Antioxidant properties of 2,3-dimethoxy-5-methyl-6-(10-hydroxydecyl)- 1,4-benzoquinone (idebenone)

Article abstract of DOI:10.1021/tx970136j

Idebenone [2,3-dimethoxy-5-methyl-6-(10-hydroxydecyl)-1,4-benzoquinone] is a synthetic analogue of coenzyme Q that is currently employed in the treatment of vascular and degenerative diseases of the central nervous system. There is some evidence to sugges

Full text of DOI:10.1021/tx970136j

54
Chem. Res. Toxicol. 1998, 11, 54-63
An tioxid a n t P r op er ties of 2,3-Dim eth oxy-5-m eth yl-
6-(10-h yd r oxyd ecyl)-1,4-ben zoqu in on e (Id eben on e)
Alvaro Mordente,*,†,∇ Giuseppe E. Martorana,† Giorgio Minotti,‡,∇ and
Bruno Giardina§
Institutes of Biological Chemistry, Pharmacology, and Chemistry and Clinical Chemistry,
Catholic University School of Medicine, Largo F. Vito 1, 00168 Rome, Italy
Received August 4, 1997X
Idebenone [2,3-dimethoxy-5-methyl-6-(10-hydroxydecyl)-1,4-benzoquinone] is a synthetic
analogue of coenzyme Q that is currently employed in the treatment of vascular and
degenerative diseases of the central nervous system. There is some evidence to suggest that
idebenone might function as an antioxidant; however, it has not been demonstrated whether
this function pertains to the quinone or hydroquinone form of idebenone. Here we demonstrate
that idebenone can scavenge a variety of free radical species, including organic radicals such
as 2,2′-azinobis(3-ethylbenzothiazoline-6-sulfonic acid) and diphenylpicrylhydrazyl, peroxyl and
tyrosyl radicals, and peroxynitrite. Idebenone can also redox couple with hypervalent species
of Mb or Hb, thus preventing lipid peroxidation promoted by these species. Likewise, idebenone
inhibits microsomal lipid peroxidation induced by ADP-iron complexes or organic hydroper-
oxides. In so doing, idebenone prevents the destruction of cytochrome P450, which otherwise
would accompany lipid peroxidation. Irrespective of the experimental system under investiga-
tion, idebenone functions by virtue of the electron-donating properties of the hydroquinone
form. Redox coupling of this hydroquinone with free radicals generates the quinone compound,
which per se lacks antioxidant activity. In many experiments, the antioxidant effects of
idebenone become appreciable at ∼2 µM, which is well in the range of plasma levels attainable
in patients given oral doses of this drug. Moreover, comparative experiments have shown
that the antioxidant efficiency of idebenone varies from no less than 50% to slightly more
than 100% of that of vitamin E or Trolox. We would therefore propose that the neuroprotective
effects of idebenone can be attributed, at least in part, to its ability to function as an antioxidant,
involving redox cycling between hydroquinone and quinone.
idebenone inhibits platelet aggregation (10) and contrib-
utes to the maintenance of vascular wall integrity and
functions (11).
In tr od u ction
Several neurodegenerative disorders are thought to
Idebenone may exist as quinone (IDB1) and hydro-
quinone (IDBH2) forms (Chart 1); however, it has not
been established which form is endowed with antioxidant
activity (7-12). The present work was therefore in-
tended to characterize the antioxidant properties of
IDBH2 versus IDB in a variety of in vitro free radical
systems. We demonstrate that idebenone is a rather
effective antioxidant which stands comparison with
coenzyme Q, vitamin E, or its water-soluble analogue
Trolox. Irrespective of the experimental system under
investigation, idebenone functions by virtue of the electron-
donating properties of IDBH2; IDB per se lacks antioxi-
dant activity and is formed as a byproduct of the redox
coupling between IDBH2 and free radicals. We would
originate from the generation of free radicals in excess
of antioxidant defenses (1, 2). Idebenone [2,3-dimethoxy-
5-methyl-6-(10-hydroxydecyl)-1,4-benzoquinone] is a syn-
thetic analogue of coenzyme Q that is currently employed
to improve cognitive status in patients with clinical
history of stroke, Alzheimer’s disease, and multiinfarct
dementia (3, 4). Despite clinical use, little is known on
the mode of action of idebenone. Some investigators have
suggested that idebenone improves cerebral energy me-
tabolism (4, 5) and stimulates nerve growth factor
synthesis (4, 6). Others (7-11) have ascribed the neu-
roprotective action of idebenone to antioxidant properties;
in fact, idebenone inhibits lipid peroxidation in brain
homogenates (7), mitochondrial membranes (8), and
neural cell lines subjected to oxidative stress (9). Ide-
benone also appears to minimize platelet formation of
thromboxane (10) as well as the toxicity of oxidized low-
density lipoprotein to endothelial cells (11). In so doing,
1
Abbreviations: IDBH2, idebenone hydroquinone; IDB, idebenone
quinone; AAPH, 2,2′-azobis(2-amidinopropane) dihydrochloride; (R)-
PE, (R)-phycoerythrin; PnA, 9,11,13,15-cis,trans,trans,cis-octadecatet-
raenoic acid, cis-parinaric acid; DHR, dihydrorhodamine-123; H2O2,
hydrogen peroxide; ABTS, 2,2′-azinobis(3-ethylbenzothiazoline-6-sul-
fonic acid); ABTS•+, ABTS cation radical; PC, phosphatidylcholine;
* To whom correspondence should be addressed. Phone: 06-3015/
5135. Fax: 06-3053598. E-mail: ibicb@rm.unicatt.it.
† Institute of Biological Chemistry.
DTPA, diethylenetriaminepentaacetic acid; DPPH, 1,1-diphenyl-2-
•
picrylhydrazyl; O2•-, superoxide anion; MbIV, ferrylmyoglobin; MbIV
,
ferrylmyoglobin radical; MbIII, metmyoglobin; MbIIO2, oxymyoglobin;
TBARS, thiobarbituric acid-reactive substances; •NO, nitric oxide;
ONOO-, peroxynitrite; SIN-1, 3-morpholinosydnonimine; CoQ10H2,
coenzyme Q10 hydroquinone; CUOOH, cumene hydroperoxide; LMW,
low molecular weight.
‡ Institute of Pharmacology.
§ Institute of Chemistry and Clinical Chemistry.
∇ A.M. and G.M. contributed equally to this paper.
X Abstract published in Advance ACS Abstracts, December 15, 1997.
S0893-228x(97)00136-7 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/19/1998

Idebenone Antioxidant Properties
Chem. Res. Toxicol., Vol. 11, No. 1, 1998 55
was generated by reacting this probe with MbIV (16). Briefly,
ABTS (150 µM) was reconstituted with MbIII (2.5 µM) in
standard buffer, at 20 °C, in the presence or absence of
antioxidants. Hydrogen peroxide (75 µM) was subsequently
included to convert MbIII to MbIV, and the formation of ABTS•+
was monitored spectrophotometrically as the absorbance in-
crease at 734 nm. The lag time from H2O2 addition to absor-
bance increase was used to calculate the antioxidant efficiency
of test compounds. In a separate set of experiments ABTS•+
was generated by electrolysis as described by Barr and Aust
(17), using a PS500XT power supply (Hoefer Scientific Instru-
ments, San Francisco, CA). ABTS•+ was stable at neutral pH,
and its concentration after electrolysis was determined by
assuming ꢀ660nm ) 12 mM-1 cm-1 (17).
Ch a r t 1. Str u ctu r es of Qu in on e (IDB) a n d
Hyd r oqu in on e (IDBH2) Id eben on e
2. P er oxyl R a d ica l Sca ven gin g. Peroxyl radicals were
generated in aqueous environments by thermal decomposition
of AAPH (18). The peroxyl radical-scavenging activity of
idebenone was evaluated from its competition with hydrophilic
or hydrophobic probes such as (R)-PE (19) and crocin (14) or
PnA (20), respectively. In a first system, 1.2 nM (R)-PE was
equilibrated for 5 min at 37 °C and subsequently reacted with
4 mM AAPH. Peroxyl radical reactivity was then determined
from the loss of (R)-PE fluorescence (emission at 575 nm,
excitation at 495 nm). Where indicated, test compounds (or
equivalent volumes of ethanol-HCl) were included 20 min after
AAPH addition, and antioxidant efficiencies were determined
from their ability to prevent fluorescence decay. In a second
system, (R)-PE was replaced with crocin, which is bleached by
AAPH-derived radicals. Briefly, the reaction mixture contained
crocin (12 µM) and increasing amounts of antioxidants or
equivalent volumes of ethanol-HCl in standard buffer, 40 °C.
The reaction was started by the addition of 5 mM AAPH, and
the bleaching of crocin was monitored as the decrease in
absorbance at 440 nm (ꢀ ) 133 mM-1 cm-1) (14). A linear
bleaching rate was detected approximately 30 s after the
addition of AAPH and lasted several minutes. Experimental
data were eventually plotted to fit the following competition
kinetic equation:
therefore propose that the neuroprotective effects of
idebenone can be attributed, at least in part, to its ability
to function as an antioxidant, involving redox cycling
between IDBH2 and IDB.
Ma ter ia ls a n d Meth od s
Ch em ica ls. Idebenone quinone (IDB) was kindly provided
by Takeda Pharmaceutical Co. (Tokyo, J apan). IDBH2 (ꢀ290nm
)
3.94 mM-1 cm-1) was prepared by chemical reduction of IDB
(ꢀ275nm ) 14 mM-1 cm-1 in absolute ethanol), essentially as
described by Rieske for short-chain coenzyme Q analogues (13).
Purity was checked by reverse-phase HPLC. Stock solutions
of IDBH2 were stored in the dark at -20 °C, in absolute ethanol
under acidic conditions (10 mM HCl).
(R)-Phycoerythrin [(R)-PE] and 2,2′-azobis(2-amidinopropane)
dihydrochloride (AAPH) were purchased from Polysciences Inc.
(Warrington, PA).
Ca u t ion : AAPH is toxic and is an irritant. Use care in
handling.
Vc/Va ) 1 + Ka/Kc[A]/[C]
cis-Parinaric acid (PnA) and dihydrorhodamine-123 (DHR)
were obtained from Molecular Probes (Eugene, OR). Hydrogen
peroxide (H2O2), 2,2′-azinobis(3-ethylbenzothiazoline-6-sulfonic
acid) (ABTS), and Trolox were from Aldrich Chemical Co.
(Milwaukee, WI). Beef liver catalase (E.C. 1.11.1.6) was from
Boehringer Mannheim (Germany) and was made free of thymol
by ultrafiltration in Diaflo YM100 membrane (Amicon, Beverly,
MA). Saffron, egg yolk phosphatidylcholine (PC), cumene
hydroperoxide (CUOOH), arachidonic acid, horse heart myo-
globin (type III, 99% purity), and bovine hemoglobin were from
Sigma Chemical Co. (St. Louis, MO). Crocin was isolated with
methanol from saffron, after treatment with ethyl ether to
remove interfering substances (14). The concentration of crocin
was determined by assuming ꢀ443nm ) 133 mM-1 cm-1 (14). All
solutions were prepared in 18 MΩ double-distilled deionized
water (Milli-Q, Millipore Co., Bedford, MA). Unless otherwise
indicated, the reactions were carried out in a standard buffer
composed of 50 mM sodium phosphate buffer, pH 7.4. Phos-
phate buffer was routinely passed through Chelex-100 (Bio-Rad)
to minimize metal contamination. DTPA (0.1 mM) was also
included in the standard buffer to prevent possible interferences
from residual trace metals.
where Vc is the crocin bleaching rate in the absence of antioxi-
dant, Va is the crocin bleaching rate in the presence of antioxi-
dant, Ka is the rate constant for the interaction of free radicals
with the antioxidant, Kc is the rate constant for the reaction
between free radicals and crocin, [A] is the antioxidant concen-
tration, and [C] is the crocin concentration. The Ka/Kc slope,
calculated from the linear regression of the plot of [A]/[C] vs
Vc/Va, indicates the relative capacity of antioxidants to interact
with peroxyl radicals.
In a third system, idebenone was incorporated in PC lipo-
somes which also contained PnA, a fluorescent polyunsaturated
fatty acid which is suitable to monitor the early phases of lipid
peroxidation in liposomal vesicles (20), ghosts, and intact
erythrocytes (21). The preparation of PC liposomes and the
incorporation of PnA, idebenone, or other antioxidants were as
described previously (22). Typical incubations (3 mL final
volume) consisted of 200 µM PC liposomes containing 1.5 µM
PnA (corresponding to a 0.75% molar ratio to phospholipids)
and increasing concentrations of antioxidants in standard buffer,
37 °C. Lipid peroxidation was initiated with 10 mM AAPH, and
the oxidative modifications of PnA were monitored as fluores-
cence decay (excitation at 324 nm, emission at 413 nm).
Assa ys. 1. DP P H a n d ABTS Ra d ica l Sca ven gin g. The
stable nitrogen-centered free radical 1,1-diphenyl-2-picrylhy-
drazyl (DPPH) was prepared by mixing 1 mL of standard buffer
with 1 mL of absolute ethanol and 0.5 mL of ethanol-dissolved
500 µM DPPH, at 25 °C (15). Following stabilization of
absorbance at 516 nm, antioxidants were included and the
concentrations inducing a decrease of 0.20 absorbance unit/10
min (IC0.20) were assumed as biochemical indices of free radical-
3. Stu d ies w ith Tyr osyl Ra d ica ls. UV irradiation of
tyrosine solutions generates tyrosyl radicals which subsequently
recombine to fluorescent dityrosines (23). L-Tyrosine (400 µM)
was incubated in standard buffer at 25 °C in the presence or
absence of idebenone. Superoxide dismutase (70 units mL-1
)
and catalase (500 units mL-1) were included to remove O2
•-
•-
scavenging efficiency. The monocation radical of ABTS (ABTS•+
)
formed during irradiation and H2O2 formed via O2 dismuta-

56 Chem. Res. Toxicol., Vol. 11, No. 1, 1998
Mordente et al.
tion, respectively. UV irradiation was carried out at 275 nm,
and dityrosine fluorescence emission was recorded at 405 nm.
The scavenging of tyrosyl radicals by test compounds was
inferred from the delay in the fluorescence buildup.
4. In ter a ction s w ith Hyp er va len t Hem ep r otein s: Sp ec-
tr a l Stu d ies a n d Lip id P er oxid a tion Exp er im en ts. The
formation of MbIV or HbIV was monitored in incubations (2 mL
final volume) containing MbIII or HbIII (both adjusted to give 50
µM heme) and H2O2 (150 µM) (24). MbIV concentration was
determined by measuring absorbance at 550 and 630 nm,
according to the following formula: [MbIV] ( µM) ) [249 ×
absorbance at 550 nm] - [367 × absorbance at 630 nm] (25).
After idebenone addition the MbIV-to-MbIII transition was
monitored as the absorbance decrease at 550 nm, that is the
wavelength at which the absorption spectra of MbIII and MbIV
differ most (24). Concurrent oxidation of IDBH2 was monitored
at 275 nm by assuming ꢀ ) 12.5 mM-1 cm-1 for [IDB] minus
[IDBH2]. At this wavelength the spectral difference between
MbIII and MbIV is minimal and does not interfere with IDB
determination. In all experiments the addition of idebenone was
routinely preceded by that of catalase (500 units mL-1). This
was intended to decompose unreacted H2O2 which otherwise
would convert MbIII back to MbIV, altering the apparent stoi-
chiometry of MbIV reduction and idebenone oxidation.
Ferrylmyoglobin- or ferrylhemoglobin-dependent lipid per-
oxidation was studied in incubations containing arachidonic acid
vesicles (0.4 mM) and MbIII or HbIII (50 µM) in standard buffer
at 37 °C. Reactions were started by adding H2O2 (150 µM), and
lipid peroxidation was assayed as the formation of thiobarbituric
acid-reactive substances (TBARS) (26), with modifications
involving butanol extraction of the TBA adduct (27). In other
experiments, lipid peroxidation was monitored as the fluores-
cence decay of PC liposome-incorporated PnA. Briefly, a small
volume of an ethanolic solution of PnA (final concentration 1
µM, corresponding to a 0.5% molar ratio to phospholipids) was
injected into reaction mixtures (3 mL final volume) containing
200 µM phospolipid in standard buffer at 20 °C. Lipid peroxi-
dation was initiated by adding MbIII (2 µM) and H2O2 (5 µM).
Where indicated, the addition of MbIII and H2O2 was preceded
by that of antioxidants. The loss of PnA fluorescence was
monitored with excitation at 312 nm and emission at 455 nm.
F igu r e 1. Interaction of idebenone with the ABTS radical.
Incubations were prepared as described under Materials and
Methods. Panel A shows the time-dependent formation of the
ABTS radical in the absence (‚‚‚) or presence of IDBH2 (s) or
Trolox (- -). In panel B the lag phases preceding radical
detection were plotted as a function of IDBH2 (9) or Trolox (0)
concentration.
between iron and either phosphate or DTPA present in the
standard buffer; similar ligand interactions may occur with
several other buffers (29). Aliquots of reaction mixtures were
taken at the indicated times and assayed separately for TBARS
formation or cytochrome P450 destruction. Although unbuf-
fered, the pH of reaction mixtures did not vary throughout the
experiment time.
Unless otherwise indicated all values are given as means (
SE. Other experimental conditions are given in the legends to
figures and tables.
5. Stu d ies w ith P er oxyn itr ite. Peroxynitrite (ONOO-) is
formed by reaction of superoxide anion (O2•-) with nitric oxide
(•NO). In this study ONOO- was generated by the spontaneous
decomposition of 3-morpholinosydnonimine (SIN-1) at neutral
•-
•
pH, yielding O2 and NO. ONOO- formation was monitored
spectrophotometrically by taking advantage of its ability to
oxidize DHR to rhodamine, having a high molar extinction
coefficient at 500 nm (74 500 M-1 cm-1) (28). Neither O2•- nor
•
H2O2 or NO per se can oxidize DHR, thus making this assay
specific for ONOO- (28). The reagent mixture (2 mL final
volume) consisted of DHR (50 µM) and SIN-1 (1 mM) in
standard buffer, 30 °C. Antioxidants or equivalent volumes of
ethanol-HCl were added 15 min after SIN-1, i.e., when the rate
of ONOO- formation was constant.
Resu lts
In ter a ction of Id eben on e w ith DP P H a n d ABTS
Ra d ica ls. 1,1-Diphenyl-2-picrylhydrazyl (DPPH) is a
stable nitrogen-centered free radical which may convert
into 1,1-diphenyl-2-picrylhydrazine, a diamagnetic mol-
ecule, upon one-electron reduction. DPPH has a strong
absorption at 516 nm which decreases after the addition
of reducing equivalents (15). Experiments with 2.5-50
µM IDBH2 showed an instantaneous and dose-dependent
decrease of DPPH absorbance, whereas similar concen-
trations of IDB were virtually ineffective. In this system,
the efficiency of IDBH2 was very similar to that of Trolox
(IC0.200 ) 12.02 ( 0.54 vs 11.98 ( 0.71 µM; n ) 7).
Furthermore, the efficiency of IDBH2 was comparable to
that of other antioxidants, such as R-tocopherol or
butylated hydroxytoluene (IC0.200 ) 12.77 ( 0.81 and
11.35 ( 0.5 µM, respectively) (34).
6. ADP -Ir on -Dep en d en t Micr osom a l Lip id P er oxid a -
tion a n d Cytoch r om e P 450 Destr u ction . Microsomes were
isolated from the livers of male Wistar rats (∼200 g body wt)
which had been given phenobarbital intragastrically for 3 days
at a daily dose of 20 mg/kg body wt. Phenobarbital treatment
had the dual advantage of inducing cytochrome P450 (a major
catalyst of microsomal lipid peroxidation) and depleting a
membrane pool of non-heme iron which participates in free
radical reactions and potentially interferes with exogenously
added iron or other pro-oxidant agents (29, 30). Microsomes
were isolated as described previously (31) and assayed for
proteins and cytochrome P450 by established procedures (32,
33). Lipid peroxidation was studied by reconstituting mi-
crosomes (1 mg of protein/mL) with ADP-Fe(II) (1 mM chelator:
0.1 mM FeSO4) or CUOOH (0.25 mM). These experiments were
performed in 0.3 M NaCl, carefully adjusted to pH 7.0 just prior
to use. This was done to avoid ligand-catalyzed interaction
As shown in Figure 1A (dotted line), the inclusion of
H2O2 in a mixture containing MbIII and ABTS produced

Idebenone Antioxidant Properties
Chem. Res. Toxicol., Vol. 11, No. 1, 1998 57
the lag phase during which (R)-PE was protected from
oxidation, we determined that one IDBH2 could trap two
peroxyl radicals (1.94 ( 0.07, n ) 7). Under the same
experimental conditions, a similar stoichiometry (2.06 (
0.08, n ) 7) was obtained with Trolox (Figure 2B, open
symbols, dotted line). In this assay neither IDB nor an
equivalent volume of ethanol-HCl could affect the AAPH-
dependent decay of (R)-PE fluorescence (see also Figure
2A). Lipophilic antioxidants, such as CoQ10H2 and R-to-
copherol, were similarly ineffective, as already reported
(35).
The ability of IDBH2 to scavenge AAPH-derived per-
oxyl radicals was further assessed by the crocin bleaching
test. This assay has the advantage of providing some
indirect information on the nature of byproducts eventu-
ally formed through reaction of antioxidants with peroxyl
radicals. In fact, the bleaching rate can paradoxically
increase if antioxidant generates byproducts that are
themselves capable to react with crocin, as demonstrated
with Ebselen (36). By contrast, the bleaching rate will
decrease if antioxidant byproducts remain unreactive or
contribute to intercept peroxyl radicals, as demonstrated
with Trolox (36). In preliminary experiments we found
that 6.25 or 12.5 µM IDBH2 decreased the rate of peroxyl
radical-dependent crocin bleaching from 0.316 to 0.236
or 0.183 nmol/(mL‚min)-1, respectively. The reaction was
then investigated within a broad range of IDBH2 con-
centrations (2.5-25 µM), and the bleaching rates were
plotted according to the competition kinetic equation
described under Materials and Methods. A Ka/Kc ratio
was calculated and used as an index of the efficiency with
which antioxidants reacted with peroxyl radicals. Under
these conditions, the Ka/Kc for IDBH2 was virtually
identical with that calculated for Trolox (0.753 ( 0.045
vs 0.788 ( 0.055, respectively; n ) 6), suggesting a
comparable efficiency in peroxyl radical scavenging.
The above-described experiments focused on the ability
of IDBH2 to protect hydrophilic probes such as (R)-PE
and crocin from peroxyl radicals generated by thermal
decomposition of AAPH. In a subsequent set of experi-
ments both the oxidizable probes and IDBH2 were
incorporated in a lipid environment. For this purpose,
we used PnA, a polyunsaturated fatty acid which be-
comes fluorescent upon incorporation in lipid bilayers.
The decay of PnA fluorescence is therefore indicative of
peroxidative processes within lipid environments. As
shown in Figure 3A, the fluorescence of PnA incorporated
in PC liposomes decreased linearly upon exposure to
AAPH. Figure 3A also shows that simultaneous incor-
poration of an increasing amount of IDBH2 in PC
liposomes prevented the decay of PnA fluorescence,
suggesting that IDBH2 protected PnA from peroxyl
radicals. To permit adequate evaluation of the antioxi-
dant activity of IDBH2 in lipid environments, compara-
tive experiments were performed with PC liposomes
fortified with vitamin E. These experiments showed that
IDBH2 was slightly but significantly more effective than
R-tocopherol in protecting PnA from AAPH-dependent
oxidation (Figure 3B).
F igu r e 2. Effect of idebenone on peroxyl radical-dependent
oxidations: studies with (R)-PE. Peroxyl radicals were gener-
ated in incubations containing AAPH, as described under
Materials and Methods. Panel A shows the effect of IDB or
ethanol-HCl (‚‚‚) versus increasing concentrations of IDBH2 (s)
on the fluorescence decay of (R)-PE (1.2 nM) induced by AAPH
(4 mM). In panel B the lag phases preceding the fluorescence
decay were plotted as a function of IDBH2 (9) or Trolox (0)
concentration.
an instantaneous and linear increase of absorbance at
734 nm, indicative of the formation of ABTS•+. In the
presence of IDBH2 (solid line), the absorbance increase
was delayed by a lag phase that increased linearly with
the concentration of this drug (Figure 1). By calculating
the slope values of straight lines obtained by plotting lag
phases versus antioxidant concentrations, we could de-
termine that IDBH2 was 50% as efficient as Trolox in
trapping ABTS•+ (Figure 1B). In selected experiments,
ABTS•+ was generated electrochemically and the decay
in absorbance at 660 nm was monitored in the presence
or absence of IDBH2 or Trolox. Under these conditions,
IDBH2 quenched ABTS•+ with a 50% efficiency as com-
pared with Trolox, thus confirming the experiments in
which this species was generated by ferrylmyoglobin.
Neither IDB nor ethanol-HCl was found to scavenge
ABTS•+ (see also Figure 1A).
In ter a ction of Id eben on e w ith P er oxyl Ra d ica ls.
AAPH is a water-soluble azo compound which produces
a constant flow of peroxyl radicals upon thermal decom-
position. As shown in Figure 2A, incubation of (R)-PE
with AAPH resulted in a linear decrease of its fluores-
cence emission (dotted line). When IDBH2 was included
20 min after AAPH addition, the decay in fluorescence
stopped and then resumed after a lag phase which
increased with the concentration of IDBH2 (see Figure
2A, solid lines, and B, closed symbols, solid line). Con-
sidering the rate constant of peroxyl radical generation
(1.1 × 10-6 [AAPH] s-1 at 37 °C) (18) and the length of
In ter a ction of Id eben on e w ith Tyr osyl Ra d ica l.
The tyrosyl radical-scavenging activity of idebenone was
studied in a model system relying on UV irradiation of
aqueous tyrosine solutions, yielding tyrosyl radicals
which rearrange to fluorescent dityrosines (23). In this
system, compounds that donate electrons to tyrosyl
radicals suppress or delay the formation of fluorescent

58 Chem. Res. Toxicol., Vol. 11, No. 1, 1998
Mordente et al.
F igu r e 4. Effect of idebenone on dityrosine fluorescence
buildup. Incubations were prepared as described under Materi-
als and Methods. Panel A shows the dityrosine fluorescence
buildup in the absence (‚‚‚) or presence of 1, 2.5, 5, or 10 µM
IDBH2 (s). In panel B the lag phases preceding the buildup of
fluorescence were plotted as a function of IDBH2 (9) or Trolox
(0) concentration.
F igu r e 3. Effect of idebenone on peroxyl radical-dependent
oxidations: studies with PnA. Incubations were prepared as
described under Materials and Methods. Panel A shows the
AAPH-dependent decrease in fluorescence of liposome-incorpo-
rated Pna in the absence (‚‚‚) or presence of 1, 2.5, or 10 µM
IDBH2 (s). In panel B the inhibition of PnA fluorescence decay
was plotted as a function of IDBH2 (9) or R-tocopherol (O)
concentration.
line) with the exception that the 632-nm peak was shifted
to 622 nm. This spectral shift has already been observed
upon reduction of MbIV by other electron donors (24, 25,
37). Under the same experimental conditions, neither
IDB nor ethanol-HCl was able to reduce MbIV (not
shown).
dityrosines. As shown in Figure 4A (dotted line), UV
irradiation of tyrosine produced an immediate and linear
increase of fluorescence emission. In the presence of
IDBH2 (solid line) the fluorescence buildup was delayed
by the appearance of a lag phase that increased with the
concentration of the antioxidant (Figure 4). The slope
values of the corresponding straight lines showed that
IDBH2 was 50% as efficient as Trolox in scavenging
tyrosyl radical. Neither IDB nor ethanol-HCl was found
to inhibit dityrosine formation (see also Figure 4A).
In ter a ction of Id eben on e w ith F er r ylh em op r o-
tein s. In the presence of H2O2, MbIII and HbIII undergo
a two-electron oxidation process yielding a transient
protein radical plus a long-lived oxoferryl moiety (FeIVdO),
which can be detected spectrophotometrically as changes
in the visible and Soret regions. As shown in Figure 5A,
the spectrum of MbIII (dashed line) was characterized by
absorption peaks at 502, 582, and 632 nm. The addition
of a 3-fold excess of H2O2 (Figure 5A, dotted line) induced
a rapid absorbance increase in the 520-600-nm region
with distinct peaks at 546 and 586 nm, plus disappear-
ance of the peak at 632 nm. These spectral changes have
Spectra recorded in the Soret region confirmed the data
obtained in the visible region. The spectrum of MbIII in
the Soret region was characterized by an absorption peak
at 408 nm (Figure 5A inset, dashed line). The addition
of a 3-fold excess of H2O2 induced a shift from 408 to 420
nm (Figure 5A inset, dotted line), which is distinctive of
MbIV (37). After catalase removal of unreacted H2O2,
reconstitution of IDBH2 with MbIV shifted the peak at
420 nm back to 408 nm (see also Figure 5A inset, solid
line), confirming the reductive action of IDBH2 on MbIV.
The MbIV-to-MbIII transition was also monitored as the
decrease in absorbance at 550 nm. As shown in Figure
6A neither IDB nor an equivalent volume of ethanol-HCl
could reduce MbIV. By contrast, IDBH2 irreversibly and
concentration-dependently reduced MbIV to MbIII (Figure
6). Moreover, the initial rate of MbIII formation was
directly proportional to the concentration of IDBH2, with
a linear relationship in the 1-10 µM range (Figure 6B
inset). The apparent second-order rate constant for the
reaction between IDBH2 and MbIV, as determined from
the linear portion of the plot d[MbIII]/dt vs [IDBH2], was
3.78 ( 0.09 × 10-4 M-1 s-1 (n ) 6). This value is close to
that reported for short-chain CoQ analogues such as
CoQ1H2 and CoQ2H2 (24). The reduction of MbIV to MbIII
•
been attributed to the formation of MbIV or an MbIV/MbIV
composite (25, 37). After catalase removal of unreacted
H2O2, the addition of IDBH2 caused spectral changes
consisting of disappearance of the peaks at 546 and 586
nm plus reappearance of the peaks at 502 and 582 nm
(Figure 5A, solid line). This spectral pattern was there-
fore similar to that of native MbIII (cf. Figure 5A, dashed

Idebenone Antioxidant Properties
Chem. Res. Toxicol., Vol. 11, No. 1, 1998 59
F igu r e 5. Reduction of MbIV by idebenone. Ferrylmyoglobin
(‚‚‚) was produced by reacting 50 µM MbIII (- -) with 150 µM
H2O2. Following the addition of catalase (500 units mL-1) to
remove unreacted H2O2, 50 µM IDBH2 was included and
repetitive scans (s) were taken every 1 min. The inset shows
the spectral changes in the Soret region under comparable
conditions, with the exception that MbIII, H2O2, and IDBH2
concentrations were lowered to 10, 30, and 10 µM, respectively.
F igu r e 6. Idebenone-dependent reduction of MbIV. Ferrylmyo-
globin was formed by reacting MbIII with H2O2 as described in
Figure 5. The reduction of MbIV to MbIII by IDBH2 versus IDB
or ethanol-HCl was monitored as the decrease in absorbance
at 550 nm. Panel A shows a time course experiment. Panel B
summarizes the dependence of MbIII formation versus idebenone
concentration.
was paralleled by the oxidation of IDBH2 to IDB, and
the stoichiometry of idebenone oxidation versus MbIV
reduction (as determined from d[MbIII]/dt vs d[oxidized
idebenone]/dt) averaged unity (1.03 ( 0.04, n ) 6).
Comparative experiments showed that IDBH2 could also
reduce HbIV. In fact, 5 min reaction of 100 µM IDBH2
with 50 µM Hb gave a mixture composed of HbIV (4 µM),
HbIII (38 µM), and HbIIO2 (8 µM).
The hypervalent states of hemoproteins are strong
oxidants that promote oxidation, peroxidation, and ep-
oxidation of various biomolecules in vitro (37). In par-
ticular, the oxoferryl moiety of MbIV, but not the protein
radical, has been found to promote the peroxidation of
fatty acids, membranes, and lipoproteins (37). Antioxi-
dants that donate electron to the oxoferryl moiety will
therefore protect against oxidative damage. This has
been demonstrated with Trolox, ascorbate, ergothioneine,
urate, thiols, and CoQ1H2 or CoQ2H2 (24, 25, 37). As
shown in Table 1, the incubation of arachidonic acid with
a mixture of MbIII plus H2O2 yielded TBARS indicative
of peroxidation. The addition of IDBH2 decreased lipid
peroxidation close to the values observed with either MbIII
or H2O2 added individually, in keeping with the ability
of IDBH2 to compete with lipids for the oxidizing equiva-
lents of H2O2-activated Mb. Fully comparable results
were obtained in experiments containing HbIII in place
of MbIII (see also Table 1).
Ta ble 1. Effects of Id eben on e on MbIV- or
HbIV-Dep en d en t Lip id P er oxid a tion
TBARS
addition
(nmol/mL)a
MbIII + arachidonic acidb
0.34 ( 0.01
0.19 ( 0.01
2.96 ( 0.17
H2O2 + arachidonic acidb
MbIII + H2O2 + arachidonic acidc
MbIII + H2O2 + IDBH2 (100 µM) + arachidonic acidc 0.44 ( 0.01
MbIII + H2O2 + IDB (100 µM) + arachidonic acidc
HbIII + arachidonic acidb
2.92 ( 0.25
0.43 ( 0.01
2.92 ( 0.24
HbIII + H2O2 + arachidonic acidc
HbIII + H2O2 + IDBH2 (100 µM) + arachidonic acidc 0.47 ( 0.01
HbIII + H2O2 + IDB (100 µM) + arachidonic acidc
2.86 ( 0.25
a
Lipid peroxidation products were determined as TBARS versus
reagent blanks lacking MbIII or HbIII. Values are means ( SE of
three separate determinations in duplicate. b Arachidonic acid (0.4
mM) was incubated for 30 min at 37 °C with either horse heart
MbIII (50 µM), bovine HbIII (50 µM), or H2O2 (150 µM). c Horse
heart MbIII (50 µM) or bovine HbIII (50 µM) was incubated with
H2O2 (150 µM) for 5 min at 20 °C. After removal of excess of H2O2
by catalase (500 units mL-1), IDBH2 or IDB was included.
Arachidonic acid (0.4 mM) was eventually added, and reaction
mixtures were incubated for 30 min at 37 °C.
H2O2 apparently accompanied with a slow decay of PnA
fluorescence; however, control experiments showed that
such decay reflected a spontaneous photobleaching of
PnA, inasmuch as it could be observed even in the
absence of MbIII or H2O2. When both MbIII and H2O2 were
included in the reaction mixtures, a rapid and extensive
fluorescence decay was observed. The addition of IDBH2
concentration-dependently prevented the decay, consis-
tent with the ability of this antioxidant to reduce the
Further evidence for a redox coupling of IDBH2 with
H2O2-activated hemoproteins was obtained in a recon-
stituted system involving PnA incorporated in PC lipo-
somes. As shown in Figure 7A, the addition of MbIII or

60 Chem. Res. Toxicol., Vol. 11, No. 1, 1998
Mordente et al.
F igu r e 7. Effect of idebenone on MbIV-dependent oxidation of
PnA incorporated in liposomes. Ferrylmyoglobin was formed by
reacting MbIII (2 µM) with H2O2 (5 µM). The fluorescence decay
of liposome-incorporated PnA was monitored in the presence of
1-3 µM IDBH2 (s) and either 3 µM IDB or an equivalent
volume of ethanol-HCl (‚‚‚), as described under Materials and
Methods. The upper trace shows that MbIII or H2O2 per se had
no effect on the decay of PnA fluorescence.
F igu r e 8. Effect of idebenone on ONOO--dependent oxidation
of DHR. Incubations were prepared as described under Materi-
als and Methods. Panel A shows the time-dependent oxidation
of DHR by SIN-1 in the presence of 10-20 µM IDBH2 (s) versus
20 µM Trolox (-‚-) and 20 µM IDB or an equivalent volume of
ethanol-HCl. In panel B the lag phases preceding the resump-
tion of absorbance were plotted as a function of IDBH2 (9) or
Trolox (0) concentration.
hypervalent species of myoglobin before it could attack
PnA. Once again, neither IDB nor ethanol-HCl could
prevent the decay of PnA fluorescence.
general agreement that ADP-Fe(II) must first oxidize
with molecular oxygen (40). In our experiments, a
chelate made of 500 µM ADP and 50 µM FeSO4 was found
to oxidize with oxygen at a linear rate of 2.432 ( 0.071
nmol/(mL‚min)-1, as measured by monitoring the disap-
pearance of bathophenanthroline-chelatable Fe(II) in an
inert medium like NaCl (41). This rate was not ap-
preciably affected by 100 µM IDB [2.469 ( 0.038 nmol/
(mL‚min)-1]; however, it was effectively decreased by
equimolar IDBH2 [0.108 ( 0.055 nmol/(mL‚min)-1]. In-
cubation of rat liver microsomes with ADP-Fe(II) re-
sulted in extensive formation of TBARS, indicative of
lipid peroxidation. This reaction was nearly abolished
by IDBH2 but not by IDB, in keeping with the different
effects they had on ADP-Fe(II) oxidation (Table 2).
Lipid hydroperoxides generated upon ADP-iron-de-
pendent initiation of peroxidative processes are known
to react with cytochrome P450. This process accompanies
destruction of the hemoprotein and liberates lipid alkoxyl
radicals which reinitiate hydrogen abstraction from
polyunsaturated fatty acid, thus propagating lipid per-
oxidation (29, 42). In this respect, Table 2 shows that
(i) ADP-Fe(II)-dependent microsomal lipid peroxidation
predictably caused the destruction of cytochrome P450
and (ii) the addition of IDBH2 prevented cytochrome P450
destruction, although less effectively as compared with
the concomitant inhibition of TBARS formation (75% vs
98%). The latter observation suggested that IDBH2
uncoupled the destruction of cytochrome P450 from the
propagation of lipid peroxidation. Inasmuch as propaga-
In ter a ction of Id eben on e w ith P er oxyn itr ite.
Peroxynitrite is the product of the nearly diffusion-
limited reaction between O2•- and NO. Peroxynitrite is
•
a powerful oxidant that damages many cellular compo-
nents (38, 39) and experimental probes such as DHR (28).
As shown in Figure 8A (dotted line), the oxidation of DHR
by ONOO- was a time-dependent process that ap-
proached zeroth-order kinetics after ∼10 min and pro-
ceeded for ∼30 min. The addition of IDBH2 15 min after
SIN-1 caused an instantaneous inhibition of DHR oxida-
tion, which resumed after a lag phase that was strictly
dependent on the antioxidant concentration (see also
Figure 8A, solid line). DHR oxidation was similarly
inhibited by Trolox, with the usual appearance of a lag
phase that increased with increasing the concentration
of the antioxidant (Figure 8). The slope values of the
straight lines obtained by plotting lag phases versus
antioxidant concentrations showed that IDBH2 was 50%
as efficient as Trolox in scavenging ONOO- (see also
Figure 8B). IDB or an equivalent volume of ethanol-HCl
did not inhibit DHR oxidation by ONOO- (see also Figure
8A).
Effects on ADP -F e(II)-Dep en d en t Micr osom a l
Lip id P er oxid a tion a n d Cytoch r om e P 450 Destr u c-
tion . ADP-Fe(II) is known to promote lipid peroxidation
in liposomal and microsomal systems (40). The precise
mechanism of initiation remains unknown, but there is

Idebenone Antioxidant Properties
Chem. Res. Toxicol., Vol. 11, No. 1, 1998 61
Ta ble 2. Effects of Id eben on e on ADP -F e(II)- a n d
CUOOH-Dep en d en t Micr osom a l Lip id P er oxid a tion a n d
Cytoch r om e P 450 Destr u ction
the free radical-scavenging efficiency of a given compound
should stand comparison with other established antioxi-
dants. According to such criteria, idebenone might be
viewed as a first-choice drug for antioxidant interventions
in free radical-mediated diseases. In fact, our studies
demonstrate that idebenone can effectively scavenge a
variety of free radical species, including DPPH and
ABTS•+ (cf. Figure 1), peroxyl and tyrosyl radicals (cf.
Figures 2-4), and ONOO- (cf. Figure 8). Idebenone can
also redox couple with hypervalent species of Mb or Hb,
thus preventing lipid peroxidation associated with the
generation of these species (cf. Figures 5-7 and Table 1).
In many cases, the antioxidant effects of idebenone
become appreciable at ∼2 µM, which is well in the range
of plasma levels attainable in patients after oral doses
of idebenone (4). When comparisons are made under
appropriate conditions, the antioxidant efficiency of ide-
benone varies from no less than 50% to slightly more
than 100% that of vitamin E or Trolox (cf. Figures 1-4
and 8). Furthermore, idebenone appears to protect
biomolecules irrespective of whether they are placed in
hydrophobic or hydrophilic environments. This is at-
tested by experiments in which this compound could
prevent peroxyl radical-dependent oxidation of water-
soluble (R)-PE or liposome-incorporated PnA (cf. Figures
2 and 3). These results suggest that idebenone may
intercept free radicals in both aqueous phases and lipid-
water interfaces. Several reasonings and lines of evi-
dence would therefore set the stage to conclude that
idebenone is a broadly active and biologically relevant
antioxidant.
TBARS
(nmol/mg
of protein)
P450 destruction
(nmol/mg
TBARS/P450
(nmol/nmol)
system
of protein)
ADP-Fe(II)a
+IDB
21.03 ( 1.16
21.04 ( 1.34
0.15 ( 0.05
7.40 ( 0.87
6.19 ( 0.50
0.11 ( 0.08
0.80 ( 0.09
0.74 ( 0.12
0.21 ( 0.06
0.93 ( 0.11
0.54 ( 0.09
0.32 ( 0.07
26.28
28.43
0.71
7.96
11.46
0.34
+IDBH2
CUOOHb
+IDB
+IDBH2
a
Incubations were prepared and assayed for TBARS and
cytochrome P450 destruction as described under Materials and
Methods. Values are those determined after 5-min incubations.
b
Incubations were prepared and assayed as in footnote a with the
exception that ADP-Fe(II) was replaced with CUOOH (0.25 mM).
tion accounts for g90% of TBARS formation (42), we
could therefore calculate the ratio of TBARS to cyto-
chrome P450 destruction as a tentative index of lipid
radical reactions mediated by the hemoprotein. As also
shown in Table 2, IDBH2 decreased this ratio by greater
than 95% in comparison with control incubations, con-
firming that IDBH2 could interfere with the propagation
of lipid peroxidation. On the other hand, IDB did not
prevent cytochrome P450 destruction nor affect the ratio
between TBARS accumulation and cytochrome P450
destruction.
The experiments described above suggested that ID-
BH2 had the potential to inhibit lipid peroxidation not
only by affecting ADP-iron reactions that initiate the
peroxidative process but also by intercepting the propa-
gating species generated upon subsequent involvement
of cytochrome P450. To better characterize the latter
possibility, we studied the effects of IDBH2 on a lipid
peroxidation system involving CUOOH in place of ADP-
Fe(II). This system relies on the homolytic cleavage of
CUOOH by cytochrome P450, yielding cumyl alkoxyl
radicals that initiate lipid peroxidation (43). The reaction
would subsequently proceed through the usual lipid
hydroperoxide- and cytochrome P450-dependent mech-
anism, although both the exogenously added and endog-
enously formed hydroperoxide may contribute to the net
destruction of the hemoprotein. As shown in Table 2,
CUOOH induced lipid peroxidation and cytochrome P450
destruction. IDBH2 prevented cytochrome P450 destruc-
tion by approximately 66%; however, some protection
(42%) was also observed with IDB, perhaps reflecting
bulky interferences with CUOOH-microsome interac-
tions. Nevertheless, only IDBH2 could significantly
inhibit the formation of TBARS and the ratio with
cytochrome P450 destruction. Collectively, these findings
showed that (i) idebenone did interfere with cytochrome
P450-dependent lipid peroxidation and (ii) the hydro-
quinone moiety was required to intercept lipid radicals
formed upon cytochrome P450 cleavage of hydroperoxides
and hemoprotein destruction.
Partially reduced species of dioxygen such as O2•- and
H2O2 are currently given attention as possible mediators
of neural dysfunctions, degenerative processes, or trau-
matic injury (1, 2). While not particularly reactive per
•-
se, both O2 and H2O2 may become cytotoxic upon
secondary reactions with other species. For example,
O2•- may become toxic by reacting with NO and forming
•
ONOO-, which is currently believed to mediate neuro-
toxicity by excitatory amino acids such as glutamate (39).
On the other hand, H2O2-dependent activation of MbIII
and HbIII may very likely occur when vascular damage
and bleeding set the premises for a leakage of these
proteins within tissues, as in the case of brain trauma,
stroke, or ischemia-reperfusion (1). Finally, H2O2 can
activate myeloperoxidase to compound I and II interme-
diates that oxidize tyrosine to tyrosyl radicals (44). In
terms of neuropathology, tyrosyl radicals might be formed
and play some noxious role under conditions of ischemia-
reperfusion and H2O2 formation by activated neutrophils
and macrophages. In experiments monitoring the reduc-
tion of nitro blue tetrazolium by xanthine oxidase-
•-
generated O2 or the stability of H2O2 at 240 nm, we
could determine that idebenone lacks reactivity with
either species (not shown). However, our data indicate
that IDBH2 readily interacts with ONOO-, tyrosyl radi-
cals, and H2O2-activated MbIII or HbIII. It follows that
•-
IDBH2 can mitigate the toxicity of O2 and H2O2 by
scavenging those reactive species which would function
as secondary and more potent mediators of oxidant
damage.
One additional mechanism of toxicity by O2•- and H2O2
appears to involve the ability of these species to delocalize
redox-active LMW iron from otherwise inactive cellular
stores. Superoxide is known to release iron from ferritin
(45) or enzymes endowed with Fe-S clusters (46).
Discu ssion
Drugs or natural compounds are very often studied in
vitro to evaluate whether they can be used as antioxi-
dants in vivo. Two major criteria should be kept in mind
when performing this type of determination. First, test
compounds should scavenge free radicals at concentra-
tions attainable in tissues and biological fluids. Second,

62 Chem. Res. Toxicol., Vol. 11, No. 1, 1998
Mordente et al.
Similarly, prolonged interactions of H2O2 with hemepro-
teins results in irreversible denaturation of the porphyrin
pocket and consequent release of the iron coordinated
therein (47). Once delocalized in a LMW form, iron
promotes lipid peroxidation through reaction mecha-
within the plasma milieu (e.g., LDL oxidation) but not
in intracellular environments. Likewise, vitamin E
penetrates the central nervous system slowly and may
thus fail to afford protection in neural cells acutely
exposed to oxidative injury (55). The clinical usefulness
of vitamin E might therefore be confined to the long-term
supplementation of patients affected by chronic degen-
erative processes, such as Alzheimer’s disease (56).
These problems are not observed with idebenone which
distributes throughout the body and readily approaches
similar concentrations in tissues and fluids (4, 57, 58).
These pharmacokinetic properties, as well as the anti-
oxidant functions described in our study, may help clarify
the neuroprotective action of idebenone and its greater
efficiency in comparison with vitamin E or CoQ10 (9, 12).
•-
nism(s) that may not depend any longer on O2 and
H2O2. This is the case for ADP-Fe(II) complexes which
initiate liposomal or microsomal lipid peroxidation by
virtue of SOD- and catalase-insensitive mechanisms (48).
ADP-iron-dependent lipid peroxidation probably pro-
ceeds through the oxidation of Fe(II) with oxygen and
•-
the formation of perferryl species [Fe(II)O2-Fe(III)O2
]
or poorly characterized Fe(II)-O2-Fe(III) complexes that
abstract hydrogen from the bis-allylic bonds of polyun-
saturated fatty acids (40, 49). In either case, lipid
peroxidation best occurs when appropriate Fe(II):Fe(III)
ratios are formed, although the molecular basis for such
a requirement has remained a matter of debate (40). Our
results demonstrate that idebenone affects the oxidation
of ADP-Fe(II) with oxygen and thus precludes the
initiation of lipid peroxidation by reactive species which
require some Fe(III) (cf. Table 2). Simultaneous deter-
minations of cytochrome P450 destruction and compara-
tive experiments in CUOOH-dependent systems suggest
that idebenone may also scavenge lipid-reactive species
which propagate lipid peroxidation and concur in mi-
crosomal damage (Table 2). Idebenone can therefore
mitigate the toxicity associated with iron delocalization
and oxidative deterioration of biological membranes.
Previous studies on the antioxidant functions of ide-
benone have not conclusively established whether it acts
in the hydroquinone or quinone form (7-12). The
experiments described in this paper would suggest that
the antioxidant effects of idebenone pertain exclusively
to its hydroquinone form, IDBH2; the quinone form, IDB,
cannot scavenge ABTS•+, peroxyl, or tyrosyl radicals, nor
can it react with ONOO-- or H2O2-activated hemopro-
teins. Likewise, IDB does not interfere with ADP-iron-
or CUOOH-dependent microsomal lipid peroxidation. The
experiments with ferrylmyoglobin also demonstrate that
reactions between IDBH2 and excited species accompany
its stoichiometric conversion to IDB. The antioxidant
effects of idebenone are therefore mediated by electron-
transfer mechanisms that involve oxidation of IDBH2 to
IDB. Similar mechanisms have been described for CoQ10
and imply that antioxidant interventions with idebenone
become effective following the reduction of IDB to IDBH2.
Unfortunately, site(s) and mechanism(s) of IDB reduction
are unknown. In principle, IDB might be reduced by the
same mitochondrial dehydrogenases that reduce CoQ10;
however, recent studies have shown that plasma mem-
branes from various cell types, including erythrocytes and
hepatocytes, contain NAD(P)H dehydrogenases which
reduce more rapidly short-chain coenzyme Q analogues
than CoQ10 (50, 51). This alternative pathway of reduc-
tion has also been described with purified DT-diaphorase
(52). As mentioned, the structure of idebenone is strongly
reminiscent of short-chain CoQ analogues; hence, both
CoQ10 and CoQ1 reductases might concur in converting
idebenone from IDB to IDBH2, making it suitable for
redox coupling with noxious radicals.
Refer en ces
(1) Halliwell, B. (1992) Reactive oxygen species and the central
nervous system. J . Neurochem. 59, 1609-1623.
(2) Olanow, C. W. (1993) A radical hypothesis for neurodegeneration.
Trends Neurosci. 16, 439-444.
(3) Bergamasco, B., Scarzella, L., and La Commare, P. (1994)
Idebenone, a new drug in the treatment of cognitive impairment
in patients with dementia of the Alzheimer type. Funct. Neurol.
9, 161-168.
(4) Gillis, J . C., Benefield, P., and McTavish, D. (1994) Idebenone. A
review of its pharmacodynamic and pharmacokinetic properties,
and therapeutic use in age-related cognitive disorders. Drugs
Aging 5, 133-152.
(5) Curti, D., Izzo, E., Brambilla, L., Facchetti, G., Sangiovanni, G.,
and Brambilla, G. (1995) Effect of a ubiquinone-like molecule on
oxidative energy metabolism in rat cortical synaptosomes at
different ages. Neurochem. Res. 20, 1001-1006.
(6) Nitta, A., Hasegawa, T., and Nabeshima, T. (1993) Oral admin-
istration of idebenone, a stimulator of NGF synthesis, recovers
reduced NGF content in aged rat brain. Neurosci. Lett. 163, 219-
222.
(7) Suno, M., and Nagaoka, A. (1984) Inhibition of lipid peroxidation
by a novel compound, idebenone (CV-2619). J pn. J . Pharmacol.
35, 196-198.
(8) Suno, M., and Nagaoka, A. (1984) Inhibition of lipid peroxidation
by a novel compound (CV-2619) in brain mitochondria and mode
of action of the inhibition. Biochem. Biophys. Res. Commun. 125,
1046-1052.
(9) Miyamoto, M., Murphy, T. H., Schnaar, R. L., and Coyle, J . T.
(1989) Antioxidants protect against glutamate-induced cytotox-
icity in a neuronal cell line. J . Pharmacol. Exp. Ther. 250, 1132-
1140.
(10) Suno, M., Terashita, Z., and Nagaoka, A. (1989) Inhibition of
platelet aggregation by idebenone and the mechanism of the
inhibition. Arch. Gerontol. Geriatr. 8, 313-321.
(11) Naito, M., Hayashi, T., Yamada, K., Asai, K., Yoshimine, N., and
Iguchi, A. (1993) Protective effects of idebenone on vascular
endothelial cells against toxicity induced by oxidatively modified
low-density lipoprotein. Artery 20, 314-323.
(12) Wieland, E., Schutz, E., Armstrong, V. W., Kuthe, F., Heller, C.,
and Oellerich, M. (1995) Idebenone protects hepatic microsomes
against oxygen radical-mediated damage in organ preservation
solutions. Transplantation 60, 444-451.
(13) Rieske, J . S. (1967) Preparation and properties of reduced
coenzyme Q-cytochrome c reductase (complex III of the respiratory
chain). Methods Enzymol. 10, 239-245.
(14) Tubaro, F., Micossi, E., and Ursini, F. (1996) The antioxidant
capacity of complex mixtures by kinetic analysis of crocin bleach-
ing inhibition. J . Am. Oil Chem. Soc. 73, 173-179.
(15) Blois, M. S. (1958) Antioxidant determinations by the use of stable
free radical. Nature 181, 1199-1200.
(16) Rice Evans, C., and Miller, N. J . (1994) Total antioxidant status
in plasma and body fluids. Methods Enzymol. 234, 279-293.
(17) Barr, D. P., and Aust, S. D. (1993) On the mechanism of
peroxidase-catalyzed oxygen production. Arch. Biochem. Biophys.
303, 377-382.
Antioxidant interventions with vitamin E or CoQ10 are
limited by pharmacokinetic constraints. Coenzyme Q10
approaches high plasma levels but enters tissues rather
poorly (53, 54). It follows that CoQ10 supplementation
may be a good strategy to ameliorate oxidant damage
(18) Terao, K., and Niki, E. (1986) Damage to biological tissues induced
by radical initiator 2,2′-azobis(2-amidinopropane) dihydrochloride
and its inhibition by chain-breaking antioxidants. J . Free Radical
Biol. Med. 2, 193-201.
(19) Glazer, A. N. (1990) Phycoerythrin fluorescence-based assay for
reactive oxygen species. Methods Enzymol. 186, 161-168.

Idebenone Antioxidant Properties
Chem. Res. Toxicol., Vol. 11, No. 1, 1998 63
(20) Kuypers, F. A., van den Berg, J . J ., Schalkwijk, C., Roelofsen, B.,
and Op den Kamp, J . A. (1987) Parinaric acid as a sensitive
fluorescent probe for the determination of lipid peroxidation.
Biochim. Biophys. Acta 921, 266-274.
(21) van den Berg, J . J ., Op den Kamp, J . A., Lubin, B. H., Roelofsen,
B., and Kuypers, F. A. (1992) Kinetics and site specificity of
hydroperoxide-induced oxidative damage in red blood cells. Free
Radical Biol. Med. 12, 487-498.
(22) Schultz, J . R., Ellerby, L. M., Gralla, E. B., Valentine, J . S., and
Clarke, C. F. (1996) Autoxidation of ubiquinol-6 is independent
of superoxide dismutase. Biochemistry 35, 6595-6603.
(23) Holler, T. P., and Hopkins, P. B. (1989) A qualitative fluorescence-
based assay for tyrosyl radical scavenging activity: ovothiol A is
an efficient scavenger. Anal. Biochem. 180, 326-330.
(24) Mordente, A., Santini, S. A., Miggiano, A. G., Martorana, G. E.,
Petitti, T., Minotti, G., and Giardina, B. (1994) The interaction
of short chain coenzyme Q analogues with different redox states
of myoglobin. J . Biol. Chem. 269, 27394-27400.
(25) Giulivi, C., Romero, F. J ., and Cadenas, E. (1992) The interaction
of Trolox C, a water-soluble vitamin E analogue, with ferrylmyo-
globin: Reduction of the oxoferryl moiety. Arch. Biochem. Bio-
phys. 299, 302-312.
(26) Buege, J . A., and Aust, S. D. (1978) Microsomal lipid peroxidation.
Methods Enzymol. 52, 302-310.
(27) Ryan, T. P., Samokyszyn, V. M., Dellis, S., and Aust, S. D. (1990)
Effects of (+)-1,2-bis(3,5-dioxopiperazin-1-yl)propane (ADR-529)
on iron-catalyzed lipid peroxidation. Chem. Res. Toxicol. 3, 384-
390.
(28) Crow, J . P., Beckman, J . S., and McCord, J . M. (1995) Sensitivity
of the essential zinc-thiolate moiety of yeast alcohol dehydroge-
nase to hypochlorite and peroxynitrite. Biochemistry 34, 3544-
3552.
(29) Minotti, G. (1992) The role of an endogenous nonheme iron in
microsomal redox reactions. Arch. Biochem. Biophys. 297, 189-
198.
(42) Svingen, B. A., Buege, J . A., O’Neal, F. O., and Aust, S. D. (1979)
The mechanism of NADPH-dependent lipid peroxidation. The
propagation of lipid peroxidation. J . Biol. Chem. 254, 5892-5899.
(43) Weiss, R. H., Arnold, J . L., and Estabrook, R. W. (1987) Trans-
formation of an arachidonic acid hydroperoxide into epoxyhydroxy
and trihydroxy fatty acids by liver microsomal cytochrome P-450.
Arch. Biochem. Biophys. 252, 334-338.
(44) Heinecke, J . W., Li, W., Francis, G. A., and Goldstein, J . A. (1993)
Tyrosyl radical generated by myeloperoxidase catalyzes the
oxidative cross-linking of proteins. J . Clin. Invest. 91, 2866-2872.
(45) Thomas, C. E., Morehouse, L. A., and Aust, S. D. (1985) Ferritin
and superoxide-dependent lipid peroxidation. J . Biol. Chem. 260,
3275-3280.
(46) Keyer, K., and Imlay, J . A. (1996) Superoxide accelerates DNA
damage by elevating free-iron levels. Proc. Natl. Acad. Sci. U.S.A.
93, 13635-13640.
(47) Puppo, A., and Halliwell, B. (1988) Formation of hydroxyl radicals
from hydrogen peroxide in the presence of iron. Is haemoglobin
a biological Fenton reagent? Biochem. J . 249, 185-190.
(48) Bucher, J . R., Tien, M., and Aust, S. D. (1983) The requirement
for ferric in the initiation of lipid peroxidation by chelated ferrous
iron. Biochem. Biophys. Res. Commun. 111, 777-784.
(49) Minotti, G., Mancuso, C., Frustaci, A., Mordente, A., Santini, S.
A., Calafiore, A. M., Liberi, G., and Gentiloni, N. (1996) Paradoxi-
cal inhibition of cardiac lipid peroxidation in cancer patients
treated with doxorubicin. Pharmacologic and molecular reap-
praisal of anthracycline cardiotoxicity. J . Clin. Invest. 98, 650-
661.
(50) Stocker, R., and Suarna, C. (1993) Extracellular reduction of
ubiquinone-1 and -10 by human Hep G2 and blood cells. Biochim.
Biophys. Acta 1158, 15-22.
(51) Valliant, F., Larm, J . A., McMullen, G. L., Wolvetang, E. J ., and
Lawen, A. (1996) Effectors of the mammalian plasma membrane
NADH-oxidoreductase system. Short-chain ubiquinone analogues
as potent stimulators. J . Bioenerg. Biomembr. 28, 531-540.
(52) Beyer, R. E., Segura Aguilar, J ., Di Bernardo, S., Cavazzoni, M.,
Fato, R., Fiorentini, D., Galli, M. C., Setti, M., Landi, L., and
Lenaz, G. (1996) The role of DT-diaphorase in the maintenance
of the reduced antioxidant form of coenzyme Q in membrane
systems. Proc. Natl. Acad. Sci. U.S.A. 93, 2528-2532.
(30) Laudicina, D. C., and Marnett, L. J . (1990) Enhancement of
hydroperoxide-dependent lipid peroxidation in rat liver mi-
crosomes by ascorbic acid. Arch. Biochem. Biophys. 278, 73-80.
(31) Minotti, G. (1989) Adriamycin-dependent release of iron from
microsomal membranes. Arch. Biochem. Biophys. 268, 398-403.
(32) Stoschek, C. (1990) Quantitation of protein. Methods Enzymol.
182, 58-60.
(33) Omura, T., and Sato, R. (1964) The carbon monoxide-binding
pigment of liver microsomes. I. Evidence for its hemoprotein
nature. J . Biol. Chem. 239, 2379-2385.
(34) Teng, C., Hsiao, G., Inoue, T., and Satoh, M. (1996) N-Allylseco-
boldine as a novel antioxidant against peroxidative damage. Eur.
J . Pharmacol. 303, 129-139.
(35) Tsuchiya, M., Scita, G., Freisleben, H. J ., Kagan, V. E., and
Packer, L. (1992) Antioxidant radical-scavenging activity of
carotenoids and retinoids compared to alpha-tocopherol. Methods
Enzymol. 213, 460-472.
(36) Maiorino, M., Roveri, A., and Ursini, F. (1992) Antioxidant effect
of Ebselen (PZ 51): Peroxidase mimetic activity on phospholipid
and cholesterol hydroperoxides vs free radical scavenger activity.
Arch. Biochem. Biophys. 295, 404-409.
(37) Giulivi, C., and Cadenas, E. (1994) Ferrylmyoglobin: Formation
and chemical reactivity toward electron-donating compounds.
Methods Enzymol. 233, 189-202.
(38) Pryor, W. A., and Squadrito, G. L. (1995) The chemistry of
peroxynitrite: A product from the reaction of nitric oxide with
superoxide. Am. J . Physiol. 268, L699-L722.
(39) Beckman, J . S., Chen, J ., Crow, J . P., and Ye, Y. Z. (1994)
Reactions of nitric oxide, superoxide and peroxynitrite with
superoxide dismutase in neurodegeneration. Prog. Brain Res. 103,
371-380.
(40) Minotti, G. (1993) Sources and role of iron in lipid peroxidation.
Chem. Res. Toxicol. 6, 134-146.
(53) Zhang, Y., Aberg, F., Appelkvist, E. L., Dallner, G., and Ernster,
L. (1995) Uptake of dietary coenzyme Q supplement is limited in
rats. J . Nutr. 125, 446-453.
(54) Lonnrot, K., Metsa Ketela, T., Molnar, G., Ahonen, J ., Latvala,
M., Peltola, J ., Pietila, T., and Alho, H. (1996) The effect of
ascorbate and ubiquinone supplementation on plasma and CSF
total antioxidant capacity. Free Radical Biol. Med. 21, 211-217.
(55) Pappert, E. J ., Tangney, C. C., Goetz, C. G., Ling, Z. D., Lipton,
J . W., Stebbins, G. T., and Carvey, P. M. (1996) Alpha-tocopherol
in the ventricular cerebrospinal fluid of Parkinson’s disease
patients: Dose-response study and correlations with plasma
levels. Neurology 47, 1037-1042.
(56) Sano, M., Ernesto, C., Thomas, R., Klauber, M. R., Schafer, K.,
Grundman, M., Woodbury, P., Growdon, J ., Cotman, C. W.,
Pfeiffer, E., and Schneider, L. S. (1997) A controlled trial of
selegiline, alpha-tocopherol, or both as treatment for Alzheimer
disease. N. Engl. J . Med. 336, 1216-1222.
(57) Torii, H., Yoshida, K., Kobayashi, T., Tsukamoto, T., and Tanaya-
ma, S. (1985) Disposition of idebenone (CV-2619), a new cerebral
metabolism improving agent, in rats and dogs. J . Pharmacobio-
dyn. 8, 457-467.
(58) Nagai, Y., Yoshida, K., Narumi, S., Tanayama, S., and Nagaoka,
A. (1989) Brain distribution of idebenone and its effect on local
cerebral glucose utilization in rats. Arch. Gerontol. Geriatr. 8,
257-272.
(41) Minotti, G., and Aust, S. D. (1987) An investigation into the
mechanism of citrate-Fe2+-dependent lipid peroxidation. Free
Radical Biol. Med. 3, 379-387.
TX970136J