Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Pharmacogenet Genomics. Author manuscript; available in PMC 2012 October 1.
Published in final edited form as:
PMCID: PMC3172373

Xenobiotic metabolizing gene variants, pesticide use, and risk of prostate cancer



To explore associations with prostate cancer and farming, it is important to investigate the relationship between pesticide use and single nucleotide polymorphisms (SNPs) in xenobiotic metabolic enzyme (XME) genes.


We evaluated pesticide-SNP interactions between 45 pesticides and 1,913 XME SNPs with respect to prostate cancer among 776 cases and 1,444 controls in the Agricultural Health Study.


We used unconditional logistic regression to estimate odds ratios (ORs) and 95% confidence intervals (CIs). Multiplicative SNP-pesticide interactions were calculated using a likelihood ratio test.


A positive monotonic interaction was observed between petroleum oil/petroleum distillate use and rs1883633 in the oxidative stress gene glutamate-cysteine ligase (GCLC) (p-interaction=1.0×10−4); men carrying at least one variant allele (minor allele) experienced an increased prostate cancer risk (OR=3.7, 95% CI: 1.9–7.3). Among men carrying the variant allele for thioredoxin reductase 2 (TXNRD2) rs4485648, microsomal epoxide hyrdolase 1 (EPHX1) rs17309872, or myeloperoxidase (MPO) rs11079344, increased prostate cancer risk was observed with high compared to no petroleum oil/petroleum distillate (OR=1.9, 95% CI: 1.1–3.2, p-interaction=0.01), (OR=2.1, 95% CI: 1.1–4.0, p-interaction=0.01), or terbufos (OR=3.0, 95% CI: 1.5–6.0 p-interaction=2.0×10−3) use, respectively. No interactions were deemed noteworthy at the false discovery rate = 0.20 level; the number of observed interactions in XMEs was comparable to the number expected by chance alone.


We observed several pesticide-SNP interactions in oxidative stress and phase I/phase II enzyme genes and risk of prostate cancer. Additional work is needed to explain the joint contribution of genetic variation in XMEs, pesticide use, and prostate cancer risk.

Keywords: Prostate cancer, pesticides, xenobiotic metabolizing enzymes, single nucleotide polymorphism, interaction


There have been few environmental factors that have been identified to alter prostate cancer risk. However, results from the Agricultural Health Study (AHS), a prospective cohort of licensed private and commercial pesticide applicators in Iowa and North Carolina, show that pesticide applicators have a significantly higher risk of prostate cancer than men in the general population of Iowa and North Carolina [1]. The metabolism and subsequent excretion of pesticides from the body requires a series of chemical reactions that depend on xenobiotic metabolic enzymes (XMEs) [2]. Phase I and phase II enzyme reactions are well-described processes for the clearance of xenobiotic substances. The enzymes responsible for mediating phase I reactions encompass members of the cytochrome P450 (CYPs) superfamily [3]. Substantial literature documents the principal involvement of CYPs in the metabolism of specific xenobiotic substrates including some herbicides and organophosphophate (OP) insecticides [48]; still, the metabolism of many pesticides is not well characterized. Phase II enzymes also play a crucial role in xenobiotic metabolism and are a necessary part of the pathway to excretion. Phase II enzymes, such as sulfotransferases (SULTs), N-acetyltransferases (NATs), UDP-glucuronosyltransferases (UGTs), and glutathione S-transferases (GSTs] catalyze conjugation reactions of intermediates directly to form detoxification products, or further metabolize other reactive intermediates for future excretion [3]. In addition, key nuclear receptors including the aryl hydrocarbon receptor (AHR), pregnane X receptor (PXR or NR1I2) and the constitutive active/androstane receptor (CAR or NR1I3) can stimulate gene transcription in response to pesticides and affect their metabolism [911]. Accumulation of toxic pesticide intermediates can induce the production of reactive oxygen species (ROS), markers of oxidative stress, which can react with cellular DNA to cause mutation or gross DNA rearrangements [12;13]. For example, the roles of enzymatic antioxidants like superoxide dismutase (SOD) and catalase (CAT), or paraoxonase 1 (PON1) in the metabolism of highly toxic OP oxon metabolites, are well-described in the defense against ROS [12;14].

There is also evidence that genetic susceptibility plays a role in prostate cancer development. For example, twin studies have estimated that 42% of prostate cancer risk may be explained by genetic factors [15]. Recent genome-wide scans of prostate cancer have also identified high susceptibility loci in various gene regions [16;17]. Although none of the genome-wide scans to date have identified susceptibility loci in XMEs, these studies have not addressed the complex interactions that XMEs have with relevant exposures. Thus, in order to fully understand the relationship between pesticide use, genetic susceptibility in XMEs, and prostate cancer risk, it is important to investigate the role of the interaction between pesticides and genetic polymorphisms. Only one study has considered the joint contribution of two single nucleotide polymorphisms (SNPs) in XME genes and pesticide use on subsequent risk of prostate cancer; this study observed an elevated but non-significant risk among carriers of the allele variants [18].

In this study we evaluated the interaction between pesticide use and 1,913 SNPs in genes that code XMEs, and risk of prostate cancer in 2,220 AHS participants.


Study population

The AHS is a prospective cohort study that includes 55,747 male licensed pesticide applicators in Iowa and North Carolina. Applicators were recruited from 1993 through 1997; a detailed description of this cohort has been previously published [19]. During a follow-up telephone interview conducted in 1999–2003, applicators were asked for a mouthwash rinse sample to provide DNA from buccal cells. Approximately 72% of all applicators who completed the follow-up interview returned a buccal sample. In addition, applicators with incident prostate cancer who did not return a sample at follow-up were asked separately to provide a mouthwash rinse sample, with 307/561 (55%) returning a sample. Men diagnosed with incident prostate cancer between 1993 and 2004 who also provided a buccal cell sample were included in the current nested-case-control study. Eligibility, inclusion and exclusion criteria have been previously described [20]. Breifly, cancer cases were coded using the International Classification of Diseases for Oncology, 2nd edition, and stage (local, regional, distant, unstaged) and grade (well differentiated, moderately differentiated, poorly differentiated, undifferentiated, missing) were abstracted by the state cancer registries in Iowa and North Carolina. Eligible controls were frequency matched 2:1 to cases by date of birth (+/− 1 year). Controls were male applicators who provided buccal cell material, were alive and not lost to follow-up at the time of case diagnosis, and had no previous cancer diagnosis except non-melanoma skin cancer. All participants for the nested case-control study are white. Based on these inclusion criteria, 841 cases (66% of total white cases in the cohort as of 2004) and 1,659 controls were identified (total N= 2,500). Due to genotyping space limitations 164 controls were excluded. Of the remaining samples, 108 were removed due to insufficient or poor DNA quality (N=20; 14 cases, 6 controls) or <90% completion rate (i.e. more than 10% of the SNP assays failed for a given sample, N=88; 47 cases, 41 controls). We further identified 5 individuals who were suspected to be non-white (<80% European ancestry using STRUCTURE software [21] or significant deviation from the first two components in principal components analysis [22]) leaving a final sample size of 776 cases and 1,444 controls. Informed consent was obtained and the study protocol was reviewed by all relevant Institutional Review Boards.

Genotyping, Gene/SNP selection, and Quality Control

DNA was extracted from buccal cells using the Autopure protocol (QIAGEN). Genotyping was performed at NCI’s Core Genotyping Facility ( [23], using the Custom Infinium® BeadChip Assays (iSelect) from Illumina Inc. as part of a collaborative genotyping effort of an array of 26,512 SNPs in 1,291 candidate genes. TagSNPs were chosen to cover candidate genes for the three ancestry populations [Caucasian (CEU), Japanese Tokyo (JPT) + Chinese Beijing (CHB) and Yoruba people of Ibadan, Nigeria (YRI)] in the HapMap Project (Data Release 20/Phase II, NCBI Build 36.1 assembly, dbSNPb126) to allow use of this custom iSelect panel for studies containing different ethnic populations. TagSNPs were chosen using a modified version of the method described by Carlson et al. [24] as implemented in the Tagzilla ( software package. For each original candidate gene, SNPs within the region spanning 20 kb 5′ of the start of transcription to 10 kb 3′ of the end of the last exon were grouped using a binning threshold of r2=0.80, and tagSNPs chosen from these bins.

From the available data, we selected 149 candidate genes in the xenobiotic metabolism pathway, defined to include genes known to play a role in the biotransformation of xenobiotic substrates, including pesticides. Selected genes include those that encode phase I and phase II enzymes, receptors mediating the induction of xenobiotic biotransforming enzymes, and key enzymes in the regulation of the intracellular redox environment (oxidative stress). Additional genotyping for two glutathione transferase variants responsible for key conjugation reactions was also conducted; copy number variation assays for GSTM1 (Ex4+10+>-) and GSTT1 (Ex5–49+>-) deletions were performed separately using Applied Biosystems TaqMan® SNP Genotyping Assays. As a result, a total of 2,192 SNPs in 149 candidate genes in the xenobiotic metabolism pathway were identified. Exclusion for SNPs with low completion rate (<90% of samples) and SNPs showing evidence of deviation from Hardy–Weinberg proportions (p<1 x10−6) were made. SNPs with MAF<0.05 in the nested case-control samples were further excluded for analysis due to limited power. The overall genotyping rate was between 96% and 100% in the resultant N=1,913 SNPs in 149 candidate genes (Supplemental Table 1). Blinded duplicate samples (2%) were also included and concordance of these samples ranged from 96–100%.

Statistical Analysis

Unconditional logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for the association between SNPs and prostate cancer, pesticides and prostate cancer and the interaction between SNPs, pesticide use and prostate cancer risk. For SNP analyses, genotypes were coded as counts of the risk allele assuming a log-additive model and models were adjusted for age (10 yr-intervals) and state (Iowa or North Carolina). Exposure to pesticides was classified from responses to two self-administered questionnaires that were completed at enrollment. These questionnaires collected comprehensive data on lifetime use of 50 pesticides; pesticides with a prevalence of use less than 5% in the current nested case-control subgroup were excluded leaving 45 pesticides for analysis (17 herbicides, 21 insecticides, 2 fumigants, and 5 fungicides). Participants were asked how many years they applied each chemical (1 year or less, 2–5, 6–10, 11–20, 21–30, or more than 30 years) and how many days the applicator personally used it in an average year (less than 5, 5–9, 10–19, 20–39, 40–59, 60–150, or more than 150 days). Pesticides were categorized by lifetime exposure days (years of use × days per year) in this analysis from AHS data release version P1REL0712.04. Lifetime exposure days were categorized as non-exposed, low, and high exposed to a given chemical using the median cut-point based on distribution of lifetime days among cases and controls together for each chemical. Pesticide models were adjusted for confounders of the pesticide-prostate cancer relationship and include age, state and family history of prostate cancer in first degree relative (no, yes, missing).

SNP-pesticide interactions were examined using a multiplicative model. The p-value for each SNP-pesticide interaction was computed by comparing nested models with and without the cross-product terms using a likelihood ratio test. SNP-pesticide interactions with p-interaction≤0.01 were carried further for exploration with stratification by genotype using a dominant model for increased power. The top five interaction results, regardless of the direction of the association, from each XME pathway are presented. We also hypothesized that pesticides increase the risk of prostate cancer; thus in order to highlight the most interesting findings, additional stratified results (with p-interaction≤0.01) are presented if a monotonic increasing risk across pesticide exposure strata was evident. All interaction models were adjusted for age and state. Additional factors that were examined but ultimately not considered in the modeling because they did not change point estimates by more than 10% were family history of prostate cancer in first degree relative (no, yes, missing), type of applicator (private or commercial), and other pesticide adjustment based on the correlations between selected pesticides. We were not able to explore aggressive prostate cancer alone due to small numbers.

We applied the false discovery rate (FDR) (Benjamini–Hochberg adjustment) method to account for the expected proportion of false discoveries [25]. FDR values were calculated separately for each pesticide and separately from the results of 1,913 tests (i.e., total number of SNPs studied) in the evaluation of the association between each SNP-pesticide interaction and the risk of prostate cancer. Interactions were deemed noteworthy at an FDR = 0.20 level.


Cases and controls selected for the nested case-control study were similar to all prostate cancer cases from the cohort in age, state of residence, applicator type, presence of familial prostate cancer, and in prostate cancer disease characteristics for cases (data not shown) [20]. A list of all 149 genes and the number of SNPs in each gene identified in the xenobiotic metabolizing pathway is presented in Supplemental Table 1.

We examined SNPs involved in regulating a number of enzymes, including 1,203 in phase I/phase II enzymes, 61 in key xenobiotic receptors, and 649 in key enzymes that regulate the intracellular redox environment Table 1 shows a summary of the SNP results, as well as the interaction results. The number of observed main effects of SNPs in the oxidative stress pathway (at the α=0.01 level) was slightly greater than the number expected by chance (1% of the 649 SNPs examined in this pathway); in contrast, we did not observe any departure from the expected numbers of main effects for SNPs in Phase I/Phase II enzyme genes or in receptor genes. When crossed with 45 pesticides, the number of observed interactions was comparable to the number expected by chance alone (at the α=0.01 level).

Table 1
Summary results table for observed and expected single nucleotide polymorphism associations and pesticide interactions in xenobiotic metabolizing enzymes.

Associations between SNPs in XME genes and prostate cancer that were statistically significant at the p ≤ 0.01 level are presented in Table 2. P-values for all 1,913 SNPs are presented in Supplemental Table 2. The strongest association was seen for rs933271 in the oxidative stress gene thioredoxin reductase 2 (TXNRD2) with ORper allele = 1.33 (95% CI: 1.16, 1.52), p-trend = 4.62×10−5. Five additional SNPs in this gene were associated with prostate cancer, rs5993882 (p-trend=0.0014), rs5746847 (p-trend=0.0036), rs4485648 (p-trend=0.0065), rs9606186 (p-trend=0.0078), rs6518591 (9-trend=0.0086). After simultaneous adjustment for these five TXNRD2 SNPs, only rs933271 remained significant (p=0.0075), suggesting a single independent signal. Associations between several other oxidative stress genes and prostate cancer were observed. Two SNPs in myeloperoxidase (MPO) were significantly associated with prostate cancer, rs11079344, ORper allele = 1.59 (95% CI: 1.22, 2.08) and rs8178406, ORper allele = 1.21 (95% CI: 1.07, 1.37). Two SNPs in microsomal epoxide hydrolase 1 (EPHX1) were also significantly associated with prostate cancer, rs2740168 (p-trend=0.0042) and rs2292566, ORper allele = 1.27 (95% CI: 1.07, 1.51). Additional associations in phase I/phase II enzyme genes including CYP2C18, and CYP2C19 were also observed.

Table 2
Selected results (p-trend ≤ 0.01) for the association between prostate cancer and SNPs in xenobiotic metabolizing enzyme genes in the AHS PNCC

Associations between pesticide use and prostate cancer are presented in Table 3. Although no statistically significant positive associations between pesticides and prostate cancer were observed, there was suggestive evidence of increased risk (OR greater than 1.0) with increasing number of days of use of petroleum oil/petroleum distillate used as herbicide, terbufos, fonofos, phorate, and methyl bromide. Among chemicals with a lower prevalence of use (<30 %) in the nested case-control study (dieldrin, lindane, 2,4,5-T and others), some significant (p-trend<0.05) inverse associations were observed. A list of all 45 pesticides, their prevalence of use, and the median level of lifetime days of use of each chemical is presented in Supplemental Table 3.

Table 3
Odds ratios and 95% CI for the association between prostate cancer and 45 pesticides in the AHS PNCC

Stratified odds ratios for the association between pesticides and prostate cancer for the top five interactions from each XME pathway are presented in Table 4. In general, the p-values for SNPs in Phase I/II enzymes and for oxidative stress SNPs were smaller compared with those for receptors. Of the fifteen interaction results presented, fourteen were qualitative (involving increased risk with exposure in one genotype stratum and decreased risk in the other) or showed a significant protective association with prostate cancer. However, one interaction showed evidence for a positive interaction with a monotonic trend of increasing risk among variant allele (minor allele) carriers; the risk of prostate cancer associated with low petroleum oil/petroleum distillate use was 1.6 times those with no use (OR=1.6, 95% CI: 0.9, 2.9) and for high petroleum oil/petroleum distillate use was 3.7 times those with no use (OR=3.7, 95% CI: 1.9–7.3) among men carrying at least one variant allele in the glutamate-cysteine ligase (GCLC) SNP rs1883633. None of the interactions presented in Table 4 were noteworthy after adjustment using the FDR method.

Table 4
Stratified odds ratios and 95% CI, adjusted for age and state, for associations between pesticides and prostate cancer for top five interactions from each pathway

In Table 5, we present all results that show a positive monotonic trend with a p-interaction for pesticides and prostate cancer <0.01, regardless of pathway. Several SNP-pesticide combinations in oxidative stress genes and phase I/phase II enzyme genes displayed this pattern, but none in the receptor pathway, although fewer tests were conducted for receptor SNPs. In addition to the positive interaction between petroleum oil/petroleum distillate and rs1883633, in the oxidative stress gene GCLC (p-interaction=1.0×10−4), other notable pesticide interactions among variants which also showed independent associations with prostate cancer were observed. Among men carrying the variant allele for TXNRD2 rs4485648, EPHX1 rs17309872, or MPO rs11079344, increased prostate cancer risk was observed with high compared to no petroleum oil/petroleum distillate (OR=1.9, 95% CI: 1.1–3.2, p-interaction=0.01), (OR=2.1, 95% CI: 1.1–4.0, p-interaction=0.01), or terbufos (OR=3.0, 95% CI: 1.5–6.0 p-interaction=2.0×10−3) use, respectively. None of the interactions presented in Table 5 were noteworthy after adjustment using the FDR method.

Table 5
Stratified odds ratios for the association between pesticides and prostate cancer stratified by oxidative stress and Phase I/Phase II enzyme genotype where p-interaction ≤ 0.01.


In this nested case-control study, we evaluated the interaction between genes coding for XMEs and pesticide use on the risk of prostate cancer. We used two approaches to present interaction results. The first approach highlights the top five interactions observed in each XME pathway (Table 4). Among these interactions, the majority showed biologically less plausible inverse associations, qualitative interactions, or unstable point estimates due to small cell counts. For the second approach, we presented interactions (Table 5) where at least one genotype stratum showed a significant positive pesticide effect on prostate cancer and a monotonic pattern across pesticide use strata (from low to high use). The interaction between petroleum oil/petroleum distillate and the promoter region SNP rs1883633, in GCLC was identified in both approaches. Several other notable interactions in oxidative stress and phase I/phase II enzyme genes based on the stratified pattern of pesticide use by genotype were observed, including several interactions with SNPs that also showed a main effect on prostate cancer in our study; however, none were deemed noteworthy at the FDR = 0.20 level. Furthermore, the number of observed interactions was comparable to the number expected by chance alone.

The GCLC gene is shown to have oxidative stress-responsive elements in the promoter/enhancer region [26;27] and polymorphisms that are associated with decreased GCLC expression are suggested to be important determinants of susceptibility to oxidative stress and DNA damage [28]. While this SNP was not independently associated with prostate cancer, there does appear to be a modifying effect of GCLC genotype on the association between petroleum oil/petroleum distillate use and prostate cancer. The use of petroleum oil/petroleum distillate has not been previously associated with prostate cancer in AHS analyses [29], however, there appears to be a non-significant but positive association with petroleum oil herbicide use in this subset of nested case-control participants. Historically, petroleum oils and distillates, hydrocarbons derived from petroleum, have been used as herbicides [30]. It is difficult to interpret the effect of the use of petroleum oil-based pesticides on prostate cancer in this study because of the wide variability in use and structure as well as the lack of specificity about its use in the questionnaire data. Nonetheless, several other interactions with this chemical were observed in oxidative stress pathway SNPs, implicating this mechanism in pesticide-related prostate carcinogenesis.

Other interactions with petroleum oil were also observed. We observed an increased risk of prostate cancer associated with increasing levels of petroleum oil use among variant allele carriers of the TXNRD2 SNP rs4485648, which also had an association with prostate cancer (SNP p-trend=0.0065). TXNRD2 is a key enzyme in the regulation of the intracellular redox environment [31]. Thus, polymorphisms in these enzymes may result in an imbalance in the oxidative stress/antioxidant status [32]. These oxygen radicals may cause damage to DNA and chromosomes, induce epigenetic alterations, and interact with oncogenes or tumor suppressor genes [33;34] and increase prostate cancer risk. It is important to consider, however, that alternative mechanisms might explain the increased prostate cancer risk observed with TXNRD2. This gene partially overlaps the catechol-O-methyltransferase (COMT) gene on chromosome 22 that has been well described in androgen and estrogen metabolism [35;36]. The role of androgen biosynthesis and metabolism in prostate cancer growth, proliferation, and progression is well established [37] and the putative role of pesticides as endocrine disruptors [38], suggests that pesticides may act via a hormonal mechanism to influence prostate cancer development.

Several of the pesticides with significant interactions in this study have been previously linked to prostate cancer in the AHS. For example, use of the fumigant methyl bromide was significantly associated with prostate cancer risk in AHS applicators in an early analysis of the cohort [29]. Also, use of the organophosphate insecticides fonofos and terbufos has been linked with excesses of prostate cancer in applicators among those with a family history of prostate cancer [29;39;40]. The association between terbufos and prostate cancer is of particular interest since it is still registered for use and has a relatively high prevalence of ever use (~40% in AHS cohort), unlike fonofos which is no longer registered for use and methyl bromide which has a low prevalence of use. In the present study terbufos use was associated with prostate cancer in the presence of the variant genotype of the MPO promoter SNP rs11079344. This SNP was also significantly associated with prostate cancer in the current study. MPO is an oxidative stress gene that codes for myeloperoxidase which has been described to influence cancer risk in response to xenobiotics, including pesticides [4143]. Another MPO promoter region SNP which was not genotyped here, with a variant resulting in decreased MPO expression [44], has been extensively studied in lung, bladder, and breast cancer [45]. Thus, our results suggest that when we consider these two potential risk factors together, the risk of prostate cancer may be potentiated.

In addition to variants in oxidative stress related genes, SNPs in Phase I/Phase II enzyme genes showed positive associations with prostate cancer. For example, among men carrying the variant allele in the EPHX1 gene rs2292566, the risk of prostate cancer in the highest category of petroleum oil use was 2.1 times those with no use. In addition, the independent association of rs2292566 appears to be an important risk factor for prostate cancer in this population. Epoxide hydrolase functions in both the activation and detoxification of epoxides and low EPHX1 activity has been associated with increased cytogenetic damage in the presence of pesticide exposure [3;41]. Variations in direct pesticide metabolizing enzymes, including those in the CYP2C family appear also to be important genetic markers of prostate cancer susceptibility in this study. Although we did not observe any positive interactions, several SNP associations were observed in CYP2C18 and CYP2C19. A variety of organophosphate and carbamate compounds have been shown to be directly metabolized by phase I enzymes CYP2C8, CYP2C9, CYP2C18, and CYP2C19 [46]. Thus, polymorphisms in the genes that code for these Phase I/Phase II enzymes may alter levels of toxic pesticide intermediates and influence prostate cancer risk.

Several strengths and limitations of our study should be recognized. High quality genotype and pesticide information is available in the AHS. For many gene-exposure studies, the key limitation is the quality of the exposure information. The quality of information on pesticide use among AHS participants is high; self-reported pesticide use information has been found to be reliable in this cohort [47;48]. Furthermore, the ability of the AHS to look at individual pesticides rather than groups (herbicides or insecticides or chemical classes) is critical because observed cancer risks appear to be chemical-specific. Although the numbers within some strata were small, to our knowledge there are no other studies with more power to examine this interaction. In addition, we considered many interactions and none of the observed interactions were deemed noteworthy at the FDR=0.20 level; it is possible, however, that multiple testing issues have masked any true interactions present in this dataset. The numbers of interactions were what we might have expected by chance alone although some interesting patterns were observed. Similarly, we observed interesting SNP associations, however, the number of significant associations for SNPs in the oxidative stress genes was only slightly greater than would be expected by chance, while the number of significant associations for SNPs in Phase I/Phase II enzyme genes and in receptor genes were as expected. Also, we were not able to examine disease aggressiveness due to small numbers. Finally, all participants in this study were white, which limits the generalizability of the results to other racial/ethnic groups.

In conclusion, we observed several positive interactions in oxidative stress and phase I/phase II enzyme genes. None, however, were deemed noteworthy at the FDR=0.20 level and the overall number of observed interactions was comparable to the number expected by chance alone. Nonetheless, some XMEs did modify the association between pesticide use and prostate cancer. Interactions of pesticides and SNPs in oxidative stress genes and phase I/phase II enzyme genes implicate these mechanisms in particular. More evidence for the pesticide-prostate association is needed to fully explain the mechanisms by which pesticides might influence cancer risk.

Supplementary Material


This research was supported by the Intramural Research Program of the NIH, National Cancer Institute, Division of Cancer Epidemiology and Genetics (Z01CP010119) and National Institute of Environmental Health Sciences (Z01ES049030). KHB was supported by National Cancer Institute grant T32 CA105666. We thank the participants in the Agricultural Health Study for their contributions in support of this research.


Single Nucleotide Polymorphism
Xenobiotic Metabolizing Enzyme
Odds Ratio
Confidence Interval
Agricultural Health Study
Prostate Nested Case-Control Study
False Discovery Rate

Reference List

1. Koutros S, Alavanja MC, Lubin JH, Sandler DP, Hoppin JA, Lynch CF, et al. An update of cancer incidence in the Agricultural Health Study. J Occup Environ Med. 2010;52(11):1098–105. [PMC free article] [PubMed]
2. Rose RL, Tang J, Choi J, Cao Y, Usmani A, Cherrington N, et al. Pesticide metabolism in humans, including polymorphisms. Scand J Work Environ Health. 2005;31 (Suppl 1):156–63. [PubMed]
3. Parkinson A, Ogilvie BW. Biotransformation of Xenobiotics. In: Klaassen CD, editor. Casarett and Doull’s Toxicology: The Basic Science of Poisons. 7. The McGraw-Hill Companies, Inc; 2008. pp. 161–304.
4. Hodgson E. In vitro human phase I metabolism of xenobiotics I: pesticides and related compounds used in agriculture and public health, May 2003. J Biochem Mol Toxicol. 2003;17(4):201–6. [PubMed]
5. Hodgson E, Rose RL. Organophosphorus chemicals: potent inhibitors of the human metabolism of steroid hormones and xenobiotics. Drug Metab Rev. 2006;38(1–2):149–62. [PubMed]
6. Rodriguez-Antona C, Ingelman-Sundberg M. Cytochrome P450 pharmacogenetics and cancer. Oncogene. 2006;25(11):1679–91. [PubMed]
7. Tamasi V, Vereczkey L, Falus A, Monostory K. Some aspects of interindividual variations in the metabolism of xenobiotics. Inflamm Res. 2003;52(8):322–33. [PubMed]
8. Usmani KA, Cho TM, Rose RL, Hodgson E. Inhibition of the human liver microsomal and human cytochrome P450 1A2 and 3A4 metabolism of estradiol by deployment-related and other chemicals. Drug Metab Dispos. 2006;34(9):1606–14. [PubMed]
9. Casabar RC, Das PC, Dekrey GK, Gardiner CS, Cao Y, Rose RL, et al. Endosulfan induces CYP2B6 and CYP3A4 by activating the pregnane X receptor. Toxicol Appl Pharmacol. 2010;245(3):335–43. [PubMed]
10. Casado S, Alonso M, Herradon B, Tarazona JV, Navas J. Activation of the aryl hydrocarbon receptor by carbaryl: Computational evidence of the ability of carbaryl to assume a planar conformation. Environ Toxicol Chem. 2006;25(12):3141–7. [PubMed]
11. Pascussi JM, Gerbal-Chaloin S, Drocourt L, Maurel P, Vilarem MJ. The expression of CYP2B6, CYP2C9 and CYP3A4 genes: a tangle of networks of nuclear and steroid receptors. Biochim Biophys Acta. 2003;1619(3):243–53. [PubMed]
12. Garrett NE, Stack HF, Waters MD. Evaluation of the genetic activity profiles of 65 pesticides. Mutat Res. 1986;168(3):301–25. [PubMed]
13. Soltaninejad K, Abdollahi M. Current opinion on the science of organophosphate pesticides and toxic stress: a systematic review. Med Sci Monit. 2009;15(3):RA75–RA90. [PubMed]
14. Lopez O, Hernandez AF, Rodrigo L, Gil F, Pena G, Serrano JL, et al. Changes in antioxidant enzymes in humans with long-term exposure to pesticides. Toxicol Lett. 2007;171(3):146–53. [PubMed]
15. Lichtenstein P, Holm NV, Verkasalo PK, Iliadou A, Kaprio J, Koskenvuo M, et al. Environmental and heritable factors in the causation of cancer--analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med. 2000;343(2):78–85. [PubMed]
16. Thomas G, Jacobs KB, Yeager M, Kraft P, Wacholder S, Orr N, et al. Multiple loci identified in a genome-wide association study of prostate cancer. Nat Genet. 2008;40(3):310–5. [PubMed]
17. Witte JS. Multiple prostate cancer risk variants on 8q24. Nat Genet. 2007;39(5):579–80. [PubMed]
18. Multigner L, Ndong JR, Giusti A, Romana M, Delacroix-Maillard H, Cordier S, et al. Chlordecone exposure and risk of prostate cancer. J Clin Oncol. 2010;28(21):3457–62. [PubMed]
19. Alavanja MC, Sandler DP, McMaster SB, Zahm SH, McDonnell CJ, Lynch CF, et al. The Agricultural Health Study. Environ Health Perspect. 1996;104(4):362–9. [PMC free article] [PubMed]
20. Koutros S, Beane Freeman LE, Berndt SI, Andreotti G, Lubin JH, Sandler DP, et al. Pesticide use modifies the association between genetic variants on chromosome 8q24 and prostate cancer. Cancer Res. 2010;70(22):9224–33. [PMC free article] [PubMed]
21. Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000;155:945–59. [PubMed]
22. Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet. 2006;38:904–9. [PubMed]
23. Packer BR, Yeager M, Burdette L, Welch R, Beerman M, Qi L, 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. [PMC free article] [PubMed]
24. Carlson CS, Eberle MA, Rieder MJ, Yi Q, Kruglyak L, Nickerson DA. Selecting a maximally informative set of single-nucleotide polymorphisms for association analyses using linkage disequilibrium. Am J Hum Genet. 2004;74(1):106–20. [PubMed]
25. Benjamini YHY. Controlling the false discovery rate: a practical and powerful approach to multiple testing. 1995. pp. 289–300.
26. Galloway DC, Blake DG, McLellan LI. Regulation of gamma-glutamylcysteine synthetase regulatory subunit (GLCLR) gene expression: identification of the major transcriptional start site in HT29 cells. Biochim Biophys Acta. 1999;1446(1–2):47–56. [PubMed]
27. Mulcahy RT, Wartman MA, Bailey HH, Gipp JJ. Constitutive and beta-naphthoflavone-induced expression of the human gamma-glutamylcysteine synthetase heavy subunit gene is regulated by a distal antioxidant response element/TRE sequence. J Biol Chem. 1997;272(11):7445–54. [PubMed]
28. Shi S, Hudson FN, Botta D, McGrath MB, White CC, Neff-LaFord HD, et al. Over expression of glutamate cysteine ligase increases cellular resistance to H2O2-induced DNA single-strand breaks. Cytometry A. 2007;71(9):686–92. [PubMed]
29. Alavanja MC, Samanic C, Dosemeci M, Lubin J, Tarone R, Lynch CF, et al. Use of agricultural pesticides and prostate cancer risk in the Agricultural Health Study cohort. Am J Epidemiol. 2003;157(9):800–14. [PubMed]
30. Ashton FM, Monaco TJ. Weed Science. 3. Wiley & Sons, Inc; 1991. p. 320.
31. Powis G, Montfort WR. Properties and biological activities of thioredoxins. Annu Rev Pharmacol Toxicol. 2001;41:261–95. [PubMed]
32. Chen Y, Cai J, Jones DP. Mitochondrial thioredoxin in regulation of oxidant-induced cell death. FEBS Lett. 2006;580(28–29):6596–602. [PMC free article] [PubMed]
33. Arsova-Sarafinovska Z, Eken A, Matevska N, Erdem O, Sayal A, Savaser A, et al. Increased oxidative/nitrosative stress and decreased antioxidant enzyme activities in prostate cancer. Clin Biochem. 2009;42(12):1228–35. [PubMed]
32. Donkena KV, Young CY, Tindall DJ. Oxidative stress and DNA methylation in prostate cancer. Obstet Gynecol Int. 2010;2010:302051. [PMC free article] [PubMed]
35. Platz EA, Giovannucci E. The epidemiology of sex steroid hormones and their signaling and metabolic pathways in the etiology of prostate cancer. J Steroid Biochem Mol Biol. 2004;92(4):237–53. [PubMed]
36. Ross RK, Pike MC, Coetzee GA, Reichardt JK, Yu MC, Feigelson H, et al. Androgen metabolism and prostate cancer: establishing a model of genetic susceptibility. Cancer Res. 1998;58(20):4497–504. [PubMed]
37. Hsing AW, Chu LW, Stanczyk FZ. Androgen and prostate cancer: is the hypothesis dead? Cancer Epidemiol Biomarkers Prev. 2008;17(10):2525–30. [PubMed]
38. Soto AM, Sonnenschein C. Environmental causes of cancer: endocrine disruptors as carcinogens. Nat Rev Endocrinol. 2010;6(7):363–70. [PubMed]
39. Bonner MR, Williams BA, Rusiecki JA, Blair A, Beane Freeman LE, Hoppin JA, et al. Occupational exposure to terbufos and the incidence of cancer in the Agricultural Health Study. Cancer Causes Control. 2010;21(6):871–7. [PMC free article] [PubMed]
40. Mahajan R, Blair A, Lynch CF, Schroeder P, Hoppin JA, Sandler DP, et al. Fonofos exposure and cancer incidence in the agricultural health study. Environ Health Perspect. 2006;114(12):1838–42. [PMC free article] [PubMed]
41. Costa C, Teixeira JP, Silva S, Roma-Torres J, Coelho P, Gaspar J, et al. Cytogenetic and molecular biomonitoring of a Portuguese population exposed to pesticides. Mutagenesis. 2006;21(5):343–50. [PubMed]
42. Huang SK, Chiu AW, Pu YS, Huang YK, Chung CJ, Tsai HJ, et al. Arsenic methylation capability, myeloperoxidase and sulfotransferase genetic polymorphisms, and the stage and grade of urothelial carcinoma. Urol Int. 2009;82(2):227–34. [PubMed]
43. Silveira VS, Canalle R, Scrideli CA, Queiroz RG, Tone LG. Role of the CYP2D6, EPHX1, MPO, and NQO1 genes in the susceptibility to acute lymphoblastic leukemia in Brazilian children. Environ Mol Mutagen. 2010;51(1):48–56. [PubMed]
44. Piedrafita FJ, Molander RB, Vansant G, Orlova EA, Pfahl M, Reynolds WF. An Alu element in the myeloperoxidase promoter contains a composite SP1-thyroid hormone-retinoic acid response element. J Biol Chem. 1996;271(24):14412–20. [PubMed]
45. Chu H, Wang M, Wang M, Gu D, Wu D, Zhang Z, et al. The MPO -463G>A polymorphism and cancer risk: a meta-analysis based on 43 case-control studies. Mutagenesis. 2010;25(4):389–95. [PubMed]
46. Usmani KA, Karoly ED, Hodgson E, Rose RL. In vitro sulfoxidation of thioether compounds by human cytochrome P450 and flavin-containing monooxygenase isoforms with particular reference to the CYP2C subfamily. Drug Metab Dispos. 2004;32(3):333–9. [PubMed]
47. Blair A, Tarone R, Sandler D, Lynch CF, Rowland A, Wintersteen W, et al. Reliability of reporting on life-style and agricultural factors by a sample of participants in the Agricultural Health Study from Iowa. Epidemiology. 2002;13(1):94–9. [PubMed]
48. Hoppin JA, Yucel F, Dosemeci M, Sandler DP. Accuracy of self-reported pesticide use duration information from licensed pesticide applicators in the Agricultural Health Study. J Expo Anal Environ Epidemiol. 2002;12(5):313–8. [PubMed]