(1-adamantyl)benzamide, triethylsilane, TFA, toluene, reflux, 48 h, 97%; (h) LiOH-H2O, 30% H2O2, THF:H2O = 4:1 (v/v), 0 ꢀC,
2.5 h then rt 3 h, 86%; (i) (S)-3-(3-phenylpropanoyl)-4-benzyloxazolidin-2-one, LiHMDS, dehydrated THF, -50 ꢀC to 0 ꢀC, 2 h, 70%.
As shown in Figure 4, (R)-7 exhibited maximum adipocyte
differentiation activity at 100 nM, while a much higher con-
centration of (S)-7 was needed for full activity.
These results are in good agreement with data reported for
other series of PPARγ agonists. For example, in the case of
tyrosine-based PPARγ agonists, exemplified by 2, the (S)-
enantiomer exhibited about 100-fold more potent binding
activity, transactivation activity, and lipogenesis activity than
the (R)-enantiomer.21 In the case of the R-alkoxyphenylpro-
panoic acid-based PPARγ agonist series, exemplified by 3,
the (S)-enantiomer exhibited about 20-fold more potent glucose
uptake activity as compared with the (R)-enantiomer.22 There-
fore, further studies were carried out to investigate the apparent
reversal of the stereochemistry-activity relationship in the
present compounds.
2.3. Binding of r-Alkylphenylpropanoic Acid Derivatives
(S)-5 and (S)-6 to PPARs-LBD. To understand the (S)-
conformational preference of our previous series, we had
examined the three-dimensional structures of these phenylpro-
panoic acid PPAR ligands complexed with the correspond-
ing PPAR subtype ligand-binding domains (LBDs) by means of
Previous SAR data on our phenylpropanoic acid deriva-
tives13-16 indicated that the (S)-enantiomer is more potent
than the (R)-enantiomer (see also Figure 3). For example, in
the case of PPARR/δ dual agonist 4, the (S)-enantiomer ex-
hibited about 13-fold and 37-fold more potent PPARR and
PPARδ transactivation activities, respectively, as compared
with the (R)-enantiomer. In the case of PPARδ-selective
agonist 5, the (S)-enantiomer exhibited about 12-fold more
potent activity than the (R)-enantiomer. As for the PPAR pan
agonist 6, the (S)-enantiomer is more potent than the (R)-
enantiomer for all PPAR subtypes (about 8-fold, 7-fold, and
2-fold more potent for PPARR, PPARδ, and PPARγ trans-
activation activities, respectively). Although the activity of 6
resided primarily in the (S)-enantiomer, the enantioselectiv-
ity was less marked in the case of PPARγ.

334 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Ohashi et al.
Figure 3. (A) Potency of (S)-7, (R)-7, and related compounds toward hPPARs. (B) Dose-response relationship of the hPPARγ transactiva-
tion activity. (A) Compounds were screened for agonist activity toward PPAR-GAL4 chimeric receptors in transiently transfected HEK-293
cells as described. The EC50 value is the molar concentration of the test compound that affords 50% of the maximal reporter activity. “ia”
means inactive (no apparent activity) at the concentration of 10 μM. (B) abbreviations used are as follows: feno, 10-5 M fenofibrate (PPARR-
selective agonist); GW, 10-7 M 2-(2-methyl-4-((4-methyl-2-(4-(trifluoromethyl)phenyl)thiazol-5-yl)methylthio)phenyl) acetic acid (GW-
501516)20 (PPARδ-selective agonist); Cig, 10-5 M ciglitazone (PPARγ-selective agonist); (S)-6, 10-6 M (S)-6 (PPAR pan agonist).
Parts E-G of Figure 5 present close-up views of com-
pounds (S)-5 and (S)-6 complexed with PPARR-, PPARδ-,
and PPARγ-LBD, respectively. Notably, these studies are
consistent with the enantioselectivity of the compounds. The
ethyl group in the (S)-enantiomer is located at the longest
binding site with the head carboxyl group, forming hydro-
phobic interactions with the surrounding hydrophobic resi-
dues. When the (R)-enantiomer approaches the ligand-
binding pocket, the terminal methyl group of the R-position
ethyl group may interfere sterically with the surrounding
residues, particularly with the side chains of the Cys residues
in the central part of the H3 helix (Cys276, Cys285, and Cys285
of PPARR-, PPARδ-, and PPARγ-LBD, respectively).
Figure 4. Dose-dependent induction of 3T3-L1 adipocyte differen-
2.4. Binding of (S)-7 and (R)-7 to PPARγ-LBD. The insight
tiation by (R)-7 and (S)-7. Data are expressed as mean ( SD (n=3).
obtained from the studies on the R-ethylphenyl propanoic
X-ray crystallographic analysis.23 We succeeded in determin-
ing the three-dimensional structures of PPARR LBD-(S)-6
complex, PPARδ LBD-(S)-5 complex, and PPARγ LBD-
(S)-6 complex, as depicted in Figure 5A-G.24 All three PPAR
LBDs have similar folding of a three-layered sandwich com-
prising mainly R-helices, as also observed in the nuclear recep-
tor LBDs (Figure 5A-D), and the overall interaction modes
of the proteins with the ligands are substantially conserved.
The ligand-binding domains form Y-shaped pockets. The
longest arm is placed behind helix 3, where the bound PPAR
ligands specifically contact both hydrophilic and hydropho-
bic residues.25 In particular, the hydrogen-bonding network
contains the Tyr residue on the AF-2 helix (helix 12), which
plays a key role in interacting with the carboxyl group of the
ligand. The second arm, which is hydrophilic, is located on
the opposite side to the first arm, against helix 3. The hydro-
phobic third arm lies at the entrance (H20 and β-turns) to the
binding pocket.
All three phenylpropanoic acid-type PPAR agonists show
very similar ligand-binding modes. The acidic head carboxyl
groups are located in the first cavity, making conventional
interactions with the polar side chains, and thus all three
PPAR LBDs adopt the conventional fully active conforma-
tion of helix 12 (Figure 5A-C). The central benzene rings and
the alkoxy groups are located in the center and the second
cavity, respectively, and the hydrophobic tail part lies at the
entrance of the LBDs.
acid series 5 and 6 could not explain the discrepancy arising
from the incorporation of a benzyl substituent at the R-position
of the phenylpropanoic acid. Therefore, we conducted an
X-ray crystallographic analysis of (R)-7 and (S)-7 com-
plexed with the PPARγLBD. PPARγLBD-(R)-7andPPARγ
LBD-(S)-7 complexes were cocrystallized under the same con-
ditions, as described in the Experimental Section. The crys-
tallographic data and refinement statistics are summarized in
Table 1.
2.5. Structural Basis for the Full-Agonistic Activity of r-
Benzylphenylpropanoic Acid. Transcriptional activation at a
nuclear receptor is a complex process, elicited initially by the
binding of a ligand. After ligand binding, conformational
change of the nuclear receptor occurs, which facilitates disso-
ciation of the corepressor complex and subsequent recruit-
ment of the coactivator complex.26 The qualitative nature of
the corepressor dissociation and coactivator recruitment was
reported to be regulated differentially by the three-dimensional
structure around the C-terminal H12 helix and the surround-
ing helixes of the nuclear receptor.27 In the case of the bind-
ing of a full agonist, such as rosiglitazone, the H12 helix
is stabilized by tight hydrogen bonding to Tyr473, and the
H12 helix is docked properly against the H3 helix and H11
helix.26 In this conformation, H12 forms part of the coacti-
vator-binding surface along with the H3 helix and H5 helix.
The positions of two distinct amino acids, Lys301 (in H3 helix)
and Glu471 (in H12 helix), are especially important because

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 335
Figure 5. (A-C) Crystal structures of PPARs LBD-TIPP complexes. (A) PPARR LBD-(S)-6 complex; (B) PPARδ LBD-(S)-5 complex;
(C) PPARγ LBD-(S)-6 complex. Proteins are represented as ribbon models and the ligands are depicted as space-filling models, with F, C, N,
and O atoms in aqua, gray, blue, and red, respectively. (D) Superposition of the main chains of each PPARs LBD. The numbering of the second
structure is also depicted. (E-G) Zoomed view of the ligand-binding mode of PPARs LBD-TIPP complexes. (E) PPARR LBD-(S)-6
complex; (F) PPARδ LBD-(S)-5 complex; (G) PPARγ LBD-(S)-6 complex. Proteins are represented as ribbon models and the ligands are
depicted as cylinder models, with F, C, N, and O atoms in aqua, gray, blue, and red, respectively.
Table 1. Crystallographic Data and Refinement Statistics (Values in
Parentheses Are for the Last Shell)
these amino acids clamp the LXXLL motif involved in the
coactivators.
PPARγ LBD-(S)-7
PPARγ LBD-(R)-7
C2
The framework of hPPARγ LBD complexed with (S)-7
or (R)-7 is depicted in Figure 6B,C, together with the struc-
ture of rosiglitazone-hPPARγ LBD complex (PDB: 2PRG,
Figure 6A). Figure 6D shows a superposition of these three
complexes (side chains of amino acids except those of Lys301
and Glu471 are omitted). The overall foldings of hPPARγ
LBD complexed with (S)-7 or (R)-7 and rosiglitazone are
well matched, except for the thermodynamically unstable
region from the end of the H2 helix to the beginning of the H3
helix. The positions of the important amino acids for coacti-
vator binding, Lys301 in the H3 helix and Glu471 in the H12
helix, overlap well in these structures. These structural-
biological findings support the functional studies depicted
in Figure 3B, indicating that both (S)-7 and (R)-7 function
as full agonists.
Data Collection
space group
unit cell constants (A)
C2
˚
a
93.192
61.061
93.391
61.271
b
c
118.618
102.82
118.871
102.82102
1
β (deg)
wavelength (A)
˚
1
a
˚
resolution (A)
50.0-2.30
(2.38-2.30)
28907(2841)
99.3 (98.7)
32.7(6.6)
3.7 (3.3)
3.0 (24.0)
50.0-2.30
(2.38-2.30)
28731(2825)
97.8 (97.5)
23.6(7.0)
3.7 (3.4)
3.5 (23.1)
no. of unique reflections
completeness (%)
I/σ(I)
redundancy
merge (%)b
R
2.6. Structural Basis for the PPARγ Selectivity and Poten-
cy of r-Benzylphenylpropanoic Acid. As mentioned above,
R-ethylphenylpropanoic acid derivative 6 exhibited trans-
activation activities toward all PPAR subtypes, while R-
benzylphenylpropanoic acid derivative 7 specifically transacti-
vated PPARγ, being a more potent activator of PPARγ than
was 6. It appears that these differences in selectivity and
activity can be mainly attributed to the difference of the inter-
action potential of the side chain at the R-position of phenyl-
propanoic acid with the surrounding amino acids. Parts
B and C of Figure 6 show superposition of the amino acids
of hPPARγ interacting with the benzyl group of (S)-7 (R-
substituent binding pocket) and the corresponding amino
acids of hPPARδ and hPPARR (Figure 6B; hPPARδ vs
hPPARR, Figure 6C; hPPARγ vs hPPARR). The volume of
the R-substituent-binding pocket of hPPARδ is smaller than
that of hPPARRdue to the bulky Met residue (Val in hPPARR)
located at the bottom of the pocket. The bulky benzyl side
Refinement
50.0-2.30
22.7/27.8
˚
resolution range (A)
50.0-2.30
23.2/28.8
d
Rworkc/Rfree
no. of atoms
protein
water
4155
132
42
4150
83
ligand
average B-factor (A)
42
˚
protein
water
47.54
45.26
69.78
45.19
40.31
62.55
ligand
rmsd
bond lengths (A)
˚
0.007
1.1
0.007
1.1
angles (deg)
PDB code
3AN3
3AN4
P
a Values in parentheses correspond to the last shell. b Rmerge=( |II
P
ÆIIæ|)/ I |II| where ÆIIæ is the mean II over symmetry-equivalent reflec-
P
P
tions. c Rwork
calculate Rfree
=
.
|F - F |/ |F | for all data excluding data to used to
O
C
O
P
P
d Rfree
=
|F - F |/ |F | for all data
O
C
O

336 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Ohashi et al.
Figure 6. (A-C) Overall crystal structures of PPARγ LBD-ligand complexes. (A) PPARγ LBD-rosiglitazone complex (PDB 2PRG);
(B) PPARγ LBD-(S)-7 complex; (C) PPARγ LBD-(R)-7 complex. Main chains of the proteins are represented, and each of the two amino
acids Lys301 and Glu473 are highlighted as space-filling models, with green (rosiglitazone complex), magenta ((S)-7 complex), and cyan ((R)-7
complex). (D) Superposition of the main chains of each PPARγ LBD. (E-F) Zoomed views of the superposed amino acid residues forming the
benzyl group-binding pockets. (E) Superposition of the PPARδ LBD (cyan) and PPARR LBD (orange). (F) Superposition of the PPARγ LBD
(red) and PPARR LBD (orange). Protein backbones are omitted, and the side-chain amino acids are represented as cylinder models. (G-H)
Zoomed views of the amino acids involved in the interaction of the side chain of the ligands. (E) Zoomed view of PPARγ LBD-(S)-7 complex.
(F) Zoomed view of PPARγ LBD-(S)-6 complex. Proteins are represented as wireframe models and the amino acids interacting with the side-
chain ethyl group of (S)-6 are depicted as yellow cylinders, and the additional amino acid side chains interacted are depicted as green cylinders.
chain of (S)-7 might not fit into the pocket of hPPARδ owing
to steric interference with the side chain of the Met residue. It
could also be considered that the width of the entrance of
hPPARR is smaller than that of hPPARδ due to the bulky
Tyr residue (His in hPPARδ) located at the entrance of the
pocket. The bulky benzyl side chain of (S)-7 might not be
able to enter the pocket of hPPARR due to steric interference
with the side chain of the Tyr residue.
These results clearly explain why the activity toward hPPARγ
is improved by replacing the side chain at the R-position of the
phenylpropanoic acids. In the case of (S)-6, the ethyl group
interacts with the hydrophobic cavity composed of Phe282,
Cys285, Ser289, Tyr327, and His449 (Tyr327 and His449 also
contribute to the hydrogen-bonding network of the acidic
carboxyl group of (S)-6) (Figure 6E). On the other hand, in
the case of (S)-7, the benzyl group is wider than the ethyl
group, and therefore the benzyl group interacts further with
the additional hydrophobic cavity composed of Gln286,
His323, Phe363, Leu453, and Tyr473. This expansion of the
interaction area might be the main reason for the increase in
the activity of (S)-7 versus (S)-6 (Figure 6D).
2.7. Comparison of the Structures of hPPARγ LBD Com-
plexed with (S)-7 and (R)-7. The overall structures of hPPARγ
LBD complexed with (S)-7 and (R)-7 are shown in parts

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 337
Figure 7. (A,C) Overall crystal structures of hPPARγ LBD-ligand complexes. (A) hPPARγ LBD-(S)-7 complex; (C) hPPARγ LBD-(R)-7
complex. (B) Superposition of the crystal structures of hPPARγ LBD-(S)-7 complex and hPPARγ LBD-(R)-7 complex. Proteins are
represented as wireframe models, and (S)-7 and (R)-7 are depicted as cylinder models in magenta ((S)-7) and cyan ((R)-7 complex).
(D) Superposition of the bound structures of (S)-7 and (R)-7. (E) Zoomed superposition of the conformations of the benzyl group in (S)-7 and
(R)-7. (F,I) Zoomed views of the amino acids of hPPARγ LBD involved in the interaction of the side-chain benzyl group of the ligands.
(F) hPPARγ LBD-(S)-7 complex; (I) hPPARγ LBD-(R)-7 complex. The amino acids and the side-chain benzyl groups are depicted as space-
filling models in orange (interacting amino acids), magenta (benzyl group of (S)-7), and cyan (benzyl group of (R)-7). The flipped Phe363s are
highlighted in green. (G) Superposition of the amino acids of hPPARγ LBDs interacting with both (S)-7 and (R)-7. The main chains are omitted
and the side chains are depicted as cylinder models in magenta ((S)-7 complex) and cyan ((R)-7 complex).
A((S)-7complex) and C ((R)-7complex) of Figure 7. Figure 7B
presents a superposition of the two complexes. As can be seen
from Figure 7D, which is a close-up view of the superposed
ligands, the three-dimensional structures of bound (S)-7 and
(R)-7 closely resemble each other. The only difference is in the
direction of the methylene chain and the configuration of the
benzene ring of the R-substituent benzyl group (Figure 7D). It
is very surprising that, although the stereochemistry is oppo-
site between (S)-7 and (R)-7, the benzyl side chains of both
ligands are located in the same hydrophobic pocket of hPPARγ
LBD formed by the H3, H5, H7, H11, and H12 helixes. This
may be the main reason why the (R)-enantiomer of7 exhibited
potent hPPARγ transactivation activity, even though the (R)-
enantiomer is generally thought to be unfavorable, based on
the results obtained with compound 6. It should also be noted
that the conformation of the side-chain benzyl group of
Phe363 is flipped between hPPARγ LBD complexed with (S)-7
and with (R)-7 (Figure 7F-I). Compared to the (S)-6-hPPARγ
LBD complex, the Phe363 of hPPARγ LBD complexed with
(S)-7 exhibits a flipped conformation. Although it is not clear
why the conformation of Phe363 is different between the two com-
plex structures, the results of the computational study described
below indicated that this conformational change apparently
did not contribute to the present stereochemical discrepancy.
2.8. Detailed Analysis of the Interactions of the Side-Chain
Benzyl Group of (R)-7 and (S)-7 Complexed with hPPARγ.
We focused on Ser289 and Leu469 in the complex of hPPARγ
LBD with (S)-7 and (R)-7 (Figure 8A-D), for Ser289 is the
closest amino acid to the methylene chain and the Leu469 is
the closest amino acid to the distal benzene ring of the side-
chain benzyl group of both (S)-7 and (R)-7. As described
above, the binding modes of the side-chain benzyl group of

338 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Ohashi et al.
Figure 8. (A) The binding mode of (S)-7 to the ligand-binding pocket of hPPARγ LBD. (B) The binding mode of (R)-7 to the ligand-binding
pocket of hPPARγ LBD. Proteins are represented as ribbon models, and the ligands are depicted as cylinder models in magenta ((S)-7) and
cyan ((R)-7). (C) Zoomed views of the benzyl-binding pocket of hPPARγ LBD-(S)-7 complex. (D) Zoomed views of the benzyl-binding
pocket of hPPARγ LBD-(R)-7 complex. The four interacting amino acids are depicted as yellow cylinders, and the benzyl side chains are
depicted in magenta ((S)-7) and cyan ((R)-7 complex). The italic numbers indicate the distance between the indicated atoms (dotted lines).
(E) Superposition of the bound structure of (S)-7 and the energy-minimized structure of (S)-7. (F) Superposition of the bound structure of (R)-7
and the energy-minimized structure of (R)-7. The calculated ligand energies of the bound ligands are described below. The yellow arrows
indicate the directions of distortion of the benzyl groups.
(S)-7 and (R)-7 are similar, but there are significant differ-
ences in the detailed protein-ligand interactions. A short
contact between the 2-position hydrogen atom of the benzyl
group of (S)-7 and the oxygen atom of the side chain of
Ser289 was observed, together with a short contact between
the 3-position hydrogen atom of the benzyl group of (S)-7
and the hydrogen atom of the side chain of Leu469, with
and somewhat distorted in the complex as compared to (R)-7.
In addition, the conformation and the direction of the side-
chain benzyl group of (S)-7 and (R)-7 are different, even
though these groups interact with the same binding pocket,
and some short contacts can be seen. We suggest that these
two factors contribute to the reversal of the stereochemistry-
activity relationship in the case of the benzyl derivative.
˚
distances of 1.4 and 1.5 A, respectively. On the other hand,
3. Conclusion
such short contacts were not observed in the case of (R)-7.
This may be one reason why (S)-7 is a weaker hPPARγ
agonist than (R)-7.
We have designed and synthesized optically active R-
benzylphenylpropanoic acid-type hPPARγ-selective agonists
in which the key feature is the introduction of a bulky benzyl
group at the R-position of the carboxyl group. Interestingly,
these compounds showed reversal of the stereochemistry-
activity relationship found with other structurally similar
R-alkylphenylpropanoic acid PPAR agonists, i.e., (R)-7is more
more potent and hPPARγ-selective than the (S)- enantiomer .
The three-point attachment model is generally accepted to
explain how receptors and/or enzymes distinguish between
enantiomers,28,29 i.e., three groups arranged around a chiral
tetrahedral carbon atom interact with three locations in the
active site. Therefore, medicinal chemists sometimes tend to
think that a consistent tendency will be observed in a series of
compounds with the same basic framework. However, the
present results show that this is not necessarily the case. So,
double-checking by making several analogues in both enan-
tiomeric series is important.
The ligand energy calculations of both (S)-7 and (R)-7 also
indicated instability of the complex of (S)-7 with hPPARγ
LBD. The calculations indicated that the ligand energy of
(S)-7 is 142.6 kcal/mol, while that of (R)-7 is 118.3 kcal/mol.
Thus, (S)-7 is energetically unstable by 24.3 kcal/mol com-
pared with (R)-7. As already mentioned, the three-dimensional
structures of the bound (S)-7 and (R)-7 are closely similar,
and the only differences are the direction of the methylene
chain and the configuration of the benzene ring of the R-
substituent benzyl group. Therefore, these ligand energy dif-
ferences might reflect distortion of the side-chain benzyl
group of (S)-7 and (R)-7 when bound to hPPARγ LBD. To
examine this possibility, we compared the differences of the
dihedral angles of the bound benzyl group of these ligands
from those of the corresponding energy-minimized structures
(Figure 8E,F). The difference of the dihedral angle of the
more potent enantiomer, (R)-7, is about 10ꢀ. This indicates
that the bound structure of (R)-7 closely resembles the
energy-minimized structure. On the contrary, in the case of
(S)-7, a large difference of the dihedral angle, 131ꢀ, is observed.
This indicated that the benzyl group is substantially distorted
when (S)-7 is bound to hPPARγ LBD, and short contacts to
Ser289 and Leu469 remain despite the distortion of the
benzyl group conformation.
Our analysis also demonstrates that X-ray crystallographic
analysis combined with computational chemistry is a power-
ful tool to investigate three-dimensional structure-activity
relationships in detail.
4. Experimental Section
4.1. Chemistry. 4.1.1. General Methods. Melting points were
determined with a Yanagimoto hot-stage melting point appa-
ratus and are uncorrected. NMR spectra were recorded on a
VarianVXR-300 (1H 300 MHz) spectrometer. Proton chemical
X-ray crystallographic study in combination with compu-
tational study indicated that (S)-7 is energetically less stable

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 339
shifts were referenced to the TMS internal standard. Elemental
analysis was carried out with a Yanagimoto MT-5 CHN recorder
elemental analyzer and results were within 0.4% of the theore-
tical values. FAB-MS was carried out with a VG70-SE. Chemi-
cal purity was determined by HPLC analysis with a Pegasil ODS
SP100 column (4.6 mm ꢀ150 mm; flow rate of 1 mL/min; sol-
vent, MeCN:0.1% TFA=3:1 v/v; detection at 254 nm), con-
firming >98% purity.
aqueous solution of ammonium chloride was added to the reac-
tion mixture, and the whole was extracted with ethyl acetate. The
extract was washed with water and brine, dried over anhydrous
sodium sulfate, and concentrated. The residue was purified by
silica gel chromatography (eluant; n-hexane:ethyl acetate=9: 2
v/v) to afford 0.87 g (64%) of the desired compound as a colorless
1
oil. H NMR (400 MHz, CDCl3) δ 7.70 (d, J = 2.4 Hz, 1H),
7.44-7.16 (m, 14H), 6.98 (dd, J=7.6, 1.6 Hz, 2H), 6.87 (d, J=
8.4 Hz, 1H), 5.29 (s, 2H), 4.60-4.52 (m, 1H), 4.42-4.36 (m, 1H),
3.97-3.88 (m, 3H), 3.78 (t, J=8.4 Hz, 1H), 3.09 (dd, J=13.6,
8.8 Hz, 1H), 3.03-2.96 (m, 2H), 2.85-2.76 (m, 2H), 2.44 (dd, J=
13.6, 9.6 Hz, 1H), 1.81-1.72 (m, 2H), 0.96 (t, J=7.4 Hz, 3H). MS
(FAB) m/z 592 (M þ Hþ).
4.1.2. Benzyl 2-Propoxy-5-formylbenzoate (9). A mixture of
5-formylsalicylic acid (8) (2.50 g, 15.0 mmol), benzyl bromide
(1.79 mL, 15.1 mmol), potassium hydrogencarbonate (1.51 g,
15.1 mmol), and 40 mL of N,N-dimethylformamide was stirred
at room temperature for 24 h. The mixture was poured into
ice-water, and stirring was continued for 1 h. The precipitate
was collected by filtration and redissolved in 100 mL of AcOEt.
This solution was dried over anhyd MgSO4, filtered, and con-
centrated to obtain 3.54 g of benzyl 5-formylsalicylate. A mix-
ture of this product (2.00 g, 7.80 mmol), iodopropane (1.14 mL,
11.7 mmol), potassium carbonate (1.62 g, 11.7 mmol), and
30 mL of N,N-dimethylformamide was stirred at 60 ꢀC for 24 h.
The reaction mixture was poured into water, and the whole was
extracted with ethyl acetate. The organic solution was washed
with water and brine, dried over anhydrous magnesium sulfate,
and concentrated. The residue was purified by silica gel column
chromatography (eluant; n-hexane:ethyl acetate=3:1 v/v) to
afford 2.53 g (quant.) of the title compound as a colorless
powder. 1H NMR (400 MHz, CDCl3) δ 9.90 (s, 1H), 8.33 (d, J=
2.0 Hz, 1H), 7.99 (dd, J=8.6, 2.2 Hz, 1H), 7.46 (dd, J=8.0,
1.2 Hz, 2H), 7.41-7.31 (m, 3H), 7.07 (d, J=8.8 Hz, 1H), 5.37
(s, 2H), 4.08 (t, J=6.4 Hz, 2H), 1.89-1.80 (m, 2H), 1.01 (t, J=
7.4 Hz, 3H). MS (FAB) m/z 299 (M þ Hþ).
4.1.6. 5-((R)-2-Benzyl-3-((S)-4-benzyl-2-oxo-oxazolidin-3-yl)-
3-oxopropyl)-2-propoxybenzoic Acid (13). Compound 12 (0.89 g,
1.50 mmol), 200 mg of 10% palladium on carbon, and 80 mL of
ethyl acetate were mixed and catalytic hydrogenation was carried
out at an initial hydrogen pressure of 98 kPa. After completion
of the reaction, the catalyst was removed by filtration and the
filtrate was washed with ethyl acetate. The reaction mixture and
the washings were combined and concentrated. The residue was
purified by silica gel chromatography (eluant; n-hexane:ethyl
acetate=5:2 v/v) to afford 0.41 g (55%) of the desired com-
1
pound as a colorless oil. H NMR (400 MHz, CDCl3) δ 10.98
(s, 1H), 8.06 (d, J=2.4 Hz, 1H), 7.52 (dd, J=8.4, 2.4 Hz, 1H),
7.24-7.22 (m, 5H), 7.20-7.16 (m, 3H), 7.03 (dd, J=7.2, 2.0 Hz,
2H), 6.97 (d, J=8.4 Hz, 1H), 4.56-4.48 (m, 1H), 4.46-4.40
(m, 1H), 4.22-4.13 (m, 2H), 3.97 (dd, J=8.8, 2.4 Hz, 1H), 3.81
(t, J=8.2 Hz, 1H), 3.16 (dd, J=13.8. 8.2 Hz, 1H), 3.08-2.98
(m, 2H), 2.84-2.76 (m, 2H), 2.54 (dd, J=13.4, 9.4 Hz, 1H),
1.96-1.88 (m, 2H), 1.08 (t, J=7.2 Hz, 3H). MS (FAB) m/z 502
(M þ Hþ).
4.1.3. Benzyl 5-Hydroxymethyl-2-propoxybenzoate (10). To a
solution of 9 (2.53 g, 8.48 mmol), and 30 mL of ethanol was added
NaBH4 (160 mg, 4.23 mmol) portionwise at 0 ꢀC, and the mix-
ture was stirred for 2 h at room temperature. The excess ethanol
was evaporated, the residue was poured into water, and the
whole was extracted with ethyl acetate. The extract was washed
with water and brine, dried over anhydrous magnesium sulfate,
and concentrated. The residue was purified by silica gel column
chromatography (eluant; n-hexane:ethyl acetate = 3:1 v/v) to
afford 2.33 g (91%) of the title compound as a colorless oil. 1H
NMR (400 MHz, CDCl3) δ 7.79 (d, J=2.4 Hz, 1H), 7.46-7.44
(m, 3H), 7.39-7.31 (m, 3H), 6.94 (d, J=8.4 Hz, 1H), 5.37 (s, 2H),
4.62 (s, 2H), 3.98 (t, J=6.4 Hz, 2H), 1.85-1.76 (m, 2H), 0.99
(t, J=7.4 Hz, 3H). MS (FAB) m/z 301 (M þ Hþ).
4.1.7. 5-((S)-2-((R)-4-Benzyl-2-oxo-oxazolidine-3-carbonyl)-
butyl)2-n-butoxybenzylaldehyde (14). To a solution of 13 (1.37 g,
3.02 mmol) and 25 mL of dehydrated tetrahydrofuran was added
dropwise a 1 mol/L solution of borane-tetrahydrofuran complex
(5.00 mL, 5.00 mmol). After completion of the addition, the mix-
ture was stirred overnight at room temperature. A saturated aqu-
eous solution of ammonium chloride was added to the reaction
mixture, and the whole was extracted with ethyl acetate. The
extract was washed with water and brine, dried over anhydrous
sodium sulfate, and concentrated. The residue was purified by
silica gel chromatography (eluant; n-hexane:ethyl acetate=
4:1 v/v) to afford 1.18 g (89%) of the intermediate hydroxy-
methyl derivative as a colorless oil. A solution of the hydroxy-
methyl derivative (1.18 g, 2.68 mmol), activated MnO2 (0.50 g,
5.75 mmol), and 30 mL of dehydrated dichloromethane was
stirred overnight at room temperature. The catalyst was col-
lected by filtration, washed with dichloromethane, and concen-
trated. The residue was purified by silica gel chromatography
(eluant; n-hexane:ethyl acetate=4:1 v/v) to afford 855 mg (73%)
of the desired compound as a colorless oil. 1H NMR (500 MHz,
CDCl3) δ 8.05 (d, J=2.3 Hz, 1H), 7.52 (dd, J=8.5, 2.3 Hz, 1H),
7.29-7.24 (m, 3H), 7.08 (d, J=6.8 Hz, 2H), 6.97 (d, J=8.5 Hz,
1H), 4.68 (m, 1H), 4.24-4.01 (m, 5H), 3.14 (dd, J=13.2, 3.4 Hz,
1H), 3.08 (dd, J=13.7, 7.7 Hz, 1H), 2.77 (dd, J=13.7, 6.8 Hz,
1H), 2.58 (dd, J=13.2, 9.8 Hz, 1H), 1.87 (m, 2H), 1.77 (m, 1H),
1.51 (m, 3H), 0.99 (t, J=7.3 Hz, 3H), 0.93 (t, J=7.3 Hz, 3H). MS
(FAB) 454 (M þ H)þ.
4.1.4. Benzyl 5-Bromomethyl-2-propoxybenzoate (11). A mix-
ture of 10 (2.33 g, 7.76 mmol), phosphorus tribromide (0.74 mL,
7.79 mmol), and 30 mL of dehydrated ether was stirred for 1.5 h
at 0 ꢀC. The reaction mixture was poured into water, and the
whole was extracted with ether. The extract was washed with
water and brine, dried over anhydrous magnesium sulfate, and
concentrated. The residue was purified by silica gel column
chromatography (eluant; n-hexane:ethyl acetate = 3:1 v/v) to
afford 1.43 g (52%) of the title compound as a colorless oil. 1H
NMR (400 MHz, CDCl3) δ 7.84 (d, J=2.4 Hz, 1H), 7.49-7.44
(m, 3H), 7.40-7.31 (m, 3H), 6.92 (d, J=8.4 Hz, 1H), 5.35 (s, 2H),
4.47 (s, 2H), 3.98 (t, J=6.4 Hz, 2H), 1.84-1.76 (m, 2H), 0.98
(t, J=7.4 Hz, 3H).
4.1.5. Benzyl 5-((R)-2-Benzyl-3-((S)-4-benzyl-2-oxo-oxazolidin-
3-yl)-3-oxopropyl)-2-propoxy-benzoate (12). (R)-3-(3-Phenylpro-
panoyl)-4-benzyloxazolidin-2-one (0.74 g, 2.33 mmol) and 20 mL
of dehydrated tetrahydrofuran were mixed under an argon atmo-
sphere and cooled to-50 ꢀC. Under stirring, a 1 mol/L solution of
sodium bis(trimethylsilyl)amide in dehydrated tetrahydrofuran
(8.00 mL, 8.00 mmol) was added dropwise. After completion of
the addition, the mixture was stirred for 1 h at -15 ꢀC and re-
cooled to -50 ꢀC, and then a solution of 11 (0.86 g, 2.37 mmol) in
dehydrated tetrahydrofuran (20 mL) was added dropwise. After
completion of the addition, the mixture was further stirred for 1 h
while being gradually heated to room temperature. A saturated
4.1.8. N-((5-((R)-2-Benzyl-3-((S)-4-benzyl-2-oxo-oxazolidin-
3-yl)-3-oxopropyl)-2-propoxy-phenyl)methyl)-4-(1-adamantyl)-
benzamide (15). A mixture of 14 (200 mg, 0.41 mmol), 4-(1-
adamantyl)benzamide (270 mg, 1.03 mmol), triethylsilane (0.16 mL,
1.03 mmol), trifluoroacetic acid (0.08 mL, 1.03 mmol), and 30 mL
of dehydrated toluene was refluxed for 2 days. The mixture was
evaporated, and the residue was purified by silica gel column
chromatography (eluant; n-hexane:ethyl acetate=2:1 v/v) to afford
290 mg (97%) of the title compound as a colorless oil. 1H
NMR (300 MHz, CDCl3) δ 7.61 (d, J=8.7 Hz, 2H), 7.35 (d,
J=8.7 Hz, 2H), 7.28-7.16 (m, 10H), 6.88-6.85 (m, 2H), 6.80

340 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Ohashi et al.
(d, J=8.4 Hz, 1H), 6.71 (t, J=5.7 Hz, 1H), 4.70-4.52 (m, 3H),
4.42-4.34 (m, 1H), 3.98-3.88 (m, 3H), 3.74 (t, J= 8.4 Hz,
1H), 3.07 (dd, J=13.8, 8.7 Hz, 1H), 3.01-2.84 (m, 2H), 2.79
(dd, J=13.8, 6.3 Hz, 2H), 2.48 (dd, J=13.2, 9.0 Hz, 1H), 2.10
(s, 3H), 1.89-1.72 (m, 14H), 1.03 (t, J = 7.5 Hz, 3H). MS
(FAB) m/z 725 (M þ Hþ).
nitrogen stream after they had been briefly soaked in cryopro-
tection buffer (1.1 mol/L sodium citrate, 100 mmol/L HEPES,
pH 7.5, and 25% (v/v) glycerol). Diffraction data were collected
at BL38B1 in SPring-8 (Harima, Japan). All data were processed
using HKL2000.31 All structures were solved by the molecular
replacement method, using the previously published structure
of the 6-hPPARγ LBD complex (PDB 2ZNO)23 as a probe
after removing the ligand part. The correctly positioned mole-
cules were refined with CNS32 and O.33 The crystallographic
data and refinement statistics are summarized in Table 1. The
coordinates of hPPARγ LBD-(S)-7 and hPPARγ LBD-(R)-7
complexes have been deposited in the Brookhaven Pro-
tein Data Bank (PDB) with the codes of 3AN3 and 3AN4,
respectively.
4.1.9. (R)-2-Benzyl-3-(3-(((4-adamantan-1-yl)benzamido)methyl)-
4-propoxyphenyl)propanoic Acid ((R)-7). Compound 15 (290 mg,
0.400 mol) was dissolved in 24 mL of tetrahydrofuran and 6 mL
of water under an argon atmosphere with ice-cooling. To this
solution was added 30% aqueous hydrogen peroxide (0.4 mL,
4.00 mmol). Then a solution of lithium hydroxide monohydrate
(100 mg, 2.38 mmol) in water (1 mL) was added, and the mixture
was stirred further for 2.5 h at 0 ꢀC, and for 3 h at room tem-
perature. An aqueous solution of sodium hydrogen sulfite (1.00
g/6 mL) was added to the mixture, and the whole was stirred for
30 min. The reaction mixture was acidified with diluted HCl and
then extracted with ethyl acetate. The extract was washed with
brine, dried over anhydrous magnesium sulfate, and concentrated.
The residue was purified by silica gel column chromatography
(eluant; n-hexane:ethyl acetate=2:1 v/v) to afford 190 mg (86%)
of the title compound as a colorless powder; mp 177-178 ꢀC. 1H
NMR (300 MHz, CDCl3) δ 7.68 (d, J=8.1 Hz, 2H), 7.39 (d, J=
8.4 Hz, 2H), 7.26-7.14 (m, 6H), 7.03 (dd, J=8.1, 1.8 Hz, 1H),
6.77-6.70 (m, 2H), 4.64-4.51 (m, 2H), 3.94 (t, J=6.5 Hz, 2H),
3.01-2.71 (m, 5H), 2.10 (s, 3H), 1.90-1.72 (m, 14H), 1.06 (t, J=
7.4 Hz, 3H). HRMS (FAB, MHþ) calcd for C36H24Cl3N2O5
566.3270, found 566.3282. [R]D -3ꢀ (c 0.10, CH3CN). Anal.
(C37H44NO4) C, H, N.
4.4. Computational Study to Calculate Strain Energies of the
(R)-7 and (S)-7. Each binding conformation of (R)-7 and (S)-7
was extracted from each X-ray complex structure. The strain
energies of the binding conformations of (R)-7 and (S)-7 were
calculated using the Tripos force field in SYBYL7.3.
Acknowledgment. This work was supported in part by the
Targeted Proteins Research Programof the JapanScience and
Technology Corporation (JST) and by the Uehara Memorial
Foundation, and the Teromo Life Science Foundation. We
are grateful toDrs. SeikiBaba andNobuhiro Mizuno for their
kind help in the X-ray diffraction data collection on BL38B1
at SPring-8. We thank Dr. Tsuyoshi Waku and Rinna Nakamori
for their help in crystallization. T.O. and K.M. were supported
by a Grant-in-Aid for Creative Scientific Research Program
(18GS0316) from the Japan Society for the Promotion of
Science (JSPS).
4.1.10. (S)-2-Benzyl-3-(3-(((4-adamantan-1-yl)benzamido)-
methyl)-4-propoxyphenyl)propanoic Acid ((S)-7). This compound
was prepared by means of a procedure similar to that used for (R)-
7; mp 177-178 ꢀC. 1H NMR (300 MHz, CDCl3) δ 7.68 (d, J=
8.1 Hz, 2H), 7.39 (d, J=8.1 Hz, 2H), 7.26-7.14 (m, 6H), 7.03 (dd,
J=8.1, 1.8 Hz, 1H), 6.77-6.70 (m, 2H), 4.64-4.51 (m, 2H), 3.94
(t, J=6.5 Hz, 2H), 3.01-2.71 (m, 5H), 2.10 (s, 3H), 1.90-1.72
(m, 14H), 1.06 (t, J=7.4 Hz, 3H). HRMS (FAB, MHþ) calcd for
Supporting Information Available: Combustion analysis data
for (R)-7 and (S)-7. This material is available free of charge via
the Internet at http://pubs.acs.org.
C
36H24Cl3N2O5 566.3270, found 566.3265. [R]D þ3ꢀ (c 0.10,
References
CH3CN). Anal. (C37H44NO4 1/4H2O) C, H, N.
3
(1) Chawla, A.; Repa, J. J.; Evans, R. M.; Mangelsdorf, D. J. Nuclear
receptors and lipid physiology: opening the X-files. Science 2001,
294, 1866–1870.
(2) Banner, C. D.; Gottlicher, M.; Widmark, E.; Sjovall, J.; Rafter,
J. J.; Gustafsson, J. A. Systematic analytical chemistry/cell assay
approach to isolate activators of orphan nuclear receptors from
biological extracts: characterization of peroxisome proliferator-
activated receptor activators in plasma. J. Lipid Res. 1993, 34,
1583–1591.
(3) Braissant, O.; Foufelle, F.; Scotto, C.; Dauca, M.; Wahli, W.
Differential expression of peroxisome proliferator-activated receptors
(PPARs): tissue distribution of PPAR-alpha, -beta, and -gamma in
the adult rat. Endocrinology. 1996, 137, 354–366.
(4) Spiegelman, B. M.; Hu, E.; Kim, J. B.; Brun, R. PPAR gamma and
the control of adipogenesis. Biochimie 1997, 79, 111–112.
(5) Okuno, A.; Tamemoto, H.; Tobe, K.; Ueki, K.; Mori, Y.; Iwamoto,
K.; Umesono, K.; Akanuma, Y.; Fujiwara, T.; Horikoshi, H.; Yazaki,
Y.; Kadowaki, T. Troglitazone increases the number of small adipo-
cytes without the change of white adipose tissue mass in obese Zucker
rats. J. Clin. Invest. 1998, 101, 1354–1361.
(6) Mendez, M.; LaPointe, M. C. PPARgamma inhibition of cyclooxy-
genase-2, PGE2 synthase, and inducible nitric oxide synthase in
cardiac myocytes. Hypertension 2003, 42, 844–850.
(7) Morrison, R. F.; Farmer, S. R. Role of PPARgamma in regulat-
ing a cascade expression of cyclin-dependent kinase inhibitors,
p18(INK4c) and p21(Waf1/Cip1), during adipogenesis. J. Biol.
Chem. 1999, 274, 17088–17097.
(8) Huang, J. W.; Shiau, C. W.; Yang, J.; Wang, D. S.; Chiu, H. C.;
Chen, C. Y.; Chen, C. S. Development of small-molecule cyclin
D1-ablative agents. J. Med. Chem. 2006, 49, 4684–4689.
(9) Chaplain, M. A.; McDougall, S. R.; Anderson, A. R. Mathemat-
ical modeling of tumor-induced angiogenesis. Annu. Rev. Biomed.
Eng. 2006, 8, 233–257.
4.2. Cell Culture and Transactivation Assays. Human embry-
onic kidney HEK293 cells were cultured in Dulbecco’s Modified
Eagle’s Medium (DMEM) containing 5% fetal bovine serum
and antibiotic-antimycotic mixture (Nacalai) at 37 ꢀC in a humid-
ified atmosphere of 5% CO2 in air. Transfections were performed
by means of the calcium phosphate coprecipitation method. Eight
hours after transfection, test compounds were added. Cells were
harvested approximately 16-20 h after the treatment, and lucif-
erase and β-galactosidase activities were measured with a lumino-
meter and a microplate reader. DNA cotransfection experiments
included 50 ng of reporter plasmid, 20 ng of pCMX-ss-galactosidase,
15 ng of each receptor expression plasmid, and pGEM carrier
DNA to make a total of 150 ng of DNA per well in a 96-well
plate. Luciferase data were normalized to an internal β-galacto-
sidase control and reported values are the means of triplicate
assays.
4.3. X-ray Crystallography of the Ligand/PPAR-LBD Com-
plex. Determination of the 4-hPPARR LBD, the 5-hPPARδ
LBD, and the 6-hPPARγ LBD cocrystal structures, including
protein purification, crystal growth, and structural refinement,
was performed as described previously.23
The human PPARγ LBD (aa 203-477) was purified as
described previously.30 All crystals were prepared through cocrys-
tallization with (R)-7 or (S)-7 under the same conditions to permit
straightforward comparison between the two structures. An ali-
quot (2 μL) of protein solution (15 mg/mL in 20 mmol/L Tris,
pH 8.0, 150 mmol/L NaCl) was mixed with an equal volume of
0.1 mmol/L ligand solution (0.8 mol/L sodium citrate, 100 mmol/L
HEPES, pH 7.5, 1% DMSO), and then crystallization of the
ligand/protein mixture was achieved by the hanging drop vapor
diffusion method at 293K. Crystals were flash-cooled in a liquid
€
€
(10) Sato, M. Peroxisome proliferator activated receptor ligands and
angiogenesis. Nippon Rinsho 2005, 63, 603–608.
(11) Lehmann, J. M.; Moore, L. B.; Smith-Oliver, T. A.; Wilkison, W. O.;
Willson, T. M.; Kliewer, S. A. An antidiabetic thiazolidinedione is a

Article
high affinity ligand for peroxisome proliferator-activated receptor
gamma (PPAR gamma). J. Biol. Chem. 1995, 270, 12953–12956.
(12) Hashimoto, Y.; Miyachi, H. Nuclear receptor antagonists designed
based on the helix-folding inhibition hypothesis. Bioorg. Med. Chem.
2005, 13, 5080–5093.
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 341
B. R.; Hull-Ryde, E. A.; Lake, D. H.; Lenhard, J. M.; Oliver, W.,
Jr.; Oplinger, J.; Pentti, M.; Parks, D. J.; Plunket, K. D.; Tong,
W. Q. N-(2-Benzoylphenyl)-L-tyrosine PPARgamma agonists. 3.
Structure-activity relationship and optimization of the N-aryl
substituent. J. Med. Chem. 1998, 41, 5055–5069.
(13) 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 PPARR subtype-selective activators. J. Med. Chem. 2003, 46,
3581–3599.
(14) Kasuga, J.; Yamasaki, D.; Araya, Y.; Nakagawa, A.; Makishima,
M.; Doi, T.; Hashimoto, Y.; Miyachi, H. Design, synthesis and
evaluation of a novel series of R-substituted phenylpropanoic acid
derivatives as human peroxisome proliferator-activated receptor
(PPAR) alpha/delta dual agonists for the treatment of metabolic
syndrome. Bioorg. Med. Chem. 2006, 14, 8405–8414.
(15) Kasuga, J.; Nakagome, I.; Aoyama, A.; Sako, K.; Ishizawa, M.;
Ogura, M.; Makishima, M.; Hirono, S.; Hashimoto, Y.; Miyachi,
H. Design, synthesis, and evaluation of potent, structurally novel
peroxisome proliferator-activated receptor (PPAR) delta-selective
agonists. Bioorg. Med. Chem. 2007, 15, 5177–5190.
(16) Kasuga, J.; Yamasaki, D.; Ogura, K.; Shimizu, M.; Sato, M.;
Makishima, M.; Doi, T.; Hashimoto, Y.; Miyachi, H. SAR-oriented
discovery of peroxisome proliferator-activated receptor pan agonist
with a 4-adamantylphenyl group as a hydrophobic tail. Bioorg. Med.
Chem. Lett. 2008, 18, 1110–1105.
(17) Xu, H. E.; Lambert, M. H.; Montana, V. G.; Plunket, K. D.;
Moore, L. B.; Collins, J. L.; Oplinger, J. A.; Kliewer, S. A.; Gampe,
R. T., Jr.; McKee, D. D.; Moore, J. T.; Willson, T. M. Structural
determinants of ligand binding selectivity between the peroxisome
proliferator-activated receptors. Proc. Natl. Acad. Sci. U.S.A.
2001, 98, 13919–13924.
(18) Evans, D. A.; Ennis, M. D.; Mathre, D. J. Asymmetric alkylation
reactions of chiral imide enolates. A practical approach to the
enantioselective synthesis of alpha-substituted carboxylic acid
derivatives. J. Am. Chem. Soc. 1982, 104, 1737–1739.
(22) Hulin, B.; Newton, L. S.; Lewis, D. M.; Genereux, P. E.; Gibbs,
E. M.; Clark, D. A. Hypoglycemic activity of a series of alpha-
alkylthio and alpha-alkoxy carboxylic acids related to ciglitazone.
J. Med. Chem. 1996, 39, 3897–3907.
(23) Oyama, T.; Toyota, K.; Waku, T; Hirakawa, Y.; Nagasawa, N.;
Kasuga, J.; Hashimoto, Y.; Miyachi, H.; Morikawa, K. Adapt-
ability and selectivity of human peroxisome proliferator-activated
receptor (PPAR) pan agonists revealed from crystal structures.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 2009, 65, 786–795.
(24) Computer-graphic work was performed using MOE (Molecular
Operating Environment) (Ryoka Systems Inc.).
(25) Nolte, R. T.; Wisely, G. B.; Westin, S.; Cobb, J. E.; Lambert, M. H;
Kurokawa, R.; Rosenfeld, M. G.; Willson, T. M.; Glass, C. K.; Milburn,
M. V. Ligand binding and co-activator assembly of the peroxisome
proliferator-activated receptor-gamma. Nature 1998, 395, 137–143.
(26) Rosenfeld, M. G.; Lunyak, V. V.; Glass, C. K. Sensors and signals:
a coactivator/corepressor/epigenetic code for integrating signal-
dependent programs of transcriptional response. Genes Dev. 2006,
20, 1405–1428.
(27) Nagy, L.; Kao, H. Y.; Love, J. D.; Li, C.; Banayo, E.; Gooch, J. T.;
Krishna, V.; Chatterjee, K.; Evans, R. M.; Schwabe, J. W. Mech-
anism of corepressor binding and release from nuclear hormone
receptors. Genes Dev. 1999, 13, 3209–3216.
(28) Ogston, A. G. Interpretation of experiments on metabolic pro-
cesses, using isotopic tracer elements. Nature 1948, 163, 963.
(29) Easson, L. H.; Stedman, E. Studies on the relationship between
chemical constitution and physiological action;Molecular dissym-
metry and physiological activity. Biochem. J. 1933, 27, 1257–1266.
(30) Waku, T.; Shiraki, T.; Oyama, T.; Fujimoto, Y.; Maebara,
K.; Kamiya, N.; Jingami, H.; Morikawa, K. Structural insight into
PPARgamma activation through covalent modification with
endogenous fatty acids. J. Mol. Biol. 2009, 385, 188–199.
(19) The enantiomeric excess was determined with a CHIRALPAK
AD-H column, with hexane/i-PrOH/trifluoroacetic acid=90:10:
0.1 v/v/v as the mobile phase, at a flow rate of 0.5 mL/min, with
detection at 254 nm.
(20) Sznaidman, M. L.; Haffner, C. D.; Maloney, P. R.; Fivush, A.;
Chao, E.; Goreham, D.; Sierra, M. L.; LeGrumelec, C.; Xu, H. E.;
Montana, V. G.; Lambert, M. H.; Willson, T. M.; Oliver, W. R., Jr.;
Sternbach, D. D. Novel selective small molecule agonists for peroxi-
some proliferator-activated receptor delta (PPARdelta); synthesis
and biological activity. Bioorg. Med. Chem. Lett. 2003, 13, 1517–1521.
(21) Cobb, J. E.; Blanchard, S. G.; Boswell, E. G.; Brown, K. K.;
Charifson, P. S.; Cooper, J. P.; Collins, J. L.; Dezube, M.; Henke,
(31) Otwinowski, Z.; Minor, W. Methods Enzymol. 1997, 276,
307–326.
(32) Brunger, A. T.; Adams, P. D.; Clore, G. M.; DeLano, W. L.; Gros,
€
P.; Grosse-Kunstleve, R. W.; Jiang, J.-S.; Kuszewski, J.; Nilges,
M.; Pannu, N. S.; Read, R. J.; Rice, L. M.; Simonson, T.; Warren,
G. L. Crystallography & NMR system: A new software suite for
macromolecular structure determination. Acta Crystallogr. 1998,
D54, 905–921.
(33) Jones, T. A.; Zou, J.-Y.; Cowan, S. W.; Kjeldgaard, M. Improved
methods for building protein models in electron density maps and
the location of errors in these models. Acta Crystallogr. 1991, A47,
110–119.