532
Chem. Res. Toxicol. 1992,5, 532-540
Comparative Metabolism of Dibenz[ a j ]anthracene and 7-Methyldibenz[a,j ]anthracene in Primary Cultures of Mouse Keratinocytes Raghunathan V. Nair,? Anusha N. Nettikumara,? Cecilia Cortez,j Ronald G. Harvey,* and John DiGiovanni*p+ Department of Carcinogenesis, Science Park-Research Division, University of Texas M. D. Anderson Cancer Center, P.O. Box 389, Smithville, Texas 78957, and The Ben May Institute, University of Chicago, 5841 Maryland Avenue, Chicago, Illinois 60637 Received February 24, 1992
The identification of several metabolites formed from dibenz[aj]anthracene (DB[aj] A) and 7-methyldibenz[a jlanthracene (7MeDB[ajlA) in primary cultures of mouse keratinocytes is presented. The metabolites were analyzed by coelution with known synthetic standards using high-pressure liquid chromatography. The metabolite identifications were further facilitated by comparisons of fluorescence excitation and emission spectra obtained for isolated metabolites with those for the synthetic standards. Both DB[ajlA and ita 7-methyl analog were converted by mouse keratinocytes to the corresponding 3,4- as well as 5,6-dihydrodiols. The 5,6-dihydrodiol of DB[ajlA was the major intracellular metabolite found a t all time points up to 24 h. By 48 h, the relative proportion of both intracellular dihydrodiols had decreased, and they were found in approximately equal proportions. In contrast, 7MeDB[aj]A-3,4-dihydrodiol was the major intracellular metabolite of 7MeDB[ajlA a t all time points examined up to 48 h. The decreased intracellular retention of DB[ajlA-3,4-dihydrodiol was due, in part, to glucuronide conjugation and subsequent excretion of this metabolite. In this regard, both the 3,4as well as the 5,6-dihydrodiols of DBEajIA were found as glucuronides in the extracellular medium, whereas no such conjugates were detected extracellularly in cultures exposed to the 7-methyl derivative. The chromatographic profiles of cell- and medium-associated metabolites from both hydrocarbons ala0 exhibited several other metabolite peaks that remain to be identified. The observed differences in dihydrodiol metabolism by mouse keratinocytes could explain, in part, the greater biological activity of 7MeDB[ajlA relative to DB[aj]A.
Introduction Several methyl-substituted polycyclic aromatic hydrocarbons (PAHs)' are known to be more potent tumor initiators than the corresponding non-methylated derivatives (1-4). In particular, previous work from our laboratory established that 7-methyldibenz[ajlanthracene (7MeDBIajIA) is approximatelythree times more potent as a tumor initiator in mouse skin when compared with dibenz[ajlanthracene (DB[ajlA) (4). The biochemical mechanism(s) underlying the differential tumorigenicity of methylated versus non-methylated PAHs is not clearly understood. Recent studies have attempted to identify correlations between modifications of specific deoxyribonucleosides in cellular DNA caused by PAHs with their tumor-initiating potencies (5-7). Such studies have led to the hypothesis that modification of deoxyadenosine (dAdo) residues in DNA may be important for the greater biological activity of PAHs such as 7J2-dimethylbenz* To whom correspondence and reprint requests should be addressed.
University of Texas M. D. Anderson Cancer Center. University of Chicago. Abbreviations: PAH, polycyclic aromatic hydrocarbon; 7MeDB[aJl A, 7-methyldibenz[aJ]anthracene;DB[aj3A, dibenz[aJ]anthracene; dAdo,deoxyadenosine;DMBA,7,12-dimethylbenz[a]anthracene;MEM, minimum essential medium; FBS, fetal bovine serum; PBS, phosphatebuffered saline; HPLC, high-pressure liquid chromatography; MC, 3methylcholanthrene; BA, benz[a]anthracene; 7MeBA, 'I-methylbenz[a]anthracene; DB[a,hlA, dibenz[a,h]anthracene; DB[ajJAC, dibenz[ aJlacridine; DB[c,hlAC, dibenz[c,hlacridine; TLC, thin-layer chromatography; THF, tetrahydrofuran; NADP, nicotinamide adenine dinucleotide phosphate; DMSO, dimethyl sulfoxide; HPLC, high-pressure liquid chromatography;DDQ,2,3-dichloro-5,6-dicyano-l,4-benzoquinone. +
[alanthracene (DMBA) (6). However, such correlations have been based primarily on comparisons between PAHs of markedly different chemical structural features. We have recently examined the covalent modification of specific deoxyribonucleosides in mouse epidermal cell DNA following the exposure of cells to DB[ajlA and the structurally analogous 7MeDBIajlA (8Ia2The spectrum of covalently modified deoxyribonucleosides produced by these hydrocarbons (8) suggested both similarities and differences in their metabolic activation pathways in mouse epidermis. Therefore, in order to understand possible differences in metabolism of methylated (7MeDB[ aj l A) versus the corresponding non-methylated (DB[ aj l A) PAHs under similar experimental conditions,we examined the metabolism of DB[aj]A and 7MeDB[ajlA by mouse keratinocytes in culture. Cultured mouse keratinocytes serve as an in vitro model system for mouse epidermis in vivo. The present results show some differences in the metabolism of DB[aj]A and 7MeDB[ajlA that could explain, in part, the differences in tumorigenic potencies between these hydrocarbons.
t
Experimental Procedures Chemicals. DB[aj]A (9, IO), 7MeDB[ajlA (9),(+)-transDB[aj]A-3,4-dihydrodiol (II), (*)-trans-7MeDB[aJlA-3,4-di*Baer-Dubowska, W., Nair, R. V., Gill, R. D., Nettikumara, A. N., Cortez, C., Harvey, R. G., and DiGiovanni, J. (1990) Analysis of the covalent deoxyribonucleoside adducts formed in SENCAR mouse epidermis followingtopical application of dibenz[aJl anthracene (submitted for publication).
0 1992 American Chemical Society
Metabolism of Dibenz[a,j]anthracene
Derivatives
hydrodiol(1I), (f)-DB[aj]A-1,2,3,4-tetrahydrotetraol(12), and (*)-7MeDB[aj] A-1,2,3,4-tetrahydrotetraol(12) were prepared according to published procedures as indicated. Tetraol markers derived from hydrolysis of DB[ajlA syn diol epoxide were used after recovery following reaction of calf thymus DNA with DB[aj]A syn diol epoxide as described (8, 12). General tritium labelings of DB[aj]A and 7MeDB[aj]A were carried out by Chemsyn Science Laboratories, Lenexa, KS. Initial specific activities were between 1.8 and 3.5 Ci/mmol, depending on the lot of [3H]DB[a,j]A, and 1.5 Ci/mmol for [3Hl-7MeDB[ajlA. Labile tritium was removed, and radiochemical purity was determined by thin-layer chromatography (TLC) on silica gel in hexane/benzene (9:l for DB[aj]A; 7:3 for 7MeDBIajlA). Radiochemical purities as supplied by Chemsyn were -84 f 5% for DB[aj]A (three different lots) and 298% for 7MeDB[ajlA. The radioactive DB [a j l A was received dissolved in methylene chloride, whereas radioactive 7MeDB[ajlA was received as a solution in benzene. The [3HlDB[aj]A prepared in this way was further purified (purity 1 9 8 %) by preparative high-pressure liquid chromatography (HPLC) as described (13). [3HlDB[ajlA was also prepared by catalytic reduction (14)of 7-bromodibenz[ajlanthracene at the Tritium Labelling Laboratory, Du Pont NEN Research Products, Boston, MA, according to the following procedure. To a solution of 7-bromodibenz[aj]anthracene (50 mg) in tetrahydrofuran (THF) were added KOH/CH30H (10 mL), 5% Pd/C (15 mg), and tritium gas (50 Ci). The reaction mixture was stirred for 0.5 h a t ambient conditions. An uptake of 1.0 cc of tritium was noted. Labile tritium was removed with THF/CHaOH (1:l). [3H]DB[aj]A prepared in this way was supplied as a solution of THF. The specific activity of this material after purification by preparative HPLC (14) was 22.8 Ci/mmol, which was subsequently adjusted to 1 Ci/mmol by combination with unlabeled DB[aj]A prior to use. The purity of all radiolabeled hydrocarbons was 298% and was confirmed by HPLC as previously described (13). 8-Glucuronidase (Escherichia coli type IX; EC 3.2.1.31) and aryl sulfatase (limpet type v; EC 3.1.6.1) were purchased from Sigma Chemical Co. (St. Louis, MO). The syntheses of other hydrocarbon derivatives used in the present study are described below. (a) 7-Bromodibenz[ajlanthracene.A solution of DB[ajlA (500 mg, 1.8mmol) and N-bromosuccinimide (320 mg, 1.8mmol) in 20 mL of CC4 was heated at reflux for 30 min. Ferric chloride hexahydrate (15 mg) was added, and heating was continued for 16 h. The solution was allowed to cool and then filtered through a short column of Florisil and evaporated to dryness. The crude product was dissolved in benzene and chromatographed on a Florisil column eluted with benzene/hexane (1:9) to yield 7-bromodibenz[ajlanthracene (630 mg, 98%)essentially pure by TLC (a single spot), mp 250-253 OC: lH NMR (500 MHz, CDC13) 6 9.94 (8, 1 H14), 8.89 (d, 2, H1,13, J = 8.22 Hz), 8.83 (d, 2, &,e, J = 9.20 Hz), 7.87 (d, 2, H4,10, J = 7.55 Hz), 7.88 (d, 2, H5,9, J 9.26 Hz), 7.71 (t, 2, H2,12, J = 7.46 Hz), 7.63 (t, 2, H3,11,J = 7.62 Hz). The position of substitution by bromine was confirmed by the disappearance of the characteristic singlet at 6 8.37 for H, of DB[aj]A (15)and the downfieldshift oftheadjacent H6,8prOtonS from 6 7.92 to 8.83. Anal. Calcd for C22H13Br: C, 73.97; H, 3.67; Br, 22.37. Found: C, 74.04; H, 3.86; Br, 22.19. (b)(f)-cis-DB[aj1A-5,6-dihydrodiol. The reaction of DB[ a j ] A (1.10 g, 3.94 mmol) with Os04 (1.0 g, 3.94 mmol) in pyridine (50mL) was carried out by the general procedure described earlier (16);the reaction time was 3 h. The crude cis-dihydrodiol (1.11g) was purified by acetylation, chromatographyon a column of Florisil, and deacetylation with sodium methoxide in methanol to yield pure cis-dihydrodiol(860 mg, 70%),mp 226-228 OC: lH NMR (300 MHz, DMSO-& + D2O) 6 9.22 (8, 1, Hid, 9.10 (d, 1, HIS), 8.35 (d, 1,HI), 8.10 ( s , l , H,), 7.40-8.01 (m, 8, Ar), 4.85 (d, 1, Hs, J = 3.01 Hz), 4.72 (d, 1, H5, J = 3.28 Hz). Anal. Calcd for C22Hl602: C, 84.59; H, 5.16. Found: C, 84.44; H, 5.31. (c)DB[ajIA-I,B-dione.A solution of the cis-dihydrodiol(500 mg, 1.6 mmol) and 2,3-dichloro-5,6-dicyano-l,4-benzoquinone (900 mg, 4 mmol) in 20 mL of 1% aqueous tetrahydrofuran was stirred at room temperature for 1 h. A brick red precipitate
Chem. Res. Toxicol., Vol. 5, No. 4, 1992 533 formed. The reaction was quenched with water, and the precipitate was filtered off, washed with water, and dried to yield the brick red quinone (480 mg, 97%), mp 326-327 OC. Anal. Calcd for C22H1202: C, 85.70; H, 3.92. Found: C, 85.76; H, 4.00. (d)(*)-trans-DB[aj1A-5,6-dihydrodioI.Reduction of the 5,g-quinone (380 mg, 1mmol) with NaBH4 (1.2 g) in ethanol (300 mL) in the presence of 0 2 was carried out by the published procedure (17). The reaction time was 72 h. The crude transdihydrodiol(280 mg, 89 %) gave a single peak on HPLC analysis on a silica gel column eluted with 30 % T H F in hexane. Further purification via acetylation, chromatography on a Florisil column, and treatment with sodium methoxide in methanol furnished the analytical sample of the trans-dihydrodiol, which softened and turned red at 234 "C and melted a t 254-256 OC: lH NMR (DMSO-ds + D20) S 9.14 (8, 1,Hid), 9.02 (d, 1,H13, J = 8.3 Hz), 8.26 (d, 1,HI, J = 7.65 Hz), 8.11 (s, 1,H,), 7.36-7.95 (m, 8, Ar), 4.64(d,1,H6,J=9.4Hz),4.54(d,1,HglJ=9.4Hz). Anal.Calcd for C22H1602: C, 84.59; H, 5.16. Found: C, 84.56; H, 5.08. (e)DB[aj1A-5,6-oxide.The synthesis of this compound from the trans-5,6-dihydrodiol (78 mg, 0.25 mmol) 5y reaction with the dimethyl acetal of dimethylformamide (1mmol) in refluxing T H F (9 mL) was carried out by the general procedure described earlier (16). The pure 5,6-oxide (27 mg) melted at 160-162 OC with decomposition: lH NMR (500 MHz, CDC13) 6 9.34 (8, 1, Hid), 8.74 (d, 1, H13, J = 8.22 Hz), 8.37 (d, 1, Hi, J = 7-98),8.07 ( 8 , 1, H,), 7.36-7.71 (m, 8, Ar), 4.68 (d, 1,He, J = 3.95 Hz), 4.56 (d, 1, H5, J = 3.96 Hz). Anal. Calcd for C22H140: C, 89.77; H, 4.79. Found: C, 89.92; H, 4.68. (f) (f)-cis-7MeDB[aj1A-5,6-dihydrodiol. The reaction of 7MeDB[aj]A (1.22 g, 4.2 mmol) with Os04 (1.0 g, 3.9 mmol) in 50 mL of pyridine was carried out by the procedure employed for DB[aj]A; the reaction time was 18days. Following a similar workup procedure, the pure cis-dihydrodiol(800 mg, 62%) was obtained, which softened and turned brown at 209 "C and melted at 224-226 OC dec: lH NMR (500 MHz, DMSO-ds + D2O) 6 9.08 (5, 1, H14), 9.02 (d, 1, H13, J = 8.26 Hz), 8.18 (d, 1,Hi, J = 7.08), 8.07 (d, 1,He, J = 9.14 Hz), 7.96 (d, 1,H4, J = 7.88), 7.85 (d, 1, Hg, J=9.14Hz),7.34-7.70(m,6,Ar),5.13(d,l,Hs,J=3.35Hz), 4.70 (d, 1, Hg, J = 2.78 Hz), 2.80 (8, 1, CH3). Anal. Calcd for C23H1802: C, 84.64; H, 5.56. Found: C, 84.50; H, 5.48. (g)7MeDB[aj'A-5,6-dione. Reaction of the cis-dihydrodiol (155 mg, 0.47 mmol) with DDQ (320 mg, 1.41 mmol) by the procedure employed for the synthesis of the unmethylated quinone furnished the orange-red quinone (150mg, 100%), mp 222224 "C. Anal. Calcd for C23H1402: C, 85.70; H, 4.38. Found: C, 85.75; H, 4.22. (h)(f)-trans-7MeDB[aj1A-5,6-dihydrodiol. Reduction of the quinone (180 mg, 0.55 mmol) with NaBH4 (200 mg) by the procedure employed for the preparation of the trans-5,6-dihydrodiol of DB[aj]A (reaction time 48 h) afforded a mixture of the cis- and trans-dihydrodiols (5050by NMR analysis of the diacetates); total yield was 71 %. lH NMR data for the transdihydrodiol (DMSO-ds + DzO): 6 9.10 (8, 1,HlJ, 9.02 (d, 1,H13, J = 8.2 Hz), 8.18 (d, 1, Hi, J = 8.3 Hz), 8.06 (d, 1, Ha, J = 9.2 Hz), 7.36-7.95 (m, 7, Ar), 5.10 (d, 1,H6, J = 3.5 Hz), 4.68 (d, 1, Hg, J = 3.1 Hz), 2.8 (9, 3, CH3). (i)7-Formyl-DB[aj1A.Oxidation of 7MeDB[aj]A (100mg, 0.34 mmol) with DDQ (231mg, 1.02 mmol) in aqueous acetic acid (8 mL of HzO in 40 mL of HOAc) by the published method (18) for 48 h furnished the crude product. Recrystallization from benzene/hexane gave pure 7-formyl-DB[ajlA (47 mg, 44 % 1, mp 200-201 OC: 'H NMR (60 MHz, CDCl3) 6 12.08 ( ~ ,CHO), l , 10.08 (s, 1, H14), 7.2-9.0 (m, 12, Ar). Anal. Calcd for C23H140: C, 90.17; H, 4.61. Found: C, 89.94; H, 4.66. (j) 7-(Hydroxymethyl)-DB[aj~A.To a suspension of 7formyl-DB[aj]A (30 mg, 0.1 mmol) in 30 mL of ethanol was added NaBH4 (37 mg, 1.0 mmol), and the mixture was stirred overnight at ambient temperature under argon. The solvent was evaporated under vacuum, the organic residue was dissolved in benzene and concentrated to a small volume, and hexane was added. Crystallization gave the pure alcohol (22 mg, 71 % ): 'H NMR (500 MHz, DMSO-&) S 10.23 ( 8 , 1, Hid, 9.36 (d, 2, HlJ3,
634 Chem. Res. Toxicol., Vol. 5, No. 4, 1992 J = 8.2 Hz), 8.36 (d, 2, He,s,J = 9.5 Hz), 7.60-8.38 (m, 8, Ar), 5.43 (d, 2, CH2). Anal. Calcd for C23HleO: C, 89.53; H, 5.23. Found: C, 89.38; H, 5.18.
The hydrocarbons and their derivatives used in the present study should be considered carcinogens and should be handled with extreme care. Cell Cultures. Primary cultures of keratinocytes from dorsal skins of adult female SENCAR mice (NCI, Frederick, MD) were prepared according to established procedures (19) in low Ca2+ modified minimum essential medium (MEM) with growth factor supplements and 1% FBS. Cultures were switched to high Ca2+ (1.4 mM) modified MEM lacking supplements 24 h prior to treatment. Cells were seeded a t a density of approximately 2.5 X 1Oeper 35 mm2or 8 X lO6per 100 mm2petri dish. Cell cultures were treated with [3H]DB[aj]A or [3Hl-7MeDB[ajlA at a final media concentration of 1.7 nM. Cells and media from duplicate dishes were collected and pooled at various times following hydrocarbon treatments. For isolation of metabolites in quantities sufficient for spectral characterization, keratinocytes cultured in 100 mmzpetri dishes were treated with 1.7 nM media concentrations of PAH for 24 h. Large-scale metabolism experiments for 7MeDB[aj]A were also carried out utilizing 3methylcholanthrene (MC) induced mouse liver S9 preparations essentially as described (20). Extraction of Intra- a n d Extracellular Metabolites. Cells and media were collected a t 6,12,24, and48 h followingtreatments with [3H]DB[aj]A or [3H]-7MeDB[ajlA. For isolation of cellassociated metabolites, PBS (1mL) was added and the cells were scraped from the dishes. The dishes were further washed twice with PBS (0.5mL), and the washings were added to the cell suspension. The combined cell suspension was homogenized using a Polytron PTlO homogenizer (Brinkmann Instruments, Inc., Westbury, NY) for 15 s (3X) at setting 6. The cell homogenates were extracted with ethyl acetate/acetone (2:1,2 vol) to afford thecell-associatedmetabolites (21). Likewise, extraction of the collected media samples as described (21) afforded the extracellular metabolites. Media samples were further analyzed for the presence of glucuronide and sulfate conjugates according to previously published procedures (22,231. The intra- and extracellular metabolites obtained were further analyzed by HPLC. The quantities of metabolites necessary for obtaining the fluorescence spectra were acquired by pooling the organic extracts of media collected from numerous preparative cell culture experiments. The quantities of 7MeDB[ajlA metabolite peak I (Figures 3 and 6) sufficient for obtaining an NMR spectrum were acquired from large-scalereactions utilizing liver S9 (9OOOg) preparations from 7-9-week-old female SENCAR mice pretreated intraperitoneally with MC (80 mg/kg body weight in olive oil on 2 consecutive days, killed 24 h later). Reactions of 100-mL volumes containing 334-369 mg of protein, 1.3 mM NADP, 0.5 mM glucose 6-phosphate, 100 units of glucose-6-phosphate dehydrogenase, and 0.15 mM MgCl2 in 50 mM Tris-HC1, pH 7.5, were preincubated at 37 "C for 5 min, 7MeDB[aj]A (5 mL of 0.5 mg/mL solution in acetone) was added, and reactions were continued a t 37 OC for an additional 45 min. The metabolites were then extracted several times with ethyl acetate/acetone (2:1, 4 vol). Individual metabolites were isolated using preparative or semipreparative HPLC. HPLC Analysis. All analytical HPLC runs were carried out on an Altex Ultrasphere Octadecylsilane (ODS) column (4.6 mm X 25 cm) using an IBM LC/9533 ternary gradient liquid chromatograph equipped with a Shimadzu CR501 chromatopac. The column flow rate was 1 mL/min. The DB[aJ]A and 7MeDB[ajlA metabolites were separated by sequential elution: 5070% methanol in water gradient, linear, 20 min; 70% methanol in water, 10 min; 70-85% methanol in water gradient, linear, 20 min; 8 5 1 0 0 % methanol in water gradient, linear, 10 min. Individual 0.5-min fractions were collected in scintillation vials using an LKB 2070 Ultrorac I1 fraction collector concurrent with injection of 100-pLsamples into the column. Following addition of scintillation fluid (Ready Value, Beckman Instruments, Inc., Fullerton, CA), the radioactivity of fractions was determined
Nair et al. Table I. Distribution of Radioactivity after Treatment of Primary Cultures of Mouse Keratinocytes with [ W ] D B [ ~ J ~ for A 6, 12.24. and 48 h* time after 5% % remaining % % 8-glucuaddition of organic in aqueous remaining as ronidase [3HlDB[ajlA (h) soluble phase DB[ajlA releasable Intracellular 6 81 19 65 b 12 66 34 43 24 69 31 27 48 68 32 22 Extracellular 80 20 83 I 58 42 61 14 24 39 61 29 18 48 33 61 9 19 a Values in the table represent percentages of the totalradioactivity (in dpm) associated with the cells (intracellular)and media (extracellular). The intracellular-associatedradioactivity represented 15, 13,14, and 19% of the total input dpm (11.2 X 10s dpm, 1Ci/mmol) at the 6,12,24, and 48 h time points, respectively. The extracellularassociated radioactivity represented 85,87,86, and 81 % of the total input dpm at the 6,12,24, and 48 h time pointa, respectively. Values not determined. 6 12
*
using a Beckman LS 1800 liquid scintillation counter. The quantities of individual metabolites necessary for spectral characterization were isolated by the separation of metabolite mixtures collected from ethyl acetate/acetone (21) extraction of media samples from several preparative cell culture experiments on a Zorbax ODS column (21.2 mm X 25 cm) using a Du Pont Model 870 pump module equipped with a Series 8800 gradient controller, Model 860 absorbance detector, and an LKB 2210 two-channel recorder. The metabolites were eluted by a linear gradient of 50-100% methanol in water over 60 min. The column flow rate was 5 mL/min. Fractions (10 s for DB[ajlA, 30 s for 7MeDB[aj]A) collected using an LKB 2211 SuperRac following the injection of 2-mL samples (in methanol) into the column were aliquoted and analyzed by radioactive scintillation counting, and the respective metabolite peaks were pooled to obtain the individual metabolites. Metabolite peak I (Figures 3 and 6) in quantities sufficient for NMR analysis was isolated from reactions of mouse liver S9 preparations with unlabeled 7MeDBIaJlA by separation on an Altex Ultrasphere ODS column (10 mm X 25 cm) according to procedures described for analytical HPLC a t the flow rate of 3 mL/min. Fluorescence spectra for the various metabolites and synthetic markers were recorded using an SLM Aminco SPF-5OOC spectrofluorometer. The proton ['HI NMR spectrum for 7MeDBiajlA metabolite peak I (Figures 3 and 6) was recorded on a Varian XL400 spectrometer.
Results DB[ajJA Metabolism. The distribution of radioactivity at various time periods following exposure of primary cultures of mouse keratinocytes to 1.7 nM [3HlDB[ajlA is shown in Table I. The data represent percentages of total radioactivity associated with cells (intracellular) or media (extracellular)when partitioned into ethyl acetatel acetone (2:l)at each time point. The percentage of total radioactivity found intracellularlyor extracellularlyat each time point is given in the footnote to Table I. Treatment of media samples with 8-glucuronidase followed by extraction with ethyl acetate/acetone (2:l) released an additional 7, 14, 18, and 19% of the total extracellular radioactivity at 6,12,24,and 48 h, respectively (Table I). HPLC analyses of the ethyl acetate/acetone soluble material from the cells and the media at 24 h following treatment provided the profiles shown in Figure 1. It is clear that [3HlDB[ajlA was metabolized extensively by primary cultures of mouse keratinocytes. Peaks labeled
Chem. Res. Toxicol., Vol. 5, No. 4, 1992 535
Metabolism of Dibenz[a,j]anthracene Derivatives
)
OH
'.?u;_I I II *
250
.
.
300
400
.
.
. 500
Wavelength (nm)
1
- b 14-
E
Q 0
-
I
8
50
Fraction 3n Number
14
.--
100
lc u*
u,
I
Figure 1. HPLC elution profiles of intracellular and extracellular metabolites of [3HlDB[aj]A produced by primary cultures of mouse keratinocytes after 24 h of exposure to 0.5 pg/mL of hydrocarbon: (a) synthetic metabolite markers; (b) ethyl acetate/ acetone soluble metabolites found intracellularly; (c) ethyl acetate/acetone soluble metabolites found extracellularly. The total radioactivity injected for each sample was between 70 OOO and 100 000 dpm. UI-UI represent unknown metabolites. I and I1 in Figure 1 showed retention times identical to those for synthetic markers (*)-trans-DB[aj]A-5,6-dihydrodiol and (f)-trans-DB[aj]A-3,4-dihydrodiol,respectively. The identity of peak I (Figure 1)as the trans5,6-dihydrodiol was further supported by comparison of the fluorescence excitation and emissionprofiles obtained for the isolated metabolite (peak I, Figure 1) with those for the synthetic standard (Figure 2). The quantities of metabolite peak I1 (Figure 1)formed in the preparative cell culture experiments were not sufficient for obtaining the fluorescence spectrum. Both dihydrodiols were found
Figure 2. Fluorescence excitation and emission spectra of (a) (*)-trcms-DB[aj]Ad,6-dihydrodiol and (b) DB[aj] Ametabolite
peak I (Figure 1)in methanol.
in higher relative proportions intracellularly (Figure lb, Table 11). Absolute levels (Le., picomoles) of free DB[ a i l A-3,4-dihydrodiol were similar intracellularly and extracellularly at 6 and 12 h, whereas intracellular levels were -2-fold higher at 24 and 48 h (Table 11). Notably, @-glucuronidasetreatment of the extracellular medium released additional quantities of both the 5,6- and the 3,4-dihydrodiols. Examination of absolute levels of DB[ajIA-3,4-diol in the extracellular medium after treatment with @-glucuronidaserevealed much higher levels in the medium relative to the intracellular levels a t all time points (ratios of 2.4, 5.2, 2.1, and 2.7 a t 6, 12, 24, and 48 h, respectively) (Table 11). Treatment of media samples with aryl sulfatase prior to metabolite extraction failed to release additional radioactivity into the organic solvent, indicating the absence of formation of any sulfate conjugates. In addition to the dihydrodiols (peaks I and 11),the HPLC chromatograms (Figure 1) also exhibited several other peaks (UdJ4) of unknown identity. Among these unknown metabolites, peak U4 eluted in the region where phenolic metabolites were expected to elute. No metabolite peaks correspondingto the tetrahydrotetraolsderived from hydrolysis of either syn or anti DB[ajlA diol epoxides were noted in these chromatograms (Figure 1).In addition, a radioactive peak, at peak fraction 20, was noted in the HPLC chromatograms at some time points (see Figure IC). However, this peak was not consistently reproducible. Furthermore, the media samples collected at the 48-h time point showed an additional peak (peak fraction 49) closely associated with the unknown peak, Us. The proportion of this peak increased in the 24-h media sample following treatment with @-glucuronidase and therefore possibly represents a phenol or dihydrodiol type metabolite. The intracellular profile a t the48-h time point also showed an additional peak at fraction 43 which was not detected a t any other time point examined. The nature of these additional radioactive peaks also remains to be determined. 7MeDB[ajJA Metabolism. The radioactivity distribution found upon the exposure of mouse keratinocyte cultures to 1.7 nM [3H]-7MeDB[ajlA is given in Table 111. Again, the data in Table I11 represent percentages of total radioactivity associated with cells (intracellular) or media (extracellular) when partitioned into ethyl acetatel acetone (2:l). The percentage of total radioactivity found intracellularly or extracellularly at each time point is also given in the footnote to Table 111. Note that, for each
Nair et al.
536 Chem. Res. Toxicol., Vol. 5, No. 4, 1992
Table 11. Formation of Metabolites of DB[aJA by Primary Cultures of Mouse Keratinocytes as a Function of Culture Time. time after addition of [3HlDB[ajlA (h) % metabolite
6
12 Intracellular 1.3
0.5 0.5 1.9 14.1 4.9 (30)c 1.9 65.0
Ulb
uz
u3
5,6-dihydrodiol 3,4-dihydrodiol u4
DB[aJ]A
u1
1.7
6.9 21.3 9.2 (40) 4.2 43.2 Extracellular 7.7 2.3 1.5 9.2
3.6 0.5 0.4 4.1 0.8 (28) 1.9 (71) 0.1 83.4
uz
u3
5,6-dihydrodiol 3,4-dihydrodiol(-@-gluc)d 3,4-dihydrodiol (+@-gluc)e u4
DB[aJlA
2.1 (54) 6.5 (207) 0.6 61.2
24 2.5 4.0 11.9 15.3
48 11.1
6.2 27.4
3.7 6.8 5.0 6.3 (40) 5.8 22.4
22.0 5.8 3.1 8.1 1.6 (27) 5.1 (127) 1.0 28.8
33.0 6.3 1.9 2.7 1.4 (19) 5.0 (107) 1.1 8.7
11.2 (60)
0 Cell and media samples, collected at each time point indicated, were extracted with ethyl acetate/acetone (2:l) as described in methods and similar to the extraction described in Table I. Total radioactivity injected per sample into the HPLC column ranged from 70 OOO to 100 OOo dpm. Values represent percentages of total dpm eluted from the HPLC columns, and recoveries from the HPLC column were 98 3%. To determine the actual amount (i.e., picomoles) of a given metabolite at any time point, use the information provided in the footnote to Table I and the following formula:
organic/100)(%dpmHPLC/100) 2.2 X lo3dpm/pmol where % dpmI ,, E = % of total input dpm found intra- or extracellularly at a given time point from footnote to Table I, % dpmI , ,E organic = % of intra- or extracellular dpm extracted by organic solvent at a given time point as presented in Table I. %dpm HPLC = %dpm corresponding to a given HPLC peak at a given time point from Table 11. UI-U~represent unknowns. e Values in parentheses represent the actual picomoles of this metabolite found intracellularly at each time point determined using the formula detailed in footnote a. d Values in parentheses represent the actual picomoles of this metabolite found unconjugated in the extracellular medium at each time point determined using the formula detailed in footnote a. e Values in parentheses represent the total picomoles of this metabolite found in the extracellular medium (unconjugated and conjugated with glucuronic acid). Again, values were determined using the formula in footnote a, except that % dpmE included the additional radioactivity released by @-glucuronidasetreatment. pmol metabolite =
(11.2
X
lo6 dpm)(%dpm,,,E/100)(%dpm~,,E
Table 111. Distribution of Radioactivity after Treatment of Primary Cultures of Mouse Keratinocytes with [JH]-7MeDB[aJlA for 6, 12,24, and 48 ha time after addition of % % remaining % % 8-glucuronidase remaining [3H]-7MeDB- organic in aqueous phase 7MeDBIaJlA releasable [ailA (h) soluble Intracellular b 6 97 4 85 12 94 6 68 10 51 24 90 23 30 48 77 6 12 24 48
88 73 52 50
Extracellular 12 27 48 50
75 55 32 17
