Benzylidenemalononitrile compounds as activators of cell resistance to oxidative stress and modulators of multiple signaling pathways. A structure-activity relationship study

Article abstract of DOI:10.1016/j.bcp.2011.05.028

Benzylidenemalononitrile (BMN) tyrphostins are well known as potent tyrosine kinase inhibitors. Moreover, in recent years it has been recognized that members of the tyrphostin family possess additional biological activities independent of their ability to inhibit protein tyrosine kinases. In this study, we examined the relationship between the structure of 49 BMNs and related compounds, and their capacity to induce heme oxygenase 1 (HO-1) gene expression in U937 human monocytic cells, to activate upstream signaling pathways and to protect cells against menadione-induced oxidative stress. It was found that the electron-withdrawing (NO₂, CN, halogen) groups in BMN molecules and double meta-MeO substituents increased the HO-1 gene induction, while the electron-donating groups in ortho/para position (OH, MeO and N-morpholino) significantly decreased it. The magnitude of activation of c-Jun, Nrf2, p38 MAPK, and p70S6K correlated with specific substitution patterns in the BMN structure. BMN-dependent maximal up-regulation of HO-1 required parallel increase in Nrf2 and phospho-c-Jun cellular levels. Liquid chromatography mass spectrometry (LC-MS) analysis revealed that BMNs can generate conjugates with one or two glutathione equivalent(s). This study supports the hypothesis that BMNs induce the expression of protective genes by alkylating sensitive cysteine residues of regulatory factors.

Full text of DOI:10.1016/j.bcp.2011.05.028

Biochemical Pharmacology 82 (2011) 535–547
Contents lists available at ScienceDirect
Biochemical Pharmacology
journal homepage: www.elsevier.com/locate/biochempharm
Benzylidenemalononitrile compounds as activators of cell resistance to oxidative
stress and modulators of multiple signaling pathways. A structure–activity
relationship study
a
a
Kyril Turpaev a,b,c, , Mikhail Ermolenko , Thierry Cresteil , Jean Claude Drapier
*
a,**
a Institut de Chimie des Substances Naturelles, UPR2301 CNRS, Centre de Recherche de Gif, 91190 Gif-sur-Yvette, France
b Vavilov Institute of General Genetics, Russian Academy of Sciences, 119991 Moscow, Russia
c Center for Theoretical Problems of Physicochemical Pharmacology, Russian Academy of Sciences, 119991 Moscow, Russia
A R T I C L E I N F O
A B S T R A C T
Article history:
Benzylidenemalononitrile (BMN) tyrphostins are well known as potent tyrosine kinase inhibitors.
Moreover, in recent years it has been recognized that members of the tyrphostin family possess
additional biological activities independent of their ability to inhibit protein tyrosine kinases. In this
study, we examined the relationship between the structure of 49 BMNs and related compounds, and
their capacity to induce heme oxygenase 1 (HO-1) gene expression in U937 human monocytic cells, to
activate upstream signaling pathways and to protect cells against menadione-induced oxidative stress. It
was found that the electron-withdrawing (NO2, CN, halogen) groups in BMN molecules and double meta-
MeO substituents increased the HO-1 gene induction, while the electron-donating groups in ortho/para
position (OH, MeO and N-morpholino) significantly decreased it. The magnitude of activation of c-Jun,
Nrf2, p38 MAPK, and p70S6K correlated with specific substitution patterns in the BMN structure. BMN-
dependent maximal up-regulation of HO-1 required parallel increase in Nrf2 and phospho-c-Jun cellular
levels. Liquid chromatography mass spectrometry (LC–MS) analysis revealed that BMNs can generate
conjugates with one or two glutathione equivalent(s). This study supports the hypothesis that BMNs
induce the expression of protective genes by alkylating sensitive cysteine residues of regulatory factors.
ß 2011 Elsevier Inc. All rights reserved.
Received 7 April 2011
Accepted 25 May 2011
Available online 2 June 2011
Keywords:
Benzylidenemalononitriles
Oxidative stress
Heme oxygenase
Nrf2
c-Jun
increase in cellular glutathione (GSH) level, inhibit expression of
1. Introduction
pro-inflammatory genes as iNOS and COX-2, and suppress TNF
a
release and activation of poly (ADP-ribose) polymerase [3,5]. These
tyrphostin-driven effects have been observed in experiments
aimed at pharmacological suppression of septic shock, ischemia,
inflammation, and injuries caused by radiation and chemotherapy
[6–10]. The exact molecular mechanisms mediating the protective
effects of BMN compounds remain unclear, but we have previously
shown that AG-126 and several other structurally related
tyrphostins are potent activators of the expression of heme
oxygenase 1 (HO-1), H-ferritin, interleukin 8 (IL-8), and several
other, mainly redox-sensitive, genes. Our study also provided
evidence that the gene induction capacity of BMN tyrphostins was
not related to inhibition of protein tyrosine kinases. We showed
that AG-126-like tyrphostins stimulate key regulators of redox-
sensitive signaling systems like MAP kinases as well as AP-1 and
Nrf2 transcription factors [11]. It is worth noting that BMN
compounds have distant structural similarity with curcumin and
chalcones. These well-known activators of protective genes have
promising anti-inflammatory properties and capacities for pre-
venting tumorigenesis and suppressing angiogenesis [12,13].
The present study was focused on the structure–activity
relationship of a wide series of synthetic BMN compounds and
Pharmacological effects of benzylidenemalononitrile (BMN)
compounds have been examined since the 1990s when several of
their derivatives, referred to as tyrphostins, were recognized as
specific inhibitors of epidermal growth factor tyrosine kinase [1,2].
Subsequent design and testing of a series of BMNs revealed new
specific inhibitors of various protein tyrosine kinases [2]. Later, it
was revealed that BMN tyrphostins possess biological activities
unrelated to inhibition of protein kinase activity. For instance,
phenolic tyrphostins are antioxidants, free radical scavengers, and
mitochondrial uncouplers [3,4]. AG-126-like tyrphostins stimulate
Abbreviations: ARE, antioxidant response element; BMN, benzylidenemalononi-
trile(s); ELSD, evaporative light scattering detection; ESI, electrospray ionization;
GSH, glutathione; HO-1, heme oxygenase 1; IL-8, interleukin 8; LC–MS, liquid
chromatography mass spectrometry; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide; p70S6K, p70 S6 kinase.
* Corresponding author. Present address: Department of Medical Cell Biology,
Uppsala University, Biomedical Center, 75123 Uppsala, Sweden.
Tel.: +46 73 742 53 23; fax: +46 18 471 40 59.
** Corresponding author. Tel.: +46 73 742 53 23; fax: +46 18 471 40 59.
E-mail addresses: kyril.turpaev@yahoo.com (K. Turpaev),
jean-claude.drapier@icsn.cnrs-gif.fr (J.C. Drapier).
0006-2952/$ – see front matter ß 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.bcp.2011.05.028

536
K. Turpaev et al. / Biochemical Pharmacology 82 (2011) 535–547
related compounds in a human monocytic cell line. BMN biological
effects were characterized by measuring HO-1 expression, the
activity of upstream signaling systems, and modulation of cell
resistance to oxidative stress. Moreover, we have determined the
structure of the products generated by reaction between BMNs and
GSH, a key component of cellular thiol redox homeostasis.
(Molecular Devices, Sunnyvale, CA) was used for quantification.
Dose–effect analyses were performed according to [16]. The data
were expressed using the median-effect equation Fa/Fu = (D/Dm)m,
where D is the concentration, Fa and Fu are the fractions of the cells
affected and unaffected, respectively, by the concentration D, Dm is
the concentration required to induce 50% cell death, and the index
m is a measure of the sigmoidicity of the dose–response curve [16].
2. Materials and methods
2.6. Western blotting
2.1. Reagents
U937 cells (ꢂ10 ꢁ 106) were washed twice with ice-cold
phosphate-buffered saline and whole cell extracts were prepared
in 10 mM Tris, pH 7.4, 1% SDS, 10 mM NaF, 1 mM Na3VO4, and
Tyrphostins AG-9 (4-methoxybenzylidene)malononitrile), AG-
10 (4-hydroxy-benzylidenemalononitrile), and AG-126 (3-hy-
droxy-4-nitrobenzylidenemalononitrile) were purchased from
Calbiochem (San Diego, CA). Chemicals were dissolved in DMSO.
In all experiments with cell cultures, the final DMSO concentration
in the medium did not exceed 0.05%.
100
m
M phenylmethylsulfonyl (Sigma–Aldrich, St. Louis, MO)
l and denatured by heating at
fluoride in a total volume of 600
m
95 8C for 5 min. Protein concentration in each lysate was assayed
colorimetrically (Bio-Rad protein reagent; Bio-Rad, Hercules, CA),
and 20
mg of protein was resolved by 10% SDS-PAGE. After
2.2. General procedure for the synthesis of BMN compounds
electrophoresis, proteins were transferred to Hybond ECL nitro-
cellulose membranes (Amersham Biosciences, UK). Membranes
were blocked in Tris-buffered saline containing 5% nonfat milk for
2 h. The following antibodies were used for Western analyses:
rabbit anti-HO-1 (Santa Cruz Biotechnology, Santa Cruz, CA), anti-
phospho-p38, anti-phospho-c-Jun, anti-phospho-p70S6 K (Cell
Signaling Technology, Beverly, MA), and anti-Nrf2 (Abcam, Cam-
bridge, UK) at 1:100 (anti-HO-1) or 1:1000 dilution (all other
antibodies) in 1ꢁ Tris-buffered saline, 0.1% Tween 20 (TTBS), and
Mostof the BMN compounds were convenientlyprepared inhigh
yield by straightforward Knoevenagel condensation of the relevant
aldehyde with malononitrile (1.2 equiv.) catalyzed by imidazole
(0.01 equiv.) at room temperature in methanol. The products
obtained were of at least 98% purity after re-crystallization of the
product from methanol, ethanol or heptane. EMK 201 and EMK 211
were prepared from EMK 171 and EMK 181, respectively, by
treatment with BBr3 (4 equiv., ꢀ40 8C to room temperature). EMK
311 and EMK 321 were prepared by condensation of 3-nitroben-
zaldehyde with the corresponding methylene compounds in the
presence of catalytic amounts of piperidine and acetic acid in boiling
benzene [14]. EMK 381 was obtained as reported [15]. All reagents
were from Sigma–Aldrich (St. Louis, MO).
5% BSA overnight. Rabbit anti-b-actin antibodies (Cell Signaling) at
a 1:10,000 dilution in 1 ꢁ TTBS were used as a control. Appropriate
horseradish peroxidase-coupled secondary goat antibodies (Dako,
Glostrup, Denmark) were used for visualization using the
chemiluminescent HRP substrate (Millipore, Billerica, MA) and
chemiluminescence detection system (Bio-Rad, Hercules, CA). For
quantification the Quantity One-4.6 software (Bio-Rad) was used.
2.3. Cell treatments
2.7. Absorption spectra of BMN–GSH conjugates
U937 cells were cultured in RPMI 1640 medium supplemented
with 25 mM HEPES, pH 7.3, 5% fetal calf serum, and gentamicin
BMNs (50
mM) were incubated with GSH (1 mM) in a total
(50
m
g/ml) (Invitrogen, Carlsbad, CA) at 37 8C under humid 5% CO2
volume of 1 ml of 50 mM Tris–HCl buffer (pH 7.5) in the presence
of 5% DMSO for 1 h at room temperature. Spectra were recorded
using a UV–vis double-beam spectrophotometer (Safas UV mc2,
Safas, Monaco). GSH and DMSO were present at the same
concentration in the reference cuvette.
atmosphere. Cells were grown in 75 cm2 flasks in a volume of
15 ml up to a density of about 0.7 ꢁ 106 cells/ml and kept in fresh
culture medium for 14 h before the experiments.
2.4. RNA extraction and quantitative real-time PCR analysis
2.8. Liquid chromatography mass spectrometry (LC–MS) analysis of
BMN–GSH conjugates
Cells were lysed and total RNA was extracted using the RNA
extraction kit together with DNA elimination by DNase (Promega,
Madison, WI), according to the manufacturer’s instructions. Two
BMN compounds (1 mM) were incubated with 5 mM GSH
m
g of RNA was used to prepare cDNAs with oligo(dT) primers and
(Sigma-Aldrich, St. Louis, MO) in a total volume of 400 ml of 50 mM
AMV reverse transcriptase (Promega). Quantitative real-time PCR
(RT-PCR) was performed using the Roche Light Cycler System and
the FastStart DNA Master SYBR Green I kit (Roche Diagnostics,
Basel, Switzerland) as previously reported [11]. Values were
normalized to the relative amounts of GAPDH cDNA. All RT-PCR
measurements were performed in triplicate.
Tris–HCl buffer (pH 7.5) in the presence of 10% MeOH for 1 h at
room temperature. The reaction mixture was analyzed using a
UPLC/MS System (Waters Acquity UPLC, Waters Corporation,
Milford, MA) connected to a Waters Acquity TQD triple quadripole
spectrometer, with electrospray ionization (ESI) and a Waters
Acquity ELSD detector. Separations were carried out on a Waters
HSS C18 (2.1 ꢁ 50 mm, 1.8
mm) column at a flow rate of 0.6 ml/
min using the gradient elution from 5% water–acetonitrile to 100%
acetonitrile (both containing 0.1% of formic acid) over 10 min.
2.5. Measurement of menadione toxicity
U937 cells were grown in 96-well microtiter plates at a density
of 15 ꢁ 103/well in a volume of 200
m
l. The protective effect of
3. Results
prior 14-h incubation with 25
m
M BMN compounds on survival of
M menadione was quantified
U937 cells exposed for 5 h to 0–60
m
3.1. Time-course and dose dependence of AG-126 induced activation
of HO-1 expression
by colorimetric measurement of MTT reduction. Menadione
(Sigma–Aldrich, St. Louis, MO) was dissolved in phosphate-
buffered saline (PBS) (pH 7.3) and added to culture medium to
the final concentrations as indicated. A SpectraMax Plus Reader
We have previously reported that the cytoprotective protein
HO-1 is highly sensitive to tyrphostin AG-126 in U937 monocytic

K. Turpaev et al. / Biochemical Pharmacology 82 (2011) 535–547
537
cells [11]. Here, we further characterized this effect by performing
quantitative dose–response and time-course experiments. The
HO-1 mRNA content increased in U937 cells together with AG-126
191, in the mono-substituted BMN series with electron-withdraw-
ing groups (NO2, CN, halogens) in ortho/meta/para position and
meta-MeO substituents (-I effect) increased HO-1 mRNA induction,
while the electron-donating groups with strong +R effect in para-
position (OH and N-morpholino) significantly decreased activity.
The effect of single methoxy-group in ortho or para position was
insignificant. In the series of di-substituted BMNs, the net
combination of the above effects could be seen, with the exception
of striking loss of activity for 3,5-dinitro-BMN EMK 281, as
compared to the mono-nitro substituted EMK 491, EMK 151 and
EMK 121, and compounds bearing a 4-OH group in combination
with electron-withdrawing groups at the ortho-position to this
group (4-OH,3-MeO, for EMK 031, 4-OH,3-F for EMK 211, 4-OH,3-
NO2 for EMK 071).
concentration up to 50
about 500-fold (Fig. 1A). The HO-1 mRNA reached its maximum
after a 3-h exposure to 25 M AG-126 (Fig. 1B). The HO-1 protein
mM, to reach a maximal accumulation of
m
level was increased up to 9 h after exposure to AG-126, reaching an
approximately 10-fold increase over the basal level (Fig. 1C).
3.2. Capacity of BMNs to induce HO-1: structure–activity relationship
As a hallmark of up-regulation of cell protective systems, the
capacity of a series of BMNs and structurally related compounds to
affect HO-1 at both the mRNA and protein levels were examined
(see Table 1). The parent benzylidenemalononitrile EMK 191
strongly activated HO-1 mRNA induction. As compared to EMK
3.3. Effect of BMNs upon signaling pathways
A main goal of our study was to provide insights into
mechanism of activation of upstream signaling systems by BMN
compounds. Therefore, we examined the effects of BMNs and
related compounds on the protein content and/or the phosphory-
lation level of key stress signaling molecules, namely the
transcription factors Nrf2 and c-Jun, p70 S6 kinase (p70S6K), an
mTOR substrate and key regulator of mRNA translation and p38
MAPK. As shown in Fig. 2A, a pilot study focusing on 6 BMNs
showed that a majority of these compounds tend to activate Nrf2
expression and the phosphorylation of c-Jun and p38, and to
reduce the phosphorylation of p70S6 K. Then, we examined the
ability of BMN compounds to promote Nrf2 accumulation and c-
Jun phosphorylation using an extended set of compounds (Table
1). Quantification of Western blots showed that the parent BMN
(EMK 191) and the nitro-substituted BMNs (EMK 121, EMK 151,
AG-126 and EMK 281) are the most effective activator of Nrf2
expression. All other substitutions attenuated the magnitude of
Nrf2 accumulation whatever the presence of electron-donating
and electron-withdrawing groups. As regards Nrf2 accumulation,
twelve para-substituted BMN compounds proved effective in
the following order: NO2 > H > morpholine > MeS > CH3 > t-
Bu > MeO > F > HO > CF3 > CN > Br, while for HO-1 gene activa-
tion, the order was: NO2 > CN > Br > CH3 > MeS > F > H > t-
Bu > MeO ꢃ HO > morpholine. The position and number of NO2
substitution were of minor importance. In contrast, the presence of
two CN groups on the lateral chain was indispensable, indicating
that the Michael reaction acceptor functionality is critical for the
Nrf2 activation (Table 1).
Stimulation of c-Jun phosphorylation was highly dependent on
the structure of BMN compounds. In U937 cells exposed to parent
BMN EMK 191, the level of phospho-c-Jun increased about 3-fold.
The rank order of twelve BMN compounds as regards c-Jun
activation was as follows: NO2 > CF3 > Br > t-Bu > MeO ꢄ F ꢄ
CN ꢄ MeS > H > CH3 > HO ꢃ morpholine. Compounds containing
para- or meta-NO2 groups, 3,5-(MeO)2 (EMK 131), 4-t-Bu (EMK
241) or halogens were effective activators of c-Jun phosphorylation
(FC > 5), whereas addition of multiple MeO groups (EMK 721, EMK
731, EMK 841, EMK 851, and EMK 861) was less effective. The
presence of a 4-OH group (EMK 031, EMK 071, EMK 211, and EMK
871) or a 4-morpholine group (EMK 251) led to complete reduction
of c-Jun phosphorylation (Table 1).
A
600
400
200
0
0
3
10
25
50
AG-126 (μM)
B
C
400
300
200
100
0
0
1.5
3
5
7
9
Hours
15
10
5
0
0
3
6
9
When HO-1 mRNA accumulation elicited by BMN treatment is
plotted against Nrf2 content (Fig. 2B) or p-c-Jun content (Fig. 2C), a
significant positive correlation can be calculated. The linear
correlation coefficients (R2) were respectively 0.346 (n = 42) and
Hours
Fig. 1. HO-1 expression in U937 cells exposed to AG-126. Dose dependence (A),
time-course (B) for HO-1 mRNA accumulation and time-course of HO-1 protein
0.383 (n = 38) whereas Pearson correlation coefficients (rX,Y) were
accumulation (C). U937 cells were incubated with 25
mM AG-126 for the indicated
respectively 0.588 and 0.619. Of interest, BMN compounds (EMK
031, EMK 071, EMK 211, EMK 251, EMK 861, and EMK 871) that
up-regulated Nrf2 accumulation but did not stimulate c-Jun
times (B and C) or for 3 h at the indicated concentrations (A). Each bar represents the
mean ꢅ SD of at least three independent experiments. HO-1 mRNA and protein levels
in untreated cells were assigned a value of 1.

538
K. Turpaev et al. / Biochemical Pharmacology 82 (2011) 535–547
Table 1
BMN compounds examined in this study. Structure and potency to induce HO-1 and to activate Nrf2 and c-Jun.
Compounds
Code
Induction/activation (Fold-change)
Structure
HO-1 mRNA
76.3
HO-1 protein
Nrf2
8.7
p-cJun
3.5
AG-9
ND
ND
8.4
CN
CN
MeO
AG-10
AG-126
28.3
8.2
2.2
5.5
CN
CN
OMe
307.1
13.1
HO
CN
CN
O2N
EMK 031
EMK 041
EMK 071
9.4
ND
6.3
ND
5.0
5.7
7.4
1.3
2.1
0.9
MeO
CN
CN
CN
CN
CN
CN
HO
HO
123.0
4.1
MeO
O2N
HO
EMK 081
EMK 091
165.7
127.9
8.0
ND
9.2
8.7
2.5
3.5
CN
CN
CN
CN
F
EMK 121
EMK 131
309.7
243.4
ND
ND
15.8
8.2
13.2
8.5
CN
CN
O2N
MeO
CN
CN
OMe
EMK 151
EMK 161
283.6
163.0
9.0
ND
13.5
9.6
13.8
3.4
O2N
CN
CN
CN
CN
MeS
MeO
EMK 171
EMK 181
118.3
133.5
ND
ND
5.5
8.8
5.0
2.8
CN
CN
F
F
CN
CN
MeO
EMK 191
EMK 201
98.4
ND
ND
12.9
7.8
2.9
3.8
CN
CN
153.5
HO
CN
CN
F
F
EMK 211
4.5
ND
4.0
1.4
CN
CN
HO

K. Turpaev et al. / Biochemical Pharmacology 82 (2011) 535–547
539
Table 1 (Continued )
Compounds
Code
Induction/activation (Fold-change)
Structure
HO-1 mRNA
188.1
HO-1 protein
Nrf2
6.4
p-cJun
EMK 221
5.4
5.5
F
CN
CN
EMK 231
EMK 241
135.7
81.3
ND
ND
5.8
8.9
5.1
5.3
MeO
t-Bu
CN
CN
CN
CN
EMK 251
EMK 281
10.0
ND
7.0
9.7
1.1
CN
CN
N
O
145.4
12.4
12.8
O2N
CN
CN
NO2
EMK 301
EMK 311
7.7
1.6
ND
ND
1.5
1.7
8.8
6.4
O2N
O2N
O2N
O2N
CO2Et
CN
CO2Et
SO2Me
SO2Me
CN
NO2
EMK 321
EMK 381
EMK 421
1.3
ND
ND
9.6
1.1
1.0
7.0
1.9
1.5
3.6
1.0
169.5
CN
CN
EMK 431
EMK 441
EMK 461
EMK 471
EMK 491
246.4
276.3
232.3
93.2
8.3
12.1
5.7
6.8
9.1
ND
6.5
11.5
6.4
4.8
ND
4.0
29.6
2.7
7.2
CN
CN
CN
CN
CN
CN
F
CN
Br
CN
CN
OMe
245.5
CN
CN
NO2
EMK 511
EMK 531
EMK 591
270.6
260.6
272.5
10.9
11.5
14.2
ND
ND
4.9
ND
ND
3.5
NC
Cl
CN
CN
CN
CN
CN
CN
NC

540
K. Turpaev et al. / Biochemical Pharmacology 82 (2011) 535–547
Table 1 (Continued )
Compounds
Code
Induction/activation (Fold-change)
Structure
HO-1 mRNA
ND
HO-1 protein
Nrf2
4.3
p-cJun
3.8
EMK 611
9.4
6.0
ND
CN
CN
Cl
Br
EMK 621
EMK 661
EMK 711
240.5
ND
4.8
5.3
4.5
8.3
CN
CN
11.7
2.1
CN
CN
F3C
136.4
9.6
2.9
OMe
OMe
MeO
CN
CN
EMK 721
25.7
4.3
1.3
CN
CN
MeO
MeO
EMK 731
EMK 841
31.5
31.3
4.1
6.3
4.8
4.2
1.6
1.2
CN
CN
MeO
MeO
OMe
CN
CN
MeO
MeO
EMK 851
EMK 861
EMK 871
40.4
2.1
4.0
1.3
1.7
3.5
2.1
2.6
1.5
1.1
1.2
CN
CN
MeO
MeO
OMe
CN
CN
MeO
MeO
OMe
2.0
CN
CN
HO
OMe
EMK 991
ND
ND
6.9
ND
4.3
7.1
3.2
5.0
CN
CN
EMK 1041
CN
CN
N
EMK 1051
EMK 1071
ND
1.0
2.9
2.3
7.9
1.0
1.8
CN
CN
N
H
40.5
CN
CN
S
U937 cells were exposed to the indicated chemicals (25 mM) for 3 h. Relative mRNA and protein levels were measured by RT-PCR and Western blotting, respectively. The
mRNA and protein (phosphoprotein) levels in untreated cells were assigned a value of 1. Data represent the mean of at least three independent experiments. BMN compounds
referred to here as EMK 031, EMK 041, EMK 071, EMK131, and EMK 251 are also known as tyrphostins A16, A15, A19, A2, and A6, respectively [1]. ND, not done.

K. Turpaev et al. / Biochemical Pharmacology 82 (2011) 535–547
541
Fig. 2. Activation of signaling systems by selected BMNs or structurally similar compounds. (A) Representative Western blot image. U937 cells were incubated with BMNs
(25 M) for 3 h. Whole cell lysates were analyzed by immunoblotting with Nrf2, phospho-cJun, phospho-p38, phospho-p70S6K, and -actin antibodies. (B) Relative
m
b
induction of HO-1 mRNA plotted against Nrf2 level. (C) Relative induction of HO-1 mRNA plotted against p-c-Jun level. Western blot or RT-PCR quantification data of three
independent experiments are shown. HO-1 mRNA and protein levels in untreated cells were assigned a value of 1.
phosphorylation were weak HO-1 inducers. Conversely, com-
pounds which actively stimulated both c-Jun phosphorylation and
Nrf2 accumulation (e.g. BMNs containing a NO2 group) are very
potent activators of HO-1 gene expression. Altogether, these data
suggest that parallel actions of Nrf2 and p-c-Jun is a prerequisite
for full-scale HO-1 induction, and subsequently for accumulation
of HO-1 protein.
selected. In control experiments, survival of U937 cells decreased
as menadione concentration increased, and at concentration over
40
mM almost no cell survived. Fig. 3A shows a representative plot
of the protective effect of prior incubation with AG-126 on survival
of U937 cells exposed to menadione. Cell death reached 95% after
exposure to 30
were pretreated with 25
m
M menadione, but declined to 55% when cells
M AG-126 for 14 h. Menadione toxicity
m
Activation of p38 phosphorylation was markedly stimulated by
compounds bearing NO2- and halogen benzene ring substituents
(Supplemental Table 1). Regarding p-p70S6 K, its level was
strongly reduced (FC = 0.16) in cells exposed to EMK 281 whereas
FC was between 0.4 and 0.5 for EMK 311, EMK 431, EMK 211, EMK
491, EMK 621, and EMK 151 (see Fig. 2A and Supplemental Table
1). However, only slight to moderate correlation was noted
between accumulation of HO-1 mRNA and either p-p38 or p-p70SK
content in U937 cells treated with BMNs (data not shown).
on U937 cells was quantified by calculating its median-effect
concentration (Dm) upon 5-h exposure (Fig. 3B) was 13.3 ꢅ 1.9 mM.
Interestingly, prior incubation of U937 cells with AG-126 for 6, 14 or
22 h caused a Dm increase of 34, 66 or 68%, respectively. U937 cells
were also pretreated with different BMN compounds (25 mM) for
14 h prior to exposure to menadione for 5 h. Even though they are
structurally different from AG-126, EMK 121 and EMK 201 had a
similar protective effect whereas BMN compounds with MeO
substitutions (EMK 041, EMK 131, EMK 721, EMK 731, EMK 841)
(Table 2) were weaker protectors. Subsequent analysis revealed a
significant correlation between HO-1 induction and the protective
properties of the analyzed BMN compounds. Pearson correlation
coefficients (rX,Y) between Dm increase and induction of HO-1 at the
mRNA and protein levels, were 0.683 (n = 12) and 0.784 (n = 9),
respectively, and the linear correlation coefficients (R2) were 0.467
(p < 0.02) and 0.615 (p < 0.02), respectively (Fig. 3C and D).
3.4. Protective effects of BMN compounds upon oxidative stress
To generate oxidative stress, we exposed U937 cells to
menadione, which induces cell death through superoxide genera-
tion and depletion of intracellular SH-containing molecules.
Twelve compounds with different capacity to induce HO-1 were

542
K. Turpaev et al. / Biochemical Pharmacology 82 (2011) 535–547
Fig. 3. Protective effects of BMNs upon oxidative stress. (A) Fraction of necrotic cells (Fa) plotted against menadione concentration. Cells were exposed to menadione for 5 h
after being pre-treated or not with 25 M AG-126 (14 h). (B) Analysis of the above data by the median effect plot. Fu is the fraction of unaffected cells. The results are analyzed
by plotting log (Fa/Fu) with respect to log of menadione concentration (0–60 M). (C) Correlations between HO-1 mRNA accumulation or (D) HO-1 protein content and the
m
m
median concentration values (Dm) of BMNs (12 and 9 compounds for C and D, respectively). Dm is the concentration of menadione required to produce a 50% toxic effect, i.e.,
when Fa = Fu. Each Dm value represents the mean of three independent experiments.
3.5. Spectroscopic evidence for reactivity of AG-126 with GSH
and EMK 611 conjugate with GSH, the most abundant cellular thiol
compound [17]. UV–vis spectrum of AG-126 (50 M) with GSH
m
In cells of different types, activation of redox-sensitive signaling
pathways is triggered mainly by modification of sensitive cysteine
residues on specific cellular targets. It was shown that chlor-
obenzylidene malononitriles referred to in this study as EMK 531
(1 mM) in 50 mM Tris HCl buffer (pH 7.4) showed a slight
hypsochromic shift from 281 to 278 nm with concomitant increase
of intensity of the absorption maximum (Fig. 4). These spectral
alterations suggest that AG-126 forms a conjugate with GSH.

K. Turpaev et al. / Biochemical Pharmacology 82 (2011) 535–547
543
Table 2
Protective effects of BMN compounds upon menadione toxicity.
Compounds
Dm
(
ꢅSD)
Dm increase (mM)
AG-126
24.4 ꢅ 2.5
20.0 ꢅ 2.2
23.0 ꢅ 2.3
25.2 ꢅ 2.7
19.9 ꢅ 2.1
19.6 ꢅ 2.2
23.4 ꢅ 2.4
20.3 ꢅ 1.9
16.2 ꢅ 1.8
18.1 ꢅ 2.5
20.8 ꢅ 2.0
18.4 ꢅ 1.9
11.1
6.7
EMK 041
EMK 081
EMK 121
EMK 131
EMK 151
EMK 201
EMK 281
EMK 721
EMK 731
EMK 841
EMK 1071
9.7
11.9
6.6
6.3
10.1
7.0
2.9
4.8
7.5
5.1
U937 cells were exposed or not to the indicated BMN compounds (25
mM) for 14 h
prior to treatment with menadione (0–60 M) for an additional 5-h period. The
m
data are expressed as the variation of the median-effect concentration (Dm) in BMN
preconditioned cells relative to untreated cells, and are the mean of at least three
independent experiments. The Dm value for cells exposed to menadione alone was
13.3 ꢅ 1.9 mM.
Fig. 4. Spectrophotometric analysis of AG-126 reaction with GSH. AG-126 (50
mM)
was incubated with (dotted line) or without (solid line) 1 mM GSH in 50 mM Tris–
HCl buffer containing 5% of DMSO (pH 7.5) for 1 h at room temperature.
that this compound reacts with a single molecule of GSH (Fig. 6B).
Likewise, reaction of EMK 131 (3,4-(MeO)2 BMN) with GSH, gave a
single product with elution retention time of 1.64 min. ESI-MS2
analysis revealed an m/z value of 827.5 corresponding to conjugate
between EMK 131 and two GSH (Fig. 6C). As regards EMK 721 (2,4-
(MeO)2 BMN), ELSD analysis detected a single product resulting
from conjugation of one GSH and one EMK 721 molecule (m/
3.6. Correlations between chemical activity of BMN compounds and
HO-1 mRNA induction
The reactivity of some BMNs and related Michael acceptors
toward thiols has previously been studied [17,18]. The reaction
proceeds via a fast reversible addition of a thiol to the conjugated
system. Equilibrium constants for reactions with n-butanethiol of
BMN derivatives referred here to as EMK 031, EMK 721, AG-9, EMK
081, EMK 471, EMK 491, EMK 191, EMK 231, EMK 621, EMK 441,
EMK 531, EMK 151, and EMK 121 range from 0.13 to 12.0 ꢁ 102
(Mꢀ1) (25 8C, pH 7.0) [18]. We analyzed the correlation between
BMN reaction rates with thiols by plotting the equilibrium
constant for the reactions between BMNs and n-butanethiol [18]
versus the capacity to induce HO-1 gene expression (our data from
Table 1). As shown in Fig. 5A, there is a clear correlation between
these two parameters (R2 = 0.674,
rX,Y = 0.821, n = 14).
Moreover, as shown in Fig. 5B, there is a positive correlation
between the hydrolysis rate of BMNs [19] and the capacity to
induce HO-1 mRNA expression (R2 = 0.742,
rX,Y = 0.861, n = 12).
Calculated half-lives of these BMN compounds in aqueous solution
range from 2.5 min to 38.3 h. Surprisingly, the short lifespan of
BMN compounds in water did not reduce their capacity to induce
HO-1 mRNA. Thus, the compounds with a short half-life such as
EMK 121, EMK 151 and EMK 531 (calculated t1/2 = 2.5, 3.2 and
9.7 min, respectively) were the most effective activators of HO-1
gene expression.
3.7. LC–MS analysis of BMN adducts with GSH
In order to further characterize GSH conjugates with BMNs, we
analyzed the reaction mixtures of a selection of BMNs with
reduced GSH (1 h, 25 8C) by LC–MS using electrospray ionization
(ESI) in positive and negative mode. The compounds tested (AG-
126, EMK 131, EMK 721, EMK 731, and EMK 1071) were selected
because they differently stimulate HO-1 expression and activate
upstream signaling systems (Table 1). As shown in Fig. 6A,
evaporative light scattering detector (ELSD) revealed a single
product of the reaction between AG-126 and GSH eluting at
1.50 min. Detailed ESI-MS2- analysis produced a peak with an m/z
value of 828.5, corresponding to the conjugate of AG-126 with two
equivalents of GSH and gave rise to fragments at 521 (R-GSH) and
306 (GSH). The thiophene malononitrile EMK 1071, after reaction
with GSH, produced a compound with an ELSD retention time of
2.00 min. ESI-MS2-analysis showed two major peaks correspond-
ing to the mono- and di-charged ions (482.2 and 241.7), suggesting
Fig. 5. Correlation between HO-1 gene expression and chemical properties of BMN
compounds. The relative levels of HO-1 mRNA (data from Table 1) were plotted
against equilibrium constants (Keq ꢁ 10ꢀ2) for reactions with n-butanethiol [18] (A)
and BMN hydrolysis first-order constants ðkH ꢁ 104Þ [19] (B).
O
2

544
K. Turpaev et al. / Biochemical Pharmacology 82 (2011) 535–547
Fig. 6. LC–MS analysis of adducts between GSH and BMNs. ELSD chromatograms and subsequent ESI-MS2 analysis of the reaction products. Bottom panels compare the
experimental m/z and calculated MW values of initial compounds and expected GSH conjugates. Compounds were incubated with 5 mM GSH in 50 mM Tris–HCl buffer
containing 10% of MeOH (pH 7.5) for 1 h at room temperature. The incubation of AG-126 (A) and EMK 1071 (B) with GSH resulted in generation of single ELSD detectable
products with retention time 1.50 and 2.00 min, respectively. ESI-MS2 analyses suggest that AG-126 conjugates with two GSH equivalents whereas EMK 1071 conjugates
with one GSH molecule. The product of EMK 131 (C) and EMK 721 (D) reactions with GSH were revealed as a single peaks (ELSD retention times 1.64 and 2.20 min,
respectively) corresponding to BMN adducts with a one molecule of GSH. The incubation of EMK 731 (E) with GSH resulted in the generation of two products revealed by ELSD
as peaks at retention times of 1.37 and 1.91 min. ESI-MS2 analysis suggests that the first peak corresponds to EMK 731 conjugated with two GSH equivalents whereas the
second one corresponds to EMK 731 conjugate with a single GSH molecule.
z = 522.2) (Fig. 6D). Lastly, the reaction between EMK 731 (3,4-
(MeO)2 BMN) and GSH generated two products with elution
retention time of 1.4 and 1.9 min consistent with the addition of
two (m/z = 829.4) and one (m/z = 522.3) molecule(s) of GSH to one
molecule of EMK 731, respectively (Fig. 6E). It is worth noting that
LC–MS analysis performed after 1-h and 24-h incubation of AG-
126 with GSH revealed the same composition of the reaction
mixture (data not shown), suggesting that these conjugates of
BMN with GSH are thermodynamically stable.
expression 90-fold [20]. Here, we point to a significant correlation
between BMN-mediated HO-1 induction and decrease in menadi-
one-induced toxicity, suggesting that HO-1 up-regulation is an
indicator of drug efficacy. In addition, our investigation of the
structure–activity relationship of 49 BMNs points out structure
substitutions that determine specific responsiveness of major
signaling pathways.
HO-1 gene expression is activated mainly at the transcriptional
level in various cell types by a broad spectrum of chemical and
physical agents including heat shock, ultraviolet light, hypoxia,
cytokines, reactive oxygen and nitrogen species, and xenobiotics
[21,22]. The regulation of HO-1 expression is mediated by a
complex network of transcriptional regulators acting on many
responsive elements located in the promoter region of the HO-1
gene: antioxidant response element (ARE), Maf recognition
element (MARE), hypoxia response elements (HREs), heat shock
element (HSE), and multiple binding sites for transcription factors
4. Discussion
In a previous study, we showed that BMN compounds are
potent activators of HO-1 expression [11]. Thus, in U937 cells
exposed to AG-126, the level of HO-1 mRNA increased to more
than 300 times the basal level, as compared with the synthetic
triterpenoid CDDO-imidazolide which stimulates HO-1 mRNA

K. Turpaev et al. / Biochemical Pharmacology 82 (2011) 535–547
545
Fig. 7. Scheme of a proposed mechanism of BMN-dependent signaling and regulation of HO-1 gene expression. In the absence of BMN, the thiol-rich sensor protein Keap1
retains Nrf2 in the cytoplasm where it is targeted for ubiquity-mediated degradation. BMN compounds alkylate reactive SH-groups of Keap1 resulting in Nrf2 dissociation,
nuclear translocation and binding to ARE regulatory elements in heteromeric combination with bZIP family transcription factors (c-Jun, small Maf proteins, ATF4). In parallel,
BMN compounds can modify other regulatory proteins, resulting in up-regulation of MAP kinases and AP-1 family transcription factors. MAP kinases can regulate Nrf2
stability and induce gene expression via the AP-1 binding sites TPA-responsive element (TRE) and c-AMP-responsive element (CRE). Multiple copies of AP-1 binding sites are
present in the regulatory region of HO-1 gene.
AP-1, C/EBP, STAT, and NF-
k
B. In brief, the master regulator Nrf2
various Michael reaction acceptors like olefins conjugated to
electron-withdrawing groups [35,36]. A common feature of these
structurally different compounds is their high reactivity with
sulfhydryl groups.
plays a central role in induction of HO-1 in partnership with other
transcription factors including small Maf proteins, c-Jun and ATF4
[23–26].
Our structure–activity relationship analysis shows that BMN
compounds have different structural requirements for activation
of Nrf2 and c-Jun cellular signaling pathways (Table 1). It is
interesting to note that with the exception of EMK 841, none of the
14 BMNs displaying either Nrf2 or p-c-Jun low expression
(FC ꢆ 1.7) was found to elicit robust (FC > 10) HO-1 mRNA
induction (Table 1). These observations suggest that the parallel
increase in cellular Nrf2 and p-c-Jun is a prerequisite for full-scale
HO-1 gene induction by BMNs. This assertion is supported by our
previous observation that U937 cell treatment with tyrphostin AG-
126 and synthetic NO donor DPTA-NO, a potent c-Jun activator,
had a cumulative effect on expression of the ferritin H and IL-8
genes [11] which, like HO-1, are under the control of Nrf2/ARE and
AP-1 signaling [27,28]. Nrf2/ARE signaling can be activated
through Nrf2 phosphorylation by MAP kinases and other
regulatory protein kinases that facilitate Nrf2 nuclear translocation
[29,30]. It is therefore possible that HO-1 induction by BMN
compounds is mediated by JNK-dependent phosphorylation of
Nrf2, as previously shown for PC-3 cells exposed to isothiocyanates
[29]. As summarized in the scheme proposed in Fig. 7, interplay
between different signaling networks shows that the BMN-
induced activation of HO-1 expression can be explained by the
binding of the Nrf2-c-Jun complex to ARE regulatory sequences
located in the enhancer region [29,30].
Nrf2 functional activity is regulated negatively by Keap1, a
cytoplasmic factor that binds Nrf2 and the actin cytoskeleton, and
has ubiquitin ligase adaptor activity. Association of Keap1 with
Nrf2 retains the latter in the cytoplasm and promotes its rapid
proteosomal degradation. Keap1 contains several critical thiol
groups, particularly Cys151, Cys273, Cys278, and their oxidation by
reactive oxygen species or alkylation by electrophiles enables
dissociation and nuclear translocation of Nrf2 [31–34]. Chemical
substances activating Keap1/Nrf2 pathway were classified by
Talalay and co-workers as 10 distinct groups that include
quinones, isothiocyanates, dimercaptans, heavy metals, and
Our results indicate that potency of BMN compounds to induce
HO-1 mRNA induction is related to their properties of Michael
reaction acceptors. Thus, the presence of two strong electron-
withdrawing CN groups conjugated to a C55C bond, is a critical
requirement for transcriptional activation of HO-1 gene. Further-
more, our data showing that reactivity of BMNs and related
Michael acceptors with thiols is correlated with their capacity to
induce HO-1 expression, suggest that BMN compounds can
modify Keap1 and others cytoplasmic regulatory factors by
linking to their critical protein sulfhydryl groups. Binding to Keap-
1 would promote release and translocation of Nrf2 to the nucleus
where it forms heterodimers with partner proteins (e.g. c-Jun),
binds to the ARE, and activate gene expression. In addition, our
finding that BMN reactivity with GSH correlates with the rate of
HO-1 mRNA induction is in good agreement with previous results
obtained by Talalay’s group showing that the reactivity of
benzylideneketone acceptors with sulfhydryl reagents is the
major factor that determines their potency to induce cell
protective genes [36]. Further,
a quantum chemical study
established significant structure–activity relationships between
electron-acceptor properties of triterpenoids and their biological
potencies [37].
LC–MS analysis revealed that reaction of GSH with BMN
compounds could generate two types of conjugates, one with two
GSH molecules and another with one GSH molecule. Remarkably,
AG-126 and EMK 131 which generate di(GSH) conjugates, induced
HO-1 mRNA expression and c-Jun activation more potently than
EMK 721, EMK 1071, and EMK 731 which generate mono-GSH
conjugates. The ability of several BMN compounds to generate
conjugates with two GSH equivalents suggests they can yield intra-
or inter-molecular bridges between low- or high-molecular weight
thiols. It is tempting to speculate that GSH, the most abundant
cellular thiol, may serve as a carrier molecule for BMNs, protecting
them from a simple hydrolysis, and liberating them upon cell
demands.

546
K. Turpaev et al. / Biochemical Pharmacology 82 (2011) 535–547
[7] Novogrodsky A, Weisspapir M, Patya M, Meshorer A, Vanichkin A. Tyrphostin
A large body of experimental data shows that under physiolog-
4-nitrobenzylidene malononitrile reduces chemotherapy toxicity without
impairing efficacy. Cancer Res 1998;58:2397–403.
[8] Vanichkin A, Patya M, Lagovsky I, Meshorer A, Novogrodsky A. 4-Nitrobenzy-
lidene malononitrile reduces apoptosis-mediated liver injury in mice. J Hepa-
tol 2002;36:631–6.
[9] Kann O, Hoffmann A, Schumann RR, Weber JR, Kettenmann H, Hanisch UK. The
tyrosine kinase inhibitor AG126 restores receptor signaling and blocks release
functions in activated microglia (brain macrophages) by preventing a chronic
rise in the intracellular calcium level. J Neurochem 2004;90:513–25.
[10] Chatterjee PK, Patel NS, Kvale EO, Brown PA, Stewart KN, Britti D, et al. The
tyrosine kinase inhibitor tyrphostin AG126 reduces renal ischemia/reperfu-
sion injury in the rat. Kidney Int 2003;64:1605–19.
[11] Turpaev K, Drapier JC. Stimulatory effect of benzylidenemalononitrile tyr-
phostins on expression of NO-dependent genes in U-937 monocytic cells. Eur J
Pharmacol 2009;606:1–8.
ical conditions, BMN compounds are able to stimulate the
resistance of cells and tissues against different harmful challenges.
For example, tyrphostin AG-1714 (4-nitrobenzylidene malononi-
trile, referred to here as EMK 121) provides significant protection
against mortality induced by chemotherapy with cisplatin and
doxorubicin in animals and in cell cultures [7]. Remarkably, AG-
1714 stimulates resistance of healthy tissues to chemotherapy
without protecting malignant tissues against drug-mediated
suppression [7]. This selectivity may be associated with increased
basal HO-1 expression in various tumor cells [22]. Experiments
performed in vivo showed that tyrphostin AG-126 reduced
development of pneumococcal meningitis, kidney acute injury
and dysfunction after ischemia and reperfusion, development of
acute pancreatitis, consequences of spinal cord trauma, and
mitochondria-mediated 7-ketocholesterol toxicity [6,10,38–41].
Our findings that BMN compounds are activators of cell protection
may provide clues to development of new drugs that mimic the
protective response.
[12] Kunnumakkara AB, Anand P, Aggarwal BB. Curcumin inhibits proliferation,
invasion, angiogenesis and metastasis of different cancers through interaction
with multiple cell signaling proteins. Cancer Lett 2008;269:199–225.
[13] Sporn MB, Suh N. Chemoprevention: an essential approach to controlling
cancer. Nat Rev Cancer 2002;2:537–43.
[14] Harper DAR, Steenson BE. Preparation of
a,b-unsaturated sulfones by the
Knoevenagel method. Synthesis 1980;806–7.
[15] Bumagin NA, Andryukhova NP, Beletskaya IP. Arylation of acrylonitrile with
aryl halides catalyzed by palladium complexes. Dokl Akad Nauk SSSR
1990;313:107–9.
In conclusion, our study discloses that full activation of HO-1
expression by BMNs requires parallel up-regulation of two major
signaling pathways, the Nrf2/Keap-1 system and the c-Jun/JNK
pathway. This finding provides a possible tool for pharmacological
modulation of intercellular signaling. Many short-distance signal-
ing molecules like prostaglandins, NO, regulatory peptides, and
cytokines activate JNK MAP kinases and c-Jun phosphorylation. It is
thus likely that BMN compounds dedicated to Nrf2 activation will
up-regulate protective systems exclusively in tissues that produce
c-Jun-activating signaling molecules. According to these guide-
lines, the most suitable molecules are 4-morphlino BMN (EMK
251), 4-OH BMN (AG-10, EMK 031, EMK 071) and thiophene
malononitrile EMK 1071. However, a potential drawback of AG-10
is its adverse effect, as a phenolic compound, on mitochondrial
membrane polarity [3]. Nonetheless, because of the different
capacities of BMN hydrolysis products to cross membranes, we
suggest that BMNs chiefly react with molecular targets located in
proximity to vasculature.
[16] Chou TC, Talalay P. Quantitative analysis of dose–effect relationships: the
combined effects of multiple drugs or enzyme inhibitors. Adv Enzyme Regul
1984;22:27–55.
[17] Rietveld EC, Hendrikx MM, Seutter-Berlage F. Glutathione conjugation of
chlorobenzylidene malononitriles in vitro and the biotransformation to mer-
capturic acids in rats. Arch Toxicol 1986;59:228–34.
[18] Pritchard RB, Lough CE, Currie DJ, Holmes HL. Equilibrium reactions of n-
butanethiol with some conjugated heteroenoid compounds. Can J Chem
1968;46:775–81.
[19] Pritchard RB, Lough CE, Reesor JB, Holmes HL, Currie DJ. The relative rates of
reaction of potassium cyanide and water with substituted benzalmalononi-
triles. Can J Chem 1967;45:775–7.
[20] Liby K, Hock T, Yore MM, Suh N, Place AE, Risingsong R, et al. The synthetic
triterpenoids, CDDO and CDDO-imidazolide, are potent inducers of heme
oxygenase-1 and Nrf2/ARE signaling. Cancer Res 2005;65:4789–98.
[21] Ryter SW, Alam J, Choi AM. Heme oxygenase-1/carbon monoxide: from basic
science to therapeutic applications. Physiol Rev 2006;86:583–650.
[22] Was H, Dulak J, Jozkowicz A. Heme oxygenase-1 in tumor biology and therapy.
Curr Drug Targets 2010;11:1551–70.
[23] Copple IM, Goldring CE, Kitteringham NR, Park BK. The Nrf2-Keap1 defense
pathway: role in protection against drug-induced toxicity. Toxicology
2008;246:24–33.
[24] Levy S, Jaiswal AK, Forman HJ. The role of c-Jun phosphorylation in EpRE
activation of phase II genes. Free Radic Biol Med 2009;47:1172–9.
[25] Jaiswal AK. Nrf2 signaling in coordinated activation of antioxidant gene
expression. Free Radic Biol Med 2004;36:1199–207.
[26] Lee AC, Murray M. Up-regulation of human CYP2J2 in HepG2 cells by butylated
hydroxyanisole is mediated by c-Jun and Nrf2. Mol Pharmacol 2010;77:987–94.
[27] Pietsch EC, Chan JY, Torti FM, Torti SV. Nrf2 mediates the induction of ferritin H
in response to xenobiotics and cancer chemopreventive dithiolethiones. J Biol
Chem 2003;278:2361–9.
Acknowledgments
This work was supported by funding from CNRS. We thank
Odile Thoison for discussion and help with performing LC/MS
`
analyses and Genevieve Aubert for excellent technical assistance.
[28] Loboda A, Stachurska A, Florczyk U, Rudnicka D, Jazwa A, Wegrzyn J, et al. HIF-
1 induction attenuates Nrf2-dependent IL-8 expression in human endothelial
cells. Antioxid Redox Signal 2009;11:1501–17.
[29] Xu C, Yuan X, Pan Z, Shen G, Kim JH, Yu S, et al. Mechanism of action of
isothiocyanates: the induction of ARE-regulated genes is associated with
activation of ERK and JNK and the phosphorylation and nuclear translocation
of Nrf2. Mol Cancer Ther 2006;5:1918–26.
Appendix A. Supplementary data
[30] Anwar AA, Li FY, Leake DS, Ishii T, Mann GE, Siow RC. Induction of heme
oxygenase 1 by moderately oxidized low-density lipoproteins in human
vascular smooth muscle cells: role of mitogen-activated protein kinases
and Nrf2. Free Radic Biol Med 2005;39:227–36.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bcp.2011.05.028.
[31] Sekhar KR, Rachakonda G, Freeman ML. Cysteine-based regulation of the CUL3
adaptor protein Keap1. Toxicol Appl Pharmacol 2010;244:21–6.
[32] Dinkova-Kostova AT, Holtzclaw WD, Kensler TW. The role of Keap1 in cellular
protective responses. Chem Res Toxicol 2005;18:1779–91.
[33] Wakabayashi N, Dinkova-Kostova AT, Holtzclaw WD, Kang MI, Kobayashi A,
Yamamoto M, et al. Protection against electrophile and oxidant stress by
induction of the phase 2 response: fate of cysteines of the Keap1 sensor
modified by inducers. Proc Natl Acad Sci USA 2004;101:2040–5.
[34] Fourquet S, Guerois R, Biard D, Toledano MB. Activation of NRF2 by nitrosative
agents and H2O2 involves KEAP1 disulfide formation. J Biol Chem 2010;285:
8463–71.
[35] Dinkova-Kostova AT, Fahey JW, Talalay P. Chemical structures of inducers of
nicotinamide quinone oxidoreductase 1 (NQO1). Methods Enzymol 2004;382:
423–48.
[36] Dinkova-Kostova AT, Massiah MA, Bozak RE, Hicks RJ, Talalay P. Potency of
Michael reaction acceptors as inducers of enzymes that protect against
carcinogenesis depends on their reactivity with sulfhydryl groups. Proc Natl
Acad Sci USA 2001;98:3404–9.
References
[1] Gazit A, Yaish P, Gilon C, Levitzki A. Tyrphostin I: synthesis and biological
activity of protein tyrosine kinase inhibitors. J Med Chem 1989;32:2344–52.
[2] Levitzki A, Mishani E. Tyrphostins and other tyrosine kinase inhibitors. Annu
Rev Biochem 2006;75:93–109.
[3] Sagara Y, Ishige K, Tsai C, Maher P. Tyrphostins protect neuronal cells from
oxidative stress. J Biol Chem 2002;277:36204–15.
[4] Soltoff SP. Evidence that tyrphostins AG10 and AG18 are mitochondrial
uncouplers that alter phosphorylation-dependent cell signaling. J Biol Chem
2004;279:10910–8.
[5] Novogrodsky A, Vanichkin A, Patya M, Gazit A, Osherov N, Levitzki A. Preven-
tion of lipopolysaccharide-induced lethal toxicity by tyrosine kinase inhibi-
tors. Science 1994;264:1319–22.
[6] HanischUK,PrinzM, AngstwurmK, Ha¨uslerKG, KannO, KettenmannH, etal. The
protein tyrosine kinase inhibitor AG126 prevents the massive microglial cyto-
kine induction by pneumococcal cell walls. Eur J Immunol 2001;31:2104–15.

K. Turpaev et al. / Biochemical Pharmacology 82 (2011) 535–547
547
[37] Bensasson RV, Zoete V, Berthier G, Talalay P, Dinkova-Kostova AT. Potency
ranking of triterpenoids as inducers of a cytoprotective enzyme and as
inhibitors of a cellular inflammatory response via their electron affinity and
their electrophilicity index. Chem Biol Interact 2010;186:118–26.
[38] Gonc¸alves S, Fernandez-Sanchez R, Sanchez-Nin˜o MD, Tejedor A, Neria F, Egido
J, et al. Tyrphostins as potential therapeutic agents for acute kidney injury.
Curr Med Chem 2010;17:974–86.
tins AG 126 and AG556 modulates murine experimental acute pancreatitis.
Surgery 2005;138:913–23.
[40] Genovese T, Mazzon E, Esposito E, Muia` C, Di Paola R, Crisafulli C, et al.
Inhibition of tyrosine kinase-mediated cellular signalling by Tyrphostins
AG126 and AG556 modulates secondary damage in experimental spinal cord
trauma. Neuropharmacology 2007;52:1454–71.
[41] Kim YJ, Lee CS. Tyrosine kinase inhibitor AG126 reduces 7-ketocholesterol-
induced cell death by suppressing mitochondria-mediated apoptotic process.
Neurochem Res 2010;35:603–12.
[39] Balachandra S, Genovese T, Mazzon E, Di Paola R, Thiemerman C, Siriwardena
AK, et al. Inhibition of tyrosine-kinase-mediated cellular signaling by tyrphos-