Influence of lifestyle choices on risks of CYP1B1 polymorphisms for prostate cancer

Abstract Cytochrome P450 1B1 (CYP1B1) converts xenobiotics to carcinogens and how lifestyle choices may interact with CYP1B1 polymorphisms and affect prostate cancer risk was assessed. Blood genomic DNA from a Caucasian population was analysed at polymorphic sites of the 5′ untranslated region of CYP1B1 using TaqMan genotyping assays. Overall, drinker status and minor alleles at rs2551188, rs2567206 and rs10175368 were associated with prostate cancer. Linkage was observed between rs2551188, rs2567206, rs2567207 and rs10175368, and the G‐C‐T‐G haplotype (major allele at respective sites) was decreased in cancer. Interestingly when classified by lifestyle factors, no associations of genotypes were found for non‐smokers and non‐drinkers, whereas on the contrary, minor type at rs2567206 and rs10175368 increased and major G‐C‐T‐G decreased risk for cancer among smokers and drinkers. Interestingly, rs2551188, rs2567206 and rs10175368 minor genotypes correlated with increased tissue CYP1B1 as determined by immunohistochemistry. Further, rs10175368 enhanced luciferase activity and mobility shift show stronger binding of nuclear factor for the minor allele. These results demonstrate smoking and alcohol consumption to modify the risks of CYP1B1 polymorphisms for prostate cancer which may be through rs10175368, and this is of importance in understanding their role in the pathogenesis and as a biomarker for this disease.

identifying risks in the carcinogenesis process is an important step towards its prevention.
Lifestyle factors such as tobacco smoking and alcohol consumption are established risks for various types of cancers. 3,4 Worldwide, tobacco smoking accounts for roughly 21% of cancer deaths with 29% in high-income countries. 3 In the USA in the year 2010, the estimated death rate of all cancers due to cigarette smoking was roughly 38% with about 112 000 deaths among men aged 35 years or older and does not include additional deaths from environmental tobacco smoke or usage of cigars, pipes or smokeless tobacco. 5 In prostate, a meta-analysis of 4 million cohort participants showed that current cigarette smoking was correlated with increased risk of cancer death (relative risk [RR]; 1.24, 95% confidence interval [CI]; 1.18-1.31) with cigarettes smoked per day having a dose-response association with cancer mortality. 6 Also, compared with non-smokers, former smokers (hazard ratio [HR]; 1.63, 95% CI; 1.30-2.04, P < .001) and current smokers (HR; 1.80, 95% CI 1.45-2.24, P < .001) had a higher risk of prostate cancer biochemical recurrence. 7 Alcohol consumption accounts for about 5% of all cancer deaths worldwide and a large proportion of cancers in Europe and America. 3 In the USA, 92% of respondents 18 years and older claimed life-time alcohol usage 8 and up to 3.7% of cancer deaths were linked to drinking in the USA. 9 In a meta-analyses study, Bagnardi et al 4 find alcohol drinking to be associated with various cancers and this effect is strongest among heavy drinkers. The effect of alcohol drinking appears to be dose-dependent as light to moderate drinking resulted in a lower risk, whereas heavy drinking caused an increased risk of certain cancers. 10 As for prostate, a dose-response was also observed for cancer risk among current drinkers (P trend < .01). 11 Tobacco smoking and alcohol thus play a causative role in the carcinogenesis process, and a major enzyme that affects this process is cytochrome P450 (CYP) 1B1. CYP1B1 is a member of the CYP superfamily involved in phase I metabolism of many xenobiotics. 12,13 CYP1B1 can metabolically convert tobacco smoke pro-carcinogens such as polycyclic aromatic hydrocarbons to reactive or carcinogenic intermediates 13,14 and result in DNA adduct formation. 14,15 In prostate, mRNA transcripts including CYP1B1 were observed along with DNA adducts after incubating with 2-amino-1-methyl-6-phenylimiazo [4,5-b]pyridine (PhIP), 2-amino-3-methylimidazo [4,5-f]quinoline (IQ) and benzo[a]pyrene (B[a]P). 14 Also in primary mammary epithelial cells, B[a]P caused DNA adducts to form as well as CYP1B1 induction. 15 Alternatively, PAHs or smoking can enhance levels of CYP1B1 expression. 13,16 As CYP1B1 is expressed in human prostate, 17 gene up-regulation and activation of pro-carcinogens may thus be influential in the prostatic carcinogenesis process.
The main form of alcohol in alcoholic beverages is ethanol, which may pose a risk even at moderate drinking amounts. 18 Levels of ethanol in blood were shown to dramatically rise above 15 mmol/L within 30 minutes after drinking whisky (0.72 g/kg ethanol) and gradually decreased over a 6 hour period. 19 Although the tumorigenicity of ethanol itself may be dependent on experimental conditions, its direct metabolic product acetaldehyde has been shown to be carcinogenic in animals. 20 Indeed, CYP1B1 was shown to metabolize ethanol into significant amounts of acetaldehyde 21 and studies in rats show inhalation and oral administration of acetaldehyde to be carcinogenic in animals. 22,23 Acetaldehyde can interfere with DNA   synthesis and repair, 24 cause point mutations, 20,24 form direct bonds   with DNA 24 and form other DNA adducts at cellular concentrations. 24,25 The International Agency for Research on Cancer (IARC) has listed acetaldehyde as carcinogenic. 18 Additionally, acetaldehyde gets metabolized into acetate via acetaldehyde dehydrogenase and in the process, radicals are formed 24 that can bind to DNA. 26 It is thus apparent that CYP1B1 can lead to cancer by activating various compounds into carcinogenic forms and of interest are genetic polymorphisms that can alter enzyme levels. A number of studies have focused on the coding region or missense variants that enhance CYP1B1 catalytic activity and show them to be associated with prostate cancer. 27,28 Also of importance are variants in the promoter or 5 0 untranslated region (5 0 UTR) as these may lead to an upregulation 29 or down-regulation 30 of RNA transcription and consequentially, enzyme expression levels. A study on promoter polymorphisms and prostate cancer did indeed show variants to be associated with progression of cancer. 31 Promoter polymorphisms may thus be indicators of disease susceptibility or factors in polygenic diseases due to alterations in enzyme expression.
To date, studies on polymorphic variants of the CYP1B1 promoter region/5 0 UTR and their risks for prostate cancer and their functional role are lacking. Additionally, the impact of lifestyle factors on the risks of these variants for prostate cancer has never been investigated. In this report, we evaluated the risks of 8 polymorphic sites in the promoter region/5 0 UTR of CYP1B1 for prostate cancer and how this is influenced by major lifestyle factors among a Caucasian population. We have been suggested that minor genotypes and alleles are associated with cancer and that tobacco smoking or alcohol consumption can increase their risk. Also, we examined the functional effects of polymorphisms and have been suggested that minor alleles can affect promoter activity as well as correlate with protein expression in prostatic cells.

| CYP1B1 genotyping
To analyse CYP1B1 polymorphisms, TaqMan genotyping assays (Applied Biosystems, Foster City, CA) were utilized according to manufacturer's instructions. In brief, a 5-lL reaction containing TaqMan Universal Master Mix (Applied Biosystems) and 0.5 ng of sample DNA was prepared. Thermal cycle conditions were 95°C for 10 minutes, followed by 40 to 60 cycles of 95°C for 15 seconds and 60°C for 1 minute. End-point fluorescent readings were analysed using the QuantStudio 7 Real-Time PCR system (Applied Biosystems). The polymorphic sites analysed were in the promoter region/5 0 UTR of the CYP1B1 gene which were reported in dbSNP (http://www.ncbi.nlm. nih.gov/projects/SNP/) and ID# (base change, distance prior to ATG start site) are as follows: rs2551188 (G to A, À263 bp), rs2567206 (C to T, À1001 bp), rs2567207 (T to C, À1112 bp), rs162556 (T to C, À3 924 bp), rs10175368 (G to A, À5331 bp), rs163090 (T to A, À11 102 bp), rs162330 (T to G, À16966 bp) and rs162331 (A to G, À170 64 bp).

| Cell culture
Caucasian prostate cancer PC3 and DU145 cell lines were obtained from American Type Cell Collection (ATCC, Manassas, VA, USA).
Cells were cultured in RPMI1640 medium supplemented with 10% foetal bovine serum in a humidified atmosphere containing 5% CO2.
These cell lines were authenticated by DNA short-tandem repeat analysis by ATCC, and experiments with cell lines were performed within 6 months of their revival.

| Immunohistochemical analysis
Immunostaining of CYP1B1 was performed on specimens of BPH.  (DAB) was added as chromogen followed by counterstaining with haematoxylin. Cellular expression levels were analysed by the intensity of positive cells using ImageJ software (http://rsb.info.nih.gov/ij) and ranked on an overall scale from 0 to 3, with 0 indicating the absence of staining; 1, weak staining; 2, moderate staining; and 3, strong staining.

| Site-directed mutagenesis and promoter reporter assay
Two sets of Gaussia luciferase CYP1B1 vectors consisting of the 5 0 UTR or promoter region at the following base pairs prior to the ATG start site: À1143 to +190 and customized À5590 to À5090 were utilized along with negative control (Gene Copoeia, Rockville, MD). These underwent site-directed mutagenesis using QuikChange   the minor allele T and A were also observed to be significantly increased in PC at rs2567206 (P = .008) and rs10175368 (P = .001),

| CYP1B1 polymorphisms and prostate cancer risk
respectively. Additionally, minor allele A at rs2551188 proved to be a risk for cancer (P = .043). No differences were found between patients with prostate cancer and control at other polymorphic sites.

| Linkage disequilibrium of CYP1B1 polymorphisms
Linkage between the polymorphic sites of CYP1B1 was calculated and

| Haplotype frequencies of CYP1B1 polymorphic sites
Haplotype frequencies of rs2551188-rs2567206-rs2567207-rs1017 5368 and rs163090-rs162330-rs162331 in patients with prostate cancer were calculated and results are shown in Table 4. The major haplotype was G-C-T-G for rs2551188-rs2567206-rs2567207-rs10175368, which was expressed in 70.7% of healthy individuals overall. Interestingly, G-C-T-G represents major allele at the respective rs sites and was significantly reduced in prostate cancer when compared to other haplotypes combined (P = .028). For rs163090-rs162330-rs162331, 2 haplotypes were similarly expressed and predominant, being T-T-G (47.7%) and A-G-A (46.6%) in overall controls. These haplotypes involving rs163090, rs162330 and rs162331, however, did not show any significant differences between cases and controls.

| CYP1B1 polymorphisms within clinical stage of prostate cancer patients
Prostate cancer samples were classified in terms of clinical stage.
There were 82 samples of unknown status. No statistical differences were observed when classified in terms of stage ≤2 (N = 253) vs ≥3 (N = 65) for all CYP1B1 polymorphic sites (data not shown).

| Influence of lifestyle factors on risks of CYP1B1 polymorphisms for prostate cancer
As lifestyle factors can affect risks of prostate cancer, interaction between choices and CYP1B1 polymorphisms were determined. Interaction between lifestyle choices and CYP1B1 haplotypes was also determined. The effect of smoker and drinker status on prostate cancer risks for haplotype frequencies of rs2551188-rs2567206-rs2567207-rs10175368 and rs163090-rs162330-rs162331 is shown in Table 4. Interestingly compared to healthy controls, major G-C-T-G of rs2551188-rs2567206-rs2567207-rs10175368 was significantly lower in cancer among smokers (P = .036) with a tendency for drinkers (P = .066), whereas no associations were observed in non-smokers and non-drinkers. Lifestyle factors did not influence the risks for any of the rs163090-rs162330-rs162331 haplotypes.

| CYP1B1 polymorphisms and protein expression among BPH specimens
As genotypes and haplotypes were observed to be a risk for prostate cancer, expression level of CYP1B1 protein was evaluated for all sites. Immunohistochemistry was performed on 83 BPH specimens and scored. Interestingly compared to major genotype, variants at rs2551188 (P = .015), rs2567206 (P =.016) and rs10175368 (P = .047) were determined to have increased CYP1B1 levels (Figure 1A). Other polymorphic sites, however, showed no correlations with CYP1B1 expression (data not shown).

| CYP1B1 polymorphisms and promoter activity
To assess functional properties, all polymorphic sites were evaluated for promoter activity by site-directed mutagenesis followed by luciferase activity. Although minor alleles of rs2551188 and rs2567206 had no effect, rs10175368. A significantly up-regulated CYP1B1 promoter activity compared to major allele G in both PC3 and DU145 cells (P < .001, Figure 1B). Other polymorphisms did not affect promoter activity (data not shown).

| Nuclear extract binding of polymorphic site
Promoter activity was enhanced due to polymorphism, and thus, the binding capability of nuclear extracts to polymorphic sites was analysed. Nuclear extract was observed to weakly bind to rs10175368 G allele motif but interestingly, a 2.1-fold larger amount of protein bound to the minor A allele form (P = .045, Figure 1C). Binding of nuclear extracts to motifs of rs2551188 and rs2567206 was not observed (data not shown). shown to enhance cancer of the lung, bladder, and head and neck. 12 Interestingly in prostate cancer tissue, Caucasian ever smokers had significantly higher PAH-DNA adducts compared to non-smokers. 32 When evaluating polymorphic sites of CYP1B1, the minor allele at rs2551188, rs2567206 and rs10175368 were an overall risk for    involving G-C at these respective sites demonstrated a reduced risk for prostate cancer, which is in agreement with our results.
Polymorphisms of rs2551188, rs2567206 and rs10175368 are thus determined to be a risk for prostate cancer and the mechanism by which they may play a role is not known. These sites are located in the promoter or 5 0 UTR which are of importance as variants in this region may lead to increased gene expression 29 and, consequently, increased enzyme or CYP1B1 levels. In concordance, results of this study demonstrate polymorphisms at these 3 sites to be associated with increased CYP1B1 protein levels as was observed in human prostatic specimens. On the contrary, only the rs10175368 minor allele showed increased luciferase activity and mobility shift demonstrated strong binding towards this variant. As these sites are linked, it may thus be through the minor allele of rs10175368 that expression levels of CYP1B1 are increased. This co-dependence with rs10175368 may be pertinent for rs2567206 as cancer risk for these sites was modified by smoker and alcohol drinker status. This is corroborated in smokers as Rotunno et al 34 observed rs10175368 to have increased mRNA expression among current smokers (P = .004) but not for never and former smokers. Thus, these lifestyle factors appear to interact at the genetic level to possibly increase CYP1B1 levels with rs10175368 playing a major role. Further experimentation is necessary to determine the identity of the factor that can bind to the rs10175368 minor allele in prostate cancer cells.
On the other hand, risk of the rs2551188 A allele was not affected by lifestyle factors and did not affect promoter activity or bind nuclear protein. Reasons for this independence are not known.
Unlike rs2567206 and rs10175368 that are in the gene promoter region, rs2551188 is located in the 5 0 UTR of intron 1. This site undergoes a G to A base change and studies suggest this transition to affect RNA stability, 42 which can consequentially lead to enhanced processing and gene or CYP1B1 expression. Alternatively, up-regulation may be caused by linkage with rs10175368.
In conclusion, polymorphisms of the promoter and 5 0 UTR region of CYP1B1 are determined to be a risk for prostate cancer that can be modified by lifestyle factors in Caucasian men. These polymorphisms that potentially are capable of increasing gene expression levels and more so due to rs10175368 are thus critical to the prostate cell as CYP1B1 plays a role in the activation of carcinogens from precursors from various sources such as tobacco smoke and alcohol. These findings thus suggest CYP1B1 and its polymorphisms as a potential biomarker and gene of importance in understanding the pathogenesis of prostate cancer. It is essential to note, however, that this study does have its limitations as total sample size consisted of 405 controls and 400 prostate cancer cases, and a larger validation study is needed to verify results in the future.

This study was supported by the United States National Cancer
Institute (R21CA185003 to YT); United States Department of Defense (W81XWH-04-1-0579 to YT); and United States Department of Veterans Affairs (Merit Review to YT).

CONFLI CT OF INTEREST
The authors confirm that there are no conflict of interests. YT wrote the manuscript. All authors read and approved the final manuscript.