Functional Implications of Antiestrogen Induction of Quinone Reductase: Inhibition of Estrogen-Induced Deoxyribonucleic Acid Damage

Functional Implications of Antiestrogen Induction ofQuinone Reductase: Inhibition of Estrogen-InducedDeoxyribonucleic Acid Damage

NICOLE R. BIANCO, GEORGE PERRY, MARK A. SMITH, DENNIS J. TEMPLETON, AND

MONICA M. MONTANO

Department of Pharmacology (N.R.B., M.M.M.) and Institute of Pathology (G.P., M.A.S.), CaseWestern Reserve University School of Medicine, Cleveland, Ohio 44106; and Department ofPathology (D.J.T.), University of Virginia Medical School, Charlottesville, Virginia 22908-0214

Recent studies have shown that the antiestrogenstamoxifen and raloxifene may protect againstbreast cancer, presumably because of a blockadeof estrogen receptor (ER)-mediated transcription.Another possible explanation is that antiestrogen-liganded ER transcriptionally induces genes thatare protective against cancer. We previously re-ported that antiestrogen-liganded ER� transcrip-tionally activates the major detoxifying enzymequinone reductase (QR) [NAD(P)H:quinone oxi-doreductase]. It has been established that metab-olites of estrogen, termed catecholestrogens, canform DNA adducts and cause oxidative DNA dam-age. We hypothesize that QR inhibits estrogen-induced DNA damage by detoxification of reactivecatecholestrogens. We report here that physiolog-ical concentrations of 17�-estradiol cause oxida-tive DNA damage, as measured by levels of 8-hydroxydeoxyguanine, in ER-positive MCF7 breastcancer cells, MDA-MB-231 breast cancer cells

(ER� negative/ER� positive) and nontumorigenicMCF10A breast epithelial cells (very low ER), whichis dependent on estrogen metabolism. Estrogen-induced 8-hydroxydeoxyguanine was inverselycorrelated to QR and ER� levels and was followedby downstream induction of the DNA repair en-zyme XPA. Trans-hydroxytamoxifen, raloxifene,and the pure antiestrogen ICI-182,780 protectedagainst estradiol-mediated damage in breast can-cer cells containing ER�. This is most likely due tothe ability of these antiestrogens to activate ex-pression of QR via ER�. We conclude that up-regulation of QR, either by overexpression or in-duction by tamoxifen, can protect breast cellsagainst oxidative DNA damage caused by estrogenmetabolites, representing a possible novel mech-anism of tamoxifen prevention against breastcancer. (Molecular Endocrinology 17: 1344–1355,2003)

ALTHOUGH IT IS widely accepted that the risk ofdeveloping breast cancer is directly related to

one’s lifetime exposure to estrogen, the precise role ofestrogen in the initiation and progression of breastcancer has yet to be determined. It has been hypoth-esized that initiation may result from induction of DNAdamage by estrogen metabolites and preexisting le-sions, although progression may be facilitated by es-trogen receptor (ER)-mediated up-regulation of mito-genic genes. There is already significant evidence thatestrogen metabolites are tumorigenic in animal mod-els (1–3). More recently, the direct action of estrogenhas been shown by its ability to transform normalbreast epithelial cells in culture (4).

In certain cell types, including both normal andbreast cancer cells, estrogens may be oxidized by

extrahepatic cytochrome P450 enzymes (mainlyCYP1A1 and CYP1B1) to hydroxy-catecholestrogensand further oxidized to the semiquinone and quinoneform (5–7). This metabolism is potentially harmful,given that the quinone-catecholestrogen can bind toDNA and form DNA adducts. Furthermore, redoxcycling between the quinone and unstable semiqui-none form causes hydroxyl radical formation thatcan lead to hydroxylated nucleotide bases [e.g. 8-hydroxydeoxyguanine (8-OHdG)] and permanent mu-tation, if not repaired (2, 3, 8). Glutathione-S-trans-ferase detoxifies these quinones by conjugation withglutathione, and catechol O-methyltransferase detoxi-fies the hydroxy-catecholestrogens by methylation (3).The estrogens and catecholestrogens can also be de-toxified by conjugation to glucuronides and sulfates,although the significance of these processes in thebreast is less clear (9). The ability of quinone reductase(QR) to detoxify quinone-catecholestrogens by reduc-tion of the reactive quinone-catecholestrogen back tothe hydroxy-catecholestrogen has been shown fortwo synthetic estrogens, diethylstilbestrol (10) and4-hydroxyequilenin-o-quinone (11). The resultinghydroxy-catecholestrogen is then available a second

Abbreviations: DMSO, Dimethylsulfoxide; DNase, deoxyri-bonuclease; E2, estradiol; EpRE, electrophile responseelement; ER, estrogen receptor; NAC, N-acetylcysteine;8-OHdG, 8-hydroxydeoxyguanine; 2- or 4-OH-E2, 2- or4-hydroxyestradiol; QR, quinone reductase; QRAS, QR anti-sense; RAL, raloxifene; RNase, ribonuclease; TOT, trans-hydroxytamoxifen; XPA, xeroderma pigmentosum comple-mentation group A.

0888-8809/03/$15.00/0 Molecular Endocrinology 17(7):1344–1355Printed in U.S.A. Copyright © 2003 by The Endocrine Society

doi: 10.1210/me.2002-0382

1344

time for conjugation and excretion. Alternatively, ifconjugation is impaired, then the metabolite may re-enter the redox cycle. Figure 8 includes a schematic ofestrogen metabolism for reference.

In 1998, the antiestrogen tamoxifen became the firstdrug to be approved for the reduction of risk for breastcancer, showing about a 50% reduction in both non-invasive and invasive breast cancer (12). A more re-cent Italian study has seen a dramatic decrease inbreast cancer specifically in high-risk women, basedon risk factors such as nulliparity and early age atmenarche, with a previous hysterectomy (13). Ralox-ifene (RAL), a similar selective ER modulator, may alsoprotect against breast cancer (14). Antiestrogens havetraditionally been thought to protect against breastcancer by blocking ER-mediated transcription of mi-togenic genes. However, it is also possible that an-tiestrogens protect cells against tumor-promotingevents by inducing detoxification enzymes such as QR.We have previously shown that antiestrogens induce QRvia transcriptional regulation by the ER (mainly ER�) atthe antioxidant response element/electrophile responseelement (EpRE) located in the promoter region of QR andother phase II detoxification genes (15, 16). Thus, ourprimary goal was to determine whether antiestrogenscould block estrogen-induced oxidative DNA damagevia activation of QR.

We report here that physiological concentrations ofestradiol (E2) cause an increase in oxidative DNA dam-age that is dependent upon estrogen metabolism inMCF-7 breast cancer cells (ER�/� positive) (17),MCF10A normal breast epithelial cells (very low ER)(18), and MDA-MB-231 cells (ER� negative/ER� pos-itive) (19). This damage is inversely related to QR andER� levels and can also be blocked by antiestrogensif ER� is present. We have also found a significantcorrelation between levels of E2-induced oxidativeDNA damage and the downstream induction of theDNA repair enzyme xeroderma pigmentosum comple-mentation group A (XPA).

RESULTS

Estrogen-Induced Oxidative DNA Damage inBreast Epithelial Cells Is Dependent on EstrogenMetabolism, But Not the ER

Although previous studies have shown estrogen-induced oxidative DNA damage in cultured breastcells, typically high concentrations of estrogen havebeen used (20, 21). Also, these studies have used invitro methods such as HPLC-electrochemical detec-tion to determine oxidative damage to DNA, which hasproduced controversial results (22). Thus, to deter-mine the effect of physiological concentrations of es-trogen, we treated breast cells with physiologicaldoses of E2 (10�10 to 10�8 M) and then measured theoxidative DNA marker 8-OHdG by quantitative immu-nocytochemistry. This is a previously established

method for quantifying relative increases in cellular8-OHdG levels (23, 24), and compared with methodsthat involve prior isolation and manipulation of DNA,this method does not create artificial oxidative modi-fication during the procedure. This method also al-lowed us to quantify relative 8-OHdG immunoreactiv-ity per cell rather than total 8-OHdG of a cellpopulation.

We see a dose-dependent increase in 8-OHdG lev-els over control after 24 h of estrogen treatment in ERpositive MCF7 breast cancer cells (Fig. 1A) with sig-nificance occurring at 10�9 M E2 (P � 0.01). Treatmentof MCF7 cells with the antioxidant N-acetylcysteine(NAC) prevents E2-induced accumulation of 8-OHdG,confirming an oxygen radical mediated process (Fig.1A). To determine whether E2-induced damage couldbe a result of ER-induced proliferation, we tested twoother cells lines, MCF10A nontumorigenic breast epi-thelial cells (very low ER) (18) and MDA-MB-231 breastcancer epithelial cells (ER� negative, ER� positive)(19). Neither cell line proliferates in response to E2 (25,26). However, E2 still induces 8-OHdG formation inthese cells (Fig. 1B). The E2-induced damage is alsotime dependent, with maximum levels of 8-OHdG oc-curring by 24 h post treatment in MCF7 and MCF10cells (Fig. 1C, top). There is generally a maximum 3- to4-fold enhancement of 8-OHdG in MCF7 cells after24 h and slightly less in MCF10A cells. Interestingly,MCF10A cells appear to accumulate oxidative dam-age much more slowly than MCF7 cells. The 8-OHdGin both cell lines drops back to basal levels by approx-imately 36 h post treatment. To determine whether thedecrease in damage at 36 h was from degradation ofthe estrogen, we supplied MCF7 cells with fresh mediaand estrogen after 36 h. The damage was not restoredafter 24 h of fresh estrogen treatment. When cells weretreated for 24 h and then removed from estrogen, themaximum levels of 8-OHdG slowly decreased back tobasal levels after approximately 24 h in both cell lines(Fig. 1C; bottom).

The 1F7 8-OHdG antibody recognizes both RNA-derived 8-hydroxyguanine and DNA-derived 8-OHdG(27). To verify that we were examining DNA oxidation,we pretreated the cells with deoxyribonuclease(DNase) or ribonuclease (RNase). We found the ma-jority of the E2-induced immunoreactivity in the nu-cleus, suggesting that DNA is the major target. Con-firming this hypothesis, DNase treatment significantlydepletes the nuclear staining, whereas RNase onlyslightly reduces the immunoreactivity (Fig. 1D).

Although the increase in 8-OHdG is independent ofER-mediated proliferation, it is dependent upon estro-gen metabolism to catecholestrogens. Treating thecells with the estrogen metabolism inhibitor �-naph-thoflavone, which inhibits CYP1A and 1B, blocked theE2 effect (Fig. 2A). �-Naphthoflavone or its solventdimethylsulfoxide (DMSO) had no effects on basal lev-els of 8-OHdG. The two hydroxy-catecholestrogenmetabolites formed in breast cells by CYP1A1 andCYP1B1 are 2- and 4-hydroxyestradiol (2- and 4-OH-

Bianco et al. • QR Inhibits E2-Induced DNA Damage Mol Endocrinol, July 2003, 17(7):1344–1355 1345

E2), respectively (3). In an effort to determine the cat-echolestrogen responsible for this damage, we treatedcells with increasing doses of either 2-OH-E2 or 4-OH-E2. Only the 4-OH-E2 metabolite was capable of in-ducing 8-OHdG in our system (Fig. 2B).

Antiestrogens and QR Protect against Estrogen-Induced Oxidative DNA Damage

Our previous studies have shown that antiestrogen-liganded ER can induce the detoxifying enzyme QR in

Fig. 1. Estrogen Induces Oxidative DNA Damage in Breast Epithelial CellsA, MCF7 breast cancer cells were grown on coverslips and treated with vehicle (control) or increasing doses of E2 for 24 h. The

cells were then immunostained for 8-OHdG using the 1F7 monoclonal antibody (1:100; Trevigen). Shown are the relative levelsof 8-OHdG after quantification. Values are the means � SE of three adjacent fields from three or more separate experiments. *,Level of significance P � 0.01 vs. control; **, level of significance P � 1 � 10�6 vs. control as determined by t test. As a controlfor oxidative damage, cells were treated with NAC (300 �M) � E2. B, Same experiments using MCF10A or MDA-MB-231 cellstreated with E2 (10�8 M). **, P � 0.001; *, P � 0.005 vs. control as determined by t test. C, Top, Time course of E2-induced oxidativedamage in MCF7 (�) and MCF10A (f) cells. Far right panel shows 36-h E2 treatment followed by 24-h treatment with fresh E2.Bottom, Withdrawal time course 24 h after E2 treatment. Cells were treated with 10�8 M E2 for 24 h, washed twice with Hanks’Balanced Saline Solution, and replaced with hormone-free media. Cells were then fixed at indicated time points and immuno-stained. D, Images of 1F7-stained MCF7 cells before and after 24-h E2 (10�8 M) treatment. As controls, cells were also treatedwith DNase or RNase before adding the primary antibody.

1346 Mol Endocrinol, July 2003, 17(7):1344–1355 Bianco et al. • QR Inhibits E2-Induced DNA Damage

ER-containing cells (15, 16). To determine the func-tional significance of this activation, we monitored theeffect of antiestrogens and QR on estrogen-inducedoxidative damage, because QR may be able to detox-ify the reactive catecholestrogen quinones (10). First,we sought to determine whether antiestrogens have aprotective effect on estrogen-induced DNA damage.For this and subsequent experiments, we used theoptimal concentration of E2 (10�8 M) required to induceoxidative DNA damage so that we may be able toaccurately test the inhibition of this process. MCF7,MDA-MB-231, and MCF10A cells were treated with E2

alone or in combination with the antiestrogens trans-hydroxytamoxifen (TOT), RAL, and ICI-182,780. An-tiestrogens show no protective effect on the levels ofE2-induced 8-OHdG in MCF10A cells, which have verylow levels of ER (Fig. 3B). In ER-positive MCF7 breastcancer cells, the antiestrogens protect against E2-mediated damage by approximately 50% (Fig. 3A).However, only the protection by TOT and RAL was

significant (P � 0.05). In MDA-MB-231 breast cancerepithelial cells (ER� negative, ER� positive), the an-tiestrogens still protect against oxidative damage, al-

Fig. 2. Estrogen-Induced Oxidative DNA Damage Is Depen-dent on Estrogen Metabolism

A, MCF7 cells were grown on coverslips and treated for24 h with 10�8 M E2 � �-naphthoflavone (�-naph) (5 �M inDMSO) and stained for 8-OHdG. As controls, cells weretreated with the vehicles ethanol (control) or DMSO. *, Levelof significance P � 0.00001 vs. control as determined by ttest. B, MCF7 cells were treated with vehicle (C) or increasingconcentrations (10�11 to 10�7 M) of 2-OH-E2 or 4-OH-E2 andstained for 8-OHdG. Shown are the relative levels of 8-OHdGafter quantification. Values are the means � SE of three ad-jacent fields from three or more separate experiments.

Fig. 3. Antiestrogens Protect against Estrogen-Induced Ox-idative DNA Damage in ER-Containing Cells

MCF7 (A), MCF10A (B), or MDA-MB-231 (C) cells grown oncoverslips were treated with vehicle (control) or E2 (10�8 M) �the antiestrogens (AE) TOT (10�7 M), ICI-182,780 (ICI; 10�7 M),or RAL (10�7 M) for 24 h and immunostained for 8-OHdG. Thedata are separated according to the AE used for that exper-iment. Shown are the relative levels of 8-OHdG after quanti-fication. Values are the means � SE of three adjacent fieldsfrom three or more separate experiments per treatmentgroup. *, Level of significance P � 0.05; **, P � 0.001 vs. E2

alone as determined by t test.


though only TOT treatment is considered significantlyprotective.

To determine whether QR can protect against E2-induced 8-OHdG, we transiently overexpressed orunderexpressed QR in MCF7 cells using the self-contained tetracycline-regulated pBPSTR1 vector (28)containing sense or antisense QR cDNA, respectively.In cells overexpressing QR (average increase, 40%),E2-induced 8-OHdG is blocked (Fig. 4A). Alternatively,cells with reduced QR (average decrease, 50%) con-tained significantly higher levels of 8-OHdG with alltreatments except for TOT (P � 0.05) (Fig. 4B). Thiseffect may be attributable to the ability of TOT toup-regulate QR or other detoxification genes also reg-ulated by the EpRE, thus negating the antisense effect.Supporting this hypothesis, TOT does not compen-sate for the increase in 8-OHdG levels in MCF10A cells(very low ER) (18) underexpressing QR (average de-crease, 40%) (Fig. 4C). Figure 5 shows the immuno-fluorescence confirming overexpression and underex-pression of QR in infected cells. We also reconfirmedTOT-induced QR expression using the same antibody(Fig. 5). For this and subsequent experiments, we onlyused TOT, because this was the antiestrogen used inthe initial studies regarding QR regulation.

ER� Is Necessary for Antiestrogen Protectionagainst Estrogen-Induced OxidativeDNA Damage

As mentioned, antiestrogens primarily activate tran-scription of QR preferentially via ER� (16). However,expression of ER� in MCF7 and other breast epithelialcell lines has been rather controversial. To specificallydetermine the levels of ER� in our cell lines, we im-munoblotted for ER� using the ER�-7B10.7 monoclo-nal antibody. As expected, ER� is detected in bothMCF7 and MDA-MB-231 cells, whereas MCF10A cellsdo not express the full-length form (Fig. 6A). Thesmaller band is similar in size to the ER-�2(�cx) splicevariant, an inactive dominant negative repressor ofER� (29, 30), although specific antibodies are neces-sary for such determination. We also confirmed theseresults using a different ER� monoclonal antibody(14C8, Novus Biologicals, Littleton, CO) (data notshown). These results are slightly different from thosereported by Fuqua et al. (17). They show a truncatedER� in MDA-MB-231 cells with their own monoclonalantibody. However, Girdler et al. (19) also observesfull-length ER� in MDA-MB-231 cells using anothernoncommercially available monoclonal antibody.

To evaluate the role of ER� in antiestrogen protectionagainst E2-induced oxidative damage, we again usedthe self-contained tetracycline-regulated pBPSTR1 vec-tor (28) to under- or overexpress ER� before E2 treat-ment in MCF7 cells. Figure 5 shows the reduced orincreased expression of ER� with ER� antisense/senseretroviral infection. Lowering ER� in MCF7 cells (averagedecrease, 40%) raises the levels of E2-induced 8-OHdG,and tamoxifen is significantly less protective against the

Fig. 4. QR Protects against Estrogen-Induced OxidativeDNA Damage

A, MCF7 cells were transiently infected with QR retrovi-ruses (�) or control retroviruses (f). The cells were thentreated with vehicle (Control), E2 (10�8 M), or the antiestrogenTOT (10�7 M) for 24 h and immunostained for 8-OHdG. *,Level of significance, P � 0.01 vs. vehicle-treated controlcells as determined by t test. B, Same experiment, exceptthat MCF7 cells were transiently infected with QRAS retrovi-ruses (�) or control retroviruses (f). a, Level of significanceP � 0.01 vs. respective control as determined by t test. b,Level of significance P � 0.05 vs. cells infected with controlretroviruses with same treatment as determined by t test. C,MCF10A cells were transiently infected with QRAS retrovi-ruses (�) or control retroviruses (f). *, Level of significanceP � 0.01 vs. cells infected with control retroviruses with sametreatment as determined by t test. Shown are the relativelevels of 8-OHdG after quantification. Values are the means �SE of three adjacent fields from two or more separate exper-iments.


damage (Fig. 6B). Interestingly, overexpressing ER� inMCF7 cells (average increase, 30%) (Fig. 5) preventsE2-induced DNA damage even without tamoxifen (Fig.6B). The same result occurs in MCF10A cells expressingER� (Fig. 6C).

Estrogen Treatment Induces XPA in BreastEpithelial Cells

To determine the possible downstream effects ofestrogen-induced DNA damage, we measured levelsof XPA, a member of the nucleotide excision repaircomplex, after E2 treatment. Although nucleotide ex-cision repair is more commonly associated with bulkyadduct repair, it has also recently been shown to repaircertain oxidative lesions, including 8-OHdG (31–36).We were able to detect about a 2-fold significant in-crease in XPA with E2 (10�8 M) (P � 0.01) in MCF7 cellsand slightly more in MCF10A cells (P � 0.05) (Fig. 7A).However, the two cell lines are not significantly differ-ent. Thus, the cells may be responding to the increasein DNA damage by activating certain repair enzymes.To determine whether induction of XPA follows8-OHdG induction, we again performed a time-coursestudy. XPA shows a slow induction with maximumlevels occurring 24 h after E2 treatment in MCF7 cells(Fig. 7B) vs. 8-OHdG, which is significantly increased4 h post treatment (Fig. 1C).

Because reduction of ER� leads to increased pro-duction of oxidative DNA damage in MCF7 cells (Fig.6), we tested what effect loss or gain of ER� wouldhave on levels of XPA in the cell. Interestingly, in ER�antisense-expressing cells, we see a significant rise inXPA levels to those of the E2-treated cells (Fig. 7C). Itshould also be noted that although TOT is able toblock the E2-induced XPA in control cells, TOT has noeffect in ER� antisense cells treated with E2. Overex-pression of ER� prevents the E2 increase in XPA.These are similar to the results obtained for 8-OHdG(Fig. 6). Taken together, this suggests that TOT re-quires ER� for protection against E2-mediated oxida-tive DNA damage and the resulting induction of XPA.

DISCUSSION

Four important conclusions can be drawn from thesestudies: 1) physiological concentrations of E2 lead tooxidative damage in MCF7 cells, MDA-MB-231 cells,

Fig. 5. Retroviral Infected Breast CellsTop, MCF7 cells treated with vehicle (Control) or TOT (10�7

M) for 24 h and immunostained for QR (1:100). Middle, first

and second rows, QRs or QRAS retrovirus-infected MCF7cells immunostained for QR (1:100). Third and fourth rows,ER�AS retrovirus-infected MCF7 cells immunostained forER� (1:100). Bottom, QRAS retrovirus-infected MCF10A cellsimmunostained for QR (1:100) and ER�S retrovirus-infectedMCF10A cells immunostained for ER� (1:100). Controlretroviruses were obtained from cells transfected with thepBPSTR1 vector alone.


and MCF10A cells; 2) antiestrogens can protectagainst such damage; 3) QR and ER� also protectagainst such damage, and ER� plays an essential rolein mediating antiestrogen protection; and 4) the E2-

induced DNA damage results in the downstream in-duction of the DNA repair enzyme XPA. Our proposedmodel for how this occurs is shown in Fig. 8.

It has been demonstrated extensively that estrogenexposure is linked to the overall risk of breast cancerin one’s lifetime. Most studies have focused on ER andthe transcriptional regulation of mitogenic genes thatcould be oncogenic directly or indirectly throughgrowth factors. Recently, there have been studiesshowing that estrogen itself may be a genotoxic mu-tagen capable of causing chromosomal mutations inanimals and cell culture (3, 4, 37), leading to tumori-genesis in various animal models (1, 3). Estrogen alsocauses MCF10 breast epithelial cells to take on atransformed phenotype independently of the ER (4).Specifically, the catecholestrogens 2-OH-E2 and4-OH-E2, which are formed in a cell-specific mannerby cytochrome P-450-catalyzed hydroxylation, areconsidered to be mutagenic (21). This is perhapsthrough the oxidation of catecholestrogens tosemiquinone radicals, and subsequently to quinones.Redox cycling between quinones and unstablesemiquinones causes hydroxyl radical formation thatcan lead to hydroxylated nucleotide bases (e.g.8-OHdG) and mutagenic DNA damage if not repaired(2, 3, 8). Also, the quinone-catecholestrogens can bindto DNA forming DNA adducts (37). Thus, initiation maybe due not only to ER-mediated proliferation, but alsoto DNA damage caused by a combination of estrogenmetabolism and preexisting lesions. Once initiated,the ER may then confer a selective advantage to thesepremalignant cells. QR may be linked to estrogen-mediated DNA damage by reducing the reactivequinone-catecholestrogen to the hydroxy-catecho-lestrogen allowing for conjugation and excretion.Thus, we chose to monitor whether QR could protectagainst estrogen-mediated oxidative DNA damage.We used 8-OHdG as a marker for oxidative damagebecause it is one of the most common oxidized basesand has demonstrated mutagenic potential (22).8-OHdG lesions result in mutational frequencies of1–5% (mainly G:C to T:A transitions) (38). 8-OHdGmay also be a prognostic marker in that both normaland malignant breast tissue from breast cancer pa-tients was shown to have higher levels of 8-OHdG thancontrol subjects (39, 40).

Before we examined the role of QR, we first neededto show that physiological concentrations of E2 inducedamage in breast cells. This was important becauseprevious studies in breast cells have used higher con-centrations of estrogen (4, 20, 21). Also, all of theanimal models have been developed using pharma-cological doses of E2 to obtain tumors in a short periodof time (3). At physiological concentrations, we areable to detect a dose-dependent increase in 8-OHdGover control, providing additional validity to the animalmodels. For reference, plasma concentrations of E2 inthe average female range between 10�10 and 10�9 M

(41). However, it should be noted that local concen-trations of E2 and its metabolites in the breast may be

Fig. 6. ER� Protects against Estrogen-Induced OxidativeDNA Damage

A, Western blot analysis for ER� in MCF10A, MDA-MB-231, and MCF7 cells. Recombinant ER� was loaded as acontrol. Ponceau S stain shows equal protein loading acrosslanes. B, MCF7 cells were transiently infected with ER�S

retroviruses (u), ER�AS retroviruses (�), or control retrovi-ruses (f). The cells were then treated with vehicle (Control),E2 (10�8 M), or the antiestrogen TOT (10�7 M) for 24 h andimmunostained for 8-OHdG. Shown are the relative levels of8-OHdG after quantification. a, Level of significance P � 0.05vs. vehicle-treated cells; b, P � 0.05 vs. E2 alone as deter-mined by t test. C, MCF10A cells were transiently infectedwith ER�S retroviruses (�) or control retroviruses (f). Thecells were then treated with vehicle (Control), E2 (10�8 M) �the antiestrogen TOT (10�7 M) for 24 h and immunostained for8-OHdG. Shown are the relative levels of 8-OHdG after quan-tification. *, P � 0.05 vs. control E2 as determined by t test.Values are the means � SE of three adjacent fields from twoor more separate experiments.


significantly higher due to production by aromataseand metabolism by the CYP1A/1B enzymes (3).

To verify an estrogen metabolism-mediated effect,we treated MCF7 cells with �-naphthoflavone, an in-hibitor of CYP1A and 1B that prevents catecholestro-gen formation. As expected, �-naphthoflavone re-verses the E2-induced damage. This is consistent withthe hamster model, in which �-naphthoflavone com-pletely suppresses E2-induced kidney carcinogenesis(42). Furthermore, our studies show that 4-OH-E2 isthe only major catecholestrogen capable of producingthe oxidative damage, because 2-OH-E2 has no effectin our system. These data support other in vitro and invivo studies showing 4-OH-E2 to be the more muta-genic and carcinogenic metabolite (1, 3, 37, 43). Thismay be due to an increased ability of the 4-OH-E2

metabolite to consume oxygen and generate DNAstrand breaks in the presence of Cu(II) (43) or thedecreased levels of inactivating conjugation com-pared with 2-OH-E2 (6). Unfortunately, we do not ob-serve as much damage from 4-OH-E2 as theoreticallyexpected on the basis of rates of estrogen metabolismin MCF7 cells. Thus, we cannot make a direct com-parison between E2- and 4-OH-E2-induced damage.Several possible reasons for this discrepancy exist,including autooxidation of applied 4-OH-E2 in the me-dia before it reaches the cell. Moreover, because ofredox cycling of the catecholestrogens, there may not

Fig. 7. Estrogen Induces XPA in Breast Epithelial CellsA, MCF7 (f) or MCF10A (�) cells grown on coverslips were

treated with vehicle (Control), E2 (10�8 M), or the antiestrogenTOT (10�7 M) for 24 h. The cells were then immunostained forXPA and quantified. Values are the means � SE of threeadjacent fields from three or more separate experiments. *,Level of significance P � 0.05; **, level of significance P �0.01 vs. respective vehicle-treated cells as determined by ttest. B, Time-course analysis of XPA induction. C, MCF7 cellswere transiently infected with ER�S retroviruses (u), ER�AS

retroviruses (�), or control retroviruses (f). The cells werethen treated with vehicle (Control), E2 (10�8 M), and/or theantiestrogen TOT (10�7 M) for 24 h and immunostained for8-OHdG. Shown are the relative levels of 8-OHdG after quan-tification. Values are the means � SE of three adjacent fieldsfrom two or more separate experiments. a, Level of signifi-cance P � 0.05 vs. respective vehicle-treated cells; b, P �

0.05 vs. E2 alone; c, level of significance P � 0.05 vs. cellsinfected with control retroviruses with same treatment asdetermined by t test.

Fig. 8. Proposed Model for Antiestrogen Protection againstE2-Induced Oxidative DNA Damage

Estrogen may be metabolized in breast cells to catecho-lestrogens. These compounds may then be further oxidizedto catecholestrogen-semiquinones (CE-SQ) and catecho-lestrogen-quinones (CE-Q). DNA damage may occur throughreactive oxygen species generated from redox cycling be-tween the CE-SQ and CE-Q or from direct reaction of theCE-Q with DNA forming DNA adducts. If not repaired, thisDNA damage could lead to tumor initiation. QR protectsagainst oxidative damage by directly reducing the CE-Q backto the hydroxy-catecholestrogen. Conjugation though meth-ylation, glucuronidation, or sulfation then occurs, allowing fordetoxification and excretion. Antiestrogens protect againstE2-induced oxidative damage by binding to the ER� andmediating transcriptional activation of the QR promoter EpREregulatory region.


be a 1:1 correspondence between 8-OHdG formed.Although we cannot rule out an ER-mediated effect inE2-induced damage, the increase in 8-OHdG was notdependent on ER-mediated proliferation becauseboth MCF10A and MDA-MB-231 cells also incur dam-age, neither of which proliferate in response to E2 (25,26). Additionally, this oxidative damage occurs in non-tumorigenic MCF10A cells, suggesting that damage isnot specific to cancer cells. Interestingly, we foundthat 8-OHdG accumulates much slower after E2 treat-ment in MCF710A cells compared with MCF7 cells.We hypothesize that this may be due to decreased E2

metabolism in MCF10A cells. Not only is the rate of E2

metabolism much slower in MCF10A cells, but theyalso contain lower amounts of CYP1B1 mRNA (7).

In 1998, the antiestrogen tamoxifen became the firstdrug to be approved for the reduction of risk for breastcancer, showing about a 50% reduction in both non-invasive and invasive breast cancer (12). Tamoxifenhas traditionally been thought to protect againstbreast cancer by blocking ER-mediated transcription.However, it is also possible that tamoxifen protectscells against tumor-promoting events by inducing de-toxification enzymes such as QR (15, 16). We did seeinhibition of E2-induced 8-OHdG with pharmacologi-cal concentrations of TOT, RAL, and ICI-182,780 butonly in cells containing ER, confirming a probable ER-mediated effect. However, we do not know whetherthe protective effect of tamoxifen is through QR aloneor other phase II detoxification genes that may also beregulated by ER at the EpRE. This is currently beingpursued in our laboratory.

Although tamoxifen does not induce damage in ER-containing MCF7 cells, there was a modest but sig-nificant induction of 8-OHdG in MCF10A cells. This isa controversial topic because one study reported8-OHdG formation by tamoxifen derivatives, althoughthis was an in vitro reaction (44). However, anotherstudy has shown that tamoxifen can reduce oxidativedamage to DNA in vitro and in mouse epidermis (45).In another similar study, estrogen was found to elevatesuperoxide anion and hydride radicals in mice uteri,whereas tamoxifen and ICI-182,780 decreased radicalformation (46). Studies wherein preisolated DNA isanalyzed using in vitro methods should be interpretedwith caution. Cell compartmentation and environment(e.g. metals) are critical to oxidative damage effects. Itis also possible that tamoxifen may cause oxidativedamage in ER-negative cells, while being protective inER-positive cells through induction of detoxificationenzymes such as QR.

As expected from previous reporter assays showingER� to be the active isoform on the EpRE (16), ER� issufficient for conferring antiestrogen protection inMDA-MB-231 cells that contain only ER�. Furtherconfirming the protective effects of ER�, antisenseinhibition of ER� leads to elevated levels of 8-OHdG inMCF7 cells, and tamoxifen was significantly less pro-tective against E2-induced 8-OHdG. Unexpectedly, byoverexpressing ER� in MCF10A or MCF7 cells, we

saw that E2-induced damage is prevented. This maybe due to low levels of ligand-independent activationof the overexpressed ER� at the EpRE, which we alsoobserve in our transfection reporter assays (data notshown). It is also possible that ER� has an additionalinnate protective role independent of ligand activation.

To more directly assess the role of QR in E2-inducedoxidative stress, we modulated the levels of QR activ-ity in MCF7 cells and then measured the levels of8-OHdG. Simply overexpressing QR inhibits E2-induced 8-OHdG. In contrast, reduction of QR activityby antisense inhibition leads to an increase in controland E2-induced 8-OHdG levels. However, 8-OHdG isnot elevated in TOT-treated QR antisense (QRAS) cells,perhaps because of compensation by TOT-inducedQR gene transcription. Further support for this hypoth-esis comes from the fact that TOT-treated QRAS cellsare susceptible to damage in MCF10A cells that havevery low ER and would not be able to compensate.Although the QR inhibitor dicoumarol increases basallevels of 8-OHdG significantly, it unexpectedly has noeffect with E2 or TOT (data not shown). We hypothe-size that this is due to additional nonspecific effects ofdicoumarol in our system, because dicoumarol inhibitsa number of enzymes (47). This may also explain why2(3)-t-butyl-4-hydroxyanisol and dicoumarol, whichstimulate or inhibit QR, respectively, both lowered tu-mor incidence by estrogen in the Syrian hamstermodel (48).

Finally, we measured levels of the DNA repair en-zyme XPA after E2 and found an increase subsequentto the increase in 8-OHdG, suggesting that the cellsare responding by increased repair. Further supportfor increased repair comes from the rapid decrease in8-OHdG levels between 24 and 36 h post treatment,which occurs promptly after XPA has reached full in-duction at 24 h. This repair readiness may also explainwhy 8-OHdG does not reaccumulate when fresh es-trogen is added after 36 h. We also found a significantrise in XPA in cells depleted of ER�, again suggestingthat ER� plays a protective role against DNA damageand the resulting induction of repair enzymes. This isimportant because the role of ER� in breast cancer isnot well established, although it may play a protectiverole against breast cancer (49, 50). The fact that thereis no further increase in XPA with E2 in ER�AS cellsrelative to untreated cells suggests to us that there isalready a maximal induction of XPA in these cells.Although the induction of XPA is rather modest, theinduction is similar to what is observed after 20 htreatment with cisplatin in an ovarian carcinoma cellline (51). XPA is involved in nucleotide excision repairof bulky nucleotide lesions and perhaps oxidative DNAdamage such as 8-OHdG (31–36). Thus, XPA may beinvolved in the repair of both oxidative damage and/orDNA adducts caused by estrogen.

Although tamoxifen significantly protects againstbreast cancer, there are also several drawbacks totamoxifen prophylactic use. These side effects haveincluded strokes, pulmonary embolisms, and an in-


creased risk of endometrial cancer (12, 52). RAL alsoappears to protect against breast cancer and estro-gen-induced DNA damage, and it may have fewer sideeffects, although thromboembolic disease is still aproblem (14). We believe the research presented heresupports two additional prophylactic strategies. First,drugs aimed at reducing extrahepatic catecholestro-gen formation (specifically 4-OH-E2) may be effectivein blocking the harmful by-products of catecholestro-gen metabolism, while still maintaining the beneficialeffects of estrogen. An additional therapy that is cur-rently being pursued is the induction of QR and otherphase II detoxification enzymes. Oltipraz is one suchdrug that is currently in clinical trials (53). We proposethat this therapy would be protective against estrogen-induced DNA damage in the breast and possibly otherestrogen target tissues that form catecholestrogens,as well as certain chemically induced cancers.

MATERIALS AND METHODS

Chemicals and Materials

Cell culture media was purchased from Life Technologies,Inc. (Grand Island, NY). Calf serum was from HyClone Lab-oratories, Inc. (Logan, UT), and fetal calf serum from AtlantaBiologicals (Norcross, GA). 17�-Estradiol, 2-OH-E2, 4-OH-E2, dicoumarol, �-naphthoflavone, NAC, and the antiestro-gens TOT and RAL were obtained from Sigma (St. Louis,MO). The antiestrogen ICI-182,780 was obtained from Tocris(Ballwin, MO).

Plasmid Construction

To make pBPSTR1-QR (sense or antisense), QR cDNA wasreleased from pMT2-QR (54) by NcoI/AflIII digestion, blunted,and inserted into BamHI-digested and blunted pBPSTR1vector (28). To make pBPSTR1-ER� (sense or antisense),ER� cDNA was released from Flag-ER��BSII-SK� (16) byNcoI/HindIII digestion, blunted, and inserted into BamHI-digested and blunted pBPSTR1 vector.

Tissue Culture and Retroviral-Mediated Transfection

Breast epithelial cells (MCF7, MDA-MB-231, and MCF10A)and PA317 amphotropic packaging cells were obtained fromAmerican Type Culture Collection (Manassas, VA) and main-tained according to their recommended protocols. Before theexperiments, breast epithelial cells were depleted of estrogenby growth in improved MEM minus phenol red containing5% charcoal dextran-treated calf serum for 5 d beforeexperiments.

Retroviruses were made by transfecting PA317 cells withthe pBPSTR1 plasmid alone or pBPSTR1 containing full-length QR or ER� in the sense or antisense orientation.Breast epithelial cell lines were infected with retrovirus-containing supernatants in the presence or absence of 3�g/ml tetracycline. The self-contained, tetracycline-regulatedretroviral vector pBPSTR1 contains both the response unit,composed of tetracycline resistance operon regulatory ele-ments (tetO) within a minimal cytomegalovirus promoter, andthe regulator unit, encoding the tTA protein (the tetracyclinerepressor fused to the transactivator protein VP16) (28). Geneexpression is inhibited by tetracycline, which binds the trans-activator protein tTA, causing it to dissociate from the tetO

minimal cytomegalovirus promoter. Changes in protein ex-pression were verified by immunofluorescence staining.Changes in protein levels of QR and ER� were also verified inMCF7 cells by Western blot analysis (data not shown).

Immunofluorescence Staining of Breast Cells

Cells grown on coverslips were fixed in 4% paraformalde-hyde. After blocking with 5% normal goat serum, sampleswere incubated with polyclonal rabbit QR antibody [1:100dilution (55)], 14C8 monoclonal ER� antibody (1:100 dilution;Novus Biologicals), or polyclonal ER� antibody (kindly pro-vided by the laboratory of Benita S. Katzenellenbogen, Uni-versity of Illinois, Champaign-Urbana, IL) and goat, anti-rabbit IgG Alexa 488 or antimouse IgG Alexa 594 fluorescentsecondary antibody (Molecular Probes, Inc., Eugene, OR). Asa negative control, cells were immunostained with nonspe-cific rabbit IgG or with secondary antibody alone. Semiquan-titative analysis was performed on a Macintosh computerusing Adobe Photoshop 6.0 software (Adobe Systems, SanJose, CA). Mean fluorescence intensity of 15–20 cells fromeach experiment was measured as previously referenced byour laboratory and others (by using the luminosity channel onthe histogram function) and averaged with background sub-tracted out from each field (56, 57). The analysis was per-formed a second time using the Zeiss KS300 Imaging System(Carl Zeiss Inc., Thornwood, NY) quantitation program withsimilar results.

Western Blot Analyses

Whole-cell extracts were prepared from breast epithelial cellsusing mammalian protein extraction reagent (Pierce Chemi-cal, Rockford, IL). Fifty micrograms of protein extract wereseparated by electrophoresis on 12% sodium dodecyl sul-fate-polyacrylamide gels and transferred electrophoreticallyonto nitrocellulose membranes. Recombinant human ER�(PanVera Corp., Madison, WI) was included as a positivecontrol. Blots were incubated with anti-ER� (long form)monoclonal antibody 7B10.7 (1:200 dilution; GeneTex, SanAntonio, TX) and goat antimouse IgG secondary antibody(1:3000 dilution) for detection by chemiluminescence (ECL,Amersham Biosciences, Inc., Piscataway, NJ).

Immunocytochemistry for 8-OHdG/XPA in Breast Cells

Cells grown on coverslips were fixed in methacarn (methanol-chloroform-acetic acid, 6:3:1) for 1 h at room temperature.Endogenous peroxidase activity in the cells was eliminatedby a 30-min incubation with 3% H2O2 in methanol, andnonspecific binding sites were blocked in a 15-min incuba-tion with 10% normal goat serum in Tris-buffered saline [150mM Tris-HCl and 150 mM NaCl (pH 7.6)]. The cells were thenpretreated with proteinase-K [20 �g/ml in PBS (pH 7.4)] for 15min at room temperature (Sigma, St. Louis, MO). To detectoxidized nucleosides, we used the anti-8-oxo-dG monoclo-nal antibody 1F7 (1:100; Trevigen, Gaithersburg, MD). As anegative control, cells were incubated without the primaryantibody. Immunostaining was developed by the peroxidase-antiperoxidase procedure. The 8-OHdG antibody also recog-nizes 8-hydroxyguanine (27), so to confirm that the oxidativedamage was mainly to DNA, some slides were preincubatedafter proteinase-K treatment with DNase I (10 U/�l in PBS for1 h at 37 C) (Roche Diagnostics, Indianapolis, IN) or RNase (5�g/ml in PBS for 1 h at 37 C) (Promega Corp., Madison, WI).For XPA staining, the XPA Ab-1 monoclonal antibody (1:50;Neomarkers, Fremont, CA) was used after the same protocolas used for 8-OHdG.


Relative Quantification of 8-OHdG/XPA

Immunoreactivity was evaluated by measuring OD as de-scribed (23, 24). The OD was assessed using a Zeiss Axio-cam digital camera (Carl Zeiss, Inc.) with a KS300 ImagingSystem quantitation program. The OD of manually outlinedcells was measured. Five cells in three adjacent fields weremeasured, and the background OD was subtracted fromeach. Each experiment was performed two or more times,and results were measured under the same optical and lightconditions. Also, an electronic shading correction was usedto compensate for any unevenness that might be present inthe illumination. Statistical analysis was performed using theStudent’s t test.

Acknowledgments

We thank Zvezdana Kubat, Sandra Siedlak, and XiaoyanSun for technical assistance. We also thank Dr. Steven A.Reeves (Massachusetts General Hospital) for the pBPSTR1retroviral vector and Dr. Anil Jaiswal (Baylor College of Med-icine) for the pMT2-QR vector.

Received November 18, 2002. Accepted April 7, 2003.Address all correspondence and requests for reprints to:

Dr. Monica M. Montano, Ph.D., Case Western Reserve Uni-versity School of Medicine, Department of Pharmacology,H. G. Wood Building W307, 2109 Adelbert Road, Cleveland,Ohio 44106. E-mail: [email protected].

This work was supported by NIH Grant CA80959(to M.M.M.) and Institutional National Research ServiceAward Predoctoral Fellowships T32 GM08056 and 5T32CA59366-10 (to N.R.B.).

REFERENCES

1. Newbold RR, Liehr JG 2000 Induction of uterine adeno-carcinoma in CD-1 mice by catechol estrogens. CancerRes 60:235–237

2. Liehr JG 1997 Dual role of oestrogens as hormones andpro-carcinogens: tumour initiation by metabolic activa-tion of oestrogens. Eur J Cancer Prev 6:3–10

3. Liehr JG 2000 Is estradiol a genotoxic mutagenic carcin-ogen? Endocr Rev 21:40–54

4. Russo J, Hu YF, Tahin Q, Mihaila D, Slater C, Lareef MH,Russo IH 2001 Carcinogenicity of estrogens in humanbreast epithelial cells. APMIS 109:39–52

5. Jefcoate CR, Liehr JG, Santen RJ, Sutter TR, Yager JD,Yue W, Santner SJ, Tekmal R, Demers L, Pauley R,Naftolin F, Mor G, Berstein L 2000 Tissue-specific syn-thesis and oxidative metabolism of estrogens. J NatlCancer Inst Monogr 27:95–112

6. Zhu BT, Conney AH 1998 Functional role of estrogenmetabolism in target cells: review and perspectives. Car-cinogenesis 19:1–27

7. Spink DC, Spink BC, Cao JQ, DePasquale JA, PentecostBT, Fasco MJ, Li Y, Sutter TR 1998 Differential expres-sion of CYP1A1 and CYP1B1 in human breast epithelialcells and breast tumor cells. Carcinogenesis 19:291–298

8. Thibodeau PA, Paquette B 1999 DNA damage inducedby catecholestrogens in the presence of copper (II): gen-eration of reactive oxygen species and enhancement byNADH. Free Rad Biol Med 27:1367–1377

9. Raftogianis R, Creveling C, Weinshilboum R, Weisz J2000 Estrogen metabolism by conjugation. J Natl CancerInst Monogr 27:113–124

10. Roy D, Liehr JG 1988 Temporary decrease in renal qui-none reductase activity induced by chronic administra-

tion of estradiol to male Syrian hamsters. Increased su-peroxide formation by redox cycling of estrogen. J BiolChem 263:3646–3651

11. Shen L, Pisha E, Huang Z, Pezzuto JM, Krol E, Alam Z,van Breemen RB, Bolton JL 1997 Bioreductive activationof catechol estrogen-ortho-quinones: aromatization ofthe B ring in 4-hydroxyequilenin markedly alters quinoidformation and reactivity. Carcinogenesis 18:1093–1101

12. Fisher B, Costantino JP, Wickerham DL, Redmond CK,Kavanah M, Cronin WM, Vogel V, Robidoux A, DimitrovN, Atkins J, Daly M, Wieand S, Tan-Chiu E, Ford L,Wolmark N 1998 Tamoxifen for prevention of breastcancer: report of the National Surgical Adjuvant Breastand Bowel Project P-1 Study. J Natl Cancer Inst 90:1371–1388

13. Veronesi U, Maisonneuve P, Rotmensz N, Costa A, Sac-chini V, Travaglini R, D’Aiuto G, Lovison F, Gucciardo G,Muraca MG, Pizzichetta MA, Conforti S, Decensi A, Rob-ertson C, Boyle P 2003 Italian randomized trial amongwomen with hysterectomy: tamoxifen and hormone-dependent breast cancer in high-risk women. J Natl Can-cer Inst 95:160–165

14. Cauley JA, Norton L, Lippman ME, Eckert S, Krueger KA,Purdie DW, Farrerons J, Karasik A, Mellstrom D, Ng KW,Stepan JJ, Powles TJ, Morrow M, Costa A, Silfen SL,Walls EL, Schmitt H, Muchmore DB, Jordan VC, Ste-Marie LG 2001 Continued breast cancer risk reduction inpostmenopausal women treated with raloxifene: 4-yearresults from the MORE trial. Multiple Outcomes of Ralox-ifene Evaluation. Breast Cancer Res Treat 65:125–134

15. Montano MM, Katzenellenbogen BS 1997 The quinonereductase gene: a unique estrogen receptor-regulatedgene that is activated by antiestrogens. Proc Natl AcadSci USA 94:2581–2586

16. Montano MM, Jaiswal AK, Katzenellenbogen BS 1998Transcriptional regulation of the human quinone reduc-tase gene by antiestrogen-liganded estrogen receptor-�and estrogen receptor-�. J Biol Chem 273:25443–25449

17. Fuqua SA, Schiff R, Parra I, Friedrichs WE, Su JL, McKeeDD, Slentz-Kesler K, Moore LB, Willson TM, Moore JT1999 Expression of wild-type estrogen receptor � andvariant isoforms in human breast cancer. Cancer Res59:5425–5428

18. Hu YF, Lau KM, Ho SM, Russo J 1998 Increased expres-sion of estrogen receptor � in chemically transformedhuman breast epithelial cells. Int J Oncol 12:1225–1228

19. Girdler F, Browell DA, Cunliffe WJ, Shenton BK, Hem-ming JD, Scorer P, Young JR, Brotherick I 2001 Use ofthe monoclonal antibody DAKO-ER� (8D5–1) to measureoestrogen receptor � in breast cancer cells. Cytometry45:65–72

20. Lavigne JA, Goodman JE, Fonong T, Odwin S, He P,Roberts DW, Yager JD 2001 The effects of catechol-O-methyltransferase inhibition on estrogen metabolite andoxidative DNA damage levels in estradiol-treated MCF-7cells. Cancer Res 61:7488–7494

21. Hiraku Y, Yamashita N, Nishiguchi M, Kawanishi S 2001Catechol estrogens induce oxidative DNA damage andestradiol enhances cell proliferation. Int J Cancer 92:333–337

22. Helbock HJ, Beckman KB, Ames BN 1999 8-Hydroxydeoxyguanosine and 8-hydroxyguanine as bi-omarkers of oxidative DNA damage. Methods Enzymol300:156–166

23. Nunomura A, Perry G, Aliev G, Hirai K, Takeda A, BalrajEK, Jones PK, Ghanbari H, Wataya T, Shimohama S,Chiba S, Atwood CS, Petersen RB, Smith MA 2001 Ox-idative damage is the earliest event in Alzheimer disease.J Neuropathol Exp Neurol 60:759–767

24. Nunomura A, Chiba S, Kosaka K, Takeda A, CastellaniRJ, Smith MA, Perry G 2002 Neuronal RNA oxidation isa prominent feature of dementia with Lewy bodies. Neu-roreport 13:2035–2039


25. Pilat MJ, Cristman JK, Brooks SC 1996 Characterizationof the estrogen receptor transfected MCF10A breast cellline 139B6. Breast Cancer Res Treat 37:253–266

26. Garcia M, Derocq D, Freiss G, Rochefort H 1992 Activa-tion of estrogen receptor transfected into a receptor-negative breast cancer cell line decreases the metastaticand invasive potential of the cells. Proc Natl Acad SciUSA 89:11538–11542

27. Nunomura A, Perry G, Pappolla MA, Wade R, Hirai K,Chiba S, Smith MA 1999 RNA oxidation is a prominentfeature of vulnerable neurons in Alzheimer’s disease.J Neurosci 19:1959–1964

28. Paulus W, Baur I, Boyce FM, Breakfield XO, Reeves SA1996 Self-contained, tetracline regulated retroviral vectorsystem for gene delivery to mammalian cells. J Virol70:62–67

29. Palmieri C, Cheng GJ, Saji S, Zelada-Hedman M, WarriA, Weihua Z, Van Noorden S, Wahlstrom T, CoombesRC, Warner M, Gustafsson J-A 2002 Estrogen receptor �in breast cancer. Endocr Relat Cancer 9:1–13

30. Tong D, Schuster E, Seifert M, Czerwenka K, Leodolte S,Zeillinger R 2002 Expression of estrogen receptor � iso-forms in human breast cancer tissues and cell lines.Breast Cancer Res Treat 71:249–255

31. Croteau DL, Bohr VA 1997 Repair of oxidative damage tonuclear and mitochondrial DNA in mammalian cells.J Biol Chem 272:25409–25412

32. Reardon JT, Bessho T, Kung HC, Bolton PH, Sancar A1997 In vitro repair of oxidative DNA damage by humannucleotide excision repair system: possible explanationfor neurodegeneration in xeroderma pigmentosum pa-tients. Proc Natl Acad Sci USA 94:9463–9468

33. Jaiswal M, Lipinski LJ, Bohr VA, Mazur SJ 1998 Efficientin vitro repair of 7-hydro-8-oxodeoxyguanosine by hu-man cell extracts: involvement of multiple pathways. Nu-cleic Acids Res 26:2184–2191

34. Driggers WJ, Grishko VI, LeDoux SP, Wilson GL 1996Defective repair of oxidative damage in the mitochondrialDNA of a xeroderma pigmentosum group A cell line.Cancer Res 56:1262–1266

35. Lipinski LJ, Hoehr N, Mazur SJ, Dianov GL, Senturker S,Dizdaroglu M, Bohr VA 1999 Repair of oxidative DNAbase lesions induced by fluorescent light is defective inxeroderma pigmentosum group A cells. Nucleic AcidsRes 27:3153–3158

36. Klein JC, Bleeker MJ, Saris CP, Roelen HC, Brugghe HF,van den Elst H, van der Marel GA, van Boom JH, WestraJG, Kriek E 1992 Repair and replication of plasmids withsite-specific 8-oxodG and 8-AAFdG residues in normaland repair-deficient human cells. Nucleic Acids Res 20:4437–4443

37. Cavalieri E, Frenkel K, Liehr JG, Rogan E, Roy D 2000Estrogens as endogenous genotoxic agents–DNA ad-ducts and mutations. J Natl Cancer Inst Monogr 27:75–93

38. Shibutani S, Grollman AP 1994 Miscoding during DNAsynthesis on damaged DNA templates catalysed bymammalian cell extracts. Cancer Lett 83:315–322

39. Li D, Zhang W, Zhu J, Chang P, Sahin A, Singletary E,Bondy M, Hazra T, Mitra S, Lau SS, Shen J, DiGiovanniJ 2001 Oxidative DNA damage and 8-hydroxy-2-deox-yguanosine DNA glycosylase/apurinic lyase in humanbreast cancer. Mol Carcinog 31:214–223

40. Musarrat J, Arezina-Wilson J, Wani AA 1996 Prognosticand aetiological relevance of 8-hydroxyguanosine inhuman breast carcinogenesis. Eur J Cancer 32A:1209–1214

41. 1998 Williams textbook of endocrinology. In: Wilson JD,Foster DW, Kronenberg H, Larsen PR, eds. 9th ed.Philadelphia: WB Saunders Company

42. Liehr JG, Gladek A, Macatee T, Randerath E, RanderathK 1991 DNA adduct formation in liver and kidney of maleSyrian hamsters treated with estrogen and/or �-naph-thoflavone. Carcinogenesis 12:385–389

43. Thibodeau PA, Kachadourian R, Lemay R, Bisson M, DayBJ, Paquette B 2002 In vitro pro- and antioxidant prop-erties of estrogens. J Steroid Biochem Mol Biol 81:227–236

44. Okubo T, Nagai F, Ushiyama K, Yokoyama Y, Ozawa S,Kano K, Tomita S, Kubo H, Kano I 1998 DNA cleavageand 8-hydroxydeoxyguanosine formation caused by ta-moxifen derivatives in vitro. Cancer Lett 122:9–15

45. Wei H, Cai Q, Tian L, Lebwohl M 1998 Tamoxifen re-duces endogenous and UV light-induced oxidative dam-age to DNA, lipid and protein in vitro and in vivo. Carci-nogenesis 19:1013–1018

46. Jain S, Saxena D, Kumar PG, Koide SS, Laloraya M 1999Effect of estradiol and selected antiestrogens on pro-and antioxidant pathways in mammalian uterus. Contra-ception 60:111–118

47. Ross D, Siegel D, Beall H, Prakash AS, Mulcahy RT,Gibson NW 1993 DT-diaphorase in activation and detox-ification of quinones. Bioreductive activation of mitomy-cin C. Cancer Metastasis Rev 12:83–101

48. Roy D, Liehr JG 1990 Inhibition of estrogen-inducedkidney carcinogenesis in Syrian hamsters by modulatorsof estrogen metabolism. Carcinogenesis 11:567–570

49. Gustafsson J, Warner M 2000 Estrogen receptor � in thebreast: role in estrogen responsiveness and develop-ment of breast cancer. J Steroid Biochem Mol Biol 74:245–248

50. Mann S, Laucirica R, Carlson N, Younes PS, Ali N,Younes A, Li Y, Younes M 2001 Estrogen receptor �expression in invasive breast cancer. Hum Pathol 32:113–118

51. Hector S, Bolanowska-Higdon W, Zdanowicz J, Hitt S,Pendyala L 2001 In vitro studies on the mechanisms ofoxaliplatin resistance. Cancer Chemother Pharmacol 48:398–406

52. Bange J, Zwick E, Ullrich A 2001 Molecular targets forbreast cancer therapy and prevention. Nat Med7:548–552

53. Kensler TW, Curphey TJ, Maxiutenko Y, Roebuck BD2000 Chemoprotection by organosulfur inducers ofphase 2 enzymes: dithiolethiones and dithiins. DrugMetabol Drug Interact 17:3–22

54. Shaw PM, Reiss A, Adesnik M, Nebert DW, Schembri J,Jaiswal AK 1991 The human dioxin-inducible NAD(P)H:quinone oxidoreductase cDNA-encoded protein ex-pressed in COS-1 cells is identical to diaphorase 4. EurJ Biochem 195:171–176

55. Raina AK, Templeton DJ, Deak JC, Perry G, Smith MA1999 Quinone reductase (NQO1), a sensitive redox indi-cator, is increased in Alzheimer’s disease. Redox Rep4:23–27

56. Bianco NR, Montano MM 2002 Regulation of prothymo-sin � by estrogen receptor �: molecular mechanisms andrelevance in estrogen-mediated breast cell growth. On-cogene 21:5233–5244

57. Chai Y, Shao J, Miller VM, Williams A, Paulson HL 2002Live-cell imaging reveals divergent intracellular dynamicsof polyglutamine disease proteins and supports a se-questration model of pathogenesis. Proc Natl Acad SciUSA 99:9310–9315


Documents

Functional Implications of Antiestrogen Induction of Quinone Reductase: Inhibition of Estrogen-Induced Deoxyribonucleic Acid Damage