Discovery of Thiazolidine-2,4-Dione/Biphenylcarbonitrile Hybrid as Dual PPAR α/γ Modulator with Antidiabetic Effect: In vitro, In Silico and In Vivo Approaches

Article abstract of DOI:10.1111/cbdd.12102

A small series of thiazolidine-2,4-dione and barbituric acid derivatives 1-4 was prepared using a short synthetic route, and all compounds were characterized by elemental analysis, mass spectrometry, and NMR (¹H, ¹³C) spectroscopy. T

Full text of DOI:10.1111/cbdd.12102

Chem Biol Drug Des 2013; 81: 474–483
Research Article
Discovery of Thiazolidine-2,4-Dione/
Biphenylcarbonitrile Hybrid as Dual PPAR a/c
Modulator with Antidiabetic Effect: In vitro, In Silico
and In Vivo Approaches†
Sergio Hidalgo-Figueroa1, Juan J.
els, compound 1 may bind into the active site of both
1
1
isoforms showing important short contacts with the
peroxisome proliferator-activated receptor c residues:
Tyr 473, His 449, Ser 289, His 323; and peroxisome
proliferator-activated receptor a residues: Tyr 464, His
440, Ser 280 and Tyr 314.
ꢀ
Ramırez-Espinosa , Samuel Estrada-Soto ,
2
2
ꢀ
ꢀ
ꢀ
Julio C. Almanza-Perez , Ruben Roman-Ramos ,
Francisco J. Alarcon-Aguilar , Jesus V.
2
ꢀ
ꢀ
3
ꢀ
Hernandez-Rosado , Hermenegilda
4
4
~
ꢀ
ꢀ
Moreno-Dıaz , Daniel Dıaz-Coutino and
1,
ꢀ
Gabriel Navarrete-Vazquez *
Key words: 2,4-thiazolidinedione, diabetes, dual agonist,
hydrogen bonds, molecular docking, PPAR
1
ꢀ
Facultad de Farmacia, Universidad Autonoma del Estado
de Morelos, Av. Universidad 1001, Chamilpa, Cuernavaca
Received 8 November 2012, revised 6 December 2012 and
accepted for publication 10 December 2012
ꢀ
Mor., 62209, Mexico
Laboratorio de Farmacologıa, Depto. Ciencias de la
2
ꢀ
ꢀ
Salud, D.C.B.S., Universidad Autonoma Metropolitana-
ꢀ
Iztapalapa, Apdo.-Postal 55–535, CP 09340, Mexico, D.F.,
Peroxisome proliferator-activated receptors (PPARs) are
ligand-activated transcription factors and members of the
nuclear hormone receptor superfamily, and exist in three
isoforms, PPARa, PPARc, and PPARd. Each of these
regulates tissue-specific target genes involved in many bio-
logical processes such as the solute carrier family 2 (facili-
tated glucose transporter), member 4 (GLUT4) (1). PPARa
and PPARc are the molecular targets of a number of mar-
keted drugs, such as the fibrate class of hypolipidemic
drugs and the thiazolidinedione class of insulin-sensitizing
drugs, respectively (2,3).
ꢀ
Mexico
Posgrado en Biologıa Experimental, D.C.B.S., Universidad
3
ꢀ
ꢀ
Autonoma Metropolitana-Iztapalapa, Apdo.-Postal 55–535,
ꢀ
ꢀ
CP 09340, Mexico, D.F., Mexico
4Universidad del Papaloapan, Campus Tuxtepec, Circuito
Central 200, Col. Parque Industrial, 68301, Tuxtepec,
ꢀ
Oaxaca, Mexico
*Corresponding author: Gabriel Navarrete-Vazquez, gabriel_
ꢀ
navarrete@uaem.mx
†This work was taken in part from the PhD thesis of S.
Hidalgo-Figueroa. He is a CONACyT fellow (209159).
Vladimir Hernandez-Rosado is also a CONACYT fellow. We
ꢀ
are grateful to Korina Sanchez-Salazar for technical
assistance.
Cardiovascular disorders are the most common cause of
morbidity and mortality in diabetes as well as dyslipidemic
conditions, and their prevalence is increasing constantly in
developing countries (4). Diabetes mellitus is a chronic and
progressive disease of metabolic deregulation character-
ized by insulin resistance in peripheral tissues (liver, mus-
cle, and adipose) and impaired insulin secretion by the
pancreas (5).
A small series of thiazolidine-2,4-dione and barbituric
acid derivatives 1–4 was prepared using
a short
synthetic route, and all compounds were characterized
by elemental analysis, mass spectrometry, and NMR
(1H, 13C) spectroscopy. Their in vitro relative expres-
sion of peroxisome proliferator-activated receptor a
and peroxisome proliferator-activated receptor c was
evaluated. Compound 1 showed an increase in the
mRNA expression of both peroxisome proliferator-acti-
vated receptor isoforms, as well as the GLUT-4 levels.
The antidiabetic activity of compound 1 was deter-
mined at 50 mg/kg single dose using a non-insulin-
dependent diabetes mellitus rat model. The results
indicated a significant decrease in plasma glucose lev-
els. Additionally, we performed a molecular docking of
compound 1 into the ligand binding pocket of peroxi-
some proliferator-activated receptor a and peroxisome
proliferator-activated receptor c. In these binding mod-
Synchronized therapies, which concurrently control diabe-
tes and inhibit progression of cardiovascular complica-
tions, may be a fascinating therapeutic option in the
treatment of diabetes. PPAR a/c dual agonists are new
class of drugs, which have been developed to target both
PPARs simultaneously to produce both antidiabetic and
hypolipidemic effects (6). The recently investigated PPAR
a/c dual agonist tesaglitazar (Figure 1) is noted to reduce
triglycerides, raise cardioprotective HDL levels and conse-
quently improve insulin sensitivity (7,8).
474
ª 2013 John Wiley & Sons A/S. doi: 10.1111/cbdd.12102

Discovery of Dual PPAR Modulator with Antidiabetic Effect
Varian Inova 400 instrument. Chemical shifts (dH, dC) and
coupling constants values (J) are given in ppm and Hz,
respectively. Standard reference was used: TMS (dH = 0,
dC = 0) in DMSO-d6. The following abbreviations are used:
s, singlet; d, doublet; q, quartet; t, triplet; bs, broad signal.
IR spectra have been recorded on a Bruker Vector 22 FT
spectrophotometer (Bruker Instruments, Billerica, MA,
USA). Mass spectra were recorded on a Jeol JMS-700
equipment (JEOL USA Inc., Peabody, MA, USA). Elemen-
tal analyses were carried out on an Elementar Vario ELIII
instrument (Elementar, Hanau, Germany).
General method of synthesis of compounds 1–4
A mixture of 4′-[(4-formylphenoxy)methyl]biphenyl-2-carbo-
nitrile or ethyl (4- formylphenoxy)acetate (0.0006 mol), 2,
4-thiazolidinedione or barbituric acid (1 equiv.), benzoic
acid (30% mol), and piperidine (30% mol) in 2 mL of tolu-
ene was refluxed with continuous removal of water formed
during the reaction using a Dean-Stark apparatus for
3–4 h. The reaction mixture was cooled, and the yellow
solid was filtered off and dried to afford the corresponding
compound.
Figure 1: Pharmacophoric pattern in peroxisome proliferator-
activated receptor (PPAR) a/c dual agonist tesaglitazar and the
requirements completed by compounds 1–4.
4′-({4-[(Z)-(2,4-dioxo-1,3-thiazolidin-5-ylidene)
methyl]phenoxy}methyl)-1,1′-biphenyl-2-
The classic pharmacophoric features of PPAR a/c dual
agonists essentially consists of three parts (9): (i) an acidic
head group, (ii) central aromatic region and, (iii) a lipophilic
side chain (Figure 1).
carbonitrile (1)
Yield: 80%, mp: 235.6–236.8 °C. 1H NMR (400 MHz,
DMSO-d6) d: 5.25 (2H, s, CH2), 7.18 (2H, d, Jo = 9.2 Hz,
H-2′, H-6′), 7.53–7.64 (5H, m, H-4, H-3′, H-5, H-3″, H-5″),
7.59 (1H, s, NH), 7.71 (1H, C=CH), 7.74–7.78 (4H, m,
H-5,H-6, H-2″, H-6″), 7.95 (1H, dd, Jo = 8.0, Jm = 1.2 Hz,
H-3, ppm; 13C NMR (100 MHz, DMSO-d6) d: 69.4 (CH2),
110.6 (C-2), 116.1 (C-2′, C-6′), 119.1 (C=C), 121.4 (C-4′),
126.3 (C-6), 128.6(C-3″, C-5″), 128.7 (CN), 129.3(C-3′,
C5′), 130.6 (C-5), 131.8 (C-4″) 132.5 (C-2″, C-6″), 134.0
(C-4), 134.3 (C-3), 137.5 (C-1″), 137.9 (C-1), 144.6
(HC=C), 160.4 (C-1″), 168.5 (O=C-S), 168.7 (O=C-N) ppm;
MS/FAB+: m/z 413 (M + H+). Anal. Calcd for
C24H16N2O3S: C, 69.89; H, 3.91; N, 6.79; S, 7.77. Found:
C, 69.92; H, 3.92; N, 6.81; S, 7.79.
The 2,4-thiazolidinedione ring in the most of glitazones is
reported to make three central H-bonds with His323,
Tyr473, and His449 that are important for the activation of
the PPARc. Several replacements of this heterocycle have
been carried out using a variety of isosteres such as free
carboxyl group, oxazolidinedione, tetrazole, and barbituric
acid (9–11).
In our ongoing research on PPAR a/c dual agonists deriva-
tives with cardioprotective and antidiabetic activities, we
report in this article the preparation of derivatives 1–4, their
in vitro relative expression of PPARa, PPARc, and GLUT4,
molecular docking of the most active compound in both
PPAR isoforms, as well as its in vivo hypoglycemic effect.
4′-({4-[(2,4,6-trioxotetrahydropyrimidin-5(2H)-
ylidene)methyl]phenoxy}methyl)biphenyl-2-
carbonitrile (2)
1
Yield: 73%, mp: 267 °C (dec). H NMR (400 MHz, DMSO-
Methods and Materials
d6) d: 5.32 (2H, s, CH2), 6.82 (2H, d, Jo = 9.2 Hz, H-2′,
H-6′), 7.17 (dd, 1H, Jo = 8.8, Jm = 2 Hz, H-4)7.94 (dd,
1H, Jo = 7.6, Jm = 1.6 Hz, H-6), 7.76–7.80 (m, 2H, H-3,
H-5), 7.57–7.64 (m, 4H, H-2, H-6″, H-3″, H-5″), 8.24 (1H,
s, C=CH), 8.36 (2H, dd, Jo = 9.2 Hz, H-3′, H-5′), 10.0
(2H, 2NH) ppm; 13C NMR (100 MHz, DMSO-d6) d: 74.4
(CH2), 115.3 (C-2), 115.4 (C-2′, C-6′), 119.9 (C=C), 120.9
(C-4′), 123.8 (C-6), 130.6 (C-3″, C-5″), 133.1 (C-5), 133.4
(C-3′, C-5′), 133.5 (CN), 133.9 (C-2″, C-6″), 135.3 (C-4),
138.8 (C-3), 139.7 (C-4″), 142.6 (C-1″). 149.3 (C=O),
Chemistry
All starting materials and reagents were obtained commer-
cially and were used as received. Melting points were
determined on an EZ-Melt MPA120 automated melting
point apparatus from Stanford Research Systems and are
uncorrected. TLC monitored reactions on 0.2 mm pre-
coated silica gel 60 F254 plates (E. Merck KGaA, Darm-
stadt, Germany). NMR studies were carried out with a
Chem Biol Drug Des 2013; 81: 474–483
475

Hidalgo-Figueroa et al.
155.4 (HC=C), 156.8 (C-1), 167.4 (N-C=O), 167.6
(N-C=O), 169.1 (O-C=O) ppm; MS/FAB+: m/z 424
(M + H+). Anal. Calcd for C25H17N3O4: C, 70.91; H, 4.05;
N, 9.92. Found: C, 70.88; H, 4.08; N, 9.97.
130.2 (C-5), 130.6 (C-6), 132.3 (C-2″, C-6″), 133.9 (C-4),
134.3 (C-3), 137.3 (C-4′), 137.9 (C-1″), 144.5 (C-1),
163.6 (C-1′), 191.7 (H-C=O) ppm; MS/FAB+: m/z 314
(M + H+).
Ethyl {4-[(Z)-(2,4-dioxo-1,3-thiazolidin-5-ylidene)
methyl]phenoxy}acetate (3)
Synthesis of Ethyl (4-formylphenoxy)acetate (9)
A
mixture
of
and
4-hydroxybenzaldehyde
potassium carbonate
(0.100 g,
(0.169 g,
Yield: 75%, mp: 191.9–193.7 °C. 1H NMR (400 MHz,
DMSO-d6) d: 1.19 (3H, t, CH3), 4.15 (2H, q, CH2), 4.85 (2H,
s, CH2), 7.07 (1H, NH), 7.07 (2H, dd, Jo = 8.8 Hz,
Jm = 1 Hz, H-2′, H-6′), 7.53 (2H, dd, Jo = 8.8 Hz,
Jm = 1 Hz, H-3′, H-5′), 7.72 (1H, s, CH) ppm; 13C NMR
(100 MHz, DMSO-d6) d: 14.5 (CH3), 61.2 (CH2), 65.1 (CH2),
115.9 (C-2, C-6), 121.2 (C=C), 126.6 (C-4), 132.1 (HC=C),
132.4 (C-3, C-5), 159.6 (C-1), 167.9 (N-C=O), 168.4 (O-
C=O), 168.8 (S-C=O) ppm; MS/FAB+: m/z 308 (M + H+).
Anal. Calcd for C14H13NO5S: C, 54.71; H, 4.26; N, 4.56; S,
10.43. Found: C, 54.73; H, 4.25; N, 4.52; S, 10.39.
0.0008 mol)
0.0012 mol, 1.5 equiv.) was dissolved in acetone (10 mL)
and was stirred at room temperature. After 30 min, ethyl
bromoacetate (0.138 g, 0.0008 mol, 1.01 equiv.) was
added dropwise and stirred for 4 h. After the completion
of the reaction, the solvent was removed under reduced
pressure, and the oily residue washed with water to
extract the formed KBr. The oily product was then purified
by column chromatography.
Yield: 65%, oil. 1H NMR (400 MHz, DMSO-d6) d: 1.62
(3H, t, CH3), 3.33 (2H, q, CH2), 5.10 (2H, s, CH2), 6.85
(2H, dd, Jo = 9.2 Hz, H-2′, H-6′), 8.38 (2H, dd,
Jo = 9.2 Hz, Jm = 1 Hz, H-3′, H-5′), 9.99 (1H, s, O=C-H)
ppm; 13C NMR (100 MHz, DMSO-d6) d: 22.5 (CH3), 69.1
(CH2), 69.6 (CH2), 115.1 (C-2, C-6), 130.5 (C-4), 134.2
(C-3, C-5), 162.8 (C-1), 164.3 (O-C=O), 206.9 (H-C=O)
ppm; MS/FAB+: m/z 209 (M + H+).
Ethyl {4-[(2,4,6-trioxotetrahydropyrimidin-
5(2H)-ylidene)methyl]phenoxy}acetate (4)
Yield: 78%, mp: 210.1–212.5 °C. 1H NMR (400 MHz,
DMSO-d6) d: 1.19 (3H, t, CH3), 4.15 (2H, q, CH2), 4.64
(2H, s, CH2), 6.69 (2H, dd, Jo = 8.8 Hz, Jm = 1 Hz, H-2′,
H-6′), 8.23 (1H, S, CH), 8.31 (2H, dd, Jo = 8.8 Hz,
Jm = 1 Hz, H-3′, H-5′), 11.26 (2H, 2NH) ppm; 13C NMR
(100 MHz, DMSO-d6) d: 14.4 (CH3), 61.3 (CH2), 65.2
(CH2), 113.9 (C-2, C-6), 116.5 (C=C), 126.2 (C-4), 137.6
(C-3, C-5), 150.7 (C=O), 152.1 (HC=C), 161.9 (C-1), 162.6
(N-C=O), 164.3 (N-C=O), 169.3 (O-C=O) ppm; MS/FAB+:
m/z 319 (M + H+). Anal. Calcd for C15H14N2O6: C, 56.60;
H, 4.43; N, 8.80. Found: C, 56.78; H, 4.43; N, 8.87.
Biological evaluation
In vitro PPAR assay
3T3-L1 fibroblasts (9 9 105 cells per well) were cultured in
6-well plates (Corning Incorporated, Corning, Corning, NY,
USA) in Dulbecco’s modified Eagle’s medium supple-
mented with 25 mM glucose, 10% fetal bovine serum (v/v),
1 mM sodium pyruvate, 2 mM glutamine, 0.1 mM non-
essential amino acids, and gentamicin, in a 5% CO2
humidified atmosphere, at 37 °C. After 2 days of conflu-
ence, the cells were differentiated to the adipocyte pheno-
type with 0.5 mM 3-isobutyl-1-methylxanthine, 0.25 lM
dexamethasone acetate, and 0.8 lM insulin, for 48 h, fol-
lowed by insulin alone for 48 h more. The culture medium
without insulin was changed every 2 days during 8 days
of differentiation (12). To determine the effect of com-
pounds on PPARs and GLUT-4 expression, the cells were
treated by 24 h.
Synthesis of 4′-[(4-formylphenoxy)methyl]-1,1′-
biphenyl-2-carbonitrile (5)
A
mixture
of
and
4-hydroxybenzaldehyde
potassium carbonate
(0.100 g,
(0.169 g,
0.0008 mol)
0.0012 mol, 1.5 equiv.) was dissolved in acetone (10 mL)
and was stirred at room temperature. After 30 min, 4′-bro-
momethylbiphenyl-2-carbonitrile (0.22 g, 0.0008 mol, 1.01
equiv.) was added in small portions and stirred for 2 h.
After the completion of the reaction, the solvent was
removed under reduced pressure, and the resulting solid
washed with water to extract the formed KBr. The crude
solid product was then recrystallized from acetone.
RNA was isolated from cultured cells using a TriPure isola-
tion reagent (Invitrogen, Paisley, UK). Absorbance was
measured at 260 and 280 nm for each RNA sample, and
the absorbance ratio (260–280 nm) was 1.9 Æ 0.2. To
confirm RNA integrity, 1 lg was run in 1% agarose gel.
RNA was stained with ethidium bromide and visualized
using an Image Gel-Logic 212 Pro (Kodak/Caaresream,
Rochester, New York, USA). Two major ribosomal bands
(28S and 18S rRNA) but no degraded RNA were detected
(data not shown). Two micrograms of total RNA were
reverse-transcripted using the ImProm II reverse transcrip-
Yield: 77%, mp: 137.1–139.6 °C. 1H NMR (400 MHz,
DMSO-d6) d: 5.32 (2H, s, CH2), 7.25 (2H, d, Jo = 8.4 Hz,
H-2′, H-6′), 7.57 (d, 1H, Jm = 1.2 Hz, H-4), 7.58–7.68
(4H, m, H-2″, H-3″, H-5″, H-6″), 7.74–7.78 (2H, dt,
Jo = 8.0, Jm = 1.2 Hz, H-5, H-6), 7.89 (2H, dd, Jo = 8.4,
H-3′, H-5′), 7.96 (d, 1H, Jo = 8.0 Hz, H-3) 9.88 (1H,
O=C-H) ppm; 13C NMR (100 MHz, DMSO-d6) d: 69.6
(CH2), 110.6 (C-2), 115.7 (C-2′, C-6′), 128.6 (C-3″, C-5″),
128.7 (C-4′), 129.3(C-3′, C-5′), 129.5 (CN), 130.1 (C-4″),
476
Chem Biol Drug Des 2013; 81: 474–483

Discovery of Dual PPAR Modulator with Antidiabetic Effect
tion system (Promega, Madison, WI, USA), the reaction
(20 lL) was incubated in a thermocycler Select Cycler
(BioProducts, West Palm Beach, FL, USA), following the
cycle program: incubation at 25 °C for 5 min, extension at
42 °C for 55 min. The enzyme was inactivated at 70 °C,
for 15 min, and finally, samples were cooled to 4 °C, for
5 min. Then 1/10 volume of each RT reaction was ampli-
fied with SYBR Green master mix (Roche Molecular Bio-
chemicals, Mannheim, Germany) containing 0.5 mM of
customized primers for PPAR-c (F-CCAGAGTCTGCT-
GATCTGCG; R-GCCACCTCTTTGCTCTGCTC; Gene Bank
NM_011146.1), PPAR-a (F- ATGCCAGTACTGCCGTTTTC;
R-GGCCTTGACCTTGTTCATGT; Gene Bank NM_011144),
GLUT-4 (F- GATTCTGCTGCCCTTCTGTC; R- ATT-
GGACGCTCTCTCTCCAA; Gene Bank NM_009204.2), as
well as Fast Start enzyme, PCR buffer and 3.5 mM MgCl2,
in a final volume of 10 lL, the reactions were measured in
a Rotor-Gene real-time (Corbett Life Science, Concorde,
NSW, Australia). PCR was conducted using the following
cycling conditions: pre-incubation and denaturation at
95 °C during 10 min. The threshold cycles (Ct) were mea-
sured in separate tubes and in duplicate (13). The identity
and purity of the amplified products was checked by elec-
trophoresis on 2% agarose gel. The melting curve was
analyzed at the end of amplification following SYBER
Green kit conditions, as indicated by the company (Roche
Molecular Biochemicals). To ensure the quality of the mea-
surements, each assay included a negative control for
each gene. The amount of mRNA for each adipokine was
normalized according to the amount of mRNA encoding
atoms were merged. All torsions were allowed to rotate
during docking. The auxiliary program AUTOGRID generated
the grid maps. Each grid was centered at the crystallo-
graphic coordinates of the crystallographic compound.
ꢁ
The grid dimensions were 60 9 60 9 60 A with points
ꢁ
separated by 0.375 A. The Lamarckian genetic algorithm
was applied for the search using default parameters. The
number of docking runs was 25. After docking, all solu-
tions were clustered into groups with RMSD lower than
ꢁ
2.0 A. The clusters were ranked by the lowest energy rep-
resentative of each cluster.
Validation was performed in AUTODOCK 4.2 using PDB pro-
teins 1I7G (PPARa) and 1I7I (PPARc). The X-ray crystal
structure (Tesaglitazar) was built and docked within the
active site region formed by PPARa residues: Cys276,
Ser280, Tyr314, Leu321, Val332, His440; and PPARc resi-
dues: Tyr464 and Cys285, Ser289, His323, Ile341,
Met364, His449, and Tyr473. This validation was carried
out based on the important interactions made by the
ligand bound within the active site residues indicating that
the parameters for docking simulations are good in repro-
ducing theses interactions between the X-ray crystal struc-
ture with both receptors.
In vivo studies
Male Wistar rats weighing 200–250 g bodyweight were
housed at standard laboratory conditions and fed with a
rodent pellet diet and water ad libitum. They were main-
tained at room temperature and at a photoperiod of 12 h
day/night cycle. Animals described as fasted were
deprived of food for 16 h but had free access to tap
water. All animal procedures were conducted in accor-
dance with our Federal Regulations for Animal Experimen-
tation and Care (SAGARPA, NOM-062-ZOO-1999,
ribosomal
protein
36B4
(F-AAGCGCGTCCTGG-
CATTGTCT; R-CCGCAGGGGCAGCAGTGGT; Gene Bank
NM_007475.2).
The DCt values were calculated in every sample for each
gene of interest as follows: Ctgene of interestÀCtreference
gene with b-actin as the reference gene (mRNA of refer-
ence remained stable throughout the experiments). Rela-
tive changes in the expression level of one specific gene
(DDCt) were calculated as DCt of the test group minus
DCt of the control group and then presented as 2ÀDDCt.
ꢀ
Mexico) and approved by the Institutional Animal Care and
Use Committee based on US National Institute of Health
publication (No. 85-23, revised 1985).
Induction of diabetes
Streptozotocin (STZ) was dissolved in citrate buffer (pH
4.5) and nicotinamide was dissolved in normal physiologi-
cal saline solution. Non-insulin-dependent diabetes mellitus
rat model was induced in overnight fasted rats by a single
intraperitoneal injection of 65 mg/kg STZ, 15 min after the
i.p. administration of 110 mg/kg of nicotinamide (15).
Hyperglycemia was confirmed by the elevated glucose
concentration in plasma, determined at 72 h by glucome-
ter. The animals with blood glucose concentration higher
than 250 mg/dL were used for the antidiabetic screening.
In silico docking studies
DISCOVERY STUDIO, version 3.5, (Accelrys Inc., San Diego,
CA, USA), and PYMOL 1.0 were used for visualization. The
crystal structure of PPARa and PPARc was retrieved from
the PDB with the accession codes 1I7G and 1I7I, respec-
tively. Docking calculations were conducted with AUTODOCK,
version 4.2 (14). In short, AUTODOCK performs an automated
docking of the ligand with user-specified dihedral flexibility
within a protein rigid binding site. The program performs
several runs in each docking experiment. Each run pro-
vides one predicted binding mode. All water molecules
and also tesaglitazar (crystallographic ligand) were
removed from the crystallographic structure, and all hydro-
gen atoms were added. For all ligands and proteins,
Gasteiger charges were assigned and non-polar hydrogen
Antidiabetic assay (non-insulin-dependent
diabetes mellitus rat model)
The diabetic animals were divided into groups of five ani-
mals each (n = 5). Rats of experimental groups were
Chem Biol Drug Des 2013; 81: 474–483
477

Hidalgo-Figueroa et al.
administered a suspension of the compound 1 (prepared
in 1% Tween 80) orally (50 mg/kg body weight). A control
group animal was also treated with 1% Tween 80. Gliben-
clamide (5 mg/kg) was used as hypoglycemic reference
drug. Blood samples were collected from the caudal vein
at 0, 1, 3, 5, and 7 h after vehicle, sample and drug
administration. Blood glucose concentration was estimated
by the enzymatic glucose oxidase method using a com-
mercial glucometer (Accu-Chek, Performa; Rocheꢀ). The
percentage variation of glycemia for each group was
calculated in relation to the initial (0 h) level, according to
the formula, %Variation of glycemia = [(GxÀG0)/G0] 9 100,
where G0 were initial glycemia values, and Gx were the
glycemia values at +1, +3, +5, and +7 h, respectively (14).
All values were expressed as mean Æ SEM. Statistical
significance was estimated by analysis of variance (ANOVA),
p < 0.05 and p < 0.01 implies significance.
Knoevenagel reaction once again, to obtain compounds 3
and 4, respectively (Scheme 3).
All compounds were recovered with 73–80% yields.
Compounds were purified by recrystalization or by column
chromatography. The chemical structures of the synthesized
compounds were confirmed on the basis of their spectro-
scopic and spectrometric data (NMR and mass spectra),
and their purity ascertained by microanalysis.
The in vitro part
Relative expression of PPARa and PPARc
Results presented in Figure 2A–C shows that compound 1
increase significantly the levels of relative expression for
PPARc (about fivefold), PPARa (sixfold), and GLUT4 (three-
fold) mRNA′s. The activation of PPARc could decrease the
glucose serum levels in diabetic patients due to the reduc-
tion in insulin resistance. Meanwhile, the activation of
PPARa is known to be effective in reducing serum triglyc-
eride levels (16). Our data also suggest that compound 1
can induce GLUT4 expression.
Results and Discussion
Chemistry
Title compounds 1–4 were prepared in a two-step reac-
tion. Starting from 4-hydroxybenzaldehyde (7), it was alkyl-
ated with 4-bromomethylbiphenyl-2-carbonitrile (6) under
SN2 conditions to give rise to the aldehyde 5 (Scheme 1).
Skeletal muscle glucose uptake is the rate-limiting step of
glucose utilization, and insulin-dependent and insulin-inde-
pendent signaling pathways physiologically regulate it,
both leading to the translocation of GLUT4 glucose trans-
porter to the plasma membrane. Several evidences indi-
cate that the levels of GLUT4 expression in skeletal
muscle are crucial for the regulation of total body glucose
homeostasis (17,18).
The aldehyde 5 was condensed with 2,4-thiazolidinedione
or barbituric acid, following Knoevenagel reaction condi-
tions to afford compounds
(Scheme 2).
1
and 2, respectively
For the rest of compounds, the treatment of aldehyde 7
with ethyl bromoacetate (8) provided ester 9, which was
reacted with 2,4-thiazolidinedione or barbituric acid, using
The designed scaffold hop from 2,4-thiazolidinedione and
barbituric acid did not retain the full biological effect,
Scheme 1: Synthesis of precur-
sor 5. a) K2CO3, acetone, reflux.
Scheme 2: Synthesis of com-
pounds 1 and 2. a) 2,4-thiazolidin-
edione,
benzoic
acid
(cat),
piperidine (cat), toluene, reflux; b)
barbituric acid¸ benzoic acid (cat),
piperidine (cat), toluene, reflux.
478
Chem Biol Drug Des 2013; 81: 474–483

Discovery of Dual PPAR Modulator with Antidiabetic Effect
Scheme 3: Synthesis of compo-
unds 3 and 4. a) K2CO3, acetone,
reflux; b) 2,4-thiazolidinedione,
benzoic acid (cat), piperidine (cat),
toluene, reflux; c) barbituric acid¸
benzoic acid (cat), piperidine (cat),
toluene, reflux.
A
B
C
Figure 2: Effect of compounds on expression of: mRNA PPARc (A), PPARa (B), and GLUT4 (C). Results are mean Æ SEM (n = 6)
*p < 0.05 compared with control (C). **p < 0.01 compared with control group. PPAR, peroxisome proliferator-activated receptor.
converting structures 1 and 2 into activity cliffs (chemical
compounds with highly similar structures but significantly
different biological activities) (19). It is possible that lack of
activity for compounds 3 and 4 in to the PPARc could be
related to low lipophilic tail showed by these structures.
docked into the catalytic site of the human PPARa and
PPARc using the program AUTODOCK 4.2. The docking
protocol was validated by re-docking of co-crystal ligand
Tesaglitazar in both PPAR′s isoforms. After re-docking, the
root mean square (RMS) deviation between the co-crystal
ꢁ
ligand and the docked structure was 0.99 A to PPARa
ꢁ
The in silico part
and 1.19 A to PPARc.
Molecular docking of compound 1 with PPARa and
PPARc
The aim of the molecular docking tools is to map the pos-
sible drug–receptor interactions to explain the activity
shown to assist to rational drug design. Compound 1 was
The compound 1 exhibited three hydrogen bonds between
oxygen and sulfur atoms from 2,4.thiazolidinedione and
several amino acids found in the active site of PPARa,
such as Ser280, Tyr314, His440, showing a binding free
energy of À11.04 Kcal/mol (Figure 3).
Chem Biol Drug Des 2013; 81: 474–483
479

Hidalgo-Figueroa et al.
A
B
Figure 3: Binding model of 1 in the active site of (A) PPARa and (B) PPARc. Stick representation of side chains (in red), shows the
residues that participate in the hydrogen bond network. Residue making van der Waals contacts is in yellow. PPAR, peroxisome
proliferator-activated receptor.
A
B
Figure 4: Comparison of docked 1 and co-crystal ligand (Tesaglitazar) into the active binding site of PPARa (A) and PPARc (Β). They
represent the lowest energy conformation estimated by AUTODOCK4.2. PPAR, peroxisome proliferator-activated receptor.
Compound 1 also showed four hydrogen interactions
between the 2,4-thiazolidinedione core and the PPARc
residues Ser289, His323, His449, and Tyr473, showing
binding free energy of À10.47 Kcal/mol. An extra p-r
interaction was found between 2-cyanophenyl group and
Ile341 (Figure 3). The molecular docking study revealed
that compound 1 docked into PPAR a and c exhibited
hydrogen bonds via O=C and -S- atoms with important
residues involved in the interaction shown by a significant
dual receptor agonist (Tesaglitazar). This could explain its
relevant activity on both receptors.
by 1, with Ser, Tyr, and His in both PPAR isoforms.
According to the models predicted by AUTODOCK4.2, the
studied derivative binds into the active site of PPARa and
PPARc.
With these results, we propose that the features could be
taking into consideration for designing novel dual PPAR a/
c modulators should consist of four parts:
(a)An acidic head group, such as thiazolidine-2,-4-dione,
carboxylic acid, or related bioisoteres.
Figure 4 shows a comparison of the binding mode of the
most active compound with tesaglitazar. In addition, com-
(b)A central aromatic backbone
pound
1
share the same binding pattern and
pharmacophoric interactions as the co-crystal ligand of
PPAR′s. It is notable the three-dimensional similarity of the
binding models of the compound 1 and Tesaglitazar.
(c)An extra-lipophilic aromatic region
(d)A flexible spacer that connects regions (b) and (c),
and allow the structure to adopt specific conformation
(Figure 6).
Figure 5 depicts the 2D interactions maps of compound
1 and tesaglitazar. Hydrogen bond network is conserved
480
Chem Biol Drug Des 2013; 81: 474–483

Discovery of Dual PPAR Modulator with Antidiabetic Effect
A
B
C
D
Figure 5: Two-dimensional interaction diagram of the predicted binding mode of compound 1 with: (A) PPARa and (B) PPARc. It shows
as reference, the corresponding 2-D interaction maps for tesaglitazar with (C) PPARa and (D) PPARc. PPAR, peroxisome proliferator-
activated receptor.
Figure 6: Overlay of bioactive conformations of Tesaglitazar (yellow: PPARa; blue: PPARc) and compound 1 (purple: PPARa; green:
PPARc). It is remarkable that both compounds share the same binding pattern and pharmacophoric interactions with PPAR′s. PPAR,
peroxisome proliferator-activated receptor.
In silico toxicology profile
Cytochrome P450 isoform CYP3A4 is the major enzyme
responsible for xenobiotic metabolism in human organism.
Inhibition of CYP3A4 at clinically relevant concentration
(IC50 < 10 lM) can lead to drug–drug interactions and
undesirable adverse effects (20). Compound 1 showed
satisfactory toxicity profiles, and its predictions of inhibition
In silico prediction of toxicity has been performed in drug
design and development to avoid the experimental study
of potentially harmful substances. The toxicity parameters
of the compound 1 and tesaglitazar were calculated
through the ACD/TOXSUITE software, v. 2.95 (Table 1).
Chem Biol Drug Des 2013; 81: 474–483
481

Hidalgo-Figueroa et al.
Table 1: Toxicity profiles predicted for compound 1 and PPAR a/c dual agonist tesaglitazar
LD50 (mg/kg)
Mouse
Probability of inhibition (IC50 or Ki <10 lM)
CYP450 isoform
Rat
i.p.
Compound
i.p.
p.o.
p.o.
3A4
2D6
2C9
1A2
hERG
1
520
440
1200
930
450
230
3200
3600
0.25
0.22
0.03
0.05
0.26
0.15
0.02
0.02
0.58
0.01
Tesaglitazar
hERG, human ether-a-go-go related gene; PPAR, peroxisome proliferator-activated receptor.
for the four isoforms of CYP450 were comparable than
the reference compound tesaglitazar.
Cardiotoxicity of drug-like compounds associated with
human ether-a-go-go (hERG) channel inhibition is becom-
ing more and more common cause of drug candidates’
attrition (21). Compound 1 showed low prediction of hERG
channel inhibition at clinically relevant concentrations
(Ki < 10 lM).
The acute toxicity of the chemical is defined as a dose that
is lethal to 50% of the treated animals (LD50). The acute
toxicity can be viewed as a ‘cumulative potential’ to cause
various acute effects and death of animals. In these pre-
dictions, compound 1 demonstrated similar calculated
LD50 than tesaglitazar, showing low toxicity profiles.
Figure 7: Effect of a single dose of 1 (50 mg/kg; intragastric,
n = 5) in streptozotocin-nicotinamide-induced diabetes rat model.
ISS, isotonic saline solution. *p < 0.05 versus ISS group.
The in vivo part
crucial for the regulation of total body glucose homeostasis
and insulin resistance.
Antidiabetic activity of compound 1
Compound 1 was evaluated for in vivo antidiabetic activity
using a STZ-nicotinamide non-insulin-dependent diabetes
mellitus rat model (NIDDM) rat model (16,22). Glibencla-
mide was taken as hypoglycemic control. The hypoglyce-
mic activity of 1 was determined using a 50 mg/kg single
dose. Compound 1 demonstrated significant hypoglycemic
activity (p < 0.05) compared with control, by lowering
glycemia ranging from 12% to 35% (Table 2). The effect
was sustained during the 7 h of experiment, and it was
comparable with the hypoglycemic action showed by gli-
benclamide (Figure 7).
Conclusions and Future Directions
We report the discovery of a thiazolidine-2,4-dione/biphe-
nylcarbonitrile hybrid (1) as a promising antidiabetic com-
pound, which shown significant increasing in the mRNA
expression of PPAR a, PPAR c, and also GLUT 4. This
compound showed a good in vivo hypoglycemic effect.
According to the docking models, compound 1 may bind
into the active site of both PPAR′s in the same manner as
tesaglitazar does. The high in vivo hypoglycemic activity
and the satisfactory toxicity profile predicted for this com-
pound makes it a suitable lead to develop single entities
with PPAR a/c dual agonist that may confer potential use
in the treatment of experimental diabetes.
In our in vivo experiments, the increase in GLUT4 expres-
sion in rat could be associated with a marked decrease in
glucose concentrations, confirming that the level of
expression of this transporter in skeletal muscle may be
Table 2: Percentage of variation of blood glucose concentration on streptozotocin-nicotinamide induced diabetic rats treated with com-
pound 1
Percentage of variation of glycemia Æ SEM (mg/dL)a
Compound
Dose (mg/kg)
Zero hour
First hour
Third hour
Fifth hour
Seventh hour
1
50
5
0 Æ 0.0
0 Æ 0.0
À12.42 Æ 7.1
À20.4 Æ 8.17
À25.55 Æ 4.5
À36.3 Æ 7.55
À35.72 Æ 10.7
À37.4 Æ 7.16
À32.36 Æ 14.8
À43.6 Æ 9.9
Glibenclamide
aValues represent the mean Æ SEM (n = 6). p < 0.05 compared with control group. The negative value indicates decrease in glycemia.
482
Chem Biol Drug Des 2013; 81: 474–483

Discovery of Dual PPAR Modulator with Antidiabetic Effect
(2008) Glycine increases mRNA adiponectin and
diminishes pro-inflammatory adipokines expression in
3T3-L1 cells. Eur J Pharmacol;587:317–321.
Acknowledgment
This work was supported in part by the Consejo Nacional
13. Almanza-Perez J.C., Alarcon-Aguilar F.J., Blancas-
Flores G., Campos-Sepulveda A.E., Roman-Ramos R.,
Garcia-Macedo R., Cruz M. (2010) Glycine regulates
inflammatory markers modifying the energetic balance
through PPAR and UCP-2. Biomed Pharmacoth-
er;64:534–540.
ꢀ
de Ciencia y Tecnologıa (CONACyT) under grant No.
100608 and Facultad de Farmacia internal grant.
References
14. Morris G.M., Huey R., Lindstrom W., Sanner M.F.,
Belew R.K., Goodsell D.S., Olson A.J. (2009) Autodock4
and AutoDockTools4: automated docking with selective
receptor flexibility. J Comput Chem;16:2785–2791.
15. Torres-Piedra M., Ortiz-Andrade R., Villalobos-Molina
R., Singh N., Medina-Franco J.L., Webster S.P., Binnie
1. Lee C.H., Olson P., Evans R.M. (2003) Minireview: lipid
metabolism, metabolic diseases, and peroxisome pro-
liferator-activated receptors. Endocrinology;144:2201–
2207.
2. Willson T.M., Brown P.J., Sternbach D.D., Henke B.R.
(2000) The PPARs: from orphan receptors to drug
discovery. J Med Chem;43:527–550.
ꢀ
M., Navarrete-Vazquez G., Estrada-Soto S. (2010) A
comparative study of flavonoid analogues on strepto-
zotocin-nicotinamide induced diabetic rats: quercetin
as a potential antidiabetic agent acting via 11b-hydrox-
ysteroid dehydrogenase type 1 inhibition. Eur J Med
Chem;45:2606–2612.
3. Sharma A.K., Sk U.H., He P., Peters J.M., Amin S.
(2010) Synthesis of isosteric selenium analog of the
PPARbeta/delta agonist GW501516 and comparison of
biological activity. Bioorg Med Chem Lett;20:4050–
4052.
16. Nevin D.K., Lloyd D.G., Fayne D. (2011) Rational tar-
geting of peroxisome proliferating activated receptor
subtypes. Curr Med Chem;18:5598–5623.
17. Tsao T.S., Li J., Chang K.S., Stenbit A.E., Galuska D.,
Anderson J.E., Zierath J.R., McCarter R.J., Charron
M.J. (2001) Metabolic adaptations in skeletal muscle
overexpressing GLUT4: effects on muscle and physical
activity. FASEB J;15:958–969.
18. Zisman A., Peroni O.D., Abel E.D., Michael M.D.,
Mauvais-Jarvis F., Lowell B.B., Wojtaszewski J.F.,
Hirshman M.F., Virkamaki A., Goodyear L.J., Kahn
C.R., Kahn B.B. (2000) Targeted disruption of the glu-
cose transporter 4 selectively in muscle causes insulin
resistance and glucose intolerance. Nat Med;6:924–
928.
4. Wild S., Roglic G., Green A., Sicree R., King H. (2004)
Global prevalence of diabetes: estimates for the year
2000 and projections for 2030. Diabetes Care;27:1047
–1053.
ꢀ
ꢀ
5. Navarrete-Vazquez G., Ramırez-Martınez M., Estrada-
Soto S., Nava-Zuazo C., Paoli P., Camici G., Escalan-
ꢀ
ꢀ
te-Garcıa J., Medina-Franco J.L., Lopez-Vallejo F.,
Ortiz-Andrade R. (2012) Synthesis, in vitro and in silico
screening of ethyl 2-(6-substituted benzo[d]thiazol-
2-ylamino)-2-oxoacetates as protein-tyrosine phos-
phatase 1B inhibitors. Eur J Med Chem;53:346–355.
6. Balakumar P., Rose M., Ganti S.S., Krishan P., Singh
M. (2007) PPAR dual agonists: are they opening Pan-
dora’s Box? Pharmacol Res;56:91–98.
7. Adeghate E., Adem A., Hasan M.Y., Tekes K., Kalasz
H. (2011) Medicinal chemistry and actions of dual and
pan PPAR modulators. Open Med Chem J;5(Suppl.
2):93–98.
8. Cox S.L. (2006) Tesaglitazar: a promising approach in
type 2 diabetes. Drugs Today (Barc);42:139–146.
9. Sundriyal S., Viswanad B., Ramarao P., Chakraborti
A.K., Bharatam P.V. (2008) New PPARgamma ligands
based on barbituric acid: virtual screening, synthesis
and receptor binding studies. Bioorg Med Chem
Lett;18:4959–4962.
10. Momose Y., Maekawa T., Yamano T., Kawada M.,
Odaka H., Ikeda H., Sohda T. (2002) Novel 5-substi-
tuted 2,4-thiazolidinedione and 2,4-oxazolidinedione
derivatives as insulin sensitizers with antidiabetic activi-
ties. J Med Chem;45:1518–1534.
ꢀ
ꢀ
19. Mendez-Lucio O., Perez-Villanueva J., Castillo R.,
Medina-Franco J.L. (2012) Activity landscape modeling
of PPAR ligands with dual-activity difference maps.
Bioorg Med Chem;20:3523–3532.
20. Xu L., Chen Y., Pan Y., Skiles G.L., Shou M. (2009)
Prediction of human drug-drug interactions from time-
dependent inactivation of CYP3A4 in primary hepato-
cytes using a population-based simulator. Drug Metab
Dispos;37:2330–2339.
21. Taboureau O., Jørgensen F.S. (2011) In silico predic-
tions of hERG channel blockers in drug discovery: from
ligand-based and target-based approaches to systems
chemical biology. Comb Chem High Throughput
Screen;14:375–387.
ꢀ
22. Moreno-Dıaz H., Villalobos-Molina R., Ortiz-Andrade R.,
~
ꢀ
Dıaz-Coutino D., Medina-Franco J.L., Webster S.P.,
11. Momose Y., Maekawa T., Odaka H., Ikeda H., Sohda
T. (2002) Novel 5-substituted-1H-tetrazole derivatives
as potent glucose and lipid lowering agents. Chem
Pharm Bull;50:100–111.
ꢀ
Binnie M., Estrada-Soto S., Ibarra-Barajas M., Leon-Ri-
vera I., Navarrete-Vazquez G. (2008) Antidiabetic activity
of N-(6-substituted-1,3-benzothiazol-2-yl)benzenesulf-
onamides. Bioorg Med Chem Lett;18:2871–2877.
ꢀ
~
12. Garcia-Macedo R., Sanchez-Munoz F., Almanza-Perez
J.C., Duran-Reyes G., Alarcon-Aguilar F., Cruz M.
Chem Biol Drug Des 2013; 81: 474–483
483