356
Chem. Res. Toxicol. 1988,1, 356-363
Nucleic Acid Binding of Arylamines during the Respiratory Burst of Human Granulocytes Michael D. Corbett*ltp* and Bernadette R. Corbettt Food Science and Human Nutrition Department and Department of Pharmacology and Therapeutics, University of Florida, Gainesville, Florida 32611 Received July 29, 1988
Following stimulation with phorbol myristate acetate, human granulocytes were found to incorporate a series of arylamines into cellular nucleic acid. No such binding occurred if the granulocytes were not induced to undergo the respiratory burst. The relative amount of covalent binding to cellular DNA and RNA was found to depend strongly on the chemical structure of the arylamine. 2-Aminofluorene gave the highest ratio of DNA/RNA binding, while 4-nitroaniline showed a very low ratio of DNA/RNA binding. 4-Nitroaniline may bind only to RNA, since the degree of binding to DNA was a t the level of detectability. Two other substrates, 4chloroaniline and 4-methylaniline, gave intermediary ratios of DNA/RNA binding. Studies on the possible role of the granulocyte enzyme myeloperoxidase in the activation and binding of these arylamines were conducted in vitro and also through the use of azide, an inhibitor of myeloperoxidase activity in cells. The results indicate that myeloperoxidase probably plays only a limited role in causing the covalent binding of arylamines to nucleic acid in human granulocytes. It is probable that other reactive oxygen species, which are not dependent upon myeloperoxidase for their production, are necessary for the bioactivation of some arylamines, especially for substrates such as 4-nitroaniline. A free-radical mechanism for arylamine bioactivation, and its potential role in arylamine toxicity, was presented in the context of the current scientific literature.
The role of metabolic N-oxidation in the bioactivation of aromatic amines and amides was discovered nearly 30 years ago (1,2). Since then, numerous studies have reported on the ability of cytochrome P-450 dependent oxygenases to carry out C- and N-hydroxylation reactions. I t is now generally believed that such N-hydroxylation reactions are an absolute requirement in order for an arylamine or aryl amide to display genotoxicity (3-5). The products of these 2e- oxidation reactions are arylhydroxylamines and hydroxamic acids, which generally require a subsequent conjugation reaction in order to form reactive metabolites capable of covalent binding to DNA or other macromolecules (4, 6). The bulk of these cytochrome P-450 dependent N-oxidation reactions undoubtedly occur in the liver; however, other organs, particularly lung, also appear to be proficient in such reactions (7). The concept of extrahepatic bioactivation of xenobiotics is a relatively recent development in the field of xenobiochemistry and has greatly broadened the approach now taken in attempts to elucidate possible mechanisms of chemical toxicity (8). The discovery of the ability of prostaglandin synthase to effect cometabolism of certain xenobiotics (9) contributed much to this concept of extrahepatic bioactivation. The ability of peroxidative enzymes, including prostaglandin synthase, to effect the bioactivation of arylamines is now widely recognized (10-15). The fungal enzyme, chloroperoxidase,was found by us to be particularly efficient in its ability to effect N-oxidation of arylamines to arylhydroxylamines and nitroso aromatic metabolites (16-18) most likely via a 2eprocess. Another peroxidase was discovered in pea seed-
* Address correspondence to this author at the University of Florida, IFAS-0163, Gainesville, FL 32611-0163. Food Science and Human Nutrition Department. Department of Pharmacology and Therapeutics.
*
0893-228x/88/2701-0356$01.50/0
lings with a similar ability to effect the 2e- oxidation reactions on arylamines (19). On the other hand, most biomolecules with peroxidative activity have been reported to oxidize arylamines via le- processes. In contrast to 2eoxidation reactions, the products of le- peroxidative reactions are much more complex in structure, which is consistent with the initial production of radical species from arylamines (14, 15, 17, 20). Our interest in peroxidative reactions of arylamines led us to an investigation of the fate of such chemicals in certain phagocytic cells. Polymorphonuclear leukocytes (granulocytes) and cells of the monocyte/macrophage series generally possess the ability to produce reactive oxygen species as the result of the "respiratory burst" (21). Such reactive oxygen species, coupled with the presence of myeloperoxidases (MP0)l in many such cell types, make the activated phagocyte a particularly interesting system in which to investigate the potential bioactivation of arylamines. We now report on the relative covalent binding of a series of four arylamines to the nucleic acids of human granulocytes as a consequence of the respiratory burst.
Experlmental Section Radiolabeled Substrates. [9-'*C]-Z-Aminofluorene was obtained from Chemsyn Science Laboratories (Lenexa, KS). [rir~g-U-'~C]-4-Chloroaniline, [rir~g-U-'~C]-4-methylaniline, and [rir~g-U-'~C]-4-nitroaniline were obtained from Pathfinder Laboratories (St. Louis, MO). After any necessary dilution with unlabeled chemical, samples weighing from 2-5 mg each were purified by chromatography on silica gel 60 (E. Merck, 70-230 mesh) columns (1.1 X 16-18 cm bed volume) by gravity elution Abbreviations: MPO, myeloperoxidase; PMA, phorboll2-myristate 13-acetate;SDS, sodium lauryl sulfate; SOD, superoxidedismutase; 2-AF, 2-aminofluorene;4-C1, 4-chloroaniline;4-CH8, 4-methylaniline; 4-N02, 4-nitroaniline. 0 1988 American Chemical Society
Chem. Res. Toxicol., Vol. 1, No. 6, 1988 357
Binding of Arylamines during the Respiratory Burst with CH2C12. Aliquots of 5 mL each were collected and analyzed by HPLC employing a C18reversed-phase column (3.9 mm X 30 cm) with 50% aqueous methanol as solvent, except for 2aminofluorene in which case 60% aqueous methanol was employed as the solvent. Appropriate fractions were combined and evaporated, and then the residues were dissolved in sufficient 95% ethanol such that each substrate solution was 10 mM. Each substrate had a radiochemical purity exceeding 99.5% as determined by HPLC analysis and liquid scintillation counting. Specific activities were determined to be as follows: 2-aminofluorene, 15.2 mCi/mmol; 4-chloroaniline, 5.8 mCi/mmol; 4methylaniline, 10.5 mCi/mmol; 4-nitroaniline, 9.7 mCi/mmol. Preparation of Human Granulocytes. The granulocyte fraction was isolated from whole human blood by using a modification of a previously described method (22). Forty milliliters of freshly drawn and heparinized whole blood was placed in a 60-mL plastic syringe containing 20 mL of 3% (w/v) Dextran T-500 in 0.9% (w/v) NaCl and 0.1% (w/v) glucose. The syringe was fitted with a plastic plunger, inverted five times to achieve thorough mixing, and then allowed to stand inverted for 45 min at room temperature. The straw-colored upper layer was extruded by means of the plunger into a 50-mL Oak Ridge polycarbonate (PC) centrifuge tube and centrifuged at 200g for 20 min at 4 "C. To lyse residual erythrocytes, 9 mL of ice-cold water was added to the pellet and the suspension mixed by vortexing for 30 s and then immediately treated with 1 mL of 9% NaCl to restore isotonicity. The sample was then centrifuged a t 200g for 5 min a t 4 "C. This procedure was repeated if any erythrocytes remained. The pellet was resuspended in 20 mL of Hank's balanced salt solution (HBSS), which was Ca2+and Mg2+free, and then layered onto 10 mL of cold (4 "C) FicolllO77 (Sigma Chemical Co.) in a 50-mL Oak Ridge PC centrifuge tube and centrifuged at 250g for 20 min at 4 OC. The pellet was resuspended in 10 mL of HBSS (Ca2+and Mg2+free) and centrifuged at 250g for 20 min. The pellet was resuspended in 20 mL of complete HBSS, and the cells were counted in a hemocytometer using Turks solution. The cells were then diluted with complete HBSS to give suspensions containing (1.8-2.0) X 106 cells/mL, which were used immediately.
Incubation of Arylamine Substrates with Granulocyte Suspensions. For studies in which both DNA and RNA were isolated from granulocytes, each incubation volume was 40 mL of granulocyte suspension [ (1.8-2.0) X lo6 cells/mL] contained in a 250-mL PC flask at 37 OC. Agitation at 180 rpm was achieved by use of a gyrotory incubator. The [14C]arylaminesubstrate as a 10 mM solution in 95% ethanol was added by means of a microliter syringe to give the desired concentration, which was 10 pM for most experiments. To induce the respiratory burst, PMA as a solution of 0.1 mg/mL in DMSO was added to give a final concentration of 0.1 pg/mL. In certain studies, various modulators were added as concentrated aqueom solutions to give the desired concentration and generally allowed to preincubate for 5 min prior to the addition of PMA. A 200-pL aliquot was removed and used to monitor for the production of superoxide anion by the cytochrome c reduction method (22). The incubation was generally terminated after 30 min by rapidly chilling the incubation and centrifugation a t 600g for 10 min at 4 "C. The pellet was resuspended in 10 mL of HBSS and then centrifuged to obtain the washed granulocyte pellet. The supernatant was analyzed by HPLC to determine the amount of substrate that remained and to identify any obvious metabolites. Isolation of DNA and RNA from Granulocytes. The pellet from the incubation was resuspended in 2 mL of 10 mM EDTA, 100 mM NaC1, and 20 mM Tris buffer, pH 8, which contained 2% SDS and 0.6 mg/mL proteinase K (Sigma type XI), and the cells were lysed by incubation for 2-4 h at 51 "C. After the cells had lysed, as evidenced by the absence of turbidity, the solution was extracted twice for 30 min each with 2 mL of a phenol reagent that consisted of 0.1% (w/v) of 8-hydroxyquinoline in phenol/ CHC13/isoamyl alcohol (25:24:1) that had been saturated with 0.1 M Tris.HC1, pH 8.0, containing 0.2% dithiothreitol. The upper aqueous layer was extracted with 2 mL of 4% (v/v) isoamyl alcohol in CHCl, and then with 2 mL of H20-saturated diethyl ether. After removal of the residual ether with a stream of N2, the aqueous solution was transferred to a 10-mL Oak Ridge PC centrifuge tube and treated with 0.1 mL of 3 M NaCl followed by an equal volume of ice-cold 2-ethoxyethanol to precipitate
DNA. After standing a t -20 "C for 30 min, the supernatant was decanted from the DNA precipitate into a 50-mL Oak Ridge PC centrifuge tube, treated with 2 volumes of ice-cold absolute ethanol, and then allowed to stand overnight to precipitate RNA, which was collected by centrifugation a t 9000g for 60 min at 4 "C. Both precipitates were gently washed with 1mL of ice-cold 80% ethanol and then dried with a stream of N2 The DNA was dissolved in 2 mL of 1 mM EDTA in 10 mM Tris-HC1, pH 7.8, and then treated with 200 pg of RNase A (Sigma type XII-A) and 2800 units of RNase T1 (Bethesda Research Laboratories) for 1 h at 37 "C. The RNA was dissolved in 1 mL of 0.1 M sodium acetate, pH 5, containing 5 mM MgClz and then treated with 8.7 units of DNase I ("RNase free", BRL) for 1h at 25 "C. Following the nuclease treatments, each sample was incubated with 20 pg/mL proteinase K for 1 h at 51 OC and then extracted twice for 5 min each with an equal volume of the phenol reagent, and then with CHC1, and with H20-saturated diethyl ether, and then the residual ether was removed with a stream of N2 The nucleic acid solutions were treated with 50 and 100 pL of 3 M NaCl and 3 and 5 mL of ice-cold absolute ethanol for RNA and DNA, respectively, and stored overnight at -20 "C. Following centrifugation, the pellets were washed with 1mL of 95% ethanol, dried with Nz, and then dissolved in H 2 0 (2.0 mL for DNA and 1.0 mL for RNA samples). UV absorbance at 260 and 280 nm was determined, and the amount of nucleic acid present was determined from Am (23). The amount of 14Clabel present was determined by liquid scintillation counting of 0.1- and 0.5-mL aliquots of RNA and DNA, respectively, in Scintiverse Bio-HP (Fisher Scientific). The potential contamination of the nucleic acid samples with protein was monitored by use of the protein-dye binding assay (24).
Enzymatic Hydrolysis of DNA. Fractions of DNA from replicate incubations were combined, treated with aqueous NaCl to make 0.15 mM, and precipitated by the addition of 2 volumes of cold ethanol. After collection and drying with a N2 stream, the DNA pellets were dissolved in 10 mM Tris.HC1, pH 7.4, which contained 0.1 mM EDTA and 10 mM MgC12, to give final concentrations of about 1mg of DNA/mL. Portions (1.0 mL) of the DNA solutions were treated with 214 units of DNase I (Sigma type 11) for 6 h at 37 "C, and then 1mL of 1 M Tris buffer, pH 9.0, was added to adjust the pH to about 8.8. Snake venom phosphodiesterase I (Sigma type VII) was added to give a concentration of 0.05 unit/mL, and the solution was incubated for 6 h at 37 "C, followed by treatment with 8.3 units of alkaline phosphatase (Sigma type VII-S) for each milliliter of solution, and incubation for an additional 12-15 h. The DNA hydrolysis solution was extracted twice with equal volumes of H20-saturated diethyl ether and then twice with equal volumes of H20-saturated 1-butanol. The aqueous fraction was made 100 mM in tetrabutylammonium phosphate and then reextracted with equal volumes of H20-saturated 1-butanol. The organic extracts were washed once with 0.12 volume of H 2 0 and centrifuged to separate the phases. LSC was employed to determine the amount of radiolabel present in the organic extracts and in the remaining aqueous portion. Myeloperoxidase-Catalyzed Binding of Arylamine Substrates to DNA. Incubations were carried out at 37 "C by placing 10 mL of 0.05 M KH2POl buffer, pH 7.0, containing 0.1 M NaCl and 1mg/mL of denatured and purified calf thymus DNA (25) in a 25-mL Erlenmeyer flask along with the "C-labeled substrate (10 pM) and 2 pg/mL human myeloperoxidase (Calbiochem,San Diego, CA). A 2-mL aliquot was taken for a T = 0 time point, and then the reaction was initiated by the addition of 40 pL of a 10 mM aqueous solution of H20z. Additional aliquots of 2 mL each were taken at appropriate times and quenched by extracting with 4 mL of ethyl acetate, which had been precooled to -20 "C. The aqueous layer was extracted with 4 mL of H20-saturated diethyl ether, then twice with 2 mL of phenol reagent for 5 min each, and then once each with 2 mL of CHC13 and 2 mL of H,O-saturated diethyl ether. After removal of residual ether with a stream of Nz, the DNA was precipitated by the addition of 5 mL of cold ethanol and collected by centrifugation. After drying with N2, the amount of '"C bound to DNA was determined as described above. Following concentration of the ethyl acetate extract, the amount of starting material that remained was determined by HPLC analysis.
358 Chem. Res. Toxicol., Vol. 1, No. 6, 1988 Table I. Covalent Binding of [14C]Arylaminesto Granulocyte Nucleic Acid covalent binding in nucleic acid grouped expts, nmol of binding re1 [14C]arylamine/mg of comDound’ nucleic acid f SDb to 2-AFc 2-AF 1.55 f 0.15 1.00 4-CH3 0.74 f 0.18 0.48 0.41 4-C1 0.64 f 0.17 4-NO2 0.24 f 0.11 0.15
Corbett and Corbett Table 11. Dependence of Nucleic Acid Binding on Incubation Time and Substrate Concentration covalent binding, nmol expt concn, incubation of [14C]arylamine/mg no. substrate pM time, min” of nucleic acidb 1 2-AF 10 0 0.00 2-AF 10 5 0.24 2-AF 10 10 0.67 2-AF 10 30 1.35 2-AF 5 30 0.61 2 4-CH3 5 30 0.62 4-CH3 10 0 0.00 4-CH3 10 10 0.70 4-CH3 10 30 1.10 3 &NO2 10 0 0.00 4-NO2 10 10 0.21 4-NO2 10 30 0.68
“Ten micromolar incubation with 2 X lo8 cells/mL for 30 min following the addition of PMA. Four independent experiments were conducted at approximately weekly intervals employing human blood obtained from the same volunteer. Each experiment was grouped since it utilized all four of the arylamine substrates with identical cell preparations. Similar relative results were obtained when granulocytes from other volunteers were employed; however, the absolute amounts of nucleic acid binding were observed to vary severalfold between volunteers.
Following treatment with PMA. Results from single grouped experiments.
Results For each experiment, which necessarily employed a fresh granulocyte preparation, the production of 02*by an aliquot of the cell suspension was monitored by the continuous method for cytochrome c reduction (22). The rate of production of 0;- seldom varied more than 20% from an average value, even though cells were obtained from several sources. Preliminary studies demonstrated that, at concentrations of 10 pM or less, none of the four arylamine substrates caused any inhibition of 0;- production. Initial experiments were conducted to determine the binding of the 14C-labeledarylamines to total nucleic acids of human granulocytes. None of these arylamines gave any detectable covalent binding in the absence of a stimulus for the respiratory burst for incubation periods up to 1h. However, in the presence of PMA, we observed the covalent binding of all four arylamine substrates to granulocyte nucleic acid (Table I). The variability in the data ranged from 10 to 45% depending upon the substrate. Most of this variability is thought to be due to differences in the nature of the individual granulocyte preparations, even though they were obtained from the same human volunteer. In the extreme case of 4-N02,slight variations in the relative amount of RNA present in the nucleic acid isolate would be expected to contribute considerably to the variability of the data (vide infra). The range in the amounts of arylamines bound to nucleic acid was found to be less than 1 order of magnitude, even though the electronic nature of the aromatic rings of the model arylamines that were employed varied over a wide range. The binding to granulocyte nucleic acid following stimulation with PMA was also found to increase with time and substrate concentration (Table 11). In these experiments (Tables I and 11), nucleic acid was isolated by a simplification of the method described under Experimental Section for DNA and RNA isolation. This involved the precipitation of “total” nucleic acid with ethanol following the removal of protein by repetitive phenol extraction. Analysis of the nucleic acid showed it to be free of protein (
