Expression of cytochromes P450 and glutathione S-transferases in human prostate, and the potential for activation of heterocyclic amine carcinogens via acetyl-coA-, PAPS- and ATP-dependent pathways
Article first published online: 4 MAY 2005
Copyright © 2005 Wiley-Liss, Inc.
International Journal of Cancer
Volume 117, Issue 1, pages 8–13, 20 October 2005
How to Cite
Di Paolo, O. A., Teitel, C. H., Nowell, S., Coles, B. F. and Kadlubar, F. F. (2005), Expression of cytochromes P450 and glutathione S-transferases in human prostate, and the potential for activation of heterocyclic amine carcinogens via acetyl-coA-, PAPS- and ATP-dependent pathways. Int. J. Cancer, 117: 8–13. doi: 10.1002/ijc.21152
- Issue published online: 2 AUG 2005
- Article first published online: 4 MAY 2005
- Manuscript Accepted: 4 FEB 2005
- Manuscript Received: 22 OCT 2004
- U.S. National Cancer Institute
- National Center for Toxicological Research
- cytochrome P450;
- glutathione S-transferase;
- heterocyclic amines;
Dietary factors appear to be involved in the high incidence of prostate cancer in “Westernized” countries, implicating dietary carcinogens such as heterocyclic amines (HAs) in the initiation of prostate carcinogenesis. We examined 24 human prostate samples with respect to their potential for activation and detoxification of HAs and the presence of DNA adducts formed in vivo. Cytochromes P450 1B1, 3A4 and 3A5 were expressed at low levels (<0.1–6.2 pmol/mg microsomal protein). N-Acetyltransferase (NAT) activities, using p-aminobenzoic acid (NAT1) and sulfamethazine (NAT2) as substrates, were <5–5,500 and <5–43 pmol/min/mg cytosolic protein, respectively. Glutathione S-transferases (GSTs) P1, M2 and M3 were expressed at 0.038–1.284, 0.005–0.126 and 0.010–0.270 μg/mg cytosolic protein, respectively; GSTM1 was expressed in all GSTM1-positive samples (0.012–0.291 μg/mg cytosolic protein); and GSTA1 was expressed at low levels (<0.01–0.11 μg/mg cytosolic protein). Binding of N-hydroxy-PhIP to DNA in vitro occurred primarily by an AcCoA-dependent process (<1–54 pmol/mg/DNA), PAPS- and ATP-dependent binding being <1–7 pmol/mg DNA. In vivo, putative PhIP- or 4-aminobiphenyl-DNA adducts were found in 4 samples (0.4–0.8 adducts/108 bases); putative hydrophobic adducts were found in 6 samples (8–64 adducts/108 bases). Thus, the prostate appears to have low potential for N-hydroxylation of HAs but greater potential for activation of N-hydroxy HAs to genotoxic N-acetoxy esters. The prostate has potential for GSTP1-dependent detoxification of ATP-activated N-hydroxy-PhIP but little potential for detoxification of N-acetoxy-PhIP by GSTA1. However, there were no significant correlations between expression/activities and DNA adducts formed in vitro or in vivo, DNA adducts in vivo possibly reflecting carcinogen exposure. © 2005 Wiley-Liss, Inc.
Prostate cancer is a major health problem in the United States, being the most frequently diagnosed male cancer and the second most common cause of cancer death among men.1 The etiology of human prostate cancer is not well understood; however, there is evidence that dietary factors are involved in the high incidence of prostate cancer in the United States and Western Europe, implicating environmental or food-derived carcinogens in the initiation of prostate carcinogenesis.2
Heterocyclic amines (HAs), aromatic amines (AAs) and polycyclic aromatic hydrocarbons (PAHs) are formed during the cooking of meat at high temperatures,3, 4 or are widely distributed in the environment as fuel products and components of tobacco smoke.5 HAs and AAs have been shown to induce cancer in various organs of the rat and mouse, such as in the mammary gland, prostate, lung, colon, skin, bladder and liver.2
These carcinogens require metabolic activation to exert their genotoxic effects, and the relative rates of activation and/or detoxification can be critical determinants in the induction, promotion and progression of tumorigenesis.6 The major pathway of HA and AA activation involves phase I hepatic cytochrome P450 (CYP)-mediated N-hydroxylation (primarily CYP1A2) followed by phase II conjugation of the N-hydroxylamines to ester derivatives that react with DNA.7N-Acetyltransferases (NATs) and sulfotransferase (SULT) are capable of activating N-hydroxy-HAs in this manner, and an ATP-dependent pathway of activation has recently been identified in prostate for the most abundantly formed HA, 2-amino-1-methyl-6-phenylimidazo(4,5-b)pyridine (PhIP).8 Of importance as far as detoxification is concerned are glutathione S-transferases (GSTs),9, 10 and UDP-glucuronosyltransferases (UGTs),11 although their action appears to be limited primarily to the detoxification of N-acetoxy- and N-hydroxy-PhIP, respectively. In addition, an enzymatic, NADH-dependent reductive process has been identified for detoxification of N-hydroxy-HAs.12
PAH carcinogens are activated by epoxidation (catalyzed primarily by CYPs 1A1, 1B1 and 3A4), epoxide hydration, and a secondary oxidation of the resultant dihydrodiols to reactive diolepoxides.5 Several of these proximate and ultimate PAH carcinogens are known to be detoxified by GSTs, namely, GSTs M1, A1 and P1.13
The purpose of our work was to investigate the metabolic capacity of human prostate tissue for activation and detoxification of the most abundantly formed HA carcinogen, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), particularly in relationship to the pathways outlined above, and to determine by means of 32P-postlabeling-HPLC the DNA adduct levels derived from PhIP, 4-aminobiphenyl (ABP) and benzo[a]pyrene (BAP).
Material and methods
Chemicals and reagents
2-Hydroxyamino-1-methyl-6-phenylimidazo(4,5-b)pyridine(N-hydroxy-PhIP) was purchased from Toronto Research Chemicals (Toronto, Ontario, Canada). [3H]N-Hydroxy-PhIP was synthesized from [3H]PhIP (53 Ci/mmol, ChemSyn, Lenexa, KS) as described by Lin et al.;14 this was diluted with unlabelled material to 40 mCi/mmol before use. Adenosine 5′-[γ-32P]triphosphate with an original specific activity of ∼5,000 Ci/mmol was obtained from ICN Biomedicals (Irvine, CA). T4 polynucleotide kinase (PNK) was from USB (Cleveland, OH). ABP, PhIP and benzo[a]pyrene (BAP) adducts of dG-3′, 5′-bisphosphate (pdGp-ABP, pdGp-PhIP and pdGp-BAP, respectively) were synthesized according to Gorlewska-Roberts et al.15
Polyclonal antibodies for cytochromes P450 (CYPs) 1A1, 1A2, 2A6, 2E1, 3A4 and 3A5 were purchased from Gentest Corporation (Woburn, MA). CYP1B1 antibody was purchased from Alpha Diagnostics (San Antonio, TX). Recombinant reference CYPs (Gentest) were purchased from BD Sciences (Franklin Lakes, NJ).
All other materials were obtained from the Sigma Chemical Co. (St. Louis, MO) or as stated in the text.
Snap-frozen prostate tissue (0.2–0.5 g) was obtained from the U.S. Cooperative Human Tissue Network (Southern Division, Montgomery AL). All tissues were benign prostatic hyperplastic (BPH) from patients who had undergone transurethral prostatectomy because of prostatic disease. Brief details of patient race, age and diagnosis (where available) are given in Table I. Tissues were stored at −80°C until use.
|Sample||Age||Race||Tissue type, histology, and patient diagnosis1|
|1||70||Caucasian||BPH (no details)|
|2||47||Caucasian||BPH (no details)|
|3||63||Caucasian||Focal, atypical adenomatous hyperplasia, chronic inflammation; no evidence of malignancy|
|4||64||Caucasian||BPH from patient with adenocarcinoma (Gleason 6)|
|5||57||Caucasian||Adenomatous and stromal hyperplasia; chronic inflammation; no tumor; patient with invasive urothelial carcinoma|
|6||65||Caucasian||Hyperplasia and chronic inflammation|
|7||67||Caucasian||BPH (no details)|
|8||56||Caucasian||BPH from uninvolved margins of tissue; patient with high-grade PIN, multifocal adenocarcinoma (Gleason 6)|
|9||69||Caucasian||BPH (no details)|
|10||84||Caucasian||Fibromuscular and glandular hyperplasia; no malignancy|
|11||57||Caucasian||BPH from uninvolved margins of tissue; patient with adenocarcinoma (Gleason 6)|
|12||59||Caucasian||BPH from uninvolved margins of tissue; patient with well-differentiated and highly localized adenocarcinoma (Gleason 7)|
|13||78||Caucasian||Benign stromal and glandular hyperplasia, no sign of malignancy|
|14||64||Caucasian||BPH (no details)|
|15||80||Caucasian||BPH (no details)|
|16||60||Caucasian||BPH (no details)|
|17||84||Caucasian||BPH tissue; patient with poorly differentiated transitional cell carcinoma|
|18||62||Caucasian||BPH tissue; patient with low grade PIN and chronic prostatitis|
|19||69||Caucasian||BPH (no details)|
|20||75||African-American||BPH (no details)|
|21||73||Not known||BPH (no details)|
|22||51||Not known||BPH (no details)|
|23||—||Not known||BPH (no details)|
|24||—||Not known||BPH (no details)|
Preparation of tissue fractions
Thawed tissues were washed in 50 mM sodium pyrophosphate, pH 7.0, and 1 mM dithiothreitol, minced and homogenized in 0.01 M Tris-HCl, pH 7.8, 0.25 M sucrose, 10 mM EDTA and 0.1 mM phenylmethylsulfonyl fluoride using a Brinkman Polytron homogenizer. The homogenate was centrifuged at 12,500g for 20 min to obtain the crude nuclear pellet for DNA isolation. The supernatant was centrifuged at 105,000g for 1 hr to obtain cytosolic and microsomal fractions. Microsomes were washed in the Tris-EDTA buffer and resuspended in 10 mM Tris acetate, pH 7.4, 0.25 M sucrose and 20% glycerol. All procedures were performed at 4°C. Protein concentrations of microsomal and cytosolic fractions were measured by the biuret reaction. All fractions were stored at −80°C until use.
Cytosolic N-acetyltransferase (NAT) activities
Specific activities towards NAT1 and NAT2 were determined by the acetylation of the NAT1- and NAT2-specific substrates p-aminobenzoic acid (PABA) and sulfamethazine (SMZ), respectively.16 Briefly, reactions were carried out using 500 μM AcCoA, 100 μM PABA or 500 μM SMZ at 37°C for 15 min; protein concentration was 1 mg/ml. The reactions were terminated by addition of cold trichloroacetic acid (final concentration 4 %). AcCoA was omitted from the blanks. The N-acetyl PABA and N-acetyl-SMZ products were quantitated by HPLC as described.17
Microsomal glucuronidation activity
Glucuronidation activity toward N-hydroxy-PhIP was determined as described by Nowell et al.11 All solutions were argon-saturated prior to use and incubations were also carried out under argon. Incubations were performed at 37°C for 30 min in 100 mM Tris-HCl, pH 7.5, 5 mM MgCl2 and 8.5 mM saccharolactone with 50 μg microsomal protein, 5 mM UDP-glucuronic acid and 100 μM [3H]N-hydroxy-PhIP in a final volume of 50 μl. Reactions were terminated by the addition of 20 μl of ethanol. Aliquots of 50 μl of the samples were spotted onto preabsorbent layer of channeled silica gel TLC plates [Baker 250Si-PA (19C); VWR Scientific, Suwanee, GA] and developed in chloroform/methanol/glacial acetic acid/water (65:25:2:4 v/v). After development, the plates were dried and subjected to autoradiography for 3 days at −80°C. Silica gel containing N-hydroxy-PhIP glucuronides and from corresponding areas in control lanes was scraped into scintillation vials, and radioactivity was determined by liquid scintillation counting.
AcCoA-dependent binding of N-hydroxy-PhIP to calf thymus DNA was performed in a manner similar to that of Flammang et al.18 Briefly, assays were performed at 37°C under argon, in 50 mM sodium pyrophosphate buffer, pH 7.0, containing 1 mM dithiothreitol, 2 mg/ml calf thymus DNA, 1.0 mg/ml cytosolic protein, 1 mM AcCoA, and 5 mM [3H] N-hydroxy-PhIP. Reactions were terminated after 15 min by the addition of 2 vol of water-saturated n-butanol and vortex mixing. The DNA was isolated by multiple solvent extractions and precipitations and the extent of covalent binding determined by liquid scintillation counting. DNA was quantitated by the diphenylamine reaction.
PAPS-dependent DNA binding catalyzed by sulfotransferases was measured as described in Chou et al.19 The assays were performed at 37°C under argon in 50 mM potassium phosphate pH 7.0, 10 mM MgCl2, 0.5 mM EDTA, 10 mM 3′-phosphoadenosine-5′-phosphosulfate (PAPS), containing 2 mg/ml calf thymus DNA, 1 mg/ml cytosolic protein and 5 mM [3H] N-hydroxy-PhIP. PAPS was omitted from the blanks. The DNA was isolated by multiple solvent extractions and precipitations and the extent of PAPS-dependent covalent binding was determined by liquid scintillation counting.
ATP kinase-catalyzed DNA binding was measured similarly to PAPS-dependent binding, using 1 mM ATP instead of PAPS. ATP was omitted from the blanks.
Quantitation of cytochromes P450
Cytochromes P450 (CYPs) 1A1, 1A2, 1B1, 2A6, 2E1, 3A4 and 3A5 were quantitated by immunoblotting as previously described.20 Microsomal protein (100 μg) was separated by SDS-polyacrylamide gel electrophoresis according to the method of Laemmli,21 using a 4–10% gradient gel, and including an appropriate reference recombinant protein and molecular weight markers. Proteins were electrophoretically transferred onto nitrocellulose membranes as described by Towbin et al.22 and probed with an appropriate primary antibody and horseradish peroxidase conjugated secondary antibody. Expression was quantitated by using ChemiGlow West (Alpha Innotech, San Leandro, CA), exposure to photographic film, development and determining the density of bands on the film.
Quantitation of GSTs, and GSTM1 genotype
GSTs were isolated and quantitated by the method of Coles et al.23 A GST pool was prepared from a known aliquot of each cytosol, corresponding to 0.1–5.0 mg of cytosolic protein, using a column of S-linked glutathione-agarose. Bound proteins were eluted from the column in 100 mM Tris, 40 mM sodium phosphate and 50 mM glutathione, pH 9.6. A portion of the GST eluate (0.250 ml) was subjected to HPLC analysis (C-18, 300 μM pore “Jupiter” column; Phenomenex, Torrance, CA) and GST subunits quantitated by reference to authentic standards. GSTM1 genotype was determined from genomic DNA (see below) by the method of Arand et al.24
Isolation and hydrolysis of DNA, adduct enrichment and 32P-postlabeling
DNA was isolated from the crude nuclear pellets by digestion with proteinase K, RNAases A and T1, according to the method of Gupta et al.25 DNA concentrations were determined spectrophotometrically using A260. DNA hydrolysis, adduct enrichment and 32P-postlabeling was performed according to Gorlewska-Roberts et al.15 Briefly, 30 μg of DNA was hydrolyzed in 10 mM sodium succinate pH 6.0 by adding a solution (1 μl/μg DNA) of 0.2 U/μl micrococcal nuclease (MN) and 3.6 mU/μl spleen phosphodiesterase (SPD) and incubation at 37°C overnight. The MN/SPD had been dialyzed in water before use to remove ammonium salts. The enrichment of hydrophobic adducted nucleotides was performed using a 1 ml HLB Oasis Sep-Pak (Waters, Milford, MA) in digestion buffer. Normal nucleotides were removed by washing with water, and adducted nucleotides were eluted in 3 column volumes of methanol. Samples were evaporated to dryness in a SpeedVac Concentrator. [5′-32P]-phosphorylation was performed using polynucleotide kinase (2 U) and [γ-32P]ATP (100 μCi and ∼2 μM) in 50 mM Tris-HCl buffer, pH 7.6,, containing 10 mM MgCl2 and 10 mM 2-mercaptoethanol in a total volume of 20 μl. Mixtures were incubated at 37°C for 45 min. HPLC analyses of the 32P-postlabeled DNA adducts were performed by injecting the total 32P-postlabeled mixture after addition of the 3 (nonradioactive) standards (pdGp-ABP, pdGp-PhIP and pdGp-BAP) as UV markers, using sequential UV detection and on-line liquid scintillation counting.
Data were analyzed for correlations between expression and DNA-binding by linear regression analysis using the program SigmaStat 3.0 (Jandel Scientific, San Rafael, CA).
PABA acetylation activity (NAT1) was detected for 21/24 samples (Table II). Significant acetylation activity was observed for SMZ (NAT2) in 19/24 samples (Table II). NAT1 and NAT2 activities were not correlated (r = 0.06, p = 0.8)
|Sample||NAT activity pmol/mg protein2||GST subunit expression μg protein/mg protein||CYP expression pmol/mg protein3||Cofactor-dependent binding of N-OH-PhIP in vitro pmol/mg DNA2||DNA adducts in vivo (adducts/108 bases)|
UGT activity for N-hydroxy-PhIP was not detected in any sample (i.e., < 20 pmol/min).
Expression of GSTs
GSTs A1, M1, M2, M3 and P1 were detected in the prostate cytosols (Fig. 1). Although expression was variable between tissue samples, GSTP1 (the major GST in all samples), M2 and M3 were detected in all samples and GST M1 was expressed in all GSTM1 positive samples (Table II). GSTA1 was present at low levels in 18/24 samples (Table II). Expression of GSTs P1, M2 and M3 were significantly correlated (Fig 2).
Metabolic activation of N-hydroxy-PhIP to DNA-binding species
AcCoA-dependent binding of N-hydroxy-PhIP to calf thymus DNA was observed in 23/24 samples (Table II). In contrast, ATP- and PAPS-dependent DNA-binding of [3H]N-Hydroxy-PhIP was low (<1-7 pmol/mg DNA/ mg cytosolic protein; Table II) with significant binding observed for 16/24 and 11/24 samples, respectively. There was no significant correlation between NAT1 activity, NAT2 activity and AcCoA-DNA-dependent binding (r <0.1, p> 0.5).
DNA adducts in prostate tissue
Examples of HPLC profiles of the 32P-postlabeling are shown in Figure 3 and adduct quantitation is given in Table II. Ten DNAs showed quantifiable adducts, of which 4 showed discrete peaks that coeluted with the synthetic pdGp-ABP and/or pdG-PhIP adduct standards (Fig. 3a) and 6 showed a broad peak of presumed hydrophobic adducts (Fig. 3b). Putative pdGp-ABP was present in 3 samples, and putative pdG-PhIP was present in 2, with 1 sample containing both adducts (Fig. 3a, Table II). pdGp-BAP was not detected in any sample. There were no correlations between NAT activity, GST expression and the presence of DNA adducts in vivo.
Several studies have shown that the human prostate is capable of activating HA carcinogens to mutagenic species or to DNA-binding species in vitro. For example, prostate tissue incubated in the presence of PhIP, IQ or BAP resulted in the formation of low levels of adducts (∼0–7 adducts/108 nucleotides) that was variable between samples.26 Similarly, prostate cells in culture incubated with PhIP or BAP formed DNA single-strand breaks,27 and prostate epithelial cells in culture metabolized the HAs, PhIP, MeIQx and Glu-P-2 to mutagenic species.28
Low expression of several CYPs in the prostate has been demonstrated, of which 1A1, 1A2, 1B1, 3A5 and 4B1 appear to be most frequently expressed on the basis of mRNA.26, 29 In our study, we showed that CYPs 1B1 and 3A5 are present as protein, and demonstrated the expression of CYP3A4 protein. However we did not detect CYP1A2 protein, despite a report of low CYP1A2 activity in 2 prostate epithelial cell cultures.28 The presence of low levels of CYPs 3A4 and 3A5 could result in the low levels of activation of PAH carcinogens to DNA-binding species, although Williams et al.26 provided evidence that activation of BAP in prostate tissue in vitro occurred primarily via CYPs 1A1 or 1B1.
In contrast, the prostate appears to be more efficient in the activation of N-hydroxy-PhIP to DNA-binding or genotoxic species, i.e., by prostate tissue in vitro,26 by prostate cells in culture,27 and by LNCaP cells.8 We found that activation of N-hydroxy-PhIP to DNA-binding species in vitro occurred primarily by an AcCoA-dependent process and that PAPS- or ATP-dependent DNA-binding was relatively low. The range of NAT1 activity reported here for whole tissue cytosol (∼0–5.5 nmol/mg cytosolic protein) includes the values obtained for 2 prostate epithelial cell lines (2.9–3.6 nmol/mg cytosolic protein28). NAT2 activity was not detected by Lawson and Kolar.28 In contrast, LNCaP cells activated N-hydroxy-PhIP by an ATP-dependent process and NAT/AcCoA-dependent binding appeared to be negligible.8
Previous studies have shown that N-acetoxy-HA esters are detoxified by GSH.9 GSTA1 catalyzes this reaction for N-acetoxy-PhIP,10 and ATP-activated N-hydroxy-PhIP is detoxified by GSTP1.9 GSTs of the alpha, mu and pi classes have been shown to be expressed in prostate tissue,30, 31, 32 although they had not been accurately identified or quantitated. We confirmed the presence of GSTs of these classes. GSTA1 was expressed at very low levels and there appears to be little potential for the prostate to detoxify N-acetoxy-PhiP by this route. GSTP1, which is commonly silenced in prostate tumors (and LNCaP cells) by hypermethylation33 was consistently the major GST in our prostate samples. GSTs M2 and M3 were also consistently (but variably) expressed, and there was a correlation between expression of GSTs P1, M2 and M3, indicating a conserved pattern of GST phenotype, similar (in degree of correlation) to that of other tissues.6 GSTM1 was expressed in all GSTM1-positive samples. Thus, GSTs of the mu class form a significant proportion (∼5–50%) of GSTs (of the alpha mu and pi classes) of the prostate.
There were no correlations between CYP expression, NAT activity, GST expression and PhIP-DNA binding in vitro, whether AcCoA-, PAPS- or ATP-dependent. The reasons for this lack of correlation and the differences between the apparent importance of AcCoA- and ATP-dependent activation between our experiments and those of Nelson et al.8 are not clear. However, it should be borne in mind that the prostate cytosols would be expected to contain nucleophilic macromolecules (or other nucleophilic components) that could compete with DNA for N-acetoxy-PhIP, and which were not standardized in our assays. Similarly GSH is known to reduce N-acetoxy-PhIP nonenzymatically to the amine.9 These competing reactions could cause the observed lack of correlation. There are also obvious differences between the procedures used here and those of Nelson et al. (i.e., incubations of prostate tissue cytosols vs. the use of LNCaP cells) that could cause the difference of result between workers. Although PhIP has been extensively used as a model for HA carcinogen metabolism, PhIP-DNA adducts were not a consistent feature of the human prostate samples in vivo. Putative pdGp-PhIP was detected in only 2 samples, others presenting no (detectable) adducts, an ill-defined spectrum of hydrophobic adducts or putative pdGp-ABP adduct. These inconsistent patterns of adduct formation may be determined by carcinogen exposure rather than being a reflection of the metabolic profile of the prostate; however, details of, e.g., smoking history and diet (that would reflect carcinogen exposure) were not available for the tissue donors.
Although prostate has little ability to N-hydroxylate HAs, N-hydroxy-HAs could be formed in the liver and transported to the prostate by the circulation where they could be activated to the carcinogenic N-acetoxy esters (i.e., similar to the mechanism proposed for exposure of the colon to N-hydroxy-HAs6). It is of interest to note that the NAT1*10 genotype (high activity) has been reported to be a risk factor for prostate cancer, particularly in the presence of “slow metabolizer” NAT2 genotypes.34 Whether these observations reflect a significant role for NAT1 in prostate carcinogenesis is not clear. However, high SULT1A1 activity (determined from blood cell platelets or deduced from genotype) has also been found to be a risk factor for prostate cancer35 even though we found PAPS-dependent activation of N-hydroxy-PhIP by prostate cytosols to be relatively low. Thus, the pathways of activation, and the importance, of dietary carcinogens in prostate carcinogenesis needs further study.
The prostate samples used were provided by the U.S. Cooperative Human Tissue Network, which is funded by the U.S. National Cancer Institute; other workers may have received samples of these same tissues. This research was supported, in part, by an appointment of O. Di Paolo to the Postgraduate Research Participation Program at the National Center for Toxicological Research administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the U. S. Department of Energy and the U. S. Food and Drug Administration. The authors thank Dr. K. Gorlewska-Roberts (National Center for Toxicological Research, U. S. Food and Drug Administration, Jefferson, AR) for assistance in 32P-postlabeling.
- 7Metabolic activation and DNA adducts of aromatic amines and nitroaromatic hydrocarbons, In: CooperCS, GroverPL, eds. Handbook of experimental pharmacology carcinogenesis and mutagenesis. Heidelberg: Springer-Verlag, 1990, 267–325., .