Synthesis and biological activities of novel aryl indole-2-carboxylic acid analogs as PPARγ partial agonists

Article abstract of DOI:10.1016/j.bmcl.2005.08.002

A series of novel aryl indole-2-carboxylic acids has been identified as potent selective PPARγ modulators. Their chemical synthesis and in vitro activities are discussed. Compound 5 was selected for in vivo testing in the db/db mouse model of type 2 diabetes and resulted in reduction of hyperglycemia at comparable plasma exposure when compared to rosiglitazone.

Full text of DOI:10.1016/j.bmcl.2005.08.002

Bioorganic & Medicinal Chemistry Letters 15 (2005) 5035–5038
Synthesis and biological activities of novel aryl
indole-2-carboxylic acid analogs as PPARc partial agonists
James F. Dropinski,* Taro Akiyama, Monica Einstein, Bahanu Habulihaz, Tom Doebber,
Joel P. Berger, Peter T. Meinke and Guo Q. Shi
Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
Received 25 April 2005; revised 29 July 2005; accepted 1 August 2005
Available online 8 September 2005
Abstract—A series of novel aryl indole-2-carboxylic acids has been identified as potent selective PPARc modulators. Their chemical
synthesis and in vitro activities are discussed. Compound 5 was selected for in vivo testing in the db/db mouse model of type 2
diabetes and resulted in reduction of hyperglycemia at comparable plasma exposure when compared to rosiglitazone.
Ó 2005 Elsevier Ltd. All rights reserved.
The peroxisome proliferator-activated receptor
c
lacks useful in vivo activity in our db/db mouse efficacy
model, presumably due to its poor pharmacokinetic
(PK) properties. A medicinal chemistry effort was
initiated based on the interesting biological profile of
compound 1a. The initial goal was to retain the
indole-2-carboxylate core while introducing chemical
diversity in a fashion that would result in unique
compounds with improved PK profiles. To this end,
two major structural changes have been introduced to
compound 1a to create the new scaffolding represented
by compound 1b (Fig. 1). First, the benzyl linkage of
compound 1a was eliminated in an effort to remove a
potentially metabolically labile moiety. Second, the
thioether linkage was replaced with an oxygen linkage
not only to eliminate another potential metabolic liabil-
ity but also to distinguish this class from the indole
compounds patented recently by Bayer.11 These efforts
have led to the identification of a new class of potent
SPPARcMs with improved in vivo efficacy. Herein, we
report the synthesis, structure–activity relationships
(SAR), and in vivo activity of this new class of
compounds.
(PPARc) is a member of the nuclear hormone receptor
superfamily of ligand-dependent transcription factors
and has been shown to be intimately involved in the
regulation of adipogenesis, insulin action, and glucose
homeostasis.1,2 PPARc is also the target protein for
two drugs currently marketed for the treatment of type
2 diabetes, rosiglitazone and pioglitazone.3,4 These
PPARc full agonists modestly ameliorate hyperglyce-
mia. However, undesirable side effects are also associat-
ed with these agents.5,6 The most significant side effects,
including peripheral edema and weight gain, are serious
enough to limit their clinical use in some diabetic
patients. Recent work in the drug discovery community
has focused on identifying selective PPARc modulators
(SPPARcMs). Such ligands are typical partial agonists
in cell-based PPARc transcriptional assays. The clinical
application of the SPPARcMs promises to surpass that
of the currently available PPARc full agonists if their
desirable efficacy profile can be maintained, while their
negative side effects are minimized.7–9
Research in our laboratories was initiated based on
a class of indole-2-carboxylates, which was identified
as a screening lead in the infancy of this project.
Compound 1a is a novel SPPARcM and was highly
selective for PPARc versus the other PPAR isoforms,
namely PPARa and PPARd.10 However, compound 1a
R4
S
O
R1
R2
Cl
COOH
Cl
COOH
R3
N
N
1a
1b
*
Corresponding author. Tel.: +73 25 94 30 13; fax: +73 25 94 95 56;
e-mail: James_Dropinski@merck.com
Figure 1. Screening lead and chosen scaffold.
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.08.002

5036
J. F. Dropinski et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5035–5038
OCH3
COOEt
O
OH
OH
R1
R2
R1
H3CO
H3CO
a, b
OMe
COOEt
COOEt
N
N
N
R2
NH2
H
a
5-position
2
3
c, d
21
20
R4
O
N
OMe
COOEt
b,c
R1
R2
CF3
R1
R2
CF3
e,f, g,
COOH
R3
N
O
O
R1O
HO
R3
d,e
COOH
COOEt
N
N
5-19
4
Scheme 1. Reagents and conditions: (a) Na2CO3, BrCH2CO2Me, 16 h,
80 °C, 90%; (b) NaOEt, Et2O, 2 h, reflux, 90%; (c) excess CH2N2,
CH3OH, rt, 95%; (d) Ar-B(OH)2, Cu(OAc)2, Et3N, 4 A molecular
˚
22
23-29
sieves, CH2Cl2, 16 h, rt, 50–70%; (e) BBr3, CH2Cl2, 1 h, 0 °C, 95%; (f)
Ar-B(OH)2, Cu(OAc)2, Et3N, 4 A sieves, CH2Cl2, 16 h, rt, 50–70%; (g)
˚
Scheme 2. Reagents and conditions: (a) 1.0 BBr3, CH2Cl2, 1 h,
À78 °C, 80%; (b) 3-(trifluoro-methyl)phenylboronic acid, Cu(OAc)2,
NaOH, CH3OH, 16 h, 60 °C, 90%.
˚
Et3N, 4 A molecular sieves, CH2Cl2, 16 h, rt, 65%; (c) BBr3, CH2Cl2,
1 h, 0 °C, 80%; (d) R1-Cl, Cs2CO3, DMF, 1 h, 80 °C, 75–85% or
Ar-B(OH)2, Cu(OAc)2, Et3N, 4 A molecular sieves, CH2Cl2, 16 h, rt,
˚
The synthetic route for the preparation of compounds
5–19 is shown in Scheme 1. Commercially available
anthranilates were converted to the 3-hydroxy-indole
core 3 in two steps.12 Sequential N-arylation and O-
arylation of compound 3 by two consecutive copper-cat-
alyzed coupling reactions using various arylboronic
acids were used as key steps for the introduction of aryl
groups on the top and bottom of the molecule. Since
there was no differentiation of reactivity between the
hydroxy and NH groups, a protection strategy was
developed to allow selective manipulation at each site.
The 3-hydroxyl group was protected as the methyl ether
by treatment with diazomethane (note: use of trimethyl-
silyldiazomethane gave no conversion). Subsequent
N-arylation was readily accomplished by a copper-cata-
lyzed coupling reaction with arylboronic acids to give
compound 4. To our knowledge, this is the first applica-
tion of this copper-mediated coupling reaction to the
arylation on an indole nitrogen.13 The 3-hydroxy group
in 4 was then deprotected and a second copper-mediated
coupling reaction was performed on the hydroxyl group.
The final step was the simple hydrolysis of the ethyl
ester, which gave final products 5–19.
50–60%; (e) NaOH, CH3OH, 16 h, 60 °C, 90%.
namely nitrogen, into the indole core. The synthesis of
this new aza-construct is shown in Scheme 3. Alkylation
of 4-t-butylaniline with ethyl bromoacetate, followed by
cyclocondensation with the appropriate chloropyridines
(31), afforded the hydroxyazaindole core (32). The latter
was converted to compounds 33–39 using either the
previously discussed copper coupling or halide
displacement chemistry.
The newly synthesized compounds were evaluated in the
PPAR SPA binding assay to ascertain c and a binding
profiles. The active analogs were also tested for function-
al activity in a PPAR-GAL4 transactivation (TA) assay,
O
OEt
NH2
HN
OEt
O
R1
N
Cl
31
a
b, c
30
To investigate the effects of chemical diversity at the 5
position of the indole scaffold, compound 20 (obtained
as an intermediate from Scheme 1) was used for further
derivatization. It was found that the 3-methyl ether in 20
could be selectively demethylated at À78 °C in the pres-
ence of the 5-methyl ether, which could then be removed
at 0 °C. The 5-hydroxyl group was either arylated
or alkylated using boronic acid coupling or halide
displacement chemistry, respectively. This route led to
5-position variants that include compounds 23–29 (see
Scheme 2).
R2
OH
O
N
COOEt
d or e, f
COOH
R1
N
N
R1
N
32
33-39
Scheme 3. Reagents and conditions: (a) BrCH2CO2Et, Et3N, DMF
16 h, 50 °C, 100%; (b) neat, 130 °C, 6 h, 50%; (c) NaOEt (21 wt% in
EtOH), Et2O, 2 h, reflux, 40%; (d) Ar-B(OH)2, Cu(OAc)2, 5.0 Et3N,
2
˚
4 A molecular sieves, CH2Cl2, 16 h, rt, 30–50%; (e) R -Cl, Cs2CO3,
A series of azaindoles was prepared to create further
chemical diversity by incorporating a heteroatom,
DMF, 1 h, 80 °C, 75–85%; (f) NaOH, CH3OH, 16 h, 60 °C, 90%.

J. F. Dropinski et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5035–5038
5037
where EC50 values, as well as percent maximal activa-
tion, were measured. One of the early compounds that
emerged from our SAR studies was compound 5, which
gave very potent c binding and transactivation activity
(Table 1) and showed modest in vivo glucose lowering
activity (vide infra). Compound 5 quickly established it-
self as a comparator for future SAR studies. Notably,
para substitution different than tBu at R3 (e.g., 9 and
10) resulted in diminished binding and functional activi-
ties. The combination of a lipophilic trifluoromethoxy
group at R2 and a chlorine at R4 (e.g., 14) led to a similar
erosion of activity when compared to compound 5. Chlo-
ro substitution at R1 or R2 retained the same binding
potency as compound 5, but a 5-fold loss in functional
activity was observed. The pyridyl analogs at R1 (e.g.,
28 and 29) were the most potent compounds prepared
in this series with respect to intrinsic in vitro activity.
Unfortunately, 28 and 29 suffered from poor PK and
low bioavailability (dose normalized AUC = 0.0 and
0.3 lM h, t1/2 = 0.8 and 1.8 h, F = 2.4% and 9.8% for
28 and 29, respectively) and thus were not extensively
studied. The detailed SAR is given in Table 1.
Observations of the potency enhancing features of a pyr-
idyl moiety in the previous set of compounds prompted
the introduction of nitrogen directly into the core. A ser-
ies of azaindoles was prepared and the SAR of analogs
33–39 is shown in Table 2. It was found that chlorine
substitution at R1 led to potency comparable to that
of compound 5. Also, as with the previous group of
compounds, the pyridyl analogs (34 and 39) again
showed the best in vitro potency. However, in the
db/db mouse model, compounds 33, 35, and 39 at
30 mpk showed no efficacy and subsequent exposure
measurements revealed that drug levels were below the
limits of detection.
Since compound 5 appeared to possess the most inter-
esting biological profile, as well as desirable pharmaco-
kinetic properties (dose normalized AUC = 9.19 lM h,
Table 1. In vitro activities of compounds 5–19 and 23–29
Compound
R1
R2
R3
R4
hPPAR c/a SPA binding
IC50 (lM)a
hPPAR c TA EC50 (lM)b
(%max at 3 lM)
Rosiglitazone
0.210
0.022 (168%)
0.055 (31%)
0.013 (31%)
(13%)*
1a
5
0.026/>15
0.012/10.575
0.049/3.577
0.049/>50
0.078/>50
0.115/>50
0.107/3.197
0.060/>15
0.120/>15
0.102/>50
0.298/>15
0.027/>15
0.019/>15
0.017/>15
0.083/>15
0.013/4.482
0.039/>15
0.318/>50
0.018/6.271
0.018/4.154
0.011/9.024
0.005/11.475
0.002/>15
H
H
p-tBu
p-tBu
p-tBu
p-tBu
p-OMe
p-CF3
m-iPr
m-iPr
m-iPr
p-tBu
p-tBu
p-tBu
p-tBu
p-tBu
p-tBu
p-tBu
p-tBu
p-tBu
p-tBu
p-tBu
p-tBu
p-tBu
m-CF3
p-OCF3
m-OEt
m-OH
m-CF3
m-CF3
m-CF3
p-CF3
6
H
H
7
H
H
0.233 (30%)
0.161 (22%)
0.334 (25%)
NT
8
H
H
9
H
H
10
11
12
13
14
15
16
17
18
19
23
24
25
26
27
28
29
H
H
H
H
NT
NT
H
H
H
H
m-OH
p-Cl
NT
H
OCF3
Cl
Cl
H
0.411 (22%)
0.106 (27%)
NT
H
m-CF3
m-iPr
H
Cl
m-CF3
3,5 di-CF3
3-CF3, 4-Cl
m-CF3
m-CF3
m-CF3
m-CF3
m-CF3
m-CF3
m-CF3
0.136 (37%)
(53%)*
(45%)*
(41%)*
NT
Cl
Cl
H
H
CH3
H
H
Et
H
H
0.034 (21%)
0.091 (34%)
0.126 (32%)
0.006 (42%)
0.002 (20%)
iBu
H
pyr-2-yl
4-CF3-pyr-2-yl
4-Cl-pyr-2-yl
H
H
H
Note: No d activity <50 lM and thus not shown here.
a Binding affinities were measured using radioligands following published procedures.14
b Agonist activities were measured in a PPAR-GAL4 construct following published procedures.14 The EC50 refers to the concentration which yields a
50% response relative to the standard; partial agonism percentages are given in parentheses.
* No EC50 was obtained; no plateau reached in titration; maximal activity only reported. NT = not tested.
Table 2. In vitro activity of compounds 33–39
Compound
R1
R2
hPPAR c/a IC50 (lM)
hPPAR c TA EC50 (lM) (%max at 3 lM)
33
34
35
36
37
38
39
H
m-CF3
0.075/>50
0.251/>50
0.021/>15
0.100/>15
0.104/>15
0.202/>50
0.058/>50
0.464 (30%)
0.034 (56%)
0.176 (23%)
0.299 (31%)
0.755 (19%)
(29%)*
H
Cl
4-CF3-pyr-2-yl
m-CF3
m-CF3
OCH3
Cl
Cl
m-OCH3
m-SO2CH3
4-CF3-pyr-2-yl
Cl
0.095 (7%)
* No EC50 was obtained; no plateau reached in titration; maximal activity only reported. NT = not tested.

5038
J. F. Dropinski et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5035–5038
700
600
500
400
300
200
100
0
700
600
500
400
300
200
100
0
-4
-2
0
2
4
6
8
10
Days on Test Compound
db Control
Compound 5 30 mpk
Rosiglitazone 10 mpk
Lean control
Figure 2. Effects of test compounds on plasma glucose levels in db/db mice. Male db/db mice (11–13 weeks old) and lean mice were dosed daily for 11
days by gavage with vehicle or the indicated doses of test compounds. Plasma glucose levels were measured before each dosing at days 4, 7, and 11.
Each data point represents the mean value ( SD) of seven individual mice.
t1/2 = 0.5 h), our efforts were directed back to further
profile compound 5 in our in vivo model.
References and notes
1. Moller, D. E. Nature 2001, 414, 821.
2. Berger, J.; Moller, D. E. Annu. Rev. Med. 2002, 53, 409.
3. Momose, Y.; Meguro, K.; Ikeda, H.; Hatanka, C.; Oi, S.;
Sodha, T. Chem. Pharm. Bull. 1991, 39, 1440.
4. Cantello, B. C.; Cawthorne, M. A.; Haigh, D.; Hindley, R.
M.; Smith, S. A.; Thurlby, P. J. Med. Chem. 1994, 37,
3977.
5. Hanefeld, M.; Belcher, G. Int. J. Clin. Pract. 2001, 27.
6. Sorbera, L. A.; Rabbasseda, X.; Castaner, J. Drugs Future
1998, 23, 977.
7. Berger, J. P.; Petro, A. E.; MacNaul, K. L.; Kelly, L. J.;
Zhang, B. B.; Richards, K.; Elbrecht, A.; Johnson, B. A.;
Zhou, G.; Doebber, T. W.; Biswas, C.; Parikh, M.;
Sharma, N.; Tanen, M. R.; Thompson, G. M.; Ventre, J.;
Adams, A. D.; Mosely, R.; Surwit, R. S.; Moller, D. E.
Mol. Endocrinol. 2003, 17, 662.
8. 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.
9. Berger, J. P.; Akiyam, P. M. T. Trends Pharmacol. Sci.
2005, 26, 244.
10. Berger, J.; Doebber, T.; Leibowitz, M.; Moller, D.;
Mosely, R.; Tolman, R.; Ventre, J.; Zhang, B.; Zhou,
G. WO 01/30343, 2001; Chem. Abstr. 2001, 134,
320871.
11. Stolle, A.; Dumas, J.; Carley, W.; Coish, P.; Magnu-
son, S.; Wang, Y.; Nagarathnam, D.; Lowe, D.; Su,
N.; Bullock, W.; Campbell, A.; Qi, N.; Baryza, J.;
Cook, J. WO 02/30895, 2002; Chem. Abstr. 2002, 1364,
309846.
Compound 5 was evaluated in the db/db mouse model,
using rosiglitazone as the comparator. Compound 5
was shown to effectively lower glucose by 54% at
30 mpk in the 11-day study with 5% body weight increase
vs the vehicle group. Rosiglitazone exhibited a lowering
of 70% glucose at 10 mpk with 8% body weight increase.
Pharmacokinetic measurements after the termination of
the study showed that compound 5 gave a plasma expo-
sure of 275 lM h. The exposure of rosiglitazone was not
measured in this assay, but it typically gave an exposure
of 300 lM h, derived from averaging values from six pre-
vious assays. Thus, compound 5 demonstrated about
80% of the glucose-lowering efficacy of rosiglitazone at
comparable plasma exposure (see Fig. 2).
In summary, chemical modifications of the indole-2-car-
boxylate core led to novel and intrinsically potent
SPPARcMs. The synthetic highlight was the extension
of copper-catalyzed boronic acid coupling to include
the indole nitrogen. This arylation strategy enabled
rapid analoging and introduction of chemical diversity
to evaluate this class better. The SAR was expanded
to include insertion of a heterocycle into the core result-
ing in an azaindole core; however, these compounds
were submitted for in vivo evaluation and showed no
measurable plasma exposure. Compound 5 was evaluat-
ed in the db/db mouse model and resulted in comparable
lowering of glucose at a similar plasma concentration as
rosiglitazone.
12. Bos, M.; Jenk, F.; Martin, J. R.; Moreau, J. L.; Mutel, V.;
Sleight, A. J.; Widmer, U. Eur. J. Med. Chem. 1997, 32,
253.
13. Evans, D. A.; Katz, J. L.; West, T. R. Tetrahedron Lett.
1998, 39, 2937.
Acknowledgments
14. Berger, J.; Leibowitz, M. D.; Doebber, T. W.; Elbrecht,
A.; Zhang, B.; Zhou, G.; Biswas, C.; Cullinan, C. A.;
Hayes, N. S.; Li, Y.; Tanen, M.; Ventre, J.; Wu, S. M.;
Berger, G. D.; Mosely, R.; Marquis, R.; Santini, C.;
Sahoo, S. P.; Tolman, R. L.; Smith, R. G.; Moller, D. E.
J. Biol. Chem. 1999, 274, 6718.
The authors greatly acknowledge all our colleagues
in the PPAR program for their technical support in
the evaluation of the compounds presented in this
manuscript.