A

(fa4cmaWwl©

K1

1

rt

C4eP2c

0d

Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 150–158

Estrogen metabolism and formation of estrogen-DNA adductsin estradiol-treated MCF-10F cells

The effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin inductionand catechol-O-methyltransferase inhibition

Fang Lu, Muhammad Zahid, Muhammad Saeed, Ercole L. Cavalieri, Eleanor G. Rogan ∗

Eppley Institute for Research in Cancer and Allied Diseases, University of Nebraska Medical Center,986805 Nebraska Medical Center, Omaha, NE 68198-6805, United States

Received 27 October 2006; accepted 21 December 2006

bstract

Formation of estrogen metabolites that react with DNA is thought to be a mechanism of cancer initiation by estrogens. The estrogens estroneE1) and estradiol (E2) can form catechol estrogen (CE) metabolites, catechol estrogen quinones [E1(E2)-3,4-Q], which react with DNA toorm predominantly depurinating adducts. This may lead to mutations that initiate cancer. Catechol-O-methyltransferase (COMT) catalyzesn inactivation (protective) pathway for CE. This study investigated the effect of inhibiting COMT activity on the levels of depurinating-OHE1(E2)-1-N3Ade and 4-OHE1(E2)-1-N7Gua adducts in human breast epithelial cells. MCF-10F cells were treated with TCDD, aytochrome P450 inducer, then with E2 and Ro41-0960, a COMT inhibitor. Estrogen metabolites and depurinating DNA adducts in cultureedium were analyzed by HPLC with electrochemical detection. Pre-treatment of cells with TCDD increased E2 metabolism to 4-OHE1(E2)

nd 4-OCH3E1(E2). Inclusion of Ro41-0960 and E2 in the medium blocked formation of methoxy CE, and depurinating adducts were observed.ith Ro41-0960, more adducts were detected in MCF-10F cells exposed to 1 �M E2, whereas without the inhibitor, no increases in adducts

ere detected with E2 ≤ 10 �M. We conclude that low COMT activity and increased formation of depurinating adducts can be critical factors

eading to initiation of breast cancer.2007 Elsevier Ltd. All rights reserved.

eywords: Estrogen metabolism; Catechol-O-methyltransferase; 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD); Depurinating estrogen-DNA adducts; MCF-

ed

0F cells

. Introduction

A variety of evidence shows that estrogens play a majorole in the etiology of breast cancer [1,2]. In extrahepaticissues, phase I [cytochrome P450 (CYP)1A1 and CYP1B1]

Abbreviations: E1(E2), estrone (estradiol); CE, catechol estrogen(s);E-3,4-Q, catechol estrogen-3,4-quinone(s); 2- and 4-OHE1(E2), 2- and-hydroxyestrone(estradiol); 2- and 4-OCH3E1(E2), 2- and 4-methoxy-strone(estradiol); COMT, catechol-O-methyltransferase; CYP, cytochrome450; Ro41-0960, 2′-fluoro-3,4-dihydroxy-5-nitrobenzophenone; TCDD,,3,7,8-tetrachlorodibenzo-p-dioxin; HPLC (ECD), high pressure liquidhromatography (with multi-channel electrochemical detection)∗ Corresponding author. Fax: +1 402 559 8068.

E-mail address: egrogan@unmc.edu (E.G. Rogan).

[4itciepNsmt

960-0760/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.oi:10.1016/j.jsbmb.2006.12.102

nzymes predominantly metabolize estrone (E1) and estra-iol (E2) to 2- and 4-catechol estrogens (CE), respectively3,4]. It has been suggested that these metabolites, 2- and-hydroxyestrone (estradiol) [2- and 4-OHE1(E2)], if formedn the breast, can become endogenous ultimate carcinogenshat damage DNA and generate mutations leading to breastancer [2,5–7]. The CE, particularly 4-OHE1(E2), maynitiate estrogen carcinogenesis after metabolism to catecholstrogen quinones (CE-Q) that can react with DNA to formredominantly the depurinating adducts, 4-OHE1(E2)-1-

3Ade and 4-OHE1(E2)-1-N7Gua, which generate apurinic

ites [2,5,6]. In turn, these apurinic sites can give rise toutations that induce cell transformation [8–10] and initiate

he series of events leading to cancer [2,5–7]. On the other

stry &

haama

sgtaamsGriesa

oaUpcChimvorCDC

F. Lu et al. / Journal of Steroid Biochemi

and, phase II [catechol-O-methyltransferase (COMT)nd NADPH quinone oxidoreductase 1 (NQO1)] enzymesre involved in the inactivation of the oxidative estrogenetabolites by COMT-catalyzed methylation of the CE [11]

nd the reduction of quinones back to CE by NQO1 [12].COMT is an intracellular enzyme that is present in both

oluble and membrane-bound forms encoded by the sameene with different transcription start sites [13] and withhe soluble form predominating in most organs [11]. COMTctivity varies among individuals [14]. This activity can beltered by either endogenous factors such as genetic poly-orphisms and levels of expression or exogenous factors

uch as induction or inhibition by environmental compounds.enetic epidemiology studies have proposed a possible cor-

elation between the low activity allele (COMTLL) and

ncreased breast cancer risk [15–17]. However, results frompidemiological studies have been inconsistent [18–20]. Aingle nucleotide substitution in codon 108 causes an aminocid transition (Val/Met), which results in a high (Val/Val)

C(

(

Fig. 1. COMT inhibition results in decreased inactivation of CEs, which in turn m

Molecular Biology 105 (2007) 150–158 151

r low (Met/Met) activity form of the COMT enzyme withthree to fourfold difference in activity [21]. One-fourth ofnited States Caucasians are homozygous for the val108metolymorphism in the COMT gene [21,22]. COMT activityan also be inhibited by substrate competition for the enzyme.ertain catechol metabolites of polychlorinated biphenylsave been shown to inhibit COMT activity [23]. Ro41-0960s a nitrocatechol-type inhibitor of COMT that fully inhibitsethylation of CE [11]. COMT inhibition decreases inacti-

ation of CE, which in turn may lead to increased formationf CE-Q and DNA damage that initiates cancer. However, theelationship has not been established between inhibition ofOMT activity and increased levels of CE and depurinatingNA adducts after exposure to E2. We hypothesize that lowOMT activity and increased formation of DNA adducts by

E-Q may be critical factors in the initiation of breast cancer

Fig. 1).The MCF-10F cell line [aromatic hydrocarbon receptor

AhR) positive, but estrogen receptor-� (ER�) negative] is a

ay lead to increased depurinating DNA adducts that can initiate cancer.

1 stry &

ggoctwom1Rnit

2

2

IwOEma[IS

2

M(thtDL[m

2

Ttbdt(daab

iapasrA

24

wocRpo(t4iwwOota2wCr5p(eTrrplpnut

2

ppw

52 F. Lu et al. / Journal of Steroid Biochemi

ood experimental model for researching estrogen carcino-enesis by a non-ER mediated pathway [8–10]. Expressionf CYP enzymes that metabolize E2 to CE in MCF-10Fells is very low, unless they are induced by 2,3,7,8-etrachlorodibenzo-p-dioxin (TCDD) [24]. Thus in this study,e pretreated these cells with TCDD to enhance formationf CE. To assess the effects of altered COMT activity on E2etabolism and formation of estrogen-DNA adducts, MCF-

0F cells were grown in the presence of the COMT inhibitoro41-0960. The profile of estrogen metabolites and depuri-ating estrogen-DNA adducts was analyzed to gain morensight into the mechanism by which COMT plays a pro-ective role in estrogen initiation of breast cancer.

. Materials and methods

.1. Chemicals

TCDD (>99% pure) was purchased from AccuStandardnc. (New Haven, CT); 2-OCH3E1(E2) and 4-OCH3E1(E2)ere obtained from Steraloids Inc. (Newport, RI). 2-HE1(E2) and 4-OHE1(E2) were synthesized by reacting2 with 2-iodoxybenzoic acid (IBX) and then separating theixture by HPLC as described [25]. The depurinating DNA

dduct standards were synthesized by published procedures26,27]. Cell culture media were purchased from Mediatechnc. (Herndon, VA). All other chemicals were obtained fromigma (St. Louis, MO).

.2. Cell lines and cell culture

MCF-10F cells were obtained from the ATCC (Rockville,D) and cultured in DMEM and Ham’s F12 medium

Mediatech, Inc.) with 20 ng/ml epidermal growth fac-or, 0.01 mg/ml insulin, 500 ng/ml hydrocortisone, 5%orse serum and 100 �g/ml penicillin/streptomycin mix-ure. Estrogen-free medium was prepared in phenol red-freeME/F12 medium with charcoal-stripped FBS (HyClone,ogan, UT). Cell viability was determined by the MTT

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bro-ide] assay [28].

.3. Western blots

Cell cultures were exposed to various concentrations ofCDD for 72 h and 4-OHE2 or Ro41-0960 for 24 h. After

reatment, cells were harvested and then lysed in RIPAuffer with protease inhibitor. Nuclei and unlysed cellularebris were removed by centrifugation. Protein concentra-ions were determined by using the BCA protein assay kitPierce Biotechnology, Rockford, IL). Western blot proce-

ures were previously described [24]. Dilutions of primarynti-CYP1A1, CYP1B1, �-actin (Genetest, Bedford, MA)nd COMT (Chemicon International, Temecula, CA) anti-odies were made in blocking solution (3% non-fat dry milk

1OTc

Molecular Biology 105 (2007) 150–158

n PBS). The blots were incubated for 3 h with primaryntibody and for 1 h with secondary antibody at room tem-erature. After each step, blots were washed with PBST (PBSnd 0.1% Tween-20), incubated with ECL solution (Amer-ham Biotech, Piscataway, NJ) for 1 min, and visualized withadiographic film. Intensities of the bands were quantified bylpha DigiDoc 1201 (Alpha Innotech, San Leandro, CA).

.4. Determination of COMT activity by HPLC using-OHE2 as the substrate

COMT enzymatic activity was determined by using HPLCith electrochemical detection (ECD), as described previ-usly [13,29] with minor modifications. Cellular COMT fromontrol and TCDD (0.1–30 nM), 4-OHE2 (0.1–30 �M) oro41-0960 (0.1–30 �M) treated MCF-10F cells was pre-ared, as described above. Enzyme reactions were carriedut in a final volume of 250 �l of 0.1 M sodium phosphatepH 7.8) containing 1 mM DTT, 5 mM MgCl2, cytosolic pro-eins (50 �g), varying amounts (0–100 �M) of CE substrate-OHE2, and 200 �M S-adenosylmethionine (SAM, saturat-ng). The incubation mixture, except for the CE substrate,as pre-incubated for 3 min at 37 ◦C. Then, the reactionas initiated by adding 5 �l of various concentrations of 4-HE2 in DMSO and terminated after 27 min by adding 25 �lf 4 M perchloric acid. Following centrifugation to precipi-ate proteins, the supernatant (245 �l) was passed through

5000 M.W. cut-off filter (Millipore, Billerica, MA), and00 �l of each sample was analyzed by HPLC. The extractsere subjected to HPLC analysis on a reverse phase Luna-218 column (250 mm × 4.6 mm, 5 �m; Phenomenex, Tor-

ance, CA) in a system equipped with dual ESA Model80 solvent delivery modules, an ESA Model 540 autosam-ler and a 12-channel CoulArray electrochemical detectorESA, Chelmsford, MA). The serial array of 12 coulometriclectrodes was set at potentials between −35 and 690 mV.he data were acquired and processed using the CoulAr-

ay software package (ESA). Peaks were identified by bothetention time and peak height ratios between the dominanteak and the peaks in the two adjacent channels. Metabo-ites and DNA adducts were quantified by comparison ofeak response ratios with known amounts of standards. Whenecessary, the identity of the analytes was confirmed byltraperformance liquid chromatography/tandem mass spec-rometry (UPLC/MS/MS).

.5. Inhibition of COMT activity by Ro41-0960 in vitro

Cytosolic proteins (50 �g from MCF-10F cell lysate) werere-incubated at 37 ◦C in assay buffer [200 mM potassiumhosphate (pH 7.8), 5 mM MgCl2, 1 mM DTT, 0.2 mM SAM]ith various concentrations of Ro41-0960 (0.1, 1, 3, and

0 �M) or DMSO for 7 min, then the substrate, 0.1 mM 4-HE2 was added and the mixture was incubated for 23 min.he reactions were terminated by adding 25 �l of 4 M per-hloric acid, and the samples were analyzed as described

stry &

awia

2d

ta1Tlww2ca(aommtPa

2

bP

3

3W

eviC41ate

Frf1fs

F. Lu et al. / Journal of Steroid Biochemi

bove. The inhibition of COMT activity by Ro41-0960as compared to the DMSO control, and the remain-

ng activity was expressed as percent of control enzymectivity.

.6. HPLC analysis of estrogen metabolites andepurinating DNA adducts

To determine the effect of TCDD on estrogen metabolism,he medium was replaced with estrogen-free medium for 12 hnd the cells exposed to 0–30 nM TCDD for 72 h. Then,�M E2 was added and incubation was continued for 24 h.o investigate the effect of inhibiting COMT activity on the

evels of CE and depurinating DNA adducts, MCF-10F cellsere pretreated with TCDD (10 nM) for 72 h, and then treatedith E2 (0.1–30 �M) with or without 3 �M Ro41-0960 for4–72 h. To directly determine the relationship between theoncentration of 4-OHE2 and formation of depurinating DNAdducts, cells were treated with increasing concentrations0–30 �M) of 4-OHE2 for 24 h. The media were collectednd 2 mM ascorbic acid added to protect E2 metabolites fromxidative degradation. The assay of metabolism of E2 was

odified from previously described procedures [30]. In brief,edia were processed by various concentration methods and

he methanol/water mixtures were applied to a Certify II Sep-ak cartridge. The extracts were subjected to HPLC analysis,s described above.

w(iTi

ig. 2. (a) Western blot of COMT protein expression in MCF-10F cells treated withepresentative immunoblots (from three replicates) demonstrate that anti-COMT anorm with the soluble form predominating in MCF-10F cells. Each lane contains 30201 and normalized to �-actin. (b) TCDD-induced metabolism of E2 to 4-OCH3E1

or 72 h and then with 1 �M E2 for 24 h. Extracts from cell culture medium were anix experiments.

Molecular Biology 105 (2007) 150–158 153

.7. Statistical analysis

The statistical significance of the results was determinedy Student’s-test and ANOVA analysis by using SAS andrizm software.

. Results

.1. Expression and activity of COMT determined byestern blot and HPLC using MCF-10F cellular extracts

The phase II enzyme COMT is considered to be a keynzyme in decreasing the effects of 4-OHE1(E2) by con-erting the CE into the corresponding methyl ether. COMTs not easily induced or suppressed. In the present study,OMT protein levels were measured in TCDD (0–30 nM),-OHE2 (0–60 �M) or Ro41-0960 (0–30 �M) treated MCF-0F cells by Western blot. COMT is expressed predominantlys S-COMT in most human tissues examined, and we foundhat S-COMT predominated in the MCF-10F cells. The lev-ls were consistent with reported levels [31]. Treatmentith TCDD did not increase the level of COMT proteins

data not shown). MCF-10F cells were treated with increas-ng concentrations of 4-OHE2 (1–60 �M) with or withoutCDD (10 nM) pretreatment for 24 h. COMT was detectable

n untreated MCF-10F cells and protein levels were not

increasing concentrations of 4-OHE2 with or without TCDD for 24 h. Thetibody recognizes both the 30 kD membrane-bound form and 25 kD soluble

�g of cell lysate. Intensity of the bands was quantified by Alpha DigiDoc(E2) and 2-OCH3E1(E2). MCF-10F cells were treated with 0–30 nM TCDDalyzed by HPLC. Data represent the mean ± S.D. of triplicate cultures from

154 F. Lu et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 150–158

Table 1Profile of estrogen metabolism in E2-treated MCF-10F cells with or without Ro41-0960a

E2 (without COMT inhibitor) (pmol/106 cells) E2 + Ro41-0960 (with COMT inhibitor) (pmol/106 cells)

0.1 �M 1 �M 10 �M 30 �M 0.1 �M 1 �M 10 �M 30 �M

E1(E2) 57 (±4)b 151 (±7) 354 (±9) 705 (±18) 44 (±5) 124 (±14) 268 (±10) 375 (±13)4-OHE1(E2) 0.11 (±0.01) 0.21 (±0.04) 1.45 (±0.24) 4.0 (±0.4) 0.17 (±0.03) 0.73 (±0.08) 2.71 (±0.25) 6.68 (±0.38)2-OHE1(E2) 0.05 (±0.01) 0.10 (±0.12) 0.66 (±0.07) 1.63 (±0.21) 0.09 (±0.01) 0.26 (±0.05) 1.28 (±0.18) 2.72 (±0.16)4-OCH3E1(E2) 1.42 (±0.16) 5.18 (±0.60) 11.97 (±1.96) 37 (±1) 0.03 (±0.01) 0.02 (±0.01) 0.13 (±0.06) 0.25 (±0.08)2-OCH3E1(E2) 0.42 (±0.04) 1.20 (±0.08) 3.64 (±0.59) 12 (±1) nd nd 0.12 (±0.04) 0.21 (±0.09)4-OHE1(E2)-1-N3Ade ndc nd nd nd nd nd 0.08 (±0.01) 0.35 (±0.08)4-OHE1(E2)-1-N7Gua nd nd nd nd nd nd 0.06 (±0.02) 0.36 (±0.06)

a 0.5 × 106 cells were plated in normal medium for 24 h, then the medium was changed to estrogen-free medium for 72 h. After changing with fresh estrogen-free medium, cells were pre-incubated with or without Ro41-0960 for 2 h before incubation with 0.1–30 �M E2 for 24 h. Media were collected, extracted byu

rrectedp

cOto

ciacscetwct5CCta

3c

1mctetiueoo

i

tctcrptoaa9OwMdbsfobcTmitiwgf6i

3e

sing a C8 cartidge and analyzed by HPLC with ECD.b The estrogen metabolite and depurinating DNA adduct levels were coresented in parentheses.c Not detected.

hanged with increasing concentrations of its substrate, 4-HE2 (Fig. 2a). Although Ro41-0960 competitively inhibits

he activity of COMT, it does not change the protein levelsf COMT (data not shown).

Although protein levels of enzymes may be similar, theyan differ in catalytic activity [32]. To investigate the activ-ty of COMT in TCDD- or 4-OHE2-treated MCF-10F cellsnd to validate findings from the Western blot experiments,atalytic activities were measured using 4-OHE2 as the sub-trate. The inhibition of COMT activity by Ro41-0960 wasompared to a DMSO control, and the remaining activity wasxpressed as percent of control enzyme activity. Data fromhe in vitro experiments showed that no significant changeas seen in COMT activity in cells treated with increasing

oncentrations of TCDD or 4-OHE2, compared with con-rol: 2.3 ± 0.6 �M 4-OHE2 was methylated to 4-OCH3E2 by0 �g of cellular COMT protein. Ro41-0960 fully inhibitedOMT activity at a concentration of 1 �M and decreasedOMT activity to 23% of the control at 0.1 �M in the reac-

ion system with saturating concentrations of SAM (200 �M)nd 4-OHE2 (100 �M).

.2. Estrogen metabolism in TCDD-induced MCF-10Fells treated with E2

To examine the profile of estrogen metabolism in MCF-0F cells, an HPLC method with ECD was modified. Eachetabolite has a distinct retention time and is oxidized as a

haracteristic cluster of oxidation peaks that are observed inwo or more adjacent channels [30]. Standard solutions ofach compound were combined to generate equimolar mix-ures containing varying concentrations of each standard andnjected onto the column. These standard solutions were thensed to generate calibration curves. Standard curves were lin-ar between 250 and 1000 pmol (data not shown). The limit

f detection under the conditions of analysis was ∼10 pmoln the column.

The profile of metabolites produced was first assessedn control or TCDD-pretreated MCF-10F cells subsequently

ntr

for recovery and normalized to cell numbers. The standard deviation is

reated with 0.1–30 �M E2. In non-TCDD-treated MCF-10Fells, metabolism of E2 was very limited. After 24 h, 95% ofhe estrogen recovered was unmetabolized; some E2 had beenonverted to E1, and the combination of other metabolitesepresented <5% of the total (Table 1). In contrast, in cellsretreated with TCDD (Table 2), metabolism was primarilyo 4-OCH3E1(E2) (28%) and 4-OHE1(E2) (1.2%). Formationf the corresponding 2-OHE1(E2) was very low; 2-OHE1(E2)nd 2-OCH3E1(E2) combined were <3.8% of the total. Theverage ratio of 4-OCH3E1(E2)/2-OCH3E1(E2) was about:1. To learn whether the TCDD-induced formation of 4-HE1(E2) and 2-OHE1(E2) is related to incubation timeith E2 after induction by TCDD (10 nM), TCDD-pretreatedCF-10F cells were incubated with 1 �M E2 for 1–72 h. The

ata show that formation of estrogen metabolites can be seeny 6 h, with a maximum at 6–24 h, and these metabolitestart to degrade after 72 h (data not shown). TCDD-inducedormation of estrogen metabolites is also related to the timef incubation with E2 (data not shown). To determine theest induction concentration and time for TCDD, MCF-10Fells were treated with various concentrations (0–30 nM) ofCDD for 24–72 h. TCDD induced formation of estrogenetabolites in a dose-dependent fashion (Fig. 2b). The max-

mal response for E2 metabolism was achieved followingreatment with 10 nM TCDD, and the optimal time for max-mum induction was 72 h. Therefore, 10 nM TCDD for 72 has used in subsequent experiments. The analysis of estro-en metabolites by HPLC also showed that the best timeor detecting catechol metabolites (4-OHE2 and 2-OHE2) is–24 h after adding E2 into the medium and for DNA adductss 48–72 h (data not shown).

.3. Effects of COMT inhibition on levels of catecholstrogens and methoxy catechol estrogens

In preliminary toxicity experiments, cell growth wasot affected by treatment with 10 �M Ro41-0960 for upo 72 h (data not shown). This concentration of inhibitoreduced COMT activity 87 and 99% at 24 and 72 h, as

F. Lu et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 150–158 155

Table 2Effect of COMT inhibition on the formation of CE and methoxy CE in TCDD-pretreated E2-treated MCF-10F cellsa

E2 (without COMT inhibitor) (pmol/106 cells) E2 + Ro41-0960 (with COMT inhibitor) (pmol/106 cells)

0.1 �M 1 �M 10 �M 30 �M 0.1 �M 1 �M 10 �M 30 �M

E1(E2) 76 (±3)b,c 2986 (±172) 4833 (±141) 5071 (±552) 13 (±2) 547 (±48) 1192 (±63) 2861 (±159)4-OHE1(E2) ndd 12.1 (±0.7) 23.1 (±0.7) 198 (±40) nd 11.1 (±0.9) 272 (±14) 331 (±60)2-OHE1(E2) nd 9.5 (±0.5) 2.3 (±0.1) 20.5 (±1.9) nd 2.7 (±0.2) 78 (±4) 36 (±5)4-OCH3E1(E2) 88 (±1.1) 445 (±26) 1650 (±48) 3265 (±206) nd nd 4.9 (±0.3) 7.6 (±0.9)2-OCH3E1(E2) 12.6 (±0.2) 111 (±6) 263 (±8) 321 (±30) nd nd 7.9 (±0.4) 3.0 (±0.6)

a Cells were plated (∼40% confluence) and pretreated with 10 nM TCDD for 72 h. Cultures were pre-incubated with or without Ro41-0960 2 h beforei d analy

normaative ex

a(bwRctts4ltOrwip

3d

ioe0fiw

Mndtie

e1ifw0

awa4Dmrchl

blTa20ition with 3 �M Ro41-0960 inhibited methoxy CE formationin MCF-10F cells by about 99%. In culture medium fromMCF-10F cells incubated with increasing concentrations of

Fig. 3. Effect of COMT inhibition on the formation of depurinating DNA

ncubation with 0.1–30 �M E2 for 24 h. Medium was collected, extracted anb The estrogen metabolite levels detected were corrected for recovery andc Data represent the mean ± S.D. of triplicate cultures from one representd Not detected.

ssessed by methylation of the COMT substrate 4-OHE2data not shown). To determine the effect of COMT inhi-ition on E2 metabolism, TCDD-pretreated MCF-10F cellsere exposed to 1 �M E2 for 24 h in the presence of 3 �Mo41-0960 (Table 2). The COMT inhibitor decreased theoncentration of 4-OCH3E1(E2) by 99% and 2-OCH3E1(E2)o undetectable levels. This was accompanied by a twoo fivefold increase in 4-OHE2 concentration, whereas noignificant change was seen in 2-OHE2 levels. Therefore,-OHE2, which represented 3% of the total estrogen metabo-ites recovered without COMT inhibition, was 80% of theotal recovered with COMT inhibition. The corresponding 4-CH3E1(E2) concentrations were 77 and <0.1% of the total,

espectively. The level of E1(E2) remaining in cells treatedith the COMT inhibitor was lower than that in cells without

nhibitor, suggesting that E2 metabolism was increased in theresence of the inhibitor (Table 2).

.4. Effect of COMT inhibition on the formation ofepurinating DNA adducts

To investigate the effects of COMT inhibition in TCDD-nduced E2-treated MCF-10F cells, the effects of Ro41-0960n CE-DNA adduct levels were studied. MCF-10F cells werexposed to 10 nM TCDD for 72 h, and then incubated with.1–30 �M E2 for 72 h with or without Ro41-0960. The pro-le of estrogen metabolites and depurinating DNA adductsas determined in cell culture medium by HPLC.Initial experiments demonstrated that in TCDD-pretreated

CF-10F cells exposed to 1 �M E2 for up to 24 h, similarly toon-E2-treated controls, no depurinating DNA adducts wereetected (data not shown). However, when the cells werereated with E2 while COMT activity was concurrently inhib-ted, depurinating DNA adduct levels continued to rise for thentire duration of treatment.

There were no significant differences in DNA adduct lev-ls in cells treated with 3 �M COMT inhibitor, TCDD alone,0 �M E2, TCDD plus 10 �M E2, or TCDD plus 3 �M

nhibitor (Fig. 3). In contrast, statistically significant three-old increases in the levels of depurinating DNA adductsere seen when TCDD-pretreated cells were incubated with.1–30 �M E2 in the presence of 3 �M Ro41-0960 (p < 0.05,

a0cat

zed by HPLC with ECD.lized to cell count.periment.

s determined by ANOVA). This increase in DNA adductsas dose-dependent with increasing concentrations of E2

nd was associated with an increase in the concentration of-OHE1(E2). Linear regression analysis of the depurinatingNA adduct levels in the medium was carried out to deter-ine their association with the levels of 4-OHE1(E2). The

esults indicate that the increased depurinating DNA adductoncentrations in cell culture media were associated withigh 4-OHE1(E2) levels (R2 = 0.78) and low 4-OCH3E1(E2)evels (data not shown).

To study whether the increase in CE-DNA adducts coulde attributed to decreased COMT activity, methoxy CEevels were determined in the culture medium by HPLC.CDD increased E2 metabolism, primarily to 4-OCH3E1(E2)nd 2-OCH3E1(E2), and formation of 4-OHE1(E2) and-OHE1(E2) was low without Ro41-0960. In contrast, Ro41-960 blocked formation of 4- and 2-OCH3E1(E2), andncreased levels of 4- and 2-OHE1(E2) were detected. Incuba-

dducts in MCF-10F cells treated with TCDD and E2 with or without Ro41-960. The cells were pretreated with TCDD and then treated with variousoncentrations of E2 in the presence or absence of Ro41-0960. The levels ofdducts in the cells treated with Ro41-0960 are significantly different fromhe untreated cells, p < 0.05 as determined by ANOVA.

156 F. Lu et al. / Journal of Steroid Biochemistry &

Fig. 4. Effect of the concentration of 4-OHE2 on the formation of depuri-ncwp

Ectiitb

ca(itiit0rslO

4

Qdbearibi

g

t4oaghCiwo

Tt(tnof(

ti4twwae(

prO1o

tauboeaiastaca

ating DNA adducts in MCF-10F cells. Linear regression analysis wasonducted by graphing individual data points (n = 3) from cells pretreatedith TCDD, then treated with increasing concentrations of 4-OHE2 in theresence of 3 �M Ro41-0960.

2 and 3 �M Ro41-0960, the COMT inhibitor decreased theoncentration of 4-OCH3E1(E2) and 2-OCH3E1(E2) to unde-ectable levels, while a concentration-dependent increasen CE formation was seen. These data indicate that thencrease in CE-DNA adducts was due to COMT inhibitionhat resulted in a decreased inactivation of the CE, as showny decreased methoxy CE formation (Table 2).

To directly determine the relationship between the con-entration of 4-OHE2 and formation of depurinating DNAdducts, cells were treated with increasing concentrations0–30 �M) of 4-OHE2 for 24 h. The DNA adduct levelsncreased with increasing concentration of 4-OHE2 (Fig. 4);hey are consistent with previous data showing ∼3-foldncreases in DNA adduct levels when COMT activity wasnhibited. The graph of the concentration of 4-OHE2 versushe amount of DNA adducts in the presence of 3 �M Ro41-960 yielded a line with a slope of 0.1 (p < 0.01 by linearegression analysis) and an R2 = 0.97 (by correlation analy-is), suggesting a direct relationship between CE metaboliteevels and DNA adducts in the absence of COMT activity and-methylated metabolites (Fig. 4).

. Discussion

When estrogen metabolism is balanced, low levels of CE-are formed and, consequently, the likelihood of DNA

amage leading to cancer initiation is low. Maintainingalanced estrogen metabolism requires interplay betweenstrogen activating enzymes, such as the cytochrome P450s,nd deactivating enzymes, such as COMT and quinoneeductase [2]. COMT activity appears to play a key rolen maintaining low levels of CE-Q. The role of COMT in

alancing estrogen metabolism was explored in MCF-10F,mmortalized human breast epithelial cells.

MCF-10F cells normally have low levels of the estro-en activating enzymes, CYP1A1 and CYP1B1. Thus, after

A

v

Molecular Biology 105 (2007) 150–158

reatment of the cells with E2, the levels of 2-OHE1(E2) and-OHE1(E2) in these cells are low (Table 1). In fact, almost allf the CE are present as 2-OCH3E1(E2) and 4-OCH3E1(E2),nd depurinating DNA adducts are not detected. When estro-en metabolism is unbalanced by inhibiting COMT, slightlyigher levels of CE are detected, and formation of methoxyE is virtually eliminated (Table 1). With COMT inhib-

ted, however, the depurinating DNA adducts can be detectedhen the MCF-10F cells are treated with higher levelsf E2.

Induction of CYP1B1 by pretreatment of the cells withCDD dramatically increases E2 metabolism, with forma-

ion of high levels of 2-OCH3E1(E2) and 4-OCH3E1(E2)Table 2). Inclusion of Ro41-0960 to inhibit COMT increaseshe levels of 2-OHE1(E2) and 4-OHE1(E2), almost elimi-ates formation of the methoxy CE and increases formationf 4-OHE1(E2)-1-N3Ade and 4-OHE1(E2)-1-N7Gua as aunction of the level of TCDD used to induce CYP1B1Figs. 2b and 3).

To explore the possibility that COMT levels might respondo substrate induction, the levels of COMT protein and activ-ty were analyzed in MCF-10F cells treated with 0–60 �M-OHE2 (Fig. 2a). No changes in the levels of COMT pro-ein or activity were observed. Although COMT activityas fully inhibited by Ro41-0960 in cell extracts incubatedith 4-OHE2 at a low concentration of 1 �M Ro41-0960,concentration of 3 �M was needed to achieve a similar

ffect when the whole cells were treated with the inhibitorTables 1 and 2).

Formation of the depurinating DNA adducts should beroportional to the concentration of E2-3,4-Q available toeact with DNA. In fact, when the cells were treated with 4-HE2, the levels of 4-OHE1(E2)-1-N3Ade and 4-OHE1(E2)--N7Gua detected in the cells were proportional to the levelf 4-OHE2 used to treat the cells (Fig. 4).

This is the first study to explore the role of COMT inhe formation of estrogen metabolites and depurinating DNAdduct levels in a normal human breast epithelial cell linender conditions in which E2 metabolism has been enhancedy TCDD. COMT activity can be limited by the presencef low activity genetic polymorphisms [15–21] or by lowxpression of the enzyme. In this study, inhibition of COMTctivity was shown to increase the formation of depurinat-ng CE-DNA adducts in MCF-10F cells. Formation of thesedducts and the concomitant apurinic sites in DNA have beenhown to induce mutations that are associated with initia-ion of breast cancer [5–7]. Therefore, a low level of COMTctivity may be a critical factor in the initiation of breast can-er by increasing metabolism of estrogens to form CE-DNAdducts.

cknowledgements

F. Lu was supported by a fellowship from the Uni-ersity of Nebraska Environmental Toxicology Graduate

stry &

PCoBlN

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

F. Lu et al. / Journal of Steroid Biochemi

rogram. This study was supported by U.S.P.H.S. grant P01A49210 from the National Cancer Institute and Departmentf Defense grant DAMD17-03-1-0299 from the U.S. Armyreast Cancer Research Program. Core support at the Epp-

ey Institute was provided by grant P30 CA36727 from theational Cancer Institute.

eferences

[1] B.E. Henderson, H.S. Feigelson, Hormonal carcinogenesis, Carcino-genesis 21 (2000) 427–433.

[2] E. Cavalieri, D. Chakravarti, J. Guttenplan, E. Hart, J. Ingle, R.Jankowiak, P. Muti, E. Rogan, J. Russo, R. Santen, T. Sutter, Cate-chol estrogen quinones as initiators of breast and other human cancers:implications for biomarkers of susceptibility and cancer prevention,BBA Rev. Cancer 1766 (2006) 63–78.

[3] B.T. Zhu, A.H. Conney, Functional role of estrogen metabolismin target cells: review and perspectives, Carcinogenesis 19 (1998)1–27.

[4] D.C. Spink, B.C. Spink, J.Q. Cao, J.A. DePasquale, B.T. Pentecost,M.J. Fasco, Y. Li, T.R. Sutter, Differential expression of CYP1A1and CYP1B1 in human breast epithelial cells and breast tumor cells,Carcinogenesis 19 (1998) 291–298.

[5] D. Chakravarti, P.C. Mailander, K.-M. Li, S. Higginbotham, H.L.Zhang, M.L. Gross, J.L. Meza, E.L. Cavalieri, E.G. Rogan, Evidencethat a burst of DNA depurination in SENCAR mouse skin induceserror-prone repair and forms mutations in the H-ras gene, Oncogene 20(2001) 7945–7953.

[6] P.C. Mailander, J.L. Meza, S. Higginbotham, D. Chakravarti, Inductionof A.T to G.C mutations by erroneous repair of depurinated DNA fol-lowing estrogen treatment of the mammary gland of ACI rats, J. SteroidBiochem. Mol. Biol. 101 (2006) 204–215.

[7] Z. Zhao, W. Kosinska, M. Khmelnitsky, E.L. Cavalieri, E.G. Rogan,D. Chakravarti, P.G. Sacks, J.B. Guttenplan, Mutagenic activity of 4-hydroxyestradiol, but not 2-hydroxyestradiol, in BB rat2 embryoniccells, and the mutational spectrum of 4-hydroxyestradiol, Chem. Res.Toxicol. 19 (2006) 475–479.

[8] J. Russo, M.H. Lareef, O. Tahin, Y.F. Hu, C. Slater, X. Ao, I.H. Russo,17Beta-estradiol is carcinogenic in human breast epithelial cells, J.Steroid Biochem. Mol. Biol. 80 (2002) 149–162.

[9] J. Russo, M.H. Lareef, G. Balogh, S. Guo, I.H. Russo, Estrogen and itsmetabolites are carcinogenic agents in human breast epithelial cells, J.Steroid Biochem. Mol. Biol. 87 (2003) 1–25.

10] M.H. Lareef, J. Garber, P.A. Russo, I.H. Russo, R. Heulings, J.Russo, The estrogen antagonist ICI-182-780 does not inhibit thetransformation phenotypes induced by 17-beta-estradiol and 4-OHestradiol in human breast epithelial cells, Int. J. Oncol. 26 (2005) 423–429.

11] P.T. Mannisto, S. Kaakkola, Catechol-O-methyltransferase (COMT):biochemistry, molecular biology, pharmacology, and clinical efficacyof the new selective COMT inhibitors, Pharmacol. Rev. 51 (1999)593–628.

12] N.W. Gaikwad, E.G. Rogan, E.L. Cavalieri, Evidence for NQO1-catalyzed reduction of estrogen ortho-quinones. Mass spectrometricdetermination of enzyme-substrate complex, Free Radic. Biol. Med.,2007, in press.

13] R.M. Hersey, K.I. Williams, J. Weisz, Catechol estrogen formationby brain tissue: characterization of a direct product isolation assayfor estrogen-2- and 4-hydroxylase activity and its application to stud-

ies of 2- and 4-hydroxyestradiol formation by rabbit hypothalamus,Endocrinology 109 (1981) 1912–1920.

14] C.K. Cohn, D.L. Dunner, J. Axelrod, Reduced catechol-O-methyltransferase activity in red blood cells of women with primaryaffective disorder, Science 70 (1970) 1323–1324.

[

Molecular Biology 105 (2007) 150–158 157

15] J.A. Lavigne, K.J. Helzlsouer, H.Y. Huang, P.T. Strickland, D.A. Bell,O. Selmin, M.A. Watson, S. Hoffman, G.W. Comstock, J.D. Yager,An association between the allele coding for a low activity variant ofcatechol-O-methyltransferase and the risk for breast cancer, CancerRes. 57 (1997) 5493–5497.

16] C.S. Huang, H.D. Chern, K.J. Chang, C.W. Cheng, S.M. Hsu, C.Y.Shen, Breast cancer risk associated with genotype polymorphism of theestrogen metabolizing genes CYP17, CYP1A1, and COMT: a multi-genic study on cancer susceptibility, Cancer Res. 59 (1999) 4870–4875.

17] D.-S. Yim, S.K. Park, K.-Y. Yoo, K.-S. Yoon, H.H. Chung, H.J. Kang,S.H. Ahn, D.Y. Noh, K.J. Choe, I.J. Jang, S.G. Shin, P.T. Strick-land, A. Hirvonen, D. Kang, Relationship between the val158metpolymorphism of catechol-O-methyl transferase and breast cancer,Pharmacogenetics 11 (2001) 1–8.

18] P.A. Thompson, P.G. Shields, J.L. Freudenheim, A. Stone, J.E. Vena,J.R. Marshall, S. Graham, R. Laughlin, T. Nemoto, F.F. Kadlubar,C.B. Ambrosone, Genetic polymorphisms in catechole-O-methyl-transferase, menopausal status, and breast cancer risk, Cancer Res. 58(1998) 2107–2110.

19] R. Millikan, G.S. Pittman, C.-K.J. Tse, E. Duell, B. Newman, D. Savitz,P.G. Moorman, R.J. Boissy, D.A. Bell, Catechol-O-methyltransferaseand breast cancer risk, Carcinogenesis 19 (1998) 1943–1947.

20] S. Wedren, T.R. Rudqvist, F. Granath, E. Weiderpass, M. Ingelman-Sundberg, I. Persson, C. Magnusson, Catechol-O-methyltransferasegene polymorphism and post-menopausal breast cancer risk, Carcino-genesis 24 (2003) 681–687.

21] H.M. Lachman, D.F. Papolos, T. Saito, Y.M. Yu, C. Szumlanski,R.M. Weinshilboum, Human catechol-O-methyltransferase pharmaco-genetics: description of a functional polymorphism and its potentialapplication to neuropsychiatric disorders, Pharmacogenetics 6 (1996)243–250.

22] P.D. Scanlon, F.A. Raymond, R.M. Weinshilboum, Catechol-O-methyltransferase: thermolabile enzyme in erythrocytes of subjectshomozygous for allele for low activity, Science 203 (1979)63–65.

23] C.E. Garner, L.T. Burka, A.E. Etheridge, H.B. Matthews, Catecholmetabolites of polychlorinated biphenyls inhibit the catechol-O-methyltransferase-mediated metabolism of catechol estrogens, Toxicol.Appl. Pharmacol. 162 (2000) 115–123.

24] Z.H. Chen, Y.J. Hurh, H.K. Na, J.H. Kim, Y.J. Chun, D.H. Kim,K.S. Kang, M.H. Cho, Y.J. Surh, Resveratrol inhibits TCDD-inducedexpression of CYP1A1 and CYP1B1 and catechol estrogen-mediatedoxidative DNA damage in cultured human mammary epithelial cells,Carcinogenesis 25 (2004) 2005–2013.

25] M. Saeed, M. Zahid, E.G. Rogan, E.L. Cavalieri, Synthesis of cat-echols of natural and synthetic estrogens by using 2-iodoxybenzoicacid (IBX) as the oxidizing agent, Steroids 70 (2005) 173–178.

26] D.E. Stack, J. Byun, M.L. Gross, E.G. Rogan, E.L. Cavalieri,Molecular characteristics of catechol estrogen quinones in reac-tions with deoxyribonucleosides, Chem. Res. Toxicol. 9 (1996) 851–859.

27] K.-M. Li, R. Todorovic, P. Devanesan, S. Higginbotham, H. Kofeler, R.Ramanathan, M.L. Gross, E.G. Rogan, E.L. Cavalieri, Metabolism andDNA binding studies of 4-hydroxyestradiol and estradiol-3,4-quinonein vitro and in female ACI rat mammary gland in vivo, Carcinogenesis25 (2004) 289–297.

28] F. Denizot, R. Lang, Rapid colorimetric assay for cell growth and sur-vival. Modifications to the tetrazolium dye procedure giving improvedsensitivity and reliability, J. Immunol. Methods 89 (1986) 271–

277.

29] J.E. Goodman, L.T. Jensen, P. He, J.D. Yager, Characterization ofhuman soluble high and low activity catechol-O-methyltransferase cat-alyzed catechol estrogen methylation, Pharmacogenetics 12 (2002)517–528.

1 stry &

[

[

(2005) 53–60.

58 F. Lu et al. / Journal of Steroid Biochemi

30] E.G. Rogan, A.F. Badawi, P. Devanesan, J.L. Meza, J.A. Edney, W.W.West, S.M. Higginbotham, E.L. Cavalieri, Relative imbalances in

estrogen metabolism and conjugation in breast tissue of women withcarcinoma: potential biomarkers of susceptibility to cancer, Carcino-genesis 24 (2003) 697–702.

31] J.D. Vidal, C.A. Vandevoort, C.B. Marcus, N.R. Lazarewicz, A.J. Con-ley, 2,3,7,8-tetrachlorodibenzo-p-dioxin induces CYP1B1 expression

[

Molecular Biology 105 (2007) 150–158

in human luteinized granulosa cells, Arch. Biochem. Biophys. 439

32] S. Dawling, N. Roodi, R.L. Mernaugh, X. Wang, F.F. Parl,Catechol-O-methyltransferase (COMT)-mediated metabolism of cat-echol estrogens: comparison of wild-type and variant COMT isoforms,Cancer Res. 61 (2001) 6716–6722.