Cancer Research Grants Metabolism
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gold, B.
Right arrow Articles by Offit, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gold, B.
Right arrow Articles by Offit, K.
[Cancer Research 64, 8891-8900, December 15, 2004]
© 2004 American Association for Cancer Research

Regular Articles

Estrogen Receptor Genotypes and Haplotypes Associated with Breast Cancer Risk

Bert Gold1, Francis Kalush2, Julie Bergeron3, Kevin Scott3, Nandita Mitra4, Kelly Wilson2, Nathan Ellis4, Helen Huang4, Michael Chen3, Ross Lippert5,6, Bjarni V. Halldorsson5, Beth Woodworth1, Thomas White2, Andrew G. Clark2, Fritz F. Parl7, Samuel Broder2, Michael Dean1 and Kenneth Offit4

1 Human Genetics Section, Laboratory of Genomic Diversity, National Cancer Institute at Frederick, Frederick, Maryland; 2 Celera Diagnostics or Celera, Inc, Rockville, Maryland or Alameda, California; 3 SAIC-Frederick, Inc, Frederick, Maryland; 4 Memorial Sloan-Kettering Cancer Center, New York, New York; 5 Applied Biosystems, Rockville, Maryland; 6 Department of Mathematics, Massachusetts Institute of Technology, Cambridge, Massachusetts; and 7 Department of Pathology, Vanderbilt University School of Medicine, Nashville, Tennessee


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Nearly one in eight US women will develop breast cancer in their lifetime. Most breast cancer is not associated with a hereditary syndrome, occurs in postmenopausal women, and is estrogen and progesterone receptor-positive. Estrogen exposure is an epidemiologic risk factor for breast cancer and estrogen is a potent mammary mitogen. We studied single nucleotide polymorphisms (SNPs) in estrogen receptors in 615 healthy subjects and 1011 individuals with histologically confirmed breast cancer, all from New York City. We analyzed 13 SNPs in the progesterone receptor gene (PGR), 17 SNPs in estrogen receptor 1 gene (ESR1), and 8 SNPs in the estrogen receptor 2 gene (ESR2). We observed three common haplotypes in ESR1 that were associated with a decreased risk for breast cancer [odds ratio (OR), ~ O.4; 95% confidence interval (CI), 0.2–0.8; P < 0.01]. Another haplotype was associated with an increased risk of breast cancer (OR, 2.1; 95% CI, 1.2–3.8; P < 0.05). A unique risk haplotype was present in ~7% of older Ashkenazi Jewish study subjects (OR, 1.7; 95% CI, 1.2–2.4; P < 0.003). We narrowed the ESR1 risk haplotypes to the promoter region and first exon. We define several other haplotypes in Ashkenazi Jews in both ESR1 and ESR2 that may elevate susceptibility to breast cancer. In contrast, we found no association between any PGR variant or haplotype and breast cancer. Genetic epidemiology study replication and functional assays of the haplotypes should permit a better understanding of the role of steroid receptor genetic variants and breast cancer risk.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Only a small fraction (<5%) of women diagnosed with breast cancer have a clear hereditary predisposition (1, 2, 3) , and of these, about one half have predisposing mutations in BRCA1, BRCA2, PTEN, TP53, or other known cancer predisposing genes. However, twin studies indicate that the heritability of breast cancer is ~30% (4) , suggesting that genes other than the well-mapped regions act as modifiers of breast cancer risk. Although it is likely low penetrance as well as high penetrance genes may be involved in the etiology, it remains unclear which genomic regions and which biochemical functions or signal transduction pathways account for additional, heritable breast cancer incidence or progression.

Abundant epidemiologic evidence suggests that estrogen plays a crucial role in most breast cancers. Nulliparous women are at significantly elevated risk, as are women who have children late in their lives, women who have early menarche or women who have late menopause. Obesity is also associated with breast cancer risk; estrogen synthesis in adipose tissue is proposed to account for this increase in risk. Whereas estrogen receptor (ER)-positive and progesterone receptor (PgR)-positive breast cancers have better short-term prognosis than those that have become hormone independent (5) , receptor status varies as a function of age and menopausal status. Younger patients are more likely receptor negative and hormonally unresponsive; older patients are more often receptor positive and hormonally responsive. Recent and complete reviews of the genetics of breast cancer and its relation to the estrogen and progesterone receptor are available (6, 7, 8, 9, 10, 11, 12) .

In this report, we seek to identify candidate steroid hormone receptor gene variants in ESR1, ESR2, and PGR that might be associated with risk of breast cancer, perhaps leading to accelerated or slower rates of neoplastic transformation.8


   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
5'-Nucleotidase Assay Designs.
Single nucleotide polymorphisms (SNPs) discovered through data mining of the Celera Discovery System, a Celera Proprietary Database, or deposits into dbSNP were chosen for assay design. Limited resequencing permitted discovery of the rare variant in ESR2 exon 4 reported here. Sequences chosen for scoring on the entire cohort were either purchased from Applied Biosystems as Assays-on-Demand (AOD) or submitted to an Applied Biosystems Assays-by-Design pipeline. One difficult assay (G393G in PGR) was designed by Raymond Stephens of Celadon Labs., College Park, MD. Propynyl T oligonucleotide probes for G393G were manufactured under special license agreement by TriLink Biotech, San Diego, CA.

5'-Nucleotidase Assay Method.
Five nanograms of genomic patient and control DNAs were aliquoted with a Hydra liquid handler (HYDRA Robbins Molecular BioProducts, San Diego, CA) into 384-well bar-coded optical thermocycler plates compatible with the ABI PRISM 7900HT sequence detection system (ABI Prism 7900 HT, Applied Biosystems, Inc., Foster City, CA). Before assay, these were rehydrated with 2.4 µL of deionized water with a Qfill2 automated pipetter (QFILL2 Genetix Ltd., Queensway, New Milton Hampshire, United Kingdom). For Assays-on-Demand products (denoted by a catalog number and paucity of sequence information in Table 1Citation Citation ), 2.5 µL of TaqMan Universal PCR master mix was added with 125 nL of Assay-on-Demand mix per well. For Assays-by-Design products or assays devised in house, 2.5 µL of PCR master mix was combined with 100 µmol of each primer (45 nl each) and 100 µmol of each probe (10 nl each). Plates were sealed and cycled at 95°C for 10 minutes, followed by 50 cycles of 95°C for 15 seconds, 58°C for 1 minute in an ABI GeneAmp PCR System 9700 thermocycler set for 9600 emulation. At the end of cycling, plates were held at 25°C until reading in a 7900 HT sequence detection system. Each plate contained controls of each genotype and no template controls. Data from plates failing any control were discarded. Manual genotype calls were done conservatively, consistent with the standards discussed in Clark et al.(13) and missing values were excluded from the analysis as detailed in Results. Aggregate indeterminate genotypes averaged 1% of total ESR1 SNPs sampled (range, 0.4–2%). Genotype tabulations and missing value details are provided as supplementary data.9


View this table:
[in this window]
[in a new window]

  		Table 1 Primers, probes, and genotyping assays

 

View this table:
[in this window]
[in a new window]

  		Table 1A Continued

 
The following SNPs were assayed in the ESR1 gene: rs851984, rs2881766, ESR1002, rs2077647, rs827421, rs9322331, rs712221, hCV1576295, rs1801132, rs1884051, rs6905370, rs926778, rs3020366, rs750686, rs2228480, rs3798577, rs910416; in the ESR2 gene: rs1255998, rs928554, rs1152579, rs4986938, rs1256049, E2EX4CorT, rs1256030, rs1271572; and in the PGR gene: rs511298, rs1042839, rs492457, rs1042838, rs2020876, rs578938, rs613120, rs566351, rs506487, rs1379130, rs3740753, PR331GORA, rs518162. These SNPs were chosen for assay based on the following criteria: (a) all nonsynonymous, splicing, and transcription factor binding site variants were examined; (b) allele frequency exceeded 10% in Caucasian samples; (c) they were available by Assay on Demand. Some novel SNPs that did not meet the frequency criteria were selected from ESR2 and PGR after resequencing. Probe and primer sequences, when available, are provided in Table 1Citation Citation , as are detailed order numbers for Assays-on-Demand products, when applicable.

A map with SNP location details is provided as Fig. 1Citation .



View larger version (38K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 1. Maps of typed steroid receptor gene SNPs. A, typed ESR1 polymorphisms. Top bar, maps all 17 typed polymorphisms with dbSNP names and including the three synonymous polymorphisms typed in the coding region (rs2077647 is ESR1390, which is also known as S10S in exon 1; rs1801132 is ESR1071 which is P325P in exon 4; and rs2228480 is ESR1031 which is T594T in exon 8). SYNE1, Nuclear Envelope Spectrin Repeat Protein 1, also known as Synaptic Nuclear Envelope Protein 1. Bottom bar, only those SNPs mapping to the transcribed (RefSeq) region of the gene. B, typed ESR2 polymorphisms. SYNE2, Nuclear Envelope Spectrin Repeat Protein 2, also known as Synaptic Nuclear Envelope Protein 2. C, typed PGR polymorphisms.

 
Statistical Methods.
SNP typing in control samples was checked for compliance with Hardy–Weinberg equilibrium with Tools For Population Genetic Analyses (TFPGA).10 Contingency table analysis for individual SNPs was carried out with SAS (SAS Institute, Inc., Cary, NC), SAS/Genetics (SAS Institute, Inc.), and SPSS (SPSS Inc., Chicago, IL). For haplotype estimation we used SNPhap,11 PHASE12 (14) , SNPEM13 (15) , MLOCUS14 (16) , an extension of the Clark (17) algorithm with expectation maximization (EM; ref. 18 ), and haplo.score15 (19) . Self-assigned demography was checked with STRUCTURE.16 Each analysis was conducted for the case–control population as a whole; stratifying by one of six ethnicities (Asian American, African American, Hispanic, Ashkenazi Jewish, Unknown, or European American) and by age (age ≤50 and age >50). To address the effect of age of menopause, a surrogate marker (age < or >50) was used. Five males with breast cancer and two male controls were included in the study, but excluded for the statistical analysis presented here. No result presented here was significantly impacted when we included or excluded these males. For the purpose of haplotype estimation, the most common alleles in the study as a whole were represented by "1" for ESR1 loci, and "A" for ESR2 loci; in addition, efforts are made to describe each haplotype by the base letter of the variants that compose it. Map coordinates provided are those from NCBI Build 33 (April 2003).17 SNP name designations are consistent with those defined in dbSNP when possible.

The strategy used to search for statistically significant haplotypes consisted of, first, using the EM algorithm or a Bayesian model to provide haplotype estimates; second, these estimated haplotypes were used to identify htSNPs (haplotype-tagging SNPs; ref. 20 ) representing a minimal informative subset of SNPs in each gene of interest (21) . Haplotype-tagging SNPs for ESR1 were chosen through use of the software program PHASEpybest.py.18 Rare haplotypes representing less than 1% of the total were deleted for the purpose of determining htSNPs. HT SNP Tester (at the same web site) was used to determine the final set of htSNPs used in the analysis, which included rare haplotypes. We used a permutation program [SNPEM (22) or PHASE 2.02 (14) ] to evaluate the statistical significance of any association observed between haplotype and disease state. Schaid’s program, haplo.score [Schaid et al. (19) ] which uses an E-M algorithm to estimate haplotypes and then tests disease association through a general linear model, was used to verify significant associations discovered with SNPEM. For sparse data, haplo.score computes simulation P-values for all score tests of association. In addition to our htSNP work, we performed two separate haplotype association analyses on blocks of strong linkage disequilibrium in ESR1.

Case–Control Sample Description.
Breast cancer cases consisted of 1,006 female patients with histologically confirmed breast cancer who presented for treatment or consultation at Memorial Sloan Kettering Cancer Center from January 2000 through December 2001. DNA was obtained from peripheral blood samples. Information was obtained on age at time of diagnosis of breast cancer, age at donation of blood sample, histologic confirmation of breast cancer diagnosis, sex, and ethnicity. All DNA samples were permanently anonymized according to an Institutional Review Board (IRB)-approved protocol. Samples were unselected for family history or any other demographic characteristic. Control subjects were drawn from the New York Cancer Study. The New York Cancer Study is a cohort study, modeled after the Framingham study, in which 18,187 individuals were enrolled from January 2000 through December 2002 (23) . Individuals were recruited at 14 sites in the New York metropolitan area. Volunteers were recruited by advertisement targeted to an ethnically diverse community. Volunteers were all individually interviewed and filled out a questionnaire that took 1 hour. Individuals were from 30 to 65 years of age. Individuals provided informed consent for use of DNA, which was obtained from 50 cc of whole blood. Information was gathered on age, sex, personal medical history, ethnicity (by 2000 Census nomenclature), family history, substance use, reproductive history and medication use. For the cohort group, a subset of 613 subjects were chosen who were female, free of breast cancer, and with age and ethnicity annotating each sample. Other information, including all personal identifiers were permanently removed, according to instructions of participating IRBs. IRB approval for the study stipulated that no clinical information beyond case or control designation, sex, ethnicity and age could be linked to samples subsequent to anonymization for DNA analysis. Controls were age, sex, and ethnically matched to cases. In both cases and controls, DNA was extracted from peripheral blood lymphocytes with Qiagen QIAamp Blood kits (Qiagen GmbH, Hilden, Germany) and then spectrophotometrically quantitated. Human subjects research approvals were obtained at Memorial Sloan-Kettering Cancer Center, and an exemption was obtained at the National Cancer Institute. Among 1,626 research subjects, 7 males were excluded from the final analysis on account of their sex. The inclusion or exclusion of these males had an insignificant impact on our statistical conclusions. The female case population was composed of 927 European Americans, 388 Ashkenazi Jews, 149 African Americans, 81 research subjects of Hispanic ethnicity, 39 Asians, and 35 subjects for whom no ethnicity or race was provided.

Verification Data Set.
On completion of genotyping and analysis from the New York Academic Medical Development Corporation (NY AMDeC) study, we were provided a set of incomplete genotyping data from 298 breast cancer cases and 94 controls from a study conducted at Vanderbilt University School of Medicine. Investigators there (24) had typed five SNPs in ESR1 during the early 1990s, but have recently added to their data set. Although these investigators have now typed seven SNPs (rs2077647, rs746432, Intron 1 PvuII (which is rs2234693 or c.454–397T>C), intron 1 XbaI (which is rs9340799 or c.454–351A>G), Exon 2 codon 160, rs1801132, and rs2228480), only four (rs746432, rs2234693, rs9340799, and rs1801132) provide sufficient information to assemble 564 useful haplotypes from 282 individuals (190 cases and 94 controls).


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Linkage Disequilibrium.
Both PGR and ESR2 SNPs have a large amount of linkage disequilibrium (Fig. 2B and C)Citation . Thus, we see a small number of haplotypes when we do frequency estimation by EM. ESR1 haplotypes indicated only moderate linkage disequilibrium in the region as represented by the D' statistic, and more modest linkage disequilibrium as measured by a correlation coefficient (R2; Fig. 2ACitation ). An estimation of the total frequencies was done by EM and PHASE as described in Materials and Methods. These showed 585 haplotypes for 1,626 individuals with 17 loci. Codominant segregation of five alleles of ESR1 (rs851984, ESR1002, rs2077647, rs1801132, and rs2228480) was confirmed by screening several Centre d’Etudes du Polymorphisme Humain (CEPH) families. European-American allele frequencies in the NY AMDeC study were compared with those of CEPH founders. Each was used separately for linkage disequilibrium estimation (data not shown). Our linkage disequilibrium measurements in ESR1 were consistent with those provided by Zuppan et al. (25) .



View larger version (42K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 2. Linkage disequilibrium measures with the ESR1, ESR2, and PGR SNPs in European-American controls. The upper right triangle plot in each panel provides a measure of D'; the lower left triangle plot in each panel provides a measure of the correlation coefficient (Pearson’s R2). A, linkage disequilibrium statistics for ESR1, along with displacement measures relating to coordinates in NCBI genome build 33 (April 2003).17 In the lower right hand corner of the figure, there is a legend for each of the disequilibrium measures. B, linkage disequilibrium statistics for ESR2. C, linkage disequilibrium statistics for PGR.

 
Analysis of Individual Single Nucleotide Polymorphisms in ESR1.
Seventeen SNPs were typed in ESR1 with an average distance of 25,606 bp. Fifteen of these 17 variants are in the public dbSNP, whereas two are unique to Celera (hCV1576295, ESR1002). The latter is a T/G SNP identified by Celera as located in the promoter region upstream of untranslated exon 1C. It is –1831 bp from the transcription start site specified by NM_000125 as elaborated in the April 2003 (Build 33, hg15) version of the human genome map.10 After checking Hardy–Weinberg equilibrium in each of the six sample groups (Asians, African Americans, Hispanics, Ashkenazi Jews, Unknown ethnicity, and European Americans), was carried out a two-by-two contingency table analysis on each SNP in every gene with affection status. Hardy–Weinberg compliance was obtained in each population in every control group, so long as cell numbers did not reach single digits. When cell numbers were very small, such as the Asian population, Fisher’s Exact Test evidenced no violations of Hardy–Weinberg in the control population. Each population was also stratified into two age categories: age >50 and age ≤50. A comparison of the age match between cases and controls is available as supplementary data at ftp://ftp.ncifcrf.gov/pub/users/goldb/in a folder labeled CANCERRESEARCH.9 Differences in the genotype distributions between cases and controls were tested with the {chi}2 test, Fisher’s Exact Test, and a Monte Carlo test. Although no SNPs were differentially distributed among aggregate cases and controls in a statistically significant way, when the population was age-stratified into an age 50-and-under and an over-50 age group, three ESR1 SNPs were statistically significantly associated (with P-values ~ 0.01, 0.001, and 0.003 by Fisher’s Exact Test) with disease in the Jewish population; these are ESR1002, rs2077647, and rs827421 respectively (see Fig. 3Citation ). Whereas ESR1002 is located at a putative promoter site, rs2077647 is a synonymous SNP in exon 1 often described as S10S, and rs827421 is located in intron 1. Among the Ashkenazi Jewish control subjects over age 50, 82, 43, and 57% were distributed with the more common allele, at the three SNPs, respectively; whereas 89, 56, and 45 of the cases, possess that allele. Both the 3 x 2 genotype distributions and the 2 x 2 allele distributions were statistically significant departures from that expected for these two SNPs. These SNPs implicate the region, denoted as the A/B region of the steroid hormone receptor, that encodes the ligand-dependent transactivation domain. As expected, a haplotype resulting from SNPs ESR1002 and rs2077647, which is the most common haplotype among the older Ashkenazi Jews, accounting for 50% of all haplotypes among them, 11 (T-C), is associated with an increased risk for breast cancer among Ashkenazi Jews over 50 years of age or older [odds ratio (OR), 1.706; 95% confidence interval (CI), 1.213–2.399; P < 0.05 from SNPEM, for the dominant model and OR = 2.916, 95% CI 1.598–5.320 for the recessive model]. The details of the contingency tables are available as supplementary data at ftp://ftp.ncifcrf.gov/pub/users/goldb/in a folder labeled CANCERRESEARCH.9



View larger version (23K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 3. Histogram and contingency table analysis of statistically significant susceptible ESR1 SNPs in presumptively postmenopausal Ashkenazi-Jewish patients and controls. A, bar graph representing the genotype counts of the putative promoter SNP ESR1002 in the older-than-age-50 Ashkenazi-Jewish case–controls under study. Case–control genotype comparison in the 3 x 2 contingency table provided a P-value of ~0.02 with Fisher’s Exact Test. B, allele counts of the ESR1002 SNP in the over-age-50 Ashkenazi-Jewish case–controls. The Fisher’s Exact Test P-value computed for this comparison is ~0.012. C, bar graph depicting the genotype counts of the S10S SNP in the over-age-50 Ashkenazi-Jewish case–controls. This case control genotype comparison provided a P-value ~0.0025 with Fisher’s Exact Test. D, allele counts of the S10S SNP in the over-age-50 Ashkenazi-Jewish case–control population. These provide a P-value of ~0.0025 with an OR of 1.706 (95% CI, 1.213–2.399) for rs2077647 alone. E, data on the distribution of ESR1 intron 1 genotypes (rs827421) in Ashkenazi-Jewish cases and controls over age 50. The {chi}2 P-value computed for this comparison is ~0.004. F, the allele counts for the same comparison of cases and controls in the over-age-50 Ashkenazi-Jewish population studied. This SNP evidences a P-value of ~0.003. Contingency table details for these SNPs and resultant haplotype can be found at ftp://ftp.ncifcrf.gov/pub/users/goldb/in a folder labeled CANCERRESEARCH.9

 
Typing of eight htSNP (haplotype tagging) sites, composed of rs851984, rs2881766, rs2077647, hCV1576295, rs1801132, rs6905370, rs2228480, and rs3798577 in the ESR1 locus, permitted identification of individual haplotypes associated with either an increased or a decreased risk of breast cancer (see Table 2Citation Citation ).


View this table:
[in this window]
[in a new window]

  		Table 2 Selected ESR1 haplotypes

 

View this table:
[in this window]
[in a new window]

  		Table 2A Continued

 
Protective Haplotypes of ESR1.
Three protective haplotypes, H4, H6 and H13, evidenced a statistically significant level of protection among overall female research subjects. When only female European-American study subjects were evaluated, only H6 conferred statistically significant protection. We sought to further define the location of the statistically protective signal through two methods: (a) redefining the haplotype with a rapid redefinition feature of SNPEM, and (b) stratifying the population by age. Research subjects over 50, bearing H6, evidenced significant protection that could be localized to the first six SNPs of the eight-SNP haplotype as H6b (see Table 2Citation Citation ). H6b was only marginally significant with the Dirichlet method in Northern Europeans. Peak significance of the protective haplotype was obtained when only the first six htSNPs were used to estimate haplotypes in the older population sample (P = 0.004). With these parameters, 1.8% of the older research subjects with a six-SNP haplotype, 112221 (G-T-T-C-C-G) provided an OR of 0.330 (95% CI, 0.136–0.799). The use of PHASE 2.02, a Bayesian haplotype estimation method, in case–control mode, affirmed the statistical significance of the association (data not shown).

Susceptible Haplotypes of ESR1.
Among all female study participants, H3 and H4 evidenced statistically significant susceptibility to breast cancer. H3 evidenced susceptibility when only older study subjects were considered. This haplotype (H3) was very rare among African Americans, and did not evidence significant susceptibility in study participants ages 50 or under. Haplotype H8: 12111111 (G-G-C-T-G-G-G-T) was not associated with an increased risk of breast cancer when the data set as a whole was considered, but was significant by two measures (P-value from SNPEM = 0.014 and P-value from haplo.score ~0.02) when only self-identified Ashkenazi-Jewish research subjects were analyzed, although the 95% CIs encompass unity when males are excluded from the analysis [OR, 3.292; 95% CI, 0.945–11.471 (OR, 3.706; 95% CI, 1.076–12.757; P < 0.01 with males included in the analysis)]. A separate susceptible haplotype localized to five (of the eight htSNPs) distal to the protective haplotype. In this analysis, rs2077647, hCV1576295, rs1801132, rs6905370, and rs2228480 formed haplotype H2a 11221 (C-T-C-A-G) that accounted for 2.6% of the study subjects with a P < 0.001 for risk, an OR of 1.776 (95% CI, 1.001–3.152). This is consistent with a susceptible haplotype mapping to exon 4 and intron 4, which might implicate the ligand-binding domain of ER{alpha}.

We reanalyzed the ESR1 data with two blocks of SNPs in clear linkage disequilibrium for association with breast cancer, i.e., SNPs 2, 3, 4, 5, 6, and 7 as one block and SNPs 10, 11, 12, and 13 as another block for association analysis. We found some susceptibility in older European-American study participants (P < 0.05) in a haplotype of 221 or G-A-C for rs1884051-rs6905370-rs926778, but the ORs varied only between 1.4 and 1.8, with each 95% CI overlapping unity, yet with significance with the SNPEM permutation algorithm (22) preserved among older European Americans (P < 0.05). This haplotype was very rare in Jews, observed only twice, both times among cases.

Results from the Vanderbilt Validation Data Set.
One haplotype, consisting of four SNPs, from a genotype complete subset of the Vanderbilt case–control study (see Materials and Methods), provided a statistically significant association with breast cancer when analyzed with SNPEM and when ORs were computed with SNPhap and SPSS. This haplotype 1211 (C-C-A-G) provided a P-value <0.05 through comparison with a permuted distribution in SNPEM, and a P < 0.004 with Fisher’s Exact Test (OR, 4.619, 95% CI, 1.378–15.481). The haplotype was present in 29 (~ 5%) of the estimated 568 haplotypes. Among cases, 26 among 345 weighted haplotypes were 1211 (C-C-A-G); among 173 weighted control haplotypes only 3 were 1211 (C-C-A-G). This data set and haplotype also evidenced a greater risk for breast cancer for bearers over age 50 (P < 0.01 by Fisher’s Exact Test); limiting the analysis to that group demonstrated an OR of 7.9 (95% CI, 1.025–60.868). With weighted probabilities and SNPhap on subjects over age 50 to determine the haplotypes, there were 14 cases and only 1 control with this haplotype. Whether 1211 (C-C-A-G) in the Vanderbilt data set overlaps with, or is identical to, one of the haplotypes in the NY AMDeC data set cannot be determined at this time, because there is only a single SNP typed in common in the two case–control data sets (rs2077547, or S10S).

Susceptible Haplotype in ESR2 among Ashkenazi Jews.
The 388 self-identified Ashkenazi Jewish females in the study were typed for the eight markers in ESR2 at 14q23.2 described in Fig. 1Citation . One haplotype, E2H1, formed from the last seven of these SNPs, AAAAAAA (T-C-G-G-T-A-C), beginning with rs1256030, manifested a statistically significant (P = 0.037) susceptibility for breast cancer (OR, 2.317; 95% CI, 1.042–5.155; see Table 3Citation ). This was the third most common ESR2 haplotype among the Ashkenazi Jews tested, making up 11.7% of the predicted haplotypes. The 382 Ashkenazim in the study for whom complete genotypes were provided by TaqMan were further investigated with SNPEM. On reduction of the haplotype under investigation to the final five 3' SNPs in ESR2, a new but related haplotype, E2H5, was statistically significant among the Ashkenazim, AAAAA, (C-G-G-T-A; P = 0.001 OR, 1.82; 95% CI, 1.213–2.737). Localization of the SNPs providing the haplotype association has been possible through the use of the SNPEM permutation algorithm (Table 3)Citation . This implicated the four distal (3') SNPs of ESR2 as giving rise to the significant haplotype because the exon 4 C or T polymorphism identified in this study is invariant among the Ashkenazi Jewish subjects.


View this table:
[in this window]
[in a new window]

  		Table 3 Significant ESR2 haplotypes in Ashkenazi-Jewish NY AMDEC research subjects

 

   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
These results suggest that a portion of hereditary predisposition to breast cancer can be accounted for by allele polymorphism in genes in the steroid hormone pathway. Individual differences in hormonal regulation may result from haplotypes that confer an increased risk or protection from risk of breast cancer in a subset of the population.

Although a linkage between late-onset breast cancer and ESR1 was published in 1991 (25) and several published studies find statistically significant associations between ESR1 polymorphisms and breast cancer (24 , 26, 27, 28, 29, 30) , other studies have not shown linkage or association (31, 32, 33) . Each of these studies either had a relative paucity of samples or chose only a few SNPs, with no efforts to generate haplotypes for haplotype estimation and association analysis. Recently, a breast cancer protective association was reported for a GT dinucleotide repeat polymorphism 6627 bp upstream from the transcription start site of ESR1 exon 1 in a large, ethnically homogeneous Han Chinese cohort (GT18 allele; OR, 0.58; 95% CI, 0.36–0.94; ref. 29 ). Previously, that same group had reported a PvuII polymorphism in intron 1 that was associated with increased breast cancer risk (genotype pp; OR, 1.4; 95% CI, 1.1–1.8; ref. 30 ). Data suggest that the most frequent allele of this common PvuII polymorphism eliminates a functional binding site for transcription factor B-myb, which may, therefore, down-regulate ESR1 transcription (34 , 35) . Although we have not typed the variants described by Cai et al. (29 , 30) , the magnitude and direction of the protective effects seen by them are consistent with the haplotype data reported here.

Since the discovery of ESR2 in 1996 (36 , 37) , several groups have characterized its unique expression profile (38, 39, 40, 41, 42, 43, 44) , but few have searched for polymorphisms associated with breast cancer risk (45, 46, 47) . Five ESR2 polymorphisms have been identified in the German population (48) , among which, one, rs1256049, provided evidence for an association with anorexia nervosa. This same ESR2 polymorphism had a highly statistically significant association with ovulatory dysfunction in a Chinese population (49) . In addition, an intragenic CA repeat polymorphism in ESR2 has been associated with bone mineral density in a Japanese research subject population (50) . More recently, the Shanghai breast study group reported an ESR2 exon 7 synonymous SNP (rs1256054, L392L) as conferring increased risk of breast cancer (OR, 2.37; 95% CI, 1.18–4.77) in a robust study (47) . They hypothesized that this SNP may act as an exonic splicing enhancer. We are currently in the process of typing this SNP; although preliminary results indicate that it is quite rare and will likely be uninformative in our European-American and Ashkenazi-Jewish populations (data not shown).

We typed six intronic and 3' SNPs in addition to six of those characterized by De Vivo et al. (51 : +44 C/T, +331 G/A, S344T, G393G, V660L, and H770H). As can be seen through inspection of Fig. 2CCitation , we found significant linkage disequilibrium throughout the PGR gene as gauged by either the D' or R2 statistic computed for the 12 SNPs that we typed. However, we found no single SNP, nor any haplotype, that was significantly associated with breast cancer when we stratified by age, ethnicity, or both. We sought to determine why we were not able to replicate the association of the functional +331 G/A polymorphism with breast cancer risk reported by De Vivo et al. (51) . Whereas they found 87% GG at this site among their cases and 90% GG among their controls, we found 93.4% and 93% among our cases and controls, respectively; and whereas they found 87% of this same genotype among their postmenopausal cases with 90% among their postmenopausal controls, we found 93% GG and 93.2% GG among our older cases and controls, respectively. We sought to determine whether this disparity could be explained by demographic differences between the Nurses Health Study participants, polled by De Vivo, and our research subjects by stratifying our results by both ethnicity and age. This analysis of +331 G/A genotypes in cases and controls also yielded no association.

This study has several limitations inherent in its study design. To gain IRB approval, we permanently anonymized DNA samples after collection of a minimum clinical annotation. Thus, it is not now possible to retrospectively examine clinical or demographic records to examine additional potential confounding variables, such as endogenous or exogenous estrogen exposure and other environmental variables. To address the effect of age of menopause, we used a surrogate marker (age < or >50).

Second, our study, although among the largest to date, still lacks statistical power to come to firm conclusions concerning the relationship of ESR1, ESR2, or PGR SNPs or haplotypes with respect to the specific populations tested. Our genotyping adhered to reproducibility and control standards published elsewhere (13) , and we meet recently published genotyping standards (52) . In addition, we demonstrated haplotype segregation in the CEPH families on the five initial SNPs provided by Celera and tested and obtained Hardy–Weinberg equilibrium in each of our control populations, even although some were quite small.

Although we made an effort to verify our findings in an independent data set, this additional analysis was also underpowered. Nonetheless, the Vanderbilt verification data set confirmed the existence of an ESR1 risk haplotype, which includes the aforementioned PvuII SNP, although it may not be precisely the same haplotype identified in the NY AMDeC study or the Shanghai breast cancer study.

With regard to the statistical limitations inherent in our study design, we used a variety of methods of imputing haplotypes to provide evidence for statistically significant disease associations. Whereas Nyholt (53) has emphasized a critical need for multiple test corrections in disease association studies to avoid Type I error, Krawczak et al., (54) and others (55 , 56) have questioned the overzealous application of the Bonferroni (57) correction. In an attempt to avoid potential errors in inference about associations stemming from the method of determination of haplotypes, we applied several different haplotype inference procedures. We observed that the methods found the common haplotypes accurately, but there was some disagreement in the rare haplotypes. With a variety of methods, some assuming underlying normal distributions, others being permutation based and still others being Bayesian, we have found similar haplotype frequencies in the genes under investigation. Such methods include those of Fallin et al.(15) and Fallin and Schork (22 , 58) , whose haplotype assignment, affection status permutation algorithm has been adopted by Schaid et al. (19) and modified by Stephens and Donnelly (59) . These methods (22) use EM haplotype estimations, which assume compliance with Hardy–Weinberg equilibrium, whereas Stephens’ Gibbs sampling algorithm bases haplotype estimations on existent allele frequencies. In all cases of significant association between haplotypes and breast cancer risk, these methods provided consistent haplotype inferences.

Using a variety of statistical and laboratory methods, we have discovered and validated the presence of common polymorphisms in three sex steroid hormone receptor genes. These candidate genes for disease association analysis have functional significance for the etiology under study and, therefore, cannot be thought of as impartially selected for statistical association testing. With cases and controls from the same geographical area and matched for age, we analyzed patterns of linkage disequilibrium and affection trait association with these genetic variants. We were able to define ESR1 haplotypes that conferred significant association with breast cancer risk in a North American population. Whereas the "protective" alleles that have been identified are quite rare in the populations studied, and their overall contribution to disease may be quite small, this analysis of SNP genotypes provides a means to associate variants in steroid hormone receptor genes and the breast cancer phenotype. Continued study of haplotypes of candidate genes in the steroid hormone receptor signal transduction pathway will provide additional insight into the biology of breast neoplasia.


   ACKNOWLEDGMENTS
 
The authors thank Drs. Vanessa Clark and Daniele Fallin for helpful comments on the statistical analysis. Drs. Robert Stephens and Matthew Stephens graciously provided recompiled software. The authors would like to acknowledge the New York Cancer Project, which, in connection with the publication of this study, made available biological samples from and information on control individuals. The New York Cancer Project is administered and funded by AMDeC Foundation, Inc. The content of this publication does not necessarily reflect the views of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the United States Government.


   FOOTNOTES
 
Grant support: Funded in whole or in part with federal funds from the Center for Cancer Research of the National Cancer Institute and the NIH under contract NO1-CO-12400, the Barbara Goldsmith Foundation, The Lymphoma Foundation, The Frankel Foundation, and the Academic Medicine Development Company. The New York Cancer Project is administered and funded by AMDeC Foundation, Inc.

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

Requests for reprints: Bert Gold, Human Genetics Section, Laboratory of Genomic Diversity, Center for Cancer Research, National Cancer Institute at Frederick, Building 560, Room 21-21, Frederick, MD 21702. Phone: (301) 846-5098; Fax: (301) 846-1909; E-mail: goldb{at}ncifcrf.gov

8 Genetic loci are identified in this publication as ESR1, ESR2 and PGR consistent with Human Genome Organization (HUGO) guidelines. ER{alpha}, ERß and PgR refer to the respective peptide products. Back

9 ftp://ftp.ncifcrf.gov/pub/users/goldb/ in a folder labeled CANCERRESEARCH. Back

10 Authored by Dr. Mark P. Miller (Mark.Miller{at}cnr.usu.edu; Utah State University, Logan, UT) and available from the web page http://bioweb.usu.edu/mpmbio/ or by writing to the author. Back

11 By David Clayton (david.clayton{at}cimr.cam.ac.uk; Cambridge Institute for Medical Research, Cambridge, United Kingdom), downloaded from http://www-gene.cimr.cam.ac.uk/clayton/software/. Back

12 Written by Matthew Stevens (stephens{at}stat.washington.edu; University of Washington, Seattle, WA) and available from his web site, http://www.stat.washington.edu/stephens/software.html. Back

13 Written by Dr. Nicholas Schork and M. Daniele Fallin, and obtained from dfallin{at}jhsph.edu Back

14 Obtained from Dr. Jeffrey C. Long (longjc{at}umich.edu; University of Michigan Medical School, Ann Arbor, MI). Back

15 The R-version of haplo.score was authored by Dr. Daniel J. Schaid (schaid{at}mayo.edu; Mayo Clinic, Rochester, MN). Back

16 Authored by Dr. Jonathan Pritchard (pritch{at}uchicago.edu; The University of Chicago, Chicago, IL) and available at his web site, http://pritch.bsd.uchicago.edu/software.html. Back

17 http://genome.ucsc.edu. Back

18 Written by Ross Lazarus at http://www.innateimmunity.net. Back

Received 4/ 9/04. Revised 10/ 1/04. Accepted 10/ 6/04.


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Newman B, Mu H, Butler LM, Millikan RC, Moorman PG, King MC Frequency of breast cancer attributable to BRCA1 in a population-based series of American women. JAMA 1998;279:915-21.[Abstract/Free Full Text]
  2. Vehmanen P, Friedman LS, Eerola H, et al A low proportion of BRCA2 mutations in Finnish breast cancer families. Am J Hum Genet 1997;60:1050-8.[Medline]
  3. Whittemore AS, Gong G, Itnyre J Prevalence and contribution of BRCA1 mutations in breast cancer and ovarian cancer: results from three U.S. population-based case-control studies of ovarian cancer. Am J Hum Genet 1997;60:496-504.[Medline]
  4. Lichtenstein P, Holm NV, Verkasalo PK, et al Environmental and heritable factors in the causation of cancer–analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med 2000;343:78-85.[Abstract/Free Full Text]
  5. Clark GM, McGuire WL, Hubay CA, Pearson OH, Marshall JS Progesterone receptors as a prognostic factor in Stage II breast cancer. N Engl J Med 1983;309:1343-7.[Abstract]
  6. Parl FF . Estrogens, estrogen receptor and breast cancer 2000 IOS Press/Ohmsha Amsterdam
  7. de Jong MM, Nolte IM, te Meerman GJ, et al Genes other than BRCA1 and BRCA2 involved in breast cancer susceptibility. J Med Genet 2002;39:225-42.[Abstract/Free Full Text]
  8. Jordan VC Selective estrogen receptor modulation: concept and consequences in cancer. Cancer Cell 2004;5:207-13.[CrossRef][Medline]
  9. Hanstein B, Djahansouzi S, Dall P, Beckmann MW, Bender HG Insights into the molecular biology of the estrogen receptor define novel therapeutic targets for breast cancer. Eur J Endocrinol 2004;150:243-55.[Abstract]
  10. Tempfer CB, Schneeberger C, Huber JC Applications of polymorphisms and pharmacogenomics in obstetrics and gynecology. Pharmacogenomics 2004;5:57-65.[CrossRef][Medline]
  11. Lymberis SC, Parhar PK, Katsoulakis E, Formenti SC Pharmacogenomics and breast cancer. Pharmacogenomics 2004;5:31-55.[CrossRef][Medline]
  12. Bland KI, Copeland EM . The breast: comprehensive management of benign and malignant disorders 3rd ed. 2004 Saunders St. Louis
  13. Clark VJ, Metheny N, Dean M, Peterson RJ Statistical estimation and pedigree analysis of CCR2-CCR5 haplotypes. Hum Genet 2001;108:484-93.[CrossRef][Medline]
  14. Stephens M, Smith NJ, Donnelly P A new statistical method for haplotype reconstruction from population data. Am J Hum Genet 2001;68:978-89.[CrossRef][Medline]
  15. Fallin D, Cohen A, Essioux L, et al Genetic analysis of case/control data using estimated haplotype frequencies: application to APOE locus variation and Alzheimer’s disease. Genome Res 2001;11:143-51.[Abstract/Free Full Text]
  16. Long JC . Multiple locus haplotype analysis (MLOCUS, OBSHAP, PAIRWISE), Software and documentation distributed by the author. Section on population genetics and linkage 2.0 ed. 1999 Laboratory of Neurogenetics, NIAAA, National Institutes of Health Bethesda, MD
  17. Clark AG Inference of haplotypes from PCR-amplified samples of diploid populations. Mol Biol Evol 1990;7:111-22.[Abstract]
  18. Fullerton SM, Clark AG, Weiss KM, et al Sequence polymorphism at the human apolipoprotein AII gene (APOA2): unexpected deficit of variation in an African-American sample. Hum Genet 2002;111:75-87.[CrossRef][Medline]
  19. Schaid DJ, Rowland CM, Tines DE, Jacobson RM, Poland GA Score tests for association between traits and haplotypes when linkage phase is ambiguous. Am J Hum Genet 2002;70:425-34.[CrossRef][Medline]
  20. Johnson GC, Esposito L, Barratt BJ, et al Haplotype tagging for the identification of common disease genes. Nat Genet 2001;29:233-7.[CrossRef][Medline]
  21. Bafna VH, BV, Schwartz RS, Clark AG, Istrail S. Haplotypes and informative SNP selection algorithms: don’t block out information. Proceedings of the Seventh Annual International Conference on Computational Molecular Biology (RECOMB) 2003:19–27.
  22. Fallin D . Haplotype-based approaches to genetic case-control studies 2001 Case Western Reserve University Cleveland, Ohio [Ph.D. Dissertation]
  23. Mitchell MK, Gregersen PK, Johnson S, Parsons R, Vlahov D The New York Cancer Project: rationale, organization, design, and baseline characteristics. J Urban Health 2004;81:301-10.[CrossRef][Medline]
  24. Roodi N, Bailey LR, Kao WY, et al Estrogen receptor gene analysis in estrogen receptor-positive and receptor-negative primary breast cancer. J Natl Cancer Inst (Bethesda) 1995;87:446-51.[Abstract/Free Full Text]
  25. Zuppan P, Hall JM, Lee MK, Ponglikitmongkol M, King MC Possible linkage of the estrogen receptor gene to breast cancer in a family with late-onset disease. Am J Hum Genet 1991;48:1065-8.[Medline]
  26. Andersen TI, Heimdal KR, Skrede M, Tveit K, Berg K, Borresen AL Oestrogen receptor (ESR) polymorphisms and breast cancer susceptibility. Hum Genet 1994;94:665-70.[Medline]
  27. Iwase H, Greenman JM, Barnes DM, Hodgson S, Bobrow L, Mathew CG Sequence variants of the estrogen receptor (ER) gene found in breast cancer patients with ER negative and progesterone receptor positive tumors. Cancer Lett 1996;108:179-84.[CrossRef][Medline]
  28. Wang M, Dotzlaw H, Fuqua SA, Murphy LC A point mutation in the human estrogen receptor gene is associated with the expression of an abnormal estrogen receptor mRNA containing a 69 novel nucleotide insertion. Breast Cancer Res. Treat 1997;44:145-51.[CrossRef][Medline]
  29. Cai Q, Gao YT, Wen W, et al Association of breast cancer risk with a GT dinucleotide repeat polymorphism upstream of the estrogen receptor-alpha gene. Cancer Res 2003;63:5727-30.[Abstract/Free Full Text]
  30. Cai Q, Shu XO, Jin F, et al Genetic polymorphisms in the estrogen receptor alpha gene and risk of breast cancer: results from the Shanghai Breast Cancer Study. Cancer Epidemiol Biomark Prev 2003;12:853-9.[Abstract/Free Full Text]
  31. Southey MC, Batten LE, McCredie MR, et al Estrogen receptor polymorphism at codon 325 and risk of breast cancer in women before age forty. J Natl Cancer Inst (Bethesda) 1998;90:532-6.[Abstract/Free Full Text]
  32. Schubert EL, Lee MK, Newman B, King MC Single nucleotide polymorphisms (SNPs) in the estrogen receptor gene and breast cancer susceptibility. J Steroid Biochem Mol Biol 1999;71:21-7.[CrossRef][Medline]
  33. Kang HJ, Kim SW, Kim HJ, et al Polymorphisms in the estrogen receptor-alpha gene and breast cancer risk. Cancer Lett 2002;178:175-80.[CrossRef][Medline]
  34. Herrington DM, Howard TD, Brosnihan KB, et al Common estrogen receptor polymorphism augments effects of hormone replacement therapy on E-selectin but not C-reactive protein. Circulation 2002;105:1879-82.[Abstract/Free Full Text]
  35. Schuit SC, Oei HH, Witteman JC, et al Estrogen receptor alpha gene polymorphisms and risk of myocardial infarction. JAMA 2004;291:2969-77.[Abstract/Free Full Text]
  36. Kuiper GG, Enmark E, Pelto-Huikko M, Nilsson S, Gustafsson JA Cloning of a novel receptor expressed in rat prostate and ovary. Proc Natl Acad Sci USA 1996;93:5925-30.[Abstract/Free Full Text]
  37. Mosselman S, Polman J, Dijkema R ER beta: identification and characterization of a novel human estrogen receptor. FEBS Lett 1996;392:49-53.[CrossRef][Medline]
  38. Zou A, Marschke KB, Arnold KE, et al Estrogen receptor beta activates the human retinoic acid receptor alpha-1 promoter in response to tamoxifen and other estrogen receptor antagonists, but not in response to estrogen. Mol Endocrinol 1999;13:418-30.[Abstract/Free Full Text]
  39. Bieche I, Parfait B, Laurendeau I, Girault I, Vidaud M, Lidereau R Quantification of estrogen receptor alpha and beta expression in sporadic breast cancer. Oncogene 2001;20:8109-15.[CrossRef][Medline]
  40. Poola I, Abraham J, Liu A Estrogen receptor beta splice variant mRNAs are differentially altered during breast carcinogenesis. J Steroid Biochem Mol Biol 2002;82:169-79.[CrossRef][Medline]
  41. Poola I, Clarke R, DeWitty R, Leffall LD Functionally active estrogen receptor isoform profiles in the breast tumors of African American women are different from the profiles in breast tumors of Caucasian women. Cancer (Phila) 2002;94:615-23.
  42. Saji S, Omoto Y, Shimizu C, et al Clinical impact of assay of estrogen receptor beta cx in breast cancer. Breast Cancer 2002;9:303-7.[Medline]
  43. Omoto Y, Eguchi H, Yamamoto-Yamaguchi Y, Hayashi S Estrogen receptor (ER) beta1 and ERbetacx/beta2 inhibit ERalpha function differently in breast cancer cell line MCF7. Oncogene 2003;22:5011-20.[CrossRef][Medline]
  44. Weihua Z, Andersson S, Cheng G, Simpson ER, Warner M, Gustafsson JA Update on estrogen signaling. FEBS Lett 2003;546:17-24.[CrossRef][Medline]
  45. Forsti A, Zhao C, Israelsson E, Dahlman-Wright K, Gustafsson JA, Hemminki K Polymorphisms in the estrogen receptor beta gene and risk of breast cancer: no association. Breast Cancer Res Treat 2003;79:409-13.[CrossRef][Medline]
  46. Hasegawa S, Miyoshi Y, Ikeda N, et al Mutational analysis of estrogen receptor-beta gene in human breast cancers. Breast Cancer Res Treat 2003;78:133-4.[CrossRef][Medline]
  47. Zheng SL, Zheng W, Chang BL, et al Joint effect of estrogen receptor beta sequence variants and endogenous estrogen exposure on breast cancer risk in Chinese women. Cancer Res 2003;63:7624-9.[Abstract/Free Full Text]
  48. Rosenkranz K, Hinney A, Ziegler A, et al Systematic mutation screening of the estrogen receptor beta gene in probands of different weight extremes: identification of several genetic variants. J Clin Endocrinol Metab 1998;83:4524-7.[Abstract/Free Full Text]
  49. Sundarrajan C, Liao WX, Roy AC, Ng SC Association between estrogen receptor-beta gene polymorphisms and ovulatory dysfunctions in patients with menstrual disorders. J Clin Endocrinol Metab 2001;86:135-9.[Abstract/Free Full Text]
  50. Ogawa S, Hosoi T, Shiraki M, et al Association of estrogen receptor beta gene polymorphism with bone mineral density. Biochem Biophys Res Commun 2000;269:537-41.[CrossRef][Medline]
  51. De Vivo I, Huggins GS, Hankinson SE, et al A functional polymorphism in the promoter of the progesterone receptor gene associated with endometrial cancer risk. Proc Natl Acad Sci USA 2002;99:12263-8.[Abstract/Free Full Text]
  52. Rebbeck TR, Ambrosone CB, Bell DA, et al SNPs, haplotypes, and cancer: applications in molecular epidemiology. Cancer Epidemiol Biomark Prev 2004;13:681-7.[Free Full Text]
  53. Nyholt DR Genetic case-control association studies—correcting for multiple testing. Hum Genet 2001;109:564-7.[CrossRef][Medline]
  54. Krawczak M, Boehringer S, Epplen JT Correcting for multiple testing in genetic association studies: the legend lives on. Hum Genet 2001;109:566-7.[CrossRef]
  55. Boehringer S, Epplen JT, Krawczak M Genetic association studies of bronchial asthma—a need for Bonferroni correction?. Hum Genet 2000;107:197[CrossRef][Medline]
  56. Perneger TV What’s wrong with Bonferroni adjustments. BMJ 1998;316:1236-8.[Free Full Text]
  57. Bonferroni CE Teoria statistica delle classi e calcolo delle probabilità. Pubblicazioni del R Instituto Superiore di Scienze Economiche e Commericiali di Firenze 1936;8:3-62.
  58. Fallin D, Schork NJ Accuracy of haplotype frequency estimation for biallelic loci, via the expectation-maximization algorithm for unphased diploid genotype data. Am J Hum Genet 2000;67:947-59.[CrossRef][Medline]
  59. Stephens M, Donnelly P A comparison of Bayesian methods for haplotype reconstruction from population genotype data. Am J Hum Genet 2003;73:1162-9.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 J. Nutr.Home page
E. Sonestedt, M. I. L. Ivarsson, S. Harlid, U. Ericson, B. Gullberg, J. Carlson, H. Olsson, H. Adlercreutz, and E. Wirfalt
The Protective Association of High Plasma Enterolactone with Breast Cancer Is Reasonably Robust in Women with Polymorphisms in the Estrogen Receptor {alpha} and {beta} Genes
J. Nutr., May 1, 2009; 139(5): 993 - 1001.
[Abstract] [Full Text] [PDF]

Home page
 Hum Mol GenetHome page
A. M. Dunning, C. S. Healey, C. Baynes, A.-T. Maia, S. Scollen, A. Vega, R. Rodriguez, N. L. Barbosa-Morais, B. A.J. Ponder, for SEARCH, et al.
Association of ESR1 gene tagging SNPs with breast cancer risk
Hum. Mol. Genet., March 15, 2009; 18(6): 1131 - 1139.
[Abstract] [Full Text] [PDF]

Home page
 CarcinogenesisHome page
S. Tchatchou, A. Jung, K. Hemminki, C. Sutter, B. Wappenschmidt, P. Bugert, B. H.F. Weber, D. Niederacher, N. Arnold, R. Varon-Mateeva, et al.
A variant affecting a putative miRNA target site in estrogen receptor (ESR) 1 is associated with breast cancer risk in premenopausal women
Carcinogenesis, January 1, 2009; 30(1): 59 - 64.
[Abstract] [Full Text] [PDF]

Home page
 Jpn J Clin OncolHome page
M. Hamaguchi, M. Nishio, T. Toyama, H. Sugiura, N. Kondo, Y. Fujii, and H. Yamashita
Possible Difference in Frequencies of Genetic Polymorphisms of Estrogen Receptor {alpha}, Estrogen Metabolism and P53 Genes Between Estrogen Receptor-positive and -negative Breast Cancers
Jpn. J. Clin. Oncol., November 1, 2008; 38(11): 734 - 742.
[Abstract] [Full Text] [PDF]

Home page
 J. Clin. Pathol.Home page
H. Grans, M. Nilsson, K. Dahlman-Wright, and B. Evengard
Reduced levels of oestrogen receptor {beta} mRNA in Swedish patients with chronic fatigue syndrome
J. Clin. Pathol., February 1, 2007; 60(2): 195 - 198.
[Abstract] [Full Text] [PDF]

Home page
 JNCI J Natl Cancer InstHome page
The Breast Cancer Association Consortium
Commonly studied single-nucleotide polymorphisms and breast cancer: results from the Breast Cancer Association Consortium.
J Natl Cancer Inst, October 4, 2006; 98(19): 1382 - 1396.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Epidemiol. Biomarkers Prev.Home page
L. Dossus, F. Canzian, R. Kaaks, A. Boumertit, and E. Weiderpass
No Association between Progesterone Receptor Gene +331G/A Polymorphism and Endometrial Cancer.
Cancer Epidemiol. Biomarkers Prev., July 1, 2006; 15(7): 1415 - 1416.
[Full Text] [PDF]

Home page
 Cancer Epidemiol. Biomarkers Prev.Home page
K. A. Pooley, C. S. Healey, P. L. Smith, P. D.P. Pharoah, D. Thompson, L. Tee, J. West, C. Jordan, D. F. Easton, B. A.J. Ponder, et al.
Association of the progesterone receptor gene with breast cancer risk: a single-nucleotide polymorphism tagging approach.
Cancer Epidemiol. Biomarkers Prev., April 1, 2006; 15(4): 675 - 682.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
G. S. Huggins, J. Y.Y. Wong, S. E. Hankinson, and I. De Vivo
GATA5 Activation of the Progesterone Receptor Gene Promoter in Breast Cancer Cells Is Influenced by the +331G/A Polymorphism
Cancer Res., February 1, 2006; 66(3): 1384 - 1390.
[Abstract] [Full Text] [PDF]

Home page
 JNCI J Natl Cancer InstHome page
M. T. Landi, P. A. Kanetsky, S. Tsang, B. Gold, D. Munroe, T. Rebbeck, J. Swoyer, M. Ter-Minassian, M. Hedayati, L. Grossman, et al.
MC1R, ASIP, and DNA Repair in Sporadic and Familial Melanoma in a Mediterranean Population
J Natl Cancer Inst, July 6, 2005; 97(13): 998 - 1007.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
G. S. Hageman, D. H. Anderson, L. V. Johnson, L. S. Hancox, A. J. Taiber, L. I. Hardisty, J. L. Hageman, H. A. Stockman, J. D. Borchardt, K. M. Gehrs, et al.
From The Cover: A common haplotype in the complement regulatory gene factor H (HF1/CFH) predisposes individuals to age-related macular degeneration
PNAS, May 17, 2005; 102(20): 7227 - 7232.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gold, B.
Right arrow Articles by Offit, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gold, B.
Right arrow Articles by Offit, K.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online