Selective PPARγ modulators with improved pharmacological profiles

Article abstract of DOI:10.1016/j.bmcl.2005.03.092

A series of metabolically robust N-benzyl-indole selective PPARγ modulators with either a 3-benzoyl or 3-benzisoxazoyl moiety have been identified. In vitro, these compounds are partial agonists and exhibit reduced adipogenesis in human adipocytes. In vivo, these SPPARγMs result in potent glucose lowering in db/db mice and attenuate increases in heart weight and brown adipose tissue that is typically observed in rats upon treatment with PPARγ full agonists.

Full text of DOI:10.1016/j.bmcl.2005.03.092

Bioorganic & Medicinal Chemistry Letters 15 (2005) 2437–2440
Selective PPARc modulators with improved
pharmacological profiles
Kun Liu, Regina M. Black, John J. Acton, III, Ralph Mosley, Sheryl Debenham,
Ramon Abola, Meng Yang, RichardTschirret-Guth, Lawrence Colwell, Cherrie Liu,
Margaret Wu, Chuanlin F. Wang, Karen L. MacNaul, Margaret E. McCann,
DavidE. Moller, Joel P. Berger, Peter T. Meinke, A. Brian Jones andHaroldB. Wood
*
Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
Received25 February 2005; revised21 March 2005; accepted23 March 2005
Available online 16 April 2005
Abstract—A series of metabolically robust N-benzyl-indole selective PPARc modulators with either a 3-benzoyl or 3-benzisoxazoyl
moiety have been identified. In vitro, these compounds are partial agonists and exhibit reduced adipogenesis in human adipocytes.
In vivo, these SPPARcMs result in potent glucose lowering in db/db mice and attenuate increases in heart weight and brown adipose
tissue that is typically observedin rats upon treatment with PPAR c full agonists.
ꢀ 2005 Elsevier Ltd. All rights reserved.
The peroxisome proliferator-activatedreceptor gamma
HO2C
(PPARc) is member of a large family of ligandactivated
nuclear transcription factors. Both rosiglitazone and
pioglitazone are full agonists that specifically activate
PPARc andare usedto treat type 2 iadbetes
(T2DM).1 However, their efficacy is limitedude to
mechanism-basedavderse events (AEs). In humans,
AEÕs associatedwith PPAR c activation include weight
gain, edema, and anemia, while in rodents cardiac
hypertrophy is also observed.2 Although the underlying
causes for these AEs are poorly understood, a PPARc
ligandthat ameliorates these AEs wouldbe desirable.
In search of PPAR ligands devoid of these AEs, new
classes of selective PPARc modulators (SPPARcMs)3
have been reported.4 We recently reportedthe discovery
of SPPARcMs that exhibit potent glucose lowering in
the db/db mouse model for T2DM.5 Interestingly, these
indole derived SPPARcMs did not cause cardiac hyper-
trophy (CH) andattenuatedbrown aidpose tissue
(BAT) increases in S–D rats.
O
HO2C
F3CO
O
F3CO
in vivo
N
O
N
H
1a (CYP2C9 IC50 = 50 nM)
H3CO
1
Figure 1. In vivo loss of the N-benzoyl group.
isoforms (a,d,c) in vitro, it was a major circulating
metabolite in plasma which accumulatedon multiple
dosing (10% of parent). Unfortunately, 1a was a potent
inhibitor of CYP2C9 (IC50 = 50 nM), a liability which
precluded further studies. Herein, we describe the identi-
fication andevaluation of secondgeneration inodle-
basedSPPAR cMs with improvedmetabolic profiles
that are inactive on the a and d PPAR isoforms.
Alternative replacements for the N-benzoyl group that
wouldbe less susceptible to metabolism were investi-
gated. Considering the importance of the benzoyl car-
bonyl for activity, both pyridyl and quinolyl analogs
(2–4) with a nitrogen positionedto replace the carbonyl
of the p-anisoyl group were appended onto the nitrogen
of the indole. Unfortunately, these derivatives exhibited
diminished in vitro potency (Table 1).
During preclinical evaluation of 1 in rats, loss of the p-
anisoyl group was observed( Fig. 1). Although the
unsubstitutedindole 1a was inactive on all three PPAR
Keywords: PPAR; Selective modulator; Partial agonists.
*
Corresponding author. Tel.: +1 732 594 7592; fax: +1 732 594
9556; e-mail: blair_wood@merck.com
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.03.092

2438
K. Liu et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2437–2440
Table 1. Human PPARc binding (SPA) IC50s andtransactivation
tion andthe propensity for the unsubstitutedindole
metabolite to accumulate on multiple dosing in rat, ana-
logs completely devoid of this liability were desired.
EC50s6
S
O
N
R
O
F3CO
O
O
To this end, metabolically more stable analogs were pre-
paredby repositioning the substituents on the indole nu-
cleus. Transposing the attachment of the benzyl moiety
from the 3-position to the nitrogen of the indole pro-
vided compounds with the desired features. With these
transposed N-benzyl indoles attaching the benzoyl or
the 1-BZX to the indole 3-position returned analogs that
were potent SPPARcMs (11, 12, Table 1). The SAR of
the transposed indole analogs tracked identically with
the N-benzoyl indoles reported previously.
CO2H
N
Het
N
N
2, 2-pyridyl
3, 2-(6-OMe)quinolyl
4, 3-(6-OMe)quinolyl
Rosiglitazone, R = H
15, R = CH3
Y
4
6
O
F3CO
F3CO
N
OR
O
2
CO2H
N
O
OCH3
N
O
Cl
Given the unique differences in the transactivation assay
with these SPPARcMs relative to full agonists, further
investigation of these SPPARcMs was undertaken.
The acidic TZD head group present in rosiglitazone is
crucial for its binding and functional activity. For exam-
ple, N-methyl rosiglitazone (15) was devoid of both
binding and functional activity. An X-ray co-crystal
structure with rosiglitazone boundto PPAR c has shown
that the acidic TZD moiety interacts with Tyr-473 on
helix 12 (AF-2).8 However, the carboxylic acidin the in-
dole SPPARcMs couldbe replacedby a nitrile ( 13) and,
most surprisingly, an acetylenic group (14). Unlike full
agonists, both of these analogs retainedin vitro activity
despite the absence of an acidic group in the molecule
(Table 1).9
11, R = (s)-lactyl
13, R = CH2CN
14, R = CH2CCH
Cmpds 5-10
Cl
F3CO
N
O
CO2H
N
O
12
OCH3
Compds
Y
SPA
Transactivation % Max
IC50, nMa EC50, nMb
activation
Rosiglitazone
1
—
—
250
20
3
100
21
—
—
32
26
—
31
24
33
21
28
24
22
19
—
1
466
2
—
—
—
—
49
195
—
3
3
135
29
4
—
H
5
33
641
Compound 12 was selectedto examine the effects of
these SPPARcMs on adipocyte differentiation. The level
of adipose fatty acid binding protein (aP2) mRNA
expression was measuredin differentiating human pre-
adipocytes.10 SPPARcM 12 was foundto induce only
15% of the maximal level of aP2 mRNA, a marker of
adipogenesis as compared to rosiglitazone treatment.
Attenuatedlipidaccumulation was observedin human
adipocytes upon treatment with 12 versus rosiglitazone
at similar concentrations as judged by oil red O staining
(Fig. 2).11
6
2-nPr
4-Cl
5-Cl
6-nPr
6-Cl
—
7
5
109
8
114
39
3
9
20
1
10
11
12
13
14
15
2
1
—
—
2
42
3
16
31
>3000
—
—
62
>50,000
a IC50 is measured by the complete displacement of a radio-labeled full
agonist indicating competitive binding to the receptor.
b EC50 is the compoundconcentration at which 50% of a given com-
poundÕs intrinsic maximal response has been reached.
Basedupon its in vitro profile, compound 12 was se-
lectedfor evaluation in vivo. This analog was metaboli-
cally stable andno metabolites were proudcedin
detectable quantities in rodents that were inhibitors of
CYP2C9. In db/db mice, SPPARcM 12 afforded
dose dependent glucose lowering comparable to that
However, the use of an N-(1-benzisoxazole, BZX) to mi-
mic the carbonyl group yielded analogs with the desired
activity but with slightly reduced potency in vitro. It was
possible to regain potency comparable to the N-benzoyl
indoles, if appropriate substitution was introduced onto
the aromatic ring at the indole 3-position. The most po-
tent derivatives were those substituted in either the 4- or
6-position of this aromatic ring, while substitution at the
2-position afforded a reduction in binding potency
(compounds 5–10, Table 1). Like the N-benzoyl analogs,
loss of the N-(1-BZX) group occurredin vivo in rats and
the parent indole was detected in plasma, albeit at signifi-
cantly reduced levels (1–3% of parent). In this case, in
vitro studies suggested that loss of the N-(1-BZX) oc-
7
curredvia reductive metabolism at the BZX moiety.
Despite the improvedmetabolic stability of the N-(1-
BZX) indole analogs, given the potent CYP2C9 inhibi-
Figure 2. Visualization of lipidaccumulation in human adipocytes by
oil redO staining.

K. Liu et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2437–2440
2439
Table 2. Comparison of 12 and rosiglitazone in male db/db mice and Sprague–Dawley rats
Compds
Db/db mice
Gluc. Corr. (%)
S–D rata
HWc (g)
Dose (mpk)
Dose (mpk)
BAT (g)
Rosiglitazone
12
10
10
30
—
77
62
78
—
150
100
1.26 0.06b
1027 466b
1.04 0.05
462 45
12
Vehicle
—
1.04 0.08
316 40
a Values are means of six animals/group standard deviation.
b For italics values P < 0.05 comparedto vehicle (Dunnett Õs t-test).
c HW = heart weight.
of rosiglitazone (Table 2) after once daily oral dosing for
11 days.12 Similar plasma exposures were obtainedfor
compound 12 at 30 mg/kg androsiglitazone at 10 mg/
kg (AUC = 660 and250–700 lM h,13 respectively).
Compound 12 was also studied in male Sprague–Daw-
ley rats. After once daily dosing for two weeks, 12
(AUC = 1030 lM h, ꢀ1.6–10· efficacy exposure) re-
sultedin less CH andBAT increases as comparedto that
of rosiglitazone (AUC = 2700 lM h, ꢀ4–10· efficacy
exposure), which was included as an internal positive
control (Table 2).14
The preparation of the indole core (16) for compounds
1–10 (Scheme 1) was accomplishedvia Fischer Õs method
from commercially available starting materials andhas
been described previously.5 N-Arylation of 16 was
achievedby either of two complementary methosd:
methodone utilizeda Pd(0) catalyzedreaction with a
heteroaryl halide15 while methodtwo involveda base
mediated reaction with a 1-chlorobenzisoxazole (1-Cl
BZX)16 (compds 2–10, Scheme 1). Preparation of the
3-benzoyl (11) and3-(1-BZX) ( 12) indole analogs was
accomplishedin the following manner. Sequential treat-
ment of commercially available 3-trifluoromethoxy ani-
line 17 with t-butyl hypochlorite, thiomethyl acetone,
andtriethyl amine in methylene chloride at 0 ꢁC pro-
vided a 3:1 mixture of isomeric 3-thiomethyl indoles.17
The crude mixture was treated with RaNi in ethanol
to afford the desired 6-trifluoromethoxy indole 18 in
50–60% yieldafter separation of the isomers. Reaction
CO2H
i. Het-X, Cs2CO3
O
O
Pd(0), t-Bu3P
or Cs2CO3, 1-ClBZX
of 18 andZnCl with EtMgBr in THF followedby the
2
addition of a benzoyl chloride resulted in the selective
aroylation at the indole 3-position to give 19 and 20 in
75–85% yield.18 The 3-(1-BZX) indole 21 was synthe-
sizedfrom 20 in five steps and10–15% overall yield.
N-Benzylation of the indole nitrogen with the appropri-
ate benzyl bromide was achieved in high yields with
Cs2CO3 in DMF to afford 13 and 14. After N-benzyla-
tion, base hydrolysis provided 11 and 12 (Scheme 1).
The requisite benzyl bromides were readily accessible
from commercially available starting materials.19
X
NH
Het
ii. BBr3
iii. lactate, DEAD, Ph3P
iv. NaOH
X
N
F3CO
F3CO
16, X = CH2 or O
Cmpds 2-10
a. tBuOCl, TEA
S
H
N
F3CO
NH2
F3CO
O
b. RaNi,
17
separate
18
c. EtMgBr, ZnCl2,
aryl chloride
H
N
H
F3CO
In summary, repositioning the 3-benzyl moiety to the in-
dole nitrogen has resulted in a metabolically stable series
of indole SPPARcMs. Interestingly, these compounds
do not require an acidic head group which is integral
for activity with traditional full agonists. This difference
suggests that these SPPARcMs possibly bindto the
receptor in a functionally active conformation distinct
from full agonists. It may be the case that a unique bind-
ing conformation leads to an altered expression of
genes. This may explain the attenuatedadipogenic activ-
ity as comparedto rosiglitazone in human adipocytes.
While these SPPARcMs afforded reductions in glucose
in db/db mice equal to that of rosiglitazone, in rats the
SPPARcMs elicit reduced heart and BAT weight in-
creases relative to rosiglitazone. Gratifyingly, these
SPPARcMs have demonstrated an improvement in the
AEs commonly observedin preclinical species. Further
characterization of these SPPARcMs continues in an ef-
fort to explore the possibility that compounds with these
features may affordbeneficial glycaemic control in hu-
mans with reduced AEs.
d. TsCl
e. Pd (0)
F3CO
N
N
O
O
f. NH2OH
g. Ac2O, pyridine
h. NaOH
Y
H3CO
19, Y = H
H3CO
21
20, Y = allyloxy
i. Cs2CO3,
benzylbromide
j. NaOH
Cl
HO2C
F3CO
RO
F3CO
O
N
N
O
N
O
H3CO
12
11, R = (s)-lactyl
13, R = CH2CN
14, R = CH2CCH
Scheme 1. Indole syntheses.

2440
K. Liu et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2437–2440
Sharma, N.; Tanen, M. R.; Thompson, G. M.; Ventre, J.;
Acknowledgements
Adams, A. D.; Mosley, R.; Surwit, R. S.; Moller, D. E.
Mol. Endocrinol. 2003, 17, 662; (h) Schupp, M.; Janke, J.;
Clasen, R.; Unger, T.; Kintscher, U. Circulation 2004,
109(17), 2054–2057; (i) Minoura, H.; Takeshita, S.; Ita,
M.; Hirosumi, J.; Mabuchi, M.; Kawamura, I.; Nakajima,
S.; Nakayama, O.; Kayakiri, H.; Oku, T.; Ohkubo-
Suzuki, A.; Fukagawa, M.; Kojo, H.; Hanioka, K.;
Yamasaki, N.; Imoto, T.; Kobayashi, Y.; Mutoh, S. Eur.
J. Pharmacol. 2004, 494, 273–281.
We wish to acknowledge the skills and hard work of the
many scientists who provided biological support includ-
ing Thomas T. Doebber, John Ventre, Roger Meurer,
Neelam Sharma, Chhabi Biswas, andRaul Alvaro.
References and notes
5. Acton, J. J.; Black, R. M.; Jones, A. B.; Moller, D. E.;
Colwell, L.; Doebber, T. W.; MacNaul, K. L.; Berger, J.;
Wood, H. B. Bioorg. Med. Chem. Lett. 2005, 15, 357–362.
6. Procedures for the SPA and transactivation assay are
described in: Berger, J. P.; Petro, A. E.; MacNaul, K.;
Kelly, L. E.; Zhang, B. B.; Richards, K.; Elbrecht, A.;
Johnson, B.; Zhou, G.; Doebber, T. W.; Biswas, C.;
Parikh, M.; Sharma, N.; Tanen, M. R.; Thompson, M.;
Ventre, J.; Adams, A. D.; Mosley, R.; Surwit, R. S.;
Moller, D. E. Mol. Endocrinol. 2003, 17, 662, and; Berger,
J.; Liebowitz, M. D.; Doebber, T. W.; Elbrecht, A.;
Zhang, B.; Zhou, G.; Biswas, C.; Cullinan, C. A.; Hayes,
N. S.; Li, Y.; Tanen, M.; Ventre, J.; Wu, M. S.; Berger, G.
D.; Mosley, R.; Marquis, R.; Santini, C.; Sahoo, S. P.;
Tolman, R. L.; Smith, R. G.; Moller, D. E. J. Biol. Chem.
1999, 274, 6718.
1. Reviews on PPARc andinsulin resistance: (a) Olefsky, J.
M.; Saltiel, A. R. Trends Endocrinol. Metab. 2000, 11, 362;
(b) Berger, J.; Wagner, J. A. Diabetes Technol. Therap.
2002, 4, 163; (c) Berger, J.; Moller, D. E. Annu. Rev. Med.
2002, 53, 409; (d) Knouff, C.; Auwerx, J. Endocrine Rev.
2004, 25, 899–918.
2. (a) Plosker, G. L.; Faulds, D. Drugs 1999, 57, 409–438; (b)
Balfour, J. A.; Plosker, G. L. Drugs 1999, 57, 921–930; (c)
Tugwoog, J. D.; Montague, C. T. Hum. Exp. Toxicol.
2002, 21, 429–437; (d) Nesto, R. W.; Bell, D.; Bonow, R.
O.; Fonseca, V.; Grundy, S. M.; Horton, E. S.; Winter, M.
L.; Porte, D.; Semenkovich, C. F.; Smith, S.; Young, L.
H.; Kahn, R. Diabetes Care 2004, 27, 256–263.
3. SPPARM refers to a compoundthat is a partial agonist in
a transcriptional assay andhas edmonstratedunique
molecular andpharmacological properties relative to
traditional agonists. It has been suggested that modulation
of the PPARs may be the preferredstrategy for drug
design. Cock, T.-A.; Houten, S. M.; Auwerx, J. EMBO
Rep. 2004, 5, 142–147.
7. Tschirret-Guth, R. A.; Wood, H. B. Drug Metab. Dispos.
2003, 31, 1000–1004.
8. 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. Nature
1998, 395, 137–143.
4. (a) Reginato, M. J.; Bailey, S. T.; Krakow, S. L.; Minami,
C.; Ishii, S.; Tanaka, H.; Lazar, M. A. J. Biol. Chem. 1998,
273, 32679–32684; (b) Oberfield, J. L.; Collins, J. L.;
Holmes, C. P.; Goreham, D. P.; Cooper, J. P.; Cobb, J. E.;
Lenhard, J. M.; Hull-Ryde, E. A.; Mohr, C. P.; Blan-
chard, S. G.; Parks, D. J.; Moore, L. B.; Lehmann, J. M.;
Plunket, K.; Miller, A. B.; Milburn, M. V.; Kliewer, S. A.;
Willson, T. M. Proc. Natl. Acad. Sci. U.S.A. 1999, 96,
6102; (c) Elbrecht, A.; Chen, Y.; Adams, A.; Berger, J.;
Griffin, P.; Klatt, T.; Zhang, B.; Menke, J.; Zhou, G.;
Smith, R. G.; Moller, D. E. J. Biol. Chem. 1999, 27, 7913;
(d) Shimaya, A.; Kurosaki, E.; Nakano, R.; Hirayama, R.;
Shibasaki, M.; Shikama, H. Metabolism 2000, 49, 411–
417; (e) Rocchi, S.; Picard, F.; Vamecq, J.; Gelman, L.;
Potier, N.; Zeyer, D.; Dubuquoy, L.; Bac, P.; Champy,
M.-F.; Plunket, K. D.; Leesnitzer, L. M.; Blanchard, S.
G.; Desreumaux, P.; Moras, D.; Renaud, J.-P.; Auwerx, J.
Mol. Cell 2001, 8, 743; (f) Thor, M.; Beierlein, K.; Dykes,
G.; Gustovsson, A.-L.; Heidrich, J.; Jendeberg, L.;
Lindqvist, B.; Pegurier, C.; Roussel, P.; Slater, M.;
Svensson, S.; Sydow-Bachmann, M.; Thornstrom, U.;
Uppenberg, J. Bioorg. Med. Chem. Lett. 2002, 12, 3565–
3567; (g) Berger, J. P.; Petro, A. E.; McNaul, K. L.; Kelly,
L. J.; Zhang, B. B.; Richards, K.; Elbrecht, A.; Johnson,
B. A.; Zhou, G.; Doebber, T. W.; Biswas, C.; Parikh, M.;
9. Ligand–receptor co-crystal structures in which the ligand
does not make contact with Tyr-473 have been reported.
See Refs. 4b and 4f.
10. Experimental details Thompson, G. M.; Trainor, D.;
Tanen, M.; Biswas, C.; LaCerte, C.; Berger, J. P.; Kelly, L.
J. Anal. Biochem. 2004, 330, 21.
11. Experimental details described in Ref. 4g.
12. db/db mice were used as described in: Berger, J.; Bailey, P.;
Biswas, C.; Cullinan, C. A.; Doebber, T. W.; Hayes, N. S.;
Saperstein, R.; Smith, R. G.; Liebowitz, M. D. Endocri-
nology 1996, 137, 4189.
13. The AUC reportedis the historical range of values
obtained after dosing at 10 mkd.
14. Male S–D rats were usedas describedin Ref. 5.
15. Hartwig, J. F.; Kawatsura, M.; Hauck, S. I.; Shaughnessy,
K. H.; Alcazar-Roman, L. M. J. Org. Chem. 1999, 64,
5575–5580.
16. Boshagen, H. Chem. Ber. 1967, 100, 3326–3330.
17. Gassman, P. G.; van Bergen, T. J. Org. Synth. 1988, 601–
605, Coll. Vol. 6.
18. Bergman, J.; Venemalm, L. Tetrahedron 1990, 46,
6061.
19. All compounds gave satisfactory 1H NMR andLC/MS
data.