University of Southern California Dissertations and Theses

Functional proteomic analysis of altered protein signaling modules in Alzheimer's disease

(USC Thesis Other)

Functional proteomic analysis of altered protein signaling modules in Alzheimer's disease

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

FUNCTIONAL PROTEOMIC ANALYSIS OF ALTERED PROTEIN SIGNALING
MODULES IN ALZHEIMER’S DISEASE

by
Stefani Nicole Cottrell Thomas

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PHARMACEUTICAL SCIENCES)

May 2007

Copyright 2007 Stefani Nicole Cottrell Thomas
ii
DEDICATION

I would like to dedicate this work to my parents, Mr. Arthur Lee Thomas and Dr.
Virginia Cottrell Thomas, who continue to provide me with tremendous love, support,
and guidance. I cannot begin to express the magnitude of my sincere gratitude to them
for instilling within me the determination, perseverance and confidence to always
continue to pursue and accomplish my goals.

iii
ACKNOWLEDGEMENTS
First and foremost, I would like to thank my advisor, Dr. Austin Yang, for providing me
with the unique opportunity to be a part of such a dynamic laboratory environment that
promotes the persistent attainment of knowledge in several scientific disciplines. My
growth as a young scientific researcher over the past few years is directly attributable to
Dr. Yang’s enthusiasm for his role as an educator and a scientist, his patience and
guidance, and his commitment to ensuring that I have the requisite foundation for a
successful career.
I would like to express my gratitude to my committee members, Drs. Wei-
Chiang Shen, David Ann, Curtis Okamoto, and P. Elyse Schauwecker, for their
encouragement and support. I would like to thank all of them for their interest in my
research as exhibited by all of their decisions to serve on my Ph.D guidance committee
and on my Ph.D. dissertation committee.
To my mentors, Drs. Joanne Berger-Sweeney of Wellesley College and Wilma
Wasco of Harvard Medical School, I would like to extend my appreciation for their
advice and words of wisdom during my journey as a pre-doctoral student. Also, I am
greatly appreciative to the board of directors of the Society for Neuroscience’s (SfN)
Minority Neuroscience Fellowship Program (MNFP) for awarding me with a federally-
funded pre-doctoral fellowship. It was through SfN’s MNFP that I was able to be
matched with such inspirational mentors, attend the annual SfN meetings in 2003 –
2006, and attend a scientific meeting in addition to the SfN meeting every year that I
was supported by the MNFP fellowship.
iv
I am extremely grateful to Dr. Phyllis Hanson of Washington University who
has so graciously provided Dr. Yang’s laboratory with various cell lines, plasmid DNA
constructs, and antibodies that have been integral to the completion of the research
included in the last chapter of my dissertation.
I would like to thank Dr. Frank Yang of Micro-Tech Scientific from whom I
have learned a considerable amount about the fundamentals and principles of nano-
scale liquid chromatography.
I am grateful for the training that Dr. Brian Soreghan gave me when I joined Dr.
Yang’s laboratory at the University of Southern California and also for the guidance he
provided me with on my initial research projects. I would like to express my gratitude
to Dr. Bingwen Lu and Young Jeng for their assistance with research projects I was
working on while at USC. Thanks to Wade Thompson-Harper and Linda Jankins of the
School of Pharmacy Graduate Affairs Office at USC, I have been able to maintain my
status as a student at USC while I complete my doctoral research at the University of
Maryland, Baltimore.
Finally I would like to express my appreciation to Diane Cripps, Noble
Nemieboka and Drs. Peter Gutierrez, Nanadakumar Madayiputhiya, Tina Tekirian and
Yunhu Wan who have provided a stimulating and nurturing research environment for
me at the University of Maryland, Baltimore. I am thankful for the insightful
discussions I had with Peter, Nandu and Noble pertaining to liquid chromatography and
mass spectrometry applications. I would like to express my gratitude to Tina for her
encouragement, guidance and assistance in the completion of the research included in
v
the final chapter of my dissertation. I am grateful for the aid and insight so graciously
provided by Yunhu in the bioinformatics and computational biology aspect of my
research. I would like to thank Diane for her knowledge of analytical chemistry, which
has been an impetus for the continued scientific progress of Dr. Yang’s laboratory. I
would also like to say thank you to Diane for always providing me with encouragement
and support ever since she joined Dr. Yang’s laboratory at USC.
I am thankful for the wonderful opportunity to learn from these individuals, who
have all been integral to the progress I have made as a pre-doctoral student.

vi
TABLE OF CONTENTS

Dedication ii
Acknowledgements iii
List of Tables viii
List of Figures ix
Abbreviations xii
Abstract xiv
Chapter 1: Introduction 1
1.1 Historical perspective of Alzheimer’s disease (AD) 1
1.2 Neuropathology of AD 2
1.3 Genetic risk factors predisposing to AD 6
1.4 Transgenic mouse models of AD 8
1.5 Amyloid cascade hypothesis 10
1.6 Endosomal and lysosomal dysfunction in AD 14
Chapter 2: Proteomic analysis of oxidatively modified (carbonylated)
proteins in brain tissue of presenilin 1 + amyloid precursor
protein (PS1+APP) transgenic mouse model of Alzheimer’s disease 17
2.1 Abstract 17
2.2 Introduction 18
2.2.1 Affinity purification of carbonylated proteins 20
2.2.2 MudPIT 24
2.2.3 Statistical validation and interpretation of
proteomic data 26
2.2.4 Systems Reconstruction technology for proteomic
data visualization 33
2.3 Methods 34
2.4 Results 37
2.5 Discussion 44

Chapter 3: Reduced neuronal expression of synaptic transmission modulator
HNK-1/neural cell adhesion molecule as a potential consequence
of amyloid β-mediated oxidative stress 51
3.1 Abstract 51

vii
3.2 Introduction 52
3.3 Methods 57
3.4 Results 64
3.5 Discussion 79

Chapter 4: The role of vacuolar protein sorting protein Vps4b in the
development of endosomal protein sorting pathology associated
with Alzheimer’s disease 85
4.1 Abstract 85
4.2 Introduction 86
4.3 Methods 90
4.4 Results 97
4.5 Discussion 120

Bibliography 128

viii
LIST OF TABLES
Table 2.1 Example of a “multi-hit” protein as identified by ProteinProphet
TM
.
Table 2.2 ProteinProphet
TM
assignment of a protein group, as opposed to a single
unique protein, due to peptide degeneracy.

Table 2.3 ProteinProphet
TM
protein identifications based on single peptides.

Table 4.1 Partial list of Vps4b(E235Q)-Myc interacting proteins identified with
high confidence.

ix
LIST OF FIGURES
Figure 1.1 Production of Aβ by β- and γ-secretase processing of APP.
Figure 1.2 Sequence of pathologic events occurring in familial AD as proposed by
the amyloid cascade hypothesis.

Figure 2.1 Schematic representation of MudPIT identification of carbonylated
proteins from mouse brain homogenate.

Figure 2.2 Derivatization of carbonyl group on carbonylated protein with EZ-
Link
TM
Biocytin Hydrazide reagent (Pierce) results in the formation of a
biocytin-hydrazone protein conjugate.

Figure 2.3 Distribution of correct and incorrect peptide discriminating scores (F
scores) as determined by PeptideProphet
TM
for one biocytin hydrazide-
streptavidin affinity-purified carbonylated protein sample from the
digested brain tissue homogenate of one PS1+APP mouse used in this
study.

Figure 2.4 ProteinProphet
TM
protein identification error rate versus sensitivity plot
of the same dataset used to construct Figure 2.3.

Figure 2.5 Functional gene ontology (GO) analysis of biocytin hydrazide-
streptavidin affinity-purified carbonylated proteins from PS1+APP mice.

Figure 2.6 Functional network analysis and reconstructed protein-protein interaction
pathways of carbonylated proteins identified in PS1+APP mouse brain.

Figure 3.1 Structure and biosynthetic pathway of HNK-1.

Figure 3.2 Impaired MTT reduction in neuronal cultures treated with FeSO
4
, Aβ1-
40 and Aβ1-42.

Figure 3.3 Identification of a GlcAT-P peptide by LC-MS/MS.

Figure 3.4 Aβ leads to the down-regulation of HNK-1/NCAM expression in
primary neurons.

Figure 3.5 NCAM expression in primary neurons does not vary with respect to the
concentration of Aβ1-40 or Aβ1-42 added exogenously to the culture
medium.

x
Figure 3.6 HNK-1 protein immunoreactivity is reduced in hippocampus, cingulated
and frontoparietal cortex brain regions of 12 month-old Tg2576 mice
relative to age-matched wild-type (wt) control mice.

Figure 3.7 NCAM immunolabeling in hippocampus does not differ between 12
month-old Tg2576 mice and age-matched wild-type (wt) controls.

Figure 3.8 HNK-1 immunoreactivity is decreased in brain regions exhibiting Aβ
plaque deposition in Tg2576 mice at 12 months of age.

Figure 3.9 HNK-1/CD57 decreases with increasing Aβ deposition in Tg2576 mice.

Figure 4.1 Vps4b catalyzes the dissociation of the ESCRT-III components
following the completion of endosomal cargo sorting into MVBs.

Figure 4.2 Time course of Vps4b(E235Q)-Myc expression in HEK293 cells.

Figure 4.3 Following 4hr of tetracycline-induced expression in HEK293 cells,
Vps4b(wt)-Myc has a diffuse cytoplasmic localization whereas
Vps4b(E235Q)-Myc is localized to the periphery of vacuolar “EQ”
compartments.

Figure 4.4 No statistically significant difference in viability of HEK293 cells
following 4hr vs. 9hr tetracycline-induced (0.75µg/mL) expression of
Vps4b(wt)-Myc and Vps4b(EQ)-Myc.

Figure 4.5 Immunoprecipitation of wild type (wt) and dominant negative (EQ)
Vps4b-Myc with c-Myc antibody.

Figure 4.6 Silver-stained Tris-Glycine SDS-PAGE gel of eluate from Myc
immunoprecipitation of Vps4b(wt)-Myc and Vps4b(E235Q)-Myc
protein interactors.

Figure 4.7 Representative chromatogram from LC-MS/MS analysis of gel segment
E of the 9hr Vps4b(E235Q)-Myc IP sample.

Figure 4.8 MS/MS spectrum of an identified Vps4b peptide (amino acid residues
169-180).

Figure 4.9 Distribution of number of Vps4b(E235Q)-Myc interacting proteins
identified according to their number of unique peptides.

Figure 4.10 Qualitative comparison of number of Vps4b(E235Q)-Myc associating
proteins identified by LC-MS/MS following 4hr and 9hr tetracycline
induction of Vps4b(E235Q)-Myc expression in HEK293 cells.
xi

Figure 4.11 Quantitative relative abundance of Vps4b and ESCRT-III proteins in the
4hr and 9hr induction c-Myc immunoprecipitation samples isolated from
post-nuclear supernatant of Vps4b(E235Q)-Myc expressing HEK293
cells.

Figure 4.12 More Vps4b(E235Q)-Myc interacting proteins involved in cell death,
protein turnover and protein folding were identified in the 9hr induction
sample than in the 4hr sample, as determined by quantitative relative
abundance analysis.

Figure 4.13 MS/MS spectrum of an identified ubiquitin peptide (amino acid residues
12-27) from the 9hr induction sample.

Figure 4.14 Expression of immature N-glycosylated and mature N+O glycosylated
APP in Vps4b(E235Q)-Myc HEK293 cells decreases with increasing
time of induction, whereas the induction of Vps4b(wt)-Myc positively
correlates with increased expression of immature N-glycosylated and
mature N+O glycosylated APP.

Figure 4.15 Production of an APP C-terminal fragment is reduced in HEK293 cells
when Vps4b(E235Q)-Myc expression is induced for 4hrs and 9hrs in
comparison to both non-induced cells (0hr) and the induction of
Vps4b(wt)-Myc.

Figure 4.16 Expression of Vps4b is reduced in frontal cortex tissue of 12 month-old
Tg2576 mice, compared to age-matched wild-type (C57BL) mice.

xii
ABBREVIATIONS
AAA ATPase associated with cellular activities
Aβ amyloid beta
AD Alzheimer’s disease
AGE advanced glycation end product
ApoE apolipoprotein E
APP amyloid precursor protein
BLAST basic local alignment search tool
CHMP charged multivesicular body protein/chromatin modifying protein
ESCRT endosomal sorting complex required for transport
GFAP glial fibrillary acidic protein
GFP green fluorescent protein
GlcAT-P glucuronyltransferase
HEK human embryonic kidney
HNE 4-hydroxynonenol
HNK-1 human natural killer
HPLC high performance liquid chromatography
ICAT isotope coded affinity tag
IL-1β interleukin-1 beta
LC-MS/MS liquid chromatography tandem mass spectrometry
LTP long term potentiation
xiii
MDA malondialdehyde
MTT 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide
MudPIT multidimensional protein identification technology
NCAM neural cell adhesion molecule
NCBI National Center for Biotechnology Information
NFT neurofibrillary tangle
NO nitrogen oxide
NOS nitrogen oxide synthase
NSP number of sibling peptides
NTT number of tryptic termini
PHF paired helical filament
PS1 presenilin 1
PS2 presenilin 2
ROS reactive oxygen species
RP reverse phase
SCX strong cation exchange
ΤΝF−α tumor necrosis factor-alpha
TOT number of MS/MS spectra matching same peptide
VPS vacuolar protein sorting
Xcorr cross-correlation score

xiv
ABSTRACT
Neuritic plaques comprised of amyloid β (Aβ) are one of the primary neuropathological
hallmarks of Alzheimer’s disease (AD). However, Aβ plaque deposition is preceded by
aberrations of the endosomal/lysosomal system, including abnormally enlarged
endosomal compartments, accumulation of protease-resistant proteins, and atypical
activation of the lysosomal system. The functional mechanisms accounting for these
abnormalities have not yet been delineated. Towards our goal of identifying the
proteins whose dysfunction contributes to the development of endosomal/lysosomal
pathology in AD, we have found that: 1) proteins implicated in regulating late
endosomal trafficking are among the targets of oxidation (carbonylation) in the brain of
a presenilin 1/amyloid precursor protein (PS1/APP) transgenic mouse model of AD; 2)
reduced neuronal expression of synaptic membrane protein HNK-1/NCAM is
associated with Aβ pathology in models of Aβ deposition in cell culture and an amyloid
precursor protein (APP) transgenic mouse, and is a potential consequence of impaired
endosomal sorting of proteins that are trafficked to synaptic terminals under non-
pathological conditions; and 3) perturbation of the function of vacuolar protein sorting
protein 4b (Vps4b) in regulating multivesicular body (MVB) biogenesis and endosomal
trafficking results in its abnormal association with proteins involved in cell death and
protein turnover.
Our functional proteomic approach relies upon the specific isolation of sub-
proteomes, or protein interaction modules, that are assembled in a sub-cellular location-
xv
and stimulus-specific manner to carry out distinct cellular tasks. The physical
association of the components of these protein interaction modules results in the
causation of a particular phenotype that may or may not have a role in disease
pathogenesis depending upon the post-translational modification, sub-cellular
localization, and relative abundances of the protein interaction module constituents. We
have applied sensitive mass spectrometry techniques and robust computational biology
methods to the qualitative and quantitative anaylsis of protein interaction modules that
are potentially involved in the causation of the endosomal/lysosomal abnormalities
associated with neurodegeneration in AD using various cell culture and transgenic
mouse models of AD pathogenesis.

1
CHAPTER 1: INTRODUCTION
1.1 Historical perspective of Alzheimer’s disease (AD)
A century ago, at a meeting in Munich, a Bavarian psychiatrist by the name of Alois
Alzheimer was the first to define the clinicopathological syndrome that now bears his
name. In a report published the following year, 1907, Alzheimer detailed the
progressive memory impairment, disordered cognitive function, paranoia, and
progressive decline in memory function that were displayed by one of his patients, a
woman by the name of Auguste D. (7). These observations typify the characteristic
features of what is currently known as Alzheimer’s disease (AD). In the same report,
Alzheimer also described the neuropathology of Auguste D. that included
neurofibrillary tangles and senile plaques. These original observations of Alzheimer
continue to be of relevance to much of the AD research that has been conducted over
the past 100 years.
Alzheimer’s disease is the most common form of dementia among the elderly,
accounting for approximately two thirds of all cases of dementia (172). In 2000, the
estimated prevalence of individuals in the United States with AD was 4.5 million (95).
The percentage of persons with AD increases by a factor of two with approximately
every five years of age; hence ~1% of 60 year-olds and ~30% of 85 year-olds are
afflicted by AD (114). Based on these statistics, and barring any revolutionary break-
through AD therapeutic treatments, the prevalence of AD is predicted to increase to
13.2 million by 2050 (51). With the annual cost of caring for patients with AD totaling

2
~$50 billion (257), there is an urgent need to develop effective therapeutic interventions
for patients with AD.

1.2 Neuropathology of AD
Historically, AD has been characterized by two major neuropathological hallmarks:
extracellular cerebral plaques comprised of 39-43 amino acid residue beta-amyloid (Aβ)
peptides and intraneuronal neurofibrillary tangles containing paired helical filaments of
microtubule-associated tau (134; 202). Initially, extracellular senile plaques comprised
of Aβ were thought to serve as the key correlates of AD-associated cognitive decline
and behavioral deficits (77). However, in recent years, it has become increasingly
apparent that the microscopically visible Aβ plaques are representative of reservoirs of
small, diffusible oligomeric assemblies of Aβ (203), thus raising a question concerning
the exact mechanism whereby Aβ contributes to AD neuropathology – particularly
whether intracellular or extracellular Aβ plays a critical role in the progression of AD
pathology. Several recent studies have suggested that soluble Aβ oligomers function as
stimuli that precede the deposition of Aβ plaques (120; 121; 135; 235; 245), thus
indicating the necessity for studies designed to identify the pathological occurrences
that precede Aβ plaque deposition.
Aβ was initially sequenced from the meningeal blood vessels of patients
diagnosed with cerebrovascular amyloidosis associated with AD (77) and the following
year Aβ was recognized as the primary constituent of the senile plaques that are

3
characteristic of AD patient brain tissue (150). It is now known that Aβ is derived from
proteolytic processing of the amyloid precursor protein (APP), a single transmembrane
glycoprotein that exists as multiple isoforms resulting from alternative splicing of a
single transcript (81; 116; 124; 184; 192; 238). APP is cotranslationally translocated
into the endoplasmic reticulum and then post-translationally modified, or matured,
through the secretory pathway where it acquires N- and O-linked sugars (255).
The gene for APP has been localized to chromosome 21 and trisomy 21
(Down’s syndrome) invariably leads to the neuropathology of AD (neuritic plaques and
neurofibrillary tangles) (18; 48; 145). Several researchers consider the proteolytic
processing of APP to be central to the pathogenesis of AD. While being trafficked
through the secretory pathway en route to the cell surface, most newly synthesized APP
molecules are cleaved into soluble APP α (sAPP- α) by α-secretase which cleaves within
the Aβ sequence, thereby precluding amyloid accumulation (Figure 1.1) (66; 175; 212).
Amyloidogenic processing of APP occurs upon its re-internalization from the plasma
membrane and delivery to endocytic compartments (41; 44; 128; 241; 256; 262) where
it is cleaved by β-secretase, and in some cases also by γ-secretase, which results in the
release of soluble 3- and 4-kDa fragments that contain all or part of the Aβ sequence
(Figure 1.1) (93; 205; 210). Hence, intracellular transport and localization are central
determinants in APP processing and also in the production of Aβ.

Figure 1.1. Production of Aβ by β- and γ-secretase processing of APP. The sequence
within APP that contains Aβ is underlined and depicted by its single-letter amino acid
code. The numbers correspond to the amino acid residue number of the APP770
isoform. Sequential cleavage of APP by β- and γ-secretase (indicated by arrows)
results in the production of Aβ1-40 or Aβ1-42. Cleavage by α-secretase (arrow)
precludes the generation of Aβ. NTF, NH
2
-terminal fragment; CTF, COOH-terminal
fragment.
The major Aβ peptide in aqueous cerebral cortical extracts from AD brain has
been reported as Aβ1-40 (160). However, the predominant form of Aβ present in senile
plaques in AD brain is Aβ1-42 (159; 194). Human neurons in AD-vulnerable brain
regions, namely cortex and hippocampus, specifically accumulate Aβ1-42, but not the
more abundantly secreted Aβ1-40 (88). Results from biochemical studies of synthetic
Aβ peptides indicate that peptides terminating at residue 42 aggregate much more
rapidly than those ending at residue 40 (26; 111). Furthermore, Aβ1-42 is resistant to
degradation and accumulates as insoluble aggregates once it has been internalized by
endocytosis, whereas Aβ1-40 peptides are degraded and do not accumulate following
endocytosis (25; 125; 268; 269). Multi-vesicular bodies (MVBs) have been implicated
4

5
as a major subcellular site of Aβ1-42 accumulation within neurons (236). The
observation that substantial Aβ1-42 accumulation occurs within synaptic compartments
and is associated with cellular pathology, provides a potential molecular basis for
clinical observations correlating the severity of dementia with markers of synaptic loss
(240).
Aside from extracellular cerebral plaques comprised of Aβ peptides, the other
classic neuropathological hallmark of AD is the presence of intraneuronal
neurofibrillary tangles (NFTs) containing paired helical filaments (PHFs) of the
microtubule-associated protein tau. By electron microscopy, PHFs, the unit fibrils of
NFTs, appear as two strands twisted around one another (hence the name “paired”
helical filaments) with a width of 10-22nm and a helical periodicity of 75-80nm (50;
258). The discovery that the highly phosphorylated tau protein is the major component
of PHFs found in the AD brain has brought attention to the nature and enzymology of
the post-translational modifications (PTMs) of this microtubule-associated protein (49;
82; 266). The hyperphosphorylation of tau appears to precede its aggregation into
NFTs in the AD brain and plays an integral role in the loss of its biological function of
stimulating microtubule assembly, gain of its toxicity, dissociation from microtubules,
and its aggregation into PHFs (15; 23; 129).
Part of the complexity of PHF formation involves the determination of multiple
potential phosphorylation sites on each of the six tau isoforms, which have all been
shown to be highly phosphorylated and present in PHFs (79; 80). It is uncertain which
kinase(s) and phosphatase(s) regulate the in vivo phosphorylation of tau and result in its

6
aggregation into insoluble PHFs. Many of the protein kinases that affect tau also affect
other substrates and it has been demonstrated that the down-regulation of certain protein
phosphatases in the AD brain may be only partially responsible for tau
hyperphosphorylation (83; 84).

1.3 Genetic risk factors predisposing to AD
Although there is some uncertainty with respect to the degree to which AD is accounted
for by genetic factors, it is known that clinically typical AD can cluster in families and
can specifically be inherited in an autosomal dominant manner, which is commonly
referred to as “familial AD” (202). However, phenotypic analyses of familial versus
non-familial, or “sporadic”, AD cases have revealed that these two forms of AD are
phenotypically quite similar and in some instances are often indistinguishable (141).
The first genetic mutations – missense mutations resulting in a change in single
amino acids and therefore a change in gene function – causing familial AD were
discovered in the APP gene (78; 101; 162). Most of these mutations are clustered at, or
in close proximity to, the sites within APP that are normally cleaved by α-, β-, and γ-
secretases. It therefore follows that these mutations promote the generation of Aβ by
favoring its amyloidogenic processing by β-, and γ-secretases (29; 42; 234). Families
harboring APP missense mutations that cause AD generally suffer from an early-onset
form of AD (onset prior to age 60) (202). The prevalence of early-onset familial AD
has been estimated as 7% (30).

7
Accounting for the majority (~60%) of early-onset familial AD are mutations in
two related genes, presenilin 1 (PS1) and presenilin 2 (PS2) (138; 209). It has been
reported that human embryonic kidney cells transfected with mutant PS1 and PS2
cDNAs sectrete two- to three-fold elevated levels of Aβ1-42 as compared with cells
transfected with wild-type presenilin genes (43). Additionally, the introduction of
mutant PS1 transgenes into transgenic mice expressing human APP also results in a
significant increase in Aβ1-42 production in mouse brains (43). Taken together, these
results suggest that the PS1/PS2 mutations lead to a specific increase in Aβ secretion
both in vitro and in vivo (199). Missense mutations in PS1 have been found to be
causative of AD in certain families with a clinical onset of AD in their 40s and 50s
(198; 209). Genetic surveys have identified as many as 75 missense mutations in
PS1 and three in PS2 as molecular causes of early-onset AD in several hundred families
from an international population (98).
A major genetic risk factor for late-onset AD (age at onset >60 years) is the ε4
allele of apolipoprotein (apoE4), a protein implicated in cholesterol metabolism that is
also involved in local circuits of lipid turnover participating in membrane repair. A
strong association between the inheritance of apoE4 and both sporadic and late-onset
AD has been described (46; 231). These studies have demonstrated that: 1) carriers of
the ε4 allele have an increased risk of developing AD in an allele-dose-dependent
manner, 2) the apoE4 genotype modulates the age of onset of AD, and 3) the
inheritance of the ε4 allele correlates with increased deposition of Aβ in blood vessels
and plaques and an increased density of senile plaques in the cerebral cortex. It should

8
be noted, however, that the presence of the ε4 allele is neither sufficient nor necessary
for AD to develop (259). Some humans homozygous for the ε4 allele still do not show
AD-like symptoms in their ninth decade of life and beyond and conversely, several
individuals develop AD without harboring ε4 alleles (202).

1.4 Transgenic mouse models of AD
Evidence for the temporal occurrence of specific AD-related pathogenic events has
come from the study of mice transgenic for mutant human APP (70; 104; 119; 131; 132;
135; 235; 239) and from mice that are doubly transgenic for APP in addition to mutant
presenilin (21; 104; 105). Such mouse models of AD display the formation of AD-
associated lesions in a temporally compressed manner, thereby facilitating the
deciphering of the cellular and protein changes that precede and/or accompany neuronal
alteration.
The doubly transgenic presenilin-1
M146L
/amyloid precursor protein
K670N, M671L

(PS1+APP) mouse model of AD has been shown to accumulate Aβ deposits with age
(104). PS1+APP mice are the result of crossing Tg2576 amyloid precursor protein
(APP) mice (human APP695 with a familial Alzheimer’s disease gene mutation
K670N/M671L) (70; 107) with line 5.1 mutant presenilin 1 (PS1
M146L
) mice (61; 104).
The addition of the mutant PS1 transgene in PS1+APP mice potentiates the deposition
of Aβ (21; 104). PS1+APP mice develop large numbers of fibrillar Aβ deposits in
cerebral cortex and hippocampus earlier than their singly transgenic Tg2576 littermates.

9
Both diffuse and fibrillar Aβ deposits accumulate over the life span of PS1+APP
mice, with the pattern of deposition resembling that which is characteristic of AD
brains. In general, amyloid deposition in the PS1+APP mouse starts in the cingulated
cortex between 10-12 weeks of age. Preliminary examinations suggest that the level of
variance in amyloid load among PS1+APP mice at this age is on the order of 25%
(104). Both the size and the number of deposits increase rapidly with age and the
affected brain regions begin to include the hippocampus, corpus callosum and
neocortex. In addition, PS1+APP mice have been generated from three different mutant
lines of PS1 mice and all double transgenic mice show elevated Aβ levels relative to the
APP parental strain (104; 105).
In PS1+APP mice, early Aβ deposits are comprised of fibrillar Aβ resembling
compact Aβ plaques and as the mice age, non-fibrillar Aβ deposits increase in number
and spread to brain regions not typically associated with amyloid plaques in AD (87).
Immunohistochemistry data indicate detectable Aβ immunostaining in PS1+APP mice
older than 3 months, with Aβ deposition increasing in both frontal cortex and
hippocampus tissue up to 9-12 months of age (87). After 12 months of age, Aβ deposits
are detected in the striatum, thalamus, and brain stem.
In contrast, detectable changes in Aβ levels in Tg2576 mice do not occur until
6-7 months of age (119). At 10 months, minimal histological evidence of Aβ
deposition is observed and diffuse plaques are not formed until the mice reach 12
months of age. Also at this age, there is a rapid increase in both diffuse and cored Aβ

10
plaques. Tg2576 mice also display microglial activation, reactive astrocytes with
increased glial fibrillary acidic protein (GFAP), and dystrophic neurites after ~10
months of age. Aβ plaque-associated reactive microglia in these mice also show
enhanced staining for tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β) (16),
two pro-inflammatory cytokines found to be elevated in microglia of brains from AD
patients (60; 92).
Although the Tg2576 and PS1+APP mouse models of AD do not perfectly
replicate the full spectrum of the disease pathology of AD, they clearly replicate at least
some of the most important features of human AD. For example, Tg2576 and
PS1+APP mice not only develop amyloid deposition, but they also concomitantly
display learning and memory deficits, dystrophic neurites, and synaptic dysfunction (9;
135; 235). Additionally, the temporal relationships between the formation of amyloid
aggregates and oxidative damage in both APP and PS1+APP double transgenic mice
have been examined by several different groups (39; 70; 152; 153). The appearance of
Aβ deposits is tightly associated with the increased accumulation of oxidative markers
within these animals.

1.5 Amyloid cascade hypothesis
Aβ is a normal product of APP metabolism throughout life, which can be measured in
tissue homogenate, cerebrospinal fluid, and plasma (94; 205; 210; 247), thus suggesting
that the production of Aβ is not a pathological event. However, the total amount of Aβ
in the brains of patients with late-onset AD has been reported to be ~100-200-fold

higher than in age-matched control patients (90), hence implicating Aβ accumulation as
a central part of the pathogenesis of AD. It has been proposed that Aβ accumulation is
“causative”, either directly or through several downstream events, in AD – an idea that
is the central tenet of the “amyloid cascade hypothesis” as originally proposed by Hardy
in 1996 (97).

Figure 1.2. Sequence of pathologic events occurring in familial AD as
proposed by the amyloid cascade hypothesis.
The main principle of the amyloid cascade hypothesis is that the extracellular
deposition of Aβ is the key initiating event in the pathogenesis of AD, ultimately
11

12
resulting in the dendritic dystrophy, loss of synapses, and the neurotransmitter deficits
that are the basis for AD-associated dementia. A representation of the sequence of
pathogenic events leading to familial AD is presented in Figure 1.2.
According to the amyloid cascade hypothesis, missense mutations in APP, PS1,
or PS2 genes account for the amyloidogenic processing of APP, resulting in increased
Aβ1-42 production and accumulation both in brain interstitial fluid and in intracellular
compartments. Aggregated Aβ1-42 then undergoes oligomerization and is deposited as
diffuse plaques. Aβ1-40 can also be present in the diffuse plaques comprised primarily
of Aβ1-42 which leads to the accumulation of plaque-associated proteins such as
complement factors and cytokines that then trigger an inflammatory response typified
by microglial activation, cytokine release, and astrocytosis. The next events in the
cascade of pathological events are the occurrences of progressive synaptic and neuritic
injury, and the disruption of neuronal metabolic and ionic homeostasis which results in
oxidative injury. Subsequently, the activities of kinases and phosphatases are altered,
thus leading to the hyperphosphorylation of microtubule-associated tau and the
formation of paired helical filaments. Among the terminal steps in the cascade are
widespread neuronal dysfunction and cell death in the hippocampal and cerebral cortex
brain regions, accompanied by progressive neurotransmitter defects. The end-point of
this cascade is dementia, whereby individuals exhibit the progressive memory
impairment, disordered cognitive function, paranoia, and progressive decline in memory
function that Alois Alzheimer originally described a century ago (7).

13
While the amyloid cascade hypothesis has as its foundation mutations in APP,
PS1 or PS2 genes that are observed in cases of familial AD, the subsequent events in
the cascade are common to both familial and sporadic AD. As mentioned earlier,
familial and sporadic AD cases are phenotypically highly similar and in some instances
are often indistinguishable (141).
One particular event in the amyloid cascade hypothesis that will be detailed
further in Chapters 2 and 3 is oxidative injury, or oxidative stress. Oxidative stress
occurs when the production of reactive oxygen species (ROS) exceeds cellular
antioxidant defenses (206; 221). It is generally understood that both aging and
oxidative stress are central risk factors in the development of late onset sporadic AD.
Numerous studies have demonstrated that Aβ accumulation is associated with markers
of oxidative stress including protein oxidation (219; 221), lipid peroxidation (147;
196), advanced glycation end products (220), and nucleic acid oxidation (171).
Although the amyloid cascade hypothesis provides a general framework to
explain the events that occur in AD pathogenesis, there are several observations which
counter its primary assumption that the pathogenic events in the progression of AD
result from the abnormal deposition of Aβ that comprise senile plaques. For example,
the number of Aβ deposits in the brain does not correlate well with the degree of
cognitive impairment experienced by AD patients (143). Additionally, the degree of
dementia in AD patients has been shown, in some cases, to be more closely correlated
with the concentration of soluble Aβ as opposed to the histologically-determined degree
of Aβ plaque formation (157; 247).

14
1.6 Endosomal and lysosomal dysfunction in AD
Among the most striking contradictions to the amyloid cascade hypothesis are
data indicating that endosomal and lysosomal abnormalities, such as Rab5-positive
endosome enlargement, increases in lysosome number, and the up-regulation of the
expression of lysosomal hydrolases, are among the earliest known neuropathological
changes in AD (3; 34; 35; 37; 170; 267). These data suggest that: 1) the deposition of
Aβ is actually a later event in the progression of AD pathogenesis than proposed by the
amyloid cascade hypothesis, and 2) endosomal abnormalities could potentially play a
causative role in Aβ-associated AD pathology.
Endosomal trafficking - trafficking of cargo by both the biosynthetic and
endocytic pathways - plays an important role in integrating inter-cellular
communication, signaling, and metabolism within the cell. In neurons, endosomal
trafficking is a highly specialized process arising from this cell population’s complex
requirements for signal transduction and its polar morphology, in which the regions
most active in signal transduction and inter-cellular communication are located at
substantial distances from the perikaryon, where many of the targets and effectors of
these signaling events are located. The significant spatial separation between various
endosomal sorting events makes neurons particularly vulnerable to modest impairments
of endosomal processes. It is therefore conceivable how defects in endosomal
trafficking events can contribute to the severe neurodegenerative pathology of AD.
Data indicating a substantial reduction in Aβ production when the endocytosis of APP

15
is inhibited or when APP internalization is blocked (128; 180; 227) highlight the
importance of the endosomal pathway to APP turnover and Aβ generation.
During the process of endocytosis, which is both a constitutive and a highly
regulated process, extracellular material and plasma membrane are internalized and
subsequently trafficked through a series of vesicular compartments. In the case of
material that is destined for lysosomal degradation, proteolysis via endocytic trafficking
to lysosomes is in part mediated by covalent modification by ubiquitin (102). The
formation of multivesicular bodies (MVBs), which are endosomal structures formed by
the inward budding of the surface membrane yielding a collection of internal vesicles, is
triggered by ubiquitin and is mediated by a complex of proteins collectively termed
endosomal sorting complex required for transport (ESCRT). Following the fusion of
the MVB with the lysosome, material that is destined for degradation is exposed to
lysosomal hydrolases and is degraded.
The impairment of lysosomal function can reduce autophagocytic capacity
resulting in the inability to efficiently degrade macromolecules, increase cellular
vulnerability to oxidative stress, and ultimately contribute to an apoptotic form of cell
death (233). The re-distribution of lysosomal hydrolases has been implicated in
contributing to the neuropathology of AD (267). Furthermore, the lysosomal
hydrolases, cathepsins D, B, and S, have all been shown to be upregulated in AD brain
as a consequence of lysosomal activation (33; 36; 38; 163; 165; 201).
The proteolytic enzymes responsible for the amyloidogenic processing of APP
have been detected in several subcellular locations, including the plasma membrane,

16
early endosomes, late endosomes, autophagic vacuoles and lysosomes (52; 176; 246).
It is therefore conceivable how endosomal and lysosomal dysfunction could markedly
alter Aβ production and localization. Accordingly, there is a need to decipher the
functional mechanisms underlying the endosomal and lysosomal abnormalities
associated with Aβ pathology as observed in the AD brain. The role of a particular
vacuolar protein sorting protein, Vps4b, in potentially contributing to such endosomal
and lysosomal abnormalities will be explored in Chapter 4.

17
CHAPTER 2: PROTEOMIC ANALYSIS OF OXIDATIVELY
MODIFIED (CARBONYLATED) PROTEINS IN BRAIN TISSUE OF
PRESENILIN 1 + AMYLOID PRECURSOR PROTEIN (PS1+APP)
TRANSGENIC MOUSE MODEL OF ALZHEIMER’S DISEASE
2.1 Abstract
Increasing evidence suggests that oxidative injury is involved in the
pathogenesis of many age-related neurodegenerative disorders, including Alzheimer’s
disease (AD). Identifying the protein targets of oxidative stress is critical to determine
which proteins may be responsible for the neuronal impairments and subsequent cell
death that occurs in AD. In this study, we have applied a high-throughput shotgun
proteomic approach to identify the targets of protein carbonylation in both aged and
PS1+APP transgenic mice. However, because of the inherent difficulties associated
with proteomic database searching algorithms, several newly developed bioinformatic
tools were implemented to ascertain a probability-based discernment between correct
protein assignments and false identifications to improve the accuracy of protein
identification. Assigning a probability to each identified peptide/protein allows one to
objectively monitor the expression and relative abundance of particular proteins from
diverse samples, including tissue from transgenic mice of mixed genetic backgrounds.
This robust bioinformatic approach also permits the comparison of proteomic data
generated by different laboratories since it is instrument- and database-independent.
Applying these statistical models to our initial studies, we detected a total of 117

18
oxidatively modified (carbonylated) proteins, 59 of which were specifically associated
with PS1+APP mice. Pathways and network component analyses suggest that there are
three major protein networks that could be potentially altered in PS1+APP mice as a
result of oxidative modifications: 1) iNOS-integrin signaling, 2) CRE/CBP transcription
regulation and 3) rab-lyst vesicular trafficking. We believe the results of these studies
will help establish an initial AD database of oxidatively modified proteins and provide a
foundation for the design of future hypothesis driven research in the areas of aging and
neurodegeneration.

2.2 Introduction
An increasing amount of data have been generated indicating the Alzheimer’s disease
brain is under increased oxidative stress which could have a causal role in the
neurodegeneration that is characteristic of this debilitating disease (130; 148; 181; 182;
253). The first studies identifying a role for oxidative stress in AD were conducted in
1988 using fibroblasts from AD patients and controls whereby several key enzyme
abnormalities were identified (74). Since then, several lines of evidence that oxidative
damage has more than an associative role in AD have been reported including increased
protein oxidation (219; 221), lipid peroxidation (147; 196), advanced glycation end
products (223), and oxidation of nucleic acids (171) in AD brains versus age-matched
control patients.
Increased reactive carbonyls were the first form of oxidative damage identified
in AD brain tissue (216); however, the detailed pathological significance of this

19
hallmark of oxidative damage remains largely unknown. Carbonylation of proteins is a
widely investigated oxidative protein modification, and the extent of protein
carbonylation is often used as a marker to determine the levels of oxidatively-damaged
proteins in the AD brain. Using combinations of Oxyblot (Western blot-based detection
of protein carbonyl groups derivatized with dinitrophenylhydrazine), 2-D gel
electrophoresis, and mass spectrometry, the identifications of some of the proteins that
are carbonylated to a greater extent in AD brain tissue versus age-matched control
brains have recently begun to be elucidated. For example, Butterfield et al. have been
successful in applying this coupled 2-D fingerprinting-immunological detection
methodology and subsequent identification of proteins by mass spectrometry to disclose
several proteins that are carbonylated in AD brain tissue, including creatine kinase BB,
glutamine synthase, ubiquitin carboxy-terminal hydrolase L-1, dihydropyrimidinease-
related protein, α-enolase, and heat shock cognate 71 (31; 32). However, because it is
known that certain proteins (very basic, acidic, transmembrane, or high molecular
weight proteins) are often excluded or unresolved using 2-D gels, our group has
developed a strategy coupling biocytin hydrazide-streptavidin with on-line
microcapillary reverse-phase liquid chromatography tandem mass spectrometry (LC-
MS/MS) as a method to enrich for and detect carbonylated proteins in the brains of aged
mice and a transgenic mouse model of AD (225; 226).
Although no single transgenic mouse model of AD can completely replicate all
aspects of the human disease, they provide excellent models for studying specific
pathological events due to the expression of the mutated human genes associated with

20
familial AD. The transgenic mouse model of AD chosen for the present study was the
doubly transgenic presenilin-1
M146L
/amyloid precursor protein
K670N, M671L
(PS1+APP)
mouse model of AD, which has been shown to accumulate Aβ deposits with age (104).
The primary advantage of using this transgenic mouse is its accelerated aging
phenotype. Markers of oxidative stress and learning deficits can be detected in these
mice as early as 6 months of age (104). Studies conducted using transgenic mouse
models of AD have demonstrated that Aβ and oxidative damage are inextricably linked
in vivo (217), thereby supporting the use of transgenic animals for the development of
antioxidant therapeutic strategies.

2.2.1 Affinity purification of carbonylated proteins
As with any biological marker of a disease state, a valid biological marker of oxidative
stress should be: 1) specific for the reactive species involved; 2) a chemically and
biologically stable product; 3) determined by an assay that is specific, sensitive, and
reproducible; 4) a major product of oxidative modification that may be directly
implicated in the onset or progression of disease; and 5) representative of the balance
between the generation of oxidative damage and clearance (54). Protein carbonyl
content is widely used as a marker to determine the level of protein oxidation caused
either by direct oxidation of amino acid side chains (e.g. proline and arginine to γ-
glutamylsemialdehyde, lysine to aminoadipic semialdehyde, and threonine to
aminoketobutyrate), or via indirect reactions with oxidative by-products including lipid

21
peroxidation derivatives such as 4-hydroxynonenal (HNE), malondialdehyde (MDA),
and advanced glycation end products (AGEs) (206; 253).
Traditionally, the identification of protein carbonylation has been through the
derivatization of carbonyl groups by 2,4-dinitrophenylhydrazine (DNPH) (218), which
leads to the formation of a stable dinitrophenyl hydrazone product that can be detected
by various immunoassays (enzyme-linked immunosorbent assay,
immunohistochemistry, or Western blot) or by spectrophotometric assay (136; 137;
164; 222). Spectrophotometric DNPH assays can be coupled to protein fractionation by
high performance liquid chromatography (HPLC) to give greater specificity and
sensitivity than measuring total carbonyls in a protein mixture (137). Providing even
more sensitivity are Western-blot assays, which can detect as little as 1pmol carbonyl in
a protein sample and require as little as 50ng protein oxidized to the extent of 0.5mol
carbonyl/mol protein (137).
The afore-mentioned methods are sufficient for detection of the protein carbonyl
content within a complex biological mixture; however two of the goals of proteomic
investigations of protein post-translational modification are the identification of proteins
that are susceptible to modification and the determination of the exact amino acid
residues where the modifications occur. Therefore, it is advantageous to employ a
protein carbonylation detection procedure that is compatible with mass spectrometry.
To this end, our group has developed a method to affinity purify carbonylated proteins
based on a biocytin hydrazide and streptavidin methodology coupled with LC-MS/MS

(Figure 2.1) (225; 226). By using this method, we have been able to identify hundreds
of carbonylated proteins in a single experiment.

Figure 2.1 Schematic representation of MudPIT identification of carbonylated
proteins from mouse brain homogenate.

In our biocytin hydrazide/streptavidin affinity purification scheme, soluble
proteins from tissue homogenate of mouse brain were reacted with biocytin hydrazide
in order to induce the formation of hydrazone conjugates with the carbonyl groups of
oxidatively modified proteins (Figure 2.2).
These biocytin-labeled conjugates were then affinity purified using streptavidin
immobilized on agarose beads. The carbonylated proteins bound to the agarose beads
were subsequently reduced with dithiothreitol, alkylated with iodoacetamide, and
digested with trypsin. The tryptic peptides generated from this approach were identified
using MudPIT, as described in section 2.2.2. (This method can also be applied to the
analysis of insoluble carbonylated proteins in tissue homogenate following treatment
with 70% formic acid).
22

Figure 2.2 Derivatization of carbonyl group on carbonylated protein with EZ-Link
TM
Biocytin Hydrazide
reagent (Pierce) results in the formation of a biocytin-hydrazone protein conjugate.
A point worthy of mention regarding this affinity purification approach is the
general concern that nonspecific binding of non-carbonylated proteins to the
streptavidin-agarose beads will occur. Recently, a mass spectrometric and isotope-
coded affinity tag (ICAT) based approach has been reported that quantitatively
addresses the issue of specificity during protein affinity purification procedures (187).
To differentiate specific complex components from non-specific co-purifying proteins,
a control purification is prepared in which the complex of interest is not enriched and
the samples are labeled with either a “light” or “heavy” stable isotope tag. Following
tryptic digestion and LC-MS/MS analysis, true components of the complex of interest
are distinguished by their increased abundance in the purified sample (protein sample
derivatized using biocytin hydrazide and affinity purified using streptavidin conjugated
to agarose) in comparison to the non-enriched control sample (non-derivatized protein
sample also subjected to streptavidin-agarose affinity purification). It has been
demonstrated that this approach distinguishes between specific complex components
23

24
and non-specific co-purifying proteins, even when the specific components are more
than 20 times less abundant than the co-purifying proteins. The application of this
quantitative mass spectrometry technique has been shown not only to permit
identification of specific protein complex components in partially purified samples, but
also to allow the detection of quantitative differences in protein abundance within the
specific complexes, thereby providing a valuable tool with which to quantitatively
evaluate protein carbonylation: our biocytin hydrazide-streptavidin affinity purification
scheme could be performed prior to the afore-mentioned ICAT procedure, followed by
MudPIT and the statistical interpretation of protein identifications, as described in
sections 2.2.2 and 2.2.3.

2.2.2 MudPIT
The most widely used method of protein resolution and identification has traditionally
been two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) followed by
mass spectrometry (250). In this technique, proteins are separated in the first dimension
by isoelectric point (pI) and in the second dimension by molecular weight (MW).
Typically individual spots on the gel are excised, enzymatically digested, and are
analyzed by mass spectrometry. However, if high throughput analysis of a particular
proteome is desired, requiring the interrogation of numerous gel spots, 2D-PAGE
approaches can become laborious and time-consuming. Furthermore, 2D-PAGE has as
its primary drawback limited pI and MW ranges for efficient protein separation. Hence,
an alternative two-dimensional protein separation scheme was established that employs

25
non-gel-based chromatography to resolve and identify thousands of proteins from
biological samples (75; 140; 156; 249). The result of these efforts is an approach
termed multidimensional protein identification technology, or MudPIT.
A standard MudPIT experiment used two-dimensional liquid chromatography
consisting of a strong cation exchange (SCX) column to first separate peptides based on
their charge, in tandem with a reverse-phase (RP) column to resolve peptides based on
their hydrophobicity. Peptides are first loaded onto an SCX column and a salt step
gradient using a buffer such as ammonium acetate is used to elute a discrete fraction of
the bound peptides onto an RP column. During this process, the peptides are retained
on the RP column while any contaminating salts from the SCX buffer are diverted to
waste. The peptides are then eluted from the RP column into the mass spectrometer
using a gradient of increasing acetonitrile. Finally, the RP column is re-equilibrated in
preparation for binding the next fraction of peptides from the SCX column. An iterative
process of gradients with increasing salt concentrations is used to displace subsequent
fractions of peptides from the SCX column onto the RP column, followed by
acetonitrile gradients to elute the bound peptides from the RP column and into the mass
spectrometer.
When applied to the large-scale analysis of the yeast Saccharomyces cerevisiae
strain BJ450, this MudPIT strategy resulted in the assignment of 5540 peptides to MS
spectra, leading to the detection and identification of ~1480 proteins (249). Additional
studies have applied MudPIT to the characterization of complex biological mixtures
such as rat liver (113), developing mouse brain phosphoproteome (14), sumoylated

26
proteins in human embryonic kidney (HEK) cells (146), human erythroid leukemia cell
line K562 (189), human heart mitochondria (73), and human epidermoid carcinoma cell
line A431 (40). MudPIT has been demonstrated to achieve a dynamic range of 10,000
to 1 between the most abundant and least abundant proteins/peptides in a complex
mixture (260). The sensitivity and dynamic range of this proteomic approach makes it
amenable for use in the discernment of certain protein post-translational modifications
in complex biological samples.
The degree of reproducibility of this type of bottom-up proteomic analysis, in
terms of protein identification, is rather high at the protein level, but in general not at
the peptide level. The issue of the reproducibility of protein identification can be
addressed from both a bioinformatic stand-point (189), (effectively minimizing the false
positive identification rate) as well as from a chromatographic point of view (260)
(optimizing the system’s dynamic range – ratio of the abundance of highly abundant
proteins to low abundance proteins).

2.2.3 Statistical validation and interpretation of proteomic data
Proteomics experiments can result in the generation of very large numbers (hundreds of
thousands) of MS/MS spectra. Computer algorithms are then used to match the
experimental MS/MS spectra with theoretical MS/MS spectra generated from a selected
protein database in order to make peptide and protein identifications. One such
algorithm is termed SEQUEST
TM
which is essentially a peptide fragmentation pattern
matching algorithm.

27
In the SEQUEST
TM
algorithm, peptides with molecular masses matching those
of the experimental MS/MS spectra are first extracted from a protein database. For
each peptide queried from the database, a theoretical MS/MS spectrum is generated.
Each peptide is then given a preliminary score based on the number of predicted
fragment ions from its theoretical MS/MS spectrum that match the experimental
MS/MS spectrum. The top 500 matching peptides are then subjected to a more
extensive ion-matching algorithm in order to generate cross-correlation (Xcorr) scores.
Ultimately, a list of identified proteins, along with their constituent peptides and their
Xcorr scores, is produced and protein identifications are generally evaluated based on
the Xcorr scores. However, one important caveat to keep in mind is that any mass
spectrometry-based protein identification is completely dependent on the sequences that
have been deposited into the database. For example, for any peptide sequence, the
SEQUEST
TM
algorithm will only identify the most likely match for a protein in the
database. As a result, if a protein does not exist in the database or if the sequence is
incorrect, the best-matched protein assigned by SEQUEST
TM
will be incorrect.
Additionally, although it is claimed that the human genome is completely sequenced,
the genomic sequences are still not fully annotated and it is unclear how much of the
remaining sequence may code for proteins (191).
The generation of large numbers of MS/MS spectra in high-throughput
proteomic experiments creates a challenge in not only protein identification, but also in
deciphering correct protein assignments from false identifications. Two probability-
based statistical models that have been developed to assess the validity of peptide and

28
protein identifications are PeptideProphet
TM
and ProteinProphet
TM
. Given that
PeptideProphet
TM
and ProteinProphet
TM
are probability-based programs, they facilitate
the objective comparison of mass spectrometry data, independent of the specific mass
spectrometer used for a particular experiment, the database used for searches, and the
database search algorithm. The use of tools such as PeptideProphet
TM
and
ProteinProphet
TM
aids in ensuring an objective evaluation of the quality of mass
spectrometry data.
Briefly, PeptideProphet
TM
first applies machine learning language to distinguish
correct and incorrect peptide assignments in the dataset being analyzed by using an
expectation maximization algorithm. It uses the observed information about each
assigned peptide in the dataset, learns to distinguish correct from incorrect peptide
assignments, and finally, computes a probability for each assignment being correct.
The program then computes discriminant scores (F scores) for correct and incorrect
peptide assignments using a set of scores generated from SEQUEST
TM
search results;
this set includes Xcorr (cross-correlation score between theoretical and experimental
MS/MS spectra), ΔC
n
(the relative difference between the first and second highest Xcorr
score for all peptides of a given protein queried from the database), Sp Rank (a measure
of how well the assigned peptide scored using a preliminary correlation metric), d
M
(the
absolute difference between the masses of the precursor peptide and the assigned
peptide), and NTT (number of tryptic termini = 0, 1, or 2). Probability scores (p
comp
) for
the peptides assigned to each acquired spectrum are then calculated based on the
distribution of discriminant (F) scores.

29
The list of peptide sequences and their respective p
comp
scores obtained from
PeptideProphet
TM
is then used to determine a minimal list of proteins that can be
correlated with the observed data and used to compute a probability (P
comp
)

that each
protein is indeed present in the original peptide mixture. This is accomplished by
employing the ProteinProphet
TM
program, which uses an expectation maximization
algorithm similar to that used by PeptideProphet
TM
. ProteinProphet
TM
uses the
following scoring system to determine the probability that a protein is present in a
specific sample: 1) NTT: number of tryptic termini, 2) NSP: number of sibling peptides
(different peptide sequences matching the same protein identification), 3) TOT: number
of MS/MS spectra matching the same peptide, and 4) peptide probability score (p
comp
).
After these analyses, the probability of each protein is then ranked from 0 (incorrect) to
1 (correct). Proteins that are represented by numerous peptides, high percentage
sequence coverage, or extremely strong single ion elution profiles are then retained and
are assigned probability scores close to 1.
Additional issues to consider in the statistical validation of mass spectrometry
datasets include: 1) the number of constituent peptides of an identified protein, 2)
repetitive peptide sequences from proteins belonging to the same gene family, 3)
protein identifications based on single peptide “hits”, and 4) borderline identifications.
These issues will be discussed herein within the context of PeptideProphet
TM
and
ProteinProphet
TM
.
Identified proteins with several constituent peptides are often referred to as
“multi-hit” proteins and typically have a high probability. Table 2.1 displays the

30
ProteinProphet
TM
identification of isoform 8 of α-tubulin. The validity of this protein
assignment is high due to its high probability (1.00), sequence coverage (22%) and
multiple constituent peptides (8). In addition, the number of tryptic termini (ntt) is two
for seven out of the eight identified peptides, and the number of sibling peptides (nsp)
and total number of MS/MS spectra matching the same peptide (tot) exceed one.
gi|8394493|ref|NP_059075.1|
probability: 1.00
coverage: 22.0 %
>gi|8394493|ref|NP_059075.1| (NM_017379) tubulin alpha 8; tubulin alpha 8 [Mus musculus]
Peptide sequence Probability Ntt Nsp Tot
FDGALNVDLTEFQTNLVPYPR 1.00 / 1.00 2 6 2
LISQIVSSITASLR 1.00 / 1.00 2 6 2
VGINYQPPTVVPGGDLAK 1.00 / 1.00 2 6 4
QLFHPEQLITGK 1.00 / 1.00 2 6 7
TIQFVDWC*PTGFK 1.00 / 1.00 2 6 3
FDLMYAK 1.00 / 1.00 2 6 7
NLDIERPTYTNLNR 1.00 / 1.00 2 6 8
PPTVVPGGDLAK 0.99 / 0.95 1 6 2
Table 2.1 Example of a “multi-hit” protein as identified by ProteinProphet
TM
. *C indicates modified
cysteine residue (carboxyamidomethylation).

[PROTEIN GROUP 1: gi_mus_musculus_ras_protein_rab33b_oncogene_dbj_source_
family_member_mouse_gtp_related_ rab3a_ref_rab_binding]
Probability: 1.00
-1 gi|346947|pir||A45384 1.00
>gi|346947|pir||A45384 GTP-binding protein rab3D - mouse
Peptide sequence Probability Ntt Nsp Tot
LLLIGNSSVGK 1.00 / 1.00 2 3 4
LQIWDTAGQER 1.00 / 1.00 2 3 4
-2 gi|8394133|ref|NP_058554.1| 0.00
>gi|8394133|ref|NP_058554.1| (NM_016858) RAB33B, member of RAS oncogene
family [Mus musculus]
LQIWDTAGQER 0.99 / 1.00 2 0 4
-3 gi|6679593|ref|NP_033027.1| 0.00
>gi|6679593|ref|NP_033027.1| (NM_009001) RAB3A, member RAS oncogene
family [Mus musculus]
LLLIGNSSVGK 0.99 / 1.00 2 0 4
Table 2.2 ProteinProphet
TM
assignment of a protein group, as opposed to a single unique protein,
due to peptide degeneracy.

31
A main reason why protein inference based upon peptide assignments is a
challenging task, even when statistical models are employed to validate those
assignments, is the issue pertaining to peptide degeneracy – instances where peptide
sequences are present in more than a single entry in the protein sequence database.
Often it is not possible to distinguish the origins of a single peptide if the peptide
sequence is located within a highly conserved domain of a particular gene family. An
example of such a case is given in Table 2.2. The peptide with the amino acid sequence
L(I)QI(L)WDTAGQER is a constituent peptide of two different proteins, GTP-binding
protein Rab 3D, and Rab33B, therefore complicating the assignment of a unique protein
to these peptide sequences. A similar situation occurs with another peptide,
LLLIGNSSVGK which is common to both GTP-binding protein Rab 3D and Rab 3A.
In both of these cases, ProteinProphet
TM
did not make a unique protein assignment, but
rather identified the name of a protein group with a probability of 1.00. The
identification of a specific valid protein will require the presence of a gene-specific
peptide in order to be positively identified.
Lastly, there is the issue pertaining to protein identifications which are based
upon single peptide “hits”. It is sometimes the case that such proteins are present in low
abundance in biological samples, hence when employing proteomic techniques such as
MudPIT, it is advantageous to undergo measures which will enrich the starting material
for a particular protein(s) of interest. Although there are instances where MS/MS data
is generated for only one peptide of a given protein, it is still possible for valid protein
assignments to be made. The statistical model used in ProteinProphet
TM
penalizes, but

32
does not exclude, peptides corresponding to single-hit proteins and rewards those
corresponding to multi-hit proteins (167). Such is the case for the first three proteins
listed in Table 2.3. The peptides identified from these proteins are all tryptic peptides
(ntt = 2) and the probabilities of the proteins are significantly higher than a filter cutoff
value of 0.5. The last two proteins listed in Table 2.3 are examples of borderline data.
As is the case with the first three proteins listed in Table 2.3, these are single-hit
proteins, but their probabilities are lower than the filter cutoff of 0.5 and are thus not
acceptable as valid protein identifications.
gi|21263432|sp|Q91VR2|ATPG_MOUSE 1.00
>gi|21263432|sp|Q91VR2|ATPG_MOUSE ATP synthase gamma chain, mitochondrial
precursor
Peptide sequence Probability Ntt Nsp Tot
THSDQFLVSFK 1.00 / 1.00 2 0 1
gi|400622|sp|P31648|S6A1_MOUSE 0.98
>gi|400622|sp|P31648|S6A1_MOUSE Sodium- and chloride-dependent GABA transporter 1
VADGQISTEVSEAPVASDKPK 0.98 / 1.00 2 0 4
gi|6678195|ref|NP_033331.1| 0.98
>gi|6678195|ref|NP_033331.1| (NM_009305) synaptophysin; Syp I [Mus musculus]
LHQVYFDAPSC*VK 0.98 / 1.00 2 0 2
gi|6755686|ref|NP_035630.1| 0.45
>gi|6755686|ref|NP_035630.1| (NM_011500) striatin, calmodulin binding protein; striatin [Mus
musculus]
AAGDGAAAAGAAR 0.45 / 0.75 1 0 1
gi|484964|pir||PN0510 0.44
>gi|484964|pir||PN0510 integrin beta-3 chain - mouse (fragment)
MC*SGHGQC*NCGDCVCDSDWTGYYC*NC*TTR 0.44 / 0.74 1 0 1
Table 2.3 ProteinProphet
TM
protein identifications based on single peptides. *C indicates modified
cysteine residue (carboxyamidomethylation).

When undertaking proteomic studies involving MudPIT, the statistical filtering
of large-scale datasets becomes a critical factor. The implementation of such statistical
tools as Peptide- and ProteinProphet
TM
which yield computed probabilities that can be
used to estimate false-positive error rates resulting from data filtering, can serve as a
reliable means of publishing large-scale datasets of protein identifications.

33
2.2.4 Systems Reconstruction technology for proteomic data visualization
Among the many available tools for the visualization of data generated from proteomics
experiments is Systems Reconstruction technology. MetaCore
TM
is a program
developed and based on Systems Reconstruction technology which contains over
11,500 human pathways (both metabolic and signal transduction), over 12,000
biochemical reactions and more than 250 manually curated maps for major functional
pathways of cellular processes (168). Curation is carried out on two levels. On the first
level, experimental data from original scientific publications is used to create individual
reactions, short pathways, and signaling interactions. On the second level, larger maps
for major functional blocks are assembled from selected review publications. The
pathways on these maps are linked into larger functional models/blocks via joint
metabolites, signaling proteins and/or regulatory effectors. In these functional signaling
network diagrams, the spatial organization of the proteins is directly proportional to
their known interactions; proteins known to directly interact with each other are
depicted in close proximity, while the distance is greater between proteins whose
interactions are mediated by second messengers, receptors, or a hierarchy of other
proteins.
MetaCore
TM
therefore represents an integrated database on human and
mammalian signaling, regulatory and metabolic pathways, that are interconnected via
associations between genes, proteins, metabolites, pathways and human diseases. These
maps and pathway networks represent the backbone for the integration of several types
of experimental data, such as mRNA expression, protein expression (2D-PAGE or mass

34
spectrometry), protein-protein interaction assays (yeast two-hybrid systems, co-
immunoprecipitation), metabolic profiles, and enzymatic activity. Tools such as
Metacore
TM
can be used to visualize proteomic data towards the elucidation of the
significance of protein-protein interactions within the context of specified signaling
pathways that are known to be perturbed in selected disease processes.

2.3 Methods
Animals
All experimental procedures were subject to the approval of the Institutional Animal
Care and Use Committee of the University of Southern California. Twelve month-old
C57BL mice were acquired from Simonsen Laboratories (Gilroy, CA). Following
anesthetization with chloroform and decapitation, whole forebrains were dissected,
immediately frozen in liquid nitrogen and stored at -80ºC until further use. Frozen
presenilin 1 + amyloid precursor protein (PS1+APP) mouse brain tissue was generously
donated from Dr. Karen Duff at the Center for Dementia Research, Nathan Kline
Institute, New York.
Tissue homogenization
Approximately 50mg of whole brain tissue per mouse (four Tg2576 mice and four age-
matched control C57BL mice) was cut into small pieces on dry ice and subsequently
homogenized in lysis buffer with a pH compatible for subsequent reaction with biocytin
hydrazide reagent [50mM sodium acetate, pH 5.5; 150mM NaCl; 1% Triton X-100;
0.1% SDS; 1mM EDTA; protease inhibitor cocktail containing AEBSF, aprotinin,
leupeptin, bestatin, pepstatin A, and E-64 (Sigma)] using a hand-held tissue disruptor.

35
Tissue homogenate was centrifuged at 18,000 x g for 30min at 4ºC. Supernatant was
used for reaction with biocytin hydrazide reagent.
Protein carbonyl labeling and affinity purification
Aliquots of 300μL of 5mM EZ-Link
TM
biocytin hydrazide (Pierce) dissolved in 100mM
sodium acetate, pH 5.5 were added to 600μL of tissue homogenate supernatant. The
hydrazone conjugation reaction was then allowed to proceed for 1hr at room
temperature (~22ºC) in the dark and 200μL of 100mM Tris, pH 7.5 was added to quench
the reaction. Sodium cyanoborohydride was added to 10mM final concentration in order
to stabilize the reaction products. ImmunoPure immobilized streptavidin agarose slurry
(100μL) (Pierce) was added and samples were rotated overnight at 4ºC. After pulse
centrifugation to pellet the agarose beads containing the bound proteins, the supernatant
was removed and the beads were washed five times with lysis buffer to remove non-
specifically bound proteins and twice with 100mM ammonium bicarbonate, pH 8.5.
Reduction, alkylation, and tryptic digestion
The beads were re-suspended in 400μL of 100mM ammonium bicarbonate, pH 8.5 and
proteins were reduced with 4μL of 1M DTT for 1hr at 56ºC with intermittent inversion
of the samples. Proteins were alkylated with 8μL of 1M Iodoacetamide for 30min at
room temperature in the dark with intermittent inversion. Alkylation reaction was
quenched by addition of 16μL of 1M DTT. Supernatant was removed following pulse
centrifugation and beads were re-suspended in 400μL of 100mM ammonium
bicarbonate, pH 8.5. Enzymatic digestion with 0.5μg of TPCK-treated trypsin (Sigma)

36
was allowed to proceed for 2hr at 37ºC with rotation. Trypsin was added in the same
manner for an additional 2hr incubation followed by a 16hr incubation. After
centrifugation to pellet the agarose beads, supernatant was collected and acetic acid was
added to 5% final vol to stop the trypsin reaction. Digested sample was subsequently
frozen, lyophilized, and stored at -80ºC until analysis by liquid chromatography tandem
mass spectrometry (LC-MS/MS).
Liquid chromatography and mass spectrometry
All samples were analyzed using an LCQ Classic quadrupole ion trap mass
spectrometer (Thermo Finnigan). One dimensional liquid chromatography (LC) was
conducted using an Ultra Plus II high performance liquid chromatography (HPLC)
binary gradient system (Micro-Tech Scientific) equipped with a 15cm x 75μm internal
diameter reverse phase (RP) capillary column with in-tube end frits and packed with
5μm, 300Å pore size C18 particles (Micro-Tech Scientific). Samples were loaded onto
the RP column with 99% solvent “A” (5% acetonitrile, 1% formic acid), and 1% solvent
“B” (95% acetonitrile, 0.8% formic acid). Peptides were eluted with a 100min linear
gradient of 1-95% solvent B. The mass spectrometer was equipped with an LCQ
nanospray ion source (Thermo Finnigan) and a 10μm internal diameter uncoated
SilicaTip
TM
PicoTip
TM
nanospray emitter (New Objective). The spray voltage of the
mass spectrometer was set to 1.5kV and the heated capillary temperature to 160ºC.
MS/MS data generation and analysis
Tandem MS/MS spectra were acquired with Xcalibur version 1.2 software using the
following data-dependent acquisition method: a full MS scan was followed by three

37
consecutive MS/MS scans of the three most abundant (intense) ions from the preceding
full MS scan with the dynamic exclusion feature enabled whereby ions were excluded
from being selected for MS/MS for 3min after being detected in the full MS scan four
times in 1.5min. Data were analyzed using Bioworks 3.1, Beta test-site version
(Thermo Finnigan) using the SEQUEST
TM
algorithm (270) to determine cross-
correlation scores between acquired MS/MS spectra and theoretical spectra generated
from a mouse protein database. Database searches were conducted using a nine node
(2cpu/node) cluster computer (Thermo Finnigan). The following parameters were used
for database searches: protease: no enzyme, molecular weight range: 400-4500,
threshold: 1000, mass: monoisotopic, precursor mass tolerance: 1.4, group scan: 10,
minimum ion count: 20, charge state: auto, and amino acid modifications: differential –
cysteine +57.0520 (carboxyamidomethylation). Results were filtered using
SEQUEST
TM
cross-correlation (Xcorr) score and charge state as follows: Xcorr ≥1.5 for
+1 ions, ≥ 2.0 for +2 ions, and ≥ 2.5 for +3 ions. The quality of peptide assignments
and protein inference was assessed by using the PeptideProphet
TM
and ProteinProphet
TM
statistical models (123; 167).

2.4 Results
As increasing evidence has suggested that oxidative damage in brain tissue is intimately
related to neurodegeneration and Alzheimer’s disease, we undertook an investigation to
identify proteins that may be specifically oxidatively modified (carbonylated) in AD
transgenic mice versus age-matched controls. Traditionally, the identification of protein

38
carbonylation has been through the derivatization of carbonyl groups by 2,4-
dinitrophenylhydrazine (DNPH), followed by immunostaining of DNP conjugated
proteins with anti-DNP antibody. The identifications of DNP-modified carbonylated
proteins are then subsequently determined by 2-D gel electrophoresis/ immunoblottting
analysis or 2-D gel electrophoresis/mass spectrometric analyses. In this study, we
utilized a biocytin hydrazide-streptavidin affinity methodology coupled with MudPIT
that we have previously applied to brain tissue from aged mice (226). However, in the
present study, we incorporated the use of statistical models, PeptideProphet
TM
and
ProteinProphet
TM
, to validate the protein identifications and to circumvent the listings of
hundreds of proteins, the majority of which could very well be false identifications.
The PeptideProphet -generated distribution of the correct versus false peptide
identifications from the digested brain tissue homogenate of one of the PS1+APP mice
used in this study is depicted in Figure 2.3. It is evident that true positives (correct
identifications, indicated by the solid line) and false positives (incorrect identifications,
denoted by the dashed line) have distinct distributions. From these distributions,
PeptideProphet assigns a probability score to each identified peptide, thereby
ascribing a measure of confidence to all of the identified peptides. For example, a
peptide probability score of 0.9 means there is a 90% chance that peptide is identified
correctly.
TM
TM

Correct
Incorrect
Correct
Incorrect

Figure 2.3 Distribution of correct and incorrect peptide discriminating
scores (F scores) as determined by PeptideProphet
TM
for one biocytin
hydrazide-streptavidin affinity-purified carbonylated protein sample
from the digested brain tissue homogenate of one PS1+APP mouse
used in this study.
The PeptideProphet
TM
-assigned peptide probability scores are then used by
ProteinProphet
TM
, along with the number of tryptic termini (NTT), number of sibling
peptides (NSP - different peptide sequences matching the same protein identification),
and total number of MS/MS spectra matching the same peptide (TOT) for each peptide
in order to generate a probability score for the identified proteins. However, as with
peptide identifications, the occurrence of incorrect protein identifications is an
important consideration. The protein identification error rate (percentage of incorrectly
identified proteins among total identified proteins) versus sensitivity (fraction of all
correct peptide assignments accepted by the user) of the same dataset used for Figure
2.3 is plotted as shown in Figure 2.4. As indicated by the arrow in Figure 2.4, for this
dataset, at a sensitivity level of 90% (0.90), the error rate is less than 5% (0.05). That is,
we have the ability to correctly identify 90% of the proteins in the sample with less than
5% of the proteins being identified as false positives.
39

Figure 2.4 ProteinProphet
TM
protein identification error
rate versus sensitivity plot of the same dataset used to
construct Figure 2.3. Arrow indicates an error rate of
~5% at a 90% level of sensitivity.
Using our biocytin hydrazide-streptavidin affinity purification of carbonylated
proteins methodology combined with MudPIT, a total of 117 statistically-validated
proteins were identified among the transgenic PS1+APP and control mice used in this
study. Of those proteins, there were 59 specific to the PS1+APP mice, 42 unique to the
control mice and 16 that were common to both groups.
Difficulties often arise from the management of the large datasets resulting from
experiments involving MudPIT. The datasets created from the presently described
study are no exception. One approach used to visualize our data was functional cellular
process categorization of the identified proteins based on a system established by the
Incyte Human Proteome Survey Database (103). Proteins in the Incyte Human
Proteome Survey Database are classified into 42 different cellular processes that are a
reflection of the biological process ontology adopted by the Gene Ontology (GO)
hierarchical classification system. First, we filtered our datasets by only accepting
proteins with a ProteinProphet
TM
probability value ≥0.5. Next, the National Center for
40

41
Biotechnology Information (NCBI) annotations of the filtered proteins were converted
to LocusLink (currently known as Entrez Gene) IDs using NCBI’s LinkOut program.
The list of LocusLink IDs was then used as input data by the publicly available
GoSurfer (http://bioinformatics.bioen.uiuc.edu/gosurfer/index.htm/) for GO analysis
(272). Finally, all proteins were assigned to one or multiple cellular process categories.
The result of this gene annotation analysis conducted using the dataset from the
PS1+APP transgenic mice is presented in Figure 2.5.
The biocytin hydrazide-streptavidin enriched carbonylated proteins from the
PS1+APP mice analyzed in this study represent 11 cellular processes: apoptosis, motor,
defense/immunity, cell adhesion, transcription regulation, transport, structure, signal
transduction, enzymatic, binding and unknown. Proteins with binding function were the
most predominant, representing 27% of the identified proteins, while those with motor
and apoptotic functions were the least represented, each accounting for only 1% of the
proteins.

unknown
21%
binding
27%
apoptosis
1%
enzyme
16%
signal
transduction
11%
structure
8%
transport
7%
transcription
regulation
4%
motor
1% defense/
immunity
2%
cell adhesion
2%

Figure 2.5 Functional gene ontology (GO) analysis of biocytin
hydrazide-streptavidin affinity-purified carbonylated proteins from
PS1+APP mice. Identified proteins with ProteinProphet
TM

probabilities larger than 0.5 were converted to LocusLink (Entrez
Gene) IDs and used for GO analysis using GoSurfer whereby
molecular functions were assigned.
As another means of data visualization, our datasets were also subjected to more
extensive pathway analyses using MetaCore
TM
.

A functional MetaCore
TM
network
analysis of the biocytin hydrazide-streptavidin affinity-purified carbonylated PS1+APP
mouse brain proteins dataset is shown in Figure 2.6, which demonstrates several key
signaling pathways comprised of such proteins as inducible nitric oxide synthase
(iNOS) and cAMP response element binding protein-binding protein (CBP) that could
potentially be altered in PS1+APP mice. This observation is consistent with previous
reports that both iNOS (58) and CREB (115; 142) are critical in maintaining synaptic
function, which becomes compromised early in the development of AD pathogenesis.
Analysis of the reconstructed protein-protein interaction pathway further suggests that
signaling pathways involving CREB are among those which may be altered in
PS1+APP mice.
42

Figure 2.6 Functional network analysis and reconstructed protein-protein
interaction pathways of carbonylated proteins identified in PS1+APP mouse
brain. Concentric circles adjacent to protein symbols indicate carbonylated
proteins identified in this study. Circled protein names represent proteins
discussed in the text.
In addition, the pathway analysis also indicates that such defects in CREB
signaling could have profound effects on rab-lyst vesicular trafficking. Although the
interaction between CREB and rab proteins is largely unknown, Systems
Reconstruction analysis postulates this CREB-rab interaction could be indirectly
mediated through a p53-mediated cell signaling pathway. The afore-mentioned proteins
are encircled in Figure 2.6. The vectors in Figure 2.6 indicate various other protein-
protein interactions, thereby facilitating the formation of hypotheses with respect to the
potential consequences of protein post-translational modifications, such as
carbonylation, that could alter the nature of protein-protein interactions in numerous
signaling pathways.

43

44
2.5 Discussion
We have used a rapid and sensitive proteomic analysis coupled with
bioinformatic statistical modeling programs to identify carbonylated proteins in brain
homogenate from 12 month-old PS1+APP mice. We utilized an enrichment procedure
that we previously developed for the purpose of identifying protein carbonylation
events in aged and young mice using a biocytin hydrazide and streptavidin affinity
methodology (226). However, in this study, we used the statistical models,
PeptideProphet
TM
and ProteinProphet
TM
(123; 167) to statistically validate the peptide
and protein identifications from our LC-MS/MS acquisition data sets. Our results
showed that we were able to identify 59 proteins unique to the transgenic animal group
(4 animals) and 42 proteins specific to the age-matched control group (4 animals). We
detected 16 proteins that were found in both groups. As a means of visualizing and
integrating the large datasets generated in this study, we employed GO analysis and
Systems Reconstruction technology, thus enabling us to focus on the identified
carbonylated proteins that are potentially involved in various cellular pathways related
to the progression of AD pathogenesis.
A large group of carbonylated proteins identified in PS1+APP mice are those
proteins involved in inflammation and the acute phase response. It has been shown that
several key inflammatory proteins can be identified in senile plaques of AD brains. In
addition, some of these proteins, such as α1-antichymotrypsin, α-macroglobulin,
apolipoprotein E and heparan sulphate proteoglycan, are also known to promote the
aggregation of Aβ. In AD, it is well documented that activated astrocytes and

45
microglia, complement pathways and accumulation of proinflammatory cytokines are
often tightly associated with amyloid plaques and dystrophic neurites (64; 65).
Therefore, it is widely accepted that activation of inflammatory pathways plays a
critical role in the early initiation and propagation of Alzheimer’s disease.
Immunopathological studies of PS1+APP mice have indicated that microglia and
astrocytes are activated by both fibrillar and soluble Aβ and that there is a direct
correlation between increased levels of C1q complement proteins and cyclooxygenase-2
expression in response to depositions of fibrillar Aβ (153). Our results suggest that
various glial markers such as GFAP, C1q and TNF-α receptor are also the targets of
oxidative stress. Although the pathological significances of these modifications remain
largely unknown at this time, earlier studies have demonstrated that C1q plays an active
role in the phagocytosis of Aβ aggregates by microglia (252). ROS modification of
C1q may enhance C1q-Aβ interaction and lead to the accumulation of fibrillar Aβ
(153). It has been shown that C1q-immunoreactivity is specifically associated with
thioflavin-S positive fibrillar staining. Although it is not clear whether C1q is
oxidatively modified in this study, it has been documented that markers of oxidative
stress, such as 4-hydroxynonenol (HNE) and 3-nitrotyrosine, are strongly associated
with fibrillar Aβ in PS1+APP mouse brain (152). The levels of several cytokines, such
as IL-1β (92) and tumor necrosis factor (TNF) (68; 89) in AD have also been reported
to be increased in AD compared to age-matched controls. We did not observe oxidative
modifications of these cytokines; however, our results indicate that various receptors
such as the TNF receptor, scavenger receptor and macrophage-stimulating protein

46
receptor (MSP receptor), are the targets of oxidative stress suggesting that alteration of
cell signaling and proinflammatory pathways could ultimately lead to brain injury and
synaptic dysfunction.
Cytoskeletal proteins, such as β-tubulin and β-actin, have recently been
identified as oxidatively modified in AD by utilizing an imunohistochemical and 2D-
PAGE analysis (5). Furthermore, the degrees of protein carbonylation on these
cytoskeletal proteins were significantly higher in AD brains than in age-matched control
brains, suggesting that there may be a causal relationship between the changes of
cytoskeletal protein functions and these oxidative stress related post-translational
modifications (5). In fact, comprehensive biochemical analyses have revealed that
modification of neurofibrillary tangles (NFT) by HNE can lead to the cross-linking of
NFT and is essential for the initial assembly of paired helical filaments (PHF)
suggesting that the oxidative modification of NFT by lipid peroxide, such as HNE, may
play an important role in the formation of NFT during the course of Alzheimer’s disease
(251).
We detected a series of carbonylated motor proteins, the myosins, which were
present specifically in PS1+APP mice. Myosin proteins are actin-based motors with a
wide variety of functions, including the trafficking and localization of proteins within
intracellular

compartments. Membrane trafficking events are critical to both

pre- and
postsynaptic processes, and knowledge of how myosins

contribute to these processes
will lead to a greater understanding

of neuronal function. In the CNS, myosin V is
highly expressed in hippocampal pyramidal cells,

localizing both to dendritic spines and

47
shafts, and is an abundant

component of the postsynaptic density (PSD) (67; 251).
Therefore, one would predict the alteration of myosin function will have a profound
effect on both pre- and post-synaptic transport.
Several proteins involved in signal transduction and cell signaling were also
identified to be the targets of oxidative stress specifically in PS1+APP mice. Calcium
(Ca
2+
)/calmodulin-dependent protein kinase (CaMKII) is a multifunctional protein
kinase and regulation of CaMKII in hippocampus has been shown

to play an important
role in neuroplasticity, gene expression,

learning, and memory (24; 59; 76; 96; 155).
Besides possible alterations in protein phosphorylation, our initial pathway analysis
study also suggests that nitrogen oxide synthase (NOS)-mediated signaling transduction
could be altered in PS1+APP mice. Although the roles of nitric oxide (NO) and
peroxynitrite (NO
3
-
) in the pathogenesis of AD remain controversial, several recent
studies have indicated that inducible NOS (iNOS) and nitrotyrosine immunoreactivity
could be detected in pyramidal-like neurons and glial cells in the brains of AD patients,
suggesting that the modification of proteins by peroxynitrite can be a key factor in the
early development of AD (133). Furthermore, it has been shown in vitro that Aβ is able
to stimulate the production of iNOS in cultured astrocytes and that this Aβ-mediated
iNOS induction is at least in part regulated by IL-1β and TNF-α-dependent cell
signaling pathways (4). Most importantly, the result of our pathway component
analysis (Figure 2.6) also suggests that there could be a cross-talk between iNOS-
mediated cell signaling and CREB-regulated gene expression in PS1+APP mice,
suggesting that increases in NO-mediated oxidative stress could also lead to changes in

48
CREB mediated gene expression. It is well documented (see (1) for review) that the
cyclic-AMP signaling system and CREB-mediated transcription play key roles in the
conversion of short- to long-term memory and synaptic plasticity. Therefore, our
results suggest that alteration of iNOS and CREB signaling pathways in PS1+APP mice
could have a profound effect on memory consolidation and the progression of AD
pathogenesis.
In addition to being one of the reactants involved in the formation of
nitrotyrosine, NO also plays a critical role in cell signaling. As suggested in Figure 2.6,
the oxidative modifications of integrins such as alpha-4/beta-7, alpha-L/beta-2, alpha-
4/beta-1, and alpha-2b/beta-3 integrin could potentially be linked to the perturbation of
cell signaling pathways that involve iNOS. Microglial cell activation plays a central
role in acute and chronic inflammatory processes associated with neurodegeneration. In
a study of activated neonatal rat microglial cells which express integrins such as CD11b
(integrin alpha b) and CD29 (integrin beta-1), iNOS activity was found to be increased
in addition to a concomitant enhancement in the production of NO (55). Additionally,
in activated macrophages, the activity of iNOS has been found to be associated with
integrin-linked kinase (ILK) (237). Within the context of the reconstructed pathway
analysis as presented in Figure 2.6, these findings could suggest that oxidative
modification of integrins or of iNOS itself could substantially alter the activity of iNOS
and subsequently the concentration of NO, hence also altering cell signaling pathways
involving the NO.

49
It is widely documented that abnormalities of the neuronal endocytic pathways
are very early markers of Alzheimer’s disease (169), and similar endosomal/lysosomal
abnormalities have also been reported in PS1+APP mice as well (3). Although the
mechanism leading to these abnormal endocytic pathways in AD is largely unknown,
we observed that neurobeachin, also known as lysosomal trafficking regulator, is also a
target of oxidative stress in these transgenic mice. It has been shown that neurobeachin
displays a high-affinity binding site (K
d
= 10nM) for the type II regulatory subunit of
protein kinase A (PKA RII) (248). Subcelluar localization studies indicate that
neurobeachin is peripherally associated with pleiomorphic tubulovesicular
endomembranes near the trans sides of Golgi stacks and throughout the cell body and
cell processes. In some cases, neurobeachin has also been observed to concentrate
within a subpopulation of the postsynaptic membrane. Genetic analysis of spontaneous
neurobeachin insertion mutant mice have indicated that the function of neurobeachin

may entail neuron-to-neuron transmission rather than the development of synapses since
the differentiation of synapses was completely normal in neurobeachin null mice (232).
Furthermore, we have also identified that a rab-related GTPase is also a target of ROS.
However, one should also note that the identification of rab-4 in this study is solely
based on a single peptide. This issue becomes critical when a protein is part of a large
protein family such as rab. Therefore, our results in this study only suggest that one of
the rab family proteins is oxidatively modified rather than rab-4 specifically since we
did not observe any MS/MS spectra that are specific for rab-4.

50
In conclusion, the present study describes the application of a high-throughput
proteomic analysis in the analysis of carbonylated proteins in the brains of a PS1+APP
transgenic mouse model of Alzheimer’s disease. By coupling this approach with
several newly developed bioinformatics and network analysis tools, we have tentatively
identified three major pathways 1) iNOS-integrin signaling pathway, 2) CRE/CBP
transcription regulation and 3) rab-lyst vesicular trafficking that could be altered
specifically in these transgenic mice. Since AD is a progressive disease, the elucidation
of the earliest protein targets of oxidation could potentially serve as early markers for
the development of early AD diagnosis during the initial processes of
neurodegeneration.
Given the number of oxidized proteins that were identified as specific to the
transgenic mice, 59, and the number of proteins specific to the age-matched control
mice, 42, the suitability of the PS1+APP mouse as a model of protein oxidation in AD
might come into question. To that end, it should be noted that the identified proteins are
not an exhaustive indication of the carbonylated proteins present in the brains of 12
month-old PS1+APP mice and should merely serve as a benchmark for more conclusive
studies designed to experimentally validate and more precisely elucidate differences in
the classes of proteins that are susceptible to carbonylation in transgenic and control
mice of various ages. The outcomes of the present study should provide an initial
baseline from which new hypothesis-driven types of research can be formed in the area
of aging and various age-related neurodegenerative diseases.

51
CHAPTER 3: REDUCED NEURONAL EXPRESSION OF SYNAPTIC
TRANSMISSION MODULATOR HNK-1/NEURAL CELL ADHESION
MOLECULE AS A POTENTIAL CONSEQUENCE OF AMYLOID β-
MEDIATED OXIDATIVE STRESS
3.1 Abstract
Oxidative stress imparted by reactive oxygen species (ROS) is implicated in the
pathogenesis of Alzheimer’s disease (AD). Given that amyloid β (Aβ) itself generates
ROS that can directly damage proteins, elucidating the functional consequences of
protein oxidation can enhance our understanding of the process of Aβ-mediated
neurodegeneration. In this study, we employed a biocytin hydrazide/streptavidin
affinity purification methodology followed by two-dimensional liquid chromatography
tandem mass spectrometry coupled with SEQUEST bioinformatics technology, to
identify the targets of Aβ-induced oxidative stress in cultured primary cortical mouse
neurons. The Golgi-resident enzyme glucuronyltransferase (GlcAT-P) was a
carbonylated target that we investigated further owing to its involvement in the
biosynthesis of HNK-1, a carbohydrate epitope expressed on cell adhesion molecules
and implicated in modulating the effectiveness of synaptic transmission in the brain.
We found that increasing amounts of Aβ, added exogenously to the culture media of
primary cortical neurons, significantly decreased HNK-1 expression. Moreover, in
vivo, HNK-1 immunoreactivity was decreased in brain tissue of a transgenic mouse

52
model of AD. We conclude that a potential consequence of Aβ-mediated oxidation of
GlcAT-P is impairment of its enzymatic function, thereby disrupting HNK-1
biosynthesis and possibly adversely affecting synaptic plasticity. Considering that AD
is partly characterized by progressive memory impairment and disordered cognitive
function, the data from our in vitro studies can be reconciled with results from in vivo
studies demonstrating that HNK-1 modulates synaptic plasticity and is critically
involved in memory consolidation.

3.2 Introduction
Alzheimer’s disease (AD) is a progressive neurodegenerative disorder characterized by
a gradual loss of cognitive function, and it is the major form of dementia affecting the
elderly. According to the amyloid cascade hypothesis (99), amyloid β (Aβ) peptide
accumulation in the brain is the primary influence driving AD pathogenesis including
synaptic, neuritic and oxidative injuries (236). Concentrations of soluble Aβ less than 1
μM have not been found to be toxic to various types of neuronal cells (185); however,
higher concentrations of fibrillar Aβ are neurotoxic in culture (183). Aβ peptides that
end at residue 42 aggregate more rapidly than those ending at residue 40 and so Aβ1–42
is more resistant to degradation (26). The resistance of Aβ1–42 to degradation explains
why it preferentially accumulates in the brain (125; 268). A structural explanation of
why Aβ1-42 is more prone to form aggregates is its higher metal affinity, in comparison
to Aβ1-40 (10). Aggregated Aβ has the potential to generate oxygen free radicals that

53
are neurotoxic to enzymes (27). Although oxidative injury is only one of the many
causative factors of AD, this sort of injury is believed to play a major role in AD
pathogenesis (130; 221).
Oxidative damage occurs when the production of reactive oxygen species
(ROS
•
) exceeds cellular anti-oxidant defenses (130; 206; 221). Evidence that the AD
brain is subject to increased oxidative stress includes: increased levels of neurotoxic
trace elements Fe
2+
, Al
3+
, and Hg
2+
in the AD brain that are capable of stimulating free
radical generation (63); increased lipid peroxidation and increased levels of 4-
hydroxynonenal (HNE), an aldehyde product of lipid peroxidation (196); the presence
of malondialdehyde (MDA) (265), carbonyls (215), peroxynitrite (PN) (85) and
superoxide dismutase-1 (SOD-1) in neurofibrillary tangles (178); and studies
demonstrating that Aβ is capable of generating free radicals (28). Moreover, with age,
the central nervous system (CNS) becomes particularly vulnerable to free radical
damage as a result of the brain's high oxygen consumption rate, its abundant lipid
content, and a relative lack of antioxidant enzymes compared with other tissues (148).
The current study used the incubation of primary cultured neurons with Aβ as a
model of oxidative stress in the AD brain; however, it should be noted that Aβ is known
to have both antioxidant and oxidant properties (126). At concentrations between 0.1
and 1.0nM, Aβ has been found to have antioxidant properties in human cerebrospinal
fluid (CSF) (127). Aβ was used at concentrations in the micromolar range (1 – 25µM)
in the present study and at these high concentrations, it has pro-oxidant properties. It
has been demonstrated previously that the addition of micromolar concentrations of Aβ

54
to cultured rat hippocampal neurons is neurotoxic and causes the accumulation of ROS
•
(86). Moreover, the neuronal vulnerability to toxic insults mediated by Aβ can be
attenuated by the application of antioxidants such as vitamin E and catalase (221). Aβ
toxicity is thought to be mediated by a direct interaction between Aβ and transition
metals with subsequent generation of ROS
•
(108). Aβ binds metals to its metal-binding
site(s) and then reduces them in its metal-reducing site in order to produce ROS
•
. In
addition, the endogenous secretion of Aβ can be increased when oxidative stress is
induced by micromolar concentrations of Aβ itself, thereby causing a detrimental
feedback loop mechanism of Aβ-mediated neurotoxicity (126).
Protein oxidation is characterized by the covalent modification of a protein
induced either directly by ROS
•
or indirectly by reaction with secondary by-products of
oxidative stress such as HNE, MDA, and PN (17). Irreversible protein oxidative
damage can lead to increased proteolytic susceptibility and decreased biological
activity. Alternatively, oxidative modifications can promote the formation of cross-
linked protein aggregates that are resistant to protease-removal (5). The accumulation
of damaged proteins occurs during the course of both normal and pathological aging.
The goal of this study was to employ a newly developed proteomic approach to
examine a subset of the proteins that become carbonylated as a result of Aβ1-40 or
Aβ1-42-mediated oxidative stress. Carbonyl formation is a widely-used detectable
marker of irreversible protein oxidation (20; 206). Carbonyl derivatives can be formed
by ROS
•
-mediated oxidation of side chains of histidine, arginine and lysine, into ketone

55
or aldehyde derivatives. Proteomic approaches to examine post-translational
modifications such as protein carbonylation predominantly consist of two-dimensional
gel electrophoresis and mass spectrometry; however, a limitation of two-dimensional
gel electrophoresis is its inability to detect low abundance proteins, membrane proteins,
and very acidic or very basic proteins (179). Therefore, we utilized two-dimensional
liquid chromatography tandem mass spectrometry (LC-MS/MS) - a sensitive and rapid
proteomic technique.
Glucuronyltransferase (GlcAT-P) (EC 2.4.1.135) was a carbonylated target that
we decided to investigate further because of its enzymatic role in the biosynthesis of
HNK-1, a neural-specific carbohydrate epitope expressed on cell adhesion molecules
and implicated in modulating the effectiveness of synaptic transmission in the brain.
Our results provide evidence for a correlation between Aβ-induced carbonylation of
GlcAT-P and decreased HNK-1 expression. GlcAT-P is a transmembrane protein that
is localized in the medial- and trans-Golgi apparatus (186) and is the principal enzyme
responsible for the biosynthesis of the sulfoglucuronylgalactose structure in the HNK-1
carbohydrate epitope in the mature brain (174; 264). HNK-1 is a sulfated trisaccharide
with the structure HSO
3
→3GlcAβ (1→3)-Galβ (1→4)-GlcNAc, and was first
recognized by a monoclonal antibody raised against human natural killer cells (2). The
involvement of GlcAT-P in the biosynthetic pathway of HNK-1 and the structure of
HNK-1 are depicted in Figure 3.1.

Figure 3.1 Structure and biosynthetic pathway of HNK-1. (a) The HNK-1
carbohydrate epitope is expressed on glycoproteins and glycolipids as a
sulfated trisacharide, HSO
3
→3GlcAβ (1 →3)-Galβ (1 →4)-GlcNAc. (b)
GlcAT-P is the principal enzyme responsible for the biosynthesis of the
sulfoglucuronylgalactose structure of HNK-1.
HNK-1 is expressed on several cell adhesion molecules such as neural cell
adhesion molecule (NCAM), L1, telencephalin, and tenascin-R (263) and is involved in
modulating the effectiveness of synaptic transmission in the brain (230). GlcAT-P -/-
mice deficient in the HNK-1 carbohydrate epitope exhibit defects in long term
potentiation (LTP) and hippocampus-dependent spatial learning (263). Because AD is
partly characterized by progressive memory impairment and disordered cognitive
function (202), the data from our in vitro studies can be reconciled with results from in
vivo studies demonstrating that HNK-1 modulates synaptic plasticity and is critically
involved in memory consolidation (204; 230; 263).

56

57
3.3 Methods
Tissue culture
All experimental procedures were subject to the approval of the Institutional Animal
Care and Use Committee (IACUC) of the University of Southern California. Timed-
pregnant Swiss Webster mice (15-17 day-old fetuses) were anesthetized with
chloroform and de-capitated. Fetuses were removed using aseptic techniques. Fetal
brains were isolated and placed in Ca
2+
/Mg
2+
-free Hanks’ balanced salt solution (HBSS)
(Cellgro). Dissected cortical tissue was rinsed with Ca
2+
/Mg
2+
-free HBSS and digested
in 5mL papain (Sigma) solution (200µg/mL in Ca
2+
/Mg
2+
-free HBSS) at 37ºC for 15
min. Papain was quenched with fetal bovine serum (Cellgro) and tissues were washed
with Leibovitz’s L-15 medium (Invitrogen) followed by Neurobasal medium
(Invitrogen) supplemented with 2% B27 (Invitrogen), 0.25% L-Glutamine (Sigma),
0.1% Glutamic acid (Sigma), 1% Penicillin/Streptomycin (Cellgro), and 0.7% BSA
(Sigma), and tissues were dissociated by repeated trituration using fire-polished Pasteur
pipettes. Dissociated cells were then plated at 1.0 x 10
6
cells/mL on poly-d-lysine
(Sigma) coated plates and cultured in supplemented Neurobasal medium (Invitrogen),
providing cultures comprised of 90-95% neurons. Cells were maintained for 4-7 days
in vitro (DIV) prior to use in experiments.
Animals
A total of 20 animals were used for quantification studies and included 5 animals per
group with the four groups as follows: 12 month-old wild-type, 24 month-old wild-
type, 12 month-old Tg2576 and 24 month-old Tg2576. Serial sections were

58
immunostained for HNK-1/CD57 and Aβ1-42. All animals were immunostained in a
single experiment to reduce variability.
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reduction assay
The quantification of the reduction of MTT to formazon crystals in cells, a measure of
the function of mitochondrial respiratory chain reactions, was performed as described
previously (154; 161). In brief, following experimental treatment of cultured neurons
(24-well plates), culture media was removed and 0.2mL of MTT solution (0.25mg/mL
in phenol red-free media) was added and allowed to incubate for 90 min at 37°C.
Isopropyl alcohol was added to dissolve the formazon product that was subsequently
transferred to 96-well plates. Absorbance at 560nm was used to quantify the amount of
soluble formazon formed.
Incubation of Aß within neuronal cell cultures
The synthesis of Αβ1-42 and Αβ1-40 peptides has been described previously
(DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVIA) (26). Peak
fractions were collected, frozen in liquid nitrogen, lyophilized, and stored dry at -80ºC
until needed. Aβ1-40 and Aβ1-42 peptides were dissolved in serum-free Neurobasal
medium (Invitrogen) supplemented with B27 without antioxidants (Invitrogen) to
achieve final concentrations of 1, 5, 10, and 25µM. Cultured mouse cortical neurons
were washed twice with 1x phosphate buffered saline (PBS) (Invitrogen) and the Aβ-
containing media was added to the cells (in triplicate) for a 16 hr incubation at 37º C in
5% CO
2
. Following incubation with the peptides, cell viability was assessed via Trypan
Blue exclusion.

59
Immunoblot analysis
Cultured cortical neurons were washed twice with ice-cold 1x PBS (Invitrogen) and
lysed in a buffer containing 1% Triton-X-100, 0.1% SDS, and 1mM EDTA in PBS (pH
7.4) and a protease-inhibitor cocktail with 4-(2-aminoethyl)benzenesulfonyl fluoride
(AEBSF), aprotinin, leupeptin, bestatin, pepstatin A, and E-64 (Sigma). Cell lysates
were subjected to centrifugation at 18,000 x g for 30 min at 4°C and the resultant
supernatants were used for subsequent analyses. Protein concentration was determined
by the bicinchoninic acid (BCA) protein assay (Pierce) and 50µg protein per sample
was used for SDS-PAGE analysis on 10% acrylamide gels unless otherwise noted.
Protein samples were electroblotted onto nitrocellulose membranes and blots were
blocked in 5% nonfat milk, 0.1% Tween-20 in PBS. Blots were incubated with
monoclonal mouse anti-HNK-1/NCAM (1:4,000) (Sigma), monoclonal mouse anti-
actin (1:1,000) (Chemicon) or polyclonal rabbit anti-NCAM (1:5,000) (Chemicon)
antibodies overnight at 4°C. Blots were then incubated in HRP-conjugated goat anti-
mouse (1:20,000) or goat anti-rabbit (1:40,000) antibodies (Pierce) followed by
chemiluminescent detection with Super Signal West Pico Chemiluminescent Substrate
(Pierce). Bands were quantified using densitometric software, UN-SCAN-IT (Silk
Scientific Corporation) and analyzed by Student’s t-test.
Immunohistochemical staining of HNK-1 and NCAM in Tg2576 mouse brain tissue
Tg2576 and wild-type mice (12 months) were sacrificed by an intraperitoneal injection
of sodium pentobarbital and transcardial perfusion with ice cold physiological saline
followed by 4% paraformaldehyde. Tissue was subsequently post-fixed at 4°C

60
overnight and transferred to PBS with 0.02% sodium azide for long-term storage. The
immunostaining procedure was performed with free-floating 50 μm vibratome sections,
and all pretreatments and incubations were at room temperature (23ºC). Sections were
washed in Tris buffered saline (TBS), pH 7.5, followed by a 30 min treatment in 3%
H
2
O
2
, 3% methanol in TBS to quench endogenous peroxidase activity. After several
washes in TBS, followed by TBS with 0.1% Triton X-100, and TBS with 0.1 % Triton
X-100 and 2% bovine serum albumin for 30 min, sections were incubated overnight in
either anti-HNK-1/CD57 (mouse monoclonal, Sigma, 1:100), anti-CD57 Clone NK-1
(mouse monoclonal, Zymed Laboratories Inc.,1:500), anti-NCAM (goat polyclonal,
Santa Cruz Biotechnology Inc., 1:500), or in polyclonal rabbit anti-Aβ1-42 (100).
After rinsing in TBS with 0.1% Triton X-100, and blocking for 30 min in the
blocking buffer mentioned above, sections were incubated in biotinylated-spacer arm
(SP)-Affinipure donkey anti-mouse IgG (Jackson Immunochemicals, 1:200) or donkey
anti-goat (Jackson Immunochemicals, 1:200) for 1 hr. Both secondary antibodies were
selected for minimal cross-reactivity with mouse serum proteins. After several washes,
sections were incubated for 1 hr in an avidin-biotin complex (Vectastain ABC kit
Vector Laboratories). Positive immunoreactivity was visualized with 3’-
diaminobenzidine and hydrogen peroxide (DAB kit, Vector Laboratories). Sections
were mounted on coated slides, allowed to dry overnight, counterstained with cresyl
violet and dehydrated in a series of increasing concentrations of alcohol. Slides were
coverslipped using Depex mounting media. Appropriate controls included sections
without primary or secondary antibodies, and all were negative.

61
Aß1-42 quantification in tissue sections
Aß1-42 quantification was by image analysis to yield load values as described
previously (100). To quantify the extent of HNK-1/CD57 labeling, images were
captured using a 10x objective and the number of positive cells counted. A region
similar to that captured by the serial section was used to quantify Aβ. Quantification
was done blind with respect to group by two independent individuals and the results
averaged.
Affinity purification of carbonylated proteins and tryptic digestion
EZ-Link
TM
biocytin hydrazide (Pierce) dissolved in 100mM sodium acetate (pH 5.5)
was added to cell lysate samples (containing 20µg protein) to a final concentration of
5mM. The samples were incubated for 1 hr at room temperature and 100mM Tris (pH
7.5), was added to each sample (20mM final concentration) to stop the reaction. Sodium
cyanoborohydride (Sigma) was added to achieve a final concentration of 10mM in order
to stabilize the reaction products. ImmunoPure immobilized streptavidin (Pierce) was
added to each sample and samples were rotated overnight at 4°C. After centrifugation
and transfer of the supernatant, agarose beads containing the bound protein were
washed twice with lysis buffer and twice with 100mM ammonium bicarbonate (pH
8.5). After addition of 400 µL 100mM ammonium bicarbonate (pH 8.5) (Sigma) to
each tube, samples were subsequently reduced with dithiothreitol (DTT) (EMD
Chemicals) and alkylated with iodoacetamide (EMD Chemicals). The samples were
then re-centrifuged. Following removal of the supernatant, ammonium bicarbonate was
added, and samples were treated with 0.5 µg tosylphenylalanylchloromethane (TPCK)-

62
treated trypsin (Sigma). The digestion reactions were incubated by rotation at 37°C for
two hours. This method was repeated twice, and the final incubation period spanned 16
hrs. After trypsin treatment, samples were centrifuged, and the supernatants containing
tryptic fragments were collected, acidified with acetic acid (5%) and subsequently
lyophilized pending LC-MS/MS analysis.
LC-MS/MS
Samples were analyzed using a Finnigan LCQ Classic ion trap mass spectrometer
(Thermo Electron). Two-dimensional liquid chromatography was conducted with an
Ultra Plus II Proteomic System (Micro-Tech Scientific) equipped with two 15cm x
75µm (internal diameter) reverse phase (RP) capillary columns (in-tube end frits packed
with 5µm C18, 300Å particles) and a strong cation exchange (SCX) column 10cm x
0.3mm (internal diameter) from Micro-Tech Scientific. Samples were loaded onto the
SCX column with 99% mobile phase solvent A (5% acetonitrile, 1% formic acid) and
1% solvent B (95% acetonitrile, 0.8% formic acid). Peptides were eluted from the SCX
column with 10 salt steps (0, 5, 10, 15, 20, 30, 40, 60, 80, and 100% of 500mM
ammonium acetate). Peptides from each salt step were eluted onto a reverse-phase
column. A linear gradient of 1-99% solvent B for 65 min was used to elute the peptides
from the RP column. Peptides were then eluted directly into the opening ion source of
the mass spectrometer via an LCQ nanospray ion source (Thermo Finnigan) and a
10µm (internal diameter) non-coated SilicaTip
TM
PicoTip
TM
nanospray emitter (New

63
Objective). The electrical contact was made through a liquid junction at the
polyetheretherketone (PEEK) union. The spray voltage of the mass spectrometer was
set to 1.5kV and the heated capillary temperature to 160ºC.
MS/MS data generation and analysis
Tandem MS/MS spectra were acquired with Xcalibur 1.2 software using the following
method: a full MS scan was followed by three consecutive MS2 scans of the top 3 ion
peaks from the preceding full scan using dynamic exclusion (4 repetitions in 1.5 min.
were excluded for 15 min.). Data were analyzed using Bioworks 3.1, Beta-test site
version from Thermo Finnigan, utilizing the SEQUEST algorithm to determine cross-
correlation scores between acquired spectra and an NCBI mouse protein FASTA
database. The following parameters were used for the TurboSEQUEST search
analyses: no enzyme was chosen for the protease as not all proteins are digested to
completion; molecular weight range: 400-4500; threshold: 1000; monoisotopic;
precursor mass: 1.4; group scan: 10; minimum ion count: 20; charge state: auto;
peptide: 1.5; fragment ions: 0; and differential amino acid modifications: Cys 57.0520
(carboxyamidomethylation). Results were filtered using SEQUEST cross-correlation
scores greater than 1.5 for +1 ions, 2.0 for +2 ions, and 2.5 for +3 ions. MS/MS spectra
were verified by manual inspection.

64
3.4 Results
Aß induces increased levels of oxidative stress in primary cultured neurons as
determined by MTT reduction
Studies have indicated that Aβ accumulation is associated with markers of oxidative
stress such as protein oxidation, and that Aβ plays a key role in oxidative stress-evoked
neuropathology within the context of AD (152; 221). Concentrations of Aβ in the
micromolar range have been demonstrated to be neurotoxic and cause the accumulation
of ROS
•
(86). To ensure that our in vitro model of Aß-induced oxidative stress resulted
in the generation of ROS
•
and created an intracellular oxidizing environment, we
assayed for MTT reduction, which can be considered as an index of cellular reducing
power or redox state (207; 208). FeSO
4
was used as a positive control as it is a well-
known inducer of O
2
•-
, H
2
O
2
and
•
OH production (112) and iron has an important
pathophysiological role as a catalyst for free radical generation because it has a loosely
bound electron and has the ability to exist in more than one valence (149). Increased
concentrations of FeSO
4
, Aβ1-40, and Aβ1-42 caused a decrease in MTT reduction
after 24 and 48h (Figure 3.2). The impairment in MTT reduction was notably
statistically significant in neurons incubated for 24h with 5µM FeSO
4
, 25µM Αβ1-40,
and 25µM Aβ1-42 (p < 0.005), thus indicating an increased level of oxidative stress in
neurons subjected to these treatments.

0
20
40
60
80
100
120
Fe 5µM Fe 50µM Aß40 1µM Aß40 25µM Aß42 1µM Aß42 25µM
MTT reduction (% of control)
24hr
48hr
*
*
*
*
**
**
**
0
20
40
60
80
100
120
Fe 5µM Fe 50µM Aß40 1µM Aß40 25µM Aß42 1µM Aß42 25µM
MTT reduction (% of control)
24hr
48hr
*
*
*
*
**
**
**
Figure 3.2 Impaired MTT reduction in neuronal cultures treated with FeSO
4
, Aβ1-
40 and Aβ1-42. Cultured primary cortical mouse neurons were incubated for 24
or 48 hr with vehicle (control), FeSO
4
(5, 50 µM), Aβ1-40 (1, 25 µM) or Aβ1-42
(1, 25 µM). Levels of MTT reduction were quantified and are expressed as
percentages of control values (mean ±SEM; n=3). *p < 0.01, **p < 0.005 versus
control value (Student’s t-test).
Following an incubation period of 24h, the average values of MTT reduction in
neurons exposed to 1 and 25µM Aβ1-40 (32.4 and 18.9% of control values,
respectively) exceeded the average values of MTT reduction in neurons exposed to 1
and 25µM Aβ1-42 (26.7 and 9.5% of control values, respectively). A similar
phenomenon was observed after a 48h incubation period: the average values of MTT
reduction in neurons exposed to 1 and 25µM Aβ1-40 (48.4 and 12.2% of control values,
respectively) were greater than those in neurons exposed to 1 and 25µM Aβ1-42 (10.3
and 5.8% of control values, respectively), suggesting that under these conditions Aβ1-
42 increases the extent of oxidative stress in primary neurons.

Identification of GlcAT-P by LC-MS/MS
In order to characterize some of the proteins that are potentially modified as a result of
Aβ-mediated oxidative damage, a biocytin hydrazide and streptavidin methodology was
65

66
employed as a means of affinity-purifying these carbonylated proteins. Our group
initially applied this method to the identification of oxidatively damaged proteins
subject to carbonylation in aged mouse brain (226). Among the carbonylated proteins
purified from aged mouse brain and identified using this high-throughput proteomic
approach were several low-abundance receptor proteins, mitochondrial proteins
involved in glucose and energy metabolism, in addition to a series of receptors and
tyrosine phosphatases known to be associated with insulin and insulin-like growth
factor metabolism and cell-signaling pathways (226).
In the present study, one of the proteins affinity purified from cell lysate treated
with 25µM Aβ1-42 and identified by the aforementioned proteomic approach was
glucuronyltransferase (GlcAT-P), an enzyme which is involved in the biosynthesis of
HNK-1, a neural specific carbohydrate epitope expressed on glycoproteins. A
representative MS/MS spectrum of an identified GlcAT-P peptide, along with the
coverage of the b-, x-, y-, and z-ions (resulting from collision-induced fragmentation
along the peptide backbone) is presented in Figure 3.3. MS/MS spectra acquired by
mass spectrometry were subjected to SEQUEST database searches and matched with
theoretical MS/MS spectra from an NCBI mouse protein FASTA database. BLAST
database searches using the amino acid sequence (DRDIVEVVR) derived from the
MS/MS spectrum interpreted to be from GlcAT-P confirmed that the sequence is unique
to GlcAT-P and has no similarity with any GlcAT homologues. Other carbonylated
proteins from neurons treated with exogenously added Aβ1-42 (25µM) also identified
by our biocytin hydrazide/streptavidin affinity purification proteomic approach include

cytoskeletal structural proteins (plastin and neurofilament medium polypeptide),
mevalonate kinase (a protein involved in cholesterol metabolism), and regulatory
proteins ( α-chain translation elongation factor eEF-1 and 60kD heat shock protein).

400 600 800 1000 1200
m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Relative Abundance
856.4
601.0
1082.4
502.1
714.1
839.5
599.5
1052.4
813.1
484.0
971.4
583.0
372.9
696.1
926.2 728.6
y3
y4
y5
b*4
b5
yº6
y6
x7
b9
z*8
b8
b6
b3
387.3
y8
985.7
b3 b4 b5 b6 b8
b*4
D R D I V E V V R
z*8 x7 yº6
y8 y7 y6 y5 y4 y3 y1
400 600 800 1000 1200
m/z
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Relative Abundance
856.4
601.0
1082.4
502.1
714.1
839.5
599.5
1052.4
813.1
484.0
971.4
583.0
372.9
696.1
926.2 728.6
y3
y4
y5
b*4
b5
yº6
y6
x7
b9
z*8
b8
b6
b3
387.3
y8
985.7
b3 b4 b5 b6 b8
b*4
D R D I V E V V R
z*8 x7 yº6
y8 y7 y6 y5 y4 y3 y1

Figure 3.3 Identification of a GlcAT-P peptide by LC-MS/MS. Mass spectra
generated by LC-MS/MS analysis (experimental spectra) were subjected to a
SEQUEST database search and matched with theoretical MS/MS spectra from an
NCBI mouse protein FASTA database. Identified b-, x-, y-, and z-ions resulting
from fragmentation of the peptide with the charge retained on the N- terminus (b-
ions) or C-terminus (x-, y- and z-ions) are indicated. * indicates an ion that has
lost an ammonia molecule (NH
3
). ° indicates an ion that has lost a water molecule
(H
2
0).
67

Figure 3.4 Aβ leads to the down-regulation of HNK-1/NCAM expression in primary neurons.
(a) Cultured mouse cortical neurons were incubated for 16h with no peptide (lane 1), Aβ1-40
[1µM (lane 2), 5µM (lane 3), 10µM (lane 4), 25µM (lane 5)] or Aβ1-42 [1µM (lane 6), 5µM (lane
7), 10µM (lane 8), 25µM (lane 9)]. At the end of the incubation, the cells were harvested and
total cell lysates were separated on a 10% Tris-glycine polyacrylamide gel and subjected to
SDS-PAGE analysis. HNK-1 expression was evaluated by immunoblot analysis using an
HNK-1/NCAM monoclonal antibody (145 and 170 kDa). As a control, the immunoblot was
stripped and re-probed with an anti-actin monoclonal antibody (43 kDa). (b) Dose-response
pattern of the effect of Aβ on the ratio of HNK-1/NCAM relative to actin expression. Neurons
were incubated for 16h with either Aβ1-40 ( ■) or Aβ1-42 ( □). HNK-1/NCAM and actin
expression were determined by immunoblotting. The bands were quantified by densitometric
analysis and expressed as values relative to control (no peptide added). Values are means
±SEM for three independent cell preparations. * p < 0.01 versus with control (Student’s t-test).
(c) Relationship between dose and cell death for Aβ1-40 and Aβ1-42. Neuronal viability was
assessed using Trypan Blue exclusion following a 16h incubation period with Aβ1-40 ( ▲) or
Aβ1-42 ( ■). Values are mean ±SEM of three independent cell preparations. *p < 0.05, **p <
0.01 versus control (Student’s t-test).

68

69
Aß leads to the down-regulation of HNK-1/NCAM expression in primary neurons
Because GlcAT-P is directly involved in the biosynthesis of the carbohydrate epitope
HNK-1 and given that the results of our affinity purification strategy and LC-MS/MS
analysis revealed that GlcAT-P is a potential target of Aβ-mediated oxidative stress, we
decided to investigate whether the function of GlcAT-P was altered as a consequence of
Aβ-mediated oxidative modification. As there is no commercially available GlcAT-P
antibody, the function of GlcAT-P was assessed by the expression of HNK-1/NCAM.
Cell lysates from Aβ1-40 and Aβ1-42-treated neurons were subjected to SDS-PAGE
and Western analysis using an anti-HNK-1/NCAM antibody that recognizes both the
145 and 170 kDa forms of HNK-1/NCAM. As shown in Figure 3.4a, there was a
marked reduction in HNK-1/NCAM expression which was directly related to the
amount of Aβ added to cultured neurons. To ensure that the decrease in HNK-
1/NCAM expression was not attributed to a reduction in total protein synthesis during
the 16 hour incubation of the cultured neurons with Aβ and to ensure that an equal
amount of protein was loaded into each lane, the immunoblot was stripped and re-
probed with an anti-actin monoclonal antibody (43 kDa). The decreased expression of
HNK-1/NCAM with respect to actin expression, for each condition of neuronal
treatment with Aβ, from three independent experiments, was quantified by
densitometric analysis and is depicted graphically in Figure 3.4b.
The ratio of HNK-1/NCAM to actin expression decreased with increasing
concentrations of Aβ, with Aβ1-42 having a more pronounced effect than Aβ1-40.
The dose-dependent decrease in HNK-1/NCAM : actin expression was statistically

70
significant (p < 0.01) at a concentration of 25µM Aβ1-42, whereas the dose-dependent
decrease in HNK-1/NCAM : actin expression in Aβ1-40-treated neurons did not reach
statistical significance at any of the conditions used for the experiment. Expression of
HNK-1/NCAM : actin was reduced by ~30% in cells treated with 10µM Aβ1-40, and by
~35% in cells treated with 10µM Aβ1-42 compared with control cells (Figure 3.4b).
The effect of Aβ1-42 on HNK-1/NCAM expression was exacerbated at a concentration
of 25µM (Figure 3.4a, lane 9). Cells treated with 25µM Aβ1-42 showed an ~80%
reduction in HNK-1/NCAM : actin expression compared with control cells (Figure
3.4b), whereas there was a minimal difference in the reduction of HNK-1/NCAM
expression between cells incubated with 25µM Aβ1-40 (Figure 3.4a, lane 5) and those
incubated with 10µM Αβ1-40 (Figure 3.4a, lane 4).
The cell death curve shown in Figure 3.4c depicts the extent of cell viability in
response to treatment with various concentrations of Aβ. Over the range of peptide
concentrations used for the present study, 1 – 25µM, cell death only reached statistical
significance following incubation with 10 and 25µM Aβ1-42. Following a 16 hour
incubation with 25µM Aβ, we assessed cell viability to be ~75% for neurons treated
with Aβ1-42 compared with ~87.5% for Aβ1-40 treated neurons; however, HNK-
1/NCAM expression was nearly abolished at this concentration of Aβ1-42 (~20% of
control expression levels), whereas ~70% of the control levels of HNK-1/NCAM
expression remained in Aβ1-40 treated neurons (Figure 3.4b). This suggests that

although cell survival is affected by Aβ, cell death probably does not account for the
observed decrease in HNK-1/NCAM expression.

Figure 3.5 NCAM expression in primary neurons does not vary with respect to the concentration
of Aβ1-40 or Aβ1-42 added exogenously to the culture medium. (a) Cultured Aβ-treated mouse
cortical neurons were harvested following a 16h incubation and cell lysates were resolved using
SDS-PAGE on 4-20% Tris-glycine polyacrylamide gels. NCAM expression was determined by
immunoblot analysis using an anti-NCAM polyclonal antibody that detects NCAM of 200-250
kDa. As a control, the immunoblot was stripped and re-probed with an anti-actin monoclonal
antibody (43 kDa). (b) Quantification of NCAM expression relative to actin expression in Aβ-
treated primary neurons. The bands were quantified by densitometric analysis and are expressed
as values relative to control (no peptide added). Values are mean ±SEM for three independent
cell preparations. There were no statistically significant differences relative to control.
To address the possibility that the observed Aβ concentration-dependent
reduction in HNK-1/NCAM expression could be attributed to a decrease in the
expression of NCAM, as opposed to a reduction in the expression of the HNK-1
carbohydrate epitope, lysate from Aβ-treated neuronal cells was subjected to
immunoblot analysis with a polyclonal NCAM antibody. As demonstrated in Figure
71

3.5a, the expression of NCAM, as opposed to that of the HNK-1 carbohydrate moiety,
did not vary with respect to the concentration of exogenously added Aβ. As described
for Figure 3.4, the immunoblot was stripped and re-probed with an actin antibody to
control for consistency in total protein synthesis during the 16 hour incubation period
with Aβ and to control for the amount of protein per sample used for electrophoretic
separation. The expression of NCAM with respect to actin expression for each
condition of neuronal treatment with Aβ, from three independent experiments, was
quantified by densitometric analysis and is depicted graphically in Figure 3.5b. There is
evidence of a relatively consistent level of NCAM expression, compared to control
values, regardless of peptide concentration. This serves to support the conclusion drawn
from the results shown in Figure 3.4, namely that the decreased HNK-1/NCAM
reactivity is indicative of reduced expression of HNK-1.

HNK-1 expression is reduced in hippocampal and cortical brain regions of Tg2576
mice at 12 months of age, relative to age-matched wild-type control mice
The expression of HNK-1/NCAM in human AD frontoparietal cortex brain tissue was
assayed by using whole tissue homogenate for Western blot analyses; however we did
not detect any difference in the expression of HNK-1/NCAM between AD and control
tissue (data not shown). We therefore concluded that the tissue-level expression of
HNK-1/NCAM would be more appropriately assessed by immunohistochemical
methods (Figures 3.6-3.9).

72

73
To validate the results obtained using our model of oxidative stress in the AD
brain – treatment of primary cortical mouse neurons with varying concentrations of
exogenous Aβ – immunohistochemical analysis of the distribution and extent of HNK-1
labeling was performed using tissue from the hippocampus, cingulate cortex and
frontoparietal cortex brain regions of Tg2576 and age-matched wild-type mice. Tg2576
mice harbor a mutation in the amyloid precursor protein (K670N, M671L) and display
markedly elevated Aβ levels at an early age. By 9-12 months of age, these mice
develop characteristic extracellular AD-type Aβ deposits in the cortex and hippocampus
(104).

Figure 3.6 HNK-1 protein immunoreactivity is reduced in hippocampus, cingulated
and frontoparietal cortex brain regions of 12 month-old Tg2576 mice relative to age-
matched wild-type (wt) control mice. Free-floating sections 50µm thick from wt
and Tg2576 mice were incubated in anti-HNK-1 antibody (brown label) and counter
stained with cresyl violet (purple) to detect differences in the distribution and extent
of neuronal labeling. (a) More neuronal labeling using an antibody against HNK-1
was present in wt mice in area CA1 than in Tg2576 mice. In addition, wt mice
showed more extensive neuronal labeling for HNK-1 in the cingulate and
frontoparietal cortices. (b) Higher magnification photographs illustrate that reduced
HNK-1 labeling in the frontoparietal cortex of Tg2576 mice appeared in both cortical
layers with layer V appearing most affected. Individual neurons also showed
differences in the distribution of HNK-1 with wt animals showing immunostaining
on neuronal membranes and dendrites (arrows) whereas immunolabeling was more
limited to the plasma membrane in Tg2576 animals. Scale bar 500μm in (a), 100µm
in (b) top panel and 20μm in (b) bottom panel.
As shown in Figure 3.6a, immunostaining for HNK-1 (labeled with anti-HNK-
1/CD57 antibody) was reduced in the hippocampus, cingulate cortex, and frontoparietal
cortex of 12 month-old Tg2576 mice as compared with wild-type control mice of the
74

75
same age. Analysis of magnified images of serial sections from the frontoparietal
cortex revealed that not only was the staining of HNK-1 decreased in Tg2576 mice at
12 months of age, but also that the specific cellular region(s) of neuronal HNK-1
staining differed (Figure 3.6b). In neurons of 12 month-old wild-type mice, HNK-1
staining was observed on the periphery of the cell soma, in addition to along the neuritic
processes. In contrast, in 12 month-old Tg2576 mice, the HNK-1 staining was localized
primarily on the periphery of the neuronal cell soma and there was a striking lack of
HNK-1 immunoreactivity along the neuritic processes. Similar results were observed
using a second commercial source of anti-HNK-1 antibody (results not shown). Given
that the formation of neuritic processes plays an integral role in learning and memory,
the observation that neurons from 12 month-old Tg2576 mice lack HNK-1 staining in
this neuronal region suggests a possible role for HNK-1 in synaptic plasticity, which is
known to be altered in the brains of patients with AD.
In order to determine the robustness of the HNK-1/CD57 staining shown in
Figure 3.6, a second set of studies was conducted using an NCAM-specific antibody.
The images in Figure 3.7 demonstrate no observable differences in hippocampal NCAM
expression between 12 month-old Tg2576 and wild-type control mice.

Figure 3.7 NCAM immunolabeling in
hippocampus does not differ between 12
month-old Tg2576 mice and age-matched
wild-type controls. Free-floating sections
50µm thick were immunostained with anti-
NCAM (brown label) to determine the level
and distribution of labeling in Tg2576 as
compared with age-matched wild type
animals. There was little difference in the
extent and intensity of immunolabeling for
NCAM. Scale bar 500 μm. Cresyl violet
counterstain (purple).

Figure 3.8 HNK-1 immunoreactivity is decreased in brain
regions exhibiting Aβ plaque deposition in Tg2576 mice at
12 months of age. Free-floating serial tissue sections 50µm
thick from hippocampus, cingulate cortex or frontoparietal
cortex were incubated in anti-HNK-1 (left column) or anti-
Aβ17-24 (right column) to determine the level and
distribution of labeling in 12 month-old Tg2576 mice.
Scale bar 500 μm. Cresyl violet counterstain (purple).
76

Figure 3.9 HNK-1/CD57 decreases with increasing Aß deposition in Tg2576
mice. (a) Wild-type (wt) mice at 12 and 24 months of age show no Aβ
accumulation and there is a progressive age-dependent rise in the extent of Aβ42
load in 12 and 24 month-old Tg2576 animals in both cingulate and
frontoparietal cortex. (b) The number of HNK-1/CD57 positive cells in the
cingulate cortex and in the frontoparietal cortex is significantly decreased in 24
month-old Tg2576 mice with extensive Aβ deposition. Data were analyzed by
ANOVA with post-hoc Bonferroni tests and are presented as mean ±SEM. (c)
Individual data points are plotted to show the inverse correlation between Aβ42
accumulation and the number of HNK-1/CD57 positive cells. The line represents
the results of a regression analysis.
The extent of Aβ1-42 load and HNK-1 labeling in the hippocampus, cingulate
cortex and frontoparietal cortex was examined by image analysis as described
previously (100). Immunostaining of serial sections from brain tissue of the
77

78
aforementioned brain regions of 12 month-old Tg2576 mice reveal a reduced amount of
HNK-1 positive staining in areas with positive staining for Aβ (Figure 3.8). To
quantify these results, an ANOVA was used to compare four groups of mice (12 and 24
month-old wild-type and Tg2576 mice, with five mice per group) for the extent of Aβ1-
42 accumulation. Tissue from one of the 24-month-old wild-type animals could not be
imaged leaving a total sample size of 19. As expected, a significant main effect of
group was observed in the cingulate cortex (F
3,18
= 17.9, p < 0.0001) and in the
frontoparietal cortex (F
3,18
= 43.91 p < 0.0001). The 24 month-old Tg2576 had the
highest Aβ loads of all the groups (Figure 3.9a). Counts of HNK-1/CD57 positive cells
in each group of animals revealed similar numbers in each group for the cingulate
cortex (F
3,19
< 1, p not significant) (Figure 3.9b). However, there was a significant
difference in the number of HNK-1/CD57 positive cells as a function of group (F
3,19
=
7.1, p < 0.003) and post hoc Bonferroni tests indicate that this reflected a decrease in the
number of HNK-1/CD57 positive cells in the 24 month-old Tg2576 mice with the
highest Aβ loads (Figure 3.9b). The decrease in HNK-1/CD57 positive cells between
wild-type 24 month-old animals and Tg2576 animals of the same age approached
significance (p < 0.096). In addition, results from regression analysis suggested a
significant inverse correlation between Aβ load and HNK-1/CD57 cell counts in the
frontoparietal cortex (r = -0.72, p < 0.0001), indicating that with more extensive Aβ
there were fewer HNK-1/CD57 positive cells (Figure 3.9c). Thus, in vitro, decreased
HNK-1 expression is correlated with increasing concentrations of exogenously added
Aβ, and in vivo decreased HNK-1 expression is associated with Aβ plaque formation.

79
3.5 Discussion
A biocytin hydrazide/streptavidin purification strategy was used to extract carbonylated
proteins from Aβ-treated whole cell lysates, followed by LC-MS/MS and bioinformatic
analyses. The results of these analyses suggest that the enzyme GlcAT-P is a target of
Aβ-induced oxidative stress in cultured neurons. Our results from Western blot
analyses examining the expression of the sulfated trisaccharide carbohydrate epitope
HNK-1 support our proteomic data and demonstrate reduced expression of two forms of
HNK-1/NCAM (145 and 170 kDa) that correlate with increasing concentrations of cell-
associated Aβ. We hypothesize that the carbonylation of GlcAT-P is a result of Aβ-
induced oxidative stress because in vitro, the expression of HNK-1/NCAM was
inversely related to the concentration of Aβ added to the neuronal culture media and in
vivo, the extent of HNK-1 immunoreactivity was markedly reduced in the brain tissue
of a mouse model of AD.
It is generally accepted that the formation of free radicals and the degree to
which they cause oxidative damage play important roles during the course of aging and
various age-related neurodegenerative diseases. Protein carbonylation content is widely
used as a marker to determine levels of protein oxidation caused either by the direct
oxidation of amino acid side chains (e.g. proline and arginine to γ-
glutamylsemialdehyde, lysine to amino-adipicsemialdehyde, and threonine to amino-
ketobutyrate) or via indirect reactions with oxidative by-products (lipid peroxidation
derivatives such as 4-hydroxynonenal, malondialdehyde and advanced glycation end
products) (182; 206). Studies demonstrating that markers of protein carbonylation can

80
be detected in AD senile plaques provide evidence that Aβ plays a pivotal role in the
initiation of oxidative modifications (5; 251). Although the exact pathological
significances of such oxidative modifications have remained largely unknown, a recent
study using a combination of Oxyblot and two-dimensional gel electrophoresis has
shown that the levels of oxidative modification of β-actin and creatine kinase-BB are
higher in AD than control brains (5). In addition, the carbonylation of both β-actin and
creatine kinase BB is observed before the neurofibrillary degeneration in neurons of
AD, suggesting that age-related oxidative stress might be a primary occurrence in the
initial stages of AD pathogenesis (6).
Our initial mass spectrometry-based study suggested that GlcAT-P is a target of
Aβ-induced protein carbonylation in cultured neurons. The results from our present
study demonstrate the advantage of using a proteomic approach to monitor the
expression of a protein that cannot be assessed by conventional Western blotting
analyses, as there is no commercially available anti-GlcAT-P antibody. To confirm our
mass spectrometric data, we monitored the activity of GlcAT-P indirectly by examining
the expression of HNK-1 epitope formation. Based upon the results shown in Figure
3.4a, we concluded that 10-25µM Aβ1-42 is sufficient to induce a degree of
carbonylation of GlcAT-P that inhibits HNK-1 synthesis, as determined by the
eradication of HNK-1/NCAM expression. It should be noted that the decrease of HNK-
1/NCAM expression could also be attributed to a loss of activity of any of the enzymes
in the biosynthetic pathway of HNK-1, such as HNK-1 sulfotransferase (HNK-1 ST) or
GlcAT-S, a GlcAT variant. GlcAT-S (also referred to as GlcAT-D, UDP-

81
glucuronosyltransferase-S), which has recently been cloned in mice, has been
demonstrated to exhibit glucuronyltransferase activity towards glycoproteins and to be
involved in the biosynthesis of HNK-1 (110).
The down-regulation of HNK-1/NCAM in Aβ-treated neurons might also be a
consequence of altered lipid metabolism. Previous studies by us and other groups have
shown that Aβ toxicity is tightly associated with Aβ-mediated lipid peroxidation and
alteration of sphingolipid metabolism (224). Biochemical studies have suggested that
the activity of GlcAT-P, a type-II transmembrane protein, can be modulated in part by
phospholipids as well as by sphingosine (56). Finally, the loss of GlcAT-P activity may
also be due to other types of oxidative modification. For example, it has been shown
that a disulfide bond formation between Cys(33) and Cys(301) within human GlcAT-I
is critical for its glucuronosyltransferase activity (177), suggesting that oxidation of
cysteine residues, rather than protein carbonylation, could account for the loss of
GlcAT-P activity in Aβ-treated cultured neurons.
Our biocytin hydrazide/streptavidin affinity purification methodology coupled
with LC-MS/MS and bioinformatic analyses is a sensitive approach to identify post-
translationally modified (carbonylated) proteins whose functions could be altered due to
the effects of Aβ-mediated oxidation. These alterations in protein function are
ultimately related to the degree of oxidative protein modification. Studies examining
the relationship between age-associated levels of protein carbonylation and the
enzymatic activity of aconitase, indicate that increased aconitase carbonylation is
associated with impaired aconitase activity (57). In addition, it has been shown that

82
0.25-0.30 mol carbonyl per mol glutamine synthetase is sufficient to abolish its
enzymatic activity (144). Hence, the degree of carbonyl content resulting from protein
oxidation can be indirectly related to enzymatic activity. The degree of protein
oxidation also has an effect on the susceptibility of proteins to degradation. Mildly
oxidized proteins are selectively recognized and rapidly degraded by the proteasome,
whereas proteins that are severely oxidized can form toxic aggregates that are resistant
to proteasomal degradation (211).
Synaptic plasticity is a process of critical importance for the structural upkeep of
the brain and for the functional adaptation of the brain to the environment (158). Cell
adhesion molecules such as NCAM, telencephalin and tenascin-R, are known to play
important roles in the abilities of synapses to change the efficiency of neurotransmission
following a specific stimulus and to strengthen this synaptic transmission by a process
known as long-term potentiation (LTP) (an electrophysiological correlate of learning
and memory) (197). Cell adhesion molecules also regulate cell-to-extracellular matrix
adhesion which is critical to the processes of learning and memory (261). In mice, cell
adhesion molecules have been associated with performance in one-trial inhibitory
avoidance tasks that are used to investigate memory consolidation (230). Several of
these cell adhesion molecules implicated in synaptic efficiency carry the HNK-1
carbohydrate epitope. Saghatelyan et al. have previously shown that HNK-1 is
involved in the modulation of perisomatic inhibition and LTP in the CA1 region of the
hippocampus. They found that incubation of hippocampal slices with antibodies against
HNK-1 resulted in a disrupted balance between excitatory and inhibitory synaptic

83
transmission and a modulation in synaptic plasticity (195). In a similar in vivo study,
C57BL/6J mice that received intrahippocampal injections of an HNK-1 antibody
following one hour of training in a step-down avoidance task exhibited a median
latency of step down that was less than that exhibited in IgG-treated control mice (230).
Alternative approaches to examining the functional role of the HNK-1 carbohydrate in
vivo have focused upon the generation of mice that lack HNK-1 as a consequence of
GlcAT-P or HNK-1 sulfotransferase (ST) genetic deficiencies. Results from water
maze tests used to evaluate spatial navigation and hippocampal-dependent memory
formation in GlcAT-P -/- and HNK-1 ST -/- mice revealed impaired spatial learning
compared with that in GlcAT-P +/+ and HNK-1 ST +/+ mice (204; 263). Collectively,
these data support a role for HNK-1 in memory consolidation.
Our present study suggests that the carbonylation of GlcAT-P impairs its
enzymatic function, as determined by a reduced expression of HNK-1/NCAM in
cultured neuronal cells. We hypothesize that the carbonylation of GlcAT-P is a result
of Aβ-induced oxidative stress because the expression of HNK-1/NCAM is inversely
related to the concentration of Aβ added to the culture media. The results from our
immunohistochemical studies also support this hypothesis given that HNK-1 expression
was decreased in hippocampal and cortical brain regions of Tg2576 mice at an age
when these mice exhibit characteristic extracellular AD-type Aβ deposits. Bearing in
mind the Aβ-associated ROS
•
model of AD neurotoxicity that implicates the generation
of free radicals resulting from Aβ accumulation and aggregation as a key factor in the
development of AD (27; 202), our data can be reconciled with characterizations of

84
HNK-1 in the aforementioned in vivo studies (204; 230; 263) that describe a role for
HNK-1 in modulating synaptic plasticity and memory consolidation. The decrease in
HNK-1 expression due to Aβ-mediated oxidative stress might partially account for the
progressive memory impairment and disordered cognitive function that are behavioral
characteristics of AD.
The impairment of synaptic plasticity and LTP observed in mice that are HNK-1
deficient (263) is highly relevant to AD neuropathology. Perturbations of synaptic
plasticity and LTP reduction occur as consequences of age and aggregated Aβ, both of
which serve as critical risk factors in AD (229; 245). In order to examine the role that
HNK-1 has in LTP, memory consolidation and spatial learning within the context of
AD, electrophysiological studies using AD brain are needed to further confirm our
observations.
Future studies could use bioinformatics software to elucidate the exact amino
acid target(s) of Aβ-mediated oxidative stress for GlcAT-P. Histidine, arginine, and
lysine are known to be the most susceptible amino acids for ROS
•
-mediated carbonyl
formation (17; 130). However, when designing drugs to attenuate the pathological
effects of oxidative stress on certain key enzymes within the context of AD and other
neurodegenerative disorders, the exact amino acid targets need to be clarified. If a
relationship between HNK-1 and synaptic plasticity or LTP is established in AD brain,
there may be potential for the development of anti-oxidant agents that target GlcAT-P
and protect this molecule from Aβ-induced oxidative stress.

85
CHAPTER 4: THE ROLE OF VACUOLAR PROTEIN SORTING
PROTEIN VPS4B IN THE DEVELOPMENT OF ENDOSOMAL
PROTEIN SORTING PATHOLOGY ASSOCIATED WITH
ALZHEIMER’S DISEASE
4.1 Abstract
Sporadic late-onset Alzheimer’s disease (AD) is characterized by abnormalities of the
endosomal/lysosomal system such as endosome enlargement, increases in lysosome
number, and the accumulation of protease-resistant and ubiquitinated proteins. These
aberrations are among the earliest known neuropathological changes in AD and they
mimic the cellular phenotypes associated with the expression of an ATP hydrolysis-
deficient mutant, Vps4b(E235Q), that is involved in multivesicular body (MVB)
biogenesis and endosomal trafficking. In our previous studies, we have demonstrated
that Vps4b is carbonylated in the brains of a presenilin 1/amyloid precursor protein
(PS1/APP) transgenic mouse model of AD (225). In the current study we present a
combined biochemical, immunological, mass spectrometry and bioinformatics approach
to objectively identify and statistically validate potential Vps4b interactors in
tetracycline-inducible HEK293 cells stably expressing wild-type Vps4b or ATP
hydrolysis-deficient dominant negative Vps4b(E235Q) without the use of interactor-
specific antibodies. Furthermore, we have developed a novel probability-based method
to quantitate the relative abundances of Vps4b interactors using spectral counting from
our LC-MS/MS data. Using this method, we are able to specifically manipulate the

86
level of Vps4b expression without perturbing the endogenous level of expression of its
binding partners. Furthermore, we have implicated a potential role for Vps4b in the: 1)
altered processing of immature and mature APP and 2) production of APP C-terminal
fragments (CTFs), thus suggesting a possible link between Vps4b function and the
amyloidogenic processing of APP. Lastly, expression of Vps4b is significantly reduced
in the frontal cortex of an APP transgenic mouse model of AD (Tg2576 mice) in
comparison with age-matched wild-type C57BL mice. Taken together, these data serve
to further our goal of implicating a role for Vps4b in the development of the
endosomal/lysosomal pathologies observed in AD.

4.2 Introduction
There is increasing evidence suggesting that endosomal sorting pathways may be
uniquely vulnerable to disease pathogenesis and recent studies have begun to reveal
disease-related defects in the regulation of protein sorting (213). In Alzheimer’s disease
(AD), abnormalities of the endosomal/lysosomal system such as endosome
enlargement, increases in lysosome number, and the accumulation of protease-resistant
and ubiquitinated proteins are among the earliest detectable neuropathological changes
(3; 34; 35; 37; 170; 267). These findings have particular relevance to the development
of amyloid beta (Aβ)-related AD pathology given that: 1) the production Aβ is
intricately linked to the endocytic pathway (128), 2) abnormally enlarged neuronal early
endosomes coincide with early rises in soluble Aβ levels in the at-risk brain regions of
AD (157; 166), and 3) Aβ generation increases substantially when either endosome

87
enlargement or increased cathepsin delivery to endosomes is reproduced in cell culture
models (91; 151); thus underscoring the need to elucidate the functional mechanisms
underlying endosomal and lysosomal abnormalities in the AD brain.
The endosomal and lysosomal aberrations of the AD brain including
endosome enlargement and the accumulation of protease-resistant and ubiquitinated
proteins are similar to the phenotypes exhibited by cells expressing an ATP hydrolysis-
deficient vacuolar protein sorting (vps) mutant, Vps4b. The function of Vps4b and
other class E vps mutants was first described in yeast using an assay to determine
defects in the sorting of the soluble resident vacuolar protein, carboxypeptidase Y
(CPY) (188). CPY is normally translocated into the ER and then transported to the
Golgi where it is sorted away from secretory proteins destined for the cell surface into a
pathway that takes it via endosomes to the vacuole. However, in the class E vps yeast
strains, CPY is mis-sorted at the late Golgi into vesicles destined for the cell surface and
is secreted. It was therefore determined that class E vps mutants disrupt multivesicular
body (MVB) sorting and cause the formation of an enlarged multilamellar endosome
adjacent to the vacuole, termed the “class E compartment” (13; 69; 188; 190). To date,
27 class E Vps proteins have been identified in mammalian cells (243).

Figure 4.1. Vps4b catalyzes the dissociation of the ESCRT-III components
following the completion of endosomal cargo sorting into MVBs. Disruption of
the ATP hydrolysis function of Vps4b perturbs various routes of endosomal
cargo sorting, as indicated by “x”s in the diagram.

Vps4b is an AAA (ATPase associated with a variety of cellular activities) type
ATPase that is 444 amino acids in length. Information from its crystal structure reveals
that Vps4b has a modular structure with an N-terminal microtubule interacting and
trafficking (MIT) domain, an AAA ATPase domain of ~220 amino acids that shares a
strong sequence conservation (~30%) with other AAA-type ATPases (244), a β domain
organized into three anti-parallel β sheets, and two C-terminal α-helices (243). Vps4b
uses its ATP-hydrolyzing function to catalyze the endosomal membrane disassembly
and dissociation of endosomal sorting complex required for transport (ESCRT)-III
following the completion of endosomal cargo sorting into MVBs (Figure 4.1). ESCRT-
88

89
III is comprised of 10 class E Vps proteins that form an endosome-associated
heterooligomeric protein complex required for the sorting of proteins into the MVB
pathway (12). Evidence that Vps4b participates in membrane transport from early
endosomes to late endosomes/lysosomes includes: 1) inhibition of ubiquitinated
epidermal growth factor receptor (EGF-R) degradation due to an accumulation of the
receptors in class E compartments (72), 2) decrease of cell surface-expressed transferrin
receptor (TfR) and its concomitant abnormal association in class E compartments (271),
and 3) a mis-sorting of and defect in the proteolytic processing of newly synthesized
lysosomal hydrolase cathepsin D (72).
There is currently no consensus as to which of the ESCRT-III proteins interact
with Vps4b as different studies report specific interactions between Vps4b and varying
subsets of ESCRT-III proteins (22). Towards our goal of implicating a role for Vps4b
in the development of the endosomal/lysosomal pathologies observed in AD, we have
developed a combined biochemical, immunological, mass spectrometry and
bioinformatics approach to objectively identify and statistically validate potential Vps4b
interactors in tetracycline-inducible HEK293 cells stably expressing wild-type (wt)
Vps4b or ATP hydrolysis-deficient dominant negative Vps4b(E235Q), without the use
of interactor-specific antibodies. Furthermore, we have developed a novel probability-
based method to quantitate the relative abundances of Vps4b interactors using spectral
counting from our LC-MS/MS data. This method permits the high precision analysis of
the stoichiometry of protein complexes and it does not rely upon the use of Vps4b
interactor-specific antibodies.

90
In order to establish a link between the endosomal/lysosomal abnormalities
associated with expression of the ATP hydrolysis-deficient dominant negative
Vps4b(E235Q) mutant and Aβ-related AD pathology, the expression of mature and
immature full-length amyloid precursor protein (APP) was analyzed in tetracycline-
inducible HEK293 cells stably expressing c-Myc tagged Vps4b(wt) or Vps4b(E235Q).
We also assayed for the production of APP C-terminal fragments (CTFs) in these cells
as a means of correlating Vps4b-associated aberrant endosomal trafficking and the
potentially amyloidogenic processing of APP. In addition, we found that Vps4b
expression is decreased in the frontal cortex of an APP transgenic mouse model of AD
in comparison to age-matched wild-type mice, thus potentially further establishing a
link between the function of Vps4b and Aβ-associated AD pathology.

4.3 Methods
Antibodies
The primary antibodies used were against Vps4b (1:1,000, kindly provided by Dr. Jerry
Kaplan, University of Utah Health Sciences Center), c-Myc (1:50 for
immunofluorescence and 1µg/mL for Western blotting, Upstate), actin (1:1,000,
Chemicon), and APP C-terminus (G369 antibody, 1:1,1000, generously provided by Dr.
Sam Gandy, Thomas Jefferson University Farber Institute for Neurosciences). The
secondary antibodies used were goat anti-mouse-HRP (1:20,000, Pierce), goat anti-
rabbit-HRP (1:40,000, Pierce), and goat anti-mouse-Texas Red (1:100, Santa Cruz
Biotechnology).

91
Animals
Left frontal cortex brain tissue from 12 month-old Tg2576 mice and age-matched wild-
type (C57BL) mice was kindly provided by Dr. Elizabeth Head of the Institute for Brain
Aging and Dementia at the University of California, Irvine. Tissue was homogenized in
100mM Tris, pH 6.8, 2% SDS, and a protease inhibitor cocktail (Sigma) using a hand-
held tissue disruptor (1.5mL extraction buffer per 0.15g tissue). Homogenates were
boiled for 5min at 95ºC prior to centrifuging at 10,000 x g three times for 30min per
centrifugation. Supernatant (soluble fraction) from all centrifugations were combined.
Protein concentration was determined by the BCA method (Pierce).
Cell culture
Tetracycline-inducible Vps4b stable cell lines were generated as described previously
(53; 139) and generously provided by Dr. Phyllis Hanson (Washington University
School of Medicine). Briefly, Vps4b(wt)-Myc or Vps4b(E235Q)-Myc cDNA was
cloned into the multiple cloning site of an inducible expression vector under zeocin
selection and the resulting construct was co-transfected with a regulatory plasmid,
pcDNA6/TR expressing high levels of the TetR gene under blasticidin selection (T-
REx
TM
system, Invitrogen), into HEK293 cells to generate cell lines stably expressing
Vps4b(wt)-Myc or Vps4b(E235Q)-Myc cells. Cell lines were maintained in
Dulbecco’s Modified Eagle’s Medium (Gibco) supplemented with 10% tetracycline-
free fetal bovine serum (Hyclone), 5µg/mL blasticidin (Invitrogen), and 125µg/mL
zeocin (Invitrogen). To induce protein expression, 0.75µg/mL tetracycline (Sigma) was
added for the indicated length of time.

92
Cell viability assay
Cell viability was assessed using a two-color fluorescence cell viability assay
(Live/Dead Viability/Cytotoxicity Assay Kit, Molecular Probes). Briefly, Vps4b(wt)-
Myc and Vps4b(E235Q)-Myc stable cells induced for 4hr or 9hr were grown on two-
well culture slides and incubated with 2µM calcein AM and 4µM EthD-1 in PBS for
30min at room temperature. Following incubation, culture wells were detached from
the culture slides and coverslips were mounted using ProLong Antifade mounting
medium (Molecular Probes). Cells were examined using an Olympus Reflected
Fluorescence System and images were acquired using Spot RT Advanced imaging
software (Diagnostic Instruments, Inc.). Four fields of at least 30 cells were counted
per condition.
Immunostaining and microscopy
For imaging analysis, Vps4b(wt)-Myc and Vps4b(E235Q)-Myc stable cells were plated
onto culture slides. Cells were fixed in 4% paraformaldehyde in PBS (Electron
Microscopy Sciences). Fixed cells were permeabilized with 0.1% Triton X-100
(Sigma) in PBS for 10min followed by blocking with 2% BSA (Sigma) in PBS for 1hr.
Cells were incubated with primary antibody (anti c-Myc, Upstate) overnight at 4ºC
followed by incubation with secondary antibody (goat anti-mouse-Texas Red, Santa
Cruz Biotechnology). Coverslips were mounted using ProLong Antifade mounting
medium (Molecular Probes). Cells were examined at 60x magnification using an
Olympus Reflected Fluorescence System and images were acquired using Spot RT
Advanced imaging software (Diagnostic Instruments, Inc.).

93
Immunoprecipitation
Vps4b(wt)-Myc and Vps4b(E235Q)-Myc stable cells were grown to 70-80%
confluency on 145 x 20mm dishes prior to induction for 4hr or 9hr. Cells were lysed
by mechanical disruption with a motorized pestle in immunoprecipitation (IP) buffer
containing 20mM Tris-HCl, pH 7.4, 150mM NaCl, 10mM EDTA, 0.2% Triton X-100,
5mM ATP, 2mM MgCl
2
, and Complete
TM
protease inhibitor (Roche Applied Science).
Nuclei were pelleted by centrifugation at 800 x g for 5min at 4ºC and 500µg – 1mg
protein from the post-nuclear supernatant was used for immunoprecipitation with
immobilized anti c-Myc agarose slurry (Pierce) overnight at 4ºC. Agarose beads were
washed three times with IP buffer containing 1% Triton X-100. Proteins were eluted
with non-reducing sample buffer (Pierce) by boiling for 5min at 95ºC. For reducing
SDS-PAGE analysis, DTT was added to a final concentration of 107mM.
Immunoblot analysis
Protein in eluted IP samples was separated by SDS-PAGE on 4-20% Tris-Glycine gels
(Bio-Rad) and transferred to nitrocellulose membranes. Blots were blocked overnight
at 4ºC in PBS, 0.1% Tween-20, 5% dry non-fat milk, probed with primary antibodies
for 1hr at room temperature, and incubated with HRP-conjugated secondary antibodies
for 1hr at room temperature. Chemiluminescent detection was with Super Signal West
Pico Chemiluminescent Substrate (Pierce). Bands were quantified using densitometric
software, Scion Image (Scion Corporation), and statistical significance was assessed by
Student’s t-test.

94
Silver staining and in-gel digestion
In-gel trypsin digestion of IP eluate separated by SDS-PAGE was performed following
protein visualization via silver staining (Silver Quest Silver Staining Kit, Invitrogen)
and the division of each gel lane into eight segments (A: proteins > 150kDa, B: 150-
250kDa, C: 100-150kDa, D: 75-100kDa, E: 50-75kDa, F: 37-50kDa, G: 25-37kDa, and
H: 10-25kDa). Excised gel segments were cut into 1×1mm pieces, washed with
ultrapure water (Burdick & Jackson), dehydrated in 100% methanol (Burdick &
Jackson), re-hydrated in 30% methanol, washed twice in ultrapure water, and washed in
100mM NH
4
HCO
3
containing 30% acetonitrile. Gel pieces were dried in a vacuum
centrifuge, swelled in 10mM TCEP (Pierce) in 100mM NH
4
HCO
3
and allowed to
incubate for 1hr at 56ºC to reduce the proteins. Proteins were alkylated with 55mM
Iodoacetamide (Sigma) in 100mM NH
4
HCO
3
for 45min at room temperature in the
dark. Gel pieces were washed with 100mM NH
4
HCO
3
, shrunken with acetonitrile
(Burdick & Jackson) and dried in a vacuum centrifuge. Re-hydration with digestion
solution [(50μL H
2
0, 50μL 100mM NH
4
HCO
3
, and 1.5μg trypsin (Promega)] was
performed on ice for 45min. Any remaining supernatant was removed and 25μL
digestion buffer (digestion solution without trypsin) was added for overnight enzymatic
cleavage at 37ºC. Peptides were extracted at 37ºC for 15min with shaking - once with
50mM NH
4
HCO
3
, pH 7.8, and twice with 5% formic acid (EM): acetonitrile.
Supernatant (containing extracted peptides) was dried in a vacuum centrifuge and stored
at -20ºC until LC-MS/MS analysis.

95
LC-MS/MS data generation
MS analysis of digested IP eluate was performed using an LTQ linear ion trap mass
spectrometer (Thermo Electron) with peptides separated by a true nanoflow Xtreme-
Simple LC system (Micro-Tech Scientific, Inc.) equipped with a 150 mm x 75 μm
internal diameter C-18 reverse-phase (RP) column (1.8μm particles with 120 Å pore
size) (Micro-Tech Scientific). Peptides were loaded directly into the sample loop with
95% solvent A (2% acetonitrile, 0.1% formic acid) and 5% solvent B (95% acetonitrile,
0.1% formic acid) and were eluted with a linear gradient of 5-40% solvent B for 60 min.
The mass spectrometer was equipped with a nanospray ion source using an uncoated
10μm–internal diameter SilicaTip
TM
PicoTip
TM
nanospray emitter (New Objective).
The spray voltage of the mass spectrometer was 2.0 kV and the heated capillary
temperature was 200°C. Spectra were acquired using Xcalibur 1.4 software. The five
most abundant ions in each MS1 scan were selected for an MS/MS event. Other mass
spectrometric data generation parameters were: full scan MS mass range: 400-1800 m/z,
collision energy: 35%, isolation width: 3.0 m/z, minimum MS signal: 500 counts,
minimum MS/MS signal: 500 counts, activation time: 30 ms, and activation q: 0.25.
Mass spectrometry data analysis
Mass spectrometry .raw files were converted to an mzXML file format and MS/MS
spectra were searched against a human protein database using the Sequest Sorcerer
TM

integrated data appliance with the following parameters – database: human (indexed for
enzymatic cleavage with trypsin allowing two missed cleavages), scoring algorithm:
Sequest
TM
, peptide mass tolerance: 1.5amu, fragment mass type: monoisotopic,

96
reversed peptide database search: enabled, and differential residue modifications:
Cysteine +57.05 (carboxyamidomethylation) and Methionine +15.99 (oxidation).
Statistical validation of identified proteins was achieved using parameters ascribed by
the ProteinProphet
TM
feature of the Trans-Proteomic Pipeline (122) which reports the
following features for each identified protein: protein probability, percent sequence
coverage, number of unique peptides, and total number of peptide occurrences.
Quantitative relative protein abundance analysis
Proteins identified with ProteinProphet
TM
probability scores ≥ 0.5 and with ≥ 2 unique
peptides were selected for quantitative relative protein abundance analysis. For all
proteins in each IP sample [4hr Vps4b(E235Q)-Myc and 9hr Vps4b(E235Q)-Myc]
passing these preliminary filtering criteria, the mean values and standard deviation s of
the following protein features were calculated using S-PLUS 6.0 statistical data mining
software: percent sequence coverage, number of unique peptides, and total number of
peptide occurrences. A normalized probability score (0 – 1) for each feature was then
calculated based on the normal distribution of each feature within each sample using the
following equation:
f(x; μ, σ) = ∫
- ∞
x
1/( √2πσ) e
-((x-μ)
2
/2σ
2
)
, where μ = mean and σ = standard deviation
A final normalized probability score was subsequently assigned to each protein based
on the average normalized probability scores of each feature. This final averaged
normalized probability score is referred to hereafter as the “quantitative relative
abundance score”. Quantitative relative abundance scores close to 1 indicate relatively
highly abundant proteins - those with high percent sequence coverage, number of

unique peptides and total number of peptide occurrences - whereas low abundance
proteins have quantitative relative abundance scores close to 0 which is indicative of
low percent sequence coverage, number of unique peptides, and total number of peptide
occurrences.

4.4 Results
ATP hydrolysis-deficient Vps4b has an altered sub-cellular localization in stably
transfected HEK293 cells
To facilitate the investigation of the function of Vps4b in the mammalian endosomal
pathway, HEK293 cell lines were generated that stably express Myc-tagged wild-type
Vps4b and an ATP hydrolysis-deficient Vps4b dominant negative mutant (E235Q)
under the control of a tetracycline-regulated promoter (139). The introduction of the
E235Q point mutation within the AAA domain of Vps4b is a well-characterized
mutation known to abrogate ATP hydrolysis, thereby permitting this dominant negative
Vps4b mutant to bind tightly and “trap” its normal substrates (53; 244). Accordingly,
Vps4b(E235Q)-Myc expressing cells were used for our biochemistry and immunology-
based methods to decipher the protein interactors of Vps4b.

Figure 4.2 Time course of Vps4b(E235Q)-Myc expression in HEK293 cells. Equal amounts of post-
nuclear supernatant (15μg) were separated by SDS-PAGE and analyzed by Western blot with a c-Myc
antibody. Samples were loaded in duplicate. Molecular weight markers (in kDa) are indicated by
arrows
97

98
The addition of tetracycline to these stably-transfected HEK293 cells induced
the expression of Vps4b(E235Q)-Myc in a time-dependent manner (Figure 4.2).
Vps4b(E235Q)-Myc expression was initially detectable by immunoblotting using an
anti c-Myc antibody following three hours of induction and the expression of
Vps4b(E235Q)-Myc increased with time of induction with tetracycline. For all
subsequent experiments, induction times of 4hr and 9hr were selected based on our
hypotheses that: 1) following 4hr of induction, the level of Vps4b(E235Q)-Myc over-
expression would not perturb the binding with its endogenous binding partners, and 2)
after 9hr of induction, the over-expression Vps4b(E235Q)-Myc would result in its
binding with a different pool of protein interactors thus stimulating the abnormal
activation of pathways other than those implicated in the impairment of late endosomal
trafficking, such as cell death pathways, that have been proposed to play a central role
in neurodegeneration (47).
To determine the cellular localization of Vps4b(wt)-Myc and Vps4b(E235Q)-
Myc, indirect immunofluorescence was used with an anti c-Myc antibody. Initial
localization studies were conducted using direct immunofluorescence with HEK293
cells stably expressing GFP-tagged Vps4b(wt) and Vps4b(E235Q) (data not shown),
however, we decided to use a smaller tag (c-Myc sequence: EQKLISEEDL) so as to
avoid any confounding factors due to potentially spurious changes in the protein
interactors of Vps4b resulting from the presence of a large fluorescent protein tag (MW
of GFP = 29kDa).

99
Wild-type Vps4b has a cytosolic distribution, whereas the expression of ATP
hydrolysis-deficient Vps4b (Vps4bE235Q) results in the formation of abnormally
enlarged multilamellar endocytic vacuoles termed “class E compartments” or “EQ
compartments” (13; 19; 47; 72) in concert with its shift to a membrane localization. By
immunofluorescence, following 4hr of induction with tetracycline, Vps4b(wt)-Myc had
a diffuse cytoplasmic localization whereas Vps4b(E235Q)-Myc was localized to the
periphery of EQ compartments (Figure 4.3). In contrast, after 9hr of induction, a pool
of Vps4b(wt)-Myc was observed localized to the rim of EQ compartments, thus
suggesting that the level of Vps4b(wt)-Myc over-expression after 9hr of induction
resulted in some features of a dominant negative phenotype. The localization of
Vps4b(E235Q)-Myc after 9hr of induction mimicked its localization following 4hr of
induction, however a portion of Vps4b(E235Q)-Myc was also observed with a
cytoplasmic localization, which suggests that after 9hr of induction, the membrane
binding sites of Vps4b(E235Q)-Myc have been saturated.

Vps4b(wt)-Myc 4hr Vps4b(EQ)-Myc 4hr
Vps4b(EQ)-Myc 9hr Vps4b(wt)-Myc 9hr
Vps4b(wt)-Myc 4hr Vps4b(EQ)-Myc 4hr
Vps4b(EQ)-Myc 9hr Vps4b(wt)-Myc 9hr

Figure 4.3 Following 4hr of tetracycline-induced expression in HEK293 cells,
Vps4b(wt)-Myc has a diffuse cytoplasmic localization whereas
Vps4b(E235Q)-Myc is localized to the periphery of vacuolar “EQ”
compartments. After 9hr of tet-induced expression, both Vps4b(wt)-Myc and
Vps4b(E235Q)-Myc are localized to the periphery of EQ compartments. All
images were acquired at 60x magnification.
The viability of the stably transfected HEK293 cells expressing Vps4b(wt)-Myc
and Vps4b(E235Q)-Myc was assessed using an assay that is based on the simultaneous
determination of live and dead cells with two probes that measure intracellular esterase
activity and plasma membrane integrity as parameters of cell viability. As
demonstrated in Figure 4.4, there was no statistically significant difference in the
viability of HEK293 cells following 4hr or 9hr of tetracycline-induced expression of
Vps4b(wt)-Myc and Vps4b(E235Q)-Myc, in comparison to the viability of HEK293
cells that were not induced. Although following 9hr of induction the viability of the
Vps4b(wt)-Myc and Vps4b(E235Q)-Myc expressing cells decreased to 87% and 92%,
respectively, the differences in viability at this time point of induction were not
statistically significantly different from the viability of the non-induced cells.
100

0
10
20
30
40
50
60
70
80
90
100
110
No
induction
4hr 9hr
Viability (% of viability of non-induced
Vps4b(wt)-Myc
Vps4b(EQ)-Myc

Figure 4.4 No statistically significant difference in viability of HEK293 cells following 4hr vs. 9hr
tetracycline-induced (0.75µg/mL) expression of Vps4b(wt)-Myc and Vps4b(EQ)-Myc. Cell viability
was assessed using a fluorescence-based Calcein-AM/Ethidium homodimer-1 assay (Molecular
Probes). Values are means ±SEM of percentage viability of induced versus non-induced cells.
Statistical significance using Student’s t-test was assessed at a level of p < 0.01.
Interaction of Vps4b(E235Q)-Myc with its endogenous protein interactors is readily
deciphered using c-Myc based immunoprecipitation coupled with LC-MS/MS
Ten human proteins with homology to the yeast ESCRT-III proteins have been
identified thus far (106; 118; 242). These proteins are named either as human
homologues of specific yeast proteins (hSnf or hVps) or chromatin modifying
proteins/charged multivesicular body proteins (CHMPs): hSnf7-1/CHMP4A, hSnf7-
2/hVps32/CHMP4B, hSnf7-3/CHMP4C, hVps20/CHMP6, hVps24/CHMP3, hVps2-
1/CHMP2A, hVps2-2/CHMP2B, hVps46-1/CHMP1A, hVps46-2/CHMP1B, and
hVps60/CHMP5. Alternatively, these human ESCRT-III proteins are named according
to their human stem/progenitor cell (HSPC) gene products, such as HSPC134/CHMP4A
and HSPC177/CHMP5. Owing to its function in using is ATP-hydroylsis activity to
101

catalyze the disassembly and dissociation of ESCRT-III components following the
completion of endosomal cargo sorting, it has been suggested that ESCRT-III
components are the most likely candidates for Vps4b binding partners (22). However,
there is currently no consensus as to which of the ESCRT-III proteins specifically
interact with Vps4b.
While certain studies have analyzed the membrane-binding behavior of
Vps4b(E235Q) and how it influences Vps4b interactions with CHMP components, we
decided to undertake an objective investigation of the potential protein interactors of
Vps4b, irrespective of the distinct sub-cellular localization of these proteins, so as to
permit an unbiased analysis of potential Vps4b binding partners. Our biochemical
method relies upon the immunoaffinity isolation of c-Myc-tagged Vps4b along with its
endogenous protein interactors through the use of a c-Myc antibody (Figure 4.5). Both
Vps4b(wt)-Myc and Vps4b(E235Q)-Myc are specifically immunoprecipitated from
induced HEK293 cells (+) in comparison to the non-induced cells (-).

-+ - +
EQ wt
-+ - +
EQ wt
75
50
37
Vps4b-Myc
a) b)
IP: Myc
Western: Myc
IP: Myc
Western: Vps4b
-+ - +
EQ wt
-+ - +
EQ wt
75
50
37
Vps4b-Myc
a) b)
IP: Myc
Western: Myc
IP: Myc
Western: Vps4b

Figure 4.5 Immunoprecipitation of wild-type (wt) and dominant negative (EQ) Vps4b-Myc with c-
Myc antibody. 500µg protein from postnuclear supernatant of non-induced (-) or induced (+: 9hr
induction with 0.75µg/mL tetracycline) HEK293 cells stably expressing Vps4b(wt)-Myc or
Vps4b(E235Q)-Myc was subjected to immunoprecipitation with a c-Myc antibody, separated by SDS-
PAGE and analyzed via Western blot with a) c-Myc or b) Vps4b antibody. Molecular weight markers
(in kDa) are indicated by arrows on the left.
102

103
The next step in our objective approach to identify the protein interactors of
Vps4b entailed the enzymatic digestion of the immunoprecipitated Vps4b-Myc
interacting proteins following their visualization on silver-stained SDS-PAGE gels.
Approximately ¼ of the volume of the eluate from each c-Myc immunoprecipitated
sample was utilized for Western blot analysis (Figure 4.5) and the remainder of the
volume (3/4) was loaded onto SDS-PAGE gels for visualization with silver staining
(Figure 4.6). As indicated by the presence of numerous bands shown in the
representative silver-stained gel in Figure 4.6, Vps4b-Myc interacts with several
proteins that have a molecular weight of less than 75kDa. Vps4b-Myc migrates slightly
higher than the 50kDa marker on the representative gel.
As mentioned previously, we used the Vps4b(E235Q)-Myc samples (4hr and
9hr induction samples) in our initial studies to determine the binding partners of Vps4b
as it is thought that this dominant negative mutant binds tightly to and “traps” its normal
substrates (53; 244). The lanes containing the electrophoresed immunoprecipitated
Vps4b(E235Q)-Myc samples from the 4hr and 9hr induced cells were divided into eight
regions (indicated as A-H in Figure 4.6) prior to in-gel trypsin digestion. The resultant
peptides were then subjected to LC-MS/MS analysis using a nanoflow liquid
chromatography system and a linear ion trap mass spectrometer with a nanospray ion
source.

A
B
C
H
D
G
F
E
A
B
C
H
D
G
F
E

Figure 4.6 Silver-stained Tris-Glycine SDS-PAGE gel of eluate
from c-Myc immunoprecipitation of Vps4b(wt)-Myc and
Vps4b(E235Q)-Myc protein interactors. HEK293 cells were
induced with tetracycline (0.75µg/mL) for 4hr or 9hr to express
wild-type (wt) Vps4b-Myc or dominant negative (EQ) Vps4b-
Myc. 500µg protein from the postnuclear supernatant of the
lysates was used for immunoprecipitation with an immobilized
c-Myc antibody, followed by SDS-PAGE separation using 4-
20% gels and Silver Staining. Each lane was then divided into
eight segments (as indicated by brackets on the right) for in-gel
digestion followed by LC-MS/MS analysis. Molecular weight
markers (in kDa) are indicated by arrows.
A representative chromatogram from the analysis of gel segment E from the 9hr
Vps4b(E235Q)-Myc immunoprecipitation sample is presented in Figure 4.7. With the
use of a reverse phase column packed with 1.8μm diameter C18 particles with a pore
size of 120Å, we were able to achieve peak widths of ~30sec, thereby permitting the
efficient separation of the peptides resulting from the digestion of the proteins in each of
the analyzed gel segments. Optimized chromatographic properties such as separation
and peak width facilitate the comprehensive interrogation of all of the components
present in a given sample subjected to LC-MS/MS analysis; better chromatographic
separation positively influences mass spectrometry-based peptide identifications.
104

0 10 20 30 40 50 60 70 80 90 100 110
Time (min)
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Relative Abundance
46.17
45.02
62.98
42.07
53.42
38.51
36.95
35.83
93.53
0 10 20 30 40 50 60 70 80 90 100 110
Time (min)
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Relative Abundance
46.17
45.02
62.98
42.07
53.42
38.51
36.95
35.83
93.53

Figure 4.7 Representative chromatogram from LC-MS/MS analysis of gel segment E of the 9hr
Vps4b(E235Q)-Myc IP sample. Peak labels are retention times.
MS/MS spectra were searched against a human protein database and identified
proteins were statistically validated by ProteinProphet
TM
which is a software program
used to distinguish “true positive” from “false positive” protein identifications by
considering parameters such as percentage of sequence coverage, number of tryptic
termini of a peptide, number of unique peptides (different peptide sequences matching
the same identified protein), total number of MS/MS spectra matching the same peptide,
and protein probability score. The identified proteins from all gel segments (A-H) for
each sample [4hr and 9hr induction of Vps4b(E235Q)-Myc expression] were combined
and those proteins identified with a ProteinProphet
TM
probability ≥ 0.5 and number of
unique peptides ≥ 2 were retained. A partial list of the identified Vps4b(E235Q)-Myc
interacting proteins from the 9hr induction sample is presented in Table 4.1. The
105

proteins are grouped into the following functional categories: endosomal membrane
trafficking and multivesicular body (MVB) formation, cytoskeletal, nuclear, and protein
turnover. An MS/MS spectrum of one of the 24 Vps4b peptides (GILLFGPPGTGK)
identified in this sample is presented in Figure 4.8.

•Endosomal membrane trafficking and MVB formation
Vps4b
CHMP2a
CHMP2b
CHMP4a
CHMP4c
ALG-2 interacting protein 1
•Cytoskeletal
Tubulin
Actin
Vimentin
Neurofilament protein
•Nuclear
Emerin
Translation elongation factor
Heterogeneous nuclear ribonuclear proteins
•Protein turnover
Ubiquitin
Heat shock 70
Heat shock cognate protein 71

Table 4.1 Partial list of Vps4b(E235Q)-Myc interacting proteins identified with
high confidence in the 9hr induction sample.

106

200 300 400 500 600 700 800 900 1000 1100
m/z
0
10
20
30
40
50
60
70
80
90
100
Relative Abundance
y6
+
556.37
y7
+
613.25
b6
+
601.20
b5
+
544.20
y8
+
760.41
y9
+
873.49
b10
+
953.38
y10
+
986.54
b11
+
1010.57
y10
2+
493.97
b4
+
397.10
b3
+
284.01
y5
+
460.26
y9
2+
437.75
169
G I L L F G P P G T G K
180
y6
+
y5
+
y7
+
y9
+
y8
+
y10
+
b5
+
b4
+
b3
+
b6
+
b10
+
b11
+
200 300 400 500 600 700 800 900 1000 1100
m/z
0
10
20
30
40
50
60
70
80
90
100
Relative Abundance
y6
+
556.37
y7
+
613.25
b6
+
601.20
b5
+
544.20
y8
+
760.41
y9
+
873.49
b10
+
953.38
y10
+
986.54
b11
+
1010.57
y10
2+
493.97
b4
+
397.10
b3
+
284.01
y5
+
460.26
y9
2+
437.75
169
G I L L F G P P G T G K
180
y6
+
y5
+
y7
+
y9
+
y8
+
y10
+
b5
+
b4
+
b3
+
b6
+
b10
+
b11
+
169
G I L L F G P P G T G K
180
y6
+
y5
+
y7
+
y9
+
y8
+
y10
+
b5
+
b4
+
b3
+
b6
+
b10
+
b11
+

Figure 4.8 MS/MS spectrum of an identified Vps4b peptide (amino acid residues 169-180). Identified
ions resulting from fragmentation of the peptide with the charge retained on the peptide’s NH
2
- terminus
(b
n
+
ions) or COOH-terminus (y
n
+
and y
n
2+
ions) are indicated.
Probability-based quantitative relative protein abundance analysis facilitates the
comparison of protein abundance of Vps4b(E235Q)-Myc interactors in HEK293 cells
induced for 4hr versus 9hr
We hypothesize that following 4hr of induction, the level of Vps4b(E235Q)-
Myc over-expression does not perturb the interaction with its endogenous binding
partners, whereas after 9hr of induction, the over-expression of Vps4b(E235Q)-Myc
results in its binding with a different pool of protein interactors thus stimulating the
abnormal activation of pathways other than those implicated in the impairment of late
endosomal trafficking, such as cell death pathways, that have been proposed to play a
central role in neurodegeneration (47). In order to facilitate the qualitative comparison
of the proteins identified in the 4hr induction sample versus the 9hr induction sample
107

we analyzed the peptide occurrence distribution in these samples, along with a control
sample: post-nuclear supernatant from non-induced HEK293 cells subjected to c-Myc
immunoprecipitation coupled with LC-MS/MS as described above.

0
5
10
15
20
25
30
35
40
2 - 3 4 - 5 6 - 7 8 - 9 10 - 11 12 - 13 14 - 15 16 -17 18 - 19 > 19
Number of unqiue peptides
Number of proteins identified
4hr
9hr
No induction

Figure 4.9 Distribution of number of Vps4b(E235Q)-Myc interacting proteins identified according to
their number of unique peptides. Data are presented from the 4hr, 9hr, and no-induction samples.
Figure 4.9 is a histogram of the frequency of the number of identified proteins
with 2-3, 4-5, 6-7, 8-9, 10-11, 12-13, 14-15, 16-17, 18-19 and >19 unique constituent
peptides. It should be noted that some proteins, such as Vps4b, were identified in more
than one gel segment and are thus represented with different numbers of unique
peptides. For example, in the 4hr induction sample, Vps4b was identified in gel
segment E with 12 unique peptides and in gel segment F with 6 unique peptides. Also
in the 4hr induction sample, tubulin α-1 chain was identified in gel segment F with 6
unique peptides and in gel segment G with 2 unique peptides. The more abundant
proteins are those with high numbers of unique peptides. Among the three samples, the
108

109
9hr Vps4b(E235Q)-Myc induction sample had the highest number of proteins identified
(51 proteins) followed by the 4hr induction sample (33 proteins) and the no induction
sample (21 proteins). However, if the least abundant proteins – those with 2-3 unique
peptides – are ignored, as they could be indicative of spurious and potentially non-
specific Vps4b(E235Q)-Myc interactors, the difference in the number of identified
proteins among the three samples becomes less distinct: the number of identified
proteins with ≥ 4 peptides in the three samples is then 14 proteins in the 9hr sample, 14
proteins in the 4hr sample and 12 proteins in the no induction sample. The majority of
the proteins identified in the no induction sample were cytoskeletal proteins such as
actin, tubulin, vimentin and actinin.
In order to further decipher the specific differences between the identified
Vps4b(E235Q)-Myc interacting proteins in the 4hr and 9hr induction samples, a
comparative analysis was conducted (Figure 4.10). Although there were 51 proteins
identified in the 9hr induction sample and 33 proteins in the 4hr sample, 28 of these
proteins were found in both samples. Subtracting the 28 common proteins, it was
determined that there were 23 proteins unique to the 9hr induction sample and only 5
proteins unique to the 4hr induction sample.

23
unique
proteins
28
common
proteins
9hr
induction
4hr
induction
5
unique
proteins
23
unique
proteins
28
common
proteins
9hr
induction
4hr
induction
5
unique
proteins

Figure 4.10 Qualitative comparison of number of Vps4b(E235Q)-Myc associating
proteins identified by LC-MS/MS following 4hr and 9hr tetracycline induction of
Vps4b(E235Q)-Myc expression in HEK293 cells. Of the 56 total identified proteins, 5
were unique to the 4hr induction sample, 23 to the 9hr induction sample and 28 proteins
were identified in both samples.

While the above qualitative analysis provided us with an indication of the
number of unique and common Vps4b(E235Q)-Myc interacting proteins between the
4hr and 9hr induction samples, it was imperative to conduct a more informative
quantitative analysis that would permit us to determine the relative abundances of the
proteins common to both samples towards our goal of identifying the pathway(s) that
are perturbed specifically in the 9hr induction sample versus the 4hr induction
sample. To achieve this, we established a “quantitative relative abundance score” for
each protein with ProteinProphet
TM
probability scores ≥ 0.5 and with ≥ 2 unique
peptides.
There is a direct linear relationship between spectral counts and protein
abundance; however, this relationship is skewed by protein size and tryptic peptides
with missed cleavages (200). Statistical modeling to establish an “adjusted spectral
110

111
count” has been used as one approach to infer protein abundance from spectral counts
(200). Our quantitative relative abundance score avoids the bias associated with
protein size and spectral counting by normalizing the following protein features when
assessing relative protein abundance: percent sequence coverage, number of unique
peptides, and total number of peptide occurrences.
The quantitative relative abundance score takes into account the following
three features of each protein identified in each IP sample [4hr Vps4b(E235Q)-Myc
and 9hr Vps4b(E235Q)-Myc]: percent sequence coverage, number of unique
peptides, and total number of peptide occurrences. First, the mean values and
standard deviations of these features were calculated and a normalized probability
score (0 – 1) for each feature was calculated based on the normal distribution of each
feature within each sample. A final normalized probability score was subsequently
assigned to each protein based on the average normalized probability scores of each
feature. This final averaged normalized probability score is what we have termed the
“quantitative relative abundance score”. Quantitative relative abundance scores near
1 indicate relatively highly abundant proteins - those with high percent sequence
coverage, number of unique peptides and total number of peptide occurrences -
whereas low abundance proteins have quantitative relative abundance scores close to
0 which is indicative of low percent sequence coverage, number of unique peptides,
and total number of peptide occurrences. This method of inferring protein abundance
from spectral counts permits the high precision analysis of the stoichiometry of
protein complexes and it does not rely on the use of interactor-specific antibodies.

0
0.2
0.4
0.6
0.8
1
1.2
Vps4b CHMP4A CHMP2B CHMP2A CHMP4C CHMP5
Quantitative relative abundance score
4hr
9hr
Figure 4.11 Quantitative relative abundance of Vps4b and ESCRT-III proteins in the 4hr and 9hr
induction c-Myc immunoprecipitation samples isolated from post-nuclear supernatant of
Vps4b(E235Q)-Myc expressing HEK293 cells. Statistical validation of protein identifications was
achieved with ProteinProphet
TM
and a quantitative relative abundance score was calculated for each
protein based on the average normalized values for the following protein features: percent sequence
coverage, number of unique peptides, and total number of occurrences.
A plot of the quantitative relative abundance scores for Vps4b and its
identified ESCRT-III component interactors - CHMPs 2A, 2B, 4A, 4C and 5 -
permits the comparative analysis of the relative abundances of these proteins (Figure
4.11). The quantitative relative abundance score of Vps4b from the 9hr induction
sample is 0.97 whereas its score in the 4hr sample is 0.76, indicating that Vps4b is
more abundant in the 9hr sample than in the 4hr sample. This supports the results
from our time-course of induction study (Figure 4.2) in that longer incubation times
with tetracycline correspond with increased induction of Vps4b(E235Q)-Myc
expression. Of particular note is the trend in the quantitative relative abundance
scores of the identified CHMPs in the 4hr versus 9hr samples. Although the scores
112

113
for the CHMPs identified in the 9hr induction sample are higher than those in the 4hr
induction sample, the difference in the scores of the CHMPs identified in both
samples (CHMPs 2A, 2B, 4A and 4C) is not as apparent as the difference in the
scores of Vps4b, thus suggesting that with this tetracycline-inducible expression
system, we are able to specifically modulate the levels of Vps4b expression without
perturbing the levels of expression of its endogenous ESCRT-III binding partners.
Of interest among the identified Vps4b(E235Q)-Myc interacting ESCRT-III
proteins is CHMP5, which was identified in the 9hr induction sample (quantitative
relative abundance score = 0.56) but not in the 4hr sample. This could potentially be
explained by the localization of Vps4b(E235Q)-Myc following 9hr of induction. It is
possible that the association of CHMP5 with Vps4b(E235Q)-Myc could be intricately
linked to the specific localization (cytosol or membrane) of Vps4b(E235Q)-Myc.
Our quantitative relative abundance score can be used not only for inter-
sample comparative analysis, e.g. comparison of the relative abundances of the same
protein identified in more than one sample, but it can also be used for intra-sample
comparative analysis whereby the relative abundances of different proteins within the
same sample can be compared. The features that contribute to the final computed
quantitative relative abundance score (percent sequence coverage, number of unique
peptides, and total number of peptide occurrences) are normalized, thereby avoiding
potential bias and confounding factors when comparing different proteins that vary
significantly in molecular weight.

114
From this intra-sample comparison, we determined that the most abundant
ESCRT-III component that co-immunoprecipitated with Vps4b(E235Q)-Myc in both
the 4hr and 9hr induction samples was CHMP4A (quantitative relative abundance
score in 4hr sample = 0.50, and in 9hr sample = 0.54) (Figure 4.11). The least
abundant Vps4b(E235Q)-Myc co-immunoprecipitating ESCRT-III component in the
4hr sample was CHMP4C (score = 0.36), whereas the least abundant in the 9hr
induction sample was CHMP2A (score = 0.42). Differences in the relative
abundances of ESCRT-III components that associate with Vps4b could have
implications for the membrane-binding behavior of this protein interaction module as
ESCRT-III contains two functionally distinct subcomplexes, one which is required
for its Vps4b-dependent dissociation, and the other which is responsible for its
membrane association (12).
Following 9hr induction of expression, Vps4b(E235Q)-Myc interacts with more
proteins involved in cell death, protein turnover and protein folding than after 4hr
induction of expression
Consistent with our hypothesis that following 4hr of induction the level of
Vps4b(E235Q)-Myc over-expression does not perturb the interaction with its
endogenous binding partners, whereas after 9hr of induction, the over-expression of
Vps4b(E235Q)-Myc results in its binding with a different pool of protein interactors
thus stimulating the abnormal activation of pathways other than those implicated in the
impairment of late endosomal trafficking that have been proposed to play a central role
in neurodegeneration, there were more proteins involved in cell death, protein turnover

and protein folding identified in the 9hr induction sample versus the 4hr induction
sample (Figure 4.12). These proteins included the 26S proteasome regulatory subunit
S1, ALG-2 (apoptosis-linked gene - 2) interacting protein 1 (AIP1), heat shock 70kDa
protein (Hsp70), heat shock cognate 71kDa protein (Hsc70), and ubiquitin. Figure 4.13
is an MS/MS spectrum of an identified ubiquitin peptide (TITLEVEPSDTIENVK) from
the 9hr induction sample. The abnormal accumulation of ubiquitinated cargo that is
trafficked through the MVB sorting pathway en route to the lysosome for degradation
has been observed in class E compartments formed in cells expressing Vps4b(E235Q),
hence implicating a role for Vps4b in the endocytic trafficking of ubiquitinated cargo
(102; 139).

0
0.1
0.2
0.3
0.4
0.5
0.6
26S proteasome
regulatory
subunit
ubiquitin Heat shock
70kDa protein
(Hsp70)
Heat shock
cognate 71kDa
protein (Hsc70)
ALG-2 interacting
protein 1 (AIP1)
Q uantitative relative abun dan ce score
4hr
9hr

Figure 4.12 More Vps4b(E235Q)-Myc interacting proteins involved in cell death, protein turnover and protein
folding were identified in the 9hr induction sample than in the 4hr sample, as determined by quantitative
relative abundance analysis.

115

AIP1, ubiquitin, Hsp70, and Hsc71 were identified with quantitative relative
abundance scores of 0.31, 0.41, 0.28, and 0.51, respectively (Figure 4.12), whereas
these proteins were not identified in the 4hr induction sample. Contrarily, the 26S
proteasome regulatory subunit S1 was identified only in the 4hr induction sample
(quantitative relative abundance score = 0.28). These data suggest that following 9hr of
induction of Vps4b(E235Q)-Myc expression, the cell death, protein turnover, and
protein folding pathways are up-regulated to a greater extent than following 4hr
induction of Vps4b(E235Q)-Myc expression.

12
T I T L E V E P S D T I E N V K
27
y6
+
y4
+ y9
+
y11
+ y10
+
y12
+
b5
+ b13
+
b7
+
b6
+ b10
+
b12
+
b15
+
y3
+
12
T I T L E V E P S D T I E N V K
27
y6
+
y4
+ y9
+
y11
+ y10
+
y12
+
b5
+ b13
+
b7
+
b6
+ b10
+
b12
+
b15
+
y3
+
300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
m/z
0
10
20
30
40
50
60
70
80
90
100
R e la tiv e A bu nda nc e
y9+
1002.48
y10+
1131.50
y11+
1230.51 b7+
787.23
b15+
1641.69
b13+
1428.63
y12+
1359.55
b5+
558.26
y4+
489.38
b12+
1300.49
y3+
360.18
y6+
703.48
b10+
1085.52
b6+
657.38
300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
m/z
0
10
20
30
40
50
60
70
80
90
100
R e la tiv e A bu nda nc e
y9+
1002.48
y10+
1131.50
y11+
1230.51 b7+
787.23
b15+
1641.69
b13+
1428.63
y12+
1359.55
b5+
558.26
y4+
489.38
b12+
1300.49
y3+
360.18
y6+
703.48
b10+
1085.52
b6+
657.38

Figure 4.13 MS/MS spectrum of an identified ubiquitin peptide (amino acid residues 12-27) from the 9hr
induction sample.
Expression of full-length APP and APP C-terminal fragments is altered by
Vps4b(E235Q) expression
In addition to establishing a method to statistically validate the identified protein
interactors of Vps4b(E235Q)-Myc along with determining their quantitative relative
116

117
abundance, we wanted to test our hypothesis that the expression of the ATP hydrolysis-
deficient Vps4b dominant negative mutant perturbs the endosomal trafficking of the
amyloid precursor protein (APP), consequently affecting its processing. Newly
synthesized APP molecules traverse the Golgi where they are N-glycoslylated
(immature APP) or N+O glycosylated (mature APP) en route to the plasma membrane,
where some are cleaved to liberate a soluble APP extracellular fragment (sAPPα) while
resulting in the concomitant formation of an 83-residue APP carboxyl-terminal
fragment (CTF) C83. The expression of both immature and mature APP was reduced in
HEK293 cells when Vps4b(E235Q)-Myc expression was induced for 4hr and 9hr in
comparison with HEK293 cells not induced to express Vps4b(E235Q)-Myc (Figure
4.14). This effect was most pronounced in the expression of mature APP in the cells
induced for 9hr to express Vps4b(E235Q)-Myc. After 9hr of induction in these cells,
there was no detectable expression of mature APP. However, this trend was reversed in
HEK cells stably expressing Vps4b(wt)-Myc. In these cells, with increasing time of
induction, there was a concomitant increase in the expression of both immature and
mature APP. These results suggest that the dominant negative Vps4b mutant possibly
has a role in retarding the trafficking of newly synthesized APP molecules in sorting
endosomal compartments after APP is trafficked from the Golgi and before it reaches
the plasma membrane. It should be noted that there is a difference in the expression of
mature and immature APP in the non-induced Vps4b(wt)-Myc HEK293 cells and in the
non-induced Vps4b(E235Q)-Myc HEK293 cells. If the expression of mature APP was

normalized to that of immature APP, the difference in APP expression between the
non-induced Vps4b(wt)-Myc and Vps4b(E235Q)-Myc cells might not be as
pronounced.

100
150
0 4 9 0 4 9
Vps4b(wt) Vps4b(E235Q)
Time of induction (in hrs):
Actin
Mature APP: N+O glycosylation
Immature APP: N-glycosylation
37
50
100
150
0 4 9 0 4 9
Vps4b(wt) Vps4b(E235Q)
Time of induction (in hrs):
Actin
Mature APP: N+O glycosylation
Immature APP: N-glycosylation
37
50

Figure 4.14 Expression of immature N-glycosylated and mature N+O glycosylated APP in Vps4b(E235Q)-
Myc HEK293 cells decreases with increasing time of induction, whereas the induction of Vps4b(wt)-Myc
positively correlates with increased expression of immature N-glycosylated and mature N+O glycosylated
APP. Protein (50µg) from post-nuclear supernatant was immunoblotted with an antibody, G369, raised
against the C-terminus of APP. As a loading control, the immunoblot was re-probed with actin. Molecular
weight markers (in kDa) are indicated by arrows.
Non-processed APP is endocytosed from the cell surface into late endosomal
compartments for processing into 89- or 99-residue CTFs (C89 and C99, respectively)
that are generated from cleavage by β-secretase. These CTFs are retained in the
membrane and can become substrates for intra-membrane cleavage mediated by γ-
secretase, yielding the amyloidogenic 40- or 42-residue peptide Aβ and the APP
intracellular domain (AICD) (Figure1.1). Endosomal vesicles such as MVBs are an
ideal location for the amyloidogenic processing of APP, since the lipid membranes,
constricted space and low pH within these vesicles favor Aβ aggregation. Therefore,
the trafficking of APP and its potential sequestration in these endosomal compartments,
events in which Vps4b could likely have a role, influence its processing. We observed
a marked decrease in the production of an APP CTF in the Vps4b(E235Q)-Myc
expressing cells (Figure 4.15). The production of this APP CTF was also diminished in
118

cells expressing Vps4b(wt)-Myc, however, the decrease in CTF production following
4hr and 9hr of induction in these cells was not as pronounced as the decrease in CTF
production in the Vps4b(E235Q)-Myc expressing cells induced for the same period of
time. Immunoblotting with CTF-specific antibodies will serve to clarify the nature of
the observed CTF in an effort to establish which secretases are responsible for these
APP cleavage events.

20
10
15
0 4 9 0 4 9
Vps4b(wt) Vps4b(E235Q)
Time of induction (in hrs):
37
50
Actin
APP C-terminal fragment
20
10
15
20
10
15
0 4 9 0 4 9
Vps4b(wt) Vps4b(E235Q)
Time of induction (in hrs):
37
50
Actin
APP C-terminal fragment

Figure 4.15 Production of an APP C-terminal fragment is reduced in HEK293 cells when Vps4b(E235Q)-Myc
expression is induced for 4hrs and 9hrs in comparison to both non-induced cells (0hr) and the induction of
Vps4b(wt)-Myc. Protein (100µg) from post-nuclear supernatant was immunoblotted with an antibody, G369,
raised against the C-terminus of APP. As a loading control, the immunoblot was re-probed with actin.
Molecular weight markers (in kDa) are indicated by arrows.
Vps4b expression is significantly reduced in frontal cortex tissue of 12 month-old
Tg2576 mouse model of Alzheimer’s disease in comparison to age-matched control
C57BL mice
We next sought to determine the expression of Vps4b in a transgenic mouse model of
Alzheimer’s disease. We selected Tg2576 mice for this study as these mice harbor a
mutation in the amyloid precursor protein (K670N, M671L) and display markedly
elevated Aβ levels at an early age. By 9-12 months of age, these mice develop
characteristic extracellular AD-type Aβ deposits in the cortex and hippocampus (104).
We observed a statistically-significant decrease in the expression of Vps4b in the frontal
cortex tissue of 12 month-old Tg2576 mice (Figure 4.16, lanes 1 – 3) in comparison to
119

the level of Vps4b expression in age-matched wild-type C57BL mice (lanes 4 – 6).
These results implicate a possible role for Vps4b in the pathogenesis of AD; however,
in order to further confirm these findings, brain tissue from younger mice with various
stages of Aβ deposition will need to be examined.

*
0.00
20.00
40.00
60.00
80.00
100.00
120.00
C57BL (wt) Tg2576
Vps4b expression
(% of wt)
*
1 2 3 4 5 6
1 – 3: Tg2576 mice
4 – 6: C57BL (wt) mice
Actin
Vps4b
a)
b)
*
0.00
20.00
40.00
60.00
80.00
100.00
120.00
C57BL (wt) Tg2576
Vps4b expression
(% of wt)
*
*
0.00
20.00
40.00
60.00
80.00
100.00
120.00
C57BL (wt) Tg2576
Vps4b expression
(% of wt)
*
1 2 3 4 5 6
1 – 3: Tg2576 mice
4 – 6: C57BL (wt) mice
Actin
Vps4b
a)
b)

Figure 4.16 Expression of Vps4b is reduced in frontal cortex tissue of 12 month-old Tg2576 mice,
compared to age-matched wild-type (C57BL) mice. (a) Frontal cortex tissue of three Tg2576 and C57BL
mice was homogenized and 40µg protein was separated by SDS-PAGE. Vps4b expression was analyzed
by Western blot analysis using a rabbit polyclonal Vps4b antibody (generously provided by the
laboratory of Dr. Jerry Kaplan, University of Utah). As a control for protein loading, the immunoblot
was re-probed with a monoclonal actin antibody. (b) Quantitation of Vps4b immunoreactive bands using
Scion Image software. Values are means ±SEM. *p < 0.01 versus control (Student’s t-test).

4.5 Discussion
Towards our goal of implicating a role for Vps4b in the development of the
endosomal/lysosomal pathologies observed in AD, we have developed a combined
biochemical, immunological, mass spectrometry and bioinformatics approach to
objectively identify and statistically validate potential Vps4b interactors in
tetracycline-inducible HEK293 cells stably expressing c-Myc tagged wild-type (wt)
Vps4b or ATP hydrolysis-deficient dominant negative Vps4b(E235Q), without the
use of interactor-specific antibodies. As with any protein interaction study, the
120

121
protein interactors of Vps4b will vary in accordance with the experimental method
used. However, as a means of interpreting the biological significance of identified
Vps4b interactors, we have developed a novel probability-based method to quantitate
their relative abundances that does not rely upon the use of Vps4b interactor-specific
antibodies.
Following Sequest Sorcerer
TM
based protein identification from LC-MS/MS
data, protein identifications were statistically validated by the software program,
ProteinProphet
TM
. Each protein was then assigned a normalized quantitative relative
abundance score which takes into account the following protein features: percent
sequence coverage, number of unique peptides, and total number of peptides. Thus,
the normalized probability score yields information pertaining to protein abundance.
There is a direct linear relationship between spectral counts and protein abundance;
however, this relationship is skewed by protein size and tryptic peptides with missed
cleavages (200). Statistical modeling to establish an “adjusted spectral count” has
been used as one approach to infer protein abundance from spectral counts (200).
Our quantitative relative abundance score avoids the bias associated with protein size
and spectral counting by normalizing the following protein features when assessing
relative protein abundance: percent sequence coverage, number of unique peptides,
and total number of peptide occurrences.
As the relative quantitative abundance score is normalized, there is a lack of
concern about confounding factors such as: 1) low molecular weight proteins with only
one or two unique peptides identified equating to a high percentage of sequence

122
coverage, 2) proteins with a high total number of peptide occurrences but with a low
number of unique peptides, and 3) high molecular weight proteins with low percent
sequence coverage that could skew measures of relative protein abundance. This
quantitative relative abundance score can be used to determine statistically significant
differences in protein abundance both within the same sample (intra-sample
comparison) and between samples (inter-sample comparison). However, in order to
conduct such an analysis, replicates of each experimental condition need to be
performed.
Among the ESCRT-III components that were identified as Vps4b(E235Q)-Myc
interacting proteins from both the 4hr and 9hr induction samples were CHMPs 2A, 2B,
4A and 4C. Other groups have used two-hybrid screening, immunocytochemistry and
biochemical methods to demonstrate that CHMPs 2A and 4A are among the
components of ESCRT-III that interact with Vps4b (71; 139). Similar to our approach,
these Vps4b-ESCRT-III component protein interaction studies were conducted using an
ATP hydrolysis-deficient Vps4b mutant, Vps4b(E235Q), as this ATP hydrolysis-
deficient mutant is thought to “trap” and bind tightly to its normal substrates.
By ascribing a relative quantitative abundance score to the identified
Vps4b(E235Q)-Myc interacting proteins from the 4hr and 9hr induction samples, in our
studies, we were able to determine that under the experimental conditions used, the
ESCRT-III component CHMP4A was the most abundant Vps4b interactor in both
samples. Furthermore, we ascertained that proteins involved in protein folding and
turnover were up-regulated in the 9hr induction sample as compared with the 4hr

123
induction sample. This suggests that the dysfunction of Vps4b is associated with
pathways that are implicated in disease pathogenesis. For example, ALG-2 (apoptosis-
linked gene - 2) interacting protein 1 (AIP1) was identified as a Vps4b(E235Q)-Myc
interacting protein with a quantitative relative abundance score of 0.31, but it was not
identified in the 4hr sample. AIP1 is capable of regulating caspase-dependent and
independent cell death (118). The detection of AIP1 in the 9hr versus 4hr sample
further suggests that the inability of Vps4b to properly regulate MVB biogenesis and
endosomal sorting (as determined by the over-expression of the ATP hydrolysis-
deficient Vps4b mutant) coincides with cellular decisions to commit to cell death
pathways that could be implicated in disease pathogenesis.
It is known that the endosomal membrane-binding behavior of Vps4b is coupled
with its ATP hydrolysis cycle; therefore it would be of interest to determine the Vps4b
protein interactors that are sensitive to its cytoplasmic versus membrane association in
an attempt to determine which proteins are involved in the regulation of Vps4b-
dependent endosomal sorting. Vps4b ATPase activity significantly influences the
membrane association of lyst, a protein known to regulate lysosomal fission (71). The
endosomal association of ESCRT-III is also sensitive to the ATPase activity of Vps4b.
In cells expressing wild-type Vps4b, ≤30% of the ESCRT-III subunits are endosome
associated due to the ESCRT-III disassembly activity of Vps4b (11). Although the
post-nuclear supernatant of induced HEK293 cells was used as the starting material for
the immunoprecipitations, there were some identified nuclear proteins, including
emerin, and heterogeneous nuclear ribonuclear proteins (Table 4.1). It has been

124
proposed that some mammalian class E Vps homologues might have dual roles in the
regulation of gene expression in the nucleus and in the regulation of MVB sorting in the
cytoplasm (118).
To confirm the Vps4b protein interactors identified in our study, we will
conduct immunoprecipitations with interactor-specific antibodies. As an indication of
the biological significance of identified Vps4b protein interactions, functional assays
need to be performed whereby the expression of the identified interactors are
systematically knocked-down by siRNA techniques and end-points such as cell surface
receptor (EGF-R or LDL-R) endosomal accumulation or cell proliferation are assayed.
The aberrant accumulation of ubiquitinated cargo that is normally trafficked
through the MVB sorting pathway en route to the lysosome for degradation has been
observed in class E compartments formed upon the expression of Vps4b(E235Q) (102;
139) hence implicating a role for Vps4b in the endocytic trafficking of ubiquitinated
cargo. This phenomenon has potential pathological consequences related to the
trafficking of synaptic cell surface receptor proteins, such as the NMDA receptor
subunits, whose internalization, ubiquitination and subsequent degradation, are essential
for its function in rapid synaptic transmission, which is compromised in the AD brain.
Among the responses to synaptic activity in the postsynaptic density (PSD – a
specialized structure containing glutamate receptors and associated scaffolding proteins
that organize signal transduction pathways at the postsynaptic membrane), is dynamic
turnover mediated by ubiquitination and degradation (62), thereby providing evidence
that the functional and molecular organization of the PSD is regulated by synaptic

125
activity-dependent protein turnover through the ubiquitin-proteasome system (UPS)
(117). Specifically, increased synaptic activity is positively correlated with ubiquitin
conjugation and NMDA receptor degradation, while deprivation of synaptic activity
reduces NMDA receptor degradation (62; 109). Given the role of Vps4b in trafficking
ubiquitinated cargo through the MVB pathway, it is possible that Vps4b is involved in
the turnover of NMDA receptor subunits. Under conditions of impaired Vps4b function
(Vps4bE235Q) or decreased Vps4b expression, internalized ubiquitinated NMDA
receptor subunits would likely abnormally accumulate in late endosomal compartments
thereby precluding their degradation and consequently resulting in an adverse effect on
synaptic signal transduction.
The UPS might also be impaired by Aβ accumulation in MVBs, which could
provide another mechanism accounting for the defects in synaptic plasticity as are
characteristic of AD. MVBs are known to reside close to synapses (45). MVBs are
also viewed as major transport organelles within neuronal processes (254), and Aβ
accumulation in their outer membranes may disrupt vital cargo transport.
Sorting mechanisms that cause APP and the secretases to co-localize in the same
endosomal compartment would be expected to play critical roles in the regulation of
Aβ production. Endosomal vesicles such as MVBs are an ideal location for Aβ
aggregation, since the lipid membranes, constricted space and low pH within these
vesicles favor Aβ aggregation. We have found that the expression of ATP hydrolysis-
deficient Vps4b is associated with a decrease in the expression of full-length mature and
immature APP and also with a decrease in the production of APP CTF. We are

126
uncertain of the nature of the CTF detected in our immunoblotting method; however,
studies using CTF-specific antibodies, in addition to experiments involving inhibitors of
β- or γ-secretase, should prove to be more conclusive.
Although the exact identity of the APP CTF detected in these studies is not
known, our results implicate a potential role for Vps4b in the trafficking and processing
of APP. Other Vps proteins such as Vps10 (8; 173; 193; 228) and Vps35 (class A vps
mutants) (214) have been implicated in regulating APP trafficking in endocytic
compartments. Identifying, statistically validating, and quantifying the relative
abundances of the potential protein interactors of these Vps proteins using a method
similar to that described in the present study should aid in the clarification of the role
that these proteins have in the endocytic trafficking of APP that influences its
potentially amyloidogenic processing.
One caveat to our current methodology used to identify the protein interactors of
Vps4b is that the efficiency of peptide extraction from SDS-PAGE gels following the
in-gel digestion of the c-Myc immunoaffinity purified proteins is negatively influenced
by the polyacrylamide gel matrix, thus potentially impairing our quantitative relative
comparison analyses. A possible alternative is to subject the immunoprecipitated
proteins to in-solution digestion, as opposed to in-gel digestion; however, the protein
sample would not be fractionated based on molecular weight, as it is by SDS-PAGE
methods, and consequently, sample complexity would be a limiting factor in our LC-
MS/MS analysis. Gel filtration chromatography is another method of fractionating
samples based on molecular weight that we could employ which would allow us to

127
subject the proteins to in-solution digestion, thereby creating conditions for optimal
peptide recovery and more accurate measures of quantitative relative protein abundance.
Gel filtration chromatography would also enable us to analyze the protein interactors of
oligomeric Vps4b. ATP binding by Vps4b results in its homooligomerization into a
complex of ~440kDa and ATP hydrolysis by the Vps4b oligomer then returns Vps4b to
its inactive dimeric state (11).
In the current study, we present a combined biochemical, immunological,
mass spectrometry and bioinformatics approach to objectively identify and
statistically validate potential Vps4b interactors. Additionally, we have developed a
probability-based method to quantitate the relative abundances of these protein
interactors. Functional cell biology-based assays used to determine the pathological
significance of these Vps4b protein interactors in neuronal cells will clarify the
potential role of Vps4b in the development of the endosomal/lysosomal pathologies
observed in AD.

128
BIBLIOGRAPHY
1. Abel, T., and Kandel, E. (1998). Positive and negative regulatory mechanisms
that mediate long-term memory storage. Brain Res Brain Res Rev 26, 360-378.
2. Abo, T., and Balch, C. M. (1981). A differentiation antigen of human NK and K
cells identified by a monoclonal antibody (HNK-1). J Immunol 127, 1024-1029.
3. Adamec, E., Mohan, P. S., Cataldo, A. M., Vonsattel, J. P., and Nixon, R. A.
(2000). Up-regulation of the lysosomal system in experimental models of
neuronal injury: implications for Alzheimer's disease. Neuroscience 100, 663-
675.
4. Akama, K. T., and Van Eldik, L. J. (2000). Beta-amyloid stimulation of
inducible nitric-oxide synthase in astrocytes is interleukin-1beta- and tumor
necrosis factor-alpha (TNFalpha)-dependent, and involves a TNFalpha receptor-
associated factor- and NFkappaB-inducing kinase-dependent signaling
mechanism. J Biol Chem 275, 7918-7924.
5. Aksenov, M. Y., Aksenova, M. V., Butterfield, D. A., Geddes, J. W., and
Markesbery, W. R. (2001). Protein oxidation in the brain in Alzheimer's disease.
Neuroscience 103, 373-383.
6. Aksenova, M. V., Aksenov, M. Y., Payne, R. M., Trojanowski, J. Q., Schmidt,
M. L., Carney, J. M., Butterfield, D. A., and Markesbery, W. R. (1999).
Oxidation of cytosolic proteins and expression of creatine kinase BB in frontal
lobe in different neurodegenerative disorders. Dement Geriatr Cogn Disord 10,
158-165.
7. Alzheimer, A. (1907). Über eine eigenartige Erkrankung der Hirnrinde ["An
unusual disease of the cerebral cortex"]. Allgemeine Zeitschrift fur Psychiatrie
64, 146-148.

8. Andersen, O. M., Reiche, J., Schmidt, V., Gotthardt, M., Spoelgen, R., Behlke,
J., von Arnim, C. A., Breiderhoff, T., Jansen, P., Wu, X., et al. (2005). Neuronal
sorting protein-related receptor sorLA/LR11 regulates processing of the amyloid
precursor protein. Proc Natl Acad Sci U S A 102, 13461-13466.

129
9. Arendash, G. W., King, D. L., Gordon, M. N., Morgan, D., Hatcher, J. M.,
Hope, C. E., and Diamond, D. M. (2001). Progressive, age-related behavioral
impairments in transgenic mice carrying both mutant amyloid precursor protein
and presenilin-1 transgenes. Brain Res 891, 42-53.
10. Atwood, C. S., Moir, R. D., Huang, X., Scarpa, R. C., Bacarra, N. M., Romano,
D. M., Hartshorn, M. A., Tanzi, R. E., and Bush, A. I. (1998). Dramatic
aggregation of Alzheimer abeta by Cu(II) is induced by conditions representing
physiological acidosis. J Biol Chem 273, 12817-12826.
11. Azmi, I., Davies, B., Dimaano, C., Payne, J., Eckert, D., Babst, M., and
Katzmann, D. J. (2006). Recycling of ESCRTs by the AAA-ATPase Vps4 is
regulated by a conserved VSL region in Vta1. J Cell Biol 172, 705-717.
12. Babst, M., Katzmann, D. J., Estepa-Sabal, E. J., Meerloo, T., and Emr, S. D.
(2002). Escrt-III: an endosome-associated heterooligomeric protein complex
required for mvb sorting. Dev Cell 3, 271-282.
13. Babst, M., Sato, T. K., Banta, L. M., and Emr, S. D. (1997). Endosomal
transport function in yeast requires a novel AAA-type ATPase, Vps4p. Embo J
16, 1820-1831.
14. Ballif, B. A., Villen, J., Beausoleil, S. A., Schwartz, D., and Gygi, S. P. (2004).
Phosphoproteomic analysis of the developing mouse brain. Mol Cell Proteomics
3, 1093-1101. Epub 2004 Sep 1002.
15. Bancher, C., Grundke-Iqbal, I., Iqbal, K., Fried, V. A., Smith, H. T., and
Wisniewski, H. M. (1991). Abnormal phosphorylation of tau precedes
ubiquitination in neurofibrillary pathology of Alzheimer disease. Brain Res 539,
11-18.

16. Benzing, W. C., Wujek, J. R., Ward, E. K., Shaffer, D., Ashe, K. H., Younkin,
S. G., and Brunden, K. R. (1999). Evidence for glial-mediated inflammation in
aged APP(SW) transgenic mice. Neurobiol Aging 20, 581-589.
17. Berlett, B. S., and Stadtman, E. R. (1997). Protein oxidation in aging, disease,
and oxidative stress. J Biol Chem 272, 20313-20316.

130
18. Beyreuther, K., Pollwein, P., Multhaup, G., Monning, U., Konig, G., Dyrks, T.,
Schubert, W., and Masters, C. L. (1993). Regulation and expression of the
Alzheimer's beta/A4 amyloid protein precursor in health, disease, and Down's
syndrome. Ann N Y Acad Sci 695, 91-102.
19. Bishop, N., and Woodman, P. (2000). ATPase-defective mammalian VPS4
localizes to aberrant endosomes and impairs cholesterol trafficking. Mol Biol
Cell 11, 227-239.
20. Blackburn, A. C., Doe, W. F., and Buffinton, G. D. (1999). Protein carbonyl
formation on mucosal proteins in vitro and in dextran sulfate-induced colitis.
Free Radic Biol Med 27, 262-270.
21. Borchelt, D. R., Ratovitski, T., van Lare, J., Lee, M. K., Gonzales, V., Jenkins,
N. A., Copeland, N. G., Price, D. L., and Sisodia, S. S. (1997). Accelerated
amyloid deposition in the brains of transgenic mice coexpressing mutant
presenilin 1 and amyloid precursor proteins. Neuron 19, 939-945.
22. Bowers, K., Lottridge, J., Helliwell, S. B., Goldthwaite, L. M., Luzio, J. P., and
Stevens, T. H. (2004). Protein-protein interactions of ESCRT complexes in the
yeast Saccharomyces cerevisiae. Traffic 5, 194-210.
23. Braak, H., Braak, E., and Strothjohann, M. (1994). Abnormally phosphorylated
tau protein related to the formation of neurofibrillary tangles and neuropil
threads in the cerebral cortex of sheep and goat. Neurosci Lett 171, 1-4.

24. Braun, A. P., and Schulman, H. (1995). The multifunctional
calcium/calmodulin-dependent protein kinase: from form to function. Annu Rev
Physiol 57, 417-445.
25. Burdick, D., Kosmoski, J., Knauer, M. F., and Glabe, C. G. (1997). Preferential
adsorption, internalization and resistance to degradation of the major isoform of
the Alzheimer's amyloid peptide, A beta 1-42, in differentiated PC12 cells.
Brain Res 746, 275-284.

131
26. Burdick, D., Soreghan, B., Kwon, M., Kosmoski, J., Knauer, M., Henschen, A.,
Yates, J., Cotman, C., and Glabe, C. (1992). Assembly and aggregation
properties of synthetic Alzheimer's A4/beta amyloid peptide analogs. J Biol
Chem 267, 546-554.
27. Butterfield, D. A. (1997). beta-Amyloid-associated free radical oxidative stress
and neurotoxicity: implications for alzheimer's disease. Chem Res Toxicol 10,
495-506.
28. Butterfield, D. A., Hensley, K., Harris, M., Mattson, M., and Carney, J. (1994).
beta-Amyloid peptide free radical fragments initiate synaptosomal
lipoperoxidation in a sequence-specific fashion: implications to alzheimer's
disease. Biochem Biophys Res Commun 200, 710-715.
29. Cai, X. D., Golde, T. E., and Younkin, S. G. (1993). Release of excess amyloid
beta protein from a mutant amyloid beta protein precursor [see comments].
Science 259, 514-516.
30. Campion, D., Dumanchin, C., Hannequin, D., Dubois, B., Belliard, S., Puel, M.,
Thomas-Anterion, C., Michon, A., Martin, C., Charbonnier, F., et al. (1999).
Early-onset autosomal dominant Alzheimer disease: prevalence, genetic
heterogeneity, and mutation spectrum. Am J Hum Genet 65, 664-670.
31. Castegna, A., Aksenov, M., Aksenova, M., Thongboonkerd, V., Klein, J. B.,
Pierce, W. M., Booze, R., Markesbery, W. R., and Butterfield, D. A. (2002).
Proteomic identification of oxidatively modified proteins in Alzheimer's disease
brain. Part I: creatine kinase BB, glutamine synthase, and ubiquitin carboxy-
terminal hydrolase L-1. Free Radic Biol Med 33, 562-571.
32. Castegna, A., Aksenov, M., Thongboonkerd, V., Klein, J. B., Pierce, W. M.,
Booze, R., Markesbery, W. R., and Butterfield, D. A. (2002). Proteomic
identification of oxidatively modified proteins in Alzheimer's disease brain. Part
II: dihydropyrimidinase-related protein 2, alpha-enolase and heat shock cognate
71. J Neurochem 82, 1524-1532.

132
33. Cataldo, A. M., Barnett, J. L., Berman, S. A., Li, J., Quarless, S., Bursztajn, S.,
Lippa, C., and Nixon, R. A. (1995). Gene expression and cellular content of
cathepsin D in Alzheimer's disease brain: evidence for early up-regulation of the
endosomal-lysosomal system. Neuron 14, 671-680.
34. Cataldo, A. M., Barnett, J. L., Mann, D. M., and Nixon, R. A. (1996).
Colocalization of lysosomal hydrolase and beta-amyloid in diffuse plaques of
the cerebellum and striatum in Alzheimer's disease and Down's syndrome. J
Neuropathol Exp Neurol 55, 704-715.
35. Cataldo, A. M., Barnett, J. L., Pieroni, C., and Nixon, R. A. (1997). Increased
neuronal endocytosis and protease delivery to early endosomes in sporadic
Alzheimer's disease: neuropathologic evidence for a mechanism of increased
beta-amyloidogenesis. J Neurosci 17, 6142-6151.
36. Cataldo, A. M., and Nixon, R. A. (1990). Enzymatically active lysosomal
proteases are associated with amyloid deposits in Alzheimer brain. Proc Natl
Acad Sci U S A 87, 3861-3865.
37. Cataldo, A. M., Peterhoff, C. M., Troncoso, J. C., Gomez-Isla, T., Hyman, B. T.,
and Nixon, R. A. (2000). Endocytic pathway abnormalities precede amyloid
beta deposition in sporadic Alzheimer's disease and Down syndrome:
differential effects of APOE genotype and presenilin mutations. Am J Pathol
157, 277-286.

38. Cataldo, A. M., Thayer, C. Y., Bird, E. D., Wheelock, T. R., and Nixon, R. A.
(1990). Lysosomal proteinase antigens are prominently localized within senile
plaques of Alzheimer's disease: evidence for a neuronal origin. Brain Res 513,
181-192.

39. Chan, S. L., Furukawa, K., and Mattson, M. P. (2002). Presenilins and APP in
neuritic and synaptic plasticity: implications for the pathogenesis of Alzheimer's
disease. Neuromolecular Med 2, 167-196.
40. Chelius, D., Zhang, T., Wang, G., and Shen, R. F. (2003). Global protein
identification and quantification technology using two-dimensional liquid
chromatography nanospray mass spectrometry. Anal Chem 75, 6658-6665.

133
41. Chyung, A. S. C., Greenberg, B. D., Cook, D. G., Doms, R. W., and Lee, V. M.
(1997). Novel beta-secretase cleavage of beta-amyloid precursor protein in the
endoplasmic reticulum/intermediate compartment of NT2N cells. Journal of Cell
Biology 138, 671-680.
42. Citron, M., Oltersdorf, T., Haass, C., McConlogue, L., Hung, A. Y., Seubert, P.,
Vigo-Pelfrey, C., Lieberburg, I., and Selkoe, D. J. (1992). Mutation of the beta-
amyloid precursor protein in familial Alzheimer's disease increases beta-protein
production. Nature 360, 672-674.
43. Citron, M., Westaway, D., Xia, W., Carlson, G., Diehl, T., Levesque, G.,
Johnson-Wood, K., Lee, M., Seubert, P., Davis, A., et al. (1997). Mutant
presenilins of Alzheimer's disease increase production of 42-residue amyloid
beta-protein in both transfected cells and transgenic mice [see comments]. Nat
Med 3, 67-72.
44. Cook, D. G., Forman, M. S., Sung, J. C., Leight, S., Kolson, D. L., Iwatsubo, T.,
Lee, V. M., and Doms, R. W. (1997). Alzheimer's A beta 1-42 is generated in
the endoplasmic reticulum/intermediate compartment of NT2N cells. Nature
Medicine 3, 1021-1023.

45. Cooney, J. R., Hurlburt, J. L., Selig, D. K., Harris, K. M., and Fiala, J. C. (2002).
Endosomal compartments serve multiple hippocampal dendritic spines from a
widespread rather than a local store of recycling membrane. J Neurosci 22,
2215-2224.

46. Corder, E. H., Saunders, A. M., Strittmatter, W. J., Schmechel, D. E., Gaskell, P.
C., Small, G. W., Roses, A. D., Haines, J. L., and Pericak-Vance, M. A. (1993).
Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer's disease
in late onset families [see comments]. Science 261, 921-923.
47. Cotman, C. W., and Anderson, A. J. (1995). A potential role for apoptosis in
neurodegeneration and Alzheimer's disease. Mol Neurobiol 10, 19-45.
48. Coyle, J. T., Oster-Granite, M. L., and Gearhart, J. D. (1986). The neurobiologic
consequences of Down syndrome. Brain Res Bull 16, 773-787.

134
49. Cripps, D., Thomas, S. N., Jeng, Y., Yang, F., Davies, P., and Yang, A. J.
(2006). Alzheimer disease-specific conformation of hyperphosphorylated paired
helical filament-Tau is polyubiquitinated through Lys-48, Lys-11, and Lys-6
ubiquitin conjugation. J Biol Chem 281, 10825-10838. Epub 12006 Jan 10827.
50. Crowther, R. A., and Wischik, C. M. (1985). Image reconstruction of the
Alzheimer paired helical filament. Embo J 4, 3661-3665.
51. Cummings, J. L. (2004). Alzheimer's disease. N Engl J Med 351, 56-67.
52. Cupers, P., Bentahir, M., Craessaerts, K., Orlans, I., Vanderstichele, H., Saftig,
P., De Strooper, B., and Annaert, W. (2001). The discrepancy between
presenilin subcellular localization and gamma-secretase processing of amyloid
precursor protein. J Cell Biol 154, 731-740.
53. Dalal, S., Rosser, M. F., Cyr, D. M., and Hanson, P. I. (2004). Distinct roles for
the AAA ATPases NSF and p97 in the secretory pathway. Mol Biol Cell 15,
637-648.
54. Dalle-Donne, I., Scaloni, A., Giustarini, D., Cavarra, E., Tell, G., Lungarella, G.,
Colombo, R., Rossi, R., and Milzani, A. (2005). Proteins as biomarkers of
oxidative/nitrosative stress in diseases: the contribution of redox proteomics.
Mass Spectrom Rev 24, 55-99.

55. Danenberg, H. D., Haring, R., Heldman, E., Gurwitz, D., Ben-Nathan, D., Pittel,
Z., Zuckerman, A., and Fisher, A. (1995). Dehydroepiandrosterone augments
M1-muscarinic receptor-stimulated amyloid precursor protein secretion in
desensitized PC12M1 cells. Ann N Y Acad Sci 774, 300-303.
56. Das, K. K., Basu, M., Li, Z. X., Basu, S., and Jungalwala, F. (1990).
Characterization of solubilized GlcAT-1 (UDP-GlcA: nLcOse4Cer beta 1-3
glucuronyltransferase) activity from embryonic chicken brain and its inhibition
by D-erythro-sphingosine. Indian J Biochem Biophys 27, 396-401.
57. Das, N., Levine, R. L., Orr, W. C., and Sohal, R. S. (2001). Selectivity of protein
oxidative damage during aging in Drosophila melanogaster. Biochem J 360,
209-216.

135
58. Dawson, V. L., and Dawson, T. M. (1998). Nitric oxide in neurodegeneration.
Prog Brain Res 118, 215-229.
59. De Koninck, P., and Schulman, H. (1998). Sensitivity of CaM kinase II to the
frequency of Ca2+ oscillations. Science 279, 227-230.
60. Dickson, D. W., Lee, S. C., Mattiace, L. A., Yen, S. H., and Brosnan, C. (1993).
Microglia and cytokines in neurological disease, with special reference to AIDS
and Alzheimer's disease. Glia 7, 75-83.
61. Duff, K., Eckman, C., Zehr, C., Yu, X., Prada, C. M., Perez-tur, J., Hutton, M.,
Buee, L., Harigaya, Y., Yager, D., et al. (1996). Increased amyloid-beta42(43)
in brains of mice expressing mutant presenilin 1. Nature 383, 710-713.
62. Ehlers, M. D. (2003). Activity level controls postsynaptic composition and
signaling via the ubiquitin-proteasome system. Nat Neurosci 6, 231-242.
63. Ehmann, W. D., Markesbery, W. R., Alauddin, M., Hossain, T. I., and Brubaker,
E. H. (1986). Brain trace elements in Alzheimer's disease. Neurotoxicology 7,
195-206.
64. Eikelenboom, P., Hack, C. E., Rozemuller, J. M., and Stam, F. C. (1989).
Complement activation in amyloid plaques in Alzheimer's dementia. Virchows
Arch [B] 56, 259-262.
65. Eikelenboom, P., and Veerhuis, R. (1996). The role of complement and
activated microglia in the pathogenesis of Alzheimer's disease. Neurobiol Aging
17, 673-680.
66. Esch, F. S., Keim, P. S., Beattie, E. C., Blacher, R. W., Culwell, A. R.,
Oltersdorf, T., McClure, D., and Ward, P. J. (1990). Cleavage of amyloid beta
peptide during constitutive processing of its precursor. Science 248, 1122-1124.

136
67. Espindola, F. S., Espreafico, E. M., Coelho, M. V., Martins, A. R., Costa, F. R.,
Mooseker, M. S., and Larson, R. E. (1992). Biochemical and immunological
characterization of p190-calmodulin complex from vertebrate brain: a novel
calmodulin-binding myosin. J Cell Biol 118, 359-368.
68. Fillit, H., Ding, W. H., Buee, L., Kalman, J., Altstiel, L., Lawlor, B., and Wolf-
Klein, G. (1991). Elevated circulating tumor necrosis factor levels in
Alzheimer's disease. Neurosci Lett 129, 318-320.
69. Finken-Eigen, M., Rohricht, R. A., and Kohrer, K. (1997). The VPS4 gene is
involved in protein transport out of a yeast pre-vacuolar endosome-like
compartment. Curr Genet 31, 469-480.
70. Frautschy, S. A., Yang, F., Irrizarry, M., Hyman, B., Saido, T. C., Hsiao, K., and
Cole, G. M. (1998). Microglial response to amyloid plaques in APPsw
transgenic mice. American Journal of Pathology 152, 307-317.

71. Fujita, H., Umezuki, Y., Imamura, K., Ishikawa, D., Uchimura, S., Nara, A.,
Yoshimori, T., Hayashizaki, Y., Kawai, J., Ishidoh, K., et al. (2004).
Mammalian class E Vps proteins, SBP1 and mVps2/CHMP2A, interact with and
regulate the function of an AAA-ATPase SKD1/Vps4B. J Cell Sci 117, 2997-
3009. Epub 2004 Jun 2991.
72. Fujita, H., Yamanaka, M., Imamura, K., Tanaka, Y., Nara, A., Yoshimori, T.,
Yokota, S., and Himeno, M. (2003). A dominant negative form of the AAA
ATPase SKD1/VPS4 impairs membrane trafficking out of endosomal/lysosomal
compartments: class E vps phenotype in mammalian cells. J Cell Sci 116, 401-
414.
73. Gaucher, S. P., Taylor, S. W., Fahy, E., Zhang, B., Warnock, D. E., Ghosh, S.
S., and Gibson, B. W. (2004). Expanded coverage of the human heart
mitochondrial proteome using multidimensional liquid chromatography coupled
with tandem mass spectrometry. J Proteome Res 3, 495-505.

137
74. Gibson, G. E., Sheu, K. F., Blass, J. P., Baker, A., Carlson, K. C., Harding, B.,
and Perrino, P. (1988). Reduced activities of thiamine-dependent enzymes in the
brains and peripheral tissues of patients with Alzheimer's disease. Arch Neurol
45, 836-840.
75. Giddings, J. C. (1984). Two-dimensional separations: concept and promise.
Anal Chem 56, 1258A-1260A, 1262A, 1264A passim.
76. Giese, K. P., Fedorov, N. B., Filipkowski, R. K., and Silva, A. J. (1998).
Autophosphorylation at Thr286 of the alpha calcium-calmodulin kinase II in
LTP and learning. Science 279, 870-873.
77. Glenner, G. G. W. C. W. (1984). Alzheimer's disease: Initial report of the
purification and characterization of a novel cerebrovascular amyloid protein.
Biochem Biophys Res Comm 120, 885-890.

78. Goate, A., Chartier-Harlin, M. C., Mullan, M., Brown, J., Crawford, F., Fidani,
L., Giuffra, L., Haynes, A., Irving, N., James, L., and et, a. (1991). Segregation
of a missense mutation in the amyloid precursor protein gene with familial
Alzheimer's disease [see comments]. Nature 349, 704-706.
79. Goedert, M., Spillantini, M. G., Cairns, N. J., and Crowther, R. A. (1992). Tau
proteins of Alzheimer paired helical filaments: abnormal phosphorylation of all
six brain isoforms. Neuron 8, 159-168.

80. Goedert, M., Spillantini, M. G., Jakes, R., Rutherford, D., and Crowther, R. A.
(1989). Multiple isoforms of human microtubule-associated protein tau:
sequences and localization in neurofibrillary tangles of Alzheimer's disease.
Neuron 3, 519-526.
81. Goldgaber, D., Lerman, M. I., McBride, O. W., Saffiotti, U., and Gajdusek, D.
C. (1987). Characterization and chromosomal localization of a cDNA encoding
brain amyloid of Alzheimer's disease. Science 235, 877-880.

138
82. Gong, C. X., Liu, F., Grundke-Iqbal, I., and Iqbal, K. (2005). Post-translational
modifications of tau protein in Alzheimer's disease. J Neural Transm 112, 813-
838. Epub 2004 Oct 2027.
83. Gong, C. X., Shaikh, S., Wang, J. Z., Zaidi, T., Grundke-Iqbal, I., and Iqbal, K.
(1995). Phosphatase activity toward abnormally phosphorylated tau: decrease in
Alzheimer disease brain. J Neurochem 65, 732-738.
84. Gong, C. X., Singh, T. J., Grundke-Iqbal, I., and Iqbal, K. (1993).
Phosphoprotein phosphatase activities in Alzheimer disease brain. J Neurochem
61, 921-927.
85. Good, P. F., Werner, P., Hsu, A., Olanow, C. W., and Perl, D. P. (1996).
Evidence of neuronal oxidative damage in Alzheimer's disease. Am J Pathol
149, 21-28.
86. Goodman, Y., and Mattson, M. P. (1994). Secreted forms of beta-amyloid
precursor protein protect hippocampal neurons against amyloid beta-peptide-
induced oxidative injury. Exp Neurol 128, 1-12.
87. Gordon, M. N., Holcomb, L. A., Jantzen, P. T., DiCarlo, G., Wilcock, D.,
Boyett, K. W., Connor, K., Melachrino, J., O'Callaghan, J. P., and Morgan, D.
(2002). Time course of the development of Alzheimer-like pathology in the
doubly transgenic PS1+APP mouse. Exp Neurol 173, 183-195.

88. Gouras, G. K., Tsai, J., Naslund, J., Vincent, B., Edgar, M., Checler, F.,
Greenfield, J. P., Haroutunian, V., Buxbaum, J. D., Xu, H., et al. (2000).
Intraneuronal Abeta42 accumulation in human brain. Am J Pathol 156, 15-20.
89. Grammas, P., and Ovase, R. (2001). Inflammatory factors are elevated in brain
microvessels in Alzheimer's disease. Neurobiol Aging 22, 837-842.
90. Gravina, S. A., Ho, L., Eckman, C. B., Long, K. E., Otvos L, Jr., Younkin, L.
H., Suzuki, N., and Younkin, S. G. (1995). Amyloid beta protein (A beta) in
Alzheimer's disease brain. Biochemical and immunocytochemical analysis with
antibodies specific for forms ending at A beta 40 or A beta 42(43). J Biol Chem
270, 7013-7016.

139
91. Grbovic, O. M., Mathews, P. M., Jiang, Y., Schmidt, S. D., Dinakar, R.,
Summers-Terio, N. B., Ceresa, B. P., Nixon, R. A., and Cataldo, A. M. (2003).
Rab5-stimulated up-regulation of the endocytic pathway increases intracellular
beta-cleaved amyloid precursor protein carboxyl-terminal fragment levels and
Abeta production. J Biol Chem 278, 31261-31268.

92. Griffin, W. S., Sheng, J. G., Royston, M. C., Gentleman, S. M., McKenzie, J. E.,
Graham, D. I., Roberts, G. W., and Mrak, R. E. (1998). Glial-neuronal
interactions in Alzheimer's disease: the potential role of a 'cytokine cycle' in
disease progression. Brain Pathol 8, 65-72.
93. Haass, C., Koo, E. H., Mellon, A., Hung, A. Y., and Selkoe, D. J. (1992).
Targeting of cell-surface beta-amyloid precursor protein to lysosomes:
alternative processing into amyloid-bearing fragments. Nature 357, 500-503.
94. Haass, C., Schlossmacher, M. G., Hung, A. Y., Vigo-Pelfrey, C., Mellon, A.,
Ostaszewski, B. L., Lieberburg, I., Koo, E. H., Schenk, D., Teplow, D. B., and
et, a. (1992). Amyloid beta-peptide is produced by cultured cells during normal
metabolism [see comments]. Nature 359, 322-325.
95. Halaban, R., Cheng, E., Zhang, Y., Moellmann, G., Hanlon, D., Michalak, M.,
Setaluri, V., and Hebert, D. N. (1997). Aberrant retention of tyrosinase in the
endoplasmic reticulum mediates accelerated degradation of the enzyme and
contributes to the dedifferentiated phenotype of amelanotic melanoma cells.
Proc Natl Acad Sci U S A 94, 6210-6215.
96. Hanson, P. I., and Schulman, H. (1992). Neuronal Ca2+/calmodulin-dependent
protein kinases. Annu Rev Biochem 61, 559-601.
97. Hardy, J. (1996). Molecular genetics of Alzheimer's disease. Acta Neurol Scand
Suppl 165, 13-17.
98. Hardy, J. (1997). The Alzheimer family of diseases: many etiologies, one
pathogenesis? Proc Natl Acad Sci U S A 94, 2095-2097.

140
99. Hardy, J., and Selkoe, D. J. (2002). The amyloid hypothesis of Alzheimer's
disease: progress and problems on the road to therapeutics. Science 297, 353-
356.
100. Head, E., McCleary, R., Hahn, F. F., Milgram, N. W., and Cotman, C. W.
(2000). Region-specific age at onset of beta-amyloid in dogs. Neurobiol Aging
21, 89-96.
101. Hendriks, L., van Duijn, C. M., Cras, P., Cruts, M., Van Hul, W., van
Harskamp, F., Warren, A., McInnis, M. G., Antonarakis, S. E., Martin, J. J., and
et, a. (1992). Presenile dementia and cerebral haemorrhage linked to a mutation
at codon 692 of the beta-amyloid precursor protein gene. Nat Genet 1, 218-221.

102. Hislop, J. N., Marley, A., and Von Zastrow, M. (2004). Role of mammalian
vacuolar protein-sorting proteins in endocytic trafficking of a non-ubiquitinated
G protein-coupled receptor to lysosomes. J Biol Chem 279, 22522-22531. Epub
22004 Mar 22515.
103. Hodges, P. E., Carrico, P. M., Hogan, J. D., O'Neill, K. E., Owen, J. J., Mangan,
M., Davis, B. P., Brooks, J. E., and Garrels, J. I. (2002). Annotating the human
proteome: the Human Proteome Survey Database (HumanPSD) and an in-depth
target database for G protein-coupled receptors (GPCR-PD) from Incyte
Genomics. Nucleic Acids Res 30, 137-141.
104. Holcomb, L., Gordon, M. N., McGowan, E., Yu, X., Benkovic, S., Jantzen, P.,
Wright, K., Saad, I., Mueller, R., Morgan, D., et al. (1998). Accelerated
Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid
precursor protein and presenilin 1 transgenes. Nat Med 4, 97-100.
105. Holcomb, L. A., Gordon, M. N., Jantzen, P., Hsiao, K., Duff, K., and Morgan,
D. (1999). Behavioral changes in transgenic mice expressing both amyloid
precursor protein and presenilin-1 mutations: lack of association with amyloid
deposits. Behav Genet 29, 177-185.
106. Howard, T. L., Stauffer, D. R., Degnin, C. R., and Hollenberg, S. M. (2001).
CHMP1 functions as a member of a newly defined family of vesicle trafficking
proteins. J Cell Sci 114, 2395-2404.

141
107. Hsiao, K., Chapman, P., Nilsen, S., Eckman, C., Harigaya, Y., Younkin, S.,
Yang, F., and Cole, G. (1996). Correlative memory deficits, Abeta elevation,
and amyloid plaques in transgenic mice. Science 274, 99-102.
108. Huang, X., Atwood, C. S., Hartshorn, M. A., Multhaup, G., Goldstein, L. E.,
Scarpa, R. C., Cuajungco, M. P., Gray, D. N., Lim, J., Moir, R. D., et al. (1999).
The A beta peptide of Alzheimer's disease directly produces hydrogen peroxide
through metal ion reduction. Biochemistry 38, 7609-7616.

109. Huh, K. H., and Wenthold, R. J. (1999). Turnover analysis of glutamate
receptors identifies a rapidly degraded pool of the N-methyl-D-aspartate
receptor subunit, NR1, in cultured cerebellar granule cells. J Biol Chem 274,
151-157.
110. Imiya, K., Ishizaki, T., Seiki, T., Saito, F., Inazawa, J., Oka, S., and Kawasaki,
T. (2002). cDNA cloning, genomic structure and chromosomal mapping of the
mouse glucuronyltransferase-S involved in the biosynthesis of the HNK-1
carbohydrate epitope. Gene 296, 29-36.
111. Jarrett, J. T., Berger, E. P., and Lansbury P.T, Jr. (1993). The carboxy terminus
of the beta amyloid protein is critical for the seeding of amyloid formation:
implications for the pathogenesis of Alzheimer's disease. Biochemistry 32,
4693-4697.
112. Jesberger, J. A., and Richardson, J. S. (1991). Oxygen free radicals and brain
dysfunction. Int J Neurosci 57, 1-17.
113. Jiang, X. S., Dai, J., Sheng, Q. H., Zhang, L., Xia, Q. C., Wu, J. R., and Zeng, R.
(2005). A comparative proteomic strategy for subcellular proteome research:
icat approach coupled with bioinformatics prediction to ascertain rat liver
mitochondrial proteins and indication of mitochondrial localization for catalase.
Mol Cell Proteomics 4, 12-34. Epub 2004 Oct 2025.
114. Jorm, A. F. (1991). Cross-national comparisons of the occurrence of Alzheimer's
and vascular dementias. Eur Arch Psychiatry Clin Neurosci 240, 218-222.

142
115. Kandel, E. R. (2001). The molecular biology of memory storage: a dialog
between genes and synapses. Biosci Rep 21, 565-611.
116. Kang, J., Lemaire, H. G., Unterbeck, A., Salbaum, J. M., Masters, C. L.,
Grzeschik, K. H., Multhaup, G., Beyreuther, K., and Muller-Hill, B. (1987). The
precursor of Alzheimer's disease amyloid A4 protein resembles a cell-surface
receptor. Nature 325, 733-736.

117. Kato, A., Rouach, N., Nicoll, R. A., and Bredt, D. S. (2005). Activity-dependent
NMDA receptor degradation mediated by retrotranslocation and ubiquitination.
Proc Natl Acad Sci U S A 102, 5600-5605. Epub 2005 Apr 5604.

118. Katoh, K., Shibata, H., Suzuki, H., Nara, A., Ishidoh, K., Kominami, E.,
Yoshimori, T., and Maki, M. (2003). The ALG-2-interacting protein Alix
associates with CHMP4b, a human homologue of yeast Snf7 that is involved in
multivesicular body sorting. J Biol Chem 278, 39104-39113.
119. Kawarabayashi, T., Younkin, L. H., Saido, T. C., Shoji, M., Ashe, K. H., and
Younkin, S. G. (2001). Age-dependent changes in brain, CSF, and plasma
amyloid (beta) protein in the Tg2576 transgenic mouse model of Alzheimer's
disease. J Neurosci 21, 372-381.
120. Kayed, R., Head, E., Thompson, J. L., McIntire, T. M., Milton, S. C., Cotman,
C. W., and Glabe, C. G. (2003). Common structure of soluble amyloid
oligomers implies common mechanism of pathogenesis. Science 300, 486-489.
121. Kayed, R., Sokolov, Y., Edmonds, B., McIntire, T. M., Milton, S. C., Hall, J. E.,
and Glabe, C. G. (2004). Permeabilization of lipid bilayers is a common
conformation-dependent activity of soluble amyloid oligomers in protein
misfolding diseases. J Biol Chem 279, 46363-46366. Epub 42004 Sep 46321.
122. Keller, A., Eng, J., Zhang, N., Li, X. J., and Aebersold, R. (2005). A uniform
proteomics MS/MS analysis platform utilizing open XML file formats. Mol Syst
Biol 1, 2005 0017.

143
123. Keller, A., Nesvizhskii, A. I., Kolker, E., and Aebersold, R. (2002). Empirical
statistical model to estimate the accuracy of peptide identifications made by
MS/MS and database search. Anal Chem 74, 5383-5392.
124. Kitaguchi, N., Takahashi, Y., Tokushima, Y., Shiojiri, S., and Ito, H. (1988).
Novel precursor of Alzheimer's disease amyloid protein shows protease
inhibitory activity. Nature 331, 530-532.

125. Knauer, M. F., Soreghan, B., Burdick, D., Kosmoski, J., and Glabe, C. G.
(1992). Intracellular accumulation and resistance to degradation of the
Alzheimer amyloid A4/beta protein. Proc Natl Acad Sci U S A 89, 7437-7441.
126. Kontush, A. (2001). Amyloid-beta: an antioxidant that becomes a pro-oxidant
and critically contributes to Alzheimer's disease. Free Radic Biol Med 31, 1120-
1131.
127. Kontush, A., Berndt, C., Weber, W., Akopyan, V., Arlt, S., Schippling, S., and
Beisiegel, U. (2001). Amyloid-beta is an antioxidant for lipoproteins in
cerebrospinal fluid and plasma. Free Radic Biol Med 30, 119-128.
128. Koo, E. H., and Squazzo, S. L. (1994). Evidence that production and release of
amyloid beta-protein involves the endocytic pathway. J Biol Chem 269, 17386-
17389.
129. Kopke, E., Tung, Y. C., Shaikh, S., Alonso, A. C., Iqbal, K., and Grundke-Iqbal,
I. (1993). Microtubule-associated protein tau. Abnormal phosphorylation of a
non-paired helical filament pool in Alzheimer disease. J Biol Chem 268, 24374-
24384.
130. Korolainen, M. A., Goldsteins, G., Alafuzoff, I., Koistinaho, J., and Pirttila, T.
(2002). Proteomic analysis of protein oxidation in Alzheimer's disease brain.
Electrophoresis 23, 3428-3433.
131. Kotilinek, L. A., Bacskai, B., Westerman, M., Kawarabayashi, T., Younkin, L.,
Hyman, B. T., Younkin, S., and Ashe, K. H. (2002). Reversible memory loss in
a mouse transgenic model of Alzheimer's disease. J Neurosci 22, 6331-6335.

144
132. Lanz, T. A., Carter, D. B., and Merchant, K. M. (2003). Dendritic spine loss in
the hippocampus of young PDAPP and Tg2576 mice and its prevention by the
ApoE2 genotype. Neurobiol Dis 13, 246-253.
133. Lee, S. C., Zhao, M. L., Hirano, A., and Dickson, D. W. (1999). Inducible nitric
oxide synthase immunoreactivity in the Alzheimer disease hippocampus:
association with Hirano bodies, neurofibrillary tangles, and senile plaques. J
Neuropathol Exp Neurol 58, 1163-1169.
134. Lee, V. M. (1996). Regulation of tau phosphorylation in Alzheimer's disease.
Ann N Y Acad Sci 777, 107-113.
135. Lesne, S., Koh, M. T., Kotilinek, L., Kayed, R., Glabe, C. G., Yang, A.,
Gallagher, M., and Ashe, K. H. (2006). A specific amyloid-beta protein
assembly in the brain impairs memory. Nature 440, 352-357.
136. Levine, R. L., Garland, D., Oliver, C. N., Amici, A., Climent, I., Lenz, A. G.,
Ahn, B. W., Shaltiel, S., and Stadtman, E. R. (1990). Determination of carbonyl
content in oxidatively modified proteins. Methods Enzymol 186, 464-478.
137. Levine, R. L., Williams, J. A., Stadtman, E. R., and Shacter, E. (1994). Carbonyl
assays for determination of oxidatively modified proteins. Methods Enzymol
233, 346-357.
138. Levy-Lahad, E., Wasco, W., Poorkaj, P., Romano, D. M., Oshima, J., Pettingell,
W. H., Yu, C. E., Jondro, P. D., Schmidt, S. D., Wang, K., and et al. (1995).
Candidate gene for the chromosome 1 familial Alzheimer's disease locus.
Science 269, 973-977.
139. Lin, Y., Kimpler, L. A., Naismith, T. V., Lauer, J. M., and Hanson, P. I. (2005).
Interaction of the mammalian endosomal sorting complex required for transport
(ESCRT) III protein hSnf7-1 with itself, membranes, and the AAA+ ATPase
SKD1. J Biol Chem 280, 12799-12809.
140. Link, A. J., Eng, J., Schieltz, D. M., Carmack, E., Mize, G. J., Morris, D. R.,
Garvik, B. M., and Yates, J. R., 3rd (1999). Direct analysis of protein complexes
using mass spectrometry. Nat Biotechnol 17, 676-682.

145
141. Lippa, C. F., Saunders, A. M., Smith, T. W., Swearer, J. M., Drachman, D. A.,
Ghetti, B., Nee, L., Pulaski-Salo, D., Dickson, D., Robitaille, Y., et al. (1996).
Familial and sporadic Alzheimer's disease: neuropathology cannot exclude a
final common pathway. Neurology 46, 406-412.
142. Lonze, B. E., and Ginty, D. D. (2002). Function and regulation of CREB family
transcription factors in the nervous system. Neuron 35, 605-623.
143. Lue, L. F., Brachova, L., Civin, W. H., and Rogers, J. (1996). Inflammation, A
beta deposition, and neurofibrillary tangle formation as correlates of Alzheimer's
disease neurodegeneration. J Neuropathol Exp Neurol 55, 1083-1088.
144. Ma, Y. S., Chao, C. C., and Stadtman, E. R. (1999). Oxidative modification of
glutamine synthetase by 2,2'-azobis(2- amidinopropane) dihydrochloride. Arch
Biochem Biophys 363, 129-134.
145. Mann, D. M. (1988). Alzheimer's disease and Down's syndrome. Histopathology
13, 125-137.
146. Manza, L. L., Codreanu, S. G., Stamer, S. L., Smith, D. L., Wells, K. S.,
Roberts, R. L., and Liebler, D. C. (2004). Global shifts in protein sumoylation in
response to electrophile and oxidative stress. Chem Res Toxicol 17, 1706-1715.
147. Mark, R. J., Lovell, M. A., Markesbery, W. R., Uchida, K., and Mattson, M. P.
(1997). A role for 4-hydroxynonenal, an aldehydic product of lipid peroxidation,
in disruption of ion homeostasis and neuronal death induced by amyloid beta-
peptide. J Neurochem 68, 255-264.
148. Markesbery, W. R. (1997). Oxidative stress hypothesis in Alzheimer's disease.
Free Radic Biol Med 23, 134-147.
149. Markesbery, W. R., and Carney, J. M. (1999). Oxidative alterations in
Alzheimer's disease. Brain Pathol 9, 133-146.

146
150. Masters, C. L., Simms, G., Weinman, N. A., Multhaup, G., and McDonald, B. L.
B. K. (1985). Amyloid plaque core protein in alzheimers disease and Down
syndrome. Proc Natl Acad Sci USA 82, 4245-4249.
151. Mathews, P. M., Guerra, C. B., Jiang, Y., Grbovic, O. M., Kao, B. H., Schmidt,
S. D., Dinakar, R., Mercken, M., Hille-Rehfeld, A., Rohrer, J., et al. (2002).
Alzheimer's disease-related overexpression of the cation-dependent mannose 6-
phosphate receptor increases Abeta secretion: role for altered lysosomal
hydrolase distribution in beta-amyloidogenesis. J Biol Chem 277, 5299-5307.
152. Matsuoka, Y., Picciano, M., La Francois, J., and Duff, K. (2001). Fibrillar beta-
amyloid evokes oxidative damage in a transgenic mouse model of Alzheimer's
disease. Neuroscience 104, 609-613.

153. Matsuoka, Y., Picciano, M., Malester, B., LaFrancois, J., Zehr, C., Daeschner, J.
M., Olschowka, J. A., Fonseca, M. I., O'Banion, M. K., Tenner, A. J., et al.
(2001). Inflammatory responses to amyloidosis in a transgenic mouse model of
Alzheimer's disease. Am J Pathol 158, 1345-1354.
154. Mattson, M. P., Barger, S. W., Begley, J. G., and Mark, R. J. (1995). Calcium,
free radicals, and excitotoxic neuronal death in primary cell culture. Methods
Cell Biol 46, 187-216.
155. Mayford, M., Bach, M. E., Huang, Y. Y., Wang, L., Hawkins, R. D., and
Kandel, E. R. (1996). Control of memory formation through regulated
expression of a CaMKII transgene. Science 274, 1678-1683.
156. McCormack, A. L., Schieltz, D. M., Goode, B., Yang, S., Barnes, G., Drubin,
D., and Yates, J. R., 3rd (1997). Direct analysis and identification of proteins in
mixtures by LC/MS/MS and database searching at the low-femtomole level.
Anal Chem 69, 767-776.
157. McLean, C. A., Cherny, R. A., Fraser, F. W., Fuller, S. J., Smith, M. J.,
Beyreuther, K., Bush, A. I., and Masters, C. L. (1999). Soluble pool of Abeta
amyloid as a determinant of severity of neurodegeneration in Alzheimer's
disease. Ann Neurol 46, 860-866.

147
158. Mesulam, M. M. (1999). Neuroplasticity failure in Alzheimer's disease: bridging
the gap between plaques and tangles. Neuron 24, 521-529.
159. Miller, D. L., Papayannopoulos, I. A., Styles, J., Bobin, S. A., Lin, Y. Y.,
Biemann, K., and Iqbal, K. (1993). Peptide compositions of the cerebrovascular
and senile plaque core amyloid deposits of Alzheimer's disease. Arch Biochem
Biophys 301, 41-52.
160. Mori, H., Takio, K., Ogawara, M., and Selkoe, D. J. (1992). Mass spectrometry
of purified amyloid beta protein in Alzheimer's disease. J Biol Chem 267,
17082-17086.
161. Mosmann, T. (1983). Rapid colorimetric assay for cellular growth and survival:
application to proliferation and cytotoxicity assays. J Immunol Methods 65, 55-
63.
162. Mullan, M., Houlden, H., Windelspecht, M., Fidani, L., Lombardi, C., Diaz, P.,
Rossor, M., Crook, R., Hardy, J., Duff, K., and et, a. (1992). A locus for familial
early-onset Alzheimer's disease on the long arm of chromosome 14, proximal to
the alpha 1-antichymotrypsin gene. Nat Genet 2, 340-342.
163. Munger, J. S., Haass, C., Lemere, C. A., Shi, G. P., Wong, W. S., Teplow, D. B.,
Selkoe, D. J., and Chapman, H. A. (1995). Lysosomal processing of amyloid
precursor protein to A beta peptides: a distinct role for cathepsin S. Biochem J
311, 299-305.
164. Nakamura, A., and Goto, S. (1996). Analysis of protein carbonyls with 2,4-
dinitrophenyl hydrazine and its antibodies by immunoblot in two-dimensional
gel electrophoresis. J Biochem (Tokyo) 119, 768-774.
165. Nakanishi, H., Amano, T., Sastradipura, D. F., Yoshimine, Y., Tsukuba, T.,
Tanabe, K., Hirotsu, I., Ohono, T., and Yamamoto, K. (1997). Increased
expression of cathepsins E and D in neurons of the aged rat brain and their
colocalization with lipofuscin and carboxy-terminal fragments of Alzheimer
amyloid precursor protein. J Neurochem 68, 739-749.

148
166. Naslund, J., Haroutunian, V., Mohs, R., Davis, K. L., Davies, P., Greengard, P.,
and Buxbaum, J. D. (2000). Correlation between elevated levels of amyloid
beta-peptide in the brain and cognitive decline. Jama 283, 1571-1577.
167. Nesvizhskii, A. I., Keller, A., Kolker, E., and Aebersold, R. (2003). A statistical
model for identifying proteins by tandem mass spectrometry. Anal Chem 75,
4646-4658.
168. Nikolsky, Y., Ekins, S., Nikolskaya, T., and Bugrim, A. (2005). A novel method
for generation of signature networks as biomarkers from complex high
throughput data. Toxicol Lett 158, 20-29.
169. Nixon, R. A., Cataldo, A. M., and Mathews, P. M. (2000). The endosomal-
lysosomal system of neurons in Alzheimer's disease pathogenesis: a review.
Neurochem Res 25, 1161-1172.
170. Nixon, R. A., Cataldo, A. M., Paskevich, P. A., Hamilton, D. J., Wheelock, T.
R., and Kanaley-Andrews, L. (1992). The lysosomal system in neurons.
Involvement at multiple stages of Alzheimer's disease pathogenesis. Ann N Y
Acad Sci 674, 65-88.
171. Nunomura, A., Perry, G., Pappolla, M. A., Wade, R., Hirai, K., Chiba, S., and
Smith, M. A. (1999). RNA oxidation is a prominent feature of vulnerable
neurons in Alzheimer's disease. J Neurosci 19, 1959-1964.
172. Nussbaum, R. L., and Ellis, C. E. (2003). Alzheimer's disease and Parkinson's
disease. N Engl J Med 348, 1356-1364.
173. Offe, K., Dodson, S. E., Shoemaker, J. T., Fritz, J. J., Gearing, M., Levey, A. I.,
and Lah, J. J. (2006). The lipoprotein receptor LR11 regulates amyloid beta
production and amyloid precursor protein traffic in endosomal compartments. J
Neurosci 26, 1596-1603.
174. Oka, S., Terayama, K., Kawashima, C., and Kawasaki, T. (1992). A novel
glucuronyltransferase in nervous system presumably associated with the
biosynthesis of HNK-1 carbohydrate epitope on glycoproteins. J Biol Chem 267,
22711-22714.

149
175. Oltersdorf, T., Ward, P. J., Henriksson, T., Beattie, E. C., Neve, R., Lieberburg,
I., and Fritz, L. C. (1990). The Alzheimer amyloid precursor protein.
Identification of a stable intermediate in the biosynthetic/degradative pathway. J
Biol Chem 265, 4492-4497.
176. Ono, Y., Kinouchi, T., Sorimachi, H., Ishiura, S., and Suzuki, K. (1997).
Deletion of an endosomal/lysosomal targeting signal promotes the secretion of
Alzheimer's disease amyloid precursor protein (APP). J Biochem (Tokyo) 121,
585-590.
177. Ouzzine, M., Gulberti, S., Netter, P., Magdalou, J., and Fournel-Gigleux, S.
(2000). Structure/function of the human Ga1beta1,3-glucuronosyltransferase.
Dimerization and functional activity are mediated by two crucial cysteine
residues. J Biol Chem 275, 28254-28260.
178. Pappolla, M. A., Omar, R. A., Kim, K. S., and Robakis, N. K. (1992).
Immunohistochemical evidence of antioxidant stress in Alzheimer's disease. Am
J Pathol 140, 621-628.
179. Peng, J., and Gygi, S. P. (2001). Proteomics: the move to mixtures. J Mass
Spectrom 36, 1083-1091.
180. Perez, R. G., Soriano, S., Hayes, J. D., Ostaszewski, B., Xia, W., Selkoe, D. J.,
Chen, X., Stokin, G. B., and Koo, E. H. (1999). Mutagenesis identifies new
signals for beta-amyloid precursor protein endocytosis, turnover, and the
generation of secreted fragments, including Abeta42. J Biol Chem 274, 18851-
18856.
181. Perry, G., Nunomura, A., Hirai, K., Zhu, X., Prez, M., Avila, J., Castellani, R. J.,
Atwood, C. S., Aliev, G., Sayre, L. M., et al. (2002). Is oxidative damage the
fundamental pathogenic mechanism of Alzheimer's and other neurodegenerative
diseases? Free Radic Biol Med 33, 1475-1479.
182. Picklo, M. J., Montine, T. J., Amarnath, V., and Neely, M. D. (2002). Carbonyl
toxicology and Alzheimer's disease. Toxicol Appl Pharmacol 184, 187-197.

150
183. Pike, C. J., Walencewicz, A. J., Glabe, C. G., and Cotman, C. W. (1991). In
vitro aging of beta-amyloid protein causes peptide aggregation and
neurotoxicity. Brain Res 563, 311-314.
184. Ponte, P., Gonzalez-DeWhitt, P., Schilling, J., Miller, J., Hsu, D., Greenberg, B.,
Davis, K., Wallace, W., Lieberburg, I., and Fuller, F. (1988). A new A4 amyloid
mRNA contains a domain homologous to serine proteinase inhibitors. Nature
331, 525-527.
185. Price, J. M., Chi, X., Hellermann, G., and Sutton, E. T. (2001). Physiological
levels of beta-amyloid induce cerebral vessel dysfunction and reduce endothelial
nitric oxide production. Neurol Res 23, 506-512.
186. Prydz, K., and Dalen, K. T. (2000). Synthesis and sorting of proteoglycans. J
Cell Sci 113 Pt 2, 193-205.
187. Ranish, J. A., Yi, E. C., Leslie, D. M., Purvine, S. O., Goodlett, D. R., Eng, J.,
and Aebersold, R. (2003). The study of macromolecular complexes by
quantitative proteomics. Nat Genet 33, 349-355.
188. Raymond, C. K., Howald-Stevenson, I., Vater, C. A., and Stevens, T. H. (1992).
Morphological classification of the yeast vacuolar protein sorting mutants:
evidence for a prevacuolar compartment in class E vps mutants. Mol Biol Cell 3,
1389-1402.
189. Resing, K. A., Meyer-Arendt, K., Mendoza, A. M., Aveline-Wolf, L. D.,
Jonscher, K. R., Pierce, K. G., Old, W. M., Cheung, H. T., Russell, S., Wattawa,
J. L., et al. (2004). Improving reproducibility and sensitivity in identifying
human proteins by shotgun proteomics. Anal Chem 76, 3556-3568.
190. Rieder, S. E., Banta, L. M., Kohrer, K., McCaffery, J. M., and Emr, S. D.
(1996). Multilamellar endosome-like compartment accumulates in the yeast
vps28 vacuolar protein sorting mutant. Mol Biol Cell 7, 985-999.
191. Riemenschneider, M., Schoepfer-Wendels, A., Friedrich, P., Konta, L., Laws, S.
M., Mueller, J. C., Kurz, A., and Forstl, H. (2006). No association of Vacuolar
protein sorting 26 polymorphisms with Alzheimer's disease. Neurobiol Aging.

151
192. Robakis, N. K., Wisniewski, H. M., Jenkins, E. C., Devine-Gage, E. A., Houck,
G. E., Yao, X. L., Ramakrishna, N., Wolfe, G., Silverman, W. P., and Brown,
W. T. (1987). Chromosome 21q21 sublocalisation of gene encoding beta-
amyloid peptide in cerebral vessels and neuritic (senile) plaques of people with
Alzheimer disease and Down syndrome [letter]. Lancet 1, 384-385.
193. Rogaeva, E., Meng, Y., Lee, J. H., Gu, Y., Kawarai, T., Zou, F., Katayama, T.,
Baldwin, C. T., Cheng, R., Hasegawa, H., et al. (2007). The neuronal sortilin-
related receptor SORL1 is genetically associated with Alzheimer disease. Nat
Genet.
194. Roher, A. E., Lowenson, J. D., Clarke, S., Woods, A. S., Cotter, R. J., Gowing,
E., and Ball, M. J. (1993). beta-Amyloid-(1-42) is a major component of
cerebrovascular amyloid deposits: implications for the pathology of alzheimer
disease. Proc Natl Acad Sci U S A 90, 10836-10840.
195. Saghatelyan, A. K., Gorissen, S., Albert, M., Hertlein, B., Schachner, M., and
Dityatev, A. (2000). The extracellular matrix molecule tenascin-R and its HNK-
1 carbohydrate modulate perisomatic inhibition and long-term potentiation in
the CA1 region of the hippocampus. Eur J Neurosci 12, 3331-3342.
196. Sayre, L. M., Zelasko, D. A., Harris, P. L., Perry, G., Salomon, R. G., and
Smith, M. A. (1997). 4-Hydroxynonenal-derived advanced lipid peroxidation
end products are increased in Alzheimer's disease. J Neurochem 68, 2092-2097.
197. Schachner, M. (1997). Neural recognition molecules and synaptic plasticity.
Curr Opin Cell Biol 9, 627-634.
198. Schellenberg, G. D., Kamino, K., Bryant, E. M., Moore, D., and Bird, T. D.
(1992). Genetic heterogeneity, Down syndrome, and Alzheimer disease. Prog
Clin Biol Res 379, 215-226.
199. Scheuner, D., Eckman, C., Jensen, M., Song, X., Citron, M., Suzuki, N., Bird, T.
D., Hardy, J., Hutton, M., Kukull, W., et al. (1996). Secreted amyloid beta-
protein similar to that in the senile plaques of Alzheimer's disease is increased in
vivo by the presenilin 1 and 2 and APP mutations linked to familial Alzheimer's
disease [see comments]. Nat Med 2, 864-870.

152
200. Schmidt, M. W., Houseman, A., Ivanov, A. R., and Wolf, D. A. (2007).
Comparative proteomic and transcriptomic profiling of the fission yeast
Schizosaccharomyces pombe. Mol Syst Biol 3, 79.
201. Schwagerl, A. L., Mohan, P. S., Cataldo, A. M., Vonsattel, J. P., and Kowall, N.
W. N. R. A. (1995). Elevated levels of the endosomal-lysosomal proteinase
cathepsin D in cerebrospinal fluid in Alzheimer disease. J Neurochem 64, 443-
446.
202. Selkoe, D. J. (2001). Alzheimer's disease: genes, proteins, and therapy. Physiol
Rev 81, 741-766.
203. Selkoe, D. J. (2003). Aging, amyloid, and Alzheimer's disease: a perspective in
honor of Carl Cotman. Neurochem Res 28, 1705-1713.
204. Senn, C., Kutsche, M., Saghatelyan, A., Bosl, M. R., Lohler, J., Bartsch, U.,
Morellini, F., and Schachner, M. (2002). Mice deficient for the HNK-1
sulfotransferase show alterations in synaptic efficacy and spatial learning and
memory. Mol Cell Neurosci 20, 712-729.
205. Seubert, P., Vigo-Pelfrey, C., Esch, F., Lee, M., Dovey, H., Davis, D., Sinha, S.,
Schlossmacher, M., Whaley, J., Swindlehurst, C., and et, a. (1992). Isolation and
quantification of soluble Alzheimer's beta-peptide from biological fluids [see
comments]. Nature 359, 325-327.
206. Shacter, E. (2000). Quantification and significance of protein oxidation in
biological samples. Drug Metab Rev 32, 307-326.
207. Shearman, M. S. (1996). Cellular MTT reduction distinguishes the mechanism
of action of beta-amyloid from that of tachykinin receptor peptides.
Neuropeptides 30, 125-132.
208. Shearman, M. S., Ragan, C. I., and Iversen, L. L. (1994). Inhibition of PC12 cell
redox activity is a specific, early indicator of the mechanism of beta-amyloid-
mediated cell death. Proc Natl Acad Sci U S A 91, 1470-1474.

153
209. Sherrington, R., Rogaev, E. I., Liang, Y., Rogaeva, E. A., Levesque, G., Ikeda,
M., Chi, H., Lin, C., Li, G., Holman, K., and et al. (1995). Cloning of a gene
bearing missense mutations in early-onset familial Alzheimer's disease [see
comments]. Nature 375, 754-760.
210. Shoji, M., Golde, T. E., Ghiso, J., Cheung, T. T., Estus, S., Shaffer, L. M., Cai,
X. D., McKay, D. M., Tintner, R., Frangione, B., and et al. (1992). Production
of the Alzheimer amyloid beta protein by normal proteolytic processing. Science
258, 126-129.
211. Shringarpure, R., and Davies, K. J. (2002). Protein turnover by the proteasome
in aging and disease. Free Radic Biol Med 32, 1084-1089.
212. Sisodia, S. S., Koo, E. H., Beyreuther, K., Unterbeck, A., and Price, D. L.
(1990). Evidence that beta-amyloid protein in Alzheimer's disease is not derived
by normal processing. Science 248, 492-495.
213. Small, S. A., and Gandy, S. (2006). Sorting through the cell biology of
Alzheimer's disease: intracellular pathways to pathogenesis. Neuron 52, 15-31.
214. Small, S. A., Kent, K., Pierce, A., Leung, C., Kang, M. S., Okada, H., Honig, L.,
Vonsattel, J. P., and Kim, T. W. (2005). Model-guided microarray implicates the
retromer complex in Alzheimer's disease. Ann Neurol 58, 909-919.
215. Smith, C. D., Carney, J. M., Starke-Reed, P. E., Oliver, C. N., Stadtman, E. R.,
Floyd, R. A., and Markesbery, W. R. (1991). Excess brain protein oxidation and
enzyme dysfunction in normal aging and in Alzheimer disease. Proc Natl Acad
Sci U S A 88, 10540-10543.
216. Smith, C. D., Carney, J. M., Tatsumo, T., Stadtman, E. R., Floyd, R. A., and
Markesbery, W. R. (1992). Protein oxidation in aging brain. Ann N Y Acad Sci
663, 110-119.
217. Smith, M. A., Hirai, K., Hsiao, K., Pappolla, M. A., Harris, P. L., Siedlak, S. L.,
Tabaton, M., and Perry, G. (1998). Amyloid-beta deposition in Alzheimer
transgenic mice is associated with oxidative stress. J Neurochem 70, 2212-2215.

154
218. Smith, M. A., Perry, G., Richey, P. L., Sayre, L. M., Anderson, V. E., Beal, M.
F., and Kowall, N. (1996). Oxidative damage in Alzheimer's. Nature 382, 120-
121.
219. Smith, M. A., Richey Harris, P. L., Sayre, L. M., Beckman, J. S., and Perry, G.
(1997). Widespread peroxynitrite-mediated damage in Alzheimer's disease. J
Neurosci 17, 2653-2657.
220. Smith, M. A., Richey, P. L., Taneda, S., Kutty, R. K., Sayre, L. M., Monnier, V.
M., and Perry, G. (1994). Advanced Maillard reaction end products, free
radicals, and protein oxidation in Alzheimer's disease. Ann N Y Acad Sci 738,
447-454.
221. Smith, M. A., Rottkamp, C. A., Nunomura, A., Raina, A. K., and Perry, G.
(2000). Oxidative stress in Alzheimer's disease. Biochim Biophys Acta 1502,
139-144.
222. Smith, M. A., Sayre, L. M., Anderson, V. E., Harris, P. L., Beal, M. F., Kowall,
N., and Perry, G. (1998). Cytochemical demonstration of oxidative damage in
Alzheimer disease by immunochemical enhancement of the carbonyl reaction
with 2,4-dinitrophenylhydrazine. J Histochem Cytochem 46, 731-735.
223. Smith, M. A., Taneda, S., Richey, P. L., Miyata, S., Yan, S. D., Stern, D., Sayre,
L. M., Monnier, V. M., and Perry, G. (1994). Advanced Maillard reaction end
products are associated with Alzheimer disease pathology [published erratum
appears in Proc Natl Acad Sci U S A 1995 Feb 28;92(5):2016]. Proc Natl Acad
Sci U S A 91, 5710-5714.
224. Soreghan, B., Thomas, S. N., and Yang, A. J. (2003). Aberrant
sphingomyelin/ceramide metabolic-induced neuronal endosomal/lysosomal
dysfunction: potential pathological consequences in age-related
neurodegeneration. Adv Drug Deliv Rev 55, 1515-1524.
225. Soreghan, B. A., Lu, B. W., Thomas, S. N., Duff, K., Rakhmatulin, E. A.,
Nikolskaya, T., Chen, T., and Yang, A. J. (2005). Using proteomics and network
analysis to elucidate the consequences of synaptic protein oxidation in a PS1 +
AbetaPP mouse model of Alzheimer's disease. J Alzheimers Dis 8, 227-241.

155
226. Soreghan, B. A., Yang, F., Thomas, S. N., Hsu, J., and Yang, A. J. (2003). High-
throughput proteomic-based identification of oxidatively induced protein
carbonylation in mouse brain. Pharm Res 20, 1713-1720.
227. Soriano, S., Chyung, A. S., Chen, X., Stokin, G. B., Lee, V. M., and Koo, E. H.
(1999). Expression of beta-amyloid precursor protein-CD3gamma chimeras to
demonstrate the selective generation of amyloid beta(1-40) and amyloid beta(1-
42) peptides within secretory and endocytic compartments. J Biol Chem 274,
32295-32300.
228. Spoelgen, R., von Arnim, C. A., Thomas, A. V., Peltan, I. D., Koker, M., Deng,
A., Irizarry, M. C., Andersen, O. M., Willnow, T. E., and Hyman, B. T. (2006).
Interaction of the cytosolic domains of sorLA/LR11 with the amyloid precursor
protein (APP) and beta-secretase beta-site APP-cleaving enzyme. J Neurosci 26,
418-428.
229. Stephan, A., Laroche, S., and Davis, S. (2001). Generation of aggregated beta-
amyloid in the rat hippocampus impairs synaptic transmission and plasticity and
causes memory deficits. J Neurosci 21, 5703-5714.
230. Strekalova, T., Wotjak, C. T., and Schachner, M. (2001). Intrahippocampal
administration of an antibody against the HNK-1 carbohydrate impairs memory
consolidation in an inhibitory learning task in mice. Mol Cell Neurosci 17,
1102-1113.
231. Strittmatter, W. J., Saunders, A. M., Schmechel, D., Pericak-Vance, M.,
Enghild, J., Salvesen, G. S., and Roses, A. D. (1993). Apolipoprotein E: high-
avidity binding to beta-amyloid and increased frequency of type 4 allele in late-
onset familial Alzheimer disease. Proc Natl Acad Sci U S A 90, 1977-1981.
232. Su, Y., Balice-Gordon, R. J., Hess, D. M., Landsman, D. S., Minarcik, J.,
Golden, J., Hurwitz, I., Liebhaber, S. A., and Cooke, N. E. (2004).
Neurobeachin is essential for neuromuscular synaptic transmission. J Neurosci
24, 3627-3636.

156
233. Suter, M., Reme, C., Grimm, C., Wenzel, A., Jaattela, M., Esser, P., Kociok, N.,
Leist, M., and Richter, C. (2000). Age-related macular degeneration. The
lipofusion component N-retinyl-N-retinylidene ethanolamine detaches
proapoptotic proteins from mitochondria and induces apoptosis in mammalian
retinal pigment epithelial cells. J Biol Chem 275, 39625-39630.

234. Suzuki, N., Cheung, T. T., Cai, X. D., Odaka, A., Otvos L, Jr., Eckman, C.,
Golde, T. E., and Younkin, S. G. (1994). An increased percentage of long
amyloid beta protein secreted by familial amyloid beta protein precursor (beta
APP717) mutants. Science 264, 1336-1340.
235. Takahashi, R. H., Almeida, C. G., Kearney, P. F., Yu, F., Lin, M. T., Milner, T.
A., and Gouras, G. K. (2004). Oligomerization of Alzheimer's beta-amyloid
within processes and synapses of cultured neurons and brain. J Neurosci 24,
3592-3599.
236. Takahashi, R. H., Milner, T. A., Li, F., Nam, E. E., Edgar, M. A., Yamaguchi,
H., Beal, M. F., Xu, H., Greengard, P., and Gouras, G. K. (2002). Intraneuronal
Alzheimer abeta42 accumulates in multivesicular bodies and is associated with
synaptic pathology. Am J Pathol 161, 1869-1879.
237. Tan, C., Mui, A., and Dedhar, S. (2002). Integrin-linked kinase regulates
inducible nitric oxide synthase and cyclooxygenase-2 expression in an NF-
kappa B-dependent manner. J Biol Chem 277, 3109-3116.
238. Tanzi, R. E., Gusella, J. F., Watkins, P. C., Bruns, G. A., St George-Hyslop, P.,
Van Keuren, M. L., Patterson, D., Pagan, S., Kurnit, D. M., and Neve, R. L.
(1987). Amyloid beta protein gene: cDNA, mRNA distribution, and genetic
linkage near the Alzheimer locus. Science 235, 880-884.
239. Terai, K., Iwai, A., Kawabata, S., Sasamata, M., Miyata, K., and Yamaguchi, T.
(2001). Apolipoprotein E deposition and astrogliosis are associated with
maturation of beta-amyloid plaques in betaAPPswe transgenic mouse:
Implications for the pathogenesis of Alzheimer's disease. Brain Res 900, 48-56.

157
240. Terry, R. D., Masliah, E., Salmon, D. P., Butters, N., DeTeresa, R., Hill, R.,
Hansen, L. A., and Katzman, R. (1991). Physical basis of cognitive alterations in
Alzheimer's disease: synapse loss is the major correlate of cognitive impairment.
Ann Neurol 30, 572-580.

241. Thinakaran, G., Teplow, D. B., Siman, R., Greenberg, B., and Sisodia, S. S.
(1996). Metabolism of the `Swedish` amyloid precursor protein variant in
neuro2a (N2a) cells. Evidence that cleavage at the `beta-secretase` site occurs in
the golgi apparatus. J Biol Chem 271, 9390-9397.
242. Tsang, H. T., Connell, J. W., Brown, S. E., Thompson, A., Reid, E., and
Sanderson, C. M. (2006). A systematic analysis of human CHMP protein
interactions: additional MIT domain-containing proteins bind to multiple
components of the human ESCRT III complex. Genomics 88, 333-346.
243. Vajjhala, P. R., Wong, J. S., To, H. Y., and Munn, A. L. (2006). The beta
domain is required for Vps4p oligomerization into a functionally active ATPase.
Febs J 273, 2357-2373.
244. Vale, R. D. (2000). AAA proteins. Lords of the ring. J Cell Biol 150, F13-19.
245. Walsh, D. M., Klyubin, I., Fadeeva, J. V., Cullen, W. K., Anwyl, R., Wolfe, M.
S., Rowan, M. J., and Selkoe, D. J. (2002). Naturally secreted oligomers of
amyloid beta protein potently inhibit hippocampal long-term potentiation in
vivo. Nature 416, 535-539.
246. Wang, H. W., Pasternak, J. F., Kuo, H., Ristic, H., Lambert, M. P., Chromy, B.,
Viola, K. L., Klein, W. L., Stine, W. B., Krafft, G. A., and Trommer, B. L.
(2002). Soluble oligomers of beta amyloid (1-42) inhibit long-term potentiation
but not long-term depression in rat dentate gyrus. Brain Res 924, 133-140.
247. Wang, J., Dickson, D. W., Trojanowski, J. Q., and Lee, V. M. (1999). The levels
of soluble versus insoluble brain Abeta distinguish Alzheimer's disease from
normal and pathologic aging. Exp Neurol 158, 328-337.

158
248. Wang, X., Herberg, F. W., Laue, M. M., Wullner, C., Hu, B., Petrasch-Parwez,
E., and Kilimann, M. W. (2000). Neurobeachin: A protein kinase A-anchoring,
beige/Chediak-higashi protein homolog implicated in neuronal membrane
traffic. J Neurosci 20, 8551-8565.
249. Washburn, M. P., Wolters, D., and Yates, J. R., 3rd (2001). Large-scale analysis
of the yeast proteome by multidimensional protein identification technology.
Nat Biotechnol 19, 242-247.
250. Washburn, M. P., and Yates, J. R., 3rd (2000). Analysis of the microbial
proteome. Curr Opin Microbiol 3, 292-297.
251. Wataya, T., Nunomura, A., Smith, M. A., Siedlak, S. L., Harris, P. L.,
Shimohama, S., Szweda, L. I., Kaminski, M. A., Avila, J., Price, D. L., et al.
(2002). High molecular weight neurofilament proteins are physiological
substrates of adduction by the lipid peroxidation product hydroxynonenal. J Biol
Chem 277, 4644-4648. Epub 2001 Dec 4643.
252. Webster, S. D., Yang, A. J., Margol, L., Garzon-Rodriguez, W., Glabe, C. G.,
and Tenner, A. J. (2000). Complement component C1q modulates the
phagocytosis of Abeta by microglia. Exp Neurol 161, 127-138.
253. Wei, Y. H., and Lee, H. C. (2002). Oxidative stress, mitochondrial DNA
mutation, and impairment of antioxidant enzymes in aging. Exp Biol Med
(Maywood) 227, 671-682.
254. Weible, M. W., 2nd, and Hendry, I. A. (2004). What is the importance of
multivesicular bodies in retrograde axonal transport in vivo? J Neurobiol 58,
230-243.
255. Weidemann, A., Konig, G., Bunke, D., Fischer, P., Salbaum, J. M., Masters, C.
L., and Beyreuther, K. (1989). Identification, biogenesis, and localization of
precursors of Alzheimer's disease A4 amyloid protein. Cell 57, 115-126.
256. Wild-Bode, C., Yamazaki, T., Capell, A., Leimer, U., Steiner, H., Ihara, Y., and
Haass, C. (1997). Intracellular generation and accumulation of amyloid beta-
peptide terminating at amino acid 42. J Biol Chem 272, 16085-16088.

159
257. Wimo, A., Jonsson, L., and Winblad, B. (2006). An estimate of the worldwide
prevalence and direct costs of dementia in 2003. Dement Geriatr Cogn Disord
21, 175-181.
258. Wischik, C. M., Crowther, R. A., Stewart, M., and Roth, M. (1985). Subunit
structure of paired helical filaments in Alzheimer's disease. J Cell Biol 100,
1905-1912.
259. Wisniewski, T., Dowjat, W. K., Permanne, B., Palha, J., Kumar, A., Gallo, G.,
and Frangione, B. (1997). Presenilin-1 is associated with Alzheimer's disease
amyloid. Am J Pathol 151, 601-610.
260. Wolters, D. A., Washburn, M. P., and Yates, J. R., 3rd (2001). An automated
multidimensional protein identification technology for shotgun proteomics. Anal
Chem 73, 5683-5690.
261. Wright, J. W., Kramar, E. A., Meighan, S. E., and Harding, J. W. (2002).
Extracellular matrix molecules, long-term potentiation, memory consolidation
and the brain angiotensin system. Peptides 23, 221-246.
262. Xu, H., Sweeney, D., Wang, R., Thinakaran, G., Lo, A. C., Sisodia, S. S.,
Greengard, P., and Gandy, S. (1997). Generation of Alzheimer beta-amyloid
protein in the trans-Golgi network in the apparent absence of vesicle formation.
Proc Natl Acad Sci U S A 94, 3748-3752.
263. Yamamoto, S., Oka, S., Inoue, M., Shimuta, M., Manabe, T., Takahashi, H.,
Miyamoto, M., Asano, M., Sakagami, J., Sudo, K., et al. (2002). Mice deficient
in nervous system-specific carbohydrate epitope HNK-1 exhibit impaired
synaptic plasticity and spatial learning. J Biol Chem 277, 27227-27231.
264. Yamamoto, S., and S., O. (2001). Glucuronyltransferases involved in the HNK-
1 carbohydrate epitope biosynthesis. Trends Glycosci Glycotech 13, 193-208.
265. Yan, S. D., Chen, X., Schmidt, A. M., Brett, J., Godman, G., Zou, Y. S., Scott,
C. W., Caputo, C., Frappier, T., Smith, M. A., and et al. (1994). Glycated tau
protein in Alzheimer disease: a mechanism for induction of oxidant stress. Proc
Natl Acad Sci U S A 91, 7787-7791.

160
266. Yan, S. D., Yan, S. F., Chen, X., Fu, J., Chen, M., Kuppusamy, P., Smith, M. A.,
Perry, G., Godman, G. C., Nawroth, P., and et, a. (1995). Non-enzymatically
glycated tau in Alzheimer's disease induces neuronal oxidant stress resulting in
cytokine gene expression and release of amyloid beta-peptide. Nat Med 1, 693-
699.

267. Yang, A. J., Chandswangbhuvana, D., Margol, L., and Glabe, C. G. (1998). Loss
of endosomal/lysosomal membrane impermeability is an early event in amyloid
Abeta1-42 pathogenesis. J Neurosci Res 52, 691-698.
268. Yang, A. J., Chandswangbhuvana, D., Shu, T., Henschen, A., and Glabe, C. G.
(1999). Intracellular accumulation of insoluble, newly synthesized abetan-42 in
amyloid precursor protein-transfected cells that have been treated with Abeta1-
42. J Biol Chem 274, 20650-20656.
269. Yang, A. J., Knauer, M., Burdick, D. A., and Glabe, C. (1995). Intracellular A
beta 1-42 aggregates stimulate the accumulation of stable, insoluble
amyloidogenic fragments of the amyloid precursor protein in transfected cells. J
Biol Chem 270, 14786-14792.
270. Yates, J. R., 3rd, Eng, J. K., McCormack, A. L., and Schieltz, D. (1995). Method
to correlate tandem mass spectra of modified peptides to amino acid sequences
in the protein database. Anal Chem 67, 1426-1436.
271. Yoshimori, T., Yamagata, F., Yamamoto, A., Mizushima, N., Kabeya, Y., Nara,
A., Miwako, I., Ohashi, M., Ohsumi, M., and Ohsumi, Y. (2000). The mouse
SKD1, a homologue of yeast Vps4p, is required for normal endosomal
trafficking and morphology in mammalian cells. Mol Biol Cell 11, 747-763.
272. Zhong, S., Storch, K. F., Lipan, O., Kao, M. C., Weitz, C. J., and Wong, W. H.
(2004). GoSurfer: a graphical interactive tool for comparative analysis of large
gene sets in Gene Ontology space. Appl Bioinformatics 3, 261-264.

Abstract (if available)

Abstract Neuritic plaques comprised of amyloid ß (Aß) are one of the primary neuropathological hallmarks of Alzheimer's disease (AD). However, Aß plaque deposition is preceded by aberrations of the endosomal/lysosomal system, including abnormally enlarged endosomal compartments, accumulation of protease-resistant proteins, and atypical activation of the lysosomal system. The functional mechanisms accounting for these abnormalities have not yet been delineated. Towards our goal of identifying the proteins whose dysfunction contributes to the development of endosomal/lysosomal pathology in AD, we have found that: 1) proteins implicated in regulating late endosomal trafficking are among the targets of oxidation (carbonylation) in the brain of a presenilin 1/amyloid precursor protein (PS1/APP) transgenic mouse model of AD

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Protein aggregation: current scenario and recent developments

PDF

Age dependent modulation of synaptic plasticity and insulin mimetic effect of lipoic acid on a 3xTg-AD mouse model of Alzheimer's disease: implications as a therapeutic/nutraceutical agent

PDF

Insulin sensitivity in cognition, Alzheimer's disease and brain aging

PDF

Protein phosphatase 2A and annexin A5: modulators of cellular functions

PDF

Macrophage-specific Stat3 deletion in a mouse model of Alzheimer’s disease

PDF

Role of CD33 structure and function in Alzheimer’s disease

PDF

Phenotypic and multi-omic characterization of novel C. elegans models of Alzheimer's disease

PDF

The role of specific gut microbiota in regulating the development of Alzheimer抯 disease

PDF

Dihydromyricetin as a potential therapeutic for Alzheimer's disease-like changes induced by social isolation

PDF

Nebula/DSCR1 upregulation preserves axonal transport and memory function in a Drosophila model for Alzheimer's disease

PDF

Exploring the relationship between menopausal hot flushes and Alzheimer's disease biomarkers: a cross-sectional analysis in postmenopausal women

PDF

The role of the locus coeruleus in Alzheimer’s disease and cerebrovascular function: insights from neuroimaging, neuropsychology, and biofluid markers

Asset Metadata

Creator Thomas, Stefani Nicole Cottrell (author)

Core Title Functional proteomic analysis of altered protein signaling modules in Alzheimer's disease

School School of Pharmacy

Degree Doctor of Philosophy

Degree Program Pharmaceutical Sciences

Publication Date 03/28/2007

Defense Date 03/07/2007

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Alzheimer's disease,amyloid beta,HNK-1/NCAM,LC-MS/MS,OAI-PMH Harvest,proteomics,Vps4b

Language English

Advisor Shen, Wei-Chiang (committee chair), Ann, David K. (committee member), Okamoto, Curtis Toshio (committee member), Schauwecker, P. Elyse (committee member), Yang, Austin (committee member)

Creator Email stefanit@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m333

Unique identifier UC193858

Identifier etd-Thomas-20070328 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-329545 (legacy record id),usctheses-m333 (legacy record id)

Legacy Identifier etd-Thomas-20070328.pdf

Dmrecord 329545

Document Type Dissertation

Rights Thomas, Stefani Nicole Cottrell

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu