Metabolism of Carbon Tetrachloride to Trichloromethyl Radical: An

Jul 16, 1999 - Extensive ESR spin-trapping studies with α-phenyl-N-tert-butylnitrone (PBN) have shown that carbon tetrachloride (CCl4) is metabolized...
17 downloads 4 Views 93KB Size
Download PDF
730

Chem. Res. Toxicol. 1999, 12, 730-736

Metabolism of Carbon Tetrachloride to Trichloromethyl Radical: An ESR and HPLC-EC Study Detcho A. Stoyanovsky and Arthur I. Cederbaum* Department of Biochemistry and Molecular Biology, Mount Sinai School of Medicine, New York, New York 10029 Received March 1, 1999

Extensive ESR spin-trapping studies with R-phenyl-N-tert-butylnitrone (PBN) have shown that carbon tetrachloride (CCl4) is metabolized to trichloromethyl radical (•CCl3). However, the ESR analysis of R-phenyl-N-tert-butylnitrone (PBN)-spin trapped •CCl3 in biological systems appears to be complicated. It has been reported that after in vivo administration of PBN and CCl4 to rats, most of the PBN-CCl3 adduct collected in the bile was ESR silent, suggesting reduction of the nitroxide to its hydroxylamine form. The PBN-CCl3 nitroxide was also shown to undergo a NADPH-dependent reduction in the presence of liver microsomes. Thus, it appears that the variability (or the absence) of the ESR signal of PBN-CCl3 nitroxide in biological systems reflects, at least in part, the fluctuations in the equilibrium between the nitroxide and hydroxylamine forms of this adduct. To test this possibility, ESR and HPLC experiments with electrochemical detection (EC) were conducted for analysis of the major redox form of the PBN-CCl3 adduct in vivo. Standard procedures for the in vitro preparation of both redox forms of PBN-CCl3 and for their HPLC-EC analysis and electrochemical profiles were established. The intensity of the initially observed ESR spectrum of PBN-CCl3 nitroxide of the liver extract from a CCl4- and PBN-treated rat was relatively constant; after an addition of K3[Fe(CN)6] to the extract, the intensity of the ESR spectrum increased by 1 order of magnitude, most likely due to the co-oxidation of ESR silent PBN-derived hydroxylamines. The addition of PBN-CCl3 nitroxide to the liver homogenate resulted in the rapid loss of the ESR signal. The HPLC-EC analysis of the liver extract revealed that the in vivo spin trapping of •CCl3 with PBN leads to a preferential formation of the ESR silent PBN-CCl3 hydroxylamine. The predominant presence of the hydroxylamine derivative was also detected in the blood of a CCl4-treated rat. The results of this work are discussed in terms of combination of the ESR spin trapping and HPLC-EC techniques for the detection of ESR silent radical adducts in biological systems.

Introduction The NADPH-dependent metabolism of CCl4 to •CCl3 was strongly suggested by the study of Fowler, who observed the formation of hexachloroethane (C2Cl6) in a CCl4-containing liver microsomal suspension (1). The initial attempts to apply the ESR spin-trapping technique for the direct detection of •CCl3 were unsuccessful as oxygen- and carbon-centered radicals formed during the liver microsomal metabolism of CCl4 interfered with the ESR assays (2, and references therein). However, with the use of R-phenyl-N-tert-butylnitrone (PBN)1 as a spin trap, the metabolic activation of CCl4 to •CCl3 was demonstrated in both model and in vivo systems (3-5). Despite the numerous reports confirming this metabolic activation, the ESR detection of PBN-spin trapped •CCl3 in biological systems does not appear to be a trivial task. * To whom correspondence should be addressed: Department of Biochemistry, Box 1020, Mount Sinai School of Medicine, One Gustave L. Levy Place, New York, NY 10029. Telephone: (212) 241-7285. Fax: (212) 996-7214. 1 Abbreviations: ESR, electron spin resonance; EC, electrochemical; HPLC, high-performance liquid chromatography; MS, mass spectrometry; GC, gas chromatography; PBN, R-phenyl-N-tert-butylnitrone; DMPO, 5,5′-dimethyl-1-pyroline N-oxide; DMPO/GS•, 2-(S-glutathionyl)-5,5-dimethyl-1-pyrroline nitroxide; HER, 1-hydroxyethyl radical; AU, absorbance units; THF, tetrahydrofuran.

Recently, the NADPH-dependent metabolism of CCl4 in PBN-containing rat liver microsomal suspension has been reinvestigated with the goal of optimizing the ESR analysis of •CCl3 (6). A considerable scatter in the spin trapping experiments was found, and only general trends could be recognized; it was reported that at least eight repeat experiments were needed under identical conditions to obtain an average value with an error of (10% (6). Under similar experimental conditions, we have observed a transient appearance of the ESR spectrum of the PBN-CCl3 nitroxide after a lag period in which the duration was also poorly reproducible (unpublished data).2 The reported intensity of the PBN-CCl3 ESR spectrum of liver extracts from CCl4-treated rats was also variable. Furthermore, it has been shown that in the bile collected from PBN- and CCl4-treated rats, the PBNCCl3 nitroxide is fully reduced to the corresponding ESR silent hydroxylamine (7). More recently, a study from the same laboratory has demonstrated that the PBN-CCl3 nitroxide is efficiently reduced by liver microsomes and NADPH (8). Thus, it appears that the variability (or the absence) of the ESR signal of PBN-CCl3 nitroxide in biological systems reflects, at least in part, the fluctuations in the equilibrium between the nitroxide and 2

D. A. Stoyanovsky and A. I. Cederbaum, unpublished data.

10.1021/tx9900371 CCC: $18.00 © 1999 American Chemical Society Published on Web 07/16/1999

ESR and HPLC-EC Analysis of CCl4 Metabolism

Chem. Res. Toxicol., Vol. 12, No. 8, 1999 731

hydroxylamine forms of this adduct. To test this possibility, ESR and HPLC-EC were carried out for analysis of PBN-CCl3 formation in vivo, with emphasis on the determination of the predominant redox form of this adduct.

Materials and Methods Reagents. All the reagents (analytical grade) were purchased from Sigma Chemical Co. (St. Louis, MO). ESR Measurements. ESR measurements were performed on a Bruker ECS106 spectrometer with 50 kHz magnetic field modulation at room temperature (25 °C). ESR spectrometer settings were as follows: 0.5 G modulation amplitude, 40 s scan time, 20 mW microwave power, and 1 × 105 to 1 × 106 receiver gain. For in vitro spin trapping of •CCl3, PBN was used at a concentration of 0.05 M in tetrahydrofuran (THF). ESR spectrum simulations were made using a program created by P. D. Morse, II, and R. Reiter (EPR Simulation System 2.01, Scientific Software Services). The hyperfine splitting constants (in gauss) used for simulation of the spectrum of the PBN-CCl3 nitroxide were as follows: AN ) 14.1 and AH ) 1.8 (9). Preparation of PBN-CCl3 Nitroxide and Its Hydroxylamine Derivative. PBN-CCl3 adducts were prepared via Fe2+dependent decomposition of a CCl4-triethanolamine complex in the presence of PBN. Fe(NH4)(SO4)2 (0.01 M) was added to a solution of PBN (0.05 M), CCl4 (0.5 M), and triethanolamine (0.5 M) in tetrahydrofuran (1 mL). After incubation for 5 min, the suspension was diluted with H2O (3 mL) and the reaction products were extracted with 3 × 3 mL of n-hexane. The hexane phases were collected, and the crystals formed after rotor evaporation of the solvent (25 °C) were redissolved in either THF or ethanol (75%). A representative ESR spectrum of the PBNCCl3 nitroxide in THF is presented in Figure 1 (AN ) 14.1 G and AH ) 1.8 G). As will be shown in Results, the latter solution contained both the nitroxide and hydroxylamine forms of PBNCCl3; therefore, an interconversion of the adducts was carried out with either potassium ferricyanide (1 mM; oxidation of the hydroxylamine for 60 min at 25 °C) or ascorbic acid (10 mM; reduction of the nitroxide for 30 min at 25 °C). In ethanolic (or THF) solution, both forms of PBN-CCl3 were stable for at least 2 months when kept at -20 °C. Preparation of Hepatic Microsomes. Male SpragueDawley rats weighing 120-150 g were sacrificed, and liver microsomes were prepared as described previously (10), resuspended in 0.125 M KCl/0.01 M potassium phosphate buffer (pH 7.4), and stored at -70 °C. All the solutions that were used during the preparation contained 2 mM dithiothreitol, 0.15 mM desferrioxamine, leupeptin (0.001 mg/mL), and phenylmethanesulfonyl fluoride (0.1 mM). Animals and Animal Treatments. For induction of cytochrome P450 2E1, Sprague-Dawley rats (n ) 6) weighing 120150 g were injected ip with 4-methylpyrazole (200 mg/kg of body weight/day, 3 days). The rats were starved overnight prior to being killed, and experiments with CCl4 were conducted as described in refs 5 and 11. Briefly, the animals were loaded with 2 mL of PBN (200 mg/kg of body weight)-containing olive oil; 1 h later, olive oil alone (control rats; n ) 2) or CCl4 was given ip as a 20% solution in olive oil at a dose of 5 mL of solution/kg of body weight (1 mL of pure CCl4/kg; n ) 4). In a typical experiment, the animal was sacrificed 1 h after the administration of CCl4, its liver was homogenized (in 20 mL of PBS), and the homogenate was subjected to extraction with CHCl3 (40 mL) and methanol (20 mL). The resulting emulsion was centrifuged (15 min at 10 000 rpm), the chloroform phase collected, and the solvent rotor evaporated at room temperature. The residue was resuspended in 5 mL of ethanol (70%) and centrifuged (30 min at 30 000 rpm). The resulting supernatant was used for ESR and HPLC-EC analysis, while the bottom layer (0.3-0.5 mL of fats) was discarded. Blood samples (0.2 mL) were collected via heart puncture and subjected to an extraction with CHCl3 (2 mL), CH3OH (1 mL), and H2O (0.8 mL) as described above.

Figure 1. EPR spectrum of PBN-CCl3 nitroxide formed by the Fe2+-dependent decomposition of CCl4 in the presence of triethanolamine. Trace 1 represents a computer simulation of the ESR spectrum of PBN-CCl3 nitroxide. The reaction was carried out at 25 °C in THF (1 mL) containing PBN (0.05 M), triethanolamine (0.5 M), and CCl4 (0.5 M; trace 2) or PBN, CCl4, and triethanolamine with Fe(NH4)2(SO4) (0.01 M) (trace 3; the ESR spectrum was recorded after a 10-fold dilution with THF of an aliquot taken from the reaction solution); traces 4 and 5 represent the ESR spectra of PBN-CCl3 nitroxide after dilution of 0.01 mL of a PBN-CCl3-containing solution in THF with 0.3 mL of 70% methanol (v/v) and 0.3 mL of PBS, respectively. HPLC-EC Measurements of PBN-CCl3 Adducts. HPLC was performed with a Waters model 510 liquid chromatograph (Milford, MA). Separation was achieved with a C-18 reverse phase column (Nucleosil, 4.6 mm × 25 cm, 5 µm, 100 A, Supelco, Inc., Bellefonte, PA). The mobile phase was saturated with helium and contained 10 mM lithium perchlorate and 70% (v/ v) methanol. All HPLC analyses were conducted at a flow rate of 1 mL/min. Electrochemical detection of PBN-CCl3 adducts was carried out at 1.0 V with a LC-4B amperometric system (Bioanalytical Systems, West Lafayette, IN) equipped with a glassy carbon electrode and a Ag/AgCl reference electrode. A 0.020 mL injection loop was used for all experiments.

Results Preparation, Stability, and HPLC-EC Analysis of PBN-CCl3 Nitroxide. Amines have been shown to form charge-transfer complexes with CCl4 that can undergo an Fe2+-dependent decomposition to C2Cl6, presumably via recombination of •CCl3 (12). To assess the stability of PBN-CCl3 in the presence of reductants, as well as for optimization of its HPLC-EC analysis, experiments were conducted for preparation of the adduct via spin trapping of •CCl3 with PBN in a CCl4- and triethanolaminecontaining solution of THF (Figure 1, ESR spectrum 2). An addition of Fe2+ resulted in the appearance of a sixline ESR spectrum with hyperfine structure with an AN of 14.1 G and an AH of 1.8 G, which allows the assignment of the adduct to be that formed by addition of •CCl3 to PBN (Figure 1, ESR spectrum 3); the experimental and computer-simulated ESR spectra of PBN-CCl3 were in

732 Chem. Res. Toxicol., Vol. 12, No. 8, 1999

Figure 2. ESR-monitored kinetics of reduction of PBN-CCl3 nitroxide in the presence of ascorbate, glutathione, and liver microsomes with NADPH. All experiments were carried out in 0.1 M phosphate buffer (pH 7.4 and 25 °C). The nitroxide was prepared and redissolved in THF as described in Materials and Methods. The final concentration of THF in the reaction solutions did not exceed 0.05% (v/v). The arrows denote the points of the corresponding additions: (9) GSH (2 mM), (O) ascorbate (3 mM), (b) liver microsomes (1 mg of protein/mL), and NADPH (1 mM), and K3[Fe(CN)6] (1 mM).

good agreement (trace 3 vs trace 1). No formation of PBN-CCl3 or other nitroxides was observed if triethanolamine (or CCl4) was omitted from the reaction solution (data not shown). The solvent effects (please see below) on the ESR spectrum of the PBN-CCl3 nitroxide are presented in traces 4 (70% methanol; AN ) 14.7 G and AH ) 2.3 G) and 5 (PBS; AN ) 15.3 G and AH ) 2.7 G). The PBN-CCl3 nitroxide was relatively stable in the absence of reductants. However, upon addition of ascorbic acid or NADPH and liver microsomes, but not GSH or NADPH in the absence of microsomes, the ESR spectrum of the nitroxide rapidly disappeared, most likely due to its reduction to the corresponding hydroxylamine (Figure 2). In support of the latter assumption, K3[Fe(CN)6], which is known to efficiently oxidize hydroxylamines to nitroxides, quantitatively restored the ESR spectrum of the PBN-CCl3 nitroxide after incubation of the nitroxide with a NADPH-containing microsomal emulsion (Figure 2). In contrast to our results showing that the PBN-CCl3 nitroxide undergoes an ascorbate-dependent reduction, Janzen et al. (13) found that this compound is stable in the presence of ascorbate; reasons for these differences are not clear but may reflect varying concentrations and reaction conditions. The preferential redox shift of PBN-CCl3 to its hydroxylamine form in the presence of cellular reductants suggests that its ESR detection in biological systems may not always be efficient. Alternatively, both redox forms of PBN-CCl3 can be quantified by HPLC-EC analysis. Figure 3 depicts the HPLC-EC profile of a solution of PBN-CCl3 nitroxide that was prepared as described in Materials and Methods. The fraction defined by peak 1 exhibited an ESR spectrum that is typical for the PBNCCl3 nitroxide in 70% methanol (peak 1 of Figure 3 vs peak 4 of Figure 1; AN ) 14.7 G and AH ) 2.3 G). The fraction collected under peak 2 did not exert any ESR activity (Figure 3, ESR spectrum 2). An addition of

Stoyanovsky and Cederbaum

Figure 3. HPLC-EC chromatograms of a PBN-CCl3-containing solution. The HPLC-EC profile of a PBN-CCl3-containing solution prepared as described in Materials and Methods and diluted 10-fold in 70% methanol is presented with solid lines (peaks 1 and 2). An addition of ascorbate (3 mM) and a subsequent incubation of 15 min resulted in the decrease of peak 1, and a parallel increase of peak 2 (dashed lines) was observed. The fractions defined by HPLC peaks 1 and 2 were collected and subjected to ESR analysis (inset). The numbering of ESR spectra 1 and 2 corresponds to the numbering of HPLC peaks 1 and 2. Trace 3 is derived from the fraction under peak 2, which was then treated with potassium ferricyanide (1 mM). Details of the chromatographic separation are given in Materials and Methods.

potassium ferricyanide (1 mM) to this fraction, however, resulted in the appearance of the ESR spectrum of the PBN-CCl3 nitroxide (Figure 3, ESR spectrum 3; AN ) 14.7 G and AH ) 2.3 G), suggesting that peak 2 reflects the elution of the hydroxylamine form of the adduct. A brief pretreatment of the PBN-CCl3 solution with either potassium ferricyanide or ascorbate resulted in an interconversion of peak 2 and peak 1 (oxidation of the hydroxylamine form; data not shown), and vice versa (reduction of the nitroxide form; Figure 3). Full reduction of the nitroxide by ascorbate occurs with increasing incubation times. We have not studied the exact mechanism of PBN-CCl3 hydroxylamine formation in the CCl4-ethanolamine system; it is likely, however, that the PBN-CCl3 nitroxide that was formed upon addition of Fe2+ undergoes a subsequent Fe2+-dependent reduction to hydroxylamine. A direct interaction between Fe2+ and the nitroxide is not likely to occur as the redox potential of the Fe3+/Fe2+ couple is 0.77 V; the redox potentials of a series of nitroxides have been reported to be within the range of 0.5-0.9 V (14). The autoxidation of Fe2+, however, would lead to the formation of superoxide anion radical that may directly reduce the PBN-CCl3 nitroxide (O2/O2•- ) -0.115 V). Both PBN-CCl3 nitroxide and hydroxylamine were subsequently used as HPLC-EC standards. The voltage dependencies (I/V curves) of the HPLCEC signals for both redox forms of PBN-CCl3 suggest that the nitroxide is more efficiently oxidized on the glassy carbon electrode (Figure 4). These curves may be used for discrimination of the CCl3 adducts from peaks with similar retention times. ESR and HPLC-EC Analysis of the in Vivo Metabolism of CCl4 to •CCl3 Radical. Figure 5 depicts the ESR spectrum of an ethanolic extract from the liver of a control rat (trace 1; receiver gain of 1 × 106) and a CCl4-

ESR and HPLC-EC Analysis of CCl4 Metabolism

Figure 4. Voltage dependence of HPLC-EC signals of the nitroxide and hydroxylamine forms of PBN-CCl3. Both nitroxide (b) and hydroxylamine (O) forms of PBN-CCl3 were subjected to HPLC-EC analysis at different holding potentials. The nitroxide and its hydroxylamine derivative were prepared as described in Materials and Methods and diluted 10-fold in 70% methanol prior to the HPLC separation.

Figure 5. Effects of potassium ferricyanide on the ESR spectrum of a liver extract from a CCl4-treated rat. The experiments were carried out in ethanol at 25 °C. Liver extracts and standard PBN-CCl3 nitroxide were prepared as described in Materials and Methods. ESR spectra of a liver extract from control rat (trace 1) and CCl4-treated rat (trace 2; receiver gain of 1 × 106). Trace 3 represents the ESR-monitored kinetics of accumulation of nitroxides after an addition of K3[Fe(CN)6] (1 mM) to the liver extract from the CCl4-treated rat. The time interval between two consecutive ESR scannings was 20 min (receiver gain of 1 × 105). The ESR spectrum at time zero is the same as in trace 2.

treated rat (trace 2; receiver gain of 1 × 106). The hyperfine structure (in guass) of the ESR spectrum presented in trace 2 is specific for the PBN-CCl3 nitroxide (in 70% ethanol, AN ) 14.15 G and AH ) 1.85 G). The intensity of the ESR spectrum was constant for a time period of 60 min. After an addition of K3[Fe(CN)6] (1 mM)

Chem. Res. Toxicol., Vol. 12, No. 8, 1999 733

Figure 6. ESR-monitored kinetics of reduction of PBN-CCl3 nitroxide in the presence of liver homogenate. The experiments were carried out in 0.1 M phosphate buffer (pH 7.4 and 25 °C) in the presence of 0.2 mM desferrioxamine. An aliquot (0.01 mL) of a standard solution of PBN-CCl3 nitroxide in THF was added to 0.2 mL of liver homogenate (2 g of liver/1 mL of PBS), and ESR spectra were recorded every 2 min (traces 1-3). The appearance of the doublet signal of the semidehydroascorbyl radical is denoted with asterisks (trace 3). Standard PBN-CCl3 nitroxide was prepared as described in Materials and Methods.

to the liver extract from the CCl4-treated rat, the intensity of the observed ESR spectrum increased as a function of time over a 60 min time period (trace 3; receiver gain of 1 × 105; the ESR spectrum at time zero is the same as that in trace 2). We could not evaluate the real amount of the PBN-CCl3 hydroxylamine which was present as the ferricyanide treatment of the liver extract led to a shift of the hyperfine structure (in gauss) of the initial ESR spectrum to an AN of 14.25 G and an AH of 2.72 (Figure 5, ESR spectrum 3). The latter shift was most likely due to a co-oxidation of multiple PBNderived hydroxylamines that were present in the liver extract. In a detailed analysis of the liver extract from CCl4-treated rats, Janzen et al. have identified PBNderived nitroxides of •CCl3 and unidentified carbon- and oxygen-centered radicals (15). No changes in the control ESR spectrum (trace 1) were observed upon addition of K3[Fe(CN)6] (data not shown). When an aliquot of a pre- or in vivo-formed PBN-CCl3 nitroxide solution was added into a liver homogenate, the ESR spectrum of the nitroxide rapidly disappeared, most likely due to its reduction to the corresponding hydroxylamine (Figure 6). The reduction of the nitroxide by liver homogenate was paralleled by the appearance of a weak doublet signal of the one-electron oxidation product of endogenous ascorbate, the semidehydroascorbyl radical (Figure 6, ESR spectrum 3); the hepatic concentration of ascorbate is within the millimolar range (16). A similar reduction of the PBN-CCl3 nitroxide was observed in the presence of 3 mM ascorbate (Figure 2). In addition to ascorbate, the hepatic reduction of the nitroxide could also be supported by thiol-containing proteins and NADHand NADPH-linked dehydrogenases. The data presented in Figure 6 suggest that the reduction potential of liver is sufficient to shift the nitroxide-hydroxylamine equilibrium almost entirely in favor of the ESR silent hydroxylamine derivative. It may even be possible that the nitroxide form of PBN-CCl3 (Figure 5, ESR spectrum 2) could have accumulated ex vivo via oxidation of the

734 Chem. Res. Toxicol., Vol. 12, No. 8, 1999

Stoyanovsky and Cederbaum

nitroxide (data not shown). In both liver and blood extracts, the hydroxylamine derivative of PBN-CCl3, compared to its nitroxide form, was the preponderant adduct.

Discussion

Figure 7. HPLC-EC analysis of liver and blood extracts from a CCl4-treated rat. Liver and blood extracts were prepared as described in Materials and Methods. The HPLC-EC profile of liver extract is presented in panel A. The fractions under peaks 1 and 2 were collected and analyzed by ESR; the numbering of ESR spectra 1 and 2 corresponds to the numbering of HPLC peaks 1 and 2 (inset in panel A). ESR spectrum 3 is derived from the fraction under peak 2, which was then treated with potassium ferricyanide (1 mM). The HPLC-EC profile of a blood extract is presented in panel B. Details for the chromatographic separation are given in Materials and Methods. Results are representative of one of four experiments; the variability did not exceed 10%.

hydroxylamine derivative during the extraction procedures. For a direct quantification of the spin-trapped •CCl3 radical in vivo, HPLC-EC separation of the nitroxide and hydroxylamine derivatives of PBN-CCl3 was attempted. Figure 7 depicts the HPLC-EC profile of an ethanolic liver extract from a CCl4-treated rat. The fraction defined by peak 1 exhibited a typical PBN-CCl3 nitroxide ESR spectrum (Figure 7A, ESR spectrum 1 vs ESR spectrum 4 in Figure 1; AN ) 14.7 G and AH ) 2.3 G). The fraction collected under peak 2 did not exert any ESR activity (Figure 7A, ESR spectrum 2), until potassium ferricyanide (1 mM) was added (Figure 7A, ESR spectrum 3; AN ) 14.7 G and AH ) 2.3 G). Figure 7B illustrates the HPLC-EC profile of an extract from a 0.2 mL blood sample; no ESR spectrum was observed when the blood extract was subjected to ESR analysis. HPLC-EC analysis indicated the presence of the PBN-CCl3 hydroxylamine (peak 2) with just a small amount of the PBN-CCl3 nitroxide present (peak 1). The identity of the PBN-CCl3 hydroxylamine was confirmed via its ferricyanide-dependent conversion into PBN-CCl3

The results from this study demonstrate that the in vivo spin trapping of •CCl3 with PBN leads to a preferential formation of the ESR silent PBN-CCl3 hydroxylamine. The presence of PBN-CCl3 hydroxylamine in liver and blood of CCl4-treated rats was demonstrated by HPLC-EC analysis. The metabolic activation of CCl4 to •CCl3 is a wellestablished reaction. Recently, Janzen et al. have applied GC-MS and HPLC-UV/EC analysis to PBN adducts extracted from CCl4- and NADPH-containing liver microsomal suspensions, and from liver of CCl4- and PBNtreated rats (13, 15). The presence of PBN-CCl3 hydroxylamine in the extracts was indeed suggested; however, a positive identification of this adduct, as well as evaluation of its amount relative to the ESR active PBN-CCl3 nitroxide, was not presented. Reinke and Janzen have reported that in the blood of a CCl4-treated rat, the doublet signal of the one-electron oxidation product of endogenous ascorbate, the semidehydroascorbyl radical, could be observed (17); after removal of the endogenous ascorbic acid from a blood sample, the ESR spectrum of PBN-CO2 (but not PBN-CCl3) nitroxide was observed. However, when blood plasma was subjected to extraction, the ESR spectrum of the PBN-CCl3 nitroxide was documented in the extract, suggesting that immobilization of the nitroxide may occur in the presence of plasma proteins (17). In support of the latter mechanism, bovine serum albumin was shown to affect the ESR spectrum of a preformed PBN-CCl3 nitroxide. The data presented in Figure 7B of a HPLC-EC profile of a blood extract from CCl4-treated rat suggest that reduction of the nitroxide form of PBN-CCl3 to the corresponding ESR silent hydroxylamine may also contribute to the lack of a PBNCCl3 ESR signal in the analyzed blood samples. The observed preponderance of the PBN-CCl3 hydroxylamine in CCl4- and PBN-treated rats in our experiments supports the notion that the reduction of radicalderived ESR active nitroxides in spin trapping experiments is an unavoidable occurrence in most in vivo systems (18):

The reduction of nitroxides by cellular reductants is a reaction that has been routinely used as an experimental tool, e.g., studies of flip-flop rates of spin-labeled lipids (19), distribution of spin-labels in sarcoplasmic reticulum (20), and optimization of the ESR imaging technique (2125). It has been reported that the redox potential of the intracellular milieu is in the range of -0.26 to 0.28 V (26). The redox potentials of a series of nitroxides have been reported to be in the range of 0.5-0.9 V (14), suggesting that in spin trapping experiments in biological systems a major reaction pathway can be the reduction of the nitroxides to their ESR “silent” hydroxylamine derivatives. Some authors have addressed the question

ESR and HPLC-EC Analysis of CCl4 Metabolism

of how this process affects the efficacy of the ESR spin trapping technique (7, 27, 28). ESR spin trapping studies in model systems has been most successful in leading to the formation of massive spectral libraries. For example, the Internet accessible ESR database of NIEHS contains hyperfine splitting constants for more than 10 000 spintrapped adducts (18), which, in combination with computer programs for ESR spectra simulation, make virtually certain the structural identification of any physiologically relevant radical intermediate. The use of this technique in in vivo experiments and with intact cells, however, has been less successful. Reports of the reoxidation of radical-derived hydroxylamine adducts formed in biological systems back to their ESR active nitroxide forms are not frequently presented (7, 27, 28). It is likely that the presence of reductants that support the reverse reaction, and/or the presence of multiple reduced radical adducts whose oxidation leads to the appearance of unidentifiable ESR spectra (Figure 5A spectrum 3), are major obstacles for the successful ESR analysis of reoxidized hydroxylamine adducts. Alternatively, the analysis of spin-trapped radical species can be carried out via GC/MS (15, 29). The latter method allows the separation of the spin adducts and deduction of their initial structure by analyzing the MSobservable molecular fragments. The sensitivity of the GC/MS technique, however, is slightly lower than that of ESR (29); the sensitivity of the traditional ESR spectrometry is in the micromolar range. It appears that a more direct and sensitive method for detection of spintrapped adducts is their HPLC-EC analysis. There are well-established HPLC-EC protocols for the separation and detection of a series of radical-derived nitroxides (30, 31). Recently, we have described the HPLC-EC detection of the hydroxylamine forms of 1-hydroxyethyl- and •CH3derived POBN and PBN adducts formed in biological systems (32, 33). When separated by reverse-phase HPLC, the hydroxylamine derivatives have, compared to their nitroxide forms, a slightly increased retention time on a C-18 matrix (Figure 3; 32, 33); they are also oxidized less efficiently on a glassy carbon electrode (Figure 4; 33). The lowest limit of EC detection for both nitroxide and hydroxylamine forms of spin-trapped adducts is within the nanomolar range (32). However, the sensitivity of the HPLC-EC method can be further increased if detection of the analytes is carried out with a Coulochem detector. The latter method is based on the use of electrodes made of porous graphite that have a larger contact area than the glassy carbon electrodes; this results in a 100% reaction with an electroactive analyte, as compared to 5-10% with the glassy carbon electrode. The use of a Choulochem electrode may provide information about spin-trapped adducts within the picomolar range. The HPLC separation of biological samples would also facilitate the discrimination of nitroxides with similar ESR spectra (e.g., DMPO-GS• and DMPO-•OH, PBN-HER and PBN-CH3C(•)dO, etc.). The HPLC peaks can be positively identified via their retention time, co-injection of preformed standards (Figure 3; 29, 31, 32), redox interconversion (Figure 3), ESR (Figures 3 and 6) and UV spectra (32, 33), and I/V curves (Figure 4; 30, 33). The use of the HPLC-EC detection of spin-trapped radical species is expected to be especially useful in studies of radical species produced in systems enriched with cellular reductants.

Chem. Res. Toxicol., Vol. 12, No. 8, 1999 735

Acknowledgment. This work was supported by U.S. Public Health Service Grant AA-09460 from The National Institute on Alcohol Abuse and Alcoholism.

References (1) Fowler, J. S. (1969) Carbon tetrachloride metabolism in the rabbit. Br. J. Pharmacol. 37, 733-737. (2) Rosen, G. M., and Rauckman, E. J. (1982) Carbon tetrachlorideinduced lipid peroxidation: a spin trapping study. Toxicol. Lett. 10, 337-344. (3) Poyer, J. L., Floyd, R. A., McCay, P. B., Janzen, E. G., and Davis, E. R. (1978) Spin-trapping of the trichloromethyl radical produced during enzymic NADPH oxidation in the presence of carbon tetrachloride or bromotrichloromethane. Biochim. Biophys. Acta 539, 402-409. (4) Kalyanaraman, B., Mason, R. P., Perez-Reyes, E., Chignell, C. F., Wolf, C. R., and Philpot, R. M. (1979) Characterization of the free radical formed in aerobic microsomal incubations containing carbon tetrachloride and NADPH. Biochem. Biophys. Res. Commun. 89, 1065-1072. (5) Lai, E. K., McCay, P. B., Noguchi, T., and Fong, K. L. (1979) In vivo spin-trapping of trichloromethyl radicals formed from CCl4. Biochem. Pharmacol. 28, 2231-2235. (6) Janzen, E. G., Poyer, J. L., West, M. S., Crossley, C., and McCay, P. B. (1994) Study of reproducibility of spin trapping results in the use of C-phenyl-N-tert-butyl nitrone (PBN) for trichloromethyl radical detection in CCl4 metabolism by rat liver microsomal dispersions. Biological spin trapping I. J. Biochem. Biophys. Methods 29, 189-205. (7) Sentjurc, M., and Mason, R. P. (1992) Inhibition of radical adduct reduction and reoxidation of the corresponding hydroxylamines in in vivo spin trapping of carbon tetrachloride-derived radicals. Free Radical Biol. Med. 13, 151-160. (8) Connor, H. D., Thurman, R. G., Chen, G., Poyer, J. L., Janzen, E. G., and Mason, R. P. (1998) Clarification of the relationship between free radical spin trapping and carbon tetrachloride metabolism in microsomal systems. Free Radical Biol. Med. 24, 1364-1368. (9) Buettner, G. R. (1987) Spin trapping: ESR parameters of spin adducts. Free Radical Biol. Med. 3, 259-303. (10) Palakodety, R. B., Clejan, L. A., Krikun, G., Feierman, D. E., and Cederbaum, A. I. (1988) Characterization and identification of a pyrazole-inducible form of cytochrome P-450. J. Biol. Chem. 263, 878-884. (11) Harisch, G., and Meyer, W. (1985) Studies on tissue distribution of glutathione and on activities of glutathione-related enzymes after carbon tetrachloride-induced liver injury. Res. Commun. Chem. Pathol. Pharmacol. 47, 399-414. (12) Lautenberger, W. J., Jones, E. V., and Miller, J. G. (1968) Reaction of amines with haloalkanes. J. Am. Chem. Soc. 90, 1110-1115. (13) Janzen, E. G., Chen, G., Bray, T. M., Reinke, L. A., Poyer, J. L., and McCay, P. B. (1993) Study of the isolation and stability of R-trichloromethylbenzyl(tert-butyl)aminoxyl, the trichloromethyl radical adduct of R-phenyl-tert-butylnitrone (PBN). J. Chem. Soc., Perkin Trans. 2, 1983-1989. (14) Morris, S., Sosnovsky, G., Hui, B., Huber, C. O., Rao, N. U., and Swartz, H. M. (1991) Chemical and electrochemical reduction rates of cyclic nitroxides (nitroxyls). J. Pharm. Sci. 80, 149-152. (15) Janzen, E. G., Towner, R. A., Krygsman, P. H., Lai, E. K., Poyer, J. L., Brueggemann, G., and McCay, P. B. (1990) Mass spectroscopy and chromatography of the trichloromethyl radical adduct of phenyl tert-butyl nitrone. Free Radical Res. Commun. 9, 353360. (16) Banhegyi, G., Braun, L., Csala, M., Puskas, F., and Mandl, J. (1997) Ascorbate metabolism and its regulation in animals. Free Radical Biol. Med. 23, 793-803. (17) Reinke, L. A., and Janzen, E. G. (1991) Detection of spin adducts in blood after administration of carbon tetrachloride to rats. Chem.-Biol. Interact. 78, 155-165. (18) DeGray, J. A., and Mason, R. P. (1994) Biological spin trapping in Electron Spin Resonance (Atherton, N. M., Davies, M. J., and Gilbert, B. C., Eds.) Vol. 14, pp 246-301, The Royal Society of Chemistry, Cambridge, U.K. (19) Kornberg, R. D., and McConnell, H. M. (1971) Inside-outside transitions of phospholipids in vesicle membranes. Biochemistry 10, 1111-1120. (20) Tonomura, Y., and Morales, M. F. (1974) Change in state of spin labels bound to sarcoplasmic reticulum with change in enzymic state, as deduced from ascorbate-quenching studies. Proc. Natl. Acad. Sci. U.S.A. 71, 3687-3691.

736 Chem. Res. Toxicol., Vol. 12, No. 8, 1999 (21) Iannone, A., Hu, H. P., Tomasi, A., Vannini, V., and Swartz, H. M. (1989) Metabolism of aqueous soluble nitroxides in hepatocytes: effects of cell integrity, oxygen, and structure of nitroxides. Biochim. Biophys. Acta 991, 90-96. (22) Iannone, A., Bini, A., Swartz, H. M., Tomasi, A., and Vannini, V. (1989) Metabolism in rat liver microsomes of the nitroxide spin probe tempol. Biochem. Pharmacol. 38, 2581-2586. (23) Iannone, A., Tomasi, A., Vannini, V., and Swartz, H. M. (1990) Metabolism of nitroxide spin labels in subcellular fraction of rat liver. I. Reduction by microsomes. Biochim. Biophys. Acta 1034, 285-289. (24) Iannone, A., Tomasi, A., Vannini, V., and Swartz, H. M. (1990) Metabolism of nitroxide spin labels in subcellular fractions of rat liver. II. Reduction in the cytosol. Biochim. Biophys. Acta 1034, 290-293. (25) Couet, W. R., Brasch, R. C., Sosnovsky, G., and Tozer, T. N. (1985) Factors affecting nitroxide reduction in ascorbate solution and tissue homogenates. Magn. Reson. Imaging 3, 83-88. (26) Gilbert, H. F. (1990) Molecular and cellular aspects of thioldisulfide exchange. Adv. Enzymol. Relat. Areas Mol. Biol. 63, 69172. (27) Bernofsky, C., and Bandara, B. M. (1995) Spin trapping endogenous radicals in MC-1010 cells: evidence for hydroxyl radical and carbon-centered ascorbyl radical adducts. Mol. Cell. Biochem. 148, 155-164.

Stoyanovsky and Cederbaum (28) Nakagawa, K., Ishida, S., Yokoyama, H., Mori, N., Niwa, S., and Tsuchihashi, N. (1994) Rapid free radical reduction in the perfused rat liver. Free Radical Res. 21, 169-176. (29) Sang, H., Janzen, E. G., Poyer, J. L., and McCay, P. B. (1997) The structure of free radical metabolites detected by EPR spin trapping and mass spectroscopy from halocarbons in rat liver microsomes. Free Radical Biol. Med. 22, 843-852. (30) Towell, J., and Kalyanaraman, B. (1991) Detection of radical adducts of 5,5-dimethyl-1-pyrroline n-oxide by the combined use of high-performance liquid chromatography with electrochemical detection and electron spin resonance. Anal. Biochem. 196, 111119. (31) Pritsos, C. A., Constantinides, P. P., Tritton, T. R., Heimbrook, D. C., and Sartorelli, A. C. (1985) Use of high-performance liquid chromatography to detect hydroxyl and superoxide radicals generated from mitomycin C. Anal. Biochem. 150, 294-299. (32) Stoyanovsky, D. A., and Cederbaum, A. I. (1998) ESR and HPLCEC analysis of ethanol oxidation to 1-hydroxyethyl radical: rapid reduction and quantification of POBN and PBN nitroxides. Free Radical Biol. Med. 25, 536-545. (33) Stoyanovsky, D. A., Melnikov, Z., and Cederbaum, A. I. (1999) ESR and HPLC-EC analysis of the interaction of hydroxyl radical with DMSO: rapid reduction and quantification of POBN and PBN nitroxides. Anal. Chem. 71, 715-721.

TX9900371