Synthesis and evaluation of curcumin analogues as potential thioredoxin reductase inhibitors

Article abstract of DOI:10.1016/j.bmc.2008.07.054

Series of curcumin derivatives were synthesized; the inhibitory activities on thioredoxin reductase (TrxR) of all analogues were evaluated by DTNB assay in vitro. It is found that most of the analogues can inhibit TrxR in the low micromolar range; Structure-activity relationship analysis reveals that analogues with furan moiety have excellent inhibitory effect on TrxR in an irreversible manner, indicating that the furan moiety may serve as a possible pharmacophore during the interaction of curcumin analogues with TrxR. The effect of selected curcuminoids on growth of different TrxR overexpressed cancer cell lines was also investigated and discussed.

Full text of DOI:10.1016/j.bmc.2008.07.054

Bioorganic & Medicinal Chemistry 16 (2008) 8035–8041
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis and evaluation of curcumin analogues as potential thioredoxin
reductase inhibitors
Xu Qiu a, , Zhong Liu a, , Wei-Yan Shao a, Xing Liu b, Da-Ping Jing a, Yan-Jun Yu a, Lin-Kun An a,
a
a
Shi-Liang Huang a, Xian-Zhang Bu a, , Zhi-Shu Huang , Lian-Quan Gu
*
a School of Pharmaceutical Sciences, Sun Yat-sen University, Guangzhou 510275, China
b Zhongshan School of Medicine, Sun Yat-sen University, Guangzhou, 510080, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Series of curcumin derivatives were synthesized; the inhibitory activities on thioredoxin reductase (TrxR)
of all analogues were evaluated by DTNB assay in vitro. It is found that most of the analogues can inhibit
TrxR in the low micromolar range; Structure-activity relationship analysis reveals that analogues with
furan moiety have excellent inhibitory effect on TrxR in an irreversible manner, indicating that the furan
moiety may serve as a possible pharmacophore during the interaction of curcumin analogues with TrxR.
The effect of selected curcuminoids on growth of different TrxR overexpressed cancer cell lines was also
investigated and discussed.
Received 18 May 2008
Revised 21 July 2008
Accepted 22 July 2008
Available online 24 July 2008
Keywords:
Curcumin derivatives
TrxR inhibitor
Furan moiety
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
the modification of TrxR by curcumin should be the mechanistic
explanation for the cancer preventive activity of curcumin.
Thioredoxin reductase (TrxR EC 1.8.1.9) belongs to a well-
known class of homodimeric pyridine nucleotide disulfide oxidore-
ductases.1 As part of a major thiol regulating system, TrxR cata-
lyzes NADPH-dependent reduction of the redox-active disulfide
in thioredoxin (Trx), which serves a wide range of functions in cell
proliferation and intracellular redox control.2,3 Many tumor cells
have shown elevated TrxR level.4–7 Moreover, it is believed that
TrxR plays an important role in tumor proliferation and drug resis-
tance of tumor cells.8–10 Inhibition of TrxR and its related redox
reactions may thus contribute to a successful cancer therapy.
Several effective natural and synthetic TrxR inhibitors are now
available, possessing antitumor potential ranging from induction
of oxidative stress to cell cycle arrest and apoptosis.11–13 Curcumin
(1,7-bis(4-hydroxy-3-methoxyphenyl)-1,6-heptadiene-3,5-dione),
a yellow spice and pigment isolated from the rhizome of Curcuma
longa, has been found useful in various bioactivities,14–17 such as
antioxidant, anti-inflammatory and anti-HIV protease activities.
Furthermore, curcumin was found to be a chemopreventive agent
of tumor initiation and proliferation.18–20 Many explanations on
the cancer preventative activity of curcumin have been pro-
posed.21,22 Arne Holmgren et al. first reported that the rat TrxR
activity could be inhibited by curcumin.23 It was proposed that
Although some reports on biological evaluation of curcumin ana-
logues have been published in recent years, studies on the curcu-
min analogues as TrxR inhibitors remain scarce up to now. To
search for useful drug candidates in the development of chemo-
therapeutic agents, we thought it would be interesting to evaluate
some symmetrical and unsymmetrical curcumin analogues with
different substituents on aromatic or heterocyclic rings for their
TrxR inhibitory effect and cancer cell growth inhibitory activity.
2. Results and discussion
2.1. Chemistry
Various classes of unsymmetrical analogues 1a–1e, 2a–2g, 3a–
3b were designed by altering the substituted group, aryl ring and
ketene moiety respectively (Scheme 1).
All of the unsymmetrical curcumin derivatives were synthe-
sized according to a solid phase synthetic route we developed in
our previous letter.24
To extend the structure diversity of curcumin analogues for fur-
ther structure–activity relationship (SAR) information, several
symmetrical curcumin derivatives 4a–4h also designed and syn-
thesized by treating the methene protected acetyl acetone with
aromatic aldehydes in the presence of tributyl borate and n-butyl-
amine (Scheme 2).
* Corresponding author. Tel./fax: +8620 39332671.
E-mail address: phsbxzh@mail.sysu.edu.cn (X.-Z. Bu).
These authors made an equal contribution to this paper.
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.07.054

8036
X. Qiu et al. / Bioorg. Med. Chem. 16 (2008) 8035–8041
OH
O
OH
O
R1
R2
R3
R2
R3
X
HO
R5
R1
R4
R4
R1 = -CH3, X = O, R2 = H, R3 = -OH, R4 = -OCH3
2a
1a
R1 = -OCH3, R2 = R3 = R4 = R5 = H
2b R1 = R2 = R4 = H, X = S, R3 = -OH
1b R1 = -OCH3, R3 = -OH, R2 = R4 = R5 = H
1c R1 = R5 = H, R2 = R3 = R4 = -OCH3
1d R1 = R2 = R3 = R4 = H, R5 = Br
1e R1 = R2 = R3 = R4 = R5 = H
2c R1 = -CH2OH, X = O, R2 = R3 = R4 = -OCH3
2d R1 = -CH2OH, X = O, R2 = R4 = H, R3 = -N(CH3)2
2e R1 = -CH2OH, X = O, R2 = H, R3 = -OH, R4 = -OCH3
2f R1 = -CH2OH, X = O, R2 = R3 = R4 = H
O
2g R1 = -CH2OH, X = O, R2 = R4 = H, R3 = -OCH3
R1
R1 = R2 = -OCH3
3b R1 = -OH, R2 = H
3a
HO
R2
OCH3
Scheme 1. Structure of unsymmetrical curcumin analogues.
OH
O
O
R3
R2
R3
R2
4a R1 = R2 = R3 = H
4b
a, b, d
a, c, d
R1 = H, R2 = -OH, R3 = -OCH3
4c R1 = R3 = H, R2 = -OH
4d
O
O
1 = R2 = R3 = -OCH3
4e R1 = R3 = H, R2 = -OCH3
4f
R1
R1
R
OH
R
1 = R2 = -OCH3, R3 = H
4g R = -CH3
4h R = H
O
O
R
R
R1
R2
O
CHO
a: B2O3; b: (n-BuO)3B, n-C4H9NH2,
c: (n-BuO)3B, n-C4H9NH2,R
R3
d: HCl
CHO
Scheme 2. Synthesis of symmetrical curcumin derivatives 4a–4h.
2.2. Evaluation of the TrxR inhibitory activities of curcumin
analogues
Noticeably, all unsymmetrical compounds with a furan moiety
(2a, 2c,2e–2g) except 2d exerted a much higher inhibitory activity
than that of the natural curcumin. Symmetrical analogues 4g and
4h, which bear furan moieties on both terminals of the molecule,
were designed and synthesized therefore based on this primary re-
sult. Expectedly, both of 4g and 4h, especially 4g, showed excellent
inhibitory activities against TrxR, indicating that the furan moiety
may serve as a possible pharmacophore. It may be also due to the
interaction of furan moiety with TrxR by hydrogen bonding, which
induced the conformational changes of TrxR, making the active sites
of TrxR easier to access. Compared to 2f, compound with thiophene
moiety (2b) did not increase the activity. Interestingly, the introduc-
tion of dimethylamino group at the benzene moiety (namely 2d) de-
creased the activity significantly compared with 2f in spite of the
furan moiety. It might be explained that the high polarity of proton-
ized 2d led to the negative inhibitory effect on TrxR.
The TrxR inhibitory activities of all curcumin analogues were
determined with 5,50-dithiobis-(2-nitrobenzoic acid) (DTNB) as-
say.25,26 The IC50 (half maximal inhibitory concentration) values
are summarized in Table 1. The results showed that most of
unsymmetrical curcumin analogues exhibited excellent inhibitory
effects on TrxR, while all the tested traditional symmetrical curcu-
min analogues (4a, 4c–4f) except the natural curcumin (4b)
showed poor activities. Compounds 4a, 4d–4f do not possess
hydroxylated moiety, compared with the TrxR inhibition activity
of other hydroxylated compounds, this indicates that the hydroxyl-
ated moiety contribute the inhibitory activity. On the other hand,
the low activity of dihydroxylated compound 4c may result from
its relatively higher polarity.
It should be noted that the IC50 value for the curcumin was
Analogues with dimethylenecyclohexanone moiety (3a, 3b) dis-
played activities against TrxR; however, no evidence supported
that dimethylenecyclohexanone moiety would be a positive or
negative factor to the inhibitory activity, since the activity of 1b/
3b and 1c/3a showed ruleless and had no apparent coherence.
determined to be 38.3 lM, which was approximately 10-fold larger
than what was reported earlier.23 We reasoned that this result was
due to the different incubation time of curcumin with TrxR during
the inhibition assay; 4 min assay was performed in current studies
and 2 h assay was performed in the earlier report. We noted a sim-
ilar result that the natural curcumin could inhibit TrxR in a time-
dependent manner (data not shown), and with an IC50 value of
2.3. Irreversible inhibition of TrxR by curcumin analogue 4g
4.2
l
M after a 2-h incubation time.
It has been well known that mammalian TrxR contains a C-ter-
minal selenocysteine active site, which is essential for catalysis and
can be irreversibly modified by many electrophilic agents.27 To
investigate the nature of the TrxR inhibition by curcumin ana-
logues, we determined the reversibility of TrxR inactivation by cur-
cumin analogue 4g. As illustrated in Figure 1, when 4g was
incubated with TrxR for 2 h and then subjected to ultrafiltration,
the TrxR activity was found unrecoverable. This suggested that
As a whole, the unsymmetrical analogues with benzene rings
showed moderate inhibitory effect on TrxR. As compared to 1e,
the introduction of a halide group Br at the R5 (namely 1d) de-
creased the inhibitory activity. In addition, the introduction of
one terminal hydroxyl or/and methoxyl to 1e did not contribute
to the inhibitory activity (1a, 1b); on the contrary, it led to a
decrease of inhibitory activity for 1a.

X. Qiu et al. / Bioorg. Med. Chem. 16 (2008) 8035–8041
8037
M)
Table 1
Inhibition of TrxR by curcumin analogues
Entry
IC50
(lM)
Entry
IC50 (l
OH
O
OH
O
R1
R2
R3
R2
R3
HO
R5
X
R4
R1
R4
R4
R1
R2
R3
R4
R5
R1
R2
R3
X
1a
1b
OCH3
OCH3
H
H
H
OH
H
H
H
H
57.0
20.0
2a
2b
2c
2d
2e
2f
CH3
H
H
H
OCH3
H
H
OH
OH
OCH3
N(CH3)2
OH
OCH3
H
OCH3
H
OCH3
H
H
O
S
0.5
20.0
2.0
62.2
1.0
CH2OH
CH2OH
CH2OH
CH2OH
CH2OH
O
O
O
O
O
1c
H
OCH3
OCH3
OCH3
H
13.8
1d
1e
H
H
H
H
H
H
H
H
Br
H
50.0
23.3
H
H
H
OCH3
1.3
0.8
2g
Entry
IC50
(l
M)
Entry
IC50 (lM)
O
OH
O
R3
R2
R3
R2
O
HO
R1
R2
R1
R1
R1
R2
R1
R2
R3
3a
3b
OCH3
OH
OCH3
H
24.9
5.1
4a
H
OCH3
H
OH
H
H
>100
38.3
4b (Curcumin)
Entry
IC50
(l
M)
4c
4d
4e
H
OCH3
H
OH
OCH3
OCH3
H
OCH3
H
>100
>100
>100
OH
O
O
O
R
R
R
4f
OCH3
OCH3
H
>100
4g
4h
CH3
H
0.3
1.6
and human MCF-7 breast adenocarcinoma cells were examined
using insulin reduction assay after the treatment with compound
4g. As shown in Fig. 2A, TrxR activity in cell lysates declined by
100
80
60
40
20
0
approximately 30% after the exposure of HeLa cells to 50 lM of
4g. Similar findings were observed in 4g treated MCF-7 cells (Fig.
2B). These results demonstrated that curcumin analogue could re-
duce TrxR activities in both HeLa and MCF-7 cancer cells at rela-
tively high concentrations.
2.5. Cytotoxicity evaluation of selected curcumin analogues
Our present study shows that the curcuminoids, especially ana-
logues bearing furan moiety, can irreversibly inhibit mammalian
TrxR under physiologic conditions. We further investigated the ef-
fect of selected curcuminoids on the growth of TrxR overexpressed
cancer cell lines including human A549 lung adenocarcinoma epi-
thelial cell line, cis-diamminedichloroplatinum (II) (CDDP) resis-
tant A549/R cell line, doxorubicin resistant MCF-7/R cell line, and
human HepG2 hepatocellular liver carcinoma cell line using MTT
method (Fig. 3 and Table 2). The results showed that analogues
2a, 2e, 2g, and 4g, which turned out to be potent inhibitors of TrxR,
exhibited stronger toxicity to A549/R cells than that of the natural
curcumin (Fig. 3B), suggesting that the cell death of A549/R in-
duced by curcuminoids may result from the inhibition of TrxR sys-
tem; unfortunately, the most potent TrxR inhibitor 4g only showed
moderate inhibitory activities on the growth of MCF-7/R, A549/R,
and A549. Compounds 2e and 2g could also inhibit the growth of
the MCF-7/R at relatively lower micromolar concentrations; how-
ever, the cytotoxicity of 2e and 2g did not increase at higher con-
centrations, suggesting that some other induction systems
10μM
40μM
Control
Figure 1. TrxR could be irreversibly inhibited by compound 4g. NADPH-reduced
TrxR was incubated with indicated concentrations of IG3 at room temperature. The
enzyme solutions were subjected to ultrafiltration after 2 h. TrxR activity was
expressed as the percentage of the control in the absence of 4g. Data are presented
as means SD of triplicate experiments. *p < 0.001 compared with the control.
TrxR was irreversibly inhibited by curcumin analogue 4g. It might
be due to the covalent modification of the nucleophilic C-terminal
active site in TrxR by a,b-unsaturated ketone moieties in curcumin
analogues, as Arne Holmgren et al. proposed.23
2.4. Effects of curcumin analogue on TrxR activity in cancer
cells
To further investigate the effect of curcumin analogues in can-
cer cells, TrxR activities in human HeLa cervical carcinoma cells

8038
X. Qiu et al. / Bioorg. Med. Chem. 16 (2008) 8035–8041
100
80
60
40
20
0
100
*
*
80
60
40
20
0
Control
10 μM
50 μM
Control
10 μM
50 μM
[4g] (μM)
[4g] (μM)
Figure 2. Effects of curcumin analogue 4g on TrxR in HeLa and MCF-7 cells. HeLa cells (A) and MCF-7 cells (B) were cultured in the absence or presence of various
concentrations of 4g for 12 h. TrxR activity in cell lysates was measured by insulin reduction assay as described in Section 4. Activities are given as the percentage of the
control. Data are means SD of at least three determinations. *p < 0.05 compared with the control.
100
1c
1e
2a
2c
2e
curcumin
2c
2d
2e
2f
40
20
0
3b
4g
4h
50
0
2g
4g
-20
10
32
100
1
3
10
32
100
Concentration μM
Concentration μM
100
80
60
40
20
0
2a
2e
2g
4e
4g
curcumin
80
60
40
20
0
2e
2g
4g
4h
4e
1
3
10
32
100
10
32
100
Concentration μM
Concentration μM
Figure 3. Effect of different curcuminoids on growth of TrxR overexpressed cancer cell lines A549 (A), cis-diamminedichloroplatinum (II) (CDDP) resistant A549/R (B), HepG2
(C), and doxorubicin resistant MCF-7/R (D). Activities are given as the percentage of the control. Data are means SD of at least three determinations.
against 2g and 2e might be functional in MCF-7/R cells. Interest-
ingly, HepG2 seemed to be more resistant to curcuminoids as com-
pared to the other cell lines (Fig. 3C) since all of the curcumin
analogues could not induce effectively cell death even at concen-
resistant cell lines have higher TrxR level than sensitive cells; it
might render the cell lines to be more dependent on the TrxR
activities. Nevertheless, further investigation should be done to
unclose this phenomenon.
tration up to 100
l
M.
In summary, we synthesized series of curcumin derivatives; the
inhibitory activities on TrxR of all analogues were evaluated by
DTNB assay. Most of the curcumin analogues were found to show
inhibitory activity against TrxR in vitro. Especially, the analogues
with furan moiety have excellent activities against TrxR in an irre-
versible manner, indicating that the furan moiety may act as a pos-
sible pharmacophore during the interaction of analogues with
TrxR. The effects of selected curcuminoids on the growth of TrxR
overexpressed cancer cell lines were investigated and discussed.
The detailed mechanism is under active exploration.
4g and 4h have moderate inhibitory activities on the growth of
both MCF-7/R and A549 (Fig. 3A). Noticeably, while compounds
with relatively higher IC50 on TrxR (curcumin, 1e and 1c) can in-
duce the A549 cell death effectively, most of the analogues with
low IC50 on TrxR (2a, 2e, 2g, and 4h) have lower toxicities to
A549 and/or HepG2 than to drug resistant cell lines A549/R and
MCF-7/R (Table 2); the reason is not clear; it probably owns to
the multi-mechanisms of cell death induced by curcuminoids.
Furthermore, it was reported28,29 that the CDDP or doxorubicin

X. Qiu et al. / Bioorg. Med. Chem. 16 (2008) 8035–8041
8039
Table 2
described above for 4a. 3.04 g (20 mmol) of 4-hydroxy-3-methoxy-
benzaldehyde was used and 2.88 g of product was obtained as yel-
low powder, yield: 78%; 1H NMR(500 MHz, CDCl3) d: 3.94 (s, 3 H),
5.79 (s, 1H, enol), 6.46 (d, 2H, J = 16 Hz), 6.92 (d, 2H, J = 8 Hz), 7.04
(d, 2H, J = 1.5 Hz), 7.11(dd, 2H, J = 8, 1.5 Hz), 7.58 (d, 2H, J = 16 Hz);
MS (ESI) m/z: 369.1 (M+H)+.
Effect of different curcuminoids on growth of TrxR overexpressed cancer cell lines
a
Compounds
TrxR inhibition (IC50
,
l
M)
Cytotoxicity (GC50
,
lM)
A549
A549/R
MCF-7/R
HepG2
1c
1e
2a
2c
2d
2e
2f
2g
3b
4e
4g
4h
13.8
23.3
0.5
2.0
62.2
1.0
1.3
0.8
5.1
>100
0.3
60.1
41.3
>100
>100
ND
>100
ND
ND
>100
ND
52.9
66.4
18.2
ND
ND
ND
ND
ND
ND
ND
39.4
ND
ND
ND
>100
>100
>100
>100
>100
ND
ND
>100
ND
3.2.2.3. 5-Hydroxy-1,7-bis(4-hydroxyphenyl)hepta-1,4,6-trien-
3-one (4c). 4c was prepared from 4-hydroxybenzaldehyde by
using the similar procedure described above for 4a. 2.44 g
(20 mmol) of 4-hydroxybenzaldehyde was used and 2.24 g of
product was obtained as yellow powder, yield: 73%; 1H
NMR(500 MHz, DMSO-d6, ppm) d:5.97 (s, 1H, enol), 6.57 (d, 2H,
J = 16 Hz), 6.77 (d, 4H, J = 9 Hz), 7.45 (d, 4H, J = 9 Hz), 7.46 (d, 2H,
J = 16 Hz); MS(ESI) m/z: 309.3 (M+H)+.
ND
ND
35.8
ND
10.6
ND
12.1
ND
10.0
ND
>100
59.6
ND
>100
48.5
33.6
ND
1.6
38.3
Curcumin
73.0
ND
ND, not determined.
3.2.2.4. 5-Hydroxy-1,7-bis(3,4,5-trimethoxyphenyl)hepta-1,4,6-
trien-3-one (4d). Compound 4d was prepared from 3,4,5-tri-
methoxybenzaldehyde by using the similar procedure described
above for 4a. 3.92 g (20 mmol) of 3,4,5-trimethoxybenzaldehyde
was used and 3.82 g of product was obtained as orange powder,
yield: 84%; 1H NMR(500 MHz, CDCl3) d:3.71 (s, 6H), 3.84 (s, 12H),
6.17(s, 1H, enol), 6.91(d, 2H, J = 16 Hz), 7.07 (s, 4H), 7.59(d, 2H,
J = 16 Hz); MS(ESI) m/z: 457.1(M+H)+.
a
The GC50 represents the compound concentration required for the reduction of
the mean cell viability to 50%.
3. Experiment
3.1. General
All reagents are available commercially. Solvents were purified
using standard techniques. Curcumin and its analogues used in this
work were all synthesized and characterized in our lab. TrxR was
purified from porcine brains as described by Cheung et al.30 The
enzyme purity was judged by SDS–PAGE and IEF. A VarianUNITY
INOVA 500 MHz spectrometer was used to record the 1H NMR
spectra in which TMS was the internal standard. The MS analysis
3.2.2.5. 5-Hydroxy-1,7-bis(4-methoxyphenyl)hepta-1,4,6-trien-
3-one (4e). Compound 4e was prepared from 4-methoxybenzal-
dehyde by using the similar procedure described above for 4a.
2.72 g (20 mmol) of 4-methoxybenzaldehyde was used and
2.94 g of product was obtained as yellow powder, yield: 87%; 1H
NMR(400 MHz, CDCl3) d: 7.62 (d, 2H, J = 15.6 Hz), 7.51 (dd, 4H,
J = 7.8,2 Hz), 6.92(dd, 4H, J = 7.8, 2 Hz), 6.50 (d, 2H, J = 15.6 Hz),
5.78(s, 1H), 3.85(s, 6H); MS(APCI) m/z: 337.1(M+H)+.
was performed on
spectrometer.
a Finnigan LCQ Deca XP ion trap mass
3.2.2.6. 1,7-Bis(3,4-dimethoxyphenyl)-5-hydroxyhepta-1,4,6-
trien-3-one (4f). Compound 4f was prepared from 3,4-dime-
thoxybenzaldehyde by using the similar procedure described
above for 4a. 3.32 g (20 mmol) of 3,4-dimethoxybenzaldehyde
was used and 3.55 g of product was obtained as orange powder,
yield: 90%; 1H NMR (400 MHz, CDCl3) d: 7.61(d, 2H, J = 16 Hz),
7.14(dd, 2H, J = 2, 8.4 Hz), 7.07(d, 2H, J = 2 Hz), 6.88(d, 2H,
J = 8.4 Hz), 6.50(d, 2H, J = 16 Hz), 5.82(s, 1H), 3.94(s, 3H), 3.92(s,
3H); MS(APCI) m/z: 397.2(M+H)+.
3.2. Synthesis of curcumin analogues
3.2.1. Synthesis of unsymmetrical curcumin analogues 1a–e,
2a–g, 3a–b
All of the unsymmetrical curcumin derivatives 1a–e, 2a–g, 3a–
b were synthesized according our previous work, and the 1H NMR,
HRMS or MS and EA have been described in our early report.24
3.2.2. Synthesis of curcumin analogues 4a–4h
3.2.2.1. 5-Hydroxy-1,7-diphenylhepta-1,4,6-trien-3-one (4a).
Compound 4a was prepared according to Pedersen method31
with slight modification. Boric anhydride (0.35 g, 5 mmol) was
suspended in 10 mL EtOAc in the presence of acetylacetone
(1.00 g, 10 mmol), and the mixture was stirred for 3 h at 70 °C.
After removing the solvent, the resultant white solid was washed
with hexane and then, 20 ml EtOAc, benzaldehyde (2.12 g,
20 mmol), and tributyl borate (4.60 g, 20 mmol) were added,
and the mixture was stirred for a further 30 min. Butylamine
(73 mg, 1 mmol) dissolved in EtOAc was added dropwise for
15 min. The mixture reacted at 70 °C for 24 h. Then 1 N HCl was
added to adjust the pH to 5, and the solution was heated for
30 min at 60 °C. EtOAc was used to extract the product from the
water layer. The 4a was purified by recrystalization from EtOAc
to yield yellow crystals: 2.21 g (80% yield); 1H NMR(500 MHz,
DMSO-d6, ppm) d:6.19 (s, 1H, enol), 6.86 (d, 2H, J = 16 Hz), 7.41
(m, 6H), 7.61(d, 2H, J = 16 Hz), 7.66 (dd, 4H, J = 8, 2 Hz); MS(ESI)
m/z: 277.1 (M+H)+.
3.2.2.7. 5-Hydroxy-1,7-bis(5-methylfuran-2-yl)hepta-1,4,6-
trien-3-one (4g). Compound 4g was identified as a new com-
pound and prepared by using the similar procedure described
above for 4a. Boric anhydride (0.35 g, 5 mmol) and acetylace-
tone (1.00 g, 10 mmol) were suspended in 10 mL EtOAc, and
the mixture was stirred for 3 h at 70 °C. After removing the sol-
vent, the resultant solid was washed with hexane and then,
20 ml of EtOAc, 5-methylfuran-2-carbaldehyde (2.20 g,
20 mmol), and tributyl borate (4.60 g, 20 mmol) were added,
and the mixture was stirred for a further 30 min. Butylamine
(73 mg, 1 mmol) dissolved in EtOAc was added dropwise in
15 min. The mixture reacted at 70 °C for 24 h. Then 1 N HCl
was added to adjust the pH to 5, and the solution was heated
for 30 min at 60 °C. EtOAc was used to extract the product from
the water layer. The 4g was purified by recrystalization from
EtOAc to yield 2.55 g red crystal, yield 90%; 1H NMR(400 MHz,
CDCl3) d: 2.36 (s, 6H), 5.70 (s, 1H), 6.08 (dd, 2H, J = 3.2 Hz), 6.42
(d, 2H, J = 15.6 Hz), 6.50 (d, 2H, J = 3.2 Hz), 7.33(d, 2H, J =
15.6 Hz); 13C NMR(100 MHz, CDCl3) d: 183.0, 155.6, 150.6,
126.7, 120.3, 116.5, 109.1, 101.9, 13.9; MS(EI) m/z:
285.1(M+H)+; HRMS-EI m/z: calcd for C17H16O4, 284.1043;
found 284.1044.
3.2.2.2. 5-Hydroxy-1,7-bis(4-hydroxy-3-methoxyphenyl)hepta-
1,4,6-trien-3-one (4b). Compound 4b was prepared from 4-hy-
droxy-3-methoxybenzaldehyde by using the similar procedure

8040
X. Qiu et al. / Bioorg. Med. Chem. 16 (2008) 8035–8041
3.2.2.8. 1,7-Di(furan-2-yl)-5-hydroxyhepta-1,4,6-trien-3-one (4h)
Compound 4h was prepared from furan-2-carbaldehyde by
conducted in quadruplicates. For each dose, the mean cell viability
was expressed as a percentage of the control. The GC50 represents
the compound concentration required for the reduction of the
mean cell viability to 50%.
.
using the similar procedure described above for 4a, 1.92 g
(20 mmol) furan-2-carbaldehyde was used and the final product
was purified by column chromatography to yield 0.76 g yellow
powder, yield: 30%; 1H NMR (400 MHz, CDCl3) d: 7.48(d, 2H,
J = 1.6 Hz), 7.40(d, 2H, J = 15.6 Hz), 6.60 (d, 2H, J = 3.2 Hz), 6.51 (d,
2H, J = 15.6 Hz), 6.47 (dd, 2H, J = 1.6, 3.2 Hz), 5.74 (s, 1H); MS(APCI)
m/z: 257.1(M+H)+.
3.7. TrxR activity in cell lysates
Cells (1.5 ꢁ 106) were seeded in 6-well plates in 5 ml of RPMI
1640 medium. Different concentrations of curcumin analogues
were incubated with the cells for 12 h, and the control cells were
treated with less than 1% DMSO. In the harvesting, cells were tryp-
sinized and lysed in cell lysis buffer (50 mM Tris–HCl, pH 7.4,
150 mM NaCl, 1 mM NaF, 1% Triton X-100, 0.5% sodium deoxycho-
late, 0.1% SDS) in the presence of protease inhibitors (Roche). After
centrifugation at 16,000g for 10 min, protein concentrations of
supernatants were determined using the Bio-Rad assay kit. TrxR
activity was determined by end point insulin reduction assay as
described by Arner et al.25 with slight modifications. Briefly,
3.3. The inhibition assay of TrxR by curcumin analogues
For determining the TrxR inhibitory activity of curcumin ana-
logues, the DTNB reduction assay was employed. All assays were
conducted at 37 °C in a total volume of 600
ment, 18 l of TrxR was added to an assay mixture containing
100 mM potassium phosphate, 2 mM EDTA, pH 7.4, and 46 l of
NADPH and 5 l of inhibitor at various concentrations. After
4 min pre-incubation, the reaction was initiated with the addition
of 18 l of DTNB. The control was incubated with the same amount
ll. In each measure-
l
l
l
38
2 mM EDTA, 400
(pH 7.4) at 37 °C for 30 min. The reaction was terminated by the
addition of 200 l of 6 M guanidine hydrochloride/1 mM DTNB.
lg of extract were incubated with 2.5 mg/ml bovine insulin,
l
l
M NADPH, 0.8 M Trx in 60 l of 100 mM HEPES
l
l
of DMSO. The increase in absorbance at 412 nm
(De412 =
13.6 mMꢀ1 cmꢀ1) was monitored in the initial 80 s. The IC50 values
were calculated to represent the TrxR inhibitory effect of
compounds.
l
The reaction mixtures with the omission of Trx were used as the
control. The absorbance was measured at 412 nm. The percentage
of TrxR activity in comparison with the control was determined.
3.4. Reversibility of inhibition of TrxR by curcumin analogue 4g
Acknowledgments
Different concentrations of 4g (10
l
M and 40
lM, respectively)
were incubated with NADPH-reduced TrxR (0.8
l
M) at room tem-
We are indebted to the Science and Technology Key Project of
The Education Ministry of PR China (105133), the Guangdong Pro-
vincial Natural Science Foundation (5001773), and The Hong Kong
Polytechnic University Area of Strategic Development Fund for
financial support of this study.
perature for 2 h. The control was treated with the same amounts of
DMSO. Compound 4g was then removed by filtering through a Mil-
lipore Centricon-30 centrifugal filter with 30,000 nominal molecu-
lar weight limits. The enzyme in the retentate was washed with
70
50
l
l of PE buffer and then resuspended in 60
ll of PE buffer;
ll of the mixture was taken out to determine the enzyme activ-
References and notes
ity as described in Table 1. The activity was given as the percentage
of the control.
1. Arner, E. S.; Holmgren, A. Eur. J. Biochem. 2000, 267, 6102–6109.
2. Mustacich, D.; Powis, G. Biochem. J. 2000, 346, 1–8.
3. Gasdaska, P.; Berggren, M.; Berry, M.; Powis, G. FEBS Lett. 1999, 442, 105–111.
4. Mohler, J. L.; Morris, T. L.; Ford, O. H.; Alvey, R. F.; Sakamoto, C. Prostate 2002,
51, 247–255.
3.5. Cell culture
5. Gan, L.; Yang, X. L.; Liu, Q.; Xu, H. B. J. Cell Biochem. 2005, 96, 653–664.
6. Yoo, M. H.; Xu, X. M.; Carlson, B. A.; Gladyshev, V. N.; Hatfield, D. L. J. Biol. Chem.
2006, 281, 13005–13008.
7. Ueno, M.; Masutani, H.; Arai, R. J.; Yamauchi, A.; Hirota, K.; Sakai, T.; Inamoto,
T.; Yamaoka, Y.; Yodoi, J.; Nikaido, T. J. Biol. Chem. 1999, 274, 35809–35815.
8. Björkhem-Bergman, L.; Jönsson, K.; Eriksson, L. C.; Olsson, J. M.; Lehmann, S.;
Paul, C.; Björnstedt, M. Biochem. Pharmacol. 2002, 63, 1875–1884.
9. Urig, S.; Becker, K. Semin. Cancer Biol. 2006, 16, 452–465.
10. Becker, K.; Herold-Mende, C.; Park, J. J.; Lowe, G.; Schirmer, R. H. J. Med. Chem.
2001, 44, 2784–2792.
11. Rigobello, M.; Messori, L.; Marcon, G.; Cinellu, M. A.; Bragadin, M.; Folda, A.;
Scutari, G.; Bindoli, A. J. Inorg. Biochem. 2004, 98, 1634–1641.
12. Cenas, N.; Nivinskas, H.; Anusevicius, Z.; Sarlauskas, J.; Lederer, F.; Arnér, E. S. J.
J. Biol. Chem. 2004, 279, 2583–2592.
13. Wipf, P.; Lynch, S. M.; Birmingham, A.; Tamayo, G.; Jimenez, A.; Campos, N.;
Powis, G. Org. Biomol. Chem. 2004, 2, 1651–1658.
Four different human cancer cell lines (human A549 lung ade-
nocarcinoma epithelial cell line (CDDP resistant A549/R) human
MCF-7 breast carcinoma cells (doxorubicin resistant MCF-7/R),
and human HEPG2 hepatocellular liver carcinoma cell line, and
human HeLa cervical cancer cells) were cultured in RPMI 1640
medium (Gibco-BRL Grand Island, NY) supplemented with 10%
fetal calf serum, 100 U/ml penicillin, 100
lg/ml streptomycin,
and 2 mM -glutamine at 37 °C in incubator with humidified
L
atmosphere of 5.0% CO2 and 95% air.
3.6. Cell growth inhibition assay
14. Nurfina, A. N.; Reksohadiprodjo, M. S.; Timmerman, H.; Jenie, U. A.; Sugiyanto,
D.; Van der Goot, H. Eur. J. Med. Chem. 1997, 32, 321–328.
15. Maheshwari, R. K.; Singh, A. K.; Gaddipati, J.; Srimal, R. C. Life Sci. 2006, 78,
2081–2087.
16. Motterlini, R.; Foresti, R.; Bassi, R.; Green, C. J. Free Radic. Biol. Med. 2000, 28,
1303–1312.
17. Manikandan, P.; Sumitra, M.; Aishwarya, S.; Manohar, B. M.; Lokanadam, B.;
Puvanakrishnan, R. Int. J. Biochem. Cell Biol. 2004, 36, 1967–1980.
18. Gafner, S.; Lee, S.-K.; Cuendet, M.; Barthelemy, S.; Vergnes, L.; Labidalle, S.;
Mehta, R. G.; Boone, C. W.; Pezzuto, J. M. Phytochemistry 2004, 65, 2849–2859.
19. Pillai, G. R.; Srivastava, A. S.; Hassanein, T. I.; Chauhan, D. P.; Carrier, E. Cancer
Lett. 2004, 208, 163–170.
The growth inhibitory effects of curcumin analogues on differ-
ent cancer cell lines were evaluated by means of the MTT (3-
(4,5-dimethylthiazol- 2-yl)-2,5-diphenyl-2H-tetrazolium bromide)
assay as described by Carmichael32 and Mossman33 with modifica-
tions. The cells were plated at a density of 1 ꢁ 104 per well in 96-
well microplates and allowed to incubate overnight. Curcumin
analogues were added to the wells at increasing concentrations
(0–100 lM). After 48 h, each well was treated with 20 ll of
5 mg/ml MTT solution, and the cells were further incubated at
37 °C for 4 h. At the end of incubation, the untransformed MTT
was removed and 150 ll of DMSO was added. The microplates
were shaken well to dissolve the formazan dye, and the absorbance
at 570 nm was measured using a microplate-reader (Bio-Rad).
Cytoviability of the control cells was taken as 100%. Tests were
20. Ohtsu, H.; Xiao, Z.; Ishida, J.; Nagai, M.; Wang, H. K.; Itokawa, H.; Su, C. Y.; Shih,
C.; Lee, Y. F.; Tsai, M. Y.; Chang, C.; Lee, K. H. J. Med. Chem. 2002, 45, 5037–5042.
21. Aggarwal, B. B.; Kumar, A.; Bharti, A. C. Anticancer Res. 2003, 23, 363–398.
22. Gururaj, A.; Belakavadi, M.; Venkatesh, D.; Marme, D.; Salimath, B. Biochem.
Biophys. Res. Commun. 2002, 297, 934–942.
23. Fang, J.; Lu, J.; Holmgren, A. J. Biol. Chem. 2005, 280, 25284–25290.

X. Qiu et al. / Bioorg. Med. Chem. 16 (2008) 8035–8041
8041
24. Shao, W. Y.; Cao, Y. N.; Yu, Z. W.; Pan, W. J.; Qiu, X.; Bu, X. Z.; An, L. K.; Huang, Z.
S.; Gu, L. Q.; Chan, A. S. C. Tetrahedron Lett. 2006, 47, 4085–4089.
25. Arner, E. S.; Zhong, L.; Holmgren, A. Methods Enzymol. 1999, 300, 226–239.
26. Gromer, S.; Merkle, H.; Schirmer, R. H.; Becker, K. Methods Enzymol. 2002, 347,
382–394.
27. Sandalova, T.; Zhong, L.; Lindqvist, Y.; Holmgren, A.; Schneider, G. Proc. Natl.
Acad. Sci. U.S.A. 2001, 98, 9533–9538.
28. Sasada, T.; Nakamura, H.; Ueda, S.; Sato, N.; Kitaoka, Y.; Gon, Y.; Takabayashi,
A.; Spyrou, G.; Holmgren, A.; Yodoi, J. Free Radic. Biol. Med. 1999, 27, 504–514.
29. Yokomizo, A.; Ono, M.; Nanri, H.; Makino, Y.; Ohga, T.; Wada, M. Cancer Res.
1995, 55, 4293–4296.
30. Cheung, P. Y.; Churchich, J. E.; Lee, K. S. Biochem. Biophys. Res. Commun. 1999,
255, 17–22.
31. Pedersen, U.; Rasmussen, P. B.; Lawesson, S. O. Liebigs Ann. Chem. 1985, 1557–
1569.
32. Carmichael, J.; DeGraff, W. G.; Gazdar, A. F.; Minna, J. D.; Mitchell, J. B. Cancer
Res. 1987, 47, 936–942.
33. Mossman, T. J. Immunol. Methods 1983, 55–63.