Chem. Res. Toxicol. 1992,5,160-764
760
Articles Oxidation of Cyclopenta[ cdlpyrene by Human and Mouse Liver Microsomes and Selected Cytochrome P450 Enzymes Hoonjeong Kwon, Yousif Sahali, Paul L. Skipper, and Steven R. Tannenbaum* Department of Chemistry and Division of Toxicology, Massachusetts Institute of Technology, Room 56-311,Cambridge, Massachusetts 02139-4307 Received June 10, 1992 The metabolism of the environmental pollutant and suspected human carcinogen, cyclopenta[cdlpyrene (CPP), was investigated. Human liver microsomes from three individuals were examined, as well as CD-1 mouse liver microsomes. Five new metabolites recently identified in our lab, 4-hydroxy-3,4-dihydroCPP,3,4-dihydroCPP-cis-3,4-diol, 4-oxo-3,4-dihydroCPP, 3,4,9,10-tetrahydroCPP-trans-3,4-trans-9,10-tetrol, and trans-3,4-dihydroCPP-3,4,x-triols, as well as the known major metabolite, 3,4-dihydroCPP-trans-3,4-diol, were all observed from the incubations of human liver microsomes and CPP. Even though all three human samples were capable of producing all the metabolites identified from the mouse liver microsomal incubations of CPP, the quantity of each metabolite varied among the microsomal samples. In an attempt to explain the variation among human liver samples, the microsomes derived from genetically engineered cells containing specific cytochrome P450 isozyme cDNAs were employed. It was found that the 3,4-cyclopenta double bond can be oxidized by the cytochrome P450 enzymes lAl,lA2, and 3A4. The 9,lO K-region double bond was not efficiently oxidized by cytochrome P450 1A1,but by P450 1A2 either from CPP or from the t-3,li-dihydrodiol. The lack of catalytic activity of 3A4 toward the t-3,4-dihydrodiol, despite its high activity toward CPP oxidation to tetrahydrotetrols, suggests the possibility of two dihydrodiol epoxides, 3,4-dihydrodiol 9,lOepoxide and 9,10-dihydrodiol3,4-epoxide,of CPP.
Introduction Polycyclic aromatic hydrocarbons (PAHs)' are widely distributed environmental pollutants, which are carcinogenic only after metabolic activations. The metabolic activation of PAHs to their ultimate carcinogenic form has been suggested to occur at the bay region of the PAH molecules (1,2). Cyclopenta[cd]pyrene (CPP) is a PAH found in high concentration in diesel exhaust and is an urban air pollutant (3,4). It has also been detected in cigarette smoke (5). Though CPP lacks the bay region (see Figure 3 for the structures), it showed high mutagenicity upon activation with Aroclor 1254-treatedrat liver postmitochondrialsupernatant (PMS)and has been shown to induce tumors and carcinomas on mouse skin and in newborn mouse lung, respectively (6,7).While it was less active in inducing tumors than benzo[a]pyrene, most of the tumors induced by CPP became malignant and metastasized. CPP thus showed a higher malignant index than benzo[a]pyrene in the aforementioned studies. CPP neither has a geometric bay region nor can it form a benzo ring diol epoxide, but CPP-3,4-epoxide, the oxidationproduct across the localized double bond on the cyclopenta ring, has been proposed as an ultimate mutagen. The t-3,4-dihydrodiol has been reported as a major Abbreviations: CPP, cyclopenta[cd]pyrene;t(or c)-3,4-dihydrodiol, 3,4-dihydroCPP-trans(or cis)-3,4-diol; tetrols, stereoisomeric pair of 3,4,9,10-tetrahydroCPP-3,4,9,lO-tetrol; triols, stereoisomericpair of 3,4dihydroCPP-3,4,r-triole;EH, epoxide hydrolase;PAHs, polycyclic aromatic hydrocarbons; PMS, postmitochondrial supernatant; P450; cytochrome P450.
metabolite with t-9,lO-dihydrodiol as a minor one in the rat liver microsomal system (8). In mouse embryo fibroblast cells, only the t-3,4-dihydrodiol was reported (9). A study with various CPP derivatives, however, showed that the bacterial mutagenicities of 3,4-dihydrogenated CPP and the c-3,4-dihydrodiol,following PMS activation, were comparable to that of the parent compound, CPP (10). The aforementionedbacterial mutation assay suggested, therefore, that oxidation at sites in addition to the 3,4-double bond may have biological significance in the activation of this mouse carcinogen. This prompted us to investigate the detailed metabolism of CPP by the mouse, in which it has been shown to be carcinogenic. A number of new metabolites including tetrahydrotetrols and dihydrotriols have been identified (11). Since the metabolism of a compound is considered one factor that determines species difference, we have investigated the metabolism of CPP by human P450 enzymes2 as well as human liver microsomes and compared them to the mouse liver microsomal system, in an effort to establish a basis for interspecies comparison.
Materials and Methods Chemicals. [G-3H]CPP(385 mCi/mmol)was obtained from the NCI Chemical Repository (Lenexa, KS) and purified by HPLC prior to use. CPP is hazardous and should be handled with protective clothing in a well-ventilatedfume hood. Glucose 6-phosphate,8-NADP+,and glucose-6-phosphatedehydrogenase
* The nomenclature used in this report was taken from ref 20. 0 1992 American Chemical Society
Metabolism of Cyclopenta[cdlpyrene were obtained from Sigma Chemical Co. (St. Louis, MO). All solvents were chromatographic grade. Preparation of trans- and cis-3,4-Dihydrodiols. The [G-3HI-labeled t- and c-3,4-dihydrodiols were prepared from [G-3H-]CPP via epoxidation with dimethyldioxirane, followed by acid hydrolysis. CPP inmethylene chloride (at a concentration of 2 mM) was treated with a 3-fold molar excess of 70 mM dimethyldioxirane in acetone, prepared freshly by the published method (12). Upon completion of the epoxidation, which can be easily detected by the disappearance of the orange color of CPP and takes no more than 10min at 4 OC, the solvent was evaporated under a stream of nitrogen. The residue was dissolved in 50/50 (v/v) THF/water and acidified to a final concentration of 0.2 M HC1with concentrated HC1. Hydrolysis was complete within 10 min at room temperature. The resulting t-and c-3,4-dihydrodiols were produced in approximately 1:l ratio and were separated by reverse-phase HPLC with a same conditions used to separate CPP metabolites (see next section). The overall yield was higher than 80% by radioactivity. Metabolism Studies. Human liver microsomes HL 106,HL 110, and HL 115 were provided by Dr. F. P. Guengerich of Vanderbilt University (Nashville,TN). Mouse liver microsomes were prepared from CD-1mice by standard methods. Microsomes containing selected P450 enzymes or epoxide hydrolase were purchased from Gentest Corp. (Woburn, MA). They were prepared from geneticallyengineered human lymphoblastoidcells (AHH-1 TK+/-) transfected with the recombinant plasmid containing the specific P450 or epoxide hydrolase cDNA (13). Microsome-containinghuman CYPlAl, M l l l a , was reported to show 7-ethoxyresorufin deethylase activity, microsome M107a containing human CYP3A4showed testosterone 6B-hydroxylase activity but negligible 7-ethoxyresorufin deethylase activity, and microsome M103a containing human CYPlA2 was also reported to show high 7-ethoxyresorufindeethylase activity and activation of aflatoxin B1. Incubations were carried out at 37 OC for 3 h in 1mL of 0.1 M phosphate buffer (pH 7.4) containing 10 mM MgC12,0.7 mM &NADP+, 5 mM glucose 6-phosphate, 0.5 unit of glucose-6phosphate dehydrogenase, 1.0 mg of microsomal protein for the recombinant microsomes or 0.5mg for the liver microsomes, and 7 pM CPP or 5 pM t-3,4-dihydrodiol. Reactions were terminated by loading onto C18 BondElut (Analytichem International, Harbor City, CA) disposable columns. After washing out unbound material with 3 mL of water, bound material was sequentially eluted with 3 mL of 50/50 (v/v) methanol/water, 3 mL of 90/10 (v/v) methanol/water, and 1mL of methanol. The eluted fractions were combined, dried under reduced pressure, and subjected to HPLC analysis. The analysis of the metabolites was performed with a Hewlett Packard HP 1090 equipped with a diode array detector (Palo Alto, CA). A pBondapak C18 (3.9- X 300-mm) column from Waters (Milford, MA) and the following gradient at the flow of 1.5mL/min were used in this study: an isocratic phase with 10% methanol/water for 3 min, changed to 37 % methanol at 5 min, to 55% methanol at 10 min, to 68.5% methanol at 25 min, and then to 100% methanol at 45 min. The eluting fractions were collected every minute, and the radioactivity of each fraction was measured in a LS-315OP Beckman scintillation counter.
Rssults The same metabolites, compounds 2-7 (Figure 31, were produced by human or mouse liver microsomal oxidation of CPP. The most abundant metabolite in each case was 3,4-dihydroCPP-tram-3,4-diol, as observed previously when CPP was metabolized by rat liver microsomes (8)or mouse embryo fibroblasts (9). The other five metabolites were t h e cis-3,4-dihydrodiol (4), 4-hydroxy- and 4-oxo3,4-dihydroCPP (3 and 2, respectively), a diastereomeric pair of trans-3,4-tram-9,10-tetrahydrotetrols (7), and two closely related 3,4-dihydro-3,4-x-triols. The identification
40
c
Chem. Res. Toxicol., Vol. 5, No. 6,1992 761
bl-
0 C
20
.-c0 3
75 4-
6
10
1
2
3
Incubation Time, IHourl
Figure 1. Generation of trans-3,4-dihydrodiol (circles) and tetrols (squares). Mouse liver microsomes were used for the experiments. Lines were drawn arbitrarily. Table 1. Distribution of Metabolites Formed from CPP by Mouse and Various Human Liver Microsomes. metabolites* 5 (t-3,4-
6
27.1 h 1.6 14.2 4.3 20.6f 2.7
1.3 h 1.5 10.1 11.9 7 . 6 t 1.8
dihydrodiol) (triols) HL 110 HL115 HL 106 mouse
7
4 (~-3,4-
2,3
(4-OH
(tetrols) dihydrodiol) and 4-one) 17.5h0.9 24.9 30.9 9.2f 1.3
2.1 hO.6 1.7 1.4 4.3 f 0 . 7
3.2 hO.4 2.5 2.4 3.7 f0.6
Numbers are the percentage of total radioactivity in the particular peak from three measurements for HL 110 and mouse microsomes (average f SD) and from a single determination for HL 115and HL 106. Each microsomal preparation (0.5 mg/mL)was incubated with 7 pM CPP for 3 h at 37 O C in the presence of an NADPH generating system. Metabolites were then separated on HPLC equipped with a pBondapak C18 column, following solid-phase extraction on a disposable C18 column. See Figure 3 for structures and numbers.
or characterization of these additional metabolites is reported in ref 11. The concentrations of the t-3,4-dihydrodiol and tetrols as a function of incubation time were determined using microsomes, and the results are illustrated in Figure 1. The formation of the dihydrodiol was rapid initially. The concentration reached a peak at about 30 min, after which it declined to nearly half the maximum value. In contrast, the concentration of tetrols appeared to increase slowly during much of the 3-h course of the incubation. No further significant change in the concentration of either metabolite was apparent after about 2 h. The next experiments were designed t o compare the yields of even the least abundant metabolites, such as the tetrols, from incubation of CPP with microsomes of different origins. Since the yield of tetrols was shown to increase for at least 2 h and in order to obtain the best precision of measurement, we chose an incubation period of 3 h for these experiments. The results are collected in Table I. All of the human microsomes tested were able to oxidize both t h e 3,4- and 9,lO-double bonds of CPP. Assays of t h e mouse microsomes were conducted within 1 week of their preparation, and different preparations gave very similar analyses as indicated by the relatively small
Kwon et al.
762 Chem. Res. Toxicol., Vol. 5, No. 6, 1992
1
L
$
.-> .-w
with EH
0 without EH
8
8
.-g 6
6
(triols) P450 lAlb 3.9 h 1.7 P4501A2b 12.7f 3.9 P450 3A4b 0.1 f 0.1
6
U K m c
0
4
4
.-cC0 2
$ 2 c .n 0
0 t-Diol
4-OH 40x0
c-Diol
t-Diol
P450 1 A l
4-OH c-Diol Q-oxo P450 1 A 2
Figure 2. Effect of epoxide hydrolase on the formation of cyclopenta ring metabolites by (left) P450 1Al and (right) P450 1A2. When geneticallyengineered microsomalprotein containing primarily epoxide hydrolase was added to the incubation mixture (hatched), the yield of t-3,4-dihydrodiol was greatly increased and the yield of all other cyclopentaring metabolite was decreased. Table 11. Distribution of Metabolites Formed from CPP by P450 Enzymes Prepared from Genetically Engineered Cells. metabolitesb 5
(t-3,4dihydrodiol) P4501A1 6.3i0.5 P450 1A2 4.3f 1.0 P4503A4 8.4i0.5
6
(triols) 0.6f0.3 1.7 i0.3 0.7 f 0.2
Table 111. Distribution of Metabolites Formed from 3,4-DihydroCPP-trans-3,4-diol by Various Microsomal Preparations.
4
293
(~-3,4(4-OH (tetrols) dihydrodiol) and I-OXO) 0.4f0.3 1.1iO.4 4.1 f0.1 2.1 fO.3 1.4f0.2 0.8i 0.1 5.0i 0.6 2.3 f0.8 2.0f0.1 7
a Average f SD for the percentage of total radioactivity in the particular peak from three measurements, adjusted for the contribution of the control microsomes. Microsomal protein containing epoxide hydrolase (0.5 mg) was also added. The same experimental conditions were used as in Table I except 1.0 mg/mL microsomal protein was used for the enzymes. b Metabolite numbers correspond to Figure 3.
.
variation, typically about 10-15 % The standard deviation of the assay measured in HL 110 as well as that of mouse microsomal assay shows the consistency of the assay system. The possibility that different P450 enzymes are involved in the oxidation at different sites on CPP prompted us to investigate the role of different enzymes in more detail. The metabolism of CPP by microsomes prepared from human lymphoblastoid cells which had been transfected with cDNAs coding for human P450 lAl,lA2, or 3A4 (13) was examined. Microsomal epoxide hydrolase for use in these studies was also prepared from a genetically engineered cell line transfected with appropriate cDNA. Microsomes were prepared from the parental cell line to determine the basal P450 activity, which was mostly negligible. Figure 2 illustrates the effect of epoxide hydrolase on the relative amounts of the t-3,4-dihydrodiol and metabolites 2-4 produced by P450 1 A l and 1A2. In both cases, epoxide hydrolase enhanced the formation of the transdihydrodiol and suppressed the formation of the other metabolites. Thus CPP-3,Gepoxide is clearly a substrate for this particular epoxide hydrolase. Table I1 compares the distribution of metabolites 2-7 produced by each of the three P450's in the presence of epoxide hydrolase. In every case, the t-3,4-dihydrodiol was the major product. P450 1A2 and P450 3A4 also exhibited substantial activity in 9,lO K-region oxidation,
7
6
7
(tetrols) (triols) (tetrols) 0.4 f 0.0 HL l l o C 9.9 9.6 4.lf 1.2 mouse 8.7f 1.3 22.5f 2.9 0.1 f 0.1
a Average f SD for the percentage of total radioactivity from three measurements, adjusted for the contribution from control microsomes. 5 pM truns-3,4-dihydrodiol was incubated with each microsomal preparation in otherwise the same experimental conditions explained in Tables I and 11. The metabolite numbers correspond to Figure 3. Microsomal protein containing epoxide hydrolase (0.5 mg) was also added. Single measurement.
as evidenced by the proportion of triols and tetrols formed. In contrast, P450 1 A l exhibited little activity in K-region oxidation. Finally, the t-3,4-dihydrodiol was used as substrate for each of the three enzymes. As shown in Table 111, P450 1A2 was the most active of the three. P450 1 A l also exhibited substantial activity. Both of these enzymes produced far more of the triols than tetrols. Omission of microsomal epoxide hydrolase did not alter the trioktetrol ratio (data not shown). Mouse and HL 110 microsomes were also capable of oxidizing the 9,lO-double bond of the t-3,4-dihydrodiol. When these microsomes were used, the proportion of tetrols equaled or exceeded the proportion of triols. The significanceof the different ratios obtained using single enzymes or liver preparations is not quite clear, but it may be an indication that whole tissues contain other forms of epoxide hydrolase which can act on 3,4dihydrodiol9,lO-epoxides. Surprisingly, P450 3A4 was almost inactive despite the fact that this enzyme is reasonably efficient at converting CPP to tetrols and, to a lesser extent, triols. The clear implication is that P450 3A4 oxidizes the K-region 9,lOdouble bond in CPP itself but not in the partially oxidized 3,4-dihydrodiol.
Discussion The results presented in this paper reflect a more complex oxidativemetabolism of CPP than was previously understood. The trans-3,4-dihydrodiol and trans-9,lOdihydrodiol have long been known as microsomal metabolites. Recently, we elucidated an alternate fate for the 3,4-epoxide, which if it is not hydrolyzed to the t-3,4dihydrodiol, may give rise to any of three other products, structures 2,3, or 4. We also described two new types of products, which are the result of oxidation of both the 3,4and 9,lO-bonds in the same molecule. The present study provides some clarification of the enzymology involved in the formation of these more highly oxidized metabolites. Our results are indicative of the following, which is summarized in Figure 3. Either of the two transdihydrodiols may be a substrate for further oxidation to a diol epoxide species. The 3,4-dihydrodiol9,10-epoxide can hydrolyze to the two diastereomeric tetrahydrotetrols, or it can rearrange to two isomeric dihydrotriols. The isomerism must be positional, and we speculate that the phenolic hydroxyl group is at either the 9- or 10-position. We have no evidencethat the 9,10-dihydrodiol3,4-epoxide suffers any other fate than hydrolysis to tetrahydrotetrols. Rearrangement would result in a phenolic benzo[e]acenaphthalene chromophore, which we have not detected.
Chem. Res. Toxicol., Vol. 5, No.6, 1992 763
Metabolism of Cyclopenta[cdlpyrene I
10
1
f 3A4,1Al,1A2
I
OH
4
OH /
/
0
OH 5
f
3A4 z 1A2,lAl f
Ho&ol r
L
J
.
+
0
- I - -
HO
1
o.
@)OH 0
L
1A2 > lAl>> 3A4
OH
OH 0
-r----
J
OH
6
OH
OH
7
Figure 3. Metabolic pathway of CPP. The relative efficiencies of enzyme P450 l A l , l A 2 , and 3A4 are indicated for each step. The epoxides are presumptive intermediates;all other structures have been isolated and characterized. The cytochrome P450 enzymes exhibit markedly different activities. Most notably, P450 3A4 is almost inactive with respect to 3,4-dihydrodiol 9,lO-epoxide formation. Nevertheless, when given CPP as substrate, it produces tetrols in substantial yield as well as triols to a much lesser extent. It should be noted that when 3,4-dihydrodi019,10epoxide was formed in this reconstituted system, the proportion of the triols was always higher than that of tetrols (Table 111). P450 3A4must,therefore, oxidize CPP at the K-region 9,lO-double bond very efficiently. The tetrols and triols could only have arisen through the 9,lOepoxide formation in the case of this enzyme. The other cytochromesP450,lAl and 1A2, both appear to metabolize CPP by oxidation of the two double bonds in either order. This is more readily apparent in the case of P450 1A2, which produces triols and tetrols in a ratio of about 3.5:l when the t-3,4-dihydrodiol is the substrate. When CPP is the substrate, the ratio is close to unity, which may be taken as an indication of tetrol formation through oxidation of the 9,lO-dihydrodiol. The 3A4 enzyme was reported to be responsible for the activation of aflatoxin B1 and trans-7,8-dihydroxy-7,8dihydrobenzo[alpyreneand 3-hydroxylationof benzo[alpyrene among others in the human liver (14-16). The
other major human hepatic P450 enzyme, 1A2, has been known to activate aromaticamines more favorably,though it also oxidizes other polycyclic hydrocarbons. In the reconstituted system we have used, it showed very high catalyticactivitytoward K-region oxidation from the t-3,4dihydrodiol. Even though we have concentrated on the regioselectivities of different P450 enzymes, it may also be expected to observe stereoselectivities of these enzymes on the oxidationof CPP. Indeed, mouse liver microsomesshowed enantioselectivity in the formation of t-3,4-dihydrodiols (11). Stereoselective formation of the metabolites may play an important role in the biological activity of CPP as in the case of benzo[alpyrene diol epoxides. The anti7,8,9,10-tetrahydro-7,8-dihydroxy-9,lO-epoxide of benzo[alpyrene, for example, is a more potent mutagen and carcinogen than the syn isomer. Thus, it would be of interest to further investigatethe stereoselectiveformation of CPP metabolites by selected P450 enzymes. Although we failed to detect it, the 9,lO-dihydrodiol is a known metabolite of CPP, but only with rat liver microsomes. The experimental conditions used for most of the present work may have precluded ita observation, but we also examined mouse liver microsomal activation
764 Chem. Res. Toxicol., Vol. 5, No.6,1992
using conditions more favorable for its detection (5-30min incubation) without success. Others using mouse embryo fibroblasts have also not detected the 9,lOdihydrodiol (9). Thus, this diol with its high electron density, fixed 3,4-doublebond is probably rapidly oxidized further to 9,lO-dihydrodiol 3,4-epoxides. The 9,10-dihydrodiol3,4-epoxideappears to be a very good substrate for epoxide hydrolase since we found no evidence for any rearrangement products analogous to structures 2-4. On the other hand, the 3,4-dihydrodiol 9,lO-epoxide seems to be a poorer substrate for epoxide hydrolase judging from the extent of 3,4-dihydrodiol x-phenol formation. The biological activity of CPP 9,lOepoxide or the 3,4-dihydrodiol9,10-epoxideis not known, but other K-region epoxides of PAH including phenanthrene 9,lO-epoxideand indeno[l,2,3-cdlpyrene l,2-oxide have been shown to be mutagenic to bacteria (17,18). Tumorigenicity of benz[clacridine 5,6-epoxide, another K-region epoxide,has also been shown in a newborn mouse assay (19).These results, taken together with the reports of bacterial mutagenesis by CPP 3,4-derivatives (10)and the results reported in this paper, suggest that 3,4dihydrodiol9,lO-epoxide species may play an important role in the biological activity of CPP.
Acknowledgment. This work was supported by NIEHS Grants ES01640and ES02109, from the National Institutes of Health.
References (1) Jerina, D. M., Lehr, R. E., Yagi, H., Hernandez, O., Dansette, P. M.,
Wislocki, P. G., Wood, A. W., Chang, R. L., Levin, W., and Conney, A. H. (1976) Mutagenicity of benzo[alpyrene derivatives and the description of a quantum mechanical model which predicts the ease ofcarboniumion formation from diolepoxides. Inlnuitro metabolic activation in mutagenesis testing (De Serres, F. J., Fouta, J. R., Bend, J. R., and Philpot, R. M., Eds.) pp 159-195, Elsevier, Amsterdam. (2) Jerina, D. M., Yagi, H., Lehr, R. E., Thakker, D. R., SchaefferRidder, M., Karle, J. M., Levin, W., Wood, A. W., Chang, R. L., and Conney, A. H. (1978) The bay-region theory of carcinogenesis by polycyclicaromatic hydrocarbons. In Polycyclic hydrocarbons and cancer (Gelboin, H. V., and Ts’o, P. 0. P., Eds.) Vol. 1,pp 173-188, Academic Press, New York. (3) Tong, H. Y.,and Karasek, F. W. (1984) Quantitation of polycyclic aromatic hydrocarbons in diesel exhaust particulate matter by highperformance liquid chromatography fractionation and high-resolution gaa chromatography. Anal. Chem. 56,2129-2134. (4) Grimmer, G., Naujack, K.-W., and Schneider, D. (1980) Changes in PAH-profile in different areas of a city during the year. In Polynuclear Aromatic Hydrocarbons: Chem. Bio. Eff. (Bjoerseth, A., and Dennis, A. J., Eds.) pp 107-125, Battelle Press, Columbus. (5) Snook, M. E., Severson, R. F., Arrendale, R. F., Higman, H. C., and Chortyk, 0. T. (1977) The identification of high molecular weight polynuclear aromatic hydrocarbons in a biologically active fraction of cigarette smoke condensate. Beitr. Tabakforsch. 9,79-101. (6) Cavalieri, E. C., Rogan, E., Toth, B., and Munhall, A. (1981) Carcinogenicityof the environmental pollutants cyclopenteno[cdl-
Kwon et al. pyrene and cyclopentano[cd]pyrene in mouse skin. Carcinogenesis 2, 277-281. (7) Busby, W. F., Jr., Stevens, E. K., Kellenbach, E. R., Cornelisse, J., and Lugtenburg, J. (1988) Doseresponse relationships of the tumorigenicity of cyclopenta[cdlpyrene, benzo[alpyrene and 6 4 trochrysene in a newborn mouse lung adenoma bioassay. Carcinogenesis 9,741-746. (8) Eisenstadt, E., Shpizner, B., and Gold, A. (1981) Metabolism of cyclopenta[cdlpyrene at the K-region by microsomes and a reconstitutedcytochrome P-450 system fromrat liver. Biochem.Biophys. Res. Commun. 100,965-971. (9) Nesnow, S., Moore, M., Gold, A., and Eisenstadt, E. (1981) Cyclopenta[cdlpyrene: Metabolism, mutagenicity, and cell transformation. In Polynuclear Aromatic Hydrocarbons: Chem. Anal. and Biol. Fate (Cooke, M., and Dennis, A. J., Eds.) pp 387-396, Battelle Press, Columbus. (10) Cavalieri, E. L., Rogan, E. G., and Thilly, W. G. (1981) Carcinogenicity, mutagenicity and binding studies of the environmental contaminant cyclopenta[cd]pyrene and some of ita derivatives. In Polynuclear Aromatic Hydrocarbons: Chem. Anal. and Biol. Fate (Cooke, M., and Dennis, A. J., Eds.) pp 487-498, Battelle Press, Columbus. (11) Sahali, Y., Kwon, H., Skipper, P. L., and Tannenbaum, S. R. (1992) Microsomal metabolism of cyclopenta[cdlpyrene: Identification of new metabolites, absolute configuration and mechanisms. Chem. Res. Toxicol. 5, 157-162. (12) Adam, W., Chan, Y.-Y., Cremer, D., Gauss, J., Scheutzow, D., and Schindler, M. (1987) Spectral and chemical properties of dimethyldioxirane as determined by experiment and ab initio calculations. J. Org. Chem. 52, 2800-2803. (13) Gonzalez, F. J., Crespi, C. L., and Gelboin, H. V. (1991) cDNAexpressed human cytochrome P 4 5 k a new age of molecular toxicology and human risk assessment. Mutat. Res. 247,113-127. (14) Raney, K. D., Shimada, T., Kim, D.-H., Groopman, J. D., Harris, T. M., and Guengerich, F. P. (1992) Oxidation of aflatoxins and sterigmatocystin by human liver microsomes: Significance of aflatoxin Q1 as a detoxication product of aflatoxin B1. Chem. Res. Toxicol. 5, 202-210. (15) Yun,C.-H.,Shimada,T.,andGuengerich,F.P. (1992)Rolesofhuman liver cytochrome P4502C and 3A enzymes in the 3-hydroxylation of benzo(a)pyrene. Cancer Res. 52, 1868-1874. (16) Shimada, T., Iwasaki, M., Martin, M. V., and Guengerich, F. P. (1989)Human liver microsomalcytochromeP450 enzymes involved in the bioactivation of procarcinogensdetected by umu gene response in salmonella typhimurium TA 1535/pSK1002. Cancer Res. 49, 3218-3228. (17) Bucker, M., Glatt, H. R., Platt, K. L., Avnir, D., Ittah, Y., Blum, J., and Oesch, F. (1979) Mutagenicity of phenanthrene and phenanthrene K-region derivatives. Mutat. Res. 66, 337-348. (18) Rice, J. E., Coleman,D. T.,Hosted,T. J., LaVoie,E. J.,McCaustland, D. J., and Wiley, J. C., Jr. (1985) Identification of mutagenic metabolites of indeno[l,2,3-cd]pyrene formed in vitro with rat liver enzymes. Cancer Res. 45,5421-5425. (19) Chang, R. L., Levin, W., Wood, A. W., Kumar, S., Yagi, H., Jerina, D. M., Lehr, R. E., and Conney, A. H. (1984) Tumorigenicity of dihydrodiols and diol-epoxidesof benz[clacridine in newborn mice. Cancer Res. 44, 5161-5164. (20) Nebert et al. (1991) The P450 superfamily: Update on new sequences, gene mapping, and recommended nomenclature. DNA Cell Biol. 10, 1-14.
Registry No. 1, 27208-37-3; 2, 73473-56-0; 3, 80010-98-6; 4, 72273-59-7;5,72273-60-0;6 3,4,9-triol,143902-00-5;6 3,4,1O-triol, 143902-01-6;7 (stereoisomer l),139606-88-5;7 (stereoisomer2), 139684-61-0;P450, 9035-51-2.
760
Articles Oxidation of Cyclopenta[ cdlpyrene by Human and Mouse Liver Microsomes and Selected Cytochrome P450 Enzymes Hoonjeong Kwon, Yousif Sahali, Paul L. Skipper, and Steven R. Tannenbaum* Department of Chemistry and Division of Toxicology, Massachusetts Institute of Technology, Room 56-311,Cambridge, Massachusetts 02139-4307 Received June 10, 1992 The metabolism of the environmental pollutant and suspected human carcinogen, cyclopenta[cdlpyrene (CPP), was investigated. Human liver microsomes from three individuals were examined, as well as CD-1 mouse liver microsomes. Five new metabolites recently identified in our lab, 4-hydroxy-3,4-dihydroCPP,3,4-dihydroCPP-cis-3,4-diol, 4-oxo-3,4-dihydroCPP, 3,4,9,10-tetrahydroCPP-trans-3,4-trans-9,10-tetrol, and trans-3,4-dihydroCPP-3,4,x-triols, as well as the known major metabolite, 3,4-dihydroCPP-trans-3,4-diol, were all observed from the incubations of human liver microsomes and CPP. Even though all three human samples were capable of producing all the metabolites identified from the mouse liver microsomal incubations of CPP, the quantity of each metabolite varied among the microsomal samples. In an attempt to explain the variation among human liver samples, the microsomes derived from genetically engineered cells containing specific cytochrome P450 isozyme cDNAs were employed. It was found that the 3,4-cyclopenta double bond can be oxidized by the cytochrome P450 enzymes lAl,lA2, and 3A4. The 9,lO K-region double bond was not efficiently oxidized by cytochrome P450 1A1,but by P450 1A2 either from CPP or from the t-3,li-dihydrodiol. The lack of catalytic activity of 3A4 toward the t-3,4-dihydrodiol, despite its high activity toward CPP oxidation to tetrahydrotetrols, suggests the possibility of two dihydrodiol epoxides, 3,4-dihydrodiol 9,lOepoxide and 9,10-dihydrodiol3,4-epoxide,of CPP.
Introduction Polycyclic aromatic hydrocarbons (PAHs)' are widely distributed environmental pollutants, which are carcinogenic only after metabolic activations. The metabolic activation of PAHs to their ultimate carcinogenic form has been suggested to occur at the bay region of the PAH molecules (1,2). Cyclopenta[cd]pyrene (CPP) is a PAH found in high concentration in diesel exhaust and is an urban air pollutant (3,4). It has also been detected in cigarette smoke (5). Though CPP lacks the bay region (see Figure 3 for the structures), it showed high mutagenicity upon activation with Aroclor 1254-treatedrat liver postmitochondrialsupernatant (PMS)and has been shown to induce tumors and carcinomas on mouse skin and in newborn mouse lung, respectively (6,7).While it was less active in inducing tumors than benzo[a]pyrene, most of the tumors induced by CPP became malignant and metastasized. CPP thus showed a higher malignant index than benzo[a]pyrene in the aforementioned studies. CPP neither has a geometric bay region nor can it form a benzo ring diol epoxide, but CPP-3,4-epoxide, the oxidationproduct across the localized double bond on the cyclopenta ring, has been proposed as an ultimate mutagen. The t-3,4-dihydrodiol has been reported as a major Abbreviations: CPP, cyclopenta[cd]pyrene;t(or c)-3,4-dihydrodiol, 3,4-dihydroCPP-trans(or cis)-3,4-diol; tetrols, stereoisomeric pair of 3,4,9,10-tetrahydroCPP-3,4,9,lO-tetrol; triols, stereoisomericpair of 3,4dihydroCPP-3,4,r-triole;EH, epoxide hydrolase;PAHs, polycyclic aromatic hydrocarbons; PMS, postmitochondrial supernatant; P450; cytochrome P450.
metabolite with t-9,lO-dihydrodiol as a minor one in the rat liver microsomal system (8). In mouse embryo fibroblast cells, only the t-3,4-dihydrodiol was reported (9). A study with various CPP derivatives, however, showed that the bacterial mutagenicities of 3,4-dihydrogenated CPP and the c-3,4-dihydrodiol,following PMS activation, were comparable to that of the parent compound, CPP (10). The aforementionedbacterial mutation assay suggested, therefore, that oxidation at sites in addition to the 3,4-double bond may have biological significance in the activation of this mouse carcinogen. This prompted us to investigate the detailed metabolism of CPP by the mouse, in which it has been shown to be carcinogenic. A number of new metabolites including tetrahydrotetrols and dihydrotriols have been identified (11). Since the metabolism of a compound is considered one factor that determines species difference, we have investigated the metabolism of CPP by human P450 enzymes2 as well as human liver microsomes and compared them to the mouse liver microsomal system, in an effort to establish a basis for interspecies comparison.
Materials and Methods Chemicals. [G-3H]CPP(385 mCi/mmol)was obtained from the NCI Chemical Repository (Lenexa, KS) and purified by HPLC prior to use. CPP is hazardous and should be handled with protective clothing in a well-ventilatedfume hood. Glucose 6-phosphate,8-NADP+,and glucose-6-phosphatedehydrogenase
* The nomenclature used in this report was taken from ref 20. 0 1992 American Chemical Society
Metabolism of Cyclopenta[cdlpyrene were obtained from Sigma Chemical Co. (St. Louis, MO). All solvents were chromatographic grade. Preparation of trans- and cis-3,4-Dihydrodiols. The [G-3HI-labeled t- and c-3,4-dihydrodiols were prepared from [G-3H-]CPP via epoxidation with dimethyldioxirane, followed by acid hydrolysis. CPP inmethylene chloride (at a concentration of 2 mM) was treated with a 3-fold molar excess of 70 mM dimethyldioxirane in acetone, prepared freshly by the published method (12). Upon completion of the epoxidation, which can be easily detected by the disappearance of the orange color of CPP and takes no more than 10min at 4 OC, the solvent was evaporated under a stream of nitrogen. The residue was dissolved in 50/50 (v/v) THF/water and acidified to a final concentration of 0.2 M HC1with concentrated HC1. Hydrolysis was complete within 10 min at room temperature. The resulting t-and c-3,4-dihydrodiols were produced in approximately 1:l ratio and were separated by reverse-phase HPLC with a same conditions used to separate CPP metabolites (see next section). The overall yield was higher than 80% by radioactivity. Metabolism Studies. Human liver microsomes HL 106,HL 110, and HL 115 were provided by Dr. F. P. Guengerich of Vanderbilt University (Nashville,TN). Mouse liver microsomes were prepared from CD-1mice by standard methods. Microsomes containing selected P450 enzymes or epoxide hydrolase were purchased from Gentest Corp. (Woburn, MA). They were prepared from geneticallyengineered human lymphoblastoidcells (AHH-1 TK+/-) transfected with the recombinant plasmid containing the specific P450 or epoxide hydrolase cDNA (13). Microsome-containinghuman CYPlAl, M l l l a , was reported to show 7-ethoxyresorufin deethylase activity, microsome M107a containing human CYP3A4showed testosterone 6B-hydroxylase activity but negligible 7-ethoxyresorufin deethylase activity, and microsome M103a containing human CYPlA2 was also reported to show high 7-ethoxyresorufindeethylase activity and activation of aflatoxin B1. Incubations were carried out at 37 OC for 3 h in 1mL of 0.1 M phosphate buffer (pH 7.4) containing 10 mM MgC12,0.7 mM &NADP+, 5 mM glucose 6-phosphate, 0.5 unit of glucose-6phosphate dehydrogenase, 1.0 mg of microsomal protein for the recombinant microsomes or 0.5mg for the liver microsomes, and 7 pM CPP or 5 pM t-3,4-dihydrodiol. Reactions were terminated by loading onto C18 BondElut (Analytichem International, Harbor City, CA) disposable columns. After washing out unbound material with 3 mL of water, bound material was sequentially eluted with 3 mL of 50/50 (v/v) methanol/water, 3 mL of 90/10 (v/v) methanol/water, and 1mL of methanol. The eluted fractions were combined, dried under reduced pressure, and subjected to HPLC analysis. The analysis of the metabolites was performed with a Hewlett Packard HP 1090 equipped with a diode array detector (Palo Alto, CA). A pBondapak C18 (3.9- X 300-mm) column from Waters (Milford, MA) and the following gradient at the flow of 1.5mL/min were used in this study: an isocratic phase with 10% methanol/water for 3 min, changed to 37 % methanol at 5 min, to 55% methanol at 10 min, to 68.5% methanol at 25 min, and then to 100% methanol at 45 min. The eluting fractions were collected every minute, and the radioactivity of each fraction was measured in a LS-315OP Beckman scintillation counter.
Rssults The same metabolites, compounds 2-7 (Figure 31, were produced by human or mouse liver microsomal oxidation of CPP. The most abundant metabolite in each case was 3,4-dihydroCPP-tram-3,4-diol, as observed previously when CPP was metabolized by rat liver microsomes (8)or mouse embryo fibroblasts (9). The other five metabolites were t h e cis-3,4-dihydrodiol (4), 4-hydroxy- and 4-oxo3,4-dihydroCPP (3 and 2, respectively), a diastereomeric pair of trans-3,4-tram-9,10-tetrahydrotetrols (7), and two closely related 3,4-dihydro-3,4-x-triols. The identification
40
c
Chem. Res. Toxicol., Vol. 5, No. 6,1992 761
bl-
0 C
20
.-c0 3
75 4-
6
10
1
2
3
Incubation Time, IHourl
Figure 1. Generation of trans-3,4-dihydrodiol (circles) and tetrols (squares). Mouse liver microsomes were used for the experiments. Lines were drawn arbitrarily. Table 1. Distribution of Metabolites Formed from CPP by Mouse and Various Human Liver Microsomes. metabolites* 5 (t-3,4-
6
27.1 h 1.6 14.2 4.3 20.6f 2.7
1.3 h 1.5 10.1 11.9 7 . 6 t 1.8
dihydrodiol) (triols) HL 110 HL115 HL 106 mouse
7
4 (~-3,4-
2,3
(4-OH
(tetrols) dihydrodiol) and 4-one) 17.5h0.9 24.9 30.9 9.2f 1.3
2.1 hO.6 1.7 1.4 4.3 f 0 . 7
3.2 hO.4 2.5 2.4 3.7 f0.6
Numbers are the percentage of total radioactivity in the particular peak from three measurements for HL 110 and mouse microsomes (average f SD) and from a single determination for HL 115and HL 106. Each microsomal preparation (0.5 mg/mL)was incubated with 7 pM CPP for 3 h at 37 O C in the presence of an NADPH generating system. Metabolites were then separated on HPLC equipped with a pBondapak C18 column, following solid-phase extraction on a disposable C18 column. See Figure 3 for structures and numbers.
or characterization of these additional metabolites is reported in ref 11. The concentrations of the t-3,4-dihydrodiol and tetrols as a function of incubation time were determined using microsomes, and the results are illustrated in Figure 1. The formation of the dihydrodiol was rapid initially. The concentration reached a peak at about 30 min, after which it declined to nearly half the maximum value. In contrast, the concentration of tetrols appeared to increase slowly during much of the 3-h course of the incubation. No further significant change in the concentration of either metabolite was apparent after about 2 h. The next experiments were designed t o compare the yields of even the least abundant metabolites, such as the tetrols, from incubation of CPP with microsomes of different origins. Since the yield of tetrols was shown to increase for at least 2 h and in order to obtain the best precision of measurement, we chose an incubation period of 3 h for these experiments. The results are collected in Table I. All of the human microsomes tested were able to oxidize both t h e 3,4- and 9,lO-double bonds of CPP. Assays of t h e mouse microsomes were conducted within 1 week of their preparation, and different preparations gave very similar analyses as indicated by the relatively small
Kwon et al.
762 Chem. Res. Toxicol., Vol. 5, No. 6, 1992
1
L
$
.-> .-w
with EH
0 without EH
8
8
.-g 6
6
(triols) P450 lAlb 3.9 h 1.7 P4501A2b 12.7f 3.9 P450 3A4b 0.1 f 0.1
6
U K m c
0
4
4
.-cC0 2
$ 2 c .n 0
0 t-Diol
4-OH 40x0
c-Diol
t-Diol
P450 1 A l
4-OH c-Diol Q-oxo P450 1 A 2
Figure 2. Effect of epoxide hydrolase on the formation of cyclopenta ring metabolites by (left) P450 1Al and (right) P450 1A2. When geneticallyengineered microsomalprotein containing primarily epoxide hydrolase was added to the incubation mixture (hatched), the yield of t-3,4-dihydrodiol was greatly increased and the yield of all other cyclopentaring metabolite was decreased. Table 11. Distribution of Metabolites Formed from CPP by P450 Enzymes Prepared from Genetically Engineered Cells. metabolitesb 5
(t-3,4dihydrodiol) P4501A1 6.3i0.5 P450 1A2 4.3f 1.0 P4503A4 8.4i0.5
6
(triols) 0.6f0.3 1.7 i0.3 0.7 f 0.2
Table 111. Distribution of Metabolites Formed from 3,4-DihydroCPP-trans-3,4-diol by Various Microsomal Preparations.
4
293
(~-3,4(4-OH (tetrols) dihydrodiol) and I-OXO) 0.4f0.3 1.1iO.4 4.1 f0.1 2.1 fO.3 1.4f0.2 0.8i 0.1 5.0i 0.6 2.3 f0.8 2.0f0.1 7
a Average f SD for the percentage of total radioactivity in the particular peak from three measurements, adjusted for the contribution of the control microsomes. Microsomal protein containing epoxide hydrolase (0.5 mg) was also added. The same experimental conditions were used as in Table I except 1.0 mg/mL microsomal protein was used for the enzymes. b Metabolite numbers correspond to Figure 3.
.
variation, typically about 10-15 % The standard deviation of the assay measured in HL 110 as well as that of mouse microsomal assay shows the consistency of the assay system. The possibility that different P450 enzymes are involved in the oxidation at different sites on CPP prompted us to investigate the role of different enzymes in more detail. The metabolism of CPP by microsomes prepared from human lymphoblastoid cells which had been transfected with cDNAs coding for human P450 lAl,lA2, or 3A4 (13) was examined. Microsomal epoxide hydrolase for use in these studies was also prepared from a genetically engineered cell line transfected with appropriate cDNA. Microsomes were prepared from the parental cell line to determine the basal P450 activity, which was mostly negligible. Figure 2 illustrates the effect of epoxide hydrolase on the relative amounts of the t-3,4-dihydrodiol and metabolites 2-4 produced by P450 1 A l and 1A2. In both cases, epoxide hydrolase enhanced the formation of the transdihydrodiol and suppressed the formation of the other metabolites. Thus CPP-3,Gepoxide is clearly a substrate for this particular epoxide hydrolase. Table I1 compares the distribution of metabolites 2-7 produced by each of the three P450's in the presence of epoxide hydrolase. In every case, the t-3,4-dihydrodiol was the major product. P450 1A2 and P450 3A4 also exhibited substantial activity in 9,lO K-region oxidation,
7
6
7
(tetrols) (triols) (tetrols) 0.4 f 0.0 HL l l o C 9.9 9.6 4.lf 1.2 mouse 8.7f 1.3 22.5f 2.9 0.1 f 0.1
a Average f SD for the percentage of total radioactivity from three measurements, adjusted for the contribution from control microsomes. 5 pM truns-3,4-dihydrodiol was incubated with each microsomal preparation in otherwise the same experimental conditions explained in Tables I and 11. The metabolite numbers correspond to Figure 3. Microsomal protein containing epoxide hydrolase (0.5 mg) was also added. Single measurement.
as evidenced by the proportion of triols and tetrols formed. In contrast, P450 1 A l exhibited little activity in K-region oxidation. Finally, the t-3,4-dihydrodiol was used as substrate for each of the three enzymes. As shown in Table 111, P450 1A2 was the most active of the three. P450 1 A l also exhibited substantial activity. Both of these enzymes produced far more of the triols than tetrols. Omission of microsomal epoxide hydrolase did not alter the trioktetrol ratio (data not shown). Mouse and HL 110 microsomes were also capable of oxidizing the 9,lO-double bond of the t-3,4-dihydrodiol. When these microsomes were used, the proportion of tetrols equaled or exceeded the proportion of triols. The significanceof the different ratios obtained using single enzymes or liver preparations is not quite clear, but it may be an indication that whole tissues contain other forms of epoxide hydrolase which can act on 3,4dihydrodiol9,lO-epoxides. Surprisingly, P450 3A4 was almost inactive despite the fact that this enzyme is reasonably efficient at converting CPP to tetrols and, to a lesser extent, triols. The clear implication is that P450 3A4 oxidizes the K-region 9,lOdouble bond in CPP itself but not in the partially oxidized 3,4-dihydrodiol.
Discussion The results presented in this paper reflect a more complex oxidativemetabolism of CPP than was previously understood. The trans-3,4-dihydrodiol and trans-9,lOdihydrodiol have long been known as microsomal metabolites. Recently, we elucidated an alternate fate for the 3,4-epoxide, which if it is not hydrolyzed to the t-3,4dihydrodiol, may give rise to any of three other products, structures 2,3, or 4. We also described two new types of products, which are the result of oxidation of both the 3,4and 9,lO-bonds in the same molecule. The present study provides some clarification of the enzymology involved in the formation of these more highly oxidized metabolites. Our results are indicative of the following, which is summarized in Figure 3. Either of the two transdihydrodiols may be a substrate for further oxidation to a diol epoxide species. The 3,4-dihydrodiol9,10-epoxide can hydrolyze to the two diastereomeric tetrahydrotetrols, or it can rearrange to two isomeric dihydrotriols. The isomerism must be positional, and we speculate that the phenolic hydroxyl group is at either the 9- or 10-position. We have no evidencethat the 9,10-dihydrodiol3,4-epoxide suffers any other fate than hydrolysis to tetrahydrotetrols. Rearrangement would result in a phenolic benzo[e]acenaphthalene chromophore, which we have not detected.
Chem. Res. Toxicol., Vol. 5, No.6, 1992 763
Metabolism of Cyclopenta[cdlpyrene I
10
1
f 3A4,1Al,1A2
I
OH
4
OH /
/
0
OH 5
f
3A4 z 1A2,lAl f
Ho&ol r
L
J
.
+
0
- I - -
HO
1
o.
@)OH 0
L
1A2 > lAl>> 3A4
OH
OH 0
-r----
J
OH
6
OH
OH
7
Figure 3. Metabolic pathway of CPP. The relative efficiencies of enzyme P450 l A l , l A 2 , and 3A4 are indicated for each step. The epoxides are presumptive intermediates;all other structures have been isolated and characterized. The cytochrome P450 enzymes exhibit markedly different activities. Most notably, P450 3A4 is almost inactive with respect to 3,4-dihydrodiol 9,lO-epoxide formation. Nevertheless, when given CPP as substrate, it produces tetrols in substantial yield as well as triols to a much lesser extent. It should be noted that when 3,4-dihydrodi019,10epoxide was formed in this reconstituted system, the proportion of the triols was always higher than that of tetrols (Table 111). P450 3A4must,therefore, oxidize CPP at the K-region 9,lO-double bond very efficiently. The tetrols and triols could only have arisen through the 9,lOepoxide formation in the case of this enzyme. The other cytochromesP450,lAl and 1A2, both appear to metabolize CPP by oxidation of the two double bonds in either order. This is more readily apparent in the case of P450 1A2, which produces triols and tetrols in a ratio of about 3.5:l when the t-3,4-dihydrodiol is the substrate. When CPP is the substrate, the ratio is close to unity, which may be taken as an indication of tetrol formation through oxidation of the 9,lO-dihydrodiol. The 3A4 enzyme was reported to be responsible for the activation of aflatoxin B1 and trans-7,8-dihydroxy-7,8dihydrobenzo[alpyreneand 3-hydroxylationof benzo[alpyrene among others in the human liver (14-16). The
other major human hepatic P450 enzyme, 1A2, has been known to activate aromaticamines more favorably,though it also oxidizes other polycyclic hydrocarbons. In the reconstituted system we have used, it showed very high catalyticactivitytoward K-region oxidation from the t-3,4dihydrodiol. Even though we have concentrated on the regioselectivities of different P450 enzymes, it may also be expected to observe stereoselectivities of these enzymes on the oxidationof CPP. Indeed, mouse liver microsomesshowed enantioselectivity in the formation of t-3,4-dihydrodiols (11). Stereoselective formation of the metabolites may play an important role in the biological activity of CPP as in the case of benzo[alpyrene diol epoxides. The anti7,8,9,10-tetrahydro-7,8-dihydroxy-9,lO-epoxide of benzo[alpyrene, for example, is a more potent mutagen and carcinogen than the syn isomer. Thus, it would be of interest to further investigatethe stereoselectiveformation of CPP metabolites by selected P450 enzymes. Although we failed to detect it, the 9,lO-dihydrodiol is a known metabolite of CPP, but only with rat liver microsomes. The experimental conditions used for most of the present work may have precluded ita observation, but we also examined mouse liver microsomal activation
764 Chem. Res. Toxicol., Vol. 5, No.6,1992
using conditions more favorable for its detection (5-30min incubation) without success. Others using mouse embryo fibroblasts have also not detected the 9,lOdihydrodiol (9). Thus, this diol with its high electron density, fixed 3,4-doublebond is probably rapidly oxidized further to 9,lO-dihydrodiol 3,4-epoxides. The 9,10-dihydrodiol3,4-epoxideappears to be a very good substrate for epoxide hydrolase since we found no evidence for any rearrangement products analogous to structures 2-4. On the other hand, the 3,4-dihydrodiol 9,lO-epoxide seems to be a poorer substrate for epoxide hydrolase judging from the extent of 3,4-dihydrodiol x-phenol formation. The biological activity of CPP 9,lOepoxide or the 3,4-dihydrodiol9,10-epoxideis not known, but other K-region epoxides of PAH including phenanthrene 9,lO-epoxideand indeno[l,2,3-cdlpyrene l,2-oxide have been shown to be mutagenic to bacteria (17,18). Tumorigenicity of benz[clacridine 5,6-epoxide, another K-region epoxide,has also been shown in a newborn mouse assay (19).These results, taken together with the reports of bacterial mutagenesis by CPP 3,4-derivatives (10)and the results reported in this paper, suggest that 3,4dihydrodiol9,lO-epoxide species may play an important role in the biological activity of CPP.
Acknowledgment. This work was supported by NIEHS Grants ES01640and ES02109, from the National Institutes of Health.
References (1) Jerina, D. M., Lehr, R. E., Yagi, H., Hernandez, O., Dansette, P. M.,
Wislocki, P. G., Wood, A. W., Chang, R. L., Levin, W., and Conney, A. H. (1976) Mutagenicity of benzo[alpyrene derivatives and the description of a quantum mechanical model which predicts the ease ofcarboniumion formation from diolepoxides. Inlnuitro metabolic activation in mutagenesis testing (De Serres, F. J., Fouta, J. R., Bend, J. R., and Philpot, R. M., Eds.) pp 159-195, Elsevier, Amsterdam. (2) Jerina, D. M., Yagi, H., Lehr, R. E., Thakker, D. R., SchaefferRidder, M., Karle, J. M., Levin, W., Wood, A. W., Chang, R. L., and Conney, A. H. (1978) The bay-region theory of carcinogenesis by polycyclicaromatic hydrocarbons. In Polycyclic hydrocarbons and cancer (Gelboin, H. V., and Ts’o, P. 0. P., Eds.) Vol. 1,pp 173-188, Academic Press, New York. (3) Tong, H. Y.,and Karasek, F. W. (1984) Quantitation of polycyclic aromatic hydrocarbons in diesel exhaust particulate matter by highperformance liquid chromatography fractionation and high-resolution gaa chromatography. Anal. Chem. 56,2129-2134. (4) Grimmer, G., Naujack, K.-W., and Schneider, D. (1980) Changes in PAH-profile in different areas of a city during the year. In Polynuclear Aromatic Hydrocarbons: Chem. Bio. Eff. (Bjoerseth, A., and Dennis, A. J., Eds.) pp 107-125, Battelle Press, Columbus. (5) Snook, M. E., Severson, R. F., Arrendale, R. F., Higman, H. C., and Chortyk, 0. T. (1977) The identification of high molecular weight polynuclear aromatic hydrocarbons in a biologically active fraction of cigarette smoke condensate. Beitr. Tabakforsch. 9,79-101. (6) Cavalieri, E. C., Rogan, E., Toth, B., and Munhall, A. (1981) Carcinogenicityof the environmental pollutants cyclopenteno[cdl-
Kwon et al. pyrene and cyclopentano[cd]pyrene in mouse skin. Carcinogenesis 2, 277-281. (7) Busby, W. F., Jr., Stevens, E. K., Kellenbach, E. R., Cornelisse, J., and Lugtenburg, J. (1988) Doseresponse relationships of the tumorigenicity of cyclopenta[cdlpyrene, benzo[alpyrene and 6 4 trochrysene in a newborn mouse lung adenoma bioassay. Carcinogenesis 9,741-746. (8) Eisenstadt, E., Shpizner, B., and Gold, A. (1981) Metabolism of cyclopenta[cdlpyrene at the K-region by microsomes and a reconstitutedcytochrome P-450 system fromrat liver. Biochem.Biophys. Res. Commun. 100,965-971. (9) Nesnow, S., Moore, M., Gold, A., and Eisenstadt, E. (1981) Cyclopenta[cdlpyrene: Metabolism, mutagenicity, and cell transformation. In Polynuclear Aromatic Hydrocarbons: Chem. Anal. and Biol. Fate (Cooke, M., and Dennis, A. J., Eds.) pp 387-396, Battelle Press, Columbus. (10) Cavalieri, E. L., Rogan, E. G., and Thilly, W. G. (1981) Carcinogenicity, mutagenicity and binding studies of the environmental contaminant cyclopenta[cd]pyrene and some of ita derivatives. In Polynuclear Aromatic Hydrocarbons: Chem. Anal. and Biol. Fate (Cooke, M., and Dennis, A. J., Eds.) pp 487-498, Battelle Press, Columbus. (11) Sahali, Y., Kwon, H., Skipper, P. L., and Tannenbaum, S. R. (1992) Microsomal metabolism of cyclopenta[cdlpyrene: Identification of new metabolites, absolute configuration and mechanisms. Chem. Res. Toxicol. 5, 157-162. (12) Adam, W., Chan, Y.-Y., Cremer, D., Gauss, J., Scheutzow, D., and Schindler, M. (1987) Spectral and chemical properties of dimethyldioxirane as determined by experiment and ab initio calculations. J. Org. Chem. 52, 2800-2803. (13) Gonzalez, F. J., Crespi, C. L., and Gelboin, H. V. (1991) cDNAexpressed human cytochrome P 4 5 k a new age of molecular toxicology and human risk assessment. Mutat. Res. 247,113-127. (14) Raney, K. D., Shimada, T., Kim, D.-H., Groopman, J. D., Harris, T. M., and Guengerich, F. P. (1992) Oxidation of aflatoxins and sterigmatocystin by human liver microsomes: Significance of aflatoxin Q1 as a detoxication product of aflatoxin B1. Chem. Res. Toxicol. 5, 202-210. (15) Yun,C.-H.,Shimada,T.,andGuengerich,F.P. (1992)Rolesofhuman liver cytochrome P4502C and 3A enzymes in the 3-hydroxylation of benzo(a)pyrene. Cancer Res. 52, 1868-1874. (16) Shimada, T., Iwasaki, M., Martin, M. V., and Guengerich, F. P. (1989)Human liver microsomalcytochromeP450 enzymes involved in the bioactivation of procarcinogensdetected by umu gene response in salmonella typhimurium TA 1535/pSK1002. Cancer Res. 49, 3218-3228. (17) Bucker, M., Glatt, H. R., Platt, K. L., Avnir, D., Ittah, Y., Blum, J., and Oesch, F. (1979) Mutagenicity of phenanthrene and phenanthrene K-region derivatives. Mutat. Res. 66, 337-348. (18) Rice, J. E., Coleman,D. T.,Hosted,T. J., LaVoie,E. J.,McCaustland, D. J., and Wiley, J. C., Jr. (1985) Identification of mutagenic metabolites of indeno[l,2,3-cd]pyrene formed in vitro with rat liver enzymes. Cancer Res. 45,5421-5425. (19) Chang, R. L., Levin, W., Wood, A. W., Kumar, S., Yagi, H., Jerina, D. M., Lehr, R. E., and Conney, A. H. (1984) Tumorigenicity of dihydrodiols and diol-epoxidesof benz[clacridine in newborn mice. Cancer Res. 44, 5161-5164. (20) Nebert et al. (1991) The P450 superfamily: Update on new sequences, gene mapping, and recommended nomenclature. DNA Cell Biol. 10, 1-14.
Registry No. 1, 27208-37-3; 2, 73473-56-0; 3, 80010-98-6; 4, 72273-59-7;5,72273-60-0;6 3,4,9-triol,143902-00-5;6 3,4,1O-triol, 143902-01-6;7 (stereoisomer l),139606-88-5;7 (stereoisomer2), 139684-61-0;P450, 9035-51-2.
