Activation of anti-oxidant Nrf2 signaling by enone analogues of curcumin

Article abstract of DOI:10.1016/j.ejmech.2017.11.048

Inflammation and oxidative stress are common in many chronic diseases. Targeting signaling pathways that contribute to these conditions may have therapeutic potential. The transcription factor Nrf2 is a major regulator of phase II detoxification and anti-oxidant genes as well as anti-inflammatory and neuroprotective genes. Nrf2 is widespread in the CNS and is recognized as an important regulator of brain inflammation. The natural product curcumin exhibits numerous biological activities including ability to induce the expression of Nrf2-dependent phase II and anti-oxidant enzymes. Curcumin has been examined in a number of clinical studies with limited success, mainly owing to limited bioavailability and rapid metabolism. Enone analogues of curcumin were examined with an Nrf2 reporter assay to identify Nrf2 activators. Analogues were separated into groups with a 7-carbon dienone spacer, as found in curcumin; a 5-carbon enone spacer with and without a ring; and a 3-carbon enone spacer. Activators of Nrf2 were found in all three groups, many of which were more active than curcumin. Dose-response studies demonstrated that a range of substituents on the aromatic rings of these enones influenced not only the sensitivity to activation, reflected in EC₅₀ values, but also the extent of activation, which suggests that multiple mechanisms are involved in the activation of Nrf2 by these analogues.

Full text of DOI:10.1016/j.ejmech.2017.11.048

Accepted Manuscript
Activation of anti-oxidant Nrf2 signaling by enone analogues of curcumin
Lorraine M. Deck, Lucy A. Hunsaker, Thomas A. Vander Jagt, Lisa J. Whalen, Robert
E. Royer, David L. Vander Jagt
PII:
S0223-5234(17)30951-0
10.1016/j.ejmech.2017.11.048
EJMECH 9923
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 22 August 2017
Revised Date: 31 October 2017
Accepted Date: 18 November 2017
Please cite this article as: L.M. Deck, L.A. Hunsaker, T.A. Vander Jagt, L.J. Whalen, R.E. Royer, D.L.
Vander Jagt, Activation of anti-oxidant Nrf2 signaling by enone analogues of curcumin, European
Journal of Medicinal Chemistry (2017), doi: 10.1016/j.ejmech.2017.11.048.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Activation of Anti-Oxidant Nrf2 Signaling by Enone
Analogues of Curcumin
Lorraine M. Deck,a,* Lucy A. Hunsaker,b Thomas A. Vander Jagt,b,c Lisa J. Whalen,a Robert E.
Royer,b David L. Vander Jagt b
aDepartment of Chemistry and Chemical Biology, University of New Mexico, MSC03 2060, 1
University of New Mexico, Albuquerque, NM 87131, USA
bDepartment of Biochemistry and Molecular Biology, University of New Mexico School of
Medicine, MSC08 4670, Fitz Hall, Room 249, 1 University of New Mexico, Albuquerque, NM
87131, USA
cPresent address: Department of Radiology, Oregon Health and Science University, Portland,
OR 97239, USA
*Corresponding author. Email address: ldeck@unm.edu (L.M. Deck)
Abstract
Inflammation and oxidative stress are common in many chronic diseases. Targeting signaling
pathways that contribute to these conditions may have therapeutic potential. The transcription
factor Nrf2 is a major regulator of phase II detoxification and anti-oxidant genes as well as anti-
1

ACCEPTED MANUSCRIPT
inflammatory and neuroprotective genes. Nrf2 is widespread in the CNS and is recognized as an
important regulator of brain inflammation. The natural product curcumin exhibits numerous
biological activities including ability to induce the expression of Nrf2-dependent phase II and
anti-oxidant enzymes. Curcumin has been examined in a number of clinical studies with limited
success, mainly owing to limited bioavailability and rapid metabolism. Enone analogues of
curcumin were examined with an Nrf2 reporter assay to identify Nrf2 activators. Analogues were
separated into groups with a 7-carbon dienone spacer, as found in curcumin; a 5-carbon enone
spacer with and without a ring; and a 3-carbon enone spacer. Activators of Nrf2 were found in all
three groups, many of which were more active than curcumin. Dose-response studies
demonstrated that a range of substituents on the aromatic rings of these enones influenced not
only the sensitivity to activation, reflected in EC50 values, but also the extent of activation, which
suggests that multiple mechanisms are involved in the activation of Nrf2 by these analogues.
Keywords
Nrf2 activation; pyridone; enone; aldol condensation
Abbreviations Used
Nrf2, nuclear factor erythroid 2 related factor 2; CNS, central nervous system; Keap1, Kelch-
like ECH associated protein 1; NF-κB, nuclear factor kappa-light-chain enhancer of activated B
cells; Maf, musculoaponeurotic fibrosarcoma oncogene; XRE, xenobiotic response element;
ARE, anti-oxidant response element; mTORC1, mammalian target of rapamycin complex 1;
GSK-3, glycogen synthase kinase 3; PKB, protein kinase B; ROS, reactive oxygen species
2

ACCEPTED MANUSCRIPT
1. Introduction
Transcription factor nuclear factor erythroid 2 related factor 2 (Nrf2), which is a member of the
cap’n’collar family of transcription factors, is the master regulator of an inducible cellular system
of cytoprotective genes. These genes code for a broad range of proteins, including phase I and II
detoxification enzymes, anti-oxidant proteins, as well as anti-inflammatory and neuroprotective
factors, growth factors and receptors, and other transcription factors. Interest has emerged in
Nrf2 as a therapeutic target, especially for treatment of chronic inflammatory diseases. [1-6] In
the absence of stress, Nrf2 forms a cytosolic complex with Kelch-like ECH associated protein 1
(Keap1) and Cul3; Cul3 is an adaptor to link Nrf2 to an E3 ubiquitin ligase complex. Newly
synthesized Nrf2 in unstressed cells is degraded by ubiquitination and proteosomal degradation,
which limits the cytosolic concentration of Nrf2. In response to oxidative and electrophilic
stresses, cysteine residues of Keap1 are modified, which alters the interaction between Nrf2 and
Keap1 and unlocks Keap1 from the Nrf2 complex. This allows newly synthesized Nrf2 to
accumulate and translocate to the nucleus where Nrf2 interacts with small Maf proteins and then
binds to promoters with anti-oxidant response element (ARE) sequences. [7-11]
Numerous Nrf2-activating chemicals have been identified, including natural products such as
curcumin (Figure 1, Structure 1). Curcumin and some other natural product phenols can activate
Nrf2 after oxidation to electrophilic quinones, which can modify select cysteine residues in
Keap1 by Michael addition. Curcumin may also modify Keap1 by Michael addition involving
the enone functionalities of curcumin. Keap1 cysteine residues 273, 288 and 151 appear to be
especially important. [12,13] Peptide inhibitors of the Keap1-Nrf2 protein-protein interaction
have been developed as well as a variety of small molecules that inhibit the Keap1-Nrf2
3

ACCEPTED MANUSCRIPT
interaction. [14-19] These studies have been aided by the availability of several crystal
structures of Nrf2. [20-22]
Curcumin, obtained from the rhizomes of Curcuma longa (turmeric), exhibits numerous
biological activities [23] including ability to induce the expression of Nrf2-dependent phase II
and anti-oxidant enzymes such as glutathione S-transferase, aldose reductase and heme
oxygenase-1. [24,25] Curcumin appears to utilize more than a single mechanism for activation
of Nrf2, including covalent modification of Keap1 [25,26] and activation of upstream kinases.
[25,27] Curcumin has been examined in a number of clinical studies with limited success, [28]
mainly owing to limited bioavailability and rapid metabolism. Attempts to improve curcumin as
a therapeutic agent include development of new formulations that may enhance bioavailability.
There is considerable interest in the development of analogues and derivatives of curcumin
with improved therapeutic potential. [29-32] There also is interest in the development of
analogues that activate anti-oxidant Nrf2 but simultaneously inhibit pro-inflammatory NF-κB
signaling, [33-35] which is consistent with the ability of curcumin to target both of these
pathways. [36,37] Here we describe the synthesis and evaluation of the Nrf2-activating potential
of enone analogues of curcumin.
2. Results
2.1 Synthesis of curcumin analogues
Curcumin (Scheme 1, Structure 1) is a symmetrical bis-phenolic compound that exists as an
equilibrium mixture of the dienone and the keto-enol tautomer; the keto-enol tautomer
predominates and is stabilized by intramolecular hydrogen bonding. [38-40] Curcumin has a 7-
carbon unsaturated spacer separating two substituted aromatic rings. The current study involves
4

ACCEPTED MANUSCRIPT
analogues of curcumin having a 7-carbon spacer, a 5-carbon spacer, a 5-carbon ring-containing
spacer and a 3-carbon spacer. We have previously reported the synthesis and characterization of
several of these compounds. [41-43]
Scheme 1 describes the synthesis of symmetrical analogues, which maintain a 7-carbon keto-
enol spacer between aromatic rings as in curcumin. The aryl rings contain different substituents
in various positions on the ring. These analogues were designed to test the importance of the type
of functional group and its location on the aromatic ring. Compounds 1-17 were synthesized
from aromatic aldehydes by condensation with 2,4-pentanedione in an aldol type reaction as
described by Pabon. [44] This involves base-catalyzed condensation in the presence of a
trialkylborate to complex with the carbonyl groups of 2,4-pentanedione to protect C-3 from a
Knoevenagel condensation, allowing the reaction to occur only on the terminal methyl groups.
Compounds 1-17 exist predominately in the keto-enol form (99-100%) as exhibited by the
spectral data. [39,40] The proton NMR spectra have a tall peak at 5.47 ppm corresponding to the
enol proton and a short peak at 3.48 ppm for the two central methylene protons of the diketo
form when it is present. There are also doublets in the aromatic region with J values of 15.5-16.5
Hz for alkene protons present in the spacer, evidence that supports the E configuration. [39,40]
The carbon spectra have a large signal at approximately 100 ppm for the enol carbon and a very
small signal at approximately 50 ppm for the central methylene carbon in the diketo form.
5

ACCEPTED MANUSCRIPT
Scheme 1
Scheme 1. Reagents for 7-carbon spacer compounds: (a) CH3C(O)CH2C(O)CH3 (for compounds
1-17) or CH3C(O)CH(CH3)C(O)CH3 (for compound 21) or CH3C(O)CH(CH2C6H5)C(O)CH3
(for compound 18) (0.5 equiv), B2O3, B(OC4H9)3, EtOAc, nBuNH2, HCl/H2O, 20-85%; (b)
C6H5CH2Br (1.0 equiv.), NH4Cl, NaOH, CH2Cl2, 61%; (c) H2, Pd/C, EtOAc, 80-90%.
Aldol condensation of benzaldehyde with 3-methyl-2,4-pentanedione or 3-benzyl-2,4-
pentanedione in the presence of a trialkylborate afforded compounds 18 and 21 respectively, [44]
which have a methyl and a benzyl group on the central methylene carbon spacer. These
compounds were designed to test the importance of the central methylene hydrogens. Use of
monosubstituted pentanediones avoided disubstitution of the methylene carbon on the spacer
which was observed when methyl iodide or benzyl bromide and base were used in an SN2
reaction on 1,7-diphenyl-1,6-heptadiene-3,5-dione. Pedersen reported monosubstitution when
reacting an alkyl halide with curcumin in the presence of a base but in our hands we observed
disubstitution. [40] When a single substituent such as a methyl or benzyl group is placed on the
6

ACCEPTED MANUSCRIPT
central methylene carbon, a pattern in the proton and carbon NMR spectra is observed that is
similar to analogues without the central group. Reaction of 18 with benzyl bromide in the
presence of base gave the dibenzylated compound 23, which exhibits no tautomerization because
the molecule is trapped in the diketo form. [40]
The saturated compounds, 19, 20, and 22 were synthesized by hydrogenation of unsaturated
compounds, 18, 1 and 21 using hydrogen in the presence of palladium on charcoal as reported by
Venkateswarlu et al. [45] In the proton and carbon NMR spectra of 22 a mixture of the diketo
(10-15%) and enol (85-90%) forms was observed due to the lack of conjugation. This is not
surprising since the observation of both forms has also been reported in 1,5-diaryl-2,4-
pentanediones. [46]
The analogues shown in Scheme 2 contain a 5-carbon spacer and two identical aryl rings,
except for 36 and 37. These compounds were designed to test the importance of the length of the
spacer between two aryl rings. Analogues 24-35 and 40-47 were prepared as described by
Masuda et al [47] by reaction of the appropriate benzaldehyde with acetone in a base-catalyzed
aldol reaction. In the case of phenolic benzaldehydes, 26 and 27, the phenol was protected with a
methoxymethyl group prior to the aldol reaction and deprotected later to give the free phenol.
[47] Compounds 38 and 39 were prepared from 1,5-bis(4-hydroxyphenyl)-1,4-pentadien-3-one
and 24 respectively by reaction with acetic anhydride in the presence of pyridine in an
esterification reaction as described by Suarez. [48] Compounds 36 and 37, which are not
symmetrical, were prepared using two consecutive base-catalyzed aldol reactions. [47] The
substituted benzaldehyde needed to prepare compound 35, which is a new compound, was
synthesized by reacting 2,3-dimethoxyethylbenzene and N-methylformanilide with phosphorus
oxychloride. [49] The proton NMR of the compounds exhibited a pair of doublets in the aromatic
7

ACCEPTED MANUSCRIPT
region with J values between 15.5 to 16.5 Hz for the alkene protons present in the spacer, again
providing evidence of (E)-stereochemistry. The carbon NMR has two signals in the aromatic
region for the alkene carbons and a downfield signal for the carbonyl carbon.
Scheme 2
Scheme 2. Reagents for 5-carbon spacer compounds: (a) CH3C(O)CH3 (0.5 equiv.), NaOH,
CH3CH2OH, H2O, 67-96%; (b) H2, Pd/C, EtOAc, 80%; (c) CH3C(O)CH3 (1.0 equiv.), NaOH,
CH3CH2OH, H2O, PhCHO, 94-99%.
Analogue 48 was prepared by hydrogenation of 47 in the presence of palladium on charcoal.
[45] Compound 48 was verified by the appearance of triplets at 2.76 ppm and 2.97 ppm for the
alkane protons in the proton NMR along with the loss of doublets in the aromatic region for the
alkene protons.
Analogues in Scheme 3 contain two identical aryl rings separated by a five carbon unsaturated
spacer having both a carbonyl and a saturated ring. They were designed to test the importance of
a ring in the spacer. Compounds 49-63 contain a cyclohexanone ring whereas compounds 64-77
8

ACCEPTED MANUSCRIPT
contain an N-methylpiperidone ring. The compounds were synthesized using a procedure
described by Masuda et al [47] by reaction of a substituted benzaldehyde with cyclohexanone or
N-methylpiperidone in a base-catalyzed aldol reaction. The phenolic benzaldehydes needed to
prepare 49 and 60 were protected before the aldol reaction and deprotected after condensation as
described earlier. The formation of products was verified by proton NMR with the appearance of
a singlet at 7.7 ppm for the alkene protons on the spacer. The carbon NMR spectra have two new
signals in the aromatic region for the alkene carbons in the spacer.
Scheme 3
Scheme 3. Reagents for 5-carbon ring containing spacer compounds: (a) Cyclohexanone (0.5
equiv), NaOH, CH3CH2OH, H2O, 75-95%; (b) 1-Methyl-4-piperidone (0.5 equiv), NaOH,
CH3CH2OH, H2O, 77-99%.
Analogues 80-82, shown in Scheme 4, contain a single aryl ring with an unsaturated three-
carbon chain and a single carbonyl. Also shown in the scheme are compounds 78, 79 and 83-87,
which contain two aryl rings with the three carbon chain. All of the compounds were designed to
test the importance of a three-carbon chain and the importance of one aryl ring versus two aryl
9

ACCEPTED MANUSCRIPT
rings. Analogues 80-82 were prepared from excess acetone and a substituted benzaldehyde in a
base-catalyzed aldol reaction. [47] The proton NMR spectra of these compounds show a pair of
doublets in the aromatic region having J values of approximately 16 Hz for the alkene protons
and a signal at 2.3 ppm for the methyl protons. The carbon NMR spectra have two signals in the
aromatic region for the alkene carbons and a carbonyl carbon signal. Compounds 78, 79 and 83-
87 were prepared in a similar manner using a substituted acetophenone and a substituted
benzaldehyde in a base-catalyzed aldol condensation. [47,50,51] Analogues 78 and 79 were
protected with a methoxymethyl group before reaction and deprotected after condensation.
Compounds 83 and 85, which contain a carboxyl group, were synthesized following a procedure
described by Cleeland. [51] The proton and carbon NMR data is similar to that reported for
compounds 80-82 except for the absence of the methyl signal.
Scheme 4
10

ACCEPTED MANUSCRIPT
Scheme 4. Reagents for 3-carbon spacer compounds: (a) CH3C(O)CH3 (1.0 equiv), NaOH,
CH3CH2OH, H2O, 30-75%; (b) PhC(O)CH3, NaOH, CH3CH2OH, H2O, 55-91%.
An additional three-carbon spacer analogue, compound 88, was prepared to test the importance
of two carbonyls in the spacer. Analogue 88, whose structure is shown in Figure 4, was prepared
by reacting acetophenone with methylbenzoate and sodium methoxide in a condensation
reaction. [52] The proton NMR has a signal for the enol proton in the spacer at 6.8 ppm and the
carbon NMR has a signal at 93 ppm for the enol carbon and a signal at 185 ppm for the carbonyl
carbon, indicating 88 exists 100% in the keto-enol form.
2.2 Nrf2 activation by curcumin analogues with a 7-carbon spacer
Curcumin (Scheme 1, Structure 1) was evaluated with an Nrf2 reporter assay. Dose-response
studies demonstrated an EC50 = 21 µM and maximum activation 23-fold above untreated cells
(Table 1). Curcumin was used as the reference compound for studies of curcumin analogues. A
series of analogues that contained 7-carbon spacers between the aromatic rings was screened
initially at 15 µM, as shown in Figure 1. Several analogues (6,8 and 9) were comparable to
curcumin as activators of Nrf2. Dose-response studies (Table 1) demonstrated that analogues 8
and 9 activated Nrf2 36-fold and 33-fold, respectively, which is considerably higher than the
activation of Nrf2 by curcumin, suggesting that more than a single mechanism is likely involved.
This is consistent with reports that curcumin can activate Nrf2 by multiple mechanisms. [25-27]
Other analogues that gave moderate activation (2,3,4,6,8,9) shared with curcumin the presence
of methoxy and/or hydroxy functional groups. However, the presence of these functional groups
11

ACCEPTED MANUSCRIPT
does not consistently predict activation of Nrf2. Analogues 5 and 7, for example, exhibit very
little activity compared with analogue 6, which is an isomer of 5 and 7. Likewise, analogues 10,
11 and 12 exhibit very little activity compared with their isomers 8 and 9. Analogues of
curcumin with other ring substituents (13-17) exhibited no significant activity. Analogues 18, 21
and 23, which contained alkyl or aralkyl groups attached to the central carbon but contained no
ring substituents, demonstrated modest activity, whereas compounds 19 and 22, which are the
products from reduction of the double bonds in 18 and 21, showed no activity. This suggests that
those analogues that exhibit activity as activators of Nrf2 function, at least in part, as Michael
acceptors, consistent with one of the mechanisms of activation of Nrf2 by curcumin. [25,26]
12

ACCEPTED MANUSCRIPT
Figure 1. Nrf2 activation by curcumin analogues with a 7-carbon spacer
13

ACCEPTED MANUSCRIPT
2.3 Nrf2 activation by curcumin analogues with a 5-carbon spacer
A series of 12 analogues of curcumin with 5-carbon spacers between the aromatic rings was
screened at 15 µM concentration of analogue and compared with curcumin (Figure 2). A number
of analogues activated Nrf2 comparable to curcumin (analogues 24, 26, 28, 29, 34, 35). Dose-
response analyses demonstrated that most of these analogues were better than curcumin, both in
terms of higher fold activations and lower EC50 values (Table 1). Analogues 28 and 29, for
example, gave 60.3 and 66.5 fold activations. Analogues 28 and 34 are the most active when
expressed as the ratio of fold activation divided by EC50. These active analogues with 5-carbon
spacers shared with active analogues with 7-carbon spacers the presence of hydroxy and/or
methoxy functional groups. Analogues with other functional groups (analogues 40-44, 46)
showed little or no activity, except for analogue 45, which contains ortho chloro functional
groups. Interestingly, analogue 47, which contains no functional groups, showed significant
activity, whereas its reduction product 48 showed no activity, consistent with the data in Figure 1
and with the conclusion that an enone functionality is essential for activity.
14

ACCEPTED MANUSCRIPT
Figure 2. Nrf2 activation by curcumin analogues with a 5-carbon spacer.
15

ACCEPTED MANUSCRIPT
2.4 Activation of Nrf2 by curcumin analogues with a 5-carbon, ring-containing spacer
A series of 29 analogues of curcumin with 5-carbon spacers, in which the spacer contained a 6-
member ring, was screened at 15 µM analogue and compared with curcumin (Figure 3). A
number of analogues (51,52,55,61,62,69,70,73,74) showed significant activity with 51,52,61 and
69 showing activity comparable to curcumin. Unlike the results in Figures 1 and 2 where active
analogues contained hydroxyl or methoxy substituents, the analogues with a 5-carbon, ring-
containing spacer included numerous examples with substituents other than hydroxyl or methoxy
(analogues 61,62,71,73,74,77). Analogue 61, for example, contains meta fluoro substituents and
is comparable to curcumin in activity. Similar to the results in Figures 1 and 2, there are
numerous analogues that are isomers yet show markedly different activities. Analogue 55 is
active whereas its isomers 53 and 59 show little or no activity. Analogues 51, 52 and 69 are
active whereas isomers 56 and 68 are inactive.
16

ACCEPTED MANUSCRIPT
Figure 3. Nrf2 activation by curcumin analogues with a 5-carbon, ring-containing spacer.
17

ACCEPTED MANUSCRIPT
2.5 Activation of Nrf2 by curcumin analogues with a 3-carbon spacer
A series of 11 analogues of curcumin with 3-carbon spacers was screened at 15 µM analogue
and compared with curcumin (Figure 4). Analogues 78,79 and 85-88, all of which contained two
aromatic rings, showed significant activity, whereas none of the analogues with a single ring
showed activity. Analogue 88, which contained no ring substituents, was considerably more
active than curcumin.
Figure 4. Nrf2 activation by curcumin analogues with a 3-carbon spacer.
18

ACCEPTED MANUSCRIPT
Table 1 summarizes the most active of all of the analogues where both fold activation and EC50
values are considered.
Table 1. Fold activation of Nrf2 by curcumin analogues and EC50 values.
Fold
Number
Structure
Activation
EC50 (µM)
O
OH
Curcumin
22.6 ± 2.3
21 ± 0.02
OH
HO
H3C
O
O
O
O
1
6
H3C
CH3
H3C
OH
O
O
20.5 ± 1.3
35.7 ± 0.9
33.3 ± 4.0
25.1 ± 3.6
14.4 ± 4
16.6 ± 0.2
7.3 ± 1.5
9.9 ± 0.6
OH
CH3
H3C
8
9
O
O
O
O
O
O
H3C
H3C
OH
CH3
O
H3C
O
O
CH3
H3C
O
O
24
OH
OCH3
HO
OCH3
OH
OH
26
28
29
18.5 ± 4.4
60.3 ± 12
66.5 ± 7.4
3.3 ± 0.5
3.3 ± 0.1
10.3 ± 0.4
OCH3
OCH3
O
OCH3
OCH3
O
H3CO
OCH3
OCH3
OCH3
O
OCH3
OCH3
34
51
18.2 ± 0.08
22.1 ± 1.1
1.32 ± 0.2
5.3 ± 0.3
H3CO
OCH3
O
OCH3
H3CO
OCH3
OCH3
OCH3
O
OCH3
52
61
69
21.1 ± 1.7
24.9 ± 0.4
30.6 ± 6.8
18.1 ± 0.9
7.6 ± 0.3
8.0 ± 0.2
OCH3
OCH3
O
F
F
OCH3
O
N
OCH3
H3CO
H3CO
OCH3
OCH3
CH3
O
CH3
86
88
29.9 ± 4.3
44.7 ± 8.3
21.1 ± 0.5
7.2 ± 1.2
H3C
O
OH
19

ACCEPTED MANUSCRIPT
3. Discussion and Conclusions
The importance of inflammation and oxidative stress in many chronic diseases supports the
concept that activation of anti-oxidant Nrf2 signaling may have therapeutic potential. A number
of Nrf2 activators have entered into clinical trials. Bardoxolone methyl, an oleanolic acid-derived
synthetic triterpene, is a potent Nrf2 activator that was evaluated for its ability to slow
progression to end-stage renal disease in patients with type 2 diabetes and stage 4 chronic kidney
disease. Bardoxolone methyl is able to modify Keap1 as a Michael acceptor. This clinical trial
was terminated in phase III owing to an increase in heart failure. [53] However, bardoxolone
methyl is currently in phase III studies for pulmonary hypertension (ClinicalTrials.gov).
Dimethyl fumarate, a simple derivative of the metabolic intermediate fumaric acid, has recently
been FDA-approved for the treatment of relapsing-remitting multiple sclerosis. [54] Dimethyl
fumarate is active as the monomethyl derivative formed by the action of non-specific esterases.
Monomethyl fumarate modifies Keap1 through electrophilic addition. Among the natural
product activators of Nrf2, sulforaphane has been extensively studied, including in a number of
clinical trials. Sulforaphane has shown promise in treatment of recurrent prostate cancer, [55,56]
schizophrenia, [57] and autism spectrum disorder. [55] Sulforaphane also modifies Keap1
through electrophilic addition. [58]
Although numerous Nrf2-activating chemicals are electrophiles that modify Keap1 by Michael
addition, there are other mechanisms for activation of Nrf2. Epigenetic modifications of CpG
methylation status of Nrf2 by the anti-cancer drug 3,3’-diindolylmethane resulted in enhanced
expression of Nrf2 and of Nrf2-target genes in cell-based studies. [55] This was suggested as an
explanation for the chemopreventive activity of this drug. The promoter of the gene encoding
Nrf2 contains XRE sequences, which can recruit the aromatic hydrocarbon receptor AhR; this
20

ACCEPTED MANUSCRIPT
allows the Nrf2 gene to be activated by polycyclic aromatic hydrocarbons. [56] The promoter of
the Nrf2 gene also contains ARE-like sequences, which allows Nrf2 to regulate its own
expression. [57] The autophagy cargo receptor and signaling adaptor protein p62 contains a
binding site for Keap1. The attraction of Keap1 to this site is further regulated by
phosphorylation by mTORC1, which leads to autophagosome–mediated destruction of Keap1
and activation of Nrf2. [58-60] This is one example of proteins other than Nrf2 that interact
specifically with Keap1, which indicates a broad role for Keap1 and a complex set of
mechanisms that impact Nrf2 levels. [61] There are also mechanisms for activation of Nrf2 that
do not appear to involve Keap1. Nrf2 activity is inhibited by the serine/threonine kinase GSK-3.
GSK-3 constitutive activity is inhibited by PKB/Akt phosphorylation. It is conceivable that
analogues of curcumin might activate Nrf2 independently of Keap1 through inhibition of GSK-3
or activation of kinases such as Akt. [61]
In the present study, we demonstrated that a wide range of curcumin analogues can activate
Nrf2. This includes dienones with a 7-carbon spacer, enones with a 5-carbon spacer, and enones
with a 3-carbon spacer (Table 1). It is likely that these analogues function to modify Keap1
through Michael addition, similar to curcumin. [25,26] However, they may utilize other
mechanisms, again similar to curcumin. [25,27] Some of these curcumin analogues have an
extensive literature. Analog 26, for example, was originally identified as a Michael acceptor with
high reactivity for sulfhydryl groups and ability to induce the expression of phase II enzymes
that protect against carcinogenesis. [62] Keap1 was subsequently identified as the source of the
critical intracellular sulfhydryl groups. [63] Analog 26, which we initially developed as an
inhibitor of the activation of pro-inflammatory NF-
κB signaling [41,42] and AP-1 signaling, [43]
was subsequently shown to down-regulate expression of the androgen receptor in human prostate
21

ACCEPTED MANUSCRIPT
cancer cells. [64] This involved an oxidative stress-mediated mechanism supported by
development of reactive oxygen species (ROS), activation of Nrf2, and induction of Nrf2-
dependent genes for NAD(P)H quinone oxidoreductase-1 and aldo-keto reductase 1C1. Analog
26 has been shown to exhibit complex concentration-dependent effects on growth of leukemia
cells. [65] Recently, it was reported that analog 26 activates Nrf2 through a Keap1-Cys151-
dependent process. [32]
The results with analog 26 raise several issues with respect to development of Nrf2 activators
as therapeutic agents. First, as mentioned above, there is the issue of selectivity for electrophilic
activators that target sulfhydryl groups in Keap1. Such agents likely would also modify other
sulfhydryl groups in numerous proteins. Second, there is the issue of activators that produce
ROS, which then leads to activation of Nrf2 through oxidative modification of Nrf2. Since the
goal is to develop therapeutic activators of Nrf2 to combat the oxidative stress associated with
many chronic inflammatory diseases, it seems counterproductive to focus on activators that
produce oxidative stress. It should be noted, however, that Nrf2 activators such as bardoxolone
methyl, dimethyl fumarate and sulforaphane also covalently modify Keap1, yet are promising
therapeutic agents. Moreover, curcumin itself has been studied extensively in clinical trials and is
considered to have low toxicity. In this study numerous analogues of the natural product
cucumin containing 7-carbon dienone spacers as in curcumin, 5-carbon enone spacers including
ring-containing spacers, and 3-carbon enone spacers were found to be activators of the anti-
oxidant transcription factor Nrf2. The presence of an enone functionality appears to be essential.
22

ACCEPTED MANUSCRIPT
4. Experimental Section
4.1 Assay Validation
The evaluation of the ability of enone analogues of curcumin to activate Nrf2 was determined
with a luciferase assay under the following conditions: 1) each analogue was evaluated in
triplicate assays with untreated cells as the negative control and curcumin-treated cells as the
positive control; 2) cells were washed extensively to remove drug before the luciferase activity
was measured; 3) before measurement of luciferase activity, the cells were lysed with buffer,
which further reduced the concentration of any remaining drug; 4) only a portion of the lysate
was used in the final assay, which diluted any remaining drug even further; 5) the measurement
of final luciferase activity was verified as a valid measurement by adding back low
concentrations of drug to determine whether this treatment either inhibited or enhanced the
luciferase activity. It was observed that this treatment with low concentrations of drug did not
alter the luciferase activity. We conclude, therefore, that the luciferase measurement conducted
under these conditions is a valid determination of the level of luciferase gene expression.
4.2 Reporter Assays
Nrf2-ARE reporter-HepG2 stable cell line (BPS Bioscience, San Diego, CA) was grown in a
humidified atmosphere at 37°C in 5% CO2/95% air. The cells were maintained in MEM
medium with Earles balanced salts and L-glutamine supplemented with 10% fetal bovine serum,
1 mM sodium pyruvate, 1% non-essential amino acids, 100 units/mL penicillin, 100 µg/mL
streptomycin, and 400 µg/mL Geneticin. One day prior to treatment, the Nrf2-ARE cells were
plated into 24-well cell culture plates at approximately 30% confluency in the above media
without Geneticin. The following day, fresh media with or without curcumin analogues was
applied to the cells. DMSO concentrations were kept at 0.1%. The cells were again placed in a
23

ACCEPTED MANUSCRIPT
humidified atmosphere at 37°C in 5% CO2/95% air for 5 hours. Plate wells were gently washed
with phosphate buffered saline (PBS) pH 7.4 and lysed with 1x passive lysis buffer.
4.3 General Chemical Procedures
Reagents were purchased from commercial sources (Aldrich, Acros). The reactions were
monitored by thin layer chromatography (TLC), which was carried out on silica gel 60F254 plates
and visualized using UV light. If necessary products were purified by column chromatography
using EM type 60 silica gel (230-400 mesh). Melting points were taken on a Thomas-Hoover
Uni-Melt capillary melting point apparatus and reported uncorrected. Infrared spectra were
obtained using a Bruker ALPHA spectrometer. Unless otherwise noted, 1H spectra were recorded
by using CDCl3 solutions at 250, 300 or 500 MHz; 13C NMR spectra were recorded in CDCl3 at
63, 75 or 125 MHz. The data were collected on a Bruker Avance III HD 500 MHz, a Bruker
Avance III 300 MHz or a Bruker AC 250 MHz spectrometer. Chemical shifts are reported in
ppm relative to TMS at 0.0 ppm for both 1H NMR and 13C NMR. Proton NMR data are reported
as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, quin =
quintet, sept = septet, dd = doublet of doublets, td = triplet of doublets, ddd = doublet of doublets
of doublets, m = multiplet), coupling constant (J in Hz) and integration. Peak assignments were
made with the aid of DEPT and HMBC spectra. The samples for accurate mass analysis (HRMS)
were run by electrospray ionization in positive mode on a Waters LCT Premier time of flight
mass spectrometer after they were dissolved in ethyl acetate with subsequent dilution into
methanol and water (037 with formic acid). High resolution mass spectra (HRMS) were obtained
at the UNM Mass Spectrometry Facility, Albuquerque, New Mexico. Compounds 1-3, 5, 6, 8, 9,
13, 18-24, 26-33, 36-40, 43-49, 78-88 were prepared and characterized according to procedures
24

ACCEPTED MANUSCRIPT
described by Weber et al. [41-43] Compounds 35, 64, 67, 69, 70, 73-77 are new analogues of
curcumin that were not included in our previous studies and were prepared and characterized
according to procedures described below. The purity of these compounds was found to be greater
than 95% by analysis with high performance liquid chromatography (HPLC). HPLC analysis
was performed with an Agilent 1290 Infinity instrument using a Phenomenex Luna 5u C18(2)
column, 250 x 4.60 mm, 100 Å, detecting at either 345 nm or 360 nm on a diode-array detector.
The solvent gradient employed was 0.1% aqueous H3PO4 to 100% acetonitrile over 15 minutes,
followed by 5 minutes elution with acetonitrile, at a flow rate of 1 mL/min. Samples were
dissolved in acetonitrile as the diluent. Detailed procedures and characterization data are
described in the supporting information.
4.4 General Procedure for Synthesis of Compounds 24-47 and 49-77 [47]
To a solution of the aromatic aldehyde (10 mmol, 2 equiv.) dissolved in 10 ml of ethanol was
added the appropriate ketone (5 mmol, 1 equiv) such as acetone (compounds 24-47),
cyclohexanone (compounds 49-63), or N-methylpiperidone (compounds 64-77) and stirred for
15 minutes. Sodium hydroxide (0.4 g) dissolved in water (10 mL) was added and the resulting
mixture was stirred at room temperature overnight. The separated solid was filtered, washed with
cold ethanol and recrystallized from ethanol or ethanol/water.
4.4.1. 1,5-Bis((E)-2-ethyl-3,4-dimethoxyphenyl)-pentadien-3-one (35): 85% yield, yellow
crystals: mp 166-167 ° δ 8.00
C; IR (ATR): νmax 1613, 1575 cm-1; 1H NMR (CDCl3, 300 MHz):
(d, J = 15.6 Hz, 2H, Ar-CH=C x 2), 7.46 (d, J = 8.7 Hz, 2H, Ar-H6 x 2), 6.91 (d, J = 15.6 Hz,
2H, =CHC(O) x 2), 6.83 (d, J = 8.7 Hz, 2H, Ar-H5 x 2), 3.91 (s, 6H, OCH3 x 2), 3.84 (s, 6H,
13
OCH3 x 2), 2.87 (q, J = 7.5 Hz, 4H, Ar-CH2 x 2), 1.19 (t, J = 7.5 Hz, 6H, CH3 x 2); C NMR
(CDCl3, 75 MHz):
δ
188.7 (C=O), 154.4 (ArC3 x 2), 147.0 (ArC4 x 2), 140.4 (Ar-CH= x 2),
139.0 (ArC2 x 2), 126.8 (ArC1 x 2), 125.5 (=CH x 2), 123.0 (ArC6H x 2), 110.0 (ArC5H x 2),
60.9 (OCH3 x 2), 55.7 (OCH3 x 2), 19.5 (CH2 x 2), 15.9 (CH3 x 2). HRMS (EI) calcd for
C25H30O5 [M+H]+: 411.2171; found, 411.2174. HPLC tR (min): 17.201. PHPLC/%: 95.78.
25

ACCEPTED MANUSCRIPT
4.4.2. 3,5-Bis((E)-3-methoxybenzylidene)-1-methylpiperidin-4-one (64): 89% yield, yellow
1
crystals; mp 98
°
C; IR (ATR): νmax 1603, 1574 cm-1; H NMR (CDCl3, 300 MHz):
δ
7.78 (s, 2H,
Ar-CH=C x 2), 7.34 (t, J = 7.9 Hz, 2H, Ar-H5 x 2), 6.99 (d, J = 7.8 Hz, 2H, Ar-H6 x 2), 6.93 (s,
2H, Ar-H2 x 2), 6.91 (d, J = 8.9 Hz, 2H, Ar-H4 x 2), 3.84 (s, 6H, OCH3 x 2), 3.76 (s, 4H,
13
CH2NCH2), 2.45 (s, 3H, NCH3); C NMR (CDCl3, 75 MHz):
δ
186.8 (C=O), 159.5 (ArC3 x 2),
136.5 (ArC1 x 2), 136.2 (CH= x 2), 133.4 (=C< x 2), 129.5 (ArC5H x 2), 122.7 (ArC6H x 2), 115.9
(ArC4H x 2), 114.5 (ArC2H x 2), 57.1 (CH2NCH2), 55.3 (OCH3 x 2), 45.8 (NCH3). HRMS (EI)
calcd for C22H23NO3 [M+H]+: 350.1756; found, 350.1754. HPLC tR (min): 10.213. PHPLC/%:
96.73.
4.4.3. 3,5-Bis((E)-2,4-dimethoxybenzylidene)-1-methylpiperidin-4-one (67): 77% yield,
1
yellow crystals; mp 131-132
δ
°
C; IR (ATR): νmax 1588, 1564 cm-1; H NMR (CDCl3, 300 MHz):
8.04 (s, 2H, Ar-CH=C x 2), 7.14 (d, J = 8.4 Hz, 2H, Ar-H6 x 2), 6.50 (dd, J = 8.5, 2.3 Hz, 2H,
Ar-H5 x 2), 6.46 (d, J = 2.3 Hz, 2H, Ar-H3 x 2), 3.83 (s, 6H, OCH3 x 2), 3.82 (s, 6H, OCH3 x 2),
3.66 (s, 4H, CH2NCH2), 2.40 (s, 3H, NCH3); 13C NMR (CDCl3, 75 MHz):
δ 186.7 (C=O), 161.9
(ArC4 x 2), 160.0 (ArC2 x 2), 131.6 (=C< x 2), 131.4 (CH= x 2), 131.2 (ArC6H x 2), 117.4 (ArC1
x 2), 104.3 (ArC5H x 2), 98.2 (ArC3H x 2), 57.3 (CH2NCH2), 55.4 (OCH3 x 2), 55.3 (OCH3 x 2),
45.7 (NCH3). HRMS (EI) calcd for C24H27NO5 [M+H]+: 410.1967; found, 410.1966. HPLC tR
(min): 9.945. PHPLC/%: 99.87.
4.4.4. 3,5-Bis((E)-2,3,4-trimethoxybenzylidene)-1-methylpiperidin-4-one (69): 95% yield,
1
yellow crystals; mp 192
°
C; IR (ATR): νmax 1593, 1577 cm-1; H NMR (CDCl3, 300 MHz):
δ
8.00 (s, 2H, Ar-CH=C x 2), 6.95 (d, J = 8.7 Hz, 2H, Ar-H6 x 2), 6.70 (d, J = 8.7 Hz, 2H, Ar-H5 x
2), 3.90 (s, 6H, OCH3 x 2), 3.89 (s, 6H, OCH3 x 2), 3.88 (s, 6H, OCH3 x 2), 3.68 (s, 4H,
CH2NCH2), 2.42 (s, 3H, NCH3); 13C NMR (CDCl3, 75 MHz):
δ
186.5 (C=O), 154.8 (ArC4 x 2),
153.7 (ArC2 x 2), 142.4 (ArC3 x 2), 132.3 (=C< x 2), 131.8 (CH= x 2), 125.2 (ArC6H x 2), 122.5
(ArC1 x 2), 106.9 (ArC5H x 2), 61.7 (OCH3 x 2), 61.0 (OCH3 x 2), 57.4 (CH2NCH2), 56.0 (OCH3
x 2), 45.7 (NCH3). HRMS (EI) calcd for C26H31NO7 [M+H]+: 470.2179; found, 470.2181. HPLC
tR (min): 9.642. PHPLC/%: 99.87.
4.4.5. 3,5-Bis((E)-3,5-dimethoxybenzylidene)-1-methylpiperidin-4-one (70): 86% yield,
yellow crystals; mp 143-144
7.72 (s, 2H, Ar-CH=C x 2), 6.53 (d, J = 2.2 Hz, 4H, Ar-H2 x 2, Ar-H6 x 2), 6.48 (d, J = 2.2 Hz,
2H, Ar-H4 x 2), 3.82 (s, 12H, OCH3 x 4), 3.75 (s, 4H, CH2NCH2), 2.44 (s, 3H, NCH3); 13C NMR
°
C; IR (ATR): νmax 1619, 1588 cm-1; 1H NMR (CDCl3, 300 MHz):
δ
(CDCl3, 75 MHz):
δ
186.8 (C=O), 160.7 (ArC3 x 2, ArC5 x 2), 137.0 (ArC1 x 2), 136.4 (CH= x
2), 133.6 (=C< x 2), 108.4 (ArC2H x 2, ArC6H x 2), 101.0 (ArC4H x 2), 57.0 (CH2NCH2), 55.4
(OCH3 x 4), 45.8 (NCH3). HRMS (EI) calcd for C24H27NO5 [M+H]+: 410.1967; found,
410.1972. HPLC tR (min): 10.483. PHPLC/%: 99.91.
4.4.6. 3,5-Bis((E)-3-trifluoromethylbenzylidene)-1-methylpiperidin-4-one (73): 80% yield,
yellow crystals; mp 142-143
δ
°
C; IR (ATR): νmax 1617, 1592 cm-1; 1H NMR (CDCl3, 300 MHz):
7.82 (s, 2H, Ar-CH=C x 2), 7.63 (m, 4H, Ar-H6 x 2, Ar-H2 x 2), 7.56 (m, 4H, Ar-H5 x 2, Ar-H4
13
x 2), 3.75 (d, J = 1.4 Hz, 4H, CH2NCH2), 2.47 (s, 3H, NCH3); C NMR (CDCl3, 75 MHz):
δ
186.3 (C=O), 135.9 (ArC1 x 2), 134.9 (CH= x 2), 134.4 (=C< x 2), 133.3 (ArC6H x 2), 131.2 (q,
J = 32.5 Hz, ArC3 x 2), 129.2 (ArC5H x 2), 126.9 (d, J = 3.6 Hz, ArC2H x 2), 125.6 (d, J = 3.3
Hz, ArC4H x 2), 120.3 (d, J = 272.4 Hz, CF3 x 2), 56.8 (CH2NCH2), 45.8 (NCH3). HRMS (EI)
26

ACCEPTED MANUSCRIPT
calcd for C22H17F6NO [M+H]+: 426.1293; found, 426.1295. HPLC tR (min): 11.245. PHPLC/%:
100.
4.4.7. 3,5-Bis((E)-4-isopropylbenzylidene)-1-methylpiperidin-4-one (74): 93% yield, yellow
1
crystals; mp 98-100
°
C; IR (ATR): νmax 1599, 1581, 1557 cm-1; H NMR (CDCl3, 300 MHz):
δ
7.80 (s, 2H, Ar-CH=C x 2), 7.35 (d, J = 8.2 Hz, 4H, Ar-H2 x 2, Ar-H6 x 2), 7.28 (d, J = 8.2 Hz,
4H, Ar-H3 x 2, Ar-H5 x 2), 3.78 (s, 4H, CH2NCH2), 2.94 (sept, J = 6.9 Hz, 2H, CH x 2), 2.46 (s,
3H, NCH3), 1.27 (d, J = 6.9 Hz, 12H, CH3 x 4); 13C NMR (CDCl3, 75 MHz):
δ 186.9 (C=O),
150.1 (ArC4 x 2), 136.3 (CH= x 2), 132.9 (=C< x 2), 132.5 (ArC1 x 2), 130.6 (ArC2H x 2, ArC6H
x 2), 126.7 (ArC3H x 2, ArC5H x 2), 57.2 (CH2NCH2), 45.8 (NCH3), 34.0 (CH x 2), 23.8 (CH3 x
4). HRMS (EI) calcd for C26H31NO [M+H]+: 374.2484; found, 374.2481. HPLC tR (min):
11.522. PHPLC/%: 97.04.
4.4.8. 3,5-Bis((E)-4-methylthiobenzylidene)-1-methylpiperidin-4-one (75): 99% yield, yellow
crystals; mp 122-123 ° δ 7.75 (s, 2H,
C; IR (ATR): νmax 1573 cm-1; 1H NMR (CDCl3, 300 MHz):
Ar-CH=C x 2), 7.33 (d, J = 8.5 Hz, 4H, Ar-H3 x 2, Ar-H5 x 2), 7.26 (d, J = 8.5 Hz, 4H, Ar-H2 x
13
2, Ar-H6 x 2), 3.75 (s, 4H, CH2NCH2), 2.51 (s, 6H, SCH3 x 2), 2.47 (s, 3H, NCH3); C NMR
(CDCl3, 75 MHz):
δ
186.6 (C=O), 140.6 (ArC4 x 2), 135.7 (CH= x 2), 132.5 (=C< x 2), 131.7
(ArC1 x 2), 130.9 (ArC2H x 2, ArC6H x 2), 125.7 (ArC3H x 2, ArC5H x 2), 57.1 (CH2NCH2),
45.9 (NCH3), 15.1 (SCH3 x 2). HRMS (EI) calcd for C22H23NOS2 [M+H]+: 382.1299; found,
382.1292. HPLC tR (min): 11.008. PHPLC/%: 99.01.
4.4.9. 3,5-Bis((E)-4-ethoxybenzylidene)-1-methylpiperidin-4-one (76): 84% yield, yellow
crystals; mp 162-163 ° δ 7.76
C; IR (ATR): νmax 1601, 1581 cm-1; 1H NMR (CDCl3, 300 MHz):
(s, 2H, Ar-CH=C x 2), 7.35 (d, J = 8.7 Hz, 4H, Ar-H2 x 2, Ar-H6 x 2), 6.92 (d, J = 8.7 Hz, 4H,
Ar-H3 x 2, Ar-H5 x 2), 4.07 (q, J = 6.9 Hz, 4H, OCH2 x 2), 3.76 (s, 4H, CH2NCH2), 2.48 (s, 3H,
13
NCH3), 1.43 (t, J = 7.1 Hz, 6H, CH3 x 2); C NMR (CDCl3, 75 MHz):
δ 186.7 (C=O), 159.6
(ArC4 x 2), 135.9 (CH= x 2), 132.3 (ArC2H x 2, ArC6H x 2), 131.3 (=C< x 2), 127.8 (ArC1 x 2),
114.5 (ArC3H x 2, ArC5H x 2)), 63.5 (OCH2 x 2), 57.2 (CH2NCH2), 45.9 (NCH3), 14.7 (CH3 x
2). HRMS (EI) calcd for C24H27NO5 [M+H]+: 378.2069; found, 378.2069. HPLC tR (min):
10.704. PHPLC/%: 100.
4.4.10. 3,5-Bis((E)-3-methylbenzylidene)-1-methylpiperidin-4-one (77): 88% yield, yellow
crystals; mp 79-80 ° δ 7.79 (s,
C; IR (ATR): νmax 1601, 1582 cm-1; 1H NMR (CDCl3, 500 MHz):
2H, Ar-CH=C x 2), 7.31 (t, J = 7.6 Hz, 2H, Ar-H5 x 2), 7.21 (s, 2H, Ar-H2 x 2), 7.20 (d, J = 8.7
Hz, 2H, Ar-H6 x 2), 7.18 (d, J = 8.3 Hz, 2H, Ar-H4 x 2), 3.76 (s, 4H, CH2NCH2), 2.46 (s, 3H,
13
NCH3), 2.39 (s, 6H, CH3 x 2); C NMR (CDCl3, 125 MHz):
δ
186.8 (C=O), 138.1 (ArC3 x 2),
136.5 (CH= x 2), 135.2 (ArC1 x 2), 133.0 (=C< x 2), 131.2 (ArC5H x 2), 129.8 (ArC4H x 2),
128.4 (ArC2H x 2), 127.3 (ArC6H x 2), 57.0 (CH2NCH2), 45.7 (NCH3), 21.4 (CH3 x 2). HRMS
(EI) calcd for C22H23NO [M+H]+: 318.1858; found, 318.1861. HPLC tR (min): 10.563. PHPLC/%:
99.05.
27

ACCEPTED MANUSCRIPT
Acknowledgments
This work was supported in part by grant BC043125 from the U.S. Army/DOD Breast Cancer
Program. HRMS data were obtained by Ken Sherrell at the UNM Mass Spectrometry Facility,
Albuquerque, New Mexico.
Appendix A. Supporting Information
General synthetic procedures for compounds 1-23, 48, 78-88 and characterization data for
compounds 4, 7, 10-12, 14-17, 25, 34, 41, 42, 50-63, 65, 66, 68, 71, 72 can be found at
References
[1] G. Joshi, J.A. Johnson, The Nrf2-ARE pathway: a valuable therapeutic target for the
treatment of neurological diseases, Recent Pat. CNS Drug Discov. 7 (2012) 218-229.
[2] M. Sandberg, J. Patil, B. D’Angelo, S.G. Weber, C. Mallard, NRF2-regulation in brain
health and disease: implications of cerebral inflammation, Neuropharmacology 79 (2014) 298-
305.
[3] L.F. Cardozo, L. M. Pedruzzi, P. Stenvinkel, M. B. Stockler-Pinto, J. B. Daleprane, M.
Leite Jr., D. Mafra, Nutritional strategies to modulate inflammation and oxidative stress
pathways via activation of the master antioxidant switch Nrf2, Biochemie 95 (2013) 1525-1533.
[4] S. Ruiz, P.E. Pergola, R.A. Zager, N.D. Vaziri, Targeting the transcription factor Nrf2 to
ameliorate oxidative stress and inflammation in chronic kidney disease, Kidney Int. 83 (2013)
1029-1041.
28