Computer-Aided Design and Synthesis of 1-{4-[(3,4-Dihydroxybenzylidene)amino]phenyl}-5-oxopyrrolidine-3-carboxylic Acid as an Nrf2 Enhancer

Article abstract of DOI:10.1002/cplu.201700539

The design and synthesis of a novel nuclear factor erythroid 2-related factor 2 (Nrf2) enhancer is reported. Using a structure-based virtual screening approach, several commercially available compounds were identified as having high probability to interac

Full text of DOI:10.1002/cplu.201700539

Accepted Article
Title: Computer-aided design and synthesis of (1-(4-((3,4-
dihydroxybenzylidene)amino)phenyl)-5-oxopyrrolidine-3-
carboxylic acid) as a novel Nrf2 enhancer
Authors: Shirin Kahremany, Ilana Babaev, Pinhas Hasin, Tigist
Tamir, Tali Ben-Zur, Guy Cohen, Sagiv Weintraub, Zhengyu
Jiang, Daniel Offen, Ben Major, Shai Rahimipour, Hanoch
Senderowitz, and Arie Gruzman
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemPlusChem 10.1002/cplu.201700539
Link to VoR: http://dx.doi.org/10.1002/cplu.201700539
A Journal of
www.chempluschem.org

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Computer-aided design and synthesis of (1-(4-((3,4-
dihydroxybenzylidene)amino)phenyl)-5-oxopyrrolidine-3-
carboxylic acid) as a novel Nrf2 enhancer
Shirin Kahremany,[a] Ilana Babaev,[a] Pinhas Hasin,[a] Tigist Y. Tamir,[b] Tali Ben-Zur,[c] Guy Cohen,[d]
Zhengyu Jiang,[e] Sagiv Weintraub,[a] Daniel Offen,[c] Shai Rahimipour,[a] M. Ben Major,[b] Hanoch
Senderowitz,[a] Arie Gruzman*[a]
[a]
Dr. S. Kahremany, Ilana Babaev, Dr. P. Hasin, Dr. S. Weintraub, Prof. S. Rahimipour, Prof. H. Senderowitz, Prof. A. Gruzman , Department of Chemistry,
Faculty of Exact Sciences, Bar-Ilan University, Ramat Gan, 5290002 (Israel)
E-mail: gruzmaa@biu.ac.il
[b]
[c]
Dr. Tigist Y. Tamir, Prof. M. Ben Major, Department of Pharmacology and the Lineberger Comprehensive Cancer Center, University of North Carolina at
Chapel Hill, Chapel Hill, NC, 27599, (USA)
Dr. T. Ben-Zur, Prof. D. Offen, Felsenstein Medical Research Center, Sackler Faculty of Medicine, Tel-Aviv University, Rabin Medical Center-Beilinson
Campus, Petah Tikva 49100 (Israel)
[d]
[e]
Dr. Guy Cohen, The Skin Research Institute. The Dead-Sea & Arava Science Center. Tamar regional council, Dead-Sea mobile post 86910, (Israel)
Dr. Zhengyu Jiang, Jiangsu Key Laboratory of Drug Design and Optimization, Department of Medicinal Chemistry, School of Pharmacy, China
Pharmaceutical University Nanjing, Jiangsu, (P. R. China)
Supporting information for this article is given via a link at the end of the document.
Abstract: We report on the design and synthesis of a novel nuclear
factor erythroid 2-related factor 2 (Nrf2) enhancer. Using a structure-
based virtual screening approach, we identified several commercially
available compounds with high probability to interact with the Nrf2
binding pocket in the Kelch-like ECH-associated protein 1 (Keap1).
Keap1 is an adaptor protein that recruits Nrf2 to a Cul3 (Cullin-3)-
dependent ubiquitin ligase complex. The identified compounds were
tested against rat pheochromocytoma PC-12 cells for their
cytoprotective activity and one compound (SKT359126)
demonstrated an Nrf2-mediated cell protective effect. Based on the
SKT359126 structure, twenty-three novel derivatives were
synthesized and evaluated. Among the screened derivatives,
compound 16 (1-(4-((3,4-dihydroxybenzylidene)amino)phenyl)-5-
oxopyrrolidine-3-carboxylic acid), demonstrated better activity than
the parent molecules in activating the Nrf2 transduction pathway in a
dose- and time-dependent manner. This compound can serve as a
promising starting point for the development of novel therapeutics for
the treatment of oxidative stress-related diseases.
The human body is constantly exposed to numerous oxidative
and electrophilic chemicals.[1] The imbalance between
biochemical processes leads to the production of oxidative
species that cause many human diseases.[2] The human
antioxidant defense system is involved in the direct
neutralization of these species.[3] The antioxidative system
includes many enzymes, cofactors and vitamins.[4, 5] The activity
of several antioxidant cytoprotective enzymes is controlled by
the Nrf2/antioxidant response element (ARE) transcription
pathway.[6] This pathway includes three main components: ARE,
Nrf2 and Keap1.[7] The antioxidative response is initiated by
Keap1. This protein is a cytoplasmic 624 amino-acid multi-
domain protein that functions as a substrate adaptor protein to
bring Nrf2 to a Cul3-dependent ubiquitin ligase complex. Keap1
contains (i) an N-terminal region (NTR, amino acids 1–60), (ii) a
BTB (broad complex/tramtrack/bric-à-brac) proteins domain
(amino acids 61–179)[8] through which Keap1 forms
a
homodimer and also interacts with Cul3, (iii) an intervening
region (IVR, amino acids 180–314), which is cysteine-rich and
contains eight cysteine residues among its 134 amino acids, (iv)
a Kelch-domain made of six-bladed β-propeller in which each
1. Introduction
1
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
blade of the propeller (I–VI) is composed of four β-strands (A–D)
and (v) a C-terminal region (amino acids 599–624).[9-11]
used.[22]
tetrahydroisoquinolines,
To
date,
compounds
that
belong
to
sulfonamides,
oxadiazoles,
Under physiological intracellular and environmental conditions,
cytosolic Nrf2 is maintained at a low basal level by constitutive
proteasomal degradation.[12] The specific binding of the Neh2
(Nrf2-ECH homology 2) regulatory domain of Nrf2 to Keap1[13]
thiopyrimidines and furancarbazones have been reported as
Keap1–Nrf2-interaction inhibitors.[22-26]
Covalent electrophilic Keap1 inhibitors that bind to thiol
residues have been also reported.[28] However, these
compounds have the potential to react with other cysteine
moieties in different proteins and may therefore cause
unpredictable side effects. Thus, the development of PPI
inhibitors of the Nrf2–Keap1 complex is a most promising route.
Here we describe the design and synthesis of compound 16 that
belongs to the oxopyrrolidine-based class of Nrf2 activators. The
compound shows a cell-protective effect under severe oxidative
stress conditions by activating the Nrf2 transduction pathway in
a dose- and time-dependent manner. Nrf2 activators that belong
to the oxopyrrolidine class of chemical compounds have not
been described yet. Our prototype molecule can serve as a
promising starting point for the development of novel
therapeutics for the treatment of oxidative stress-related
diseases.
ensures the cytoplasmic sequestration and ubiquitination of
14]
Nrf2,[11,
thereby leading to Nrf2 proteasomal regulated
degradation.[15] Keap1 has highly reactive thiol groups in its 27
cysteine residues causing the protein to be very sensitive to
oxidative environments.[16]
Upon exposure to high levels of reactive oxygen species (ROS),
Nrf2 is released from Keap1 and translocates into the
nucleus.[15, 17] Inside the nucleus, the Nrf2 forms heterodimers
with other basic leucine zipper domain (bZIP) transcription
factors, including small Maf proteins, c-Jun, activating
transcription factor-4, c-Fos, Fos-related antigen-1 (Fra-1) and
setera.[18] These formed complexes bind to ARE in targeted
gene promoters and upregulates their expression.[19] These
genes [for example, glutathione S-transferases (GSTs),
NADP(H), quinone oxidoreductase
1
(NQO1) and heme-
oxygenase-1(HO1)] are critical for effective antioxidative stress
response.
2. Results and Discussion
The Keap1 structure was used by our team as a basis for the
molecular modeling of potential Nrf2 enhancers. Keap1 forms a
homodimer, and each dimer binds one molecule of Nrf2, via its
two Kelch domains, with one low-affinity (DLG motif, residues
24–31, latch) and one high-affinity binding site (ETGE motif,
residues 78–82, hinge).[20]
The activity of Nrf2 is negatively regulated by binding to the
Kelch domain of Keap1. Thus, compounds that interact with the
Kelch domain and disrupt Nrf2 binding lead to the intracellular
release of Nrf2. Free Nrf2 translocates to the nuclei and
constantly activates the transcription of many antioxidant genes.
As mentioned above, interactions between Keap1 and Nrf2 are
mediated through the ETGE motif to a much greater extent than
through the DLG one. Thus, we decided to focus on the ETGE
motif as a target for the design of a competitive ligand that is
able to release Nfr2. Based on the available crystal structures of
the human Keap1 Kelch domain, as well as on the 16-mer-Nrf2
ETGE peptide (PDB codes 1ZGK and 2FLU, respectively) a
pharmacophore model was therefore generated using the
LigandScout software package[29] and used to identify features
for potential small-molecule inhibitors of the Keap1−Nrf2 PPI.
However, due to the large size of the Keap1-Nrf2 interface, the
resulting model was found to be very complex to fit the size of
small molecules. Thus, to adjust the model, we omitted two
negative and one hydrophobic feature, based on visual
inspection. The resulting pharmacophore presented in Figure 2
contains three negatively charged features, one hydrogen bond
donor feature, a hydrophobic center, and 15 excluded volumes
reflected potential steric restriction and corresponded to the
Although the crystal structures of full Nrf2 or full Nrf2-Keap1
complex is not available yet, the crystallographic structures of
the Keap1 Kelch domain in the presence of short peptides
containing conserved ETGE and DLG motifs located in the Neh2
domain of Nrf2 have been solved.[21]
Despite the difference in the binding affinity, both the ETGE
and DLG peptides display electrostatic interaction with Arg415
and Arg483 of Keap1, indicating the significant contribution of
these residues to Nrf2 recognition. Therefore, both interaction
sites were used as targets for Nrf2 activator design.[22-26] We
hypothesized that binding of a small molecule to Keap1 instead
of Nrf2 will prevent the binding of Keap1 to Nrf2 and therefore
increase the nuclear concentration of the latter. The first inhibitor
of the Nrf2–Keap1 protein−protein interaction (PPI) was
developed by Wells group at 2012.[27] Almost all known
protein−protein interaction (PPI) inhibitors of the Nrf2–Keap1
complex have a negatively charged group (Figure 1). In most
cases, a carboxylic acid moiety or its isosteres have been
2
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
positions that are sterically claimed by the macromolecular
environment surrounding the 16-mer-Nrf2 ETGE peptide.
The potential cytoprotective effect of the test compounds on
the cell viability following
a 24-h incubation period was
investigated. Oxidative stress was induced by H2O2 at a
concentration of 150 µM. Compounds and H2O2 were added
simultaneously. Figure 5 shows that three out of the five tested
compounds (MCULEPO1277, SKT656274 and SKT359126)
were able to prevent the cytotoxic effect of H2O2 in the PC-12
cells. Cell viability increased by 40-50% for compounds
MCULEPO1277, SKT656274 and SKT359126.
This pharmacophore model was applied to virtual screening
using the ZINC database.[30] A total of 7,377,031 (which are in
stock) compounds were retrieved and processed using the
Screen Library protocol, as implemented in LigandScout.[29]
Based on this protocol, 25 conformations were generated for
each compound. These conformations were subsequently fitted
into the pharmacophore model using the LigandScout default
parameters. Only 509 hits were obtained based on the features
defined as essential (negative charge features), which constitute
the basis for salt bridge formation. Five compounds with the
highest binding score were selected for purchase (Figure 3).
To prove that the cytoprotective effect of the compounds
mediated through Nrf2 activation, a Dual-Luciferase Reporter
Assay was conducted.[35] This assay is based on measuring the
intracellular signal from two individual reporter enzymes within a
single system. In the assay, the activities of the firefly and renilla
luciferases are measured sequentially. Using this method, the
direct activation of the promoter in PC-12 cells, which is usually
activated by Nrf2, can be tested. As shown in Figure 6, only one
out of three test compounds (SKT359126) significantly activates
the Nrf2 promoter at the low pharmacological concentration of
10 µM. These results were further confirmed by the positive
effect of the known Nrf2 activator tert-butylhydroquinone
(tBHQ).[36]
As discussed earlier, Keap1−Nrf2 ETGE interacts via the
formation of an electrostatic interactions network between
conserved arginine (Arg380, Arg415 and Arg483) residues of
Keap1 and acidic residues of the Nrf2 ETGE motif. In order to
examine the ability of these five compounds to fit into the Keap1
binding site and to form these crucial interactions, they were first
energy minimized and then flexibly docked into an energy-
minimized Keap1 structure using Glide. The binding scores of
the five top compounds are presented in S.Table 1 in the
Supplemental Information.
Interestingly, the three-cyclic molecule SKT359126 was
found to be active, whereas compound LTOO724252, in which
two rings were replaced by a linear moiety, was found to be
inactive. This fact emphasizes the importance of molecular
rigidity for the biological activity. We hypothesized that the
interaction between the negative charge of the carboxylic acids
(SKT359126) and the positively charged arginines in the active
center of Keap1 are stabilized by the rigid structure of the
molecule. Thus, only the derivatives of compound SKT359126
were designed and synthesized.
The docking site was chosen according to the position of the
peptide ligand cocrystallized at the binding site of Keap1 (PDB
code: 2FLU). The results are summarized in Figure 4 and show
that all five compounds were able to dock into the site by
forming interactions with the conserved arginine binding site
residues, similar to those formed by the Nrf2 ETGE motif,
suggesting that they may potentially inhibit the Keap1–Nrf2
interaction.
In total, 23 novel derivatives of compound SKT359126 were
designed, based on two parameters: 1) the existence of
negatively charged moieties in the distal part of the molecules
and 2) the structure’s similarity to the parent compound.
SKT359126 is commercially available and its synthesis was
described by Paytash et al. in 1950. Based on this study, we
synthesized nine of its amide pyrrolidine derivatives, as outlined
in Scheme 1.[37] We did not applied a chiral control in the
synthesis and also chiral resolution was not used. Thus, all
compounds were a mixture of 4 diastereomers. The cyclization
These five molecules were purchased and biologically
evaluated in PC-12 cells. This cell line derives from the neural
crest of the pheochromocytoma of rat adrenal medulla of
embryonic origin, which has a mixture of neuroblastic and
eosinophilic cells.[31] Thus, these cells could easily differentiate
into neuron-like cells even though they are not considered adult
neurons. As a result, this model is used for screening the
possible neuron-related biological effects of test compounds.[32-
34]
In our case, we used them as a general model for cellular
oxidative stress.
of p-phenylenediamine with itaconic acid resulted in
a
corresponding dicarboxylic-acid derivative.[37] Refluxing of
SKT359126 with appropriate amines in dry DMF in the presence
3
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
of carbonyldiimidazole (CDI) yielded its novel amide derivatives,
namely, 1,1'-(1,4-phenylene)bis(5-oxo-N-(thiazol-2-yl)pyrrolidine-
3-carboxamide) 1, 1,1'-(1,4-phenylene)bis(N-(5-methyl-1,3,4-
thiadiazol-2-yl)-5-oxopyrrolidine-3-carboxamide) 2, 1,1'-(1,4-
phenylene)bis(N-(3-(1H-imidazol-1-yl)propyl)-5-oxopyrrolidine-3-
carboxamide) 3, 1-(4-(4-((3-(1H-imidazol-1-yl)propyl)carbamoyl)-
2-oxopyrrolidin-1-yl)phenyl)-5-oxopyrrolidine-3-carboxylic acid 4,
(3-hydroxybenzyl)-1-(4-(4-((4-hydroxybenzyl)carbamoyl)-2-
oxopyrrolidin-1-yl)phenyl)-5-oxopyrrolidine-3-carboxamide 5, 1-
(4-(4-((4-ethoxyphenyl)carbamoyl)-2-oxopyrrolidin-1-yl)phenyl)-
5-oxopyrrolidine-3-carboxylic acid 6, 1,1'-(1,4-phenylene)bis(4-
The imine derivatives were synthesized as shown in Scheme 3.
The first step of the synthesis was conducted in the same
manner as described for the synthesis of amide pyrrolidine
derivatives (Scheme 1). However, in this case, only the
synthesis of the monosubstituted pyrrolidine molecule 12 was
sufficient for the second substitution, and not the double
substituted molecule (SKT359126). A new purification procedure
was developed to obtain 12. Briefly, the equivalent ratio between
the itaconic acid and phenyldiamine was different from the ratio
used for the synthesis of SKT359126. A ratio of 1:0.9 instead of
1:2 was used for the preparation of 12. The compound was
purified using several extraction and purification steps. First,
using water at basic pH, unreacted itaconic acid and 12 were
separated from the leftovers of the phenyldiamine. In the second
step, by adding water at acidic pH, the itaconic acid and 12 were
separated. Finally, using preparative HPLC, 12 was purified. The
(morpholine-4-carbonyl)pyrrolidin-2-one)
phenylene)bis(5-oxopyrrolidine-1,3-diyl-3-
7,
2,2'-((1,1'-(1,4-
carbonyl))bis(azanediyl))bis(ethane-1-sulfonic acid) 8 and 5-oxo-
1-(4-(2-oxo-4-(pyridin-3-ylcarbamoyl)pyrrolidin-1-
yl)phenyl)pyrrolidine-3-carboxylic acid 9. All these compounds
are more extended than the parent compound. However, the
binding site of Keap1 is known to be large enough to allow for
proper interaction with the compounds.[38] Out of all the
synthesized molecules, compounds 4, 6 and 9 were obtained as
monosubstituted ones, and were served as control compounds
for testing the critical role of two negative charges/electron rich
areas on the molecular skeleton. It is important to mention that,
even without the second substitution, the molecule still has a
negative charge, although at a shorter distance from the first
negative moiety. In addition, we completely replaced the
negatively charged and electron rich moieties by potentially two
positively charged (in physiological conditions) domains in
compound 3. Finally, in compound 7 both negative charges
were eliminated and replaced by a morpholine moiety. Both
compounds: 3 and 7 were used as important controls for
structure-activity relationships studies.
synthesis of 12 was also reported by Kolobov et al. and
[39-41]
Rutkauskas et al.,
however they used a completely
different synthetic strategy and purification methods with only
54% yields.
Compound 12 was used as a starting material for the
synthesis of 13–20. The appropriate aldehydes were coupled
with 12 to obtain eight mixture of enantiomers of imine
derivatives:
oxopyrrolidine-3-carboxylic
carboxybenzylidene)amino)phenyl)-5-oxopyrrolidine-3-carboxylic
acid 14, 4-(((4-(4-(methoxycarbonyl)-2-oxopyrrolidin-1
yl)phenyl)imino)methyl)benzoic acid 15, 1-(4-((3,4-
1-(4-((2,4-dihydroxybenzylidene)amino)phenyl)-5-
acid 13, 1-(4-((4-
dihydroxybenzylidene)amino)phenyl)-5-oxopyrrolidine-3-
carboxylic acid 16 (only for this compound 2D NMR (NOESY)
was conducted and the imine double bond configuration was
determined as E), 1-(4-((3-hydroxybenzylidene)amino)phenyl)-5-
An additional diamide compound 10 with linear and aromatic
substitutions in p-phenylenediamine was synthesised: (2-((4-(3-
carboxyacrylamido)phenyl)carbamoyl)benzoic acid 10. Briefly, p-
phenylenediamine was reacted with phthalimide to afford the
known intermediate 2-((4-aminophenyl)carbamoyl)benzoic acid.
The obtained compound was reacted with furan-2-5-dione,
resulting in 10. Finally, two 2,4-dihydroxybenzoic acid were
conjugated with p-phenylenediamine via amide bond formation,
to obtain the symmetric compound ((1,4-phenylene)bis(2,4-
dihydroxybenzamide) 11 (Scheme 2).
oxopyrrolidine-3-carboxylic
acid
17,
1-(4-((2-
hydroxybenzylidene)amino)phenyl)-5-oxopyrrolidine-3-carboxylic
acid 18 and 1-(4-((3-carboxybenzylidene)amino)phenyl)-5-
oxopyrrolidine-3-carboxylic acid 19.
In addition, compound 12 was converted to the
corresponding methyl ester to obtain methyl 1-(4-aminophenyl)-
5-oxopyrrolidine-3-carboxylate, which was coupled with 4-
carboxy benzaldehyde to yield compound 15.
Two compounds (14 and 19) out of eight were prepared in
the presence of a carboxylic acid moiety in both distal parts of
the molecule. The rest of the imine analogs of SKT359126 were
phenolic or polyphenolic molecules.
Moreover, in order to prevent free rotation and the formation
of more rigid compounds, a double bond (imine) was introduced
into the backbone of intermediate: 1-(4-aminophenyl)-5-
oxopyrrolidine-3-carboxylic acid 12 to generate compounds 13-
20. This also decreases the length of the prepared compounds.
Compound 15 was prepared to investigate the role of the
negative charge in the pyrrolidine domain. If the presence of
such a moiety is critical for the biological activity, the elimination
4
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
of the negative charge in the pyrimidine part of the molecule is
effective concentration of 16 depends on the time of incubation.
After only 6 h, the maximal effective concentration of 16 was just
10 µM. The longer incubation time (24 h) only moderately
increased the effect of 16 on the PC-12 viability.
expected to reduce activity.
The possible cytoprotective
effect of all synthesized compounds was investigated on PC-12
cell viability. The ROS generation method was changed.
Glucose oxidase (GO) was used this time as an inducer of the
oxidative stress instead of the direct supply of H2O2.[42] We
preferred to use a more tuneable and predictable system to
generate H2O2 than its direct addition to the medium at a high
dose. GO is a well-characterized ROS-generating system that
continuously produces constant concentrations of H2O2 and
O2.[43] GO catalyses two electron reduction of oxygen to H2O2,
using reducing equivalents from the oxidation of glucose to
glycolic acid. Glucose is present in the medium at a high
concentration. Therefore, the system never runs out of substrate
for the enzymatic reaction. We also changed the order of
addition of test compounds and ROS generation. In the first set
of screening, H2O2 and compounds were added to the cells
simultaneously, but in the second set of the experiments,
compounds were suppled first to allow them to reach Nrf2-
Keap1 complex and after then ROS were added to the
experimental system. In this case, cells were already protected
against oxidative stress, even before ROS attacked them. Cells
were incubated for 2.5 h with the test compounds (1 and 10 µM)
under physiological conditions and then further 1.5 h incubation
under the oxidative stress environment. Cell survival was then
determined by the MTT assay. Figure 7 shows that compound
16 exhibited a moderate but significant cell-protective effect at a
concentration of 10 µM (17%), as compared to the control cells,
treated by a the vehicle (containing 0.1%DMSO).
An additional control experiment was conducted to determine
whether the lead compound has a positive effect on cell
proliferation. Such positive effect on the number of live cells in
antioxidant experiments may mimic the increased cell viability,
which we attributed to the antioxidant activity of compound 16. A
proliferation assay (following a 24-h exposure to 16 at a
concentration of 50 µM) was performed with PC-12 cells using
the MTT test (Figure 8C). These results proved that 16 has no
significant effect on the proliferation rate of the cells and
demonstrate that the observed antioxidant cellular effect of 16 is
only related to its antioxidative properties.
Next, we evaluated the role of various moieties of 16 on its
activity. First, the importance of the pyrrolidine ring and the
catechol moiety for the biological activity was investigated.
Compounds
20-22
,
(hydroxybenzylidene)amino)phenyl)carbamoyl)benzoic acid 20,
2-((4-((4-carboxybenzylidene)amino)phenyl)carbamoyl)benzoic
acid
21
and
1-(4-((4-cyanobenzylidene)amino)phenyl)-5-
oxopyrrolidine-3-carboxylic acid 22, were synthesised by
replacing the pyrrolidine ring in 16 by 2-(carbamoyl)benzoic acid,
and the catechol moiety by either p-phenol in 20, p-carboxylic
acid in 21 or p-nitrile moiety in 22. These three catechol
replacements were chosen because of the similarity in terms of
the partially negative charge (δ-) on the catechol and nitrile or
phenol groups. In the case of carboxylic acid, the molecule is
negatively charged at the physiological pH due to anion
formation. All three compounds were synthesized from 2-((4-
aminophenyl)carbamoyl)benzoic acid, as shown in Scheme 4.
Four-hydroxybenzaldehyde, 4-formylbenzoic acid, or 4-
To verify that the effect of 16 is indeed pharmacologically
relevant, dose- and time-dependent curves of its activity were
generated. Figure 8A shows that 16 dose-dependently protects
PC-12 cells under oxidative stress conditions. The minimal
effective concentration was determined as 10 µM, as we already
concluded from the screening experiments. The maximal
effective concentration was 100 µM, and 40% cell protective
effect was achieved. Two additional time points of the exposure
to the compound 16 (6 h and 24 h) were added to the 4-h
interval of cell incubation with H2O2, which was used in the
screening experiments (Figure 8B). At this point, we were forced
to replace the GO system, which we used for the 4-h incubation,
by H2O2 (150 µM), due to massive cell death during longer
incubation periods with GO. The effect of 16 at the lowest
concentration (10 µM) at both time points was much larger than
that at the same concentration after only 4 h (50–55% versus
17%). An identical cytoprotective effect (50–55%) was observed
using 50 µM of 16. Thus, we concluded that the maximal
formylbenzonitrile
were
reacted
with
2-((4-
aminophenyl)carbamoyl)benzoic acid, respectively, to obtain the
resulting compounds 20, 21 and 22 (Scheme 4A).
Finally, compound 23 was synthesized from itaconic acid
and commercially available 5-amino-2-(carboxymethyl)benzoic
acid, to obtain a mixture of enantiomers as shown in Scheme 4B.
By synthesizing and testing this molecule, we planned to
determine the role of catechol moiety in the biological activity of
16, by replacing the two phenolic hydroxyls by two carboxylic
acids and eliminating the aromatic ring.
All four compounds were screened in PC-12 cells at two
concentrations, 10 and 50 µM and their activity was compared to
that of compound 16 (Figure 9). At 10 µM, all four compounds
(20-23) failed to show significant protecting effect on the cell
5
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
viability, although compound 16 significantly protected PC-12
cells against oxidative stress at 10 M. At a concentration of 50
µM, however, compound 20 (the molecule in which the catechol
moderate, but significant reduction of DCF signal starting from
10 µM concentration (Figure 10B).
Finally, the most important evidence showing that 16 induces
an antioxidant effect via activating the Nrf2 pathway was
obtained by examining the direct effect of the test compound on
the rate of Nrf2 nuclear translocation, using commercially
available kits as described in the Experimental part.
Physiological activation of Nrf2 by ROS lead to translocation of
the former from the cytosol to nuclei, where Nrf2 is activate
transcription of antioxidant genes. Thus, increased amount of
Nrf2 in PC-12 cellular nuclear fraction after expose the cells to
16 might be an indication for the induced by the compounds Nrf2
dissociation from Keap1. The nuclear fraction of treated by 16
and control cells were isolated and the activation of ARE was
investigated by testing amount of Nrf2 in nuclear fractions by
Nrf2-ARE binding kit as described in the Experimental part. This
ELISA kit is measuring binding activity of Nrf2 to ARE
immobilized sequence using sensitive colorimetric readout. The
intensity of the signal is directly correlated with the amount of
Nrf2 which interacted with ARE Nrf2. The amount of Nrf2 is
depends on the translocation level of this protein to nuclei.
moiety was replaced by
a
phenol group and the
carboxypyrollidine ring was changed to a carboxybenzyl cycle)
exhibited significant protecting activity although its activity was
considerably lower than that of compound 16 (17% versus 40%).
Based on these results, we concluded that a tri-cyclic backbone,
including an oxopyrrolidine ring substituted by carboxylic acid,
phenylmethanimine and catechol moieties, are critical for the
biological activity. The carboxylic acid and catechol are most
likely responsible for the interaction with the Arg415 and Arg483
residues of Keap1. Remarkably, non-active compounds such as
11 and 21 do not have the oxopyrrolidine ring. In addition, the
non-active compound 10 has an ester substitution on the
oxopyrrolidine ring instead of free carboxylic acid. This
difference may prevent the formation of salt bridges between the
free carboxylic acid with arginine residues in the binding site of
Keap1. Compound 19 has a phenol ring, in which the hydroxyl
group is located at the ortho position relative to the iminic bond.
This change may be crucial for the activity, because the distance
between the hydroxyl group and the arginine residues of Keap1
is too short for the binding interaction to take place.
Compound 16 significantly increased the translocation of Nrf2
into the nucleus after 2 h of incubation (Figure 11A) without
inducing the oxidative stress. Known Nrf2 inducer (Al-1) was
used as a positive control.[47] These results indirectly support our
hypothesis that compound 16 can interrupt Nrf2 interaction with
Keap1 and increase its release from the complex without
involving the main trigger of this process, oxidative stress.
Interestingly, when cells pre-treated with 16 were exposed to
GO, the compound also exhibited a stimulatory effect on the
Nrf2 translocation (Figure 11B). It is worth mentioning that 16 is
active in the absence of GO and that its effect is additive to that
of the natural activator of the Nrf2 released from the Keap1 –
ROS, when incubated together, suggesting that the mechanisms
of Nrf2 activation by 16 and by ROS might be distinct. In another
hand, this data does not necessarily suggest that the mode of
action of ROS and 16 are separate - it is a possibility that they
have additive effects acting via the identical mechanism
(assuming neither 16 saturates the ROS activating mechanism).
ROS causes the oxidation of thiol groups in Keap1, and this
effect is the main trigger for the Nrf2 release and nuclear
translocation.[12] Compound 16 was designed as an inhibitor of
the Kelch domain binding site, which is responsible for Nrf2
binding. Such inhibition should lead to an increased non-binding
fraction of Nrf2. Although, we did not prove the proposed
mechanism in the molecular level, the activity of 16 in the
absence of ROS may support our hypothesis that compound 16
The catechol moiety is part of the structure of compound 16,
which is known for its direct antioxidant activity.[44] Thus, to
evaluate whether the antioxidant activity of 16 is related to the
activation of some intracellular mechanisms (we assumed that
this is the Nrf2 activation) or direct quenching of ROS, the
antioxidant activity of 16 was measured by the 2’,7’-
dichlorofluorescin (DCFH) method. In this method, the
nonfluorescent DCFH is oxidized by a ROS, such as H2O2, to
form fluorescent 2,7-dichlorofluorescein (DCF; excitation 485 nm,
emission 535 nm)[45]. Besides compound 16, four additional
molecules were used as a control, including two monophenols
(18 and 20), one polyphenols (11) and the molecule without the
phenol moiety (15). All five compounds were inactive in the cell-
protection assay. The concentration chosen for the DCFH assay
was relevant to the observed biological effect (10 µM). Figure
10A shows that none of the five compounds reduced the
oxidation of DCFH to DCF, indicating that the antioxidant effect
of 16 in PC-12, is not related to their direct ability to quench
ROS as could have been expected based on the presence of
the phenolic or catechol moieties in four out of five molecules.
Moreover, at the tested concentration, all five compounds with
phenol moieties including 16, which was active in the cellular
assay showed no significant direct scavenger activity. In contrast
known polyphenolic free radical scavenger tBHQ[46] showed
6
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
mediates its biological activity by direct stimulation of the Nrf2
pathway. In one hand, many Nrf2 enhancers are electrophilic
(including polyphenols) that directly, or after metabolic activation
bind to thiol residues in Keap1. Compound 16 is catechol and
might be oxidized in the physiological environment. After such
change, the molecule is able to induce Nrf2 activity by binding to
electrophile response element.[48, 49] In another hand, two other
polyphenols (compounds 11 and 13) have not shown significant
biological effect, despite the presence of the identical
experimental conditions. Thus, based on this data, we at least
partially might conclude that compound 16 activates Nrf2 via
non-dependent to thiols interactions.
Figure 3. Virtual screening results. (A) Structures of the top five
compounds obtained from the virtual screening. (B) Fitting of the
compounds to the pharmacophore model, 2D Binding mode
interactions. Features are color-coded, as described in Figure
2C.
Figure 4. Docking of the five selected compounds at the Keap1
binding site and their 2D ligand interaction diagram. Residues
indicated by pink- and green-colored spheres represent
electrostatic and van der Waals interactions, respectively.
Hydrogen bonds are indicated by dashed lines.
Figure 5. Screening of the five compounds with the highest
fitting score for their cytoprotecting activity. PC-12 cells were
grown as described in the Experimental Section. The cells were
exposed to 150 µM H2O2 and either to DMSO (C, 0.1%) or to the
test compounds SKT716764 (SKT7), MCULEP01277 (1277),
LTOO724252 (LTOO), SKT656274 (SKT6) and SKT359126
(SKT3). All compounds were added at a concentration of 10 µM..
After that, a standard MTT analysis was conducted after 24 h. *p
≤ 0.05, n = 3, mean±SE.
In addition, docking simulations suggest that compound 16
interacts with the arginine-rich area in the Kelch domain of
Keap1 in the same mode as the Nrf2–ETGT motif (Figure 12).
Analysis of the binding mode of 16 reveals that the interactions
are hydrophobic and electrostatic. There are two π–π stacking
interactions between the compound and protein sidechains:
Tyr334 and Tyr572. Moreover, 16 forms several H-bonds with
the side chains Arg415, Arg483, Ser508, Ser555 and Tyr334.
Thus, positive correlation between in silico modeling, the
cell-protective effect and the subsequent nuclear translocation of
Nrf2 caused by compound 16 indicates that its molecular
mechanism is indeed through Nrf2 release.
Figure 6. Stimulatory effect of SKT3 on the Nrf2 promoter.
Three active compounds in the cell-protection assay: SKT3,
SKT6 and 1277 and DMSO (C) were introduced into PC-12 cells
at a concentration of 10 µM for 24 h. tBHQ (15 µM) was used as
a positive control. A double luciferase method was applied, as
described in the Experimental Section. *p ≤ 0.05, n = 3,
mean±SE.
Figure 7. Effect of synthesized compounds on PC-12 viability.
PC-12 cells were grown as described in the Experimental
Section. The cells were incubated with the test compounds for
2.5 h at concentrations of 10 µM (black columns) and 1 µM (gray
columns). After that, GO (250 mUnits per ml) was added for an
additional 1.5-h incubation period. The control cells were treated
by DMSO [red color columns (0.1%)]. A standard MTT analysis
was conducted. *p ≤ 0.05, mean±SE (n = 3).
Compound 16 may provide
a structural basis for the
development of more potent and effective Keap1–Nrf2
interaction inhibitors. It is important to mention that mimicking
the ETGT–Nrf2 domain by a small organic molecule is a
relatively novel approach that has attracted much attention over
the past three years,[50] whereas the traditional approach is to
target the thiol groups in Keap1 by covalent modifications.
Further in vivo work is needed to determine whether compound
16 can be used as a drug candidate in severe human disorders
related to oxidative stress, such as cancer, neurodegenerative
diseases and diabetes.
Figure 8. Pharmacological characterization of the lead
compound. (A) Dose-response analysis of the effect of 16. PC-
12 cells were grown and treated as described in Figure 7. *p ≤
0.05, mean±SE (n = 3). (B) Time-course analysis of the effect of
16. The cells were initially exposed to increasing concentrations
of 16, from 10 to 50 µM, and then exposed to 150 µM H2O2 for 6
and 24 h. Following the exposure, a standard MTT analysis was
conducted. *p ≤ 0.05, mean±SE (n = 3). (C) Lack of effect of 16
on proliferation rate of PC-12. Following exposure to 16 for 24 h
at a concentration of 50 µM, a standard MTT analysis was
conducted. n = 3.
Figure 9. Dose-response analysis of the effect of 16 and its
derivatives on PC-12 viability. The cells were treated as
previously described (Figure 7). They were then exposed to
increasing concentrations of the compounds, 10 µM (red
columns) and 50 µM (blue columns). DMSO was used as a
control (0.1%, black color). After 4 h, MTT analysis was
conducted as described in the Experimental Section. *p ≤ 0.05,
mean±SE (n = 3).
LEGENDS
Figure 1. Structure of known Keap1–Nrf2 PPI inhibitors.
Figure 2. Pharmacophore design. (A) Crystal structures of
Keap1 and Nrf2 ETGE peptide (PDB code: 2FLU). Residues
depicted as sticks show the interaction of the Nrf2 ETGE side
chain with Arg 380, 415 and 483 of Keap1. (B) Structure-based
pharmacophore model generated with LigandScout from the
crystal structure of the Nrf2 ETGE peptide and the Kelch domain
of Keap1. (C) The final pharmacophore which was used for
screening. Features are color-coded as follows: negative charge,
red; hydrogen bond donor, green; hydrophobic center, yellow;
excluded volumes, gray.
Figure 10. Lack of the DCF-scavenging effect of the test
compounds. (A). The test compounds were diluted at a ratio of
1:10 with 40 mM Tris (pH 7.4) to give 10 µM final concentration
and then loaded with 10 µM of DCFH (in methanol) for 15 min at
37 °C. After exposure to a solution containing 30% H2O2, the
formation of the fluorescent-oxidized derivatives of DCF was
monitored in a cuvette holder (thermostatically maintained at
37 °C) as follows: DCF, excitation wavelength (488 nm) and
7
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
emission wavelength (525 nm). DCF formation was quantified
from separate standard curves in methanol (0.05–1 lm). 0.1%
DMSO was added in the control treatment. (B). Scavenging
effect of tBHQ was compared with compound 16 in identical
conditions which were described in panel A. The results are
shown as the mean ± SE. Control cells were treated by DMSO
(0.1%) *p < 0.05.
All chemical reagents, solvents, and acids were purchased from
Acros Organic, Alfa Aesar, Bio-Lab Ltd., Merck, or IU-CHEM Ltd.,
and all were used as received. Analytical and preparative
HPLCs (Young Lin Instruments, Anyang, Korea) were performed
on LUNA C18 preparative (10 µm, 100ª30 mm) or analytical (5
µm, 250ª4.6 mm) columns, both from Phenomenex, Inc.
(Torrance, CA, USA). HPLC purification was carried out with an
increasing linear gradient of CH3CN in H2O. Purity of the
synthesized compounds was confirmed by HPLC analysis.
Analytical TLC was carried out on pre-coated silica gel 60 F254
(Merck) sheets using UV absorption and iodine physical
adsorption for visualization. Mass spectra were recorded on a
Finnigan Model 400 instrument using a QToF microspectrometer
(Micromass, Milford, MA, USA), by electrospray ionization (ESI)
in the positive/negative ion modes. The data were processed
using mass LynX ver. 4.1 calculations and de-convolution
software (Waters Corp., Milford, MA, USA). High-resolution
mass spectra (HRMS) were obtained using an LTQ Orbitrap XL
(Thermo Scientific, Waltham, MA, USA). Melting points were
Figure 11. Compound 16 activated the Nrf2 anti-oxidant system.
(A) PC-12 cells were incubated with 50 µM of 16 or Al-1 during 2
h and then nuclear extract was prepared using a commercially
available nuclear extraction kit. The nuclear translocation of Nrf2
was quantified by the Nrf2 activity kit. (B) PC-12 cells were
initially exposed to 50 µM of 16 for 2 h and then 300 µM of GO
were added for additional 2 h. The induced nuclear translocation
of Nrf2 was quantified by the Nrf2 activity kit. The results are
shown as the mean ± SE. Control cells were treated by DMSO
(0.1%) *p < 0.05.
Figure 12. The binding mode between the active conformation
of compound 16 and Keap1 provided by Glide docking. (A)
Keap1 surface with compound 16. (B) 3D diagram of the
interactions between compound 16 and the Keap1 binding site.
Only interacting residues are displayed. H-bonds are displayed
as dotted green arrows and the π–π stacking interactions as
pink dashed arrows. (C) 2D diagram of the interactions between
16 and the Keap1 binding site. H-bonds are displayed as dotted
green arrows.
measured with
a
Fisher-Johns melting-point apparatus
Scheme 1. Synthesis of nine novel amide pyrrolidine
derivatives: a) Benzene-1,4-diamine, 2-methylenesuccinic acid,
H2O, reflux 1 h. b) 1) CDI, dry DMF, 80 °C 1 h. 2) Reflux, 6–7 h.
(Waltham, MA, USA). Elemental analysis was conducted using
Thermo Flash 2000 CHN-O Elemental Analyzer (Waltham,
Massachusetts,
Phenylmethanesulfonyl fluoride (PMSF), sodium orthovanadate,
sodium-β-glycerophosphate and sodium pyrophosphate
decahydrate were purchased from Alfa Aesar (Karlsruhe,
Germany). Ethylene glycol-bis-N,N,N′,N′-tetraacetic acid
United
States).
β-Mercaptoethanol,
Scheme 2. Synthesis of 10 and 11: a) THF, RT 4 h. b) EtOH,
aromatic aldehyde, RT overnight. c) Dry DMF, CDI, reflux 6 h. d)
H2O, reflux 1 h.
Scheme 3. Synthesis of 13–19: a) H2O, phenylenediamine,
itaconic acid reflux 1 h. b) EtOH, 12, aromatic aldehyde, RT
overnight. c) MeOH, H2SO4, reflux 5 h.
tetrasodium salt (EGTA) was received from AppliChem
(Darmstadt, Germany). Dulbecco’s modified Eagle’s medium
(DMEM), fetal calf serum (FCS), horse serum (HS), L-glutamine,
penicillin/streptomycin and trypsin were purchased from
Biological Industries (Beth-Haemek, Israel). Methylthiazolyl blue
(MT) was purchased from Chem-Impex International (Wood
Dale, IL, USA). DMSO, EGTA, IGEPAL, NaCl, PBS tablets, DCF,
Tris·HCl and protease inhibitor cocktail for mammalian cells
were purchased from Merk (Rehovot, Israel). The Nuclear
Extract Kit and Nrf2-DNA binding activity kit were purchased
from Active Motif (Carlsbad, CA, USA).
Scheme 4. Synthesis of 20–23: a) THF, RT 4 h. b) EtOH,
aromatic aldehyde, RT overnight. c) Dry DMF, CDI, reflux 6 h. d)
H2O, reflux 1 h.
3. Conclusions
Based on the crystal structure of the Keap1–Nrf2 peptide
complex, a pharmacophore model was designed and, using
virtual screening, five initial hits were chosen. Based on one of
them, 1,1'-(1,4-phenylene)bis(5-oxopyrrolidine-3-carboxylic acid
(SKT359126), 23 different novel derivatives were designed,
synthesized and biologically evaluated. Compound 1-(4-((3,4-
dihydroxybenzylidene)amino)phenyl)-5-oxopyrrolidine-3-
carboxylic acid (16) exhibited an in vitro cell protective effect via
the activation of the Nfr2 pathway.
All synthetic procedures and spectroscopic characterization of
described in the manuscript compounds are shown in Supporting
Information.
Molecular modelling
Pharmacophore modelling
Experimental Section
The structure-based pharmacophore model was generated
using the LigandScout software package.[29] LigandScout is able
Materials
8
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
to generate the 3D pharmacophores from structural data of
macromolecule–ligand complexes. The program is taking for the
calculations: H-bond donors and acceptors as directed vectors,
positive and negative ionizable areas, and finally lipophilic
fragments represented by spheres. The LigandScout model is
very selective due to including in the calculations spatial
parameters about potentially inaccessible for any kind of
interactions areas for any ligand. This important detail in the
pharmacophore formation is taking into account any potential
steric restrictions. Excluded volume spheres are also included to
the model based on the coordinates defined by amino acids side
chain atoms to depict the inaccessible areas for any potential
ligand.[51] This pharmacophore model was used as a query in
virtual screening of ZINC database using the Screen Library
protocol, as implemented in LigandScout.
DCF (in methanol) for 15 min at 37 °C. After exposure to a
solution containing 30% H2O2, the formation of the fluorescent-
oxidized derivatives of the DCF was monitored in a cuvette
holder (thermostatically maintained at 37 °C) as follows: DCF,
excitation wavelength (488 nm) and emission wavelength (525
nm). The DCF formation was quantified from separate-standard
curves in methanol (0.05–1 lm).
Luciferase assay
The assay was conducted as described.[59]
Subcellular fractionation
Cellular nuclear fraction was prepared from PC-12 cells using a
Nuclear Extract Kit (Active Motif, Carlsbad, CA, USA). The
protein concentration was determined by the Lowry protein
assay.[60]
Docking simulation
Docking simulations were performed using Glide[52-54] as
implemented in Maestro V9.0.[55] For docking into crystal
structures, Glide’s grid box was centered on the coordinates of
the ligand in the complex. A docking grid was generated within
the docking box and ligands were docked into the binding sites
using Glide’s Extra Precision (XP) option with default
parameters.
Nrf2-DNA binding activity
The activation of Nrf2 was investigated by quantifying the
binding of Nrf2 to ARE using a binding kit (TransAM Nrf2 DNA;
Active Motif) according to the manufacturer’s instructions. Briefly,
nuclear protein (5–20 μg) was incubated for 1 hour in a 96-well
plate coated with ARE sequence oligonucleotide. The bound
Nrf2 was captured by anti-Nrf2 antibody and visualized by
colorimetric reaction using secondary antibody. The resultant
color was measured spectrophotometrically at 450/655 nm.
Cell culture and viability assay
PC-12 cells were grown in DMEM with 10% FCS, 5% horse
serum and 1% (v/v) penicillin/streptomycin/nystatin, maintained
at 37 °C in a humidified 5% CO2 incubator. The cells (10,000
cells/well) were plated in 96-well tissue-culture plates in the
medium (100 mL) and incubated overnight for attachment. The
compounds were added to the cells for either 2.5, 6 or 24 h
before exposure to oxidative stress (GO or H2O2).
Statistical analysis
In vitro results are given as mean±SE. The statistical
significance (p < 0.05) was calculated among the experimental
groups using the two-tailed Student’s t-test. The QuickCalcs
online
service
Graph-Pad
Software,
was
found
used
at
The cell viability was evaluated by an MTT assay.[57] The cells
www.graphpad.com/quickcalcs/ttest1.cfm,
statistical evaluations.
for
were
incubated
with
3-(4,5-
dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide (MTT) (2 mg/ml) in growth medium
for 30 min at 37 °C. The medium was then aspirated, and DMSO
was added to solubilize the colored crystals. The absorbance at
570 nm was measured in an ELISA reader. The amount of color
produced is directly proportional to the number of viable cells.
Acknowledgements
This study was supported by a Bar-Ilan University new faculty
grant for A.G. Bar-Ilan University, Ramat Gan and Rabin
Medical Center, Petah-Tikva, both in Israel, provided a grant for
A.G and D.O. (Collaborative Grant in Biomedical Research –
2013). In addition, S. K. is thankful for the support of her work by
the Wolf Foundation. We would like to thank Nechama-Sara
Cohen for the English editing of the manuscript.
Assay for reactive-oxygen species
Quantification of the formation rate of ROS was determined as
previously described.[58] Briefly, the compounds were diluted at a
ratio of 1:10 with 40 mm Tris (pH 7.4) and loaded with 10 µM of
9
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Keywords: Keap1 • Nrf2 • oxidative stress • oxopyrrolidine
derivatives • computer modelling
[22] H.-P. Sun, Z.-Y. Jiang, M.-Y. Zhang, M.-C. Lu, T.-T. Yang, Y.
Pan, H.-Z. Huang, X.-J. Zhang, Q.-d. You, MedChemComm. 2014, 5,
93-98.
References:
[23] L. Hu, S. Magesh, L. Chen, L. Wang, T. A. Lewis, Y. Chen, C.
Khodier, D. Inoyama, L. J. Beamer, T. J. Emge, J. Shen, J. E.
Kerrigan, A. N. Kong, S. Dandapani, M. Palmer, S. L. Schreiber, B.
Munoz, Bioorg Med Chem Lett. 2013, 23, 3039-3043.
[1] E. Birben, U. M. Sahiner, C. Sackesen, S. Erzurum, O. Kalayci,
World Allergy Organ J. 2012, 5, 9-19.
[2] M. Valko, K. Jomova, C. J. Rhodes, K. Kuca, K. Musilek, Arch
Toxicol. 2016, 90, 1-37.
[24] Z. Y. Jiang, M. C. Lu, L. L. Xu, T. T. Yang, M. Y. Xi, X. L. Xu, X.
K. Guo, X. J. Zhang, Q. D. You, H. P. Sun, J Med Chem. 2014, 57,
2736-2745.
[3] X. Wang, C. Hai, Mol Biol Rep. 2016, 43, 607-628.
[25] E. Jnoff, C. Albrecht, J. J. Barker, O. Barker, E. Beaumont, S.
Bromidge, F. Brookfield, M. Brooks, C. Bubert, T. Ceska, V. Corden,
G. Dawson, S. Duclos, T. Fryatt, C. Genicot, E. Jigorel, J. Kwong, R.
Maghames, I. Mushi, R. Pike, Z. A. Sands, M. A. Smith, C. C.
Stimson, J.-P. Courade, ChemMedChem. 2014, 9, 699-705.
[4] M. Rutkowski, T. Matuszewski, J. Kedziora, M. Paradowski, K.
Klos, A. Zakrzewski, Pol Merkur Lekarski. 2010, 29, 377-381.
[5] O. Blokhina, E. Virolainen, K. V. Fagerstedt, Ann Bot. 2003, 91
Spec No, 179-194.
[6] C. Espinosa-Diez, V. Miguel, D. Mennerich, T. Kietzmann, P.
Sanchez-Perez, S. Cadenas, S. Lamas, Redox Biol. 2015, 6, 183-
197.
[26] D. Marcotte, W. Zeng, J.-C. Hus, A. McKenzie, C. Hession, P.
Jin, C. Bergeron, A. Lugovskoy, I. Enyedy, H. Cuervo, D. Wang, C.
Atmanene, D. Roecklin, M. Vecchi, V. Vivat, J. Kraemer, D. Winkler,
V. Hong, J. Chao, M. Lukashev, L. Silvian, Bioorg Med Chem. 2013,
21, 4011-4019.
[7] T. Nguyen, P. Nioi, C. B. Pickett, J Biol Chem. 2009, 284, 13291-
13295.
[27] R. Hancock, H. C. Bertrand, T. Tsujita, S. Naz, A. El-Bakry, J.
Laoruchupong, J. D. Hayes, G. Wells, Free Radic Biol Med. 2012,
52, 444-451.
[8] L. M. Zipper, R. T. Mulcahy, J Biol Chem. 2002, 277, 36544-
36552.
[9] S. B. Cullinan, J. D. Gordan, J. Jin, J. W. Harper, J. A. Diehl, Mol
Cell Biol. 2004, 24, 8477-8486.
[28] D. D. Zhang, M. Hannink, Mol Cell Biol. 2003, 23, 8137-8151.
[29] G. Wolber, T. Langer, J Chem Inf Model. 2005, 45, 160-169.
[30] B. K. Shoichet, Nature. 2004, 432, 862-865.
[10] M. Furukawa, Y. Xiong, Mol Cell Biol. 2005, 25, 162-171.
[11] A. Kobayashi, M. I. Kang, H. Okawa, M. Ohtsuji, Y. Zenke, T.
Chiba, K. Igarashi, M. Yamamoto, Mol Cell Biol. 2004, 24, 7130-
7139.
[31] L. A. Greene, A. S. Tischler, Proc Natl Acad Sci U S A. 1976, 73,
2424-2428.
[12] E. Kansanen, S. M. Kuosmanen, H. Leinonen, A.-L. Levonen,
[32] A. M. Bokare, A. K. Praveenkumar, M. Bhonde, Y. Nayak, R.
Redox Biol. 2013, 1, 45-49.
Pal, R. Goel, Neurochem Res. 2017.
[13] K. Itoh, J. Mimura, M. Yamamoto, Antioxid Redox Signal. 2010,
13, 1665-1678.
[33] H. Elkon, E. Melamed, D. Offen, Cell Mol Neurobiol. 2001, 21,
771-781.
[14] D. D. Zhang, S. C. Lo, J. V. Cross, D. J. Templeton, M. Hannink,
[34] Y. Zhang, X. G. Gong, Z. Z. Wang, H. M. Sun, Z. Y. Guo, J. H.
Mol Cell Biol. 2004, 24, 10941-10953.
Hu, L. Ma, P. Li, N. H. Chen, Eur J Neurosci. 2016, 43, 1379-1388.
[15] T. W. Kensler, N. Wakabayashi, S. Biswal, Annu Rev
Pharmacol Toxicol. 2007, 47, 89-116.
[35] I. Bobilev, V. Novik, I. Levi, O. Shpilberg, J. Levy, Y. Sharoni, G.
P. Studzinski, M. Danilenko, Cancer Biol Ther. 2011, 11, 317-329.
[16] T. Suzuki, H. Motohashi, M. Yamamoto, Trends Pharmacol Sci.
2013, 34, 340-346.
[36] B. R. Imhoff, J. M. Hansen, Cell Biol Toxicol. 2010, 26, 541-551.
[37] P. L. Paytash, E. Sparrow, J. C. Gathe, J. Am. Chem. Soc. 1950,
72, 1415-1416.
[17] Z. Hochberg, Sci Signal. 2010, 3, pt1.
[18] W. Li, S. Yu, T. Liu, J. H. Kim, V. Blank, H. Li, A. N. Kong,
[38] G. Guntas, S. M. Lewis, K. M. Mulvaney, E. W. Cloer, A.
Tripathy, T. R. Lane, M. B. Major, B. Kuhlman, Protein Eng Des Sel.
2016, 29, 1-9.
Biochim Biophys Acta. 2008, 1783, 1847-1856.
[19] S. K. Niture, J. W. Kaspar, J. Shen, A. K. Jaiswal, Toxicol Appl
Pharmacol. 2010, 244, 37-42.
[39] A. V. Kolobov, S. T. Panfilov, E. R. Kofanov, Bashkirskii
Khimicheskii Zhurnal. 2006, 13, 168-169.
[20] M. Zhang, C. An, Y. Gao, R. K. Leak, J. Chen, F. Zhang, Prog
Neurobiol. 2013, 100, 30-47.
[40] K. Rutkauskas, V. Mickevicius, K. Kantminiene, M. Stasevych,
O. Komarovska-Porokhnyavets, R. Musyanovych, V. Novikov,
Chemija. 2013, 24, 74-80.
[21] P. Canning, F. J. Sorrell, A. N. Bullock, Free Radic Biol Med.
2015, 88, 101-107.
10
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
[41] K. Rutkauskas, E. Jakiene, Z. J. Beresnevicius, Chemine
Technologija. 2003, 68-73.
[42] M. Eckshtain-Levi, R. Lavi, D. S. Yufit, B. Daniel, O. Green, O.
Fleker, M. Richman, S. Rahimipour, A. Gruzman, L. Benisvy, Chem
Commun (Camb). 2016, 52, 2350-2353.
[43] N. Kozlovsky, A. Rudich, R. Potashnik, N. Bashan, Free Radical
Bio Med. 1997, 23, 859-869.
[44] A. Barzegar, Mol Biol Res Commun. 2016, 5, 87-95.
[45] R. Shapira, S. Rudnick, B. Daniel, O. Viskind, V. Aisha, M.
Richman, K. R. Ayasolla, A. Perelman, J. H. Chill, A. Gruzman, S.
Rahimipour, J Med Chem. 2013, 56, 6709-6718.
[46] J. D. Yeo, M. K. Jeong, C. U. Park, J. Lee, Journal of Food
Science. 2010, 75, C258-262.
[47] W. Hur, Z. Sun, T. Jiang, D. E. Mason, E. C. Peters, D. D.
Zhang, H. Luesch, P. G. Schultz, N. S. Gray, Chemistry & biology.
2010, 17, 537-547.
[48] H. Erlank, A. Elmann, R. Kohen, J. Kanner, Free Radic Biol Med.
2011, 51, 2319-2327.
[49] R. Sirota, D. Gibson, R. Kohen, Redox Biol. 2015, 4, 48-59.
[50] B. G. Richardson, A. D. Jain, T. E. Speltz, T. W. Moore, Bioorg
Med Chem Lett. 2015, 25, 2261-2268.
[51] M. Rella, C. A. Rushworth, J. L. Guy, A. J. Turner, T. Langer, R.
M. Jackson, J Chem Inf Model. 2006, 46, 708-716.
[52] R. A. Friesner, R. B. Murphy, M. P. Repasky, L. L. Frye, J. R.
Greenwood, T. A. Halgren, P. C. Sanschagrin, D. T. Mainz, J Med
Chem. 2006, 49, 6177-6196.
[53] T. A. Halgren, R. B. Murphy, R. A. Friesner, H. S. Beard, L. L.
Frye, W. T. Pollard, J. L. Banks, J Med Chem. 2004, 47, 1750-1759.
[54] R. A. Friesner, J. L. Banks, R. B. Murphy, T. A. Halgren, J. J.
Klicic, D. T. Mainz, M. P. Repasky, E. H. Knoll, M. Shelley, J. K.
Perry, D. E. Shaw, P. Francis, P. S. Shenkin, J Med Chem. 2004, 47,
1739-1749.
[55] Schrödinger. 2009.
[56] M. De Rosa, G. Vigliotta, G. Palma, C. Saturnino, A. Soriente,
Molecules. 2015, 20, 22044-22057.
[57] A. Munder, L. L. Israel, S. Kahremany, R. Ben-Shabat-Binyamini,
C. Zhang, M. Kolitz-Domb, O. Viskind, A. Levine, H. Senderowitz, S.
Chessler, J. P. Lellouche, A. Gruzman, ACS Appl Mater Interfaces.
2017, 9, 1189-1206.
[58] C. P. Le Bel, H. Ischiropoulos, S. C. Bondy, Chem Res Toxicol.
1992, 5, 227-231.
[59] Y. Riahi, Y. Sin-Malia, G. Cohen, E. Alpert, A. Gruzman, J.
Eckel, B. Staels, M. Guichardant, S. Sasson, Diabetes. 2010, 59,
808-818.
[60] G. L. Peterson, Anal Biochem. 1977, 83, 346-356.
11
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Figure 1: Structure of known Keap1–Nrf2 PPI inhibitors
12
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Figure 2: Pharmacophore design
13
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Figure 3: Virtual screening results
14
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Figure 4: Docking of the five selected compounds at the Keap1 binding site and their 2D ligand interaction diagram
15
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Figure 5: Screening of the five compounds with the highest fitting score
16
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Figure 6: Stimulatory effect of SKT3 on the Nrf2 promoter
17
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Figure 7: Effect of synthesized compounds on PC-12 viability
18
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Figure 8: Pharmacological characterization of the lead compound
19
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Figure 9: Dose-response analysis of the effect of 16 and its derivatives on PC-12 viability
20
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Figure 10: Lack of the DCF-scavenging effect of the test compounds
A
B
120
120
100
80
60
40
20
0
100
80
60
40
20
0
*
*
25 µM
25
C
11
15
16
18
20
5
10
10 µM
5 µM
21
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Figure 11: Compound 16 activates the Nrf2 anti-oxidant system
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0.6
A
B
*
*
*
0.5
0.4
0.3
0.2
0.1
0
Ca
v
16
c
Al-1
C
GO
16
16
GO
+
22
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Figure 12: The binding mode between the active conformation of compound 16 and Keap1 provided by the Glide docking
23
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Scheme 1: Synthesis of nine novel amide pyrrolidine derivatives
24
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Scheme 2: Synthesis of 10 and 11
25
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Scheme 3: Synthesis of 13–19
26
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Scheme 4: Synthesis of 20–23
27
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700539
ChemPlusChem
FULL PAPER
Entry for the Table of Contents
KEAP1
Nrf2
Antioxidant genes
Using a structure-based virtual screening and the lead optimization, compound 16 (1-(4-((3,4-dihydroxybenzylidene)amino)phenyl)-5-
oxopyrrolidine-3-carboxylic acid) was synthesized. The compound activated the Nrf2 transduction pathway in a dose- and time-dependent
manner and protected P-12 cells against oxidative stress. A 16 can serve as a starting point for the development of therapeutics for the
treatment of oxidative stress-related diseases.
28
This article is protected by copyright. All rights reserved.