Structure-based design, synthesis, PPAR-γ activation, and molecular docking of N-substituted phthalimides

Article abstract of DOI:10.1007/s00044-017-1867-0

N-substituted phthalimides showed peroxisome proliferator-activated receptors-γ activation in rat liver epithelial Ac2F cells in our previous study. In order to explore better peroxisome proliferator-activated receptors-γ agonists, new N-substituted phthalimide derivatives were designed and synthesized based on a pharmacophore study of natural peroxisome proliferator-activated receptors-γ agonist paecilocin A and synthetic leads. Peroxisome proliferator-activated receptors-γ activation by the new derivatives was evaluated using rat liver epithelial Ac2F cells at a concentration of 10 μM (same as previous study). All the new derivatives showed comparable or better activities than that of rosiglitazone, in which 3-hydroxy-N-(p-methoxy-phenethyl) phthalimide (compound 6) appeared as the best. Molecular docking suggested that the free hydroxyl group on the phthalimide head, a proper hydrophobic tail including a phenyl linker, were beneficial for peroxisome proliferator-activated receptors-γ activation. These N-substituted phthalimide derivatives are valuable as scaffolds for new peroxisome proliferator-activated receptors-γ agonists.

Full text of DOI:10.1007/s00044-017-1867-0

MEDICINAL
CHEMISTRY
RESEARCH
Med Chem Res
DOI 10.1007/s00044-017-1867-0
ORIGINAL RESEARCH
Structure-based design, synthesis, PPAR-γ activation, and
molecular docking of N-substituted phthalimides
1
Shumin Wang1 Zhanfei She1 Qingfeng Cao1 Na Zhao1
●
●
●
●
●
Bin Xiao
Xiangrong Tian2 Yixin Su1
●
Received: 12 October 2015 / Accepted: 4 March 2017
© Springer Science+Business Media New York 2017
●
●
●
Abstract N-substituted phthalimides showed peroxisome
proliferator-activated receptors-γ activation in rat liver epi-
thelial Ac2F cells in our previous study. In order to explore
better peroxisome proliferator-activated receptors-γ ago-
nists, new N-substituted phthalimide derivatives were
designed and synthesized based on a pharmacophore study
of natural peroxisome proliferator-activated receptors-γ
agonist paecilocin A and synthetic leads. Peroxisome
proliferator-activated receptors-γ activation by the new
derivatives was evaluated using rat liver epithelial Ac2F
cells at a concentration of 10 μM (same as previous study).
All the new derivatives showed comparable or better
activities than that of rosiglitazone, in which 3-hydroxy-N-
(p-methoxy-phenethyl) phthalimide (compound 6) appeared
as the best. Molecular docking suggested that the free
hydroxyl group on the phthalimide head, a proper hydro-
phobic tail including a phenyl linker, were beneficial for
peroxisome proliferator-activated receptors-γ activation.
These N-substituted phthalimide derivatives are valuable as
scaffolds for new peroxisome proliferator-activated recep-
tors-γ agonists.
Keywords PPAR-γ Diabetes Phthalimide Luciferase
●
assay Docking simulation
Introduction
Peroxisome proliferator-activated receptors (PPARs) belong
to the nuclear receptor superfamily of ligand-activated
transcription factors, and comprise three members: PPAR-α,
β/δ, and γ (Mangelsdorf et al. 1995; Nuclear Preceptors
Nomenclature Committee 1999; Berger and Moller 2002;
Evans et al. 2004). PPAR-γ is predominantly expressed in
adipose tissue, colon, and macrophages, and play an
important role in the regulation of insulin sensitivity, fatty
acid storage, adipogenesis, and glucose metabolism (Spie-
gelman 1998). Although PPAR-γ agonists such as thiazo-
lidinediones (TZDs, e.g., rosiglitazone (Ros) and
troglitazone) are used to treat type II diabetes mellitus, the
adverse effects including an increased risk of heart attack,
weight gain, edema, and fluid retention remain challenges.
New PPAR-γ ligands with less adverse effects are still in
demand for treatment of type II diabetes mellitus (Willson
et al. 2000).
A ternary complex structure (2PRG) composed of the
PPAR-γ ligand-binding domain (LBD), Ros, and human
steroid receptor co-activating factor-1 was reported, and we
summarized the key pharmacophore concept of Ros
(Fig. 1a) (Nolte et al. 1998; Xiao et al. 2012). In our pre-
vious study, based on a pharmacophore study of paecilocin
A (a natural PPAR-γ agonist paecilocin A, Fig. 2a) and
synthetic PPAR-γ agonists (TZDs and Tyrosine derivatives,
Fig. 2a), we proposed that the 3-hydrophthalic moiety of
paecilocin A function as a hydrophilic head and form
Electronic supplementary material The online version of this article
(doi:10.1007/s00044-017-1867-0) contains supplementary material,
which is available to authorized users.
* Yixin Su
suyixin8888@163.com
1
Ordos School of Clinical Medicine, Inner Mongolia Medical
University, Ordos 017000, China
2
College of Plant Protection, Northwest A&F University, Yangling
712100, China

Med Chem Res
Fig. 1 a Graphical illustration of
the key pharmacophore concept
of Ros. The Ros skeleton was
considered a combination of a
head, linker, and tail. Carbonyl
oxygen and nitrogen in the head
form hydrogen bonds with
residues H323, S289, H449, Q286
,
and Y473, which are crucial for
PPAR-γ activation; central
aromatic ring and oxygen
occupied a hydrophobic pocket
behind Helix 3 as a linker; tail
group occupied a larger pocket
between Helix 3 and β-sheet, in
which the nitrogen forms a key
hydrogen bond with water
molecule H2O604. b
Superposition of hPPAR-γ LBD
binding with Ros (docking
simulation). The protein is
represented as a cartoon (cyan)
and Ros as sticks (purple) (color
figure online)
H-bonds with key amino acid residues of PPAR-γ LBD:
Tyr473, His449, His323, and Ser289 (Xiao et al. 2012).
The easily accessible phthalimide moiety has often been
employed as a pharmacophore in drug development (Pessoa
et al. 2010; Motoshima et al. 2011). In our further study,
phthalimide-derived molecules were designed (Fig. 2b),
produced, and evaluated with respect to PPAR-γ activation in
rat liver Ac2F cells, and these phthalimide derivatives were
investigated as a new class of PPAR-γ ligands (Xiao et al.
2014). Previously we had studied the dose-dependent activity
of the N-substituted phthalimide skeleton (Xiao et al. 2012;
Xiao et al. 2014), in the present study, further derivation of
N-substituted phthalimide was performed and PPAR-γ acti-
vation by the new derivatives was evaluated using rat liver
epithelial Ac2F cells at a concentration of 10 μM.
Material and methods
General
1H and 13C NMR spectra were recorded on a Varian Unity
400 MHz NMR spectrometer, and chemical shifts are
reported with respect to respective residual solvent or
deuterated solvent peaks (δH 3.30 and δC 49.0 for CD3OD,
δH 7.24 and δC 76.8 for CDCl3). FAB MS data was obtained
using a JEOL JMS SX-102A spectrometer. High perfor-
mance liquid chromatography (HPLC) was performed using
a YMC ODS-H80 column (250 × 10 mm, 4 μm, 80 Å) or a
C18-5E Shodex packed column (250 × 10 mm, 5 μm, 100
Å) and a Shodex RI-71 detector.
Fig. 2 a Simplified topology of paecilocin A and of typical synthetic
PPAR-γ agonists. Paecilocin A contains a hydrophilic 3-hydroxy
phthalide moiety and a hydrophobic octyl chain; both TZD and tyr-
osine derivatives employ a phenol moiety as linker, connecting a
hydrophilic head and a hydrophobic tail. b Molecular design of the N-
substituted phthalimide skeleton of PPAR-γ agonists; a 3-hydroxy
phthalimide moiety acts as the head, a phenol moiety as the central
linker, a hydrophobic substituent acts as the tail
solvents used were from commercial source. All synthe-
sized compounds were purified to be 99% pure by a RP-
HPLC using 90% aqueous MeCN as eluent.
General synthetic methods for compounds 1–7 are
shown in Scheme 1. Unless otherwise noted, reagents and

Med Chem Res
13C NMR (CD3OD, 100 MHz): δ 168.5 (C, C=O), 155.9
(C, C–OH), 134.1 (C, C–para phenol), 132.1 (C, C-3a),
129.6 (C, C-5), 129.0 (C, C-4), 122.8 (C, C–meta phenol),
115.1 (C, C–ortho phenol), 39.4 (C, phen–CH2–CH2–),
33.2 (C, phen–CH2–CH2–); FABMS m/z 268 [M + H]+.
O
O
a
H2N
N
O
+
O
O
1
N-(p-Bromo-phenethyl) phthalimide (3) White powder;
1H NMR (CDCl3, 400 MHz): δ 2.93 (t, J = 7.2 Hz, 2H,
phen–CH2–CH2–), 3.87 (t, J = 7.6 Hz, 2H, phen–CH2–
CH2–), 7.11 (d, J = 8.4 Hz, 2H, H–ortho phen–Br), 7.36 (d,
J = 8.4 Hz, 2H, H–meta phen–Br), 7.78 (m, 4H, Ar); 13C
NMR (CDCl3, 100 MHz): δ 168.3 (C, C=O), 137.1 (C,
C–para phen–Br), 134.2 (C, C-3a), 132.2 (C, C-5), 131.8
(C, C–ortho phen–Br), 130.8 (C, C–meta phen–Br), 123.5
(C, C-4), 120.8 (C, C–Br), 39.1 (C, phen–CH2–CH2–), 34.2
(C, phen–CH2–CH2–); FABMS m/z 330 [M + H]+.
O
O
H2N
a
N
O
+
R2
R2
O
O
R1
R1
R1 = H/OH
R2 = OH/Br
2
3
4
5
R1 = H
1 = H
1 = OH R2 = Br
1 = OH R2 = OH
R2 = OH
R2 = Br
R
R
R
O
N
O
N
O
b
OH
3-Hydroxy-N-(p-bromo-phenethyl) phthalimide (4) White
powder; 1H NMR (CDCl3, 400 MHz): δ 2.92 (t, J = 7.6 Hz,
2H, phen–CH2–CH2–), 3.84 (m, 2H, phen–CH2–CH2–),
7.09 (d, J = 8.4 Hz, 2H, H–meta phen–Br), 7.13 (d, J = 8.8
Hz, 1H, H-4), 7.34 (d, J = 7.2 Hz, 1H, H-6), 7.38 (d, J =
8.4 Hz, 2H, H–ortho phen–Br), 7.55 (t, J = 7.6 Hz, 1H, H-
5), 7.57 (s, 1H, OH); 13C NMR (CDCl3, 100 MHz): δ 170.3
(C, β–C=O, –OH), 167.9 (C, γ–C=O, –OH), 154.9 (C,
C–OH), 137.0 (C, C–para phen–Br), 136.6 (C, C-6a), 132.2
(C, C-5), 131.9 (C, C–ortho phen–Br), 130.7 (C, C–meta
phen–Br), 122.9 (C, C–Br), 120.9 (C, C-6), 116.2 (C, C-
3a), 114.7 (C, C-4), 39.0 (C, phen–CH2–CH2–), 34.2 (C,
phen–CH2–CH2–); FABMS m/z 346 [M + H]+.
R
O
OH
OH
5
6 R = OMe
7 R = OEt
Scheme 1 Synthesis of phthalimide derivatives (1–7). Reagents and
conditions: a CH3COOH, 85 °C, overnight; b RI, Ag2O, stir, CH3CN,
reflux for 12 h
General procedure for the synthesis of the compounds 1–5
A mixture of amine (1.2 equiv) and phthalic anhydride in
aqueous glacial acetic acid (1 M) was stirred and heated
under reflux overnight. Products was precipitated by adding
water, filtered, and washed thoroughly with water. Residues
were diluted with MeOH, dried with MgSO4, and evapo-
rated to provide the crude products 1–5 (yield ~90%).
3-Hydroxy-N-(p-hydroxy-phenethyl) phthalimide (5)
1
White powder; H NMR (CD3OD, 400 MHz): δ 2.81 (t,
J = 7.2 Hz, 2H, phen–CH2–CH2–), 3.74 (t, J = 7.6 Hz, 2H,
phen–CH2–CH2–), 6.63 (d, J = 8.4 Hz, 2H, H–ortho
HO–phenethyl), 6.98 (d, J = 8.4 Hz, 2H, H–meta
HO–phenethyl), 7.09 (d, J = 8.0 Hz, 1H, H-4), 7.27 (d, J =
7.2 Hz, 1H, H-5), 7.52 (t, J = 7.6 Hz, 1H, H-6); 13C NMR
(CD3OD, 100 MHz): δ 168.4 (C, β–C=O, –OH), 168.1 (C,
γ–C=O, –OH), 155.9 (C, C-3), 155.2 (C, C–OH phe-
nethyl), 135.8 (C, C-6a), 133.6 (C, C-5), 129.6 (C, C–para
HO–phenethyl), 129.2 (C, C–meta HO–phenethyl), 122.9
(C, C-6), 115.1 (C, C-3a), 115.0 (C, C-4), 114.4 (C,
C–ortho HO–phenethyl), 39.1 (C, phen–CH2–CH2–), 33.3
(C, phen–CH2–CH2–); FABMS m/z 284 [M + H]+.
N-Oleyl phthalimide (1) White powder; 1H NMR (CDCl3,
400 MHz): δ 0.85 (t, J = 7.2 Hz, 3H, CH3), 1.20 (m, 22H,
CH2), 1.63 (m, 2H,=N–CH2–CH2–), 1.98 (m, 4H,=
CH–CH2–), 3.60 (t, J = 7.6 Hz, 2H,=N–CH2–CH2–), 5.30
(m, 2H,=CH–), 7.78 (m, 4H, Ar); 13C NMR (CDCl3, 100
MHz): δ 168.7 (C, C=O), 134.0 (C, C-3a, 7a), 132.4 (C, C-
5, 6), 132.3 (C, C-9′, 10′), 123.3 (C, C-4, 7), 38.3 (C, C-1′),
32.8 (C, C-16′), 32.1 (C, C-14′), 30.0 (C, C-6′), 29.9 (C, C-
13′), 29.9 (C, C-5′), 29.7 (C, C-12′), 29.6 (C, C-7′), 29.5 (C,
C-4′), 29.4 (C, C-15′), 28.8 (C, C-2′), 27.4 (C, C-8′), 27.4
(C, C-11′), 27.1 (C, C-3′), 22.9 (C, C-17′), 14.3 (C, C-18′);
FABMS m/z 398 [M + H]+.
General procedure for the synthesis of the compounds 6–7
N-(p-Hydroxy-phenethyl) phthalimide (2) White powder;
1H NMR (CD3OD, 400 MHz): δ 2.84 (t, J = 7.2 Hz, 2H,
phen–CH2–CH2–), 3.81 (t, J = 7.6 Hz, 2H, phen–
CH2–CH2–), 6.62 (d, J = 8.4 Hz, 2H, H–ortho phenol),
6.98 (d, J = 8.4 Hz, 2H, H–meta phenol), 7.77 (m, 4H, Ar);
To a solution of 3-hydroxy-N-(p-hydroxy-phenethyl)
phthalimide 5 (13 mg, 0.046 mmol) in CH3CN (1.5 ml), RI
(CH3I: 6.0 μl, ca. 0.09 mmol; CH3CH2I: 7.0 μl, ca.
0.09 mmol) and Ag2O (10 mg, 0.04 mmol) were added. The
mixture was then heated under reflux with stirring for 12 h.

Med Chem Res
Solid material was removed by filtration, solvent was
removed by evaporation, and the solid material obtained
was purified by RP HPLC using 90% aqueous MeOH as
eluant to give 6–7 (yield ~75%).
was used to determine the significances of differences
between groups, whereas differences between the mean of
individual groups were assessed using Fisher’s protected
LSD post hoc test. P values of <0.05 were considered
statistically significant.
3-Hydroxy-N-(p-methoxy-phenethyl) phthalimide (6)
1
White powder; H NMR (CD3OD, 400 MHz): δ 2.82 (t,
Molecular docking
J = 7.6 Hz, 2H, phen–CH2–CH2–), 3.76 (m, 2H,
phen–CH2–CH2–), 3.96, (s, 3H, OMe), 6.63 (d, J = 8.4 Hz,
2H, H–ortho HO–phenethyl), 6.98 (d, J = 8.8 Hz, 2H,
H–meta HO–phenethyl), 7.35 (d, J = 7.2 Hz, 1H, H-4), 7.36
(d, J = 8.8 Hz, 1H, H-5), 7.70 (t, J = 7.6 Hz, 1H, H-6); 13C
NMR (CDCl3, 100 MHz): δ 168.1 (C, β–C=O, –OH), 167.2
(C, γ–C=O, –OH), 156.8 (C, C–OMe), 154.4 (C, C–OH),
136.3 (C, C-6a), 136.3 (C, C-5), 134.4 (C, C–para –OMe),
130.5 (C, C–meta –OMe), 130.2 (C, C-6), 117.6 (C, C-3a),
115.7 (C, C-4), 115.6 (C, C–ortho –OMe), 56.5 (C, –OMe),
39.5 (C, phen–CH2–CH2–), 33.8 (C, phen–CH2–CH2–);
FABMS m/z 298 [M + H]+.
Protein coordinates were downloaded from the Protein Data
Bank (accession code: 2PRG) (Berman et al. 2000). Chain
A was prepared for docking within the molecular modeling
software package Chimera 1.5.3 (National Institutes of
Health, Bethesda, MD, USA) (Pettersen et al. 2004). Polar
hydrogen and setting grid box parameters were added using
MGLTools 1.5.4 (The Scripps Research Institute, La Jolla,
CA, USA) (Michel and Sanner 1999; Morris et al. 2009).
Docking calculations were performed using AutoDock Vina
1.1.2 software (The Scripps Research Institute, La Jolla,
CA, USA) (Trott and Olson 2010). Default settings and the
Vina scoring function were applied. For ligand preparation,
Chem3D Ultra 8.0 software (CambridgeSoft Corporation,
Cambridge, MA, USA) was used to convert the 2D struc-
tures of candidates into 3D structural data. The analysis and
visual investigation of ligand-protein interactions of dock-
ing poses were performed using PyMol v1.7.4 (The
PyMOL Molecular Graphics System, Schrodinger LLC,
New York, NY, USA).
3-Hydroxy-N-(p-ethoxy-phenethyl) phthalimide (7) White
1
powder; H NMR (CD3OD, 400 MHz): δ 1.44 (t, J = 7.2
Hz, 3H, O–CH2CH3), 2.82 (t, J = 7.6 Hz, 2H,
phen–CH2–CH2–), 3.76 (m, 2H, phen–CH2–CH2–), 4.24 (q,
J = 6.8, 7.2 Hz, 2H, O–CH2CH3), 6.63 (d, J = 8.4 Hz, 2H,
H–ortho –OEt), 6.98 (d, J = 8.8 Hz, 2H, H–meta –OEt),
7.33 (d, J = 7.2 Hz, 1H, H-4), 7.34 (d, J = 8.8 Hz, 1H, H-5),
7.67 (t, J = 7.6 Hz, 1H, H-6); 13C NMR (CDCl3, 100 MHz):
δ 168.2 (C, β–C=O, –OH), 167.2 (C, γ–C=O, –OH), 156.8
(C, C–OEt), 154.4 (C, C–OH), 136.2 (C, C-6a), 136.2 (C,
C-5), 134.5 (C, C–para –OEt), 130.4 (C, C–meta –OEt),
130.2 (C, C-6), 118.7 (C, C-3a), 115.6 (C, C-4), 115.5 (C,
C–ortho –OMe), 65.2 (C, –O–CH2–CH3), 39.5 (C,
phen–CH2–CH2–), 33.9 (C, phen–CH2–CH2–), 14.8 (C,
–O–CH2–CH3); FABMS m/z 312 [M + H]+.
For validation of docking simulation, Ros was re-docked
into 2PRG complex. Chain A, chain B, and several
important water molecules were tested for generating top-
ranked and reproducible binding modes that are close to
those of the 2PRG crystal, finally chain A and H2O308
,
H2O339, H2O444, and H2O467 were selected for PPAR-γ
agonist docking workflow, and a search grid box was set to
Biological assay
For luciferase assays, plasmids were transfected into Ac2F
cells in a 48-well plate (5 × 104 cells/well) with effector
plasmids and the TK-PPRE × 3-luciferase reporter plas-
mid (1 μg/well) plus pcDNA3(0.1 μg/well) or pFlag-
PPAR-γ1 (0.1 μg/well) using Lipofectamine™ 2000
(Invitrogen Co., USA), according to the manufacturer’s
instructions. After transfection for 4 h, conditioned media
was replaced with complete medium, and cells were
incubated for an additional 20 h. The medium was then
removed, and cells were exposed in serum-free media to
Ros or test compounds for 6 h, washed with PBS and
assayed using the ONE-Glo™ Luciferase Assay System
(Promega, Madison, WI, USA). Luciferase activities were
measured using a GloMax®-Multi Microplate Multimode
Reader (Promega Co., Sunny Vale, CA, USA). ANOVA
Fig. 3 In vitro assay of PPAR-γ activation by phthalimides 1–7 and by
Ros at a concentration of 10 μM in rat liver epithelial Ac2F cells. No
receptor control (without plasmid transfection), negative control
(transfected with a plasmid containing PPRE and pcDNA), Ros. Ros
was used as positive control to monitor luciferase activations.
Compound-treated cells were transiently transfected with PPRE plus
pFlag-PPAR-γ1. Luciferase expressions (relative light units (RLU)/
well) are the means + SDs of five independent results.

Med Chem Res
Fig. 4 The 3D putative binding
modes of Ros and 1–7 with
PPAR-γ LBD. The key amino
acid residues including Tyr473
His449, His343, Ser289, and
Glu286 were labeled and the
,
formed H bonds were shown in
the pictures. a Ros interacts with
key amino acid residues (Tyr473
His449, His343, Ser289, and
Glu286) in the PPAR-γ binding
pocket (−8.2 kcal/mol); b
Binding mode of 1 (−6.3 kcal/
mol); c Binding mode of 2
(−7.6 kcal/mol); d Binding
mode of 3 (−7.8 kcal/mol); e
Binding mode of 4 (−8.3 kcal/
mol); f Binding mode of 5
(−8.4 kcal/mol); g Binding
mode of 6 (−8.5 kcal/mol); h
Binding mode of 7 (−8.6 kcal/
mol)
,
cover the whole surface of the protein to collect all possible
orientations and conformations of the ligand paired with the
protein. Synthesized compounds were compared with Ros
in four aspects; affinity of the top-ranked mode (AT),
average affinity of the nine generated modes (AM), the
number of modes located in the binding pocket (NM), the
average number of H-bond interactions with key amino acid
residues (AI).
Results and discussion
Synthesis
N-substituted phthalimides can be prepared by heating
phthalic anhydride with various N-containing reagents
(Pluempanupat et al. 2007). Compounds 1–5 were gener-
ated by heating phthalic anhydride or 3-hydroxyphthalic

Med Chem Res
Table 1 Docking analysis results for compounds 1–7 and Ros
anhydride with oleyl, p-hydroxy-phenethyl or p-bromo-
phenethyl amine in the presence of acetic acid to afford a
90% yield. Compound 1 was to imitate the skeleton of
lipidic PPAR-γ agonists and paecilocin A (Fig. 2a), while
compounds 2–5 were to imitate the topology of synthetic
PPAR-γ agonists (e.g., Ros, farglitazar). Short alkyl chains
methyl and ethyl were added to the phenyl moiety by
etherification, and the major products obtained were
monoalkylated on the phenol group (6–7) at a yield around
75%. All of their structures were characterized.
Compounds
ATa
AMb
NMc
AId
Log P
1
−6.3
−7.6
−7.8
−8.3
−8.4
−8.5
−8.6
−8.2
−6.178
−7.144
−7.300
−7.189
−7.356
−7.400
−7.356
−7.033
2
5
8
3
4
4
6
4
0.00
0.40
1.00
1.67
1.75
2.00
1.17
1.50
9.54 0.36
4.52 0.36
3.01 0.26
3.78 0.79
2.28 0.75
2.93 0.75
3.46 0.75
2.56 0.79
2
3
4
5
6
7
Ros
a
AT, affinity of the top-ranked mode
PPAR-γ activation
b
c
d
AM, average affinity of the nine generated modes
NM, the number of modes located in the binding pocket
Synthesized compounds were subsequently evaluated for
PPAR-γ activation in rat liver epithelial Ac2F cells tran-
siently transfected with pcDNA3 + PPRE or pFlag-PPAR-
γ1 + PPRE. As shown in Fig. 3, all compounds showed
PPAR-γ activation at the concentration of 10 μM, the
potencies of compounds 1 and 2 were comparable to that of
rosiglitazone, while compounds 3, 4, 6, and 7 showed
greater potencies than that of Ros, and compound 5 showed
moderate potency. Etherization of compound 5 (6 and 7)
while keeping a free hydroxyl group on the phthalimide
moiety significantly improved the PPAR-γ activation of
compound 5. Compound 6, which has a methyl tail,
exhibited most potent activity.
Different Log P values of synthetic analogues could have
led to cell permeability differences in Ac2F cells (Kubinyi
1979; Lipinski et al. 2001). Furthermore, the mechanism of
PPAR-γ activation is complicated and different ligands are
likely to alter the conformation of PPAR-γ protein in dif-
ferent way, which could affect co-activator recruitment
(Itoh et al. 2008; Waku et al. 2009; Waku et al. 2010;
Porcelli et al. 2012). Other factors might also modulate the
bioavailabilities of these molecules in vivo. Therefore,
further intensive optimization and in vivo evaluations of
these molecules is required.
AI, the average number of H-bond interactions with key amino acid
residues
this H-bond network might stabilize PPAR-γ in the con-
formation required for successful co-activator recruitment
and PPAR-γ activation (Nolte et al. 1998; Kuhn et al.
2006). According to Linpinski’s Rule of 5, compound 1
with high Log P value is less druggable. Log P value of
compound 5 is lower than Ros, moderate cell permeability
might lead to its moderate cell-based PPAR-γ activation.
Compared with 5, methylated (6) and ethylated (7)
derivatives significantly increased PPAR-γ activation. Pre-
vious studies have shown that a hydrophobic substituent
provides a good tail for PPAR-γ agonists (Santini et al.
2003). Here, we confirmed that connecting the phenoxyl
moiety with a hydrophobic tail designed to occupy the
hydrophobic binding pocket in PPAR-γ LBD enhanced
binding between 3-hydroxy-N-phenethyl phthalimide (5)
and PPAR-γ LBD. Methyl group activated PPAR-γ better
than ethyl group, Log P values of synthetic analogues (6
and 7) and other factors could have led to PPAR-γ activa-
tion differences in Ac2F cells.
Molecular docking and SAR
Conclusion
According to docking simulation, these compounds bind to
PPAR-γ LBD in the same manner as Ros (Fig. 4). Binding
affinities of compounds 1–3 were lower than Ros while that
of 4–7 were higher than Ros (Table 1), however, com-
pounds 1–3 showed comparable PPAR-γ activation to that
of Ros might because the relatively high Log P values of
compounds 1–3 led to good cell permeability, in which
compound 3 with relatively higher binding affinity but
lower Log P value appeared better than 1 and 2. The 3-
hydroxy-phthalimide head of compounds 4–7 formed
several hydrogen bonds with key amino acid residues
(especially Tyr473 on helix 12) of PPAR-γ LBD (Fig. 4e–h),
In conclusion, we synthesized a new group of N-substituted
phthalimides based on the results of a pharmacophore study
of natural (paecilocin A) and synthetic PPAR-γ agonists as
well as our previous study. The incorporation of proper
hydrophilic and hydrophobic groups into a phthalimide
skeleton yielded novel PPAR-γ agonists, which showed
cell-based activities comparable to that of Ros. A free
hydroxyl group on the phthalimide moiety and hydrophobic
tail groups (especially the methoxyl group) were found to
be beneficial for PPAR-γ activation. These N-substituted
phthalimide derivatives are valuable as scaffolds for new

Med Chem Res
Ligand binding and co-activator assembly of the peroxisome
proliferator-activated receptor-gamma. Nature 395:137–143
PPAR-γ agonists, further optimization and in vivo evalua-
tion of these molecules would provide valuable data.
Nuclear Receptors Nomenclature Committee (1999)
A unified
nomenclature system for the nuclear receptor superfamily. Cell
97:161–163
Compliance with ethical standards
Pessoa C, Ferreira PMP, Lotufo LV, Moraes MO, Cavalcanti SMT,
Coêlho LCD, Hernandes MZ, Leite ACL, DeSimone CA,
Costa VMA (2010) Discovery of phthalimides as immunomo-
dulatory and antitumor drug prototypes. Chem Med Chem
5:523–528
Conflicts of Interest The authors declare that they have no com-
peting interests.
Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM,
Meng EC, Ferrin TE (2004) UCSF Chimera-a visualization sys-
tem for exploratory research and analysis. J Comput Chem
13:1605–1612
References
Berger J, Moller DE (2002) The mechanisms of action of PPARs.
Annu Rev Med 53:409–435
Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H,
Shindyalov IN, Bourne PE (2000) The protein data bank. Nucl
Acids Res 28:235–242
Evans RM, Barish GD, Wang YX (2004) PPARs and the complex
journey to obesity. Nat Med 10:355–361
Itoh T, Fairall L, Amin K, Inaba Y, Szanto A, Balint BL, Nagy L,
Yamamoto K, Schwabe JWR (2008) Structural basis for the
activation of PPARγ by oxidized fatty acids. Nat Struct Mol Biol
15:924–931
Kubinyi H (1979) Nonlinear dependence of biological activity on
hydrophobic character: the bilinear model. Farmaco Sci
34:248–276
Kuhn B, Hilpert H, Benz J, Binggeli A, Grether U, Humm R, Märki
HP, Meyer M, Mohr P (2006) Structure-based design of indole
propionic acids as novel PPARalpha/gamma co-agonists. Bioorg
Med Chem Lett 16:4016–4020
Lipinski CA, Lombardo F, Dominy BW, Feeney PJ (2001) Experi-
mental and computational approaches to estimate solubility and
permeability in drug discovery and development settings. Adv
Drug Deliv Rev 46:3–26
Mangelsdorf DJ, Thummel C, Beato M, Herrlich P, Schötz G, Ume-
sono K, Blumberg B, Kastner P, Mark M, Chambon P (1995) The
nuclear receptor superfamily: the second decade. Cell
83:835–839
Pluempanupat W, Adisakwattana S, Yibchok-Anun S, Chavasiri W
(2007) Synthesis of N-phenylphthalimide derivatives as alpha-
glucosidase inhibitors. Arch Pharm Res 30:1501–1506
Porcelli L, Gilardi L, Laghezza A, Piemontese L, Mitro N, Azzariti A,
Altieri F, Cervoni L, Fracchiolla G, Giudici M, Guerrini U,
Lavecchia A, Montanari R, Di Giovanni C, Paradiso A, Pochetti
G, Simone GM, Tortorella P, Crestani M, Loiodice F (2012)
Synthesis, characterization and biological evaluation of
ureidofibrate-like derivatives endowed with peroxisome
proliferator-activated receptor activity. J Med Chem 55:37–54
Santini C, Berger GD, Han W, Mosley R, MacNaul K, Berger J,
Doebber T, Wu M, Moller DE, Tolman RL, Sahoo SP (2003)
Phenylacetic acid derivatives as hPPAR agonists. Bioorg Med
Chem 13:1277–1280
Spiegelman BM (1998) PPAR-γ: Adipogenic regulator and thiazoli-
dinedione receptor. Diabetes 47:507–514
The PyMOL Molecular Graphics System, Version 1.7.4 Schrödinger,
LLC.
Trott O, Olson AJ (2010) AutoDock Vina: Improving the speed and
accuracy of docking with a new scoring function, efficient opti-
mization, and multithreading. J Comput Chem 31:455–461
Waku T, Shiraki T, Oyama T, Maebara K, Nakamori R, Morikawa K
(2010) The nuclear receptor PPARγ individually responds to
serotonin- and fatty acid-metabolites. EMBO J 29:3395–3407
Waku T, Shiraki T, Oyama T, Morikawa K (2009) Atomic structure of
mutant PPARgamma LBD complexed with 15d-PGJ2: novel
modulation mechanism of PPARgamma/RXRalpha function by
covalently bound ligands. FEBS Lett 583:320–324
Michel F, Sanner (1999) Python: a programming language for soft-
ware integration and development.
17:57–61
J Mol Graphics Mod
Morris GM, Huey R, Lindstrom W, Sanner MF, Belew RK, Goodsell
DS, Olson AJ (2009) Autodock4 and AutoDockTools4: auto-
mated docking with selective receptor flexiblity. J Comput Chem
16:2785–2791
Motoshima K, Ishikawa M, Hashimoto Y, Sugita K (2011) Peroxi-
some proliferator-activated receptor agonists with phenethylphe-
nylphthalimide skeleton derived from thalidomide-related liver X
receptor antagonists. Bioorg Med Chem 19:3156–3172
Nolte RT, Wisely GB, Westin S, Cobb JE, Lambert MH, Kurokawa R,
Rosenfeld MG, Willson TM, Glass CK, Milburn MV (1998)
Willson TM, Brown PJ, Sternbach DD, Henke BR (2000) The PPARs:
From orphan receptors to drug discovery.
J Med Chem
43:527–550
Xiao B, Su MZ, Kim EL, Hong JK, Chung HY, Kim HS, Yin J, Jung
JH (2014) Synthesis of PPAR-γ activators inspired by the marine
natural product, paecilocin A. Mar Drugs 12:926–939
Xiao B, Yin J, Park MH, Liu J, Li JL, Kim EL, Hong JK, Chung HY,
Jung JH (2012) Design and synthesis of marine fungal phthalide
derivatives as PPAR-γ agonists. Bioorg Med Chem
20:4954–4961