Design, Synthesis, and Evaluation of Triazole Derivatives That Induce Nrf2 Dependent Gene Products and Inhibit the Keap1-Nrf2 Protein-Protein Interaction

Article abstract of DOI:10.1021/acs.jmedchem.5b00602

The transcription factor Nrf2 regulates the expression of a large network of cytoprotective and metabolic enzymes and proteins. Compounds that directly and reversibly inhibit the interaction between Nrf2 and its main negative regulator Keap1 are potential pharmacological agents for a range of disease types including neurodegenerative conditions and cancer. We describe the development of a series of 1,4-diphenyl-1,2,3-triazole compounds that inhibit the Nrf2-Keap1 protein-protein interaction (PPI) in vitro and in live cells and up-regulate the expression of Nrf2-dependent gene products.

Full text of DOI:10.1021/acs.jmedchem.5b00602

Subscriber access provided by UNIV OF CAMBRIDGE
Article
The design, synthesis and evaluation of triazole
derivatives that induce Nrf2 dependent gene products
and inhibit the Keap1-Nrf2 protein-protein interaction
Hélène Bertrand, Marjolein Schaap, Liam Baird, Nikolaos D. Georgakopoulos, Adrian
Fowkes, Clarisse Thiollier, Hiroko Kachi, Albena T. Dinkova-Kostova, and Geoffrey Wells
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b00602 • Publication Date (Web): 08 Sep 2015
Downloaded from http://pubs.acs.org on September 12, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The design, synthesis and evaluation of triazole
derivatives that induce Nrf2 dependent gene
products and inhibit the Keap1ꢀNrf2 proteinꢀprotein
interaction
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Hélène C. Bertrand1,†, Marjolein Schaap1, Liam Baird2,†† , Nikolaos D. Georgakopoulos, Adrian
Fowkes1, Clarisse Thiollier1, Hiroko Kachi1, Albena T. Dinkova-Kostova2,3, Geoff Wells1,*
1UCL School of Pharmacy, University College London, 29/39 Brunswick Square, London,
WC1N 1AX, United Kingdom
2Jacqui Wood Cancer Centre, Division of Cancer Research, Medical Research Institute,
University of Dundee, Dundee, Scotland, United Kingdom
3Departments of Medicine and Pharmacology and Molecular Sciences, Johns Hopkins University
School of Medicine, Baltimore, MD, USA
KEYWORDS: Keap1, NQO1, Nrf2, proteinꢀprotein interactions
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 36
1
2
3
4
5
6
7
8
9
ABSTRACT: The transcription factor Nrf2 regulates the expression of a large network of
cytoprotective and metabolic enzymes and proteins. Compounds that directly and reversibly
inhibit the interaction between Nrf2 and its main negative regulator Keap1 are potential
pharmacological agents for a range of disease types including neurodegenerative conditions and
cancer. We describe the development of a series of 1,4ꢀdiphenylꢀ1,2,3ꢀtriazole compounds that
inhibit the Nrf2ꢀKeap1 proteinꢀprotein interaction (PPI) in vitro and in live cells and upꢀregulate
the expression of Nrf2ꢀdependent gene products.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Introduction Sustained oxidative stress and inflammation have been proposed as crucial steps in
the aetiology of chronic neurodegenerative conditions and cancer. A promising strategy to
mitigate these effects and curtail the development of these conditions involves inducing the
expression of cytoprotective proteins under the control of the transcription factor nuclear factor
erythroid 2 p45ꢀrelated factor 2 (Nrf2). Nrf2 is a cap ‘n’ collar basic leucine zipper transcription
factor that regulates the expression of genes with antioxidant response element (ARE) sequences
in their promoter regions. The genes controlled by Nrf2 encode a group of around 500 proteins
that have roles in metabolism (e.g. NAD(P)H quinone oxidoreductaseꢀ1 (NQO1), heme
oxygenaseꢀ1 (HO1), glutathione conjugation enzymes), redox homeostasis (e.g. glutathione
biosynthetic and reducing enzymes, thioredoxin, thioredoxin reductase) and autophagy (e.g.
sequestosomeꢀ1/p62) amongst other processes.1ꢀ3 In particular, upꢀregulation of phase II drug
metabolism enzymes plays a role in removing the polar, reactive products of primary (phase I)
metabolites, thereby limiting their genotoxic potential.
ACS Paragon Plus Environment
2

Page 3 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
There has been considerable progress recently in understanding the regulation of Nrf2 in the
cellular context and this has provided a number of intervention opportunities to control its
activity. The activity of Nrf2 under basal conditions is limited by its short halfꢀlife (< 10 min).4, 5
This is controlled largely by the rapid ubiquitination and proteosomal destruction of Nrf2, a
process that is orchestrated by the ubiquitination facilitator protein Kelchꢀlike ECHꢀassociated
protein 1 (Keap1).6ꢀ9 A ‘hinge and latch’ mechanism has been proposed in which two sites in the
Nrf2 Nꢀterminal Neh2 domain interact with Keap1, a high affinity 79ETGE82 ‘hinge’ motif (Kd ~
5 nM) that tethers Nrf2 to one arm of the Keap1 dimer and a lower affinity 26QDIDLG31 ‘latch’
motif (Kd ~ 1000 nM).10 Keap1 uses a cyclical mechanism to target Nrf2 for degradation,
whereby Nrf2 binds sequentially to the Keap1 dimer, first through the high affinity motif to form
the open conformation of the Keap1ꢀNrf2 protein complex, and then through the lower affinity
motif to form the closed conformation of the complex.11 The closed conformation allows
ubiquitination of lysine residues in the intervening Neh2 domain sequence of Nrf2 via a pendant
Cullin 3 ubiquitination complex.12ꢀ15 The ubiquitinated Nrf2 is released from Keap1 and
degraded by the 26S proteasome, free Keap1 is regenerated and able to bind to newly translated
Nrf2, and the cycle of Keap1ꢀmediated degradation of Nrf2 begins again.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Under conditions of oxidative stress, or on exposure to electrophilic compounds, cysteine
sensors within Keap1 (e.g. Cys151, 273, 288) are oxidized or conjugated, leading to
accumulation of the closed conformation of the Keap1ꢀNrf2 complex. It is proposed that the
resulting structural change inactivates the ubiquitination process, Nrf2 is not released, and free
Keap1 is not regenerated, allowing newly synthesized Nrf2 to accumulate within the cell and
exert its transcriptional activity.1, 16ꢀ18 A number of electrophilic Nrf2 inducers have been widely
studied, including the broccoliꢀderived isothiocyanate compound sulforaphane 1 and the
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 36
1
2
synthetic compound oltipraz 2 (Figure 1).1, 16ꢀ19 Sulforaphane upꢀregulates the expression of a
range of AREꢀresponsive genes in treated cells and has cytoprotective effects in vivo including
reducing the number of tumors that develop in carcinogen exposure studies.20, 21 Furthermore, in
humans, intervention with sulforaphaneꢀrich broccoli sprouts accelerates the detoxification of
aflatoxin and airborne pollutants, suggesting frugal means to counteract the health risk associated
with longꢀterm exposures of environmental toxins.22, 23 Sulforaphane also has promise for
treatment of autism spectrum disorder – a recent placeboꢀcontrolled, doubleꢀblind, randomized
clinical trial has shown that young men with autism who received daily oral doses of broccoli
sprouts delivering sulforaphane (50ꢀ150 µmol) for 18 weeks had significant improvement in
social interaction, abnormal behavior, and verbal communication.24
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
An alternative approach to enhance Nrf2 activity involves direct disruption of the Keap1ꢀNrf2
interaction at the proteinꢀprotein interface. Thus compounds that interact with the Keap1 Kelch
domain have the potential to disrupt Nrf2 binding in a reversible manner and inhibit its
ubiquitination. Such compounds may have advantages over electrophilic inducers that include
improved specificity and reduced offꢀtarget effects due to the lack of a covalent interaction. We
and others showed that peptides (e.g. 3ꢀ5) that mimic Nrf2 were able to disrupt the Keap1ꢀNrf2
interaction.25ꢀ29 A few small molecule Nrf2 activators that inhibit the Keap1ꢀNrf2 proteinꢀprotein
interaction (PPI) have been reported recently, confirming the potential of a direct targeting
strategy. These direct small molecules inhibitors were identified by virtual screening (6, 7), high
throughput screening (8ꢀ10) and subsequent SAR studies, or by structureꢀbased design and
molecular determinants analysis (11) (Figure 1).30ꢀ35
In this study we have used the available crystal structures of the human Keap1 Kelch domain,
both alone36 and in complex14 with peptides to identify potential small molecule inhibitors of the
ACS Paragon Plus Environment
4

Page 5 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Keap1ꢀNrf2 proteinꢀprotein interaction using an in silico fragmentꢀbased approach. Analysis of
the binding poses and proteinꢀinteractions of the most recurrent molecular scaffolds led us to
propose the 1,4ꢀdiarylꢀ1,2,3ꢀtriazole scaffold as a candidate for SAR investigations. As docking
of a prototypical ligand of this new series demonstrated encouraging scores, we synthesized
small libraries of 1,4ꢀdiarylꢀ1,2,3ꢀtriazoles and evaluated their ability to disrupt the Keap1ꢀNrf2
PPI using a fluorescence polarization assay that we developed and validated previously.25 In
parallel, we investigated their ability to induce the expression of the Nrf2 target enzyme NQO1
in Hepa1c1c7 cells, an assay that has been widely applied in the discovery and evaluation of
Nrf2 inducing agents.37, 38 Further studies in dose response experiments allowed us to extract
favorable features for biological activity and to identify a narrow set of compounds with
promising activity. Compound 22s was not cytotoxic over a wide concentration range, showed
upꢀregulation of Nrf2 protein levels and induced the expression of downstream gene products in
a manner comparable to that of sulforaphane 1. Intriguingly, 22s appears to disrupt the PPI in
live cells in a manner that is distinct from sulforaphane and other electrophilic inducers. Recently
we have shown that compound 22s (PMI, HB229) is a promising tool compound in the area of
mitochondrial autophagy (mitophagy).39 It appears to induce mitophagy in a p62 and Nrf2
dependent manner, but its activity is only partially dependent upon Parkin and PINK1. In this
respect the compound has intriguing functional differences from the prototype inducers of
mitophagy, the ionophores CCP and FCCP, and from 1 which lacks equivalent mitophagyꢀ
inducing activity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound Design Compounds from a subset of the ZINC database (“clean fragments”,
~178,000 compounds)40 were docked into the human Keap1 Cꢀterminal Kelch domain structure
(PDB entry: 2FLU) in the Nrf2 ‘ETGE’ peptide binding site14 and ranked using the Autodock 4.2
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 36
1
2
and DOCK 6.6 software packages.41, 42 After further refinement of the highest scoring ligands
(Autodock ꢁGbinding ≤ ꢀ7.5 kcal/mol) using extended docking protocols, the 364 highestꢀscoring
molecules (Autodock ꢁGbinding ≤ ꢀ8.0 kcal/mol) were selected and analyzed. The results indicated
that a relatively small group of molecular scaffolds dominated the compound set; common
features could be identified in these recurrent scaffolds and in hits recently identified by
structureꢀbased virtual screening.31 Two examples, compounds 12 and 13, are shown in Figure 2.
Although compound 12 contains a thiazolidinone moiety, a known PAINS (PanꢀAssay
Interference Compounds)43 substructure, it was included in our initial analyses because related 2ꢀ
oxoꢀrhodanine derivatives have been shown to interact reversibly with Keap1 in fluorescence
anisotropy assays (the authors did not present data regarding cellꢀbased activity).31 However, we
did not pursue SAR studies with 12 beyond the in silico evaluation. Analysis of the calculated
binding modes of the highestꢀscoring compounds suggested that carboxylate or nitro substituents
formed favorable electrostatic and hydrogen bond interactions with the Arg 380, 415, 483 and
Asn 382 residues of Keap1, in some cases the scaffold formed additional hydrogen bond
interactions with, for example, the side chain of Ser 602 within the binding site (Figure 2). Using
the distances between the favorable substituent features of the highestꢀscoring docked ligands
and the distance between the glutamate side chains of the Nrf2 ETGE sequence as a guide, we
designed a new series of ligands based upon a simple 1,4ꢀdiarylꢀ1,2,3ꢀtriazole scaffold that can
be both easily synthesized and functionalized to mimic the pharmacophores identified from the
docking exercise. Our initial docking studies with compounds from this series suggested that the
calculated binding energies should be similar to those identified by virtual screening and that
they could form similar interactions within the binding pocket (Figure 2, e.g. 24d).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
6

Page 7 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis and Fixed-Dose Screening We initially synthesized a small combinatorial library of
36 1,4ꢀdiarylꢀ1,2,3ꢀtriazoles by reacting ethynylꢀ and azidoꢀbenzene intermediates bearing meta
or para nitro, carboxylic acid or carboxamide groups (21ꢀ26aꢀf, Scheme 1). The compounds
were synthesized by copper catalyzed alkyneꢀazide cycloadditions under microwave irradiation
(Scheme 1). The azidobenzene intermediates 14 were prepared using standard azidation
conditions from the corresponding anilines or by direct substitution of a halogenꢀsubstituted
benzene (using NaN3 in DMF). The ethynylbenzene precursors 15ꢀ20 were prepared in two steps
from the corresponding halogenated derivatives by Sonogashira coupling with
trimethylsilylacetylene followed by cleavage of the TMS group by treatment with K2CO3 in
MeOH.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The compounds were screened in vitro using a fluorescence polarization assay that we
developed previously in which we evaluated the ability of 21ꢀ26aꢀf to disrupt the interaction
between the Kelch domain of Keap1 and a fluorescein labelled peptide based on the high affinity
‘ETGE’ motif from Nrf2 (FITCꢀβꢀDEETGEFꢀOH, β for βꢀalanine).25 We have demonstrated that
the native Nrf2 Neh2 domain is capable of displacing the fluorescent peptide with a Ki of ~ 25
nM, consistent with the Kd determined using isothermal calorimetry of 5.3 nM.25 The
competition experiment was initially carried out using a fixed 100 µM concentration of inhibitor
(21ꢀ26aꢀf, Table 1). Five of the six compounds with a metaꢀcarboxylic acid substituent on the
triazole 4ꢀphenyl ring exhibited encouraging activity (24a,cꢀf). Compounds bearing a 4ꢀ
carboxamide substituent in the 4ꢀphenyl ring (25aꢀe) appeared to interfere consistently with the
FP signal (FP value greater than the control), despite the molecules lacking innate fluorescence
properties. However, otherwise there was not a significant discrimination between the meta or
para substitution pattern on either the 1ꢀ (R1) or 4ꢀphenyl (R2) substituent.
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 36
1
2
3
4
5
6
7
8
9
In addition to the fixed dose FP assays we screened the 36ꢀmember library to evaluate Nrf2
dependent transcription using an established colorimetric NQO1 induction assay in Hepa1c1c7
mouse hepatoma cells.37, 38 Initially we tested the compounds at a fixed concentration of 10 µM,
with an exposure time of 24 h (Table 1). Results from the NQO1 induction experiments were
more discriminating than the fixed dose FP assay; six compounds induced NQO1 expression >
1.5ꢀfold. Of these compounds four had a 4ꢀ(3ꢀnitrophenyl)triazole motif (22, Scheme 1). All of
the active compounds had a nitro/nitro or nitro/carboxamide substitution pattern. In contrast
many of the compounds with carboxylic acid substituents were inactive (relative induction ~1).
The lack of activity of the carboxylate derivatives may relate to poorer cell penetration due to
their net negative charge at physiological pH. Based upon the observations from the fixed dose
FP and NQO1 assays, including the relatively good activity in the FP assay, combined with
promising cellꢀbased activity (Table 1) we chose a metaꢀnitro substitution on the 4ꢀ
phenyltriazole moiety for the second generation library.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The second series (22gꢀt) was synthesized using 1ꢀethynylꢀ3ꢀnitrobenzene 16 and an array of
azidobenzenes 14gꢀt bearing substituents in the 3ꢀposition with a range of electronic and steric
properties (Scheme 2). Compounds 22gꢀt showed consistent activity in the FP assay with
percentage inhibitions in the 60ꢀ80% range (100 µM fixed dose) with the exception of
compounds 22p (R1 = F) and 22t (R1 = CN), which were less active. All of the compounds
showed promising activity in the NQO1 induction assay with compound 22s (R1 = I)39
demonstrating a robust 4ꢀfold induction at 10 µM (Table 2).
As the carboxylic acid function appeared to confer favorable activity in the FP assays, but was
detrimental in the cellꢀbased NQO1 assays, we prepared a third series of compounds to examine
the influence of a tetrazole moiety as a carboxylate isostere. We first envisioned the synthesis of
ACS Paragon Plus Environment
8

Page 9 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
tetrazoles from the corresponding carboxamide compounds synthesized in the first library, using
SiCl4 and NaN3 in a mixture MeCN/DMF.44 No reaction occurred in these conditions, which led
us to consider the synthesis of the tetrazole ring from a cyano precursor. Using 1ꢀethynylꢀ3ꢀ
cyanobenzene 27 and a small series of 3ꢀsubstitutedꢀphenylazides selected from the previous
screening exercise (14b,c,h,l,q,s,t), a series of 1,2,3ꢀtriazoles 28 was prepared. This series was
evaluated along with the corresponding tetrazoles 29 that were synthesized in good yields by
[3+2] cycloaddition of the cyano compounds with NaN3 in a sealed tube, using ammonium
chloride and catalytic lithium chloride in refluxing DMF (Scheme 3, Table 3).45
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The tetrazoles and their cyano precursors were tested in the fixed dose FP assay at a
concentration of 100 µM (Table 3); only two of the compounds showed significant activity, 29b
and 29c (~70% inhibition). The EC50s of these two tetrazoloꢀcompounds, determined in a doseꢀ
response experiment (vide infra), were >50 µM. The NQO1 induction ability (Table 3) was
generally poor indicating that they were not effective Nrf2 inducers and that in this context the
tetrazole substituents were relatively poor carboxylate isosteres.
Dose-Response Evaluation of Selected Analogues Based upon the fixed dose assessments, a
selection of compounds from the 22aꢀt series was tested in FP and NQO1 dose response assays.
The EC50s of compounds were in the range 5ꢀ39 µM (Table 4). Compounds 22d (R1 = CO2H),
22h (R1 = Me), 22q (R1 = Cl) and 22s (R1 = I) were the most active with EC50 values in the range
5ꢀ10 µM. A further five compounds displayed intermediate activity with EC50 values in the 11ꢀ
15 µM range: 22b (R1 = NO2), 22l (R1 = OEt), 22o (R1 = NMe2), 22k (R1 = OMe), 22m (R1 =
1,3ꢀdioxolane). The results indicate that a range of substituents is tolerated at this position, with
the possible exception of the tꢀbutyl (22i, 37.8 µM) and thiomethyl (22n, 38.5 µM) substituents.
Compound 22s was evaluated further in the FP assay. Experiments in the presence of varying
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 36
1
2
3
4
5
6
7
8
9
concentrations of the detergent Triton X100 (0 – 1% v/v) showed no change in the EC50 of the
compound (Figure S2). Treatment of the Keap1 protein with 22s for 2.5 h followed by dialysis
yielded protein that demonstrated a similar Kd to Keap1 treated with buffer alone (Figure S3).
Taken together these data provide support for the proposed reversible binding of 22s to Keap1.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The in vitro activity of the most active compounds determined in the doseꢀresponse competitive
FP assay is close to the inhibitory potency of peptides described previously that mimic the
‘ETGE’ and the ‘DLG’ motifs (3 and 5, EC50 = 5.4 µM and 17.1 µM respectively, Figure 3).25 In
the ‘hinge and latch’ type binding mechanism suggested for the interaction of Nrf2 with Keap1,
both motifs must be bound to allow ubiquitination of Nrf2.11, 17 In principle, disruption of the
interaction of the low affinity ‘DLG’ motif alone would thus be sufficient to prevent the
ubiquitination of Nrf2 and allow its release from Keap1 tethering.
The same group of compounds was evaluated along with the prototype Nrf2 inducing agents
sulforaphane 1 and oltipraz 2 (positive controls) in a dose response assay to determine the
38
concentration range over which NQO1 induction occurred.37,
The results (Table 4) are
expressed as the concentration required to induce a twoꢀfold increase in NQO1 enzymatic
activity (CD value) after 24 h. Five compounds exhibited CD values below 1.5 µM, indicating
activities that are comparable to 1 (0.25 µM) and 2 (10.2 µM). There is a relatively good
correlation between the FP assay EC50 concentrations and the NQO1 assay CD values with
compounds 22h, 22q and 22s showing promising activity in both experiments.
Stabilization of Nrf2 and induction of Nrf2-dependent genes We selected three compounds
with differing activity in the FP and NQO1 assays for further evaluation (22h, 22i and 22s).
Compound 22i was less active in the FP assay and did not significantly induce NQO1 enzymatic
ACS Paragon Plus Environment
10

Page 11 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
activity, whereas 22h and 22s showed similar activity in the FP experiments and induction of
NQO1 in the low micromolar range (Table 4, Figure 3). The effects of the compounds on Nrf2
activity were evaluated in Hepa1c1c7 cells and compared with sulforaphane 1 (Figure 3). The
cells were incubated with the compounds at a fixed concentration of 10 µM with different
exposure times. Treatment with 1, 22h and 22s resulted in a marked increase in Nrf2 protein
levels (Figure 3), comparable after 6 h to those produced by 1; the effect from 22i was reduced
compared to the other compounds. The increase in Nrf2 levels was timeꢀdependent; the highest
level of induction occurred after 3ꢀ6 h for 22s whereas for 1, maximal induction was achieved
after 1 h. Upꢀregulation of the expression of the Nrf2ꢀdependent enzymes HOꢀ1 and NQO1 was
also observed. For these proteins, the time course was the same for both 22h, 22s and 1; peak
HOꢀ1 induction occurred after 6 h, but NQO1 protein was not detected before 24 h, consistent
with previously reported differences in kinetics of their transcripts.46 The effect of 22i on HOꢀ1
expression was significant although upꢀregulation of NQO1 was close to undetectable. This
suggests that the compounds may have differential effects on downstream targets, an observation
that warrants further investigation.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Disruption of the Keap1-Nrf2 interaction in live cells As 22s was the most robust inducer in
these series, we next evaluated the ability of this compound to disrupt the Keap1ꢀNrf2 interaction
in live cells by use of a FRETꢀbased assay system in which Keap1 was labelled with the
fluorophore mCherry and Nrf2 was labelled with EGFP, as previously described.11, 47 These
proteins when ectopically expressed into HEK293 cells recapitulate the native Keap1ꢀNrf2
interaction.11 Fluorescence lifetime imaging microscopy (FLIM) was then used to determine the
timescale over which EGFP fluorescence emission occurs. Shorter lifetimes reflect a more
efficient energy transfer to the mCherry FRET partner, whereas longer timescales reflect
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 36
1
2
3
4
5
6
7
8
9
dissociation of the complex. The distribution of FRET efficiencies (FE) associated with
individual pixels in the microscopy images is related to the proximity of the two fluorophores.
Previous experiments have characterized two FE populations, a low (~13%) FE population in
which the fluorophores are relatively distant from each other, corresponding to Nrf2 bound to
Keap1 by only the high affinity ETGE motif (termed ‘open’ conformation of the Keap1ꢀNrf2
complex), and a high (~21%) FE population in which the fluorophores are closer together, in
which Nrf2 is bound to Keap1 by both the low affinity DLG and high affinity ETGE motifs
(termed ‘closed’ conformation of the Keap1ꢀNrf2 complex).11
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Exposure to electrophilic inducers such as sulforaphane 1, causes accumulation of the complex
in the ‘closed’ conformation, and it was proposed that, in this electrophileꢀinactivated ‘closed’
complex, Nrf2 is not ubiquitinated or released by Keap1, which means that the free Keap1 dimer
is not regenerated, allowing newly synthesized Nrf2 to accumulate.11 In contrast to sulforaphane
1 and other electrophilic inducers, when 22s (10 µM) was administered to HEK293 cells that had
been coꢀtransfected with EGFPꢀNrf2 and Keap1ꢀmCherry, the high FE population was
diminished and the low FE population predominated (Figure 4 and Tables S1, S2). This suggests
that the compound disrupts the interaction between the Nrf2 DLG motif and Keap1, but not the
ETGE motif. This observation is consistent with direct inhibition of the Keap1ꢀNrf2 proteinꢀ
protein interaction.
Cytotoxicity We then compared the cytotoxicity of 22s and 1 in Hepa1c1c7 cells using a
sulforhodamine B cytotoxicity assay. 22s was not cytotoxic (cell viability ~100%) up to a
concentration of 200 µM over 24 and 48 h periods. 1 exhibited cytotoxicity at a concentration
greater than 10ꢀ30 µM, indicating that 22s has a broader therapeutic window for NQO1
ACS Paragon Plus Environment
12

Page 13 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
induction (Figure S1). For 1 the therapeutic (chemopreventive) index ([cytotoxicity GI50]/[NQO1
CD]) was ~250 in the Hepa1c1c7 cell line, compared to that of 22s which was >1000.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conclusion In this study we developed a series of compounds with a 1,4ꢀdiarylꢀ1,2,3ꢀtriazole
scaffold using input from virtual screening of a fragment library combined with structural and
distances considerations. We have synthesized three small libraries of triazole derivatives
bearing a range of substituents on the 1ꢀ and 4ꢀaryl substituents. Amongst the substitution
patterns evaluated the best combinations for cellꢀbased activity were a metaꢀnitro group on the 4ꢀ
phenyl ring and a metaꢀnitro, methyl or halogen on the 1ꢀphenyl unit. The ability of three
compounds 22h, 22i and 22s to stabilize Nrf2 and induce the expression of its target genes
NQO1 and HOꢀ1 in a concentrationꢀ and timeꢀdependent manner correlated with the activity of
the compounds in the PPI FP assay. The most active compound 22s binds reversibly to Keap1
and was not cytotoxic over a wide concentration range, but had similar cellꢀbased activity to
sulforaphane 1. Intriguingly, live cellꢀbased imaging experiments suggest that 22s drives the
formation of the ‘open’ conformation of the Keap1ꢀNrf2 complex. Overall, the molecular
modeling and experimental results are consistent with direct targeting of the Keap1ꢀNrf2
interaction. Further detailed investigations are underway to elucidate the precise underlying
mode of action. Despite the structural simplicity of this compound, it has intriguing properties as
both a cell permeable biochemical probe for Nrf2 activation and mitophagy,39 and as a template
for further structural refinement.
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 36
1
2
3
4
FIGURES
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Structures of known indirect (e.g. 1, 2) and direct (3ꢀ11) Keap1ꢀNrf2 inhibitors with
their EC50 values for inhibition of the Keap1ꢀNrf2 interaction.30ꢀ35
ACS Paragon Plus Environment
14

Page 15 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Scaffolds 12 and 13 identified using in silico docking calculations (top) and the diaryl
triazole scaffold (e.g. 24d) docked in the Keap1 Kelch domain.
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 36
1
2
3
4
5
6
7
8
9
120
100
80
100
80
60
40
20
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
60
40
20
0.1
1
10
100
0.1
1
10
100
Concentration (ꢀM)
Concentration (ꢀM)
Figure 3. Dose response curves (top) for the fluorescence polarization assays of 3 (▲) and 5 (■)
and 22h (○), 22i (∆) and 22s (□). Western Blot analysis (bottom) demonstrating upregulation of
Nrf2 and its downstream targets HOꢀ1 and NQO1 in Hepa1c1c7 cells after exposure to 10 µM
concentrations of 1, 22h, 22i and 22s over a 24 h period.
ACS Paragon Plus Environment
16

Page 17 of 36
Journal of Medicinal Chemistry
1
2
3
4
EGFP Image
EGFP Lifetime
EGFP Lifetime
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A. Cell 1
1.9 ns
2.6 ns
B. Cell 1 + 1 hr
1.9 ns
2.6 ns
C. Cell 2
1.9 ns
2.6 ns
D. Cell 2 + 1 h 22s
1.9 ns
2.6 ns
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 36
1
2
3
4
5
6
7
8
EGFP Image
FRET Efficiency FRET Efficiency
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
E. Cell 3
0%
30%
F. Cell 3 + 1 hr 1
0%
30%
G. Cell 4
0%
30%
H. Cell 4 + 1 h 22s
0%
30%
ACS Paragon Plus Environment
18

Page 19 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 4. The lifetime and FRET efficiency in the cytoplasm of EGFPꢀNrf2 transfected cells
imaged before and after treatment with 1 or 22s for 1 h. HEK293 cells were transfected with
EGFPꢀNrf2 + Keap1ꢀmCherry, and the lifetime (AꢀD) and FRET efficiency (EꢀH) in the
cytoplasmic compartment were calculated. A, B show the fluorescence lifetime data from a
single cell imaged twice, once at time 0 (A), and once 1 h later (B). C, D show the fluorescence
lifetime data from a single cell imaged twice, once at the basal state (C), and once 1 h after
treatment with 10 ꢂM 22s (D). The left column shows the EGFP image from which the lifetime
data are derived. The middle column shows a pictorial representation of the EGFP lifetime where
the color of the cell corresponds to the lifetime of EGFP, ranging from 1.9 ns to 2.6 ns as
indicated on the legend below the image. The right column shows the lifetime data from each
pixel of the image plotted on a graph, with lifetime on the xꢀaxis and frequency on the yꢀaxis. Eꢀ
H show the FRET efficiency data for individual EGFPꢀNrf2 + Keap1ꢀmCherry coꢀtransfected
cells which were imaged twice, once in the basal state (E, G) and once after 1 h treatment with 5
ꢂM 1 (F) or 10 ꢂM 22s (H). The left column shows the EGFP image from which the FRET
efficiency data are derived. The middle column shows a pictorial representation of the FRET
efficiency where the color of the cell corresponds to the FRET efficiency, ranging from 0% to
30% as indicated on the legend below the image. The right column shows the FRET efficiency
from each pixel of the image plotted on a graph, with FRET efficiency on the xꢀaxis and
frequency on the yꢀaxis. The FRET efficiency graphs (E, F) show that the FRET efficiency
distribution is altered by 1, which leads to an increase in the interaction at 21% FRET efficiency.
The FRET efficiency distributions are shown pictorially in the central column of E, F, where an
increase in the amount of yellow relative to green can be seen in response to 1 treatment. The
FRET efficiency graphs (G, H) show that the FRET efficiency distribution is altered by 22s,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 36
1
2
3
4
5
6
7
8
9
which leads to an increase in the interaction at 13% FRET efficiency. The FRET efficiency
distributions are shown pictorially in the central column of G,H, where an increase in the amount
of green relative to yellow can be seen in response to 22s treatment. White scale bars shown in
the figures represent a distance of 10 ꢂm.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
20

Page 21 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
SCHEMES
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthesis of the 36ꢀmember library of 1,4ꢀbiarylꢀ1,2,3ꢀtriazole derivatives.
Reagents and conditions: i. CuSO4.5H2O (3.3 mol %), ascorbic acid (10.0 mol %), tertꢀ
butanol/H2O 1/1, MW, 130 °C, 30 min.
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 36
1
2
3
4
Scheme 2. Synthesis of the refined click library 22g-t.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Reagents and conditions: i. see Scheme 1.
ACS Paragon Plus Environment
22

Page 23 of 36
Journal of Medicinal Chemistry
1
2
3
4
Scheme 3. Synthesis of the cyanoꢀ and tetrazoleꢀsubstituted series.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Reagents and conditions: i. see Scheme 1; ii. NH4Cl (1.25 eq.), NaN3 (1.25 eq.), LiCl (cat.),
DMF, reflux, 24 h.
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 36
1
2
3
4
TABLES.
5
6
7
8
9
Table 1. Fixed dose inhibition of the Keap1ꢀNrf2 interaction (FP) and induction of NQO1 by
compounds 21ꢀ26
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
FPa
21
22
23
24
25
26
72.6 74.8 83.8 82.3 ꢀb
48.0
a
46.5 81.0 87.5 63.8 2.34 32.0
b
81.3 39.4 77.3 92.8 ꢀb
61.0 80.1 89.6 82.3 ꢀb
46.1 64.5 76.5 87.6 ꢀb
70.7
78.6
ꢀb
c
d
e
62.2 65.3 15.4 96.3 43.7 53.3
f
NQO1c
21
22
23
24
25
26
1.00 3.38 0.89 1.14 1.02 1.05
1.81 2.30 1.06 0.85 1.33 2.06
a
b
c
d
e
f
0.99 ꢀd
0.90 0.78 0.79 0.81
0.66 0.83 0.47 0.86 0.79 0.71
1.07 2.08 0.86 0.91 0.98 0.72
0.98 2.09 0.95 1.07 1.02 1.02
Notes: a. Percentage inhibition at 100 ꢂM; b. Fluorescence interference; c. Fold induction at 10
ꢂM; d. Not determined.
ACS Paragon Plus Environment
24

Page 25 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
Table 2. Fixed dose inhibition of the Keap1ꢀNrf2 interaction (FP) and induction of NQO1 by
compounds 22b,d,fꢀt
7
8
9
Cpd
22b
22d
22f
R1
FPa
NQO1b
2.30
0.83
2.09
1.95
2.20
1.93
2.54
2.56
2.48
2.24
2.95
2.91
2.27
2.50
2.94
4.00
2.33
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NO2
CO2H
CONH2
H
81.0
80.1
65.3
66.9
68.9
69.8
66.9
68.2
69.3
67.2
77.7
65.4
42.6
60.4
68.2
67.2
36.8
22g
22h
22i
Me
tBu
CF3
OMe
OEt
3,4ꢀOCH2O
SMe
NMe2
F
22j
22k
22l
22m
22n
22o
22p
22q
22r
22s
22t
Cl
Br
I
CN
Notes: a. Percentage inhibition at 100 ꢂM; b. Fold induction at 10 ꢂM.
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 36
1
2
3
4
5
6
Table 3. Fixed dose inhibition of the Keap1ꢀNrf2 interaction (FP) and induction of NQO1 by
compounds 28 and 29
7
8
9
Cpd R1
FPa NQO1b Cpd FPa NQO1b
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3ꢀNO2
59.4 2.1
71.1 1.75
70.0 1.00
30.4 0.88
39.6 1.31
40.7 1.24
51.5 1.03
28b
28c
28h
28l
29b
29c
29h
29l
4ꢀCO2H 52.8 0.98
3ꢀMe
3ꢀOEt
3ꢀCl
3ꢀI
49.7 1.69
24.5 1.30
19.9 2.39
21.9 1.53
58.8 1.92
28q
28s
28t
29p
29s
3ꢀCN
29uc 13.7 0.84
Notes: a. Percentage inhibition at 100 ꢂM; b. Fold induction at 10 ꢂM; c. R1 = CN4H.
Table 4. EC50 values determined by FP assay and CD values from the NQO1 induction assay for
selected compounds.
Compound R1
FP EC50
(ꢂM)
NQO1
CD
(ꢂM)
3ꢀNO2
11.5 (± 1.4) ꢀ
22b
22d
22g
22h
22i
3ꢀCO2H
H
5.0 (± 1.9) >10
20.6 (± 5.6) >10
10.0 (± 2.7) 1.3
37.8 (± 7.1) >10
22.6 (± 6.4) ꢀ
3ꢀMe
3ꢀtBu
3ꢀCF3
3ꢀOMe
3ꢀOEt
22j
22k
22l
15 (± 5.8)
4.7
11.3 (± 3.3) ꢀ
ACS Paragon Plus Environment
26

Page 27 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
3,4ꢀ
OCH2O
15 (± 3.4)
ꢀ
22m
6
7
8
9
3ꢀSMe
38.5 (± 9.5) ꢀ
22n
22o
22q
22r
22s
1
3ꢀN[Me]2 12.7 (± 3.1) 2.0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3ꢀCl
3ꢀBr
3ꢀI
ꢀ
8.8 (± 1.7) 0.7
24.5 (± 7.1) 1.2
7.1 (± 0.7) 0.6
ꢀ
ꢀ
0.3
ꢀ
10.2
2
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 36
1
2
3
4
ASSOCIATED CONTENT
5
6
7
8
9
Supporting Information. Detailed chemical synthesis and experimental procedures for
molecular modeling, fluorescence polarization assays, NQO1 assays and Western blot
experiments. This material is available free of charge via the Internet at http://pubs.acs.org.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Author
*UCL School of Pharmacy, University College London, London WC1N 1AX. Email:
g.wells@ucl.ac.uk, Tel: +44 (0)20 7753 5935, Fax: +44 (0)20 7753 5964.
Present Addresses
† Ecole Normale Supérieureꢀ PSL research University, Sorbonne Universités ꢀ UPMC Univ Paris
06, CNRS ꢀ UMR7203, Laboratoire des Biomolécules, 24 rue Lhomond, 75005 Paris, France
††Department of Medical Biochemistry, Tohoku University School of Medicine, 2ꢀ1 Seiryoꢀ
machi, Aobaꢀku, Sendai 980ꢀ8575, Japan.
Author Contributions
All authors have given approval to the final version of the manuscript.
Funding Sources
The authors acknowledge Cancer Research UK (C9344/A10268) (H.B., A.F., M.S., G.W.) and
(C20953/A10270) (A.DꢀK), the MRC (L.B. and A.DꢀK), the BBSRC (BB/L01923X/1) (A.DꢀK
and G.W) and the School of Pharmacy (G.W.) for financial support.
ACKNOWLEDGMENT
ACS Paragon Plus Environment
28

Page 29 of 36
Journal of Medicinal Chemistry
1
2
3
4
5
6
We thank Dr Tadayuki Tsujita and Prof. John Hayes (University of Dundee) for providing the
purified Keap1 Kelch domain protein.
7
8
9
ABBREVIATIONS
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
FP, fluorescence polarization; GI, growth inhibition.
REFERENCES
1.
stresses via the Keap1ꢀNrf2ꢀARE pathway. Annu Rev Pharmacol Toxicol 2007, 47, 89ꢀ116.
2. Hayes, J. D.; McMahon, M.; Chowdhry, S.; DinkovaꢀKostova, A. T. Cancer
Kensler, T. W.; Wakabayashi, N.; Biswal, S. Cell survival responses to environmental
chemoprevention mechanisms mediated through the Keap1ꢀNrf2 pathway. Antioxid Redox
Signal 2010, 13, 1713ꢀ1748.
3.
Cheng, X.; Siow, R. C.; Mann, G. E. Impaired redox signaling and antioxidant gene
expression in endothelial cells in diabetes: a role for mitochondria and the nuclear factorꢀE2ꢀ
related factor 2ꢀKelchꢀlike ECHꢀassociated protein 1 defense pathway. Antioxid Redox Signal
2011, 14, 469ꢀ487.
4.
Hong, F.; Sekhar, K. R.; Freeman, M. L.; Liebler, D. C. Specific patterns of electrophile
adduction trigger Keap1 ubiquitination and Nrf2 activation. J Biol Chem 2005, 280, 31768ꢀ
31775.
5.
against electrophiles and reactive oxygen species. Adv Enzyme Regul 2006, 46, 113ꢀ140.
6. Itoh, K.; Wakabayashi, N.; Katoh, Y.; Ishii, T.; Igarashi, K.; Engel, J. D.; Yamamoto, M.
Kobayashi, M.; Yamamoto, M. Nrf2ꢀKeap1 regulation of cellular defense mechanisms
Keap1 represses nuclear activation of antioxidant responsive elements by Nrf2 through binding
to the aminoꢀterminal Neh2 domain. Gene Dev 1999, 13, 76ꢀ86.
ACS Paragon Plus Environment
29