35
Cancer Letters, 65 (1992) 35 -41 Elsevier Scientific Publishers Ireland Ltd.
Detection of PAH:DNA adducts from auto-oxidation using 32P-postlabeling Piotr Bryla and Eric H. Weyand Rutgers, The State Uniuersity of New
Jersey College of Pharmacy, P.O.Box 789, Piscataway, NJ 088550789
(USA)
(Received 13 December 1992) (Revision received 28 April 1992) (Accepted 30 April 1992)
Summary The binding of benzo[a]pyrene, 7,12dimethylbene[a]anthracene, S-methylcholanthrene, bene[a]anthracene, dibenz[a,c]anthracene and phenanthrene to calf thymus DNA in vitro in the absence of enzymatic or chemical activation was investigated using the 32Ppostlabeling assay. Reactions were performed in the dark or under white light in 1 ml of Tris - HCI buffer (pH 7.51, containing 150 mM KCI, 250 pg of DNA and 0.12 nmol- 600 nmol of hydrocarbon. Reactions were incubated for 1 h at 37OC and the extent of hydrocarbon:DNA adduct formation was determined. With the exception of phenanthrene, all of the hydrocarbons investigated formed DNA adducts that were easily detected with the 32P-postlabeling assay. The multiplicity and level of hydrocarbon:DNA adducts varied for each hydrocarbon. A dose related increase in adduct formation was observed. Adduct levels ranged from 0.07 to 15.28 ad-
ducts per IO7 nucleotides. Highest adduct levels were detected with 7,12-dimethylbenz[a]anthracene (DMBA) and benzo[a]pyrene (B[a]P). Hydrocarbon: DNA adduct formation was enhanced when reactions were performed under white light. A comparison of the adduct levels formed from auto-oxidation and enzymatic activation suggests that 0.05 and 0.26% of the adducts detected in the enzymatic activation of B[a]P and DMBA, can be attributed to auto-oxidation, respectively. These data demonstrate that in the absence of enzymatic or chemical activation, polycyclic aromatic hydrocarbons can undergo autooxidation in vitro and form hydrocarbon:DNA adducts that are detectable with the 32Pposttabeling assay.
Keywords: polycyclic aromatic hydrocarbons; DNA adducts; 32P-postlabeling assay Introduction
Correspondence to: Eric H. Weyand, Rutgers, The State University of New Jersey College of Pharmacy, P.O.Box 789, Piscataway, NJ 08855-0789, USA. Abbreviations: B[a]P, benzo[a]pyrene; DMBA, 7,12dimethylbenz[a]anthracene; 3-MC, 3-methylcholanthrene; B[a]A, benz[a]anthracene; DB[o,c]A, dibenz[a,c]anthracene; PEI-cellulose TLC, polyethyleneimine-cellulose thin-layer chromatography; PAH, polycyclic aromatic hydrocarbon. 0304.3835/92/$05.00 Printed and Published
0 1992 Elsevier Scientific Publishers in Ireland
Extensive studies have been directed towards the elucidation of macromolecular adducts that are critical for the initiation of chemical carcinogenesis. In particular, emphasis has been placed on chemical:DNA adduct formation since these modifications have Ireland Ltd
36
been correlated to mutation or malignant transformation [1,2]. The importance of detecting and identifying chemical:DNA adducts has led to the development of sensitive methods for the analysis of DNA adducts. Fluorescence [3] and immunoassay [4] techniques have proved to be valuable methods for the detection of specific DNA adducts. More recently, the introduction of the 32P-postlabeling assay for the detection of DNA adducts has proven to be an extremely sensitive method which is applicable to a wide variety of compounds [5]. Extensive studies have used the 32P-pastlabeling assay to detect chemical:DNA adducts in tissues isolated from in vivo systems such as laboratory animals and humans [6 - 81. In addition, studies have also used the 32Ppostlabeling assay to evaluate DNA adduct formation using various in vitro systems [9, lo]. Masento and co-workers [ll] recently demonstrated a potential limitation of the 32Ppostlabeling assay when used for detecting DNA adducts formed in vitro. These investigators determined that non-adducted tetrol derivatives of polycyclic aromatic hydrocarbons (PAHs) give rise to 32P-labeled products during 32P-postlabeling analysis. These data clearly demonstrate the need for extensive DNA purification prior to 32P-postlabeling analysis. This is particularly important when DNA is modified by reactive electrophiles in vitro. We have recently reported that B[a]P can undergo auto-oxidation and form DNA adducts that can be detected with the 32Ppostlabeling assay [12]. The formation and detection of unanticipated hydrocarbon:DNA adducts, formed as a result of non-enzymatic activation, could influence the interpretation and value of data obtained from in vitro activation systems. Thus, the present study was undertaken to provide insight into the ability of the 32P-postlabeling assay to detect PAH: DNA adduct formation in the absence of enzymatic or chemical activation. Materials and Methods Chemicals
Analytical
grade
benzo[a]pyrene
(B[a]P),
7,12-dimethylbenz[a]anthracene (DMBA), 3methylcholanthrene (3-MC), benz[a]anthracene, (B[a]A), dibenz[a,c]anthracene (DB[a,c]A) and phenanthrene, were purchased from Aldrich Company (Milwaukee, WI). B[a]P and DMBA were further purified on an alumina column [13] and determined to be greater than 96% pure using high performance liquid chromatography. All buffer solutions were prepared with 18 rn0 ultra purified water from a five bowl Milli-Q Water Purification System (Millipore Company, NJ) and purified through a Chelex 100 ion exchange resin (Bio-Rad Laboratories, Richmond, CA) to remove trace metals [14,15]. Horseradish peroxidase, type II (Rz = 1.47)) calf thymus DNA (type I) and 30% hydrogen peroxide solution were purchased from Sigma Chemical Co. (St. Louis, MO). Machery Nagel polyethyleneiminecellulose (PEI-cellulose) TLC plates were purchased from Bodman Chemicals (Aston, PA) while all biochemicals used for 32Pas previously postlabeling were obtained described [7]. Calf thymus DNA was determined to be iron free by atomic absorption spectroscopy which was provided by Dr. Thomas Medwick and Mr. Kin T. Tang in the Department of Pharmaceutical Chemistry, Rutgers University. Reaction
conditions
and adduct
analysis
Reaction mixtures (1 ml) contained 150 mM KCI, 50 mM Tris - HCI (pH 7.5) and 250 pg DNA. Reactions were initiated by the addition of 600 nmol of each hydrocarbon in 10 ,uI DMSO or 120, 12 or 0.12 nmol of each hydrocarbon in 5 ~1 DMSO. Reactions were incubated in the dark or with white light, with shaking at 37OC for 1 h. Reactions were stopped by the addition of 3.2 ml of phenol. Reaction mixtures were vortexed and the aqueous phase was removed and extracted with an equal volume of phenol:chloroform:isoamyl alcohol (50:48:2, by vol.) followed by an equal volume of chloroform:isoamyl alcohol (24:1, v/v). DNA was isolated by ethanol precipitation and quantified spectrophotometrically at 260 nm. The horseradish peroxidase/hydrogen peroxide enzymatic activation
37
of B[a]P and DMBA were performed using a slightly modified procedure to that described by Cavalieri and co-workers [16,17]. Mixtures (1 ml) containing 150 mM KCI, 50 mM TrisHCI buffer (pH 7.0)) 250 pg calf thymus DNA, 250 c(g HRP and 120 nmol B[a]P or DMBA were preincubated for 3 min at 37OC. Reactions were initiated by the addition of 0.5 pmol HzOz and incubations continued for 1 h. Reactions were stopped by the addition of 100 ~1 Tris-HCI (pH 8.0), 35 111100 mM EDTA, 100 ~1 10% SDS and 3.2 ml buffer saturated phenol. DNA was isolated by ethanol precipitation and quantified spectrophotometrically at 260 nm. Az60/zm ratios were in the range of 1.65- 1.70. Hydrocarbon: DNA adduct formation was evaluated using the Nuclease Pr variant of the 32P-postlabeling method [l&19]. In brief, DNA (20 pg) was hydrolyzed to deoxyribonucleotides with micrococcal endonuclease and spleen phosphodiesterase followed by digestion with nuclease Pi to convert unmodified nucleotides to nucleosides. Hydrocarbondeoxyribonucleotides (5 pg) were 32P-labeled using 25 &i of carrier free [T-~~P]ATP. 32Plabeled deoxyribonucleoside-3’ ,5’-bisphosphates were chromatographed on 10 cm x 10 cm PEI-cellulose TLC plates using a modified four solvent developing system. Solvents used were Dl: 1 .O M sodium phosphate (pH 6.5), D2: not performed, D3: 5.3 M lithium formate (pH 3.5) containing 8.5 M urea, D4: 1.2 M lithium chloride, 0.5 M Tris- HCI (pH 8.0) containing 8.5 M urea, D5: 1.7 M sodium phosphate (pH 6.0). Following development, spots of radioactivity were located by autoradiography using Kodak XOmat AR film with intensifying screens. Spots of radioactivity were removed from TLC plates and the amount of radioactivity in each determined by liquid scintillation counting. Background levels of radioactivity were determined by counting blank areas of each chromatogram. Background radioactivity was subtracted from radioactivity associated with adducts and relative adduct levels were calculated as described by Reddy and Randerath [19]. The specific activity of the
[T-~~P]ATP was determined by measuring the incorporation of radioactivity into a known amount of dAp. Results and Discussion
The reaction of B[a]P, DMBA, 3-MC, B[a]A, or DB[a,c]A with calf thymus DNA in a buffer system resulted in hydrocarbon:DNA adducts that were easily detected with the 32Ppostlabeling assay. In contrast, phenanthrene was the only hydrocarbon evaluated that did not form DNA adducts when reacted with calf thymus DNA in vitro (maps not shown). The lack of adduct formation with phenanthrene may be related to its high ionization potential (8.19 eV) and in ability to form radical cations. Cavalieri and Rogan have demonstrated that polycyclic aromatic hydrocarbons with ionization potentials above 7.57 eV do not readily form radical cations in vitro [23]. This is in contrast to hydrocarbons with lower ionization potentials which are capable of forming radical cations. The detection of DNA adducts with B[a]P, DMBA, 3-MC and B[a]A is consistent with ionization potential below 7.57 and the ability of these compounds to form radical cations. The patterns of DNA adducts detected varied considerably between hydrocarbons (Fig. 1). The most complex adduct pattern was observed with DMBA. It is interesting to note that the adduct profiles observed for both B[a]P and DMBA in the present study, are similar to previously reported adduct profiles generated using enzymatic activation systems. In the case of B[a]P, adduct profiles are similar to those previously reported by Bodell and coworkers [20] who used a horseradish peroxidase/hydrogen peroxide (HRP/HzO,) in vitro activation system (Fig. 1, map 2 and 8). This similarity is not surprising since the mechanism of B[a]P activation by autooxidation or HRP/H202 catalysis is likely similar with the involvement of reactive oxygen species. Thus, one would expect to form similar types of B[a]P:DNA adducts. In the case of DMBA, however, the resemblance between adduct patterns observed in this study
PEI-cellulose TLC maps of hydrocarbon:DNA adducts formed in vitro in the absence of enzymatic or chemical activation. Maps performed under white light while maps 8 - 14 represent reactions performed in the dark. The origin is located at the bottom chromatogram and was excised prior to autoradiography. D3 solvent development was in the direction from left to right while was in the direction from bottom to top. Autoradiography was performed at -80°C for the time indicated above.
6: 24hr 12: 48 hr 5: 24 hr 11: 48 hr 4: 3 hr 10: 3 hr
3: 3hr 9: 48 hr
2: 3hr 8: 48 hr
1: 24 hr 7: 24 hr
Fis. 1. Representative 1 - 7 represent reactions left-hand corner of each D4 solvent development
DB[a,c]A
BMA 3-MC
DMBA
JWP
Control
39
and others [ZZ], does not necessarily suggest a similarity in activation mechanism (Fig. 1, map 3 and 9). DMBA:DNA adduct profiles reported by other investigators are considered products of microsomal activation (a twoelectron oxidation mechanism) while the adducts detected in the present study most likely from a one-electron oxidation arise mechanism. Any resemblance of adduct profiles may be due to the low resolution capacity of PEI-cellulose TLC for hydrocarbon:DNA adducts derived from PAH with similar molecular weights. A dose related increase in hydrocarbon:DNA adduct formation was observed when hydrocarbon levels were varied between 0.12 nmol and 600 nmol per reaction (Table 1). The level of adducts detected when reactions were performed in the dark ranged from 0.07 to 1.70 adducts per lo7 nucleotides. Hydrocarbon: DNA adduct formation was enhanced when reactions were performed under white light. Adduct levels in these reactions ranged from 0.09 to 15.28 adducts per lo7 nucleotides. The most notable enhancement, was observed with 120 nmol B[a]P and 12 nmol DMBA. The enhancement of adduct formation by white light was anticipated since PAH are known to undergo extensive photooxidation [23,24]. The lower levels of adducts observed in reactions with 600 nmol hydrocarbon may be due to hydrocarbon precipitation or the increased volume of DMSO in reactions. DMSO is known to be a scavenger of hydroxy radicals which may be involved in the auto-oxidation and binding of hydrocarbons to DNA in our reactions. The relative contribution of adduct formation from auto-oxidation in an in vitro enzymatic activation system was investigated for B[a]P and DMBA. Reactions were performed in the dark with 120 nmol of B[a]P or DMBA using a HRP/H202 activation system. B[a]P formed 139 adducts per lo7 nucleotides while DMBA formed 148 adducts per lo7 nucleotides, respectively. Comparing these results with the levels of adducts formed in the absence of enzymatic activation suggests that
Table 1. Dose (nmol)
Levels of PAH:DNA adducts formed in vitro. Adducts/ 10’ Dark
Nucleotides Light
Benzo[a]pyrene 0.56 5.45 600 0.07 2.21 120 0.19 1.37 12 ND 0.17 0.12 7,12-Dimethylbenz[a]anthracene 1.70 15.28 600 0.39 10.50 120 0.21 6.84 12 0.17 0.18 0.12 3-Methylcholanthrene 1.32 1.21 600 0.38 0.54 120 0.12 0.58 12 0.11 0.12 0.12 Benz[a]anthracene 0.45 0.95 600 0.12 0.80 120 0.43 0.09 12 ND 0.15 0.12 Dibenz[a,c]anthracene 1.14 1.76 600 0.32 0.57 120 0.13 0.06 12 0.08 ND 0.12
Adduct ratio Light/dark
9.7
31.6 7.2 9.0 26.9 33.6 1.1 0.9 1.4 4.8 1.1 2.1 6.7 0.2 1.5 1.8 0.5 -
ND = no adduct formation detected. Note: No phenanthrene:DNA adducts were detected when 120 Fmol-600 nmol of phenanthrene was reacted with calf thymus DNA in vitro.
between 0.05 and 0.26 percent of the adducts formed in the enzymatic system may arise from auto-oxidation of B[a]P and DMBA, respectively. If reactions were performed under white light the contribution of auto-oxidation to adduct formation could account for 1.6 and 7.1% of total adducts, respectively. The levels and profiles of hydrocarbon:DNA adducts formed following enzymatic activation in our studies were consistent with results reported by other investigators [21,25]. HRP/HsOs activation of B[a]P or DMBA has been reported to result in 161 and 136 adducts per lo7
40
nucleotides, respectively. The extent of DNA binding following microsomal activation has also been reported. Microsomal activation of B[a]P results in 118 adducts per lo7 nuclcotides while DMBA activation formed 1.2 adducts per lo7 nucleotides. The mechanism responsible for hydrocarbon activation and DNA adduct formation in reactions performed in the dark is not fully understood. A series of events such as PAH auto-oxidation and subsequent formation of reactive oxygen species is one possible explanation that could account for hydrocarbon oxidation and DNA binding. Recently, we have reported that Oz :, OH*, HzOz and t O2 are involved in B[u]P auto-oxidation and binding to DNA in vitro in the absence of enzymatic or chemical activation [12]. It is likely that these reactive oxygen species are also involved in the mechanism responsible for the hydrocarbon:DNA adducts observed in the present study with DMBA, 3-MC, B[a]A and DB[a,c]A. Recent studies have demonstrated that hydrocarbon oxidation by one-electron oxidation results in both stable and labile hydrocarbon:DNA adducts [21,25]. The labile hydrocarbon:DNA adducts would not be detected in our experiments due to the limitation of the 32P-postlabeling assay. Since oneelectron oxidation is inherent in an autooxidation mechanism, the extent of hydrocarbon:DNA adducts formed as a result of auto-oxidation may be greater, since labile DNA adducts would have went undetected in our studies. However, these results clearly demonstrated that auto-oxidation does contribute to stable hydrocarbon:DNA adduct formation. In summary, we have demonstrated that polycyclic aromatic hydrocarbons can undergo auto-oxidation and form DNA adducts that are readily detected with the 32P-postlabeling assay. Adduct formation is greatly enhanced in the presence of white light. Hydrocarbon autooxidation may account for up to 0.26% of total DNA adduct formation when hydrocarbon binding to DNA is evaluated in the dark using an in vitro enzymatic activation system.
The presence of white light may greatly enhance the contribution of auto-oxidation to hydrocarbon: DNA adduct formation. Hydrocarbon:DNA adduct patterns that result from PAH auto-oxidation may resemble adduct patterns observed from enzymatic activation systems. Thus, the potential for detecting PAH:DNA adducts as a result of autooxidation should be taken into account when investigating the activation and DNA binding of polycyclic aromatic hydrocarbons in vitro. Acknowledgments This work was supported by USPHS Grant R29 CA49826 from the National Cancer Institute. References Miller, J.A. and Miller, E.C. (1981) Searches for ultimate chemical carcinogens and their reactions with cellular macromolecules. Cancer, 47, 2327 - 2345. Wogan, G.N. and Gorelick, N.J. (1985) Chemical and biochemical dosimetry of exposure of genotoxic chemicals. Environ. Health Perspect., 62, 5- 18. Vanhakangas, K., Haugen, A. and Harris, C.C. (1985) An applied synchronous fluorescence spectrophotometric assay to study benzo[o]pyrene diolepoxide-DNA adducts. Carcinogenesis, 6, 1109 - 1116. Santella, R.M., Gasparo, F. and Hsieh, L. (1987) Quantitation of carcinogen-DNA adducts with monoclonal antibodies. Prog. Exp. Tumor Res., 31, 63 - 75. Gupta, R.C., Reddy, M.V. and Randerath, K. (1982) 32P-Postlabeling analysis of non-radioactive aromatic carcinogen-DNA adducts. Carcinogenesis. 3, 10811092. Phillips, D.H., Hewer, A., Martin, C.N., Gamer, R.C. and King, M.M. (1988) Correlation of DNA adduct levels in human lung with cigarette smoking. Nature, 336, 790 - 792. Weyand, E.H., Rice, J.E. and LaVoie, E.J. (1987) 32Ppostlabeling analysis of DNA adducts from non-alternate PAH using thin-layer and high performance liquid chromatography. Cancer Lett., 37, 257 - 266. Sophie, L., Ni She, M., Hewer, A., Grover, P.L., Platt, K.L., Oesch, F. and Phillips, D.H. (1991) The metabolic activation of dibenz[a,h]anthracene in mouse skin examined by 32P-postlabeiling: minor contribution of the 3, 4dial 1,2-oxides to DNA binding. Carcinogenesis, 12, 1079 - 1083. Canella, K., Peltonen, K. and Dipple, A. (1991) Identification of (+) and (-1 anti-benzo[a]pyrene dihydrodiol epoxide-nucleic acid adducts by the 32P-postlabeling assay. Carcinogenesis, 12, 1109- 1114.
41
10
11
12
13
14
15
16
17
18
Bodell, W.J., Devanesan, P.D., Rogan, E.G. and 32P-Postlabeling analysis of Cavalieri, E.L. (1989) benzo[a]pyrene-DNA adducts formed in vitro and in vivo. Chem. Res. Toxicol., 2, 312-315. Masento, M.S., Hewer, A., Grover, P.L. and Phillips, D.H. (1989) Enzyme-mediated phosphorylation of polycyclic hydrocarbon metabolites: detection of nonassay. Caradduct compounds in the 32P-postlabeling cinogenesis, 10, 1557 - 1559. Bryla, P. and Weyand, E.H. (1991) Role of activated oxygen species in benzo[a]pyrene:DNA adduct formation in vitro. Free Rad. Biol. Med., 11, 17 - 24. VanCanntfort, J., DeGraeve, J. and Gielen, J.E. (1977) Radioactive assay for aryl hydrocarbon hydroxylase: improved method and biological importance. Biochem. Biophys. Res. Commun., 79, 505-512. Minotti, G. and Aust, S.D. (1989) The role of iron in oxygen radical mediated lipid peroxidation. Chem.-Biol. fnteract., 71, l- 19. Minotti, G. and Aust, S.D. (1987) Superoxide-dependent redox cycling of citrate-Fe 3+. Evidence for a superoxide dismutase like activity. Arch. Bioch. Biophys., 253, 257 - 267. Rogan, E.G., Katomski, P.A., Roth, R.W. and Cavalieri, E.L. (1979) Horseradish peroxidase/hydrogen peroxidecatalyzed binding of aromatic hydrocarbons to DNA. J. Biol. Chem., 254, 7055- 7059. Cavalieri, E.L., Rogan, E.G., Roth, R.W., Sauger, R.K. and Hakam, A. (1983) The relationship between ionization potential and horseradish peroxidase/hydrogen peroxidecatalyzed binding of aromatic hydrocarbons to DNA. Chem.-Biol. Interact., 47, 87 - 109. Dunn, B.P. and Stich, H.F. (1986) 32P-Postlabeling analysis of aromatic DNA adducts in human oral mucosal cells. Carcinogenesis, 7, 115 - 1120.
19
20
21
22
23
24
25
Reddy, M.V. and Randerath, K. (1986) Nuclease Plmediated enhancement of sensitivity of 32P-postlabeling test for structurally diverse DNA adducts. Carcinogenesis, 7, 1543- 1551. Cavalieri, E. and Rogan, E. (1985) Role of radical cations in aromatic hydrocarbon carcinogenesis. Environ. Health Perspect., 64, 69-84. Bodell, W.J., Devanesan, P.D., Rogan. E.G. and 32P-Postlabeling analysis of Cavalieri, E.L. (1989) benzo[a]pyrene-DNA adducts formed in vitro and in vivo. Chem. Res. Toxicol., 2, 312-315. Schmeiser, H., Dipple, A., Schurdak, M.E., Randerath, E. and Randerath, K., (1988) Comparison of 32Ppostlabeling and high pressure liquid chromatographic analyses for 7, 12-dimethylbenz[a]anthracene-DNA adducts, Carcinogenesis, 9, 633 - 638. Shevchuk, 1. and Goubergrits, M. (1991) Reactivity of alternate polyarenes under photooxidation, In: Polycyclic Aromatic Hydrocarbons Measurements, Means and Metabolism, pp. 831- 838. Editors: M. Cooke, K. Loening and J. Menit, Battelle Press, Columbus, Ohio. Vu-Du, T. and Huynh, C.K. (1991) Photodecomposition rates of adsorbed PAH and mutagenic activity of irradiated synthetic mixtures, In: Polycyclic Aromatic Hydrocarbons Measurements, Means and Metabolism, pp. 979 - 994. Editors: M. Cooke, K. Loening and J. Merrit, Battelle Press, Columbus, Ohio. RamaKrishna, N.V.S., Devanesan, P.D., Rogan, E.G., Cavalieri, E.L., Jeong, X., Jankowiak, R. and Small, G.J. (1992) Mechanism of metabolic activation of the potent carcinogen 7,12-dimethylbenz[a]anthracene. Chem. Res. Toxicol., 5, 220 - 226.
Cancer Letters, 65 (1992) 35 -41 Elsevier Scientific Publishers Ireland Ltd.
Detection of PAH:DNA adducts from auto-oxidation using 32P-postlabeling Piotr Bryla and Eric H. Weyand Rutgers, The State Uniuersity of New
Jersey College of Pharmacy, P.O.Box 789, Piscataway, NJ 088550789
(USA)
(Received 13 December 1992) (Revision received 28 April 1992) (Accepted 30 April 1992)
Summary The binding of benzo[a]pyrene, 7,12dimethylbene[a]anthracene, S-methylcholanthrene, bene[a]anthracene, dibenz[a,c]anthracene and phenanthrene to calf thymus DNA in vitro in the absence of enzymatic or chemical activation was investigated using the 32Ppostlabeling assay. Reactions were performed in the dark or under white light in 1 ml of Tris - HCI buffer (pH 7.51, containing 150 mM KCI, 250 pg of DNA and 0.12 nmol- 600 nmol of hydrocarbon. Reactions were incubated for 1 h at 37OC and the extent of hydrocarbon:DNA adduct formation was determined. With the exception of phenanthrene, all of the hydrocarbons investigated formed DNA adducts that were easily detected with the 32P-postlabeling assay. The multiplicity and level of hydrocarbon:DNA adducts varied for each hydrocarbon. A dose related increase in adduct formation was observed. Adduct levels ranged from 0.07 to 15.28 ad-
ducts per IO7 nucleotides. Highest adduct levels were detected with 7,12-dimethylbenz[a]anthracene (DMBA) and benzo[a]pyrene (B[a]P). Hydrocarbon: DNA adduct formation was enhanced when reactions were performed under white light. A comparison of the adduct levels formed from auto-oxidation and enzymatic activation suggests that 0.05 and 0.26% of the adducts detected in the enzymatic activation of B[a]P and DMBA, can be attributed to auto-oxidation, respectively. These data demonstrate that in the absence of enzymatic or chemical activation, polycyclic aromatic hydrocarbons can undergo autooxidation in vitro and form hydrocarbon:DNA adducts that are detectable with the 32Pposttabeling assay.
Keywords: polycyclic aromatic hydrocarbons; DNA adducts; 32P-postlabeling assay Introduction
Correspondence to: Eric H. Weyand, Rutgers, The State University of New Jersey College of Pharmacy, P.O.Box 789, Piscataway, NJ 08855-0789, USA. Abbreviations: B[a]P, benzo[a]pyrene; DMBA, 7,12dimethylbenz[a]anthracene; 3-MC, 3-methylcholanthrene; B[a]A, benz[a]anthracene; DB[o,c]A, dibenz[a,c]anthracene; PEI-cellulose TLC, polyethyleneimine-cellulose thin-layer chromatography; PAH, polycyclic aromatic hydrocarbon. 0304.3835/92/$05.00 Printed and Published
0 1992 Elsevier Scientific Publishers in Ireland
Extensive studies have been directed towards the elucidation of macromolecular adducts that are critical for the initiation of chemical carcinogenesis. In particular, emphasis has been placed on chemical:DNA adduct formation since these modifications have Ireland Ltd
36
been correlated to mutation or malignant transformation [1,2]. The importance of detecting and identifying chemical:DNA adducts has led to the development of sensitive methods for the analysis of DNA adducts. Fluorescence [3] and immunoassay [4] techniques have proved to be valuable methods for the detection of specific DNA adducts. More recently, the introduction of the 32P-postlabeling assay for the detection of DNA adducts has proven to be an extremely sensitive method which is applicable to a wide variety of compounds [5]. Extensive studies have used the 32P-pastlabeling assay to detect chemical:DNA adducts in tissues isolated from in vivo systems such as laboratory animals and humans [6 - 81. In addition, studies have also used the 32Ppostlabeling assay to evaluate DNA adduct formation using various in vitro systems [9, lo]. Masento and co-workers [ll] recently demonstrated a potential limitation of the 32Ppostlabeling assay when used for detecting DNA adducts formed in vitro. These investigators determined that non-adducted tetrol derivatives of polycyclic aromatic hydrocarbons (PAHs) give rise to 32P-labeled products during 32P-postlabeling analysis. These data clearly demonstrate the need for extensive DNA purification prior to 32P-postlabeling analysis. This is particularly important when DNA is modified by reactive electrophiles in vitro. We have recently reported that B[a]P can undergo auto-oxidation and form DNA adducts that can be detected with the 32Ppostlabeling assay [12]. The formation and detection of unanticipated hydrocarbon:DNA adducts, formed as a result of non-enzymatic activation, could influence the interpretation and value of data obtained from in vitro activation systems. Thus, the present study was undertaken to provide insight into the ability of the 32P-postlabeling assay to detect PAH: DNA adduct formation in the absence of enzymatic or chemical activation. Materials and Methods Chemicals
Analytical
grade
benzo[a]pyrene
(B[a]P),
7,12-dimethylbenz[a]anthracene (DMBA), 3methylcholanthrene (3-MC), benz[a]anthracene, (B[a]A), dibenz[a,c]anthracene (DB[a,c]A) and phenanthrene, were purchased from Aldrich Company (Milwaukee, WI). B[a]P and DMBA were further purified on an alumina column [13] and determined to be greater than 96% pure using high performance liquid chromatography. All buffer solutions were prepared with 18 rn0 ultra purified water from a five bowl Milli-Q Water Purification System (Millipore Company, NJ) and purified through a Chelex 100 ion exchange resin (Bio-Rad Laboratories, Richmond, CA) to remove trace metals [14,15]. Horseradish peroxidase, type II (Rz = 1.47)) calf thymus DNA (type I) and 30% hydrogen peroxide solution were purchased from Sigma Chemical Co. (St. Louis, MO). Machery Nagel polyethyleneiminecellulose (PEI-cellulose) TLC plates were purchased from Bodman Chemicals (Aston, PA) while all biochemicals used for 32Pas previously postlabeling were obtained described [7]. Calf thymus DNA was determined to be iron free by atomic absorption spectroscopy which was provided by Dr. Thomas Medwick and Mr. Kin T. Tang in the Department of Pharmaceutical Chemistry, Rutgers University. Reaction
conditions
and adduct
analysis
Reaction mixtures (1 ml) contained 150 mM KCI, 50 mM Tris - HCI (pH 7.5) and 250 pg DNA. Reactions were initiated by the addition of 600 nmol of each hydrocarbon in 10 ,uI DMSO or 120, 12 or 0.12 nmol of each hydrocarbon in 5 ~1 DMSO. Reactions were incubated in the dark or with white light, with shaking at 37OC for 1 h. Reactions were stopped by the addition of 3.2 ml of phenol. Reaction mixtures were vortexed and the aqueous phase was removed and extracted with an equal volume of phenol:chloroform:isoamyl alcohol (50:48:2, by vol.) followed by an equal volume of chloroform:isoamyl alcohol (24:1, v/v). DNA was isolated by ethanol precipitation and quantified spectrophotometrically at 260 nm. The horseradish peroxidase/hydrogen peroxide enzymatic activation
37
of B[a]P and DMBA were performed using a slightly modified procedure to that described by Cavalieri and co-workers [16,17]. Mixtures (1 ml) containing 150 mM KCI, 50 mM TrisHCI buffer (pH 7.0)) 250 pg calf thymus DNA, 250 c(g HRP and 120 nmol B[a]P or DMBA were preincubated for 3 min at 37OC. Reactions were initiated by the addition of 0.5 pmol HzOz and incubations continued for 1 h. Reactions were stopped by the addition of 100 ~1 Tris-HCI (pH 8.0), 35 111100 mM EDTA, 100 ~1 10% SDS and 3.2 ml buffer saturated phenol. DNA was isolated by ethanol precipitation and quantified spectrophotometrically at 260 nm. Az60/zm ratios were in the range of 1.65- 1.70. Hydrocarbon: DNA adduct formation was evaluated using the Nuclease Pr variant of the 32P-postlabeling method [l&19]. In brief, DNA (20 pg) was hydrolyzed to deoxyribonucleotides with micrococcal endonuclease and spleen phosphodiesterase followed by digestion with nuclease Pi to convert unmodified nucleotides to nucleosides. Hydrocarbondeoxyribonucleotides (5 pg) were 32P-labeled using 25 &i of carrier free [T-~~P]ATP. 32Plabeled deoxyribonucleoside-3’ ,5’-bisphosphates were chromatographed on 10 cm x 10 cm PEI-cellulose TLC plates using a modified four solvent developing system. Solvents used were Dl: 1 .O M sodium phosphate (pH 6.5), D2: not performed, D3: 5.3 M lithium formate (pH 3.5) containing 8.5 M urea, D4: 1.2 M lithium chloride, 0.5 M Tris- HCI (pH 8.0) containing 8.5 M urea, D5: 1.7 M sodium phosphate (pH 6.0). Following development, spots of radioactivity were located by autoradiography using Kodak XOmat AR film with intensifying screens. Spots of radioactivity were removed from TLC plates and the amount of radioactivity in each determined by liquid scintillation counting. Background levels of radioactivity were determined by counting blank areas of each chromatogram. Background radioactivity was subtracted from radioactivity associated with adducts and relative adduct levels were calculated as described by Reddy and Randerath [19]. The specific activity of the
[T-~~P]ATP was determined by measuring the incorporation of radioactivity into a known amount of dAp. Results and Discussion
The reaction of B[a]P, DMBA, 3-MC, B[a]A, or DB[a,c]A with calf thymus DNA in a buffer system resulted in hydrocarbon:DNA adducts that were easily detected with the 32Ppostlabeling assay. In contrast, phenanthrene was the only hydrocarbon evaluated that did not form DNA adducts when reacted with calf thymus DNA in vitro (maps not shown). The lack of adduct formation with phenanthrene may be related to its high ionization potential (8.19 eV) and in ability to form radical cations. Cavalieri and Rogan have demonstrated that polycyclic aromatic hydrocarbons with ionization potentials above 7.57 eV do not readily form radical cations in vitro [23]. This is in contrast to hydrocarbons with lower ionization potentials which are capable of forming radical cations. The detection of DNA adducts with B[a]P, DMBA, 3-MC and B[a]A is consistent with ionization potential below 7.57 and the ability of these compounds to form radical cations. The patterns of DNA adducts detected varied considerably between hydrocarbons (Fig. 1). The most complex adduct pattern was observed with DMBA. It is interesting to note that the adduct profiles observed for both B[a]P and DMBA in the present study, are similar to previously reported adduct profiles generated using enzymatic activation systems. In the case of B[a]P, adduct profiles are similar to those previously reported by Bodell and coworkers [20] who used a horseradish peroxidase/hydrogen peroxide (HRP/HzO,) in vitro activation system (Fig. 1, map 2 and 8). This similarity is not surprising since the mechanism of B[a]P activation by autooxidation or HRP/H202 catalysis is likely similar with the involvement of reactive oxygen species. Thus, one would expect to form similar types of B[a]P:DNA adducts. In the case of DMBA, however, the resemblance between adduct patterns observed in this study
PEI-cellulose TLC maps of hydrocarbon:DNA adducts formed in vitro in the absence of enzymatic or chemical activation. Maps performed under white light while maps 8 - 14 represent reactions performed in the dark. The origin is located at the bottom chromatogram and was excised prior to autoradiography. D3 solvent development was in the direction from left to right while was in the direction from bottom to top. Autoradiography was performed at -80°C for the time indicated above.
6: 24hr 12: 48 hr 5: 24 hr 11: 48 hr 4: 3 hr 10: 3 hr
3: 3hr 9: 48 hr
2: 3hr 8: 48 hr
1: 24 hr 7: 24 hr
Fis. 1. Representative 1 - 7 represent reactions left-hand corner of each D4 solvent development
DB[a,c]A
BMA 3-MC
DMBA
JWP
Control
39
and others [ZZ], does not necessarily suggest a similarity in activation mechanism (Fig. 1, map 3 and 9). DMBA:DNA adduct profiles reported by other investigators are considered products of microsomal activation (a twoelectron oxidation mechanism) while the adducts detected in the present study most likely from a one-electron oxidation arise mechanism. Any resemblance of adduct profiles may be due to the low resolution capacity of PEI-cellulose TLC for hydrocarbon:DNA adducts derived from PAH with similar molecular weights. A dose related increase in hydrocarbon:DNA adduct formation was observed when hydrocarbon levels were varied between 0.12 nmol and 600 nmol per reaction (Table 1). The level of adducts detected when reactions were performed in the dark ranged from 0.07 to 1.70 adducts per lo7 nucleotides. Hydrocarbon: DNA adduct formation was enhanced when reactions were performed under white light. Adduct levels in these reactions ranged from 0.09 to 15.28 adducts per lo7 nucleotides. The most notable enhancement, was observed with 120 nmol B[a]P and 12 nmol DMBA. The enhancement of adduct formation by white light was anticipated since PAH are known to undergo extensive photooxidation [23,24]. The lower levels of adducts observed in reactions with 600 nmol hydrocarbon may be due to hydrocarbon precipitation or the increased volume of DMSO in reactions. DMSO is known to be a scavenger of hydroxy radicals which may be involved in the auto-oxidation and binding of hydrocarbons to DNA in our reactions. The relative contribution of adduct formation from auto-oxidation in an in vitro enzymatic activation system was investigated for B[a]P and DMBA. Reactions were performed in the dark with 120 nmol of B[a]P or DMBA using a HRP/H202 activation system. B[a]P formed 139 adducts per lo7 nucleotides while DMBA formed 148 adducts per lo7 nucleotides, respectively. Comparing these results with the levels of adducts formed in the absence of enzymatic activation suggests that
Table 1. Dose (nmol)
Levels of PAH:DNA adducts formed in vitro. Adducts/ 10’ Dark
Nucleotides Light
Benzo[a]pyrene 0.56 5.45 600 0.07 2.21 120 0.19 1.37 12 ND 0.17 0.12 7,12-Dimethylbenz[a]anthracene 1.70 15.28 600 0.39 10.50 120 0.21 6.84 12 0.17 0.18 0.12 3-Methylcholanthrene 1.32 1.21 600 0.38 0.54 120 0.12 0.58 12 0.11 0.12 0.12 Benz[a]anthracene 0.45 0.95 600 0.12 0.80 120 0.43 0.09 12 ND 0.15 0.12 Dibenz[a,c]anthracene 1.14 1.76 600 0.32 0.57 120 0.13 0.06 12 0.08 ND 0.12
Adduct ratio Light/dark
9.7
31.6 7.2 9.0 26.9 33.6 1.1 0.9 1.4 4.8 1.1 2.1 6.7 0.2 1.5 1.8 0.5 -
ND = no adduct formation detected. Note: No phenanthrene:DNA adducts were detected when 120 Fmol-600 nmol of phenanthrene was reacted with calf thymus DNA in vitro.
between 0.05 and 0.26 percent of the adducts formed in the enzymatic system may arise from auto-oxidation of B[a]P and DMBA, respectively. If reactions were performed under white light the contribution of auto-oxidation to adduct formation could account for 1.6 and 7.1% of total adducts, respectively. The levels and profiles of hydrocarbon:DNA adducts formed following enzymatic activation in our studies were consistent with results reported by other investigators [21,25]. HRP/HsOs activation of B[a]P or DMBA has been reported to result in 161 and 136 adducts per lo7
40
nucleotides, respectively. The extent of DNA binding following microsomal activation has also been reported. Microsomal activation of B[a]P results in 118 adducts per lo7 nuclcotides while DMBA activation formed 1.2 adducts per lo7 nucleotides. The mechanism responsible for hydrocarbon activation and DNA adduct formation in reactions performed in the dark is not fully understood. A series of events such as PAH auto-oxidation and subsequent formation of reactive oxygen species is one possible explanation that could account for hydrocarbon oxidation and DNA binding. Recently, we have reported that Oz :, OH*, HzOz and t O2 are involved in B[u]P auto-oxidation and binding to DNA in vitro in the absence of enzymatic or chemical activation [12]. It is likely that these reactive oxygen species are also involved in the mechanism responsible for the hydrocarbon:DNA adducts observed in the present study with DMBA, 3-MC, B[a]A and DB[a,c]A. Recent studies have demonstrated that hydrocarbon oxidation by one-electron oxidation results in both stable and labile hydrocarbon:DNA adducts [21,25]. The labile hydrocarbon:DNA adducts would not be detected in our experiments due to the limitation of the 32P-postlabeling assay. Since oneelectron oxidation is inherent in an autooxidation mechanism, the extent of hydrocarbon:DNA adducts formed as a result of auto-oxidation may be greater, since labile DNA adducts would have went undetected in our studies. However, these results clearly demonstrated that auto-oxidation does contribute to stable hydrocarbon:DNA adduct formation. In summary, we have demonstrated that polycyclic aromatic hydrocarbons can undergo auto-oxidation and form DNA adducts that are readily detected with the 32P-postlabeling assay. Adduct formation is greatly enhanced in the presence of white light. Hydrocarbon autooxidation may account for up to 0.26% of total DNA adduct formation when hydrocarbon binding to DNA is evaluated in the dark using an in vitro enzymatic activation system.
The presence of white light may greatly enhance the contribution of auto-oxidation to hydrocarbon: DNA adduct formation. Hydrocarbon:DNA adduct patterns that result from PAH auto-oxidation may resemble adduct patterns observed from enzymatic activation systems. Thus, the potential for detecting PAH:DNA adducts as a result of autooxidation should be taken into account when investigating the activation and DNA binding of polycyclic aromatic hydrocarbons in vitro. Acknowledgments This work was supported by USPHS Grant R29 CA49826 from the National Cancer Institute. References Miller, J.A. and Miller, E.C. (1981) Searches for ultimate chemical carcinogens and their reactions with cellular macromolecules. Cancer, 47, 2327 - 2345. Wogan, G.N. and Gorelick, N.J. (1985) Chemical and biochemical dosimetry of exposure of genotoxic chemicals. Environ. Health Perspect., 62, 5- 18. Vanhakangas, K., Haugen, A. and Harris, C.C. (1985) An applied synchronous fluorescence spectrophotometric assay to study benzo[o]pyrene diolepoxide-DNA adducts. Carcinogenesis, 6, 1109 - 1116. Santella, R.M., Gasparo, F. and Hsieh, L. (1987) Quantitation of carcinogen-DNA adducts with monoclonal antibodies. Prog. Exp. Tumor Res., 31, 63 - 75. Gupta, R.C., Reddy, M.V. and Randerath, K. (1982) 32P-Postlabeling analysis of non-radioactive aromatic carcinogen-DNA adducts. Carcinogenesis. 3, 10811092. Phillips, D.H., Hewer, A., Martin, C.N., Gamer, R.C. and King, M.M. (1988) Correlation of DNA adduct levels in human lung with cigarette smoking. Nature, 336, 790 - 792. Weyand, E.H., Rice, J.E. and LaVoie, E.J. (1987) 32Ppostlabeling analysis of DNA adducts from non-alternate PAH using thin-layer and high performance liquid chromatography. Cancer Lett., 37, 257 - 266. Sophie, L., Ni She, M., Hewer, A., Grover, P.L., Platt, K.L., Oesch, F. and Phillips, D.H. (1991) The metabolic activation of dibenz[a,h]anthracene in mouse skin examined by 32P-postlabeiling: minor contribution of the 3, 4dial 1,2-oxides to DNA binding. Carcinogenesis, 12, 1079 - 1083. Canella, K., Peltonen, K. and Dipple, A. (1991) Identification of (+) and (-1 anti-benzo[a]pyrene dihydrodiol epoxide-nucleic acid adducts by the 32P-postlabeling assay. Carcinogenesis, 12, 1109- 1114.
41
10
11
12
13
14
15
16
17
18
Bodell, W.J., Devanesan, P.D., Rogan, E.G. and 32P-Postlabeling analysis of Cavalieri, E.L. (1989) benzo[a]pyrene-DNA adducts formed in vitro and in vivo. Chem. Res. Toxicol., 2, 312-315. Masento, M.S., Hewer, A., Grover, P.L. and Phillips, D.H. (1989) Enzyme-mediated phosphorylation of polycyclic hydrocarbon metabolites: detection of nonassay. Caradduct compounds in the 32P-postlabeling cinogenesis, 10, 1557 - 1559. Bryla, P. and Weyand, E.H. (1991) Role of activated oxygen species in benzo[a]pyrene:DNA adduct formation in vitro. Free Rad. Biol. Med., 11, 17 - 24. VanCanntfort, J., DeGraeve, J. and Gielen, J.E. (1977) Radioactive assay for aryl hydrocarbon hydroxylase: improved method and biological importance. Biochem. Biophys. Res. Commun., 79, 505-512. Minotti, G. and Aust, S.D. (1989) The role of iron in oxygen radical mediated lipid peroxidation. Chem.-Biol. fnteract., 71, l- 19. Minotti, G. and Aust, S.D. (1987) Superoxide-dependent redox cycling of citrate-Fe 3+. Evidence for a superoxide dismutase like activity. Arch. Bioch. Biophys., 253, 257 - 267. Rogan, E.G., Katomski, P.A., Roth, R.W. and Cavalieri, E.L. (1979) Horseradish peroxidase/hydrogen peroxidecatalyzed binding of aromatic hydrocarbons to DNA. J. Biol. Chem., 254, 7055- 7059. Cavalieri, E.L., Rogan, E.G., Roth, R.W., Sauger, R.K. and Hakam, A. (1983) The relationship between ionization potential and horseradish peroxidase/hydrogen peroxidecatalyzed binding of aromatic hydrocarbons to DNA. Chem.-Biol. Interact., 47, 87 - 109. Dunn, B.P. and Stich, H.F. (1986) 32P-Postlabeling analysis of aromatic DNA adducts in human oral mucosal cells. Carcinogenesis, 7, 115 - 1120.
19
20
21
22
23
24
25
Reddy, M.V. and Randerath, K. (1986) Nuclease Plmediated enhancement of sensitivity of 32P-postlabeling test for structurally diverse DNA adducts. Carcinogenesis, 7, 1543- 1551. Cavalieri, E. and Rogan, E. (1985) Role of radical cations in aromatic hydrocarbon carcinogenesis. Environ. Health Perspect., 64, 69-84. Bodell, W.J., Devanesan, P.D., Rogan. E.G. and 32P-Postlabeling analysis of Cavalieri, E.L. (1989) benzo[a]pyrene-DNA adducts formed in vitro and in vivo. Chem. Res. Toxicol., 2, 312-315. Schmeiser, H., Dipple, A., Schurdak, M.E., Randerath, E. and Randerath, K., (1988) Comparison of 32Ppostlabeling and high pressure liquid chromatographic analyses for 7, 12-dimethylbenz[a]anthracene-DNA adducts, Carcinogenesis, 9, 633 - 638. Shevchuk, 1. and Goubergrits, M. (1991) Reactivity of alternate polyarenes under photooxidation, In: Polycyclic Aromatic Hydrocarbons Measurements, Means and Metabolism, pp. 831- 838. Editors: M. Cooke, K. Loening and J. Menit, Battelle Press, Columbus, Ohio. Vu-Du, T. and Huynh, C.K. (1991) Photodecomposition rates of adsorbed PAH and mutagenic activity of irradiated synthetic mixtures, In: Polycyclic Aromatic Hydrocarbons Measurements, Means and Metabolism, pp. 979 - 994. Editors: M. Cooke, K. Loening and J. Merrit, Battelle Press, Columbus, Ohio. RamaKrishna, N.V.S., Devanesan, P.D., Rogan, E.G., Cavalieri, E.L., Jeong, X., Jankowiak, R. and Small, G.J. (1992) Mechanism of metabolic activation of the potent carcinogen 7,12-dimethylbenz[a]anthracene. Chem. Res. Toxicol., 5, 220 - 226.
