UROLOGICAL ONCOLOGY Association of AKR1C3 Polymorphisms with Bladder Cancer N. Ozan Tiryakioğlu, Nagehan Ersoy Tunalı* Purpose: Polymorphisms in the genes coding for the carcinogen metabolizing enzymes may affect enzyme activ- ities and alter the activation and detoxification rates of the carcinogens. AKR1C3 is one of the very polymorphic xenobiotic metabolizing enzymes involved in the bioactivation process. Here we aimed to investigate the associ- ation of two single nucleotide polymorphisms in AKR1C3, rs12529 (c.15C > G) and rs1937920 (12259 bp 3’ of STP A > G) with urinary bladder cancer (UBC). Materials and Methods: Two-hundred fifty UBC cases and 250 control subjects were genotyped using the Pol- ymerase Chain Reaction and Restriction Fragment Length method. Associations of the genotypes with UBC risk and tumor characteristics were assessed using logistic regression and Fisher’s exact test. The results are corrected for multiple testing. Results: We identified strong associations between the studied AKR1C3 variants and UBC risk. The homozygous variant genotype of rs12529 was found to be inversely associated with UBC, and rs1937920 was shown to be as- sociated with increased risk of UBC. None of the genotypes were found to be significantly associated with tumor characteristics. Conclusion: We provided evidence that rs12529 and rs1937920 are significant in the molecular pathogenesis of UBC. However, the results presented here should be regarded as preliminary and might represent a first step of future larger studies aiming to better elucidate the role of AKR1C3 polymorphisms in the susceptibility to bladder cancer. Keywords: genetic association studies; humans; polymorphism; single nucleotide; urinary bladder neoplasms; genetics; 3-hydroxysteroid dehydrogenases; aryl hydrocarbon hydroxylases; aryl hydrocarbon receptor nuclear translocator. INTRODUCTION Urinary Bladder Cancer (UBC) is the ninth most common cancer worldwide, with approximate- ly 400.000 new cases each year.(1) Despite increased awareness against certain risk factors and recent promising developments in the diagnosis and treat- ment, worldwide incidence of UBC increases steadily. (2) Bladder cancer is three to five times higher in men than women, and its incidence increases with age for both sexes, peaking at the seventh decade.(3). The main molecular pathways leading to UBC development are altered xenobiotic metabolism and mutations in genes involved in DNA repair, tumor suppression and cell pro- liferation mechanisms.(4-8) The most common risk factor for UBC is cigarette smoking, followed by occupation- al exposure to aromatic amines used in the production of dyes, rubber and textiles and to polycyclic aromatic hydrocarbons (PAHs) formed during combustion of fossil and carbon-containing fuels.(9,10) These well-stud- ied and mutagenic carcinogens should undergo bioac- tivation in order to exert their carcinogenic potential. Alterations in the activation and following detoxifica- tion mechanisms of these carcinogenic chemicals result in the formation of DNA adducts, which eventually lead to DNA damage. This process constitutes the basis for genetic susceptibility to UBC.(4) Therefore, studies investigating genetic susceptibility to UBC focus on genes involved in xenobiotic metabolism and DNA re- pair. The ultimate function of the xenobiotic metabo- lism is detoxification of carcinogens, where chemical modifications catalyzed by Phase I enzymes result in the activation of carcinogens, and those catalyzed by Phase II enzymes facilitate the excretion of these com- pounds. Polymorphisms in the genes coding for the xe- nobiotic enzymes may induce altered enzyme activities that confer susceptibility to various cancers, including UBC, due to altered activation/detoxification rates of carcinogens.(3) Almost two thirds of all the UBC cases are shown to be associated with smoking. PAHs, tobac- co-specific N-nitrosamines and aromatic amines are the Department of Molecular Biology and Genetics, Haliç University, Istanbul, Turkey. *Correspondence: Department of Molecular Biology and Genetics, Haliç University, Istanbul, 34445 Turkey. Tel: +90 212 9242444. E-mail: nagehan.ersoy@gmail.com. Received January 2016 & Accepted March 2016 Urological Oncology 2615 Vol 13 No 02 March-April 2016 2616 significant contributors to the carcinogenic effects of tobacco smoke via DNA adducts or formation of reac- tive oxygen species (ROS) leading to oxidative damage and ultimately DNA mutations.(11-13) The balance be- tween the activation and detoxification of these carcino- gens by Phase I and Phase II enzymes affect the amount of accumulating DNA damage. Aldoketoreductases (AKRs) are Phase I enzymes that belong to a superfami- ly of NAD(P)H-linked oxidoreductases and catalyze the conversion of carbonyl group-containing xenobiotics to alcohols for conjugation reactions.(14) Since they have been shown to activate polycyclic aromatic hydrocar- bons (PAHs) and nitrosamines in exhaust and cigarette smoke, AKR family members are considered important for smoking-related cancers.(15) The human AKR genes (AKR 1, 6, 7) are highly polymorphic, where variant genotypes may cause alterations in respective protein functions, and act as modifiers of UBC risk. AKR1C3 on chromosome 10p15-p14 is known to act on various subtrates including hormones and PAHs.(12) In addition to that, in vitro studies have shown significant induction of AKR1C3 upon exposure to cigarette smoke conden- sate, diesel exhaust and PAHs.(13-16) On this background we hypothesized that genetic varia- tions in the AKR1C3 gene may contribute to UBC risk. AKR polymorphisms have been previously investigat- ed in a total of six studies for their possible involve- ment in various cancer susceptibilities,(17-22) however, there is only one study investigating the involvement of AKR1C3 polymorphisms in bladder cancer.(17) In this work we aimed to investigate the association of two sin- gle nucleotide polymorphisms (SNPs), rs12529 (c.15C > G; p.His5Gln) and rs1937920 (12259 bp 3’ of STP A > G), with the risk of developing UBC. rs12529 re- sults in an amino acid substitution and has previously been shown to be inversely associated with bladder can- cer risk.(17) There is no enough data to conclude on the physiological effects of this SNP, however, according to current literature it is possible that homozygous vari- ant genotype may induce AKR1C3 expression to facil- itate an efficient bioactivation process. rs1937920, re- siding in the 3’UTR of the AKR1C3 gene, has also been implicated previously in an association study(17) and has the potential to have regulatory roles in AKR1C3 ex- pression. MATERIALS AND METHODS Study Population Total of 500 Caucasian subjects of Turkish origin were included in the study. The clinical and demograph- ic data of the patients and controls were provided in Table 1. Cases include 250 first-time diagnosed and histologically confirmed transitional cell carcinoma patients who had not received previous chemotherapy or radiotherapy. Patient selection was not subjected to age, gender or tumor stage/grade restrictions. Patients who had received previous radiotherapy, chemotherapy or radical cystectomy, patients with previous diagnosis of cancer, metastasized cancer, serum prostate-specif- ic antigen (PSA) > 2.5 ng/mL, and those with bladder tumors secondary to other malignancies were excluded from the study. Clinical diagnosis of UBC was given by expert urologists based on urine cytology, urinary tract imaging and cystoscopic analysis. When abnormal tissue is recognized during cystoscopy, transurethral re- section of the bladder tumor (TURBT) was performed. Tumors were classified according to the histologic tu- mor grading system of the World Health Organization (WHO) guidelines.(23) Pathological staging of the tum- ors was performed according to the TNM classification of malignant tumors,(24) where pTa and pT1 tumors were grouped as superficial and pT2-pT4 tumors were considered invasive. Tumors were graded as low grade and high grade tumors. The control group includes age- and gender-matched healthy individuals admitted to the hospitals for routine check-up examination with no present or previous his- tory of cancer. Subjects with any degree of hematuria, urinary symptoms, benign prostate hyperplasia (BPH), history of prostatitis and pre-cancerous lesions were ex- cluded from the study. The study protocol was approved by Haliç University Human Research Ethics Committee. All subjects have provided a written informed consent prior to their in- clusion in the study in accordance with Helsinki Dec- laration, revised in 2000. Data regarding age, gender, Table 1. Clinical and demographic data of the study population. Variables Controls Patients P Value Age ± SD (years ) 64.2 ± 12.82 66.4 ± 13.21 .059 Male, no (%) 200 (80) 198 (79.2) .92 Female, no (%) 50 (20) 52 (20.8) BMI ± SD (kg/m2) 27.49 ± 4.22 28.3 ± 3.46 .019 Non-Smoker, no (%) 169 (67.6) 49 (19.6) < .001 Smoker, no (%) 81 (32.4) 201 (80.4) Tumor grade, (%) Low - 61.2 - High - 38.8 - Tumor stage, (%) Superficial - 62.8 - Invasive - 37.2 - Abbreviations: BMI, body mass index; SD, standard deviations. AKR1C3 Polymorphisms in Bladder Cancer-Tiryakioğlu et al. ethnicity, occupation, weight, height, family history of cancer, personal medical history and cigarette con- sumption were collected upon completion of standard- ized questionnaires. Family history of bladder cancer was considered positive when first or second degree relatives of the subjects had given bladder cancer diag- nosis. None of the subjects included in the study were under occupational exposure to hazardous carcinogens related to bladder cancer. Subjects were classified as smokers (who had smoked more than 100 cigarettes in his lifetime) and non-smokers. Ten mL blood samples were collected from patients in EDTA-containing tubes Table 2. Genotype and allele distributions for rs12529 and rs1937920. Variables rs12529 CC CG GG C G no (%) no (%) no (%) no (%) no (%) Patients 32 (12.8) 200 (80) 18 (7.2) 264 (52.8) 236 (47.2) Controls 40 (16) 146 (58.4) 64 (25.6) 226 (45.2) 274 (54.8) P value for HWE .007152 rs1937920 AA AG GG A G no (%) no (%) no (%) no (%) no (%) Patients 101 (40.4) 121 (48.4) 28 (11.2) 323 (64.6) 177 (35.4) Controls 120 (48) 115 (46) 15 (6) 355 (71) 145 (29) P value for HWE .089977 r2 for LD .435795 Abbreviations: HWE, Hardy-Weinberg equilibrium; LD, linkage disequilibrium. Genotype/Allele OR 95% CI P Value P* Value rs12529 CC Referent CG 1.712 1.027-2.856 .038 .076 GG 0.352 0.175-0.708 .003 .006 CC vs (CG+GG) 1.298 0.785-2.144 .308 .616 (CC+CG) vs GG 4.435 2.540-7.744 < .0001 < .0001 C Ref. G 0.737 0.575-0.946 .016 .033 CAT 0.606 - .002 - rs1937920 AA Referent AG 1.250 0.865-1.806 .234 .468 GG 2.218 1.123-4.381 .020 .040 AA vs (AG+GG) 1.362 0.956-1.940 .087 .174 (AA+GG) vs GG 0.506 0.263-0.973 .038 .076 A Referent G 1.342 1.028-1.751 .030 .061 CAT 1.425 - .024 - Abbreviations: CAT, Cochran -Armitage Trend test; OR, odds ratio; CI, confidence interval. All P values were adjusted for age, gender, BMI and smoking status. P*: P values after Bonferroni correction. Table 3. Associations of rs12529 and rs1937920 genotypes and alleles with bladder cancer risk. AKR1C3 Polymorphisms in Bladder Cancer-Tiryakioğlu et al. Urological Oncology 2617 Vol 13 No 02 March-April 2016 2618 and stored at 4oC until DNA extraction. The control subjects have provided cheek epithelial cells on cotton swabs as the source of DNA. Genotyping DNA was isolated from the whole blood samples of patients using High Pure PCR Template Preparation Kit (Roche, Germany) and from buccal swabs us- ing Gentra Puregene Buccal Cell kit (Qiagen, Mary- land, USA). Isolated DNA samples were quantified by spectrophotometry and loaded on 1% agarose gels to confirm their integrity. Both single nucleotide pol- ymorphisms (SNPs) were genotyped using Polymer- ase Chain Reaction (PCR) and Restriction Fragment Length Polymorphism (RFLP) methods. PCR-RFLP assays were designed with Biology Workbench using Primer3 and TCAG modules.(25) rs12529 polymor- phic region amplified using the forward and reverse primers, 5’-CCTCCTACATGCCATTGGTT-3’ and 5’-CAACCCAATACGGGTTTCAC-3’, respectively, and the samples were genotyped by Bts-I digestion of the PCR products. rs1937920 polymorphism was geno- typed by TaqI restriction of the PCR products amplified using 5’-AGGCAGGCGAACAGAAACTA-3’ and 5’-GAAAAACACTTGCATTCGCA-3’ primers. The specificity of the PCR products were analyzed by 2% agarose gel electrophoresis, and the digestion products were separated on 3% agarose gels. Genotypes were Table 4. Associations of rs12529 and rs1937920 genotypes with tumor grade and stage Genotype/ Grade OR 95 % CI P Value P* Value Stage OR 95 % CI P Value P* Value Allele Low High Superficial Invasive rs12529 CC 11 21 Referent 18 14 Referent CG 132 68 0.270 0.123-0.592 .00064 .00192 128 72 0.723 0.654-1.352 .399 1.0 GG 10 8 0.419 0.129-1.366 .145 .436 11 7 0.818 0.252-2.655 .738 1.0 C 154 110 Referent 164 100 Referent G 152 84 0.774 0.539-1.111 .164 .492 150 86 0.940 0.654-1.352 .740 1.0 rs1937920 AA 61 40 Referent 65 39 Referent AG 72 49 1.038 .605-1.780 .892 1.0 70 50 1.190 0.695-2.039 .525 1.0 GG 20 8 0.610 .245-1.518 .285 .855 22 8 0.606 0.246-1.493 .273 .819 A 194 129 Referent 200 128 Referent G 112 65 .873 .598-1.274 .48052 1.0 114 66 0.905 0.621-1.317 .60086 1.0 Abbreviations: OR, odds ratio; CI, confidence interval. All P values were calculated for allelic and additive model and adjusted for age, gender, BMI and smoking status. P*: P values after Bonferroni correction. Figure 1. Bts-I (A) and TaqI (B) digestion products run on 3% agarose gels for 15 minutes at 120 volts (50bp ladder, ThermoFisher, California-USA). AKR1C3 Polymorphisms in Bladder Cancer-Tiryakioğlu et al. determined according to the RFLP patterns on the gels (Figure 1) and confirmed by Sanger sequencing(26) of 50 randomly chosen samples (Figure 2). Statistical Analysis The control samples were tested for Hardy-Weinberg equilibrium (HWE) with Chi-square test. The associa- tions between UBC risk and genotypes were examined with allelic, additive, dominant and recessive models using logistic regression. While Cochran-Armitage Trend test was used to obtain a general odds ratio (OR) for both SNPs, it was essentially used for rs12529, since genotype distribution was found to be departed from HWE. The demographic data have shown normal dis- tribution among patients and controls according to the Shapiro-Wilk test results (P < .05). Therefore, differ- ences between the continuous variables were assessed using t test. Fisher's exact test was used for the assess- ment of categorical variables. All the reported P values are two-tailed and adjusted for confounding factors in- cluding age, gender, body mass index (BMI) and smok- ing status. To account for multiple comparisons, Bon- ferroni correction was applied. Statistical Package for the Social Science (SPSS Inc, Chicago, Illinois, USA) version 21.0 was used for all statistical analyses and P < .05 was considered statistically significant. Statistical power of 80% was considered as the lowest acceptable power score. Power calculation was performed accord- ing to the method suggested by Kelsey and colleagues. (27) Previously reported genotype frequencies for the Caucasian population were used as hypothetical expo- sure rates. The sample size for the minimally acceptable power was calculated for rs12529 and rs1937920 as 65 cases and 65 controls with 130 subjects, and 71 cases and 71 controls with 142 subjects, respectively. RESULTS The allelic and genotypic distributions for rs12529 and rs1937920 SNPs, together with HWE and linkage dise- quilibrium (LD) calculations are shown in Table 2. Ac- cording to the results, the two studied SNPs were not in LD (r2 = .44). Both SNPs were shown to be significant- ly associated with UBC risk (Table 3). Our results point out rs12529 as a protective variant (OR trend = 0.606; p trend = .002) and rs1937920 as a risk variant (OR trend = 1.425; p trend = .024) for UBC. For the rs12529 SNP, frequency of the G allele was found to be significantly different between patients and controls (P* = .016). The homozygous variant GG gen- otype is shown to decrease the risk of UBC (95% CI: .175-.708, OR = 0.352, P* = .006). This effect is also apparent when homozygous wild type and heterozy- gous genotypes (CC+CG) were tested against the ho- mozygous variant (95% CI: 2.540-7.744, OR = 4.435, P* < .0001). The frequency of the rs1937920 variant allele was also significantly different between patients and controls (P* = .03). The variant genotype GG was identified to be a significant risk factor for UBC (95% CI: 1.123-4.381, OR = 2.218, P* = .04), where the GG genotype raises the odds of developing UBC by 2.2 folds. The associations of the alleles and genotypes with tumor grades and stages are shown in Table 4. The heterozygous genotype for rs12529 appears to be significantly associated with tumor grade (P* = .002). However, since the significance decreases in the dom- inant model, this seems to be an artificial effect due to high heterozygous genotype frequency and could be negated in a study with larger sample size. There was no other significant associations between the tumor grades/stages and the studied polymorphisms. DISCUSSION Our results identified the 3’ UTR SNP, rs1937920, as a risk variant and the nonsynonymous rs12529 SNP as a protective variant for UBC, thereby provided evi- dence that AKR1C3 locus is important in the molecular pathogenesis of UBC. The number of studies investi- gating the association of AKR1C3 polymorphisms with cancer is limited to a few. In three of the six studies prostate cancer samples were analyzed,(20-22) and the rest are three individual studies on non-Hodgkin lymphoma, lung cancer and urinary bladder cancer.(17-19) rs1937920 polymorphism has been identified in a Span- ish Bladder Cancer Study using Golden Gate Assay(17) and minor allele frequency was reported to be 0.29 in the control population. To the best of our knowledge this is the first study investigating the rs1937920 SNP Figure 2. Sanger sequencing results for rs12529. Electropherogram for the wild type homozygous (A), heterozygous (B) and homozygous variant (C) genotypes. AKR1C3 Polymorphisms in Bladder Cancer-Tiryakioğlu et al. Urological Oncology 2619 Vol 13 No 02 March-April 2016 2620 in relation to bladder cancer. According to our results, homozygous variant genotype carriers of the Turkish population are at 2.2 folds more risk of developing UBC. It can be speculated that this SNP may result in decreased AKR1C3 expression and/or activity, leading to inefficient bioactivation process. Further association studies with large and different ethnic groups and tis- sue/in vitro expression studies are required in order to clarify the role of this polymorphism in UBC. We have identified the nonsynonymous rs12529 SNP to be inversely associated with UBC risk. To the best of our knowledge, our results constitute the second con- tribution to the literature in terms of the involvement of rs12529 in UBC. Here we report the association of the homozygous variant GG genotype with decreased risk for UBC (OR GG = 0.352; 95% CI: 0.175 - 0.708; P* = .006). Our results corroborated the findings of Figuroea and colleagues(17) who previously reported an inverse association of UBC with this polymorphism in the Spanish population (95 % CI : 0.52-1.18, OR = 0.78, P = .04). These two consistent data in the literature sup- port the idea that AKR1C3 c.15 C > G polymorphism may contribute significantly to UBC pathogenesis via induction of AKR1C3 expression to facilitate an effi- cient bioactivation process. However, this hypothesis should be confirmed with AKR1C3 expression analyses in UBC patients. Interestingly, the heterozygous gen- otype for rs12529 shows a trend towards significance (P* = .076) in terms of its association with increased UBC risk in our study. However, since the significance of this effect decreases in dominant model, this seems to be an artificial effect as a result of high heterozygous genotype frequency and could be negated in a study with larger sample size. Regarding the associations of the alleles and genotypes with tumor characteristics, the only significant association was between the het- erozygous genotype for rs12529 and tumor grade (P* = .00192). However, this is possibly due to high rep- resentation of the heterozygotes. We could not detect any other associations between the tumor characteris- tics and the studied genotypes. The association of rs12529 was investigated previous- ly in Chinese lung cancer patients by Lan et al.(18) who reported two folds increased risk for the homozygous variant genotype (GG). In another study by Lan and colleagues, no association between AKR1C3 variants and non-Hodgkin lymphoma was reported.(19) Cun- nigham and colleagues on the other hand, investigated AKR1C3 variants in sporadic and familial prostate can- cer, but could not find an association between rs12529 and prostate cancer risk.(20) Two other studies on pros- tate cancer have revealed significant association of AKR1C3 variants and prostate cancer risk when gene interactions were in taken into consideration.(21,22) While the discrepancy on the role of rs12529 SNP can be attributed to differences in the pathogeneses of lung and urinary bladder cancers, a unique gene-gene or gene-environment effects specific to Chinese popula- tion is more likely. This is also supported by the fact that rs12529 genotype distributions in Asian popula- tions are significantly different than that of Caucasians such that Asians have the frequency of 0.770 for the G allele, which was designated as the minor allele for other populations. CONCLUSIONS In conclusion, our results suggest that AKR1C3 rs12529 and rs1937920 variants have significant contri- butions to the risk of developing UBC. Because of the case–control design of this study, population stratifica- tion and admixture effects may increase the type I error rate of association; therefore, the results presented here should be regarded as preliminary and might represent a first step of future larger studies aiming to better elu- cidate the role of AKR1C3 polymorphisms in the sus- ceptibility to bladder cancer. ACKNOWLEDGEMENTS The authors thank Haliç University for financial sup- port. CONFLICT OF INTEREST None declared. REFERENCES 1. Ferlay J, Soerjomataram I, Ervik M, Dikshit R, Eser S, Mathers C, Rebelo M, Parkin DM, Forman D, Bray, F. GLOBOCAN 2012 v1.1, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 11 [Internet]. Lyon, France: International Agency for Research on Cancer; 2014. Available from: http:// globocan.iarc.fr, accessed on 13/01/2016. 2. Burger M, Catto JW, Dalbagni G, et al. Epidemiology and risk factors of urothelial bladder cancer. Eur Urol. 2013;63:234-41. 3. Franekova M, Halasova E, Bukovska E, Luptak J, Dobrota D. Gene polymorphism in bladder cancer. Urol Oncol. 2008;26:1-8. 4. Bartsch H, Rojas M, Nair U, Nair J, Alexandrov K. Genetic cancer susceptibility and DNA adducts: studies in smokers, tobacco chewers, and coke oven workers. Cancer Detect Prev. 1999;23:445-53. 5. Kalantari MR, Ahmadnia H. P53 overexpression in bladder urothelial AKR1C3 Polymorphisms in Bladder Cancer-Tiryakioğlu et al. neoplasms: new aspect of World Health Organization/International Society of Urological Pathology classification. Urol J. 2007;4:230-3. 6. Jalali Nadoushan MR, Taheri T, Jouian N, Zaeri F. Overexpression of HER-2/neu oncogene and transitional cell carcinoma of bladder. Urol J. 2007;4:151-4. 7. Karimianpour N, Mousavi-Shafaei P, Ziaee AA, et al. Mutations of RAS gene family in specimens of bladder cancer. Urol J. 2008;5:237-42. 8. Nanda MS, Sameer AS, Syeed N, et al. Genetic aberrations of the K-ras proto-oncogene in bladder cancer in Kashmiri population. Urol J. 2010;7:168-73. 9. Zeegers MP, Dorant E, Goldblum RA, Van den Brandt PA. The impact of characteristics of cigarette smoking on urinary tract cancer risk: a meta-analysis of epidemiologic studies. Cancer. 2000;89:630-9. 10. Rota M, Bosetti C, Boccia S, Boffetta P, La Vecchia C. Occupational exposures to polycyclic aromatic hydrocarbons and respiratory and urinary tract cancers: an updated systematic review and a meta-analysis to 2014. Arch Toxicol. 2014;88:1479-90. 11. Hecht SS. Tobacco carcinogens, their biomarkers and tobacco-induced cancer. Nat Rev Cancer. 2003;3:733–44. 12. Palackal NT, Lee SH, Harvey RG, Blair IA, Penning TM. Activation of polycyclic aromatic hydrocarbon trans-dihydrodiol proximate carcinogens by human aldo-keto reductase (AKR1C) enzymes and their functional overexpression in human lung carcinoma (A549) cells. J Biol Chem. 2002;277:24799- 808. 13. Nagaraj NS, Beckers S, Mensah JK, Waigel S, Vigneswaran N, Zacharias W. Cigarette smoke condensate induces cytochromes P450 and aldo-keto reductases in oral cancer cells. Toxicol Lett. 2006;165:182-94. 14. Penning TM, Jin Y. Aldo-Keto reductases and bioactivation/detoxication. Annu Rev Pharmacol Toxicol. 2007;47:263-92. 15. Penning TM, Drury JE. Human Aldo-Keto Reductases: Function, gene regulation and single nucleotide polymorphisms. Arch Biochem Biophys.. 2007;464:241-250. 16. Courter LA, Pereira C, Baird WM. Diesel exhaust influences carcinogenic PAH-induced genotoxicity and gene expression in human breast epithelial cells in culture. Mutat Res. 2007;625:72-82. 17. Figuroea JD, Malats N, Garci-Closas M, et al. Bladder cancer risk and genetic variation in AKR1C3 and other metabolizing genes. Carcinogenesis. 2008;29:1955-62. 18. Lan Q, Mumford JL, Shen M, et al. Oxidative damage-related genes AKR1C3 and OGG1 modulate risks for lung cancer due to exposure to PAH-rich coal combustion emissions. Carcinogenesis. 2004;25:2177-81. 19. Lan Q, Zheng T, Shen M, et al. Genetic polymorphisms in the oxidative stress pathway and susceptibility ton on-Hodgkin lymphoma. Hum Genet. 2007;121:162-8. 20. Cunningham JM, Hebbring SJ, McDonnel SK, et al. Evaluation of genetic variations in the androgen and estrogen metabolic pathways as risk factors for sporadic and familial prostate cancer. Cancer Epidemiol Biomarkers Prev. 2007;16:969-78. 21. Vaarala, MH, Mattila H, Ohtonen P, Tammela TL, Paavonen TK, Schleutker J. The interaction of CYP3A5 polymorphisms along the androgen metabolism pathway in prostate cancer. Int J Cancer. 2008;122: 2511-6. 22. Kwon EM, Holt SK, Fu R, et al. Androgen metabolism and JAK/STAT pathway genes and prostate cancer risk. Cancer Epidemiol. 2012;36:347-53. 23. Miyamoto H, Miller JS, Fajardo DA, Lee TK, Netto GJ, Epstein JI. Non-invasive papillary urothelial neoplasms: the 2004 WHO/ISUP classification system. Pathol Int. 2010;60:1-8. 24. Hermanek P and Sobin L (eds.) TNM classification of malignant tumors. Berlin: Springer-Verlag. 1987;131-146. 25. Subramaniam S. The biology workbench--a seamless database and analysis environment for the biologist. Proteins. 1998;32:1-2. 26. Database of Single Nucleotide Polymorphisms (dbSNP). Bethesda (MD): National Center for Biotechnology Information, National Library of Medicine. dbSNP accession:{rs12529}, (dbSNP Build ID: {144}). Available from:http://www.ncbi.nlm.nih.gov/SNP/ 27. Kelsey JL, Whittemore AS, Evans AS, Thompson WD. Methods in Observational Epidemiology (2nd Edition) 1996. Tables 12- 15. Oxford University Press. AKR1C3 Polymorphisms in Bladder Cancer-Tiryakioğlu et al. Urological Oncology 2621