Parallel in vitro and in silico investigations into anti-inflammatory effects of non-prenylated stilbenoids

Article abstract of DOI:10.1016/j.foodchem.2019.01.128

Stilbenoids represent a large group of bioactive compounds, which occur in food and medicinal plants. Twenty-five stilbenoids were screened in vitro for their ability to inhibit COX-1, COX-2 and 5-LOX. Piceatannol and pinostilbene showed activity comparable to the zileuton and ibuprofen, respectively. The anti-inflammatory potential of stilbenoids was further evaluated using THP-1 human monocytic leukemia cell line. Tests of the cytotoxicity on the THP-1 and HCT116 cell lines showed very low toxic effects. The tested stilbenoids were evaluated for their ability to attenuate the LPS-stimulated activation of NF-κB/AP-1. Most of the tested substances reduced the activity of NF-κB/AP-1 and later attenuated the expression of TNF-α. The effects of selected stilbenoids were further investigated on inflammatory signaling pathways. Non-prenylated stilbenoids regulated attenuation of NF-?B/AP-1 activity upstream by inhibiting the phosphorylation of MAPKs. A docking study used to in silico analyze the tested compounds confirmed their interaction with NF-?B, COX-2 and 5-LOX.

Full text of DOI:10.1016/j.foodchem.2019.01.128

Accepted Manuscript
Analytical Methods
Parallel in vitro and in silico investigations into anti-inflammatory effects of
non-prenylated stilbenoids
Veronika Leláková, Karel Šmejkal, Karolina Jakubczyk, Ondřej Veselý,
Přemysl Landa, Jiř í Václavík, Pavel Bobá ľ, Hana Pí žová, Veronika Temml,
Theresa Steinacher, Daniela Schuster, Sebastian Granica, Zuzana Hanáková,
Jan Hošek
PII:
S0308-8146(19)30203-1
DOI:
https://doi.org/10.1016/j.foodchem.2019.01.128
FOCH 24223
Reference:
To appear in:
Food Chemistry
Received Date:
Revised Date:
Accepted Date:
25 May 2018
7 December 2018
25 January 2019
Please cite this article as: Leláková, V., Šmejkal, K., Jakubczyk, K., Veselý, O., Landa, P., Václavík, J., Bobá ľ, P.,
Pí žová, H., Temml, V., Steinacher, T., Schuster, D., Granica, S., Hanáková, Z., Hošek, J., Parallel in vitro and in
silico investigations into anti-inflammatory effects of non-prenylated stilbenoids, Food Chemistry (2019), doi:
https://doi.org/10.1016/j.foodchem.2019.01.128
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Parallel in vitro and in silico investigations into anti-
inflammatory effects of non-prenylated stilbenoids
a,b
a,*
d
d,e
d
Veronika Leláková , Karel Šmejkal , Karolina Jakubczyk , OndřejVeselý , Přemysl Landa ,
a
c
c
f
f
Jiří Václavík , Pavel Bobáľ , Hana Pížová , Veronika Temml , Theresa Steinacher , Daniela
f
g
a
b
Schuster , Sebastian Granica , Zuzana Hanáková , Jan Hošek
a
b
Department of Natural Drugs, Department of Molecular Biology and Pharmaceutical
c
Biotechnology, Department of Chemical Drugs, Faculty of Pharmacy, University of Veterinary
and Pharmaceutical Sciences Brno, Palackého třída 1946/1, 61242 Brno, Czech Republic.
dLaboratory of Plant Biotechnologies, Institute of Experimental Botany, Czech Academy of
Sciences, Rozvojová 263, 16502 Prague, Czech Republic.
eDepartment of Quality of Agricultural Products, Faculty of Agrobiology, Food and Natural
Resources, The Czech University of Life Sciences Prague, Kamýcka129, 165 21 Prague 6 –
Suchdol, Czech Republic.
fDepartment of Pharmacy/Pharmacognosy and Center of Molecular Biosciences (CMBI),
University of Innsbruck, Innrain 80-82, 6020 Innsbruck, Austria.
gDepartment of Pharmacognosy and Molecular Basis of Phytotherapy, Faculty of Pharmacy,
Medical University of Warsaw, Banacha 1, 02-097, Warsaw, Poland.
*
corresponding author
1

KEYWORDS
Anti-inflammatory; cyclooxygenase; molecular docking simulations; lipoxygenase; macrophages;
NF-κB; stilbenes
ABSTRACT
Stilbenoids represent a large group of bioactive compounds, which occur in food and medicinal
plants.
Twenty-five stilbenoids were screened in vitro for their ability to inhibit COX-1, COX-2 and 5-
LOX. Piceatannol and pinostilbene showed activity comparable to the zileuton and ibuprofen,
respectively.
The anti-inflammatory potential of stilbenoids was further evaluated using THP-1 human
monocytic leukemia cell line. Tests of the cytotoxicity on the THP-1 and HCT116 cell lines
showed very low toxic effects. The tested stilbenoids were evaluated for their ability to attenuate
the LPS-stimulated activation of NF-κB/AP-1. Most of the tested substances reduced the activity
of NF-κB/AP-1 and later attenuated the expression of TNF-α.
The effects of selected stilbenoids were further investigated on inflammatory signaling pathways.
Non-prenylated stilbenoids regulated attenuation of NF-ĸB/AP-1 activity upstream by inhibiting
the phosphorylation of MAPKs.
A docking study used to in silico analyze the tested compounds confirmed their interaction with
NF-ĸB, COX-2 and 5-LOX.
2

1
. Introduction
Stilbenoids form a wide group of naturally occurring compounds, the best-known member of
which is trans-resveratrol (1) (Giacomini, Rupiani, Guidotti, Recanatini, & Roberti, 2016; Shen,
Wang, & Lou, 2009). Stilbenoids are phenolic compounds displaying two aromatic rings linked
by an ethylene bridge, structurally characterized by the presence of a 1,2-diarylethylene nucleus
with hydroxyl substitution on the aromatic rings (Fresco, Borges, Diniz, & Marques, 2006). In
some plants, including food and medicinal plants, they are very variable substituted, mainly with
hydroxy, methoxy, methyl, or prenyl groups. The diarylethylene core can be further modified to
form 2-arylbenzofurans or lactone ring containing phtalides.
Stilbenoids are distributed in higher plants in the form of monomers or oligomers, and are
commonly present as glycosides (Han, Shen, & Lou, 2007). Trans-resveratrol (1), the most
investigated stilbene, is present in a variety of berries, peanuts, and medicinal plants together with
its glycoside piceid in common grapes (Vitis vinifera), and particularly in grapes infected by
Botrytis cinerea (used for production of Tokaji wines), or in grapes of V. rotundifolia used for
manufacturing of muscadine wine. The high content of resveratrol (1) can be found also in peanut
butter and dark chocolate (Baur & Sinclair, 2006). Pterostilbene (2) is primarily originated from
blueberries (Vaccinium spp.), its content can vary depending to the type of berry (McCormack &
McFadden, 2013). Moreover, the presence of stilbene derivatives is also observed in many
medicinal plants (Sut, Maggi, & Dall'Acqua, 2017; Xiao, Zhang, Xuan, Zhang, Xu, & Bai, 2008).
After consumption and absorption into blood stream, stilbenoids interact with different plasma and
body proteins and this interaction could influence their pharmacokinetic profile and biological
effect (Cao, Shi, Jia, Xiao, & Chen, 2016; Tang, Xie, Cao, Yang, Chen, & Xiao, 2017). There are
3

numerous studies reporting the positive effects of stilbenoids supplemented as dietary agents or
nutraceuticals, mostly on cardiovascular health and cancer chemoprevention (Chachay,
Kirkpatrick, Hickman, Ferguson, Prins, & Martin, 2011). The presence of stilbenoids in food and
medicinal plants is therefore associated with the multiple health benefits, including anti-
inflammatory activity.
Inflammation is a natural reaction of an organism, generally directed against any type of damage,
such as infection, irritation, or injury. It is characterized more or less specifically by activation of
the immune system, which leads to vasodilatation and increased blood flow, release of
inflammatory mediators, elevated cellular metabolism and the influx of various immune-system
cells into the tissues (Brezáni, Šmejkal, Hošek, & Tomášová, 2018; Medzhitov, 2008). Upon
activation, cells involved in the inflammatory reaction produce and release various inflammatory
mediators to control the reaction of the organism. Under normal circumstances, the reaction is
limited to the affected tissue and is strictly controlled. In some cases, connected with so-called
civilization diseases, the inflammation expands from the acute phase, and becomes uncontrolled
and continuous, switching into a chronic phase (Murakami & Hirano, 2012). Well-known
examples of diseases associated with such chronic inflammation include diabetes mellitus (type 2),
certain types of cancer, inflammatory bowel diseases, and some neurodegenerative disorders like
Alzheimer’s disease and Parkinson’s disease (Bontempo et al., 2015; Chen, Zhang, & Huang,
2
016; Khansari, Shakiba, & Mahmoudi, 2009; Reuter, Gupta, Chaturvedi, & Aggarwal, 2010).
Consumption of a well-balanced diet rich in natural phenolic compounds, such as stilbenoids,
flavonoids, anthocyanins and others, can represent an approach resulting in health-promoting
effects and prevention against different types of inflammatory and cancer diseases (Bontempo et
al., 2015; Pandey & Rizvi, 2009).
4

There is no certain way to cure chronic inflammation. Several pharmacotherapeutic methods act
on different levels. In general, non-steroidal anti-inflammatory drugs (NSAIDs) can be
distinguished from steroids (corticosteroids), and so-called biological treatments such as TNF-α
antibodies are also available (Tabas & Glass, 2013). Anti-inflammatory therapy based on NSAIDs
involves inhibiting enzymes of the arachidonic acid metabolic pathway (cyclooxygenase, COX;
lipoxygenase, LOX; phospholipase A, PLA), or reducing the production of various pro-
inflammatory proteins (cytokines), like IL-1β, IL-6 or TNF-α, whose expression is controlled by
NF-ĸB (Nuclear Factor kappa-light-chain-enhancer of activated B cells) transcriptional factor
(Golan-Goldhirsh & Gopas, 2014). Anti-inflammatory treatment is usually associated with
problems caused by side effects of the drugs used in long-termed therapeutic schemes or with
pleiotropic effects of some drugs. Scientists are therefore looking for new compounds with anti-
inflammatory potential, which target different cellular proteins, such as transcriptional factors
responsible for cytokines release and which intervene into their signaling pathways. The inhibition
of the release of enzymes and mediators from neutrophils, macrophages and mast cells, inhibition
of inducible nitric oxide synthase (iNOS) expression, nitric oxide (NO) and reactive oxygen
species (ROS) release represent also possible targets of new anti-inflammatory molecules (Brezáni
et al., 2018; Krakauer, 2004; Laveti et al., 2013).
Information about the anti-inflammatory activity of stilbenoids is available, especially as it
concerns trans-resveratrol (1), but it is well-nigh impossible to compare the activities of
compounds using different assays (Dvorakova & Landa, 2017). For this reason, we systematically
analyzed the ability of 25 stilbenoids to inhibit COX-1, COX-2, 5-LOX and the production of pro-
inflammatory cytokines in lipopolysaccharide (LPS)-activated THP-1 macrophages. These effects
5

were also evaluated in silico by molecular docking simulations of the ability of tested compounds
to interact with active sites of COX-2 and 5-LOX, and to NF-κB.
2
2
. Experimental Section
.1. General Experimental Procedures
Twenty-five tested stilbenoids (1–25) (Figure 1) were obtained commercially (1–7, 12–19, 22–
3), isolated from plant sources (8–11, 24, 25) or synthesized by our group (20, 21). Most of them
2
are naturally occurring compounds (1–11, 15, 17, 18, 20, 22–25) (Dvorakova & Landa, 2017;
Granica et al., 2015; Li et al., 2015; Moodie et al., 2017; Pan & Lungren, 1995; Weng et al., 2016;
Woo, Min, & Ryoo, 2010; Xiao et al., 2008), the resting compounds are of a synthetic origin (12–
1
4, 16, 19, 21). Compounds 9–11, 24 and 25 are atypical compounds obtained from Tragopogon
tommasinii Sch.Bip. (Asteraceae, Cichorieae) (Granica et al., 2015), oxyresveratrol (8) was
isolated from Morus alba L. (Moraceae) root bark and identified by comparison of its spectral data
with literature (Qiu et al., 1996). Reagents and solvents used for isolation and synthesis were
purchased from Sigma-Aldrich (Steinheim, Germany) and used as received, if not stated otherwise.
Anhydrous solvents and reagents were absolutized as usual and distilled prior to use. The progress
of the reactions and the purity of the intermediates were monitored using TLC plates with a normal
phase silica gel 60 F254 (Merck, Darmstadt, Germany).
R1
R2
R3
R4
R5
R6
6

Trans-resveratrol (1)
Pterostilbene (2)
Piceatannol (3)
OH
OCH
OH
H
H
H
H
H
H
OH
OH
OH
OH
OH
OCH
OH
H
H
H
H
3
3
3
OH
H
Pinostilbene (4)
OCH
OH
Piceatannol-3´-O-β-
glucopyranoside (5)
OH
H
O-Glc
OH
OH
H
Thunalbene (7)
Oxyresveratrol (8)
OCH
OH
H
3
H
OH
H
OH
H
H
OH
H
OH
OH
H
H
H
Trans-stilbene (12)
Cis-stilbene (13)
H
H
H
H
H
H
H
H
4
-Stilbenecarboxylic acid (14)
Pinostilbenoside (15)
H
H
H
COOH
O-Glc
H
H
H
OCH
H
3
3
H
H
OH
H
H
Trans-α-methylstilbene (16)
H
H
CH
3
Pinosylvin-monomethylether
OCH
OH
H
H
H
H
OH
OH
H
H
H
(
17)
Isorhapontigenin (18)
OCH
H
3
OH
2
,4,3´,5´-tetramethoxystilbene
OCH
3
3
OCH
3
OCH
OCH
3
3
3
OCH
3
3
(
19)
3
,4´,5-trimethoxy-trans-
stilbene (20)
OCH
H
H
OCH
H
3
´-bromo-3,4´,5-trimethoxy-
trans-stilbene (21)
OCH
OCH
3
3
H
H
Br
H
OCH
H
OCH
OCH
3
3
H
H
3
,5-dimethoxystilbene (22)
1
-(2,4-dihydroxyphenyl)-2-(4-hydroxyphenyl)-ethanone (23)
7

R1
R2
R3
Batatasin III (6)
H
OH
H
2
-carboxyl-3-O-methyl-4´-β-D-glucopyranosyl-
dihydroresveratrol (10)
COOH
H
O-Glc
R1
R2
7
-β-glucopyranosyloxy-3-(4-hydroxybenzyl)-5-methoxyphtalide (9)
H
OH
7
-(1→6)-α-rhamnosyl-β-glucopyranosyloxy-3-(4-hydroxybenzyl)-5-
methoxyphtalide (11)
Rha
OCH
3
R
3
-O-caffeoyl-(9→5)-β-apiosyl-(1→6)-β-glucopyranosyl-3´-O-
methyldihydropiceatannol (24)
OH
3
-O-caffeoyl-(9→5)-β-apiosyl-(1→6)-β-glucopyranosyl-5,3´-O-
dimethyldihydropiceatannol (25)
OCH
3
8

Figure 1. Structures of the test compounds 1–25.
Melting points were measured using a Böetius apparatus (Franz KüstnerNachf. KG, Germany) and
were not corrected. IR spectra were recorded on a SmartMIRacle ATR Zn/Se for Nicolet Impact
4
10 FT-IR spectrometer (Thermo Scientific Corp., MA, USA). NMR spectra were collected on a
JEOL ECZR-400 MHz instrument (Jeol Corp., Japan). All of the NMR measurements were done
at 25 °C. Chemical shifts δ are reported in parts per million (ppm) and J values in Hz. The signal
of TMS or the residual solvent signals of CDCl
3
or DMSO-d were used as an internal reference.
6
We have used the modified procedures described by Davis & Groshens (2012) to prepare
trimethoxystilbenes, as shown in Figure 2. The total syntheses of 3,4´,5-trimethoxy-trans-stilbene
(
20) and 3´-bromo-3,4´,5-trimethoxy-trans-stilbene (21) were carried out in a four-step linear
sequence to provide the products in satisfactory overall yields. Direct reduction of
,5-dimethoxybenzoic acid by sodium bis(2-methoxyethoxy)aluminumhydride (SMEAH) in
3
toluene yielded benzyl alcohol. Chlorination of the alcohol with thionyl chloride in diethyl ether
provided an intermediate which upon Michaelis–Arbuzov reaction with triethylphosphite yielded
the key phosphonate. Finally, Horner–Wadsworth–Emmons reaction of the intermediate with
4
-methoxybenzaldehyde or 3-bromo-4-methoxybenzaldehyde using sodium tert-pentoxide as a
base produced 3,4´,5-trimethoxy-trans-stilbene (20) or 3´-bromo-3,4´,5-trimethoxy-trans-stilbene
21) in 32 % and 54 % yields, respectively.
(
9

H CO
H CO
H CO
3
3
3
OH
Cl
SOCl2
SMEAH
MePh
COOH
DMFcat, Et O
2
H CO
H CO
H CO
3
3
3
P(OEt)3
KI
R
R
H CO
CHO
3
H CO
H CO
O
P
3
3
OCH CH3
2
OCH3
6
X = H, 7 X = Br
OCH CH3
2
t-PentONa
THF
H CO
H CO
3
3
2
2
0 R = H
1 R = Br
Figure 2. The total synthesis of (E)-3,4´,5-trimethoxystilbene (20) and (E)-3´-bromo-3,4´,5-
trimethoxystilbene (21).
2
.2. COX-1 and COX-2 Assays
An enzymatic in vitro assay was used to test inhibitory activity of the stilbenoids. COX-1 (1
unit/reaction) from ram seminal vesicles and human recombinant COX-2 (0.5 unit/reaction; both
from Sigma-Aldrich, USA) were added to 180 µL of a mixture consisting of 100 mM Tris buffer
(
pH 8.0), 5 μM porcine hematin, 18 mM L-epinephrine, and 50 μM Na EDTA. Tested compounds
2
were dissolved in DMSO (pure DMSO as blank) to obtain 20 µM concentration for an initial
screening. Subsequently, 0.032, 0.16, 0.8, and 4 µM concentrations were prepared for selected
compounds. Ten microliters of the compounds to be tested were incubated with reaction mixture
for 5 min at room temperature. The reaction was then initiated by adding 5 µL of arachidonic acid
(
10 μM) and the mixture was incubated for 20 min at 37 °C. The reaction was stopped by adding
0 µL of formic acid (10% v/v). (S)-(+)-ibuprofen (Sigma-Aldrich, USA) was used as a reference
inhibitor. The inhibition of COX enzymes was evaluated as the potential to reduce the synthesis
2
1
0

of prostaglandin E
NY, USA). Samples were diluted 1:15 in the assay buffer provided in the kit, before being used
with the kit according to the manufacturer’s instructions. The concentration of PGE was estimated
according to the absorbance measured at 405 nm by a Tecan Infinite M200 microplate reader
Tecan Group, Switzerland). The inhibitory activity was expressed as percent inhibition compared
2
(PGE
2
), quantified using a Prostaglandin E
2
ELISA kit (Enzo Life Sciences,
2
(
to the blank. The experiments were repeated at least two times with at least two replicates in each
test.
2
.3. 5-LOX Assay
In case of the LOX assays, the inhibitory potential of each stilbenoid was tested using an
enzymatic in vitro assay. Human recombinant 5-LOX (Cayman Chemical, MI, USA) was added
1 unit/reaction) to 180 μL of an incubation mixture consisting of phosphate buffer saline (pH 7.4),
mM Na EDTA, and 200 μM ATP. Tested compounds were dissolved in DMSO (pure DMSO as
(
1
2
blank) to obtain 5 µM concentration for initial screening. Subsequently, 0.008, 0.04, 0.2, and 1 µM
concentrations were prepared for selected compounds. Ten microliters of the compounds to be
tested were incubated with reaction mixture for 10 min at 4°C. Then, 5 μL of 80 mM CaCl was
2
added, the reaction was initiated by adding 5 µL of arachidonic acid (800 μM) and the mixture
was incubated for 10 min at 37 °C. The reaction was stopped by adding 20 µL of formic acid (10%
v/v). Zileuton (Farmak a.s., Czech Republic) was used as a reference inhibitor. Inhibition of the 5-
LOX enzyme was evaluated as the potential to reduce the synthesis of leukotriene B
4
(LTB ).
4
LTB was quantified using a Leukotriene B ELISA kit (Enzo Life Sciences). Samples were
4
4
diluted 1:15 in the assay buffer provided in kit before being used with the kit according to the
manufacturer’s instructions. The measurement of the absorbance and the determination of the
inhibitory activity were the same as for the COX assays.
1
1

2
.4. Cell Culture and Differentiation to Macrophages
The THP-1 human monocytic leukemia cell line and HCT116 colon carcinoma cell line were
purchased from the European Collection of Authenticated Cell Cultures (Salisbury, UK). The
TM
THP-1-XBlue -MD2-CD14 cell line was supplied by Invivogen (San Diego, CA, USA). THP-1
TM
and THP-1-XBlue -MD2-CD14 cells were cultured in RPMI 1640 medium containing stabilized
2
mM L-glutamine (Biosera, France), whereas HCT116 cells were cultured in DMEM high glucose
medium w/L-glutamine and w/sodium pyruvate (Biosera, France), both supplemented with
antibiotics [100 U/mL penicillin and 100 mg/mL streptomycin (Biosera, France)], and 10% FBS
(HyClone, UT, USA). Cells were kept in an incubator at 37 °C in a water-saturated atmosphere of
TM
air containing 5% CO . The suspensions of THP-1 and THP-1-XBlue -MD2-CD14 cells were
2
passaged approximately twice a week, adherent HCT116 cells were passaged in 2–3 days intervals
when the culture reached 80-90 % of confluence.
To differentiate the THP-1 monocytes to macrophages, cells were split into microtiter plates at a
concentration of 500 000 cells/mL and stimulated by phorbol myristate acetate (PMA), as
described previously (Hanáková et al., 2017) with some modifications. Briefly, after
differentiation by PMA and 24 h incubation, the culture medium of adherent macrophages was
replaced with a fresh complete medium. After another 24 h, the cells were washed with PBS, put
into serum-free medium for 2 h, and then used for subsequent experiments.
2
.5. Cytotoxicity Testing
The undifferentiated floating THP-1 cells (500 000 cells/mL) were transferred into a serum-free
RPMI 1640 medium and seeded into 96-well plates (100 µL/well) in triplicates. After 2 h, the test
compounds (1–25) dissolved in DMSO were added in increasing concentrations (0.313, 0.625,
0
.988, 1.25, 1.48, 2.22, 2.50, 3.33, 5, 6, 10, 15, and 20 µM). Measurements were taken after 24 h
1
2

incubation with each of the test compounds. The adherent HCT116 cells (500 000 cells/mL)
transferred into a serum-free DMEM high glucose medium were treated in the same manner
described above. The tested concentrations of compounds (1–25) for this cell line were 5 μM and
1
0 μM. The maximum concentration of DMSO in the assays never exceeded 0.1% (v/v). Cell
viability was measured using the Cell Proliferation Reagent kit WST-1 (Roche Diagnostics, Basel,
Switzerland), according to the manufacturer’s instruction. The amount of formazan created (which
corresponds to the number of metabolically active cells in the culture) was calculated as
a percentage of the control cells, which were treated only with DMSO, assigned as 100 %.
The resultant viability curves were used to calculate the IC50 values using four parameters logistic
(4PL) analysis.
2
.6. Determination of the NF-ĸB/AP-1 activity
The activity of NF-ĸB/AP-1 was evaluated on the THP-1-XBlue -MD2-CD14 cell line
Invivogen; San Diego, CA, USA) as described previously (Miláčková, Kapustová, Mučaji, &
Hošek, 2017) and slightly modified. Briefly, cells were transferred into serum-free medium
500 000 cells/mL) and the test compounds (1–25) at the concentration of 2 μM were added 2 h
TM
(
(
later. After a 1 h of incubation of the treated cells with the compounds, cells were stimulated by
LPS (1 µg/mL). LPS stimulation of the Toll-like 4 receptors in this cell line induced signaling
cascades leading to the activation of NF-ĸB/AP-1 and the subsequent production of the reporter
protein SEAP was evaluated by a QUANTI-Blue™ assay. After 24 h of incubation, 20 µL of
cultivation medium was mixed with 175 µL of Quanti-Blue™ reagent (Invivogen, San Diego,
USA) and incubated at 37 °C for 30 – 40 min according to the manufacturer’s instructions. The
activity of NF-ĸB/AP-1 was determined spectrophotometrically in a Fluostar Omega Microplate
Reader (BMG Labtech, Ortenberg, Germany) and compared to control.
1
3

2
.7. Signaling Pathway Analysis
TM
The level of IĸB was determined on the THP-1-XBlue -MD2-CD14 cell line, in a way similar to
that described previously (Miláčková et al., 2017). The mitogen-activated protein (MAP) kinase
pathways mediated by p38, ERK1/2, and JNK were measured. Briefly, cells cultivated in serum-
free medium were pretreated with the test compounds (3, 4, 7, 10, and 24) at a concentration of
2
µM. One hour later, 1 µg/mL of LPS was added and after 30 min the cells were collected into
ice-cold lysis buffer (50 mM Tris-HCl pH 7.5, 1 mM EGTA, 1 mM EDTA, 1 mM sodium
orthovanadate, 50 mM sodium fluoride, 5 mM sodium pyrophosphate, 270 mM sucrose) with
protease inhibitors (Roche Diagnostics, Basel, Switzerland). Cell lysates were centrifuged 3 000
g/5 min/4 °C and supernatants were mixed with 5x Laemmli buffer [250 mM Tris-HCl pH 6.8,
1
0% (w/v) SDS, 30% (v/v) glycerol, 5% (v/v) β-mercaptolethanol, 0.04% (w/v) bromphenol blue]
and incubated 5 min/70 °C. To separate the proteins, 9 μg of the proteins were loaded onto a 12%
polyacrylamide gel and then transferred electrophoretically to PVDF (polyvinylidene fluoride)
membranes that were subsequently blocked using 5% BSA dissolved in TBST buffer [150 mM
NaCl, 10 mM Tris base, and 0.1% (v/v) Tween-20]. The membranes were incubated at 4 °C for
1
6 h with primary mouse anti-IĸB-α antibody and primary rabbit anti-phospho-SAPK/JNK, anti-
p44/42 MAPK (ERK1/2), anti-phospho-p44/42 MAPK (ERK1/2), anti-p38 MAPK, and anti-
phospho-p38 MAPK antibodies (Cell Signaling, Danvers, MA, USA; products No. 4814, 4668,
4
695, 4370, 9212 and 4511) at a dilution of 1:500 (anti-IĸB-α), 1:2000 (anti-phospho-p44/42
MAPK) and 1:1000 (other antibodies). Two other primary mouse antibodies were used, anti-β-
actin (Abcam, Cambridge, UK; product No. 8226) at a dilution of 1:5000 and anti-SAPK/JNK
(Sigma-Aldrich, product No. SAB4200176) at a dilution of 1:1000. After washing, the secondary
anti-mouse and anti-rabbit IgG antibodies (Sigma-Aldrich, products No. A0168 and A0545) were
1
4

diluted 1:2000 and applied to the membranes, which were then incubated for 1 h at the laboratory
temperature (∼22 °C). The amount of bound secondary antibody was determined using ECL
reagent (Bio-Rad, USA). Chemiluminescence was detected using a Syngene PXi4
chemiluminescence imaging system (Cambridge, UK).
2
.8. Evaluation of Cytokine Secretion
Differentiated THP-1 macrophages, prepared as described above, were pretreated for 1 h with
μM solutions of the test compounds or prednisone dissolved in DMSO 0.1% or with a 0.1%
2
DMSO solution alone (the vehicle). The concentrations of the test compounds (1–25) used did not
have cytotoxic effects (cell viability ≥ 95%). An inflammatory-like response was triggered by
adding 1 μg/mL of LPS isolated from Escherichia coli 0111:B4 (Sigma-Aldrich, USA) and
dissolved in water to the pretreated macrophages; the control cells were cultured without LPS
treatment. After 24 h, the medium was collected and the concentration of TNF-α was measured
using a Human TNFα ELISA Kit (Diaclone, Besançon, France) according to the manufacturer’s
instruction. Each experiment was performed in triplicate.
2
.9. Molecular docking
The simulations of ligand-protein docking on NF-κB were performed with AutoDockVina (Trott
&
Olson, 2010) using a PyRx virtual screening tool. Molecular dynamics were performed with a
CUDA version of NAMD (Phillips et al., 2005), and VMD was used to prepare the molecular
visualization and protein preparation. Docking simulations were performed using system with
Core I7 and NVIDIA GTX 850. The docking simulations on 5-LOX and COX-2 were carried out
in Gold 5.2 (CCDC, www.ccdc.cam.ac.uk, 2015). Docking poses and protein-ligand interactions
were visualized in LigandScout 4.1 (Inte:Ligand, http://www.inteligand.com/, 2017).
1
5

In case of 5-LOX, the docking simulation was carried out on a structure based on PDB entry
O8Y (Gilbert, Bartlett, Waight, Neau, Boeglin, & Brash, 2011), which was virtually mutated to
3
wild type 5-LOX and energetically minimized in Discovery Studio 4.0 (Biovia). The binding site
was defined in an 8 Å radius around the γ2 carbon atom (CG2) of ILE167. ChemScore Kinase was
selected as a configuration template and CHEMPLP was selected as a scoring function. These
settings were optimized in a previous SAR study on 5-LOX (Pein et al., 2018).
For the docking on COX-2 the X-ray crystal structure 5IKV (Orlando & Malkowski, 2016), a
structure of human COX-2 co-crystalized with flufenamic acid was employed. The ligand from
the A chain was extracted and re-docked to validate the workflow (RMSD to the original ligand
was 0.295Å). The binding site was defined in a radius of 6 Å around the extracted ligand
(A_1;A_1:5IKV). ChemScore Kinase was selected as a configuration template and CHEMPLP
was selected as a scoring function.
The procedure described previously (Liu et al., 2015) was used for a protein energy minimization
calculation. Besides, an additional GPU accelerator with NAMD software package was applied
for the acceleration of the computation process. The crystallographic coordinates of NF-ĸB bound
to DNA were obtained from the RCSB Protein Data Bank with ID 3GUT. Only chain A (p65) and
B (p50) were used for simulations and all other coordinates were removed. These two protein
chains were minimized using molecular dynamics simulation with an NAMD package. The system
of proteins was neutralized and the concentration of NaCl was set to 0.15 mol/L. Minimization
used 5000 steps (Liu et al., 2015). After minimization, all heteroatoms were removed from the file
and the proteins were subjected to docking using a PyRx docking tool with AutoDockVina with
the exhaustiveness set to 20. All tested compounds meet the conditions of Lipinski rule of five
(except of compounds 24 and 25, which possess higher molecular weight) suggesting that tested
1
6

compounds can penetrate cell membranes, pass into cell compartments, and have several chemical
groups forming hydrogen bonds. All compounds formed complexes on various binding sites of
target protein structure with high binding energy (ΔG < -7.0 kcal/mol). In case that binding site of
best hit from docking result competed with DNA binding into DNA protein complex, the non-
specific result was treated as positive.
2
.10. Statistical Evaluation
Statistical analyses were carried out using GraphPad Prism 6.01 software (San Diego, CA,
USA). The data were graphed as the mean ± SEM. Comparisons between groups were made using
a one-way ANOVA test, followed by a Bonferroni’s multiple comparisons test.
3
3
. Results and Discussion
.1. COX/LOX Studies
All of the compounds (1–25) were tested for their potential to inhibit the catalytic activity of
COX-1, COX-2, and 5-LOX in vitro. Cell-free assay using human recombinant enzymes was used
for the primary screening of activity and the reaction products were quantified using ELISA kits.
Based on the results of screening, IC50 values were calculated for the most promising compounds
(Table 1). Only 5 of the 25 tested compounds showed significant activity against COX-1 and/or
COX-2. Trans-resveratrol (1) showed even more activity against both COX-1 and COX-2 than
ibuprofen used as a standard, although it was much more selective for COX-1 than COX-2
(COX-1/COX-2 selectivity ratio 0.22). The inhibitory activity of stilbenoids against COX enzymes
is well-known. A report describing the activity of trans-resveratrol (1), pterostilbene (2), and
piceatannol (3) on COX-1, COX-2, and 5-LOX enzymes accords with the observations we present
1
7

here (Kutil et al., 2015). Comparison of the compounds described here shows that any big change
in the substitution pattern of trans-resveratrol (1), such as methoxylation, glycosylation, or
modification of the carbon linkage connecting rings diminishes or eliminates the activity. On the
other hand, pinostilbene (4) demonstrated the greatest selectivity for COX-2 (COX-1/COX-2
selectivity ratio 5.48), even greater than ibuprofen (COX-1/COX-2 ratio 3.41). A noteworthy is
also the effect of pinosylvin monomethylether (17) and isorhapontigenin (18) against both COX-
1
0
and COX-2 enzymes. However, a high selectivity to COX-1 (COX-1/COX-2 ratios 0.28 and
.88, respectively), comparable with that of trans-resveratrol (1), is often associated with side
effects like gastrointestinal damage, characteristic for traditional non-selective NSAIDs. Although
the attention is aimed mostly at the development of selective COX-2 inhibitors, Perrone, Scilimati,
Simone & Vitale (2010) highlight the important role of COX-1 in the organism under normal and
pathological conditions and suggest that selective COX-1 inhibition might be highly relevant for
several diseases.
A similar situation was observed for the anti-5-LOX effect, where piceatannol (3) and
pterostilbene (2) (and to limited extent also pinostilbene (4)) showed promising activity,
comparable to the zileuton used as a positive control (Table 1). The specific activities of 4-
stilbenecarboxylic acid (14) and 1-(2,4-dihydroxyphenyl)-2-(4-hydroxyphenyl)-ethanone (23)
against 5-LOX are described here for the first time.
Table 1: Calculated IC50 values for COX-1, COX-2, and 5-LOX, and the COX-1/COX-2
selectivity ratio for selected stilbenoid compounds and corresponding standards, as determined
from a series of measurements in triplicates.
1
8

Compound/
Assay
COX-1
IC50 [μM]
0.20 ± 0.1
5.99 ± 1.57
inactive
COX-2
IC50 [μM]
0.90 ± 0.32
6.67 ± 2.21
inactive
COX-1/COX-2
5-LOX
selectivity ratio IC50 [μM]
Trans-resveratrol (1)
Pterostilbene (2)
Piceatannol (3)
0.22
0.90
-
inactive
1.63 ± 0.44
0.39 ± 0.09
4.01 ± 0.58
5
Pinostilbene (4)
1.37 ± 0.55
inactive
0.25 ± 0.06
inactive
5.48
-
4
-stilbenecarboxylic acid (14)
Pinosylvin monomethylether
17)
1
.66 ± 0.19
5.86 ±1.25
0.28
0.88
inactive
inactive
(
Isorhapontigenin (18)
8.77 ± 1.23 10.01 ± 3.15
1
-(2,4-dihydroxyphenyl)-2-
(
4-hydroxyphenyl)ethanone
23)
inactive
inactive
-
5
(
Ibuprofen
Zileuton
6.92 ±2.74
Not tested
2.03 ± 1.11
Not tested
3.41
-
Not tested
0.11 ± 0.06
In in silico simulation carried out on COX-2 the investigated compounds were docked to the
substrate binding site. The docking studies in general confirmed the results of in vitro assays. The
most active compounds all showed a clear binding pattern that involves hydrogen bonds with key
residues of COX-2. The hydroxyl groups of the most active compound pinostilbene (4) form
hydrogen bonds with Tyr355, Tyr385 and Ser530 which act as a carboxyl binding site for
arachidonic acid (AA) (Figure 3A) (Kiefer et al., 2000).
1
9

The compounds were also docked in an allosteric binding site on 5-LOX that has been shown to
accommodate phenolic 5-LOX inhibitors (Schaible et al., 2016). It has lately been shown that some
5
-LOX inhibitors do not act in competition with AA, but instead inhibit the enzyme independently
of substrate concentration (Mohamed et al, 2014; Srivastava, Vyas, Variya, Patel, Qureshi, &
Ghate, 2016; Cheungal et al., 2018). We therefore docked the compounds to the allosteric binding
site that is currently being investigated. The most active compounds dock to the binding site with
good scores and form hydrogen bond interactions that would suggest a stable placement. The
hydroxyl groups of the most active compound piceatannol (3) are predicted to form hydrogen
bonds with Phe402 and Arg138 (Figure 3B).
Figure 3. Docking pose of pinostilbene (4) in the substrate binding site of COX-2 (A),
piceatannol (3) in the allosteric binding site of 5-LOX (B).
3
1
.2. Determination of Cytotoxicity
The THP-1 cell line was used to further evaluate the anti-inflammatory potential of compounds
–25. Moreover, the human colon carcinoma cell line HCT116 was selected to verify the non-
toxic effect of the tested stilbenoids. A WST-1 assay was used to measure the viability of cells
after treatment with increasing concentrations of stilbenoids 1–25 in order to determine the
2
0

tolerable non-toxic concentrations for use in subsequent in vitro experiments and potentially
calculate the values of IC50. These assays showed relatively low toxic effects of tested compounds,
with IC50 values greater than 10 µM for both cell lines (the viability was > 70% for THP-1 at the
concentration of 20 μM and for HCT116 it was > 75 % at 10 μM). The only exception was
piceatannol (3) with an IC50 value of 6.7 ± 1.0 µM for the THP-1 cell line. The viability of the
HCT116 cell line was not markedly affected by the tested compounds. Cytotoxicity assays showed
relatively safe profiles for all of the compounds tested which enabled us to continue our study of
cell-based anti-inflammatory activity. The low toxicity of the stilbenoids we tested may be related
to the presence of several hydroxyl groups attached directly on ring A and/or ring B. The absence
of lipophilic substituents (Sirerol, Rodríguez, Mena, Asensi, Estrela, & Ortega, 2016) that could
increase the ability of the compounds to penetrate the membranes of cells and interact with the
internal cell environment might be also a factor (Hanáková et al. 2017), but subsequent
experiments showed that at non-toxic concentrations stilbenoids are able to interact well with
intracellular mechanisms. Based on the cytotoxicity assays, the following bioactivity tests were
performed on each of the compounds 1–25 at non-toxic concentrations of 2 µM.
3
.3. Anti-inflammatory Potential
Although the anti-inflammatory effect of stilbenoids is well-known, only a few compounds have
been in detail evaluated for this activity (Dvorakova & Landa, 2017), trans-resveratrol (1),
piceatannol (3), and pterostilbene (2) having been studied the most. Our study evaluated the
activities of 25 stilbenoids all at the same pharmacologically relevant concentration of 2 μM and
all under the same conditions in order to provide a reliable comparison of their effect. All of
stilbenoids 1–25 were first evaluated for their ability to attenuate the LPS-stimulated activation of
2
1

NF-κB/AP-1 in THP1-XBlue™-MD2-CD14 cells in a primary screening for anti-inflammatory
effects. LPS stimulation of the TLR4 receptors in this cell line induced signaling cascades that led
to the activation of both the NF-κB and AP-1 transcription factors. Most of the tested compounds
reduced the activity of NF-κB/AP-1 when the cells were stimulated by LPS (Figure 4A).
Piceatannol (3), thunalbene (7), oxyresveratrol (8), 2-carboxyl-3-O-methyl-4´-β-D-
glucopyranosyl-dihydroresveratrol
(10)
and
3-O-caffeoyl-(9→5)-β-apiosyl-(1→6)-β-
glucopyranosyl-3´-O-methyldihydropiceatannol (24) affected the activity of NF-κB/AP-1 in a
statistically significant manner, comparable to that of the conventionally used drug prednisone.
These data indicate that hydroxylation of aromatic rings increases the inhibition activity against
NF-κB/AP-1. Piceatannol (3) and oxyresveratrol (8), differing from trans-resveratrol (1) only in
one additional aromatic hydroxyl group, have shown greater activity against NF-κB/AP-1 than 1
itself. Likewise, compound 24 has been more active than his O-methylated analogue (25). The
position of hydroxy groups influences also the activity of stilbenes in order para<ortho<meta.
Compounds 3 and 7, having a hydroxyl group in meta-position of ring B, showed greater inhibitory
effect on NF-κB/AP-1 activation than their analogues 8 and 4, having a hydroxyl group in ortho-
and para- position, respectively. Interestingly, hydroxyl group in para-position seems to enhance
the NF-κB/AP-1 activity in the case of compound 4 compared to 7. On the other hand,
modifications of ethylene bridge, for example its substitution (compound 16, 23 referred to 12),
attenuates the inhibition of activity against NF-κB/AP-1. Similarly, the hydrogenation of ethylene
bridge in the case of batatasin III (6) caused a decrease of activity in comparison with thunalbene
(
7). The most active compound piceatannol (3), in agreement with previous studies, showed a
greater effect on the inhibition of the NF-κB/AP-1 pathway than resveratrol (1) and oxyresveratrol
8) (Ashikawa, Majumdar, Banerjee, Bharti, Shishodia, & Aggarwal, 2002; Liu, Kim, Park, Na, &
(
2
2

Surh, 2009; Son, Park, Na, Jue, Kim, & Surh, 2010). Furthermore, this is the first report to describe
the attenuating effects of thunalbene (7) and compounds (10 and 24) from T. tommasinii on the
NF-κB/AP-1 activity.
2
3

Figure 4. Effects of compounds 1–25 on LPS-stimulated NF-κB/AP-1 activity (A) and effect of
compounds 3, 4, 7, 10 and 24 on the signaling cascade in the THP1-XBlue™-MD2-CD14 cell
line (B). (A) Cells of the THP1-XBlue™-MD2-CD14 cell line were pre-treated with the test
compounds (2 μM, dissolved in DMSO), or prednisone (P; 2 µM, dissolved in DMSO) for 1 h.
Subsequently, LPS (1 µg/mL) was added [except for the control cells (C or CTRL)] to trigger the
activation of NF-κB/AP-1 signaling. After 24 h, the NF-κB/AP-1 activity was evaluated based
on the spectrophotometric measurement of amount of alkaline phosphatase secreted. The results
are expressed as the mean ± S.E. for six independent experiments. (B) The level of MAPKs, IκB
and β-actin in cell lysate was detected 30 min after LPS stimulation by Western blot and
immunodetection. The blots shown represent the results of three independent experiments. V or
Veh. represent the vehicle treated cells. * Indicates a significant difference of p < 0.05 in
comparison with cells treated only with the vehicle, ** indicates a significant difference of p <
0
.01 in comparison with cells treated only with the vehicle, and **** indicates a significant
difference of p < 0.0001 in comparison with cells treated only with the vehicle.
Compounds possessing significant effect on NF-κB/AP-1 activity, piceatannol (3), pinostilbene
4) (due to its observed effect on LPS-stimulated NF-κB activity), thunalbene (7), 2-carboxyl-3-
(
O-methyl-4´-β-D-glucopyranosyl-dihydroresveratrol (10), and 3-O-caffeoyl-(9→5)-β-apiosyl-
1→6)-β-glucopyranosyl-3´-O-methyldihydropiceatannol (24), were selected for more detailed
(
analysis of their effect on NF-κB and on the signaling pathways of mitogen-activated protein
kinases (MAPKs). MAPK cascades, represented by p38, ERK, and JNK kinases, can be triggered
by phosphorylation events in their activation loops during inflammation or in response to an
extracellular stimulus, such as LPS. This activates the AP-1 transcription factor. As for MAPKs,
2
4

extracellular stimuli activate transcription factor NF-κB by inducing degradation of its inhibitor
IκBα (Lee & Kim, 2007). Piceatannol (3), pinostilbene (4), thunalbene (7), 2-carboxyl-3-O-
methyl-4´-β-D-glucopyranosyl-dihydroresveratrol (10) inhibited the phosphorylation of MAPKs
p38, ERK1/2, and JNK (Figure 4B). Only compound 24 did not affected MAPKs, possibly because
the size and shape of molecule would not fit into an active center of these proteins. Several previous
studies have demonstrated attenuation of MAPKs phosphorylation by stilbenoids, for example
pterostilbene (2) (Paul et al., 2009), desoxyrhapontigenin (Choi et al., 2014), trans-resveratrol (1)
(Oh et al., 2009), or piceatannol (3) (Ko et al., 2013). However, it is worth noting that most of
these studies applied much greater concentrations of stilbenoids than 2 μM used in current study,
levels that do not correspond to concentrations likely to be found in living organism, e.g., for
resveratrol (1) the maximal plasma concentrations were ≈ 2 µM after oral administration (Cottart,
Nivet-Antoine, Languillier-Morizot, & Beaudeux, 2010; Walle, Hsieh, DeLegge, Oatis, & Walle,
2
004). None of the tested compounds influenced the degradation of IκBα at a concentration of 2
μM, moreover, compound 7 notably decreased its level. Similar negative results have been found
for pterostilbene (2) using different models (Paul et al., 2009) and at low concentrations (< 10 μM)
for piceatannol (3) (Liu et al., 2009) and trans-resveratrol (1) (Oh et al., 2009). It can be
hypothesized that if thunalbene (7) attenuates NF-κB activity, but the level of IκB is low, 7 could
directly interact with activated NF-κB and inhibit its activity. Several studies report about targeting
the NF-κB signaling pathway by small natural molecules at various levels, e.g. p65 subunit
acetylation, methylation and DNA binding. Similarly, NSAIDs have been found to block NF-κB,
but their effects are highly pleiotropic (Gupta, Sundaram, Reuter, & Aggarwal, 2010).
These experiments show that some stilbenoids attenuate the activity of NF-κB and AP-1 by
inhibiting the MAP kinases p38, ERK1/2, and JNK, rather than by inhibiting the degradation of
2
5

IκBα. Interestingly, compound 24 affected neither the MAPKs nor IκBα, but was nevertheless able
to attenuate the activity of NF-κB/AP-1 and the secretion of TNF-α. To investigate another
possible mechanism of action, all of the test stilbenoids were docked into the DNA binding pocket
of the NF-κB heterodimer p65/p50 (Figure S1). The bond of compound 24 to NF-κB had one of
the highest binding energies (ΔG = -6.8 kcal/mol) (Table S1).
A mechanism of competitive inhibition has been described for trans-resveratrol (1) and its
derivatives (Mulakayala et al., 2013) and may explain why 24 inhibits NF-κB/AP-1 activity, even
though does not affect MAPKs or NF-κB/AP-1 signaling cascades. Interestingly, compound 25,
which differs from 24 only by a single methoxy group, had a great binding energy (ΔG = -7.3
kcal/mol), but its inhibitory activity against NF-κB/AP-1 was slightly lower [14.7 ± 5.0 % for 25
vs. 20.0 ± 6.0 % for 24]. Piceatannol-3-O-glucoside (5) had the same binding energy (ΔG = -7.3
kcal/mol), but its inhibition of NF-κB/AP-1 was relatively weak (11.3 ± 6.8 %).
The next experiments were focused on the influence of stilbenoids on LPS-activated TNF-α
expression, which is under the control of NF-κB transcription. Piceatannol (3), thunalbene (7),
oxyresveratrol (8), 2-carboxyl-3-O-methyl-4´-β-D-glucopyranosyl-dihydroresveratrol (10) and 3-
O-caffeoyl-(9→5)-β-apiosyl-(1→6)-β-glucopyranosyl-3´-O-methyldihydropiceatannol (24) were
selected for this analysis based on their significant effects on the inhibition of NF-κB/AP-1.
Pinostilbene (4) and resveratrol (1) were added to the experimental set because they were the most
potent COX and 5-LOX inhibitors, as seen from previous experiments. Compounds 1, 3, 8, 24,
and 10 diminished the production of TNF-α more than the prednisone that was used as a positive
control, but only compounds 10 and 24 influenced the LPS-stimulated TNF-α secretion in a
statistically significant way (p < 0.05) (Figure 5). These results substantiate the previous study of
Granica et al. (2015), in which compounds 10 and 24 significantly attenuated the secretion of
2
6

TNF-α from human neutrophils. The ability of trans-resveratrol (1), piceatannol (3) and
oxyresveratrol (8) to diminish the TNF-α expression has been already described in several studies
(
Dvorakova & Landa, 2017; Mouihate, Horn, & Pittman, 2006). Interestingly, trans-resveratrol
1) did not influence the NF-κB activity, which indicates that it interacts with some post-
(
transcription and/or post-translation regulatory mechanism. In contrast, thunalbene (7) reduced the
activity of NF-κB/AP-1, but did not influence the expression of TNF-α. Possible interaction of 7
with other pro-inflammatory cytokines being under NF-κB transcription control, would be a
subject of further investigation. It is worth noting that pinostilbene (4) moderately increased the
production of TNF-α, probably because of its greater effect on NF-κB/AP-1. Compound 4 showed
the greatest inhibitory effect on COX-2 and moderate influence on 5-LOX.
2
7

Figure 5. The effect of compounds 1, 3, 4, 7, 8, 10 and 24 on the secretion of TNF-α. THP-1
macrophages were pre-treated with each of the compounds (2 μM, dissolved in DMSO) or
prednisone (P; 2 µM, dissolved in DMSO) for 1 h. LPS (1 µg/mL) was then added [except for
the control cells (C)] to trigger the secretion of the pro-inflammatory cytokine TNF-α. After 24
h, the amount of TNF-α was evaluated by ELISA. The results are expressed as the mean ± S.E.
for six independent experiments. * Indicates a significant difference, p < 0.05 in comparison with
the cells treated only with the vehicle (V).
All of the above-mentioned mechanisms - direct inhibition of the NF-κB/AP-1 activity and/or
MAPKs signaling cascades, the interaction with the DNA binding pocket of NF-κB, the inhibition
of the secretion of TNF-α, and the inhibition of the COX and LOX enzymes - could contribute to
overall anti-inflammatory effect of stilbenoids. This study confirms that stilbenoids should be
considered as potent multi-target anti-inflammatory agents and perspective candidates for further
drug discovery.
Acknowledgement. The financial support of GACR no. 16-07193S project is gratefully
acknowledged by K.Š., V. L., P. B., and J.H. V.T. is funded by the Austrian Science Fund (FWF)
project T942-B30.
Conflict of interest. Authors do not have any conflict of interest.
References:
2
8

Ashikawa, K., Majumdar, S., Banerjee, S., Bharti, A. C., Shishodia, S., & Aggarwal, B. B.
2002). Piceatannol Inhibits TNF-Induced NF-κB Activation and NF-κB-Mediated Gene
(
Expression Through Suppression of IκBα Kinase and p65 Phosphorylation. Journal of
Immunology, 169(11), 6490-6497.doi:10.4049/jimmunol.169.11.6490.
Baur, J. A., & Sinclair, D. A. (2006). Therapeutic potential of resveratrol: the in vivo
evidence. Nature Reviews Drug Discovery, 5, 493-506. doi:10.1038/nrd2060.
Bontempo, P., De Masi, L., Carafa, V., Rigano, D., Scisciola, L., Iside, C., Grassi, R.,
Molinari, A.M., Aversano, R., Nebbiosso, A., Carputo, D., Altucci, L. (2015). Anticancer
activities of anthocyanin extract from genotyped Solanum tuberosum L. “Vitelotte”. Journal
of Functional Foods, 19, 584-593. doi:10.1016/j.jff.2015.09.063.
Brezáni, V., Šmejkal, K., Hošek, J., & Tomášová, V. (2018). Anti-inflammatory Natural
Prenylated Phenolic Compounds - Potential Lead Substances. Current Medicinal Chemistry,
2
5, 1094-1159. doi:10.2174/0929867324666170810161157.
Cao H., Shi J., Jia, X. P., Xiao J. B., & Chen, X. Q. (2016). Non-covalent interaction between
dietary stilbenoids and human serum albumin: Structure-affinity relationship, and its
influence on the stability, free radical scavenging activity and cell uptake of stilbenoids.
Food Chemistry, 202, 383-388. doi:10.1016/j.foodchem.2016.02.003.
Chachay, V. S., Kirkpatrick, C. M. J., Hickman, I. J., Ferguson, M., Prins, J. B., & Martin, J.
H. (2011). Resveratrol – pills to replace a healthy diet? British Journal of Clinical
Pharmacology, 72, 27-38. doi:10.1111/j.1365-2125.2011.03966.x.
2
9