University of Southern California Dissertations and Theses

An overview of Alzheimer’s disease and the potential for APP-beta secretase-interaction as a therapeutic target

(USC Thesis Other)

An overview of Alzheimer’s disease and the potential for APP-beta secretase-interaction as a therapeutic target

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Copyright 2018 Evan Feeley
1
An Overview of Alzheimer’s Disease and the Potential for APP-Beta Secretase-Interaction
as a Therapeutic Target
By Evan Feeley
A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
requirements for the degree
MASTER OF SCIENCE
(BIOCHEMISTRY AND MOLECULAR BIOLOGY)
August 2018

Copyright 2018 Evan Feeley
2

Acknowledgement
I would like to thank Dr. Sita Reddy, Dr. Lucio Comai and the rest of the Reddy lab for
my experience at USC. I am happy to have come away from this experience with the knowledge
of the techniques that I have learned at the bench.
I would like to thank Dr. Amy Merrill and Dr. Young Hong for their support and
guidance as I prepared my original research at the bench.
I would also like to thank my committee chair, Dr. Zoltan Tokes for guiding me to the
topic of Alzheimer’s Disease, which I have enjoyed writing and have learned so much about.
And, thank you to the rest of my committee, Dr. Baruch Frenkel and Dr. Joseph Hacia, for
graciously helping to guide me through the process of researching and writing this thesis.
I would like to thank Dr. Judd Rice and Dr. Pragna Patel for helping me transition away
from the bench and into this written thesis.
And, finally, I would like to thank Monica Pan for helping coordinate the progression of
my degree and the submission of this document.

Copyright 2018 Evan Feeley
3

Table of Contents
1. Abstract…………………………………………………………………………….………5
2. Introduction and Epidemiology……………………………………………….………….6
3. The Amyloid Hypothesis and APP Metabolism…………………………………………7
4. Tau Pathology…………………………………………………………………………….12
5. Alzheimer’s Disease Genetic Risk Factors……………………………………………...16
6. Biomarkers and Diagnosis……………………………………………………………….22
7. Ageing, Sirtuins, and Alzheimer’s Disease……………………………………………...23
8. A History of Research……………………………………………………………………25
9. Clinical Trials…………………………………………………………………………….36
10. Discussion…………………………………………………………………………………41
11. Conclusion………………………………………………………………………………...48
12. References………………………………………………………………………………...49

Copyright 2018 Evan Feeley
4

Abbreviations
AD: Alzheimer’s Disease
EOAD: Early Onset Alzheimer’s Disease
LOAD: Late onset Alzheimer’s Disease
APP: amyloid precursor protein
NFT: neurofibrillary tangles
PHF: paired helical filaments
AICD: APP intracellular domain
PS1: presenilin 1
PS2: presenilin 2
ApoE: Apolipoprotein E
p-tau: phosphorylated tau
CSF: cerebrospinal fluid
AGE: advanced glycation end products
RAGE: receptors for advanced glycation end products
PRS: polygenic risk score
BDNF: Brain-derived neurotrophic factor
Sirt: sirtuin
NSC: neural stem cell
NMN: Nicotinamide mononucleotide

Copyright 2018 Evan Feeley
5

An Overview of Alzheimer’s Disease and the Potential for APP-Beta Secretase-Interaction
as a Therapeutic Target
Evan Feeley
Department of Biochemistry and Molecular Biology, University of Southern California,
CA 90033
Abstract
Since Alzheimer’s disease was first described by Alois Alzheimer in 1906, little progress
has been made in the development of therapeutics. Early onset Alzheimer’s disease has clear
heritable dominant mutations that lead to pathology, but sporadic or late onset Alzheimer’s
disease, like obesity or heart disease, is not as simple. Late onset Alzheimer’s is highly
multifactorial, influenced by a combination of many genetic and environmental factors as a
person ages. Here, I attempt to give an overview of the disease genetics, biomarkers, and the
two hallmarks of Alzheimer’s, amyloid plaques and neurofibrillary tangles. The two leading
hypotheses, the amyloid cascade hypothesis and the tau hypothesis must be intimately
interlinked in order to cause pathology. Mouse models do not effectively represent Alzheimer’s
disease in humans and new methods to target amyloid precursor protein metabolism need to be
developed. From my broad investigation into Alzheimer’s disease, I suggest that the interaction
between the amyloid precursor protein and beta secretase be a key focus for the development of
future therapeutics, but beta secretase must not be inhibited.

KEY WORDS: Alzheimer’s Disease, APP metabolism, Tau pathology, Genetics, Beta secretase

Copyright 2018 Evan Feeley
6

Introduction and Epidemiology
Every 66 seconds someone in the United States develops Alzheimer’s disease (AD)
dementia. As an age-related disease, the impact of AD on society becomes greater the more
people in a population reach the age of 65. In the United States (US), the first generation of baby
boomers turned 70 in 2016. The number of individuals in the US over the age of 65 will continue
to increase dramatically and is expected to nearly double from 48 million to 88 million by 2050.
In 2017, about 5.5 million Americans were estimated to be living with AD dementia. 5.3 million
of these diagnoses are sporadic late onset AD (LOAD) and about 200,000 were early onset AD
(EOAD). Women are predominantly diagnosed with AD compared to men. About two thirds of
all diagnosed are women. However, when proportions of men and women with AD are compared
to populations of people 65 and older, the proportions are not statistically significantly different.
Even though these estimates seem staggering AD is underdiagnosed and underreported.
AD is the sixth leading cause of death in the US and years of morbidity before AD takes
a person’s life are substantial. People 65 and older survive, on average, four to eight years after
their diagnosis. It is difficult to determine how many people exactly have died from AD, but
based on reports from 2014, about 93,541 people died in the US that year. Difficulties in
determining this number reflect that AD can cause acute illness that leads to death, which is
reported on the death certificate in place of AD. Therefore, deaths from AD are also
underreported.
The burden on society due to AD is massive because of the extent of care that AD
patients require. Most patients are expected to be in nursing homes. The cost of healthcare and
long-term care for AD dementia is one of the costliest to society. Total payments in 2017 were
estimated to be 259 billion US dollars, with medicare and medicaid covering about 175 billion

Copyright 2018 Evan Feeley
7

[3]. Given that the burden of AD is so great it has been a heavy focus in research since the 1980s.
However, the mechanism of AD is still not fully understood and there are no AD pathology
altering therapeutics.
The Amyloid Hypothesis and APP Metabolism
The dogma of AD for over the past 25 years is the “Amyloid Cascade Hypothesis”. In a
nutshell, the amyloid precursor protein (APP) is cleaved into various metabolites by enzymes
called secretases and then these pieces are exported from the cell where they are normally
degraded or removed. In AD, there is an inability to degrade the APP metabolites effectively.
Amyloid beta then accumulates into plaques on the brain. The plaques are believed to cause
neurotoxicity and induce tau pathology that leads to neurofibrillary tangles (NFT) which cause
neurodegeneration and eventually cell death.
APP, first identified in 1987, is part of a highly conserved protein family that resembles
cell surface receptors and is encoded by a gene on chromosome 21. The transmembrane APP
protein has three variants: 695, 751, and 770- each proposed to have a slightly different effect on
the brain. In the AD brain, APP770 has been shown to be increased and APP695 reduced.
Structurally, eighty-eight percent of the total mass of the APP protein is in the N terminus
extracellular domain. While, the cytoplasmic C terminus has forty-seven amino acids, is the most
conserved region, and mainly controls the function of the protein.

Copyright 2018 Evan Feeley
8

The function of APP is not
completely known, but research suggests
that APP mediates cell adhesion and
migration at a point in early phases of
development and during the formation of
synaptic junctions. In addition, knockout
experiments in mice demonstrate that
APP has an important role in nervous
system development. In the brain, it has
been suggested to be involved in
synaptic plasticity and maturation, and
brain network function. APP may also
be important in synaptogenesis, dendritic
spine formation, synaptic vesicle and
transmitter release, and behavior [1].
APP travels down the axons and
dendrites to the terminals where, alpha,
beta, and gamma secretases
constitutively cleave the protein and
then amyloid beta is released into the
synapse. Amyloid beta is 39 to 43
amino acids long. The most common peptide lengths are either 40 or 42 amino acids. Amyloid
Figure [1]: Flowchart describing the events of the amyloid
cascade hypothesis. Genetic findings that mutations in APP,
PS1, and PS2 led to increased confidence in the hypothesis
and research on this mechanism [28}.

Copyright 2018 Evan Feeley
9

beta 42 is considered to be the most toxic and easy to aggregate. It is not certain, but if amyloid
aggregation is toxic it is most likely that the aggregation causes damage through synaptotoxicity
and might affect the function of neighboring cells like brain glia and endothelial cells. Amyloid
beta can bind APP creating a positive feedback loop where amyloid beta production can even
cause dysfunction of APP.
The endoproteolytic processing happens through two major pathways, non-
amyloidogenic and amyloidogenic. The non-amyloidogenic pathway is mediated by alpha
secretases that are mostly found in the postsynaptic region of the excitatory synapse. It cleaves
the region inside amyloid beta and hinders its release. This pathway generates soluble APP
fragment alpha and the membrane bound, alpha C terminal fragment. The C terminal fragment is
cut by gamma secretase and generates the APP intracellular domain (AICD) and p3 peptide that
is rapidly degraded.
The amyloidogenic pathway is mediated by beta secretase, BACE1. Beta secretase
cleaves the ectodomain and generates soluble APP beta and the membrane bound APP beta C
terminal fragment. The C terminal fragment is then cleaved by gamma secretase in the lipid
bilayer and produces amyloid beta and AICD. Beta secretase has multiple cellular functions and
knocking it out has been shown to be hazardous. Reduction of beta secretase, however, can be
beneficial to AD patients [1].
The formation of ecto-fragments, soluble APP beta or soluble APP alpha, and the APP
intracellular domain (AICD) depends upon the secretase that processes the extracellular APP
domain. If the beta secretase has access to APP and cleaves it, the process is known as
amyloidogenic. If APP is cleaved by the alpha secretase, the process in known as non-

Copyright 2018 Evan Feeley
10

amyloidogenic. It is possible that AD may be a result of an imbalance of alpha and beta secretase
cleavage.
All of the APP metabolites have supposed physiological functions in the brain. Both
soluble APP alpha and soluble APP beta fragments have protective effects (soluble APP beta is
less protective) against AD and are involved in neurite outgrowth. Soluble APP alpha, however,
Figure [2]: Schematic of both the amyloidogenic and non-amyloidogenic pathways of APP metabolism. The
amyloidogenic pathway, initiated by beta secretase, results in the C terminal fragment that eventually is
cleaved by gamma secretase to yield amyloid beta.
[38]

Copyright 2018 Evan Feeley
11

also helps in long term potentiation unlike soluble APP beta. Amyloid beta monomers are
protective similarly to soluble APP alpha and are involved in neural progenitor cell proliferation
and synaptic transmission. The AICD is released inside of the cell and acts as a transcription
factor for many genes including APP, neprilysin, which degrades amyloid, and aquaporin 1.
APP695, a variant of APP, has been shown to epigenetically regulate the gene, transthyretin,
which binds and clears amyloid beta peptides.
It is important to note that both pathways are important for normal brain function and that
APP metabolism is largely dependent on the colocalization of APP with secretases. The
trafficking of APP may be altered in AD and have a profound effect on metabolism. Deficiencies
in proteins like AP2, a clathrin mediated endocytosis adapter protein, important for trafficking
secretases, could contribute to disease. Beta secretase cleavage occurs preferentially in the early
or late endosomes or lysosomes and APP and beta secretase follow similar trafficking routes.
APP can reciprocally regulate trafficking of gamma secretase components, such as presenilin 1
(PS1) and presenilin 2 (PS2) [1].
APP trafficking can be regulated by APP phosphorylation and LDL receptor family
proteins that also act as receptors for apolipoprotein E (ApoE) and PS1. In GWAS, genome wide
association studies, the variants associated with AD are often in proteins that play a role in APP,
cholesterol, and lipid metabolism [1]. APP is believed to be associated with other processes
important for synaptic function such as cholesterol turnover, biosynthesis, and hydroxylation.
Cell membrane lipid composition may also be important in APP metabolism. The amyloidogenic
pathway occurs in cholesterol rich regions such as lipid rafts and depletion of cholesterol
prevents amyloidogenic processing. Statins, a drug that will reduce cholesterol, remain

Copyright 2018 Evan Feeley
12

controversial as there are discrepancies in benefits to AD patients. But, other drugs targeting
cholesterol promote the non-amyloidogenic pathway [1].
The amyloid cascade hypothesis will remain relevant considering that amyloid plaques
remain a hallmark of the disease. But, the focus will shift to other aspects of APP metabolism,
considering APP is processed into many different metabolites. One product, the C terminal
fragment, is cleaved by gamma secretase and its accumulation is closely associated with AD
pathology. APP C terminal fragments build up in dystrophic neurites in AD brains and impair
vesicular trafficking. In mouse models APP C terminal fragments have been linked to synaptic
failure and memory impairment and overexpression of these fragments in mice caused them to
develop AD like symptoms including phosphorylated Tau. APP C terminal fragments also
attenuate the cAMP/PKA/CREB pathway by causing the breakdown of cAMP [18].
Tau Pathology
It is important to point out that amyloid beta and amyloid plaques are not the only
hallmark of AD. From the beginning, in post mortem autopsies, it has been observed that
amyloid plaques can be extensive in people who did not experience cognitive impairment and
aged individuals without AD can have large amounts of plaques. Therefore, amyloid plaques
may be a result of normal aging. The reverse is also true, where patients with AD barely had any
plaques at all. Mouse models have shown that even though amyloid beta can cause cognitive
defects it does not cause neurodegeneration.
Multiple mouse models manipulated to overexpress amyloid beta have shown amyloid
plaques but did not show NFT, the other hallmark of AD, and cell death was not observed. This
suggests that amyloid beta does not induce tau pathology and amyloid beta on its own does not
cause AD. Mice that overexpressed amyloid beta 42 had plaques, but the degeneration of nerve

Copyright 2018 Evan Feeley
13

cells, neuronal loss, and cognitive impairment was not observed. Immunotherapies used to treat
mouse models of AD have been shown to effectively clear AB, but these therapies did not
relieve the symptoms of AD [18].
Recent focus has shifted to the tau hypothesis and tau pathology is believed to be more
directly correlated with AD. Tau is highly present in the axon, the somatodendrite, the nucleus
and dendrites. The tau protein is unfolded and highly soluble. There are six isoforms of tau that
vary in size and they are differently expressed during brain development. Tau plays an important
role in tubulin assembly and microtubule stability, and thus supports the normal function of
neurons. It not only stabilizes microtubules, but soluble tau also regulates APP trafficking.
In AD, tau is hyperphosphorylated and this hyperphosphorylation changes the
conformation of tau from a monomer to a paired helical filament (PHF) which leads to the
formation of NFT. The NFTs themselves may not be toxic but the tau oligomer, the most toxic
form, causes synaptic impairment [6]. Abnormal tau can convert normal tau to a disease form
and sometimes can incorrectly bind to filamentous actin to induce its destabilization, causing
synaptic impairment and defects in the integrity of mitochondria [18].

Copyright 2018 Evan Feeley
14

Tau directly helps regulate cargo up and down the microtubule. When there are
deficiencies in tau, microtubules breakdown. This means that the transport of cargo up and down
the microtubules between the cell bodies and the axons is impaired. Tau oligomers can be
transmitted from one region of the AD brain to another. In drosophila, overexpression of tau
results in neuronal loss before NFTs form. The underlying mechanism of how tau is toxic is
Figure [3]: Tree diagram flow chart showing how post translation modifications can lead to the formation of tau
oligomers that then assemble into paired helical filaments. Paired helical filaments can be broken down and
formation blocked by methylene blue, fulvic acid, and potentially kinase inhibitors. However, if left untreated,
paired helical filaments will form neurofibrillary tangles that are believed so cause pathology in neurons. [39]

Copyright 2018 Evan Feeley
15

unknown and the question of, why tau is hyperphosphorylated in the first place, still persists.
[18].
Tau pathology comes in stages and spreading of tau pathology is strongly correlated with
AD symptoms. It spreads in a predicted pattern. First tau appears in the transentorhinal region
(stages 1 and 2). Then it appears in the limbic region (stages 3 and 4). And, finally, it spreads to
the neocortical areas (stages 5 and 6). Tauopathy can be seen before amyloid beta even begins to
accumulate and is present in other neurodegenerative diseases such as: frontotemporal dementia,
Parkinson’s disease, progressive supranuclear palsy, corticobasal degeneration, argyrophilic
grain disease, tangle only dementia, and chronic traumatic encephalopathies [18].
There are several post translational modifications of tau that are associated with AD. The
most significant is hyperphosphorylation. Normal phosphorylation is necessary for tau to bind to
microtubules and perform normal function. The actual causes that lead to hyperphosphorylation
are still under debate. Tau phosphorylation at proline rich regions blocked microtubule assembly
and promoted self-aggregation. Increased levels of tau phosphorylated at tyrosines also played a
role in the neurodegeneration process.
Tau can also be truncated leading to increased tau aggregation. In the AD brain it was
found that caspase 6 cleaved the N terminus of tau to truncate it. Caspase 3 will cleave tau at
residue aspartic acid 421. It also can be cleaved at glutamic acid 39. This truncated form
aggregated more readily than full length tau.
Glycosylation can affect tau function. N-glycosylation stabilizes the structure of PHF,
facilitates kinases to phosphorylate tau, and suppresses the action of phosphatases. Non
hyperphosphorylated tau was glycosylated in AD brain but not in normal control brains and
glycosylation precedes hyperphosphorylation. O-glycosylation, however, helps regulate tau

Copyright 2018 Evan Feeley
16

phosphorylation in a site-specific manner. Treating AD mouse models with an insulin sensitizer,
such as rosiglitazone, can restore O-glycosylation of tau to prevent hyperphosphorylation. This
attenuated learning and memory deficits, but this treatment was not recommended to treat AD
patients.
Tau can be subjected to glycation, a non-enzymatic post translational modification.
Together they form advanced glycation end-products (AGE). Methylglyoxal can induce tau
hyperphosporylation through AGE formation, upregulation of receptors of advanced glycation
end-products (RAGE), and activation of GSK-3B and p38 MAPK. Synaptic dysfunction in AD
is associated with the p38alpha MAPK pathway.
Other post translational modifications involved in tau aggregation are lysine methylation,
sumoylation, ubiquitination, and nitration. Lysine methylation makes tau more likely to promote
microtubule formation which weakens its ability to aggregate. Sumoylation induces
hyperphosphorylation which inhibits ubiquitination. Interaction of the carboxy terminus of the
HSC70-interacting protein, CHIP, with HSP70/90 induces ubiquitination of tau and causes tau to
aggregate. Abnormal nitration of tau was also found in the NFT of AD brains, but the role of tau
nitration remains unclear. Nitrated tau undergoes conformational changes that keep it from
binding to microtubules [6].
Alzheimer’s Disease Genetic Risk Factors
Familial Alzheimer’s Disease is caused by known heritable dominant mutations in either
the exons 16 and 7 of the amyloid precursor protein (APP), presenilin 1 (PS1), or presenilin 2
(PS2). These mutations are linked to EOAD (before age 65) and are correlated with an increase
in amyloid beta production [20][5]. PS1 and PS2 are both a part of the gamma secretase complex
which cleaves APP secondary to alpha and beta secretase [1].

Copyright 2018 Evan Feeley
17

Regarding sporadic LOAD, the ApoE gene is the most highly correlated genetic risk
factor. There are a few variants of the ApoE gene with ApoE4 being the Alzheimer’s Disease
risk variant, ApoE3 being neutral, and ApoE2 offering a protective effect [5]. ApoE4 is
normally made and released by astrocytes and microglia. It binds to high density lipoproteins to
move cholesterol and phospholipids to LDL receptors. ApoE4 transgenic mice do not develop
AD but they show reduced amyloid beta clearance, aberrant cholesterol movement, and
disruptions in the blood brain barrier. It is suggested to cause the activation of proinflammatory
cyclophilin A- matrix metalloprotease 9 pathway that leads to the breakdown of the blood brain
barrier and neurodegeneration. As such, the blood brain barrier needs to be considered in future
mouse models [24]. The odds ratio of developing Alzheimer’s Disease increases in ApoE4
heterozygotes and then increases again four-fold in ApoE4 homozygotes [20]. The more copies
of the ApoE4 allele an AD patient has the lower the age of onset of the disease. ApoE4 carriers
show the most reduced ability to clear amyloid beta [2].
ApoE is secreted by astrocytes in the central nervous system where it binds to microglia
and is taken up into nerve cells by an ApoE receptor during development or repair periods after
damage. ApoE affects the clearance of amyloid beta and affects inflammatory receptor signaling.
In mouse models ApoE is involved in tau regulated neurodegeneration and inflammation and
having ApoE4 causes more damage while knockout of ApoE was protective against AD [18].
The R47H allele of the TREM2 gene is also closely associated with AD. Additionally, it
has been associated with Parkinson’s, frontotemporal dementia, and amyotrophic lateral
sclerosis. APP/PS1 transgenic mice with one copy of TREM2 have altered microglial responses
but do not show an increase in the amounts of amyloid plaque [24]. TREM2 recognizes ApoE
and is expressed in microglia. It is important in injury response and helps allow the microglial

Copyright 2018 Evan Feeley
18

response to tau damage in the brain protecting against inflammation and brain atrophy. When the
TREM2 and ApoE pathway is activated, the microglia lose the ability to regulate brain
homeostasis [18]. TREM2 deficient APP/PS1 transgenic mice show a prevention of blood
derived myeloid cells from entering the brain and reduced burden of amyloid plaques. TREM2 is
somehow altering microglia responses but there is no mouse model with human R47H TREM2
yet to test this [24].
ApoE and TREM2 however may not be the only genes that can be used to evaluate the
risk of developing AD. Apostolova, L.G. et al looked at the top 20 Alzheimer’s Disease risk
variants associated with brain amyloidosis. A GWAS was run that revealed 20 AD amyloidosis
genetic risk loci. These 20 well- established Alzheimer’s Disease risk genes were identified and
validated in the largest Alzheimer Disease GWAS to date [2].
ABCA7 has one of the strongest association with amyloid deposition. The ABC protein
family is responsible for transport of molecules, primarily lipids, across cell membranes. Loss of
function of ABCA7 is associated with increased beta secretase cleavage of APP. Out of fifteen
ABCA7 variants, three variants were associated with brain amyloidosis but not brain atrophy and
one variant of ABCA7 is a rare missense mutation that resulted in significant protection against
Alzheimer’s Disease.

Copyright 2018 Evan Feeley
19

FERMT2 codes for a 680 amino acid protein that is involved in the scaffolding of
extracellular matrices important for cell adhesion. There may be a stage dependent association of
FERMT2 with amyloidosis. It is a coactivator of B3-integrin, a microglial and reactive astrocyte
marker that helps in post stroke brain recovery and has been shown to modify tau neurotoxicity
in Drosophila. FERMT2 is upregulated in atherosclerotic plaques suggesting that it maintains a
possible role in inflammation and leukocyte leakage.
CLU, an extracellular chaperone protein of 427 amino acids, encodes for a clusterin. It is
highly expressed in neurons and ependymal cells and involved in synaptic maintenance and
apoptosis. It has been shown to associate with reduced aggregation and promotes clearance of
amyloid beta.
Figure [4]: Graph showing the correlation of genes and their impact on AD risk vs the allele frequency. Certain
dominant genetic mutation can result in familial or early onset Alzheimer’s disease, while a cumulation of small
common risk factors can lead to sporadic or late onset Alzheimer’s disease. [20]

Copyright 2018 Evan Feeley
20

DSG2 is a cell adhesion desmosome cadherin protein that binds to plaque proteins and
intermediate filaments, possibly playing a role in inflammation. However, the mechanistic
explanation and relation to Alzheimer’s disease has yet to be explained. It is possible that it has a
late modulatory effect on amyloid deposition and it is abundantly present in the corpus callosum.
EPHA1 encodes a 976 amino acid protein that is a receptor tyrosine kinase. It is
responsible for contact dependent signaling in the nervous system during development. It is
highly expressed in the cerebral cortex and the hippocampus and has been observed to have a
negative effect on brain amyloidosis.
PICALM codes for a 652 amino acid protein that binds to the heavy chain of clathrin and
assists in vehicle assembly for endocytosis. It colocalizes with APP and knockdown of PICALM
resulted in the reduction of APP internalized and a reduction in amyloid beta generation. It was
found to modulate the clearance of tau and thus must play a role in autophagy. There is a
negative association between PICALM and the prefrontal brain volume and working memory, as
well as, hippocampal, amygdalar and white matter lesion volume and entorhinal,
parahippocampal, and temporal pole cortical thickness.
SORL1 encodes a large protein of 2,186 amino acids and is in the LDL receptor family. It
readily binds to ApoE and lipoprotein lipase and is likely also involved in mediating endocytosis,
specifically, APP trafficking and recycling. It is downregulated in lymphoblasts and cortical
pyramidal neurons in patients with Alzheimer’s Disease. It could be used to monitor cognitive
decline and the conversion of mild cognitive impairment to Alzheimer’s Disease. The levels of
this protein correlate with the soluble APP products produced from beta secretase cleavage.

Copyright 2018 Evan Feeley
21

ZCWPW1 codes for a 648 amino acid protein and is a possible risk gene for late onset
Alzheimer’s Disease. It is involved in epigenetic regulation of gene expression [2]. Cluster of
Differentiation 33 (CD33) was also shown to be a strong genetic locus associated with AD [8].
Given the identified possible risk variants it is reasonable to assume that Alzheimer’s
Disease is complex, and many alleles may contribute to pathogenesis. Chaudhury, S. et al looked
at the ability to determine a polygenic risk score (PRS) based on genes that may have roles in
Alzheimer’s Disease pathogenesis. TRIP4, SPPL2A, and ABI3 were identified as additional
Alzheimer’s Disease risk genes [5] and R47H and TREM2 were identified as risky rare variants
[8].
AD genetic risk scores were calculated by multiplying each individual GWAS allele
effect size using the beta coefficients from a previous data set. The idea is to collect many
polymorphisms of small effect size to give a whole polygenic risk score. Using this combination
of genetic risk factors the model was successfully able to distinguish between controls and
Alzheimer’s Disease patients with 75.5% accuracy and there was a significantly higher average
PRS for AD patients than controls. The ApoE locus alone has a predictive value of 65.2%.
Combined, if the risk prediction from ApoE2 and ApoE4 and sex is used with the predictive
value of the other SNPs, the predictive value is raised another 3% to 78.5% predictive value [8].
Epigenetics may also play a crucial role in AD as it relates to aging. Epigenetic changes
and histone markers related to normal aging can be traced and histone modifications are
suggested to be dysregulated in AD. Acetylation is often a histone marker related to the
activation of genes. In mouse models of AD, histone acetylation is reduced in areas pertaining to
memory and neuronal genes and these acetylation marks have been observed to be associated
with age related memory deficits and learning. Upon treatment of these mouse models with

Copyright 2018 Evan Feeley
22

nonspecific deacetylase inhibitors the synaptic and cognitive plasticity of these mice was
restored.
In model organisms such as C. elegans and yeast, H4K16ac plays a role in whole
organism aging and cell senescence. H4K16ac regulates chromatin compaction, gene expression,
stress responses, and DNA damage repair. Across the epigenome of a normal brain there is a
gain of H4K16ac as the brain ages. However, in the aging brain of AD patients there is a bias
towards a loss of H4K16ac. Changes in H4K16ac during normal aging are negatively correlated
with changes in AD and the changes associated with age in AD correlated with nearby gene
expression. It stands to reason that normal aspects of aging either fail to happen or are
dysregulated in the AD brain. Specifically, there is a similar age-related dysregulation pattern
with a transcriptional corepressor, REST, that normally increases with age but decreases in AD.
However, a genome wide assessment of the effects of REST has not yet been performed in the
human brain [23].
Biomarkers and Diagnosis
Genetics can be used to assess the risk of Alzheimer’s disease but upon onset of cognitive
impairment Alzheimer’s Disease must be properly diagnosed. The markers, tau and amyloid beta
42 are the most tested markers and best represent neurofibrillary tangles and amyloid plaques.
These markers are evaluated in the blood and the cerebrospinal fluid (CSF). When neurons
degenerate they release tau, increasing the levels of tau in the blood. Amyloid beta 42, Tau, and
phosphorylated Tau (p-Tau), are also used to reflect the progression of Alzheimer’s Disease.
Certain mild cognitive impairment patients who converted to AD showed that total Tau and p-
Tau were elevated but that amyloid beta was decreased in CSF. Additional ways of diagnosing

Copyright 2018 Evan Feeley
23

Alzheimer’s might come in the form of using microRNAs (miR). Cell free MiR-125b reduced in
serum of AD patients was associated with AD patients with 82% accuracy [8].
Other potential biomarkers that may be useful for Alzheimer’s Disease diagnosis are
proteins involved in neuroinflammation. Chitinase-3-like protein1, YKL-40, IL6, IL8, TRAIL,
and Visinin-like protein-1 (VILIP1) play important roles in neuroinflammation and their CSF
values could contribute to differentiate symptomatic AD patients from controls.
Peripheral biomarkers could also be beneficial in diagnosing AD before serious
symptoms manifest. Abnormal mitochondrial respiration, altered nucleotide pools, and decreased
DNA repair activity in peripheral blood mononuclear cells may all be signs of AD. Other
peripheral markers could be those involved in cell cycle deregulation in peripheral lymphocytes
of AD patients. Cytokines like CDK4, CDK6, cyclin B, and cyclin D are all highly expressed in
AD patients. The classical biomarkers in combination with peripheral ones can diagnose AD
with 80-85% accuracy and have proven useful. The challenge however remains in diagnosing in
the preclinical stages. [8].
Ageing, Sirtuins, and Alzheimer's
Ageing is the greatest risk factor for neurodegenerative diseases. As a person ages their
brain functions such as cognition, emotion, circadian rhythm, and autonomic function decline.
Even if cognitive function is normal, glucose metabolism in people over the age of 70 becomes
abnormal and strongly correlates with depression and anxiety. There is increasing evidence that
the microglia play a major role in the ageing process of the brain. These microglia are tiny brain
immune cells that induce inflammation and if they are constitutively activated or primed, even at
low levels, this can cause chronic damage to the brain. Recent studies suggest that sirtuins

Copyright 2018 Evan Feeley
24

participate in the process of microglial cytokine production and that sirtuin 1 (Sirt1) mediates
interactions between neurons and the microglia.
Sirtuins are a conserved family of NAD+ dependent deacetylases that help control the
ageing process in several model organisms from yeast to mice. They interact with other
pathways known to be involved in ageing such as the FOXO pathway and the mTOR pathway.
When the brain ages, brain atrophy occurs most extensively in the prefrontal cortex and the
hippocampus and plaques and NFT continue to form. Sirt1 is known to be protective against the
symptoms of AD and increase cognition in mouse models. It promotes neurite outgrowth, axon
development, dendritic branching, long term potentiation, learning, and memory. As the brain
ages there is also an observable decrease in neurogenesis and sirt1 and sirt2 have been connected
to the regulation of neural stem and progenitor cells.
Sirt1 can also attenuate cognitive decline associated with age by regulating synaptic
plasticity and adult neurogenesis in the hippocampus. Knockout experiments of sirt1 in mice are
conflicting. Some showed impaired memory associated with long term potentiation in the
hippocampus [26]. Others, even with very large hybridization signals for sirt1 in the
hippocampus, showed that genetic deletion of sirt1 in mice did not result in abnormalities in the
anatomy of the brain. Some sirt1 knockout experiments did reduce dendritic branching and
density and led to reduced production of neurotrophic factors [36]. In the hippocampus sirt1
positively regulates synaptic plasticity by interacting with a repressor that contains a
transcription factor, YY1, that represses a specific microRNA, miR-134. MiR-134 is a brain
specific microRNA that down regulates cAMP-responsive element-binding protein (CREB) and
brain-derived neurotrophic factor (BDNF) expression. CREB and BDNF are involved in synapse
formation and long-term potentiation.

Copyright 2018 Evan Feeley
25

Sirt1 also plays a role in the neural stem cell (NSC) pool. Nicotinamide mononucleotide
(NMN), a NAD+ intermediate that enhances sirtuin activity, significantly increased the NSC
pool in the dentate gyrus when given to old mice over a long-term period. Mice that over
expressed brain specific sirt1 had extended life spans and their skeletal muscles maintained a
young morphology. Administration of NMN attenuates cognitive deterioration in AD mouse
models but perhaps does not completely halt deterioration [26].
In the human AD brain, the levels of sirt1 are reduced but no significant differences were
recorded between the control and AD groups in post mortem brain analyses. Other sirt1
overexpression experiments in mice in the brain have shown increased life spans and rescued
ApoE4 mediated reduction of soluble APP alpha secretion. Sirt1 has been observed to interact
with both tau and APP, is involved in the CREB pathway, and sirt1 promotes healthy APP
processing. Resveratrol, which stimulates sirtuin activity, was shown to have protective effects
against neurodegeneration. Sirt1 has been seen to enhance alpha secretase activity and decrease
the amount of amyloid beta. Sirt1 can also suppress inflammation and is able to reduce oxidative
stress in astrocytes via transcriptional activity upregulating anti-oxidative enzymes and
repressing the expression of senescence associated secretory phenotype genes, deacetylating
histones at their promoter regions. [36].
A History of Research
AD was first described by Alois Alzheimer in 1906. He presented his findings, supported
by histology, from a case-study of a 51-year-old woman who had early onset dementia and died
four years after the onset of symptoms [33]. Silver and Nissl staining of the brain resulted in
distinct patterns of plaques and tangles in the one brain [11]. He found that the woman’s brain
was largely atrophic, had neurofibrillary deposits, and that some areas of the cerebrum were

Copyright 2018 Evan Feeley
26

resistant to staining [33]. Plaques were described before Alzheimer in elderly patients with
epilepsy, but Alzheimer was the first to describe NFT [11]. His findings were published in 1907,
but the disease was not named for him until after his death in 1916 by a colleague [33]. From the
first description of AD until the 1960s the focus on AD research was to categorize and cement
the clinical pathology of the disease. For the first six decades this research was supported by case
studies of the symptoms of patients followed by staining and microscope work on the brains of
these patients post mortem.
The earliest case studies repeatedly were able to find rigid or sclerotic plaques in different
areas of the brain. In 1913, a case study under the disease name, Presbyophrenia, later identified
as AD, followed a 68-year-old woman. Microscopic examination of the brain post mortem
revealed that there were large amounts of plaques specifically in the cerebral cortex [30]. These
Figure [5]: Rough timeline of events and themes in Alzheimer’s disease research since it’s first description in
1906.

Copyright 2018 Evan Feeley
27

case studies continued sporadically up and throughout the 1930s and by the mid-1930s there
were about 90 well described cases of AD. These case studies became more detailed with the
invention of the electron microscope in 1931, allowing for a more in-depth analysis of the brain.
In 1936, a case study of a 53-year-old woman, was concerned about making a distinction
between EOAD and senility. Prior to her death she needed to be kept in a mental institution
where she received mental examinations and close record of descriptions were kept of her well-
being. The only treatment she was given were sedatives. The post-mortem brain was weighed
and analyzed. After fixation, it could be seen that there was severe brain atrophy and had
characteristic plaques, which were subjected to deeper microscopic analysis. It was noticed that
nerve cells were rare in the cortex and neurofibrils appeared to be broken up. Also, the number
of astrocytes appeared to increase in specific regions [13]. By 1939, as seen in a case study of a
53-year-old woman, accompanied by physical tests of well-being, electroencephalography tests
were used to determine patterns in the brain. But, even still, at this stage, diagnosis was tricky
and ultimately was a process of elimination [21].
Throughout the 1940s and 50s there was an effort to compare AD and Pick’s disease,
another familial disease which also characteristically resulted in severe atrophy of the brain. The
debate in the 1940s largely lumped the two diseases together but became more controversial
throughout the 1950s. In 1941, in a case study of another 53-year-old woman with severe
memory loss, symptoms of intellectual deterioration appeared around the age of 41. She died in a
psych ward at 59. A common theme among these case studies are that patients are almost always
confined to psych wards, macroscopic observations were dutifully taken, and negative
Wassermann tests for syphilis were always noted. Nissl staining was performed to look at cell
loss and silver staining revealed enormous amounts of senile plaques. Myelin staining of this

Copyright 2018 Evan Feeley
28

patient’s brain suggested that there was significant degeneration and there was intense gliosis,
especially in areas around the plaques, which was noted to be typical of the disease at this time
[15].
The debate to fully characterize, categorize, and identify AD continued throughout the
1960s. There was a concerted effort at this time to determine if there was a difference between
AD and senile dementia. Which, looking at it now, really is identifying the difference between
EOAD, LOAD, and other forms of senile dementia. Woodard, J.S. using data from 1,000
necropsy cases from three state mental hospitals, proposed in 1966 that AD was not just the
disease to cause early onset dementia but was also largely responsible for dementia in those
patients 65 years and older. He felt that AD was underdiagnosed in older populations and
mentioned that trisomy 21 could be an important genetic component contributing to AD.
He importantly noted that senile plaques do not necessarily correlate well with AD after
the age of 65. Even, in 1966 it was noticed that senile plaques can appear even without a history
of or symptoms of dementia. He determined that the requirements of AD diagnosis were as
follows: granulovacuolar neuronal degeneration- the accumulation of cytoplasm in small
vacuoles in hippocampal pyramidal neurons measuring up to five microns in diameter that
contain a small granule, neurofibrillary degeneration- described by Alzheimer as thickening and
contortion of fibrils that look like tangles in the neuronal cytoplasm, neuronal degeneration, and
senile plaques, which, at this time, could only be described as particles with variable patterns of
organization. He argued that the best correlate to AD was the granulovacuolar degeneration of
the brain, specifically, in the hippocampus [37].
The two hallmarks of AD, neurofibrillary tangles and amyloid plaques, were described
from the very beginning and due to observations that the tangles correlated better with AD than

Copyright 2018 Evan Feeley
29

the plaques, research on neurofibrillary tangles was much more popular than amyloid plaques
throughout the 1970s. By 1976, using tilt stage electron microscopy and x-ray images, it was
determined that neurofibrillary tangles are made of paired helical filaments, an expansion on the
discovery first made by Kidd in 1963 that these filaments consisted of a bifilar helix [34]. By,
1978, neurofibrillary tangles could be immuno-stained and the proteins in the paired helical
filaments were observed and thought to be very similar to beta tubulin [17]. By 1979, PHF are
seen to uniquely form in the hippocampus and are thought not to disrupt any organelles in the
cell but merely displace them. At this time, it was not completely clear whether or not PHF
directly led to cell death. It was seen that if neurons were treated with chemicals that block the
formation of the fibers that make up the PHF that microtubules would break down.
Tau was discovered in 1975 by Weingarten as a factor essential for microtubule
assembly. It is observed that tau is able to be phosphorylated and it is hypothesized that proteins
that make up the microtubule also make up the paired helical filaments. PHFs could be induced
by exposing human cerebral cortical neurons to extracts from AD patient brains. This, in addition
to tangles being restricted to certain types of neurons, led to the thought that a virus could be
involved in the formation of PHF and NFT [35]. But, the AD extract experiment that caused
tangles in cultured neurons failed to be reproduced in other laboratories [32]. Aluminum
persisted as a theme in AD research throughout the 1970s and it was thought that elevated
amounts of aluminum were associated with neurofibrillary degeneration [17]. But, this was
effectively debunked by 1979 as being part of the AD pathogenesis [35].
Also, important to come out of the 1970s was the hypothesis that the cholinergic system
played an important role in AD pathology. It was noticed that the choline acetyltransferase in the
hippocampus of AD brains was depleted. It was thought that perhaps there is a problem with the

Copyright 2018 Evan Feeley
30

cholinergic receptor pathway as anticholinergic drugs could create memory problems [29]. The
reduction in activity of choline acetyltransferases was noticed to correlate with the amount of
plaque on the brain and the amount of cognitive impairment [32]. The cholinergic hypothesis is
the basis of many AD therapies and are the only drugs to have any safe effect on slowing the
progression of AD.
By the late 1970s, AD was recognized to be the most common cause of dementia.
Pedigree analysis and work in twins by Kallman and Sander back in 1949 led to an
understanding that there was a dominant genetic inheritance pattern in early onset dementia. By
the end of the 1970s, it was understood that there was a dominant mode of inheritance for
EOAD, but dementia was sporadic in older patients with AD. Also based on neurotubule
research in the 70s, it was noticed that there were short side arm proteins and it was thought they
were tau protein. Similarities between PHF and proteins associated with microtubules continued
to be a point of investigation as immunostaining of PHF became more sophisticated.
Interestingly, extracts of microtubules from humans were used to make antibodies in rabbits.
These antibodies were then used to stain normal and AD brain sections and what was observed
was that both microtubules and PHFs would be stained.
Up until 1980 still not much was known about amyloid plaques. It was thought that there
may be an astrocyte component to it based on immunohistochemistry. It was observed that
plaques had increased oxidative enzyme and acid hydrolase activity. Researchers found that the
core of the plaque is made of amyloid and is surrounded by large numbers of unmyelinated
neurites that also seem to contain large numbers of mitochondria. There was an occasional
presence of microglia in and around the plaques and plaques could be induced after causing brain
damage in animal models [32].

Copyright 2018 Evan Feeley
31

By, 1988 research on PHFs began to slow because it was difficult to further characterize
them because they are very insoluble [10], but it was known that whatever protein was making
up PHF was phosphorylated [4]. Not only that but in 1984 amyloid beta was isolated and
characterized by Glenner and Wong from, first, cerebral vessels, and then later from senile
plaques in both AD and Down’s syndrome patients. This paved the way for an explosion in
research on amyloid beta and amyloid beta metabolism. Once amyloid beta was characterized the
amino acid sequence was used to find the gene coding for it. Then, through hybridization
techniques using cDNA libraries the gene for amyloid beta was localized to chromosome 21 by
four different groups.
This localization pointed to the fact that amyloid beta is part of a much larger gene
coding for a 695 amino acid protein from the open reading frame. This research ruled out the
possibility that amyloid beta was an infectious plaque like those of prion diseases and
Creutzfeld-Jakob disease. For years it had been recognized that the plaques in AD were the same
as those in patients with Down’s syndrome, this research confirmed what was already previously
hypothesized based on the similarities between plaques in AD and Down’s syndrome and their
relationship with chromosome 21 [10]. This gene was cloned and dubbed amyloid precursor
protein or APP. It is known by 1989 that the C terminus of APP can cause neural degeneration
and intense investigation ensued into APP metabolism and amyloid beta. It was not yet known
how amyloid beta was formed, though, it was hypothesized that it occurred by proteolytic
cleavage. It was not known exactly how it was cleaved yet or by what, and there were no
effective animal models to test this on. Up until this point it was well recognized that amyloid
beta treatment would not be effective at treating AD [4].

Copyright 2018 Evan Feeley
32

Also, by 1989, amyloid beta was characterized as a transmembrane protein and that
cleavage of the C terminal fragment might be the first step to producing amyloid beta. It was
hypothesized, due to its orientation in the membrane, that APP cleavage must happen before
APP is inserted into the membrane. Because of the relationship with Down’s syndrome and
chromosome 21 it was thought that AD could be caused by increased doses of APP since
Down’s syndrome is caused by an extra chromosome 21. The APP gene was mapped and
sequenced in other mammals and led researchers to believe that it was highly conserved and
there must be a critical function of APP in the brain. It was also reported by this time that
different variants of APP correlated with certain parts of the brain. APP695 was increased two to
three-fold in certain parts of the brain and was especially abundant in fetal brains, suggesting
perhaps it played a crucial role in development.
Tau was found in 1986, only two years after the characterization of amyloid beta, to
make up PHF through immunochemical means as it had been shown that anti-neurofilament
antibodies that were able to label PHF also reacted specifically to phosphorylated regions of tau
[31]. Because of the genetic evidence however provided by the role of APP in patients with
familial AD, the idea persisted that, since NFT were involved in the pathology of other diseases,
they were not a direct or primary cause of AD, and thus the amyloid cascade hypothesis
continued to be the primary focus of AD research for the next couple decades.
The 1990s was dominated by research that focused on the amyloid cascade hypothesis
and elucidating the mechanism of APP metabolism. The amyloid cascade hypothesis was
proposed in 1991 and is the most researched explanation of AD [27]. Further genetic
discoveries related to familial AD and the production of amyloid beta supported the amyloid
hypothesis. Missense mutations in PS1 and PS2, which were found to localize in the smooth and

Copyright 2018 Evan Feeley
33

rough endoplasmic reticulum, and the Golgi, in families with EOAD resulted in an increase in
the more amyloidogenic 42 amino acid amyloid beta. The missense mutations were thought to
result in a gain of function because even mutated PS1 could rescue the defects in PS1 knockout
mice. Delta and beta catenin were found to co-immunoprecipitate with PS1. It was known that
alpha, beta, and gamma secretases were responsible for the proteolytic cleavage of APP, but it
was not known that presenilins were a part of the gamma secretase complex. The genetic
discoveries of the missense mutations in the presenilins added to the amyloid hypothesis by
observing that not only do these mutations cause AD but that they also cause an increase in
amyloid beta 42. It was hypothesized that presenilins must cause a change in gamma secretase
activity since amyloid beta 42 is produced under normal conditions but was increased when
presenilins were mutated. However, evidence from mouse models expressing the mutated human
PS1 gene was conflicting. This mutation in mice did not cause any AD symptoms or
abnormalities other than an increase in amyloid beta 42 [12].
By the end of the 1990s the composition of the two hallmarks of AD, amyloid plaques
and NFT, and four AD related genes: APP, PS1, PS2, and ApoE4 were identified. The ApoE4
allele was observed as the first major genetic risk factor for sporadic AD, while ApoE2 might be
protective. Mutant human APP expressed in mice was the first working animal model of AD. All
subsequent AD transgenic mouse models created from mutations in APP, PS1, and PS2, resulted
in an increase in amyloid beta. The most aggressive AD phenotypes were observed when mice
had both the presenilin mutations and mutations in APP. The advancements in the understanding
of AD, however, did not seem to gain more clarity. Instead, AD was beginning to be recognized
as a multifactorial syndrome rather than a disease due to a single cause because of the many risk
factors associated with it. It was found that other post translational modifications of tau other

Copyright 2018 Evan Feeley
34

than phosphorylation could also lead to the formation of paired helical filaments such as
ubiquitination.
Amyloid beta, as it was understood, was not associated with dystrophic neurons or
activating microglia as long as it was diffuse and in 1992 it was discovered that amyloid beta is
constitutively secreted under normal metabolic conditions. But, alas, the evidence garnered at
this time was used to support the idea that the formation of NFT is a response to the
accumulation of amyloid beta and other APP metabolites. By the end of 1990s arguments for tau
pathology being the primary mode of the disease remained uncommon as a relationship between
mutations involving the tau gene and AD remained unidentified.
Mutations in tau had been implicated in other diseases where tangles arise without the
formation of amyloid beta, such as in frontotemporal dementia, leading to the argument that
tangles definitely do not cause plaques, but it can be also said that tau pathology on its own can
cause neural degeneration and cell death. In Down’s syndrome, however, it was argued that NFT
do not appear until much later than amyloid plaques.
The deletion of APP in mice did not cause problems with viability. Amyloid beta can be
generated from endocytosed APP. The mutations in APP that affect AD occur specifically in
regions where secretases cleave the protein and do not occur in other sites. It was also paramount
to uncover the relationship between ApoE4 and APP metabolism. ApoE4, when co-expressed
with mutant APP in mice resulted in either the increased aggregation or decreased clearance of
amyloid beta but it did not affect APP processing [28].
By 2005 it is still not known what the endogenous function of APP is because mice that
lack APP have only small observable brain defects. Amyloid beta has been the focus of research
for the last decade. Finally, it was found that Tau is encoded by MAPT gene on chromosome 17

Copyright 2018 Evan Feeley
35

but this gene encoding tau still had not been linked to AD, even though mutations in MAPT and
tau has been linked to other neurodegenerative diseases. The 2000s also saw the invention of the
triple transgenic mouse model of AD. It was created by microinjection of mutated APP and tau
genes into the embryo of a PS1 mutated mouse. This mouse exhibits both amyloid beta and tau
pathology and in this model amyloid beta deposition precedes tau pathology.
Using this mouse model, Oddo was able to garner substantial evidence for the amyloid
hypothesis. By removing amyloid beta using immunotherapy, Oddo observed that there was a
stop in early tau pathology and when the antibody was removed amyloid beta plaques would
come back followed by the formation of neurofibrillary tangles. These triple transgenic mice
have also shown that cognitive impairments begin to occur before amyloid plaques or NFT take
hold. From Oddo’s studies in transgenic mice the evidence was compelling that amyloid beta and
tau somehow interact and that amyloid beta makes tau pathology worse. Mutant PS1 and tau
transgenic mice without mutated APP did not develop amyloid beta pathology. They did develop
tau pathology, but one that was less severe than the triple transgenic mouse [19].
Also, in the 2000s was more sophisticated research on the ApoE gene and its variants,
with the understanding that ApoE2 had protective effects, ApoE3 was normal, and ApoE4 was a
risk factor. Presenilin polymorphisms were also seen to be intimately linked to AD as a specific
polymorphism was found to be associated with a decrease risk of AD in people with the ApoE4
allele. By the late 2000s the genetic impacts of tau mutations were linked to AD and it was seen
that problems with the gene encoding tau could cause pathological phenotypes in the brains of
people with frontotemporal Parkinson's, sporadic Pick’s and AD. Many of these such mutation
lead to an imbalance of 3R to 4R tau ratios with 4R being significantly higher than 3R, but still
the majority of genetic abnormalities related to AD were still directly related to APP metabolism.

Copyright 2018 Evan Feeley
36

As such in depth research for the support of the amyloid hypothesis continued. It was
found that in cultured cortical neurons amyloid beta can induce apoptosis through the JNK-c-
Jun-FasL-caspase-dependent extrinsic pathway and that amyloid beta causes oxidative stress
through the mitochondria. Amyloid beta itself was seen to be able to cause issues with RACK1
distribution, impairing muscarinic, acetylcholine, PKC, and GABAergic transmission and
regulation.
Tau phosphorylation gained support as being part of tau pathology. Tau was recognized
to be phosphorylated at serine residues 396 and 404 and that phosphorylation at these sights
reduced its ability to stabilize microtubules. Tau phosphorylation was deemed to be the first step
in tau aggregation but the reason of why tau is being phosphorylated has not been elucidated.
The hypothesis with the most attention at the time was that there is a deregulation of kinases due
to oxidative stress and abnormalities in metabolic homeostasis [22].
Clinical Trials
Failure rates of drug trials for AD are high with reviewers saying that anywhere from
100-172 development failures occurred up until 2008 and 101 failures from 1998 to 2013. The
first drugs approved to treat AD were drugs that were based off the cholinergic hypothesis of AD
like cholinesterase inhibitors and memantine. Drug development has been most influenced by the
cholinergic hypothesis. Cholinesterase inhibitors have been around since the late 70s and 80s
ever since Drachman and Leavitt came up with the idea that memory was dependent on the
cholinergic system and age in 1974, which findings were later supported by two independent
British groups. They found that the severity of dementia is correlated with the loss of cholinergic
neurons specifically in the nucleus basalis of Meynert, which has wide projections to the
neocortex in the forebrain.

Copyright 2018 Evan Feeley
37

The cholinergic hypothesis was a focus of drug development throughout the 1980s and
1990s. From 1986 to 1996 trials based on the cholinergic system were mainly cholinesterase
inhibitors. One of the main challenges with these drugs was figuring out the proper dosages.
During this time M1 muscarinic receptor agonists began to gain support. They were effective at
improving cognition but had side effects that made the drug inefficacious, such as
gastrointestinal issues, salivation, sweating, and frequent urination. From 1986 to 1996 the most
important drugs tested were cholinesterase inhibitors: tacrine, valnacrine, sustained release
physostigmine, eptastimine, metrifonate, donepezil, rivastigmine, and galantamine. Tacrine,
donepezil, rivastigmine, and galantamine were all marketed in the United States [27]. Donepezil,
galantamine, rivastigmine, and tacrine are all cholinesterase inhibitors that just slightly slow the
progression of AD. 5-HT6 receptor agonists and idalopirdine are thought to also increase
cholinergic activity [16]. Focus on the cholinergic hypothesis for the treatment of AD, to this day
still has not been abandoned because even though it does not target AD at the source it can
mildly improve memory in patients. The cholinergic hypothesis continues to be considered for
small molecule therapies such as nicotinic receptor modulators. The last drug approved for
market in the United states was an N-methyl-D-aspartate (NMDA) receptor agonist that was
shown to be useful in patients with moderate to severe AD in 2003. Once again, these drugs have
only shown modest and short-term effects on patients.
After the amyloid cascade hypothesis was proposed in 1991, drugs have been designed to
target the processing of APP, including beta and gamma secretases. Drug development based on
the amyloid hypothesis has been complicated due to different parts of APP metabolism having
different toxicities and effects on inflammation. 2001 was the first time that a drug based on the

Copyright 2018 Evan Feeley
38

amyloid hypothesis had made it to the late stages, but, as of 2014, not a single drug target has
been successfully developed based on the amyloid hypothesis.
Because drug development had been so difficult researchers began to lower their
expectations. There was an increased effort to market drugs that were supposed to slow the rate
of cognitive decline, instead of attaining an improvement in cognition. These drugs include anti-
inflammatory agents, neuroprotective agents, and metabolic enhancers. In 12-month trials from
1994 to 2010, conjugated estrogens, an amyloid beta vaccine, gamma secretase modulators, and
agents that stimulate growth hormone secretion were all tried.
Once a risk state for AD could be identified in the 1980s there was motivation for
preventative trials. Vitamin E and vitamin B complexes were tried to assess possible AD
prevention, but these trials also came up negative. Anti-hypertension medications, conjugated
estrogens, HMG-CoA reductase inhibitor, selenium, gingko biloba, and NSAIDs, non-steroidal
anti-inflammatory drugs, have all been used in prevention trials but none have worked.
In 18-month trials from 2001-2013 there was a focus on disease modification.
Therapeutics in clinical trial to attenuate cognitive decline were anti-inflammatory
hydrochloroquine, HMG-CoA reductase inhibitor to reduce cholesterol, combinations of vitamin
B and docosahexaenoic acid to lower homocysteine, and a 5-HT1a agonist. Drugs targeting the
amyloid hypothesis were multiple AB aggregation inhibitors, gamma secretase modulators,
gamma secretase inhibitors, a RAGE inhibitor, and antibodies targeting amyloid beta, but all
ended in failure.
The amyloid hypothesis has been the most prominent focus in drug development since
1993. In an effort to manipulate this pathway several secretase inhibitors and modulators, and
active and passive immunization techniques with antibodies targeting different epitopes of

Copyright 2018 Evan Feeley
39

amyloid beta monomers, oligomers and fibrils have all be developed, along with fibrillization
inhibitors, and antiaggregants. Vaccines for amyloid beta such as, AN1792, tramiprosate, and
bapineuzumab were able to decrease amyloid beta and tau CSF biomarkers and increase
hippocampal volume but did not have any significant effects on actual clinical symptoms.
There has been a growing interest recently in tau-based approaches to tau-based therapy,
including inhibitors of GSK-3B, which has been observed to phosphorylate tau, tau aggregation
inhibitors, such as methylene blue, microtubule stabilizers, and inhibitors of tau N-
glcNacylation, which include small molecules antibodies, and vaccines [27]. Microtubule
stabilizers have shown that they could be helpful in the treatment of AD, but so far, the
stabilizers have proven to be too toxic. Tau aggregation inhibitors could also be effective but the
first-generation tau inhibitors have always failed because of toxic side effects. [16].
Reducing Tau has had several effects on synaptic function. Tau-tau aggregation can be
blocked and the stability of PHF can be reduced. Stabilizing microtubules could be a good
therapeutic for AD but drugs like paclitaxel and other taxanes, which stabilize microtubules, do
not readily pass through the blood brain barrier. Currently, a drug, davunetide, promotes
microtubule stability and reduces tau hyperphosphorylation in pre-clinical studies. In a phase II
clinical trial, intranasal administration of a drug, NAP, did not produce significant side effects
and improved cognitive function in people with MCI. Since, elevated Fyn kinase has been shown
to be linked to tau phosphorylation the drug, Saracatinib, a fyn kinase inhibitor, is now in phase
II of clinical trials.
The enzyme, B-N-acetylglucosaminidase (OGA), is responsible for removing O-
GlcNAcylation and inhibiting this enzyme resulted in reduced NFT and neuronal loss in mice.
OGA inhibitors may be a promising candidate for further clinical development. Immunotherapy

Copyright 2018 Evan Feeley
40

could also be a very promising avenue. Antibodies, DC8E8 or AADvac1, target diseased tau and
inhibit tau-tau interactions. Using this immunotherapy, hyperphosphorylated tau was reduced by
95% in rats and this therapy is currently in phase II clinical trial [6].
Since, AD drug trials have been so tricky and the drug development failure rate so high,
99.6% from 2002 to 2012, there is a lot we can learn from these failed trials and many questions
we must ask when developing future therapies. Are the animal models that we are using for
testing appropriate? Does the drug or therapeutic cross the blood brain barrier? It is important to
know that the compound excluding mechanism of the central nervous system is different in mice
and humans. For example, the drug, Tarenflurbil, had good outcomes in AD mouse models but
likely failed in clinical trials because sufficient amounts were not able to cross the blood brain
barrier. Monoclonal antibodies are often largely excluded from the central nervous system with
maybe 1 out of every 1000 crossing the blood brain barrier.
With the difficulties of crossing the blood brain barrier in mind, drugs like Solanezumab
tested the peripheral sink hypothesis. Since antibodies don’t pass through the blood brain barrier
well, Solanezumab bound peripheral amyloid beta outside the brain. The thought was that if
amyloid beta is depleted outside of the brain it would passively be drawn out into the periphery
through the blood brain barrier, but this drug did not meet the necessary clinical outcomes.
Several immunotherapies: AN1792, bapineuzumab, and gantenerumab, show target engagement
with amyloid beta but did not significantly reduce cognitive deficits or prove a significant
clinical benefit. [7] CAD106 is another immunotherapy that tries to target specific epitopes of
amyloid beta. It causes a strong immune response without causing an inflammatory t cell
response and is currently in phase 3 of clinical trials [16].

Copyright 2018 Evan Feeley
41

First generation beta secretase inhibitors have failed in clinical trial because of low oral
bioavailability and an inability to effectively penetrate the blood brain barrier. Second generation
beta secretase inhibitors failed because of toxicity in the liver. Third generation beta secretase
inhibitors are currently being tested in clinical trial. Gamma secretase inhibitors have previously
failed because of worsening of symptoms in some patients and have also shown not to be good
drug targets because gamma secretase is also critical for the notch signaling pathway.
Avagacestat, a gamma secretase inhibitor, failed in phase two because of side effects like brain
bleeding and nonmelanoma skin cancer. EVP-0962, a small molecule that pushes production
towards shorter forms of amyloid beta without affecting notch pathway signaling, was also
developed but still failed in phase 2 [16].
Discussion
Because AD is an age-related disease as our population ages AD will continuously
become a greater burden on society and these numbers will also continue to climb as more
sophisticated diagnostic technology arises. Women are diagnosed with AD more often than men
[3], but more women also survive past the age of 65. This creates a bias in diagnosis and I do not
think that gender is a risk factor for AD.
Inhibiting and knocking out beta secretase is not the answer to AD therapy as it is most
likely playing other critical roles in the brain. Increased amounts of APP can also lead to AD in
patients with Down’s syndrome. If there is an increased amount of APP, then APP would act as a
competitive inhibitor binding beta secretase. GWAS studies suggest a link between AD, APP
metabolism, and lipid metabolism. It is reasonable to suspect that beta secretase plays a role in
cholesterol and lipid metabolism. If beta secretase is being predominantly used up by the cell in
APP metabolism then perhaps cholesterol and lipid metabolism is not being carried out properly.

Copyright 2018 Evan Feeley
42

It has been noticed that amyloidogenic processing of APP occurs in cholesterol rich
regions and drugs targeting cholesterol promote alpha secretase cleavage [1], perhaps, suggesting
a different intimate link between beta secretase and lipid metabolism where cholesterol increases
beta secretase’s
affinity for APP
and once again
keeps beta secretase
from performing
functions elsewhere
in the cell.
Variants of
AP2 have also
shown to be
possible risk
factors for AD and
are important in the trafficking of secretases. Since, APP and beta secretase follow similar
trafficking routes, if beta secretase is being inappropriately localized with APP, then beta
secretase is unlikely to be able to perform other functions.
Research has shown that the C terminal fragment of APP processing is the most toxic
metabolite [1][31]. The C terminal fragment is a direct result of beta secretase-mediated
cleavage, supporting the idea that beta secretase could be the most crucial part of AD pathology.
These results could suggest that the C terminal fragment itself is toxic or it could suggest the
cell’s response to unavailable beta secretase.
Figure [6]: Diagram showing the possible connection between tauopathy and
amyloidosis. The possible bridge between the two pathways must occur upstream in
APP metabolic dysfunction. Issues with trafficking and C-terminal fragment
accumulation is a prominent suspect. [18]

Copyright 2018 Evan Feeley
43

However, since overexpression of the C terminal fragments themselves cause AD like
symptoms in mice including the phosphorylation of tau then perhaps the increased affinity of
beta secretase for APP would play a dual role. On one hand, the increased affinity for APP
causes beta secretase to be unavailable for other functions, causing certain symptoms of AD, and
then, its direct product, the C terminal fragment leads to other aspects of AD pathology, through
avenues such as attenuation of the cAMP/PKA/CREB pathway by causing the breakdown of
cAMP. The importance of this pathway in AD is supported by observations that sirt1 can
indirectly repress elements that downregulate CREB and protect against AD symptoms in mouse
models.
Amyloid plaques may occur as a normal process of aging, not because of a problem with
APP metabolism, but an inability or inefficiency of amyloid plaque clearance. In rare patients
with AD that do not experience plaques, there may be issues with beta secretase and APP
metabolism, but these individuals could have robust systems of amyloid beta clearance.
There is an undiscovered relationship between APP and beta secretase trafficking,
localization, and activity, APP metabolism, lipid metabolism, and tau pathology. Soluble tau has
been observed to regulate APP trafficking. So, it is possible that the phosphorylation or other
post translational modifications of tau might affect its ability to perform this function correctly
and maybe increase its likelihood to colocalize with beta secretase.
Work in drosophila and mice gives conflicting evidence about which pathology, amyloid
beta or tau pathology, happens first. It is useless to continue to predict biochemical pathway
pathology based on the amyloid plaque and NFT end points, since it is clear that the mechanism
is upstream of the physical lesions in causing cell death. The biochemical mechanism upstream
of the two lesions may be more closely interconnected than previously realized. Overexpression

Copyright 2018 Evan Feeley
44

of tau in drosophila results in neuronal loss before NFT form [18]. This supports the idea that
abnormalities in tau might cause an increased localization of APP with beta secretase, leading to
the formation of the C terminal fragment that then, might in turn be associated with the
phosphorylation of tau.
ApoE variants having an intimate relationship with AD supports the idea of aberrant lipid
metabolism playing a key role in AD pathology. ApoE has been observed to move cholesterol to
LDL receptors to facilitate their degradation. If ApoE4 is not able to facilitate cholesterol,
lipoprotein, and phospholipid movement as efficiently as normal conditions, this may result in an
increased interaction between cholesterol and junctions in APP metabolism, namely, beta
secretase cleavage of APP. Research on ApoE might also suggests a link between lipid
Figure [7]: Alteration of beta secretase activity on APP is a key aspect of Alzheimer’s disease pathology because
of it’s potential use elsewhere in the cell and its ability to generate C terminal fragments. My hypothesis is that
there are many ways in which the activity of beta secretase can be altered, but the more APP cleavage occurs
via beta secretase the worse disease pathology would become.

Copyright 2018 Evan Feeley
45

metabolism and responses to damage and inflammation, since it is taken up during development
and repair periods. It is not known whether inflammation is a cause or a response to AD
pathology, but it is intercorrelated with the disease.
ABCA7 being a strongly correlated genetic risk variant also lends support to the
importance of beta secretase cleavage, the trafficking of lipids, and lipid metabolism in
pathology, because ABCA7 is responsible for the movement of lipids across cell membranes and
loss of function of ABCA7 is associated with increased beta secretase cleavage of APP [2].
Risk factors like SORL1 suggest that issues with APP trafficking may lead to disease in a
different way. If AD is one of APP build up and a decrease in trafficking and flow throughout
the cell, SORL1’s ability to bind ApoE and facilitate APP movement and recycling is significant.
The levels of SORL1 have been observed to correlate with the C terminal fragments produced
from beta secretase cleavage [2], suggesting that the continual endocytosis of APP from the cell
membrane is essential in maintaining healthy metabolism. Otherwise beta secretase could have
an increased opportunity to cleave APP and produce C terminal fragments in the cell membrane.
Genetic evidence provided by mutations in PS1 and PS2 [12] could weaken my
hypothesis that beta secretase and C terminal fragments are the pathology causing part of APP
metabolism, unless this idea about secretases and competition persists. Missense mutations in
PS1 and PS2 are thought to result in a gain of function of gamma secretase. It is possible that
these missense mutations result in an increased affinity for the notch intracellular domain and
other targets of gamma secretase. This would result in a buildup of C terminal fragments. It is
still possible, however, that these mutations lead to AD merely by increasing the amounts of
amyloid beta, stimulating inflammation pathways, and weakening of the blood brain barrier,

Copyright 2018 Evan Feeley
46

since experiments in mice showed that the PS1 mutation increased amyloid beta 42 but did not
develop other AD symptoms or abnormalities.
Genetic evidence of mutations in APP could be more helpful. The mutations in APP that
occur in AD are only found in regions important for APP cleavage by secretases [28]. These
mutations could increase the affinity of beta secretase for APP. It is possible that AD may be a
result of an imbalance of alpha and beta secretase cleavage [1]. If mutations in APP are causing
an increase in the affinity of beta secretase to cleave instead of alpha secretase, then
competitively APP will be binding with beta secretase.
A reassessment of appropriate biomarkers needs to happen in order to properly monitor
AD. The most common markers are AB42 and tau in the CSF [8], but because cognitive
impairment can occur before either of the two lesions take hold, evaluating levels of C-terminal
fragments might be a better solution.
Drugs like cholinesterase inhibitors and other agonists that affect the cholinergic system
are not worthwhile therapies because improvement in symptoms are too mild. Drug targeting
should focus on hypotheses that will alter AD at its source, which, again, I propose to be APP
and beta secretase interaction.
Gamma secretase inhibitors are not an effective therapeutic because of its important role
in notch signaling in the brain and elsewhere in the body. Beta secretase inhibitors have also
shown to be toxic [16]. I predict that beta secretase has multiple other functions in the cell and
elsewhere in the body. As such, inhibiting secretases is not the answer. Instead, therapies need to
be explored where the cleavage sites on APP are blocked from interacting with beta secretase.
Then beta secretase would be free to perform other functions. This approach, from my
knowledge has not yet been tried. Designing drugs to cross the blood brain barrier is difficult. An

Copyright 2018 Evan Feeley
47

antibody that will bind the specific epitope of APP at the beta secretase cleavage site could be
useful, but antibodies don’t pass through the blood brain barrier well [7]. A small molecule that
could bind APP would be ideal. Work in triple transgenic mice has shown that cognitive
impairments occur before plaques or tangles take hold [19]. This is important because it supports
the idea that amyloid beta plaques and NFT are not good drug targets for AD pathology as they
are not the source of what causes dementia and brain atrophy.
Mouse models have been able to recapitulate certain parts of AD pathology such as
tangles and amyloid plaques using genes related to familial AD like APP, presenilins, and Mapt
but they have not been successful in translation to the clinic. Current models struggle to
demonstrate the significant cell loss seen in AD patients. Mouse models of early onset AD suffer
Figure [8]: Schematic diagram of my hypothesis as presented in this discussion on the interaction between
cholesterol metabolism, APP metabolism, formation of NFT, and endocytosis. I believe that if the interaction
between APP and beta secretase can be blocked without inhibiting beta secretase from performing its other
functions, then AD could perhaps be reversed as C- terminal fragments would no longer accumulate.

Copyright 2018 Evan Feeley
48

from a mis or overexpression of the transgenic protein compared to the endogenous protein,
developmental effects from knocked out or overexpressed genes, unknown disruption of
endogenous genes, and the timing of expression of transgenic genes that mimics actual
pathology. Knock in models using genes associated with familial AD have been used, but only
end up showing relatively mild, late onset phenotypes.
Inducible models using these genes have also shown interesting results. They have the
ability to show what happens after a mutant familial AD allele has been turned off. These studies
have suggested that AD pathology is reversible. Models that utilize one mutant gene have been
good for assessing its role and effect on pathways and AD pathology, but the triple transgenic is
currently the best, of many insufficient options, for testing drugs. Most of the mouse models
currently develop amyloidosis and tauopathy together and separately, and therefore much of the
related phenotypes, but there are currently no animal models that demonstrate significant
neuronal loss. All mouse models to date have not been shown to have high predictive value [24].
Conclusion
AD research and therapeutic development might still have a long way to go, but I believe
we are on the brink of efficacious drugs. The two leading hypotheses, amyloid cascade and tau,
need to be reassessed. Instead of one lesion causing the other, focus needs to be placed on the
intimate link between them but, both hypotheses do not need to be abandoned all together.
Mouse models will continue to be a struggle due to AD’s multifactorial nature, but I believe
blocking the interaction of APP with beta secretase without inhibiting beta secretase could
produce disease reversing effects.

Copyright 2018 Evan Feeley
49

References
1. Agostinho, P., et al. "Localization and Trafficking of Amyloid-Β Protein Precursor and
Secretases: Impact on Alzheimer's Disease." J Alzheimers Dis 45.2 (2015): 329-47.
2. Apostolova, L. G., et al. "Associations of the Top 20 Alzheimer Disease Risk Variants
with Brain Amyloidosis." JAMA Neurol (2018).
3. Alzheimer's Association. "2017 Alzheimer’s Disease Facts and Figures." Alzheimers
Dement.13 (2017): 325-73.
4. Caputo, C. B., and A. I. Salama. "The Amyloid Proteins of Alzheimer's Disease as
Potential Targets for Drug Therapy." Neurobiol Aging 10.5 (1989): 451-61.
5. Chaudhury, S., et al. "Polygenic Risk Score in Postmortem Diagnosed Sporadic Early-
Onset Alzheimer's Disease." Neurobiol Aging 62 (2018): 244.e1-44.e8.
6. Chong, F. P., et al. "Tau Proteins and Tauopathies in Alzheimer's Disease." Cell Mol
Neurobiol (2018).
7. Cummings, J. "Lessons Learned from Alzheimer Disease: Clinical Trials with Negative
Outcomes." Clin Transl Sci 11.2 (2018): 147-52.
8. El Kadmiri, N., et al. "Biomarkers for Alzheimer Disease: Classical and Novel
Candidates' Review." Neuroscience 370 (2018): 181-90.
9. Epis, R., et al. "Alpha, Beta-and Gamma-Secretases in Alzheimer's Disease." Front
Biosci (Schol Ed) 4 (2012): 1126-50.
10. Glenner, G. G. "Alzheimer's Disease: Its Proteins and Genes." Cell 52.3 (1988): 307-8.
11. Goedert, M., and M. G. Spillantini. "A Century of Alzheimer's Disease." Science
314.5800 (2006): 777-81.

Copyright 2018 Evan Feeley
50

12. Haass, C., and B. De Strooper. "The Presenilins in Alzheimer's Disease--Proteolysis
Holds the Key." Science 286.5441 (1999): 916-9.
13. Hannah, J. A. "A Case of Alzheimer's Disease with Neuropathological Findings." Can
Med Assoc J 35.4 (1936): 361-6.
14. Hata, S. "[Molecular Pathogenesis of Sporadic Alzheimer's Disease (Ad) and
Pharmaceutical Research to Develop a Biomarker for Ad Diagnosis]." Yakugaku Zasshi
135.9 (2015): 1023-7.
15. Hemphill, R. E., and E. Stengel. "Alzheimer's Disease with Predominating Crossed
Cerebro-Cerebellar Hemiatrophy." J Neurol Psychiatry 4.2 (1941): 97-106.
16. Hung, S. Y., and W. M. Fu. "Drug Candidates in Clinical Trials for Alzheimer's Disease."
J Biomed Sci 24.1 (2017): 47.
17. Iqbal, K., et al. "Chemical Relationship of the Paired Helical Filaments of Alzheimer's
Dementia to Normal Human Neurofilaments and Neurotubules." Brain Res 142.2 (1978):
321-32.
18. Kametani, F., and M. Hasegawa. "Reconsideration of Amyloid Hypothesis and Tau
Hypothesis in Alzheimer's Disease." Front Neurosci 12 (2018): 25.
19. LaFerla, F. M., and S. Oddo. "Alzheimer's Disease: Abeta, Tau and Synaptic
Dysfunction." Trends Mol Med 11.4 (2005): 170-6.
20. Lane, C. A., J. Hardy, and J. M. Schott. "Alzheimer's Disease." Eur J Neurol 25.1 (2018):
59-70.
21. Longridge, R. G. "Alzheimer's Disease." Proc R Soc Med 32.3 (1939): 222-4.
22. Mondragón-Rodríguez, S., et al. "Causes Versus Effects: The Increasing Complexities of
Alzheimer's Disease Pathogenesis." Expert Rev Neurother 10.5 (2010): 683-91.

Copyright 2018 Evan Feeley
51

23. Nativio, R., et al. "Dysregulation of the Epigenetic Landscape of Normal Aging in
Alzheimer's Disease." Nat Neurosci (2018).
24. Onos, K. D., et al. "Toward More Predictive Genetic Mouse Models of Alzheimer's
Disease." Brain Res Bull 122 (2016): 1-11.
25. Sanabria-Castro, A., I. Alvarado-Echeverría, and C. Monge-Bonilla. "Molecular
Pathogenesis of Alzheimer's Disease: An Update." Ann Neurosci 24.1 (2017): 46-54.
26. Satoh, A., S. I. Imai, and L. Guarente. "The Brain, Sirtuins, and Ageing." Nat Rev
Neurosci 18.6 (2017): 362-74.
27. Schneider, L. S., et al. "Clinical Trials and Late-Stage Drug Development for Alzheimer's
Disease: An Appraisal from 1984 to 2014." J Intern Med 275.3 (2014): 251-83.
28. Selkoe, D. J. "Translating Cell Biology into Therapeutic Advances in Alzheimer's
Disease." Nature 399.6738 Suppl (1999): A23-31.
29. Smith, C. M., and M. Swash. "Possible Biochemical Basis of Memory Disorder in
Alzheimer Disease." Ann Neurol 3.6 (1978): 471-3.
30. Stoddart, W. H. "Presbyophrenia (Alzheimer's Disease)." Proc R Soc Med 6. Sect Psych
(1913): 13-4.
31. Tanzi, R. E., P. H. St George-Hyslop, and J. F. Gusella. "Molecular Genetic Approaches
to Alzheimer's Disease." Trends Neurosci 12.4 (1989): 152-8.
32. Terry, R. D., and P. Davies. "Dementia of the Alzheimer Type." Annu Rev Neurosci 3
(1980): 77-95.
33. Vishal, S., A. Sourabh, and S. Harkirat. "Alois Alzheimer (1864-1915) and the Alzheimer
Syndrome." J Med Biogr 19.1 (2011): 32-3.

Copyright 2018 Evan Feeley
52

34. Wisniewski, H. M., and D. Soifer. "Neurofibrillary Pathology: Current Status and
Research Perspectives." Mech Ageing Dev 9.1-2 (1979): 119-42.
35. Wiśniewski, H. M., H. K. Narang, and R. D. Terry. "Neurofibrillary Tangles of Paired
Helical Filaments." J Neurol Sci 27.2 (1976): 173-81.
36. Wong, S. Y., and B. L. Tang. "Sirt1 as a Therapeutic Target for Alzheimer's Disease."
Rev Neurosci 27.8 (2016): 813-25.
37. Woodard, J. S. "Alzheimer's Disease in Late Adult Life." Am J Pathol 49.6 (1966): 1157-
69.
38. Grimm, M. O., et al. "Neprilysin and Aβ Clearance: Impact of the App Intracellular
Domain in Nep Regulation and Implications in Alzheimer's Disease." Front Aging
Neurosci 5 (2013): 98.
39. Guzman-Martinez, Leonardo & Farías, Gonzalo & Maccioni, Ricardo. (2013). Tau
Oligomers as Potential Targets for Alzheimer’s Diagnosis and Novel Drugs. Frontiers in
neurology. 4. 167. 10.3389/fneur.2013.00167.

Abstract (if available)

Abstract Since Alzheimer’s disease was first described by Alois Alzheimer in 1906, little progress has been made in the development of therapeutics. Early onset Alzheimer’s disease has clear heritable dominant mutations that lead to pathology, but sporadic or late onset Alzheimer’s disease, like obesity or heart disease, is not as simple. Late onset Alzheimer’s is highly multifactorial, influenced by a combination of many genetic and environmental factors as a person ages. Here, I attempt to give an overview of the disease genetics, biomarkers, and the two hallmarks of Alzheimer’s, amyloid plaques and neurofibrillary tangles. The two leading hypotheses, the amyloid cascade hypothesis and the tau hypothesis must be intimately interlinked in order to cause pathology. Mouse models do not effectively represent Alzheimer’s disease in humans and new methods to target amyloid precursor protein metabolism need to be developed. From my broad investigation into Alzheimer’s disease, I suggest that the interaction between the amyloid precursor protein and beta secretase be a key focus for the development of future therapeutics, but beta secretase must not be inhibited.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Understanding the role of APP and DYRK1A in human brain pericytes

PDF

From risk mitochondrial and metabolic phenotype towards a precision medicine approach for Alzheimer's disease

PDF

Role of CD33 structure and function in Alzheimer’s disease

PDF

Long-term blood pressure variability across the clinical and biomarker spectrum of Alzheimer’s disease

PDF

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

PDF

The contribution of arachidonate 5-lipoxygenase to inflammatory response in coronary artery disease

PDF

TREM2 and C1q signaling regulates immunoproteostasis in Alzheimer's disease

PDF

Using neuroinformatics to identify genomic and proteomic markers of suboptimal aging and Alzheimer's disease

PDF

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

PDF

Neural crest-derived cranial pericytes and their dysfunction in disease

PDF

Interaction of monoclonal antibodies MW1 and PHP1 with huntingtin exon1 protein

PDF

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

PDF

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

PDF

The effects of a fasting mimicking diet (FMD) on mouse models of Alzheimer's and Parkinson's disease

PDF

Antibody discovery and structural studies of the viral oncogene ORF74

PDF

Charcot-Marie-Tooth disease type 1E: follow-up studies on a screen for therapeutic small molecules and PMP22-interacting proteins

PDF

Derivation, expansion and characterization of human hippocampal primordial cells from normal and diseased iPSCs

PDF

Beclin 2-mediated autophagic degradation of the pathogenic proteins in neurodegenerative diseases

PDF

The role of endoplasmic reticulum chaperone glucose-regulated 78-kilodalton (GRP78) in lung cancer

PDF

Characterization of metabolically distinct muscle resident fibro-adipogenic subpopulations reveals a potentially exploitable mechanism of skeletal muscle aging and disease

Asset Metadata

Creator Feeley, Evan (author)

Core Title An overview of Alzheimer’s disease and the potential for APP-beta secretase-interaction as a therapeutic target

School Keck School of Medicine

Degree Master of Science

Degree Program Biochemistry and Molecular Biology

Publication Date 08/06/2018

Defense Date 08/01/2018

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

Tag Alzheimer’s disease,APP metabolism,beta secretase,genetics,OAI-PMH Harvest,tau pathology

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Tokes, Zoltan (committee chair), Frenkel, Baruch (committee member), Hacia, Joseph (committee member)

Creator Email efeeley@usc.edu,evanfeeley@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-64800

Unique identifier UC11671940

Identifier etd-FeeleyEvan-6590.pdf (filename),usctheses-c89-64800 (legacy record id)

Legacy Identifier etd-FeeleyEvan-6590.pdf

Dmrecord 64800

Document Type Thesis

Format application/pdf (imt)

Rights Feeley, Evan

Type texts

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

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA