Design, synthesis and evaluation of the inhibitory selectivity of novel trans-resveratrol analogues on human recombinant CYP1A1, CYP1A2 and CYP1B1

Article abstract of DOI:10.1016/j.bmc.2012.07.012

A series of trans-stilbene derivatives containing 4′-methylthio substituent were synthesized and evaluated for inhibitory activities on human recombinant cytochrome P450(s): CYP1A1, CYP1A2, and CYP1B1. CYP1A2-related metabolism of stilbene derivatives was estimated by using NADPH oxidation assay. Additionally, for CYP1A2 and CYP1B1 molecular docking analysis was carried out to provide information on enzyme-ligand interactions and putative site of metabolism. 3,4,5-Trimethoxy-4′-methylthio-trans-stilbene, an analogue of DMU-212 (3,4,5,4′-tetramethoxy-trans-stilbene) was an effective inhibitor of all CYP1 enzymes. On the other hand, 2,3,4-trimethoxy-4′-methylthio- trans-stilbene, appeared to be the most selective inhibitor of the isozymes CYP1A1 and CYP1B1, displaying extremely low affinity towards CYP1A2. Molecular modeling suggested that the most probable binding poses of the methylthiostilbene derivatives in CYP1A2 active sites are those with the methylthio substituent directed towards the heme iron. Products of CYP1A2-catalyzed oxidation of 2,4,5-trimethoxy-4′-methylthiostilbene and 3,4,5-trimethoxy-4′-methylthiostilbene were identified as monohydroxylated compounds. Other studied derivatives appeared to be poor substrates of CYP1A2. Structure-activity relationship analysis rendered better understanding of the mechanism of action of xenobiotic-metabolizing enzymes crucial at the early stage of carcinogenesis.

Full text of DOI:10.1016/j.bmc.2012.07.012

Bioorganic & Medicinal Chemistry 20 (2012) 5117–5126
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Design, synthesis and evaluation of the inhibitory selectivity
of novel trans-resveratrol analogues on human recombinant
CYP1A1, CYP1A2 and CYP1B1
Renata Mikstacka a, , Agnes M. Rimando , Zbigniew Dutkiewicz , Tomasz Stefanski , Stanisław Sobiak
⇑
b
a
a
a
´
a Department of Chemical Technology of Drugs, Poznan University of Medical Sciences, Grunwaldzka 6, 60-780 Poznan, Poland
b United States Department of Agriculture, Agricultural Research Service, Natural Products Utilization Research Unit, PO Box 8048, University, MS 38677, USA
a r t i c l e i n f o
a b s t r a c t
A series of trans-stilbene derivatives containing 40-methylthio substituent were synthesized and evalu-
ated for inhibitory activities on human recombinant cytochrome P450(s): CYP1A1, CYP1A2, and CYP1B1.
CYP1A2-related metabolism of stilbene derivatives was estimated by using NADPH oxidation assay. Addi-
tionally, for CYP1A2 and CYP1B1 molecular docking analysis was carried out to provide information on
enzyme–ligand interactions and putative site of metabolism.
Article history:
Received 30 March 2012
Revised 2 July 2012
Accepted 6 July 2012
Available online 17 July 2012
3,4,5-Trimethoxy-40-methylthio-trans-stilbene, an analogue of DMU-212 (3,4,5,40-tetramethoxy-trans-
stilbene) was an effective inhibitor of all CYP1 enzymes. On the other hand, 2,3,4-trimethoxy-40-methyl-
thio-trans-stilbene, appeared to be the most selective inhibitor of the isozymes CYP1A1 and CYP1B1, dis-
playing extremely low affinity towards CYP1A2. Molecular modeling suggested that the most probable
binding poses of the methylthiostilbene derivatives in CYP1A2 active sites are those with the methylthio
substituent directed towards the heme iron. Products of CYP1A2-catalyzed oxidation of 2,4,5-trimeth-
oxy-40-methylthiostilbene and 3,4,5-trimethoxy-40-methylthiostilbene were identified as monohydroxy-
lated compounds. Other studied derivatives appeared to be poor substrates of CYP1A2. Structure–activity
relationship analysis rendered better understanding of the mechanism of action of xenobiotic-metaboliz-
ing enzymes crucial at the early stage of carcinogenesis.
Keywords:
P450
CYP1A1
CYP1A2
CYP1B1
trans-Stilbene derivatives
Molecular docking
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
strate better bioavailability than the parent compound that under-
goes fast biotransformation to glucuronides and sulfates.11
Chemopreventive, cardioprotective and neuroprotective activi-
ties of trans-resveratrol (3,40,5-trihydroxystilbene), the best known
natural stilbene derivative, has been documented in numerous
studies on animal models.1 Its therapeutic action in cancer diseases
is also under intensive preclinical and clinical studies.2–4 In the last
decade, other natural compounds with stilbene backbone were
shown to possess promising activity concerning cancer preven-
tion.2,5 New stilbene derivatives are designed and synthesized in
order to find resveratrol analogues with cancer chemopreventive
and/or therapeutic activities superior to that of the parent com-
pound. The compounds are examined with respect to different tar-
gets such as estrogen receptors,6 COX 1/2,7 signal transduction
pathways,8,9 cell proliferation,10 and inhibitory activity towards
cytochromes P450 (CYPs). The studies are focused on structure–
activity relationship in order to identify structural determinants
responsible for beneficial activities of stilbene derivatives. Other
investigations are aimed to find resveratrol analogues that demon-
Methoxylation of hydroxyl groups is supposed to prevent polyphe-
nol metabolism and enhance stilbene bioactivity.12,13
Cytochrome P450 family 1 includes xenobiotic-metabolizing en-
zymes: CYP1A1, CYP1A2 and CYP1B1 that catalyze the activation of
environmental procarcinogens such as polycyclic aromatic hydro-
carbons, aromatic and heterocyclic amines to carcinogenic forms.
Additionally, CYP1B1 metabolizes 17b-estradiol to 4-hydroxyestra-
diol that is oxidized by peroxidase to estradiol-3,4-quinone, and
forms a quinone-DNA adducts responsible for estrogen-related car-
cinogenesis.14 For these reasons, there is a demand for inhibitors
that demonstrate high selectivity toward specific forms of cyto-
chrome 450; namely, CYP1A1 and CYP1B1 for blocking CYP-related
carcinogenesis. CYP1A2 metabolizes numerous xenobiotics includ-
ing drugs such as cafeine, theophylline, methadone, verapamil, pro-
pranolol, warfarin, and tamoxifen. Not much is known about
CYP1A2; hence its properties, the structure of its active site and
substrate specificity need thorough investigation. All members of
CYP1A family are expressed in extrahepatic tissues; however,
CYP1A2 is the only constitutive form of liver enzyme. In humans
CYP1B1 is overexpressed in tumor cells, and this has important
⇑
Corresponding author. Tel./fax: +48 618 546 625.
E-mail address: rmikstac@ump.edu.pl (R. Mikstacka).
0968-0896/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2012.07.012

5118
R. Mikstacka et al. / Bioorg. Med. Chem. 20 (2012) 5117–5126
implications for tumor development and progression. Thus, regula-
tors of the expression and catalytic activity of Family 1 cytochromes
are suggested to play an important role in cancer chemoprevention
mainly at the early stage of tumorigenesis; specifically, develop-
ment of anticancer prodrugs specifically activated by CYP1B1 is a
promising strategy in cancer chemotherapy.15 Moreover, it was dis-
covered recently that CYP1B1 generates reactive oxygen species
(ROS) leading to hypertension,16 therefore, CYP1B1 inhibitors might
be of therapeutic importance in cardiovascular diseases.
The diverse activities of resveratrol include moderate inhibition
of CYP1A1 and CYP1B1, and rather weak inactivation of CYP1A2.17
Natural resveratrol methyl ethers appeared to be more specific and
potent inhibitors of cytochromes P450 (CYP) Family 1 in compari-
son to the parent compound.18–20 Many synthetic methoxy stil-
bene derivatives were tested for inhibitory activity toward CYP1
isozymes. 3,5,20,40-Tetramethoxystilbene and 2,6,20,40-tetrameth-
oxystilbene appeared to be potent and specific inhibitors of
CYP1B1;21,22 while natural resveratrol analogue, rhapontigenin
(3,5,30-trihydroxy-40-methoxystilbene) was found to inhibit selec-
tively CYP1A1.23 Results from our previous study revealed that
some of stilbene derivatives with methylthio substituent are selec-
tive and potent inhibitors of CYP1 enzymes.24 Among the series of
compounds investigated earlier, 2-methoxy-40-methylthiostilbene
and 3-methoxy-40-methylthiostilbene were the most potent com-
petitive inhibitors of CYP1A1 and CYP1B1 activities.24 Methylthio-
stilbene derivatives merit attention in view of the fact that
substitution of the 40-oxygen atom with the less electronegative
sulfur atom reduces toxicity to HEK 293 cells, and as was observed
by Yang et al.25 enhances the ability of the compound to activate
human Sirtuin-1.
information on mechanism of enzyme inhibition at the molecular
level; make analysis of binding orientations of inhibitors to en-
zyme active sites possible and allow prediction of putative sites
of metabolism.26 These procedures are helpful in structure–activity
relationship analysis and as such, are employed for verification of
hypothesis based on experimental results. Moreover, in silico tools
are useful in the design of more efficient and selective synthetic
inhibitors of enzymes. Models of cytochrome P450 enzymes as
key enzymes metabolizing exogenous and endogenous substrates
have been created and well characterized.27 Molecular docking of
different CYP1 substrates has been reported, for example, rutaecar-
pine derivatives,28 7-alkoxylresorufins,29 coumarins,30 and meth-
oxyflavonoids;31 however, computational analysis of stilbenes as
substrates of CYP1 enzymes is still lacking.
In our present study, a series of 40-methylthiostilbenes differing
in methoxy groups patterns in ring A (Table 1) were synthesized
and examined for their inhibitory effects on activities of human re-
combinant CYPs: CYP1A1, CYP1A2 and CYP1B1. In order to eluci-
date the inhibition selectivity of some 40-methylthiostilbenes,
which exerted potent and selective inhibition of CYP1B1, molecu-
lar docking approach was used. We found two 40-methylthiostil-
bene derivatives as substrates of CYP1A2 and identified the
products of their metabolism. Additionally, on the basis of CYP1A2
computational docking, the site of metabolism was predicted and
the results of virtual screening were compared with experimental
data. The specific interactions that govern orientation of an inhib-
itor/substrate in the enzyme active site are discussed.
2. Results and discussion
2.1. Chemistry
In this study we extended our studies of 40-methylthiostilbene
derivatives and synthesized seven new compounds differing in
the substitution pattern and performed computational docking of
synthesized compounds in CYP1A2 and CYP1B1 active sites. Com-
putational modeling and docking procedures provide additional
The synthesis of compounds is outlined in Scheme 1. The key
synthetic step for the construction of these compounds involves
the generation of phosphonate ester
2
(diethyl-4-meth-
Table 1
Summary of IC50 values for EROD inhibition by a series of 40-methythiolstilbene derivatives
R
3
2
1
4
B
vin'
5
6'
5'
4'
1'
2'
vin
6
A
H3C
3'
S
IC50
(
lM)
Ratio
NADPH oxidizedb
CYP1A1
CYP1A2
CYP1B1
CYP1A2/CYP1A1
CYP1A2/CYP1B1
Monomethoxy-40-methylthiostilbenes
2-M-40-MTSa
3-M-40-MTSa
4-M-40-MTSa
1.0
1.1
28.0
7.5
11.5
23.0
0.3
0.5
5.5
7.5
10.5
0.8
25.0
23.0
4.2
<0.3
<0.3
<0.3
Dimethoxy-40-methylthiostilbenes
2,3-DiM-40-MTSa
2,4-DiM-40-MTS (S1)
2,5-DiM-40-MTS (S2)
3,5-DiM-40-MTSa
1.8
2.4
3.8
1.4
22.0
8.1
39.5
5.0
1.1
2.0
1.1
1.5
12.2
3.4
10.4
3.6
20.0
4.1
35.9
3.3
<0.3
<0.3
<0.3
<0.3
Trimethoxy-40-methylthiostilbenes
2,3,4-TriM-40-MTS (S3)
2,4,5-TriM-40-MTS (S4)
2,4,6-TriM-40-MTS (S5)
3,4,5-TriM-40-MTS (S6)
0.9
0.8
0.4
3.6
>50
15.0
6.5
1.0
0.9
0.5
2.6
>55.5
18.7
16.3
1.9
>50
16.7
17.0
2.7
<0.6
2.91
<0.3
2.22
7.0
2-Chloro-40-MTS (S7)
1.0
14.5
0.3
14.5
48.33
Not determined
Each IC50 value was calculated by plotting the inhibition of EROD activity against the inhibitor concentration. The determination was performed in three parallel experiments.
The coefficient of variation did not exceed 15%.
a
Compounds studied earlier (Mikstacka et al., 2008). M = methoxy; MTS = methylthiosilbene.
Values are in M/pmol CYP1A2/min. Stilbenes were dissolved in DMSO; the amount of DMSO did not exceed 0.5%. The concentration of stilbenes in reaction mixture was
b
10 lM. The level of NADPH oxidation when only DMSO as a vehicle was added was equal 0.3 lM/pmol CYP1A2/min.

R. Mikstacka et al. / Bioorg. Med. Chem. 20 (2012) 5117–5126
5119
O
Cl
P(OEt)2
OH
(a)
(b)
S1: R1 = R3 = OCH3 ; R2 = R4 = R5 = H
S2: R1 = R4 = OCH3 ; R2 = R3 = R5 = H
S3: R1 = R2 = R3 = OCH3 ; R4 = R5 = H
S4: R1 = R3 = R4 = OCH3 ; R2 = R5 = H
S5: R1 = R3 = R5 = OCH3 ; R2 = R4 = H
S6: R2 = R3 = R4 = OCH3 ; R1 = R5 = H
S7: R1 = Cl ; R2 = R3 = R4 = R5 = H
SCH3
SCH3
SCH3
1
2
R2
R1
R3
R4
(c)
R5
H3CS
S1-S7
Scheme 1. General synthetic route for the synthesis of the trans-methylthiostilbene analogues (S1–S7). Reagents and conditions: (a) SOCl2, toluene, 0.5 h, room temp, 84%;
(b) P(OEt)3, 130 °C, 2 h, 80%; (c) (R1–R5)PhCHO, NaH, DMF, 0 °C to room temp, 2 h, 48–75%.
ylthiobenzylphosphonate) as an intermediate. This was prepared
from commercially available 4-methylthiobenzyl alcohol in two
steps. First, 4-methylthiobenzyl alcohol was converted to the chlo-
ride 1 using SOCl2 in toluene at room temperature. Then, through
the Michaelis–Arbuzov reaction of 1 with triethylphosphite at
130 °C the corresponding phosphonate ester 2 was obtained.
trans-methylthiostilbene analogues were prepared by Wittig–
Horner reaction of 2 with the commercially available benzalde-
hydes in DMF using sodium hydride as a base. Under these reaction
conditions trans-isomers were obtained exclusively. Geometries of
these compounds were confirmed by their characteristic 1H NMR
coupling constants for the olefinic protons of about 16–16.5 Hz.
Intermediates were characterized by NMR spectroscopy. Final
products were characterized by NMR, and mass spectrum, which
were in full accordance with depicted structures. One of the stud-
ied compounds, S6 was reported earlier by Cushman et al.8
tion to each other seem to have better affinity to the CYP1A2 active
site (thus lower IC50 values for S1 and 3,5-dimethoxy-40-MTS, see
Table 1); however, to confirm that observation a bigger series of
compounds should be tested.
All stilbene derivatives used in this study showed moderate to
weak inhibitory activity towards CYP1A2 in comparison with
CYP1A1 and CYP1B1. However, basing on the experimental inhibi-
tion data it is noticeable that CYP1A2 is the most susceptible to
variations in the pattern of methoxy substituents. The 40-methyl-
thiostilbenes may be divided into two distinct groups: relatively
good (IC50 <10 lM) and very weak (IC50 >10 lM) inhibitors. Inter-
estingly, the motif 3,4,5-trimethoxy did not diminish the affinity
of this trimethoxy derivative to CYP1A2, while it influenced signif-
icantly the inhibitory potency towards CYP1A1 and CYP1B1. It may
be concluded that analogues where the substituents in ring B are
arranged in such a way where half of ring B (when an imaginary
line is drawn from C1 to C4) mirrors the other half of the ring pro-
vide relatively strong inhibition of CYP1A2.
2.2. Inhibitory effect on CYP1A1, CYP1A2 and CYP1B1 activities
S3, S4, and S5 appeared to be the most potent inhibitors of
CYP1A1 (IC50 <1 lM). Comparing the potency of test compounds
towards different CYP1 isoforms it may be concluded that IC50 val-
ues determined for CYP1A1 and CYP1B1 correlate with each other
with few exceptions of monomethoxy-40-methylthiostilbenes and
S2.
Theinhibitoryactivitiesofsynthesized40-methylthiostilbene(40-
MTS) analogues 2,4-dimethoxy-40-MTS (S1), 2,5-dimethoxy-40-MTS
(S2), 2,3,4-trimethoxy-40-MTS (S3), 2,4,5-trimethoxy-40-MTS (S4),
2,4,6-trimethoxy-40-MTS (S5) 3,4,5-trimethoxy-40-MTS (S6), and
2-chloro-40-MTS (S7) on human recombinant cytochrome P450(s)
CYP1A1, CYP1A2 and CYP1B1 were evaluated and compared with
those of 40-methylthiostilbenes previously investigated.24 IC50 val-
ues for the series of polymethoxy 40-methylthiostilbenes are sum-
marized in Table 1. In the CYP1B1 inhibition assay, S7 and two 40-
MTS analogues studied earlier, that is, 2-methoxy-40-MTS and 3-
methoxy-40-MTS, were the potent inhibitors with IC50 values of
2.3. Metabolism studies
To examine whether the stilbene derivatives were metabolized
in CYP1A2-catalyzed reaction, NADPH oxidation in the presence of
CYP1A2 was measured spectrophotometrically. Significant
CYP1A2-related metabolism was observed only for S6 and S4 (Ta-
ble 1). The effect of NADPH oxidation was observed at lower stil-
0.3, 0.5 and 0.3 lM, respectively. Interestingly, substitution of 2-
methoxy group with chlorine atom did not change the inhibitory
activity of 40-MTS derivative toward CYP1A1 and CYP1B1. However,
the substituent at the position 2 seems to play a crucial role in
CYP1B1 inhibition, which is in agreement with earlier reports.21,22
All the stilbenes studied appeared to be strong or moderate
bene concentrations (610 lM); at higher concentrations NADPH
was oxidized less effectively (data not shown). It appeared that
at higher concentrations inhibitory effect of S6 and S4 on CYP1A2
activity might be prevailing. However these speculations need to
be elucidated by further experimental explorations.
inhibitors of CYP1B1 with IC50 <2.6 l
M. 2,3,4-Trimethoxy-40-MTS
(S3) was found to be the most selective inhibitor of the enzymes
CYP1A1 and CYP1B1, and exerted very low affinity to CYP1A2.
Experimental data suggested that there was not a single structural
determinant for selective inhibition of the 40-methylthiostilbenes;
rather, the pattern of substitution on ring A of the stilbene scaffold
determined the inhibitory properties. For example, among the
dimethoxy derivatives, those having the substituents in meta posi-
The products from CYP1A2-catalyzed metabolism of S6 and S4
were analyzed by GC–MS, which revealed monohydroxylated
metabolites. The total ion chromatograms (TIC) of pure samples
of S6 and S4 showed single peaks (retention times 11.58 min and
11.92; Figs. 1A and 2A, respectively) and molecular ion peaks of
316 amu (Figs. 1B and 2B, respectively). Upon reaction with
CYP1A2, the S6 reaction mixture showed a peak (Fig. 1C1 and

5120
R. Mikstacka et al. / Bioorg. Med. Chem. 20 (2012) 5117–5126
A
70.0x106
60.0x106
50.0x106
40.0x106
30.0x106
20.0x106
10.0x106
A
11.58
B
1x106
315.8
316.9
301.0
168.1
1x106
211.0
226.1
139.1
93.5
800x103
158.1
600x103
400x103
200x103
317.9
195.0
181.1
128.1
107.5
241.0
78.1
286.0
270.2
258.1
6
8
10
12
14
16
18
20
22
24
45x106
40x106
35x106
30x106
25x106
20x106
15x106
10x106
5x106
0
80
120
160
200
240
m/z
280
320
360
C1
C2
m/z
11.8
D
350x103
300x103
250x103
200x103
150x103
100x103
50x103
A
316.9
175x103
150x103
125x103
100x103
75x103
50x103
25x103
0
70.6
11.80
97.1
83.7
301.1
332.1
181.3
168.0
152.0
111.1
125.2
140.1
211.1
195.0
253.0
226.1
241.1
285.0
269.0
80
120
160
200
240
m/z
280
320
360
6
8
10
12
14
16
18
20
22
24
Time (min)
Figure 1. GC–MS analysis of 3,4,5-trimethoxy-40-thiomethyl-trans-stilbene (S6). (A) Total ion chromatogram (TIC) of pure S6, retention time = 11.58 min. (B) Mass spectrum
of S6 showing molecular ion m/z 316 and characteristic fragmentation. (C1) TIC of the ethyl acetate extract of the CYP1A2 and S6 reaction mixture, showing a single peak
when mass-defined at m/z 332 (C2) indicative of addition of one hydroxyl group. (D) Mass spectrum of peak in panel C2 at 11.80 min showing molecular ion at m/z 332 and
characteristic fragmentations.
C2) with mass 332 amu (Fig. 1D). Similarly, reaction of CYP1A2
with S4 showed a peak (Fig. 2C1 and C2) with mass 332 amu
(Fig. 2D). Mass spectral data indicated that hydroxylation occurred
on ring B evident from the presence of fragmentation peaks m/z
140 and 152, as well as intact ring A fragment peaks at m/z 168,
181, and 195 for S6 (Fig. 1D, see also inset). Similarly, for S4
hydroxylation on ring B was evident from the presence of fragmen-
tation peaks m/z 139, 152 and 165, as well as intact ring A frag-
ment peaks at m/z 181 and 195 (Fig. 2D, see also inset).
Considering that 3,5,30,40-tetrahydroxystilbene (piceatannol) is a
product of resveratrol 30-hydroxylation catalyzed by CYP1A2,32 it
is most probable is that the hydroxy group was introduced at the
position 30 of B ring. In our experiments, the structure of identified
products confirms the orientation of stilbenes in the CYP1A2 active
site cavity predicted by means of molecular docking. Moreover,
these metabolites might be inhibitors of CYP1A2 activity. We can
only speculate from the molecular modeling, that it might be diffi-
cult for such bulky molecules to leave the enzyme cavity through
an entrance canal or that the binding of the hydroxy metabolites
to the CYP1A2 active site is irreversible.
conducted in order to elucidate the selectivity of CYP1 inhibition
by stilbenes differing in substitution pattern in the A ring. We ex-
plored the possible binding modes for the compounds that differed
in inhibitory potency toward CYP1A2 and CYP1B1. Basing on the
evalution of molecular docking programs35,36 we chose LigandFit
as the appropriate and we validated its performance with 7,8-ben-
zoflavone and 7-ethoxyresorufin as ligands. With the use of
LigandFit docking procedure, we analyzed orientations of the test
molecules in the enzyme active sites, to choose those that are
the most probable.
Among all poses of the studied compounds in the active site
cavity of CYP1A2 the most frequent occurring were the poses with
B ring localized close to the heme. Docking procedure found some
exceptions when A ring was directed towards the heme: one of two
poses found for 2-methoxy-40-MTS and two poses found for S7. Li-
gands in the enzyme active site were stabilized by a single
p–p
stacking interaction between Phe226 and ring B and A in ‘reversed’
poses (Fig. 3A). Some exceptions were observed for 3-methoxy-40-
MTS when two
ring and for S2 when two
curred with A and B ring and additionally a
tion of Phe260 with ring was found. However, these
p–p
stacking interactions of Phe226 with A and B
stacking interactions of Phe226 oc-
stacking interac-
p–p
p–p
2.4. Molecular docking
A
exceptional orientations seem to be less effective because of in-
creased distances of ligand to the heme. More importantly, for
S2, S4, S5, and S6, but not S3, docking procedure has found poses
Molecular docking for the series of 40-MTS derivatives included
in this study into the active sites of CYP1A233 and CYP1B134 was

R. Mikstacka et al. / Bioorg. Med. Chem. 20 (2012) 5117–5126
5121
A
70x106
60x106
50x106
40x106
30x106
20x106
10x106
A
B
11.92
1.4e+6
222.1
226.1
1.2e+6
152.1
316.0
301.0
115.1
211.1
270.1
254.1
239.1
137.0
1.0e+6
8.0e+5
6.0e+5
4.0e+5
2.0e+5
128.1
165.1
168.1
195.1
181.1
285.0
0
5x106
C1
C2
80
120
160
200
240
280
320
360
4x106
3x106
2x106
1x106
8.57
m/z
90x103
80x103
70x103
60x103
50x103
40x103
30x103
20x103
10x103
D
A
316.6
80000
70000
60000
50000
40000
30000
20000
10000
0
8.57
332.6
158.4
273.6
301.6
152.4
195.5
165.4
211.5
139.4
253.5
181.5
6
8
10
12
14
16
18
20
22
24
80
120
160
200
240
280
320
360
Time (min)
m/z
Figure 2. GC–MS analysis of 2,4,5-trimethoxy-40-thiomethyl-trans-stilbene (S4). (A) Total ion chromatogram (TIC) of pure S4, retention time = 11.92 min. (B) Mass spectrum
of S4 showing molecular ion m/z 316 and characteristic fragmentation. (C1) TIC of the ethyl acetate extract of the CYP1A2 and S4 reaction mixture, showing a single peak
when mass-defined at m/z 332 (C2) indicative of addition of one hydroxyl group. (D) Mass spectrum of peak in panel C2 at 8.58 min showing molecular ion at m/z 332 and
characteristic fragmentations.
stabilized by two
and Phe260. Docking of S3 in CYP1A2 cavity resulted in three dis-
tinct poses; however in one of them, the lowest in DockScore rank-
ing, a p–p stacking interaction was not found. Moreover, the best
pose of S3 was classified at the lowest position among the best
poses of the studied series of compounds, while its strain energy
achieved the highest value (103.09 kcal/mol; for more data see
Supplementary data); these data support weak inhibitory activity
demonstrated experimentally. The ideal poses for metabolized S4
p
–
p
stacking interactions of ring A with Phe226
Identified by GC–MS, oxidized products of S6 and S4 provided
additional evidence that the most effective poses are those when
the ring B is directed towards the heme (Fig. 4). The stilbenes that
are not metabolized by CYP1A2 probably inhibit O-deethylation of
7-ethoxyresorufin by binding to a site that only partially involves
the catalytic site. Binding to the regulatory sites that causes a
change in the cytochrome structure and shape may be taken into
account as a mechanism of enzyme inhibition. The other explana-
tion for the lack of metabolism observed for several derivatives is
that the molecules did not sufficiently come close to the heme.
and S6 possessed two p–p interaction within Phe260 and Phe226
and additionally, for S6 hydrophobic interaction with Gly316 for
ring A has occurred (Fig. 4). The orientations let the molecules be
close to the heme and promote the oxidation of ring B in the vicin-
For effective enzymatic reaction, the distance of a substrate to
0
the heme iron should not exceed 6 ÅA.37 However, specific interac-
tions between an inhibitor and amino acid residues in the enzyme
cavity govern orientation of a molecule in the active site. For exam-
ple, methoxy substituents on the A ring may influence better fit of
the molecule to the binding site cavity by hydrophobic interactions
or hydrogen bonding. Schematic representation of ligand interac-
tions with amino acids residues visualizes the predominant role
of hydrophobic interactions (Fig. 3A). Moreover, these interactions
may produce an effect of shortening of the distance between B ring
and the iron. It is worth mentioning that for the metabolized deriv-
atives S4 and S6 docked in CYP1A2 binding site, the distances be-
ity of 40-methylthio group. The distances between Fe atom and C
0
atoms in 30 and 50 positions for S4 were 5.44 and 4.460 ÅA, respec-
tively. Corresponding values for S6 were 5.37 and 4.38 ÅA (the data
concerning the other test stilbenes are in Supplementary data).
Interestingly, for S2 and S5 two
p–p interactions within Phe226,
Phe260 and the ring A did not bring the molecules closer to the
heme what was in agreement with poor oxidation of these com-
pounds; contrary, these
p–p interactions caused the significant in-
crease of distances of C atoms in 30 and 50 positions to Fe atom.

5122
R. Mikstacka et al. / Bioorg. Med. Chem. 20 (2012) 5117–5126
Figure 3. The amino acid residues surrounding 2,4,6-trimethoxy-40-methylthio-trans-stilbene (S5) in (A) CYP1A2 and (B) CYP1B1 binding sites. Residues involved in van der
Waals and polar interactions are colored in green and magenta, respectively. Solid blue lines represent interactions and dashed blue lines represent hydrogen bonds.
p–p
Figure 4. Human CYP1A2 active site in complex with (A) 2,4,5-trimethoxy-40-methylthio-trans-stilbene (S4) and (B) 3,4,5-trimethoxy-40-methylthio-trans-stilbene (S6).
Solid blue lines represent interactions. Heme is represented as a stick model in red.
p–p
tween Fe atom and C atoms in 30 and 50 positions were shorter than
5.5 and 4.5 ÅA, respectively.
and CYP1B1 and demonstrates the conformational changes of the
ligand in the enzyme cavity.
0
Analysis of the three-dimensional structures of polymethoxy-
stilbenes showed that in in S3, both 2- and 3-methyl groups are
displaced from the aromatic ring A plane and are on opposite sides
of this plane. However distortion of methyl group is not considered
a decisive conformational change, it seems to us that in case of
bulky trimethoxy derivative it may be a reason for a very low affin-
ity of this compound to CYP1A2 active site, which is described as
narrow and suitable for rather flat ligand molecules. Furthermore,
fitting of this trimethoxystilbene to CYP1A2 active site leads to the
significant increase of strain energy and seems not to be energeti-
cally favorable. Figure 5 illustrates the best poses of S3 in CYP1A2
In the active site of CYP1B1 the orientation of 40-MTS deriva-
tives was the same as in CYP1A2 with B ring directed towards
the heme with some exceptions: 3 of 4 poses found by LigandFit
for 4-methoxy-40-MTS and 1 of 4 poses found for 2,3-dimethoxy-
40-MTS were reversed, with A ring oriented towards the heme.
However, the 40-MTS derivatives exerted much more potent inhib-
itory activity towards CYP1B1 in comparison to CYP1A2. Higher
affinity of 40-MTS to CYP1B1 than CYP1A2 is well supported by
DockScore function values which are significantly higher in case
of CYP1B1 (see Supplementary Table S1). This function was deter-
mined for the most selective CYP1B1 inhibitor, S3, to be 7-fold

R. Mikstacka et al. / Bioorg. Med. Chem. 20 (2012) 5117–5126
5123
higher as compared with CYP1A2. For the best pose of this trimeth-
3. Conclusions
oxy derivative strain energy decreased from 103.09 in CYP1A2
binding site to 40.7 kcal/mol in CYP1B1 binding site.
A series of di- and tri-methoxy-40-MTS derivatives were synthe-
sized using efficient methods. The compounds were evaluated for
their inhibitory potency towards CYP1A1, CYP1A2 and CYP1B1
activities. 2,3,4-Trimethoxy-40-MTS, S3, was the most selective
inhibitor of CYP1A1 and CYP1B1, while displaying extremely low
affinity toward CYP1A2, which was explained by in silico studies.
Substitution of 2-methoxy group with chlorine atom did not
change the inhibitory activity of 40-MTS derivative toward CYP1A1
For CYP1B1 two
A and B rings of studied stilbenes were found (Fig. 3B) with some
minor exceptions when only one stacking interaction oc-
p–p stacking interactions between Phe231 and
p–p
curred. Additionally, for some of ligands docked in the CYP1B1 cav-
ity hydrogen bonds were found. For monitoring hydrogen bonds
the default value of 2.5 Å as a distance criterion was used. The abil-
ity of hydrogen bond formation by Gln332 in the CYP1B1 active
site may influence high inhibitory activity of stilbenes substituted
on position 2.21,22 An exceptional behavior of S4 is worth present-
ing. This stilbene was oriented by docking procedure with 5-meth-
oxy group pointed toward Gln332 residue and involved in
hydrogen bonding that made the molecule come closer to the
and CYP1B1. Molecular modeling studies revealed several
p–p
stacking interactions for CYP1A2 and CYP1B1 active site; addition-
ally for CYP1B1 hydrogen bonds with Gln332 were found for 2-
MTS, S2, S4, S5, and S6. The residues Phe226 and Phe260 for
CYP1A2, and Phe231 for CYP1B1 were involved in the p–p stacking
heme; as a result, a
sible to occur.
p–p stacking interaction of B ring was impos-
interactions with aromatic rings of polymethoxy-stilbene deriva-
tives. Whatever interaction found for the least active analogue S3
in CYP1A2 cavity, together with DockScore function, characterized
this compound with very low value. Conversely, in CYP1B1 active
site the orientation of this trimethoxy derivative was stabilized
Summarizing, the docking study has shown that specific amino
acid residues for each CYP1 play an important role in the interac-
tions of the 40-MTS derivatives in the enzyme pocket. The critical
residues identified as participating in
p–p stacking interactions
by
p–p stacking interaction with Phe231 and its strain energy
are Phe226 and Phe260 for CYP1A2, and Phe231 for CYP1B1. How-
ever, Phe231 in CYP1B1 active site interacts with the both aromatic
rings of the stilbenes, while in the case of CYP1A2 the most fre-
quent interaction is only with one of the rings, namely A ring. Addi-
tionally, in the case of CYP1B1 Gln332 takes part in hydrogen
bonding of methoxy substituents. Comparing the influence of a
series of 40-MTS derivatives on CYP1B1 activity the highest inhibi-
tory effect was observed for 2-methoxy- and 3-methoxy-40-MTS
analogues; these compounds inhibited CYP1B1 activity in a com-
was 2.5-fold diminished.
CYP1A2 effectively metabolized S4 and S6 to products with hy-
droxyl group substituted on B ring as identified by means of GC–
MS. Other stilbene derivatives studied were not metabolized by
CYP1A2; however, CYP1A2-catalyzed O-deethylation of 7-ethoxy-
resorufin was inhibited in their presence. In this case, the stilbenes
probably bind to a site other than the catalytic site of enzyme.
Molecular docking studies suggested the preferred orientation of
the 40-methylthiostilbene derivatives with the thiomethyl substi-
tuent on B ring directed towards the iron heme. The stilbene mol-
ecules might also be oriented with the A ring towards the heme;
however, that position was not energetically favorable. In fact,
the products hydroxylated at A ring were not found.
petitive manner with Ki values 0.04 and 0.03 l
M, respectively.24
However, all stilbenes tested were relatively potent CYP1B1 inhib-
itors, it may be speculated that in the case of other members of this
series, additional methoxy groups did not help in achieving better
affinity to CYP1B1 active site. On the other hand, polymethoxylated
A ring influences significantly the affinity of stilbene derivatives to-
wards CYP1A2 and in this way improves the selectivity of CYP1
inhibition.
Summarizing, study of the inhibitory activity of stilbene deriv-
atives enriched our knowledge of CYPs properties and mechanism
of CYP-related catalysis. Computational screening of compounds
with the use of enzyme model supported experimental data, and
Figure 5. Comparison of the best poses of 2,3,4-trimethoxy-40-methylthio-trans-stilbene (S3) in (A) CYP1A2 and (B) CYP1B1 binding cavities and (C) minimized structure of
S3. Solid blue lines represent interactions. Heme is represented as a wireframe model in red.
p–p

5124
R. Mikstacka et al. / Bioorg. Med. Chem. 20 (2012) 5117–5126
would help in the design of molecule(s) with desirable properties
for future studies.
1.27 (t, J = 7.2, 6H), 2.49 (s, 3H), 3.13 (d, J = 21.6, 3H), 4.04 (q,
J = 7.6, 4H); 7.23 (s, 4H).
4.3. General method of synthesis of compounds 40-
methylthiostilbene derivatives
4. Experimental
4.1. Chemicals
A solution of 2 (1.5 g, 5.47 mmol) in dry DMF (10 mL) was
added to a magnetically stirred solution of NaH (60% dispersion
in mineral oil, 0.22 g, 5.47 mmol) in dry DMF under nitrogen
(10 mL) at 0 °C, and the solution was stirred for 30 min. A solution
of corresponding benzaldehyde (4.56 mmol) in dry DMF (10 mL)
were added at 0 °C, and the reaction mixture was allowed to warm
to room temperature over a period of 1.5 h. The mixture was
poured slowly onto 250 mL ice-water and the precipitated solid
was filtered, washed with water, dried, and crystallized from di-
luted or pure ethanol to yield white crystals.
Supersomes, microsomes from baculovirus-infected insect cells
coexpressing NADPH-CYP reductase and CYP1A1, CYP1A2 or
CYP1B1, were purchased from GENTEST (Woburn, MA, USA). The
total CYP content was provided by the supplier. Glucose-6-phos-
phate dehydrogenase, nicotinamide adenine dinucleotide phos-
phate (NADP+), 7-ethoxyresorufin and resorufin were obtained
from Sigma (St. Louis, MO, USA). All the other chemicals and re-
agents were of the highest grade available. Reactions which in-
volved air or moisture sensitive reagents were performed in
oven-dried glassware under an argon atmosphere, unless other-
wise stated. Before use, solvents for chromatography and isolation
(ethyl acetate, hexanes) were purified by distillation. 4-meth-
ylthiobenzyl alcohol, methoxylated benzaldehydes and 2-chloro-
benzaldehyde and other reagents and anhydrous solvents used in
synthesis were generally purchased from Aldrich Chemical Com-
pany. All reaction mixtures were magnetically stirred and moni-
tored by TLC.
NMR spectra for intermediates were recorded on a Varian Gem-
ini 300 MHz model spectrometer (300 MHz for 1H and 75 MHz for
13C) and NMR spectra for final products were recorded on Bruker
Avance II 400 (400 MHz for 1H and 101 MHz for 13C) with TMS as
an internal standard in CDCl3 unless otherwise specified. Chemical
shifts are expressed in ppm (d), and peaks are listed as singlet (s),
doublet (d), triplet (t), quintet (q), multiplet (m), with coupling
constants (J) expressed in Hertz. Melting points were determined
in capillary tubes on a Stuart SMP10 micro melting point apparatus
and are uncorrected. The LRMS (EI) spectra were recorded on Bru-
ker 320MS/420GC mass spectrometer and HRMS (ESI) spectra were
recorded on a Intectra Mass AMD 402 or 604 mass spectrometer.
TLC was run on the silica gel coated plastic sheets (Silica Gel 60,
230–400 mesh, Merck, Germany) and visualized in UV light (254
or 365 nm).
4.3.1. 2,4-Dimethoxy-40-methylthio-trans-stilbene (S1)
Yield: 0.91 g (70%); mp 107–108 °C; 1H NMR (400 MHz, CDCl3)
ppm (d): 7.51 (d, J = 8.50, 1H, C6-H), 7.45 (d, J = 8.40, 2H, C20-H, C60-
H), 7.37 (d, J = 16.50, 1H, Cvin-H), 7.24 (d, J = 8.40, 2H, C30-H, C50-H),
6.98 (d, J = 16.50, 1H, Cvin0-H), 6.53 (d, J = 8.5, 1H, C5-H), 6.49 (s,
1H, C3-H), 3.89 (s, 3H, C4-OCH3), 3.85 (s, 3H, C2-OCH3), 2.51 (s,
3H, C40-SCH3).
13C NMR (101 MHz, CDCl3) ppm (d): 160.46 (C2), 157.96 (C4),
136.76 (C40), 135.41 (C10), 127.12 (C6), 126.82 (Cvin), 126.66
(C20, C60), 126.30 (C30, C50), 122.72 (Cvin0), 119.48 (C1), 104.97
(C5), 98.46 (C3), 55.37 (C4-OCH3), 55.37 (C2-OCH3), 15.99 (C40-
SCH3). LRMS (EI) m/z 286 [M+, 100%]. HRMS (ESI) m/z Calcd for
C
17H18O2S [M+H]+: 286.1028. Found: 286.1010.
4.3.2. 2,5-Dimethoxy-40-methylthio-trans-stilbene (S2)
Yield: 0.85 g (65%); mp 58–60 °C; 1H NMR (400 MHz, CDCl3)
ppm (d): 7.47 (d, J = 8.30, 2H, C20-H, C60-H), 7.43 (d, J = 16.50, 1H,
Cvin-H), 7.25 (d, J = 8.40, 2H, C30-H, C50-H), 7.16 (s, 1H, C6-H),
7.06 (d, J = 16.37, 1H, Cvin0-H), 6.85 (d, J = 8.9, 1H, C3-H), 6.81 (d,
J = 8.9, 1H, C4-H), 3.86 (s, 3H, C2-OCH3), 3.83 (s, 3H, C5-OCH3),
2.51 (s, 3H, C40-SCH3).
13C NMR (101 MHz, CDCl3) ppm (d):153.73 (C2), 151.38 (C5),
137.55 (C40), 134.76 (C10), 128.64 (Cvin), 127.20 (C20, C60), 126.96
(C1), 126.65 (Cvin0), 122.64 (C30, C50), 113.62 (C4), 112.25 (C6),
111.51 (C3), 56.22 (C2-OCH3), 55.74 (C5-OCH3), 15.81 (C40-SCH3).
LRMS (EI) m/z 286 [M+, 100%]. HRMS (ESI) m/z Calcd for
4.2. Diethyl 4-(methylthiobenzyl) phosphonate (2)
4.2.1. 4-Methylthiobenzyl chloride (1)
C
17H18O2S [M+H]+: 286.1028. Found: 286.1030.
To a solution of 4-methylthiobenzyl alcohol (13.0 g, 84.2 mmol)
in anhydrous toluene (150 mL) was added dropwise neat thionyl
chloride (7.6 mL, 12.4 g, 104 mmol) at room temperature. The
resultant pale yellow solution was stirred at room temperature
for 30 min, and ice-cold brine (150 mL) was added. The organic
phase was separated, washed with ice-cold brine (5 ꢀ 150 mL),
dried over MgSO4 and filtered. The solvent was evaporated under
reduced pressure. The residue (14.0 g) was distilled in vacuum at
115 °C/2.5 mmHg to yield 1 (12.27 g, 84.3%) of viscous oil which
crystallized easily in refrigerator. 1H NMR (300 MHz, CDCl3) ppm
(d): 7.20–7.35 (m, 4H), 4.56 (s, 2H), 2.49 (s, 3H). 13C NMR
(75 MHz, CDCl3) ppm (d): 139.3, 134.3, 129.2, 126.7, 46.1, 15.8.
4.3.3. 2,3,4-Trimethoxy-40-methylthio-trans-stilbene (S3)
Yield: 0.87 g (60%); mp 144–146 °C; 1H NMR (400 MHz, CDCl3)
ppm (d): 7.44 (d, J = 8.30, 2H, C20-H, C60-H), 7.30 (d, J = 16.30, 1H,
Cvin-H), 7.30 (d, J = 8.8, 1H, C6-H), 7.24 (d, J = 8.30, 2H, C30-H,
C50-H), 6.99 (d, J = 16.50, 1H, Cvin0-H), 6.70 (d, J = 8.8, 1H, C5-H),
3.91 (s, 3H, C3-OCH3), 3.91 (s, 3H, C4-OCH3), 3.88 (s, 3H, C2-
OCH3), 2.50 (s, 3H, C40-SCH3).
13C NMR (101 MHz, CDCl3) ppm (d): 153.17 (C2), 151.64 (C4),
142.38 (C3), 137.26 (C40), 134.93 (C10), 127.21 (C20, C60), 126.72
(Cvin), 126.71 (C30, C50), 124.42 (Cvin0), 122.31 (C6), 120.56 (C1),
107.76 (C5), 61.26 (C3-OCH3), 60.85 (C2-OCH3), 55.99 (C4-OCH3),
15.85 (C40-SCH3). LRMS (EI) m/z 316 [M+, 100%]. HRMS (ESI) m/z
Calcd for C18H12O3S [M+H]+: 316.1133. Found: 316.1135.
4.2.2. Diethyl 4-(methylthiobenzyl) phosphonate (2)
To a triethyl phosphite (14.27 mL, 13.82 g, 83.2 mmol) which
was heated at reflux, 4-methylthiobenzyl chloride (12.27 g,
71.06 mmol) was added dropwise with stirring, at such a rate that
a gentle reflux was maintained. When the addition was completed,
the reaction was refluxed for an additional 2 h. The mixture was
cooled to 25 °C, and the product mixture was fractionally distilled
under vacuum to yield 15.6 g (80%) of water clear, viscous liquid:
bp 142–145 °C/0.025 mmHg; 1H NMR (300 MHz, CDCl3) ppm (d):
4.3.4. 2,4,5-Trimethoxy-40-methylthio-trans-stilbene (S4)
Yield: 0.91 g (63%); mp 113–115 °C; 1H NMR (400 MHz, CDCl3)
ppm (d): 7.46 (d, J = 8.30, 2H, C20-H, C60-H), 7.40 (d, J = 16.50, 1H,
Cvin-H), 7.24 (d, J = 8.40, 2H, C30-H, C50-H), 7.13 (s, 1H, C6-H),
6.94 (d, J = 16.40, 1H, Cvin0-H), 6.55 (s, 1H, C3-H), 3.93 (s, 6H, C4-
OCH3, C5-OCH3), 3.88 (s, 3H, C2-OCH3), 2.51 (s, 3H, C40-SCH3).

R. Mikstacka et al. / Bioorg. Med. Chem. 20 (2012) 5117–5126
5125
13C NMR (101 MHz, CDCl3) ppm (d): 151.68 (C2), 149.59 (C4),
143.44 (C5), 136.93 (C40), 135.19 (C10), 126.79 (C20, C60), 126.67
(C30, C5), 126.16 (Cvin), 122.41 (Cvin0), 118.24 (C1), 109.34 (C6),
97.71 (C3), 56.68 (C4-OCH3), 56.53 (C5-OCH3), 56.06 (C2-OCH3),
15.93 (C40-SCH3). LRMS (EI) m/z 316 [M+, 100%]. HRMS (ESI) m/z
Calcd for C18H12O3S [M+H]+: 316.1133. Found: 316.1146.
The concentrations of compounds required for 50% inhibition of
catalytic activities (IC50) were determined graphically by plotting
percent of control enzyme activity versus inhibitor concentration.
All assays were performed in three parallel experiments. The coef-
ficient of variation did not exceed 15%.
4.5. NADPH oxidation
4.3.5. 2,4,6-Trimethoxy-40-methylthio-trans-stilbene (S5)
Yield: 0.69 g (48%); mp 109–110 °C; 1H NMR (400 MHz, CDCl3)
ppm (d): 7.49 – 7.41 (m, 3H, C20-H, C60-H, Cvin-H), 7.38 (d,
J = 16.60, 1H, Cvin0-H), 7.24 (d, J = 8.50, 2H, C30-H, C50-H), 6.19 (s,
2H, C3-H, C5-H), 3.90 (s, 6H, C2-OCH3, C6-OCH3), 3.86 (s, 3H, C4-
OCH3), 2.51 (s, 3H, C40-SCH3).
To establish whether 40-methylthiostilbene derivatives are
metabolized by CYP1A2, NADPH oxidation was estimated spectro-
photometrically at 340 nm in the presence of enzyme according to
the procedure described by Dávila-Borja et al. with minor modifi-
cation.39 Reaction mixture contained 5 pmol/ml CYP1A2 in Tris
buffer, pH 7.4, and a stilbene derivative at the concentration in
13C NMR (101 MHz, CDCl3) ppm (d): 160.17 (C4), 159.45 (C2,
C6), 136.90 (C40), 136.07 (C10), 129.18 (Cvin), 126.93 (C20, C60),
126.55 (C30, C50), 119.37 (Cvin0), 108.10 (C1), 90.79 (C3, C5),
55.75 (C2-OCH3, C6-OCH3), 55.29 (C4-OCH3), 16.16 (C40-SCH3).
LRMS (EI) m/z 316 [M+, 100%]. HRMS (ESI) m/z Calcd for
C18H12O3S [M+H]+: 316.1133. Found: 316.1146.
the range 1–50 lM. The reaction was initiated by adding of
0.2 mM NADPH. Incubation at 37 °C was carried on for 3 min and
NADPH oxidation was spectrophotometrically recorded every
15 s. The amount of NADPH oxidized was calculated by using the
extinction coefficient 6.22 [mMꢁ1 cmꢁ1] at 340 nm. As a positive
control, 7-ethoxyresorufin was used; at the concentration of
4.3.6. 3,4,5-Trimethoxy-40-methylthio-trans-stilbene (S6)
Yield: 1.08 g (75%); mp 111–113 °C [lit. 109–111 °C8]; 1H NMR
(400 MHz, CDCl3) ppm (d): 7.22 (d, J = 8.30, 2H, C20-H, C60-H), 7.14
(d, J = 8.30, 2H, C30-H, C50-H), 6.51 (d, J = 16.50, 1H, Cvin-H), 6.48
(d, J = 16.30, 1H, Cvin0-H), 6.48 (s, 2H, C2-H, C6-H), 3.84 (s, 3H,
C4-OCH3), 3.68 (s, 6H, C3-OCH3, C5-OCH3), 2.46 (s, 3H, C40-SCH3).
13C NMR (101 MHz, CDCl3) ppm (d): 152. 91 (C3, C5), 137.36
(C4), 137.21 (C40), 134.04 (C10), 132.55 (C1), 129.88 (Cvin),
129.41 (C20, C60), 129.28 (Cvin0), 126.16 (C30, C50), 105.94 (C2,
C6), 60.90 (C4-OCH3), 55.89 (C3-OCH3, C5-OCH3), 15.74 (C40-
SCH3). LRMS (EI) m/z 301 (70), 316 [M+, 100%]. HRMS (ESI) m/z
Calcd for C18H12O3S [M+H]+: 316.1133. Found: 316.1139.
2 lM increase in NADPH oxidation was observed in the full range
of time (3 min).
4.6. CYP1A2 metabolism
Reaction mixture containing 40 pmol human recombinant
CYP1A2, 1.3 mM NADP, 3.3 mM glucose-6-phosphate, 1.5 U/ml
glucose-6-phosphate dehydrogenase and 20 lM stilbene deriva-
tive: 3,4,5-trimethoxy-40-methylthiostilbene or 2,4,5-trimethoxy-
40-methylthiostilbene as a substrate, was incubated at 37 °C for
30 min.
4.7. GC–MS analysis of metabolic products of 3,4,5-trimethoxy-
40-MTS and 2,4,5-trimethoxy-40-MTS
4.3.7. 2-Chloro-40-methylthio-trans-stilbene (S7)
Yield: 0.87 g (73%); mp 90–92 °C; 1H NMR (400 MHz, CDCl3)
ppm (d): 7.68 (d, J = 7.80, 1H, C6-H), 7.48 (d, J = 16.20, 1H, Cvin-
H), 7.48 (d, J = 8.40, 2H, C20-H, C60-H), 7.39 (d, J = 7.90, 1H, C3-H),
7.26 (t, J = 6.30, 1H, C5-H), 7.26 (d, J = 8.50, 2H, C30-H, C50-H),
7.19 (t, J = 7.60, 1H, C4-H), 7.04 (d, J = 16.20, 1H, Cvin0-H), 2.51 (s,
3H, C40-SCH3).
Reference chromatograms and mass spectra of neat samples
(50 lg/ml ethyl acetate) of S6 and S4 were obtained by GC–MS
(Figs. 1 and 2, respectively). For the determination of metabolic
products, 3 mg of lyophilized reaction solution was re-dissolved
in 200
(100 l, twice). The ethyl acetate layers were combined and dried
under a stream of nitrogen, reconstituted to 50 l of ethyl acetate,
ll of distilled water then partitioned with ethyl acetate
13C NMR (101 MHz, CDCl3) ppm (d): 138.50 (C40), 135.38 (C10),
133.95 (C1), 133.33 (C2), 130.56 (C20, C60), 129.81 (C3), 128.41
(C4), 127.18 (Cvin), 126.88 (C6), 126.58 (C5), 126.32 (C30, C50),
124.03 (Cvin0), 15.70 (C40-SCH3). LRMS (EI) m/z 178 (100), 260
[M+, 93%]. HRMS (ESI) m/z Calcd for C15H13ClS [M+H]+: 260.4948.
Found: 260.4936.
l
l
and used for GC–MS analysis. GC–MS was performed on a JEOL
gcmate II Instrument (JEOL USA Inc., Peabody, MA) using a J&W
DB-5 capillary column (0.25 mm internal diameter, 0.25 lM film
thickness, and 30 m length; Agilent Technologies, Foster City,
CA). The GC was run under the following temperature program:
initial 250 °C held for 1 min, increased to 275 at 10 °C/min rate, in-
creased to 300 °C at the rate of 3 °C/min and held at this tempera-
ture for 3.0 min. The carrier gas was ultrahigh purity helium, at
1 ml/min flow rate. The injection port, GC–MS interface and ioniza-
tion chamber were at 250, 230, and 230 °C, respectively. The vol-
4.4. 7-Ethoxyresorufin O-deethylation (EROD) enzyme assay
To assess CYP1 enzyme activities 7-ethoxyresorufin O-deethy-
lase (EROD) activity was measured according to the method of
Burke et al.38 The test compounds were freshly dissolved in
dimethylsulfoxide (DMSO) on the day of experiment.
ume of injection was 1 ll, splitless injection. Mass spectra were
The reaction mixture (1 ml total volume) contained the various
concentrations of a test compound, 1.3 mM NADP+, 3.3 mM glu-
cose-6-phosphate, 0.5 U/ml glucose-6-phosphate dehydrogenase,
acquired in positive electron impact (70 ev), low-resolution mode.
Under these conditions S6 had a retention time of 11.58 min
(Fig. 1A) and S4 had a retention time of 11.92 min (Fig. 2A). The
monohydroxylated products were revealed from a reconstructed
ion chromatogram specifying m/z 332 (Figs. 1C and 2C). Using m/
z 348, no peak appeared indicating absence of dihydroxylated
metabolites. GC–MS analyses were done in duplicates.
3.3 mM magnesium chloride and
2 lM 7-ethoxyresorufin in
100 mM potassium phosphate (pH 7.4). The reactions were initi-
ated by the addition of human recombinant cytochromes
(1.25 pmole CYP1A1, 5 pmole CYP1A2 or 5 pmole CYP1B1) and
samples were incubated at 37 °C for 15 min. The fluorescence of
the product was determined on a HITACHI Model F 2500 fluores-
cence spectrophotometer (kex550 and kem585). The quantitation
of the deethylated metabolite was based on comparison of its fluo-
rescence with resorufin as a standard. Control incubations did not
contain the test compounds.
4.8. Docking
Analyzed molecules were docked to the active sites of CYP1A2
(PDB: 2HI4) and CYP1B1 (PDB: 3PM0) with the use of Accelrys Dis-
covery Studio 2.5.5 suite of programs by LigandFit procedure.40 For

5126
R. Mikstacka et al. / Bioorg. Med. Chem. 20 (2012) 5117–5126
5. Saiko, P.; Szakmary, A.; Jaeger, W.; Szekeres, T. Mutat. Res. 2008, 658, 68.
6. Lappano, R.; Rosano, C.; Madeo, A.; Albanio, L.; Plastina, P.; Gabriele, B.; Forti, L.;
Stivala, L. A.; Iacopetta, D.; Dolce, V.; Andò, S.; Pezzi, V.; Maggiolini, M. Mol.
Nutr. Food Res. 2009, 53, 845.
7. Kang, S. S.; Cuendet, M.; Endringer, D. C.; Croy, V. L.; Pezzuto, J. M.; Lipton, M. A.
Bioorg. Med. Chem. 2009, 17, 1044.
these two targets all water molecules were removed and hydrogen
atoms were added. While coordinates of all heavy atoms were kept
frozen, hydrogen atoms positions were optimized with the use of
adopted basis Newton-Raphson minimization algorithm (max
steps = 500, RMS gradient = 0.01 and distance-dependent dielec-
trics as implicit solvent model). The binding site was defined from
8. Cushman, M.; Nagarathnam, D.; Gopal, D.; He, H.-M.; Lin, C. M.; Hamel, E. J.
Med. Chem. 1992, 35, 2293.
9. Ma, Z.; Molavi, O.; Haddadi, A.; Lai, R.; Gossage, R. A.; Lavasanifar, A. Cancer
Chemother. Pharmacol. 2008, 63, 27.
10. Mazué, F.; Colin, D.; Gobbo, J.; Wegner, M.; Rescifina, A.; Spatafora, C.; Fasseur,
D.; Delmas, D.; Meunier, P.; Tringali, C.; Latruffe, N. Eur. J. Med. Chem. 2010, 45,
2972.
11. Walle, T.; Hsieh, F.; DeLegge, M. H.; Oatis, J. E., Jr.; Walle, U. K. Drug Metab.
Dispos. 2004, 32, 1377.
12. Walle, T.; Ta, N.; Kawamori, T.; Wen, X.; Tsuji, P. A.; Walle, U. K. Biochem.
Pharmacol. 2007, 73, 1288.
13. Wilson, M. A.; Rimando, A. M.; Wolkow, C. A. BMC Pharmacol. 2008, 8, 15.
14. Liehr, J. G.; Ricci, M. J. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 3294.
15. Bruno, R. D.; Njar, V. C. O. Bioorg. Med. Chem. 2007, 15, 5047.
16. Jennings, B. L.; Sahan-Firat, S.; Estes, A. M.; Das, K.; Farjana, N.; Fang, X. R.;
Gonzales, F. J.; Malik, K. U. Hypertension 2010, 56, 667.
17. Chang, T. K. H.; Chen, J.; Lee, W. B. K. JPET 2001, 299, 874.
18. Guengerich, F. P.; Chun, Y.-J.; Kim, D.; Gillam, E. M. J.; Shimada, T. Mutat. Res.
2003, 523–524, 173.
19. Mikstacka, R.; Rimando, A. M.; Szalaty, K.; Stasik, K.; Baer-Dubowska, W.
Xenobiotica 2006, 36, 269.
20. Mikstacka, R.; Przybylska, D.; Rimando, A. M.; Baer-Dubowska, W. Mol. Nutr.
Food Res. 2007, 51, 517.
21. Chun, Y. J.; Kim, S.; Kim, D.; Lee, S. K.; Guengerich, F. P. Cancer Res. 2001, 61,
8164.
22. Chun, Y. J.; Lim, C.; Ohk, S. O.; Lee, J. M.; Lee, J. H.; Choi, S.; Kim, S. Med. Chem.
Commun. 2011, 2, 402.
23. Chun, Y. J.; Ryu, S. Y.; Leong, T. C.; Kim, M. Y. Drug Metab. Dispos. 2001, 29, 389.
24. Mikstacka, R.; Baer-Dubowska, W.; Wieczorek, M.; Sobiak, S. Mol. Nutr. Food
Res. 2008, 52, S77.
25. Yang, H.; Baur, J. A.; Chen, A.; Miller, C.; Sinclair, D. A. Aging Cell 2007, 6, 35.
26. Vasanthanathan, P.; Hritz, J.; Taboureau, O.; Olsen, L.; Jorgensen, F. S.;
Vermeulen, N. P. E.; Oostenbrink, C. J. Chem. Inf. Model. 2009, 49, 43.
27. Smith, D. A.; Ackland, M. J.; Jones, B. C. Drug Discov. Today 1997, 2, 406.
28. Don, M.-J.; Lewis, D. F. V.; Wang, S.-Y.; Tsai, M.-W. Bioorg. Med. Chem. Lett.
2003, 13, 2535.
the volume of a cocrystallized ligand (a-naphthoflavone). The grid
resolution was set to default value of 0.5 Å.
For the conformational search of the ligands a Monte Carlo
method was employed. During this search internal ligand electro-
static energy was not included and the variable number of trials,
depending on the number of rotatable ligand bonds, was used
according to non-modified default settings for LigandFit procedure.
In all calculations for proteins and ligands CHARMm forcefield with
MMFF94 partial charge rules were used. For docking we have used
default values of parameters, except for DockScore function, where
CFF force field was used instead of default Dreiding.
Twenty poses were saved for each ligand after docking and then
minimized in situ with use of adopted basis Newton-Raphson algo-
rithm. During each docking run poses were scored with seven scor-
ing functions DockScore (CFF), LigScore1, LigScore2, PLP1, PLP2,
Jain and PMF (Table S1–S3). For further evaluation DockScore CFF
function (DockScore CFF force field version) was chosen as exhibit-
ing the best correlation with experimental data. CFF force field was
used also for LigScore calculations. For saved poses we applied pro-
tocol ‘Analyze ligand poses’ to calculate RMS deviation using first
pose of each ligand as a reference and to find hydrogen bonds
and close contacts. After visual inspection of poses and analysis
of RMSD values, repeated poses were discarded. For the best poses,
these characterized by the highest values of DockScore function,
molecular volume and strain energy were calculated.
29. Lewis, D. F. V.; Lake, B. G.; Dickins, M. Xenobiotica 2004, 34, 501.
30. Lewis, D. F. V.; Ito, Y.; Lake, B. G. Toxicol. In Vitro 2006, 20, 256.
31. Takemura, H.; Itoh, T.; Yamamoto, K.; Sakakibara, H.; Shimoi, K. Bioorg. Med.
Chem. 2010, 18, 6310.
Acknowledgment
´
This work was supported by funding from Poznan University of
32. Piver, B.; Fer, M.; Vitrac, X.; Merillon, J.-M.; Dreano, Y.; Berthou, F.; Lucas, D.
Biochem. Pharmacol. 2004, 68, 773.
Medical Sciences (Grant No 501-01-03313427-08032).
33. Sansen, S.; Yano, J. K.; Reynald, R. L.; Schoch, G. A.; Griffin, K. J.; Stout, C. D.;
Johnson, E. F. J. Biol. Chem. 2007, 282, 14348.
Supplementary data
34. Wang, A.; Savas, U.; Stout, C. D.; Johnson, E. F. J. Biol. Chem. 2011, 286, 5736.
35. Montes, M.; Miteva, M. A.; Villoutreix, B. O. Proteins 2007, 68, 712.
36. Li, X.; Li, Y.; Cheng, T.; Liu, Z.; Wang, R. J. Comp. Chem. 2010, 31, 2109.
37. de Graaf, C.; Oostenbrink, C.; Keizers, P. H.; van der Wijst, W. T.; Jongejan, A.;
Vermeulen, N. P. E. J. Med. Chem. 2006, 49, 2417.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2012.07.012.
38. Burke, M. D.; Thompson, S.; Elcombe, C. R.; Halpert, J.; Haaparanta, T.; Mayer, R.
T. Biochem. Pharmacol. 1985, 34, 3337.
References and notes
39. Dávila-Borja, V. M.; Belmont, J. A.; Espinosa, J. J.; Moreno-Sanchez, R.; Albores,
A.; Montero, R. D. Arch. Toxicol. 2007, 81, 697.
40. Venkatachalam, C. M.; Jiang, X.; Oldfield, T.; Waldman, M. J. Mol. Graph. Model.
2003, 21, 289.
1. Goswami, S. K.; Das, D. K. Cancer Lett. 2009, 284, 1.
2. Rimando, A.; Suh, N. Planta Med. 2008, 74, 1635.
3. Bishayee, A.; Politis, T.; Darvesh, A. S. Cancer Treat. Rev. 2010, 36, 43.
4. Kundu, J.; Surh, Y.-J. Cancer Lett. 2008, 269, 243.