Antioxidant and antiviral activities of silybin fatty acid conjugates

Article abstract of DOI:10.1016/j.ejmech.2009.11.056

Two selective acylation methods for silybin esterification with long-chain fatty acids were developed, yielding a series of silybin 7-O- and 23-O-acyl-derivatives of varying acyl chain lengths. These compounds were tested for their antioxidant (inhibition of lipid peroxidation and DPPH-scavenging) and anti-influenza virus activities. The acyl chain length is an important prerequisite for both biological activities, as they improved with increasing length of the acyl moiety.

Full text of DOI:10.1016/j.ejmech.2009.11.056

European Journal of Medicinal Chemistry 45 (2010) 1059–1067
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Antioxidant and antiviral activities of silybin fatty acid conjugates
a
a
a
a
a
b
ˇ
ꢀ
´
ˇ
´
´
ˇ
´
Radek Gazak , Katerina Purchartova , Petr Marhol , Lucie Zivna , Petr Sedmera , Katerina Valentova ,
Nobuo Kato c, Hiroyo Matsumura c, Kunihiro Kaihatsu c , Vladimır Kren
,
a,
*
**
´
ˇ
a
ˇ
´
Institute of Microbiology, Academy of Sciences of the Czech Republic, V´ıdenska 1083, CZ-142 20 Prague, Czech Republic
b
ꢁ
ˇ
´
Department of Medical Chemistry and Biochemistry, Faculty of Medicine and Dentistry, Palacky University, Hnevot´ınska 3, CZ-775 15 Olomouc, Czech Republic
c The Institute of Scientific and Industrial Research, Department of Organic Fine Chemicals, Osaka University, 8-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Two selective acylation methods for silybin esterification with long-chain fatty acids were developed,
yielding a series of silybin 7-O- and 23-O-acyl-derivatives of varying acyl chain lengths. These
compounds were tested for their antioxidant (inhibition of lipid peroxidation and DPPH-scavenging) and
anti-influenza virus activities. The acyl chain length is an important prerequisite for both biological
activities, as they improved with increasing length of the acyl moiety.
Received 14 August 2009
Received in revised form
30 October 2009
Accepted 27 November 2009
Available online 6 December 2009
Ó 2009 Elsevier Masson SAS. All rights reserved.
Keywords:
Silybin
Silibinin
Influenza virus
Lipophilic derivatives
Lipid peroxidation
Antiviral
1. Introduction
the identification of suitable sites for derivatization of the silybin
molecule without losing the biological activity of the resulting
The flavonolignan silybin (1) is the major biologically active
component of an extract from the seeds of the milk thistle (Silybum
marianum (L.) Gaertn.) known as silymarin. Silybin is an active
component in numerous phytopreparations used in the prevention
and treatment of various liver diseases and as a protectant against
a number of hepatotoxins and mycotoxins [1–3]. Natural silybin (1)
is a nearly equimolar mixture of two diastereoisomers (Fig.1) [4–7].
The cytoprotectivity of 1 is based on several mechanisms
operating at various cell levels [8]. Silybin acts mainly as an effec-
tive radical scavenger (antilipoperoxidant), and also as an antioxi-
dant [9]. The role of the individual hydroxyl groups of silybin and
molecular mechanisms of its antiradical and antioxidant action
have been explained recently [10,11]. The radical scavenging
activity of 1 could be also partly involved in cell regulatory path-
ways based on reactive oxygen species (ROS) [12]. Detailed char-
acterization of the role of the respective hydroxyl groups and
moieties of the silybin molecule in these processes [11] has enabled
conjugates.
In the last decade, silybin has been in the spotlight due to its
multiple beneficial activities [8,13,14] not directly related to its
hepatoprotective and/or antioxidant (radical scavenging) activities.
These mostly include anticancer and chemopreventive behavior.
The prevention and treatment of prostate hyperplasia involving
adenocarcinoma is a very recent and important silymarin applica-
tion, which is also documented by current successful clinical tests
(clinical phase II) [15].
The bioavailability and therapeutic efficacy of silybin is rather
limited by its very low solubility in water (430 mg/L). A number of
silybin water-soluble semisynthetic derivatives were designed to
overcome this problem, e.g., silybin bis-hemisuccinate (Legalon)
[16], silybin-23-O-phosphate [17], silybin 23-O-
b-glycosides [6],
and silybinic acid [9]. However, modifications of silybin leading to
an increase in its water-solubility usually led to an impairment of
its antioxidant (antiradical) activity in the lipophilic milieu [9].
On the contrary, some lipophilic preparations of silybin, e.g., its
non-covalent complex with phosphatidyl choline – (Silipide, IdB
1016, Indena, IT) [18,19], possessed not only better bioavailability
than silybin [20,21] but also exhibited higher antioxidant activities
[22,23]. Moreover, Silipide also exhibits promising anticancer
activities as was demonstrated by its significant inhibition of the
* Corresponding author. Tel.: þ420 296 442 510; fax: þ420 296 442 509.
** Corresponding author.
E-mail addresses: kunihiro@sanken.osaka-u.ac.jp (K. Kaihatsu), kren@biomed.
ˇ
cas.cz (V. Kren).
0223-5234/$ – see front matter Ó 2009 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2009.11.056

ꢀ
´
R. Gazak et al. / European Journal of Medicinal Chemistry 45 (2010) 1059–1067
1060
growth of human ovarian cancer xenografts [3]. Furthermore,
silybin-containing liposomes composed of phospholipids and
cholesterol were found to be much more effective hepatoprotectors
in the murine model than silybin itself [24].
lipophilization of the silybin molecule by its conjugation to fatty
acids can not only improve its antioxidant but also its anti-influ-
enza virus activities.
The aim of the present study was to develop general method(s)
for selective acylation of the silybin molecule with monocarboxylic
acids and preliminary testing of these derivatives for their antioxi-
dant and anti-influenza virus activities compared to 1. Detailed
knowledge of the effect of particular silybin OH groups on the
antiradical mechanism [10,11] enabled us to rationally determine
the most suitable sites for substitution. A further aim of this
work was to evaluate the effect of the acyl chain length on the
antioxidant and anti-influenza-virus activities of silybin and its
lipoderivatives.
Such a positive effect of lipophilic formulations of silybin clearly
demonstrates that the membrane-stabilizing function is one of the
basic molecular mechanisms of silybin activity [25,26] and that
silybin–phospholipid complexes are more effective than silybin.
Nevertheless, current knowledge on the interactions of silybin with
lipid bilayers is very poor. According to a study on the influence of
silybin on liver microsomal membranes, it is believed that silybin
molecules are incorporated into the hydrophobic–hydrophilic
interface of membranes [27] and mainly interact with the polar
head group region of the lipid bilayers, although some of the silybin
molecules can partition into the hydrophobic region of the
membrane [28].
2. Results and discussion
The conjugation of the water-soluble antioxidant to a long-chain
fatty acid has been demonstrated to improve the incorporation of
2.1. Chemistry
antioxidants into liposomes (used as
a model for cellular
According to our recent studies [10,11], the most suitable posi-
tions for silybin modification are 7-OH or 23-OH of 1, as they are not
responsible for its major antioxidant and antiradical action (23-OH)
or can even act as a pro-oxidant moiety (7-OH). Although a selec-
tive enzymatic procedure for the acylation of silybin’s primary
alcoholic group (23-OH) was developed recently [39], producing
reasonable yields of the corresponding esters of silybin with several
short chain monocarboxylic acids, this procedure requires rather
long reaction times, activation of the corresponding carboxylic
acids (e.g., vinyl esters) and provides rather low yields with long-
chain acids. As an enzymatic approach leads exclusively to the
esters at 23-OH of 1, the development of new procedures for the
selective esterification of other positions of 1 are required. More-
over, the development of a suitable general esterification method
for the primary alcoholic group of 1 will be required to obtain
sufficient yields of the corresponding esters of 1 with long-chain
carboxylic acids.
membranes) [29]. An improvement of the antioxidant activity in
heterogeneous systems by lipophilization of the hydrophilic anti-
oxidant has previously been observed: the antioxidant efficiency of
retinyl ascorbate (an ester of lipophilic retinoic acid and the water-
soluble antioxidant ascorbic acid) [30] as well as ferulates, dihydro-
ferulates, caffeates, dihydrocaffeates and gallates were generally
found to exhibit a larger antioxidative effect than the parent
compounds [31–33]. Moreover, most of the previous studies proved
that the antioxidant activity of these conjugates is dependent on
the acyl (and/or alkyl) chain length and is generally better for
derivatives with a longer aliphatic chain. The conjugation of epi-
catechin with fatty acids was also shown to change its biological
activities, such as an increase in DNA polymerase inhibition and
inhibition of angiogenesis [34].
There are several reports that describe the anti-hepatitis C virus
(HCV) effect of medicinal herbs in clinical trials. Although the most
preparations did not show positive effects on the clearance of
serum HCV RNA or anti-HCV antibodies or on serum liver enzymes,
silybin exhibited a significant effect, reducing serum aspartate
aminotransferase and gamma-glutamyltranspeptidase activities
[35,36]. The antiviral mechanism of silybin action is still unclear
because its chemical and pharmacological properties may change
when it is administered orally. However, silybin A, silybin B, iso-
Initially, esterification reactions of silybin (1) using free
carboxylic acids utilizing various well known activation reagents and
procedures (1,10-carbonyldiimidazole (CDI), N,N0-dicyclohexyl-
carbodiimid/4-dimethylaminopyridine (DCC/DMAP)
– Steglich
esterification, diethyl azodicarboxylate/triphenylphosphine (DEAD/
Ph3P) – Mitsunobu reaction) were tested, however, these attempts
failed due to their low selectivity (DCC/DMAP) or did not work at all
(CDI, Mitsunobu reaction). As a result, more reactive acylating
reagents (namely acyl chlorides) were chosen.
Selective acylation of the 7-OH position of silybin was accom-
plished with the respective acyl chloride (1.5 equiv) in pyridine
(Scheme 1). As the 7-OH of 1 represents its most acidic hydroxyl
group, the achieved selectivity can be explained by an exclusive
phenolate generation at this position caused by pyridine as a weak
base. On the contrary, the reaction of 1 with acyl chloride in the
presence of a Lewis acid (BF3$Et2O) led to the selective esterifica-
tion of the 23-OH group of 1 (Scheme 1). Interestingly, the
silybin A, and isosilybin B elicited the strongest anti-NF-
kB and
anti-HCV activities of the standardized extracts of Silybum maria-
num in vitro [37]. Based on these findings, the antiviral effect of
silybin derivatives may be ascribed to their antioxidative activities.
Peterhands et al. demonstrated that influenza virus infection
induces oxidative stress and the production of ROS from phago-
cytes [38]. These agens may modify proteins and lipids, damage
DNA, and alter mitochondrial membrane potential, producing
apoptotic cytotoxicity. Effective delivery of silybin to the cellular
membrane may increase its antiviral efficacy. Therefore,
23
16
CH2OH
CH2OH
10
11
O
O
O
D
B
17
HO
7
O
C
19 OCH3
HO
O
O
OCH3
OH
2
3
14
O
E
A
20
OH
OH
OH
5
OH
O
OH
silybin B (2R, 3R, 10S, 11S)
silybin A (2R, 3R, 10R, 11R)
Fig. 1. Isomers of natural silybin (1).

ꢀ
´
R. Gazak et al. / European Journal of Medicinal Chemistry 45 (2010) 1059–1067
1061
23
23
R
O
O
O
CH2OH
O
O
CH2OH
HO
7
O
O
19 OCH3
O
7
O
O
19 OCH3
A
20
3
20
3
OH
OH
OH
OH
5
O5H
OH
1
2 a - g
B
a R = Pr
O
b
R = CH (CH2)5CH3
2
23
CH2O
R
O
O
c
R = CH (CH2)9CH3
2
7
19
HO
O
OCH3
OH
d R = CH (CH2)13CH3
2
eR = CH2-(CH2)6-CH=CH-(CH ) -CH3
2 7
20
3
OH
O5H
O
f R = CH -(CH2)5-(CH2-CH=CH) -(CH2)4-CH3
2
2
g R = CH -(CH2)8-CH=CH-(CH ) -CH3
3 a - g
2
2 7
Scheme 1. Reagents and conditions. A. Acyl chloride (1.5 equiv), Py, 0 ꢀC, 1 h (29–45%); B. Acyl chloride (1 equiv), BF3$Et2O (2.4 equiv), CH2Cl2/MeCN (1:1, v/v), 2 h at 0 ꢀC, then 1 h at
r.t. (28–32%).
formation of 23-O-acetylsilybin as a side product was observed in
this reaction (less than 10%) despite no acetyl donor being present
in the reaction mixture, suggesting that acetonitrile (used as a co-
solvent) is able to act as a weak acetylation reagent under these
conditions.
Although the yields of the corresponding esters were moderate,
the reaction conditions were optimized to reduce side product
formation (di- and tri-esters) enabling a convenient separation of
the product from the unreacted 1 using a short and rapid column
chromatography. Despite the complex nature of silybin (five
hydroxyl groups) the fast reaction and selectivity make these
methods a versatile approach for silybin esterification.
Structures of silybin esters were determined by NMR spectros-
copy. All OH signals are visible in the 1H NMR spectra of acyl sily-
bins measured in DMSO, their experimental assignment was
proved by HMBC. The absence of some of them (compared to the
parent compound) provides the first hint of the acylation site. The
downfield acylation shift is observable with H-3 or H-23 protons.
The heteronuclear coupling between H-3 or H-23 and the acyl
carbonyl is a direct proof of the acylation at these positions. An
acylation at C-7 is manifested by chemical shift changes of C-6, C-7,
C-8, and C-4a.
inhibition activity of antioxidants against lipid peroxidation.
Therefore, a series of 7-O- and 23-O-acyl-derivatives was synthe-
sized with the aim of determining which of these positions on
silybin is more suitable for the substitution and to evaluate the
effect of the length of aliphatic acyl chain upon its lipid perox-
idation inhibitory activity. Although the inhibition of lipid perox-
idation represents an assay that better reflects the ability of
antioxidants to protect against oxidative damage in biological
systems, the radical-scavenging activities of acyl-derivatives
against DPPH (1,1-diphenyl-2-picrylhydrazyl) radicals were also
measured for comparison with other previously published data on
silybin. Esters of silybin with unsaturated acids were also prepared,
however due to their enormous susceptibility to auto-oxidation
observed during their structural analysis, these compounds were
excluded from biological testing.
The results obtained (Table 1) clearly demonstrate that the
length of the acyl chain is decisive for lipid peroxidation inhibitory
activity. Surprisingly, acyl derivatives with short acyl chains (C4–C8)
are inferior inhibitors than the parent silybin. On the other hand,
longer acyl derivatives (C12 and mainly C16) are better inhibitors.
This observation is consistent with the behavior of 2,2,5,7,8-pen-
tamethyl-6-chromanol (an analogue of
a methyl instead of the phytyl group) and
compounds exert essentially the same antioxidant activity in an
organic solution, but -tocopherol possesses a higher activity
a-tocopherol having only
a
-tocopherol – both
2.2. Antioxidant activity
a
The role of the individual hydroxyl groups of silybin on its
antiradical action was recently determined [10,11]. These studies
not only defined the positions, which react with radicals directly
(20-OH of 1), but also evaluated the effect of other OH groups
substitutions on the antioxidant activity of 1. Despite substantial
amounts of useful knowledge being obtained, the effect of substi-
tution was only studied with the methyl group. Such an approach
enables the SAR-study of similar compounds but cannot evaluate
the effect of the substitution in terms of the change in physico-
chemical properties, e.g. their lipophilicity. As the lipid peroxi-
dation of biomembranes caused by free radicals is one of the most
serious pathological events in organisms, most studies have been
aimed at evaluating the inhibitory activity of antioxidants against
this process. Lipophilicity is a crucial factor, which influences the
against lipid peroxidation in the membranes and low density
lipoproteins due to higher mobility within and between
membranes and particles [40–42]. The esterification of silybin by
short chain acids probably improves its interaction with bio-
membranes compared with silybin itself, however, it simulta-
neously reduces its mobility and hinder its ability for self-
stabilization by dimerization [11,43]. This ability is probably
restored with longer chain esters, which, due to their higher lipo-
philicity and better mobility within membranes, exhibited a higher
inhibitory activity in both types of silybin substitutions.
As far as the site of silybin acylation is concerned, the more
suitable position for its substitution is the 7-OH group, mainly with
long-chain fatty acids. A plausible explanation of this behavior
probably lies in the recently described [11] pro-oxidative activity of

ꢀ
´
1062
R. Gazak et al. / European Journal of Medicinal Chemistry 45 (2010) 1059–1067
Table 1
Table 2
Antioxidant properties of silybin and its fatty acid esters (LPX – inhibition of lipid
peroxidation).
Cytotoxicity and anti-influenza virus A/PR8/34/(H1N1) activity of silybin and its
fatty acid esters.
Antiradical activity
DPPHa
Antioxidant activity
LPXb
Cytotoxicity
Antiviral
activity EC50
Selectivity
SIc
a
b
CC50
(
m
mol/L)
(mmol/L)
IC50 (mmol/L)
IC50 (mmol/L)
Silybin (1)
>800
300
240
>560
NDd
300
30
120
>480
ND
14
6
–
Silybin (1)
2.1 ꢁ 0.1
2.3 ꢁ 0.1
4.0 ꢁ 0.2
3.0 ꢁ 0.2
3.6 ꢁ 0.3
3.4 ꢁ 0.6
3.2 ꢁ 0.4
2.8 ꢁ 0.8
3.2 ꢁ 0.6
58.1 ꢁ 3.6
100.6 ꢁ 29.4
116.7 ꢁ 36.0
44.7 ꢁ 0.5
28.9 ꢁ 1.0
91.7 ꢁ 5.2
61.1 ꢁ 5.9
50.5 ꢁ 4.5
51.9 ꢁ 3.8
7-O-Butanoylsilybin (2a)
7-O-Octanoylsilybin (2b)
7-O-Dodecanoylsilybin (2c)
7-O-Palmitoylsilybin (2d)
23-O-Butanoylsilybin (3a)
23-O-Octanoylsilybin (3b)
23-O-Dodecanoylsilybin (3c)
23-O-Palmitoylsilybin (3d)
21.4
40.0
>93.3
–
7-O-Butanoylsilybin (2a)
7-O-Octanoylsilybin (2b)
7-O-Dodecanoylsilybin (2c)
7-O-Palmitoylsilybin (2d)
23-O-Butanoylsilybin (3a)
23-O-Octanoylsilybin (3b)
23-O-Dodecanoylsilybin (3c)
23-O-Palmitoylsilybin (3d)
6
NDd
24
18
3
12.5
1.7
40.0
>160
3
a
CC50 represents the concentration of compound required to reduce cell viability
by 50% relative to the control well without test compound.
EC50 represents the concentration of compound required to reduce plaque
number by 50% relative to the control well without test compound.
Data are expressed as mean values from three measurements in three independent
experiments and standard deviations.
DPPH – 1,1-diphenyl-2-picrylhydrazyl radical scavenging; IC50 (mM) – the
concentration of the tested compound required to reduce the absorbance of DPPH
by 50%.
b
a
c
SI (Selectivity index) is the ratio of CC50 to EC50
.
d
CC50 and EC50 of 2d were not determined due to its low water solubility.
b
LPX – Inhibition of microsomal lipid peroxidation–the activity was calculated as
the concentration of the tested compound inhibiting the color reaction with thio-
barbiturate by 50% (IC50).
observations with methyl-derivatives of silybin, where 7-OH
methylation slightly lowered the DPPH-scavenging activity of the
corresponding derivative in comparison with silybin [11]. However,
7-OH methylation led to a very low decrease in scavenging activity
compared to acylation. The effect of silybin 23-OH acylation was
much more striking, since 23-OH was considered to be ineffective
for radical scavenging. Nevertheless, a similar decrease in capacity
of lipophilic conjugates of water-soluble antioxidants to scavenge
DPPH radicals in a homogenous system (EtOH or MeOH solution)
was observed in most of the previous studies, e.g. for esters of
dihydrocaffeic acid [44] and gallic acid esters [31,33]. Additionally,
this moiety. Accordingly, substitution of the 7-OH can switch off
this undesirable activity (as far as antioxidant activity is concerned)
and in long-chain acyl-derivatives leads to more effective inhibitors
compared to the corresponding 23-O-acyl-derivatives. Neverthe-
less, the impact of esterification of both silybin positions on the
spatial interactions with lipid bilayers can play an important role in
the resulting antioxidant activity of the conjugate during lipid
oxidation in heterogeneous systems [31].
Our results on the DPPH-scavenging activity for 7-O-acyl-
derivatives (Table 1) are partially consistent with our recent
Table 3
1H NMR data (400 MHz, d6-DMSO, 30 ꢀC) of 7-O-acyl silybins 2a-d.
Proton
2a
2b
2c
2d
2
3
5.234 d (11.6)
4.777 dd (11.6,6.0)
4.768 dd (11.6,6.0)
6.358 d (2.1)
6.317 d (2.1)
6.311 d (2.1)
4.181 m
5.232 d (11.6)
5.229 d (11.6)
4.772 dd (11.6,6.3)
4.761 dd (11.6,6.3)
6.346 d (2.1)
6.305 d (2.1)
6.299 d (2.1)
4.176 ddd (7.8,4.7,2.4)
4.171 ddd (7.8,4.7,2.4)
4.919 d (7.8)
7.111 d (1.8)
7.105 d (1.8)
7.036 dd (8.3,1.8)
7.032 dd (8.3,1.8)
6.984 d (8.3)
6.981 d (8.3)
7.016 d (2.0)
5.225 d (11.6)
4.768 dd (11.6,6.4)
4.758 dd (11.6,6.4)
6.342 d (2.1)
6.301 d (2.1)
6.295 d (2.1)
4.172 ddd (7.9,4.9,2.5)
4.167 ddd (7.9,4.9,2.5)
4.911 d (7.9)
7.105 d (2.1)
7.099 d (2.1)
7.031 dd (8.3,2.1)
7.028 dd (8.3,2.1)
6.980 d (8.3)
6.978 d (8.3)
7.013 d (2.0)
4.776 dd (11.6,6.3)
4.765 dd (11.6,6.3)
6.349 d (2.0)
6.308 d (2.0)
6.303 d (2.0)
4.180 ddd (7.8,4.6,1.7)
4.173 ddd (7.8,4.6,1.7)
4.918 d (7.8)
7.113 d (2.0)
7.107 d (2.0)
7.037 dd (8.2,2.0)
7.034 dd (8.3,2.0)
6.985 d (8.2)
6.983 d (8.3)
7.021 d (2.0)
7.017 d (2.0)
6.807 d (8.1)
6.805 d (8.1)
6
8
10
11
13
4.920 d (7.8)
7.114 d (2.0)
7.109 d (2.0)
7.039 dd (8.0,2.0)
7.036 dd (8.0,2.0)
6.986 d (8.0)
6.983 d (8.0)
7.021 d (1.7)
7.017 d (1.7)
6.807 d (8.1)
15
16
18
21
22
7.009 d (2.0)
6.800 d (8.2)
6.806 d (8.1)
6.805 d (8.1)
6.869 dd (8.1,2.0)
6.871 dd (8.1,1.7)
6.870 dd (8.1,2.0)
6.869 dd (8.1,2.0)
3.550 ddd (12.3,4.5,1.7)
3.356 ddd (12.3,4.6,4.5)
3.782 s
6.863 dd (8.2,2.0)
23 d
23 u
19-OMe
3.551 dm (11.7)
3.352 m
3.782 s
3.551 ddd (12.2,4.4,2.5)
3.355 ddd (12.2,4.7,4.4)
3.782 s
3.543 ddd (12.2,5.3,2.5)
3.348 ddd (12.2,5.3,4.9)
3.776 s
3.781 s
3-OH
5-OH
7-OH
20-OH
23-OH
5.940 d (6.0)
11.660 s
–
9.116 s
4.924 t (4.5)
5.930 d (6.3)
11.664 s
–
9.114 s
4.924 t (4.5)
5.932 d (6.3)
11.669 s
–
9.109 s
4.923 t (4.4)
5.937 d (6.4)
11.890 s
–
9.128 s
4.953 t (5.3)
Additional signals – 2a: 2.541 (t, 2 H, J ¼ 7.2 Hz, 2 ꢂ H-20), 1.636 (m, 2 H, 2 ꢂ H-30), 0.947 (t, 3 H, J ¼ 7.4 Hz, 3 ꢂ H-30); 2b: 2.551 (t, 2 H, J ¼ 7.4 Hz, 2 ꢂ H-20), 1.610 (m, 2 H, 2 ꢂ H-
30), 0.862 (t, 3 H, J ¼ 6.8 Hz, 3 ꢂ H-80); 2c: 2.546 (t, 2 H, J ¼ 7.3 Hz, 2 ꢂ H-20), 1.605 (m, 2 H, 2 ꢂ H-30), 0.851 (t, 3 H, J ¼ 6.8 Hz, 3 ꢂ H-120); 2d: 2.544 (t, 2 H, J ¼ 7.3 Hz, 2 ꢂ H-20),
1.600 (m, 2 H, 2 ꢂ H-30), 0.847 (t, 3 H, J ¼ 7.0 Hz, 3 ꢂ H-160).

ꢀ
´
R. Gazak et al. / European Journal of Medicinal Chemistry 45 (2010) 1059–1067
1063
these studies demonstrated that the DPPH-scavenging ability in
a homogenous system is independent of the aliphatic chain length
and usually lower than that of the parent non-conjugated
antioxidant.
a concentration of 800
m
M. However, the cytotoxicities of 7-O-
and 23-O-acyl-derivatives increased as a function of acyl length.
Notably, the 7-O-octanoyl (2b), 23-O-octanoyl (3b), and 23-O-
dodecanoyl (3c) silybin derivatives exhibited relatively high
cytotoxicities (Table 2). This may be due to their increased cell
2.3. Cytotoxicities of silybin and its acyl-derivatives
membrane permeabilities. A similar phenomenon was also
reported in the case of acyl derivatives of epigallocatechin-3-O-
gallate [45]. On the other hand, 7-O-dodecanoyl (2c), 7-O-pal-
mitoyl (2d), and 23-O-palmitoyl (3d) esters exhibited consider-
ably lower cytotoxicities, probably due to their poor water
solubilities.
The cytotoxic effects of silybin and its 7-O- and 23-O-acyl-
derivatives to Madin-Darby Canine Kidney (MDCK) cells, given as
CC50, were evaluated by cell viability assay. As shown in Table 2,
silybin (1) did not display an apparent cytotoxic effect even at
Table 4
13C NMR data (100 MHz, d6-DMSO, 30 ꢀC).
C
2
1
2a
82.81
82.76
71.81
2b
82.79
82.75
71.80
2c
82.80
82.76
71.81
2d
82.77
82.57
71.81
71.52
199.47
3a
82.52
82.48
71.52
3b
82.52
82.47
71.53
3c
82.51
82.47
71.53
3d
82.52
82.47
71.54
82.65
82.62
71.56
71.49
197.76
197.74
100.55
163.38
96.16
166.90
95.11
162.55
162.54
78.23
3
4
71.75
71.74
71.75
71.45
71.45
71.46
71.47
199.51
199.49
104.81
161.99
102.94
158.09
101.68
161.86
161.85
78.17
199.49
199.47
104.80
161.98
102.91
158.09
101.65
161.85
161.84
78.17
199.50
199.48
104.79
161.98
102.90
158.09
101.63
161.85
161.83
78.19
197.72
197.69
100.50
163.35
96.13
166.90
95.08
162.49
162.48
75.05
197.69
197.66
100.48
163.35
96.14
166.96
95.09
162.49
162.47
75.05
197.71
197.68
100.49
163.34
96.12
166.88
95.07
162.48
162.47
75.05
197.65
197.61
100.45
163.36
96.15
167.01
95.11
162.48
4a
5
6
7
8
104.80
163.34
102.92
158.10
101.66
161.98
161.84
78.19
8a
10
75.06
78.20
78.17
75.03
75.03
11
12a
75.93
75.89
143.36
143.33
116.71
116.75
129.72
129.68
121.44
121.32
116.52
116.38
143.83
143.80
127.53
127.51
111.83
111.75
147.69
147.68
147.09
147.08
115.38
75.89
143.32
75.88
75.90
143.50
143.32
116.64
75.99
75.97
143.13
143.09
116.81
116.69
130.57
130.54
121.57
121.37
116.44
116.38
143.25
143.23
126.63
75.96
75.96
143.25
143.22
116.80
116.67
130.57
130.54
121.54
121.32
116.44
116.37
143.21
143.09
126.63
143.33
143.32
116.67
116.58
130.15
130.11
121.37
121.22
116.42
116.37
143.73
143.70
127.59
143.33
143.31
116.69
116.62
129.70
129.67
121.41
121.27
116.41
116.37
143.82
143.80
127.51
143.13
143.10
116.79
116.72
130.57
130.53
121.57
121.41
116.46
116.41
143.25
143.24
126.63
143.12
143.08
116.80
116.68
130.56
130.53
121.55
121.33
116.43
116.37
143.24
143.22
126.63
126.62
111.77
111.72
147.82
147.81
147.39
147.38
115.44
13
14
15
16
16a
17
18
19
20
21
22
116.70
116.63
129.71
129.67
121.42
121.28
116.42
129.68
121.31
121.30
116.38
143.82
143.80
127.51
143.80
143.67
127.54
111.86
111.81
147.73
147.72
147.71
147.05
114.44
115.43
120.59
111.79
111.73
147.67
111.78
111.73
147.68
147.67
147.08
147.06
115.36
111.80
147.69
111.80
111.75
147.81
147.82
147.39
147.38
115.48
115.47
120.68
120.66
62.42
55.74
172.46
35.12
17.87
13.40
111.78
111.74
147.83
147.82
147.39
111.78
111.73
147.83
147.82
147.40
147.07
115.36
120.55
147.16
147.07
115.38
115.46
115.44
120.67
120.64
62.43
55.74
172.60
33.26
24.40
28.36
28.40
31.14
22.05
13.94
115.45
120.65
120.56
120.53
120.56
120.65
120.63
62.43
55.73
172.59
33.25
24.39
28.44
23
OMe
10
60.28
55.81
60.22
55.75
170.76
35.33
17.17
13.31
60.21
55.74
170.89
33.50
24.18
28.28
28.32
31.09
22.03
13.93
60.21
55.74
170.90
33.50
24.17
28.31
60.23
55.74
170.91
33.52
24.17
28.30
62.42
55.74
172.60
33.27
24.40
28.44
20
30
40
50
60
70
80
90
100
110
120
13
140
150
160
31.30
22.10
13.94
31.32
22.11
13.96
31.31
22.11
13.96
13.96
Additional signals – 2c: 28.99 (2 C), 28.28, 28.71, 28.64; 2d: 29.04, 28.96, 28.84, 28.72, 28.63; 3c: 28.01 (2 C), 28.92, 28.72, 28.70. 3d: 29.08, 29.07, 29.04, 29.01, 28.91, 28.73.

ꢀ
´
R. Gazak et al. / European Journal of Medicinal Chemistry 45 (2010) 1059–1067
1064
Table 5
1H NMR data (400 MHz, d6-DMSO, 30 ꢀC) of 23-O-acyl silybins 3a-d.
Proton
3a
3b
3c
3d
2
3
5.095 d (11.2)
5.093 d (11.2)
4.602 m
5.093 d (11.0)
4.594 m
5.087 d (11.2)
4.595 dd (11.2,6.3)
4.582 dd (11.2,6.3)
5.915 d (2.1)
5.871 d (2.1)
5.864 d (2.1)
4.510 ddd (7.9,5.0,2.9)
4.505 ddd (7.9,5.0,2.9)
4.924 d (7.9)
7.112 d (1.9)
7.100 d (1.9)
7.031 dd (8.3,1.9)
7.022 dd (8.3,1.9)
6.978 d (8.3)
6.973 d (8.3)
7.020 d (2.0)
6799 d (8,1)
4.606 dd (11.2,5.2)
4.596 dd (11.2,5.2)
5.920 d (2.1)
5.879 d (2.1)
5.874 d (2.1)
4.520 ddd (7.9,4.9,2.8)
4.510 ddd (7.9,4.9,2.8)
4.932 d (7.9)
7.119 d (1.9)
7.112 d (1.9)
7.039 dd (8.3,1.9)
7.035 dd (8.3,1.9)
6.992 d (8.3)
6.989 d (8.3)
7.029 d (1.9)
6.810 d (8.2)
6.869 dd (8.2,1.9)
4.144 dd (12.4,2.8)
6
8
5.915 m
5.871 m
5.918 m
5.874 m
10
4.514 m
4.513 m
11
13
4.929 d (7.7)
7.115 m
4.929 d (7.8)
7.110 d (2.1)
15
16
7.035 m
7.032 m
6.974 d (8.2)
6.981 d (8.0)
18
21
22
23d
7.024 m
6.805 d (7.9)
6.863 m
7.022 m
6.804 d (7.9)
6.862 m
6.857 dd (8.2,2.0)
4.136 dd (12.3,2.9)
4.132 dd (12.3,2.9)
3.929 dd (12.3,5.0)
3.774 s
4.141 dd (12.2,2.5)
4.138 dd (12.1,2.5)
23 u
19-OMe
3.949 dd (12.4,4.9)
3.780 s
3.935 dd (12.2,4.6)
3.778 s
3.936 dd (12.1,4.5)
3.777 s
3.778 s
3.771 s
3-OH
5-OH
7-OH
20-OH
23-OH
5.790 d (5.2)
11.873 s
10.791 s
9.177 s
5.787 br s
11.873 s
10.826 s
9.181 s
–
5.788 d (5.7)
11.872 s
10.809 s
9.171 s
–
5.794 d (6.3)
11.870 s
10.810 br s
9.182 s
–
–
Additional signals – 3a: 2.273 (m, 2 H, 2 ꢂ H-20), 1.529 (m, 2 H, 2 ꢂ H-30), 0.877 (t, 3 H, J ¼ 7.4 Hz, 3 ꢂ H-40); 3b: 2.303 (m, 2 H, 2 ꢂ H-20), 1.501 (m, 2 H, 2 ꢂ H-30), 0.851 (t, 3 H,
J ¼ 6.8 Hz, 3 ꢂ H-80); 3c: 2.302 (m, 2 H, 2 ꢂ H-20), 1.502 (m, 2 H, 2 ꢂ H-30), 0.848 (t, 3 H, J ¼ 6.2 Hz, 3 ꢂ H-120); 3d: 2.298 (m, 2 H, 2 ꢂ H-20), 1.945 (m, 2 H, 2 ꢂ H-30), 1.240 (m, 2 H,
2 ꢂ H-40), 0.843 (t, 3 H, J ¼ 7.0 Hz, 3 ꢂ H-160).
2.4. Anti-influenza virus activities of silybin and its acyl-derivatives
with 7-OH esterification than with 23-OH. This observation is in
agreement with the previously suggested pro-oxidant function of
the 7-OH of silybin [11]. However, the rather contradictory results
of silybin esters in the DPPH assay, together with the absence of
precise information on the interaction of these lipophilic deriva-
tives with biomembranes and their behavior in solution (possi-
bility of micelles formation) calls for further studies of these
promising semisynthetic lipophilic derivatives of silybin.
According to the cytotoxic effect of silybin esters, the octanoyl
ester seemed to have the highest affinity to cells. However,
dodecanoyl and palmitoyl esters exhibited more potent antiviral
effects. This can be explained by both their antioxidant properties
and cell membrane permeabilities. The antiviral mechanisms of
silybin derivatives remain unclear, however, our synthetic study is
showing the way to the new leads in the preparation of anti-
influenza drugs, which is a very topical issue at present (e.g., swine
flu outbreak).
The anti-influenza virus activities of silybin and its 7-O- and 23-
O-acyl-derivatives were studied by plaque formation assay in an
MDCK monolayer. Silybin did not exhibit an antiviral effect (Table
2). In contrast, 7-O- and 23-O-acyl silybin derivatives showed
potent antiviral activities (Table 2). An exception is 7-O-palmi-
toylsilybin (2d) because it suffers from poor water solubility in the
cell culture medium. The antiviral effect of 7-O- and 23-O-acyl
silybin derivatives was also increased as a function of acyl length
(Table 2). Based on their cytotoxicities, we expected that the
introduction of octanoyl or dodecanoyl groups to silybin should
increase their cell permeability and increase their antiviral activity.
In fact, 7-O-octanoyl (2b), 7-O-dodecanoyl (2c), 23-O-dodecanoyl
(3c), and 23-O-palmitoyl (3d) silybin derivatives displayed potent
antiviral effects and high selectivity index values (Table 2). These
silybin derivatives could probably inhibit ROS generated by influ-
enza virus infection and inhibited the virus-induced cytopathic
effect.
4. Experimental protocols
3. Conclusion
4.1. General methods
Two selective acylation methods for silybin esterification with
monocarboxylic acids were developed, yielding a series of silybin
7-O- and 23-O-acyl-derivatives of different acyl chain lengths. The
results of experiments focusing on the inhibition by the tested
esters of the lipid peroxidation of microsomes induced by tert-
butylhydroperoxide proved that the length of the acyl chain of the
corresponding ester plays an important role in this activity.
However, the site of silybin acylation was found to be less signif-
icant in achieving a high inhibitory activity. The palmitates of
silybin were found to be the best inhibitors of lipid peroxidation
from our tested series; a more pronounced effect was connected
Silybin (mixture of A and B, ca 1:1) was kindly provided by
Dr. L. Cvak (TAPI Galena, IVAX Pharmaceuticals, Opava, CZ). Acyl
chlorides (octanoyl chloride, dodecanoyl chloride) were prepared
by reacting the corresponding carboxylic acid with an excess of
oxalyl chloride and used immediately after preparation without
further purification. All other chemicals were purchased from
Sigma–Aldrich and used without further purification. The reactions
were monitored by TLC on F254 silica gel (Merck) and the spots were
visualized with UV light and by charring with 5% H2SO4 in ethanol.
1H and 13C NMR spectra were recorded at 30 ꢀC in DMSO on
Varian Inova 400 or Bruker Avance III 400 NMR spectrometer (400

ꢀ
´
R. Gazak et al. / European Journal of Medicinal Chemistry 45 (2010) 1059–1067
1065
and 100 MHz, respectively). The residual solvent signal (dH 2.500, dC
4.2.5. 2-[2,3-dihydro-3-(4-hydroxy-3-methoxyphenyl)-2-
39.60) served as an internal reference. All 2D NMR experiments
were performed using the standard manufacturer’s software
(ChemPack 3.2 or TopSpin 2.1).
(hydroxymethyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-3,5-dihydroxy-
7-(cis-9-octadecenoyl)-oxy-4H-1-benzopyran-4-one (2e)
Compound 2e was prepared according to general method
A as a yellowish amorphous solid (34%). LRMS (ESI): m/z
769.4 (M þ Naþ). For 1H and 13C NMR data see Supplemen-
tary data.
HRMS – measurements were performed on a commercial
APEX-Ultra FTMS instrument equipped with a 9.4 T super-
conducting magnet and a Dual II ESI/MALDI ion source (Bruker
Daltonics, Billerica, USA) using MALDI ion source, positive-ion
mode. The interpretation of mass spectra was done using the
software package DataAnalysis version 3.4 (Bruker Daltonics,
Billerica, USA). Positive-ion electrospray ionization (ESI) mass
spectra were recorded on a LCQDECA spectrometer (ThermoQuest,
San Jose, USA). HPLC-analyses were carried out on a Spectra
Physics analytical system (San Jose, USA) comprised of an SP 8800
ternary gradient pump, an SP 8880 autosampler and a Spectra
Focus scanning UV/VIS detector. The columns employed were a EC
NUCLEOSIL 100-5 C18 AB, 125 ꢂ 3 mm (Macherey–Nagel, DE) or
Chromolite Speed ROD, RP-18e, 50 ꢂ 4.6 mm monolithic column
(Merck, DE).
4.2.6. 2-[2,3-dihydro-3-(4-hydroxy-3-methoxyphenyl)-2-
(hydroxymethyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-3,5-dihydroxy-
7-(cis,cis-9,12-octadecadienoyl)-oxy-4H-1-benzopyran-4-one (2f)
Compound 2f was prepared according to general method
A as a yellowish amorphous solid (29%). LRMS (ESI): m/z
767.4 (M þ Naþ). For 1H and 13C NMR data see Supple-
mentary data.
4.2.7. 2-[2,3-dihydro-3-(4-hydroxy-3-methoxyphenyl)-2-
(hydroxymethyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-3,5-dihydroxy-
7-(cis-11-eicosenoyl)-oxy-4H-1-benzopyran-4-one (2g)
Compound 2g was prepared according to general method
A as a yellowish amorphous solid (29%). LRMS (ESI): m/z
797.4 (M þ Naþ). For 1H and 13C NMR data see Supplemen-
tary data.
4.2. General method A – preparation of silybin 7-O-acyl derivatives
The corresponding acyl chloride (0.311 mmol, 1.5 equiv) was
added to a cooled solution (0 ꢀC) of silybin (1; 100 mg, 0.207 mmol)
in dry pyridine (7 mL) and the reaction mixture stirred at 0 ꢀC for
1 h. The reaction was stopped by diluting the reaction mixture with
ice-cold HCl (50 mL, 5% solution in water, v/v) and after brief stir-
ring the solution extracted with ethyl acetate (3 ꢂ 30 mL), the
organic phase washed with brine, dried over anhydrous Na2SO4 and
evaporated under reduced pressure.
4.3. General method B – preparation of silybin 23-O-acyl
derivatives
Silybin (1;100 mg; 0.207 mmol) was dissolved in a CH3CN/CH2Cl2
mixture (1:1, 10 mL, v/v). Acyl chloride (0.207 mmol, 1 equiv) and
ꢃ
BF3 Et2O (0.062 mL, 0.248 mmol, 50% solution in diethyl ether; v/v)
Flash chromatography on silica gel (chloroform/acetone/formic
acid 95:5:1) yielded the corresponding 7-O-acyl-silybin (2a-g).
were added and the mixture was stirred for 2 h at 0 ꢀC; then
additional BF3$Et2O (0.062 mL, 0.248 mmol, 50% solution in diethyl
ether, v/v) was added and the stirring continued for 1 h at room
temperature. The mixture was then diluted with a saturated ice-cold
solution of NaHCO3 (15 mL) and briefly stirred (approx. 10 min). The
solution was extracted with ethyl acetate (2 ꢂ 15 mL), the organic
phase washed with brine, dried over anhydrous Na2SO4 and evap-
orated under reduced pressure.
4.2.1. 2-[2,3-dihydro-3-(4-hydroxy-3-methoxyphenyl)-2-
(hydroxymethyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-7-
butanoyloxy-3,5-dihydroxy-4H-1-benzopyran-4-one (2a)
Compound 2a was prepared according to general method A as
a white amorphous solid (45%). HR-MS (MALDI) calcd for C29H28O11
(Mþ): 552.1632, found 552.1622. For 1H and 13C NMR data see
Tables 3 and 4.
Flash chromatography on silica gel (chloroform/acetone/formic
acid 95:5:1) yielded the corresponding 23-O-acyl-silybin (3a-g).
4.2.2. 2-[2,3-dihydro-3-(4-hydroxy-3-methoxyphenyl)-2-
(hydroxymethyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-3,5-dihydroxy-
7-octanoyloxy-4H-1-benzopyran-4-one (2b)
4.3.1. 2-[2,3-dihydro-2-(butanoyloxymethyl)-3-(4-hydroxy-3-
methoxyphenyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-3,5,7-
trihydroxy-4H-1-benzopyran-4-one (3a)
Compound 2b was prepared according to general method A as
a white amorphous solid (41%). HR-MS (MALDI) calcd for C33H36O11
(Mþ): 608.2258, found 608.2253. For 1H and 13C NMR data see
Tables 3 and 4.
Compound 3a was prepared according to general method B as
a white amorphous solid (28%). HR-MS (MALDI) calcd for C29H28O11
(Mþ): 552.1632, found 552.1615. For 1H and 13C NMR data see
Tables 4 and 5.
4.2.3. 2-[2,3-dihydro-3-(4-hydroxy-3-methoxyphenyl)-2-
(hydroxymethyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-3,5-dihydroxy-
7-dodecanoyloxy-4H-1-benzopyran-4-one (2c)
4.3.2. 2-[2,3-dihydro-3-(4-hydroxy-3-methoxyphenyl)-2-
(octanoyloxymethyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-3,5,7-
trihydroxy-4H-1-benzopyran-4-one (3b)
Compound 2c was prepared according to general method A as
a white amorphous solid (32%). HR-MS (MALDI) calcd for C37H44O11
(Mþ): 664.2884, found 664.2880. For 1H and 13C NMR data see
Tables 3 and 4.
Compound 3b was prepared according to general method B as
a white amorphous solid (30%). HR-MS (MALDI) calcd for C33H36O11
(Mþ): 608.2258, found 608.2241. For 1H and 13C NMR data see
Tables 4 and 5.
4.2.4. 2-[2,3-dihydro-3-(4-hydroxy-3-methoxyphenyl)-2-
(hydroxymethyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-3,5-dihydroxy-
7-hexadecanoyloxy-4H-1-benzopyran-4-one (2d)
4.3.3. 2-[2,3-dihydro-2-(dodecanoyloxymethyl)-3-(4-hydroxy-3-
methoxyphenyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-3,5,7-
trihydroxy-4H-1-benzopyran-4-one (3c)
Compound 2d was prepared according to general method A as
a white amorphous solid (36%). HR-MS (MALDI) calcd for C41H52O11
(Mþ): 720.3510, found 720.3509. For 1H and 13C NMR data see
Tables 3 and 4.
Compound 3c was prepared according to general method B as
a white amorphous solid (30%). HR-MS (MALDI) calcd for C37H44O11
(Mþ): 664.2884, found 664.2879. For 1H and 13C NMR data see
Tables 4 and 5.

ꢀ
´
R. Gazak et al. / European Journal of Medicinal Chemistry 45 (2010) 1059–1067
1066
4.3.4. 2-[2,3-dihydro-2-(hexadecanoyloxymethyl)-3-(4-hydroxy-3-
methoxyphenyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-3,5,7-
trihydroxy-4H-1-benzopyran-4-one (3d)
Compound 3d was prepared according to general method B as a white
amorphous solid (32%). HR-MS (MALDI) calcd for C41H52O11 (Mþ):
720.3510, found 720.3511. For 1H and 13C NMR data see Tables 4 and 5.
with 0.2% dimethyl sulfoxide containing respective silybin deriva-
tive sample at various concentrations was added to each well and
the cell cultures were incubated for 2 h. Then the cells were washed
twice with D-PBS, and maintained in D-MEM containing 0.2% BSA
for 24 h at 37 ꢀC in 5% CO2. The cytotoxicity of each silybin deriv-
atives was evaluated by methyltetrazolium (MTT) reduction assay
according to manufacture’s protocol (Promega).
4.3.5. 2-[2,3-dihydro-3-(4-hydroxy-3-methoxyphenyl)-2-((cis-9-
octadecenoyl)-oxymethyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-
3,5,7-trihydroxy-4H-1-benzopyran-4-one (3e)
4.6. Antiviral activity
Compound 3e was prepared according to general method B as
4.6.1. Evaluation of direct antiviral inhibitory effects of silybin
derivatives on influenza A/PR8/34 (H1N1)
a
yellowish amorphous solid (32%). LRMS (ESI): m/z 769.4
(M þ Naþ). For 1H and 13C NMR data see Supplementary data.
Each silybin derivative was mixed with the influenza virus in
Opti-MEM I reduced serum medium containing 0.2% DMSO. After
being incubated for 30 min at room temperature, the mixed solu-
tion was applied to a confluent monolayer of MDCK cells in a 6-well
plate (MOI ¼ 2.5 ꢂ 10-4) and incubated for 1 h at room temperature.
Then, after the solution was removed from each well, the cell sheets
were washed with D-PBS. After that, overlay medium (DMEM
containing 0.8% Oxoid agar No.1, 6.0 ꢂ 10-4% trypsin, and 0.2% BSA)
was added to each well. After incubating for approx. 2 days at 37 ꢀC
in 5% CO2, the cell sheets were fixed with 5% glutaraldehyde and
stained with methylene blue solution. The plaques formed in each
well were counted and the inhibitory effects of each sample on
virus infection were evaluated.
4.3.6. 2-[2,3-dihydro-3-(4-hydroxy-3-methoxyphenyl)-2-((cis,cis-
9,12-octadecadienoyl)-oxymethyl)-1,4-benzodioxin-6-yl]-2,3-
dihydro-3,5,7-trihydroxy-4H-1-benzopyran-4-one (3f)
Compound 3f was prepared according to general method B as
a
yellowish amorphous solid (29%). LRMS (ESI): m/z 767.4
(M þ Naþ). For 1H and 13C NMR data see Supplementary data.
4.3.7. 2-[2,3-dihydro-2-((cis-11-eicosenoyl)-oxymethyl)-3-(4-
hydroxy-3-methoxyphenyl)-1,4-benzodioxin-6-yl]-2,3-dihydro-
3,5,7-trihydroxy-4H-1-benzopyran-4-one (3g)
Compound 3g was prepared according to general method B as
a
yellowish amorphous solid (27%). LRMS (ESI): m/z 797.4
(M þ Naþ). For 1H and 13C NMR data see Supplementary data.
Acknowledgements
4.4. Antioxidant activity
This work was supported by grant KJB400200701 from the
Grant Agency of the Academy of Sciences of Czech Republic and by
institutional research concept AV0Z50200510 and grants OC08049,
LC06010 and MSM6198959216 from the Ministry of Education,
Youth and Sports of the Czech Republic.
4.4.1. DPPH (1,1-diphenyl-2-picrylhydrazyl radical) scavenging
The absorbance change in DPPH (stable radical) was measured
in the reaction mixture containing a solution of the compound
tested (65 mL, 0–20.0 mM, DMSO/MeOH 1:9, v/v) and 65 mL of DPPH
(0.1 mM, DMSO/MeOH 1:9, v/v) at 540 nm for 10 min [46]. The
antiradical activity of the compound tested is expressed as the
concentration required to reduce the absorbance by 50% (IC50).
Supplementary data
Dose response curves for cytotoxicities and antiviral activities of
the tested compounds, supporting analytical data (NMR-spectra
and HPLC) of the compounds 2a-g and 3a-g, and structural char-
acterization of the compounds 2e-g and 3e-g are available as
supplementary data. Supplementary data associated with this
article can be found in the online version at doi:10.1016/j.ejmech.
2009.11.056.
4.4.2. Inhibition of microsomal lipid peroxidation
Microsomes were isolated from rat liver of the strain Wistar as
described previously [47] and resuspended in 50 mM Tris-HCl
buffer with 100 mM KCl and 0.1 mM EDTA (pH 7.4). The protein
concentration in the microsomal suspension was determined using
the Bradford method. This suspension (400
mL) was then mixed with the compounds tested (50
in DMSO) and incubated in the presence of tert-butylhydroperoxide
(50
L, 10 mM in DMSO) at 37 ꢀC for 1 h. The products of lipid
m
L, 0.625 mg protein/
m
L, 0–100
mM
References
m
[1] P. Morazzoni, E. Bombardelli, Fitoterapia 66 (1995) 3–42.
[2] K. Flora, M. Hahn, H. Rosen, K. Benner, Am. J. Gastroenterol. 93 (1998)
139–143.
[3] D. Gallo, S. Giacomelli, C. Ferlini, G. Raspaglio, P. Apollonio, S. Prisley,
A. Riva, P. Morazzoni, E. Bombardelli, G. Scambia, Eur. J. Cancer 39 (2003)
2403–2410.
peroxidation were determined via a standard reaction with thio-
barbituric acid: addition of ice-cold mixture (0.7 mL) of thio-
barbituric acid (26 mM) and trichloroacetic acid (918 mM), heating
(90 ꢀC for 15 min), cooling, separation by centrifugation (10 min,
10 000 rpm, 4 ꢀC) and absorbance measurement at 535 nm. The
activity was calculated as the concentration of the tested
compound that inhibited the color reaction with thiobarbiturate
(without the tested compound) by 50 % (IC50).
[4] A. Pelter, R. Ha¨nsel, Tetrahedron Lett. 25 (1968) 2911–2916.
[5] H. Wagner, L. Ho¨rhammer, R. Mu¨ nster, Arzneim. Forsch. 18 (1968) 688–695.
ˇ
ˇ
´
[6] V. Kren, J. Kubisch, P. Sedmera, P. Halada, V. Prikrylova, A. Jegorov, L. Cvak,
ˇ
R. Gebhardt, J. Ulrichova´, V. Sima´nek, J. Chem. Soc. Perkin Trans. 1 (1997)
2467–2474.
[7] D.Y.-W. Lee, Y. Liu, J. Nat. Prod. 66 (2003) 1171–1174.
ꢀ
ꢀ
ˇ
[8] R. Gaza´k, D. Walterova´, V. Kren, Curr. Med. Chem. 14 (2007) 315–338.
4.5. Cytotoxicity
´
´
´
ˇ
´
´
[9] R. Gazak, A. Svobodova, J. Psotova, P. Sedmera, V. Prikrylova, D. Walterova,
ˇ
V. Kren, Bioorg. Med. Chem. 12 (2004) 5677–5687.
[10] P. Trouillas, P. Marsal, A. Svobodova´, J. Vosta´lova´, R. Gaza´k, J. Hrba´c, P. Sedmera,
ꢀ
ˇ
4.5.1. Evaluation of cytotoxicities of silybin derivatives on Madin-
Darby Canine Kidney (MDCK) cells
ˇ
´
V. Kren, R. Lazzaroni, J.-L. Duroux, D. Walterova, J. Phys. Chem. A 112 (2008)
1054–1063.
MDCK cells were seeded into 96-well culture plate at
1.5 ꢂ 104 cells/well in Dulbecco’s Modified Eagle Medium (DMEM)
with 10% fetal calf serum and left for 4–5 h. Afterwards the cells
were washed twice with Dulbecco’s Phosphate-Buffered Salines (D-
PBS); then the Opti-MEM I reduced serum medium (Gibco BRL)
ꢀ
ꢁ
´
´
´
[11] R. Gazak, P. Sedmera, M. Vrbacky, J. Vostalova, Z. Drahota, P. Marhol,
ˇ
D. Walterova´, V. Kren, Free Radic. Biol. Med. 46 (2009) 745–758.
[12] R.J. Williams, J.P.E. Spencer, C. Rice-Evans, Free Radic. Biol. Med. 36 (2004)
838–849.
[13] M.C. Comelli, U. Mengs, C. Schneider, M. Prosdocimi, Integr. Cancer Ther. 6
(2007) 120–129.

ꢀ
´
R. Gazak et al. / European Journal of Medicinal Chemistry 45 (2010) 1059–1067
1067
[14] G. Deep, R. Agarwal, Integr. Cancer Ther. 6 (2007) 130–145.
[15] R.P. Singh, R. Agarwal, Mol. Carcinogen. 45 (2006) 436–442.
[16] R. Braatz, K. Gurler, G. Bergish, G. Halbach, H. Soicke, K. Schmidt, Czech. Patent
273 610, 1985; Chem. Abstr. 105, (1985) P127476b.
[17] G. Pifferi, R. Pace, M. Conti, Il Farmaco 49 (1994) 75–76.
[18] M. Conti, S. Malandrino, M.J. Magistretti, Jpn. J. Pharmacol. 60 (1992) 315–321.
[19] G. Buzzelli, S. Moscarella, A. Gusti, A. Duchini, C. Marena, M. Lampertico, Int. J.
Clin. Pharmacol. Ther. Toxicol. 31 (1993) 456–460.
[20] P. Morazzoni, M.J. Magistretti, C. Giachetti, G. Zanolo, Eur. J. Drug Metab.
Pharmokinet. 17 (1992) 39–44.
[21] N. Barzaghi, F. Crema, G. Gatti, G. Pifferi, E. Perucca, Eur. J. Drug Metab.
Pharmokinet. 15 (1990) 333–338.
[31] H. Kikuzaki, M. Hisamoto, K. Hirose, K. Akiyama, H. Taniguchi, J. Agric. Food
Chem. 50 (2002) 2161–2168.
[32] Z. Lu, G. Nie, P.S. Belton, H. Tang, B. Zhao, Neurochem. Int. 48 (2006) 263–274.
[33] Ch.S. Hunneche, M.N. Lund, L.H. Skibsted, J. Nielsen, J. Agric. Food Chem. 56
(2008) 9258–9268.
[34] K. Matsubara, A. Saito, A. Tanaka, N. Nakajima, R. Akagi, M. Mori, Y. Mizushina,
Life Sci. 80 (2007) 1578–1585.
[35] J.P. Liu, E. Manheimer, K. Tsutani, C. Gluud, Cochrane Database Syst. Rev. 4
(2001) CD003183.
[36] J.P. Liu, E. Manheimer, K. Tsutani, C. Gluud, Am. J. Gastroenterol. 98 (2003)
538–544.
[37] S.J. Polyak, C. Morishima, M.C. Shuhart, C.C. Wang, Y. Liu, D.Y. Lee, Gastroen-
terology 132 (2007) 1925–1936.
[38] E. Peterhans, Biochem. Biophys. Res. Commun. 91 (1979) 383–392.
[39] E. Theodosiou, M.H. Katsoura, H. Loutrari, K. Purchartova´, V. Krˇen, F.N. Kolisis,
H. Stamatis, Biocat. Biotrans. 27 (2009) 161–169.
[22] R. Carini, A. Comoglio, E. Albano, G. Poli, Biochem. Pharmacol. 43 (1992) 2111–
2115.
[23] A. Comoglio, A. Tomasi, S. Malandrino, G. Poli, E. Albano, Biochem. Pharmacol.
50 (1995) 1313–1316.
[24] H. Maheshwari, R. Agarwal, C. Patil, O.P. Katare, Arzneim. Forsch. 53 (2003)
420–427.
[40] N. Gotoh, N. Noguchi, J. Tsuchiya, K. Morita, H. Sakai, H. Shimasaki, E. Niki, Free
Radic. Res. 24 (1996) 123–134.
[25] A. Valenzuela, A. Garrido, Biol. Res. 27 (1994) 105–112.
[26] P. Letteron, G. Labbe, C. Degott, A. Berson, B. Fromenty, M. Delaforge, D. Larrey,
D. Pessayre, Biochem. Pharmacol. 39 (1990) 2027–2034.
[27] T. Parasassi, A. Martellucci, F. Conti, B. Messina, Cell Biochem. Funct. 2 (1984)
85–88.
[41] E. Niki, N. Noguchi, Acc. Chem. Res. 37 (2004) 45–51.
[42] V.W. Bowry, K.U. Ingold, Acc. Chem. Res. 32 (1999) 27–34.
ꢀ
ˇ
[43] R. Gaza´k, P. Sedmera, M. Marzorati, S. Riva, V. Kren, J. Mol. Catal. B, Enzym. 50
(2008) 87–92.
[44] F.A. Silva, F. Borges, C. Guimares, J.L. Lima, C. Matos, S. Reis, J. Agric. Food Chem.
48 (2000) 2122–2126.
´
[28] O. Weso1owska, B. qania-Pietrzak, M. Kuzdza1, K. Stanczak, D. Mosia˛dz,
P. Dobryszycki, A. Ozyhar, M. Komorowska, A.B. Hendrich, K. Michalak, Acta
Pharmacol. Sin. 28 (2007) 296–306.
[45] S. Mori, S. Miyake, T. Kobe, T. Nakaya, S.D. Fuller, N. Kato, K. Kaihatsu, Bioorg.
Med. Chem. Lett. 18 (2008) 4249–4252.
[29] T. Nakayama, K. Ono, K. Hashimoto, Biosci. Biotechnol. Biochem. 62 (1998)
1005–1007.
[30] K. Abdulmajed, C. McGuigan, C.M. Heard, Free Radic. Res. 39 (2005) 491–498.
[46] C. Deby, G. Margotteaux, C.R. Seances Soc. Biol. Fil. 164 (1970) 2675–2681.
[47] H. Haraguchi, T. Saito, N. Okamura, A. Yagi, Planta Med. 61 (1995)
333–336.