Quantitative Proteomic Analysis of the Hippocampus in the 5XFAD Mouse Model at Early Stages of Alzheimer's Disease Pathology

Abstract

Alzheimer's disease (AD) is characterized by progressive memory loss accompanied by synaptic and neuronal degeneration. Although research has shown that substantial neurodegeneration occurs even during the early stages of AD, the detailed mechanisms of AD pathogenesis are largely unknown because of difficulties in diagnosis and limitations of the analytical methods. The 5XFAD mouse model harbors five early-onset familial AD (FAD) mutations and displays substantial amyloid plaques and neurodegeneration. Here, we use quantitative mass spectrometry to identify proteome-wide changes in the 5XFAD mouse hippocampus during the early stages of AD pathology. A subset of the results was validated with immunoblotting. We found that the 5XFAD mice display higher expression of ApoE, ApoJ (clusterin), and nicastrin, three important proteins in AD that are known to participate in amyloid-β processing and clearance, as well as the neurological damage/glial marker protein GFAP and other proteins. A large subset of the proteins that were up- or downregulated in 5XFAD brains have been implicated in neurological disorders and cardiovascular disease, suggesting an association between cardiovascular disease and AD. Common upstream regulator analysis of upregulated proteins suggested that the XBP1, NRF2, and p53 transcriptional pathways were activated, as was IGF-1R signaling. Protein interactome analysis revealed an interconnected network of regulated proteins, with two major sub-networks centered on AβPP processing membrane complexes and mitochondrial proteins. Together with a recent study on the transcriptome of 5XFAD mice, our study allows a comprehensive understanding of the molecular events occurring in 5XFAD mice during the early stages of AD pathology.

Full text

Journal of Alzheimer’s Disease 36 (2013) 321–334
DOI 10.3233/JAD-130311
IOS Press
321
Quantitative Proteomic Analysis of the
Hippocampus in the 5XFAD Mouse Model at
Early Stages of Alzheimer’s Disease
Pathology
Ingie Honga,1, Taewook Kangb,c,1, YongCheol Yooc,1, Royun Parkd,1, Junuk Leea, Sukwon Leea,
Jeongyeon Kima, Boemjong Songa, Se-Young Kimc, Minho Moond,KiNaYun
c, Jin Young Kimc,
Inhee Mook-Jungd,∗, Young Mok Parkb,c,∗and Sukwoo Choia,∗
aSchool of Biological Sciences, College of Natural Sciences, Seoul National University, Seoul, Korea
bGraduate School of Analytical Science and Technology (GRAST), Chungnam National University, Daejeon, Korea
cMass Spectrometry Research Center, Korea Basic Science Institute, Ochang, Korea
dDepartment of Biochemistry and Molecular Biology, Seoul National University College of Medicine, Seoul, Korea
Handling Associate Editor: Bhumsoo Kim
Accepted 18 March 2013
Abstract. Alzheimer’s disease (AD) is characterized by progressive memory loss accompanied by synaptic and neuronal
degeneration. Although research has shown that substantial neurodegeneration occurs even during the early stages of AD, the
detailed mechanisms of AD pathogenesis are largely unknown because of difficulties in diagnosis and limitations of the analytical
methods. The 5XFAD mouse model harbors five early-onset familial AD (FAD) mutations and displays substantial amyloid
plaques and neurodegeneration. Here, we use quantitative mass spectrometry to identify proteome-wide changes in the 5XFAD
mouse hippocampus during the early stages of AD pathology. A subset of the results was validated with immunoblotting. We
found that the 5XFAD mice display higher expression of ApoE, ApoJ (clusterin), and nicastrin, three important proteins in AD
that are known to participate in amyloid-␤processing and clearance, as well as the neurological damage/glial marker protein
GFAP and other proteins. A large subset of the proteins that were up- or downregulated in 5XFAD brains have been implicated in
neurological disorders and cardiovascular disease, suggesting an association between cardiovascular disease and AD. Common
upstream regulator analysis of upregulated proteins suggested that the XBP1, NRF2, and p53 transcriptional pathways were
activated, as was IGF-1R signaling. Protein interactome analysis revealed an interconnected network of regulated proteins, with
two major sub-networks centered on A␤PP processing membrane complexes and mitochondrial proteins. Together with a recent
study on the transcriptome of 5XFAD mice, our study allows a comprehensive understanding of the molecular events occurring
in 5XFAD mice during the early stages of AD pathology.
Keywords: Alzheimer’s disease, amyloid-␤, memory, mild cognitive impairment, PS1, proteomics, transgenic mice
1These authors contributed equally to this work.
∗Correspondence to: Sukwoo Choi, Ph.D., School of Biological
Sciences, College of Natural Sciences, Seoul National Uni-
versity, Seoul 151-742, Korea. E-mail: sukwoo12@snu.ac.kr.;
Young Mok Park, Ph.D., Mass Spectrometry Research Center,
Korea Basic Science Institute, Ochang 363-883, Korea. E-mail:
ympark@kbsi.re.kr.; Inhee Mook-Jung, Ph.D., Department of Bio-
chemistry and Biomedical Sciences, Seoul National University
College of Medicine, 28 Yungun-dong, Jongno-gu, Seoul 110–799,
Korea. E-mail: inhee@snu.ac.kr.
ISSN 1387-2877/13/$27.50 © 2013 – IOS Press and the authors. All rights reserved

322 I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice
INTRODUCTION
Alzheimer’s disease (AD) is the most common form
of irreversible dementia in elderly people, tragically
affecting more than 26 million people worldwide [1].
Clinically, the disease is characterized by progres-
sive memory loss and a decline in cognitive abilities
(such as language, attention, and executive function).
A definitive diagnosis requires postmortem assess-
ment of the two histological hallmarks of AD: neuritic
amyloid-␤(A␤) plaques and neurofibrillary tangles
(NFTs) [2, 3]. Several symptomatic treatments exist
for AD; however, no disease-modifying therapies are
currently available [4–6]. Interestingly, the onset of AD
is preceded by a period of mild cognitive impairment
(MCI) during which individuals display only mild
disturbances in cognitive functions such as episodic
memory [7]. A␤accumulation and aggregation, along
with various synaptic deficits, are widely found in
MCI, but no clinical method is available for identi-
fying or treating prodromal AD [8]. A recent study
even shows that A␤deposition may start as early
as 15 years before symptoms begin [9]. Disease-
modifying treatments for AD, when they become
available, will thus need to intervene early in the course
of the disease, before neurodegeneration is too severe
and thus irreversible, which underscores the impor-
tance of understanding the early pathogenesis of AD
[6].
Ample genetic and biochemical data support the
hypothesis that A␤accumulation and aggregation in
the brain are early and important events in the patho-
genesis of AD [10, 11]. Mutations associated with
early-onset familial AD (FAD) are dominantly inher-
ited and are found in the amyloid-␤protein precursor
(A␤PP) gene itself or in the presenilin 1 (PS1) and
PS2 genes [3], the products of which, together with
nicastrin, Aph1, and Pen-2, are essential components
of the ␥-secretase protein complex [12]. The ␧4 allele
of the apolipoprotein E (APOE) gene has been iden-
tified as the major risk factor for the more common
late-onset AD [13]. The presence of an APOE ␧4 allele
increases the risk of AD by approximately fourfold,
presumably by diminishing protection or augmenting
toxicity compared to other alleles [14].
Neuroinflammation is another hallmark of AD, and
activated astrocytes and microglia have been found sur-
rounding amyloid plaques [15, 16]. Due to the complex
nature of AD, it is challenging to determine whether
a given pathological structure or biochemical change
drives the disease, is a neutral bystander, or simply
represents an unsuccessful defense mechanism.
The development of quantitative proteomics has
greatly accelerated our understanding of large-scale
protein networks and the identification of impor-
tant biomarkers [17–19]. Because processes such
as alternative splicing, protein processing, and post-
translational modification are central to generating
the full complexity of life, mass-spectrometry-based
proteomics offers an unprecedented level of detail in
the biochemical assessment of biological processes.
Recent developments in chemical peptide labeling with
isobaric tags such as TMT and iTRAQ have enabled
comparison of the expression levels of thousands of
proteins across complex samples. AD pathology has
also been studied using proteomic approaches, mostly
with cerebrospinal fluid [20] or postmortem patient
samples [21]. However, there are substantial difficul-
ties with interpreting data from postmortem tissue
and from individuals of diverse genetic and patho-
logical backgrounds. For instance, the analysis of
postmortem brain tissue cannot discriminate between
changes that are specifically involved in AD versus
those that are simply a consequence of neuronal degen-
eration. Here, we employ TMT tagging to uncover the
proteome-wide changes in 5XFAD transgenic mice, a
mouse model that co-expresses five FAD mutations in
A␤PP and PS1 [22] and develops early A␤42 accu-
mulation (1.5 months), amyloid deposition and gliosis
(2 months), neuron loss (2 months) [23], synapse
degeneration (4 months), spatial learning deficits (4–6
months), and increased p25 levels (9–12 months) [22,
24]. We focus on the early phase of pathology, when
synapse and memory deficits are first detected [22],
to elucidate mechanisms that are involved in the onset
and development of AD. This study broadly comple-
ments our recent study on RNA expression in 5XFAD
mice, which used next-generation RNA-Seq technol-
ogy [25].
MATERIALS AND METHODS
Experimental animals
Transgenic AD model mice with five famil-
ial AD mutations (5XFAD) were purchased from
Jackson Laboratories (strain: B6SJL-Tg [APP-
SwFlLon,PS1*M146L*L286V] 6799Vas/J, stock no.
006554) and maintained by crossing hemizygous
transgenic mice with B6SJL F1 mice [22]. Founder
transgenic mice were identified by polymerase chain
reaction (PCR), and non-transgenic littermate mice
served as controls. All mice were housed in groups

I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice 323
of three to five per cage, with food and water avail-
able ad libitum under standardized environmental
conditions.
5XFAD mice express both mutant human amyloid-␤
protein precursor (APP695) with the Swedish muta-
tion (K670N, M671L), Florida mutation (I716V), and
London mutation (V717I) and human presenilin 1
harboring two FAD mutations (M146L and L286V).
Both transgenes are expressed under the control of the
mouse Thy1 promoter to induce overexpression in the
brain [22]. These mice exhibit AD-related pathology
earlier than other animal models, and amyloid depo-
sition starts in the deep cortex and subiculum at 2
months of age. Synaptic marker proteins decrease at
4–9 months, and memory deficits are detected from 4–6
months of age [26, 27]. Importantly, amyloid plaques
and neuronal death are not detected in the hippocampus
at 4 months. Because we focused on the early events
of AD pathology, before severe neurodegeneration has
occurred, all animals in this study were killed at 4
months of age.
Animal treatment and maintenance were performed
in accordance with the Principles of Laboratory Ani-
mal Care (NIH publication No. 85–23, revised 1985)
and the Animal Care and Use Guidelines of Seoul
National University, Seoul, Korea. All efforts were
made to minimize animal suffering and to reduce the
number of mice used.
Crude synaptosome fractionation and sample
preparation
Mice were anesthetized and decapitated to extract
brain tissue. Three biological replicates from three dif-
ferent mice were generated for both the 5XFAD and
wild-type littermate groups. Hippocampal brain tissue
samples were homogenized in ice-cold homogeniza-
tion buffer containing 10 mM Tris (pH 7.6), 320 mM
sucrose, 5 mM NaF, 1 mM Na3VO4, 1 mM EDTA,
1 mM EGTA,and protease/phosphatase inhibitor cock-
tail. The homogenates were centrifuged at 1,000×
gat 4◦C for 10 min to remove nuclei and large
debris. The resultant supernatant was further cen-
trifuged at 10,000×gat 4◦C for 30 min to obtain a
crude synaptosomal fraction, which was lysed in mod-
ified RIPA buffer containing 50 mM Tris (pH 7.6),
150 mM NaCl, 5 mM NaF, 1 mM Na3VO4, 0.5% Tri-
ton X-100, 0.5% sodium deoxycholate, 0.1% SDS,
and protease/phosphatase inhibitor cocktail. Samples
were sonicated and spun down at 15,000 ×gat4
◦C
for 15 min. Protein content was measured by a bicin-
choninic acid assay (Thermo Scientific) following the
manufacturer’s protocol, and confirmed by SDS-PAGE
and silver staining.
TMT labeling
Samples were subsequently tagged with tan-
dem mass tags for quantitative mass spectrometry
(TMTsixplex™ Isobaric Mass Tagging Kit, Thermo
Scientific). Briefly, 100␮g was taken from each
sample and was reduced with 500 mM tris(2-
carboxyethyl)phosphine (TCEP) at 55◦C for 1 h and
then alkylated with 300 mM iodoacetamide (IAA)
at 37◦C in the dark for 30 min. The samples were
desalted using a 10,000 MW-cutoff membrane filter
and dissolved in 100 mM triethylammonium bicarbon-
ate (TEAB) buffer to a final concentration of 1 ␮g/␮L.
Sequencing-grade trypsin (Promega, Madison, WI,
USA) was added at 1:100 (w/w) into proteins in TEAB
buffer and incubated overnight at 37◦C. Three sam-
ples of both transgenic 5XFAD mice and littermates
were individually labeled using TMT-126, 128, 130
(5XFAD mice), and TMT-127, 129, 131 (wild-type
littermates) following the manufacturer’s instructions.
Aqueous hydroxylamine solution (5% w/v) was added
to quench the reaction. The six samples were then
combined, speed-vacuum dried, and then dissolved in
50 ␮L of water containing 0.1% formic acid for LC-
MS/MS analysis.
2D-LC-MS/MS
The TMT-labeled samples were analyzed using
a 2D-LC-MS/MS system consisting of a nanoAC-
QUITY UltraPerformance LC System (Waters, USA)
and an LTQ Orbitrap Elite mass spectrometer (Thermo
Scientific, USA) equipped with a nano-electrospray
source. A detailed description of 2D-LC-MS/MS anal-
ysis can be found in the literature [28, 29]. Briefly,
a strong cation exchange (5 ␮m, 3 cm) column was
placed just before the C18 trap column (id 180 ␮m,
length 20 mm, and particle size 5 ␮m; Waters). Pep-
tide solutions were loaded in 5 ␮L aliquots for each
run. Peptides were displaced from the strong cation
exchange phase to the C18 phase by a salt gradient that
was introduced through an autosampler loop and then
desalted for 10 min at a flow rate of 4 ␮L/min. Then, the
trapped peptides were separated on a 200 mm home-
made microcapillary column consisting of C18 (Aqua;
particle size 3 ␮m) packed into 100 ␮m silica tubing
with an orifice id of 5 ␮m.
An eleven-step salt gradient was performed using
3␮L of 0, 25, 50, 100, 250, and 500 mM ACN (0.1%

324 I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice
formic acid in water) and 4, 5, 9, and an additional
9␮L at 500 mM ACN (0.1% formic acid in 30% ACN).
The mobile phases, A and B, were composed of 0 and
100% acetonitrile, respectively, and each contained
0.1% formic acid. The LC gradient began with 5%
B for 1 min and was ramped to 20% B over 5 min, to
45% B over 90 min, to 95% B over 1 min, and remained
at 95% B over 13 min and was then ramped to 5%
B for another 5 min. The column was re-equilibrated
with 5% B for 15 min before the next run. The voltage
applied to produce an electrospray was 2.0 kV. Dur-
ing the chromatographic separation, the LTQ Orbitrap
Elite was operated in a data-dependent mode. The MS
data were acquired using the following parameters:
four data-dependent CID-high energy collision disso-
ciation (CID-HCD) dual MS/MS scans per full scan;
CID scans were acquired in LTQ with two-microscan
averaging; full scans and HCD scans were acquired in
Orbitrap at resolutions of 60,000 and 15,000, respec-
tively, with two-microscan averaging; 35% normalized
collision energy in CID and 45% normalized collision
energy in HCD; ±1 Da isolation window. Previously
fragmented ions were excluded for 60 s. In CID-HCD
dual scans, each selected parent ion was first frag-
mented by CID and then by HCD.
Protein identification, quantification, and
statistical analysis
The resultant MS/MS spectra were analyzed
against the latest murine IPI database (IPI.MOUSE.
7.26.2012). Protein identification, quantification, and
analysis were performed with Integrated Proteomics
Pipeline - IP2 (Integrated Proteomics Applications)
using ProLuCID, DTASelect2 and Census. The rate
of decoy hits in the combined forward and reversed
database was less than 1% of the forward hits. Pro-
LuCID [27] was used to identify the peptides with
the following parameters: a precursor mass error tol-
erance of 25 ppm and a fragment ion mass error of
600 ppm. The enzyme was specified as trypsin, and
three potential missed cleavages were allowed. TMT
modification at the N-terminus and lysine residues and
carbamidomethylation at cysteine residues were cho-
sen as static modifications. Oxidation at methionine
was chosen as the variable modification.
The CID and HCD tandem MS spectra from the
same precursor ion are often combined by software
to allow better peptide identification and quantifica-
tion [30]. We used in-house software in which reporter
ions from the HCD spectrum were inserted into the
CID spectrum with the same precursor ion at the pre-
vious scan. Reporter ions were extracted from small
windows (±20 ppm) around their expected m/z in the
HCD spectrum. The output data files were filtered
and sorted to compose the protein list using DTAS-
elect [31] with two or more peptide assignments for
a protein identification. Quantitative analysis was con-
ducted using Census [32], and the intensity at a reporter
ion channel for a protein was calculated as the average
of this reporter ion’s intensities from all constituent
peptides from the identified protein. The resulting
ratios were logarithmized (base = 2) to achieve a nor-
mal distribution. The median and standard deviation
were calculated, and ratio values were corrected for the
median to account for variability among different pairs
[33]. Ratios were averaged, and proteins with ratio val-
ues beyond p< 0.01 in normal distribution were defined
as significantly regulated. To further assess the indi-
vidual statistical significance of the expression level
change in each protein, one-sample t-tests were used.
Gene annotation and upstream regulator analysis
Gene Ontology annotation enrichment analysis was
performed using the DAVID Bioinformatics Resource
(v6.7) developed by NIAID, at the National Insti-
tutes of Health [34]. DAVID analysis enabled the
enrichment of functional-related gene groups and cel-
lular compartments. Ingenuity Pathway Analysis (IPA;
Ingenuity Systems, http://www.ingenuity.com) was
used to functionally annotate genes implicated in dis-
ease and search for common upstream regulators.
Network analysis for the regulated proteins
The proteins of our study were searched against the
STRING database version 9 [35] for protein-protein
interactions. Only interactions between the proteins
belonging to the increased or decreased groups were
selected. STRING defines a metric called “confidence
score” to define interaction confidence; we selected all
interactions for our regulated dataset that had a con-
fidence score ≥0.7 (high confidence). The resulting
interactome had 16 nodes and 30 interactions, which
we call the 5XFAD interaction network (FADIN)
(Fig. 5).
Immunoblotting
Immunochemistry was carried out to quantify spe-
cific proteins present in the samples and to validate

I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice 325
TMT analysis. Samples that were saved for valida-
tion after TMT analysis were loaded and separated on
polyacrylamide gels and transferred onto PVDF mem-
branes, blocked with 10% (w/v) milk, and incubated
with primary antibody for 2 h at room temperature.
The following primary antibodies were diluted in
Tris-buffered saline with 0.1% Tween-20 (TBST)
with 3% (w/v) BSA and 0.001% sodium azide:
anti-nicastrin (1:1,000; MAB5556, Millipore), anti-
GFAP (1:1,000; 18-0063, Invitrogen) antibodies were
used. Goat anti-mouse HRP or goat anti-rabbit HRP
secondary antibodies were applied for 1 h at room
temperature before staining by chemiluminescence.
The optical density for each band was quantified with
ImageJ (NIH). Quantification was performed in the
linear range of band intensity to ensure accurate quan-
tification.
RESULTS
Quantification of proteomic changes in the
hippocampus of 5XFAD mice
The hippocampi of 4-month-old 5XFAD mice and
their wild-type littermates were dissected, and pro-
teins were extracted. After TMT-labeling, the samples
were mixed, and the ratios of individual proteins were
examined by quantitative proteomics. A total of 1481
proteins were identified and quantified from 10,766
unique peptides. Tables 1 and 2 show the proteins
whose levels were strongly increased or decreased in
5XFAD mice compared to wild-type littermates. The
overexpressed protein A␤PP showed a strong increase,
as did other well-known AD-related proteins such as
the apolipoproteins ApoE and clusterin (ApoJ), the
␥-secretase protein complex subcomponent nicastrin
(these three proteins are known to contribute to A␤pro-
cessing and clearance [36–38]), and the glial marker
protein GFAP (69 increased proteins in total). The lev-
els of proteins including Shroom2, centaurin alpha 1
(Adap1, Arf-GAP with dual PH domain-containing
protein 1), Opcml, Mapre2 (APC protein-binding
EB1 gene family homologue), and CB1 cannabinoid
receptor-interacting protein 1 (Cnrip1) were strongly
decreased in 5XFAD mice (28 decreased proteins in
total). The MS/MS spectra for two representative pro-
teins, ApoE and GFAP, are shown in Fig. 1A and B,
where the level of 5XFAD-labeled mass tags (126, 128,
130) is significantly higher than in control littermates
(127, 129, 131).
Validation of AβPP overexpression in 5XFAD
transgenic mice
Because 5XFAD mice overexpress a human mutant
form of A␤PP (K670N, M671L, I716V, V717I), we
were able to use the resolution of mass spectrometry
to discern mouse native A␤PP. A trypsinized peptide
unique to mouse native A␤PP showed similar quantifi-
cation levels in 5XFAD mice and wild-type littermates
(p= 0.1111, unpaired t-test), whereas total A␤PP was
overexpressed by approximately threefold in 5XFAD
mice (p< 0.0023), in good agreement with our prior
RNA expression study [25]. Unfortunately, the other
overexpressed protein PS1 was not quantified in our
experiments. These results verify the overexpression of
mutant A␤PP in 5XFADmice and support the accuracy
of our quantitative analysis.
Validation of quantitative analysis with western
blot
We next validated the quantification of our
TMT analysis in a small subset of proteins, using
immunoblotting. Western blots for GFAP and nicas-
trin verified strong increases in protein expression,
correlating well with the TMT analysis (Fig. 3A, B)
and further validating our quantitative analysis with
orthogonal methodology.
Enrichment analysis of increased and decreased
proteins in 5XFAD mice
We used DAVID [34] with Gene Ontology to
functionally categorize the proteins regulated in
5XFAD mice by biological process, cellular compo-
nent, and molecular function (Fig. 4). The results
for biological process showed strong enrichment
of oxidative stress-related proteins, along with ion
homeostasis, protein folding, nerve impulse transmis-
sion, and axon ensheathment-related proteins. Cellular
component analysis showed strong enrichment of
endoplasmic reticulum (ER), pigment granule, vesi-
cle, membrane-enclosed lumen, and mitochondrial
proteins. Molecular function annotation revealed that
protein disulfide isomerase (PDI) activity, which is
mostly localized to the ER, is enriched in the regulated
proteins of 5XFAD mice, along with oxidoreduc-
tase activity, dehydrogenase activity, cofactor binding,
unfolded protein binding, myelin sheath constituents,
and copper ion binding.

326 I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice
Table 1
Upregulated proteins in the 5XFAD mouse hippocampus
IPI GENE Average No of IPI GENE Average No of
SYMBOL fold change Peptides SYMBOL fold change Peptides
(TG/WT) (TG/WT)
IPI00230494 App 3.39∗22 IPI00124291 Vps45 1.20 2
IPI00323571 Apoe 3.29∗18 IPI00263013 Plp1 1.20 90
IPI00123059 Scg5 1.74 3 IPI00758024 Prdx6 1.20 32
IPI00320420 Clu 1.73∗5 IPI00134058 Erp44 1.20 2
IPI00130758 Itm2b 1.68∗2 IPI00316989 Epdr1 1.20∗4
IPI00649033 Gfap 1.63∗23 IPI00395177 Abr 1.19 6
IPI00121514 Stip1 1.37 16 IPI00312018 Mlec 1.19∗2
IPI00133218 Arl8b 1.36∗2 IPI00128826 Cacng8 1.19 8
IPI00230145 Fth1 1.35 3 IPI00853902 Mllt4 1.19∗2
IPI00404551 Ctsd 1.32 18 IPI00123342 Hyou1 1.19 4
IPI00752486 Cst3 1.28 9 IPI00135686 Ppib 1.19 7
IPI00853896 Tmsb4x 1.27∗9 IPI01027185 Txn2 1.18 32
IPI00112129 Gatm 1.27 2 IPI00130589 Sod1 1.18 5
IPI00885558 Pdia3 1.26 6 IPI00137227 Rab2a 1.18∗13
IPI00944803 Prpsap2 1.26 2 IPI00113052 Tsfm 1.18 3
IPI00857249 Clstn3 1.26 5 IPI00222496 Pdia6 1.18 10
IPI00129526 Hsp90b1 1.24 32 IPI00830254 Ppp1r12b 1.18 2
IPI00130624 Pld3 1.24 12 IPI00828469 Lap3 1.18 8
IPI00118674 Ncstn 1.24∗2 IPI00221454 Prdx6-ps1 1.17 17
IPI00989331 Wnk4 1.24 6 IPI00399449 Nsfl1c 1.17∗16
IPI00845833 Ppp3r1 1.23 50 IPI00113386 Ethe1 1.17 2
IPI00474959 Rcn2 1.23 8 IPI00775948 Rpl7 1.17∗7
IPI00230108 Pdia3 1.23∗38 IPI00652358 Arsb 1.17∗6
IPI00129178 Oat 1.23∗10 IPI00399958 Calu 1.17 4
IPI00458048 Sorbs1 1.22 3 IPI00320241 Dnajb11 1.17 2
IPI00123639 Calr 1.21 11 IPI00337975 Scn8a 1.16 2
IPI00223594 Mbp 1.21 138 IPI00125509 Fxc1 1.16∗3
IPI00118832 Erp29 1.21 9 IPI00776047 Aifm1 1.16 5
IPI00124115 S100a13 1.20 6 IPI00223377 Mbp 1.16 139
IPI00464317 Gls 1.20 46 IPI00130640 Hrsp12 1.16 17
IPI00271951 Pdia4 1.20 4 IPI00331549 Dhrs1 1.16∗11
IPI00230737 Aqp4 1.20∗3 IPI00762452 Idh1 1.16 11
IPI00329953 Stx7 1.20∗3 IPI00170212 Cox16 1.16 2
IPI00109108 Stt3a 1.20∗3 IPI00153107 Blmh 1.15 5
IPI00133522 P4hb 1.20 11
Average fold change was calculated by averaging three ratio values of 5XFAD mice/wild-type littermates. Asterisks indicate proteins with ratios
that were significantly different from 1 (one-sample t-test, p< 0.05).
Disease-related functional annotation of proteins
with increased and decreased levels in 5XFAD
mice
The functional annotation and upstream regula-
tor analysis were performed using Ingenuity Pathway
Analysis. The significance of the canonical pathway
and functional annotation was determined with p-
values (as an index of the confidence in the overlap)
and/or ratios (as an index of the amount of over-
lap). Functional annotation indicated that 28 of the
97 regulated proteins in 5XFAD mice had previously
been implicated in neurological disorders. The impli-
cated neurological disorders included not only AD
but also Huntington’s disease, Parkinson’s disease,
and amyotrophic lateral sclerosis, suggesting that var-
ious neurodegenerative mechanisms are activated in
5XFAD mice (Table 3). Another large subset of the
regulated proteins (14 molecules) was related to car-
diovascular disease (CVD), verifying our previous
RNA-Seq results [25] and supporting the hypothesis
that there is a strong association between CVD and
AD [39].
Upstream regulator analysis of proteins
upregulated in 5XFAD mice
To identify upstream signaling pathways that may
underlie the increased expression of proteins in
5XFAD mice, we used the upstream regulator anal-
ysis in IPA. Three transcriptional regulators (XBP1,
NFE2L2, TP53) and one transmembrane receptor
signaling pathway (IGF-1R) showed significant acti-
vation z-scores, suggesting that the activation of these

I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice 327
Table 2
Down-regulated proteins in the 5XFAD mouse hippocampus
IPI GENE Average No of IPI GENE Average No of
SYMBOL Fold Change Peptides SYMBOL Fold Change Peptides
(TG/WT) (TG/WT)
IPI00848690 Ly6h 0.87 8 IPI00857911 Opcml 0.84∗19
IPI00308976 Me3 0.87 5 IPI00874497 Pde1a 0.83∗5
IPI00122975 Gpm6b 0.86 6 IPI00620222 Myo1c 0.83 3
IPI00857239 Atp6v0c 0.86 5 IPI00131695 Alb 0.82 3
IPI00115977 Me2 0.86 3 IPI00890092 Aldh7a1 0.82 4
IPI00311175 Tuba8 0.85∗407 IPI00323166 Nnt 0.81 5
IPI00407339 Hist1h4 0.85 9 IPI00128857 Me1 0.81 4
IPI00756510 Smap2 0.85 2 IPI00988950 Hbb-b1 0.81 14
IPI00108330 Cnrip1 0.85∗5 IPI00378768 Adap1 0.80 3
IPI00880852 Vamp2 0.85 51 IPI00121443 Cox6a1 0.80 3
IPI00474073 Mapre2 0.85∗4 IPI00463074 Shroom2 0.79∗5
IPI00408059 Ppp2r5c 0.85 2 IPI00387416 Ubqln2 0.79 3
IPI00126551 Diras2 0.85 2 IPI00469114 Hba-a 0.78 2
IPI00463489 Opcml 0.84∗24 IPI00626782 Arfgef1 0.69 2
Average fold change was calculated by averaging three ratio values of 5XFAD mice/wild-type littermates. Asterisks indicate proteins with ratios
that were significantly different from 1 (one-sample t-test, p< 0.05). No of peptides shows number of independent spectra used for quantification.
Table 3
5XFAD-regulated molecules implicated in neurological disorders
Functional annotation p-value # of Molecules Molecules
Movement Disorders 2.76E-05 16 AIFM1,APOE,App,AQP4,CLU,CTSD,ETHE1,FTH1,
GFAP,MBP,PLP1,PPP3R1,PRDX6,SCN8A,SOD1,
TMSB10/TMSB4X
Neuromuscular disease 1.23E-05 15 AIFM1,APOE,AQP4,CLU,CST3,DNAJB11,FTH1,
GFAP,MBP,PDIA3,PPP3R1,PRDX6,SCN8A,SOD1,
TMSB10/TMSB4X
Progressive motor neuropathy 2.40E-07 12 AIFM1,APOE,AQP4,CLU,CST3,DNAJB11,FTH1,
GFAP,MBP,PDIA3,SCN8A,SOD1
Disorder of basal ganglia 2.91E-04 12 AIFM1,APOE,AQP4,CLU,FTH1,GFAP,MBP,PPP3R1,
PRDX6,SCN8A,SOD1,TMSB10/TMSB4X
Dementia 2.66E-05 11 AIFM1,APOE,BLMH,CLU,CST3,CTSD,GFAP,GLS,
ITM2B,NCSTN,SOD1
Tauopathy 1.78E-04 10 AIFM1,APOE,BLMH,CLU,CST3,CTSD,GFAP,
NCSTN,SCN8A,SOD1
Neurological signs 1.14E-03 10 AIFM1,APOE,App,AQP4,CLU,GFAP,PPP3R1,
PRDX6,SCN8A,TMSB10/TMSB4X
alzheimer’s disease 5.55E-04 9 AIFM1,APOE,BLMH,CLU,CST3,CTSD,GFAP,
NCSTN,SOD1
huntington’s Disease 2.57E-03 9 AIFM1,APOE,AQP4,CLU,GFAP,PPP3R1,
PRDX6,SCN8A,TMSB10/TMSB4X
Cerebrovascular dysfunction 7.75E-06 7 AIFM1,APOE,AQP4,CLU,CST3,HYOU1,ITM2B
Parkinson’s disease 4.26E-05 7 AIFM1,APOE,AQP4,FTH1,GFAP,MBP,SOD1
Gliosis 4.91E-06 6 ABR,APOE,App,ITM2B,PLP1,SOD1
Amyotrophic lateral sclerosis 3.01E-05 6 AIFM1,APOE,CLU,GFAP,SCN8A,SOD1
Inflammatory demyelinating disease 6.04E-05 6 AQP4,CST3,DNAJB11,MBP,PDIA3,PLP1
Deposition of amyloid fibrils 1.31E-06 4 APOE,App,CLU,ITM2B
Astrocytosis 1.08E-04 4 APOE,App,PLP1,SOD1
Ischemia of brain 2.51E-04 4 AQP4,CLU,CST3,HYOU1
Damage of brain 4.37E-04 4 AQP4,CST3,HYOU1,SOD1
Multiple Sclerosis 4.35E-03 4 CST3,DNAJB11,MBP,PDIA3
Ataxia 6.98E-03 4 App,CTSD,PLP1,SCN8A
Neurodegeneration 8.33E-03 4 APOE,GFAP,PLP1,SOD1
Sporadic amyotrophic lateral sclerosis 2.67E-05 3 CLU,GFAP,SOD1
Cerebral amyloid angiopathy 3.27E-05 3 APOE,CST3,ITM2B
Ischemic injury of brain 2.06E-04 3 AQP4,CST3,HYOU1
Hereditary CNS demyelinating disease 3.04E-04 3 APOE,GFAP,PLP1
Cognitive impairment 2.42E-03 3 APOE,App,CST3
Paralysis 5.23E-03 3 MBP,PLP1,SOD1

328 I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice
A
B

I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice 329
Fig. 2. Validation of A␤PP expression in 5XFAD mice. A) Using a single peptide unique to mouse native A␤PP (TEEISEVK), the expression
level of endogenous A␤PP was analyzed. Note the similar expression levels in wild-type (WT) and transgenic (TG) mice. B) Total A␤PP
expression levels (based on multiple peptides) including both native mouse A␤PP and over-expressed human mutant A␤PP were compared. TG
mice overexpressed A␤PP by approximately threefold.
Table 4
Activated upstream regulator analysis of 5XFAD mice
Upstream Molecule Predicted Activation p-value Target molecules
Regulator Type Activation State z-score of overlap in dataset
XBP1 transcriptional regulator Activated 2.919 3.88E-11 CALR,DNAJB11,ERP29,ERP44,HSP90B1,HYOU1,
PDIA3,PDIA4,PDIA6,PPIB,SOD1
NFE2L2 transcriptional regulator Activated 2.938 2.50E-10 CTSD,DNAJB11,ERP29,FTH1,HSP90B1,OAT,PDIA3,
PDIA4,PDIA6,PPIB,S100A13,SOD1,STIP1
TP53 transcriptional regulator Activated 2.766 7.13E-06 APOE,App,CALU,CLU,CTSD,GATM,IDH1,OAT,
P4HB,PDIA6,PRDX6,SOD1,SORBS1,
STIP1,TMSB10/TMSB4X
IGF1R transmembrane receptor Activated 2.236 5.10E-04 AQP4,CALU,CLU,FTH1,RPL7
pathways contributes to the increased protein expres-
sion in 5XFAD mice (Table 4). Notably, NFE2L2 is
the key component of the NRF2-mediated oxidative
stress response, which is known to be neuroprotective
in AD. Further, TP53 (tumor protein p53) is known to
be upregulated in AD [40], and insulin/IGF-1 signaling
is distorted in AD [41–43].
Analysis of the 5XFAD interaction network
To find protein-protein interaction networks among
regulated proteins, we employed STRING (Search
Tool for the Retrieval of Interacting Genes/Proteins)
analysis [35] for the 87 proteins that were upreg-
ulated (69) or downregulated (28) in 5XFAD mice
Fig. 1. The CID and HCD MS/MS spectra of two representative peptides. The peaks in the CID reveal precursor peptide sequence based on
fragmentation patterns, and the HCD MS/MS spectrum detects reporter ions of m/z 126 to 131, which are used for quantification of the precursor
peptide. A) Tandem MS spectra of peptide LQAEIFQAR from ApoE. HCD shows strong increases in transgenic samples (labeled 126, 128, 130)
compared to wild-type (127, 129, 131). B) Tandem MS spectra of peptide LALDIEIATYR from GFAP, which was also increased in transgenic
samples.

330 I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice
Fig. 3. Validation of quantitative analysis with western blotting. A)
Western blot results for GFAP and nicastrin. B) TMT analysis and
western blotting show quantification of GFAP and nicastrin in good
agreement.
(Tables 1 and 2). In the global STRING-generated
protein-protein network that we named the 5XFAD
interaction network (FADIN, Fig. 5A), several com-
plexes and cellular functions formed prominent, tightly
connected clusters divided by cellular localization. As
shown in Fig. 5B, one network is centered on A␤PP
and its processing machinery in the cell membrane.
Another network of mainly ER-localized proteins is
shown in Fig. 5C.
DISCUSSION
Here, we used quantitative mass spectrometry to
identify proteome-wide changes in the 5XFAD mouse
hippocampus at early stages of AD pathology and val-
idated a subset of the results using immunoblotting.
We found that the FAD mutant A␤PP and PS1 that are
overexpressed in 5XFAD mice induce increases in the
expression of ApoE, ApoJ (clusterin), and nicastrin,
three important proteins in AD that are known to par-
ticipate in A␤processing and clearance, as well as in
the expression of other proteins, including the neu-
rological damage/glial marker protein GFAP. A large
subset of the proteins that were strongly upregulated or
downregulated in 5XFAD brains have been implicated
in neurological disorders such as AD, Parkinson’s
disease, Huntington’s disease, and amyotrophic lat-
eral sclerosis, suggesting a substantial overlap of the
pathological mechanisms of these conditions. The
second-largest subset of regulated proteins was related
to CVD, supporting the hypothesis that there is a
strong association between CVD and AD [39]. Grow-
ing literature indicates that CVD and CVD risk factors
such as hypertension, high LDL cholesterol, low HDL
cholesterol, and diabetes are associated with increased
risk of AD, and together with our results suggest that
the pathology of both diseases might share common
molecular mechanisms.
Common upstream regulator analysis of increased
proteins suggested that the NRF2 and p53 tran-
scriptional pathways were activated, as was IGF-1
signaling. Considerable data points to a potential role
of the insulin/IGF-1 pathway in the pathogenesis of
AD [42]. In particular, AD patients show changes in
insulin/IGF-1 levels as well as insulin receptor/IGF-
1 receptor levels, and their response to insulin is
defective [43]. Our finding that targets of IGF-1R are
upregulated supports the overactivation of this path-
way in AD, which is still a matter of debate. Targets of
the XBP1 pathway were also upregulated, suggesting
the activation of the unfolded protein response [44].
Protein interactome analysis revealed an intercon-
nected network of regulated proteins (FADIN), with
two major sub-networks centered on A␤PP process-
ing membrane complexes and mitochondrial proteins,
respectively. Together with a recent study on the
transcriptome of 5XFAD mice, our study provides a
comprehensive understanding of the molecular events
occurring in 5XFAD mice during the early stages of
AD pathology.
Our results point to strong homeostatic and neu-
roprotective responses in the brains of 5XFAD mice
after injury. ApoE and ApoJ (clusterin), which are
the main chaperones of A␤and are known to coop-
eratively suppress A␤levels [36, 37], had strongly
increased expression in 5XFAD brains. The increased
expression of the ␥-secretase protein complex compo-
nent nicastrin in 5XFAD mice may have occurred in
response to the overexpressed substrate A␤PP or to
match the overexpressed PS1, another component of
the ␥-secretase complex. Although the other subunits

I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice 331
Fig. 4. DAVID Gene Ontology enrichment analysis of proteins increased and decreased in 5XFAD mice. A) Enrichment analysis by biological
process. B) Enrichment analysis by cellular component C) Enrichment analysis by molecular function.
of the ␥-secretase protein complex were not quanti-
fied in mass spectrometry, it is possible that nicastrin,
together with Aph-1 and Pen-2, are co-regulated with
the overexpressed PS1 in a coordinated fashion to
maintain stoichiometry of the functional enzyme com-
plex [45]. The most notable cellular processes among
the proteins that were up- or downregulated in 5XFAD
mice were redox homeostasis and oxidative stress
response, and many proteins were of mitochondrial
origin, suggesting substantial oxidative stress and acti-
vation of defense mechanisms. Moreover, the activated
transcription regulator Nrf2 is the primary cellular
defense against the cytotoxic effects of oxidative stress.
Several mitochondrial molecules important in oxida-
tive stress, such as Nnt [46] and Cox6A1 [47], are
decreased in 5XFAD mice, possibly contributing to
mitochondrial dysfunction. A previous 2D-gel pro-
teomics study with the Tg2576 mouse model also
found several proteins of the respiratory chain reg-
ulated in synaptosomal/mitochondrial fractions [48],
further supporting the role of mitochondrial abnor-
malities in AD. Activated microglia and astrocytes
are abundant in the proximity of neuritic plaques, and
gliosis is known to contribute to the progression of
AD. Although plaques are not readily detected in the
hippocampus of 5XFAD mice at the age we sampled,
proteins known to be expressed by glial cells, such
as ApoE, GFAP, Plp1, and SOD1, were increased in
5XFAD samples, suggesting that gliosis may partially
precede plaque formation.
The 5XFAD mice used in this study overexpress
mutant A␤PP, which leads to global increases in A␤.
A␤exposure is known to block long-term potentia-
tion and exaggerate or induce long-term depression
in the brain [49–51]. Because the reversal of memory-
related long-term potentiation can account for memory
weakening [26, 52–57], we hypothesize that proteins
that were upregulated or downregulated in the present
results may contribute to the process of synapse and
memory impairment. Functional studies targeting pro-
teins that were regulated in this study but have not
been thoroughly explored previously may lead to the
discovery of novel molecular roles in AD and memory.
Because the samples analyzed for this study were
from identical genetic backgrounds except for famil-
ial AD transgenes, our results possess considerable

332 I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice
AB
C
Fig. 5. The 5XFAD interaction network (FADIN) shows an intricate protein-protein interaction network among regulated proteins in 5XFAD
mice (A). The color of label shows nature of regulation (red-increase; blue-decrease), and more strongly regulated proteins are presented with
larger labels. Two prominent sub-networks are evident, divided by cellular localization. B) One network is centered on A␤PP and its processing
machinery in the cell membrane. C) Another network of mainly ER-localized proteins.
advantages over human samples in discerning the
causal pathology of AD. Nevertheless, it is still diffi-
cult to determine whether a given biochemical change
drives the disease or represents a defensive response,
as well as which individual molecular interactions are
key to pathology. Our study will provide a foundation
for further studies exploring the detailed and causal
relationships of AD pathology.
ACKNOWLEDGMENTS
This work was supported by National Research
Foundation of Korea (NRF) grants (2011-0019226 and
2011-0018209) and by a Korea Basic Science Institute
grant (T33615 to YMP, KHK, and SWC). I.H. was
supported by NRF grant # 2012R1A6A3A01019438
and I. M-J was supported by SNUH Research grant #
03-2012-0120.
Authors’ disclosures available online (http://www.j-
alz.com/disclosures/view.php?id=1723).
REFERENCES
[1] Brookmeyer R, Johnson E, Ziegler-Graham K, Arrighi HM
(2007) Forecasting the global burden of Alzheimer’s disease.
Alzheimers Dement 3, 186-191.
[2] Dubois B, Feldman HH, Jacova C, Dekosky ST, Barberger-
Gateau P, Cummings J, Delacourte A, Galasko D, Gauthier
S, Jicha G, Meguro K, O’Brien J, Pasquier F, Robert P,
Rossor M, Salloway S, Stern Y, Visser PJ, Scheltens P (2007)
Research criteria for the diagnosis of Alzheimer’s disease:
Revising the NINCDS-ADRDA criteria. Lancet Neurol 6,
734-746.
[3] Ittner LM, Gotz J (2011) Amyloid-beta and tau–a toxic pas
de deux in Alzheimer’s disease. Nat Rev Neurosci 12, 65-72.
[4] Citron M (2010) Alzheimer’s disease: Strategies for disease
modification. Nat Rev Drug Discov 9, 387-398.
[5] Selkoe DJ (2011) Resolving controversies on the path to
Alzheimer’s therapeutics. Nat Med 17, 1060-1065.
[6] Gandy S, Dekosky ST (2013) Toward the treatment and pre-
vention of Alzheimer’s disease: Rational strategies and recent
progress. Annu Rev Med 64, 367-383.
[7] Petersen RC (2003) Mild cognitive impairment clinical trials.
Nat Rev Drug Discov 2, 646-653.
[8] Blennow K, Hampel H, Weiner M, Zetterberg H (2010) Cere-
brospinal fluid and plasma biomarkers in Alzheimer disease.
Nat Rev Neurol 6, 131-144.
[9] Bateman RJ, Xiong C, Benzinger TL, Fagan AM, Goate A,
Fox NC, Marcus DS, Cairns NJ, Xie X, Blazey TM, Holtz-
man DM, Santacruz A, Buckles V, Oliver A, Moulder K,
Aisen PS, Ghetti B, Klunk WE, McDade E, Martins RN,
Masters CL, Mayeux R, Ringman JM, Rossor MN, Schofield
PR, Sperling RA, Salloway S, Morris JC, Dominantly Inher-
ited Alzheimer N (2012) Clinical and biomarker changes in
dominantly inherited Alzheimer’s disease. N Engl J Med 367,
795-804.
[10] Hardy J, Selkoe DJ (2002) The amyloid hypothesis of
Alzheimer’s disease: Progress and problems on the road to
therapeutics. Science 297, 353-356.

I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice 333
[11] Straub JE, Thirumalai D (2011) Toward a molecular theory of
early and late events in monomer to amyloid fibril formation.
Annu Rev Phys Chem 62, 437-463.
[12] DeStrooper B, Annaert W (2010) Novel research horizons for
presenilins and gamma-secretases in cell biology and disease.
Annu Rev Cell Dev Biol 26, 235-260.
[13] Corder EH, Saunders AM, Strittmatter WJ, Schmechel DE,
Gaskell PC, Small GW, Roses AD, Haines JL, Pericak-Vance
MA (1993) Gene dose of apolipoprotein E type 4 allele and
the risk of Alzheimer’s disease in late onset families. Science
261, 921-923.
[14] Bu G (2009) Apolipoprotein E and its receptors in
Alzheimer’sdisease: Pathways, pathogenesis and therapy. Nat
Rev Neurosci 10, 333-344.
[15] Griffin WS, Sheng JG, Roberts GW, Mrak RE (1995)
Interleukin-1 expression in different plaque types in
Alzheimer’s disease: Significance in plaque evolution. J Neu-
ropathol Exp Neurol 54, 276-281.
[16] Sheng JG, Mrak RE, Griffin WS (1997) Glial-neuronal inter-
actions in Alzheimer disease: Progressive association of
IL-1alpha+ microglia and S100beta+astrocytes with neurofib-
rillary tangle stages. J Neuropathol Exp Neurol 56, 285-290.
[17] Bantscheff M, Schirle M, Sweetman G, Rick J, Kuster B
(2007) Quantitative mass spectrometry in proteomics: A crit-
ical review. Anal Bioanal Chem 389, 1017-1031.
[18] Domon B, Aebersold R (2010) Options and considera-
tions when selecting a quantitative proteomics strategy. Nat
Biotechnol 28, 710-721.
[19] Cox J, Mann M (2011) Quantitative, high-resolution pro-
teomics for data-driven systems biology. Annu Rev Biochem
80, 273-299.
[20] Zhang J, Goodlett DR, Quinn JF, Peskind E, Kaye JA, Zhou
Y, Pan C, Yi E, Eng J, Wang Q, Aebersold RH, Montine TJ
(2005) Quantitative proteomics of cerebrospinal fluid from
patients with Alzheimer disease. J Alzheimers Dis 7, 125-133.
discussion 173-180
[21] Schonberger SJ, Edgar PF, Kydd R, Faull RL, Cooper GJ
(2001) Proteomic analysis of the brain in Alzheimer’s dis-
ease: Molecular phenotype of a complex disease process.
Proteomics 1, 1519-1528.
[22] Oakley H, Cole SL, Logan S, Maus E, Shao P, Craft J,
Guillozet-Bongaarts A, Ohno M, Disterhoft J, Van Eldik L,
Berry R, Vassar R (2006) Intraneuronal beta-amyloid aggre-
gates, neurodegeneration, and neuron loss in transgenic mice
with five familial Alzheimer’s disease mutations: Potential
factors in amyloid plaque formation. J Neurosci 26, 10129-
10140.
[23] Moon M, Hong HS, Nam DW, Baik SH, Song H, Kook
SY, Kim YS, Lee J, Mook-Jung I (2012) Intracellular
amyloid-beta accumulation in calcium-binding protein-
deficient neurons leads to amyloid-beta plaque formation in
animal model of Alzheimer’s disease. J Alzheimers Dis 29,
615-628.
[24] Kimura R, Ohno M (2009) Impairments in remote memory
stabilization precede hippocampal synaptic and cognitive fail-
ures in 5XFAD Alzheimer mouse model. Neurobiol Dis 33,
229-235.
[25] KimKH, Moon M, Yu SB, Mook-Jung I, Kim JI (2012) RNA-
Seq analysis of frontal cortex and cerebellum from 5XFAD
mice at early stage of disease pathology. J Alzheimers Dis 29,
793-808.
[26] Kim J, Lee S, Park K, Hong I, Song B, Son G, Park H,
Kim WR, Park E, Choe HK, Kim H, Lee C, Sun W, Kim
K, Shin KS, Choi S (2007) Amygdala depotentiation and fear
extinction. Proc Natl Acad SciUSA104, 20955-20960.
[27] Xu T, Venable JD, Park SK, Cociorva D, Lu B, Liao L,
Wohlschlegel J, Hewel J, Yates JR (2006) ProLuCID, a fast
and sensitive tandem mass spectra-based protein identifica-
tion program. Mol Cell Proteomics 5, S174-S174.
[28] Washburn MP, Wolters D, Yates JR, 3rd (2001) Large-scale
analysis of the yeast proteome by multidimensional protein
identification technology. Nat Biotechnol 19, 242-247.
[29] Byun K, Young Kim J, Bayarsaikhan E, Kim D, Jeong GB,
Yun KN, Kyeong Min H, Kim SU, Yoo JS, Lee B (2012)
Quantitative proteomic analysis reveals that lipopolysac-
charide induces mitogen-activated protein kinase-dependent
activation in human microglial cells. Electophoresis 33,
3756-3763.
[30] Li Z, Adams RM, Chourey K, Hurst GB, Hettich RL, Pan
C (2012) Systematic comparison of label-free, metabolic
labeling, and isobaric chemical labeling for quantitative
proteomics on LTQ Orbitrap Velos. J Proteome Res 11, 1582-
1590.
[31] Cociorva D, D LT, Yates JR (2007) Validation of tandem mass
spectrometry database search results using DTASelect. Curr
Protoc Bioinformatics Chapter 13, Unit 13 14.
[32] ParkSK, Venable JD, Xu T, Liao L, YatesJR (2006) A tool for
quantitative analysis of high-throughput mass spectrometry
data. Mol Cell Proteomics 5, S199-S199.
[33] Raso C, Cosentino C, Gaspari M, Malara N, Han X,
McClatchy D, Park SK, Renne M, Vadala N, Prati U, Cuda G,
Mollace V, Amato F, Yates JR, 3rd (2012) Characterization of
breast cancer interstitial fluids by TmT labeling, LTQ-orbitrap
velos mass spectrometry, and pathway analysis. J Proteome
Res 11, 3199-3210.
[34] Huang DW, Sherman BT, Lempicki RA (2009) Systematic
and integrative analysis of largegene lists using DAVIDbioin-
formatics resources. Nat Protoc 4, 44-57.
[35] Szklarczyk D, Franceschini A, Kuhn M, Simonovic M, Roth
A, Minguez P, Doerks T, Stark M, Muller J, Bork P, Jensen LJ,
von Mering C (2011) The STRING database in 2011: Func-
tional interaction networks of proteins, globally integrated and
scored. Nucleic Acids Res 39, D561-D568.
[36] DeMattos RB, Cirrito JR, Parsadanian M, May PC, O’Dell
MA, Taylor JW, Harmony JA, Aronow BJ, Bales KR, Paul
SM, Holtzman DM (2004) ApoE and clusterin cooperatively
suppress Abeta levels and deposition: Evidence that ApoE
regulates extracellular Abeta metabolism in vivo.Neuron 41,
193-202.
[37] JiangQ, Lee CY, Mandrekar S, WilkinsonB, Cramer P, Zelcer
N, Mann K, Lamb B, Willson TM, Collins JL, Richardson JC,
Smith JD, Comery TA, Riddell D, Holtzman DM, Tontonoz
P, Landreth GE (2008) ApoE promotes the proteolytic degra-
dation of Abeta. Neuron 58, 681-693.
[38] Yu G, Nishimura M, Arawaka S, Levitan D, Zhang L, Tan-
don A, Song YQ, Rogaeva E, Chen F, Kawarai T, Supala
A, Levesque L, Yu H, Yang DS, Holmes E, Milman P,
Liang Y, Zhang DM, Xu DH, Sato C, Rogaev E, Smith M,
Janus C, Zhang Y, Aebersold R, Farrer LS, Sorbi S, Bruni
A, Fraser P, St George-Hyslop P (2000) Nicastrin modu-
lates presenilin-mediated notch/glp-1 signal transduction and
betaAPP processing. Nature 407, 48-54.
[39] Stampfer MJ (2006) Cardiovascular disease and Alzheimer’s
disease: Common links. J Intern Med 260, 211-223.
[40] Hooper C, Meimaridou E, Tavassoli M, Melino G, Lovestone
S, Killick R (2007) p53 is upregulated in Alzheimer’s disease
and induces tau phosphorylation in HEK293a cells. Neurosci
Lett 418, 34-37.
[41] Vardy ER, Rice PJ, Bowie PC, Holmes JD, Grant PJ, Hooper
NM (2007) Increased circulating insulin-like growth factor-1

334 I. Hong et al. / Quantitative Proteomic Analysis of 5XFAD Mice
in late-onset Alzheimer’s disease. J Alzheimers Dis 12, 285-
290.
[42] Freude S, Schilbach K, Schubert M (2009) The role of IGF-1
receptor and insulin receptor signaling for the pathogene-
sis of Alzheimer’s disease: From model organisms to human
disease. Curr Alzheimer Res 6, 213-223.
[43] Moloney AM, Griffin RJ, Timmons S, O’Connor R, Ravid R,
O’Neill C (2010) Defects in IGF-1 receptor, insulin receptor
and IRS-1/2 in Alzheimer’s disease indicate possible resis-
tance to IGF-1 and insulin signalling. Neurobiol Aging 31,
224-243.
[44] Iwakoshi NN, Lee AH, Vallabhajosyula P, Otipoby KL,
Rajewsky K, Glimcher LH (2003) Plasma cell differentiation
and the unfolded protein response intersect at the transcription
factor XBP-1. Nat Immunol 4, 321-329.
[45] Sato T, Diehl TS, Narayanan S, Funamoto S, Ihara Y, De
Strooper B, Steiner H, Haass C, Wolfe MS (2007) Active
gamma-secretase complexes contain only one of each com-
ponent. J Biol Chem 282, 33985-33993.
[46] Huang TT, Naeemuddin M, Elchuri S, Yamaguchi M, Kozy
HM, Carlson EJ, Epstein CJ (2006) Genetic modifiers of the
phenotype of mice deficient in mitochondrial superoxide dis-
mutase. Hum Mol Genet 15, 1187-1194.
[47] Grant CM, MacIver FH, Dawes IW (1997) Mitochondrial
function is required for resistance to oxidative stress in the
yeast Saccharomyces cerevisiae. FEBS Lett 410, 219-222.
[48] Gillardon F, Rist W, Kussmaul L, Vogel J, Berg M, Danzer K,
Kraut N, Hengerer B (2007) Proteomic and functional alter-
ations in brain mitochondria from Tg2576 mice occur before
amyloid plaque deposition. Proteomics 7, 605-616.
[49] Walsh DM, Klyubin I, Fadeeva JV, Cullen WK, Anwyl R,
Wolfe MS, Rowan MJ, Selkoe DJ (2002) Naturally secreted
oligomers of amyloid beta protein potently inhibit hippocam-
pal long-term potentiation in vivo.Nature 416, 535-539.
[50] Hsieh H, Boehm J, Sato C, Iwatsubo T, Tomita T, Sisodia S,
Malinow R (2006) AMPAR removal underlies Abeta-induced
synaptic depression and dendritic spine loss. Neuron 52, 831-
843.
[51] Haass C, Selkoe DJ (2007) Soluble protein oligomers in
neurodegeneration: Lessons from the Alzheimer’s amyloid
beta-peptide. Nat Rev Mol Cell Biol 8, 101-112.
[52] Hong I, Kim J, Lee J, Park S, Song B, Kim J, An B, Park K,
Lee HW, Lee S, Kim H, Park SH, Eom KD, Lee S, Choi S
(2011) Reversible plasticity of fear memory-encoding amyg-
dala synaptic circuits even after fear memory consolidation.
PLoS One 6, e24260.
[53] HongI, Kim J, Song B, Park K, Shin K, Eom KD, Han PL, Lee
S, Choi S (2012) Fear conditioning occludes late-phase long-
term potentiation at thalamic input synapses onto the lateral
amygdala in rat brain slices. Neurosci Lett 506, 121-125.
[54] Hong I, Song B, Lee S, Kim J, Choi S (2009) Extinction of
cued fear memory involves a distinct form of depotentiation
at cortical input synapses onto the lateral amygdala. Eur J
Neurosci 30, 2089-2099.
[55] Pascoli V, Turiault M, Luscher C (2012) Reversal of cocaine-
evoked synaptic potentiation resets drug-induced adaptive
behaviour. Nature 481, 71-75.
[56] Hong I, Kim J, Song B, Park S, Lee J, Kim J, An B, Lee S,
Choi S (2011) Modulation of fear memory by retrieval and
extinction: A clue for memory deconsolidation. Rev Neurosci
22, 205-229.
[57] Kim J, Song B, Hong I, Lee J, Park S, Eom JY, Lee CJ,
Lee S, Choi S (2010) Reactivation of fear memory renders
consolidated amygdala synapses labile. J Neurosci 30, 9631-
9640.