- Email: [email protected]
Microsomal epoxide hydrolase polymorphism and susceptibility to ovarian cancer
Cancer Letters 177 (2002) 75–81 www.elsevier.com/locate/canlet
Microsomal epoxide hydrolase polymorphism and susceptibility to ovarian cancer S.W. Baxter, D.Y.H. Choong, I.G. Campbell* VBCRC Cancer Genetics Laboratory, Peter MacCallum Cancer Institute, Locked Bag No. 1 A’Beckett Street, Melbourne, Victoria, 8006, Australia Received 6 July 2001; received in revised form 8 September 2001; accepted 14 September 2001
Abstract Polymorphic variants of microsomal epoxide hydrolase (mEPHX) with altered enzyme activity have been associated with an increased risk for ovarian cancer. We assessed the frequency of exon 3 and exon 4 variants of mEPHX among 291 ovarian cancers and 257 controls from a UK-based population. The distribution of the exon 3 alleles among both the cancer and control groups was significantly different from that expected under Hardy–Weinberg equilibrium suggesting that the PCR restriction fragment length polymorphism (PCR-RFLP) genotyping assay might be flawed. The codon 113 polymorphism was reassessed using a two-color allele-specific PCR-based assay. We found that a codon 119 G . A polymorphism, present in 20% of the British population and linked to the wild-type exon 3 allele, resulted in some Tyr113/His113 heterozygotes being falsely classified as His113/His113 homozygotes when using the PCR-RFLP assay. Consequently, we reassessed all our codon 113 data using the new allele-specific assay. We found no evidence of an association of ovarian cancer risk with the exon 3 Tyr113 . His113 variant. Similarly the frequencies of the exon 4 His139 . Arg139 genotypes were not significantly different between cases and controls. Stratifying the genotyping data according to the predicted mEPHX activity revealed a highly significant decrease in high mEPHX activity among the serous ovarian cancers (P ¼ 0:01) suggesting that high mEPHX activity may be protective for this histological sub-type. Furthermore previous disease association studies of exon 3 alleles which utilized the PCR-RFLP assay may be compromised by the existence of a codon 119 G . A polymorphism which may be common in Caucasian populations. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Ovarian cancer; Microsomal epoxide hydrolase; Cancer predisposition; Polymorphism
1. Introduction It is likely that inherited genetic factors contribute to the development of ovarian cancer since familial clustering of this disease is well documented. Polymorphic variants in genes involved in detoxifying endogenous or exogenous genotoxic chemicals may * Corresponding author. Tel.: 161-3-96561803; fax: 161-396561411. E-mail address: [email protected] (I.G. Campbell).
represent an important class of low penetrance disease predisposing alleles. Microsomal epoxide hydrolase (mEPHX) is involved in the activation and detoxification of exogenous chemicals and may also be involved in steroidogenesis [1,2]. Two variants of mEPHX with different enzyme activity have been described [3]. A tyrosine (Tyr113) to histidine (His113) substitution at codon 113 in exon 3 decreases enzyme activity by approximately 40% while a histidine (His139) to arginine (Arg139) substitution at codon 139 in exon 4 increases enzyme activity by more than 25%. The
0304-3835/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0304-383 5(01)00782-0
76
S.W. Baxter et al. / Cancer Letters 177 (2002) 75–81
His113 slow activity genotype has been associated with an increased risk of colon cancer [4] and hepatocellular cancer [5], whereas the high activity Tyr113 genotype has been associated with an increase in risk for smoking-related lung [6,7] and upper aero-digestive tract cancers [8]. The high activity Tyr113 genotype has also been shown to increase ovarian cancer risk in a small study of 75 ovarian cancers and 73 controls [9] although Spurdle et al. [10] did not observe any association in a much larger study. We report here an investigation of the frequency of both the exon 3 and exon 4 mEPHX polymorphic variants in a retrospective case control study of 291 epithelial ovarian cancers and 257 controls. 2. Materials and methods 2.1. Subjects Details of cases and controls have been described previously [11,12]. Briefly, incident cases of ovarian tumors were ascertained from women undergoing primary surgery in hospitals from southern England between 1993 and 1998. A specialist gynecological pathologist confirmed the histological diagnosis for each tumor and only malignant epithelial ovarian tumors (n ¼ 291) were included in the study. The histological sub-types of the tumors included in the study were serous (127), mucinous (40), endometrioid (78), clear cell (12) and undifferentiated (34). The controls consisted of 257 white female volunteers who were either staff at the Princess Anne Hospital, Southampton, UK or patients attending for nonneoplastic disease conditions. Epidemiological data such as reproductive factors, oral contraceptive use, smoking and obesity were not available for either the cases or controls. However, both groups were resident in and around the Southampton metropolitan area, which is predominantly an Anglo-Saxon population. The age range of the ovarian cancer patients was 23– 90 with a mean of 62 years and for the controls the age range was 20–78 with a mean of 42 years. 2.2. mEPHX genotyping The exon 3 and exon 4 variants were detected using the PCR restriction fragment length polymorphism (RFLP) assay described by Smith and Harrison [13].
Briefly, the codon 113 T . C variant (Tyr113 to His113) was detected using a FAM-labeled forward primer (5 0 -GATCGATAAGTTCCGTTTCACC-3 0 ) and a reverse primer (5 0 -ATCCTTAGTCTTGAAGTGAGGaT-3 0 ) which included an engineered base change of an A for a G (shown in lower case) which produces an EcoRV restriction enzyme site in the Tyr113 allele. The codon 139 A . G variant (His139 . Arg139) produces an RsaI restriction enzyme site and was detected using the forward primer 5 0 -ACATCCACTTCATCCACGT-3 0 and a HEXlabeled reverse primer 5 0 -ATGCCTCTGAGAAGCCAT-3 0 . PCRs were performed in a reaction volume of 10 ml containing 10–200 ng of genomic DNA, primers (50 ng of each), 1 £ reaction buffer (Applied Biotechnologies, UK), 200 nM of dATP, dTTP, dGTP, and dCTP (Promega, UK) and 0.2 units of Taq DNA polymerase (Red Hot Taq, Applied Biotechnologies, UK). PCR conditions consisted of an initial denaturation at 948C for 5 min followed by 35 cycles of 948C for 30 s, 558C for 30 s, and 728C for 60 s followed by one cycle of 728C for 5 min. PCRs and restriction digest were carried out as multiplex reactions and the products were separated on 2.5% agarose gels. A scanning laser fluorescence imager (Bio-Rad Molecular Imager FX) was used to discriminate the exon 3 and exon 4 alleles using 488 and 532 nm lasers, respectively. A substitution polymorphism (G to A) at codon 119 has been reported [14]. This polymorphism may affect the accuracy of the exon 3 genotyping and to circumvent this potential problem we developed a dual-colour allele-specific assay for the codon 113 polymorphism. We amplified each allele using a FAM-labeled forward primer (5 0 -CAGGTGGAGATTCTCAACAGAC-3 0 ) specific for the codon 113 ‘C’ allele and a HEX-labeled forward primer (5 0 CAGGTGGAGATTCTCAACAGAT-3 0 ) specific for the ‘T’ allele and a common unlabeled reverse primer (5 0 -CCTGCCTAGCTCTAAAGATGGA-3 0 ). The frequency of the codon 119 polymorphism was assessed using a PCR-RFLP assay based on the fact that the G . A substitution creates a Tsp509 I restriction enzyme recognition site. A 220 bp fragment encompassing codons 113 and 119 was amplified using the forward primer 5 0 -TGGAAGAAGCAGGTGGAGAT-3 0 and the reverse primer 5 0 GCAATATTTTTGCATTTAGTAAGGTT-3 0 . Since this fragment also includes a non-polymorphic
S.W. Baxter et al. / Cancer Letters 177 (2002) 75–81
77
Table 1 mEPHX exon 4 His139 . Arg139 genotype frequencies in ovarian cancer and control groups Group
Controls All ovarian cancer Serous Endometrioid/clear cell Mucinous Undifferentiated a
Number
257 291 127 90 40 34
OR (95% CI) a (Arg139/Arg139)
Number of cases with specified genotype (%) His139/His139 wt/wt
His139/Arg139 wt/high
Arg139/Arg139 high/high
152 (59.1) 193 (66.3) 85 (67.0) 57 (63.3) 28 (70.0) 23 (67.7)
92 (35.8) 82 (28.2) 37 (29.1) 26 (28.9) 9 (22.5) 10 (29.4)
13 16 5 7 3 1
(5.1) (5.5) (3.9) (7.7) (7.5) (2.9)
1.09 (0.51–2.31) 0.76 (0.26–2.20) 1.58 (0.61–4.10) 1.52 (0.41–5.59) 0.56 (0.07–4.49)
The odds ratio (OR) and 95% confidence intervals (CI) were calculated using the combined wt/wt and wt/high genotypes as a reference.
Tsp509 I site 59 bp from the 3 0 end, digestion with Tsp509 I produces fragments of 161 and 59 bp with the uncut ‘G’ allele and fragments of 50, 111 and 59 bp with the cut ‘A’ allele. 2.3. Statistical analysis Comparisons of frequency were analyzed using Fisher’s exact test. All statistical tests were based on two-tailed probability. 3. Results The distribution of the exon 4 alleles in the control and cancer groups was not significantly different to that predicted under Hardy–Weinberg equilibrium
(P ¼ 0:99 and P ¼ 0:41, respectively) (Table 1). However, the distribution of the exon 3 alleles in both the control and cancer groups was found to be highly significantly different to that predicted under Hardy–Weinberg equilibrium (P ¼ 0:007 and P ¼ 0:01, respectively) (Table 2). Specifically, in both groups there were more His113/His113 homozygotes and fewer His113/Tyr113 heterozygotes than expected. This suggested that the conventional PCRRFLP genotyping assay [13] might be flawed. Yoshikawa et al. [14] have reported the existence of a codon 119 G . A polymorphism which is contained within the reverse primer used for codon 113 genotyping. They reported that the codon 119 ‘A’ allele is in strong linkage disequilibrium with the Tyr113 allele. In Tyr113/His113 heterozygotes this primer mismatch
Table 2 mEPHX exon 3 Tyr113 . His113 genotype frequencies in ovarian cancer and control groups Group
Number
Controls b
257
All ovarian cancer
291
Serous Endometrioid/clear cell Mucinous Undifferentiated a
127 90 40 34
OR (95% CI) a (Tyr113/Tyr113)
Number of cases with specified genotype (%) His113/His113 slow/slow
Tyr113/His113 wt/slow
Tyr113/Tyr113 wt/wt
28 45 35 51 14 16 3 2
100 (38.9) 83 (32.3) 114 (39.2) 98 (33.7) 51 (40.2) 32 (35.5) 17 (42.5) 14 (41.2)
129 (50.2) 129 (50.2) 142 (48.8) 142 (48.8) 62 (48.8) 42 (46.7) 20 (50.0) 18 (52.9)
(10.9) (17.5) (12.0) (17.5) (11.0) (17.8) (7.5) (5.9)
0.95 (0.68–1.32) 0.95 (0.62–1.45) 0.87 (0.54–1.41) 0.99 (0.51–1.93) 1.12 (0.55–2.29)
The odds ratio (OR) and 95% confidence intervals (CI) were calculated using the pooled slow/slow and wt/slow genotype as a reference. Figures in italics are the genotype frequencies based on the PCR-RFLP assay. These are shown only for comparison with the allele-specific assay and were not used in the statistical calculations. b
78
S.W. Baxter et al. / Cancer Letters 177 (2002) 75–81
results in the failure of the Tyr113 allele to amplify, leading to a false His113/His113 genotype designation. We assessed the frequency of the codon 119 polymorphism among the control group utilizing the fact that the G . A substitution creates a Tsp509 I restriction site (GATT . AATT), an example of which is shown in Fig. 1A. The frequency of the ‘A’ allele in the control population was 0.12 with five A/A homozygotes, 49 A/G heterozygotes and 203 G/G homozygotes. We then reassessed all the controls for the frequency of the codon 113 alleles using a two-colour allele-specific assay with a primer that did not span the codon 119 polymorphism. As predicted, some of the His113/His113 homozygotes (17 individuals) were found to be Tyr113/His113 heterozygotes while all the Tyr113/Tyr113 homozygotes were confirmed as such in this independent assay (Table 2). A comparison of the codon 113 genotyping using the PCR-RFLP assay and the dual-colour allele-specific assay for some selected cases is shown in Fig. 1B,C, respectively. Genotyping the codon 119 polymorphism demonstrated that a false His113/ His113 genotype was recorded in the PCR-RFLP assay only in the presence of the codon 119 ‘A’ allele (Fig. 1A). Several individuals with a false His113/ His113 genotype were sequenced and in each case these were confirmed to be His113/Tyr113 heterozygotes and each contained a codon 119 ‘A’ allele (Fig.
Fig. 1. Examples of the genotyping results for codon 119 and codon 113 in ten control individuals. (A) Codon 119 PCR-RFLP genotyping assay. (B,C) Comparison of the mEPHX exon 3 genotyping using the PCR-RFLP assay (B) and the allele-specific assay described in this study (C). The two sets of lanes in (C) were derived from the same gel but scanned with a 488 laser (FAM) and a 532 laser (HEX). The asterisks in lanes 5–8 indicate individuals recording a His113/His113 genotype using the PCR-RFLP assay (B) but a His113/Tyr113 genotype using the allele-specific assay (C). Note that each of these individuals were heterozygous for the codon 119 A allele (A).
Fig. 2. Sequence of codon 113 and 119 of mEPHX for the individual from lane 5 of Fig. 1. Note that this individual was apparently a His113/His113 homozygote based on the PCR-RFLP assay but the sequence clearly shows that this individual is heterozygous at codon 113 and also heterozygous at codon 119.
2). Consistent with the findings of Yoshikawa et al. [14] we observed tight linkage between the codon 119 ‘A’ allele and the Tyr113 allele since all the codon 119 A/A homozygotes were also homozygous for Tyr113 and only one of the 49 codon 119 A/G heterozygotes was homozygous for His113. Having established that only the His113 homozygote cases may have been falsely classified, we reassessed these among the ovarian cancer cases. We found that 16 cases designated as His113 homozygotes using the PCR-RFLP assay were in fact Tyr113/His113 heterozygotes based on the allelespecific PCR assay. All the analyses of allele frequencies described below have been based on this more reliable genotyping assay. The distribution of the exon 3 and exon 4 alleles among the control population is similar to previous studies [4,6–9,15,16]. The distribution of EPHX genotypes among the ovarian cancers was similar to that observed in the controls for both the exon 3 and exon 4 polymorphisms (Tables 1 and 2). There was no evidence of an association of the Tyr113/Tyr113 genotype with ovarian cancer either as a single group (OR 0.95, 95% CI 0.68–1.32) or stratified according to histological sub-type. Similarly, the Arg139/Arg139 genotype was not associated with ovarian cancer (OR 1.09, 95% CI 0.51–2.31). Table 3 shows the three mEPHX activity levels which were assigned to each genotype according to current functional knowledge of the in vitro expression of the variant alleles [7]. The distribution of the
S.W. Baxter et al. / Cancer Letters 177 (2002) 75–81
79
Table 3 Distribution of predicted mEPHX activity among ovarian cancers and controls Group
Controls All ovarian cancer Serous Endometrioid/clear cell Mucinous Undifferentiated
Number
257 291 127 90 40 34
OR (95% CI) a (high)
Predicted mEPHX activity Slow b
Intermediate c
High d
97 (41.9) 109 (37.5) 47 (37.0) 37 (41.1) 14 (35.0) 11 (27.5)
87 (30.1) 128 (44.0) 59 (46.5) 32 (35.6) 20 (50.0) 17 (42.5)
73 (27.3) 54 (18.5) 21 (16.5) 21 (23.3) 6 (15) 6 (15)
One reference 0.57 (0.38–0.86) 0.50 (0.29–0.86) 0.77 (0.44–1.34) 0.44 (0.18–1.10) 0.54 (0.21–1.36)
a
The odds ratio (OR) and 95% confidence intervals (CI) were calculated using the pooled slow and intermediate genotypes as a reference. Predicted slow activity genotypes: exon 3 slow/slow 1 exon 4 wt/wt genotypes (23 controls, 26 cancers), exon 3 slow/slow 1 exon 4 wt/ high genotypes (5 controls, 8 cancers), exon 3 wt/slow 1 exon 4 wt/wt genotypes (69 controls, 75 cancers). c Predicted intermediate activity genotypes: exon 3 wt/wt 1 exon 4 wt/wt genotypes (60 controls, 92 cancers), exon 3 wt/slow 1 exon 4 wt/ high genotypes (27 controls, 33 cancers), exon 3 slow/slow 1 exon 4 high/high genotypes (0 controls, 3 cancers). d Predicted high activity genotypes: exon 3 wt/wt 1 exon 4 wt/high genotypes (60 controls, 40 cancers), exon 3 wt/wt 1 exon 4 high/high genotypes (9 controls, 7 cancers), exon 3 wt/slow 1 exon 4 high/high genotypes (4 controls, 7 cancers). b
predicted mEPHX activity was significantly different between the ovarian cancers and controls with an increase in the predicted intermediate mEPHX activity phenotype and corresponding decrease in the high activity phenotype (P , 0:01, OR 0.57, 95% CI 0.38– 0.86). A reduction in the high activity phenotype was evident among all the histological sub-types but this was only statistically significant among the serous tumors (P ¼ 0:01, OR 0.50, 95% CI 0.29–0.86). Interestingly, the frequency of low activity genotypes was also lower among all groups compared to the controls suggesting the possibility of a co-dominant effect. Compared to the combined high and slow activity groups, intermediate EPHX activity was associated with an OR of 1.53 (95% CI 1.08–2.17).
4. Discussion mEPHX is an important component in the activation and detoxification of exogenous chemicals as well as the metabolism of epoxides and endogenous steroids. Lancaster et al. [9] has reported an association of the exon 3 high activity Tyr113 allele with an increased risk of ovarian cancer. mEPHX is an attractive ovarian cancer-predisposing gene given its putative role in steroidogenesis. In particular, Hattori et al. [1] have recently shown that mEPHX is involved in estrogen production in the ovary. Never the less the study of Lancaster et al. [9] was small, only involving
73 patients and 75 controls, and the association must necessarily be viewed with caution. Indeed, a more recent large study failed to detect any association of the codon 119 polymorphism with ovarian cancer unstratified for histological sub-type [10]. Our investigation of 291 epithelial ovarian cancers has also failed to identify any association of the exon 3 variant with ovarian cancer. The frequency of the exon 4 variant has not previously been reported in ovarian cancer but again we did not observe any statistically significant association with ovarian cancer. As both polymorphisms affect mEPHX function, stratification based upon all possible exon 3 and exon 4 genotype combinations may provide a more accurate reflection of the influence of mEPHX activity on ovarian cancer risk. However, as this would result in very small sub-groups most investigators have combined the genotype data to provide an index of predicted activity based upon the in vitro expression of the four possible haplotypes [3]. We classified our cases and controls as slow, intermediate or high activity according to the categorization of Benhamou et al. [7]. Based upon this classification we observe a significant decrease in the number of ovarian carcinomas with high mEPHX activity and a corresponding increase in the number with intermediate mEPHX activity. This association was only observed among the serous ovarian cancers, although there was a similar but non-significant trend among the other histological sub-types. Our data suggest that
80
S.W. Baxter et al. / Cancer Letters 177 (2002) 75–81
higher mEPHX activity may be protective for serous ovarian cancer. The association of predicted low mEPHX activity with serous ovarian tumors could be viewed as inconsistent with the absence of any measurable association with either the exon 3 or exon 4 genotypes. In addition, this association is in the opposite direction to that reported by Lancaster who reported that the high activity exon 3 Tyr113 allele was associated with an increased risk of ovarian cancer [9]. Consequently, our conclusions need to be interpreted cautiously particularly as the predicted mEPHX activity has been based on in vitro expression studies that may not necessarily accurately predict in vivo expression [16]. In addition, other genetic variants of mEPHX have recently been identified which may contribute to functional variation of in vivo mEPHX expression [17]. Previous studies of mEPHX exon 3 alleles have utilized a PCR-RFLP-based assay that Yoshikawa et al. [14] and ourselves have shown to be flawed because of a G . A polymorphism at codon 119 which is included in the reverse primer used in the conventional PCR-RFLP assay [13]. Yoshikawa et al. [14] suggested that this polymorphism is unlikely to exist in Caucasian populations but we have shown that it is relatively common with 20% of the British population carrying at least one ‘A’ allele. The interference of this polymorphism on the reliability of the PCRRFLP assay resulted in a significant skewing of the genotype distributions. To our knowledge the majority of previous studies of the mEPHX codon 113 polymorphism, apart from the study of Spurdle et al. [10], have been conducted using the PCR-RFLP assay and consequently the validity of these studies must be questioned. In summary, our findings do not support an association of individual exon 3 and exon 4 genotypes with ovarian cancer risk. However, based on predicted mEPHX activity our data do suggest that higher mEPHX activity may be protective for serous ovarian cancer. Confounding due to differences in ethnicity is unlikely because both the cases and controls were drawn from the same geographical area and the population is predominantly Anglo-Saxon in origin. Although the controls were significantly younger than the cases this is unlikely to affect our results given that ovarian cancer is a very rare disease in
Caucasian and other population groups. Nevertheless, we recognize that our conclusions are based upon multiple comparisons of small sub-groups and it was not possible with our data set to adjust for known risk factors for ovarian cancer or other lifestyle factors. Consequently, it will be important to verify our conclusion in independent studies. We have also shown that previous disease association studies of exon 3 alleles which utilized the PCR-RFLP assay may be compromised by the existence of a codon 119 G . A polymorphism which is common in Caucasian and Japanese populations.
References [1] N. Hattori, H. Fujiwara, M. Maeda, S. Fujii, M. Ueda, Epoxide hydrolase affects estrogen production in the human ovary, Endocrinology 141 (2000) 3353–3365. [2] J. Seidegard, J.W. DePierre, Microsomal epoxide hydrolase. Properties, regulation and function, Biochim. Biophys. Acta 695 (1983) 251–270. [3] C. Hassett, L. Aicher, J.S. Sidhu, C.J. Omiecinski, Human microsomal epoxide hydrolase: genetic polymorphism and functional expression in vitro of amino acid variants, Hum. Mol. Genet. 3 (1994) 421–428. [4] D.J. Harrison, A.L. Hubbard, J. Mac Millan, A.H. Wyllie, C.A. Smith, Microsomal epoxide hydrolase gene polymorphism and susceptibility to colon cancer, Br. J. Cancer 79 (1999) 168–171. [5] K.A. McGlynn, E.A. Rosvold, E.D. Lustbader, Y. Hu, M.L. Clapper, T. Zhou, C.P. Wild, X.L. Xia, A. Baffoe-Bonnie, D. Ofori-Adjei, Susceptibility to hepatocellular carcinoma is associated with genetic variation in the enzymatic detoxification of aflatoxin B1, Proc. Natl. Acad. Sci. USA 92 (1995) 2384–2387. [6] S.J. London, J. Smart, A.K. Daly, Lung cancer risk in relation to genetic polymorphisms of microsomal epoxide hydrolase among African-Americans and Caucasians in Los Angeles County, Lung Cancer 28 (2000) 147–155. [7] S. Benhamou, M. Reinikainen, C. Bouchardy, P. Dayer, A. Hirvonen, Association between lung cancer and microsomal epoxide hydrolase genotypes, Cancer Res. 58 (1998) 5291– 5293. [8] N. Jourenkova-Mironova, K. Mitrunen, C. Bouchardy, P. Dayer, S. Benhamou, A. Hirvonen, High-activity microsomal epoxide hydrolase genotypes and the risk of oral, pharynx, and larynx cancers, Cancer Res. 60 (2000) 534–536. [9] J.M. Lancaster, H.A. Brownlee, D.A. Bell, P.A. Futreal, J.R. Marks, A. Berchuck, R.W. Wiseman, J.A. Taylor, Microsomal epoxide hydrolase polymorphism as a risk factor for ovarian cancer, Mol. Carcinog. 17 (1996) 160–162. [10] A.B. Spurdle, D.M. Purdie, P.M. Webb, X. Chen, A. Green, G. Chenevix-Trench, The microsomal epoxide hydrolase
S.W. Baxter et al. / Cancer Letters 177 (2002) 75–81
[11]
[12]
[13]
[14]
Tyr113His polymorphism: association with risk of ovarian cancer, Mol. Carcinog. 30 (2001) 71–78. S.J. Morland, X. Jiang, A. Hitchcock, E.J. Thomas, I.G. Campbell, Mutation of galactose-1-phosphate uridyl transferase and its association with ovarian cancer and endometriosis, Int. J. Cancer 77 (1998) 825–827. T.P. Manolitsas, P. Englefield, D.M. Eccles, I.G. Campbell, No association of a 306-bp insertion polymorphism in the progesterone receptor gene with ovarian and breast cancer, Br. J. Cancer 75 (1997) 1398–1399. C.A. Smith, D.J. Harrison, Association between polymorphism in gene for microsomal epoxide hydrolase and susceptibility to emphysema, Lancet 350 (1997) 630–633. M. Yoshikawa, K. Hiyama, S. Ishioka, H. Maeda, A. Maeda, M. Yamakido, Microsomal epoxide hydrolase genotypes and
81
chronic obstructive pulmonary disease in Japanese, Int. J. Mol. Med. 5 (2000) 49–53. [15] N.A. Wong, F. Rae, A. Bathgate, C.A.D. Smith, D.J. Harrison, Polymorphisms of the gene for microsomal epoxide hydrolase and susceptibility to alcoholic liver disease and hepatocellular carcinoma in a Caucasian population, Toxicol. Lett. 115 (2000) 17–22. [16] N.R. Kitteringham, C. Davis, N. Howard, M. Pirmohamed, B.K. Park, Interindividual and interspecies variation in hepatic microsomal epoxide hydrolase activity: studies with cis-stilbene oxide, carbamazepine 10, 11-epoxide and naphthalene, J. Pharmacol. Exp. Ther. 278 (1996) 1018–1027. [17] S. Raaka, C. Hassett, C.J. Omiencinski, Human microsomal epoxide hydrolase: 5 0 -flanking region genetic polymorphisms, Carcinogenesis 19 (1998) 387–393.
Microsomal epoxide hydrolase polymorphism and susceptibility to ovarian cancer S.W. Baxter, D.Y.H. Choong, I.G. Campbell* VBCRC Cancer Genetics Laboratory, Peter MacCallum Cancer Institute, Locked Bag No. 1 A’Beckett Street, Melbourne, Victoria, 8006, Australia Received 6 July 2001; received in revised form 8 September 2001; accepted 14 September 2001
Abstract Polymorphic variants of microsomal epoxide hydrolase (mEPHX) with altered enzyme activity have been associated with an increased risk for ovarian cancer. We assessed the frequency of exon 3 and exon 4 variants of mEPHX among 291 ovarian cancers and 257 controls from a UK-based population. The distribution of the exon 3 alleles among both the cancer and control groups was significantly different from that expected under Hardy–Weinberg equilibrium suggesting that the PCR restriction fragment length polymorphism (PCR-RFLP) genotyping assay might be flawed. The codon 113 polymorphism was reassessed using a two-color allele-specific PCR-based assay. We found that a codon 119 G . A polymorphism, present in 20% of the British population and linked to the wild-type exon 3 allele, resulted in some Tyr113/His113 heterozygotes being falsely classified as His113/His113 homozygotes when using the PCR-RFLP assay. Consequently, we reassessed all our codon 113 data using the new allele-specific assay. We found no evidence of an association of ovarian cancer risk with the exon 3 Tyr113 . His113 variant. Similarly the frequencies of the exon 4 His139 . Arg139 genotypes were not significantly different between cases and controls. Stratifying the genotyping data according to the predicted mEPHX activity revealed a highly significant decrease in high mEPHX activity among the serous ovarian cancers (P ¼ 0:01) suggesting that high mEPHX activity may be protective for this histological sub-type. Furthermore previous disease association studies of exon 3 alleles which utilized the PCR-RFLP assay may be compromised by the existence of a codon 119 G . A polymorphism which may be common in Caucasian populations. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Ovarian cancer; Microsomal epoxide hydrolase; Cancer predisposition; Polymorphism
1. Introduction It is likely that inherited genetic factors contribute to the development of ovarian cancer since familial clustering of this disease is well documented. Polymorphic variants in genes involved in detoxifying endogenous or exogenous genotoxic chemicals may * Corresponding author. Tel.: 161-3-96561803; fax: 161-396561411. E-mail address: [email protected] (I.G. Campbell).
represent an important class of low penetrance disease predisposing alleles. Microsomal epoxide hydrolase (mEPHX) is involved in the activation and detoxification of exogenous chemicals and may also be involved in steroidogenesis [1,2]. Two variants of mEPHX with different enzyme activity have been described [3]. A tyrosine (Tyr113) to histidine (His113) substitution at codon 113 in exon 3 decreases enzyme activity by approximately 40% while a histidine (His139) to arginine (Arg139) substitution at codon 139 in exon 4 increases enzyme activity by more than 25%. The
0304-3835/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0304-383 5(01)00782-0
76
S.W. Baxter et al. / Cancer Letters 177 (2002) 75–81
His113 slow activity genotype has been associated with an increased risk of colon cancer [4] and hepatocellular cancer [5], whereas the high activity Tyr113 genotype has been associated with an increase in risk for smoking-related lung [6,7] and upper aero-digestive tract cancers [8]. The high activity Tyr113 genotype has also been shown to increase ovarian cancer risk in a small study of 75 ovarian cancers and 73 controls [9] although Spurdle et al. [10] did not observe any association in a much larger study. We report here an investigation of the frequency of both the exon 3 and exon 4 mEPHX polymorphic variants in a retrospective case control study of 291 epithelial ovarian cancers and 257 controls. 2. Materials and methods 2.1. Subjects Details of cases and controls have been described previously [11,12]. Briefly, incident cases of ovarian tumors were ascertained from women undergoing primary surgery in hospitals from southern England between 1993 and 1998. A specialist gynecological pathologist confirmed the histological diagnosis for each tumor and only malignant epithelial ovarian tumors (n ¼ 291) were included in the study. The histological sub-types of the tumors included in the study were serous (127), mucinous (40), endometrioid (78), clear cell (12) and undifferentiated (34). The controls consisted of 257 white female volunteers who were either staff at the Princess Anne Hospital, Southampton, UK or patients attending for nonneoplastic disease conditions. Epidemiological data such as reproductive factors, oral contraceptive use, smoking and obesity were not available for either the cases or controls. However, both groups were resident in and around the Southampton metropolitan area, which is predominantly an Anglo-Saxon population. The age range of the ovarian cancer patients was 23– 90 with a mean of 62 years and for the controls the age range was 20–78 with a mean of 42 years. 2.2. mEPHX genotyping The exon 3 and exon 4 variants were detected using the PCR restriction fragment length polymorphism (RFLP) assay described by Smith and Harrison [13].
Briefly, the codon 113 T . C variant (Tyr113 to His113) was detected using a FAM-labeled forward primer (5 0 -GATCGATAAGTTCCGTTTCACC-3 0 ) and a reverse primer (5 0 -ATCCTTAGTCTTGAAGTGAGGaT-3 0 ) which included an engineered base change of an A for a G (shown in lower case) which produces an EcoRV restriction enzyme site in the Tyr113 allele. The codon 139 A . G variant (His139 . Arg139) produces an RsaI restriction enzyme site and was detected using the forward primer 5 0 -ACATCCACTTCATCCACGT-3 0 and a HEXlabeled reverse primer 5 0 -ATGCCTCTGAGAAGCCAT-3 0 . PCRs were performed in a reaction volume of 10 ml containing 10–200 ng of genomic DNA, primers (50 ng of each), 1 £ reaction buffer (Applied Biotechnologies, UK), 200 nM of dATP, dTTP, dGTP, and dCTP (Promega, UK) and 0.2 units of Taq DNA polymerase (Red Hot Taq, Applied Biotechnologies, UK). PCR conditions consisted of an initial denaturation at 948C for 5 min followed by 35 cycles of 948C for 30 s, 558C for 30 s, and 728C for 60 s followed by one cycle of 728C for 5 min. PCRs and restriction digest were carried out as multiplex reactions and the products were separated on 2.5% agarose gels. A scanning laser fluorescence imager (Bio-Rad Molecular Imager FX) was used to discriminate the exon 3 and exon 4 alleles using 488 and 532 nm lasers, respectively. A substitution polymorphism (G to A) at codon 119 has been reported [14]. This polymorphism may affect the accuracy of the exon 3 genotyping and to circumvent this potential problem we developed a dual-colour allele-specific assay for the codon 113 polymorphism. We amplified each allele using a FAM-labeled forward primer (5 0 -CAGGTGGAGATTCTCAACAGAC-3 0 ) specific for the codon 113 ‘C’ allele and a HEX-labeled forward primer (5 0 CAGGTGGAGATTCTCAACAGAT-3 0 ) specific for the ‘T’ allele and a common unlabeled reverse primer (5 0 -CCTGCCTAGCTCTAAAGATGGA-3 0 ). The frequency of the codon 119 polymorphism was assessed using a PCR-RFLP assay based on the fact that the G . A substitution creates a Tsp509 I restriction enzyme recognition site. A 220 bp fragment encompassing codons 113 and 119 was amplified using the forward primer 5 0 -TGGAAGAAGCAGGTGGAGAT-3 0 and the reverse primer 5 0 GCAATATTTTTGCATTTAGTAAGGTT-3 0 . Since this fragment also includes a non-polymorphic
S.W. Baxter et al. / Cancer Letters 177 (2002) 75–81
77
Table 1 mEPHX exon 4 His139 . Arg139 genotype frequencies in ovarian cancer and control groups Group
Controls All ovarian cancer Serous Endometrioid/clear cell Mucinous Undifferentiated a
Number
257 291 127 90 40 34
OR (95% CI) a (Arg139/Arg139)
Number of cases with specified genotype (%) His139/His139 wt/wt
His139/Arg139 wt/high
Arg139/Arg139 high/high
152 (59.1) 193 (66.3) 85 (67.0) 57 (63.3) 28 (70.0) 23 (67.7)
92 (35.8) 82 (28.2) 37 (29.1) 26 (28.9) 9 (22.5) 10 (29.4)
13 16 5 7 3 1
(5.1) (5.5) (3.9) (7.7) (7.5) (2.9)
1.09 (0.51–2.31) 0.76 (0.26–2.20) 1.58 (0.61–4.10) 1.52 (0.41–5.59) 0.56 (0.07–4.49)
The odds ratio (OR) and 95% confidence intervals (CI) were calculated using the combined wt/wt and wt/high genotypes as a reference.
Tsp509 I site 59 bp from the 3 0 end, digestion with Tsp509 I produces fragments of 161 and 59 bp with the uncut ‘G’ allele and fragments of 50, 111 and 59 bp with the cut ‘A’ allele. 2.3. Statistical analysis Comparisons of frequency were analyzed using Fisher’s exact test. All statistical tests were based on two-tailed probability. 3. Results The distribution of the exon 4 alleles in the control and cancer groups was not significantly different to that predicted under Hardy–Weinberg equilibrium
(P ¼ 0:99 and P ¼ 0:41, respectively) (Table 1). However, the distribution of the exon 3 alleles in both the control and cancer groups was found to be highly significantly different to that predicted under Hardy–Weinberg equilibrium (P ¼ 0:007 and P ¼ 0:01, respectively) (Table 2). Specifically, in both groups there were more His113/His113 homozygotes and fewer His113/Tyr113 heterozygotes than expected. This suggested that the conventional PCRRFLP genotyping assay [13] might be flawed. Yoshikawa et al. [14] have reported the existence of a codon 119 G . A polymorphism which is contained within the reverse primer used for codon 113 genotyping. They reported that the codon 119 ‘A’ allele is in strong linkage disequilibrium with the Tyr113 allele. In Tyr113/His113 heterozygotes this primer mismatch
Table 2 mEPHX exon 3 Tyr113 . His113 genotype frequencies in ovarian cancer and control groups Group
Number
Controls b
257
All ovarian cancer
291
Serous Endometrioid/clear cell Mucinous Undifferentiated a
127 90 40 34
OR (95% CI) a (Tyr113/Tyr113)
Number of cases with specified genotype (%) His113/His113 slow/slow
Tyr113/His113 wt/slow
Tyr113/Tyr113 wt/wt
28 45 35 51 14 16 3 2
100 (38.9) 83 (32.3) 114 (39.2) 98 (33.7) 51 (40.2) 32 (35.5) 17 (42.5) 14 (41.2)
129 (50.2) 129 (50.2) 142 (48.8) 142 (48.8) 62 (48.8) 42 (46.7) 20 (50.0) 18 (52.9)
(10.9) (17.5) (12.0) (17.5) (11.0) (17.8) (7.5) (5.9)
0.95 (0.68–1.32) 0.95 (0.62–1.45) 0.87 (0.54–1.41) 0.99 (0.51–1.93) 1.12 (0.55–2.29)
The odds ratio (OR) and 95% confidence intervals (CI) were calculated using the pooled slow/slow and wt/slow genotype as a reference. Figures in italics are the genotype frequencies based on the PCR-RFLP assay. These are shown only for comparison with the allele-specific assay and were not used in the statistical calculations. b
78
S.W. Baxter et al. / Cancer Letters 177 (2002) 75–81
results in the failure of the Tyr113 allele to amplify, leading to a false His113/His113 genotype designation. We assessed the frequency of the codon 119 polymorphism among the control group utilizing the fact that the G . A substitution creates a Tsp509 I restriction site (GATT . AATT), an example of which is shown in Fig. 1A. The frequency of the ‘A’ allele in the control population was 0.12 with five A/A homozygotes, 49 A/G heterozygotes and 203 G/G homozygotes. We then reassessed all the controls for the frequency of the codon 113 alleles using a two-colour allele-specific assay with a primer that did not span the codon 119 polymorphism. As predicted, some of the His113/His113 homozygotes (17 individuals) were found to be Tyr113/His113 heterozygotes while all the Tyr113/Tyr113 homozygotes were confirmed as such in this independent assay (Table 2). A comparison of the codon 113 genotyping using the PCR-RFLP assay and the dual-colour allele-specific assay for some selected cases is shown in Fig. 1B,C, respectively. Genotyping the codon 119 polymorphism demonstrated that a false His113/ His113 genotype was recorded in the PCR-RFLP assay only in the presence of the codon 119 ‘A’ allele (Fig. 1A). Several individuals with a false His113/ His113 genotype were sequenced and in each case these were confirmed to be His113/Tyr113 heterozygotes and each contained a codon 119 ‘A’ allele (Fig.
Fig. 1. Examples of the genotyping results for codon 119 and codon 113 in ten control individuals. (A) Codon 119 PCR-RFLP genotyping assay. (B,C) Comparison of the mEPHX exon 3 genotyping using the PCR-RFLP assay (B) and the allele-specific assay described in this study (C). The two sets of lanes in (C) were derived from the same gel but scanned with a 488 laser (FAM) and a 532 laser (HEX). The asterisks in lanes 5–8 indicate individuals recording a His113/His113 genotype using the PCR-RFLP assay (B) but a His113/Tyr113 genotype using the allele-specific assay (C). Note that each of these individuals were heterozygous for the codon 119 A allele (A).
Fig. 2. Sequence of codon 113 and 119 of mEPHX for the individual from lane 5 of Fig. 1. Note that this individual was apparently a His113/His113 homozygote based on the PCR-RFLP assay but the sequence clearly shows that this individual is heterozygous at codon 113 and also heterozygous at codon 119.
2). Consistent with the findings of Yoshikawa et al. [14] we observed tight linkage between the codon 119 ‘A’ allele and the Tyr113 allele since all the codon 119 A/A homozygotes were also homozygous for Tyr113 and only one of the 49 codon 119 A/G heterozygotes was homozygous for His113. Having established that only the His113 homozygote cases may have been falsely classified, we reassessed these among the ovarian cancer cases. We found that 16 cases designated as His113 homozygotes using the PCR-RFLP assay were in fact Tyr113/His113 heterozygotes based on the allelespecific PCR assay. All the analyses of allele frequencies described below have been based on this more reliable genotyping assay. The distribution of the exon 3 and exon 4 alleles among the control population is similar to previous studies [4,6–9,15,16]. The distribution of EPHX genotypes among the ovarian cancers was similar to that observed in the controls for both the exon 3 and exon 4 polymorphisms (Tables 1 and 2). There was no evidence of an association of the Tyr113/Tyr113 genotype with ovarian cancer either as a single group (OR 0.95, 95% CI 0.68–1.32) or stratified according to histological sub-type. Similarly, the Arg139/Arg139 genotype was not associated with ovarian cancer (OR 1.09, 95% CI 0.51–2.31). Table 3 shows the three mEPHX activity levels which were assigned to each genotype according to current functional knowledge of the in vitro expression of the variant alleles [7]. The distribution of the
S.W. Baxter et al. / Cancer Letters 177 (2002) 75–81
79
Table 3 Distribution of predicted mEPHX activity among ovarian cancers and controls Group
Controls All ovarian cancer Serous Endometrioid/clear cell Mucinous Undifferentiated
Number
257 291 127 90 40 34
OR (95% CI) a (high)
Predicted mEPHX activity Slow b
Intermediate c
High d
97 (41.9) 109 (37.5) 47 (37.0) 37 (41.1) 14 (35.0) 11 (27.5)
87 (30.1) 128 (44.0) 59 (46.5) 32 (35.6) 20 (50.0) 17 (42.5)
73 (27.3) 54 (18.5) 21 (16.5) 21 (23.3) 6 (15) 6 (15)
One reference 0.57 (0.38–0.86) 0.50 (0.29–0.86) 0.77 (0.44–1.34) 0.44 (0.18–1.10) 0.54 (0.21–1.36)
a
The odds ratio (OR) and 95% confidence intervals (CI) were calculated using the pooled slow and intermediate genotypes as a reference. Predicted slow activity genotypes: exon 3 slow/slow 1 exon 4 wt/wt genotypes (23 controls, 26 cancers), exon 3 slow/slow 1 exon 4 wt/ high genotypes (5 controls, 8 cancers), exon 3 wt/slow 1 exon 4 wt/wt genotypes (69 controls, 75 cancers). c Predicted intermediate activity genotypes: exon 3 wt/wt 1 exon 4 wt/wt genotypes (60 controls, 92 cancers), exon 3 wt/slow 1 exon 4 wt/ high genotypes (27 controls, 33 cancers), exon 3 slow/slow 1 exon 4 high/high genotypes (0 controls, 3 cancers). d Predicted high activity genotypes: exon 3 wt/wt 1 exon 4 wt/high genotypes (60 controls, 40 cancers), exon 3 wt/wt 1 exon 4 high/high genotypes (9 controls, 7 cancers), exon 3 wt/slow 1 exon 4 high/high genotypes (4 controls, 7 cancers). b
predicted mEPHX activity was significantly different between the ovarian cancers and controls with an increase in the predicted intermediate mEPHX activity phenotype and corresponding decrease in the high activity phenotype (P , 0:01, OR 0.57, 95% CI 0.38– 0.86). A reduction in the high activity phenotype was evident among all the histological sub-types but this was only statistically significant among the serous tumors (P ¼ 0:01, OR 0.50, 95% CI 0.29–0.86). Interestingly, the frequency of low activity genotypes was also lower among all groups compared to the controls suggesting the possibility of a co-dominant effect. Compared to the combined high and slow activity groups, intermediate EPHX activity was associated with an OR of 1.53 (95% CI 1.08–2.17).
4. Discussion mEPHX is an important component in the activation and detoxification of exogenous chemicals as well as the metabolism of epoxides and endogenous steroids. Lancaster et al. [9] has reported an association of the exon 3 high activity Tyr113 allele with an increased risk of ovarian cancer. mEPHX is an attractive ovarian cancer-predisposing gene given its putative role in steroidogenesis. In particular, Hattori et al. [1] have recently shown that mEPHX is involved in estrogen production in the ovary. Never the less the study of Lancaster et al. [9] was small, only involving
73 patients and 75 controls, and the association must necessarily be viewed with caution. Indeed, a more recent large study failed to detect any association of the codon 119 polymorphism with ovarian cancer unstratified for histological sub-type [10]. Our investigation of 291 epithelial ovarian cancers has also failed to identify any association of the exon 3 variant with ovarian cancer. The frequency of the exon 4 variant has not previously been reported in ovarian cancer but again we did not observe any statistically significant association with ovarian cancer. As both polymorphisms affect mEPHX function, stratification based upon all possible exon 3 and exon 4 genotype combinations may provide a more accurate reflection of the influence of mEPHX activity on ovarian cancer risk. However, as this would result in very small sub-groups most investigators have combined the genotype data to provide an index of predicted activity based upon the in vitro expression of the four possible haplotypes [3]. We classified our cases and controls as slow, intermediate or high activity according to the categorization of Benhamou et al. [7]. Based upon this classification we observe a significant decrease in the number of ovarian carcinomas with high mEPHX activity and a corresponding increase in the number with intermediate mEPHX activity. This association was only observed among the serous ovarian cancers, although there was a similar but non-significant trend among the other histological sub-types. Our data suggest that
80
S.W. Baxter et al. / Cancer Letters 177 (2002) 75–81
higher mEPHX activity may be protective for serous ovarian cancer. The association of predicted low mEPHX activity with serous ovarian tumors could be viewed as inconsistent with the absence of any measurable association with either the exon 3 or exon 4 genotypes. In addition, this association is in the opposite direction to that reported by Lancaster who reported that the high activity exon 3 Tyr113 allele was associated with an increased risk of ovarian cancer [9]. Consequently, our conclusions need to be interpreted cautiously particularly as the predicted mEPHX activity has been based on in vitro expression studies that may not necessarily accurately predict in vivo expression [16]. In addition, other genetic variants of mEPHX have recently been identified which may contribute to functional variation of in vivo mEPHX expression [17]. Previous studies of mEPHX exon 3 alleles have utilized a PCR-RFLP-based assay that Yoshikawa et al. [14] and ourselves have shown to be flawed because of a G . A polymorphism at codon 119 which is included in the reverse primer used in the conventional PCR-RFLP assay [13]. Yoshikawa et al. [14] suggested that this polymorphism is unlikely to exist in Caucasian populations but we have shown that it is relatively common with 20% of the British population carrying at least one ‘A’ allele. The interference of this polymorphism on the reliability of the PCRRFLP assay resulted in a significant skewing of the genotype distributions. To our knowledge the majority of previous studies of the mEPHX codon 113 polymorphism, apart from the study of Spurdle et al. [10], have been conducted using the PCR-RFLP assay and consequently the validity of these studies must be questioned. In summary, our findings do not support an association of individual exon 3 and exon 4 genotypes with ovarian cancer risk. However, based on predicted mEPHX activity our data do suggest that higher mEPHX activity may be protective for serous ovarian cancer. Confounding due to differences in ethnicity is unlikely because both the cases and controls were drawn from the same geographical area and the population is predominantly Anglo-Saxon in origin. Although the controls were significantly younger than the cases this is unlikely to affect our results given that ovarian cancer is a very rare disease in
Caucasian and other population groups. Nevertheless, we recognize that our conclusions are based upon multiple comparisons of small sub-groups and it was not possible with our data set to adjust for known risk factors for ovarian cancer or other lifestyle factors. Consequently, it will be important to verify our conclusion in independent studies. We have also shown that previous disease association studies of exon 3 alleles which utilized the PCR-RFLP assay may be compromised by the existence of a codon 119 G . A polymorphism which is common in Caucasian and Japanese populations.
References [1] N. Hattori, H. Fujiwara, M. Maeda, S. Fujii, M. Ueda, Epoxide hydrolase affects estrogen production in the human ovary, Endocrinology 141 (2000) 3353–3365. [2] J. Seidegard, J.W. DePierre, Microsomal epoxide hydrolase. Properties, regulation and function, Biochim. Biophys. Acta 695 (1983) 251–270. [3] C. Hassett, L. Aicher, J.S. Sidhu, C.J. Omiecinski, Human microsomal epoxide hydrolase: genetic polymorphism and functional expression in vitro of amino acid variants, Hum. Mol. Genet. 3 (1994) 421–428. [4] D.J. Harrison, A.L. Hubbard, J. Mac Millan, A.H. Wyllie, C.A. Smith, Microsomal epoxide hydrolase gene polymorphism and susceptibility to colon cancer, Br. J. Cancer 79 (1999) 168–171. [5] K.A. McGlynn, E.A. Rosvold, E.D. Lustbader, Y. Hu, M.L. Clapper, T. Zhou, C.P. Wild, X.L. Xia, A. Baffoe-Bonnie, D. Ofori-Adjei, Susceptibility to hepatocellular carcinoma is associated with genetic variation in the enzymatic detoxification of aflatoxin B1, Proc. Natl. Acad. Sci. USA 92 (1995) 2384–2387. [6] S.J. London, J. Smart, A.K. Daly, Lung cancer risk in relation to genetic polymorphisms of microsomal epoxide hydrolase among African-Americans and Caucasians in Los Angeles County, Lung Cancer 28 (2000) 147–155. [7] S. Benhamou, M. Reinikainen, C. Bouchardy, P. Dayer, A. Hirvonen, Association between lung cancer and microsomal epoxide hydrolase genotypes, Cancer Res. 58 (1998) 5291– 5293. [8] N. Jourenkova-Mironova, K. Mitrunen, C. Bouchardy, P. Dayer, S. Benhamou, A. Hirvonen, High-activity microsomal epoxide hydrolase genotypes and the risk of oral, pharynx, and larynx cancers, Cancer Res. 60 (2000) 534–536. [9] J.M. Lancaster, H.A. Brownlee, D.A. Bell, P.A. Futreal, J.R. Marks, A. Berchuck, R.W. Wiseman, J.A. Taylor, Microsomal epoxide hydrolase polymorphism as a risk factor for ovarian cancer, Mol. Carcinog. 17 (1996) 160–162. [10] A.B. Spurdle, D.M. Purdie, P.M. Webb, X. Chen, A. Green, G. Chenevix-Trench, The microsomal epoxide hydrolase
S.W. Baxter et al. / Cancer Letters 177 (2002) 75–81
[11]
[12]
[13]
[14]
Tyr113His polymorphism: association with risk of ovarian cancer, Mol. Carcinog. 30 (2001) 71–78. S.J. Morland, X. Jiang, A. Hitchcock, E.J. Thomas, I.G. Campbell, Mutation of galactose-1-phosphate uridyl transferase and its association with ovarian cancer and endometriosis, Int. J. Cancer 77 (1998) 825–827. T.P. Manolitsas, P. Englefield, D.M. Eccles, I.G. Campbell, No association of a 306-bp insertion polymorphism in the progesterone receptor gene with ovarian and breast cancer, Br. J. Cancer 75 (1997) 1398–1399. C.A. Smith, D.J. Harrison, Association between polymorphism in gene for microsomal epoxide hydrolase and susceptibility to emphysema, Lancet 350 (1997) 630–633. M. Yoshikawa, K. Hiyama, S. Ishioka, H. Maeda, A. Maeda, M. Yamakido, Microsomal epoxide hydrolase genotypes and
81
chronic obstructive pulmonary disease in Japanese, Int. J. Mol. Med. 5 (2000) 49–53. [15] N.A. Wong, F. Rae, A. Bathgate, C.A.D. Smith, D.J. Harrison, Polymorphisms of the gene for microsomal epoxide hydrolase and susceptibility to alcoholic liver disease and hepatocellular carcinoma in a Caucasian population, Toxicol. Lett. 115 (2000) 17–22. [16] N.R. Kitteringham, C. Davis, N. Howard, M. Pirmohamed, B.K. Park, Interindividual and interspecies variation in hepatic microsomal epoxide hydrolase activity: studies with cis-stilbene oxide, carbamazepine 10, 11-epoxide and naphthalene, J. Pharmacol. Exp. Ther. 278 (1996) 1018–1027. [17] S. Raaka, C. Hassett, C.J. Omiencinski, Human microsomal epoxide hydrolase: 5 0 -flanking region genetic polymorphisms, Carcinogenesis 19 (1998) 387–393.