Design and synthesis of potent and subtype-selective PPARα agonists

Article abstract of DOI:10.1016/j.bmcl.2005.12.022

Beginning with a moderately potent PPARγ agonist 9, a series of potent and highly subtype-selective PPARα agonists was identified through a systematic SAR study. Based on the results of the efficacy studies in the hamster and dog models of dyslipidemia and the desired pharmacokinetic data, the optimized compound 39 was selected for further profiling.

Full text of DOI:10.1016/j.bmcl.2005.12.022

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1673–1678
Design and synthesis of potent and subtype-selective
PPARa agonists
Ranjit C. Desai,a,* Edward Metzger,a Conrad Santini,a Peter T. Meinke,a James V. Heck,a
Joel P. Berger,b Karen L. MacNaul,b Tian-quan Cai,c Samuel D. Wright,c Arun Agrawal,d
David E. Mollerb and Soumya P. Sahooa,*
aDepartment of Medicinal Chemistry, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
bDepartment of Metabolic Disorders, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
cDepartment of Cardiovascular Diseases, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
dDepartment of Drug Metabolism, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
Received 16 November 2005; revised 2 December 2005; accepted 2 December 2005
Available online 27 December 2005
Abstract—Beginning with a moderately potent PPARc agonist 9, a series of potent and highly subtype-selective PPARa agonists
was identified through a systematic SAR study. Based on the results of the efficacy studies in the hamster and dog models of dysl-
ipidemia and the desired pharmacokinetic data, the optimized compound 39 was selected for further profiling.
Ó 2006 Elsevier Ltd. All rights reserved.
Coronary heart disease (CHD) is the leading cause of
death in the US and much of the developed world. An
estimated 13 million Americans have CHD. The direct
and indirect costs for CHD in US are quite staggering
approaching more than $130 billion per year.1 Among
the risk factors for CHD, two of the most significant
ones are elevated LDL cholesterol and low HDL choles-
terol. In addition, hypertriglyceridemia is also a likely
contributor to CHD risk. The statin class of HMG-
CoA reductase inhibitors (Zocor, Lipitor) and the fi-
brate class of antidyslipidemic drugs such as fenofibrate
or clofibrate (Fig. 1) have found wide acceptance for the
clinical management of dyscholesterolemia and dyslipi-
demia. Statins effectively lower serum LDL-c levels but
their HDL-raising effect is marginal. Fibrates, on the
other hand, are quite effective at lowering serum triglyc-
erides, LDL-c and also raise HDL cholesterol levels.
The triglycerides-lowering and HDL-raising effects of fi-
brates are attributed to the activation of PPARa recep-
tors. This results in an increase in lipoprotein lipase gene
expression and transrepression of apoC-III thereby
increasing lipoprotein catabolism.2
The HDL-cholesterol elevation observed with fibrates
arises in part from the up-regulation of apoA-I and
apoA-II.3 Though effective, for dyslipidemia, fibrates
are weak PPARa agonists and their subtype selectivity
is poor. Therefore, a potent and subtype-selective hu-
man PPARa agonist could offer a superior alternative
for the management of dyslipidemia. The first reported
examples of selective PPARa agonists were GW9578
34 and GW7647 4.5 This was followed by publications
from researchers at Kyorin 56 and Lilly 6.7 The PPARa
selectivity of compounds 3–6 ranged from 20- to 200-
fold over human PPARc and PPARd receptors. Recent-
ly, researchers from these laboratories have reported
examples of potent and highly subtype-selective PPARa
agonists as exemplified with cyclic fibrates 7 and 8.8,9 We
have earlier disclosed work in the PPARa/c dual agonist
area10,11 and as an extension of that work described
below are the results of our efforts in identifying potent
and highly subtype-selective (>1000-fold) PPARa
agonists.
With the objective of looking at the outcome of replac-
ing a TZD headpiece on a moderately potent PPARc
agonist 9 with propionic acid as found in classical
fibrates such as fenofibrate, we synthesized compound
10. In the binding assay, this analog displayed affinity
for all three human PPAR receptors.
Keyword: PPARa agonist.
*
Corresponding authors. Fax: +1 732 594 9556 (R.C.D); e-mail:
ranjit_desai@merck.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.12.022

1674
R. C. Desai et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1673–1678
O
R1
NHR2
O
O
Cl
N
HO
N
H
PriO
HO
EtO
O
O
OMe
CF3
O
Cl
S
O
O
O
5
3 R1 = n-Pr, R2 = 2,4-difluorophenyl
4 R1 = R2 = cyclohexyl
1
Fenofibrate
2
Clofibrate
H
N
OCH2CF3
HO
OCF3
O
O
HO
N
HO
N
O
O
O
O
O
Cl
O
O
8
O
O
Cl
7
6
Me
Figure 1. PPARa agonists.
Substituting the propyl chain in 10 with a chloro substi-
tuent as in compound 11 showed a 4-fold improvement
in PPARa potency,12 but more importantly, in the func-
tional transactivation (TA) assay this analog displayed
only PPARa activity (Fig. 2). Encouraged by these re-
sults, we initiated an SAR study to investigate the role
of the methylene spacer and the orientation of the pro-
pionic acid side chain. As seen from the TA data in
Figure 3, extending the methylene bridge by an extra
unit results in PPARa,c dual agonist 12, whereas reduc-
ing the methylene spacer by one carbon (compound 13)
leaves it unchanged when compared to 11. Similarly, the
meta-oriented propionic acid analog 14 is indistinguish-
able from 11, whereas the corresponding ortho-linked
derivative 15 was devoid of useful activity. Considering
the potential for the metabolic oxidation at the para car-
bon atom in the meta-linked analog 14, we elected to
pursue only the para-linked series. Next, we probed
the effects of alkyl group substitutions on the propionic
acid side chain to optimize the substituents (Table 1). In
the case of mono-substituted analogs 16–20, there was
not much variance in binding activity. The Me/Et-
substituted analog 21 was found to be the most potent
analog displaying around 250-fold PPARa selectivity
in the binding assay. The diethyl-substituted analog 22
was essentially indistinguishable from 21. However,
due to synthetic ease in making 21, we pursued only this
substitution pattern for further SAR. Having estab-
lished the optimum substituents, the next endeavor
was to separate the enantiomers and determine if there
was a difference in the activity. This was conveniently
achieved by a chiral resolution of the starting phenol
46 using Chiracel OJ column and converting the enanti-
omers to the final targets 24 and 25. The absolute
O
O
O
HN
OH
S
OH
O
O
O
O
O
O
O
O
F
O
O
F
O
O
F
Cl
11
10
9
hPPAR α/γ/δ SPA EC50 (nM) = 30/981/2390
hPPAR TA EC50 (nM) = 20/ND/>3000
hPPAR α/γ/δ SPA EC50 (nM) = NA/735/NA
hPPAR TA EC50(nM) = NA/80/NA
hPPAR α/γ/δ SPA EC50(nM) = 110/1087/106
hPPAR TA EC50 (nM) = 51/ND/55
Figure 2. Lead development.
O
O
O
O
O
F
O
O
O
O
F
OH
O
OH
2
12
Cl
Cl
13
hPPAR α/δ/γ SPAEC50 (nM) = 66/7690.545
hPPAR α/δ/γ SPAEC50 (nM) = 87/NA/NA
hPPAR TA EC50 (nM) = 35/ND/284
hPPAR TA EC50 (nM) = 62/NA/NA
O
O
HO
O
O
F
O
O
O
F
O
O
Cl
O
Cl
14
15
hPPAR α/δ/γ SPA EC50 (nM) = 36/NA/2230
HO
hPPAR α/δ/γ SPA EC50 (nM) = NA
hPPAR TA EC50 (nM) = 21/ND/>3000
Figure 3. Role of methylene tether and orientation of acid side chain.

R. C. Desai et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1673–1678
Table 1. In vitro human PPAR activities of compounds 16–25
1675
subtype selectivity. In particular, the trifluoromethyleth-
yl containing analog 31 displayed subnanomolar bind-
ing affinity for the PPARa receptor and >5000-fold
subtype selectivity making it perhaps the most potent
agonist reported to date. Unfortunately, 31 was also
HOOC
R
R
1
2
R
O
O
Cl
found to be
a
potent inhibitor of Cyp2C9
Compound R1, R2
Binding IC50
(nM)17
Transactivation
EC50 (nM)16
(IC50 < 0.05 lM), thus precluding it from further evalu-
ation. In general, all compounds in Table 2 suffered to a
varying degree from P450 enzyme inhibition issues
making it necessary to explore different phenolic ring
systems as coupling partners with the western phenols
containing 1,3-propylidene linker.
a
d
c
a
d
c
16
17
18
19
20
11
21
22
23
24
25
H, H
675 >50 6440
208 4070 NA
156 1170 NA
H, Me
H, Et
H, Pr
H, Ph
130 1130 NA 454
262 1340 NA
30 981 2390 20
Toward that end, drawing from our earlier experience
with PPARc agonists,13 we first synthesized compound
34, a trifluoromethyl-substituted coumarin derivative
(Table 3). Though not as potent as some of the analogs
described earlier, 34 nonetheless maintained the PPARa
subtype selectivity. Introduction of the chloro substitu-
ent at the 6 position (35) improved potency almost
20-fold furnishing a 28 nM PPARa-selective agonist.
Interestingly, moving chlorine to 8 position as in 36
caused an almost 6-fold drop in the PPARa potency
and also introduced some binding activity for PPARd
when compared to 34. Unlike coumarin derivatives 34,
the corresponding lactam analog 38 was found to be
inactive at the PPAR receptors, thus highlighting the
need for the oxygen atom for PPAR activity. Based on
the superior in vitro potency, compound 35 was selected
for in vivo evaluation of hypolipidemic efficacy. As
shown in Table 5, oral administration of compound 35
at a dose of 1 mpk for 9 days lowered serum triglycer-
ides and cholesterol by 35% and 41%, respectively. For
comparison, fenofibrate achieved 30% and 31% reduc-
tion of triglycerides and cholesterol at a much higher
dose of 100 mpk. Compound 35 was also evaluated in
a dog model to assess its lipid-lowering profile. The
dog model was chosen because dogs exhibit strong lipid
lowering in response to fibrates and statins, and have
been used as the principal preclinical species during
the development of simvastatin (Zocor).12,13 Compound
35 was orally dosed at 1 mpk for 14 days. For compar-
ison purposes, in this study simvastatin and fenofibrate
were dosed po at 4 and 50 mpk, respectively. As seen
from the data in Table 6, average serum cholesterol
was reduced by 22%, whereas simvastatin showed an
average decrease of 19%. Interestingly, an additive effect
of cholesterol lowering was observed when 35 was
co-dosed with simvastatin. A combined dose of 35 at
3 mpk and simvastatin at 4 mpk lowered cholesterol
by 32% which is greater than by either 35 or simvastatin
at the corresponding dose (Table 6). Based on the results
of superior in vivo efficacy in the two animal models, 35
was characterized in pharmacokinetic studies in three
preclinical animal species (Table 7).
Me, Me
Me, Et
Et, Et
ND >3000
5 NA 1240
22 NA NA
2.5 ND NA
2.7 ND NA
Cyclobutyl 116 1350 NA
Me, Et R
Me, Et S
15
ND NA
ND NA
ND >3000
6 1090 NA
101 2414 NA
3
20
NA, not active; ND, not determined.
configuration of the slower eluting isomer was deter-
mined to be (S) based on the X-ray crystallographic data
on the corresponding (2R)-2-phenyloxazoline amide
derivative (Scheme 2). As seen from the data, the (R)
enantiomer 24 is around 7-fold more potent that the cor-
responding (S) isomer 25.
Having established the optimum substituents on the
propionic acid side chain as well as identifying the de-
sired (R) enantiomer as the preferred isomer, the stage
was now set for the further expansion of SAR on the
eastern phenols. Earlier, we had identified the 4 position
as a suitable site for substituent introduction. Accord-
ingly, a variety of substituents were introduced at the
4 position and as seen from the data in Table 2, with
the exception of compounds 26–27 and 30 which showed
PPARa/d agonism in the binding assay, all other ana-
logs displayed potent PPARa activity and an excellent
Table 2. In vitro human PPAR activities of compounds 26–33
HOOC
R
O
O
Cl
Compound
R
H
Binding IC50
(nM)17
Transactivation
EC50 (nM)16
a
d
c
a
d
c
26
27
28
29
30
31
42 4630
742
851
NA
NA
NA
NA
54.1 ND NA
2.1 ND NA
7.8 ND NA
11.4 ND NA
2.9 ND NA
OCF3
3
OCH2CF3
OSO2Me
CF3
10 NA
30 NA
3
603
CH2CF3
<1 >70000 >6000 <1
ND NA
Overall, 35 exhibited low plasma clearance, good oral
bioavailability, and systemic exposure across the species
Encouraged by these results, we looked at the Pan Labs
assay18 for the off-target activity for 35. No significant
off-target activity was observed for compound 35.
Unfortunately, the results for the stability studies of 35
indicated the lactone ring stability issues. It was
32
33
2
NA
1851
3610
1
ND >3000
Me
Me
22 NA
27.5 ND >3000
NA, not active; ND, not determined.

1676
R. C. Desai et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1673–1678
Table 3. In vitro human PPAR activities of compounds 34–38
HOOC
O
OR
Compound
R
Binding IC50 (nM)17
Transactivation EC50 (nM)16
a
d
c
a
d
c
CF3
34
537
NA
NA
3270
134
NA
NA
NA
NA
ND
ND
ND
O
O
CF3
Cl
35
36
37
28
157
49
44
ND
ND
ND
>3000
ND
O
O
O
CF3
ND
10
O
Cl
CF3
ND
O
O
Me
Me
38
6090
NA
NA
ND
ND
ND
N
H
O
NA, not active; ND, not determined.
Table 4. In vitro human PPAR activities of compounds 39–42
HOOC
O
OR
Compound
R
Binding IC50 (nM)17
Transactivation EC50 (nM)16
a
d
c
a
d
c
CF
O
3
Cl
39
20
NA
NA
NA
NA
7.3
ND
NA
ND
ND
CF
3
Cl
Cl
40
41
125
19
NA
ND
ND
ND
ND
O
1845
O
CF
3
42
NA
NA
NA
10
ND
ND
O
NA, not active; ND, not determined.

R. C. Desai et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1673–1678
1677
Table 5. In vivo efficacy of compounds 35 and 39 on serum cholesterol
and triglycerides in hamster14
observed that at pH 7.5, 35 exists as an equilibrium mix-
ture of the closed and the open forms (Scheme 1). This
posed a significant developmental problem. To address
the instability issue, we transformed the lactone
carbonyl into a gem-dialkyl-substituted derivative. 39
(Table 4). To our delight, 39 not only retained the po-
tent PPARa-binding activity but also showed 6-fold
potency improvement in the functional TA assay. The
corresponding gem-diethyl analog 40 gave up some
potency, whereas analog 41 displayed binding activity
at the PPARa and PPARd receptors. In the in vivo effi-
cacy studies, 39 showed robust lowering of serum tri-
glycerides and cholesterol in the two animal models,
hamster and dog, at a dose of 0.3 mpk. Also, like its
predecessor 35, compound 39 showed an additive
cholesterol-lowering effect when co-administered with
simvastatin (Table 6). Compound 39 was also character-
ized in pharmacokinetic studies in three preclinical
animal species and exhibited low plasma clearance, good
oral bioavailability and systemic exposure across the
species (Table 7). No significant off-target activity was
observed for 39 in the Pan Labs counterscreens.
Compound Dose (mpk) Cholesterol (%) Triglyceride (%)
35
39
1
0.3
À41
À21
À31
5
3
3
À35
À14
À30
6
6
4
Fenofibrate 100
Table 6. Cholesterol lowering by compounds 35, 39, fenofibrate, and
simvastatin in male beagle dogs15
Compound
Dose (mpk)
Cholesterol (%)
35
39
1
À22
À19
À18
À19
À32
À26
4
3
4
2
2
5
0.3
50
Fenofibrate
Simvastatin
35 + simvastatin
39 + simvastatin
4
3 + 4
0.3 + 4
Table 7. Pharmacokinetic data for compounds 35 and 39
Compound Species Cl AUC (po) t1/2
(mL minÀ1 kgÀ1
(lM h/mL) (h) (%)
1.6
1.45
The chemistry used in the preparation of analogs 11–
42 is illustrated with the synthesis of compound 39
(Scheme 2). Commercially available 4-benzyloxyphenol
was first alkylated with methyl 2-bromobutyrate to
give 44 which on alkylation with MeI followed by
removal of the benzyl-protecting group provided the
bis alkylated derivative 46. Chiral resolution of 46
using Chiracel OJ column furnished enantiomers 47
and 48. Alkylation of slower eluting S enantiomer
with benzyl 3-bromopropylbromide followed by
hydrogenolysis gave the desired alcohol derivative
F
)
35
39
Rat
17
2.3 84
3.6 99
2.5 39
Dog
30.6
Monkey 9.45
3.95
Rat
Dog
2.3
1.5
25.4
8.7
2.4 94
7.7 35
5.2 24
Monkey 7.8
3.95
Doses used were 0.5 mg/kg, iv and 2 mg/kg, po (n = 3, except monkey,
where, n = 2).
CF3
OH
O
CF3
NaOOC
HOOC
O
Cl
O
Cl
ONa
NaOH -MeOH
O
O
O
O
O
O
35
Scheme 1. Lactone ring stability of compound 35.
HO
MeOOC
O
MeOOC
O
b
a
C6H5
c
Et Me
Et
O
O
C6H5
O
C6H5
43
45
44
R
Ph
O
O
MeOOC
MeOOC
O
MeOOC
O
O
N
d
O
R
S
Et Me
O
OH
Faster eluting
OH
OH
Slower eluting
47
48
S
46
OMe
X-ray20
CF3
MeOOC
HOOC
O
O
Cl
O
e, f
g, h
47
O
OH
O
O
49
39
Scheme 2. Reagents: (a) COOMeCHBrCH2CH3, Cs2CO3, CH3CN; (b) LDA, MeI, THF, À78 °C; (c) Pd/C, H2; (d) Chiracel OJ; (e) Br(CH2)3OBn,
Cs2CO3, DMF; (f) Pd/C, H2; (g) (C6H5)3P, EtOOCN@NCOOEt, 53, THF; (h) NaOH–MeOH. See above mentioned reference for further
information.

1678
R. C. Desai et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1673–1678
CF3
CF3
O
CF3
O
Cl
Cl
Cl
Cl
a
b
c
OH
HO
O
HO
OH
HO
OH
HO
50
51
53
52
Scheme 3. Reagents: (a) CF3COCH2COOMe, H2SO4; (b) MeMgBr; (c) pTSA.
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.
J. Med. Chem. 2003, 46, 5121.
49. Mitsunobu reaction of 49 with phenol 53 followed
by hydrolysis of the methyl ester furnished the final
target 39. The preparation of phenol 53 is described
in Scheme 3.19
8. Koyam, H.; Boueres, J. K.; Miller, D. A.; Berger, J. P.;
MacNaul, K. L.; Wang, P.; Ippolito, M. C.; Wright, D.
D.; Agrawal, A. K.; Moller, D. A.; Sahoo, S. P. Bioorg.
Med. Chem. Lett. 2005, 15, 3347.
9. 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.; Hernandez, M.; Wright, S. D.;
Moller, D. E.; Heck, J. V.; Meinke, P. T. J. Med. Chem.
2005, 48, 5589.
10. Desai, R. C.; Han, W.; Metzger, E. J.; Bergman, J. P.;
Gratale, D. F.; MacNaul, K. L.; Berger, J. P.; Doebber, T.
W.; Leung, K.; Moller, D. E.; Heck, J. V.; Sahoo, S. P.
Bioorg. Med. Chem. Lett. 2005, 13, 2795.
11. Desai, R. C.; Gratale, D. F.; Han, W.; Koyama, H.;
Metzger, E. J.; Lombardo, V. K.; MacNaul, K. L.; Berger,
J. P.; Doebber, T. W.; Leung, K.; Franklin, R.; Moller, D.
E.; Heck, J. V.; Sahoo, S. P. Bioorg. Med. Chem. Lett.
2005, 13, 3541.
Condensation of 4-chlororesorcinol with methyl trifluo-
roacetoacetate in the presence of sulfuric acid gave
coumarin 51, which on treatment with excess methyl
magnesium bromide gave compound 52. pTSA cata-
lyzed ring closure of 52 furnished phenol 53.
In summary, starting with a weak PPARc lead, we have
developed through a systematic SAR studies a series of
potent and subtype-selective PPARa agonists. Based on
its excellent in vivo efficacy as well as the desirable phar-
macokinetic data, compound 39 was selected for further
evaluation. Finally, the additive cholesterol-lowering
effect seen on co-administering 39 with simvastatin in
the dog model offers an interesting possibility of
combining a potent PPARa agonist with statins to more
effectively manage the lipid profile in high risk patients
with chronic heart disease.
12. We have consistently noted in several cases 4- to 10-fold
improvements in PPARa potency by replacing propyl
chain with chloro substituent.
13. Unpublished results from these laboratories. The couma-
rin containing analogs 34–37 did not have cytochrome
P450 enzyme inhibition issues.
Acknowledgments
14. Male Golden Syrian hamsters (120–150 g weight, n = 10)
were fed normal rodent chow with free access to water and
received once-a-day oral dosing of the sodium salts of the
tested compounds by gavage with vehicle (0.5% methyl-
cellulose) for 9 days.
15. Male beagle dogs (12–18 kg weight, n = 5) were fed a
cholesterol-free chow diet ad libitum with free access to
water. Test compounds were suspended in 0.5% methyl-
cellulose and gavaged daily for 14 days. Mean values are
shown. Data at the final day were p < 0.05 against vehicle
control.
We thank Neelam Sharma, Marc Ippolito, Frank Xia-
odong Gan, Melba Hernandez, Ying Li, and Raul Alva-
ro for the technical support. We also thank Nancy Tsou
and Richard G. Ball for X-ray crystallography data.
Thanks are also due to Dr. Timothy Blizzard for help
with the manuscript preparation.
References and notes
16. Berger, J. P.; Leibowitz, M. D.; Doebber, T. W.; Elbrecht,
A.; Zhang, B.; Zhou, G.; Biswas, C.; Cullinan, C. A.;
Hayes, N. S.; Li, Y.; Tanem, M.; Venture, J.; Wu, S. M.;
Berger, G. D.; Mosley, R.; Marquis, R.; Santini, C.;
Sahoo, S. P.; Tolman, R. L.; Smith, R. G.; Moller, D. E.
J. Biol. Chem. 1999, 274, 6718.
17. Adams, A. D.; Yuen, W.; Hu, Z.; Santini, C.; Jones, B. A.;
MacNaul, K. L.; Berger, J. P.; Doebber, T. B.; Moller, D.
E. Bioorg. Med. Chem. Lett. 2003, 12, 931.
18. These assays were performed by MDS Pharma Services.
19. The preparation of compound 39 has been described:
Desai, R. C.; Sahoo, S. P. WO Patent 2004/010992 A1.
20. CCDC 291597 contains the supplementary crystallograph-
ic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
1. Anum, E. A.; Adera, T. Ann. Epidemiol. 2004, 14, 705.
2. Staels, B.; Auwerx, J. Curr. Pharm. Des. 1997, 3, 1.
3. Staels, B.; Dallongeville, J.; Auwerx, J.; Schoonjans, K.;
Leitersdorf, E.; Fruchart, J. C. Circulation 1998, 89, 2088.
4. 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.; Kliewer, S. A.; Wilson, T. M.
J. Med. Chem. 1999, 423, 3785.
5. Brown, P. J.; Stuart, L. W.; Hurley, K. P.; Lewis, M. C.;
Winegar, D. A.; Wilson, J. G.; Wilkinson, W. O.; Ittoop, O.
R.; Wilson, T. M. Bioorg. Med. Chem. Lett. 2001, 11, 1225.
6. Nomura, M.; Tanase, T.; Ide, T.; Tsunoda, M.; Suzuki,
M.; Uchiki, H.; Murakami, K.; Miyachi, H. J. Med.
Chem. 2003, 46, 3581.
7. Xu, Y.; Mayhugh, D.; Saeed, A.; Wang, X.; Thompson,
R. C.; Dominianni, S. J.; Kauffman, R. F.; Singh, J.; Bean,