DOI:http://dx.doi.org/10.7314/APJCP.2012.13.5.2311 Association between Polymorphisms in UGT1A6 and 1A7 and Colorectal Cancer Risk RESEARCH COMMUNICATION Association between Polymorphisms in UDP- glucuronosyltransferase 1A6 and 1A7 and Colorectal Cancer Risk Kayo Osawa1,7*, Chiaki Nakarai1,7, Minami Akiyama1,7, Ryuta Hashimoto1,7, $NLPLWVX7VXWRX1,7-XUR7DNDKDVKL1,7<XNR7DNDRND2, Shiro Kawamura3,7, Etsuji Shimada3,7.HQLFKL7DQDND, Masaya Kozuka5,7, MasahiroYamamoto5,7, Yoshiaki Kido1,6,7 Abstract Genetic polymorphisms of uridine diphosphate-glucuronosyltransferases 1A6 (UGT1A6) and 1A7 (UGT1A7) may lead to genetic instability and colorectal cancer carcinogenesis. Our objective was to measure the interaction between polymorphisms of these repair genes and tobacco smoking in colorectal cancer (CRC). A total of 68 individuals with CRC and 112 non-cancer controls were divided into non-smoker and smoker groups according to pack-years of smoking. Genetic polymorphisms of UGT1A6 and UGT1A7 were examined using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). We found a weak association of UGT1A6SRO\PRUSKLVPVZLWK&5&ULVN FUXGHRGGVUDWLR>25@FRQÀGHQFHLQWHUYDO>&,@ 3 DGMXVWHG25&,3 7KH25VIRUWKHUGT1A7 polymorphisms were statistically VLJQLÀFDQW FUXGH25&,3 DGMXVWHG25&,3 7KH joint effect of tobacco exposure and UGT1A6SRO\PRUSKLVPVZDVVLJQLÀFDQWO\DVVRFLDWHGZLWKFRORUHFWDOFDQFHU ULVNLQQRQVPRNHUV FUXGH25&,3 DGMXVWHG25&,3 ,Q FRQFOXVLRQRXUÀQGLQJVVXJJHVWWKDWUGT1A6 and UGT1A7 gene polymorphisms are associated with CRC risk LQWKH-DSDQHVHSRSXODWLRQ,QSDUWLFXODU8*7$SRO\PRUSKLVPVPD\VWURQJO\LQFUHDVH&5&ULVNWKURXJK the formation of carcinogens not associated with smoking. .H\ZRUGVGene polymorphisms - colorectal cancer (CRC) - UGT1A6 - UGT1A7 $VLDQ3DFLÀF-&DQFHU3UHY 13, 2311-2314 Introduction 1993). The locus codes of nine functional proteins of the UGT1A family differ only in their N-terminus as a Colorectal cancer (CRC) is the third most commonly result of alternate splicing of independent exon 1 regions diagnosed cancer, and the leading cause of cancer death to a shared C-terminus encoded by exons 2 to 5 (Nagar in the world (Siegel et al., 2012). Genetic susceptibility et al., 2006).UGT1A6 is expressed in several human to this disease may result from inherited mutations in tissues, and catalyses the glucuronidation of small planar genes involved in carcinogen metabolism and DNA phenols and primary aromatic amines (Nagar et al. 2006). repair (Paz-Elizur et al., 2008; Joshi et al., 2009; Zhao et Two functionally important SNPs arise in exon 1 of the al., 2012). Uridine diphosphate-glucuronosyltransferases 8*7$ gene, which result in T181A and R184S amino (UGTs) play an important role in glucuronidation of acid changes (8*7$ ), although alleles carrying various endogenous and exogenous compounds such as only the T181A polymorphism (8*7$ ) or the bilirubin, steroid hormones, bile acids, anticancer drugs, R184S polymorphism (8*7$ ) have been described WREDFFRVSHFLÀFFDUFLQRJHQVDQGGLHWDU\SURFDUFLQRJHQV (http://www.pharmacogenomics.pha.ulaval.ca/files/ including heterocyclic amines (HCAs) and polycyclic content/sites/pharmacogenomics/files/Nomenclature/ aromatic hydrocarbons (PAHs) (Ciotti et al., 1997; van UGT1A/UGT1A6.htm). Metabolic rates of phenols by der Logt et al., 2004; Nagar et al., 2006). UGTs are recombinant 8*7$ were lower than those of the FODVVLÀHGLQWRVHYHUDOIDPLOLHVDQGVXEIDPLOLHVEDVHG most commonly occurring enzyme (Ciotti et al., 1997). on their structural and sequence homology (Jin et al., 8*7$FDWDO\VHVWKHFRQMXQFWLRQDQGGHWR[LÀFDWLRQ 1)DFXOW\RI+HDOWK6FLHQFHV.REH8QLYHUVLW\*UDGXDWH6FKRRORI+HDOWK6FLHQFHV2&OLQLFDO/DERUDWRU\2WHPDH+RVSLWDO2VDND -DSDQ3'HSDUWPHQWRI6XUJHU\1DWLRQDO+RVSLWDO2UJDQL]DWLRQ.REH0HGLFDO&HQWHU4'HSDUWPHQWRI6XUJHU\+\RJR&DQFHU&HQWHU $NDVKL-DSDQ5'HSDUWPHQWRI6XUJHU\.REH5RVDL+RVSLWDO6'LYLVLRQRI'LDEHWHVDQG(QGRFULQRORJ\'HSDUWPHQWRI,QWHUQDO 0HGLFLQH.REH8QLYHUVLW\*UDGXDWH6FKRRORI0HGLFLQH7.REH-DSDQ )RUFRUUHVSRQGHQFHRVDZDN#NREHXDFMS $VLDQ3DFLÀF-RXUQDORI&DQFHU3UHYHQWLRQ9RO 2311 Kayo Osawa et al of several tobacco-related carcinogens (Nagar et al., (pack-years = 0), and smokers (pack-years >0). The 2006), and polymorphisms have been reported to result gene-smoking interaction was also computed, and was in the amino-acid changes N129K /R131K(8*7$ ), adjusted for age and gender. Hardy-Weinberg equilibrium N129K/R131K/W208R (8*7$ ), and W208R ZDVWHVWHGXVLQJWKHJRRGQHVVRIÀW&KLVTXDUHWHVWWR (8*7$ ) (http://www.pharmacogenomics.pha. compare the observed genotype frequencies with the ulaval.ca/files/content/sites/pharmacogenomics/files/ expected genotype frequencies among the control subjects. Nomenclature/UGT1A/UGT1A7.htm). UGT1A7 activity The statistical analysis was performed with PASW for resulting from these polymorphisms were moderate to low Windows version 17.0 (SPSS Japan Inc., Tokyo, Japan). compared with that of wild type 8*7$ (Guillemette et al., 2000; Zheng et al., 2001). Results and Discussion We have already reported that genetic polymorphisms 100.0 of 1$7 and &<3$ in metabolic processes, and of We investigated the potential impact of genetic 087<+ and $3(; DNA repair genes, contribute polymorphisms located in 8*7$ and 8*7$ on to CRC susceptibility in relation to smoking status in CRC risk. UGTs are thought to play a critical role in75.0 Japanese populations (Yoshida et al., 2007; Kasahara et al., CRC because they are key enzymes in the metabolism 2008). We conducted a case-control study to evaluate the of endogenous and exogenous compounds, including relevance to CRC risk by means of examining 8*7$ steroid hormones, xenobiotics and drugs (Nager et al., and 8*7$ gene polymorphisms, and assessed the effect 2006). Many genetic polymorphisms in UGTs have50.0 of these genetic polymorphisms on CRC susceptibility in been described, and some have been associated with relation to smoking status. increased CRC susceptibility (Fang & Lazarus, 2004; 7DQJHWDO 7KHSUHVHQWVWXG\LVWKHÀUVWWLPH25.0 Materials and Methods that the association between 8*7$ and 8*7$ polymorphisms and CRC has been studied in a Japanese 6WXG\6XEMHFWV population. A total of 67 CRC patients (38 with colon cancer, 24 Table 1 summarizes the distribution of characteristics 0 with rectal cancer and 5 with other carcinomas) were of CRC patients and controls. There was no difference in recruited between October 2003 and March 2005 at the the gender distribution (P=0.518) between males (patients, None Kobe Medical Center and Kobe Rosai Hospital in Kobe, 53.7%; controls, 63.4%) and females (patients, 38.3%; Remission Japan. The populations were included in previous studies controls, 36.6%). The mean age was 67.3 ± 11.0 years Radiotherapy that investigated genetic polymorphisms of metabolic or for patients and 67.6 ± 6.8 years for controls, respectively Chemotherapy DNA repair enzymes (Yoshida et al., 2007; Kasahara et al., (P=0.859). Never-smokers comprised 55.2% of patients 2008). The controls consisted of 112 individuals who had and 45.5% of controls, and smokers comprised 35.8% of Persistence or recurrence or Persistence not been currently or previously diagnosed with cancer, patients and 51.8% of controls. Subtypes of the cases were chemoradiation Concurrent and were recruited between November 2002 and March as follows: colon, 56.7%; rectal, 35.8% and others, 7.5%. 2003. Informed consent was obtained from all patients Genotype distributions of 8*7$ adjusted for withtreatment diagnosed Newly and controls, and all samples were coded after collection gender, age, and smoking habit, and corresponding withouttreatment diagnosed Newly of blood and data. The amount of smoke exposure was allele frequencies are shown in Table 2. The genotype calculated as pack-years, the product of number of years distributions and allele frequencies in patients are based an individual had smoked and the average number of RQ3&5²5)/3$OOJHQRW\SHGLVWULEXWLRQVWHVWHGIXOÀOOHG cigarettes smoked per day, converting to a standard pack of the Hardy-Weinberg criteria. The distribution of variant 20 cigarettes. The study design was approved by the Ethics 8*7$DOOHOHVGHPRQVWUDWHGDERUGHUOLQHVLJQLÀFDQW Review Committee on Genetic and Genomic Research, association with CRC risk overall when 8*7$ Kobe University Graduate School of Medicine. 7DEOH&KDUDFWHULVWLFVRI&RORUHFWDO&DQFHU&DVHV *HQRW\SLQJ and Control Subjects Genomic DNA was isolated from leukocytes in Item Patients Controls P-value the previous studies (Yoshida et al., 2007; Kasahara n % n % et al., 2008). The genotypes of 8*7$ and 8*7$ were determined by PCR-RFLP analysis, as previously Number 67 112 Gender:males 36 53.7 71 63.4 0.518a described (Logt et al., 2004; Chen et al., 2006). females 26 38.3 41 36.6 unknown 5 7.5 0 0.0 6WDWLVWLFDO$QDO\VLV Age: Mean ±S.D. 67.3 ±11.0 67.6 ±6.8 0.859b Logistic regression analysis was performed to Smoking status (Pack-years) assess the association between each genotype and CRC. Never 37 55.2 51 45.5 0.109a Associations were expressed as odd ratios (OR) with Ever 24 35.8 58 51.8 FRQÀGHQFHLQWHUYDOV &, DQG3YDOXHVRI unknown 6 9.0 3 2.7 ZHUHFRQVLGHUHGVWDWLVWLFDOO\VLJQLÀFDQW7KH25VZHUH Histological type computed to estimate the association between certain colon 38 56.7 rectum 24 35.8 genotypes and CRC, and were adjusted for age, gender, unknown 5 7.5 and smoking habits. The subjects were divided into two groups according to pack-years of smoking: never-smoked a:Ƶ2 analysis, bStudent’s T-test 2312 $VLDQ3DFLÀF-RXUQDORI&DQFHU3UHYHQWLRQ9RO DOI:http://dx.doi.org/10.7314/APJCP.2012.13.5.2311 Association between Polymorphisms