A mitochondria-targeted macrocyclic Mn(II) superoxide dismutase mimetic

Article abstract of DOI:10.1016/j.chembiol.2012.08.005

Superoxide (O₂^s-) is the proximal mitochondrial reactive oxygen species underlying pathology and redox signaling. This central role prioritizes development of a mitochondria-targeted reagent selective for controlling O₂^s-. We have conjugated a mitochondria-targeting triphenylphosphonium (TPP) cation to a O₂ ^s--selective pentaaza macrocyclic Mn(II) superoxide dismutase (SOD) mimetic to make MitoSOD, a mitochondria-targeted SOD mimetic. MitoSOD showed rapid and extensive membrane potential-dependent uptake into mitochondria without loss of Mn and retained SOD activity. Pulse radiolysis measurements confirmed that MitoSOD was a very effective catalytic SOD mimetic. MitoSOD also catalyzes the ascorbate-dependent reduction of O₂^s-. The combination of mitochondrial uptake and O₂^s- scavenging by MitoSOD decreased inactivation of the matrix enzyme aconitase caused by O₂^s-. MitoSOD is an effective mitochondria-targeted macrocyclic SOD mimetic that selectively protects mitochondria from O₂^s- damage.

Full text of DOI:10.1016/j.chembiol.2012.08.005

Chemistry & Biology
Article
A Mitochondria-Targeted Macrocyclic
Mn(II) Superoxide Dismutase Mimetic
1
2
Geoffrey F. Kelso, Andrej Maroz, Helena M. Cocheme,3,4 Angela Logan,3 Tracy A. Prime,3 Alexander V. Peskin,5
´
Christine C. Winterbourn,5 Andrew M. James,3 Meredith F. Ross,3 Sally Brooker,6 Carolyn M. Porteous,7
Robert F. Anderson,2 Michael P. Murphy,3, and Robin A.J. Smith
*
1
1Centre for Green Chemistry, Monash University, Victoria 3800, Australia
2School of Chemical Sciences, The University of Auckland, Private Bag 92019, Auckland 1142, New Zealand
3Medical Research Council Mitochondrial Biology Unit, Hills Road, Cambridge CB2 0XY, UK
4Department of Genetics, Evolution, and Environment, Institute of Healthy Ageing, University College London, Gower Street,
London WC1E 6BT, UK
5Department of Pathology, University of Otago Christchurch, P.O. Box 4345, Christchurch, New Zealand
6Department of Chemistry
7Department of Biochemistry
University of Otago, PO Box 56, Dunedin 9054, New Zealand
*Correspondence: mpm@mrc-mbu.cam.ac.uk
http://dx.doi.org/10.1016/j.chembiol.2012.08.005
SUMMARY
against oxidative damage and explore redox signaling path-
ways, mitochondria-targeted antioxidants have been developed
Superoxide (O2,-) is the proximal mitochondrial that comprise a lipophilic triphenylphosphonium (TPP) cation
conjugated to an antioxidant moiety (Murphy and Smith, 2007;
Smith et al., 2011). The large lipophilic surface of the TPP cation
reactive oxygen species underlying pathology and
redox signaling. This central role prioritizes develop-
enables such compounds to move rapidly through biological
membranes, while their positive charge causes them to be
ment of a mitochondria-targeted reagent selective
for controlling O2,-. We have conjugated a mitochon-
accumulated several 100-fold inside mitochondria in vivo driven
dria-targeting triphenylphosphonium (TPP) cation to
by the large membrane potential of up to 180 mV (negative
inside) (Ross et al., 2005). This dramatic mitochondrial accumu-
a O2,--selective pentaaza macrocyclic Mn(II) super-
oxide dismutase (SOD) mimetic to make MitoSOD,
lation makes targeted antioxidants far more protective against
oxidative damage than untargeted equivalents in cultured cells,
a mitochondria-targeted SOD mimetic. MitoSOD
showed rapid and extensive membrane potential-
dependent uptake into mitochondria without loss
in animal models of pathologies, and in human clinical trials
(Gane et al., 2010; Murphy and Smith, 2007, 2010; Snow et al.,
of Mn and retained SOD activity. Pulse radiolysis 2010). A targeted version of ubiquinone (MitoQ) is the most
extensively studied of these mitochondria-targeted antioxidants
(Kelso et al., 2001; Murphy and Smith, 2007, 2010). Many other
measurements confirmed that MitoSOD was a
very effective catalytic SOD mimetic. MitoSOD also
antioxidants have been targeted to mitochondria by conjugation
to the TPP cation, including vitamin E (Smith et al., 1999), plasto-
quinone (Skulachev et al., 2009), Ebselen (Filipovska et al., 2005),
catalyzes the ascorbate-dependent reduction of
O2,-. The combination of mitochondrial uptake and
O2,- scavenging by MitoSOD decreased inactivation
lipoic acid (Brown et al., 2007), nitroxides (Dhanasekaran et al.,
2005; Trnka et al., 2008), and nitrones (Murphy et al., 2003).
,-
of the matrix enzyme aconitase caused by O2
.
MitoSOD is an effective mitochondria-targeted
macrocyclic SOD mimetic that selectively protects
mitochondria from O2,- damage.
,-
The central role of O2 in mitochondrial metabolism requires
the development of a mitochondria-targeted antioxidant that is
selective for O2,-. A number of mitochondria-targeted antioxi-
dants do react with O2,-. The nitroxide MitoTEMPOL has SOD
activity in vitro but is rapidly reduced to the hydroxylamine within
mitochondria, and thus its protective properties are not attribut-
INTRODUCTION
,-
able to its reactions with O2 (Trnka et al., 2008). The related
The mitochondrial respiratory chain is a major source of super- molecule, incorporating an amide-linked TPP, MitoTEMPO,
oxide (O2,-), which can go on to form other reactive oxygen also has SOD activity in vitro, as well as protective effects in vivo
species (ROS), such as hydrogen peroxide (H2O2), the hydroxyl (Dikalova et al., 2010); however, its biological activity is unlikely
radical (,OH), and peroxynitrite (ONOO-) (Murphy, 2009). The to be due to its SOD activity as it is also reduced to the hydrox-
production of O2,- and derived ROS leads to oxidative damage ylamine in vivo (Soule et al., 2007). The mitochondria-targeted
that disrupts mitochondrial function and thereby contributes to nitroxide Mito-Carboxy Proxyl (MitoCP) protects against oxida-
a range of pathologies, including neurodegenerative diseases, tive damage to mitochondria far more effectively than the untar-
,-
diabetes, and ischemia-reperfusion injury (Finkel, 2005). Mito- geted version and has a O2 dismutation rate of ꢀ105 Mꢁ1 sꢁ1
chondrial ROS production is also an important component of in vitro (Dhanasekaran et al., 2005); however, it is likely that
redox signaling (Collins et al., 2012; Murphy, 2009). To protect protection by MitoCP is largely due to other antioxidant reactions
Chemistry & Biology 19, 1237–1246, October 26, 2012 ª2012 Elsevier Ltd All rights reserved 1237

Chemistry & Biology
MitoSOD
,-
NHAc
1.
it would be difficult to distinguish the effects of depleting O2
from that of degrading H2O2 (Doctrow et al., 1997).
A
1. PBr5
2. MeOH
3. NaBH4
Br
Br
+
N
O
BF4
O
Although there are a number of antioxidant compounds tar-
geted to mitochondria that do react with O2,-, they are not
O
O
N
N
2. ethylene glycol,
p-TSA
,-
HOOC
N
COOH
selective, and furthermore their rates of reaction with O2 are
OH
OH
O
O
typically ꢀ105 Mꢁ1sꢁ1, considerably lower than that of MnSOD,
the mitochondrial SOD isoform (ꢀ2 3 109 Mꢁ1 sꢁ1; Cadenas and
Davies, 2000). As the biological effects of these compounds are
unlikely to be due solely to their reactivity with O2,-, they cannot
5
6
7
P+
Br
Br
1.HO
NaH
SH,
S
N
S
Ph3P
,-
be used to infer the role of mitochondrial O2 in oxidative
damage or redox signaling. Consequently, there is a need to
2. HBr, H2SO4
N
,-
O
O
O
O
modulate mitochondrial O2 selectively using a synthetic mito-
chondria-targeted SOD mimetic. The importance of this chal-
lenge is supported by the observation that selectively decreasing
8
4
R
N
Cl
1. MnCl2
2. 4 or 2,6-pyridinedialdehyde
NH2 H2N
,-
mitochondrial O2 by overexpressing MnSOD enzyme gives
3. NaBH4
4. sat. NaCl
Cl
N
H
N
H
therapeutically relevant protection in animal models of disease
(Chen et al., 1998) and modulates redox signaling (Kim et al.,
2004; Murphy, 2009). To address this need, we have synthesized
a mitochondria-targeted SOD mimetic, MitoSOD (1), based on
the established pentaaza macrocyclic Mn(II) SOD mimetic
M40403 (2) (Salvemini et al., 1999, 2002) (Figure 1A). This mole-
cule was chosen from a wide variety of known Mn-based SOD
NH HN
Mn
HN OH
2 NH
3
+
1, R = S(CH2)3PPh3 Cl
2, R = H
B
ꢀ
mimetics (Batinic-Haberle et al., 2010; Iranzo, 2011) because it
P(1)
is an effective SOD mimetic in vitro and has been shown to be
stable and protective in vivo (Thompson et al., 2010). Our
previous study of M40403 reaction kinetics using pulse radiol-
ysis demonstrated that in a phosphate environment it catalyzes
O2,- dismutation to H2O2 with kcat = 3.55 3 106 Mꢁ1 sꢁ1 at 22ꢂC
(pH 7.4) (Maroz et al., 2008), in good agreement with the value
more recently reported using the stopped flow method (2.3
0.1 3 106 Mꢁ1sꢁ1 [pH 7.4], 21ꢂC; Friedel et al., 2012). The cata-
lytic SOD cycle involves an initial fast reaction of M40403 with
O2,- (k1 = 9.0 0.5 3 107 Mꢁ1 sꢁ1, 22ꢂC [pH 8.0]) to produce a
S(1)
N(1)
O(80)
Cl(1)
N(2)
N(5)
Cl(3)
Mn(1)
Cl(2)
[MnIII(L)ClnO2](2-n)+ intermediate containing
a
coordinated
peroxide (Maroz et al., 2008). This rapid reaction underlies the
O(1)
N(3)
N(4)
,-
O2 selectivity of M40403 and its lack of reactivity with other
ROS, such as ONOO-, H2O2 or HOCl (Salvemini et al., 2002).
Whereas a triethylenetetraamine analog of M40403 reacts with
O(70)
5
ꢀ
NO (Filipovic et al., 2008, 2010), the rate constant is about 10
Figure 1. Synthesis and Structure of MitoSOD
times smaller than that for the initial reaction of M40403 with
,-
(A) MitoSOD (1) was obtained by Mn(II) template cyclization of the tetraamine
3 and the 2,6-pyridine dialdehyde derivative 4 followed by in situ borohydride
reduction. Compound 4 was obtained from chelidamic acid (5) via the bro-
modiol (6) and oxidized derivatives (7, 8). M40403 (2) was obtained similarly
from 3 and 2,6-pyridine dialdehyde.
(B) A projection view of MitoSOD [Mn(II)(L+)(Cl)(OH2)](Cl)2$CH3CH2OH (1(OH2)$
CH3CH2OH) derived from single crystal X-ray diffraction data. The dashed
lines represent hydrogen bonds. O(70) and O(80) represent the oxygen atoms
of half occupancy ethanol molecules. O(1) represents the oxygen atom of
a coordinated H2O molecule. Only the major occupancy site of Cl(2) is shown.
Hydrogen atoms, other than those on chiral centers or involved in hydrogen
bonding, have been omitted for clarity.
ꢀ
O2 . The SOD mimetic Mn(III) metalloporphyrins (Batinic-
Haberle, 2002) have high rates of reaction with O2,-, but they
are not chemoselective as they also react at comparable rates
with ONOO- and CO3,- (Salvemini et al., 2002). The Mn(III)-salen
compounds have SOD activity but also show catalase activity
(Doctrow et al., 1997). Here, we report the synthesis of a mito-
chondria-targeted SOD mimetic, MitoSOD, that is accumulated
,-
by mitochondria and selectively degrades mitochondrial O2
,
protecting against oxidative damage. The chemical and biolog-
ical properties of MitoSOD establish an opportunity to control
and assess O2,- production within mitochondria.
See also Tables S1–S3.
RESULTS AND DISCUSSION
of this nitroxide and its hydroxylamine reaction products (Kagan
et al., 2007). A TPP-conjugated version of the salen-Mn(III) Design and Synthesis of MitoSOD
compound EUK-134 has only limited protection against oxida- MitoSOD was designed by attaching a TPP lipophilic cation to
tive damage; however, its uptake into mitochondria was not the pyridine ring of M40403 so as to drive its selective accu-
,-
demonstrated (Dessolin et al., 2002). Furthermore, as this class mulation and ultimate dismutation of O2 within mitochondria.
of Mn(III) complex has both SOD and catalase mimetic activity, MitoSOD (1) and the untargeted SOD mimetic macrocycle
1238 Chemistry & Biology 19, 1237–1246, October 26, 2012 ª2012 Elsevier Ltd All rights reserved

Chemistry & Biology
MitoSOD
A
60
50
40
30
20
10
0
Table 1. Superoxide Dismutase Activity of MitoSOD and M40403
Measured by Pulse Radiolysis
kcat 3 10ꢁ6 (Mꢁ1 sꢁ1
)
6 µM
4 µM
Compound
MitoSOD
M40403
pH 7.4
pH 8.0
2.5 0.2
5.3 0.2
2.6 0.2
3.6 0.2
2 µM
1 µM
peroxide, as has been demonstrated for the parent M40403
macrocycle (Maroz et al., 2008). When catalytic (1–6 mM)
amounts of MitoSOD were present at pH 8.0, reflecting that
20
30
40
50
60
[O2•-] (µM)
,-
within mitochondria, the rate of O2 decay (20–60 mM) was
,-
considerably faster compared to O2 alone and also followed
B
40
30
20
10
0
mixed first- and second-order kinetics, which indicated that
,-
O2 was decaying by spontaneous (kself) and MitoSOD-cata-
lyzed (kobs) dismutation. To determine kobs, the first half-life
(t1/2) of O2,- decay was measured over a range of O2,- concentra-
tions with different concentrations of MitoSOD, and then plots of
1/t1/2 versus [O2,-] were constructed (Figure 2A), and values of
kobs were calculated from the y-intercepts, as previously
described (Anderson et al., 2003). A plot of kobs versus MitoSOD
concentration was linear (Figure 2B), demonstrating that the
decay of O2,- was first order in both O2,- and MitoSOD, indicative
of catalytic dismutation. The rate constant for the catalyzed
dismutation (kcat) was calculated from the slope of the fitted
function and is presented in Table 1 along with the value
measured at pH 7.4 and our previously determined values for
M40403 (Maroz et al., 2008). MitoSOD and M40403 were found
to have similar SOD activities, but that of MitoSOD decreased
at pH 7.4 (Table 1), presumably because of a pH-dependent
effect of the (3-mercaptopropyl) triphenylphosphonium substit-
uent on the reactivity of the intermediate Mn(III) complex. In
summary, MitoSOD is an effective SOD mimetic, and its initial
0
1
2
3
4
5
6
[MitoSOD] (µM)
Figure 2. Kinetics of MitoSOD Activity Measured by Pulse Radiolysis
,-
(A) Kinetic plots showing the dependence of the reciprocal first half-life of O2
,-
on the concentration of O2 at pH 8.0, measured at 270 nm, in the absence
(control) and with increasing concentrations of MitoSOD. Data were obtained
following the pulse radiolysis (35–100 Gy in 1.5 ms) of oxygen-saturated
aqueous solutions containing sodium formate (10 mM), phosphate buffer
(10 mM), and EDTA (0.1 mM).
,-
(B) Dependence of rate constants for the reaction of MitoSOD with O2
(derived from intercepts of A) on the concentration of MitoSOD. The slope of
the fitted line provides the value for kcat shown in Table 1.
,-
fast reaction with O2 underlies the selectivity for this ROS in
a biological environment.
M40403 (2) were synthesized as shown in Figure 1A, and the
structure of MitoSOD was determined by X-ray crystallography
(Figure 1B). The structure of MitoSOD was similar to that of
M40401, the active dimethyl analog of M40403 (Aston et al.,
2001), and the addition of a TPP cation appendage did not distort
the macrocycle conformation, which is crucial for SOD-mimetic
activity.
To evaluate whether MitoSOD could also act as a SOD
,-
mimetic under the steady-state O2 flux that occurs within
,-
mitochondria, O2 was generated continuously by xanthine
oxidase (XO) and monitored by the rate of cytochrome c reduc-
,-
tion. The reactivity of MitoSOD with O2 was determined from
its competition with cytochrome c reduction (Figures 3A and
3B) (Loschen et al., 1974). These experiments showed that
MitoSOD effectively blocked cytochrome c reduction in a
dose-dependent manner by competing for O2,-, whereas a
Kinetics of Superoxide Dismutation by MitoSOD
To determine whether MitoSOD was an effective SOD mimetic, control TPP cation without any bioactive cargo, methyltriphenyl-
we measured its rate of catalytic dismutation of O2,-. This was phosphonium bromide (TPMP), was ineffective (Figures 3A and
done by generating O2,- (1–60 mM in 1.5 ms) by the pulse radiol- 3B). A plot of the reciprocal of the cytochrome c reduction rate
ysis of phosphate buffer (pH 7.4 or 8.0), containing sodium versus compound concentration from Figure 3B was linear
,-
formate and EDTA, and measuring the decay of O2 spectro- and, by slope comparison, indicated that 1.1 nmol MitoSOD
photometrically (Bielski et al., 1985). In the absence of MitoSOD, corresponded to 1 U Cu, ZnSOD. In a separate approach, we
,-
O2 decay followed mixed first- and second-order kinetics, assessed the ability of MitoSOD to intercept a steady-state
,-
which is interpreted as the spontaneous dismutation of O2,- (kself
and background catalytic decomposition, presumably by trace Winterbourn, 2000). Here, we found that to get 50% inhibition
)
flux of O2 using the WST-1 reduction assay (Peskin and
,-
metal impurities. When [MitoSOD] [ [O2,-], the decay of O2
of WST-1 reduction required 2.4 0.1 ng (0.14 pmol) Cu,ZnSOD
28 ng (197 pmol) M40403; and 10 5 ng
was predominantly pseudo-first-order because of the fast reac- monomer; 96
tion of the macrocycle with O2,- (k1 = 1.57 0.04 3 108 Mꢁ1 sꢁ1 (20 pmol) MitoSOD (means SD, n = 3). So by this assay, Mito-
[pH 8.0]). We propose that the product of this reaction is SOD is ꢀ10-fold more effective than M40403 and ꢀ140-fold
a [MnIII(L-TPP)ClnO2](3-n)+ intermediate containing a coordinated less effective than Cu,ZnSOD.
Chemistry & Biology 19, 1237–1246, October 26, 2012 ª2012 Elsevier Ltd All rights reserved 1239

Chemistry & Biology
MitoSOD
Figure 3. Steady-State Activity of MitoSOD
(A) O2,- was generated by the addition of xanthine
oxidase (XO). Where indicated (second arrow)
MitoSOD (2–15 mM) or control compound (meth-
yltriphenylphosphonium bromide; TPMP, 15 mM)
was added, and the rate of reduction of cyto-
A
B
100
TPMP
2
80
60
40
20
0
compound
5
[MitoSOD]
(µM)
,-
chrome c by O2 was assessed at 550 nm. Data
15
shown are typical traces of experiments repeated
in triplicate.
XO
(B) The slope of cytochrome c reduction by XO/
hypoxanthine in the presence of the indicated
additions was divided by that in the absence of
additions to give the relative SOD activity. Data are
means SD, n = 3.
0.5
1
2
5
10 15
1 min
[MitoSOD] (µM)
(C) Ascorbate oxidation and cytochrome c reduc-
tion in the presence of M40403 and MitoSOD.
C
D
Ascorbate oxidation
Cyt c reduction
,-
[MnII(L-TPP)Cln](3-n)+
100
A flux of O2 was generated by addition of XO
H2O2
fast
O2•-
and acetaldehyde (10 mM), and then Cu,ZnSOD
(60 nM) or various concentrations of M40403 and
MitoSOD were added. Then either cytochrome c
was added and its rate of reduction was monitored
at 550 nm, or ascorbate 30 mM was added and
its rate of oxidation was monitored at 266 nm.
Note that MitoSOD has a greater effect on cyto-
chrome c reduction in (C) than in (A) and (B)
because the flux of superoxide generated is about
3-fold lower in (C). Data are means ranges for 2–3
determinations.
80
60
40
20
0
k1 = 1.57 x
108 M-1 s-1
ascorbyl
radical ascorbate
2H+
[MnII(L-TPP)ClnO2](1-n)+
[MnIII(L-TPP)ClnO2](2-n)+
rate-determining
kcat = 5.3 x 106 M-1 s-1
1
2
1 10
1
2
5 10
(D) The proposed mechanism for the MitoSOD-
catalyzed dismutation of O2,- is shown. In addition,
an alternative path for the rate-limiting reduction of
the Mn(III) species to Mn(II) by ascorbate is shown
as a dashed line.
O2
O2•-
The Mn(III) porphyrin SOD mimetics were more effective at and as an ascorbate/superoxide oxidoreductase catalyst,
,-
protecting against O2,- within cells than their in vitro SOD activity thereby enhancing their ability to catalytically decompose O2
would predict (Faulkner et al., 1994). Because SOD catalysis by The overall kinetic scheme supported by these data and our
.
M40403 involves a rate-limiting reduction of an intermediate previous study (Maroz et al., 2008) is presented in Figure 3D.
Mn(III) complex by another O2,- (Maroz et al., 2008), in a biolog- To summarize, MitoSOD-catalyzed dismutation of O2,- involves
,-
ical environment this intermediate could be reduced by reduc- an initial fast inner sphere oxidation of the macrocycle by O2
tants other than O2,-. The mitochondrial matrix contains a signif- (k1 = 1.57 0.04 3 108 Mꢁ1 sꢁ1 [pH 8.0]) to form an [MnIII(L-
icant amount of ascorbate (ꢀ0.5 mM) (Li et al., 2001); therefore, TPP)ClnO2](2-n)+ intermediate that then undergoes a relatively
,-
we indirectly assessed ascorbate reduction of the MitoSOD and slow reduction (kcat) with further O2 to give a putative short-
M40403 Mn(III) intermediates by measuring ascorbate oxida- lived [MnII(L-TPP)ClnO2](1-n)+ intermediate that rapidly decom-
tion in the presence of SOD mimetic and O2,- (Figure 3C). As ex- poses to the parent [MnII(L-TPP)Cln](3-n)+ macrocycle (Maroz
pected from the rate constant (3 3 105
M
ꢁ1sꢁ1) for the direct et al., 2008). The essential observation is that MitoSOD has
,-
reaction of ascorbate with O2 (Gotoh and Niki, 1992), when similar SOD activity to M40403, demonstrating the structural
we generated O2,- by XO there was rapid oxidation of ascorbate. modification incorporated to facilitate transport into mitochon-
However, in contrast to the more catalytically active Cu, Zn SOD dria did not disrupt the desired catalytic activity. In addition,
,-
enzyme, neither MitoSOD nor M40403 decreased ascorbate MitoSOD and M40403 catalyze the reduction of O2 to H2O2
oxidation (Figure 3C). This result suggests that in the presence using electrons from ascorbate (Faulkner et al., 1994), as is
of O2,-, the Mn(III) intermediate within the catalytic cycle of illustrated in Figure 3D. We conclude that MitoSOD should be
,-
MitoSOD or M40403 SOD activity oxidizes ascorbate (Fig- effective at selectively catalyzing the degradation of O2 within
ure 3D), as the rate constants for the initial reaction of O2,- with mitochondria.
MitoSOD (k1 = 1.38 0.08 3 108 [pH 7.2]) or M40403 (8.9
0.4 3 107 Mꢁ1 sꢁ1 [pH 7.4]) (Maroz et al., 2008) are much higher Stability, Charge, and Hydrophobicity of MitoSOD
,-
than that of O2 with ascorbate. Neither GSH nor NADH were To be biologically useful MitoSOD has to be stable. In 1 M HCl
,-
able to accelerate the steady-state dismutation of O2 by MitoSOD rapidly dissociated to metal-free macrocycle and
MitoSOD when measured as in Figure 3A (data not shown), Mn(II), as detected by positive ion electrospray mass spectrom-
suggesting these biological-reducing agents do not react etry (data not shown). The rates of MitoSOD dissociation from
with the Mn(III) intermediate. In summary our results suggest pH 3.8 to 4.3 were modest and had a linear dependence on
that in a biological environment MitoSOD and M40403 will [H+] (Figures S1A and S1B available online) (Aston et al., 2001),
,-
selectively decompose O2 by acting as both a SOD mimetic as reported for M40403 (Salvemini et al., 1999). This gave kdis
1240 Chemistry & Biology 19, 1237–1246, October 26, 2012 ª2012 Elsevier Ltd All rights reserved

Chemistry & Biology
MitoSOD
[MnII(L-TPP)Cl2]+. The molar conductivity of MitoSOD in water
was 294 S cm2 molꢁ1 at 20ꢂC, indicative of a 1:3 electrolyte (Fel-
tham and Hayter, 1964), indicating that both chloro ligands and
the TPP chloride counter anion dissociated to give [MnII(L-
TPP)]3+. However, when MitoSOD in 90% acetonitrile was
infused into a mass spectrometer, only species at m/z = 391
and 817 were detected, corresponding to [MnII(L-TPP)Cl]2+
and [MnII(L-TPP)Cl2]+, and there was no peak at m/z = 249 as
expected for [MnII(L-TPP)]3+ (Figure S2A). This suggests that
under biological conditions of [Clꢁ] (ꢀ120 mM), the [MnII(L-
TPP)Cl]2+ and [MnII(L-TPP)Cl2]+ forms will predominate. To
extend this we assessed the effect of [MitoSOD] on the potential
of an ion-selective electrode sensitive to TPP cations (Asin-
Cayuela et al., 2004). In 120 mM KCl, the electrode potential
was a linear function of log10[MitoSOD] with a slope of 29.7
2.3 mV/decade (Figure 4A), consistent with a Nernstian response
of the electrode to a dication. Together, these data suggest
that under biological conditions the predominant form of
MitoSOD is [MnII(L-TPP)Cl]2+ although [MnII(L-TPP)Cl2]+ is also
likely to be present.
mitochondria
A
B
-10
Slope = 29.7 2.3 mV
succinate
-20
-30
-40
-50
5x 1 µM MitoSOD
1 min
FCCP
-5.5
-5.0
-4.5
Log10 [MitoSOD]
5
4
3
2
1
0
C
D
1600
1200
800
400
0
[3H]-MitoSOD
[3H]-MitoSOD + FCCP
The uptake of lipophilic cations across phospholipid bilayers
is largely determined by the balance between their charge and
hydrophobicity (Ross et al., 2005). The charge raises the Born
energy for movement into the bilayer, and this is counteracted
by increasing hydrophobicity. Consequently, only relatively
hydrophobic TPP-conjugated compounds with low charge are
readily accumulated by mitochondria (Ross et al., 2005). Mito-
SOD was more lipophilic than the parent macrocycle M40403,
as shown by their octanol/PBS partition coefficients of 2.2
0.03 and 0.63 0.03, respectively (n = 3, means SD). The
MitoSOD partition coefficient is higher than that for many TPP
monocations, such as TPMP (partition coefficient = 0.33), which
is readily taken up into mitochondria, indicating that the monoca-
tion form of MitoSOD, [MnII(L-TPP)Cl2]+, should be taken up by
energized mitochondria. The double charge on the dication
form of MitoSOD, [MnII(L-TPP)Cl]2+, raises its Born energy
considerably, and studies of bisTPP dications have shown that
they have to be more hydrophobic than the corresponding
monocation to be taken up by mitochondria (Ross et al., 2006,
2005). As lipophilic TPP dications with a partition coefficient
>0.8 are taken up (Ross et al., 2006), the doubly charged form
of MitoSOD, [MnII(L-TPP)Cl]2+, may be sufficiently hydrophobic
to pass through the mitochondrial inner membrane. However,
the form of MitoSOD present within the octanol phase may
be the moncation, and consequently it is not possible to
conclude from partition coefficient measurements alone that
the MitoSOD dication is sufficiently hydrophobic to cross a
[3H]-M40403
0
5
10
15
0
200
400
600
[3H]-TPMP ACR
Time (min)
Figure 4. Uptake of MitoSOD by Energized Mitochondria
(A) The ion-selective electrode response to ten sequential aliquots of 3 mM
MitoSOD at 30ꢂC was plotted versus log10[MitoSOD]. The straight line repre-
sents a linear function fitted to the data by least-squares regression analysis.
(B) Ion-selective electrode measurement of MitoSOD uptake by energized
mitochondria. Incubation medium supplemented with rotenone (4 mg mlꢁ1
)
was added to the incubation chamber at 37ꢂC, and the electrode was then
calibrated by five sequential additions of 1 mM MitoSOD. Liver mitochondria
(2 mg protein mlꢁ1) were then added, followed by succinate (2 mM) and FCCP
(500 nM) where indicated.
(C) Time dependence of uptake of [3H]MitoSOD and [3H]M40403 by isolated
mitochondria. Rat liver mitochondria were incubated with 5 mM [3H]MitoSOD
or [3H]-M40403 at 25ꢂC and the uptake determined at various times. The data
are means of three separate experiments SEM. For the FCCP experiment,
500 nM FCCP was present.
(D) The effect of the mitochondrial membrane potential on the uptake of [3H]-
MitoSOD. Mitochondria were incubated with 800 nM MitoSOD and 1 mM
TPMP and supplemented with 1–16 mM malonate or FCCP (330 nM). Then
either 200 nM [3H]MitoSOD or 100 nM [3H]TPMP was added, and the ACR
values of [3H]MitoSOD or [3H]TPMP were determined in parallel. The data
are means of duplicate incubations of three separate experiments SEM. The
equation of the line is y = 1.727X + 415, R2 = 0.98.
See also Figure S1.
for MitoSOD = 2.64 0.11 3 10ꢁ2 Mꢁ1 sꢁ1 at 18ꢂC, correspond- biological membrane. We conclude that MitoSOD is stable to
ing to a half-life at pH 7.2 of z13 years. This was corroborated by dissociation under typical biological pH and exists predomi-
incubating 150 mM MitoSOD at pH 8.0, which, by high-perfor- nantly as [MnII(L-TPP)Cl]2+ with small amounts of [MnII(L-TPP)
mance liquid chromatography monitoring, showed that it was Cl2]+. In addition, the monocation and possibly the dication
stable for at least two weeks. In contrast, at pH 1.0 the half-life are sufficiently hydrophobic to be taken up across biological
is ꢀ3.8 min. Therefore, MitoSOD is stable under most biological membranes.
conditions but would dissociate rapidly in stomach acid.
For MitoSOD to act as a mitochondria-targeted antioxidant, it Uptake of MitoSOD by Energized Mitochondria
must pass easily through phospholipid bilayers, which will For MitoSOD to be an effective mitochondria-targeted SOD
depend critically on its overall charge. The Mn(II) cation is bound mimetic, it must be selectively accumulated by mitochondria
to the macrocycle (L-TPP) and 0, 1, or 2 chloro ligands, giving in response to the membrane potential. To assess this, we
three possible species: [MnII(L-TPP)]3+, [MnII(L-TPP)Cl]2+, or used an ion-selective electrode that is responsive to lipophilic
Chemistry & Biology 19, 1237–1246, October 26, 2012 ª2012 Elsevier Ltd All rights reserved 1241

Chemistry & Biology
MitoSOD
TPP cations (Asin-Cayuela et al., 2004) and that responds to membrane than for TPMP, as expected from its greater hydro-
MitoSOD (Figure 4A). The electrode response was calibrated phobicity (partition coefficient: MitoSOD, 2.2; TPMP, 0.33).
by additions of 1 to 5 mM MitoSOD, and then unenergized mito-
It is well established that the uptake of TPMP reflects the mito-
chondria were added, resulting in a lowering of the solution chondrial membrane potential and is satisfactorily described
concentration of MitoSOD (Figure 4B). The observation of exten- by the Nernst equation for a monocation. The uptake of
sive binding of MitoSOD to the unenergized mitochondria is MitoSOD (Figure 4D) is about twice that of TPMP. However, if
consistent with the behavior of other hydrophobic TPP cations MitoSOD crosses the membrane as a dication then, from the
(Asin-Cayuela et al., 2004). Addition of the respiratory substrate Nernst equation, its uptake would be expected to be orders of
succinate generated a membrane potential across the inner magnitude greater than that of TPMP (Ross et al., 2006). Further-
membrane, leading to uptake of MitoSOD into the mitochondria, more, the ACR for MitoSOD is a linear function of that for TPMP,
as indicated by its loss from the incubation medium, and this consistent with the uptake of MitoSOD as a monocation. Earlier
uptake was reversed when the uncoupler carbonyl cyanide-p- we showed that the dication form of MitoSOD predominates
trifluoromethoxyphenylhydrazone (FCCP) was added to dissi- in biological solutions. For a dication to be taken up across
pate the membrane potential (Figure 4B). In this experiment a phospholipid bilayer the high Born energy for transport due
ꢀ9 nmol of MitoSOD was bound to the unenergized mitochon- to the double charge must be balanced by elevated hydropho-
dria, and on energization a further ꢀ4.5 nmol of MitoSOD was bicity. The octanol/PBS partition coefficient for MitoSOD was,
taken up, corresponding to an intramitochondrial MitoSOD by comparision with other lipohilic dications, sufficient for it to
concentration of ꢀ1.5 mM, assuming an intramitochondrial be taken up into mitochondria. However, a limitation to this
volume of 0.5 ml mg proteinꢁ1 (Brown and Brand, 1985). As the comparision is that the form of MitoSOD that partitions into the
external concentration of MitoSOD at this point was then octanol phase may be the monocation, in which case the dica-
ꢀ0.5 mM, this indicates an accumulation of approximately tion form of MitoSOD may be too hydrophilic to pass through
3,000-fold, consistent with membrane potential-dependent a phospholipid bilayer. This possibility, in conjunction with the
uptake and binding of MitoSOD within mitochondria.
data from Figure 4, lead us to conclude that MitoSOD is taken
To further assess MitoSOD uptake by mitochondria and up by mitochondria in response to the membrane potential
compare it with that of M40403, we next investigated the uptake and, even though the dominant form of MitoSOD in solution is
of [3H]MitoSOD and [3H]M40403. When [3H]MitoSOD was a dication, the monocation is likely to be the major form passing
incubated with energized mitochondria, there was rapid and through the membrane bilayer.
extensive uptake of MitoSOD into mitochondria that was stable
For MitoSOD to be an effective SOD mimetic within mito-
for up to 15 min and was decreased by incubation with the chondria, it is important that it is taken up as the intact Mn(II)
uncoupler FCCP (Figure 4C). Conversely, when [3H]M40403 macrocycle. The apo form of MitoSOD, in which Mn has been
was incubated with energized mitochondria, there was negligible removed by acid treatment, was found to be readily taken up
accumulation of this macrocycle into the mitochondrial pellet by energized mitochondria (data not shown). Consequently,
(Figure 4C). The far greater uptake of MitoSOD over M40403 is the uptake of MitoSOD as seen in Figure 4 could be explained
consistent with its uptake into mitochondria in response to the by prior dissociation of Mn from MitoSOD, followed by uptake
membrane potential. Overall, by considering the interaction of apo-MitoSOD. To determine whether MitoSOD was accu-
with unenergized mitochondria (Figure 4B), with uncoupled mito- mulated intact, we measured, by inductively coupled plasma
chondria (Figure 4C), and the value of its partition coefficient, it mass spectrometry, the extent of Mn uptake into mitochondria
is clear that MitoSOD is extensively bound to unenergized during incubations with MitoSOD. Incubation with
5 mM
mitochondria by adsorption to membranes. To better distinguish MitoSOD led to the rapid uptake of ꢀ2.5–3.5 nmol Mn (Figure 5A),
between simple binding and membrane potential-driven with similar kinetics and extent of uptake as seen for MitoSOD
accumulation into mitochondria, we expressed the uptake of itself (Figure 4). To see if it was possible for apo-MitoSOD to
[3H]MitoSOD into the pellet as an accumulation ratio (ACR: mol be taken up by mitochondria and then incorporate Mn to form
lꢁ1 in mitochondria/mol lꢁ1 in supernatant). This ACR of [3H] MitoSOD in situ, we incubated apo-MitoSOD with MnCl2 and
MitoSOD was compared with that of [3H]TPMP in parallel exper- found no observable Mn incorporation to give MitoSOD, as
iments over a range of mitochondrial membrane potentials set assessed by mass spectrometry (Figure S2B). Incubation of
by different concentrations of the respiratory inhibitor malonate apo-MitoSOD with mitochondria did not lead to MitoSOD
(Figure 4D) (Ross et al., 2008). As expected from Figures 4B assembly (data not shown), and incubation of apo-MitoSOD
and 4C, it is evident that with uncoupled mitochondria, there is and MnCl2 with mitochondria did not lead to uptake of Mn into
extensive binding of MitoSOD, and as the membrane potential mitochondria (Figure 5A). Therefore, we conclude that MitoSOD
increases there is an increase in ACR. The maximal ACR for is taken up into mitochondria as an intact complex and the
MitoSOD was found to be ꢀ1,300, whereas the ACR for TPMP uptake of apo-MitoSOD alone does not lead to the assembly
was ꢀ550 under comparable conditions. Furthermore, when of MitoSOD within the mitochondrial matrix.
mitochondria were fully uncoupled with FCCP, the MitoSOD
ACR was still significant at ꢀ500. Together, these findings are MitoSOD Protects Mitochondrial Aconitase
consistent with extensive nonspecific binding of the hydro- from Superoxide Damage
phobic MitoSOD to mitochondria. The slope of the plot of the To assess the ability of MitoSOD to detoxify O2 within mito-
,-
,-
ACR of MitoSOD against that of TPMP (Figure 4D) is greater chondria, we used the redox cycler paraquat to induce O2
than one, indicating that within mitochondria there is more ex- formation in isolated heart mitochondria and monitored the
tensive binding of MitoSOD to the matrix face of the inner O2,--dependent inactivation of the intramitochondrial enzyme
1242 Chemistry & Biology 19, 1237–1246, October 26, 2012 ª2012 Elsevier Ltd All rights reserved

Chemistry & Biology
MitoSOD
,-
´
aconitase, which is particularly sensitive to O2 (Cocheme and
Murphy, 2008). Paraquat decreased the activity of mitochondrial
aconitase to 18% of control, and addition of MitoSOD partially
protected aconitase from this damage (Figures 5B and 5C). In
contrast, M40403 was ineffective, presumably because of its
lack of uptake by mitochondria (Figure 5C). Although the extent
of protection by MitoSOD is modest, this is expected because
A
4
3
2
no additions
MitoSOD
MnCl2
,-
the rate constant of MitoSOD-catalyzed O2 dismutation (5 3
106 Mꢁ1 sꢁ1 at pH 8.0, 22ꢂC) is ꢀ400-fold lower than kcat for
the mitochondrial enzyme MnSOD (2 3 109 Mꢁ1 sꢁ1 at 25ꢂC;
Bull et al., 1991). However, the large amount of MitoSOD taken
up by mitochondria in response to the membrane potential
ApoMitoSOD + MnCl2
1
0
0
2
4
6
8
10
,-
(ꢀ1.5 mM) significantly increases the O2 defense capacity,
Time (min)
although as the concentration of MnSOD in the mitochondrial
matrix is ꢀ10–20 mM, the accumulated MitoSOD provides at
most 20%–30% of the SOD activity in the mitochondrial matrix
under these conditions. Higher concentrations of MitoSOD
provided no further protection because at higher concentrations
TPP cations start to affect the membrane potential, thereby
limiting further accumulation.
B
no additions
100
80
paraquat + MitoSOD
paraquat
60
40
20
0
SIGNIFICANCE
0
2
4
6
8
Antioxidants targeted to mitochondria by conjugation to
lipophilic cations have proven useful as tools for investi-
gating oxidative damage and redox signaling, and as poten-
tial therapies for diseases involving mitochondrial oxidative
Time (min)
C
100
80
60
40
20
0
damage (Murphy and Smith, 2007, 2010). Although some
,-
P < 0.05
mitochondria-targeted antioxidants can react with O2
,
they are not selective. Consequently, it is not possible to
infer if their action is due to depleting O2,- or to some other
antioxidant reaction. We have resolved this issue by invent-
ing a mitochondria-targeted antioxidant that is selective for
O2,-. This was achieved by conjugating the core function-
ality of the Mn(II) macrocycle SOD mimetic M40403 to the
mitochondria-targeting lipophilic cation TPP. The resultant
compound, MitoSOD, was a very effective catalytic SOD
+ paraquat
,-
mimetic and also had catalytic O2 reductase activity in
the presence of the biological reductant ascorbate. The
Figure 5. MitoSOD Increases Mitochondrial Mn Uptake and
Decreases Aconitase Inactivation
,-
rate constant of MitoSOD-catalyzed O2 dismutation (5 3
106 M–1 s–1 at pH 8.0, 22ꢂC) is lower than kcat for the mito-
chondrial enzyme MnSOD (2 3 109 M–1 s–1 at 25ꢂC; Bull
et al., 1991). However, the large amount of MitoSOD taken
up by mitochondria in response to the membrane potential
Rat liver mitochondria (1 mg protein mlꢁ1) were incubated in 0.5 ml incubation
medium supplemented with 10 mM succinate, 8 mg mlꢁ1 rotenone at 37ꢂC for
the indicated times with 5 mM MitoSOD, no additions, 5 mM MnCl2, or 1 mM
apo-MitoSOD + 5 mM MnCl2. Then the mitochondria were pelleted, and the
concentration of Mn in the pellets was determined by inductively coupled
plasma (ICP) analysis. Data are means SD of quadruplicate determinations.
(B) Rat heart mitochondria (2 mg protein mlꢁ1) were incubated in incubation
medium (600–800 ml) containing glutamate/malate (10 mM) and bovine serum
albumin (5% w/v), with or without paraquat (500 mM) or MitoSOD (5 mM) or
M40403 (5 mM), at 37ꢂC for the indicated times. Samples (100 ml), taken
immediately before paraquat addition and at the indicated times were
snap-frozen on dry ice and assessed for aconitase activity. Aconitase activity
is expressed as a percentage of the activity immediately before paraquat
addition. Data correspond to the means SD of three independent experi-
ments.
,-
significantly increases the O2 defense capacity within
mitochondria above that provided by the endogenous
MnSOD enzyme alone, indicating that this is a viable antiox-
idant strategy and that developing analogs of MitoSOD
,-
with greater rates of O2 dismutation is the best path to
development of future SOD mimetics. To our knowledge,
MitoSOD constitutes the first example of a selective macro-
cyclic Mn(II) SOD mimetic that is taken up into mitochondria,
driven by the membrane potential, and exhibits a protective
,-
effect against O2 within mitochondria. This work estab-
(C) Mitochondria were incubated and processed as in (B) for 6 min with or
without paraquat (500 mM), MitoSOD (5 mM), or M40403 (5 mM). Statistical
significance was calculated by a two-tailed, paired Student’s t test.
See also Figure S2.
lishes the baseline parameters for the rational development
of bioactive mitochondria-targeted SOD mimetics, as well
as other mitochondria-targeted macrocycle and metal-
dependent moieties, with potential for pharmaceutical
application.
Chemistry & Biology 19, 1237–1246, October 26, 2012 ª2012 Elsevier Ltd All rights reserved 1243

Chemistry & Biology
MitoSOD
O2 + H$
/
HO$2
k = 1:2 3 1010 Mꢁ1 sꢁ1
H+ + O$2ꢁ
k = 3:2 3 109 Mꢁ1 sꢁ1
k = 2:1 3 108 Mꢁ1 sꢁ1
k = 4:2 3 109 Mꢁ1 sꢁ1
;
(3)
(4)
(5)
(6)
(7)
EXPERIMENTAL PROCEDURES
Synthetic Chemistry
HO2$
!
;
The target compounds MitoSOD (1) and M40403 (2) were obtained by Mn(II)
template cyclization of the tetraamine 3 and 4 or 2,6-pyridinedialdehyde,
respectively, followed by in situ borohydride reduction of the intermediate
Schiff-base macrocycle (Figure 1A). General synthetic methods and details
of syntheses are given in the Supplemental Experimental Procedures.
HCOꢁ2 + HO$
HCOꢁ2 + H$
/
/
/
CO$2ꢁ + H2O
CO$2ꢁ + H2
O$2ꢁ + CO2
;
;
O2 + CO$ꢁ
:
X-Ray Crystallography
2
X-ray data were collected by Assoc. Prof. C.E.F. Rickard, The University of
The decay in the optical absorbance due to O2,-, for various radiation doses,
was monitored at 270 nm, and the first half-lives were obtained directly from
the absorbance traces.
Auckland, New Zealand, on a Bruker SMART diffractometer using graphite-
˚
monochromated Mo-Ka radiation (l = 0.71073 A). MitoSOD crystallized in
the chiral monoclinic space group P21 with one molecule of coordinated
water and two half occupancy noncoordinated molecules of ethanol per
MnLCl3, that is, as [MnLCl(OH2)]Cl2$EtOH. The data were corrected for
Lorentz and polarization effects, and a semiempirical absorption correction
(SADABS) was applied. The structure was solved by direct methods
(SHELXS-97) (Sheldrick, 1990, 1997) and refined using SHELXL-97 (Sheldrick
and Schneider, 1997). All nonhydrogen atoms were made anisotropic.
Hydrogen atoms were inserted at calculated positions and rode on the atoms
to which they are attached, except for those on N2, N3, N4, N5, and O1,
which were located from difference maps and fixed (AFIX 01). The two half
occupancy EtOH solvent molecules are both well behaved. One of the two
noncoordinating chloride ions is modeled as disordered over two sites as
0.85:0.15 Cl2:Cl20 (an attempt to model this as a Cl:Br disorder resulted in
0.9:0.1 occupancy; this option, of the presence of a trace of Br, cannot be
excluded although it was not employed here). Mn(II) is coordinated by the
five N atoms of the macrocycle in a near-planar arrangement, as evident
from the sum of the five N-Mn-N chelate angles (363ꢂ). A chloride ion and
water molecule are also coordinated to Mn(II), with a ClꢁMn-O angle of
176.8ꢂ; the geometry of the coordination sphere is best described as pentag-
onal bipyramidal. The alternating axial disposition of the cyclohexyl methine
hydrogens confirmed the stereochemistry of the four chiral centers of the
macrocycle. Despite the existence of M40403 being known for at least a
decade, its X-ray structure remains unreported. An attempt by us to prepare
crystals of M40403 resulted in crystal twinning, and an X-ray structure was
unable to be obtained. Hence, the MitoSOD structure was compared to the
X-ray structure of a pentagonal bipyramidal polymorph of M40401, an active
1,1’-(pyridine-2,6-diyl)diethanamine analog of M40403 (Aston et al., 2001).
The MitoSOD and M40401 structures showed similar Mn-N bond lengths
and N-Mn-N chelate angles.
Determination of SOD Activity
For Figures 3A and 3B, a mixture of acetylated ferricytochrome c (25 mM) in
potassium phosphate (50 mM) and EDTA (100 mM) at pH 7.8 was supple-
mented with hypoxanthine (50 mM), held at 25ꢂC in a 1 ml cuvette, and the
absorbance at 550 nm was monitored (Loschen et al., 1974). After 30 s
,-
incubation, O2 production was initiated by addition of xanthine oxidase
(XO, 0.2 U mlꢁ1) to give an O2,- generation rate of ꢀ6 mM minꢁ1, and then after
a further 90 s the test compounds were added from a DMSO stock solution.
The rate of cytochrome c reduction at 550 nm was recorded for a further
3 min. Monitoring the formation of uric acid at 295 nm confirmed that MitoSOD
up to at least 10 mM did not inhibit XO activity. Demetallation of MitoSOD by
treatment with acid gave a metal-free residue that had no SOD activity. The
comparison of the specific SOD activity for Cu,ZnSOD, MitoSOD, and
M40403 was assessed by the WST-1 assay as described (Peskin and Winter-
bourn, 2000). For Figure 3C, ascorbate oxidation and cytochrome c reduction
were measured at 22ꢂC in 50 mM phosphate buffer (pH 7.4) containing 50 mM
DTPA and 5 mg mlꢁ1 catalase. A flux of O2,- was generated by addition of XO
(ꢀ0.1 U mlꢁ1 to give a O2,- generation rate of 2 mM minꢁ1) and acetaldehyde
(10 mM). Cu,ZnSOD (60 nM), or various concentrations of M40403 and
MitoSOD, was added to this. Then either cytochrome c (20 mM) was added
and its rate of reduction was monitored at 550 nm, or ascorbate (30 mM)
was added and its rate of oxidation was monitored at 266 nm. Data are
means ranges for 2–3 determinations. The oxidation rate of ascorbate alone,
which was ꢀ2% of the O2,--dependent reaction, was subtracted.
Uptake of MitoSOD by Mitochondria
An ion-selective electrode was prepared that responds to the triphenylphos-
phonium (TPP) moiety (Asin-Cayuela et al., 2004) and inserted into a stirred
solution of incubation medium (120 mM KCl, 10 mM HEPES [pH 7.2], and
1 mM EGTA). The uptake of [3H]MitoSOD and [3H]M40403 by isolated rat liver
mitochondria (1 mg protein mlꢁ1) was determined by suspension in Incubation
Medium supplemented with 10 mM succinate, 8 mg mlꢁ1 rotenone, 4 mM
MitoSOD or M40403, and 1 mM [3H]MitoSOD (28 Ci molꢁ1) or [3H]-M40403
(13 Ci molꢁ1) at 25ꢂC. The uptake of [3H]MitoSOD and [3H]M40403 was deter-
mined at various times by isolating the mitochondria by centrifugation and
then determining the amount of [3H]macrocycle in the mitochondrial pellet by
scintillation counting and using the specific activity of the starting solution to
calculate the amount associated with the mitochondria (Kelso et al., 2001;
Smith et al., 1999, 2004). The data are means of three separate experiments
SEM. For the FCCP experiment, 500 nM FCCP was present. To determine the
effect of the mitochondrial membrane potential on the uptake of [3H]-MitoSOD
rat liver mitochondria (1 mg protein mlꢁ1) were incubated in incubation medium
containing 10 mM succinate, 8 mg mlꢁ1 rotenone, 100 mg mlꢁ1 nigericin, 800 nM
MitoSOD, 1 mM TPMP, and 1–16 mM malonate, or to measure the uptake of
3H-labeled compounds in the complete absence of a membrane potential,
FCCP (330 nM) was added to the incubation medium. These were supple-
mented with either 200 nM [3H]MitoSOD (28 Ci molꢁ1) or 100 nM [3H]TPMP
(500 Ci molꢁ1) for 3 min at 25ꢂC. The ACR of [3H]MitoSOD or [3H]TPMP was
determined at various times in parallel to enable the ACR for the two com-
pounds to be determined under identical conditions by isolating the mito-
chondria by centrifugation and then determining the amount of [3H]MitoSOD
or [3H]TPMP in the mitochondrial pellet and in the supernatant by scintillation
Pulse Radiolysis
The
4 MeV linear accelerator (LINAC) of The University of Auckland,
New Zealand, was used to irradiate sample solutions contained in a quartz
cuvette. The LINAC has been adapted to produce variable amounts of
radiation (0–100 Gy in 1.5 ms) to the sample. Dosimetry was carried out
,-
using aerated KSCN solution assuming the (SCN)2 radical produced has
a radiation chemical yield (G) of 0.29 mM Gyꢁ1 and an extinction coefficient
of 7,580 Mꢁ1 cmꢁ1 at 475 nm (Schuler et al., 1980). Changes in absorbance
were measured using a PC-controlled custom-built optical detection system
(Anderson et al., 1997). All experiments were performed at 22ꢂC 1ꢂC.
The radiolysis of water with high-energy electrons produces both primary
radical species and products as shown in Equation (1). The G-values of the
different species (in mM Gyꢁ1) are shown in parentheses.
H2O,eꢁaqð0:28Þ+ $OHð0:28Þ+ $Hð0:055Þ+ H2ð0:04Þ+ H2O2ð0:07Þ+ H3O+ ð0:28Þ
(1)
Quantitative amounts of O2,- were produced in aerated solutions containing
sodium formate (10 mM), phosphate buffer (10 mM [pH 7.4 or 8.0]), and EDTA
(0.1 mM). Under these conditions, all the primary radical species are quickly
converted into O2,-, as described by Equations (2)–(7), resulting in a yield
,-
of O2 of 0.62 mM Gyꢁ1 by the end of the pulse (Bielski et al., 1985; Ilan and
Rabani, 1976).
O2 + eꢁaq
/
O$2ꢁ
k = 1:9 3 1010 Mꢁ1 sꢁ1
;
(2)
counting and taking the mitochondrial volume as 0.5 ml mg proteinꢁ1
.
1244 Chemistry & Biology 19, 1237–1246, October 26, 2012 ª2012 Elsevier Ltd All rights reserved

Chemistry & Biology
MitoSOD
Determination of Aconitase Activity
Cocheme´ , H.M., and Murphy, M.P. (2008). Complex I is the major site of mito-
Aliquots (10 ml) of the snap-frozen mitochondrial incubtions samples were
incubated in assay buffer (190 ml) comprising Tris-HCl (50 mM [pH 7.4]),
MnCl2 (0.6 mM), sodium citrate (5 mM [pH 7.0]), NADP+ (0.2 mM), Triton
X-100 (0.1% v/v), and isocitrate dehydrogenase (0.4 U mlꢁ1) in quintuplicate
at 30ꢂC in a 96-well plate, and the appearance of NADPH (ε340 = 6.22 3 103
chondrial superoxide production by paraquat. J. Biol. Chem. 283, 1786–1798.
´
Collins, Y., Chouchani, E.T., James, A.M., Menger, K.E., Cocheme, H.M., and
Murphy, M.P. (2012). Mitochondrial redox signalling at a glance. J. Cell Sci.
125, 801–806.
Dessolin, J., Schuler, M., Quinart, A., De Giorgi, F., Ghosez, L., and Ichas, F.
(2002). Selective targeting of synthetic antioxidants to mitochondria: towards
a mitochondrial medicine for neurodegenerative diseases? Eur. J. Pharmacol.
447, 155–161.
Mꢁ1 cmꢁ1) was monitored over 10 min at 340 nm (Cocheme and Murphy,
´
2008).
SUPPLEMENTAL INFORMATION
Dhanasekaran, A., Kotamraju, S., Karunakaran, C., Kalivendi, S.V., Thomas,
S., Joseph, J., and Kalyanaraman, B. (2005). Mitochondria superoxide dismu-
tase mimetic inhibits peroxide-induced oxidative damage and apoptosis: role
of mitochondrial superoxide. Free Radic. Biol. Med. 39, 567–583.
Supplemental Information includes two figures, three tables, and Supple-
mental Experimental Procedures and can be found with this article online at
http://dx.doi.org/10.1016/j.chembiol.2012.08.005.
Dikalova, A.E., Bikineyeva, A.T., Budzyn, K., Nazarewicz, R.R., McCann, L.,
Lewis, W., Harrison, D.G., and Dikalov, S.I. (2010). Therapeutic targeting of
mitochondrial superoxide in hypertension. Circ. Res. 107, 106–116.
ACKNOWLEDGMENTS
Doctrow, S.R., Huffman, K., Marcus, C.B., Musleh, W., Bruce, A., Baudry, M.,
and Malfroy, B. (1997). Salen-manganese complexes: combined superoxide
dismutase/catalase mimics with broad pharmacological efficacy. Adv.
Pharmacol. 38, 247–269.
This work was supported by Grant UOA 0407 of the Marsden Fund from the
Royal Society of New Zealand and by the Medical Research Council (UK).
Received: May 30, 2012
Revised: July 27, 2012
Faulkner, K.M., Liochev, S.I., and Fridovich, I. (1994). Stable Mn(III) porphyrins
mimic superoxide dismutase in vitro and substitute for it in vivo. J. Biol. Chem.
269, 23471–23476.
Accepted: August 4, 2012
Published: October 25, 2012
Feltham, R.D., and Hayter, R.G. (1964). The electrolyte type of ionized
complexes. J. Chem. Soc., 0, 4587–4591.
REFERENCES
ꢀ
ꢀ
Filipovic, M.R., Duerr, K., Mojovic, M., Simeunovic, V., Zimmermann, R.,
ꢀ
ꢀ
ꢀ
ꢀ
Niketic, V., and Ivanovic-Burmazovic, I. (2008). NO dismutase activity of
seven-coordinate manganese(II) pentaazamacrocyclic complexes. Angew.
Chem. Int. Ed. Engl. 47, 8735–8739.
Anderson, R.F., Denny, W.A., Li, W., Packer, J.E., Tercel, M., and Wilson, W.R.
(1997). Pulse radiolysis studies on the fragmentation of arylmethyl quaternary
nitrogen mustards by one-electron reduction in aqueous solution. J. Phys.
Chem. A 101, 9704–9709.
ꢀ
Filipovic, M.R., Koh, A.C., Arbault, S., Niketic, V., Debus, A., Schleicher, U.,
ꢀ
ꢀ
ꢀ
Bogdan, C., Guille, M., Lemaıˆtre, F., Amatore, C., and Ivanovic-Burmazovic,
I. (2010). Striking inflammation from both sides: manganese(II) pentaazama-
crocyclic SOD mimics act also as nitric oxide dismutases: a single-cell study.
Angew. Chem. Int. Ed. Engl. 49, 4228–4232.
Anderson, R.F., Shinde, S.S., Hay, M.P., Gamage, S.A., and Denny, W.A.
(2003). Activation of 3-amino-1,2,4-benzotriazine 1,4-dioxide antitumor
agents to oxidizing species following their one-electron reduction. J. Am.
Chem. Soc. 125, 748–756.
Filipovska, A., Kelso, G.F., Brown, S.E., Beer, S.M., Smith, R.A., and Murphy,
Asin-Cayuela, J., Manas, A.R., James, A.M., Smith, R.A., and Murphy, M.P.
(2004). Fine-tuning the hydrophobicity of a mitochondria-targeted antioxidant.
FEBS Lett. 571, 9–16.
M.P. (2005). Synthesis and characterization of
a triphenylphosphonium-
conjugated peroxidase mimetic. Insights into the interaction of ebselen with
mitochondria. J. Biol. Chem. 280, 24113–24126.
Aston, K., Rath, N., Naik, A., Slomczynska, U., Schall, O.F., and Riley, D.P.
(2001). Computer-aided design (CAD) of Mn(II) complexes: superoxide
dismutase mimetics with catalytic activity exceeding the native enzyme.
Inorg. Chem. 40, 1779–1789.
Finkel, T. (2005). Radical medicine: treating ageing to cure disease. Nat. Rev.
Mol. Cell Biol. 6, 971–976.
ꢀ
ꢀ
Friedel, F.C., Lieb, D., and Ivanovic-Burmazovic, I. (2012). Comparative
studies on manganese-based SOD mimetics, including the phosphate effect,
by using global spectral analysis. J. Inorg. Biochem. 109, 26–32.
ꢀ
Batinic-Haberle, I. (2002). Manganese porphyrins and related compounds as
mimics of superoxide dismutase. Methods Enzymol. 349, 223–233.
Gane, E.J., Weilert, F., Orr, D.W., Keogh, G.F., Gibson, M., Lockhart, M.M.,
Frampton, C.M., Taylor, K.M., Smith, R.A., and Murphy, M.P. (2010). The mito-
chondria-targeted anti-oxidant mitoquinone decreases liver damage in
a phase II study of hepatitis C patients. Liver Int. 30, 1019–1026.
ꢀ
ꢀ
Batinic-Haberle, I., Rebouc¸ as, J.S., and Spasojevic, I. (2010). Superoxide
dismutase mimics: chemistry, pharmacology, and therapeutic potential.
Antioxid. Redox Signal. 13, 877–918.
Bielski, B.H.J., Cabelli, D.E., and Arudi, R. (1985). Reactivity of HO2,/O2
,-
Gotoh, N., and Niki, E. (1992). Rates of interactions of superoxide with vitamin
E, vitamin C and related compounds as measured by chemiluminescence.
Biochim. Biophys. Acta 1115, 201–207.
radicals in aqueous solution. J. Phys. Chem. Ref. Data 14, 1041–1100.
Brown, G.C., and Brand, M.D. (1985). Thermodynamic control of electron flux
through mitochondrial cytochrome bc1 complex. Biochem. J. 225, 399–405.
Ilan, Y., and Rabani, J. (1976). Some fundamental reactions in radiation-
chemistry: nanosecond pulse radiolysis. Int. J. Radiat. Phys. Chem. 8,
609–611.
Brown, S.E., Ross, M.F., Sanjuan-Pla, A., Manas, A.R., Smith, R.A., and
Murphy, M.P. (2007). Targeting lipoic acid to mitochondria: synthesis and
characterization of a triphenylphosphonium-conjugated alpha-lipoyl deriva-
tive. Free Radic. Biol. Med. 42, 1766–1780.
Iranzo, O. (2011). Manganese complexes displaying superoxide dismutase
activity: a balance between different factors. Bioorg. Chem. 39, 73–87.
Bull, C., Niederhoffer, E.C., Yoshida, T., and Fee, J.A. (1991). Kinetic studies of
superoxide dismutases: properties of the manganese-containing protein
from Thermus thermophilus. J. Am. Chem. Soc. 113, 4069–4076.
Kagan, V.E., Jiang, J., Bayir, H., and Stoyanovsky, D.A. (2007). Targeting
nitroxides to mitochondria: location, location, location, and ...concentration:
highlight commentary on ‘‘Mitochondria superoxide dismutase mimetic
inhibits peroxide-induced oxidative damage and apoptosis: role of mitochon-
drial superoxide’’. Free Radic. Biol. Med. 43, 348–350.
Cadenas, E., and Davies, K.J. (2000). Mitochondrial free radical generation,
oxidative stress, and aging. Free Radic. Biol. Med. 29, 222–230.
Chen, Z., Siu, B., Ho, Y.S., Vincent, R., Chua, C.C., Hamdy, R.C., and Chua,
B.H. (1998). Overexpression of MnSOD protects against myocardial
ischemia/reperfusion injury in transgenic mice. J. Mol. Cell. Cardiol. 30,
2281–2289.
Kelso, G.F., Porteous, C.M., Coulter, C.V., Hughes, G., Porteous, W.K.,
Ledgerwood, E.C., Smith, R.A.J., and Murphy, M.P. (2001). Selective targeting
of a redox-active ubiquinone to mitochondria within cells: antioxidant and
antiapoptotic properties. J. Biol. Chem. 276, 4588–4596.
Chemistry & Biology 19, 1237–1246, October 26, 2012 ª2012 Elsevier Ltd All rights reserved 1245

Chemistry & Biology
MitoSOD
Kim, A., Zhong, W., and Oberley, T.D. (2004). Reversible modulation of cell
cycle kinetics in NIH/3T3 mouse fibroblasts by inducible overexpression of
mitochondrial manganese superoxide dismutase. Antioxid. Redox Signal. 6,
489–500.
Salvemini, D., Riley, D.P., and Cuzzocrea, S. (2002). SOD mimetics are coming
of age. Nat. Rev. Drug Discov. 1, 367–374.
Schuler, R.H., Patterson, L.K., and Janata, E. (1980). Yield for the scavenging
of hydroxyl radicals in the radiolysis of N2O-saturated aqueous solutions.
J. Phys. Chem. 84, 2088–2089.
Li, X., Cobb, C.E., Hill, K.E., Burk, R.F., and May, J.M. (2001). Mitochondrial
uptake and recycling of ascorbic acid. Arch. Biochem. Biophys. 387, 143–153.
Sheldrick, G.M. (1990). Phase annealing in SHELX-90: direct methods for
´
Loschen, G., Azzi, A., Richter, C., and Flohe, L. (1974). Superoxide radicals as
larger structures. Acta Cryst. A46, 467–473.
precursors of mitochondrial hydrogen peroxide. FEBS Lett. 42, 68–72.
Sheldrick, G.M. (1997). Patterson superposition and ab initio phasing. In
Macromolecular Crystallography, Pt A, C. Carter, Jr. and R. Sweet, eds.
(San Diego: Academic Press), pp. 628–641.
Maroz, A., Kelso, G.F., Smith, R.A., Ware, D.C., and Anderson, R.F. (2008).
Pulse radiolysis investigation on the mechanism of the catalytic action of
Mn(II)-pentaazamacrocycle compounds as superoxide dismutase mimetics.
J. Phys. Chem. A 112, 4929–4935.
Sheldrick, G.M., and Schneider, T.R. (1997). SHELXL: high-resolution refine-
ment. In Macromolecular Crystallography, Pt B, C. Carter, Jr. and R. Sweet,
eds. (San Diego: Academic Press), pp. 319–343.
Murphy, M.P. (2009). How mitochondria produce reactive oxygen species.
Biochem. J. 417, 1–13.
Skulachev, V.P., Anisimov, V.N., Antonenko, Y.N., Bakeeva, L.E., Chernyak,
B.V., Erichev, V.P., Filenko, O.F., Kalinina, N.I., Kapelko, V.I., Kolosova,
N.G., et al. (2009). An attempt to prevent senescence: a mitochondrial
approach. Biochim. Biophys. Acta 1787, 437–461.
Murphy, M.P., and Smith, R.A. (2007). Targeting antioxidants to mitochondria
by conjugation to lipophilic cations. Annu. Rev. Pharmacol. Toxicol. 47,
629–656.
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., et al.
(2003). Superoxide activates uncoupling proteins by generating carbon-
centered radicals and initiating lipid peroxidation: studies using a mitochon-
dria-targeted spin trap derived from alpha-phenyl-N-tert-butylnitrone.
J. Biol. Chem. 278, 48534–48545.
Smith, R.A., and Murphy, M.P. (2010). Animal and human studies with the
mitochondria-targeted antioxidant MitoQ. Ann. N Y Acad. Sci. 1201, 96–103.
Smith, R.A., Kelso, G.F., James, A.M., and Murphy, M.P. (2004). Targeting
coenzyme Q derivatives to mitochondria. Methods Enzymol. 382, 45–67.
Smith, R.A., Hartley, R.C., and Murphy, M.P. (2011). Mitochondria-targeted
small molecule therapeutics and probes. Antioxid. Redox Signal. 15, 3021–
3038.
Peskin, A.V., and Winterbourn, C.C. (2000). A microtiter plate assay for super-
oxide dismutase using a water-soluble tetrazolium salt (WST-1). Clin. Chim.
Acta 293, 157–166.
Smith, R.A.J., Porteous, C.M., Coulter, C.V., and Murphy, M.P. (1999).
Selective targeting of an antioxidant to mitochondria. Eur. J. Biochem. 263,
709–716.
´
Ross, M.F., Kelso, G.F., Blaikie, F.H., James, A.M., Cocheme, H.M.,
Filipovska, A., Da Ros, T., Hurd, T.R., Smith, R.A., and Murphy, M.P. (2005).
Lipophilic triphenylphosphonium cations as tools in mitochondrial bioener-
getics and free radical biology. Biochemistry (Mosc.) 70, 222–230.
Snow, B.J., Rolfe, F.L., Lockhart, M.M., Frampton, C.M., O’Sullivan, J.D., Fung,
V., Smith, R.A., Murphy, M.P., and Taylor, K.M.; Protect Study Group. (2010). A
double-blind, placebo-controlled study to assess the mitochondria-targeted
antioxidant MitoQ as a disease-modifying therapy in Parkinson’s disease.
Mov. Disord. 25, 1670–1674.
Ross, M.F., Da Ros, T., Blaikie, F.H., Prime, T.A., Porteous, C.M., Severina, I.I.,
Skulachev, V.P., Kjaergaard, H.G., Smith, R.A., and Murphy, M.P. (2006).
Accumulation of lipophilic dications by mitochondria and cells. Biochem. J.
400, 199–208.
Soule, B.P., Hyodo, F., Matsumoto, K., Simone, N.L., Cook, J.A., Krishna,
M.C., and Mitchell, J.B. (2007). The chemistry and biology of nitroxide
compounds. Free Radic. Biol. Med. 42, 1632–1650.
Ross, M.F., Prime, T.A., Abakumova, I., James, A.M., Porteous, C.M., Smith,
R.A.J., and Murphy, M.P. (2008). Rapid and extensive uptake and activation
of hydrophobic triphenylphosphonium cations within cells. Biochem. J. 411,
633–645.
Thompson, J.S., Chu, Y., Glass, J., Tapp, A.A., and Brown, S.A. (2010). The
manganese superoxide dismutase mimetic, M40403, protects adult mice
from lethal total body irradiation. Free Radic. Res. 44, 529–540.
Salvemini, D., Wang, Z.-Q., Zweier, J.L., Samouilov, A., Macarthur, H., Misko,
T.P., Currie, M.G., Cuzzocrea, S., Sikorski, J.A., and Riley, D.P. (1999).
A nonpeptidyl mimic of superoxide dismutase with therapeutic activity in
rats. Science 286, 304–306.
Trnka, J., Blaikie, F.H., Smith, R.A.J., and Murphy, M.P. (2008). A mitochon-
dria-targeted nitroxide is reduced to its hydroxylamine by ubiquinol in mito-
chondria. Free Radic. Biol. Med. 44, 1406–1419.
1246 Chemistry & Biology 19, 1237–1246, October 26, 2012 ª2012 Elsevier Ltd All rights reserved