The benzothiophene selective estrogen receptor modulators (SERM) raloxifene and arzoxifene are in clinical use and clinical trials for chemoprevention of breast cancer and other indications. These SERMs are “oxidatively labile” and therefore have potential to activate antioxidant responsive element (ARE) transcription of genes for cytoprotective phase II enzymes such as NAD(P)H-dependent quinone oxidoreductase 1 (NQO1). To study this possible mechanism of cancer chemoprevention, a family of benzothiophene SERMs was developed with modulated redox activity, including arzoxifene and its metabolite desmethylarzoxifene (DMA). The relative antioxidant activity of these SERMs was assayed and correlated with induction of NQO1 in murine and human liver cells. DMA was found to induce NQO1 and to activate ARE more strongly than other SERMs, including raloxifene and 4-hydroxytamoxifen. Livers from female, juvenile rats treated for 3 days with estradiol and/or with the benzothiophene SERMs arzoxifene, DMA, and F-DMA showed substantial induction of NQO1 by the benzothiophene SERMs. No persuasive evidence in this assay or in MCF-7 breast cancer cells was obtained of a major role for the estrogen receptor in induction of NQO1 by the benzothiophene SERMs. These results suggest that arzoxifene might provide chemopreventive benefits over raloxifene and other SERMs via metabolism to DMA and stimulation of ARE-mediated induction of phase II enzymes. The correlation of SERM structure with antioxidant activity and NQO1 induction also suggests that oxidative bioactivation of SERMs may be modulated to enhance chemopreventive activity. [Mol Cancer Ther 2007;6(9):2418–28]

Selective estrogen receptor modulators (SERM) are clinically important for treatment and prevention of breast cancer and for therapy of postmenopausal symptoms. The therapeutic mechanism of action of SERMs is widely believed to be tissue- and cell-selective agonist or antagonist activity at the estrogen receptor (ER; ref. 1). The optimization of ligand binding to the two ER isoforms, ERα and ERβ, is often used to guide SERM design. A major drug discovery goal is the “ideal SERM” that is antiestrogenic in breast and endometrial tissue but proestrogenic in the vasculature and brain, which would be of use in cancer chemoprevention and represents an attractive alternative to hormone replacement therapy. The majority of SERMs, because of common chemical structural elements, can be oxidatively bioactivated to reactive metabolites. These metabolites are not only potentially genotoxic but also have the potential to induce cytoprotective phase II enzymes such as NAD(P)H-dependent quinone oxidoreductase (NQO1). However, there has been little attention to structural modifications of SERMs designed to control oxidative bioactivation and via ER-independent mechanisms thereby to enhance chemopreventive activity and to attenuate toxicity (2).

The ER accommodates planar polycyclic phenols, and hence all SERMs, or their active metabolites, have polyaromatic phenolic scaffolds; SERMs are therefore generally redox active and susceptible to oxidative metabolism in vivo (36). The active metabolite of the archetype SERM tamoxifen (7) is 4-hydroxytamoxifen (HOT), a triphenylethylene phenol formed on oxidative metabolism (Fig. 1A; ref. 8). Further oxidative metabolism of HOT yields quinoid reactive metabolites: a quinone methide and an o-quinone (3). Similarly, other SERMs are metabolized to reactive quinoids, including the diquinone methides formed by the benzothiophene SERMs raloxifene and arzoxifene (Fig. 1B; refs. 4, 5). These quinoids are able to act as electrophiles and oxidants leading to modification of cellular biomolecules and thus potentially to cytotoxicity and genotoxicity (3). Tamoxifen remains an effective clinical therapeutic for breast cancer, but long-term treatment is known to be associated with increased risk of endometrial cancer (9).

Figure 1.

A, structures of SERMs and other compounds used in this study. B, single-electron oxidation of benzothiophene SERMs and BTC leads to a semiquinone or phenoxy radical, but a second one-electron oxidation to yield a quinoid is blocked for all derivatives except 4′-OH (DMA), NH2-DMA, and BTC.

Figure 1.

A, structures of SERMs and other compounds used in this study. B, single-electron oxidation of benzothiophene SERMs and BTC leads to a semiquinone or phenoxy radical, but a second one-electron oxidation to yield a quinoid is blocked for all derivatives except 4′-OH (DMA), NH2-DMA, and BTC.

Close modal

Cellular defense mechanisms that respond to damage from oxidative stress and electrophiles, the key causes of malignant transformation, represent a target for chemopreventive agents (10). It has been known for some time that compounds that possess weak carcinogenic activity have the potential to protect against carcinogenesis (11). The balance between the deleterious effects and chemopreventive potential of compounds that are able to form electrophilic and oxidative metabolites is determined by their metabolic bioactivation and ultimately by their chemical structure (Fig. 2).

Figure 2.

Beneficial and detrimental consequences of SERM oxidative metabolism resulting primarily from formation of reactive oxygen species (ROS) and electrophilic quinoid metabolites capable of covalent modification of biomolecules, most importantly by Michael-type additions to thiols.

Figure 2.

Beneficial and detrimental consequences of SERM oxidative metabolism resulting primarily from formation of reactive oxygen species (ROS) and electrophilic quinoid metabolites capable of covalent modification of biomolecules, most importantly by Michael-type additions to thiols.

Close modal

Induction of cytoprotective phase II enzymes such as NQO1 is mediated largely at the antioxidant/electrophile responsive element (ARE) that promotes transcriptional activation of phase II genes (12). Prochaska et al. (13, 14) suggested that the “oxidative lability” of compounds created the inducer signal, and work on induction by quinones led to the search for a sulfhydryl-rich sensor protein able to transduce modification of sulfhydryl groups to give ARE activation. Current thinking holds that the cytosolic regulatory protein Kelch-like ECH-associated protein 1 (Keap1) is the sulfhydryl-rich sensor that responds to modification by oxidants/electrophiles. Under basal conditions, Keap1 associates with nuclear factor erythroid 2–related factor 2 (Nrf2) and targets it for degradation by the ubiquitin-proteasome proteolysis system (15, 16). When Keap1 is modified by oxidative/electrophilic reagents, proteolysis is inhibited and dissociated Nrf2 translocates to the nucleus, binds to ARE, and promotes phase II enzyme gene expression (17).

Reactive metabolites formed from SERMs may cause cell stress, alternatively initiating genotoxic/cytotoxic pathways or contributing to cytoprotection via the Keap1/Nrf2 pathway. The opportunity therefore exists to modulate metabolic bioactivation toward cytoprotection by structural modification. Benzothiophene SERMs were selected because of the clinical importance of raloxifene in women's health (18) and the potential importance of arzoxifene, an improved next-generation SERM targeted at chemoprevention (19). A homologous series of arzoxifene analogues has been developed (Fig. 1) and studied in vitro and in vivo, showing that ER-independent chemoprevention is a realistic target for SERM therapeutics.

Reagents

BTC {the “benzothiophene core” 2-(4-hydroxyphenyl)benzo[b]thiophen-6-ol} was synthesized following literature procedures and the X-DMA SERMs using our improved methods (20, 21). Other reagents were purchased from Sigma unless stated otherwise. Cell culture media and supplements, gels, and buffers were obtained from Invitrogen.

Peroxyl Radical Trapping

Net antioxidant capacity was assayed by the BODIPY11 fluorescence decay method. X-DMA (10 μmol/L) and BODIPY11 stock solutions prepared in 40% acetonitrile/60% 10 mmol/L PBS were added to a 96-well plate and the reaction was initiated by addition of small volumes (10 μL) of azobisamidinopropane stock solution in PBS. The fluorescence decay of the probe was monitored to completion. The emission intensity in relative fluorescence units was measured with time and normalized; with 100% corresponding to relative fluorescence units immediately before addition of initiator. The net antioxidant capacity [AUC = (AUCinh − AUCun) / AUCun] was calculated from the area under the curve of the BODIPY11 fluorescence decay in the presence (AUCinh) and absence (AUCun) of an antioxidant (22).

Cell Culture

Hepa 1c1c7 murine hepatoma cells were supplied by Dr. J.P. Whitlock, Jr. (Standford University, Stanford, CA) and were cultured in α-MEM with 1% penicillin-streptomycin and 10% fetal bovine serum (Atlanta Biologicals). HepG2 cells stably transfected with ARE-luciferase reporter were kindly provided by Dr. A.N. Kong (Rutgers University, Piscataway, NJ) and cultured in F-12 medium with 10% fetal bovine serum, 1% penicillin-streptomycin, 1% essential amino acids, and 0.2 mg/mL insulin (23). ERα-positive, ERβ-positive, and ER-negative MCF-7 cells were generously supplied by Drs. D.B. Lubahn and W.V. Welshons (University of Missouri, Columbia, MO) and were cultured in MEM plus 2.5 mol/L HEPES, 6 μg/mL insulin, 2 mmol/L l-glutamine, 5% stripped calf serum, and 2% penicillin-streptomycin with or without geneticin. All cells were incubated in 5% CO2 at 37°C and medium was changed every 2 to 3 days.

Animal Treatment

The study protocol was approved by the institutional Animal Care and Use Committee of University of Illinois at Chicago. Female juvenile Sprague-Dawley rats were received at 12 days of age from Harlan and 11 pups were housed in a cage with a nursing dam. Animals had free access to standard rat chow (Harlan Teklad) and water and were allowed 1 week to acclimate. Nineteen- to 21-day-old rats were treated with 17β-estradiol (E2 as the benzoate, 0.1 mg/kg/d), DMA (10 mg/kg/d), arzoxifene (10 mg/kg/d), or F-DMA (10 mg/kg/d) s.c., either alone or challenged with E2 (0.1 mg/kg/d), for 3 days. On the fourth day, animals were sacrificed and livers were excised, rapidly frozen in liquid nitrogen, and stored at −80°C until NQO1 assay. The rats were maintained on a controlled light cycle of 14:10 h (light/dark) at 25°C with water and food ad libitum. All animal procedures were reviewed and approved by the University of Illinois at Chicago before implementation.

NQO1 Activity Assay

For rat liver samples, liver was homogenized in 0.25 mol/L sucrose and homogenate was centrifuged at 15,000 × g for 20 min at 4°C (24). Supernatant was collected and one-fifth volume of 0.1 mol/L CaCl2/0.25 mol/L sucrose was added, incubated on ice for 30 min, and centrifuged as before. Clear cytosolic supernatant was used immediately or stored at −80°C. Concentrated supernatant was diluted 50 times in PBS and triplicates were assayed as below. Protein concentration was determined by the bicinchoninic acid method (Pierce).

For cultured liver cells, Hepa 1c1c7 cells were plated on 96-well plates, incubated overnight, and treated with test sample, DMSO, or bromoflavone. After 48 h, the NQO1 activity was assayed as previously described (25). Induction of NQO1 activity was calculated by comparing the NQO1 specific activity of sample-treated cells with that of solvent-treated cells. CD values represent the concentration required to double NQO1 induction. Chemopreventive index (CI = IC50/CD) is calculated from IC50 for inhibition of cell growth (24).

For cultured MCF-7 cells, NQO1 activity was assayed using 2,6-dichlorophenolindophenol as the substrate according to Lhoste et al. (26) with minor modifications. The final reaction mixture contained 25 mmol/L Tris-HCl (pH 7.4), 0.7 mg/mL bovine serum albumin, 0.01% Tween 20 (w/v), 5 μmol/L FAD, 200 μmol/L NADH, and 40 μmol/L 2,6-dichlorophenolindophenol, and assays were carried out in the presence and absence of 20 μmol/L dicumarol. The presence of 20 μmol/L dicumarol completely inhibited enzyme-dependent 2,6-dichlorophenolindophenol reduction. NQO1 activity was described as the dicumarol-inhibitable decrease in absorbance of 2,6-dichlorophenolindophenol for 5 min at 600 nm on a microplate reader. NQO1 activity was normalized to the protein concentration, which was determined by the bicinchoninic acid assay using bovine serum albumin as a standard. Induction of activity was calculated as described above.

Immunoassay

Briefly, 10 μg of tissue or cell lysis protein were separated by electrophoresis in NuPAGE Novex 4% to 12% Bis-Tris gel (Invitrogen) and transferred onto nitrocellulose membranes (Bio-Rad). Membranes were blocked with 5% nonfat dry milk and incubated with primary polyclonal antibodies to NQO1 (1:500) or β-actin (1:500) and then with horseradish peroxidase–linked chicken anti-goat immunoglobulin G (1:1,000; all from Santa Cruz Biotechnology). Western blots were detected with the ECL kit (Amersham Biosciences).

ARE-Luciferase Reporter Assay

HepG2-ARE-Luc cells were plated on a six-well plate overnight; treated with DMSO, test samples, or positive control bromoflavone on the second day; and incubated for additional 24 h. The luciferase activity was determined as previously described (27). The data were obtained from three separate experiments and expressed as fold induction over control [treated cells/DMSO-treated cells (± SD)].

Cell Growth Determination

Cells were plated and treated as described in the NQO1 assay. The IC50 values were determined as previously described (27).

Statistics

One-way ANOVA analysis with Dunnett's posttest or Tukey's multiple comparison test was done using GraphPad Prism version 4.00 for Windows, GraphPad Software.

Net Antioxidant Capacity of X-DMA SERMs

To test the redox and antioxidant activities of the X-DMA SERMs, net antioxidant capacity was assayed using the BODIPY11 fluorescent probe. Decay of fluorescence reflects the formation and reaction of the probe with peroxyl radicals in the aqueous acetonitrile medium, formed from the thermal decay of the azo initiator azobisamidinopropane (22). This assay of peroxyl radical trapping is relevant to inhibition of lipid peroxidation, wherein net antioxidant activity can be quantified by the area under the curve. X-DMA SERMs (10 μmol/L) were incubated with BODIPY11 and azobisamidinopropane initiator, the observed retardation of fluorescence decay in the presence of SERMs indicating a modest ability to inhibit BODIPY oxidation by peroxyl radical trapping (Fig. 3). The X-DMA SERMs that are able to act as traditional phenolic antioxidants (X = MeO, F, Br, SO2CH3) modestly retarded the rate of reporter oxidation, with no discernable dependence in this assay on the electron-withdrawing nature of the 4′-substituent; however, the two SERMs that are able to undergo two-electron oxidation to a quinoid (DMA, NH2-DMA) were significantly better antioxidants.

Figure 3.

Net antioxidant capacity toward peroxyl radicals of X-DMA SERMs (10 μmol/L), assayed by measurement of azobisamidinopropane-induced fluorescence decay of BODIPY11 reporter in 40% acetonitrile in 10 mmol/L phosphate buffer at 37°C. A, relative net antioxidant capacity as given by AUC [AUC = (AUCinh − AUCun) / AUCun]. B, representative curves for the time course of BODIPY11 oxidation as measured by relative fluorescence units (RFU), normalized relative to 100% relative fluorescence units immediately before addition of initiator.

Figure 3.

Net antioxidant capacity toward peroxyl radicals of X-DMA SERMs (10 μmol/L), assayed by measurement of azobisamidinopropane-induced fluorescence decay of BODIPY11 reporter in 40% acetonitrile in 10 mmol/L phosphate buffer at 37°C. A, relative net antioxidant capacity as given by AUC [AUC = (AUCinh − AUCun) / AUCun]. B, representative curves for the time course of BODIPY11 oxidation as measured by relative fluorescence units (RFU), normalized relative to 100% relative fluorescence units immediately before addition of initiator.

Close modal

Induction of NQO1 in Liver Cells

Inducible cytoprotective enzymes such as NQO1 have commonly been labeled as phase II enzymes, and this label is used herein, although the appropriateness of this label is problematic and recognized (28). X-DMA SERMs were administrated to Hepa 1c1c7 cells in different concentrations and NQO1 activity was measured. The Hepa 1c1c7 cell system is widely used to screen chemopreventive agents because the NQO1 enzyme is readily induced to measurable levels that allow comparison of agents (29). Induction of NQO1 activity in the presence of SERMs was compared with solvent and to bromoflavone as a positive control (Table 1; ref. 24). All SERMs, but not E2, induced NQO1 in Hepa 1c1c7 cells. At the lower concentration of the SERMs studied, induction by DMA, Br-DMA, and NH2-DMA was significant (P < 0.05), whereas at the higher concentration, all seven X-DMA SERMs reached significance. Interestingly, BTC was seen to be a potent inducer of NQO1, giving 3-fold induction at 2.5 μmol/L, whereas DMA only reached 2.2-fold induction at twice the concentration. The DMA series SERMs as a family seem to have greater potential for chemoprevention than raloxifene and HOT, as determined by the CD measured (Table 1).

Table 1.

Inhibition of cell growth and relative NQO1 induction by SERMs in Hepa 1c1c7 cells

CompoundsIC50 (μmol/L), Hepa 1c1c7IC50 (μmol/L), HepG2Relative induction, 5.0 μmol/L*Relative induction, 2.5 μmol/L*CD (μmol/L)CIEHOMO (eV)
DMSO (0.5%) — — 1.01 ± 0.03 1.0 ± 0.05 — — — 
Bromoflavone (0.67 μmol/L) — — — 5.80 ± 0.50 — — — 
BTC 27.1 ± 6.0 22.1 ± 1.8 3.10 ± 0.45 2.90 ± 0.40 1.7 15.9 — 
DMA 5.6 ± 1.4 13.5 ± 1.0 2.21 ± 0.44 1.65 ± 0.38 4.1 1.4 −5.49 
H-DMA 4.7 ± 1.1 8.7 ± 2.1 1.90 ± 0.57 1.26 ± 0.05 5.9 0.8 −5.69 
MeO-DMA (arzoxifene) 6.5 ± 2.3 10.6 ± 1.4 1.84 ± 0.35 1.36 ± 0.10 6.1 1.1 −5.45 
Br-DMA 4.3 ± 1.4 8.6 ± 2.8 1.73 ± 0.43 1.51 ± 0.50 6.5 0.7 −5.77 
F-DMA 5.6 ± 1.3 9.1 ± 2.6 1.68 ± 0.37 1.35 ± 0.13 7.3 0.8 −5.72 
NH2-DMA 12.3 ± 5.4 19.2 ± 4.0 1.61 ± 0.20 1.51 ± 0.22 7.6 1.6 −5.20 
SO2Me-DMA 8.4 ± 2.2 6.4 ± 0.6 1.63 ± 0.12 1.50 ± 0.10 7.8 1.1 −6.01 
HOT 7.3 ± 1.0 14.5 ± 2.2 1.24 ± 0.21 1.12 ± 0.13 21 0.3 — 
Raloxifene 10.2 ± 3.3 24.7 ± 6.3 1.31 ± 0.17 1.20 ± 0.14 16 0.6 — 
E2 22.6 ± 7.1 29.8 ± 1.7 1.09 ± 0.31 0.90 ± 0.20 53 0.4 — 
CompoundsIC50 (μmol/L), Hepa 1c1c7IC50 (μmol/L), HepG2Relative induction, 5.0 μmol/L*Relative induction, 2.5 μmol/L*CD (μmol/L)CIEHOMO (eV)
DMSO (0.5%) — — 1.01 ± 0.03 1.0 ± 0.05 — — — 
Bromoflavone (0.67 μmol/L) — — — 5.80 ± 0.50 — — — 
BTC 27.1 ± 6.0 22.1 ± 1.8 3.10 ± 0.45 2.90 ± 0.40 1.7 15.9 — 
DMA 5.6 ± 1.4 13.5 ± 1.0 2.21 ± 0.44 1.65 ± 0.38 4.1 1.4 −5.49 
H-DMA 4.7 ± 1.1 8.7 ± 2.1 1.90 ± 0.57 1.26 ± 0.05 5.9 0.8 −5.69 
MeO-DMA (arzoxifene) 6.5 ± 2.3 10.6 ± 1.4 1.84 ± 0.35 1.36 ± 0.10 6.1 1.1 −5.45 
Br-DMA 4.3 ± 1.4 8.6 ± 2.8 1.73 ± 0.43 1.51 ± 0.50 6.5 0.7 −5.77 
F-DMA 5.6 ± 1.3 9.1 ± 2.6 1.68 ± 0.37 1.35 ± 0.13 7.3 0.8 −5.72 
NH2-DMA 12.3 ± 5.4 19.2 ± 4.0 1.61 ± 0.20 1.51 ± 0.22 7.6 1.6 −5.20 
SO2Me-DMA 8.4 ± 2.2 6.4 ± 0.6 1.63 ± 0.12 1.50 ± 0.10 7.8 1.1 −6.01 
HOT 7.3 ± 1.0 14.5 ± 2.2 1.24 ± 0.21 1.12 ± 0.13 21 0.3 — 
Raloxifene 10.2 ± 3.3 24.7 ± 6.3 1.31 ± 0.17 1.20 ± 0.14 16 0.6 — 
E2 22.6 ± 7.1 29.8 ± 1.7 1.09 ± 0.31 0.90 ± 0.20 53 0.4 — 

NOTE: Data shown are mean ± SD for at least triplicate measurements.

*

Induction of NQO1 activity relative to DMSO control.

P < 0.01, treatment significantly different from the control (one-way ANOVA with Dunnett's posttest).

P < 0.05, treatment significantly different from the control (one-way ANOVA with Dunnett's posttest).

CD and chemopreventive index (CI = IC50/CD) have been used to identify new chemopreventive agents. The X-DMA SERMs showed IC50 values for inhibition of cell growth in murine and human liver cells of 4 to 20 μmol/L (Table 1). The best agents were BTC, DMA, and NH2-DMA with CIs of 16, 1.4, and 1.6, respectively. For comparison, the values reported for bromoflavone are CD = 0.01 μmol/L and CI = 17,000, whereas for the well-studied chemopreventive agent sulforaphane, CD = 0.43 μmol/L and CI = 26 (24). Although cytotoxicity was observed for benzothiophene SERMs in liver cancer cells, liver toxicity is not revealed in clinical data on raloxifene and arzoxifene (30).

The SERMs were also assayed in the human liver cell line HepG2, which is important because of the availability of HepG2 cells transfected with the ARE-luciferase reporter (27, 31). BTC was confirmed as a strong inducer of NQO1 and DMA as a weaker but significant inducer (Fig. 4A). The observations in the two cell lines indicated that X-DMA SERMs have the potential to induce NQO1 in the liver.

Figure 4.

A, incubation of SERMs (2.5 μmol/L) with HepG2 cells for 24 h showing fold induction of NQO1 relative to control as measured by enzyme activity. The measured IC50 values in this cell line are annotated to the plot for reference. B, incubation of SERMs (2.5 μmol/L) with HepG2 cells for 24 h showing fold induction of ARE-luciferase reporter activity relative to vehicle control.

Figure 4.

A, incubation of SERMs (2.5 μmol/L) with HepG2 cells for 24 h showing fold induction of NQO1 relative to control as measured by enzyme activity. The measured IC50 values in this cell line are annotated to the plot for reference. B, incubation of SERMs (2.5 μmol/L) with HepG2 cells for 24 h showing fold induction of ARE-luciferase reporter activity relative to vehicle control.

Close modal

Induction of NQO1 In vivo

Three of the X-DMA SERMs were selected for study in vivo: DMA, arzoxifene, and F-DMA. To study the role of ER signal transduction in the action of the X-DMA SERMs, E2 was included, alone and in coadministration arms. The juvenile rat model, commonly used in study of estrogenic activity, using 19- to 21-day-old female Sprague-Dawley rats, was selected (8). The organs of rats at this age are fully estrogen responsive, but the hormone is not yet being produced in the ovaries, so this rat model can offer a clear hormone background when studying SERM activity (32). Eight groups of immature female rats were treated daily with E2 (0.1 mg/kg) or vehicle control, with or without X-SERMs (10 mg/kg), for 3 days. Animals were sacrificed on the 4th day and individual NQO1 activity was analyzed in the excised liver. The selected dosage compares to 6 or 20 mg/kg arzoxifene daily in a chemoprevention study in mice (33) and 10 to 50 mg/kg arzoxifene daily in phase I human clinical studies, which reported no adverse events compared with placebo (30).

The cumulative data showed that induction of NQO1 was observed in all animals treated with X-DMA SERMs: arzoxifene, DMA, and F-DMA induced NQO1 activity to about 1.6-, 1.5-, and 1.3-fold that of the control animals, respectively (Fig. 5A and B). Fold induction was significant for all three X-DMA SERMs relative to the E2 treatment group and was significant for arzoxifene relative to the vehicle control group (P < 0.05). Although not reaching significance, the E2 treatment group showed a trend toward reduced levels of hepatic NQO1 relative to control.

Figure 5.

Induction of liver NQO1 activity in juvenile female rats treated for 3 d (10 mg/kg qd SERMs or 0.1 mg/kg qd E2 or vehicle control). A, distribution of individual rats in treatment groups receiving arzoxifene (Arz), DMA, F-DMA, or E2. B, cumulative data showing statistical significance of NQO1 induction by arzoxifene relative to control and significant induction by arzoxifene (P < 0.001), DMA (P < 0.005), and F-DMA (P < 0.05) relative to the E2-treated group. C, Western blot showing different levels of NQO1 expression in rat liver tissue of different groups. D, cumulative data showing the effects of 10 mg/kg qd X-DMA SERMs in antagonizing and reversing the effects of 0.1 mg/kg qd E2 on cotreatment. Data were analyzed by one-way ANOVA using Tukey's post hoc analysis.

Figure 5.

Induction of liver NQO1 activity in juvenile female rats treated for 3 d (10 mg/kg qd SERMs or 0.1 mg/kg qd E2 or vehicle control). A, distribution of individual rats in treatment groups receiving arzoxifene (Arz), DMA, F-DMA, or E2. B, cumulative data showing statistical significance of NQO1 induction by arzoxifene relative to control and significant induction by arzoxifene (P < 0.001), DMA (P < 0.005), and F-DMA (P < 0.05) relative to the E2-treated group. C, Western blot showing different levels of NQO1 expression in rat liver tissue of different groups. D, cumulative data showing the effects of 10 mg/kg qd X-DMA SERMs in antagonizing and reversing the effects of 0.1 mg/kg qd E2 on cotreatment. Data were analyzed by one-way ANOVA using Tukey's post hoc analysis.

Close modal

Adult female Sprague-Dawley rats have been reported to show elevated NQO1 in response to E2 treatment (32), but the effects of E2, which are well known to be highly dose sensitive, are expected to be more complex in adult animals endogenously producing E2. A study on NQO1 levels in the tissue of ovariectomized mice treated with E2 over 14 days showed high tissue variability but significantly reduced uterine NQO1 (34). The observed enzyme activity was substantiated by immunoassay of NQO1 protein expression by Western blot (Fig. 5C), confirming that the three X-DMA SERMs induced elevated levels of NQO1 protein relative to control in rat liver, in contrast to E2. Finally, in female juvenile rats, the influence of E2 coadministration on the putative chemopreventive actions of the X-DMA SERMs was not significant (Fig. 5D). The simplest explanation for this observation is that NQO1 induction by the SERMs in the liver is not ER mediated.

Induction of NQO1 in Breast Cancer Cells

To explore the influence of SERMs on NQO1 induction in estrogen-sensitive cells, ERα, ERβ, and ER(−) MCF-7 breast cancer cells were studied. Jiang et al. (35) have assessed phase II enzyme induction in seven cell lines in response to sulforaphane and other chemopreventive agents, reporting that Hepa 1c1c7 cells were highly inducible, but that others, including MCF-7 and MDA-MB-231 breast cancer cells, were poorly responsive to chemopreventive agents (e.g., sulforophane induced NQO1 2-fold in Hepa 1c1c7 cells, but <1.2-fold in MCF-7 cells). Background NQO1 activity was observed to be significantly higher in MCF-7 than in Hepa 1c1c7 cells (data not shown). Nevertheless, one group has reported NQO1 induction by SERMs in MCF-7 cells, stating that HOT induced NQO1 via an ERβ-dependent pathway (36). Another group has reported NQO1 induction in both ERα- and ERβ-transfected cell lines derived from MCF-7 cells by tamoxifen and HOT (34). Thus, we initially studied the MCF-7 cell line stably transfected with ERβ, but observed only very weak induction of NQO1 by the X-DMA SERMs raloxifene and HOT (Supplementary data).1

1

Supplementary material for this article is available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/).

At concentrations from 1 nmol/L to 2 μmol/L, no substantial increase in induction was observed for HOT (Supplementary data).1 Selected SERMs were then studied in the MCF-7 ERα cell line, with raloxifene and DMA being observed to induce NQO1 by 1.25-fold over control, but again effects did not reach significance (Supplementary data).1 In ER(−) MCF-7 cells, the induction of NQO1 by DMA was significant compared with both control and HOT-treated cells (Supplementary data).1 The relative insensitivity of MCF-7 cells to NQO1 induction is shown by the weak response to the positive control bromoflavone, which gave 1.3-fold induction compared with 1.2-fold by DMA. Ansell et al. (34) have emphasized that estrogenic actions mediated through the ARE are highly dependent on cellular environment to account for differences in cell culture observations in the literature. The idiosyncrasies of MCF-7 cell sublines, especially toward estrogens and antiestrogens, have been noted (37). Nevertheless, the results presented herein are compatible with an ER-independent mechanism for NQO1 induction by DMA in MCF-7 cells.

Activation of ARE-Luciferase

To investigate the possible mechanism of NQO1 induction by X-DMA SERMs, HepG2-ARE-C8 cells were treated with SERMs and luciferase induction was assayed. HepG2-ARE-C8 is a human hepatocarcinomal cell line that was stably transfected with the pARE-T1-luciferase construct. Luciferase induction in this cell line reflects the ability to activate the endogenous ARE, which regulates expression of many phase II enzymes. A good correlation between ARE-luciferase induction and NQO1 induction is anticipated and indeed is shown in Fig. 4B. BTC was observed to strongly activate ARE, giving 13-fold luciferase induction relative to control. X-DMA SERMs raloxifene and HOT gave 1.5-fold induction, and again DMA was observed to be an exceptional ARE activator, giving almost 3-fold induction of ARE-luciferase activity. These data strongly imply that in liver cells, the observed induction of NQO1 is mediated via ARE activation.

SERMs are polyaromatic phenols susceptible to oxidative metabolism to electrophilic quinoid metabolites (Fig. 1), which may cause either genotoxicity/cytotoxicity or cytoprotection via triggering of cellular defense mechanisms (Fig. 2). The quinoid metabolites, including o-quinones, quinone methides, and diquinone methides, have variously been shown to generate reactive oxygen species and to both oxidize and alkylate biomolecules such as proteins and nucleic acids; this has been argued to contribute to cytotoxicity and genotoxicity (3). However, sulfhydryl group oxidation or alkylation by quinoids, which are Michael acceptor electrophiles, can trigger chemopreventive mechanisms, in particular via reaction with Keap1 and subsequent Nrf2/ARE activation of induction of cytoprotective phase II enzymes such as NQO1 (10, 12, 13, 38). The balance between the detrimental and beneficial drug properties associated with oxidative bioactivation and potential outcomes are depicted in Fig. 2.

The specific semiquinone and electrophilic quinoid formed from bioactivation of benzothiophene SERMs are depicted in Figs. 1B and 2. When the 4′-MeO group of arzoxifene is replaced by groups such as F, Br, H, and SO2Me, oxidation to a quinoid is blocked (Fig. 1B; ref. 39). The potential detrimental effects of quinoid formation are hence prevented, but this raises the question about whether the putative beneficial effects, notably chemoprevention via induction of phase II enzymes, are also lost. A family of benzothiophene SERMs has been developed (20) to answer the following questions: Does benzothiophene SERM bioactivation lead to induction? Is this property lost if bioactivation to a quinoid is structurally blocked? The 4′-substituted desmethoxyarzoxifene (X-DMA) family includes arzoxifene (X = MeO) and the arzoxifene metabolite DMA (X = OH). Previous comparison of DMA and F-DMA showed that DMA is oxidatively bioactivated to a diquinone methide, which is a Michael acceptor depleting cellular reduced glutathione (Fig. 2), in contrast to F-DMA, which is also an ER ligand but is not bioactivated to a quinoid and does not deplete cellular reduced glutathione (Fig. 1B; refs. 5, 39).

SERMs and Oxidative Bioactivation

Tamoxifen is oxidatively metabolized to HOT and other electrophilic reactive intermediates leading to formation of DNA adducts and genotoxicity (40). The increased risk of endometrial cancer in long-term tamoxifen therapy is associated, at least in part, with this chemical carcinogenesis pathway (3, 9). Subsequent generations of SERMs in clinical use and preclinical development including the benzothiophene SERMs raloxifene and arzoxifene are based on polyaromatic phenolic scaffolds and have been observed to form quinoids in vitro and in vivo (36, 39). Many of these quinoid metabolites have been shown to act as electrophiles and/or to generate reactive oxygen species, although, as yet, there is no evidence that the newer-generation SERMs are carcinogenic. The chemical character of oxidative metabolites formed from individual SERMs is varied, and this is likely to strongly influence the balance between the potential detrimental and beneficial effects. Induction of the phase II enzymes aryl sulfotransferase and hydroxysteroid sulfotransferase has been reported in rat liver and intestine, in both male and female Sprague-Dawley rats, after tamoxifen treatment for 1 to 2 weeks (41). In separate studies, based largely on work in MCF-7 cells, HOT and raloxifene have been reported to induce NQO1 via an ERβ-dependent mechanism (36). In further studies, NQO1 induction by tamoxifen and HOT was reported in both ERα- and ERβ-positive cells derived from the MCF-7 line (34). Thus, in the present study, HOT and raloxifene were included with the X-DMA SERMs for comparison.

NQO1 Induction via ARE Activation

Measurement of NQO1 activity in vitro and in vivo provides an efficient assay for discovery of chemopreventive agents (29), and NQO1 is a biomarker for cancer chemoprevention (42). NQO1, as other inducible phase II enzymes such as glutathione S-transferases and UDP-glucuronyl transferase, provides cytoprotection by detoxification of small molecules; however, NQO1 also binds and regulates the stability of the tumor suppressor protein p53, inhibiting p53 degradation (43). Epidemiologic evidence indicates that genetically deficient NQO1 is a risk factor for development of malignancies, and, regardless of the exact mechanism of chemoprevention, NQO1 has been proved as a useful biomarker for discovery of agents such as sulforaphane that have shown efficacy in chemoprevention of breast and other cancers (44). Sulforaphane, Michael acceptors, and other classes of chemopreventive agents have been shown to chemically modify the cysteine residues of the redox sensor protein Keap1, leading to activation of ARE by Nrf2 (16). Ex vivo experiments using Keap1, Keap1/Nrf2, and Nrf2 knockouts showed ablated induction of NQO1 by sulforaphane and other agents, emphasizing the importance of the Keap1/Nrf2/ARE pathway for chemoprevention (45). However, multiple kinase cascades converge on Nrf2 phosphorylation and both actin modification and gene transactivation may also contribute to overall ARE-mediated induction (31, 46).

Expectations from X-DMA Redox Activity

In comparing the properties of the X-DMA SERMs, the simplistic prediction would be that only DMA and possibly NH2-DMA, which are susceptible to bioactivation to quinoids, will be capable of oxidation and electrophilic modification of Keap1 leading to NQO1 induction via Nrf2/ARE. If confirmed by experiment, this would indicate that (a) oxidative bioactivation of SERMs is a contributor to their chemopreventive activity, and (b) new SERMs should incorporate oxidative lability as a design parameter.

Not all chemopreventive agents are electrophiles. Early work on the phenolic antioxidant 2,3-t-butyl-4-hydroxyanisole showed that the observed chemopreventive actions correlated with induction of NQO1, levels of which were elevated in nearly all tissues (47). More recent work on 2,3-t-butyl-4-hydroxyanisole implicated activation of kinase cascades converging on Nrf2 in NQO1 induction (31). The X-DMA family members that are blocked from diquinone methide formation (X = H, F, Br, OMe, SO2Me) were observed in the 1,1-diphenyl-2-picrylhydrazyl radical scavenging assay to act as phenolic antioxidants (20). Measurement herein of the classic antioxidant activity of the X-DMA SERMs in peroxyl radical trapping showed that the superior antioxidants were those able to undergo two-electron oxidation to quinoids (DMA, NH2-DMA; Table 1; Fig. 3); however, this assay did not distinguish between the remaining members of the family, which were poorer antioxidants. In the 1,1-diphenyl-2-picrylhydrazyl assay, antioxidant activity correlated linearly with calculated EHOMO values for the phenolic antioxidant X-DMA SERMs (Table 1; ref. 20). Zoete et al. (48) have previously correlated CD values for a family of NQO1 inducers with EHOMO values. Thus, the X-DMA redox activity data predict the relative induction of NQO1: DMA >> NH2-DMA >> arzoxifene > H-DMA > F-DMA > Br-DMA > SO2Me-DMA. This prediction was first tested in vitro in liver cell lines.

In vitro X-DMA Activity

In Hepa 1c1c7 cells, DMA was indeed observed to be an inducer of NQO1 and was observed to be the strongest of the X-DMA SERMs, and several other X-DMA SERMs reached significance versus control (Table 1). At higher concentration, all SERMs except raloxifene and HOT reached significance (P < 0.05); E2 had no effect on NQO1 activity. CD values also confirmed the prediction of DMA as the strongest NQO1 inducer; however, other X-DMA SERMs were also inducers and not readily differentiated.

It was postulated above that after bioactivation to a diquinone methide, the 2-phenyl-benzothiophene core of the X-DMA SERMs would provide the reactivity toward the Keap1 sensor required to induce NQO1. Therefore, BTC, constituting the core of the benzothiophene SERMs, was studied for comparison. In Hepa 1c1c7 cells, BTC was indeed observed to be a potent inducer of NQO1, with a CI value comparable to the chemopreventive agent sulforaphane (Table 1). A potential explanation for the weaker induction by DMA and raloxifene, both of which contain the BTC sub-structure, is that for the quinoid-forming compounds, direct interaction of the quinoid with the sensor protein is required for activation and induction; the steric hindrance due to the SERM side arm would inhibit interaction with the sensor protein relative to BTC.

The observation of significant activation of ARE by both BTC and DMA and induction of NQO1 in HepG2 cells (Fig. 4) strongly supports the concept that appropriate benzothiophene SERMs can act as chemopreventive agents via the Keap1/Nrf2/ARE pathway. It should be noted that the character of the quinoid metabolite is expected to be of importance: DMA and raloxifene are bioactivated to diquinone methides, which have different chemical characteristics to other quinoids. The results confirm that engineering of the SERM structure to modulate redox activity affects biological activity of potential therapeutic significance.

In vivo X-DMA Activity

On the basis of the in vitro data collected in murine and human hepatocellular culture, DMA would be predicted to act in vivo as a SERM with chemopreventive activity, in addition to any ER-dependent properties, derived from phase II enzyme induction via the Keap1/Nrf2/ARE pathway. DMA, arzoxifene, and F-DMA were selected for comparison of induction of hepatic NQO1 in the juvenile rat model, which is commonly used in assessing estrogenic and antiestrogenic activity of new SERMs (8, 49). It was found that all three X-DMA SERMs at 10 mg/kg/d induced NQO1 after only 3 days of treatment, with arzoxifene being the strongest inducer, elevating NQO1 activity 1.6-fold over vehicle control (Fig. 5B). Arzoxifene was originally developed as an improved SERM with increased bioavailability over raloxifene, and because DMA is known to be an arzoxifene metabolite, it is not surprising that arzoxifene and DMA induced NQO1 in vivo. In the same model, treatment with E2 showed a trend toward decreased NQO1 activity relative to vehicle. Increased NQO1 induction in response to all three X-DMA SERMs was significant relative to E2 treatment; furthermore, comparison of each SERM treatment group with the relevant SERM + E2 treatment group showed no significant difference (Fig. 5D). Hence, the induction of hepatic NQO1 in vivo by the X-DMA SERMs does not show estrogenic or antiestrogenic character and is therefore unlikely to be ER mediated. In accord with this assessment, the observed induction of NQO1 by X-DMA SERMs in MCF-7 cell lines stably transfected with ERα or ERβ was very modest and was not ER dependent. For example, SO2Me-DMA, which is a ligand for ERα (IC50 = 27 nmol/L) but not for ERβ (20), gave 1.2-fold induction of NQO1 in both ERα- and ERβ-positive MCF-7 cells. In ER(−) MCF-7 cells, DMA showed significant induction of NQO1, which is compatible with its ability to induce NQO1 via ER-independent activation of the ARE. The chemopreventive properties of arzoxifene toward ER(−) breast cancer have previously been reported in a mouse model (33).

This is the first report that the benzothiophene SERM arzoxifene can induce NQO1 in vivo. Given the interest in stimulation of the Nrf2/ARE signaling pathway and induction of NQO1 in cancer prevention, this result is potentially significant for the clinical use of arzoxifene as a chemopreventive agent. Comparison of the in vitro and in vivo data suggests that the arzoxifene metabolite DMA is responsible for the observed NQO1 induction via Nrf2/ARE activation, suggesting that arzoxifene might provide chemopreventive benefit over other SERMs via metabolism to DMA. In all assays done, the arzoxifene metabolite DMA was superior to raloxifene. The high activities of DMA and BTC support the hypothesis that oxidative bioactivation of benzothiophene SERMs to a quinoid metabolite can stimulate Nrf2/ARE signaling, likely via direct interaction with the Keap1 sensor. However, induction of NQO1, although weaker, was also observed for X-DMA SERMs, such as F-DMA, which cannot form quinoids but which are phenolic antioxidants. Although the evidence supports the stimulation of Nrf2/ARE signaling in induction of NQO1 by the X-DMA SERMs, further research is needed to differentiate the exact mechanisms invoked by quinoid-forming versus phenolic antioxidant SERMs, including the roles of mitogen-activated protein kinase and protein kinase C pathways. Induction of cytoprotective defenses in tissues to which SERMs are traditionally directed as cancer chemopreventive agents also warrants further study.

Grant support: NIH grants CA102590, CA102590-S1, and CA79870.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: B. Yu and B.M. Dietz contributed equally to this work.

We thank Rezene T. Asghodom for assistance in the animal study.

1
Katzenellenbogen BS, Katzenellenbogen JA. Biomedicine. Defining the “S” in SERMs.
Science
2002
;
295
:
2380
–1.
2
Evans DC, Watt AP, Nicoll-Griffith DA, Baillie TA. Drug-protein adducts: an industry perspective on minimizing the potential for drug bioactivation in drug discovery and development.
Chem Res Toxicol
2004
;
17
:
3
–16.
3
Dowers TS, Qin ZH, Thatcher GR, Bolton JL. Bioactivation of selective estrogen receptor modulators (SERMs).
Chem Res Toxicol
2006
;
19
:
1125
–37.
4
Yu L, Liu H, Li W, et al. Oxidation of raloxifene to quinoids: potential toxic pathways via a diquinone methide and o-quinones.
Chem Res Toxicol
2004
;
17
:
879
–88.
5
Liu H, Liu J, van Breemen RB, Thatcher GRJ, Bolton JL. Bioactivation of the selective estrogen receptor modulator desmethylated arzoxifene to quinoids: 4′-fluoro substitution prevents quinoid formation.
Chem Res Toxicol
2005
;
18
:
162
–73.
6
Bolton JL. Quinoids, quinoid radicals, and phenoxyl radicals formed from estrogens and antiestrogens.
Toxicology
2002
;
177
:
55
–65.
7
Jordan VC. Tamoxifen: a most unlikely pioneering medicine.
Nat Rev Drug Discov
2003
;
2
:
205
–13.
8
Jordan VC, Collins MM, Rowsby L, Prestwich G. A monohydroxylated metabolite of tamoxifen with potent antioestrogenic activity.
J Endocrinol
1977
;
75
:
305
–16.
9
Bergman L, Beelen ML, Gallee MP, Hollema H, Benraadt J, van Leeuwen FE. Risk and prognosis of endometrial cancer after tamoxifen for breast cancer. Comprehensive Cancer Centres' ALERT Group. Assessment of liver and endometrial cancer risk following tamoxifen.
Lancet
2000
;
356
:
881
–7.
10
Holtzclaw WD, Dinkova-Kostova AT, Talalay P. Protection against electrophile and oxidative stress by induction of phase 2 genes: the quest for the elusive sensor that responds to inducers.
Adv Enzyme Regul
2004
;
44
:
335
–67.
11
Riegel B, Wartman WB, Hill WT, Reeb BB, Shubik P, Stanger DW. Delay of methylcholanthrene skin carcinogenesis in mice by 1,2,5,6-dibenzofluorene.
Cancer Res
1951
;
11
:
301
–3.
12
Prestera T, Talalay P. Electrophile and antioxidant regulation of enzymes that detoxify carcinogens.
Proc Natl Acad Sci U S A
1995
;
92
:
8965
–9.
13
Prochaska HJ, De Long MJ, Talalay P. On the mechanisms of induction of cancer-protective enzymes: a unifying proposal.
Proc Natl Acad Sci U S A
1985
;
82
:
8232
–6.
14
Prochaska HJ, Bregman HS, De Long MJ, Talalay P. Specificity of induction of cancer protective enzymes by analogues of tert-butyl-4-hydroxyanisole (BHA).
Biochem Pharmacol
1985
;
34
:
3909
–14.
15
Nguyen T, Sherratt PJ, Huang HC, Yang CS, Pickett CB. Increased protein stability as a mechanism that enhances Nrf2-mediated transcriptional activation of the antioxidant response element. Degradation of Nrf2 by the 26 S proteasome.
J Biol Chem
2003
;
278
:
4536
–41.
16
Zhang DD, Hannink M. 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
2003
;
23
:
8137
–51.
17
Kwak MK, Wakabayashi N, Itoh K, Motohashi H, Yamamoto M, Kensler TW. Modulation of gene expression by cancer chemopreventive dithiolethiones through the Keap1-2 pathway. Identification of novel gene clusters for cell survival.
J Biol Chem
2003
;
278
:
8135
–45.
18
Delmas PD, Bjarnason NH, Mitlak BH, et al. Effects of raloxifene on bone mineral density, serum cholesterol concentrations, and uterine endometrium in postmenopausal women.
N Engl J Med
1997
;
337
:
1641
–7.
19
Suh N, Glasebrook AL, Palkowitz AD, et al. Arzoxifene, a new selective estrogen receptor modulator for chemoprevention of experimental breast cancer.
Cancer Res
2001
;
61
:
8412
–5.
20
Qin Z, Kastrati I, Chandrasena RE, et al. Benzothiophene selective estrogen receptor modulators with modulated oxidative activity and receptor affinity.
J Med Chem
2007
;
50
:
2682
–92.
21
Palkowitz AD, Glasebrook AL, Thrasher KJ, et al. Discovery and synthesis of [6-hydroxy-3-[4-2-(1-piperidinyl)ethoxy]phenoxy]-2-(4-hydroxyphenyl)]b enzo[b]thiophene: a novel, highly potent, selective estrogen receptor modulator.
J Med Chem
1997
;
40
:
1407
–16.
22
Nicolescu AC, Li Q, Brown L, Thatcher GRJ. Nitroxidation, nitration, and oxidation of a BODIPY fluorophore by RNOS and ROS.
Nitric Oxide
2006
;
15
:
163
–76.
23
Chen C, Yu R, Owuor ED, Kong AN. Activation of antioxidant-response element (ARE), mitogen-activated protein kinases (MAPKs) and caspases by major green tea polyphenol components during cell survival and death.
Arch Pharm Res
2000
;
23
:
605
–12.
24
Song LL, Kosmeder JW II, Lee SK, et al. Cancer chemopreventive activity mediated by 4′-bromoflavone, a potent inducer of phase II detoxification enzymes.
Cancer Res
1999
;
59
:
578
–85.
25
Prochaska HJ, Santamaria AB. Direct measurement of NAD(P)H:quinone reductase from cells cultured in microtiter wells: a screening assay for anticarcinogenic enzyme inducers.
Anal Biochem
1988
;
169
:
328
–36.
26
Lhoste EF, Gloux K, De Waziers I, et al. The activities of several detoxication enzymes are differentially induced by juices of garden cress, water cress and mustard in human HepG2 cells.
Chem Biol Interact
2004
;
150
:
211
–9.
27
Dietz BM, Kang YH, Liu G, et al. Xanthohumol isolated from Humulus lupulus inhibits menadione-induced DNA damage through induction of quinone reductase.
Chem Res Toxicol
2005
;
18
:
1296
–305.
28
Josephy DP, Guengerich PF, Miners JO. “Phase I and Phase II” drug metabolism: terminology that we should phase out?
Drug Metab Rev
2005
;
37
:
575
–80.
29
Kang YH, Pezzuto JM. Induction of quinone reductase as a primary screen for natural product anticarcinogens.
Methods Enzymol
2004
;
382
:
380
–414.
30
Fabian CJ, Kimler BF, Anderson J, et al. Breast cancer chemoprevention phase I evaluation of biomarker modulation by arzoxifene, a third generation selective estrogen receptor modulator.
Clin Cancer Res
2004
;
10
:
5403
–17.
31
Yuan X, Xu C, Pan Z, et al. Butylated hydroxyanisole regulates ARE-mediated gene expression via Nrf2 coupled with ERK and JNK signaling pathway in HepG2 cells.
Mol Carcinog
2006
;
45
:
841
–50.
32
Sanchez RI, Mesia-Vela S, Kauffman FC. Induction of NAD(P)H quinone oxidoreductase and glutathione S-transferase activities in livers of female August-Copenhagen Irish rats treated chronically with estradiol: comparison with the Sprague-Dawley rat.
J Steroid Biochem Mol Biol
2003
;
87
:
199
–206.
33
Liby K, Rendi M, Suh N, et al. The combination of the rexinoid, LG100268, and a selective estrogen receptor modulator, either arzoxifene or acolbifene, synergizes in the prevention and treatment of mammary tumors in an estrogen receptor-negative model of breast cancer.
Clin Cancer Res
2006
;
12
:
5902
–9.
34
Ansell PJ, Espinosa-Nicholas C, Curran EM, et al. In vitro and in vivo regulation of antioxidant response element-dependent gene expression by estrogens.
Endocrinology
2004
;
145
:
311
–7.
35
Jiang ZQ, Chen C, Yang B, Hebbar V, Kong AN. Differential responses from seven mammalian cell lines to the treatments of detoxifying enzyme inducers.
Life Sci
2003
;
72
:
2243
–53.
36
Bianco NR, Perry G, Smith MA, Templeton DJ, Montano MM. Functional implications of antiestrogen induction of quinone reductase: inhibition of estrogen-induced deoxyribonucleic acid damage.
Mol Endocrinol
2003
;
17
:
1344
–55.
37
Osborne CK, Hobbs K, Trent JM. Biological differences among MCF-7 human breast cancer cell lines from different laboratories.
Breast Cancer Res Treat
1987
;
9
:
111
–21.
38
Kwak MK, Wakabayashi N, Kensler TW. Chemoprevention through the Keap1-2 signaling pathway by phase 2 enzyme inducers.
Mutat Res
2004
;
555
:
133
–48.
39
Liu H, Bolton JL, Thatcher GRJ. Chemical modification modulates estrogenic activity, oxidative reactivity, and metabolic stability in 4′F-DMA, a new benzothiophene selective estrogen receptor modulator.
Chem Res Toxicol
2006
;
19
:
779
–87.
40
Moorthy B, Sriram P, Pathak DN, Bodell WJ, Randerath K. Tamoxifen metabolic activation: comparison of DNA adducts formed by microsomal and chemical activation of tamoxifen and 4-hydroxytamoxifen with DNA adducts formed in vivo.
Cancer Res
1996
;
56
:
53
–7.
41
Maiti S, Chen G. Tamoxifen induction of aryl sulfotransferase and hydroxysteroid sulfotransferase in male and female rat liver and intestine.
Drug Metab Dispos
2003
;
31
:
637
–44.
42
Cuendet M, Oteham CP, Moon RC, Pezzuto JM. Quinone reductase induction as a biomarker for cancer chemoprevention.
J Nat Prod
2006
;
69
:
460
–3.
43
Asher G, Lotem J, Cohen B, Sachs L, Shaul Y. Regulation of p53 stability and p53-dependent apoptosis by NADH quinone oxidoreductase 1.
Proc Natl Acad Sci U S A
2001
;
98
:
1188
–93.
44
Cornblatt BS, Ye L, Dinkova-Kostova AT, et al. Preclinical and clinical evaluation of sulforaphane for chemoprevention in the breast.
Carcinogenesis
2007
;
28
:
1485
–90.
45
Wakabayashi N, Dinkova-Kostova AT, Holtzclaw WD, et al. 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 U S A
2004
;
101
:
2040
–5.
46
Kang KW, Lee SJ, Kim SG. Molecular mechanism of nrf2 activation by oxidative stress.
Antioxid Redox Signal
2005
;
7
:
1664
–73.
47
Benson AM, Hunkeler MJ, Talalay P. Increase of NAD(P)H:quinone reductase by dietary antioxidants: possible role in protection against carcinogenesis and toxicity.
Proc Natl Acad Sci U S A
1980
;
77
:
5216
–20.
48
Zoete V, Rougee M, Dinkova-Kostova AT, Talalay P, Bensasson RV. Redox ranking of inducers of a cancer-protective enzyme via the energy of their highest occupied molecular orbital.
Free Radic Biol Med
2004
;
36
:
1418
–23.
49
Geiser AG, Hummel CW, Draper MW, et al. A new selective estrogen receptor modulator with potent uterine antagonist activity, agonist activity in bone, and minimal ovarian stimulation.
Endocrinology
2005
;
146
:
4524
–35.