Synthesis, cytotoxic and combined cDDP activity of new stable curcumin derivatives

Article abstract of DOI:10.1016/j.bmc.2009.03.016

New curcumin derivatives are synthesized in order to improve chemical properties of curcumin. The aromatic ring glycosylation of curcumin provides more water-soluble compounds with a greater kinetic stability which is a fundamental feature for drug bioavailability. The glycosylation reaction is quite simple, low cost, with high yield and minimum waste. NMR data show that the ability of curcumin to coordinate metal ion, in particular Ga(III), is maintained in the synthesized products. Although the binding of glucose to curcumin reduces the cytotoxicity of the derivatives towards cisplatin (cDDP)-sensitive and -resistant human ovarian carcinoma cell lines, the compounds display a good selectivity since they are much less toxic against non-tumourigenic Vero cells. The combination of cDDP with the most active glycosyl-curcuminoid drug against both cDDP-sensitive and -resistant as well as against Vero cell lines is tested. The results show an improvement of cDDP efficacy with higher selectivity towards cancer cells than non-cancer cells. These studies indicate the need for developing new valid components of drug treatment protocols to cDDP-resistant cells as well.

Full text of DOI:10.1016/j.bmc.2009.03.016

Bioorganic & Medicinal Chemistry 17 (2009) 3043–3052
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis, cytotoxic and combined cDDP activity of new stable curcumin
derivatives
Erika Ferrari a, Sandra Lazzari a, Gaetano Marverti b, Francesca Pignedoli a, Ferdinando Spagnolo a,
Monica Saladini a,
*
a Department of Chemistry, University of Modena and Reggio Emilia, Via Campi 183, 41100 Modena, Italy
b Department of Biomedical Science, University of Modena and Reggio Emilia, Via Campi 287, 41100 Modena, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
New curcumin derivatives are synthesized in order to improve chemical properties of curcumin. The aro-
matic ring glycosylation of curcumin provides more water-soluble compounds with a greater kinetic sta-
bility which is a fundamental feature for drug bioavailability. The glycosylation reaction is quite simple,
low cost, with high yield and minimum waste. NMR data show that the ability of curcumin to coordinate
metal ion, in particular Ga(III), is maintained in the synthesized products. Although the binding of glucose
to curcumin reduces the cytotoxicity of the derivatives towards cisplatin (cDDP)-sensitive and -resistant
human ovarian carcinoma cell lines, the compounds display a good selectivity since they are much less
toxic against non-tumourigenic Vero cells. The combination of cDDP with the most active glycosyl-curc-
uminoid drug against both cDDP-sensitive and -resistant as well as against Vero cell lines is tested. The
results show an improvement of cDDP efficacy with higher selectivity towards cancer cells than non-can-
cer cells. These studies indicate the need for developing new valid components of drug treatment proto-
cols to cDDP-resistant cells as well.
Received 12 November 2008
Revised 24 February 2009
Accepted 7 March 2009
Available online 14 March 2009
Keywords:
Curcuminoidic compounds
Cytotoxicity
Kinetic stability
Metal complexes
Cisplatin
Ovarian carcinoma
Vero cells
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
to possess higher ROS scavenging ability than curcumin.7 Previ-
ous investigation assessed the ability of curcumin to form stable
Cisplatin (cDDP) is a compound in clinical use for the treatment
of malignancies of the urogenital tract and other cancers.1 The
overall clinical success of chemotherapy with cDDP is complicated
by toxicity and diminished by intrinsic and acquired tumour resis-
tance to this drug. cDDP-resistance is multifactorial and involves
decreased drug accumulation, increased glutathione and metallo-
thioneines content and enhanced DNA repair activity.2 The devel-
opment of new chemotherapeutic agents and new combination
regimen is thus highly desirable.
complexes with a fundamental metal ion such as Fe3+ 8
, and iron
chelation was related to biological activity of curcumin.9 Since tu-
mour cells growth requires higher iron level, iron chelators are
used as antitumour therapies.10 Due to its excellent pharmacody-
namic profile, curcumin proceeded onto clinical trials however its
use is limited by a poor bioavailability.11 The main drawbacks in
the clinical use of curcumin are: low solubility, high rate of
metabolism, inactivity of metabolic products and/or rapid elimi-
nation from the body. In order to improve curcumin water solu-
bility and drug-delivery, different approaches have been
reported such as curcumin conjugation with nucleosides12 and
biopolymers.13,14
This paper reports a simple synthetic pathway able to give new
glycosyl curcuminoids (Scheme 1) with high yield and low reaction
time; the compounds are completely characterized. The chemical
stability of synthesized compounds is also evaluated and compared
with curcumin as kinetic stability is an important factor that influ-
ences serum level of active molecules. 1H NMR and UV–vis spec-
troscopy are here used to evaluate metal chelating ability of
studied compounds.
Curcumin
[1,7-bis(3-hydroxy-4-methoxyphenyl)hepta-1,6-
diene-3,5-dione] is a well-known dietary pigment derived from
Curcuma longa L. It has been shown to inhibit growth of several
types of malignant cells both in vivo and in vitro, various mecha-
nisms of action have been proposed which may be correct.3 Recent
studies demonstrated that curcumin is also able to reduce the pro-
liferation of human ovarian carcinoma cells with potency compa-
rable to cisplatin4 by inducing G2/M phase cell-cycle arrest.5
Moreover curcumin demonstrated a great ability in chelating
essential metal ions such as Cu(II)6 and these complexes showed
These curcumin derivatives are tested against a cDDP-sensitive
human ovarian carcinoma cell line, 2008 cells, and its -resistant
counterpart, C13* cells.
* Corresponding author. Tel.: +39 0592055040; fax: +39 059373543.
E-mail addresses: gaetano.marverti@unimore.it (G. Marverti), monica.saladini@
unimore.it (M. Saladini).
0968-0896/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2009.03.016

3044
E. Ferrari et al. / Bioorg. Med. Chem. 17 (2009) 3043–3052
chemical agents. In addition, this line is non-tumourigenic but
immortalized, allowing to culture cells for longer than normal cell
line.
H
O
O
R2
R1
R2
R1
6
9
4
2
2. Results and discussion
2.1. Chemistry
7
8
5
3
1
10
Existing reports detail the synthesis of compound 12,20 however
the low yield and the formation of a mixture of glycosilated prod-
ucts suggested to try a new synthetic pathway extensible to all gly-
cosilated derivatives.
The SN2 reaction of 3-substituted 4-hydroxyl-benzaldehyde on
2,3,4,6-tetracetyl-a-D-glucopyranosyl bromide (1) is performed
Compound
R1
OH
OAc
R2
5
6
OCH3
OCH3
by activation of phenolic moiety using an aqueous NaOH solution;
the glycosylated products (2–4) are well recrystallised by EtOH and
only b-anomer is detected. The reactivity of 3,4-dihydroxyl-benzal-
dehyde and vanillin with respect to 4-hydroxyl-benzaldehyde is
due to the electron withdrawing effect of meta-substituent. There-
fore the increasing phenolic acidity activates the phenolate ion in
the SN reaction, overcoming the steric hindrance of the vicinal
substituent.
The general synthesis for curcumin analogs is shown in Scheme
2. The new synthetic pathway represents an improvement of Pa-
bon reaction21 as it concerns yield, reaction time, costs and waste.
Boric anhydride is added to form a complex with 2,4-pentane-
dione (acac) in order to protect C-1 from Knoevenagel condensa-
tion and lead to aldol condensation. The proper solvent is DMF,
which is able to provide high solubility for both reactants and
intermediates together with a medium of suitable polarity for
the process. DMF also facilitates the isolation and separation of
the curcuminoids (5–9) from by-products.
O-
16
O
AcO
15
11
12
14
7
H
AcO
13
OAc
OAc
(AcGlcO-)
AcGlcO-
8
9
OH
AcGlcO-
OCH3
O-
16
O
HO
15
14
11
12
10
H
HO
13
OH
OH
(GlcO-)
GlcO-
GlcO-
11
12
OH
Due to depletion of diketone by side reactions, the 1,8/1 benzal-
dehyde/diketone stoichiometric ratio is preferred and n-butyl-
amine was chosen as condensation catalyst.
OCH3
Scheme 1.
During the reaction, water is produced upon formation of the
acac complex as well as upon formation of curcuminoid itself.
Since water can generate side reactions and substantially reduce
the yield, the reaction is carried out under anhydrous conditions,
and tributylborate (nBuO)3B is added as drying agent.
The pure products separate from reaction mixture and NMR
data support a purity greater than 98%, therefore no chromato-
graphic purification is needed. This one pot reaction is character-
ized by quite good yields (ꢀ80%), short time (6 h) and small
volumes of implied solvents (1 ml of DMF for 0.5 mmol of product).
The final deacetylation step to obtain compound 10–12, per-
formed by use of CH3ONa, is a quantitative reaction and no prob-
lems of decomposition are observed for all the products.
A fundamental requisite for the clinical efficacy of anticancer
drugs is selectivity of action, that means a good therapeutic index,
therefore it would be helpful to test cytotoxic activity of these cur-
cumin derivatives towards normal cell line. In the present study,
the antitumour activity of the compounds has been tested in par-
allel with the investigation of the possible changes in growth and
cell proliferation patterns induced by these curcumin analogs on
Vero cells, an immortal, non-tumourigenic fibroblastic cell
line.15,16 It is well known that curcumin has a antiproliferative ef-
fect both on ovarian cancer cells and fibroblast cells, although it in-
duces different phase cell-cycle arrest.5,17 Even if the Vero cell
lineage is obtained from a different species (African green monkey
kidney), it shares a common embryonic origin (mesoderm) with
cells from human genital tract.18 Vero cells are usually considered
as control cells19 as they present very well defined properties
which allow to easily observe possible growth modification by
2.2. NMR spectroscopy
Fe-sequestering ability is often connected to drugs anticancer
activity22 therefore we tested metal affinity of curcumin deriva-
1. B2O3, DMF, 80°C, 30’
2. (nBuO)3B, 80°C, 30’
CHO
O
OH
3.
, nBuNH2, 4h
R1
O
O
R2
H3C
CH3
R1
R1
4. HCl, r.t., 30’
R2
R2
Scheme 2.

E. Ferrari et al. / Bioorg. Med. Chem. 17 (2009) 3043–3052
3045
tives by means of NMR technique which provides useful informa-
tion on coordination sites and complexes stoichiometry.
of the complex originated from metal chelation by keto-enolic
moiety. In line with the behaviour of curcumin,8 the phenolic
group of 11 does not interact with the metal ion. Small changes
for the aromatic protons are due to delocalization of the metal
charge along the aliphatic chain and aromatic rings. As more Ga(III)
is added a new downfield species appears hinting the formation of
another complex species characterized by 1/1 Me/L molar ratio,
independently on the ligand. No evidence of an interaction be-
tween sugar moiety and metal ion is observed by NMR spectros-
copy in acid media, although we cannot exclude a possible
glycoside interaction with the metal ion under physiological condi-
tions, as reported in literature in the solid state.27–29
All the molecules show a typical spectral pattern of a keto-eno-
lic moiety,8,23 the enolic proton is mobile in fast exchange with
residual HOD and implied in a strong intra-molecular hydrogen
bond that makes it equally shared by the two oxygen atoms as con-
firmed by crystal structure of analogous compounds.24 The ability
to chelate metal ions is tested by adding Ga(NO3)3 to a solution of
the ligand; diamagnetic Ga(III) is used as NMR probe instead of
paramagnetic Fe(III).25,26
The addition of Ga(NO3)3 in 1/2 metal to ligand Me/L molar ratio
to a solution of compounds 10, 11, 12 (Scheme 1), at acidic pH,
immediately originates new signals, in slow chemical exchange
in NMR time scale, whose spectral pattern resembles the one of
the free ligand but strongly downfield shifted (Fig. 2, Tables 1
and 2). Since little reductions are observed in the pH value (ꢀ0.5
pH units), the formation of the new sets of signals is not a conse-
quence of the decreased pH induced by Ga(III) hydrolysis but is
more reasonably attributed to the formation of a complex species
with a 1/2 Me/L molar ratio. The greatest downfield shift is for
the H-4 signal; H-4 probably ‘feels’ the delocalized positive charge
2.3. UV–vis spectroscopy
Similarly to curcumin, glucosyl curcuminoids maintain their
dye properties showing a absorbance maximum in the 300–
400 nm spectral range. By adding Ga(III) solution to the free li-
gands in acidic condition (pH ꢀ 4.5) a general blue shift is observed
due to the interaction between the metal ion and the keto-enolic
moiety, Figure 1 reports gallium titration of 12 at k = 480 nm the
0.24
0.22
0.20
0.18
0.16
0.14
0.12
0.10
0.65
0.50
0.35
A
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
M/L
0.20
0.001
250
280
320
360
400
440
480
520
560
600
nm
Figure 1. Spectrophotometric titration of 12 (2.5 ꢁ 10ꢂ5 M) with Ga(NO3)3 in aqueous solution. The inset shows the plot of absorbance versus metal to ligand molar ratio
(Me/L) at k = 480 nm.
Aromatic
ring
Sugar
moiety
H-4
H-3
B
A
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
ppm
Figure 2. 1H NMR spectra of compound 11 (A) and 11/Ga(NO3)3 2:1 molar ratio system (B) in CD3OD.

3046
E. Ferrari et al. / Bioorg. Med. Chem. 17 (2009) 3043–3052
Table 1
1H chemical shifts (d ppm) of ligands, their gallium(III) complexes and
D
d (dcomplex ꢂ dligand) registered at 300 K in CD3OD
H-1
H-3
H-4
H-6
H-7
H-9
H-10
Compound 10
Me/L 1:2
Me/L 1:1
6.01
6.06 (0.05)
6.13 (0.12)
6.69
6.83 (0.14)
6.87 (0.18)
7.62
7.80 (0.18)
7.88 (0.26)
7.6
7.62 (0.02)
7.67 (0.07)
7.14
7.16 (0.02)
7.19 (0.05)
7.14
7.16 (0.02)
7.19 (0.05)
7.6
7.62 (0.02)
7.67 (0.07)
Compound 11
Me/L 1:2
Me/L 1:1
6.03
6.04 (0.01)
6.12 (0.09)
6.65
6.76 (0.11)
6.81 (0.16)
7.57
7.72 (0.15)
7.81 (0.24)
7.17
7.17 (0.00)
7.22 (0.05)
—
—
7.24
7.10
7.22 (ꢂ0.02)
7.09 (ꢂ0.01)
7.27 (0.03)
7.16 (0.06)
Compound 12
Me/L 1:2
Me/L 1:1
6.02
6.04 (0.02)
6.13 (0.11)
6.72
6.84 (0.12)
6.90 (0.18)
7.60
7.76 (0.16)
7.86 (0.26)
7.12
7.15 (0.03)
7.18 (0.06)
7.18
7.18
7.17 (ꢂ0.01)
7.16 (ꢂ0.02)
7.23 (0.05)
7.25 (0.07)
Table 2
13C chemical shifts (d ppm) of ligands, their gallium(III) complexes and
D
d (dcomplex ꢂ dligand) registered at 300 K in CD3OD
C-1
C-2
C-3
C-4
C-5
C-6
C-7
C-8
C-9
C-10
Compound 10
Me/L 1:2
Me/L 1:1
102.8
102.7 (ꢂ0.1)
185.6
186.8 (1.2)
186.8 (1.2)
124.5
126.7 (2.2)
125.5 (1.0)
142.1
143.6 (1.5)
144.8 (2.7)
131.6
131.6
131.9 (0.3)
131.8 (0.2)
118.9
119.0 (0.1)
118.6 (ꢂ0.3)
161.6
161.8 (0.2)
162.1 (0.5)
118.9
119.0 (0.1)
118.6 (ꢂ0.3)
131.6
131.9 (0.3)
131.8 (0.2)
131.5 (ꢂ0.1)
132.0 (0.4)
Compound 11
Me/L 1:2
Me/L 1:1
102.8
104.3 (1.5)
185.4
186.5 (1.1)
186.8 (1.4)
124.5
126.7 (2.2)
126.3 (1.8)
142.1
143.4 (1.3)
144.5 (2.4)
132.6
132.6 (0.0)
132.5 (ꢂ0.1)
116.2
116.9 (0.7)
117.2 (1.0)
147.7
148.3 (0.6)
149.6 (1.9)
149.4
149.4 (0.0)
150.2 (0.8)
118.8
118.8 (0.0)
119.0 (0.2)
122.8
123.4 (0.6)
123.7 (0.9)
Compound 12
Me/L 1:2
Me/L 1:1
102.8
103.6 (0.8)
185.6
186.8 (1.2)
186.9 (1.3)
124.5
126.8 (2.3)
126.3 (1.8)
142.1
143.5 (1.4)
145.0 (2.9)
132.1
132.2 (0.1)
132.0 (ꢂ0.1)
112.1
112.6 (0.5)
112.8 (0.7)
149.6
150.1 (0.5)
150.8 (1.2)
150.1
150.2 (0.1)
150.8 (0.7)
117.3
117.3 (0.0)
117.2 (ꢂ0.1)
122.5
122.8 (0.3)
122.8 (0.3)
absorbance increases with Me/L molar ratio reaching a maximum
at Me/L value of 0.5 corresponding to a ML2 complex, confirming
NMR data (insert Fig. 1).
stability of cyclodextrin complexed curcumin (t1/2 >100 h).31 If
compared to curcumin the presence of acetyl groups in compound
6 also increases kinetic stability (4% decomposed after 1 h) sug-
gesting that the fast degradation of curcumin is strongly influenced
by phenolic groups. The kinetic stability of curcumin derivatives is
also confirmed by NMR data. In fact no relevant changes in 1H
chemical shifts and integrated areas are observed during the first
72 h. The nature of meta-substituent does not drive the degrada-
tion process in glycosylated products, although compound 10 is
the more stable one during the first 2 h.
2.4. Kinetic study
Chemical stability is a fundamental drug feature therefore we
studied the degradation process of our compounds following the
time dependent diminishing in UV–vis k max absorbance. By plot-
ting ln(At/A0) versus time at 37 °C and constant ionic strength a lin-
ear regression for all the compounds is observed hinting a first
order kinetic process. Figure 3 shows that degradation is extremely
slow at pH 7 for all glycosylated compounds (10, 11 and 12) being
the percentage of decomposed compounds less than 30% during
the first hour, while curcumin is 50% decomposed in the same
experimental conditions; previous study reported that curcumin
is almost completely decomposed after 1 h.30 The presence of su-
gar moiety delays degradation processes being the t1/2, estimated
from curves in Figure 2, in the range 150–200 h, resembling the
2.5. Cytotoxic assay
In order to compare the growth inhibitory effects of curcumin
and its derived drugs, we tested the cytotoxicity of curcumin
against a cDDP-sensitive human ovarian carcinoma cell line and
its acquired cDDP-resistant counterpart. cDDP-resistant cell line
shows a 3.4-fold cross-resistance to curcumin, since its IC50 value
(IC50 = drug concentration that reduces cell growth by 50%) is
17
1 l lM in 2008 cells. At 10 lM
M in C13* cells and 5.0 0.4
curcumin inhibits 2008 cell growth at 90% while cDDP-resistant
cells still survive to this concentration by a nearly 60% and display
a more dose-related sensitivity to the compound than the parental
line (data not shown). According to the reports of Adams et al.,32
100
80
60
40
20
0
Table 3
IC50 values deduced from the dose–response curves in 2008 and C13* cells
Compounds
Cell lines (IC50
,
l
M)
C13* cells
17
40 3.7
RFI
Vero cell
RFII
2008 cells
Curcumin
6
10
11
12
cDDP
5.0 0.4
9.8 0.8
22 1.9
480 37
220 16
1.5 0.1
1
3.4
4.1
3.8
>2
3.8
12.3
22
45
330
ND*
ND*
10
1.3
1.1
3.97
83
ND
840 68
18.4 1.6
7
0
2
4
6
8
Time (h)
Figure 3. Plotting of residual compound percentage versus time (h) for com-
pounds: 5 (d), 6 ( ), 10 (N), 11 (ꢀ), 12 (j). Spectrophotometric measurements were
performed in buffer solution (pH 7) at 37 °C.
*
Resistant factors: RFI = IC50 resistant/IC50 parent line, RFII = IC50 Vero cell/ IC50
resistant. (*ND = not determined).

E. Ferrari et al. / Bioorg. Med. Chem. 17 (2009) 3043–3052
3047
curcumin cytotoxicity against the cell model of our study is also
confirmed to be comparable to that of cDDP, as reported in Table
3. Five micromolar curcumin concentration (IC50 value in 2008
cells) blocks Vero cell growth by only 10%. Ten micromolar curcu-
min concentration inhibits 2008 and C13* cell growth by about 90%
and 45%, respectively and decreases Vero cell proliferation by
about 25% (data not shown).
Figure 4 shows the cytotoxicity of compounds 6, 10, 11, 12, in
2008 and C13* cells, and Table 3 reports the corresponding IC50
values.
fold lower effectiveness of 6 with comparison to curcumin is
also maintained against Vero cells, whose growth is 50% inhib-
ited by 45
shown).
5
l
M 6 (Fig. 5) and 22
2 lM curcumin (data not
With the purpose to improve curcumin water-solubility the
phenyl groups have been functionalized with glucose in position
4, obtaining compound 12. As shown in Figure 4 (panel D), com-
pound 12 is a less potent inhibitor than curcumin against either
2008 and C13* cell growth. After 72 h-exposure the IC50 values ob-
tained were 220 16 lM and 840 68 lM, respectively; thus dis-
The acetylation of the phenolic groups of curcumin provides
compound 6, whose cytotoxicity (Fig. 4, panel A) is almost re-
verted to that of curcumin, being compound 6 only twofold less
potent than the lead compound. As a result, the IC50 values of 6
playing a 3.8-fold cross-resistance to the analogue in C13* cells.
Cell growth inhibition of 12 is accompanied with a very scant cyto-
toxicity against the non-neoplastic monkey fibroblasts (Fig. 5),
whose growth is inhibited by a maximum of 20% even up to
are 9.8 0.8
respectively. The dose–response curve of cell survival to
shows a plateau from 20 M to higher concentrations only for
l
M
and 40
4
lM
in 2008 and in C13* cells,
750 lM.
6
To increase the antitumour activity of glycosyl derivatives in
comparison to curcumin, we have replaced methoxyl groups in po-
sition 3 with hydroxylic group, producing compound 11. Figure 4
(panel C) shows that cDDP-sensitive cells are also more responsive
l
2008 cells. This response trend is very similar to that shown
by the same cells treated with curcumin. In addition, the two-
100
100
A
B
2008 cells
C13* cells
2008 cells
C13* cells
75
75
50
25
0
50
25
0
0
25
50
75
100
0
20
40
60
Compound [10] µM
Compound [6] µM
100
75
50
25
0
100
75
50
25
0
2008 cells
C13* cells
C
D
2008 cells
C13* cells
0
250
500
750
0
200
400
600
800
1000
Compound [11] µM
Compound [12] µM
Figure 4. Cytotoxicity of b-diketones in 2008 (s) and C13* (d) cells. 24 h after seeding, cells were exposed for 3 days to the indicated concentrations of the drugs and then
cytotoxicity was assessed by the crystal violet method. Results represent the mean of three separate experiments performed in duplicate. The standard error was within 10%
for each data point.

3048
E. Ferrari et al. / Bioorg. Med. Chem. 17 (2009) 3043–3052
cently it has been shown that curcumin treatment improved the
100
75
50
25
0
efficacy of cDDP-based chemotherapy against human laryngeal
carcinoma cells.33
In this regard we have tested the combination of cDDP with
the most active drug among glycosyl-curcuminoids, 10, against
both sensitive and resistant cell lines. The nature of the drug
interaction between cDDP and 10 is investigated by median-ef-
fect analysis, according to Chou and Talalay,34 which calculates
the combination indexes (CI), a measure of drug interaction ef-
fect. A CI <1 indicates synergism, CI = 1 additivity and CI >1
antagonism. Compound 10 and cDDP were combined at a ratio
of 20:1 for 2008 cells and 5:1 for C13* cells. Scheduling exper-
iments were carried out by adding both agents either simulta-
neously or sequentially to determine the optimum conditions.
As shown in Figure 6 the concurrent drug treatment schedule
appears to be the most effective combination against 2008 cell
line, being the CI values always lower than 1 (panel A). Sequen-
tial combinations do not improve the interactive effects of the
drugs since a 24 h-pretreatment with 10 causes, at most, su-
pra-additive effects with 2 combinations and slight synergistic
11
12
10
6
0
100 200 300 400 500 600 700
[drugs] µM
inhibition with 5 lM cDDP (panel E). In addition, when cDDP
was added 24 h before 10, antagonistic interactions is observed
with all combinations tested (panel C).
All treatment schedules with the two drugs bring about similar
interactive positive effects against the resistant line; in fact, almost
Figure 5. Dose–response effects of the b-diketones on growth of Vero cells. 24 h
after seeding, cells were exposed for 3 days to the indicated concentrations of the
drugs and then cytotoxicity was assessed by the crystal violet method. Results
represent the mean of three separate experiments performed in duplicate. The
standard error was within 10% for each data point.
additive effects at 5 and 10
20 M are displayed. In particular, the lowest CI (0.35), indicative
of a strong synergy, is observed when 10 is combined concurrently
with 20 M cDDP (panel B). These results indicate that 10 may im-
prove the efficacyof cDDP in the sensitive line more than in the resis-
tant line. On the contrary, 10 post-treatment cause a detrimental
effect on cDDP action. Despite their resistance to each drug alone,
C13* cell growth could be successfully inhibited by the appropriate
combinations of the two drugs.
In all the schedules tested, the interactive effects of the two drugs
resulted highly antagonistic against the non-tumourigenic lineage,
Vero cells (Fig. 7). As it appears, antagonism of the combinations in-
creases with cDDP concentrations, reaching very high CI values at
the doses of cDDP used in C13* cells. These data indicate the selectiv-
ity of the interactive effects of cDDP with 10 that, when properly
combined, might be more potent inhibitors against carcinoma cells
than against non-tumourigenic control cells.
lM cDDP and synergistic cell killing at
to 11 than their resistant counterpart. While the IC50 value of 11 is
l
about 500
the IC50 value in the concentration range investigated. At 750
2008 and C13* cells are inhibited by 70% and only 15%, respec-
tively. However, the cytotoxicity of 11 against C13* cells parallels
a poor inhibition of Vero cell growth, which survive by nearly 80
l
M in 2008 cells, in C13* cells was not possible to obtain
lM,
l
% up to 750 lM 11 (Fig. 5).
The elimination of meta-substituent on the phenyl ring gives
compound 10. Figure 4 (panel B) provides evidence that this re-
moval is actually effective since the IC50 values are 10-fold de-
creased both in 2008 and in C13* cells (Table 3), when compared
to 12 compound. Again, C13* cells are also 3.8-fold cross-resistant
to 10, as deduced from IC50 values and resistant factor RFI
(RFI = IC50 resistant/IC50 parent line) reported in Table 3. Com-
pound 10 displays a good selectivity since it is 15- and 4-fold less
toxic against Vero cells (IC50 value 330 31 lM) than against 2008
and C13* cells, respectively (Fig. 5). The resistant factor RFII value
(RFII = IC50 Vero cell/ IC50 resistant) (Table 1) shows a better selec-
tivity of 10 towards C13* cells than against Vero cells when com-
pared to curcumin and compound 6, indicating that even cDDP-
resistant cells are somehow more responsive to the derivative than
control non-tumourigenic line.
3. Conclusion
The curcumin derivatives are easily synthesized with high yield
and low costs; they demonstrate a greater kinetic stability if com-
pared to curcumin, maintaining the metal-ligating ability through
b-diketo moiety. Among curcumin derivatives compound 6 and 10
demonstrated the better cytotoxicity towards 2008 and C13* cell
lines. In particular compound 10 displays a good selectivity since
it is much less toxic against non-tumourigenic Vero cells. Compound
10 improves also the efficacy of cDDP in sensitive cell line and also in
resistant cell line while the combination of the two drugs is highly
antagonistic against non-tumourigenic cell line indicating an im-
proved selectivity.
These data suggest that binding a glucose molecule to curcu-
min greatly reduces the cytotoxicity of the derivatives, as indi-
cated by the increased values of IC50 both in sensitive and
resistant cell lines. This diminished effectiveness could be as-
cribed to the greater molecular complexity and bulkiness com-
pared to curcumin, which may limit cellular penetration of the
compounds. According to this hypothesis the replacement of
glucose with acetoxy group restores the cytotoxicity almost to
the level of curcumin, being the related IC50 value about only
doubled with respect to the lead molecule. The growth-inhibi-
tory activity increases by decreasing the electron withdrawing
effect of meta-substituent, as indicated by the sequence of po-
tency 11 < 12 < 10. The absence of meta-substituents on the aro-
matic ring may facilitate the electron conjugation and
delocalization on the sp2 moiety, which may be responsible of
the enhanced cytotoxicity observed in both lines.
4. Experimental
4.1. Synthesis
4.1.1. 2,3,4,6-Tetracetyl-
Compound 1 was synthesized quantitatively by reaction of
1,2,3,4,6-pentaacetyl-b- -glucopyranose with HBr 30% in glacial
acetic acid, at room temperature for 4 h.35
a-D-glucopyranosyl bromide (1)
Cisplatin (cDDP) is frequently used in combination with other
drugs in order to minimize the side-effects of each compound. Re-
D

E. Ferrari et al. / Bioorg. Med. Chem. 17 (2009) 3043–3052
3049
C13* cells
2008 cells
3.0
2.5
2.0
1.5
1.0
0.5
0.0
3.0
2.5
2.0
1.5
1.0
0.5
0.0
cDDP+10
cDDP+10
A
B
1
2
3
4
5
5
10
15
20
[cDDP] µM
[cDDP] µM
3.0
3.0
2.5
2.0
1.5
1.0
0.5
0.0
cDDP-24h+ 10
cDDP-24h +10
C
D
2.5
2.0
1.5
1.0
0.5
0.0
1
2
3
4
5
5
10
15
20
[cDDP] µM
[cDDP] µM
3.0
2.5
2.0
1.5
1.0
0.5
0.0
3.0
2.5
2.0
1.5
1.0
0.5
0.0
10-24h+cDDP
10-24h+cDDP
F
E
1
2
3
4
5
5
10
15
20
[cDDP] µM
[cDDP] µM
Figure 6. Median-effect analysis of cisplatin combined with compound 10 at the constant ratio of 1:20 for the 2008 cells (left panels) and of 1:5 for the C13* cells (right
panels). Cells were allowed to attach overnight and then simultaneously (upper panels) or sequentially (middle and lower panels) treated with 10 and with the indicated
concentrations of cDDP in 24-well plates. Following a 3-day exposure, cells were stained with crystal violet. Results are expressed as combination indexes (CI), where CI <1
indicate synergistic effects, CI >1 indicate antagonistic effects, CI = 1 indicate additivity. Results represent the mean of three separate experiments performed in duplicate. The
standard error was within 10% for each data point.
4.1.2. 4-(2,3,4,6-Tetracetyl-b-
(2); 3-hydroxy-4-(2,3,4,6-tetracetyl-b-
aldehyde (3); 3-methoxy-4-(2,3,4,6-tetracetyl-b-
yloxy)benzaldehyde (4)
The appropriate benzaldehyde were solubilized in NaOH,
and then slowly dropped in an equimolar solution of 1 in ace-
tone. The reaction was kept under stirring overnight at room
D
-glucopyranos-1-yloxy)benzaldehyde
-glucopyranos-1-yloxy)benz-
-glucopyranos-1-
5.52, yield 80%; Anal. Calcd for C21H23O12 (3): C, 53.91; H,
4.96. Found: C, 54.22; H, 4.28, yield 85%; Anal. Calcd for
C22H25O12 (4): C, 54.84; H, 5.23. Found: C, 55.22; H, 5.82, yield
80%; NMR chemical shifts (ppm in CDCl3): (2) H-11 5.31, H-12
5.21, H-13 5.30, H-14 5.17, H-15 3.92, H-16 4.28, H-160 4.18,
H-6 7.85, H-7 7.10, H-9 7.10, H-10 7.85; (3) H-11 5.08, H-12
5.29, H-13 5.36, H-14 5.17, H-15 3.92, H-16 4.32, H-160 4.20,
H-6 7.46, H-9 7.08, H-10 7.40; (4) H-11 5.31, H-12 5.11, H-
13 5.31, H-14 5.17, H-15 3.84, H-16 4.27, H-160 4.19, H-6
7.43, H-9 7.21, H-10 7.41.
D
D
temperature, and
a precipitate splits up when water was
added. The isolated product was recrystallized from EtOH. Anal.
Calcd for C21H23O11 (2): C, 55.83; H, 5.14. Found: C, 56.02; H,

3050
E. Ferrari et al. / Bioorg. Med. Chem. 17 (2009) 3043–3052
added. 0.4 ꢁ 10ꢂ3 mol of n-butylamine in 0.5 ml of DMF were
Vero cells
slowly dropped (1 h). The solution kept under stirring at 80 °C
for 4 h developed a yellow-orange colour. The solution was acidi-
fied with 8 ml of HCl 0.5 M. During the acidification, the solution
was cooled to room temperature. The crude product initially ap-
peared as heavy oil; but as stirring continued, it slowly trans-
formed into an orange solid. After about 1 h of stirring, the solid
was filtered and resuspended in water at room temperature. The
crude solid was collected and dried in vacuum. Anal. Calcd for
C21H20O6 (5): C, 68.47; H, 5.47. Found: C, 68.41; H, 5.52, yield
60%; Anal. Calcd for C25H24O8 (6): C, 66.36; H, 5.35. Found: C,
65.91; H, 5.73, yield 87%; Anal. Calcd for C47H60O22 (7): C, 57.78;
H, 6.19. Found: C, 58.57; H, 5.79, yield 81%; Anal. Calcd for
C47H60O24 (8): C, 55.95; H, 5.99. Found: C, 56.35; H, 6.07, yield
91%; Anal. Calcd for C49H64O24 (9): C, 56.75; H, 6.22. Found: C,
57.09; H, 5.31, yield 77%. NMR chemical shifts (ppm in CD3OD):
(5) H-1 5.95, H-3 6.61, H-4 7.56, H-6 7.20, H-9 6.82, H-10 7.09;
C-1 100.2, C-2 183.7, C-3 120.6, C-4 140.4, C-5 127.3, C-6 110.1,
C-7 147.8, C-8 149.0, C-9 115.0, C-10 122.6; (6) H-1 5.94, H-3
6.59, H-4 7.57, H-6 7.49, H-7 6.82, H-9 6.82, H-10 7.49; C-1
100.3, C-2 183.6, C-3 120.4, C-4 140.2, C-5 126.5, C-6 129.3, C-7
115.3, C-8 160.1, C-9 115.3, C-10 129.3. NMR chemical shifts
(ppm in CDCl3): (7) H-1 5.80, H-3 6.52, H-4 7.61, H-6 7.00, H-7
7.50, H-9 7.50, H-10 7.00, H-11 5.14, H-12 5.29, H-13 5.28, H-14
5.18, H-15 3.89, H-16 4.29, H-160 4.18; C-1 99.5, C-2 180.7, C-3
120.7, C-4 125.9, C-5 133.2, C-6 135.2, C-7 114.7, C-8 162.3, C-9
114.7, C-10 135.2, C-11 96.3, C-12 68.2, C-13 70.7, C-14 66.0, C-
15 69.1, C-16 58.9; (8) H-1 5.80, H-3 6.51, H-4 7.55, H-6 7.18, H-
9 6.96, H-10 7.02, H-11 5.00, H-12 5.28, H-13 5.32, H-14 5.17, H-
15 3.88, H-16 4.31, H-160 4.20; C-1 99.6, C-2 180.7, C-3 121.3, C-
4 127.8, C-5 134.0, C-6 127.1, C-7 114.7, C-8 155.6, C-9 114.7, C-
10 127.1, C-11 96.1, C-12 68.6, C-13 70.1, C-14 65.7, C-15 69.7,
C-16 59.4; (9) H-1 5.82, H-3 6.51, H-4 7.58, H-6 7.10, H-9 7.04,
H-10 7.05, H-11 5.02, H-12 5.27, H-13 5.28, H-14 5.16, H-15
3.80, H-16 4.26, H-160 4.17; C-1 99.1, C-2 180.6, C-3 120.9, C-4
137.5, C-5 129.1, C-6 109.2, C-7 148.3, C-8 145.2, C-9 117.1, C-10
119.1, C-11 99.7, C-12 68.6, C-13 70.0, C-14 65.8, C-15 69.6, C-16
59.4.
cDDP+10
50
40
30
20
10
0
5
10
15
15
15
20
20
20
[cDDP] µM
30
25
20
15
10
5
cDDP-24h+10
0
5
10
[cDDP] µM
40
35
30
25
20
15
10
5
10-24h+cDDP
4.1.4. 1,7-Bis(4-b-d-glucopyranos-1-yloxophenyl)hepta-1,6-
diene-3,5-dione (10); 1,7-bis(3-hydroxy-4-b-
yloxophenyl)hepta-1,6-diene-3,5-dione (11); 1,7-bis(3-
methoxy-4-b- -glucopyranos-1-yloxophenyl)hepta-1,6-diene-
3,5-dione (12)
D-glucopyranos-1-
0
5
10
D
[cDDP] µM
The acetylated compounds, solubilized in CH3ONa 0.1 M meth-
anolic solution, were kept under stirring for 2 h; pH was brought to
neutrality by use of a DOWEX ion exchange resin, which was final-
ly filtered off, and the solutions were evaporated under vacuum.
Anal. Calcd for C31H36O14 (10): C, 58.86; H, 5.74. Found: C, 56.50;
H, 5.76. Anal. Calcd for C31H36O16 (11): C, 56.02; H, 5.46. Found:
C, 54.53; H, 6.79; Anal. Calcd for C33H40O16 (12): C, 57.22; H,
5.82. Found: C, 53.83; H, 7.02.
Figure 7. Median-effect analysis of cisplatin combined with compound 10 against
Vero cells. Cells were allowed to attach overnight and then simultaneously (upper
panel) or sequentially (middle and lower panel) treated with 10 and with the
indicated concentrations of cDDP in 24-well plates. Following a 3-day exposure,
cells were stained with crystal violet. Results are expressed as combination indexes
(CI), where CI <1 indicate synergistic effects, CI >1 indicate antagonistic effects,
CI = 1 indicate additivity. Results represent the mean of three separate experiments
performed in duplicate. The standard error was within 10% for each data point.
NMR chemical shifts (ppm in CD3OD): (10) H-1 6.01, H-3 6.69,
H-4 7.62, H-6 7.14, H-7 7.61, H-9 7.60, H-10 7.14, H-11 4.98, H-
12 3.52, H-13 3.47, H-14 3.42, H-15 3.45, H-16 3.91, H-160 3.71;
C-1 102.8, C-2 185.6, C-3 124.5, C-4 142.1, C-5 131.6, C-6 131.6C-
7 118.9, C-8 161.6, C-9 118.9, C-10 131.6, C-11 96.5, C-12 68.4,
C-13 71.2, C-14 66.7, C-15 68,7, C-16 58.3; (11) H-1 6.01, H-3
6.63, H-4 7.53, H-6 7.14, H-9 7.20, H-10 7.07, H-11 4.97, H-12
3.53, H-13 3.47, H-14 3.44, H-15 3.45, H-16 3.90, H-160 3.71; C-1
102.8, C-2 185.4, C-3 124.5, C-4 142.1, C-5 132.6, C-6 116.2, C-7
147.7, C-8 149.4, C-9 118.8, C-10 122.8, C-11 96.3, C-12 68.1, C-
13 71.4, C-14 65.0, C-15 70.1, C-16 59.2; (12) H-1 6.02, H-3 6.72,
H-4 7.60, H-6 7.12, H-9 7.18, H-10 7.18, H-11 4.58, H-12 3.53, H-
13 3.48, H-14 3.43, H-15 3.45, H-16 3.89, H-160 3.71; C-1 102.8,
C-2 185.6, C-3 124.5, C-4 142.1, C-5 132.1, C-6 112.1, C-7 149.6,
4.1.3. 1,7-Bis[3-methoxy-4-hydroxyphenyl]hepta-1,6-diene-
3,5-dione (5); 1,7-bis[3-methoxy-4-acetylphenyl]hepta-1,6-
diene-3,5-dione (6); 1,7-bis[4-(2,3,4,6-tetracetyl-b-d-
glucopyranos-1-yloxy)phenyl]hepta-1,6-diene-3,5-dione (7);
1,7-bis[3-hydroxy-4-(2,3,4,6-tetracetyl-b-d-glucopyranos-1-
yloxy)phenyl]hepta-1,6-diene-3,5-dione (8); 1,7-bis[3-methoxy-
4-(2,3,4,6-tetracetyl-b-D-glucopyranos-1-yloxy)phenyl]hepta-
1,6-diene-3,5-dione (9)
A suspension of 1.0 ꢁ 10ꢂ3 mol of B2O3 and 1.0 ꢁ 10ꢂ3 mol of
acetylacetone in DMF (1.5 ml) was stirred for 30 min at 80 °C, fol-
lowed by the addition of 4.0 ꢁ 10ꢂ3 mol of tributylborate. After
30 min 1.8 ꢁ 10ꢂ3 mol of the appropriate benzaldehyde was

E. Ferrari et al. / Bioorg. Med. Chem. 17 (2009) 3043–3052
3051
C-8 150.1, C-9 117.3, C-10 122.5, C-11 99.4, C-12 68.1, C-13 70.5, C-
14 66.2, C-15 69.7, C-16 59.2.
crystal violet solution in 80% absolute ethanol for at least 30 min.
After washing several times with distilled water to remove the
dye excess, the cells were let to dry. The incorporated dye was sol-
ubilized in acidic isopropanol (1 N HCl/2-propanol, 1:10). After
appropriate dilution, dye was determined spectrophotometrically
at 540 nm. The extracted dye was proportional to cell number. Per-
centage of cytotoxicity was calculated by comparing the absor-
bance of exposed to non-exposed (control) cultures.
4.2. Spectroscopy
Spectrophotometric measurements were performed using Jasco
V-570 spectrophotometer at 25 0.1 °C in the 200–600 nm spec-
tral range employing a 1 cm quartz cell. 2.5 ꢁ 10ꢂ5 M water solu-
tion of 10, 11 and 12 were investigated; constant additions
(10
l
l) of Ga(NO3)3 solution (10ꢂ2 M) using a micropipette in order
4.6. Median-effect analysis
to reach different Me/L molar ratios.
NMR Spectra were recorded on a Bruker Avance AMX-400 spec-
trometer with a Broad Band 5 mm probe (inverse detection). Nom-
inal frequencies are 100.13 MHz for 13C and 400.13 MHz for 1H.
The typical acquisition parameters for 1H are as follows: 20 ppm
Median-effect analysis was used to determine the nature of the
combination between cisplatin and compound 10 combined at a
fixed ratio.34 Computer analysis by Calcusyn software, Biosoft,
Cambridge, UK) of the dose–response curves was used to calculate
the combination index (CI) at increasing levels of cell kill. CI values
of less than or greater than 1 indicate synergy and antagonism,
respectively, whereas a CI value of 1 indicates additivity of the
drugs.
spectral bandwidth (SW), 6.1 ls pulse width (90° pulse hard pulse
on 1H), 0.5–1 s pulse delay, 216–512 number of scans. 0.5 ml of a
CD3OD 10ꢂ2 M solution of each ligand was prepared, then addition
(10
l
l) of Ga(NO3)3 (5 ꢁ 10ꢂ2 M in CD3OD) was performed using a
micropipette.
Acknowledgement
4.3. Kinetic studies
We are thankful to ‘Centro Interdipartimentale Grandi Strum-
enti (CIGS)’ of the University of Modena and Reggio Emilia, which
supplied NMR Spectrometer.
The chemical stability of the ligands was evaluated as a change
in absorbance in the 385–440 nm range. The aqueous solutions of
the ligands, concentration from 1 ꢁ 10ꢂ5 to 5 ꢁ 10ꢂ5 M ,were pre-
pared in 10ꢂ2 M phosphate buffered solution (pH 7). A constant io-
nic strength of 0.1 M (NaNO3) was maintained in all experiments.
The cell was stored at 37 °C. The spectra were recorded every
30 min for 8 h.
References and notes
1. Wong, E.; Giandomenico, C. M. Chem. Rev. 1999, 99, 2451.
2. Kelland, L. R. Drugs 2000, 59, 1.
3. Hatchera, H.; Planalpb, R.; Chob, J.; Tortia, F. M. Cell. Mol. Life Sci. 2008, 65, 1631.
4. Adams, B. K.; Ferstl, E. M.; Davis, M. C.; Herold, M.; Kurtkaya, S.; Camalier, R. F.;
Hollingshead, M. G.; Kaur, G.; Sausville, E. A.; Rickles, F. R.; Snyder, J. P.; Liotta,
D. C.; Shoji, M. Bioorg. Med. Chem. 2004, 12, 3871.
4.4. Cell lines
5. Weir, N. M.; Selvendiran, K.; Kutala, V. K.; Tong, L.; Vishwanath, S.; Rajaram, M.;
Tridandapani, S.; Anant, S.; Kuppusamy, P. Cancer Biol. Ther. 2007, 6, 178.
6. John, V. D.; Kuttan, G.; Krishnankutty, K. J. Exp. Clin. Cancer Res. 2002, 21, 219.
7. Shen, L.; Ji, H. F. Spectrochim. Acta, Part A 2007, 67, 619.
8. Borsari, M.; Ferrari, E.; Grandi, R.; Saladini, M. Inorg. Chim. Acta 2002, 328, 61.
9. Jiao, Y.; Wilkinson, J., IV; Pietsch, E. C.; Buss, J. L.; Wang, W.; Planalp, R.; Torti, F.
M.; Torti, S. V. Free Radical Biol. Med. 2006, 40, 1152.
10. Bussl, J. L.; Torti, F. M.; Torti, S. V. Curr. Med. Chem. 2003, 10, 1021.
11. Anand, P.; Kunnumakkara, A. B.; Newman, R. A.; Aggarwal, B. B. Mol.
Pharmaceutics 2007, 4, 807.
12. Kumar, S.; Narain, U.; Tripathi, S.; Misra, K. Bioconjugate Chem. 2001, 12, 464.
13. Shi, W.; Dolai, S.; Rizk, S.; Hussain, A.; Tariq, H.; Averick, S.; L’Amoreaux, W.; El
Idrissi, A.; Banerjee, P.; Raja, K. Org. Lett. 2007, 9, 5461.
14. Safavy, A.; Raisch, K. P.; Mantena, S.; Sanford, L. L.; Sham, S. W.; Rama Krishna,
N.; Bonner, J. A. J. Med. Chem. 2007, 50, 6284.
15. Bianchi, N. O.; Ayres, J. Exp. Cell. Res. 1971, 68, 253.
16. Santos, A. R., Jr.; Dolder, H.; Wada, M. L. F. J. Submicrosc. Cytol. Pathol. 2003, 35,
35.
17. Zhang, M.; Bian, F.; Wen, C.; Hao, N. J. Huazhong Univ. Sci. Technol. Med. Sci.
2007, 27, 339.
18. Menezo, Y. J. R.; Guerin, J. F.; Czyba, J. C. Biol. Reprod. 1990, 42, 301.
19. Rossi, T.; Coppi, A.; Bruni, E.; Ruberto, A.; Santachiara, S.; Baggio, G. Anticancer
Res. 2007, 27, 2555.
The 2008 cell line, established from a patient with serous cyst-
adenocarcinoma of the ovary and the cDDP-resistant C13* subline,
generated as previously described,36 were grown as monolayers in
RPMI 1640 medium (BioWhittaker Europe) containing 10% heat-
inactivated fetal bovine serum (BioWhittaker Europe) and 50 lg/
ml gentamycin sulfate. Cultures were equilibrated with humidified
5% CO2 in air at 37 °C. All studies were performed in Mycoplasma
negative cells, as routinely determined with the Mycotest detection
kit (Euroclone, Switzerland).
Vero cells, established from kidney cells of the African green
monkey (Cercopithecus aethiops), were chosen as a control cell
line18 These cells were obtained from the Istituto Zooprofilattico
(Brescia, Italy), and maintained in MEM medium (Whittaker Bio-
products, Walkersville, MD, USA) supplemented with 10% heat-
inactivated fetal calf serum (Gibco/BRL, Gaithersburg, MD, USA)
and penicillin G (100 lg/ml), streptomycin (100 lg/ml) (Sigma
Chemical Company, MO, USA). Cultures were equilibrated with
humidified 5% CO2 in air at 37 °C. The medium was renewed at
48 h intervals, and the cells were always subcultured when the
monolayers become confluent.
20. Mishra, S.; Narain, U.; Mishra, R.; Misra, K. Bioorg. Med. Chem. 2005, 13, 1477.
21. Pabon, H. J. J. Rec. Trav. Chim. 1964, 83, 379.
22. Buss, J. L.; Torti, F. M.; Torti, S. V. Curr. Med. Chem. 2003, 10, 1021.
ˇ
ˇ
23. Solcániova, E.; Hrnciar, P.; Liptaj, T. Org. Magn. Res. 1982, 18, 55.
24. Arrieta, A. F.; Mostad, A. Acta Crystallogr., Sect. E 2001, 57, o1198.
25. Atkinson, R. A.; Salah El Din, A. L.; Kieffer, B.; Lefevre, J. F.; Abdallah, M. A.
Biochemistry 1998, 37, 15965.
4.5. Cytotoxicity assay
26. Hara, Y.; Shen, L.; Tsubouchi, A.; Akiyama, M.; Umemoto, K. Inorg. Chem. 2000,
39, 5074.
27. Hegetschweiler, K.; Kradolfer, T.; Gramlich, V.; Hancock, R. D. Chem. Eur. J.
1995, 1, 74.
28. Hegetschweiler, K.; Hausherr-Primo, L.; Koppenol, W. H.; Gramlich, V.; Odier,
L.; Meyer, W.; Winkler, H.; Trautwein, A. X. Angew. Chem., Int. Ed. Engl. 1995, 34,
2242.
29. Hegetschweiler, K.; Ghisletta, M.; Hausherr-Primo, L.; Kradolfer, T.; Schmalle,
H. W.; Gramlich, V. Inorg. Chem. 1995, 34, 1950.
30. Wang, Y. J.; Pan, M. H.; Cheng, A. L.; Lin, L. I.; Ho, Y. S.; Hsieh, C. Y.; Lin, J. K. J.
Pharm. Biomed. Anal. 1997, 15, 1867.
31. Tomren, M. A.; Másson, M.; Loftsson, T.; Tønnensen, H. H. Int. J. Pharm. 2007,
338, 27–34.
Cell growth in the absence or in the presence of the compounds
was assessed by a modification of the crystal violet dye assay.37
Cells were seeded at a density of 4 ꢁ 104 cells/well of a 24-well cul-
ture plate (Sarstedt, Numbrecht, Germany) and cultured for 72 h at
37 °C (control cells). Treated cells were incubated with different
concentrations of the compounds and were grown under the same
conditions as non-treated (control) cells. After incubation, the cul-
ture medium was removed and the cells were washed with 0.1 ml
of 0.1 M phosphate-buffered saline (PBS), pH 7.4 at 37 °C, then cell
monolayer was fixed with methanol prior to staining with 0.1%

3052
E. Ferrari et al. / Bioorg. Med. Chem. 17 (2009) 3043–3052
32. Adams, B. K.; Ferstl, E. M.; Davis, M. C.; Herold, M.; Kurtkaya, S.; Camalier, R. F.;
Hollingshead, M. G.; Kaur, G.; Sausville, E. A.; Rickles, F. R.; Snyder, J. P.; Liotta,
D. C.; Shoji, M. Bioorg. Med. Chem. 2004, 12, 3871–3883.
34. Chou, T. C.; Talalay, P. Adv. Enzyme Regul. 1984, 27–55.
35. Fischer, E. Chem. Ber. 1911, 44, 1898–1904.
36. Andrews, P. A.; Albright, K. D. Cancer Res. 1992, 52, 1895–1901.
37. Kueng, W.; Siber, E.; Eppenberger, U. Anal. Biochem. 1989, 182, 16.
33. Kuhar, M.; Imran, S.; Singh, N. J. Cancer Mol. 2007, 3, 121–128.