Chem. Res. Toxicol. 1993,6,345-350
345
Acetylation of Phenolic Derivatives of 7H-Dibenzo[c,g]carbazole: Identification and Quantitation of Major Metabolites by Rat Liver Microsomes Weiling Xue, Joanne Schneider, Koka Jayasimhulu, and David Warshawsky* Department of Environmental Health, University of Cincinnati Medical Center, 3223 E d e n Avenue, Cincinnati, Ohio 45267-0056 Received September 21, 1992
Acetylation stabilized the phenolic metabolites of 7H-dibenzo[cg]carbazole (DBC) and made it possible to accumulate greater amounts of metabolites for comprehensive chemical structural elucidation and quantification without the use of radiolabeled DBC. High-resolution mass spectral data and lH NMR and fluorescence spectra were used to confirm the existence of 5-OH-DBC, 3-OH-DBC, 1-OH-DBC, and the oxidative dimer, 6,6’-bis-(5-OH-DBC), in the acetylated metabolite mixture formed in vitro by 3-methylcholanthrene-induced rat liver microsomes. Using the synthesized acetoxy-DBC derivatives as standards, the HPLC external standard method was employed for quantitation of the major DBC metabolites after acetylation. The quantities of 5-OH-DBC, 3-OH-DBC, 1-OH-DBC, and DBC in the metabolite mixture determined using the external standard method were found to agree with those calculated using the radiometric method. Acetylation is a promising nonradiometric qualitative and quantitative technique for further metabolism studies of DBC and analogues which produce unstable monohydroxylated metabolites.
Introduction The metabolism of 7H-dibenzo[cglcarbazole (DBC),’ an N-heterocyclic polynuclear aromatic carcinogen, has been studied through the last decade in an attempt to understand its carcinogenic mechanism (1-4). The main metabolites formed in vitro by rat or mouse liver microsomes have been found to be monohydroxylated derivatives of DBC. Among them, 5-hydroxy-DBC (5OH-DBC) and 3-hydroxy-DBC (3-OH-DBC) are the two major metabolites (1, 3 , 4 ) . Neither of them has shown significantsarcomatogenic and hepatocarcinogenic effects as the parent DBC has (5-7). N-Hydroxy-DBC (N-OHDBC) was expected to be the active metabolite, as was the case for the carcinogenic homocyclic and heterocyclic aromatic amines ( 2 , 5 , 8 , 9). Since N-OH-DBC appears extremely unstable and the instabilities of other phenolic derivatives are also well-known, it is difficult to collect sufficient pure metabolite samples from biologicalsystems. Hence, there is insufficient analytical data for structural confirmation of many metabolites found in metabolism studies. In many instances, the metaboliteb) ultimately responsible for carcinogenesis has(have) not been totally elucidated. Moreover, for quantitative studies of metabolite distribution, radiolabeled DBC is usually employed in an enzymatic system and the quantity of individual metabolites is calculated on the basis of the radioactivity of the correspondingfraction followingHPLC separation. In consideration of the shortage and the biohazard of radioactive compounds, the development of nonradiometric techniques is necessary. * To whom correspondence and reprint requests should be addressed.
j Abbreviations: DBC, ‘iH-dibenzo[cglcarbazole: 1-OH-DBC, l-hydroxy-DBC; 3-OH-DBC, 3-hydroxy-DBC; 5-OH-DBC, 5-hydroxy-DBC; 1-AcO-DBC, l-acetoxy6,6’-bis(5-OH-DBC), 6,6’-bis-(5-hydroxy-DBC); DBC; 2-AcO-DBC,2-acetoxy-DBC; 3-AcO-DBC,3-acetoxy-DBC;4-Ac0DBC, 4-acetoxy-DBC;5-AcO-DBC,5acetoxyDBC; 6-AcO-DBC,6-acetoxyDBC; 6,6’-bis-(5-AcO-DBC), 6,6’-bis(5-acetoxy-DBC); N-OH-DBC, N-hydroxy-DBC; 3-MC, 3-methylcholanthrene; SFS, synchronous fluorescence spectroscopy.
Acetylation has been widely used in derivatization of amino, hydroxy, and thiol compounds for chromatographic analysis (10). Some of the acetylated phenolic derivatives of DBC were reported stable ( 1 , I I ) . We found that the hydroxylated DBC derivatives were able to be converted quantitatively to the acetoxy-DBC compounds (11)that could be employed in analysis of DBC metabolites. In this article, we describe the application of the acetylation technique to structure elucidation and quantitation of major DBC metabolites formed in vitro by 3-methylcholanthrene (3-MC)induced rat liver microsomes. The acetylated metabolites of DBC were stable enough to be separated by HPLC and collected in -0.l-mg quantities for more comprehensivestructural analysis using 1H NMR and MS in addition to UV and fluorescence spectroscopy and to be quantified by application of the HPLC external standard method (12).
Materials and Methods Chemicals. (1,4,5,9,10,13,13b,13~-~~C)DBC (2.7 mCi/mmol) was purchased from Amersham (Arlington Heights, IL). The syntheses of unlabeled DBC, N-acetyl-DBC, 5-AcO-DBC, and 6,6’-bis(5-AcO-DBC)were previously reported by this laboratory (11). 1-,2-, 3-, 4-, and 6-AcO-DBC were prepared by acetylation of the corresponding hydroxylated DBC (synthesized in this laboratory) with acetic anhydride/pyridine following the procedures described (11). The acetylated DBC derivatives were recrystallized from benzenelhexane or hexane. The purities of these compounds were estimated to be >98% by peak area normalization (UV detector a t 254 nm) following HPLC separation. The stock solutions of the standard compounds were prepared in chloroform and stored a t -20 “C. The stock solutions of mixtures, when necessary, were generated by combining the solutions of the individual compounds. Analytical Instruments. lH NMR spectra were obtained using a Bruker NR/300 AF and a Bruker AMX 400 instrument in deuterated chloroform,and chemical shifts (6)were with respect
Q893-228~/93/2706-Q345$04.QQIQ 0 1993 American Chemical Society
Xue et al.
346 Chem. Res. Toxicol., Vol. 6, No. 3, 1993 to tetramethylsilane (0 ppm; s for singlet; d, doublet; dd, double doublet; t, triplet; m, multiplet). The coupling constants (Jvalue) were reported in hertz (Hz). High-resolution MS data were collected on a Kratos Model MS 80 instrument using a n internal standard of perfluorokerosine. Fluorescence spectra were recorded using a Perkin-Elmer MPF-66 fluorescence spectrophotometer with a 7700 professional computer. SFS spectra were produced when excitation and emission monochromators were scanned simultaneously with a fixed-wavelength difference between emission and excitation (AX) of 10 nm. Radioactivity of samples in Scintiverse scintillation cocktail (Fisher Scientific Co., Cincinnati, OH) was determined using a Packard Tri Carb Model 460 liquid scintillation counter. Metabolism. The liver microsomes were prepared from male Sprague-Dawley rats pretreated with 3-MC as previously reported (4). The aryl hydrocarbon hydroxylase enzyme activity of the microsomal protein was 2859 units/mg of protein as measured by picomoles of 3-OH-BaP formed (13). Metabolites were produced by incubating DBC [0.1 pCi of (14C)DBC was diluted with unlabeled DBC to total 0.24 pmol in 40 pL of methanol per assay] in an NADPH generating system (total volume of 4 mL) with 0.4 mg of protein/mL for 1 h as described previously ( 4 ) . The reaction was quenched with ice-cold acetone and extracted twice with 4 mL of ethyl acetate. The latter was evaporated to dryness under nitrogen for acetylation or was stored a t -20 "C, in the dark, for further HPLC analysis. Acetylation of Metabolites. The metabolites obtained from the incubation described above were mixed with acetic anhydride (0.4 mL) and pyridine (50 pL) a t ambient temperature in the dark for 3 h. After neutralization, extraction with ethyl acetate of the reaction mixture, and evaporation of the solvent described before ( I I ) , the dry samples were redissolved in 180 pL of chloroform and an aliquot of 18 pL was used for each HPLC analysis. For NMR and MS analysis, 10 microsomal assays were combined for acetylation and the products were dissolved in 300 pL of chloroform for multiple injections onto an analytical HPLC column (see below). Repeated collections of the appropriate fractions were evaporated to dryness under nitrogen a t room temperature. HPLC. DBC metabolites and the acetylated DBC metabolites, respectively, were separated by a Waters HPLC system using multistage linear methanol/water gradients on an analytical reversed-phase column (Whatman Partisil-10 ODS-2,10 pm, 25X 0.45-cm i.d.) with a flow rate of 1mL/min at 20 "C. For DBC metabolites, the HPLC gradient was 76-79% methanol over 6 min, 79-85% methanol over 3 min, 85-100% methanol over 18 min, and 100% methanol for 11 min. For the acetylated DBC metabolites, the following gradient programs were employed: (a) 8 0 % methanol for 2 min, 80-85% methanol over 5 min, isocratic for 5 min, 85-90% methanol over 6 min, isocratic for 5 min, and 90-80% methanol over 4 min; (b) 8 2 % methanol for 8 min, 82-85 % methanol over 2 min, isocratic for 5 min, 85-88 % methanol over 6 min, isocratic for 5 min, and 88-82% methanol over 4 min; (c) 70-90% methanol over 35 min, isocratic for 5 min, 90-100% methanol over 2 min, and 100-70% methanol over 5 min. The UV absorption of the eluent was monitored a t 254 nm using a Waters 484 tunable absorbance detector. Chromatographic peak integration was performed with a Maxima 820 chromatographyworkstation (Waters Dynamic Solution Division of Millipore, Ventura, CA), and the peak area was reported in units of millivolt second (mV-s). Fractions corresponding to HPLC peaks were collected manually for fluorescence spectroscopy. For 'H NMR and MS analysis, multiple collections of the fractions of interest were performed and then evaporated to dryness. For radioactivity determinations, 1-mL fractions were collected on an ISCO Model 2111 fraction collector.
Results and Discussion The course of the acetylation reaction of hydroxy-DBC using acetic anhydride/ pyridine was monitored by HPLC, and the reaction was usually completed in 2.5 h; the
Table I. Melting Point and Mass Spectral Data of Acetoxy-DBC Derivatives MS and relative intensity ( % ) compound mp" ("C) molecular ionb major fragmentsi 1-AcO-DBC 215-7 325.1083 (39) 283.1012 (100) 2-AcO-DBC
183-5 (185)
3-AcO-DBC
208-10 (211) 325.1099 (79)
4-AcO-DBC
184-6 (185)
325.1102 (77)
5-AcO-DBC
26C-2 (11)
325.1096 (24)
6-AcO-DBC
169-71 (163) 325.1133 (79)
325.1131 (68)
6,6'-bis(5-AcO-DBC) 207-10 (11) fraction 1
64gd 325.1129 (11)
fraction 2
325.1108 (31)
254.0930 (22) 283.0964 (100) 254.1017 (18) 283.1016 (100) 254.0946 (53) 283.1016 (100) 254.0977 (55) 283.0958 (100) 254.0949 (31) 283.0956 (100) 254.0996 (81) 283.0974 (100) 254.1008 (52) 283.1017 (100) 254.0927 (10)
a Numbersinparenthesesarevaluesinref1. Calcdfor C22H15N02, 325.1103. Calcd for C ~ O H ~ ~283.0997; N O , calcd for C19H12N,254.0970. Chemical ionization mass, M + 1.
products were pure and stable. N-Acetylation products were not found. The procedures in aqueous sodium hydroxide solution with acetic anhydride ( I ) were tested yielding incomplete acetylation. The physical properties and spectral data of the acetoxy-DBC are reported in Tables I and 11. Different HPLC programs were tested for the separation of the acetoxy-DBC derivatives and employed for cross comparison of retention times of standards and metabolites. The HPLC retention data of acetoxy-DBC and major acetylated metabolites are shown in Table 111. 1-AcO-DBCalways eluted first and was clearly separated from the rest of the AcO-DBC derivatives. 5-AcO-DBC was well resolved from 6-AcO-DBC. Before acetylation, 1-OH-DBC and 2-OH-DBC overlapped with close retention times; 5-OH-DBC and 6-OH-DBC had identical retention times and could not be separated under different HPLC conditions2 (4). Conversely, the physical and chemical similarity between 2-, 3-, and 4-AcO-DBC resulted in a lack of chromatographic separation of the three compounds. However, without acetylation, a good separation of 2-, 3-, and 4-hydroxy-DBC had been previously reached and reported (3, 4). As reported3 ( 4 ) ,the major metabolite of DBC formed from in vitro microsomal incubations cochromatographed with 5-OH-DBC and 6-OH-DBC. After acetylation of the metabolites, the major fraction (peak 3 in Figure 1A) cochromatographed only with 5-AcO-DBC, and not with 6-AcO-DBC. In addition, the fluorescencespectra of peak 3 were identical to those of 5-AcO-DBC and not 6-Ac0DBC (shown in Figure 2). One of the DBC metabolites (fraction 1 in Figure 1B) had been assigned as a 1-OHDBC-dominated mixture with minor 2-OH-DBC based on its fluorescence spectra3. Acetylation of this fraction resulted in fluorescence spectra and retention time identical to those of 1-AcO-DBC and not 2-AcO-DBC (see Figure 3). After acetylation, one previously unknown fraction in the chromatogram of DBC metabolites (peak X in Figure 1B) eluted behind DBC (fraction 4 in Figure 1A). Its retention time and spectra matched those of 6,6'bis(5-AcO-DBC). The fluorescence spectral data of ac3
W. Xue, J. Schneider, and D. Warshawsky, in preparation. J. Schneider, W. Xue, and D. Warshawsky, unpublished data.
Chem. Res. Toxicol., Vol. 6, No. 3, 1993 347
Acetylation of DBC Metabolites
Table 11. ‘H NMR Spectral Data of AcO-DBC Derivatives chemical shifts (ppm) compound COCH3 1-AcO-DBC 2.17 (5)
H1
H2
a
H3
H4
H5
7.93 (d) a 7.82 (d) =78 J 5 . 6 = 8.8 8.90 (9) 7.24 (d) 7.96 (d) 7.51 (d) 7.61 (d) J 3 , 4 = 8.7 J 5 , 6 = 8.7 9.12 (d) 7.54 (d) C 7.57 (d) c 51.2 = 8.4 J 5 , 6 = 8.7 9.03 (d) 7.51 (t) 7.27 (d) d d J1,2= 8.6 J 2 , 3 = 6.8 9.2 (dd) e 7.54 (t) 8.00 (d) 7.27 ( 8 ) J 1 , z = 8.3 J 3 , 4 = 7.8 9.12 (d) 7.71 (t) 7.56 (t) 8.03 (d) 7.53 (s) J 1 , z = 8.4 J 3 , 4 = 8.0 9.24 (d) 7.84 (t) 7.57 (t) 8.03 (d) 51.2 = 8.5 J 2 , 3 = 7.4 J 3 , 4 = 7.8 a
J3,4
2-AcO-DBC 2.44 ( 8 ) 3-AcO-DBC 2.43 (9) 4-AcO-DBC 2.54 ( 8 ) 6-AcO-DBC 2.55 ( 8 ) fraction 3
2.58 ( 8 )
fraction 4
2.05 ( 8 )
0
H6
H7
H8
H9
H10
8.82 (s) 7.83 (d) a J s , g = 8.8 8.53 (s) 7.83 (d) b J 8 , g = 8.7 8.83 (s) 7.86 (d) c J 8 , g = 8.7 8.83 (s) 7.90 (d) 7.86 (d) J 8 , 9 = 8.9 8.91 (s) 7.90 (d) e J 8 , g = 8.7 8.81 (s) 7.86 (d) 7.66 (d) J s , g = 8.7 8.68 (s) 7.80 (d) 7.47 (d) J 8 , g = 8.8
H11
7.93 (d) a JIOJI = 7.8 8.02 (d) 7.52 (t) JIO,II 8.0 8.03 (d) 7.41 (dd) J i o , n = 7.8 8.03 (d) d J i o , i i = 7.9 8.04 (d) 7.54 (t) JIOJI = 7.9 8.07 (d) 7.57 (t) J i o , i i = 7.5 8.07 (d) 7.66 (t) JIOJI = 8.0
H12
H13
a
8.94 (d) = 8.2 b 9.11 (d) 512.13 = 8.2 c 9.19 (d) J12,13 = 9.0 d 9.07 (d) J12,13 = 8.4 e 9.20 (dd) J12,13 E 8.3 7.66 9.21 (d) J12,13 = 8.4 7.74 9.39 (d) J12,13 = 8.5 J12.13
7.47-7.58 (m). 7.66-7.70 (m). 7.63-7.71 (m). 7.61-7.68 (m). e 7.68-7.76 (m). Table 111. HPLC Separation and Fluorescence Spectral Data of AcO-DBC Derivatives compound
1-AcO-DBC 2-Ac0-DBC 3-AcO-DBC 4-AcO-DBC 5-AcO-DBC 6-AcO-DBC 6,6’-bis(5-AcO)-DBC DBC fraction 3 fraction 4 0
A
HPLC retention time (min) for program: 1 2 3 12.3 10.4 24.8 17.0 14.8 30.2 17.3 15.0 30.9 16.8 14.7 30.9 19.7 17.2 32.9 21.6 19.8 35.5 27.4 25.0 22.2 37.6 24.1 20.0 17.1 27.4 25.1
fluorescence spectra (nm) exa 280,301,350 243,280,302,347 241,279,302,350 242,279,302,350 241,281,302,352 241,277,301,344 282,305,355 241,279,300,348 241,279,302,351 281,305,354
emb 383,402 372,390,411 375,394,412 375,394,412 378,397 369,386,403 401,422 371,389,410 379,397 401,421
sync ( A i = 10 nm)
373 363 367 367 369 360 390 364 369 390
Excitation spectra peak wavelengths. Emission spectra peak wavelengths. Synchronous spectra peaks reported as excitation wavelengths.
r
5-AC0
DBC
6,6-bis(5-Aca-DBC)
W
z m -
$
