Analysis of cigarette-smoke-induced DNA adducts by butanol extraction and nuclease P1-enhanced 32P-postlabeling in human lymphocytes and granulocytes.

In an earlier study, we analyzed the aromatic DNA adducts separated from lymphocytes and granulocytes of smokers and nonsmokers using the nuclease P1-enhanced 32P-postlabeling assay. Here we compare the butanol extraction and nuclease P1-enhanced procedure on the same kind of samples. The DNA adducts of 42 per 10(8) nucleotides from smokers' lymphocytes were statistically higher (p < 0.05) than those of 11 from nonsmokers', when analyzed by the nuclease P1 treatment, but not by the 1-butanol extraction. The radioactivity obtained from the DNA digests on the TLC plates was lower in butanol-treated DNA samples when compared to those of nuclease P1 digestion. Lymphocytes appear to be a suitable test tissue for determining aromatic carcinogen exposure when detecting smoking-related DNA adducts by the nuclease P1-enhanced 32P-postlabeling analysis. ImagesFIGURE 1.FIGURE 2.FIGURE 3.


Introduction
Carcinogen-DNA adducts are currently considered a useful biological marker in risk assessment for potential environmental carcinogens (1). Cigarette smoking is a major cause of lung cancer and other diseases, and it is also an important source of exposure to polycyclic aromatic hydrocarbons (PAHs), aromatic amines, and tobacco-specific nitrosoamines (2)(3)(4). Carcinogenic PAHs of cigarette smoke were analyzed by the 32P-postlabeling assay from various human tissues, and the complexity of the adduct formation was shown in target and nontarget tissues (5)(6)(7)(8)(9). Although the chromatographic profiles obtained by the 32P-postlabeling technique were characteristic for aromatic compounds [benzo[a]pyrene; (8,10)], their identification is only well verified in animals (11). In an earlier study we separated lymphocytes and granulocytes of smokers and nonsmokers and analyzed aromatic DNA adducts by the nuclease P1-enhanced 32P-postlabeling assay (12). This study was carried out with the same lymphocyte and granulocyte DNA samples to investigate whether there are differences between the butanol extraction and the nuclease P1 enhancement procedures before 'Institute of Occupational Health, Topeliuksenkatu 41aA,  Helsinki, Finland.

Materials and Methods
Smokers and controls blood sampling, DNA isolation, nuclease P1-enhanced 32P-postlabeling analysis, solvents used for the chromatographic purification, and the quantification of adduct levels are described elsewhere (12). Four micrograms of lymphocyte DNA from 10 smokers and 8 nonsmokers and granulocyte DNA from 10 smokers and from 9 nonsmokers were treated with the micrococcal nuclease (MN; 0.16 U) and spleen phosphodiesterase (SPD; 1.6 ,ug) at 37°C for 3.5 hr and treated with the nuclease P1 (3 ,ug at 370C for 40 min) according to the method of Reddy and Randerath (13). Butanol extraction was carried out according to the procedure of Gupta (14). After the incubation with MN and SPD, the DNA digests were diluted to 40 ,L1 by adding water (DNA concentration 0.1 ,ug/,uL). For the extraction of adducts, 5 ,ul of 10 mM tetrabutylammonium chloride (TBA), 5 ,LL of 100 mM ammonium formate, pH 3.5, and 80 ,uL water were added (total volume 130 ,LL). The mixture was extracted two times with water-saturated 1-butanol (130 ,L) by mixing in a vortex 30 sec. The phases were separated by centrifuging for 30 sec in a microcentrifuge, and the butanol phases were combined (total volume about 260 ,LL). The butanol extract, first back-extracted two times with the 1-butanol-saturated water, was adjusted with 2.5 pLL of FIGURE 1. Autoradiograms of32P-labeled DNA adducts from lymphocytes and granulocytes of smoker. Lymphocyte DNA ofmale smokers analyzed by (a,b) butanol extraction and (c,d) nuclease P1 treatment. Autoradiography was at -70°C for 3.5 days. Circles marked "x" on panels a and b are the background spots, and the three open circles on panel c are smoking-specific adducts also obtained from the samples in a previous study (12).  Figure 1 shows the autoradiograms obtained by the butanoland nuclease P1-enriched 32P-postlabeling analysis from the same male smokers' lymphocyte and granulocyte DNA, respectively. Lymphocyte DNA obtained by butanol extraction showed a radioactive zone going up to the right corner, consisting of several faint spots (Fig. la). The granulocyte DNA of the same smoker showed only few radioactive spots on the autoradiograms when DNA was analyzed by the butanol extraction (Fig. lb). The intensity of radioactivity was higher in smokers' lympho-  Figure lc shows the smoking-specific adducts detected in the previous study also (12). The radioactivity was very low on the autoradiograms obtained from granulocyte DNA of smokers by nuclease P1 enhanced 32P-postlabeling (Fig. ld). A circle marked as an x-spot in the middle of the autoradiograms from smoker's lymphocytes indicates the background spot ( Fig. la, b). These x-spots were also obtained when calf thymus DNA was analyzed by the butanol extraction procedure (Fig. 2a), but not by the nuclease P1 treatment (Fig. 2b).

Results
The intensity of the spots on the autoradiograms obtained from the lymphocyte and granulocyte DNA of nonsmokers showed very low radioactivity when treated both with nuclease P1 and butanol extraction before 32ppostlabeling ( Fig. 3a-d). Nonsmokers' lymphocyte DNA (Fig. 3a), but not granulocyte DNA, also showed the similar background x-spot that was obtained from smokers' lymphocyte and granulocyte DNA extracted by butanol (Fig. la, b).
The levels of adducts ranged from 4 to 32 per 108 nucleotides in lymphocytes of both smokers and nonsmokers after butanol extraction and from 16 to 98 in smokers and from 15 to 28 in nonsmokers when analyzed by nuclease P1-enhanced 32P-postlabeling ( Table 1). The levels of adducts were low in granulocytes, ranging from 4 to 13 per 108 nucleotides in smokers and from 3 to 13 in nonsmokers after butanol extraction, and from 4 to 14 in smokers and from 5 to 19 in nonsmokers after nuclease P1enhanced procedure ( Table 1). The mean number of DNA adducts/108 nucleotides detected fronm smokers' and nonsmokers' lymphocytes and granulocytes are presented in Figure 4. The error bars indicate the standard deviation of the two to three mean parallel 32P-postlabeled experiments. The total DNA adduct levels of 42 + 7.5/108 nucleotides from smokers' lymphocytes were significantly higher than those of 22 ± 1.9/108 nucleotides from nonsmokers, when analyzed by the nuclease P1-enhanced postlabeling assay ( Table 2). The butanol extraction analysis yielded from the smokers' lymphocytes 20 + 7.5 and from those nonsmokers' lymphocytes 13 + 3.1 DNA adducts per 108 nucleotides ( Table 1). The total DNA adduct levels in granulocytes, when analyzed by both the butanol extraction and nuclease P1-enhanced 32ppostlabeling, were statistically lower (p < 0.05) compared to those oflymphocytes, but showed no statistically significant difference between smokers' and nonsmokers' granulocyte DNA adduct levels ( Table 2). Figure 5 shows the comparison of DNA adduct levels from the lymphocytes of 10 smokers when analyzed by nuclease P1-enhanced 32P-postlabeling in years 1990 and 1991. The relationship of the adduct levels was relatively good, although not statistically significant (y = 0.74x + 20.5, r = 0.55, n = 10).

Discussion
As the nuclease P1 enzyme is thought to dephosphorylate the adducted mononucleotides formed by tobaccospecific nitrosoamines and aromatic amines bound to the C-8 position of guanine, butanol extraction is an alternative method of detecting adducts formed by cigarette smoke (14)(15)(16). When butanol extraction was applied to the DNA samples, all backgrounds of the TLC plates were clean (Fig. la,b compared to c,d), suggesting that butanol did not extract some minor adducts. Overall, lower DNA adduct levels were obtained from smokers' lymphocytes  and granulocytes by the butanol extraction than by the nuclease P1 treatment, perhaps suggesting lower levels of P1-sensitive nitroaromatic and aromatic amine DNA adducts than stable PAH-type adducts (16). Possible reasons for lower adduct levels could also be the loss of material during the butanol extraction steps and discriminating any adducted dinucleotides. Yet, as long as the 32ppostlabeled adducts remain unidentified, it is difficult to assess the selectivity of the two methods (14,16,17).
A statistically significant difference was obtained between the total DNA adduct levels between the smokers' and nonsmokers' lymphocytes analyzed by the nuclease P1 procedure only ( Table 2). The relationship of the adduct data from smokers' lymphoctes was studied, with the results obtained with the same samples in 1990 (Fig. 5). The results of this study confirm the same levels of adducts determined by the nuclease P1-enhanced 32Ppostlabeling analysis in the 2 years (12). The adduct data are also in accordance, irrespective of the pretreatment, with the higher DNA adduct levels in lymphocytes than in granulocytes of smokers. The 32P-postlabeled total white blood cell DNA adducts has been used in several occupational studies as an indicator of PAH exposure. By contrast, no effect of smoking was seen in these studies (18)(19)(20). The human lymphocytes are thus more suitable for use in determining the effect of smoking or, perhaps, in humans occupationally or environmentally exposed to PAHs (12,21).    Adducts/10 nucleotides in 1990 FIGURE 5. Relationship between the DNA adducts/108 nucleotides ana-*lyzed from smokers' lymphocytes in 1990 and 1 year later. The pair of dotted lines represent the 95% confidence limit. Degrees of freedom = 8, correlation coefficients = 0.55, p = 0.07.
We thank the Work Environment Fund of Finland for the financial support.