Metabolism of acetaldehyde to methyl and acetyl radicals: in vitro and in vivo electron paramagnetic resonance spin-trapping studies

Free Radical Biology & Medicine, Vol. 29, No. 8, pp. 721–729, 2000 Copyright © 2000 Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5849/00/$–see front matter

PII S0891-5849(00)00374-9

Original Contribution METABOLISM OF ACETALDEHYDE TO METHYL AND ACETYL RADICALS: IN VITRO AND IN VIVO ELECTRON PARAMAGNETIC RESONANCE SPIN-TRAPPING STUDIES LIA S. NAKAO,* MARIA B. KADIISKA,† RONALD P. MASON,† MERCEDES T. GRIJALBA,*

and

OHARA AUGUSTO*

*Departamento de Bioquı´mica, Instituto de Quı´mica, Universidade de Sa˜o Paulo, Sa˜o Paulo, Brazil; and †Laboratory of Pharmacology and Chemistry, National Institute of Environmental Health Sciences, Research Triangle Park, NC, USA (Received 19 April 2000; Revised 13 June 2000; Accepted 22 June 2000)

Abstract—Acetaldehyde oxidation by enzymes and cellular fractions has been previously shown to produce radicals that have been characterized as superoxide anion, hydroxyl, and acetyl radicals. Here, we report that acetaldehyde metabolism by xanthine oxidase, submitochondrial particles and whole rats produces both the acetyl and the methyl radical, although only the latter was unambiguously identified in vivo. Electron paramagnetic resonance (EPR) characterization of both radicals was possible by the use of two spin traps, 5,5-dimethyl 1-pyrroline N-oxide (DMPO) and ␣-(4-pyridyl 1-oxide)-N-t-butylnitrone (POBN), and of acetaldehyde labeled with 13C. The POBN-acetyl radical adduct proved to be unstable, but POBN was employed to monitor acetaldehyde metabolism by Sprague-Dawley rats because previous studies have shown its usefulness for in vivo spin trapping. EPR analysis of the bile collected from treated and control rats showed the presence of the POBN-methyl and of an unidentified, biomolecule-derived, POBN adduct. Because decarbonylation of the acetyl radical is one of the routes for methyl radical formation from acetaldehyde, detection of the latter in bile provides strong evidence for the production of both radicals in vivo. The results may be relevant to understanding the toxic effects of acetaldehyde itself and of its more relevant biological precursor, ethanol. © 2000 Elsevier Science Inc. Keywords—Acetaldehyde metabolism, Ethanol metabolism, Methyl radical, Acetyl radical, Xanthine oxidase, Submitochondrial particles, Spin trap, Free radicals

INTRODUCTION

tract, and 90% of it is metabolized in the liver. Its primary metabolite is acetaldehyde, which is produced by the action of alcohol dehydrogenase in hepatocytes. Further oxidation of the aldehyde by aldehyde dehydrogenase produces acetate [8]. This route is considered to be a detoxification pathway because acetate is an innocuous compound. However, there are other metabolic routes available to ethanol and acetaldehyde. For instance, ethanol oxidation by hepatic microsomal P450 enzymes becomes more pronounced after chronic alcohol consumption or after a high acute dose of the compound [9]. Such a route has been associated with the production of free radicals [10]. Relevantly, free radicals such as the superoxide anion [11,12] and the 1-hydroxyethyl radical have been demonstrated to be produced during ethanol metabolism, both in vitro and in vivo [13–19]. Similarly, oxidation of acetaldehyde to acetate may be accompanied by the generation of free radicals, as has been demonstrated to occur during its oxidation by some chemical [20,21] and enzymatic systems [22–26].

The toxicological importance of acetaldehyde is derived from its ubiquitous presence in the environment, the continuous exposure of humans resulting from occupational and lifestyle factors, and the reactivity of the compound [1–3]. As a potent electrophile, acetaldehyde readily reacts with nucleophiles in proteins, phospholipids, and nucleic acids to produce adducts, some of which have been detected in alcoholic patients [4 –7]. Indeed, a particularly important source of human endogenous exposure to acetaldehyde is the ingestion of alcoholic beverages, and the aldehyde is considered to mediate some of the adverse effects associated with chronic alcohol consumption [4 –7]. Ethanol is rapidly absorbed in the gastrointestinal Address correspondence to: Dr. Ohara Augusto, Instituto de Quı´mica, Universidade de Sa˜o Paulo, Caixa Postal 26077, 05513-970, Sa˜o Paulo, SP, Brazil; Tel: 55 (11) 3 818-3873; Fax: 55 (11) 3 818-2186 and 55 (11) 3 815-5579; E-Mail: [email protected]. 721

722

L. S. NAKAO et al.

Enzymes and cellular fractions that have been shown to oxidize acetaldehyde to free radical intermediates include xanthine oxidase [22,23], aldehyde oxidase [24], mitochondria [25], and microsomes [26]. In most of these studies, free radical formation was inferred from indirect measurements, and acetaldehyde-derived free radicals were not identified. Exceptions were the spin-trapping studies indicating acetyl radical production from acetaldehyde oxidation by xanthine oxidase [23] and rat liver and brain microsomes [26]. These systems are known producers of active oxygen species, such as the superoxide anion radical and hydrogen peroxide, that, in the presence of transition metal ions, generate the hydroxyl radical. Consequently, it has been assumed that formation of the acetyl radical arises from acetaldehyde attack by the hydroxyl radical (Eqn. 1) [27], a mechanism that predicts the parallel formation of methyl radicals from acetyl radical decarbonylation (Eqn. 2) [28,29].

spectively. Other chemicals were from Merck (Darmstadt, Germany). DMPO was purified by distillation, and its concentration was determined by ultraviolet spectroscopy in ethanol (⑀234 ⫽ 7700 M⫺1cm⫺1). Xanthine oxidase and all other chemicals used in the enzymatic experiments were treated overnight with Chelex-100. Enzyme activity was determined by measuring the formation of uric acid at 290 nm (⑀ ⫽ 12,200 M⫺1cm⫺1) from 0.1 mM xanthine in phosphate buffer, pH 7.6.

CH 3 CHO ⫹ • OH 3 CH 3 C • O ⫹ H 2 O

EPR spin-trapping experiments in vitro

k ⫽ 3.6 ⫻ 10 9 M ⫺1 s ⫺1 CH 3 C • O 3 • CH 3 ⫹ CO

(1) (2)

Diverse values have been reported for the rate constant of acetyl radical decomposition (Eqn. 2) (k ⫽ 10⫺2⫺1 s⫺1; see [29]), but whatever the time scale of the process, it allowed the EPR-spin–trapping detection of both the acetyl and methyl radicals during the oxidation of acetaldehyde by hydrogen peroxide-iron (II)-EDTA [20]. In contrast, the methyl radical has never been detected during the oxidation of acetaldehyde by enzymatic systems. The importance of free radical mechanisms in the toxic effects of ethanol is becoming increasingly recognized (reviewed in [30,31]). Because acetaldehyde is an important ethanol metabolite, it is relevant to fully characterize its free radical metabolites. We report here that both acetyl and methyl radicals are produced during acetaldehyde metabolism by xanthine oxidase, submitochondrial particles, and rats. MATERIALS AND METHODS

Chemicals DMPO, POBN, Chelex-100, xanthine oxidase from buttermilk (grade III), allopurinol, antimycin A, and EDTA were obtained from Sigma Chemical Co. (St. Louis, MO, USA). Desferrioxamine was obtained from Ciba-Geigy (Summit, NJ, USA). Acetaldehyde and labeled acetaldehyde (1,2-13C2, 99%) were purchased from Aldrich Chemical Co. (Milwaukee, WI, USA) and Cambridge Isotopes Laboratories (Andover, MA, USA), re-

SMPS preparation Beef heart mitochondria were isolated according to Vercesi et al. [32]. The SMPS were obtained from frozen mitochondria by sonication, as described elsewhere [33]. Protein concentration was determined by the Bradford method. SMPS were kept frozen at ⫺70°C.

EPR spectra were recorded at room temperature using a Bruker EMX spectrometer. Xanthine oxidase reaction mixtures were incubated in a shaker at 36°C for 15 min, and the EPR spectra were recorded at room temperature within 2 min after incubation. SMPS particles were reactivated by thawing at 31°C for 10 min. EPR spectra were recorded within 1 min after the addition of the last reagent. The concentration of radical adducts was estimated by double integration of the EPR spectra using known concentrations of 4-hydroxy-2,2,6,6-tetramethyl1-piperidinyloxy radical as a standard. Computer simulation analyses of some spectra were performed by using a program written by Duling [34]. EPR spin-trapping experiments in vivo Sprague-Dawley male rats (300 – 450 g; Charles River Breeding Laboratories, Raleigh, NC, USA) fed a standard chow mix (NIH open formula, Ziegler Brothers, Inc., Gardner, PA, USA) were used in all experiments. Nonfasted rats were anesthetized (with Nembutal at 50 mg per kilogram of body weight) and bile ducts were cannulated with a segment of PE 10 tubing (Becton Dickinson, Sparks, MD, USA). All animals were given an ip injection of the spin trap POBN (1 g/kg) dissolved in water. Experimental groups were given an intragastric injection of acetaldehyde (1 g/kg) as an aqueous solution (1:1, v/v). Bile samples (⬃400 ␮l) were collected every 20 min for 100 min into plastic Eppendorf tubes containing a 50 ␮l solution of 2,2⬘-dipyridyl (30 mM) and bathocuproine-disulfonic acid, disodium salt hydrate (30 mM). Bile samples were frozen on dry ice immediately after collection and were stored at ⫺70°C. The samples

Free radicals from acetaldehyde

723

were thawed immediately before EPR analysis. Direct EPR analysis of the bile fractions collected from the rats treated with POBN and acetaldehyde presented a pronounced spectrum of the ascorbyl radical and traces of POBN radical adducts, possibly indicating that most of the produced adducts have been reduced to their corresponding hydroxylamines [35,36]. Consequently, as soon as they were thawed, most of the bile fractions were treated with 1 mM ferricyanide for 1 h [35,36] before scanning the EPR spectra. The spectra were recorded at room temperature with a Bruker EMX instrument equipped with a Super High Q cavity. RESULTS

Oxidation by xanthine oxidase/Fe(III)-EDTA Xanthine oxidase is a molybdo-flavoenzyme that may play a role in acetaldehyde metabolism because its conversion from the dehydrogenase to the oxidase form is increased during ethanol intoxication [37]. Acetaldehyde oxidation by xanthine oxidase has been demonstrated to produce superoxide anion, hydrogen peroxide, and acetyl radicals [11,22,23]. The latter species was identified with EPR spin-trap studies using POBN that produces adducts that are difficult to characterize by their EPR parameters [38]. In addition, the low acetaldehyde concentrations employed (up to 2 mM) produced low radical yields, a fact that also may compromise radical identification. Here, we show that incubations of 80 mM acetaldehyde with 6 mU/ml of xanthine oxidase and 5 ␮M iron(III)EDTA in the presence of the spin trap DMPO led to an EPR composite spectrum of three radical adducts (Fig. 1A). These adducts were identified by computer simulation of the experimental spectrum as the DMPO/•CH3 (aN ⫽ 1.62mT and aH ⫽ 2.32 mT), DMPO/•COCH3 (aN ⫽ 1.52 mT and aH ⫽ 1.87 mT), and DMPO/•OH (aN ⫽ 1.49 mT and aH ⫽ 1.49 mT) radical adducts [20]. Control experiments showed that radical production was completely dependent on the presence of both substrate (Fig. 1B) and enzyme (Fig. 1C). Accordingly, the EPR signal was barely detectable in the presence of allopurinol, a known inhibitor of xanthine oxidase (Fig. 1D). In the absence of added iron(III)-EDTA, the signal intensity was reduced to about 60% of the original signal (data not shown). Even in the absence of added transition metal ion complex, and in the presence of 1 mM desferrioxamine, the EPR signal was not completely abolished (Fig. 1E). These results indicate that transition metal ions are required for free radical formation and that the enzyme is likely to be contaminated with traces of metal ions that are not removed by CHELEX pretreatment (see Experimental Procedures) or by incubation with desferrioxamine (Fig. 1E) [39].

Fig. 1. EPR spectra of DMPO radical adducts obtained during incubation of acetaldehyde with xanthine oxidase. The spectra were obtained after incubation of 80 mM acetaldehyde with 6 U/ml xanthine oxidase, 5 ␮M iron III-EDTA, and 80 mM DMPO for 15 min at 36°C in phosphate buffer, pH 7.8. (A) Complete system; (B) same as A, in the absence of acetaldehyde; (C) same as (A), in the absence of xanthine oxidase; (D) same as (A), in the presence of 1.5 mM allopurinol; (E) same as (A), in the absence of Fe(III)-EDTA and in the presence of 1 mM desferrioxamine. The composite spectrum of panel (A) is labeled to show its components: (●) DMPO/•CH3, (E) DMPO/•COCH3, and (䊐) DMPO/•OH radical adducts. Instrumental conditions: microwave power, 20 mW; time constant, 327 ms; scan rate, 0.0298 mT/s; modulation amplitude, 0.1 mT; gain, 1⫻ 106.

The above data appeared to be consistent with an earlier proposed mechanism whereby the aldehyde is both the enzyme substrate for the formation of superoxide anion and hydrogen peroxide and the target of these species [11,22,23,40]. In the presence of transition metal ions, they produce the hydroxyl radical which attacks acetaldehyde to produce the acetyl radical that decomposes to the methyl radical (Eqns. 1 and 2). As expected from simple competition for the hydroxyl radical between the aldehyde (Eqn. 2) and DMPO (Eqn. 3) [41], the yields of the DMPO/•OH radical adduct decreased,

724

L. S. NAKAO et al.

DMPO concentration, making it unlikely that all methyl radicals are produced from acetyl radical decarbonylation (Eqn. 2). Taken together, these results suggest that the Fenton mechanism (Eqns. 1 and 2) is not the only route to produce acetaldehyde-derived radicals in the xanthine oxidase system. DMPO ⫹ • OH 3 DMPO/ • OH k ⫽ 3.4 ⫻ 10 9 M ⫺1 s ⫺1

(3)

Oxidation by SMPS

Fig. 2. Effects of acetaldehyde (A) and DMPO (B) concentrations on the yields of DMPO radical adducts produced during the oxidation of acetaldehyde by xanthine oxidase. Experimental and instrumental conditions were the same as those of Fig. 1 except for acetaldehyde (A) and DMPO (B) concentrations, which were varied as specified in the figure. Radical adduct concentrations correspond as follows: (●) DMPO/•CH3, (E) DMPO/•COCH3 and (䊐) DMPO/•OH.

whereas the methyl and acetyl radical adduct yields increased with acetaldehyde concentration up to saturation (Fig. 2A). At high aldehyde concentration, however, the yields of the DMPO adducts did not respond to DMPO concentration as predicted from a simple Fenton mechanism (Fig. 2B). For instance, at 80 mM aldehyde, the yield of hydroxyl radicals is expected to be maximum (Fig. 2A), with about half of it trapped by acetaldehyde and the other half by 80 mM DMPO, as predicted from the nearly identical rate constants of these reactions (Eqns. 1 and 3). Above 80 mM, DMPO is expected to trap more hydroxyl radical than the aldehyde, and high enough concentrations of the spin trap should be able to inhibit the production of the acetaldehyde-derived radicals. It has been previously reported that hydroxyl radical scavengers such as mannitol and benzoate were potent inhibitors of acetyl radical formation [22]. This trend, however, was not observed under our experimental conditions (Fig. 2B). In addition, the yields of both DMPO/ • CH3 and DMPO/•COCH3 adducts increased with

Acetaldehyde-dependent chemiluminescence in mitochondria and SMPS incubations has been used as evidence for the formation of unidentified free radicals [25]. Here, we show that incubation of 100 mM acetaldehyde with succinate-supplemented SMPS in the presence of 0.3 mM antimycin A and 50 mM DMPO produced an EPR spectrum (Fig. 3A) very similar to those obtained during acetaldehyde oxidation by xanthine oxidase (Fig. 1). Simulation of the experimental spectrum (Fig. 3B) confirmed that it is a composite of the spectra of the DMPO/•CH3, DMPO/•COCH3, and DMPO/•OH radical adducts [20]. Decreasing the acetaldehyde to 50 mM decreased the EPR signal intensity to 70% of the original signal. In the absence of antimycin A, no EPR signal was observed (Fig. 3C). In the absence of acetaldehyde, only the DMPO/•OH radical adduct was detectable, but it decayed rapidly (Fig. 3D), probably because of its reduction to the corresponding EPR-silent hydroxylamine. These results demonstrate that beef heart SMPS are able to metabolize acetaldehyde to methyl and acetyl radicals in a process that depends on the leakage of electrons from the electron transport chain because the presence of antimycin A is essential for the detection of acetaldehyde-derived radicals (Fig. 3). Antimycin A is an inhibitor of complex III, which prevents the normal flow of electrons into cytochrome oxidase, thereby increasing the production of superoxide anion and hydrogen peroxide by mitochondria and SMPS [42,43]. Characterization of thePOBN–acetyl radical adduct DMPO, which was important to demonstrate the formation of both the acetyl and methyl radicals from acetaldehyde in vitro (Figs. 1–3), has been of limited usefulness in in vivo experiments [44]. In contrast, POBN, whose adducts are difficult to discriminate by EPR spectroscopy, has been particularly useful for the detection of free radical metabolites in vivo [17,45– 47]. This success is probably because of the high doses of POBN that can be administered to rats and because of the stability of POBN radical adducts [17,45– 47]. Because previous studies with POBN

Free radicals from acetaldehyde

Fig. 3. EPR spectra of DMPO radical adducts obtained during incubation of acetaldehyde with SMPS. The spectra were obtained after 1 min incubation of 100 mM acetaldehyde with 12 mg/ml beef heart SMPS, 10 mM succinate, 0.3 mM antimycin A, and 50 mM DMPO in phosphate-buffered saline, pH 7.4. (A) Complete system; (B) computer simulation of (A); (C) the same as (A), in the absence of antimycin A; (D) the same as (A), in the absence of acetaldehyde. Frozen SMPS were allowed to thaw for 10 min at 31°C. The spectra were obtained 1 min after the addition of the last reagent. The composite spectrum of panel A is labeled to show its components: (●) DMPO/•CH3, (E) DMPO/•COCH3, and (䊐) DMPO/•OH radical adducts. Instrumental conditions: microwave power, 31 mW; time constant, 655 ms; scan rate, 0.015 mT/s; modulation amplitude, 0.15 mT; gain, 1.42 ⫻ 106.

and acetaldehyde oxidation reported only one adduct, identified as POBN/•COCH3 [23,26], we first analyzed the radicals trapped by POBN in vitro during the oxidation of acetaldehyde by hydrogen-peroxide iron(II) that was previously shown to produce DMPO/•OH, DMPO/•COCH3, and DMPO/•CH3 radical adducts [20]. Incubation of 80 mM acetaldehyde with 5 mM hydrogen peroxide and 0.1 mM iron(II)-EDTA at pH 5.0 in the presence of POBN led to the detection of an EPR spectrum that is characteristic of at least two radical adducts with similar hyperfine splitting

725

Fig. 4. EPR spectra of POBN radical adducts obtained during incubation of acetaldehyde with hydrogen peroxide and iron(II)-EDTA. The spectra were obtained after the specified time of incubation of 100 mM acetaldehyde with 5 mM hydrogen peroxide, 0.5 mM iron(II)-EDTA and 100 mM POBN in acetate buffer, pH 5.0. (A) 3 min incubation; (B) computer simulation of (A); and (C) 20 min incubation. The composite spectrum of panel A is labeled to show two of its components: (●) POBN/•CH3 and (E) POBN/•COCH3 radical adducts. Instrumental conditions: microwave power, 20 mW; time constant, 164 ms; scan rate, 0.095 mT/s; modulation amplitude, 0.05 mT; gain, 3.99 ⫻ 104.

constants (Fig. 4A). Computer simulation of the experimental spectrum (Fig. 4B) was consistent with three radical adducts, the POBN/•COCH3 (aN ⫽ 1.50 mT and aH ⫽ 0.30 mT), POBN/•CH3 (aN ⫽ 1.60 mT, and aH ⫽ 0.27 mT), and POBN/•OH (aN ⫽ 1.49 mT and aH ⫽ 0.19 mT) [38] radical adducts in relative yields of 57, 37, and 6%, respectively. The EPR spectrum changed with time, and after a 30 min incubation, it was dominated by the POBN/•CH3 radical adduct (74%; Fig. 4C). Similar experiments were run with commercially available acetaldehyde labeled with 13C at both carbon atoms (Fig. 5). As expected, additional lines resulting from 13C (I⫽1/2) were evident. Again, at initial incubation times, at least two labeled radical adducts were observed (Fig. 5A). Computer simulation of the experimental spectrum (Fig. 5B) was consistent with the presence of three radical adducts, POBN/•13CO13CH3 (aN ⫽ 1.49 mT, aH⫽ 0.29 mT, and a13C ⫽ 0.58 mT), POBN/•13CH3 (aN ⫽

726

L. S. NAKAO et al.

Fig. 5. EPR spectra of POBN radical adducts obtained during incubation of acetaldehyde labeled with 13C with hydrogen peroxide and iron (II)-EDTA. The spectra were obtained after the specified time of incubation of 100 mM acetaldehyde with 5 mM hydrogen peroxide; 0.5 mM iron(II)-EDTA; and 100 mM POBN in acetate buffer, pH 5.0. (A) 7 min incubation; (B) computer simulation of (A); and (C) 20 min incubation. Two of the POBN/•13CO13CH3 radical adduct peaks that do not overlap with the POBN/•13CH3 radical adduct peaks were labeled (E) in panels (A) and (C) to show their time-dependent decay. Instrumental conditions: microwave power, 20 mW; time constant, 164 ms; scan rate, 0.095 mT/s; modulation amplitude, 0.1 mT; gain, 3.99 ⫻ 104.

1.60 mT, aH ⫽ 0.28 mT, and a13C ⫽ 0.50 mT) and POBN/ • OH (aN ⫽ 1.50 mT and aH ⫽ 0.18mT) in relative yields of 38, 35, and 27%, respectively. This simulation characterized the EPR parameters of the POBN/•COCH3 radical adduct for the first time. Again, this adduct was not stable, and at 20 min of incubation, the EPR spectrum was dominated by the methyl radical adduct (63%; Fig. 5C). The results show that although the POBN/•COCH3 radical adduct was of limited stability, the spin trap could be useful to study acetaldehyde metabolism in vivo. In vivo metabolism Bile fractions collected from rats administered acetaldehyde and POBN presented six-line EPR signals (Fig. 6B) that were 3.5 to 5 times more intense than those present in the corresponding bile fractions of control animals (Fig.

Fig. 6. EPR spectra of POBN radical adducts present in the bile of rats treated with acetaldehyde (1 g/kg) and POBN (1 g/kg). The bile fractions were those collected from 60 to 80 min as described in Materials and Methods. (A) POBN-treated rats; (B) acetaldehyde plus POBN–treated rats; (C) POBN-treated rats plus addition of 30 mM acetaldehyde to bile and 1 h incubation; (D) the same as (C) after the addition of 1 mM ferricyanide and a further 1 h incubation. Instrumental conditions: microwave power, 20 mW; time constant, 1310 ms; scan rate, 0.012 mT/s; modulation amplitude, 0.1 mT; gain, 2⫻ 105.

6A). The EPR signals shown in Fig. 6 correspond to bile fractions collected from 60 – 80 min after treatment, but fractions collected at earlier times presented the same, although less intense, signals (data not shown). The EPR spectra of the bile fractions from control animals did not increase significantly after the addition of 30 mM acetaldehyde and 1 h of incubation (Fig. 6C) nor after a further 1 h of incubation after the addition of 1 mM ferricyanide (Fig. 6D). As noted in Materials and Methods, the EPR spectra shown (Figs. 6A and 6B) were obtained after the addition of 1 mM ferricyanide to the bile samples because this treatment increased the six-line signal resulting from POBN adducts in the bile fractions of treated animals. This indicates that the adducts were either excreted as their corresponding reduced hydroxylamines or were reduced to them by the reductants present in the bile [35,36]. However, the experiments in Figs. 6A and 6B, together with the ex vivo

Free radicals from acetaldehyde

727

•

COCH3 (Figs. 4 and 5) will preclude its EPR detection in vivo. Nevertheless, because decarbonylation of the acetyl radical is one of the possible routes for methyl radical formation from acetaldehyde (Eqns. 1 and 2) [25,26], detection of the latter in the bile of treated rats (Fig. 7) provides strong evidence for the production of both radicals during acetaldehyde metabolism in vivo. Also, the results demonstrate that the aldehyde promotes metabolic imbalances that produce a biomolecule-derived radical adduct or adducts (Fig. 7). The yield of the unidentified POBN radical adduct(s) was always higher in the bile from treated rats (Fig. 6). The presence of biomolecule-derived adducts in the bile of animals treated with drugs and transition metal ions has been reported before [45– 47]. The nature of these adducts remains to be established, although lipid-derived radicals such as pentyl or ethyl radical have been tentatively identified in some cases [47]. DISCUSSION

Fig. 7. EPR spectra of POBN radical adducts present in the bile of rats 80 min after treatment with acetaldehyde labeled with 13C (1 g/kg) and POBN (1 g/kg). (A) Experimental spectrum; (B) computer simulation of (A) obtained by addition of the spectral components present in (C) and (D); (C) computer simulation of the unidentified POBN/carboncentered radical adduct; and (D) computer simulation of the POBN/ •13 CH3 radical adduct. Instrumental conditions: microwave power, 20 mW; time constant, 1310 ms; scan rate, 0.012 mT/s; modulation amplitude, 0.1 mT; gain, 2⫻ 105.

controls (Figs. 6C and 6D), establish that most of the POBN radical adducts detected in the bile of treated rats are due to the metabolism of acetaldehyde in vivo. However, because all the EPR spectra present similar hyperfine splitting constants (Fig. 6), it is impossible to distinguish the adducts resulting from acetaldehyde itself from those resulting from biomolecules eventually oxidized by metabolic imbalances caused by the aldehyde [48]. To approach this problem, we replaced normal 12C acetaldehyde with the compound labeled with 13C at both carbon atoms. In this case, the EPR spectrum of the collected bile showed the extra splitting expected from 13 C (I⫽1/2) (Fig. 7). Computer simulation of the experimental spectrum gave the best fit by assuming two radical adducts, POBN/•13CH3 (aN ⫽ 1.58 mT, aH ⫽ 0.28 mT, and a13C ⫽ 0.46 mT) and an unlabeled and unidentified POBN carbon-centered radical adduct (aN ⫽ 1.56 mT and aH ⫽ 0.27 mT) in relative yields of 15 and 85%, respectively (Figs. 7B–7D). This spectrum is difficult to fully analyze. The instability of the POBN/

Our results indicate that acetaldehyde is metabolized to both the acetyl and the methyl radical by enzymatic systems (Figs. 1–3) and whole rats (Figs. 6 and 7), although in the latter case, only the methyl radical was unambiguously identified. Most likely, the instability of the POBN/•COCH3 (Figs. 4 and 5) precluded its EPR detection in vivo. Identification of both radicals was possible by exploring the diverse properties of the spin traps, DMPO and POBN, the former being more useful to differentiate among different adducts (Figs. 1–3) and the latter being more useful to perform in vivo experiments (Figs. 6 and 7). In addition, the use of acetaldehyde labeled with 13C was important to distinguish between radicals derived from acetaldehyde itself and from biomolecule oxidation from metabolic imbalances caused by the aldehyde in vivo (Figs. 6 and 7), [48]. Although the biomolecule-derived radicals produced in vivo from acetaldehyde (this work) and other treatments [45– 47] remain to be identified, their formation indicates the occurrence of free radical chain reactions triggered by xenobiotic-derived radicals. Considering all the possible reactions of these primary radicals, it is indeed remarkable that they and their products can be trapped in whole animals. This only argues the importance of in vivo spin-trapping studies to the understanding of the biological effects of toxic compounds [44 – 47]. The results described here cannot provide unambiguous information about the enzymes that metabolize acetaldehyde to free radicals in vivo nor about the mechanisms involved. In vitro, xanthine oxidase (Figs. 1 and 2) and SMPS (Fig. 3) were able to oxidize acetaldehyde to produce acetyl and methyl radicals. Both systems are likely to be relevant to acetaldehyde metabolism in vivo, particularly

728

L. S. NAKAO et al.

Fig. 8. Schematic representation of some possible routes for the formation of methyl radicals from acetaldehyde. Route (a), hypothetical escape of acetyl radicals from enzymatic oxidation of acetaldehyde; route (b), Fenton mechanism [11,22,23]; route (c), nucleophilic addition of a peroxo anion XOO⫺ to acetaldehyde. XOO⫺ corresponds to ONOO⫺ and HOO⫺ for (1) and (2), respectively [20,21].

during ethanol or acetaldehyde intoxication [37]. Moreover, xanthine oxidase is related to aldehyde oxidase, which is considered to metabolize acetaldehyde in vivo [24]. Acetaldehyde oxidation to acetate by xanthine oxidase is paralleled by the production of superoxide anion and hydrogen peroxide [11]. However, a Fenton mechanism (Fig. 8, route b) is unlikely to be the only route for acetaldehyde-derived free radical production because the yields of the DMPO/•CH3 and DMPO/•COCH3 were not inhibited by increasing concentrations of DMPO, particularly the methyl radical adduct (Fig. 2). This fact may be relevant to the metabolism of acetaldehyde in vivo. Indeed, it is possible that during the enzymatic oxidation of acetaldehyde to acetate, a small fraction of the product escapes as the acetyl radical (Fig. 8, route a). Also, enzymatic routes can be proposed for methyl radical formation by analogy with the mechanism that operates during acetaldehyde oxidation by peroxides such as hydrogen peroxide and peroxynitrite [20,21]. In these cases, detailed studies have shown that a nucleophilic addition of the peroxide to acetaldehyde produces an intermediate that cleaves to methyl radical in the absence or in the presence of transition metal ions, depending on the peroxide nature (Fig. 8, route c). Peroxynitrite peroxo bond spontaneously cleaves to produce free radicals, whereas hydrogen peroxide peroxo bond homolysis requires the assistance of transition metal ions (Fig. 8, route c, parts 1 and 2) [20,21]. Amino acid residues can facilitate peroxide deprotonation, and prosthetic transition metal ions can cleave peroxo bonds. In addition, enzymes such as xanthine oxidase that reduce oxygen to hydrogen peroxide may produce deprotonated peroxide in situ as long as electron transfer is faster than proton transfer. Consequently, it may be speculated that acetaldehyde metabolism to free radicals can occur through hydroxyl radical formation and through nucleophilic addition of peroxides (Fig. 8), among other possible mech-

anisms. In this context, it is important to note that acetaldehyde oxidation by mitochondria has been suggested to produce 1-hydroxyethylperoxide [25]. In conclusion, although metabolic routes and mechanisms remain to be established, our results demonstrate that acetaldehyde is metabolized both in vitro and in vivo to acetyl and methyl radicals. Such species are likely to play a role in the toxic effects associated with acetaldehyde itself or with the most studied of its biological precursors, ethanol. Acknowledgements — This work was partially supported by grants from the Fundac¸a˜o de Amparo a` Pesquisa do Estado de Sa˜o Paulo (FAPESP), Conselho Nacional de Desenvolvimento Cientı´fico e Tecnolo´gico (CNPq), and Financiadora de Estudos e Projetos (FINEP) to O.A.

REFERENCES [1] International Agency for Research on Cancer. IARC monograph on the evaluation of the carcinogenic risk of chemicals to humans: allyl compounds, aldehydes, epoxides and peroxides, Vol. 36. Lyon, France: International Agency for Research on Cancer; 1985. [2] Dellarco, V. L. A mutagenicity assessment of acetaldehyde. Mutat. Res. 195:1–20; 1988. [3] Feron, V. J.; Til, H. P.; de Vrijer, F.; Woutersen, R. A.; Cassee, F. R.; van Bladeren, P. J. Aldehydes: occurrence, carcinogenic potential, mechanism of action and risk assessment. Mutat. Res. 259:363–385; 1991. [4] Niemela, O.; Klajner, F.; Orrego, H.; Vidnis, E.; Blendis, L.; Israel, Y. Antibodies against acetaldehyde-modified proteins epitopes in human alcoholics. Hepatology 7:1210 –1214; 1987. [5] Lin, R. C.; Smith, R. S.; Lumeng, L. Detection of a proteinacetaldehyde adduct in the liver of rats fed alcohol chronically. J. Clin. Invest. 81:615– 619; 1988. [6] Fang, J.-L.; Vaca, C. E. Detection of DNA adducts of acetaldehyde in peripheral white blood cells of alcohol abusers. Carcinogenesis 18:627– 632; 1997. [7] Birt, J. E. E. C.; Shucker, D. E. G.; Farmer, P. B. Stable acetaldehyde-protein adducts as biomarkers of alcohol exposure. Chem. Res. Toxicol. 11:136 –142; 1998. [8] Lieber, C. S. Biochemical factors in alcoholic liver diseases. Semin. Liver Dis. 13:136 –153; 1993. [9] Krikun, J. G.; Lieber, C. S.; Cederbaum, A. I. Increased micro-

Free radicals from acetaldehyde

[10] [11] [12]

[13] [14] [15] [16]

[17]

[18]

[19]

[20]

[21]

[22]

[23] [24] [25] [26]

[27] [28] [29]

somal oxidation of ethanol by cytochrome P-450 and hydroxyl radical-dependent pathways after ethanol consumption. Biochem. Pharmacol. 33:3306 –3309; 1984. Kuthan, H.; Ulbrich, V. Oxidase and oxygenase function of the microsomal cytochrome P-450 monoxygenase system. Eur. J. Biochem. 126:583–588; 1982. Fridovich, I. Oxygen radicals from acetaldehyde. Free Radic. Biol. Med. 7:557–558; 1989. Nakano, M.; Kikuyama, M.; Hasegawa, T.; Ito, T.; Sakurai, K.; Hiraishi, K.; Hashimura, E.; Adachi, M. The first observation of O2⫺ generation at real time in vivo from non-Kupffer sinusoidal cells in perfused rat liver during acute ethanol intoxication. FEBS Lett. 372:140 –143; 1995. Albano, E.; Tomasi, A.; Goria-Gatti, L.; Dianzani, M. U. Spintrapping of free radical species produced during microsomal metabolism of ethanol. Chem. Biol. Interact. 65:223–234; 1988. Knecht, K T.; Bradford, B. U.; Mason, R. P.; Thurman, R. G. In vivo formation of a free radical metabolite of ethanol. Mol. Pharmacol. 38:26 –30; 1990. Moncada, C.; Torres, V.; Varghese, G.; Albano, E.; Israel, Y. Ethanol-derived immunoreactive species formed by free radical mechanisms. Mol. Pharmacol. 46:786 –791; 1994. Clot, P.; Bellomo, G.; Tabone, M.; Arico´, S.; Albano, E. Detection of antibodies against protein modified by hydroxyethyl free radicals in patients with alcoholic cirrhosis. Gastroenterology 108:201–207; 1995. Moore, D. R.; Reinke, L. A.; McCay, P. B. Metabolism of ethanol to 1-hydroxyethyl radicals in vivo: detection with intravenous administration of ␣-(4-pyridyl-1-oxide)-N-t-butylnitrone. Mol. Pharmacol. 47:1224 –1230; 1995. Knecht, K. T.; Adachi, Y.; Bradford, B. U.; Iimuro, Y.; Kadiiska, M.; Qun-Hui, X.; Thurman, R. G. Free radical adducts in the bile of rats treated chronically with intragastric alcohol: inhibition by destruction of Kupffer cells. Mol. Pharmacol. 47:1028 –1034; 1995. Rao, D. N. R.; Yang, M.-X.; Lasker, J. M.; Cederbaum, A I. 1-Hydroxyethyl radical formation during NADPH- and NADHdependent oxidation of ethanol by human liver microsomes. Mol. Pharmacol. 49:814 – 821; 1996. Nakao, L. S.; Oushi, D.; Augusto, O. Oxidation of acetaldehyde by peroxynitrite and hydrogen peroxide/iron (III). Production of acetate, formate, and methyl radical. Chem. Res. Toxicol. 12: 1010 –1018; 1999. Hodges, G. R.; Ingold, K. U. Cage-escape of geminate radical pairs can produce peroxynitrate from peroxynitrite under a variety of experimental conditions. J. Am. Chem. Soc. 121:10695–10701; 1999. Puntarulo, S.; Cederbaum, A. I. Chemiluminescence from acetaldehyde oxidation by xanthine oxidase involves generation of and interactions with hydroxyl radicals. Alcohol Clin. Exp. Res. 13: 84 –90; 1989. Albano, E.; Clot, P.; Comoglio, A.; Dianzani, M. U.; Tomasi, A. Free radical activation of acetaldehyde and its role in protein alkylation. FEBS Lett. 348:65– 69; 1994. Rajasinghe, H.; Jayatilleke, E.; Raw, S. DNA cleavage during ethanol metabolism: role of superoxide radicals and catalytic iron. Life Sci. 47:807– 814; 1990. Boh, E. E.; Baricos, W. H.; Bernofsky, C.; Steele, R. H. Mitochondrial chemiluminescence elicited by acetaldehyde. J. Bioenerg. Biomembr. 14:115–133; 1982. Gonthier, B.; Jeunet, A.; Barret, L. Electron spin resonance study of free radicals produced from ethanol and acetaldehyde after exposure to a Fenton system or to brain and liver microsomes. Alcohol 8:369 –375; 1991. Schuchmann, M. N.; von Sonntag, C. The rapid hydration of the acetyl radical. A pulse radiolysis study of acetaldehyde in aqueous solution. J. Am. Chem. Soc. 110:5698 –5701; 1988. Harris, E. F. P.; Waters, W. A. Thiol catalysis of the homolytic decomposition of aldehydes. Nature (Lond.) 170:212–213; 1952. Fischer, H.; Paul, H. Rate constants for some prototype radical

[30] [31] [32] [33]

[34] [35]

[36] [37]

[38] [39] [40] [41] [42] [43]

[44] [45] [46]

[47] [48]

729

reactions in liquids by kinetic electron spin resonance. Accounts Chem. Res. 20:200 –206; 1987. Bondy, S. C. Ethanol toxicity and oxidative stress. Toxicol. Lett. 63:231–241; 1992. Ishii, H.; Kurose, I.; Kato, S. Pathogenesis of alcoholic liver disease with particular emphasis on oxidative stress. J. Gastroenterol. Hepatol. 12:S272–S282; 1997. Vercesi, A. E.; Reynafarje, B.; Lehninger, A. L. Stoichiometry of H⫹ ejection and Ca2⫹ uptake coupled to electron transport in rat heart mitochondria. J. Biol. Chem. 253:6379 – 6385; 1978. Beltran, C.; Gomez-Puyou, M. T.; Darzon, A.; Gomez-Puyou, A. Simultaneous synthesis and hydrolysis of ATP regulated by the inhibitor protein in submitochondrial particles. Eur. J. Biochem. 160:163–168; 1986. Duling, D. R. Simulation of multiple isotropic spin-trap EPR spectra. J. Magn. Reson. 104:105–110; 1994. Sentjurc, M.; Mason, R. P. Inhibition of radical adduct reduction and reoxidation of the corresponding hydroxylamines in in vivo spin trapping of carbon-tetrachloride-derived radicals. Free Radic. Biol. Med. 13:151–160; 1992. Stoyanovsky, D. A.; Cederbaum, A. I. Metabolism of carbon tetrachloride to trichloromethyl radical: an ESR and HPLC-EC study. Chem. Res. Toxicol. 12:730 –736; 1999. Batteli, M. G.; Abbondanza, A.; Stirpe, F. Effects of hypoxia and ethanol on xanthine oxidase of isolated rat hepatocytes: conversion from D to O form and leakage from cells. Chem. Biol. Interact. 83:73– 84; 1992. Li, A. S. W.; Cummings, K. B.; Roethling, H. P.; Buettner, G. R.; Chignell, C. F. A spin trapping data base implemented on the IBM PC/AT. J. Magn. Reson. 79:140 –142; 1988. Lloyd, R. V.; Mason, R. P. Evidence against transition metalindependent hydroxyl radical generation by xantine oxidase. J. Biol. Chem. 265:16733–16736; 1990. Fridovich, I. The mechanism of the enzymatic oxidation of aldehydes. J. Biol. Chem. 241:3126 –3128; 1966. Finkelstein, E.; Rosen, G. M.; Rauckman, E. J. Spin trapping— kinetics of the reaction of superoxide and hydroxyl radicals with nitrones. J. Am. Chem. Soc. 102:4994 – 4999; 1980. Boveris, A.; Chance, B. The mitochondrial generation of hydrogen peroxide. General properties and effect of hyperbaric oxygen. Biochem. J. 134:707–716; 1973. Giuvili, C.; Boveris, A.; Cadenas, E. Hydroxyl radical generation during mitochondrial electron transfer and the formation of 8-hydroxydesoxyguanosine in mitochondrial DNA. Arch. Biochem. Biophys. 316: 909 –916; 1995. Knecht, K. T.; Mason, R. P. In vivo spin trapping of xenobiotic free-radical metabolites. Arch. Biochem. Biophys. 303:185–194; 1993. Kadiiska, M. B.; Qun-Hui, X. H.; Mason, R. P. In vivo free radical generation by chromium (IV): an electron-spin-resonance spintrapping investigation. Chem. Res. Toxicol. 7:800 – 805; 1994. Kadiiska, M. B.; Morrow, J. D.; Awad, J. A.; Roberts, L. J.; Mason, R. P. Identification of free radical formation and F2isoprostanes in vivo Cr (IV) poisoning. Chem. Res. Toxicol. 11:1516 –1520; 1998. Kadiiska, M. B.; Mason, R. P. Acute intoxication by methanol causes free radical generation in rats: an ESR spin-trapping investigation. Free Radic. Biol. Med. 28:1106 –1114; 2000. Wright, R. M.; MacManaman, J. L.; Repine, J. E. Alcoholinduced breast cancer: a proposed mechanism. Free Radic. Biol. Med. 26:348 –354; 1999.

ABBREVIATIONS

DMPO—5,5-dimethyl-1-pyrroline-N-oxide POBN—␣-(4-pyridyl 1-oxide)- N-t-butylnitrone SMPS—submitochondrial particles