One of the greatest challenges facing human geneticists is the identification and characterization of susceptibility genes for common complex multifactorial human diseases. This challenge is partly due to the limitations of parametric-statistical methods for detection of gene effects that are dependent solely or partially on interactions with other genes and with environmental exposures. We introduce multifactor-dimensionality reduction (MDR) as a method for reducing the dimensionality of multilocus information, to improve the identification of polymorphism combinations associated with disease risk. The MDR method is nonparametric (i.e., no hypothesis about the value of a statistical parameter is made), is model-free (i.e., it assumes no particular inheritance model), and is directly applicable to case-control and discordant-sib-pair studies. Using simulated case-control data, we demonstrate that MDR has reasonable power to identify interactions among two or more loci in relatively small samples. When it was applied to a sporadic breast cancer case-control data set, in the absence of any statistically significant independent main effects, MDR identified a statistically significant high-order interaction among four polymorphisms from three different estrogen-metabolism genes. To our knowledge, this is the first report of a four-locus interaction associated with a common complex multifactorial disease.

The correlation reported previously, as well as our current findings, suggest that further investigations are warranted to understand the possible linkage of the ER gene locus to hereditary breast cancer.

The oxidative metabolism of estrogens has been implicated in the development of breast cancer; yet, relatively little is known about the mechanism by which estrogens cause DNA damage and thereby initiate mammary carcinogenesis. To determine how the metabolism of the parent hormone 17B-estradiol (E 2 ) leads to the formation of DNA adducts, we used the recombinant, purified phase I enzyme, cytochrome P450 1B1 (CYP1B1), which is expressed in breast tissue, to oxidize E 2 in the presence of 2 ¶-deoxyguanosine or 2 ¶-deoxyadenosine. We used both gas and liquid chromatography with tandem mass spectrometry to measure E 2 , the 2-and 4-catechol estrogens (2-OHE 2 , 4-OHE 2 ), and the depurinating adducts 4-OHE 2 -1(A,B)-N7-guanine (4-OHE 2 -N7-Gua) and 4-OHE 2 -1(A,B)-N3-adenine (4-OHE 2 -N3-Ade). CYP1B1 oxidized E 2 to the catechol 4-OHE 2 and the labile quinone 4-hydroxyestradiol-quinone to produce 4-OHE 2 -N7-Gua and 4-OHE 2 -N3-Ade in a time-and concentration-dependent manner. Because the reactive quinones were produced as part of the CYP1B1-mediated oxidation reaction, the adduct formation followed MichaelisMenten kinetics. Under the conditions of the assay, the 4-OHE 2 -N7-Gua adduct (K m , 4.6 F 0.7 Mmol/L; k cat , 45 F

Oxidative metabolites of estrogens have been implicated in the development of breast cancer, yet relatively little is known about the metabolism of estrogens in the normal breast. We developed a mathematical model of mammary estrogen metabolism based on the conversion of 17B-estradiol (E 2 ) by the enzymes cytochrome P450 (CYP) 1A1 and CYP1B1, catechol-O-methyltransferase (COMT), and glutathione S-transferase P1 into eight metabolites [i.e., two catechol estrogens, 2-hydroxyestradiol (2-OHE 2 ) and 4-hydroxyestradiol (4-OHE 2 ); three methoxyestrogens, 2-methoxyestradiol, 2-hydroxy-3-methoxyestradiol, and 4-methoxyestradiol; and three glutathione (SG)-estrogen conjugates, 2-OHE 2 -1-SG, 2-OHE 2 -4-SG, and 4-OHE 2 -2-SG]. When used with experimentally determined rate constants with purified enzymes, the model provides for a kinetic analysis of the entire metabolic pathway. The predicted concentration of each metabolite during a 30-minute reaction agreed well with the experimentally derived results. The model also enables simulation for the transient quinones, E 2 -2,3-quinone (E 2 -2,3-Q) and E 2 -3,4-quinone (E 2 -3,4-Q), which are not amenable to direct quantitation. Using experimentally derived rate constants for genetic variants of CYP1A1, CYP1B1, and COMT, we used the model to simulate the kinetic effect of enzyme polymorphisms on the pathway and identified those haplotypes generating the largest amounts of catechols and quinones. Application of the model to a breast cancer case-control population identified a subset of women with an increased risk of breast cancer based on their enzyme haplotypes and consequent E 2 -3,4-Q production. This in silico model integrates both kinetic and genomic data to yield a comprehensive view of estrogen metabolomics in the breast. The model offers the opportunity to combine metabolic, genetic, and lifetime exposure data in assessing estrogens as a breast cancer risk factor. (Cancer Epidemiol Biomarkers Prev 2006;15(9):1620 -9)

Oxidative metabolites of estrogens have been implicated in the development of breast cancer, yet relatively little is known about the metabolism of estrogens in the normal breast. We developed an experimental in vitro model of mammary estrogen metabolism in which we combined purified, recombinant phase I enzymes CYP1A1 and CYP1B1 with the phase II enzymes COMT and GSTP1 to determine how 17β-estradiol (E2) is metabolized. We employed both gas and liquid chromatography with mass spectrometry to measure the parent hormone E2 as well as eight metabolites, that is, the catechol estrogens, methoxyestrogens, and estrogen–GSH conjugates. We used these experimental data to develop an in silico model, which allowed the kinetic simulation of converting E2 into eight metabolites. The simulations showed excellent agreement with experimental results and provided a quantitative assessment of the metabolic interactions. Using rate constants of genetic variants of CYP1A1, CYP1B1, and COMT, the model further allowed examination of the kinetic impact of enzyme polymorphisms on the entire metabolic pathway, including the identification of those haplotypes producing the largest amounts of catechols and quinones. Application of the model to a breast cancer case-control population defined the estrogen quinone E2-3,4-Q as a potential risk factor and identified a subset of women with an increased risk of breast cancer based on their enzyme haplotypes and consequent E2-3,4-Q production. Our in silico model integrates diverse types of data and offers the exciting opportunity for researchers to combine metabolic and genetic data in assessing estrogenic exposure in relation to breast cancer risk.