Chem. Res. Toxicol. 1996, 9, 439-444
439
Cigarette Smoking and Urinary 3-Alkyladenine Excretion in Man Virginie Prevost† and David E. G. Shuker*,†,‡ International Agency for Research on Cancer, 150 cours Albert Thomas, 69372 Lyon Cedex 08, France Received June 15, 1995X
A recently developed technique to measure several 3-alkyladenines (3-alkAde) simultaneously in urine has been applied to study alkylating agent exposure arising from cigarette smoke. A volunteer who was a moderate smoker (10-32 cigarettes per day) excreted significantly more 3-methyladenine (3-MeAde) on days when he smoked than when he did not smoke. In contrast, there was no significant difference in 3-MeAde excretion in a light smoker (less than 17 cigarettes per day) on smoking compared to nonsmoking days. However, in volunteers who consumed a standardized diet low in preformed 3-MeAde, there was a smoking-related increase in 3-MeAde excretion even at low levels of cigarette use (less than 11 cigarettes per day). In the same volunteers, no evidence of smoking-related excretion of 3-(2-hydroxyethyl)adenine could be seen. In contrast, levels of urinary 3-ethyladenine (3-EtAde) increased slightly in smokers on standardized diets. Furthermore, since the normal background level of 3-EtAde was low, an exposure-dependent increase in this adduct was seen in two smokers on freechoice diets over a 15 day period. The agent(s) responsible for the increase in 3-EtAde excretion have not been identified, but preliminary results suggest that a direct-acting ethylating compound is present in tobacco smoke.
Introduction Tobacco smoke is causally related to the occurrence of several human cancers at different sites including the respiratory tract, upper digestive tract, bladder, renal pelvis, and pancreas (1). As tobacco smoke is a complex mixture of many thousands of structurally diverse components, it is clear that attribution of cancer risk to a single agent or group of agents is difficult. Nonetheless, a number of tobacco smoke constituents are known to be carcinogenic and include polyaromatic hydrocarbons, tobacco-specific nitrosamines, and some radioactive metal isotopes (1). Moreover, compounds such as the nornicotine derivative 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (MNPB),1 which are metabolically activated to alkylating agents, appear to have a spectrum of carcinogenic activity which strongly suggests that they are primarily responsible for lung carcinogenesis associated with tobacco smoke (2). In order to evaluate the relative importance of the various alkylating agent exposures arising as a consequence of tobacco smoke inhalation, methods for the detection of a range of DNA adducts are required. The 32P-postlabeling approach for “bulky” adducts has demonstrated that exposure to tobacco smoke leads to multiple DNA adducts, characterized by a diagonal * To whom correspondence should be addressed at MRC Toxicology Unit, Hodgkin Building, University of Leicester, P.O. Box 138, Lancaster Rd., Leicester LE1 9HN, U.K. Tel: +44 116 252 5573; Fax: +44 116 252 5616; E-mail:
[email protected]. † IARC. ‡ Present address: MRC Toxicology Unit, Hodgkin Building, University of Leicester, P.O. Box 138, Lancaster Rd., Leicester LE1 9HN, U.K. X Abstract published in Advance ACS Abstracts, February 1, 1996. 1 Abbreviations: 3-alkyladenines (3-alkAde); 3-methyladenine (3MeAde); 3-(2-hydroxyethyl)adenine (3-HOEtAde); 3-ethyladenine (3EtAde); 3-benzyladenine (3-BzAde); phosphate buffered saline (PBS); ethylene oxide (EO); 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (MNPB); N7-(2-hydroxyethyl)guanine (7-HOEtGua), immunoaffinity purification-gas chromatography-mass spectrometry (IA-GC-MS).
0893-228x/96/2709-0439$12.00/0
radioactive zone on two dimensional TLC (3). Exposure of the lungs of smokers to individual polyaromatic hydrocarbons such as benzo[a]pyrene has also been demonstrated (4). Recent improvements in 32P-postlabeling methods for low molecular weight adducts have also shown that tobacco smoke exposure gives rise to methyl- and ethyl-DNA adducts (5, 6). Tobacco smoke contains methylating agents which give rise to increased levels of 7-methylguanine in DNA of total white blood cells, granulocytes, and lymphocytes (6). However, levels of O6-methylguanine were not significantly different in placental DNA from smoking and nonsmoking mothers (7). Numerous experimental studies have shown that tobacco-specific nitrosamines such as NMPB give rise to both methyl and 4-(3-pyridyl)-3-oxobutyl adducts (8). There is clear evidence that specific MNPB-derived DNA adducts are substantially elevated in peripheral lung and tracheobronchial tissues from smokers compared to nonsmokers (9). The objective of the present study was to determine if exposure to tobacco smoke resulted in increased levels of urinary 3-alkyladenines (3-alkAde). 3-AlkAde are formed in DNA by many alkylating carcinogens, such as those present in tobacco smoke, and are removed from DNA either by spontaneous depurination or by the action of glycosylases and excreted in urine (10). The measurement of urinary 3-alkAde is therefore a noninvasive way of detecting and quantifying recent exposure to alkylating carcinogens (10). We have recently developed an immunoaffinity purification-gas chromatography-mass spectrometry (IA-GC-MS) method which allows several 3-alkAde to be determined simultaneously in urine samples and used the method to show that most of the adducts are excreted intact in human urine (11). Some 3-alkAde such as 3-methyladenine (3-MeAde) are present preformed in the diet and contribute substantially to urinary levels (12). This exposure to dietary 3-MeAde © 1996 American Chemical Society
440 Chem. Res. Toxicol., Vol. 9, No. 2, 1996 Scheme 1. Diagrammatic Summaries of the Two Study Designs Used To Investigate the Effect of Cigarette Smoking on Urinary 3-AlkAde Excretion
would be likely to mask any effect of low-level exposures to methylating agents, although there is some evidence that urinary levels of methylated purines are elevated in smokers compared to nonsmokers (14). Consumption of strictly controlled diets, which are low in preformed 3-alkAde, over several days results in the excretion of stable low levels of the adducts (11, 13) and offers the possibility of examining the effect of exposure to low levels of alkylating agents, such as those in tobacco smoke, in volunteers. Preliminary accounts of parts of this work have been published previously (15, 16).
Materials and Methods Recruitment of Subjects. Four subjects were recruited from laboratory staff and nonlaboratory personnel. The criteria for recruitment were that they were regular smokers in good general health and not undergoing any medical treatment during the period of the study. Two men (subject 1, aged 28 years, and subject 4, aged 23 years) and two women (subject 2, aged 29 years, and subject 3, aged 28 years) were recruited, and the study design (see below) was explained to them at a faceto-face interview. The study was approved by the Ethical Committee of the IARC. Study Designs. Two different types of study design were employed during the course of this work and are described in detail below. Study Design 1 (The Effect of Smoking on Urinary 3-AlkAde Excretion in Subjects Consuming Standardized Diets). Three volunteers (subjects 1, 2, and 3) consented to collect consecutive 24-h urine samples during a 10 day period. On days 1, 2, 9, and 10 the volunteers consumed a free-choice diet and were allowed to smoke according to their normal habit. On days 3-8 the volunteers consumed one of two commercially available, nutritionally-balanced, semiliquid diets which had previously been found to contain only low levels of preformed 3-alkAde (11) and bottled water. On days 5 and 6 the volunteers resumed smoking (4-11 cigarettes/day). The study design is summarized in Scheme 1. Study Design 2 (The Effect of Smoking on Urinary 3-AlkAde Excretion in Subjects Consuming Free-Choice Diets). Two volunteers (subjects 1 and 4) were asked to collect consecutive 24-h urine samples for 15 days and to note the number of cigarettes smoked during each 24 h period. They consented to stop smoking for 5 days in the middle of the study. In fact, one volunteer (subject 4) found it difficult to stop smoking for 5 consecutive days and resumed smoking on the fifth day (day 10 of the study). The subjects were allowed to
Prevost and Shuker continue with their normal diet during the study period. The study design is summarized in Scheme 1. Urine Collection. The protocol for the collection, treatment, and storage of 24-h urines (8:00 a.m.-8:00 a.m. the following day) has been described in detail elsewhere (11). Briefly, the urine collection bottles contained sodium azide (sufficient for a final concentration of 0.05% w/v assuming an average of 1 L of urine per day per subject) to inhibit bacterial growth. At the end of the collection period the total volume was noted and two 50 mL aliquots were taken and frozen at -20 °C for subsequent analysis. The remainder of the urine sample was discarded. Analysis of Urinary 3-AlkAde. Methods for the analysis of 3-MeAde, 3-ethyladenine (3-EtAde), 3-(2-hydroxyethyl)adenine (3-HOEtAde), and 3-benzyladenine (3-BzAde) have been described in detail elsewhere (11). Briefly, the pH of aliquots (5 mL) of urine was adjusted to 7.4, and a mixture of four deuterium labeled internal standards of 3-alkAde (d3-3-MeAde, d5-3-EtAde, d4-3-HOEtAde, and d7-3-BzAde, 50 pmol of each) was added. The samples were eluted through a precolumn of Sepharose CL4B directly onto an immunoaffinity column (monoclonal antibody EM-6-47 bound to Sepharose CL4B). After several washing steps, 3-alkAde were eluted from the columns with aqueous acetic acid (1 M). The dried 3-alkAde were converted to volatile tert-butyldimethylsilyl derivatives for analysis by GC-MS. The individual 3-alkAde were quantitated by reference to their respective internal standards. The variation in 3-alkAde levels for repeated analysis of samples using the above method has been found to be less than 5% (13). Data Analysis. The assumption used in the statistical analysis of the results was that the 3-alkAde content of the 24-h urines is an independent variable. This is true if the formation and metabolism of 3-alkAde are rapid. Preliminary studies established that excretion of 3-alkAde in human urine was essentially quantitative within 24 h following administration of deuterium labeled analogues (11), and repair of 3-alkAde in DNA is known to be extremely rapid in many experimental animals (10). Thus, the two experimental designs used in this study correspond to a standard “within-subjects” or “repeatedmeasures” design (17) which is amenable to straightforward analysis of the data. Statistical significance of the results was therefore determined by Student’s t-test using Sigmaplot for Windows (v1.0) software (Jandel Scientific, Erkrath, Germany). Analysis of Urinary Cotinine. Urinary cotinine was determined using a commercially available competitive ELISA assay (Coti-Traq, Serex Inc., Tenafly, NJ, USA) according to the manufacturer’s instructions. Analysis of Tobacco Smoke Condensates for Preformed 3-AlkAde. Each subject was asked to provide a number of cigarettes corresponding to the type used during the study period, and these were analyzed for the presence of preformed 3-alkAde as follows: Tobacco smoke from cigarettes (1-5) was drawn through a solution of phosphate buffered saline (PBS, pH 7.4, 6 mL) using an all-glass apparatus (18) which allowed cigarettes to be burned at approximately the same rate as cigarettes were smoked by the volunteers on the study. The resulting solutions were analyzed for the presence of 3-alkAde by the same method as for urinary analysis (11). Formation of 3-EtAde in Calf Thymus DNA Treated with Tobacco Smoke. Aliquots of calf thymus DNA solutions (5 mg/mL in PBS, pH 7.4, 6 mL) were treated with tobacco smoke (from 1-5 cigarettes) as described above. The treated DNA solutions were heated at 100 °C for 30 min to depurinate alkylated bases. The resulting solutions (5 mL) were subjected to IA-GC-MS analysis for 3-EtAde content as described above.
Results 3-MeAde Excretion in Smokers on a Free-Choice Diet. Twenty-four-hour urine samples were collected for 15 days from a smoker (subject 1) consuming a normal diet who agreed to stop smoking for 5 days in the middle of this period (Study Design 2). 3-MeAde was determined in the urines (Figure 1A), and no significant difference
Cigarette Smoking and 3-Alkyladenines
Chem. Res. Toxicol., Vol. 9, No. 2, 1996 441
to 32 cigarettes a day during the 15 day period of Study Design 2. For four (subject 4) or five (subject 1) consecutive days in the middle of the 15-day period, the subjects did not smoke any cigarettes. Subject 4 excreted significantly more 3-EtAde on days when he smoked than when he did not smoke (p 0.5). However, a significant difference was seen for subject 4 (Figure 1B) who was a heavier smoker then subject 1 (smoking days, 3-MeAde ) 119 ( 20 nmol/day, and nonsmoking days, 3-MeAde ) 76 ( 26 nmol/day, p
