AltCHIVES
OF HIO(‘HF%lISTKY
Vol. 299, No. 2. December,
AND
tlIOPHYS1CS
pp. 313-319,
1992
Oxidation Mechanism of Vitamin E Analogue (Trolox C, 6-Hydroxy-2,2,5,7,8-pentamethylchroman) and Vitamin E by Horseradish Peroxidase and Myoglobin Masao Nakamura’
and Takaaki
Hayashi
Biophysics, Research Institute for Electronic Science, Hokkaido and Hokkaido Institute of Public Health, N-19 W-12, Kita-ku,
Received
April
27, 1992, and in revised
form
July
Academic Press, Inc.
Much attention has been paid to studies of the chemical properties of vitamin E, because vitamin E has been found to prevent the radical chain reactions which lead to lipid peroxidation (Z-4). However, vitamin E is so insoluble correspondence
should
be addressed.
Sapporo
060;
31, 1992
The oxidation of 6-hydroxy-2,2,5,7,8-pentamethylchroman, Trolox C, and a-tocopherol by horseradish peroxidase was examined by stopped-flow and ESR experiments. The catalytic intermediate of horseradish peroxidase during the oxidation of vitamin E analogues and vitamin E was invariably Compound II, and rate constants for the rate-determining step decreased in the order 6-hydroxy-2,2,5,7,8-pentamethylchroman> Trolox C > a-tocopherol. The formation of phenoxyl radicals from substrates was verified with ESR and was followed optically. Resulting 6-hydroxy-2,2,5,7,8-pentamethylchroman and Trolox C radicals decayed through a dismutation reaction, followed by formation of the quinoid form via a transient intermediate. The sequence of events after formation of 6-hydroxy-2,2,5,7,8-pentamethylchroman and Trolox C radicals was similar to that observed by pulse radiolysis (Thomas, M. J., and Bielski, B. H. J. (1989). J. Am. Chem. Sot. 111, 3315-3319). Final oxidation products of 6-hydroxy-2,2,5,7,8-pentamethylchroman and Trolox C were identified as the quinoid forms and were obtained quantitatively whether or not the analogue had a carboxyl or methyl group at the 2-position of chroman ring. In contrast, enzymatic oxidation of a-tocopherol gave a-tocopherol quinone in very low yield. Conversion of 6-hydroxy-2,2,5,7,8-pentamethylchroman, Trolox C, and cu-tocopherol to the corresponding quinones was also catalyzed by metmyoglobin ~11s~ in a reaction completely inhibited by ascorbate.
’ To whom 7861.
University, Kita-ku, Sapporo, Japan
Fax: (Japan)
11.756.
that extensive research has been conducted with watersoluble vitamin E analogue (1,5-S). Using pulse radiolysis, oxidations of vitamin E and its analogues have been examined in detail (1, 5, 9, 10). It has been suggested that the antioxidant activity of vitamin E is ascribable to its hydrogen atom donation from the phenolic hydroxyl group to lipid radicals (2, 3, 11). Synergistic antioxidant effects of vitamin E and vitamin C upon lipid peroxidation have been observed in model systems (2-4, 12-14). In this reaction, lipid radicals are scavenged by vitamin E located in liposomes, generating vitamin E radicals. The chroman ring, which is the chemically reactive head of the vitamin, has been assumed to be directed toward water so that the resulting radicals are accessible to ascorbate to regenerate vitamin E. Although a detectable amount of a-tocopherol quinone (l-4 nmol/g) relative to the u-tocopherol (5-67 nmol/g) has been found in rat liver, the oxidative conversion mechanism is not clearly understood (15,16). Conversion to the quinoid form has been demonstrated when a-tocopherol is exposed to oxygen radicals generated by irradiation (10) or from the xanthine oxidase system (17), though no quinoid form is obtained after reaction of CYtocopherol with organic peroxide (7). In previous papers (18, 19), it was reported that the phenolic hydroxyl group of n-tocopherol and Trolox C’ is selectively oxidized by peroxidases by way of a oneelectron transfer, and the final oxidation product of Trolox C is Trolox C quinone (18). Thomas and Bielski have delineated the sequence of events after one-electron oxidation of Trolox C by pulse radiolysis (1). They further pointed out that a carboxyl group at the 2-position of
‘Abbreviations used: Trolox C, 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid, Trolox C quinone, 2-hydroxy-2-carboxy-4.(3,5,6trimethylbenzoquinone-2-yl) butane; PMC, 6-hydroxy-2,2,5,7&pentamethylchroman; PMC quinone, 2-hydroxy-2-methyl-4-(3,S$trimethylbenzoquinone-2-yl) butane. 313
314
NAKAMURA
Troiox
AND
C : R=COOH
PMC : R=CH3
0
I
a-Tocopherdquinone I
H
: R=
3
cd Trolox
ti 0
C quinone
PMC quinone
: R=COOH
: R=CH3
SCHEME 1
Trolox C is needed to form the corresponding quinone and that this analogue is not a good model for the study of a-tocopherol. In this study, we further investigated the enzymatic oxidation mechanism of PMC which has a methyl group at the 2-position instead of a carboxyl group (Scheme l), and found that the oxidation rate was higher than that of Trolox C and a-tocopherol. Therefore, it is of considerable interest to determine the final oxidation products of PMC, Trolox C, and a-tocopherol after one-electron oxidation of the phenolic hydroxyl group. Quantitative analysis of the final oxidation products was also conducted by HPLC in the presence or absence of ascorbate. MATERIALS
AND
METHODS
Horseradish peroxidase was purified by the method of Shannon et al. (20) from a crude preparation purchased from Toyobo Co. (Osaka). The
A
HAYASHI
enzyme used was isoenzyme C (A,,,/Azso = 3.4) and the concentration was calculated on the basis of a value for t of 100 mM- ‘cm ’ at 403 nm. Horse heart metmyoglobin was obtained from Sigma. Trolox C and PMC were purchased from Aldrich and Waco Pure Chemicals (Osaka), respectively. Vitamin E (DL-tu-tocopherol) obtained from Sigma was further purified with HPLC before use. The other reagents used were of the highest purity available. Stopped-flow measurements were performed using a Union Giken Model RA-1300 rapid reaction analyzer, and spectral changes were made with a Shimadzu UV300 dual-beam spectrophotometer. ESR spectra were recorded on a Varian E-109B spectrometer equipped with a stoppedflow apparatus. The oxidation products of Trolox C, PMC, and u-tocopherol were analyzed with a Hitachi 655 HPLC system equipped with a Radial Pak 8NVC18 column (8 mm X 10 cm, Waters). Trolox C quinone, PMC quinone, and a-tocopherol quinone were obtained from the reactions of Trolox C, PMC, and a-tocopherol with FeC13 (21). The oxidation products were purified with HPLC and identified as corresponding quinones by NMR and by mass and infrared spectrometry. The reactions were carried out in 0.1 M potassium phosphate (pH 7.4) at 25°C unless otherwise noted.
RESULTS
Oxidations of vitamin E analogues (Trolox C, PMC) and vitamin E (ol-tocopherol) by horseradish peroxidase were followed in the uv region ranging from 350 to 220 nm. Figure 1A shows the increase in absorbance at 270 nm assignable to production of Trolox C quinone (enzyme concentration = 30 nM). Using the extinction coefficient of Trolox C quinone, the molar ratio of quinoid form per added H,Oz was estimated to be 1.0. With PMC, the same spectral changes were observed (enzyme concentration = 5 nM) (Fig. 1B). As can be seen in Fig. lC, only a slight decrease in absorbance at 290 nm was observed during oxidation of n-tocopherol by horseradish peroxidase in alcoholic solvent. The result suggests that the reaction
B
C
8 ZO.O! D 8 :: a
250 Wa”elengthRLtl)
FIG. 1. Spectral changes of Trolox C (A), PMC (B), and n-tocopherol (C) catalyzed by horseradish peroxidase. (A) Concentrations were 100 PM Trolox C, 50 @M H,O*, and 30 nM horseradish peroxidase. Reaction was carried out in 0.1 M phosphate buffer (pH 7.4). (B) Concentrations were 100 FM PMC, 50 FM H,O*, and 5 nM horseradish peroxidase. Reaction was carried out in 0.1 M phosphate buffer (pH 7.4). (C) Concentrations were 20 pM n-tocopherol, 80 pM H,O,, and 30 nM horseradish peroxidase. Reaction was carried out in 20 mM phosphate buffer containing 50% (v/v) methanol with continuous stirring. Using HPLC, the complete recovery of a-tocopherol was confirmed before reaction. The peroxidase activity of the enzyme was reduced by 65% in the alcoholic solvent (19). Spectra were obtained at the times indicated.
OXIDATION
OF VITAMIN
E AND
ITS
ANALOGUES
315
BY PEROXIDASE
B
FIG. 2. Difference spectra observed during the oxidation of Trolox C (Aa) and PMC (Ba) in the absence or presence of ascorbate. The difference spectra were plotted from stopped-flow traces at varying wavelengths at the times indicated. (A) Reactions were carried out in the solution contained 100 FM Trolox C, 12 j&M H,O,, and 1.2 pM horseradish peroxidase in the presence (0) or absence (0) of 100 pM ascorbate. Time courses of stopped-flow traces at 432 nm in the presence or absence of ascorbate are shown in b. (B) Reactions were carried out in solution containing 50 pM PMC, 12 FM H,Os, and 1.2 NM horseradish peroxidase in the presence (0) or absence (0) of 100 pM ascorbate.
gave the oxidation products in low yield. Thomas and Bielski (1) have concluded that a quinoid form could be produced from vitamin E analogues that have a carboxyl moiety at the 2-position. However, the spectral change in PMC in the enzyme reaction mixture suggests, by analogy with the Trolox C-HZO,-peroxidase system (Fig. lA), that the oxidation product of PMC is a quinoid form. From stopped-flow kinetics and ESR experiments it has been inferred that peroxidases catalyze a one-electron oxidation of Trolox C and a-tocopherol (18, 19). Figure 2 shows the difference spectra reconstructed from stopped-flow traces at different wavelengths. The increase in absorbance at 411 nm, which is an isosbestic wavelength between ferric and Compound II of the enzyme, represents the transiently formed intermediate (Fig. 2A). The primary oxidation products of phenolic compounds by horseradish peroxidase have been confirmed as phenoxyl radicals (22-25). Therefore, the oxidation of PMC by the enzyme was performed in the presence of 100 PM ascorbate, since phenoxyl radicals should be scavenged by ascorbate and GSH (18,24-27). The second-order rate constants of the reaction of ascorbate with horseradish peroxidase Compound II and Trolox C radicals are 1.5 X 10’ (25) and 8 X lo6 M-r s-l (8), respectively. Therefore, the results indicate that Compound II of horseradish peroxidase could be obtained in the steady state during the oxidation of PMC (Fig. 2Ba), and the rate constant for the reaction of Compound II with PMC was estimated to be 1.3 X lo6 M-’ so’ in the presence of ascorbate (Fig.
FIG. 3. Spectra of Trolox C and PMC radicals during the oxidation of Trolox C and PMC by horseradish peroxidase. Spectra were plotted from stopped-flow traces at varying wavelengths at the times indicated. (A) Concentrations were 1 mM Trolox C, 50 GM H,O1, and 0.29 +M horseradish peroxidase. (B) Concentrations were 0.5 mM PMC, 75 FM H202, and 0.29 KM horseradish peroxidase. The absorbance changes below 440 nm were corrected by subtracting the absorbance change due to the formation of horseradish peroxidase Compound II.
2Bb). As shown in Figs. 3A and 3B, the spectra in the presence of 0.29 PM enzyme represent Trolox C radicals and PMC radicals, respectively, and are identical to that of Trolox C radicals formed by pulse radiolysis and enzymatic reaction (1,5,18). The results enabled us to follow the PMC radicals at 440 nm under these conditions. The concentrations of PMC radicals in the steady state were proportional to the square root of the PMC concentration (data not shown). Since the concentration of PMC radicals is calculated to be 11 FM (Fig. 4a), the dismutation constant of the radicals is estimated to be 1.6 X lo6 M-’ s-l (5, 18). No radicals were observed in the presence of 100 PM ascorbate (Fig. 4e). To confirm the formation of PMC radicals during the peroxidase reaction, ESR experiments were performed.
Time(s)
FIG. 4. Effect of ascorbate concentrations upon PMC radicals in the steady state during the oxidation of PMC by horseradish peroxidase system. Concentrations were 100 @M PMC, 50 pM Hz02, and 0.29 pM horseradish peroxidase. Reactions were carried out in the presence of 0 (a), 25 (b), 45 (c), 60 (d), and 100 (e) IJM ascorbate.
316
NAKAMURA
AND
HAYASHI
C
a
b
b '40
‘56
'ho
c FIG. 5. ESR spectra observed during the oxidation of Trolox C (A), PMC (B), and cu-tocopherol (C) by horseradish peroxidase system in the absence (a) or presence (b) of ascorbate. (A) Concentrations were 1 mM Trolox C, 0.5 mM H,OZ, and 1.1 pM horseradish peroxidase. Reactions were carried out in the absence (a) or presence (b) of 0.5 mM ascorbate. Instrumental conditions were gain, 2.0 X lo*; power, 20 mW; modulation amplitude, 1 G; time constant, 0.125 s; scan rate, 100 G/min. Spectra were obtained at 30 s after reactions were started. (B) Concentrations were 0.5 mM PMC, 0.5 mM HzOz, and 1.1 pM horseradish peroxidase. Instrumental conditions were the same as in A. (C) Concentrations in the flow cell were 8.0 mM oc-tocopherol, 0.2 mM H,O*, and 9.1 pM horseradish peroxidase. Reactions were carried out in solution containing 50% (v/v) methanol in the absence (a) or presence (b) of 1 mM ascorbate. Instrumental conditions were gain, 5 X 104; power, 63 mW; modulation amplitude, 1 G; time constant, 0.125 s; scan rate, 50 G/min. Spectra were taken during flow (0.2 ml/s).
The seven-line hyperfine pattern characteristic of phenoxyl radicals such as Trolox C and vitamin E radicals formed by peroxidase reactions (Figs. 5Aa and 5Ca) was observed during the reaction (Fig. 5Ba). The addition of ascorbate gave rise to a change of spectra indicating the formation of ascorbate radicals (Figs. 5b). The reaction of phenoxyl radicals with ascorbate has been examined by pulse radiolysis and the reaction rate has been documented (8). Since the rate of the reaction of PMC radicals with ascorbate and the dismutation of PMC radicals are nearly the same, ascorbate scavenged PMC radicals under the experimental conditions. The direct observation of PMC-mediated oxidation of ascorbate was monitored at 284 nm (Fig. 6), which is an isosbestic wavelength between that of PMC and PMC quinone (Fig. 1B). From the initial decrease in absorbance, the reaction rate for a rate-determining step of the reaction is calculated to be 1.2 X lo6 M-l s1 (Table I). The reaction rate was compatible with the rate obtained for the reaction of Compound II with PMC. Ascorbate radicals decay through a dismutation reaction with a value of 1 X lo5 M-’ s-l at pH 7.2 (28-30). These kinetic results supported the accumulation of ascorbate radicals during the PMC-mediated oxidation of ascorbate by horseradish peroxidase (Fig. 5Bb).
Horseradish peroxidase catalyzed the one-electron oxidation of PMC, Trolox C, and a-tocopherol (18, 19). Thomas and Bielski have observed the sequence of reactions after Trolox C radical formation (1). Trolox C radicals undergo dismutation reaction to form a crossconjugated keto diene, which is hydrolyzed to Trolox C quinone. Figure 7 shows the time courses of stopped-flow
E I
FIG. 6. Oxidation of ascorbate catalyzed by PMC-horseradish peroxidaseeHZ02 system. Concentrations were 40 pM PMC, 0.2 mM ascorbate, 20 or 40 pM H202 and 30 nM horseradish peroxidase. The oxidation of ascorbate was followed optically at 284 nm, which is an isosbestic wavelength between PMC and corresponding quinone. The oxidation of ascorbate by Trolox C system was observed similarly.
OXIDATION TABLE
OF VITAMIN
E AND
I
Rate Constants (Mm’ se‘) of the Reaction of Horseradish Peroxidase Compound II with a-Tocopherol and Vitamin E Analogue and Dismutation Constants of Corresponding Radicals
ITS
ANALOGUES BY PEROXIDASE
317
PMC, Trolox C, and cu-tocopherol radicals can be repaired by ascorbate. DISCUSSION
The phenolic hydroxyl groups of Trolox C, PMC, and a-tocopherol are readily oxidized by peroxidases. The rate PMC a-Tocopherol constant of the reaction of peroxidase Compound II with Carboxylic PMC is lOO-fold faster than that with Trolox C. However, Phytyl” acid” Methyl” no clear difference in the steady-state concentration of the radicals is observed at the substrate concentrations Rate constants (M-’ s-‘) used (Figs. 2Aa and 2Ba). The dismutation constants of 5.1 x IOZd 1.1 x 10”’ 1.3 x 10” Stopped-flow 1.5 x 10” c 1.2 x lo6 Overallh Trolox C (1, 18) and PMC radicals are 2.1 X lo4 and 1.6 Dismutation constants of X 10” Mu i s- ‘, respectively. Since these radicals disappear 2.1 x lo4 c 1.6 X 10” the radicals (Mm’ s ‘1 only through the dismutation reaction, the results are explained in terms of the increase in the dismutation con’ A substituent at the 2position. stant of PMC radicals. * The rate constants were estimated from the initial velocities of Trolox C- or PMC-mediated oxidation of ascorbate by peroxidase. Synergistic inhibitory effects of vitamin E and ascor’ Values were given in our previous experiments (18, 19). bate upon lipid peroxidation have been documented (2“The reaction was carried out in the same buffer containing 50% 4). Reactions of peroxyl radicals with vitamin E were ex(v/v) methanol (10). amined by an indirect method (3, 31). Chamulitra and Mason (32) observed that the concentration of peroxyl radicals generated by the lipoxygenase system decreased traces at various wavelengths during the oxidation of in the presence of vitamin E or Trolox C, resulting in the Trolox C by the enzyme. Similar increases in absorbance formation of vitamin E and Trolox C radicals. It has been at 440 and 330 nm indicate the formation of phenoxyl accepted that vitamin E could be regenerated through the radicals. ESR signals were observed to have similar time courses under the same conditions. Time courses at 284 reaction of vitamin E radicals with ascorbate and GSH, and 233 nm, the isosbestic wavelength between Trolox C which are present in vivo at high concentration (33). Several lines of evidence have suggested that hemogloand Trolox C quinone, are quite different. Since it has bin and myoglobin catalyze the peroxidation of unsatubeen shown that a cross-conjugated keto diene compound, rated fatty acids and that the ferry1 form is implicated in an intermediate formed through the dismutation of Trolox the reaction (34-37). The chemical properties of ferrylC radicals, has a strong absorbance at 235-240 nm (l), myoglobin are similar to those of horseradish peroxidase the time course at 233 nm indicates the formation and Compound II, which is an oxidant that reacts with subdecay of the intermediate. The initial lag phase was seen in the time course of the formation of Trolox C quinone at 270 nm, which reached a maximal rate at 13 s after the A start of the reaction. The maximal level in the time course at 233 nm was obtained at almost the same time (Fig. 7Ad). Although there is no conclusive evidence that the observed intermediate is identical to that obtained by pulse radiolysis (l), the results confirmed the assumption that the 233-nm band was a precursor of the final product, 40 Time(s) Trolox C quinone. The final oxidation products of PMC, Trolox C, and cY-tocopherol by horseradish peroxidase were examined with HPLC (Table II). With PMC and Trolox C, the molar ratio of quinoid form to added hydrogen peroxide was nearly 1.0, provided that the reactions were carried out in the presence of a limiting amount of H,02 (25 @I) (Table II). The molar ratio was compatible with that calculated from the increase of absorbance at 270 nm (Figs. FIG. 7. Stopped-flow kinetics for the oxidation products of Trolox C catalyzed by horseradish peroxidase. Concentrations were 250 pM Trolox 1A and 1B) using the molar extinction coefficient of Trolox C quinone (1). No formation of quinoid form was ob- C, 25 WM H202, and 0.29 pM horseradish peroxidase. (A) The stoppedflow traces were observed at 440 (a), 330 (b), 284 (c), and 233 (d) nm. served upon the addition of 100 yM ascorbate. A similar (B) The increase in absorbance at 270 nm indicates the formation of pattern was obtained when horseradish peroxidase was Trolox C quinone (Fig. 1A). The kinetic traces on PMC resembled those replaced by metmyoglobin. The results indicate that on Trolox C and are omitted in the figure. Trolox
C
318
NAKAMURA TABLE
II
Concentrations of Quinoid Form after Oxidations of Vitamin E Analogue and a-Tocopherol by Horseradish Peroxidase and Metmyoglobin in the Presence or Absence of Ascorbate
Oxidation by
HRP 60 nM HRP 60 nM HRP 60 nM HRP 60 nM Mb 1.3 PM Mb 1.3 PM Mb 1.3 /LM Mb 1.3 uM HRP 60 nM HRP 60 nM HRP 60 nM HRP 60 nM Mb 1.3 jtM Mb 1.3 /iM Mb 1.3 /AM Mb 1.3 /iM HRP 1.9 ELM HRP 1.9 PM HRP 1.9 /.tM HRP 1.9 FM Mb 5.2 @M Mb 5.2 /AM Mb 5.2 /LM Mb 5.2 /LM
cu-Tocopherol or vitamin E analogue (PM) Trolox C 50 50 50 50 50 50 50 50 PMC 50 50 50 50 50 50 50 50 ol-Tocopherol 30 30 30 30 30 30 30 30
HA (PM)
Ascorbate (PM)
Quinoid form (PM)
0 25 50 50 0 25 50 50
0 0 0 100 0 0 0 100
0.0 24.5 49.3 0.0 0.0 12.6 24.3 0.0
0 25 50 50 0 25 50 50
0 0 0 100 0 0 0 100
0.0 25.2 50.5 0.0 0.0 13.0 25.5 0.0
0 25 50 50 0 25 50 50
0 0 0 100 0 0 0 100
0.0 0.22 0.40 0.0 0.0 0.22 0.25 0.0
Note. Reactions were carried out in 20 mM phosphate buffer (pH 7.4) containing 50 pM Trolox C, 50 pM PMC, or 30 FM Lu-tocopherol, various concentrations of H,O,, and various concentrations of HRP or metmyoglobin in the presence or absence of 100 FM ascorbate. All reactions were carried out in a single phase and the oxidation products were analyzed by HPLC after incubation for 30 min at 25°C. For the determination of final oxidation products, the following HPLC conditions were used: (A) Trolox C and Trolox C quinone-mobile phase, methanol(pH 2.3) (1:l); flow rate, 2.0 ml/min; detector wave10 mM NH,H*PO, length, 270 nm; retention times of Trolox C and Trolox C quinone, 9.8 and 7.3 min, respectively. (B) PMC and PMC quinone-mobile phase, methanol-water (65:35); flow rate, 2.5 ml/min; detector wavelength, 285 nm; retention times of PMC and PMC quinone, 8.1 and 4.3 min, respectively. (C) n-Tocopherol and n-tocopherol quinone-mobile phase, methanol-water (97:3); flow rate, 3.0 ml/min; detector wavelength, 270 nm; retention times of ru-tocopherol and cu-tocopherol quinone, 6.9 and 4.5 min, respectively.
to yield the ferric form and substrate strate molecules radicals. Therefore, myoglobin and peroxidase share common characteristics with regard to peroxidase activity, although the activity of myoglobin is very low. Trolox C and PMC quinone were detected when incubations were performed in the presence of myoglobin instead of horseradish peroxidase. Formation of the quinoid form depended on the ascorbate concentration. Neither Trolox
AND
HAYASHI
C nor PMC quinone was formed when the reactions were carried out in the presence of 100 yM ascorbate (Table II). When the concentration of ascorbate was decreased to 50 PM, no Trolox C quinone was found but about 20% of the PMC quinone production remained. This observation was anticipated because the dismutation constant of PMC radicals was faster than that of Trolox C radicals. Although metmyoglobin shows a weak ascorbate peroxidase activity, the concentration of ascorbate radicals in the steady state was increased by the addition of Trolox C and PMC (data not shown). Results confirmed that ferrylmyoglobin oxidizes vitamin E analogues by way of a one-electron transfer, causing the phenoxyl radical-mediated oxidation of ascorbate. Ferrylmyoglobin may be formed during ischemia and is implicated in reperfusion injury (38). Ferrylmyoglobin would be derived from the reaction of metmyoglobin and deoxymyoglobin with peroxides. It has been pointed out that oxymyoglobin is oxidized by drugs to metmyoglobin and also autoxidizes to metmyoglobin (34, 38, 39). Ferrylmyoglobin has been observed in perfused rat heart exposed to ethylhydroperoxide (40), and ferrylhemoglobin is also detected in red blood cells (41). No main oxidation products of a-tocopherol after one-electron oxidation by peroxidase and metmyoglobin could be identified (Table II), whereas loss of LYtocopherol was observed with the increase in hydrogen peroxide. The amount of a-tocopherol quinone was slightly increased when the reactions were carried out under acidic pH. Fukuzawa and Gebicki have reported that a-tocopherol quinone was formed from a-tocopherol after reaction with oxygen radicals in acidic solutions, irrespective of the location of the vitamin in model systems (10). Nishikimi et al. (17) observed the intermediate during the reaction of a-tocopherol with superoxide at neutral pH, which was spontaneously converted to its quinoid form considered to be 8a-hydroxytocopherone. The results imply that a substantial amount of tocopherol quinone was obtained in the reaction of a-tocopherol with oxygen radicals. Analysis of the oxidation products of a~tocopherol with various oxidants has suggested that the mechanism of a-tocopherol oxidation varies with the nature of oxidants (16). For the reasons outlined, peroxidase reactions provide a means of overcoming this problem. Since Trolox C, PMC, and n-tocopherol radicals produced by peroxidase reaction are liberated from the enzyme (18), the amount of quinoid formed after a one-electron oxidation of a-tocopherol and its analogues is depended upon the properties of the corresponding phenoxyl radicals. It is important to note that both a-tocopherol radicals and 8a-hydroxychromanone were reduced back to cu-tocopherol by ascorbate (17). The final oxidation products of phenolic compounds such as tyrosine or methoxyphenol after the one-electron oxidation by peroxidase have been identified as dimers (42, 43). The main products of (Ytocopherol, which has methyl groups attached to the aromatic ring, might be dimer products (1, 7). Therefore, it
OXIDATION
OF
VITAMIN
E AND
is reasonable to assume that no main products of a-tocopherol after one-electron oxidation by the enzymes could be detected under the same HPLC conditions (Table II). It has been generally accepted that quinoid forms found in antitumor drugs undergo redox cycling by NAD(P)H-dependent enzymes (44). Tocopherol quinone has been found in animal tissues, although at quite low concentrations; however, it is not clear whether the quinoid form exists as a reduced or an oxidized form in viuo because the reduced form is autoxidizable (15). It has been confirmed that PMC and Trolox C exhibit antioxidant as well as oc-tocopherol activity in homogeneous and liposome systems (32,45). Convincing evidence exists that the phytyl side chain of a-tocopherol at the 2-position facilitates the incorporation and proper retention of the vitamin in biomembranes (46). The present results have shown that one-electron oxidation of the phenolic hydroxyl group initiates the reactions which lead to the formation of PMC and Trolox C quinone, while the phytyl side chain of the vitamin at the 2-position decreases or prevents the reactions. In contrast, a different reaction path has been postulated when n-tocopherol is exposed to superoxide (10, 17). In this path, superoxide may act as an oxidant and, subsequently, a nucleophilic reagent or reductant.
ITS
ANALOGUES
D., and Kelley,
15. Bieri,
J. G., and Tolliver, G. T. (1989)
17. Nishikimi,
M., Yamada, 101-108.
We thank Dr. Richard S. Magliozzo (Albert Einstein College of Medicine of Yeshiva University) for reading the manuscript. This work was supported by a grant from the Suhara Memorial Foundation.
REFERENCES 1. Thomas, M. J., and Bielski, 3315-3319. K. U. (1961)
4. Niki, 5. Bisby,
Chem.
G. W., and Ingold, E. (1987)
J. Am. Chem. Sot. 111,
B. H. J. (1989) Reu. 61,563-589. K. U. (1986)
Act. Chem. Res. 19, 194-
Chem. Phys. Lipids 44, 227-253.
R. H., Ahmed,
S., and Cundall,
R. B. (1984)
Biophys.
Biochem. Biophys. Res. Commun.
8. Davies, M. ,J., Form. 255, 513522.
L. G., and Willson,
9. Packer, J. E., Slater, 737-738.
T. F., and Willson,
K., and Gebicki,
,J. (1983)
C., and
13. Mukai, K., Nishimura, 266, 274-278.
Biochim. Biophys.
K. (1980)
M. (1990)
C. K., Draper,
H. H., Csallany,
A. S., and Chiu,
M. (1967)
22. Yamazaki, I. (1977) Vol. 3, pp. 183-218,
in Free Radicals in Biology (Pryor, Academic Press, New York.
W. A., Ed.),
23. Mason, R. P. (1980) Vol. 5, pp. 161-222,
in Free Radicals in Biology (Pryor, Academic Press, New York.
W. A., Ed.),
24. Ohnishi, T., Yamazaki, zaki, I. (1969) Biochim. 25. Nakamura,
H., Iyanagi,
T., Nakamura,
T., and Yama-
Biophys. Acta 172, 357-369.
M., Yamazaki,
I., Kotani,
T., and Ohtaki,
S. (1985)
J.
T., and Ohtaki,
S. (1989)
J.
Biol. Chem. 260, 13546613552. 26. Nakamura,
M., Yamazaki, I., Kotani, 12909-12913.
Biol. Chem. 264,
27. Ross, D., Norbeck, 15028-15032.
K., and Moldeus,
28. Yamazaki, I., Mason, 235, 244462449. 29. Yamazaki,
P. (1985)
H. S., and Piette,
I., and Piette,
L. (1961)
J. Biol. Chem. 260,
J. Biol. Chem.
L. (1960)
Biochim. Biophys. Acta 50, 62-
69. B. H. ,J., Comstock,
D. A., and Bowen,
J. Am.
R. A. (1971)
R. L. (1988) R. L. (1979)
P. T.
Biochem. J. Nature
278,
Arch. Biochem. Biophys. 226,
Martell,
Y., and Ishizu,
H. A. (1941)
M., and Kikuchi,
K. (1986)
R&l.
J. Am. Chem. Sot. 63, S. (1991)
J. Rid. Chem.
J. Am. Chem. Sot. 103,
31. Burton, G. W., and Ingold, 647226477.
K. U. (1981)
32. Chamulitrat,
R. P. (1989) J. Biol. Chem. 264,20968-
W., and Mason,
20973. 33. Kosower, N. S., and Kosower, E. M. (1976) in Free Biology (Pryor, W. A., Ed.), Vol. 2, pp. 55-84, Academic York.
Radicals in Press, New
S. (1985) Arch. Biochem. Biophys. 237,314-
34. Kanner, 321.
J., and Harel,
35. Kanner,
J., and Harel,
S. (1985)
36. Galaris, D., Mira, D., Sevanian, (1988) Arch. Biochem. Biophys. D., Sevanian,
Lippi& 20, 625-628 A., Cadenas, E., and Hochstein, 262, 221-231.
A., Cadenas, E., and 163-169.
Hochstein,
P.
P. (1990)
Arch. Biochem. Biophys. 281, D., Cadenas,
E., and Hochstein,
P. (1989)
Arch. Biochem.
Biophys. 273, 497-504. 39. Tajima, 12605.
G., and Shikama.
40. Tamura, M., 1019-1031. 41. Giulivi, 19460.
Oshino,
K. (1987)
N., and Chance,
C., and Davies,
K. J. A. (1990)
42. Nonhebel, D. C., and Walton, p. 327, Cambridge Iiniversity 43. Ohtaki,
K., Watanabe, Y., Uemoto, Sot. Jpn. 59, 3113-3116.
12. Golumbic, 127991280.
H., and Yagi,
Lipids 2, 390-396.
38. Galaris,
A. (1985)
7. Suarna, C., Craig, D. C., Cross, K. J., and Southwell-Keely, (1988) J. Org. Chem. 53, 1281L1284.
11. Mukai, Chem.
16, 7777779.
M. (1991)
20. Shannon, 2166-2172.
37. Galaris,
B&hem.
Res. Commun. 119,245-251. 6. Ozawa, T., and Hanaki, 126, 873878.
IO. Fukuzawa, 2422251.
Lipids
Lipids 24, 299-304.
Chem. ?;oc. 93, 5624-5629.
ACKNOWLEDGMENTS
3. Burton, 201.
T. J. (1981)
J. Biochem. 108, 245-249. J. Biochem. 110, 595-597. L. M., Key, E., and Lew, J. Y. (1966) J. Bid. Chem. 241,
18. Nakamura, 19. Nakamura,
21. Chow,
FEBS Lett. 284, 147-151.
P. M. (1991)
16. Vatassery,
Acta 627,
319
PEROXIDASE
14. Njus,
30. Bielski,
2. Ingold,
BY
S., Nakagawa,
J. Biol. Chem. 262, B. (1982)
J. B&hem.
92,
J. Bid. Chem. 265, 19453-
J. C. (1974) in Free-Radical Press, London.
H., Nakamura,
12603-
Chemistry,
M., and Yamazaki,
I. (1982)
J. Biol. Chem. 257,
13398-13403. 44. Powis, G. (1989) Free Radical Bid. Med. 6, 63-101 45. Kagan, V. E., Serbinova, E., and Packer, L. (1990) Arch. Biochem. Biophys. 282, 221-225. 46. Niki, E., Kawakami, Kamiya, Y. (1985)
A., Saito,
M., Yamamoto, Y., Tsuchiya, 2191-2196.
J. Biol. Chcm. 260,
J., and
OF HIO(‘HF%lISTKY
Vol. 299, No. 2. December,
AND
tlIOPHYS1CS
pp. 313-319,
1992
Oxidation Mechanism of Vitamin E Analogue (Trolox C, 6-Hydroxy-2,2,5,7,8-pentamethylchroman) and Vitamin E by Horseradish Peroxidase and Myoglobin Masao Nakamura’
and Takaaki
Hayashi
Biophysics, Research Institute for Electronic Science, Hokkaido and Hokkaido Institute of Public Health, N-19 W-12, Kita-ku,
Received
April
27, 1992, and in revised
form
July
Academic Press, Inc.
Much attention has been paid to studies of the chemical properties of vitamin E, because vitamin E has been found to prevent the radical chain reactions which lead to lipid peroxidation (Z-4). However, vitamin E is so insoluble correspondence
should
be addressed.
Sapporo
060;
31, 1992
The oxidation of 6-hydroxy-2,2,5,7,8-pentamethylchroman, Trolox C, and a-tocopherol by horseradish peroxidase was examined by stopped-flow and ESR experiments. The catalytic intermediate of horseradish peroxidase during the oxidation of vitamin E analogues and vitamin E was invariably Compound II, and rate constants for the rate-determining step decreased in the order 6-hydroxy-2,2,5,7,8-pentamethylchroman> Trolox C > a-tocopherol. The formation of phenoxyl radicals from substrates was verified with ESR and was followed optically. Resulting 6-hydroxy-2,2,5,7,8-pentamethylchroman and Trolox C radicals decayed through a dismutation reaction, followed by formation of the quinoid form via a transient intermediate. The sequence of events after formation of 6-hydroxy-2,2,5,7,8-pentamethylchroman and Trolox C radicals was similar to that observed by pulse radiolysis (Thomas, M. J., and Bielski, B. H. J. (1989). J. Am. Chem. Sot. 111, 3315-3319). Final oxidation products of 6-hydroxy-2,2,5,7,8-pentamethylchroman and Trolox C were identified as the quinoid forms and were obtained quantitatively whether or not the analogue had a carboxyl or methyl group at the 2-position of chroman ring. In contrast, enzymatic oxidation of a-tocopherol gave a-tocopherol quinone in very low yield. Conversion of 6-hydroxy-2,2,5,7,8-pentamethylchroman, Trolox C, and cu-tocopherol to the corresponding quinones was also catalyzed by metmyoglobin ~11s~ in a reaction completely inhibited by ascorbate.
’ To whom 7861.
University, Kita-ku, Sapporo, Japan
Fax: (Japan)
11.756.
that extensive research has been conducted with watersoluble vitamin E analogue (1,5-S). Using pulse radiolysis, oxidations of vitamin E and its analogues have been examined in detail (1, 5, 9, 10). It has been suggested that the antioxidant activity of vitamin E is ascribable to its hydrogen atom donation from the phenolic hydroxyl group to lipid radicals (2, 3, 11). Synergistic antioxidant effects of vitamin E and vitamin C upon lipid peroxidation have been observed in model systems (2-4, 12-14). In this reaction, lipid radicals are scavenged by vitamin E located in liposomes, generating vitamin E radicals. The chroman ring, which is the chemically reactive head of the vitamin, has been assumed to be directed toward water so that the resulting radicals are accessible to ascorbate to regenerate vitamin E. Although a detectable amount of a-tocopherol quinone (l-4 nmol/g) relative to the u-tocopherol (5-67 nmol/g) has been found in rat liver, the oxidative conversion mechanism is not clearly understood (15,16). Conversion to the quinoid form has been demonstrated when a-tocopherol is exposed to oxygen radicals generated by irradiation (10) or from the xanthine oxidase system (17), though no quinoid form is obtained after reaction of CYtocopherol with organic peroxide (7). In previous papers (18, 19), it was reported that the phenolic hydroxyl group of n-tocopherol and Trolox C’ is selectively oxidized by peroxidases by way of a oneelectron transfer, and the final oxidation product of Trolox C is Trolox C quinone (18). Thomas and Bielski have delineated the sequence of events after one-electron oxidation of Trolox C by pulse radiolysis (1). They further pointed out that a carboxyl group at the 2-position of
‘Abbreviations used: Trolox C, 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid, Trolox C quinone, 2-hydroxy-2-carboxy-4.(3,5,6trimethylbenzoquinone-2-yl) butane; PMC, 6-hydroxy-2,2,5,7&pentamethylchroman; PMC quinone, 2-hydroxy-2-methyl-4-(3,S$trimethylbenzoquinone-2-yl) butane. 313
314
NAKAMURA
Troiox
AND
C : R=COOH
PMC : R=CH3
0
I
a-Tocopherdquinone I
H
: R=
3
cd Trolox
ti 0
C quinone
PMC quinone
: R=COOH
: R=CH3
SCHEME 1
Trolox C is needed to form the corresponding quinone and that this analogue is not a good model for the study of a-tocopherol. In this study, we further investigated the enzymatic oxidation mechanism of PMC which has a methyl group at the 2-position instead of a carboxyl group (Scheme l), and found that the oxidation rate was higher than that of Trolox C and a-tocopherol. Therefore, it is of considerable interest to determine the final oxidation products of PMC, Trolox C, and a-tocopherol after one-electron oxidation of the phenolic hydroxyl group. Quantitative analysis of the final oxidation products was also conducted by HPLC in the presence or absence of ascorbate. MATERIALS
AND
METHODS
Horseradish peroxidase was purified by the method of Shannon et al. (20) from a crude preparation purchased from Toyobo Co. (Osaka). The
A
HAYASHI
enzyme used was isoenzyme C (A,,,/Azso = 3.4) and the concentration was calculated on the basis of a value for t of 100 mM- ‘cm ’ at 403 nm. Horse heart metmyoglobin was obtained from Sigma. Trolox C and PMC were purchased from Aldrich and Waco Pure Chemicals (Osaka), respectively. Vitamin E (DL-tu-tocopherol) obtained from Sigma was further purified with HPLC before use. The other reagents used were of the highest purity available. Stopped-flow measurements were performed using a Union Giken Model RA-1300 rapid reaction analyzer, and spectral changes were made with a Shimadzu UV300 dual-beam spectrophotometer. ESR spectra were recorded on a Varian E-109B spectrometer equipped with a stoppedflow apparatus. The oxidation products of Trolox C, PMC, and u-tocopherol were analyzed with a Hitachi 655 HPLC system equipped with a Radial Pak 8NVC18 column (8 mm X 10 cm, Waters). Trolox C quinone, PMC quinone, and a-tocopherol quinone were obtained from the reactions of Trolox C, PMC, and a-tocopherol with FeC13 (21). The oxidation products were purified with HPLC and identified as corresponding quinones by NMR and by mass and infrared spectrometry. The reactions were carried out in 0.1 M potassium phosphate (pH 7.4) at 25°C unless otherwise noted.
RESULTS
Oxidations of vitamin E analogues (Trolox C, PMC) and vitamin E (ol-tocopherol) by horseradish peroxidase were followed in the uv region ranging from 350 to 220 nm. Figure 1A shows the increase in absorbance at 270 nm assignable to production of Trolox C quinone (enzyme concentration = 30 nM). Using the extinction coefficient of Trolox C quinone, the molar ratio of quinoid form per added H,Oz was estimated to be 1.0. With PMC, the same spectral changes were observed (enzyme concentration = 5 nM) (Fig. 1B). As can be seen in Fig. lC, only a slight decrease in absorbance at 290 nm was observed during oxidation of n-tocopherol by horseradish peroxidase in alcoholic solvent. The result suggests that the reaction
B
C
8 ZO.O! D 8 :: a
250 Wa”elengthRLtl)
FIG. 1. Spectral changes of Trolox C (A), PMC (B), and n-tocopherol (C) catalyzed by horseradish peroxidase. (A) Concentrations were 100 PM Trolox C, 50 @M H,O*, and 30 nM horseradish peroxidase. Reaction was carried out in 0.1 M phosphate buffer (pH 7.4). (B) Concentrations were 100 FM PMC, 50 FM H,O*, and 5 nM horseradish peroxidase. Reaction was carried out in 0.1 M phosphate buffer (pH 7.4). (C) Concentrations were 20 pM n-tocopherol, 80 pM H,O,, and 30 nM horseradish peroxidase. Reaction was carried out in 20 mM phosphate buffer containing 50% (v/v) methanol with continuous stirring. Using HPLC, the complete recovery of a-tocopherol was confirmed before reaction. The peroxidase activity of the enzyme was reduced by 65% in the alcoholic solvent (19). Spectra were obtained at the times indicated.
OXIDATION
OF VITAMIN
E AND
ITS
ANALOGUES
315
BY PEROXIDASE
B
FIG. 2. Difference spectra observed during the oxidation of Trolox C (Aa) and PMC (Ba) in the absence or presence of ascorbate. The difference spectra were plotted from stopped-flow traces at varying wavelengths at the times indicated. (A) Reactions were carried out in the solution contained 100 FM Trolox C, 12 j&M H,O,, and 1.2 pM horseradish peroxidase in the presence (0) or absence (0) of 100 pM ascorbate. Time courses of stopped-flow traces at 432 nm in the presence or absence of ascorbate are shown in b. (B) Reactions were carried out in solution containing 50 pM PMC, 12 FM H,Os, and 1.2 NM horseradish peroxidase in the presence (0) or absence (0) of 100 pM ascorbate.
gave the oxidation products in low yield. Thomas and Bielski (1) have concluded that a quinoid form could be produced from vitamin E analogues that have a carboxyl moiety at the 2-position. However, the spectral change in PMC in the enzyme reaction mixture suggests, by analogy with the Trolox C-HZO,-peroxidase system (Fig. lA), that the oxidation product of PMC is a quinoid form. From stopped-flow kinetics and ESR experiments it has been inferred that peroxidases catalyze a one-electron oxidation of Trolox C and a-tocopherol (18, 19). Figure 2 shows the difference spectra reconstructed from stopped-flow traces at different wavelengths. The increase in absorbance at 411 nm, which is an isosbestic wavelength between ferric and Compound II of the enzyme, represents the transiently formed intermediate (Fig. 2A). The primary oxidation products of phenolic compounds by horseradish peroxidase have been confirmed as phenoxyl radicals (22-25). Therefore, the oxidation of PMC by the enzyme was performed in the presence of 100 PM ascorbate, since phenoxyl radicals should be scavenged by ascorbate and GSH (18,24-27). The second-order rate constants of the reaction of ascorbate with horseradish peroxidase Compound II and Trolox C radicals are 1.5 X 10’ (25) and 8 X lo6 M-r s-l (8), respectively. Therefore, the results indicate that Compound II of horseradish peroxidase could be obtained in the steady state during the oxidation of PMC (Fig. 2Ba), and the rate constant for the reaction of Compound II with PMC was estimated to be 1.3 X lo6 M-’ so’ in the presence of ascorbate (Fig.
FIG. 3. Spectra of Trolox C and PMC radicals during the oxidation of Trolox C and PMC by horseradish peroxidase. Spectra were plotted from stopped-flow traces at varying wavelengths at the times indicated. (A) Concentrations were 1 mM Trolox C, 50 GM H,O1, and 0.29 +M horseradish peroxidase. (B) Concentrations were 0.5 mM PMC, 75 FM H202, and 0.29 KM horseradish peroxidase. The absorbance changes below 440 nm were corrected by subtracting the absorbance change due to the formation of horseradish peroxidase Compound II.
2Bb). As shown in Figs. 3A and 3B, the spectra in the presence of 0.29 PM enzyme represent Trolox C radicals and PMC radicals, respectively, and are identical to that of Trolox C radicals formed by pulse radiolysis and enzymatic reaction (1,5,18). The results enabled us to follow the PMC radicals at 440 nm under these conditions. The concentrations of PMC radicals in the steady state were proportional to the square root of the PMC concentration (data not shown). Since the concentration of PMC radicals is calculated to be 11 FM (Fig. 4a), the dismutation constant of the radicals is estimated to be 1.6 X lo6 M-’ s-l (5, 18). No radicals were observed in the presence of 100 PM ascorbate (Fig. 4e). To confirm the formation of PMC radicals during the peroxidase reaction, ESR experiments were performed.
Time(s)
FIG. 4. Effect of ascorbate concentrations upon PMC radicals in the steady state during the oxidation of PMC by horseradish peroxidase system. Concentrations were 100 @M PMC, 50 pM Hz02, and 0.29 pM horseradish peroxidase. Reactions were carried out in the presence of 0 (a), 25 (b), 45 (c), 60 (d), and 100 (e) IJM ascorbate.
316
NAKAMURA
AND
HAYASHI
C
a
b
b '40
‘56
'ho
c FIG. 5. ESR spectra observed during the oxidation of Trolox C (A), PMC (B), and cu-tocopherol (C) by horseradish peroxidase system in the absence (a) or presence (b) of ascorbate. (A) Concentrations were 1 mM Trolox C, 0.5 mM H,OZ, and 1.1 pM horseradish peroxidase. Reactions were carried out in the absence (a) or presence (b) of 0.5 mM ascorbate. Instrumental conditions were gain, 2.0 X lo*; power, 20 mW; modulation amplitude, 1 G; time constant, 0.125 s; scan rate, 100 G/min. Spectra were obtained at 30 s after reactions were started. (B) Concentrations were 0.5 mM PMC, 0.5 mM HzOz, and 1.1 pM horseradish peroxidase. Instrumental conditions were the same as in A. (C) Concentrations in the flow cell were 8.0 mM oc-tocopherol, 0.2 mM H,O*, and 9.1 pM horseradish peroxidase. Reactions were carried out in solution containing 50% (v/v) methanol in the absence (a) or presence (b) of 1 mM ascorbate. Instrumental conditions were gain, 5 X 104; power, 63 mW; modulation amplitude, 1 G; time constant, 0.125 s; scan rate, 50 G/min. Spectra were taken during flow (0.2 ml/s).
The seven-line hyperfine pattern characteristic of phenoxyl radicals such as Trolox C and vitamin E radicals formed by peroxidase reactions (Figs. 5Aa and 5Ca) was observed during the reaction (Fig. 5Ba). The addition of ascorbate gave rise to a change of spectra indicating the formation of ascorbate radicals (Figs. 5b). The reaction of phenoxyl radicals with ascorbate has been examined by pulse radiolysis and the reaction rate has been documented (8). Since the rate of the reaction of PMC radicals with ascorbate and the dismutation of PMC radicals are nearly the same, ascorbate scavenged PMC radicals under the experimental conditions. The direct observation of PMC-mediated oxidation of ascorbate was monitored at 284 nm (Fig. 6), which is an isosbestic wavelength between that of PMC and PMC quinone (Fig. 1B). From the initial decrease in absorbance, the reaction rate for a rate-determining step of the reaction is calculated to be 1.2 X lo6 M-l s1 (Table I). The reaction rate was compatible with the rate obtained for the reaction of Compound II with PMC. Ascorbate radicals decay through a dismutation reaction with a value of 1 X lo5 M-’ s-l at pH 7.2 (28-30). These kinetic results supported the accumulation of ascorbate radicals during the PMC-mediated oxidation of ascorbate by horseradish peroxidase (Fig. 5Bb).
Horseradish peroxidase catalyzed the one-electron oxidation of PMC, Trolox C, and a-tocopherol (18, 19). Thomas and Bielski have observed the sequence of reactions after Trolox C radical formation (1). Trolox C radicals undergo dismutation reaction to form a crossconjugated keto diene, which is hydrolyzed to Trolox C quinone. Figure 7 shows the time courses of stopped-flow
E I
FIG. 6. Oxidation of ascorbate catalyzed by PMC-horseradish peroxidaseeHZ02 system. Concentrations were 40 pM PMC, 0.2 mM ascorbate, 20 or 40 pM H202 and 30 nM horseradish peroxidase. The oxidation of ascorbate was followed optically at 284 nm, which is an isosbestic wavelength between PMC and corresponding quinone. The oxidation of ascorbate by Trolox C system was observed similarly.
OXIDATION TABLE
OF VITAMIN
E AND
I
Rate Constants (Mm’ se‘) of the Reaction of Horseradish Peroxidase Compound II with a-Tocopherol and Vitamin E Analogue and Dismutation Constants of Corresponding Radicals
ITS
ANALOGUES BY PEROXIDASE
317
PMC, Trolox C, and cu-tocopherol radicals can be repaired by ascorbate. DISCUSSION
The phenolic hydroxyl groups of Trolox C, PMC, and a-tocopherol are readily oxidized by peroxidases. The rate PMC a-Tocopherol constant of the reaction of peroxidase Compound II with Carboxylic PMC is lOO-fold faster than that with Trolox C. However, Phytyl” acid” Methyl” no clear difference in the steady-state concentration of the radicals is observed at the substrate concentrations Rate constants (M-’ s-‘) used (Figs. 2Aa and 2Ba). The dismutation constants of 5.1 x IOZd 1.1 x 10”’ 1.3 x 10” Stopped-flow 1.5 x 10” c 1.2 x lo6 Overallh Trolox C (1, 18) and PMC radicals are 2.1 X lo4 and 1.6 Dismutation constants of X 10” Mu i s- ‘, respectively. Since these radicals disappear 2.1 x lo4 c 1.6 X 10” the radicals (Mm’ s ‘1 only through the dismutation reaction, the results are explained in terms of the increase in the dismutation con’ A substituent at the 2position. stant of PMC radicals. * The rate constants were estimated from the initial velocities of Trolox C- or PMC-mediated oxidation of ascorbate by peroxidase. Synergistic inhibitory effects of vitamin E and ascor’ Values were given in our previous experiments (18, 19). bate upon lipid peroxidation have been documented (2“The reaction was carried out in the same buffer containing 50% 4). Reactions of peroxyl radicals with vitamin E were ex(v/v) methanol (10). amined by an indirect method (3, 31). Chamulitra and Mason (32) observed that the concentration of peroxyl radicals generated by the lipoxygenase system decreased traces at various wavelengths during the oxidation of in the presence of vitamin E or Trolox C, resulting in the Trolox C by the enzyme. Similar increases in absorbance formation of vitamin E and Trolox C radicals. It has been at 440 and 330 nm indicate the formation of phenoxyl accepted that vitamin E could be regenerated through the radicals. ESR signals were observed to have similar time courses under the same conditions. Time courses at 284 reaction of vitamin E radicals with ascorbate and GSH, and 233 nm, the isosbestic wavelength between Trolox C which are present in vivo at high concentration (33). Several lines of evidence have suggested that hemogloand Trolox C quinone, are quite different. Since it has bin and myoglobin catalyze the peroxidation of unsatubeen shown that a cross-conjugated keto diene compound, rated fatty acids and that the ferry1 form is implicated in an intermediate formed through the dismutation of Trolox the reaction (34-37). The chemical properties of ferrylC radicals, has a strong absorbance at 235-240 nm (l), myoglobin are similar to those of horseradish peroxidase the time course at 233 nm indicates the formation and Compound II, which is an oxidant that reacts with subdecay of the intermediate. The initial lag phase was seen in the time course of the formation of Trolox C quinone at 270 nm, which reached a maximal rate at 13 s after the A start of the reaction. The maximal level in the time course at 233 nm was obtained at almost the same time (Fig. 7Ad). Although there is no conclusive evidence that the observed intermediate is identical to that obtained by pulse radiolysis (l), the results confirmed the assumption that the 233-nm band was a precursor of the final product, 40 Time(s) Trolox C quinone. The final oxidation products of PMC, Trolox C, and cY-tocopherol by horseradish peroxidase were examined with HPLC (Table II). With PMC and Trolox C, the molar ratio of quinoid form to added hydrogen peroxide was nearly 1.0, provided that the reactions were carried out in the presence of a limiting amount of H,02 (25 @I) (Table II). The molar ratio was compatible with that calculated from the increase of absorbance at 270 nm (Figs. FIG. 7. Stopped-flow kinetics for the oxidation products of Trolox C catalyzed by horseradish peroxidase. Concentrations were 250 pM Trolox 1A and 1B) using the molar extinction coefficient of Trolox C quinone (1). No formation of quinoid form was ob- C, 25 WM H202, and 0.29 pM horseradish peroxidase. (A) The stoppedflow traces were observed at 440 (a), 330 (b), 284 (c), and 233 (d) nm. served upon the addition of 100 yM ascorbate. A similar (B) The increase in absorbance at 270 nm indicates the formation of pattern was obtained when horseradish peroxidase was Trolox C quinone (Fig. 1A). The kinetic traces on PMC resembled those replaced by metmyoglobin. The results indicate that on Trolox C and are omitted in the figure. Trolox
C
318
NAKAMURA TABLE
II
Concentrations of Quinoid Form after Oxidations of Vitamin E Analogue and a-Tocopherol by Horseradish Peroxidase and Metmyoglobin in the Presence or Absence of Ascorbate
Oxidation by
HRP 60 nM HRP 60 nM HRP 60 nM HRP 60 nM Mb 1.3 PM Mb 1.3 PM Mb 1.3 /LM Mb 1.3 uM HRP 60 nM HRP 60 nM HRP 60 nM HRP 60 nM Mb 1.3 jtM Mb 1.3 /iM Mb 1.3 /AM Mb 1.3 /iM HRP 1.9 ELM HRP 1.9 PM HRP 1.9 /.tM HRP 1.9 FM Mb 5.2 @M Mb 5.2 /AM Mb 5.2 /LM Mb 5.2 /LM
cu-Tocopherol or vitamin E analogue (PM) Trolox C 50 50 50 50 50 50 50 50 PMC 50 50 50 50 50 50 50 50 ol-Tocopherol 30 30 30 30 30 30 30 30
HA (PM)
Ascorbate (PM)
Quinoid form (PM)
0 25 50 50 0 25 50 50
0 0 0 100 0 0 0 100
0.0 24.5 49.3 0.0 0.0 12.6 24.3 0.0
0 25 50 50 0 25 50 50
0 0 0 100 0 0 0 100
0.0 25.2 50.5 0.0 0.0 13.0 25.5 0.0
0 25 50 50 0 25 50 50
0 0 0 100 0 0 0 100
0.0 0.22 0.40 0.0 0.0 0.22 0.25 0.0
Note. Reactions were carried out in 20 mM phosphate buffer (pH 7.4) containing 50 pM Trolox C, 50 pM PMC, or 30 FM Lu-tocopherol, various concentrations of H,O,, and various concentrations of HRP or metmyoglobin in the presence or absence of 100 FM ascorbate. All reactions were carried out in a single phase and the oxidation products were analyzed by HPLC after incubation for 30 min at 25°C. For the determination of final oxidation products, the following HPLC conditions were used: (A) Trolox C and Trolox C quinone-mobile phase, methanol(pH 2.3) (1:l); flow rate, 2.0 ml/min; detector wave10 mM NH,H*PO, length, 270 nm; retention times of Trolox C and Trolox C quinone, 9.8 and 7.3 min, respectively. (B) PMC and PMC quinone-mobile phase, methanol-water (65:35); flow rate, 2.5 ml/min; detector wavelength, 285 nm; retention times of PMC and PMC quinone, 8.1 and 4.3 min, respectively. (C) n-Tocopherol and n-tocopherol quinone-mobile phase, methanol-water (97:3); flow rate, 3.0 ml/min; detector wavelength, 270 nm; retention times of ru-tocopherol and cu-tocopherol quinone, 6.9 and 4.5 min, respectively.
to yield the ferric form and substrate strate molecules radicals. Therefore, myoglobin and peroxidase share common characteristics with regard to peroxidase activity, although the activity of myoglobin is very low. Trolox C and PMC quinone were detected when incubations were performed in the presence of myoglobin instead of horseradish peroxidase. Formation of the quinoid form depended on the ascorbate concentration. Neither Trolox
AND
HAYASHI
C nor PMC quinone was formed when the reactions were carried out in the presence of 100 yM ascorbate (Table II). When the concentration of ascorbate was decreased to 50 PM, no Trolox C quinone was found but about 20% of the PMC quinone production remained. This observation was anticipated because the dismutation constant of PMC radicals was faster than that of Trolox C radicals. Although metmyoglobin shows a weak ascorbate peroxidase activity, the concentration of ascorbate radicals in the steady state was increased by the addition of Trolox C and PMC (data not shown). Results confirmed that ferrylmyoglobin oxidizes vitamin E analogues by way of a one-electron transfer, causing the phenoxyl radical-mediated oxidation of ascorbate. Ferrylmyoglobin may be formed during ischemia and is implicated in reperfusion injury (38). Ferrylmyoglobin would be derived from the reaction of metmyoglobin and deoxymyoglobin with peroxides. It has been pointed out that oxymyoglobin is oxidized by drugs to metmyoglobin and also autoxidizes to metmyoglobin (34, 38, 39). Ferrylmyoglobin has been observed in perfused rat heart exposed to ethylhydroperoxide (40), and ferrylhemoglobin is also detected in red blood cells (41). No main oxidation products of a-tocopherol after one-electron oxidation by peroxidase and metmyoglobin could be identified (Table II), whereas loss of LYtocopherol was observed with the increase in hydrogen peroxide. The amount of a-tocopherol quinone was slightly increased when the reactions were carried out under acidic pH. Fukuzawa and Gebicki have reported that a-tocopherol quinone was formed from a-tocopherol after reaction with oxygen radicals in acidic solutions, irrespective of the location of the vitamin in model systems (10). Nishikimi et al. (17) observed the intermediate during the reaction of a-tocopherol with superoxide at neutral pH, which was spontaneously converted to its quinoid form considered to be 8a-hydroxytocopherone. The results imply that a substantial amount of tocopherol quinone was obtained in the reaction of a-tocopherol with oxygen radicals. Analysis of the oxidation products of a~tocopherol with various oxidants has suggested that the mechanism of a-tocopherol oxidation varies with the nature of oxidants (16). For the reasons outlined, peroxidase reactions provide a means of overcoming this problem. Since Trolox C, PMC, and n-tocopherol radicals produced by peroxidase reaction are liberated from the enzyme (18), the amount of quinoid formed after a one-electron oxidation of a-tocopherol and its analogues is depended upon the properties of the corresponding phenoxyl radicals. It is important to note that both a-tocopherol radicals and 8a-hydroxychromanone were reduced back to cu-tocopherol by ascorbate (17). The final oxidation products of phenolic compounds such as tyrosine or methoxyphenol after the one-electron oxidation by peroxidase have been identified as dimers (42, 43). The main products of (Ytocopherol, which has methyl groups attached to the aromatic ring, might be dimer products (1, 7). Therefore, it
OXIDATION
OF
VITAMIN
E AND
is reasonable to assume that no main products of a-tocopherol after one-electron oxidation by the enzymes could be detected under the same HPLC conditions (Table II). It has been generally accepted that quinoid forms found in antitumor drugs undergo redox cycling by NAD(P)H-dependent enzymes (44). Tocopherol quinone has been found in animal tissues, although at quite low concentrations; however, it is not clear whether the quinoid form exists as a reduced or an oxidized form in viuo because the reduced form is autoxidizable (15). It has been confirmed that PMC and Trolox C exhibit antioxidant as well as oc-tocopherol activity in homogeneous and liposome systems (32,45). Convincing evidence exists that the phytyl side chain of a-tocopherol at the 2-position facilitates the incorporation and proper retention of the vitamin in biomembranes (46). The present results have shown that one-electron oxidation of the phenolic hydroxyl group initiates the reactions which lead to the formation of PMC and Trolox C quinone, while the phytyl side chain of the vitamin at the 2-position decreases or prevents the reactions. In contrast, a different reaction path has been postulated when n-tocopherol is exposed to superoxide (10, 17). In this path, superoxide may act as an oxidant and, subsequently, a nucleophilic reagent or reductant.
ITS
ANALOGUES
D., and Kelley,
15. Bieri,
J. G., and Tolliver, G. T. (1989)
17. Nishikimi,
M., Yamada, 101-108.
We thank Dr. Richard S. Magliozzo (Albert Einstein College of Medicine of Yeshiva University) for reading the manuscript. This work was supported by a grant from the Suhara Memorial Foundation.
REFERENCES 1. Thomas, M. J., and Bielski, 3315-3319. K. U. (1961)
4. Niki, 5. Bisby,
Chem.
G. W., and Ingold, E. (1987)
J. Am. Chem. Sot. 111,
B. H. J. (1989) Reu. 61,563-589. K. U. (1986)
Act. Chem. Res. 19, 194-
Chem. Phys. Lipids 44, 227-253.
R. H., Ahmed,
S., and Cundall,
R. B. (1984)
Biophys.
Biochem. Biophys. Res. Commun.
8. Davies, M. ,J., Form. 255, 513522.
L. G., and Willson,
9. Packer, J. E., Slater, 737-738.
T. F., and Willson,
K., and Gebicki,
,J. (1983)
C., and
13. Mukai, K., Nishimura, 266, 274-278.
Biochim. Biophys.
K. (1980)
M. (1990)
C. K., Draper,
H. H., Csallany,
A. S., and Chiu,
M. (1967)
22. Yamazaki, I. (1977) Vol. 3, pp. 183-218,
in Free Radicals in Biology (Pryor, Academic Press, New York.
W. A., Ed.),
23. Mason, R. P. (1980) Vol. 5, pp. 161-222,
in Free Radicals in Biology (Pryor, Academic Press, New York.
W. A., Ed.),
24. Ohnishi, T., Yamazaki, zaki, I. (1969) Biochim. 25. Nakamura,
H., Iyanagi,
T., Nakamura,
T., and Yama-
Biophys. Acta 172, 357-369.
M., Yamazaki,
I., Kotani,
T., and Ohtaki,
S. (1985)
J.
T., and Ohtaki,
S. (1989)
J.
Biol. Chem. 260, 13546613552. 26. Nakamura,
M., Yamazaki, I., Kotani, 12909-12913.
Biol. Chem. 264,
27. Ross, D., Norbeck, 15028-15032.
K., and Moldeus,
28. Yamazaki, I., Mason, 235, 244462449. 29. Yamazaki,
P. (1985)
H. S., and Piette,
I., and Piette,
L. (1961)
J. Biol. Chem. 260,
J. Biol. Chem.
L. (1960)
Biochim. Biophys. Acta 50, 62-
69. B. H. ,J., Comstock,
D. A., and Bowen,
J. Am.
R. A. (1971)
R. L. (1988) R. L. (1979)
P. T.
Biochem. J. Nature
278,
Arch. Biochem. Biophys. 226,
Martell,
Y., and Ishizu,
H. A. (1941)
M., and Kikuchi,
K. (1986)
R&l.
J. Am. Chem. Sot. 63, S. (1991)
J. Rid. Chem.
J. Am. Chem. Sot. 103,
31. Burton, G. W., and Ingold, 647226477.
K. U. (1981)
32. Chamulitrat,
R. P. (1989) J. Biol. Chem. 264,20968-
W., and Mason,
20973. 33. Kosower, N. S., and Kosower, E. M. (1976) in Free Biology (Pryor, W. A., Ed.), Vol. 2, pp. 55-84, Academic York.
Radicals in Press, New
S. (1985) Arch. Biochem. Biophys. 237,314-
34. Kanner, 321.
J., and Harel,
35. Kanner,
J., and Harel,
S. (1985)
36. Galaris, D., Mira, D., Sevanian, (1988) Arch. Biochem. Biophys. D., Sevanian,
Lippi& 20, 625-628 A., Cadenas, E., and Hochstein, 262, 221-231.
A., Cadenas, E., and 163-169.
Hochstein,
P.
P. (1990)
Arch. Biochem. Biophys. 281, D., Cadenas,
E., and Hochstein,
P. (1989)
Arch. Biochem.
Biophys. 273, 497-504. 39. Tajima, 12605.
G., and Shikama.
40. Tamura, M., 1019-1031. 41. Giulivi, 19460.
Oshino,
K. (1987)
N., and Chance,
C., and Davies,
K. J. A. (1990)
42. Nonhebel, D. C., and Walton, p. 327, Cambridge Iiniversity 43. Ohtaki,
K., Watanabe, Y., Uemoto, Sot. Jpn. 59, 3113-3116.
12. Golumbic, 127991280.
H., and Yagi,
Lipids 2, 390-396.
38. Galaris,
A. (1985)
7. Suarna, C., Craig, D. C., Cross, K. J., and Southwell-Keely, (1988) J. Org. Chem. 53, 1281L1284.
11. Mukai, Chem.
16, 7777779.
M. (1991)
20. Shannon, 2166-2172.
37. Galaris,
B&hem.
Res. Commun. 119,245-251. 6. Ozawa, T., and Hanaki, 126, 873878.
IO. Fukuzawa, 2422251.
Lipids
Lipids 24, 299-304.
Chem. ?;oc. 93, 5624-5629.
ACKNOWLEDGMENTS
3. Burton, 201.
T. J. (1981)
J. Biochem. 108, 245-249. J. Biochem. 110, 595-597. L. M., Key, E., and Lew, J. Y. (1966) J. Bid. Chem. 241,
18. Nakamura, 19. Nakamura,
21. Chow,
FEBS Lett. 284, 147-151.
P. M. (1991)
16. Vatassery,
Acta 627,
319
PEROXIDASE
14. Njus,
30. Bielski,
2. Ingold,
BY
S., Nakagawa,
J. Biol. Chem. 262, B. (1982)
J. B&hem.
92,
J. Bid. Chem. 265, 19453-
J. C. (1974) in Free-Radical Press, London.
H., Nakamura,
12603-
Chemistry,
M., and Yamazaki,
I. (1982)
J. Biol. Chem. 257,
13398-13403. 44. Powis, G. (1989) Free Radical Bid. Med. 6, 63-101 45. Kagan, V. E., Serbinova, E., and Packer, L. (1990) Arch. Biochem. Biophys. 282, 221-225. 46. Niki, E., Kawakami, Kamiya, Y. (1985)
A., Saito,
M., Yamamoto, Y., Tsuchiya, 2191-2196.
J. Biol. Chcm. 260,
J., and
