Metalloenzyme-like Activity of Alzheimer’s Disease ?-Amyloid
Cu-DEPENDENT CATALYTIC CONVERSION OF DOPAMINE, CHOLESTEROL, AND BIOLOGICAL REDUCING
AGENTS TO NEUROTOXIC H2O2*
Received for publication, June 28, 2002, and in revised form, August 12, 2002
Published, JBC Papers in Press, August 20, 2002, DOI 10.1074/jbc.M206428200
Carlos Opazo‡, Xudong Huang§, Robert A. Cherny¶, Robert D. Moir?, Alex E. Roher**,
Anthony R. White¶, Roberto Cappai¶, Colin L. Masters¶, Rudolph E. Tanzi?,
Nibaldo C. Inestrosa‡, and Ashley I. Bush§¶‡‡
From the ‡Centro de Regulacio ´n Celular y Patologı ´a, Departamento de Biologı ´a Celular y Molecular,
Facultad de Ciencias Biolo ´gicas, Pontificia Universidad Cato ´lica de Chile, Santiago 114-D, Chile,
§Laboratory for Oxidation Biology, Genetics and Aging Research Unit, and Department of Psychiatry,
Harvard Medical School, Massachusetts General Hospital, Charlestown, Massachusetts 02129,
¶Oxidation Disorders Laboratory, Mental Health Research Institute of Victoria and Department of Pathology,
the University of Melbourne, Parkville, Victoria 3052, Australia, ?Genetics and Aging Research Unit and Department
of Neurology, Harvard Medical School, Massachusetts General Hospital, Charlestown, Massachusetts 02129, and the
**Haldeman Laboratory for Alzheimer Disease Research, Sun Health Research Institute, Sun City, Arizona 85351
?-Amyloid (A?) 1–42, implicated in the pathogenesis of
Alzheimer’s disease, forms an oligomeric complex that
binds copper at a CuZn superoxide dismutase-like bind-
ing site. A??Cu complexes generate neurotoxic H2O2from
O2through Cu2?reduction, but the reaction mechanism
has been unclear. We now report that A?1–42, when bind-
ing up to 2 eq of Cu2?, generates the H2O2catalytically by
recruiting biological reducing agents as substrates under
conditions where the Cu2?or reducing agents will not
form H2O2themselves. Cholesterol is an important sub-
strate for this activity, as are vitamin C, L-DOPA, and
dopamine (Vmaxfor dopamine ? 34.5 nM/min, Km? 8.9 ?M).
The activity was inhibited by anti-A? antibodies, Cu2?
chelators, and Zn2?. Toxicity of A? in neuronal culture
was consistent with catalytic H2O2production. A? was
not toxic in cell cultures in the absence of Cu2?, and do-
pamine (5 ?M) markedly exaggerated the neurotoxicity of
200 nM A?1–42?Cu. Therefore, microregional catalytic
H2O2production, combined with the exhaustion of reduc-
ing agents, may mediate the neurotoxicity of A? in Alzhei-
mer’s disease, and inhibitors of this novel activity may be
of therapeutic value.
A?1characteristically collects in the neocortex in AD.
A?1–40 is the major soluble species (1), and A?1–42 is a minor
species but is enriched in plaque amyloid (2). Familial AD-
linked mutations of amyloid protein precursor, presenilin-1
and presenilin-2, increase the concentration of A?1–42 (3).
A?1–42 is toxic in primary neuronal culture at ?M concentra-
tions (4), by an unclear mechanism likely to be mediated by
reactive oxygen species generation (5, 6). We recently reported
that A? binds Cu2?with very high affinity (KA?1–42? 8 atto-
molar) (7, 8), forming an allosterically cooperative Cu2?coor-
dination site that resembles superoxide dismutase 1 (9). The
A??Cu2?complex is redox-active and produces H2O2from O2
through the reduction of Cu2?(10, 11). The H2O2that A?
directly generates contributes to the neurotoxicity of the pep-
tide in primary neuronal cultures (11). This is important be-
cause of the marked H2O2-mediated damage to neocortical
tissue in AD (12, 13) and in amyloid-bearing transgenic mice
The metal dependence of the generation of H2O2by A? may
be a target for AD therapeutics. Cu and Zn are markedly
elevated in amyloid plaques (15, 16). Therefore it is significant
that H2O2generation by A? in vitro is abolished by chelators
(10, 11), such as clioquinol, which we have reported recently
has in vivo efficacy in blocking brain A? accumulation in
Tg2576 mice (17).
The chemical origin of the electrons in the A?-mediated
reduction of O2to generate H2O2(10, 11, 18) remained to be
clarified. Here we elaborate this reaction mechanism, demon-
strating the donation of electrons from biological reducing
agents (e.g. cholesterol, catecholamines, vitamin C). We further
demonstrate that this reaction is catalytic and that the A??Cu
complex is a cuproenzyme that can be purified from the human
brain. The neurotoxicity of A? in cell culture is shown to be
completely dependent upon both the binding of Cu2?and the
simultaneous presence of a reducing agent. Taken together,
these data argue that A? toxicity is because of abnormal enzy-
matic activity, which represents an important target for med-
ical chemistry approaches for the treatment of AD.
Reagents—A? peptides were synthesized by the W. Keck Laboratory,
Yale University, New Haven, CT. Confirmatory data were obtained by
* This work was supported in part by National Health & Medical
Research Council, National Institutes of Health Grants 2R01AG12685
(to A. I. B.) and 1K01MH02001 (to X. H.), Prana Biotechnology Ltd.,
and Grants Fondo de Investigacio ´n Avanzada en A´reas Prioritarias
number 13980001, CIMM-ICA/006 (to N. C. I.) and Fondo Nacional de
Desarrollo Cientı ´fico y Tecnolo ´gico number 2990087, Direccio ´n de In-
vestigacio ´n y Postgrado (to C. O.). The costs of publication of this article
were defrayed in part by the payment of page charges. This article must
therefore be hereby marked “advertisement” in accordance with 18
U.S.C. Section 1734 solely to indicate this fact.
‡‡ To whom correspondence should be addressed: Laboratory for Ox-
idation Biology, Genetics and Aging Research Unit, Massachusetts
General Hospital, Bldg. 114, 16th St., Charlestown, MA 02129. Tel.:
1The abbreviations used are: A?, ?-amyloid; AD, Alzheimer’s dis-
ease; DA, dopamine; DCF, 2?,7?-dichlorofluorescin diacetate; TCEP,
tris(2-carboxyethyl)phosphine hydrochloride; HRP, horseradish perox-
idase; PBS, phosphate-buffered saline; CDTA, 1,2-cyclohexylenedini-
L-DOPA, levo-dihydroxyphenylalanine; MTS,
nyl)-2H-tetrazolium; DTPA, diethylene triamine pentaacetic acid;
DETC, diethyl dithiocarbamate.
THE JOURNAL OF BIOLOGICAL CHEMISTRY
© 2002 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 277, No. 43, Issue of October 25, pp. 40302–40308, 2002
Printed in U.S.A.
This paper is available on line at http://www.jbc.org
by guest on December 26, 2015
employing A? peptides synthesized in the laboratory of Dr. C. Glabe,
University of California, Irvine, CA. A? peptide stock solutions were
prepared in water treated with Chelex (Bio-Rad) and quantified accord-
ing to established procedures (19). Cu2?-Gly (1:6) was prepared as
described previously (8). Monoclonal anti-A? antibodies 4G8 (which
detects A? residues 17–21) and 6E10 (which detects residues 9–16)
were obtained from Senetek, and 10H3 (immunogen is residues 1–28 of
A?) were obtained from Pierce Endogen. The other reagents were ob-
tained from Sigma unless otherwise noted.
Hydrogen Peroxide Assay—Dichlorofluorescein diacetate (DCF; Mo-
lecular Probes, Eugene, OR) was dissolved (5 mM) in 100% dimethyl
sulfoxide (argon purged for 2 h at 20 °C), deacetylated with 0.25 M
NaOH for 30 min, and then neutralized, pH 7.4, to a final concentration
of 1 mM. Horseradish peroxidase (HRP) stock solution was prepared to
1 ?M in PBS, pH 7.4. The reactions were carried out in PBS, pH 7.4,
under ambient conditions in a 96-well plate (250 ?l/well) containing
freshly prepared synthetic peptide (up to 1 ?M), Cu-Gly (up to 2 ?M),
reducing agents (up to 10 ?M), deacetylated DCF (100 ?M), and HRP
(0.1 ?M), incubated at 37 °C. 10 ?M EDTA was included to prevent
reactions with contaminating concentrations (?0.2 ?M) of free Cu2?.
The concentrations of A? used varied to bring readout values into a
convenient target range. For example, in studies to determine which
reducing agents would promote H2O2production (see Fig. 1) relatively
higher peptide concentrations (1 ?M) were studied so that low produc-
tion of H2O2would not be overlooked. In most experiments, though, A?
was used at 200 nM to reflect the soluble concentration in the AD brain
(20). Studies were completed on the day of reagent preparation. Reac-
tions were conducted in the dark to avoid photodynamic effects. The
signal specific for H2O2was the decrease in fluorescence of parallel
samples coincubated with catalase (4000 units/ml; 10 ?M). Fluorescent
readings were recorded by a Packard Fluorocount plate reader (485 nm
excitation, 530 nm emission), against a standard curve of reagent grade
H2O2in PBS, pH 7.4.
Primary Neuronal Cultures—Cortical neuronal cultures were pre-
pared as described previously (18) with some modifications. Briefly,
embryonic day 18 BL6Jx129sv mouse cortices were removed, dissected
free of meninges, and dissociated in 0.025% (w/v) trypsin. Dissociated
cells were plated in poly-L-lysine-coated 48-well culture plates (Nunc) at
a density of 1 ? 105cells/cm2in plating medium (minimum Eagle’s
medium with 10% fetal calf serum and 5% horse serum). Cultures were
maintained at 37 °C in 5% CO2for 2 h before the plating medium was
replaced with Neurobasal growth medium containing B27 supplements.
This method resulted in cultures highly enriched for neurons (?5%
glia). After 6 days in culture, the medium was replaced with fresh
Neurobasal medium supplemented with B27 lacking antioxidants (to
minimize scavenging of H2O2generated in the medium) or Locke’s
buffer, which contained the following (in mM): NaCl, 154; KCl, 5.6;
CaCl2, 2.3; MgCl2, 1; NaHCO3, 3.6; glucose, 5; Hepes, pH 7.4, 5, into
which experimental compounds were added.
For treatment of neuronal cultures fresh A? stock solution was
diluted (final concentration 0.2–2.5 ?M) in Neurobasal medium or
Locke’s buffer and coincubated (15 min, 20 °C) ? Cu-Gly (0.4–5 ?M), DA
(5–20 ?M), catalase (1000 unit/ml), or Me2SO (98 ?M). The mixtures
were then added to neuronal cultures for up to 24 h, and cell viability
was then assayed.
Cell Viability Assays—Cell survival was monitored by phase contrast
microscopy, and cell viability was quantitated using the MTS assay as
described previously (21). Briefly, cells were washed two times with 250
?l of Locke’s buffer, then placed in Neurobasal medium (250 ?l), and 25
?l of MTS (Promega) was added to each well and incubated for 4 h in a
48-well plate. Absorbance (490 nm) was determined using a Wallac
Victor Multireader and compared with MTS incubated in cell-free me-
dium. The data were calculated as a percentage of values from un-
TCEP, used in our previous studies to detect H2O2, may have
served as the source of electrons to form H2O2from O2, because
it is also a reducing agent (10, 11). To evaluate this possibility
we coincubated A? complexed with Cu2?(A??Cu) ? TCEP
using the DCF assay to measure H2O2. A ratio of 2Cu:A? was
chosen for initial studies, because A?1–42 can bind up to two
Cu atoms at pH 7.4 (8). TCEP was indeed found to be a signif-
icant substrate for H2O2formation by A??Cu (Fig. 1A); however
H2O2formation by A??Cu alone was not observed. HRP is used
in the DCF-based H2O2assay to catalyze the reaction of H2O2
with the DCF detection reagent. As negative controls, we con-
firmed that A??Cu did not increase DCF fluorescence in the
absence of HRP and that catalase abolished DCF fluorescence
in these and subsequent assays (not shown). Therefore, DCF is
a suitable reagent to detect H2O2production by A?. Further-
more, in the absence of A?, the Cu2?(?2 ?M) alone did not
generate H2O2in these and subsequent experiments, indicat-
ing that EDTA at this concentration (10 ?M, present in all
subsequent experiments) blocks H2O2generation by any pos-
sible free Cu?.
Having established that a non-biological reducing agent
(TCEP) will serve as a reservoir substrate for the generation of
H2O2by A??Cu, we next surveyed a variety of neurochemicals
to identify candidate biological substrates for the same reaction
(Fig. 1B). H2O2production was greatest for the stronger bio-
chemical reducing agents, vitamin C (1.9 ?M) ? DA (1.7 ?M) ?
L-DOPA (1.3 ?M). Other agents that promoted significant H2O2
production included cholesterol, progesterone (0.5 ?M), epi-
nephrine (0.4 ?M), norepinephrine, serotonin, and NADPH (0.3
?M). Equivocal H2O2levels were generated by NADH and ?-es-
tradiol. Acetylcholine, histamine, L-phenylalanine, L-tyrosine,
and methionine were all unable to generate H2O2in the pres-
ence of A??Cu.
To characterize the enzymology of catalytic H2O2production
by A??Cu in vitro, we elected DA as the test substrate, because
its redox chemistry is well understood. We first optimized H2O2
production as a product of the stoichiometry of Cu bound to
A?1–42. Under the same conditions as in Fig. 1B, we incubated
FIG. 1. Characterization of H2O2production by A??Cu. A, TCEP
is a substrate for H2O2production by the A??Cu complex. A?1–42 (1 ?M)
and Cu2?-Gly (2 ?M) (A?Cu2) were incubated ? TCEP (10 ?M), and
H2O2levels were measured by DCF assay. Control incubations were
performed in the absence of A?. B, differential effect of reducing agents.
A?Cu2was incubated with different reducing agents (10 ?M), and H2O2
levels were measured. C, optimal stoichiometry of A?1–42?Cu complex.
Synthetic human or rat A?1–42 (200 nM) was incubated with increasing
concentrations of Cu2?-Gly (0–800 nM) in the presence of DA (5 ?M),
and H2O2levels were measured. D, dependence of H2O2production
upon A?Cu2concentration. Increasing concentrations of A?Cu2were
incubated with DA (5 ?M), and H2O2produced was measured as a
product of A? concentration. All incubations were in PBS, pH 7.4, 10 ?M
EDTA for 60 min at 37 °C. Data are means ? S.E.
Catalysis of Hydrogen Peroxide Production by A?
by guest on December 26, 2015
A?1–42 (200 nM) with Cu2?(0 to 1000 nM) and assayed result-
ant H2O2levels generated in the presence of DA (5 ?M) and
EDTA (Fig. 1C). H2O2production by A??Cu saturated at a
molar ratio of 2Cu:1A? (A?Cu2) suggesting that there the two
Cu binding sites coordinated by each A? monomer at pH 7.4 (8)
both support redox activity. The saturation of H2O2production
in this experiment confirms that only the Cu2?bound to A?,
and not free Cu2?, contributed to the production of H2O2.
We studied the dynamic range of H2O2production from
A?Cu2by incubating the complex (0–900 nM A?) with DA (5
?M). H2O2production was linearly dependent on A?Cu2con-
centrations between 100 and 600 nM (Fig. 1D). To determine
the Michaelis-Menten relationships between A?Cu2and DA,
we demonstrated that the rate of H2O2production by A?Cu2
(200 nM A?) was dependent upon DA concentration (0 to 10 ?M)
(Fig. 2A). The data obtained at 60 min of incubation were
plotted using a Lineweaver-Burk equation (Fig. 2B). This re-
vealed a linear relationship (R2? 0.998) between 1/v and 1/DA
and a Vmaxof 34.5 nM/min and Kmof 8.9 ?M. 60 min was chosen
as the interval for the Lineweaver-Burk transformation, be-
cause it was midpoint of the 120-min study, and because H2O2
production was occurring at a constant rate for all DA concen-
trations during the incubation period. The transformation
yielded similar values for data from the 30-, 90-, and 120-min
incubation values (not shown).
Because reaction with DA may alter the redox activity of
A??Cu, we repeated some previous observations of A?-mediated
H2O2production, this time in the presence of DA. A?Cu2(0.5
?M) ? DA (5 ?M) generated no detectable superoxide (O2.) using
dihydroethidium (20 ?M; Molecular Probes) as an indicator
following 1 h of incubation. Catalytic H2O2was readily de-
tected from A?Cu2at 200 nM (Fig. 1, B and C), so lack of
detectable O2.from a 2.5-fold greater concentration of A?Cu2
suggests that H2O2is not generated by disproportionation of
O2., in agreement with our previous report (10). The H2O2
catalytic activity of variant A? peptides (prepared as A?Cu2),
using DA as substrate, was again found to exhibit the same
relative redox activities (A?1–42 ? A?1–40 ? rat A?1–42 ?
rat A?1–40) (see Figs. 1C and 3A) as reported previously,
correlating with the respective abilities of the peptides to re-
duce Cu2?and their toxicity in neuronal culture (10, 11). The
lower H2O2catalytic activity of rat A?1–42 compared with
human A?1–42 did not appear to be because of decreased Cu2?
affinity or stoichiometry, because excess Cu2?added to rat
A?1–42 did not enhance its activity to the same level as human
A?1–42, saturating at a binding ratio of 2Cu:1A? (Fig. 1C).
Importantly, cerebral amyloid deposits are scarce in aged rats
and mice (22), which possess a homologue with three amino
acid substitutions (23). A?1–28 and A?25–35, as well as control
peptides insulin and apotransferrin, were again found to be
redox inert in the presence of DA (Fig. 3A) (10, 11).
To appraise the affinity of the active Cu2?in catalytic H2O2
production, we studied the effects of chelators of higher Cu2?
affinity than EDTA. DTPA and CDTA have both been observed
previously to deplete Cu2?from A?1–42 (8) and were found to
abolish H2O2production in the presence of DA (Fig. 3B). Sim-
ilarly, DETC, which inhibits Cu-mediated superoxide dis-
mutase activity, inhibited H2O2production. In contrast, mela-
tonin, which has little affinity for Cu2?, was unable to inhibit
H2O2production (data not shown).
The effects of stereochemical occlusion of the A?Cu2active
site were studied using anti-A? monoclonal antibodies. A?Cu2
was preincubated (30 min) with various antibodies, and H2O2
generation upon incubation with DA (5 ?M) was determined.
All three anti-A? antibodies significantly inhibited H2O2pro-
duction (10H3 ? 4G8 ? 6E10), whereas a nonspecific anti-IgG
antibody present at 5-fold greater excess than the anti-A?
antibodies had no inhibitory effect (Fig. 3C).
Cytotoxic H2O2production by A?, which is augmented by Cu
in cultures (11, 18), could potentially be driven by reducing
agents in the culture medium. To test this, we studied the
effects of A?1–42, Cu2?, and DA together and individually,
upon the survival of E18 primary neurons studied in media
that lacked H2O2scavengers. To achieve this, the cells were
first grown in Neurobasal medium plus B-27 supplement with
antioxidants, and after 6–8 days the medium was changed to
Locke’s buffer without antioxidants for the experiments. We
treated the cultures with A?1–42 (2.5 ?M) ? Cu2?-Gly (5 ?M),
in the presence or absence of DA (20 ?M), and assayed for
survival at 14 and 16 h (Fig. 4A). Because DA rapidly oxidizes
in culture medium, it was necessary to abbreviate the onset of
toxicity, which was achieved by using higher concentrations of
reagents than were employed in the H2O2assay studies.
A?1–42 by itself was not toxic (Fig. 4A), but when bound to Cu
A?1–42 became rapidly toxic (85% survival at 14 h, 60% sur-
vival at 16 h; intervals chosen after pilot studies showed that
cell death commenced after 12 h and was near total by 24 h).
The time dependence of toxicity in the absence of DA was
compatible with the catalytic H2O2production by A?Cu2uti-
lizing reducing agents released by the cells and accumulating
in the culture medium. Compatible with H2O2production me-
diating cell death, A?Cu2toxicity was markedly exaggerated
by the addition of exogenous DA. Neither Cu alone nor DA
FIG. 2. Catalytic H2O2production by A?Cu2using dopamine as
substrate. A, A?Cu2(200 nM) was incubated with increasing DA con-
centrations for 120 min at 37 °C in PBS, pH 7.4, 10 ?M EDTA, and H2O2
levels were measured at 30-min intervals. Data are means ? S.E. B,
Lineweaver-Burk transformation of the data obtained at 60 min from
the kinetic experiment described in A.
Catalysis of Hydrogen Peroxide Production by A?
by guest on December 26, 2015
alone was toxic (Fig. 4A). To ensure that the neurotoxicity was
not because of unbound Cu2?, EDTA (10 ?M) was present in the
We investigated the H2O2-related toxicity of A?1–42 at
lower peptide concentration (200 nM) to test whether toxicity
will occur at physiological peptide concentrations (20) and
when the cells are under less acute stress. Because toxicity
evolved more slowly at this lower concentration, a longer incu-
bation was needed. Therefore, to avoid toxicity due to serum
deprivation, Neurobasal medium (lacking B27) was employed
instead of Locke’s buffer. Under these conditions, we observed
that 200 nM A?Cu2was markedly toxic in the presence of 5 ?M
DA (?45% survival after 24 h; see Fig. 4B). The toxicity of
A?Cu2was completely rescued (100% viability) by the selective
H2O2scavenging enzyme, catalase, but only partially rescued
by the radical scavenger Me2SO (Fig. 4B). Hence, the toxicity of
A? could be attributed entirely to catalytic H2O2production.
To ascertain whether these in vitro reactions could occur in
FIG. 5. Catalytic H2O2generation by A? metalloprotein puri-
fied from Alzheimer’s disease amyloid. Synthetic A? (200 nM) with
no added metals (apo-A?), A?Cu2(200 nM), 200 nM A?Cu2? 1600 nM
Zn2?(A?Cu2Zn8), and A? purified from AD-affected brain tissue (A?AD;
200 nM, which is metallated with 1.7 eq Zn, 0.4 eq Cu, 0.2 eq Al, 0.1 eq
Fe) were incubated with DA (5 ?M) ? EDTA (10 ?M) in PBS, pH 7.4, and
H2O2levels were assayed after 1 h.
FIG. 3. Characterization of A?Cu2-mediated catalytic H2O2
generation by activity mapping and inhibition. A, synthetic A?
fragments and variants or control peptides (apotransferrin and insulin)
were incubated (200 nM) with Cu2?-Gly (400 nM) in the presence of DA
(5 ?M) for 60 min at 37 °C in PBS, pH 7.4, and H2O2levels were
measured. B, inhibition by copper chelators. A?Cu2(200 nM) was incu-
bated in PBS, pH 7.4, containing DA (5 ?M), EDTA (10 ?M) ? melatonin
or chelators DETC, CDTA, or DTPA (10 ?M) for 60 min at 37 °C, and
H2O2levels were measured. C, inhibition by anti-A? antibodies. A?Cu2
(100 nM) was preincubated for 30 min in PBS ? anti-A? antibodies (100
nM) or excess purified mouse IgG (500 nM). The mixtures were then
incubated with DA (5 ?M) for 60 min at 37 °C in PBS, pH 7.4, 10 ?M
EDTA, and H2O2levels were measured. Data are means ? S.E. t test
results of effects of antibody treatment are indicated by * (p ? 0.05), **
(p ? 0.01).
FIG. 4. Catalytic H2O2production by A?Cu2mediates toxicity
in neuronal culture. A, dopamine potentiation of A?Cu2neurotoxic-
ity. Primary cortical neurons were incubated in Locke’s buffer ? EDTA
(10 ?M) for 14 or 16 h. The effects of A?1–42 (2.5 ?M), Cu2?-Gly (5 ?M),
and DA (20 ?M), alone and in combination, upon cell viability compared
with untreated cells were determined. B, catalase rescue of A?Cu2-
dopamine neurotoxicity. Primary cortical neurons in Neurobasal me-
dium ? EDTA (10 ?M) were incubated with A?Cu2(200 nM) ? DA (5
?M) ? catalase (Cat; 1000 units/ml) or Me2SO (98 ?M) for 24 h, and cell
viability was then determined. Data are means ? S.E.
Catalysis of Hydrogen Peroxide Production by A?
by guest on December 26, 2015
AD, we measured catalytic H2O2production from A? purified
from AD plaque (24). This A? preparation is neurotoxic in
neuronal primary cell culture (24). We found that the purified
A? generated a significant amount of H2O2(?270 nM; see Fig.
5). Inductively coupled plasma mass spectrometry found that
the only metals it contained were Zn (1.7 eq), Cu (0.4 eq), Al
(0.2 eq), and Fe (0.1 eq). By comparison, synthetic A?Cu2
generated ?630 nM H2O2, whereas A?Cu2incubated with a
4-fold excess of Zn2?generated a similar amount of H2O2as the
human-derived A? (which also bound an ?4-fold excess of Zn2?
to Cu2?). Zn2?binding inhibits A?-mediated H2O2production
(18). However, the A? within amyloid is not bound to sufficient
Zn2?to prevent Cu2?-mediated H2O2production.
The major finding in this report is that A? forms an H2O2-
producing cuproenzyme-like activity utilizing O2and biological
reducing agents as substrates. The Kmfor DA as the substrate
(8.9 ?M) indicates that this enzyme-like reaction is likely to
occur under physiological conditions. Because H2O2could not
be generated by free Cu in this assay system (which was com-
plexed by EDTA), A? must therefore generate a specific struc-
ture that presents redox-active Cu to O2and certain reducing
substrates. This structure could be similar to the Cu binding
site on CuZn-superoxide dismutase (9), which also catalyzes
the reduction of O2(25).
A proposed reaction pathway for the sequence of electron
transfers that converts O2into H2O2is described in Fig. 6. In
this model, an electron is initially transferred from A? to re-
duce Cu2?to Cu?. It is also possible that two electrons reduce
two Cu2?atoms bound to a single molecule of A? (11). After
transferring the electron(s) to Cu2?, the A? peptide forms a
positively charged radical (A???). The oxidized site of A? that
supports this electron transfer is uncertain, but the sole Met
residue at position 35 is a candidate for this reaction. This
residue is essential for Cu2?reduction, which is abolished
when the peptide buries this residue by membrane insertion
(9), and is also abolished in A?1–28, which lacks Met (11).
Met-35 is also essential for the toxicity of A? (26). The strong
reducing potential of the A??Cu2?complex (?550 mV) (11)
appears to impel the reduction of O2to O2
as a double electron transfer from 2A???Cu?or from A???Cu?
although the formation of Cu3?has also been considered pos-
sible (11). The dependence of this reaction step upon the con-
centration of O2has been demonstrated previously (11). The
formation of a O2.intermediate, and subsequent disproportion-
ation to H2O2, was not observed currently or previously (10,
11). After electron donation to O2, the radicalized A???Cu2?
complex might be restored to A??Cu2?by electron transfer from
2?. This step occurs
biological reducing agents (e.g. vitamin C and DA; see Fig. 1B).
It is possible that the oxidized products of this reaction step
might themselves be neurotoxic (e.g. dopaminochrome).
Our findings indicate that DCF is more suitable than TCEP
as a detection reagent for H2O2production by A??Cu, because,
unlike TCEP, DCF does not act as an electron donor in this
system (Fig. 1A). Our current findings therefore provide a more
complete explanation for the reaction sequence in our original
report (10); H2O2was generated by A??Cu using TCEP as a
substrate for O2reduction (the formation reaction), whereas a
proportion of the TCEP (which was abundant in the incuba-
tion) reacted simultaneously with H2O2(in the detection
Our data also indicate that several biological reducing
agents could potentially drive the generation of H2O2by A? in
the neocortex in extracellular and intracellular compartments.
Cholesterol was found to be a significant substrate for H2O2
production (Fig. 1B), suggesting a mechanism for how statin
use might reduce the risk for AD (27). A? penetrates lipid
membranes when bound to Cu2?or Zn2?(9) where it could
potentially convert membrane cholesterol into H2O2. Scaveng-
ing of cholesterol by A??Cu activity may also contribute to the
membrane thinning and pathology seen in AD cortical tissue
(28). Similarly, the availability of progesterone as a substrate
for H2O2generation by A? (Fig. 1B) is intriguing because of the
increased incidence of AD in women. Vitamin C, L-DOPA, and
DA were excellent substrates for A??Cu-mediated production of
H2O2(Fig. 1B). A? could readily encounter vitamin C in intra-
cellular and extracellular compartments in vivo. Although A?
accumulation is principally an extracellular feature of AD neu-
ropathology, intraneuronal A? accumulation has been pro-
posed as an early pathophysiological event in AD (29) where it
may contribute to A? accumulation in neurofibrillary tangle-
bearing neurons (2). Intracellular A? could potentially recruit
catecholamine precursors, L-DOPA and DA, to produce cyto-
toxic H2O2within cortical noradrenergic neurons. This may
contribute to the loss of noradrenergic input to the hippocam-
pus in AD, where noradrenergic axonal abnormalities are as-
sociated with ?-amyloid deposits (30). Several other untested
biological reducing agents could also be substrates for A?-
mediated H2O2production, and they await further investiga-
tion. A systematic survey of concentrations of the reducing
agents in vulnerable brain regions in AD and control tissue also
awaits investigation. Cholesterol, however, has already been
found in A? plaques (31).
A?1–42 has such high affinity for Cu2?(attomolar) (8) that
it is likely to be metallated in vivo, and indeed we found that A?
purified from AD plaque is mainly bound to Zn (1.7 eq per mol
of A?) and Cu (0.4 eq). The excess of Zn binding to the human
A? probably reflects the plaque environment where Zn levels
are markedly elevated (15, 16, 32). The binding of Zn and Cu to
human A? is in agreement with our previous finding that there
are selective sites on A? (8) that favor binding Zn and Cu over
other metals (7).
Our findings are the first to link the toxicity of A? solely to
the generation of a neurotoxin (H2O2) catalytically generated
by the peptide itself, by a mechanism that is driven by the
consumption of normal metabolites (e.g. dopamine and vitamin
C). This would appear to be a simple explanation for the tox-
icity of A? and provides a novel alternative target for a drug
that would neutralize only toxic (rogue enzyme) forms of what
is otherwise a normal protein. We found that A? can be neu-
rotoxic at much lower concentrations (e.g. 200 nM; see Fig. 4B)
than those usually reported (usually micromolar). Our data
would be germane to the neurochemical environment of the
brain in AD, where the toxicity of A? has been linked to soluble
FIG. 6. Model for catalytic H2O2production by A?Cu2. A? re-
duces Cu2?to Cu?by transferring an electron from the peptide back-
bone, generating a peptide radical. The Cu?donates two electrons to
O2, generating O2
radical is reconstituted by reaction with a reducing agent. Zn2?inhibits
the generation of H2O2possibly by inhibiting Cu reduction.
2?, which becomes protonated to form H2O2. The A?
Catalysis of Hydrogen Peroxide Production by A?
by guest on December 26, 2015
forms of the protein that are present at concentrations of about
200 nM (20, 33). A? toxicity was completely rescued by catalase
(Fig. 4B), which does not penetrate the cells. Therefore, the
origin of the neurotoxic H2O2is probably extracellular A?Cu2
rather than from a downstream intracellular event such as
mitochondrial dysfunction. H2O2is freely permeable across
tissue boundaries, so the H2O2formed by A? could be the
primary source of the abundant intracellular and lipid peroxi-
dative damage generated in AD-affected brain (12, 13, 34). A
secondary neurochemical stress that could be caused by A??Cu
enzyme activity could be the consumption and depletion of
normal metabolites such as vitamin C, catecholamines, choles-
terol, and progesterone.
A caveat with our experimental paradigm, as with most
reports of A? toxicity in cell culture, is that toxicity is acute
(?24 h), whereas neuronal demise in actual AD is a far slower
process. In actual AD brain tissue, up-regulated H2O2scaveng-
ing enzymes (35) may help protect the brain from acute dam-
age. The induction of these defenses may be gradual, because
the buildup of A? takes many years. However, in our experi-
ments, the increase of H2O2caused by A? may be too rapid for
cells in culture to induce a sufficient up-regulation of their
antioxidant systems. The cell death that we observed was par-
ticularly abrupt because of the removal of reactive oxygen
species scavengers and glutathione precursors from the culture
medium. Some level of H2O2can be cleared by neuronal cata-
lase, but beyond the threshold of what can be cleared by cata-
lase, glutathione will be consumed to clear H2O2. Energy is
expended to reconstitute reduced glutathione. To some extent
the cell can also provide cysteine, required for glutathione
synthesis, by breakdown of polypeptides like metallothionein.
However, in the presence of continuous catalytic H2O2produc-
tion by A?, the nutriture of the cell may be insufficient to
maintain glutathione levels and keep up with clearing the
H2O2. H2O2then permeates into all tissue compartments caus-
ing oxidative toxicity and inducing apoptosis. This model would
explain why the threshold for cellular resistance to A? toxicity
in cell culture has been reported to be proportional to the
activities of catalase and glutathione peroxidase (36). A further
caveat with our cell toxicity data is that we performed studies
using embryonal neuronal cultures, whereas the AD process
affects adult neurons. Adult neurons might be even more vul-
nerable to H2O2-mediated toxicity, because there is evidence
that as neurons differentiate they lose glutathione peroxidase
and catalase activity and become markedly more vulnerable to
oxidative stress in cell culture (37). Finally, intact cortical
tissue may have a qualitatively different resistance compared
with cells in culture to A?-mediated H2O2production. Never-
theless, injection of synthetic A? into brain parenchyma in
mutant tau transgenic mice induces AD-like tau cytopathology
(38). It is possible that catalytic H2O2production may mediate
that result following recruitment of parenchymal Cu and re-
ducing agents by the peptide. The role of H2O2in the toxicity of
A? might be further tested in transgenic models of AD, where
genetic or pharmacological ablation of peroxide scavengers (e.g.
catalase and glutathione peroxidase) would be expected to ex-
aggerate the neuropathology associated with A? accumulation.
Critical neurochemical factors in AD that could contribute to
the abnormal generation of H2O2by A? would include the
increases in cortical Cu (15) and reducing agents (39). Elevated
brain-reducing equivalents in AD are thought to represent a
compensation for increased oxidative stress, which leads to
up-regulation of glucose-6-phosphate dehydrogenase (39, 40).
Our current data suggest that this compensatory increase in
reducing equivalents could paradoxically promote H2O2gener-
ation by A??Cu, leading to a vicious biochemical cycle.
The inhibition of H2O2production from A?Cu2by anti-A?
antibodies (Fig. 3C) is also of interest, because vaccination with
A?1–42, or passive vaccination with anti-A? antibodies, is
being studied currently with a view to clinical utility, following
the success of the approach in preventing amyloid accumula-
tion in transgenic mice (41). The mechanisms by which these
A?-centered immune reactions achieve this effect are still un-
clear. Our data indicate that antibody binding to A? could
inhibit abnormal H2O2production by the peptide. This would
make the antibody act, in a similar manner to a Cu2?chelator
(Fig. 3B), by blocking the active site on A? from transferring
electrons to O2.
The role of Zn2?in the pathophysiology of A? is pleiotropic
(42). Cobinding of Zn2?inhibited catalytic H2O2production by
A?Cu, but not completely, despite the presence of Zn2?in 4-fold
excess to Cu2?(Fig. 5). 10-Fold excess of Zn2?was observed
previously to abolish H2O2formation completely by A? (18).
Hence the ability of Zn2?to quench the cuproenzyme activity of
A? appears to be relatively inefficient and may explain why
coincubation of synthetic A? with Zn2?rescues the neurotox-
icity of A??Cu incompletely (18). Similarly, the A? purified from
AD brain was found to bind Zn in molar excess to Cu but in
insufficient excess to abolish the catalytic generation of H2O2
(Fig. 5). Previous studies have shown that similar preparations
of purified AD A? are markedly neurotoxic in cell culture (43).
Zn2?, released during neurotransmission from cortical gluta-
matergic synapses, precipitates A? to form amyloid plaques
(16, 44, 45). Because Zn2?suppresses H2O2production by A?
(18) (Fig. 5), we proposed that plaque might be less damaging
than soluble A? or diffuse deposits of A?. However, our current
data indicate that the A? from plaque is insufficiently loaded
with Zn2?to completely abolish catalytic activity (Fig. 5), and
therefore, Zn2?-induced plaque formation might not be a
wholly effective means of preventing H2O2production by A?.
Indeed, despite an inverse correlation between oxidative dam-
age and plaque burden in AD, oxidative adducts are still ab-
normally elevated even in the cases where plaque burden is
heaviest (18). Precipitation of A? by Zn2?may also inhibit
clearance and catabolism of A? (46). Therefore, interdiction of
both the Zn2?and the Cu2?interactions with A? may be
beneficial in a therapeutic compound, such as clioquinol, which
is markedly effective at inhibiting A? accumulation in vivo (17)
and has shown efficacy in a recent phase 2 clinical trial of AD
(47). Future studies will determine whether inhibition of cata-
lytic H2O2production by A? in vitro fulfills its potential to
become a readout system that predicts efficacy of small mole-
cules like clioquinol in clinical trials.
Acknowledgments—We thank Dr. C. Glabe, University of California,
Irvine for providing synthetic peptide and Gulcan Kocak of the Univer-
sity of Melbourne for assistance with cell cultures.
1. Vigo-Pelfrey, C., Lee, D., Keim, P., Lieberburg, I., and Schenk, D. B. (1993)
J. Neurochem. 61, 1965–1968
2. Masters, C. L., Multhaup, G., Simms, G., Pottgiesser, J., Martins, R. N., and
Beyreuther, K. (1985) EMBO J. 4, 2757–2763
3. Suzuki, N., Cheung, T. T., Cai, X. D., Odaka, A., Otvos, L., Jr., Eckman, C.,
Golde, T. E., and Younkin, S. G. (1994) Science 264, 1336–1340
4. Yankner, B. A., Duffy, L. K., and Kirschner, D. A. (1990) Science 250, 279–282
5. Behl, C., Davis, J. B., Lesley, R., and Schubert, D. (1994) Cell 77, 817–827
6. Harris, M. E., Hensley, K., Butterfield, D. A., Leedle, R. A., and Carney, J. M.
(1995) Exp. Neurol. 131, 193–202
7. Atwood, C. S., Moir, R. D., Huang, X., Bacarra, N. M. E., Scarpa, R. C.,
Romano, D. M., Hartshorn, M. A., Tanzi, R. E., and Bush, A. I. (1998)
J. Biol. Chem. 273, 12817–12826
8. Atwood, C. S., Scarpa, R. C., Huang, X., Moir, R. D., Jones, W. D., Fairlie, D. P.,
Tanzi, R. E., and Bush, A. I. (2000) J. Neurochem. 75, 1219–1233
9. Curtain, C., Ali, F., Volitakis, I., Cherny, R., Norton, R., Beyreuther, K.,
Barrow, C., Masters, C., Bush, A., and Barnham, K. (2001) J. Biol. Chem.
10. Huang, X., Atwood, C. S., Hartshorn, M. A., Multhaup, G., Goldstein, L. E.,
Scarpa, R. C., Cuajungco, M. P., Gray, D. N., Lim, J., Moir, R. D., Tanzi,
R. E., and Bush, A. I. (1999) Biochemistry 38, 7609–7616
Catalysis of Hydrogen Peroxide Production by A?
by guest on December 26, 2015
11. Huang, X., Cuajungco, M. P., Atwood, C. S., Hartshorn, M. A., Tyndall, J.,
Hanson, G. R., Stokes, K. C., Leopold, M., Multhaup, G., Goldstein, L. E.,
Scarpa, R. C., Saunders, A. J., Lim, J., Moir, R. D., Glabe, C., Bowden, E. F.,
Masters, C. L., Fairlie, D. P., Tanzi, R. E., and Bush, A. I. (1999) J. Biol.
Chem. 274, 37111–37116
12. Smith, C. D., Carney, J. M., Starke-Reed, P. E., Oliver, C. N., Stadtman, E. R.,
Floyd, R. A., and Markesbery, W. R. (1991) Proc. Natl. Acad. Sci. U. S. A.
13. Nunomura, A., Perry, G., Pappolla, M. A., Wade, R., Hirai, K., Chiba, S., and
Smith, M. A. (1999) J. Neurosci. 19, 1959–1964
14. Smith, M. A., Hirai, K., Hsiao, K., Pappolla, M. A., Harris, P., Siedlak, S.,
Tabaton, M., and Perry, G. (1998) J. Neurochem. 70, 2212–2215
15. Lovell, M. A., Robertson, J. D., Teesdale, W. J., Campbell, J. L., and
Markesbery, W. R. (1998) J. Neurol. Sci. 158, 47–52
16. Suh, S. W., Jensen, K. B., Jensen, M. S., Silva, D. S., Kesslak, J. P., Danscher,
G., and Frederickson, C. J. (2000) Brain Res. 852, 274–278
17. Cherny, R. A., Atwood, C. S., Xilinas, M. E., Gray, D. N., Jones, W. D., McLean,
C. A., Barnham, K. J., Volitakis, I., Fraser, F. W., Kim, Y.-S., Huang, X.,
Goldstein, L. E., Moir, R. D., Lim, J. T., Beyreuther, K., Zheng, H., Tanzi,
R. E., Masters, C. L., and Bush, A. I. (2001) Neuron 30, 665–676
18. Cuajungco, M. P., Goldstein, L. E., Nunomura, A., Smith, M. A., Lim, J. T.,
Atwood, C. S., Huang, X., Farrag, Y. W., Perry, G., and Bush, A. I. (2000)
J. Biol. Chem. 275, 19439–19442
19. Huang, X., Atwood, C. S., Moir, R. D., Hartshorn, M. A., Vonsattel, J.-P., Tanzi,
R. E., and Bush, A. I. (1997) J. Biol. Chem. 272, 26464–26470
20. McLean, C., Cherny, R., Fraser, F., Fuller, S., Smith, M., Beyreuther, K., Bush,
A., and Masters, C. (1999) Annal. Neurol. 46, 860–866
21. White, A. R., Huang, X., Jobling, M. F., Barrow, C. J., Beyreuther, K., Masters,
C. L., Bush, A. I., and Cappai, R. (2001) J. Neurochem. 76, 1509–1520
22. Vaughan, D. W., and Peters, A. (1981) J. Neuropathol. Exp. Neurol. 40,
23. Shivers, B. D., Hilbich, C., Multhaup, G., Salbaum, M., Beyreuther, K., and
Seeburg, P. H. (1988) EMBO J. 7, 1365–1370
24. Roher, A. E., Chaney, M. O., Kuo, Y. M., Webster, S. D., Stine, W. B.,
Haverkamp, L. J., Woods, A. S., Cotter, R. J., Tuohy, J. M., Krafft, G. A.,
Bonnell, B. S., and Emmerling, M. R. (1996) J. Biol. Chem. 271,
25. Estevez, A. G., Crow, J. P., Sampson, J. B., Reiter, C., Zhuang, Y., Richardson,
G. J., Tarpey, M. M., Barbeito, L., and Beckman, J. S. (1999) Science 286,
26. Varadarajan, S., Yatin, S., Kanski, J., Jahanshahi, F., and Butterfield, D. A.
(1999) Brain Res. Bull. 50, 133–141
27. Wolozin, B., Kellman, W., Ruosseau, P., Celesia, G. G., and Siegel, G. (2000)
Arch. Neurol. 57, 1439–1443
28. Mason, R. P., Shoemaker, W. J., Shajenko, L., Chambers, T. E., and Herbette,
L. G. (1992) Neurobiol. Aging 13, 413–419
29. Gouras, G. K., Tsai, J., Naslund, J., Vincent, B., Edgar, M., Checler, F.,
Greenfield, J. P., Haroutunian, V., Buxbaum, J. D., Xu, H., Greengard, P.,
and Relkin, N. R. (2000) Am. J. Pathol. 156, 15–20
30. Powers, R. E., Struble, R. G., Casanova, M. F., O’Connor, D. T., Kitt, C. A., and
Price, D. L. (1988) Neuroscience 25, 401–417
31. Mori, T., Paris, D., Town, T., Rojiani, A. M., Sparks, D. L., Delledonne, A.,
Crawford, F., Abdullah, L. I., Humphrey, J. A., Dickson, D. W., and Mullan,
M. J. (2001) J. Neuropathol. Exp. Neurol. 60, 778–785
32. Lee, J.-Y., Mook-Jung, I., and Koh, J.-Y. (1999) J. Neurosci. 19, 1–5
33. Cherny, R. A., Legg, J. T., McLean, C. A., Fairlie, D., Huang, X., Atwood, C. S.,
Beyreuther, K., Tanzi, R. E., Masters, C. L., and Bush, A. I. (1999) J. Biol.
Chem. 274, 23223–23228
34. Pratico, D., Clark, C. M., Lee, V. M., Trojanowski, J. Q., Rokach, J., and
FitzGerald, G. A. (2000) Ann. Neurol. 48, 809–812
35. Aksenov, M. Y., Tucker, H. M., Nair, P., Aksenova, M. V., Butterfield, D. A.,
Estus, S., and Markesbery, W. R. (1998) J. Mol. Neurosci. 11, 151–164
36. Sagara, Y., Dargusch, R., Klier, F. G., Schubert, D., and Behl, C. (1996)
J. Neurosci. 16, 497–505
37. Tamagno, E., Aragno, M., Parola, M., Parola, S., Brignardello, E., Boccuzzi, G.,
and Danni, O. (2000) Neuroreport 11, 1865–1869
38. Gotz, J., Chen, F., van Dorpe, J., and Nitsch, R. M. (2001) Science 293,
39. Russell, R. L., Siedlak, S. L., Raina, A. K., Bautista, J. M., Smith, M. A., and
Perry, G. (1999) Arch. Biochem. Biophys. 370, 236–239
40. Martins, R. N., Harper, C. G., Stokes, G. B., and Masters, C. L. (1986)
J. Neurochem. 46, 1042–1045
41. Schenk, D., Barbour, R., Dunn, W., Gordon, G., Grajeda, H., Guido, T., Hu, K.,
Huang, J., Johnson-Wood, K., Khan, K., Kholodenko, D., Lee, M., Liao, Z.,
Lieberburg, I., Motter, R., Mutter, L., Soriano, F., Shopp, G., Vasquez, N.,
Vandevert, C., Walker, S., Wogulis, M., Yednock, T., Games, D., and
Seubert, P. (1999) Nature 400, 173–177
42. Bush, A. I., and Tanzi, R. E. (2002) Proc. Natl. Acad. Sci. U. S. A. 99,
43. Kuo, Y. M., Emmerling, M. R., Vigo-Pelfrey, C., Kasunic, T. C., Kirkpatrick,
J. B., Murdoch, G. H., Ball, M. J., and Roher, A. E. (1996) J. Biol. Chem.
44. Bush, A. I., Pettingell, W. H., Multhaup, G., Paradis, M. D., Vonsattel, J. P.,
Gusella, J. F., Beyreuther, K., Masters, C. L., and Tanzi, R. E. (1994)
Science 265, 1464–1467
45. Lee, J.-Y., Cole, T. B., Palmiter, R. D., Suh, S. W., and Koh, J.-Y. (2002) Proc.
Natl. Acad. Sci. U. S. A. 99, 7705–7710
46. Bush, A. I., Pettingell W. H., Jr., Paradis, M. D., and Tanzi, R. E. (1994) J. Biol.
Chem. 269, 12152–12158
47. Masters, C. L. (2002) 7th International Geneva/Springfield Alzheimer’s Sym-
posium, Geneva, April 3, 2002
Catalysis of Hydrogen Peroxide Production by A?
by guest on December 26, 2015
Inestrosa and Ashley I. Bush Download full-text
2002, 277:40302-40308. J. Biol. Chem.
Masters, Rudolph E. Tanzi, Nibaldo C.
Anthony R. White, Roberto Cappai, Colin L.
Cherny, Robert D. Moir, Alex E. Roher,
Carlos Opazo, Xudong Huang, Robert A.
BIOLOGICAL REDUCING AGENTS TO
DOPAMINE, CHOLESTEROL, AND
CATALYTIC CONVERSION OF
Metalloenzyme-like Activity of Alzheimer's
ENZYME CATALYSIS AND
doi: 10.1074/jbc.M206428200 originally published online August 20, 2002
Find articles, minireviews, Reflections and Classics on similar topics on the
10.1074/jbc.M206428200Access the most updated version of this article at doi:
. JBC Affinity Sites
When a correction for this article is posted•
When this article is cited •
to choose from all of JBC's e-mail alertsClick here
This article cites 46 references, 20 of which can be accessed free at
by guest on December 26, 2015