Endometriosis is a common disease affecting up to 15% of women of reproductive age and is a known cause of infertility and chronic pelvic pain.1,2 Endometriosis is a hormone-dependent disease and a critical role of proestrogenic and antiestrogenic hormonal stimuli in the biology and clinical course of this disease has been established.3 Estrogen and progesterone receptors are expressed in endometriotic lesions and the biology of endometriosis is driven by estrogen exposure. Therefore, antiestrogenic hormonal treatment with danazol, progestins, and gonadotropin-releasing hormone (GnRH) analogues is widely used to treat endometriosis-associated symptoms.4,5 In addition, aromatase inhibitors have been proposed as novel treatment option for women with endometriosis.6
Various factors contribute to the susceptibility to and development of endometriosis, including hormonal and immunologic factors as well as genetic factors. Endometriosis has a familial component and it has been suggested that endometriosis has a genetic basis. For example, the incidence of endometriosis is significantly increased in first-degree relatives compared with controls (6.9% compared with 0.9%, respectively).4 Concordance in twins has also been demonstrated.7 Individual genetic variation has also been advocated as a significant contributor to the biology of endometriosis. A number of single nucleotide polymorphisms (SNPs) have been found to be associated with the clinical course of and the susceptibility to the disease. For example, our group found a 306–base pair insertion polymorphism in intron G of the progesterone receptor to be associated with endometriosis8 and the IL-6 promoter polymorphism -174 G/C to be associated with ovarian endometriosis cysts.9 Other groups recently described the vascular endothelial growth factor +405 C/G polymorphism,10 the Interleukin-2 receptor β (IL-2Rbeta)–627*C polymorphism,11 and the AhRR codon 185 polymorphism to be associated with susceptibility to and severity of endometriosis. In view of the variability of genetic influences, a polygenetic model of endometriosis seems more likely to explain these data compared with dominant or recessive single gene models.12
The susceptibility to endometriosis is also strongly influenced by ecogenetic interactions. For example, countries with a high exposure to organic pollutants such as dioxins and polychlorinated biphenyls such as Belgium have a high prevalence of endometriosis.13 Also, some have found that immunocompetence against exotoxins is characteristically impaired among women with endometriosis.14 It is in accordance with these data that Baranova et al15 detected an association between endometriosis and polymorphic variations of enzymes involved in phase II exotoxin metabolization, including the N-acetyltransferase 2 gene and glutathione S-transferase M.
We previously developed a microarray system allowing for large-scale genoptyping of SNPs with multiplex polymerase chain reaction (PCR).16 This technique was used for the parallel analysis of 10 SNPs within 7 human genes involved in the estrogen metabolism, ie, catechol-O-methyltransferase (COMT) Val158Met G->A, 17-β-hydroxysteroid dehydrogenase type 1 (HSD17) vlV A->C, cytochrome P450 (CYP) 17 A2 allele T->C, CYP1A1–1 MspI RFLP T->C, CYP1A1–2 Ile462Val A->G, CYP19–2 Arg264Cys C->T, CYP19–3 C1558T C->T, CYP 1B1 Leu432Val, CYP1B1 Asn453Ser, and estrogen receptor (ER) alpha IVS1 –401>C.
The above listed SNPs were selected because the respective gene products are involved in the estrogen metabolism and can thus be seen as candidate genes for endometriosis. For example, variant COMT, CYP17, and CYP19 enzymes have been noted to be associated with significant changes in serum hormone concentrations.17–19 Variations of the ER alpha gene leads to significant alterations in the response to therapy with estrogen.21 HSD17 catalyzes the final step of estradiol (E2) biosynthesis, ie, the conversion of estrone to E2. The functional property of the HSD17 vlV A->C SNP is unclear, but presence of this SNP has been found to be associated with elevated E2 levels and breast cancer, an estrogen-dependent disease.21,22 We performed a case–control study estimating the association of 10 estrogen metabolizing gene SNPs and endometriosis in a white population of middle-European origin. In this study, we attempted to establish a multiple genetic model based on gene–gene interactions to define potentially critical SNP combinations in the susceptibility for endometriosis.
MATERIALS AND METHODS
Approval for this study was obtained by the Institutional Review Board at the Medical University of Vienna, Vienna, Austria. Thirty-two women with surgically and histologically proven endometriosis were included in our study. All patients were of white origin and consented to participating in the study. As controls, we enrolled healthy white women from the same geographical area without endometriosis in their personal histories based on personal interview (n = 790). All of these women had no clinical signs or symptoms of endometriosis, such as dysmenorrhea, chronic pelvic pain, or infertility. Women with menstrual pain requiring pain medication were not included in the control group. All women were recruited consecutively from the outpatient service of the Department of Obstetrics and Gynecology, Medical University of Vienna and from private gynecologic offices in the area between June 2001 and December 2004.
We previously developed a microarray system combining liquid phase amplification of genomic DNA with allele-specific solid-phase PCR into a 1-step reaction on a microarray platform allowing for large-scale genoptyping of SNPs.16 This integrated genotyping strategy was subsequently extended by a multiplex PCR reaction, enabling the parallel analysis of 10 polymorphisms within 7 human genes involved in the estrogen metabolism: COMT Val158Met G->A, HSD17 vlV A->C, CYP 17 A2 allele T->C, CYP1A1 MspI RFLP T->C, CYP1A1 Ile462Val A->G, CYP19 Arg264Cys C->T, CYP19 C1558T C->T, CYP 1B1 Leu432Val, CYP1B1 Asn453Ser, and ER alpha IVS1 –401>C. In brief, the procedure was as follows:
DNA was extracted from patients' blood or buccal swabs using the QIAamp DNA Blood Midi Kit (Qiagen, Hilden, Germany) protocol. All oligonucleotides were synthesized in house with an Expedite 8909 Nucleic Acid Synthesizer (PerSeptive Biosystems, Foster City, CA) using standard phosphoramidite chemistry. Oligonucleotides for microarray attachment were synthesized with a 5′ terminal (CH2)6-NH2 modification (Cruachem Ltd, Glasgow, UK) and purified by perfusion chromatography on a BioCAD Sprint system (PerSeptive Biosystems).
Amplification of 10 fragments of the genomic region containing the SNP was performed in a multiplex PCR assay in a 25-μL volume containing 100 ng genomic DNA, 1.5 x HotStar TaqPCR buffer, 75 μM each dNTP, 0.75 u HotStar TaqDNA Polymerase (Qiagen, Hilden, Germany), and 0.1–0.35 μM of each multiplex PCR primer. Reactions were carried out in a Primus 96 thermal cycler (MWG Biotech, Ebersberg, Germany) according to the following scheme: 95°C for 15 minutes, 30 cycles at 95°C for 45 seconds, 55°C for 30 seconds, 72°C for 40 seconds, and 72°C for 3 minutes.
After purification of the PCR products using the QIA PCR purification kit (Qiagen), the generation of single strands was combined with the incorporation of a fluorescent dye in a primer extension reaction. The 25-μL primer extension mix contained 10 μL (equal to 1/3) of purified PCR product, 1 x Vent (exo-) polymerase buffer, 1.5 units Vent (exo-) polymerase (New England Biolabs, Beverly, MA), 0.25 μM fragment specific nested primers, 12.5 μM each dNTP and 5 μM Cy3-dCTP (Amersham Pharmacia Biotech Europe, Freiburg, Germany). Primer extension was cycled as follows: 95°C for 3 minutes, 25 cycles at 95°C for 20 seconds, 60°C for 20 seconds, 72°C for 20 seconds, and 72°C for 3 minutes.
The fluorescently labeled single strands were hybridized to an allele-specific oligonucleotide array. The 15-mer ASO probes carried a 5′ T-tail23 and contained the mutant nucleotide in the middle, at position 8.24 Four microliters of the primer extension reaction was mixed with 1 μL 5×hybridization buffer (20x sodium sulfate citrate [SSC], 0.5% sarcosyl) denatured for 3 minutes at 98°C and hybridized to the ASO array under a glass cover slip for 45 minutes at 55°C in a hybridization chamber (Telechem International, Sunnyvale, CA). Slides were then washed in 2×SSC, 0.1% sodium dodecyl sulfate [SDS] for 1 minute, followed by a second and third washing step in 0.2×SSC for 1 minute at ambient temperature and a final stringent washing step in 0.1×SSC for 15 minutes at 55°C.
The slides were scanned using an Affymetrix GMS 418 Scanner (Affymetrix, Santa Clara, CA), and median fluorescent intensities were calculated using the GenePix 3.0 software (Axon Instruments, Foster City, CA). The genotype was deduced from comparison of hybridization rates to the allele-specific oligonucleotides.
For the preparation of the ASO arrays, oligonucleotides with a 5′ amino modification were printed on 3D-Link slides (Motorola Life Sciences, Northbrook, IL) using an Affymetrix 417 Arrayer (Affymetrix, Santa Clara, CA) according to the manufacturer's instructions.
Preparation of the slides and covalent coupling of the solid-phase PCR primers was done as described previously.25 Briefly, glass slides were treated with a trimethoxysilane derivatized polyethylenimine and subsequently activated using the amine reactive cross-linker EGS. The amine modified PCR primers were spotted to the activated glass slides using an Affymetrix GMS 417 Arrayer. Spotted glass slides can be stored at least 3 months until used in on-chip PCR.
A 10-μL PCR reaction mix containing 2×HotStar TaqPCR-buffer, 50 μM each dNTP, 20 μM Cy3-dCTP (Amersham Pharmacia Biotech Europe), 2.5 μg/μL BSA, 25 % (volume to volume) Self-Seal Reagent (MJ Research, Waltham, MA), 3 units HotStar TaqDNA Polymerase (Qiagen) and the multiplex PCR primers at 0.025–0.4 μM was prepared. Before starting thermal cycling, human genomic DNA (150 ng) in sterile water was added. After pipetting the reaction mix onto the oligonucleotide array, a glass cover slip (22 × 22 mm) was mounted to seal the reaction. Glass slides were put into a PTC 200 In Situ slide thermocycler (MJ Research), and cycling was carried out according to the following scheme: 80°C for 10 minutes, 95°C for 5 minutes, 50 cycles at 95°C for 30 seconds, and 64°C for 1 minute. After cycling, the slides were placed in 0.1×SSC, 0.1 % SDS for 10 minutes with gentle agitation. After removing the cover slips the slides were washed again for 10 minutes in SSC, then washed with deionized water and dried under an air stream.
Incorporation of Cy3-dCTP during the DNA amplification reaction was measured using the ScanArray 4000 Scanner (Packard BioScience, Meriden, CT). Fluorescence intensities (medians after local background subtraction) were calculated using the Genepix 3.0 software (Axon Instruments). Automated data analysis and interpretation of the genotype was performed using a scripting extension of the Genepix software programmed in Visual Basic Script Version 5.5 (Microsoft Corporation, Redmond, WA). The script automatically calculates ratios from median values extracted from the program's results tab and assigns genotype information to the ratios according to the following scheme of predefined thresholds: 3-fold and above: wild-type, 0.33-fold and below: mutant, lower than 1.5-fold but higher than 0.67-fold: heterozygous. Ratio values not matching to this scheme were flagged for visual inspection by the user.
Throughout statistical analysis, SNPs were considered binary variables using a dominant gene model, ie, homozygous wild-type compared with heterozygous and homozygous mutant (wild type to wild type, compared with wild type to mutant and mutant to mutant). This model was used for methodologic purposes and is not meant to classify the SNPs investigated in this study. The association of SNPs with the presence of endometriosis was evaluated by univariate logistic regression or χ2 tests and expressed as odds ratios (OR) and 95% confidence intervals (CI).
To correct for multiple comparisons, stepwise logistic regression was used to find variables that independently discriminate cases, ie, endometriosis, from controls. All 10 SNPs were used as candidate variables. In a second stepwise logistic regression analysis, we additionally considered all 2-way interactions among all candidate variables. Based on the genotype distributions, we performed a power analysis. With the given sample size of 765 probands, we achieved a power between 7.9% and 91.9% for the 10 SNPs (COMT 30.6%; HSD17 91.9%, CYP17 32.1%, CYP1A1 11.0%, CYP1A1 30.6%, CYP19 51.8%, CYP19 7.9%, CYP1B1 16.4%, CYP1B1 7.9%, and ER alpha 55.0%) with an α of .05. The Mann-Whitney U test was used to compare continuous variables. P values less than .05 were considered statistically significant. We used the statistical software SAS System for Windows 8.2 (SAS Institute Inc., Cary, NC) for statistical analysis.
Table 1 shows the clinical characteristics of enrolled women. Based on chart review and surgery reports, the 32 enrolled women had endometriosis defined as revised American Fertility Society score26 I, II, III, and IV in 0, 21, 10, and 1 cases, respectively. Table 2 shows genotype frequencies of all investigated SNPs in patients with endometriosis and controls. A univariate analysis demonstrated that the presence of the mutant HSD17 vlV C allele was significantly associated with endometriosis (P = .004; OR 3.9, 95% CI 1.6–9.8). Among women with endometriosis, carriage of the mutant allele was found in 7 of 32 (28%) cases compared with 52 of 726 (7%) of controls. No association was ascertained between the other investigated SNPs and the presence or absence of endometriosis.
Using a stepwise logistic regression analysis, we ascertained that the presence of a mutant HSD17 allele remained to be an independent predictor for endometriosis (P = .002; OR 2.8, 95% CI 1.8–8.9; regression coefficient [standard error] 0.200 [0.053]). Using a second stepwise logistic regression analysis testing for all 2-way interactions among SNPs considered as candidate variables, no significant interactions between any SNP combination and the presence of endometriosis was found. Thus, a multiple model based on gene–gene interactions could not be established.
Genetic testing for SNPs is beginning to be implemented into clinical practice in some medical subspecialities such as hematology and clinical pharmacology. With respect to endometriosis, a considerable number of studies on SNPs of candidate genes focusing on various pathways of the biology of endometriosis have been published, among them COMT, intron G of the progesterone receptor, IL-6, vascular endothelial growth factor, IL-2Rbeta, and AhRR.8–27 Most studies, however, investigated a single SNP and thus did not allow for the calculation of interaction models or correction of multiple comparisons.
Additive and multiplicatory effects of the simultaneous carriage of multiple SNPs have been described.28,29 Identifying critical combinations of SNPs conferring additive or multiplicatory risks of developing endometriosis would be of clinical interest for defining high-risk populations or for identifying women most likely to benefit from pharmacologic interventions related to the gene product of the respective SNPs, eg, antiestrogens. Thus, we attempted to establish such a combination by simultaneously investigating 10 SNPs involved in the estrogen metabolism. Our study demonstrates that the presence of the HSD17 vlV A->C SNP is associated with susceptibility to endometriosis, but no interactions between the investigated SNPs were detected in this population.
In our case–control study, we present an analysis of 10 SNPs of estrogen metabolizing genes in a population of white women investigating over 7,000 genotypes in patients with endometriosis and controls. Based on our study, we support previously published data that individual genetic variation based on the presence of some SNPs modulated the susceptibility to endometriosis. In particular, we found that genetic variation of the HSD17 gene may play a significant role in this respect. The enzyme HSD17 is a gate keeper of E2 production, and the HSD17 vlV A->C SNP is associated with elevated E2 levels and breast cancer, an estrogen-dependent disease.21,22 Thus, it seems plausible that the presence of this SNP promotes the development of endometriosis. On the other hand, the results of our study also indicate that not every genetic alteration providing a proestrogenic environment is necessarily associated with endometriosis. This may be explained in different ways. For example, the other investigated enzymes and their respective genetic variants not found to be associated with endometriosis may have only minimal effects on endometriotic cell growth not detectable by this study. Also, some enzymes involved in the estrogen metabolism may play no role in the cell growth of endometriosis due to an altered microenvironment in endometriotic lesions. Moreover, our panel of SNPs is limited to the estrogen metabolism and is therefore not suitable to assess the influence of progestin biology on the cause of endometriosis. As demonstrated by Osteen et al,30 changes in progestin biology play a critical role in the development of ectopic lesions, and selective progesterone receptor modulators are currently under investigation as pharmacotherapeutics for endometriosis.31
Our results are at odds with a previous report from Taiwan on the association between the CYP17 A2 allele T->C SNP and endometriosis.32 This difference underlines the importance of the ethnic background in SNP association studies, a phenomenon well known in this field of research. Thus, differences in genotype frequencies as well as genotype–phenotype correlations must be acknowledged when comparing the results of various association studies.
A limitation of our study is its low power. Therefore, the results of our study have to be interpreted with caution. Although we can rule out major additive or multiplicatory effects of the investigated SNPs, larger studies may be able to detect subtle differences or revised American Fertility Society stage–specific associations. Also, we cannot rule out that some women in the control group had endometriosis in view of data in the literature that endometriosis may be present in up to 15% of women.1 A strength of our study is the large control group, reducing problems associated with selection bias. Also, care has been taken to include only women of white origin, because the allele and genotype frequencies of SNPs strongly depend on ethnic background.
Our data demonstrate an association between a SNP of the HSD17 gene and endometriosis, indicating that the HSD17 vlV A->C polymorphism is a candidate genetic susceptibility marker of this disease. In our study, however, we were not able to establish a multiple genetic model of the susceptibility for endometriosis.
1. Olive DL Schwartz LB. Endometriosis. N Engl J Med 1993;328:1759–69.
2. Berkley KJ, Rapkin AJ, Papka RE. The pains of endometriosis. Science 2005;308:1587–9.
3. Bergqvist IA. Hormonal regulation of endometriosis and the rationales and effects of gonadotrophin-releasing hormone agonist treatment: a review. Hum Reprod 1995;10:446–52.
4. Lessey BA. Medical management of endometriosis and infertility. Fertil Steril 2000;73:1089–96.
5. Emmanuel KR, Davis C. Outcomes and treatment options in rectovaginal endometriosis. Curr Opin Obstet Gynecol 2005;17:399–402.
6. Amsterdam LL, Gentry W, Jobanputra S, Wolf M, Rubin SD, Bulun SE. Anastrazole and oral contraceptives: a novel treatment for endometriosis. Fertil Steril 2005;84:300–4.
7. Bischoff FZ, Simpson JL. Heritability and molecular genetic studies of endometriosis. Hum Reprod Update 2000;6:37–44.
8. Wieser F, Schneeberger C, Tong D, Tempfer C, Huber JC, Wenzl R. PROGINS receptor gene polymorphism is associated with endometriosis. Fertil Steril 2002;77:309–12.
9. Wieser F, Fabjani G, Tempfer C, Schneeberger C, Sator M, Huber J, et al. Analysis of an interleukin-6 gene promoter polymorphism in women with endometriosis by pyrosequencing. J Soc Gynecol Investig 2003;10:32–6.
10. Hoon Kim S, Min Choi Y, Ha Choung S, Kwan Jun J, Gu Kim J, Yong Moon S. Vascular endothelial growth factor gene +405 C/G polymorphism is associated with susceptibility to advanced stage endometriosis. Hum Reprod. 2005 Jun 24; [Epub ahead of print]
11. Hsieh YY, Chang CC, Tsai FJ, Hsu CM, Lin CC, Tsai CH. Interleukin-2 receptor beta (IL-2Rbeta)-627*C homozygote but not IL-12Rbeta1 codon 378 or IL-18 105 polymorphism is associated with higher susceptibility to endometriosis. Fertil Steril 2005;84:510–2.
12. Wenzl R, Kiesel L, Huber JC, Wieser F. Endometriosis: a genetic disease. Drugs Today (Barc) 2003;39 (12):961–72.
13. Fierens S, Mairesse H, Heilier JF, De Burbure C, Focant JF, Eppe G, et al. Dioxin/polychlorinated biphenyl body burden, diabetes and endometriosis: findings in a population-based study in Belgium. Biomarkers 2003;8 (6):529–34.
14. Wieser F, Dogan S, Klingel K, Diedrich K, Taylor RN, Hornung D. Expression and regulation of CCR1 in peritoneal macrophages from women with and without endometriosis. Fertil Steril 2005;83:1878–81.
15. Baranova H, Canis M, Ivaschenko T, Albuisson E, Bothorishvilli R, Baranov V, et al. Possible involvement of arylamine N-acetyltransferase 2, glutathione S-transferases M1 and T1 genes in the development of endometriosis. Mol Hum Reprod 1999;5:636–41.
16. Huber M, Mundlein A, Dornstauder E, Schneeberger C, Tempfer CB, Mueller MW, et al. Accessing single nucleotide polymorphisms in genomic DNA by direct multiplex polymerase chain reaction amplification on oligonucleotide microarrays. Anal Biochem 2002;303:25–33.
17. Worda C, Sator MO, Schneeberger C, Jantschev T, Ferlitsch K, Huber JC. Influence of the catechol-O-methyltransferase (COMT) codon 158 polymorphism on estrogen levels in women. Hum Reprod 2003;18:262–6.
18. Small CM, Marcus M, Sherman SL, Sullivan AK, Manatunga AK, Feigelson HS. CYP17 genotype predicts serum hormone levels among pre-menopausal women. Hum Reprod 2005;20:2162–7.
19. Tworoger SS, Chubak J, Aiello EJ, Ulrich CM, Atkinson C, PotterJD,et al. Association of CYP17, CYP19, CYP1B1, and COMT polymorphisms with serum and urinary sex hormone concentrations in postmenopausal women. Cancer Epidemiol Biomarkers Prev 2004;13:94–101.
20. Herrington DM, Howard TD, Hawkins GA, Reboussin DM, Xu J, Zheng SL, et al. Estrogen-receptor polymorphisms and effects of estrogen replacement on high-density lipoprotein cholesterol in women with coronary disease. N Engl J Med 2002;346 (13):967–74.
21. Wu AH, Seow A, Arakawa K, Van Den Berg D, Lee HP, Yu MC. HSD17B1 and CYP17 polymorphisms and breast cancer risk among Chinese women in Singapore. Int J Cancer 2003;104:450–7.
22. Setiawan VW, Hankinson SE, Colditz GA, Hunter DJ, De Vivo I. HSD17B1 gene polymorphisms and risk of endometrial and breast cancer. Cancer Epidemiol Biomarkers Prev 2004;13:213–9.
23. Guo Z, Guilfoyle RA, Thiel A, Wang R, Smith LM. Direct fluorescence analysis of genetic polymorphisms by hybridization with oligonucleotide arrays on glass supports. Nucleic Acids Res 1994;22:5456–65.
24. Pease AC, Solas D, Sullivan EJ, Cronin MT, Holmes CP, Fodor SP. Light-generated oligonucleotide arrays for rapid DNA sequence analysis. Proc. Natl. Acad. Sci 1994;91:5022–6.
25. Huber M, Losert D, Hiller R, Harwanegg C, Mueller MW, Schmidt WM. Detection of single base alterations in genomic DNA by solid phase polymerase chain reaction on oligonucleotide microarrays. Anal Biochem 2001;299 (1):24–30.
26. Revised American Fertility Society classification of endometriosis: 1985. Fertil Steril 1985;43:351-2.
27. Tsuchiya M, Katoh T, Motoyama H, Sasaki H, Tsugane S, Ikenoue T. Analysis of the AhR, ARNT, and AhRR gene polymorphisms: genetic contribution to endometriosis susceptibility and severity. Fertil Steril 2005;84:454–8.
28. Ribalta J, Halkes CJ, Salazar J, Masana L, Cabezas MC. Additive effects of the PPARgamma, APOE, and FABP-2 genes in increasing daylong triglycerides of normolipidemic women to concentrations comparable to those in men. Clin Chem 2005;51:864–71.
29. Gerhardt A, Scharf RE, Beckmann MW, Struve S, Bender HG, Pillny M, et al. Prothrombin and factor V mutations in women with a history of thrombosis during pregnancy and the puerperium. N Engl J Med 2000;342:374–80.
30. Osteen KG, Bruner-Tran KL, Ong D, Eisenberg E. Paracrine mediators of endometrial matrix metalloproteinase expression: potential targets for progestin-based treatment of endometriosis. Ann N Y Acad Sci 2002;955:139–46.
31. Chwalisz K, Perez MC, Demanno D, Winkel C, Schubert G, Elger W. Selective progesterone receptor modulator development and use in the treatment of leiomyomata and endometriosis [Published erratum appears in Endocr Rev 2005;26:703]. Endocr Rev 2005;26 (3):423–38.
© 2005 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
32. Hsieh YY, Chang CC, Tsai FJ, Lin CC, Tsai CH. Estrogen receptor alpha dinucleotide repeat and cytochrome P450c17alpha gene polymorphisms are associated with susceptibility to endometriosis. Fertil Steril 2005;83:567–72.