Design, synthesis, and evaluation of curcumin derivatives as Nrf2 activators and cytoprotectors against oxidative death

Article abstract of DOI:10.1016/j.ejmech.2017.04.008

Activation of nuclear factor erythroid-2-related factor 2 (Nrf2) has been proven to be an effective means to prevent the development of cancer, and natural curcumin stands out as a potent Nrf2 activator and cancer chemopreventive agent. In this study, we synthesized a series of curcumin analogs by introducing the geminal dimethyl substituents on the active methylene group to find more potent Nrf2 activators and cytoprotectors against oxidative death. The geminally dimethylated and catechol-type curcumin analog (compound 3) was identified as a promising lead molecule in terms of its increased stability and cytoprotective activity against the tert-butyl hydroperoxide (t-BHP)-induced death of HepG2 cells. Mechanism studies indicate that its cytoprotective effects are mediated by activating the Nrf2 signaling pathway in the Michael acceptor- and catechol-dependent manners. Additionally, we verified by using copper and iron ion chelators that the two metal ion-mediated oxidations of compound 3 to its corresponding electrophilic o-quinone, contribute significantly to its Nrf2-dependent cytoprotection. This work provides an example of successfully designing natural curcumin-directed Nrf2 activators by a stability-increasing and proelectrophilic strategy.

Full text of DOI:10.1016/j.ejmech.2017.04.008

Accepted Manuscript
Design, synthesis, and evaluation of curcumin derivatives as Nrf2 activators and
cytoprotectors against oxidative death
Zhi-Shan Tu, Qi Wang, Dan-Dan Sun, Fang Dai, Bo Zhou
PII:
S0223-5234(17)30263-5
10.1016/j.ejmech.2017.04.008
EJMECH 9355
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 15 January 2017
Revised Date: 1 April 2017
Accepted Date: 4 April 2017
Please cite this article as: Z.-S. Tu, Q. Wang, D.-D. Sun, F. Dai, B. Zhou, Design, synthesis, and
evaluation of curcumin derivatives as Nrf2 activators and cytoprotectors against oxidative death,
European Journal of Medicinal Chemistry (2017), doi: 10.1016/j.ejmech.2017.04.008.
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
1
2
Design, synthesis, and evaluation of curcumin derivatives as Nrf2 activators and
cytoprotectors against oxidative death
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Zhi-Shan Tu, Qi Wang, Dan-Dan Sun, Fang Dai*, Bo Zhou*
State Key Laboratory of Applied Organic Chemistry, Lanzhou University, 222
Tianshui Street S., Lanzhou, Gansu 730000, China
1
1
1
1
1
1
1
1
1
1
2
______________________________
* Corresponding author.
E-mail: bozhou@lzu.edu.cn; daifang@lzu.edu.cn
Fax: +86-931-8915557
1

ACCEPTED MANUSCRIPT
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Abstract
Activation of nuclear factor erythroid-2-related factor 2 (Nrf2) has been proven to
be an effective means to prevent the development of cancer, and natural curcumin
stands out as a potent Nrf2 activator and cancer chemopreventive agent. In this study,
we synthesized a series of curcumin analogs by introducing the geminal dimethyl
substituents on the active methylene group to find more potent Nrf2 activators and
cytoprotectors against oxidative death. The geminally dimethylated and catechol-type
curcumin analog (compound 3) was identified as a promising lead molecule in terms
of its increased stability and cytoprotective activity against the tert-butyl hydroperox-
ide (t-BHP)-induced death of HepG2 cells. Mechanism studies indicate that its cyto-
protective effects are mediated by activating the Nrf2 signaling pathway in the Mi-
chael acceptor- and catechol-dependent manners. Additionally, we verified by using
copper and iron ion chelators that the two metal ion-mediated oxidations of compound
3 to its corresponding electrophilic o-quinone, contribute significantly to its
Nrf2-dependent cytoprotection. This work provides an example of successfully de-
signing natural curcumin-directed Nrf2 activators by a stability-increasing and proe-
lectrophilic strategy.
Keywords: Nrf2, curcumin, cytoprotective activity, Michael acceptor, catechol
2

ACCEPTED MANUSCRIPT
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
1. Introduction
Nuclear factor erythroid-2-related factor 2 (Nrf2) is an imperative regulator of cel-
lular resistance against electrophilic and oxidative stress by activating phase II detox-
ifying and antioxidant enzymes such as hemeoxygenase-1(HO-1), glutathione
S-transferase (GST), γ-glutamylcysteine synthetase (γGCS) and γ-glutamyl cysteine
ligase (γ-GCL) [1]. In its inactive state, the basic leucine zipper protein (bZIP) tran-
scription factor is associated with a Kelch-like ECH-associated protein 1 (Keap1) and
sequestered in the cytoplasm, which facilitates its ubiquitination and proteasomal
degradation. Once cells are exposed to oxidative or electrophilic stress, the active
cysteine residues in Keap1 are covalently modified, resulting in Nrf2 disassociation
from Keap1, thereby causing it stabilization and translocation to the nucleus. Accord-
ingly, the Nrf2 accumulates in the nucleus, dimerizes with obligatory partner sMaf,
and subsequently transactivates antioxidant response element (ARE)-driven phase II
detoxifying and antioxidant genes [2-4]. In addition, numerous studies also indicate
that diverse protein kinases including mitogen-activated protein kinase (MAPK),
phosphatidylinositol 3-kinase (PI3K) and protein kinase C are involved in the Nrf2
transcriptional activation [5, 6].
Nrf2 has been identified as a potential molecular target for cancer chemoprevention
and this has sparked the increasing interest in identifying and developing natural
product-derived chemopreventive agents by targeting the keap1-Nrf2/ARE pathway
[4, 7]. Curcumin (Fig. 1) is a renowned natural polyphenolic compound that isolated
3

ACCEPTED MANUSCRIPT
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
from the rhizome of Curcuma longa. It has shown diverse biological activities in-
cluding antioxidant, anti-inflammatory and anticancer activities [8, 9]. Although the
precise mechanisms involved in its various activities remains to be elucidated, the an-
tioxidant capacity of this molecule seems to be an element underlying its versatile bi-
ological properties. By virtue of its two phenolic groups and the β-diketone moiety,
curcumin is known to directly scavenge the reactive oxygen species (ROS) [10, 11].
Moreover, due to the electrophilic characteristics of the Michael acceptor moieties, it
can also act as an indirect antioxidant by activating the Keap1-Nrf2/ARE signaling
pathway, which is crucial for cytoprotection against various forms of stress [12, 13].
Unfortunately, the clinical application of curcumin has been retarded by its poor bio-
availability due to its low solubility in aqueous solutions and instability under physi-
ological conditions [9, 14]. To deal with these drawbacks, numerous approaches have
been undertaken including changing the keto-enol moiety into monocarbonyl or het-
erocyclic moiety, and substitution at central methylene carbon and so on [9, 15, 16].
Among these approaches, the geminal dimethylation on the central methylene carbon
has been verified as an efficient modification strategy to improve pharmacological
activities of curcumin [16-18]. Specifically, introduction of two hydrophobic methyl
groups results in increased stability and anticancer activity [16], antioxidative activity
[17] and water-solubility [18].
Thus, we synthesized a panel of curcumin analogs by insertion of dimethyl substit-
uents on the active methylene group of curcumin (Fig. 1) to find more effective Nrf2
4

ACCEPTED MANUSCRIPT
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
activators than the parent molecule. Moreover, we have previously identified a cate-
chol-type resveratrol analog as a proelectrophile that is converted intracellularly into
its corresponding electrophilic o-quinone by copper ions. The copper-ion mediated
oxidation is indispensible for its cytoprotective activity against the oxidative death by
activating Nrf2 [19]. Therefore, the catechol moiety was also considered to introduce
into the geminal dimethyl substituted-curcumin analogs. The further modification
generated compound 3 (Fig. 1), a catechol-type and geminally dimethylated curcumin
analog. In view of this molecule containing two active sites (the Michael acceptor and
catechol units), we further synthesized its reduced analog (3-H) and catechol-omitting
analog (10) to clarify whether the two units are crucial for its Nrf2-dependent cyto-
protective activity.
Figure 1 here
2. Results
2.1 Chemical Synthesis
Compounds 1-10 were synthesized according to the route shown in Scheme 1, with
100 commercially available aromatic aldehydes as the starting materials. The aromatic al-
101 dehydes containing relatively active phenolic hydroxyl group were first protected by
102 methoxymethyl chloride (MOMCl) or tert-butyldimethylsilyl chloride (TBSCl). The
103 classical 1, 3-diketones (intermediates 2) were synthesized by the aldol condensation
104 of the modified benzaldehydes with acetylacetone according to the Pabon reaction
5

ACCEPTED MANUSCRIPT
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
05 [20]. Then the final products 2, 4–8 and 10 were synthesized by the substitution of
06 iodomethane. Compounds 1, 3 and 9 bearing the relatively active phenolic hydroxyl
07 group were obtained by removing the MOM or TBS groups in the final step. Com-
08 pound 3-H was synthesized by catalytic hydrogenation of compound 3 over Pd/C.
09
Scheme 1 here
10 2.2 Stability assay of curcumin and its analogs
11
12
13
14
15
16
17
18
19
20
21
Stability of the selected compounds (40 µM) in PBS (pH 7.4) was assessed by
monitoring their UV–visible absorption changes at 25 °C. As shown in Fig. 2, cur-
cumin was rapidly degraded in PBS within 30 min, while the geminally dimethylat-
ed curcumin, compound 1, showed no signs of decomposition over a 30 min period.
These results demonstrate that geminal dimethylation on the central methylene group
improves the stability of curcumin. Notably, compound 3 was also stable under the
same conditions, whereas methylation and dehydroxylation of the two catechol moi-
eties in compound 3 to generate compounds 2 and 10, respectively, led to significant
decomposition. These data suggest that the catechol moieties of compound 3 are also
essential for its stability.
Figure 2 here
22 2.3 SAR analysis for the cytoprotection of curcumin and its analogs
23 The t-BHP-induced oxidative death of HepG2 cells was employed as a model to
24 evaluate the cytoprotective activity of curcumin and its analogs [21, 22]. Specifically,
25 the cells were pretreated with the test compounds at nontoxic concentration (5 µM)
6

ACCEPTED MANUSCRIPT
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
26 (Fig. 3A) for 24 h followed by their removal and a subsequent challenge with 900 µM
27 t-BHP. As shown in Fig 3B, exposure of the cells to t-BHP for 6 h resulted in about
28 70% cell death, but pretreatment with 5 µM curcumin and its analogs attenuated cell
29 death induced by t-BHP. Most of the analogs were more potent cytoprotective agents
30 than the parent curcumin. A comparison of cytoprotection between curcumin and
31 compound 1 clearly indicates that geminal dimethylation on the central methylene
32 carbon of curcumin enhanced significantly its activity. Furthermore, demethylation of
33 compound 1 to produce the catechol-type compound 3 resulted in an enhanced cyto-
34 protection. On the contrary, abrogating the catechol (compound 10) or Michael ac-
35 ceptor (3-H) moieties of compound 3 led to an obvious reduction in the cytoprotec-
36 tion. These results indicate that both the catechol and Michael acceptor moieties are
37 essential for recapitulating the cytoprotective activity observed for compound 3.
38
Figure 3 here
39 2.4. Compound 3 attenuated dose-dependently the cell death and ROS accumulation
40 induced by t-BHP
41
Considering that compound 3 was the most prominent one among the test mole-
42 cules, its cytoprotective activity at different concentrations was subsequently meas-
43 ured by the MTT assay. As shown in Fig. 4A, treatment with t-BHP induced about
44 70% cell death, but pretreatment of the cells with compound 3 (1.25, 2.5 and 5 µM)
45 protected does-dependently the cells from oxidative death. Since t-BHP can be me-
46 tabolized to free radical intermediates by cytochrome P-450, ultimately resulting in
7

ACCEPTED MANUSCRIPT
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
47 oxidative stress [22], the extent of ROS generation induced by t-BHP was monitored
48 by flow cytometry using DCFH-DA. Treatment with 900 µM t-BHP for 3 h resulted
49 in 2.5-fold increase in the ROS levels relative to the control. In contrast, compound 3
50 suppressed dose-dependently the ROS accumulation (Fig. 4B). Remarkably, the test
51 procedure, pretreatment with compound 3 followed by its removal, hints at the possi-
52 bility that this molecule can penetrate into the cells and function as an indirect anti-
53 oxidant to activate endogenous antioxidative defense system.
54
55 2.5. Compound 3 activated Nrf2 and its downstream genes in HepG2 cells
56 Under activation in response to oxidative and electrophilic stress, Nrf2 detaches
Figure 4 here
57 from Keap1, translocates to the nucleus and ultimately activates transcription of phase
58 II detoxifying enzymes [1]. Therefore, we next examined the expression of Nrf2 and
59 its downstream phase II detoxifying enzymes induced by compound 3. It can be seen
60 from Fig. 5A that after exposing the cells to 5 µM compound 3, both nuclear and cy-
61 tosolic levels of Nrf2 were increased as early as 1 h, to a peak value at 3–6 h, and then
62 gradually decreased to a basal level after 12 h. Additionally, 3 h treatment with com-
63 pound 3 under different concentrations (1.25, 2.5 and 5 µM) increased
64 dose-dependently the Nrf2 expression (Fig. 5B). Likewise, a time- and
65 dose-dependent effects could be also observed for the expression of phase II detoxi-
66 fying enzymes (including HO-1, GCLC and GCLM) induced by compound 3 (Fig. 5C
67 and D). Furthermore, tert-butyl hydroquinone (t-BHQ), a well-known Nrf2 activator,
8

ACCEPTED MANUSCRIPT
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
68 was employed as a positive control. When compared with t-BHQ, compound 3 was a
69 more effective cytoprotector and Nrf2 activator since only 5 µM compound 3 showed
70 a similar effect observed for 20µM t-BHQ in the cytoprotection (Fig. 5E) and Nrf2
71 activation (Fig. 5F) and its downstream gene expression (Fig. 5G).
72
73
Figure 5 here
To further confirm the cytoprotection of compound 3 toward HepG2 cells depend
74 on the Nrf2 activation and its downstream gene expression, the cytoprotection assay
75 was checked again in the presence of trigonelline (TRG) (an Nrf2 inhibitor) [23, 24],
76 Zinc protoporphyrin IX (ZnPP) (a HO-1 inhibitor) or L-buthionine-(S,R)-sulfoximine
77 (BSO) (a GCL inhibitor). Pretreatment with TRG, ZnPP and BSO attenuated obvi-
78 ously the cytoprotective activity of compound 3 (Fig. 6). These above results support
79 the conclusion that compound 3 ameliorates the t-BHP-induced death of HepG2 cells
80 through activating Nrf2, highlighting its role as an efficient Nrf2 activator and thereby
81 as an indirect antioxidant.
82
83 2.6 Compound 3 induced phosphorylation of Akt and JNK
84 It is known that MAPKs including JNKs and ERKs, as well as PI3K/Akt pathways
Figure 6 here
85 play a role in the Nrf2 transcriptional activation [5, 6]. The influence of compound 3
86 on the MAPKs and PI3K/Akt pathways was thus studied. It can be seen from Western
87 bolt results that treatment with compound 3 showed no benefit on phosphorylation of
88 ERK, but increased time-dependently that of p-Akt within 120 min. In addition,
9

ACCEPTED MANUSCRIPT
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
89 compound 3 stimulated a time-dependent increase of JNK phosphorylation and a
90 maximal response occurred at 30 min (Fig. 7A). In contrast, pretreatment with
91 SP600125 (a JNK inhibitor) or LY294002 (a specific PI3K/Akt inhibitor) inhibited
92 markedly the cytoprotection (Fig. 7B) and the expression of phase II detoxifying en-
93 zymes including HO-1, GCLC and GCLM (Fig. 7C) induced by compound 3. Base on
94 the above results, we can conclude that the cytoprotection of compound 3 is mediated,
95 at least in part, by a JNK and PI3K/Akt-dependent activation of Nrf2.
96
97 2.7 Compound 3 activated Nrf2 in a Keap1-dependent manner
98 Normally, Nrf2 is sequestered by the Keap1 protein in the cytoplasm. Once cells
Figure 7 here
99 are exposed to oxidative or electrophilic stress, the active cysteine residues in Keap1
00 are covalently modified, thereby facilitating the release of Nrf2 and its transcriptional
01 activation [25, 26]. Previous reports have shown that the Michael acceptors could di-
02 rectly conjugate with the active cysteine residues of Keap1, resulting in dissociation
03 and nuclear translocation of Nrf2 [4, 27]. Usually, the reaction can be quenched by
04 DTT, a sulfhydryl-containing Michael acceptor abrogating agent [28, 29]. To ascertain
05 whether compound 3 induces Nrf2 activation in Michael acceptor- and
06 Keap1-dependent manners, we tested initially the expression of Nrf2 by Western blot-
07 ting coupled with the use of DTT. The increased Nrf2 expression induced by com-
08 pound 3 can be reversed by pretreatment with DTT (Fig. 8A), implying that com-
09 pound 3 induces Nrf2 activation at least partially in a Michael acceptor-dependent
1
0

ACCEPTED MANUSCRIPT
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
10 manner. Furthermore, the decreased Keap1 band in response to the treatment of com-
11 pound 3 can be reversed obviously by DTT (Fig. 8B). This is in tune with the results
12 from Li and coworkers, showing that an increase in the expression of Nrf2 based on
13 exposure of 2-tert-butyl-1,4-hydroquinone was accompanied with a decreased levels
14 of Keap1 [30], and highlights that the Nrf2 activation is also Keap1-dependent.
15
16 2.8 Compound 3 enhanced Nrf2 stability
17 Under basal conditions, Nrf2 is constantly targeted for Keap1-dependent ubiquiti-
Figure 8 here
18 nation and subsequent proteasomal degradation to maintain its low levels in the cells
19 [26]. Molecules bearing Michael acceptor unit can react with the active cysteine resi-
20 dues of Keap1 and thus suppress Nrf2 ubiquitination and subsequent proteasomal
21 degradation [4, 27, 31]. To investigate whether compound 3 could stabilize Nrf2 lev-
22 els, the expression of total Nrf2 was tested by Western blotting. Compound 3 stimu-
23 lated a time-dependent expression of Nrf2 in HepG2 cells with a maximal response
24 within 3 h (Fig. 8C). Moreover, we quantified directly the turnover rate of Nrf2 by
25 using cycloheximide (a protein synthesis inhibitor) as a means to block protein syn-
26 thesis [32]. As shown in Fig. 8D, after exposing the cells to CHX, the total Nrf2 levels
27 were significantly decreased in an hour, whereas pretreatment with compound 3
28 slowed obviously the turnover rate. A single-phase decay model [32] to fit the data
29 revealed that pretreatment of compound 3 prolonged the half-life of Nrf2 protein from
30 22 to 57 min, approximately 2.6 times longer than the control group (Fig. 8E). These
1
1

ACCEPTED MANUSCRIPT
31 results demonstrate that compound 3 can significantly improve the Nrf2 stability
32 through suppressing its degradation.
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
33 2.9 Both the catechol and Michael acceptor moieties are important for the cytoprotec-
34 tion and Nrf2 activation
35
In the presence of oxygen and transition metal ions, catechol proelectrophiles can
36 be easily oxidized to its corresponding o-quinone electrophiles, which subsequently
37 undergo Michael addition with Keap1 and activate the Keap1-Nrf2/ARE pathway [4,
38 33]. Redox-active metal such as copper and iron are essential to all organisms living
39 and crucial for the function of several metalloenzymes involved in oxidation metabo-
40 lism [19, 33-35]. Our previous study has shown that a catechol-type resveratrol analog
41 could be converted intracellularly into its corresponding o-quinone by copper ions.
42 The oxidative conversion is required for its Nrf2 activation [19]. Therefore, effect of
43 various metal ion chelating agents on the cytoprotection of compound 3 was investi-
44 gated to clarify the mechanism by which compound 3 is oxidized, and to confirm
45 whether its corresponding o-quinone is responsible for the cytoprotection. As shown
46 in Fig. 9A, pretreatment with the wide variety of metal ion chelator ethylenedia-
47 minetetraacetic acid (EDTA) attenuated significantly the cytoprotection, suggesting a
48 role of metal ions involved in oxidation of compound 3. Furthermore, both the iron
49 chelator desferrioxamine (DFO) (Fig. 9A) and the copper chelator neocuproine (NEO)
50 (Fig. 9B) inhibited obviously its cytoprotection. These results demonstrate that both
51 the copper and iron ions could mediate oxidations of the catechol-type compound 3 to
1
2

ACCEPTED MANUSCRIPT
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
52 its corresponding electrophilic o-quinone, thereby contributing significantly to the
53 Nrf2-dependent cytoprotection. This is further supported by relatively weaker Nrf2
54 activation and expression of its downstream genes by the catechol-omitting analog
55 (compound 10) than by compound 3 (Fig. 9C).
56
Additionally, a comparison of cytoprotection (Fig. 3B) along with protein expres-
57 sion of Nrf2 (Fig. 9C) and its downstream genes (Fig. 9D) between compound 3 and
58 its reduced analog (3-H), clearly indicates that the Michael acceptor units are neces-
59 sary for the cytoprotection. To further clarify the role of the Michael acceptor moieties,
60 the effect of DTT on the compound 3-induced cytoprotection and Nrf2 activation was
61 evaluated. Pretreatment with DTT abrogated remarkably the cytoprotection (Fig. 9B)
62 and Nrf2 activation (Fig. 8A). Based on these data, we can infer that both the catechol
63 and Michael acceptor moieties are important for the Nrf2-dependent cytoprotection.
64
Figure 9 here
65
66 3. Discussion
67
Nrf2, a potential molecular target for cancer prevention, has attracted much at-
68 tention in discovering and developing natural product-derived chemopreventive
69 agents by targeting the Keap1-Nrf2/ARE pathway [4, 7]. Most of Nrf2 activators be-
70 long to the chemical class of electrophiles [7]. However, the intrinsic reactivity of
71 electrophiles is also associated with augmented metabolic degradation and toxicity,
72 due to reactions with nontarget proteins and other compounds such as glutathione [4,
1
3

ACCEPTED MANUSCRIPT
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
73 36]. Therefore, a proelectrophilic strategy to design Nrf2 activators by introduction of
74 the catechol or hydroquinone moiety is suggested because the proelectrophiles mani-
75 fest fewer side effects [37, 38]. Our previous study shows a catechol-type resveratrol
76 analog intracellularly converted into its o-quinone electrophile by copper ions, and the
77 copper-mediated oxidation is necessary to its Nrf2 activation and subsequent cyto-
78 protection [19]. Accordingly, we designed a series of natural curcumin (a potent Nrf2
79 activator [13, 39]) analogs by introduction of the geminal dimethyl groups and the
80 catechol moieties to find more effective Nrf2 activators based on a stability-increasing
81 and proelectrophilic strategy.
82
In tune with the previous result [16], introduction of the geminal dimethyl substit-
83 uents does improve the stability of curcumin (Fig. 2). Furthermore, a striking feature
84 of our data is that the catechol moieties can further stabilize the curcumin analogs as
85 exemplified in compound 3. We reasoned that the phenolic hydroxyl of compound 3
86 can readily undergo deprotonation to yield phenoxide anion under physiological pH
87 conditions, resulting in delocalization of negative charge over its cinnamoyl part and
88 subsequent prevention from attacking by H O.
2
89
Subsequently, this molecule turned out to be superior to other compounds in terms
90 of its increased stability and cytoprotective activity against the oxidative death.
91 Mechanism studies indicate that the excellent cytoprotection of compound 3 is at-
92 tributed to its facilitating nuclear translocation of Nrf2 (Fig. 5A and B) and subse-
93 quent expression of Nrf2-driven antioxidants including HO-1, GCLM, GCLC (Fig.
1
4

ACCEPTED MANUSCRIPT
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
94 5C and D). These results were further supported by using the Nrf2 inhibitor (TRG)
95 and HO-1, GCL inhibitor (ZnPP and BSO, respectively) (Fig. 6). Western blot results
96 indicate that compound 3 induces Nrf2 activation in a Keap1-dependent manner (Fig.
97 8B) and increases remarkably its stability (Fig. 8D). This allows Nrf2 to escape form
98 degradation and translocate to nuclear, resulting the Nrf2 target gene expression.
99 Moreover, the cytoprotection of compound 3 is mediated, at least in part, by JNK and
00 PI3K/Akt-dependent activation of Nrf2 (Fig. 7A), which is supported by the reduced
01 cytopretection in the presence of a JNK inhibitor (SP600125) or an Akt inhibitor
02 (LY294002) (Fig. 7B and C).
03
The essential question considered here is to clarify the role of the Michael acceptor
04 and catechol units for compound 3 that make it viable Nrf2 activator. The MTT data
05 (Fig. 3B) coupled with the Western blot results (Fig. 9C and D) suggest that the Mi-
06 chael acceptor units are crucial for maintaining its Nrf2-dependent cytoprotection be-
07 cause reduction of olefin (3-H) or pretreatment of DTT (Fig. 9B) lead to a significant
08 decrease in the potency. Furthermore, the oxidation of the catechol-type compound 3
09 by intracellular copper and iron ions is also required for its Nrf2-dependent cytopro-
10 tection. This could be explained by the attenuated cytoprotction by using NEO and
11 DFO (Fig. 9A and B) and the reduced potency in the case of compound 10 (Fig. 3B,
12 9C and D). it should be pointed out that this result distinguishes our previous result in
13 the case of a catechol-type resveratrol analog where the intracellular copper rather
o
14 than iron ions are responsible for its oxidation [19]. In fact, Cu(II)/Cu(I) (E = +0.15
1
5

ACCEPTED MANUSCRIPT
o
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
15 V) redox couple is relatively easier to favor redox cycle than the Fe(III)/Fe(II) (E =
16 +0.77 V) redox couple [40]. Therefore, involvement of iron ions should be com-
17 pound-dependent, probably due to the lower redox potential of compound 3 compared
18 with that of the catechol-type resveratrol analog. This is in concert with the previous
19 studies that iron ions catalyze the wine polyphenols containing catechol or pyrogallol
20 moieties to form o-quinones [41, 42]. Generally, the hydroxylated curcuminoids work
21 as antioxidants but as prooxidants under special conditions such as high concentra-
22 tions and in the presence of cupric ions [43]. In fact, this study highlights the fact that
23 catechol-type compound 3 makes use of endogenous copper and iron ions to construct
24 an efficient prooxidant system, resulting in formation of electrophilic o-quinone and
25 Nrf2-dependent cytoprotection. In other words, in the current case, compound 3 plays
26 finally a role of indirect antioxidant by its prooxidant activity.
27
28 4. Conclusion
29
This work provides useful information for using a stability-increasing and proelec-
30 trophilic strategy to develop a curcumin-inspired Nrf2 inducer. Based on this strategy,
31 the catechol-type compound 3 has been identified as a promising lead molecule in
32 terms of its increased stability and excellent cytoprotective activity. It could effec-
33 tively disrupt the Nrf2-Keap1 complex and thereby increase the expression of phase II
34 detoxifying enzymes in the Michael acceptor- and catechol-dependent manners (Fig.
35 10). Using copper and iron ion chelating agents, we verified that this molecule could
1
6

ACCEPTED MANUSCRIPT
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
36 make use of the two intracellular ions to produce its corresponding o-quinone, thereby
37 contributing to its Nrf2-dependent cytoprotection. In addition, the activation of the
38 JNK and PI3K pathways as well as stabilization of Nrf2 by this molecule was also
39 responsible for its Nrf2-dependent cytoprotection.
40
Figure 10 here
41
42 5. Experimental section
43 5.1 General information
44
Roswell Park Memorial Institute (RPMI)-1640, zinc protoporphyrin IX (ZnPP),
45 L-buthionine-(S,R)-sulfoximine (BSO), SP600125 and cycloheximide (CHX) were
46 obtained from Sigma (St. Louis, MO, USA). LY294002, Cell lysis buffer for Western
47 and IP, BCA protein assay kit, 2′,7′-Dichlorofluorescin diacetate (DCFHDA), 3-(4,
48 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and phenylme-
49 thanesulfonyl fluoride were from Beyotime Institute of Biotechnology (Jiangsu, Chi-
50 na). Keap1, β-actin, Nrf2 and Lamin A were purchased from Santa Cruz Biotechnol-
51 ogy (CA, USA). Rabbit polyclonal antibody to HO-1 was purchased from Enzo Life
52 Sciences (NY, USA). Rabbit antibodies against glutamate cysteine ligase catalytic
53 subunit (GCLC) and glutamate cysteine ligase modulatory subunit (GCLM) were
54 purchased from Abcam (Shanghai, China). All the phosphorylated, nonphosphory-
55 lated kinase antibodies used in this study and glyceraldehyde 3-phosphate dehydro-
56 genase (GAPDH) were obtained from Cell Signaling Technology (Danvers, MA,
1
7

ACCEPTED MANUSCRIPT
57 USA). HRP-labeled secondary antibody was obtained from TransGen Biotech Co.,
58 Ltd. (Beijing, China). All other chemicals were of the highest quality available.
59 5.2 Synthesis
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
60
Synthesis of curcumin and its designed analogs was conducted according to the
61 route described in Scheme 1 and the details are described as below. Their structures
1
13
62 were confirmed by H and C NMR spectra and ESI-MS analysis (See also the sup-
63 plementary data).
64 5.2.1. TBS protection of hydroxyl groups [44]
65
Tert-butyldimethylsilyl chloride (TBSCl) (1.2 eq) and imidazol (3 eq) were added
66 to a solution of isovanillin or 3,4-dihydroxybenzaldehyde in CH Cl (0.25 M) at 0 °C.
2
2
67 The mixture was warmed to room temperature and stirred for 6 h. The solution was
68 washed with saturated Na CO brine and dried over Na SO and concentrated under
2
3,
2
4,
69 reduced pressure to afford the crude product as a white solid which was used to next
70 step directly.
71 5.2.2. MOM protection of hydroxyl groups [45]
72
To a solution of 2-hydroxybenzaldehyde (1.0 eq) in anhydrous THF (0.25 M) was
73 added a suspension of 60% w/w NaH (1.2 eq) in THF, the mixture was stirred at room
74 temperature for 1 h. Then chloromethyl methyl ether (MOMCl) (1.2 eq) was added,
75 and the mixture was stirred at room temperature for 2 h followed by quenching and
76 extraction with ethyl acetate. The organic layer was washed with brine, 5% NaOH,
77 dried over Na SO and concentrated under reduced pressure to afford the crude prod-
2
4,
1
8

ACCEPTED MANUSCRIPT
78 uct as a white solid which was used to next step directly.
79 5.2.3.General procedure for synthesis of intermediates 1, 3 and 4 [20]
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
80
To a solution of acetyl acetone (2 eq) in ethyl acetate was added boric anhydride
81 (1.2 eq) and tributyl borate (4.8 eq). The mixture was heated to 50 °C and stirred for
82 30 min. Then a solution of appropriate benzaldehyde (4 eq) in ethyl acetate was added,
83 followed by slow addition of n-butylamine (4 eq). After refluxing for 6 h, the solution
84 was cooled to room temperature and diluted with 1 N HCl. The yellow-orange solid
85 was suspended in water, filtered, and dried under vacuum. All crude compounds were
86 recrystallized in ethyl acetate to give the pure products.
87 5.2.4. General procedure for synthesis of compounds 2, 4-8 and 10 [20]
88
To a stirring solution of intermediate 3 (1.0 eq) in CH Cl (0.25 M) was added a
2 2
89 mixture of NaOH (2.4 eq) and tetrabutylammonium chloride in water under ice bath.
90 The mixture was warmed to room temperature and stirred for 30 min followed by ad-
91 dition of methyl iodide (3.6 eq). After refluxing for 2 h, the solution was cooled to
92 room temperature and poured into water, adjusted pH to 4~5 with 1 N HCl and ex-
93 tracted with CH Cl . The organic layer was washed with brine, dried over Na SO ,
2
2
2
4
94 and concentrated under reduced pressure. The residue was purified by silica gel col-
95 umn chromatography petroleum ether/acetone (3/1, v/v) to afford the target molecule.
96 (1E, 6E)-1, 7-bis(3, 4-dimethoxyphenyl)-4, 4-dimethylhepta-1,6-diene-3, 5-dione (2).
97 Compound 2 was prepared according to the general procedure, using (1E, 4Z, 6E)-1,
98 7-bis(3, 4-dimethoxyphenyl)-5-hydroxyhepta-1, 4, 6-trien-3-one (396.4 mg, 1.0 mmol)
1
9

ACCEPTED MANUSCRIPT
1
3
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
99 to obtain pure compound 2 as a yellow solid (186.8 mg, yield 44%); mp 69-71 °C; H
00 NMR (400 MHz, Acetone-d ): δ = 7.63 (d, J = 15.6 Hz, 2H), 7.34 (d, J = 2.0 Hz, 2H),
6
01 7.25 (dd, J = 8.0 Hz, 2.0 Hz, 2H), 6.98 (d, J = 8.0 Hz, 2H), 6.92 (d, J = 15.6 Hz, 2H),
13
02 3.85 (s, 6H),3.83 (s, 6H), 1.42 (s, 6H); C NMR (100 MHz, Acetone-d )：δ = 197.5,
6
03 152.0, 149.6, 143.4, 127.4, 123.4, 119.6, 111.4, 110.6, 56.0, 55.2, 55.2, 20.6; MS
+
04 (ESI): (m/z) = 424.9 [M + H] .
05 (1E, 6E)-4, 4-dimethyl-1, 7-bis(3, 4, 5-trimethoxyphenyl)hepta-1, 6-diene-3, 5-dione
06 (4). Compound 4 was prepared according to the general procedure, using (1E, 4Z,
07 6E)-5-hydroxy-1, 7-bis(3, 4, 5-trimethoxyphenyl)hepta-1, 4, 6-trien-3-one (456.5 mg,
08 1.0 mmol) to obtain pure compound 4 as a yellow solid (266.5 mg, yield 51%); mp
1
09 83-85 °C; H NMR (400 MHz, Acetone-d ): δ = 7.62 (d, J = 15.6 Hz, 2H),7.06 (s,
6
1
3
10 4H), 7.00 (d, J = 15.6 Hz, 2H), 3.85 (s, 12H), 3.75 (s, 6H), 1.43 (s, 6H); C NMR
11 (100 MHz, Acetone-d )：δ = 197.5, 153.7, 143.5, 140.7, 130.0, 121.1, 106.2, 60.0,
6
+
12 59.7, 55.6, 20.5; MS (ESI): (m/z) = 484.4 [M + H] .
13 (1E, 6E)-1, 7-bis(3-methoxyphenyl)-4, 4-dimethylhepta-1, 6-diene-3, 5-dione (5).
14 Compound 5 was prepared according to the general procedure, using (1E, 4Z,
15 6E)-5-hydroxy-1, 7-bis(3-methoxyphenyl)hepta-1, 4, 6-trien-3-one (336.1 mg, 1.0
16 mmol) to obtain pure compound 5 as a yellow solid (160.3 mg, yield 44%); mp
1
17 63-65 °C; H NMR (400 MHz, Acetone-d ): δ = 7.66 (d, J = 15.6 Hz, 2H), 7.33 (t, J
6
18 = 7.6 Hz, 2H), 7.29-7.27 (m, 4H), 7.08 (d, J = 15.6 Hz, 2H),7.00 (ddd, J = 7.6, 2.4, 1.2
1
3
19 Hz, 2H), 3.82 (s, 6H), 1.47 (s, 6H); C NMR (100 MHz, Acetone-d )：δ = 197.7,
6
2
0

ACCEPTED MANUSCRIPT
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
20 160.2, 143.1, 136.0, 129.9, 122.2, 121.1, 116.6, 113.3, 60.1, 54.8, 20.4; MS (ESI):
+
21 (m/z) = 365.0 [M + H] .
22
(1E,6E)-1,7-bis(2,6-dimethoxyphenyl)-4,4-dimethylhepta-1,6-diene-3,5-dione (6).
23 Compound 6 was prepared according to the general procedure, using
24 2,6-methoxybenzaldehyde (396.4 mg, 1.0 mmol) to obtain pure compound 6 as a yel-
1
25 low oil (199.5 mg, yield 47%); mp 165-167 °C; H NMR (400 MHz, CDCl ): δ = 8.15
3
26 (d, J = 16 Hz, 2H), 7.28-7.20 (m, 6H), 6.49 (d, J = 8.4 Hz, 4H), 3.81 (s, 12H),1.44 (s,
1
3
27 6H); C NMR (100 MHz, CDCl )：δ= 200.2, 160.4, 134.0, 131.3, 125.3, 112.7, 103.6,
3
+
28 60.7, 55.8, 21.3; MS (ESI): (m/z) = 425.0 [M + H] .
29 (1E, 6E)-1, 7-bis(2-methoxyphenyl)-4, 4-dimethylhepta-1, 6-diene-3, 5-dione (7).
30 Compound 7 was prepared according to the general procedure, using (1E, 4Z,
31 6E)-5-hydroxy-1, 7-bis(2-methoxyphenyl)hepta-1, 4, 6-trien-3-one (336.4 mg, 1.0
32 mmol) to obtain pure compound 7 as a yellow solid (156.7 mg, yield 43%); mp
1
33 124-125 °C; H NMR (400 MHz, Acetone-d ): δ = 8.03 (d, J = 15.6 Hz, 2H), 7,69 (dd,
6
34 J = 7.6 Hz, 1.6 Hz, 2H), 7.41 (ddd, J = 8.4, 7.6, 1.6 Hz, 2H), 7.07 (d, J = 8.4 Hz, 2H),
1
3
35 7.05 (d, J = 15.6 Hz, 2H), 6.97 (t, J = 7.6 Hz, 2H), 3.91 (s, 6H), 1.45 (s, 6H); C
36 NMR (100 MHz, Acetone-d )：δ = 197.8, 158.8, 138.0, 132.1, 128.6, 123.0, 122.2,
6
+
37 120.7, 111.5, 60.2, 55.1, 20.5; MS (ESI): (m/z) = 365.1 [M + H] .
38 (1E, 6E)-4, 4-dimethyl-1, 7-bis(2-(trifluoromethyl)phenyl)hepta-1, 6-diene-3, 5-dione
39 (8). Compound 8 was prepared according to the general procedure, using (1E, 4Z,
40 6E)-5-hydroxy-1, 7-bis(2-(trifluoromethyl)phenyl)hepta-1, 4, 6-trien-3-one (412.3 mg,
2
1

ACCEPTED MANUSCRIPT
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
41 1.0 mmol) to obtain pure compound 8 as a yellow solid (140.9 mg, yield 32%); mp
1
42 61-63 °C; H NMR (400 MHz, Acetone-d ): δ = 8.06 (d, J = 7.6 Hz, 2H), 8.02 (d, J =
6
43 15.6 Hz, 2H), 7.82 (d, J = 7.6 Hz, 2H), 7.72 (t, J = 7.6 Hz, 2H),7.65 (t, J = 7.6 Hz, 2H),
1
3
44 7.19 (d, J = 15.6 Hz, 2H), 1.54 (s, 6H); C NMR (100 MHz, Acetone-d )：δ = 197.54,
6
+
45 137.75, 137.73; MS (ESI): (m/z) = 440.9 [M + H] .
46 (1E, 6E)-4, 4-dimethyl-1, 7-diphenylhepta-1, 6-diene-3, 5-dione (10). Compound 10
47 was prepared according to the general procedure, using (1E, 4Z,
48 6E)-5-hydroxy-1,7-diphenylhepta-1, 4, 6-trien-3-one (552.6 mg, 2.0 mmol) to obtain
1
49 pure compound 10 as a yellow solid (182.6 mg yield 30%); mp 51-53 °C; H NMR
50 (400 MHz, Acetone-d ): δ = 7.72-7.68 (m, 6 H), 7.44-7.40 (m, 6H), 7.08 (d, J = 15.6
6
1
3
51 Hz, 2H), 1.49 (s, 6H); C NMR (100 MHz, Acetone-d )：δ = 197.7, 143.1, 134.6,
6
+
52 130.6, 128.9, 128.6, 121.9, 60.2, 20.4; MS (ESI): (m/z) = 305.9 [M + H] .
53 5.2.5. General procedure for synthesis of compounds 1 and 3 [43]
54
To a stirring solution of TBS-protected compound 1 or 3 in THF was added drop-
55 wise a solution of tetrabutylammonium fluoride (1M) in THF at 0 °C. Once finished,
56 the mixture was warmed to room temperature for 1 h. The mixture was poured into
57 ice water, adjusted pH to 3~4 with 1 N HCl, then extracted with ethyl acetate. The
58 organic layer was washed with brine, dried over Na SO and concentrated under re-
2
4,
59 duced pressure. The residue was purified by silica gel column chromatography with
60 petroleum ether/acetone (3/1, v/v) to afford the product.
61 (1E,
6E)-1,7-bis(4-hydroxy-3-methoxyphenyl)-4,4-dimethylhepta-1,
6-diene-3,
2
2

ACCEPTED MANUSCRIPT
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
62 5-dione (1). Compound 1 was prepared according to the general procedure, using (1E,
63 4Z, 6E)-1, 7-bis(4-((tert-butyldimethylsilyl)oxy)-3-methoxyphenyl)-5-hydroxyhepta-1,
64 4, 6-trien-3-one (1.194 g, 2.0 mmol) to obtain pure compound 1 as a yellow solid
1
65 (261.6 mg, yield 33%); mp 142-144 °C; H NMR (400 MHz, Acetone-d ): δ = 8.28
6
66 (s, 1H), 7,62 (d, J = 15.6 Hz, 2H), 7.35 (d, J = 2.0 Hz, 2H), 7.19 (dd, J = 8.0 Hz, 2.0
67 Hz, 2H), 6.87 (d, J = 15.6 Hz, 2H), 6.86 (d, J = 8.0 Hz, 2H), 3.87 (s, 6H), 1.42 (s, 6H);
1
3
68
C NMR (100 MHz, Acetone-d )：δ = 197.4, 149.5, 147.8, 143.6, 126.6, 123.7, 119.1,
6
+
69 115.2, 111.0, 60.0, 55.4, 20.6; MS (ESI): (m/z) = 397.0 [M + H] .
70 (1E, 6E)-1,7-bis(3, 4-dihydroxyphenyl)-4, 4-dimethylhepta-1, 6-diene-3, 5-dione (3).
71 Compound 3 was prepared according to the general procedure, using (1E, 4Z, 6E)-1,
72 7-bis(3, 4-bis((tert-butyldimethylsilyl)oxy)phenyl)-5-hydroxyhepta-1, 4, 6-trien-3-one
73 (797.4 mg, 1.0 mmol) to obtain pure compound 3 as a yellow solid (84.7 mg, yield
1
74 23%); mp 191-193 °C; H NMR (400 MHz, Acetone-d ): δ = 8.36 (s, 4H), 7.56 (d, J =
6
75 15.6 Hz, 2H), 7.16 (d, J = 2.0 Hz, 2H), 7.07 (dd, J = 8.0 Hz, 2.0 Hz, 2H), 6.85 (d, J =
1
3
76 8.0 Hz, 2H), 6.76 (d, J = 15.6 Hz, 2H), 1.42 (s, 6H); C NMR (100 MHz, Ace-
77 tone-d )：δ = 197.4, 148.3, 145.4, 143.5, 126.7, 122.5, 118.9, 115.5, 114.5, 60.0, 20.6;
6
+
78 MS (ESI): (m/z) = 369.0 [M + H] .
79 5.2.6. General procedure for synthesis of compound 9 [45]
80
To a solution of MOM-protected compound 9 in MeOH (0.25 M) was added conc.
81 HCl (0.1 M) over 10 min in ice bath. Then the mixture was warmed to room temper-
82 ature, stirred for 2 h and concentrated under reduced pressure. The crude product was
2
3

ACCEPTED MANUSCRIPT
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
5
5
5
5
83 further purified by silica gel column chromatography with petroleum ether /acetone
84 (3/1, v/v) to give compound 9 as a yellow solid.
85 (1E, 6E)-1, 7-bis(2-hydroxyphenyl)-4, 4-dimethylhepta-1, 6-diene-3, 5-dione (9).
86 Compound 9 was prepared according to the general procedure, using (1E, 4Z,
87 6E)-5-hydroxy-1, 7-bis(2-(methoxymethoxy)phenyl)hepta-1, 4, 6-trien-3-one (396.4
88 mg, 1.0 mmol) to obtain pure compound 9 as a yellow solid (117.7 mg, yield 35%);
1
89 mp 161-163 °C; H NMR (400 MHz, Acetone-d ): δ = 9.21 (s, 2H), 8.04 (d, J = 15.6
6
90 Hz, 2H), 7.62 (dd, J = 8.0 Hz, 1.6 Hz, 2H), 7.25 (td, J = 8.0, 16 Hz, 2H), 7.11 (d, J =
1
3
91 15.6 Hz, 2H), 6.95 (d, J = 8.0 Hz, 2H), 6.87 (t, J = 7.6 Hz, 2H), 1.45 (s, 6H); C
92 NMR (100 MHz, Acetone-d )：δ = 197.9, 157.1, 138.6, 131.8, 129.2, 121.8, 121.5,
6
+
93 120.0, 116.2, 60.2, 20.6; MS (ESI): (m/z) = 337.2 [M + H] .
94 5.2.7. General procedure for synthesis of compound 3-H [46].
95
A solution of compound 3 in methanol was added to a mixture of catalytic amount
96 Pd/C (10%) in methanol under hydrogen atmosphere. Then the reaction mixture was
97 stirred at room temperature for 8 h. The mixture was filtrated and the filtrate was
98 evaporated. The residue was further purified by silica gel column chromatography
99 with ethyl acetate/petroleum ether (5/1, v/v) to afford compound 3-H as a yellow oil.
00 1,7-bis(3,4-dihydroxyphenyl)-4,4-dimethylheptane-3,5-dione (3-H).
01
Compound 3-H was prepared according to the general procedure, using
02 (1E,6E)-1,7-bis(3,4-dihydroxyphenyl)-4,4-dimethylhepta-1,6-diene-3,5-dione (368.4
1
03 g, 1.0 mmol) to obtain pure compound 3-H as a yellow oil. (52.6 mg, yield 14%); H
2
4

ACCEPTED MANUSCRIPT
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
04 NMR (400 MHz, Acetone-d ): δ = 7.91 (s, 4H), 6.70 (d, J = 8.0 Hz, 2H), 6.64 (d, J =
6
13
05 2.0 Hz, 2H), 6.48 (dd, J = 8.0 Hz, 2.0 Hz, 2H), 2.63 (s, 8H), 1.24 (s, 6H); C NMR
06 (150 MHz, Acetone-d )：δ = 208.7, 144.9, 143.2, 132.9, 119.5, 115.4, 115.1, 61.7, 40.2,
6
-
07 20.5; MS (ESI): (m/z) = 371.2 [M - H] .
08 5.3 Stability assay
09
The stability of curcumin and its analogs (40 µM) were assessed by monitoring
10 their UV-visible absorption changes (interval = 2 min) in phosphate buffer solution
11 (PBS) (pH 7.4) at 25 °C by a TU-1901 UV/Vis spectrophotometer (Beijing Purkinje
12 General Instrument Co., Ltd., Beijing, China), Spectra were run against blanks con-
13 taining the solution and DMSO.
14 5.4 Cell culture
15
Human hepatoma cell line HepG2 were purchased from the Shanghai Institute of
16 Biochemistry and Cell Biology, Chinese Academy of Sciences and cultivated in 1640
17 medium supplemented with NaHCO (2 g/L), fetal bovine serum (10%, v/v), penicil-
3
18 lin (100 kU/L) and streptomycin(100 kU/L) at 37 °C in a humidified atmosphere
19 with 95% air and 5% CO2.
20 5.5 Cell viability assay
3
21
The cytotoxicity was measured by the MTT assay. Briefly, HepG2 cells ((5 × 10
22 cells/well) were seeded in 96-well plates and cultured for 24 h, treated with indicated
23 concentrations of test compounds for 24 h, followed by its removal and incubation
24 with 900 µM t-BHP for another 6 h at 37 °C. Then a solution of MTT (0.5 mg/mL) in
2
5

ACCEPTED MANUSCRIPT
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
25 fresh medium was added to each well and incubated in a CO incubator for further 4 h.
2
26 DMSO (100 µL) was added to dissolve formazan crystals, and the absorbance was
27 analyzed in a multiwall-plate reader (Bio-Rad M680) at 570 nm. In the case of inhibi-
28 tion, the cells were pretreated with TRG (100 nM), ZnPP (3 µM), BSO (10 µM),
29 LY294002 (15 µM), DTT (400 µM), SP600125 (25 µM), LY294002 (15µM), NEO
30 (100 nM), DFO (100 µM) or EDTA (500 µM) for 1 h before the addition of the test
31 compounds.
32 5.6 Measurement for the intracellular ROS levels
5
33
HepG2 cells (4 × 10 cells/well) were incubated with the test compounds (1.25, 2.5
34 or 5 µM) or control (DMSO) for 24 h followed by exposing cells to 900µM t-BHP for
35 3 h. Then the cells were collected, stained with DCFHA-DA (3 µM) for 30 min at
36 37 °C in the dark, and washed with PBS. The relative fluorescent intensities of FITC
37 were analyzed immediately using a FACSCanto flow cytometer (Becton-Dickinson,
38 San Jose, CA, USA). ROS levels of treatment groups were normalized to the relative
39 control group.
40 5.7 Western Blot Analysis
6
41
HepG2 cells (2 × 10 cells/well) were treated with the test compounds for the indi-
42 cated concentrations and time points, the total protein was lysed and collected by us-
43 ing Cell lysis buffer. Nuclear and cytoplasmic protein extracts were prepared using a
44 Nuclear and Cytoplasmic Protein Extraction Kit (Viagene Biotech, Zhejiang, China)
45 according to the manufacturer’s protocol. Concentrations of sample proteins were de-
2
6

ACCEPTED MANUSCRIPT
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
46 tected by BCA protein assay kit. For western blot analysis, equal amounts (40 µg/lane)
47 of proteins were resolved by 12% SDS-PAGE and electroblotted onto nitrocellulose
48 membranes (Bio-Rad, Hercules, CA) at 4 °C. The membranes were blocked with 5%
49 defatted milk for 1 h at room temperature, washed, and incubated with appropriate
50 specific antibody at 4 °C overnight followed by incubation with the corresponding
51 horseradish peroxidase conjugated secondary antibodies (1/5000) in TBST for another
52 1 h at room temperature. After washing, the protein bands were detected with an en-
53 hanced chemiluminescence (ECL) Western blot detection kit using an enhanced Im-
54 ageQuant chemiluminescence system. In the case of inhibition, the cells were pre-
55 treated with LY294002 (15 µM), SP600125 (25 µM), DTT (400 µM) and CHX (5
56 µg/mL) for 1 h before the addition of the test compounds.
57 5.8 Assay for Nrf2 half-life
58
To measure the half-life of Nrf2, the cells were pretreated with or without 5 µM
59 compound 3 for 3 h, followed by treatment with 5 µg/mL CHX (a protein synthesis
60 inhibitor) for the indicated periods. Total cell lysates were collected and subjected to
61 Western blot analysis with an Nrf2 antibody. The densitometric intensity of total Nrf2
62 acquired by blots was performed by using the Image J software (NIH, Bethesda, MD)
63 and plotted against the CHX treatment time to calculate the half-life of Nrf2 [19].
64 5.9 Statistical analysis
65
Data are expressed as the mean ± SD. Statistical comparisons among the results
66 were performed using analysis of variance. Significant differences (P < 0.05) between
2
7

ACCEPTED MANUSCRIPT
67 the means of two groups were analyzed by Student’s t test.
68
5
5
5
5
5
5
69 Acknowledgements
70 This work was supported by the National Natural Science Foundation of China (Grant
71 Nos. 21372109 and 31100607), and Gansu key technologies R & D program
72 (143NKDA028).
573 Conflict of Interest
5
5
5
5
5
74 The authors declare no conflict of interest.
75
76 Appendix A. Supplementary data
77 Supplementary data associated with this article can be found, in the online version, at
78
2
8