Amphiphilic/bipolar metallocorroles that catalyze the decomposition of reactive oxygen and nitrogen species, rescue lipoproteins from oxidative damage,and attenuate atherosclerosis in mice

Article abstract of DOI:10.1002/anie.200801149

(Chemical Equation Presented) Antioxidans that work! The iron corrole 1-Fe (see picture) is a potetnt catalyst for decomposition of reactive oxygen and nitrogen species that binds selectively to lipoproteins. The complex also affects cholesterol levels and its cellular efflux. 1-Fe is more effective that natural antioxidants in reducing atherosclerosis development in mice. LDL = low-density lipoprotein.

Full text of DOI:10.1002/anie.200801149

Communications
DOI: 10.1002/anie.200801149
Catalytic Antioxidants
Amphiphilic/Bipolar Metallocorroles That Catalyze the Decomposition
of Reactive Oxygen and Nitrogen Species,Rescue Lipoproteins from
Oxidative Damage,and Attenuate Atherosclerosis in Mice**
Adi Haber, Atif Mahammed, Bianca Fuhrman, Nina Volkova, Raymond Coleman, Tony Hayek,
Michael Aviram,* and Zeev Gross*
Accumulating evidence points towards the involvement of
reactive oxygen and nitrogen species (ROS and RNS,
respectively) in a very wide variety of diseases.[1] The
relationship between atherosclerosis and oxidative/nitrosa-
tive stress—the imbalance between the desired production of
ROS/RNS and the capability of biological systems to detoxify
excess amounts thereof—is often analyzed in terms of
damage to the cholesterol-transporting high- and low-density
lipoproteins (HDL and LDL, respectively).[2] Oxidized LDL
are taken up at an enhanced rate by arterial macrophages in a
noncontrolled fashion and modified HDL are less effective in
performing cellular cholesterol efflux, thus leading to the
formation of cholesterol-loaded foam cells, the hallmark of
early atherosclerosis. Although the athero-
against it, and most natural antioxidants are very poor
peroxynitrite scavengers.[7] This calls for the development of
synthetic reagents that could act on and neutralize peroxyni-
trite by one or more of the following ways: 1) interfere with its
formation by eliminating its precursors (superoxide anion and
nitric oxide); 2) decompose it to biologically benign prod-
ucts;[8] and 3) repair the damage caused by it.
The motivation for the current investigations was the fact
that the particular iron(III) and manganese(III) corroles
shown in Scheme 1A were disclosed as potent catalysts for
decomposition of peroxynitrite in purely chemical systems,
with the former acting very fast and the latter operating
through the novel mechanism shown in Scheme 1B.[9] The
sclerosis-preventing effect of certain natural
antioxidants is in harmony with this hypoth-
esis,[3] it is seriously challenged by the fact
that many potent in vitro antioxidants fail to
display beneficial in vivo effects.[4]
Recent
focus
on
peroxynitrite
(HOONO) revealed that this molecule is
involved (directly and through COH and
CNO2 derived from it) in the damage of a
wide variety of molecules,[5] including those
that are of vital importance for proper
cardiovascular function.[6] The problem
unique to peroxynitrite is that, in contrast
Scheme 1. A) Structures of the corroles used. B) The disproportionation mechanism for
to other ROS and RNS and their precursors, catalytic decomposition of hydrogen peroxide and peroxynitrite by 1-Mn and the balanced
equations for catalytic decomposition of peroxynitrite by 1-Mn and 1-Fe.
there is no known biological defense system
[*] Dr. B. Fuhrman, N. Volkova, Prof. Dr. T. Hayek, Prof. Dr. M. Aviram
same metallocorroles (of unique amphiphilicity and bipolar-
Lipid Research Laboratory, Rappaport Faculty of Medicine
ity resulting from the positioning of sulfonic acid head groups
on the otherwise lipophilic corrole) were also shown to have a
large affinity to various proteins,[10] a very important factor
that may be used for selective delivery purposes.
Technion—Israel Institute of Technology
Rambam Medical Center, Haifa 31096 (Israel)
E-mail: aviram@tx.technion.ac.il
A. Haber, Dr. A. Mahammed, Prof. Dr. Z. Gross
Schulich Faculty of Chemistry
Technion—Israel Institute of Technology
Technion City, Haifa 32000 (Israel)
E-mail: ch10zg@tx.technion.ac.il
Herein, we show that the iron corrole 1-Fe (Scheme 1A)
rescues small molecules from ROS-induced oxidation more
efficiently than the manganese complex 1-Mn, that the
opposite holds for arresting RNS-induced nitration, and that
1-Fe and 1-Mn are anti- and pro-oxidants, respectively, with
regard to their in vitro effects on oxidative damage to LDL/
HDL. An important finding with regard to targeting is that
both metallocorroles, as well as the non-transition-metal
complex 1-Ga, bind to lipoproteins more strongly than to all
other serum components. All three corroles also affect
cholesterol levels and distribution in plasma, and 1-Fe, but
Prof. Dr. R. Coleman
Department of Anatomy and Cell Biology, Rappaport Faculty of
Medicine, Technion—Israel Institute of Technology
Rambam Medical Center, Haifa 31096 (Israel)
[**] This research was supported by a generous grant from the Gurwin
Foundation to Z.G.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200801149.
7896
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 7896 –7900

Angewandte
Chemie
not 1-Mn or 1-Ga, increases cellular cholesterol efflux from
macrophages. The iron complex is more effective than natural
antioxidants for attenuation of atherosclerosis development
in mice, which is attributable to the synergetic effects that
were deduced from the in vitro investigations.
These findings are consistent with the differences deter-
mined so far in the modes of action of these complexes on
ROS and RNS. No nitrating species are released in the
catalytic cycle for decomposition of peroxynitrite by 1-Mn
[Scheme 1B, Eq. (1)], [9a] which is most likely not true for 1-Fe
[Scheme 1B, Eq. (2)].[15] On the other hand, the catalytic rate
for decomposition of hydrogen peroxide (the precursor of
COH in Cu-induced oxidations) by 1-Fe appears to be larger
than that of 1-Mn,[16] thus explaining the better antioxidant
properties of the former.
For initial appreciation regarding possible superiority of
the synthetic catalysts relative to natural antioxidants, we
determined the effect of excess (45–135 mm) punicalagin—the
active polyphenolic hydrolyzable tannin ingredient of pom-
egranate juice[11]—on the decomposition of peroxynitrite
(40 mm). This information, together with the previously
determined rate constants for catalytic decomposition of
peroxynitrite (385 mm) by 1-Fe and 1-Mn (5–20 mm;
Scheme 1),[9a] were used for calculating the minimal concen-
trations required for reducing the half-life of peroxynitrite by
50%. The results revealed that 250 mol% punicalagin,
2.5 mol% 1-Mn, and 0.05 mol% 1-Fe are needed for that
purpose, that is, the catalytically acting 1-Fe and 1-Mn are
5000 and 100 times, respectively, more effective than the most
potent dietary antioxidant, pomegranate punicalagin, which
acts in a sacrificial mode.
Extension of purely chemical findings into any in vivo
system is very challenging,[18] and could be particularly severe
for metal complexes of corrole 1 that were shown not to enter
cells in protein-free medium and to form tightly bound
noncovalent conjugates with serum proteins.[10] Accordingly,
mutual interactions between the complexes and lipoproteins
were checked to test the probability of the former reaching
the place where the latter are oxidized—the arterial wall.
Recording of the electronic spectra of aqueous (protein-free)
solutions of metal complexes of corrole 1 before and after
addition of LDL or HDL, as well as after dialysis, revealed the
binding stoichiometry: 40 Æ 5 and about 10 corrole molecules
bind with high affinity to each LDL and HDL particle,
respectively, which is significantly more than the number of
natural antioxidants present therein.[2c] The interpretation of
these data is, however, quite restricted because they do not
provide information about the dissociation constants of the
corroles from the lipoproteins, or about the affinity relative to
other serum proteins. These concerns were addressed by
density-gradient ultracentrifugation of whole serum preincu-
bated with metallocorroles. Figure 1clearly shows that these
corroles preferably bind to lipoproteins at the expense of all
serum proteins,[19] and spectral analysis of the separated
fractions (see the Supporting Information) implies that 65–
70% of the corroles are bound to HDL. These results suggest
that the lipoproteins might carry the metallocorroles all the
way to the arterial wall, where the antioxidant properties of
the latter are needed.
Another early investigation focused on the kind of
peroxynitrite-induced damage that may be eliminated by
the metallocorroles, by utilizing small molecules that repre-
sent targets for hydroxyl radicals and nitrogen dioxide
(Table 1).[12,13] Catalytic amounts of 1-Fe or 1-Mn rescued
Table 1: The inhibitory effect of metallocorroles on oxidation/nitration of
small molecules by peroxynitrite; pH 7.4, T=258C.
Reactive Substrate
species
Product
Yield [%] relative
to peroxynitrite
no additive 1-Mn 1-Fe
COH
COH
CNO2
CNO2
DMSO
formaldehyde
11.5
0
0
0
0
0
0
deoxyribose malondialdehyde 1.7
fluorescein
l-tyrosine
nitrofluorescein
3-nitro-l-tyrosine 11
30[a]
4[a]
11
[a] Yield relative to fluorescein.
dimethyl sulfoxide (DMSO) and deoxyribose from being
oxidized by peroxynitrite-derived hydroxyl radicals, but only
1-Mn completely eliminated CNO2-accredited reactions (1-Fe
had no effect on the nitration of tyrosine, and was only
partially effective in inhibiting that of fluorescein). With
CuSO4/glutathione as the initiator of ROS (by the complex
Fenton-type oxidation that produces the hydroxyl radical by
the involvement of a superoxide anion radical and hydrogen
peroxide),[14] 1-Fe eliminated the oxidation of DMSO to
formaldehyde completely, whereas 1-Mn did it only partially
(65% inhibition). These investigations clearly point towards
the following conclusions: 1) both corrole metal complexes
serve very well for preventing the formation of the hydroxyl
radical from peroxynitrite; 2) the iron complex is more
efficient in rescuing molecules from copper-induced forma-
tion of hydroxyl radicals; and 3) the manganese complex is
more efficient in inhibiting the formation of RNS from
peroxynitrite.
Figure 1. Density-gradient ultracentrifuged serum with a) no additive,
b) 20 mm 1-Mn, c) 40 mm 1-Mn, d) 20 mm 1-Fe, and e) 40 mm 1-Fe. The
yellow rings arise from different serum fractions and the green color
from the associated corroles. VLDL= very-low-density lipoprotein;
LPDS=lipoprotein-deficient serum.
Angew. Chem. Int. Ed. 2008, 47, 7896 –7900
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7897

Communications
Examination of the different fractions isolated from the
above-mentioned experiments (Figure 1) further revealed
that the amount of cholesterol in LDL from corrole-treated
serum was lowered by about 20%, whereas it was increased in
the HDL and LPDS fractions. This finding suggests that
strong binding of the metallocorroles does affect cholesterol
distribution. A further indication of a favorable alteration of
cellular cholesterol transport by metallocorroles was obtained
by studying cholesterol efflux from J-774 macrophages in a
lipoprotein-free medium. The iron (but not the manganese or
gallium) complex of corrole 1 increased the efflux by up to
20% in a dose-dependent fashion.
The effect of the metallocorroles on oxidative damage to
LDL and HDL solutions, tested by established methods,[20,21]
revealed the following results. The formation of conjugated
dienes (the earliest intermediates of lipid peroxidation) by
treatment of LDL with SIN-1, a reagent that slowly generates
peroxynitrite,[22] was increased by 1-Mn and completely
eliminated by 1-Fe (see the Supporting Information). Experi-
ments performed according to the most commonly used
copper-induced protocol revealed that in the presence of 1-
Mn, the delay time in formation of conjugated dienes (as a
result of consumption of “free radicals” by endogenous
antioxidants) was shortened from 30 to 15 min (Figure 2A),
the concentration of hydroperoxides rapidly reached a
maximum at 45 min (because of subsequent transformation
to the final aldehyde products, Figure 2B), and aldehydes
were formed faster and to a much larger extent than in the
control experiment (Figure 2C). On the other hand, practi-
cally no damage was observed in the presence of 1-Fe by all the
above three criteria (Figure 2A–C), and even after 24 h no
thiobarbituric acid reactive substances (TBARS) were
detected. Similar results were obtained for copper-induced
oxidation of HDL. Importantly, approved drugs, such as
carvedilol, do not suppress the oxidation of isolated LDL
induced by peroxyl radicals or cupric ions.[17]
Figure 2. Effect of corrole (2.5 mm) on the kinetics of CuSO4-induced
(5 mm) oxidation of LDL (100 mg proteinLÀ1). A) Conjugated diene
formation followed at 234 nm; B) lipid peroxide (PD) formation;
C) aldehyde formation followed by TBARS measurement; and
^
D) TBARS formation after 2 h for various corrole concentrations.
~
&
with 1-Mn; with 1-Fe, and without corrole for (A)–(C) and with
1-Ga for (D). MDA=malondialdehyde.
The dependence of LDL oxidation on the corrole
concentration was examined by analyses performed 2 h
after the copper-initiated oxidation (Figure 2D). The non-
redox-active gallium(III) complex 1-Ga, examined to elimi-
nate a possible effect of the corrole macrocycle itself,
provided exactly the same results as those obtained without
any corrole-based additive. In the presence of 1-Mn, up to
three times more TBARS than without any additive was
obtained in a concentration-dependent manner. On the
contrary, the iron complex 1-Fe caused a minor increase in
TBARS at a concentration of 0.5 mm, but complete inhibition
was obtained at concentrations of 2.5 mm and higher
(unmatched by any previously reported complex, vide infra).
The dramatic differences between the two transition-metal
complexes may be explained based on the mechanisms shown
in Scheme 1: 1-Mn efficiently traps hydrogen peroxide (the
precursor of COH in Cu-initiated oxidations), but the (oxo)-
manganese(V) intermediate (k1) reacts with lipid components
(k3) faster than with another hydrogen peroxide molecule (k2)
because of the very large differences in their local concen-
trations. On the other hand, the overall catalytic rate of the
iron corrole is much faster and there is no evidence for the
buildup of any oxidizing reaction intermediate.
The encouraging in vitro data with the metal complexes of
corrole 1 were extended towards an in vivo investigation on
apolipoprotein E-deficient (E0) mice, the most common
murine model used for atherosclerosis development studies.
E0 mice are hypercholesterolemic and develop spontaneous
atherosclerotic lesions similar in development and morphol-
ogy to those in humans.[23] The mice were divided into four
groups that differed only in the type of supplied drinking
water: no additive, or water containing either 1-Mn, 1-Fe, or
1-Ga (0.2 mg per mouse per day). Histopathological develop-
ment of lesions in the aorta obtained from mice sacrificed
10 weeks after treatment started revealed that the average
lesion area was very high for untreated mice and for those that
received 1-Ga (Figure 3A). A small decrease in average
lesion area (17% relative to the control group) was obtained
for the 1-Mn group, while the results obtained with 1-Fe were
truly dramatic. Two of the mice receiving that treatment did not
develop atherosclerotic lesions at all, and the average lesion
area for the other four mice was 60% smaller than in the
control group. These results are much better than those
obtained by other compounds, including E0 mice that were
treated by the most potent natural antioxidants (48 and 44%
7898 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 7896 –7900

Angewandte
Chemie
presence of reductants, such as ascorbate[27] or uric acid, at
concentrations that are 40 times larger than those needed for
full inhibition by 1-Fe.[28] On the other extreme, hemin and
structurally related iron porphyrins that associate very
strongly with lipoproteins are pro- rather than antioxidants.[29]
Differences in the mechanism and the rate of decomposition
of ROS and RNS are apparently responsible for the anti- and
pro-oxidant properties of 1-Fe and 1-Mn, respectively, as well
as for 1-Mn being more efficient in eliminating RNS. In fact,
the somewhat positive in vivo effect of 1-Mn may possibly be
traced to its effect on nitrosative stress. Future studies will be
devoted to tuning of the catalystꢀs ability regarding efficient
decomposition of other ROS/RNS and better suppression of
nitration, as well as to additional biochemical investigations.
One example of the last aspect is that the preliminary results
are indicative of increased macrophage lactonase activity
(probably related to the cellular antioxidant enzyme, para-
oxonase 2)[2a] in mice that consumed the metallocorroles.
Apparently, the combination of several favorable effects is
responsible for the quite dramatic effect displayed by the
specific iron corrole described herein as regards the rescue of
LDL/HDL from oxidative damage and the attenuation of
atherosclerosis in mice.
Figure 3. Effect of corrole consumption on the formation of lesions in
E0 mice. A) Size of atherosclerotic lesions in the different groups.
Values are means Æ standard error of the mean, n=6. For the 1-Fe
groupthe mean is for n=4 because two of the mice in the groupdid
not show formation of a lesion at all. B) Cross-section from the aortic
arch of a control mouse. C) Cross-section from the aortic arch of a
mouse that consumed 1-Fe. The lipid components are stained a
brown-black color, thus highlighting the foam cells.
Received: March 10, 2008
Revised: June 30, 2008
Published online: September 16, 2008
reduction for consumption of red wine and pomegranate
juice, respectively).[3]
Examination of the sacrificed mice revealed another
important result: the total blood cholesterol levels in mice
that received the metallocorroles were significantly reduced
(1-Fe: 40%, 1-Mn: 26%, 1-Ga: 20%) relative to the control
animals. These ex vivo observations are nicely correlated with
the in vitro results described earlier, which revealed that the
amount of cholesterol in the LDL fractions of corrole-treated
serum was lowered by about 20% and that 1-Fe even affected
cellular efflux of cholesterol.
All together, the observations obtained in the mouse
model of atherosclerosis appear to be very much consistent
with the in vitro results, which disclosed that the iron complex
1-Fe is a very potent catalytic antioxidant that also affects
cholesterol efflux from macrophages and lowers LDL cho-
lesterol levels. The unique bipolarity of corrole 1 is apparently
responsible for the selective binding to lipoproteins, which is
very important for their protection from oxidative damage
and may serve as a mechanism for delivering the complexes to
the arterial wall. This finding, together with the very fast
catalytic rates of 1-Fe, is of particular relevance in the context
of the comparison with dietary antioxidants: the noncatalytic
activity of polyphenols was shown to occur at concentrations
at least one order of magnitude higher than their bioavaila-
bility.[24]
Keywords: antioxidants · atherosclerosis · corroles · iron ·
manganese
.
[1] K. J. Barnham, C. L. Masters, A. I. Bush, Nat. Rev. Drug
Discovery 2004, 3, 205 – 214.
[2] a) M. Aviram, M. Rosenblat, Free Radical Biol. Med. 2004, 37,
1304 – 1316; b) D. Steinberg, S. Parthasarathy, T. E. Carew, J. C.
Khoo, J. L. Witztum, N. Engl. J. Med. 1989, 320, 915 – 924;
c) Oxidative Stress, Lipoproteins and Cardiovascular Dysfunc-
tion (Eds.: C. Rice-Evans, K. R. Bruckdorfer), Portland,
London, 1995.
[3] a) T. Hayek, B. Fuhrman, J. Vaya, M. Rosenblat, P. Belinky, R.
Coleman, A. Elis, M. Aviram, Arterioscler. Thromb. Vasc. Biol.
1997, 17, 2744 – 2752; b) M. Aviram, L. Dornfeld, M. Rosenblat,
N. Volkova, M. Kaplan, R. Coleman, T. Hayek, D. Presser, B.
Fuhrman, Am. J. Clin. Nutr. 2000, 71, 1062 – 1076.
[4] a) R. Stocker, J. F. Keaney, J. Thromb. Haemostasis 2005, 3,
1825 – 1834; b) G. M. Puddu, E. Cravero, G. Arnone, A. Muscari,
P. Puddo, J. Biomed. Sci. 2005, 12, 839 – 853; c) K. J. Williams,
E. A. Fisher, Curr. Opin. Clin. Nutr. Metab. Care 2005, 8, 139 –
146.
[5] C. Szabó, H. Ischiropoulos, R. Radi, Nat. Rev. Drug Discovery
2007, 6, 662 – 680.
[6] G. Peluffo, R. Radi, Cardiovasc. Res. 2007, 75, 291– 302.
[7] a) D. Bartlett, D. F. Church, P. L. Bounds, W. H. Koppenol, Free
Radical Biol. Med. 1995, 18, 85 – 92; b) M. Aviram, M. Kaplan,
M. Rosenblat, B. Fuhrman, Handb. Exp. Pharmacol. 2005, 170,
263 – 300; c) B. Olas, P. Nowak, J. Kolodziejczyk, M. Ponczek, B.
Wachowicz, J. Nutr. Biochem. 2006, 17, 96 – 102; d) N. Noguchi,
M. Takahashi, J. Tsuchiya, H. Yamashita, E. Komuro, E. Niki,
Biochem. Pharmacol. 1998, 55, 785 – 791.
The structure of corrole 1 and its metal complexes is also
distinctly different from that of all metalloporphyrin-based
decomposition catalysts of peroxynitrite reported to date,[25]
wherein the symmetrical distribution of (either negative or
positive) charges is likely to preclude their association with
lipoproteins.[26] In fact, such synthetic iron porphyrins do not
display potency for inhibiting LDL oxidation, and manganese
porphyrins protect LDL from peroxynitrite only in the
[8] J. A. Hunt, J. Lee, J. T. Groves, Chem. Biol. 1997, 4, 845 – 858.
[9] a) A. Mahammed, Z. Gross, Angew. Chem. 2006, 118, 6694 –
6697; Angew. Chem. Int. Ed. 2006, 45, 6544 – 6547; b) Z. Gersh-
Angew. Chem. Int. Ed. 2008, 47, 7896 –7900
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7899

Communications
man, I. Goldberg, Z. Gross, Angew. Chem. 2007, 119, 4398 –
lou, J. A. Parkinson, K. J. Barnham, A. Tucker, J. Woodrow, P. J.
Sadler, J. Biol. Chem. 1998, 273, 14721 – 14730.
4402; Angew. Chem. Int. Ed. 2007, 46, 4320 – 4324.
[10] a) A. Haber, H. Agadjanian, L. K. Medina-Kauwe, Z. Gross, J.
Inorg. Biochem. 2008, 102, 446 – 457; b) H. Agadjanian, J. J.
Weaver, A. Mahammed, A. Rentsendorj, S. Bass, J. Kim, I. J.
Dmochowski, R. Margalit, H. B. Gray, Z. Gross, L. K. Medina-
Kauwe, Pharm. Res. 2006, 23, 367 – 377; c) A. Mahammed, H. B.
Gray, J. J. Weaver, K. Sorasaenee, Z. Gross, Bioconjugate Chem.
2004, 15, 738 – 746.
[11] R. Tzulker, I. Glazer, I. Bar-Ilan, D. Holland, M. Aviram, R.
Amir, J. Agric. Food Chem. 2007, 55, 9559 – 9570.
[12] R. Shimanovich, J. T. Groves, Arch. Biochem. Biophys. 2001,
387, 307 – 317.
[13] Oxidation of small molecules: all reactions were performed with
no additive or with 1-Fe or 1-Mn in a phosphate buffer solution
(pH 7.4). a) Formaldehyde from DMSO (160 mm catalyst):
DMSO (140 mm) was reacted with peroxynitrite (1.25 mm);
b) malondialdehyde from deoxyribose (1 mm catalyst): deoxyri-
bose (10 mm) was reacted with peroxynitrite (390 mm); c) nitra-
tion of fluorescein (5 mm catalyst): fluorescein (5 mm) was reacted
with peroxynitrite (25 mm); d) nitration of l-tyrosine (50 mm
catalyst): l-tyrosine (80 mm) was reacted with peroxynitrite
(220 mm); e) copper-induced oxidation of DMSO (50 mm cata-
lyst): CuSO4 and phenanthroline (8 mm each) were added to
DMSO (140 mm), followed by sodium ascorbate (500 mm) or
glutathione (1m m).
[19] The Kd for the HDL–corrole conjugates may be estimated as
much lower than 10À9 mÀ1 from the relative concentrations of
HDL and albumin in serum and the previously determined Kd of
serum albumin–corrole conjugates (reference [10c]).
[20] M. Aviram, J. Vaya, Methods Enzymol. 2001, 335, 244 – 256.
[21] Oxidation of LDL: LDL (100 mg proteinLÀ1) in phosphate-
buffered saline was incubated for 30 min without any additive or
with various concentrations of 1-Mn, 1-Fe, or 1-Ga. a) By
peroxynitrite: oxidation was induced by addition of SIN-1
(250 mm) and conjugated dienes were continuously monitored
for 4 h at 378C; b) by copper sulfate: a freshly prepared CuSO4
solution (5 mm) was added and the mixture was incubated at
378C. Oxidation was determined by measuring TBARS, con-
jugated dienes, and lipid peroxides by previously described
methods (reference [23]).
[22] F. J. Martin-Romero, Y. Gutiꢁrrez-Martin, F. Henao, C. Gutiꢁr-
rez-Merino, J. Fluoresc. 2004, 14, 17 – 23.
[23] R. Coleman, T. Hayek, S. Keidar, M. Aviram, Acta Histochem.
2006, 108, 415 – 424.
[24] F. Ferroni, A. Maccaglia, D. Pietraforte, L. Turco, M. Minetti, J.
Agric. Food Chem. 2004, 52, 2866 – 2874.
[25] T. P. Misko, M. K. Highkin, A. W. Veenhuizen, P. T. Manning,
I. K. Stern, M. G. Currie, D. Salvemini, J. Biol. Chem. 1998, 273,
15646 – 15653.
[26] P. C. de Smidt, A. J. Versluis, T. J. C. van Berkel, Biochemistry
1993, 32, 2916 – 2922.
[27] J. Lee, J. A. Hunt, J. T. Groves, Bioorg. Med. Chem. Lett. 1997, 7,
2913 – 2918.
[28] a) B. J. Day, I. Batinic-Haberle, J. D. Crapo, Free Radical Biol.
Med. 1999, 26, 730 – 736; b) A. Trostchansky, G. Ferrer-Sueta, C.
Batthyany, H. Botti, I. Batinic-Haberle, R. Radi, H. Rubbo, Free
Radical Biol. Med. 2003, 35, 1293 – 1300.
[29] G. Camejo, C. Halberg, A. Manschik-Lundin, E. Hurt-Camejo,
B. Rosengren, H. Olsson, G. I. Hansson, G. B. Forsberg, B.
Ylhen, J. Lipid Res. 1998, 39, 755 – 766.
[14] N. K. Urbanski, A. Beresewicz, Acta Biochim. Pol. 2000, 47,
951– 962.
[15] J. L. Bourassa, E. P. Ives, A. L. Marqueling, R. Shimanovich, J. T.
Groves, J. Am. Chem. Soc. 2001, 123, 5142 – 5143.
[16] A. Mahammed, Z. Gross, J. Am. Chem. Soc. 2005, 127, 2883 –
2887.
[17] N. Noguchi, K. Nishino, E. Niki, Biochem. Pharmacol. 2000, 59,
1069 – 1076.
[18] a) C. R. Schlieve, C. J. Lieven, L. A. Levin, Invest. Ophthalmol.
Visual Sci. 2006, 47, 3878 – 3886; b) A. I. Ivanov, J. Christodou-
7900 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 7896 –7900