Benzo(a)pyrene metabolism by lymphocytes from normal individuals and individuals carrying the mediterranean variant of glucose‐6‐phosphate dehydrogenase

zyxwvutsrqp zyxwvutsrq zyxwvut zyxwvutsrq zyxwv zyxw zyxwvutsrqp zyxwv Int. J. Cancer: 39, 560-564 (1987) 0 1987 Alan R. Liss, Inc. Publication of the International Union Against Cancer Publication de I’Union lnternationale Contre le Cancer BENZO(a)PYRENE METABOLISM BY LYMPHOCYTES FROM NORMAL INDIVIDUALS AND INDIVIDUALS CARRYING THE MEDITERRANEAN VARIANT OF GLUCOSE-6-PHOSPHATE DEHYDROGENASE F. FEO’, M.E. RUGGIU,L. LENZEFSNI, R. GARCEA,L. DAINO, S. FRASSETTO, V. ADDIS,L. GASPAand R. PASCALE Istituto di Patologia generale dell ’Universitb di Sassari, Via P. Manzella 4 , 07100, Sassari, Italy. In vitro growing human lymphocytes (HL) and fibroblasts, isolated from glucose-6-phosphate dehydrogenase (G6PD).de. ficient subjects (Mediterranean variant), show a sharp decrease in this enzymatic activity and in NADPH:NADP+ ratio. These cells are less able than controls to hydroxylate benzo(a)pyrene (BaP) when tested in the absence of an exogenous NADPH-generating system. They exhibit great resistance to the toxic effect of BaP. G6PD-deficient fibroblasts are less prone than controls to in vitro transformation by BaP. To investigate whether this depends on a decreased production of active BaP metabolites and 6aP:DNA adducts by G6PD-deficient cells, BaP metabolism was studied in G6PDdeficient H L cultured in vitro in the presence of mitogens and treated with BaP for 24 hr. HPLC profiles of organo- and water-soluble metabolites revealed that both types of benzo(a)anthracene (6aA)-induced HL produced: 4,5-, 7.8-, 9,IO-diols. l,3-, 3,6-quinones, 3-, 9-hydroxy and 2 peaks of more polar metabolites. There was a 25-76% decrease of organoand water-soluble metabolites in the G6PD-deficient cells. When H L were incubated with 7,8-diol, the formation of metabolites mutagenic for Salmonella typhimurium (His-) was very low in G6PD-deficient cells. 6aP:deoxyadenosine (dAde) and BaPdeoxyaguanosine (dGua) adducts were identified after incubation of both types of H L with BaP. There was a 31.79% fall in adduct formation by G6PD-deficient cells. Our results indicate that G6PD-deficient human lymphocytes are less able to metabolize BaP than normal lymphocytes. We suggest that the NADPH pool is inadequate, in deficient cells, for active BaP metabolism. Most carcinogens are activated by one or more NADPHdependent reactions catalyzed by the mixed function monooxygenase system (Gelboin, 1967). G6PD activity plays an essential role as a regulator of the hexose monophosphate shunt (HMS) (Horecker, 1967). The latter is the major source of cellular NADPH. Previous work (Feo et al., 1981, 1982, 1984a,b; Pirisi et al., 1982) has shown that human skin fibroblasts and HL carrying the Mediterranean variant of G6PD exhibit a great decrease in HMS as well as in the NADPH/NADP+ ratio. G6PD deficiency protects in vitro growing fibroblasts and HL from the toxic effect of BaP. G6PD-deficient human skin fibroblasts are less prone than normal cells to in vitro transformation by BaP (Feo et al., 1984a,b). NADPH-cytochrome C reductase and aryl-hydrocarbon hydroxylase (AHH) activities are lower in homogenates of G6PD-deficient HL, when tested in the absence of an exogenous NADPH-generating system (Feo et al., 1984b). G6PD-deficient cells also produce lower amounts of water-soluble and mutagenic metabolites than controls after incubation with BaP (Feo et al., 1984a,b). Watersoluble metabolites, formed during detoxification reactions, may arise from proximate carcinogens, such as BaP epoxides and diols, as well as from phenols (Gelboin, 1980). Consequently, the determination of these metabolites does not represent a substantial evaluation of the ultimate BaP metabolite production. Further, the mutagenicity of genotoxic compounds does not always correspond to their carcinogenicity (DeFlora, 1984; Saffiotti et al., 1984). Therefore, a better knowledge of the effect of G6PD deficiency on the transformation process could be obtained by evaluation of the ability of G6PD-deficient cells to produce BaP metabolites which form adducts with DNA. In the current report, the BaP metabolite pattern and the production of BaP:DNA adducts during incubation of G6PD-deficient HL with the carcinogen are studied. Preliminary results have been reported elsewhere (Pirisi et al., 1987). MATERIAL AND METHODS Medium and chemicals Eagle’s minimum essential medium (GIBCO, Mascia Brunelli, Milan, Italy) was supplemented with 24 mM sodium bicarbonate, heparin (60 U/ml), antibiotics (100 IU penicillin/ ml, 100 pg streptomycin/ml and 25 IU nystatiniml) and 10% heat-inactivated fetal bovine serum (Flow, Milan). Phytohemagglutinin and pokeweed mitogen (GIBCO) were added to the culture medium at a final concentration of 1% each. BaP, BaA, DNase (Type II), and snake venom phosphodiesterase (Type IV) were supplied by Sigma (St. Louis, MO). dAde and dGua, alkaline phosphatase (Grade I, lyophilized), RNase A and proteinase K were obtained from Boehringer (Mannheim, FRG). [3H]BaP (52.6 Ci/mmol) was obtained from New England Nuclear (NEN, Dreieichenhain, FRG). BaP and BaA were purified (Van Cantfort et al., 1977) prior to use and added to the medium at a final concentration of 0.01 mM and 0.009 mM, respectively. Some BaP standards: 9,lO-, 4 3 - and 7,8-diols, 7,8-diol-9,1O-epoxide (BaPDE), 1,3 and 3,6-quinones, 9- and 3-hydroxy, were obtained from the Chemical Repository, NCI (Bethesda, MD) and some were a generous gift from Dr. H.V. Gelboin (NCI). [1’,2’-3H]dGua (35 Ci/ mmol) was obtained from Amersham (Little Chalfont, UK) and [8-I4C]dAde (54.9 mCi/mmol) from NEN. They were diluted with the unlabelled nucleosides before use. [1-I4C]glucose (7 mCi/mmol) was purchased from NEN. All other chemicals were of purest grade available. Radioactivity measurements were performed in an LS 1800 liquid scintillation system (Beckman, Geneva, Switzerland) with a Beckman MP solvent. Cell isolation and culture methods Blood samples were obtained from normal males or from males carrying the Mediterranean variant of G6PD. Donors, aged between 20 and 50, were healthy and not undergoing any therapy when samples were collected (between 8 and 9 A.M.). These samples were kept at 0°C for no more than 6 hr before HL isolation. Sixty-six donors for each G6PD activity group were examined for G6PD, HMS, and AHH activities. Two 28-individual subgroups were used for analysis of BaP metabolite pattern and adduct formation, respectively. Total BaP metabolites were determined in both subgroups (56 donors). In order to avoid bias in selecting the subgroups of 28 from the total group studied, donors were chosen each week for one of the two types of analysis (metabolite pattern or adduct formation), alternatively. HL were isolated (Feo et al., 1984b), suspended at 600,000/ml of medium and cultured in 25-cm2 ‘To whom reprint requests should be sent. Received: September 3, 1986 and in revised form November 26, 1986. zyxwvutsrqp zyxwvuts zyxwvutsrqp zyxwvutsrq zyxwvutsr zyxwvu BaP METABOLISM AND G6PD DEFICIENCY 56 1 covalently bound to BaP metabolites were isolated from free nucleosides by elution of a Sephadex LH 20 column with water and then with methanol (Baird and Diamond, 1976). The material eluting with methanol was separated into individual derivatized nucleosides by HPLC using a C18 column maintained at room temperature, with a 1 ml/min constant flow rate. The column was eluted with a 20-min gradient of 5599% methanol. Effluent was monitored at 254 nm and fractions were collected every 20 sec. For the preparation of the ( + )anti-BaPDE-dGua or ( +)anti-BaPDE-dAde markers, BaPDE (3.3 pmol) was dissolved in 1 ml of methanol and 0.5 Assays ml of labelled nucleoside (1.6 pmol; 10 mCi/mmol) in 100 mM G6PD was determined by following NADP+ reduction at TrisCl buffer (PH 7.4) were added. After 16 hr incubation at 340 nm, at 37°C (Feo et al., 1984u), and HMS activity was 37 “ C , the hydrocarbon-deoxynucleosideadducts were isolated determined (Feo et al., 1984a) by measuring the release of (Ashurst and Cohen, 1981). 14C02 from [1-14C] glucose. AHH activity was determined by the radioactive assay developed by Van Cantfort et al. (1977), Mutagenicity experiments Mutagenicity experiments were performed by using His- S. using HL suspensions (lo8 cells/ml). typhimurium TA-100 (Ames et al., 1960). Bacteria were grown Determination of BaP metabolites overnight in Oxoid Nutrient Broth 2, centrifuged for 10 min at HL suspended in culture medium were extracted into 3 2,000 g and resuspended in Hanks’ medium free of C a + + and volumes of ethyl acetate/acetone (2: 1 by vol) containing Mg++ to a titer of 3.5 X 107/ml. The incubations with intact 0.001 % butylated hydroxy-toluene. Additional extractions with HL (4 X 106/ml) or liver S9 fraction (2 mg protein), (Ames et 3 vols of ethyl-acetate were performed until no further radio- al., 1960) were performed for 3 hr at 37°C in the dark, in the active material was released (generally 2 extractions were presence of the indicated amounts of BaP-7,8-diol and 50 pl of sufficient). The organic phases were combined and evaporated bacterial suspension, in a final volume of 1 ml. When BaPDE under vacuum. The residue was dissolved in a small volume was substituted for BaP-7,8-diol, HL and S9 fraction were of methanol and subjected to HPLC analysis. The radioactivity omitted. At the end of incubation the bacterial suspensions of total organo-soluble metabolites was determined in the ex- were plated and the His+ revertant colonies were counted tracts after removal of unreacted BaP by TLC (Selkirk et al., (Feo et al., 1984b). Controls incubated without BaP metabo1975). For determination of water-soluble metabolites, HL lites, HL and S9 fraction were used to calculate the spontasuspensions were homogenized using a Polytron PT-10 ho- neous reversed mutation. mogenizer at full speed (3 bursts of 5 sec each). Homogenates Proteins were determined as previously published (Feo et were combined with 0.2 vols of 1 M acetate buffer (PH 5.5) al., 1984~). and incubated for 30 min at 37°C with 0.5 IU/ml of 0glucuronidase and 0.5 IU/ml of aryl-sulfatase. The reaction RESULTS was stopped by adding 2 vols of ice-cold ethyl acetate:acetone (2: 1) and extracting as above. HPLC profiles of organo-solu- BaP metabolism ble metabolites from untreated samples were subtracted from The G6PD activity of freshly isolated normal and G6PDthose of enzyme-treated samples. The radioactivity of the deficient HL was respectively 26.7 f 4.96 and 0.88 k 0.06 material remaining after extraction of glucuronide and sulfate nmol of NADPH produced per min/mg of protein (n=66 for conjugates as well as of organo-soluble metabolites was taken each type of HL, SD). The HMS activity was respectively 823 as the amount of other conjugated metabolites (for instance, 36 and 242 rt 21 dpm of 14C02 released/min/lO6 cells (n glutathione conjugates). The BaP metabolites were separated = 66 for each type of HL, SD). G6PD and HMS activities did on HPLC (Perkin Elmer Series 10) with an 84-S UV detector not change significantly during 48 hr of culture in the presence and a Perkin Elmer Analytical C18 column (25 X 0.46 cm, of mitogens and BaA. i.d.), and eluted with a 15-min gradient of 50-75% methanol The 2 HL populations used in this study exhibited the same followed by 75% methanol for 15 min and then a 15-min gradient of 75-99% methanol (flow rate: 1 ml/min). Effluent range of AHH activities (1.5-9.0 pmol of BaP hydroxylated/ was monitored at 254 nm and fractions were collected every min/mg of protein, for BaA-induced AHH and 0.63-3.7 pmol/ min/mg protein for non-induced AHH). No differences in 20 sec and used to determine radioactivity. AHH activity were found between the 2 28-individual Isolation and adduct analysis of DNA subgroups (see “Material and Methods”) of both control and In order to isolate DNA from harvested cells, HL were experimental HL, used for analysis of metabolite pattern and suspended in a solution containing 10 mM TrisCl (PH 7), 10 adduct formation. Organo-soluble and water-soluble metabolites produced by mM EDTA and 0.5% sodium dodecylsulfate, incubated for 15 min at 37°C and then for 12 hr with proteinase K (750 pglml). both types of HL were: 9.10-, 4 3 - and 7,8-diols, 1,3- and DNA was extracted once with phenol/chloroform/isoamyl al- 3,6-quinones, 9- and 3-hydroxy (Table I). Two peaks that cohol (25:24: l, by vol), and twice with chloroform/isoamyl eluted before diols, referred to as more polar, are probably alcohol (24:l). It was then precipitated from the collected mixtures of triols and tetrols (Yang et al., 1977). It cannot be organic phases with 0.1 vol of 2 M NaCl and 2 vol of ethanol. excluded, however, that small amounts of ethyl-acetate-exDNA from lo7 cells was dissolved in 1 ml of a solution tractable, sulfate-conjugated metabolites belong to this fraccontaining 10 mM TrisCl (PH 8) and 10 mM EDTA, incubated tion. Total organo-soluble metabolites represented 61-71 % in for 1 hr at 37°C with 1 U / d of heat-inactivated RNase, for 2 both normal and G6PD-deficient HL. Total organo-soluble and hr with proteinase K (350 pg/ml) and then extracted and water-soluble metabolites were about twice as low in the defiprecipitated as above. Purified DNA was dissolved in 10 mM cient cells as in controls. The data in Table I do not take into TrisCl buffer (PH 7.2) and digested to the constituent nucleo- account the fraction of water-soluble metabolites not extractasides with DNase, snake venom phosphodiesterase, and alka- ble with ethyl acetate after digestion with 0-glucuronidase and line phosphatase (Baird and Cohen, 1981). Nucleosides aryl sufatase (see “Material and Methods”). This fraction flasks at 37°C in a humidified atmosphere of 5% COz, in the presence of mitogens and BaA. After 24 hr, HL were harvested by centrifugation (10 min at 320 g) and suspended at 1.2 X lo7 cells per ml in fresh medium containing labelled BaP. Twenty-four hours later the cultures (medium plus cells) were used to extract BaP metabolites. Alternatively, the cells were harvested by centrifugation in phosphate-buffered saline and suspended at 5-7 x lo6 cellslml. The suspensions of washed cells were used for the determination of enzymatic activities and BaP:DNA adducts. zyxwvuts zyxwvut zyxwvut 562 zyxwvutsrqp zyxwvutsrqpo zyxwvutsrqponm FEO ET AL. zyxwvu zyxwvuts zyxwvuts TABLE 1 - FORMATION OF BENZO(a)PYRENE METABOLITES lN,CULTURES OF NORMAL AND G6PDDEFICIENT LYMPHOCYTES pmoli5 x 10’ cells’ Metabolites Normal Total Organo-soluble Water-soluble Organo-soluble More Polar I More Polar I1 9,lO-Diol 4,5-Diol 7,8-Diol 1,6-Quinone 3,6-Quinone 9-Hydroxy 3-H ydroxy Water-soluble More Polar I More Polar I1 9,IO-Diol 4.5-Diol 7,8-Diol 1,6-Quinone 3.6-Quinone 9-Hyhroxy 3-Hydroxy Variant’ 1,456k293 984 172 (67.4) 476k225 (32.6) 721 2244 498k115 (68.0) 234k101 (32.0) 128.2k22.7 (12.9) 63.9k18.6 (6.4) 107.4f20.5 (10.8) 31.2+7.9 (3.1) 77.6k8.6 (7.8) 155.3f26.8 (15.7) 1I5 .O 24.5 (1 I . 6) 88.1 q16.9 (8.9) ’ 233.2k44.7 (22.5) 54.3+ 17.2 (10.6) 23.6+12.7 (4.6) 34.8k 10.6 (6.8) 20.4+9.5 (4.0) 38.4k15.1 (5.5) 62.6 + 3 1.7 (14.2) 90.2+31.8 (17.6) 66.5z28.5 (13.0) 120.5+25.7 (23.6) 112.O k25.3 (25.4) 45.2c26.3 (10.2) 7.2k2.9 (1.6) 29.0+8.1 (6.6) 28.776.7 ( 6 s j 58.7+14.7 (13.3) 112.6& 65.4 (25.5) 15.4k5.2 (3.5) 32.4+ 10.6 (7.3) 34.3k9.8 (15.6) 31.0k6.4 (14.1) 1.4k0.7 (0.6) 7.0+2.6 (3.2) 7.0+2.1 (3.2) 43.3k7.3 (19.48) 68.4k13.6 (31.2) 7.4k1.1 (3.4) 19.5k9.0 (8.9) + + zyxwvu zyxwvutsr ‘Cultures of HL in MEM containing 10%inactivated FBS, mitogens and 2 pg of BaAiml, were exposed to 0.01 SD of 56 determinations, for total metabolites, and of 28 determinations for the metabolite pattern of each type of HL. In parentheses, percentages of total metabolites or percentages of the total organo- and water-soluble metabolites. -”‘t”-test: variant vs. normal, significant for at least p < 0.05 for total, organo-soluble and water-soluble. mM BaP for 24 hr and the metabolites were isolated as described in “Material and Methcds”.-2Means k TABLE 11 - BENZO(a)PYRENE-DNAADDUCTS FORMED IN CYLTURES OF NORMAL AND G6PD-DEFICIENT LYMPHOCYTES Adducts Overall binding I I1 ~~ I11 IV V pmolI5 lo7 cells’ Variant’ Normil 7.8- D 10 L9.10 - E PO X ID E W + 7.8-OIOL 7,8 - D IOL 1 2 1 1 ~2 L 3000 LYMPHOCYTES zyxwvuts 15.23k0.99 1.56+0.15 6.63 1.31 4.2971.71 2.73 k 1.26 1.05 f0.08 + X LOO 6.31+1.51 1.08k0.27 3.34 0.54 0.90 0.06 1.18k0.24 + + 4 a 300 WITHOUT S9 W I T H S9 G6PD* G6PD- [L W a ln g 20( 2 E ‘HL were grown for 24 hr in MEM supplemented with 10% inactivated FBS, W > mitogens and 2 pgiml of BaA. The cultures then received 0.01 miv BaP and 24 hr W later DNA was isolated, purified and hydrolyzed to hydrocarbon-deoxyribonucleo- .x side adducts separated by HPLC.-*Means f S D of 28 determinations with lymphocytes of each type-3“r”-test: Variant vs. normal, significant for at leastp < 0.01 for overall binding and adducts. lo( t FIGURE 1 - Induction of reversed mutation of S. iyphimurium TA100 (His-) by benzo(a)pyrene metabolites. 7,8-dihydrodiol-9,1O-epoxide was present at a final concentration of 2 p g l m l of reaction meamounted to 22.13 3.27 and 17.95 2.95 pmols of BaP dium. 7,8-diol was present at concentrations of 1 , 2 and 4 pg/rnl in the metabolite per 5 x lo7 cells (n = 28 for each type of HL, experiments with fraction S9 and at a concentration of 4 pg/ml in those SD) for normal and deficient HL, respectively, which corre- with lymphocytes. After incubation the bacteria were plated and the sponds to 1.5% and 2.3% of total metabolites produced by number of revertant colonies was counted after 72 hr at 37°C. The normal and deficient cells. It may be calculated from the data mutagenic activity of BaPDE was tested in the absence of fraction S9 in Table I that the percent distribution of organo- and water- and lymphocytes. Data are means of 2 triplicate determinations or means + SD of 5 triplicate determinations. The number of revertants soluble metabolites did not exhibit any major difference be- per plate for the BaPDE-treated bacteria is indicated on the appropritween the 2 types of HL. However, there was a lower percent- ate column. Abbreviations: G6PD+, normal lymphocytes; G6PD-, age of 9,lO-diol among organo-soluble metabolites in the G6PD-deficient lymphocytes. G6PD-deficient HL. A lower percentage of more polar fraction I and diols was also found for water-soluble metabolites. Comparison of the relative amounts of each metabolite showed a 25-67% decrease for organo-soluble metabolites and a 27- Mutagenesis experiments 76 % decrease for the water-soluble ones in the G6PD-deficient The ability of HL to transform the proximate carcinogen HL as compared to controls. 7,8-diol to a mutagenic compound was studied by the Ames zyxwvu zyxwvut 563 BaP METABOLISM AND G6PD DEFICIENCY zyxwvutsrqp zyxwvutsr zyxwvuts zyxwv zyxwvutsr test. As expected (Fig. I), BaPDE was highly mutagenic for S. typhimurium TA-100 (His-) without preliminary activation by the rat liver S9 fraction. In contrast, this activation was necessary to render the BaP-7,8-diol mutagenic. When intact HL were used as a source of the activating enzyme system, a significant reversed mutation was only observed with normal HL . Adduct formation The DNA adducts formed in normal and G6PD-deficient HL are shown in Table 11. It appears that 5 adducts were found with normal HL. Peaks I1 and 111 co-chromatographed with standards derived from ( +)-antiBaPDE and tritium-labelled dGua, while peak IV co-chromatographed with standards derived from ( +)-antiBaPDE and carbonium-labelled dAdE. The other peaks were not identified. Only 4 adducts were formed with G6PD-deficient HL, which corresponded to the first 4 peaks obtained with control cells. Peaks I1 and 111 made up, in both types of HL, about 70% of total organo-soluble adducts, while peak IV constituted 18-19%, for normal and deficient HL. A comparison between normal and G6PD-deficient HL, as concerns the amount of BaP:DNA adducts formed after incubation with BaP, shows that overall adduct formation decreased about 60% in G6PD-deficient HL in relation to controls. Decreases of 31 % , 50%, 79% and 57 % were recorded for peaks I,II,III and IV, respectively, in the deficient cells. DISCUSSION Our results show that the relative percentages of the different organo-soluble and water-soluble metabolites are similar for normal and G6PD-deficient HL, and correspond roughly to those found in normal human HL by Selkirk et al. (1975). This might indicate that no variations of the BaP metabolic patterns take place as a consequence of G6PD deficiency. Further, the decrease in organo-soluble metabolite production by G6PD-deficient cells incubated with BaP does not appear to depend on variations of the ability of these cells to conjugate BaP metabolites. We suggest that the decreased production by G6PD-deficient HL of both organo- and water-soluble metabolites depends on a reduced capacity of these cells to metabolize BaP. This appears to be confirmed by the low ability of the deficient HL to activate BaP-7,8-diol to a compound mutagenic for S. typhimurium TA-100 (His-). This compound is presumably BaPDE, a carcinogen that does not need further activation to give rise to S. typhimurium (His-) reversed mutation (Gelboin, 1980; see also Fig. 1). The decrease of G6PD activity in HL carrying the Mediterranean enzyme variant is coupled with that of the NADPH/ NADP+ ratio and of HMS activity (Feo et al., 1984~).G6PD and HMS, and then presumably also the NADPH level, are apparently not influenced by the stimulation of the monooxygenase system by BaA. This system is involved in different steps of BaP metabolism (Gelboin, 1980). The decrease of the NADPH pool, in G6PD-deficient HL, could explain the great reduction of BaP metabolism in these cells compared to normal HL. In agreement with this interpretation it has recently been observed that ethanol infusion into rats, which causes liver NADPH oxidation (Reinke et al., 1980) strongly inhibits phenol production by BaP (Reinke et al., 1982). In addition, recent evidence, obtained from a permeabilized whole-cell system, suggests that NADPH is rate-limiting in the mixed function oxidation of BaP (Sadowski et al., 1985). The major adduct detected in BaP-treated rodent (see Phillips, 1983) and human (Autrup et al., 1985; Stampfer et al., 1981) cell cultures is the (+)anti-BaPDE:dGua adduct. This represents about 70% of total adducts produced by both G6PDdeficient and normal HL. The other adduct identified was ( +)anti-BaPDE:dAde adduct. Although a correlation between the amount of DNA adducts and the individual carcinogen metabolite yields has not always been observed (Bartley and Stampfer, 1985), it may be that the pattern of carcinogen metabolism is of value in predicting the extent of DNA damage (Jones et al., 1983). In HL carrying the Mediterranean G6PD variant, a decrease of about 50%in organo-soluble BaP metabolite production is associated with a 60% decrease in DNA adduct formation. Many factors may influence the extent of DNA adduct formation, such as the extent of conjugation, the production of 7,8-diol and its subsequent epoxidation and DNA repair. As already noted, metabolite conjugation presumably does not undergo major changes in G6PD-deficient cells as compared to controls, while 7,8-diol production and its subsequent epoxidation appear to be considerably reduced in the deficient cells. However, the influence of DNA repair on the observed phenomena in the 2 types of HL is not known. The decrease in BaP metabolism and DNA adduct formation in G6PD-deficient cells could explain the resistance of these cells to the BaP toxic effect in vitro (Feo et al., 1984b) as well as that of in vitro growing human fibroblasts to BaP toxicity and transforming activity (Feo et al., 1981, 1982, 1984~).In vitro experiments do not completely reproduce in vivo conditions. However, some epidemiological evidence of a decreased tumor incidence in G6PD-deficient subjects has been described (Beaconfield et al., 1965; Naik and Anderson, 1971; Sulis, 1972; Mbensa et al., 1978). In addition, long-term per os treatment with dehydroepiandrosterone, a strong G6PD inhibitor (Lopez and Rene, 1973) which mimics many of the effects of genetically transmitted G6PD deficiency as concerns resistance to carcinogens (Schwartz and Perantoni, 1975; Feo et al., 1981, 1982, 1984a), inhibits the formation of spontaneous (Schwartz, 1979) as well as chemically induced (Schwartz and Tannen, 1981; Nyce et al., 1984) tumors in mice. ACKNOWLEDGEMENTS This investigation was supported by grants from the “Progetto Finalizzato Ingegneria Genetica e Basi Molecolari delle Malattie Ereditarie del CNR” , the “Minister0 Pubblica Istruzione” (programs 40% and 60%) and the “Regione Autonoma Sardegna”. REFERENCES BAIRD,W.M., and COHEN,G.M., The formation and persistence of benzo(a)pyrenemetabolite-deoxyribonucleosideadducts in rat skin in vivo. Int. J. Cancer, 28, 387-392 (1981). L., Effect of 7,s-benoflavone on the forASHURST, S.W., and COHEN,G . M . , In vitro formation of benzo(a)pyrene BAIRD,W.M., and DIAMOND, diol epoxide-deoxyadenosine adducts in the skin of mice susceptible to mation of benzo(a)pyrene-DNA-bound products in hamster embryo cells. benzo(a)pyrene-induced carcinogenesis. Int. J. Cancer, 27, 357-363 Chem.-Biol. Interac., 13,67-75 (1976). (1981). BARTLEY, J.C., and STAMPFER, M . R . , Factors influencing benzo(a)pyrene metabolism in human mammary epithelial cells in culture. CurcinogeneAUTRUP,H., SEREMET, T., ARENHOLT, D . , DRAGSTED, L., and JEPSEN, sis, 6, 1017-1022 (1985). A , , Metabolism of benzo(a)pyrene by cultured rat and human buccal BEACONFIELD, P.,RAINSBURG, R . , and KALTON,G., Glucose-6-phosphate mucosa cells. Carcinogenesis, 6, 1761-1765 (1985). AMES,B.N., MCCANN, J., and YAMASAKI, E . , Methods for detecting carcinogens and mutagens with Sulmonellulmammalianmicrosome mutagenicity test. Mut. Res., 31, 347-364 (1960). 564 zyxwvutsrqpo zyxwvuts zyxwvutsrq zy FEO ET AL dehydrogenase deficiency and the incidence of cancer. Oncologia, 19, 11- tion of mutagenic and carcinogenic metabolites derived from benzo(a)pyrene. Proceedings of the Third Sardinian International Meet19 (1965). S., Detoxification of genotoxic compounds as a threshold mech- ing: Agents and Processes in Chemical Carcinogenesis. Toxicol. Puthol. , DEFLORA, anism limiting their carcinogenicity. Toxicol. Pathol., 12,337-343 (1984). (1987) (In press). L., PASCALE, R., DAINO,L., FRASSETTO, s., LASPINA,v., Z A FEO,F., PIRISI,L., PASCALE, R., DAINO,L., FRASSETTO, S., and GARCEA, PIRISI, NETTI, S., GASPA,L., LEDDA,G.M., GARCEA, R., and FEO, F., Effect of R., Modulatory effect of glucosed-phosphate dehydrogenase deficiency on the benzo(a)pyrene toxicity and transforming activity for in vitro glucose-6-phosphate dehydrogenase deficiency on the benzo(a)pyrene toxicity for in vitro cultured human skin fibroblasts. Res. Cornmun. Chem. cultured human skin fibroblasts. Cancer Res., 38, 3419-3425 (1984~). Puthol. Phurmucol., 38, 301-311 (1982). FEO, F., PIRISI,L., PASCALE, R., DAINO,L., FRASSETTO, S . , ZANETTI, F.C., BELINSKY, S.A., and THURMAN, R.G., S., and GARCEA, R., Modulatory mechanisms of chemical carcinogenesis: REINKE,L.A., KAUFFMAN, the role of the NADPH pool in the benzo(a)pyrene activation. Toxicol. Interactions between ethanol metabolism and mixed-function oxidation in perfused rat liver: inhibition of p-nitroanisole 0-demethylation. J. PhurPathd., 12, 262-268 (1984b). macol. Exp. Tlter., 213,70-78 (1980). FEO, F., PIRISI, L., PASCALE, R., DAINO,L., ZANETTI, S., LASPINA, V., P., KAUFFMAN, F.C., and THURMAN, R.G., and PANI,P., Modulatory mechanisms of chemical carcinogenesis: the REINKE,L.A., MCMANUS, role of the mixed function mono-oxygenase system. In: P. Pani, F. Feo Benzo(a)-pyrene phenol production by perfused rat liver and its inhibition and A. Columbano (eds.), Recent trends in chemical carcinogenesis. by ethanol. Cancer Res., 4, 1681-1685 (1982). I.J., WRIGHT,J.A., and ISRAELS, L.G., A perrneabilized cell Proceedings of the First Sardinian International Meeting on Chemical SADOWSKI, system for studying regulation of aryl hydrocarbon hydroxylase: NADPH Carcinonesis, pp. 179-196, E.S.A., Cagliari (1981). R., ZANETTI, S., DAINO,L, and LASPINA,as rate limiting factor in benzo(a)pyrene metabolism. Int. J. Biochem., FEO, F., PIRISI,L., PASCALE, V., Modulatory effects of glucose-6-phosphate dehydrogenase deficiency 17, 1023-1025 (1985). U., BIGNAMI, M.. BERTOLERO, F., CORTESI,E., FICOKELLA, on the carcinogen activation by in vitro growing human fibroblasts. I n : T. SAFFIOTTI, M.E., Studies on chemically induced neoplastic transforGaleotti, A. Cittadini and S. Papa (eds.), Membranes in rumour growth, C., and KAIGHN, mation and mutation in the BALBi3T3 CI A31-1-1 cell line in relation to pp. 459-557, Elsevier, Amsterdam (1982). the quantitative evaluation of carcinogens. Toxicol. Pathol., 12, 383-390 GELBOIN, H.V., Carcinogens, enzyme induction and gene action. Advanc. ( 1984). Cancer Res., 10, 1-81 (1967). SCHWARTL, A.G., Inhibition of spontaneous breast cancer formation in GELBOIN, H.V., Benzo(a)pyrene metabolism, activation, and carcinogenesis: role and regulation of mixed-function oxidases and related enzymes. female C3H(AVY)mice by long-term treatment with dehydroepiandrosterone. CancerRes., 39, 1129-1131 (1979). Physiol. Rev., 60, 1107-1166 (1980). SCHWARTZ, A.G., and PERANTONI, A,, Protective effect of dehydroepianHORECKER. B.L.. Glucose-6-~hos~hate dehvdroeenase. the Dentose ohos- drosterone against aflatoxin B 1 and 7,12-dimethylbenzanthracene-induced phate cycle, and 'its place in carbohydrate metabuolism.' A m h . J. Puihol., cytotoxicity and transformation in cultured cells. Cancer Res., 35, 248247,261-281 (1967). 2487 (1975). JONES,C.A., SANTELLA, R.M., HUBERMAN, E., SELKIRK, J.K., and SCHWARTL,A.G., and TANNEN. R.H., Inhibition of 7.12-dimethylGRUNBERG, D., Cell specific activation of benzo(a)pyrene by fibroblasts benz(a)anthracene- and urethan-induced lung tumor formation in AiJ mice and hepatocytes. Carcinogenesis, 4, 1351-1357 (1983). by long-term treatment with dehydroepiandrosterone. Curcinogenesis, 2, LOPEZ, S.A., and RENE,A , , Effect of 17-ketosteroids on glucose-6- 1335-1337 (1981). phosphate dehydrogenase activity (G6PD) and on G6PD isoenzymes. SELKIRK, J.K., CROY,R.G., WHITLOCK, J.P., JR., and GELBOIN, H.V., In Proc. SOC. exp. Biol. ( N . Z ) , 142, 258-261 (1973). vitro metabolism of benzo(a)pyrene by human liver microsomes and lymMBENSA, M., RWAKUNDA, C., and VERWILIGHEN, R.L., Glucosed-phos- phocytes. Cancer Res., 35, 3651-3655 (1975). phate dehydrogenase deficiency and malignant hepatoma in Bantu popu- STAMPFER, J.C., M.T., BARTHOLOMEW, I.C., SMITH,H . S . , and BARTLEY, lation. E. Afr. med. J . , 55, 17-19 (1978). Metabolism of benzo(a)pyrene by human mammary epithelial cells: toxNAIK,S.N., and ANDERSON, D.E.. The association between glucose-6- icity and DNA adduct formation. Proc. nut. Acud. Sci. (Wush.), 78, phosphate dehydrogenase deficiency and cancer in American Negroes. 6251-6255 (1981). Oncologia, 25, 356-401 (1971). SULIS,E., Glucose-6-phosphate dehydrogenase and cancer. Luncet, I1 NYCE,J.N., MAGEE,P.N., WILLIAMS, J.R., SOBEL,E.L., and SCHWARTZ1985 (1972). A.G., Inhibition of 1,2-dimethylhydrazine induced colon tumorigenesis VANCANTFORT, J., DE GRAEVE,J., and GIELEN, J.E., Radioactive assay in BALBlc mice by dehydroepiandrosterone. Carcinogenesis, 5 , 57-62 for aryl hydrocarbon hydroxylase. Improved method and biological im(1984). portance. Biochim. biophys. Res. Commun., 79, 505-512 (1977). PHILLIPS, D.H., Fifty years of benzo(a)pyrene. Nature ( L m d . ) ,303, 468- YANG,S.K., ROLLER, P.P., and GELBOIN, H.V., Enzymatic mechanism 472 (1983). of benzo(a)pyrene conversion to diols and phenols and an improved highPIRISI,L., GARCEA,R., PASCALE,R., RUGGIU,M.E., and FEO, F., pressure liquid chromatographic separation of benzo(a)pyrene derivatives. Control of glucose-6-phosphate dehydrogenase deficiency on the forma- Biochemistry, 16, 3680-3687 (1977). zyxwvuts zyxw zyxwvutsrqp zyxwvutsrqpo