Skip to main content
PMC home page
Carcinogenesis logoLink to Carcinogenesis
. 2008 Nov 17;30(1):71–77. doi: 10.1093/carcin/bgn258

Polymorphisms in estrogen- and androgen-metabolizing genes and the risk of gastric cancer

Neal D Freedman 1,*, Jiyoung Ahn 1, Lifang Hou 1, Jolanta Lissowska 2, Witold Zatonski 2, Meredith Yeager 1, Stephen J Chanock 1, Wong Ho Chow 1, Christian C Abnet 1
PMCID: PMC2639034  PMID: 19015200

Abstract

Androgens and estrogens may play a role in gastric cancer etiology. To investigate the association of gastric cancer with single-nucleotide polymorphisms (SNPs) in six genes (COMT, CYP1B1, CYP17A1, CYP19A1, HSD17B1 and SHBG) involved in estrogen and androgen synthesis and metabolism, 58 haplotype-tagging SNPs were genotyped in 295 gastric cancer cases and 415 controls from a population-based study in Poland. We assessed differences in haplotype frequency between cases and controls using a global score test and calculated multivariate odds ratios (ORs) and 95% confidence intervals (CIs) for individual haplotypes using logistic regression. We found associations in one linkage disequilibrium (LD) block containing the 3′ untranslated region of COMT (rs9332377, rs165728, rs165849 and rs1110478), global score test (df = 4, P = 0.033). Relative to the most frequent GATA haplotype, the GATG haplotype was associated with statistically significant increased gastric cancer risk (OR = 1.50, 95% CI: 1.06–2.12; false discovery rate (FDR) value = 0.459) and the AACA haplotype with borderline increased risk (OR = 1.36, 95% CI = 1.00–1.85; FDR = 0.50). We also found associations for the LD block containing part of the SHBG coding region (rs6258, rs6259, rs2955617, rs1641544 and rs1641537). The CACCC haplotype was associated with statistically significant lower gastric cancer risk relative to the referent CGACC haplotype (OR = 0.55, 95% CI = 0.34–0.90; FDR = 0.459), but the overall score test was statistically non-significant. No other statistically significant associations were observed. In summary, we found possible associations between gastric cancer and polymorphisms in COMT, involved in estrogen inactivation, and SHBG, a modulator of hormone bioavailability. These findings should be interpreted cautiously until replicated in other studies.

Introduction

Gastric cancer is the second leading cause of cancer death worldwide (1). Incidence rates vary widely by geographic area. Men are two times more likely to develop gastric cancer than women in both high- and low-incidence areas (2). The ratio of male to female cases is greatest during the reproductive years and become more similar after the age of 60, corresponding with menopause (2). As normal and cancerous gastric tissue expresses steroid hormone receptors that bind to estrogens and androgens (3,4), one hypothesis suggests that sex hormones may play an etiological role in gastric cancer (2). The overall results from nearly 10 epidemiologic studies that investigated associations between menstrual and reproductive factors and gastric cancer risk (5–12) suggest an association between age of menopause, years of fertility or hormone replacement therapy and gastric cancer risk, although the associations did not reach statistical significance and case numbers were small in most studies. Two types of clinical data also support the hypothesis. Prostrate cancer patients treated with estrogen had reduced risk of developing gastric cancer (13) and women treated with the estrogen antagonist tamoxifen had a statistically non-significant increased risk of gastric cancer (14,15). Finally, male rats are more likely to develop gastric cancer than female rats; but tumor numbers become similar after male rats are treated with estrogen. Castrating male rats also lowers gastric cancer risk (16). Similar results have been found in a mouse model of gastric cancer (17).

A number of genes encode enzymes that synthesize and metabolize androgens and estrogens. CYP17A1 converts pregnenolone to dehydroepiandrosterone and progesterone to 17 alpha-hydroxyprogesterone (18). CYP19A1 converts the androstenedione to estrone and testosterone to estradiol (19). HSD17B1 converts androstenedione to testosterone and estrone to estradiol (19). SHBG encodes a plasma glycoprotein that transports androgens and estrogens in the circulation and that regulates the availability of estrogens and androgens (20). Among other functions, CYP1B1 and COMT convert estrogens to compounds that lack estrogenic activity. CYP1B1 catalyzes the C4 hydroxylation of estradiol to 4-hydroxy catechol estrogens, which can form adducts with DNA. COMT further metabolizes 4-hydroxy catechol estrogens into stable methylated catechol estrogens that do not form DNA adducts (21,22).

As androgens and estrogens may be related to gastric cancer risk, we hypothesized that genetic polymorphisms in these genes might be associated with gastric cancer.

Materials and methods

Study design

The methods of this population-based case–control study of gastric cancer have been described in detail previously (23). Cases are Caucasian residents of Warsaw, Poland, aged 21–79 years, who were newly diagnosed with gastric cancer (ICD-0 151) between 1994 and 1996 and identified by physicians in each of the 22 hospitals serving the study area. All the cases were confirmed to be gastric adenocarcinoma by study pathologists. Controls were randomly selected from a computerized registry of Warsaw residents and frequency matched to cases by sex and 5 year age groups. Informed consent was obtained from all participants and Institutional Review Boards of the USA National Cancer Institute (NCI) and the M. Sklodowska-Curie Memorial Cancer Center and Institute of Oncology, Warsaw, Poland, approved the study.

Trained interviewers collected demographic, lifestyle and medical information, including tobacco use, alcohol consumption, family history of gastric cancer, childhood living conditions, history of selected medical conditions, medication use and occupational history, and typical diet before 1990. Among 515 eligible cases and 549 eligible controls, in-person interviews were conducted with 324 cases (62.9%) and 480 controls (87.4%). Proxy interviews were conducted with the next of kin for 140 cases (27.2%) who were too ill to be interviewed or deceased. Of 464 gastric cancer cases and 480 controls with successful interviews, genomic DNA was obtained from peripheral blood lymphocytes of 305 cases (65.7%) and 427 controls (90.0%) as described previously (24).

Selection of haplotype tagging single-nucleotide polymorphisms

Data from the Breast and Prostate Cohort Consortium (BPC3) (25) were used to identify haplotype tagging single-nucleotide polymorphisms (SNPs) for common haplotypes in COMT, CYP1B1, CYP17A1, CYP19A1, HSD17B1 and SHBG. A two-stage approach was used by the Consortium to comprehensively measure genetic variation in these genes. First, a genomic region beginning 20 kb upstream of the transcription start site and ending 10 kb downstream of the stop codon was chosen. In addition to SNPs in the dbSNP database, SNPs in these regions were identified by sequencing a multiethnic panel of 190 breast and prostate cancer cases. Allele frequencies in Caucasians were calculated, and haplotypes were imputed by the expectation–maximization algorithm (26). We genotyped previously published haplotype-tagging SNPs identified by the BPC3 to predict common haplotypes in Caucasians for CYP17A1 [two linkage disequilibrium (LD) blocks and nine SNPs, Rh2 ≥ 0.85] (27), CYP19A1 (four LD blocks and 20 SNPs, Rh2 ≥ 0.75) (28) and HSD17B1 (one LD block and four SNPs, Rh2 ≥ 0.8) (29,30). Haplotype-tagging SNPs for CYP1B1 (two haplotype blocks and nine SNPs, Rh2 ≥ 0.7) and COMT (four LD blocks and 10 SNPs, Rh2 ≥ 0.9) have been selected by the BPC3, but have not yet been published. Details can be found on the website http://cgf1.nci.nih.gov/cohort.cfm. Two of three SNPs in LD block #1 of COMT failed validation and were not included. We did not analyze haplotypes in this block separately; to improve statistical efficiency without losing predictive power, we combined this block with LD blocks 2 and 3 into a single block. A previous report also analyzed LD blocks 1–3 as a single entity (31).

As BPC3 had not yet chosen tagging SNPs for SHBG at the time of genotyping, nine SNPs were chosen by the NCI Core Genotyping Facility from BPC3 LD data using a pair-wise R2 of ≥0.8 and >5% minor allele frequency in Caucasians with Tagzilla (http://tagzilla.nci.nih.gov/). The BPC3 identified three LD blocks in SHBG. To improve statistical efficiency, we combined SHBG LD blocks #1 and #2 into a single block for analysis. One assay in SHBG failed validation (rs2908809). Of successfully genotyped SNPs in SHBG, rs12150660 had the highest pair-wise LD with rs2908809 (R2 = 0.43), coverage of the remaining SNPs in SHBG did not change.

Genotyping

In total, 61 SNPs were genotyped in these six genes. All assays were performed at the NCI’s Core Genotyping Facility; description, methods and validation data for the assays can be found at http://snp500cancer.nci.nih.gov. Each assay was sequence validated in the 102 samples of the NCI SNP500Cancer project (32) and additionally validated against 210 HapMap control subjects. Complete concordance was observed for all assays, except for rs174674 and rs5748489 in LD block #1 of COMT and rs2908809 in LD block #1 of SHBG. These assays, therefore, were not included in this study.

Of the 58 validated assays, 38 polymorphisms were genotyped by SNPlex assay (Applied Biosystems, Foster City, CA) and the remaining 20 SNPs were genotyped by TaqMan (Applied Biosystems). At the laboratory, case–control status was masked and we included quality control samples from the NCI SNP500 project and blinded duplicate samples from our study. Concordance for the blinded quality control samples (∼10% of samples) was >97%. Genotypes were successfully obtained from >94% of participants for all SNPs with a median completion rate of 98%.

Table I presents details on genotyped SNPs, their chromosomal region and their refSNP number. Although we also genotyped rs605059 (HSD17B1), rs284849 (CYP17A1), rs2445765 (CYP19A1), rs28566535 (CYP19A1) and rs6493494 (CYP19A1), these SNPs were in strong LD (R2 > 0.9) with rs598126, rs2486758, rs2446405, rs1902586 and rs749292, respectively, and were not further analyzed. We had sufficient DNA to genotype polymorphisms by TaqMan assays in 710 participants (415 controls and 295 cases), but SNPlex assays required more DNA and therefore these assays were restricted to 676 participants (391 controls and 285 cases). Associations between other covariates and gastric cancer did not change when analyses were restricted to those with a DNA sample (data not shown).

Table I.

Polymorphisms examined in this study

Gene Chromosomal region LD blocka RS # Nearest gene to SNP, nucleotide changeb Genotyping platform MAF in controls
COMT 22q11.21 Block 1 rs4485648 TXNRD2, IVS3+475A>G TaqMan 0.168
rs7290221 COMT, IVS1−6042C>G SNPlex 0.435
rs4646312 COMT, IVS1−385T>C SNPlex 0.341
rs4633 COMT, Ex3−104C>T TaqMan 0.453
Block 2 rs9332377 ARVCF, 3058 bp 3′ of STP G>A SNPlex 0.159
rs165728 ARVCF, 1727 bp 3′ of STP G>A TaqMan 0.092
rs165849 ARVCF, IVS19+70C>T SNPlex 0.312
rs1110478 ARVCF, IVS11+440G>A SNPlex 0.195
CYP1B1 2p21 Block 1 rs10175368 CYP1B1, −5329G>A SNPlex 0.321
rs162556 CYP1B1, −3922C>T SNPlex 0.432
rs162557 CYP1B1, −2919C>T SNPlex 0.275
Block 2 rs1056836 CYP1B1, Ex3+251G>C TaqMan 0.449
rs1800440 CYP1B1, Ex3+315A>G TaqMan 0.14
rs162562 CYP1B1, Ex3+939C>A TaqMan 0.25
rs9341266 CYP1B1, Ex3−1249C>T TaqMan 0.029
rs163086 FAM82A, IVS10−1363C>T SNPlex 0.199
rs163077 FAM82A, IVS10−8520T>C SNPlex 0.262
CYP17A1 10q24.3 Block 1 rs7097872 CYP17A1, −9517G>A SNPlex 0.444
rs2486758 CYP17A1, −361G>A SNPlex 0.2
rs743572 CYP17A1, Ex1+27T>C SNPlex 0.416
rs4919687 CYP17A1, IVS1−99C>T TaqMan 0.298
Block 2 rs4919682 CYP17A1, 6128 bp 3′ of STP G>A TaqMan 0.302
rs10883782 CYP17A1, 6526 bp 3′ of STP C>T TaqMan 0.115
rs619824 CYP17A1, 9170 bp 3′ of STP G>T SNPlex 0.465
rs6892 CYP17A1, 14588 bp 3′ of STP C>T SNPlex 0.179
CYP19A1 15q21.1 Block 1 rs2446405 CYP19A1, −111683A>T SNPlex 0.149
rs2470144 CYP19A1, IVS1−5604A>G SNPlex 0.386
rs2445762 CYP19A1, IVS1−1587A>G SNPlex 0.228
rs1004984 CYP19A1, IVS2+2484C>T SNPlex 0.278
rs1902584 CYP19A12, IVS2+4359T>A SNPlex 0.039
Block 2 rs3751591 CYP19A1, IVS2+9303T>C SNPlex 0.145
rs2445759 CYP19A1, IVS2+15172C>A TaqMan 0.059
rs936306 CYP19A1, IVS2+36415G>A SNPlex 0.139
rs1902586 CYP19A1, IVS2−35706C>T SNPlex 0.058
Block 3 rs749292 CYP19A1, IVS2−23584C>T SNPlex 0.494
rs1008805 CYP19A1, IVS2−14452C>T SNPlex 0.376
Block 4 rs727479 CYP19A1, IVS3+418G>T TaqMan 0.325
rs28757184 CYP19A1, Ex6−27C>T TaqMan 0.017
rs700519 CYP19A1, Ex8+47C>T TaqMan 0.048
rs17601241 CYP19A1, IVS8+26C>T SNPlex 0.071
rs10046 CYP19A1, Ex11+268C>T SNPlex 0.423
rs4646 CYP19A1, Ex11+410T>G SNPlex 0.221
HSD17B1 17q11–q21 Block 1 rs676387 HSD17B1, IVS4−150C>A SNPlex 0.205
rs598126 COASY, Ex3+57A>G SNPlex 0.475
rs2010750 MLX, IVS7−174C>T TaqMan 0.437
SHBG 17p13–p12 Block 1 rs12150660 FXR2, −4064C>A TaqMan 0.248
rs13894 SAT2, Ex6+31C>T SNPlex 0.085
rs858521 SAT2, IVS4−36C>G SNPlex 0.373
Block 2 rs6258 SHBG, Ex4−2C>T TaqMan 0.01
rs6259 SHBG, Ex8+6G>A TaqMan 0.074
rs2955617 SHBG, 2114 bp 3′ of STP C>A TaqMan 0.324
rs1641544 SHBG, 3195 bp 3′ of STP T>C SNPlex 0.052
rs1641537 ATP1B2, −9115T>C TaqMan 0.127
Open in a new tab

MAF, minor allele frequency; RS, refSNP.

a

SNPs are listed 5′–3′ across each gene. LD blocks were defined using data from the BPC3 (25).

b

SNPs in the regulatory regions 20 kb upstream and 10 kb downstream of each gene of interest may include neighboring gene sequences. Further information on the nearest gene and nucleotide change of each SNP can be found at http://snp500cancer.nci.nih.gov.

Statistical analysis

We tested for Hardy–Weinberg (HW) equilibrium in controls using an exact test (33). All SNPs passed except for rs1110478 (P-value < 0.001) in COMT that was also out of HW equilibrium in Caucasians of the NCI SNP500 reference population (32).

We examined pair-wise LD in controls using Haploview (34). We used Haplo.Stats (version 1.3.1) (35) to compute common haplotype frequencies (those with 5% or greater frequency), odds ratios (ORs) and 95% confidence intervals (CIs) for haplotypes, assuming an additive model. Uncommon haplotypes (<5% prevalence) were included in the models as a single ‘rare haplotypes category’. A global score test, adjusted for age, gender, education and smoking (never, former, current and pack-years among smokers) was used to evaluate the overall difference in haplotypes between cases and controls (36). Alternative LD blocks defined within controls of our study (using a solid spline definition with D′ < 0.80) gave similar results to those presented in this manuscript using BPC3-defined LD blocks.

For individual SNPs, we used unconditional logistic regression in SAS version 9.1 to estimate ORs and 95% CIs for the gastric cancer risk associated with a single copy or two copies of the minor allele relative to carrying no copies. We tested the null hypothesis of no association using a likelihood ratio test that compared models with and without the SNP in question.

All ORs were adjusted for age, gender, education (less than high school, high school or technical training and some college/college graduate) and smoking (never, former, current and pack-years among smokers). Further adjustment for other potential confounding variables including alcohol, caloric intake, intake of fruits, vegetables, sausages, red meats or preserved vegetables, family history of gastric cancer or other cancers, gastroesophageal reflux disease or Helicobacter pylori serology did not alter risk estimates. To examine the possible effect modification by sex, we inspected data for men and women separately and formally tested for differences using a likelihood ratio test. All tests were two sided and an alpha level of <0.05 was considered statistically significant.

We evaluated the robustness of our findings by calculating the false discovery rate (FDR) (37) for both P-values from LD blocks and the P-value for trend of individual SNPs. Among all SNPs analyzed, FDR is the expected ratio of erroneous rejections of the null hypothesis to the total number of rejected hypotheses.

Results

By design, controls matched cases in their age and sex distribution. Cases smoked more and had less education than controls (Table II).

Table II.

Distribution of selected covariates in cases and controls from Warsaw, Poland

Gastric cancer cases (n = 295)
 Controls (n = 415)
 P-valuea
No. % No. %
Age (years)b 0.97
    <50 37 12.5 51 12.3
    50–59 54 18.3 72 17.4
    60–69 117 39.7 163 39.3
    ≥70 87 29.5 129 31.1
Genderb 0.69
    Male 196 62.4 269 64.8
    Female 99 33.6 146 35.2
Education 0.002
    Less than high school 140 47.5 157 37.8
    High school or technical training 94 31.9 123 29.6
    Some college/college graduate 61 20.7 135 32.5
Smoking statusc <0.001
    Never 85 29.0 168 40.5
    Former 85 29.0 132 31.8
    Current 123 42.0 115 27.7
Pack-years among ever smokers 0.014
    ≤16.5 47 22.6 62 25.1
    >16.5–28 38 18.3 64 25.9
    >28–40 48 23.1 65 26.3
    >40 75 36.1 56 22.7
Open in a new tab
a

Fisher’s exact test for distribution of differences between gastric cancer cases and control participants.

b

The study design frequency-matched controls by age and gender to the cases.

c

Totals do not add up to 295 cases or 415 controls due to missing data.

We found no association with LD blocks in CYP1B1, CYP17A1, CYP19A1 or HSD17B1 (Table III). In contrast, we found associations in one of two LD blocks in COMT and one of two LD blocks in SHBG. In COMT, a LD block (block #2) containing the 3′ untranslated region (UTR) (rs9332377, rs165728, rs165849 and rs1110478) had a statistically significant global score test (P = 0.033). This block had four common haplotypes with >5% prevalence. Relative to the most common haplotype (GATA), the GATG haplotype was associated with statistically significant increased risk of gastric cancer (OR: 1.50, 95% CI: 1.06–2.12), whereas the AACA haplotype had borderline increased risk (OR: 1.36, 95% CI: 1.00–1.85). For SHBG, LD block #2 (rs6258, rs6259, rs2955617, rs1641544 and rs1641537), which contains part of the SHBG coding region, had a statistically non-significant global score test (P = 0.14), but the CACCC haplotype had statistically significant lower risk than the referent CGACC haplotype (OR: 0.55, 95% CI: 0.34–0.90).

Table III.

Haplotype analysis of COMT, CYP1B1, CYP17A1, CYP19A1, HSD17B1 and SHBG and gastric cancer risk

Gene LD block # Haplotypea % of controls % of cases Global Pb OR-1c 95% CI OR-2d 95% CI
COMT Block #1 ACTT 35.8 29.3 0.22 1.00 Reference 1.00 Reference
AGCC 24.0 23.1 1.16 0.87–1.56 1.11 0.82–1.51
GGTT 9.6 9.7 1.17 0.76–1.79 1.16 0.75–1.79
ACTC 9.0 9.4 1.25 0.80–1.96 1.25 0.79–1.98
ACCC 7.8 9.9 1.48 0.96–2.30 1.45 0.92–2.29
Block #2 GATA 63.2 55.7 0.033 1.00 Reference 1.00 Reference
AACA 14.9 17.5 1.33 0.99–1.80 1.36 1.00–1.85
GGCG 9.6 8.2 0.94 0.63–1.42 0.97 0.64–1.47
GATG 5.2 8.9 1.43 1.02–2.01 1.50 1.06–2.12
CYP1B1 Block #1 TAG 32.1 32.5 0.27 1.00 Reference 1.00 Reference
CGG 23.0 26.1 1.12 0.83–1.51 1.07 0.78–1.46
CGA 20.3 17.3 0.75 0.44–1.28 0.80 0.57–1.12
CAG 17.4 18.5 1.06 0.76–1.47 1.08 0.77–1.51
CAA 7.3 5.6 0.75 0.44–1.28 0.71 0.41–1.23
Block #2 GTTGGG 30.2 30.8 0.52 1.00 Reference 1.00 Reference
CTTGAG 20.1 18.6 0.91 0.66–1.24 0.90 0.65–1.24
CTGGGA 18.1 17.6 0.95 0.69–1.31 0.88 0.63–1.23
GCTGGG 14.0 14.9 1.05 0.74–1.49 0.97 0.68–1.39
GTTGGA 7.4 9.8 1.33 0.86–2.07 1.32 0.84–2.08
CTGGGG 6.2 4.5 0.71 0.40–1.26 0.69 0.39–1.24
CYP17A1 Block #1 GATC 35.0 33.9 0.63 1.00 Reference 1.00 Reference
AACT 30.4 32.5 1.11 0.86–1.45 1.04 0.80–1.36
GGTC 19.4 20.6 1.12 0.82–1.53 1.05 0.76–1.45
AACC 12.1 11.2 0.98 0.68–1.42 0.93 0.63–1.37
Block #2 ATTT 30.0 30.0 0.92 1.00 Reference 1.00 Reference
GTGT 23.4 21.1 0.90 0.67–1.22 0.97 0.72–1.32
GTTT 17.2 16.8 0.98 0.70–1.37 1.00 0.71–1.41
GTGC 17.0 18.4 1.09 0.79–1.50 1.11 0.80–1.54
GCGT 10.9 11.9 1.10 0.77–1.59 1.13 0.78–1.65
CYP19A1 Block #1 TAACT 61.3 61.7 0.48 1.00 Reference 1.00 Reference
TGGTT 16.2 14.7 0.89 0.66–1.22 0.87 0.64–1.20
AGGCT 6.9 5.5 0.80 0.50–1.28 0.78 0.48–1.27
Block #2 TCGC 78.7 76.9 0.85 1.00 Reference 1.00 Reference
CCGC 6.9 7.5 1.16 0.75–1.81 1.13 0.72–1.77
Block #3 TT 49.3 46.7 0.88 1.00 Reference 1.00 Reference
CC 37.8 38.6 1.08 0.85–1.37 1.07 0.84–1.37
CT 12.8 14.4 1.19 0.85–1.68 1.10 0.77–1.56
Block #4 TCCCTG 53.9 51.5 0.41 1.00 Reference 1.00 Reference
GCCCCG 15.1 15.2 1.04 0.77–1.42 1.09 0.79–1.49
GCCCCT 14.3 15.8 1.15 0.84–1.59 1.14 0.83–1.58
TCCTCT 6.6 4.7 0.73 0.43–1.22 0.70 0.41–1.20
HSD17B1 Block #1 CAT 42.8 41.2 0.37 1.00 Reference 1.00 Reference
CGC 31.7 30.4 1.00 0.78–1.30 0.93 0.71–1.21
AGC 20.3 22.1 1.12 0.83–1.51 1.06 0.78–1.44
CAC 4.6 6.2 1.40 0.86–2.28 1.30 0.79–2.14
SHBG Block #1 CCG 37.5 39.7 0.46 1.00 Reference 1.00 Reference
CCC 29.0 27.2 0.88 0.67–1.16 0.82 0.62–1.09
ACC 24.9 25.4 0.98 0.74–1.29 0.96 0.72–1.27
CTC 8.6 7.4 0.79 0.51–1.22 0.76 0.49–1.18
Block #2 CGACC 67.4 70.3 0.14 1.00 Reference 1.00 Reference
CGCCC 11.0 9.2 0.79 0.55–1.15 0.78 0.53–1.14
CGCCT 7.8 9.2 1.15 0.78–1.70 1.01 0.68–1.51
CACCC 7.5 5.1 0.63 0.39–1.01 0.55 0.34–0.90
Open in a new tab
a

Haplotypes include SNPs listed in Table I and are presented 5′–3′. Haplotypes with <5% frequency were grouped together and included in the models as one category, but are not presented here.

b

A global score test was used to evaluate the overall difference in haplotype frequency in cases and controls.

c

Model-1: estimates are adjusted for the matching variables (age and gender).

d

Model-2: estimates are adjusted for age, gender, education and smoking (never, former, current and pack-years among smokers).

We next examined the individual SNPs within the two LD blocks that showed evidence for an association with gastric cancer risk (block #2 in COMT and block #2 in SHBG) (Table IV). For COMT, we found evidence for an association between gastric cancer risk and rs1110478 (OR per variant allele of 1.31, 95% CI: 1.02–1.67) and statistically non-significant increased risk with rs9332377 (OR per variant allele of 1.24, 95% CI: 0.93–1.67) and rs165849 (OR per variant allele of 1.23, 95% CI: 0.97–1.56). For SHBG, we found an association between rs6259 and gastric cancer risk (OR for GA/AA: 0.59, 95% CI: 0.36–0.96) and statistically non-significant reduced risk with rs2955617 (OR per variant allele of 0.81, 95% CI: 0.64–1.03). Individual data for all SNPs genotyped are presented in supplementary Table 1 (available at Carcinogenesis Online). We observed a statistically significant association between rs162557 (−2919C>T of CYP1B1) and gastric cancer risk (OR per variant allele of 0.74, 95% CI: 0.56–0.96). But, we did not observe evidence for an association of the LD block containing this SNP and gastric cancer risk (P = 0.27).

Table IV.

Association of selected SNPs and gastric cancer risk

Gene LD block # RS # Genotype No. of controls (%) No. of cases (%) Global Pa ORb 95% CI
COMT Block #2c rs9332377 GG 272 (71) 185 (67) 0.18 1.00 Reference
GA 102 (27) 76 (28) 1.10 0.76–1.58
AA 10 (3) 14 (5) 2.23 0.94–5.28
Per allele risk 1.24 0.93–1.67
rs165728 AA 328 (81) 240 (83) 0.44 1.00 Reference
AG 77 (19) 47 (16) AG/GG 0.85 0.57–1.28
GG 2 (0.5) 2 (1)
rs165849 TT 183 (48) 114 (41) 0.23 1.00 Reference
TC 162 (42) 126 (46) 1.27 0.90–1.78
CC 39 (10) 36 (13) 1.47 0.87–2.50
Per allele risk 1.23 0.97–1.56
rs1110478 AA 255 (68) 165 (61) 0.08 1.00 Reference
AG 94 (25) 74 (28) 1.18 0.81–1.71
GG 26 (7) 30 (11) 1.91 1.08–3.41
Per allele risk 1.31 1.02–1.67
SHBG Block #2d rs6258 CC 405 (98) 288 (98) 0.80 1.00 Reference
CT 8 (2) 5 (2) 0.87 0.27–2.74
rs6259 GG 348 (85) 261 (90) 0.029 1.00 Reference
GA 59 (15) 29 (10) GA/AA 0.59 0.36–0.96
AA 1 (0) 1 (0)
rs2955617 AA 187 (45) 147 (50) 0.17 1.00 Reference
AC 188 (45) 123 (42) 0.75 0.54–1.03
CC 40 (10) 25 (8) 0.74 0.42–1.30
Per allele risk 0.81 0.64–1.03
rs1641544 CC 345 (90) 249 (90) 0.86 1.00 Reference
CT 39 (10) 27 (10) CT/TT 0.96 0.56–1.63
TT 1 (0) 0 (0)
rs1641537 CC 314 (76) 215 (74) 0.88 1.00 Reference
CT 92 (22) 70 (24) 1.04 0.72–1.50
TT 6 (2) 6 (2) 1.34 0.41–4.40
Per allele risk 1.07 0.78–1.48
Open in a new tab

RS, refSNP.

a

Calculated from a likelihood ratio test comparing models with indicator variables for one or two copies of the minor allele to models without.

b

Estimates are based on an additive effects model and adjusted for age, gender, education and smoking (never, former, current and pack-years among smokers).

c

Pair-wise R2: rs9332377/rs165728 (0.01); rs9332377/rs165849 (0.41); rs9332377/rs1110478 (0.01); rs165728/rs165849 (0.22); rs165728/rs1110478 (0.42); rs165849/rs1110478 (0.15).

d

Pair-wise R2: rs6258/rs6259 (0); rs6258/rs2955617 (0.02); rs6258/rs1641544 (0); rs6258/rs1641537 (0); rs6259/rs2955617 (0.16); rs6259/rs1641544 (0); rs6259/rs1641537 (0.01); rs2955617/rs1641544 (0.10); rs2955617/rs1641537 (0.28); rs1641544/rs1641537 (0.31).

We evaluated the robustness of our findings using the FDR method. The FDR value for LD block #2 in COMT was 0.429. For haplotypes, the FDR value was 0.459 for the GATG haplotype of COMT and the CACCC haplotype of SHBG; the AACA haplotype of COMT had a FDR of 0.50. Whereas for individual SNPs, the FDR value was 0.583 for rs1110478 (COMT), rs6259 (SHBG) and rs162557 (CYP1B1).

Risk estimates did not differ by sex (P for interaction between haplotypes or individual genetic variants and sex was all >0.279), but we had low power for these tests. For example, the OR per variant allele of rs1110478 and gastric cancer risk was 1.12 (0.74–1.69) in women and 1.43 (1.04–1.95) in men; the OR for rs6259 (GA/AA versus GG) was 0.51 (0.17–1.50) in women and 0.62 (0.36–1.08) in men.

Discussion

We systematically investigated genetic variation in six genes (COMT, CYP1B1, CYP17A1, CYP19A1, HSD17B1 and SHBG) related to the synthesis and metabolism of estrogen and androgen using haplotype-tagging SNPs (25). We examined associations with gastric cancer risk using haplotype blocks and individual SNPs were merited. We found evidence for an association of gastric cancer risk with genetic polymorphisms in COMT and SHBG, but not with LD blocks in the other genes we examined.

CYP1B1 and COMT sequentially metabolize estrogens into methylated catechol estradiols that lack estrogenic activity. We found a statistically significant global P-test for a haplotype block in COMT (block #2 in our analysis) and gastric cancer risk. These findings were consistent with the results of individual SNPs in this block, with one SNP (rs1110478) having a statistically significant positive association and two other SNPs (rs9332377 and rs165849) showing some evidence for an association. rs1110478 was out of HW equilibrium in both controls and cases of this study and also in the SNP500 Caucasian reference panel, with more individuals homozygous for the minor allele than expected (32). This deviation from HW equilibrium could indicate a genotyping error and may increase the possibility that this association is due to chance (38). But, the genotyping assay for rs1110478 was validated against sequencing in DNA samples from >300 individuals and found to be 100% concordant, genotyping completion rates were >95% and there were no discordant pairs for the blinded quality control samples (∼10% of total samples), suggesting that these results are unlikely to be due to a genotyping error. By chance, we would expect that 5% of genotyped SNPs (two of 58 in this study) would be out of HW equilibrium. Two other SNPs (rs9332377 and rs165849) in this block showed increased but not statistically significant association with gastric cancer risk, providing some evidence that the association observed with rs1110478 is not due to a HW error.

No previous studies have examined the association of SNPs in LD block #2 of COMT and gastric cancer risk. Specific haplotypes in block #2 were previously shown to be associated with increased risk of breast cancer (31,39) as well as other chronic diseases (40). This LD block contains the 3′ UTR of COMT. Though no in vitro data are available for these SNPs, 3′ UTRs are important for the stability of many messenger RNA transcripts and thus could affect COMT expression. This genomic region also contains the 3′ UTR of the neighboring armadillo repeat deleted in the velocardiofacial syndrome (ARVCF) gene that helps maintain adherens junctions (41), another process that may be relevant to gastric cancer (42).

We did not find associations between gastric cancer risk and LD block #1 of COMT, which overlaps the gene’s coding region. SNPs within this region have been associated with increased risk of breast cancer in some studies, though the summary estimate from a recent meta-analysis does not support an association (43). Recent data from a case–control study in Japan also did not observe an association between SNPs in this block and gastric cancer risk (44).

SHBG is a plasma glycoprotein that transports androgens and estrogens while in circulation and regulates the availability of free estrogens and androgens (20). Although the global P test was not statistically significant, one haplotype in LD block #2 in SHBG (CACCC) was associated with reduced risk of gastric cancer. This result was consistent with risk estimates for individual SNPs in this block, with those carrying the minor allele of rs6259 showing a statistically significant association. Though no previous studies have examined this SNP for gastric cancer, carrying the minor allele of rs6259 has been linked to decreased endometrial (45) and breast cancer risk (46) and increased risk of prostate cancer (47) in some studies, but not in others (48–50). One study examined the association with ovarian cancer and found no association (51).

rs6259 is a putatively functional SNP (D356N) that inserts an additional N-glycosylation site (52) and has been associated with modestly increased circulating levels of SHBG in those homozygous for the minor allele (50,53,54). But, the association observed in this study was among heterozygotes, where circulating SHBG levels appear similar to those with the major allele (50,53,54). Only one case and one control in this study were homozygous for the minor allele. rs6259 does not affect hormone binding (55), but may affect other aspects including the transport of hormones into the cell (56). SHBG may also regulate steroid signaling via alternative pathways, including regulating cyclic adenosine monophosphate levels by binding to its plasma membrane bound receptor (SHBG-R) (57,58). Alternatively, rs6259 could be in high LD with another functional polymorphism in SHBG or a neighboring gene.

We found no association between LD blocks in CYP1B1, CYP17A1, CYP19A1 and HSD17B1 and gastric cancer risk. Previous studies with hormonally associated cancers, such as breast or prostate, have not demonstrated consistent associations (27–30,49,50). One recent Japanese study observed statistically significant associations between several SNPs in CYP17A1 (rs619824 and rs743572) and CYP19A1 (rs4646 and rs1902586) and gastric cancer risk (44) that were also genotyped in our study. Our results did not replicate these findings.

Strengths of this study include its population-based design and comprehensive assessment of selected LD blocks, as well as use of haplotypes to limit the number of comparisons. Limitations include the exclusion of a larger percentage of cases (36%) than controls (14%) due to lack of DNA availability. Cases with a blood sample for DNA extraction had a lower tumor stage and grade than cases without a blood sample. If the association between genotypes or other exposures differed between more and less advanced cases, then the risk estimates observed in our study may not reflect the entire distribution of gastric cancer cases. But, the distribution of genotypes did not vary by tumor stage or grade in our study. In addition, demographic and lifestyle factors did not differ between cases and controls with and without a DNA sample, arguing against selection bias as a major determinant of our results.

An additional limitation is that our results are from a single modestly sized study. As we examined genetic variation in 13 LD blocks, statistically significant findings could be due to multiple comparisons and chance. Indeed, the most statistically significant LD block had a FDR of 0.429, the most statistically significant haplotype had a FDR of 0.459 and the most statistically significant individual SNPs had a FDR of 0.583. Therefore, our findings should be interpreted with caution and confirmed in additional studies. We also had limited power to examine associations in men and women separately.

In conclusion, in the first study to comprehensively examine the association of genetic variation in six genes of the steroid hormone metabolism pathway and gastric cancer risk, we found possible associations with genetic variation in COMT, involved in estrogen inactivation, and SHBG, a modulator of sex hormone bioavailability. We found no association with haplotypes in CYP1B1, CYP17A1, CYP19A1 and HSD17B1.

Supplementary material

Supplementary Table 1 can be found at http://carcin.oxfordjournals.org/

Funding

USA National Cancer Institute (N02-CP-40501, N01-CP-05626, N02-CP-71103); National Cancer Institute Office of Women’s Health; Intramural Research Program of National Institutes of Health/National Cancer Institute, Division of Cancer Epidemiology and Genetics.

Supplementary Material

[Supplementary Data]
bgn258_index.html (739B, html)

Acknowledgments

We are grateful to the medical, pathology and nursing staff in the Warsaw hospitals for case recruitment and diagnosis and to the Survey Centre Institute of the Institute of Philosophy and Sociology at the Polish Academy of Science in Warsaw for assistance in training and coordination of interviewers.

Conflict of Interest Statement: None declared.

Glossary

Abbreviations

BPC3

Breast and Prostate Cohort Consortium

CI

confidence interval

FDR

false discovery rate

HW

Hardy–Weinberg

LD

linkage disequilibrium

NCI

National Cancer Institute

OR

odds ratio

SNP

single-nucleotide polymorphism

UTR

untranslated region

References

  • 1.Parkin DM, et al. Global cancer statistics, 2002. CA Cancer J. Clin. 2005;55:74–108. doi: 10.3322/canjclin.55.2.74. [DOI] [PubMed] [Google Scholar]
  • 2.Sipponen P, et al. Delayed rise in incidence of gastric cancer in females results in unique sex ratio (M/F) pattern: etiologic hypothesis. Gastric Cancer. 2002;5:213–219. doi: 10.1007/s101200200037. [DOI] [PubMed] [Google Scholar]
  • 3.Kominea A, et al. Androgen receptor (AR) expression is an independent unfavorable prognostic factor in gastric cancer. J. Cancer Res. Clin. Oncol. 2004;130:253–258. doi: 10.1007/s00432-003-0531-x. [DOI] [PubMed] [Google Scholar]
  • 4.Takano N, et al. Expression of estrogen receptor-alpha and -beta mRNAs in human gastric cancer. Cancer Lett. 2002;176:129–135. doi: 10.1016/s0304-3835(01)00739-x. [DOI] [PubMed] [Google Scholar]
  • 5.Freedman ND, et al. Menstrual and reproductive factors and gastric cancer risk in a large prospective study of women. Gut. 2007;56:1671–1677. doi: 10.1136/gut.2007.129411. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Frise S, et al. Menstrual and reproductive risk factors and risk for gastric adenocarcinoma in women: findings from the Canadian national enhanced cancer surveillance system. Ann. Epidemiol. 2006;16:908–916. doi: 10.1016/j.annepidem.2006.03.001. [DOI] [PubMed] [Google Scholar]
  • 7.Heuch I, et al. Menstrual and reproductive factors and risk of gastric cancer: a Norwegian cohort study. Cancer Causes Control. 2000;11:869–874. doi: 10.1023/a:1008975817061. [DOI] [PubMed] [Google Scholar]
  • 8.Inoue M, et al. Height, weight, menstrual and reproductive factors and risk of gastric cancer among Japanese postmenopausal women: analysis by subsite and histologic subtype. Int. J. Cancer. 2002;97:833–838. doi: 10.1002/ijc.10149. [DOI] [PubMed] [Google Scholar]
  • 9.Kaneko S, et al. Menstrual and reproductive factors and the mortality risk of gastric cancer in Japanese menopausal females. Cancer Causes Control. 2003;14:53–59. doi: 10.1023/a:1022596104796. [DOI] [PubMed] [Google Scholar]
  • 10.La Vecchia C, et al. Menstrual and reproductive factors and gastric-cancer risk in women. Int. J. Cancer. 1994;59:761–764. doi: 10.1002/ijc.2910590609. [DOI] [PubMed] [Google Scholar]
  • 11.Lindblad M, et al. Hormone replacement therapy and risks of oesophageal and gastric adenocarcinomas. Br. J. Cancer. 2006;94:136–141. doi: 10.1038/sj.bjc.6602906. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Palli D, et al. Reproductive history and gastric cancer among post-menopausal women. Int. J. Cancer. 1994;56:812–815. doi: 10.1002/ijc.2910560609. [DOI] [PubMed] [Google Scholar]
  • 13.Lindblad M, et al. Estrogen and risk of gastric cancer: a protective effect in a nationwide cohort study of patients with prostate cancer in Sweden. Cancer Epidemiol. Biomarkers Prev. 2004;13:2203–2207. [PubMed] [Google Scholar]
  • 14.Matsuyama Y, et al. Second cancers after adjuvant tamoxifen therapy for breast cancer in Japan. Ann. Oncol. 2000;11:1537–1543. doi: 10.1093/oxfordjournals.annonc.a010406. [DOI] [PubMed] [Google Scholar]
  • 15.Rutqvist LE, et al. Adjuvant tamoxifen therapy for early stage breast cancer and second primary malignancies. Stockholm Breast Cancer Study Group. J. Natl Cancer Inst. 1995;87:645–651. doi: 10.1093/jnci/87.9.645. [DOI] [PubMed] [Google Scholar]
  • 16.Furukawa H, et al. Effect of sex hormones on carcinogenesis in the stomachs of rats. Cancer Res. 1982;42:5181–5182. [PubMed] [Google Scholar]
  • 17.Ohtani M, et al. Protective role of 17 beta-estradiol against the development of Helicobacter pylori-induced gastric cancer in INS-GAS mice. Carcinogenesis. 2007;28:2597–2604. doi: 10.1093/carcin/bgm150. [DOI] [PubMed] [Google Scholar]
  • 18.Sharp L, et al. CYP17 gene polymorphisms: prevalence and associations with hormone levels and related factors. a HuGE review. Am. J. Epidemiol. 2004;160:729–740. doi: 10.1093/aje/kwh287. [DOI] [PubMed] [Google Scholar]
  • 19.Tsuchiya Y, et al. Cytochrome P450-mediated metabolism of estrogens and its regulation in human. Cancer Lett. 2005;227:115–124. doi: 10.1016/j.canlet.2004.10.007. [DOI] [PubMed] [Google Scholar]
  • 20.Hammond GL. Potential functions of plasma steroid-binding proteins. Trends Endocrinol. Metab. 1995;6:298–304. doi: 10.1016/1043-2760(95)00162-x. [DOI] [PubMed] [Google Scholar]
  • 21.Yue W, et al. Genotoxic metabolites of estradiol in breast: potential mechanism of estradiol induced carcinogenesis. J. Steroid Biochem. Mol. Biol. 2003;86:477–486. doi: 10.1016/s0960-0760(03)00377-7. [DOI] [PubMed] [Google Scholar]
  • 22.Zhu BT, et al. Functional role of estrogen metabolism in target cells: review and perspectives. Carcinogenesis. 1998;19:1–27. doi: 10.1093/carcin/19.1.1. [DOI] [PubMed] [Google Scholar]
  • 23.Chow WH, et al. Risk of stomach cancer in relation to consumption of cigarettes, alcohol, tea and coffee in Warsaw, Poland. Int. J. Cancer. 1999;81:871–876. doi: 10.1002/(sici)1097-0215(19990611)81:6<871::aid-ijc6>3.0.co;2-#. [DOI] [PubMed] [Google Scholar]
  • 24.Lan Q, et al. Glutathione S-transferase genotypes and stomach cancer in a population-based case-control study in Warsaw, Poland. Pharmacogenetics. 2001;11:655–661. doi: 10.1097/00008571-200111000-00003. [DOI] [PubMed] [Google Scholar]
  • 25.Hunter DJ, et al. A candidate gene approach to searching for low-penetrance breast and prostate cancer genes. Nat. Rev. Cancer. 2005;5:977–985. doi: 10.1038/nrc1754. [DOI] [PubMed] [Google Scholar]
  • 26.Excoffier L, et al. Maximum-likelihood estimation of molecular haplotype frequencies in a diploid population. Mol. Biol. Evol. 1995;12:921–927. doi: 10.1093/oxfordjournals.molbev.a040269. [DOI] [PubMed] [Google Scholar]
  • 27.Setiawan VW, et al. CYP17 genetic variation and risk of breast and prostate cancer from the National Cancer Institute Breast and Prostate Cancer Cohort Consortium (BPC3) Cancer Epidemiol. Biomarkers Prev. 2007;16:2237–2246. doi: 10.1158/1055-9965.EPI-07-0589. [DOI] [PubMed] [Google Scholar]
  • 28.Haiman CA, et al. Genetic variation at the CYP19A1 locus predicts circulating estrogen levels but not breast cancer risk in postmenopausal women. Cancer Res. 2007;67:1893–1897. doi: 10.1158/0008-5472.CAN-06-4123. [DOI] [PubMed] [Google Scholar]
  • 29.Kraft P, et al. Genetic variation in the HSD17B1 gene and risk of prostate cancer. PLoS Genet. 2005;1:e68. doi: 10.1371/journal.pgen.0010068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Feigelson HS, et al. Haplotype analysis of the HSD17B1 gene and risk of breast cancer: a comprehensive approach to multicenter analyses of prospective cohort studies. Cancer Res. 2006;66:2468–2475. doi: 10.1158/0008-5472.CAN-05-3574. [DOI] [PubMed] [Google Scholar]
  • 31.Gaudet MM, et al. Comprehensive assessment of genetic variation of catechol-O-methyltransferase and breast cancer risk. Cancer Res. 2006;66:9781–9785. doi: 10.1158/0008-5472.CAN-06-1294. [DOI] [PubMed] [Google Scholar]
  • 32.Packer BR, et al. SNP500Cancer: a public resource for sequence validation, assay development, and frequency analysis for genetic variation in candidate genes. Nucleic Acids Res. 2006;34:D617–D621. doi: 10.1093/nar/gkj151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Wigginton JE, et al. A note on exact tests of Hardy-Weinberg equilibrium. Am. J. Hum. Genet. 2005;76:887–893. doi: 10.1086/429864. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Barrett JC, et al. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21:263–265. doi: 10.1093/bioinformatics/bth457. [DOI] [PubMed] [Google Scholar]
  • 35.Lake SL, et al. Estimation and tests of haplotype-environment interaction when linkage phase is ambiguous. Hum. Hered. 2003;55:56–65. doi: 10.1159/000071811. [DOI] [PubMed] [Google Scholar]
  • 36.Schaid DJ, et al. Score tests for association between traits and haplotypes when linkage phase is ambiguous. Am. J. Hum. Genet. 2002;70:425–434. doi: 10.1086/338688. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Benjamini Y, et al. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. B. 1995;57:289–300. [Google Scholar]
  • 38.Schaid DJ, et al. Biased tests of association: comparisons of allele frequencies when departing from Hardy-Weinberg proportions. Am. J. Epidemiol. 1999;149:706–711. doi: 10.1093/oxfordjournals.aje.a009878. [DOI] [PubMed] [Google Scholar]
  • 39.Gaudet MM, et al. Catechol-O-methyltransferase haplotypes and breast cancer among women on Long Island, New York. Breast Cancer Res. Treat. 2006;99:235–240. doi: 10.1007/s10549-006-9205-0. [DOI] [PubMed] [Google Scholar]
  • 40.Bray NJ, et al. A haplotype implicated in schizophrenia susceptibility is associated with reduced COMT expression in human brain. Am. J. Hum. Genet. 2003;73:152–161. doi: 10.1086/376578. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Hatzfeld M. The p120 family of cell adhesion molecules. Eur. J. Cell Biol. 2005;84:205–214. doi: 10.1016/j.ejcb.2004.12.016. [DOI] [PubMed] [Google Scholar]
  • 42.Conlin VS, et al. Helicobacter pylori infection targets adherens junction regulatory proteins and results in increased rates of migration in human gastric epithelial cells. Infect. Immun. 2004;72:5181–5192. doi: 10.1128/IAI.72.9.5181-5192.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Wen W, et al. Cytochrome P450 1B1 and catechol-O-methyltransferase genetic polymorphisms and breast cancer risk in Chinese women: results from the Shanghai breast cancer study and a meta-analysis. Cancer Epidemiol. Biomarkers Prev. 2005;14:329–335. doi: 10.1158/1055-9965.EPI-04-0392. [DOI] [PubMed] [Google Scholar]
  • 44.Ikeda S, et al. Screening of 214 single nucleotide polymorphisms in 44 candidate cancer susceptibility genes: a case-control study on gastric and colorectal cancers in the Japanese population. Am. J. Gastroenterol. 2008;103:1476–1487. doi: 10.1111/j.1572-0241.2008.01810.x. [DOI] [PubMed] [Google Scholar]
  • 45.Kataoka N, et al. Association of endometrial cancer risk with a functional polymorphism (Asp(327)Asn) in the sex hormone-binding globulin gene. Cancer. 2007;109:1296–1302. doi: 10.1002/cncr.22531. [DOI] [PubMed] [Google Scholar]
  • 46.Cui Y, et al. Association of breast cancer risk with a common functional polymorphism (Asp327Asn) in the sex hormone-binding globulin gene. Cancer Epidemiol. Biomarkers Prev. 2005;14:1096–1101. doi: 10.1158/1055-9965.EPI-04-0721. [DOI] [PubMed] [Google Scholar]
  • 47.Berndt SI, et al. Variant in sex hormone-binding globulin gene and the risk of prostate cancer. Cancer Epidemiol. Biomarkers Prev. 2007;16:165–168. doi: 10.1158/1055-9965.EPI-06-0689. [DOI] [PubMed] [Google Scholar]
  • 48.Becchis M, et al. The additionally glycosylated variant of human sex hormone-binding globulin (SHBG) is linked to estrogen-dependence of breast cancer. Breast Cancer Res. Treat. 1999;54:101–107. doi: 10.1023/a:1006100929670. [DOI] [PubMed] [Google Scholar]
  • 49.Cunningham JM, et al. Evaluation of genetic variations in the androgen and estrogen metabolic pathways as risk factors for sporadic and familial prostate cancer. Cancer Epidemiol. Biomarkers Prev. 2007;16:969–978. doi: 10.1158/1055-9965.EPI-06-0767. [DOI] [PubMed] [Google Scholar]
  • 50.Dunning AM, et al. Polymorphisms associated with circulating sex hormone levels in postmenopausal women. J. Natl Cancer Inst. 2004;96:936–945. doi: 10.1093/jnci/djh167. [DOI] [PubMed] [Google Scholar]
  • 51.Garcia-Closas M, et al. Ovarian cancer risk and common variation in the sex hormone-binding globulin gene: a population-based case-control study. BMC Cancer. 2007;7:60. doi: 10.1186/1471-2407-7-60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Power SG, et al. Molecular analyses of a human sex hormone-binding globulin variant: evidence for an additional carbohydrate chain. J. Clin. Endocrinol. Metab. 1992;75:1066–1070. doi: 10.1210/jcem.75.4.1400872. [DOI] [PubMed] [Google Scholar]
  • 53.Cousin P, et al. Human variant sex hormone-binding globulin (SHBG) with an additional carbohydrate chain has a reduced clearance rate in rabbit. J. Clin. Endocrinol. Metab. 1998;83:235–240. doi: 10.1210/jcem.83.1.4515. [DOI] [PubMed] [Google Scholar]
  • 54.Haiman CA, et al. Common genetic variation in the sex steroid hormone-binding globulin (SHBG) gene and circulating shbg levels among postmenopausal women: the Multiethnic Cohort. J. Clin. Endocrinol. Metab. 2005;90:2198–2204. doi: 10.1210/jc.2004-1417. [DOI] [PubMed] [Google Scholar]
  • 55.Petra PH, et al. Complete enzymatic deglycosylation of native sex steroid-binding protein (SBP or SHBG) of human and rabbit plasma: effect on the steroid-binding activity. Protein Sci. 1992;1:902–909. doi: 10.1002/pro.5560010708. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Hammes A, et al. Role of endocytosis in cellular uptake of sex steroids. Cell. 2005;122:751–762. doi: 10.1016/j.cell.2005.06.032. [DOI] [PubMed] [Google Scholar]
  • 57.Rosner W, et al. Sex hormone-binding globulin mediates steroid hormone signal transduction at the plasma membrane. J. Steroid Biochem. Mol. Biol. 1999;69:481–485. doi: 10.1016/s0960-0760(99)00070-9. [DOI] [PubMed] [Google Scholar]
  • 58.Tadokoro T, et al. The inhibitory effect of androgen and sex-hormone-binding globulin on the intracellular cAMP level and tyrosinase activity of normal human melanocytes. Pigment Cell Res. 2003;16:190–197. doi: 10.1034/j.1600-0749.2003.00019.x. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

[Supplementary Data]
bgn258_index.html (739B, html)
bgn258_carcin-2008-00561-File002_bgn258.doc (531.5KB, doc)

Articles from Carcinogenesis are provided here courtesy of Oxford University Press

RESOURCES