Metabolism of benzo [a] pyrene and benzo [a] pyrene-7, 8-diol by ...

Carcinogenesis vol.19 no.10 pp.1847–1853, 1998

Metabolism of benzo[a]pyrene and benzo[a]pyrene-7,8-diol by human cytochrome P450 1B1

James H.Kim, Kevin H.Stansbury, Nigel J.Walker, Michael A.Trush, Paul T.Strickland and Thomas R.Sutter1 Department of Environmental Health Sciences, The Johns Hopkins School of Hygiene and Public Health, 615 North Wolfe Street, Baltimore, MD 21205, USA 1To

whom correspondence should be addressed

Benzo[a]pyrene (B[a]P), a ubiquitous environmental, tobacco and dietary carcinogen, has been implicated in human cancer etiology. The role of human cytochrome P450 1B1 in the metabolism of B[a]P is poorly understood. Using microsomal preparations of human P450 1A1, 1A2 and 1B1 expressed in baculovirus-infected insect cells, as well as human and rat P450 1B1 expressed in yeast, we have determined the metabolism of B[a]P, with and without the addition of exogenous epoxide hydrolase, and B[a]P7,8-dihydrodiol (7,8-diol), each substrate at a concentration of 10 µM. HPLC analysis detected eight major metabolites of B[a]P and four metabolites of the 7,8-diol. The results of these studies indicate that cytochrome P450 1B1 carries out metabolism of B[a]P along the pathway to the postulated ultimate carcinogen, the diol epoxide 2, at rates much higher than P450 1A2 but less than P450 1A1. The rates of formation of the 7,8-diol metabolite in incubations with epoxide hydrolase are 0.17 and 0.38 nmol/min/nmol P450 for human P450 1B1 and 1A1, respectively, and undetectable for 1A2. The rates of total tetrol metabolite formation from the 7,8-diol, which are indicative of diol epoxide formation, are 0.60, 0.43 and 2.58 nmol/min/nmol P450 for 1B1, 1A2 and 1A1 respectively. In agreement with other reports of rat P450 1B1 activity, our data show this rat enzyme to be very active for B[a]P and 7,8-diol, with rates higher than human P450 1B1. In addition to the established role of P450 1A1 in B[a]P metabolism, P450 1B1 may significantly contribute to B[a]P and 7,8-diol metabolism and carcinogenesis in rodent tumor models and in humans. Introduction Benzo[a]pyrene (B[a]P) is a ubiquitous environmental pollutant produced during combustion processes. The majority of Abbreviations: B[a]P, benzo[a]pyrene; DE 1, (6)-benzo[a]pyrene-r-7,t-8dihydrodiol-c-9,10-epoxide; DE 2, (6)-benzo[a]pyrene-r-7,t-8-dihydrodiol-t9,10-epoxide; 4,5-diol, (6)-benzo[a]pyrene-trans-4,5-dihydrodiol; 7,8-diol, (–)-benzo[a]pyrene-7R-trans-7,8-dihydrodiol; 9,10-diol, benzo[a]pyrenetrans-9,10-dihydrodiol; DMBA, dimethylbenz[a]anthracene; DMSO, dimethyl sulfoxide; G6P, glucose 6-phosphate; 3-OH, 3-hydroxybenzo[a]pyrene; 7-OH, 7-hydroxybenzo[a]pyrene; 9-OH, 9-hydroxybenzo[a]pyrene; mEH, microsomal epoxide hydrolase; PAH, polycyclic aromatic hydrocarbon; P450, cytochrome P450; 1,6-quinone, benzo[a]pyrene-1,6-dione; 3,6-quinone, benzo[a]pyrene-3,6-dione; 6,12-quinone, benzo[a]pyrene-6,12-dione; RTCC, (6)-benzo[a]pyrene-r-7,t-8,c-9,c-10-tetrahydrotetrol; RTCT, (6)-benzo[a]pyrene-r-7,t-8,c-9,t-10-tetrahydrotetrol; RTTC, (6)-benzo[a]pyrene-r-7,t-8,t9,c-10-tetrahydrotetrol; RTTT, (6)-benzo[a]pyrene-r-7,t-8,t-9,t-10-tetrahydrotetrol. © Oxford University Press

human exposure occurs by ingestion, inhalation and dermal absorption, for example, during consumption of char-broiled foods and smoking (1). B[a]P must be bioactivated by enzymes such as cytochrome P450 (P450) in order to acquire its mutagenic and carcinogenic properties (2–7). The first step in the metabolic activation of B[a]P is formation of the B[a]P7,8-epoxide, followed by hydrolysis by microsomal epoxide hydrolase (mEH) to the (–)-benzo[a]pyrene-7R-trans-7,8dihydrodiol (7,8-diol) metabolite. This metabolite is further metabolized by P450 to the mutagenic (6)-benzo[a]pyrene-r7,t-8-dihydrodiol-t-9,10-epoxide species, also known as diolepoxide 2 (DE 2). DE 2 is extremely reactive and can bind to macromolecules such as DNA, RNA and protein. Also, DE 2 can undergo spontaneous hydrolysis to two products, (6)benzo[a]pyrene-r-7,t-8,t-9,c-10-tetrahydrotetrol (RTTC) and (6)-benzo[a]pyrene-r-7,t-8,t-9,t-10-tetrahydrotetrol (RTTT). These two tetrol metabolites, which are more stable in aqueous solution, are indicative of DE 2 formation. In addition, (6)benzo[a]pyrene-r-7,t-8,c-9,t-10-tetrahydrotetrol (RTCT) and (6)-benzo[a]pyrene-r-7,t-8,c-9,c-10-tetrahydrotetrol (RTCC) metabolites are indicative of (6) benzo[a]pyrene-r-7,t-8dihydrodiol-c-9,10-epoxide, also known as diol-epoxide 1 (DE 1), formation. Common phenolic metabolites are 3-hydroxy-, 7-hydroxy- and 9-hydroxybenzo[a]pyrene (3-OH, 7-OH and 9-OH). These metabolites arise in part by rearrangement of epoxides such as the B[a]P-2,3-epoxide, B[a]P-7,8-epoxide and B[a]P-9,10-epoxide. The addition of mEH forms dihydrodiol metabolites, such as the 7,8-diol and benzo[a]pyrene-trans9,10-dihydrodiol (9,10-diol), from the initial epoxides (5). In addition, quinone metabolites are also found, namely B[a]P1,6-dione, B[a]P-3,6-dione and B[a]P-6,12-dione (1,6-quinone, 3,6-quinone and 6,12-quinone). Formation of phenol and quinone metabolites is thought to be along pathways towards conjugation and detoxification (2). However, there is evidence of B[a]P quinone metabolites forming DNA adducts when they undergo 1e– reduction by cytochrome P450 reductase to form unstable semiquinone species (8,9). The P450 enzymes of the 1A subfamily are known to be inducible by polycyclic aromatic hydrocarbons (PAH), such as B[a]P, and polyhalogenated hydrocarbons, such as 2,3,7,8tetrachlorodibenzo-p-dioxin, and have also been shown to metabolize B[a]P and related PAHs (5,10,11). Recombinant human P450 1A1 and 1A2 have been studied extensively in the metabolism of compounds such as B[a]P, 7,8-diol, 7,12dimethylbenz[a]anthracene (DMBA) and 2-acetylaminofluorene (12–16). Similarly, P450 1B1 is inducible by 2,3,7,8tetrachlorodibenzo-p-dioxin in human keratinocyte cultures (17). P450 1A1 and 1B1 are both expressed in many human tissues (17–19) and are presumably inducible through activation of the aryl hydrocarbon receptor in these tissues. Human P450 1B1 gene characterization has shown that, unlike the genes for P450 1A1 and 1A2, it lacks a consensus TATA box and CCAAT motif, has three exons instead of seven and is located on chromosome 2 instead of chromosome 15 (20,21). 1847

J.H.Kim et al.

Current research has focused on the role of human P450 1B1 in procarcinogen activation and endogenous steroid metabolism (19,22,23). It has been demonstrated that P450 1B1 is a catalytically efficient 17β-estradiol 4-hydroxylase, and a 2hydroxylase (23). In addition, human P450 1B1 has been shown to metabolize environmental and dietary carcinogens such as benzo[c]phenanthrene and 2-amino-1-methyl-6phenylimidazo[4,5-b]pyridine (24,25). Human P450 1B1 has been shown to bioactivate a battery of procarcinogens, such as PAHs and arylamines, indicating that the contribution of P450 1B1 to extra-hepatic xenobiotic metabolism may be significant (19). However, rates of B[a]P metabolite formation and metabolite profiles have not been reported previously for human P450 1B1. Previous studies of P450 1B1 in rat and mouse have shown that this enzyme is able to metabolize PAHs such as B[a]P and DMBA to diol and phenol metabolites (26–28). These studies were conducted with microsomal preparations of rat and mouse tissues. Expression of P450 1B1 determined by northern blot and immunoblot analysis has been found in many rat tissues, such as high constitutive expression in adrenal glands and inducibility by 2,3,7,8-tetrachlorodibenzo-p-dioxin in heart, intestine, kidney, liver, lung, ovary and testes (26,27,29). Metabolism of PAHs by recombinant rat P450 1B1 has not yet been examined. Here we report on the rates of formation and metabolite distribution of B[a]P and 7,8-diol metabolism by human recombinant P450 1A1, 1A2 and 1B1 expressed at high content and specific activity. Metabolism experiments were conducted in the presence and absence of mEH so that the full spectrum of metabolites could be examined. We have also expressed recombinant rat P450 1B1 in yeast so that comparisons could be made between human and rat P450 1B1 for B[a]P and 7,8diol metabolism. Materials and methods Chemicals and reagents [3H]B[a]P (generally labeled, sp. act. 78.0 Ci/mmol, purity .98%) was purchased from Amersham Life Science (Arlington Heights, IL) and further purified through a column of silica gel type G using hexane as the eluant. The hexane was removed by evaporation under nitrogen gas. B[a]P metabolite standards 1,6-quinone, 3,6-quinone, 6,12-quinone, 3-OH, 7-OH, 9-OH, (6)benzo[a]pyrene-trans-4,5-dihydrodiol (4,5-diol), 7,8-diol, 9,10-diol, RTTC, RTTT, RTCT and RTCC were purchased from the National Cancer Institute Chemical Carcinogen Repository (Kansas City, MO). Silica gel type G, βNADP1, glucose 6-phosphate (G6P) and G6P dehydrogenase were purchased from Sigma (St Louis, MO). HPLC grade methanol, ethyl acetate, hexane and acetone solvents were purchased from J.T.Baker (Phillipsburg, NJ). Microsomal preparations of recombinant human cytochromes P450 1A1, 1A2 and 1B1 expressed in baculovirus-infected insect Sf9 cells and lymphoblastoid cell microsomes containing mEH were purchased from GENTEST (Woburn, MA). P450 content was determined spectrophotometrically (30). Microsomal preparations of recombinant human cytochrome P450 1B1 expressed in yeast, ∆0, were prepared as previously described (23). B[a]P and 7,8-diol metabolism Incubations contained 100 mM sodium phosphate, pH 7.4, 5 mM MgCl2, 15 mM G6P, 1 U/ml G6P dehydrogenase, 1.4 mM NADP1 and 5 pmol P450 in a final volume of 200 µl. Recombinant human mEH in the form of microsomal protein from a lymphoblastoid cell line was added at a final concentration of 250 µg/ml to those reaction mixtures where indicated. Reactions were preincubated at 37°C for 2 min and the reaction was initiated by addition of B[a]P or 7,8-diol substrate at a final concentration of 10 µM. Substrates were dissolved in dimethyl sulfoxide (DMSO), yielding a final DMSO concentration of 1% in the reaction mixture. Incubations were carried out for 15 min and were terminated by addition of 1 vol ice-cold acetone. B[a]P, 7,8-diol, and their metabolites were extracted twice with 2 vols ethyl acetate, vortexed and centrifuged for 10 min at 3600 g. The organic solvent

1848

extracts were combined and dried down under nitrogen. The residue was dissolved in 1 ml methanol and the sample was analyzed by HPLC. In all cases, extraction efficiency was .90% as determined by the amount of B[a]P and B[a]P metabolites which were recovered. HPLC analysis of B[a]P and 7,8-diol oxidation products Analysis of B[a]P and its metabolites by HPLC was performed with a liquid chromatograph consisting of a Waters U6K injector and two Waters Model 501 solvent delivery systems (Waters Associates, Milford, MA). Absorbance was monitored at 254 nm using a Waters Model 490 multiwavelength detector. Separations performed with a Prodigy ODS3 (Phenomenex, St Torrance, CA) reverse phase column (5 µm, 4.63250 mm) were carried out at ambient temperature with a flow rate of 1 ml/min using a methanol–water gradient system. Initial solvent conditions were 65% methanol with a linear gradient to 73% methanol in 20 min, then a linear gradient from 73 to 78% methanol in 5 min, a linear gradient from 78 to 82.4% methanol in 22 min, followed by a linear gradient from 82.4 to 100% methanol in 8 min. In addition, samples were analyzed on a 55% methanol isocratic solvent program for detection of 9,10-diol as well as tetrol metabolites. The identity of oxidation products was determined by co-elution of authentic standards. Radiochemical analysis was performed by a Radiomatic FLO-ONE\β Model A-200 detector system (Packard, Meriden, CT). Product amounts were calculated based on the specific activity of [3H]B[a]P substrate and quench factors determined using an internal standard. Analysis of 7,8-diol metabolism products was performed on a Varian HPLC (Sugar Land, TX) using Shimadzu CLASS VP software (Columbia, MD). Separations performed with a Waters µBondapak C18 reverse phase column (5 µm, 3.93300 mm) were carried out at ambient temperature with a flow rate of 1 ml/min. Tetrol metabolites and 7,8-diol were separated on a 50% methanol isocratic program. Products were detected using a Dynamax Fluorescence Detector Model FL-2 (Rainin, Emeryville, CA) with excitation at λ 5 340 nm and emission at λ 5 402 nm. Quantitation was by comparison with a standard curve using authentic standards. Expression of rat cytochrome P450 1B1 in yeast Construction of the rat P450 1B1 expression plasmid closely followed the strategy implemented for human P450 1B1 (23). To prepare the rP4501B1∆3S plasmid, the SacI–EcoRI fragment of the rat P450 1B1 cDNA clone 14 (29) was inserted into the SacI/EcoRI-digested yeast expression vector pYES2 (Invitrogen, Carlsbad, CA). This intermediate plasmid was digested with KpnI and RsrII and the 7.17 kb product, lacking the non-coding sequence of rat P450 1B1 and 345 nt of the coding sequence of the corresponding N-terminal end, was isolated. A DNA product encoding a modified rat P450 1B1 N-terminal sequence (deletion of amino acid residues 2–4) was generated by PCR amplification of clone 14 DNA as described (23). Primer 1 (59-GGG GTA CCA ACA GAT CAT GCT TAG TGC AGA CAG TCC A-39) modified this 59-region of the ATG, creating a KpnI site for cloning and deleting amino acid residues 2–4. The reverse complement primer 2 (59-TCC TTC CAG CGC TCT GAG TAG T-39) corresponded to the rat P450 1B1 cDNA sequence located 39 of the RsrII site. The PCR product was digested with KpnI and RsrII and cloned into the 7.17 kb product of the intermediate plasmid described above, to produce plasmid pTS3-1. Primer 2 was used to verify the sequence of the resultant plasmid, which corresponds to the PCR-generated DNA. In order to make pTS3-1 more closely resemble the human P450 1B1 expression plasmid ∆3 (23), plasmid pTS3-1 was digested with KpnI, treated with mung bean nuclease (Promega, Madison, WI) and religated to produce plasmid rP4501B1∆3S. Both the specific and modified rat P450 1B1 DNA sequences of this plasmid were verified by sequence analysis using specific synthetic primers as described previously (29). Plasmid rP4501B1∆3S was transformed into Saccharomyces cerevisiae strain JL 20. Expression of this recombinant rat P450 1B1 protein, microsome preparation and P450 quantitation followed methods described previously (23). Immunoblot analysis Microsomal proteins were separated by SDS–PAGE, electroblotted to Hybond ECL (Amersham, Arlington Heights, IL) and incubated with rabbit anti-P450 1B1 antibody at 5 µg IgG/ml for 1 h at room temperature as described previously (31). Bound antibody was detected by incubation with a horseradish peroxidase-conjugated goat anti-rabbit IgG (Promega, Madison, WI) at 0.03 µg/ ml and an enhanced chemiluminescence method (Super Signal; Pierce, Rockford, IL). Statistical analysis The rates and percent metabolite distributions presented in Tables I and II were analyzed by one way analysis of variance within samples and between samples using SigmaStat for Windows Version 1.0 published by Jandel Corporation (Chicago, IL). P-values of these pairwise comparisons are

B[a]P metabolism by human P450 1B1

Table I. Metabolism of benzo[a]pyrene by human P450s 1A1, 1A2, 1B1 and rat P450 1B1 Enzyme(s)a

1A1c 1A1 1

mEHc

1A2c 1A2 1

mEHc

1B1c 1B1 1

mEHc

1B1d 1B1 1

mEHd

1B1e 1B1 1 mEHe

Metabolites formed (nmol/min/nmol P450)b 9,10-diol

7,8-diol

1,6-quinone

3,6-quinone

6,12-quinone

9-OH

3- and 7-OH

Total

0.47 6 0.08 (12.8%) 0.44 6 0.15 (16.5%) ,0.02 (,3.3%) ,0.02 (,3.0%) 0.11 6 0.03 (10.3%) 0.13 6 0.04 (12.3%) 0.10 6 0.02 (7.3%) 0.21 6 0.07 (14.7%) 0.21 6 0.02 (10.0%) 0.56 6 0.03 (20.6%)

0.57 6 0.10 (15.5%) 0.38 6 0.12 (14.2%) ,0.02 (,3.3%) ,0.02 (,3.0%) 0.18 6 0.04 (16.8%) 0.17 6 0.06 (16.0%) 0.04 6 0.02 (2.9%) 0.31 6 0.08 (21.7%) ,0.02 (,1.0%) 0.47 6 0.01 (17.3%)

0.42 6 0.05 (11.4%) 0.27 6 0.08 (10.1%) 0.10 6 0.01 (16.7%) 0.08 6 0.01 (12.1%) 0.11 6 0.01 (10.3%) 0.11 6 0.02 (10.4%) 0.15 6 0.01 (11.0%) 0.16 6 0.01 (11.2%) 0.21 6 0.01 (10.0%) 0.16 6 0.01 (5.9%)

0.73 6 0.24 (19.8%) 0.48 6 0.24 (18.0%) 0.12 6 0.01 (20.0%) 0.12 6 0.01 (18.2%) 0.17 6 0.04 (15.9%) 0.17 6 0.04 (16.0%) 0.13 6 0.04 (9.5%) 0.07 6 0.01 (4.9%) 0.28 6 0.02 (13.2%) 0.30 6 0.03 (11.0%)

0.30 6 0.07 (8.1%) 0.22 6 0.06 (8.2%) 0.15 6 0.01 (25.0%) 0.18 6 0.04 (27.3%) 0.18 6 0.04 (16.8%) 0.19 6 0.02 (17.9%) 0.18 6 0.02 (13.2%) 0.16 6 0.01 (11.2%) 0.30 6 0.02 (14.2%) 0.26 6 0.03 (9.5%)

0.52 6 0.08 (14.1%) 0.33 6 0.05 (12.3%) 0.07 6 0.04 (11.7%) 0.10 6 0.01 (15.2%) 0.14 6 0.02 (13.1%) 0.12 6 0.03 (11.3%) 0.31 6 0.05 (22.8%) 0.20 6 0.04 (14.0%) 0.42 6 0.04 (20.0%) 0.32 6 0.01 (11.8%)

0.67 6 0.12 (18.2%) 0.55 6 0.02 (20.6%) 0.12 6 0.06 (20.0%) 0.14 6 0.03 (21.2%) 0.18 6 0.04 (16.8%) 0.17 6 0.03 (16.0%) 0.45 6 0.03 (33.1%) 0.32 6 0.07 (22.4%) 0.67 6 0.05 (31.6%) 0.65 6 0.06 (23.9%)

3.68 6 0.41 2.68 6 0.62 0.60 6 0.11 0.66 6 0.06 1.07 6 0.19 1.06 6 0.20 1.36 6 0.15 1.33 6 0.25 2.12 6 0.08 2.73 6 0.07

aMicrosomal

preparations of 5 pmol of recombinant P450 enzymes were incubated with and without microsomal epoxide hydrolase supplementation for 15 min at 37°C in 100 mM sodium phosphate pH 7.4 in the presence of 10 µM B[a]P, as described in Materials and methods. were analyzed by HPLC as described in Materials and methods. Rate values are the means 6 SEs of three separate experiments. The percentages of metabolite distribution are reported in the parentheses below the rate values. cMicrosomal preparations of human recombinant P450 enzymes expressed in baculovirus-infected insect cells. dMicrosomal preparations of human recombinant P450 1B1 expressed in yeast. eMicrosomal preparations of rat recombinant P450 1B1 expressed in yeast. bMetabolites

Table II. Metabolism of benzo[a]pyrene-7,8-diol by human P450s 1A1, 1A2, 1B1 and rat P450 1B1 Enzymea

1A1c 1A2c 1B1c 1B1d 1B1e

Metabolites formed (nmol/min/nmol P450)b RTTC

RTTT

Total DE 2

RTCT

RTCC

Total DE 1

DE 2:DE 1

1.75 6 0.29 (67.7%) 0.21 6 0.02 (46.5%) 0.30 6 0.02 (52.2%) 0.59 6 0.04 (49.2%) 0.77 6 0.10 (53.3%)

0.11 6 0.01 (4.3%) 0.04 6 0.02 (9.7%) 0.05 6 0.02 (8.0%) 0.07 6 0.01 (6.3%) 0.08 6 0.01 (5.3%)

1.86 6 0.30

0.11 6 0.01 (4.2%) 0.05 6 0.02 (10.6%) 0.05 6 0.02 (8.7%) 0.10 6 0.01 (8.7%) 0.15 6 0.03 (10.7%)

0.61 6 0.10 (23.7%) 0.15 6 0.01 (33.1%) 0.18 6 0.02 (31.0%) 0.43 6 0.03 (35.8%) 0.44 6 0.05 (30.6%)

0.72 6 0.11

2.58

0.19 6 0.04

1.26

0.25 6 0.02

1.40

0.54 6 0.04

1.24

0.59 6 0.07

1.44

0.24 6 0.04 0.35 6 0.04 0.67 6 0.04 0.85 6 0.10

aMicrosomal

preparations of 5 pmol of recombinant P450 enzymes were incubated for 15 min at 37°C in 100 mM sodium phosphate pH 7.4 in the presence of 10 µM benzo[a]pyrene-7,8-diol, as described in Materials and methods. bMetabolites were analyzed by HPLC as described in Materials and methods. Rate values are the means 6 SEs of three separate experiments. The percentages of metabolite distribution are reported in the parentheses below the rate values. cMicrosomal preparations of human recombinant P450 enzymes expressed in baculovirus-infected insect cells. dMicrosomal preparations of human recombinat P450 1B1 expressed in yeast. eMicrosomal preparations of rat recombinant P450 1B1 expressed in yeast. presented in the Results section and were determined using the Student– Newman–Keuls test.

Results B[a]P metabolism by human cytochrome P450 Separation of B[a]P oxidation products by HPLC is shown in Figure 1. The solid line indicates the detection of standards by UV absorbance at 254 nm and the dashed line is a sample radioactive trace of metabolites produced by P450 1A1 metabolism of B[a]P. The 4,5-diol, 7,8-diol, 1,6-quinone, 3,6quinone, 6,12-quinone and 9-OH metabolites were separated

under these conditions, while the 3-OH and 7-OH metabolites co-eluted. The 9,10-diol metabolite was detected using a second 55% methanol isocratic program, as shown by the inset in Figure 1. This program also separated the four tetrol metabolites and 9,10-diol metabolite, although no tetrol metabolites were detected. The relative rates of B[a]P metabolite production by recombinant human P450 1A1, 1A2 and 1B1, with and without mEH supplementation, are shown in Table I (upper). Total metabolism of B[a]P by P450 1A1 was 3.4-fold greater than the overall metabolism by P450 1B1 (Table I, summed rates 1849

J.H.Kim et al.

Fig. 1. Representative chromatogram of benzo[a]pyrene oxidation products by human P450 1A1 analyzed by HPLC. Ultraviolet absorbance at 254 nm is represented by the solid line (left, y-axis) and the radioactivity trace is represented by the dashed line (right, y-axis). The arrow marks the co-elution of B[a]P-9,10-diol and tetrol metabolites, which were later analyzed on a 55% methanol isocratic gradient (boxed inset).

on far right). P450 1A2 produced metabolites at a rate near 50% of P450 1B1, though this difference was not statistically significant. However, while the formation of phenol and quinone metabolites were similar between P450 1A2 and 1B1, P450 1B1 produced the diol metabolites, which were less than the limits of detection in P450 1A2 sample incubations. Total quinone metabolite formation was 0.37 and 0.46 nmol/min/ nmol P450 for P450 1A2 and 1B1 respectively. Total phenol metabolite formation was 0.19 and 0.32 nmol/min/nmol P450 for P450 1A2 and 1B1 respectively. Diol metabolite formation by P450 1B1 was comparable with phenol metabolite formation at 0.29 nmol/min/nmol P450. The rates of metabolism by P450 1B1 and 1A2 in insect Sf9 cell microsomes were not greatly affected by the addition of mEH to the sample incubations. P450 1A1 produced greater quantities of every metabolite compared with P450 1A2 and 1B1. Rates of formation of 9,10-diol, 7,8-diol, 1,6-quinone, 3-OH, 7-OH and 9-OH metabolites by P450 1A1 were greater (P , 0.05) than those for P450s 1A2 and 1B1. By percent metabolite distribution, 6,12quinone was produced as a minor product by P450 1A1 compared with P450s 1A2 and 1B1 (P , 0.05). Supplementation of incubations of P450 1A1 expressed in insect Sf9 cells with mEH was used to examine the effect of mEH on the rates and percent formation of diol metabolites. However, no effect was observed with mEH supplementation in insect Sf9 cells. Incubations with control microsomes not expressing P450 were negative for B[a]P metabolism, as were P450 incubations performed without NADP1 (data not shown). 7,8-Diol metabolism by human cytochrome P450 Separation of 7,8-diol oxidation products detected by fluorescence is shown in Figure 2. The lower trace in Figure 2A is a chromatogram showing separation of the four tetrol standards and 7,8-diol. The upper trace in Figure 2A is a representative chromatogram of the 7,8-diol oxidation products in a P450 1A1 incubation. Unknown oxidation products are indicated by asterisks. The chromatogram in Figure 2B is a 10-fold amplification of the upper trace (Figure 2A) illustrating the 1850

Fig. 2. Representative chromatogram of benzo[a]pyrene-7,8-diol oxidation products analyzed by HPLC with fluorescence detection. (A) Separation of authentic standards is shown in the lower trace. The upper trace is a representative chromatogram of 7,8-diol metabolism by P450 1A1 as described in Table II. Asterisks mark the detection of unknown oxidation products of 7,8-diol metabolism. (B) Chromatogram of a 10-fold amplification of the upper trace in (A), showing the B[a]P tetrol metabolites as well as unknown oxidation products.

B[a]P metabolism by human P450 1B1

Fig. 3. Expression of recombinant rat P450 1B1 in S.cerevisiae. Immunoblot analysis of microsomal protein is described in Materials and methods. Microsomal protein loadings were: lane 1, 5.6 µg microsomal protein from yeast expressing human P450 1B1; lane 2, 14.7 µg microsomal protein from yeast expressing rat P450 1B1; lane 3, 2.5 µg microsomal protein from rat adrenal glands. The position of molecular weight standards is shown on the left.

detection and separation of the tetrol metabolites, as well as unknown fluorescent 7,8-diol oxidation products. In incubations of P450 1B1 and 1A2, these unknown products were minor products or were not detected (data not shown). Each of the three human enzymes examined, P450 1A1, 1A2 and 1B1, produced the tetrol metabolites of the 7,8-diol substrate. The rates of formation of these metabolites are presented in Table II (upper). P450 1A1 had the greatest total rate for 7,8-diol metabolism (2.58 nmol/min/nmol P450), followed by P450 1B1 (0.60) and P450 1A2 (0.43). The DE 2-derived RTTC metabolite was formed at the greatest rate by each P450 enzyme, followed by formation of the DE 1derived RTCC metabolite. Much less of the RTCT and RTTT metabolites were formed. P450 1A1 produced significantly greater amounts of both the DE 2 and DE 1 metabolites compared with P450s 1A2 and 1B1 (P , 0.05). The DE 2:DE 1 ratio presented in Table II indicates a preference for P450 1A1 in forming DE 2. However, the difference in production of DE 2 and DE 1 metabolites by P450 1A1 was not statistically significant. P450 1A2 and 1B1 appear to produce almost equal amounts of DE 2 and DE 1 metabolites. The addition of mEH had no effect on 7,8-diol metabolism (data not shown). Incubations with control microsomes not expressing P450 were negative for 7,8-diol metabolism, as were P450 incubations performed without NADP1 (data not shown). Expression of rat P450 1B1 in yeast Rat P450 1B1 was expressed in S.cerevisiae. Microsomes prepared were spectrally determined to have a specific content of 13.6 pmol P450/mg microsomal protein. Immunoblot analysis of microsomal protein from yeast expressing human P450 1B1, rat P450 1B1 and rat adrenal glands, which have been shown to have high constitutive expression of P450 1B1 (26,31), is shown (Figure 3). Human P450 1B1 expressed in yeast was isolated at a specific content of 35.5 pmol P450/ mg microsomal protein. At these P450 expression levels, endogenous P450 reductase in this strain of yeast has previously been shown to not be limiting (23,31). Anti-P450 1B1 IgG was highly specific and no other bands were detected (31). Human P450 1B1 migrates as a 52 kDa protein and rat P450 1B1 as a 56 kDa protein in this SDS–PAGE analysis. B[a]P metabolism by human and rat cytochrome P450 1B1 The rates of formation of B[a]P oxidation products by human P450 1B1 and rat P450 1B1 are presented in Table I (lower). Rat P450 1B1 was approximately twice as active as human P450 1B1. Pairwise comparisons of total rates between human and rat P450 1B1, with and without mEH supplementation, were statistically significant (P , 0.05). For human P450 1B1, the phenol metabolites were formed in the greatest amounts, at 0.76 nmol/min/nmol P450. Quinone metabolites were formed at a total rate of 0.46 nmol/min/nmol P450. The two diol

metabolites, 7,8-diol and 9,10-diol, were formed in the least amounts, at rates of 0.04 and 0.10 nmol/min/nmol P450 respectively. Supplementation of yeast-expressed human P450 1B1 incubations with mEH resulted in a decrease in phenol metabolite production along with an increase in diol metabolite production. The production of 9-OH was decreased by 35% by mEH, to 0.20, while 9,10-diol production increased by 110%, to 0.21 nmol/min/nmol P450. The production of 3-OH and 7-OH was decreased by 29%, to 0.32 nmol/min/nmol P450, though 7,8-diol production increased by almost 800%, to 0.31. The increase in 7,8-diol production with mEH supplementation was statistically significant (P , 0.05). B[a]P metabolism by rat P450 1B1 produced the phenol metabolites in the greatest amounts, at a total rate of 1.09 nmol/ min/nmol P450. Rates of formation of quinone metabolites were also substantial, at a total rate of 0.79 nmol/min/nmol P450. In the absence of mEH, the 7,8-diol metabolite was undetected, though the 9,10-diol was produced at 0.21 nmol/ min/nmol P450. Supplementation of the sample incubations with mEH resulted in only slight decreases in phenol production, but large increases in diol production. 3-OH and 7OH metabolite formation was unaffected; however, 7,8-diol production increased from undetectable to 0.47 nmol/min/ nmol P450. 9-OH production decreased by 24%, to 0.32 nmol/ min/nmol P450, with a 160% significant (P , 0.05) increase in 9,10-diol production, to 0.56 nmol/min/nmol P450. Quinone production was slightly decreased in 1,6-quinone and 6,12quinone formation, though 3,6-quinone production showed a slight increase in the presence of mEH. Overall metabolism was significantly (P , 0.05) increased by mEH in rat P450 1B1 samples expressed in yeast, from 2.12 to 2.73 nmol/min/ nmol P450. 7,8-Diol metabolism by human and rat P450 1B1 The rates of formation of 7,8-diol oxidation products by human and rat cytochrome P450 1B1 are given in Table II (lower). Total metabolism by the rat enzyme was slightly greater than metabolism by human P450 1B1. The DE 2-derived metabolites were produced in greater quantity than the DE 1-derived metabolites. DE 2 metabolites had rates of formation of 0.67 nmol/min/nmol P450 for human P450 1B1 and 0.85 nmol/ min/nmol P450 for rat P450 1B1. These rates were comparable with the rates of formation of DE 1-derived metabolites. The DE 2:DE 1 ratio indicates a slight preference for formation of DE 2 by human and rat P450 1B1. The amounts of individual tetrol metabolite formation followed the same order of preference as human P450s 1A1, 1A2 and 1B1: RTTC . RTCC . RTCT 5 RTTT. Discussion Metabolism of B[a]P leads to formation of epoxide-derived diol and phenol metabolites and quinone metabolites (5). Phenol metabolites are less or non-mutagenic compared with the diol metabolites and have been thought to be pathways towards detoxification by glutathione S-transferase, UDP glucuronyl transferase and sulfotransferase (5). Quinone metabolites have also been thought to be non-mutagenic, however, there is recent evidence to indicate that they are able to form adducts with DNA (8,9). Mutagenic metabolites, such as the diol epoxides, have been found to cause chromatid breaks in vitro in lymphocyte cultures, which have been associated with increased risk of lung cancer in a case–control study (32). Diol epoxides, specifically DE 2, have been shown to form 1851

J.H.Kim et al.

DNA adducts in the p53 tumor suppressor gene in vitro and in HeLa cells and normal human bronchial epithelial cells (33,34). Enzymes of the P450 1 family are the principal catalysts of metabolism of B[a]P to the 7,8-diol and then the diol epoxides, which is the major route of formation of mutagenic species (5,18). The metabolism of B[a]P and 7,8-diol by recombinant human P450 1A1 and 1A2 has been studied extensively. There are several reports of the rates of diol, quinone and phenol formation by P450 1A1 and 1A2, as well as other P450 isoforms (13,15,16). A table summarizing these data has been presented (16) and there are differences between studies in the rates of formation of B[a]P metabolites, which are probably due to the use of different expression systems and reaction conditions. Here, we used recombinant P450 enzymes, expressed in comparable systems, to draw comparisons of metabolic activities. In this report, human P450 1A1 was shown to have the greatest capacity for B[a]P metabolism. Human P450 1B1 and 1A2 were less active towards B[a]P; 30 and 15% of P450 1A1 activity respectively. However, in contrast to P450 1B1, P450 1A2 was unable to form B[a]P metabolites of the mutagenic diol-epoxide pathway. Although the 4,5-diol metabolite was detected in other reports as a product of P450 1A1 and 1A2 metabolism (13,15) under our reaction conditions and method of analysis, we could not detect formation of 4,5diol by P450 1A1, 1A2 or 1B1. P450 1B1 produced the 7,8-diol metabolite at approximately half the rate of P450 1A1. The majority of B[a]P metabolism by P450 1A1 and 1B1 was to the phenol and quinone metabolites. Formation of the 7,8-diol was ~14 and 16% of the total metabolite production for P450s 1A1 and 1B1 respectively. Considering these data, it appears that 7,8-diol formation is proportionately similar between P450 1A1 and 1B1, despite the greater rate of formation by P450 1A1. In the baculovirus–insect cell system used for these comparisons, mEH did not appear to have a large effect on the metabolite distribution of B[a]P, suggesting that these microsomal preparations contain endogenous mEH activity. Each enzyme, P450 1A1, 1A2 and 1B1, was able to form the tetrol metabolites of the 7,8-diol substrate and exhibited a similar stereoselectivity, producing the RTTC metabolite at the greatest rate. This indicates a preference for formation of DE 2, which is the ultimate carcinogen (5), from 7,8-diol. By percent metabolite distribution, DE 1-derived metabolites were formed in the greatest amount by P450 1A2, followed by 1B1 and then by 1A1. In a previous study, it was reported that P450 1A1 produced DE 2-derived tetrol products in ~2-fold greater amount than DE 1-derived tetrols (13), while P450 1A2 produced similar amounts of the DE 2- and DE 1-derived tetrols. Our data are in agreement with this metabolite pattern, though our rates of tetrol metabolite formation are slightly higher. Here, we observed that P450 1B1 and 1A2 formed slightly more of the DE 2-derived tetrol products compared with the DE 1-derived tetrol products, as indicated by the DE 2:DE 1 ratio (Table II). P450 1A1 7,8-diol metabolism exhibited formation of large quantities of unknown metabolites not detected in reactions containing P450 1B1 and 1A2. The majority of these unknown compounds are highly polar and may include triol metabolites which were not examined here, further oxidation products of 7,8-diol metabolites or insect cell-derived conjugates. Human and rat P450 1B1 expressed at comparable levels 1852

in the same yeast strain were compared for B[a]P and 7,8diol metabolism. For B[a]P metabolism, addition of mEH had a substantial effect on the metabolite distribution. For both human and rat P450 1B1, the addition of mEH shifted formation away from phenols to diol metabolites. In the rat P450 1B1 incubations, exogenous mEH also increased the overall rate of metabolism by 30%. This is due mostly to the large increases in diol formation. In the presence of mEH, rat P450 1B1 exhibited 2-fold greater activity for B[a]P compared with human P450 1B1. Previous work on rat P450 1B1 has shown this enzyme to form the diols preferentially compared with the phenol metabolites in the presence of mEH (27). We found that the diol metabolites were formed in slightly larger amounts than the phenol metabolites but much greater than the quinone metabolites by rat P450 1B1. It was also reported that rat P450 1B1 could form small quantities of the 4,5-diol metabolite when exogenous mEH was added to the incubation mixture and that rat P450 1A1 produced the 4,5-diol predominantly compared with the other diol and phenol metabolites. In microsomes prepared from adrenals of male rats, which express P450 1B1, supplemented with exogenous mEH, 4,5-diol production was 8% of total metabolism (27). In the current study, we were unable to detect formation of the 4,5-diol. For 7,8diol metabolism, the differences between rat and human P450 1B1 were less dramatic. Rat P450 1B1 was only 19% more active than human. By metabolite distribution, rat and human P450 1B1 formed comparable percentages of the DE 2-derived tetrol metabolites. The results presented here indicate that both P450 1A1 and 1B1 participate in B[a]P metabolism at the first step in the pathway to DE 2 by formation of the 7,8-diol metabolite. In contrast, the 7,8-diol was not detected as an oxidation product of P450 1A2 metabolism of B[a]P, whereas phenol and quinone metabolites were formed. Formation of non-mutagenic phenols and mutagenic quinone metabolites was catalyzed by all three human enzymes. Regarding the second step in the production of DE 2, all three members of the P450 1 family catalyzed this reaction, although P450 1A1 exhibited the greatest rate of activity and percentage of RTTC formation. Given that human and rat P450 1B1 metabolize B[a]P to mutagenic species, this enzyme may play a significant role in in vivo B[a]P metabolism and carcinogenesis in humans and rat tumor models. Acknowledgements This study was supported by NIH grants ES08148, ES03760 and ES06052, Training Grant ES07141 and Center Grant ES03819. K.H.S. and N.J.W. were supported by NRSA fellowships ES05766 and ES05655 respectively. The authors thank Carrie Hayes, Ying Li and Frances Crofts for helpful discussions and technical assistance.

References 1. Greenberg,A., Hsu,C., Rothman,N. and Strickland,P.T. (1993) PAH profiles of charbroiled hamburgers: pyrene/B[a]P ratios and presence of reactive PAH. Polycyclic Aromatic Compounds, 3, 101–110. 2. Wood,A.W., Goode,R.L., Chang,R.L., Levin,W., Conney,A.H., Yagi,H., Dansette,P.M. and Jerina,D.M. (1975) Mutagenic and cytotoxic activity of benzo[a]pyrene 4,5-, 7,8-, and, 9,10-oxides and the six corresponding phenols. Proc. Natl Acad. Sci. USA, 72, 3176–3180. 3. Wood,A.W., Levin,W., Lu,A.Y.H., Yagi,H., Hernandez,O., Jerina,D.M. and Conney,A.H. (1976) Metabolism of benzo[a]pyrene and benzo[a]pyrene derivatives to mutagenic products by highly purified hepatic microsomal enzymes. J. Biol. Chem., 251, 4882–4890. 4. Wislocki,P.G., Wood,A.W., Chang,R.L., Levin,W., Yagi,H., Hernandez,O., Jerina,D.M. and Conney,A.H. (1976) High mutagenicity and toxicity of a

B[a]P metabolism by human P450 1B1 diol epoxide derived from benzo[a]pyrene. Biochem. Biophys. Res. Commun., 68, 1006–1012. 5. Gelboin,H.V. (1980) Benzo[a]pyrene metabolism, activation, and carcinogenesis: role and regulation of mixed-function oxidases and related enzymes. Physiol. Rev., 60, 1107–1166. 6. Pelkonen,O. and Nebert,D.W. (1982) Metabolism of polycyclic aromatic hydrocarbons: etiologic role in carcinogenesis. Pharmacol. Rev., 34, 189–222. 7. Conney,A.H., Chang,R.L., Jerina,D.M. and Wei,S.J.C. (1994) Studies on the metabolism of benzo[a]pyrene and dose-dependent differences in the mutagenic profile of its ultimate carcinogenic metabolite. Drug Metab. Rev., 26, 125–163. 8. Chesis,P.L., Levin,D.E., Smith,M.T., Ernster,L. and Ames,B.N. (1984) Mutagenicity of quinones: pathways of metabolic activation and detoxification. Proc. Natl Acad. Sci. USA, 81, 1696–1700. 9. Joseph,P. and Jaiswal,A.K. (1994) NAD(P)H:quinone oxidoreductase1 (DT diaphorase) specifically prevents the formation of benzo[a]pyrene quinone– DNA adducts generated by cytochrome P4501A1 and P450 reductase. Proc. Natl Acad. Sci. USA, 91, 8413–8417. 10. Hankinson,O. (1995) The aryl hydrocarbon receptor complex. Annu. Rev. Pharmacol. Toxicol., 35, 307–340. 11. Yang,S.K., Selkirk,J.K., Plotkin,E.V. and Gelboin,H.V. (1975) Kinetic analysis of the metabolism of benzo[a]pyrene to phenols, dihydrodiols, and quinones by high-pressure chromatography compared to analysis by aryl hydrocarbon hydroxylase assay, and the effect of enzyme induction. Cancer Res., 35, 3642–3650. 12. McManus,M.E., Burgess,W.M., Veronese,M.E., Huggett,A., Quattrochi,L.C. and Tukey,R.H. (1990) Metabolism of 2-acetylaminofluorene and benzo[a]pyrene and activation of food derived heterocyclic amine mutagens by human cytochromes P-450. Cancer Res., 50, 3367–3376. 13. Shou,M., Korzekwa,K.R., Crespi,C.L., Gonzalez,F.J. and Gelboin,H.V. (1994) The role of 12 cDNA-expressed human, rodent, and rabbit cytochromes P450 in the metabolism of benzo[a]pyrene and benzo[a]pyrene trans-7,8-dihydrodiol. Mol. Carcinog., 10, 159–168. 14. Shou,M., Korzekwa,K.R., Krausz,K.W., Buters,J.T., Grogan,J., Goldfarb,I., Hardwick,J.P., Gonzalez,F.J. and Gelboin,H.V. (1996) Specificity of cDNAexpressed human and rodent cytochrome P450s in the oxidative metabolism of the potent carcinogen, 7,12-dimethylbenz[a]anthracene. Mol. Carcinog., 17, 241–249. 15. Bauer,E., Guo,Z., Ueng,Y.F., Bell,L.C., Zeldin,D. and Guengerich,F.P. (1995) Oxidation of benzo[a]pyrene by recombinant human cytochrome P450 enzymes. Chem. Res. Toxicol., 8, 136–142. 16. Gautier,J.C., Lecoeur,S., Cosme,J., Perret,A., Urban,P., Beaune,P. and Pompon,D. (1996) Contribution of human cytochrome P450 to benzo[a]pyrene and benzo[a]pyrene-7,8-dihydrodiol metabolism, as predicted from heterologous expression in yeast. Pharmacogenetics, 6, 489–499. 17. Sutter,T.R., Tang,Y.M., Hayes,C.L., Wo,Y.P., Jabs,E.W., Li,X., Yin,H., Cody,C.W. and Greenlee,W.F. (1994) Complete cDNA sequence of a human dioxin-inducible mRNA identifies a new gene subfamily of cytochrome P450 that maps to chromosome 2. J. Biol. Chem., 269, 13092–13099. 18. Rendic,S. and Di Carlo,F.J. (1997) Human cytochrome P450 enzymes: a status report summarizing their reactions, substrates, inducers, and inhibitors. Drug Metab. Rev., 29, 413–580. 19. Shimada,T., Hayes,C.L., Yamazaki,H., Amin,S., Hecht,S.S., Guengerich,F.P. and Sutter,T.R. (1996) Activation of chemically diverse procarcinogens by human cytochrome P-450 1B1. Cancer Res., 56, 2979–2984. 20. Tang,Y.M., Wo,Y.Y.P., Stewart,J., Hawkins,A.L., Griffin,C.A., Sutter,T.R. and Greenlee,W.F. (1996) Isolation and characterization of the human cytochrome P450 CYP1B1 gene. J. Biol. Chem., 271, 28324–28330. 21. Wo,Y.Y.P., Stewart,J. and Greenlee,W.F. (1997) Functional analysis of the promoter for the human CYP1B1 gene. J. Biol. Chem., 272, 26702–26707. 22. Crespi,C.L., Penman,B.W., Steimel,D.T., Smith,T., Yang,C.S. and Sutter,T.R. (1997) Development of a human lymphoblastoid cell line constitutively expressing human CYP1B1 cDNA: substrate specificity with model substrates and promutagens. Mutagenesis, 12, 83–89. 23. Hayes,C.L., Spink,D.C., Spink,B.C., Cao,J.Q., Walker,N.J. and Sutter,T.R. (1996) 17β-Estradiol hydroxylation catalyzed by human cytochrome P4501B1. Proc. Natl Acad. Sci. USA, 93, 9776–9781. 24. Einolf,H.J., Story,W.T., Marcus,C.B. et al. (1997) Role of cytochrome P450 enzyme induction in the metabolic activation of benzo[c]phenanthrene in human cell lines and mouse epidermis. Chem. Res. Toxicol., 10, 609–617. 25. Crofts,F.G., Strickland,P.T., Hayes,C.L. and Sutter,T.R. (1997) Metabolism

of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) by human cytochrome P450 1B1. Carcinogenesis, 18, 1793–1798. 26. Bhattacharyya,K.K., Brake,P.B., Etom,S.E., Otto,S.A. and Jefcoate,C.R. (1995) Identification of rat adrenal cytochrome P450 active in polycyclic hydrocarbon metabolism as rat CYP1B1. J. Biol. Chem., 270, 11595–11602. 27. Otto,S., Bhattacharyya,K.K. and Jefcoate,C.R. (1992) Polycyclic aromatic hydrocarbon metabolism in rat adrenal, ovary, and testis microsomes is catalyzed by the same novel cytochrome P450 (P450RAP). Endocrinology, 131, 3067–3076. 28. Savas,U., Christou,M. and Jefcoate,C.R. (1993) Mouse endometrium stromal cells express a polycyclic aromatic hydrocarbon-inducible cytochrome P450 that closely resembles the novel P450 in mouse embryo fibroblasts (P450EF). Carcinogenesis, 14, 2013–2018. 29. Walker,N.J., Gastel,J.A., Costa,L.T., Clark,G.C., Lucier,G.W. and Sutter,T.R. (1995) Rat CYP1B1: an adrenal cytochrome P450 that exhibits sex-dependent expression in livers and kidneys of TCDD-treated animals. Carcinogenesis, 16, 1319–1327. 30. Omura,T. and Sato,R. (1964) The carbon monoxide-binding pigment of liver microsomes. I. Evidence for its hemoprotein nature. J. Biol. Chem., 239, 2370–2378. 31. Walker,N.J., Crofts,F.G., Li,Y., Lax,S.F., Hayes,C.L., Strickland,P.T., Lucier,G.W. and Sutter,T.R. (1998) Induction and localization of cytochrome P450 1B1 (CYP1B1) protein in the livers of TCDD-treated rats: detection using polyclonal antibodies raised to histidine-tagged fusion proteins produced and purified from bacteria. Carcinogenesis, 19, 395–402. 32. Wei,Q., Gu,J., Cheng,L., Bondy,M.L., Jiang,H., Hong,W.K. and Spitz,M.R. (1996) Benzo[a]pyrene diol epoxide-induced chromosomal aberrations and risk of lung cancer. Cancer Res., 56, 3975–3979. 33. Denissenko,M.F., Pao,A., Tang,M. and Pfeifer,G.P. (1996) Preferential formation of benzo[a]pyrene adducts at lung cancer mutational hotspots in p53. Science, 274, 430–432. 34. Puisieux,A., Lim,S., Groopman,J. and Ozturk,M. (1991) Selective targeting of p53 gene mutational hotspots in human cancers by etiologically defined carcinogens. Cancer Res., 51, 6185–6189. Received on March 2, 1998; revised on June 1, 1998; accepted on June 2, 1998

1853