Characterization of Common UGT1A8, UGT1A9, and UGT2B7 Variants with Different Capacities to Inactivate Mutagenic 4-Hydroxylated Metabolites of Estradiol and Estrone

Research Article

Jean Thibaudeau,1 Johanie Le´pine,1 Jelena Tojcic,1 Yannick Duguay,1 Georges Pelletier,2 Marie Plante,3 Jacques Brisson,4 Bernard Teˆtu,5 Simon Jacob,6 Louis Perusse,7 Alain Be´langer,2 and Chantal Guillemette1

1Canada Research Chair in Pharmacogenomics, Oncology and Molecular Endocrinology Research Center, CHUQ Research Center, Faculty of Pharmacy, 2Oncology and Molecular Endocrinology Research Center, CHUQ Research Center, Faculty of Medicine, 3Gynecologic Oncology Service, Hoˆtel-Dieu de Que´bec,Faculty of Medicine, 4Population Health Research Unit, Hoˆpital Saint-Sacrement du CHA de Que´becand Faculty of Medicine, 5Department of Pathology, Hoˆtel-Dieu de Que´bec,Faculty of Medicine, 6Anatomic-Pathological and Cytological Laboratory, Hoˆpital Saint-Sacrement, Faculty of Medicine and 7Division of Kinesiology, Department of Social and Preventive Medicine, Faculty of Medicine, Laval University, Quebec, Canada

Abstract In addition to estrogens from the circulation, the uterine and mammary gland tissues possess the required enzymes for estrogen The oxidative metabolism of estrone (E1) and estradiol (E2)to form carcinogenic 4-hydroxy-catecholestrogens (4-OHCE) is synthesis from circulating C19 steroid precursors. Besides, the associated with uterine and breast carcinogenesis. In this contribution of ovarian and in situ synthesis varies during aging and the later source of estrogens becomes more significant at study, we conducted functional analyses of genetic variants in the UDP-glucuronosyltransferase UGT1A8, UGT1A9, and menopause. It is also quite evident that in situ estrogen UGT2B7 enzymes primarily involved in the inactivation of concentrations are also dependent on local metabolism. Transfor- 4-OHCEs. Compared with UGT2B7*2 (H268Y), UGT2B7*1 mation of estrogens to hydroxylated metabolites, i.e., catecholes- exhibited a 2-fold lower efficiency (intrinsic clearance) at trogens, has recently received attention because of their association conjugating 4-hydroxyestrone and 4-hydroxyestradiol at posi- with breast and uterine tumorigenesis (3–5). tions 3 and 4 caused by altered capacities (V ) and affinities Estrogens are substrates of cytochromes P450 CYP1A1/1A2 and max CYP1B1, that generates predominantly 2- and 4-hydroxy-catecho- (Km). The 79 G>A promoter variation, characterizing the UGT2B7*2g haplotype, leads to a 50% reduction of transcrip- lestrogens (2- and 4-OHCE) derivatives, respectively (6). These tion (P < 0.001) in human endometrial carcinoma-1B cells. metabolic pathways lead to the formation of metabolites with Furthermore, a >12-fold decreased intrinsic clearance of the distinctive biological activities compared with parent estradiol (E2) *1 proteins was induced by selected amino acid substitutions and estrone (E1) in a number of tissues including uterus and in UGT1A8 (*3 C277Y) and UGT1A9 (*3 M33T). Frequencies of the mammary gland (7, 8). Both 2-OHCEs and 4-OHCEs may be low-activity alleles in Caucasians were 45% for UGT2B7*1, subsequently converted to corresponding semiquinone and qui- 5% for the 79A promoter variant, 1.2% for UGT1A8*3, and none forms and then undergo a CYP450/peroxidase catalyzed free 2.2% for UGT1A9*3. Supporting a protective role in two organs radicals generating metabolic redox cycling. These quinones and sensitive to 4-OHCE–induced damages, the expression of UGT free radicals can mediate DNA adducts and cellular oxidation, enzymes was shown by immunohistochemistry in normal leading to possible mutagenesis (9–12). Quinone intermediates of breast and endometrial tissues and confirmed by Western 4-OHCEs are particularly carcinogenic because they form unstable blotting in a subset of samples. Altogether, findings suggest DNA adducts and induce tumor formation in animal models that specific polymorphisms in UGT genes may modulate the (13, 14). In opposition, 2-OHCEs generate stable DNA adducts and they possess antiproliferative and antiangiogenic properties once exposure to carcinogenic metabolites of E2 and potentially lead to an altered risk of breast and endometrial cancers in women transformed to methoxy derivatives, mostly 2-methoxyestradiol, by carrying the variant alleles. (Cancer Res 2006; 66(1): 125-33) the action of catechol-O-methyltransferase (15). One of the major pathways of conjugation for E2 and its oxidized and methoxylated metabolites is through UDP-glucuronosyltrans- Introduction ferases (UGT), which catalyze the covalent addition of glucuronic Breast and endometrial cancers have been associated with acid (UDPGA). In our recent study, we showed that glucuronidation higher exposure to endogenous or exogenous estrogens (1, 2). of E2 and E1, as well as their catecholestrogen and methoxyestrogen metabolites, is mostly catalyzed by six UGTs, i.e., UGT1A1, 1A3, 1A8, 1A9, 1A10, and 2B7 (16). The study of Lepine et al. further showed that among these, UGT1A8, 1A9, and 2B7 are highly Note: J. Thibaudeau and J. Le´pinecontributed equally to this work and should be efficient toward the carcinogenic 4-OHCEs (16). regarded as joint first authors. For UGT allele nomenclature description, refer to: UDP-Glucuronosyltransferase Limited attention was given to the expression of estrogen- (UGT) Alleles Nomenclature Home Page, Quebec, Canada, Pharmacogenomics conjugating UGT enzymes in steroid target tissues, but the Laboratory, c2005. Available from: http://pha.ulaval.ca/labocg/alleles/alleles.html. Requests for reprints: Chantal Guillemette, Canada Research Chair in available data indicate their presence in breast and uterine tissues Pharmacogenomics, Pharmacogenomics Laboratory, CHUQ Research Center, T3-67, (16–18). In turn, the significant amounts of estrogen-glucuronides 2705 Boulevard Laurier, Quebec, Canada G1V 4G2. Phone: 418-654-2296; Fax: 418-654- in the breast cyst fluid supports a local inactivation of estrogens by 2761; E-mail: [email protected]. I2006 American Association for Cancer Research. UGT enzymes (19, 20). In support of a role of the glucuronidation doi:10.1158/0008-5472.CAN-05-2857 pathway in the metabolism of estrogens in target cells is the www.aacrjournals.org 125 Cancer Res 2006; 66: (1). January 1, 2006

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research. Cancer Research consistent association of UGT polymorphisms with altered transcriptional activity of the UGT2B7 79G>A promoter poly- circulating hormones and modified risk of endometrial, breast, morphism in human endometrial carcinoma (HEC-1B) cells using and ovarian cancers (18, 21–24). Altogether, these results suggest pGL3 reporter gene constructs. Finally, to gain information into the that the metabolic pathway mediated by UGT enzymes is protective role of UGT against carcinogenic E2 metabolites, we implicated in estrogen inactivation and that the activity of these determined the pattern of expression of UGT1A8, UGT1A9, and enzymes may play a critical role in the modulation of in situ UGT2B7 enzymes in healthy endometrium and breast samples estrogen exposure. from postmenopausal women. Given the central role of UGT1A8, UGT1A9, and UGT2B7 in the The main findings of this study are that several forms of UGT metabolism of 4-OHCEs, we studied the influence of polymor- enzymes are associated with reduced expression or decreased phisms in these genes to identify those which modify enzyme inactivation of 4-OHE2 and 4-OHE1. In addition, expression analyses catalytic activity and the level of expression. Four nonsynonymous in the uterus and breast, two target tissues of 4-OHCE-induced polymorphisms of UGT1A8 and UGT1A9 exons 1 have been damages, suggest a potential alteration of the lifetime exposure to associated with changes in phenotypes (Fig. 1). Three alleles of these genotoxic metabolites in women carrying low-activity alleles. UGT1A9, *1, *2 (C3Y), and *3 (M33T), and three alleles of UGT1A8, 173 277 *1, *2 (A G), and *3 (C Y), are present in Caucasian and African- Materials and Methods American populations (25, 26). For UGT2B7, only one nonsynon- Chemicals. 4-OHE and 4-OHE were purchased from Steraloids ymous polymorphism was found, UGT2B7*2 (H268Y; refs. 27, 28). 1 2 (Newport, RI). The glucuronides derivatives of 4-OHE1 and 4-OHE2 were Functional studies revealed that some forms of UGT proteins obtained as previously described (16). High-performance liquid chromatog- present modified activity against a variety of xenobiotics but their raphy (HPLC) grade methanol, acetonitrile, 1-chlorobutane, and iso-amyl- effects on estrogen glucuronidation has never been investigated alcohol were purchased from VWR Canlab (Montre´al,Quebec, Canada). (25–29). In addition, a functional variation in the UGT2B7 promoter Ammonium hydroxide was obtained from Fisher Scientific (Nepean, Ontario, is located at position 79 G>A and is in linkage disequilibrium Canada). All chemicals were of the highest grade available. with the codon 268 variation and is referred as haplotype III, DNA samples and tissue procurement. DNA samples from 258 UGT2B7*2g (ref. 30; Fig. 1). This promoter polymorphism was unrelated Caucasian subjects were obtained from the Quebec Family Study associated with a significant decrease of transcriptional activity in (31). These samples had been anonymized prior to their receipt in our hepatoma and colon cells, two sites of expression of the UGT2B7 laboratory. All subjects provided written consent for the use of their DNA for experimental purposes, and the present study was reviewed and enzyme, whereas its functional effect in estrogen-dependent cells approved by the Institutional Review Boards (CHUL Research Center and remains to be established (30). We initially investigated the Laval University). Nonmalignant uterine and breast tissues were obtained functional effect of common polymorphisms by in vitro enzymatic from postmenopausal women who had not received hormone replacement assays with 4-hydroxyestrone (4-OHE1) and 4-hydroxyestradiol therapy for at least 3 weeks and who had no menstrual bleeding for at least (4-OHE2) using individual overexpressed UGT1A8, UGT1A9, and 1 year. All subjects provided written consent for use of their specimens, and UGT2B7 allozymes to determine the maximal velocity (Vmax), the the present study was reviewed and approved by Institutional Review affinity (Km), and catalytic efficiencies (CLint). We also assessed the Boards. Fresh tissue from patients were collected by pathologists and

Figure 1. Schematic representation of the UGT1A (A) and UGT2B7 (B) genes indicating the locations of the single amino acid substitutions studied. Frequency of each allele is indicated. p, pseudogenes.

Cancer Res 2006; 66: (1). January 1, 2006 126 www.aacrjournals.org

Protein expression analysis assessed by immunohistochemistry. Immunohistochemistry analyses were done as previously described (16). Briefly, serial cuts of nonmalignant human uterine and breast tissues from postmenopausal women (n = 5 for each tissue) were incubated with human UGT1A8/1A9 (#519) and UGT2B7 (#1809) antisera diluted 1:1,000 and 1:250, respectively. Sections were washed and incubated with peroxidase- conjugated goat anti-rabbit immunoglobulins (DakoCytomation, Missis- sauga, Ontario, Canada) diluted 1:200. Endogenous peroxidase activity was eliminated and peroxidase intensity was controlled under the microscope. Sections were counterstained with hematoxylin and the control experiments were done by substituting preimmune rabbit serum. Enzymatic assays. Prior to enzymatic assays, UGT proteins were stably expressed in HEK-293 cell systems and microsomal fractions were prepared as previously described (16). For a more accurate assessment of the quantitative differences in glucuronidation activity for variant allozymes, HEK293-derived cell lines were characterized for UGT protein expression by Western blot analysis (data not shown) as previously described (16). To establish the functional effect of nonsynonymous polymorphisms of Figure 2. Western blot analysis of UGT1A8, UGT1A9, and UGT2B7 UGT1A8, UGT1A9, and UGT2B7 on 4-OHE and 4-OHE glucuronidation, expression in healthy endometrium. To determine the specificity of our 1 2 polyclonal antibodies, microsomal proteins (10 Ag) of HK293 cells stably we conducted enzymatic assays with microsomal proteins obtained from expressing each UGT were separated on a 10% SDS-PAGE gel. Both HEK-293 cells stably overexpressing each variant protein. Enzymatic assays antibodies were used at a dilution of 1:2,000. Human liver (HL) was used as were done with 40 to 60 Ag total microsomal proteins in 100 AL reaction positive control. A, the anti-UGT1A8/1A9 (#519) binds only to UGT1A8, volume containing; 50 mmol/L Tris-HCl (pH 7.5), 10 mmol/L MgCl ,5Ag/ UGT1A9, and to HL through UGT1A1, 1A3, 1A4, 1A6, 1A7, 1A8, 1A9, and 1A10. 2 B, the anti-UGT2B7 (#1809) binds only to UGT2B7 and to HL proteins toward mL pepstatin, 0.5 Ag/mL leupeptin, 10 Ag/mL phosphatidylcholine, and 1 UGT2B4, 2B7, 2B10, 2B11, 2B15, 2B17, and 2B28 proteins. C, microsomal mmol/L UDPGA. Reactions were started by adding varying concentrations fractions (10 Ag) of five endometrium samples from postmenopausal women of 4-OHE1 and 4-OHE2 ranging from 1 to 200 Amol/L and incubating at were tested to evaluate the expression of UGT1A8/1A9 and UGT2B7 (#519 and 37jC. After 3 hours, the assays were stopped by adding 100 AL of ice-cold #1809 antibodies, respectively, at a dilution of 1:2,000). methanol and than centrifuged at 14,000 g for 10 minutes in order to remove proteins before mass spectrometry analysis. Relative glucuronida- deposited in liquid nitrogen within 30 minutes of surgery pending transfer tion activities are expressed as pmol/min/UGT level, which was obtained to a freezer at 80jC. after dividing the absolute activity of UGT (pmol/min/mg) by the protein Genotyping and haplotype analyses of UGT1A8. PCR was used to expression level obtained by Western blot analyses. amplify the first exon of the UGT1A8 gene, and primers were designed to Mass spectrometry analysis. Glucuronide formation was assessed by amplify overlapping fragments, as previously described for UGT1A9 (26). mass spectrometry analyses as previously described (16). Briefly, incubation The primers used for UGT1A8 were: #174F-175R, 5V-ctggaccgggaattcatgga medium was diluted with 0.1 mL methanol/water (50:50; vol/vol), vortexed, and 5V-gtggctgtagagatcatatgct; #1338F-1322R, 5V-ttcgccaggggaatag and 5V- and then transferred into a conical vial for injection into the mass atttgctctagggggtc. Amplicons were sequenced with an ABI 3700 automated spectrometer. The HPLC and tandem mass spectrometry system consisted sequencer using Big Dye (Perkin-Elmer, Boston, MA) dye primer chemistry. of a mass spectrometer (model API 3000; Perkin-Elmer Sciex, Thornhill, Sequences were analyzed with Staden preGap4 and Gap4 programs. Canada) equipped with an electrospray ionization source in the negative ion Haplotypes for UGT1A8 were determined using Phase v2.1 program mode and a HPLC pump plus autosampler (model 2690, Waters, Milford, (26, 32). The linkage between the different polymorphisms was determined MA). The mass spectrometer was operated in the multiple reaction 8 with the LDplotter tool program. monitoring mode. Protein expression analysis assessed by Western blot analysis. An Enzyme kinetic analysis. Determination of the kinetic variables was antibody (#519) was raised against amino acids 3 to 118 of UGT1A9 and its done using Sigmaplot 8.02a with Enzyme Kinetic 1.1 (Systat Software, Point characterization revealed that it also recognized UGT1A8 (95% of Richmond, CA). Best kinetic enzyme model was chosen after visual homology). The specificity of this antibody (1:2,000) was studied by Western inspection of the fitted curve, the Eadie-Hofstee plot and the Akaike blot experiments as previously described (Fig. 2A; ref. 16). The character- Information Criterion values. Coherent models were the Michaelis-Menten ization of the polyclonal anti-UGT2B7 (#1809) has already been described model [V =(Vmax S)/(Km + S)], the substrate activation model [Hill n n n (ref. 16; Fig. 2B). UGT1A8, UGT1A9, and UGT2B7 protein expression in five equation, V =(Vmax S )/(Km + S )] and the uncompetitive substrate healthy endometrium samples were assessed by Western blot analysis. inhibition model (V =[Vmax S]/[Km + S (1 + S/Ks)]), where Vmax is the Microsomes were extracted as follows: 200 mg of frozen tissue were first maximal velocity, Km is the substrate concentration at half-maximal homogenized using polytron in microsome buffer [20 mmol/L KH2PO4, velocity, ‘‘n’’ is the degree of curve sigmoidicity, and Ks is an inhibition 20 mmol/L K2HPO4, 20% glycerol, 1 mmol/L DTT (pH 7.0)], sonicated, and constant. Based on these models, the intrinsic clearance (CLint) estimation j then centrifuged at 14,000 g at 4 C for 20 minutes. The supernatant was values were calculated using the following equations: Michealis-Menten centrifuged at 125,000 g for 60 minutes at 4jC, and finally, the pellet and substrate inhibition, CLint = Vmax/Km; substrate activation, CLmax = was homogenized in the same microsome buffer. The quantity of protein 1/n (Vmax/Km) ([n 1] / [n (n 1) ]) (33, 34). Statistical analysis was was determined using the Bradford method (Bio-Rad, Mississauga, Canada) done with a two-tailed Student’s t test. using bovine serum albumin as standard. Ten micrograms of the Cell culture and transfection of UGT2B7 promoters. HEC-1B cells microsomal proteins were loaded on a 10% SDS-polyacrylamide gel (SDS- were obtained from American Type Culture Collection (Manassas, VA) and PAGE) and transferred onto nitrocellulose (Bio-Rad). The expressions of cultured as recommended by the supplier in a humidified 5% carbon UGT1A8 and UGT1A9 were monitored with the anti-UGT1A8/1A9 (#519) at dioxide incubator at 37jC. Cells were seeded 24 hours before transfection a dilution of 1:2,000, whereas UGT2B7 was probed with antibody #1809 at a assays in a 24-well plate with 60,000 cells per well (corresponding to 50% of dilution of 1:2,000 (16). Bands were revealed using enhanced chemilumi- confluence). Plasmids were obtained as previously described (30). Trans- nescence (Perkin-Elmer, Boston, MA). fections of 1,000 ng plasmid constructs with pRLnull as an internal transfection efficiency control were carried out in MEM using 4 AL of ExGen 500 transfection reagent (MBI Fermentas, Burlington, Ontario, Canada). 8 https://innateimmunity.net/. Supplemented serum was added after 3 hours. Cells were harvested www.aacrjournals.org 127 Cancer Res 2006; 66: (1). January 1, 2006

24 hours after transfection and promoter activity was monitored using the UGT level; P < 0.001; Fig. 3) whereas a 13-fold decreased velocity Dual-Luciferase Reporter Assay System as stated by the manufacturer was also observed for the conjugation of 4-OHE1 at position 4 (Promega, Madison, WI). Luciferase activity was measured in a 96-well plate (158 versus 2,113 pmol/min/UGT level; P < 0.001). Despite better on a LB96V microplate luminometer (EG&G Berthold, Bad Wildbad, affinities (K ) of the UGT1A8*3 allozyme for the conjugation of Germany) with 30 AL of cell lysates. m 4-OHE1 at both positions 3 and 4 (29 and 35 versus 290 and 162 Amol/L, respectively), this variant had an intrinsic clearance Results diminished by at least 2- to 3-fold (P < 0.05) compared with the 277 Enzymatic activities of UGT allozymes. Estrogens are poten- UGT1A8*1 protein. For the conjugation of 4-OHE2, the C Y*3 tially conjugated at the 2-, 3-, 4- or 17-hydroxyl positions. Our variation also had a remarkable effect on the velocity of the previous work showed that glucuronidation occurs predominantly enzyme for the conjugation of the 3-hydroxyl (4 versus 163 at the 3- and 4-hydroxyl positions of 4-OHCEs (16). Therefore, both pmol/min/UGT level; P < 0.01) and the 4-hydroxyl (265 versus positions were investigated. 7,730 pmol/min/UGT level; P < 0.01; Fig. 3). Similarly to 4-OHE1, UGT1A8. The codon 277 cysteine for a tyrosine *3 variation the Km values were also enhanced for the UGT1A8*3 allozyme of UGT1A8 had a dramatic effect on the catalytic activity of the compared with the UGT1A8*1 at both positions 3 and 4 (20 and enzyme towards the 4-hydroxylated estrogens. Compared with 63 Amol/L versus 103 and 133 Amol/L, respectively). As a result, the UGT1A8*1 enzyme, the velocity of conjugation of 4-OHE1 the CLint values were significantly reduced by 8- to 13-fold (3- was reduced at position 3 by 30-fold (8 versus 246 pmol/min/ hydroxyl, 0.2 versus 1.6 AL/min/mg; P < 0.01; 4-hydroxyl, 4.2

Figure 3. Kinetic profiles for the glucuronidation of 4-OHE1 (A and C) and 4-OHE2 (B and D) at position 4 by UGT1A8 (A and B) and UGT1A9 (C and D). Microsomes, produced from HEK-293 cells as described in Materials and Methods, were incubated for 60 minutes at 37jC with concentrations of 4-OHE1 or 4-OHE2 ranging from 1 to 200 Amol/L. Quantification of estrogen-glucuronides was done by HPLC tandem mass spectrometry and absolute glucuronidation activities were divided by UGT protein expression assessed by Western blot analysis. Points, mean of two independent experiments done in duplicate; bars, F SEM.

Cancer Res 2006; 66: (1). January 1, 2006 128 www.aacrjournals.org

Table 1. Kinetic variables for the glucuronidation of 4-OHE1 and 4-OHE2 by UGT1A8, UGT1A9, and UGT2B7

4-OHCEs and -3G -4G UGT

Apparent Km Relative Vmax CLint Apparent Km Relative Vmax CLint (Amol/L) (pmol/min/UGT level) (AL/min/mg) (Amol/L) (pmol/min/UGT level) (AL/min/mg)

4-OHE1* 1A8*1 290 F 15 246 F 6 0.9 162 F 3 2,113 F 22 7.1 c 1A8*2 207 F 51 328 F 1 1.8 216 F 26 2,192 F 63 6.1 b c b c 1A8*3 29 F 2 8 F 1x 0.3 35 F 8 158 F 15x 2.9 1A9*1 111 F 770F 5 0.6 73 F 2 412 F 40 5.7 1A9*2 139 F 33 75 F 2 0.6 85 F 1 402 F 111 4.8 c c c c c 1A9*3 392 F 59 18 F 4 0.05 99 F 963F 18 0.6 2B7*1 — — — 45 F 3 7,317 F 745 163.6 c b 2B7*2 — — — 25 F 0 11,637 F 868 475.0 k 4-OHE2 1A8*1 103 F 6 163 F 18 1.6 133 F 22 7,730 F 355 56.5 c 1A8*2 82 F 17 110 F 11 1.4 87 F 12 4,236 F 395 49.6 b b b c b c 1A8*3 20 F 4 4 F 1 0.2 63 F 14 265 F 42 4.2 1A9*1 33 F 332F 0 0.4 33 F 1 3,135 F 260 47.4 1A9*2 33 F 334F 1 0.5 28 F 1 3,069 F 232 54.8 b b b 1A9*3 36 F 38F 0x 0.1 26 F 2 422 F 44 8.2 2B7*1 117 F 258F 2 0.5 50 F 1 2,426 F 283 49.0 c c 2B7*2 90 F 294F 1 1.1 47 F 1 4,398 F 490 93.2

F NOTE: —, not detected; CLint values were calculated as described in Materials and Methods. Results are expressed with mean SEM of two independent experiments done in duplicate.

*For 4-OHE1 glucuronidation, all allozymes displayed a hyperbolic profile for both positions 3 and 4 except for UGT1A8 alleles, which showed a sigmoid profile at position 4. cP < 0.05. bP < 0.01. x P < 0.001. k For 4-OHE2 glucuronidation at positions 3 and 4, UGT1A8*1 and *2 had a hyperbolic profile, whereas UGT1A8*3 showed a substrate inhibition profile. All UGT1A9 alleles showed a sigmoid profile, whereas UGT2B7*1 and *2 had a substrate inhibition profile for the glucuronidation of 4-OHE2 at positions 3 and 4.

versus 56.5 AL/min/mg; P < 0.05). In contrast, the change pmol/min/UGT level, respectively; P < 0.01). This leads to a of UGT1A8 codon 173 alanine to a glycine (UGT1A8*2) had significant decrease of intrinsic clearances by 4-fold at position 3 no influence on enzymatic activity using 4-OHE1 or 4-OHE2 (0.1 versus 0.4 AL/min/mg; P <0.01) and by almost 6-fold at position (Table 1). 4 (8.2 versus 47.4 AL/min/mg; P < 0.01; Table 1). UGT1A9. The codon 3 cysteine to a tyrosine change, character- UGT2B7. The UGT2B7*1 enzyme with a histidine at codon izing the *2 allele of UGT1A9, had no significant effect on enzyme 268 had almost a 3-fold lower intrinsic clearance for 4-OHE1 at kinetics compared with UGT1A9*1, both for the glucuronidation position 4 (163.6 AL/min/mg; P < 0.01) compared with the at positions 3 and 4 of 4-OHE1 and 4-OHE2 (Fig. 3). However, UGT2B7*2 allozyme (475.0 AL/min/mg), characterized by a compared with the UGT1A9*1 protein, the *3 allozyme with a tyrosine at codon 268 (Fig. 4). The altered clearance is the threonine instead of a methionine at codon 33, had a Vmax almost result of a change in the affinity of the variant *2 enzyme, 4-fold lower for the conjugation of 4-OHE1 at position 3 (18 versus which was decreased by almost 2-fold for the glucuronidation at 70 pmol/min/UGT level; P < 0.05) and 6.5-fold lower for the position 4 (25 versus 45 Amol/L; P < 0.05). The velocity was also conjugation of 4-OHE1 at position 4 (63 versus 412 pmol/min/UGT increased by a factor of 1.6 compared with UGT2B7*1 (11,637 268 level; P < 0.05). The Km values for both UGT1A9*1 and *3 proteins versus 7,317 pmol/min/UGT level). Similarly, UGT2B7*2 Y is were similar, except for the conjugation at position 3 of 4-OHE1 more efficient at conjugating 4-OHE2 at both positions compared that was increased by 3.5-fold (392 versus 111 Amol/L; P < 0.05). with the UGT2B7*1 allozyme (P < 0.05; Fig. 4). In this case, a Intrinsic clearances were significantly decreased by >9- and >12- higher Vmax for the conjugation rate of 4-OHE2 at both positions 268 fold (P < 0.05) for the glucuronidation of 4-OHE1 at positions 3 and 3and4forUGT2B7*2Y is observed compared with 4 by the *3 protein, respectively. In parallel, the effect of the UGT2B7*1 H268 (94 and 4,398 versus 58 and 2,426 pmol/min/ 33 UGT1A9*3 (M T) variation on the glucuronidation of 4-OHE2 was UGT level, respectively; Table 1). exclusively within the capacity of the enzyme. Vmax values for the To assess the effect of the promoter 79 G>A variant conjugation at positions 3 and 4 were significantly reduced by the transcriptional activity, we did transient cell transfection using presence of a threonine at codon 33 (8 and 422 versus 32 and 3,135 luciferase reporter promoter constructs covering the 190/34 www.aacrjournals.org 129 Cancer Res 2006; 66: (1). January 1, 2006

Figure 4. Kinetic profiles of UGT2B7 and transcriptional activity of UGT2B7 promoter. A and B, kinetic profiles for the glucuronidation of 4-OHE1 (A) and 4-OHE2 (B) at position 4 by UGT2B7. Microsomes, produced from HEK-293 cells as described in Materials and Methods, were incubated for 60 minutes at 37jC with concentrations of 4-OHE1 or 4-OHE2 ranging from 1 to 200 Amol/L. Quantification of estrogen-glucuronides was done by HPLC tandem mass spectrometry and absolute glucuronidation activities were divided by UGT protein expression assessed by Western blot analysis. Points, mean of two independent experiments done in duplicate; bars, F SE. C, transcriptional activities of UGT2B7 promoter constructs in HEC-1B endometrial cells. Promoter constructs of UGT2B7 covering the 1,362/34 and the 190/34 regions (relative to the transcription starting site) were transfected in HEC-1B endometrial cells and promoter activities were measured by a luciferase reporter plasmid assay. *, P < 0.05; ***, P < 0.001.

and 1,362/34 UGT2B7 promoter regions (positions are relative different exon 1 sequences on the locus is associated with four to the transcriptional start site; ref. 30). HEC-1B cells were used downstream exons common to all UGT1A isoforms (Fig. 1; refs. as a cellular model. Transcription levels of the 79A variant 35, 36). It is thus likely that a linkage disequilibrium might be plasmids were significantly reduced by 50% (P < 0.001) and by present between UGT1A8 and UGT1A9. However, our data 30% (P < 0.05) using short or long promoter constructs, obtained in 516 chromosomes indicate a prevalence of 1.2% respectively (Fig. 4). and 2.2% for UGT1A8*3 and UGT1A9*3 alleles, respectively, and Linkage disequilibrium in the UGT1 gene. The importance no significant linkage between both loci (R2= 0.0003, DV= 1.00). of understanding linkage disequilibrium in candidate loci is Expression of UGT1A8, UGT1A9, and UGT2B7 in estrogen- becoming increasingly evident, and future epidemiologic studies responsive tissues. Immunohistochemistry of UGTs was done should consider the haplotype of UGT1A8 and UGT1A9 because in normal breast and endometrial tissues obtained from five of the genomic structure of the UGT1A gene. Indeed, the entire postmenopausal women. Representative results are shown in UGT1 family is derived from a single gene locus (UGT1A) which Fig. 5. In breast tissues, immunostaining of UGT1A8/1A9 and is composed of 17 exons. To synthesize the final protein, 1 of 13 UGT2B7 is observed in the cytoplasm of ductal epithelial cells

Cancer Res 2006; 66: (1). January 1, 2006 130 www.aacrjournals.org

(Fig. 5A-C). For the uterus, immunostaining of UGT1A8/1A9 and inactivation of genotoxic endogenous catecholestrogens. We UGT2B7 was detected in the cytoplasm of both epithelial cells of present the evaluation of the kinetic properties of UGT1A8, endometrial glands and those lining the uterine cavity and some UGT1A9, and UGT2B7 variants towards 4-OHCE inactivation and stromal cells were also labeled (Fig. 5B-D). In the myometrium, the identification of UGT enzymes with low enzymatic activity. In smooth muscle cells had weak immunolabeling (data not addition, the biological relevance of UGT enzymes is further shown). When the nonimmune serum was used, no staining supported by their expression in the breast and the uterus, two could be detected (data not shown). Western blot analyses using organs sensitive to 4-OHCE-induced damage supporting the microsomal preparations of endometrial tissues also confirmed protective role of this metabolic pathway. the expression of UGT1A8/1A9 and UGT2B7 proteins (Fig. 2C). We previously showed that, out of all known human UGTs, Because the antibody (#519) used in the immunohistochemistry UGT2B7, UGT1A8, and UGT1A9 are mainly involved in the experiments recognizes both UGT1A8/1A9 proteins, we con- glucuronidation of 4-OHE2 and 4-OHE1 (16). Particularly, UGT2B7 firmed mRNA expression of individual genes by specific reverse represents the most efficient UGT for the glucuronidation of 4- transcription-PCR in these patients. Both UGT1A8 and UGT1A9 OHE1 with a 25-fold higher intrinsic clearance for the formation transcripts were detected in samples from postmenopausal of the predominant glucuronide at position 4, compared with women (n = 5) although with variability between samples UGT1A8 and UGT1A9, respectively (163.6 versus 7.1 and 5.7 AL/ (data not shown), despite the limited number of samples min/mg). However, all three enzymes share similar intrinsic analyzed. clearances for the glucuronidation of 4-OHE2 at position 4 (49.0, 56.5, and 47.4 AL/min/mg; Table 1). In this study, we show a significant influence of variants UGT1A8*3, UGT1A9*3,and Discussion UGT2B7*2 on the formation of 4-OHE1 and 4-OHE2 glucuronides. One potential mechanism of estrogen carcinogenesis in breast Two of these mutations in the UGT1A8 and UGT1A9 genes (the and endometrial tissues involves the formation of catechol *3 alleles) severely reduce enzyme function, whereas the metabolites, especially the 4-oxidized metabolites of E2 and E1. UGT2B7*2 variation is associated with a significant 2-fold Several authors have reported that these estrogen metabolites increase in clearance. induced genotoxicity and participate in tumor initiation (10, 11, The A173G (*2) and C277Y (*3) polymorphisms in UGT1A8 are 13). The elimination and inactivation of these metabolites has the only amino acid substitutions described for this gene (25). thus far received little attention, and the observation of the high A previous functional study revealed that the capacity of the specificity of UGT enzymes for these substrates attracted our UGT1A8*2 protein in transient transfections is similar to the *1 interest (16). UGT represents a metabolic pathway that participates for various molecules (25). From the data presented here, a in blocking the production of reactive oxygen species induced by modest effect of the UGT1A8*2 is also predicted because this the catecholestrogen-redox cycle and the production of DNA protein shares similar catalytic properties (Km, Vmax, and CLint) adducts by enhancing their elimination and preventing their with UGT1A8*1. In contrast, the near loss of activity for accumulation in tissues. The present study assessed the influence UGT1A8*3 was previously reported for a number of molecules of common polymorphisms in the glucuronidation pathway for the suggesting an effect independent of the nature of the substrate (25, 29). For 4-OHCEs, a drastic reduction of activity for the *3 allozyme was observed at positions 3 and 4. Unexpectedly, the affinity of UGT1A8*3 was significantly enhanced by the presence of a tyrosine at codon 277 compared with the UGT1A8*1 protein with a cysteine. On the other hand, due to a drastic decreased velocity (13- to 40-fold) of the UGT1A8*3 protein compared with *1, the intrinsic clearances were between 3- and 13-fold lower for both estrogenic substrates. The UGT1A8*3 polymorphism corre- sponds to a single base pair mutation at the very highly conserved cysteine to a tyrosine in the substrate-binding region at codon 277. It was postulated that C277 might be involved in a disulfide bridge to another cysteine essential for the protein conformation or catalysis (25). Similarly, we observed a reduced capacity of the UGT1A9*3 variant to inactivate carcinogenic 4-OHE1 and 4-OHE2 metabolites. This amino acid substitution, located in the substrate-binding domain, was revealed as a loss of function mutation for selected, but not all, substrates (26, 29). Although the activity of UGT1A9*3 was drastically decreased for the anticancer agent SN-38, its activity was maintained for the glucuronidation of another Figure 5. Expression of UGT1A8, UGT1A9, and UGT2B7 in healthy antineoplasic drug flavopiridol and was partially decreased for endometrium and breast tissues from postmenopausal women by the immunosuppressive agent mycophenolic acid (26, 29). Thus, it immunohistochemistry. A and C, immunolabeling of breast for UGT1A8/1A9 and UGT2B7 is observed in the cytoplasm of ductal epithelial cells (arrows). seems that the effect of this amino acid substitution, corresponding B and D, immunolabeling of endometrium for UGT1A8/1A9 and UGT2B7 is to a nonconservative amino acid change, is substrate-dependent. observed in the cytoplasm of epithelial cells of endometrial glands (G) and those UGT1A8*3 and UGT1A9*3 were found in f1.2% and 2.2% of the covering the uterine cavity (C). Some stromal cells are also labelled (arrows). When the nonimmune serum was used, no staining could be detected (data not population, respectively, consistent with frequencies found in shown). Original magnification is 300. previous studies (25, 26). www.aacrjournals.org 131 Cancer Res 2006; 66: (1). January 1, 2006

An enhancement of the catalytic efficiency was observed for the concentrations (10- to 50-fold higher), it is plausible that UGT amino acid substitution at codon 268 of the UGT2B7 protein. The may have a significant physiologic role in uterine and breast UGT2B7*2 H268Y allozyme presented a significant increase on both tissues (3). Therefore, individuals with low UGT activity alleles affinity and velocity compared with UGT2B7*1. Previous studies might present lower inactivation activities for estrogen metabo- reported similar activities for the glucuronidation of 2-OHCEs and lites and being exposed to higher concentrations of these 4-OHCEs (37). Our study has several distinctions compared with mutagenic metabolites. However, the relative contribution of previous studies explaining those controversial observations. First, individual UGTs in the tissular conjugation of 4-OHCEs, as the glucuronide formation was quantified by HPLC coupled with determined by their kinetic properties and level of expression, is tandem mass spectrometry using authentic standards compared still unknown and deserves further investigation. with the nonspecific thin layer chromatographic technique used in Supporting the hypothesis that glucuronidation modulate this previous study. Our approach has the particularity to quantify estrogen bioavailability within target cells, we recently observed glucuronides at both positions 3 and 4. Second, the enzymatic a decreased risk to develop endometrial cancer by 60% to 80% activities were normalized by UGT protein expression in the associated with a genetic variation in UGT1A1, depending on heterologous system accounting for the variable expression of the menopausal status (18). The association of a low UGT1A1 variant allozymes, which was not the case in this study. The activity allele and the protection towards endometrial cancer is functional analyses reveal a substrate-selective effect because explained by the lower expression of UGT1A1 in the endome- earlier studies showed that the UGT2B7*2 protein has no effect on trium that would reduce the excretion of 2-OHCEs and 2- the in vitro glucuronidation of androsterone, menthol, opioids, MeOCEs, the antiproliferative metabolites of E2 (15). Suscepti- propranolol, and epirubicin (27, 38, 39), whereas a reduced activity bility to breast cancer associated with the polymorphic UGT1A1 for lithocholic acid, hyodeoxycholic acid, 17h-estradiol and 3V- promoter was also shown (21–23). In addition, the results of azido-3V-deoxythymidine is observed (40, 41). In addition, we report Gestl and coauthors support the role of UGT on mammary hor- that the codon 268 amino acid substitution influences the affinity monal status because they have shown the formation of 4-OHE1 of the protein for the glucuronidation of 4-OHE1 at position glucuronides by normal breast tissues (17). These observations 4-hydroxyl suggesting that this variation affects the function and were reinforced by the demonstration of the expression of the recognition of the enzyme for selected substrates. Thus, the UGT2B7 in the ductal epithelial cells of the mammary glands, a UGT2B7*1 allele represents a low activity with regard to 4-OHCE finding confirmed in this study. glucuronidation and a high activity allele for the conjugation of In conclusion, this work shows the variable efficiency of specific other molecules. genetic variants of UGT enzymes expressed in breast and uterine The functionality of the UGT2B7 79 G>A promoter tissues to inactivate the mutagenic metabolites of E2. The pattern polymorphism was also revealed in endometrial cells and leads of expression for UGT1A8, UGT1A9, and UGT2B7, similar to that to a significant decrease of the transcriptional activity of the of the CYP1B1, involved in the formation of 4-OHCEs, further gene. Consistent with these results, we previously showed that supports a biological inactivating role for the UGT pathway in favor this polymorphism is associated with a 2.5- to 7-fold decrease of of the conjugation of 4-OHCEs with glucuronide acid prior to their transcription in colon and liver cells, respectively (30). Because release into the circulation. This is reinforced by the fact that the 79A variation is in strong linkage disequilibrium with the normal breast tissues are able to form 4-OHCEs glucuronides (17), 268 high-activity UGT2B7*2 (Y ) allele, prediction of the physiologic a finding that was recently confirmed in our preliminary studies for effect of this haplotype (UGT2B7*2g) remains to be determined. normal endometrial tissues (data not shown). Thus, we propose a This haplotype occurs in 5% of the population, whereas the possible influence of the reduced capacity to inactivate E2 268 frequency of the UGT2B7*2a (79G and Y ) allele is much carcinogenic metabolites by UGT1A8*3, UGT1A9*3, and UGT2B7*1 higher, accounting for >50% of the Caucasian population (30). variant proteins that may compromise the individual’s protection Due to the limited number of samples analyzed, and the low against carcinogenic estrogens and estrogen-induced carcinogen- frequency of the 79 UGT2B7 variant, its relationship with esis. Hence, women with specific UGT variants may present a UGT2B7 expression in breast and uterine tissues could not be variable lifetime exposure to these mutagenic estrogen metabolites. assessed in this study. It is thus hypothesized that an impaired UGT-mediated 4-OHCE The biological relevance of the kinetic differences observed inactivation pathway may have an effect on population-based risk needs to be evaluated in relation to enzyme expression in tissues estimates of breast and endometrial cancers. Further investigations where the 4-OHCEs are formed. The immunohistochemical to determine the influence of UGT1A8, UGT1A9, and UGT2B7 analyses revealed predominant staining of UGTs in endometrial polymorphisms on the risk of cancer in case-control studies are glandular cells and in epithelial cells covering the uterine cavity currently ongoing. and in the ductal epithelial cells of mammary glands. Other studies previously showed a similar expression pattern of CYP1B1, the enzyme responsible for the formation of 4-OHCEs Acknowledgments in these tissues (42–44). The expression pattern of 4-OHCEs metabolizing UGTs further supports their role in the local Received 8/11/2005; revised 9/23/2005; accepted 10/21/2005. Grant support: Canadian Institutes of Health Research (CIHR MOP-68964). inactivation and modulation of concentrations in E2 and its J. Thibaudeau, J. Le´pine,and Y. Duguay are the recipients of graduate studentship metabolites in uterine and breast tissues. In addition, whereas awards from the Natural Sciences and Engineering Research Council of Canada, CIHR, and the Canada’s Research-Based Pharmaceutical Companies-CIHR, respectively. the Km values of UGT enzymes for 4-OHCEs are in the C. Guillemette is the chairholder of the Canada Research Chair in Pharmacogenomics. micromolar range, other phase I and phase II biotransforming The costs of publication of this article were defrayed in part by the payment of page enzymes share similar affinities for estrogens (45–47). In addition, charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. given that higher concentrations of estrogens in tissues, such as We thank Patrick Caron, Lyne Villeneuve, and Louise De´syfor their expertise and the mammary gland, are observed compared with serum technical assistance.

Cancer Res 2006; 66: (1). January 1, 2006 132 www.aacrjournals.org

References 18. Duguay Y, McGrath M, Lepine J, et al. The functional 33. Venkatakrishnan K, Von Moltke LL, Greenblatt DJ. UGT1A1 promoter polymorphism decreases endometri- Human drug metabolism and the cytochromes P450: 1. Key TJ, Appleby PN, Reeves GK, et al. Body mass index, al cancer risk. Cancer Res 2004;64:1202–7. application and relevance of in vitro models. J Clin serum sex hormones, and breast cancer risk in 19. Belanger A, Caron S, Labrie F, Naldoni C, Dogliotti Pharmacol 2001;41:1149–79. postmenopausal women. J Natl Cancer Inst 2003;95: L, Angeli A. Levels of eighteen non-conjugated and 34. Houston JB, Kenworthy KE. In vitro-in vivo scaling of 1218–26. conjugated steroids in human breast cyst fluid: CYP kinetic data not consistent with the classical 2. Kaaks R, Lukanova A, Kurzer MS. Obesity, endogenous relationships with cyst type. Eur J Cancer 1990;26: Michaelis-Menten model. Drug Metab Dispos 2000;28: hormones, and endometrial cancer risk: a synthetic 277–81. 246–54. review. Cancer Epidemiol Biomarkers Prev 2002;11:1531–43. 20. Belanger A, Labrie F, Angeli A. Unconjugated and 35. Gong QH, Cho JW, Huang T, et al. Thirteen 3. Jefcoate CR, Liehr JG, Santen RJ, et al. Tissue-specific glucuronide steroid levels in human breast cyst fluid. UDPglucuronosyltransferase genes are encoded at the synthesis and oxidative metabolism of estrogens. J Natl Ann N Y Acad Sci 1990;586:93–100. human UGT1 gene complex locus. Pharmacogenetics Cancer Inst Monogr 2000;27:95–112. 21. Adegoke OJ, Shu XO, Gao YT, et al. Genetic 2001;11:357–68. 4. Zhu BT, Conney AH. Functional role of estrogen polymorphisms in uridine diphospho-glucuronosyl- 36. Owens IS, Ritter JK. Gene structure at the human metabolism in target cells: review and perspectives. transferase 1A1 (UGT1A1) and risk of breast cancer. UGT1 locus creates diversity in isozyme structure, Carcinogenesis 1998;19:1–27. Breast Cancer Res Treat 2004;85:239–45. substrate specificity, and regulation. Prog Nucleic Acid 5. Raftogianis R, Creveling C, Weinshilboum R, Weisz J. 22. Guillemette C, De Vivo I, Hankinson SE, et al. Res Mol Biol 1995;51:305–38. Estrogen metabolism by conjugation. J Natl Cancer Inst Association of genetic polymorphisms in UGT1A1 with 37. Cheng Z, Rios GR, King CD, et al. Glucuronidation of Monogr 2000;27:113–24. breast cancer and plasma hormone levels. Cancer catechol estrogens by expressed human UDP-glucur- 6. Lee AJ, Cai MX, Thomas PE, Conney AH, Zhu BT. Epidemiol Biomarkers Prev 2001;10:711–4. onosyltransferases (UGTs) 1A1, 1A3, and 2B7. Toxicol Sci Characterization of the oxidative metabolites of 17h- 23. Guillemette C, Millikan RC, Newman B, Housman DE. 1998;45:52–7. estradiol and estrone formed by 15 selectively expressed Genetic polymorphisms in uridine diphospho-glucuro- 38. Coffman BL, King CD, Rios GR, Tephly TR. The human cytochrome p450 isoforms. Endocrinology 2003; nosyltransferase 1A1 and association with breast cancer glucuronidation of opioids, other xenobiotics, and 144:3382–98. among African Americans. Cancer Res 2000;60:950–6. androgens by human UGT2B7Y(268) and UGT2B7H(268). 7. Sasaki M, Kaneuchi M, Fujimoto S, Tanaka Y, Dahiya 24. Cecchin E, Russo A, Corona G, et al. UGT1A1*28 Drug Metab Dispos 1998;26:73–7. R. CYP1B1 gene in endometrial cancer. Mol Cell polymorphism in ovarian cancer patients. Oncol Rep 39. Innocenti F, Iyer L, Ramirez J, Green MD, Ratain MJ. Endocrinol 2003;202:171–6. 2004;12:457–62. Epirubicin glucuronidation is catalyzed by human UDP- 8. Iscan M, Klaavuniemi T, Coban T, Kapucuoglu N, 25. Huang YH, Galijatovic A, Nguyen N, et al. Identifica- glucuronosyltransferase 2B7. Drug Metab Dispos 2001; Pelkonen O, Raunio H. The expression of cytochrome tion and functional characterization of UDP-glucurono- 29:686–92. P450 enzymes in human breast tumours and normal syltransferases UGT1A8*1, UGT1A8*2 and UGT1A8*3. 40. Barbier O, Turgeon D, Girard C, et al. 3V-azido-3V- breast tissue. Breast Cancer Res Treat 2001;70:47–54. Pharmacogenetics 2002;12:287–97. deoxythimidine (AZT) is glucuronidated by human 9. Liehr JG, Ricci MJ, Jefcoate CR, Hannigan EV, 26. Villeneuve L, Girard H, Fortier LC, Gagne JF, UDP-glucuronosyltransferase 2B7 (UGT2B7). Drug Hokanson JA, Zhu BT. 4-Hydroxylation of estradiol by Guillemette C. Novel functional polymorphisms in the Metab Dispos 2000;28:497–502. human uterine myometrium and myoma microsomes: UGT1A7 and UGT1A9 glucuronidating enzymes in 41. Gall WE, Zawada G, Mojarrabi B, et al. Differential implications for the mechanism of uterine tumorigen- Caucasian and African-American subjects and their glucuronidation of bile acids, androgens and estrogens esis. Proc Natl Acad Sci U S A 1995;92:9220–4. impact on the metabolism of 7-ethyl-10-hydroxycamp- by human UGT1A3 and 2B7. J Steroid Biochem Mol Biol 10. Liehr JG. Is estradiol a genotoxic mutagenic carcin- tothecin and flavopiridol anticancer drugs. J Pharmacol 1999;70:101–8. ogen? Endocr Rev 2000;21:40–54. Exp Ther 2003;307:117–28. 42. Bofinger DP, Feng L, Chi LH, et al. Effect of TCDD 11. Cavalieri EL, Rogan EG. A unifying mechanism in the 27. Bhasker CR, McKinnon W, Stone A, et al. Genetic exposure on CYP1A1 and CYP1B1 expression in explant initiation of cancer and other diseases by catechol polymorphism of UDP-glucuronosyltransferase 2B7 cultures of human endometrium. Toxicol Sci 2001;62: quinones. Ann N Y Acad Sci 2004;1028:247–57. (UGT2B7) at amino acid 268: ethnic diversity of alleles 299–314. 12. Liehr JG, Ricci MJ. 4-Hydroxylation of estrogens as and potential clinical significance. Pharmacogenetics 43. Tsuchiya Y, Nakajima M, Kyo S, Kanaya T, marker of human mammary tumors. Proc Natl Acad Sci 2000;10:679–85. Inoue M, Yokoi T. Human CYP1B1 is regulated by U S A 1996;93:3294–6. 28. Holthe M, Rakvag TN, Klepstad P, et al. Sequence estradiol via estrogen receptor. Cancer Res 2004;64: 13. Liehr JG, Fang WF, Sirbasku DA, Ari-Ulubelen A. variations in the UDP-glucuronosyltransferase 2B7 3119–25. Carcinogenicity of catechol estrogens in Syrian ham- (UGT2B7) gene: identification of 10 novel single 44. Muskhelishvili L, Thompson PA, Kusewitt DF, Wang sters. J Steroid Biochem 1986;24:353–6. nucleotide polymorphisms (SNPs) and analysis of their C, Kadlubar FF. In situ hybridization and immunohis- 14. Newbold RR, Liehr JG. Induction of uterine adeno- relevance to morphine glucuronidation in cancer tochemical analysis of cytochrome P450 1B1 expression carcinoma in CD-1 mice by catechol estrogens. Cancer patients. Pharmacogenomics J 2003;3:17–26. in human normal tissues. J Histochem Cytochem 2001; Res 2000;60:235–7. 29. Bernard O, Guillemette C. The main role of UGT1A9 49:229–36. 15. Lakhani NJ, Sarkar MA, Venitz J, Figg WD. 2- in the hepatic metabolism of mycophenolic acid and the 45. Hanna IH, Dawling S, Roodi N, Guengerich FP, Parl Methoxyestradiol, a promising anticancer agent. Phar- effects of naturally occurring variants. Drug Metab FF. Cytochrome P450 1B1 (CYP1B1) pharmacogenetics: macotherapy 2003;23:165–72. Dispos 2004;32:775–8. association of polymorphisms with functional differ- 16. Lepine J, Bernard O, Plante M, et al. Specificity and 30. Duguay Y, Baar C, Skorpen F, Guillemette C. A novel ences in estrogen hydroxylation activity. Cancer Res regioselectivity of the conjugation of estradiol, estrone, functional polymorphism in the uridine diphosphate- 2000;60:3440–4. and their catecholestrogen and methoxyestrogen glucuronosyltransferase 2B7 promoter with significant 46. Dawling S, Roodi N, Mernaugh RL, Wang X, Parl metabolites by human uridine diphospho-glucuronosyl- impact on promoter activity. Clin Pharmacol Ther 2004; FF. Catechol-O-methyltransferase (COMT)-mediated transferases expressed in endometrium. J Clin Endo- 75:223–33. metabolism of catechol estrogens: comparison of crinol Metab 2004;89:5222–32. 31. Lanouette CM, Giacobino JP, Perusse L, et al. wild-type and variant COMT isoforms. Cancer Res 17. Gestl SA, Green MD, Shearer DA, Frauenhoffer E, Association between uncoupling protein 3 gene and 2001;61:6716–22. Tephly TR, Weisz J. Expression of UGT2B7, a UDP- obesity-related phenotypes in the Quebec Family Study. 47. Adjei AA, Thomae BA, Prondzinski JL, Eckloff BW, glucuronosyltransferase implicated in the metabolism of Mol Med 2001;7:433–41. Wieben ED, Weinshilboum RM. Human estrogen 4-hydroxyestrone and all-trans retinoic acid, in normal 32. Stephens M, Smith NJ, Donnelly P. A new statistical sulfotransferase (SULT1E1) pharmacogenomics: gene human breast parenchyma and in invasive and in situ method for haplotype reconstruction from population resequencing and functional genomics. Br J Pharmacol breast cancers. Am J Pathol 2002;160:1467–79. data. Am J Hum Genet 2001;68:978–89. 2003;139:1373–82.

www.aacrjournals.org 133 Cancer Res 2006; 66: (1). January 1, 2006

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 2006 American Association for Cancer Research. Characterization of Common UGT1A8, UGT1A9, and UGT2B7 Variants with Different Capacities to Inactivate Mutagenic 4-Hydroxylated Metabolites of Estradiol and Estrone

Jean Thibaudeau, Johanie Lépine, Jelena Tojcic, et al.

Cancer Res 2006;66:125-133.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/66/1/125

Cited articles This article cites 47 articles, 16 of which you can access for free at: http://cancerres.aacrjournals.org/content/66/1/125.full#ref-list-1

Citing articles This article has been cited by 16 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/66/1/125.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/66/1/125. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.