A metabolomic study of the CRND8 transgenic mouse model of Alzheimer's disease

Article (PDF Available)inNeurochemistry International 56(8):937-47 · July 2010with153 Reads
DOI: 10.1016/j.neuint.2010.04.001 · Source: PubMed
Alzheimer's disease is the most common neurodegenerative disease of the central nervous system characterized by a progressive loss in memory and deterioration of cognitive functions. In this study the transgenic mouse TgCRND8, which encodes a mutant form of the amyloid precursor protein 695 with both the Swedish and Indiana mutations and develops extracellular amyloid beta-peptide deposits as early as 2-3 months, was investigated. Extract from eight brain regions (cortex, frontal cortex, cerebellum, hippocampus, olfactory bulb, pons, midbrain and striatum) were studied using (1)H NMR spectroscopy. Analysis of the NMR spectra discriminated control from APP695 tissues in hippocampus, cortex, frontal cortex, midbrain and cerebellum, with hippocampal and cortical region being most affected. The analysis of the corresponding loading plots for these brain regions indicated a decrease in N-acetyl-L-aspartate, glutamate, glutamine, taurine (exception hippocampus), gamma-amino butyric acid, choline and phosphocholine (combined resonances), creatine, phosphocreatine and succinate in hippocampus, cortex, frontal cortex (exception gamma-amino butyric acid) and midbrain of affected animals. An increase in lactate, aspartate, glycine (except in midbrain) and other amino acids including alanine (exception frontal cortex), leucine, iso-leucine, valine and water soluble free fatty acids (0.8-0.9 and 1.2-1.3 ppm) were observed in the TgCRND8 mice. Our findings demonstrate that the perturbations in metabolism are more widespread and include the cerebellum and midbrain. Furthermore, metabolic perturbations are associated with a wide range of metabolites which could improve the diagnosis and monitoring of the progression of Alzheimer's disease.
A metabolomic study of the CRND8 transgenic mouse model of Alzheimer’s
Reza M. Salek
, Jing Xia
, Amy Innes
, Brian C. Sweatman
, Robert Adalbert
Suzanne Randle
, Eileen McGowan
, Piers C. Emson
, Julian L. Griffin
Department of Biochemistry, The Hopkins Building, Tennis Court Road, University of Cambridge, Cambridge CB2 1QW, UK
Laboratory of Molecular Neuroscience, The Babraham Institute, Babraham, Cambridge CB22 3AT, UK
Safety Assessment, GlaxoSmithKline, Park Road, Ware, Herts SG12 ODP, UK
Department of Neuroscience, Mayo Clinic, Jacksonville, FL 32224, USA
Laboratory of Molecular Signalling, The Babraham Institute, Babraham, Cambridge CB22 3AT, UK
1. Introduction
Alzheimer’s disease (AD) is the most common cause of
dementia, affecting 7% of the US population over the age of 65,
and is characterized by an accumulation of pathological amyloid
peptide (A
) and tau lesions (Hebert et al., 2003). With the increase
in longevity in most populations, the prevalence of AD is predicted
to rise dramatically over the next 40 years (Suh and Checler, 2002).
AD is a progressive neurodegenerative disease described by
broad cognitive decline involving memory, reason, judgment and
orientation. There are two major hypotheses to explain the
neurodegeneration: (i) the amyloid cascade hypothesis (Hardy,
2006; Hardy et al., 1998; Hardy and Higgins, 1992); and (ii) the
neuronal cytoskeleton degeneration hypothesis (Braak and Braak,
1991). In the former the amyloid precursor protein (APP), a plasma
membrane spanning protein, is inappropriately cleaved giving rise
to an increased concentration of A
then aggregates
extracellularly to form A
plaques (senile plaques or neuritic
plaques) leading to neurodegeneration. Mutations within APP
itself (Citron et al., 1992; Goate et al., 1991; Haass et al., 1995) and
in the presenilins that make up part of the
-secretase complex
that cleave APP (Selkoe and Wolfe, 2000) are common pathogenic
mutations in familial AD. The formation of intracellular neurofi-
brillary tangles in various brain regions of AD sufferers also occurs
(Selkoe, 2001).
Abnormalities in cerebral metabolism in neurodegenerative
diseases including AD has been documented widely over the past
50 years, with decreased cerebral metabolism preceding clinical or
Neurochemistry International 56 (2010) 937–947
Article history:
Received 26 October 2009
Received in revised form 4 April 2010
Accepted 6 April 2010
Available online 14 April 2010
Animal models
Amyloid precursor protein
Alzheimer’s disease is the most common neurodegenerative disease of the central nervous system
characterized by a progressive loss in memory and deterioration of cognitive functions. In this study the
transgenic mouse TgCRND8, which encodes a mutant form of the amyloid precursor protein 695 with
both the Swedish and Indiana mutations and develops extracellular amyloid
-peptide deposits as early
as 2–3 months, was investigated. Extract from eight brain regions (cortex, frontal cortex, cerebellum,
hippocampus, olfactory bulb, pons, midbrain and striatum) were studied using
H NMR spectroscopy.
Analysis of the NMR spectra discriminated control from APP695 tissues in hippocampus, cortex, frontal
cortex, midbrain and cerebellum, with hippocampal and cortical region being most affected. The analysis
of the corresponding loading plots for these brain regions indicated a decrease in N-acetyl-
glutamate, glutamine, taurine (exception hippocampus),
-amino butyric acid, choline andphosphocho-
line (combined resonances), creatine , phosphocreatine and succinate in hippocampus, cortex, frontal
cortex (exception
-amino butyric acid) and midbrain of affected animals. An increase in lactate,
aspartate, glycine (except in midbrain) and other amino acids including alanine (exception frontal
cortex), leucine, iso-leucine, valine and water soluble free fatty acids (0.8–0.9 and 1.2–1.3 ppm) were
observed in the TgCRND8 mice. Our findings demonstrate that the perturbations in metabolism are more
widespread and include the cerebellum and midbrain. Furthermore, metabolic perturbations are
associated with a wide range of metabolites which could improve the diagnosis and monitoring of the
progression of Alzheimer’s disease.
ß 2010 Elsevier Ltd. All rights reserved.
* Corresponding author at: Department of Biochemistry, University of Cambridge,
Building O, Downing Site, Tennis Court Road, Cambridge CB2 1QW, UK.
Tel.: +44 01223 333667; fax: +44 01223 766002.
E-mail address: jlg40@mole.bio.cam.ac.uk (J.L. Griffin).
Contents lists available at ScienceDirect
Neurochemistry International
journal homepage: www.elsevier.com/locate/neuint
0197-0186/$ see front matter ß 2010 Elsevier Ltd. All rights reserved.
neuroanatomical development of the disease for a number of
disorders (Blass, 1993a, 1993b; Gibson et al., 1998; Hoyer, 1997).
Alterations in mitochondrial enzymes, arising from a genetic lesion
or resulting from the pathology of AD, are one of the factors that
characterize the AD metabolic deficit, including enzymes such as
pyruvate dehydrogenase complex,
-ketoglutarate dehydrogenase
complex and cytochrome oxidase (Gibson et al., 1998). Genetic
abnormalities in other genes including APP and the presenilins also
appear to lead to mitochondrial abnormalities in AD (Cruts et al.,
1998; Rubinsztein, 1997). In severe forms of AD and in age-related
sporadic late-onset dementia, significant reduction in glucose
metabolism has also been reported (Hoyer, 1996; Ishii et al., 1997).
Other reports of metabolic abnormalities in AD have described
abnormal lipid metabolism in Alzheimer’s disease patients (Klunk
et al., 1994; Pettegrew et al., 1988; Svennerholm and Gottfries,
1994; Kanfer et al., 1996). In a recent article Du and colleagues (Du
et al., 2008) studied mice lacking Cyclophilin D providing
substantial evidence that mitochondria serves as direct targets
for A
mediated neuronal toxicity.
The observations that A
progressively accumulates in cortical
mitochondria from AD patients and in brains from transgenic mouse
models of the disease suggest the role of mitochondrial A
in the
pathogenesisor development of the disease. Furthermore, reduction
in mitochondrial superoxide dismutase modulates AD-like pathol-
ogy and accelerates the onset of behavioural changes in human
amyloid precursor protein transgenic mice (Esposito et al., 2006).
This evidence together suggests that an approach like metabolomics
which focuses on metabolic changes in a pathology should prove to
be highly discriminatory for monitoring the progression of AD.
Several transgenic mouse models expressing human mutant
APP develop robust amyloid plaque pathology and dense core
plaques in an age dependant manner e.g. Tg2576 (K670N,M671L)
mice, TgCRND8(KM670/671NL + V717F) mice and PSAPP bigenic
transgenic mice (Chishti et al., 2001; Games et al., 1995; Hsiao
et al., 1996; Janus et al., 2001; Mucke et al., 2000). Despite
formation of age-dependent amyloid pathology none of these
transgenic mice developed widespread neuronal loss (Citron et al.,
1997; Irizarry et al., 1997) nor marked neurofibrillary tangle
pathology, although several of these models have tau positive
dystrophic neurites (see reviews by Borchelt et al. (1998) and
McGowan et al. (2006)). The TgCRND8 transgenic mouse expresses
a double mutant form of human APP 695 isoform (KM670/
671NL + V717F) under control of the hamster prion protein
promoter, and expresses amyloid deposits as early as 2–3 months
(Chishti et al., 2001). CRND8 mutant mice have high levels of A
by 6 months of age. With ageing, the mice have larger plaques,
multicentric dense-cored deposits and neuritic changes similar to
those seen in AD (Chishti et al., 2001).
In this study, we have implemented a metabolomic approach to
describe the metabolic perturbations that accompany neurode-
generation in various brain regions of the TgCRND8 mouse in an
age dependant manner. Given the profound changes that occur in
mitochondrial metabolism we hypothesized that metabolomics
would prove to be a highly profitably approach to the identification
of metabolic markers of the disease which may in the future by
used for diagnosis and/or assessment of AD disease progress.
2. Experimental procedures
2.1. Sample collection
Transgenic CRND8 APP 695 and their non-transgenic littermates (controls) were
maintained on a B6C3 hybrid strain background. Mice were weaned at 21 days
postnatally. The mice were maintained within a specific pathogen free barrier
facility on a 12-h light–dark cycle and were permitted free access to food and water.
Animal care was in accordance with IACUC (Institutional Animal Care and Use
Committee, USA). The genotype of the mice was established after birth and
confirmed after death by PCR analysis of genomic DNA extracted from tail-snips.
In total, 163 samples from 7 different brain regions and two age groups of young
(2–3 months) and old (12–13 months) were analysed. The young mice were all
female (4 control and 5 transgenic mice) and older animals were both male (3
control and 2 transgenic mice) and female (5 control and 3 transgenic mice). Mice
were cervically dislocated, brains rapidly removed and hemisected. One
hemisphere was immersion fixed in 10% formalin and then processed for paraffin
embedding using standard techniques. The other hemisphere was rapidly dissected
into cortex, frontal cortex, cerebellum, hippocampus, striatum, pons, midbrain and
olfactory bulb. Tissues were rapidly frozen on dry ice and stored at 80 8C.
2.2. Tissue extraction
Frozen brain tissue (50 mg; cortex, cerebellum and hippocampus, 20 mg;
striatum and midbrain 10 mg; pons and olfactory bulb) samples were ground on
dry ice using a pestle and mortar and transferred into micro-centrifuge tubes before
adding methanol and chloroform in a ratio of 2:1 (v/v; 600
l) (Le Belle et al., 2002).
Homogenates were sonicated for 15 min in a water bath, after which 200
l each of
chloroform and distilled water were added to form an emulsion. The phases were
separated by centrifugation, and the upper aqueous phase was removed and dried
in an evacuated centrifuge (Eppendorf, Hamburg, Germany).
H NMR spectroscopy
Brain tissue aqueous extracts were reconstituted in 600
O containing
50 mM sodium phosphate (pH 7.4) and 0.25 mM sodium (3-trimethylsilyl)-2,2,3,3-
tetradeuteriopropionate (TSP; as a chemical shift reference at 0 ppm) (Cambridge
Isotope Laboratories, Inc., USA). High resolution solution state
H NMR spectra were
recorded using a DRX700 Bruker Avance spectrometer, equipped with a 5 mm TXI
ATMA probe (Bruker BioSpin GmbH, Rheinstetten, Germany) at a proton frequency
of 699.9 MHz. All one-dimensional (1D) spectra were acquired at 27 8C with a
spectral width of 20.00 ppm using a conventional pre-saturation pulse sequence for
water suppression based on the first increment of the nuclear. Overhauser effect
spectroscopy (NOESY) pulse sequence [RD (relaxation delay)–
acquire, where t
= mixing time] (Macura and Huang, 1981). The water resonance
was irradiated during the relaxation delay (3 s) and mixing time (t
= 100 ms), with
fixed at 4
s. Each spectrum was acquired with 128 scans collected into 64k data
points with an acquisition time of 2.34 s. All spectra were processed using 1D NMR
Manager software (version 9, Advanced Chemistry Development Inc., Toronto,
Canada), zero-filled to twice the number of points and multiplied by an exponential
weighting function corresponding to a line broadening of 0.3 Hz, Fourier
transformed, phased, baseline corrected and referenced to TSP. Spectra were
segmented into 0.04 ppm chemical shift buckets (frequency windows) between 0.2
and 4.55 ppm (excluding water resonance and aromatic region of the spectra) using
the Intelligent Bucketing facility (a variable length bucket integration system
designed to ensure peaks do not straddle buckets). To account for differences in
sample amounts each integrated region was normalized to the total spectral area.
Spectra were assigned by comparison with previous literature (Govindaraju et al.,
2000; Lindon et al., 1999; Pears et al., 2005), two-dimensional NMR spectroscopy
and using Chenomix NMR Suite version 5.1 (Chenomix, Inc., 2007).
2.4. Chemometric analysis of the data
Multivariate pattern recognition techniques were used to analyse the NMR data
set, since this method is capable of handling multiple variables simultaneously and
to cope with numerous co-linearities and missing variables unlike univariate
approaches. This was carried out using SIMCA-P+ 11.5 (Umetrics AB, Umea
Sweden). Data were mean-centred and Pareto scaled prior to analysis. Pareto
scaling augments the representation of low concentration metabolites in statistical
models by dividing each variable by the square root of the standard deviation of the
variable, without overly increasing the contribution noise makes to the model.
Two multivariate statistical techniques were used within the SIMCA package;
principal components analysis (PCA) and projection to latent structures by partial
least squares discriminant analysis (PLS-DA). PCA is an unsupervised technique that
describes observations (in this study spectra from mouse brain tissue) with regards to
one or more latent variables termed principal components (PCs), which are linear
combinations of theoriginal variables(NMRbuckets). The techniquegenerates a set of
PCs to replace the original variables,each accountingfor decreasingproportions of the
total variation in the dataset. The information not captured in the first component
forms the residuals through which the second component is calculated and placed
orthogonal to the first PC. Each observation is then assigned a score along each PC.
Similar observations cluster together and are distinct from other clusters of
observations. The weight given to each original NMR variable within any given PC
describes how influential that particular variable is and the relation (or correlation) of
the variable to others. To determine which of the original variables are responsible for
this separation the loading scores of the PCs can be analysed (Jackson, 1991; Wold
et al., 1984). To determine which metabolites were most important for a given
classification only those that significantly contributed to a component as determined
by a jack-knifing procedure within SIMCA were considered.
When the total variation in the dataset is obscured, the variation between two
prescribed classes (e.g. wild-type and transgenic mice in this study) can be
R.M. Salek et al. / Neurochemistry International 56 (2010) 937–947
identified by a supervised pattern recognition approach such as PLS-DA. This
approach is similar to PCA but it uses variables discriminately that correlate to class
membership. In PLS-DA, the NMR data are set as an ‘X’ matrix with dummy
variables representing the class of each sample forming the ‘Y’ matrix. This allows
the analysis of the NMR variables which contribute most strongly to the separation
of the classes (Ho
skuldsson, 1996; Wold et al., 1984).
2.5. Model validation
The parameters R
and Q
were used to evaluate the performance of each model.
The R
score indicates how much of the total variation in the dataset is described by
a particular component and R
indicates the variation described by all the
components in the model (scored 0–1). Q
is a measure of how accurately the model
can predict class membership, and hence is more relevant to supervised
approaches. Q
estimates the predictive ability of the model by leaving out
observations (1/7th) from the building of the model and then predicting their class
membership or trend variable. Q
> 0.08 is indicative of a model that is better than
chance and scores above 0.7 demonstrate a highly robust trend or separation
(Eriksson et al., 1999). The validity and the degree of over fit for the models were
assessed by the criteria and validation routine built-in the SIMCA package.
2.6. Univariate data analysis
Univariate statistical tests, using Matlab (version 2007a, The Mathworks Inc.)
were also carried out on the bucket integrated NMR spectra, comparing the mutant
to the control mice for young and old mice in the hippocampus, cortex, frontal
cortex, midbrain and cerebellum brain regions. Each bucket was treated as an
independent variable. Statistically significant changes between the distributions of
the affected animal and the control group were assessed using a parametric
Student’s t-test (based on assumption of normal distribution) as well as the non-
parametric Kruskal–Wallis (non-parametric generalization of ANOVA test with no
assumption on the form of the distribution) and Kolmogorov–Smirnov (based on
comparison of distribution shape) tests.
2.7. Quantification of amyloid deposition
Hemibrains were immersed, fixed in 10% formalin and processed for paraffin
embedding. Brain tissue sections (5
m) were immunostained with anti-total A
antibody (33.1.1, 1:1000; a gift from T. Golde, Mayo Clinic) on a Dako (Glostrup,
Denmark) autostainer using standard techniques. Sections were counterstained
with hematoxylin. A
plaque burden in the hippocampus, piriform cortex and
cerebellum was determined using Meta-Morph image analysis software (Molecular
Devices, Palo Alto, CA). All of the above analyses were performed in a blinded
For more detailed histological analysis the CRND8 mice were crossed with YFP-H
mice (Feng et al., 2000) this generated transgenic mice in which yellow
fluorescent protein is selectively expressed in neurons to follow changes in
neuronal morphology. In addition fixed heterozygous CRND8/YFP-H mice brains
were also processed for A
immunoreactivity and stained with thioflavin S to
visualise plaques. Images were captured using a Leica TCS-NT-UV confocal
3. Results
3.1. Metabolites coverage by
H NMR spectroscopy
High resolution solution state
H NMR spectra were collected
from hippocampus, cortex, frontal cortex, cerebellum, midbrain,
pons, olfactory bulb and striatum from wild-type and APP695
mutant mouse. In each brain region, 23 metabolites were
detected (Fig. 1A shows an example NMR spectrum from the cortex
tissue of a 2 months mouse). Comparing all the spectra, the
dominant trend in the metabolomic dataset was associated with
brain region, with both PCA and PLS-DA analyses separating
different brain regions robustly, regardless of gender, age and
genotype of the mice. A PLS-DA model separated eight different
brain regions (not shown, Q
= 0.65, R
= 0.91 and A = 11) from
which cortical, cerebellar, midbrain and pons samples were
distinctly separated from hippocampal, olfactory bulb and striatal
samples along PLS-DA component 1 and from each other along
PLS-DA component 2. Other brain regions were also separated by
lower components (e.g. PLS-DA component 3 separated striatum,
hippocampus, cerebellum from cortex, midbrain and olfactory
bulb samples). A PLS-DA model using cortical, cerebellar and
olfactory bulb samples separated all three regions as shown in
Fig. 1B(Q
= 0.96). The most discriminatory metabolites in this
model were increased concentrations of taurine, GABA and
NAAG + NAG (N-acetyl-aspartyl-glutamate and N-acetyl-gluta-
mate; the two metabolites cannot be distinguished separately in
a 1D NMR spectrum and so the relevant resonance is labelled with
both metabolites) in the olfactory bulb, increased concentrations
of creatine + phosphocreatine and myo-inositol in the cerebellum
and increased concentrations of lactate, glutamate and alanine in
the cortex. Fig. 1C shows a PLS-DA model that separates samples
from the hippocampal, olfactory bulb cerebellum and midbrain
= 0.96). The metabolites contributing to the separation of the
brain region in this model are increased concentrations of taurine,
NAA, NAAG + NAG and GABA in the olfactory bulb, increased
concentrations of glutamate, glutamine, and creatine in hippo-
campus and increased concentrations of lactate and myo-inositol
in the midbrain region. Given the large differences in metabolic
profiles from different tissues in subsequent analyses the
individual brain regions were considered separately.
3.2. Gender and age differences
Both PCA and PLS-DA models were built to model gender
differences. No female and male separation was observed using
multivariate data analysis when the entire dataset was chosen or
during comparisons of equal numbers of male and female samples
within the same age group (12–13 months). Similarly, PLS-DA for
each brain region failed to separate male from female for either the
entire group or the 12–13 months group, hence all male and female
samples were considered together in subsequent analyses.
Metabolic profiles of tissues from young mice (2–3 months)
could be separated from that from older mice (12–13 months)
using both PCA and PLS-DA, regardless of the genotype for each
brain region (e.g. Fig. 2A, separation of the young from older mice
for the hippocampus brain region). Metabolites that discriminated
between old and young mouse tissue, regardless of genotype, were
lactate, alanine, lysine and N-acetyl-aspartyl-glutamate (NAAG)
which decreased with age and glutamine, creatine, myo-inositol,
and malonate which increased as mice aged.
To examine how ageing affected the metabolic profile of the
transgenic mice, this was compared in the young and old groups in
different brain regions and then cross-compared with the changes
detected in control mice. The loadings plots (and hence metabo-
lites) discriminating the wild-type age differences and APP 695
mutant age differences for different brain regions were distinctly
different (Fig. 2B and C). For example, the most important
metabolites discriminating the cortex of ageing control mice were
increased creatine, taurine and malonate and decreased NAAG,
alanine and lysine. In the cortex of the transgenic mice lactate,
myo-inositol, aspartate and glutamine were increased while NAAG
and glutamate were decreased in the aged mice. Similar
differences were also observed for the hippocampus and midbrain
regions with regard to genotype ageing profiles. The decrease in
NAAG + NAG was consistent for the mice ageing for both genotypes
for the brain regions measured. Thus, this suggests that while there
were some similarities between the metabolic consequences of
ageing, differences occurred between the two different genotypes
as the mice aged as would be expected for a progressive disorder
such as AD.
3.3. TgCRND8 APP 695 mutant model
In order to prevent over-fitting of the pattern recognition
models due to the small sample size for each brain region, a two-
tier approach was used. Firstly, both ages were modelled together
to compare control and mutant mice. Then individual time points
were considered for each brain region. The
H NMR spectral
R.M. Salek et al. / Neurochemistry International 56 (2010) 937–947
profiles derived from control and CRND8 transgenic mice samples
were distinguished using both PCA and PLS-DA analysis for the
hippocampus, cortex, frontal cortex, midbrain and cerebellar
regions (Fig. 3). All models built were generally highly robust and
predictable (0.4 < Q
< 0.80 for PLS-DA models using either one or
2 components) with hippocampus, frontal cortex and cortex being
the most robust models (0.50 < Q
). While separation was not
detected for the striatum, pons and olfactory bulb, it should be
noted that the size of samples was relatively low and therefore
spectral quality was reduced. Table 1a summarises the multivari-
ate data analysis results for the metabolites that were altered in the
individual brain regions of Tg mice, relative to control mice for both
young and older mice as well as the combined age models. In
addition, Table 1b summarises the univariate data analysis results
for the affected brain regions and the metabolites found to be
NAA was relatively decreased in the hippocampus, cortex,
frontal cortex and midbrain of the affected mice compared to the
wild-type. This was such a dominant trend that it was visually
apparent in the spectra. Fig. 4 shows the reduced concentration of
NAA in the cortex of a 12 months affected mouse compared to the
control animal at the same age. NAG + NAAG was also decreased in
concentration in the hippocampus as well as the midbrain and
cerebellum of the older mice.
There were marked decreases in glutamate and glutamine
across the hippocampus, cortex, frontal cortex and mid brain but
not the cerebellum, and these changes were consistent with the
alterations in NAA concentration. Although not as consistent
across the age ranges examined GABA also decreased in
concentration across the different brain regions examined. The
concentrations of taurine and creatine were also reduced in the
cortex and frontal cortex of the older mice as well as hippocampus.
For a number of other amino acids, including aspartate, leucine,
iso-leucine, valine, arginine, lysine, alanine and glycine the broad
trend was for increased concentrations in the mutant mouse.
Lactate was markedly increased in all brain regions and at both
time points. Short and medium chain free fatty acid partly soluble
in water, represented by the spectral region between 0.85–0.90
and 1.25–1.3 ppm chemical shifts and corresponding to terminal
groups and the long chain CH
groups of fatty acids
respectively, showed an increase in concentration in all the
affected brain regions with the exception of the cerebellum where
it was decreased.
The remaining changes in metabolites between wild-type and
mutant were more age and region specific and included myo-
inositol which was relatively decreased in the midbrain and
increased in cortical brain regions from young transgenic mice.
Choline and phosphocholine concentrations were decreased in all
Fig. 1. (A) High resolution solution state
H NMR spectrum, 700 MHz, of an aqueous extract of cortex tissue taken from a 2-month-old control mouse. Peak 1, valine, leucine,
isoleucine; peak 2, fatty acids; peak 3, lactate; peak 4, alanine; peak 5,
-amino butyric acid (GABA); peak 6, acetate; peak 7, N-acetyl-
-aspartate (NAA); peak 8, glutamate; peak
9, N-acetyl-aspartyl-glutamate (NAAG); peak 10, glutamate and glutamine; peak 11, glutamine; peak 12, succinate; peak 13, aspartate; peak 14, creatine/phosphor-creatine;
peak 15, malonate; peak 16, ethanolamine; peak 17, choline; peak 18, phosphocholine; peak 19, taurine and myo-inositol; peak 20, scyllo-myo-inositol; peak 21, taurine; peak 22,
myo-inositol; peak 23, glycine peak; peak 24, O-Phosphoethanolamide. (B) PLS-DA score plot separation of cortex (*), cerebellum (~) and olfactory bulb (
= 0.96,
= 0.76 and A (PC Components fitted) = 3]. (C) PLS-DA score plot separation of hippocampus (!), midbrain (&) and olfactory bulb (^)[Q
= 0.96, R
=0.78andA = 4].
R.M. Salek et al. / Neurochemistry International 56 (2010) 937–947
Fig. 2. Age-related changes in the metabolic profiles of tissue from different brain regions. (A) A PLS-DA model separating young (*) form old (^) mice, regardless of the
genotype, for the hippocampus (SD = standard deviation from the mean, Q
= 0.54 for the model). Comparison of the metabolic profiles as represented by the loadings plot of
PLS-DA models separating young and old mice for the cortex (B) and midbrain (C) for wild-type and mutant mice considered separately. Key: transgenic mice (grey long dash);
wild-type (black solid line). The x-axis is the chemical shift of an integral region and the y-axis is the weighted coefficient loading for that region for PLS-DA component 1.
Clear differences are apparent between these two loadings scores.
R.M. Salek et al. / Neurochemistry International 56 (2010) 937–947
brain regions (for the hippocampus only in older mice) in the
transgenic mice with the exception of the increase in choline in the
midbrain of young mutant mice.
3.4. Histopathology results
Amyloid deposition in the hippocampus, piriform cortex and
cerebellum for the TgCRND8 APP659 mice for both older and
younger aged mice were evaluated with the results summarised in
Table 2. The amyloid burden was determined by calculating the
total percentage of tissue area taken up by amyloid deposition. The
highest immunoreactivity and plaque density were observed in the
posterior and the anterior of the hippocampal formation followed
by the anterior piriform cortex and to a less extent in the posterior
piriform cortex. A relatively weaker signal was observed in the
cerebellum brain region. The amyloid plaque burden analysis
showed increased plaques in the older TgCRND8 compared with
few plaques in the 2–3-month-old mice (Fig. 5A–D). The large
Fig. 3. (A) Multivariate data analysis of the cortex separates wild-type (*) from the transgenic mice samples (&)[Q
= 0.54, A = 2] irrespective of ages and gender. (B) The
relevant loading plots for the cortex model shown as a pseudo-NMR spectrum with each point shown in colour corresponding to the weight of the loading in the model. This
represents the correlation of the NMR loading variable with the discrimination between effected and the control mouse with positive values being metabolites that are
relatively increased in the APP 695 mouse and negative values decreased (Cloarec et al., 2005). (C) Hippocampus separation of wild-type (~) from transgenic (
= 0.60,
A = 2] all models include both ages. (D) The relevant pseudo-NMR spectra loading plots for the hippocampus model.
Table 1a
Summary of the multivariate data analysis of the relative metabolite changes for the TgCRND8 APP 695 mutants compared to wild-type in five affected brain regions (for the
combined and separated age groups). The Q
for each model is show at the bottom of the table.
Relative metabolites
change in TgCRND8
Free fatty acids """"""" " " """###
Leu-ile/val """ –––– "" # #
Lactate """"""" "
" """""
Leu/lys/Arg """"""" " """# #
Alanine """ " " # #"" ##
GABA ## # # ––### ##
NAA ##
Glutamate ##
## #
Glutamine ###
# ######
# ––
Creatine/P-creatine ### # # ## ### #"
Taurine #### # ##––
–––– ## ##
myo-Inositol # #
"# " # ###
Glycine " """"" " ###––
Aspartate " """"" " " """""
Choline/P-choline –/# ––/# –/# –/# –/# –/# –/## –/#" # –/# –/#
Succinate ###
#### # # –––###
Serine ""––––" ––
Malonate ## ## ## ## ##
Model Q
0.64 0.7 0.57 0.54 0.47 0.42 0.74
0.8 0.5 0.53
0.68 0.53 0.4 0.24 0.21
Two component used for the model; P-creatine = phosphocreatine, P-choline = phosphocholine.
Significant change in multivariate analysis.
R.M. Salek et al. / Neurochemistry International 56 (2010) 937–947
amyloid plaques contained dystrophic axons and dendrites
amongst the A
immunoreactive and thioflavin S stained material
(Fig. 5E). Other hallmarks include loss of vesicles and disruption to
the synaptic transmission mechanism.
4. Discussion
In order for animal models to contribute to the understanding
and treatment of AD, it is vital that they exhibit specific symptoms
or pathology of AD. In this metabolomic study of an animal model
of A
deposition, we have focussed on the CRND8 transgenic
mouse expressing APP695 with both the Swedish and Indiana
mutations. In this mouse, formation of elevated A
42 levels and
development of plaques as early as 2–3 months of age has been
reported, although no neuronal cell loss is apparent even in older
mice (Chishti et al., 2001; Janus et al., 2001). The TgCRND8 mice
exhibit a high basal synthesis of the concentration of A
, skewed
towards the production of A
42, which has been suggested to
contribute to early onset of amyloid deposition in the hippocam-
pus and cerebral cortex. This is rarely found in the striatum of
young TgCRND8 mice, although older CRND8 have striatal
pathology (Woodhouse et al., 2009). With ageing, the plaques
become larger and at about 5 months of age the mature plaques in
the TgCRND8 mice have associated dystrophic neurites throughout
the hippocampus, cortex, corpus callosum, and in some cases in the
striatum of TgCRND8 mice (Chishti et al., 2001; Dudal et al., 2004),
similar to aged human sufferers of AD (Woodhouse et al., 2009).
The aged TgCRND8 mice exhibit amyloid angiopathy, in addition to
plaque consistent with the similar observation in the neocortex
of human AD cases (Adlard and Vickers, 2002). In addition it has
previously been recorded that TgCRND8 mice at 11 weeks show
significant impairment in the acquisition of spatial information
relative to non-transgenic littermates (Chishti et al., 2001).
In the present study, our results demonstrated that both
histologically and metabolically, the hippocampus and cortical
regions were affected in the mutant mice with an increase in the
Table 1b
Summary of the univariate data analysis for the Student’s t-test (t-test), Kruskal–Wallis (KW-test) and Kolmogorov–Smirnov (KS-test) tests carried out for the TgCRND8 APP
695 mutants compared to wild-type in the five affected brain regions. The majority of the results correspond to the multivariate analysis results in Table 1a.
Brain regions Hippocampus old Hippocampus young Cortex old Cortex young Frontal cortex old
Univariate stats t-test KS-test KW-test t-test KS-test KW-test t-test KS-test KW-test t-test KS-test KW-test t-test KS-test KW-test
GABA 0.05 0.01 0.02
NAA 0.004 0.01 0.02 0.03 0.13 0.05 0.02 0.02 0.03 0.01 0.05 0.02
Glutamate 0.002 0.01 0.03 0.02 0.05 0.02 0.05 0.02 0.01 0.02
Glutamine 0.01 0.02 0.004 0.006 0.01 0.05 0.05
Cr/PCr 0.006 0.01 0.02 0.02 0.05 0.03
Taurine 0.02 0.05
myo-Inositol 0.05 0.05 0.006 0.01 0.02
Brain regions Frontal cortex young Midbrain old Midbrain young Cerebellum old Cerebellum young
Univariate stats t-test KS-test KW-test t-test KS-test KW-test t-test KS-test KW-test t-test KS-test KW-test t-test KS-test KW-test
GABA 0.1 0.04 0.05 0.04
NAA 0.04 0.03
Glutamate 0.02 0.04 0.001 0.008 0.016
Glutamine 0.04 0.008 0.01 0.05 0.05 0.05
Cr/PCr 0.001 0.008 0.016 0.05
Taurine 0.01 0.03 0.03
myo-Inositol 0.04 0.03 0.05 0.02 0.003 0.003 0.008
Table 2
Plaque burden analysis percentage in older mice and numbers of plaques per section observed in younger mice. All the other brains regions had no pathology.
APP 695 mutant mice Anterior hippocampus Posterior hippocampus Anterior piriform cortex Posterior piriform cortex Cerebellum
Male 13m 3.81% 4.94% 2.72% 1.96% 0.44%
Male 13m 4.72% 4.73% 3.50% 5.81% 0.41%
Female 12m 3.81% 3.76% 3.03% 2.29% 0.29%
Female 3m
31 12 4 0
Female 3m
33 6 8 0
Female 3m
00 0 2 0
Female 3m
31 7 4 0
Female 3m
13 7 6 0
Number of plaques per section since there were too few for plaques for burden analysis in the young mice.
Fig. 4. Comparisons of NMR spectra from cortex of a control mouse (I) to cortex of
the APP 695 mutant mouse (II) both at 12 months of age. Decrease in the NAA peak,
highlighted using the arrows. Both spectra were normalized to total spectral area in
order to compensate for potential difference in the sample size.
R.M. Salek et al. / Neurochemistry International 56 (2010) 937–947
severity as the mice aged. In addition, two other brain regions,
midbrain and cerebellum, were also partly affected, especially in
older mice, with a metabolic pattern similar to that of hippocampal
and cortical regions being detected in the midbrain, but the
cerebellum having quite distinct metabolic changes in comparison
to other brain regions. No significant difference between the
striatum of the control mice compared to the affected mutant was
detected (combined model Q
< 0.17), which is likely to reflect the
relatively small sample size of the striatum brain region, and hence
reduced spectral quality for low concentration metabolites.
One of the most noticeable changes in the dataset was a
decrease in NAA in the hippocampus, cortex and frontal cortex,
with detectable changes even in young transgenic mice compared
to the control animals. NAA, under normal conditions, is
synthesized in the mitochondria of neurons, but not in glial cells
(Baslow, 2003). Decreases in NAA of the cortical brain region
(frontal and the mesial temporal) of patients with AD compared to
healthy elderly individuals has been reported numerously during
in vivo
H magnetic resonance spectroscopy/imaging studies
(MRS) (Chen et al., 2000; Hancu et al., 2005; Huang et al., 2001;
Fig. 5. Histology of the A
deposition in the TgCRND8 APP695 young and old transgenic mice. (A) Cerebral cortex of a 3-month-old female mouse, with an occasional stained
plaque (arrowhead). (B) Cerebral cortex plaque from a 13-month-old male mouse note the numerous A immunoreactive plaques (arrowheads). (C) A immunoreactive plaques
present in the hippocampus of a male mouse at 13 months of age (arrowheads). (D) Comparison of coronal sections at different levels through the TgCRND8 APP695 mouse
brain. Note the striking accumulation of A
immunoreactivity (brown peroxidase positive material) in the 13-month-old mouse (i) as compared to a 3-month-old mouse (ii).
immunoreactivity is particularly concentrated in the hippocampus and cerebral cortex with much lower amounts of A
immunoreactive material in thalamus, striatum
and cerebellum. Abbreviations used: DG, dentate gyrus; CA1, CA1 region of hippocampus. Scale bars: A–C 200
m; D, 0.5 cm. (E) Confocal image of the hippocampal pyramidal
cell layer of a 5-month-old YFP-H/CRND8 mouse. Note the strong A
-immunoreactive (red) and thioflavin S stained plaques. Arrowheads indicate the axons of the pyramidal
cells and arrows the dendrites. The A
immunoreactive material accumulates particularly amongst the dendritic trees of the pyramidal cells. Scale bar 20 m. (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
R.M. Salek et al. / Neurochemistry International 56 (2010) 937–947
Jessen et al., 2005; Jones and Waldman, 2004; Kantarci et al., 2002)
and were shown to correlate with recognition memory in health,
and NFTs in Alzheimer’s disease patients (Mohanakrishnan et al.,
1995). In many previous studies the concentration of NAA has been
reported as either a ratio to creatine, myo-inositol or choline, even
though any of these metabolites can change with age or disease,
and ignores the tissue composition of the regions measured,
despite differences in NAA concentration in grey matter and white
matter (Petroff et al., 1995; Pfefferbaum et al., 1999). In the present
study, NAA is represented as a ratio to the combined sum of
metabolites present in the tissue extracts. This approach is rapid
but is also robust against changes not being detected in a ratio
because two metabolites change in a similar direction (as occurs in
the present study for NAA and creatine).
Furthermore, the decrease in cortical NAA, characteristic of the
transgenic mouse was apparent at 2–3 months of age, prior to any
detectable neuronal cell loss detectable by histology. This, coupled
with decrease in the concentration of Cr + PCr suggests that NAA
may not be a marker for intact neuronal cells and total neuronal
cell density per se, but instead reflects the energy status of
neurones, providing an early stage marker of dysfunction.
Similarly, a reduced concentration of NAA, in the absence of
neuronal death but with the presence of nuclear inclusion bodies
inside neurones has been reported in the transgenic mouse model
of Huntington’s disease (Jenkins et al., 2000). In another transgenic
APP mutant mice, the APP Tg2576, in vitro and in vivo magnetic
resonance spectroscopy showed that NAA, glutamate and gluta-
thione were decreased while taurine was increased in the cerebral
cortex of APP transgenic mice at 19 months of age when there is
widespread A
deposits with no significant neuronal loss
(Dedeoglu et al., 2004). However, these were in mice much older
than those examined in the present study. Similarly, in a separate
transgenic mouse model (APP-PS1) with A
mutant co-expressing
with human presenilin 1 mutation, a reduction in NAA and
glutamate and an increase in myo-inositol compared to total
concentration of creatine was reported in aged mice in a
study of mice aged from 2 to 30 months (Marjanska et al., 2005).
The concentration of myo-inositol was decreased relatively in
the cortical region of the young transgenic mice, but increased as
the transgenic mice aged (this was most significant in the frontal
cortex). A decrease in NAA and an increase in myo-inositol both
occur during neuronal cell loss/dysfunction and the associated
gliosis, and these changes occurred in anatomic distributions
which reflect pathological involvement and atrophy patterns as
reported in AD (Meyerhoff et al., 1993; Rothman et al., 2003).
However, in the present study there is no evidence of neuronal cell
loss and that subsequent gliosis has occurred and thus our changes
most likely reflect early stage neuronal dysfunction and glial
activation, and also suggest that a simple ratio of NAA/myo-inositol
cannot be used as a proxy to neuronal cell loss and gliosis.
The concentration of glutamate in the brain is controlled by the
balance between neuronal and glial cell metabolism as part of the
glutamate/glutamine cycle between these cell types (Attwell,
1994; Fonnum, 1984). Glutamate is also the immediate precursor
for synthesis of GABA in neurons (Schousboe et al., 1997;
Sonnewald and Kondziella, 2003). Glutamate induced excitotoxi-
city may play a role in acute neuronal death and in the
pathogenesis of chronic neurodegenerative disorders such as AD
(Hugon et al., 1996). Decreases in glutamate and glutamine ratio in
AD patients have been detected in vivo using
H MRS (Antuono
et al., 2001). However, it is difficult to resolve these metabolites in
in vivo MRS spectroscopy, since the glutamate and glutamine
resonance in MRS spectroscopy are usually overlapped. Using
solution based
H NMR we discriminated between these two
metabolites, demonstrating a decreased concentration of gluta-
mate and glutamine in all affected brain regions, except the
cerebellum. The most significant changes for glutamate were in the
hippocampus and the cortex, whereas glutamine decreased most
in the hippocampus of young animals and midbrain of old mice
(see Tables 1a and 1b). GABA was also reduced in all of the effected
brain regions with similar reports observed in the cortex and
frontal cortex of the AD patients (Garcia-Alloza et al., 2006; Lanctot
et al., 2007; Mohanakrishnan et al., 1995). However, other reports
have suggested little net change in the aged and diseased brain for
the GABA in the AD tissue specially in the hippocampus (Rissman
et al., 2007). Our study detected net decreases in glutamate,
glutamine and GABA which indicate that excitotoxicity is unlikely
to be a major part of the pathology of this particular transgenic
mouse, although steady state total concentrations of metabolites
do not determine the localisation of amino acids.
Taurine is highly concentrated in rodent brain compared to
humans and acts as an osmoregulatory, antioxidant and neuro-
modulator (Burg et al., 1997; Huxtable, 1989). In AD patients a
decrease in the concentration of taurine has been observed in the
cortex and cerebrospinal fluid. Similarly, in this study we observed
a relative decrease in the concentration of taurine in the frontal
cortex and midbrain of the APP 695 transgenic mice compared the
The interpretation of the increased concentrations of lactate
detected in all brain regions of transgenic mice requires some
caution in terms of interpretation. A large proportion of this lactate
is produced post-mortem during the 30–40 s required to dissect
out tissue, prior to freezing. Thus, the increase in lactate may
represent an increase in brain glucose or glycogen in vivo, with
these metabolites rapidly being metabolized to lactate as part of
anaerobic glycolysis. Importantly, since dissection rates were the
same for control and mutant animals, the detected difference in the
concentration of lactate did not represent a post-mortem delay in
the sampling of one group. It should also be noted that the
observed low concentration of acetate, derived from the break-
down of NAA, is an indicator of minimal post-mortem sample
degradation, although the degradation rate of NAA is much slower
than that of glucose or glycogen in the brain (McIlwain and
Bachelard, 1985).
In this study, metabolomics enabled us to compare the brain
metabolic profile of an animal model of AD for different brain
regions, investigating the impact of disease on regions not primarily
associated with AD. We detected early changes in the metabolic
profile of affected brain regions before accumulation of plaques and
major histology changes and at an early stage of the disease. These
changes could be readily compared with observations in human AD
patients. Finally, early detection of metabolic changes related to AD
might be beneficial at a stage when substantial behavioural or
locomotor activity changesusingstandard testshavenot occurred in
animal models, or indeed patients where a definitive antemortem
diagnosis is not yet available.
This work was supported by the NIH grant R21 DK070288-01,
USA (RMS & JLG), The Royal Society, UK (a University Research
Fellowship to JLG) and NIA R01AG020216-01A2 to EM. We thank
Monica Castanedes-Casey, Linda Rousseau and Virginia Phillips for
expert histology. We thank Dr. David Westaway for supplying
breeding pairs of the CRND8 mice and Dr Robert Adalbert for the
provision of Fig. 5E. Finally we thank, Drs Denis Rubtsov and Aalim
Weljie for the Matlab source codes.
Adlard, P.A., Vickers, J.C., 2002. Morphologically distinct plaque types differentially
affect dendritic structure and organisation in the early and late stages of
Alzheimer’s disease. Acta Neuropathol. 103, 377–383.
R.M. Salek et al. / Neurochemistry International 56 (2010) 937–947
Antuono, P.G., Jones, J.L., Wang, Y., Li, S.J., 2001. Decreased glutamate + glutamine in
Alzheimer’s disease detected in vivo with
H MRS at 0.5 T. Neurology 56, 737–
Attwell, D., 1994. Glia and neurons in dialogue. Nature 369, 707–708.
Baslow, M.H., 2003. N-acetylaspartate in the vertebrate brain: metabolism and
function. Neurochem. Res. 28, 941–953.
Blass, J.P., 1993a. Pathophysiology of the Alzheimer’s Syndrome. Lippincott Wil-
liams & Wilkins, Hagerstown, MD, pp. S25–S38, ETATS-UNIS.
Blass, J.P., 1993b. Metabolic alterations common to neural and non-neural cells in
Alzheimer’s disease. Hippocampus 3 (Spec No), 45–53.
Borchelt, D.R., Wong, P.C., Sisodia, S.S., Price, D.L., 1998. Transgenic mouse models of
Alzheimer’s disease and amyotrophic lateral sclerosis. Brain Pathol. 8, 735–757.
Braak, H., Braak, E., 1991. Neuropathological stageing of Alzheimer-related changes.
Acta Neuropathol. (Berl.) 82, 239–259.
Burg, M.B., Kwon, E.D., Kultz, D., 1997. Regulation of gene expression by hyperto-
nicity. Annu. Rev. Physiol. 59, 437–455.
Chen, J.G., Charles, H.C., Barboriak, D.P., Doraiswamy, P.M., 2000. Magnetic reso-
nance spectroscopy in Alzheimer’s disease: focus on N-acetylaspartate. Acta
Neurol. Scand. Suppl. 176, 20–26.
Chishti, M.A., Yang, D.S., Janus, C., Phinney, A.L., Horne, P., Pearson, J., Strome, R.,
Zuker, N., Loukides, J., French, J., Turner, S., Lozza, G., Grilli, M., Kunicki, S.,
Morissette, C., Paquette, J., Gervais, F., Bergeron, C., Fraser, P.E., Carlson, G.A.,
George-Hyslop, P.S., Westaway, D., 2001. Early-onset amyloid deposition and
cognitive deficits in transgenic mice expressing a double mutant form of
amyloid precursor protein 695. J. Biol. Chem. 276, 21562–21570.
Citron, M., Oltersdorf, T., Haass, C., McConlogue, L., Hung, A.Y., Seubert, P., Vigo-
Pelfrey, C., Lieberburg, I., Selkoe, D.J., 1992. Mutation of the beta-amyloid
precursor protein in familial Alzheimer’s disease increases beta-protein pro-
duction. Nature 360, 672–674.
Citron, M., Westaway, D., Xia, W., Carlson, G., Diehl, T., Levesque, G., Johnson-Wood,
K., Lee, M., Seubert, P., Davis, A., Kholodenko, D., Motter, R., Sherrington, R.,
Perry, B., Yao, H., Strome, R., Lieberburg, I., Rommens, J., Kim, S., Schenk, D.,
Fraser, P., St George Hyslop, P., Selkoe, D.J., 1997. Mutant presenilins of Alzhei-
mer’s disease increase production of 42-residue amyloid beta-protein in both
transfected cells and transgenic mice. Nat. Med. 3, 67–72.
Cloarec, O., Dumas, M.E., Craig, A., Barton, R.H., Trygg, J., Hudson, J., Blancher, C.,
Gauguier, D., Lindon, J.C., Holmes, E., Nicholson, J., 2005. Statistical total
correlation spectroscopy: an exploratory approach for latent biomarker iden-
tification from metabolic 1H NMR data sets. Anal Chem 77, 1282–1289.
Cruts, M., van Duijn, C.M., Backhovens, H., Van den Broeck, M., Wehnert, A., Serneels,
S., Sherrington, R., Hutton, M., Hardy, J., St George-Hyslop, P.H., Hofman, A., Van
Broeckhoven, C., 1998. Estimation of the genetic contribution of presenilin-1
and -2 mutations in a population-based study of presenile Alzheimer disease.
Hum. Mol. Genet. 7, 43–51.
Dedeoglu, A., Choi, J.K., Cormier, K., Kowall, N.W., Jenkins, B.G., 2004. Magnetic
resonance spectroscopic analysis of Alzheimer’s disease mouse brain that
express mutant human APP shows altered neurochemical profile. Brain Res.
1012, 60–65.
Du, H., Guo, L., Fang, F., Chen, D., Sosunov, A.A., McKhann, G.M., Yan, Y., Wang, C.,
Zhang, H., Molkentin, J.D., Gunn-Moore, F.J., Vonsattel, J.P., Arancio, O., Chen,
J.X., Yan, S.D., 2008. Cyclophilin D deficiency attenuates mitochondrial and
neuronal perturbation and ameliorates learning and memory in Alzheimer’s
disease. Nat. Med. 14, 1097–1105.
Dudal, S., Krzywkowski, P., Paquette, J., Morissette, C., Lacombe, D., Tremblay, P.,
Gervais, F., 2004. Inflammation occurs early during the Abeta deposition
process in TgCRND8 mice. Neurobiol. Aging 25, 861–871.
Eriksson, L., Johansson, E., Kettaneh-Wold, N., Wold, S., 1999. Introduction to Multi-
and Megavariate Data Analysis Using Projection Methods (PCA and PLS).
Umetrics, Umea, Sweden.
Esposito, L., Raber, J., Kekonius, L., Yan, F., Yu, G.Q., Bien-Ly, N., Puolivali, J., Scearce-
Levie, K., Masliah, E., Mucke, L., 2006. Reduction in mitochondrial superoxide
dismutase modulates Alzheimer’s disease-like pathology and accelerates the
onset of behavioral changes in human amyloid precursor protein transgenic
mice. J. Neurosci. 26, 5167–5179.
Feng, G., Mellor, R.H., Bernstein, M., Keller-Peck, C., Nguyen, Q.T., Wallace, M.,
Nerbonne, J.M., Lichtman, J.W., Sanes, J.R., 2000. Imaging neuronal subsets in
transgenic mice expressing multiple spectral variants of GFP. Neuron 28, 41–51.
Fonnum, F., 1984. Glutamate: a neurotransmitter in mammalian brain. J. Neuro-
chem. 42, 1–11.
Games, D., Adams, D., Alessandrini, R., Barbour, R., Berthelette, P., Blackwell, C., Carr,
T., Clemens, J., Donaldson, T., Gillespie, F., et al., 1995. Alzheimer-type neuro-
pathology in transgenic mice overexpressing V717F beta-amyloid precursor
protein. Nature 373, 523–527.
Garcia-Alloza, M., Tsang, S.W., Gil-Bea, F.J., Francis, P.T., Lai, M.K., Marcos, B., Chen,
C.P., Ramirez, M.J., 2006. Involvement of the GABAergic system in depressive
symptoms of Alzheimer’s disease. Neurobiol. Aging 27, 1110–1117.
Gibson, G.E., Sheu, K.F.R., Blass, J.P., 1998. Abnormalities of mitochondrial enzymes
in Alzheimer disease. J. Neural Transm. 105, 855–870.
Goate, A., Chartier-Harlin, M.C., Mullan, M., Brown, J., Crawford, F., Fidani, L., Giuffra,
L., Haynes, A., Irving, N., James, L., et al., 1991. Segregation of a missense
mutation in the amyloid precursor protein gene with familial Alzheimer’s
disease. Nature 349, 704–706.
Govindaraju, V., Young, K., Maudsley, A.A., 2000. Proton NMR chemical shifts and
coupling constants for brain metabolites. NMR Biomed. 13, 129–153.
Haass, C., Lemere, C.A., Capell, A., Citron, M., Seubert, P., Schenk, D., Lannfelt, L.,
Selkoe, D.J., 1995. The Swedish mutation causes early-onset Alzheimer’s disease
by beta-secretase cleavage within the secretory pathway. Nat. Med. 1, 1291–
Hancu, I., Zimmerman, E.A., Sailasuta, N., Hurd, R.E., 2005. 1H MR spectroscopy
using TE averaged PRESS: a more sensitive technique to detect neurodegenera-
tion associated with Alzheimer’s disease. Magn. Reson. Med. 53, 777–782.
Hardy, J., 2006. Alzheimer’s disease: the amyloid cascade hypothesis: an update and
reappraisal. J. Alzheimers Dis. 9, 151–153.
Hardy, J., Duff, K., Hardy, K.G., Perez-Tur, J., Hutton, M., 1998. Genetic dissection of
Alzheimer’s disease and related dementias: amyloid and its relationship to tau.
Nat. Neurosci. 1, 355–358.
Hardy, J.A., Higgins, G.A., 1992. Alzheimer’s disease: the amyloid cascade hypothe-
sis. Science 256, 184–185.
Hebert, L.E., Scherr, P.A., Bienias, J.L., Bennett, D.A., Evans, D.A., 2003. Alzheimer
disease in the US population: prevalence estimates using the 2000 census. Arch.
Neurol. 60, 1119–1122.
skuldsson, A., 1996. Prediction Methods in Science and Technology. Thor Pub-
lishing, Copenhagen.
Hoyer, S., 1996. Oxidative metabolism deficiencies in brains of patients with
Alzheimer’s disease. Acta Neurol. Scand. Suppl. 165, 18–24.
Hoyer, S., 1997. Models of Alzheimer’s disease: cellular and molecular aspects. J.
Neural Transm. Suppl. 49, 11–21.
Hsiao, K., Chapman, P., Nilsen, S., Eckman, C., Harigaya, Y., Younkin, S., Yang, F., Cole,
G., 1996. Correlative memory deficits, Abeta elevation, and amyloid plaques in
transgenic mice. Science 274, 99–102.
Huang, W., Alexander, G.E., Chang, L., Shetty, H.U., Krasuski, J.S., Rapoport, S.I.,
Schapiro, M.B., 2001. Brain metabolite concentration and dementia severity in
Alzheimer’s disease: a (1)H MRS study. Neurology 57, 626–632.
Hugon, J., Vallat, J.M., Dumas, M., 1996. Role of glutamate and excitotoxicity in
neurologic diseases. Rev. Neurol. (Paris) 152, 239–248.
Huxtable, R.J., 1989. Taurine in the central nervous system and the mammalian
actions of taurine. Prog. Neurobiol. 32, 471–533.
Irizarry, M.C., McNamara, M., Fedorchak, K., Hsiao, K., Hyman, B.T., 1997. APPSw
transgenic mice develop age-related A beta deposits and neuropil abnormali-
ties, but no neuronal loss in CA1. J. Neuropathol. Exp. Neurol. 56, 965–973.
Ishii, K., Sasaki, M., Kitagaki, H., Yamaji, S., Sakamoto, S., Matsuda, K., Mori, E., 1997.
Reduction of cerebellar glucose metabolism in advanced Alzheimer’s disease. J.
Nucl. Med. 38, 925–928.
Jackson, J.E., 1991. A User’s Guide to Principal Components. John Wiley, New York.
Janus, C., Phinney, A.L., Chishti, M.A., Westaway, D., 2001. New developments in
animal models of Alzheimer’s disease. Curr. Neurol Neurosci. Rep. 1, 451–457.
Jenkins, B.G., Klivenyi, P., Kustermann, E., Andreassen, O.A., Ferrante, R.J., Rosen, B.R.,
Beal, M.F., 2000. Nonlinear decrease over time in N-acetyl aspartate levels in the
absence of neuronal loss and increases in glutamine and glucose in transgenic
Huntington’s disease mice. J. Neurochem. 74, 2108–2119.
Jessen, F., Traeber, F., Freymann, N., Maier, W., Schild, H.H., Heun, R., Block, W., 2005. A
comparative study of the different N-acetylaspartate measures of the medial
temporal lobe in Alzheimer’s disease. Dement. Geriatr. Cogn. Disord. 20, 178–183.
Jones, R.S., Waldman, A.D., 2004.
H MRS evaluation of metabolism in Alzheimer’s
disease and vascular dementia. Neurol Res. 26, 488–495.
Kanfer, J.N., Singh, I.N., Pettegrew, J.W., McCartney, D.G., Sorrentino, G., 1996.
Phospholipid metabolism in Alzheimer’s disease and in a human cholinergic
cell. J. Lipid Mediat. Cell Signal. 14, 361–363.
Kantarci, K., Xu, Y., Shiung, M.M., O’Brien, P.C., Cha, R.H., Smith, G.E., Ivnik, R.J., Boeve,
B.F., Edland, S.D., Kokmen, E., Tangalos, E.G., Petersen, R.C., Jack Jr., C.R., 2002.
Comparative diagnostic utility of different MR modalities in mild cognitive
impairment and Alzheimer’s disease. Dement. Geriatr. Cogn. Disord. 14, 198–207.
Klunk, W.E., Xu, C.J., Panchalingam, K., McClure, R.J., Pettegrew, J.W., 1994. Analysis
of magnetic resonance spectra by mole percent: comparison to absolute units.
Neurobiol. Aging 15, 133–140.
Lanctot, K.L., Herrmann, N., Rothenburg, L., Eryavec, G., 2007. Behavioral correlates
of GABAergic disruption in Alzheimer’s disease. Int. Psychogeriatr. 19, 151–158.
Le Belle, J.E., Harris, N.G., Williams, S.R., Bhakoo, K.K., 2002. A comparison of cell and
tissue extraction techniques using high-resolution
H NMR spectroscopy. NMR
Biomed. 15, 37–44.
Lindon, J.C., Nicholson, J.K., Everett, J.R., 1999. NMR spectroscopy of biofluids,
Annual Reports on NMR Spectroscopy, vol. 38, Annual Reports on NMR Spec-
troscopy, vol. 38, pp. 1–88.
Macura, S., Huang, Y., 1981. Two-dimensional chemical exchange and cross-relax-
ation spectroscopy of coupled nuclear spins. J. Magn. Reson. 43, 259–281.
Marjanska, M., Curran, G.L., Wengenack, T.M., Henry, P.G., Bliss, R.L., Poduslo, J.F.,
Jack Jr., C.R., Ugurbil, K., Garwood, M., 2005. Monitoring disease progression in
transgenic mouse models of Alzheimer’s disease with proton magnetic reso-
nance spectroscopy. Proc. Natl. Acad. Sci. U.S.A. 102, 11906–11910.
McGowan, E., Eriksen, J., Hutton, M., 2006. A decade of modeling Alzheimer’s
disease in transgenic mice. Trends Genet. 22, 281–289.
McIlwain, H., Bachelard, H.S., 1985. Biochemistry and the Central Nervous Systems.
Churchill Livingstone, Edinburgh, New York.
Meyerhoff, D.J., MacKay, S., Bachman, L., Poole, N., Dillon, W.P., Weiner, M.W., Fein,
G., 1993. Reduced brain N-acetylaspartate suggests neuronal loss in cognitively
impaired human immunodeficiency virus-seropositive individuals: in vivo
magnetic resonance spectroscopic imaging. Neurology 43, 509–515.
Mohanakrishnan, P., Fowler, A.H., Vonsattel, J.P., Husain, M.M., Jolles, P.R., Liem, P.,
Komoroski, R.A., 1995. An in vitro 1H nuclear magnetic resonance study of the
temporoparietal cortex of Alzheimer brains. Exp. Brain Res. 102, 503–510.
Mucke, L., Masliah, E., Yu, G.Q., Mallory, M., Rockenstein, E.M., Tatsuno, G., Hu, K.,
Kholodenko, D., Johnson-Wood, K., McConlogue, L., 2000. High-level neuronal
R.M. Salek et al. / Neurochemistry International 56 (2010) 937–947
expression of abeta 1–42 in wild-type human amyloid protein precursor
transgenic mice: synaptotoxicity without plaque formation. J. Neurosci. 20,
Pears, M.R., Cooper, J.D., Mitchison, H.M., Mortishire-Smith, R.J., Pearce, D.A., Griffin,
J.L., 2005. High resolution
H NMR-based metabolomics indicates a neurotrans-
mitter cycling deficit in cerebral tissue from a mouse model of Batten disease. J.
Biol. Chem. 280, 42508–42514.
Petroff, O.A., Pleban, L.A., Spencer, D.D., 1995. Symbiosis between in vivo and in vitro
NMR spectroscopy: the creatine, N-acetylaspartate, glutamate, and GABA con-
tent of the epileptic human brain. Magn. Reson. Imaging 13, 1197–1211.
Pettegrew, J.W., Panchalingam, K., Moossy, J., Martinez, J., Rao, G., Boller, F., 1988.
Correlation of phosphorus-31 magnetic resonance spectroscopy and morpho-
logic findings in Alzheimer’s disease. Arch. Neurol. 45, 1093–1096.
Pfefferbaum, A., Adalsteinsson, E., Spielman, D., Sullivan, E.V., Lim, K.O., 1999. In
vivo brain concentrations of N-acetyl compounds, creatine, and choline in
Alzheimer disease. Arch. Gen. Psychiatry 56, 185–192.
Rissman, R.A., De Blas, A.L., Armstrong, D.M., 2007. GABA(A) receptors in aging and
Alzheimer’s disease. J. Neurochem. 103, 1285–1292.
Rothman, D.L., Behar, K.L., Hyder, F., Shulman, R.G., 2003. In vivo NMR studies of the
glutamate neurotransmitter flux and neuroenergetics: implications for brain
function. Annu. Rev. Physiol. 65, 401–427.
Rubinsztein, D.C., 1997. The genetics of Alzheimer’s disease. Prog. Neurobiol. 52,
Schousboe, A., Westergaard, N., Waagepetersen, H.S., Larsson, O.M., Bakken, I.J.,
Sonnewald, U., 1997. Trafficking between glia and neurons of TCA cycle inter-
mediates and related metabolites. Glia 21, 99–105.
Selkoe, D.J., 2001. Alzheimer’s disease: genes, proteins, and therapy. Physiol. Rev.
81, 741–766.
Selkoe, D.J., Wolfe, M.S., 2000. In search of gamma-secretase: presenilin at the
cutting edge. Proc. Natl. Acad. Sci. U.S.A. 97, 5690–5692.
Sonnewald, U., Kondziella, D., 2003. Neuronal glial interaction in different neuro-
logical diseases studied by ex vivo
C NMR spectroscopy. NMR Biomed. 16,
Suh, Y.H., Checler, F., 2002. Amyloid precursor protein, presenilins, and alpha-
synuclein: molecular pathogenesis and pharmacological applications in Alz-
heimer’s disease. Pharmacol. Rev. 54, 469–525.
Svennerholm, L., Gottfries, C.G., 1994. Membrane lipids, selectively diminished in
Alzheimer brains, suggest synapse loss as a primary event in early-onset form
(type I) and demyelination in late-onset form (type II). J. Neurochem. 62, 1039–
Wold, S., Albano, C., Dunn, W.J., Edlund, U., Esbensen, K., Geladi, P., Hellberg, S.,
Johansson, E., Lindberg, W., Sjo
m, M., 1984. Multivariate Data Analysis in
Chemistry. D. Reidel Publishing Company, Holland.
Woodhouse, A., Vickers, J.C., Adlard, P.A., Dickson, T.C., 2009. Dystrophic neurites in
TgCRND8 and Tg2576 mice mimic human pathological brain aging. Neurobiol.
Aging 30, 864–874.
R.M. Salek et al. / Neurochemistry International 56 (2010) 937–947
    • "In addition, the two first components (vectors) contributed positively to the model (value of Q 2 positive = 66.5 %), and the variation of the proteins was 97.5 % (R 2 ). Values of Q 2 > 0.08 indicates that a model is better than chance, and scores of 0.7 or higher, demonstrate a very robust trend or separation [40]. The protein SCCCRZ3002G10 of unknown function was the one that contributed the most to the separation of the groups, being found in higher amount using the ND Method (Fig. 3a, c). "
    [Show abstract] [Hide abstract] ABSTRACT: Background Sugarcane has been used as the main crop for ethanol production for more than 40 years in Brazil. Recently, the production of bioethanol from bagasse and straw, also called second generation (2G) ethanol, became a reality with the first commercial plants started in the USA and Brazil. However, the industrial processes still need to be improved to generate a low cost fuel. One possibility is the remodeling of cell walls, by means of genetic improvement or transgenesis, in order to make the bagasse more accessible to hydrolytic enzymes. We aimed at characterizing the cell wall proteome of young sugarcane culms, to identify proteins involved in cell wall biogenesis. Proteins were extracted from the cell walls of 2-month-old culms using two protocols, non-destructive by vacuum infiltration vs destructive. The proteins were identified by mass spectrometry and bioinformatics. Results A predicted signal peptide was found in 84 different proteins, called cell wall proteins (CWPs). As expected, the non-destructive method showed a lower percentage of proteins predicted to be intracellular than the destructive one (33 % vs 44 %). About 19 % of CWPs were identified with both methods, whilst the infiltration protocol could lead to the identification of 75 % more CWPs. In both cases, the most populated protein functional classes were those of proteins related to lipid metabolism and oxido-reductases. Curiously, a single glycoside hydrolase (GH) was identified using the non-destructive method whereas 10 GHs were found with the destructive one. Quantitative data analysis allowed the identification of the most abundant proteins. Conclusions The results highlighted the importance of using different protocols to extract proteins from cell walls to expand the coverage of the cell wall proteome. Ten GHs were indicated as possible targets for further studies in order to obtain cell walls less recalcitrant to deconstruction. Therefore, this work contributed to two goals: enlarge the coverage of the sugarcane cell wall proteome, and provide target proteins that could be used in future research to facilitate 2G ethanol production. Electronic supplementary material The online version of this article (doi:10.1186/s12870-015-0677-0) contains supplementary material, which is available to authorized users.
    Full-text · Article · Dec 2016
    • "Brain regional variation should also be considered, 125 as a comprehensive perspective of brain pathophysiology can be obtained from the integrated 126 analysis of multiple brain regions (e.g. Ivanisevic et al., 2014; Salek et al., 2010). 127 "
    [Show abstract] [Hide abstract] ABSTRACT: Metabolism being a fundamental part of molecular physiology, elucidating the structure and regulation of metabolic pathways is crucial for obtaining a comprehensive perspective of cellular function and understanding the underlying mechanisms of its dysfunction(s). Therefore, quantifying an accurate metabolic network activity map under various physiological conditions is among the major objectives of systems biology in the context of many biological applications. Especially for CNS, metabolic network activity analysis can substantially enhance our knowledge about the complex structure of the mammalian brain and the mechanisms of neurological disorders, leading to the design of effective therapeutic treatments. Metabolomics has emerged as the high-throughput quantitative analysis of the concentration profile of small molecular weight metabolites, which act as reactants and products in metabolic reactions and as regulatory molecules of proteins participating in many biological processes. Thus, the metabolic profile provides a metabolic activity fingerprint, through the simultaneous analysis of tens to hundreds of molecules of pathophysiological and pharmacological interest. The application of metabolomics is at its standardization phase in general, and the challenges for paving a standardized procedure are even more pronounced in brain studies. In this review, we support the value of metabolomics in brain research. Moreover, we demonstrate the challenges of designing and setting up a reliable brain metabolomic study, which, among other parameters, has to take into consideration the sex differentiation and the complexity of brain physiology manifested in its regional variation. We finally propose ways to overcome these challenges and design a study that produces reproducible and consistent results.
    Full-text · Article · May 2016
    • "The metabolite changes observed here (fig. 10) only show a limited overlap with those detected in mouse (Salek, et al., 2010) and man (Trushina and Mielke, 2013). There are multiple explanations: We use whole body parts rather than organs or body fluids, we separate effects of Aβ-expression and Aβtoxicity , and we do of course use flies, and as it seems, the dominating effects here are those of decreased oxidative phosphorylation and increased oxidative stress. "
    [Show abstract] [Hide abstract] ABSTRACT: Amyloid beta peptide (Aβ) aggregation is linked to the initiation of Alzheimer's disease; accordingly, aggregation-prone isoforms of Aβ, expressed in the brain, shorten the lifespan of Drosophila melanogaster. However, the lethal effects of Aβ are not apparent until after day 15. We used shibireTS flies that exhibit a temperature-sensitive paralysis phenotype as a reporter of proteostatic robustness. In this model we found that increasing age, but not Aβ expression, lowered the flies’ permissive temperature, suggesting that Aβ did not exert its lethal effects by proteostatic disruption. Instead we observed that chemical challenges, in particular oxidative stressors, discriminated clearly between young (robust) and old (sensitive) flies. Using nuclear magnetic resonance spectroscopy in combination with multivariate analysis, we compared water-soluble metabolite profiles at various ages in flies expressing Aβ in their brains. We observed two genotype-linked metabolomic signals, the first reported the presence of any Aβ isoform and the second the effects of the lethal Arctic Aβ. Lethality was specifically associated with signs of oxidative respiration dysfunction and oxidative stress.
    Full-text · Article · Feb 2016
Show more