Free Radical Biology & Medicine, Vol. 37, No. 6, pp.

755 767, 2004

Copyright D 2004 Elsevier Inc.
Printed in the USA. All rights reserved
0891-5849/$-see front matter

doi:10.1016/j.freeradbiomed.2004.05.034

Serial Review: The Powerhouse Takes Control of the Cell: the Role of
Mitochondria in Signal Transduction
Serial Review Editor: Victor Darley-Usmar
MITOCHONDRIAL SUPEROXIDE: PRODUCTION, BIOLOGICAL EFFECTS,
AND ACTIVATION OF UNCOUPLING PROTEINS
MARTIN D. BRAND, CHARLES AFFOURTIT, TELMA C. ESTEVES, KATHERINE GREEN,
ADRIAN J. LAMBERT, SATOMI MIWA, JULIAN L. PAKAY, and NADEENE PARKER
MRC Dunn Human Nutrition Unit, Cambridge CB2 2XY, UK
Available
online2004; Accepted 28 May 2004)
(Received 26 March 2004; Revised
24 May
Available online 23 June 2004

AbstractMitochondria are potent producers of cellular superoxide, from complexes I and III of the electron transport
chain, and mitochondrial superoxide production is a major cause of the cellular oxidative damage that may underlie
degradative diseases and aging. This superoxide production is very sensitive to the proton motive force, so it can be
strongly decreased by mild uncoupling. Superoxide and the lipid peroxidation products it engenders, including
hydroxyalkenals such as hydroxynonenal, are potent activators of proton conductance by mitochondrial uncoupling
proteins such as UCP2 and UCP3, although the mechanism of activation has yet to be established. These observations
suggest a hypothesis for the main, ancestral function of uncoupling proteins: to cause mild uncoupling and so diminish
mitochondrial superoxide production, hence protecting against disease and oxidative damage at the expense of a small
loss of energy. We review the growing evidence for this hypothesis, in mitochondria, in cells, and in vivo. More recently
evolved roles of uncoupling proteins are in adaptive thermogenesis (UCP1) and perhaps as part of a signaling pathway to
regulate insulin secretion in pancreatic h cells (UCP2). D 2004 Elsevier Inc. All rights reserved.
KeywordsMitochondria, Complex I, Proton motive force, Superoxide, Lipid peroxidation, Hydroxynonenal,
Uncoupling protein, Aging, Thermogenesis, Insulin secretion, Free radicals
Contents
1.
2.

3.

4.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Superoxide production by mitochondria . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1. Methods for measuring superoxide production . . . . . . . . . . . . . . . . . . .
2.2. Superoxide production by complex I . . . . . . . . . . . . . . . . . . . . . . . .
2.3. The effect of mild uncoupling on mitochondrial superoxide production from
complex I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.4. Superoxide production from sites in the electron transport chain other than complex I
Biological actions of superoxide produced by mitochondria . . . . . . . . . . . . . . . .
3.1. The significance of superoxide in biological systems . . . . . . . . . . . . . . . .
3.2. Superoxide, disease, and aging . . . . . . . . . . . . . . . . . . . . . . . . . . .
Activation of the proton conductance of mitochondrial carriers by superoxide. . . . . . .
4.1. Superoxide activation of the proton conductance of uncoupling proteins (UCPs). .
4.2. Activation of the proton conductance of mitochondrial carriers by other
radical-related species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3. Possible mechanisms of activation of mitochondrial carriers by superoxide . . . .

756
756
756
756
757
757
757
757
758
759
759
759
759

This article is part of a series of reviews on The Powerhouse Takes Control of the Cell: the Role of Mitochondria in Signal Transduction. The
full list of papers may be found on the home page of the journal.
Address correspondence to: Martin Brand, MRC Dunn Human Nutrition Unit, Hills Road, Cambridge CB2 2XY, UK. Fax: +44-1223-252805;
E-mail: martin.brand@mrc-dunn.cam.ac.uk.
755

756

M. D. BRAND et al.

5.

The role of superoxide activation in the function of uncoupling proteins . . . . . . . . .

5.1 Mitochondrial uncoupling proteins have an ancestral function of protection against
superoxide production and consequent damage . . . . . . . . . . . . . . . . . . .
5.2 Specialized functions of mitochondrial uncoupling proteins: UCP1 and adaptive
thermogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.3 Specialized functions of mitochondrial uncoupling proteins: UCP2 and regulation
of insulin secretion in pancreatic h cells . . . . . . . . . . . . . . . . . . . . . . .
5.4 Knockout mice, uncoupling proteins, and superoxide . . . . . . . . . . . . . . . .
4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

INTRODUCTION

The main function of mitochondria is ATP synthesis.

Electrons from reduced substrates are passed from complexes I and II of the electron transport chain through
complexes III and IV to oxygen, forming water and
causing protons to be pumped across the mitochondrial
inner membrane. The proton motive force set up by proton
pumping drives protons back through the ATP synthase in
the inner membrane, forming ATP from ADP and phosphate. There are two major side reactions that are relevant
here: electrons may leak from the respiratory chain and
react inappropriately with oxygen to form superoxide, and
pumped protons may leak back across the inner membrane, diverting the conserved energy away from ATP
synthesis and into heat production.
Mitochondrial superoxide production is a major cause
of the cellular oxidative damage that may underlie
degradative diseases and aging. In this review we outline
our current picture of the sites and determinants of
superoxide production by the electron transport chain
and then consider its cellular and physiological effects,
with particular emphasis on one of the recently identified
targets: the proton conductance of mitochondrial uncoupling proteins. These considerations lead to a model for
the function of uncoupling proteins in attenuating mitochondrial superoxide production. We consider the evidence for this model and the physiological and
pathological situations to which it may be relevant.
SUPEROXIDE PRODUCTION BY MITOCHONDRIA

Methods for measuring superoxide production

Measurement of mitochondrial superoxide production
is reviewed in [1,2]. Superoxide production by isolated
respiratory chain complexes or submitochondrial particles
is usually followed spectrophotometrically by monitoring
reduction of epinephrine or of acetylated cytochrome c.
Direct measurement of superoxide production in intact
isolated mitochondria is more problematic due to its
inability to cross the inner membrane and its rapid conversion to hydrogen peroxide by superoxide dismutase

760
760
762
762
764
764
764

(SOD) and other enzymes. Most direct methods use spin

traps with electron paramagnetic resonance spectroscopy.
Indirect methods rely on fluorometric measurement of
external hydrogen peroxide, which readily diffuses out of
the mitochondrial matrix. Inclusion of SOD in the external
medium ensures that superoxide generated on either side
of the inner membrane is detected; in the absence of SOD
most of the signal comes from superoxide generated in the
matrix and converted to hydrogen peroxide in situ [3].
Superoxide production by complex I
The relative contributions of the different complexes of
the electron transport chain to total superoxide production
may depend on the tissue examined [4]. However, there is
growing evidence that complex I produces most of the
superoxide generated by intact mammalian mitochondria
in vitro; the contribution of other complexes and sites is
low. When mitochondria oxidize the complex II substrate
succinate in the absence of electron transport chain inhibitors, the rate of superoxide production is high. However,
this rate is almost abolished when the complex I inhibitor
rotenone is added [5 12]. Therefore, high superoxide
production from complex I occurs during reverse electron
transport from succinate to NAD+ driven by the high proton motive force generated from proton pumping by
complexes III and IV. This superoxide production is
primarily on the matrix side of the inner membrane [3,12].
Curiously, forward electron transport into complex I
from NAD-linked substrates produces little superoxide.
When complex I inhibitors such as rotenone are added,
this low rate increases, but not usually to the rate seen with
reverse electron transport [5 7,11]. This asymmetry of
superoxide production between forward and reverse electron transport has been investigated in this laboratory (A.J.
Lambert and M.D. Brand, unpublished observations),
leading to the conclusion that high rates of superoxide
production from complex I in vitro require an adequate
supply of electrons, a high pH gradient across the membrane [13], and a particular configuration achieved by
reverse electron transport or by inhibition of complex I
with high concentrations of myxothiazol (more usually
used as a potent complex III inhibitor).

Mitochondrial superoxide and UCPs

The sites of superoxide production within complex I,

and their relative importance, remain uncertain (Fig. 1).
Each of the electron-transferring components (FMN,
FeS centers, Q) within the complex has been suggested
as a superoxide-producing site [7,9,14 17]. This lack of
consensus may reflect the protocols used; it is likely
that different rates will be obtained from different sites
depending on the conditions [13]. In addition, many of
the inhibitors used to dissect out the sites of production
of superoxide within complex I are poorly characterized
in intact mitochondria. They include FMN reagents such
as diphenyleneiodonium and FeS center inhibitors such
as p-chloromercuribenzoate and ethoxyformic anhydride
and a variety of Q site inhibitors [18]. In our hands
(with the exception of rotenone and piericidin), these
complex I inhibitors display nonspecific effects, such as
inhibition of succinate-linked respiration, uncoupling,
and proton motive force-dependent accumulation into
the mitochondrial matrix (A.J. Lambert and M.D.
Brand, unpublished observations). Our results suggest
that the major site involved in high rates of superoxide
production involves FeS center N2 and the associated
ubiquinone binding site.
The effect of mild uncoupling on mitochondrial
superoxide production from complex I
Mild uncoupling using synthetic uncouplers increases
the conductance of the mitochondrial membrane to
protons and so decreases both of the components of
the proton motive force: membrane potential and pH
gradient. Several reports link superoxide production rate
to the magnitude of the membrane potential
[6,11,19,20], and the high superoxide production from
complex I during reverse electron transport is particularly sensitive to membrane potential and to mild
uncoupling [12]. More recently, we found that superoxide production by complex I is even more sensitive to
the pH gradient than it is to membrane potential, since
conversion of pH gradient to membrane potential at
constant proton motive force strongly depressed superoxide production [13]. Therefore mild uncoupling,
which slightly reduces both membrane potential and
pH gradient, very effectively decreases the high superoxide production that occurs from complex I during
reverse electron transport.
Superoxide production from sites in the electron
transport chain other than complex I
Other sites in the electron transport chain are also
capable of producing superoxide, particularly in the
presence of inhibitors of electron transport.
The importance of ubiquinone in superoxide production has long been noted, and the quinone pool was
originally suggested to be a major site [21]. Currently, it

757

is accepted that along with complex I, complex III also

has a large capacity to produce superoxide (e.g.,
[12,22,23]), at least in the presence of antimycin, a
specific inhibitor of center i of complex III.1 Myxothiazol (an inhibitor of center o of complex III)
decreases superoxide production [3,12], suggesting that
center o is responsible for its production within complex
III. Structurally, center o faces the cytosolic side of the
membrane [24,25], but it seems that the superoxide
produced at this site appears at both sides of the inner
membrane [12,22].
Complex II can produce superoxide in the presence
of complex II inhibitors [26,27], but it is not normally a
significant site. However, fumarate reductase, which
catalyzes the reverse reaction in bacteria, can produce
significant amounts of superoxide when oxygen is
present [28], suggesting that complex II has evolved
to generate little superoxide. Glycerol-3-phosphate dehydrogenase produces significant amounts of superoxide in the native state, mostly to the cytosolic side of the
membrane [12,29]. Its distribution is limited in mammals to tissues such as brown adipose and brain, where
it is a potentially important site of superoxide production [30]. In addition, two other enzymes involved in
fatty acid oxidation, electron transfer flavoprotein and
electron transfer flavoprotein quinone oxidoreductase,
may also produce superoxide, in the matrix [3].
BIOLOGICAL ACTIONS OF SUPEROXIDE PRODUCED BY
MITOCHONDRIA

The significance of superoxide in biological systems

Superoxide is a reactive molecule but it can be
converted to hydrogen peroxide by superoxide dismutase (Mn-SOD in the matrix, Cu/Zn-SOD in the cytosol)
and then to oxygen and water by catalase or glutathione
peroxidase. However, these antioxidant systems are not
perfect, and superoxide that evades them (together with
the secondary reactive oxygen species it generates) can
damage proteins, lipids, and DNA directly. Although the
hydrogen peroxide produced by SOD is relatively
unreactive, it can form highly reactive hydroxyl radicals
in the presence of ferrous ion via Fenton chemistry and
these hydroxyl radicals can initiate lipid peroxidation
cascades in membranes. Furthermore, the products of
sugar, protein, and lipid oxidation can cause secondary
damage to proteins, which may lose catalytic function

Center i (also known as center N) faces the inside

(electronegative side) of the mitochondrial inner membrane and is the
site of ubiquinone reduction. Center o (also known as center P)
faces the outside (electropositive side) of the mitochondrial inner
membrane and is the site of ubiquinol oxidation.

758

M. D. BRAND et al.

Fig. 1. Superoxide (O2 ) production by complex I of the electron transport chain. Forward and reverse electron transport is shown by
solid and dashed black arrows, respectively. Production of superoxide from the flavin (FMN), iron sulfur centers (N1 N5), and
possible Q binding sites (A, B, and C) is shown as gray arrows. A large Q binding pocket with three quinone binding sites has been
proposed [18,110]. Hydrogen peroxide (H2O2) production by complex I in intact mitochondria is independent of external SOD, and thus
superoxide is produced in the matrix exclusively [3,12]. SDH, succinate dehydrogenase.

and undergo selective degradation [31,32]. Thus mitochondrially produced superoxide can be a major source
of cellular damage (oxidative stress). The significance of superoxide production in the mitochondrial
matrix is clearly demonstrated in two different models
of Mn-SOD knockout mice. Although the pathological
phenotypes of the two models differ, in each case life
span is severely curtailed, to either 10 d [33] or 3 weeks
[34].
Superoxide, disease, and aging
There are many pathologies related to oxidative
stress [35], including atherosclerosis, hypertension, ischemia reperfusion, inflammation, cystic fibrosis, cancer, diabetes, Parkinsons disease, and Alzheimers
disease. The general concept that oxidative stress is
involved in many diseases appears robust, but lacks
detail. Less understood is the proposed relationship
between superoxide production, oxidative stress, and
aging. According to the free radical theory of aging
[36 38], the progressive decline in cellular function
with age results from accumulation of damage to
cellular constituents. Reactive oxygen species (ROS;
e.g., superoxide, hydroperoxyl radical, hydrogen peroxide, and hydroxyl radical) are of particular importance,
since much damage is oxidative.
If mitochondrial superoxide production is important
in determining the rate of aging, then long-lived animals
should produce less. This is true in mammalian mitochondria: the rates of hydrogen peroxide and superoxide
production by mitochondria from ox (maximum life
span 30 years) are approximately 20% of those from
mouse (3.5 years) [39]. The confounding effects of

body size and metabolic rate can be avoided by comparing animals that have similar body sizes and metabolic rates, such as pigeons and rats. Mitochondria from
pigeon (maximum life span 35 years) produce superoxide at significantly lower rates than mitochondria from
rat (4 years) [40]; the difference is at complex I of the
electron transport chain [41]. Even better comparisons
would be between bats or naked mole rats (with
relatively long life spans) and similar-sized short-lived
mammals. Initial data suggest that mitochondria from
long-lived bats do produce ROS at lower rates than their
shorter lived counterparts [42].
Caloric restriction has clear effects on survival of
laboratory rodents and other species: it extends mean
and maximum life span, delays the onset and lowers the
frequency of age-related diseases, and retards the agerelated decline in several physiological systems [43].
Mitochondria isolated from rats and mice on caloric
restriction produce superoxide at lower rates than agematched fully fed controls [44 47]. This lowering of
ROS production tends to correlate with lowered oxidative damage in tissues of calorically restricted rodents,
and this has been proposed as a mechanism of retarded
aging [48,49].
Although both longevity and caloric restriction are
associated with lower mitochondrial superoxide production, they have no consistent effect on its removal. SOD
activity correlates with life span in some (but not all)
tissues of mammals and birds [50 52], but caloric
restriction has no consistent effect on SOD activity
[31]. Taking other antioxidant enzymes (catalase, glutathione peroxidase) into account, there are no consistent
correlations with maximum life span and protection

Mitochondrial superoxide and UCPs

against ROS. Therefore, it would seem that if superoxide contributes to aging, selection and caloric restriction extend life span, at least in part, by lowering its
production. Of course, many other changes could explain the observed effects, and there is no direct
evidence that lowering superoxide production increases
life span.
Do manipulations that enhance the removal of superoxide extend life span? Overexpression of Cu/ZnSOD in Drosophila had no effect on life span in one
study [53] but a positive effect in another [54]. However, the effects of overexpression of antioxidant
enzymes on life span may be dependent on the longevity of the control strain used [55]. Treatment of
Caenorhabditis elegans with the Eukarion SOD/catalase
mimetics EUK-8 and EUK-134 increased life span [56],
but it seems that this effect can be obtained only under
highly specific culture conditions [57]. When supplied
in the drinking water of the housefly, Musca domestica,
neither EUK-8 nor EUK-134 had any effect on life span
[58].
Overproduction or inadequate removal of superoxide
can give rise to severe phenotypes, molecular damage,
and pathology, but it remains to be seen if normal
superoxide production by mitochondria is a major factor
in aging.

759

muscle mitochondria from UCP3 / [66] and kidney

mitochondria from UCP2 / [67] mice. UCPs can be
activated by extramitochondrially generated superoxide
[64,65] and by superoxide generated in the mitochondrial matrix under nonphysiological [65] or physiological [8] conditions. In summary, superoxide produced
by the electron transport chain can cause mild uncoupling of mitochondria by activating the proton conductance of UCPs.
Activation of the proton conductance of mitochondrial
carriers by other radical-related species

MITOCHONDRIAL CARRIERS BY SUPEROXIDE

Insight into the mechanism by which superoxide

activates UCPs came from the finding that the lipid
peroxidation product 4-hydroxy-trans-2-nonenal (HNE)
induces uncoupling of mitochondria through UCP1,
UCP2, and UCP3 and also through the adenine nucleotide
translocase (ANT) [68]. Further progress came from the
use of MitoPBN, a novel mitochondrially targeted derivative of the spin trap a-phenyl-N-tert-butylnitrone that
reacts rapidly with carbon-centered radicals but not with
superoxide or lipid peroxidation products [69]. MitoPBN
prevented activation of UCPs by superoxide and by a
chemical generator of carbon-centered radicals, AAPH,
but not by HNE [69]. These observations support a model
in which endogenous superoxide production generates
carbon-centered radicals that initiate lipid peroxidation,
producing alkenals (e.g., HNE) that activate UCPs and
ANT [69,70] (Fig. 2).

Superoxide activation of the proton conductance of

uncoupling proteins (UCPs)

Possible mechanisms of activation of mitochondrial

carriers by superoxide

An initial connection between UCPs and reactive

oxygen species was suggested in 1997 when GDP, the
classic nucleotide inhibitor of UCP1, was shown to
increase membrane potential and hydrogen peroxide
production in mitochondria containing UCP1 or UCP2,
but not in mitochondria lacking UCPs [59]. Reactive
oxygen species were first implicated as potential activators of uncoupling by UCPs in the late 1990s when
studies suggested that they caused ATP-inhibited uncoupling in plant mitochondria [60,61].
Added ubiquinone stimulates GDP-sensitive proton
conductance in mammalian mitochondria, but this stimulation is prevented by superoxide dismutase, suggesting that activation of UCPs occurs via superoxide
production [62]. Indeed, ubiquinone itself is not
required for UCP activity in mitochondria [63]. Subsequent work revealed that superoxide increases mitochondrial proton conductance through UCP1, UCP2,
and UCP3 [64,65]. This activation is prevented by
purine nucleotides and requires uncoupling proteins
since it is present in yeast mitochondria ectopically
expressing mammalian UCP1 but absent in skeletal

Peroxidation of the polyunsaturated fatty acyl chains

of phospholipids generates a complex mixture of shortchain aldehydes: 2-alkenals such as 2-hexenal, 4-hydroxy-2-alkenals such as HNE, and ketoaldehydes such
as malondialdehyde [71,72]. Initially, these aldehydes
were believed to cause only general cytotoxic effects
associated with oxidative stress, but evidence is accumulating that they also have specific signaling roles
[73].
The 4-hydroxy-2-alkenals, particularly HNE, are the
most bioactive, probably through reaction with proteins
at the aldehyde group and across the C2C3 double
bondmade more electropositive by the 4-hydroxy
group. The double bond and either an acyl or a carbonyl
group are essential for activation of UCPs [68]. Structurally related compounds containing both these elements, such as other 2-alkenals, trans-retinoic acid, and
the retinoid TTPNB, are also UCP activators (Table 1).
Retinoic acid and TTPNB were previously thought to
act as fatty acid analogs [74], but the presence of the
CC double bond and acyl group suggests they may
activate in a manner similar to that of HNE.

ACTIVATION OF THE PROTON CONDUCTANCE OF

760

M. D. BRAND et al.

Fig. 2. Model for activation of mitochondrial carriers (UCPs and ANT) by superoxide, through initiation of lipid peroxidation.
Superoxide (O2 ) and hydroperoxyl radical (HO2 ) are generated primarily by the electron transport chain (ETC). Superoxide is
dismutated to hydrogen peroxide (H2O2) by Mn-superoxide dismutase (MnSOD). Some superoxide inactivates enzymes (e.g., aconitase)
that contain iron sulfur centers, releasing ferrous iron, which catalyzes production of hydroxyl radicals ( OH) from the H2O2 by the
Fenton reaction. The hydroxyl and hydroperoxyl radicals can both extract hydrogen atoms from polyunsaturated fatty acyl chains of
membrane phospholipids, generating carbon-centered fatty acyl radicals that react with oxygen to form peroxyl radicals. These
propagate lipid peroxidation cascades that produce a complex mixture of species including reactive alkenals such as 4-hydroxy-2nonenal (shown here). Reactive alkenals activate UCPs and ANT, increasing the proton conductance of the mitochondrial inner
membrane. Known activators and their structural homologs are shown in Table 1. This induced uncoupling is part of the proposed
regulatory mechanism to decrease mitochondrial generation of superoxide when it is too high (Fig. 3).

How do HNE and its structural analogs interact with

proteins? 4-Hydroxyalkenals can react with most free
amino acids [71]. However, the majority of HNE
protein adducts are with specific cysteine, lysine, or
histidine residues through Michael additions and
Schiffs bases [72,75 77]. The presence of two functional groups in alkenals, each capable of forming stable
covalent adducts, means that they can also create interor intraprotein cross-links.
So far, a definitive link between the covalent modification of UCPs or ANT by lipid peroxidation products and the induction of mild uncoupling has not been
made. However, ANT is inhibited by HNE [78] and
preferentially oxidatively modified in mitochondria exposed to superoxide [79]. Modification was conformation specific, as carboxyatractylate protected but
bongkrekic acid did not [80]. In houseflies, ANT
exhibited HNE adducts and was the only protein in
mitochondrial membranes showing an age-associated
increase in carbonyls [81]. The initial steps in determining how aldehydic lipid peroxidation products induce proton conductance through UCPs and ANT will

involve verifying the existence and then characterizing

the nature and location of adducts.
THE ROLE OF SUPEROXIDE ACTIVATION IN THE
FUNCTION OF UNCOUPLING PROTEINS

Mitochondrial uncoupling proteins have an ancestral

function of protection against superoxide production and
consequent damage
The thermogenic function of UCP1 has been well
characterized, but a function for its homologs (UCP2,
UCP3, avian UCP, and plant UCP) has yet to be
unambiguously defined. A possible physiological function for UCPs has been proposed [68 70] based on
two observations discussed above. First, matrix superoxide production from complex I is highly dependent
on the magnitude of the proton motive force, and
second, matrix superoxide activates the proton conductance of UCPs and induces mild uncoupling. In the
proposed model, UCPs respond to overproduction of
matrix superoxide by catalyzing mild uncoupling,
which lowers proton motive force and decreases super-

Mitochondrial superoxide and UCPs

Table 1. Reactive Alkenals and Homologs That Activate the Proton
Conductance of Uncoupling Proteins (UCPs) and the Adenine
Nucleotide Translocase (ANT)

Compound

Structure

Shown
for
References
UCP1
UCP2
UCP3
ANT

[68,69]
[68,69]
[68,69]
[68]

UCP2
ANT

[68]
[68]

UCP2
ANT

[68]
[68]

trans-Retinoic
acid

UCP1
UCP2
ANT

[68,74,109]
[68,74]
[68]

trans-Retinal

UCP2
ANT

[68]
[68]

UCP2
ANT

[68]
[68]

UCP2

[74,82]

4-Hydroxy-2nonenal

trans-2-Nonenal

trans-2-Nonenoic
acid

Cinnamic acid

TTNPB

oxide production from the electron transport chain,

attenuating superoxide-mediated damage, at the cost
of slightly lowered efficiency of oxidative phosphorylation (Fig. 3). This negative feedback loop to protect
the cell from ROS-induced damage may represent the
ancestral function of all UCPs.
If this postulated ancestral function of UCPs is
correct, then UCP-containing mitochondria should display UCP-specific mild uncoupling under conditions
allowing endogenous superoxide production. This has
been demonstrated in isolated mitochondria [8] and

761

intact cells [67]. Isolated skeletal muscle mitochondria

incubated with succinate exhibited high rates of endogenous superoxide production that was 75% inhibitable
by rotenone, showing that most of it was produced by
reverse electron flow from succinate dehydrogenase into
complex I. There was a significant GDP-sensitive component to proton conductance, indicating that UCP3 can
be activated by endogenous superoxide. Decreasing the
rate of superoxide production with rotenone partially
attenuated the GDP-sensitive proton conductance. Decreasing superoxide concentrations further with a mitochondrially targeted antioxidant abolished the GDPsensitive component [8]. In cells, a substantial proportion of the mitochondrial proton conductance observed
in isolated thymocytes is lost in UCP2-ablated thymocytes [82], an effect reproduced by addition of a cellpermeant SOD mimetic [67], strongly suggesting that
endogenous superoxide activates UCP2 significantly in
intact thymocytes.
If the model in Fig. 3 is correct, then conversely,
inhibition or ablation of UCP should increase mitochondrial membrane potential, ROS production, and
oxidative damage. This has been demonstrated in mitochondria and in cells. Mitochondria isolated from rat
skeletal muscle have significantly higher membrane
potential and superoxide production when GDP is added,
showing that UCP3 can function to limit the rate of
endogenous superoxide production [8]. In the same way,
GDP stimulates H2O2 production by brown adipose
tissue mitochondria and by those from UCP2-containing
nonparenchymal liver cells, spleen, and thymus [59]. In
addition, skeletal muscle mitochondria isolated from
UCP3 / mice have higher ROS levels [83] and increased markers of ROS-dependent damage [84] than
wild type. In macrophages and endothelial cells, UCP2
knockout raises membrane potential, superoxide levels,
and markers of oxidative stress [85,86]. In pancreatic
islets, when UCP2 activity is suppressed by addition of a
cell-permeant SOD mimetic, by adenoviral overexpression of Mn-SOD, or by UCP2 knockout, mitochondrial
membrane potential and islet superoxide production,
ATP levels, and insulin secretion are raised significantly
[67].
Thus the UCPs can be stimulated by endogenous
superoxide in isolated mitochondria and in cells to
increase mitochondrial proton conductance, lower mitochondrial proton motive force, and decrease mitochondrial ROS production and oxidative damage, in
accordance with the model in Fig. 3.
Mitochondrial hydrogen peroxide and ROS production may be controlled processes that are important in
cellular signaling [87]. In addition to its role in protection against oxidative stress, the feedback pathway
through the UCPs that is described here provides a

762

M. D. BRAND et al.

Fig. 3. Model of the ancestral function of UCPs. Mild uncoupling mediated by UCPs decreases the production of mitochondrial ROS via
a negative feedback loop, which serves as a protective mechanism against the deleterious effects of ROS. Matrix superoxide production
is exquisitely sensitive to proton motive force (PMF). A high PMF across the mitochondrial inner membrane leads to an increase in oneelectron donors capable of reducing oxygen to superoxide. Superoxide is the precursor of other forms of ROS leading to oxidative chain
reactions generating carbon-centered radicals, which in turn initiate lipid peroxidation (Fig. 2). These ROS cause oxidative damage to
the cell. Reactive aldehydes produced by ROS-mediated lipid peroxidation activate UCPs to induce mild uncoupling. Mild uncoupling
increases oxygen consumption, leading to a decrease in local oxygen concentration, and decreases PMF, leading to a lower concentration
of one-electron reductants. These effects combine to close the negative feedback loop that attenuates production of mitochondrial ROS.

potential mechanism for the regulation of such cellular

signaling pathways.
Specialized functions of mitochondrial uncoupling
proteins: UCP1 and adaptive thermogenesis
The main function of UCP1 is to catalyze adaptive
thermogenesis in brown adipose tissue in mammals. This
thermogenic function, presumably evolved rela-tively
recently (because it is found only in mammals), would
have been achieved partly by a massive increase in the
expression of UCP1 in the cold. One to three percent of
brown adipose tissue mitochondrial protein in warmadapted rodents is UCP1 and in cold-adapted animals
this rises to 3 5% [88], whereas only 0.01% of muscle
or brown adipose mitochondrial protein is UCP3 [89]
and only 0.02 0.03% of lung or spleen mitochondrial
protein is UCP2 [90], a concentration difference of up to
500-fold. This massive concentration difference explains
why UCP2 and UCP3 normally make a negligible
contribution to thermogenesis and basal metabolic rate.
Increased expression of UCP1 would have been accompanied by increased capacity for fat oxidation in brown
adipocytes and by strengthened dependence of UCP1catalyzed proton conductance on free fatty acid concentration, to allow regulation of the thermogenic response.
It is possible that avian UCP has evolved to fulfill a
similar adaptive thermogenic role in bird muscle [91].

Specialized functions of mitochondrial uncoupling

proteins: UCP2 and regulation of insulin secretion
in pancreatic b cells
UCP2 is expressed in pancreatic h cells [92], and
mitochondria from clonal h cells show marked superoxide-stimulated, GDP-inhibited UCP2 proton conductance [64]. When the plasma glucose concentration
rises, h cells increase their oxidative metabolism, resulting in an increased ATP/ADP ratio, closure of KATP
channels, depolarization of the plasma membrane potential, influx of Ca2+, and secretion of insulin (Fig. 4; [93]).
This signaling pathway requires that mitochondrial electron transfer is tightly coupled to ATP synthesis. In this
respect, the presence of UCP2 in h cell mitochondria is
surprising, since its activity will decrease such coupling
and consequently impair glucose-stimulated insulin secretion (GSIS).
Several studies have investigated the role of UCP2 in h
cells. Adenovirus-mediated overexpression of UCP2
decreases GSIS in pancreatic islets [94] and cultured
insulinoma cells [95]. However, these results could reflect
a UCP2 overexpression artifact that compromises mitochondrial integrity, rather than the native function of
UCP2. More persuasively, ablation of UCP2 results in
mouse islets with improved GSIS [96]. Furthermore,
attenuating endogenously generated superoxide with a

Mitochondrial superoxide and UCPs

763

Fig. 4. Regulation of insulin secretion by glucose and fatty acids. Glucose enters the h cell via a glucose transporter (GLUT2). Glucose
oxidation establishes a proton motive force (PMF) that drives ATP synthesis, increasing the ATP/ADP ratio. This causes closure of KATP
channels, depolarization of the plasma membrane potential (Dcp), and Ca2 + flux into the cell, triggering insulin release. UCP2 activity
dissipates the proton motive force, lowering ATP/ADP. At permissive glucose concentrations, fatty acids potentiate glucose-stimulated
insulin secretion through a G-protein-linked receptor (GPR40) pathway [99].

cell-permeant SOD mimetic blunts UCP2-mediated proton leak in h cells and improves GSIS [67]. These
observations suggest a role for UCP2 in the pathological
development of diabetes [67,94,96]. This may be correct,
but it is unlikely that the evolved function of UCP2 is
pathological.
Physiologically, h cells face a fluctuating mixture of
glucose and fatty acids. We suggest that superoxide
regulation of UCP2 activity normally coordinates the
appropriate response to changes in nutrient supply
[64,97]. Activation of the proton conductance of UCP2
by superoxide and the consequent decrease in mitochondrial proton motive force could have two main purposes:
to attenuate further superoxide production and reduce
oxidative damage, as discussed above (Fig. 3), or to
lower the ATP/ADP ratio, providing a physiological
signal to reduce insulin secretion (Fig. 4).
A signaling role of UCP2 could be important during
periods of low glucose concentration (e.g., sleep). Such
conditions favor h-oxidation of fatty acids, resulting in
superoxide production [3,98]. Activation by superoxide
of mild uncoupling mediated by UCP2 will ensure that

the h cell responds properly to the lack of glucose, as it

will attenuate insulin secretion despite the availability of
an adequate substrate for ATP production. A protective
role of UCP2 against ROS formation may be relevant
when both glucose and fatty acid concentrations are
high (e.g., after a meal). Superoxide activation of UCP2
will decrease the proton motive force, which will reduce
potential oxidative damage, but impair insulin secretion.
The potentiation of GSIS by fatty acids observed under
these conditions can be explained by fatty acid activation of a specific G-protein-linked receptor ([99]; Fig.
4). The only pathological effect of UCP2 predicted by
our model occurs during sustained hyperglycemia when
superoxide activation of UCP2 results in impaired
insulin secretion [67,100]. Such hyperglycemic conditions in humans, however, are predominantly associated
with recent Western diets and are therefore of limited
evolutionary relevance.
A protective role of UCP2 against oxidative damage is
at variance with studies that show that UCP2 knockout
mice exhibit improved glucose homeostasis for an extended period of time and do not suffer h cell loss, despite

764

M. D. BRAND et al.

elevated ROS levels (see [67]). In general, however, it is

difficult to assess whether laboratory mice have maintained a function for UCP2 in h cells and, if so, to what
extent normal husbandry necessitates its presence. Moreover, the evolutionary advantage provided by UCP2
effects on glucose homeostasis may be hard to detect by
conventional experimental approaches.
Knockout mice, uncoupling proteins, and superoxide
Mitochondria from UCP knockout mice have proved
to be crucial in showing that the GDP-sensitive proton
conductance activated by superoxide or reactive alkenals
in mammalian mitochondria is indeed mediated by UCP3
[64,66,68] or UCP2 [67], because the effects of the
activators are lost in mitochondria from the knockout
animals. Knockouts are, of course, immune from the
criticism of artifactual UCP function.
Similarly, cells isolated from UCP2 knockout mice
have been used to validate the important conclusions that
UCP2-dependent uncoupling can occur in situ in response to endogenous superoxide [67] and to the retinoic
acid analog TTNPB [82].
The whole-animal phenotypes of UCP knockout
mice are instructive. The phenotypes of UCP2 / and
UCP3 / mice show that these UCPs are not required
for body-weight regulation, exercise tolerance, or (unlike UCP1) cold-induced thermogenesis [83,85,101].
Consistent with the model in Fig. 3, however, these
mice show signs of higher levels of ROS. UCP2 may
have a fundamental role in protection against diseases
involving oxidative damage, such as atherosclerosis
[102,103]. On the other hand, activation of UCP2 by
ROS could be of hindrance in diabetes (discussed
above), in acute pathogen infection [85], and perhaps
in drug-induced hyperthermia [104].
UCP2-ablated mice are reported to have increased
ROS in macrophages compared to wild type [85] and, in
parallel, exhibit accelerated atherosclerotic plaque development [102], consistent with the view that oxidative
damage is involved in atherogenesis [105,106]. It seems
that in atherosclerosis, UCP2 expression in macrophages
confers protection against disease progression. However,
higher levels of ROS are associated with cytolytic
activity [107] and can confer increased toxoplasmacidal
and bactericidal activity in macrophages. Indeed, UCP2ablated mice are resistant to normally lethal Toxoplasma
gondii infection and this effect disappears upon addition
of a ROS scavenger [85], suggesting that attenuation of
ROS production by UCP2 may be counterproductive in
such infections.
UCP3 / mice have a decreased thermogenic response
to the drug MDMA (3,4-methylenedioxymethamphetamine, ecstasy), which can cause fatal hyperthermia,
suggesting that MDMA can activate UCP3-mediated

thermogenesis in skeletal muscle [104]. MDMA does

not activate directly the proton conductance of UCP3 in
isolated mitochondria (J.L. Pakay and M.D. Brand, unpublished results), but we can speculate that metabolites of
MDMA, which share some structural characteristics with
aldehydic lipid peroxidation products [108], may do so.
The damaging effects of amphetamines, including
MDMA, may be mediated by free radicals [108], so
MDMA might activate UCP3 by causing directly or
indirectly an increase in ROS.
These studies in knockout mice show that UCPs may
have an important role in long-term protection against
ROS. Conversely, the high levels of ROS generated
during hyperglycemia or possibly administration of drugs
such as MDMA cause overactivation of UCPs that could
potentially be fatal or lead to development of conditions
such as diabetes.

CONCLUSIONS

Although much is known about superoxide production by mitochondria, we lack a consensus on where it is
produced in the electron transport chain, on the mechanism of its production, and on how it is regulated. There
is a vast literature about effects of superoxide and other
ROS on cells, but there is still no unambiguous mechanistic description of how mitochondrial superoxide production may cause disease or aging. Research into the
interactions of superoxide with mitochondrial uncoupling
proteins is in its infancy, but the early indications are that
it represents an effective way to attenuate overproduction
of ROS by mitochondria and so protect against disease
and aging, as well as a potential signaling mechanism in
some cells. Much further work will be needed to explore
these possibilities.
REFERENCES
[1] Barja, G. The quantitative measurement of H2O2 generation in
isolated mitochondria. J. Bioenerg. Biomembr. 34:227 233;
2002.
[2] Degli Esposti, M. Measuring mitochondrial reactive oxygen species. Methods Enzymol. 26:335 340; 2002.
[3] St-Pierre, J.; Buckingham, J. A.; Roebuck, S. J.; Brand, M. D.
Topology of superoxide production from different sites in the
mitochondrial electron transport chain. J. Biol. Chem. 277:
44784 44790; 2002.
[4] Barja, G. Mitochondrial oxygen radical generation and leak:
sites of production in states 4 and 3, organ specificity, and
relation to aging and longevity. J. Bioenerg. Biomembr. 31:
347 366; 1999.
[5] Han, D.; Canali, R.; Rettori, D.; Kaplowitz, N. Effect of glutathione depletion on sites and topology of superoxide and hydrogen peroxide production in mitochondria. Mol. Pharmacol. 64:
1136 1144; 2003.
[6] Hansford, R.G.; Hogue, B. A.; Mildaziene, V. Dependence
of H2O2 formation by rat heart mitochondria on substrate
availability and donor age. J. Bioenerg. Biomembr. 29:89 95;
1997.

Mitochondrial superoxide and UCPs

[7] Liu, Y.; Fiskum, G.; Schubert, D. Generation of reactive oxygen
species by the mitochondrial electron transport chain. J. Neurochem. 80:780 787; 2002.
[8] Talbot, D. A.; Lambert, A. J.; Brand, M. D. Production of endogenous matrix superoxide from mitochondrial complex I leads
to activation of uncoupling protein 3. FEBS Lett. 556:111 115;
2004.
[9] Turrens, J. F.; Boveris, A. Generation of superoxide anion by the
NADH dehydrogenase of bovine heart mitochondria. Biochem.
J. 191:421 427; 1980.
[10] Venditti, P.; De Rosa, R.; Di Meo, S. Effect of thyroid state on
H2O2 production by rat liver mitochondria. Mol. Cell. Endocrinol. 205:185 192; 2003.
[11] Votyakova, T. V.; Reynolds, I. J. Dcm-dependent and -independent production of reactive oxygen species by rat brain mitochondria. J. Neurochem. 79:266 277; 2001.
[12] Miwa, S.; St-Pierre, J.; Partridge, L.; Brand, M. D. Superoxide
and hydrogen peroxide production by Drosophila mitochondria.
Free Radic. Biol. Med. 35:938 948; 2003.
[13] Lambert, A. J.; Brand, M. D. Superoxide production by NADH:ubiquinone oxidoreductase (complex I) depends on the pH
gradient across the mitochondrial inner membrane. Biochem. J.
(in press); 2004.
[14] Herrero, A.; Barja, G. Localization of the site of oxygen radical
generation inside the complex I of heart and nonsynaptic brain
mammalian mitochondria. J. Bioenerg. Biomembr. 32:609 615;
2000.
[15] Cadenas, E.; Boveris, A.; Ragan, C. I.; Stoppani, A. O. Production of superoxide radicals and hydrogen peroxide by NADH
ubiquinone reductase and ubiquinol cytochrome c reductase
from beef-heart mitochondria. Arch. Biochem. Biophys. 180:
248 257; 1977.
[16] Genova, M. L.; Ventura, B.; Giuliano, G.; Bovina, C.; Formiggini, G.; Parenti Castelli, G.; Lenaz, G. The site of production of
superoxide radical in mitochondrial Complex I is not a bound
ubisemiquinone but presumably iron sulfur cluster N2. FEBS
Lett. 505:364 368; 2001.
[17] Kushnareva, Y.; Murphy, A. N.; Andreyev, A. Complex I-mediated reactive oxygen species generation: modulation by cytochrome c and NAD(P)+ oxidation reduction state. Biochem. J.
368:545 553; 2002.
[18] Degli Esposti, M. Inhibitors of NADH ubiquinone reductase:
an overview. Biochim. Biophys. Acta 1364:222 235; 1998.
[19] Korshunov, S. S.; Skulachev, V. P.; Starkov, A. A. High protonic
potential actuates a mechanism of production of reactive oxygen
species in mitochondria. FEBS Lett. 416:15 18; 1997.
[20] Liu, S. S. Generating, partitioning, targeting and functioning of
superoxide in mitochondria. Biosci. Rep. 17:259 272; 1997.
[21] Boveris, A.; Cadenas, E.; Stoppani, A. O. Role of ubiquinone in
the mitochondrial generation of hydrogen peroxide. Biochem. J.
156:435 444; 1976.
[22] Turrens, J. F.; Alexandre, A.; Lehninger, A. L. Ubisemiquinone
is the electron donor for superoxide formation by complex III of
heart mitochondria. Arch. Biochem. Biophys. 237:408 414;
1985.
[23] Herrero, A.; Barja, G. Sites and mechanisms responsible for the
low rate of free radical production of heart mitochondria in the
long-lived pigeon. Mech. Ageing Dev. 98:95 111; 1997.
[24] Zhang, Z.; Huang, L.; Shulmeister, V. M.; Chi, Y. I.; Kim, K. K.;
Hung, L. W.; Crofts, A. R.; Berry, E. A.; Kim, S. H. Electron
transfer by domain movement in cytochrome bc1. Nature 392:
677 684; 1998.
[25] Iwata, S.; Lee, J. W.; Okada, K.; Lee, J. K.; Iwata, M.; Rasmussen, B.; Link, T. A.; Ramaswamy, S.; Jap, B. K. Complete
structure of the 11-subunit bovine mitochondrial cytochrome
bc1 complex. Science 281:64 71; 1998.
[26] McLennan, H. R.; Degli Esposti, M. The contribution of
mitochondrial respiratory complexes to the production of reactive oxygen species. J. Bioenerg. Biomembr. 32:153 162;
2000.
[27] Gredilla, R.; Barja, G.; Lopez-Torres, M. Effect of short-term
caloric restriction on H2O2 production and oxidative DNA dam-

[28]

[29]
[30]

[31]
[32]
[33]

[34]

[35]
[36]
[37]
[38]
[39]

[40]

[41]

[42]

[43]
[44]

[45]

[46]

[47]

765

age in rat liver mitochondria and location of the free radical

source. J. Bioenerg. Biomembr. 33:279 287; 2001.
Messner, K. R.; Imlay, J. A. Mechanism of superoxide and
hydrogen peroxide formation by fumarate reductase, succinate
dehydrogenase, and aspartate oxidase. J. Biol. Chem. 277:
42563 42571; 2002.
Sohal, R. S. Aging, cytochrome oxidase activity, and hydrogen
peroxide release by mitochondria. Free Radic. Biol. Med. 14:
583 588; 1993.
Drahota, Z.; Chowdhury, S. K.; Floryk, D.; Mracek, T.; Wilhelm,
J.; Rauchova, H.; Lenaz, G.; Houstek, J. Glycerophosphate-dependent hydrogen peroxide production by brown adipose tissue
mitochondria and its activation by ferricyanide. J. Bioenerg.
Biomembr. 34:105 113; 2002.
Sohal, R. S.; Weindruch, R. Oxidative stress, caloric restriction,
and aging. Science 273:59 63; 1996.
Shigenaga, M. K.; Hagen, T. M.; Ames, B. N. Oxidative damage
and mitochondrial decay in aging. Proc. Natl. Acad. Sci. USA
91:10771 10778; 1994.
Li, Y.; Huang, T. T.; Carlson, E. J.; Melov, S.; Ursell, P. C.;
Olson, J. L.; Noble, L. J.; Yoshimura, M. P.; Berger, C.; Chan,
P. H.; Wallace, D. C.; Epstein, C. J. Dilated cardiomyopathy and
neonatal lethality in mutant mice lacking manganese superoxide
dismutase. Nat. Genet. 11:376 381; 1995.
Lebovitz, R. M.; Zhang, H.; Vogel, H.; Cartwright, J., Jr.; Dionne, L.; Lu, N.; Huang, S.; Matzuk, M. M. Neurodegeneration,
myocardial injury, and perinatal death in mitochondrial superoxide dismutase-deficient mice. Proc. Natl. Acad. Sci. USA
93:9782 9787; 1996.
Halliwell, B.; Gutteridge, J. M. C. Free radicals in biology and
medicine. New York: Oxford Univ. Press; 1999.
Beckman, K. B.; Ames, B. N. The free radical theory of aging
matures. Physiol. Rev. 78:547 581; 1998.
Harman, D. Aging: a theory based on free radical and radiation
chemistry. J. Gerontol. 11:298 300; 1956.
Harman, D. The biologic clock: the mitochondria? J. Am. Geriatr. Soc. 20:145 147; 1972.
Ku, H. H.; Brunk, U. T.; Sohal, R. S. Relationship between
mitochondrial superoxide and hydrogen peroxide production
and longevity of mammalian species. Free Radic. Biol. Med.
15:621 627; 1993.
Barja, G.; Cadenas, S.; Rojas, C.; Perez-Campo, R.; LopezTorres, M. Low mitochondrial free radical production per unit
O2 consumption can explain the simultaneous presence of high
longevity and high aerobic metabolic rate in birds. Free Radic.
Res. 21:317 327; 1994.
Barja, G.; Herrero, A. Localization at complex I and mechanism
of the higher free radical production of brain nonsynaptic mitochondria in the short-lived rat than in the longevous pigeon.
J. Bioenerg. Biomembr. 30:235 243; 1998.
Brunet-Rossinni, A. K. Reduced free-radical production and
extreme longevity in the little brown bat (Myotis lucifugus)
versus two non-flying mammals. Mech. Ageing Dev. 125:
11 20; 2004.
Weindruch, R.; Walford, R. L. The retardation of aging and
disease by dietary restriction. Springfield (IL): Thomas; 1988.
Judge, S.; Judge, A.; Grune, T.; Leeuwenburgh, C. Short-term
CR decreases cardiac mitochondrial oxidant production but increases carbonyl content. Am. J. Physiol. 286:R254 R259;
2004.
Gredilla, R.; Sanz, A.; Lopez-Torres, M.; Barja, G. Caloric restriction decreases mitochondrial free radical generation at complex I and lowers oxidative damage to mitochondrial DNA in the
rat heart. FASEB J. 15:1589 1591; 2001.
Lambert, A. J.; Merry, B. J. Effect of caloric restriction on
mitochondrial reactive oxygen species production and bioenergetics: reversal by insulin. Am. J. Physiol. 286:R71 R79;
2004.
Sohal, R. S.; Ku, H. H.; Agarwal, S.; Forster, M. J.; Lal, H.
Oxidative damage, mitochondrial oxidant generation and antioxidant defenses during aging and in response to food restriction in
the mouse. Mech. Ageing Dev. 74:121 133; 1994.

766

M. D. BRAND et al.

[48] Merry, B. J. Molecular mechanisms linking calorie restriction

and longevity. Int. J. Biochem. Cell Biol. 34:1340 1354; 2002.
[49] Yu, B. P. Aging and oxidative stress: modulation by dietary
restriction. Free Radic. Biol. Med. 21:651 668; 1996.
[50] Sohal, R. S.; Sohal, B. H.; Brunk, U. T. Relationship between
antioxidant defenses and longevity in different mammalian species. Mech. Ageing Dev. 53:217 227; 1990.
[51] Tolmasoff, J. M.; Ono, T.; Cutler, R. G. Superoxide dismutase:
correlation with life-span and specific metabolic rate in primate
species. Proc. Natl. Acad. Sci. USA 77:2777 2781; 1980.
[52] Barja, G.; Cadenas, S.; Rojas, C.; Lopez-Torres, M.; Perez-Campo, R. A decrease of free radical production near critical targets
as a cause of maximum longevity in animals. Comp. Biochem.
Physiol. Biochem. Mol. Biol. 108:501 512; 1994.
[53] Seto, N. O.; Hayashi, S.; Tener, G. M. Overexpression of Cu Zn
superoxide dismutase in Drosophila does not affect life-span.
Proc. Natl. Acad. Sci. USA 87:4270 4274; 1990.
[54] Sun, J.; Folk, D.; Bradley, T. J.; Tower, J. Induced overexpression
of mitochondrial Mn-superoxide dismutase extends the life span
of adult Drosophila melanogaster. Genetics 161:661 672; 2002.
[55] Orr, W. C.; Mockett, R. J.; Benes, J. J.; Sohal, R. S. Effects of
overexpression of copper zinc and manganese superoxide dismutases, catalase, and thioredoxin reductase genes on longevity
in Drosophila melanogaster. J. Biol. Chem. 278:26418 26422;
2003.
[56] Melov, S.; Ravenscroft, J.; Malik, S.; Gill, M. S.; Walker, D. W.;
Clayton, P. E.; Wallace, D. C.; Malfroy, B.; Doctrow, S. R.;
Lithgow, G. J. Extension of life-span with superoxide dismutase/catalase mimetics. Science 289:1567 1569; 2000.
[57] Keaney, M.; Gems, D. No increase in lifespan in Caenorhabditis
elegans upon treatment with the superoxide dismutase mimetic
EUK-8. Free Radic. Biol. Med. 34:277 282; 2003.
[58] Bayne, A. C.; Sohal, R. S. Effects of superoxide dismutase/
catalase mimetics on life span and oxidative stress resistance
in the housefly, Musca domestica. Free Radic. Biol. Med. 32:
1229 1234; 2002.
[59] Negre-Salvayre, A.; Hirtz, C.; Carrera, G.; Cazenave, R.; Troly,
M.; Salvayre, R.; Penicaud, L.; Casteilla, L. A role for uncoupling protein-2 as a regulator of mitochondrial hydrogen peroxide generation. FASEB J. 11:809 815; 1997.
[60] Kowaltowski, A. J.; Costa, A. D.; Vercesi, A. E. Activation of
the potato plant uncoupling mitochondrial protein inhibits reactive oxygen species generation by the respiratory chain. FEBS
Lett. 425:213 216; 1998.
[61] Pastore, D.; Fratianni, A.; Di Pede, S.; Passarella, S. Effects of
fatty acids, nucleotides and reactive oxygen species on durum
wheat mitochondria. FEBS Lett. 470:88 92; 2000.
[62] Echtay, K. S.; Brand, M. D. Coenzyme Q induces GDP-sensitive
proton conductance in kidney mitochondria. Biochem. Soc.
Trans. 29:763 768; 2001.
[63] Esteves, T. C.; Echtay, K. S.; Jonassen, T.; Clarke, C. F.; Brand,
M. D. Ubiquinone is not required for proton conductance by
uncoupling protein 1 in yeast mitochondria. Biochem. J. 379:
309 315; 2004.
[64] Echtay, K. S.; Roussel, D.; St-Pierre, J.; Jekabsons, M. B.;
Cadenas, S.; Stuart, J. A.; Harper, J. A.; Roebuck, S. J.; Morrison, A.; Pickering, S.; Clapham, J. C.; Brand, M. D. Superoxide activates mitochondrial uncoupling proteins. Nature 415:
96 99; 2002.
[65] Echtay, K. S.; Murphy, M. P.; Smith, R. A.; Talbot, D. A.; Brand,
M. D. Superoxide activates mitochondrial uncoupling protein 2
from the matrix side: studies using targeted antioxidants. J. Biol.
Chem. 277:47129 47135; 2002.
[66] Cadenas, S.; Echtay, K. S.; Harper, J. A.; Jekabsons, M. B.; Buckingham, J. A.; Grau, E.; Abuin, A.; Chapman, H.; Clapham, J. C.;
Brand, M. D. The basal proton conductance of skeletal muscle
mitochondria from transgenic mice overexpressing or lacking uncoupling protein-3. J. Biol. Chem. 277:2773 2778; 2002.
[67] Krauss, S.; Zhang, C. Y.; Scorrano, L.; Dalgaard, L. T.; StPierre, J.; Grey, S. T.; Lowell, B. B. Superoxide-mediated activation of uncoupling protein 2 causes pancreatic beta cell dysfunction. J. Clin. Invest. 112:1831 1842; 2003.

[68] Echtay, K. S.; Esteves, T. C.; Pakay, J. L.; Jekabsons, M. B.;

Lambert, A. J.; Portero-Otin, M.; Pamplona, R.; Vidal-Puig,
A. J.; Wang, S.; Roebuck, S. J.; Brand, M. D. A signalling
role for 4-hydroxy-2-nonenal in regulation of mitochondrial
uncoupling. EMBO J. 22:4103 4110; 2003.
[69] Murphy, M. P.; Echtay, K. S.; Blaikie, F. H.; Asin-Cayuela, J.;
Cocheme, H. M.; Green, K.; Buckingham, J. A.; Taylor, E. R.;
Hurrell, F.; Hughes, G.; Miwa, S.; Cooper, C. E.; Svistunenko,
D. A.; Smith, R. A.; Brand, M. D. Superoxide activates uncoupling proteins by generating carbon-centered radicals and initiating lipid peroxidation: studies using a mitochondria-targeted spin
trap derived from a-phenyl-N-tert-butylnitrone. J. Biol. Chem.
278:48534 48545; 2003.
[70] Brand, M. D.; Buckingham, A. J.; Esteves, T. C.; Green, K.;
Lambert, A. J.; Miwa, S.; Murphy, M. P.; Pakay, J. L.; Talbot,
D. A.; Echtay, K. S. Mitochondrial superoxide and aging: uncoupling-protein activity and superoxide production. Biochem.
Soc. Symp. 71:203 213; 2004.
[71] Esterbauer, H.; Schaur, R. J.; Zollner, H. Chemistry and biochemistry of 4-hydroxynonenal, malonaldehyde and related aldehydes. Free Radic. Biol. Med. 11:81 128; 1991.
[72] Uchida, K. 4-Hydroxy-2-nonenal: a product and mediator of
oxidative stress. Prog. Lipid Res. 42:318 343; 2003.
[73] Parola, M.; Bellomo, G.; Robino, G.; Barrera, G.; Dianzani,
M. U. 4-Hydroxynonenal as a biological signal: molecular
basis and pathophysiological implications. Antioxid. Redox Signal. 1:255 284; 1999.
[74] Rial, E.; Gonzalez-Barroso, M.; Fleury, C.; Iturrizaga, S.; Sanchis, D.; Jimenez-Jimenez, J.; Ricquier, D.; Goubern, M.; Bouillaud, F. Retinoids activate proton transport by the uncoupling
proteins UCP1 and UCP2. EMBO J. 18:5827 5833; 1999.
[75] Crabb, J. W.; ONeil, J.; Miyagi, M.; West, K.; Hoff, H. F.
Hydroxynonenal inactivates cathepsin B by forming Michael
adducts with active site residues. Protein Sci. 11:831 840;
2002.
[76] Musatov, A.; Carroll, C. A.; Liu, Y. C.; Henderson, G. I.; Weintraub, S. T.; Robinson, N. C. Identification of bovine heart cytochrome c oxidase subunits modified by the lipid peroxidation
product 4-hydroxy-2-nonenal. Biochemistry 41:8212 8220;
2002.
[77] Ishii, T.; Tatsuda, E.; Kumazawa, S.; Nakayama, T.; Uchida, K.
Molecular basis of enzyme inactivation by an endogenous electrophile 4-hydroxy-2-nonenal: identification of modification
sites in glyceraldehyde-3-phosphate dehydrogenase. Biochemistry 42:3474 3480; 2003.
[78] Chen, J. J.; Bertrand, H.; Yu, B. P. Inhibition of adenine nucleotide translocator by lipid peroxidation products. Free Radic. Biol.
Med. 19:583 590; 1995.
[79] Zwizinski, C. W.; Schmid, H. H. Peroxidative damage to cardiac
mitochondria: identification and purification of modified adenine
nucleotide translocase. Arch. Biochem. Biophys. 294:178 183;
1992.
[80] Giron-Calle, J.; Schmid, H. H. Peroxidative modification of a
membrane protein: conformation-dependent chemical modification of adenine nucleotide translocase in Cu2+/tert-butyl hydroperoxide treated mitochondria. Biochemistry 35:15440 15446;
1996.
[81] Yan, L. J.; Sohal, R. S. Mitochondrial adenine nucleotide translocase is modified oxidatively during aging. Proc. Natl. Acad.
Sci. USA 95:12896 12901; 1998.
[82] Krauss, S.; Zhang, C. Y.; Lowell, B. B. A significant portion of
mitochondrial proton leak in intact thymocytes depends on expression of UCP2. Proc. Natl. Acad. Sci. USA 99:118 122;
2002.
[83] Vidal-Puig, A. J.; Grujic, D.; Zhang, C. Y.; Hagen, T.; Boss, O.;
Ido, Y.; Szczepanik, A.; Wade, J.; Mootha, V.; Cortright, R.;
Muoio, D. M.; Lowell, B. B. Energy metabolism in uncoupling
protein 3 gene knockout mice. J. Biol. Chem. 275:16258 16266;
2000.
[84] Brand, M. D.; Pamplona, R.; Portero-Otin, M.; Requena, J. R.;
Roebuck, S. J.; Buckingham, J. A.; Clapham, J. C.; Cadenas, S.
Oxidative damage and phospholipid fatty acyl composition in

Mitochondrial superoxide and UCPs

[85]

[86]

[87]
[88]

[89]

[90]

[91]

[92]
[93]
[94]

[95]

[96]

[97]
[98]

[99]

[100]

skeletal muscle mitochondria from mice underexpressing or overexpressing uncoupling protein 3. Biochem. J. 368:597 603;
2002.
Arsenijevic, D.; Onuma, H.; Pecqueur, C.; Raimbault, S.; Manning, B. S.; Miroux, B.; Couplan, E.; Alves-Guerra, M. C.;
Goubern, M.; Surwit, R.; Bouillaud, F.; Richard, D.; Collins,
S.; Ricquier, D. Disruption of the uncoupling protein-2 gene in
mice reveals a role in immunity and reactive oxygen species
production. Nat. Genet. 26:435 439; 2000.
Duval, C.; Negre-Salvayre, A.; Dogilo, A.; Salvayre, R.; Penicaud, L.; Casteilla, L. Increased reactive oxygen species production with antisense oligonucleotides directed against uncoupling
protein 2 in murine endothelial cells. Biochem. Cell Biol. 80:
757 764; 2002.
Allen, R. G.; Tresini, M. Oxidative stress and gene regulation.
Free Radic. Biol. Med. 28:463 499; 2000.
Stuart, J. A.; Harper, J. A.; Brindle, K. M.; Jekabsons, M. B.;
Brand, M. D. A mitochondrial uncoupling artifact can be caused
by expression of uncoupling protein 1 in yeast. Biochem. J. 356:
779 789; 2001.
Harper, J. A.; Stuart, J. A.; Jekabsons, M. B.; Roussel, D.; Brindle, K. M.; Dickinson, K.; Jones, R. B.; Brand, M. D. Artifactual
uncoupling by uncoupling protein 3 in yeast mitochondria at the
concentrations found in mouse and rat skeletal-muscle mitochondria. Biochem. J. 361:49 56; 2002.
Pecqueur, C.; Alves-Guerra, M. C.; Gelly, C.; Levi-Meyrueis, C.;
Couplan, E.; Collins, S.; Ricquier, D.; Bouillaud, F.; Miroux, B.
Uncoupling protein 2, in vivo distribution, induction upon oxidative stress, and evidence for translational regulation. J. Biol.
Chem. 276:8705 8712; 2001.
Talbot, D. A.; Hanuise, N.; Rey, B.; Rouanet, J. L.; Duchamp,
C.; Brand, M. D. Superoxide activates a GDP-sensitive proton
conductance in skeletal muscle mitochondria from king penguin
(Aptenodytes patagonicus). Biochem. Biophys. Res. Commun.
312:983 988; 2003.
Saleh, M. C.; Wheeler, M. B.; Chan, C. B. Uncoupling protein-2:
evidence for its function as a metabolic regulator. Diabetologia
45:174 187; 2002.
Rutter, G. A. Nutrient-secretion coupling in the pancreatic islet
h-cell: recent advances. Mol. Aspects Med. 22:247 284; 2001.
Chan, C. B.; De Leo, D.; Joseph, J. W.; McQuaid, T. S.; Ha,
X. F.; Xu, F.; Tsushima, R. G.; Pennefathner, P. S.; Salapatek,
A. M. F.; Wheeler, M. B. Increased uncoupling protein-2
levels in beta-cells are associated with impaired glucosestimulated insulin secretionmechanism of action. Diabetes
50:1302 1310; 2001.
Hong, Y.; Fink, B. D.; Dillon, J. S.; Sivitz, W. I. Effects of
adenoviral overexpression of uncoupling protein-2 and -3 on
mitochondrial respiration in insulinoma cells. Endocrinology
142:249 256; 2001.
Zhang, C. Y.; Baffy, G.; Perret, P.; Krauss, S.; Peroni, O.;
Grujic, D.; Hagen, T.; Vidal-Puig, A. J.; Boss, O.; Kim, Y.
X.; Zheng, X. X.; Wheeler, M. B.; Shulman, G. I.; Chan, C.
B.; Lowell, B. B. Uncoupling protein-2 negatively regulates
insulin secretion and is a major link between obesity, beta cell
dysfunction, and type 2 diabetes. Cell 105:745 755; 2001.
Green, K.; Brand, M. D.; Murphy, M. P. Prevention of mitochondrial oxidative damage as a therapeutic strategy in diabetes.
Diabetes 53 (Suppl. 1):S108 S110; 2004.
Koshkin, V.; Wang, X. L.; Scherer, P. E.; Chan, C. B.; Wheeler, M. B. Mitochondrial functional state in clonal pancreatic
beta-cells exposed to free fatty acids. J. Biol. Chem. 278:
19709 19715; 2003.
Itoh, Y.; Kawamata, Y.; Harada, M.; Kobayashi, M.; Fujii, R.;
Fukusumi, S.; Ogi, K.; Hosoya, M.; Tanaka, Y.; Uejima, H.;
Tanaka, H.; Maruyama, M.; Satoh, R.; Okubo, S.; Kizawa, H.;
Komatsu, H.; Matsumura, F.; Noguchi, Y.; Shinobara, T.; Hinuma, S.; Fujisawa, Y.; Fujino, M. Free fatty acids regulate insulin
secretion from pancreatic beta cells through GPR40. Nature 422:
173 176; 2003.
Sakai, K.; Matsumoto, K.; Nishikawa, T.; Suefuji, M.; Nakamaru, K.; Hirashima, Y.; Kawashima, J.; Shirotani, T.; Ichinose,

[101]

[102]

[103]

[104]
[105]
[106]

[107]

[108]

[109]
[110]

767

K.; Brownlee, M.; Araki, E. Mitochondrial reactive oxygen species reduce insulin secretion by pancreatic beta-cells. Biochem.
Biophys. Res. Commun. 300:216 222; 2003.
Gong, D. W.; Monemdjou, S.; Gavrilova, O.; Leon, L. R.;
Marcus-Samuels, B.; Chou, C. J.; Everett, C.; Kozak, L. P.;
Li, C.; Deng, C.; Harper, M. E.; Reitman, M. L. Lack of
obesity and normal response to fasting and thyroid hormone
in mice lacking uncoupling protein-3. J. Biol. Chem. 275:
16251 16257; 2000.
Blanc, J.; Alves-Guerra, M. C.; Esposito, B.; Rousset, S.;
Gourdy, P.; Ricquier, D.; Tedgui, A.; Miroux, B.; Mallat, Z.
Protective role of uncoupling protein 2 in atherosclerosis. Circulation 107:388 390; 2003.
Alves-Guerra, M. C.; Rousset, S.; Pecqueur, C.; Mallat, Z.;
Blanc, J.; Tedgui, A.; Bouillaud, F.; Cassard-Doulcier, A. M.;
Ricquier, D.; Miroux, B. Bone marrow transplantation reveals
the in vivo expression of the mitochondrial uncoupling protein 2
in immune and nonimmune cells during inflammation. J. Biol.
Chem. 278:42307 42312; 2003.
Mills, E. M.; Banks, M. L.; Sprague, J. E.; Finkel, T. Pharmacology: uncoupling the agony from ecstasy. Nature 426:403 404;
2003.
Niki, E. Antioxidants and atherosclerosis. Biochem. Soc. Trans.
32:156 159; 2004.
Shatrov, V. A.; Brune, B. Induced expression of manganese
superoxide dismutase by non-toxic concentrations of oxidized
low-density lipoprotein (oxLDL) protects against oxLDL-mediated cytotoxicity. Biochem. J. 374:505 511; 2003.
Murray, H. W.; Juangbhanich, C. W.; Nathan, C. F.; Cohn,
Z. A. Macrophage oxygen-dependent antimicrobial activity. II.
The role of oxygen intermediates. J. Exp. Med. 150:950 964;
1979.
Green, A. R.; Mechan, A. O.; Elliott, J. M.; OShea, E.; Colado,
M. I. The pharmacology and clinical pharmacology of 3,4methylenedioxymethamphetamine (MDMA, ecstasy). Pharmacol. Rev. 55:463 508; 2003.
Tomas, P.; Ledesma, A.; Rial, E. Photoaffinity labeling of the
uncoupling protein UCP1 with retinoic acid:ubiquinone favors
binding. FEBS Lett. 526:63 65; 2002.
Okun, J. G.; Lummen, P.; Brandt, U. Three classes of inhibitors
share a common binding domain in mitochondrial complex I
(NADH:ubiquinone oxidoreductase). J. Biol. Chem. 274:
2625 2630; 1999.

ABBREVIATIONS

AAPH2,2V-azobis(2-methylpropionamidine) dihydrochloride
ANT adenine nucleotide translocase
FeS iron sulfur
FMN flavin mononucleotide
GSIS glucose-stimulated insulin secretion
HNE 4-hydroxy-trans-2-nonenal
MDMA3,4-methylenedioxymethamphetamine
MitoPBN mitochondrially targeted derivative of a-phenyl-N-tert-butylnitrone ([4-[4-[[(1,1-dimethylethyl)oxidoimino]methyl]phenoxy]butyl]triphenylphosphonium bromide)
Q ubiquinone
ROS reactive oxygen species
SOD superoxide dismutase
TTNPB (E)-4-[2-(5,6,7,8-tetrahydro-5,5,8,8-tetramethyl-2-naphthalenyl)-1-propenyl]-benzoic acid
UCP uncoupling protein