Activation of the NRF2 Signaling Pathway by Copper-Mediated Redox Cycling of Para- and Ortho-Hydroquinones

Article abstract of DOI:10.1016/j.chembiol.2009.12.013

Transcription factor NF-E2 p45-related factor 2 (Nrf2) mediates adaptation to oxidants and electrophiles through up-regulating genes that contain antioxidant response elements (AREs) in their promoters. Using the stably transfected human AREc32 reporter cell line, we found that copper and other transition metals enhanced induction of ARE-driven luciferase by 2-tert-butyl-1,4-hydroquinone (tBHQ) as a result of increased oxidation to 2-tert-butyl-1,4-benzoquinone (tBQ). Following exposure to tBHQ for 30 min, ARE-luciferase activity measured after 24 hr was dependent on the presence of Cu²⁺. In contrast, tBQ-induced activity was Cu²⁺-independent. The metal-catalyzed oxidation of tBHQ to tBQ occured rapidly and stoichiometrically. Compounds that share para- or ortho-hydroquinone structures, such as catechol estrogens, dopamine, and l-DOPA, also induced ARE-driven luciferase in a Cu²⁺-dependent manner. Thus, the oxidation of para- and ortho-hydroquinones to quinones represents the rate-limiting step in the activation of Nrf2.

Full text of DOI:10.1016/j.chembiol.2009.12.013

Chemistry & Biology
Article
Activation of the NRF2 Signaling Pathway by
Copper-Mediated Redox Cycling of Para- and
Ortho-Hydroquinones
1,2
Xiu Jun Wang,1,3 John D. Hayes,1 Larry G. Higgins,1 C. Roland Wolf,1, and Albena T. Dinkova-Kostova
*
1Cancer Research UK Molecular Pharmacology Unit, Biomedical Research Institute, Ninewells Hospital and Medical School,
University of Dundee, Dundee DD1 9SY, Scotland, UK
2Division of Clinical Pharmacology, Departments of Medicine and Pharmacology and Molecular Sciences, Johns Hopkins University School
of Medicine, Baltimore, MD 21205, USA
3Present address: Department of Pharmacology, College of Medicine, Zhejiang University, Hangzhou, PR China
*Correspondence: c.r.wolf@dundee.ac.uk
DOI 10.1016/j.chembiol.2009.12.013
SUMMARY
During oxidative or electrophilic stress, Keap1 is inactivated as
a consequence of modification of its highly reactive cysteine
Transcription factor NF-E2 p45-related factor 2 (Nrf2) residues (Eggler et al., 2005; Zhang, 2006). Under such circum-
stances, the stability of Nrf2 increases and it accumulates in the
nucleus, where after dimerization with a small Maf (Musculo-
mediates adaptation to oxidants and electrophiles
through up-regulating genes that contain antioxidant
response elements (AREs) in their promoters. Using
the stably transfected human AREc32 reporter cell
Aponeurotic Fibrosarcoma) protein (Motohashi and Yamamoto,
2004; Nioi et al., 2003), it transactivates genes that contain an
ARE in their promoter regions.
line, we found that copper and other transition metals
Nrf2 has been widely studied from a pharmacology and toxi-
enhanced induction of ARE-driven luciferase by
2-tert-butyl-1,4-hydroquinone (tBHQ) as a result of
cology perspective. This body of work has revealed that the
CNC-bZIP transcription factor is activated by many thiol-reac-
increased oxidation to 2-tert-butyl-1,4-benzoqui-
tive electrophilic xenobiotics and can mediate resistance against
these compounds, in part because it controls GSH synthesis.
Evidence also suggests that Nrf2 is activated by redox-cycling
none (tBQ). Following exposure to tBHQ for 30 min,
ARE-luciferase activity measured after 24 hr was
dependent on the presence of Cu2+. In contrast, agents and can confer protection against this class of chemicals,
tBQ-induced activity was Cu2+-independent. The
but that the resistance, at least in the case of menadione, is inde-
pendent of GSH (Higgins et al., 2009). Relatively little is known
metal-catalyzed oxidation of tBHQ to tBQ occured
rapidly and stoichiometrically. Compounds that
about how quinones and hydroquinones activate Nrf2. This is,
however, an important issue because several endogenous
share para- or ortho-hydroquinone structures, such
hydroquinones exist, such as 4-hydroxyestradiol and dopamine,
as catechol estrogens, dopamine, and L-DOPA, also
that could influence Nrf2 activity. Thus, 4-hydroxyestradiol and
dopamine could serve as endogenous second messengers to
induced ARE-driven luciferase in a Cu2+-dependent
manner. Thus, the oxidation of para- and ortho-
hydroquinones to quinones represents the rate-
control the expression of antioxidant genes in mammary or
neural tissues.
limiting step in the activation of Nrf2.
The findings that 1,2-hydroquinones (catechols, ortho-hydro-
quinones) and 1,4-hydroquinones (para-hydroquinones), but not
1,3-hydroquinones (resorcinols, meta-hydroquinones), induce
ARE-driven gene expression suggested that the oxidative lability
INTRODUCTION
Nrf2 (NF-E2 p45-related factor 2) is a cap‘n’collar (CNC) basic- of these compounds is essential for inducer activity (Prochaska
region leucine zipper (bZIP) transcription factor that allows et al., 1985; Rushmore et al., 1991). On the basis of molecular
adaptation to oxidative and electrophilic stress. It regulates orbital calculations, Bensasson et al. (2008) proposed a two-
a battery of genes for antioxidant enzymes and detoxifying step mechanism of induction of Nrf2-dependent enzymes by
proteins, including glutamate-cysteine ligase (comprising cata- oxidizable diphenols involving (1) oxidation of the diphenols to
lytic [GCLC] and modifier [GCLM] subunits), NAD(P)H:quinone their quinone derivatives, and (2) oxidation by these quinones
oxidoreductase 1 (NQO1), glutathione S-transferases, and aldo- of critical cysteine residues in Keap1 that are essential for its
keto reductases, through antioxidant response elements (AREs) ubiquitin ligase substrate adaptor activity, and thus repression
in the regulatory regions of these genes. Under homeostatic of Nrf2.
conditions, Keap1 (Kelch-like ECH-associated protein 1) sup-
Copper (Cu) is an essential trace element. It is present in many
presses the activity of Nrf2 by acting as an E3 ubiquitin ligase tissues at micromolar concentrations, and in some tissues,
substrate adaptor for Cul3/Rbx1 and targeting the CNC-bZIP including kidney and liver, its concentrations exceed 100 mmol/l
transcription factor for proteasomal degradation (Hayes and (Linder and Hazegh-Azam, 1996). In the body, copper exists
McMahon, 2009; Kensler et al., 2007; Nguyen et al., 2009). in both oxidation states, oxidized Cu2+ and reduced Cu+.
Chemistry & Biology 17, 75–85, January 29, 2010 ª2010 Elsevier Ltd All rights reserved 75

Chemistry & Biology
Copper-Dependent Activation of Nrf2 by Diphenols
Moreover, it is a redox-active transition metal and can catalyze
To confirm that Cu2+ can increase induction of endogenous
the oxidative activation of a number of phenolic compounds ARE-driven genes by tBHQ, western blotting was performed,
via a Cu2+/ Cu+ redox cycling mechanism that produces reactive and the expression of the endogenous ARE-dependent genes
oxygen species (ROS) (Li et al., 2002; Li and Trush, 1993).
encoding AKR1C, AKR1B10, and heme oxygenase 1 (HO-1)
Although tBHQ is a prototypic inducing agent, the initiating was evaluated (Devling et al., 2005; Lou et al., 2006; Reichard
events or species responsible for the activation of Nrf2 by the et al., 2007; MacLeod et al., 2009). Induction of AKR1C and
hydroquinone have not been well defined. Electron spin reso- AKR1B10 by 5 mM tBHQ was markedly increased by CuCl2
nance spectroscopy has revealed that in the presence of O2, (Figure 1C, lanes 2 and 4), and HO-1 only became detectable
Cu2+ mediates the one electron oxidation of tBHQ to a semiqui- in the presence of CuCl2, confirming the potentiating effect of
none anion radical (Li et al., 2002), which, in turn, reacts with Cu2+ on the activation of Nrf2 by tBHQ.
a second molecule of O2 to give O2$ꢀ and tBQ, with the subse-
$ꢀ
quent reoxidation of Cu+ by O2 to produce H2O2 and regen- Low Oxygen Inhibits Cu2+-Enhanced Induction
erate Cu2+ (Nakamura et al., 2003); 1,4-Hydroquinone can also of Luciferase Activity by tBHQ
be oxidized by Cu2+ in a similar way to tBHQ (Li and Trush, Oxidation of tBHQ by Cu2+ requires O2 (Li et al., 2002; Nakamura
1993). In the presence of cytochrome P450-dependent monoox- et al., 2003). When AREc32 cells were cultured in DMEM under
ygenases, 2- and 4-hydroxy estrogens are converted into semi- 1% (vol) O2, without the addition of Cu2+, 5 mM tBHQ induced
quinone radicals and eventually to their corresponding estrogen luciferase activity to a similar degree as it did under 21% O2 (Fig-
quinones. Conversion of these catechol estrogens to quinones ure 2A); this is probably because the hydroquinone is still
can also occur through a nonenzymatic mechanism involving capable of undergoing slow auto-oxidation to the quinone
Cu2+ (Li et al., 1994; Thibodeau and Paquette, 1999).
even under low oxygen conditions. In addition, under low oxygen
The finding that Cu2+ can promote oxidation of tBHQ has conditions, CuCl2 was still able to enhance the induction of lucif-
prompted us to question whether this redox-active transition erase activity by tBHQ ꢁ4-fold (from 5-fold to 20-fold). Most
metal might either augment or inhibit induction of ARE-driven importantly however, this potentiation was significantly lower
gene expression by hydroquinones. Furthermore, we tested than that observed under aerobic conditions (from 5-fold to
whether Cu2+ has an effect on induction of gene expression by 50-fold). The low oxygen conditions did not affect the induction
endogenous hydroquinones formed from estradiol or by dopa- of luciferase activity by SFN significantly (Figure 2B). These
mine. Using AREc32 cells (Wang et al., 2006), we demonstrate results indicate that Cu2+-enhanced activation of Nrf2 by tBHQ
that oxidation of diphenols to quinones is a critical step in the is dependent on the oxygen concentration, whereas activation
activation of Nrf2. In particular, Cu2+ efficiently mediates the of Nrf2 by SFN is not increased by Cu2+ nor is it oxygen-
redox-cycling of oxidizable diphenols. The concentration of dependent.
Cu2+ in the cell culture medium correlates with the potency of
these compounds as inducers of ARE-driven luciferase activity. 1,2- and 1,4-Diphenols Are Not Direct Inducers
To our knowledge, we provide the first experimental evidence of ARE-Driven Genes
that the para- and ortho-hydroquinones themselves are unable Quinones are electrophilic compounds that can readily react
to activate Nrf2 directly and that their corresponding quinones with thiols. To avoid interference by thiol compounds in the
react with Keap1 and are the direct activators of Nrf2. In addition, culture medium, we exposed AREc32 cells to 5 mM tBHQ in
we show that in the presence of Cu2+, the catechol estrogens PBS for only 30 min and then continued to grow the cells for
2-hydroxyestradiol, 4-hydroxyestradiol, 4-hydroxyestrone, as a further 24 hr in the growth medium to which the hydroquinone
well as dopamine, are potent Nrf2 activators.
had not been added. This short-time exposure to tBHQ had
no effect on the expression of luciferase, as determined by
measuring its activity 24 hr later (Figure 3 and Table 1). However,
when the 30 min exposure to 5 mM tBHQ was performed in the
presence of 50 mM CuCl2 or 100 mM CuCl, the ARE-driven lucif-
erase activity was induced ꢁ30-fold. Importantly, induction was
RESULTS
Cu2+ and tBHQ Synergistically Enhance ARE-Driven
Gene Expression
It has been reported that tBHQ can be oxidized by Cu2+ (Li et al., completely blocked by the metal chelators DFX (100 mmol/l) and
2002; Nakamura et al., 2003). To test whether copper has any EDTA (100 mmol/l) (Figure 3), indicating that Cu2+-mediated
effect on the activation of Nrf2 by tBHQ, CuCl2 was added to oxidation of tBHQ is essential to activate ARE-dependent gene
the culture medium of AREc32 cells. CuCl2 enhanced the induc- expression. Under identical conditions, the quinone tBQ induced
tion of ARE-driven luciferase activity by tBHQ in a dose-depen- luciferase activity in the absence of Cu2+, and this induction was
dent manner (Figure 1A). In the absence of CuCl2, treatment not affected by DFX or EDTA. These data further demonstrate
for 24 hr with 5 mM tBHQ resulted in ꢁ5-fold induction of lucif- that (1) Cu2+-mediated oxidation of tBHQ to the quinone tBQ
erase activity. However, in the presence of 10 mM CuCl2, the occurs rapidly, (2) oxidation of tBHQ to tBQ is an essential
induction by 5 mM tBHQ was doubled to 10-fold, and in the pres- step in inducing ARE-dependent gene expression, and (3) tBQ,
ence of 100 mM CuCl2, it was increased to ꢁ40-fold. In addition, and not tBHQ, is the ultimate inducing agent.
CuCl also enhanced the induction of ARE-driven luciferase
The conversion of 1,4-hydroquinones to quinones can be
activity by tBHQ in a dose-dependent manner (Figure 1B). In mediated by Cu2+ (Li et al., 1994; Thibodeau and Paquette,
contrast, Cu2+ had no effect on the induction by sulforaphane 1999). We therefore examined whether Cu2+ can potentiate the
(SFN), an isothiocyanate that activates Nrf2, but is not redox- ability of other oxidizable diphenols to activate ARE-driven
active.
reporter gene activity. As shown in Table 1, 1,4-hydroquinone
76 Chemistry & Biology 17, 75–85, January 29, 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
Copper-Dependent Activation of Nrf2 by Diphenols
Figure 1. Cu2+ and Cu+ Enhance tBHQ-Induced ARE-Driven Gene Expression
(A and B) AREc32 cells were exposed to (A) CuCl2 or (B) CuCl (10 to 200 mmol/l) in DMEM for 24 hr in the presence of 0.1% (v/v) DMSO, 5 mM tBHQ or 5 mM SFN.
The luciferase activity of cells treated with DMSO (control) was arbitrarily set at 1.
(C) Induction of AKR1C, AKR1B10, and HO-1 by tBHQ is enhanced by CuCl2. Cells were cultured in DMEM containing 5 mM tBHQ in the presence or absence of
50 mM CuCl2 at 21% (v/v) O2 for 24 hr. Whole cell extracts (30 mg of protein) were separated by 15% SDS-PAGE and the levels of AKR1C, AKR1B10, and HO-1
analyzed by western blotting. Immunoblotting for actin was used as the loading control. Results are representative of three separate experiments.
and catechol were only able to induce ARE-luciferase in
Surprisingly, Cu2+ further enhanced induction of ARE-lucif-
the presence of CuCl2. In contrast, after treatment with 20 mM erase activity by quinones (Figure 3 and Table 1). To characterize
1,4-benzoquinone (1,4-BQ) for 30 min, luciferase activity this further, the effect of metal chelators on the Cu2+-dependent
was induced 16-fold 24 hr later. 1,3-Diphenol (resorcinol), which potentiation of induction by quinones of ARE-driven gene
is unable to undergo oxidation to a quinone, was inactive both expression was examined. In contrast to the complete suppres-
in the presence and in the absence of CuCl2; 1,3-diphenol sion of the inducer activity of hydroquinones by DFX and EDTA,
was inactive as an inducer in both murine Hepalclc7 cells the metal chelators only partially blocked the induction of lucif-
(Prochaska et al., 1985) and in human HepG2 cells (Rushmore erase activity by the cotreatment with tBQ and CuCl2 (Figure 3),
et al., 1991). These data indicate that the formation of qui- to yield levels of induction similar to those observed for tBQ
nones is the essential step in the activation of Nrf2 by 1,2- and alone. This finding suggests that the potentiation by Cu2+ of
1,4-diphenols.
tBQ-induced luciferase activity could be due to the activity of
Chemistry & Biology 17, 75–85, January 29, 2010 ª2010 Elsevier Ltd All rights reserved 77

Chemistry & Biology
Copper-Dependent Activation of Nrf2 by Diphenols
We tested whether other transition metals could increase
induction by tBHQ. Exposure of AREc32 cells to 50 mmol/l of
CoCl2, NiSO4, or FeCl2 for 24 hr enhanced induction of luciferase
activity by tBHQ (10 mmol/l) from 20-fold to 35-, 37-, and 38-fold,
respectively. This implied that, in addition to Cu2+, other redox-
active metals can also mediate the oxidation of tBHQ to tBQ.
Of note, serum reduced the Cu2+-mediated potentiation of the
tBHQ-induced luciferase activity (see Figure S1A available
online), possibly as a result of chelating or sequestering the
Cu2+ by serum proteins, suggesting that such processes will
affect the intracellular copper levels.
Cu2+ Is a Very Efficient Catalyst of the Oxidation of tBHQ
In Tris-HCl buffer (pH 8.0), tBHQ has a UV absorption maximum
at 290 nm (am ꢁ3000 Mꢀ1 cmꢀ1) (Figure 4A, black line). Addition
of CuCl2 caused a 38-nm blue shift, to 252 nm, and a hyper-
chromism (Figure 4A, red line), suggesting that Cu2+ catalyzed
the oxidation of tBHQ to tBQ according to the reaction
shown in the inset. Indeed, examination of the UV spectrum of
authentic tBQ showed that it absorbs maximally at 252 nm
Figure 2. Hypoxia Inhibits the Potentiation by CuCl2 of tBHQ-
Induced Luciferase Activity
AREc32 cells were cultured in DMEM containing 5 mM tBHQ (A) or 2 mM SFN
(B) in the presence or absence of 50 mM CuCl2 at 21% (v/v) O2 or 1% (v/v) O2 for
24 hr. The luciferase activity of cells treated with 0.1% DMSO (control) and
cultured at 21% (v/v) O2 was set at 1.0. Values are mean SD. Results are
representative of three separate experiments. **p < 0.001.
NQO1, which can catalyze the reduction of a portion of the with a molar absorptivity of ꢁ20,000 Mꢀ1 cmꢀ1 such that the
tBQ to yield tBHQ, thereby increasing the effective concentra-
spectra of tBHQ upon addition of CuCl2 and the spectrum of
tion of the hydroquinone. According to this hypothesis, the tBQ were virtually identical (Figure 4A, red line and Figure 4B).
newly formed tBHQ will be susceptible to Cu2+-dependent We reasoned that, under aerobic conditions, the oxidation of
tBHQ to tBQ should also occur when Cu+ is used in place of
Cu2+, since in the presence of O2, Cu+ is easily oxidized to
oxidation and thus capable of activating Nrf2. Indeed, high
levels of NQO1 activity were detected in AREc32 cells (1004.8
135 nmol/min/mg protein), consistent with the proposal that Cu2+. Indeed, although the reaction initially proceeded at
NQO1 catalyzes the obligatory 2 electron-reduction of tBQ to a slightly slower rate, Cu+ was also a very efficient catalyst
tBHQ (Talalay and Dinkova-Kostova, 2004). Attempts were (Figure S2). Bubbling N2 reduced the rate of Cu+-dependent
made to reduce the NQO1 activity by using siRNA or the inhibi- formation of tBQ by 40% (Figure S3A), confirming that O2 is
tor dicumarol. Unfortunately, both of these strategies resulted required to accept electrons from Cu+ and generate Cu2+, which
in either cell death, or no significant inhibition of the enzyme in turn oxidizes tBHQ. As expected, bubbling N2 had no effect
activity at concentrations that were well-tolerated by the cells on the initial rate of the Cu2+-dependent formation of tBQ
through mechanisms that are not understood.
(Figure S3B).
Figure 3. The Metal Chelators DFX and EDTA Have
Distinct Effects on the Induction of Luciferase
Activities by tBHQ and tBQ
AREc32 cells were exposed to 5 mM tBHQ or 2.5 mM tBQ
in PBS with or without 50 mM CuCl2 in the presence or
absence of 100 mM DFX or 100 mM EDTA. After 30 min,
the PBS was replaced with DMEM plus serum. The cells
were further incubated for 24 hr before measuring
luciferase activity. The luciferase activity of cells treated
with 0.1% DMSO (control) was set to 1. Values are
mean SD. Results are representative of three separate
experiments. **p < 0.001.
78 Chemistry & Biology 17, 75–85, January 29, 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
Copper-Dependent Activation of Nrf2 by Diphenols
Table 1. The Effect of CuCl2 on the Induction of Luciferase Activity by Diphenols
Phenolic
compound
Con.
(µmol/l)
Structure
Phenolic
compound
Con.
(µmol/l)
Structure
ARE-Luciferase activity
(fd to control)
ARE-Luciferase activity
(fd to control)
- CuCl2
+ CuCl2
- CuCl2
+ CuCl2
OH
OH
OH
OH
O
CH3
OH
tBHQ
5
5
E2
10
5
1.1 0.1
1.0 0.1
32.6 2.5 **
1.3 0.1 1.5 0.1
HO
OH
OH
CH3
dtBHQ
2-OH-E2
1.3 0.1
24.7 4.5 *
1.5 0.1
1.5 0.1
1.2 0.1
6.3 1.3 **
HO
HO
OH
O
CH3
tBQ
Resorcinol
1,4-BQ
2.5
20
20
20
4-OH-E2
E2-3,4-Q
5
25
5
4.1 0.3 *
1.3 0.1
6.9 1.5 *
HO
O
OH
CH3
OH
5.8 0.4** 12.5 2.7 *
OH
O
O
O
CH3
4-OH-E1
Dopamine
16 3.8 ** 33 0.7 **
1.5 0.1
6.3 0.2 *
5.5 1.9 *
HO
O
OH
OH
1,4-HQ
12.5
1.4
0
39 13 *
1.0 0.02
HO
HO
NH2
OH
OH
OH
O
HO
HO
OH
Catechol
20
L-DOPA
50
1.5 0.2
11 2.6 **
1.0 0.07
2.3 0.6*
NH2
AREc32 cells were incubated for 30 min in PBS containing each test compound in the presence or absence of 50 mM CuCl2. Thereafter, the PBS was
removed and replaced with DMEM plus serum. The cells were further incubated for 24 hr before luciferase activity was measured. The concentration of
tBHQ, dtBHQ, tBQ, 2-OHE2, 4-OHE2, and 4-OH-E1 in PBS was 5 mmol/l; the concentration of E2 was 10 mmol/l; the concentration of E2-3,4-Q was
2.5 mmol/l; and the concentration of all of the other compounds was 20 mmol/l. The value of luciferase activity of cells treated with vehicle (control) was
set at 1. Values are mean SD. *p > 0.05; **p > 0.005.
The possibility that other (than O2) endogenous electron tBQ but Not tBHQ Reacts with Cys Residues
acceptors from Cu+ could be important for this reaction was in Glutathione and Keap1
next considered. Perhaps the most relevant one in vivo is We next examined the ability of tBHQ and tBQ to react with sulf-
H2O2, as it can efficiently generate Cu2+ from Cu+ according to hydryl groups. When GSH was added to the solution containing
the reaction: Cu+ + H2O2 / Cu2+ + OH + OHꢀ (Halliwell and tBHQ and CuCl2, a time-dependent 50-nm red shift, to 303 nm,
d
Gutteridge, 2007). Adding H2O2 to the mixture of tBHQ and and a hypochromism were noted, presumably reflecting the
Cu+ under N2 accelerated the reaction even further than when formation of a thioether bond between the quinone and the
the same reaction was carried out at ambient conditions (Fig- cysteine residue of glutathione (Figure 4A, purple and brown
ure S4), confirming that H2O2, in addition to O2, constitutes lines); this observation was confirmed in a reaction of tBQ with
another physiologically relevant electron acceptor contributing GSH (Figure 4B, blue line). Titration of small aliquots of substoi-
to the copper-catalyzed oxidation of hydroquinones.
chiometric quantities of GSH into a solution containing a constant
We also asked whether other divalent cations could catalyze amount of tBQ resulted in incremental increases in the absor-
the oxidation of tBHQ to tBQ and found that Mn2+ was also effec- bance at 303 nm, proportional to each aliquot that was added
tive (Figure S5). In contrast, Fe2+ was only weakly active, and (Figure 4C). Once the concentration of GSH was equal to the
Zn2+ was essentially inactive. Thus, it appears that both in vitro concentration of tBQ, there was no further change, indicating
and in cell culture, copper is the most efficient catalyst of this a reaction stoichiometry of 1:1. In contrast to the reaction of
reaction, but other transition metals are also capable of medi- tBQ with GSH, no spectral changes were observed when
ating the oxidation of tBHQ.
tBHQ was used in place of tBQ, showing that tBHQ itself
Chemistry & Biology 17, 75–85, January 29, 2010 ª2010 Elsevier Ltd All rights reserved 79

Chemistry & Biology
Copper-Dependent Activation of Nrf2 by Diphenols
Figure 4. tBQ but Not tBHQ Reacts with Gluta-
thione and Cysteine Residues in Keap1
(A) Absorption spectra in 20 mM Tris-HCl/0.005% Tween
20 (pH 8.0) at 25ꢂC of 100 mM tBHQ before (black line)
and after (red line) addition of 100 mM CuCl2, and following
incubation with 100 mM GSH for 1 min (purple line) and
5 min (brown line).
(B) Absorption spectra of 100 mM tBQ before (red line) and
after 5 min incubation with 200 mM GSH (blue line) in the
same buffer.
(C) Titration of 2.0-ml aliquots each delivering 5 nmol of
GSH into a solution of 100 mM tBQ in a final volume of
500 ml under the same reaction conditions.
(D) Absorption spectrum kinetics of the reaction of 50 mM
tBQ with 5 mM Keap1. Spectra were recorded every 30 s.
(E) Titration of 0.5-ml aliquots each delivering 2.5 nmol of
tBQ into a solution of 5 mM Keap1 in a final volume of
500 ml.
12-dioxooleana-1,9(11)-dien-28-onitrile] leads
to changes in the spectra of these molecules
that are very similar to the changes observed
when they react with sulfhydryl groups in GSH
and DTT (Dinkova-Kostova et al., 2002, 2005).
In the present study, we also observed sim-
ilar time-dependent spectral changes when
tBQ was allowed to react with recombinant
murine Keap1 (Figure 4D). Titration experiments
revealed that the stoichiometry of the reaction
of Keap1 with tBQ is 1:4, indicating that four
cysteine residues were modified by the quinone
under these reaction conditions (Figure 4E).
Taken together, our experiments demonstrate
that Cu2+ catalyzes the oxidation of tBHQ to
tBQ, and it is the resultant quinone that subse-
quently reacts with cysteine residues in Keap1.
ROS are generated during Cu2+-mediated
redox-cycling of tBHQ (Li et al., 2002) and the
closely related 2,5-di-tert-butylhydroquinone
(dtBHQ) (Nakamura et al., 2003). dtBHQ can
undergo a similar one-electron oxidation as
tBHQ, including conversion to its benzoquinone
form, and Cu2+-dependent superoxide genera-
tion, but unlike tBQ, it has been reported not
to react with GSH (Nakamura et al., 2003). As
described above, the addition of GSH to a solu-
tion containing CuCl2 and tBHQ leads to a rapid
decrease in absorbance at 252 nm, the lmax for
does not react with GSH (Figure S6). Overnight incubation of tBQ tBQ, and an increase in absorbance at 303 nm, reflecting the
with GSH resulted in the formation of two mono- and one bis- formation of quinone-glutathione conjugate (Figure 4A and
substituted glutathione conjugates, in agreement with a previous Figure S8A). In sharp contrast, the corresponding decrease in
report (van Ommen et al., 1992). The conjugates were separated absorbance was much less pronounced when dtBHQ was
by HPLC and analyzed by mass spectrometry, revealing two used in place of tBHQ (Figure S8B), indicating that, under these
identical molecular ions at m/z 470 for the mono-thioethers, reaction conditions, formation of a glutathione conjugate with
and a molecular ion at m/z 775 for the bis-thioether (Figure S7). dtBHQ is inefficient, most likely due to steric hindrance of the
ThereactionofKeap1withinducersofARE-drivengeneexpres- second tert-butyl group. These findings are consistent with the
sion [e.g., SFN, the Michael acceptors bis(2-hydroxybenzylidene) reported chemoprotective activity of tBHQ against benzo[a]py-
acetone, and the oleanane dicyanotriterpenoid 2-cyano-3, rene-induced neoplasia of the forestomach of ICR/Ha mice,
80 Chemistry & Biology 17, 75–85, January 29, 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
Copper-Dependent Activation of Nrf2 by Diphenols
Figure 5. Cellular GSH Levels Influence the Cu2+-Potentiated Activation of Nrf2 by Oxidizable Diphenols
(A) GSH levels after 24 hr pretreatment with 20 mM BSO or 2 mM SFN. Cells were incubated with 20 mM BSO or 2 mM SFN in DMEM plus serum for 24 hr. After this
pretreatment period, cells were exposed to PBS containing 5 mM tBHQ and 50 mM CuCl2 for 30 min. The PBS was then removed and the cells were further incu-
bated in DMEM plus serum for 24 hr before luciferase activity was measured. The GSH level of the cells pretreated with 0.1% DMSO (control) was set to 100%.
(B) Cellular glutathione levels affect the potentiation by CuCl2 of tBHQ-, 4OHE2-, and dopamine induced luciferase activity. AREc32 cells were incubated with
20 mM BSO or 2 mM SFN in DMEM plus serum for 24 hr. After this pretreatment period, cells were exposed to PBS containing 5 mM tBHQ, 5 mM 4OHE2, or 50 mM
dopamine in the presence of 50 mM CuCl2 for 30 min. The PBS was then removed and the cells were further incubated in DMEM plus serum for 24 hr before
luciferase activity was measured. The value from the cells pretreated and treated with 0.1% DMSO (control) was set to 1.0. Values are mean SD. Results
are representative of three separate experiments. *p < 0.05; **p < 0.005.
and the lack of protection by dtBHQ in the same model (Watten- other hand, a 200% increase in the level of intracellular GSH
(following 24 h-exposure to 2 mM SFN) decreased the induction
berg et al., 1980).
We then evaluated the relative contributions of ROS and of luciferase activity by 5 mM tBHQ in the presence of 50 mM
the electrophilic quinone reacting with intracellular thiols in the CuCl2, from 26-fold to 9-fold (Figure 5B). Although these data
induction of ARE-dependent gene expression by tBHQ in the could also be explained by competition of sulforaphane and
presence of CuCl2. Under our culture conditions, where the cells tBQ for binding to Keap1, we think that this interpretation is
were treated with 5 mmol/l or 10 mmol/l of either tBHQ or dtBHQ in unlikely, on the basis of the known instability of Keap1-sulfora-
the presence of CuCl2, ROS were not detectable using the fluo- phane complexes and the rapid metabolism and excretion of
rescent probe 20,7-dichlorodihydrofluorescein diacetate, though sulforaphane (Ye et al., 2002; Zhang and Callaway, 2002), the
they were detected when the cells were treated with 10-fold administration of which preceded exposure to tBHQ and CuCl2
higher concentrations of tBHQ or dtBHQ (data not shown). by 24 hr. These data suggest that the intracellular thiol concen-
Furthermore, only tBHQ was able to induce luciferase activity tration influences the ability of tBHQ to activate the Keap1/
in the presence of CuCl2, whereas dtBHQ was essentially inac- Nrf2/ARE pathway, providing further support for the hypothesis
tive as an inducing agent in the absence of copper and only that the thiol-reactive quinone metabolite represents the actual
very weakly active in the presence of copper (Table 1). Taken proximal inducing agent.
together, our results strongly suggest that the major species
that stimulates Nrf2-mediated gene induction is the electrophilic Induction of ARE-Reporter Luciferase Activity
quinone, and that ROS that are formed during the redox cycling by Catechol Estrogens and Catecholamines Requires
play a much smaller role in the induction process.
Oxidation to Their Quinone Derivatives
Cu2+ mediates the oxidation of catechol estrogens (Li et al.,
1994; Thibodeau and Paquette, 1999). Lee et al. (2003) reported
that 2-OHE2 and 4-OHE2 induced ARE-driven gene expression.
Cellular GSH Levels Influence
the Cu2+-Potentiated Activation of Nrf2
As demonstrated above, the quinone forms of oxidizable diphe- On the basis of our findings, we predicted that Cu2+ will poten-
nols react stoichiometrically with GSH in vitro. To examine the tiate the activation of Nrf2 by catechol estrogens. To test this
significance of this reaction in cells, we evaluated the effect hypothesis, AREc32 cells were treated with 2-OHE2, 4-OH-E1,
that manipulating the intracellular GSH level had on the ability and 4-OHE2 in PBS for 30 min. As expected, 2-OHE2,
of tBHQ to induce ARE-driven genes. Thus, we first pretreated 4-OHE2, and 4-OH-E1 were unable to induce ARE-luciferase
the cells with BSO to deplete GSH. As shown in Figure 5A, expo- activity in the absence of CuCl2 (Table 1). However, in the
sure to 20 mM BSO for 24 hr reduced the intracellular GSH level presence of 50 mM CuCl2, each of these catechol estrogens
by 50%. This decrease in GSH markedly increased the induction induced the luciferase activity markedly, and at a concentration
of luciferase activity by treatment with 5 mM tBHQ for 30 min in of 5 mmol/l, they induced luciferase activity >6-fold. To determine
the presence of 50 mM CuCl2 from 26-fold to 70-fold. On the whether the quinone formed from catechol estrogens was the
Chemistry & Biology 17, 75–85, January 29, 2010 ª2010 Elsevier Ltd All rights reserved 81

Chemistry & Biology
Copper-Dependent Activation of Nrf2 by Diphenols
Nrf2/ARE signaling pathway. Under aerobic conditions, Cu2+
catalyzes the oxidation of ortho- and para-hydroquinones to
their corresponding quinones, which, in turn, react with cysteine
residues in Keap1, resulting in derepression of Nrf2. This study
also highlights the importance of transition metals and oxygen
in the activation of the Keap1/Nrf2/ARE pathway by oxidizable
phenolic compounds that are ubiquitous in our environment.
We used tBHQ as a model compound to demonstrate that its
oxidation to tBQ is the rate-limiting step that enables the hydro-
quinone to activate the Keap1/Nrf2/ARE pathway. Furthermore,
we have shown that induction of ARE-driven gene expression by
a series of compounds that contain para- or ortho-hydroquinone
moieties is similarly potentiated by Cu2+. Oxidizable diphenols
are a ubiquitous class of compounds, to which we are exposed
continuously. Many are metabolic products of dietary plant con-
stituents or of environmental pollutants such as polycyclic
aromatic hydrocarbons (Jin and Penning, 2007; Palackal et al.,
2002; Park et al., 2008). Some are produced during metabolism
of endogenous precursors such as catechol estrogens and
dopamine. The ability of Cu2+ and other transition metals to
potentiate the activation of the Keap1/Nrf2/ARE pathway implies
that any alteration in the homeostasis of copper or other transi-
tion metals will have a significant impact on the biological effects
of hydroquinones and quinones in vivo.
Figure 6. CuCl2 Is Required for the Induction of AKR1C by 4OHE2
Cells were exposed to PBS containing DMSO, 5 mM 4OHE2, 50 mM CuCl2, or
5 mM 4OHE2 and 50 mM CuCl2 in the presence or absence of 100 mM DFX for
30 min. The PBS was then removed and the cells were further incubated in
DMEM plus serum for 24 hr. Whole cell extracts (30 mg of protein) were sepa-
rated by 15% SDS-PAGE and the levels of AKR1C analyzed by western blot-
ting. Immunoblotting for actin was used as the loading control. Results are
representative of two separate experiments.
active species, we obtained E2-3,4-Q by oxidation of 4-OHE2.
E2-3,4-Q induced ARE-luciferase in the absence of CuCl2 by
ꢁ6-fold (Table 1). Immunoblotting revealed that 4-OHE2 induced
endogenous AKR1C expression only in the presence of CuCl2.
However, induction of AKR1C by this steroid was completely
abolished by DFX (Figure 6). Our results demonstrate that, like
the synthetic diphenols, endogenous catechol estrogens are
not inducers, but oxidation to their quinone derivatives is the
requisite step in the activation of Nrf2.
We have demonstrated that the capacity of hydroquinones to
activate the Keap1/Nrf2/ARE pathway is closely correlated with
the efficiency of their conversion to eletrophilic metabolites and
their potencies as inducers are dependent on the concentration
of Cu2+ and oxygen. As the copper content in some tissues
exceeds 100 mmol/l (Linder and Hazegh-Azam, 1996), and the
doses used in the present study are within physiological levels,
it seems likely that a significant fraction of the hydroquinones
will be converted to their respective quinones, especially in
tissues, which are normally exposed to high levels of oxygen.
Recent studies have shown that para-hydroquinones, such as
tBHQ and the ortho-hydroquinone carnosic acid, protect
neuronal cells against oxidative stress in a manner associated
with activation of ARE-driven gene expression (Satoh et al.,
2006, 2008, 2009). Our findings indicate that the oxidation of
hydroquinones is a prerequisite for their protective effects.
It should be noted, however, that the overproduction of
quinones has detrimental effects, including acute cytotoxicity,
immunotoxicity, and carcinogenicity, and our findings imply
that high levels of oxygen and transition metals are important
determinants. Thus, aldo-keto reductase catalyzed formation
of the electrophilic and redox-active ortho-quinones 7,12-dime-
thylbenz[a]anthracene-3,4-dione and benzo[a]pyrene-7,8-dione
has been shown to occur in human lung adenocarcinoma cells
exposed to the polycyclic aromatic hydrocarbons 7,12-dime-
thylbenz[a]anthracene and benzo[a]pyrene, causing reactive
oxygen species generation, oxidative DNA damage, and DNA
strand breaks (Jin and Penning, 2007; Palackal et al., 2002;
Park et al., 2008). Similarly, the ortho-quinone products of dopa-
mine have been implicated in the cytotoxicity of this neurotrans-
mitter in dopaminergic neuronal cells (Asanuma et al., 2004;
Miyazaki and Asanuma, 2009). Patients with Wilson’s disease
(carrying a genetic mutation in the copper transporter that results
High concentrations of dopamine activate Nrf2 in astrocytes
and meningeal cells (Shih et al., 2007) and in PC12 dopaminergic
neuronal cells (Jia et al., 2008). Because dopamine and its
precursor L-DOPA, possess two hydroxyl groups on the benzene
ring and can undergo oxidation to become catechol quinones,
we postulated that Cu2+ might increase the activation of Nrf2
by dopamine and L-DOPA. We exposed AREc32 cells to dopa-
mine or L-DOPA for 30 min and found that neither of the com-
pounds had any effect on the luciferase activity in the absence
of Cu2+. However, in the presence of 50 mmol/l CuCl2, 30 min
treatment with 12.5 mmol/l dopamine resulted in 6-fold increase
in the luciferase activity (Table 1), and 30 min treatment with
50 mmol/l L-DOPA produced a 2-fold increase in the luciferase
activity (Table 1). In addition, pretreatment of AREc32 cells
with BSO that decreased the intracellular glutathione by 50%
nearly doubled the induction of luciferase activity by dopamine
and CuCl2 (Figure 5B); in contrast, an ꢁ2-fold increase in
the cellular glutathione content, achieved by pretreatment with
SFN, diminished the induction by nearly 50% (Figure 5B). Our
data suggest that oxidation of dopamine and L-DOPA is the
key step in determining their ability to activate Nrf2-regulated
neuroprotective pathways.
DISCUSSION
This study establishes the chemical and molecular mechanisms in copper accumulation) have high frequency of Parkinsonism
whereby hydroquinone-type compounds activate the Keap1/ (Taly et al., 2007). In addition, the fact that iron and copper levels
82 Chemistry & Biology 17, 75–85, January 29, 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
Copper-Dependent Activation of Nrf2 by Diphenols
are relatively high in the alveolar epithelial cells in the breast knowledge, this is the first experimental evidence that
(Vorherr, 1974) may partly explain why 4OHE2 plays a role in oxidizable hydroquinones are not Nrf2 activators them-
the etiology of breast cancer (Cavalieri et al., 2000). Copper selves, but their corresponding quinones are the ultimate
metabolism is altered in tumor-bearing mice and rats and also inducers. Our data also imply that hydroquinones will not
in humans (Apelgot et al., 1986; Semczuk and Pomykalski, stimulate Nrf2-mediated gene induction in the absence of
1973; Tani and Kokkola, 1972). Additionally, serum and tumor transition metals and oxygen. These findings are of partic-
copper levels are significantly elevated in patients with cancer, ular importance for certain human conditions in which the
compared with healthy subjects (Kuo et al., 2002), and positively levels of transition metals are increased, such as Wilson’s
correlate with tumor incidence, burden, malignant progression, disease (genetic mutation in the copper transporter result-
and recurrence in many human and animal cancers. On the ing in copper accumulation) or patients with cancer (who
basis of our findings, we hypothesize that elevated endogenous have increased plasma and tumor copper levels), and espe-
copper levels will result in enhanced activation of the Keap1/ cially in tissues that are exposed to relatively high concen-
Nrf2/ARE pathway by oxidizable diphenols from both exogenous trations of oxygen, such as the lung, the brain, and the alve-
and endogenous origin in an effort to protect against potential olar epithelium of the breast.
deleterious effects of quinones. A prominent example is the
up-regulation of NQO1, a prototypical Nrf2-regulated enzyme
that catalyzes the obligatory 2-electron reduction of ortho- and
para-quinones, thereby diverting them from participating in
redox cycling and sulfhydryl depletion reactions. Notably,
although under our experimental conditions copper is most effi-
cient in the hydroquinone to quinone conversion, we do not think
that our findings are restricted to diseases associated with
increased copper concentration, but will also be valid for other
pathologic abnormalities linked to excessive exposures to
metals from contaminated water or food.
EXPERIMENTAL PROCEDURES
Chemicals and Reagents
All chemicals were of analytical grade and were from Sigma-Aldrich (Dorset,
UK.). E2-3,4-Q was prepared as described elsewhere (Zahid et al., 2006).
The media supplements for cell culture were from Life Technologies. The anti-
sera against AKR1C and AKR1B10 have been described elsewhere (MacLeod
et al., 2009; O’Connor et al., 1999). The antibody against HO-1 was from
BIOMOL International.
Cell Culture and Evaluation of ARE-Driven Luciferase Activity
The stable human mammary ARE-reporter cell line, AREc32 (Wang et al.,
2006), was maintained in growth medium (Dulbecco’s MEM with glutamax
[DMEM] supplemented with 10% fetal bovine serum [FBS] and penicillin-
streptomycin) containing 0.8 mg/ml G418, at 37ꢂC, in 95% air and 5% CO2.
4-OHE1, 4-OHE2, and tBQ were dissolved in acetonitrile. L-buthionine-
S,R-sulfoximine (BSO), CuCl2, and CuCl were prepared in H2O. Dopamine
and 3,4-dihydroxy-L-phenylalanine (L-DOPA) were freshly prepared in ethanol
and 0.5 N HCl, respectively. All other chemicals were dissolved in DMSO. For
xenobiotic treatments, AREc32 cells were seeded in 96-well plates at a density
of 1.2 3 104 cells/well and were allowed 24 hr to recover. Typically, cells were
treated with the xenobiotic of interest for further 24 hr in DMEM. However, on
occasions, treatments were for 30 min in PBS. For treatments in medium, cells
were left for 24 hr in the presence of the xenobiotic before being harvested, and
the firefly luciferase activity in cell lysates was measured using a luminometer
(Turner Designs Model TD-20/20, Promega) following addition of Luciferase
Assay Reagent (Promega). For short-term treatment in PBS, the culture
medium was replaced with PBS containing xenobiotics (in 0.1% vehicle by
vol.). After 30 min, PBS was replaced with growth medium, and the cells
were further incubated for 24 hr. In experiments involving metal chelators,
each chelator was mixed with CuCl2 prior to being mixed with tBHQ or tBQ.
To culture cells under low O2 conditions, a hypoxia workstation (Biotrace
Fred Baker) gassed with 1% (v/v) O2, and 5% (v/v) CO2 balanced with
ꢁ94% N2 was used. Cells were seeded at 1.2 3 104 cells per well in growth
medium in 96-well plates and first incubated for 24 hr at 95% air and 5%
CO2. Cells and culture medium were equilibrated in the hypoxic workstation
for 2 hr prior to induction for further 24 hr.
In summary, the ability of hydroquinones to undergo redox
cycling and to form quinones provides the basis for their biolog-
ical activity—either to activate the Nrf2 cytoprotective pathway
or, at high concentrations, to create oxidative stress and damage
macromolecules, ultimately causing toxicity and/or carcinoge-
nicity. In this study, we demonstrate an alternative, nonenzy-
matic, oxygen- and copper-dependent pathway for the activa-
tion of the Keap1/Nrf2/ARE pathway by hydroquinones. This
knowledge is essential for designing novel chemopreventive
strategies and for developing new approaches against drug
resistance.
SIGNIFICANCE
Para- and ortho-hydroquinones represent a ubiquitous
class of oxidizable diphenols. Among them are metabolites
of exogenous dietary constituents and environmental pollut-
ants, as well as endogenous catechol estrogens and dopa-
mine. Paradoxically, both cytoprotective and cytotoxic
properties have been attributed to such molecules. At low
concentrations, they activate transcription factor Nrf2 that
controls the expression of a battery of >100 cytoprotective
proteins, whereas at high concentrations, they lead to acute
cytotoxicity and carcinogenicity. In this study, we demon-
strate that a large series of oxidizable diphenols are able
to activate Nrf2-dependent gene expression only in the pres-
ence, but not in the absence, of transition metals and
oxygen. Under aerobic conditions, Cu2+ catalyzes the oxida-
tion of ortho- and para-hydroquinones to their correspond-
ing quinones, which subsequently react with cysteine resi-
dues of the protein sensor for inducers Keap1, resulting in
activation of Nrf2. Furthermore, oxidation to quinones is
both necessary and sufficient for activation of the Keap1/
Nrf2 pathway by ortho- and para-hydroquinones. To our
Purification of Keap1 and Binding of Quinones
Recombinant Keap1 was expressed and purified as described elsewhere
(Dinkova-Kostova et al., 2002). Binding studies were performed at 25ꢂC by
adding a 5-ml aliquot of a solution containing the relevant compound (dissolved
in acetonitrile) to a 500 ml solution of 5 mM Keap1 in 20 mM Tris$HCl/0.005%
(v/v) Tween 20 (pH 8.0) in 0.5-cm pathlength quartz cuvettes. Immediately after
mixing, the UV-VIS spectra were recorded by using a double-beam spectro-
photometer. To determine the reaction stoichiometry, tBQ was titrated in
0.5-ml aliquots delivering 2.5 nmol of tBQ per aliqout into 5 mM Keap1 in the
same buffer in a final 500-ml reaction volume. The increase in absorbance at
303 nm was monitored after addition of each aliquot. Similar experiments
were performed with reduced glutathione in place of Keap1.
Chemistry & Biology 17, 75–85, January 29, 2010 ª2010 Elsevier Ltd All rights reserved 83

Chemistry & Biology
Copper-Dependent Activation of Nrf2 by Diphenols
Determination of GSH
Dinkova-Kostova, A.T., Liby, K.T., Stephenson, K.K., Holtzclaw, W.D., Gao, X.,
Suh, N., Williams, C., Risingsong, R., Honda, T., Gribble, G.W., et al. (2005).
Extremely potent triterpenoid inducers of the phase 2 response: correlations
of protection against oxidant and inflammatory stress. Proc. Natl. Acad. Sci.
USA 102, 4584–4589.
Reduced glutathione was measured as described elsewhere (Kamencic et al.,
2000). Cells were seeded at 0.5 3 105 cells per well in 24-well plates. After
overnight incubation, cells were treated with 20 mM BSO or 2 mM SFN for
24 hr and then were lysed. Cell lysates (100 ml) were incubated with 100 ml
PBS containing 80 mM monochlorobimane (mCB) and 1 U/ml glutathione
S-transferase for 1 hr at 25ꢂC. Formation of the GS-mCB adduct was quanti-
fied by its fluorescence with excitation at 390 nm and emission at 490 nm.
Eggler, A.L., Liu, G., Pezzuto, J.M., van Breemen, R.B., and Mesecar, A.D.
(2005). Modifying specific cysteines of the electrophile-sensing human
Keap1 protein is insufficient to disrupt binding to the Nrf2 domain Neh2.
Proc. Natl. Acad. Sci. USA 102, 10070–10075.
Western Blot Analysis
Halliwell, B., and Gutteridge, J.M.C. (2007). Free radicals in biology and medi-
cine, 4th ed. (New York: Oxford University Press).
Whole-cell extracts were prepared as described elsewhere (Wang et al., 2006).
Briefly, cells were lysed in 0.1 M HEPES (pH 7.4), containing 0.5 M KCl, 5 mM
MgCl2, 0.5 mM EDTA, and 20% glycerol that was supplemented with
protease inhibitors (Roche Diagnostics). Proteins (30 mg) were separated
by SDS-PAGE. Immunoblotting was performed as described elsewhere
(O’Connor et al., 1999).
Hayes, J.D., and McMahon, M. (2009). NRF2 and KEAP1 mutations: perma-
nent activation of an adaptive response in cancer. Trends Biochem. Sci. 34,
176–188.
Higgins, L.G., Kelleher, M.O., Eggleston, I.M., Itoh, K., Yamamoto, M., and
Hayes, J.D. (2009). Transcription factor Nrf2 mediates an adaptive response
to sulforaphane that protects fibroblasts in vitro against the cytotoxic effects
of electrophiles, peroxides and redox-cycling agents. Toxicol. Appl. Pharma-
col. 237, 267–280.
Statistical Analysis
Statistical comparisons were performed by the unpaired Student’s t test.
Jia, Z., Zhu, H., Misra, B.R., Li, Y., and Misra, H.P. (2008). Dopamine as a potent
inducer of cellular glutathione and NAD(P)H:quinone oxidoreductase 1 in PC12
neuronal cells: a potential adaptive mechanism for dopaminergic neuroprotec-
tion. Neurochem. Res. 33, 2197–2205.
SUPPLEMENTAL INFORMATION
Supplemental Information includes eight figures and can be found with this
article online at doi:10.1016/j.chembiol.2009.12.013.
Jin, Y., and Penning, T.M. (2007). Aldo-keto reductases and bioactivation/
detoxication. Annu. Rev. Pharmacol. Toxicol. 47, 263–292.
ACKNOWLEDGMENTS
Kamencic, H., Lyon, A., Paterson, P.G., and Juurlink, B.H. (2000). Monochloro-
bimane fluorometric method to measure tissue glutathione. Anal. Biochem.
286, 35–37.
We thank Nobunao Wakabayashi (Johns Hopkins University) for providing the
Keap1 expression plasmid, Stewart Finlayson for expression and purification
of recombinant Keap1, and Ian M. Eggleston and Sebastien Ronseaus for
assistance with mass spectrometry. We thank Rene´ V. Bensasson (Labora-
toire de Chimie des Substances Naturelles, MNHN, Paris), Mike McMahon,
and Colin J. Henderson for helpful discussions. This work was supported by
Cancer Research UK (C4639/A5661 and C20953/A10270), Research Councils
UK, the Royal Society, and the American Cancer Society (RSG-07-157-01-
CNE).
Kensler, T.W., Wakabayashi, N., and Biswal, S. (2007). Cell survival responses
to environmental stresses via the Keap1-Nrf2-ARE pathway. Annu. Rev. Phar-
macol. Toxicol. 47, 89–116.
Kuo, H.W., Chen, S.F., Wu, C.C., Chen, D.R., and Lee, J.H. (2002). Serum and
tissue trace elements in patients with breast cancer in Taiwan. Biol. Trace
Elem. Res. 89, 1–11.
Lee, J.M., Anderson, P.C., Padgitt, J.K., Hanson, J.M., Waters, C.M., and
Johnson, J.A. (2003). Nrf2, not the estrogen receptor, mediates catechol
estrogen-induced activation of the antioxidant responsive element. Biochim.
Biophys. Acta 1629, 92–101.
Received: September 15, 2009
Revised: December 12, 2009
Accepted: December 15, 2009
Published: January 28, 2010
Li, Y., and Trush, M.A. (1993). Oxidation of hydroquinone by copper: chemical
mechanism and biological effects. Arch. Biochem. Biophys. 300, 346–355.
Li, Y., Trush, M.A., and Yager, J.D. (1994). DNA damage caused by reactive
oxygen species originating from
a copper-dependent oxidation of the
REFERENCES
2-hydroxy catechol of estradiol. Carcinogenesis 15, 1421–1427.
Li, Y., Seacat, A., Kuppusamy, P., Zweier, J.L., Yager, J.D., and Trush, M.A.
(2002). Copper redox-dependent activation of 2-tert-butyl(1,4)hydroquinone:
formation of reactive oxygen species and induction of oxidative DNA damage
in isolated DNA and cultured rat hepatocytes. Mutat. Res. 518, 123–133.
Apelgot, S., Coppey, J., Fromentin, A., Guille, E., Poupon, M.F., and Roussel,
A. (1986). Altered distribution of copper (64Cu) in tumor-bearing mice and rats.
Anticancer Res. 6, 159–164.
Asanuma, M., Miyazaki, I., Diaz-Corrales, F.J., and Ogawa, N. (2004). Quinone
formation as dopaminergic neuron-specific oxidative stress in the pathogen-
esis of sporadic Parkinson’s disease and neurotoxin-induced parkinsonism.
Acta Med. Okayama 58, 221–233.
Linder, M.C., and Hazegh-Azam, M. (1996). Copper biochemistry and molec-
ular biology. Am. J. Clin. Nutr. 63, 797S–811S.
Lou, H., Du, S., Ji, Q., and Stolz, A. (2006). Induction of AKR1C2 by phase II
inducers: identification of a distal consensus antioxidant response element
regulated by NRF2. Mol. Pharmacol. 69, 1662–1672.
Bensasson, R.V., Zoete, V., Dinkova-Kostova, A.T., and Talalay, P. (2008).
Two-step mechanism of induction of the gene expression of a prototypic
cancer-protective enzyme by diphenols. Chem. Res. Toxicol. 21, 805–812.
MacLeod, A.K., McMahon, M., Plummer, S.M., Higgins, L.G., Penning, T.M.,
Igarashi, K., and Hayes, J.D. (2009). Characterization of the cancer chemopre-
ventive NRF2-dependent gene battery in human keratinocytes: demonstration
that the KEAP1-NRF2 pathway, and not the BACH1-NRF2 pathway, controls
cytoprotection against electrophiles as well as redox-cycling compounds.
Carcinogenesis 30, 1571–1580.
Cavalieri, E., Frenkel, K., Liehr, J.G., Rogan, E., and Roy, D. (2000). Estrogens
as endogenous genotoxic agents—DNA adducts and mutations. J. Natl.
Cancer Inst. Monogr. 27, 75–93.
Devling, T.W., Lindsay, C.D., McLellan, L.I., McMahon, M., and Hayes, J.D.
(2005). Utility of siRNA against Keap1 as a strategy to stimulate a cancer che-
mopreventive phenotype. Proc. Natl. Acad. Sci. USA 102, 7280–7285A.
Miyazaki, I., and Asanuma, M. (2009). Approaches to prevent dopamine
quinone-induced neurotoxicity. Neurochem. Res. 34, 698–706.
Dinkova-Kostova, A.T., Holtzclaw, W.D., Cole, R.N., Itoh, K., Wakabayashi, N.,
Katoh, Y., Yamamoto, M., and Talalay, P. (2002). Direct evidence that sulfhy-
dryl groups of Keap1 are the sensors regulating induction of phase 2 enzymes
that protect against carcinogens and oxidants. Proc. Natl. Acad. Sci. USA 99,
11908–11913.
Motohashi, H., and Yamamoto, M. (2004). Nrf2-Keap1 defines a physiologi-
cally important stress response mechanism. Trends Mol. Med. 10, 549–557.
Nakamura, Y., Kumagai, T., Yoshida, C., Naito, Y., Miyamoto, M., Ohigashi, H.,
Osawa, T., and Uchida, K. (2003). Pivotal role of electrophilicity in glutathione
84 Chemistry & Biology 17, 75–85, January 29, 2010 ª2010 Elsevier Ltd All rights reserved

Chemistry & Biology
Copper-Dependent Activation of Nrf2 by Diphenols
S-transferase induction by tert-butylhydroquinone. Biochemistry 42, 4300–
Semczuk, B., and Pomykalski, M. (1973). Serum copper levels in patients with
4309.
largyngeal carcinoma [in Polish]. Otolaryngol. Pol. 27, 17–23.
Nguyen, T., Nioi, P., and Pickett, C.B. (2009). The Nrf2-antioxidant response
element signaling pathway and its activation by oxidative stress. J. Biol.
Chem. 284, 13291–13295.
Shih, A.Y., Erb, H., and Murphy, T.H. (2007). Dopamine activates Nrf2-regu-
lated neuroprotective pathways in astrocytes and meningeal cells. J. Neuro-
chem. 101, 109–119.
Nioi, P., McMahon, M., Itoh, K., Yamamoto, M., and Hayes, J.D. (2003). Iden-
tification of a novel Nrf2-regulated antioxidant response element (ARE) in the
mouse NAD(P)H:quinone oxidoreductase 1 gene: reassessment of the ARE
consensus sequence. Biochem. J. 374, 337–348.
Talalay, P., and Dinkova-Kostova, A.T. (2004). Role of nicotinamide quinone
oxidoreductase 1 (NQO1) in protection against toxicity of electrophiles and
reactive oxygen intermediates. Methods Enzymol. 382, 355–364.
Taly, A.B., Meenakshi-Sundaram, S., Sinha, S., Swamy, H.S., and Arunodaya,
G.R. (2007). Wilson disease: description of 282 patients evaluated over 3
decades. Medicine (Baltimore) 86, 112–121.
O’Connor, T., Ireland, L.S., Harrison, D.J., and Hayes, J.D. (1999). Major differ-
ences exist in the function and tissue-specific expression of human aflatoxin
B1 aldehyde reductase and the principal human aldo-keto reductase AKR1
family members. Biochem. J. 343, 487–504.
Tani, P., and Kokkola, K. (1972). Serum iron, copper, and iron-binding capacity
in bronchogenic pulmonary carcinoma. Scand. J. Respir. Dis. Suppl. 80,
121–128.
Palackal, N.T., Lee, S.H., Harvey, R.G., Blair, I.A., and Penning, T.M. (2002).
Activation of polycyclic aromatic hydrocarbon trans-dihydrodiol proximate
carcinogens by human aldo-keto reductase (AKR1C) enzymes and their func-
tional overexpression in human lung carcinoma (A549) cells. J. Biol. Chem.
277, 24799–24808.
Thibodeau, P.A., and Paquette, B. (1999). DNA damage induced by catecho-
lestrogens in the presence of copper (II): generation of reactive oxygen species
and enhancement by NADH. Free Radic. Biol. Med. 27, 1367–1377.
van Ommen, B., Koster, A., Verhagen, H., and van Bladeren, P.J. (1992). The
glutathione conjugates of tert-butyl hydroquinone as potent redox cycling
agents and possible reactive agents underlying the toxicity of butylated
hydroxyanisole. Biochem. Biophys. Res. Commun. 189, 309–314.
Park, J.H., Mangal, D., Tacka, K.A., Quinn, A.M., Harvey, R.G., Blair, I.A., and
Penning, T.M. (2008). Evidence for the aldo-keto reductase pathway of polycy-
clic aromatic trans-dihydrodiol activation in human lung A549 cells. Proc. Natl.
Acad. Sci. USA 105, 6846–6851.
Prochaska, H.J., De Long, M.J., and Talalay, P. (1985). On the mechanisms of
induction of cancer-protective enzymes: a unifying proposal. Proc. Natl. Acad.
Sci. USA 82, 8232–8236.
Vorherr, H. (1974). The Breast. Morphology, Physiology, and Lactation (New
York: Academic).
Wang, X.J., Hayes, J.D., and Wolf, C.R. (2006). Generation of a stable antiox-
idant response element-driven reporter gene cell line and its use to show
redox-dependent activation of nrf2 by cancer chemotherapeutic agents.
Cancer Res. 66, 10983–10994.
Reichard, J.F., Motz, G.T., and Puga, A. (2007). Heme oxygenase-1 induction
by NRF2 requires inactivation of the transcriptional repressor BACH1. Nucleic
Acids Res. 35, 7074–7086.
Rushmore, T.H., Morton, M.R., and Pickett, C.B. (1991). The antioxidant
responsive element. Activation by oxidative stress and identification of the
DNA consensus sequence required for functional activity. J. Biol. Chem.
266, 11632–11639.
Wattenberg, L.W., Coccia, J.B., and Lam, L.K. (1980). Inhibitory effects of
phenolic compounds on benzo(a)pyrene-induced neoplasia. Cancer Res.
40, 2820–2823.
Ye, L., Dinkova-Kostova, A.T., Wade, K.L., Zhang, Y., Shapiro, T.A., and Tala-
lay, P. (2002). Quantitative determination of dithiocarbamates in human
plasma, serum, erythrocytes and urine: pharmacokinetics of broccoli sprout
isothiocyanates in humans. Clin. Chim. Acta 316, 43–53.
Satoh, T., Okamoto, S.I., Cui, J., Watanabe, Y., Furuta, K., Suzuki, M.,
Tohyama, K., and Lipton, S.A. (2006). Activation of the Keap1/Nrf2 pathway
for neuroprotection by electrophilic phase II inducers. Proc. Natl. Acad. Sci.
USA 103, 768–773.
Zahid, M., Kohli, E., Saeed, M., Rogan, E., and Cavalieri, E. (2006). The greater
reactivity of estradiol-3,4-quinone vs estradiol-2,3-quinone with DNA in the
formation of depurinating adducts: implications for tumor-initiating activity.
Chem. Res. Toxicol. 19, 164–172.
Satoh, T., Kosaka, K., Itoh, K., Kobayashi, A., Yamamoto, M., Shimojo, Y.,
Kitajima, C., Cui, J., Kamins, J., Okamoto, S., et al. (2008). Carnosic acid,
a catechol-type electrophilic compound, protects neurons both in vitro and
in vivo through activation of the Keap1/Nrf2 pathway via S-alkylation of tar-
geted cysteines on Keap1. J. Neurochem. 104, 1116–1131.
Zhang, D.D. (2006). Mechanistic studies of the Nrf2-Keap1 signaling pathway.
Drug Metab. Rev. 38, 769–789.
Satoh, T., Saitoh, S., Hosaka, M., and Kosaka, K. (2009). Simple ortho- and
para-hydroquinones as compounds neuroprotective against oxidative stress
in a manner associated with specific transcriptional activation. Biochem. Bio-
phys. Res. Commun. 379, 537–541.
Zhang, Y., and Callaway, E.C. (2002). High cellular accumulation of sulphora-
phane, a dietary anticarcinogen, is followed by rapid transporter-mediated
export as a glutathione conjugate. Biochem. J. 364, 301–307.
Chemistry & Biology 17, 75–85, January 29, 2010 ª2010 Elsevier Ltd All rights reserved 85