Free Radical Biology & Medicine 40 (2006) 850  862
www.elsevier.com/locate/freeradbiomed
Original Contribution
Aging sensitizes toward ROS formation and lipid peroxidation
in PS1M146L transgenic mice
a, a a a a
*
Katrin Schuessel , Claudia Frey , Claudia Jourdan , Uta Keil , Claudia C. Weber ,
b a a,b
Franz Müller-Spahn , Walter E. Müller , Anne Eckert
a
Department of Pharmacology, Biocentre, University of Frankfurt, 60439 Frankfurt am Main, Germany
b
Neurobiology Research Laboratory, Psychiatry University Hospital, CH-4025 Basel, Switzerland
Received 24 June 2005; revised 13 September 2005; accepted 10 October 2005
Available online 15 November 2005
Abstract
Mutations in the presenilins (PS) account for the majority of familial Alzheimer disease (FAD) cases. To test the hypothesis that oxidative
stress can underlie the deleterious effects of presenilin mutations, we analyzed lipid peroxidation products (4-hydroxynonenal (HNE) and
malondialdehyde) and antioxidant defenses in brain tissue and levels of reactive oxygen species (ROS) in splenic lymphocytes from transgenic
mice bearing human PS1 with the M146L mutation (PS1M146L) compared to those from mice transgenic for wild-type human PS1 (PS1wt) and
nontransgenic littermate control mice. In brain tissue, HNE levels were increased only in aged (19 22 months) PS1M146L transgenic animals
compared to PS1wt mice and not in young (3 4 months) or middle-aged mice (13 15 months). Similarly, in splenic lymphocytes expressing the
transgenic PS1 proteins, mitochondrial and cytosolic ROS levels were elevated to 142.1 and 120.5% relative to controls only in cells from aged
PS1M146L animals. Additionally, brain tissue HNE levels were positively correlated with mitochondrial ROS levels in splenic lymphocytes,
indicating that oxidative stress can be detected in different tissues of PS1 transgenic mice. Antioxidant defenses (activities of antioxidant enzymes
Cu/Zn-SOD, GPx, or GR) or susceptibility to in vitro oxidative stimulation was unaltered. In summary, these results demonstrate that the
PS1M146L mutation increases mitochondrial ROS formation and oxidative damage in aged mice. Hence, oxidative stress caused by the combined
effects of aging and PS1 mutations may be causative for triggering neurodegenerative events in FAD patients.
© 2005 Elsevier Inc. All rights reserved.
Keywords: Aging; Alzheimer disease; Amyloid h; Brain; Antioxidant enzyme; Lipid peroxidation; Lymphocyte; Mitochondria; Hydroxynonenal; Oxidative stress;
Presenilin; Reactive oxygen species; Transgenic mouse; Transgenic; Free radical
Alzheimer_s disease (AD) is the most frequent form of or PS2, which have provided valuable insight into putative
dementia and among the leading causes of death in the elderly. pathogenic mechanisms. These mutations cause inheritable
Although the precise mechanisms by which neurodegeneration familial forms of AD (FAD) showing an early onset of
in AD patients is triggered remain largely speculative, the past cognitive symptoms in contrast to the much more common
decades of research have identified rare genetic mutations in sporadic form of AD with onset of symptoms usually over the
the amyloid precursor protein (APP) or in the presenilins PS1 age of 60, for which aging represents the main risk factor [1].
Mutations in the presenilins PS1 and PS2 account for the
majority of FAD cases, and clinical onset in some carriers of
Abbreviations: AD, Alzheimer disease; Ah, amyloid h; APP, amyloid
PS1 mutations is extremely early, occasionally as soon as in the
precursor protein; DAF-2-DA, diaminofluorescein-2-diacetate; DHE, dihy-
third decade of life [2,3]. In recent years, the number of
droethidium; DHR, dihydrorhodamine 123; FAD, familial Alzheimer disease;
GPx, glutathione peroxidase; GR, glutathione disulfide reductase; HBSS, mutations identified in the PS genes has been ever increasing,
Hanks_ balanced salt solution; H2DCF-DA, dichlorodihydrofluorescein diace-
especially in PS1, in which more than a hundred mutations
tate, HNE, 4-hydroxynonenal; MDA, malondialdehyde; Mio, million cells;
have been found so far [4]. Presenilins are a component of the
PS1, presenilin 1; PS2, presenilin 2; R123, rhodamine 123; ROS, reactive
g-secretase complex, which is involved in the formation of
oxygen species; SOD, superoxide dismutase.
amyloid h (Ah) from its precursor protein APP [5]. Several
* Corresponding author. Fax: [phone].
E-mail address: [e-mail] (K. Schuessel). different PS mutations have been shown to alter g-secretase
0891-5849/$ - see front matter © 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.freeradbiomed.2005.10.041
K. Schuessel et al. / Free Radical Biology & Medicine 40 (2006) 850  862 851
processing of APP to yield higher levels of the Ah1 42 isoform was analyzed for lipid peroxidation products malondialdehyde
[6 10] and a higher production of Ah1 42 correlated with an (MDA) and 4-hydroxynonenal (HNE) as markers of oxidative
earlier clinical onset of Alzheimer dementia in carriers of PS1 damage. Oxidative stress results from an imbalance between
mutations [11]. With the exception of one deletion mutant, all the production and the detoxification of reactive oxygen
these mutations are single amino acid mutations, which are species (ROS), and we have previously identified an impair-
spread across the whole length of PS1 protein. This suggests ment of the antioxidant enzyme Cu/Zn-dependent superoxide
that all PS1 mutations trigger the development of FAD via a dismutase as a possible causative factor for oxidative damage
common toxic mechanism caused by only subtle alterations of in mutant APP transgenic mice [31,33]. In order to test whether
the protein structure, consistent with a recent report that a similar deficiency in antioxidant defense may be present in
different PS1 mutations all cause a similar structural change in PS1 transgenic mice, we assayed antioxidant enzymatic
protein conformation and interaction with APP [12]. Similar to activities of Cu/Zn-dependent superoxide dismutase (Cu/Zn-
PS mutations, FAD mutations in APP result in elevated SOD), glutathione peroxidase (GPx), and glutathione reductase
formation of Ah species [13,14]. The high amounts of Ah (GR). The expression of transgenic PS1 under the control of
deposited as amyloid plaques, which are a characteristic feature the HMG-CoA reductase promoter leads to high expression in
of brains from AD patients, suggest a pathological role for this brain tissue, but also in peripheral tissues from PS1 transgenic
peptide and led to the development of the amyloid hypothesis mice [34,35]. Therefore, splenic lymphocytes could be isolated
of AD [15,16]. It states that the increased formation and and employed for a direct study of mutant presenilin effects on
accumulation of Ah, and especially the Ah1 42 isoform, ROS levels in living cells by staining with various ROS-
triggers neurodegenerative events in AD patients. Therefore, sensitive fluorescent dyes and selective analysis by flow
the increased formation of Ah1 42 formation is proposed to be cytometry. In order to additionally assess a putative interaction
the causative mechanism for neurotoxicity induced by pre- between aging and the effects of the PS1 mutation, transgenic
senilin mutations. Consistent with this hypothesis, proapoptotic mice of different age groups were studied, i.e., young, middle-
effects are induced by expression of mutant PS1 in cell culture aged, and aged mice (3 4, 13 15, and 19 22 months of age,
systems [17 19] and transgenic mice [20]. Similarly, in respectively).
primary cultured neurons of transgenic rats expressing wild-
type PS1, increased apoptosis correlated with the amount of Materials and methods
PS1 expression [21]. Moreover, mice transgenic for human
PS1 with FAD mutations display increased formation of rodent Animals
Ah1 42 and neuronal degeneration [22,23] and show cognitive
deficits [24]. PS1 transgenic mice were generated as described before
However, it remains speculative how neurodegeneration [21,36]. Mice were bred on a C57BL/6J background (Iffa
may be mediated by Ah and whether similar events take place Credo, France), and for all experiments, heterozygous trans-
during the development of AD. It has been suggested that genic mice bearing human wild-type presenilin 1 (PS1wt) or
oxidative stress plays an important role, which is mainly based human presenilin 1 with the M146L mutation (PS1M146L) and
on two observations: (1) Aging is the most important risk factor the respective nontransgenic control animals from the same
for the development of sporadic AD and also plays an litter (non-tg) were used. Mice were maintained on a 12-
important role in the familial forms of AD, as even in FAD h dark light cycle with pelleted food and tap water ad libitum.
cases only very few patients show onset of cognitive symptoms Animals were handled according to the French and German
before the age of 30 [2,3]. The phenomenon of aging has guidelines for animal care.
previously been closely linked to the accumulation of oxidative Transgene expression of either PS1wt or PS1M146L is
stress [25], which could also be demonstrated in brains from under the control of a HMG-CoA reductase promoter. This
sporadic and familial AD patients [26 28]. (2) In accordance results in a strong neuronal expression but also ubiquitous
with the amyloid hypothesis, numerous studies of Ah toxicity expression in peripheral tissues [34,35]. PS1 mutations lead to
in cell culture [29] or APP transgenic mice [30,31] have increased amyloidogenic processing of endogenous mouse
provided evidence that Ah can induce neuronal cell death by APP; however, no Ah plaques can be detected in the brains
eliciting oxidative damage. from these mice, which seems to be due to differences in
In order to test the hypothesis that presenilin mutations can aggregation properties of rodent compared to human Ah [37].
similarly provoke the development of FAD by inducing Animals were studied at 3 4 (young), 13 15 (middle-aged),
oxidative stress in vivo, we analyzed several oxidative stress- and 19 22 (aged) months of age. Brain tissue and splenic
related parameters in transgenic mice expressing human PS1 lymphocytes were studied in a total of 11 nontransgenic control,
with a FAD mutation. As controls, nontransgenic littermate 11 PS1wt, and 11 PS1M146L transgenic mice at 3 4 months of
mice as well as mice expressing human wild-type PS1 were age; 27 nontransgenic control, 18 PS1wt, and 13 PS1M146L
employed. The comparison with PS1wt transgenic mice allows transgenic mice at 13 15 months of age; as well as 27
us to detect effects that are caused specifically by the presence nontransgenic control, 10 PS1wt, and 12 PS1M146L transgenic
of the M146L FAD mutation in PS1, as the expression of wild- mice at 19 22 months of age. An additional cohort of 7
type PS1 alone has been reported to exhibit cytotoxic effects in nontransgenic and 7 PS1M146L transgenic mice were employed
some cell culture models [18,32]. Brain tissue from these mice at an age of 21 months for detailed analysis of splenic lym-
852 K. Schuessel et al. / Free Radical Biology & Medicine 40 (2006) 850  862
phocyte subgroups and apoptosis quantification. Gender was available Superoxide Dismutase Assay Kit from Calbiochem.
distributed equally between the groups. The assay of GPx activity (cytosolic GPx, EC 1.11.1.9),
based on the reaction described by Paglia et al. [42] with tert-
Isolation of mouse brain tissue and preparation of splenic butylhydroperoxide as substrate, and the assay of GR activity
lymphocytes (EC 1.8.1.7), based on the method of Mizuno and Ohta [43],
were performed utilizing the commercially available Cellular
Mice were quickly killed by cervical dislocation, and brains Glutathione Peroxidase Assay Kit and Glutathione Reductase
were removed and dissected on ice. Cerebellum and brain stem Assay Kit from Calbiochem, respectively. All assays were
were removed and the forebrain was dissected midsagittally. performed as described in more detail previously [31].
Both hemispheres were immediately snap frozen in liquid
nitrogen and stored at 80-C until homogenization. Splenic Isolation of mRNA from splenic lymphocytes and RT-PCR for
lymphocytes were prepared by mechanical dissociation of detection of human PS1 expression in transgenic mice
individual spleens and erythrocyte lysis as described previously
[38]. For determination of ROS with fluorescent dyes, live cells RNA was isolated from splenic lymphocytes utilizing the
were counted via trypan blue exclusion (Biochrom AG, Berlin, High Pure RNA Isolation Kit from Boehringer (Mannheim,
Germany) in a Neubauer chamber and adjusted to 1 million Germany) as described in the manufacturer_s protocol. Sixty-
cells per ml (1 Mio/ml). four nanograms of RNA was used for reverse transcriptase
PCR, and RT reactions were performed with the SuperScript II
Preparation of mouse brains for determination of lipid Kit from Invitrogen (Karlsruhe, Germany) according to the
peroxidation products and antioxidant enzyme activities supplier_s instructions. cDNA synthesis was performed
from RNA samples with oligo(dT) primers. PCR of PS1
One mouse brain hemisphere was minced in 1 ml of cold 20 DNA was conducted with the Eppendorf Master Taq Kit
mM Tris-buffered saline with 10 strokes in a Potter homog- (Cologne, Germany) utilizing the primers sense, 5V-TAA-
enizer at 1200 rpm. An aliquot of this homogenate was diluted TTGGTCCATAAAAGGC-3V, and antisense, 5V-GCACAGAA-
1:1 with Tris-buffered saline and centrifuged at 3000g and 4-C AGGGAGTCACAAG-3V, which yields a 425-bp product
for 10 min and supernatants were collected and stored at specific for human presenilin 1. PCR products were detected
80-C for lipid peroxidation assays. The remaining homog- after electrophoresis in ethidium-stained agarose gels (VWR
enate was centrifuged at 8500g and 4-C for 10 min and International).
supernatants were collected and stored at 80-C for determi-
nation of antioxidant enzyme activities. Measurement of reactive oxygen species in splenic lymphocytes
with oxidation-sensitive fluorescent dyes and flow cytometry
Assay of lipid peroxidation products analysis
Lipid peroxidation products MDA and HNE were deter- Splenic lymphocytes were incubated at 1 Mio/ml with ROS-
mined by a photometrical method utilizing the lipid peroxida- sensitive dyes at 37-C, washed twice, and resuspended for
tion assay kit from Calbiochem (Schwalbach, Germany), which flow-cytometric analysis with Becton Dickinson (Heidelberg,
is based on the method of Esterbauer and Cheeseman [39]. The Germany) FACSCalibur utilizing CellQuest Pro software. A
colorimetric reaction is a condensation of the respective minimum of 8000 events were recorded per single measure-
aldehyde with 1-methyl-2-phenylindole yielding chromophores ment. Cells were analyzed immediately after staining and
with absorption maxima at 586 nm. Both aldehydes react with always kept on ice in the dark until measurement.
1-methyl-2-phenylindole in the presence of methanesulfonic Dihydrorhodamine 123 (DHR), rhodamine 123 (R123),
acid, whereas substitution of methanesulfonic acid by hydro- dichlorodihydrofluorescein diacetate (H2DCF-DA), and dihy-
chloric acid allows measurement of MDA alone due to an droethidium (DHE) (5 mM stabilized solution in DMSO) were
irreversible cyclization reaction of hydroxyalkenals [40]. HNE purchased from Molecular Probes (Leiden, Netherlands).
levels were calculated from absorbance levels obtained by Diaminofluorescein-2-diacetate (DAF-2-DA) (5 mM solution
subtracting the absorbance of samples after hydrochloric acid in DMSO) was obtained from Calbiochem.
reaction from absorbance after methanesulfonic acid reaction. Basal levels of ROS were assayed immediately after
Basal levels of MDA and HNE were assayed after incubation lymphocyte isolation in RPMI medium containing 5% fetal
of samples at 37-C for 30 min. Stimulated MDA levels were calf serum (Sigma, Taufkirchen, Germany). Serum withdraw-
assayed after incubation in the presence of 50 AM FeCl3 (VWR al was conducted by transferring lymphocytes to HBSS
International, Darmstadt, Germany). (Sigma) buffer containing 10 mM Hepes, 1 mM CaCl2, and
0.5 mM MgSO4 (all VWR International), pH 7.4. Cells were
Antioxidant enzyme activities: Cu/Zn-SOD, GPx, and GR oxidatively stimulated in HBSS buffer by addition of
activity hydrogen peroxide (Sigma) and incubation at 37-C for 15
min. After stimulation, cells were washed, resuspended in
The assay of Cu/Zn-SOD activity (EC 1.15.1.1) is based RPMI medium, and stained with ROS-sensitive dyes as
on the method of Nebot et al. [41] utilizing the commercially described below.
K. Schuessel et al. / Free Radical Biology & Medicine 40 (2006) 850  862 853
DHR is the nonfluorescent reduced form of R123. Inside the FITC-conjugated anti-CD3 antibody and fluorescence detec-
cell, it is oxidized mainly by mitochondrial ROS to the tion in FL-1 or PE-conjugated anti-CD4 antibody and
positively charged fluorescent rhodamine R123 [44]. Because fluorescence detection in FL-2. All antibodies were purchased
the fluorescence signal obtained after staining of lymphocytes from Becton Dickinson.
with DHR depends on two parameters, i.e., (1) oxidation of the
dye and (2) incorporation of R123 into mitochondria, Chemicals
lymphocytes were also stained with the oxidized form of
DHR, R123, to correct for possible mitochondrial defects that NH4Cl, KHCO3, hydrochloric acid, ethanol, and chloro-
would affect uptake of oxidized DHR. Furthermore, R123 form were purchased from VWR International. All aqueous
staining also controls for effects of P-glycoprotein [45], as solutions were prepared with deionized and filtered water
R123 is a substrate for P-glycoprotein, which is expressed in (Millipore).
murine lymphocytes and can affect R123 staining of cells [46].
DHR and R123 were used at a final concentration of 10 and 1 Protein content
AM, respectively, and an incubation time of 15 min. H2DCF-
DA is hydrolyzed and oxidized by various ROS to the Protein levels in brain homogenates were determined
fluorescent DCF [47]. H2DCF-DA does not stain mitochondria according to the method of Lowry et al. [56] using Bio-Rad
[48], but is oxidized mainly by peroxides in the cytosol [49]. Protein Assay solutions (Munich, Germany) and bovine serum
H2DCF-DA was used at a concentration of 10 AM and 30 min albumin (Sigma Aldrich) as standard.
incubation time. Both DHR and H2DCF have been described to
be also oxidized by peroxynitrite [50]. Peroxynitrite is formed Statistics
in a reaction of the superoxide radical anion with nitric oxide
[51]. In order to assess a possible contribution of peroxynitrite Student_s t test was calculated for comparison between two
to DHR and H2DCF oxidation in our studies, we also groups. For simultaneous estimation of the effects of two
monitored superoxide and nitric oxide levels. Superoxide parameters, two-way ANOVA was calculated. For calculation
anions are readily detected by oxidation of DHE [52], which of correlation between two parameters assayed in samples from
was used at a concentration of 5 AM and 30 min of incubation. the same animal, linear regression assuming Gaussian distri-
Nitric oxide levels were measured with DAF-2-DA [53] at a bution was performed. All calculations were done with Prism
concentration of 2.5 AM and 30 min of incubation time. In the GraphPad software and p 0.05 was assessed as significant.
presence of nitric oxide within cells, DAF-2-DA is hydrolyzed All data in bar graphs are represented as means + SEM.
and converted to the fluorescent triazole derivate DAF-2T.
The oxidized forms of all ROS-sensitive dyes as well as Results
R123 could be excited with the 488 nm laser of the Becton
Dickinson FACSCalibur flow cytometer. Emission of R123, Increased oxidative damage in brains from aged PS1M146L
DCF, and DAF-2T fluorescence was detected in channel FL-1 transgenic mice
(530-nm filter, 30-nm bandpass); emission of ethidium
fluorescence after DHE oxidation was quantified in channel Brain tissue from nontransgenic littermate control mice as
FL-2 (585-nm filter, 42-nm bandpass). well as PS1wt and PS1M146L transgenic mice was studied for
For quantification of mitochondrial ROS in lymphocyte the levels of lipid peroxidation products MDA and HNE as
subgroups, cells were double stained with DHR and anti-CD3 , markers of oxidative damage. Whereas no differences in lipid
anti-CD4 , or anti-CD8 PE antibodies (all from Becton peroxidation levels were found in young (3 4 months) or
Dickinson), and double stained cells were gated according to middle-aged (13 15) mice, increased levels of HNE could be
their fluorescence signals in channels FL-1 (R123 signal) and detected in aged PS1M146L mice (19 22 months) (see Fig.
FL-2 (PE signal). 1A). Interestingly, HNE levels were not increased in PS1wt
transgenic mice, indicating that merely the transgenic PS1
Apoptosis measurement in splenic lymphocytes protein does not induce oxidative damage. Rather, the presence
of the M146L mutation is responsible for increased lipid
The apoptotic status of lymphocytes was quantified with the peroxidation. Of note, whereas HNE levels were increased,
DNA dye 7-actino-antimycin (7-AAD; Molecular Probes) as MDA levels were not different between the mouse models
described by [54]. 7-AAD stains cells that have lost membrane (data not shown), suggesting that predominantly HNE is
integrity. For quantification of apoptosis, 1 Mio/ml splenic formed in brains from these mice.
lymphocytes were incubated with 6 AM 7-AAD for 15 min at As impaired antioxidant defenses can account for ROS
room temperature. Afterward, cells were washed, resuspended accumulation and oxidative damage, we analyzed enzymatic
in phosphate-buffered saline (PAA, Cölbe, Germany), and activities of Cu/Zn-SOD, GPx, and GR in brains from these
immediately analyzed by flow cytometry. Apoptotic cells were mice. However, none of these enzyme activities was impaired in
gated and quantified based on the signal_s forward scatter and PS1M146L transgenic mice in any age group analyzed (data not
7-AAD fluorescence detected in FL-3 (650-nm filter) [55]. shown). As an additional indicator of antioxidant capacity, lipid
Lymphocyte subsets were identified by double staining with peroxidation in brain homogenates was stimulated with ferric
854 K. Schuessel et al. / Free Radical Biology & Medicine 40 (2006) 850  862
promoter leads to strong expression in brain tissue but also
in peripheral cells [34,35]. Hence, these cells could be used to
conveniently study ROS levels directly by flow cytometry.
Similar to the situation in brain tissue from PS1 transgenic
mice, only cells from aged transgenic mice bearing the
PS1M146L mutation displayed increased ROS levels as
measured by increased oxidation of DHR (see Fig. 3A). Of
note, DHR oxidation was not increased in lymphocytes from
PS1wt transgenic mice, indicating that this effect is specifically
caused by the presence of the FAD mutation PS1M146L.
In order to specify putative sources and/or types of ROS,
splenic lymphocytes were stained with further ROS-sensitive
Fig. 1. Lipid peroxidation products in brains from aged mice (19  22 months of
age) transgenic for human wild-type presenilin 1 (PS1wt; n = 9  10) or human
dyes. Apart from DHR, which detects mainly mitochondrial
mutant presenilin 1 (PS1M146L; n = 10) and nontransgenic age-matched
ROS production, H2DCF-DA for indication of cytosolic
littermate controls (non-tg, n = 18  20). (A) Levels of HNE (nmol/mg protein)
peroxides and DHE detecting superoxide radicals were
are increased in brains from PS1M146L transgenic mice compared to PS1wt
employed. Additionally, confounding factors like incorporation
mice, *p < 0.05, Student_s t test. (B) Levels of MDA (nmol/mg protein)
of oxidized DHR into mitochondria or oxidation of DHR and
formation after incubation of brain homogenates with 50 AM FeCl3 for 30 min
at 37-C. Differences between groups are not statistically significant.
H2DCF-DA by peroxynitrite [50] were controlled for by staining
with R123, i.e., the oxidized form of DHR, and DAF2-DA, a
iron in vitro. Again, no difference in the formation of MDA could
nitric oxide-sensitive dye, respectively. The results demonstrate
be observed between nontransgenic and PS1 transgenic mice
that mainly mitochondrial ROS production (142.1% DHR
(Fig. 1B). Hence, antioxidant activity is seemingly not impaired
oxidation relative to nontransgenic controls, see Fig. 3A) and
by the presence of mutant PS1, indicating that oxidative damage
in brains from PS1M146L mice stems from overproduction
rather than from impaired detoxification of ROS.
Increased ROS levels in splenic lymphocytes expressing mutant
PS1
In order to analyze ROS production in PS1 transgenic mice
in more detail, ROS levels were measured directly in living
cells from these mice. To this purpose, splenic lymphocytes
were isolated from nontransgenic and PS1wt and PS1M146L
transgenic mice and analyzed for expression of PS1 mRNA by
RT-PCR. The expression of transgenic human PS1 could be
detected in splenic lymphocytes from PS1wt or PS1M146L
transgenic mice (see Fig. 2), which confirms previously
published results showing that the HMG-CoA reductase
Fig. 3. Increased ROS formation in PS1M146L transgenic mice. Splenic
lymphocytes expressing the transgenic protein were isolated and stained with
different ROS-sensitive fluorescent dyes and analyzed by flow cytometry. Data
are presented as mean fluorescence intensity levels (MFI) relative to the means
of nontransgenic control mice (100%). n = 20 nontransgenic (non-tg), n = 10
PS1wt, and n = 11 PS1M146L transgenic mice were studied. (A) Mitochondrial
ROS formation (oxidation of DHR to R123) is increased in cells from
PS1M146L mice (142.1% relative to nontransgenic controls). *p < 0.05,
Fig. 2. Expression of human presenilin mRNA in lymphocytes from PS1 Student_s t test, PS1M146L vs non-tg. (B) Cytosolic ROS formation (oxidation
transgenic mice. Total RNA was isolated from splenic lymphocytes and of H2DCF-DA to DCF) is increased in cells from PS1M146L mice (120.5%
subjected to RT-PCR with oligo(dT) primers, and human presenilin 1 cDNA relative to nontransgenic controls). **p < 0.01, Student_s t test, PS1M146L vs
sequences were amplified by PCR with primers specific for human presenilin 1. non-tg. (C) Nitric oxide levels (oxidation of DAF-2-DA to DAF-2-T) are not
Lane 1, PS1 PCR positive control (DNA isolated from PS1 transgenic mouse); significantly different between groups. (D) Mitochondrial uptake of R123 is
lanes 2, 3, 5, and 8, samples from nontransgenic animals; lane 4, PS1M146L slightly increased in cells from PS1M146L mice (107.2% relative to
transgenic mouse; lanes 6 and 7, PS1wt transgenic mouse. nontransgenic controls). *p < 0.05, Student_s t test, PS1M146L vs non-tg.
K. Schuessel et al. / Free Radical Biology & Medicine 40 (2006) 850  862 855
to a lesser extent cytosolic ROS production (120.5% H2DCF
oxidation relative to nontransgenic controls, see Fig. 3B) are
increased in cells from PS1M146L mice. Of note, oxidation of
DHE (data not shown) and DAF-2-DA-staining (Fig. 3C) were
unchanged, indicating that the oxidation of DHR and H2DCF is
unlikely to be confounded by increased levels of peroxynitrite.
Furthermore, increased staining of cells with DHR seems to be
specific for oxidation of the dye, because staining with R123 as
control for dye retention due to mitochondrial uptake was only
slightly increased in PS1M146L cells (Fig. 3D). Although this
difference was significant, the extent of R123 staining of cells
was much lower (107.2% relative to controls) compared to the
extent of DHR oxidation in cells from PS1M146L mice. Hence,
these results indicate that mainly mitochondrial and secondarily
cytosolic ROS are increased by the PS1M146L mutation,
whereas formation of nitric oxide and superoxide as well as
severe impairment of mitochondrial membrane integrity as
sources of ROS are negligible.
Importantly, the observed alterations in brain tissue HNE
Fig. 4. Susceptibility of PS1-expressing splenic lymphocytes toward secondary
and peripheral ROS levels are specific for the PS1M146L insults. ROS formation was quantified by oxidation of DHR in cells from
nontransgenic controls (non-tg) and PS1M146L transgenic mice (PS1M146L)
mutation, because no differences were observed in any of the
in (A) aged animals (19  22 months of age) and (B and C) middle-aged
tested parameters between nontransgenic mice and PS1wt
animals (13  15 months of age). Data are presented as mean fluorescence
mice. This allows also for a direct comparison of PS1M146L
intensity levels (MFI) relative to the means of age-matched nontransgenic
transgenic mice with nontransgenic littermate control mice in
control mice (100%). (A) Cells from aged mice react similarly upon serum
the following experiments. withdrawal and additional stimulation with rising concentrations of hydrogen
peroxide for 15 min at 37-C. Two-way ANOVA: p < 0.01, effect of
As an additional indicator of antioxidant metabolism, ROS
PS1M146L transgene; p < 0.0001, effect of hydrogen peroxide stimulation;
production in cells from these mice was also studied after
interaction n.s. Serum withdrawal *p < 0.05, Student_s t test, PS1M146L (n =
stimulation with hydrogen peroxide in vitro. Although DHR
7) vs non-tg (n = 7). (B) In cells from middle-aged mice, basal levels of ROS
oxidation was increased in PS1M146L-expressing cells from
(DHR oxidation) are not significantly different between nontransgenic (n =
aged mice (see Fig. 4A), this reflects the basal increased ROS 22) and PS1M146L transgenic (n = 9) mice. (C) After serum withdrawal for
15 min, more ROS are formed in cells from middle-aged PS1M146L
production and not an additional effect of oxidative stimula-
transgenic mice compared to nontransgenic mice. *p < 0.05, Student_s t test
tion, indicating that cells from the different transgenic animals
PS1M146L (n = 8) vs non-tg (n = 21).
react similarly upon ROS accumulation. This observation
supports our findings on brain tissue that oxidative damage (see Fig. 5A). However, differences could be detected between
in PS1M146L transgenic mice is caused by overproduction CD4+ and CD8+ cells in that DHR oxidation was increased in
rather than by impaired detoxification of ROS. CD4+ cells compared to CD8+, which was independent of the
Interestingly, in middle-aged mice (13 15 months of age), transgene. Hence, CD4+ cells accumulate higher ROS levels.
basal ROS levels were unaltered (Fig. 4B), but DHR oxidation Apoptosis levels in the same cells also indicated that CD4+ cells
was increased in cells from PS1M146L mice after serum react more sensitively toward this type of cell death. Although
withdrawal (Fig. 4C). Serum withdrawal represents a mild stress differences failed to reach statistical significance (see Fig. 5B),
condition, indicating that the presence of an additional stimulus these results confirm previous findings from our group [35].
can trigger the ROS-elevating effects of the PS1M146L Moreover, the levels of DHR oxidation correlated with the
mutation. Therefore, the effect of the PS1 mutation in middle- amount of apoptosis in CD4+ cells from the same animal,
aged mice is evident only after a secondary insult, whereas in indicating that ROS accumulation and apoptotic cell death are
aged mice basal levels of ROS are already increased. interrelated (see Fig. 5C).
ROS formation correlates with apoptosis in CD4+ T cells Correlation of mitochondrial ROS levels in lymphocytes with
HNE levels in brain tissue
Previous studies by our group on lymphocytes from AD
patients found that a different vulnerability toward cell death Although peripheral cells from AD patients have been
exists in specific cell types. Accordingly, we analyzed lympho- widely used for studies of putative pathogenic cell death
cyte subpopulations, i.e., the CD3-positive pool (T lymphocytes mechanisms, evidence that relevant findings in these cells can
only) and the CD4+ and CD8+ subpool of T cells, for a putatively be transferred to the situation in human brain tissue could not
different accumulation of ROS. In general, the ROS-elevating be provided so far as noninvasive measures of apoptotic cell
effect of the PS1 mutation was similar in all lymphocyte death or oxidative stress in human brain tissue in vivo are
subgroups studied, indicating that the presence of mutant PS1 unavailable. In order to test whether measurement of ROS
leads to increased oxidative stress independent of the cell type levels in peripheral cells can be used as an indicator of
856 K. Schuessel et al. / Free Radical Biology & Medicine 40 (2006) 850  862
levels as a marker for lipid peroxidation in brain tissue as well
as increased ROS levels in splenic lymphocytes expressing the
mutant human transgenic protein. Interestingly, the effect of the
FAD mutation is dependent on the natural aging process, as
only aged mice display increased ROS levels and oxidative
damage. This is in close analogy to the situation in humans, in
which carriers of these mutations live without cognitive
disabilities until the disease strikes after some decades [57].
Hence, PS1M146L transgenic mice can represent a suitable
and valuable model to study FAD-relevant pathogenic mechan-
isms in vivo.
Oxidative damage in our PS1M146L transgenic mouse
model parallels previous studies on PS1M146V knock-in mice,
in which synaptosomes display increased levels of oxidatively
modified proteins [58] and hippocampal neurons show
increased vulnerability toward Ah-induced cell death mediated
by superoxide radical formation [20]. The selectively increased
HNE but not MDA levels in brains from PS1M146L transgenic
mice are furthermore in close analogy to previously published
Fig. 5. Increased ROS formation (DHR oxidation) correlated with apoptotic cell results from our group on Thy1-APP transgenic mice, a mouse
death induced by the PS1M146L mutation in CD4+ T lymphocytes from aged
model of Alzheimer disease based on expression of human
mice (19  22 months of age). (A) The PS1M146L mutation increases ROS
APP with FAD mutations. In brain tissue from these mice only
formation (DHR oxidation) independent of the cell type. Increased mean
HNE and not MDA levels were similarly elevated [31], and
fluorescence intensity (MFI) after staining with DHR was detected in CD3+ cells
mitochondrial dysfunction could be demonstrated [59]. In
(total T cells) and in the subpools of CD4+ and CD8+ T cells from PS1M146L
mice. CD4+ T cells accumulate higher ROS levels compared to CD8+ T cells. addition, mitochondrial dysfunction and increased levels of
Two-way ANOVA in CD4+ and CD8+ cells: p < 0.01, effect of PS1M146L
HNE-modified protein were also reported in a different
transgene; p < 0.0001, difference between CD4+ and CD8+ subsets; interaction
PS1M146L transgenic mouse model [60,61]. These findings
n.s. *p 0.05, Student_s t test, PS1M146L (n = 7) vs non-tg (n = 7) in the same
in animal models furthermore correlate well with a study of
subset (CD3+, CD4+); (*)p = 0.07, Student_s t test, PS1M146L (n = 7) vs non-tg
lipid peroxidation levels in blood samples from AD patients, in
(n = 6) in CD8+ subset; ++p < 0.01, Student_s t test, CD4+ cells vs CD8+ cells
from nontransgenic mice. (B) Percentage of apoptotic cells in splenic which only HNE and not MDA levels were specifically
lymphocytes from non-tg (n = 7) and PS1M146L transgenic mice (n = 7). Both
increased compared to nondemented controls [62]. These
CD3+ and CD4+ T lymphocytes from PS1M146L mice exhibit increased levels
results suggest that increased HNE levels are a specific marker
of apoptosis, but differences are not statistically significant. (C) Correlation
for lipid peroxidation processes in consequence to mitochon-
between DHR oxidation and percentage of apoptotic cells in lymphocyte samples
drial dysfunction and oxidative stress in AD patients, and this
isolated from the same individual animal. Correlation calculated by linear
regression: PS1M146L transgenic mice (n = 7), p < 0.05, Pearson_s correlation marker can also be detected in relevant transgenic mouse
coefficient rp = 0.8134; non-tg mice (n = 7), p = 0.05, Pearson_s correlation
models of the disease. As several cytotoxic properties have
coefficient rp = 0.7466; all (non-tg and PS1M146L mice, n = 14), p < 0.01,
been described for HNE [63], it is feasible that HNE can cause
Pearson_s correlation coefficient rp = 0.7548.
neurodegeneration and cognitive deficits, which were de-
scribed in mutant PS1 transgenic mice [24].
oxidative stress in brain tissue, we analyzed a possible The causal mechanism for oxidative stress in PS1M146L
correlation between DHR oxidation in peripheral lymphocytes transgenic mice may not stem from an impairment of
and HNE levels in brain tissue as markers of oxidative stress in
PS1 transgenic mice and littermate controls. As shown in Fig.
6, a significant and positive linear correlation was found
between DHR oxidation in splenic lymphocytes and HNE
levels in brain tissue homogenates in these animals. Hence,
measurement of oxidative stress levels in peripheral cells can
be successfully used as an indicator of central nervous system
oxidative status in our animal model.
Discussion
Fig. 6. In aged mice (19  22 months of age), ROS levels (DHR oxidation) in
Oxidative stress in PS1M146L mice is caused by
splenic lymphocytes are positively correlated with brain tissue levels of lipid
overproduction rather than insufficient detoxification of ROS
peroxidation (HNE). n = 18 nontransgenic controls (non-tg), n = 10 PS1wt
mice, and n = 9 PS1M146L mice were studied. Correlation calculated by
Our results demonstrate the presence of oxidative stress in
linear regression in all samples (n = 37): p < 0.05, Pearson_s correlation
PS1M146L transgenic mice as assessed by increased HNE coefficient rp = 0.3376.
K. Schuessel et al. / Free Radical Biology & Medicine 40 (2006) 850  862 857
antioxidant defenses but rather from an increased production of transgenic mice (19 22 months) displayed increased HNE
ROS. This is supported by our observations that activities of levels. This is paralleled by the alterations observed in splenic
several antioxidant enzymes in brain tissue Cu/Zn-dependent lymphocytes from these mice: whereas no differences in ROS
superoxide dismutase, glutathione peroxidase, and glutathione levels were observed in cells derived from young animals,
reductase are unchanged. Additionally, in vitro stimulation of neither under basal nor under stimulated conditions, lympho-
lipid peroxidation in brain homogenates by addition of ferric cytes from middle-aged PS1M146L animals showed unaltered
iron, which induces hydroxyl radical formation via Haber basal ROS levels but increased ROS levels after serum
Weiss and Fenton reactions [64], did not lead to a different deprivation, which represents a mild stress condition leading
extent of MDA formation in PS1 transgenic animals and to apoptosis induction in lymphocytes [71,72]. Finally, in cells
controls. Additionally, no exacerbation of ROS formation from aged animals, even basal levels of ROS were increased.
could be detected in splenic lymphocytes after oxidative Thus, the aging process seems to sensitize toward the ROS-
stimulation with hydrogen peroxide, which leads to hydroxyl elevating effect of the PS1M146L mutation. Of note, the
radical formation via Fenton reactions and mitochondrial presence of PS1wt in transgenic mice did not alter levels of
damage [65]. Hence, an impairment of antioxidant defense lipid peroxidation products in brain tissue nor ROS levels in
mechanisms as underlying causal mechanism for oxidative lymphocytes, suggesting that oxidative stress is specifically
stress in PS1M146L mice seems unlikely. caused by the presence of the M146L mutation. It can be
In contrast, impairment of antioxidant enzymes could be speculated that the aging process triggers the effect of the PS1
identified in previous studies in PS1M5 mice bearing mutant mutation. A slight but not significant elevation of mitochon-
human presenilin with five different FAD mutations [34] and in drial ROS levels was observed in murine splenic lymphocytes
Thy1-APP transgenic mice [31]. Thy1-APP transgenic mice during aging (data not shown), and other groups have reported
accumulate high levels of Ah with aging, and the presence of increased levels of oxidative damage in lymphocytes from aged
multiple FAD mutations in mutant PS1 shows additive effects rodents and humans [73,74]. Thus, an elevation of ROS levels
on Ah formation [66]. Therefore, in PS1M5 and in Thy1-APP with aging may lower the threshold of oxidative homeostasis,
transgenic mice, high levels of Ah may lead to impairment of leading to exaggerated ROS accumulation in PS1M146L-
antioxidant defenses possibly via an interaction with trace expressing cells from aged mice.
elements essential for antioxidant activity [33]. In contrast,
lower levels of Ah formation may not be sufficient to elicit Specific elevation of mitochondrial and cytosolic ROS but not
significant impairment of antioxidant defenses, and such a superoxide radicals caused by the PS1 mutation
deficit was also not identified in PS1M146L mice. Neverthe-
less, we found markers of oxidative stress in PS1M146L As various ROS are produced in living cells and can interact
transgenic mice, suggesting that already low levels of Ah can in complex reactions, we additionally stained lymphocytes with
have deleterious effects apart from the impairment of several different ROS-sensitive dyes for an estimation of the
antioxidant enzymes. exact type and the source of ROS that accounts for increased
Of note, PS1M146L mice do not develop amyloid plaques DHR oxidation caused by PS1M146L. The increased oxidation
in brain tissue with aging similar to other mutant PS1 of DHR (142.1% relative to controls) is indicative of elevated
transgenic mice [22,67]. Therefore, the presence of Ah mitochondrial ROS levels [44,75]. Furthermore, cytosolic ROS
deposits is not necessarily a prerequisite for triggering levels measured by oxidation of H2DCF-DA [48] were also
oxidative damage. This is in accordance with studies on human increased in PS1M146L mice (120.5% relative to controls).
brain tissue, which suggest that oxidative stress is a very early However, superoxide levels as detected by DHE oxidation and
event in AD patients whereas amyloid plaques form at later nitric oxide levels as assessed by DAF-2-DA oxidation were
stages [68,69]. Also, oxidative damage could be identified in not altered, suggesting that peroxynitrite is not a major
Thy1-APP transgenic mice from as early as 3 months of age contributor to the increased DHR and H2DCF-DA oxidation
[31]. These findings support the notion that already low levels in lymphocytes from PS1M146L transgenic mice.
of Ah can result in oxidative stress even in the absence of The relatively higher rate of DHR oxidation (142.1%
amyloid plaques. Moreover, neurodegeneration was identified relative to controls) compared to H2DCF-DA oxidation
in mutant PS1 transgenic mice not showing amyloid plaque (120.5% relative to controls) suggests that ROS are primarily
deposition [22,23] and in transgenic mice expressing mouse produced by mitochondria in PS1M146L-expressing cells.
amyloid h [70], suggesting that the formation of mouse Mitochondrial ROS formation may be a consequence of
amyloid h itself does not require the deposition of amyloid deleterious effects of Ah on mitochondria as described
plaques for provoking deleterious effects. previously [59,76], which is supported by findings that Ah
and both APP and active g-secretase complexes containing PS1
Aging sensitizes toward the effect of the PS1 mutation can localize to mitochondria [77 79]. Mitochondrial ROS
formation in PS1M146L transgenic mice is in close analogy to
Of note, the aging process sensitizes toward the deleterious other reports of mitochondrial toxicity and oxidative stress in
effects of the PS1 mutation: whereas no differences in HNE several AD mouse models based on APP mutations
levels were found in brain tissue from young (3 4 months) or [31,59,80,81] and double transgenic APP/PS1 mice [82].
middle-aged animals (13  15 months), aged PS1M146L Moreover, mitochondrial dysfunction was identified in cybrids
858 K. Schuessel et al. / Free Radical Biology & Medicine 40 (2006) 850  862
bearing mitochondrial DNA from AD patients [83]. These mitochondrial membrane potential, was not impaired. For these
findings and our observation that mitochondrial ROS formation reasons, it is unlikely that increased DHR oxidation observed
can also be detected in PS1M146L transgenic mice add further in splenic lymphocytes from PS1M146L transgenic mice is a
importance to hypotheses proposing mitochondrial dysfunction secondary event due to loss of mitochondrial membrane
as central to the pathogenesis of AD [84,85]. integrity and apoptotic processes. Rather, increased ROS
Mitochondrial membrane potential was not disrupted in formation can be considered the trigger for apoptosis in
PS1M146L cells, as R123 uptake into mitochondria was even PS1M146L-expressing CD4+ lymphocytes.
slightly but significantly increased. This is consistent with
previous findings from our group in which PC12 cells Use of peripheral cells to study effects in brain tissue
transfected with FAD mutant APP display an increased
mitochondrial membrane potential [59], which can be consid- So far, numerous studies have been conducted on peripheral
ered a compensatory reaction to deleterious effects of Ah on cell models like lymphocytes, platelets, fibroblasts, or olfactory
mitochondria. Furthermore, cultured cells from AD patients neurons from Alzheimer disease patients, in which increased
bearing PS1 mutations do not show a decreased mitochondrial apoptosis and markers of oxidative damage could be identified
membrane potential [86]. The extent of increased R123 [35,83,88,95 100]. These results suggest that inherent Alzhei-
incorporation into mitochondria at 7.2% relative to controls mer-specific properties of cells exist independent of the cell
is, however, not sufficient to explain the 42.1% increase in type which can become evident also in the periphery.
DHR fluorescence. Therefore, the major part of increased DHR Moreover, cybrid cell lines containing mitochondrial DNA
fluorescence must be due to oxidation of the dye. derived from platelets of AD patients show mitochondrial
defects and increased ROS production [101], suggesting that it
Functional relevance of mitochondrial ROS (DHR oxidation) is the mitochondria from AD patients which harbor the relevant
for apoptotic cell death in CD4+ lymphocytes from PS1M146L defects. As any viable cell contains mitochondria, peripheral
mice cells may therefore be used to monitor pathogenic events in
brain tissue. However, evidence for such a correlation between
As previous studies from our group and others had peripheral cells and central nervous system tissue in humans is
identified a higher vulnerability of the CD4+ lymphocyte currently lacking. Our data in PS1M146L transgenic mice
subset to oxidative stress and apoptosis during aging in show that DHR oxidation in splenic lymphocytes was
PS1M146L transgenic mice and in AD patients [35,87,88], positively correlated with HNE levels in brain tissue from
we analyzed specific T lymphocyte subsets from transgenic each individual animal. Although direct assessment of HNE
mice for a putatively different formation of ROS. The effects of levels in splenic lymphocytes was precluded due to the need
the PS1 mutation were not different in CD4+ or CD8+ lym- for high tissue amounts, the correlation between peripheral
phocyte subsets, suggesting that the transgene leads to similar ROS and brain tissue HNE levels suggests that oxidative stress
effects independent of the cell type. However, we could in brain tissue can be mirrored in peripheral cells. This adds
observe that DHR oxidation was specifically higher in the further importance and relevance to the numerous studies
CD4+ lymphocyte subset compared to CD8+ lymphocytes, and conducted on peripheral cell models in neurodegenerative
DHR oxidation was positively correlated with cell death in diseases so far. Moreover, as DHR oxidation detects increased
CD4+ cells. Hence, in CD4+ lymphocytes, increased mito- mitochondrial ROS levels, it can be speculated that mitochon-
chondrial ROS formation may trigger apoptosis. This is in drial ROS formation underlies the formation of HNE in
good accordance with the important role of mitochondria as the PS1M146L mice. This would be in analogy to previous studies
central executioners in apoptotic cell death signaling [89,90]. from our group on Thy1-APP transgenic mice as a different
Mitochondrial damage can trigger apoptosis probably via mouse model of AD. In these mice, similarly increased levels
glutathione depletion [91] followed by oxidation of mito- of HNE [31], reactive oxygen species [80], and mitochondrial
chondrial thiols and loss of mitochondrial membrane potential dysfunction [59] could be detected.
[92], providing a functional link between increased mitochon-
drial DHR oxidation in CD4+ lymphocytes and apoptosis. Summary of mutant PS1-induced effects
Although it is often difficult to determine whether ROS
initiate apoptosis or are a by-product of the apoptotic process, In summary, our results on changes in oxidative stress
our results suggest that increased ROS levels are a trigger of parameters observed in brain tissue and peripheral cells support
apoptotic cell death in lymphocytes in our experiments. First, the hypothesis of oxidative toxicity caused specifically by the
only viable lymphocytes, not apoptosing cells, were gated by mutant PS1M146L compared to PS1wt transgenic or nontrans-
flow cytometry analysis for quantification of ROS levels. genic mice. Moreover, oxidative toxicity of the PS1M146L
Second, a decrease in mitochondrial membrane potential has mutation is evident only in aged animals, arguing for an
been consistently reported to result in overproduction of additive effect of aging the most important risk factor for the
superoxide radicals [93,94]. However, we did not detect development of sporadic AD and the FAD mutation on
increased superoxide formation in PS1M146L-expressing cells elevated sensitivity toward ROS accumulation. In brain tissue,
as evidenced by unaltered DHE oxidation. Furthermore, R123 increased lipid peroxidation as assessed by elevated HNE
uptake into mitochondria, which depends upon maintained levels could be observed in aged PS1M146L mice. In parallel,
K. Schuessel et al. / Free Radical Biology & Medicine 40 (2006) 850  862 859
Yager, D.; Slunt, H. H.; Wang, R.; Seeger, M.; Levey, A. I.; Gandy,
mitochondrial ROS were 42.1% and cytosolic ROS 20.5%
S. E.; Copeland, N. G.; Jenkins, N. A.; Price, D. L.; Younkin, S. G.;
higher in PS1M146L cells relative to cells from nontransgenic
Sisodia, S. S. Familial Alzheimer_s disease-linked presenilin 1
littermate control mice, pointing to mitochondria as the main
variants elevate Abeta1-42/1-40 ratio in vitro and in vivo. Neuron
source of ROS. Increased mitochondrial ROS formation may
17:1005  1013; 1996.
trigger apoptotic cell death in vulnerable cells, as DHR [9] Citron, M.; Westaway, D.; Xia, W.; Carlson, G.; Diehl, T.;
Levesque, G.; Johnson-Wood, K.; Lee, M.; Seubert, P.; Davis, A.;
oxidation and levels of apoptosis were positively correlated
Kholodenko, D.; Motter, R.; Sherrington, R.; Perry, B.; Yao, H.;
in CD4+ cells. The mechanism of ROS accumulation in
Strome, R.; Lieberburg, I.; Rommens, J.; Kim, S.; Schenk, D.;
PS1M146L transgenic mice is probably related to increased
Fraser, P.; St George, H. P.; Selkoe, D. J. Mutant presenilins of
ROS formation but not impaired detoxification, as we observed
Alzheimer_s disease increase production of 42-residue amyloid beta-
neither an impairment of antioxidant enzyme activities in brain protein in both transfected cells and transgenic mice. Nat. Med.
3:67  72; 1997.
tissue nor a different susceptibility toward oxidative stimuli in
[10] Lemere, C. A.; Lopera, F.; Kosik, K. S.; Lendon, C. L.; Ossa, J.; Saido,
brain tissue and in lymphocytes.
T. C.; Yamaguchi, H.; Ruiz, A.; Martinez, A.; Madrigal, L.; Hincapie,
In conclusion, our results support the hypothesis that the
L.; Arango, J. C.; Anthony, D. C.; Koo, E. H.; Goate, A. M.; Selkoe, D.
combined effects of aging and the presence of PS1 mutations
J.; Arango, J. C. The E280A presenilin 1 Alzheimer mutation produces
can lead to oxidative stress, which may be a causative event for increased A beta 42 deposition and severe cerebellar pathology. Nat.
Med. 2:1146 1150; 1996.
triggering neurodegeneration in familial AD patients. Further
[11] Duering, M.; Grimm, M. O.; Grimm, H. S.; Schroder, J.; Hartmann, T.
studies will be conducted to enlighten the exact mechanisms
Mean age of onset in familial Alzheimer_s disease is determined by
underlying the oxidative stress cascade in this model.
amyloid beta 42. Neurobiol. Aging 26:785  788; 2005.
[12] Berezovska, O.; Lleo, A.; Herl, L. D.; Frosch, M. P.; Stern, E.
Acknowledgments A.; Bacskai, B. J.; Hyman, B. T. Familial Alzheimer_s disease
presenilin 1 mutations cause alterations in the conformation of
presenilin and interactions with amyloid precursor protein.
We thank Dr. Christian Czech (Hoffmann La Roche AD,
J. Neurosci. 25:3009  3017; 2005.
Basel, Switzerland) and Dr. Laurent Pradier (Sanofi-Aventis,
[13] Suzuki, N.; Cheung, T. T.; Cai, X. D.; Odaka, A.; Otvos, L., Jr.;
Vitry-sur-Seine, France) for providing the PS1 transgenic
Eckman, C.; Golde, T. E.; Younkin, S. G. An increased percentage of
mouse models. This work was supported by grants from long amyloid beta protein secreted by familial amyloid beta protein
precursor (beta APP717) mutants. Science 264:1336  1340; 1994.
Alzheimer Forschung Initiative (No. 01808) and from Hes-
[14] Citron, M.; Oltersdorf, T.; Haass, C.; McConlogue, L.; Hung, A. Y.;
sisches Kultusministerium (K.S.).
Seubert, P.; Vigo-Pelfrey, C.; Lieberburg, I.; Selkoe, D. J. Mutation of
the beta-amyloid precursor protein in familial Alzheimer_s disease
References
increases beta-protein production. Nature 360:672  674; 1992.
[15] Hardy, J. A.; Higgins, G. A. Alzheimer_s disease: the amyloid cascade
[1] George-Hyslop, P. H. Molecular genetics of Alzheimer_s disease. Biol. hypothesis. Science 256:184  185; 1992.
Psychiatry 47:183  199; 2000. [16] Hardy, J.; Selkoe, D. J. The amyloid hypothesis of Alzheimer_s
[2] Campion, D.; Flaman, J. M.; Brice, A.; Hannequin, D.; Dubois, B.; disease: progress and problems on the road to therapeutics. Science
Martin, C.; Moreau, V.; Charbonnier, F.; Didierjean, O.; Tardieu, S. 297:353  356; 2002.
Mutations of the presenilin I gene in families with early-onset [17] Guo, Q.; Sopher, B. L.; Furukawa, K.; Pham, D. G.; Robinson, N.;
Alzheimer_s disease. Hum. Mol. Genet. 4:2373  2377; 1995. Martin, G. M.; Mattson, M. P. Alzheimer_s presenilin mutation
[3] Dowjat, W. K.; Kuchna, I.; Wisniewski, T.; Wegiel, J. A novel highly sensitizes neural cells to apoptosis induced by trophic factor withdrawal
pathogenic Alzheimer presenilin-1 mutation in codon 117 (Pro117S- and amyloid beta-peptide: involvement of calcium and oxyradicals.
er): comparison of clinical, neuropathological and cell culture J. Neurosci. 17:4212  4222; 1997.
phenotypes of Pro117Leu and Pro117Ser mutations. J. Alzheimers [18] Wolozin, B.; Alexander, P.; Palacino, J. Regulation of apoptosis by
Dis. 6:31  43; 2004. presenilin 1. Neurobiol. Aging 19:S23  S27; 1998.
[4] Lleo, A.; Berezovska, O.; Growdon, J. H.; Hyman, B. T. Clinical, [19] Weihl, C. C.; Ghadge, G. D.; Kennedy, S. G.; Hay, N.; Miller, R. J.;
pathological, and biochemical spectrum of Alzheimer disease associ- Roos, R. P. Mutant presenilin-1 induces apoptosis and downregulates
ated with PS-1 mutations. Am. J. Geriatric Psychiatry 12:146  156; Akt/PKB. J. Neurosci. 19:5360  5369; 1999.
2004. [20] Guo, Q.; Sebastian, L.; Sopher, B. L.; Miller, M. W.; Ware, C. B.;
[5] De Strooper, B. Aph-1, Pen-2, and nicastrin with presenilin generate an Martin, G. M.; Mattson, M. P. Increased vulnerability of hippocampal
active gamma-secretase complex. Neuron 38:9  12; 2003. neurons from presenilin-1 mutant knock-in mice to amyloid beta-
[6] Scheuner, D.; Eckman, C.; Jensen, M.; Song, X.; Citron, M.; Suzuki, peptide toxicity: central roles of superoxide production and caspase
N.; Bird, T. D.; Hardy, J.; Hutton, M.; Kukull, W.; Larson, E.; Levy- activation. J. Neurochem. 72:1019  1029; 1999.
Lahad, E.; Viitanen, M.; Peskind, E.; Poorkaj, P.; Schellenberg, G.; [21] Czech, C.; Lesort, M.; Tremp, G.; Terro, F.; Blanchard, V.; Schombert,
Tanzi, R.; Wasco, W.; Lannfelt, L.; Selkoe, D.; Younkin, S. Secreted B.; Carpentier, N.; Dreisler, S.; Bonici, B.; Takashima, A.; Moussaoui,
amyloid beta-protein similar to that in the senile plaques of S.; Hugon, J.; Pradier, L. Characterization of human presenilin 1
Alzheimer_s disease is increased in vivo by the presenilin 1 and 2 transgenic rats: increased sensitivity to apoptosis in primary neuronal
and APP mutations linked to familial Alzheimer_s disease. Nat. Med. cultures. Neuroscience 87:325  336; 1998.
2:864  870; 1996. [22] Chui, D. H.; Tanahashi, H.; Ozawa, K.; Ikeda, S.; Checler, F.; Ueda, O.;
[7] Duff, K.; Eckman, C.; Zehr, C.; Yu, X.; Prada, C. M.; Perez-tur, J.; Suzuki, H.; Araki, W.; Inoue, H.; Shirotani, K.; Takahashi, K.; Gallyas,
Hutton, M.; Buee, L.; Harigaya, Y.; Yager, D.; Morgan, D.; Gordon, F.; Tabira, T. Transgenic mice with Alzheimer presenilin 1 mutations
M. N.; Holcomb, L.; Refolo, L.; Zenk, B.; Hardy, J.; Younkin, S. show accelerated neurodegeneration without amyloid plaque formation.
Increased amyloid-beta42(43) in brains of mice expressing mutant Nat. Med. 5:560  564; 1999.
presenilin 1. Nature 383:710  713; 1996. [23] Sadowski, M.; Pankiewicz, J.; Scholtzova, H.; Ji, Y.; Quartermain, D.;
[8] Borchelt, D. R.; Thinakaran, G.; Eckman, C. B.; Lee, M. K.; Jensen, C. H.; Duff, K.; Nixon, R. A.; Gruen, R. J.; Wisniewski, T.
Davenport, F.; Ratovitsky, T.; Prada, C. M.; Kim, G.; Seekins, S.; Amyloid-beta deposition is associated with decreased hippocampal
860 K. Schuessel et al. / Free Radical Biology & Medicine 40 (2006) 850  862
glucose metabolism and spatial memory impairment in APP/PS1 mice. Spectrophotometric assay of superoxide dismutase activity based on the
J. Neuropathol. Exp. Neurol. 63:418  428; 2004. activated autoxidation of a tetracyclic catechol. Anal. Biochem.
[24] Pak, K.; Chan, S. L.; Mattson, M. P. Presenilin-1 mutation sensitizes 214:442  451; 1993.
oligodendrocytes to glutamate and amyloid toxicities, and exacerbates [42] Paglia, D. E.; Valentine, W. N. Studies on the quantitative and
white matter damage and memory impairment in mice. Neuromol. Med. qualitative characterization of erythrocyte glutathione peroxidase.
3:53 64; 2003. J. Lab. Clin. Med. 70:158  169; 1967.
[25] Harman, D. Free radical involvement in aging: pathophysiology and [43] Mizuno, Y.; Ohta, K. Regional distributions of thiobarbituric acid-
therapeutic implications. Drugs Aging 3:60  80; 1993. reactive products, activities of enzymes regulating the metabolism of
[26] Lovell, M. A.; Ehmann, W. D.; Butler, S. M.; Markesbery, W. R. oxygen free radicals, and some of the related enzymes in adult and aged
Elevated thiobarbituric acid-reactive substances and antioxidant rat brains. J. Neurochem. 46:1344  1352; 1986.
enzyme activity in the brain in Alzheimer_s disease. Neurology [44] Royall, J. A.; Ischiropoulos, H. Evaluation of 2V,7V-dichlorofluorescin
45:1594  1601; 1995. and dihydrorhodamine 123 as fluorescent probes for intracellular H2O2
[27] Schuessel, K.; Leutner, S.; Cairns, N. J.; Muller, W. E.; Eckert, A. in cultured endothelial cells. Arch. Biochem. Biophys. 302:348  355;
Impact of gender on upregulation of antioxidant defence mechanisms in 1993.
Alzheimer_s disease brain. J. Neural Transm. 111:1167  1182; 2004. [45] Efferth, T.; Lohrke, H.; Volm, M. Reciprocal correlation between
[28] Nunomura, A.; Chiba, S.; Lippa, C. F.; Cras, P.; Kalaria, R. N.; Takeda, expression of P-glycoprotein and accumulation of rhodamine 123 in
A.; Honda, K.; Smith, M. A.; Perry, G. Neuronal RNA oxidation is a human tumors. Anticancer Res. 9:1633  1637; 1989.
prominent feature of familial Alzheimer_s disease. Neurobiol. Dis. [46] Bommhardt, U.; Cerottini, J. C.; MacDonald, H. R. Heterogeneity in P-
17:108  113; 2004. glycoprotein (multidrug resistance) activity among murine peripheral T
[29] Behl, C.; Davis, J. B.; Lesley, R.; Schubert, D. Hydrogen peroxide cells: correlation with surface phenotype and effector function. Eur. J.
mediates amyloid beta protein toxicity. Cell 77:817  827; 1994. Immunol. 24:2974  2981; 1994.
[30] Pratico, D.; Uryu, K.; Leight, S.; Trojanoswki, J. Q.; Lee, V. M. [47] Hempel, S. L.; Buettner, G. R.; O_Malley, Y. Q.; Wessels, D. A.;
Increased lipid peroxidation precedes amyloid plaque formation in an Flaherty, D. M. Dihydrofluorescein diacetate is superior for detecting
animal model of Alzheimer amyloidosis. J. Neurosci. 21:4183  4187; intracellular oxidants: comparison with 2V,7V-dichlorodihydrofluorescein
2001. diacetate, 5(and 6)-carboxy-2V,7V-dichlorodihydrofluorescein diacetate,
[31] Schuessel, K.; Schafer, S.; Bayer, T. A.; Czech, C.; Pradier, L.; Muller- and dihydrorhodamine 123. Free Radic. Biol. Med. 27:146  159; 1999.
Spahn, F.; Muller, W. E.; Eckert, A. Impaired Cu/Zn-SOD activity [48] Diaz, G.; Liu, S.; Isola, R.; Diana, A.; Falchi, A. M. Mitochondrial
contributes to increased oxidative damage in APP transgenic mice. localization of reactive oxygen species by dihydrofluorescein probes.
Neurobiol. Dis. 18:89  99; 2005. Histochem. Cell Biol. 120:319  325; 2003.
[32] Hashimoto, Y.; Ito, Y.; Arakawa, E.; Kita, Y.; Terashita, K.; Niikura, T.; [49] Walrand, S.; Valeix, S.; Rodriguez, C.; Ligot, P.; Chassagne, J.; Vasson,
Nishimoto, I. Neurotoxic mechanisms triggered by Alzheimer_s disease- M. P. Flow cytometry study of polymorphonuclear neutrophil oxidative
linked mutant M146L presenilin 1: involvement of NO synthase via a burst: a comparison of three fluorescent probes. Clin. Chim. Acta
novel pertussis toxin target. J. Neurochem. 80:426  437; 2002. 331:103  110; 2003.
[33] Bayer, T. A.; Schafer, S.; Simons, A.; Kemmling, A.; Kamer, T.; Tepest, [50] Crow, J. P. Dichlorodihydrofluorescein and dihydrorhodamine 123 are
R.; Eckert, A.; Schussel, K.; Eikenberg, O.; Sturchler-Pierrat, C.; sensitive indicators of peroxynitrite in vitro: implications for intracel-
Abramowski, D.; Staufenbiel, M.; Multhaup, G. Dietary Cu stabilizes lular measurement of reactive nitrogen and oxygen species. Nitric
brain superoxide dismutase 1 activity and reduces amyloid Abeta Oxide 1:145  157; 1997.
production in APP23 transgenic mice. Proc. Natl. Acad. Sci. USA [51] Huie, R. E.; Padmaja, S. The reaction of NO with superoxide. Free
100:14187 14192; 2003. Radic. Res. Commun. 18:195  199; 1993.
[34] Leutner, S.; Czech, C.; Schindowski, K.; Touchet, N.; Eckert, A.; [52] Rothe, G.; Valet, G. Flow cytometric analysis of respiratory burst
Muller, W. E. Reduced antioxidant enzyme activity in brains of mice activity in phagocytes with hydroethidine and 2V,7V-dichlorofluorescin.
transgenic for human presenilin-1 with single or multiple mutations. J. Leukocyte Biol. 47:440  448; 1990.
Neurosci. Lett. 292:87  90; 2000. [53] Kojima, H.; Nakatsubo, N.; Kikuchi, K.; Kawahara, S.; Kirino, Y.;
[35] Schindowski, K.; Kratzsch, T.; Peters, J.; Steiner, B.; Leutner, S.; Nagoshi, H.; Hirata, Y.; Nagano, T. Detection and imaging of nitric
Touchet, N.; Maurer, K.; Czech, C.; Pradier, L.; Frolich, L.; Muller, W. oxide with novel fluorescent indicators: diaminofluoresceins. Anal.
E.; Eckert, A. Impact of aging: sporadic, and genetic risk factors on Chem. 70:2446  2453; 1998.
vulnerability to apoptosis in Alzheimer_s disease. Neuromol. Med. [54] Schmid, I.; Uittenbogaart, C. H.; Giorgi, J. V. Sensitive method for
4:161  178; 2003. measuring apoptosis and cell surface phenotype in human thymocytes
[36] Czech, C.; Delaere, P.; Macq, A. F.; Reibaud, M.; Dreisler, S.; Touchet, by flow cytometry. Cytometry 15:12 20; 1994.
N.; Schombert, B.; Mazadier, M.; Mercken, L.; Theisen, M.; Pradier, [55] Lecoeur, H.; Oliveira-Pinto, L. M.; Gougeon, M. L. Multiparametric
L.; Octave, J. N.; Beyreuther, K.; Tremp, G. Proteolytical processing of flow cytometric analysis of biochemical and functional events
mutated human amyloid precursor protein in transgenic mice. Brain associated with apoptosis and oncosis using the 7-aminoactinomycin
Res. Mol. Brain Res. 47:108 116; 1997. D assay. J. Immunol. Methods 265:81  96; 2002.
[37] Dyrks, T.; Dyrks, E.; Masters, C. L.; Beyreuther, K. Amyloidogeni- [56] Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. J.
city of rodent and human beta A4 sequences. FEBS Lett. 324:231  Protein measurement with the folin phenol reagent. J. Biol. Chem.
236; 1993. 193:265  275; 1951.
[38] Eckert, A.; Schindowski, K.; Leutner, S.; Luckhaus, C.; Touchet, N.; [57] Czech, C.; Tremp, G.; Pradier, L. Presenilins and Alzheimer_s disease:
Czech, C.; Muller, W. E. Alzheimer_s disease-like alterations in biological functions and pathogenic mechanisms. Prog. Neurobiol.
peripheral cells from presenilin-1 transgenic mice. Neurobiol. Dis. 60:363  384; 2000.
8:331  342; 2001. [58] LaFontaine, M. A.; Mattson, M. P.; Butterfield, D. A. Oxidative stress in
[39] Esterbauer, H.; Cheeseman, K. H. Determination of aldehydic lipid synaptosomal proteins from mutant presenilin-1 knock-in mice: impli-
peroxidation products: malonaldehyde and 4-hydroxynonenal. Methods cations for familial Alzheimer_s disease. Neurochem. Res. 27:417  421;
Enzymol. 186:407  421; 1990. 2002.
[40] Gerard-Monnier, D.; Erdelmeier, I.; Regnard, K.; Moze-Henry, N.; Yadan, J. [59] Keil, U.; Bonert, A.; Marques, C. A.; Scherping, I.; Weyermann, J.;
C.; Chaudiere, J. Reactions of 1-methyl-2-phenylindole with malondialde- Strosznajder, J. B.; Muller-Spahn, F.; Haass, C.; Czech, C.; Pradier, L.;
hyde and 4-hydroxyalkenals: analytical applications to a colorimetric assay Muller, W. E.; Eckert, A. Amyloid-beta induced changes in nitric oxide
of lipid peroxidation. Chem. Res. Toxicol. 11:1176 1183; 1998. production and mitochondrial activity lead to apoptosis. J. Biol. Chem.
[41] Nebot, C.; Moutet, M.; Huet, P.; Xu, J. Z.; Yadan, J. C.; Chaudiere, J. 279:50310 50320; 2004.
K. Schuessel et al. / Free Radical Biology & Medicine 40 (2006) 850  862 861
[60] Begley, J. G.; Duan, W.; Chan, S.; Duff, K.; Mattson, M. P. Altered mutation: oxidative stress, caspases, and the JNK pathway. J. Biol.
calcium homeostasis and mitochondrial dysfunction in cortical Chem. 278:28294 28302; 2003.
synaptic compartments of presenilin-1 mutant mice. J. Neurochem. [77] Crouch, P. J.; Blake, R.; Duce, J. A.; Ciccotosto, G. D.; Li, Q. X.;
72:1030  1039; 1999. Barnham, K. J.; Curtain, C. C.; Cherny, R. A.; Cappai, R.; Dyrks, T.;
[61] Mohmmad, A. H.; Wenk, G. L.; Gramling, M.; Hauss-Wegrzyniak, B.; Masters, C. L.; Trounce, I. A. Copper-dependent inhibition of human
Butterfield, D. A. APP and PS-1 mutations induce brain oxidative stress cytochrome c oxidase by a dimeric conformer of amyloid-beta 1-42.
independent of dietary cholesterol: implications for Alzheimer_s J. Neurosci. 25:672 679; 2005.
disease. Neurosci. Lett. 368:148  150; 2004. [78] Anandatheerthavarada, H. K.; Biswas, G.; Robin, M. A.; Avadhani,
[62] McGrath, L. T.; McGleenon, B. M.; Brennan, S.; McColl, D.; McILroy, N. G. Mitochondrial targeting and a novel transmembrane arrest of
S.; Passmore, A. P. Increased oxidative stress in Alzheimer_s disease as Alzheimer_s amyloid precursor protein impairs mitochondrial func-
assessed with 4-hydroxynonenal but not malondialdehyde. Q. J. Med. tion in neuronal cells. J. Cell Biol. 161:41 54; 2003.
94:485  490; 2001. [79] Hansson, C. A.; Frykman, S.; Farmery, M. R.; Tjernberg, L. O.;
[63] Kruman, I.; Bruce-Keller, A. J.; Bredesen, D.; Waeg, G.; Mattson, M. P. Nilsberth, C.; Pursglove, S. E.; Ito, A.; Winblad, B.; Cowburn, R. F.;
Evidence that 4-hydroxynonenal mediates oxidative stress-induced Thyberg, J.; Ankarcrona, M. Nicastrin, presenilin, APH-1, and PEN-2
neuronal apoptosis. J. Neurosci. 17:5089  5100; 1997. form active gamma-secretase complexes in mitochondria. J. Biol.
[64] Gutteridge, J. M. Lipid peroxidation initiated by superoxide-dependent Chem. 279:51654  51660; 2004.
hydroxyl radicals using complexed iron and hydrogen peroxide. FEBS [80] Scherping, I.; Czech, C.; Pradier, L.; Muller, W. E.; Eckert, A.
Lett. 172:245  249; 1984. Mitochondrial dysfunction in amyloid precursor protein transgenic
[65] Dumont, A.; Hehner, S. P.; Hofmann, T. G.; Ueffing, M.; Droge, mouse model of Alzheimer_s disease. Naunyn Schmiedeberg_s Arch.
W.; Schmitz, M. L. Hydrogen peroxide-induced apoptosis is CD95- Pharmacol. 369 (Suppl. 1):R15; 2004.
independent, requires the release of mitochondria-derived reactive [81] Aliev, G.; Seyidova, D.; Lamb, B. T.; Obrenovich, M. E.; Siedlak, S.
oxygen species and the activation of NF-kappaB. Oncogene 18: L.; Vinters, H. V.; Friedland, R. P.; LaManna, J. C.; Smith, M. A.; Perry,
747  757; 1999. G. Mitochondria and vascular lesions as a central target for the
[66] Citron, M.; Eckman, C. B.; Diehl, T. S.; Corcoran, C.; Ostaszewski, B. development of Alzheimer_s disease and Alzheimer disease-like
L.; Xia, W.; Levesque, G.; St George, H. P.; Younkin, S. G.; Selkoe, D. J. pathology in transgenic mice. Neurol. Res. 25:665  674; 2003.
Additive effects of PS1 and APP mutations on secretion of the 42- [82] Blanchard, V.; Moussaoui, S.; Czech, C.; Touchet, N.; Bonici, B.;
residue amyloid beta-protein. Neurobiol. Dis. 5:107 115; 1998. Planche, M.; Canton, T.; Jedidi, I.; Gohin, M.; Wirths, O.; Bayer, T. A.;
[67] Holcomb, L.; Gordon, M. N.; McGowan, E.; Yu, X.; Benkovic, S.; Langui, D.; Duyckaerts, C.; Tremp, G.; Pradier, L. Time sequence of
Jantzen, P.; Wright, K.; Saad, I.; Mueller, R.; Morgan, D.; Sanders, S.; maturation of dystrophic neurites associated with Abeta deposits in
Zehr, C.; O_Campo, K.; Hardy, J.; Prada, C. M.; Eckman, C.; Younkin, APP/PS1 transgenic mice. Exp. Neurol. 184:247  263; 2003.
S.; Hsiao, K.; Duff, K. Accelerated Alzheimer-type phenotype in [83] Cardoso, S. M.; Santana, I.; Swerdlow, R. H.; Oliveira, C. R.
transgenic mice carrying both mutant amyloid precursor protein and Mitochondria dysfunction of Alzheimer_s disease cybrids enhances
presenilin 1 transgenes. Nat. Med. 4:97  100; 1998. Abeta toxicity. J. Neurochem. 89:1417  1426; 2004.
[68] Cuajungco, M. P.; Goldstein, L. E.; Nunomura, A.; Smith, M. A.; Lim, [84] Eckert, A.; Keil, U.; Marques, C. A.; Bonert, A.; Frey, C.; Schussel, K.;
J. T.; Atwood, C. S.; Huang, X.; Farrag, Y. W.; Perry, G.; Bush, A. I. Muller, W. E. Mitochondrial dysfunction, apoptotic cell death, and
Evidence that the beta-amyloid plaques of Alzheimer_s disease Alzheimer_s disease. Biochem. Pharmacol. 66:1627  1634; 2003.
represent the redox-silencing and entombment of Abeta by zinc. [85] Zhu, X.; Smith, M. A.; Perry, G.; Aliev, G. Mitochondrial failures
J. Biol. Chem. 275:19439  19442; 2000. in Alzheimer_s disease. Am. J. Alzheimers Dis. Other Dement. 19:
[69] Nunomura, A.; Perry, G.; Aliev, G.; Hirai, K.; Takeda, A.; Balraj, E. K.; 345  352; 2004.
Jones, P. K.; Ghanbari, H.; Wataya, T.; Shimohama, S.; Chiba, S.; [86] Huang, H. M.; Fowler, C.; Xu, H.; Zhang, H.; Gibson, G. E.
Atwood, C. S.; Petersen, R. B.; Smith, M. A. Oxidative damage is the Mitochondrial function in fibroblasts with aging in culture and/or
earliest event in Alzheimer disease. J. Neuropathol. Exp. Neurol. Alzheimer_s disease. Neurobiol. Aging 26:839 848; 2005.
60:759  767; 2001. [87] Schindowski, K.; Frohlich, L.; Maurer, K.; Muller, W. E.; Eckert, A.
[70] LaFerla, F. M.; Tinkle, B. T.; Bieberich, C. J.; Haudenschild, C. C.; Jay, Age-related impairment of human T lymphocytes_ activation: specific
G. The Alzheimer_s A beta peptide induces neurodegeneration and differences between CD4(+) and CD8(+) subsets. Mech. Ageing Dev.
apoptotic cell death in transgenic mice. Nat. Genet. 9:21  30; 1995. 123:375  390; 2002.
[71] Maucher, C.; Weissinger, E. M.; Kremmer, E.; Baccarini, M.; Procyk, [88] Grossmann, A.; Kukull, W. A.; Jinneman, J. C.; Bird, T. D.; Villacres,
K.; Henderson, D. W.; Wolff, L.; Kolch, W.; Kaspers, B.; Mushinski, E. C.; Larson, E. B.; Rabinovitch, P. S. Intracellular calcium response
J. F.; Mischak, H. Activation of bcl-2 suppressible 40 and 44 kDa is reduced in CD4+ lymphocytes in Alzheimer_s disease and in older
p38-like kinases during apoptosis of early and late B lymphocytic persons with Down_s syndrome. Neurobiol. Aging 14:177  185; 1993.
cell lines. FEBS Lett. 427:29 35; 1998. [89] Susin, S. A.; Zamzami, N.; Kroemer, G. Mitochondria as regulators of
[72] Kawanishi, M. Expression of Epstein-Barr virus latent membrane apoptosis: doubt no more. Biochim. Biophys. Acta 1366:151  165;
protein 1 protects Jurkat T cells from apoptosis induced by serum 1998.
deprivation. Virology 228:244  250; 1997. [90] Kroemer, G.; Zamzami, N.; Susin, S. A. Mitochondrial control of
[73] Tian, L.; Cai, Q.; Bowen, R.; Wei, H. Effects of caloric restriction on apoptosis. Immunol. Today 18:44  51; 1997.
age-related oxidative modifications of macromolecules and lymphocyte [91] Macho, A.; Hirsch, T.; Marzo, I.; Marchetti, P.; Dallaporta, B.; Susin,
proliferation in rats. Free Radic. Biol. Med. 19:859  865; 1995. S. A.; Zamzami, N.; Kroemer, G. Glutathione depletion is an early
[74] Mendoza-Nunez, V. M.; Sanchez-Rodriguez, M. A.; Retana-Ugalde, and calcium elevation is a late event of thymocyte apoptosis.
R.; Vargas-Guadarrama, L. A.; Altamirano-Lozano, M. A. Total J. Immunol. 158:4612 4619; 1997.
antioxidant levels, gender, and age as risk factors for DNA damage in [92] Marchetti, P.; Decaudin, D.; Macho, A.; Zamzami, N.; Hirsch, T.;
lymphocytes of the elderly. Mech. Ageing Dev. 122:835  847; 2001. Susin, S. A.; Kroemer, G. Redox regulation of apoptosis: impact of
[75] Ischiropoulos, H.; Gow, A.; Thom, S. R.; Kooy, N. W.; Royall, J. A.; thiol oxidation status on mitochondrial function. Eur. J. Immunol.
Crow, J. P. Detection of reactive nitrogen species using 2,7-dichlor- 27:289 296; 1997.
odihydrofluorescein and dihydrorhodamine 123. Methods Enzymol. [93] Castedo, M.; Macho, A.; Zamzami, N.; Hirsch, T.; Marchetti, P.; Uriel,
301:367  373; 1999. J.; Kroemer, G. Mitochondrial perturbations define lymphocytes under-
[76] Marques, C. A.; Keil, U.; Bonert, A.; Steiner, B.; Haass, C.; going apoptotic depletion in vivo. Eur. J. Immunol. 25:3277  3284;
Muller, W. E.; Eckert, A. Neurotoxic mechanisms caused by the 1995.
Alzheimer_s disease-linked Swedish amyloid precursor protein [94] Zamzami, N.; Marchetti, P.; Castedo, M.; Decaudin, D.; Macho, A.;
862 K. Schuessel et al. / Free Radical Biology & Medicine 40 (2006) 850  862
Sheehan, J. P.; Susin, S. A.; Petit, P. X.; Mignotte, B.; Kroemer, G. [98] Mecocci, P.; Polidori, M. C.; Cherubini, A.; Ingegni, T.; Mattioli, P.;
Sequential reduction of mitochondrial transmembrane potential and Catani, M.; Rinaldi, P.; Cecchetti, R.; Stahl, W.; Senin, U.; Beal, M. F.
generation of reactive oxygen species in early programmed cell death. Lymphocyte oxidative DNA damage and plasma antioxidants in
J. Exp. Med. 182:367  377; 1995. Alzheimer disease. Arch. Neurol. 59:794  798; 2002.
[95] Gibson, G. E.; Huang, H. M. Oxidative processes in the brain and non- [99] Parshad, R. P.; Sanford, K. K.; Price, F. M.; Melnick, L. K.; Nee, L. E.;
neuronal tissues as biomarkers of Alzheimer_s disease. Front. Biosci. Schapiro, M. B.; Tarone, R. E.; Robbins, J. H. Fluorescent light-induced
7:d1007  d1015; 2002. chromatid breaks distinguish Alzheimer disease cells from normal cells
[96] Eckert, A.; Oster, M.; Zerfass, R.; Hennerici, M.; Muller, W. E. in tissue culture. Proc. Natl. Acad. Sci. USA 93:5146  5150; 1996.
Elevated levels of fragmented DNA nucleosomes in native and [100] Ghanbari, H. A.; Ghanbari, K.; Harris, P. L.; Jones, P. K.; Kubat, Z.;
activated lymphocytes indicate an enhanced sensitivity to apoptosis in Castellani, R. J.; Wolozin, B. L.; Smith, M. A.; Perry, G. Oxidative
sporadic Alzheimer_s disease: specific differences to vascular dementia. damage in cultured human olfactory neurons from Alzheimer_s disease
Dement. Geriatr. Cognit. Disord. 12:98  105; 2001. patients. Aging Cell 3:41  44; 2004.
[97] Cecchi, C.; Fiorillo, C.; Sorbi, S.; Latorraca, S.; Nacmias, B.; Bagnoli, [101] Swerdlow, R. H.; Parks, J. K.; Miller, S. W.; Tuttle, J. B.; Trimmer, P. A.;
S.; Nassi, P.; Liguri, G. Oxidative stress and reduced antioxidant Sheehan, J. P.; Bennett, J. P., Jr.; Davis, R. E.; Parker, W. D., Jr. Cybrids
defenses in peripheral cells from familial Alzheimer_s patients. Free in Alzheimer_s disease: a cellular model of the disease? Neurology
Radic. Biol. Med. 33:1372  1379; 2002. 49:918  925; 1997.