Synthesis, biological evaluation and docking studies of trans-stilbene methylthio derivatives as cytochromes P450 family 1 inhibitors

Article abstract of DOI:10.1111/cbdd.13042

Cytochromes P450 family 1 (CYP1) are responsible for the metabolism of procarcinogens, for example polycyclic aromatic hydrocarbons and aromatic and heterocyclic amines. The inhibition of CYP1 activity is examined in terms of chemoprevention and cancer chemotherapy. We designed and synthesized a series of trans-stilbene derivatives possessing a combination of methoxy and methylthio functional groups attached in different positions to the trans-stilbene skeleton. We determined the effects of synthesized compounds on the activities of human recombinant CYP1A1, CYP1A2 and CYP1B1 and, to explain the variation of inhibitory potency of methoxystilbene derivatives and their methylthio analogues, we employed computational analysis. The compounds were docked to CYP1A1, CYP1A2 and CYP1B1 binding sites with the use of Accelrys Discovery Studio 4.0 by the CDOCKER procedure. For CYP1A2 and CYP1B1, values of scoring functions correlated well with inhibitory potency of stilbene derivatives. All compounds were relatively poor inhibitors of CYP1A2 that possess the most narrow and flat enzyme cavity among CYP1s. For the most active CYP1A1 inhibitor, 2-methoxy-4′-methylthio-trans-stilbene, a high number of molecular interactions was observed, although the interaction energies were not distinctive.

Full text of DOI:10.1111/cbdd.13042

DR. MARCIN WIERZCHOWSKI (Orcid ID : 0000-0003-2619-0466)
DR. ZBIGNIEW DUTKIEWICZ (Orcid ID : 0000-0002-6141-3322)
Article type : Research Article
Synthesis, Biological Evaluation, and Docking Studies of trans-Stilbene Methylthio
Derivatives as Cytochromes P450 Family 1 Inhibitors
a
Marcin Wierzchowski§ , Zbigniew Dutkiewicz§ * a, Agnieszka Gielara-Korzańskaa, Artur Korzańskib,
Anna Teubert c, Artur Teżyk d ,Tomasz Stefańskia, Wanda Baer-Dubowskae, Renata Mikstackaf
a Department of Chemical Technology of Drugs, Faculty of Pharmacy, Poznan University of Medical Sciences,
Grunwaldzka 6, 60-780 Poznań, Poland
b Department of Crystallography, Adam Mickiewicz University, Umultowska 89B, 61-614 Poznań, Poland
c Institute of Bioorganic Chemistry, Polish Academy of Sciences, Noskowskiego 12/14, 61-704 Poznań, Poland
d Department of Forensic Medicine, Poznan University of Medical Sciences, Święcickiego 6, 60-781 Poznań,
Poland
eDepartment of Pharmaceutical Biochemistry, Poznan University of Medical Sciences, Święcickiego 4, 60-781
Poznań, Poland
fDepartment of Inorganic and Analytical Chemistry, Ludwik Rydygier Collegium Medicum, Nicolaus Copernicus
University in Toruń, Dr A. Jurasza 2, 85-089 Bydgoszcz, Poland
* Corresponding author: Zbigniew Dutkiewicz, Department of Chemical Technology of Drugs, Faculty of
Pharmacy, Poznań University of Medical Sciences, Grunwaldzka 6, 60-780 Poznań, Poland, phone:
+48 61 854 66 30; fax: +48 61 854 66 39, E-mail: zdutkie@ump.edu.pl
§ These authors contributed equally to the work presented here and should therefore be regarded as
equivalent authors
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1111/cbdd.13042
This article is protected by copyright. All rights reserved.

Keywords Cytochromes P450 family 1; trans-resveratrol analogues, molecular docking
Abstract
Cytochromes P450 family 1 (CYP1) are responsible for the metabolism of procarcinogens,
e.g. polycyclic aromatic hydrocarbons and aromatic and heterocyclic amines. The inhibition
of CYP1 activity is examined in terms of chemoprevention and cancer chemotherapy.
We designed and synthesized a series of trans-stilbene derivatives possessing a combination
of methoxy and methylthio functional groups attached in different positions to the trans-
stilbene skeleton. We determined the effects of synthesized compounds on the activities of
human recombinant CYP1A1, CYP1A2 and CYP1B1 and, to explain the variation of
inhibitory potency of methoxystilbene derivatives and their methylthio analogs, we employed
computational analysis. The compounds were docked to CYP1A1, CYP1A2 and CYP1B1
binding sites with the use of Accelrys Discovery Studio 4.0 by the CDOCKER procedure.
For CYP1A2 and CYP1B1, values of scoring functions correlated well with inhibitory
potency of stilbene derivatives. All compounds were relatively poor inhibitors of CYP1A2
that possess the most narrow and flat enzyme cavity among CYP1s. For the most active
CYP1A1 inhibitor, 2-methoxy-4’-methylthio-trans-stilben, a high number of molecular
interactions was observed, although the interaction energies were not distinctive.
Introduction
The cytochrome P450 superfamily comprises the enzymes which contain a heme prosthetic
group catalysing oxidation reactions and N- and O-dealkylations of numerous xenobiotics.
The cytochrome P450 family 1 (CYP1) represents isoforms CYP1A1, CYP1A2 and
CYP1B1, which are responsible for the phase I metabolism of endogenous and exogenous
This article is protected by copyright. All rights reserved.

substrates including drug compounds and procarcinogens, e.g. polycyclic aromatic
hydrocarbons and aromatic and heterocyclic amines. Isozyme CYP1B1 is involved in
pathogenesis of the hormone-induced cancers, being responsible for metabolism of 17-alpha-
estradiol (E2) to highly mutagenic and carcinogenic 4-hydroxy-E2 (1). Recently, the role of
CYP1B1 in cancer progression and metastasis was reported (2). The inhibition of CYP1B1
activity by natural compounds present in human diet constitutes one of the chemopreventive
strategies, while in cancer therapy the inhibition of CYP1A1 and CYP1B1 activities may be
helpful in avoiding drug resistance caused by the metabolism of chemotherapeutics catalysed
by CYP1s (3,4). Trans-resveratrol (3,4’,5-trihydroxy-trans-stilbene) is a natural polyphenol
found in grapes, berries and peanuts, showing well-characterized beneficial bioactivities. It
efficiently and selectively inhibits CYP1B1 activity (5), although its bioavailability in
humans was determined as poor (6) due to conjugation reactions with sulphuric and
glucuronic acid. In the last decade, the natural and synthetic trans-resveratrol analogues were
studied in the context of their interaction with CYP1s. The substitution of hydroxyl groups
with methoxy substituents efficiently improved bioavailability by preventing polyphenol
metabolism and concomitantly influenced the affinity of compounds to active sites of
cytochromes. Hence, polymethoxy derivatives of trans-stilbene are a class of compounds
worth to be studied with regard to their interaction with CYP1 enzymes.
Chun and coworkers reported a strong inhibitory potency of 3,5,2’4’-tetramethoxy-
trans-stilbene and 2,4,2’,6’-tetramethoxy-trans-stilbene against CYP1B1 activity (7-9). In
our previous studies, we found 3,4,2’-trimethoxy-trans-stilbene to be an efficient inhibitor of
CYP1B1 activity at nanomolar level (10). The effect of polymethoxy derivatives of 4-
methylthio-trans-stilbene on CYP1A1, CYP1A2 and CYP1B1 activities was studied,
indicating a strong dependence of the inhibitory effect and selectivity of action on the pattern
This article is protected by copyright. All rights reserved.

of substituents linked to the ligand core (11). To obtain more information on the ligand-
enzyme interaction molecular modelling and docking were used. All the previous studies
revealed the essential role of substituents in the positions 2 and 4 (7-11).
In the current study, we designed and synthesized a series of trans-stilbene derivatives
possessing methoxy and methylthio functional groups attached in 2, 3’ and 4’ positions to the
trans-stilbene skeleton (Table 1). We determined the effect of synthesized compounds on the
activities of human recombinant CYP1A1, CYP1A2 and CYP1B1 and, to explain the
variation of inhibitory potency of methoxystilbene derivatives and their methylthio
analogues, we employed computational analysis.
Methods
Chemicals
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-phosphate
dehydrogenase, nicotinamide adenine dinucleotide phosphate (NADP+), 7-ethoxyresorufin
and resorufin were obtained from Sigma (St. Louis, MO, USA). All the other chemicals and
reagents were of the highest grade available.
Chemical synthesis
A two-step synthesis of trans-stilbene compounds S1–S6 started from proper methoxybenzyl
or methylthiobenzyl chloride (1a–1c) via a Michaelis-Arbuzov reaction (triethyl phosphite,
solvent free, 24h, 140°C), resulting in corresponding diethyl phosphonates (2a – 2b)
according to the method described by Murias at al. (12). Next, phosphonates and methoxy or
This article is protected by copyright. All rights reserved.

methylthioaldehydes (3a – 3b) are transformed in Horner-Wadsworth-Emmons reaction
conditions (base - sodium methoxide, solvent - DMF, 24h, 0 next 100°C, inert gas - N2) to
planned compounds S1–S6 . The structures of all compounds are confirmed by NMR, mass
spectra, elementary analysis and of two compounds by crystallographic techniques.
R
R1
R
R1
H3C
O
A
R1
R
H3C
ii)
i)
O
+
+
R2
P
O
P
O
O
CH3
R2
O
Cl
O
H3C
B
CH3
R, R1, R2 = H- or CH3O- or CH3S-
(1a-1c)
(2a-2c)
(3a-3b)
(S1-S6)
Scheme 1: Synthesis scheme for compound S1-S6: i) solvent free, 140°C, 24h, ii) NaOCH3, DMF,
100°C, 24h. Synthesis and characteristics of 2-methoxy-4’-methylthio-trans-stilbene S2 and
2,3’,4’-trimethoxy-trans-stilbene S5 was described earlier (10,13).
The yield varied between 24-80% (Table 1). Generally, the yield of a reaction decreases with
an increase in the mass of the aldehyde applied. The presence of a sulphur atom in the final
molecule additionally decreases the yield of a reaction. This phenomenon may also be
associated with a higher volume of the methylthio group than the methoxy group. Horner-
Wadsworth-Emmons reaction is a selective way to possess E isomer. We have not observed
any traces of Z isomers during realization of synthetic procedure.
General procedures of synthesis of compounds S1, S3, S4 and S6
Proper phosphonate and 10 mL of dry DMF were placed in round-flask and cooled to 0°C in
ambient atmosphere (under nitrogen). Next sodium methoxide and proper benzaldehyde were
added and resulting reaction mixture was strongly stirred under nitrogen for 1h in room
temperature and next 1.5 h in 100°C. After cooling to room temperature reaction mixture was
poured into 500 mL of water with crushed ice and left overnight. Resulting precipitate was
This article is protected by copyright. All rights reserved.

filtered off, washed with distilled water (3×25 mL) and purified by crystallization from 95%
ethanol. It gave planned compounds S1, S3, S4, S6. Details are described below.
2,4’-dimethoxy-trans-stilbene (S1)
In the reaction applied diethyl 4-methoxybenzylphosphonate 2a (1.50 g, 5.81 mmol), sodium
methoxide (0.63 g, 11.62 mmol) and 2-methoxybenzaldehyde 3a (0.79 g, 5.81 mmol). It gave
2,4’-dimethoxy-trans-stilbene (S1) as white crystalline solid (1.12 g, 80%). M.p. = 91-92 °C.
TLC Rf (n-Hex:EtOAc 7:1)= 0.44. 1H NMR (500 MHz, DMSO-d6) δ 7.62 (dd, J = 7.5, 1.0
Hz, 1H), 7.49 (d, J = 8.5 Hz, 2H), 7.27 (d, J = 16.5 Hz, 1H), 7.25 – 7.20 (m, 1H), 7.15 (d, J =
16.5 Hz, 1H), 7.02 (d, J = 8.0 Hz, 1H), 6.96 (to, J = 7.5 Hz, 1H), 6.94 (do, J = 8.5 Hz, 2H),
3.84 (s, 3H), 3.77 (s, 3H). 13C NMR (125 MHz, DMSO-d6) δ 158.9, 156.3, 130.1, 128.4,
128.4, 127.6, 125.9, 125.8, 120.6, 120.6, 114.2, 111.3, 55.4, 55.1 (graphical presentation of
signal-structure allocation - Figure 6S). LRMS(EI) m/z 240 [M]+. Anal. cal. for C16H16O2
C(79.97%), H(6.71%). Found C(79.65%), H(7.09%). Crystal data for S1 (Supporting
Information, Tables 1S–7S, Figures 1S–2S), CCDC 1482227.
2-methylthio-4’-methoxy-trans-stilbene (S3)
In the reaction applied diethyl 4-methoxybenzylphosphonate 2a (0.52 g, 2.00 mmol), sodium
methoxide (0.22 g, 4.00 mmol) and 2-methylthiobenzaldehyde 3b (0.30 g, 2.00 mmol). It
gave 2-methylthio-4’-methoxy-trans-stilbene (S3) as white crystalline solid (179 mg, 35%).
M.p. = 46-48 °C. TLC Rf (n-Hex:EtOAc 7:1)= 0.56. 1H NMR (500 MHz, DMSO-d6) δ 7.63
(dd, J = 7.5, 0.9 Hz, 1H), 7.51 (d, J = 8.5 Hz, 2H), 7.35–7.23 (m, 3H), 7.19 (td, J = 7.5, 1.0
Hz, 1H), 7.11 (d, J = 16.0 Hz, 1H), 6.96 (d, J = 8.5 Hz, 2H), 3.77 (s, 3H), 2.47 (s, 3H). 13C
NMR (125 MHz, DMSO-d6) δ 159.6, 136.9, 135.9, 130.5, 130.3, 130.0, 128.3, 126.6, 125.8,
125.7, 123.0, 114.7, 55.6, 15.7 (graphical presentation of signal-structure allocation - Figure
7S). HRMS(EI) [m/z]: calculated for C16H16OS 256.09219. Found 256.09245. Anal. cal. for
C16H16OS C(74.96%), H(6.29%), S(12.51%). Found C(74.70%), H(6.19%), S(12.41%).
2,4’-dimethylthio-trans-stilbene (S4)
In the reaction applied diethyl (4-methylthiobenzyl)phosphonate 2b (0.55 g, 2.00 mmol),
sodium methoxide (0.22 g, 4.00 mmol) and 2-methylthiobenzaldehyde 3b (0.30 g, 2.00
mmol. It gave 2,4’-dimethylthio-trans-stilbene (S4) as white crystalline solid (150 mg, 28%)
M.p. = 69-70 °C. TLC Rf (n-Hex:EtOAc 7:1)= 0.54. 1H NMR (500 MHz, DMSO-d6) δ 7.67
This article is protected by copyright. All rights reserved.

(d, J = 7.5 Hz, 1H), 7.53 (d, J = 8.0 Hz, 2H), 7.42 (d, J = 16.0 Hz, 1H), 7.33 (t, J = 7.0 Hz,
1H), 7.31-7.25 (m, 3H), 7.21 (t, J = 7.0 Hz, 1H), 7.15 (d, J = 16.0 Hz, 1H), 2.50 (so, 3H),
2.49 (s, 3H). 13C NMR (125 MHz, DMSO-d6) δ 138.0, 136.7, 135.2, 133.6, 129.9, 128.2,
127.0, 126.2, 126.1, 125.6, 125.2, 124.0, 15.3, 14.6 (graphical presentation of signal-structure
allocation - Figure 8S). HRMS(EI) [m/z]: calculated for C16H16S2 272.06934. Found
272.06956. Anal. cal. for C16H16S2 C(70.54%), H(5.92%), S(23.54%). Found C(70.58%),
H(5.93%), S(23.63%).
2-methylthio-3’,4’-dimethoxy-trans-stilbene (S6)
In the reaction applied diethyl (3,4-dimethoxybenzyl)phosphonate 2c (0.58 g, 2.00 mmol),
sodium methoxide (0.22 g, 4.00 mmol) and 2-methylthiobenzaldehyde 3b (0.30 g, 2.00
mmol). It gave 2-methylthio-3’,4’-dimethoxy-trans-stilbene (S6) as white crystalline solid
(137 mg, 24%). M.p. = 99-100 °C. TLC Rf (n-Hex:EtOAc 7:1)= 0.27. 1H NMR (500 MHz,
DMSO-d6) δ 7.62 (d, J = 7.5 Hz, 1H), 7.34–7.24 (m, 3H), 7.23–7.14 (m, 2H), 7.14–7.05 (m,
2H), 6.97 (d, J = 8.5 Hz, 1H), 3.81 (s, 3H), 3.78 (s, 3H), 2.48 (s, 3H). 13C NMR (126 MHz,
DMSO) δ 149.0, 148.9, 136.5, 135.5, 130.6, 129.9, 127.9, 126.0, 125.5, 125.2, 122.8, 119.7,
111.9, 109.6, 55.5, 55.5, 15.2 (graphical presentation of signal-structure allocation - Figure
4). HRMS(EI) [m/z]: calculated for C16H16O2S 286.10275. Found 28610186. Anal. cal. for
C16H16O2S C(71.30%), H(6.34%), S(11.20%). Found C(71.16%), H(6.48%), S(11.10%).
Crystal data for S6 (Supporting Information, Tables 8S–14S, Figures 3S–5S) CCDC
1482226.
Crystallographic data
The identification of compounds S1 and S6 was confirmed by X-ray crystallography and data
are included in Supporting Information.
The most active CYP1B1 inhibitor 3,4,2'-trimethoxy-trans-stilbene S5 (crystallographic
structure in (10)) and new designed analogue 2-methylthio-3',4'-dimethoxystilbene S6
crystalized in monoclinic crystal system in P21/c and P21 space groups respectively. Both tri-
substituted structures reveal similarity. Structure of 2-methylthio-3',4'-dimethoxystilbene S6
This article is protected by copyright. All rights reserved.

is very similar to 3,4,2'-trimethoxy-trans-stilbene S5 where the sulphur atom is replaced with
oxygen atom. The angle between planes created by benzene rings for S6 is bigger (6,4(1)°)
than this one for S5 (3,05(8)°). Angle between C9-C8-C7 in S5 is smaller (124,27°) than
equivalent C4-C9-C10 (126,83°) in S6 and adequately C8-C7-C4 angle (128,4°) in S5 is
bigger than equivalent C9-C10-C11 (127,04°) in S6.
As seen in Figures 1 and 2, the crystal packing of S5 and S6 is very similar, and the main
difference between these packings lies in the directions of molecules arrangement.
NMR study
Structure of synthetized trans-stilbene compounds were confirmed and analyzed using
various NMR techniques. Figure 4 presents annotated signals 1H NMR and 13C NMR of
molecule S6. 2D NMR techniques as 1H-1H COSY (Correlation SpectroscopY), HSQC
(Heteronuclear Single Quantum Coherence) and HMBC (Heteronuclear Multiple-Bond
Correlation spectroscopy) enabled the allocation of all carbon peaks present in the 13C NMR
experiments. Exact position of overlapped and hidden signals were recognized thanks to a
presence of corresponding spots in 2D NMR experiments. Spectra analysis shown
characteristic signals of trans-stilbenes. Two doublets of group -CH=CH- have been
observed with coupling constants 3J between 16.0–16.5 Hz. Protons signals of methoxy
groups were observed as singlets in range 3.77–3.84 ppm while methylthio group protons
were shown as singlets in range 2.47–2.50 ppm. 13C NMR carbons signals of CH3O- are
observed in range 55.1–55.6 ppm. Methylthio group carbons reveal signals between 14.6–
15.7 ppm. Figures 6S–8S in Supporting Information shown annotated signals for compounds
S1, S3, and S4.
This article is protected by copyright. All rights reserved.

7-ethoxyresorufin O-deethylation (EROD) enzyme assay
The activity of cytochromes was measured spectrofluorometrically according to the method
of Burke et al. (14) with the use of 7-ethoxyresorufin as described earlier (10). Briefly, the
reaction mixture (1 ml total volume) contained various concentrations of a test compound,
1.3 mM NADP+, 3.3 mM glucose-6-phosphate, 0.5 U/ml glucose-6-phosphate
dehydrogenase, 3.3 mM magnesium chloride, 0.5 µM 7-ethoxyresorufin in 100 mM
potassium phosphate (pH 7.4), and human recombinant cytochromes: 1.25 pmole CYP1A1, 5
pmole CYP1A2 or 5 pmole CYP1B1. Samples were incubated at 37°C for 15 min. The
fluorescence of the product was determined on a HITACHI Model F 2500 fluorescence
spectrophotometer (ex550 and em585). The quantitation of the deethylated product was
based on the comparison of its fluorescence with resorufin as a standard.
Docking studies
The studied methoxylated and methylthiolated derivatives of trans-stilbene and reference
ligand -naphthoflavone (ANF) were docked to CYP1A1 (PDB: 4I8V) (15), CYP1A2 (PDB:
2HI4) (16) and CYP1B1 (PDB: 3PM0) (17) binding sites with the use of the CDOCKER
procedure (18) implemented in Accelrys Discovery Studio 4.0a. Receptors were prepared for
docking by the automatic ‘Prepare protein’ procedure, which removed water molecules,
added hydrogen atoms and protonated amino acid residues at the specified pH (pH = 7.4).
All calculations were performed with the use of the CHARMm forcefield. Partial charges for
receptors were set according to the Momany-Rone (MR) method.
This article is protected by copyright. All rights reserved.

For all targets, the binding site was defined around the centroid of the cocrystallized
ligand (-naphthoflavone). The radius of the sphere was set to the value of 10 Å. During
docking, MMFF94 partial charge rules were used for ligands. For each ligand, 20 initial
conformations were generated randomly using high-temperature grid-based molecular
dynamics, and then translated into the binding site. Ten orientations of each conformation
were used for docking and further simulated annealing and minimization. For the molecular
dynamics and simulated annealing, default values of parameters were used. Finally, poses
were minimized with full potential, the complete CHARMm forcefield expression. Twenty
poses were saved for each ligand after docking.
Binding free energies for selected receptor-ligand complexes, namely for poses
characterized by the highest values of –CDOCKER energy (the sum of the interaction energy
and ligand strain), were calculated (Table 2). In ‘Calculate binding energies’ protocol
Poisson-Boltzmann with non-polar surface area (PBSA) an implicit solvent model was used,
the parameter ‘use non-polar surface area’ was set to ‘True’, and CFF charges were used for
ligands. Dielectric constants used in calculations for water as a solvent and applied to the
entire space of the molecular system (protein) were equal to 80 and 4, respectively.
Selected poses were also scored with additional ten scoring functions: LigScore1 (CFF)
LigScore2 (CFF), PLP1, PLP2, Jain, PMF, PMF04, Ludi1, Ludi2 and Ludi3 (Supporting
Information, Tables 15S–17S).
This article is protected by copyright. All rights reserved.

Preparing CYP1A1, CYP1A2, CYP1B1 for
docking (Prepare Protein procedure, CHARMm
forcefield, radius of the binding sphere = 10 Å)
Sketching and preparing ligands
(trans-stilbene derivatives)
Docking by CDOCKER procedure
(CDOCKER energy, CDOCKER interaction energy)
Choosing ‘best fitted’ poses
(with the highest values of –CDOCKER energy)
Calculation of G
(PBSA implicit solvent model)
Additional scoring by 10 scoring functions
Analysis of ligand poses
(ligand-protein interactions)
Scheme 2: Flowchart: docking of trans-stilbene derivatives to CYP1A1, CYP1A2 and
CYP1B1 and post-docking analysis of protein-ligand complexes.
Results
Biological in vitro assay
In order to examine the affinity of synthesized methoxy/methylthio trans-stilbene derivatives
to CYP1A1, CYP1A2 and CYP1B1 active sites, the effect of the compounds on the CYP1-
dependent 7-ethoxyresorufin O-deethylase activity was evaluated. The inhibitory effect of
trans-stilbene derivatives on CYP1A1, CYP1A2 and CYP1B1 activities was determined
using the human recombinant cytochromes CYP1A1, CYP1A2 and CYP1B1with human
This article is protected by copyright. All rights reserved.

NADPH-P450 reductase. All studied compounds potently inhibited CYP1A1 and CYP1B1
activities and showed moderate inhibitory activity against CYP1A2 (Table 2). For a series of
tested compounds, the values of IC50 for CYP1A2 inhibition were the highest and the less
diversified (3.32 – 8.0 µM), while for CYP1A1 a 10-fold difference was observed (0.076
versus 0.756 µM). For CYP1B1 the effect of compounds S1–S4 and S6 ranged between
0.029 – 0.066 µM, while S5 strongly inhibited CYP1B1 at the nanomolar concentration.
Computational study
To find out the molecular basis of ligand interactions with CYP1 isozymes, a computational
study was employed. Binding free energies (ΔG) and interaction energies for the first ranked
poses (characterized by the highest –CDOCKER_ENERGY scores) of compounds S1–S6
docked to CYP1A1, CYP1A2 and CYP1B1 active sites were calculated (Table 2) and
interactions of the studied trans-stilbene derivatives with amino acid residues in binding sites
were analyzed.
To validate the docking procedure, the cocrystallized ligand -naphthoflavone (ANF)
was redocked into CYP1A1, CYP1A2 and CYP1B1 binding sites. For each cytochrome
a conformation of ANF with the highest –CDOCKER energy value was selected as the most
probable binding conformation and was compared with X-ray pose of ANF. RMSD for all
heavy atoms between experimental (X-ray) conformation and predicted pose of ligand was
found to be 0.44, 0.47 and 0.53 Å for CYP1A1, CYP1A2 and CYP1B1, respectively. For
docked poses binding free energies (G) were also calculated (Table 2).
In the molecular docking analysis ANF was used as a reference molecule. However,
the inhibitory potency of ANF toward CYP1A1 and CYP1A2 is evidently stronger,
methylthio-trans-stilbenes appeared to be more selective CYP1B1 inhibitors than ANF. For
the most potent CYP1B1 inhibitor in the series (S5), the ratios of inhibitory activities
This article is protected by copyright. All rights reserved.

CYP1A1/CYP1B1 and CYP1A2/CYP1B1 were 106 and 1450, respectively (the
corresponding values for ANF are 9.2 and 2.9). The inhibitory activities of S5 and ANF
towards CYP1B1 were of the same order of magnitude (0.0035 nM and 0.0013 nM,
respectively).
The experimentally determined inhibitory activity of the studied compounds against
CYP1A1 weakly correlated with the values of binding energy and interaction energy
calculated by the CDOCKER procedure (Table 2). To find out if the gaps in the CYP1A1
structure may have an impact on ligand affinity, we also docked the ligands to the structures
B, C and D presented by Walsh et al. (15). Slight differences in the conformations of the side
chains of amino acid residues in the enzyme cavity did not change the results of docking
(Table 21S in the Supporting Information).
All the studied ligands docked to CYP1A1 were oriented with 4’ substituent to the heme. The
substitution of oxygen with a sulphur atom in S2 caused only a slight shift of a molecule in
the enzyme cavity (Figure 5). However, the inhibitory potency of methylthio derivative S2
against CYP1A1 was 10-fold higher than its methoxy analogue S1. Moreover, the multiple
molecular non-bonded interactions of S2 with amino acids residues were observed. The 4’-
methylthio group of S2 participated in the π-sulphur, π-alkyl and alkyl-alkyl interactions with
the heme, as well as in alkyl-alkyl interactions with Val382, Leu496 and Ile386 (Figure 5).
CDOCKER interaction energies calculated for CYP1A2 were between -42.53 and -48.24
kcal/mol, while IC50 for all the studied compounds was between 3.5 and 8.4 µM. When
comparing CYP1A1 and CYP1A2 CDOCKER interaction energies calculated for particular
compound S1–S6, it is noticeable that for CYP1A1 the energies are higher, which is in
agreement with higher inhibitory activity. In the CYP1A2 binding site, ligands which possess
a 4’-substituent (S1–S4) are oriented with A ring towards the heme, while S5 and S6 with
This article is protected by copyright. All rights reserved.

3,4-dimethoxyphenyl demonstrate both orientations with ring A or B directed to the heme.
CDOCKER interaction energies for these two orientations are similar, while ΔG is more
differentiated and suggests that a preferable orientation is that with a 2-methoxy/2-methylthio
substituent directed towards the heme.
The most potent inhibitor of CYP1B1 activity, S5 was characterized by the highest
negative values of (CDOCKER) interaction energy and ΔG (Table 2) calculated for the first
ranked pose of the series of compounds. Interestingly, the CDOCKER interaction energy
values calculated for CYP1B1 did not differ significantly from the respective values for
CYP1A1, despite the fact that the IC50 values for CYP1B1 were more than 10 times lower
than that for CYP1A1. In comparison to CDOCKER interaction energy, ΔG values
demonstrated a better correlation with the differentiated inhibitory activity of compounds
against CYP1B1 and CYP1A1.
The first ranked poses of the ligands in the CYP1A1, CYP1A2 and CYP1B1 binding
sites are presented on the aligned images for S1–S4 and S5–S6, separately (Figure 6). In
CYP1A1 and CYP1B1 binding sites the ligands S5 and S6 were oriented with 3,4-
dimethoxyphenyl to the heme moiety in the way which guaranties the shortest distances
between both oxygen atoms of methoxy groups and the iron atom of heme (Supporting
Information, Table 22S).
Discussion
The results concerning the inhibitory potency and selectivity of the studied stilbenes
demonstrate variations highly dependent on the substitution pattern. We found S2 to be a
potent inhibitor of CYP1A1 with the IC50 value of 0.076 µM exhibiting similar potency to
natural stilbene rhapontigenin (3,3’,5-trihydroxy-4’-methoxy-trans-stilbene) that inhibited
This article is protected by copyright. All rights reserved.

human recombinant CYP1A1 with the IC50 value of 0.04 µM (19). Its high inhibitory activity
may be due to the multiple interactions with crucial amino acid residues. The SCH3 group
participates in the bonding with alkyl groups of amino acid residues and also with the heme;
between the heme and the methylthio group π-sulphur interaction may appear. The increased
radius of a sulphur atom in comparison to oxygen one and the larger number of electrons may
change the position of ligand in the active site effectively influencing the energy of
interaction between ligand and surrounding amino acid residues. However, the binding of S2
to CYP1A1 is not favorable energetically when compared with the other compounds of a
series. Because of a low correlation of CDOCKER interaction energy and other scoring
functions with the effect of stilbenes on CYP1A1 activity, we additionally employed the
LigandFit procedure (20). Nonetheless, the results obtained with this method of docking have
also been unsatisfying (Supporting Information, Table 18S). By contrast, for the orientations
of ligands in CYP1A2 and CYP1B1 binding sites selected on the basis of the highest
CDOCKER score, a correlation between inhibitory activity and ΔG as well as CDOCKER
interaction energy was found. In the case of CYP1A2, the values of LigScore2_CFF and
PMF scoring functions correlated well with the inhibitory activity of the tested compounds
(the three most active compounds were estimated high in the ranking). The most active
compound against CYP1B1, S5 was also the best scored ligand by LigScore1_CFF, PLP1
and Ludi1. For CYP1A1, an analogous correlation did not appear; conversely, none of the
scoring functions used for binding affinity estimation of the poses obtained in CDOCKER
run correlated with the inhibitory activity of the tested compounds; in LigandFit only PMF
scoring function well estimated the inhibitory potency of stilbenes (Supporting Information,
Tables 15S and 18S). Surprisingly, rhapontigenin, as a known potent CYP1A1 inhibitor
docked along with the studied stilbenes, was characterized by the lowest energy in the series
(binding energy -33.59 kcal/mol). Thus, none of the procedures used in our studies could
This article is protected by copyright. All rights reserved.

explain the selectivity of tested stilbenes against CYP1A1. The reason for the discrepancies
observed may be elucidated by the fact that CDOCKER and LigandFit do not provide the
plasticity of the binding center of an enzyme that in the case of CYP1A1 might play an
important role. Moreover, we must remember that in silico models currently used in studies
of ligand-cytochrome P450 interactions are based on approximation and do not include a
protein environment which may have a great influence on the effectiveness of biochemical
processes (21). For example, in the CYP1A1 cavity, the molecule of water can participate in
the ligand-enzyme interaction exerting an influence on the energetic parameters of ligand
binding.
Significantly higher IC50 values determined for the studied stilbenes docked to CYP1A2
are explained by the structure of the CYP1A2 active site that is the most narrow and flat
among CYP1s. Meta-analysis of the selectivity of CYP1 inhibitors revealed that stilbenes are
weak CYP1A2 inhibitors due to the shape and length of their molecules (22). The volumes of
active sites of CYP1A1, CYP1A2 and CYP1B1, which are 524, 375 and 398 Å3,
respectively, (15,16,17) and crucial amino acid residues in the enzyme cavity determine the
shape of the best fitted ligands. A triangle with side length 9.3, 8.7 and 7.2 Å was proposed as
a contour of selective CYP1A2 inhibitors (22). Stilbene derivatives better fit in the CYP1A1
binding site, exhibiting the shape of a long strip with the length of 12.3 Å characteristic of
efficient inhibitors of CYP1A1 activity (22). Moreover, amino acid residues in the vicinity to
the heme, Val382 in CYP1A1 and Val395 in CYP1B1, are replaced by the branched Leu382
in CYP1A2, which results in narrowing the CYP1A2 cavity. For this reason, the CDOCKER
procedure distinguished the orientations of phenyl with 2-methylthio and 2-methoxy
substituents towards the heme moiety in CYP1A2 cavity as favorable.
This article is protected by copyright. All rights reserved.

One of the most potent CYP1B1 inhibitors, S5 displayed a high selectivity for the
CYP1B1 inhibitor over CYP1A1, as demonstrated in our previous report (10). Until now, S5
binding to the CYP1B1 active site was analyzed with the LigandFit procedure, while in the
current studies we used the CDOCKER method to confirm the particular affinity of S5 to
CYP1B1. The values of binding free energy and interaction energy calculated by the
CDOCKER procedure for the first ranked pose of S5 were significantly higher than those of
other analogues. However, CDOCKER interaction energies calculated for CYP1A1 and
CYP1B1 for S5 were not significantly different: -51.60 and -52.32 kcal/mol, respectively. A
more significant difference was observed when ΔG were compared (-41.59 kcal/mol versus -
50.32 kcal/mol).
Molecular basis of inhibitory selectivity relies on the differences in the structure of
CYP1 binding sites namely the type of amino acid residues and their orientation in the active
site space. The presence of hydrophilic Gln332 in CYP1B1 instead of Phe319 in CYP1A1
might be the most prominent difference that influences the affinity of stilbenes to these
CYP1s. Atoms of oxygen or sulphur in the position 2 of ligands S1-S4 form the hydrogen
bonds with NH2 group of Gln332, whereas, S5 and S6 interact with Gln332 by means of π
electrons of stilbene phenyl ring, which participate as acceptors in the polar hydrogen-π
interaction. This type of bonds was described by Du et al. in his theoretical study (23).
For S1-S6 docked in CYP1B1 binding site there are interactions between peptide
bond Gly329/Ala330 and ligand π electrons. The π-amide stacking is considered as
meaningful when the interaction distance is less than 5 Å. For CYP1A1 the π-amide stacking
interaction distances were significantly greater than 5 Å. As we observed, interactions of
This article is protected by copyright. All rights reserved.

CYP1A1 with ligands were mostly hydrophobic in nature (e.g. π-π interactions with Phe224
and Phe258, π-alkyl interactions with Ala317, Leu312 and Leu496). Moreover, for S1 and
S2, weak hydrogen bonds between Asn228 and polarized CH3 group in position 2 (C-
H···O=C) occur in CYP1B1 binding site, whereas, in the CYP1A1 cavity the orientation of
Asn222 prevents this type of bonding. Additionally, for increased affinity of S5 and S6 which
are oriented with 3,4-substituted phenyl to the heme, the effect of slightly different position
of Leu509 in CYP1B1 comparing to Leu496 in CYP1A1 may play an important role
resulting in more space in the vicinity of the heme, that makes possible more effective T-
shaped π-π interactions.
It should be taken into consideration that the recognition process of the ligands in
CYP1 enzymes is dependent on two consecutive events, one involved with substrate access
channels, and the other event within the catalytic cavity (24). Additionally, the substrates and
inhibitors use different channels of CYP1 enzymes, depending on the size of molecules. The
efficiency of channel penetration by CYP1 ligands may influence their inhibitory potency to
a great extent, while the computational procedure used in our study analyzed only the
interactions of ligands in the binding sites of enzymes. Noticeably, experimentally
determined IC50 values reflect all determinants/factors that have an impact on the activity of
inhibitors including the fitting to the enzyme binding site, the specific and reciprocal
interactions between a ligand and the enzyme, and the energetic parameters.
This article is protected by copyright. All rights reserved.

Conclusions
1. The affinity of ligands to active sites of CYP1s is dependent primarily on the geometry
of the molecule, influenced by the number and position of substituents, which both
decide on the conformation of ligands in enzyme cavity.
2. The substitution of an oxygen atom with sulphur in trans-stilbene derivatives causes
an increase in the number of non-bonded interactions formed in the CYP1A1 active
site, although the effectiveness of SCH3/OCH3 substitution depends on substituent’s
position. A methylthio group in the position 2, not in the position 4, increased 10
times the affinity of the derivative to CYP1A1 binding site.
3. CDOCKER and LigandFit procedures achieved good results in determining the
affinity of compounds to CYP1A2 and CYP1B1, showing a satisfactory correlation of
scoring functions with the inhibitory activity of the studied stilbenes, although they
failed in the evaluation of the affinity of the tested stilbene derivatives to CYP1A1.
Acknowledgments The studies were supported by funding from Poznan University of
Medical Sciences number 502-01-03313427-08870.
Conflict of interest
All authors declare no financial/commercial conflict of interests.
This article is protected by copyright. All rights reserved.

References
1. Gajjar K., Martin-Hirsch P.L., Martin F.L. (2012) CYP1B1 and hormone-induced
cancer. Cancer Lett;324:13-30.
2. Kwon Y.-J., Baek H.-S., Ye D.-J., Shin S., Kim D., Chun Y.-J. (2016) CYP1B1
enhances cell proliferation and metastasis through induction of EMT and
activation of Wnt/β-catenin signalling via Sp1 upregulation. PLoS
ONE;11:e0151598
3. Cui J., Meng Q., Zhang X., Cui Q., Zhou W., Li S. (2015) Design and synthesis of
new α-naphthoflavones as cytochrome P450 (CYP) 1B1 inhibitors to overcome
docetaxel-resistance associated with CYP1B1 overexpression. J Med
Chem;58:3534-3547.
4. Bruno R.D., Njar V.C.O. (2007) Targeting cytochrome P450 enzymes: a new
approach in anti-cancer drug development. Bioorg Med Chem;15:5047-5060.
5. Chang T.K.H., Lee W.B., Ko H.H. (2000) Trans-resveratrol modulates the
catalytic activity and m-RNA expression of the procarcinogen-activating human
cytochrome P450 1B1. Can J Physiol Pharmacol;78:874-881.
6. Walle T., Hsieh F., DeLegge M.H., Oatis J.E., Walle U.K. (2004) High absorption
but very low bioavailability of oral resveratrol in humans. Drug Metab
Dispos;32:1377-1382.
7. Chun Y.-J., Kim S., Kim D., Lee S.-K., Guengerich F.P. (2001) A new selective
and potent inhibitor of human cytochrome P450 1B1 and its application to
antimutagenesis. Cancer Res;61:8164-8170.
8. Chun Y.-J., Lim C., Ohk S.O., Lee J.M., Lee J.H., Choi S., Kim S. (2011) trans-
Stilbenoids: potent and selective inhibitors for human cytochrome P450 1B1.
MedChemComm;2:402-405.
This article is protected by copyright. All rights reserved.

9. Chun Y.-J., Oh Y.-K., Kim B.J., Kim D., Kim S.S., Choi H.-K., Kim M.-Y.
(2009) Potent inhibition of human cytochrome P450 1B1 by tetramethoxystilbene.
Toxicol Lett;189:84-89.
10. Mikstacka R., Wierzchowski M., Dutkiewicz Z., Gielara-Korzańska A., Korzański
A., Teubert A., Sobiak S., Baer-Dubowska W. (2014) 3,4,2’-Trimetoxy-trans-
stilbene – a potent CYP1B1 inhibitor. MedChemComm;5:496-501.
11. Mikstacka R., Rimando A.M., Dutkiewicz Z., Stefański T., Sobiak S. (2012)
Design, synthesis and evaluation of the inhibitory selectivity of novel trans-
resveratrol analogues on human recombinant CYP1A1, CYP1A2 and CYP1B1.
Bioorg Med Chem;20:5117-5126.
12. Murias M., Handler N., Erker T., Pleban K., Ecker G., Saiko P., Szekeres T., Jäger
W. (2004) Resveratrol analogues as selective cyclooxygenase-2 inhibitors:
synthesis and structure-activity relationship. Bioorg Med Chem;12:5571- 5578.
13. Mikstacka R., Baer-Dubowska W., Wieczorek M., Sobiak S. (2008)
Thiomethylstilbenes as inhibitors of CYP1A1, CYP1A2 and CYP1B1 activities.
Mol Nutr Food Res;52:S77-S83.
14. Burke M.D., Thompson S., Elcombe C.R., Halpert J., Haaparanta T., Mayer R.T.
(1985) Ethoxy-, pentoxy-, and benzyloxyphenoxazones and homologues: A series
of substrates to distinguish between different induced cytochromes P-450.
Biochem Pharmacol;34:3337-3345.
15. Walsh A.A., Szklarz G.D., Scott E.E. (2013) Human cytochrome P450 1A1
structure and utility in understanding drug and xenobiotic metabolism. J Biol
Chem;288:12932-12943.
16. Sansen S., Yano J.K., Reynald R.L., Schoch G.A., Griffin K.J., Stout C.D.,
Johnson E.F. (2007) Adaptations for the oxidation of polycyclic aromatic
This article is protected by copyright. All rights reserved.

hydrocarbons exhibited by the structure of human P450 1A2. J Biol
Chem;282:14348-14355.
17. Wang A., Savas U., Stout C.D., Johnson E.F. (2011) Structural characterization of
the complex between α-naphthoflavone and human cytochrome P450 1B1. J Biol
Chem;286:5736-5743.
18. Wu G., Robertson D.H., Brooks III C.L., Vieth M. (2003) Detailed analysis of
grid-based molecular docking: A case study of CDOCKER–A CHARMm-based
MD docking algorithm. J Comp Chem;24:1549-1562.
19. Chun Y.J., Ryu S.Y., Jeong T.C., Kim M.Y. (2001) Mechanism-based inhibition
of human cytochrome P450 1A1 by rhapontigenin. Drug Metab Dispos;29:389-
393.
20. Venkatachalam C.M., Jiang X., Oldfield T., Waldman M. (2003) LigandFit: a
novel method for the shape-directed rapid docking of ligands to protein active
sites. J Mol Graph Model;21:289-307.
21. Olsen L., Oostenbrink C., Jorgensen F.S. (2015) Prediction of cytochrome P450
mediated metabolism. Adv Drug Deliver Rev;86:61-71.
22. Liu J., Sridhar J., Foroozesh M. (2013) Cytochrome P450 Family 1 inhibitors and
structure-activity relationship. Molecules;18:14470-14495.
23. Du Q.-S., Wang Q.-Y., Du L.-Q., Chen D., Huang R.-B. (2013) Theoretical study
on the polar hydrogen-π (Hp- π) interactions between protein side chains. Chem
Cent J;7:92
24. Urban P., Truan G., Pompon D. (2015) Access channels to the buried active site
control substrate specificity in CYP1 P450 enzymes. Biochim Biophys
Acta;1850:696-707.
This article is protected by copyright. All rights reserved.

Notes
a Accelrys Software Inc., Discovery Studio Modeling Environment, Release 4.0, San Diego:
Accelrys Software Inc., 2014.
Supporting Information
Additional Supporting Information may be found in the online version of this article:
Appendix 1S. Synthesis and chemical characterizations of compounds.
Appendix 2S. Crystal data for compounds S1 and S6.
Table 1S. Crystal data and structure refinement for S1.
Table 2S. Atomic coordinates (× 104) and equivalent isotropic displacement parameters (Å2 x
103) for S1.
Table 3S. Bond lengths [Å] and angles [deg] for S1.
Table 4S. Anisotropic displacement parameters (Å2 × 103) for S1.
Table 5S. Hydrogen coordinates (× 104) and isotropic displacement parameters (A2 × 103) for
S1.
Table 6S. Torsion angles [deg] for S1.
Table 7S. Hydrogen-bonds for S1 [Å and deg].
Table 8S. Crystal data and structure refinement for S6.
4
2
Table 9S. Atomic coordinates (× 10 ) and equivalent isotropic displacement parameters (Å
3
× 10 ) for S6.
Table 10S. Bond lengths [Å] and angles [deg] for S6.
2
3
Table 11S. Anisotropic displacement parameters (Å × 10 ) for S6.
4
2
3
Table 12S. Hydrogen coordinates (× 10 ) and isotropic displacement parameters (Å × 10 )
for S6.
Table 13S. Torsion angles [deg] for S6.
Table 14S. Hydrogen-bonds for S6 [Å and deg].
This article is protected by copyright. All rights reserved.

Figure 1S. Anisotropic ellipsoid representation of 2,4'-dimetoksystilben (S1) , showing the
atom-labelling scheme. Displacement ellipsoids are drawn at the 50% probability level and H
atoms are shown as small spheres of arbitrary radii.
Figure 2S. The crystal packing of S1 as seen along [001] direction.
Figure 3S. Anisotropic ellipsoid representation of 2-methylthio-3’,4’-dimethoxy-trans-
stilbene S6, showing the atom-labelling scheme. Displacement ellipsoids are drawn at the
50% probability level and H atoms are shown as small spheres of arbitrary radii.
Figure 4S. The crystal packing of S6 as seen along [010] direction.
Figure 5S. The crystal packing of S6 as seen along [101] direction.
Appendix 3S. 1H, 13C NMR, 1H-1H COSY, 1H-13C HSQC and 1H-13C HMBC spectra of the
synthesized compounds.
Figure 6S. Annotated signals of S1 1H and 13C NMR spectra according to 2D experiments
(1H-1H COSY, 1H-13C HSQC and 1H-13C HMBC).
Figure 7S. Annotated signals of S3 1H and 13C NMR spectra according to 2D experiments
(1H-1H COSY, 1H-13C HSQC and 1H-13C HMBC).
Figure 8S. Annotated signals of S4 1H and 13C NMR spectra according to 2D experiments
(1H-1H COSY, 1H-13C HSQC and 1H-13C HMBC).
Appendix 4S. Results of docking.
Table 15S. Estimated binding energies (G), CDOCKER interaction energies (IE) and
values of scoring functions for top scored poses of S1–S6 docked to CYP1A1 binding site
with CDOCKER procedure.
Table 16S. Estimated binding energies (G), CDOCKER interaction energies (IE) and
values of scoring functions for top scored poses of S1–S6 docked to CYP1A2 binding site
with CDOCKER procedure.
Table 17S. Estimated binding energies (G), CDOCKER interaction energies (IE) and
values of scoring functions for top scored poses of S1–S6 docked to CYP1B1 binding site
with CDOCKER procedure.
Table 18S. Estimated binding energies (G) and values of scoring functions for top scored
poses of S1–S6 docked to CYP1A1 binding site with LigandFit procedure.
Table 19S. Estimated binding energies (G) and values of scoring functions for top scored
poses of S1–S6 docked to CYP1A2 binding site with LigandFit procedure.
Table 20S. Estimated binding energies (G) and values of scoring functions for top scored
poses of S1–S6 docked to CYP1B1 binding site with LigandFit procedure.
This article is protected by copyright. All rights reserved.

Table 21S. Values of CDOCKER interaction energies (IE) for top scored poses of docked
ligands in CYP1A1 active site of chains A, B, C and D.
Table 22S. Distances (in Ǻngström) between O atom of methoxy substituent in 4’ or 3’
position and Fe atom in heme for the top scored poses in CYP1A1, CYP1A2 and CYP1B1
active sites.
Table 1: Positions of substituents and the reaction yield of compounds S1–S6. Data with
asterisks taken from earlier described procedures. (*(13); **(10)).
Substituent position
Compound
Yield [%]
R
R1
R2
S1
S2
S3
S4
S5
S6
4’-OCH3
4’-SCH3
4’-OCH3
4’-SCH3
4’-OCH3
4’-OCH3
H
OCH3
OCH3
SCH3
SCH3
OCH3
SCH3
80
H
H
41*
35
H
28
3’-OCH3
3’-OCH3
32**
24
This article is protected by copyright. All rights reserved.

Table 2: Inhibitory effect of methoxy/methylthio trans-stilbene derivatives and ANF on
CYP1A1, CYP1A2 and CYP1B1 activities and values of CDOCKER interaction energies
(IE) and estimated binding free energies (ΔG)
CYP1A1
CYP1A2
CYP1B1
Compound
IC50
IE
IC50
IE
IC50
IE
G
G
G
[kcal/mol]
[kcal/mol]
[kcal/mol]
[M]
[kcal/mol]
[M]
[kcal/mol]
[M]
[kcal/mol]
S1
0.75
(0.64-
0.90)
–44.93
–45.14
–43.90
–44.08
–51.60
–50.56
–43.98
–36.67
–37.45
–36.91
–36.49
–41.59
–40.01
–39.13
4.13
(3.58-
4.78)
–44.72
–43.05
–44.35
–42.53
–48.24
–46.27
–47.61
–38.00
–35.81
–36.86
–35.22
–43.93
–41.67
–41.90
0.066
(0.057-
0.076)
–45.55
–45.60
–44.53
–45.01
–52.32
–51.40
–47.40
–44.60
–44.01
–42.33
–42.07
–50.32
–48.87
–46.66
S2
0.076
(0.067-
0.085)
8.38
(6.39-
11.0)
0.034
(0.029-
0.040)
S3
0.49
(0.40-
0.60)
7.69
(6.85-
8.62)
0.028
(0.022-
0.035)
S4
0.60
(0.54-
0.65)
8.0
(5.3-
12.1)
0.026
(0.023-
0.029)
S5
0.36
(0.27-
0.48)
4.93
(4.43-
5.50)
0.0034
(0.0016-
0.0072)
S6
0.76
(0.68-
0.84)
3.54
(3.11-
4.04)
0.029
(0.022-
0.039)
ANF*)
0.012
0.0038
0.0013
*) IC50 values taken from [3]
Figure legend
Figure 1: Anisotropic ellipsoid representation of 2-methylthio-3’,4’-dimethoxy-trans-
stilbene S6 (left) and 2,4’-dimethoxy-trans-stilbene S1 (right), showing the atom-labelling
scheme. Displacement ellipsoids are drawn at the 50% probability level and H atoms are
shown as small spheres of arbitrary radii.
Figure 2: The crystal packing of S5 (left) and S6 (right) as seen along [010] direction.
Figure 3: The crystal packing of S5 (left) as seen along [100] and S6 (right) as seen along
[101] direction.
This article is protected by copyright. All rights reserved.

Figure 4: Annotated Nuclear Magnetic Resonance signals of molecule S6. 1H (bold) and 13C
NMR spectra according to 2D experiments (1H-1H COSY, 1H-13C HSQC and 1H-13C
HMBC). Abbreviations s – singlet, d – doublet, t – triplet, o – overlapped, h – hidden.
Figure 5: Non-bonded interactions of S1 (a) and S2 (b) with CYP1A1 binding site. Heme is
represented as a stick model in red. Interaction types: alkyl-alkyl and -alkyl (grey dashed
lines), - (magenta dashed lines) and -sulphur (yellow dashed lines).
Figure 6: Orientations of the ligands towards the heme: S1–S4 in binding sites of CYP1A1
(a), CYP1A2 (c), CYP1B1 (e) and S5–S6 in binding sites of CYP1A1 (b), CYP1A2 (d),
CYP1B1 (f). S5 – carbon atoms in black, S6 – carbon atoms in magenta.
This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.