Chem Biol Drug Des 2010; 75: 475–480
ª 2010 John Wiley & Sons A/S doi: 10.1111/j.1747-0285.2010.00955.x
Research Article
Identification and Characterization of Novel Nrf2 Inducers Designed to Target the Intervening Region of Keap1 Jian H. Wu1,2, Weimin Miao1,3, Liang G. Hu1 and Gerald Batist1,2,* 1
Montreal Centre for Experimental Therapeutics in Cancer, Segal Cancer Center; Lady Davis Institute for Medical Research, Sir Mortimer B. Davis-Jewish General Hospital, McGill University, 3755 Cote-Ste-Catherine, Rd., Montreal, QC H3T 1E2, Canada 2 Department of Oncology, McGill University, 3755 CoteSte-Catherine, Rd., Montreal, QC H3T 1E2, Canada 3 The University of Tennessee Graduate School of Medicine, 1924 Alcoa Highway, Knoxville, TN 37920, USA *Corresponding author: Gerald Batist,
[email protected] Transcription factor Nrf2 regulates a battery of genes encoding detoxifying enzymes. Under basal conditions, Nrf2 is sequestered in the cytoplasm by a protein known as Keap1. In response to oxidative stress, Keap1-mediated ubiquitination of Nrf2 is decreased significantly and the Nrf2 pathway is turned on. Residues C273 and C288 at the intervening region (IVR) domain of Keap1 are necessary for Keap1 to repress Nrf2, indicating a critical role of the IVR domain in the functional interaction of Keap1 with Nrf2. To identify chemical modulator targeting the IVR domain of Keap1, we built a 3D structural model of the Keap1 IVR domain and demonstrated this structural model is effective in retrieving novel Nrf2 inducers from chemical databases, BM10, 31, and 40 increase concentration of nuclear Nrf2, with a potency comparable to that of sulforaphane. We showed C297S mutation partially abolished the Nrf2inducing effect of BM31, suggesting BM31 may target C297 in the IVR domain of Keap1. Further, BM31 and BM40 potently induce expression of ARE-regulated enzyme gamma-glutamylcysteine synthetase. We demonstrated that BM31 provides protections for the MCF-7 cells from cytotoxic damage of carcinogen benzo[a]pyrene. Key words: chemoprevention, screening
Keap1,
Nrf2
inducers,
virtual
Abbreviations: 3D, three-dimensional; ARE, Antioxidant response element; B[a]P, Benzo[a]pyrene; EQ, Ethoxyquin; GCS, c-glutamylcysteine synthetase; IVR, Intervening region; SFN, sulforaphane. Received 8 August 2009; revised 1 January 2010 and accepted for publication 28 January 2010
As chemical carcinogenesis likely plays a role in cancer development, cytoprotective enzyme induction is believed to be an important means of cancer chemoprevention. Transcription factor Nrf2 regulates a battery of genes encoding carcinogen-detoxifying enzymes and antioxidant proteins by binding to the antioxidant response element (ARE) promoter regulatory sequence. Under basal conditions, in which the redox homeostasis is maintained in cells, Nrf2 is sequestered in the cytoplasm by a protein known as Keap1, which targets Nrf2 for ubiquitination and degradation by the proteasome, and thus controls both the subcellular localization and steady-state levels of Nrf2. In response to oxidative stress or chemopreventive compounds, Keap1-mediated ubiquitination of Nrf2 is decreased significantly and the Nrf2 pathway is turned on. Thus, Keap1 is a molecular switch that senses various stimuli and turns the Nrf2 pathway on and off (1,2). Administration of Nrf2-inducing agents has been shown to result in decreased carcinogenesis in animal models and altered carcinogen metabolism in humans (1). Clinical interventions have shown that Nrf2 inducers increase cytoprotective enzyme expression, resulting in modulation of aflatoxin disposition (3). Keap1 contains the BTB domain mediating Keap1 homodimer formation, the 'intervening region' (IVR) (amino acids 180–314), and the Kelch repeat domain that mediates binding to the Neh2 domain of Nrf2 (4). Human Keap1 contains 27 cysteine residues. Three key cysteine residues (C151, C273, and C288) have been identified. C151 is required for several Nrf2 inducers, such as sulforaphane (SFN) and tert-butylhydroquinone, to manifest their effect. Importantly, residues C273 and C288 at the IVR domain are necessary for Keap1 to repress Nrf2. A single cysteine to serine mutation C273S or C288S render Keap1 unable to repress Nrf2 (5,6). The transgenic expression of mutant Keap1(C273A) and ⁄ or Keap1(C288A) protein in Keap1 null mice failed to reverse constitutive Nrf2 activation, indicating that cysteine residues at positions 273 and 288 are essential for Keap1 to repress Nrf2 activity in vivo (7). This suggests a critical role of the IVR domain in the functional interaction of Keap1 with Nrf2. It appears that identification of small molecules to interact with the IVR domain represents an attractive strategy for the induction of Nrf2 and the cytoprotective enzymes. In the present work, we combined in silico screening with cellbased assays to identify novel agents that modulate the Keap1– Nrf2 signaling pathway. We built a three-dimensional (3D) structural model of the Keap1 IVR domain and utilized the 3D model to virtually screen the chemical databases for putative Nrf2 inducers, which were subsequently verified by luciferase-based reporter gene 475
Wu et al.
assays and Western blot analyses. Among the novel Nrf2 inducers, we identified BM10, BM31, and BM40 triggered Nrf2 nuclear translocation and induced expression of ARE-regulated enzymes, such as c-glutamylcysteine synthetase (GCS). We showed that BM31 in the low dose range (50–200 nM) provides protections for the MCF-7 cells from cytotoxic damage of carcinogen benzo[a]pyrene (B[a]P).
Materials and Methods Construction of the 3D model of Keap1 IVR and virtual screening Using a novel fold recognition technique as implemented in program Prospect (Oak Ridge National Laboratory, Oak Ridge, TN, USA) (8), we have identified the crystal structure of FadR (PDB entry: 1e2x), a fatty acid-responsive transcription factor, as a structure template for the construction of the 3D model of Keap1 IVR (residues 180–314). The percentage of the sequence identify and strong similarity between Keap1 IVR (residues 180–314) and the template (1e2x, residues 25– 225) is 33.65%. Based on this template, 100 structural models of Keap1 IVR were generated using software MODELLER v8 (9) and the model with lowest objective function was selected for further study. Docking of ethoxyquin (EQ) and virtual screening of chemical database were performed using software GOLD v3.0 (The Cambridge Crystallographic Data Centre, Cambridge, UK) (10). The ligand was treated as fully flexible whereas the protein was kept rigid except that each serine, threonine, and tyrosine hydroxyl group was allowed to rotate to optimize hydrogen bonding. The scoring function GoldScore and the GOLD standard default parameter setting were selected.
Chemicals and antibodies EQ was purchased from Sigma–Aldrich. SFN was obtained from LKT laboratory. All BM compounds were obtained from the Drug Synthesis and Chemistry Branch, Developmental Therapeutics Program, NCI, Bethesda, MD, USA. The Nrf2, Keap1, and GCS antibody were from Santa Cruz Biotechnology (Santa Cruz, CA, USA).
Cell culture The HepG2, MCF-7, and 293T cell lines were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). The HepG2 cells were grown in alpha-minimal essential media supplemented with non-essential amino acids, sodium pyruvate, 90% Earle's balanced salt solution, and 10% fetal bovine serum. MCF-7 cells were maintained in RPMI-1640 medium. The 293T cells were grown in DMEM supplemented with 10% fetal bovine serum.
Luciferase assay One representative ARE (A2-ARE) from GSTA2 was used in this work, and A2-ARE were cloned into the KpnI and MluI sites of pGL3-promoter vector, which contains a heterologous SV40 promoter and a firefly luciferase reporter gene. Cells were seeded at 1 · 105 per well using 24-well plate, grown overnight before being transfect476
ed with pGL3 and other designated constructs using LipofectAMINE (Invitrogen) for 5 h. Plasmids expressing Keap1 mutants (C257S, C273S, C288S, and C297S) were generously provided by Dr Mark Hannink (5). The plasmid pRL, containing a renilla reniformis luciferase reporter gene, was cotransfected as internal control to correct transfection efficiency. After transfection, cells were treated with test compounds at designated concentrations for 24 h prior to harvesting. The cells were then washed twice with phosphate-buffered saline and harvested in 100 lL of 1· passive lysis buffer (Promega, Madison, WI, USA). The luciferase activities were analyzed in 20-lL cell extracts with the Dual luciferase assay kit (Promega) on a Lumat LB 9507 luminometer (Berthold Technologies, Oak Ridge, TN, USA). The related luciferase activities are expressed as a ratio of the pGL3 reporter activity to that of the control plasmid pRL. The fold of induction is expressed as ratio of induction from the treated cells versus the untreated. Values represent measurements from three independent experiments performed in triplicate and are presented as the mean € SEM.
Western blot analysis Cells were treated with test compounds at designated concentrations and were harvested at 24 h. Sixty microgram of cell-lysate protein and 35 lg of nuclear extract were loaded into each lane, and electrophoresed through a 4–20% gradient SDS–PAGE gel (Bio-Rad, Mississauga, ON, Canada), and transferred onto a nitrocellulose membrane. The membrane was incubated with primary antibody and subsequently incubated with HRP-conjugated secondary antibody. Finally, ECL detection was carried out, and the results were recorded on X-ray film by autophotography.
Cell Viability and capability of compounds in protecting cells from carcinogen Cells were seeded at 4 · 103 per well using 96-well plate, grown overnight before being treated with test compounds for 72 h. Viable cells were measured by MTT assays. To evaluate the ability of test compounds in protecting MCF-7 cells from the damage of B[a]P, MCF-7 cells were seeded at 2 · 103 per well using 96-well plate, grown overnight before being treated with test compounds at designed concentration for 48 h. Next, cells were treated with 160 nM B[a]P in the presence of test compound for another 72 h. The viable cells were measured using MTT assays. Experiments were performed in triplicate and repeated twice.
Statistical analysis GRAPHPAD PRISM 5 software (San Diego, CA, USA) was used to perform statistical analysis by t-test. p < 0.05 is considered statistical significant.
Results and Discussion Structural model of Keap1 IVR domain and virtual screening Residues C257, C273, C288, and C297, all of which are in IVR region, have been proposed as the direct sensors of inducers of the Chem Biol Drug Des 2010; 75: 475–480
Identification of Novel Nrf2 Inducers
Next, the Keap1 IVR ⁄ EQ complex was employed to virtually screen 90 000 compounds extracted from the NCI-3D chemical database (NCI, US) for putative Nrf2 inducers. The chemical structures and the predicted binding modes for each of the 50 top-score compounds were visually inspected. We have adapted a multiple-cycle screening strategy (12). In the first cycle, we obtained chemical samples for 23 top-score compounds. Of the 23 compounds, we identified seven hits using ARE-driven luciferase assays. In the second cycle, we screened the NCI-3D database for the analogues of the initial hits by a ligand-based approach using the tanimoto index (with a cutoff of 0.9). We obtained the chemical samples of 17 analogues of the hits from the first cycle and then subjected them to cell-based assays. The two cycles of screenings led to identification of 9 novel Nrf-2 inducers (Figure 2).
Figure 1: Structural model of the intervening region (IVR, amino acids 180–314) of human Keap1. Residues Cys-297 and Cys-288 (red surface) are inside the cavities, whereas Cys-273 (red surface) and Cys-257 (on the other side of the surface, not shown) are on the surface but not inside a cavity. Ethoxyquin (in green sticks) was docked into the surface cavity that includes Cys-297.
ARE-regulated enzymes (11). According to the 3D IVR model, residues C288 and C297 were found to be inside surface cavities whereas C273 and C257 are on the surface, but they are not inside a cavity (Figure 1). The Keap1 IVR model suggested residues C297 and C288 as two possible direct sensors of Nrf2 inducers. We have docked EQ, an effective inducer of ARE-regulated enzymes, into the IVR structural model. Without any preset preference between cavities C288 and C297, EQ was docked into cavity C297 automatically using software GOLD (10) probably because of the fact that the cavity around C288 is shallow.
ARE-driven luciferase assays and Western blot analyses of the top-score compounds have identified novel Nrf2 inducers The 23 top-score molecules from the first round of virtual screening were initially tested in the cell-based ARE-driven luciferase assay. Specifically, HepG2 cell line was transfected with ARE-driven pGL3 luciferase vector and then exposed to individual test compounds at 10, 50 lM. Classical Nrf2 inducers EQ and SFN were used as the positive controls. Our results showed that 7 out of 23 top-score compounds are significantly active in inducing ARE-driven luciferase activities (Figure 3). To examine the effect of these active compounds on Nrf2 nuclear translocation, we used Western blot analysis to measure the nuclear level of Nrf2 after cells were exposed to these compounds. Among the 7 hits, compound BM10 is most potent in inducing Nrf2 nuclear translocation (Figure S1, Supplementary Material). However, although compounds BM4, BM6, and BM18 potently induce ARE-driven luciferase activities (Figure 3), we found they only have modest activity in inducing Nrf2 nuclear translocation (Figure S1), suggesting the Western blot analysis of
Figure 2: Chemical structures of the novel Nrf2 inducers identified in this work. Chem Biol Drug Des 2010; 75: 475–480
477
Wu et al.
BM10, 31, 40 (Figure 4B). The potencies of BM31 and BM40 are comparable to that of SFN and EQ. It is of note that despite BM10 potently induces nuclear translocation of Nrf2, it shows modest potency in inducing the GCS. With our current data, we cannot explain this phenomenon. We therefore focused our further study on BM31. In addition, our initial data indicated the BM compounds do not activate the XRE pathway (data not shown).
Figure 3: The ARE-driven luciferase activity in the cell-based system exposed to the DMSO vehicle, EQ at 10 lM, SFN at 10 lM, and compounds BM4, 5, 6, 9, 10, 18, and 19 at 10 and 50 lM. Nrf2 nuclear translocation is a more reliable approach to detect Nrf2 inducers. Consequently, we have directly subjected the 17 compounds from the second cycle of screening to Western blot analysis, which led to identification of BM31 and BM40 as potent Nrf2 inducers. Overall, our results show BM10, BM31, and BM40 potently elevate the nuclear Nrf2 in HepG2 cells, and their potencies are comparable to that of SFN (Figure 4A). However, it is not clear at this stage why BM40 showed more pronounced effect on Nrf2 nuclear translocation at low concentration (10 lM) than the high concentration (50 lM).
Compounds BM10, 31, and 40 induce downstream carcinogen-detoxifying enzymes Nrf2 is a master transcription factor regulating multiple ARE-regulated antioxidant and detoxifying enzymes, such as GCS, glutathione S-transferase, NAD(P)H quinone oxidoreductase-1, and UDP glucuronosyltransferase. We examined the effects of our lead compounds on GCS. Our results demonstrated that GCS was indeed induced by
Impact of Keap1 mutation on the Nrf2-inducing potency of BM31 and SFN Keap1 is multi-domain protein, in which the IVR domain contains 4 cysteine residues, C257, C273, C288, and C297. The known Nrf2 inducers may interact with these cysteine residues in the IVR domain, resulting in conformational change of Keap1 and ultimately activating Nrf2. To investigate the role of these active cysteine residues in the Nrf2-inducing activity of BM31, we obtained 4 Keap1 mutants (C257S, C273S, C288S, and C297S) from our collaborator, Dr Mark Hannink (5). BM31 is selected for this and further investigation because the dose response is consistent for its capability in inducing Nrf2 nuclear translocation and induction of enzyme GCS (Figure 4). We have performed co-transfection experiments to investigate the effects of Keap1 IVR cysteine mutants on ARE-driven luciferase activities. Specifically, we co-transfected the 293T cells with ARE-driven pGL3 luciferase report vector, Keap1 vector with and without the Nrf2 expression vector (Figure 5). After transfection, the cells were exposed to BM31 at 50 lM for 16 h before being harvested. The cell lysates were used for measurement of ARE-driven luciferase activities. The data demonstrated that compound BM31 potently induces the ARE-driven luciferase activities when cells were co-transfected with ARE-pGL3, Nrf2, and the C257S or C273S or C288S mutant Keap1 vectors. The results demonstrate that the 273S and 288S Keap1 mutations led to the partial loss of basal level inhibition on Nrf2 by Keap1 and could constitutively activate the ARE-driven luciferase activities (Figure 5), consistent with previous findings (5). However, we found that the C273S and C288S mutations could not completely abolish the activity of BM31 or SFN in activating the ARE-driven luciferase activities (Figure 5). In particular, our data demonstrated that Keap1 297S mutant limits the potency of BM31 but not SFN in activating the
A
B
478
Figure 4: (A) Nrf2 protein detected in nuclear proteins by Western blotting; (B) Endogenous GCS, a Nrf2-regulated DME protein, in whole cell protein extracts, detected by Western blotting. HepG2 cells were treated with EQ, SFN, DMSO, BM10, BM31, and BM40 at designated concentrations for 24 h. Chem Biol Drug Des 2010; 75: 475–480
Identification of Novel Nrf2 Inducers
1
DMSO 2
BM31 3
SFN 4
Nrf2
Keap1 IP: Keap1
BM31 100
**
M
Chem Biol Drug Des 2010; 75: 475–480
*
60
D
Figure 6: Lane 1: Nrf2 transfection; Lanes 2, 3,4: Nrf2 and Keap1 co-transfection. Lanes 2, 3, 4: treated with DMSO, BM31 (50 lM), and SFN (5 lM), respectively. The 293T cell lysates were immunoprecipitated with Keap1 antibody, and eluted proteins were immunoblotted with Nrf2 antibody.
**
80
SO
IB: Nrf2 and Keap1
120
0
BM31 and SFP enhance the Keap1–Nrf2 association To study the effect of BM31 on the interaction of Keap1 with Nrf2, we performed Co-IP experiments. As the endogenous Keap1 and Nrf2 are low in these cells, we used ectopic expression of Keap1 and Nrf2. We first transfected Nrf2 expression vector into MCF-7 cells and found it very hard to detect Nrf2 expression, likely as a
00
ARE-driven luciferase activities (Figure 5). Our preliminary conclusion is that BM31 may target the binding site around C297 in the IVR domain of Keap1 and mutation of residue C297 could partially abolish the Nrf2-inducing effect of BM31.
BM31 protects MCF-7 cells from the cytotoxic damage of the carcinogen B[a]P To test the capability of BM compounds in protecting the cytotoxic effect of B[a]P, we have performed cellular assays using MCF-7 cells. Incubation of MCF-7 cells with various concentrations of BM31 for 48 h was followed by incubation with 160 nM B[a]P + the BM31 for 72 h. MTT assays were performed to evaluate the protection effect of the BM31. DMSO without B[a]P was used as a control. As shown in Figure 7, in the low dose range (from 50 to 200 nM), BM31 provides significant protection for the MCF-7 cells from the cytotoxic damage of the carcinogen B[a]P (Figure 7). Further work is needed to verify whether this protection effect of BM31 at nanomolar concentrations is medicated by induction of Nrf2-regulated detoxifying enzymes. In addition, we found that BM10, BM31, and BM40 do not have cytotoxic effects on the proliferation of MCF-7 cells in a broad dose range (10 nM–10 lM), which distinguishes them from SFN (Figure S2, Supplemental Material).
10
Figure 5: ARE-driven luciferase assays in 293T cells. Demonstrating Keap1 C297S mutant did not affect the potency of SFN but abolished the Nrf-2 induction activity of BM31.
0
C257S C297S C257S C273S C288S C297S + – + + + +
50
– +
0
– –
20
0 Keap1 Nrf2
10
1000
50
2000
Cell survival (% of control)
Relative luciferase
SFN–5 µM
20
BM31–50 µM 3000
10
DMSO
result of the presence of greatly increased expression of Cul3 E3ligase, which leads to active proteasomal degradation of Nrf2 (13). We therefore used 293T cells and successfully achieved high Nrf2 expression in this cell line. In the co-IP experiments, we co-transfected 293T cells with Nrf2 and Keap1 expression vectors. After transfection, the cells are treated with SFN and BM31 for 24 h before harvesting. The cell lysates were used for IP with Keap1 antibody. The eluted proteins were immunoblotted with Nrf2 antibody. The results revealed that treatment with BM31 or SFN increased the Nrf2 ⁄ Keap1 complex (Figure 6), indicating the BM31 might induce Nrf2 by modulating interaction between Keap1 and another protein partner in Keap1–Nrf2 pathway. For example, arsenic induces the Nrf2-dependent response through markedly enhancing the interaction between Keap1 and Cul3, subunits of the E3 ubiquitin ligase for Nrf2, which led to impaired dynamic assembly ⁄ disassembly of the E3 ubiquitin ligase and thus decreased its ligase activity, resulting in inhibition of Nrf2 ubiquitination and degradation (14). Our further work will determine whether BM31 induces Nrf2 via a mechanism similar to that of arsenic.
0
4000
Concentration (nM)
Figure 7: Protection effect of BM31 for MCF-7 cells against the cytotoxic effect of 160 nM B[a]P. DMSO without B[a]P is included as a control. *p < 0.05, **p < 0.006 when compared with control 0 (DMSO with B[a]P). Experiments were performed in triplicate. 479
Wu et al.
On the other hand, tumor cells could hijack the Nrf2 pathway to their survival advantage, conferring resistance to chemotherapeutic agents. Indeed, recent study indicates the involvement of Nrf2 activation in resistance to 5-fluorouracil in human colon cancer HT-29 cells (15). This highlights that Nrf2 inducer is a double-edge sword.
Conclusion and perspective We built a 3D structural model of the Keap1 IVR domain and demonstrated this structural model is effective in retrieving novel Nrf2 inducers from chemical databases. We discovered three novel compounds, BM10, BM31, and BM40, which significantly elevate the nuclear Nrf2 and induce ARE-regulated enzymes. We found that BM10, BM31, and BM40 do not have cytotoxic effects on the proliferation of MCF-7 cells in a broad dose range (10 nM–10 lM), which distinguishes them from SFN. Further, BM31 at nanomolar concentrations provides protections for MCF-7 cells from the cytotoxic damage of carcinogen B[a]p. In particular, strong experimental evidence suggests the involvement of photo-oxidative stress mediated by reactive oxygen species as a crucial mechanism of solar damage relevant to human skin photoaging and photocarcinogenesis. The Nrf2 inducers BM10, 31, and 40 could be further optimized as novel photo-chemopreventive agents, targeting skin cell photo-oxidative stress (16).
Acknowledgments This work was supported by the Canadian Institutes of Health Research (G.B.) and the U.S. Army Medical Research and Material Command under W81XWH-04-1-065 (J.W. & G.B.). We thank Dr Mark Hannick for providing Keap1-expressing plasmids. We are grateful for the chemical samples as a gift from the Drug Synthesis and Chemistry Branch, Developmental Therapeutics Program, NCI, Bethesda, MD, USA. J.W. is a FRSQ investigator.
References 1. Osburn W.O., Kensler T.W. (2008) Nrf2 signaling: an adaptive response pathway for protection against environmental toxic insults. Mutat Res;659:31–39. 2. Eggler A. L., Gay K. A., Mesecar A. D. (2008) Molecular mechanisms of natural products in chemoprevention: induction of cytoprotective enzymes by Nrf2. Mol Nutr Food Res;52:S84–S94. 3. Yates M.S., Kensler T.W. (2007) Keap1 eye on the target: chemoprevention of liver cancer. Acta Pharmacol Sin;28:1331–1342. 4. Kobayashi M., Yamamoto M. (2005) Molecular mechanisms activating the Nrf2-Keap1 pathway of antioxidant gene regulation. Antioxid Redox Signal;7:385–394. 5. Zhang D.D., Hannink M. (2003) Distinct cysteine residues in Keap1 are required for Keap1-dependent ubiquitination of Nrf2 and for stabilization of Nrf2 by chemopreventive agents and oxidative stress. Mol Cell Biol;23:8137–8151. 6. Wakabayashi N., Dinkova-Kostova A.T., Holtzclaw W.D., Kang M.I., Kobayashi A., Yamamoto M., Kensler T.W., Talalay P. (2004)
480
Protection against electrophile and oxidant stress by induction of the phase 2 response: fate of cysteines of the Keap1 sensor modified by inducers. Proc Natl Acad Sci USA;101:2040–2045. 7. Yamamoto T., Suzuki T., Kobayashi A., Wakabayashi J., Maher J., Motohashi H., Yamamoto M. (2008) Physiological significance of reactive cysteine residues of Keap1 in determining Nrf2 activity. Mol Cell Biol;28:2758–2770. 8. Xu Y., Xu D. (2000) Protein threading using PROSPECT: design and evaluation. Proteins;40:343–354. 9. Marti-Renom M.A., Stuart A.C., Fiser A., Sanchez R., Melo F., Sali A. (2000) Comparative protein structure modeling of genes and genomes. Annu Rev Biophys Biomol Struct;29:291–325. 10. Nissink J.W., Murray C., Hartshorn M., Verdonk M.L., Cole J.C., Taylor R. (2002) A new test set for validating predictions of protein-ligand interaction. Proteins;49:457–471. 11. Dinkova-Kostova A.T., Holtzclaw W.D., Cole R.N., Itoh K., Wakabayashi N., Katoh Y., Yamamoto M., Talalay P. (2002) Direct evidence that sulfhydryl groups of Keap1 are the sensors regulating induction of phase 2 enzymes that protect against carcinogens and oxidants. Proc Natl Acad Sci USA;99:11908–11913. 12. Hamilton D., Wu J.H., Batist G. (2007) Structure-based identification of novel human gamma-glutamylcysteine synthetase inhibitors. Mol Pharmacol;71:1140–1147. 13. Loignon M., Miao W., Hu L., Bier A., Bismar T.A., Scrivens P.J., Mann K., Basik M., Bouchard A., Fiset P.O., Batist Z., Batist G. (2009) Cul3 overexpression depletes Nrf2 in breast cancer and is associated with sensitivity to carcinogens, to oxidative stress, and to chemotherapy. Mol Cancer Ther;8:2432–2440. 14. Wang X.J., Sun Z., Chen W., Li Y., Villeneuve N.F., Zhang D.D. (2008) Activation of Nrf2 by arsenite and monomethylarsonous acid is independent of Keap1-C151: enhanced Keap1-Cul3 interaction. Toxicol Appl Pharmacol;230:383–389. 15. Akhdar H., Loyer P., Rauch C., Corlu A., Guillouzo A., Morel F. (2009) Involvement of Nrf2 activation in resistance to 5-fluorouracil in human colon cancer HT-29 cells. Eur J Cancer;45:2219– 2227. 16. Wondrak G.T., Cabello C.M., Villeneuve N.F., Zhang S., Ley S., Li Y., Sun Z., Zhang D.D. (2008) Cinnamoyl-based Nrf2-activators targeting human skin cell photo-oxidative stress. Free Radic Biol Med;45:385–395.
Supporting Information Additional Supporting Information may be found in the online version of this article: Figure S1. (A–C) Nrf2 protein detected in nuclear proteins by Western blotting. Figure S2. MCF-7 cell survival curves for BM10, BM31, BM40 and SFN. Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.
Chem Biol Drug Des 2010; 75: 475–480
