Effects of Multiple Risk Factors for Hepatocellular Carcinoma on Formation of Aflatoxin -DNA

Covalent binding of aflatoxin B1 (AFB1) with hepatic DNA may be a critical step in hepatocarcinogenesis. The extent of the AFB1 binding to DNA may depend on various endogenous factors and concurrent exposure to other environmental agents. This study was performed to determine whether any individual characteristics correlated with the formation of AFB1-DNA adducts. The major AFB1-DNA adduct, AFB1-N -guanine, was measured using a high performance liquid chromatographic assay in urine samples from 43 asymptomatic hepatitis B virus surface antigen earners and 43 noncarriers randomly selected from a cohort study in Taiwan. The total aflatoxin metabolite level was associated with the detection rate of urinary AFB1-N7guanine adducts in a dose-dependent manner. The AFB1DNA adduct excreted in the urine was detectable in 60% of individuals who smoked cigarettes but abstained from alcohol, 64% of individuals who had a habit of drinking alcohol but not smoking cigarettes, and only 29% of those who neither smoked nor drank alcohol. The association between urinary AFB1-DNA adduct level and habits of smoking cigarettes and drinking alcohol remained statistically significant when adjustment was made for potential confounders. There was a significant increase with age for the detection rate of urinary AFB5N7-guanine adducts. Age and habits of cigarette smoking and alcohol drinking were also found to be associated with a higher percentage of AFB1-N -guanine in total AFB1 metabolite excretion, indicating an increased activation of AFB1. No significant association with the AFB1-DNA adduct level was observed for hepatitis B virus surface antigen carrier status, educational level, and ethnicity. These data suggest a potential role of age, cigarette smoking, and alcohol drinking in AFB1-induced hepatocarcinogenesis. Received 1/5/96: revised 5/6/96; accepted 5/7/96. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked ads’ertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This study was supported by Grants NSC 83-04l2-B-002-256 and NSC 84233 l-B-002-183 from the National Science Council of the Republic of China. 2 To whom requests for reprints should be addressed, at Institute of Epidemiology, College of Public Health, National Taiwan University. I Jen-Ai Road, Section 1, Taipei 10018, Taiwan. Fax: (886) 2-3511955. Introduction Considerable evidence indicates that HCC3 is multifactorial in origin. Our previous epidemiobogical studies have linked many risk factors to the development of HCC (18). One of the most striking epidemiobogical characteristics of HCC is its remarkable geographic variation. High-incidence areas cluster in the Far East and tropical Africa, where the annual incidence ranges up to 500 cases per 100,000 population (9). Although chronic HBV infection has been well documented as the most important risk factor for HCC in these areas, ingestion of AFB1 has long been implicated as another major cause ( 10). This mycotoxin is the most potent hepatocarcinogen in a variety of animal species (1 1). Field studies in Africa and Southeast Asia have shown a strong correlation between high incidence of HCC and contamination of foodstuffs by AFs (12, 13). FB1 is metabolized by the microsomal mixed-function oxygenase enzyme system to a variety of reduced and oxidized derivatives, including an unstable reactive epoxide, AFB -8,9epoxide, which can form adducts with nucleophilic sites in DNA (14). The latter process has been shown to be critical for the carcinogenesis induced by AFB1 in animals (15). The major product formed by the interaction of FB1 with hepatic DNA in vivo is AFB1-N7-guanine (16). This compound is lost rapidly from DNA and excreted in the urine of AFB treated animals. The quantity of the adducts excreted in the urine following FB1 administration was shown to be proportional to the amount of AFB 1-N7-guanine initially formed in the liver DNA (17). The measurement of these adducts in human urine provides a noninvasive method of estimating DNA binding by I at the target site and a useful surrogate dosimeter for HCC risk following AFB1 exposure. The impact of chemoprotection on urinary levels of FB1 -N7-guanine adducts in rats has been demonstrated to be comparable to that on the hepatic DNA modification levels by FB1 (18). A case-control study nested in a cohort study indicated that the presence of urinary AFB 1-DNA adducts was the most important predictor of developing AFB 1-related HCC among urinary AF biomarkers examined (19). The complete elimination of exposure to AFB1 is not possible in high HCC risk areas. Recent experimental works have focused on the chemopreventive effects of a number of synthetic and natural compounds that may modulate the FB1 binding to DNA in vitro and in vivo (15, 18, 20, 21). On the other hand, there may be considerable interindividual variation in the DNA adduct levels formed by I for a given dietary exposure to this mycotoxin. There is a strong interaction between chronic HBV infection and AF exposure on the devel3 The abbreviations used are: HCC, hepatocellular carcinoma; HBV, hepatitis B virus; AF, aflatoxin; AFB1 , aflatoxin ; AFB -N7-guanine. 8,9-dihydro-8-(N7guanyl)-9-hydroxyaflatoxin I HBsAg. HBV surface antigen; anti-HCV, antibodies against hepatitis C virus; HPLC, high performance liquid chromatography: CI, confidence interval. on August 27, 2017. © 1996 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from 614 Risk Factors for Formation of AFB1-DNA Adducts opment of HCC (19). One possible mechanism responsible for this interaction, among several that have been proposed. is that chronic HBV infection may alter AF metabolism, either the activation to the reactive epoxide or the detoxification. However, epidemiobogical data on the association between HBsAg carrier status and the serum level of the FB1 -albumin adduct, another surrogate estimator of hepatic DNA damage induced by AFB1, have been inconsistent (22-25). Other factors that may influence the binding of FB1 to hepatic DNA have yet to be identified. Cigarette smoking and alcohol drinking have also been linked to the development of HCC (2, 4, 8). They may interact with other HCC risk factors, such as , in the pathogenesis of HCC. This study was carried out to evaluate the correlations of multiple HCC risk factors with the formation of AFB1-DNA adducts, using urinary AFB1-N7-guanine as a biomarker for assessing AFB1-DNA binding in the liver. Materials and Methods Subjects. This study included a total of 43 male HBsAg carriers and 43 male age-matched (within 5 years) noncarriers who were randomly selected from the cohort members without HCC in a prospective study on the multifactorial etiology of HCC in Taiwan. The cohort characteristics and method of follow-up have been described in a previous study (8). At entry into this cohort study, all the study participants were tested for HBsAg by RIA (Abbott Laboratories, North Chicago, IL). Those who were positive for HBsAg were scheduled to undergo both ultrasonography and a-fetoprotein measurement every 6-12 months. Individuals with ultrasonographic images compatible with liver cirrhosis or HCC and/or elevated a-fetoprotein levels of more than 20 ng/ml were referred to hospitals for further confirmatory examinations, including fine-needle aspiration cytology and image diagnosis. All the chronic HBsAg carriers included in this study remained asymptomatic throughout a 5-year follow-up period, except two who were subsequently affected with liver cirrhosis. This study was conducted with the approval of the Department of Health and in compliance with regulations for the protection of human research subjects. Urine samples were collected in disposable containers from study subjects at the initial recruitment examination during a period from August 1988 through June 1992. The urine samples were stored at -30#{176}C until they were used for analysis of AF metabolites. At the time of urine collection, each study subject was also personally interviewed to obtain information on demographic characteristics, habits of cigarette smoking and alcohol drinking, and the frequency (meals per month) of consuming peanuts and fermented bean products, which were thought to be major AF-contaminated foodstuffs in Taiwan within the 6 months before questionnaire interview, as well as personal and family history of various chronic diseases. A habit of cigarette smoking was defined as having smoked cigarettes more than 4 days a week for at least 1 year. Alcohol drinking was defined as having consumed alcohol more than 1 day a week for at least 1 year. All but seven study subjects were also tested for anti-HCV. Anti-HCV was examined by a secondgeneration enzyme immunoassay (Abbott Laboratories, North Chicago, IL). Analysis of Urinary AF Metabolites. Authentic AFB 1-N7guanine was synthesized according to the method of Martin and Garner (26). Standards of AFB1, AFM1, and AFP1 (Sigma Chemical Co., St. Louis, MO) were prepared in bezene:acetonitrile (98:2, v/v), and AFB1-N7-guanine was prepared in methanol. For extraction of I metabolites in urine, an aliquot of 5 ml urine sample was incubated with 5 ml ofO.l M Na2SO4 and 5 ml of 0.1 M acetic acid containing 710 units f3-glucuronidase for 1 8 h at 37#{176}C. Precipitated material was removed by centrifugation of the sample at 3000 rpm for 10 mm. The supernatant was then treated with 5 ml chloroform, and the mixture was vortexed for 1 mm. The sample was centrifuged for 5 mm at 3000 rpm to aid in the solvent separation. Following centrifugation, the chloroform layer was removed and treated with 5 ml deionized water. The mixture was vortexed and centrifuged at 3000 rpm for 5 mm. The chloroform layer was removed and evaporated to dryness under a stream of N2 gas, and then n-hexane (2 ml) and trifluoroacetic acid (200 p.1) were added. After 1 h at 40#{176}C, 0.8 ml deionized water:acetonitrile (9: 1 , v/v) was added, the solution was vortexed for 1 mm, and the lower aqueous layer was transferred to a vial for HPLC analysis. FB1 metabolites excreted in the urine were analyzed by reversed-phase HPLC. HPLC was performed with a Waters system incoporating two M510 pumps, an M680 system controller, an M7 17 autoinjector, and an M470 scanning fluorescence detector. Quantitation of various metabolites was accomplished with a Millennium 2010 chromatography manager (Waters Associates, Milford, MA). The HPLC column used was a C18 l0.am (300 X 3.9 mm) p.Bondpak column (Waters). Gradient elution was used to improve the resolution of very similar structures of FB1 metabolites. An aliquot of 25 l extracted samples was injected into the HPLC system. For analysis of AFM1, AFB3-N7-guanine, and FB1, chromatographic separation was obtained by elution for 12 mm with 15% acetonitrile:water, and then the system was automatically switched to a 22% acetonitrile:water solvent mix. The flow rates were 1 .5, 0.8, 0.3, and 1.0 mm/mI at 0-12, 1 3-21 , 22-35, and 36-41 mm, respectively. Elutes were measured by fluorescence detection with 365 nm excitation and a 430-nm emission wavelength for AFM1 and with 500 nm emission for AFB 1-N7-guanine and FB1 . AFM , AFB 1-N7-guanine, and AFB1 eluted at 8.9, 21.5, and 36.4 mm, respectively. The HPLC analysis for AFP1 was done with a 30-mm elution with 12% acetonitrile:water, followed by a 12-22% acetonitrile linear gradient generated over 12 mm and then at 22% acetonitrile. The flow rates were 1.5, 0.3, 1.5, 0.3, 1.0, and 0.3 mb/mm at 0-17, 18-19, 20-29, 30-41, 42-44, and 45-52 mm, respectively. AFP1 chromatographed with a retention time of43.1 mm by monitoring the fluorescence emission at 500 nm with the excitation wavelength at 365 nm. All aqueous mobile phases before use were adjusted by orthophosphoric acid and triethylammonium formate buffer to pH 3.0. Statistical Methods. The relationship between monthly consumption frequency of peanuts and fermented bean products and the excretion of total AF metabolites (including AFM5, AFP1 , FB1 -N7-guanine, and AFB ) in urine was examined by using Pearson’s correlation coefficient. Odds ratios and their 95% CIs were computed to examine the associations of the positivity of urinary AFB1-N7-guanine with various variables. Urinary levels of total AF metabolites were trichotomized according to the tertile distribution of the levels of all study subjects. Mantel’s x test for a trend was performed to examine the dose-response relationship for the odds ratios of the positivity of AFB1-N7-guanine in relation to age and the total AF metabolites excreted. Multivariate-adjusted odds ratios were estimated by modeling the data through logistic regression. In the analysis, consecutive scores of 1 , 2, 3, and 4 were assigned to the age groups of 30-39, 40-49, 50-59, and 60 years. All on August 27, 2017. © 1996 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from