Inhibitory effects of multi-substituted benzylidenethiazolidine-2,4-diones on LDL oxidation

Article abstract of DOI:10.1016/j.bmc.2004.06.001

Multi-substituted benzylidenethiazolidine-2,4-diones 3a-h were synthesized by Knoevenagel condensation of di- or tri-substituted 4-hydroxybenzaldehydes [or 1-(3,5-di-tert-butyl-4-hydroxyphenyl)ethanone] 1 with thiazolidine-2,4-dione (2) and evaluated for

Full text of DOI:10.1016/j.bmc.2004.06.001

Bioorganic & Medicinal Chemistry 12 (2004) 4017–4023
Inhibitory effects of multi-substituted
benzylidenethiazolidine-2,4-diones on LDL oxidation
Tae-Sook Jeong,a, Ju-Ryoung Kim,a, Kyung Soon Kim,a,b Kyung-Hyun Cho,a
Ki-Hwan Baeb and Woo Song Leea,*
aNational Research Laboratory of Lipid Metabolism and Atherosclerosis,
Korea Research Institute of Bioscience and Biotechnology, Daejeon 305-333, Republic of Korea
bCollege of Pharmacy, Chungnam National University, Daejeon 305-764, Republic of Korea
Received 13 April 2004; revised 31 May 2004; accepted 1 June 2004
Abstract—Multi-substituted benzylidenethiazolidine-2,4-diones 3a–h were synthesized by Knoevenagel condensation of di- or tri-
substituted 4-hydroxybenzaldehydes [or 1-(3,5-di-tert-butyl-4-hydroxyphenyl)ethanone] 1 with thiazolidine-2,4-dione (2) and eval-
uated for antioxidant activities of Cu2þ-induced oxidation of human low-density lipoproteins (LDL). Among compounds 3a–h, 3a
was superior to probucol in LDL-antioxidant activities and found to be ninefold more active than probucol. Due to its potency,
compound 3a was tested for complementary in vitro investigations, such as TBARS assay (IC50 ¼ 0.1 lM), lag time (240 min at
1.5 lM), relative electrophoretic mobility (REM) of ox-LDL (inhibition of 83% at 10 lM), fragmentation of apoB-100 (inhibition of
61% at 5 lM), and radical DPPH scavenging activity on copper-mediated LDL oxidation. In macrophage-mediated LDL oxidation,
the TBARS formation was also inhibited by compound 3a.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
a PPARc ligand.6 Therefore, compound 3a and its
analogues acting as both an antioxidant and a PPARc
ligand may be of value for the prevention and/or the
treatment of atherosclerosis by attenuating foam cell
formation.
The syntheses and biological activities of thiazolidinones
have been extensively reviewed.1 Especially, thiazolid-
inediones (TZDs) were found to exhibit antidiabetic
activities as PPARc ligands2 and anti-inflammatory
activities as 5-lipoxygenase and cyclooxygenase inhibi-
tors.3 In addition, it has been known that PPARc is
highly expressed in macrophage foam cells of human
and mouse atherosclerotic lesions and TZDs inhibit the
development of atherosclerosis in Ldlrꢀ=ꢀ mice.4 In
general, oxidized low-density lipoproteins (ox-LDLs)
play a key role in the early stages of atherosclerosis.5 In
connection with our ELISA system (based on the
interaction between PPARc and SRC-1) for searching a
PPARc ligand, we found that 3,5-di-tert-butyl-4-
hydroxybenzylidene-thiazolidin-2,4-dione (3a) acted as
For the last two decades, antioxidants have exhibited
anti-atherogenic activities by inhibiting the foam
cell formation in animal model.7 Among various
antioxidants, probucol and vitamin E are well known
to lower lipid levels and the risk of coronary heart
disease incidence.8–13 However, both probucol and
vitamin E have an untoward effect of lowering
serum high-density lipoprotein (HDL)-cholesterol
levels.
This paper deals with the study of antioxidant activities
of 3a–h on LDL using thiobarbituric acid reactive sub-
stances (TBARS) formation. Also, oxidation of LDL
for 3a was monitored by various systems, such as con-
jugated diene formation, relative electrophoretic mobil-
ity (REM) of ox-LDL, fragmentation of apoB-100,
radical DPPH scavenging activity, and macrophage-
mediated LDL oxidation.
Keyword: Antioxidant.
* Corresponding author. Tel.: +82-42-860-4278; fax: +82-42-861-2675;
e-mail: wslee@kribb.re.kr
Both authors contributed equally to the work.
0968-0896/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.06.001

4018
T.-S. Jeong et al. / Bioorg. Med. Chem. 12 (2004) 4017–4023
Table 1. Inhibition of copper-induced lipid peroxidation in the
TBARS assay by benzylidenethiazlidine-2,4-diones 3a–h
2. Results and discussion
2.1. Chemistry
R1 R4
O
R2
5-(3,5-Di-tert-butyl-4-hydroxybenzylidene)thiazolidine-
2,4-dione (3a) was prepared by Knoevenagel con-
densation of 3,5-di-tert-butylbenzaldehyde (1) with
thiazolidine-2,4-dione (2) in 73% yield. Similar reactions
of 3,5-di-i-propyl-, methyl-, or 2,3,5-trimethylbenzalde-
hydes 1b–d with 2 gave benzylidenethiazoles 3b–d in
moderate yields. Benzaldehydes 1e–g substituted with
electron rich functional groups, such as hydroxy, fluoro,
and methoxy groups, were subjected to Knoevenagel
condensation conditions to give benzylidenethiazoles
3e–g in good yields (method A) (Scheme 1). In order
to prepare an analog containing a methyl substituent
on the benzylidene double bond, Knoevenagel con-
densation (method B) of the di-tert-butyl ketone 1h
with 2 gave 5-[1-(3,5-di-tert-butyl-4-hydroxyphenyl)eth-
ylidene]thiazolidine-2,4-dione (3h) in 78% yield. The
geometry of all benzylidenethiazolidine-2,4-diones 3a–h
was determined to be (Z)-olefinic products by compar-
ing their spectroscopic analysis with reported data pre-
viously.3
NH
S
HO
O
R3
Compds
R1
R2
R3
R4
IC50 (lM)
3a
3b
3c
3d
3e
3f
H
H
H
Me
H
H
H
H
t-Bu
i-Pr
Me
t-Bu
i-Pr
Me
Me
H
H
H
H
H
H
H
H
Me
0.1
0.9
0.3
4.5
Me
OH
F
2.3
F
16.0
25.5
0.3
3g
3h
Pa
OMe
t-Bu
OMe
t-Bu
0.9
a Probucol was used as a reference antioxidant.
are reducing agents that reduce transition metal or
interact with free radicals to result in the formation of
antioxidant-derived free radicals. Also, Ziouzenkova
et al.16 reported that copper ions are bound to high as
well as to low affinity sites to generate lipid peroxyl
radicals (LOOÅ), which can be effectively inhibited by a-
tocopherol. Therefore, these results may be rationalized
that bulky substituents on phenol ring may stabilize the
phenoxy radical formed to show more antioxidant
activity against LDL, which agreed with earlier work.17
It also demonstrated that 2,6-di-tert-butylphenol is more
reactive toward radical source to afford 2,6-di-tert-butyl-
phenoxy radical in inhibition of free radical reaction.18
Even though 3e has more reactive to DPPH (Fig. 4), 3e
showed less antioxidant activity than 3a. It may con-
clude that 3e might protect Cu2þ-mediated LDL oxi-
dation as a Cu2þ-chelator,19 not free radical scavenger.
In order to confirm, which part of phenol or amide
moiety of 3a concerns in free radical reaction, 1a and
thizolidine-2,4-dione 2 were tested for their antioxidant
activities to show the values with an IC50 of 0.4 lM for
1a and an IC50 of >100 lM for 2. However, hybrid
molecule 3a appeared to be a remarkably potent anti-
oxidant with an IC50 of 0.1 lM, which observed in the
MDA test (at least nine times more potent than pro-
bucol itself). Furthermore, the steric and electronic
factors of substituents to stabilize phenoxy radical
formed from phenolic hydroxy group may influence
antioxidant activities for human LDL.
2.2. Biology
2.2.1. Lipid peroxidation. Various di-, tri-, or tetra-
substituted benzylidenethiazoles 3a–h were synthesized
and examined in vitro for their abilities to protect hu-
man LDL against Cu2þ-induced peroxidation. The
prescreening test as preliminary evaluation to select the
best candidate among 3a–h is based on a malondialde-
hyde (MDA) dosage. The results are expressed as IC50
(concentration inhibiting 50% of the Cu2þ-induced lipid
peroxidation), determined by the TBARS assay, and
expressed as a MDA equivalent. Then, initial concen-
tration of a-tocopherol in a native LDL was 1.8 lM,
which was determined by HPLC.14 The LDL-antioxi-
dant activities of the compounds 3a–h were confirmed
by the positive control with probucol. Compounds 3a–e
containing functional groups (i-Pr, Me, and OH)
showed a potent antioxidant activities, whereas 3f and
3g involving electron donor groups (fluoro and meth-
oxy) exhibited less antioxidant activity. Compound 3h
including an additional methyl group at benzylidene
double bond exhibited a high inhibitory activity (Table
1). According to Lichtenberg’s results,15 antioxidants
R1
O
R1
R4
O
R2
R2
R4
2.2.2. Detection of conjugated diene formation. The
extent of lag time was interpreted as oxidation resistant
capacity of LDL. The formation of conjugated dienes
during LDL oxidation represents the early peroxidation
of LDL. The oxidation of LDL was determined by
measuring conjugated diene formation at 234 nm every
10 min for 4 h. The effect of compound 3a on the pro-
longing of lag time was shown in Figure 1. First, the
LDL was incubated with CuSO4 alone to have a lag time
(80 min). While 1.0 lM of 3a appeared to have moderate
2
NH
i
S
HO
HO
R3
O
R3
H
1
3
N
O
O
2 : TZD =
S
Scheme 1. Reagents and reaction conditions: (i) method A: piperidine
(cat.), EtOH, reflux; method B: NH4OAc, toluene, reflux.

T.-S. Jeong et al. / Bioorg. Med. Chem. 12 (2004) 4017–4023
4019
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
control
Probucol 3 µM
Probucol 5 µM
3a 1 µM
3a 1.5 µM
Figure 3. Inhibition of Cu2þ-mediated apoB-100 fragmentation in
LDL by 3a. LDL (240 lg protein/mL) was incubated for with 10 lM
CuSO4. Lane 1: standard marker, lane 2: native LDL (absence of
CuSO4), lane 3: ox-LDL, lane 4: 3a (20 lM), lane 5: 3a (10 lM), lane 6:
3a (5 lM), lane 7: 3a (1 lM), lane 8: probucol (20 lM), lane 9: probucol
(10 lM), lane 10: probucol (5 lM), lane 11: probucol (1 lM).
the LDL oxidation has been evaluated by the frag-
mentation of apoB-100 through the electrophoretic
analysis on polyacrylamide gel in the presence of 4%
sodium dodecylsulfate (SDS-PAGE). The band of
apoB-100 was observed on the native LDL (lane 2), but
the band completely disappeared when the LDL was
incubated with CuSO4 (lane 3). The fragmentation of
apoB-100 was inhibited in 86%, 80%, and 61%, respec-
tively, in the presence of 20, 10, and 5 lM of 3a (lanes 4–
6), whereas addition of 1 lM of 3a did not protect apoB-
100 fragmentation (lane 7). At the same concentration
of probucol, as a positive control, the fragmentation of
apoB-100 was inhibited in 61%, 58%, and 36%, respec-
tively, (lanes 8–10), whereas 1 lM of probucol also did
not protect apoB-100 fragmentation (lane 11), as shown
in Figure 3. As a result, 3a was more potent than pro-
bucol in protection of LDL against Cu2þ-induced per-
oxidation.
0
20
40
60
80 100 120 140 160 180 200 220 240 260
Time (min)
Figure 1. Effect of 3a on Cu2þ-mediated LDL oxidation. LDL (100 lg
protein/mL) was incubated with 5 lM of CuSO4 in the presence or
absence of antioxidant, 3a. Conjugated diene formation was measured
by determining the absorbance at 234 nm every 10 min for 4 h. Pro-
bucol was used as a reference antioxidant.
effect on the lag time extension (190 min), 1.5 lM of 3a
strongly delayed the lag time to 240 min. However, 3.0
and 5.0 lM of probucol mildly extended the lag time to
132 and 162 min, respectively. Thus, 3a inhibited much
more the formation of conjugated diene formation
during Cu2þ-induced LDL oxidation than did probucol.
2.2.3. Relative electrophoretic mobility (REM) of
ox-LDL. The effects of various concentration of 3a on
Cu2þ-mediated oxidation of LDL were determined by
REM assay. As shown in Figure 2, the 10 lM of CuSO4
alone was employed to oxidize LDL for 12 h (lane 2).
The REM of ox-LDL was lowered to 83%, 83%, and
10% (lanes 3–5), respectively, in the presence of 20, 10,
and 5 lM of 3a, whereas treatment of 1 lM of 3a did not
protect LDL oxidation (lane 6), compared to that of
oxidized LDL. The mobility of LDL in probucol as a
positive control was reduced dose-dependently to 86%
(20 lM), 70% (10 lM), and 7% (5 lM) (lanes 7–9),
whereas treatment with 1 lM of probucol did not pro-
tect LDL oxidation (lane 10).
2.2.5. Radical DPPH scavenging activity. 1,1-Diphenyl-
2-picrylhydrasyl (DPPH) is a very stable free radical and
widely used for evaluation of antioxidant activities.21 In
order to confirm whether compounds 3a–h act as a free
radical scavenger, we tested radical DPPH scavenging
activity of representative 3a, 3e, 3f, and 3g at 100 lM.
Then, activities of 3a, 3e, 3f, and 3g were measured as
decolorizing activity following the trapping of the un-
paired electron of DPPH.22 These results are shown in
Figure 4. At the earliest time, the DPPH radical was
decreased rapidly by 100 lM of 3a and 3e, but decreased
slowly by 100 lM of 3f and 3g. After 20 min, 34% and
9% DPPH radicals, respectively, were remained at the
presence of compounds 3a and 3e, whereas 3f and 3g
having electron donor groups such as fluoro and meth-
oxy exhibited 93% and 87% DPPH radicals remaining at
100 lM, respectively. Therefore, these compounds 3a–h
have proven to have radical DPPH scavenging capacity.
2.2.4. ApoB-100 fragmentation. Since it has been
reported that radical reaction of LDL causes fragmen-
tation of apoB-100 that is a major component of LDL,20
2.2.6. Macrophage-mediated LDL oxidation. Next, we
were interested to perform an antioxidant activity of 3a
in macrophage-mediated oxidation of LDL. The cellular
oxidative modification of LDL to a form recognized by
the scavenger receptor requires the presence of transi-
tion metal ions in the medium.23 In macrophage-medi-
ated LDL oxidation, the TBARS formation in the
harvested medium was also inhibited by 3a at a similar
order of activity to that obtained in Cu2þ-induced LDL
oxidation. These results are summarized in Table 2. The
Figure 2. Effect of 3a on electrophoretic mobility of LDL. LDL
(240 lg protein/mL) was incubated with 10 lM CuSO4. Lane 1: native
LDL (absence of CuSO4), lane 2: ox-LDL, lane 3: 3a (20 lM), lane 4:
3a (10 lM), lane 5: 3a (5 lM), lane 6: 3a (1 lM), lane 7: probucol
(20 lM), lane 8: probucol (10 lM), lane 9: probucol (5 lM), lane 10:
probucol (1 lM). Probucol is a positive control.

4020
T.-S. Jeong et al. / Bioorg. Med. Chem. 12 (2004) 4017–4023
LDL protein, respectively. As a result, antioxidant
activity of 3a was much higher than that of probucol on
macrophage-mediated LDL oxidation (Table 2).
100
80
60
40
20
0
Control
3a
3. Conclusions
3e
3f
3g
In this study, we synthesized multi-substituted
benzylidenethiazolidine-2,4-diones 3a–h by Knoevena-
gel condensation of di- or tri-substituted 4-hydroxy-
benzaldehydes [or 1-(3,5-di-tert-butyl-4-hydroxyphen-
yl)ethanone] 1 with thiazolidine-2,4-dione (2) and
demonstrated that they exhibited antioxidant activities.
Previously, Unangst et al.3 reported that 3a and 3h were
novel dual 5-lipoxygenase and cyclooxygenase inhibitors
with anti-inflammatory activities. Furthermore, we
demonstrated the most potent antioxidant activity of 3a
in various LDL oxidation tools (e.g., TBARS assay,
conjugated diene formation, REM of ox-LDL, frag-
mentation of apoB-100 by SDS-PAGE, radical DPPH
scavenging activity, and macrophage-mediated LDL
oxidation). Further studies on in vivo efficacy test of
cholesterol-lowering and anti-atherogenic activities and
the mechanism of antioxidant action of compound 3a
are underway.
0
5
10
15
20
Time (min)
Figure 4. Effects of 3a, 3e, 3f, and 3g on radical DPPH scavenging.
Compounds 3a, 3e, 3f, and 3g (100 lM) were incubated with 100 lM of
DPPH in methanol at room temperature for 20 min. The absorbance at
517 nm of each compound solution was measured. The antiradical
activity was expressed by the remaining DPPH percentage.
Table 2. Effects of 3a and probucol on LDL oxidation by macro-
phages
Incubation conditionsa
MDA nmol/mg
LDL proteinb
LDL + Cu2þ
10.4 0.7
80.0 2.6
2.0 0.2
3.5 0.1
4.8 0.1
10.9 1.3
15.2 0.7
20.1 2.1
22.3 1.6
31.0 2.0
4. Experimental
4.1. Chemistry
LDL + cell + Cu2þ
LDL + cell + Cu2þ + 40 lM 3a
LDL + cell + Cu2þ + 20 lM 3a
LDL + cell + Cu2þ + 10 lM 3a
LDL + cell + Cu2þ + 5 lM 3a
LDL + cell + Cu2þ + 40 lM probucol
LDL + cell + Cu2þ + 20 lM probucol
LDL + cell + Cu2þ + 10 lM probucol
LDL + cell + Cu2þ + 5 lM probucol
Melting points were measured on a Thomas–Hoover
capillary apparatus and are uncorrected. Proton and
carbon NMR spectra were measured downfield relative
to tetramethylsilane in CDCl3 and d-DMSO unless
otherwise noted (value in ppm); coupling constants J are
1
reported in Hertz; H NMR and 13C NMR were con-
a LDL (120 lg protein/mL) was incubated for 24 h at 37 ꢁC in RPMI
1640 medium with 2 lM of Cu2þ in 12-well plate containing macro-
phages, in the absence (control) or presence of increasing concen-
trations of compound 3a tested (5–40 lM).
ducted on Bruker avance 300 spectrometer. High reso-
nance mass spectra (HRMS) were recorded on JMS-700
(Jeol, Japan) and measured at Gyeongsang National
University (Chinju, Korea). HPLC analyses were per-
formed on Shimadzu CLASS-VP using a porasil (silica,
Waters) column eluted with hexane/i-PrOH (9:1) at a
flow rate of 1 mL/min.
b The extent of LDL oxidation was determined by directly in the
harvested medium using the TBARS assay. Data are shown as
means SD from two independent experiments performed in dupli-
cate.
LDL was only incubated with 2 lM of CuSO4 to give
the value of ox-LDL (10.4 0.7 MDA nmol/mg LDL
protein), whereas the content of ox-LDL by incubation
of LDL and 2 lM CuSO4 in the presence of macrophage
was 80.0 2.6 MDA nmol/mg LDL protein. The value
of ox-LDL by macrophage-mediated LDL oxidation
has proven to be much higher than that by Cu2þ-in-
duced LDL oxidation. This result coincides with the
previous reports.24 Therefore, antioxidant activities of
3a and probucol were tested by macrophage-mediated
LDL oxidation at dose-dependent concentration rang-
ing from 5 to 40 lM. In the presence of 5, 10, 20, and
40 lM of 3a, the content of ox-LDL was 10.9 1.3,
4.8 0.1, 3.5 0.1, and 2.0 0.2 MDA nmol/mg LDL
protein, respectively. At the same concentration of
probucol, the content of ox-LDL was 31.0 2.0,
22.3 1.6, 20.1 2.1, and 15.2 0.7 MDA nmol/mg
4.1.1. Typical procedure for the preparation of (Z)-5-(3,5-
di-tert-butyl-4-hydroxybenzylidene)-thiazolidine-2,4-dione
(3a) [method A]. To a solution of 3,5-di-tert-butyl-4-
hydroxybenzaldehyde (1a) (7 g, 29.9 mmol) in EtOH
(100 mL) was added thiazolidine-2,4-dione (2) (4.2 g,
35.8 mmol) in the presence of catalytic amounts of
piperidine (0.9 mL, 9.0 mmol) at room temperature.
After heating at 70 ꢁC for 24 h, the reaction mixture was
cooled to 0 ꢁC, and 1 N HCl (0.5 mL) and H2O (100 mL)
were added to precipitate a yellow powder, which was
filtered to give the crude compound 3a. The collected 3a
was washed with EtOAc and dried on air to obtain the
pure compound 3a (7.3 g, 73%) as a yellow powder. Mp
1
240–242 ꢁC (lit.3 mp 238–240 ꢁC); H NMR (300 MHz,
CDCl3) d 1.47 (18H, s), 5.71 (1H, s, –OH), 7.37 (2H, s),
7.85 (1H, s), 9.22 (1H, br, –NH); 13C NMR (75 MHz,

T.-S. Jeong et al. / Bioorg. Med. Chem. 12 (2004) 4017–4023
4021
CDCl3) d 167.8, 167.2, 156.7, 136.9, 135.9, 128.3, 124.4,
118.0, 34.5, 30.1; HPLC analysis [hexane/i-PrOH (9:1)]
tR ¼ 1:0 min (99.96%); HRMS calcd for C18H23NO3S:
333.1399; found: 333.1393.
281 (Mþ); HRMS calcd for C12H11NO5S: 281.0358;
found: 281.0356.
4.1.8. (Z)-5-[1-[3,5-Bis(1,1-dimethylethyl)-4-hydroxyphen-
yl]ethylene]-2,4-thiazolidinedione (3h) [method B]. A
mixture of 1h (1.0 g, 4.03 mmol), 2 (0.94 g, 8.05 mmol),
and NH4OAc (0.6 g, 8.05 mmol) in toluene (4 mL) was
refluxed for 48 h. The reaction mixture was cooled and
evaporated. The residue was digested briefly in warm
MeOH (10 mL), cooled, and filtered to give 1.1 g (78%)
of 3h as a yellow powder. Mp 254–255 ꢁC (lit.3 mp 253–
254 ꢁC); 1H NMR (300 MHz, d-DMSO) d 1.39 (18H, s),
2.65 (3H, s), 7.20 (2H, s), 7.45 (1H, br, –OH), 12.18 (1H,
br, –NH); 13C NMR (75 MHz, d-DMSO) d 21.3, 30.1,
34.7, 120.0, 123.7, 132.9, 139.0, 150.5, 155.1, 166.6,
168.0.
4.1.2. (Z)-5-(4-Hydroxy-3,5-diisopropylbenzylidene)-
thiazolidine-2,4-dione (3b). 89%, a yellow powder, mp
1
202–203 ꢁC (lit.3 mp 201–203 ꢁC); H NMR (300 MHz,
d-DMSO) d 1.18 (12H, d, J ¼ 6:6 Hz), 3.32 (2H, m),
7.26 (2H, s), 7.73 (1H, s), 9.01 (1H, s, –OH), 12.44 (1H,
br, –NH); 13C NMR (75 MHz, d-DMSO) d 22.7, 26.1,
118.8, 124.6, 126.1, 133.1, 135.9, 153.7, 167.4, 168.1. MS
m=z 305 (Mþ); HRMS calcd for C16H19NO3S: 305.1086;
found: 305.1085.
4.1.3. (Z)-5-(4-Hydroxy-3,5-dimethylbenzylidene)thiazol-
idine-2,4-dione (3c). 80%, a yellow powder, mp 266–
1
268 ꢁC; H NMR (300 MHz, d-DMSO) d 2.20 (6H, s),
4.2. Biology
7.17 (2H, s), 7.60 (s, 1H), 9.14 (1H, s, –OH), 12.42 (1H,
br, –NH); 13C NMR (75 MHz, d-DMSO) d 16.6, 118.8,
123.9, 125.1, 130.9, 132.5, 156.1, 167.5, 168.1; MS m=z
249 (Mþ); HRMS calcd for C12H11NO3S: 249.0460;
found: 249.0440.
4.2.1. Isolation of LDL. Blood was collected from nor-
malipidemic volunteers. EDTA was used as anti-coa-
gulant (1.5 mg/mL of blood). After low-speed
centrifugation of the whole blood to obtain plasma,
EDTA (0.1%), NaN3 (0.05%), and PMSF (0.015%) were
added to prevent lipoprotein modification. LDL was
isolated from the plasma by discontinuous density gra-
dient ultra centrifugation.25 Briefly, the plasma was
centrifuged at 100,000g at 4 ꢁC for 20 h. After the top
layers containing chylomicron and very low-density
lipoprotein (VLDL) were removed, the density of
remaining plasma fractions was increased to 1.064 with
NaBr solution and they were recentrifuged at 100,000g
for an additional 24 h. The LDL fraction in the top of
the tube was collected and dialyzed overnight against
three changes of phosphate buffer (pH 7.4) containing
NaCl (150 mM), in the dark at 4 ꢁC to remove NaBr and
EDTA. The LDL in PBS was stored at 4 ꢁC and used
within 4 weeks. The purity of the fraction was confirmed
by agarose gel electrophoresis and SDS-PAGE as de-
scribed elsewhere.26 Concentration of LDL protein was
determined by using bovine serum albumin (BSA) as a
standard.26
4.1.4. (Z)-5-(4-Hydroxy-2,3,5-trimethylbenzylidene)-
thiazolidine-2,4-dione (3d). 98%, a yellow powder, mp
1
257–259 ꢁC; H NMR (300 MHz, d-DMSO) d 2.11 (3H,
s), 2.18 (3H, s), 2.22 (3H, s), 7.01 (1H, s), 7.89 (1H, s),
8.88 (1H, s, –OH), 12.45 (1H, br, –NH); 13C NMR
(75 MHz, d-DMSO) d 12.6, 15.6, 16.8, 121.3, 122.1,
123.3, 124.3, 127.1, 131.1, 136.6, 155.3, 167.3, 168.5; MS
m=z 263 (Mþ); HRMS calcd for C13H13NO3S: 263.0616;
found: 263.0619.
4.1.5. (Z)-5-(3,4-Dihydroxybenzylidene)thiazolidine-2,4-
dione (3e). 78%, a yellow powder, mp >300 ꢁC; 1H
NMR (300 MHz, d-DMSO) d 6.85–7.00 (3H, m), 7.60
(1H, s), 9.46 (1H, s, –OH), 9.83 (1H, s, –OH), 12.42 (1H,
br, –NH); 13C NMR (75 MHz, d-DMSO) d 98.0, 116.3,
118.7, 124.0, 124.3, 132.7, 145.9, 148.6, 167.5, 168.2; MS
m=z 237 (Mþ); HRMS calcd for C10H7NO4S: 237.0096;
found: 237.0090.
4.2.2. Measurement of a-tocopherol in LDL. LDL was
added to the PBS buffer (10 mM, pH 7.4) at a final
concentration of 150 lg/mL. The LDL solution (1 mL)
was chilled in ice and 1 mL of ethanol containing 2 lg of
a-tocopherol acetate as an internal standard was added
to the LDL solution. The mixture was immediately ex-
tracted using 2 mL of hexane. After vortexing and cen-
trifugation at 1000g for 5 min at 4 ꢁC, 1.8 mL of organic
phase (hexane) was removed under nitrogen stream to
give the residue, which was dissolved in 80 lL of etha-
nol. Ethanol extracts (20 lL) was analyzed by HPLC
(Shimadzu SCL-10A, Tokyo, Japan) on C-18 column
(YMC-Pack Pro C18, 150 · 4.6 mm I.D., S-5 lM, YMC
Co., Ltd, Japan) with MeOH/H2O (98:2) as the mobile
phase at a flow rate 1 mL/min. a-Tocopherol was
monitored using a diode array detector at 200 or 280 nm
with retention time of 11.07 0.27 min and quantified
4.1.6. (Z)-5-(3,5-Difluoro-4-hydroxybenzylidene)thiazol-
idine-2,4-dione (3g). 75%, a yellow powder, mp 269–
1
270 ꢁC. H NMR (300 MHz, d-DMSO) d 7.28 (2H, dd,
J ¼ 7:8, 8.7 Hz), 7.68 (1H, s), 12.43 (1H, br, –NH); 13C
NMR (75 MHz, d-DMSO) d 113.7, 122.8, 123.4, 130.1,
150.5, 153.7, 167.3, 167.6; MS m=z 257 (Mþ); HRMS
calcd for C10H5F2NO3S: 256.9958; found: 256.9955.
4.1.7. (Z)-5-(4-Hydroxy-3,5-dimethoxybenzylidene)thi-
azolidine-2,4-dione (3f). 78%, a yellow powder, mp 250–
1
252 ꢁC; H NMR (300 MHz, d-DMSO) d 3.81 (6H, s),
6.88 (2H, s), 7.71 (1H, s), 9.34 (1H, s), 12.47 (1H, br,
–NH); 13C NMR (75 MHz, d-DMSO) d 56.0, 108.0,
119.6, 123.2, 132.9, 138.6, 148.2, 167.4, 168.0; MS m=z

4022
T.-S. Jeong et al. / Bioorg. Med. Chem. 12 (2004) 4017–4023
the amount of a-tocopherol using a standard plot of
peak area ratio (a-tocopherol/a-tocopherol acetate)
against a-tocopherol concentration. The concentration
of a-tocopherol in the native LDL solution (150 lg/mL)
was found to be 1.8 lM.
Diagnostics, Palo Alto, CA).30 Briefly, the LDL (120 lg/
mL) in phosphate-buffered saline, pH 7.4, was oxidized
with 10 lM of CuSO4 for 12 h at 37 ꢁC with or without
3a or probucol. Thereafter, the agarose gel (0.7% aga-
rose) was electrophoresed (85 V) in buffer containing
40 mM/L Tris, 40 mM/L glacial acetic acid, and 1 mM/L
EDTA for 1 h. After electrophoresis, lipoprotein bands
were stained with Coomassie Brilliant Blue, and REM
was defined as ratio of the distances migrated from the
origin by ox-LDL versus native LDL.
4.2.3. Cu2+-induced oxidation of LDL. Human plasma
LDL (120 lg protein/mL) was oxidized in 10 mM
phosphate-buffered saline with 10 lM CuSO4 and then
stopped by addition of 1 mM EDTA. In our experi-
ments, oxidation was carried out in the presence of 3a–h.
After the incubation, TBARS formation, conjugated
diene formation, relative electrophoretic mobility
(REM), fragmentation of the apoB-100 of LDL, and
radical DPPH scavenging activity were measured as
described below.
4.2.7. Electrophoresis of the apoB-100 fragmentation on
SDS-PAGE. After the oxidation with or without 3a,
sample were denatured with 3% SDS, 10% glycerol, and
2-mercaptoethanol at 95 ꢁC for 10 min. SDS-polyacryl-
amide gel electrophoresis (SDS-PAGE, 4%) was per-
formed to detect the apoB-100 fragmentation. The
electrophoresis was processed at 100 V for 80 min. After
the electrophoresis, the gel was dried and stained
with Coomassie Brilliant Blue R250.31 Density of each
spot was measured at 550 nm by a Shimadzu dual-
wavelength flying-spot scanner CS-9000.
4.2.4. TBARS assay. The LDL oxidation was deter-
mined spectrophotometrically by measuring the amount
of thiobarbituric acid reactive substances (TBARS).27
One milliliter of 20% (w/v) trichloroacetic acid (TCA)
and 1 mL of 0.67% (w/v) thiobarbituric acid (TBA) were
added to the post incubation mixture. The mixture was
heated at 100 ꢁC for 15 min and then cooled. After
centrifugation at 1500g for 15 min to remove precipi-
tated protein, the absorbance of the supernatant was
measured at 540 nm. The concentration of TBARS was
expressed as equivalents of 1,1,3,3-tetraethoxypropane
that was used as standard.
4.2.8. Radical DPPH scavenging activity. Radical DPPH
scavenging activity was studied by methanol solution of
compounds 3a, 3e, 3f, and 3g at 100 lM. Freshly made
DPPH radical solution (2 mL) was added into each 1 mL
of 3a, 3e, 3f, and 3g to result in 34%, 9%, 93%, and 87%
DPPH radicals remaining. Then, the final concentration
was 100 lM for DPPH radicals. The absorbance was
measured at 517 nm against a blank of pure methanol
including only DPPH radical for 20 min using the UV–
visible spectrophotometer at room temperature. Radical
DPPH scavenging capacity was calculated from the
difference in absorbance with each concentration of 3a,
3e, 3f, and 3g and expressed as percent DPPH radical
remaining, according to the following equation:
The inhibition ratio (%) was calculated using the fol-
lowing formula:
Inhibition ratio ð%Þ ¼ ðA ꢀ A1Þ=A ꢁ 100
where A was the absorbance of the control and A1 was
the absorbance of the tested sample.
% DPPH remaining ¼ 100 · (absorbance of sample/
4.2.5. Detection of conjugated dienes. The formation of
conjugated dienes was measured by monitoring of the
absorbance at 234 nm using an UV–vis spectrophoto-
meter (Hewlett Packard model 8453, Agilent Techno-
logies, Germany).28 Briefly, 3 mL of an LDL solution
(120 lg/mL) in phosphate-buffered saline, at pH 7.4, was
incubated 10 lM of CuSO4 at 37 ꢁC in the presence or
absence of 3a, thereafter the absorbance at 234 nm was
measured every 10 min. Probucol was used as a reference
antioxidant. The plot of absorbance against time pro-
duces three phases: (a) a lag phase, (b) a propagation
phase, and (c) a decomposition phase. The lag time (the
extent to which the compounds protected LDL from
oxidation was reflected by the prolongation of the lag
phase compared to that of control) was measured as the
intercept between the baseline and the tangent of the
absorbance curve during the propagation phase.29
absorbance of control).
4.3. Cell culture
Human monocytic THP-1 cells (ATCC) were cultured in
RPMI 1640 medium (Gibco/BRL) with phenol red
containing 10% fetal bovine serum (Gibco/BRL), 100 U/
mL penicillin, and 100 lg/mL streptomycin at 37 ꢁC
under 5% CO2 in air. Cells in RPMI 1640 medium with
serum and antibiotics were plated in 12-well plate
(1 · 106 cell/well in 1 mL). Differentiation of THP-1 cells
to macrophage was induced by treatment of phobal 12-
myristrate 13-acetate (PMA, 150 ng/mL, Sigma) for
3 days.32
4.3.1. Cell-mediated oxidation. THP-1 macrophage cells
were washed three times with serum-free RPMI 1640
media. The LDL (120 lg/mL) was added to the culture
medium and the medium was supplemented with 2 lM
CuSO4 to catalyze cell-mediated LDL oxidation in the
4.2.6. Relative electrophoretic mobility (REM). The
electrophoretic mobility of native or oxidized LDL was
detected by agarose gel electrophoresis (Ciba Corning

T.-S. Jeong et al. / Bioorg. Med. Chem. 12 (2004) 4017–4023
4023
13. Rimm, E. B.; Stampfer, M. J.; Ascherio, A.; Giovannucci,
E.; Colditz, G. A.; Willett, W. C. N. Engl. J. Med. 1993,
328, 1450.
14. (a) Zhu, Q. Y.; Huang, Y.; Tsang, D.; Chen, Z. Y. J.
Agric. Food Chem. 1999, 47, 2020; (b) Buttries, J. L.;
Diplock, A. T. Methods Enzymol. 1984, 105, 131.
15. Pinchuk, I.; Lichtenberg, D. Prog. Lipid Res. 2002, 41,
279, and references cited therein.
16. Ziouzenkova, O.; Sevanian, A.; Abuja, P. M.; Ramos, P.;
Esterbauer, H. Free Rad. Bio. Med. 1998, 24, 607.
17. Lazer, E. S.; Wong, H. C.; Possanza, G. J.; Graham,
A. G.; Farina, P. R. J. Med. Chem. 1989, 32, 100.
18. Mahony, L. R.; DaRooge, M. A. J. Am. Chem. Soc. 1967,
89, 5619.
presence or absence of 3a or probucol. Compound 3a
was dissolved in ethanol, in which ethanol concentration
was not in excess at least 0.5% of the medium. Equiva-
lent amounts of ethanol were added to control cell cul-
tures. Incubation was then carried out at 37 ꢁC for 24 h,
in a humidified atmosphere containing 5% CO2. After
supernatants were collected and centrifuged, the extent
of LDL oxidation was determined directly in the har-
vested medium using TBARS assay.
Acknowledgements
19. Brown, J. E.; Khodr, H.; Hider, R. C.; Rice-Evans, C. A.
Biochem. J. 1998, 330, 1173.
This work is supported by a grant (National Research
Laboratory Program) from the Ministry of Science and
Technology (M1-0302-00-0033), Korea.
20. (a) Tanaka, K.; Iguchi, H.; Taketani, S.; Nakata, R.;
Tokumaru, S.; Sugimoto, T.; Kojo, S. J. Biochem. 1999,
125, 173; (b) Hashimoto, R.; Matsukawa, N.; Nariyama,
Y.; Ogiri, Y.; Hamagawa, E.; Tanaka, K.; Usui, Y.;
Nakano, S.; Maruyama, T.; Kyotani, S.; Tsushima, M.;
Kojo, S. Biochim. Biophys. Acta 2002, 1584, 123; (c)
Matsukawa, N.; Nariyama, Y.; Hashimoto, R.; Kojo, S.
Bioorg. Med. Chem. 2003, 11, 4009.
References and notes
21. Blois, M. S. Nature 1958, 181, 1199.
1. Brown, F. C. Chem. Rev. 1961, 61, 463.
2. Momosw, Y.; Maekawa, T.; Yamano, T.; Kawada, M.;
Odaka, H.; Ikeda, H.; Sohda, T. J. Med. Chem. 2002, 45,
1518.
3. Unangst, P. C.; Connor, D. T.; Cetenko, W. A.; Sorenson,
R. J.; Sircar, J. C.; Wright, C. D.; Schrier, D. J.; Dyer,
R. D. Bioorg. Med. Chem. Lett. 1993, 3, 1729.
4. Li, A. C.; Glass, C. K. Nat. Med. 2002, 8, 1235.
5. Steinberg, D.; Parthasarathy, S.; Carew, T. E.; Khoo,
J. C.; Witztum, J. L. N. Engl. J. Med. 1989, 320, 915.
6. These results will be published.
22. Zhou, K.; Yu, L. J. Agri. Food Chem. 2004, 52, 1112.
23. (a) Heinecke, J. W.; Rosen, H.; Suzuki, L. A.; Chait, A. J.
Biol. Chem. 1987, 262, 10098; (b) Heinecke, J. W.; Baker,
L.; Rosen, H.; Chait, A. J. Clin. Invest. 1984, 74, 1890; (c)
Steinbrecher, U. P.; Parthasarathy, S.; Leake, D. S.;
Witztum, J. L.; Steinberg, D. Proc. Natl. Acad. Sci. U.S.A.
1984, 81, 3883.
24. (a) Sparrow, C. P.; Olszewski, J. J. Lipid Res. 1993, 34,
1219; (b) Rifici, V. A.; Schneider, S. H.; Khachadurian, A.
K. J. Nutr. 2002, 132, 2532.
25. Havel, R. J.; Edger, H. A.; Bradgon, J. H. J. Clin. Invest.
1955, 34, 1345.
26. Markwell, M. A. K.; Hass, S. M.; Tolbert, N. E.; Bieber,
L. L. Methods Enzymol. 1981, 72, 296.
7. Bjorkhem, I.; Henriksson-Freyschuss, A.; Breuer, O.;
Diczfalusy, U.; Berglund, L.; Henriksson, P. Arterioscler.
Thromb. 1991, 11, 15.
8. Kita, T.; Nagano, Y.; Yokode, M.; Ishii, K.; Kume, N.;
Ooshima, A.; Yoshida, H.; Kawai, C. Proc. Natl. Acad.
Sci. U.S.A. 1987, 84, 5928.
27. Liu, F.; Ng, T. B. Life Sci. 2000, 66, 725.
28. Esterbauer, H.; Striegl, G.; Puhl, H.; Rotheneder, M. Free
Radic. Res. Commun. 1989, 6, 67.
9. Carew, T. E.; Schwenke, D. C.; Steinberg, D. Proc. Natl.
Acad. Sci. U.S.A. 1987, 84, 7725.
29. Khursheed, P. N.; Enrique, B.; Charles, S. L. Atheroscle-
rosis 2000, 152, 89.
10. Nagano, Y.; Nakamura, T.; Matsuzawa, Y.; Cho, M.;
Ueda, Y.; Kita, T. Atherosclerosis 1992, 92, 131.
11. Daugherty, A.; Zweifel, B. S.; Schonfeld, G. Br. J.
Pharmacol. 1989, 98, 612.
30. Reid, V. C.; Mitchinson, M. J. Atherosclerosis 1993, 98,
17.
31. Miura, S.; Watanabe, J.; Tomita, T.; Sano, M.; Tomita, I.
Biol. Pharm. Bull. 1994, 17, 1567.
12. Sasahara, M.; Raines, E. W.; Chait, A.; Carew, T. E.;
Steinberg, D.; Wahl, P. W.; Ross, R. J. Clin. Invest. 1994,
94, 155.
32. Philippe, L.; Christiane, D.; Laure, P.; Martine, M.;
Sabine, G.; Brudi, M.; John, C. Arterioscler., Thromb.,
Vasc. Biol. 1997, 17, 979.