Professional Documents
Culture Documents
PIIS0021925817390488
PIIS0021925817390488
PIIS0021925817390488
11446-11450,1985
Printed in U.S.A.
Volker Fischer, PaulR. West+, SidneyD. Nelson§, Peter J. Harvisont, andRonald P. Mason
From the National Instituteof Environmental Health Sciences, Laboratory of Molecular Biophysics, Research Triangle Park,
North Carolina 27709, the $Departmentof Chemistry, University of Victoria, Victoria, British Columbia V8W2Y2, and the
§Department of Medicinal Chemistry, University of Washington, Seattle, Washington98195
N-Acetyl-p-benzoquinone imine, a hepatic metabo- dramatically in their reactions with the tripeptide glutathione
lite of acetaminophen, andits analogue, N-acetyl-3,5- (4). Whereas glutathione acts solely as a reductant towardN -
dimethyl-p-benzoquinone imine, were metabolized by acetyl-3,5-dimethyl-p-benzoquinone imine, onlya glutathione
rat livermicrosomes and NADPH to their correspond- adduct of N-acetyl-2,6-dimethyl-p-benzoquinone imine is
ing 4-aminophenoxyl free radicals.ESR spectra were formed. N-Acetyl-p-benzoquinone imineitself undergoes both
recorded and unambiguously identified. As indicated reactions with glutathione,yielding acetaminophen and 34s-
by the purple color and confirmed by UV and mass glutathiony1)acetaminophen (4). The absence of covalent
spectroscopy, indophenols were formed as final prod- bond formation by 3’,5’-dimethyl-4’-hydroxyacetanilide(as
ucts. The4-aminophenoxyl freeradicalformation reflected by both radical dimerization and GSH conjugate
could be suppressed by the deacetylase inhibitors,so- formation) suggests a redox mechanism of toxicity for acet-
dium fluoride and paraoxon. Microsomal incubations
of N-acetyl-2,6-dimethyl-p-benzoquinone imine and aminophen (5).
NADPH do not result in a detectable radical concen- Both acetaminophen and its 3‘,5’-dimethylated analogue
tration; in addition, no indophenol was found. Substi- are oxidized by horseradish peroxidase to theircorresponding
tution of NADPH-cytochrome P-450 reductase for rat N-acetyl-p-benzoquinone imines via free radical intermedi-
liver microsomes eliminates the deacetylase activity ates (5, 6). Reduction of N-acetyl-p-benzoquinoneimine
andresultsindirectreduction of N-acetyl-3,5-di- through radical intermediates also has been suggested (2, 4,
methyl-p-benzoquinone imine to the N-acetyl-2,6-di- 7), although no clear evidence for the formationof 4-acetyla-
methyl-4-aminophenoxyl free radical. Neither the in- minophenoxyl free radicals has beenfound. ESR spectrawere
cubation of N-acetyl-p-benzoquinone imine northat of reported previously from the reduction of N-acetyl-p-benzo-
N-acetyl-2,6-dimethyl-p-benzoquinoneimine with quinone imines (4,7), but these spectra either could not be
NADPH-cytochrome P-450 reductaseyielded a detect- assigned unambiguously or were misinterpreted.
able concentrationof the corresponding phenoxy1 free It is extremely important that ESR spectra be accurately
radical. When starting material that had been exposed characterized in order to identify the different free radical
to the atmosphere was used, a previously reported free metabolites which may be formed. For example, deacetylation
radical with a splitting constant of -2 G was formed. of N-acetyl-p-benzoquinone imine in microsomal systems
Thisspectrumisidenticalwiththat of the2,6-di- shouldbe considered. Incombination witha one-electron
methyl-p-benzosemiquinonefree radical, implying hy- reduction, this pathway could lead to the 4-aminophenoxyl
drolysis of the starting material. Neither the N-acetyl- free radical, whereas two-electron reduction would lead to 4-
4-aminophenoxylnorthe N-acetyl-2,6-dimethyl-4- aminophenol, a known nephrotoxic agent (8) (Scheme 1).To
aminophenoxyl radical reduces oxygento form super- investigate this, we studied the reductionof N-acetyl-p-ben-
oxide or react with oxygen in any other detectable zoquinone imine, N-acetyl-3,5-dimethyl-p-benzoquinone im-
way.
OH 0 0
I I I
phen have been found to cause markedly different extents of
hepatic necrosisin rodents (3). 3’,5’-Dimethyl-4’-hydroxy-
acetanilide is approximately equipotent to acetaminophen as
a hepatotoxin, whereas 2‘,6’-dimethyl-4’-hydroxyacetanilide r
is not hepatotoxic. Interestingly,theN-acetyl-p-benzoqui-
noneimines of these two acetaminophenanalogues differ
6
5 4 32
* This work was partially supported by Grant GM 25418 (to S.D.
N.) from the National Institutes of Health. The costs of publication
of this article were defrayed in part by the payment of page charges. NH2 NH2
Thisarticlemusttherefore be hereby marked “aduertisement” in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact. SCHEME
1
11446
c=o
50 ’
C
‘ 1‘ I I
““ ‘ ’ ‘
n
50
DISCUSSION
ESR spectroscopy allowed us to detect two classes of free
radical metabolites from N-acetyl-p-benzoquinone imine and
its analogue, N-acetyl-3,5-dimethyl-p-benzoquinone imine, in
systems containing either NADPH-cytochrome P-450 reduc-
tase or rat liver microsomes. Free radical metabolites of N -
acetyl-2,6-dimethyl-p-benzoquinone imine were not detected.
NADPH-cytochrome P-450 reductase reduced two of the
three N-acetyl-p-benzoquinone imines investigated to yield
detectable free radicals. We found only a broad line signal for
the reduction of N-acetyl-p-benzoquinone imine itself. The
accompanying polymer formation is consistent with the reac-
tivity of the expected primary metabolite, the N-acetyl-4-
aminophenoxyl free radical. This radical reacts on a milli-
second time scale, subsequently forming paramagnetic, mel-
anin-like polymers which account for the signal observed in
C Y static systems (6). Powis et al. (7) reported asimilar spectrum
in a chemical system with dimethyl sulfoxide, but did not
observe this secondary radical with reductase in an aqueous
phase.
The observation of N-acetyl-2,6-dimethyl-4-aminophen-
oxy1 free radical during the reduction of N-acetyl-3,5-di-
methyl-p-benzoquinone imine in purified P-450 reductase
systems is also consistent with our expectations; this radical
is stable in aerobic solutions and does not form a polymer (5).
E V In contrast, the N-acetyl-3,5-dimethyl-4-aminophenoxyl, pos-
I
sibly formed during the reduction of N-acety1-2,6-dimethyl-
-
p-benzoquinone imine, may react further to form a polymer
since it is known to form GSH adducts (4). In any case, we
" 1OG
were unable to detect any type of free radical with N-acetyl-
2,6-dimethyl-p-benzoquinone imine.
Rosen et al. (4) failed to detect N-acetyl-2,6-dimethy1-4-
FIG.5. ESR spectra of incubations containing 100 mM spin aminophenoxyl by direct ESR spectroscopy, but reported an
trap DMPO and possible superoxide-generating systems in ESR spectrum which was assigned to the N-acetyl-3,5-di-
phosphate buffer, pH 7.4, containing 1 mM DTPA. A, 1.0mM methyl-4-aminophenoxyl free radical. This spectrum shows
dihydroxyfumarate, 0.5 mM HzOz, and 2.8 units/ml horseradish per- seven equidistant lines separated by -2 G (4) andcould result
oxidase; B, 1.0 mM dihydroxyfumarate and 2.8 units/ml horseradish from a spin three system, or an even higher spin system, in
peroxidase; C, 1.0 mM dihydroxyfumarate; D,0.5 mM Hz02 and 2.8 case additional lines are hidden in the noise. In contrast, the
units/mlhorseradish peroxidase; E, 1.5 mM 3',5'-dimethyl-4'-hy-
droxyacetanilide and 1.5 mM N-acetyl-3,5-dimethyl-p-benzoquinoneortho position of phenoxyl free radicals is of high spin density
imine; F, 5.0 mM 3,5-dimethyl-4-hydroxyacetanilide, 2.5 mM HZ@, (6), and one would expect the ESRspectrum of N-acetyl-3,5-
and 2.8 units/ml horseradish peroxidase; G, 10.0 mM acetaminophen, dimethyl-4-aminophenoxylfree radical to show a dominant
5.0 mM HzOZ, and 2.8 units/ml horseradish peroxidase. triplet with an intensity ratio of 1:2:1 due to protons in the 2-
and 6-positions. Ortho protons in other N-acetyl-4-amino-
oxide-generating system containing dihydroxyfumarate, substituted phenoxyls have a hyperfine splitting of -5 G (5,
horseradish peroxidase, and hydrogen peroxide (24). The ESR 6). On the other hand, the hyperfine coupling constants as
spectrum consistsof two components which can be identified measured from the spectrum reported in Ref. 4 are identical
as the superoxide adduct and its decomposition product, the to the values reported for 2,6-dimethyl-p-benzosemiquinone
DMPO hydroxyl adduct (22, 25). Omission of hydrogen per- (27), which,canalso be obtained by reduction of 2,6-dimethyl-
oxide yields the same spectrum (Fig. 5B), because this system p-benzoquinone with cytochrome P-450 reductase and
generates hydrogen peroxide (24). Fig. 5, C and D,shows the NADPH (Fig. 6). Hydrolysis of the N-acetyl-3,5-dimethyl-p-
dependence of the signal on both dihydroxyfumarate and benzoquinone imine starting material could account for this
horseradish peroxidase. result.
As shown in Fig. 5, E-G, no detectable DMPO superoxide Neither N-acetyl-4-aminophenoxylnor N-acetyl-2,6-&-
adduct is formed in systems where N-acetyl-4-aminophenoxyl methyl-4-aminophenoxyl was found to generate superoxide.
(Fig. 5G) (6) or N-acetyl-2,6-dimethyl-4-aminophenoxyl free This is consistent with oxygraph experiments, where no de-
radicals (5) are present (Fig. 5, E and F ) . The N-acetyl-2,6- tectable oxygen consumption could be found (26),and is
dimethyl-4-aminophenoxylfree radical generated either in a typical for a phenoxyl free radical, which has a very low
peroxidase system (Fig. 5E) or by comproportionation (Fig. reactivity toward oxygen (28). This result is not surprising in
5F) is stable and does not react with oxygen. In rapid-flow viewof the slow reaction of benzosemiquinone itself with
experiments, N-acetyl-4-aminophenoxylalso does not react oxygen, even though a stable quinone is formed. In the case
with oxygen at a detectable rate, because the steady-state of the N-acetyl-4-aminophenoxylradicals, the initial product
radical concentration is invariant with oxygen saturation (26). would be a cation, with the charge next to the acetyl group
In a static incubation, N-acetyl-4-aminophenoxyl polymerizes making electron transfer to oxygen extremely difficult, as
11450 4-Aminophenoxyl, a Metabolite of N-Acetyl-p-benzoquinone Imine
imine is covalently bound to microsomal protein with loss of
d
20% of the acetyl groupas acetamide. Our results demonstrate
enzymatic deacetylation and the formation of 4-aminophen-
oxyl. This consequentlysuggests thatanyexplanation of
acetaminophen toxicity should consider both theelectrophilic
N-acetyl-4-aminophenoxyl and N-acetyl-p-benzoquinone im-
ine, as well as their deacetylated metabolites, i.e. 4-amino-
phenoxy1 and p-benzoquinone imine. The latter may be the
active intermediates responsible for the toxicity of 4-amino-
phenol. The metabolitesof 4-aminophenol bindcovalently to
protein in kidney and liver and deplete glutathione levels in
the kidney (30). The fact that we donotdetectthe 3,5-
dimethyl-4-aminophenoxylfree radical, which should be rel-
atively stable, suggests that N-acetyl-2,6-dimethyl-p-benzo-
quinone imine is nota good substrate for microsomal deace-
tylation. Thus, the much greater toxicity of acetaminophen
and 3’,5’-dimethyl-4’-hydroxyacetanilide compared to 2‘,6‘-
dimethyl-4’-hydroxyacetanilide ( 3 )could be due to differences
in the rateof deacetylation. In support of this concept is the
FIG. 6. ESR spectrum of 2,6-dimethyl-p-benzosemiquinone recent report that arylation of renal macromolecules in vivo
obtained by reduction of 1.4 mM 2,6-dimethyl-p-benzoqui- by reactive intermediates from p-aminophenol correlateswith
none with 0.7 mM NADPH and 33 units/ml cytochrome P-450
reductase in phosphate buffer, pH 7.4, containing 1 m M acetaminophen-induced nephrotoxicity(31).
DTPA. The spectrometer parameters were: 1-s time constant; 0.66- REFERENCES
G modulation amplitude; 20-mW microwave power; and 2.5-G/min 1. Miner, D. J., and Kissinger, P. T. (1979) Biochem. Pharrnacol. 28, 3285-
scan rate. 3290
2. Dahlin, D. C., Miwa, G. T., Lu, A. Y. H., and Nelson, S.D. (1984) Proc.
Natl. Acad. Sci. CJ. S. A. 81, 1327-1331
shown by our data. DMPO superoxide spectra, like the one 3. Fernando, C. R., Calder, I. C., and Ham, K. N. (1980) J. Med. Chem. 23,
- - - ---
115.1-1158
reported by Rosen et al. (4) in an N-acetyl-p-benzoquinone 4. Rosen, ..
G. M., Raukman, E. J., Ellington, S. P., Dahlin, D. C., Christie, J.
imine/acetaminophen system, can normally be obtained only L., and Nelson, S.D. (1984) Mol. Pharmacol. 2 5 , 151-157
5. Fischer, V., and Mason, R. P. (1984) J . Bid. Chem. 2 5 9 , 10284-10288
in fast redox cycling systems, e.g. the methyl viologen, rat 6. West,P. R., Harman, L. S.,Josephy, P. D., andMason, R. P. (1984)
liver microsomes, and NADPH system (see Fig. 4 of Ref. 29). Biochern. Pharmacol. 3 3 , 2933-2936
7. Powis, G., Svingen, B. A,, Dahlin, D. C., and Nelson, S.D. (1984) Biochem.
We were able to demonstrate thepossibility of spin-trapping Pharmacol. 3 3 , 2367-2370
superoxide radicals with DMPO in a peroxidase system using 8. Newton, J. F., Kuo, C. H., Gemhorys, M. W., Mudge, G. H., and Hook, J.
B. (1982) Toxicol. Appl. Phrmacol. 65,336-344
dihydroxyfumarateas a substrate.TheDMPO superoxide 9. Lauterburg, B. H., Smith, C. V., Hughes, H., and Mitchell, J. R. (1984) J.
Clin. Inuest. 73, 124-133
radical adduct at the concentration present must, therefore, 10. Adams, J. D., Jr., Lauterburg, B. H., and Mitchell, J. R. (1983) J. Pharmacol.
be apoor substrate for eitheroxidation or reduction by Exp. Ther. 227,749-754
horseradish peroxidase. 11. Buettner, G. R., andOberley, L.W. (1978) Biochem. Biophys. Res. Commun.
83.69-74
4-Aminophenoxylfree radicals were obtained in incuba- 12. Dahlin, D. C., and Nelson, S.D. (1982) J . Med. Chem. 25,885-886
tions of N-acetyl-p-benzoquinone imine or N-acetyl-3,5-di- 13. Docampo, R., Mason, R. P., Mottley, C., and Muniz, R. P. A. (1981) J.
Biol. Chem. 256. 10930-10933
methyl-p-benzoquinone imine with ratliver microsomes. It is 14. Yasukochi: Y., and’Masters, B . - S . S. (1976) J. Biol. Chem. 251,5337-5344
C. J., Wolf, C. R., and Philpot, R. M. (1979) J . Bid. Chem.
not clear why this relatively stable free radical was not de- 15. Serabjit-Singh,
254,9901-9907
tected in a similar experiment ( 7 ) . These 4-aminophenoxyl 16. Jackson, R. A. (1983) J. Chem. Soc. Perkin Trans. 11523-529
17. Josephy, P. D., Eling, T. E., and Mason, R. P. (1982) Mol. Pharmacol. 2 3 ,
free radicals will be in a rapid equilibrium with the electro- .” -dfifi
4fil _ _ I
philic p-benzoquinoneiminesand4-aminophenols, 18. Brown, K. C., and Corbett, J . F. (1979) J. Soc. Cosmet. Chem. 30,191-211
which 19. Josephy, P. D., and Van Damme, A. (1984) Biochem. Pharmacol. 33,1155-
react with each other to form the corresponding indophenols. 1156
Three pathwayscould be responsible for the formationof the 20. Boyd, J. A,, and Eling, T. E. (1984) J. Biol. Chem. 2 5 9 , 1388-13896
21. Patton, C. J., and Crouch, S.R. (1977) A d . Chem. 49,464-469
4-aminophenoxyl freeradical (Scheme 1). The N-acetyl-4- 22. Finkelstein, E., Rosen, G. M., Rauckman, E. J., and Paxton,J. (1979) Mol.
Pharmacol. 16.676-685
aminophenoxyl free radical is highly reactive and is not de- 23. Yamaguchi, T., Nakano, T., and Kimoto, E (1984) Biochem. Biophys. Res.
tectable in the experiments with microsomes. Consequently, Commun. 1 2 0 , 534-539
Yamazaki, I., and Piette, L. H. (1963) Biochim. Biophys. Acta 77,47-64
its deacetylation should notbe a major pathway. Our data do 24.25. Harbour, J. R., Chow, V., and Bolton, J. R. (1974) Can. J. Chem. 52,3547-
not allow us to distinguish between direct deacetylationof the 3553
26. Fischer, V., West, P. R., Harman, L. S., and Mason, R. P. (1985) Enuiron.
N-acetyl-p-benzoquinone imine followed by reduction or its Health Perspect., in press
reduction to acetaminophen prior to deacetylation and back- 28.27. Pedersen, J. A. (1973) J . Chem. Soc. Perkin Trans. II424-431
Doba. T.. Burton. G. W.. Ingold. K. U.. and Matsuo. M. (1984) J. Chem.
oxidation to the 4-aminophenoxylfree radical by the parent Soc. Chem. Commun. 461-262
quinone imine. Experiments to investigate this latter mecha- 29. Finkelstein. E.. Rosen. G. M.. and Rauckman. E. J. (1980) Arch. Biochem.
Biophys. 200, 1-16’
nism further are currently inprogress in our laboratory. 30. Crowe, C. A,, Yong, A. C., Calder, I.C., Ham, K. N., and Tange, J. D.
(1979) Chem. Biol. Interact. 27,235-243
The binding data reportedby Dahlin et al. (2) indicate that 31. Newton. J. F.. Pasino. D. A.. and Hook. J . B. (1985) Toxicol. ADD/.P h r -
in mouse liver microsomal systems N-acetyl-p-benzoquinone macoi 7 8 , 39-46