Polymorphisms in sulfotransferases SULT1A1 and SULT1A2 are not related to colorectal cancer

Dear Sir, Cytosolic sulfotransferases (SULT) play an important role in Phase II metabolism of several xenobiotics. Although sulfonation in general exerts detoxifying effects, it may also lead to the bioactivation of certain promutagens and procarcinogens including heterocyclic amines that can be formed when meats are cooked at high temperature.1,2 Several sulfotransferases have been identified in humans, which show important tissue distribution variability and different affinities for some substrates. For SULT1A1 and SULT1A2, 2 polymorphisms have been detected that lead to altered enzyme functions. SULT1A1 is highly expressed in numerous tissues including colon mucosa. Its variant SULT1A1*His was associated with decreased enzyme activity and thermostability in platelets and reduced activation of some pro mutagens in vitro3 compared to SULT1A1*Arg. There are no published studies demonstrating the presence of SULT1A2 protein in any tissue. Using a modified protocol for electrophoresis, our group could separate SULT1A2 from other cross-immunoreactive SULT forms, including SULT1A1, and unambiguously detect the presence of moderate levels of SULT1A2 protein in some (but not all) colon mucosa samples investigated as well as in the colon carcinoma cell line Caco-2 (W. Meinl, W. Teubner, A. Lampen and H. R. Glatt, unpublished data). In vitro studies have shown that SULT1A2 can activate procarcinogens.4 The gene SULT1A2 is usually in linkage disequilibrium with SULT1A1, and high-activity alloenzymes (SULT1A1*Arg and SULT1A2* Asn) are associated together.5 SULT1A1*His has been associated to a protective effect in colorectal cancer,6 although other studies have not confirmed this finding.7–10 We have tested the main hypothesis that polymorphisms in SULT1A1 and SULT1A2 modify the risk of colorectal cancer. We have also explored whether these polymorphisms interact with exposures such as diet, alcohol, tobacco or drugs and with genetic alterations in the tumors. A hospital-based case-control study was conducted to assess risk factors of colorectal cancer and gene-environment interactions. Details about the study population, interviews and collection of biological samples have been published elsewhere.11 Briefly, cases were patients with a new diagnosis of colorectal adenocarcinoma attending a University Hospital in Barcelona, Spain, between January 1996 and December 1998. Controls were selected from admissions to the same hospital during this period. To avoid selection bias of controls, the reason for admission had to be a disease not previously diagnosed for that patient. Main diagnosis groups of controls were: acute digestive surgery 19%, urology 17%, gastroenterology 16% and orthopedic surgery 15%. The study protocol was cleared by the Ethics Committee of the hospital and all individuals gave written informed consent to participate and to perform genetic analysis on their samples. Our study is based on 293 cases (67% of eligible) and 272 controls (63% of eligible) that provided a blood sample to perform the genetic analyses. Cases and controls not included were similar to those included in demographic variables and main exposures (diet, alcohol, smoking and drugs). Cases and controls were interviewed by trained personnel using a structured questionnaire. A dietary history questionnaire, developed previously and validated in the framework of the EPIC study, focused on average food consumption 1 year before diagnosis. Detailed cooking methods were recorded for meat and fish products. Other risk factors measured were body mass index (BMI) at diagnosis and 10 years before, parity in women and life-long history of non-steroidal anti-inflammatory drugs (NSAID), tobacco and alcohol use. The polymorphic SULT1A1 and SULT1A2 sequences were amplified in separate reactions as described previously.5,12 Genotyping was carried out by restriction fragment length polymorphism (RFLP) analyses after incubation with Bsp143II for SULT1A1 and BpiI for SULT1A2, respectively, and subsequent separation of the digestion products on agarose gels. Fresh tumour tissue and normal mucosa samples were obtained from surgically extracted specimens of cases. K-ras gene mutations in codons 12 and 13 were studied by PCR followed by single strand conformation polymorphism (SSCP) analysis. Similarly, p53 exons 4–9 were analyzed by PCR/SSCP and direct sequencing whenever necessary. Expression of the p53 protein was determined by immunohistochemistry, using commercially available antibodies (ab-6 Pantropic, Oncogene Research Products, VWR International GmbH, Darmstadt, Germany). Microsatellite instability was studied analyzing 5 microsatellite sequences. Each polymorphism was tested to evaluate if genotype frequencies followed Hardy-Weinberg equilibrium. Haplotypes were estimated with the EM-algorithm. Odds ratios (OR), adjusted for age and gender, and 95% confidence intervals (CI)