Structure-based de novo design, synthesis, and biological evaluation of the indole-based PPARγ ligands (I)

Article abstract of DOI:10.1016/j.bmcl.2006.06.093

MCSS and LeapFrog, two de novo drug design programs, were used for the novel indole-based PPARγ ligands' study. The designed compounds were synthesized and tested for the PPARγ protein binding activities in vitro. Out of the compounds that were synthesize

Full text of DOI:10.1016/j.bmcl.2006.06.093

Bioorganic & Medicinal Chemistry Letters 16 (2006) 5913–5916
Structure-based de novo design, synthesis, and biological
evaluation of the indole-based PPARc ligands (I)
Xiaochun Dong,a Zhenshan Zhang,b Ren Wen,a,* Jianhua Shen,b
Xu Shenb and Hualiang Jiangb
aDepartment of Medicinal Chemistry, Fudan University, Shanghai 200032, China
bShanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai 201203, China
Received 28 April 2006; revised 12 June 2006; accepted 30 June 2006
Abstract—MCSS and LeapFrog, two de novo drug design programs, were used for the novel indole-based PPARc ligands’ study.
The designed compounds were synthesized and tested for the PPARc protein binding activities in vitro. Out of the compounds that
were synthesized, two molecules (compounds 14d and 7d) possessed potent PPARc protein binding activity close to rosiglitazone
in vitro.
ꢀ 2006 Elsevier Ltd. All rights reserved.
The peroxisome proliferator-activated receptor (PPAR)
belongs to the family of nuclear receptors, which play an
important role in regulating the expression of a large
number of genes involved in lipid metabolism and ener-
gy balance.1 Synthetic PPARc agonists for the treatment
of type II diabetes have been proved successful for glu-
cose control, for example, the marketed drugs rosiglitaz-
one (1) and pioglitazone (2). They belong to the class of
thiazolidinedione (TZD) antidiabetic agents (Fig. 1) and
control the blood glucose level in type II diabetes by an
insulin sensitizing mechanism.2,3
Figure 1. Structures of marketed PPARc agonists.
Recently more compounds in clinical and preclinical
research have been reviewed.4 In addition to the TZDs,
other structurally diverse synthetic PPARc agonists
have been identified such as a-alkoxy-b-phenyl propa-
noic acids (ragaglitazar, 35) and tyrosine derivatives
(farglitazar, 46) (Fig. 2). A typical PPARc agonist con-
sists of an acidic head attached to an aromatic scaffold,
a linker, and a hydrophobic tail.
Lately several indole-based compounds have been
reported to be potent PPARc agonists like 5 and 6
(Fig. 2).7,8 It is shown that indole ring perhaps improves
the binding activity for these compounds.
Keywords: De novo drug design; PPARc ligand; Type II diabetes;
Indole compounds.
*
Figure 2. Structures of some PPARc agonists in clinical and preclinical
research.
Corresponding author. Tel.: +86 021 54237560; fax: +86 021
64389178; e-mail: rwen@shmu.edu.cn
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.06.093

5914
X. Dong et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5913–5916
In this letter, we report the structure-based de novo
design9 by MCSS and LeapFrog programs, synthesis,
and biological evaluation of new indole-based PPARc
ligands.
Thus, the suitable position of 3-methyl indole obtained
by MCSS calculations was used along with the JOIN
move in LeapFrog program to generate the novel
ligands. The new ligands were checked for alternative
orientations using FLY move and were completely
minimized using TWIST move to evaluate the binding
energy for the minimum energy conformation of the
ligand. The process was repeated to generate the
different ligands.
The X-ray crystal structure of the human apo-PPARc
ligand-binding domain (LBD) was first reported in
1998.10 Up to now, there are six different co-crystal
structures of PPARc with the synthetic agonists. The
protein model used in our work was constructed accord-
ing to the crystal structure of PPARc–LBD–ragaglitazar
complex from the Brookhaven Protein Data Bank
(PDB), entry 1NYX.11
The MCSS and LeapFrog design result indicated that 3-
(6-benzyloxy-1H-indol-3-yl)-2-acylaminopropionic acid
7a–d were nicely accommodated by PPARc ligand-bind-
ing pocket. Then the advanced docking program
AutoDock 3.013 was used to determine the lowest energy
position in the active site for the above lead molecules.
The calculated binding free energy and LeapFrog score
of each molecule are given in Table 1.
The multiple copy simultaneous search (MCSS) pro-
gram12 was employed to calculate the energetically
favorable position and the orientation of indole nucleus
in the active site of the receptor. The functional group
chosen for the MCSS calculation was trpr (tryptophan
side chain: 3-methyl indole). Replicas of the given func-
tional group were randomly distributed inside the bind-
ing site and then simultaneously and independently
energy-minimized.
Table 2. Biological activity of compounds 14a–e and 7a–e compared
to the marketed compound rosiglitazone
Compounds
RU (10À5mol/L)
KD (mol/L)
14a
7a
9.36
9.20
na
na
14b
7b
16.23
14.16
13.46
20.00
35.29
73.42
11.58
16.69
na
na
na
Table 1. LeapFrog scores and AutoDock calculated binding free
energy for the complexes of the lead compounds
14c
7c
na
Compounds LeapFrog score Binding free energy (Kcal/mol)
na
8.69 · 10À6
6.86 · 10À6
na
7a
7b
7c
7d
À15.13
À21.02
À24.67
À27.09
À7.94
À10.06
À10.58
À11.07
À10.43
14d
7d
14e
7e
na
4.98 · 10À6
Ragaglitazar na
Rosiglitazone
Scheme 1. Reagents and conditions: (a) N3CH2COOC2H5, C2H5ONa, C2H5OH, 85%; (b) xylene, reflux, 80%; (c) 2 N NaOH, EtOH, reflux, 91%;
(d) Cu, quinoline, reflux, 82%; (e) (CH3)2NH, HCHO, CH3COOH, 91%; (f) RNHCH(COOC2H5)2, NaOH, toluene, reflux, 31–62%; (g) 10% NaOH,
reflux, 73–86%; (h) H2O, reflux, 72–83%.

X. Dong et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5913–5916
5915
Figure 3. Superimposition of compound 7d (white) on the PPARc-bound conformation of ragaglitazar (blue). Residues involved in hydrogen
bonding action to 7d are also indicated.
The designed ligands were synthesized as shown in
Scheme 1. 4-Benzyloxybenzaldehyde was condensed
with ethyl azidoacetate in the presence of sodium ethox-
ide to obtain azidocinnamate 8, which was heated in xy-
lene to provide ester 9. The subsequent hydrolysis and
Cu/quinoline mediated decarboxylation afforded the
6-benzyloxyindole 11. By the Mannich reaction,
compound 11 was transformed to 3-(dimethylaminom-
ethyl)-6-benzyloxyindole 12, which acted with 2-acyl-
amino malonic acid diethyl ester in the presence of
sodium hydroxide and toluene to afford 13a–e. The
compounds 13a–e were saponifed under basic condition
to afford diacid 14a–e, and decarboxylated to obtain
target compounds 7a–e.
ring of benzyloxy is situated in the hydrophobic pocket
formed by Leu333, Glu343 and Ser342. Other residues
with hydrophobic action include Phe282, Phe363,
His449, Met364, Tyr327, Ile326 and Leu330.
In summary, MCSS and LeapFrog programs were
used successfully to design novel indole-based PPARc
ligands. Among our designed and synthesized
compounds, in fact two new indole molecules (com-
pounds 7d and 14d) possessed potent PPARc binding
activity close to rosiglitazone in in vitro biological as-
say. Currently, further detailed study and in vivo
pharmacological evaluation of these compounds are
under way.
The newly synthesized 3-(6-benzyloxy-1H-indol-3-yl)-2-
acylaminopropionic acid 7a–e and their synthetic pre-
cursors 14a–e were tested through the receptor/ligand
binding assay in vitro.14 The result shows that (1) the
Response Unit (RU) values of the compounds 14a–c,
14e, 7a–c, and 7e were bellow 20 at 10À5 mol/L concen-
tration, which indicated that their binding activities to
PPARc were weak; (2) the compounds 14d and
7d exerted significant binding activities. As shown in
\Table 2, the KD values of 14d and 7d were close to that
of the marketed drug rosiglitazone.
References and notes
1. Bogacka, I.; Xie, H.; Bray, G. A.; Smith, S. R. Diab. Care
2004, 27, 1660.
2. Momose, Y.; Meguro, K.; Ikeda, H.; Hatanaka, C.; Oi, S.;
Sohda, T. Chem. Pharm. Bull. 1991, 39, 1440.
3. Cantello, B. C. C.; Cawthorne, M. A.; Cottam, G. P.; Du,
P. T.; Haigh, D.; Hindley, R. M.; Lister, C. A.; Smith, S.
A.; Thurlby, P. L. J. Med. Chem. 1994, 37, 3977.
4. Henke, B. R. J. Med. Chem. 2004, 47, 4118.
5. Lohray, B. B.; Lohray, V. B.; Bajji, A. C.; Kalchar, S.;
Poondra, R. R.; Padakanti, S.; Chakrabarti, R.; Vik-
ramadithyan, R. K.; Misra, P.; Juluri, S.; Mamidi, N.
V. S. R.; Rajagopalan, R. J. Med. Chem. 2001, 44,
2675.
6. Henke, B. R.; Blanchard, S. G.; Brackeen, M. F.; Brown,
K. K.; Cobb, J. E.; Collins, J. L.; Harrington, W. W.;
Hashim, M. A.; Hull-Ryde, E. A.; Kaldor, I.; Kliewer, S.
A.; Lake, D. H.; Leesnitzer, L. M.; Lehmann, J. M.;
Lenhard, J. M.; Orband-Miller, L. A.; Miller, J. F.; Mook,
R. A.; Noble, S. A.; Oliver, W.; Parks, D. J.; Plunket, K.
D.; Szewczyk, J. R.; Willson, T. M. J. Med. Chem. 1998,
41, 5020.
As our Docking study prediction, the compound 7d was
nicely accommodated by PPARc-ligand binding pocket.
The calculated binding free energy was À11.07 Kcal/
mol, better than other designed compounds 7a–c. As
seen in Figure 3, the amino acid group of 7d as the
hydrophilic group head is involved in hydrogen bonds
formation with Cys285 and Ser289. The indole heterocy-
cle as the flat aromatic group also forms hydrogen
bonds interaction with Cys285. Arg288 forms H-bond
action with the O atom of the benzyloxy. The benzene

5916
X. Dong et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5913–5916
7. James, F. D.; Akiyama, T.; Einstein, M.; Habulihaz, B.;
Doebber, T.; Berger, J. P.; Meinke, P. T.; Shi, G. Q.
Bioorg. Med. Chem. Lett. 2005, 15, 5035.
11. The ligand-binding pocket of the receptor was defined as
the collection of the amino acids enclosed within a sphere
˚
of 5 A radius around the bound ligand.
8. Henke, B. R.; Adkisom, K. K.; Blanchard, S. G.;
Leesnitzer, L. M.; Mook, R. A.; Plunket, J. K. D.; Ray,
J. A.; Roberson, C.; Unwalla, R.; Willson, T. M. Bioorg.
Med. Chem. Lett. 1999, 9, 3329.
9. All molecular modeling and docking simulations were
performed on a Silicon Graphics Origin3200 workstation
using the SYBYL 6.8 molecular modeling software from
Tripos Inc. and Insight II 2000 molecular modeling
software from Accelrys Inc.
10. Nolte, R. T. G.; Wisely, B.; Westin, S.; Cobb, J. E.;
Lambert, M. H.; Kurokawa, Ri.; Rosenfeld, M. G.;
Willson, T. M.; Glass, C. K.; Milburn, M. V. Nature 1998,
395, 137.
12. MCSS calculations were performed using the CHARMM
22 force field and InsightII/MCSS 2.1program.
13. The docking parameters were set on as follow: not only the
atom types but also the generations and the number of runs
for the LGA algorithm were edited and properly assigned
according to the requirement of the Amber force field. The
number of generation, energy evaluation, and docking runs
was set to 370,000, 1,500,000, and 10, respectively. The
kinds of atomic charges were taken as Kollman-all-atom for
PPARc and Gasteiger–Huckel for ligands.
14. Yu, C. Y.; Chen, L. L.; Luo, H. B.; Chen, J.; Cheng, F.;
Gui, C. S.; Zhang, R. H.; Shen, J. H.; Chen, K. X.; Jiang,
H. L.; Shen, X. Eur. J. Biochem. 2004, 271, 386.